[2021-04-14 17:01:04 swin_tiny_patch4_window7_224] (main.py 96): INFO number of params: 28288354 [2021-04-14 17:01:04 swin_tiny_patch4_window7_224] (main.py 99): INFO number of GFLOPs: 4.49440512 [2021-04-14 17:01:05 swin_tiny_patch4_window7_224] (main.py 123): INFO no checkpoint found in , ignoring auto resume [2021-04-14 17:01:05 swin_tiny_patch4_window7_224] (main.py 141): INFO Start training [2021-04-14 17:01:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][0/1251] eta 3:07:40 lr 0.000001 time 9.0014 (9.0014) loss 6.9944 (6.9944) grad_norm 1.3503 (1.3503) [2021-04-14 17:01:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][10/1251] eta 0:23:52 lr 0.000001 time 0.2715 (1.1542) loss 7.0452 (6.9999) grad_norm 1.3224 (1.3454) [2021-04-14 17:01:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][20/1251] eta 0:15:05 lr 0.000002 time 0.2936 (0.7356) loss 6.9762 (6.9864) grad_norm 1.3055 (1.3188) [2021-04-14 17:01:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][30/1251] eta 0:11:55 lr 0.000002 time 0.2739 (0.5860) loss 6.9401 (6.9785) grad_norm 1.2990 (1.3118) [2021-04-14 17:01:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][40/1251] eta 0:10:17 lr 0.000003 time 0.2735 (0.5100) loss 6.9028 (6.9729) grad_norm 1.3277 (1.3059) [2021-04-14 17:01:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][50/1251] eta 0:09:17 lr 0.000003 time 0.2643 (0.4640) loss 6.9049 (6.9705) grad_norm 1.2854 (1.3043) [2021-04-14 17:01:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][60/1251] eta 0:08:35 lr 0.000003 time 0.2607 (0.4328) loss 6.8786 (6.9667) grad_norm 1.2584 (1.2967) [2021-04-14 17:01:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][70/1251] eta 0:08:04 lr 0.000004 time 0.2588 (0.4106) loss 6.9867 (6.9642) grad_norm 1.3565 (1.2964) [2021-04-14 17:01:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][80/1251] eta 0:07:40 lr 0.000004 time 0.2458 (0.3935) loss 6.9551 (6.9626) grad_norm 1.2073 (1.2926) [2021-04-14 17:01:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][90/1251] eta 0:07:21 lr 0.000005 time 0.2634 (0.3807) loss 6.9917 (6.9636) grad_norm 1.2589 (1.2877) [2021-04-14 17:01:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][100/1251] eta 0:07:05 lr 0.000005 time 0.2654 (0.3700) loss 6.9564 (6.9611) grad_norm 1.1933 (1.2819) [2021-04-14 17:01:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][110/1251] eta 0:06:52 lr 0.000005 time 0.2784 (0.3616) loss 6.9462 (6.9568) grad_norm 1.2204 (1.2770) [2021-04-14 17:01:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][120/1251] eta 0:06:40 lr 0.000006 time 0.2691 (0.3540) loss 6.9257 (6.9545) grad_norm 1.1992 (1.2720) [2021-04-14 17:01:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][130/1251] eta 0:06:30 lr 0.000006 time 0.3121 (0.3484) loss 6.9216 (6.9531) grad_norm 1.2322 (1.2658) [2021-04-14 17:01:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][140/1251] eta 0:06:21 lr 0.000007 time 0.2555 (0.3430) loss 6.9195 (6.9519) grad_norm 1.1285 (1.2589) [2021-04-14 17:01:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][150/1251] eta 0:06:12 lr 0.000007 time 0.2937 (0.3385) loss 6.9369 (6.9496) grad_norm 1.1152 (1.2543) [2021-04-14 17:01:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][160/1251] eta 0:06:05 lr 0.000007 time 0.2486 (0.3347) loss 6.8838 (6.9482) grad_norm 1.1315 (1.2491) [2021-04-14 17:02:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][170/1251] eta 0:05:57 lr 0.000008 time 0.2684 (0.3310) loss 6.9598 (6.9463) grad_norm 1.0862 (1.2420) [2021-04-14 17:02:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][180/1251] eta 0:05:51 lr 0.000008 time 0.2592 (0.3278) loss 6.8804 (6.9447) grad_norm 1.0872 (1.2371) [2021-04-14 17:02:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][190/1251] eta 0:05:44 lr 0.000009 time 0.2538 (0.3250) loss 6.9036 (6.9430) grad_norm 1.1741 (1.2307) [2021-04-14 17:02:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][200/1251] eta 0:05:38 lr 0.000009 time 0.2826 (0.3224) loss 6.9005 (6.9411) grad_norm 1.1484 (1.2247) [2021-04-14 17:02:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][210/1251] eta 0:05:32 lr 0.000009 time 0.2397 (0.3196) loss 6.9322 (6.9390) grad_norm 1.0956 (1.2185) [2021-04-14 17:02:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][220/1251] eta 0:05:27 lr 0.000010 time 0.2729 (0.3178) loss 6.9389 (6.9381) grad_norm 1.0823 (1.2129) [2021-04-14 17:02:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][230/1251] eta 0:05:22 lr 0.000010 time 0.2737 (0.3159) loss 6.9271 (6.9365) grad_norm 1.1214 (1.2066) [2021-04-14 17:02:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][240/1251] eta 0:05:17 lr 0.000011 time 0.2654 (0.3140) loss 6.8758 (6.9347) grad_norm 1.0682 (1.2009) [2021-04-14 17:02:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][250/1251] eta 0:05:12 lr 0.000011 time 0.2614 (0.3126) loss 6.9027 (6.9325) grad_norm 1.1073 (1.1953) [2021-04-14 17:02:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][260/1251] eta 0:05:08 lr 0.000011 time 0.2588 (0.3111) loss 6.9722 (6.9313) grad_norm 1.0610 (1.1895) [2021-04-14 17:02:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][270/1251] eta 0:05:03 lr 0.000012 time 0.2747 (0.3096) loss 6.9002 (6.9297) grad_norm 1.1150 (1.1847) [2021-04-14 17:02:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][280/1251] eta 0:04:59 lr 0.000012 time 0.2617 (0.3089) loss 6.8935 (6.9284) grad_norm 1.0796 (1.1794) [2021-04-14 17:02:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][290/1251] eta 0:04:55 lr 0.000013 time 0.2949 (0.3080) loss 6.9205 (6.9276) grad_norm 1.0083 (1.1742) [2021-04-14 17:02:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][300/1251] eta 0:04:51 lr 0.000013 time 0.2674 (0.3066) loss 6.9514 (6.9264) grad_norm 0.9970 (1.1689) [2021-04-14 17:02:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][310/1251] eta 0:04:47 lr 0.000013 time 0.2744 (0.3058) loss 6.9036 (6.9258) grad_norm 1.0546 (1.1641) [2021-04-14 17:02:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][320/1251] eta 0:04:43 lr 0.000014 time 0.2712 (0.3048) loss 6.8810 (6.9244) grad_norm 0.9710 (1.1593) [2021-04-14 17:02:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][330/1251] eta 0:04:40 lr 0.000014 time 0.2666 (0.3040) loss 6.7968 (6.9233) grad_norm 0.9553 (1.1538) [2021-04-14 17:02:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][340/1251] eta 0:04:36 lr 0.000015 time 0.2798 (0.3031) loss 6.8848 (6.9223) grad_norm 0.9909 (1.1489) [2021-04-14 17:02:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][350/1251] eta 0:04:32 lr 0.000015 time 0.2737 (0.3023) loss 6.8510 (6.9209) grad_norm 0.9839 (1.1441) [2021-04-14 17:02:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][360/1251] eta 0:04:28 lr 0.000015 time 0.2679 (0.3015) loss 6.9096 (6.9201) grad_norm 0.9430 (1.1394) [2021-04-14 17:02:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][370/1251] eta 0:04:25 lr 0.000016 time 0.2709 (0.3010) loss 6.8275 (6.9185) grad_norm 0.9511 (1.1348) [2021-04-14 17:03:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][380/1251] eta 0:04:21 lr 0.000016 time 0.2796 (0.3001) loss 6.8382 (6.9172) grad_norm 0.9516 (1.1301) [2021-04-14 17:03:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][390/1251] eta 0:04:17 lr 0.000017 time 0.2837 (0.2995) loss 6.9063 (6.9165) grad_norm 0.9653 (1.1259) [2021-04-14 17:03:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][400/1251] eta 0:04:14 lr 0.000017 time 0.2852 (0.2987) loss 6.8317 (6.9148) grad_norm 0.9626 (1.1215) [2021-04-14 17:03:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][410/1251] eta 0:04:10 lr 0.000017 time 0.2650 (0.2981) loss 6.9114 (6.9140) grad_norm 0.9098 (1.1178) [2021-04-14 17:03:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][420/1251] eta 0:04:07 lr 0.000018 time 0.2707 (0.2974) loss 6.9037 (6.9133) grad_norm 0.9682 (1.1137) [2021-04-14 17:03:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][430/1251] eta 0:04:03 lr 0.000018 time 0.2954 (0.2970) loss 6.8653 (6.9125) grad_norm 0.9847 (1.1103) [2021-04-14 17:03:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][440/1251] eta 0:04:00 lr 0.000019 time 0.2544 (0.2965) loss 6.9066 (6.9115) grad_norm 0.9396 (1.1060) [2021-04-14 17:03:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][450/1251] eta 0:03:57 lr 0.000019 time 0.2946 (0.2961) loss 6.9231 (6.9108) grad_norm 0.8852 (1.1018) [2021-04-14 17:03:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][460/1251] eta 0:03:53 lr 0.000019 time 0.2751 (0.2955) loss 6.8529 (6.9095) grad_norm 0.8896 (1.0982) [2021-04-14 17:03:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][470/1251] eta 0:03:50 lr 0.000020 time 0.2578 (0.2950) loss 6.8900 (6.9082) grad_norm 0.9263 (1.0948) [2021-04-14 17:03:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][480/1251] eta 0:03:47 lr 0.000020 time 0.2828 (0.2948) loss 6.8791 (6.9074) grad_norm 0.8919 (1.0914) [2021-04-14 17:03:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][490/1251] eta 0:03:44 lr 0.000021 time 0.2623 (0.2944) loss 6.8644 (6.9067) grad_norm 0.9018 (1.0881) [2021-04-14 17:03:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][500/1251] eta 0:03:40 lr 0.000021 time 0.2728 (0.2941) loss 6.9042 (6.9059) grad_norm 1.0181 (1.0852) [2021-04-14 17:03:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][510/1251] eta 0:03:37 lr 0.000021 time 0.2506 (0.2936) loss 6.8823 (6.9050) grad_norm 0.8803 (1.0827) [2021-04-14 17:03:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][520/1251] eta 0:03:34 lr 0.000022 time 0.2757 (0.2932) loss 6.7821 (6.9042) grad_norm 0.8545 (1.0792) [2021-04-14 17:03:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][530/1251] eta 0:03:31 lr 0.000022 time 0.2861 (0.2929) loss 6.9051 (6.9031) grad_norm 0.8255 (1.0760) [2021-04-14 17:03:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][540/1251] eta 0:03:27 lr 0.000023 time 0.2728 (0.2925) loss 6.8754 (6.9022) grad_norm 0.8650 (1.0727) [2021-04-14 17:03:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][550/1251] eta 0:03:24 lr 0.000023 time 0.2882 (0.2922) loss 6.8638 (6.9014) grad_norm 0.8805 (1.0694) [2021-04-14 17:03:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][560/1251] eta 0:03:21 lr 0.000023 time 0.2785 (0.2919) loss 6.8898 (6.9011) grad_norm 0.9420 (1.0660) [2021-04-14 17:03:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][570/1251] eta 0:03:18 lr 0.000024 time 0.2725 (0.2915) loss 6.8420 (6.8999) grad_norm 0.9469 (1.0634) [2021-04-14 17:03:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][580/1251] eta 0:03:15 lr 0.000024 time 0.2650 (0.2912) loss 6.9154 (6.8990) grad_norm 0.8985 (1.0604) [2021-04-14 17:03:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][590/1251] eta 0:03:12 lr 0.000025 time 0.2673 (0.2909) loss 6.9259 (6.8985) grad_norm 0.9013 (1.0580) [2021-04-14 17:04:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][600/1251] eta 0:03:09 lr 0.000025 time 0.2764 (0.2906) loss 6.8004 (6.8977) grad_norm 0.8896 (1.0551) [2021-04-14 17:04:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][610/1251] eta 0:03:06 lr 0.000025 time 0.2520 (0.2903) loss 6.8734 (6.8972) grad_norm 0.8504 (1.0522) [2021-04-14 17:04:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][620/1251] eta 0:03:02 lr 0.000026 time 0.2628 (0.2900) loss 6.8651 (6.8967) grad_norm 0.8971 (1.0495) [2021-04-14 17:04:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][630/1251] eta 0:02:59 lr 0.000026 time 0.2718 (0.2898) loss 6.8935 (6.8961) grad_norm 0.9111 (1.0475) [2021-04-14 17:04:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][640/1251] eta 0:02:56 lr 0.000027 time 0.2834 (0.2895) loss 6.7450 (6.8951) grad_norm 0.9334 (1.0453) [2021-04-14 17:04:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][650/1251] eta 0:02:53 lr 0.000027 time 0.2829 (0.2892) loss 6.8672 (6.8942) grad_norm 0.9755 (1.0437) [2021-04-14 17:04:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][660/1251] eta 0:02:50 lr 0.000027 time 0.2711 (0.2889) loss 6.7075 (6.8928) grad_norm 0.8929 (1.0417) [2021-04-14 17:04:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][670/1251] eta 0:02:47 lr 0.000028 time 0.2937 (0.2888) loss 6.7455 (6.8917) grad_norm 0.9257 (1.0401) [2021-04-14 17:04:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][680/1251] eta 0:02:44 lr 0.000028 time 0.2976 (0.2887) loss 6.8718 (6.8912) grad_norm 0.8704 (1.0380) [2021-04-14 17:04:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][690/1251] eta 0:02:41 lr 0.000029 time 0.2851 (0.2884) loss 6.8539 (6.8904) grad_norm 0.8773 (1.0362) [2021-04-14 17:04:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][700/1251] eta 0:02:38 lr 0.000029 time 0.2534 (0.2882) loss 6.8489 (6.8897) grad_norm 0.9358 (1.0355) [2021-04-14 17:04:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][710/1251] eta 0:02:35 lr 0.000029 time 0.2492 (0.2880) loss 6.8824 (6.8890) grad_norm 0.9443 (1.0340) [2021-04-14 17:04:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][720/1251] eta 0:02:32 lr 0.000030 time 0.2706 (0.2878) loss 6.9212 (6.8883) grad_norm 0.9511 (1.0326) [2021-04-14 17:04:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][730/1251] eta 0:02:29 lr 0.000030 time 0.2768 (0.2876) loss 6.8178 (6.8878) grad_norm 0.9722 (1.0315) [2021-04-14 17:04:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][740/1251] eta 0:02:26 lr 0.000031 time 0.2534 (0.2874) loss 6.8643 (6.8872) grad_norm 1.1172 (1.0311) [2021-04-14 17:04:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][750/1251] eta 0:02:23 lr 0.000031 time 0.2756 (0.2873) loss 6.8527 (6.8865) grad_norm 0.8614 (1.0299) [2021-04-14 17:04:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][760/1251] eta 0:02:20 lr 0.000031 time 0.2676 (0.2871) loss 6.8533 (6.8860) grad_norm 0.8665 (1.0294) [2021-04-14 17:04:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][770/1251] eta 0:02:18 lr 0.000032 time 0.2928 (0.2870) loss 6.8925 (6.8854) grad_norm 1.2236 (1.0284) [2021-04-14 17:04:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][780/1251] eta 0:02:15 lr 0.000032 time 0.2771 (0.2868) loss 6.8729 (6.8849) grad_norm 0.9501 (1.0273) [2021-04-14 17:04:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][790/1251] eta 0:02:12 lr 0.000033 time 0.2891 (0.2866) loss 6.7559 (6.8839) grad_norm 1.0223 (1.0271) [2021-04-14 17:04:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][800/1251] eta 0:02:09 lr 0.000033 time 0.2518 (0.2863) loss 6.8624 (6.8831) grad_norm 1.0292 (1.0273) [2021-04-14 17:04:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][810/1251] eta 0:02:06 lr 0.000033 time 0.2822 (0.2861) loss 6.7593 (6.8819) grad_norm 0.9780 (1.0285) [2021-04-14 17:05:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][820/1251] eta 0:02:03 lr 0.000034 time 0.2537 (0.2859) loss 6.8350 (6.8812) grad_norm 1.0397 (1.0308) [2021-04-14 17:05:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][830/1251] eta 0:02:00 lr 0.000034 time 0.2769 (0.2858) loss 6.7821 (6.8804) grad_norm 1.4438 (1.0334) [2021-04-14 17:05:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][840/1251] eta 0:01:57 lr 0.000035 time 0.2569 (0.2856) loss 6.8040 (6.8796) grad_norm 0.9804 (1.0346) [2021-04-14 17:05:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][850/1251] eta 0:01:54 lr 0.000035 time 0.2694 (0.2855) loss 6.9177 (6.8787) grad_norm 1.8132 (1.0368) [2021-04-14 17:05:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][860/1251] eta 0:01:51 lr 0.000035 time 0.2748 (0.2855) loss 6.8453 (6.8777) grad_norm 1.0434 (1.0393) [2021-04-14 17:05:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][870/1251] eta 0:01:48 lr 0.000036 time 0.2867 (0.2854) loss 6.7598 (6.8770) grad_norm 1.3571 (1.0424) [2021-04-14 17:05:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][880/1251] eta 0:01:45 lr 0.000036 time 0.2646 (0.2852) loss 6.7836 (6.8763) grad_norm 1.1242 (1.0434) [2021-04-14 17:05:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][890/1251] eta 0:01:42 lr 0.000037 time 0.2729 (0.2851) loss 6.6962 (6.8750) grad_norm 1.0459 (1.0442) [2021-04-14 17:05:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][900/1251] eta 0:01:40 lr 0.000037 time 0.2721 (0.2850) loss 6.8686 (6.8741) grad_norm 1.1377 (1.0479) [2021-04-14 17:05:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][910/1251] eta 0:01:37 lr 0.000037 time 0.2764 (0.2849) loss 6.8187 (6.8734) grad_norm 1.4067 (1.0503) [2021-04-14 17:05:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][920/1251] eta 0:01:34 lr 0.000038 time 0.2524 (0.2848) loss 6.8002 (6.8724) grad_norm 1.3676 (1.0523) [2021-04-14 17:05:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][930/1251] eta 0:01:31 lr 0.000038 time 0.2869 (0.2847) loss 6.7870 (6.8718) grad_norm 1.2930 (1.0547) [2021-04-14 17:05:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][940/1251] eta 0:01:28 lr 0.000039 time 0.2767 (0.2846) loss 6.8021 (6.8707) grad_norm 1.4340 (1.0588) [2021-04-14 17:05:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][950/1251] eta 0:01:25 lr 0.000039 time 0.2839 (0.2846) loss 6.7537 (6.8691) grad_norm 1.2392 (1.0614) [2021-04-14 17:05:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][960/1251] eta 0:01:22 lr 0.000039 time 0.2665 (0.2844) loss 6.8242 (6.8681) grad_norm 1.3403 (1.0654) [2021-04-14 17:05:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][970/1251] eta 0:01:19 lr 0.000040 time 0.2727 (0.2843) loss 6.8711 (6.8676) grad_norm 1.1287 (1.0671) [2021-04-14 17:05:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][980/1251] eta 0:01:17 lr 0.000040 time 0.2829 (0.2842) loss 6.8385 (6.8668) grad_norm 1.0450 (1.0685) [2021-04-14 17:05:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][990/1251] eta 0:01:14 lr 0.000041 time 0.2759 (0.2841) loss 6.7892 (6.8660) grad_norm 1.2169 (1.0711) [2021-04-14 17:05:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][1000/1251] eta 0:01:11 lr 0.000041 time 0.2462 (0.2839) loss 6.8259 (6.8650) grad_norm 1.4554 (1.0730) [2021-04-14 17:05:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][1010/1251] eta 0:01:08 lr 0.000041 time 0.2583 (0.2838) loss 6.8402 (6.8645) grad_norm 2.5297 (1.0788) [2021-04-14 17:05:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][1020/1251] eta 0:01:05 lr 0.000042 time 0.3026 (0.2838) loss 6.7809 (6.8637) grad_norm 1.1613 (1.0820) [2021-04-14 17:05:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][1030/1251] eta 0:01:02 lr 0.000042 time 0.2665 (0.2837) loss 6.8272 (6.8626) grad_norm 1.0303 (1.0845) [2021-04-14 17:06:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][1040/1251] eta 0:00:59 lr 0.000043 time 0.2734 (0.2835) loss 6.8312 (6.8617) grad_norm 1.5744 (1.0873) [2021-04-14 17:06:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][1050/1251] eta 0:00:56 lr 0.000043 time 0.2843 (0.2834) loss 6.6566 (6.8606) grad_norm 1.2993 (1.0909) [2021-04-14 17:06:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][1060/1251] eta 0:00:54 lr 0.000043 time 0.2944 (0.2834) loss 6.7530 (6.8598) grad_norm 1.6495 (1.0950) [2021-04-14 17:06:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][1070/1251] eta 0:00:51 lr 0.000044 time 0.2676 (0.2833) loss 6.8030 (6.8588) grad_norm 1.2940 (1.0969) [2021-04-14 17:06:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][1080/1251] eta 0:00:48 lr 0.000044 time 0.2649 (0.2832) loss 6.6563 (6.8577) grad_norm 1.3687 (1.0983) [2021-04-14 17:06:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][1090/1251] eta 0:00:45 lr 0.000045 time 0.2782 (0.2831) loss 6.6968 (6.8562) grad_norm 1.1389 (1.1015) [2021-04-14 17:06:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][1100/1251] eta 0:00:42 lr 0.000045 time 0.2525 (0.2830) loss 6.7468 (6.8551) grad_norm 1.3334 (1.1041) [2021-04-14 17:06:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][1110/1251] eta 0:00:39 lr 0.000045 time 0.2832 (0.2830) loss 6.8240 (6.8541) grad_norm 2.5684 (1.1088) [2021-04-14 17:06:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][1120/1251] eta 0:00:37 lr 0.000046 time 0.2600 (0.2830) loss 6.6921 (6.8534) grad_norm 1.0225 (1.1112) [2021-04-14 17:06:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][1130/1251] eta 0:00:34 lr 0.000046 time 0.2633 (0.2829) loss 6.9559 (6.8525) grad_norm 1.4511 (1.1143) [2021-04-14 17:06:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][1140/1251] eta 0:00:31 lr 0.000047 time 0.2763 (0.2828) loss 6.7313 (6.8513) grad_norm 1.0877 (1.1165) [2021-04-14 17:06:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][1150/1251] eta 0:00:28 lr 0.000047 time 0.2670 (0.2828) loss 6.8022 (6.8505) grad_norm 1.1120 (1.1194) [2021-04-14 17:06:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][1160/1251] eta 0:00:25 lr 0.000047 time 0.2758 (0.2827) loss 6.7772 (6.8495) grad_norm 1.9082 (1.1219) [2021-04-14 17:06:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][1170/1251] eta 0:00:22 lr 0.000048 time 0.2658 (0.2826) loss 6.6564 (6.8482) grad_norm 2.1344 (1.1251) [2021-04-14 17:06:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][1180/1251] eta 0:00:20 lr 0.000048 time 0.2706 (0.2825) loss 6.7728 (6.8475) grad_norm 1.5160 (1.1284) [2021-04-14 17:06:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][1190/1251] eta 0:00:17 lr 0.000049 time 0.3048 (0.2824) loss 6.6751 (6.8466) grad_norm 1.4668 (1.1301) [2021-04-14 17:06:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][1200/1251] eta 0:00:14 lr 0.000049 time 0.2926 (0.2824) loss 6.7172 (6.8455) grad_norm 1.2001 (1.1323) [2021-04-14 17:06:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][1210/1251] eta 0:00:11 lr 0.000049 time 0.2696 (0.2823) loss 6.5749 (6.8446) grad_norm 1.5306 (1.1340) [2021-04-14 17:06:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][1220/1251] eta 0:00:08 lr 0.000050 time 0.2771 (0.2822) loss 6.7366 (6.8435) grad_norm 1.4217 (1.1359) [2021-04-14 17:06:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][1230/1251] eta 0:00:05 lr 0.000050 time 0.2826 (0.2822) loss 6.8514 (6.8427) grad_norm 1.5669 (1.1404) [2021-04-14 17:06:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][1240/1251] eta 0:00:03 lr 0.000051 time 0.2503 (0.2820) loss 6.7135 (6.8417) grad_norm 1.8412 (1.1462) [2021-04-14 17:06:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [0/300][1250/1251] eta 0:00:00 lr 0.000051 time 0.2482 (0.2818) loss 6.6932 (6.8405) grad_norm 2.2573 (1.1500) [2021-04-14 17:06:59 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 0 training takes 0:05:53 [2021-04-14 17:06:59 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_0.pth saving...... [2021-04-14 17:07:15 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_0.pth saved !!! [2021-04-14 17:07:16 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.248 (1.248) Loss 6.4004 (6.4004) Acc@1 1.758 (1.758) Acc@5 5.957 (5.957) [2021-04-14 17:07:18 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.130 (0.232) Loss 6.3504 (6.3438) Acc@1 1.465 (1.749) Acc@5 5.566 (5.886) [2021-04-14 17:07:20 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.132 (0.232) Loss 6.3202 (6.3390) Acc@1 2.930 (1.981) Acc@5 6.250 (6.148) [2021-04-14 17:07:22 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.108 (0.226) Loss 6.3383 (6.3350) Acc@1 2.246 (1.937) Acc@5 6.934 (6.370) [2021-04-14 17:07:24 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.285 (0.206) Loss 6.3227 (6.3366) Acc@1 2.051 (1.917) Acc@5 6.641 (6.298) [2021-04-14 17:07:26 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 1.976 Acc@5 6.344 [2021-04-14 17:07:26 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 2.0% [2021-04-14 17:07:26 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 1.98% [2021-04-14 17:07:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][0/1251] eta 1:19:38 lr 0.000051 time 3.8199 (3.8199) loss 6.6686 (6.6686) grad_norm 1.6749 (1.6749) [2021-04-14 17:07:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][10/1251] eta 0:12:12 lr 0.000051 time 0.2854 (0.5901) loss 6.6285 (6.6958) grad_norm 1.3831 (1.5152) [2021-04-14 17:07:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][20/1251] eta 0:08:58 lr 0.000052 time 0.2756 (0.4377) loss 6.5696 (6.6971) grad_norm 1.6575 (1.5164) [2021-04-14 17:07:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][30/1251] eta 0:07:47 lr 0.000052 time 0.2735 (0.3825) loss 6.7368 (6.7095) grad_norm 1.4513 (1.5776) [2021-04-14 17:07:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][40/1251] eta 0:07:10 lr 0.000053 time 0.2548 (0.3551) loss 6.6135 (6.7133) grad_norm 1.8527 (1.5557) [2021-04-14 17:07:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][50/1251] eta 0:06:46 lr 0.000053 time 0.2776 (0.3384) loss 6.5558 (6.7064) grad_norm 1.5683 (1.5858) [2021-04-14 17:07:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][60/1251] eta 0:06:35 lr 0.000053 time 0.2750 (0.3320) loss 6.7153 (6.7017) grad_norm 1.7683 (1.5870) [2021-04-14 17:07:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][70/1251] eta 0:06:23 lr 0.000054 time 0.2765 (0.3248) loss 6.6225 (6.6905) grad_norm 1.5825 (1.5752) [2021-04-14 17:07:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][80/1251] eta 0:06:12 lr 0.000054 time 0.2652 (0.3183) loss 6.5659 (6.6895) grad_norm 1.5846 (1.5517) [2021-04-14 17:07:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][90/1251] eta 0:06:03 lr 0.000055 time 0.2679 (0.3133) loss 6.5658 (6.6937) grad_norm 1.6956 (1.5582) [2021-04-14 17:07:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][100/1251] eta 0:05:56 lr 0.000055 time 0.2942 (0.3095) loss 6.6901 (6.6970) grad_norm 1.4660 (1.5661) [2021-04-14 17:08:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][110/1251] eta 0:05:49 lr 0.000055 time 0.2505 (0.3062) loss 6.7395 (6.6955) grad_norm 1.9643 (1.5892) [2021-04-14 17:08:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][120/1251] eta 0:05:43 lr 0.000056 time 0.2980 (0.3038) loss 6.7090 (6.6951) grad_norm 2.7206 (1.6065) [2021-04-14 17:08:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][130/1251] eta 0:05:37 lr 0.000056 time 0.3074 (0.3014) loss 6.6530 (6.6928) grad_norm 1.6207 (1.6227) [2021-04-14 17:08:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][140/1251] eta 0:05:32 lr 0.000057 time 0.2704 (0.2992) loss 6.4141 (6.6905) grad_norm 1.3687 (1.6321) [2021-04-14 17:08:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][150/1251] eta 0:05:27 lr 0.000057 time 0.2745 (0.2975) loss 6.8422 (6.6909) grad_norm 1.5078 (1.6391) [2021-04-14 17:08:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][160/1251] eta 0:05:22 lr 0.000057 time 0.2979 (0.2957) loss 6.6337 (6.6906) grad_norm 2.3537 (1.6654) [2021-04-14 17:08:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][170/1251] eta 0:05:18 lr 0.000058 time 0.2747 (0.2943) loss 6.7236 (6.6901) grad_norm 1.7420 (1.6762) [2021-04-14 17:08:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][180/1251] eta 0:05:14 lr 0.000058 time 0.2859 (0.2935) loss 6.6612 (6.6877) grad_norm 1.4361 (1.6728) [2021-04-14 17:08:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][190/1251] eta 0:05:10 lr 0.000059 time 0.2911 (0.2926) loss 6.8383 (6.6855) grad_norm 1.2576 (1.6619) [2021-04-14 17:08:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][200/1251] eta 0:05:06 lr 0.000059 time 0.2967 (0.2916) loss 6.6545 (6.6834) grad_norm 1.4201 (1.6620) [2021-04-14 17:08:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][210/1251] eta 0:05:02 lr 0.000059 time 0.2797 (0.2905) loss 6.5936 (6.6829) grad_norm 1.4156 (1.6536) [2021-04-14 17:08:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][220/1251] eta 0:04:59 lr 0.000060 time 0.2683 (0.2901) loss 6.7105 (6.6809) grad_norm 1.8728 (1.6566) [2021-04-14 17:08:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][230/1251] eta 0:04:55 lr 0.000060 time 0.2735 (0.2894) loss 6.5056 (6.6786) grad_norm 1.7832 (1.6607) [2021-04-14 17:08:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][240/1251] eta 0:04:51 lr 0.000061 time 0.2836 (0.2886) loss 6.7048 (6.6768) grad_norm 1.4232 (1.6676) [2021-04-14 17:08:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][250/1251] eta 0:04:48 lr 0.000061 time 0.2874 (0.2880) loss 6.6598 (6.6742) grad_norm 1.5733 (1.6618) [2021-04-14 17:08:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][260/1251] eta 0:04:44 lr 0.000061 time 0.2722 (0.2874) loss 6.6268 (6.6741) grad_norm 1.8022 (1.6643) [2021-04-14 17:08:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][270/1251] eta 0:04:41 lr 0.000062 time 0.2501 (0.2867) loss 6.6284 (6.6723) grad_norm 1.3320 (1.6582) [2021-04-14 17:08:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][280/1251] eta 0:04:37 lr 0.000062 time 0.2649 (0.2861) loss 6.6783 (6.6696) grad_norm 1.3751 (1.6580) [2021-04-14 17:08:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][290/1251] eta 0:04:34 lr 0.000063 time 0.2913 (0.2858) loss 6.6632 (6.6693) grad_norm 1.7753 (1.6629) [2021-04-14 17:08:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][300/1251] eta 0:04:31 lr 0.000063 time 0.2680 (0.2851) loss 6.4885 (6.6668) grad_norm 1.5455 (1.6723) [2021-04-14 17:08:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][310/1251] eta 0:04:27 lr 0.000063 time 0.2830 (0.2847) loss 6.3299 (6.6649) grad_norm 1.7026 (1.6735) [2021-04-14 17:08:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][320/1251] eta 0:04:24 lr 0.000064 time 0.2647 (0.2843) loss 6.7175 (6.6624) grad_norm 1.5152 (1.6840) [2021-04-14 17:09:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][330/1251] eta 0:04:21 lr 0.000064 time 0.2640 (0.2839) loss 6.6797 (6.6609) grad_norm 1.5124 (1.6832) [2021-04-14 17:09:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][340/1251] eta 0:04:18 lr 0.000065 time 0.2658 (0.2835) loss 6.8140 (6.6609) grad_norm 2.1521 (1.6818) [2021-04-14 17:09:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][350/1251] eta 0:04:15 lr 0.000065 time 0.2692 (0.2833) loss 6.6832 (6.6617) grad_norm 1.7242 (1.6788) [2021-04-14 17:09:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][360/1251] eta 0:04:12 lr 0.000065 time 0.2551 (0.2831) loss 6.7321 (6.6606) grad_norm 1.7078 (1.6774) [2021-04-14 17:09:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][370/1251] eta 0:04:09 lr 0.000066 time 0.2559 (0.2828) loss 6.7222 (6.6587) grad_norm 1.6985 (1.6834) [2021-04-14 17:09:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][380/1251] eta 0:04:06 lr 0.000066 time 0.2762 (0.2825) loss 6.6269 (6.6571) grad_norm 1.6178 (1.6874) [2021-04-14 17:09:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][390/1251] eta 0:04:02 lr 0.000067 time 0.2795 (0.2822) loss 6.6833 (6.6555) grad_norm 1.4431 (1.6864) [2021-04-14 17:09:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][400/1251] eta 0:04:00 lr 0.000067 time 0.2565 (0.2820) loss 6.4134 (6.6512) grad_norm 1.7455 (1.6937) [2021-04-14 17:09:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][410/1251] eta 0:03:57 lr 0.000067 time 0.2634 (0.2821) loss 6.5697 (6.6481) grad_norm 1.4964 (1.6937) [2021-04-14 17:09:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][420/1251] eta 0:03:54 lr 0.000068 time 0.2843 (0.2817) loss 6.4311 (6.6470) grad_norm 2.1469 (1.7006) [2021-04-14 17:09:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][430/1251] eta 0:03:51 lr 0.000068 time 0.2667 (0.2816) loss 6.5445 (6.6454) grad_norm 2.1868 (1.7031) [2021-04-14 17:09:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][440/1251] eta 0:03:48 lr 0.000069 time 0.2499 (0.2815) loss 6.6195 (6.6447) grad_norm 1.8369 (1.7051) [2021-04-14 17:09:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][450/1251] eta 0:03:45 lr 0.000069 time 0.2686 (0.2812) loss 6.7480 (6.6447) grad_norm 2.1284 (1.7102) [2021-04-14 17:09:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][460/1251] eta 0:03:42 lr 0.000069 time 0.2760 (0.2810) loss 6.6020 (6.6429) grad_norm 1.3659 (1.7123) [2021-04-14 17:09:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][470/1251] eta 0:03:39 lr 0.000070 time 0.2582 (0.2808) loss 6.6896 (6.6398) grad_norm 1.7177 (1.7105) [2021-04-14 17:09:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][480/1251] eta 0:03:36 lr 0.000070 time 0.2685 (0.2805) loss 6.7306 (6.6404) grad_norm 2.5551 (1.7173) [2021-04-14 17:09:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][490/1251] eta 0:03:33 lr 0.000071 time 0.2638 (0.2803) loss 6.6871 (6.6388) grad_norm 1.4029 (1.7197) [2021-04-14 17:09:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][500/1251] eta 0:03:30 lr 0.000071 time 0.2790 (0.2801) loss 6.5054 (6.6372) grad_norm 1.7532 (1.7216) [2021-04-14 17:09:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][510/1251] eta 0:03:27 lr 0.000071 time 0.2571 (0.2801) loss 6.6676 (6.6362) grad_norm 1.8075 (1.7242) [2021-04-14 17:09:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][520/1251] eta 0:03:24 lr 0.000072 time 0.2861 (0.2799) loss 6.4356 (6.6353) grad_norm 1.7711 (1.7237) [2021-04-14 17:09:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][530/1251] eta 0:03:21 lr 0.000072 time 0.2858 (0.2798) loss 6.5078 (6.6346) grad_norm 2.1196 (1.7261) [2021-04-14 17:09:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][540/1251] eta 0:03:18 lr 0.000073 time 0.2783 (0.2797) loss 6.6174 (6.6325) grad_norm 1.6862 (1.7286) [2021-04-14 17:10:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][550/1251] eta 0:03:15 lr 0.000073 time 0.2579 (0.2795) loss 6.6643 (6.6322) grad_norm 2.0569 (1.7337) [2021-04-14 17:10:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][560/1251] eta 0:03:13 lr 0.000073 time 0.2596 (0.2795) loss 6.5517 (6.6310) grad_norm 2.2164 (1.7371) [2021-04-14 17:10:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][570/1251] eta 0:03:10 lr 0.000074 time 0.2658 (0.2794) loss 6.2701 (6.6290) grad_norm 1.5789 (1.7383) [2021-04-14 17:10:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][580/1251] eta 0:03:07 lr 0.000074 time 0.2694 (0.2792) loss 6.6437 (6.6287) grad_norm 1.7047 (1.7353) [2021-04-14 17:10:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][590/1251] eta 0:03:04 lr 0.000075 time 0.2696 (0.2792) loss 6.3765 (6.6262) grad_norm 1.4841 (1.7352) [2021-04-14 17:10:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][600/1251] eta 0:03:01 lr 0.000075 time 0.2675 (0.2791) loss 6.4872 (6.6240) grad_norm 2.1664 (1.7332) [2021-04-14 17:10:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][610/1251] eta 0:02:58 lr 0.000075 time 0.2730 (0.2791) loss 6.6310 (6.6233) grad_norm 1.9055 (1.7362) [2021-04-14 17:10:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][620/1251] eta 0:02:56 lr 0.000076 time 0.2565 (0.2790) loss 6.5132 (6.6211) grad_norm 1.8457 (1.7394) [2021-04-14 17:10:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][630/1251] eta 0:02:53 lr 0.000076 time 0.2712 (0.2790) loss 6.5333 (6.6208) grad_norm 2.1478 (1.7418) [2021-04-14 17:10:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][640/1251] eta 0:02:50 lr 0.000077 time 0.2762 (0.2792) loss 6.3358 (6.6183) grad_norm 1.9716 (1.7444) [2021-04-14 17:10:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][650/1251] eta 0:02:47 lr 0.000077 time 0.2811 (0.2792) loss 6.6687 (6.6173) grad_norm 1.9866 (1.7461) [2021-04-14 17:10:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][660/1251] eta 0:02:44 lr 0.000077 time 0.2696 (0.2790) loss 6.6148 (6.6178) grad_norm 1.9566 (1.7493) [2021-04-14 17:10:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][670/1251] eta 0:02:42 lr 0.000078 time 0.2624 (0.2790) loss 6.5514 (6.6159) grad_norm 1.7973 (1.7499) [2021-04-14 17:10:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][680/1251] eta 0:02:39 lr 0.000078 time 0.2655 (0.2789) loss 6.2265 (6.6137) grad_norm 1.9921 (1.7529) [2021-04-14 17:10:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][690/1251] eta 0:02:36 lr 0.000079 time 0.3156 (0.2789) loss 6.5988 (6.6122) grad_norm 1.6052 (1.7521) [2021-04-14 17:10:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][700/1251] eta 0:02:33 lr 0.000079 time 0.2617 (0.2787) loss 6.6122 (6.6108) grad_norm 1.5951 (1.7519) [2021-04-14 17:10:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][710/1251] eta 0:02:30 lr 0.000079 time 0.2932 (0.2787) loss 6.7389 (6.6093) grad_norm 1.9909 (1.7517) [2021-04-14 17:10:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][720/1251] eta 0:02:27 lr 0.000080 time 0.2945 (0.2786) loss 6.5554 (6.6069) grad_norm 1.6522 (1.7520) [2021-04-14 17:10:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][730/1251] eta 0:02:25 lr 0.000080 time 0.2670 (0.2785) loss 6.4333 (6.6057) grad_norm 1.6066 (1.7535) [2021-04-14 17:10:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][740/1251] eta 0:02:22 lr 0.000080 time 0.2533 (0.2785) loss 6.4299 (6.6046) grad_norm 2.2588 (1.7571) [2021-04-14 17:10:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][750/1251] eta 0:02:19 lr 0.000081 time 0.2543 (0.2784) loss 6.3953 (6.6026) grad_norm 1.4663 (1.7555) [2021-04-14 17:10:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][760/1251] eta 0:02:16 lr 0.000081 time 0.2808 (0.2783) loss 6.4958 (6.6025) grad_norm 2.3822 (1.7587) [2021-04-14 17:11:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][770/1251] eta 0:02:13 lr 0.000082 time 0.2732 (0.2783) loss 6.6142 (6.6014) grad_norm 1.3908 (1.7595) [2021-04-14 17:11:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][780/1251] eta 0:02:11 lr 0.000082 time 0.2681 (0.2782) loss 6.7682 (6.6011) grad_norm 1.8471 (1.7621) [2021-04-14 17:11:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][790/1251] eta 0:02:08 lr 0.000082 time 0.2690 (0.2781) loss 6.5551 (6.6001) grad_norm 1.6596 (1.7652) [2021-04-14 17:11:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][800/1251] eta 0:02:05 lr 0.000083 time 0.2789 (0.2780) loss 6.5535 (6.5989) grad_norm 1.9902 (1.7680) [2021-04-14 17:11:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][810/1251] eta 0:02:02 lr 0.000083 time 0.2775 (0.2779) loss 6.2180 (6.5972) grad_norm 2.6156 (1.7700) [2021-04-14 17:11:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][820/1251] eta 0:01:59 lr 0.000084 time 0.2571 (0.2778) loss 6.5701 (6.5963) grad_norm 2.4709 (1.7733) [2021-04-14 17:11:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][830/1251] eta 0:01:56 lr 0.000084 time 0.2621 (0.2778) loss 6.1939 (6.5940) grad_norm 2.0055 (1.7756) [2021-04-14 17:11:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][840/1251] eta 0:01:54 lr 0.000084 time 0.2696 (0.2777) loss 6.2009 (6.5930) grad_norm 2.4908 (1.7816) [2021-04-14 17:11:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][850/1251] eta 0:01:51 lr 0.000085 time 0.2645 (0.2777) loss 6.2855 (6.5913) grad_norm 1.7945 (1.7838) [2021-04-14 17:11:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][860/1251] eta 0:01:48 lr 0.000085 time 0.2718 (0.2776) loss 6.3364 (6.5903) grad_norm 2.0777 (1.7851) [2021-04-14 17:11:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][870/1251] eta 0:01:45 lr 0.000086 time 0.2775 (0.2775) loss 6.4097 (6.5891) grad_norm 1.9209 (1.7857) [2021-04-14 17:11:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][880/1251] eta 0:01:42 lr 0.000086 time 0.2648 (0.2774) loss 6.5947 (6.5886) grad_norm 2.2944 (1.7864) [2021-04-14 17:11:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][890/1251] eta 0:01:40 lr 0.000086 time 0.2689 (0.2774) loss 6.6072 (6.5881) grad_norm 2.0992 (1.7875) [2021-04-14 17:11:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][900/1251] eta 0:01:37 lr 0.000087 time 0.2735 (0.2773) loss 6.6475 (6.5871) grad_norm 1.8526 (1.7887) [2021-04-14 17:11:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][910/1251] eta 0:01:34 lr 0.000087 time 0.2971 (0.2772) loss 6.4638 (6.5859) grad_norm 2.5455 (1.7896) [2021-04-14 17:11:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][920/1251] eta 0:01:31 lr 0.000088 time 0.2860 (0.2772) loss 6.1440 (6.5851) grad_norm 1.4481 (1.7920) [2021-04-14 17:11:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][930/1251] eta 0:01:28 lr 0.000088 time 0.2617 (0.2771) loss 6.5730 (6.5844) grad_norm 1.8559 (1.7920) [2021-04-14 17:11:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][940/1251] eta 0:01:26 lr 0.000088 time 0.2690 (0.2770) loss 6.3476 (6.5838) grad_norm 2.1673 (1.7971) [2021-04-14 17:11:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][950/1251] eta 0:01:23 lr 0.000089 time 0.2604 (0.2769) loss 6.4453 (6.5822) grad_norm 1.4899 (1.7993) [2021-04-14 17:11:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][960/1251] eta 0:01:20 lr 0.000089 time 0.2551 (0.2769) loss 6.5589 (6.5810) grad_norm 1.6672 (1.7999) [2021-04-14 17:11:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][970/1251] eta 0:01:17 lr 0.000090 time 0.2690 (0.2768) loss 6.5383 (6.5802) grad_norm 2.0450 (1.8010) [2021-04-14 17:11:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][980/1251] eta 0:01:15 lr 0.000090 time 0.2605 (0.2768) loss 6.1270 (6.5782) grad_norm 1.9444 (1.8013) [2021-04-14 17:12:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][990/1251] eta 0:01:12 lr 0.000090 time 0.2441 (0.2767) loss 6.5645 (6.5770) grad_norm 2.6764 (1.8060) [2021-04-14 17:12:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][1000/1251] eta 0:01:09 lr 0.000091 time 0.2502 (0.2767) loss 6.6226 (6.5755) grad_norm 1.8465 (1.8072) [2021-04-14 17:12:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][1010/1251] eta 0:01:06 lr 0.000091 time 0.2940 (0.2768) loss 6.6015 (6.5744) grad_norm 1.8555 (1.8079) [2021-04-14 17:12:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][1020/1251] eta 0:01:03 lr 0.000092 time 0.2644 (0.2767) loss 6.5118 (6.5740) grad_norm 1.9440 (1.8079) [2021-04-14 17:12:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][1030/1251] eta 0:01:01 lr 0.000092 time 0.2576 (0.2767) loss 6.7293 (6.5734) grad_norm 2.0468 (1.8082) [2021-04-14 17:12:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][1040/1251] eta 0:00:58 lr 0.000092 time 0.2812 (0.2766) loss 6.4025 (6.5719) grad_norm 1.8076 (1.8090) [2021-04-14 17:12:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][1050/1251] eta 0:00:55 lr 0.000093 time 0.2867 (0.2766) loss 6.2915 (6.5698) grad_norm 1.7895 (1.8083) [2021-04-14 17:12:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][1060/1251] eta 0:00:52 lr 0.000093 time 0.2831 (0.2765) loss 6.5372 (6.5685) grad_norm 2.5378 (1.8112) [2021-04-14 17:12:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][1070/1251] eta 0:00:50 lr 0.000094 time 0.2796 (0.2765) loss 6.7444 (6.5687) grad_norm 2.4355 (1.8119) [2021-04-14 17:12:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][1080/1251] eta 0:00:47 lr 0.000094 time 0.2740 (0.2764) loss 6.7201 (6.5687) grad_norm 1.6357 (1.8129) [2021-04-14 17:12:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][1090/1251] eta 0:00:44 lr 0.000094 time 0.2567 (0.2764) loss 6.4917 (6.5676) grad_norm 2.2677 (1.8152) [2021-04-14 17:12:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][1100/1251] eta 0:00:41 lr 0.000095 time 0.2698 (0.2763) loss 6.2619 (6.5663) grad_norm 2.1501 (1.8174) [2021-04-14 17:12:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][1110/1251] eta 0:00:38 lr 0.000095 time 0.2492 (0.2763) loss 6.4939 (6.5653) grad_norm 1.9868 (1.8207) [2021-04-14 17:12:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][1120/1251] eta 0:00:36 lr 0.000096 time 0.2883 (0.2763) loss 6.4928 (6.5640) grad_norm 2.0408 (1.8235) [2021-04-14 17:12:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][1130/1251] eta 0:00:33 lr 0.000096 time 0.2667 (0.2762) loss 6.6045 (6.5636) grad_norm 2.4161 (1.8267) [2021-04-14 17:12:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][1140/1251] eta 0:00:30 lr 0.000096 time 0.2557 (0.2762) loss 6.4311 (6.5623) grad_norm 2.0647 (1.8268) [2021-04-14 17:12:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][1150/1251] eta 0:00:27 lr 0.000097 time 0.2750 (0.2762) loss 6.6418 (6.5611) grad_norm 1.6985 (1.8264) [2021-04-14 17:12:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][1160/1251] eta 0:00:25 lr 0.000097 time 0.2917 (0.2761) loss 6.2554 (6.5603) grad_norm 3.0200 (1.8310) [2021-04-14 17:12:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][1170/1251] eta 0:00:22 lr 0.000098 time 0.2761 (0.2761) loss 6.2569 (6.5584) grad_norm 2.0898 (1.8323) [2021-04-14 17:12:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][1180/1251] eta 0:00:19 lr 0.000098 time 0.2802 (0.2760) loss 6.4110 (6.5579) grad_norm 2.2035 (1.8338) [2021-04-14 17:12:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][1190/1251] eta 0:00:16 lr 0.000098 time 0.2690 (0.2760) loss 6.4418 (6.5569) grad_norm 2.0703 (1.8337) [2021-04-14 17:12:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][1200/1251] eta 0:00:14 lr 0.000099 time 0.2840 (0.2760) loss 6.5079 (6.5561) grad_norm 1.6548 (1.8319) [2021-04-14 17:13:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][1210/1251] eta 0:00:11 lr 0.000099 time 0.2879 (0.2760) loss 6.1022 (6.5550) grad_norm 1.8929 (1.8312) [2021-04-14 17:13:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][1220/1251] eta 0:00:08 lr 0.000100 time 0.2517 (0.2761) loss 6.1776 (6.5539) grad_norm 2.6291 (1.8355) [2021-04-14 17:13:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][1230/1251] eta 0:00:05 lr 0.000100 time 0.2619 (0.2761) loss 6.2897 (6.5527) grad_norm 2.0976 (1.8386) [2021-04-14 17:13:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][1240/1251] eta 0:00:03 lr 0.000100 time 0.2476 (0.2759) loss 6.4323 (6.5521) grad_norm 2.4377 (1.8421) [2021-04-14 17:13:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [1/300][1250/1251] eta 0:00:00 lr 0.000101 time 0.2579 (0.2758) loss 6.2539 (6.5510) grad_norm 1.9677 (1.8431) [2021-04-14 17:13:12 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 1 training takes 0:05:46 [2021-04-14 17:13:12 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_1.pth saving...... [2021-04-14 17:13:29 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_1.pth saved !!! [2021-04-14 17:13:31 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.273 (1.273) Loss 5.6614 (5.6614) Acc@1 7.422 (7.422) Acc@5 16.309 (16.309) [2021-04-14 17:13:33 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.102 (0.287) Loss 5.6330 (5.6070) Acc@1 4.980 (5.815) Acc@5 14.941 (17.196) [2021-04-14 17:13:34 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.088 (0.219) Loss 5.6710 (5.6105) Acc@1 5.762 (5.952) Acc@5 17.188 (17.188) [2021-04-14 17:13:36 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.148 (0.219) Loss 5.5580 (5.6016) Acc@1 6.250 (6.055) Acc@5 18.848 (17.421) [2021-04-14 17:13:38 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.212) Loss 5.6157 (5.6024) Acc@1 5.176 (6.024) Acc@5 14.258 (17.259) [2021-04-14 17:13:40 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 5.986 Acc@5 17.204 [2021-04-14 17:13:40 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 6.0% [2021-04-14 17:13:40 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 5.99% [2021-04-14 17:13:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][0/1251] eta 1:14:56 lr 0.000101 time 3.5943 (3.5943) loss 6.2886 (6.2886) grad_norm 1.8836 (1.8836) [2021-04-14 17:13:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][10/1251] eta 0:12:14 lr 0.000101 time 0.2940 (0.5917) loss 6.2307 (6.3828) grad_norm 1.8733 (1.8122) [2021-04-14 17:13:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][20/1251] eta 0:09:01 lr 0.000102 time 0.2941 (0.4401) loss 6.1508 (6.4185) grad_norm 1.6973 (1.7900) [2021-04-14 17:13:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][30/1251] eta 0:07:54 lr 0.000102 time 0.2658 (0.3887) loss 6.1543 (6.4311) grad_norm 1.5890 (1.8134) [2021-04-14 17:13:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][40/1251] eta 0:07:15 lr 0.000102 time 0.2711 (0.3595) loss 6.6883 (6.4218) grad_norm 1.5907 (1.8532) [2021-04-14 17:13:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][50/1251] eta 0:06:51 lr 0.000103 time 0.2844 (0.3426) loss 6.7260 (6.4406) grad_norm 2.2242 (1.8806) [2021-04-14 17:14:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][60/1251] eta 0:06:34 lr 0.000103 time 0.2740 (0.3311) loss 6.4354 (6.4608) grad_norm 1.8369 (1.9237) [2021-04-14 17:14:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][70/1251] eta 0:06:21 lr 0.000104 time 0.2874 (0.3233) loss 6.4024 (6.4711) grad_norm 1.9464 (1.9537) [2021-04-14 17:14:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][80/1251] eta 0:06:11 lr 0.000104 time 0.2691 (0.3170) loss 6.4484 (6.4759) grad_norm 2.6136 (1.9664) [2021-04-14 17:14:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][90/1251] eta 0:06:02 lr 0.000104 time 0.2708 (0.3122) loss 6.5475 (6.4650) grad_norm 2.0042 (1.9568) [2021-04-14 17:14:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][100/1251] eta 0:05:54 lr 0.000105 time 0.2924 (0.3084) loss 6.4588 (6.4633) grad_norm 1.8838 (1.9430) [2021-04-14 17:14:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][110/1251] eta 0:05:47 lr 0.000105 time 0.2589 (0.3049) loss 6.0285 (6.4473) grad_norm 2.0863 (1.9379) [2021-04-14 17:14:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][120/1251] eta 0:05:41 lr 0.000106 time 0.2770 (0.3021) loss 6.5528 (6.4465) grad_norm 2.0047 (1.9406) [2021-04-14 17:14:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][130/1251] eta 0:05:36 lr 0.000106 time 0.2718 (0.3001) loss 6.5291 (6.4456) grad_norm 1.7362 (1.9444) [2021-04-14 17:14:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][140/1251] eta 0:05:31 lr 0.000106 time 0.2831 (0.2982) loss 6.1808 (6.4449) grad_norm 1.8039 (1.9587) [2021-04-14 17:14:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][150/1251] eta 0:05:26 lr 0.000107 time 0.2654 (0.2962) loss 5.8338 (6.4370) grad_norm 1.5606 (1.9520) [2021-04-14 17:14:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][160/1251] eta 0:05:21 lr 0.000107 time 0.2932 (0.2948) loss 6.3267 (6.4382) grad_norm 2.6225 (1.9621) [2021-04-14 17:14:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][170/1251] eta 0:05:16 lr 0.000108 time 0.2513 (0.2932) loss 6.5414 (6.4403) grad_norm 1.9691 (1.9672) [2021-04-14 17:14:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][180/1251] eta 0:05:12 lr 0.000108 time 0.2613 (0.2919) loss 6.4052 (6.4403) grad_norm 1.8358 (1.9765) [2021-04-14 17:14:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][190/1251] eta 0:05:08 lr 0.000108 time 0.2670 (0.2908) loss 6.2867 (6.4366) grad_norm 1.7018 (1.9890) [2021-04-14 17:14:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][200/1251] eta 0:05:04 lr 0.000109 time 0.2568 (0.2898) loss 6.5764 (6.4418) grad_norm 1.5012 (1.9895) [2021-04-14 17:14:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][210/1251] eta 0:05:00 lr 0.000109 time 0.2661 (0.2888) loss 6.4076 (6.4430) grad_norm 1.7765 (1.9774) [2021-04-14 17:14:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][220/1251] eta 0:04:57 lr 0.000110 time 0.2860 (0.2881) loss 6.4882 (6.4398) grad_norm 2.1771 (1.9836) [2021-04-14 17:14:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][230/1251] eta 0:04:53 lr 0.000110 time 0.2646 (0.2876) loss 6.5844 (6.4407) grad_norm 1.7023 (1.9855) [2021-04-14 17:14:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][240/1251] eta 0:04:50 lr 0.000110 time 0.2658 (0.2872) loss 6.4865 (6.4333) grad_norm 2.0264 (1.9785) [2021-04-14 17:14:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][250/1251] eta 0:04:46 lr 0.000111 time 0.2663 (0.2865) loss 6.0409 (6.4286) grad_norm 1.5928 (1.9749) [2021-04-14 17:14:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][260/1251] eta 0:04:43 lr 0.000111 time 0.2498 (0.2860) loss 6.5616 (6.4257) grad_norm 1.6481 (1.9661) [2021-04-14 17:14:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][270/1251] eta 0:04:40 lr 0.000112 time 0.2717 (0.2855) loss 6.2166 (6.4268) grad_norm 1.9028 (1.9628) [2021-04-14 17:15:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][280/1251] eta 0:04:36 lr 0.000112 time 0.2904 (0.2852) loss 6.6072 (6.4232) grad_norm 2.2942 (1.9645) [2021-04-14 17:15:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][290/1251] eta 0:04:33 lr 0.000112 time 0.2521 (0.2848) loss 5.9588 (6.4160) grad_norm 1.4173 (1.9612) [2021-04-14 17:15:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][300/1251] eta 0:04:30 lr 0.000113 time 0.2701 (0.2844) loss 6.4821 (6.4162) grad_norm 2.0709 (1.9606) [2021-04-14 17:15:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][310/1251] eta 0:04:27 lr 0.000113 time 0.2886 (0.2840) loss 6.4518 (6.4154) grad_norm 2.3949 (1.9703) [2021-04-14 17:15:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][320/1251] eta 0:04:23 lr 0.000114 time 0.2864 (0.2836) loss 6.3400 (6.4132) grad_norm 1.6907 (1.9696) [2021-04-14 17:15:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][330/1251] eta 0:04:20 lr 0.000114 time 0.2570 (0.2832) loss 6.5146 (6.4111) grad_norm 2.3241 (1.9725) [2021-04-14 17:15:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][340/1251] eta 0:04:17 lr 0.000114 time 0.2814 (0.2831) loss 6.2199 (6.4094) grad_norm 2.0494 (1.9744) [2021-04-14 17:15:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][350/1251] eta 0:04:14 lr 0.000115 time 0.2457 (0.2828) loss 6.3510 (6.4084) grad_norm 1.9608 (1.9709) [2021-04-14 17:15:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][360/1251] eta 0:04:11 lr 0.000115 time 0.2745 (0.2825) loss 6.0413 (6.4011) grad_norm 2.3628 (1.9752) [2021-04-14 17:15:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][370/1251] eta 0:04:08 lr 0.000116 time 0.2763 (0.2823) loss 6.5394 (6.4003) grad_norm 2.0080 (1.9769) [2021-04-14 17:15:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][380/1251] eta 0:04:05 lr 0.000116 time 0.2893 (0.2820) loss 6.1247 (6.3993) grad_norm 2.0373 (1.9788) [2021-04-14 17:15:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][390/1251] eta 0:04:02 lr 0.000116 time 0.2620 (0.2817) loss 6.5104 (6.3970) grad_norm 1.6440 (1.9760) [2021-04-14 17:15:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][400/1251] eta 0:03:59 lr 0.000117 time 0.2912 (0.2815) loss 6.1337 (6.3963) grad_norm 2.9266 (1.9793) [2021-04-14 17:15:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][410/1251] eta 0:03:56 lr 0.000117 time 0.2524 (0.2814) loss 6.4974 (6.3930) grad_norm 2.1288 (1.9872) [2021-04-14 17:15:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][420/1251] eta 0:03:54 lr 0.000118 time 0.2694 (0.2818) loss 6.2888 (6.3921) grad_norm 2.5080 (1.9952) [2021-04-14 17:15:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][430/1251] eta 0:03:51 lr 0.000118 time 0.2489 (0.2814) loss 6.1857 (6.3922) grad_norm 2.1133 (1.9959) [2021-04-14 17:15:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][440/1251] eta 0:03:48 lr 0.000118 time 0.2999 (0.2812) loss 5.9507 (6.3919) grad_norm 2.1040 (1.9982) [2021-04-14 17:15:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][450/1251] eta 0:03:45 lr 0.000119 time 0.2934 (0.2810) loss 6.4910 (6.3884) grad_norm 1.8176 (1.9995) [2021-04-14 17:15:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][460/1251] eta 0:03:42 lr 0.000119 time 0.2921 (0.2808) loss 6.0800 (6.3857) grad_norm 1.9951 (2.0034) [2021-04-14 17:15:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][470/1251] eta 0:03:39 lr 0.000120 time 0.2674 (0.2807) loss 6.0942 (6.3862) grad_norm 1.8402 (2.0085) [2021-04-14 17:15:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][480/1251] eta 0:03:36 lr 0.000120 time 0.2768 (0.2805) loss 6.0763 (6.3843) grad_norm 2.1984 (2.0053) [2021-04-14 17:15:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][490/1251] eta 0:03:33 lr 0.000120 time 0.2615 (0.2802) loss 6.4825 (6.3851) grad_norm 1.6898 (2.0048) [2021-04-14 17:16:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][500/1251] eta 0:03:30 lr 0.000121 time 0.2792 (0.2801) loss 6.5533 (6.3840) grad_norm 1.8869 (2.0077) [2021-04-14 17:16:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][510/1251] eta 0:03:27 lr 0.000121 time 0.2814 (0.2799) loss 6.4008 (6.3819) grad_norm 2.3543 (2.0090) [2021-04-14 17:16:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][520/1251] eta 0:03:24 lr 0.000122 time 0.2821 (0.2798) loss 6.2198 (6.3808) grad_norm 2.0268 (2.0128) [2021-04-14 17:16:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][530/1251] eta 0:03:21 lr 0.000122 time 0.2670 (0.2796) loss 6.5227 (6.3790) grad_norm 1.9331 (2.0126) [2021-04-14 17:16:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][540/1251] eta 0:03:18 lr 0.000122 time 0.2808 (0.2795) loss 6.4989 (6.3777) grad_norm 2.5881 (2.0146) [2021-04-14 17:16:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][550/1251] eta 0:03:15 lr 0.000123 time 0.2729 (0.2794) loss 6.2252 (6.3776) grad_norm 2.0150 (2.0168) [2021-04-14 17:16:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][560/1251] eta 0:03:13 lr 0.000123 time 0.2679 (0.2793) loss 6.0579 (6.3761) grad_norm 2.2172 (2.0160) [2021-04-14 17:16:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][570/1251] eta 0:03:10 lr 0.000124 time 0.2600 (0.2793) loss 6.4539 (6.3742) grad_norm 1.8569 (2.0125) [2021-04-14 17:16:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][580/1251] eta 0:03:07 lr 0.000124 time 0.2782 (0.2792) loss 5.9761 (6.3716) grad_norm 2.5456 (2.0114) [2021-04-14 17:16:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][590/1251] eta 0:03:04 lr 0.000124 time 0.2968 (0.2792) loss 6.5419 (6.3716) grad_norm 2.0847 (2.0099) [2021-04-14 17:16:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][600/1251] eta 0:03:01 lr 0.000125 time 0.2678 (0.2791) loss 6.1689 (6.3709) grad_norm 1.4865 (2.0051) [2021-04-14 17:16:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][610/1251] eta 0:02:58 lr 0.000125 time 0.2582 (0.2790) loss 5.8456 (6.3697) grad_norm 1.6933 (2.0011) [2021-04-14 17:16:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][620/1251] eta 0:02:56 lr 0.000126 time 0.2923 (0.2790) loss 6.4068 (6.3688) grad_norm 2.0890 (2.0009) [2021-04-14 17:16:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][630/1251] eta 0:02:53 lr 0.000126 time 0.2884 (0.2789) loss 6.4479 (6.3680) grad_norm 2.1837 (2.0019) [2021-04-14 17:16:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][640/1251] eta 0:02:50 lr 0.000126 time 0.2831 (0.2788) loss 6.2136 (6.3651) grad_norm 2.1923 (2.0063) [2021-04-14 17:16:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][650/1251] eta 0:02:47 lr 0.000127 time 0.2565 (0.2786) loss 6.4972 (6.3640) grad_norm 2.4593 (2.0087) [2021-04-14 17:16:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][660/1251] eta 0:02:44 lr 0.000127 time 0.2709 (0.2786) loss 6.4317 (6.3618) grad_norm 2.0402 (2.0073) [2021-04-14 17:16:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][670/1251] eta 0:02:41 lr 0.000128 time 0.2824 (0.2785) loss 6.6378 (6.3619) grad_norm 2.1336 (2.0088) [2021-04-14 17:16:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][680/1251] eta 0:02:38 lr 0.000128 time 0.2758 (0.2784) loss 5.8453 (6.3612) grad_norm 2.1901 (2.0102) [2021-04-14 17:16:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][690/1251] eta 0:02:36 lr 0.000128 time 0.2843 (0.2784) loss 6.4108 (6.3596) grad_norm 2.1799 (2.0106) [2021-04-14 17:16:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][700/1251] eta 0:02:33 lr 0.000129 time 0.2892 (0.2783) loss 6.5271 (6.3601) grad_norm 1.6970 (2.0108) [2021-04-14 17:16:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][710/1251] eta 0:02:30 lr 0.000129 time 0.2580 (0.2782) loss 6.2789 (6.3590) grad_norm 1.9591 (2.0102) [2021-04-14 17:17:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][720/1251] eta 0:02:27 lr 0.000130 time 0.2805 (0.2781) loss 6.3694 (6.3572) grad_norm 2.0164 (2.0102) [2021-04-14 17:17:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][730/1251] eta 0:02:24 lr 0.000130 time 0.2692 (0.2781) loss 6.4773 (6.3578) grad_norm 2.0823 (2.0128) [2021-04-14 17:17:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][740/1251] eta 0:02:22 lr 0.000130 time 0.2661 (0.2780) loss 5.9267 (6.3587) grad_norm 1.7079 (2.0125) [2021-04-14 17:17:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][750/1251] eta 0:02:19 lr 0.000131 time 0.2915 (0.2780) loss 6.1081 (6.3570) grad_norm 2.6715 (2.0153) [2021-04-14 17:17:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][760/1251] eta 0:02:16 lr 0.000131 time 0.2643 (0.2779) loss 6.4972 (6.3558) grad_norm 1.7870 (2.0157) [2021-04-14 17:17:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][770/1251] eta 0:02:13 lr 0.000132 time 0.2657 (0.2779) loss 6.1979 (6.3536) grad_norm 1.7912 (2.0167) [2021-04-14 17:17:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][780/1251] eta 0:02:10 lr 0.000132 time 0.2665 (0.2778) loss 6.5452 (6.3518) grad_norm 1.9252 (2.0157) [2021-04-14 17:17:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][790/1251] eta 0:02:08 lr 0.000132 time 0.2842 (0.2778) loss 6.4266 (6.3514) grad_norm 1.9535 (2.0131) [2021-04-14 17:17:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][800/1251] eta 0:02:05 lr 0.000133 time 0.2656 (0.2777) loss 6.3259 (6.3506) grad_norm 1.8367 (2.0148) [2021-04-14 17:17:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][810/1251] eta 0:02:02 lr 0.000133 time 0.2782 (0.2776) loss 6.3675 (6.3501) grad_norm 2.6665 (2.0175) [2021-04-14 17:17:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][820/1251] eta 0:01:59 lr 0.000134 time 0.2729 (0.2776) loss 5.8437 (6.3490) grad_norm 2.2502 (2.0200) [2021-04-14 17:17:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][830/1251] eta 0:01:56 lr 0.000134 time 0.2440 (0.2775) loss 6.2255 (6.3472) grad_norm 1.8424 (2.0244) [2021-04-14 17:17:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][840/1251] eta 0:01:54 lr 0.000134 time 0.2631 (0.2774) loss 6.3177 (6.3461) grad_norm 1.8496 (2.0244) [2021-04-14 17:17:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][850/1251] eta 0:01:51 lr 0.000135 time 0.2709 (0.2774) loss 6.3185 (6.3456) grad_norm 1.7591 (2.0235) [2021-04-14 17:17:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][860/1251] eta 0:01:48 lr 0.000135 time 0.2476 (0.2773) loss 6.3953 (6.3436) grad_norm 2.0014 (2.0229) [2021-04-14 17:17:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][870/1251] eta 0:01:45 lr 0.000136 time 0.2792 (0.2772) loss 6.0647 (6.3429) grad_norm 1.7541 (2.0211) [2021-04-14 17:17:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][880/1251] eta 0:01:42 lr 0.000136 time 0.2573 (0.2771) loss 6.5984 (6.3429) grad_norm 1.8624 (2.0208) [2021-04-14 17:17:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][890/1251] eta 0:01:40 lr 0.000136 time 0.2609 (0.2771) loss 6.1212 (6.3427) grad_norm 2.2632 (2.0197) [2021-04-14 17:17:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][900/1251] eta 0:01:37 lr 0.000137 time 0.2742 (0.2770) loss 6.4126 (6.3411) grad_norm 2.0518 (2.0193) [2021-04-14 17:17:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][910/1251] eta 0:01:34 lr 0.000137 time 0.2642 (0.2770) loss 6.1044 (6.3407) grad_norm 2.3518 (2.0207) [2021-04-14 17:17:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][920/1251] eta 0:01:31 lr 0.000138 time 0.2541 (0.2770) loss 6.4503 (6.3408) grad_norm 2.1782 (2.0224) [2021-04-14 17:17:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][930/1251] eta 0:01:28 lr 0.000138 time 0.2597 (0.2769) loss 6.4085 (6.3393) grad_norm 2.3023 (2.0242) [2021-04-14 17:18:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][940/1251] eta 0:01:26 lr 0.000138 time 0.2950 (0.2769) loss 6.3729 (6.3378) grad_norm 2.1343 (2.0274) [2021-04-14 17:18:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][950/1251] eta 0:01:23 lr 0.000139 time 0.2664 (0.2769) loss 6.3872 (6.3368) grad_norm 2.0240 (2.0307) [2021-04-14 17:18:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][960/1251] eta 0:01:20 lr 0.000139 time 0.2843 (0.2768) loss 6.4354 (6.3357) grad_norm 2.5703 (2.0328) [2021-04-14 17:18:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][970/1251] eta 0:01:17 lr 0.000140 time 0.2746 (0.2768) loss 6.1960 (6.3331) grad_norm 2.2660 (2.0329) [2021-04-14 17:18:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][980/1251] eta 0:01:14 lr 0.000140 time 0.2617 (0.2767) loss 5.8213 (6.3322) grad_norm 2.1687 (2.0321) [2021-04-14 17:18:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][990/1251] eta 0:01:12 lr 0.000140 time 0.2965 (0.2767) loss 6.5635 (6.3321) grad_norm 2.2016 (2.0328) [2021-04-14 17:18:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][1000/1251] eta 0:01:09 lr 0.000141 time 0.2710 (0.2768) loss 6.5163 (6.3310) grad_norm 1.9011 (2.0328) [2021-04-14 17:18:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][1010/1251] eta 0:01:06 lr 0.000141 time 0.2790 (0.2768) loss 6.3801 (6.3296) grad_norm 1.9496 (2.0344) [2021-04-14 17:18:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][1020/1251] eta 0:01:03 lr 0.000142 time 0.2837 (0.2768) loss 6.5036 (6.3277) grad_norm 2.1887 (2.0354) [2021-04-14 17:18:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][1030/1251] eta 0:01:01 lr 0.000142 time 0.2893 (0.2767) loss 6.6221 (6.3271) grad_norm 2.0922 (2.0367) [2021-04-14 17:18:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][1040/1251] eta 0:00:58 lr 0.000142 time 0.2706 (0.2767) loss 5.7443 (6.3258) grad_norm 2.7370 (inf) [2021-04-14 17:18:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][1050/1251] eta 0:00:55 lr 0.000143 time 0.2527 (0.2766) loss 6.3839 (6.3243) grad_norm 2.1274 (inf) [2021-04-14 17:18:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][1060/1251] eta 0:00:52 lr 0.000143 time 0.2879 (0.2766) loss 6.3804 (6.3239) grad_norm 1.6514 (inf) [2021-04-14 17:18:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][1070/1251] eta 0:00:50 lr 0.000144 time 0.2706 (0.2765) loss 6.4855 (6.3233) grad_norm 2.0132 (inf) [2021-04-14 17:18:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][1080/1251] eta 0:00:47 lr 0.000144 time 0.2556 (0.2765) loss 6.0934 (6.3225) grad_norm 1.9082 (inf) [2021-04-14 17:18:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][1090/1251] eta 0:00:44 lr 0.000144 time 0.2913 (0.2765) loss 5.7816 (6.3221) grad_norm 2.5421 (inf) [2021-04-14 17:18:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][1100/1251] eta 0:00:41 lr 0.000145 time 0.2668 (0.2765) loss 5.5635 (6.3202) grad_norm 2.9369 (inf) [2021-04-14 17:18:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][1110/1251] eta 0:00:38 lr 0.000145 time 0.2958 (0.2765) loss 6.2195 (6.3206) grad_norm 2.0614 (inf) [2021-04-14 17:18:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][1120/1251] eta 0:00:36 lr 0.000146 time 0.2511 (0.2764) loss 6.0240 (6.3202) grad_norm 2.3820 (inf) [2021-04-14 17:18:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][1130/1251] eta 0:00:33 lr 0.000146 time 0.2724 (0.2764) loss 6.2307 (6.3200) grad_norm 1.6535 (inf) [2021-04-14 17:18:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][1140/1251] eta 0:00:30 lr 0.000146 time 0.2675 (0.2764) loss 6.4467 (6.3187) grad_norm 2.2821 (inf) [2021-04-14 17:18:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][1150/1251] eta 0:00:27 lr 0.000147 time 0.2636 (0.2763) loss 6.2621 (6.3183) grad_norm 1.5750 (inf) [2021-04-14 17:19:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][1160/1251] eta 0:00:25 lr 0.000147 time 0.2834 (0.2763) loss 6.0484 (6.3172) grad_norm 1.9869 (inf) [2021-04-14 17:19:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][1170/1251] eta 0:00:22 lr 0.000148 time 0.2784 (0.2762) loss 6.1630 (6.3164) grad_norm 2.5389 (inf) [2021-04-14 17:19:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][1180/1251] eta 0:00:19 lr 0.000148 time 0.2793 (0.2762) loss 6.4394 (6.3153) grad_norm 1.9923 (inf) [2021-04-14 17:19:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][1190/1251] eta 0:00:16 lr 0.000148 time 0.2567 (0.2762) loss 5.8929 (6.3141) grad_norm 2.6789 (inf) [2021-04-14 17:19:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][1200/1251] eta 0:00:14 lr 0.000149 time 0.2981 (0.2761) loss 6.2937 (6.3134) grad_norm 2.7906 (inf) [2021-04-14 17:19:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][1210/1251] eta 0:00:11 lr 0.000149 time 0.2877 (0.2761) loss 6.5363 (6.3126) grad_norm 1.8759 (inf) [2021-04-14 17:19:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][1220/1251] eta 0:00:08 lr 0.000150 time 0.2997 (0.2761) loss 6.4577 (6.3121) grad_norm 1.9566 (inf) [2021-04-14 17:19:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][1230/1251] eta 0:00:05 lr 0.000150 time 0.2747 (0.2761) loss 5.9902 (6.3115) grad_norm 1.8112 (inf) [2021-04-14 17:19:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][1240/1251] eta 0:00:03 lr 0.000150 time 0.2506 (0.2761) loss 6.1222 (6.3108) grad_norm 1.9575 (inf) [2021-04-14 17:19:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [2/300][1250/1251] eta 0:00:00 lr 0.000151 time 0.2591 (0.2759) loss 6.4008 (6.3094) grad_norm 1.9247 (inf) [2021-04-14 17:19:26 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 2 training takes 0:05:46 [2021-04-14 17:19:26 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_2.pth saving...... [2021-04-14 17:19:48 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_2.pth saved !!! [2021-04-14 17:19:49 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.363 (1.363) Loss 5.0082 (5.0082) Acc@1 11.328 (11.328) Acc@5 28.027 (28.027) [2021-04-14 17:19:51 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.108 (0.257) Loss 4.9213 (4.9529) Acc@1 12.988 (11.985) Acc@5 29.492 (28.169) [2021-04-14 17:19:54 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.081 (0.261) Loss 4.9519 (4.9420) Acc@1 12.109 (12.142) Acc@5 29.590 (28.544) [2021-04-14 17:19:55 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.074 (0.222) Loss 4.9054 (4.9453) Acc@1 11.621 (11.914) Acc@5 29.395 (28.399) [2021-04-14 17:19:57 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.264 (0.220) Loss 4.9095 (4.9384) Acc@1 14.258 (11.900) Acc@5 31.055 (28.647) [2021-04-14 17:19:59 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 11.882 Acc@5 28.718 [2021-04-14 17:19:59 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 11.9% [2021-04-14 17:19:59 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 11.88% [2021-04-14 17:20:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][0/1251] eta 1:21:03 lr 0.000151 time 3.8876 (3.8876) loss 6.1097 (6.1097) grad_norm 1.9171 (1.9171) [2021-04-14 17:20:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][10/1251] eta 0:12:27 lr 0.000151 time 0.2625 (0.6027) loss 6.4758 (6.2304) grad_norm 2.6075 (2.2371) [2021-04-14 17:20:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][20/1251] eta 0:09:09 lr 0.000152 time 0.2735 (0.4467) loss 5.8320 (6.1704) grad_norm 2.6808 (2.2538) [2021-04-14 17:20:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][30/1251] eta 0:07:57 lr 0.000152 time 0.2647 (0.3915) loss 6.0729 (6.1641) grad_norm 2.1884 (2.2565) [2021-04-14 17:20:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][40/1251] eta 0:07:20 lr 0.000152 time 0.2846 (0.3641) loss 5.9351 (6.1719) grad_norm 1.6667 (2.2500) [2021-04-14 17:20:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][50/1251] eta 0:06:55 lr 0.000153 time 0.2774 (0.3460) loss 5.9914 (6.1782) grad_norm 2.1510 (2.2041) [2021-04-14 17:20:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][60/1251] eta 0:06:38 lr 0.000153 time 0.2733 (0.3344) loss 5.6465 (6.1506) grad_norm 2.2056 (2.2474) [2021-04-14 17:20:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][70/1251] eta 0:06:24 lr 0.000154 time 0.2522 (0.3257) loss 6.3125 (6.1762) grad_norm 2.1113 (2.2225) [2021-04-14 17:20:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][80/1251] eta 0:06:14 lr 0.000154 time 0.2691 (0.3195) loss 6.0420 (6.1837) grad_norm 2.3376 (2.2232) [2021-04-14 17:20:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][90/1251] eta 0:06:04 lr 0.000154 time 0.2545 (0.3143) loss 5.8594 (6.1669) grad_norm 1.6782 (2.2223) [2021-04-14 17:20:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][100/1251] eta 0:05:56 lr 0.000155 time 0.2490 (0.3099) loss 6.0656 (6.1802) grad_norm 2.0256 (2.1988) [2021-04-14 17:20:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][110/1251] eta 0:05:50 lr 0.000155 time 0.2951 (0.3068) loss 6.0648 (6.1676) grad_norm 2.5786 (2.2451) [2021-04-14 17:20:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][120/1251] eta 0:05:44 lr 0.000156 time 0.3047 (0.3042) loss 6.4975 (6.1688) grad_norm 2.4411 (2.2584) [2021-04-14 17:20:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][130/1251] eta 0:05:38 lr 0.000156 time 0.2639 (0.3016) loss 6.5346 (6.1563) grad_norm 2.2265 (2.2508) [2021-04-14 17:20:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][140/1251] eta 0:05:32 lr 0.000156 time 0.2822 (0.2994) loss 6.4008 (6.1529) grad_norm 1.8475 (2.2451) [2021-04-14 17:20:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][150/1251] eta 0:05:27 lr 0.000157 time 0.2814 (0.2977) loss 6.2852 (6.1482) grad_norm 2.8663 (2.2582) [2021-04-14 17:20:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][160/1251] eta 0:05:23 lr 0.000157 time 0.2750 (0.2961) loss 5.9724 (6.1479) grad_norm 2.0603 (2.2726) [2021-04-14 17:20:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][170/1251] eta 0:05:18 lr 0.000158 time 0.2671 (0.2947) loss 6.2999 (6.1440) grad_norm 2.5821 (2.2737) [2021-04-14 17:20:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][180/1251] eta 0:05:14 lr 0.000158 time 0.2644 (0.2935) loss 6.4031 (6.1457) grad_norm 3.1547 (2.2776) [2021-04-14 17:20:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][190/1251] eta 0:05:10 lr 0.000158 time 0.2459 (0.2924) loss 6.1737 (6.1402) grad_norm 2.4483 (2.2912) [2021-04-14 17:20:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][200/1251] eta 0:05:06 lr 0.000159 time 0.2555 (0.2914) loss 6.4675 (6.1361) grad_norm 1.6736 (2.2843) [2021-04-14 17:21:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][210/1251] eta 0:05:03 lr 0.000159 time 0.2757 (0.2915) loss 6.3189 (6.1366) grad_norm 2.0638 (2.2758) [2021-04-14 17:21:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][220/1251] eta 0:05:00 lr 0.000160 time 0.2752 (0.2914) loss 5.8640 (6.1427) grad_norm 1.7789 (2.2691) [2021-04-14 17:21:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][230/1251] eta 0:04:56 lr 0.000160 time 0.2822 (0.2905) loss 5.9250 (6.1434) grad_norm 2.4954 (2.2746) [2021-04-14 17:21:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][240/1251] eta 0:04:52 lr 0.000160 time 0.2716 (0.2897) loss 6.2286 (6.1436) grad_norm 2.5785 (2.2851) [2021-04-14 17:21:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][250/1251] eta 0:04:49 lr 0.000161 time 0.2593 (0.2889) loss 6.4323 (6.1418) grad_norm 2.6913 (2.2878) [2021-04-14 17:21:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][260/1251] eta 0:04:45 lr 0.000161 time 0.2549 (0.2882) loss 6.3473 (6.1385) grad_norm 2.6543 (2.2929) [2021-04-14 17:21:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][270/1251] eta 0:04:42 lr 0.000162 time 0.2975 (0.2878) loss 6.0764 (6.1415) grad_norm 2.1214 (2.2856) [2021-04-14 17:21:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][280/1251] eta 0:04:38 lr 0.000162 time 0.2490 (0.2872) loss 5.9295 (6.1382) grad_norm 2.4280 (2.2839) [2021-04-14 17:21:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][290/1251] eta 0:04:35 lr 0.000162 time 0.2734 (0.2866) loss 6.4035 (6.1456) grad_norm 2.9718 (2.2863) [2021-04-14 17:21:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][300/1251] eta 0:04:32 lr 0.000163 time 0.2656 (0.2861) loss 5.7934 (6.1497) grad_norm 2.9727 (2.2973) [2021-04-14 17:21:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][310/1251] eta 0:04:28 lr 0.000163 time 0.2958 (0.2857) loss 5.8242 (6.1509) grad_norm 2.4569 (2.2906) [2021-04-14 17:21:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][320/1251] eta 0:04:25 lr 0.000164 time 0.2821 (0.2851) loss 5.8717 (6.1428) grad_norm 2.3452 (2.2915) [2021-04-14 17:21:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][330/1251] eta 0:04:22 lr 0.000164 time 0.2702 (0.2847) loss 5.8814 (6.1375) grad_norm 1.9604 (2.2981) [2021-04-14 17:21:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][340/1251] eta 0:04:18 lr 0.000164 time 0.2632 (0.2841) loss 6.3244 (6.1380) grad_norm 1.9475 (2.2965) [2021-04-14 17:21:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][350/1251] eta 0:04:15 lr 0.000165 time 0.2480 (0.2838) loss 6.2557 (6.1331) grad_norm 2.6760 (2.2934) [2021-04-14 17:21:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][360/1251] eta 0:04:12 lr 0.000165 time 0.2579 (0.2835) loss 5.9780 (6.1352) grad_norm 2.0211 (2.2948) [2021-04-14 17:21:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][370/1251] eta 0:04:09 lr 0.000166 time 0.2592 (0.2833) loss 6.3478 (6.1347) grad_norm 2.1855 (2.2907) [2021-04-14 17:21:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][380/1251] eta 0:04:06 lr 0.000166 time 0.2562 (0.2829) loss 5.3958 (6.1334) grad_norm 2.3361 (2.2884) [2021-04-14 17:21:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][390/1251] eta 0:04:03 lr 0.000166 time 0.2684 (0.2827) loss 5.5348 (6.1342) grad_norm 2.1622 (2.2884) [2021-04-14 17:21:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][400/1251] eta 0:04:00 lr 0.000167 time 0.2893 (0.2823) loss 5.4790 (6.1306) grad_norm 2.0006 (2.2918) [2021-04-14 17:21:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][410/1251] eta 0:03:57 lr 0.000167 time 0.2702 (0.2820) loss 6.5336 (6.1326) grad_norm 2.5405 (2.2868) [2021-04-14 17:21:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][420/1251] eta 0:03:54 lr 0.000168 time 0.2668 (0.2817) loss 6.4260 (6.1341) grad_norm 1.9059 (2.2860) [2021-04-14 17:22:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][430/1251] eta 0:03:51 lr 0.000168 time 0.2577 (0.2816) loss 6.2865 (6.1341) grad_norm 1.9548 (2.2841) [2021-04-14 17:22:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][440/1251] eta 0:03:48 lr 0.000168 time 0.2592 (0.2815) loss 5.6942 (6.1289) grad_norm 1.8743 (2.2839) [2021-04-14 17:22:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][450/1251] eta 0:03:45 lr 0.000169 time 0.2594 (0.2814) loss 6.0064 (6.1288) grad_norm 2.5711 (2.2823) [2021-04-14 17:22:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][460/1251] eta 0:03:42 lr 0.000169 time 0.2770 (0.2812) loss 6.2471 (6.1288) grad_norm 2.1112 (2.2796) [2021-04-14 17:22:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][470/1251] eta 0:03:39 lr 0.000170 time 0.2887 (0.2810) loss 5.6170 (6.1295) grad_norm 1.9894 (2.2814) [2021-04-14 17:22:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][480/1251] eta 0:03:36 lr 0.000170 time 0.2889 (0.2808) loss 6.2138 (6.1266) grad_norm 1.9647 (2.2833) [2021-04-14 17:22:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][490/1251] eta 0:03:33 lr 0.000170 time 0.2614 (0.2806) loss 6.3819 (6.1272) grad_norm 1.8493 (2.2813) [2021-04-14 17:22:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][500/1251] eta 0:03:30 lr 0.000171 time 0.2637 (0.2804) loss 6.2837 (6.1267) grad_norm 2.4763 (2.2798) [2021-04-14 17:22:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][510/1251] eta 0:03:27 lr 0.000171 time 0.2726 (0.2802) loss 5.8612 (6.1257) grad_norm 2.0217 (2.2746) [2021-04-14 17:22:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][520/1251] eta 0:03:24 lr 0.000172 time 0.2521 (0.2800) loss 6.3344 (6.1249) grad_norm 1.8935 (2.2679) [2021-04-14 17:22:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][530/1251] eta 0:03:21 lr 0.000172 time 0.2604 (0.2799) loss 5.5541 (6.1211) grad_norm 2.0909 (2.2648) [2021-04-14 17:22:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][540/1251] eta 0:03:18 lr 0.000172 time 0.2621 (0.2796) loss 6.7218 (6.1219) grad_norm 2.4726 (2.2737) [2021-04-14 17:22:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][550/1251] eta 0:03:15 lr 0.000173 time 0.2606 (0.2794) loss 6.2751 (6.1177) grad_norm 2.3645 (2.2753) [2021-04-14 17:22:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][560/1251] eta 0:03:12 lr 0.000173 time 0.2700 (0.2792) loss 5.2079 (6.1179) grad_norm 2.2625 (2.2755) [2021-04-14 17:22:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][570/1251] eta 0:03:10 lr 0.000174 time 0.2770 (0.2791) loss 6.2307 (6.1189) grad_norm 2.8582 (2.2801) [2021-04-14 17:22:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][580/1251] eta 0:03:07 lr 0.000174 time 0.2832 (0.2790) loss 5.6104 (6.1174) grad_norm 2.4131 (2.2804) [2021-04-14 17:22:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][590/1251] eta 0:03:04 lr 0.000174 time 0.2751 (0.2789) loss 5.6310 (6.1147) grad_norm 2.1013 (2.2775) [2021-04-14 17:22:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][600/1251] eta 0:03:01 lr 0.000175 time 0.2757 (0.2787) loss 6.0237 (6.1155) grad_norm 2.5620 (2.2808) [2021-04-14 17:22:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][610/1251] eta 0:02:58 lr 0.000175 time 0.2602 (0.2787) loss 5.3341 (6.1144) grad_norm 1.9108 (2.2821) [2021-04-14 17:22:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][620/1251] eta 0:02:55 lr 0.000176 time 0.2866 (0.2787) loss 6.0371 (6.1161) grad_norm 2.0393 (2.2808) [2021-04-14 17:22:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][630/1251] eta 0:02:52 lr 0.000176 time 0.2879 (0.2786) loss 5.9376 (6.1144) grad_norm 2.1210 (2.2814) [2021-04-14 17:22:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][640/1251] eta 0:02:50 lr 0.000176 time 0.2711 (0.2784) loss 5.9671 (6.1144) grad_norm 2.8859 (2.2828) [2021-04-14 17:23:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][650/1251] eta 0:02:47 lr 0.000177 time 0.2679 (0.2783) loss 6.3078 (6.1145) grad_norm 1.9993 (2.2820) [2021-04-14 17:23:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][660/1251] eta 0:02:44 lr 0.000177 time 0.2708 (0.2783) loss 6.1392 (6.1140) grad_norm 2.4412 (2.2821) [2021-04-14 17:23:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][670/1251] eta 0:02:41 lr 0.000178 time 0.2549 (0.2782) loss 5.6778 (6.1132) grad_norm 2.1318 (2.2815) [2021-04-14 17:23:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][680/1251] eta 0:02:38 lr 0.000178 time 0.2663 (0.2781) loss 6.4011 (6.1116) grad_norm 1.6798 (2.2812) [2021-04-14 17:23:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][690/1251] eta 0:02:35 lr 0.000178 time 0.2746 (0.2780) loss 6.3894 (6.1127) grad_norm 1.8735 (2.2820) [2021-04-14 17:23:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][700/1251] eta 0:02:33 lr 0.000179 time 0.2638 (0.2779) loss 5.8180 (6.1116) grad_norm 1.9673 (2.2807) [2021-04-14 17:23:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][710/1251] eta 0:02:30 lr 0.000179 time 0.2664 (0.2778) loss 6.3277 (6.1094) grad_norm 2.1041 (2.2792) [2021-04-14 17:23:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][720/1251] eta 0:02:27 lr 0.000180 time 0.2945 (0.2778) loss 6.3609 (6.1127) grad_norm 2.3534 (2.2781) [2021-04-14 17:23:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][730/1251] eta 0:02:24 lr 0.000180 time 0.2798 (0.2777) loss 6.1162 (6.1127) grad_norm 1.7933 (2.2762) [2021-04-14 17:23:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][740/1251] eta 0:02:21 lr 0.000180 time 0.2676 (0.2777) loss 6.3055 (6.1104) grad_norm 2.4132 (2.2777) [2021-04-14 17:23:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][750/1251] eta 0:02:19 lr 0.000181 time 0.2637 (0.2775) loss 5.4439 (6.1092) grad_norm 2.1641 (2.2775) [2021-04-14 17:23:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][760/1251] eta 0:02:16 lr 0.000181 time 0.2917 (0.2775) loss 5.7159 (6.1101) grad_norm 2.1454 (2.2755) [2021-04-14 17:23:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][770/1251] eta 0:02:13 lr 0.000182 time 0.2767 (0.2775) loss 6.0659 (6.1104) grad_norm 1.7712 (2.2757) [2021-04-14 17:23:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][780/1251] eta 0:02:10 lr 0.000182 time 0.2491 (0.2776) loss 6.1300 (6.1105) grad_norm 3.0701 (2.2763) [2021-04-14 17:23:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][790/1251] eta 0:02:08 lr 0.000182 time 0.2424 (0.2777) loss 6.1423 (6.1100) grad_norm 1.9924 (2.2751) [2021-04-14 17:23:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][800/1251] eta 0:02:05 lr 0.000183 time 0.2916 (0.2778) loss 6.2163 (6.1097) grad_norm 1.8829 (2.2723) [2021-04-14 17:23:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][810/1251] eta 0:02:02 lr 0.000183 time 0.2744 (0.2777) loss 6.5093 (6.1103) grad_norm 2.1507 (2.2700) [2021-04-14 17:23:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][820/1251] eta 0:01:59 lr 0.000184 time 0.2843 (0.2776) loss 6.1625 (6.1093) grad_norm 2.1963 (2.2707) [2021-04-14 17:23:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][830/1251] eta 0:01:56 lr 0.000184 time 0.2672 (0.2775) loss 5.8937 (6.1093) grad_norm 2.5552 (2.2715) [2021-04-14 17:23:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][840/1251] eta 0:01:54 lr 0.000184 time 0.2465 (0.2774) loss 6.0398 (6.1082) grad_norm 2.1484 (2.2770) [2021-04-14 17:23:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][850/1251] eta 0:01:51 lr 0.000185 time 0.2941 (0.2774) loss 6.1583 (6.1073) grad_norm 1.8007 (2.2775) [2021-04-14 17:23:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][860/1251] eta 0:01:48 lr 0.000185 time 0.2752 (0.2773) loss 6.4134 (6.1050) grad_norm 2.8393 (2.2780) [2021-04-14 17:24:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][870/1251] eta 0:01:45 lr 0.000186 time 0.2657 (0.2773) loss 6.3640 (6.1051) grad_norm 2.4378 (2.2779) [2021-04-14 17:24:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][880/1251] eta 0:01:42 lr 0.000186 time 0.2655 (0.2774) loss 6.2459 (6.1039) grad_norm 2.1022 (2.2764) [2021-04-14 17:24:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][890/1251] eta 0:01:40 lr 0.000186 time 0.2906 (0.2773) loss 6.1408 (6.1039) grad_norm 1.8207 (2.2734) [2021-04-14 17:24:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][900/1251] eta 0:01:37 lr 0.000187 time 0.2531 (0.2772) loss 6.2335 (6.1036) grad_norm 1.9626 (2.2738) [2021-04-14 17:24:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][910/1251] eta 0:01:34 lr 0.000187 time 0.2853 (0.2771) loss 5.3059 (6.1035) grad_norm 2.8930 (2.2739) [2021-04-14 17:24:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][920/1251] eta 0:01:31 lr 0.000188 time 0.2682 (0.2770) loss 6.1816 (6.1033) grad_norm 2.1045 (2.2756) [2021-04-14 17:24:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][930/1251] eta 0:01:28 lr 0.000188 time 0.2921 (0.2770) loss 6.5384 (6.1031) grad_norm 2.8631 (2.2827) [2021-04-14 17:24:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][940/1251] eta 0:01:26 lr 0.000188 time 0.2707 (0.2769) loss 5.9643 (6.1037) grad_norm 2.2681 (2.2860) [2021-04-14 17:24:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][950/1251] eta 0:01:23 lr 0.000189 time 0.2598 (0.2768) loss 6.0600 (6.1043) grad_norm 2.0906 (2.2839) [2021-04-14 17:24:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][960/1251] eta 0:01:20 lr 0.000189 time 0.2731 (0.2768) loss 6.2323 (6.1033) grad_norm 3.3120 (2.2847) [2021-04-14 17:24:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][970/1251] eta 0:01:17 lr 0.000190 time 0.2630 (0.2768) loss 5.6898 (6.1037) grad_norm 2.7560 (2.2853) [2021-04-14 17:24:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][980/1251] eta 0:01:14 lr 0.000190 time 0.2567 (0.2767) loss 6.0284 (6.1022) grad_norm 2.2701 (2.2871) [2021-04-14 17:24:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][990/1251] eta 0:01:12 lr 0.000190 time 0.2612 (0.2767) loss 5.5271 (6.1018) grad_norm 2.3733 (2.2864) [2021-04-14 17:24:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][1000/1251] eta 0:01:09 lr 0.000191 time 0.2664 (0.2766) loss 6.1801 (6.1008) grad_norm 2.3707 (2.2865) [2021-04-14 17:24:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][1010/1251] eta 0:01:06 lr 0.000191 time 0.2602 (0.2765) loss 6.3489 (6.1000) grad_norm 2.1521 (2.2853) [2021-04-14 17:24:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][1020/1251] eta 0:01:03 lr 0.000192 time 0.2591 (0.2765) loss 6.0710 (6.0975) grad_norm 2.9074 (2.2871) [2021-04-14 17:24:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][1030/1251] eta 0:01:01 lr 0.000192 time 0.2700 (0.2765) loss 6.2153 (6.0977) grad_norm 2.0562 (2.2849) [2021-04-14 17:24:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][1040/1251] eta 0:00:58 lr 0.000192 time 0.2872 (0.2765) loss 6.1180 (6.0970) grad_norm 2.2993 (2.2864) [2021-04-14 17:24:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][1050/1251] eta 0:00:55 lr 0.000193 time 0.2823 (0.2768) loss 6.2122 (6.0965) grad_norm 1.8596 (2.2884) [2021-04-14 17:24:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][1060/1251] eta 0:00:52 lr 0.000193 time 0.2636 (0.2767) loss 6.1690 (6.0953) grad_norm 2.0675 (2.2900) [2021-04-14 17:24:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][1070/1251] eta 0:00:50 lr 0.000194 time 0.2589 (0.2767) loss 5.6927 (6.0931) grad_norm 2.2014 (2.2892) [2021-04-14 17:24:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][1080/1251] eta 0:00:47 lr 0.000194 time 0.2546 (0.2766) loss 6.2597 (6.0925) grad_norm 2.3502 (2.2893) [2021-04-14 17:25:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][1090/1251] eta 0:00:44 lr 0.000194 time 0.2675 (0.2766) loss 6.2704 (6.0916) grad_norm 2.3964 (2.2888) [2021-04-14 17:25:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][1100/1251] eta 0:00:41 lr 0.000195 time 0.2806 (0.2766) loss 6.3623 (6.0894) grad_norm 2.7392 (2.2898) [2021-04-14 17:25:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][1110/1251] eta 0:00:38 lr 0.000195 time 0.2815 (0.2765) loss 5.9324 (6.0861) grad_norm 1.9612 (2.2908) [2021-04-14 17:25:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][1120/1251] eta 0:00:36 lr 0.000196 time 0.2926 (0.2765) loss 6.4209 (6.0844) grad_norm 1.9832 (2.2917) [2021-04-14 17:25:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][1130/1251] eta 0:00:33 lr 0.000196 time 0.2742 (0.2765) loss 5.2955 (6.0830) grad_norm 3.7305 (2.2954) [2021-04-14 17:25:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][1140/1251] eta 0:00:30 lr 0.000196 time 0.2631 (0.2765) loss 6.0100 (6.0823) grad_norm 2.0562 (2.2965) [2021-04-14 17:25:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][1150/1251] eta 0:00:27 lr 0.000197 time 0.3029 (0.2765) loss 5.7140 (6.0806) grad_norm 2.0213 (2.2946) [2021-04-14 17:25:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][1160/1251] eta 0:00:25 lr 0.000197 time 0.2814 (0.2764) loss 5.7524 (6.0797) grad_norm 2.1305 (2.2934) [2021-04-14 17:25:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][1170/1251] eta 0:00:22 lr 0.000198 time 0.2521 (0.2764) loss 5.7618 (6.0788) grad_norm 2.1769 (2.2915) [2021-04-14 17:25:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][1180/1251] eta 0:00:19 lr 0.000198 time 0.3029 (0.2764) loss 6.5407 (6.0793) grad_norm 1.9560 (2.2935) [2021-04-14 17:25:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][1190/1251] eta 0:00:16 lr 0.000198 time 0.2752 (0.2763) loss 5.9700 (6.0792) grad_norm 2.2739 (2.2937) [2021-04-14 17:25:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][1200/1251] eta 0:00:14 lr 0.000199 time 0.2754 (0.2763) loss 6.2589 (6.0791) grad_norm 2.2753 (2.2940) [2021-04-14 17:25:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][1210/1251] eta 0:00:11 lr 0.000199 time 0.2433 (0.2763) loss 5.7375 (6.0763) grad_norm 2.6799 (2.2946) [2021-04-14 17:25:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][1220/1251] eta 0:00:08 lr 0.000200 time 0.2593 (0.2763) loss 6.1331 (6.0757) grad_norm 2.2322 (2.2925) [2021-04-14 17:25:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][1230/1251] eta 0:00:05 lr 0.000200 time 0.2546 (0.2762) loss 6.2623 (6.0767) grad_norm 2.3096 (2.2973) [2021-04-14 17:25:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][1240/1251] eta 0:00:03 lr 0.000200 time 0.2506 (0.2761) loss 5.6294 (6.0753) grad_norm 2.1575 (2.2998) [2021-04-14 17:25:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [3/300][1250/1251] eta 0:00:00 lr 0.000201 time 0.2481 (0.2759) loss 6.4730 (6.0763) grad_norm 2.4216 (2.3010) [2021-04-14 17:25:46 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 3 training takes 0:05:46 [2021-04-14 17:25:46 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_3.pth saving...... [2021-04-14 17:26:10 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_3.pth saved !!! [2021-04-14 17:26:11 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.277 (1.277) Loss 4.2883 (4.2883) Acc@1 19.922 (19.922) Acc@5 41.895 (41.895) [2021-04-14 17:26:12 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.422 (0.249) Loss 4.3194 (4.3434) Acc@1 19.434 (18.759) Acc@5 41.797 (40.385) [2021-04-14 17:26:14 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.133 (0.221) Loss 4.3263 (4.3446) Acc@1 18.848 (18.829) Acc@5 42.383 (40.137) [2021-04-14 17:26:16 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.111 (0.209) Loss 4.2535 (4.3492) Acc@1 19.727 (18.766) Acc@5 41.992 (39.941) [2021-04-14 17:26:18 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.215) Loss 4.3676 (4.3491) Acc@1 17.383 (18.852) Acc@5 38.672 (39.853) [2021-04-14 17:26:20 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 18.788 Acc@5 39.846 [2021-04-14 17:26:20 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 18.8% [2021-04-14 17:26:20 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 18.79% [2021-04-14 17:26:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][0/1251] eta 1:23:18 lr 0.000201 time 3.9956 (3.9956) loss 6.3340 (6.3340) grad_norm 1.9468 (1.9468) [2021-04-14 17:26:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][10/1251] eta 0:12:49 lr 0.000201 time 0.2562 (0.6204) loss 6.3174 (5.9851) grad_norm 1.8368 (2.2433) [2021-04-14 17:26:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][20/1251] eta 0:09:19 lr 0.000202 time 0.2710 (0.4545) loss 6.2437 (5.9529) grad_norm 1.8282 (2.1950) [2021-04-14 17:26:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][30/1251] eta 0:08:02 lr 0.000202 time 0.2722 (0.3949) loss 5.4500 (5.8851) grad_norm 2.3699 (2.2662) [2021-04-14 17:26:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][40/1251] eta 0:07:22 lr 0.000202 time 0.2846 (0.3656) loss 5.9873 (5.9127) grad_norm 1.7039 (2.3250) [2021-04-14 17:26:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][50/1251] eta 0:06:56 lr 0.000203 time 0.2424 (0.3464) loss 6.0924 (5.9105) grad_norm 2.5043 (2.3706) [2021-04-14 17:26:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][60/1251] eta 0:06:38 lr 0.000203 time 0.2832 (0.3347) loss 5.7531 (5.9323) grad_norm 2.1120 (2.3576) [2021-04-14 17:26:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][70/1251] eta 0:06:25 lr 0.000204 time 0.2609 (0.3264) loss 5.9475 (5.9140) grad_norm 1.8337 (2.3890) [2021-04-14 17:26:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][80/1251] eta 0:06:13 lr 0.000204 time 0.2730 (0.3194) loss 5.4472 (5.9024) grad_norm 1.8811 (2.3662) [2021-04-14 17:26:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][90/1251] eta 0:06:04 lr 0.000204 time 0.2400 (0.3143) loss 6.0083 (5.8996) grad_norm 2.3061 (2.3689) [2021-04-14 17:26:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][100/1251] eta 0:05:57 lr 0.000205 time 0.2819 (0.3104) loss 6.4226 (5.9203) grad_norm 1.7605 (2.3555) [2021-04-14 17:26:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][110/1251] eta 0:05:50 lr 0.000205 time 0.2705 (0.3072) loss 6.0224 (5.9195) grad_norm 2.4335 (2.3630) [2021-04-14 17:26:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][120/1251] eta 0:05:43 lr 0.000206 time 0.2559 (0.3039) loss 6.3607 (5.9127) grad_norm 1.8235 (2.3759) [2021-04-14 17:27:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][130/1251] eta 0:05:37 lr 0.000206 time 0.2863 (0.3014) loss 6.1157 (5.9200) grad_norm 2.6958 (2.3611) [2021-04-14 17:27:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][140/1251] eta 0:05:32 lr 0.000206 time 0.2720 (0.2993) loss 6.0462 (5.9255) grad_norm 1.9929 (2.3466) [2021-04-14 17:27:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][150/1251] eta 0:05:28 lr 0.000207 time 0.2985 (0.2981) loss 5.8764 (5.9176) grad_norm 2.3321 (2.3446) [2021-04-14 17:27:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][160/1251] eta 0:05:23 lr 0.000207 time 0.2655 (0.2964) loss 5.7220 (5.9247) grad_norm 2.0823 (2.3352) [2021-04-14 17:27:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][170/1251] eta 0:05:18 lr 0.000208 time 0.2567 (0.2949) loss 5.7562 (5.9151) grad_norm 1.7998 (2.3323) [2021-04-14 17:27:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][180/1251] eta 0:05:14 lr 0.000208 time 0.2731 (0.2937) loss 5.8847 (5.9177) grad_norm 2.1315 (2.3218) [2021-04-14 17:27:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][190/1251] eta 0:05:10 lr 0.000208 time 0.2859 (0.2928) loss 5.7945 (5.9128) grad_norm 2.4126 (2.3173) [2021-04-14 17:27:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][200/1251] eta 0:05:06 lr 0.000209 time 0.2560 (0.2919) loss 6.3239 (5.9161) grad_norm 2.5327 (2.3280) [2021-04-14 17:27:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][210/1251] eta 0:05:02 lr 0.000209 time 0.2695 (0.2908) loss 6.2187 (5.9202) grad_norm 2.1332 (2.3307) [2021-04-14 17:27:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][220/1251] eta 0:04:58 lr 0.000210 time 0.2537 (0.2899) loss 5.4248 (5.9220) grad_norm 1.7645 (2.3219) [2021-04-14 17:27:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][230/1251] eta 0:04:55 lr 0.000210 time 0.2730 (0.2891) loss 6.1024 (5.9182) grad_norm 2.8935 (inf) [2021-04-14 17:27:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][240/1251] eta 0:04:51 lr 0.000210 time 0.2503 (0.2887) loss 6.1413 (5.9251) grad_norm 2.2754 (inf) [2021-04-14 17:27:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][250/1251] eta 0:04:48 lr 0.000211 time 0.2685 (0.2880) loss 6.3484 (5.9265) grad_norm 3.1437 (inf) [2021-04-14 17:27:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][260/1251] eta 0:04:44 lr 0.000211 time 0.2743 (0.2873) loss 6.2371 (5.9230) grad_norm 3.9059 (inf) [2021-04-14 17:27:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][270/1251] eta 0:04:41 lr 0.000212 time 0.2754 (0.2866) loss 5.7464 (5.9211) grad_norm 2.5078 (inf) [2021-04-14 17:27:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][280/1251] eta 0:04:38 lr 0.000212 time 0.2699 (0.2863) loss 6.2691 (5.9241) grad_norm 2.3270 (inf) [2021-04-14 17:27:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][290/1251] eta 0:04:34 lr 0.000212 time 0.2505 (0.2860) loss 5.2527 (5.9233) grad_norm 1.7587 (inf) [2021-04-14 17:27:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][300/1251] eta 0:04:31 lr 0.000213 time 0.2547 (0.2855) loss 5.8995 (5.9299) grad_norm 1.7961 (inf) [2021-04-14 17:27:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][310/1251] eta 0:04:28 lr 0.000213 time 0.2646 (0.2851) loss 6.0656 (5.9265) grad_norm 2.1060 (inf) [2021-04-14 17:27:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][320/1251] eta 0:04:25 lr 0.000214 time 0.2775 (0.2848) loss 5.8871 (5.9231) grad_norm 4.2801 (inf) [2021-04-14 17:27:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][330/1251] eta 0:04:22 lr 0.000214 time 0.2626 (0.2846) loss 5.5637 (5.9206) grad_norm 3.9180 (inf) [2021-04-14 17:27:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][340/1251] eta 0:04:19 lr 0.000214 time 0.2833 (0.2846) loss 5.9696 (5.9248) grad_norm 1.9842 (inf) [2021-04-14 17:28:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][350/1251] eta 0:04:16 lr 0.000215 time 0.2488 (0.2842) loss 6.4123 (5.9286) grad_norm 2.3038 (inf) [2021-04-14 17:28:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][360/1251] eta 0:04:12 lr 0.000215 time 0.2576 (0.2838) loss 5.8796 (5.9273) grad_norm 1.8764 (inf) [2021-04-14 17:28:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][370/1251] eta 0:04:09 lr 0.000216 time 0.2871 (0.2836) loss 5.8885 (5.9254) grad_norm 2.4976 (inf) [2021-04-14 17:28:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][380/1251] eta 0:04:06 lr 0.000216 time 0.2952 (0.2834) loss 5.8882 (5.9214) grad_norm 1.9208 (inf) [2021-04-14 17:28:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][390/1251] eta 0:04:03 lr 0.000216 time 0.2711 (0.2833) loss 6.0152 (5.9217) grad_norm 2.4637 (inf) [2021-04-14 17:28:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][400/1251] eta 0:04:00 lr 0.000217 time 0.2505 (0.2829) loss 6.2460 (5.9245) grad_norm 2.5838 (inf) [2021-04-14 17:28:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][410/1251] eta 0:03:57 lr 0.000217 time 0.2546 (0.2827) loss 6.0825 (5.9248) grad_norm 2.2368 (inf) [2021-04-14 17:28:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][420/1251] eta 0:03:54 lr 0.000218 time 0.2723 (0.2823) loss 5.7394 (5.9219) grad_norm 2.5558 (inf) [2021-04-14 17:28:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][430/1251] eta 0:03:51 lr 0.000218 time 0.3048 (0.2822) loss 6.0731 (5.9205) grad_norm 1.8061 (inf) [2021-04-14 17:28:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][440/1251] eta 0:03:48 lr 0.000218 time 0.2779 (0.2819) loss 6.4420 (5.9216) grad_norm 1.7894 (inf) [2021-04-14 17:28:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][450/1251] eta 0:03:45 lr 0.000219 time 0.2782 (0.2817) loss 5.7185 (5.9213) grad_norm 2.3646 (inf) [2021-04-14 17:28:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][460/1251] eta 0:03:42 lr 0.000219 time 0.2668 (0.2815) loss 6.0511 (5.9219) grad_norm 2.2837 (inf) [2021-04-14 17:28:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][470/1251] eta 0:03:39 lr 0.000220 time 0.2851 (0.2813) loss 6.2069 (5.9238) grad_norm 2.2362 (inf) [2021-04-14 17:28:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][480/1251] eta 0:03:36 lr 0.000220 time 0.2693 (0.2811) loss 5.4159 (5.9214) grad_norm 2.0622 (inf) [2021-04-14 17:28:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][490/1251] eta 0:03:33 lr 0.000220 time 0.2656 (0.2809) loss 6.2058 (5.9225) grad_norm 2.2024 (inf) [2021-04-14 17:28:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][500/1251] eta 0:03:30 lr 0.000221 time 0.2770 (0.2806) loss 6.2444 (5.9215) grad_norm 2.4147 (inf) [2021-04-14 17:28:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][510/1251] eta 0:03:27 lr 0.000221 time 0.2726 (0.2804) loss 5.7799 (5.9177) grad_norm 2.8036 (inf) [2021-04-14 17:28:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][520/1251] eta 0:03:24 lr 0.000222 time 0.2865 (0.2804) loss 6.0829 (5.9201) grad_norm 2.0038 (inf) [2021-04-14 17:28:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][530/1251] eta 0:03:22 lr 0.000222 time 0.2591 (0.2804) loss 5.1880 (5.9151) grad_norm 2.2252 (inf) [2021-04-14 17:28:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][540/1251] eta 0:03:19 lr 0.000222 time 0.2523 (0.2803) loss 5.1174 (5.9145) grad_norm 2.1871 (inf) [2021-04-14 17:28:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][550/1251] eta 0:03:16 lr 0.000223 time 0.2927 (0.2802) loss 6.1814 (5.9121) grad_norm 2.6018 (inf) [2021-04-14 17:28:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][560/1251] eta 0:03:13 lr 0.000223 time 0.2930 (0.2801) loss 5.5123 (5.9121) grad_norm 2.0929 (inf) [2021-04-14 17:29:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][570/1251] eta 0:03:10 lr 0.000224 time 0.2643 (0.2799) loss 5.1358 (5.9091) grad_norm 2.7815 (inf) [2021-04-14 17:29:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][580/1251] eta 0:03:08 lr 0.000224 time 0.2575 (0.2803) loss 6.1894 (5.9094) grad_norm 1.9894 (inf) [2021-04-14 17:29:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][590/1251] eta 0:03:05 lr 0.000224 time 0.2824 (0.2805) loss 6.3612 (5.9109) grad_norm 2.3620 (inf) [2021-04-14 17:29:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][600/1251] eta 0:03:02 lr 0.000225 time 0.2728 (0.2804) loss 5.5091 (5.9089) grad_norm 2.4472 (inf) [2021-04-14 17:29:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][610/1251] eta 0:02:59 lr 0.000225 time 0.2732 (0.2803) loss 6.1005 (5.9079) grad_norm 2.1387 (inf) [2021-04-14 17:29:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][620/1251] eta 0:02:56 lr 0.000226 time 0.2923 (0.2802) loss 6.2580 (5.9066) grad_norm 2.3122 (inf) [2021-04-14 17:29:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][630/1251] eta 0:02:53 lr 0.000226 time 0.2771 (0.2801) loss 6.0232 (5.9091) grad_norm 1.9700 (inf) [2021-04-14 17:29:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][640/1251] eta 0:02:51 lr 0.000226 time 0.2598 (0.2801) loss 5.1769 (5.9083) grad_norm 2.4141 (inf) [2021-04-14 17:29:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][650/1251] eta 0:02:48 lr 0.000227 time 0.2723 (0.2799) loss 5.8856 (5.9066) grad_norm 2.7954 (inf) [2021-04-14 17:29:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][660/1251] eta 0:02:45 lr 0.000227 time 0.2791 (0.2798) loss 5.7996 (5.9042) grad_norm 2.8318 (inf) [2021-04-14 17:29:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][670/1251] eta 0:02:42 lr 0.000228 time 0.2842 (0.2798) loss 6.1012 (5.9041) grad_norm 2.3016 (inf) [2021-04-14 17:29:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][680/1251] eta 0:02:39 lr 0.000228 time 0.2717 (0.2796) loss 5.9974 (5.9042) grad_norm 2.0400 (inf) [2021-04-14 17:29:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][690/1251] eta 0:02:36 lr 0.000228 time 0.2480 (0.2795) loss 4.6241 (5.9029) grad_norm 2.5465 (inf) [2021-04-14 17:29:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][700/1251] eta 0:02:33 lr 0.000229 time 0.2813 (0.2794) loss 5.7107 (5.9026) grad_norm 2.5176 (inf) [2021-04-14 17:29:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][710/1251] eta 0:02:31 lr 0.000229 time 0.2634 (0.2793) loss 6.3829 (5.9005) grad_norm 2.6182 (inf) [2021-04-14 17:29:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][720/1251] eta 0:02:28 lr 0.000230 time 0.3016 (0.2793) loss 5.4052 (5.8977) grad_norm 1.8359 (inf) [2021-04-14 17:29:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][730/1251] eta 0:02:25 lr 0.000230 time 0.2909 (0.2793) loss 6.2982 (5.8962) grad_norm 2.5060 (inf) [2021-04-14 17:29:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][740/1251] eta 0:02:22 lr 0.000230 time 0.2765 (0.2792) loss 5.8652 (5.8959) grad_norm 3.1808 (inf) [2021-04-14 17:29:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][750/1251] eta 0:02:19 lr 0.000231 time 0.2779 (0.2792) loss 5.6956 (5.8964) grad_norm 2.5286 (inf) [2021-04-14 17:29:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][760/1251] eta 0:02:17 lr 0.000231 time 0.2732 (0.2792) loss 5.2434 (5.8972) grad_norm 3.1293 (inf) [2021-04-14 17:29:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][770/1251] eta 0:02:14 lr 0.000232 time 0.3026 (0.2793) loss 5.3500 (5.8963) grad_norm 2.3493 (inf) [2021-04-14 17:29:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][780/1251] eta 0:02:11 lr 0.000232 time 0.2859 (0.2792) loss 6.0413 (5.8957) grad_norm 1.7260 (inf) [2021-04-14 17:30:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][790/1251] eta 0:02:08 lr 0.000232 time 0.2807 (0.2791) loss 6.4919 (5.8944) grad_norm 2.1573 (inf) [2021-04-14 17:30:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][800/1251] eta 0:02:05 lr 0.000233 time 0.3043 (0.2790) loss 6.2478 (5.8934) grad_norm 2.1105 (inf) [2021-04-14 17:30:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][810/1251] eta 0:02:03 lr 0.000233 time 0.2861 (0.2790) loss 5.5812 (5.8926) grad_norm 2.2710 (inf) [2021-04-14 17:30:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][820/1251] eta 0:02:00 lr 0.000234 time 0.2706 (0.2789) loss 5.7749 (5.8933) grad_norm 2.4433 (inf) [2021-04-14 17:30:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][830/1251] eta 0:01:57 lr 0.000234 time 0.2734 (0.2789) loss 6.0517 (5.8931) grad_norm 2.2436 (inf) [2021-04-14 17:30:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][840/1251] eta 0:01:54 lr 0.000234 time 0.2897 (0.2788) loss 6.2630 (5.8917) grad_norm 2.9035 (inf) [2021-04-14 17:30:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][850/1251] eta 0:01:51 lr 0.000235 time 0.2702 (0.2787) loss 6.1812 (5.8915) grad_norm 3.5936 (inf) [2021-04-14 17:30:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][860/1251] eta 0:01:48 lr 0.000235 time 0.2775 (0.2787) loss 6.3102 (5.8950) grad_norm 2.4095 (inf) [2021-04-14 17:30:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][870/1251] eta 0:01:46 lr 0.000236 time 0.2861 (0.2787) loss 6.0707 (5.8975) grad_norm 1.9529 (inf) [2021-04-14 17:30:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][880/1251] eta 0:01:43 lr 0.000236 time 0.2852 (0.2788) loss 6.2141 (5.8981) grad_norm 4.3144 (inf) [2021-04-14 17:30:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][890/1251] eta 0:01:40 lr 0.000236 time 0.2813 (0.2787) loss 5.5796 (5.8971) grad_norm 1.7896 (inf) [2021-04-14 17:30:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][900/1251] eta 0:01:37 lr 0.000237 time 0.2555 (0.2786) loss 6.1885 (5.8959) grad_norm 2.1157 (inf) [2021-04-14 17:30:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][910/1251] eta 0:01:34 lr 0.000237 time 0.2946 (0.2786) loss 5.2559 (5.8961) grad_norm 2.3420 (inf) [2021-04-14 17:30:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][920/1251] eta 0:01:32 lr 0.000238 time 0.2682 (0.2785) loss 5.8217 (5.8954) grad_norm 2.4740 (inf) [2021-04-14 17:30:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][930/1251] eta 0:01:29 lr 0.000238 time 0.2781 (0.2785) loss 6.4444 (5.8962) grad_norm 2.2262 (inf) [2021-04-14 17:30:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][940/1251] eta 0:01:26 lr 0.000238 time 0.2911 (0.2786) loss 5.4900 (5.8972) grad_norm 3.1606 (inf) [2021-04-14 17:30:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][950/1251] eta 0:01:23 lr 0.000239 time 0.2609 (0.2785) loss 5.5000 (5.8956) grad_norm 1.7972 (inf) [2021-04-14 17:30:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][960/1251] eta 0:01:21 lr 0.000239 time 0.2563 (0.2785) loss 6.2431 (5.8936) grad_norm 2.1237 (inf) [2021-04-14 17:30:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][970/1251] eta 0:01:18 lr 0.000240 time 0.2644 (0.2785) loss 5.6844 (5.8938) grad_norm 1.8414 (inf) [2021-04-14 17:30:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][980/1251] eta 0:01:15 lr 0.000240 time 0.2844 (0.2784) loss 5.7784 (5.8933) grad_norm 1.7649 (inf) [2021-04-14 17:30:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][990/1251] eta 0:01:12 lr 0.000240 time 0.2656 (0.2784) loss 5.6512 (5.8913) grad_norm 1.6867 (inf) [2021-04-14 17:30:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][1000/1251] eta 0:01:09 lr 0.000241 time 0.2912 (0.2783) loss 6.2263 (5.8912) grad_norm 2.1494 (inf) [2021-04-14 17:31:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][1010/1251] eta 0:01:07 lr 0.000241 time 0.2779 (0.2782) loss 5.7766 (5.8914) grad_norm 1.9114 (inf) [2021-04-14 17:31:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][1020/1251] eta 0:01:04 lr 0.000242 time 0.2617 (0.2782) loss 5.3455 (5.8892) grad_norm 2.2785 (inf) [2021-04-14 17:31:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][1030/1251] eta 0:01:01 lr 0.000242 time 0.2495 (0.2781) loss 5.9215 (5.8864) grad_norm 2.4784 (inf) [2021-04-14 17:31:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][1040/1251] eta 0:00:58 lr 0.000242 time 0.2661 (0.2782) loss 6.2440 (5.8866) grad_norm 1.8706 (inf) [2021-04-14 17:31:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][1050/1251] eta 0:00:55 lr 0.000243 time 0.2741 (0.2781) loss 5.6568 (5.8850) grad_norm 2.3703 (inf) [2021-04-14 17:31:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][1060/1251] eta 0:00:53 lr 0.000243 time 0.2643 (0.2780) loss 6.2456 (5.8849) grad_norm 1.9719 (inf) [2021-04-14 17:31:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][1070/1251] eta 0:00:50 lr 0.000244 time 0.2793 (0.2780) loss 6.3981 (5.8855) grad_norm 2.0769 (inf) [2021-04-14 17:31:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][1080/1251] eta 0:00:47 lr 0.000244 time 0.2595 (0.2779) loss 5.3116 (5.8852) grad_norm 3.2895 (inf) [2021-04-14 17:31:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][1090/1251] eta 0:00:44 lr 0.000244 time 0.2562 (0.2778) loss 5.3086 (5.8848) grad_norm 2.5283 (inf) [2021-04-14 17:31:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][1100/1251] eta 0:00:41 lr 0.000245 time 0.2555 (0.2778) loss 5.4967 (5.8832) grad_norm 2.3539 (inf) [2021-04-14 17:31:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][1110/1251] eta 0:00:39 lr 0.000245 time 0.2798 (0.2777) loss 6.1564 (5.8799) grad_norm 2.1174 (inf) [2021-04-14 17:31:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][1120/1251] eta 0:00:36 lr 0.000246 time 0.2675 (0.2776) loss 5.9129 (5.8796) grad_norm 2.4351 (inf) [2021-04-14 17:31:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][1130/1251] eta 0:00:33 lr 0.000246 time 0.2786 (0.2776) loss 5.6789 (5.8800) grad_norm 2.7500 (inf) [2021-04-14 17:31:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][1140/1251] eta 0:00:30 lr 0.000246 time 0.2508 (0.2776) loss 5.6131 (5.8769) grad_norm 2.4525 (inf) [2021-04-14 17:31:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][1150/1251] eta 0:00:28 lr 0.000247 time 0.2767 (0.2776) loss 6.1291 (5.8767) grad_norm 2.1286 (inf) [2021-04-14 17:31:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][1160/1251] eta 0:00:25 lr 0.000247 time 0.2696 (0.2777) loss 5.9833 (5.8761) grad_norm 2.9752 (inf) [2021-04-14 17:31:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][1170/1251] eta 0:00:22 lr 0.000248 time 0.2633 (0.2777) loss 6.1999 (5.8770) grad_norm 1.8487 (inf) [2021-04-14 17:31:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][1180/1251] eta 0:00:19 lr 0.000248 time 0.2720 (0.2777) loss 5.9846 (5.8762) grad_norm 1.8924 (inf) [2021-04-14 17:31:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][1190/1251] eta 0:00:16 lr 0.000248 time 0.2704 (0.2776) loss 5.4319 (5.8761) grad_norm 2.7871 (inf) [2021-04-14 17:31:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][1200/1251] eta 0:00:14 lr 0.000249 time 0.2569 (0.2775) loss 5.5162 (5.8763) grad_norm 2.4179 (inf) [2021-04-14 17:31:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][1210/1251] eta 0:00:11 lr 0.000249 time 0.2842 (0.2775) loss 5.1922 (5.8728) grad_norm 2.3518 (inf) [2021-04-14 17:31:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][1220/1251] eta 0:00:08 lr 0.000250 time 0.2716 (0.2775) loss 5.9811 (5.8728) grad_norm 1.9846 (inf) [2021-04-14 17:32:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][1230/1251] eta 0:00:05 lr 0.000250 time 0.2634 (0.2774) loss 5.8447 (5.8715) grad_norm 2.0398 (inf) [2021-04-14 17:32:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][1240/1251] eta 0:00:03 lr 0.000250 time 0.2524 (0.2773) loss 5.3665 (5.8704) grad_norm 2.2716 (inf) [2021-04-14 17:32:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [4/300][1250/1251] eta 0:00:00 lr 0.000251 time 0.2478 (0.2771) loss 5.2496 (5.8708) grad_norm 2.0214 (inf) [2021-04-14 17:32:09 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 4 training takes 0:05:48 [2021-04-14 17:32:09 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_4.pth saving...... [2021-04-14 17:32:32 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_4.pth saved !!! [2021-04-14 17:32:33 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.392 (1.392) Loss 3.8840 (3.8840) Acc@1 26.758 (26.758) Acc@5 48.633 (48.633) [2021-04-14 17:32:35 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.112 (0.232) Loss 3.9067 (3.9051) Acc@1 24.902 (25.000) Acc@5 47.656 (47.576) [2021-04-14 17:32:37 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.117 (0.218) Loss 3.7959 (3.8990) Acc@1 27.637 (24.981) Acc@5 48.633 (47.619) [2021-04-14 17:32:39 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.451 (0.232) Loss 3.9432 (3.8853) Acc@1 22.754 (24.893) Acc@5 45.605 (47.981) [2021-04-14 17:32:41 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.214) Loss 3.8870 (3.8873) Acc@1 26.953 (24.900) Acc@5 48.828 (48.056) [2021-04-14 17:32:43 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 24.738 Acc@5 47.960 [2021-04-14 17:32:43 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 24.7% [2021-04-14 17:32:43 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 24.74% [2021-04-14 17:32:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][0/1251] eta 1:20:40 lr 0.000251 time 3.8694 (3.8694) loss 5.8796 (5.8796) grad_norm 2.0928 (2.0928) [2021-04-14 17:32:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][10/1251] eta 0:12:26 lr 0.000251 time 0.2540 (0.6018) loss 6.1771 (5.8788) grad_norm 2.6996 (2.5542) [2021-04-14 17:32:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][20/1251] eta 0:09:11 lr 0.000252 time 0.2882 (0.4478) loss 5.5937 (5.8489) grad_norm 2.6527 (2.4089) [2021-04-14 17:32:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][30/1251] eta 0:07:56 lr 0.000252 time 0.2712 (0.3906) loss 5.6329 (5.8357) grad_norm 1.9602 (2.4551) [2021-04-14 17:32:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][40/1251] eta 0:07:18 lr 0.000252 time 0.2942 (0.3625) loss 6.3509 (5.8941) grad_norm 3.0469 (2.4132) [2021-04-14 17:33:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][50/1251] eta 0:06:54 lr 0.000253 time 0.2433 (0.3450) loss 5.9770 (5.9017) grad_norm 2.1343 (2.4236) [2021-04-14 17:33:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][60/1251] eta 0:06:36 lr 0.000253 time 0.2697 (0.3327) loss 5.7172 (5.8935) grad_norm 1.9029 (2.4137) [2021-04-14 17:33:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][70/1251] eta 0:06:22 lr 0.000254 time 0.2636 (0.3241) loss 4.8780 (5.8902) grad_norm 2.0297 (2.3720) [2021-04-14 17:33:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][80/1251] eta 0:06:12 lr 0.000254 time 0.2558 (0.3177) loss 6.0789 (5.8712) grad_norm 2.2344 (2.3984) [2021-04-14 17:33:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][90/1251] eta 0:06:03 lr 0.000254 time 0.2843 (0.3130) loss 5.9801 (5.8651) grad_norm 2.4305 (2.4013) [2021-04-14 17:33:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][100/1251] eta 0:05:55 lr 0.000255 time 0.2450 (0.3088) loss 5.5331 (5.8411) grad_norm 2.8163 (2.4218) [2021-04-14 17:33:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][110/1251] eta 0:05:48 lr 0.000255 time 0.2855 (0.3055) loss 6.2460 (5.8498) grad_norm 2.0805 (2.4086) [2021-04-14 17:33:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][120/1251] eta 0:05:42 lr 0.000256 time 0.2772 (0.3027) loss 6.0510 (5.8516) grad_norm 2.1435 (2.3897) [2021-04-14 17:33:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][130/1251] eta 0:05:36 lr 0.000256 time 0.2429 (0.3001) loss 6.1474 (5.8386) grad_norm 2.3677 (2.3708) [2021-04-14 17:33:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][140/1251] eta 0:05:31 lr 0.000256 time 0.2542 (0.2982) loss 5.5137 (5.8355) grad_norm 1.9548 (2.3600) [2021-04-14 17:33:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][150/1251] eta 0:05:26 lr 0.000257 time 0.2678 (0.2965) loss 6.3939 (5.8347) grad_norm 2.3104 (2.3622) [2021-04-14 17:33:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][160/1251] eta 0:05:21 lr 0.000257 time 0.2735 (0.2949) loss 5.8979 (5.8295) grad_norm 2.1959 (2.3417) [2021-04-14 17:33:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][170/1251] eta 0:05:17 lr 0.000258 time 0.2634 (0.2933) loss 5.8254 (5.8332) grad_norm 2.1399 (2.3302) [2021-04-14 17:33:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][180/1251] eta 0:05:12 lr 0.000258 time 0.2646 (0.2920) loss 5.7637 (5.8171) grad_norm 3.4806 (2.3490) [2021-04-14 17:33:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][190/1251] eta 0:05:08 lr 0.000258 time 0.2802 (0.2908) loss 4.3678 (5.8050) grad_norm 1.9920 (2.3466) [2021-04-14 17:33:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][200/1251] eta 0:05:04 lr 0.000259 time 0.2427 (0.2900) loss 4.8251 (5.7946) grad_norm 1.9055 (2.3317) [2021-04-14 17:33:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][210/1251] eta 0:05:01 lr 0.000259 time 0.2860 (0.2894) loss 4.9184 (5.7885) grad_norm 2.0506 (2.3311) [2021-04-14 17:33:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][220/1251] eta 0:04:57 lr 0.000260 time 0.2619 (0.2887) loss 5.8168 (5.7802) grad_norm 2.4030 (2.3361) [2021-04-14 17:33:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][230/1251] eta 0:04:53 lr 0.000260 time 0.2919 (0.2879) loss 6.2518 (5.7735) grad_norm 1.8043 (2.3396) [2021-04-14 17:33:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][240/1251] eta 0:04:50 lr 0.000260 time 0.2986 (0.2873) loss 5.8987 (5.7712) grad_norm 1.8162 (2.3404) [2021-04-14 17:33:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][250/1251] eta 0:04:46 lr 0.000261 time 0.2548 (0.2865) loss 6.0090 (5.7665) grad_norm 2.2507 (2.3345) [2021-04-14 17:33:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][260/1251] eta 0:04:43 lr 0.000261 time 0.2725 (0.2860) loss 6.0230 (5.7734) grad_norm 2.1018 (2.3312) [2021-04-14 17:34:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][270/1251] eta 0:04:40 lr 0.000262 time 0.2838 (0.2855) loss 6.0600 (5.7723) grad_norm 2.3794 (2.3300) [2021-04-14 17:34:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][280/1251] eta 0:04:36 lr 0.000262 time 0.2778 (0.2850) loss 6.0340 (5.7706) grad_norm 1.9633 (2.3281) [2021-04-14 17:34:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][290/1251] eta 0:04:33 lr 0.000262 time 0.2714 (0.2844) loss 5.9733 (5.7741) grad_norm 2.0645 (2.3266) [2021-04-14 17:34:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][300/1251] eta 0:04:30 lr 0.000263 time 0.2731 (0.2840) loss 5.3563 (5.7718) grad_norm 1.8844 (2.3427) [2021-04-14 17:34:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][310/1251] eta 0:04:26 lr 0.000263 time 0.2565 (0.2835) loss 5.5023 (5.7700) grad_norm 2.5661 (2.3465) [2021-04-14 17:34:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][320/1251] eta 0:04:23 lr 0.000264 time 0.2616 (0.2832) loss 4.5893 (5.7709) grad_norm 1.7695 (2.3490) [2021-04-14 17:34:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][330/1251] eta 0:04:20 lr 0.000264 time 0.2557 (0.2829) loss 5.8133 (5.7695) grad_norm 2.4422 (2.3616) [2021-04-14 17:34:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][340/1251] eta 0:04:17 lr 0.000264 time 0.2931 (0.2826) loss 5.9414 (5.7712) grad_norm 2.7178 (2.3604) [2021-04-14 17:34:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][350/1251] eta 0:04:14 lr 0.000265 time 0.2757 (0.2823) loss 5.0778 (5.7630) grad_norm 3.0838 (2.3746) [2021-04-14 17:34:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][360/1251] eta 0:04:11 lr 0.000265 time 0.2668 (0.2825) loss 6.0620 (5.7619) grad_norm 3.2926 (2.3776) [2021-04-14 17:34:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][370/1251] eta 0:04:08 lr 0.000266 time 0.2733 (0.2826) loss 5.7855 (5.7668) grad_norm 2.0626 (2.3713) [2021-04-14 17:34:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][380/1251] eta 0:04:05 lr 0.000266 time 0.2811 (0.2824) loss 5.9437 (5.7632) grad_norm 2.5922 (2.3715) [2021-04-14 17:34:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][390/1251] eta 0:04:02 lr 0.000266 time 0.2801 (0.2822) loss 5.7562 (5.7631) grad_norm 3.1305 (2.3744) [2021-04-14 17:34:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][400/1251] eta 0:03:59 lr 0.000267 time 0.2770 (0.2820) loss 5.8330 (5.7593) grad_norm 2.2932 (2.3747) [2021-04-14 17:34:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][410/1251] eta 0:03:56 lr 0.000267 time 0.2690 (0.2818) loss 5.9606 (5.7598) grad_norm 2.4131 (2.3742) [2021-04-14 17:34:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][420/1251] eta 0:03:53 lr 0.000268 time 0.2639 (0.2815) loss 6.0455 (5.7555) grad_norm 2.2500 (2.3683) [2021-04-14 17:34:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][430/1251] eta 0:03:50 lr 0.000268 time 0.2689 (0.2813) loss 5.6790 (5.7505) grad_norm 2.5789 (2.3640) [2021-04-14 17:34:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][440/1251] eta 0:03:47 lr 0.000268 time 0.2674 (0.2810) loss 6.2383 (5.7495) grad_norm 2.7642 (2.3631) [2021-04-14 17:34:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][450/1251] eta 0:03:44 lr 0.000269 time 0.2609 (0.2807) loss 5.9633 (5.7529) grad_norm 2.9112 (2.3615) [2021-04-14 17:34:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][460/1251] eta 0:03:41 lr 0.000269 time 0.2651 (0.2805) loss 6.4726 (5.7554) grad_norm 2.3986 (2.3589) [2021-04-14 17:34:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][470/1251] eta 0:03:38 lr 0.000270 time 0.2748 (0.2803) loss 5.3795 (5.7574) grad_norm 3.0063 (2.3561) [2021-04-14 17:34:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][480/1251] eta 0:03:36 lr 0.000270 time 0.2546 (0.2803) loss 5.1660 (5.7578) grad_norm 3.5805 (2.3619) [2021-04-14 17:35:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][490/1251] eta 0:03:33 lr 0.000270 time 0.2767 (0.2802) loss 5.1920 (5.7518) grad_norm 2.3232 (2.3618) [2021-04-14 17:35:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][500/1251] eta 0:03:30 lr 0.000271 time 0.2876 (0.2802) loss 6.2941 (5.7500) grad_norm 2.7093 (2.3562) [2021-04-14 17:35:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][510/1251] eta 0:03:27 lr 0.000271 time 0.2948 (0.2800) loss 5.7767 (5.7459) grad_norm 2.9168 (2.3537) [2021-04-14 17:35:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][520/1251] eta 0:03:24 lr 0.000272 time 0.2790 (0.2798) loss 6.2012 (5.7460) grad_norm 2.0580 (2.3505) [2021-04-14 17:35:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][530/1251] eta 0:03:21 lr 0.000272 time 0.2570 (0.2796) loss 5.2568 (5.7449) grad_norm 1.5729 (2.3429) [2021-04-14 17:35:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][540/1251] eta 0:03:18 lr 0.000272 time 0.2875 (0.2795) loss 5.9514 (5.7490) grad_norm 2.3597 (2.3405) [2021-04-14 17:35:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][550/1251] eta 0:03:15 lr 0.000273 time 0.2688 (0.2793) loss 4.8214 (5.7438) grad_norm 2.3095 (2.3402) [2021-04-14 17:35:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][560/1251] eta 0:03:12 lr 0.000273 time 0.2713 (0.2792) loss 5.4018 (5.7439) grad_norm 1.6110 (2.3374) [2021-04-14 17:35:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][570/1251] eta 0:03:10 lr 0.000274 time 0.3007 (0.2792) loss 5.7162 (5.7431) grad_norm 2.0150 (2.3357) [2021-04-14 17:35:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][580/1251] eta 0:03:07 lr 0.000274 time 0.2414 (0.2791) loss 4.9505 (5.7426) grad_norm 2.0452 (2.3360) [2021-04-14 17:35:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][590/1251] eta 0:03:04 lr 0.000274 time 0.2758 (0.2789) loss 6.0692 (5.7439) grad_norm 1.7579 (2.3356) [2021-04-14 17:35:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][600/1251] eta 0:03:01 lr 0.000275 time 0.2800 (0.2788) loss 5.4806 (5.7374) grad_norm 1.8498 (2.3382) [2021-04-14 17:35:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][610/1251] eta 0:02:58 lr 0.000275 time 0.2618 (0.2787) loss 5.2195 (5.7342) grad_norm 2.5294 (2.3359) [2021-04-14 17:35:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][620/1251] eta 0:02:55 lr 0.000276 time 0.2902 (0.2786) loss 5.2973 (5.7339) grad_norm 2.7866 (2.3400) [2021-04-14 17:35:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][630/1251] eta 0:02:52 lr 0.000276 time 0.2897 (0.2785) loss 6.3070 (5.7343) grad_norm 1.8023 (2.3368) [2021-04-14 17:35:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][640/1251] eta 0:02:50 lr 0.000276 time 0.2707 (0.2784) loss 6.1534 (5.7341) grad_norm 2.1642 (2.3363) [2021-04-14 17:35:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][650/1251] eta 0:02:47 lr 0.000277 time 0.2419 (0.2782) loss 4.6114 (5.7305) grad_norm 2.4422 (2.3421) [2021-04-14 17:35:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][660/1251] eta 0:02:44 lr 0.000277 time 0.2660 (0.2782) loss 5.4132 (5.7307) grad_norm 2.8472 (2.3379) [2021-04-14 17:35:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][670/1251] eta 0:02:41 lr 0.000278 time 0.2777 (0.2783) loss 5.3808 (5.7296) grad_norm 2.1324 (2.3355) [2021-04-14 17:35:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][680/1251] eta 0:02:38 lr 0.000278 time 0.2556 (0.2781) loss 6.1151 (5.7302) grad_norm 2.2068 (2.3340) [2021-04-14 17:35:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][690/1251] eta 0:02:35 lr 0.000278 time 0.2840 (0.2781) loss 5.8194 (5.7258) grad_norm 3.3002 (2.3346) [2021-04-14 17:35:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][700/1251] eta 0:02:33 lr 0.000279 time 0.2486 (0.2780) loss 5.4601 (5.7217) grad_norm 1.7699 (2.3358) [2021-04-14 17:36:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][710/1251] eta 0:02:30 lr 0.000279 time 0.2728 (0.2779) loss 5.4854 (5.7217) grad_norm 2.1358 (2.3368) [2021-04-14 17:36:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][720/1251] eta 0:02:27 lr 0.000279 time 0.2668 (0.2779) loss 6.0572 (5.7198) grad_norm 2.1413 (2.3384) [2021-04-14 17:36:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][730/1251] eta 0:02:24 lr 0.000280 time 0.2737 (0.2778) loss 5.1594 (5.7179) grad_norm 2.2990 (2.3355) [2021-04-14 17:36:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][740/1251] eta 0:02:21 lr 0.000280 time 0.2944 (0.2777) loss 5.2317 (5.7151) grad_norm 2.1844 (2.3331) [2021-04-14 17:36:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][750/1251] eta 0:02:19 lr 0.000281 time 0.2651 (0.2776) loss 5.0919 (5.7085) grad_norm 3.4288 (2.3314) [2021-04-14 17:36:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][760/1251] eta 0:02:16 lr 0.000281 time 0.2636 (0.2775) loss 6.3234 (5.7094) grad_norm 2.5056 (2.3326) [2021-04-14 17:36:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][770/1251] eta 0:02:13 lr 0.000281 time 0.2740 (0.2774) loss 5.8683 (5.7083) grad_norm 1.9270 (2.3316) [2021-04-14 17:36:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][780/1251] eta 0:02:10 lr 0.000282 time 0.2675 (0.2774) loss 6.1039 (5.7092) grad_norm 2.3246 (2.3313) [2021-04-14 17:36:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][790/1251] eta 0:02:07 lr 0.000282 time 0.2610 (0.2773) loss 6.0080 (5.7070) grad_norm 1.8117 (2.3299) [2021-04-14 17:36:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][800/1251] eta 0:02:05 lr 0.000283 time 0.2734 (0.2772) loss 5.9406 (5.7050) grad_norm 2.3159 (2.3302) [2021-04-14 17:36:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][810/1251] eta 0:02:02 lr 0.000283 time 0.2657 (0.2771) loss 5.7083 (5.7072) grad_norm 2.5057 (2.3291) [2021-04-14 17:36:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][820/1251] eta 0:01:59 lr 0.000283 time 0.2587 (0.2771) loss 5.3847 (5.7064) grad_norm 1.9479 (2.3310) [2021-04-14 17:36:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][830/1251] eta 0:01:56 lr 0.000284 time 0.2498 (0.2770) loss 4.9752 (5.7036) grad_norm 2.4408 (2.3308) [2021-04-14 17:36:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][840/1251] eta 0:01:53 lr 0.000284 time 0.2438 (0.2769) loss 5.4970 (5.7048) grad_norm 3.5885 (2.3328) [2021-04-14 17:36:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][850/1251] eta 0:01:51 lr 0.000285 time 0.2645 (0.2769) loss 5.1369 (5.7050) grad_norm 2.8882 (2.3336) [2021-04-14 17:36:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][860/1251] eta 0:01:48 lr 0.000285 time 0.2998 (0.2768) loss 5.8729 (5.7039) grad_norm 1.9312 (2.3339) [2021-04-14 17:36:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][870/1251] eta 0:01:45 lr 0.000285 time 0.2426 (0.2767) loss 6.2383 (5.7042) grad_norm 2.4212 (2.3316) [2021-04-14 17:36:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][880/1251] eta 0:01:42 lr 0.000286 time 0.2609 (0.2766) loss 5.5427 (5.7050) grad_norm 2.2273 (2.3298) [2021-04-14 17:36:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][890/1251] eta 0:01:39 lr 0.000286 time 0.2651 (0.2766) loss 5.9417 (5.7036) grad_norm 2.8208 (2.3343) [2021-04-14 17:36:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][900/1251] eta 0:01:37 lr 0.000287 time 0.2769 (0.2766) loss 5.7904 (5.7069) grad_norm 2.1236 (2.3342) [2021-04-14 17:36:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][910/1251] eta 0:01:34 lr 0.000287 time 0.2686 (0.2765) loss 5.6660 (5.7086) grad_norm 1.9035 (2.3316) [2021-04-14 17:36:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][920/1251] eta 0:01:31 lr 0.000287 time 0.2517 (0.2765) loss 6.0156 (5.7093) grad_norm 2.4510 (2.3291) [2021-04-14 17:37:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][930/1251] eta 0:01:28 lr 0.000288 time 0.2477 (0.2766) loss 5.4729 (5.7061) grad_norm 1.8729 (2.3276) [2021-04-14 17:37:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][940/1251] eta 0:01:25 lr 0.000288 time 0.2674 (0.2765) loss 6.1782 (5.7053) grad_norm 2.3405 (2.3290) [2021-04-14 17:37:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][950/1251] eta 0:01:23 lr 0.000289 time 0.2746 (0.2766) loss 5.1777 (5.7050) grad_norm 2.7560 (2.3309) [2021-04-14 17:37:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][960/1251] eta 0:01:20 lr 0.000289 time 0.2529 (0.2765) loss 5.8135 (5.7040) grad_norm 2.5479 (2.3320) [2021-04-14 17:37:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][970/1251] eta 0:01:17 lr 0.000289 time 0.2711 (0.2765) loss 5.6343 (5.7051) grad_norm 2.1252 (2.3302) [2021-04-14 17:37:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][980/1251] eta 0:01:14 lr 0.000290 time 0.2828 (0.2764) loss 5.1690 (5.7044) grad_norm 2.4943 (2.3283) [2021-04-14 17:37:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][990/1251] eta 0:01:12 lr 0.000290 time 0.2676 (0.2764) loss 5.2803 (5.7045) grad_norm 3.0183 (2.3264) [2021-04-14 17:37:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][1000/1251] eta 0:01:09 lr 0.000291 time 0.2772 (0.2763) loss 5.7789 (5.7036) grad_norm 2.0707 (2.3249) [2021-04-14 17:37:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][1010/1251] eta 0:01:06 lr 0.000291 time 0.2533 (0.2762) loss 6.0714 (5.7049) grad_norm 2.8166 (2.3254) [2021-04-14 17:37:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][1020/1251] eta 0:01:03 lr 0.000291 time 0.2628 (0.2762) loss 4.4964 (5.7050) grad_norm 1.8517 (2.3229) [2021-04-14 17:37:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][1030/1251] eta 0:01:01 lr 0.000292 time 0.2558 (0.2762) loss 5.7218 (5.7010) grad_norm 2.2106 (2.3235) [2021-04-14 17:37:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][1040/1251] eta 0:00:58 lr 0.000292 time 0.2925 (0.2762) loss 5.6979 (5.7004) grad_norm 1.7996 (2.3228) [2021-04-14 17:37:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][1050/1251] eta 0:00:55 lr 0.000293 time 0.2663 (0.2761) loss 5.4941 (5.7002) grad_norm 1.7129 (2.3210) [2021-04-14 17:37:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][1060/1251] eta 0:00:52 lr 0.000293 time 0.2699 (0.2760) loss 5.4833 (5.7002) grad_norm 2.0909 (2.3184) [2021-04-14 17:37:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][1070/1251] eta 0:00:49 lr 0.000293 time 0.2422 (0.2760) loss 4.6616 (5.6989) grad_norm 3.3449 (2.3181) [2021-04-14 17:37:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][1080/1251] eta 0:00:47 lr 0.000294 time 0.2951 (0.2761) loss 5.6421 (5.6968) grad_norm 2.1685 (2.3171) [2021-04-14 17:37:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][1090/1251] eta 0:00:44 lr 0.000294 time 0.2737 (0.2760) loss 5.9798 (5.6980) grad_norm 2.2849 (inf) [2021-04-14 17:37:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][1100/1251] eta 0:00:41 lr 0.000295 time 0.2607 (0.2759) loss 5.9277 (5.6985) grad_norm 2.7348 (inf) [2021-04-14 17:37:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][1110/1251] eta 0:00:38 lr 0.000295 time 0.2636 (0.2759) loss 5.8751 (5.6980) grad_norm 1.8293 (inf) [2021-04-14 17:37:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][1120/1251] eta 0:00:36 lr 0.000295 time 0.2736 (0.2758) loss 6.1771 (5.6971) grad_norm 2.1242 (inf) [2021-04-14 17:37:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][1130/1251] eta 0:00:33 lr 0.000296 time 0.2663 (0.2758) loss 6.1086 (5.6966) grad_norm 2.2652 (inf) [2021-04-14 17:37:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][1140/1251] eta 0:00:30 lr 0.000296 time 0.2741 (0.2758) loss 4.6329 (5.6958) grad_norm 2.0389 (inf) [2021-04-14 17:38:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][1150/1251] eta 0:00:27 lr 0.000297 time 0.2623 (0.2757) loss 5.3654 (5.6934) grad_norm 1.8881 (inf) [2021-04-14 17:38:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][1160/1251] eta 0:00:25 lr 0.000297 time 0.2981 (0.2757) loss 5.9998 (5.6928) grad_norm 1.8600 (inf) [2021-04-14 17:38:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][1170/1251] eta 0:00:22 lr 0.000297 time 0.2994 (0.2757) loss 5.9883 (5.6931) grad_norm 2.1787 (inf) [2021-04-14 17:38:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][1180/1251] eta 0:00:19 lr 0.000298 time 0.2615 (0.2757) loss 5.1846 (5.6920) grad_norm 2.1712 (inf) [2021-04-14 17:38:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][1190/1251] eta 0:00:16 lr 0.000298 time 0.2667 (0.2757) loss 5.4451 (5.6900) grad_norm 1.9529 (inf) [2021-04-14 17:38:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][1200/1251] eta 0:00:14 lr 0.000299 time 0.2664 (0.2756) loss 6.1953 (5.6899) grad_norm 2.5673 (inf) [2021-04-14 17:38:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][1210/1251] eta 0:00:11 lr 0.000299 time 0.2690 (0.2756) loss 5.6584 (5.6900) grad_norm 3.5675 (inf) [2021-04-14 17:38:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][1220/1251] eta 0:00:08 lr 0.000299 time 0.2709 (0.2756) loss 5.2277 (5.6899) grad_norm 2.3993 (inf) [2021-04-14 17:38:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][1230/1251] eta 0:00:05 lr 0.000300 time 0.2620 (0.2755) loss 5.2482 (5.6893) grad_norm 2.2110 (inf) [2021-04-14 17:38:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][1240/1251] eta 0:00:03 lr 0.000300 time 0.2498 (0.2755) loss 4.9809 (5.6873) grad_norm 2.4935 (inf) [2021-04-14 17:38:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [5/300][1250/1251] eta 0:00:00 lr 0.000301 time 0.2479 (0.2753) loss 5.4736 (5.6868) grad_norm 2.2968 (inf) [2021-04-14 17:38:29 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 5 training takes 0:05:46 [2021-04-14 17:38:29 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_5.pth saving...... [2021-04-14 17:38:54 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_5.pth saved !!! [2021-04-14 17:38:56 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.317 (1.317) Loss 3.5943 (3.5943) Acc@1 28.027 (28.027) Acc@5 52.051 (52.051) [2021-04-14 17:38:57 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.103 (0.258) Loss 3.5505 (3.5146) Acc@1 29.199 (30.380) Acc@5 54.883 (54.998) [2021-04-14 17:38:59 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.153 (0.236) Loss 3.5299 (3.5199) Acc@1 30.273 (30.115) Acc@5 54.688 (54.790) [2021-04-14 17:39:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.107 (0.234) Loss 3.5811 (3.5257) Acc@1 27.930 (29.996) Acc@5 54.199 (54.546) [2021-04-14 17:39:03 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.208) Loss 3.5921 (3.5258) Acc@1 30.273 (29.909) Acc@5 52.539 (54.371) [2021-04-14 17:39:05 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 30.004 Acc@5 54.510 [2021-04-14 17:39:05 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 30.0% [2021-04-14 17:39:05 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 30.00% [2021-04-14 17:39:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][0/1251] eta 1:09:46 lr 0.000301 time 3.3465 (3.3465) loss 5.1638 (5.1638) grad_norm 1.9254 (1.9254) [2021-04-14 17:39:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][10/1251] eta 0:11:46 lr 0.000301 time 0.2566 (0.5691) loss 5.3203 (5.4744) grad_norm 2.0744 (2.1514) [2021-04-14 17:39:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][20/1251] eta 0:08:47 lr 0.000301 time 0.2424 (0.4286) loss 5.3556 (5.4951) grad_norm 2.0216 (2.1346) [2021-04-14 17:39:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][30/1251] eta 0:07:41 lr 0.000302 time 0.2878 (0.3779) loss 5.5453 (5.4870) grad_norm 2.5805 (2.1811) [2021-04-14 17:39:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][40/1251] eta 0:07:05 lr 0.000302 time 0.2540 (0.3513) loss 5.6239 (5.5446) grad_norm 3.8450 (2.2320) [2021-04-14 17:39:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][50/1251] eta 0:06:43 lr 0.000303 time 0.2823 (0.3361) loss 4.5700 (5.5105) grad_norm 2.4292 (2.2722) [2021-04-14 17:39:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][60/1251] eta 0:06:28 lr 0.000303 time 0.2960 (0.3260) loss 5.7184 (5.4644) grad_norm 2.0944 (2.2584) [2021-04-14 17:39:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][70/1251] eta 0:06:16 lr 0.000303 time 0.2771 (0.3184) loss 5.4889 (5.5038) grad_norm 1.8714 (2.2666) [2021-04-14 17:39:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][80/1251] eta 0:06:06 lr 0.000304 time 0.2409 (0.3127) loss 5.7950 (5.5351) grad_norm 1.9160 (2.2829) [2021-04-14 17:39:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][90/1251] eta 0:05:57 lr 0.000304 time 0.2671 (0.3079) loss 6.0872 (5.5415) grad_norm 1.8161 (2.3420) [2021-04-14 17:39:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][100/1251] eta 0:05:50 lr 0.000305 time 0.2991 (0.3044) loss 6.1880 (5.5586) grad_norm 2.3395 (2.3152) [2021-04-14 17:39:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][110/1251] eta 0:05:44 lr 0.000305 time 0.2610 (0.3020) loss 6.3147 (5.5550) grad_norm 2.0873 (2.3002) [2021-04-14 17:39:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][120/1251] eta 0:05:38 lr 0.000305 time 0.2553 (0.2992) loss 5.6690 (5.5611) grad_norm 1.9195 (2.3074) [2021-04-14 17:39:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][130/1251] eta 0:05:33 lr 0.000306 time 0.3343 (0.2979) loss 5.1560 (5.5684) grad_norm 2.8676 (2.3056) [2021-04-14 17:39:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][140/1251] eta 0:05:30 lr 0.000306 time 0.2718 (0.2972) loss 5.9894 (5.5671) grad_norm 1.9076 (2.3138) [2021-04-14 17:39:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][150/1251] eta 0:05:26 lr 0.000307 time 0.2570 (0.2962) loss 5.9924 (5.5732) grad_norm 2.2033 (2.3004) [2021-04-14 17:39:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][160/1251] eta 0:05:21 lr 0.000307 time 0.2854 (0.2948) loss 5.6174 (5.5848) grad_norm 1.7772 (2.2826) [2021-04-14 17:39:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][170/1251] eta 0:05:17 lr 0.000307 time 0.2441 (0.2934) loss 5.4599 (5.5750) grad_norm 1.9891 (2.2659) [2021-04-14 17:39:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][180/1251] eta 0:05:12 lr 0.000308 time 0.2763 (0.2921) loss 5.3761 (5.5736) grad_norm 1.8954 (2.2662) [2021-04-14 17:40:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][190/1251] eta 0:05:08 lr 0.000308 time 0.2734 (0.2911) loss 5.7160 (5.5791) grad_norm 2.2530 (2.2592) [2021-04-14 17:40:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][200/1251] eta 0:05:04 lr 0.000309 time 0.2794 (0.2900) loss 5.4347 (5.5849) grad_norm 3.1242 (2.2559) [2021-04-14 17:40:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][210/1251] eta 0:05:00 lr 0.000309 time 0.2713 (0.2890) loss 5.8988 (5.5852) grad_norm 2.1004 (2.2556) [2021-04-14 17:40:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][220/1251] eta 0:04:57 lr 0.000309 time 0.2557 (0.2883) loss 5.5078 (5.5655) grad_norm 1.6881 (2.2505) [2021-04-14 17:40:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][230/1251] eta 0:04:53 lr 0.000310 time 0.2682 (0.2875) loss 5.9495 (5.5678) grad_norm 2.1111 (2.2494) [2021-04-14 17:40:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][240/1251] eta 0:04:49 lr 0.000310 time 0.2588 (0.2868) loss 6.0611 (5.5672) grad_norm 1.7922 (2.2492) [2021-04-14 17:40:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][250/1251] eta 0:04:46 lr 0.000311 time 0.2728 (0.2862) loss 5.8519 (5.5642) grad_norm 3.6018 (2.2717) [2021-04-14 17:40:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][260/1251] eta 0:04:42 lr 0.000311 time 0.2721 (0.2856) loss 5.9121 (5.5648) grad_norm 2.6408 (2.2749) [2021-04-14 17:40:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][270/1251] eta 0:04:39 lr 0.000311 time 0.2669 (0.2853) loss 6.0029 (5.5619) grad_norm 2.0187 (2.2681) [2021-04-14 17:40:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][280/1251] eta 0:04:36 lr 0.000312 time 0.2632 (0.2848) loss 6.0498 (5.5641) grad_norm 2.0306 (2.2734) [2021-04-14 17:40:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][290/1251] eta 0:04:33 lr 0.000312 time 0.2636 (0.2846) loss 5.8873 (5.5681) grad_norm 2.1108 (2.2654) [2021-04-14 17:40:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][300/1251] eta 0:04:30 lr 0.000313 time 0.2708 (0.2841) loss 5.7815 (5.5676) grad_norm 1.7700 (2.2554) [2021-04-14 17:40:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][310/1251] eta 0:04:27 lr 0.000313 time 0.2954 (0.2839) loss 5.9776 (5.5669) grad_norm 2.2931 (2.2523) [2021-04-14 17:40:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][320/1251] eta 0:04:23 lr 0.000313 time 0.2786 (0.2834) loss 5.2886 (5.5629) grad_norm 2.2845 (2.2504) [2021-04-14 17:40:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][330/1251] eta 0:04:20 lr 0.000314 time 0.2638 (0.2830) loss 6.3922 (5.5686) grad_norm 1.7085 (2.2465) [2021-04-14 17:40:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][340/1251] eta 0:04:17 lr 0.000314 time 0.2417 (0.2827) loss 5.7473 (5.5670) grad_norm 1.9529 (2.2490) [2021-04-14 17:40:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][350/1251] eta 0:04:14 lr 0.000315 time 0.2869 (0.2825) loss 5.7702 (5.5599) grad_norm 3.9080 (2.2548) [2021-04-14 17:40:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][360/1251] eta 0:04:11 lr 0.000315 time 0.2757 (0.2822) loss 5.7804 (5.5628) grad_norm 2.0685 (2.2584) [2021-04-14 17:40:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][370/1251] eta 0:04:08 lr 0.000315 time 0.2690 (0.2819) loss 5.5132 (5.5636) grad_norm 1.7670 (2.2566) [2021-04-14 17:40:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][380/1251] eta 0:04:05 lr 0.000316 time 0.2896 (0.2816) loss 5.9743 (5.5618) grad_norm 1.7270 (2.2591) [2021-04-14 17:40:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][390/1251] eta 0:04:02 lr 0.000316 time 0.2544 (0.2813) loss 5.1617 (5.5649) grad_norm 2.3190 (2.2578) [2021-04-14 17:40:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][400/1251] eta 0:03:59 lr 0.000317 time 0.2953 (0.2810) loss 6.0346 (5.5712) grad_norm 2.1580 (2.2547) [2021-04-14 17:41:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][410/1251] eta 0:03:56 lr 0.000317 time 0.2826 (0.2807) loss 5.1736 (5.5671) grad_norm 2.5115 (2.2516) [2021-04-14 17:41:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][420/1251] eta 0:03:53 lr 0.000317 time 0.2729 (0.2805) loss 5.7615 (5.5655) grad_norm 2.1070 (2.2489) [2021-04-14 17:41:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][430/1251] eta 0:03:50 lr 0.000318 time 0.2824 (0.2804) loss 5.7760 (5.5655) grad_norm 2.3183 (2.2510) [2021-04-14 17:41:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][440/1251] eta 0:03:47 lr 0.000318 time 0.2613 (0.2801) loss 5.6013 (5.5694) grad_norm 2.0559 (2.2448) [2021-04-14 17:41:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][450/1251] eta 0:03:44 lr 0.000319 time 0.2538 (0.2799) loss 5.9332 (5.5657) grad_norm 1.7452 (2.2473) [2021-04-14 17:41:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][460/1251] eta 0:03:41 lr 0.000319 time 0.2775 (0.2796) loss 4.5300 (5.5622) grad_norm 1.6881 (2.2438) [2021-04-14 17:41:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][470/1251] eta 0:03:38 lr 0.000319 time 0.2686 (0.2795) loss 5.3303 (5.5603) grad_norm 1.8826 (2.2423) [2021-04-14 17:41:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][480/1251] eta 0:03:35 lr 0.000320 time 0.2836 (0.2793) loss 5.8560 (5.5580) grad_norm 1.8130 (2.2395) [2021-04-14 17:41:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][490/1251] eta 0:03:32 lr 0.000320 time 0.2826 (0.2792) loss 4.8639 (5.5576) grad_norm 2.6028 (2.2416) [2021-04-14 17:41:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][500/1251] eta 0:03:29 lr 0.000321 time 0.2590 (0.2792) loss 5.8096 (5.5577) grad_norm 2.4915 (2.2411) [2021-04-14 17:41:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][510/1251] eta 0:03:26 lr 0.000321 time 0.2919 (0.2792) loss 5.2689 (5.5559) grad_norm 2.2249 (2.2491) [2021-04-14 17:41:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][520/1251] eta 0:03:23 lr 0.000321 time 0.2967 (0.2791) loss 5.1481 (5.5513) grad_norm 2.5115 (2.2499) [2021-04-14 17:41:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][530/1251] eta 0:03:21 lr 0.000322 time 0.2849 (0.2789) loss 5.9386 (5.5536) grad_norm 2.9025 (2.2526) [2021-04-14 17:41:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][540/1251] eta 0:03:18 lr 0.000322 time 0.2645 (0.2788) loss 5.7400 (5.5504) grad_norm 2.6013 (2.2534) [2021-04-14 17:41:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][550/1251] eta 0:03:15 lr 0.000323 time 0.2672 (0.2786) loss 4.4209 (5.5443) grad_norm 2.1215 (2.2528) [2021-04-14 17:41:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][560/1251] eta 0:03:12 lr 0.000323 time 0.2803 (0.2785) loss 5.8717 (5.5442) grad_norm 2.2029 (2.2504) [2021-04-14 17:41:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][570/1251] eta 0:03:09 lr 0.000323 time 0.2793 (0.2784) loss 6.1984 (5.5485) grad_norm 1.9362 (2.2552) [2021-04-14 17:41:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][580/1251] eta 0:03:06 lr 0.000324 time 0.2877 (0.2782) loss 5.6595 (5.5474) grad_norm 2.4087 (2.2535) [2021-04-14 17:41:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][590/1251] eta 0:03:03 lr 0.000324 time 0.2934 (0.2781) loss 5.7297 (5.5527) grad_norm 2.0409 (2.2511) [2021-04-14 17:41:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][600/1251] eta 0:03:00 lr 0.000325 time 0.2596 (0.2779) loss 6.0113 (5.5490) grad_norm 2.2697 (2.2479) [2021-04-14 17:41:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][610/1251] eta 0:02:58 lr 0.000325 time 0.2538 (0.2778) loss 5.1570 (5.5462) grad_norm 2.0163 (2.2463) [2021-04-14 17:41:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][620/1251] eta 0:02:55 lr 0.000325 time 0.2725 (0.2777) loss 6.2023 (5.5503) grad_norm 1.6468 (2.2464) [2021-04-14 17:42:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][630/1251] eta 0:02:52 lr 0.000326 time 0.2908 (0.2777) loss 5.2880 (5.5466) grad_norm 2.1708 (2.2439) [2021-04-14 17:42:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][640/1251] eta 0:02:49 lr 0.000326 time 0.2604 (0.2775) loss 4.9463 (5.5485) grad_norm 2.5253 (2.2444) [2021-04-14 17:42:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][650/1251] eta 0:02:46 lr 0.000327 time 0.2756 (0.2774) loss 5.2187 (5.5489) grad_norm 1.8774 (2.2421) [2021-04-14 17:42:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][660/1251] eta 0:02:43 lr 0.000327 time 0.2505 (0.2773) loss 5.8559 (5.5451) grad_norm 2.7828 (2.2429) [2021-04-14 17:42:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][670/1251] eta 0:02:41 lr 0.000327 time 0.2603 (0.2772) loss 5.8984 (5.5458) grad_norm 2.3883 (2.2399) [2021-04-14 17:42:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][680/1251] eta 0:02:38 lr 0.000328 time 0.2904 (0.2772) loss 5.9556 (5.5475) grad_norm 2.0291 (2.2369) [2021-04-14 17:42:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][690/1251] eta 0:02:35 lr 0.000328 time 0.2526 (0.2771) loss 5.2987 (5.5456) grad_norm 2.4872 (2.2399) [2021-04-14 17:42:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][700/1251] eta 0:02:32 lr 0.000329 time 0.2432 (0.2772) loss 6.2732 (5.5461) grad_norm 2.5914 (2.2417) [2021-04-14 17:42:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][710/1251] eta 0:02:29 lr 0.000329 time 0.2737 (0.2771) loss 6.0287 (5.5440) grad_norm 1.8811 (2.2450) [2021-04-14 17:42:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][720/1251] eta 0:02:27 lr 0.000329 time 0.2825 (0.2773) loss 4.8247 (5.5408) grad_norm 2.2288 (2.2444) [2021-04-14 17:42:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][730/1251] eta 0:02:24 lr 0.000330 time 0.2534 (0.2774) loss 4.7809 (5.5366) grad_norm 2.0162 (2.2466) [2021-04-14 17:42:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][740/1251] eta 0:02:21 lr 0.000330 time 0.2783 (0.2773) loss 5.9590 (5.5380) grad_norm 1.9175 (2.2447) [2021-04-14 17:42:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][750/1251] eta 0:02:18 lr 0.000331 time 0.2663 (0.2773) loss 6.0073 (5.5382) grad_norm 2.1289 (2.2448) [2021-04-14 17:42:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][760/1251] eta 0:02:16 lr 0.000331 time 0.2841 (0.2772) loss 5.9284 (5.5412) grad_norm 1.9031 (2.2442) [2021-04-14 17:42:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][770/1251] eta 0:02:13 lr 0.000331 time 0.2689 (0.2771) loss 6.0993 (5.5428) grad_norm 1.8076 (2.2424) [2021-04-14 17:42:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][780/1251] eta 0:02:10 lr 0.000332 time 0.2663 (0.2770) loss 5.5018 (5.5426) grad_norm 2.0355 (2.2445) [2021-04-14 17:42:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][790/1251] eta 0:02:07 lr 0.000332 time 0.2518 (0.2769) loss 5.3501 (5.5422) grad_norm 2.2115 (2.2422) [2021-04-14 17:42:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][800/1251] eta 0:02:04 lr 0.000333 time 0.2602 (0.2769) loss 4.7767 (5.5390) grad_norm 2.6341 (2.2399) [2021-04-14 17:42:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][810/1251] eta 0:02:02 lr 0.000333 time 0.2876 (0.2769) loss 5.5560 (5.5376) grad_norm 4.1139 (2.2465) [2021-04-14 17:42:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][820/1251] eta 0:01:59 lr 0.000333 time 0.2797 (0.2768) loss 4.8371 (5.5378) grad_norm 1.8469 (2.2478) [2021-04-14 17:42:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][830/1251] eta 0:01:56 lr 0.000334 time 0.2613 (0.2767) loss 5.1955 (5.5380) grad_norm 2.0152 (2.2449) [2021-04-14 17:42:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][840/1251] eta 0:01:53 lr 0.000334 time 0.2998 (0.2766) loss 5.5908 (5.5338) grad_norm 2.2318 (2.2451) [2021-04-14 17:43:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][850/1251] eta 0:01:50 lr 0.000335 time 0.2518 (0.2765) loss 6.1746 (5.5349) grad_norm 1.8617 (2.2428) [2021-04-14 17:43:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][860/1251] eta 0:01:48 lr 0.000335 time 0.2821 (0.2765) loss 4.4769 (5.5337) grad_norm 2.0499 (2.2418) [2021-04-14 17:43:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][870/1251] eta 0:01:45 lr 0.000335 time 0.2590 (0.2765) loss 5.4144 (5.5343) grad_norm 2.6915 (2.2403) [2021-04-14 17:43:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][880/1251] eta 0:01:42 lr 0.000336 time 0.2737 (0.2765) loss 5.8124 (5.5358) grad_norm 2.5139 (2.2421) [2021-04-14 17:43:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][890/1251] eta 0:01:39 lr 0.000336 time 0.2747 (0.2764) loss 4.9645 (5.5352) grad_norm 3.8526 (2.2464) [2021-04-14 17:43:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][900/1251] eta 0:01:37 lr 0.000337 time 0.2730 (0.2764) loss 5.2879 (5.5340) grad_norm 2.2796 (2.2466) [2021-04-14 17:43:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][910/1251] eta 0:01:34 lr 0.000337 time 0.2871 (0.2763) loss 5.1273 (5.5321) grad_norm 2.7986 (2.2474) [2021-04-14 17:43:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][920/1251] eta 0:01:31 lr 0.000337 time 0.2684 (0.2763) loss 6.0670 (5.5316) grad_norm 2.2201 (2.2480) [2021-04-14 17:43:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][930/1251] eta 0:01:28 lr 0.000338 time 0.2554 (0.2762) loss 5.9080 (5.5310) grad_norm 2.2943 (2.2468) [2021-04-14 17:43:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][940/1251] eta 0:01:25 lr 0.000338 time 0.2814 (0.2762) loss 5.2206 (5.5328) grad_norm 2.0639 (2.2490) [2021-04-14 17:43:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][950/1251] eta 0:01:23 lr 0.000339 time 0.2571 (0.2761) loss 5.7422 (5.5343) grad_norm 4.2558 (2.2516) [2021-04-14 17:43:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][960/1251] eta 0:01:20 lr 0.000339 time 0.2678 (0.2761) loss 5.4315 (5.5349) grad_norm 3.0696 (2.2517) [2021-04-14 17:43:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][970/1251] eta 0:01:17 lr 0.000339 time 0.2733 (0.2760) loss 5.1378 (5.5333) grad_norm 2.1336 (2.2485) [2021-04-14 17:43:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][980/1251] eta 0:01:14 lr 0.000340 time 0.2723 (0.2760) loss 5.9441 (5.5323) grad_norm 2.4192 (2.2502) [2021-04-14 17:43:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][990/1251] eta 0:01:12 lr 0.000340 time 0.2730 (0.2759) loss 6.1619 (5.5354) grad_norm 3.3581 (2.2508) [2021-04-14 17:43:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][1000/1251] eta 0:01:09 lr 0.000341 time 0.2702 (0.2759) loss 6.1930 (5.5371) grad_norm 2.2881 (2.2495) [2021-04-14 17:43:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][1010/1251] eta 0:01:06 lr 0.000341 time 0.2690 (0.2759) loss 4.7233 (5.5355) grad_norm 2.0578 (2.2487) [2021-04-14 17:43:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][1020/1251] eta 0:01:03 lr 0.000341 time 0.2795 (0.2758) loss 4.8938 (5.5353) grad_norm 2.4175 (2.2470) [2021-04-14 17:43:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][1030/1251] eta 0:01:00 lr 0.000342 time 0.2659 (0.2758) loss 5.4848 (5.5354) grad_norm 2.0290 (2.2471) [2021-04-14 17:43:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][1040/1251] eta 0:00:58 lr 0.000342 time 0.2771 (0.2757) loss 5.1684 (5.5361) grad_norm 1.9187 (2.2473) [2021-04-14 17:43:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][1050/1251] eta 0:00:55 lr 0.000343 time 0.2757 (0.2757) loss 5.5244 (5.5374) grad_norm 2.4785 (2.2483) [2021-04-14 17:43:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][1060/1251] eta 0:00:52 lr 0.000343 time 0.2717 (0.2757) loss 5.7855 (5.5366) grad_norm 2.7136 (2.2468) [2021-04-14 17:44:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][1070/1251] eta 0:00:49 lr 0.000343 time 0.2832 (0.2757) loss 4.7570 (5.5352) grad_norm 2.3866 (2.2460) [2021-04-14 17:44:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][1080/1251] eta 0:00:47 lr 0.000344 time 0.2845 (0.2756) loss 6.2385 (5.5340) grad_norm 3.1174 (2.2471) [2021-04-14 17:44:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][1090/1251] eta 0:00:44 lr 0.000344 time 0.2836 (0.2756) loss 5.2516 (5.5357) grad_norm 2.1463 (2.2483) [2021-04-14 17:44:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][1100/1251] eta 0:00:41 lr 0.000345 time 0.2827 (0.2757) loss 6.0782 (5.5360) grad_norm 2.3115 (2.2494) [2021-04-14 17:44:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][1110/1251] eta 0:00:38 lr 0.000345 time 0.2626 (0.2756) loss 5.4256 (5.5364) grad_norm 2.2549 (2.2478) [2021-04-14 17:44:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][1120/1251] eta 0:00:36 lr 0.000345 time 0.2694 (0.2756) loss 4.9161 (5.5371) grad_norm 2.9288 (2.2476) [2021-04-14 17:44:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][1130/1251] eta 0:00:33 lr 0.000346 time 0.2947 (0.2756) loss 5.2851 (5.5358) grad_norm 2.5083 (2.2474) [2021-04-14 17:44:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][1140/1251] eta 0:00:30 lr 0.000346 time 0.2743 (0.2756) loss 4.6628 (5.5371) grad_norm 2.0152 (2.2461) [2021-04-14 17:44:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][1150/1251] eta 0:00:27 lr 0.000347 time 0.2494 (0.2755) loss 4.9826 (5.5352) grad_norm 1.7919 (2.2434) [2021-04-14 17:44:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][1160/1251] eta 0:00:25 lr 0.000347 time 0.2689 (0.2755) loss 6.3013 (5.5347) grad_norm 2.4829 (2.2468) [2021-04-14 17:44:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][1170/1251] eta 0:00:22 lr 0.000347 time 0.2718 (0.2754) loss 4.8791 (5.5341) grad_norm 2.8908 (2.2493) [2021-04-14 17:44:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][1180/1251] eta 0:00:19 lr 0.000348 time 0.2463 (0.2754) loss 5.6306 (5.5326) grad_norm 3.3007 (2.2496) [2021-04-14 17:44:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][1190/1251] eta 0:00:16 lr 0.000348 time 0.2796 (0.2754) loss 5.1939 (5.5307) grad_norm 2.0133 (2.2496) [2021-04-14 17:44:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][1200/1251] eta 0:00:14 lr 0.000349 time 0.2897 (0.2754) loss 5.3960 (5.5308) grad_norm 1.9354 (2.2473) [2021-04-14 17:44:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][1210/1251] eta 0:00:11 lr 0.000349 time 0.2775 (0.2754) loss 5.3272 (5.5298) grad_norm 1.7083 (2.2467) [2021-04-14 17:44:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][1220/1251] eta 0:00:08 lr 0.000349 time 0.2610 (0.2753) loss 4.2072 (5.5303) grad_norm 2.1247 (2.2457) [2021-04-14 17:44:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][1230/1251] eta 0:00:05 lr 0.000350 time 0.2824 (0.2753) loss 6.1817 (5.5311) grad_norm 2.0144 (2.2465) [2021-04-14 17:44:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][1240/1251] eta 0:00:03 lr 0.000350 time 0.2481 (0.2752) loss 6.0188 (5.5310) grad_norm 2.4054 (2.2464) [2021-04-14 17:44:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [6/300][1250/1251] eta 0:00:00 lr 0.000351 time 0.2480 (0.2750) loss 5.8557 (5.5324) grad_norm 2.1671 (2.2465) [2021-04-14 17:44:51 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 6 training takes 0:05:46 [2021-04-14 17:44:51 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_6.pth saving...... [2021-04-14 17:45:13 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_6.pth saved !!! [2021-04-14 17:45:15 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.312 (1.312) Loss 3.1992 (3.1992) Acc@1 35.059 (35.059) Acc@5 59.082 (59.082) [2021-04-14 17:45:16 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.096 (0.224) Loss 3.2527 (3.2772) Acc@1 34.668 (33.141) Acc@5 59.277 (58.656) [2021-04-14 17:45:18 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.273 (0.211) Loss 3.2642 (3.2645) Acc@1 32.520 (33.580) Acc@5 59.668 (58.822) [2021-04-14 17:45:21 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.737 (0.240) Loss 3.3606 (3.2607) Acc@1 29.980 (33.729) Acc@5 57.422 (58.858) [2021-04-14 17:45:22 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.198 (0.219) Loss 3.3011 (3.2525) Acc@1 33.789 (33.703) Acc@5 57.422 (58.951) [2021-04-14 17:45:24 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 33.690 Acc@5 58.988 [2021-04-14 17:45:24 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 33.7% [2021-04-14 17:45:24 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 33.69% [2021-04-14 17:45:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][0/1251] eta 1:13:58 lr 0.000351 time 3.5476 (3.5476) loss 5.7396 (5.7396) grad_norm 2.3630 (2.3630) [2021-04-14 17:45:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][10/1251] eta 0:12:01 lr 0.000351 time 0.2681 (0.5815) loss 6.1757 (5.6536) grad_norm 1.8843 (2.0048) [2021-04-14 17:45:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][20/1251] eta 0:08:54 lr 0.000351 time 0.2727 (0.4346) loss 6.1306 (5.5662) grad_norm 2.7529 (2.1024) [2021-04-14 17:45:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][30/1251] eta 0:07:45 lr 0.000352 time 0.2710 (0.3815) loss 5.5562 (5.5300) grad_norm 1.9682 (2.0920) [2021-04-14 17:45:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][40/1251] eta 0:07:10 lr 0.000352 time 0.2669 (0.3557) loss 5.4223 (5.5445) grad_norm 2.3244 (2.1310) [2021-04-14 17:45:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][50/1251] eta 0:06:48 lr 0.000353 time 0.2870 (0.3399) loss 5.1018 (5.4928) grad_norm 2.1101 (2.2329) [2021-04-14 17:45:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][60/1251] eta 0:06:31 lr 0.000353 time 0.2948 (0.3285) loss 5.2105 (5.4985) grad_norm 2.6562 (2.2775) [2021-04-14 17:45:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][70/1251] eta 0:06:18 lr 0.000353 time 0.2787 (0.3206) loss 5.6431 (5.4902) grad_norm 2.5578 (2.2358) [2021-04-14 17:45:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][80/1251] eta 0:06:08 lr 0.000354 time 0.2610 (0.3150) loss 5.2922 (5.4809) grad_norm 2.3941 (2.2482) [2021-04-14 17:45:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][90/1251] eta 0:06:00 lr 0.000354 time 0.2707 (0.3105) loss 4.5130 (5.4866) grad_norm 1.7069 (2.2299) [2021-04-14 17:45:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][100/1251] eta 0:05:53 lr 0.000355 time 0.2719 (0.3068) loss 5.2007 (5.4878) grad_norm 2.2951 (2.2007) [2021-04-14 17:45:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][110/1251] eta 0:05:46 lr 0.000355 time 0.2878 (0.3037) loss 4.7942 (5.4767) grad_norm 1.7710 (2.1875) [2021-04-14 17:46:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][120/1251] eta 0:05:40 lr 0.000355 time 0.2963 (0.3010) loss 5.6151 (5.4818) grad_norm 1.9823 (2.1783) [2021-04-14 17:46:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][130/1251] eta 0:05:34 lr 0.000356 time 0.2880 (0.2986) loss 5.8692 (5.4812) grad_norm 2.0196 (2.1748) [2021-04-14 17:46:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][140/1251] eta 0:05:29 lr 0.000356 time 0.2463 (0.2965) loss 5.5686 (5.4685) grad_norm 2.7908 (2.2136) [2021-04-14 17:46:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][150/1251] eta 0:05:24 lr 0.000357 time 0.2651 (0.2949) loss 5.7223 (5.4535) grad_norm 1.9895 (2.2257) [2021-04-14 17:46:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][160/1251] eta 0:05:20 lr 0.000357 time 0.2689 (0.2934) loss 5.4406 (5.4598) grad_norm 2.0612 (2.2201) [2021-04-14 17:46:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][170/1251] eta 0:05:15 lr 0.000357 time 0.2802 (0.2922) loss 5.4775 (5.4664) grad_norm 1.8565 (2.2014) [2021-04-14 17:46:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][180/1251] eta 0:05:11 lr 0.000358 time 0.2480 (0.2910) loss 5.8837 (5.4790) grad_norm 1.6311 (2.1838) [2021-04-14 17:46:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][190/1251] eta 0:05:07 lr 0.000358 time 0.2509 (0.2899) loss 5.9478 (5.4677) grad_norm 1.8999 (2.1706) [2021-04-14 17:46:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][200/1251] eta 0:05:03 lr 0.000359 time 0.2685 (0.2890) loss 5.4824 (5.4642) grad_norm 1.8443 (2.1730) [2021-04-14 17:46:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][210/1251] eta 0:05:00 lr 0.000359 time 0.2520 (0.2884) loss 5.9568 (5.4883) grad_norm 2.3231 (2.1822) [2021-04-14 17:46:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][220/1251] eta 0:04:56 lr 0.000359 time 0.2790 (0.2878) loss 5.4666 (5.4799) grad_norm 2.1479 (2.1784) [2021-04-14 17:46:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][230/1251] eta 0:04:53 lr 0.000360 time 0.2880 (0.2874) loss 6.0831 (5.4876) grad_norm 2.6904 (2.1804) [2021-04-14 17:46:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][240/1251] eta 0:04:49 lr 0.000360 time 0.2598 (0.2867) loss 4.5268 (5.4756) grad_norm 2.0017 (2.1756) [2021-04-14 17:46:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][250/1251] eta 0:04:46 lr 0.000361 time 0.2877 (0.2861) loss 4.6348 (5.4596) grad_norm 2.0546 (2.1776) [2021-04-14 17:46:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][260/1251] eta 0:04:42 lr 0.000361 time 0.2697 (0.2855) loss 5.6748 (5.4585) grad_norm 1.7603 (2.1685) [2021-04-14 17:46:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][270/1251] eta 0:04:39 lr 0.000361 time 0.2868 (0.2850) loss 6.2169 (5.4622) grad_norm 1.6242 (2.1713) [2021-04-14 17:46:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][280/1251] eta 0:04:36 lr 0.000362 time 0.2518 (0.2845) loss 5.2466 (5.4587) grad_norm 1.9076 (2.1718) [2021-04-14 17:46:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][290/1251] eta 0:04:33 lr 0.000362 time 0.2707 (0.2848) loss 4.5233 (5.4510) grad_norm 2.3281 (2.1745) [2021-04-14 17:46:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][300/1251] eta 0:04:30 lr 0.000363 time 0.2474 (0.2848) loss 5.1644 (5.4481) grad_norm 2.1063 (2.1819) [2021-04-14 17:46:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][310/1251] eta 0:04:27 lr 0.000363 time 0.2965 (0.2844) loss 5.5034 (5.4441) grad_norm 1.9663 (2.1752) [2021-04-14 17:46:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][320/1251] eta 0:04:24 lr 0.000363 time 0.2594 (0.2839) loss 4.8959 (5.4461) grad_norm 2.4425 (2.1716) [2021-04-14 17:46:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][330/1251] eta 0:04:21 lr 0.000364 time 0.2747 (0.2836) loss 4.7444 (5.4433) grad_norm 1.7246 (2.1696) [2021-04-14 17:47:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][340/1251] eta 0:04:18 lr 0.000364 time 0.2565 (0.2833) loss 5.0795 (5.4409) grad_norm 1.9283 (2.1668) [2021-04-14 17:47:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][350/1251] eta 0:04:14 lr 0.000365 time 0.2656 (0.2830) loss 5.9321 (5.4462) grad_norm 2.1463 (2.1735) [2021-04-14 17:47:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][360/1251] eta 0:04:11 lr 0.000365 time 0.2833 (0.2827) loss 4.5135 (5.4432) grad_norm 2.0717 (2.1791) [2021-04-14 17:47:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][370/1251] eta 0:04:08 lr 0.000365 time 0.2730 (0.2824) loss 5.2258 (5.4403) grad_norm 2.1339 (2.1765) [2021-04-14 17:47:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][380/1251] eta 0:04:05 lr 0.000366 time 0.2547 (0.2822) loss 5.3252 (5.4366) grad_norm 2.2494 (2.1763) [2021-04-14 17:47:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][390/1251] eta 0:04:02 lr 0.000366 time 0.2683 (0.2820) loss 5.4296 (5.4366) grad_norm 2.1349 (2.1804) [2021-04-14 17:47:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][400/1251] eta 0:03:59 lr 0.000367 time 0.2831 (0.2818) loss 5.3303 (5.4316) grad_norm 2.2579 (2.1767) [2021-04-14 17:47:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][410/1251] eta 0:03:56 lr 0.000367 time 0.2629 (0.2815) loss 5.6466 (5.4342) grad_norm 1.6439 (2.1735) [2021-04-14 17:47:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][420/1251] eta 0:03:53 lr 0.000367 time 0.2916 (0.2814) loss 5.4566 (5.4271) grad_norm 2.1625 (2.1720) [2021-04-14 17:47:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][430/1251] eta 0:03:50 lr 0.000368 time 0.2927 (0.2811) loss 5.8247 (5.4241) grad_norm 1.8347 (2.1753) [2021-04-14 17:47:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][440/1251] eta 0:03:47 lr 0.000368 time 0.2694 (0.2809) loss 5.9740 (5.4274) grad_norm 2.1780 (2.1775) [2021-04-14 17:47:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][450/1251] eta 0:03:44 lr 0.000369 time 0.2797 (0.2807) loss 5.8599 (5.4240) grad_norm 2.7845 (2.1782) [2021-04-14 17:47:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][460/1251] eta 0:03:41 lr 0.000369 time 0.2504 (0.2805) loss 5.5112 (5.4281) grad_norm 1.7312 (2.1743) [2021-04-14 17:47:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][470/1251] eta 0:03:39 lr 0.000369 time 0.2709 (0.2805) loss 5.6533 (5.4289) grad_norm 2.3289 (2.1752) [2021-04-14 17:47:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][480/1251] eta 0:03:36 lr 0.000370 time 0.2828 (0.2803) loss 5.8640 (5.4246) grad_norm 2.1440 (2.1744) [2021-04-14 17:47:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][490/1251] eta 0:03:33 lr 0.000370 time 0.2660 (0.2801) loss 5.1263 (5.4196) grad_norm 2.5005 (2.1706) [2021-04-14 17:47:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][500/1251] eta 0:03:30 lr 0.000371 time 0.2832 (0.2800) loss 5.0798 (5.4195) grad_norm 1.9550 (2.1677) [2021-04-14 17:47:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][510/1251] eta 0:03:27 lr 0.000371 time 0.2776 (0.2798) loss 5.9926 (5.4215) grad_norm 1.8900 (2.1668) [2021-04-14 17:47:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][520/1251] eta 0:03:24 lr 0.000371 time 0.2736 (0.2797) loss 6.0047 (5.4213) grad_norm 2.9305 (2.1698) [2021-04-14 17:47:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][530/1251] eta 0:03:21 lr 0.000372 time 0.2749 (0.2795) loss 5.2386 (5.4216) grad_norm 2.2164 (2.1728) [2021-04-14 17:47:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][540/1251] eta 0:03:18 lr 0.000372 time 0.2438 (0.2794) loss 5.1828 (5.4180) grad_norm 1.7776 (2.1722) [2021-04-14 17:47:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][550/1251] eta 0:03:15 lr 0.000373 time 0.2825 (0.2793) loss 5.3806 (5.4193) grad_norm 1.5641 (2.1674) [2021-04-14 17:48:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][560/1251] eta 0:03:12 lr 0.000373 time 0.2625 (0.2792) loss 5.6584 (5.4161) grad_norm 2.2715 (2.1670) [2021-04-14 17:48:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][570/1251] eta 0:03:10 lr 0.000373 time 0.2686 (0.2793) loss 4.7351 (5.4115) grad_norm 1.8873 (2.1672) [2021-04-14 17:48:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][580/1251] eta 0:03:07 lr 0.000374 time 0.2632 (0.2794) loss 6.0643 (5.4090) grad_norm 3.7729 (2.1677) [2021-04-14 17:48:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][590/1251] eta 0:03:04 lr 0.000374 time 0.2903 (0.2793) loss 5.6657 (5.4067) grad_norm 3.1628 (2.1771) [2021-04-14 17:48:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][600/1251] eta 0:03:01 lr 0.000375 time 0.2489 (0.2791) loss 4.9552 (5.4070) grad_norm 2.1296 (2.1819) [2021-04-14 17:48:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][610/1251] eta 0:02:58 lr 0.000375 time 0.2652 (0.2791) loss 5.5670 (5.4128) grad_norm 2.2018 (2.1802) [2021-04-14 17:48:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][620/1251] eta 0:02:56 lr 0.000375 time 0.3000 (0.2790) loss 5.5006 (5.4093) grad_norm 1.7840 (2.1817) [2021-04-14 17:48:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][630/1251] eta 0:02:53 lr 0.000376 time 0.2529 (0.2789) loss 5.1546 (5.4087) grad_norm 1.8521 (2.1809) [2021-04-14 17:48:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][640/1251] eta 0:02:50 lr 0.000376 time 0.2674 (0.2788) loss 5.1289 (5.4114) grad_norm 1.5818 (2.1797) [2021-04-14 17:48:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][650/1251] eta 0:02:47 lr 0.000377 time 0.2715 (0.2787) loss 5.8559 (5.4118) grad_norm 2.5197 (2.1834) [2021-04-14 17:48:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][660/1251] eta 0:02:44 lr 0.000377 time 0.2654 (0.2786) loss 5.6789 (5.4141) grad_norm 1.8825 (2.1841) [2021-04-14 17:48:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][670/1251] eta 0:02:41 lr 0.000377 time 0.2725 (0.2785) loss 5.4747 (5.4177) grad_norm 2.4639 (2.1856) [2021-04-14 17:48:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][680/1251] eta 0:02:38 lr 0.000378 time 0.2687 (0.2784) loss 5.5816 (5.4187) grad_norm 2.4956 (2.1839) [2021-04-14 17:48:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][690/1251] eta 0:02:36 lr 0.000378 time 0.2461 (0.2783) loss 5.8519 (5.4207) grad_norm 1.7937 (inf) [2021-04-14 17:48:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][700/1251] eta 0:02:33 lr 0.000379 time 0.2620 (0.2782) loss 5.1447 (5.4174) grad_norm 2.0380 (inf) [2021-04-14 17:48:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][710/1251] eta 0:02:30 lr 0.000379 time 0.2706 (0.2782) loss 4.7597 (5.4186) grad_norm 2.7406 (inf) [2021-04-14 17:48:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][720/1251] eta 0:02:27 lr 0.000379 time 0.2732 (0.2782) loss 4.3441 (5.4193) grad_norm 1.7792 (inf) [2021-04-14 17:48:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][730/1251] eta 0:02:24 lr 0.000380 time 0.2715 (0.2782) loss 5.8785 (5.4230) grad_norm 2.2916 (inf) [2021-04-14 17:48:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][740/1251] eta 0:02:22 lr 0.000380 time 0.2565 (0.2780) loss 4.6109 (5.4225) grad_norm 1.5814 (inf) [2021-04-14 17:48:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][750/1251] eta 0:02:19 lr 0.000381 time 0.2913 (0.2779) loss 5.1249 (5.4159) grad_norm 1.9057 (inf) [2021-04-14 17:48:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][760/1251] eta 0:02:16 lr 0.000381 time 0.2797 (0.2778) loss 4.9563 (5.4130) grad_norm 2.0773 (inf) [2021-04-14 17:48:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][770/1251] eta 0:02:13 lr 0.000381 time 0.3126 (0.2778) loss 6.0567 (5.4144) grad_norm 2.0242 (inf) [2021-04-14 17:49:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][780/1251] eta 0:02:10 lr 0.000382 time 0.2844 (0.2777) loss 5.1999 (5.4159) grad_norm 2.1166 (inf) [2021-04-14 17:49:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][790/1251] eta 0:02:07 lr 0.000382 time 0.2640 (0.2776) loss 5.2244 (5.4155) grad_norm 1.9615 (inf) [2021-04-14 17:49:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][800/1251] eta 0:02:05 lr 0.000383 time 0.2667 (0.2775) loss 6.0814 (5.4151) grad_norm 1.8982 (inf) [2021-04-14 17:49:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][810/1251] eta 0:02:02 lr 0.000383 time 0.2612 (0.2774) loss 4.7338 (5.4131) grad_norm 2.4798 (inf) [2021-04-14 17:49:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][820/1251] eta 0:01:59 lr 0.000383 time 0.2709 (0.2774) loss 4.4455 (5.4125) grad_norm 1.7080 (inf) [2021-04-14 17:49:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][830/1251] eta 0:01:56 lr 0.000384 time 0.2626 (0.2773) loss 4.9468 (5.4121) grad_norm 1.7159 (inf) [2021-04-14 17:49:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][840/1251] eta 0:01:53 lr 0.000384 time 0.2738 (0.2772) loss 5.2121 (5.4086) grad_norm 1.8340 (inf) [2021-04-14 17:49:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][850/1251] eta 0:01:51 lr 0.000385 time 0.2665 (0.2771) loss 5.2028 (5.4088) grad_norm 1.6190 (inf) [2021-04-14 17:49:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][860/1251] eta 0:01:48 lr 0.000385 time 0.2917 (0.2770) loss 5.4565 (5.4070) grad_norm 1.9512 (inf) [2021-04-14 17:49:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][870/1251] eta 0:01:45 lr 0.000385 time 0.2514 (0.2770) loss 4.2190 (5.4087) grad_norm 2.2990 (inf) [2021-04-14 17:49:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][880/1251] eta 0:01:42 lr 0.000386 time 0.2602 (0.2769) loss 5.3112 (5.4072) grad_norm 2.5696 (inf) [2021-04-14 17:49:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][890/1251] eta 0:01:39 lr 0.000386 time 0.2664 (0.2768) loss 5.7281 (5.4037) grad_norm 1.9772 (inf) [2021-04-14 17:49:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][900/1251] eta 0:01:37 lr 0.000387 time 0.2791 (0.2768) loss 4.7561 (5.4043) grad_norm 2.0151 (inf) [2021-04-14 17:49:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][910/1251] eta 0:01:34 lr 0.000387 time 0.2826 (0.2767) loss 5.9593 (5.4074) grad_norm 2.6232 (inf) [2021-04-14 17:49:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][920/1251] eta 0:01:31 lr 0.000387 time 0.2716 (0.2766) loss 5.4118 (5.4086) grad_norm 2.1262 (inf) [2021-04-14 17:49:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][930/1251] eta 0:01:28 lr 0.000388 time 0.2735 (0.2766) loss 5.7916 (5.4112) grad_norm 2.2806 (inf) [2021-04-14 17:49:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][940/1251] eta 0:01:26 lr 0.000388 time 0.2637 (0.2766) loss 5.7188 (5.4116) grad_norm 2.2608 (inf) [2021-04-14 17:49:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][950/1251] eta 0:01:23 lr 0.000389 time 0.2724 (0.2765) loss 5.5190 (5.4137) grad_norm 1.7773 (inf) [2021-04-14 17:49:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][960/1251] eta 0:01:20 lr 0.000389 time 0.2873 (0.2765) loss 5.7872 (5.4148) grad_norm 1.7837 (inf) [2021-04-14 17:49:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][970/1251] eta 0:01:17 lr 0.000389 time 0.2970 (0.2764) loss 5.0289 (5.4119) grad_norm 2.3326 (inf) [2021-04-14 17:49:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][980/1251] eta 0:01:14 lr 0.000390 time 0.2500 (0.2764) loss 5.7296 (5.4116) grad_norm 3.0608 (inf) [2021-04-14 17:49:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][990/1251] eta 0:01:12 lr 0.000390 time 0.2833 (0.2763) loss 5.1848 (5.4112) grad_norm 1.7098 (inf) [2021-04-14 17:50:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][1000/1251] eta 0:01:09 lr 0.000391 time 0.2917 (0.2763) loss 5.7795 (5.4098) grad_norm 1.5781 (inf) [2021-04-14 17:50:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][1010/1251] eta 0:01:06 lr 0.000391 time 0.2609 (0.2763) loss 5.6779 (5.4081) grad_norm 1.7768 (inf) [2021-04-14 17:50:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][1020/1251] eta 0:01:03 lr 0.000391 time 0.2661 (0.2762) loss 5.0602 (5.4070) grad_norm 1.8446 (inf) [2021-04-14 17:50:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][1030/1251] eta 0:01:01 lr 0.000392 time 0.2814 (0.2762) loss 5.4654 (5.4072) grad_norm 2.2718 (inf) [2021-04-14 17:50:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][1040/1251] eta 0:00:58 lr 0.000392 time 0.2732 (0.2762) loss 5.0575 (5.4072) grad_norm 3.6216 (inf) [2021-04-14 17:50:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][1050/1251] eta 0:00:55 lr 0.000393 time 0.2916 (0.2761) loss 5.2167 (5.4060) grad_norm 1.9396 (inf) [2021-04-14 17:50:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][1060/1251] eta 0:00:52 lr 0.000393 time 0.2779 (0.2761) loss 4.5010 (5.4039) grad_norm 2.1058 (inf) [2021-04-14 17:50:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][1070/1251] eta 0:00:49 lr 0.000393 time 0.2566 (0.2760) loss 5.8218 (5.4042) grad_norm 2.0895 (inf) [2021-04-14 17:50:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][1080/1251] eta 0:00:47 lr 0.000394 time 0.2589 (0.2760) loss 5.6184 (5.4032) grad_norm 1.6514 (inf) [2021-04-14 17:50:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][1090/1251] eta 0:00:44 lr 0.000394 time 0.2781 (0.2760) loss 5.7730 (5.4021) grad_norm 2.2310 (inf) [2021-04-14 17:50:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][1100/1251] eta 0:00:41 lr 0.000395 time 0.2514 (0.2759) loss 5.5709 (5.3990) grad_norm 2.1742 (inf) [2021-04-14 17:50:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][1110/1251] eta 0:00:38 lr 0.000395 time 0.2546 (0.2759) loss 5.1835 (5.3982) grad_norm 2.2431 (inf) [2021-04-14 17:50:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][1120/1251] eta 0:00:36 lr 0.000395 time 0.2971 (0.2759) loss 5.4491 (5.3974) grad_norm 1.8313 (inf) [2021-04-14 17:50:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][1130/1251] eta 0:00:33 lr 0.000396 time 0.3084 (0.2759) loss 5.6218 (5.3970) grad_norm 1.9034 (inf) [2021-04-14 17:50:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][1140/1251] eta 0:00:30 lr 0.000396 time 0.2772 (0.2759) loss 5.2630 (5.3967) grad_norm 2.5727 (inf) [2021-04-14 17:50:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][1150/1251] eta 0:00:27 lr 0.000397 time 0.2922 (0.2760) loss 5.3657 (5.3965) grad_norm 2.2724 (inf) [2021-04-14 17:50:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][1160/1251] eta 0:00:25 lr 0.000397 time 0.2815 (0.2761) loss 5.6108 (5.3979) grad_norm 1.9410 (inf) [2021-04-14 17:50:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][1170/1251] eta 0:00:22 lr 0.000397 time 0.2762 (0.2761) loss 5.4939 (5.3971) grad_norm 2.1997 (inf) [2021-04-14 17:50:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][1180/1251] eta 0:00:19 lr 0.000398 time 0.2718 (0.2761) loss 4.3099 (5.3955) grad_norm 2.2162 (inf) [2021-04-14 17:50:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][1190/1251] eta 0:00:16 lr 0.000398 time 0.2806 (0.2761) loss 4.8621 (5.3962) grad_norm 1.8069 (inf) [2021-04-14 17:50:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][1200/1251] eta 0:00:14 lr 0.000399 time 0.2900 (0.2761) loss 3.9866 (5.3942) grad_norm 2.2298 (inf) [2021-04-14 17:50:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][1210/1251] eta 0:00:11 lr 0.000399 time 0.2753 (0.2761) loss 5.8000 (5.3938) grad_norm 2.1588 (inf) [2021-04-14 17:51:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][1220/1251] eta 0:00:08 lr 0.000399 time 0.2645 (0.2760) loss 3.9248 (5.3923) grad_norm 2.0326 (inf) [2021-04-14 17:51:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][1230/1251] eta 0:00:05 lr 0.000400 time 0.2726 (0.2760) loss 5.5441 (5.3917) grad_norm 2.1140 (inf) [2021-04-14 17:51:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][1240/1251] eta 0:00:03 lr 0.000400 time 0.2476 (0.2759) loss 5.1708 (5.3910) grad_norm 2.1940 (inf) [2021-04-14 17:51:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [7/300][1250/1251] eta 0:00:00 lr 0.000401 time 0.2583 (0.2757) loss 5.0205 (5.3909) grad_norm 1.6484 (inf) [2021-04-14 17:51:11 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 7 training takes 0:05:46 [2021-04-14 17:51:11 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_7.pth saving...... [2021-04-14 17:51:35 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_7.pth saved !!! [2021-04-14 17:51:36 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.297 (1.297) Loss 2.9416 (2.9416) Acc@1 39.355 (39.355) Acc@5 64.551 (64.551) [2021-04-14 17:51:37 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.105 (0.246) Loss 2.9971 (3.0166) Acc@1 38.281 (37.234) Acc@5 61.816 (62.305) [2021-04-14 17:51:40 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.124 (0.242) Loss 3.0123 (3.0066) Acc@1 35.645 (37.249) Acc@5 65.137 (62.891) [2021-04-14 17:51:42 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.142 (0.224) Loss 3.1512 (3.0140) Acc@1 35.742 (37.125) Acc@5 59.082 (62.749) [2021-04-14 17:51:43 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.214 (0.214) Loss 3.0618 (3.0151) Acc@1 37.012 (37.031) Acc@5 61.523 (62.662) [2021-04-14 17:51:45 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 36.952 Acc@5 62.636 [2021-04-14 17:51:45 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 37.0% [2021-04-14 17:51:45 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 36.95% [2021-04-14 17:51:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][0/1251] eta 1:07:10 lr 0.000401 time 3.2221 (3.2221) loss 5.7297 (5.7297) grad_norm 2.5938 (2.5938) [2021-04-14 17:51:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][10/1251] eta 0:11:49 lr 0.000401 time 0.2755 (0.5717) loss 4.9838 (5.4880) grad_norm 2.2189 (2.2991) [2021-04-14 17:51:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][20/1251] eta 0:08:53 lr 0.000401 time 0.2854 (0.4334) loss 6.1880 (5.5541) grad_norm 2.3369 (2.2175) [2021-04-14 17:51:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][30/1251] eta 0:07:44 lr 0.000402 time 0.2714 (0.3805) loss 5.6874 (5.5652) grad_norm 1.6257 (2.1571) [2021-04-14 17:52:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][40/1251] eta 0:07:08 lr 0.000402 time 0.2755 (0.3542) loss 5.2377 (5.4560) grad_norm 1.7361 (2.1744) [2021-04-14 17:52:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][50/1251] eta 0:06:46 lr 0.000403 time 0.2715 (0.3382) loss 5.7253 (5.4966) grad_norm 1.8844 (2.1304) [2021-04-14 17:52:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][60/1251] eta 0:06:28 lr 0.000403 time 0.2537 (0.3262) loss 5.0300 (5.4708) grad_norm 2.0474 (2.0854) [2021-04-14 17:52:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][70/1251] eta 0:06:16 lr 0.000403 time 0.2718 (0.3185) loss 5.2148 (5.4674) grad_norm 1.7502 (2.0939) [2021-04-14 17:52:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][80/1251] eta 0:06:06 lr 0.000404 time 0.2619 (0.3129) loss 4.4645 (5.4462) grad_norm 2.8813 (2.1125) [2021-04-14 17:52:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][90/1251] eta 0:05:57 lr 0.000404 time 0.2591 (0.3083) loss 5.5528 (5.4418) grad_norm 1.8762 (2.1568) [2021-04-14 17:52:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][100/1251] eta 0:05:50 lr 0.000405 time 0.2624 (0.3047) loss 4.7888 (5.4318) grad_norm 1.5441 (2.1299) [2021-04-14 17:52:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][110/1251] eta 0:05:44 lr 0.000405 time 0.2620 (0.3017) loss 5.9665 (5.4394) grad_norm 1.9368 (2.1301) [2021-04-14 17:52:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][120/1251] eta 0:05:38 lr 0.000405 time 0.2985 (0.2992) loss 5.0180 (5.4336) grad_norm 1.8932 (2.1093) [2021-04-14 17:52:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][130/1251] eta 0:05:32 lr 0.000406 time 0.2952 (0.2969) loss 5.8288 (5.4297) grad_norm 2.2931 (2.1121) [2021-04-14 17:52:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][140/1251] eta 0:05:27 lr 0.000406 time 0.2774 (0.2952) loss 4.0987 (5.4034) grad_norm 1.9350 (2.0974) [2021-04-14 17:52:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][150/1251] eta 0:05:23 lr 0.000407 time 0.2814 (0.2940) loss 4.8975 (5.3985) grad_norm 2.2677 (2.0882) [2021-04-14 17:52:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][160/1251] eta 0:05:19 lr 0.000407 time 0.2761 (0.2927) loss 5.8495 (5.4022) grad_norm 1.9808 (2.0907) [2021-04-14 17:52:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][170/1251] eta 0:05:14 lr 0.000407 time 0.2669 (0.2913) loss 5.8828 (5.4100) grad_norm 2.2581 (2.0951) [2021-04-14 17:52:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][180/1251] eta 0:05:11 lr 0.000408 time 0.2730 (0.2905) loss 5.5561 (5.3991) grad_norm 1.9121 (2.0775) [2021-04-14 17:52:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][190/1251] eta 0:05:07 lr 0.000408 time 0.2698 (0.2895) loss 5.7124 (5.4020) grad_norm 2.4308 (2.0702) [2021-04-14 17:52:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][200/1251] eta 0:05:03 lr 0.000409 time 0.2601 (0.2889) loss 4.5050 (5.3998) grad_norm 2.1544 (2.0875) [2021-04-14 17:52:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][210/1251] eta 0:05:00 lr 0.000409 time 0.2593 (0.2885) loss 4.5417 (5.3947) grad_norm 1.9600 (2.0994) [2021-04-14 17:52:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][220/1251] eta 0:04:56 lr 0.000409 time 0.2736 (0.2879) loss 4.4377 (5.3800) grad_norm 2.0550 (2.0902) [2021-04-14 17:52:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][230/1251] eta 0:04:53 lr 0.000410 time 0.2832 (0.2872) loss 4.2239 (5.3753) grad_norm 1.9599 (2.0896) [2021-04-14 17:52:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][240/1251] eta 0:04:49 lr 0.000410 time 0.2606 (0.2866) loss 5.9377 (5.3772) grad_norm 2.3416 (2.0934) [2021-04-14 17:52:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][250/1251] eta 0:04:46 lr 0.000411 time 0.2881 (0.2862) loss 5.7406 (5.3844) grad_norm 2.4141 (2.1014) [2021-04-14 17:53:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][260/1251] eta 0:04:43 lr 0.000411 time 0.2627 (0.2856) loss 5.5700 (5.3864) grad_norm 2.0974 (2.0978) [2021-04-14 17:53:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][270/1251] eta 0:04:39 lr 0.000411 time 0.2950 (0.2852) loss 6.1092 (5.3744) grad_norm 1.7799 (2.0914) [2021-04-14 17:53:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][280/1251] eta 0:04:36 lr 0.000412 time 0.2687 (0.2847) loss 5.8273 (5.3733) grad_norm 1.5894 (2.0805) [2021-04-14 17:53:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][290/1251] eta 0:04:33 lr 0.000412 time 0.2812 (0.2843) loss 5.8223 (5.3678) grad_norm 2.1941 (2.0790) [2021-04-14 17:53:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][300/1251] eta 0:04:29 lr 0.000413 time 0.2813 (0.2838) loss 4.7453 (5.3752) grad_norm 2.7149 (2.0871) [2021-04-14 17:53:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][310/1251] eta 0:04:26 lr 0.000413 time 0.2697 (0.2833) loss 5.2617 (5.3630) grad_norm 1.6194 (2.0902) [2021-04-14 17:53:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][320/1251] eta 0:04:23 lr 0.000413 time 0.2652 (0.2831) loss 5.5567 (5.3597) grad_norm 2.0545 (2.0881) [2021-04-14 17:53:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][330/1251] eta 0:04:20 lr 0.000414 time 0.2891 (0.2828) loss 5.8933 (5.3588) grad_norm 2.6726 (2.0979) [2021-04-14 17:53:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][340/1251] eta 0:04:17 lr 0.000414 time 0.2572 (0.2825) loss 4.7391 (5.3582) grad_norm 2.0831 (2.0986) [2021-04-14 17:53:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][350/1251] eta 0:04:14 lr 0.000415 time 0.4331 (0.2828) loss 5.3755 (5.3566) grad_norm 1.9393 (2.0927) [2021-04-14 17:53:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][360/1251] eta 0:04:11 lr 0.000415 time 0.2838 (0.2825) loss 5.6060 (5.3587) grad_norm 1.9331 (2.1029) [2021-04-14 17:53:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][370/1251] eta 0:04:09 lr 0.000415 time 0.2790 (0.2827) loss 4.4502 (5.3484) grad_norm 2.2765 (2.0993) [2021-04-14 17:53:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][380/1251] eta 0:04:06 lr 0.000416 time 0.2790 (0.2826) loss 5.2491 (5.3426) grad_norm 1.7435 (2.0936) [2021-04-14 17:53:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][390/1251] eta 0:04:03 lr 0.000416 time 0.2712 (0.2823) loss 4.9202 (5.3398) grad_norm 1.9905 (2.1003) [2021-04-14 17:53:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][400/1251] eta 0:04:00 lr 0.000417 time 0.2827 (0.2820) loss 5.5510 (5.3371) grad_norm 1.8065 (2.0961) [2021-04-14 17:53:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][410/1251] eta 0:03:57 lr 0.000417 time 0.2717 (0.2819) loss 4.9958 (5.3377) grad_norm 1.7370 (2.0977) [2021-04-14 17:53:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][420/1251] eta 0:03:54 lr 0.000417 time 0.2546 (0.2816) loss 4.8223 (5.3305) grad_norm 2.4581 (2.0977) [2021-04-14 17:53:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][430/1251] eta 0:03:51 lr 0.000418 time 0.2810 (0.2814) loss 4.2619 (5.3257) grad_norm 1.8697 (2.0963) [2021-04-14 17:53:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][440/1251] eta 0:03:48 lr 0.000418 time 0.2646 (0.2812) loss 5.0497 (5.3285) grad_norm 2.0742 (2.0962) [2021-04-14 17:53:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][450/1251] eta 0:03:45 lr 0.000419 time 0.2576 (0.2810) loss 4.9880 (5.3259) grad_norm 1.8458 (2.0924) [2021-04-14 17:53:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][460/1251] eta 0:03:42 lr 0.000419 time 0.2647 (0.2807) loss 5.4965 (5.3237) grad_norm 2.0252 (2.0919) [2021-04-14 17:53:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][470/1251] eta 0:03:39 lr 0.000419 time 0.2597 (0.2806) loss 5.3656 (5.3255) grad_norm 2.1235 (2.0892) [2021-04-14 17:54:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][480/1251] eta 0:03:36 lr 0.000420 time 0.2730 (0.2804) loss 4.5352 (5.3230) grad_norm 1.7328 (2.0930) [2021-04-14 17:54:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][490/1251] eta 0:03:33 lr 0.000420 time 0.2519 (0.2802) loss 5.8641 (5.3242) grad_norm 1.9134 (2.0887) [2021-04-14 17:54:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][500/1251] eta 0:03:30 lr 0.000421 time 0.2789 (0.2800) loss 5.4397 (5.3275) grad_norm 2.2088 (2.0836) [2021-04-14 17:54:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][510/1251] eta 0:03:27 lr 0.000421 time 0.2590 (0.2798) loss 5.5953 (5.3284) grad_norm 2.0018 (2.0829) [2021-04-14 17:54:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][520/1251] eta 0:03:24 lr 0.000421 time 0.2547 (0.2797) loss 5.6133 (5.3290) grad_norm 2.4371 (2.0844) [2021-04-14 17:54:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][530/1251] eta 0:03:21 lr 0.000422 time 0.3020 (0.2796) loss 4.2845 (5.3229) grad_norm 2.1732 (2.0823) [2021-04-14 17:54:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][540/1251] eta 0:03:18 lr 0.000422 time 0.2566 (0.2794) loss 5.7136 (5.3259) grad_norm 2.0347 (2.0787) [2021-04-14 17:54:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][550/1251] eta 0:03:15 lr 0.000423 time 0.2915 (0.2792) loss 4.4324 (5.3231) grad_norm 1.8575 (2.0737) [2021-04-14 17:54:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][560/1251] eta 0:03:12 lr 0.000423 time 0.2675 (0.2792) loss 5.7813 (5.3157) grad_norm 1.8170 (2.0737) [2021-04-14 17:54:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][570/1251] eta 0:03:10 lr 0.000423 time 0.2898 (0.2791) loss 6.1707 (5.3169) grad_norm 2.0300 (2.0726) [2021-04-14 17:54:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][580/1251] eta 0:03:07 lr 0.000424 time 0.2725 (0.2789) loss 4.3123 (5.3166) grad_norm 1.9134 (2.0734) [2021-04-14 17:54:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][590/1251] eta 0:03:04 lr 0.000424 time 0.2594 (0.2788) loss 5.0415 (5.3105) grad_norm 2.4802 (2.0730) [2021-04-14 17:54:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][600/1251] eta 0:03:01 lr 0.000425 time 0.2855 (0.2788) loss 5.7118 (5.3100) grad_norm 2.5659 (2.0699) [2021-04-14 17:54:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][610/1251] eta 0:02:58 lr 0.000425 time 0.2937 (0.2787) loss 5.0116 (5.3075) grad_norm 2.0538 (2.0760) [2021-04-14 17:54:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][620/1251] eta 0:02:55 lr 0.000425 time 0.2842 (0.2786) loss 5.3665 (5.3046) grad_norm 1.6581 (2.0744) [2021-04-14 17:54:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][630/1251] eta 0:02:52 lr 0.000426 time 0.2709 (0.2785) loss 4.8586 (5.3097) grad_norm 1.6803 (2.0733) [2021-04-14 17:54:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][640/1251] eta 0:02:50 lr 0.000426 time 0.2746 (0.2784) loss 5.6792 (5.3069) grad_norm 2.0561 (2.0685) [2021-04-14 17:54:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][650/1251] eta 0:02:47 lr 0.000427 time 0.2723 (0.2783) loss 5.2583 (5.3040) grad_norm 2.0115 (2.0674) [2021-04-14 17:54:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][660/1251] eta 0:02:44 lr 0.000427 time 0.2839 (0.2782) loss 5.5693 (5.3049) grad_norm 1.4662 (2.0696) [2021-04-14 17:54:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][670/1251] eta 0:02:41 lr 0.000427 time 0.2697 (0.2781) loss 5.4490 (5.3047) grad_norm 1.7239 (2.0662) [2021-04-14 17:54:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][680/1251] eta 0:02:38 lr 0.000428 time 0.2644 (0.2780) loss 5.6960 (5.3091) grad_norm 1.8010 (2.0664) [2021-04-14 17:54:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][690/1251] eta 0:02:35 lr 0.000428 time 0.2688 (0.2780) loss 5.6878 (5.3102) grad_norm 1.9809 (2.0685) [2021-04-14 17:55:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][700/1251] eta 0:02:33 lr 0.000429 time 0.2834 (0.2779) loss 4.5082 (5.3100) grad_norm 2.4127 (2.0664) [2021-04-14 17:55:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][710/1251] eta 0:02:30 lr 0.000429 time 0.2819 (0.2779) loss 4.3924 (5.3088) grad_norm 1.6830 (2.0660) [2021-04-14 17:55:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][720/1251] eta 0:02:27 lr 0.000429 time 0.2791 (0.2778) loss 5.5354 (5.3086) grad_norm 2.0611 (2.0652) [2021-04-14 17:55:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][730/1251] eta 0:02:24 lr 0.000430 time 0.2812 (0.2777) loss 6.2205 (5.3101) grad_norm 1.3977 (2.0615) [2021-04-14 17:55:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][740/1251] eta 0:02:21 lr 0.000430 time 0.2686 (0.2776) loss 5.0570 (5.3105) grad_norm 2.0638 (2.0583) [2021-04-14 17:55:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][750/1251] eta 0:02:19 lr 0.000431 time 0.2574 (0.2775) loss 4.7468 (5.3114) grad_norm 1.7327 (2.0571) [2021-04-14 17:55:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][760/1251] eta 0:02:16 lr 0.000431 time 0.2515 (0.2774) loss 5.2333 (5.3111) grad_norm 2.0800 (2.0570) [2021-04-14 17:55:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][770/1251] eta 0:02:13 lr 0.000431 time 0.2818 (0.2774) loss 5.3862 (5.3099) grad_norm 2.1959 (2.0592) [2021-04-14 17:55:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][780/1251] eta 0:02:10 lr 0.000432 time 0.2563 (0.2773) loss 5.4941 (5.3096) grad_norm 1.8531 (2.0611) [2021-04-14 17:55:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][790/1251] eta 0:02:07 lr 0.000432 time 0.2661 (0.2773) loss 4.4884 (5.3052) grad_norm 1.5709 (2.0587) [2021-04-14 17:55:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][800/1251] eta 0:02:05 lr 0.000433 time 0.2621 (0.2772) loss 5.9210 (5.3049) grad_norm 1.7919 (2.0568) [2021-04-14 17:55:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][810/1251] eta 0:02:02 lr 0.000433 time 0.2713 (0.2772) loss 5.3841 (5.3056) grad_norm 2.3166 (2.0557) [2021-04-14 17:55:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][820/1251] eta 0:01:59 lr 0.000433 time 0.2875 (0.2771) loss 4.8196 (5.3019) grad_norm 1.7527 (2.0539) [2021-04-14 17:55:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][830/1251] eta 0:01:56 lr 0.000434 time 0.2855 (0.2770) loss 5.2878 (5.3029) grad_norm 2.5524 (2.0598) [2021-04-14 17:55:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][840/1251] eta 0:01:53 lr 0.000434 time 0.2649 (0.2769) loss 5.6847 (5.3019) grad_norm 1.7862 (2.0578) [2021-04-14 17:55:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][850/1251] eta 0:01:51 lr 0.000435 time 0.2452 (0.2768) loss 5.4233 (5.3014) grad_norm 3.8147 (2.0592) [2021-04-14 17:55:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][860/1251] eta 0:01:48 lr 0.000435 time 0.2713 (0.2768) loss 4.9145 (5.3033) grad_norm 2.5562 (2.0620) [2021-04-14 17:55:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][870/1251] eta 0:01:45 lr 0.000435 time 0.2755 (0.2767) loss 5.9371 (5.3052) grad_norm 2.2033 (2.0630) [2021-04-14 17:55:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][880/1251] eta 0:01:42 lr 0.000436 time 0.2751 (0.2766) loss 5.4784 (5.3039) grad_norm 2.5007 (2.0639) [2021-04-14 17:55:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][890/1251] eta 0:01:39 lr 0.000436 time 0.2654 (0.2766) loss 4.9351 (5.3017) grad_norm 1.8393 (2.0663) [2021-04-14 17:55:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][900/1251] eta 0:01:37 lr 0.000437 time 0.2728 (0.2766) loss 5.1297 (5.3022) grad_norm 1.9226 (2.0674) [2021-04-14 17:55:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][910/1251] eta 0:01:34 lr 0.000437 time 0.2817 (0.2766) loss 5.5924 (5.2999) grad_norm 3.0893 (2.0679) [2021-04-14 17:56:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][920/1251] eta 0:01:31 lr 0.000437 time 0.2635 (0.2766) loss 4.4467 (5.2989) grad_norm 1.7956 (2.0694) [2021-04-14 17:56:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][930/1251] eta 0:01:28 lr 0.000438 time 0.2594 (0.2767) loss 4.9465 (5.2976) grad_norm 2.3570 (2.0678) [2021-04-14 17:56:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][940/1251] eta 0:01:26 lr 0.000438 time 0.2778 (0.2769) loss 5.1382 (5.2983) grad_norm 2.1400 (2.0662) [2021-04-14 17:56:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][950/1251] eta 0:01:23 lr 0.000439 time 0.2820 (0.2768) loss 4.1267 (5.2968) grad_norm 1.7526 (2.0634) [2021-04-14 17:56:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][960/1251] eta 0:01:20 lr 0.000439 time 0.2498 (0.2767) loss 5.3379 (5.2984) grad_norm 1.8011 (2.0595) [2021-04-14 17:56:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][970/1251] eta 0:01:17 lr 0.000439 time 0.2721 (0.2767) loss 5.8044 (5.2974) grad_norm 2.0218 (2.0601) [2021-04-14 17:56:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][980/1251] eta 0:01:14 lr 0.000440 time 0.2955 (0.2767) loss 5.3741 (5.2966) grad_norm 2.2310 (2.0629) [2021-04-14 17:56:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][990/1251] eta 0:01:12 lr 0.000440 time 0.2591 (0.2768) loss 4.5226 (5.2965) grad_norm 1.6602 (2.0604) [2021-04-14 17:56:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][1000/1251] eta 0:01:09 lr 0.000441 time 0.2713 (0.2768) loss 5.4183 (5.2981) grad_norm 1.8245 (2.0582) [2021-04-14 17:56:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][1010/1251] eta 0:01:06 lr 0.000441 time 0.2741 (0.2768) loss 5.4449 (5.2994) grad_norm 1.7842 (2.0581) [2021-04-14 17:56:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][1020/1251] eta 0:01:03 lr 0.000441 time 0.2529 (0.2767) loss 4.7963 (5.2990) grad_norm 2.6367 (2.0587) [2021-04-14 17:56:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][1030/1251] eta 0:01:01 lr 0.000442 time 0.2902 (0.2766) loss 5.3239 (5.2984) grad_norm 1.7831 (2.0561) [2021-04-14 17:56:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][1040/1251] eta 0:00:58 lr 0.000442 time 0.2670 (0.2766) loss 4.6455 (5.2997) grad_norm 2.0288 (2.0535) [2021-04-14 17:56:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][1050/1251] eta 0:00:55 lr 0.000443 time 0.2576 (0.2765) loss 5.9346 (5.2983) grad_norm 2.0146 (2.0554) [2021-04-14 17:56:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][1060/1251] eta 0:00:52 lr 0.000443 time 0.2644 (0.2765) loss 5.5572 (5.3004) grad_norm 1.8952 (2.0567) [2021-04-14 17:56:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][1070/1251] eta 0:00:50 lr 0.000443 time 0.2428 (0.2764) loss 5.0848 (5.2990) grad_norm 2.1028 (2.0597) [2021-04-14 17:56:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][1080/1251] eta 0:00:47 lr 0.000444 time 0.2867 (0.2764) loss 4.8879 (5.2974) grad_norm 1.7723 (2.0580) [2021-04-14 17:56:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][1090/1251] eta 0:00:44 lr 0.000444 time 0.2652 (0.2764) loss 5.5673 (5.2994) grad_norm 2.2109 (2.0586) [2021-04-14 17:56:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][1100/1251] eta 0:00:41 lr 0.000445 time 0.3025 (0.2764) loss 5.4431 (5.2960) grad_norm 2.0636 (2.0622) [2021-04-14 17:56:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][1110/1251] eta 0:00:38 lr 0.000445 time 0.2888 (0.2763) loss 5.4838 (5.2959) grad_norm 2.2473 (2.0606) [2021-04-14 17:56:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][1120/1251] eta 0:00:36 lr 0.000445 time 0.2858 (0.2763) loss 5.4709 (5.2965) grad_norm 1.8548 (2.0583) [2021-04-14 17:56:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][1130/1251] eta 0:00:33 lr 0.000446 time 0.2853 (0.2762) loss 5.4611 (5.2955) grad_norm 1.9357 (2.0585) [2021-04-14 17:57:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][1140/1251] eta 0:00:30 lr 0.000446 time 0.2748 (0.2762) loss 5.4642 (5.2960) grad_norm 1.8995 (2.0569) [2021-04-14 17:57:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][1150/1251] eta 0:00:27 lr 0.000447 time 0.2765 (0.2761) loss 5.1147 (5.2955) grad_norm 2.1982 (2.0554) [2021-04-14 17:57:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][1160/1251] eta 0:00:25 lr 0.000447 time 0.2750 (0.2761) loss 4.4299 (5.2953) grad_norm 2.2543 (2.0550) [2021-04-14 17:57:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][1170/1251] eta 0:00:22 lr 0.000447 time 0.2585 (0.2761) loss 4.8833 (5.2929) grad_norm 2.7523 (2.0548) [2021-04-14 17:57:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][1180/1251] eta 0:00:19 lr 0.000448 time 0.2850 (0.2760) loss 5.4288 (5.2917) grad_norm 2.0771 (2.0540) [2021-04-14 17:57:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][1190/1251] eta 0:00:16 lr 0.000448 time 0.2808 (0.2760) loss 5.8652 (5.2915) grad_norm 2.0147 (2.0562) [2021-04-14 17:57:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][1200/1251] eta 0:00:14 lr 0.000449 time 0.2899 (0.2760) loss 4.5036 (5.2910) grad_norm 2.0816 (2.0552) [2021-04-14 17:57:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][1210/1251] eta 0:00:11 lr 0.000449 time 0.2701 (0.2760) loss 5.5783 (5.2921) grad_norm 1.7972 (2.0538) [2021-04-14 17:57:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][1220/1251] eta 0:00:08 lr 0.000449 time 0.2854 (0.2759) loss 4.7503 (5.2901) grad_norm 1.9396 (2.0510) [2021-04-14 17:57:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][1230/1251] eta 0:00:05 lr 0.000450 time 0.2643 (0.2759) loss 5.2423 (5.2900) grad_norm 1.8279 (2.0513) [2021-04-14 17:57:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][1240/1251] eta 0:00:03 lr 0.000450 time 0.2480 (0.2758) loss 4.5259 (5.2875) grad_norm 1.9704 (2.0500) [2021-04-14 17:57:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [8/300][1250/1251] eta 0:00:00 lr 0.000451 time 0.2480 (0.2756) loss 5.4882 (5.2880) grad_norm 1.5737 (2.0479) [2021-04-14 17:57:32 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 8 training takes 0:05:46 [2021-04-14 17:57:32 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_8.pth saving...... [2021-04-14 17:57:55 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_8.pth saved !!! [2021-04-14 17:57:56 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.264 (1.264) Loss 2.8404 (2.8404) Acc@1 39.746 (39.746) Acc@5 67.090 (67.090) [2021-04-14 17:57:57 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.175 (0.235) Loss 2.7531 (2.7774) Acc@1 41.406 (41.531) Acc@5 66.895 (67.694) [2021-04-14 17:58:00 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.074 (0.235) Loss 2.8938 (2.7822) Acc@1 39.551 (41.509) Acc@5 64.941 (67.304) [2021-04-14 17:58:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.139 (0.224) Loss 2.7757 (2.7893) Acc@1 42.480 (41.283) Acc@5 67.090 (67.096) [2021-04-14 17:58:04 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.111 (0.220) Loss 2.7889 (2.7810) Acc@1 41.797 (41.347) Acc@5 66.504 (67.266) [2021-04-14 17:58:05 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 41.428 Acc@5 67.238 [2021-04-14 17:58:05 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 41.4% [2021-04-14 17:58:05 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 41.43% [2021-04-14 17:58:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][0/1251] eta 1:17:18 lr 0.000451 time 3.7076 (3.7076) loss 4.6945 (4.6945) grad_norm 1.8703 (1.8703) [2021-04-14 17:58:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][10/1251] eta 0:12:16 lr 0.000451 time 0.2737 (0.5934) loss 4.9496 (5.1974) grad_norm 2.5676 (1.9008) [2021-04-14 17:58:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][20/1251] eta 0:09:00 lr 0.000451 time 0.2625 (0.4392) loss 4.1078 (5.0927) grad_norm 2.6045 (2.0530) [2021-04-14 17:58:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][30/1251] eta 0:07:50 lr 0.000452 time 0.2888 (0.3857) loss 5.0505 (5.1968) grad_norm 1.9024 (2.0788) [2021-04-14 17:58:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][40/1251] eta 0:07:13 lr 0.000452 time 0.2951 (0.3580) loss 4.0725 (5.2109) grad_norm 2.0662 (2.0678) [2021-04-14 17:58:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][50/1251] eta 0:06:49 lr 0.000453 time 0.2863 (0.3413) loss 4.3934 (5.2309) grad_norm 1.7791 (2.0881) [2021-04-14 17:58:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][60/1251] eta 0:06:32 lr 0.000453 time 0.2827 (0.3297) loss 4.5302 (5.2005) grad_norm 2.6616 (2.0775) [2021-04-14 17:58:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][70/1251] eta 0:06:19 lr 0.000453 time 0.2757 (0.3216) loss 5.3353 (5.2305) grad_norm 1.8155 (2.0785) [2021-04-14 17:58:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][80/1251] eta 0:06:09 lr 0.000454 time 0.2892 (0.3156) loss 5.5240 (5.2278) grad_norm 1.7330 (2.0601) [2021-04-14 17:58:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][90/1251] eta 0:06:01 lr 0.000454 time 0.2532 (0.3118) loss 5.7313 (5.2431) grad_norm 2.3199 (2.0687) [2021-04-14 17:58:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][100/1251] eta 0:05:54 lr 0.000455 time 0.2584 (0.3077) loss 5.4085 (5.2498) grad_norm 2.0755 (2.0596) [2021-04-14 17:58:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][110/1251] eta 0:05:47 lr 0.000455 time 0.2789 (0.3046) loss 5.0409 (5.2512) grad_norm 2.6192 (2.0626) [2021-04-14 17:58:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][120/1251] eta 0:05:41 lr 0.000455 time 0.2843 (0.3019) loss 5.0773 (5.2575) grad_norm 2.0515 (2.0750) [2021-04-14 17:58:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][130/1251] eta 0:05:37 lr 0.000456 time 0.2966 (0.3010) loss 5.4232 (5.2396) grad_norm 2.2738 (2.0632) [2021-04-14 17:58:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][140/1251] eta 0:05:33 lr 0.000456 time 0.2573 (0.3001) loss 6.1077 (5.2462) grad_norm 1.8968 (2.0519) [2021-04-14 17:58:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][150/1251] eta 0:05:28 lr 0.000457 time 0.2836 (0.2985) loss 4.1726 (5.2247) grad_norm 1.8760 (2.0443) [2021-04-14 17:58:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][160/1251] eta 0:05:23 lr 0.000457 time 0.2821 (0.2967) loss 5.3553 (5.2460) grad_norm 2.0046 (2.0387) [2021-04-14 17:58:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][170/1251] eta 0:05:18 lr 0.000457 time 0.2730 (0.2950) loss 5.5267 (5.2386) grad_norm 2.2499 (2.0466) [2021-04-14 17:58:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][180/1251] eta 0:05:14 lr 0.000458 time 0.2639 (0.2936) loss 5.1863 (5.2346) grad_norm 1.9450 (2.0485) [2021-04-14 17:59:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][190/1251] eta 0:05:10 lr 0.000458 time 0.2537 (0.2924) loss 5.4180 (5.2332) grad_norm 1.5196 (2.0348) [2021-04-14 17:59:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][200/1251] eta 0:05:06 lr 0.000459 time 0.2612 (0.2912) loss 5.2401 (5.2313) grad_norm 2.1188 (2.0318) [2021-04-14 17:59:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][210/1251] eta 0:05:02 lr 0.000459 time 0.2714 (0.2904) loss 4.2025 (5.2323) grad_norm 1.8094 (inf) [2021-04-14 17:59:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][220/1251] eta 0:04:58 lr 0.000459 time 0.2678 (0.2896) loss 5.5356 (5.2330) grad_norm 1.9281 (inf) [2021-04-14 17:59:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][230/1251] eta 0:04:54 lr 0.000460 time 0.2507 (0.2888) loss 5.4302 (5.2476) grad_norm 1.6784 (inf) [2021-04-14 17:59:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][240/1251] eta 0:04:51 lr 0.000460 time 0.3092 (0.2882) loss 5.3824 (5.2610) grad_norm 1.7118 (inf) [2021-04-14 17:59:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][250/1251] eta 0:04:47 lr 0.000461 time 0.2638 (0.2875) loss 5.3939 (5.2690) grad_norm 1.7378 (inf) [2021-04-14 17:59:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][260/1251] eta 0:04:44 lr 0.000461 time 0.2722 (0.2868) loss 4.3984 (5.2671) grad_norm 2.3982 (inf) [2021-04-14 17:59:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][270/1251] eta 0:04:40 lr 0.000461 time 0.2612 (0.2863) loss 3.8006 (5.2547) grad_norm 2.0286 (inf) [2021-04-14 17:59:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][280/1251] eta 0:04:37 lr 0.000462 time 0.2698 (0.2858) loss 5.5962 (5.2539) grad_norm 1.8729 (inf) [2021-04-14 17:59:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][290/1251] eta 0:04:34 lr 0.000462 time 0.2731 (0.2852) loss 5.7772 (5.2586) grad_norm 2.3562 (inf) [2021-04-14 17:59:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][300/1251] eta 0:04:30 lr 0.000463 time 0.2632 (0.2847) loss 4.6756 (5.2471) grad_norm 1.5163 (inf) [2021-04-14 17:59:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][310/1251] eta 0:04:27 lr 0.000463 time 0.2961 (0.2845) loss 4.7685 (5.2420) grad_norm 1.8854 (inf) [2021-04-14 17:59:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][320/1251] eta 0:04:24 lr 0.000463 time 0.2802 (0.2840) loss 5.7236 (5.2376) grad_norm 2.3075 (inf) [2021-04-14 17:59:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][330/1251] eta 0:04:21 lr 0.000464 time 0.2721 (0.2836) loss 4.6290 (5.2278) grad_norm 1.7869 (inf) [2021-04-14 17:59:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][340/1251] eta 0:04:17 lr 0.000464 time 0.2660 (0.2832) loss 5.2867 (5.2273) grad_norm 1.7591 (inf) [2021-04-14 17:59:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][350/1251] eta 0:04:14 lr 0.000465 time 0.2781 (0.2829) loss 5.5989 (5.2259) grad_norm 1.6199 (inf) [2021-04-14 17:59:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][360/1251] eta 0:04:11 lr 0.000465 time 0.2842 (0.2826) loss 5.5444 (5.2272) grad_norm 1.7585 (inf) [2021-04-14 17:59:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][370/1251] eta 0:04:08 lr 0.000465 time 0.2677 (0.2822) loss 4.9053 (5.2191) grad_norm 1.8368 (inf) [2021-04-14 17:59:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][380/1251] eta 0:04:05 lr 0.000466 time 0.2901 (0.2819) loss 5.7161 (5.2212) grad_norm 2.6041 (inf) [2021-04-14 17:59:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][390/1251] eta 0:04:02 lr 0.000466 time 0.2883 (0.2817) loss 5.7168 (5.2200) grad_norm 1.9291 (inf) [2021-04-14 17:59:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][400/1251] eta 0:03:59 lr 0.000467 time 0.2864 (0.2815) loss 4.5084 (5.2203) grad_norm 1.5669 (inf) [2021-04-14 18:00:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][410/1251] eta 0:03:56 lr 0.000467 time 0.3114 (0.2813) loss 4.5564 (5.2203) grad_norm 2.1975 (inf) [2021-04-14 18:00:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][420/1251] eta 0:03:53 lr 0.000467 time 0.2792 (0.2812) loss 4.6905 (5.2244) grad_norm 1.8907 (inf) [2021-04-14 18:00:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][430/1251] eta 0:03:50 lr 0.000468 time 0.3324 (0.2810) loss 4.8167 (5.2214) grad_norm 2.8898 (inf) [2021-04-14 18:00:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][440/1251] eta 0:03:47 lr 0.000468 time 0.2653 (0.2808) loss 5.3178 (5.2228) grad_norm 1.9519 (inf) [2021-04-14 18:00:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][450/1251] eta 0:03:44 lr 0.000469 time 0.2901 (0.2807) loss 5.0816 (5.2210) grad_norm 1.6858 (inf) [2021-04-14 18:00:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][460/1251] eta 0:03:41 lr 0.000469 time 0.2604 (0.2804) loss 4.2116 (5.2132) grad_norm 1.9751 (inf) [2021-04-14 18:00:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][470/1251] eta 0:03:38 lr 0.000469 time 0.2633 (0.2803) loss 5.6199 (5.2110) grad_norm 1.9803 (inf) [2021-04-14 18:00:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][480/1251] eta 0:03:36 lr 0.000470 time 0.3041 (0.2802) loss 5.8819 (5.2114) grad_norm 1.9013 (inf) [2021-04-14 18:00:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][490/1251] eta 0:03:33 lr 0.000470 time 0.2811 (0.2800) loss 4.1754 (5.2050) grad_norm 1.9933 (inf) [2021-04-14 18:00:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][500/1251] eta 0:03:30 lr 0.000471 time 0.2697 (0.2798) loss 5.5183 (5.2055) grad_norm 1.5281 (inf) [2021-04-14 18:00:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][510/1251] eta 0:03:27 lr 0.000471 time 0.2702 (0.2796) loss 5.6306 (5.2065) grad_norm 2.0700 (inf) [2021-04-14 18:00:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][520/1251] eta 0:03:24 lr 0.000471 time 0.2831 (0.2795) loss 5.0563 (5.2069) grad_norm 2.3967 (inf) [2021-04-14 18:00:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][530/1251] eta 0:03:21 lr 0.000472 time 0.2811 (0.2794) loss 4.5368 (5.2061) grad_norm 1.9320 (inf) [2021-04-14 18:00:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][540/1251] eta 0:03:18 lr 0.000472 time 0.2737 (0.2792) loss 5.3849 (5.2037) grad_norm 2.3823 (inf) [2021-04-14 18:00:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][550/1251] eta 0:03:15 lr 0.000473 time 0.2713 (0.2791) loss 5.9253 (5.2063) grad_norm 2.2054 (inf) [2021-04-14 18:00:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][560/1251] eta 0:03:12 lr 0.000473 time 0.2653 (0.2790) loss 4.4632 (5.2079) grad_norm 2.0251 (inf) [2021-04-14 18:00:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][570/1251] eta 0:03:09 lr 0.000473 time 0.2571 (0.2789) loss 4.2603 (5.2056) grad_norm 1.8519 (inf) [2021-04-14 18:00:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][580/1251] eta 0:03:07 lr 0.000474 time 0.2709 (0.2788) loss 5.6343 (5.2037) grad_norm 2.6087 (inf) [2021-04-14 18:00:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][590/1251] eta 0:03:04 lr 0.000474 time 0.2672 (0.2786) loss 5.9762 (5.2036) grad_norm 1.8127 (inf) [2021-04-14 18:00:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][600/1251] eta 0:03:01 lr 0.000475 time 0.2797 (0.2785) loss 4.5314 (5.2018) grad_norm 1.5682 (inf) [2021-04-14 18:00:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][610/1251] eta 0:02:58 lr 0.000475 time 0.2597 (0.2784) loss 5.5680 (5.2016) grad_norm 1.7620 (inf) [2021-04-14 18:00:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][620/1251] eta 0:02:55 lr 0.000475 time 0.2693 (0.2784) loss 5.5059 (5.2010) grad_norm 1.6936 (inf) [2021-04-14 18:01:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][630/1251] eta 0:02:52 lr 0.000476 time 0.2950 (0.2784) loss 5.6464 (5.1949) grad_norm 1.7576 (inf) [2021-04-14 18:01:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][640/1251] eta 0:02:50 lr 0.000476 time 0.2489 (0.2784) loss 5.8197 (5.2002) grad_norm 1.6505 (inf) [2021-04-14 18:01:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][650/1251] eta 0:02:47 lr 0.000477 time 0.2782 (0.2783) loss 4.8119 (5.2005) grad_norm 1.9271 (inf) [2021-04-14 18:01:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][660/1251] eta 0:02:44 lr 0.000477 time 0.2947 (0.2782) loss 5.0193 (5.2006) grad_norm 2.3122 (inf) [2021-04-14 18:01:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][670/1251] eta 0:02:41 lr 0.000477 time 0.2818 (0.2783) loss 5.0592 (5.2030) grad_norm 2.0657 (inf) [2021-04-14 18:01:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][680/1251] eta 0:02:38 lr 0.000478 time 0.2734 (0.2782) loss 4.6321 (5.1999) grad_norm 2.2403 (inf) [2021-04-14 18:01:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][690/1251] eta 0:02:36 lr 0.000478 time 0.2817 (0.2781) loss 5.2234 (5.2031) grad_norm 1.9519 (inf) [2021-04-14 18:01:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][700/1251] eta 0:02:33 lr 0.000478 time 0.2767 (0.2780) loss 5.6948 (5.2032) grad_norm 1.7465 (inf) [2021-04-14 18:01:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][710/1251] eta 0:02:30 lr 0.000479 time 0.2735 (0.2781) loss 4.8625 (5.2017) grad_norm 3.0293 (inf) [2021-04-14 18:01:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][720/1251] eta 0:02:27 lr 0.000479 time 0.2648 (0.2783) loss 5.5026 (5.2029) grad_norm 2.3770 (inf) [2021-04-14 18:01:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][730/1251] eta 0:02:24 lr 0.000480 time 0.2596 (0.2781) loss 3.8548 (5.1995) grad_norm 1.6746 (inf) [2021-04-14 18:01:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][740/1251] eta 0:02:22 lr 0.000480 time 0.2800 (0.2781) loss 5.6823 (5.1994) grad_norm 2.2780 (inf) [2021-04-14 18:01:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][750/1251] eta 0:02:19 lr 0.000480 time 0.2787 (0.2780) loss 5.1690 (5.2008) grad_norm 1.8157 (inf) [2021-04-14 18:01:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][760/1251] eta 0:02:16 lr 0.000481 time 0.2815 (0.2779) loss 4.9263 (5.1994) grad_norm 1.5182 (inf) [2021-04-14 18:01:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][770/1251] eta 0:02:13 lr 0.000481 time 0.2802 (0.2778) loss 4.6628 (5.1980) grad_norm 1.8356 (inf) [2021-04-14 18:01:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][780/1251] eta 0:02:10 lr 0.000482 time 0.2532 (0.2777) loss 4.4508 (5.1986) grad_norm 2.0767 (inf) [2021-04-14 18:01:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][790/1251] eta 0:02:07 lr 0.000482 time 0.2457 (0.2776) loss 5.4527 (5.1988) grad_norm 1.6598 (inf) [2021-04-14 18:01:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][800/1251] eta 0:02:05 lr 0.000482 time 0.2774 (0.2776) loss 4.2630 (5.1976) grad_norm 1.7423 (inf) [2021-04-14 18:01:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][810/1251] eta 0:02:02 lr 0.000483 time 0.2800 (0.2776) loss 5.7623 (5.2004) grad_norm 1.7844 (inf) [2021-04-14 18:01:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][820/1251] eta 0:01:59 lr 0.000483 time 0.2789 (0.2775) loss 4.9390 (5.2003) grad_norm 2.2880 (inf) [2021-04-14 18:01:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][830/1251] eta 0:01:56 lr 0.000484 time 0.2431 (0.2774) loss 4.4266 (5.1993) grad_norm 1.9325 (inf) [2021-04-14 18:01:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][840/1251] eta 0:01:53 lr 0.000484 time 0.2955 (0.2773) loss 5.0671 (5.1960) grad_norm 1.8144 (inf) [2021-04-14 18:02:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][850/1251] eta 0:01:51 lr 0.000484 time 0.2535 (0.2773) loss 5.4069 (5.1938) grad_norm 2.7877 (inf) [2021-04-14 18:02:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][860/1251] eta 0:01:48 lr 0.000485 time 0.2548 (0.2772) loss 4.8815 (5.1911) grad_norm 1.8092 (inf) [2021-04-14 18:02:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][870/1251] eta 0:01:45 lr 0.000485 time 0.2798 (0.2771) loss 5.5628 (5.1896) grad_norm 2.3632 (inf) [2021-04-14 18:02:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][880/1251] eta 0:01:42 lr 0.000486 time 0.2902 (0.2770) loss 5.2114 (5.1860) grad_norm 1.7598 (inf) [2021-04-14 18:02:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][890/1251] eta 0:01:39 lr 0.000486 time 0.2594 (0.2769) loss 5.5576 (5.1857) grad_norm 2.4237 (inf) [2021-04-14 18:02:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][900/1251] eta 0:01:37 lr 0.000486 time 0.2681 (0.2769) loss 5.4597 (5.1839) grad_norm 2.2625 (inf) [2021-04-14 18:02:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][910/1251] eta 0:01:34 lr 0.000487 time 0.2513 (0.2769) loss 5.3767 (5.1851) grad_norm 1.7900 (inf) [2021-04-14 18:02:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][920/1251] eta 0:01:31 lr 0.000487 time 0.2459 (0.2768) loss 5.3827 (5.1838) grad_norm 2.1166 (inf) [2021-04-14 18:02:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][930/1251] eta 0:01:28 lr 0.000488 time 0.3015 (0.2767) loss 5.5757 (5.1807) grad_norm 2.1690 (inf) [2021-04-14 18:02:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][940/1251] eta 0:01:26 lr 0.000488 time 0.2746 (0.2767) loss 4.4423 (5.1771) grad_norm 2.2670 (inf) [2021-04-14 18:02:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][950/1251] eta 0:01:23 lr 0.000488 time 0.2685 (0.2766) loss 5.7531 (5.1757) grad_norm 1.9373 (inf) [2021-04-14 18:02:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][960/1251] eta 0:01:20 lr 0.000489 time 0.2704 (0.2766) loss 4.8368 (5.1744) grad_norm 1.6254 (inf) [2021-04-14 18:02:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][970/1251] eta 0:01:17 lr 0.000489 time 0.2781 (0.2766) loss 5.7470 (5.1754) grad_norm 1.4645 (inf) [2021-04-14 18:02:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][980/1251] eta 0:01:14 lr 0.000490 time 0.2731 (0.2765) loss 5.4758 (5.1748) grad_norm 1.7430 (inf) [2021-04-14 18:02:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][990/1251] eta 0:01:12 lr 0.000490 time 0.2941 (0.2764) loss 5.3378 (5.1733) grad_norm 1.8892 (inf) [2021-04-14 18:02:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][1000/1251] eta 0:01:09 lr 0.000490 time 0.2617 (0.2764) loss 5.6779 (5.1741) grad_norm 1.5352 (inf) [2021-04-14 18:02:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][1010/1251] eta 0:01:06 lr 0.000491 time 0.2723 (0.2763) loss 4.4340 (5.1736) grad_norm 1.7314 (inf) [2021-04-14 18:02:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][1020/1251] eta 0:01:03 lr 0.000491 time 0.2596 (0.2763) loss 5.5367 (5.1723) grad_norm 1.7116 (inf) [2021-04-14 18:02:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][1030/1251] eta 0:01:01 lr 0.000492 time 0.2545 (0.2763) loss 5.1125 (5.1730) grad_norm 2.0854 (inf) [2021-04-14 18:02:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][1040/1251] eta 0:00:58 lr 0.000492 time 0.2565 (0.2762) loss 6.0266 (5.1732) grad_norm 1.4185 (inf) [2021-04-14 18:02:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][1050/1251] eta 0:00:55 lr 0.000492 time 0.2775 (0.2762) loss 5.0748 (5.1719) grad_norm 2.1678 (inf) [2021-04-14 18:02:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][1060/1251] eta 0:00:52 lr 0.000493 time 0.2670 (0.2761) loss 5.7007 (5.1720) grad_norm 1.9288 (inf) [2021-04-14 18:03:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][1070/1251] eta 0:00:49 lr 0.000493 time 0.2817 (0.2761) loss 5.2336 (5.1726) grad_norm 1.8671 (inf) [2021-04-14 18:03:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][1080/1251] eta 0:00:47 lr 0.000494 time 0.2575 (0.2760) loss 5.3489 (5.1713) grad_norm 1.5834 (inf) [2021-04-14 18:03:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][1090/1251] eta 0:00:44 lr 0.000494 time 0.2661 (0.2760) loss 5.1998 (5.1712) grad_norm 2.2899 (inf) [2021-04-14 18:03:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][1100/1251] eta 0:00:41 lr 0.000494 time 0.2760 (0.2759) loss 4.9639 (5.1694) grad_norm 2.3799 (inf) [2021-04-14 18:03:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][1110/1251] eta 0:00:38 lr 0.000495 time 0.2844 (0.2759) loss 4.2728 (5.1650) grad_norm 1.9054 (inf) [2021-04-14 18:03:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][1120/1251] eta 0:00:36 lr 0.000495 time 0.2662 (0.2758) loss 5.2210 (5.1642) grad_norm 1.4890 (inf) [2021-04-14 18:03:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][1130/1251] eta 0:00:33 lr 0.000496 time 0.2644 (0.2758) loss 4.9719 (5.1638) grad_norm 1.5999 (inf) [2021-04-14 18:03:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][1140/1251] eta 0:00:30 lr 0.000496 time 0.2869 (0.2758) loss 4.5369 (5.1618) grad_norm 2.4994 (inf) [2021-04-14 18:03:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][1150/1251] eta 0:00:27 lr 0.000496 time 0.2798 (0.2758) loss 5.3427 (5.1625) grad_norm 1.5290 (inf) [2021-04-14 18:03:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][1160/1251] eta 0:00:25 lr 0.000497 time 0.2703 (0.2757) loss 5.3825 (5.1604) grad_norm 1.5024 (inf) [2021-04-14 18:03:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][1170/1251] eta 0:00:22 lr 0.000497 time 0.2739 (0.2757) loss 4.3965 (5.1587) grad_norm 1.6077 (inf) [2021-04-14 18:03:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][1180/1251] eta 0:00:19 lr 0.000498 time 0.3002 (0.2757) loss 4.7800 (5.1559) grad_norm 1.7374 (inf) [2021-04-14 18:03:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][1190/1251] eta 0:00:16 lr 0.000498 time 0.3108 (0.2757) loss 4.2686 (5.1547) grad_norm 1.6376 (inf) [2021-04-14 18:03:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][1200/1251] eta 0:00:14 lr 0.000498 time 0.2979 (0.2756) loss 4.4998 (5.1538) grad_norm 2.1276 (inf) [2021-04-14 18:03:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][1210/1251] eta 0:00:11 lr 0.000499 time 0.2844 (0.2756) loss 5.3239 (5.1552) grad_norm 2.6029 (inf) [2021-04-14 18:03:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][1220/1251] eta 0:00:08 lr 0.000499 time 0.2655 (0.2756) loss 5.1882 (5.1545) grad_norm 1.4945 (inf) [2021-04-14 18:03:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][1230/1251] eta 0:00:05 lr 0.000500 time 0.2728 (0.2755) loss 5.2848 (5.1539) grad_norm 2.1383 (inf) [2021-04-14 18:03:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][1240/1251] eta 0:00:03 lr 0.000500 time 0.2414 (0.2755) loss 4.1006 (5.1533) grad_norm 1.6200 (inf) [2021-04-14 18:03:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [9/300][1250/1251] eta 0:00:00 lr 0.000500 time 0.2480 (0.2753) loss 5.3153 (5.1541) grad_norm 1.7004 (inf) [2021-04-14 18:03:52 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 9 training takes 0:05:46 [2021-04-14 18:03:52 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_9.pth saving...... [2021-04-14 18:04:17 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_9.pth saved !!! [2021-04-14 18:04:18 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.412 (1.412) Loss 2.6020 (2.6020) Acc@1 43.750 (43.750) Acc@5 69.824 (69.824) [2021-04-14 18:04:20 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.109 (0.308) Loss 2.5994 (2.6300) Acc@1 42.871 (43.173) Acc@5 71.777 (69.727) [2021-04-14 18:04:22 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.097 (0.236) Loss 2.6490 (2.6183) Acc@1 43.652 (43.597) Acc@5 69.727 (69.829) [2021-04-14 18:04:25 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.109 (0.251) Loss 2.5606 (2.6092) Acc@1 45.410 (44.065) Acc@5 69.727 (69.774) [2021-04-14 18:04:26 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.215) Loss 2.6477 (2.6122) Acc@1 44.629 (44.074) Acc@5 67.188 (69.831) [2021-04-14 18:04:28 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 43.988 Acc@5 69.724 [2021-04-14 18:04:28 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 44.0% [2021-04-14 18:04:28 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 43.99% [2021-04-14 18:04:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][0/1251] eta 1:19:03 lr 0.000501 time 3.7921 (3.7921) loss 5.3961 (5.3961) grad_norm 2.2096 (2.2096) [2021-04-14 18:04:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][10/1251] eta 0:12:15 lr 0.000501 time 0.2547 (0.5928) loss 5.5156 (5.4523) grad_norm 1.9850 (1.9046) [2021-04-14 18:04:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][20/1251] eta 0:09:04 lr 0.000501 time 0.2555 (0.4426) loss 4.1953 (5.3491) grad_norm 2.1464 (1.8955) [2021-04-14 18:04:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][30/1251] eta 0:07:52 lr 0.000502 time 0.2581 (0.3873) loss 5.6661 (5.2660) grad_norm 1.8186 (1.8760) [2021-04-14 18:04:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][40/1251] eta 0:07:14 lr 0.000502 time 0.2658 (0.3591) loss 5.3730 (5.2548) grad_norm 1.6626 (1.9108) [2021-04-14 18:04:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][50/1251] eta 0:06:50 lr 0.000502 time 0.2704 (0.3415) loss 4.4121 (5.2249) grad_norm 2.3983 (1.9425) [2021-04-14 18:04:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][60/1251] eta 0:06:33 lr 0.000503 time 0.2481 (0.3305) loss 5.1898 (5.1885) grad_norm 1.9619 (1.9756) [2021-04-14 18:04:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][70/1251] eta 0:06:21 lr 0.000503 time 0.2795 (0.3231) loss 4.2239 (5.1968) grad_norm 1.7942 (1.9397) [2021-04-14 18:04:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][80/1251] eta 0:06:12 lr 0.000504 time 0.2893 (0.3178) loss 5.6060 (5.2228) grad_norm 1.8408 (1.9431) [2021-04-14 18:04:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][90/1251] eta 0:06:03 lr 0.000504 time 0.2639 (0.3127) loss 5.5301 (5.2046) grad_norm 1.6333 (1.9164) [2021-04-14 18:04:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][100/1251] eta 0:05:55 lr 0.000504 time 0.2786 (0.3092) loss 5.1652 (5.1866) grad_norm 1.6521 (1.9081) [2021-04-14 18:05:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][110/1251] eta 0:05:48 lr 0.000505 time 0.2616 (0.3057) loss 4.3246 (5.1868) grad_norm 1.8082 (1.9131) [2021-04-14 18:05:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][120/1251] eta 0:05:42 lr 0.000505 time 0.2750 (0.3031) loss 5.3676 (5.1845) grad_norm 1.9080 (1.9263) [2021-04-14 18:05:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][130/1251] eta 0:05:37 lr 0.000506 time 0.2691 (0.3011) loss 5.4678 (5.1799) grad_norm 1.8149 (1.9046) [2021-04-14 18:05:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][140/1251] eta 0:05:32 lr 0.000506 time 0.2852 (0.2990) loss 4.9425 (5.1706) grad_norm 1.6450 (1.8917) [2021-04-14 18:05:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][150/1251] eta 0:05:26 lr 0.000506 time 0.2613 (0.2969) loss 5.4088 (5.1713) grad_norm 1.7665 (1.8921) [2021-04-14 18:05:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][160/1251] eta 0:05:21 lr 0.000507 time 0.2735 (0.2951) loss 5.5370 (5.1741) grad_norm 1.8874 (1.8905) [2021-04-14 18:05:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][170/1251] eta 0:05:17 lr 0.000507 time 0.2734 (0.2938) loss 5.5128 (5.1823) grad_norm 2.1640 (1.8819) [2021-04-14 18:05:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][180/1251] eta 0:05:13 lr 0.000508 time 0.2707 (0.2929) loss 5.7428 (5.1837) grad_norm 1.5144 (1.8780) [2021-04-14 18:05:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][190/1251] eta 0:05:09 lr 0.000508 time 0.2603 (0.2919) loss 5.6864 (5.1869) grad_norm 3.7084 (1.8911) [2021-04-14 18:05:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][200/1251] eta 0:05:05 lr 0.000508 time 0.2582 (0.2909) loss 5.0592 (5.1824) grad_norm 1.7789 (1.8879) [2021-04-14 18:05:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][210/1251] eta 0:05:01 lr 0.000509 time 0.2796 (0.2898) loss 5.8615 (5.1883) grad_norm 1.5557 (1.8789) [2021-04-14 18:05:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][220/1251] eta 0:04:57 lr 0.000509 time 0.2560 (0.2889) loss 4.9026 (5.1879) grad_norm 1.7166 (1.8731) [2021-04-14 18:05:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][230/1251] eta 0:04:54 lr 0.000510 time 0.2686 (0.2881) loss 5.9311 (5.1876) grad_norm 1.6427 (1.8691) [2021-04-14 18:05:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][240/1251] eta 0:04:50 lr 0.000510 time 0.2916 (0.2877) loss 4.8552 (5.1730) grad_norm 1.9284 (1.8703) [2021-04-14 18:05:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][250/1251] eta 0:04:47 lr 0.000510 time 0.2650 (0.2868) loss 4.0412 (5.1720) grad_norm 1.6913 (1.8769) [2021-04-14 18:05:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][260/1251] eta 0:04:43 lr 0.000511 time 0.2713 (0.2863) loss 5.4346 (5.1785) grad_norm 2.4385 (1.8694) [2021-04-14 18:05:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][270/1251] eta 0:04:40 lr 0.000511 time 0.2904 (0.2858) loss 5.0100 (5.1698) grad_norm 1.5838 (1.8716) [2021-04-14 18:05:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][280/1251] eta 0:04:37 lr 0.000512 time 0.2981 (0.2853) loss 5.2570 (5.1588) grad_norm 1.6634 (1.8657) [2021-04-14 18:05:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][290/1251] eta 0:04:33 lr 0.000512 time 0.2734 (0.2847) loss 5.3000 (5.1536) grad_norm 1.8554 (1.8673) [2021-04-14 18:05:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][300/1251] eta 0:04:30 lr 0.000512 time 0.2882 (0.2843) loss 5.5535 (5.1486) grad_norm 1.7182 (1.8700) [2021-04-14 18:05:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][310/1251] eta 0:04:27 lr 0.000513 time 0.2659 (0.2838) loss 5.7064 (5.1509) grad_norm 2.4306 (1.8698) [2021-04-14 18:05:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][320/1251] eta 0:04:23 lr 0.000513 time 0.2779 (0.2834) loss 4.9597 (5.1537) grad_norm 1.8442 (1.8686) [2021-04-14 18:06:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][330/1251] eta 0:04:20 lr 0.000514 time 0.2925 (0.2829) loss 4.6242 (5.1479) grad_norm 1.5330 (1.8673) [2021-04-14 18:06:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][340/1251] eta 0:04:17 lr 0.000514 time 0.2704 (0.2827) loss 5.1205 (5.1534) grad_norm 1.5482 (1.8652) [2021-04-14 18:06:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][350/1251] eta 0:04:14 lr 0.000514 time 0.2745 (0.2825) loss 4.3329 (5.1480) grad_norm 2.2528 (1.8664) [2021-04-14 18:06:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][360/1251] eta 0:04:11 lr 0.000515 time 0.2636 (0.2822) loss 5.4859 (5.1517) grad_norm 1.5468 (1.8646) [2021-04-14 18:06:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][370/1251] eta 0:04:08 lr 0.000515 time 0.2719 (0.2819) loss 5.5134 (5.1508) grad_norm 1.9504 (1.8609) [2021-04-14 18:06:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][380/1251] eta 0:04:05 lr 0.000516 time 0.2600 (0.2815) loss 5.8142 (5.1500) grad_norm 1.8167 (1.8664) [2021-04-14 18:06:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][390/1251] eta 0:04:02 lr 0.000516 time 0.2733 (0.2813) loss 5.7480 (5.1513) grad_norm 1.9137 (1.8680) [2021-04-14 18:06:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][400/1251] eta 0:03:59 lr 0.000516 time 0.2860 (0.2810) loss 4.5036 (5.1560) grad_norm 1.5474 (1.8644) [2021-04-14 18:06:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][410/1251] eta 0:03:56 lr 0.000517 time 0.2795 (0.2808) loss 5.2498 (5.1614) grad_norm 1.6879 (1.8628) [2021-04-14 18:06:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][420/1251] eta 0:03:53 lr 0.000517 time 0.2602 (0.2805) loss 4.8215 (5.1589) grad_norm 1.6536 (1.8596) [2021-04-14 18:06:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][430/1251] eta 0:03:50 lr 0.000518 time 0.2797 (0.2804) loss 5.3959 (5.1559) grad_norm 2.4957 (1.8664) [2021-04-14 18:06:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][440/1251] eta 0:03:47 lr 0.000518 time 0.2671 (0.2802) loss 5.5425 (5.1518) grad_norm 2.8992 (1.8686) [2021-04-14 18:06:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][450/1251] eta 0:03:44 lr 0.000518 time 0.2491 (0.2802) loss 5.1701 (5.1516) grad_norm 1.4816 (1.8682) [2021-04-14 18:06:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][460/1251] eta 0:03:41 lr 0.000519 time 0.2620 (0.2800) loss 4.0380 (5.1532) grad_norm 2.4547 (1.8711) [2021-04-14 18:06:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][470/1251] eta 0:03:38 lr 0.000519 time 0.2637 (0.2798) loss 4.4935 (5.1486) grad_norm 2.0362 (1.8731) [2021-04-14 18:06:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][480/1251] eta 0:03:35 lr 0.000520 time 0.2658 (0.2796) loss 5.3418 (5.1433) grad_norm 2.6413 (1.8739) [2021-04-14 18:06:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][490/1251] eta 0:03:32 lr 0.000520 time 0.2598 (0.2793) loss 4.0371 (5.1389) grad_norm 1.6548 (1.8708) [2021-04-14 18:06:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][500/1251] eta 0:03:29 lr 0.000520 time 0.2952 (0.2792) loss 5.6408 (5.1396) grad_norm 1.4250 (1.8665) [2021-04-14 18:06:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][510/1251] eta 0:03:26 lr 0.000521 time 0.2613 (0.2790) loss 5.2461 (5.1438) grad_norm 1.4482 (1.8637) [2021-04-14 18:06:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][520/1251] eta 0:03:23 lr 0.000521 time 0.2788 (0.2789) loss 3.9911 (5.1444) grad_norm 1.6210 (1.8609) [2021-04-14 18:06:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][530/1251] eta 0:03:21 lr 0.000522 time 0.2753 (0.2788) loss 5.3014 (5.1451) grad_norm 1.9725 (1.8607) [2021-04-14 18:06:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][540/1251] eta 0:03:18 lr 0.000522 time 0.2762 (0.2790) loss 5.2531 (5.1478) grad_norm 1.9714 (1.8601) [2021-04-14 18:07:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][550/1251] eta 0:03:15 lr 0.000522 time 0.2490 (0.2788) loss 4.1776 (5.1449) grad_norm 2.9037 (1.8640) [2021-04-14 18:07:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][560/1251] eta 0:03:12 lr 0.000523 time 0.2758 (0.2787) loss 5.2890 (5.1432) grad_norm 2.8094 (1.8698) [2021-04-14 18:07:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][570/1251] eta 0:03:09 lr 0.000523 time 0.2545 (0.2785) loss 5.4673 (5.1369) grad_norm 1.8831 (1.8708) [2021-04-14 18:07:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][580/1251] eta 0:03:06 lr 0.000524 time 0.2584 (0.2787) loss 4.0727 (5.1387) grad_norm 1.7775 (1.8706) [2021-04-14 18:07:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][590/1251] eta 0:03:04 lr 0.000524 time 0.2564 (0.2785) loss 5.8310 (5.1396) grad_norm 2.4666 (1.8715) [2021-04-14 18:07:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][600/1251] eta 0:03:01 lr 0.000524 time 0.2635 (0.2782) loss 4.9462 (5.1398) grad_norm 2.5139 (1.8753) [2021-04-14 18:07:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][610/1251] eta 0:02:58 lr 0.000525 time 0.2958 (0.2782) loss 5.4623 (5.1350) grad_norm 2.6193 (1.8742) [2021-04-14 18:07:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][620/1251] eta 0:02:55 lr 0.000525 time 0.2954 (0.2781) loss 4.1541 (5.1346) grad_norm 1.7409 (1.8789) [2021-04-14 18:07:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][630/1251] eta 0:02:52 lr 0.000526 time 0.2571 (0.2779) loss 4.3812 (5.1356) grad_norm 1.6983 (1.8811) [2021-04-14 18:07:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][640/1251] eta 0:02:49 lr 0.000526 time 0.2697 (0.2778) loss 5.5796 (5.1396) grad_norm 1.9006 (1.8827) [2021-04-14 18:07:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][650/1251] eta 0:02:46 lr 0.000526 time 0.2560 (0.2776) loss 4.6415 (5.1377) grad_norm 2.8747 (1.8825) [2021-04-14 18:07:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][660/1251] eta 0:02:44 lr 0.000527 time 0.2532 (0.2775) loss 4.5814 (5.1338) grad_norm 1.6015 (1.8802) [2021-04-14 18:07:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][670/1251] eta 0:02:41 lr 0.000527 time 0.2721 (0.2774) loss 5.1460 (5.1361) grad_norm 1.8133 (1.8762) [2021-04-14 18:07:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][680/1251] eta 0:02:38 lr 0.000528 time 0.2675 (0.2774) loss 4.9076 (5.1324) grad_norm 1.9578 (1.8739) [2021-04-14 18:07:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][690/1251] eta 0:02:35 lr 0.000528 time 0.2964 (0.2774) loss 5.4859 (5.1332) grad_norm 1.7176 (1.8732) [2021-04-14 18:07:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][700/1251] eta 0:02:32 lr 0.000528 time 0.2691 (0.2773) loss 5.5724 (5.1358) grad_norm 2.0023 (1.8723) [2021-04-14 18:07:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][710/1251] eta 0:02:29 lr 0.000529 time 0.2604 (0.2772) loss 4.7406 (5.1384) grad_norm 2.0873 (1.8723) [2021-04-14 18:07:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][720/1251] eta 0:02:27 lr 0.000529 time 0.2674 (0.2771) loss 5.9246 (5.1398) grad_norm 1.8601 (1.8701) [2021-04-14 18:07:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][730/1251] eta 0:02:24 lr 0.000530 time 0.2798 (0.2770) loss 4.3382 (5.1368) grad_norm 1.7360 (1.8691) [2021-04-14 18:07:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][740/1251] eta 0:02:21 lr 0.000530 time 0.2902 (0.2769) loss 5.2746 (5.1332) grad_norm 1.6857 (1.8659) [2021-04-14 18:07:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][750/1251] eta 0:02:18 lr 0.000530 time 0.2466 (0.2768) loss 4.9778 (5.1314) grad_norm 1.7075 (1.8639) [2021-04-14 18:07:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][760/1251] eta 0:02:15 lr 0.000531 time 0.2536 (0.2767) loss 4.0522 (5.1292) grad_norm 1.4926 (1.8635) [2021-04-14 18:08:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][770/1251] eta 0:02:13 lr 0.000531 time 0.2898 (0.2767) loss 5.6698 (5.1269) grad_norm 1.4099 (1.8606) [2021-04-14 18:08:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][780/1251] eta 0:02:10 lr 0.000532 time 0.2607 (0.2767) loss 5.4391 (5.1285) grad_norm 2.5000 (1.8597) [2021-04-14 18:08:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][790/1251] eta 0:02:07 lr 0.000532 time 0.2712 (0.2766) loss 4.6699 (5.1259) grad_norm 1.9310 (1.8580) [2021-04-14 18:08:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][800/1251] eta 0:02:04 lr 0.000532 time 0.2715 (0.2765) loss 3.7874 (5.1197) grad_norm 1.4183 (1.8561) [2021-04-14 18:08:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][810/1251] eta 0:02:01 lr 0.000533 time 0.2632 (0.2765) loss 5.3422 (5.1176) grad_norm 1.9925 (1.8575) [2021-04-14 18:08:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][820/1251] eta 0:01:59 lr 0.000533 time 0.2633 (0.2764) loss 6.0146 (5.1181) grad_norm 1.7991 (1.8601) [2021-04-14 18:08:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][830/1251] eta 0:01:56 lr 0.000534 time 0.2710 (0.2763) loss 5.7766 (5.1174) grad_norm 1.3687 (1.8599) [2021-04-14 18:08:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][840/1251] eta 0:01:53 lr 0.000534 time 0.2764 (0.2763) loss 5.2623 (5.1179) grad_norm 1.3599 (1.8575) [2021-04-14 18:08:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][850/1251] eta 0:01:50 lr 0.000534 time 0.2439 (0.2762) loss 4.1810 (5.1163) grad_norm 1.3008 (1.8574) [2021-04-14 18:08:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][860/1251] eta 0:01:47 lr 0.000535 time 0.2727 (0.2761) loss 5.4592 (5.1181) grad_norm 1.7854 (1.8566) [2021-04-14 18:08:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][870/1251] eta 0:01:45 lr 0.000535 time 0.2754 (0.2761) loss 5.2885 (5.1212) grad_norm 1.5333 (1.8593) [2021-04-14 18:08:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][880/1251] eta 0:01:42 lr 0.000536 time 0.2657 (0.2760) loss 4.8386 (5.1195) grad_norm 2.3641 (1.8591) [2021-04-14 18:08:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][890/1251] eta 0:01:39 lr 0.000536 time 0.2924 (0.2760) loss 4.1859 (5.1143) grad_norm 1.8480 (1.8582) [2021-04-14 18:08:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][900/1251] eta 0:01:36 lr 0.000536 time 0.2922 (0.2760) loss 4.0185 (5.1073) grad_norm 1.6075 (1.8569) [2021-04-14 18:08:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][910/1251] eta 0:01:34 lr 0.000537 time 0.2854 (0.2760) loss 5.0157 (5.1073) grad_norm 1.9044 (1.8552) [2021-04-14 18:08:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][920/1251] eta 0:01:31 lr 0.000537 time 0.2739 (0.2760) loss 4.9049 (5.1080) grad_norm 2.0678 (1.8589) [2021-04-14 18:08:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][930/1251] eta 0:01:28 lr 0.000538 time 0.2827 (0.2759) loss 4.6828 (5.1118) grad_norm 1.8300 (1.8582) [2021-04-14 18:08:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][940/1251] eta 0:01:25 lr 0.000538 time 0.2680 (0.2758) loss 4.5241 (5.1127) grad_norm 2.5313 (1.8588) [2021-04-14 18:08:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][950/1251] eta 0:01:23 lr 0.000538 time 0.2577 (0.2757) loss 4.4026 (5.1125) grad_norm 1.4119 (1.8577) [2021-04-14 18:08:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][960/1251] eta 0:01:20 lr 0.000539 time 0.2782 (0.2758) loss 5.1946 (5.1136) grad_norm 2.0159 (1.8585) [2021-04-14 18:08:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][970/1251] eta 0:01:17 lr 0.000539 time 0.2783 (0.2758) loss 4.5435 (5.1115) grad_norm 1.4041 (1.8594) [2021-04-14 18:08:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][980/1251] eta 0:01:14 lr 0.000540 time 0.2753 (0.2758) loss 5.3535 (5.1104) grad_norm 1.7134 (1.8590) [2021-04-14 18:09:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][990/1251] eta 0:01:11 lr 0.000540 time 0.2751 (0.2757) loss 5.4109 (5.1099) grad_norm 2.4858 (1.8600) [2021-04-14 18:09:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][1000/1251] eta 0:01:09 lr 0.000540 time 0.2707 (0.2757) loss 5.5440 (5.1104) grad_norm 1.5891 (1.8592) [2021-04-14 18:09:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][1010/1251] eta 0:01:06 lr 0.000541 time 0.2903 (0.2757) loss 4.2310 (5.1064) grad_norm 2.0045 (1.8589) [2021-04-14 18:09:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][1020/1251] eta 0:01:03 lr 0.000541 time 0.2728 (0.2756) loss 5.0890 (5.1039) grad_norm 1.7800 (1.8581) [2021-04-14 18:09:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][1030/1251] eta 0:01:00 lr 0.000542 time 0.2809 (0.2757) loss 4.4976 (5.1021) grad_norm 2.0142 (1.8582) [2021-04-14 18:09:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][1040/1251] eta 0:00:58 lr 0.000542 time 0.2633 (0.2757) loss 4.1030 (5.1011) grad_norm 1.4879 (1.8562) [2021-04-14 18:09:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][1050/1251] eta 0:00:55 lr 0.000542 time 0.2856 (0.2756) loss 5.2246 (5.1005) grad_norm 1.4771 (1.8551) [2021-04-14 18:09:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][1060/1251] eta 0:00:52 lr 0.000543 time 0.2773 (0.2756) loss 5.1049 (5.0978) grad_norm 1.6926 (1.8560) [2021-04-14 18:09:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][1070/1251] eta 0:00:49 lr 0.000543 time 0.2896 (0.2755) loss 5.0486 (5.0969) grad_norm 1.3306 (1.8541) [2021-04-14 18:09:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][1080/1251] eta 0:00:47 lr 0.000544 time 0.2700 (0.2755) loss 4.6526 (5.0966) grad_norm 1.9741 (1.8529) [2021-04-14 18:09:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][1090/1251] eta 0:00:44 lr 0.000544 time 0.2841 (0.2755) loss 3.9611 (5.0936) grad_norm 1.6388 (1.8516) [2021-04-14 18:09:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][1100/1251] eta 0:00:41 lr 0.000544 time 0.2920 (0.2755) loss 5.2528 (5.0924) grad_norm 1.6973 (1.8498) [2021-04-14 18:09:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][1110/1251] eta 0:00:38 lr 0.000545 time 0.2728 (0.2754) loss 5.6251 (5.0919) grad_norm 1.9269 (1.8483) [2021-04-14 18:09:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][1120/1251] eta 0:00:36 lr 0.000545 time 0.2670 (0.2755) loss 5.2572 (5.0927) grad_norm 2.1975 (1.8476) [2021-04-14 18:09:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][1130/1251] eta 0:00:33 lr 0.000546 time 0.2498 (0.2754) loss 4.5174 (5.0899) grad_norm 1.6326 (1.8468) [2021-04-14 18:09:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][1140/1251] eta 0:00:30 lr 0.000546 time 0.2611 (0.2754) loss 4.9389 (5.0859) grad_norm 1.9427 (1.8471) [2021-04-14 18:09:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][1150/1251] eta 0:00:27 lr 0.000546 time 0.2611 (0.2755) loss 4.2049 (5.0832) grad_norm 1.4559 (1.8480) [2021-04-14 18:09:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][1160/1251] eta 0:00:25 lr 0.000547 time 0.2889 (0.2754) loss 4.7640 (5.0843) grad_norm 1.6062 (1.8475) [2021-04-14 18:09:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][1170/1251] eta 0:00:22 lr 0.000547 time 0.2912 (0.2754) loss 5.2760 (5.0856) grad_norm 1.4226 (1.8489) [2021-04-14 18:09:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][1180/1251] eta 0:00:19 lr 0.000548 time 0.2485 (0.2754) loss 4.3868 (5.0840) grad_norm 1.8226 (1.8470) [2021-04-14 18:09:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][1190/1251] eta 0:00:16 lr 0.000548 time 0.2780 (0.2753) loss 4.2907 (5.0829) grad_norm 2.2245 (1.8477) [2021-04-14 18:09:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][1200/1251] eta 0:00:14 lr 0.000548 time 0.2717 (0.2753) loss 5.6671 (5.0836) grad_norm 1.9915 (1.8482) [2021-04-14 18:10:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][1210/1251] eta 0:00:11 lr 0.000549 time 0.2605 (0.2752) loss 5.2623 (5.0836) grad_norm 1.3672 (1.8472) [2021-04-14 18:10:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][1220/1251] eta 0:00:08 lr 0.000549 time 0.2508 (0.2751) loss 5.0866 (5.0826) grad_norm 2.0095 (1.8481) [2021-04-14 18:10:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][1230/1251] eta 0:00:05 lr 0.000550 time 0.2800 (0.2751) loss 4.8117 (5.0823) grad_norm 1.7526 (1.8490) [2021-04-14 18:10:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][1240/1251] eta 0:00:03 lr 0.000550 time 0.2487 (0.2750) loss 5.2502 (5.0839) grad_norm 1.9373 (1.8465) [2021-04-14 18:10:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [10/300][1250/1251] eta 0:00:00 lr 0.000550 time 0.2484 (0.2748) loss 4.9384 (5.0843) grad_norm 1.4771 (1.8446) [2021-04-14 18:10:14 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 10 training takes 0:05:45 [2021-04-14 18:10:14 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_10.pth saving...... [2021-04-14 18:10:32 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_10.pth saved !!! [2021-04-14 18:10:33 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.271 (1.271) Loss 2.4788 (2.4788) Acc@1 47.949 (47.949) Acc@5 73.340 (73.340) [2021-04-14 18:10:35 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.110 (0.271) Loss 2.4784 (2.4945) Acc@1 46.777 (46.191) Acc@5 73.340 (72.328) [2021-04-14 18:10:37 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.099 (0.227) Loss 2.4040 (2.4862) Acc@1 47.266 (46.238) Acc@5 73.828 (72.368) [2021-04-14 18:10:39 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.136 (0.233) Loss 2.4536 (2.4861) Acc@1 47.168 (46.399) Acc@5 73.242 (72.240) [2021-04-14 18:10:41 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.216) Loss 2.5613 (2.4823) Acc@1 45.020 (46.461) Acc@5 70.703 (72.168) [2021-04-14 18:10:43 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 46.622 Acc@5 72.170 [2021-04-14 18:10:43 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 46.6% [2021-04-14 18:10:43 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 46.62% [2021-04-14 18:10:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][0/1251] eta 1:24:34 lr 0.000550 time 4.0562 (4.0562) loss 5.4498 (5.4498) grad_norm 1.7059 (1.7059) [2021-04-14 18:10:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][10/1251] eta 0:12:48 lr 0.000551 time 0.2693 (0.6191) loss 5.9288 (5.0781) grad_norm 1.9120 (1.7305) [2021-04-14 18:10:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][20/1251] eta 0:09:19 lr 0.000551 time 0.2680 (0.4545) loss 5.2845 (5.2228) grad_norm 1.6894 (1.7638) [2021-04-14 18:10:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][30/1251] eta 0:08:02 lr 0.000552 time 0.2704 (0.3953) loss 4.4810 (5.0855) grad_norm 1.9091 (1.8164) [2021-04-14 18:10:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][40/1251] eta 0:07:21 lr 0.000552 time 0.2707 (0.3648) loss 4.8508 (5.0538) grad_norm 2.1527 (1.8026) [2021-04-14 18:11:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][50/1251] eta 0:06:56 lr 0.000552 time 0.2837 (0.3467) loss 4.8719 (5.0054) grad_norm 1.5207 (1.7699) [2021-04-14 18:11:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][60/1251] eta 0:06:37 lr 0.000553 time 0.2785 (0.3338) loss 5.8523 (4.9889) grad_norm 1.5740 (1.7644) [2021-04-14 18:11:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][70/1251] eta 0:06:23 lr 0.000553 time 0.2635 (0.3250) loss 3.9145 (4.9735) grad_norm 1.7026 (1.7595) [2021-04-14 18:11:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][80/1251] eta 0:06:13 lr 0.000554 time 0.2716 (0.3187) loss 5.6608 (5.0087) grad_norm 1.9978 (1.7551) [2021-04-14 18:11:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][90/1251] eta 0:06:03 lr 0.000554 time 0.2770 (0.3133) loss 4.8575 (4.9925) grad_norm 1.7300 (1.7489) [2021-04-14 18:11:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][100/1251] eta 0:05:56 lr 0.000554 time 0.2740 (0.3098) loss 5.1028 (5.0042) grad_norm 2.1855 (1.7599) [2021-04-14 18:11:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][110/1251] eta 0:05:49 lr 0.000555 time 0.2766 (0.3062) loss 5.3347 (5.0128) grad_norm 1.6233 (1.7569) [2021-04-14 18:11:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][120/1251] eta 0:05:43 lr 0.000555 time 0.2968 (0.3038) loss 5.3992 (4.9992) grad_norm 1.5367 (1.7502) [2021-04-14 18:11:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][130/1251] eta 0:05:37 lr 0.000556 time 0.2691 (0.3013) loss 5.7903 (5.0152) grad_norm 1.4896 (1.7540) [2021-04-14 18:11:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][140/1251] eta 0:05:32 lr 0.000556 time 0.2779 (0.2995) loss 4.8223 (5.0072) grad_norm 1.8254 (1.7691) [2021-04-14 18:11:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][150/1251] eta 0:05:27 lr 0.000556 time 0.2412 (0.2975) loss 5.1163 (5.0163) grad_norm 2.5664 (1.7811) [2021-04-14 18:11:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][160/1251] eta 0:05:22 lr 0.000557 time 0.2966 (0.2960) loss 5.4738 (4.9938) grad_norm 2.1528 (1.7946) [2021-04-14 18:11:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][170/1251] eta 0:05:18 lr 0.000557 time 0.2611 (0.2944) loss 4.9977 (4.9900) grad_norm 1.5268 (1.7971) [2021-04-14 18:11:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][180/1251] eta 0:05:13 lr 0.000558 time 0.2569 (0.2931) loss 5.1990 (4.9904) grad_norm 1.7480 (1.7902) [2021-04-14 18:11:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][190/1251] eta 0:05:09 lr 0.000558 time 0.2824 (0.2920) loss 5.1201 (4.9747) grad_norm 1.4067 (1.7827) [2021-04-14 18:11:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][200/1251] eta 0:05:05 lr 0.000558 time 0.2763 (0.2909) loss 4.8185 (4.9706) grad_norm 1.4525 (1.7865) [2021-04-14 18:11:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][210/1251] eta 0:05:02 lr 0.000559 time 0.2870 (0.2901) loss 5.0921 (4.9764) grad_norm 1.5474 (1.7847) [2021-04-14 18:11:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][220/1251] eta 0:04:58 lr 0.000559 time 0.2538 (0.2892) loss 5.5881 (4.9755) grad_norm 1.7379 (1.7793) [2021-04-14 18:11:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][230/1251] eta 0:04:55 lr 0.000560 time 0.2708 (0.2890) loss 5.5760 (4.9926) grad_norm 1.7631 (1.7766) [2021-04-14 18:11:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][240/1251] eta 0:04:51 lr 0.000560 time 0.2821 (0.2883) loss 5.8280 (5.0032) grad_norm 1.9546 (1.7772) [2021-04-14 18:11:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][250/1251] eta 0:04:47 lr 0.000560 time 0.2510 (0.2876) loss 4.9650 (5.0121) grad_norm 1.7077 (1.7772) [2021-04-14 18:11:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][260/1251] eta 0:04:44 lr 0.000561 time 0.2524 (0.2871) loss 4.8494 (5.0095) grad_norm 1.5307 (1.7793) [2021-04-14 18:12:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][270/1251] eta 0:04:41 lr 0.000561 time 0.2757 (0.2866) loss 4.2596 (5.0083) grad_norm 1.6543 (1.7772) [2021-04-14 18:12:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][280/1251] eta 0:04:37 lr 0.000562 time 0.2654 (0.2860) loss 4.8035 (5.0034) grad_norm 1.9753 (1.7799) [2021-04-14 18:12:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][290/1251] eta 0:04:34 lr 0.000562 time 0.2778 (0.2856) loss 5.3086 (5.0092) grad_norm 1.7907 (1.7831) [2021-04-14 18:12:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][300/1251] eta 0:04:31 lr 0.000562 time 0.2649 (0.2851) loss 5.4761 (5.0067) grad_norm 2.5685 (1.7841) [2021-04-14 18:12:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][310/1251] eta 0:04:27 lr 0.000563 time 0.2679 (0.2845) loss 5.7416 (5.0176) grad_norm 1.2885 (1.7767) [2021-04-14 18:12:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][320/1251] eta 0:04:24 lr 0.000563 time 0.2645 (0.2844) loss 5.7602 (5.0221) grad_norm 1.6171 (1.7708) [2021-04-14 18:12:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][330/1251] eta 0:04:21 lr 0.000564 time 0.2851 (0.2843) loss 4.1179 (5.0259) grad_norm 1.4046 (1.7651) [2021-04-14 18:12:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][340/1251] eta 0:04:18 lr 0.000564 time 0.2824 (0.2841) loss 5.3333 (5.0288) grad_norm 2.1883 (1.7644) [2021-04-14 18:12:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][350/1251] eta 0:04:15 lr 0.000564 time 0.2910 (0.2840) loss 3.5535 (5.0183) grad_norm 2.1759 (1.7637) [2021-04-14 18:12:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][360/1251] eta 0:04:13 lr 0.000565 time 0.2780 (0.2843) loss 5.0193 (5.0163) grad_norm 1.9363 (1.7604) [2021-04-14 18:12:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][370/1251] eta 0:04:10 lr 0.000565 time 0.2723 (0.2840) loss 4.6922 (5.0198) grad_norm 1.4717 (1.7625) [2021-04-14 18:12:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][380/1251] eta 0:04:07 lr 0.000566 time 0.3027 (0.2839) loss 5.5585 (5.0221) grad_norm 1.7944 (1.7624) [2021-04-14 18:12:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][390/1251] eta 0:04:04 lr 0.000566 time 0.2593 (0.2836) loss 4.9520 (5.0256) grad_norm 1.7732 (1.7624) [2021-04-14 18:12:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][400/1251] eta 0:04:01 lr 0.000566 time 0.2553 (0.2832) loss 5.5908 (5.0266) grad_norm 1.8591 (1.7691) [2021-04-14 18:12:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][410/1251] eta 0:03:57 lr 0.000567 time 0.2646 (0.2829) loss 5.4400 (5.0228) grad_norm 1.6922 (1.7707) [2021-04-14 18:12:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][420/1251] eta 0:03:54 lr 0.000567 time 0.2739 (0.2827) loss 5.3077 (5.0193) grad_norm 1.4995 (1.7732) [2021-04-14 18:12:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][430/1251] eta 0:03:52 lr 0.000568 time 0.2895 (0.2826) loss 5.2061 (5.0245) grad_norm 2.1856 (1.7741) [2021-04-14 18:12:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][440/1251] eta 0:03:49 lr 0.000568 time 0.2709 (0.2824) loss 4.0536 (5.0276) grad_norm 1.2459 (1.7737) [2021-04-14 18:12:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][450/1251] eta 0:03:46 lr 0.000568 time 0.2711 (0.2822) loss 4.7679 (5.0199) grad_norm 1.9056 (1.7740) [2021-04-14 18:12:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][460/1251] eta 0:03:43 lr 0.000569 time 0.2773 (0.2820) loss 5.1309 (5.0167) grad_norm 1.7204 (1.7734) [2021-04-14 18:12:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][470/1251] eta 0:03:40 lr 0.000569 time 0.2879 (0.2818) loss 5.9140 (5.0140) grad_norm 1.3714 (1.7705) [2021-04-14 18:12:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][480/1251] eta 0:03:37 lr 0.000570 time 0.2586 (0.2817) loss 5.6740 (5.0165) grad_norm 1.7958 (1.7710) [2021-04-14 18:13:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][490/1251] eta 0:03:34 lr 0.000570 time 0.2743 (0.2814) loss 5.1101 (5.0184) grad_norm 1.5868 (inf) [2021-04-14 18:13:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][500/1251] eta 0:03:31 lr 0.000570 time 0.2664 (0.2812) loss 5.1242 (5.0147) grad_norm 1.4819 (inf) [2021-04-14 18:13:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][510/1251] eta 0:03:28 lr 0.000571 time 0.2899 (0.2811) loss 5.1917 (5.0128) grad_norm 1.6097 (inf) [2021-04-14 18:13:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][520/1251] eta 0:03:25 lr 0.000571 time 0.2873 (0.2809) loss 5.2349 (5.0098) grad_norm 2.1056 (inf) [2021-04-14 18:13:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][530/1251] eta 0:03:22 lr 0.000572 time 0.2745 (0.2807) loss 4.9948 (5.0119) grad_norm 1.6854 (inf) [2021-04-14 18:13:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][540/1251] eta 0:03:19 lr 0.000572 time 0.2650 (0.2806) loss 4.7521 (5.0114) grad_norm 1.8476 (inf) [2021-04-14 18:13:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][550/1251] eta 0:03:16 lr 0.000572 time 0.2904 (0.2805) loss 5.0186 (5.0054) grad_norm 1.5812 (inf) [2021-04-14 18:13:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][560/1251] eta 0:03:13 lr 0.000573 time 0.2624 (0.2804) loss 5.0411 (5.0102) grad_norm 2.0120 (inf) [2021-04-14 18:13:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][570/1251] eta 0:03:10 lr 0.000573 time 0.2798 (0.2803) loss 5.3343 (5.0109) grad_norm 1.5590 (inf) [2021-04-14 18:13:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][580/1251] eta 0:03:07 lr 0.000574 time 0.2629 (0.2800) loss 5.7885 (5.0118) grad_norm 1.6521 (inf) [2021-04-14 18:13:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][590/1251] eta 0:03:05 lr 0.000574 time 0.2775 (0.2799) loss 5.1695 (5.0115) grad_norm 1.5281 (inf) [2021-04-14 18:13:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][600/1251] eta 0:03:02 lr 0.000574 time 0.2831 (0.2799) loss 5.5034 (5.0091) grad_norm 2.0235 (inf) [2021-04-14 18:13:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][610/1251] eta 0:02:59 lr 0.000575 time 0.2844 (0.2797) loss 5.0055 (5.0108) grad_norm 1.3440 (inf) [2021-04-14 18:13:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][620/1251] eta 0:02:56 lr 0.000575 time 0.2683 (0.2797) loss 4.8401 (5.0126) grad_norm 1.7790 (inf) [2021-04-14 18:13:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][630/1251] eta 0:02:53 lr 0.000576 time 0.2828 (0.2797) loss 5.3529 (5.0131) grad_norm 1.7051 (inf) [2021-04-14 18:13:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][640/1251] eta 0:02:50 lr 0.000576 time 0.2606 (0.2796) loss 4.7032 (5.0171) grad_norm 1.5997 (inf) [2021-04-14 18:13:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][650/1251] eta 0:02:47 lr 0.000576 time 0.2721 (0.2795) loss 5.6035 (5.0191) grad_norm 1.4943 (inf) [2021-04-14 18:13:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][660/1251] eta 0:02:45 lr 0.000577 time 0.2608 (0.2794) loss 5.4863 (5.0165) grad_norm 1.5777 (inf) [2021-04-14 18:13:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][670/1251] eta 0:02:42 lr 0.000577 time 0.2730 (0.2792) loss 4.8015 (5.0211) grad_norm 1.6928 (inf) [2021-04-14 18:13:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][680/1251] eta 0:02:39 lr 0.000578 time 0.2779 (0.2792) loss 5.2301 (5.0208) grad_norm 2.0265 (inf) [2021-04-14 18:13:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][690/1251] eta 0:02:36 lr 0.000578 time 0.2735 (0.2790) loss 5.4486 (5.0225) grad_norm 1.6202 (inf) [2021-04-14 18:13:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][700/1251] eta 0:02:33 lr 0.000578 time 0.2869 (0.2790) loss 5.3190 (5.0247) grad_norm 1.9498 (inf) [2021-04-14 18:14:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][710/1251] eta 0:02:30 lr 0.000579 time 0.2603 (0.2788) loss 5.3953 (5.0288) grad_norm 1.5968 (inf) [2021-04-14 18:14:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][720/1251] eta 0:02:28 lr 0.000579 time 0.2830 (0.2787) loss 4.8675 (5.0268) grad_norm 1.5507 (inf) [2021-04-14 18:14:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][730/1251] eta 0:02:25 lr 0.000580 time 0.2665 (0.2787) loss 5.7797 (5.0319) grad_norm 1.8159 (inf) [2021-04-14 18:14:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][740/1251] eta 0:02:22 lr 0.000580 time 0.2732 (0.2785) loss 5.1256 (5.0302) grad_norm 1.8968 (inf) [2021-04-14 18:14:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][750/1251] eta 0:02:19 lr 0.000580 time 0.2865 (0.2784) loss 5.4723 (5.0294) grad_norm 1.5022 (inf) [2021-04-14 18:14:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][760/1251] eta 0:02:16 lr 0.000581 time 0.2828 (0.2783) loss 5.0560 (5.0309) grad_norm 1.5107 (inf) [2021-04-14 18:14:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][770/1251] eta 0:02:13 lr 0.000581 time 0.2420 (0.2781) loss 5.0710 (5.0312) grad_norm 2.4867 (inf) [2021-04-14 18:14:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][780/1251] eta 0:02:10 lr 0.000582 time 0.2722 (0.2781) loss 4.1028 (5.0296) grad_norm 1.4278 (inf) [2021-04-14 18:14:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][790/1251] eta 0:02:08 lr 0.000582 time 0.2569 (0.2779) loss 5.3991 (5.0310) grad_norm 2.2124 (inf) [2021-04-14 18:14:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][800/1251] eta 0:02:05 lr 0.000582 time 0.2731 (0.2779) loss 4.9343 (5.0320) grad_norm 1.3910 (inf) [2021-04-14 18:14:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][810/1251] eta 0:02:02 lr 0.000583 time 0.2804 (0.2779) loss 4.5850 (5.0308) grad_norm 1.5140 (inf) [2021-04-14 18:14:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][820/1251] eta 0:01:59 lr 0.000583 time 0.2779 (0.2778) loss 5.7285 (5.0313) grad_norm 1.4526 (inf) [2021-04-14 18:14:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][830/1251] eta 0:01:56 lr 0.000584 time 0.2740 (0.2778) loss 5.6527 (5.0321) grad_norm 1.6646 (inf) [2021-04-14 18:14:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][840/1251] eta 0:01:54 lr 0.000584 time 0.2686 (0.2777) loss 5.4262 (5.0329) grad_norm 1.7857 (inf) [2021-04-14 18:14:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][850/1251] eta 0:01:51 lr 0.000584 time 0.2899 (0.2776) loss 5.0738 (5.0363) grad_norm 1.8385 (inf) [2021-04-14 18:14:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][860/1251] eta 0:01:48 lr 0.000585 time 0.2809 (0.2776) loss 4.5315 (5.0378) grad_norm 1.6162 (inf) [2021-04-14 18:14:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][870/1251] eta 0:01:45 lr 0.000585 time 0.2786 (0.2775) loss 4.8092 (5.0378) grad_norm 1.9000 (inf) [2021-04-14 18:14:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][880/1251] eta 0:01:42 lr 0.000586 time 0.2635 (0.2775) loss 5.1431 (5.0337) grad_norm 2.3095 (inf) [2021-04-14 18:14:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][890/1251] eta 0:01:40 lr 0.000586 time 0.2658 (0.2776) loss 4.2085 (5.0313) grad_norm 2.1051 (inf) [2021-04-14 18:14:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][900/1251] eta 0:01:37 lr 0.000586 time 0.4341 (0.2777) loss 4.1084 (5.0328) grad_norm 1.9198 (inf) [2021-04-14 18:14:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][910/1251] eta 0:01:34 lr 0.000587 time 0.2754 (0.2777) loss 5.7667 (5.0319) grad_norm 1.5589 (inf) [2021-04-14 18:14:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][920/1251] eta 0:01:31 lr 0.000587 time 0.2889 (0.2776) loss 5.1641 (5.0292) grad_norm 1.9748 (inf) [2021-04-14 18:15:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][930/1251] eta 0:01:29 lr 0.000588 time 0.2842 (0.2776) loss 4.2926 (5.0259) grad_norm 1.6966 (inf) [2021-04-14 18:15:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][940/1251] eta 0:01:26 lr 0.000588 time 0.2705 (0.2776) loss 5.7614 (5.0212) grad_norm 1.7054 (inf) [2021-04-14 18:15:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][950/1251] eta 0:01:23 lr 0.000588 time 0.2557 (0.2776) loss 5.2116 (5.0198) grad_norm 2.6056 (inf) [2021-04-14 18:15:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][960/1251] eta 0:01:20 lr 0.000589 time 0.2682 (0.2775) loss 3.6265 (5.0168) grad_norm 1.5356 (inf) [2021-04-14 18:15:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][970/1251] eta 0:01:17 lr 0.000589 time 0.2571 (0.2774) loss 5.4084 (5.0166) grad_norm 1.4540 (inf) [2021-04-14 18:15:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][980/1251] eta 0:01:15 lr 0.000590 time 0.2646 (0.2773) loss 4.1358 (5.0153) grad_norm 1.3934 (inf) [2021-04-14 18:15:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][990/1251] eta 0:01:12 lr 0.000590 time 0.2596 (0.2772) loss 5.0449 (5.0146) grad_norm 2.0360 (inf) [2021-04-14 18:15:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][1000/1251] eta 0:01:09 lr 0.000590 time 0.2752 (0.2772) loss 5.3208 (5.0115) grad_norm 1.8793 (inf) [2021-04-14 18:15:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][1010/1251] eta 0:01:06 lr 0.000591 time 0.2637 (0.2771) loss 3.5484 (5.0081) grad_norm 2.1624 (inf) [2021-04-14 18:15:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][1020/1251] eta 0:01:04 lr 0.000591 time 0.2612 (0.2771) loss 5.6624 (5.0075) grad_norm 1.7541 (inf) [2021-04-14 18:15:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][1030/1251] eta 0:01:01 lr 0.000592 time 0.2748 (0.2770) loss 5.0783 (5.0086) grad_norm 1.6196 (inf) [2021-04-14 18:15:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][1040/1251] eta 0:00:58 lr 0.000592 time 0.2724 (0.2770) loss 4.9327 (5.0094) grad_norm 1.9959 (inf) [2021-04-14 18:15:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][1050/1251] eta 0:00:55 lr 0.000592 time 0.2736 (0.2769) loss 4.4569 (5.0085) grad_norm 2.2153 (inf) [2021-04-14 18:15:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][1060/1251] eta 0:00:52 lr 0.000593 time 0.2537 (0.2768) loss 4.3109 (5.0068) grad_norm 1.9554 (inf) [2021-04-14 18:15:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][1070/1251] eta 0:00:50 lr 0.000593 time 0.2592 (0.2768) loss 3.9369 (5.0062) grad_norm 1.6895 (inf) [2021-04-14 18:15:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][1080/1251] eta 0:00:47 lr 0.000594 time 0.2438 (0.2767) loss 4.7040 (5.0077) grad_norm 1.8474 (inf) [2021-04-14 18:15:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][1090/1251] eta 0:00:44 lr 0.000594 time 0.2403 (0.2767) loss 3.9567 (5.0073) grad_norm 1.8262 (inf) [2021-04-14 18:15:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][1100/1251] eta 0:00:41 lr 0.000594 time 0.2798 (0.2767) loss 5.4028 (5.0067) grad_norm 1.6450 (inf) [2021-04-14 18:15:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][1110/1251] eta 0:00:39 lr 0.000595 time 0.2723 (0.2767) loss 4.6151 (5.0065) grad_norm 1.4227 (inf) [2021-04-14 18:15:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][1120/1251] eta 0:00:36 lr 0.000595 time 0.2659 (0.2766) loss 5.4336 (5.0080) grad_norm 1.4747 (inf) [2021-04-14 18:15:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][1130/1251] eta 0:00:33 lr 0.000596 time 0.2799 (0.2766) loss 5.2158 (5.0085) grad_norm 2.9135 (inf) [2021-04-14 18:15:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][1140/1251] eta 0:00:30 lr 0.000596 time 0.2922 (0.2765) loss 4.1042 (5.0066) grad_norm 1.4965 (inf) [2021-04-14 18:16:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][1150/1251] eta 0:00:27 lr 0.000596 time 0.2732 (0.2765) loss 5.2994 (5.0070) grad_norm 1.3462 (inf) [2021-04-14 18:16:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][1160/1251] eta 0:00:25 lr 0.000597 time 0.2598 (0.2764) loss 5.2902 (5.0079) grad_norm 1.5919 (inf) [2021-04-14 18:16:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][1170/1251] eta 0:00:22 lr 0.000597 time 0.2875 (0.2764) loss 4.7225 (5.0042) grad_norm 1.4002 (inf) [2021-04-14 18:16:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][1180/1251] eta 0:00:19 lr 0.000598 time 0.2811 (0.2764) loss 4.2419 (5.0048) grad_norm 1.5713 (inf) [2021-04-14 18:16:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][1190/1251] eta 0:00:16 lr 0.000598 time 0.2629 (0.2764) loss 5.5764 (5.0053) grad_norm 1.7200 (inf) [2021-04-14 18:16:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][1200/1251] eta 0:00:14 lr 0.000598 time 0.2692 (0.2763) loss 4.9322 (5.0044) grad_norm 1.9297 (inf) [2021-04-14 18:16:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][1210/1251] eta 0:00:11 lr 0.000599 time 0.2789 (0.2763) loss 3.8492 (5.0019) grad_norm 1.4699 (inf) [2021-04-14 18:16:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][1220/1251] eta 0:00:08 lr 0.000599 time 0.2710 (0.2763) loss 5.1672 (5.0010) grad_norm 1.8665 (inf) [2021-04-14 18:16:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][1230/1251] eta 0:00:05 lr 0.000600 time 0.2774 (0.2763) loss 5.2262 (5.0004) grad_norm 1.8006 (inf) [2021-04-14 18:16:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][1240/1251] eta 0:00:03 lr 0.000600 time 0.2479 (0.2762) loss 5.2366 (5.0013) grad_norm 2.8865 (inf) [2021-04-14 18:16:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [11/300][1250/1251] eta 0:00:00 lr 0.000600 time 0.2503 (0.2759) loss 5.1918 (5.0001) grad_norm 1.2632 (inf) [2021-04-14 18:16:30 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 11 training takes 0:05:46 [2021-04-14 18:16:30 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_11.pth saving...... [2021-04-14 18:16:42 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_11.pth saved !!! [2021-04-14 18:16:43 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.264 (1.264) Loss 2.4367 (2.4367) Acc@1 47.656 (47.656) Acc@5 73.047 (73.047) [2021-04-14 18:16:44 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.156 (0.213) Loss 2.4339 (2.4143) Acc@1 47.168 (47.976) Acc@5 73.438 (73.695) [2021-04-14 18:16:46 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.088 (0.208) Loss 2.4709 (2.4086) Acc@1 47.363 (48.261) Acc@5 73.633 (74.047) [2021-04-14 18:16:49 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.120 (0.228) Loss 2.3341 (2.4090) Acc@1 50.488 (48.321) Acc@5 75.391 (73.907) [2021-04-14 18:16:50 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.085 (0.208) Loss 2.3680 (2.4048) Acc@1 48.438 (48.423) Acc@5 74.902 (74.007) [2021-04-14 18:16:52 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 48.484 Acc@5 74.040 [2021-04-14 18:16:52 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 48.5% [2021-04-14 18:16:52 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 48.48% [2021-04-14 18:16:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][0/1251] eta 1:14:02 lr 0.000600 time 3.5515 (3.5515) loss 5.5529 (5.5529) grad_norm 1.5223 (1.5223) [2021-04-14 18:16:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][10/1251] eta 0:12:19 lr 0.000601 time 0.4204 (0.5956) loss 4.9566 (4.8987) grad_norm 1.2318 (1.6730) [2021-04-14 18:17:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][20/1251] eta 0:09:02 lr 0.000601 time 0.2557 (0.4404) loss 4.8866 (4.8227) grad_norm 1.5923 (1.6969) [2021-04-14 18:17:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][30/1251] eta 0:07:51 lr 0.000602 time 0.2844 (0.3863) loss 5.4164 (4.8696) grad_norm 1.9163 (1.6571) [2021-04-14 18:17:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][40/1251] eta 0:07:15 lr 0.000602 time 0.2685 (0.3596) loss 5.5604 (4.8306) grad_norm 2.0384 (1.6779) [2021-04-14 18:17:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][50/1251] eta 0:06:51 lr 0.000602 time 0.2695 (0.3427) loss 5.3813 (4.8138) grad_norm 1.4024 (1.6809) [2021-04-14 18:17:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][60/1251] eta 0:06:34 lr 0.000603 time 0.2786 (0.3314) loss 5.3558 (4.7774) grad_norm 2.4308 (1.7053) [2021-04-14 18:17:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][70/1251] eta 0:06:21 lr 0.000603 time 0.2912 (0.3232) loss 5.0720 (4.8237) grad_norm 1.4747 (1.6977) [2021-04-14 18:17:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][80/1251] eta 0:06:10 lr 0.000604 time 0.2675 (0.3165) loss 5.2064 (4.8417) grad_norm 1.6924 (1.6869) [2021-04-14 18:17:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][90/1251] eta 0:06:02 lr 0.000604 time 0.2838 (0.3118) loss 5.4813 (4.8467) grad_norm 1.5744 (1.6840) [2021-04-14 18:17:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][100/1251] eta 0:05:56 lr 0.000604 time 0.2623 (0.3099) loss 4.7157 (4.8532) grad_norm 1.5993 (1.6805) [2021-04-14 18:17:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][110/1251] eta 0:05:50 lr 0.000605 time 0.4215 (0.3076) loss 5.3527 (4.8637) grad_norm 2.0627 (1.6922) [2021-04-14 18:17:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][120/1251] eta 0:05:44 lr 0.000605 time 0.2874 (0.3049) loss 4.7198 (4.8535) grad_norm 1.9845 (1.6978) [2021-04-14 18:17:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][130/1251] eta 0:05:39 lr 0.000606 time 0.2894 (0.3027) loss 4.3669 (4.8677) grad_norm 1.6247 (1.6998) [2021-04-14 18:17:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][140/1251] eta 0:05:34 lr 0.000606 time 0.2539 (0.3011) loss 5.3050 (4.9013) grad_norm 1.9929 (1.7095) [2021-04-14 18:17:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][150/1251] eta 0:05:30 lr 0.000606 time 0.2611 (0.3003) loss 4.7487 (4.8991) grad_norm 1.5085 (1.7147) [2021-04-14 18:17:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][160/1251] eta 0:05:26 lr 0.000607 time 0.2916 (0.2989) loss 5.1138 (4.9039) grad_norm 1.7195 (1.7101) [2021-04-14 18:17:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][170/1251] eta 0:05:21 lr 0.000607 time 0.2696 (0.2976) loss 4.9180 (4.9147) grad_norm 1.3566 (1.7032) [2021-04-14 18:17:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][180/1251] eta 0:05:17 lr 0.000608 time 0.2827 (0.2962) loss 3.7869 (4.9044) grad_norm 1.7358 (1.7008) [2021-04-14 18:17:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][190/1251] eta 0:05:12 lr 0.000608 time 0.2667 (0.2950) loss 4.7296 (4.9040) grad_norm 2.3720 (1.7332) [2021-04-14 18:17:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][200/1251] eta 0:05:08 lr 0.000608 time 0.2620 (0.2938) loss 5.1312 (4.9132) grad_norm 1.5251 (1.7394) [2021-04-14 18:17:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][210/1251] eta 0:05:05 lr 0.000609 time 0.2742 (0.2931) loss 5.2094 (4.9205) grad_norm 1.3210 (1.7417) [2021-04-14 18:17:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][220/1251] eta 0:05:01 lr 0.000609 time 0.2764 (0.2923) loss 5.8944 (4.9277) grad_norm 1.8729 (1.7376) [2021-04-14 18:17:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][230/1251] eta 0:04:57 lr 0.000610 time 0.2698 (0.2918) loss 3.6802 (4.9205) grad_norm 1.7331 (1.7425) [2021-04-14 18:18:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][240/1251] eta 0:04:54 lr 0.000610 time 0.2883 (0.2911) loss 3.7216 (4.9275) grad_norm 1.6164 (1.7352) [2021-04-14 18:18:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][250/1251] eta 0:04:50 lr 0.000610 time 0.2696 (0.2905) loss 5.4627 (4.9262) grad_norm 1.5315 (1.7383) [2021-04-14 18:18:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][260/1251] eta 0:04:47 lr 0.000611 time 0.2811 (0.2899) loss 5.3484 (4.9291) grad_norm 2.1521 (1.7343) [2021-04-14 18:18:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][270/1251] eta 0:04:43 lr 0.000611 time 0.2643 (0.2891) loss 4.9325 (4.9380) grad_norm 1.7981 (1.7251) [2021-04-14 18:18:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][280/1251] eta 0:04:40 lr 0.000612 time 0.2777 (0.2885) loss 5.5533 (4.9416) grad_norm 1.7357 (1.7224) [2021-04-14 18:18:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][290/1251] eta 0:04:36 lr 0.000612 time 0.2633 (0.2879) loss 4.6088 (4.9422) grad_norm 1.4651 (1.7216) [2021-04-14 18:18:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][300/1251] eta 0:04:33 lr 0.000612 time 0.3029 (0.2876) loss 5.2584 (4.9394) grad_norm 1.8005 (1.7216) [2021-04-14 18:18:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][310/1251] eta 0:04:30 lr 0.000613 time 0.2519 (0.2870) loss 4.7179 (4.9299) grad_norm 2.3469 (1.7290) [2021-04-14 18:18:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][320/1251] eta 0:04:26 lr 0.000613 time 0.2778 (0.2865) loss 5.1707 (4.9219) grad_norm 1.4288 (1.7289) [2021-04-14 18:18:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][330/1251] eta 0:04:23 lr 0.000614 time 0.2591 (0.2860) loss 5.4170 (4.9207) grad_norm 1.8629 (1.7363) [2021-04-14 18:18:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][340/1251] eta 0:04:20 lr 0.000614 time 0.3015 (0.2856) loss 5.3822 (4.9266) grad_norm 2.0357 (1.7316) [2021-04-14 18:18:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][350/1251] eta 0:04:16 lr 0.000614 time 0.2666 (0.2852) loss 4.8828 (4.9261) grad_norm 1.8800 (1.7297) [2021-04-14 18:18:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][360/1251] eta 0:04:13 lr 0.000615 time 0.2632 (0.2848) loss 4.9702 (4.9274) grad_norm 1.4816 (1.7332) [2021-04-14 18:18:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][370/1251] eta 0:04:10 lr 0.000615 time 0.2954 (0.2849) loss 5.3659 (4.9298) grad_norm 1.5456 (1.7323) [2021-04-14 18:18:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][380/1251] eta 0:04:07 lr 0.000616 time 0.2645 (0.2845) loss 3.5928 (4.9174) grad_norm 1.8604 (1.7274) [2021-04-14 18:18:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][390/1251] eta 0:04:04 lr 0.000616 time 0.2612 (0.2841) loss 5.3752 (4.9184) grad_norm 1.2905 (1.7284) [2021-04-14 18:18:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][400/1251] eta 0:04:01 lr 0.000616 time 0.2728 (0.2838) loss 5.0066 (4.9217) grad_norm 1.8704 (1.7266) [2021-04-14 18:18:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][410/1251] eta 0:03:58 lr 0.000617 time 0.2683 (0.2836) loss 4.5899 (4.9237) grad_norm 1.7764 (1.7301) [2021-04-14 18:18:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][420/1251] eta 0:03:55 lr 0.000617 time 0.2792 (0.2834) loss 4.1178 (4.9180) grad_norm 1.8576 (1.7260) [2021-04-14 18:18:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][430/1251] eta 0:03:52 lr 0.000618 time 0.2686 (0.2833) loss 4.7313 (4.9191) grad_norm 1.4662 (1.7224) [2021-04-14 18:18:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][440/1251] eta 0:03:49 lr 0.000618 time 0.2880 (0.2833) loss 4.0002 (4.9185) grad_norm 1.3274 (1.7256) [2021-04-14 18:19:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][450/1251] eta 0:03:46 lr 0.000618 time 0.2745 (0.2831) loss 4.6949 (4.9171) grad_norm 1.3938 (1.7246) [2021-04-14 18:19:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][460/1251] eta 0:03:43 lr 0.000619 time 0.2916 (0.2829) loss 5.4677 (4.9088) grad_norm 1.5668 (1.7254) [2021-04-14 18:19:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][470/1251] eta 0:03:40 lr 0.000619 time 0.2982 (0.2828) loss 4.5493 (4.9068) grad_norm 1.5822 (1.7212) [2021-04-14 18:19:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][480/1251] eta 0:03:38 lr 0.000620 time 0.2904 (0.2828) loss 5.4625 (4.9126) grad_norm 1.9868 (1.7205) [2021-04-14 18:19:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][490/1251] eta 0:03:35 lr 0.000620 time 0.2577 (0.2826) loss 5.2896 (4.9136) grad_norm 1.7261 (1.7192) [2021-04-14 18:19:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][500/1251] eta 0:03:32 lr 0.000620 time 0.2779 (0.2825) loss 5.5068 (4.9147) grad_norm 1.5082 (1.7192) [2021-04-14 18:19:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][510/1251] eta 0:03:29 lr 0.000621 time 0.3037 (0.2825) loss 4.7356 (4.9208) grad_norm 1.4339 (1.7168) [2021-04-14 18:19:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][520/1251] eta 0:03:26 lr 0.000621 time 0.2902 (0.2824) loss 4.5592 (4.9177) grad_norm 1.4371 (1.7161) [2021-04-14 18:19:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][530/1251] eta 0:03:23 lr 0.000622 time 0.2778 (0.2822) loss 5.3437 (4.9178) grad_norm 1.4119 (1.7130) [2021-04-14 18:19:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][540/1251] eta 0:03:20 lr 0.000622 time 0.2823 (0.2820) loss 5.2906 (4.9168) grad_norm 1.9386 (1.7140) [2021-04-14 18:19:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][550/1251] eta 0:03:17 lr 0.000622 time 0.2590 (0.2822) loss 5.1947 (4.9186) grad_norm 1.6387 (1.7124) [2021-04-14 18:19:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][560/1251] eta 0:03:15 lr 0.000623 time 0.2746 (0.2823) loss 5.6693 (4.9235) grad_norm 1.4278 (1.7108) [2021-04-14 18:19:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][570/1251] eta 0:03:12 lr 0.000623 time 0.2639 (0.2821) loss 4.1060 (4.9216) grad_norm 1.7877 (1.7093) [2021-04-14 18:19:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][580/1251] eta 0:03:09 lr 0.000624 time 0.2721 (0.2819) loss 4.4192 (4.9180) grad_norm 1.3996 (1.7070) [2021-04-14 18:19:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][590/1251] eta 0:03:06 lr 0.000624 time 0.2807 (0.2821) loss 4.8984 (4.9137) grad_norm 1.4235 (1.7056) [2021-04-14 18:19:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][600/1251] eta 0:03:03 lr 0.000624 time 0.2806 (0.2818) loss 5.4523 (4.9173) grad_norm 1.3868 (1.7022) [2021-04-14 18:19:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][610/1251] eta 0:03:00 lr 0.000625 time 0.2738 (0.2817) loss 4.7506 (4.9190) grad_norm 1.9277 (1.7003) [2021-04-14 18:19:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][620/1251] eta 0:02:57 lr 0.000625 time 0.2729 (0.2815) loss 4.6730 (4.9203) grad_norm 1.8074 (1.7028) [2021-04-14 18:19:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][630/1251] eta 0:02:54 lr 0.000626 time 0.2955 (0.2814) loss 3.7011 (4.9174) grad_norm 1.9844 (1.7002) [2021-04-14 18:19:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][640/1251] eta 0:02:51 lr 0.000626 time 0.2729 (0.2812) loss 5.0153 (4.9141) grad_norm 1.7046 (1.6994) [2021-04-14 18:19:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][650/1251] eta 0:02:48 lr 0.000626 time 0.2617 (0.2811) loss 3.8490 (4.9161) grad_norm 1.6870 (1.6984) [2021-04-14 18:19:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][660/1251] eta 0:02:46 lr 0.000627 time 0.2558 (0.2809) loss 5.3146 (4.9165) grad_norm 1.4413 (1.6998) [2021-04-14 18:20:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][670/1251] eta 0:02:43 lr 0.000627 time 0.2793 (0.2808) loss 5.3007 (4.9161) grad_norm 1.7578 (1.6981) [2021-04-14 18:20:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][680/1251] eta 0:02:40 lr 0.000628 time 0.2817 (0.2809) loss 5.5369 (4.9135) grad_norm 1.6681 (1.6975) [2021-04-14 18:20:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][690/1251] eta 0:02:37 lr 0.000628 time 0.2532 (0.2809) loss 5.3181 (4.9145) grad_norm 1.4441 (1.6957) [2021-04-14 18:20:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][700/1251] eta 0:02:34 lr 0.000628 time 0.2731 (0.2807) loss 5.0408 (4.9138) grad_norm 1.4349 (1.6960) [2021-04-14 18:20:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][710/1251] eta 0:02:31 lr 0.000629 time 0.2606 (0.2806) loss 4.9870 (4.9136) grad_norm 1.7454 (1.6959) [2021-04-14 18:20:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][720/1251] eta 0:02:28 lr 0.000629 time 0.2811 (0.2805) loss 5.6885 (4.9140) grad_norm 1.4045 (1.6939) [2021-04-14 18:20:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][730/1251] eta 0:02:26 lr 0.000630 time 0.2871 (0.2805) loss 5.2009 (4.9127) grad_norm 1.5821 (1.6906) [2021-04-14 18:20:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][740/1251] eta 0:02:23 lr 0.000630 time 0.2665 (0.2804) loss 5.3064 (4.9144) grad_norm 1.5195 (1.6899) [2021-04-14 18:20:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][750/1251] eta 0:02:20 lr 0.000630 time 0.2903 (0.2804) loss 4.5427 (4.9140) grad_norm 1.4935 (1.6904) [2021-04-14 18:20:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][760/1251] eta 0:02:17 lr 0.000631 time 0.2461 (0.2803) loss 4.8639 (4.9126) grad_norm 1.4425 (1.6878) [2021-04-14 18:20:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][770/1251] eta 0:02:14 lr 0.000631 time 0.2832 (0.2802) loss 4.1983 (4.9093) grad_norm 1.8954 (1.6849) [2021-04-14 18:20:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][780/1251] eta 0:02:11 lr 0.000632 time 0.2763 (0.2801) loss 5.7774 (4.9114) grad_norm 2.0135 (1.6850) [2021-04-14 18:20:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][790/1251] eta 0:02:09 lr 0.000632 time 0.2876 (0.2800) loss 4.8621 (4.9107) grad_norm 1.7983 (1.6877) [2021-04-14 18:20:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][800/1251] eta 0:02:06 lr 0.000632 time 0.3009 (0.2801) loss 4.5595 (4.9090) grad_norm 1.5723 (1.6880) [2021-04-14 18:20:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][810/1251] eta 0:02:03 lr 0.000633 time 0.2616 (0.2800) loss 5.5524 (4.9119) grad_norm 1.8096 (1.6855) [2021-04-14 18:20:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][820/1251] eta 0:02:00 lr 0.000633 time 0.3020 (0.2800) loss 5.3530 (4.9138) grad_norm 1.1692 (1.6835) [2021-04-14 18:20:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][830/1251] eta 0:01:57 lr 0.000634 time 0.2853 (0.2799) loss 5.1237 (4.9151) grad_norm 1.2827 (1.6819) [2021-04-14 18:20:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][840/1251] eta 0:01:55 lr 0.000634 time 0.2700 (0.2799) loss 3.9504 (4.9128) grad_norm 1.7323 (1.6829) [2021-04-14 18:20:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][850/1251] eta 0:01:52 lr 0.000634 time 0.2657 (0.2798) loss 4.8985 (4.9164) grad_norm 1.2452 (1.6823) [2021-04-14 18:20:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][860/1251] eta 0:01:49 lr 0.000635 time 0.2846 (0.2797) loss 4.4079 (4.9168) grad_norm 1.5768 (1.6803) [2021-04-14 18:20:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][870/1251] eta 0:01:46 lr 0.000635 time 0.2807 (0.2796) loss 4.5622 (4.9206) grad_norm 1.6355 (1.6806) [2021-04-14 18:20:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][880/1251] eta 0:01:43 lr 0.000636 time 0.2724 (0.2795) loss 5.2108 (4.9225) grad_norm 1.3783 (1.6813) [2021-04-14 18:21:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][890/1251] eta 0:01:40 lr 0.000636 time 0.2869 (0.2794) loss 4.0341 (4.9201) grad_norm 1.5303 (1.6798) [2021-04-14 18:21:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][900/1251] eta 0:01:38 lr 0.000636 time 0.2969 (0.2794) loss 5.0614 (4.9209) grad_norm 1.5963 (1.6813) [2021-04-14 18:21:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][910/1251] eta 0:01:35 lr 0.000637 time 0.2728 (0.2793) loss 5.7729 (4.9183) grad_norm 1.4511 (1.6822) [2021-04-14 18:21:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][920/1251] eta 0:01:32 lr 0.000637 time 0.2661 (0.2792) loss 4.6083 (4.9123) grad_norm 1.4172 (1.6819) [2021-04-14 18:21:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][930/1251] eta 0:01:29 lr 0.000638 time 0.2705 (0.2791) loss 5.7030 (4.9131) grad_norm 1.3602 (1.6795) [2021-04-14 18:21:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][940/1251] eta 0:01:26 lr 0.000638 time 0.2793 (0.2791) loss 4.8222 (4.9110) grad_norm 1.9494 (1.6781) [2021-04-14 18:21:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][950/1251] eta 0:01:23 lr 0.000638 time 0.2582 (0.2790) loss 5.6073 (4.9105) grad_norm 1.7396 (1.6774) [2021-04-14 18:21:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][960/1251] eta 0:01:21 lr 0.000639 time 0.2838 (0.2790) loss 5.2622 (4.9102) grad_norm 1.5816 (1.6782) [2021-04-14 18:21:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][970/1251] eta 0:01:18 lr 0.000639 time 0.2789 (0.2789) loss 4.5686 (4.9094) grad_norm 1.8883 (1.6775) [2021-04-14 18:21:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][980/1251] eta 0:01:15 lr 0.000640 time 0.2989 (0.2789) loss 5.0749 (4.9072) grad_norm 1.9083 (1.6799) [2021-04-14 18:21:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][990/1251] eta 0:01:12 lr 0.000640 time 0.2801 (0.2789) loss 5.2717 (4.9093) grad_norm 1.2706 (1.6792) [2021-04-14 18:21:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][1000/1251] eta 0:01:10 lr 0.000640 time 0.2610 (0.2789) loss 4.8996 (4.9078) grad_norm 1.8280 (1.6792) [2021-04-14 18:21:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][1010/1251] eta 0:01:07 lr 0.000641 time 0.2449 (0.2788) loss 4.1109 (4.9083) grad_norm 1.5995 (1.6786) [2021-04-14 18:21:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][1020/1251] eta 0:01:04 lr 0.000641 time 0.2616 (0.2787) loss 5.3505 (4.9071) grad_norm 1.2211 (1.6791) [2021-04-14 18:21:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][1030/1251] eta 0:01:01 lr 0.000642 time 0.2447 (0.2786) loss 4.4574 (4.9068) grad_norm 1.6027 (1.6779) [2021-04-14 18:21:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][1040/1251] eta 0:00:58 lr 0.000642 time 0.2859 (0.2787) loss 3.8857 (4.9061) grad_norm 1.7837 (1.6765) [2021-04-14 18:21:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][1050/1251] eta 0:00:56 lr 0.000642 time 0.2676 (0.2787) loss 5.5513 (4.9064) grad_norm 1.7545 (1.6762) [2021-04-14 18:21:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][1060/1251] eta 0:00:53 lr 0.000643 time 0.2696 (0.2787) loss 5.2177 (4.9071) grad_norm 1.7601 (1.6747) [2021-04-14 18:21:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][1070/1251] eta 0:00:50 lr 0.000643 time 0.2591 (0.2786) loss 5.3918 (4.9084) grad_norm 1.7988 (1.6760) [2021-04-14 18:21:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][1080/1251] eta 0:00:47 lr 0.000644 time 0.2943 (0.2786) loss 4.4518 (4.9067) grad_norm 1.5400 (1.6752) [2021-04-14 18:21:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][1090/1251] eta 0:00:44 lr 0.000644 time 0.2808 (0.2786) loss 5.3477 (4.9049) grad_norm 1.3053 (1.6771) [2021-04-14 18:21:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][1100/1251] eta 0:00:42 lr 0.000644 time 0.3173 (0.2787) loss 4.5640 (4.9051) grad_norm 1.7599 (1.6765) [2021-04-14 18:22:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][1110/1251] eta 0:00:39 lr 0.000645 time 0.2803 (0.2786) loss 4.0008 (4.9049) grad_norm 1.6272 (1.6744) [2021-04-14 18:22:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][1120/1251] eta 0:00:36 lr 0.000645 time 0.2976 (0.2786) loss 4.6515 (4.9046) grad_norm 1.5580 (1.6720) [2021-04-14 18:22:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][1130/1251] eta 0:00:33 lr 0.000646 time 0.2809 (0.2785) loss 4.3950 (4.9070) grad_norm 1.8084 (1.6717) [2021-04-14 18:22:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][1140/1251] eta 0:00:30 lr 0.000646 time 0.2408 (0.2785) loss 4.4612 (4.9086) grad_norm 1.3175 (1.6714) [2021-04-14 18:22:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][1150/1251] eta 0:00:28 lr 0.000646 time 0.2942 (0.2785) loss 4.0744 (4.9088) grad_norm 1.8783 (1.6704) [2021-04-14 18:22:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][1160/1251] eta 0:00:25 lr 0.000647 time 0.2936 (0.2785) loss 4.7718 (4.9084) grad_norm 1.4742 (1.6701) [2021-04-14 18:22:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][1170/1251] eta 0:00:22 lr 0.000647 time 0.2863 (0.2786) loss 4.7584 (4.9051) grad_norm 1.6101 (1.6699) [2021-04-14 18:22:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][1180/1251] eta 0:00:19 lr 0.000648 time 0.2834 (0.2787) loss 5.2755 (4.9045) grad_norm 1.5936 (1.6696) [2021-04-14 18:22:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][1190/1251] eta 0:00:16 lr 0.000648 time 0.2693 (0.2786) loss 5.8146 (4.9072) grad_norm 2.1033 (1.6690) [2021-04-14 18:22:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][1200/1251] eta 0:00:14 lr 0.000648 time 0.2810 (0.2785) loss 4.2171 (4.9082) grad_norm 1.7393 (1.6700) [2021-04-14 18:22:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][1210/1251] eta 0:00:11 lr 0.000649 time 0.2599 (0.2785) loss 5.1234 (4.9070) grad_norm 1.6549 (1.6709) [2021-04-14 18:22:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][1220/1251] eta 0:00:08 lr 0.000649 time 0.2747 (0.2785) loss 5.3261 (4.9083) grad_norm 1.4879 (1.6700) [2021-04-14 18:22:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][1230/1251] eta 0:00:05 lr 0.000650 time 0.2573 (0.2785) loss 4.3707 (4.9048) grad_norm 1.4697 (1.6700) [2021-04-14 18:22:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][1240/1251] eta 0:00:03 lr 0.000650 time 0.2474 (0.2783) loss 4.5671 (4.9055) grad_norm 1.6985 (1.6693) [2021-04-14 18:22:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [12/300][1250/1251] eta 0:00:00 lr 0.000650 time 0.2481 (0.2781) loss 5.5056 (4.9066) grad_norm 1.4517 (1.6677) [2021-04-14 18:22:42 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 12 training takes 0:05:49 [2021-04-14 18:22:42 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_12.pth saving...... [2021-04-14 18:22:55 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_12.pth saved !!! [2021-04-14 18:22:57 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.362 (1.362) Loss 2.1977 (2.1977) Acc@1 50.781 (50.781) Acc@5 77.832 (77.832) [2021-04-14 18:22:58 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.128 (0.270) Loss 2.2136 (2.2413) Acc@1 51.270 (51.278) Acc@5 76.660 (76.234) [2021-04-14 18:23:00 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.460 (0.247) Loss 2.1759 (2.2321) Acc@1 51.660 (51.260) Acc@5 76.660 (76.535) [2021-04-14 18:23:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.100 (0.229) Loss 2.2218 (2.2330) Acc@1 52.148 (51.386) Acc@5 77.051 (76.345) [2021-04-14 18:23:04 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.221) Loss 2.2645 (2.2365) Acc@1 49.609 (51.405) Acc@5 75.293 (76.250) [2021-04-14 18:23:06 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 51.234 Acc@5 76.194 [2021-04-14 18:23:06 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 51.2% [2021-04-14 18:23:06 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 51.23% [2021-04-14 18:23:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][0/1251] eta 1:20:40 lr 0.000650 time 3.8696 (3.8696) loss 3.6938 (3.6938) grad_norm 1.3404 (1.3404) [2021-04-14 18:23:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][10/1251] eta 0:12:27 lr 0.000651 time 0.2730 (0.6020) loss 3.4454 (4.6031) grad_norm 1.1754 (1.5192) [2021-04-14 18:23:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][20/1251] eta 0:09:08 lr 0.000651 time 0.2782 (0.4453) loss 5.4241 (4.7600) grad_norm 1.5119 (1.6138) [2021-04-14 18:23:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][30/1251] eta 0:07:56 lr 0.000652 time 0.2584 (0.3900) loss 4.6329 (4.7480) grad_norm 1.6371 (1.6013) [2021-04-14 18:23:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][40/1251] eta 0:07:17 lr 0.000652 time 0.2656 (0.3614) loss 5.4085 (4.7438) grad_norm 1.7335 (1.6101) [2021-04-14 18:23:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][50/1251] eta 0:06:53 lr 0.000652 time 0.2767 (0.3442) loss 5.2288 (4.7566) grad_norm 1.5761 (1.6257) [2021-04-14 18:23:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][60/1251] eta 0:06:35 lr 0.000653 time 0.2723 (0.3321) loss 4.8337 (4.8097) grad_norm 1.4263 (1.6336) [2021-04-14 18:23:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][70/1251] eta 0:06:22 lr 0.000653 time 0.2685 (0.3238) loss 4.3880 (4.7928) grad_norm 1.7058 (1.6191) [2021-04-14 18:23:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][80/1251] eta 0:06:10 lr 0.000654 time 0.2458 (0.3166) loss 3.9926 (4.8024) grad_norm 1.3738 (1.6156) [2021-04-14 18:23:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][90/1251] eta 0:06:01 lr 0.000654 time 0.2596 (0.3117) loss 3.5351 (4.7924) grad_norm 1.5667 (1.6136) [2021-04-14 18:23:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][100/1251] eta 0:05:54 lr 0.000654 time 0.2765 (0.3079) loss 3.8854 (4.7943) grad_norm 1.5740 (1.6129) [2021-04-14 18:23:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][110/1251] eta 0:05:47 lr 0.000655 time 0.2569 (0.3044) loss 5.2956 (4.7932) grad_norm 1.8506 (1.6352) [2021-04-14 18:23:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][120/1251] eta 0:05:41 lr 0.000655 time 0.2725 (0.3018) loss 4.6297 (4.7846) grad_norm 1.4753 (1.6338) [2021-04-14 18:23:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][130/1251] eta 0:05:35 lr 0.000656 time 0.2584 (0.2994) loss 5.3918 (4.7996) grad_norm 2.0029 (1.6494) [2021-04-14 18:23:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][140/1251] eta 0:05:30 lr 0.000656 time 0.2561 (0.2978) loss 5.2129 (4.7917) grad_norm 1.7182 (1.6583) [2021-04-14 18:23:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][150/1251] eta 0:05:25 lr 0.000656 time 0.2759 (0.2960) loss 5.0668 (4.7971) grad_norm 1.5509 (1.6559) [2021-04-14 18:23:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][160/1251] eta 0:05:21 lr 0.000657 time 0.2848 (0.2949) loss 4.6547 (4.7835) grad_norm 1.4434 (1.6455) [2021-04-14 18:23:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][170/1251] eta 0:05:17 lr 0.000657 time 0.2846 (0.2934) loss 4.0672 (4.7811) grad_norm 1.1712 (1.6442) [2021-04-14 18:23:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][180/1251] eta 0:05:13 lr 0.000658 time 0.2747 (0.2923) loss 4.8159 (4.8123) grad_norm 1.4803 (1.6368) [2021-04-14 18:24:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][190/1251] eta 0:05:08 lr 0.000658 time 0.2863 (0.2912) loss 4.1814 (4.8016) grad_norm 1.7191 (1.6414) [2021-04-14 18:24:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][200/1251] eta 0:05:05 lr 0.000658 time 0.2627 (0.2904) loss 4.3355 (4.7934) grad_norm 1.4678 (1.6394) [2021-04-14 18:24:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][210/1251] eta 0:05:01 lr 0.000659 time 0.2633 (0.2897) loss 5.1649 (4.8024) grad_norm 1.5635 (1.6402) [2021-04-14 18:24:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][220/1251] eta 0:04:57 lr 0.000659 time 0.2736 (0.2889) loss 5.0818 (4.7999) grad_norm 1.3745 (1.6367) [2021-04-14 18:24:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][230/1251] eta 0:04:54 lr 0.000660 time 0.2813 (0.2881) loss 5.7169 (4.7997) grad_norm 1.5249 (1.6353) [2021-04-14 18:24:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][240/1251] eta 0:04:50 lr 0.000660 time 0.2845 (0.2874) loss 4.6228 (4.8090) grad_norm 1.5446 (1.6307) [2021-04-14 18:24:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][250/1251] eta 0:04:46 lr 0.000660 time 0.2610 (0.2867) loss 4.7723 (4.8151) grad_norm 2.3879 (1.6391) [2021-04-14 18:24:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][260/1251] eta 0:04:43 lr 0.000661 time 0.2570 (0.2860) loss 5.3438 (4.8139) grad_norm 1.4656 (1.6349) [2021-04-14 18:24:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][270/1251] eta 0:04:40 lr 0.000661 time 0.2718 (0.2855) loss 4.1240 (4.8048) grad_norm 2.3276 (1.6393) [2021-04-14 18:24:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][280/1251] eta 0:04:36 lr 0.000662 time 0.2513 (0.2850) loss 4.8918 (4.8015) grad_norm 2.1103 (1.6386) [2021-04-14 18:24:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][290/1251] eta 0:04:33 lr 0.000662 time 0.2704 (0.2847) loss 5.2641 (4.7936) grad_norm 2.6436 (1.6447) [2021-04-14 18:24:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][300/1251] eta 0:04:30 lr 0.000662 time 0.2822 (0.2842) loss 5.0258 (4.7864) grad_norm 1.5283 (1.6400) [2021-04-14 18:24:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][310/1251] eta 0:04:26 lr 0.000663 time 0.2787 (0.2837) loss 4.7515 (4.7870) grad_norm 1.4756 (1.6416) [2021-04-14 18:24:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][320/1251] eta 0:04:23 lr 0.000663 time 0.3270 (0.2835) loss 5.7471 (4.7943) grad_norm 1.5070 (1.6382) [2021-04-14 18:24:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][330/1251] eta 0:04:20 lr 0.000664 time 0.2867 (0.2833) loss 5.6754 (4.8046) grad_norm 1.5731 (1.6356) [2021-04-14 18:24:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][340/1251] eta 0:04:17 lr 0.000664 time 0.2536 (0.2831) loss 4.7643 (4.8007) grad_norm 1.5980 (1.6322) [2021-04-14 18:24:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][350/1251] eta 0:04:14 lr 0.000664 time 0.2877 (0.2830) loss 4.3234 (4.7957) grad_norm 2.1330 (1.6354) [2021-04-14 18:24:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][360/1251] eta 0:04:11 lr 0.000665 time 0.2732 (0.2826) loss 5.0012 (4.7938) grad_norm 1.5933 (1.6356) [2021-04-14 18:24:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][370/1251] eta 0:04:09 lr 0.000665 time 0.2445 (0.2828) loss 4.5218 (4.8012) grad_norm 1.5844 (1.6336) [2021-04-14 18:24:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][380/1251] eta 0:04:06 lr 0.000666 time 0.2526 (0.2825) loss 5.1532 (4.8011) grad_norm 1.2425 (1.6288) [2021-04-14 18:24:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][390/1251] eta 0:04:02 lr 0.000666 time 0.2897 (0.2822) loss 5.2540 (4.7980) grad_norm 1.6307 (1.6292) [2021-04-14 18:24:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][400/1251] eta 0:03:59 lr 0.000666 time 0.2527 (0.2819) loss 5.3425 (4.8003) grad_norm 1.4294 (1.6269) [2021-04-14 18:25:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][410/1251] eta 0:03:56 lr 0.000667 time 0.2648 (0.2817) loss 4.0519 (4.7959) grad_norm 1.6828 (1.6238) [2021-04-14 18:25:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][420/1251] eta 0:03:53 lr 0.000667 time 0.2828 (0.2814) loss 4.2631 (4.7960) grad_norm 1.6857 (1.6195) [2021-04-14 18:25:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][430/1251] eta 0:03:50 lr 0.000668 time 0.2585 (0.2811) loss 4.7422 (4.7987) grad_norm 1.3203 (1.6174) [2021-04-14 18:25:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][440/1251] eta 0:03:47 lr 0.000668 time 0.2877 (0.2809) loss 4.7416 (4.8058) grad_norm 1.7763 (1.6173) [2021-04-14 18:25:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][450/1251] eta 0:03:44 lr 0.000668 time 0.2702 (0.2807) loss 5.2524 (4.8106) grad_norm 1.4280 (1.6148) [2021-04-14 18:25:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][460/1251] eta 0:03:41 lr 0.000669 time 0.2666 (0.2805) loss 4.9857 (4.8056) grad_norm 2.1298 (1.6160) [2021-04-14 18:25:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][470/1251] eta 0:03:38 lr 0.000669 time 0.2883 (0.2803) loss 5.2243 (4.8051) grad_norm 1.8144 (1.6183) [2021-04-14 18:25:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][480/1251] eta 0:03:36 lr 0.000670 time 0.2480 (0.2803) loss 5.4438 (4.8077) grad_norm 1.5018 (1.6190) [2021-04-14 18:25:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][490/1251] eta 0:03:33 lr 0.000670 time 0.2665 (0.2802) loss 5.7500 (4.8136) grad_norm 1.8508 (1.6187) [2021-04-14 18:25:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][500/1251] eta 0:03:30 lr 0.000670 time 0.2719 (0.2801) loss 5.4935 (4.8190) grad_norm 1.5265 (1.6171) [2021-04-14 18:25:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][510/1251] eta 0:03:27 lr 0.000671 time 0.2663 (0.2799) loss 4.5605 (4.8236) grad_norm 2.4334 (1.6199) [2021-04-14 18:25:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][520/1251] eta 0:03:24 lr 0.000671 time 0.2725 (0.2797) loss 5.0203 (4.8240) grad_norm 1.3588 (1.6190) [2021-04-14 18:25:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][530/1251] eta 0:03:21 lr 0.000672 time 0.2635 (0.2796) loss 5.0685 (4.8179) grad_norm 1.4003 (1.6195) [2021-04-14 18:25:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][540/1251] eta 0:03:18 lr 0.000672 time 0.2960 (0.2794) loss 4.8554 (4.8177) grad_norm 1.7739 (1.6216) [2021-04-14 18:25:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][550/1251] eta 0:03:15 lr 0.000672 time 0.2733 (0.2793) loss 4.9764 (4.8144) grad_norm 1.6383 (1.6225) [2021-04-14 18:25:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][560/1251] eta 0:03:12 lr 0.000673 time 0.2740 (0.2791) loss 4.9289 (4.8127) grad_norm 1.2952 (1.6246) [2021-04-14 18:25:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][570/1251] eta 0:03:10 lr 0.000673 time 0.2541 (0.2792) loss 4.2631 (4.8146) grad_norm 1.4945 (1.6257) [2021-04-14 18:25:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][580/1251] eta 0:03:07 lr 0.000674 time 0.2741 (0.2791) loss 4.9789 (4.8195) grad_norm 1.4129 (1.6242) [2021-04-14 18:25:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][590/1251] eta 0:03:04 lr 0.000674 time 0.2821 (0.2789) loss 5.1395 (4.8231) grad_norm 1.5180 (1.6229) [2021-04-14 18:25:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][600/1251] eta 0:03:01 lr 0.000674 time 0.2870 (0.2788) loss 5.4150 (4.8215) grad_norm 1.4599 (1.6224) [2021-04-14 18:25:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][610/1251] eta 0:02:58 lr 0.000675 time 0.2670 (0.2786) loss 5.2867 (4.8224) grad_norm 1.3203 (1.6229) [2021-04-14 18:25:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][620/1251] eta 0:02:55 lr 0.000675 time 0.2664 (0.2786) loss 5.7037 (4.8254) grad_norm 1.5336 (1.6228) [2021-04-14 18:26:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][630/1251] eta 0:02:52 lr 0.000676 time 0.2860 (0.2785) loss 4.4452 (4.8250) grad_norm 1.3707 (1.6218) [2021-04-14 18:26:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][640/1251] eta 0:02:50 lr 0.000676 time 0.2565 (0.2783) loss 5.6137 (4.8228) grad_norm 1.4352 (1.6197) [2021-04-14 18:26:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][650/1251] eta 0:02:47 lr 0.000676 time 0.2879 (0.2783) loss 4.9272 (4.8191) grad_norm 1.6933 (1.6187) [2021-04-14 18:26:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][660/1251] eta 0:02:44 lr 0.000677 time 0.2808 (0.2782) loss 4.6792 (4.8210) grad_norm 1.5090 (1.6193) [2021-04-14 18:26:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][670/1251] eta 0:02:41 lr 0.000677 time 0.2442 (0.2781) loss 3.7609 (4.8179) grad_norm 1.6736 (1.6183) [2021-04-14 18:26:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][680/1251] eta 0:02:38 lr 0.000678 time 0.2543 (0.2780) loss 5.1095 (4.8174) grad_norm 1.5369 (1.6192) [2021-04-14 18:26:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][690/1251] eta 0:02:35 lr 0.000678 time 0.2644 (0.2780) loss 3.8560 (4.8201) grad_norm 1.5262 (1.6179) [2021-04-14 18:26:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][700/1251] eta 0:02:33 lr 0.000678 time 0.2743 (0.2779) loss 5.6348 (4.8198) grad_norm 1.4055 (1.6175) [2021-04-14 18:26:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][710/1251] eta 0:02:30 lr 0.000679 time 0.2929 (0.2778) loss 5.2782 (4.8172) grad_norm 1.3834 (1.6167) [2021-04-14 18:26:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][720/1251] eta 0:02:27 lr 0.000679 time 0.2638 (0.2777) loss 4.4359 (4.8173) grad_norm 1.4989 (1.6162) [2021-04-14 18:26:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][730/1251] eta 0:02:24 lr 0.000679 time 0.2539 (0.2776) loss 4.7110 (4.8166) grad_norm 1.5680 (1.6170) [2021-04-14 18:26:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][740/1251] eta 0:02:21 lr 0.000680 time 0.2704 (0.2775) loss 4.6480 (4.8186) grad_norm 1.4275 (1.6187) [2021-04-14 18:26:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][750/1251] eta 0:02:19 lr 0.000680 time 0.2642 (0.2775) loss 4.7546 (4.8180) grad_norm 2.3917 (1.6187) [2021-04-14 18:26:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][760/1251] eta 0:02:16 lr 0.000681 time 0.3002 (0.2775) loss 5.2849 (4.8151) grad_norm 1.2621 (1.6196) [2021-04-14 18:26:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][770/1251] eta 0:02:13 lr 0.000681 time 0.2596 (0.2773) loss 4.0014 (4.8124) grad_norm 1.4049 (1.6184) [2021-04-14 18:26:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][780/1251] eta 0:02:10 lr 0.000681 time 0.2650 (0.2773) loss 5.1153 (4.8119) grad_norm 1.8198 (1.6174) [2021-04-14 18:26:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][790/1251] eta 0:02:07 lr 0.000682 time 0.2681 (0.2772) loss 3.8820 (4.8089) grad_norm 1.4924 (1.6154) [2021-04-14 18:26:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][800/1251] eta 0:02:04 lr 0.000682 time 0.2656 (0.2771) loss 4.5778 (4.8106) grad_norm 1.2657 (1.6146) [2021-04-14 18:26:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][810/1251] eta 0:02:02 lr 0.000683 time 0.2692 (0.2770) loss 4.7139 (4.8089) grad_norm 2.5666 (1.6167) [2021-04-14 18:26:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][820/1251] eta 0:01:59 lr 0.000683 time 0.2901 (0.2770) loss 4.8271 (4.8102) grad_norm 1.9849 (1.6180) [2021-04-14 18:26:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][830/1251] eta 0:01:56 lr 0.000683 time 0.2789 (0.2770) loss 4.6657 (4.8096) grad_norm 1.6857 (1.6175) [2021-04-14 18:26:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][840/1251] eta 0:01:53 lr 0.000684 time 0.2716 (0.2769) loss 5.2100 (4.8125) grad_norm 1.7001 (1.6201) [2021-04-14 18:27:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][850/1251] eta 0:01:51 lr 0.000684 time 0.3211 (0.2768) loss 4.8297 (4.8105) grad_norm 2.0045 (1.6211) [2021-04-14 18:27:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][860/1251] eta 0:01:48 lr 0.000685 time 0.2566 (0.2768) loss 4.8724 (4.8130) grad_norm 1.7843 (1.6197) [2021-04-14 18:27:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][870/1251] eta 0:01:45 lr 0.000685 time 0.2491 (0.2767) loss 5.3599 (4.8124) grad_norm 1.3598 (1.6181) [2021-04-14 18:27:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][880/1251] eta 0:01:42 lr 0.000685 time 0.2411 (0.2766) loss 4.1052 (4.8112) grad_norm 1.6612 (1.6193) [2021-04-14 18:27:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][890/1251] eta 0:01:39 lr 0.000686 time 0.3019 (0.2766) loss 4.2721 (4.8107) grad_norm 1.4749 (1.6179) [2021-04-14 18:27:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][900/1251] eta 0:01:37 lr 0.000686 time 0.2825 (0.2765) loss 4.2467 (4.8126) grad_norm 1.8959 (1.6171) [2021-04-14 18:27:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][910/1251] eta 0:01:34 lr 0.000687 time 0.2837 (0.2765) loss 5.3237 (4.8137) grad_norm 1.6024 (1.6163) [2021-04-14 18:27:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][920/1251] eta 0:01:31 lr 0.000687 time 0.2771 (0.2764) loss 5.1702 (4.8098) grad_norm 1.7585 (1.6170) [2021-04-14 18:27:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][930/1251] eta 0:01:28 lr 0.000687 time 0.2779 (0.2764) loss 4.7461 (4.8056) grad_norm 1.6636 (1.6168) [2021-04-14 18:27:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][940/1251] eta 0:01:25 lr 0.000688 time 0.2831 (0.2764) loss 4.8513 (4.8076) grad_norm 1.4408 (1.6161) [2021-04-14 18:27:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][950/1251] eta 0:01:23 lr 0.000688 time 0.2787 (0.2765) loss 4.7717 (4.8085) grad_norm 1.5499 (1.6155) [2021-04-14 18:27:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][960/1251] eta 0:01:20 lr 0.000689 time 0.2559 (0.2764) loss 5.1428 (4.8070) grad_norm 1.9968 (1.6159) [2021-04-14 18:27:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][970/1251] eta 0:01:17 lr 0.000689 time 0.2665 (0.2764) loss 3.3827 (4.8054) grad_norm 1.6684 (1.6173) [2021-04-14 18:27:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][980/1251] eta 0:01:14 lr 0.000689 time 0.2608 (0.2763) loss 5.2868 (4.8073) grad_norm 1.8974 (1.6164) [2021-04-14 18:27:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][990/1251] eta 0:01:12 lr 0.000690 time 0.2773 (0.2763) loss 4.6632 (4.8077) grad_norm 1.5944 (1.6184) [2021-04-14 18:27:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][1000/1251] eta 0:01:09 lr 0.000690 time 0.2986 (0.2763) loss 4.8262 (4.8092) grad_norm 1.3202 (1.6169) [2021-04-14 18:27:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][1010/1251] eta 0:01:06 lr 0.000691 time 0.2753 (0.2763) loss 5.5744 (4.8082) grad_norm 1.4432 (1.6156) [2021-04-14 18:27:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][1020/1251] eta 0:01:03 lr 0.000691 time 0.2874 (0.2763) loss 5.2416 (4.8074) grad_norm 1.7024 (1.6152) [2021-04-14 18:27:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][1030/1251] eta 0:01:01 lr 0.000691 time 0.2711 (0.2762) loss 4.7300 (4.8086) grad_norm 1.5836 (1.6150) [2021-04-14 18:27:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][1040/1251] eta 0:00:58 lr 0.000692 time 0.2654 (0.2762) loss 5.2712 (4.8098) grad_norm 1.8896 (1.6158) [2021-04-14 18:27:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][1050/1251] eta 0:00:55 lr 0.000692 time 0.4135 (0.2763) loss 4.6474 (4.8112) grad_norm 1.6639 (1.6167) [2021-04-14 18:27:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][1060/1251] eta 0:00:52 lr 0.000693 time 0.2937 (0.2762) loss 5.1018 (4.8107) grad_norm 1.4134 (1.6173) [2021-04-14 18:28:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][1070/1251] eta 0:00:49 lr 0.000693 time 0.2861 (0.2761) loss 5.1795 (4.8121) grad_norm 1.4258 (1.6159) [2021-04-14 18:28:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][1080/1251] eta 0:00:47 lr 0.000693 time 0.2858 (0.2761) loss 5.0094 (4.8087) grad_norm 1.2408 (1.6151) [2021-04-14 18:28:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][1090/1251] eta 0:00:44 lr 0.000694 time 0.2689 (0.2761) loss 4.8039 (4.8096) grad_norm 1.8152 (1.6140) [2021-04-14 18:28:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][1100/1251] eta 0:00:41 lr 0.000694 time 0.2912 (0.2761) loss 5.3951 (4.8081) grad_norm 1.3645 (1.6130) [2021-04-14 18:28:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][1110/1251] eta 0:00:38 lr 0.000695 time 0.2628 (0.2760) loss 5.2248 (4.8068) grad_norm 1.4396 (inf) [2021-04-14 18:28:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][1120/1251] eta 0:00:36 lr 0.000695 time 0.2788 (0.2760) loss 5.0075 (4.8032) grad_norm 2.2258 (inf) [2021-04-14 18:28:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][1130/1251] eta 0:00:33 lr 0.000695 time 0.2604 (0.2759) loss 5.0639 (4.8050) grad_norm 1.4959 (inf) [2021-04-14 18:28:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][1140/1251] eta 0:00:30 lr 0.000696 time 0.2598 (0.2759) loss 4.2701 (4.8058) grad_norm 1.9138 (inf) [2021-04-14 18:28:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][1150/1251] eta 0:00:27 lr 0.000696 time 0.2553 (0.2760) loss 5.0636 (4.8067) grad_norm 1.8807 (inf) [2021-04-14 18:28:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][1160/1251] eta 0:00:25 lr 0.000697 time 0.2684 (0.2759) loss 4.8443 (4.8095) grad_norm 1.3339 (inf) [2021-04-14 18:28:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][1170/1251] eta 0:00:22 lr 0.000697 time 0.2618 (0.2759) loss 4.9436 (4.8086) grad_norm 1.5380 (inf) [2021-04-14 18:28:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][1180/1251] eta 0:00:19 lr 0.000697 time 0.2712 (0.2758) loss 5.0524 (4.8084) grad_norm 1.7912 (inf) [2021-04-14 18:28:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][1190/1251] eta 0:00:16 lr 0.000698 time 0.2537 (0.2758) loss 5.1402 (4.8062) grad_norm 1.4619 (inf) [2021-04-14 18:28:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][1200/1251] eta 0:00:14 lr 0.000698 time 0.2596 (0.2758) loss 4.8170 (4.8049) grad_norm 1.5339 (inf) [2021-04-14 18:28:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][1210/1251] eta 0:00:11 lr 0.000699 time 0.2644 (0.2758) loss 4.8797 (4.8057) grad_norm 1.7076 (inf) [2021-04-14 18:28:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][1220/1251] eta 0:00:08 lr 0.000699 time 0.2752 (0.2757) loss 5.4383 (4.8052) grad_norm 1.4094 (inf) [2021-04-14 18:28:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][1230/1251] eta 0:00:05 lr 0.000699 time 0.2665 (0.2757) loss 4.5262 (4.8052) grad_norm 1.5228 (inf) [2021-04-14 18:28:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][1240/1251] eta 0:00:03 lr 0.000700 time 0.2480 (0.2756) loss 4.3612 (4.8025) grad_norm 1.1365 (inf) [2021-04-14 18:28:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [13/300][1250/1251] eta 0:00:00 lr 0.000700 time 0.2476 (0.2754) loss 4.6470 (4.8035) grad_norm 1.7323 (inf) [2021-04-14 18:28:53 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 13 training takes 0:05:46 [2021-04-14 18:28:53 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_13.pth saving...... [2021-04-14 18:29:08 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_13.pth saved !!! [2021-04-14 18:29:09 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.292 (1.292) Loss 2.0890 (2.0890) Acc@1 50.684 (50.684) Acc@5 77.734 (77.734) [2021-04-14 18:29:10 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.134 (0.217) Loss 2.1021 (2.1316) Acc@1 54.102 (52.459) Acc@5 77.637 (77.770) [2021-04-14 18:29:12 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.126 (0.218) Loss 2.0370 (2.1063) Acc@1 55.078 (53.106) Acc@5 79.004 (78.060) [2021-04-14 18:29:15 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.133 (0.223) Loss 2.1750 (2.1177) Acc@1 51.367 (52.813) Acc@5 77.441 (77.927) [2021-04-14 18:29:17 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.221) Loss 2.1853 (2.1252) Acc@1 52.539 (52.622) Acc@5 76.367 (77.722) [2021-04-14 18:29:19 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 52.684 Acc@5 77.712 [2021-04-14 18:29:19 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 52.7% [2021-04-14 18:29:19 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 52.68% [2021-04-14 18:29:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][0/1251] eta 1:16:50 lr 0.000700 time 3.6856 (3.6856) loss 5.0960 (5.0960) grad_norm 1.9518 (1.9518) [2021-04-14 18:29:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][10/1251] eta 0:12:12 lr 0.000701 time 0.2683 (0.5901) loss 4.9679 (4.9484) grad_norm 1.3677 (1.6725) [2021-04-14 18:29:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][20/1251] eta 0:08:58 lr 0.000701 time 0.2484 (0.4372) loss 4.7381 (4.8286) grad_norm 1.2971 (1.6233) [2021-04-14 18:29:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][30/1251] eta 0:07:52 lr 0.000701 time 0.2777 (0.3870) loss 4.1838 (4.8439) grad_norm 1.9321 (1.6226) [2021-04-14 18:29:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][40/1251] eta 0:07:13 lr 0.000702 time 0.2757 (0.3583) loss 4.7622 (4.7796) grad_norm 1.1756 (1.6417) [2021-04-14 18:29:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][50/1251] eta 0:06:50 lr 0.000702 time 0.2760 (0.3418) loss 5.2698 (4.7721) grad_norm 1.3843 (1.6278) [2021-04-14 18:29:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][60/1251] eta 0:06:32 lr 0.000703 time 0.2733 (0.3299) loss 4.6921 (4.8142) grad_norm 1.7049 (1.6190) [2021-04-14 18:29:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][70/1251] eta 0:06:20 lr 0.000703 time 0.2772 (0.3219) loss 3.9344 (4.8061) grad_norm 1.1863 (1.6194) [2021-04-14 18:29:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][80/1251] eta 0:06:09 lr 0.000703 time 0.2837 (0.3159) loss 3.9133 (4.8122) grad_norm 1.6183 (1.6200) [2021-04-14 18:29:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][90/1251] eta 0:06:01 lr 0.000704 time 0.2607 (0.3112) loss 4.7055 (4.8107) grad_norm 1.9525 (1.6274) [2021-04-14 18:29:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][100/1251] eta 0:05:54 lr 0.000704 time 0.2984 (0.3079) loss 4.4040 (4.8107) grad_norm 1.6218 (1.6155) [2021-04-14 18:29:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][110/1251] eta 0:05:47 lr 0.000705 time 0.2624 (0.3046) loss 4.9205 (4.8048) grad_norm 1.1371 (1.6004) [2021-04-14 18:29:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][120/1251] eta 0:05:41 lr 0.000705 time 0.2520 (0.3019) loss 5.3930 (4.7956) grad_norm 1.7502 (1.5983) [2021-04-14 18:29:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][130/1251] eta 0:05:36 lr 0.000705 time 0.2601 (0.2998) loss 3.9592 (4.8044) grad_norm 2.0076 (1.6010) [2021-04-14 18:30:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][140/1251] eta 0:05:31 lr 0.000706 time 0.2950 (0.2983) loss 5.3080 (4.8011) grad_norm 1.9525 (1.6012) [2021-04-14 18:30:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][150/1251] eta 0:05:26 lr 0.000706 time 0.2917 (0.2965) loss 4.4851 (4.7891) grad_norm 1.7873 (1.5945) [2021-04-14 18:30:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][160/1251] eta 0:05:21 lr 0.000707 time 0.2696 (0.2948) loss 5.1330 (4.8062) grad_norm 1.2978 (1.6055) [2021-04-14 18:30:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][170/1251] eta 0:05:18 lr 0.000707 time 0.2741 (0.2942) loss 3.9568 (4.7822) grad_norm 1.3265 (1.5976) [2021-04-14 18:30:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][180/1251] eta 0:05:13 lr 0.000707 time 0.2789 (0.2931) loss 4.3980 (4.7498) grad_norm 1.5822 (1.5934) [2021-04-14 18:30:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][190/1251] eta 0:05:09 lr 0.000708 time 0.2838 (0.2920) loss 5.0090 (4.7596) grad_norm 2.0523 (1.5923) [2021-04-14 18:30:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][200/1251] eta 0:05:05 lr 0.000708 time 0.2779 (0.2908) loss 5.2655 (4.7594) grad_norm 1.5392 (1.5870) [2021-04-14 18:30:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][210/1251] eta 0:05:01 lr 0.000709 time 0.2781 (0.2899) loss 4.8116 (4.7562) grad_norm 1.5967 (1.5858) [2021-04-14 18:30:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][220/1251] eta 0:04:58 lr 0.000709 time 0.2612 (0.2892) loss 4.2284 (4.7518) grad_norm 1.5813 (1.5876) [2021-04-14 18:30:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][230/1251] eta 0:04:54 lr 0.000709 time 0.2564 (0.2884) loss 4.9399 (4.7612) grad_norm 1.6772 (1.5889) [2021-04-14 18:30:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][240/1251] eta 0:04:50 lr 0.000710 time 0.2517 (0.2877) loss 5.0166 (4.7707) grad_norm 1.4305 (1.5840) [2021-04-14 18:30:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][250/1251] eta 0:04:47 lr 0.000710 time 0.2674 (0.2871) loss 5.5712 (4.7702) grad_norm 1.7277 (1.5839) [2021-04-14 18:30:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][260/1251] eta 0:04:44 lr 0.000711 time 0.3519 (0.2867) loss 4.4458 (4.7719) grad_norm 1.4868 (1.5825) [2021-04-14 18:30:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][270/1251] eta 0:04:40 lr 0.000711 time 0.2569 (0.2861) loss 4.1207 (4.7689) grad_norm 1.9934 (1.5816) [2021-04-14 18:30:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][280/1251] eta 0:04:37 lr 0.000711 time 0.2409 (0.2856) loss 4.5850 (4.7737) grad_norm 1.3877 (1.5813) [2021-04-14 18:30:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][290/1251] eta 0:04:34 lr 0.000712 time 0.2682 (0.2854) loss 4.9867 (4.7703) grad_norm 2.0804 (1.5827) [2021-04-14 18:30:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][300/1251] eta 0:04:30 lr 0.000712 time 0.2829 (0.2850) loss 4.4022 (4.7729) grad_norm 1.3289 (1.5777) [2021-04-14 18:30:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][310/1251] eta 0:04:27 lr 0.000713 time 0.2741 (0.2845) loss 4.7385 (4.7681) grad_norm 1.8856 (1.5789) [2021-04-14 18:30:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][320/1251] eta 0:04:24 lr 0.000713 time 0.2606 (0.2840) loss 4.9431 (4.7681) grad_norm 1.7868 (1.5822) [2021-04-14 18:30:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][330/1251] eta 0:04:21 lr 0.000713 time 0.3084 (0.2837) loss 5.3531 (4.7739) grad_norm 1.4564 (1.5804) [2021-04-14 18:30:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][340/1251] eta 0:04:18 lr 0.000714 time 0.2556 (0.2833) loss 4.5585 (4.7653) grad_norm 1.4424 (1.5795) [2021-04-14 18:30:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][350/1251] eta 0:04:14 lr 0.000714 time 0.2718 (0.2829) loss 3.6540 (4.7616) grad_norm 1.3588 (1.5776) [2021-04-14 18:31:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][360/1251] eta 0:04:12 lr 0.000715 time 0.2747 (0.2832) loss 5.4502 (4.7656) grad_norm 1.3673 (1.5759) [2021-04-14 18:31:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][370/1251] eta 0:04:09 lr 0.000715 time 0.2792 (0.2830) loss 5.2838 (4.7590) grad_norm 2.1500 (1.5750) [2021-04-14 18:31:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][380/1251] eta 0:04:06 lr 0.000715 time 0.2584 (0.2828) loss 4.1014 (4.7610) grad_norm 1.7362 (1.5753) [2021-04-14 18:31:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][390/1251] eta 0:04:03 lr 0.000716 time 0.2838 (0.2825) loss 5.5817 (4.7627) grad_norm 1.5991 (1.5753) [2021-04-14 18:31:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][400/1251] eta 0:04:00 lr 0.000716 time 0.2671 (0.2823) loss 4.7284 (4.7654) grad_norm 1.4680 (1.5725) [2021-04-14 18:31:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][410/1251] eta 0:03:57 lr 0.000717 time 0.2776 (0.2820) loss 5.1914 (4.7679) grad_norm 1.4338 (1.5719) [2021-04-14 18:31:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][420/1251] eta 0:03:54 lr 0.000717 time 0.2815 (0.2817) loss 4.3457 (4.7617) grad_norm 1.4121 (1.5691) [2021-04-14 18:31:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][430/1251] eta 0:03:51 lr 0.000717 time 0.2444 (0.2815) loss 5.1054 (4.7646) grad_norm 1.6032 (1.5693) [2021-04-14 18:31:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][440/1251] eta 0:03:48 lr 0.000718 time 0.2930 (0.2813) loss 4.9443 (4.7589) grad_norm 1.5532 (1.5694) [2021-04-14 18:31:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][450/1251] eta 0:03:45 lr 0.000718 time 0.2609 (0.2811) loss 4.7682 (4.7571) grad_norm 1.4101 (1.5662) [2021-04-14 18:31:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][460/1251] eta 0:03:42 lr 0.000719 time 0.2805 (0.2809) loss 4.9723 (4.7629) grad_norm 1.4244 (1.5628) [2021-04-14 18:31:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][470/1251] eta 0:03:39 lr 0.000719 time 0.2744 (0.2809) loss 5.6455 (4.7569) grad_norm 1.6829 (1.5627) [2021-04-14 18:31:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][480/1251] eta 0:03:36 lr 0.000719 time 0.2573 (0.2807) loss 4.3595 (4.7562) grad_norm 1.8271 (1.5600) [2021-04-14 18:31:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][490/1251] eta 0:03:33 lr 0.000720 time 0.2830 (0.2805) loss 5.1558 (4.7596) grad_norm 1.4163 (1.5588) [2021-04-14 18:31:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][500/1251] eta 0:03:30 lr 0.000720 time 0.2585 (0.2803) loss 5.2234 (4.7600) grad_norm 1.2818 (1.5610) [2021-04-14 18:31:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][510/1251] eta 0:03:27 lr 0.000721 time 0.2930 (0.2802) loss 4.2073 (4.7617) grad_norm 1.8592 (1.5632) [2021-04-14 18:31:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][520/1251] eta 0:03:24 lr 0.000721 time 0.2648 (0.2800) loss 5.1739 (4.7597) grad_norm 1.9101 (1.5636) [2021-04-14 18:31:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][530/1251] eta 0:03:21 lr 0.000721 time 0.2604 (0.2798) loss 5.0902 (4.7575) grad_norm 1.3358 (1.5624) [2021-04-14 18:31:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][540/1251] eta 0:03:18 lr 0.000722 time 0.2557 (0.2796) loss 3.9739 (4.7515) grad_norm 1.7903 (1.5638) [2021-04-14 18:31:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][550/1251] eta 0:03:15 lr 0.000722 time 0.2813 (0.2795) loss 3.4639 (4.7513) grad_norm 1.4552 (1.5629) [2021-04-14 18:31:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][560/1251] eta 0:03:12 lr 0.000723 time 0.2687 (0.2793) loss 4.0791 (4.7560) grad_norm 1.6045 (1.5613) [2021-04-14 18:31:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][570/1251] eta 0:03:10 lr 0.000723 time 0.2717 (0.2791) loss 4.6883 (4.7567) grad_norm 1.3573 (1.5618) [2021-04-14 18:32:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][580/1251] eta 0:03:07 lr 0.000723 time 0.2566 (0.2789) loss 5.8368 (4.7621) grad_norm 2.0225 (1.5639) [2021-04-14 18:32:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][590/1251] eta 0:03:04 lr 0.000724 time 0.2443 (0.2788) loss 3.6415 (4.7670) grad_norm 1.5166 (1.5649) [2021-04-14 18:32:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][600/1251] eta 0:03:01 lr 0.000724 time 0.2678 (0.2787) loss 4.8463 (4.7672) grad_norm 1.2590 (1.5648) [2021-04-14 18:32:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][610/1251] eta 0:02:58 lr 0.000725 time 0.2661 (0.2785) loss 5.1521 (4.7655) grad_norm 1.5896 (1.5635) [2021-04-14 18:32:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][620/1251] eta 0:02:55 lr 0.000725 time 0.2673 (0.2784) loss 4.7960 (4.7633) grad_norm 1.3377 (1.5629) [2021-04-14 18:32:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][630/1251] eta 0:02:52 lr 0.000725 time 0.2755 (0.2784) loss 5.3932 (4.7668) grad_norm 1.3475 (1.5616) [2021-04-14 18:32:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][640/1251] eta 0:02:50 lr 0.000726 time 0.2629 (0.2784) loss 4.5601 (4.7661) grad_norm 1.4756 (1.5597) [2021-04-14 18:32:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][650/1251] eta 0:02:47 lr 0.000726 time 0.2554 (0.2783) loss 3.6013 (4.7655) grad_norm 1.3887 (1.5574) [2021-04-14 18:32:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][660/1251] eta 0:02:44 lr 0.000727 time 0.2834 (0.2783) loss 4.2318 (4.7692) grad_norm 1.8846 (1.5565) [2021-04-14 18:32:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][670/1251] eta 0:02:41 lr 0.000727 time 0.2725 (0.2783) loss 4.4668 (4.7686) grad_norm 1.4466 (1.5551) [2021-04-14 18:32:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][680/1251] eta 0:02:38 lr 0.000727 time 0.2806 (0.2781) loss 4.2695 (4.7688) grad_norm 1.3136 (1.5553) [2021-04-14 18:32:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][690/1251] eta 0:02:35 lr 0.000728 time 0.2715 (0.2780) loss 3.9140 (4.7686) grad_norm 1.8874 (1.5565) [2021-04-14 18:32:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][700/1251] eta 0:02:33 lr 0.000728 time 0.2846 (0.2780) loss 4.5831 (4.7699) grad_norm 1.7354 (1.5560) [2021-04-14 18:32:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][710/1251] eta 0:02:30 lr 0.000729 time 0.2695 (0.2779) loss 4.0200 (4.7710) grad_norm 1.9670 (1.5558) [2021-04-14 18:32:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][720/1251] eta 0:02:27 lr 0.000729 time 0.2554 (0.2778) loss 5.0442 (4.7744) grad_norm 1.2896 (1.5546) [2021-04-14 18:32:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][730/1251] eta 0:02:24 lr 0.000729 time 0.2796 (0.2778) loss 5.0949 (4.7777) grad_norm 1.5882 (1.5520) [2021-04-14 18:32:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][740/1251] eta 0:02:21 lr 0.000730 time 0.2863 (0.2778) loss 5.3164 (4.7810) grad_norm 1.4352 (1.5520) [2021-04-14 18:32:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][750/1251] eta 0:02:19 lr 0.000730 time 0.2735 (0.2779) loss 4.6914 (4.7837) grad_norm 1.4744 (1.5518) [2021-04-14 18:32:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][760/1251] eta 0:02:16 lr 0.000731 time 0.2416 (0.2778) loss 4.2608 (4.7861) grad_norm 1.6839 (1.5544) [2021-04-14 18:32:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][770/1251] eta 0:02:13 lr 0.000731 time 0.2474 (0.2777) loss 5.2601 (4.7874) grad_norm 1.2710 (1.5538) [2021-04-14 18:32:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][780/1251] eta 0:02:10 lr 0.000731 time 0.2594 (0.2776) loss 4.4436 (4.7856) grad_norm 1.5416 (1.5537) [2021-04-14 18:32:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][790/1251] eta 0:02:07 lr 0.000732 time 0.2826 (0.2776) loss 5.4907 (4.7883) grad_norm 1.4201 (1.5518) [2021-04-14 18:33:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][800/1251] eta 0:02:05 lr 0.000732 time 0.2837 (0.2775) loss 3.8508 (4.7866) grad_norm 1.5315 (1.5499) [2021-04-14 18:33:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][810/1251] eta 0:02:02 lr 0.000733 time 0.2837 (0.2774) loss 5.1377 (4.7889) grad_norm 1.6260 (1.5497) [2021-04-14 18:33:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][820/1251] eta 0:01:59 lr 0.000733 time 0.2816 (0.2773) loss 5.3449 (4.7878) grad_norm 1.2815 (1.5480) [2021-04-14 18:33:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][830/1251] eta 0:01:56 lr 0.000733 time 0.2850 (0.2773) loss 5.0824 (4.7864) grad_norm 1.2908 (1.5462) [2021-04-14 18:33:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][840/1251] eta 0:01:53 lr 0.000734 time 0.2744 (0.2773) loss 3.6399 (4.7824) grad_norm 1.4850 (1.5441) [2021-04-14 18:33:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][850/1251] eta 0:01:51 lr 0.000734 time 0.2673 (0.2772) loss 5.0772 (4.7816) grad_norm 1.7758 (1.5434) [2021-04-14 18:33:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][860/1251] eta 0:01:48 lr 0.000735 time 0.2519 (0.2771) loss 5.2764 (4.7800) grad_norm 1.1808 (1.5429) [2021-04-14 18:33:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][870/1251] eta 0:01:45 lr 0.000735 time 0.2705 (0.2770) loss 5.6958 (4.7822) grad_norm 1.2197 (1.5419) [2021-04-14 18:33:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][880/1251] eta 0:01:42 lr 0.000735 time 0.2846 (0.2770) loss 4.2565 (4.7813) grad_norm 1.3465 (1.5412) [2021-04-14 18:33:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][890/1251] eta 0:01:39 lr 0.000736 time 0.2776 (0.2769) loss 5.0373 (4.7821) grad_norm 1.7095 (1.5401) [2021-04-14 18:33:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][900/1251] eta 0:01:37 lr 0.000736 time 0.2833 (0.2769) loss 3.9646 (4.7825) grad_norm 1.4554 (1.5394) [2021-04-14 18:33:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][910/1251] eta 0:01:34 lr 0.000737 time 0.2693 (0.2768) loss 4.2564 (4.7833) grad_norm 1.4695 (1.5388) [2021-04-14 18:33:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][920/1251] eta 0:01:31 lr 0.000737 time 0.2732 (0.2767) loss 3.5361 (4.7816) grad_norm 1.2603 (1.5390) [2021-04-14 18:33:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][930/1251] eta 0:01:28 lr 0.000737 time 0.2704 (0.2767) loss 5.0555 (4.7784) grad_norm 2.4293 (1.5413) [2021-04-14 18:33:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][940/1251] eta 0:01:26 lr 0.000738 time 0.2674 (0.2769) loss 4.2205 (4.7757) grad_norm 1.5567 (1.5421) [2021-04-14 18:33:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][950/1251] eta 0:01:23 lr 0.000738 time 0.2507 (0.2768) loss 4.4517 (4.7736) grad_norm 1.9999 (1.5425) [2021-04-14 18:33:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][960/1251] eta 0:01:20 lr 0.000739 time 0.2665 (0.2768) loss 5.3639 (4.7761) grad_norm 1.8466 (1.5441) [2021-04-14 18:33:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][970/1251] eta 0:01:17 lr 0.000739 time 0.2774 (0.2767) loss 3.6434 (4.7767) grad_norm 1.5442 (1.5442) [2021-04-14 18:33:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][980/1251] eta 0:01:14 lr 0.000739 time 0.2537 (0.2767) loss 5.5667 (4.7778) grad_norm 1.3161 (1.5439) [2021-04-14 18:33:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][990/1251] eta 0:01:12 lr 0.000740 time 0.2724 (0.2766) loss 3.9190 (4.7788) grad_norm 1.5109 (1.5426) [2021-04-14 18:33:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][1000/1251] eta 0:01:09 lr 0.000740 time 0.2589 (0.2765) loss 4.7778 (4.7794) grad_norm 1.1870 (1.5421) [2021-04-14 18:33:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][1010/1251] eta 0:01:06 lr 0.000741 time 0.2608 (0.2765) loss 3.7342 (4.7795) grad_norm 1.7664 (1.5421) [2021-04-14 18:34:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][1020/1251] eta 0:01:03 lr 0.000741 time 0.2704 (0.2764) loss 3.4850 (4.7794) grad_norm 1.7658 (1.5414) [2021-04-14 18:34:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][1030/1251] eta 0:01:01 lr 0.000741 time 0.2730 (0.2765) loss 4.8276 (4.7742) grad_norm 1.4422 (1.5399) [2021-04-14 18:34:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][1040/1251] eta 0:00:58 lr 0.000742 time 0.2670 (0.2764) loss 5.3593 (4.7760) grad_norm 1.1918 (1.5384) [2021-04-14 18:34:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][1050/1251] eta 0:00:55 lr 0.000742 time 0.2751 (0.2764) loss 3.5037 (4.7736) grad_norm 1.8830 (1.5386) [2021-04-14 18:34:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][1060/1251] eta 0:00:52 lr 0.000743 time 0.2666 (0.2763) loss 3.6304 (4.7729) grad_norm 1.4681 (1.5380) [2021-04-14 18:34:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][1070/1251] eta 0:00:50 lr 0.000743 time 0.2633 (0.2763) loss 4.9106 (4.7719) grad_norm 1.4840 (1.5361) [2021-04-14 18:34:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][1080/1251] eta 0:00:47 lr 0.000743 time 0.2888 (0.2762) loss 3.7277 (4.7714) grad_norm 1.9776 (1.5360) [2021-04-14 18:34:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][1090/1251] eta 0:00:44 lr 0.000744 time 0.2730 (0.2762) loss 4.2945 (4.7709) grad_norm 1.3701 (1.5359) [2021-04-14 18:34:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][1100/1251] eta 0:00:41 lr 0.000744 time 0.2811 (0.2761) loss 4.9748 (4.7713) grad_norm 1.3093 (1.5389) [2021-04-14 18:34:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][1110/1251] eta 0:00:38 lr 0.000745 time 0.2746 (0.2762) loss 4.3513 (4.7687) grad_norm 1.4370 (1.5382) [2021-04-14 18:34:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][1120/1251] eta 0:00:36 lr 0.000745 time 0.2950 (0.2761) loss 4.8712 (4.7654) grad_norm 1.2722 (1.5374) [2021-04-14 18:34:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][1130/1251] eta 0:00:33 lr 0.000745 time 0.2544 (0.2761) loss 5.2927 (4.7664) grad_norm 1.8229 (1.5379) [2021-04-14 18:34:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][1140/1251] eta 0:00:30 lr 0.000746 time 0.2828 (0.2761) loss 4.7687 (4.7659) grad_norm 1.4410 (1.5371) [2021-04-14 18:34:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][1150/1251] eta 0:00:27 lr 0.000746 time 0.2851 (0.2761) loss 4.8230 (4.7648) grad_norm 1.6078 (1.5369) [2021-04-14 18:34:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][1160/1251] eta 0:00:25 lr 0.000747 time 0.2571 (0.2760) loss 3.9440 (4.7637) grad_norm 1.2606 (1.5367) [2021-04-14 18:34:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][1170/1251] eta 0:00:22 lr 0.000747 time 0.2549 (0.2760) loss 3.5620 (4.7640) grad_norm 1.3388 (1.5359) [2021-04-14 18:34:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][1180/1251] eta 0:00:19 lr 0.000747 time 0.2868 (0.2760) loss 5.3638 (4.7664) grad_norm 2.1151 (1.5356) [2021-04-14 18:34:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][1190/1251] eta 0:00:16 lr 0.000748 time 0.2793 (0.2760) loss 4.3555 (4.7657) grad_norm 1.8954 (1.5345) [2021-04-14 18:34:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][1200/1251] eta 0:00:14 lr 0.000748 time 0.2587 (0.2759) loss 4.7186 (4.7633) grad_norm 1.5922 (1.5341) [2021-04-14 18:34:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][1210/1251] eta 0:00:11 lr 0.000749 time 0.2623 (0.2759) loss 4.3605 (4.7616) grad_norm 1.1961 (1.5342) [2021-04-14 18:34:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][1220/1251] eta 0:00:08 lr 0.000749 time 0.2781 (0.2759) loss 5.2480 (4.7594) grad_norm 1.3106 (1.5336) [2021-04-14 18:34:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][1230/1251] eta 0:00:05 lr 0.000749 time 0.2768 (0.2758) loss 4.3718 (4.7589) grad_norm 1.2107 (1.5329) [2021-04-14 18:35:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][1240/1251] eta 0:00:03 lr 0.000750 time 0.2483 (0.2757) loss 4.7983 (4.7571) grad_norm 1.4428 (1.5326) [2021-04-14 18:35:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [14/300][1250/1251] eta 0:00:00 lr 0.000750 time 0.2480 (0.2755) loss 4.9332 (4.7564) grad_norm 1.3669 (1.5330) [2021-04-14 18:35:05 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 14 training takes 0:05:46 [2021-04-14 18:35:05 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_14.pth saving...... [2021-04-14 18:35:20 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_14.pth saved !!! [2021-04-14 18:35:21 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.282 (1.282) Loss 2.2364 (2.2364) Acc@1 50.781 (50.781) Acc@5 77.344 (77.344) [2021-04-14 18:35:22 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.112 (0.241) Loss 2.1709 (2.1334) Acc@1 53.027 (53.667) Acc@5 78.613 (78.338) [2021-04-14 18:35:25 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.214 (0.253) Loss 2.1037 (2.1412) Acc@1 53.516 (53.046) Acc@5 77.930 (78.283) [2021-04-14 18:35:27 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.163 (0.238) Loss 2.1300 (2.1394) Acc@1 52.832 (53.283) Acc@5 78.711 (78.320) [2021-04-14 18:35:29 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.110 (0.219) Loss 2.0678 (2.1327) Acc@1 53.320 (53.425) Acc@5 79.199 (78.401) [2021-04-14 18:35:31 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 53.472 Acc@5 78.472 [2021-04-14 18:35:31 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 53.5% [2021-04-14 18:35:31 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 53.47% [2021-04-15 15:02:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][0/1251] eta 1:42:33 lr 0.000750 time 4.9190 (4.9190) loss 5.5906 (5.5906) grad_norm 1.3799 (1.3799) [2021-04-15 15:02:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][10/1251] eta 0:16:05 lr 0.000751 time 0.2678 (0.7782) loss 4.2012 (5.1537) grad_norm 2.3689 (1.5480) [2021-04-15 15:02:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][20/1251] eta 0:11:06 lr 0.000751 time 0.2667 (0.5414) loss 4.7032 (5.0102) grad_norm 1.9725 (1.5593) [2021-04-15 15:02:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][30/1251] eta 0:09:15 lr 0.000751 time 0.2616 (0.4549) loss 3.7082 (5.0246) grad_norm 1.3818 (1.5663) [2021-04-15 15:03:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][40/1251] eta 0:08:18 lr 0.000752 time 0.2664 (0.4117) loss 4.8181 (4.9704) grad_norm 1.3052 (1.5406) [2021-04-15 15:03:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][50/1251] eta 0:07:44 lr 0.000752 time 0.2865 (0.3864) loss 5.1781 (4.9435) grad_norm 1.2127 (1.5073) [2021-04-15 15:03:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][60/1251] eta 0:07:20 lr 0.000753 time 0.3059 (0.3698) loss 4.6017 (4.9010) grad_norm 2.3248 (1.5487) [2021-04-15 15:03:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][70/1251] eta 0:07:01 lr 0.000753 time 0.2895 (0.3569) loss 4.2639 (4.8666) grad_norm 1.5999 (1.5601) [2021-04-15 15:03:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][80/1251] eta 0:06:46 lr 0.000753 time 0.2515 (0.3474) loss 4.3639 (4.8533) grad_norm 1.3109 (1.5509) [2021-04-15 15:03:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][90/1251] eta 0:06:34 lr 0.000754 time 0.2672 (0.3397) loss 5.8318 (4.8529) grad_norm 1.3593 (1.5346) [2021-04-15 15:03:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][100/1251] eta 0:06:23 lr 0.000754 time 0.2917 (0.3333) loss 5.0537 (4.8571) grad_norm 1.4453 (1.5241) [2021-04-15 15:03:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][110/1251] eta 0:06:14 lr 0.000755 time 0.2524 (0.3282) loss 4.2371 (4.8601) grad_norm 1.4602 (1.5280) [2021-04-15 15:03:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][120/1251] eta 0:06:06 lr 0.000755 time 0.2512 (0.3243) loss 3.9881 (4.8567) grad_norm 1.5960 (1.5152) [2021-04-15 15:03:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][130/1251] eta 0:05:59 lr 0.000755 time 0.2780 (0.3210) loss 5.1507 (4.8443) grad_norm 1.5932 (1.5185) [2021-04-15 15:03:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][140/1251] eta 0:05:53 lr 0.000756 time 0.2738 (0.3177) loss 5.0035 (4.8327) grad_norm 2.0007 (1.5153) [2021-04-15 15:03:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][150/1251] eta 0:05:48 lr 0.000756 time 0.2762 (0.3161) loss 4.1601 (4.8201) grad_norm 1.2670 (1.5116) [2021-04-15 15:03:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][160/1251] eta 0:05:43 lr 0.000757 time 0.2571 (0.3148) loss 4.9885 (4.8198) grad_norm 1.6413 (1.5068) [2021-04-15 15:03:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][170/1251] eta 0:05:37 lr 0.000757 time 0.2728 (0.3126) loss 4.5554 (4.8112) grad_norm 1.2790 (1.5042) [2021-04-15 15:03:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][180/1251] eta 0:05:32 lr 0.000757 time 0.2799 (0.3108) loss 4.8977 (4.7936) grad_norm 1.8403 (1.5174) [2021-04-15 15:03:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][190/1251] eta 0:05:27 lr 0.000758 time 0.2717 (0.3090) loss 4.1451 (4.7896) grad_norm 1.7353 (1.5255) [2021-04-15 15:03:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][200/1251] eta 0:05:23 lr 0.000758 time 0.3095 (0.3075) loss 4.6466 (4.7801) grad_norm 1.6089 (1.5291) [2021-04-15 15:03:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][210/1251] eta 0:05:18 lr 0.000759 time 0.2762 (0.3058) loss 4.7936 (4.7709) grad_norm 1.8168 (1.5309) [2021-04-15 15:03:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][220/1251] eta 0:05:13 lr 0.000759 time 0.2868 (0.3045) loss 4.9507 (4.7659) grad_norm 1.6763 (1.5255) [2021-04-15 15:03:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][230/1251] eta 0:05:09 lr 0.000759 time 0.2486 (0.3033) loss 3.7165 (4.7642) grad_norm 1.5391 (1.5212) [2021-04-15 15:03:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][240/1251] eta 0:05:05 lr 0.000760 time 0.2750 (0.3022) loss 4.8749 (4.7653) grad_norm 1.2569 (1.5226) [2021-04-15 15:03:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][250/1251] eta 0:05:01 lr 0.000760 time 0.2955 (0.3013) loss 4.8195 (4.7557) grad_norm 1.7248 (1.5279) [2021-04-15 15:04:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][260/1251] eta 0:04:57 lr 0.000761 time 0.2553 (0.3004) loss 4.4140 (4.7514) grad_norm 1.4505 (1.5286) [2021-04-15 15:04:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][270/1251] eta 0:04:53 lr 0.000761 time 0.2698 (0.2995) loss 5.1718 (4.7470) grad_norm 1.4250 (1.5244) [2021-04-15 15:04:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][280/1251] eta 0:04:50 lr 0.000761 time 0.2851 (0.2992) loss 4.6481 (4.7524) grad_norm 1.0138 (1.5194) [2021-04-15 15:04:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][290/1251] eta 0:04:46 lr 0.000762 time 0.2697 (0.2983) loss 3.7544 (4.7434) grad_norm 1.5309 (1.5155) [2021-04-15 15:04:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][300/1251] eta 0:04:42 lr 0.000762 time 0.2687 (0.2975) loss 4.7843 (4.7295) grad_norm 1.2444 (1.5178) [2021-04-15 15:04:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][310/1251] eta 0:04:39 lr 0.000763 time 0.2797 (0.2968) loss 5.1215 (4.7293) grad_norm 1.3828 (1.5177) [2021-04-15 15:04:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][320/1251] eta 0:04:35 lr 0.000763 time 0.2900 (0.2962) loss 5.4194 (4.7405) grad_norm 1.5413 (1.5171) [2021-04-15 15:04:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][330/1251] eta 0:04:32 lr 0.000763 time 0.2751 (0.2960) loss 3.6923 (4.7396) grad_norm 1.8700 (1.5166) [2021-04-15 15:04:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][340/1251] eta 0:04:29 lr 0.000764 time 0.2778 (0.2955) loss 5.3741 (4.7435) grad_norm 1.6158 (1.5222) [2021-04-15 15:04:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][350/1251] eta 0:04:25 lr 0.000764 time 0.2841 (0.2950) loss 5.2526 (4.7442) grad_norm 1.3843 (1.5231) [2021-04-15 15:04:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][360/1251] eta 0:04:22 lr 0.000765 time 0.2733 (0.2945) loss 4.7312 (4.7434) grad_norm 1.2804 (1.5197) [2021-04-15 15:04:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][370/1251] eta 0:04:19 lr 0.000765 time 0.2888 (0.2940) loss 3.7609 (4.7391) grad_norm 1.4732 (1.5160) [2021-04-15 15:04:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][380/1251] eta 0:04:15 lr 0.000765 time 0.3011 (0.2936) loss 3.7368 (4.7343) grad_norm 1.5294 (1.5111) [2021-04-15 15:04:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][390/1251] eta 0:04:12 lr 0.000766 time 0.2958 (0.2932) loss 5.5616 (4.7298) grad_norm 1.4148 (1.5079) [2021-04-15 15:04:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][400/1251] eta 0:04:09 lr 0.000766 time 0.2793 (0.2928) loss 5.1696 (4.7344) grad_norm 1.5707 (1.5059) [2021-04-15 15:04:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][410/1251] eta 0:04:05 lr 0.000767 time 0.2814 (0.2925) loss 4.8066 (4.7388) grad_norm 1.3220 (1.5046) [2021-04-15 15:04:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][420/1251] eta 0:04:02 lr 0.000767 time 0.2668 (0.2921) loss 5.2825 (4.7378) grad_norm 1.4374 (1.5056) [2021-04-15 15:04:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][430/1251] eta 0:03:59 lr 0.000767 time 0.2888 (0.2918) loss 5.2224 (4.7451) grad_norm 1.1780 (1.5020) [2021-04-15 15:04:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][440/1251] eta 0:03:56 lr 0.000768 time 0.2510 (0.2914) loss 5.2581 (4.7485) grad_norm 1.1910 (1.5012) [2021-04-15 15:04:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][450/1251] eta 0:03:53 lr 0.000768 time 0.2698 (0.2911) loss 4.3622 (4.7463) grad_norm 1.5414 (1.5014) [2021-04-15 15:04:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][460/1251] eta 0:03:49 lr 0.000769 time 0.2585 (0.2907) loss 4.5272 (4.7414) grad_norm 1.2653 (1.4985) [2021-04-15 15:05:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][470/1251] eta 0:03:46 lr 0.000769 time 0.2693 (0.2905) loss 4.4928 (4.7343) grad_norm 1.3256 (1.4993) [2021-04-15 15:05:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][480/1251] eta 0:03:43 lr 0.000769 time 0.2594 (0.2901) loss 4.9567 (4.7320) grad_norm 1.4605 (1.4964) [2021-04-15 15:05:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][490/1251] eta 0:03:40 lr 0.000770 time 0.2737 (0.2898) loss 4.7716 (4.7376) grad_norm 2.0440 (1.4966) [2021-04-15 15:05:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][500/1251] eta 0:03:37 lr 0.000770 time 0.2747 (0.2895) loss 5.2240 (4.7361) grad_norm 1.6601 (1.4961) [2021-04-15 15:05:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][510/1251] eta 0:03:34 lr 0.000771 time 0.2956 (0.2892) loss 5.5300 (4.7406) grad_norm 1.1673 (1.4957) [2021-04-15 15:05:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][520/1251] eta 0:03:31 lr 0.000771 time 0.3008 (0.2890) loss 4.3069 (4.7398) grad_norm 1.7538 (1.4971) [2021-04-15 15:05:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][530/1251] eta 0:03:28 lr 0.000771 time 0.2784 (0.2887) loss 4.9615 (4.7349) grad_norm 1.5878 (1.4993) [2021-04-15 15:05:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][540/1251] eta 0:03:25 lr 0.000772 time 0.2693 (0.2885) loss 4.9983 (4.7335) grad_norm 1.6540 (1.4998) [2021-04-15 15:05:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][550/1251] eta 0:03:22 lr 0.000772 time 0.2533 (0.2884) loss 5.4011 (4.7321) grad_norm 1.2633 (1.5000) [2021-04-15 15:05:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][560/1251] eta 0:03:19 lr 0.000773 time 0.2650 (0.2883) loss 4.9046 (4.7368) grad_norm 1.1902 (1.4981) [2021-04-15 15:05:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][570/1251] eta 0:03:16 lr 0.000773 time 0.2667 (0.2881) loss 5.0647 (4.7392) grad_norm 1.4757 (1.4963) [2021-04-15 15:05:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][580/1251] eta 0:03:13 lr 0.000773 time 0.3039 (0.2879) loss 5.4658 (4.7398) grad_norm 1.8526 (1.4959) [2021-04-15 15:05:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][590/1251] eta 0:03:10 lr 0.000774 time 0.2868 (0.2878) loss 5.0762 (4.7404) grad_norm 1.5331 (1.4970) [2021-04-15 15:05:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][600/1251] eta 0:03:07 lr 0.000774 time 0.3059 (0.2875) loss 3.3548 (4.7376) grad_norm 1.3441 (1.4967) [2021-04-15 15:05:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][610/1251] eta 0:03:04 lr 0.000775 time 0.2790 (0.2873) loss 4.6896 (4.7377) grad_norm 1.2805 (1.4961) [2021-04-15 15:05:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][620/1251] eta 0:03:01 lr 0.000775 time 0.2559 (0.2871) loss 5.4252 (4.7402) grad_norm 1.2130 (1.4947) [2021-04-15 15:05:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][630/1251] eta 0:02:58 lr 0.000775 time 0.2671 (0.2870) loss 5.3221 (4.7420) grad_norm 1.3128 (1.4944) [2021-04-15 15:05:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][640/1251] eta 0:02:55 lr 0.000776 time 0.2589 (0.2868) loss 3.6592 (4.7413) grad_norm 1.3971 (1.4954) [2021-04-15 15:05:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][650/1251] eta 0:02:52 lr 0.000776 time 0.3128 (0.2868) loss 5.5148 (4.7414) grad_norm 1.8362 (1.4992) [2021-04-15 15:05:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][660/1251] eta 0:02:49 lr 0.000777 time 0.2767 (0.2867) loss 3.6851 (4.7364) grad_norm 1.5416 (1.4987) [2021-04-15 15:05:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][670/1251] eta 0:02:46 lr 0.000777 time 0.2697 (0.2866) loss 5.6285 (4.7403) grad_norm 1.5768 (1.4998) [2021-04-15 15:05:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][680/1251] eta 0:02:43 lr 0.000777 time 0.2745 (0.2864) loss 5.3642 (4.7411) grad_norm 1.2622 (1.4989) [2021-04-15 15:06:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][690/1251] eta 0:02:40 lr 0.000778 time 0.2556 (0.2863) loss 3.9286 (4.7380) grad_norm 1.7709 (1.4999) [2021-04-15 15:06:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][700/1251] eta 0:02:37 lr 0.000778 time 0.2655 (0.2861) loss 5.1712 (4.7361) grad_norm 1.3070 (1.5022) [2021-04-15 15:06:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][710/1251] eta 0:02:34 lr 0.000779 time 0.2939 (0.2860) loss 3.8180 (4.7344) grad_norm 1.3066 (1.5002) [2021-04-15 15:06:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][720/1251] eta 0:02:31 lr 0.000779 time 0.4211 (0.2861) loss 5.3930 (4.7315) grad_norm 1.7208 (1.5001) [2021-04-15 15:06:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][730/1251] eta 0:02:28 lr 0.000779 time 0.2618 (0.2859) loss 4.7360 (4.7308) grad_norm 1.3690 (1.4982) [2021-04-15 15:06:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][740/1251] eta 0:02:26 lr 0.000780 time 0.2969 (0.2859) loss 5.0089 (4.7368) grad_norm 1.2917 (1.5009) [2021-04-15 15:06:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][750/1251] eta 0:02:23 lr 0.000780 time 0.2676 (0.2858) loss 5.0663 (4.7343) grad_norm 1.5063 (1.5021) [2021-04-15 15:06:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][760/1251] eta 0:02:20 lr 0.000781 time 0.2982 (0.2857) loss 5.3030 (4.7350) grad_norm 2.1361 (1.5026) [2021-04-15 15:06:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][770/1251] eta 0:02:17 lr 0.000781 time 0.2794 (0.2856) loss 5.6387 (4.7369) grad_norm 1.3397 (1.5029) [2021-04-15 15:06:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][780/1251] eta 0:02:14 lr 0.000781 time 0.2670 (0.2854) loss 5.3680 (4.7368) grad_norm 1.4388 (1.5028) [2021-04-15 15:06:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][790/1251] eta 0:02:11 lr 0.000782 time 0.2741 (0.2853) loss 3.9303 (4.7332) grad_norm 1.3680 (1.5035) [2021-04-15 15:06:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][800/1251] eta 0:02:08 lr 0.000782 time 0.2823 (0.2851) loss 5.4462 (4.7343) grad_norm 1.5558 (1.5025) [2021-04-15 15:06:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][810/1251] eta 0:02:05 lr 0.000783 time 0.2870 (0.2851) loss 3.4822 (4.7268) grad_norm 1.5403 (1.5029) [2021-04-15 15:06:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][820/1251] eta 0:02:02 lr 0.000783 time 0.2847 (0.2850) loss 4.8839 (4.7276) grad_norm 1.4160 (1.5037) [2021-04-15 15:06:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][830/1251] eta 0:01:59 lr 0.000783 time 0.2619 (0.2849) loss 3.9780 (4.7247) grad_norm 1.4290 (1.5043) [2021-04-15 15:06:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][840/1251] eta 0:01:57 lr 0.000784 time 0.2534 (0.2848) loss 5.1795 (4.7244) grad_norm 1.5809 (1.5031) [2021-04-15 15:06:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][850/1251] eta 0:01:54 lr 0.000784 time 0.2672 (0.2847) loss 5.4699 (4.7239) grad_norm 1.1150 (1.5010) [2021-04-15 15:06:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][860/1251] eta 0:01:51 lr 0.000785 time 0.2704 (0.2846) loss 5.0665 (4.7247) grad_norm 1.1169 (1.4991) [2021-04-15 15:06:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][870/1251] eta 0:01:48 lr 0.000785 time 0.2687 (0.2845) loss 5.1190 (4.7243) grad_norm 1.3870 (1.4981) [2021-04-15 15:06:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][880/1251] eta 0:01:45 lr 0.000785 time 0.2553 (0.2843) loss 5.0924 (4.7222) grad_norm 1.1875 (1.4971) [2021-04-15 15:06:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][890/1251] eta 0:01:42 lr 0.000786 time 0.2695 (0.2843) loss 4.0300 (4.7200) grad_norm 1.8299 (1.4983) [2021-04-15 15:07:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][900/1251] eta 0:01:39 lr 0.000786 time 0.2659 (0.2842) loss 5.2935 (4.7205) grad_norm 1.4546 (1.4964) [2021-04-15 15:07:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][910/1251] eta 0:01:36 lr 0.000787 time 0.2513 (0.2840) loss 4.0337 (4.7196) grad_norm 1.5245 (1.4958) [2021-04-15 15:07:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][920/1251] eta 0:01:33 lr 0.000787 time 0.2865 (0.2840) loss 5.2598 (4.7215) grad_norm 1.4474 (1.4944) [2021-04-15 15:07:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][930/1251] eta 0:01:31 lr 0.000787 time 0.2718 (0.2838) loss 5.2050 (4.7253) grad_norm 1.4489 (1.4934) [2021-04-15 15:07:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][940/1251] eta 0:01:28 lr 0.000788 time 0.2939 (0.2838) loss 4.9929 (4.7231) grad_norm 1.1570 (1.4911) [2021-04-15 15:07:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][950/1251] eta 0:01:25 lr 0.000788 time 0.2654 (0.2837) loss 3.8603 (4.7176) grad_norm 1.6528 (1.4913) [2021-04-15 15:07:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][960/1251] eta 0:01:22 lr 0.000789 time 0.2796 (0.2836) loss 5.0895 (4.7179) grad_norm 1.6304 (1.4904) [2021-04-15 15:07:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][970/1251] eta 0:01:19 lr 0.000789 time 0.2678 (0.2836) loss 4.7039 (4.7190) grad_norm 1.2650 (1.4890) [2021-04-15 15:07:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][980/1251] eta 0:01:16 lr 0.000789 time 0.2616 (0.2835) loss 4.9152 (4.7196) grad_norm 1.5133 (1.4885) [2021-04-15 15:07:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][990/1251] eta 0:01:13 lr 0.000790 time 0.2680 (0.2834) loss 4.6182 (4.7195) grad_norm 1.2694 (1.4894) [2021-04-15 15:07:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][1000/1251] eta 0:01:11 lr 0.000790 time 0.2900 (0.2833) loss 4.8248 (4.7210) grad_norm 1.2611 (1.4879) [2021-04-15 15:07:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][1010/1251] eta 0:01:08 lr 0.000791 time 0.2753 (0.2833) loss 5.2957 (4.7240) grad_norm 1.5014 (1.4880) [2021-04-15 15:07:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][1020/1251] eta 0:01:05 lr 0.000791 time 0.2789 (0.2832) loss 5.0323 (4.7238) grad_norm 1.4153 (1.4866) [2021-04-15 15:07:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][1030/1251] eta 0:01:02 lr 0.000791 time 0.2657 (0.2832) loss 4.9087 (4.7203) grad_norm 1.3455 (1.4877) [2021-04-15 15:07:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][1040/1251] eta 0:00:59 lr 0.000792 time 0.2562 (0.2831) loss 5.1894 (4.7237) grad_norm 1.6858 (1.4881) [2021-04-15 15:07:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][1050/1251] eta 0:00:56 lr 0.000792 time 0.2803 (0.2830) loss 4.0501 (4.7226) grad_norm 1.2647 (1.4879) [2021-04-15 15:07:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][1060/1251] eta 0:00:54 lr 0.000793 time 0.2581 (0.2830) loss 4.3515 (4.7237) grad_norm 1.5971 (1.4883) [2021-04-15 15:07:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][1070/1251] eta 0:00:51 lr 0.000793 time 0.2914 (0.2830) loss 5.3555 (4.7221) grad_norm 1.4479 (1.4877) [2021-04-15 15:07:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][1080/1251] eta 0:00:48 lr 0.000793 time 0.2745 (0.2829) loss 3.7712 (4.7209) grad_norm 1.1772 (1.4864) [2021-04-15 15:07:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][1090/1251] eta 0:00:45 lr 0.000794 time 0.2710 (0.2829) loss 4.6450 (4.7194) grad_norm 1.5399 (1.4862) [2021-04-15 15:07:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][1100/1251] eta 0:00:42 lr 0.000794 time 0.2863 (0.2828) loss 4.8675 (4.7182) grad_norm 1.4143 (1.4856) [2021-04-15 15:07:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][1110/1251] eta 0:00:39 lr 0.000795 time 0.2486 (0.2827) loss 5.1505 (4.7178) grad_norm 1.2747 (1.4846) [2021-04-15 15:08:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][1120/1251] eta 0:00:37 lr 0.000795 time 0.2922 (0.2827) loss 3.4992 (4.7196) grad_norm 1.3474 (1.4840) [2021-04-15 15:08:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][1130/1251] eta 0:00:34 lr 0.000795 time 0.2680 (0.2827) loss 5.2765 (4.7219) grad_norm 1.3930 (1.4831) [2021-04-15 15:08:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][1140/1251] eta 0:00:31 lr 0.000796 time 0.2792 (0.2826) loss 3.7273 (4.7199) grad_norm 1.1784 (1.4831) [2021-04-15 15:08:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][1150/1251] eta 0:00:28 lr 0.000796 time 0.2609 (0.2826) loss 5.0437 (4.7207) grad_norm 1.6846 (1.4832) [2021-04-15 15:08:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][1160/1251] eta 0:00:25 lr 0.000797 time 0.2755 (0.2825) loss 5.7355 (4.7199) grad_norm 1.4169 (1.4847) [2021-04-15 15:08:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][1170/1251] eta 0:00:22 lr 0.000797 time 0.2803 (0.2824) loss 4.4489 (4.7201) grad_norm 1.4057 (1.4868) [2021-04-15 15:08:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][1180/1251] eta 0:00:20 lr 0.000797 time 0.2922 (0.2824) loss 4.5246 (4.7232) grad_norm 1.1682 (1.4860) [2021-04-15 15:08:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][1190/1251] eta 0:00:17 lr 0.000798 time 0.2868 (0.2824) loss 4.1431 (4.7245) grad_norm 1.6820 (1.4851) [2021-04-15 15:08:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][1200/1251] eta 0:00:14 lr 0.000798 time 0.2575 (0.2823) loss 5.2007 (4.7255) grad_norm 1.3605 (1.4844) [2021-04-15 15:08:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][1210/1251] eta 0:00:11 lr 0.000799 time 0.2754 (0.2823) loss 3.6747 (4.7238) grad_norm 1.3520 (1.4834) [2021-04-15 15:08:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][1220/1251] eta 0:00:08 lr 0.000799 time 0.2663 (0.2822) loss 4.0804 (4.7230) grad_norm 1.4062 (1.4825) [2021-04-15 15:08:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][1230/1251] eta 0:00:05 lr 0.000799 time 0.2937 (0.2822) loss 5.3887 (4.7227) grad_norm 1.4052 (1.4815) [2021-04-15 15:08:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][1240/1251] eta 0:00:03 lr 0.000800 time 0.2488 (0.2820) loss 5.2502 (4.7220) grad_norm 1.2947 (1.4810) [2021-04-15 15:08:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [15/300][1250/1251] eta 0:00:00 lr 0.000800 time 0.2489 (0.2818) loss 4.7058 (4.7205) grad_norm 1.3629 (1.4807) [2021-04-15 15:08:38 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 15 training takes 0:05:54 [2021-04-15 15:08:38 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_15.pth saving...... [2021-04-15 15:08:53 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_15.pth saved !!! [2021-04-15 15:08:55 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.207 (1.207) Loss 1.9721 (1.9721) Acc@1 54.785 (54.785) Acc@5 81.152 (81.152) [2021-04-15 15:08:56 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.132 (0.251) Loss 1.9826 (2.0123) Acc@1 56.348 (55.442) Acc@5 80.469 (79.474) [2021-04-15 15:08:58 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.142 (0.235) Loss 1.9157 (2.0166) Acc@1 55.469 (54.976) Acc@5 81.641 (79.436) [2021-04-15 15:09:01 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.113 (0.227) Loss 2.0030 (2.0139) Acc@1 54.688 (55.094) Acc@5 80.664 (79.609) [2021-04-15 15:09:03 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.301 (0.222) Loss 2.0375 (2.0163) Acc@1 55.859 (55.057) Acc@5 79.199 (79.628) [2021-04-15 15:09:04 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 55.260 Acc@5 79.816 [2021-04-15 15:09:04 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 55.3% [2021-04-15 15:09:04 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 55.26% [2021-04-15 15:09:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][0/1251] eta 1:17:47 lr 0.000800 time 3.7310 (3.7310) loss 4.4554 (4.4554) grad_norm 1.4116 (1.4116) [2021-04-15 15:09:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][10/1251] eta 0:12:29 lr 0.000801 time 0.2649 (0.6042) loss 4.8509 (4.7767) grad_norm 1.4029 (1.4575) [2021-04-15 15:09:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][20/1251] eta 0:09:11 lr 0.000801 time 0.2696 (0.4477) loss 4.1093 (4.7842) grad_norm 1.4543 (1.5060) [2021-04-15 15:09:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][30/1251] eta 0:08:03 lr 0.000801 time 0.3016 (0.3956) loss 5.2145 (4.8638) grad_norm 1.8111 (1.5020) [2021-04-15 15:09:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][40/1251] eta 0:07:23 lr 0.000802 time 0.2752 (0.3658) loss 4.2936 (4.8584) grad_norm 1.6072 (1.5153) [2021-04-15 15:09:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][50/1251] eta 0:06:58 lr 0.000802 time 0.2779 (0.3487) loss 4.2906 (4.8287) grad_norm 1.2270 (1.4905) [2021-04-15 15:09:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][60/1251] eta 0:06:41 lr 0.000803 time 0.2871 (0.3372) loss 4.7548 (4.7765) grad_norm 1.0789 (1.4532) [2021-04-15 15:09:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][70/1251] eta 0:06:28 lr 0.000803 time 0.2769 (0.3285) loss 4.0377 (4.7058) grad_norm 1.6807 (1.4562) [2021-04-15 15:09:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][80/1251] eta 0:06:17 lr 0.000803 time 0.2878 (0.3227) loss 4.5339 (4.7187) grad_norm 1.2867 (1.4577) [2021-04-15 15:09:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][90/1251] eta 0:06:08 lr 0.000804 time 0.2774 (0.3177) loss 3.6871 (4.7070) grad_norm 1.3784 (1.4642) [2021-04-15 15:09:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][100/1251] eta 0:06:00 lr 0.000804 time 0.2760 (0.3134) loss 5.1515 (4.7230) grad_norm 1.2083 (1.4565) [2021-04-15 15:09:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][110/1251] eta 0:05:53 lr 0.000805 time 0.2731 (0.3100) loss 4.9947 (4.7439) grad_norm 1.6356 (1.4496) [2021-04-15 15:09:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][120/1251] eta 0:05:47 lr 0.000805 time 0.2949 (0.3072) loss 5.0522 (4.7650) grad_norm 2.0010 (1.4500) [2021-04-15 15:09:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][130/1251] eta 0:05:41 lr 0.000805 time 0.2734 (0.3047) loss 4.2297 (4.7647) grad_norm 1.3506 (1.4608) [2021-04-15 15:09:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][140/1251] eta 0:05:36 lr 0.000806 time 0.2810 (0.3031) loss 3.6259 (4.7619) grad_norm 1.7231 (1.4616) [2021-04-15 15:09:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][150/1251] eta 0:05:31 lr 0.000806 time 0.2739 (0.3015) loss 5.5468 (4.7683) grad_norm 1.2542 (1.4567) [2021-04-15 15:09:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][160/1251] eta 0:05:27 lr 0.000807 time 0.2889 (0.3002) loss 4.5320 (4.7720) grad_norm 1.2578 (1.4672) [2021-04-15 15:09:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][170/1251] eta 0:05:23 lr 0.000807 time 0.2722 (0.2990) loss 5.3597 (4.7743) grad_norm 1.2996 (1.4666) [2021-04-15 15:09:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][180/1251] eta 0:05:18 lr 0.000807 time 0.2677 (0.2978) loss 5.1717 (4.7744) grad_norm 1.3782 (1.4603) [2021-04-15 15:10:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][190/1251] eta 0:05:14 lr 0.000808 time 0.2521 (0.2965) loss 5.2862 (4.7647) grad_norm 1.2540 (1.4669) [2021-04-15 15:10:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][200/1251] eta 0:05:11 lr 0.000808 time 0.2801 (0.2960) loss 5.0921 (4.7690) grad_norm 1.4357 (1.4640) [2021-04-15 15:10:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][210/1251] eta 0:05:07 lr 0.000809 time 0.2707 (0.2952) loss 5.4158 (4.7755) grad_norm 1.3528 (1.4667) [2021-04-15 15:10:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][220/1251] eta 0:05:03 lr 0.000809 time 0.2985 (0.2945) loss 4.9346 (4.7664) grad_norm 2.4456 (1.4741) [2021-04-15 15:10:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][230/1251] eta 0:04:59 lr 0.000809 time 0.2858 (0.2938) loss 3.8457 (4.7465) grad_norm 1.9810 (1.4781) [2021-04-15 15:10:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][240/1251] eta 0:04:56 lr 0.000810 time 0.2950 (0.2932) loss 4.8205 (4.7333) grad_norm 1.2799 (1.4765) [2021-04-15 15:10:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][250/1251] eta 0:04:52 lr 0.000810 time 0.2761 (0.2926) loss 5.0534 (4.7237) grad_norm 1.5316 (1.4779) [2021-04-15 15:10:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][260/1251] eta 0:04:49 lr 0.000811 time 0.2916 (0.2921) loss 5.6248 (4.7315) grad_norm 1.5649 (1.4770) [2021-04-15 15:10:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][270/1251] eta 0:04:46 lr 0.000811 time 0.2688 (0.2916) loss 4.9211 (4.7348) grad_norm 1.2128 (1.4784) [2021-04-15 15:10:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][280/1251] eta 0:04:42 lr 0.000811 time 0.2891 (0.2911) loss 4.7882 (4.7252) grad_norm 1.1711 (1.4766) [2021-04-15 15:10:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][290/1251] eta 0:04:39 lr 0.000812 time 0.2885 (0.2909) loss 4.9556 (4.7218) grad_norm 1.4235 (1.4720) [2021-04-15 15:10:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][300/1251] eta 0:04:37 lr 0.000812 time 0.2872 (0.2916) loss 3.7066 (4.7109) grad_norm 1.2762 (1.4672) [2021-04-15 15:10:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][310/1251] eta 0:04:33 lr 0.000813 time 0.2721 (0.2910) loss 3.8573 (4.7068) grad_norm 1.3028 (1.4685) [2021-04-15 15:10:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][320/1251] eta 0:04:30 lr 0.000813 time 0.2851 (0.2908) loss 5.0906 (4.7029) grad_norm 1.1984 (1.4647) [2021-04-15 15:10:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][330/1251] eta 0:04:27 lr 0.000813 time 0.2866 (0.2905) loss 4.6918 (4.7014) grad_norm 1.3189 (1.4625) [2021-04-15 15:10:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][340/1251] eta 0:04:24 lr 0.000814 time 0.2992 (0.2901) loss 4.8363 (4.7058) grad_norm 1.1774 (1.4622) [2021-04-15 15:10:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][350/1251] eta 0:04:20 lr 0.000814 time 0.2806 (0.2896) loss 4.3922 (4.7127) grad_norm 1.2797 (1.4581) [2021-04-15 15:10:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][360/1251] eta 0:04:17 lr 0.000815 time 0.2555 (0.2892) loss 4.8695 (4.7129) grad_norm 1.2210 (1.4554) [2021-04-15 15:10:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][370/1251] eta 0:04:14 lr 0.000815 time 0.2530 (0.2889) loss 4.7021 (4.7080) grad_norm 1.2686 (1.4548) [2021-04-15 15:10:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][380/1251] eta 0:04:11 lr 0.000815 time 0.2591 (0.2886) loss 5.3019 (4.7126) grad_norm 1.3261 (1.4503) [2021-04-15 15:10:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][390/1251] eta 0:04:08 lr 0.000816 time 0.2770 (0.2883) loss 4.8391 (4.7147) grad_norm 1.2844 (1.4475) [2021-04-15 15:11:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][400/1251] eta 0:04:05 lr 0.000816 time 0.2948 (0.2882) loss 3.7654 (4.7027) grad_norm 1.5830 (1.4508) [2021-04-15 15:11:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][410/1251] eta 0:04:02 lr 0.000817 time 0.2787 (0.2879) loss 4.4006 (4.6945) grad_norm 1.5880 (1.4504) [2021-04-15 15:11:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][420/1251] eta 0:03:58 lr 0.000817 time 0.2899 (0.2875) loss 3.8161 (4.6938) grad_norm 2.0623 (1.4477) [2021-04-15 15:11:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][430/1251] eta 0:03:55 lr 0.000817 time 0.2634 (0.2874) loss 3.6951 (4.6931) grad_norm 1.3053 (1.4494) [2021-04-15 15:11:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][440/1251] eta 0:03:53 lr 0.000818 time 0.2726 (0.2873) loss 5.7734 (4.6952) grad_norm 1.7022 (1.4486) [2021-04-15 15:11:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][450/1251] eta 0:03:50 lr 0.000818 time 0.2920 (0.2871) loss 4.9898 (4.7054) grad_norm 1.3626 (1.4470) [2021-04-15 15:11:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][460/1251] eta 0:03:46 lr 0.000819 time 0.2882 (0.2869) loss 4.9149 (4.7028) grad_norm 1.1103 (1.4456) [2021-04-15 15:11:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][470/1251] eta 0:03:43 lr 0.000819 time 0.2926 (0.2867) loss 5.2292 (4.6973) grad_norm 1.5471 (1.4442) [2021-04-15 15:11:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][480/1251] eta 0:03:40 lr 0.000819 time 0.2839 (0.2866) loss 5.4279 (4.7016) grad_norm 1.8455 (1.4462) [2021-04-15 15:11:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][490/1251] eta 0:03:37 lr 0.000820 time 0.2928 (0.2863) loss 5.0486 (4.6990) grad_norm 1.6695 (1.4470) [2021-04-15 15:11:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][500/1251] eta 0:03:34 lr 0.000820 time 0.2776 (0.2862) loss 4.6807 (4.6947) grad_norm 1.5737 (1.4489) [2021-04-15 15:11:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][510/1251] eta 0:03:31 lr 0.000821 time 0.2612 (0.2859) loss 5.1595 (4.6946) grad_norm 1.1824 (1.4500) [2021-04-15 15:11:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][520/1251] eta 0:03:28 lr 0.000821 time 0.2771 (0.2858) loss 4.3909 (4.6955) grad_norm 1.9453 (1.4488) [2021-04-15 15:11:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][530/1251] eta 0:03:25 lr 0.000821 time 0.2710 (0.2856) loss 4.7522 (4.6981) grad_norm 1.0977 (1.4481) [2021-04-15 15:11:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][540/1251] eta 0:03:22 lr 0.000822 time 0.2688 (0.2854) loss 4.8431 (4.6949) grad_norm 1.0829 (1.4491) [2021-04-15 15:11:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][550/1251] eta 0:03:20 lr 0.000822 time 0.2810 (0.2854) loss 4.5351 (4.6991) grad_norm 1.7407 (1.4502) [2021-04-15 15:11:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][560/1251] eta 0:03:17 lr 0.000823 time 0.2596 (0.2852) loss 5.4330 (4.6977) grad_norm 1.2773 (1.4490) [2021-04-15 15:11:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][570/1251] eta 0:03:14 lr 0.000823 time 0.2514 (0.2850) loss 4.0558 (4.6984) grad_norm 1.4745 (1.4500) [2021-04-15 15:11:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][580/1251] eta 0:03:11 lr 0.000823 time 0.2668 (0.2854) loss 4.9693 (4.7017) grad_norm 1.8605 (1.4526) [2021-04-15 15:11:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][590/1251] eta 0:03:08 lr 0.000824 time 0.2491 (0.2851) loss 3.7694 (4.6976) grad_norm 1.6439 (1.4534) [2021-04-15 15:11:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][600/1251] eta 0:03:05 lr 0.000824 time 0.2678 (0.2850) loss 4.3508 (4.6982) grad_norm 1.6433 (1.4547) [2021-04-15 15:11:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][610/1251] eta 0:03:02 lr 0.000825 time 0.2760 (0.2849) loss 5.0645 (4.7003) grad_norm 1.2314 (1.4543) [2021-04-15 15:12:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][620/1251] eta 0:02:59 lr 0.000825 time 0.2884 (0.2848) loss 4.1983 (4.6947) grad_norm 1.3138 (1.4533) [2021-04-15 15:12:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][630/1251] eta 0:02:56 lr 0.000825 time 0.2986 (0.2846) loss 4.9977 (4.7011) grad_norm 1.2692 (1.4540) [2021-04-15 15:12:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][640/1251] eta 0:02:53 lr 0.000826 time 0.3171 (0.2845) loss 4.0040 (4.6969) grad_norm 1.5451 (1.4537) [2021-04-15 15:12:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][650/1251] eta 0:02:50 lr 0.000826 time 0.2775 (0.2844) loss 4.7773 (4.6965) grad_norm 1.7188 (1.4523) [2021-04-15 15:12:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][660/1251] eta 0:02:48 lr 0.000827 time 0.2773 (0.2843) loss 4.7566 (4.6979) grad_norm 1.6658 (1.4531) [2021-04-15 15:12:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][670/1251] eta 0:02:45 lr 0.000827 time 0.2713 (0.2841) loss 5.1822 (4.6926) grad_norm 1.5774 (1.4560) [2021-04-15 15:12:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][680/1251] eta 0:02:42 lr 0.000827 time 0.2929 (0.2840) loss 3.5595 (4.6889) grad_norm 1.5315 (1.4566) [2021-04-15 15:12:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][690/1251] eta 0:02:39 lr 0.000828 time 0.2821 (0.2839) loss 4.6713 (4.6874) grad_norm 1.1981 (1.4538) [2021-04-15 15:12:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][700/1251] eta 0:02:36 lr 0.000828 time 0.2741 (0.2838) loss 4.9247 (4.6851) grad_norm 1.2099 (1.4531) [2021-04-15 15:12:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][710/1251] eta 0:02:33 lr 0.000829 time 0.2966 (0.2837) loss 5.5274 (4.6832) grad_norm 1.3924 (1.4530) [2021-04-15 15:12:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][720/1251] eta 0:02:30 lr 0.000829 time 0.2837 (0.2836) loss 4.5151 (4.6762) grad_norm 1.5299 (1.4542) [2021-04-15 15:12:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][730/1251] eta 0:02:27 lr 0.000829 time 0.2731 (0.2835) loss 3.5769 (4.6727) grad_norm 1.5348 (1.4548) [2021-04-15 15:12:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][740/1251] eta 0:02:24 lr 0.000830 time 0.2848 (0.2834) loss 4.8571 (4.6743) grad_norm 1.4131 (1.4542) [2021-04-15 15:12:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][750/1251] eta 0:02:21 lr 0.000830 time 0.2538 (0.2833) loss 4.2893 (4.6702) grad_norm 1.6214 (1.4541) [2021-04-15 15:12:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][760/1251] eta 0:02:19 lr 0.000831 time 0.2975 (0.2832) loss 4.5097 (4.6756) grad_norm 1.2986 (1.4552) [2021-04-15 15:12:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][770/1251] eta 0:02:16 lr 0.000831 time 0.2914 (0.2832) loss 5.0131 (4.6747) grad_norm 1.1574 (1.4544) [2021-04-15 15:12:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][780/1251] eta 0:02:13 lr 0.000831 time 0.3134 (0.2831) loss 5.2496 (4.6776) grad_norm 2.0656 (1.4552) [2021-04-15 15:12:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][790/1251] eta 0:02:10 lr 0.000832 time 0.3055 (0.2831) loss 4.4433 (4.6776) grad_norm 1.1993 (1.4540) [2021-04-15 15:12:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][800/1251] eta 0:02:07 lr 0.000832 time 0.2889 (0.2829) loss 4.6498 (4.6744) grad_norm 1.4958 (1.4563) [2021-04-15 15:12:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][810/1251] eta 0:02:04 lr 0.000833 time 0.2865 (0.2829) loss 3.9402 (4.6724) grad_norm 1.3911 (1.4573) [2021-04-15 15:12:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][820/1251] eta 0:02:01 lr 0.000833 time 0.2533 (0.2828) loss 5.1705 (4.6761) grad_norm 1.1330 (1.4581) [2021-04-15 15:12:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][830/1251] eta 0:01:59 lr 0.000833 time 0.2735 (0.2828) loss 3.2544 (4.6727) grad_norm 1.2035 (1.4567) [2021-04-15 15:13:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][840/1251] eta 0:01:56 lr 0.000834 time 0.2645 (0.2827) loss 3.9490 (4.6752) grad_norm 1.7447 (1.4566) [2021-04-15 15:13:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][850/1251] eta 0:01:53 lr 0.000834 time 0.2950 (0.2826) loss 4.6509 (4.6752) grad_norm 1.3390 (1.4571) [2021-04-15 15:13:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][860/1251] eta 0:01:50 lr 0.000835 time 0.2687 (0.2825) loss 3.9527 (4.6751) grad_norm 1.2698 (1.4558) [2021-04-15 15:13:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][870/1251] eta 0:01:47 lr 0.000835 time 0.2678 (0.2824) loss 4.2916 (4.6746) grad_norm 1.1991 (1.4549) [2021-04-15 15:13:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][880/1251] eta 0:01:44 lr 0.000835 time 0.3032 (0.2823) loss 5.2503 (4.6770) grad_norm 1.4891 (1.4525) [2021-04-15 15:13:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][890/1251] eta 0:01:41 lr 0.000836 time 0.2798 (0.2823) loss 4.9203 (4.6803) grad_norm 1.4129 (1.4515) [2021-04-15 15:13:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][900/1251] eta 0:01:39 lr 0.000836 time 0.2719 (0.2822) loss 5.4204 (4.6801) grad_norm 1.3005 (1.4500) [2021-04-15 15:13:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][910/1251] eta 0:01:36 lr 0.000837 time 0.2840 (0.2822) loss 4.5646 (4.6798) grad_norm 1.3976 (1.4496) [2021-04-15 15:13:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][920/1251] eta 0:01:33 lr 0.000837 time 0.2780 (0.2821) loss 3.7393 (4.6817) grad_norm 1.1729 (1.4487) [2021-04-15 15:13:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][930/1251] eta 0:01:30 lr 0.000837 time 0.2707 (0.2821) loss 5.4015 (4.6814) grad_norm 1.3253 (1.4475) [2021-04-15 15:13:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][940/1251] eta 0:01:27 lr 0.000838 time 0.2870 (0.2820) loss 4.2257 (4.6830) grad_norm 1.2881 (1.4462) [2021-04-15 15:13:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][950/1251] eta 0:01:24 lr 0.000838 time 0.2657 (0.2819) loss 4.9650 (4.6800) grad_norm 1.6570 (1.4460) [2021-04-15 15:13:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][960/1251] eta 0:01:22 lr 0.000839 time 0.2710 (0.2819) loss 5.5307 (4.6796) grad_norm 1.3623 (1.4466) [2021-04-15 15:13:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][970/1251] eta 0:01:19 lr 0.000839 time 0.2770 (0.2818) loss 4.5316 (4.6771) grad_norm 1.3221 (1.4461) [2021-04-15 15:13:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][980/1251] eta 0:01:16 lr 0.000839 time 0.3085 (0.2818) loss 3.4958 (4.6749) grad_norm 1.6490 (1.4464) [2021-04-15 15:13:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][990/1251] eta 0:01:13 lr 0.000840 time 0.2617 (0.2817) loss 4.6974 (4.6737) grad_norm 1.7853 (1.4461) [2021-04-15 15:13:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][1000/1251] eta 0:01:10 lr 0.000840 time 0.2470 (0.2816) loss 5.2024 (4.6719) grad_norm 1.7747 (1.4458) [2021-04-15 15:13:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][1010/1251] eta 0:01:07 lr 0.000841 time 0.2687 (0.2816) loss 4.6386 (4.6698) grad_norm 1.2500 (1.4451) [2021-04-15 15:13:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][1020/1251] eta 0:01:05 lr 0.000841 time 0.2712 (0.2815) loss 5.1602 (4.6704) grad_norm 1.3378 (1.4438) [2021-04-15 15:13:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][1030/1251] eta 0:01:02 lr 0.000841 time 0.2789 (0.2815) loss 5.1611 (4.6723) grad_norm 1.3892 (1.4444) [2021-04-15 15:13:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][1040/1251] eta 0:00:59 lr 0.000842 time 0.2700 (0.2814) loss 4.4800 (4.6708) grad_norm 1.2571 (1.4434) [2021-04-15 15:14:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][1050/1251] eta 0:00:56 lr 0.000842 time 0.2618 (0.2814) loss 3.4256 (4.6679) grad_norm 1.3657 (1.4424) [2021-04-15 15:14:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][1060/1251] eta 0:00:53 lr 0.000843 time 0.2778 (0.2813) loss 4.7569 (4.6670) grad_norm 1.1717 (1.4407) [2021-04-15 15:14:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][1070/1251] eta 0:00:50 lr 0.000843 time 0.2859 (0.2813) loss 5.0244 (4.6681) grad_norm 1.7370 (1.4421) [2021-04-15 15:14:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][1080/1251] eta 0:00:48 lr 0.000843 time 0.2895 (0.2812) loss 5.7312 (4.6703) grad_norm 1.2885 (1.4422) [2021-04-15 15:14:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][1090/1251] eta 0:00:45 lr 0.000844 time 0.2709 (0.2811) loss 4.6224 (4.6676) grad_norm 2.5005 (1.4432) [2021-04-15 15:14:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][1100/1251] eta 0:00:42 lr 0.000844 time 0.2613 (0.2811) loss 4.4745 (4.6679) grad_norm 1.5763 (1.4431) [2021-04-15 15:14:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][1110/1251] eta 0:00:39 lr 0.000845 time 0.2635 (0.2810) loss 4.7410 (4.6664) grad_norm 1.7851 (1.4428) [2021-04-15 15:14:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][1120/1251] eta 0:00:36 lr 0.000845 time 0.2569 (0.2810) loss 5.0503 (4.6661) grad_norm 1.4933 (1.4421) [2021-04-15 15:14:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][1130/1251] eta 0:00:33 lr 0.000845 time 0.2764 (0.2809) loss 4.6591 (4.6683) grad_norm 1.4694 (1.4409) [2021-04-15 15:14:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][1140/1251] eta 0:00:31 lr 0.000846 time 0.2682 (0.2809) loss 4.1093 (4.6647) grad_norm 1.5146 (1.4412) [2021-04-15 15:14:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][1150/1251] eta 0:00:28 lr 0.000846 time 0.3913 (0.2809) loss 5.2098 (4.6643) grad_norm 1.2006 (1.4409) [2021-04-15 15:14:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][1160/1251] eta 0:00:25 lr 0.000847 time 0.2753 (0.2810) loss 4.3365 (4.6650) grad_norm 1.6874 (1.4404) [2021-04-15 15:14:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][1170/1251] eta 0:00:22 lr 0.000847 time 0.3018 (0.2809) loss 4.2684 (4.6630) grad_norm 1.4524 (1.4409) [2021-04-15 15:14:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][1180/1251] eta 0:00:19 lr 0.000847 time 0.2723 (0.2809) loss 4.6841 (4.6636) grad_norm 1.4521 (1.4398) [2021-04-15 15:14:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][1190/1251] eta 0:00:17 lr 0.000848 time 0.2817 (0.2808) loss 4.7125 (4.6634) grad_norm 1.1926 (1.4384) [2021-04-15 15:14:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][1200/1251] eta 0:00:14 lr 0.000848 time 0.2983 (0.2808) loss 5.1470 (4.6660) grad_norm 1.5744 (1.4395) [2021-04-15 15:14:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][1210/1251] eta 0:00:11 lr 0.000849 time 0.2780 (0.2808) loss 3.8341 (4.6660) grad_norm 1.2420 (1.4395) [2021-04-15 15:14:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][1220/1251] eta 0:00:08 lr 0.000849 time 0.2872 (0.2808) loss 3.9845 (4.6658) grad_norm 1.3418 (1.4388) [2021-04-15 15:14:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][1230/1251] eta 0:00:05 lr 0.000849 time 0.2883 (0.2807) loss 4.3309 (4.6644) grad_norm 1.6071 (1.4387) [2021-04-15 15:14:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][1240/1251] eta 0:00:03 lr 0.000850 time 0.2488 (0.2806) loss 4.5530 (4.6658) grad_norm 1.1647 (1.4376) [2021-04-15 15:14:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [16/300][1250/1251] eta 0:00:00 lr 0.000850 time 0.2486 (0.2804) loss 3.9997 (4.6641) grad_norm 1.6006 (1.4367) [2021-04-15 15:14:57 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 16 training takes 0:05:52 [2021-04-15 15:14:57 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_16.pth saving...... [2021-04-15 15:15:12 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_16.pth saved !!! [2021-04-15 15:15:13 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.188 (1.188) Loss 1.9609 (1.9609) Acc@1 58.203 (58.203) Acc@5 81.250 (81.250) [2021-04-15 15:15:14 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.085 (0.210) Loss 2.0063 (1.9614) Acc@1 56.543 (56.747) Acc@5 80.273 (80.913) [2021-04-15 15:15:17 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.108 (0.243) Loss 2.0190 (1.9696) Acc@1 55.664 (56.655) Acc@5 79.980 (80.878) [2021-04-15 15:15:19 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.167 (0.224) Loss 1.8879 (1.9723) Acc@1 59.668 (56.540) Acc@5 82.812 (80.743) [2021-04-15 15:15:21 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.217) Loss 1.8822 (1.9747) Acc@1 58.105 (56.388) Acc@5 81.738 (80.669) [2021-04-15 15:15:23 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 56.378 Acc@5 80.800 [2021-04-15 15:15:23 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 56.4% [2021-04-15 15:15:23 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 56.38% [2021-04-15 15:15:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][0/1251] eta 1:09:47 lr 0.000850 time 3.3472 (3.3472) loss 4.3193 (4.3193) grad_norm 1.4653 (1.4653) [2021-04-15 15:15:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][10/1251] eta 0:12:06 lr 0.000851 time 0.2900 (0.5855) loss 3.6103 (4.5560) grad_norm 1.2042 (1.4789) [2021-04-15 15:15:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][20/1251] eta 0:09:06 lr 0.000851 time 0.3228 (0.4439) loss 3.6516 (4.6388) grad_norm 1.4507 (1.4272) [2021-04-15 15:15:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][30/1251] eta 0:07:57 lr 0.000851 time 0.2996 (0.3910) loss 3.9443 (4.6805) grad_norm 1.5049 (1.3967) [2021-04-15 15:15:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][40/1251] eta 0:07:19 lr 0.000852 time 0.2691 (0.3627) loss 5.6455 (4.6762) grad_norm 1.2765 (1.3989) [2021-04-15 15:15:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][50/1251] eta 0:06:54 lr 0.000852 time 0.2464 (0.3454) loss 4.8367 (4.6917) grad_norm 1.5598 (1.4346) [2021-04-15 15:15:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][60/1251] eta 0:06:38 lr 0.000853 time 0.2703 (0.3346) loss 4.4981 (4.7348) grad_norm 1.5239 (1.4238) [2021-04-15 15:15:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][70/1251] eta 0:06:25 lr 0.000853 time 0.2531 (0.3265) loss 4.5965 (4.7430) grad_norm 1.7927 (1.4335) [2021-04-15 15:15:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][80/1251] eta 0:06:15 lr 0.000853 time 0.2889 (0.3210) loss 4.3987 (4.7348) grad_norm 1.7627 (1.4241) [2021-04-15 15:15:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][90/1251] eta 0:06:07 lr 0.000854 time 0.2788 (0.3162) loss 4.9406 (4.6987) grad_norm 1.3013 (1.4246) [2021-04-15 15:15:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][100/1251] eta 0:06:00 lr 0.000854 time 0.2836 (0.3129) loss 4.7894 (4.7118) grad_norm 1.3565 (1.4231) [2021-04-15 15:15:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][110/1251] eta 0:05:54 lr 0.000855 time 0.2915 (0.3103) loss 4.0321 (4.6781) grad_norm 1.4667 (1.4174) [2021-04-15 15:16:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][120/1251] eta 0:05:48 lr 0.000855 time 0.3137 (0.3084) loss 5.0352 (4.6816) grad_norm 1.1242 (1.4182) [2021-04-15 15:16:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][130/1251] eta 0:05:42 lr 0.000855 time 0.3011 (0.3059) loss 5.6667 (4.6883) grad_norm 1.3190 (1.4169) [2021-04-15 15:16:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][140/1251] eta 0:05:36 lr 0.000856 time 0.2669 (0.3033) loss 4.2787 (4.6943) grad_norm 1.0790 (inf) [2021-04-15 15:16:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][150/1251] eta 0:05:32 lr 0.000856 time 0.2572 (0.3017) loss 3.6762 (4.6765) grad_norm 1.1986 (inf) [2021-04-15 15:16:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][160/1251] eta 0:05:27 lr 0.000857 time 0.2863 (0.3002) loss 4.7876 (4.6823) grad_norm 1.2803 (inf) [2021-04-15 15:16:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][170/1251] eta 0:05:23 lr 0.000857 time 0.2965 (0.2989) loss 4.7486 (4.6827) grad_norm 1.3554 (inf) [2021-04-15 15:16:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][180/1251] eta 0:05:19 lr 0.000857 time 0.2789 (0.2981) loss 4.5828 (4.6873) grad_norm 1.3148 (inf) [2021-04-15 15:16:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][190/1251] eta 0:05:14 lr 0.000858 time 0.2781 (0.2968) loss 3.4628 (4.6839) grad_norm 1.1951 (inf) [2021-04-15 15:16:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][200/1251] eta 0:05:10 lr 0.000858 time 0.2712 (0.2958) loss 5.1492 (4.6914) grad_norm 1.5658 (inf) [2021-04-15 15:16:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][210/1251] eta 0:05:06 lr 0.000859 time 0.2743 (0.2949) loss 4.9795 (4.6989) grad_norm 1.8943 (inf) [2021-04-15 15:16:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][220/1251] eta 0:05:03 lr 0.000859 time 0.2817 (0.2941) loss 5.1447 (4.6930) grad_norm 1.3455 (inf) [2021-04-15 15:16:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][230/1251] eta 0:04:59 lr 0.000859 time 0.2912 (0.2934) loss 5.5692 (4.7064) grad_norm 1.1638 (inf) [2021-04-15 15:16:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][240/1251] eta 0:04:56 lr 0.000860 time 0.2725 (0.2928) loss 4.8783 (4.6898) grad_norm 1.2368 (inf) [2021-04-15 15:16:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][250/1251] eta 0:04:52 lr 0.000860 time 0.2694 (0.2923) loss 3.6736 (4.6824) grad_norm 1.3370 (inf) [2021-04-15 15:16:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][260/1251] eta 0:04:49 lr 0.000861 time 0.2724 (0.2917) loss 5.2454 (4.6803) grad_norm 1.5162 (inf) [2021-04-15 15:16:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][270/1251] eta 0:04:45 lr 0.000861 time 0.2797 (0.2911) loss 5.1358 (4.6880) grad_norm 1.4960 (inf) [2021-04-15 15:16:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][280/1251] eta 0:04:42 lr 0.000861 time 0.2786 (0.2905) loss 5.4106 (4.6823) grad_norm 1.3729 (inf) [2021-04-15 15:16:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][290/1251] eta 0:04:38 lr 0.000862 time 0.2881 (0.2900) loss 3.7936 (4.6669) grad_norm 1.5007 (inf) [2021-04-15 15:16:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][300/1251] eta 0:04:35 lr 0.000862 time 0.2622 (0.2894) loss 5.3806 (4.6675) grad_norm 1.2808 (inf) [2021-04-15 15:16:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][310/1251] eta 0:04:32 lr 0.000863 time 0.2792 (0.2891) loss 5.2264 (4.6662) grad_norm 1.2405 (inf) [2021-04-15 15:16:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][320/1251] eta 0:04:28 lr 0.000863 time 0.2645 (0.2888) loss 4.5935 (4.6678) grad_norm 1.2392 (inf) [2021-04-15 15:16:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][330/1251] eta 0:04:25 lr 0.000863 time 0.2640 (0.2886) loss 5.0295 (4.6662) grad_norm 1.1065 (inf) [2021-04-15 15:17:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][340/1251] eta 0:04:22 lr 0.000864 time 0.2832 (0.2883) loss 3.5313 (4.6645) grad_norm 1.4262 (inf) [2021-04-15 15:17:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][350/1251] eta 0:04:19 lr 0.000864 time 0.2837 (0.2879) loss 4.9036 (4.6620) grad_norm 1.1012 (inf) [2021-04-15 15:17:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][360/1251] eta 0:04:17 lr 0.000865 time 0.2850 (0.2885) loss 3.7490 (4.6467) grad_norm 1.5583 (inf) [2021-04-15 15:17:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][370/1251] eta 0:04:13 lr 0.000865 time 0.2686 (0.2881) loss 5.0502 (4.6439) grad_norm 1.1315 (inf) [2021-04-15 15:17:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][380/1251] eta 0:04:10 lr 0.000865 time 0.2669 (0.2877) loss 4.3096 (4.6435) grad_norm 1.5641 (inf) [2021-04-15 15:17:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][390/1251] eta 0:04:07 lr 0.000866 time 0.2865 (0.2874) loss 5.2329 (4.6400) grad_norm 1.3587 (inf) [2021-04-15 15:17:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][400/1251] eta 0:04:04 lr 0.000866 time 0.3023 (0.2873) loss 3.3458 (4.6414) grad_norm 1.7763 (inf) [2021-04-15 15:17:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][410/1251] eta 0:04:01 lr 0.000867 time 0.2946 (0.2870) loss 5.3716 (4.6381) grad_norm 1.4787 (inf) [2021-04-15 15:17:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][420/1251] eta 0:03:58 lr 0.000867 time 0.2854 (0.2869) loss 5.1397 (4.6383) grad_norm 1.3319 (inf) [2021-04-15 15:17:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][430/1251] eta 0:03:55 lr 0.000867 time 0.2868 (0.2866) loss 4.7641 (4.6431) grad_norm 1.2753 (inf) [2021-04-15 15:17:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][440/1251] eta 0:03:52 lr 0.000868 time 0.2580 (0.2864) loss 3.4124 (4.6476) grad_norm 1.4233 (inf) [2021-04-15 15:17:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][450/1251] eta 0:03:49 lr 0.000868 time 0.2642 (0.2862) loss 5.2957 (4.6413) grad_norm 1.5955 (inf) [2021-04-15 15:17:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][460/1251] eta 0:03:46 lr 0.000869 time 0.2931 (0.2860) loss 3.7635 (4.6370) grad_norm 1.3448 (inf) [2021-04-15 15:17:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][470/1251] eta 0:03:43 lr 0.000869 time 0.2826 (0.2858) loss 4.2225 (4.6402) grad_norm 1.5368 (inf) [2021-04-15 15:17:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][480/1251] eta 0:03:40 lr 0.000869 time 0.2887 (0.2857) loss 4.1108 (4.6361) grad_norm 1.1830 (inf) [2021-04-15 15:17:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][490/1251] eta 0:03:37 lr 0.000870 time 0.2722 (0.2855) loss 5.2935 (4.6397) grad_norm 1.3052 (inf) [2021-04-15 15:17:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][500/1251] eta 0:03:34 lr 0.000870 time 0.2530 (0.2852) loss 4.8305 (4.6380) grad_norm 1.7076 (inf) [2021-04-15 15:17:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][510/1251] eta 0:03:31 lr 0.000871 time 0.2766 (0.2851) loss 5.1212 (4.6373) grad_norm 1.1222 (inf) [2021-04-15 15:17:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][520/1251] eta 0:03:28 lr 0.000871 time 0.2899 (0.2850) loss 4.8110 (4.6376) grad_norm 1.3711 (inf) [2021-04-15 15:17:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][530/1251] eta 0:03:25 lr 0.000871 time 0.2752 (0.2848) loss 5.2151 (4.6344) grad_norm 1.4277 (inf) [2021-04-15 15:17:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][540/1251] eta 0:03:22 lr 0.000872 time 0.2947 (0.2847) loss 4.6681 (4.6356) grad_norm 1.1623 (inf) [2021-04-15 15:18:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][550/1251] eta 0:03:19 lr 0.000872 time 0.2830 (0.2845) loss 4.5195 (4.6371) grad_norm 1.1268 (inf) [2021-04-15 15:18:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][560/1251] eta 0:03:16 lr 0.000873 time 0.2681 (0.2843) loss 3.7666 (4.6324) grad_norm 1.5161 (inf) [2021-04-15 15:18:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][570/1251] eta 0:03:13 lr 0.000873 time 0.2665 (0.2842) loss 4.5870 (4.6312) grad_norm 1.3551 (inf) [2021-04-15 15:18:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][580/1251] eta 0:03:10 lr 0.000873 time 0.2914 (0.2842) loss 3.4922 (4.6276) grad_norm 1.2666 (inf) [2021-04-15 15:18:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][590/1251] eta 0:03:07 lr 0.000874 time 0.2745 (0.2840) loss 5.0370 (4.6272) grad_norm 1.3143 (inf) [2021-04-15 15:18:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][600/1251] eta 0:03:04 lr 0.000874 time 0.2801 (0.2839) loss 4.0470 (4.6280) grad_norm 1.3777 (inf) [2021-04-15 15:18:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][610/1251] eta 0:03:01 lr 0.000875 time 0.2858 (0.2838) loss 3.7139 (4.6280) grad_norm 1.1265 (inf) [2021-04-15 15:18:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][620/1251] eta 0:02:59 lr 0.000875 time 0.2793 (0.2837) loss 5.1580 (4.6316) grad_norm 1.6462 (inf) [2021-04-15 15:18:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][630/1251] eta 0:02:56 lr 0.000875 time 0.2746 (0.2836) loss 5.3522 (4.6334) grad_norm 1.2935 (inf) [2021-04-15 15:18:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][640/1251] eta 0:02:53 lr 0.000876 time 0.2930 (0.2835) loss 5.2458 (4.6285) grad_norm 1.3855 (inf) [2021-04-15 15:18:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][650/1251] eta 0:02:50 lr 0.000876 time 0.2783 (0.2834) loss 4.7994 (4.6282) grad_norm 1.3468 (inf) [2021-04-15 15:18:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][660/1251] eta 0:02:47 lr 0.000877 time 0.2730 (0.2833) loss 5.5738 (4.6275) grad_norm 1.0972 (inf) [2021-04-15 15:18:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][670/1251] eta 0:02:44 lr 0.000877 time 0.2790 (0.2833) loss 4.9563 (4.6280) grad_norm 1.0706 (inf) [2021-04-15 15:18:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][680/1251] eta 0:02:41 lr 0.000877 time 0.2758 (0.2832) loss 3.8014 (4.6276) grad_norm 1.3343 (inf) [2021-04-15 15:18:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][690/1251] eta 0:02:38 lr 0.000878 time 0.2729 (0.2831) loss 4.8565 (4.6244) grad_norm 1.2416 (inf) [2021-04-15 15:18:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][700/1251] eta 0:02:35 lr 0.000878 time 0.2986 (0.2830) loss 5.3627 (4.6250) grad_norm 1.1838 (inf) [2021-04-15 15:18:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][710/1251] eta 0:02:33 lr 0.000878 time 0.2581 (0.2829) loss 4.8202 (4.6262) grad_norm 1.3773 (inf) [2021-04-15 15:18:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][720/1251] eta 0:02:30 lr 0.000879 time 0.2792 (0.2829) loss 4.2634 (4.6257) grad_norm 1.2695 (inf) [2021-04-15 15:18:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][730/1251] eta 0:02:27 lr 0.000879 time 0.2957 (0.2828) loss 5.2284 (4.6294) grad_norm 1.3860 (inf) [2021-04-15 15:18:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][740/1251] eta 0:02:24 lr 0.000880 time 0.2748 (0.2827) loss 3.9438 (4.6334) grad_norm 1.2554 (inf) [2021-04-15 15:18:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][750/1251] eta 0:02:21 lr 0.000880 time 0.3124 (0.2826) loss 3.7654 (4.6339) grad_norm 1.7695 (inf) [2021-04-15 15:18:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][760/1251] eta 0:02:18 lr 0.000880 time 0.2966 (0.2826) loss 5.3459 (4.6352) grad_norm 1.2384 (inf) [2021-04-15 15:19:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][770/1251] eta 0:02:15 lr 0.000881 time 0.2642 (0.2826) loss 4.8534 (4.6336) grad_norm 1.3873 (inf) [2021-04-15 15:19:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][780/1251] eta 0:02:13 lr 0.000881 time 0.2756 (0.2825) loss 5.3109 (4.6328) grad_norm 1.2031 (inf) [2021-04-15 15:19:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][790/1251] eta 0:02:10 lr 0.000882 time 0.2743 (0.2825) loss 4.8168 (4.6337) grad_norm 1.5007 (inf) [2021-04-15 15:19:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][800/1251] eta 0:02:07 lr 0.000882 time 0.2723 (0.2824) loss 4.8075 (4.6349) grad_norm 1.1918 (inf) [2021-04-15 15:19:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][810/1251] eta 0:02:04 lr 0.000882 time 0.2739 (0.2823) loss 4.7138 (4.6358) grad_norm 2.3363 (inf) [2021-04-15 15:19:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][820/1251] eta 0:02:01 lr 0.000883 time 0.2796 (0.2823) loss 3.5162 (4.6358) grad_norm 1.2723 (inf) [2021-04-15 15:19:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][830/1251] eta 0:01:58 lr 0.000883 time 0.2825 (0.2823) loss 4.7456 (4.6337) grad_norm 1.3722 (inf) [2021-04-15 15:19:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][840/1251] eta 0:01:55 lr 0.000884 time 0.2708 (0.2821) loss 5.0729 (4.6315) grad_norm 1.2398 (inf) [2021-04-15 15:19:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][850/1251] eta 0:01:53 lr 0.000884 time 0.2947 (0.2820) loss 5.4966 (4.6331) grad_norm 1.6052 (inf) [2021-04-15 15:19:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][860/1251] eta 0:01:50 lr 0.000884 time 0.3037 (0.2820) loss 4.7489 (4.6326) grad_norm 1.5124 (inf) [2021-04-15 15:19:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][870/1251] eta 0:01:47 lr 0.000885 time 0.3088 (0.2820) loss 4.6026 (4.6322) grad_norm 1.2014 (inf) [2021-04-15 15:19:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][880/1251] eta 0:01:44 lr 0.000885 time 0.2696 (0.2819) loss 5.3851 (4.6322) grad_norm 1.3780 (inf) [2021-04-15 15:19:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][890/1251] eta 0:01:41 lr 0.000886 time 0.2769 (0.2819) loss 4.8774 (4.6358) grad_norm 1.1366 (inf) [2021-04-15 15:19:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][900/1251] eta 0:01:38 lr 0.000886 time 0.2673 (0.2818) loss 4.6054 (4.6339) grad_norm 1.4597 (inf) [2021-04-15 15:19:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][910/1251] eta 0:01:36 lr 0.000886 time 0.2551 (0.2817) loss 4.3058 (4.6343) grad_norm 1.2617 (inf) [2021-04-15 15:19:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][920/1251] eta 0:01:33 lr 0.000887 time 0.2807 (0.2816) loss 5.0994 (4.6359) grad_norm 1.3166 (inf) [2021-04-15 15:19:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][930/1251] eta 0:01:30 lr 0.000887 time 0.4408 (0.2818) loss 4.9348 (4.6340) grad_norm 1.4138 (inf) [2021-04-15 15:19:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][940/1251] eta 0:01:27 lr 0.000888 time 0.2724 (0.2818) loss 4.8838 (4.6330) grad_norm 1.3006 (inf) [2021-04-15 15:19:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][950/1251] eta 0:01:24 lr 0.000888 time 0.2711 (0.2817) loss 4.9701 (4.6322) grad_norm 1.3351 (inf) [2021-04-15 15:19:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][960/1251] eta 0:01:21 lr 0.000888 time 0.2742 (0.2816) loss 5.1047 (4.6330) grad_norm 1.2753 (inf) [2021-04-15 15:19:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][970/1251] eta 0:01:19 lr 0.000889 time 0.2792 (0.2816) loss 4.5558 (4.6292) grad_norm 1.6493 (inf) [2021-04-15 15:19:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][980/1251] eta 0:01:16 lr 0.000889 time 0.2632 (0.2815) loss 3.5230 (4.6285) grad_norm 1.1813 (inf) [2021-04-15 15:20:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][990/1251] eta 0:01:13 lr 0.000890 time 0.2981 (0.2815) loss 4.8532 (4.6295) grad_norm 2.4737 (inf) [2021-04-15 15:20:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][1000/1251] eta 0:01:10 lr 0.000890 time 0.2968 (0.2814) loss 5.5029 (4.6288) grad_norm 1.4193 (inf) [2021-04-15 15:20:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][1010/1251] eta 0:01:07 lr 0.000890 time 0.2737 (0.2813) loss 5.7565 (4.6296) grad_norm 1.1265 (inf) [2021-04-15 15:20:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][1020/1251] eta 0:01:04 lr 0.000891 time 0.2449 (0.2813) loss 5.2640 (4.6275) grad_norm 1.4371 (inf) [2021-04-15 15:20:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][1030/1251] eta 0:01:02 lr 0.000891 time 0.2742 (0.2812) loss 5.3969 (4.6275) grad_norm 1.3080 (inf) [2021-04-15 15:20:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][1040/1251] eta 0:00:59 lr 0.000892 time 0.2933 (0.2812) loss 3.6417 (4.6254) grad_norm 2.1001 (inf) [2021-04-15 15:20:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][1050/1251] eta 0:00:56 lr 0.000892 time 0.2741 (0.2811) loss 5.2049 (4.6240) grad_norm 1.1030 (inf) [2021-04-15 15:20:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][1060/1251] eta 0:00:53 lr 0.000892 time 0.2852 (0.2810) loss 5.3093 (4.6264) grad_norm 1.3958 (inf) [2021-04-15 15:20:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][1070/1251] eta 0:00:50 lr 0.000893 time 0.2685 (0.2810) loss 4.7445 (4.6259) grad_norm 1.0447 (inf) [2021-04-15 15:20:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][1080/1251] eta 0:00:48 lr 0.000893 time 0.3061 (0.2810) loss 4.7911 (4.6272) grad_norm 1.3226 (inf) [2021-04-15 15:20:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][1090/1251] eta 0:00:45 lr 0.000894 time 0.2722 (0.2809) loss 3.9253 (4.6284) grad_norm 1.8201 (inf) [2021-04-15 15:20:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][1100/1251] eta 0:00:42 lr 0.000894 time 0.2916 (0.2809) loss 3.4727 (4.6264) grad_norm 1.1925 (inf) [2021-04-15 15:20:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][1110/1251] eta 0:00:39 lr 0.000894 time 0.2934 (0.2809) loss 4.6066 (4.6279) grad_norm 1.2606 (inf) [2021-04-15 15:20:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][1120/1251] eta 0:00:36 lr 0.000895 time 0.2782 (0.2809) loss 4.3947 (4.6288) grad_norm 1.3618 (inf) [2021-04-15 15:20:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][1130/1251] eta 0:00:33 lr 0.000895 time 0.2643 (0.2808) loss 4.6730 (4.6305) grad_norm 1.4405 (inf) [2021-04-15 15:20:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][1140/1251] eta 0:00:31 lr 0.000896 time 0.2737 (0.2808) loss 5.0513 (4.6287) grad_norm 1.4026 (inf) [2021-04-15 15:20:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][1150/1251] eta 0:00:28 lr 0.000896 time 0.2682 (0.2807) loss 4.7676 (4.6297) grad_norm 1.1022 (inf) [2021-04-15 15:20:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][1160/1251] eta 0:00:25 lr 0.000896 time 0.2653 (0.2807) loss 4.9637 (4.6287) grad_norm 1.8058 (inf) [2021-04-15 15:20:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][1170/1251] eta 0:00:22 lr 0.000897 time 0.2767 (0.2806) loss 4.2177 (4.6298) grad_norm 1.5634 (inf) [2021-04-15 15:20:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][1180/1251] eta 0:00:19 lr 0.000897 time 0.2742 (0.2806) loss 4.8495 (4.6301) grad_norm 1.8363 (inf) [2021-04-15 15:20:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][1190/1251] eta 0:00:17 lr 0.000898 time 0.2660 (0.2806) loss 3.9213 (4.6302) grad_norm 1.2956 (inf) [2021-04-15 15:21:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][1200/1251] eta 0:00:14 lr 0.000898 time 0.2856 (0.2806) loss 4.8229 (4.6294) grad_norm 1.4122 (inf) [2021-04-15 15:21:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][1210/1251] eta 0:00:11 lr 0.000898 time 0.2909 (0.2806) loss 5.4627 (4.6290) grad_norm 1.9306 (inf) [2021-04-15 15:21:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][1220/1251] eta 0:00:08 lr 0.000899 time 0.2779 (0.2805) loss 4.8039 (4.6302) grad_norm 1.5012 (inf) [2021-04-15 15:21:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][1230/1251] eta 0:00:05 lr 0.000899 time 0.2662 (0.2805) loss 4.2324 (4.6306) grad_norm 1.3183 (inf) [2021-04-15 15:21:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][1240/1251] eta 0:00:03 lr 0.000900 time 0.2485 (0.2804) loss 4.6640 (4.6316) grad_norm 1.0236 (inf) [2021-04-15 15:21:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [17/300][1250/1251] eta 0:00:00 lr 0.000900 time 0.2486 (0.2801) loss 5.3372 (4.6301) grad_norm 1.4646 (inf) [2021-04-15 15:21:15 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 17 training takes 0:05:52 [2021-04-15 15:21:15 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_17.pth saving...... [2021-04-15 15:21:36 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_17.pth saved !!! [2021-04-15 15:21:37 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.202 (1.202) Loss 1.9722 (1.9722) Acc@1 57.812 (57.812) Acc@5 79.883 (79.883) [2021-04-15 15:21:39 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.161 (0.257) Loss 1.9488 (1.9293) Acc@1 58.105 (58.088) Acc@5 80.664 (81.303) [2021-04-15 15:21:41 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.091 (0.254) Loss 1.8900 (1.9274) Acc@1 58.398 (58.064) Acc@5 83.594 (81.520) [2021-04-15 15:21:43 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.106 (0.225) Loss 1.9709 (1.9327) Acc@1 57.324 (57.872) Acc@5 80.762 (81.467) [2021-04-15 15:21:45 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.110 (0.223) Loss 2.0096 (1.9422) Acc@1 55.957 (57.660) Acc@5 79.883 (81.302) [2021-04-15 15:21:47 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 57.540 Acc@5 81.250 [2021-04-15 15:21:47 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 57.5% [2021-04-15 15:21:47 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 57.54% [2021-04-15 15:21:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][0/1251] eta 1:28:12 lr 0.000900 time 4.2304 (4.2304) loss 4.3455 (4.3455) grad_norm 1.3102 (1.3102) [2021-04-15 15:21:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][10/1251] eta 0:13:07 lr 0.000900 time 0.3105 (0.6348) loss 5.1869 (4.8053) grad_norm 1.6928 (1.5158) [2021-04-15 15:21:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][20/1251] eta 0:09:32 lr 0.000901 time 0.2854 (0.4648) loss 4.0104 (4.6217) grad_norm 1.1229 (1.4573) [2021-04-15 15:21:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][30/1251] eta 0:08:15 lr 0.000901 time 0.2833 (0.4055) loss 4.6586 (4.5701) grad_norm 1.1983 (1.4135) [2021-04-15 15:22:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][40/1251] eta 0:07:34 lr 0.000902 time 0.2827 (0.3752) loss 4.6946 (4.5953) grad_norm 1.2205 (1.4099) [2021-04-15 15:22:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][50/1251] eta 0:07:08 lr 0.000902 time 0.2930 (0.3568) loss 4.6232 (4.6141) grad_norm 1.4888 (1.4109) [2021-04-15 15:22:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][60/1251] eta 0:06:49 lr 0.000902 time 0.2897 (0.3441) loss 3.7220 (4.5625) grad_norm 1.4837 (1.4040) [2021-04-15 15:22:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][70/1251] eta 0:06:36 lr 0.000903 time 0.2807 (0.3354) loss 4.7247 (4.5851) grad_norm 1.3920 (1.4142) [2021-04-15 15:22:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][80/1251] eta 0:06:23 lr 0.000903 time 0.2717 (0.3278) loss 4.2312 (4.5974) grad_norm 1.5970 (1.4213) [2021-04-15 15:22:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][90/1251] eta 0:06:14 lr 0.000904 time 0.2613 (0.3224) loss 4.5687 (4.5679) grad_norm 1.1929 (1.4134) [2021-04-15 15:22:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][100/1251] eta 0:06:06 lr 0.000904 time 0.2631 (0.3185) loss 4.0377 (4.5967) grad_norm 1.5367 (1.4084) [2021-04-15 15:22:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][110/1251] eta 0:05:59 lr 0.000904 time 0.2599 (0.3150) loss 4.4698 (4.5637) grad_norm 1.4127 (1.4096) [2021-04-15 15:22:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][120/1251] eta 0:05:52 lr 0.000905 time 0.2873 (0.3118) loss 5.0274 (4.5589) grad_norm 1.7851 (1.4176) [2021-04-15 15:22:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][130/1251] eta 0:05:46 lr 0.000905 time 0.2712 (0.3093) loss 5.2392 (4.5463) grad_norm 1.4799 (1.4283) [2021-04-15 15:22:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][140/1251] eta 0:05:43 lr 0.000906 time 0.3727 (0.3087) loss 5.2515 (4.5433) grad_norm 1.2395 (1.4178) [2021-04-15 15:22:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][150/1251] eta 0:05:37 lr 0.000906 time 0.3030 (0.3067) loss 4.6856 (4.5413) grad_norm 1.0440 (1.4112) [2021-04-15 15:22:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][160/1251] eta 0:05:32 lr 0.000906 time 0.2912 (0.3047) loss 4.4431 (4.5468) grad_norm 1.6818 (1.4039) [2021-04-15 15:22:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][170/1251] eta 0:05:27 lr 0.000907 time 0.2570 (0.3031) loss 5.0312 (4.5510) grad_norm 1.5489 (1.3980) [2021-04-15 15:22:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][180/1251] eta 0:05:23 lr 0.000907 time 0.3008 (0.3020) loss 5.1984 (4.5452) grad_norm 1.4218 (1.3977) [2021-04-15 15:22:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][190/1251] eta 0:05:18 lr 0.000908 time 0.2772 (0.3005) loss 5.1659 (4.5328) grad_norm 1.6960 (1.3972) [2021-04-15 15:22:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][200/1251] eta 0:05:14 lr 0.000908 time 0.2689 (0.2993) loss 5.3163 (4.5330) grad_norm 1.0906 (1.3975) [2021-04-15 15:22:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][210/1251] eta 0:05:10 lr 0.000908 time 0.2826 (0.2984) loss 5.2043 (4.5394) grad_norm 1.3383 (1.3949) [2021-04-15 15:22:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][220/1251] eta 0:05:06 lr 0.000909 time 0.2660 (0.2973) loss 3.6256 (4.5429) grad_norm 1.4585 (1.3901) [2021-04-15 15:22:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][230/1251] eta 0:05:02 lr 0.000909 time 0.2766 (0.2968) loss 4.7212 (4.5466) grad_norm 1.3844 (1.3876) [2021-04-15 15:22:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][240/1251] eta 0:04:59 lr 0.000910 time 0.2892 (0.2959) loss 5.1276 (4.5539) grad_norm 1.3290 (1.3810) [2021-04-15 15:23:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][250/1251] eta 0:04:55 lr 0.000910 time 0.2674 (0.2951) loss 5.0543 (4.5494) grad_norm 1.2056 (1.3784) [2021-04-15 15:23:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][260/1251] eta 0:04:52 lr 0.000910 time 0.2848 (0.2948) loss 4.7797 (4.5422) grad_norm 1.2609 (1.3766) [2021-04-15 15:23:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][270/1251] eta 0:04:48 lr 0.000911 time 0.2877 (0.2942) loss 4.5070 (4.5480) grad_norm 1.4504 (1.3780) [2021-04-15 15:23:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][280/1251] eta 0:04:45 lr 0.000911 time 0.2625 (0.2936) loss 4.0754 (4.5464) grad_norm 1.2320 (1.3758) [2021-04-15 15:23:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][290/1251] eta 0:04:41 lr 0.000912 time 0.2855 (0.2929) loss 5.0286 (4.5601) grad_norm 1.1980 (1.3742) [2021-04-15 15:23:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][300/1251] eta 0:04:38 lr 0.000912 time 0.2943 (0.2925) loss 4.2740 (4.5666) grad_norm 1.1845 (1.3730) [2021-04-15 15:23:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][310/1251] eta 0:04:34 lr 0.000912 time 0.2589 (0.2919) loss 3.6126 (4.5679) grad_norm 1.4547 (1.3702) [2021-04-15 15:23:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][320/1251] eta 0:04:31 lr 0.000913 time 0.2900 (0.2915) loss 4.0793 (4.5556) grad_norm 1.5379 (1.3696) [2021-04-15 15:23:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][330/1251] eta 0:04:28 lr 0.000913 time 0.3000 (0.2910) loss 3.5883 (4.5462) grad_norm 1.6216 (1.3733) [2021-04-15 15:23:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][340/1251] eta 0:04:24 lr 0.000914 time 0.2854 (0.2906) loss 4.8179 (4.5504) grad_norm 1.5854 (1.3741) [2021-04-15 15:23:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][350/1251] eta 0:04:21 lr 0.000914 time 0.2691 (0.2902) loss 5.0447 (4.5388) grad_norm 1.7317 (1.3760) [2021-04-15 15:23:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][360/1251] eta 0:04:18 lr 0.000914 time 0.2629 (0.2897) loss 4.4113 (4.5400) grad_norm 1.0797 (1.3746) [2021-04-15 15:23:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][370/1251] eta 0:04:14 lr 0.000915 time 0.2456 (0.2893) loss 4.9456 (4.5422) grad_norm 1.1956 (1.3731) [2021-04-15 15:23:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][380/1251] eta 0:04:11 lr 0.000915 time 0.2720 (0.2890) loss 3.3024 (4.5382) grad_norm 1.2358 (1.3696) [2021-04-15 15:23:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][390/1251] eta 0:04:08 lr 0.000916 time 0.2792 (0.2887) loss 3.7926 (4.5403) grad_norm 1.2355 (1.3700) [2021-04-15 15:23:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][400/1251] eta 0:04:05 lr 0.000916 time 0.2880 (0.2885) loss 3.7155 (4.5396) grad_norm 1.3017 (1.3715) [2021-04-15 15:23:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][410/1251] eta 0:04:02 lr 0.000916 time 0.3062 (0.2882) loss 5.2833 (4.5420) grad_norm 1.2503 (1.3698) [2021-04-15 15:23:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][420/1251] eta 0:03:59 lr 0.000917 time 0.2842 (0.2880) loss 5.2231 (4.5512) grad_norm 1.2225 (1.3716) [2021-04-15 15:23:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][430/1251] eta 0:03:56 lr 0.000917 time 0.3044 (0.2877) loss 4.9900 (4.5567) grad_norm 1.1148 (1.3684) [2021-04-15 15:23:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][440/1251] eta 0:03:53 lr 0.000918 time 0.2668 (0.2874) loss 3.7936 (4.5517) grad_norm 1.3873 (1.3673) [2021-04-15 15:23:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][450/1251] eta 0:03:50 lr 0.000918 time 0.2937 (0.2873) loss 4.1926 (4.5532) grad_norm 1.2373 (1.3686) [2021-04-15 15:23:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][460/1251] eta 0:03:47 lr 0.000918 time 0.2858 (0.2872) loss 4.6809 (4.5528) grad_norm 1.7485 (1.3672) [2021-04-15 15:24:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][470/1251] eta 0:03:44 lr 0.000919 time 0.2943 (0.2869) loss 3.7047 (4.5547) grad_norm 1.1989 (1.3664) [2021-04-15 15:24:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][480/1251] eta 0:03:41 lr 0.000919 time 0.2892 (0.2868) loss 4.9939 (4.5513) grad_norm 1.4688 (1.3662) [2021-04-15 15:24:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][490/1251] eta 0:03:38 lr 0.000920 time 0.2945 (0.2865) loss 4.9236 (4.5547) grad_norm 1.4225 (1.3682) [2021-04-15 15:24:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][500/1251] eta 0:03:35 lr 0.000920 time 0.2535 (0.2864) loss 4.8202 (4.5562) grad_norm 1.7725 (1.3682) [2021-04-15 15:24:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][510/1251] eta 0:03:32 lr 0.000920 time 0.2875 (0.2863) loss 3.4351 (4.5531) grad_norm 1.0406 (1.3659) [2021-04-15 15:24:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][520/1251] eta 0:03:29 lr 0.000921 time 0.2965 (0.2861) loss 5.2498 (4.5519) grad_norm 1.1103 (1.3644) [2021-04-15 15:24:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][530/1251] eta 0:03:26 lr 0.000921 time 0.2818 (0.2860) loss 3.7935 (4.5512) grad_norm 1.3557 (1.3632) [2021-04-15 15:24:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][540/1251] eta 0:03:23 lr 0.000922 time 0.2812 (0.2858) loss 5.2920 (4.5563) grad_norm 1.1779 (1.3642) [2021-04-15 15:24:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][550/1251] eta 0:03:20 lr 0.000922 time 0.2874 (0.2856) loss 4.5600 (4.5524) grad_norm 1.2406 (1.3639) [2021-04-15 15:24:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][560/1251] eta 0:03:17 lr 0.000922 time 0.2704 (0.2854) loss 3.0073 (4.5527) grad_norm 1.2037 (1.3624) [2021-04-15 15:24:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][570/1251] eta 0:03:14 lr 0.000923 time 0.2615 (0.2854) loss 4.3657 (4.5572) grad_norm 1.1182 (1.3595) [2021-04-15 15:24:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][580/1251] eta 0:03:11 lr 0.000923 time 0.2734 (0.2853) loss 4.0693 (4.5549) grad_norm 1.2731 (1.3570) [2021-04-15 15:24:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][590/1251] eta 0:03:08 lr 0.000924 time 0.2649 (0.2851) loss 3.9405 (4.5504) grad_norm 1.9714 (1.3595) [2021-04-15 15:24:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][600/1251] eta 0:03:05 lr 0.000924 time 0.3036 (0.2850) loss 5.0488 (4.5511) grad_norm 1.3380 (1.3593) [2021-04-15 15:24:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][610/1251] eta 0:03:02 lr 0.000924 time 0.2662 (0.2848) loss 3.6843 (4.5497) grad_norm 1.3044 (1.3603) [2021-04-15 15:24:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][620/1251] eta 0:02:59 lr 0.000925 time 0.2845 (0.2847) loss 4.5850 (4.5537) grad_norm 1.2616 (1.3596) [2021-04-15 15:24:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][630/1251] eta 0:02:56 lr 0.000925 time 0.2615 (0.2846) loss 4.1655 (4.5504) grad_norm 1.6443 (1.3577) [2021-04-15 15:24:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][640/1251] eta 0:02:53 lr 0.000926 time 0.2788 (0.2846) loss 4.3431 (4.5529) grad_norm 1.1798 (1.3567) [2021-04-15 15:24:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][650/1251] eta 0:02:50 lr 0.000926 time 0.2633 (0.2845) loss 5.2649 (4.5595) grad_norm 1.1525 (1.3555) [2021-04-15 15:24:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][660/1251] eta 0:02:48 lr 0.000926 time 0.2476 (0.2843) loss 4.5435 (4.5596) grad_norm 1.3539 (1.3551) [2021-04-15 15:24:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][670/1251] eta 0:02:45 lr 0.000927 time 0.2784 (0.2842) loss 3.5275 (4.5574) grad_norm 1.4426 (1.3576) [2021-04-15 15:25:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][680/1251] eta 0:02:42 lr 0.000927 time 0.2795 (0.2841) loss 5.4384 (4.5570) grad_norm 1.3636 (1.3567) [2021-04-15 15:25:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][690/1251] eta 0:02:39 lr 0.000928 time 0.2862 (0.2839) loss 4.8853 (4.5584) grad_norm 1.8528 (1.3580) [2021-04-15 15:25:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][700/1251] eta 0:02:36 lr 0.000928 time 0.3044 (0.2838) loss 3.9779 (4.5574) grad_norm 1.4267 (1.3589) [2021-04-15 15:25:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][710/1251] eta 0:02:33 lr 0.000928 time 0.2491 (0.2837) loss 5.1612 (4.5558) grad_norm 1.2180 (1.3585) [2021-04-15 15:25:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][720/1251] eta 0:02:30 lr 0.000929 time 0.2786 (0.2838) loss 5.8942 (4.5636) grad_norm 1.2916 (1.3562) [2021-04-15 15:25:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][730/1251] eta 0:02:27 lr 0.000929 time 0.2806 (0.2837) loss 4.9210 (4.5647) grad_norm 1.0837 (1.3561) [2021-04-15 15:25:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][740/1251] eta 0:02:24 lr 0.000930 time 0.2536 (0.2836) loss 5.5581 (4.5603) grad_norm 1.3958 (1.3563) [2021-04-15 15:25:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][750/1251] eta 0:02:22 lr 0.000930 time 0.2766 (0.2835) loss 3.5397 (4.5590) grad_norm 1.1126 (1.3540) [2021-04-15 15:25:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][760/1251] eta 0:02:19 lr 0.000930 time 0.3005 (0.2835) loss 4.1225 (4.5614) grad_norm 1.3293 (1.3526) [2021-04-15 15:25:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][770/1251] eta 0:02:16 lr 0.000931 time 0.2719 (0.2833) loss 4.9814 (4.5622) grad_norm 1.4246 (1.3518) [2021-04-15 15:25:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][780/1251] eta 0:02:13 lr 0.000931 time 0.2622 (0.2833) loss 4.7716 (4.5614) grad_norm 1.7867 (1.3534) [2021-04-15 15:25:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][790/1251] eta 0:02:10 lr 0.000932 time 0.2757 (0.2833) loss 5.2572 (4.5621) grad_norm 1.3005 (1.3512) [2021-04-15 15:25:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][800/1251] eta 0:02:07 lr 0.000932 time 0.2767 (0.2832) loss 4.9161 (4.5628) grad_norm 1.3381 (1.3510) [2021-04-15 15:25:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][810/1251] eta 0:02:04 lr 0.000932 time 0.2629 (0.2831) loss 5.2389 (4.5647) grad_norm 1.3702 (1.3515) [2021-04-15 15:25:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][820/1251] eta 0:02:01 lr 0.000933 time 0.2642 (0.2830) loss 4.7163 (4.5645) grad_norm 1.7683 (1.3538) [2021-04-15 15:25:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][830/1251] eta 0:01:59 lr 0.000933 time 0.3080 (0.2830) loss 4.1368 (4.5656) grad_norm 1.4367 (1.3534) [2021-04-15 15:25:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][840/1251] eta 0:01:56 lr 0.000934 time 0.2937 (0.2829) loss 4.5365 (4.5678) grad_norm 1.6329 (1.3549) [2021-04-15 15:25:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][850/1251] eta 0:01:53 lr 0.000934 time 0.2831 (0.2829) loss 5.0225 (4.5693) grad_norm 1.9729 (1.3565) [2021-04-15 15:25:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][860/1251] eta 0:01:50 lr 0.000934 time 0.2930 (0.2829) loss 5.4767 (4.5673) grad_norm 1.1736 (1.3571) [2021-04-15 15:25:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][870/1251] eta 0:01:47 lr 0.000935 time 0.2718 (0.2827) loss 4.7104 (4.5693) grad_norm 1.0356 (1.3559) [2021-04-15 15:25:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][880/1251] eta 0:01:44 lr 0.000935 time 0.2838 (0.2827) loss 5.1392 (4.5690) grad_norm 1.3927 (1.3549) [2021-04-15 15:25:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][890/1251] eta 0:01:42 lr 0.000936 time 0.2681 (0.2827) loss 4.7679 (4.5708) grad_norm 1.2451 (1.3541) [2021-04-15 15:26:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][900/1251] eta 0:01:39 lr 0.000936 time 0.2651 (0.2827) loss 5.1883 (4.5719) grad_norm 1.2501 (1.3532) [2021-04-15 15:26:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][910/1251] eta 0:01:36 lr 0.000936 time 0.2924 (0.2826) loss 3.0084 (4.5729) grad_norm 1.1564 (1.3530) [2021-04-15 15:26:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][920/1251] eta 0:01:33 lr 0.000937 time 0.2939 (0.2825) loss 5.0529 (4.5745) grad_norm 1.3614 (1.3516) [2021-04-15 15:26:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][930/1251] eta 0:01:30 lr 0.000937 time 0.3144 (0.2825) loss 5.4534 (4.5760) grad_norm 1.4139 (1.3507) [2021-04-15 15:26:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][940/1251] eta 0:01:27 lr 0.000938 time 0.2798 (0.2825) loss 4.5045 (4.5786) grad_norm 1.1346 (1.3495) [2021-04-15 15:26:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][950/1251] eta 0:01:24 lr 0.000938 time 0.2978 (0.2824) loss 4.7230 (4.5821) grad_norm 1.5913 (inf) [2021-04-15 15:26:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][960/1251] eta 0:01:22 lr 0.000938 time 0.3009 (0.2824) loss 5.0260 (4.5801) grad_norm 1.3067 (inf) [2021-04-15 15:26:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][970/1251] eta 0:01:19 lr 0.000939 time 0.2859 (0.2823) loss 3.8257 (4.5815) grad_norm 1.2979 (inf) [2021-04-15 15:26:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][980/1251] eta 0:01:16 lr 0.000939 time 0.2551 (0.2822) loss 4.5965 (4.5800) grad_norm 1.1329 (inf) [2021-04-15 15:26:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][990/1251] eta 0:01:13 lr 0.000940 time 0.2671 (0.2822) loss 3.9610 (4.5813) grad_norm 1.3370 (inf) [2021-04-15 15:26:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][1000/1251] eta 0:01:10 lr 0.000940 time 0.2853 (0.2822) loss 4.7141 (4.5802) grad_norm 1.1426 (inf) [2021-04-15 15:26:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][1010/1251] eta 0:01:07 lr 0.000940 time 0.2841 (0.2821) loss 5.3319 (4.5809) grad_norm 1.2901 (inf) [2021-04-15 15:26:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][1020/1251] eta 0:01:05 lr 0.000941 time 0.2673 (0.2821) loss 4.1558 (4.5788) grad_norm 1.1352 (inf) [2021-04-15 15:26:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][1030/1251] eta 0:01:02 lr 0.000941 time 0.2995 (0.2820) loss 4.9611 (4.5778) grad_norm 1.2794 (inf) [2021-04-15 15:26:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][1040/1251] eta 0:00:59 lr 0.000942 time 0.2653 (0.2820) loss 4.9957 (4.5780) grad_norm 1.2475 (inf) [2021-04-15 15:26:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][1050/1251] eta 0:00:56 lr 0.000942 time 0.2731 (0.2819) loss 4.8564 (4.5794) grad_norm 1.1266 (inf) [2021-04-15 15:26:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][1060/1251] eta 0:00:53 lr 0.000942 time 0.2740 (0.2819) loss 4.5461 (4.5797) grad_norm 1.2743 (inf) [2021-04-15 15:26:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][1070/1251] eta 0:00:51 lr 0.000943 time 0.2813 (0.2818) loss 3.8997 (4.5767) grad_norm 1.0288 (inf) [2021-04-15 15:26:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][1080/1251] eta 0:00:48 lr 0.000943 time 0.2572 (0.2817) loss 4.8958 (4.5780) grad_norm 1.3660 (inf) [2021-04-15 15:26:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][1090/1251] eta 0:00:45 lr 0.000944 time 0.2667 (0.2817) loss 4.9341 (4.5785) grad_norm 1.5025 (inf) [2021-04-15 15:26:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][1100/1251] eta 0:00:42 lr 0.000944 time 0.2749 (0.2817) loss 5.5485 (4.5762) grad_norm 1.2167 (inf) [2021-04-15 15:27:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][1110/1251] eta 0:00:39 lr 0.000944 time 0.2495 (0.2816) loss 4.5869 (4.5736) grad_norm 1.2073 (inf) [2021-04-15 15:27:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][1120/1251] eta 0:00:36 lr 0.000945 time 0.2822 (0.2816) loss 5.2821 (4.5721) grad_norm 1.1199 (inf) [2021-04-15 15:27:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][1130/1251] eta 0:00:34 lr 0.000945 time 0.2710 (0.2816) loss 3.4531 (4.5706) grad_norm 1.5423 (inf) [2021-04-15 15:27:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][1140/1251] eta 0:00:31 lr 0.000946 time 0.2780 (0.2815) loss 3.8837 (4.5709) grad_norm 1.1170 (inf) [2021-04-15 15:27:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][1150/1251] eta 0:00:28 lr 0.000946 time 0.2864 (0.2815) loss 4.1121 (4.5691) grad_norm 1.2293 (inf) [2021-04-15 15:27:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][1160/1251] eta 0:00:25 lr 0.000946 time 0.2546 (0.2815) loss 4.5358 (4.5687) grad_norm 1.3006 (inf) [2021-04-15 15:27:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][1170/1251] eta 0:00:22 lr 0.000947 time 0.2888 (0.2815) loss 3.7215 (4.5700) grad_norm 1.0575 (inf) [2021-04-15 15:27:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][1180/1251] eta 0:00:19 lr 0.000947 time 0.2889 (0.2814) loss 4.9517 (4.5711) grad_norm 1.5502 (inf) [2021-04-15 15:27:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][1190/1251] eta 0:00:17 lr 0.000948 time 0.2846 (0.2813) loss 4.8094 (4.5734) grad_norm 1.1837 (inf) [2021-04-15 15:27:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][1200/1251] eta 0:00:14 lr 0.000948 time 0.2876 (0.2813) loss 4.9756 (4.5744) grad_norm 1.2924 (inf) [2021-04-15 15:27:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][1210/1251] eta 0:00:11 lr 0.000948 time 0.2885 (0.2812) loss 4.3483 (4.5719) grad_norm 1.5121 (inf) [2021-04-15 15:27:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][1220/1251] eta 0:00:08 lr 0.000949 time 0.2904 (0.2812) loss 4.4587 (4.5722) grad_norm 1.4232 (inf) [2021-04-15 15:27:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][1230/1251] eta 0:00:05 lr 0.000949 time 0.2547 (0.2812) loss 5.1937 (4.5745) grad_norm 1.1734 (inf) [2021-04-15 15:27:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][1240/1251] eta 0:00:03 lr 0.000950 time 0.2487 (0.2811) loss 3.7162 (4.5729) grad_norm 1.2059 (inf) [2021-04-15 15:27:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [18/300][1250/1251] eta 0:00:00 lr 0.000950 time 0.2472 (0.2809) loss 5.3795 (4.5757) grad_norm 1.5747 (inf) [2021-04-15 15:27:40 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 18 training takes 0:05:53 [2021-04-15 15:27:40 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_18.pth saving...... [2021-04-15 15:28:00 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_18.pth saved !!! [2021-04-15 15:28:01 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.214 (1.214) Loss 1.7924 (1.7924) Acc@1 57.715 (57.715) Acc@5 83.594 (83.594) [2021-04-15 15:28:03 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.116 (0.229) Loss 1.8996 (1.8707) Acc@1 58.203 (57.830) Acc@5 82.031 (82.102) [2021-04-15 15:28:05 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.169 (0.217) Loss 1.8561 (1.8494) Acc@1 58.008 (58.617) Acc@5 81.055 (82.078) [2021-04-15 15:28:07 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.108 (0.232) Loss 1.9096 (1.8534) Acc@1 57.715 (58.335) Acc@5 82.520 (82.022) [2021-04-15 15:28:10 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.332 (0.232) Loss 1.9108 (1.8521) Acc@1 58.008 (58.244) Acc@5 82.422 (82.086) [2021-04-15 15:28:12 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 58.204 Acc@5 82.068 [2021-04-15 15:28:12 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 58.2% [2021-04-15 15:28:12 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 58.20% [2021-04-15 15:28:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][0/1251] eta 1:22:46 lr 0.000950 time 3.9697 (3.9697) loss 4.9481 (4.9481) grad_norm 1.3972 (1.3972) [2021-04-15 15:28:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][10/1251] eta 0:12:50 lr 0.000950 time 0.2782 (0.6212) loss 5.3229 (4.5565) grad_norm 1.1656 (1.3328) [2021-04-15 15:28:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][20/1251] eta 0:09:25 lr 0.000951 time 0.2804 (0.4593) loss 4.8053 (4.5137) grad_norm 1.2432 (1.2729) [2021-04-15 15:28:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][30/1251] eta 0:08:10 lr 0.000951 time 0.2911 (0.4020) loss 3.8840 (4.4127) grad_norm 1.5862 (1.2809) [2021-04-15 15:28:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][40/1251] eta 0:07:30 lr 0.000952 time 0.2692 (0.3721) loss 4.6547 (4.5088) grad_norm 1.2688 (1.2950) [2021-04-15 15:28:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][50/1251] eta 0:07:05 lr 0.000952 time 0.2600 (0.3545) loss 4.9455 (4.5106) grad_norm 1.5820 (1.3100) [2021-04-15 15:28:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][60/1251] eta 0:06:47 lr 0.000952 time 0.3152 (0.3423) loss 4.1530 (4.5266) grad_norm 1.4929 (1.3054) [2021-04-15 15:28:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][70/1251] eta 0:06:34 lr 0.000953 time 0.2972 (0.3337) loss 4.8656 (4.5077) grad_norm 1.0820 (1.3234) [2021-04-15 15:28:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][80/1251] eta 0:06:22 lr 0.000953 time 0.2637 (0.3268) loss 3.0571 (4.4919) grad_norm 1.2864 (1.3139) [2021-04-15 15:28:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][90/1251] eta 0:06:13 lr 0.000954 time 0.2872 (0.3218) loss 4.5620 (4.4761) grad_norm 1.4060 (1.3148) [2021-04-15 15:28:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][100/1251] eta 0:06:05 lr 0.000954 time 0.2770 (0.3175) loss 5.3566 (4.5059) grad_norm 2.0050 (1.3208) [2021-04-15 15:28:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][110/1251] eta 0:05:58 lr 0.000954 time 0.2686 (0.3141) loss 4.5670 (4.5015) grad_norm 1.2483 (1.3189) [2021-04-15 15:28:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][120/1251] eta 0:05:52 lr 0.000955 time 0.2994 (0.3114) loss 5.3235 (4.4947) grad_norm 1.0305 (1.3147) [2021-04-15 15:28:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][130/1251] eta 0:05:46 lr 0.000955 time 0.2777 (0.3087) loss 4.7438 (4.5116) grad_norm 1.1323 (1.3108) [2021-04-15 15:28:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][140/1251] eta 0:05:40 lr 0.000956 time 0.2520 (0.3064) loss 4.8668 (4.5310) grad_norm 1.2044 (1.3103) [2021-04-15 15:28:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][150/1251] eta 0:05:35 lr 0.000956 time 0.2816 (0.3046) loss 4.3191 (4.5271) grad_norm 1.5464 (1.3068) [2021-04-15 15:29:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][160/1251] eta 0:05:30 lr 0.000956 time 0.2451 (0.3029) loss 4.7020 (4.5450) grad_norm 1.0656 (1.3059) [2021-04-15 15:29:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][170/1251] eta 0:05:25 lr 0.000957 time 0.2643 (0.3014) loss 4.3343 (4.5305) grad_norm 1.1121 (1.3032) [2021-04-15 15:29:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][180/1251] eta 0:05:21 lr 0.000957 time 0.2833 (0.3003) loss 5.3030 (4.5421) grad_norm 1.2701 (1.3045) [2021-04-15 15:29:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][190/1251] eta 0:05:17 lr 0.000958 time 0.2912 (0.2992) loss 3.9197 (4.5388) grad_norm 2.1458 (1.3149) [2021-04-15 15:29:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][200/1251] eta 0:05:13 lr 0.000958 time 0.2746 (0.2980) loss 5.0647 (4.5490) grad_norm 1.1196 (1.3158) [2021-04-15 15:29:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][210/1251] eta 0:05:09 lr 0.000958 time 0.2902 (0.2971) loss 5.3714 (4.5628) grad_norm 1.0454 (1.3114) [2021-04-15 15:29:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][220/1251] eta 0:05:05 lr 0.000959 time 0.3105 (0.2965) loss 3.7216 (4.5644) grad_norm 1.3788 (1.3102) [2021-04-15 15:29:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][230/1251] eta 0:05:02 lr 0.000959 time 0.2850 (0.2958) loss 4.7355 (4.5551) grad_norm 1.5194 (1.3104) [2021-04-15 15:29:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][240/1251] eta 0:04:58 lr 0.000960 time 0.2951 (0.2952) loss 4.7581 (4.5668) grad_norm 1.0665 (1.3117) [2021-04-15 15:29:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][250/1251] eta 0:04:54 lr 0.000960 time 0.2677 (0.2945) loss 5.4983 (4.5719) grad_norm 1.5579 (1.3110) [2021-04-15 15:29:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][260/1251] eta 0:04:51 lr 0.000960 time 0.2621 (0.2938) loss 5.3309 (4.5722) grad_norm 1.4832 (1.3156) [2021-04-15 15:29:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][270/1251] eta 0:04:47 lr 0.000961 time 0.2725 (0.2934) loss 4.6711 (4.5710) grad_norm 1.2227 (1.3133) [2021-04-15 15:29:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][280/1251] eta 0:04:44 lr 0.000961 time 0.2805 (0.2930) loss 5.4161 (4.5782) grad_norm 1.3150 (1.3112) [2021-04-15 15:29:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][290/1251] eta 0:04:41 lr 0.000962 time 0.2827 (0.2927) loss 3.3011 (4.5755) grad_norm 1.1142 (1.3114) [2021-04-15 15:29:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][300/1251] eta 0:04:37 lr 0.000962 time 0.2936 (0.2923) loss 4.1458 (4.5793) grad_norm 1.6190 (1.3091) [2021-04-15 15:29:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][310/1251] eta 0:04:34 lr 0.000962 time 0.2688 (0.2918) loss 4.2781 (4.5726) grad_norm 1.3077 (1.3104) [2021-04-15 15:29:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][320/1251] eta 0:04:31 lr 0.000963 time 0.2890 (0.2914) loss 4.5155 (4.5721) grad_norm 1.4341 (1.3099) [2021-04-15 15:29:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][330/1251] eta 0:04:28 lr 0.000963 time 0.2676 (0.2912) loss 4.0399 (4.5638) grad_norm 1.3614 (1.3061) [2021-04-15 15:29:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][340/1251] eta 0:04:24 lr 0.000964 time 0.2935 (0.2907) loss 4.4522 (4.5694) grad_norm 1.2248 (1.3013) [2021-04-15 15:29:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][350/1251] eta 0:04:21 lr 0.000964 time 0.2573 (0.2903) loss 4.8595 (4.5730) grad_norm 1.3249 (1.2991) [2021-04-15 15:29:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][360/1251] eta 0:04:18 lr 0.000964 time 0.2639 (0.2899) loss 4.2073 (4.5667) grad_norm 1.0592 (1.2993) [2021-04-15 15:29:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][370/1251] eta 0:04:15 lr 0.000965 time 0.2505 (0.2897) loss 4.4428 (4.5620) grad_norm 1.5887 (1.2993) [2021-04-15 15:30:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][380/1251] eta 0:04:11 lr 0.000965 time 0.2648 (0.2893) loss 5.0196 (4.5577) grad_norm 1.4830 (1.2982) [2021-04-15 15:30:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][390/1251] eta 0:04:08 lr 0.000966 time 0.2891 (0.2889) loss 4.5275 (4.5573) grad_norm 1.2390 (1.2997) [2021-04-15 15:30:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][400/1251] eta 0:04:05 lr 0.000966 time 0.2475 (0.2886) loss 4.7658 (4.5637) grad_norm 1.2051 (1.2989) [2021-04-15 15:30:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][410/1251] eta 0:04:02 lr 0.000966 time 0.2638 (0.2883) loss 4.9548 (4.5664) grad_norm 1.0313 (1.3015) [2021-04-15 15:30:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][420/1251] eta 0:03:59 lr 0.000967 time 0.2718 (0.2880) loss 4.0839 (4.5617) grad_norm 1.1397 (1.3021) [2021-04-15 15:30:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][430/1251] eta 0:03:56 lr 0.000967 time 0.2542 (0.2878) loss 4.6631 (4.5589) grad_norm 1.1734 (1.3034) [2021-04-15 15:30:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][440/1251] eta 0:03:53 lr 0.000968 time 0.2625 (0.2876) loss 5.1097 (4.5590) grad_norm 0.9898 (1.3020) [2021-04-15 15:30:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][450/1251] eta 0:03:50 lr 0.000968 time 0.2710 (0.2873) loss 3.7484 (4.5563) grad_norm 1.5346 (1.3005) [2021-04-15 15:30:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][460/1251] eta 0:03:46 lr 0.000968 time 0.2435 (0.2870) loss 4.7105 (4.5612) grad_norm 1.1099 (1.2985) [2021-04-15 15:30:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][470/1251] eta 0:03:43 lr 0.000969 time 0.2601 (0.2867) loss 4.8417 (4.5625) grad_norm 1.1897 (1.2963) [2021-04-15 15:30:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][480/1251] eta 0:03:40 lr 0.000969 time 0.2685 (0.2865) loss 3.7239 (4.5576) grad_norm 1.3025 (1.2967) [2021-04-15 15:30:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][490/1251] eta 0:03:38 lr 0.000970 time 0.3209 (0.2865) loss 5.1175 (4.5594) grad_norm 1.2549 (1.2961) [2021-04-15 15:30:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][500/1251] eta 0:03:34 lr 0.000970 time 0.2682 (0.2863) loss 5.4669 (4.5592) grad_norm 1.1796 (1.2968) [2021-04-15 15:30:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][510/1251] eta 0:03:31 lr 0.000970 time 0.2825 (0.2861) loss 4.1464 (4.5543) grad_norm 1.7535 (1.2953) [2021-04-15 15:30:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][520/1251] eta 0:03:29 lr 0.000971 time 0.3148 (0.2861) loss 5.5029 (4.5577) grad_norm 1.1164 (1.2938) [2021-04-15 15:30:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][530/1251] eta 0:03:26 lr 0.000971 time 0.2685 (0.2860) loss 3.6349 (4.5538) grad_norm 1.0863 (1.2932) [2021-04-15 15:30:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][540/1251] eta 0:03:23 lr 0.000972 time 0.2783 (0.2858) loss 3.5030 (4.5512) grad_norm 0.9447 (1.2928) [2021-04-15 15:30:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][550/1251] eta 0:03:20 lr 0.000972 time 0.2589 (0.2856) loss 4.3341 (4.5489) grad_norm 1.2523 (1.2938) [2021-04-15 15:30:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][560/1251] eta 0:03:17 lr 0.000972 time 0.2616 (0.2856) loss 4.2467 (4.5507) grad_norm 0.9501 (1.2927) [2021-04-15 15:30:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][570/1251] eta 0:03:14 lr 0.000973 time 0.2702 (0.2857) loss 3.6193 (4.5474) grad_norm 1.3165 (1.2930) [2021-04-15 15:30:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][580/1251] eta 0:03:11 lr 0.000973 time 0.3003 (0.2858) loss 4.4845 (4.5491) grad_norm 1.3275 (1.2941) [2021-04-15 15:31:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][590/1251] eta 0:03:08 lr 0.000974 time 0.2605 (0.2857) loss 5.3666 (4.5548) grad_norm 1.3363 (1.2953) [2021-04-15 15:31:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][600/1251] eta 0:03:05 lr 0.000974 time 0.2821 (0.2856) loss 3.7556 (4.5488) grad_norm 1.2557 (1.2950) [2021-04-15 15:31:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][610/1251] eta 0:03:03 lr 0.000974 time 0.2834 (0.2856) loss 5.1309 (4.5497) grad_norm 1.5074 (1.2957) [2021-04-15 15:31:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][620/1251] eta 0:03:00 lr 0.000975 time 0.2894 (0.2857) loss 5.0716 (4.5493) grad_norm 1.2795 (1.2970) [2021-04-15 15:31:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][630/1251] eta 0:02:57 lr 0.000975 time 0.2762 (0.2856) loss 4.8077 (4.5543) grad_norm 1.1509 (1.2982) [2021-04-15 15:31:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][640/1251] eta 0:02:54 lr 0.000976 time 0.2749 (0.2854) loss 3.2642 (4.5530) grad_norm 1.1215 (1.2978) [2021-04-15 15:31:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][650/1251] eta 0:02:51 lr 0.000976 time 0.2757 (0.2852) loss 4.7705 (4.5526) grad_norm 1.5022 (1.2975) [2021-04-15 15:31:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][660/1251] eta 0:02:48 lr 0.000976 time 0.2612 (0.2851) loss 4.4444 (4.5500) grad_norm 1.1559 (1.2963) [2021-04-15 15:31:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][670/1251] eta 0:02:45 lr 0.000977 time 0.3168 (0.2851) loss 5.6412 (4.5493) grad_norm 1.5649 (1.2981) [2021-04-15 15:31:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][680/1251] eta 0:02:42 lr 0.000977 time 0.2772 (0.2850) loss 4.6837 (4.5507) grad_norm 1.4437 (1.2991) [2021-04-15 15:31:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][690/1251] eta 0:02:39 lr 0.000978 time 0.2553 (0.2848) loss 2.8898 (4.5513) grad_norm 1.1392 (1.2988) [2021-04-15 15:31:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][700/1251] eta 0:02:36 lr 0.000978 time 0.2963 (0.2848) loss 4.0625 (4.5529) grad_norm 1.3166 (1.2976) [2021-04-15 15:31:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][710/1251] eta 0:02:34 lr 0.000978 time 0.2689 (0.2847) loss 5.3743 (4.5514) grad_norm 1.2213 (1.2983) [2021-04-15 15:31:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][720/1251] eta 0:02:31 lr 0.000979 time 0.2804 (0.2846) loss 3.7768 (4.5488) grad_norm 1.3265 (1.2994) [2021-04-15 15:31:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][730/1251] eta 0:02:28 lr 0.000979 time 0.2788 (0.2845) loss 4.7967 (4.5457) grad_norm 1.0621 (1.2981) [2021-04-15 15:31:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][740/1251] eta 0:02:25 lr 0.000980 time 0.2706 (0.2844) loss 4.4194 (4.5450) grad_norm 1.1797 (1.2969) [2021-04-15 15:31:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][750/1251] eta 0:02:22 lr 0.000980 time 0.2736 (0.2843) loss 4.2732 (4.5460) grad_norm 1.2030 (1.2948) [2021-04-15 15:31:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][760/1251] eta 0:02:19 lr 0.000980 time 0.2642 (0.2843) loss 3.8103 (4.5491) grad_norm 1.1379 (1.2947) [2021-04-15 15:31:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][770/1251] eta 0:02:16 lr 0.000981 time 0.2651 (0.2842) loss 4.0026 (4.5494) grad_norm 1.2859 (1.2947) [2021-04-15 15:31:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][780/1251] eta 0:02:13 lr 0.000981 time 0.2531 (0.2841) loss 4.7306 (4.5489) grad_norm 1.1537 (1.2926) [2021-04-15 15:31:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][790/1251] eta 0:02:10 lr 0.000982 time 0.2616 (0.2840) loss 5.6026 (4.5472) grad_norm 1.5557 (1.2930) [2021-04-15 15:31:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][800/1251] eta 0:02:08 lr 0.000982 time 0.2905 (0.2839) loss 5.2709 (4.5475) grad_norm 1.2060 (1.2925) [2021-04-15 15:32:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][810/1251] eta 0:02:05 lr 0.000982 time 0.2713 (0.2838) loss 4.0586 (4.5475) grad_norm 1.3377 (1.2921) [2021-04-15 15:32:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][820/1251] eta 0:02:02 lr 0.000983 time 0.2950 (0.2837) loss 4.4310 (4.5499) grad_norm 1.8878 (1.2928) [2021-04-15 15:32:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][830/1251] eta 0:01:59 lr 0.000983 time 0.3168 (0.2837) loss 5.3972 (4.5526) grad_norm 1.1380 (1.2922) [2021-04-15 15:32:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][840/1251] eta 0:01:56 lr 0.000984 time 0.2862 (0.2836) loss 5.2676 (4.5517) grad_norm 1.3592 (1.2922) [2021-04-15 15:32:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][850/1251] eta 0:01:53 lr 0.000984 time 0.2909 (0.2836) loss 4.9293 (4.5504) grad_norm 1.1002 (1.2923) [2021-04-15 15:32:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][860/1251] eta 0:01:50 lr 0.000984 time 0.2849 (0.2835) loss 5.2890 (4.5537) grad_norm 1.1993 (1.2919) [2021-04-15 15:32:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][870/1251] eta 0:01:47 lr 0.000985 time 0.2713 (0.2834) loss 5.1484 (4.5593) grad_norm 1.3703 (1.2913) [2021-04-15 15:32:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][880/1251] eta 0:01:45 lr 0.000985 time 0.2668 (0.2834) loss 5.1619 (4.5612) grad_norm 1.4736 (1.2912) [2021-04-15 15:32:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][890/1251] eta 0:01:42 lr 0.000986 time 0.2929 (0.2833) loss 4.2681 (4.5586) grad_norm 1.3927 (1.2924) [2021-04-15 15:32:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][900/1251] eta 0:01:39 lr 0.000986 time 0.2814 (0.2832) loss 4.7838 (4.5558) grad_norm 1.2274 (1.2931) [2021-04-15 15:32:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][910/1251] eta 0:01:36 lr 0.000986 time 0.2779 (0.2832) loss 3.8824 (4.5566) grad_norm 1.4683 (1.2930) [2021-04-15 15:32:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][920/1251] eta 0:01:33 lr 0.000987 time 0.2985 (0.2831) loss 4.9359 (4.5573) grad_norm 1.3916 (1.2938) [2021-04-15 15:32:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][930/1251] eta 0:01:30 lr 0.000987 time 0.2738 (0.2830) loss 4.6919 (4.5582) grad_norm 1.0327 (1.2930) [2021-04-15 15:32:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][940/1251] eta 0:01:27 lr 0.000988 time 0.2684 (0.2829) loss 4.1285 (4.5600) grad_norm 1.2855 (1.2930) [2021-04-15 15:32:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][950/1251] eta 0:01:25 lr 0.000988 time 0.2684 (0.2829) loss 4.3672 (4.5593) grad_norm 1.1936 (1.2933) [2021-04-15 15:32:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][960/1251] eta 0:01:22 lr 0.000988 time 0.2773 (0.2829) loss 5.2397 (4.5571) grad_norm 1.2645 (1.2940) [2021-04-15 15:32:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][970/1251] eta 0:01:19 lr 0.000989 time 0.2767 (0.2828) loss 5.1554 (4.5601) grad_norm 1.1410 (1.2930) [2021-04-15 15:32:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][980/1251] eta 0:01:16 lr 0.000989 time 0.2773 (0.2827) loss 4.0613 (4.5605) grad_norm 1.2720 (1.2927) [2021-04-15 15:32:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][990/1251] eta 0:01:13 lr 0.000990 time 0.2914 (0.2827) loss 4.5355 (4.5602) grad_norm 1.1097 (1.2927) [2021-04-15 15:32:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][1000/1251] eta 0:01:10 lr 0.000990 time 0.2677 (0.2827) loss 4.8930 (4.5616) grad_norm 1.3512 (1.2930) [2021-04-15 15:32:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][1010/1251] eta 0:01:08 lr 0.000990 time 0.2659 (0.2826) loss 4.8619 (4.5633) grad_norm 1.0783 (1.2921) [2021-04-15 15:33:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][1020/1251] eta 0:01:05 lr 0.000991 time 0.2689 (0.2826) loss 4.2213 (4.5623) grad_norm 1.3419 (1.2922) [2021-04-15 15:33:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][1030/1251] eta 0:01:02 lr 0.000991 time 0.2531 (0.2825) loss 4.6027 (4.5586) grad_norm 1.5621 (1.2917) [2021-04-15 15:33:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][1040/1251] eta 0:00:59 lr 0.000992 time 0.2799 (0.2825) loss 4.8773 (4.5579) grad_norm 1.7769 (1.2923) [2021-04-15 15:33:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][1050/1251] eta 0:00:56 lr 0.000992 time 0.2969 (0.2825) loss 4.7606 (4.5582) grad_norm 1.2512 (1.2925) [2021-04-15 15:33:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][1060/1251] eta 0:00:53 lr 0.000992 time 0.2622 (0.2824) loss 4.9368 (4.5582) grad_norm 1.3937 (1.2927) [2021-04-15 15:33:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][1070/1251] eta 0:00:51 lr 0.000993 time 0.2722 (0.2823) loss 5.2812 (4.5590) grad_norm 1.1481 (1.2923) [2021-04-15 15:33:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][1080/1251] eta 0:00:48 lr 0.000993 time 0.2696 (0.2823) loss 3.6499 (4.5593) grad_norm 1.1570 (1.2927) [2021-04-15 15:33:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][1090/1251] eta 0:00:45 lr 0.000994 time 0.2916 (0.2824) loss 3.9152 (4.5592) grad_norm 1.4902 (1.2924) [2021-04-15 15:33:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][1100/1251] eta 0:00:42 lr 0.000994 time 0.2831 (0.2824) loss 3.2844 (4.5577) grad_norm 1.1346 (1.2936) [2021-04-15 15:33:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][1110/1251] eta 0:00:39 lr 0.000994 time 0.2579 (0.2823) loss 4.9220 (4.5547) grad_norm 1.1985 (1.2946) [2021-04-15 15:33:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][1120/1251] eta 0:00:36 lr 0.000995 time 0.2624 (0.2823) loss 4.4456 (4.5544) grad_norm 1.3034 (1.2951) [2021-04-15 15:33:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][1130/1251] eta 0:00:34 lr 0.000995 time 0.2679 (0.2822) loss 3.9584 (4.5562) grad_norm 1.1824 (1.2947) [2021-04-15 15:33:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][1140/1251] eta 0:00:31 lr 0.000996 time 0.2711 (0.2822) loss 3.8754 (4.5524) grad_norm 1.5603 (1.2946) [2021-04-15 15:33:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][1150/1251] eta 0:00:28 lr 0.000996 time 0.2827 (0.2822) loss 4.8391 (4.5533) grad_norm 1.6904 (1.2952) [2021-04-15 15:33:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][1160/1251] eta 0:00:25 lr 0.000996 time 0.2659 (0.2823) loss 4.5002 (4.5520) grad_norm 1.3868 (1.2950) [2021-04-15 15:33:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][1170/1251] eta 0:00:22 lr 0.000997 time 0.2720 (0.2823) loss 4.9405 (4.5541) grad_norm 1.1456 (1.2955) [2021-04-15 15:33:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][1180/1251] eta 0:00:20 lr 0.000997 time 0.2646 (0.2823) loss 5.4544 (4.5535) grad_norm 1.1867 (1.2951) [2021-04-15 15:33:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][1190/1251] eta 0:00:17 lr 0.000998 time 0.3147 (0.2823) loss 4.3058 (4.5542) grad_norm 1.1449 (1.2941) [2021-04-15 15:33:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][1200/1251] eta 0:00:14 lr 0.000998 time 0.2621 (0.2822) loss 4.0664 (4.5551) grad_norm 1.1746 (1.2941) [2021-04-15 15:33:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][1210/1251] eta 0:00:11 lr 0.000998 time 0.2787 (0.2822) loss 3.4907 (4.5506) grad_norm 1.4678 (1.2945) [2021-04-15 15:33:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][1220/1251] eta 0:00:08 lr 0.000999 time 0.2853 (0.2822) loss 4.5739 (4.5514) grad_norm 1.2122 (1.2937) [2021-04-15 15:33:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][1230/1251] eta 0:00:05 lr 0.000999 time 0.2736 (0.2821) loss 4.4647 (4.5500) grad_norm 1.0659 (1.2948) [2021-04-15 15:34:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][1240/1251] eta 0:00:03 lr 0.001000 time 0.2495 (0.2820) loss 3.8271 (4.5517) grad_norm 1.4746 (1.2948) [2021-04-15 15:34:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [19/300][1250/1251] eta 0:00:00 lr 0.001000 time 0.2495 (0.2818) loss 3.7071 (4.5538) grad_norm 1.1168 (1.2951) [2021-04-15 15:34:06 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 19 training takes 0:05:54 [2021-04-15 15:34:06 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_19.pth saving...... [2021-04-15 15:34:32 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_19.pth saved !!! [2021-04-15 15:34:33 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.103 (1.103) Loss 1.8214 (1.8214) Acc@1 60.059 (60.059) Acc@5 83.105 (83.105) [2021-04-15 15:34:34 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.141 (0.218) Loss 1.7596 (1.8395) Acc@1 61.133 (59.197) Acc@5 83.398 (82.635) [2021-04-15 15:34:37 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.122 (0.241) Loss 1.8773 (1.8508) Acc@1 57.422 (58.877) Acc@5 81.445 (82.399) [2021-04-15 15:34:39 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.115 (0.241) Loss 1.9542 (1.8599) Acc@1 57.227 (58.562) Acc@5 80.566 (82.343) [2021-04-15 15:34:40 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.218) Loss 1.8555 (1.8580) Acc@1 57.324 (58.699) Acc@5 82.520 (82.358) [2021-04-15 15:34:42 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 58.754 Acc@5 82.538 [2021-04-15 15:34:42 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 58.8% [2021-04-15 15:34:42 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 58.75% [2021-04-15 15:34:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][0/1251] eta 1:37:01 lr 0.000989 time 4.6531 (4.6531) loss 4.6401 (4.6401) grad_norm 1.0930 (1.0930) [2021-04-15 15:34:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][10/1251] eta 0:13:57 lr 0.000989 time 0.3170 (0.6748) loss 5.3226 (4.7018) grad_norm 1.1051 (1.1865) [2021-04-15 15:34:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][20/1251] eta 0:10:01 lr 0.000989 time 0.2446 (0.4889) loss 4.6686 (4.6416) grad_norm 1.1418 (1.1861) [2021-04-15 15:34:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][30/1251] eta 0:08:36 lr 0.000989 time 0.2644 (0.4227) loss 4.1726 (4.6332) grad_norm 1.1756 (1.1888) [2021-04-15 15:34:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][40/1251] eta 0:07:49 lr 0.000989 time 0.2599 (0.3879) loss 4.7394 (4.6950) grad_norm 1.0660 (1.1840) [2021-04-15 15:35:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][50/1251] eta 0:07:20 lr 0.000989 time 0.2715 (0.3666) loss 5.1668 (4.7355) grad_norm 1.3274 (1.1870) [2021-04-15 15:35:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][60/1251] eta 0:07:00 lr 0.000989 time 0.2916 (0.3528) loss 4.6304 (4.7265) grad_norm 1.0895 (1.1855) [2021-04-15 15:35:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][70/1251] eta 0:06:46 lr 0.000989 time 0.2759 (0.3439) loss 3.7379 (4.7134) grad_norm 1.3764 (1.1989) [2021-04-15 15:35:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][80/1251] eta 0:06:33 lr 0.000989 time 0.2966 (0.3358) loss 5.1670 (4.6805) grad_norm 1.4356 (1.2143) [2021-04-15 15:35:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][90/1251] eta 0:06:22 lr 0.000989 time 0.3015 (0.3294) loss 4.9245 (4.6724) grad_norm 1.5098 (1.2244) [2021-04-15 15:35:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][100/1251] eta 0:06:13 lr 0.000989 time 0.2986 (0.3245) loss 4.2008 (4.6367) grad_norm 1.1189 (1.2263) [2021-04-15 15:35:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][110/1251] eta 0:06:05 lr 0.000989 time 0.2636 (0.3204) loss 5.1196 (4.6439) grad_norm 1.2437 (1.2223) [2021-04-15 15:35:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][120/1251] eta 0:05:58 lr 0.000989 time 0.2700 (0.3169) loss 5.0256 (4.6480) grad_norm 1.1386 (1.2182) [2021-04-15 15:35:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][130/1251] eta 0:05:52 lr 0.000989 time 0.2560 (0.3141) loss 4.4709 (4.6355) grad_norm 1.0552 (1.2139) [2021-04-15 15:35:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][140/1251] eta 0:05:46 lr 0.000989 time 0.2748 (0.3120) loss 4.2080 (4.6306) grad_norm 1.2633 (1.2250) [2021-04-15 15:35:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][150/1251] eta 0:05:41 lr 0.000989 time 0.2886 (0.3099) loss 5.4425 (4.6315) grad_norm 1.3269 (1.2273) [2021-04-15 15:35:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][160/1251] eta 0:05:36 lr 0.000989 time 0.3006 (0.3081) loss 4.7508 (4.6221) grad_norm 1.1253 (1.2329) [2021-04-15 15:35:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][170/1251] eta 0:05:31 lr 0.000989 time 0.2524 (0.3066) loss 4.1074 (4.6172) grad_norm 1.1532 (1.2305) [2021-04-15 15:35:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][180/1251] eta 0:05:27 lr 0.000989 time 0.2756 (0.3053) loss 4.9074 (4.5995) grad_norm 1.0873 (1.2285) [2021-04-15 15:35:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][190/1251] eta 0:05:22 lr 0.000989 time 0.2597 (0.3039) loss 3.7396 (4.5963) grad_norm 1.1953 (1.2244) [2021-04-15 15:35:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][200/1251] eta 0:05:18 lr 0.000989 time 0.2830 (0.3026) loss 3.0734 (4.5858) grad_norm 1.0192 (1.2260) [2021-04-15 15:35:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][210/1251] eta 0:05:14 lr 0.000989 time 0.2870 (0.3019) loss 3.3050 (4.5753) grad_norm 1.4732 (1.2257) [2021-04-15 15:35:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][220/1251] eta 0:05:10 lr 0.000989 time 0.2936 (0.3010) loss 4.6749 (4.5616) grad_norm 1.3874 (1.2274) [2021-04-15 15:35:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][230/1251] eta 0:05:06 lr 0.000989 time 0.2739 (0.3003) loss 4.9926 (4.5528) grad_norm 1.2444 (1.2331) [2021-04-15 15:35:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][240/1251] eta 0:05:02 lr 0.000989 time 0.2821 (0.2992) loss 4.5415 (4.5563) grad_norm 1.3908 (1.2360) [2021-04-15 15:35:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][250/1251] eta 0:04:58 lr 0.000989 time 0.2930 (0.2985) loss 5.1074 (4.5503) grad_norm 0.9295 (1.2350) [2021-04-15 15:36:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][260/1251] eta 0:04:55 lr 0.000989 time 0.2713 (0.2979) loss 4.4869 (4.5521) grad_norm 1.3547 (1.2357) [2021-04-15 15:36:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][270/1251] eta 0:04:51 lr 0.000989 time 0.2890 (0.2973) loss 4.8315 (4.5514) grad_norm 1.2619 (1.2398) [2021-04-15 15:36:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][280/1251] eta 0:04:47 lr 0.000989 time 0.2736 (0.2966) loss 4.9745 (4.5523) grad_norm 1.0543 (1.2381) [2021-04-15 15:36:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][290/1251] eta 0:04:44 lr 0.000989 time 0.2950 (0.2961) loss 4.4833 (4.5591) grad_norm 0.9593 (1.2399) [2021-04-15 15:36:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][300/1251] eta 0:04:41 lr 0.000989 time 0.2456 (0.2955) loss 4.2333 (4.5582) grad_norm 1.6871 (1.2423) [2021-04-15 15:36:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][310/1251] eta 0:04:37 lr 0.000989 time 0.2850 (0.2954) loss 4.3724 (4.5568) grad_norm 1.2558 (1.2424) [2021-04-15 15:36:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][320/1251] eta 0:04:34 lr 0.000989 time 0.3030 (0.2950) loss 3.2124 (4.5579) grad_norm 0.9927 (1.2424) [2021-04-15 15:36:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][330/1251] eta 0:04:31 lr 0.000989 time 0.2941 (0.2946) loss 4.9165 (4.5550) grad_norm 1.2942 (1.2434) [2021-04-15 15:36:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][340/1251] eta 0:04:27 lr 0.000989 time 0.3001 (0.2941) loss 5.1393 (4.5585) grad_norm 1.3835 (1.2435) [2021-04-15 15:36:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][350/1251] eta 0:04:25 lr 0.000989 time 0.4791 (0.2942) loss 3.5700 (4.5485) grad_norm 1.0946 (1.2429) [2021-04-15 15:36:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][360/1251] eta 0:04:21 lr 0.000989 time 0.2553 (0.2935) loss 4.6927 (4.5456) grad_norm 1.1746 (1.2408) [2021-04-15 15:36:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][370/1251] eta 0:04:18 lr 0.000989 time 0.2539 (0.2934) loss 4.5644 (4.5505) grad_norm 1.1459 (1.2401) [2021-04-15 15:36:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][380/1251] eta 0:04:15 lr 0.000989 time 0.2814 (0.2931) loss 4.7005 (4.5477) grad_norm 0.9545 (1.2388) [2021-04-15 15:36:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][390/1251] eta 0:04:12 lr 0.000989 time 0.2786 (0.2929) loss 4.3852 (4.5486) grad_norm 1.2871 (1.2370) [2021-04-15 15:36:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][400/1251] eta 0:04:08 lr 0.000989 time 0.2647 (0.2925) loss 4.5118 (4.5421) grad_norm 1.2200 (1.2368) [2021-04-15 15:36:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][410/1251] eta 0:04:05 lr 0.000989 time 0.2636 (0.2921) loss 5.0243 (4.5391) grad_norm 1.0195 (1.2368) [2021-04-15 15:36:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][420/1251] eta 0:04:02 lr 0.000989 time 0.2782 (0.2918) loss 5.0590 (4.5349) grad_norm 1.2231 (1.2398) [2021-04-15 15:36:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][430/1251] eta 0:03:59 lr 0.000989 time 0.2594 (0.2915) loss 4.7995 (4.5294) grad_norm 1.1979 (1.2411) [2021-04-15 15:36:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][440/1251] eta 0:03:56 lr 0.000989 time 0.2800 (0.2912) loss 5.0453 (4.5251) grad_norm 1.0156 (1.2397) [2021-04-15 15:36:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][450/1251] eta 0:03:52 lr 0.000989 time 0.2862 (0.2908) loss 4.8367 (4.5309) grad_norm 1.5276 (1.2388) [2021-04-15 15:36:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][460/1251] eta 0:03:49 lr 0.000989 time 0.3066 (0.2905) loss 5.2193 (4.5334) grad_norm 1.0891 (1.2384) [2021-04-15 15:36:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][470/1251] eta 0:03:46 lr 0.000989 time 0.2773 (0.2904) loss 3.9225 (4.5378) grad_norm 1.2238 (1.2362) [2021-04-15 15:37:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][480/1251] eta 0:03:43 lr 0.000989 time 0.2838 (0.2902) loss 3.5880 (4.5358) grad_norm 1.3223 (1.2376) [2021-04-15 15:37:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][490/1251] eta 0:03:40 lr 0.000989 time 0.2878 (0.2901) loss 3.9526 (4.5287) grad_norm 1.1566 (1.2382) [2021-04-15 15:37:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][500/1251] eta 0:03:37 lr 0.000989 time 0.2776 (0.2898) loss 5.1788 (4.5277) grad_norm 1.2364 (1.2396) [2021-04-15 15:37:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][510/1251] eta 0:03:34 lr 0.000989 time 0.2957 (0.2895) loss 4.3127 (4.5228) grad_norm 1.1129 (1.2410) [2021-04-15 15:37:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][520/1251] eta 0:03:31 lr 0.000989 time 0.3032 (0.2893) loss 5.0454 (4.5247) grad_norm 1.0029 (1.2396) [2021-04-15 15:37:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][530/1251] eta 0:03:28 lr 0.000989 time 0.2765 (0.2890) loss 4.1829 (4.5274) grad_norm 1.0740 (inf) [2021-04-15 15:37:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][540/1251] eta 0:03:25 lr 0.000989 time 0.2752 (0.2888) loss 5.2282 (4.5348) grad_norm 1.4358 (inf) [2021-04-15 15:37:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][550/1251] eta 0:03:22 lr 0.000989 time 0.2799 (0.2885) loss 3.1835 (4.5288) grad_norm 1.1931 (inf) [2021-04-15 15:37:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][560/1251] eta 0:03:19 lr 0.000989 time 0.2762 (0.2883) loss 4.3755 (4.5295) grad_norm 1.8519 (inf) [2021-04-15 15:37:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][570/1251] eta 0:03:16 lr 0.000989 time 0.2808 (0.2881) loss 4.1639 (4.5300) grad_norm 1.5537 (inf) [2021-04-15 15:37:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][580/1251] eta 0:03:13 lr 0.000989 time 0.2775 (0.2880) loss 3.3862 (4.5264) grad_norm 1.1674 (inf) [2021-04-15 15:37:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][590/1251] eta 0:03:10 lr 0.000989 time 0.2712 (0.2878) loss 4.8203 (4.5286) grad_norm 1.4103 (inf) [2021-04-15 15:37:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][600/1251] eta 0:03:07 lr 0.000989 time 0.2747 (0.2875) loss 3.7740 (4.5211) grad_norm 1.0560 (inf) [2021-04-15 15:37:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][610/1251] eta 0:03:04 lr 0.000989 time 0.2581 (0.2874) loss 3.8090 (4.5187) grad_norm 1.0912 (inf) [2021-04-15 15:37:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][620/1251] eta 0:03:01 lr 0.000989 time 0.2772 (0.2872) loss 4.0229 (4.5211) grad_norm 1.5917 (inf) [2021-04-15 15:37:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][630/1251] eta 0:02:58 lr 0.000989 time 0.2637 (0.2871) loss 5.3048 (4.5212) grad_norm 1.2676 (inf) [2021-04-15 15:37:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][640/1251] eta 0:02:55 lr 0.000989 time 0.2970 (0.2870) loss 5.1562 (4.5206) grad_norm 1.2003 (inf) [2021-04-15 15:37:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][650/1251] eta 0:02:52 lr 0.000989 time 0.2630 (0.2869) loss 3.2382 (4.5150) grad_norm 1.3814 (inf) [2021-04-15 15:37:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][660/1251] eta 0:02:49 lr 0.000989 time 0.2702 (0.2867) loss 4.1065 (4.5161) grad_norm 1.3321 (inf) [2021-04-15 15:37:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][670/1251] eta 0:02:46 lr 0.000989 time 0.2789 (0.2866) loss 3.6220 (4.5156) grad_norm 1.0486 (inf) [2021-04-15 15:37:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][680/1251] eta 0:02:43 lr 0.000989 time 0.2581 (0.2865) loss 5.0742 (4.5184) grad_norm 1.8723 (inf) [2021-04-15 15:38:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][690/1251] eta 0:02:40 lr 0.000989 time 0.2987 (0.2864) loss 4.7766 (4.5130) grad_norm 1.3426 (inf) [2021-04-15 15:38:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][700/1251] eta 0:02:37 lr 0.000989 time 0.2610 (0.2862) loss 4.6964 (4.5091) grad_norm 1.1639 (inf) [2021-04-15 15:38:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][710/1251] eta 0:02:34 lr 0.000989 time 0.2801 (0.2861) loss 3.8720 (4.5094) grad_norm 1.7461 (inf) [2021-04-15 15:38:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][720/1251] eta 0:02:31 lr 0.000989 time 0.2839 (0.2861) loss 4.8359 (4.5073) grad_norm 1.3507 (inf) [2021-04-15 15:38:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][730/1251] eta 0:02:28 lr 0.000989 time 0.2776 (0.2859) loss 3.9422 (4.5076) grad_norm 1.5106 (inf) [2021-04-15 15:38:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][740/1251] eta 0:02:26 lr 0.000989 time 0.2603 (0.2858) loss 3.7551 (4.5054) grad_norm 1.1277 (inf) [2021-04-15 15:38:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][750/1251] eta 0:02:23 lr 0.000989 time 0.2616 (0.2857) loss 3.8568 (4.4993) grad_norm 0.9608 (inf) [2021-04-15 15:38:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][760/1251] eta 0:02:20 lr 0.000989 time 0.2813 (0.2856) loss 5.3049 (4.5008) grad_norm 1.1003 (inf) [2021-04-15 15:38:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][770/1251] eta 0:02:17 lr 0.000989 time 0.2673 (0.2855) loss 4.8824 (4.5007) grad_norm 1.4338 (inf) [2021-04-15 15:38:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][780/1251] eta 0:02:14 lr 0.000989 time 0.2616 (0.2854) loss 5.2049 (4.5012) grad_norm 1.0304 (inf) [2021-04-15 15:38:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][790/1251] eta 0:02:11 lr 0.000988 time 0.2633 (0.2853) loss 5.0862 (4.4990) grad_norm 1.2393 (inf) [2021-04-15 15:38:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][800/1251] eta 0:02:08 lr 0.000988 time 0.2600 (0.2852) loss 4.7743 (4.4987) grad_norm 1.6568 (inf) [2021-04-15 15:38:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][810/1251] eta 0:02:05 lr 0.000988 time 0.3126 (0.2852) loss 4.3877 (4.5019) grad_norm 1.1225 (inf) [2021-04-15 15:38:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][820/1251] eta 0:02:02 lr 0.000988 time 0.2539 (0.2850) loss 3.8441 (4.4998) grad_norm 1.0498 (inf) [2021-04-15 15:38:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][830/1251] eta 0:01:59 lr 0.000988 time 0.2815 (0.2850) loss 3.5159 (4.4972) grad_norm 1.5703 (inf) [2021-04-15 15:38:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][840/1251] eta 0:01:57 lr 0.000988 time 0.2861 (0.2850) loss 4.0157 (4.4978) grad_norm 1.6333 (inf) [2021-04-15 15:38:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][850/1251] eta 0:01:54 lr 0.000988 time 0.2978 (0.2848) loss 4.4096 (4.4988) grad_norm 1.1472 (inf) [2021-04-15 15:38:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][860/1251] eta 0:01:51 lr 0.000988 time 0.2947 (0.2848) loss 4.3657 (4.4979) grad_norm 1.3210 (inf) [2021-04-15 15:38:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][870/1251] eta 0:01:48 lr 0.000988 time 0.2816 (0.2847) loss 5.4055 (4.4979) grad_norm 1.2179 (inf) [2021-04-15 15:38:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][880/1251] eta 0:01:45 lr 0.000988 time 0.4111 (0.2848) loss 4.9698 (4.4978) grad_norm 1.3713 (inf) [2021-04-15 15:38:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][890/1251] eta 0:01:42 lr 0.000988 time 0.3122 (0.2847) loss 4.9112 (4.4970) grad_norm 1.2253 (inf) [2021-04-15 15:38:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][900/1251] eta 0:01:39 lr 0.000988 time 0.2920 (0.2846) loss 5.1233 (4.5025) grad_norm 1.1835 (inf) [2021-04-15 15:39:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][910/1251] eta 0:01:37 lr 0.000988 time 0.2921 (0.2845) loss 3.5793 (4.5055) grad_norm 1.3516 (inf) [2021-04-15 15:39:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][920/1251] eta 0:01:34 lr 0.000988 time 0.2676 (0.2845) loss 5.2634 (4.5083) grad_norm 1.1917 (inf) [2021-04-15 15:39:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][930/1251] eta 0:01:31 lr 0.000988 time 0.2768 (0.2845) loss 4.1978 (4.5046) grad_norm 1.2193 (inf) [2021-04-15 15:39:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][940/1251] eta 0:01:28 lr 0.000988 time 0.2658 (0.2846) loss 5.0337 (4.5045) grad_norm 1.0423 (inf) [2021-04-15 15:39:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][950/1251] eta 0:01:25 lr 0.000988 time 0.2860 (0.2845) loss 3.9515 (4.5042) grad_norm 1.1450 (inf) [2021-04-15 15:39:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][960/1251] eta 0:01:22 lr 0.000988 time 0.2873 (0.2844) loss 4.8821 (4.5031) grad_norm 1.1580 (inf) [2021-04-15 15:39:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][970/1251] eta 0:01:19 lr 0.000988 time 0.2688 (0.2843) loss 4.4232 (4.5029) grad_norm 1.3157 (inf) [2021-04-15 15:39:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][980/1251] eta 0:01:17 lr 0.000988 time 0.2943 (0.2843) loss 3.8070 (4.5018) grad_norm 1.3896 (inf) [2021-04-15 15:39:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][990/1251] eta 0:01:14 lr 0.000988 time 0.2886 (0.2842) loss 4.0645 (4.5037) grad_norm 1.1494 (inf) [2021-04-15 15:39:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][1000/1251] eta 0:01:11 lr 0.000988 time 0.2659 (0.2841) loss 4.4329 (4.5024) grad_norm 1.5917 (inf) [2021-04-15 15:39:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][1010/1251] eta 0:01:08 lr 0.000988 time 0.2691 (0.2840) loss 4.8484 (4.5055) grad_norm 1.2422 (inf) [2021-04-15 15:39:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][1020/1251] eta 0:01:05 lr 0.000988 time 0.2783 (0.2840) loss 3.5218 (4.5056) grad_norm 1.7636 (inf) [2021-04-15 15:39:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][1030/1251] eta 0:01:02 lr 0.000988 time 0.2748 (0.2839) loss 4.5339 (4.5025) grad_norm 1.1804 (inf) [2021-04-15 15:39:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][1040/1251] eta 0:00:59 lr 0.000988 time 0.2808 (0.2838) loss 4.2884 (4.5023) grad_norm 1.3466 (inf) [2021-04-15 15:39:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][1050/1251] eta 0:00:57 lr 0.000988 time 0.2605 (0.2837) loss 4.2712 (4.5038) grad_norm 1.5238 (inf) [2021-04-15 15:39:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][1060/1251] eta 0:00:54 lr 0.000988 time 0.2877 (0.2837) loss 3.9649 (4.5046) grad_norm 1.0252 (inf) [2021-04-15 15:39:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][1070/1251] eta 0:00:51 lr 0.000988 time 0.3039 (0.2836) loss 3.2958 (4.5029) grad_norm 0.9357 (inf) [2021-04-15 15:39:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][1080/1251] eta 0:00:48 lr 0.000988 time 0.2715 (0.2836) loss 4.7740 (4.5018) grad_norm 1.3847 (inf) [2021-04-15 15:39:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][1090/1251] eta 0:00:45 lr 0.000988 time 0.2762 (0.2835) loss 5.2592 (4.5036) grad_norm 1.5533 (inf) [2021-04-15 15:39:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][1100/1251] eta 0:00:42 lr 0.000988 time 0.2748 (0.2835) loss 4.8871 (4.5041) grad_norm 1.0030 (inf) [2021-04-15 15:39:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][1110/1251] eta 0:00:39 lr 0.000988 time 0.2871 (0.2834) loss 4.7930 (4.5053) grad_norm 1.4869 (inf) [2021-04-15 15:40:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][1120/1251] eta 0:00:37 lr 0.000988 time 0.2747 (0.2833) loss 5.4030 (4.5051) grad_norm 1.1329 (inf) [2021-04-15 15:40:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][1130/1251] eta 0:00:34 lr 0.000988 time 0.2767 (0.2833) loss 4.8305 (4.5047) grad_norm 1.4618 (inf) [2021-04-15 15:40:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][1140/1251] eta 0:00:31 lr 0.000988 time 0.2922 (0.2833) loss 3.3340 (4.5047) grad_norm 1.1998 (inf) [2021-04-15 15:40:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][1150/1251] eta 0:00:28 lr 0.000988 time 0.2863 (0.2832) loss 4.5350 (4.5027) grad_norm 1.2985 (inf) [2021-04-15 15:40:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][1160/1251] eta 0:00:25 lr 0.000988 time 0.2730 (0.2832) loss 4.5798 (4.5020) grad_norm 1.4271 (inf) [2021-04-15 15:40:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][1170/1251] eta 0:00:22 lr 0.000988 time 0.2980 (0.2832) loss 4.6072 (4.5037) grad_norm 1.1633 (inf) [2021-04-15 15:40:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][1180/1251] eta 0:00:20 lr 0.000988 time 0.2701 (0.2831) loss 3.8450 (4.5026) grad_norm 1.4130 (inf) [2021-04-15 15:40:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][1190/1251] eta 0:00:17 lr 0.000988 time 0.2662 (0.2831) loss 4.6224 (4.5001) grad_norm 1.1644 (inf) [2021-04-15 15:40:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][1200/1251] eta 0:00:14 lr 0.000988 time 0.2846 (0.2830) loss 5.1138 (4.5016) grad_norm 1.0400 (inf) [2021-04-15 15:40:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][1210/1251] eta 0:00:11 lr 0.000988 time 0.2823 (0.2830) loss 4.3018 (4.5013) grad_norm 1.1832 (inf) [2021-04-15 15:40:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][1220/1251] eta 0:00:08 lr 0.000988 time 0.2656 (0.2829) loss 4.2206 (4.5030) grad_norm 1.1190 (inf) [2021-04-15 15:40:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][1230/1251] eta 0:00:05 lr 0.000988 time 0.2674 (0.2829) loss 4.1886 (4.5031) grad_norm 1.2410 (inf) [2021-04-15 15:40:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][1240/1251] eta 0:00:03 lr 0.000988 time 0.2485 (0.2828) loss 3.8522 (4.5022) grad_norm 1.2738 (inf) [2021-04-15 15:40:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [20/300][1250/1251] eta 0:00:00 lr 0.000988 time 0.2486 (0.2825) loss 4.5487 (4.5027) grad_norm 1.0122 (inf) [2021-04-15 15:40:38 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 20 training takes 0:05:55 [2021-04-15 15:40:38 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_20.pth saving...... [2021-04-15 15:40:55 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_20.pth saved !!! [2021-04-15 15:40:56 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.212 (1.212) Loss 1.8492 (1.8492) Acc@1 58.301 (58.301) Acc@5 81.641 (81.641) [2021-04-15 15:40:57 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.107 (0.241) Loss 1.7917 (1.7703) Acc@1 60.449 (60.272) Acc@5 83.008 (83.461) [2021-04-15 15:40:59 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.174 (0.216) Loss 1.7735 (1.7828) Acc@1 58.594 (59.752) Acc@5 82.812 (83.394) [2021-04-15 15:41:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.093 (0.241) Loss 1.8255 (1.7839) Acc@1 58.105 (59.703) Acc@5 83.105 (83.326) [2021-04-15 15:41:04 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.213) Loss 1.6557 (1.7839) Acc@1 62.891 (59.835) Acc@5 86.035 (83.346) [2021-04-15 15:41:05 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 59.812 Acc@5 83.466 [2021-04-15 15:41:05 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 59.8% [2021-04-15 15:41:05 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 59.81% [2021-04-15 15:41:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][0/1251] eta 1:21:12 lr 0.000988 time 3.8948 (3.8948) loss 3.7719 (3.7719) grad_norm 1.1526 (1.1526) [2021-04-15 15:41:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][10/1251] eta 0:12:41 lr 0.000988 time 0.2755 (0.6136) loss 4.1150 (4.2879) grad_norm 0.9625 (1.1962) [2021-04-15 15:41:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][20/1251] eta 0:09:18 lr 0.000988 time 0.2769 (0.4540) loss 4.5059 (4.3551) grad_norm 1.5323 (1.2451) [2021-04-15 15:41:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][30/1251] eta 0:08:07 lr 0.000988 time 0.2741 (0.3996) loss 4.7783 (4.3931) grad_norm 1.0000 (1.2242) [2021-04-15 15:41:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][40/1251] eta 0:07:28 lr 0.000988 time 0.3001 (0.3700) loss 4.1727 (4.4609) grad_norm 1.5745 (1.2573) [2021-04-15 15:41:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][50/1251] eta 0:07:04 lr 0.000988 time 0.2865 (0.3533) loss 3.9636 (4.4291) grad_norm 1.0456 (1.2368) [2021-04-15 15:41:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][60/1251] eta 0:06:47 lr 0.000988 time 0.2856 (0.3420) loss 4.5014 (4.3838) grad_norm 1.2253 (1.2280) [2021-04-15 15:41:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][70/1251] eta 0:06:36 lr 0.000988 time 0.3033 (0.3356) loss 4.4306 (4.4237) grad_norm 1.1283 (1.2181) [2021-04-15 15:41:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][80/1251] eta 0:06:24 lr 0.000988 time 0.2981 (0.3281) loss 4.6650 (4.4580) grad_norm 1.6362 (1.2313) [2021-04-15 15:41:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][90/1251] eta 0:06:14 lr 0.000988 time 0.2909 (0.3228) loss 5.5359 (4.4610) grad_norm 1.0384 (1.2378) [2021-04-15 15:41:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][100/1251] eta 0:06:06 lr 0.000988 time 0.2766 (0.3183) loss 4.7241 (4.4620) grad_norm 1.2857 (1.2413) [2021-04-15 15:41:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][110/1251] eta 0:05:59 lr 0.000988 time 0.2788 (0.3152) loss 5.4155 (4.4458) grad_norm 1.3590 (1.2538) [2021-04-15 15:41:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][120/1251] eta 0:05:53 lr 0.000988 time 0.2751 (0.3121) loss 5.1606 (4.4597) grad_norm 1.5051 (1.2609) [2021-04-15 15:41:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][130/1251] eta 0:05:49 lr 0.000988 time 0.3477 (0.3115) loss 3.8388 (4.4660) grad_norm 1.1322 (1.2649) [2021-04-15 15:41:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][140/1251] eta 0:05:45 lr 0.000988 time 0.2717 (0.3110) loss 4.8498 (4.4647) grad_norm 1.4624 (1.2583) [2021-04-15 15:41:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][150/1251] eta 0:05:39 lr 0.000988 time 0.2881 (0.3086) loss 4.8488 (4.4684) grad_norm 1.1592 (1.2545) [2021-04-15 15:41:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][160/1251] eta 0:05:34 lr 0.000988 time 0.2770 (0.3068) loss 4.7159 (4.4878) grad_norm 1.2184 (1.2483) [2021-04-15 15:41:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][170/1251] eta 0:05:29 lr 0.000988 time 0.2954 (0.3052) loss 4.2179 (4.4697) grad_norm 1.3587 (1.2485) [2021-04-15 15:42:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][180/1251] eta 0:05:25 lr 0.000988 time 0.2831 (0.3036) loss 3.7289 (4.4649) grad_norm 1.0998 (1.2421) [2021-04-15 15:42:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][190/1251] eta 0:05:20 lr 0.000988 time 0.2923 (0.3022) loss 4.6090 (4.4769) grad_norm 1.0980 (1.2437) [2021-04-15 15:42:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][200/1251] eta 0:05:16 lr 0.000988 time 0.2548 (0.3011) loss 4.5377 (4.4861) grad_norm 1.2444 (1.2391) [2021-04-15 15:42:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][210/1251] eta 0:05:12 lr 0.000988 time 0.2543 (0.2998) loss 4.6672 (4.4815) grad_norm 1.5097 (1.2404) [2021-04-15 15:42:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][220/1251] eta 0:05:08 lr 0.000988 time 0.2604 (0.2989) loss 4.5588 (4.4517) grad_norm 1.2664 (1.2459) [2021-04-15 15:42:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][230/1251] eta 0:05:04 lr 0.000988 time 0.2662 (0.2980) loss 5.3247 (4.4578) grad_norm 1.1623 (1.2472) [2021-04-15 15:42:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][240/1251] eta 0:05:00 lr 0.000988 time 0.2652 (0.2971) loss 5.4703 (4.4540) grad_norm 1.1003 (1.2469) [2021-04-15 15:42:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][250/1251] eta 0:04:56 lr 0.000988 time 0.2679 (0.2966) loss 4.4512 (4.4529) grad_norm 1.3086 (1.2448) [2021-04-15 15:42:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][260/1251] eta 0:04:53 lr 0.000988 time 0.2511 (0.2958) loss 4.8460 (4.4497) grad_norm 1.1786 (1.2451) [2021-04-15 15:42:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][270/1251] eta 0:04:49 lr 0.000988 time 0.2927 (0.2950) loss 5.1673 (4.4528) grad_norm 1.3450 (1.2431) [2021-04-15 15:42:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][280/1251] eta 0:04:46 lr 0.000988 time 0.3058 (0.2945) loss 5.1639 (4.4486) grad_norm 1.1872 (1.2438) [2021-04-15 15:42:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][290/1251] eta 0:04:42 lr 0.000988 time 0.2846 (0.2938) loss 4.7378 (4.4511) grad_norm 1.1068 (1.2408) [2021-04-15 15:42:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][300/1251] eta 0:04:39 lr 0.000988 time 0.3011 (0.2934) loss 4.2956 (4.4503) grad_norm 1.3821 (1.2395) [2021-04-15 15:42:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][310/1251] eta 0:04:35 lr 0.000988 time 0.2893 (0.2929) loss 4.6895 (4.4495) grad_norm 0.9160 (1.2374) [2021-04-15 15:42:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][320/1251] eta 0:04:32 lr 0.000988 time 0.2836 (0.2924) loss 4.0018 (4.4466) grad_norm 1.4283 (1.2365) [2021-04-15 15:42:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][330/1251] eta 0:04:28 lr 0.000988 time 0.2739 (0.2920) loss 4.9708 (4.4497) grad_norm 1.3438 (1.2374) [2021-04-15 15:42:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][340/1251] eta 0:04:25 lr 0.000988 time 0.2681 (0.2916) loss 4.4532 (4.4497) grad_norm 1.1590 (1.2355) [2021-04-15 15:42:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][350/1251] eta 0:04:22 lr 0.000988 time 0.2888 (0.2912) loss 4.7769 (4.4410) grad_norm 1.1512 (1.2315) [2021-04-15 15:42:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][360/1251] eta 0:04:19 lr 0.000988 time 0.2683 (0.2908) loss 4.9723 (4.4479) grad_norm 1.6567 (1.2357) [2021-04-15 15:42:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][370/1251] eta 0:04:15 lr 0.000988 time 0.2841 (0.2905) loss 4.4408 (4.4483) grad_norm 1.0578 (1.2372) [2021-04-15 15:42:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][380/1251] eta 0:04:12 lr 0.000988 time 0.2732 (0.2900) loss 4.8600 (4.4450) grad_norm 1.1666 (1.2362) [2021-04-15 15:42:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][390/1251] eta 0:04:09 lr 0.000988 time 0.2691 (0.2899) loss 4.2747 (4.4482) grad_norm 1.1899 (1.2367) [2021-04-15 15:43:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][400/1251] eta 0:04:06 lr 0.000988 time 0.2671 (0.2896) loss 5.0492 (4.4543) grad_norm 1.2475 (1.2346) [2021-04-15 15:43:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][410/1251] eta 0:04:03 lr 0.000988 time 0.2707 (0.2894) loss 3.7034 (4.4478) grad_norm 1.4009 (1.2367) [2021-04-15 15:43:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][420/1251] eta 0:04:00 lr 0.000988 time 0.2949 (0.2891) loss 4.9322 (4.4471) grad_norm 1.4286 (1.2371) [2021-04-15 15:43:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][430/1251] eta 0:03:57 lr 0.000988 time 0.2678 (0.2887) loss 4.6613 (4.4462) grad_norm 1.1995 (1.2395) [2021-04-15 15:43:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][440/1251] eta 0:03:54 lr 0.000988 time 0.2755 (0.2888) loss 4.5671 (4.4487) grad_norm 1.4795 (1.2393) [2021-04-15 15:43:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][450/1251] eta 0:03:51 lr 0.000988 time 0.2863 (0.2885) loss 4.8349 (4.4461) grad_norm 0.9214 (1.2382) [2021-04-15 15:43:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][460/1251] eta 0:03:48 lr 0.000988 time 0.2765 (0.2883) loss 3.6795 (4.4444) grad_norm 1.2431 (1.2376) [2021-04-15 15:43:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][470/1251] eta 0:03:45 lr 0.000988 time 0.2716 (0.2882) loss 4.4194 (4.4434) grad_norm 1.4884 (1.2388) [2021-04-15 15:43:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][480/1251] eta 0:03:41 lr 0.000988 time 0.2642 (0.2879) loss 4.5860 (4.4381) grad_norm 1.3624 (1.2396) [2021-04-15 15:43:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][490/1251] eta 0:03:39 lr 0.000988 time 0.2736 (0.2879) loss 3.5719 (4.4365) grad_norm 1.1710 (1.2402) [2021-04-15 15:43:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][500/1251] eta 0:03:36 lr 0.000988 time 0.2556 (0.2877) loss 4.9715 (4.4375) grad_norm 0.9792 (1.2395) [2021-04-15 15:43:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][510/1251] eta 0:03:33 lr 0.000988 time 0.2861 (0.2875) loss 4.1172 (4.4358) grad_norm 1.0642 (1.2391) [2021-04-15 15:43:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][520/1251] eta 0:03:30 lr 0.000988 time 0.2721 (0.2873) loss 4.4118 (4.4315) grad_norm 1.2052 (1.2397) [2021-04-15 15:43:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][530/1251] eta 0:03:27 lr 0.000988 time 0.2737 (0.2871) loss 4.8518 (4.4330) grad_norm 1.1618 (1.2394) [2021-04-15 15:43:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][540/1251] eta 0:03:24 lr 0.000988 time 0.3099 (0.2871) loss 4.9280 (4.4317) grad_norm 1.1203 (1.2368) [2021-04-15 15:43:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][550/1251] eta 0:03:21 lr 0.000988 time 0.2784 (0.2870) loss 3.7244 (4.4236) grad_norm 1.0674 (1.2358) [2021-04-15 15:43:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][560/1251] eta 0:03:18 lr 0.000988 time 0.2769 (0.2868) loss 4.8919 (4.4244) grad_norm 1.2850 (1.2374) [2021-04-15 15:43:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][570/1251] eta 0:03:15 lr 0.000988 time 0.2596 (0.2866) loss 5.1067 (4.4285) grad_norm 1.4030 (1.2373) [2021-04-15 15:43:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][580/1251] eta 0:03:12 lr 0.000988 time 0.2666 (0.2865) loss 4.6455 (4.4293) grad_norm 1.1993 (1.2390) [2021-04-15 15:43:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][590/1251] eta 0:03:09 lr 0.000988 time 0.2667 (0.2864) loss 4.9992 (4.4369) grad_norm 1.2023 (1.2388) [2021-04-15 15:43:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][600/1251] eta 0:03:06 lr 0.000988 time 0.2954 (0.2862) loss 4.6120 (4.4334) grad_norm 1.1540 (1.2391) [2021-04-15 15:44:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][610/1251] eta 0:03:03 lr 0.000988 time 0.2791 (0.2861) loss 4.0875 (4.4316) grad_norm 1.2210 (1.2368) [2021-04-15 15:44:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][620/1251] eta 0:03:00 lr 0.000988 time 0.3009 (0.2859) loss 5.0110 (4.4343) grad_norm 1.2078 (1.2354) [2021-04-15 15:44:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][630/1251] eta 0:02:57 lr 0.000988 time 0.2916 (0.2858) loss 4.3762 (4.4328) grad_norm 1.3054 (1.2366) [2021-04-15 15:44:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][640/1251] eta 0:02:54 lr 0.000987 time 0.2713 (0.2858) loss 3.6208 (4.4339) grad_norm 1.0143 (1.2354) [2021-04-15 15:44:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][650/1251] eta 0:02:51 lr 0.000987 time 0.2781 (0.2856) loss 4.5822 (4.4347) grad_norm 1.3493 (1.2363) [2021-04-15 15:44:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][660/1251] eta 0:02:48 lr 0.000987 time 0.2516 (0.2854) loss 4.6533 (4.4319) grad_norm 1.3000 (1.2358) [2021-04-15 15:44:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][670/1251] eta 0:02:45 lr 0.000987 time 0.2759 (0.2853) loss 4.3911 (4.4330) grad_norm 1.2569 (1.2342) [2021-04-15 15:44:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][680/1251] eta 0:02:42 lr 0.000987 time 0.2590 (0.2852) loss 5.0882 (4.4369) grad_norm 1.2177 (1.2332) [2021-04-15 15:44:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][690/1251] eta 0:02:39 lr 0.000987 time 0.2781 (0.2851) loss 4.6367 (4.4357) grad_norm 1.4913 (1.2329) [2021-04-15 15:44:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][700/1251] eta 0:02:37 lr 0.000987 time 0.2862 (0.2850) loss 5.2031 (4.4385) grad_norm 1.4555 (1.2331) [2021-04-15 15:44:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][710/1251] eta 0:02:34 lr 0.000987 time 0.2738 (0.2851) loss 5.0100 (4.4391) grad_norm 1.0663 (1.2345) [2021-04-15 15:44:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][720/1251] eta 0:02:31 lr 0.000987 time 0.2686 (0.2852) loss 3.5563 (4.4344) grad_norm 1.2039 (1.2333) [2021-04-15 15:44:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][730/1251] eta 0:02:28 lr 0.000987 time 0.2917 (0.2852) loss 3.8547 (4.4329) grad_norm 1.3269 (1.2327) [2021-04-15 15:44:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][740/1251] eta 0:02:25 lr 0.000987 time 0.2548 (0.2851) loss 4.9317 (4.4339) grad_norm 1.0972 (1.2318) [2021-04-15 15:44:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][750/1251] eta 0:02:22 lr 0.000987 time 0.2673 (0.2851) loss 5.0927 (4.4341) grad_norm 1.1765 (1.2322) [2021-04-15 15:44:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][760/1251] eta 0:02:19 lr 0.000987 time 0.2750 (0.2849) loss 4.9568 (4.4395) grad_norm 1.5311 (1.2331) [2021-04-15 15:44:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][770/1251] eta 0:02:16 lr 0.000987 time 0.2827 (0.2848) loss 5.1224 (4.4406) grad_norm 1.1181 (1.2334) [2021-04-15 15:44:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][780/1251] eta 0:02:14 lr 0.000987 time 0.2702 (0.2847) loss 4.3321 (4.4411) grad_norm 1.0767 (1.2319) [2021-04-15 15:44:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][790/1251] eta 0:02:11 lr 0.000987 time 0.3101 (0.2847) loss 4.4657 (4.4403) grad_norm 1.1757 (1.2305) [2021-04-15 15:44:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][800/1251] eta 0:02:08 lr 0.000987 time 0.2624 (0.2846) loss 3.4104 (4.4384) grad_norm 1.3989 (1.2302) [2021-04-15 15:44:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][810/1251] eta 0:02:05 lr 0.000987 time 0.3007 (0.2845) loss 4.7973 (4.4379) grad_norm 1.2526 (1.2304) [2021-04-15 15:44:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][820/1251] eta 0:02:02 lr 0.000987 time 0.2819 (0.2844) loss 3.3932 (4.4382) grad_norm 1.1370 (1.2288) [2021-04-15 15:45:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][830/1251] eta 0:01:59 lr 0.000987 time 0.2769 (0.2843) loss 4.2359 (4.4387) grad_norm 1.0665 (1.2279) [2021-04-15 15:45:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][840/1251] eta 0:01:56 lr 0.000987 time 0.2454 (0.2842) loss 4.6028 (4.4360) grad_norm 1.1254 (1.2282) [2021-04-15 15:45:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][850/1251] eta 0:01:53 lr 0.000987 time 0.2771 (0.2841) loss 5.2184 (4.4360) grad_norm 1.1445 (1.2277) [2021-04-15 15:45:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][860/1251] eta 0:01:51 lr 0.000987 time 0.2824 (0.2840) loss 3.6078 (4.4358) grad_norm 1.7300 (1.2280) [2021-04-15 15:45:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][870/1251] eta 0:01:48 lr 0.000987 time 0.2702 (0.2840) loss 4.4283 (4.4364) grad_norm 1.2168 (1.2274) [2021-04-15 15:45:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][880/1251] eta 0:01:45 lr 0.000987 time 0.2844 (0.2839) loss 4.4421 (4.4379) grad_norm 1.3598 (1.2285) [2021-04-15 15:45:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][890/1251] eta 0:01:42 lr 0.000987 time 0.2649 (0.2838) loss 4.2390 (4.4388) grad_norm 1.2178 (1.2286) [2021-04-15 15:45:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][900/1251] eta 0:01:39 lr 0.000987 time 0.3046 (0.2837) loss 4.5013 (4.4387) grad_norm 1.1719 (1.2282) [2021-04-15 15:45:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][910/1251] eta 0:01:36 lr 0.000987 time 0.2599 (0.2836) loss 4.6624 (4.4371) grad_norm 1.0192 (1.2282) [2021-04-15 15:45:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][920/1251] eta 0:01:33 lr 0.000987 time 0.2585 (0.2836) loss 4.9504 (4.4377) grad_norm 0.9644 (1.2278) [2021-04-15 15:45:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][930/1251] eta 0:01:30 lr 0.000987 time 0.2469 (0.2834) loss 5.1598 (4.4381) grad_norm 1.0675 (1.2268) [2021-04-15 15:45:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][940/1251] eta 0:01:28 lr 0.000987 time 0.2737 (0.2834) loss 3.9447 (4.4399) grad_norm 1.3524 (1.2271) [2021-04-15 15:45:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][950/1251] eta 0:01:25 lr 0.000987 time 0.2804 (0.2832) loss 5.0577 (4.4427) grad_norm 1.1672 (1.2268) [2021-04-15 15:45:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][960/1251] eta 0:01:22 lr 0.000987 time 0.3021 (0.2832) loss 4.7101 (4.4431) grad_norm 0.9648 (1.2264) [2021-04-15 15:45:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][970/1251] eta 0:01:19 lr 0.000987 time 0.3020 (0.2832) loss 4.0966 (4.4421) grad_norm 1.1809 (1.2260) [2021-04-15 15:45:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][980/1251] eta 0:01:16 lr 0.000987 time 0.2846 (0.2831) loss 4.2918 (4.4402) grad_norm 1.3002 (1.2264) [2021-04-15 15:45:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][990/1251] eta 0:01:13 lr 0.000987 time 0.2610 (0.2831) loss 4.9768 (4.4442) grad_norm 1.1667 (1.2254) [2021-04-15 15:45:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][1000/1251] eta 0:01:11 lr 0.000987 time 0.2831 (0.2830) loss 4.9647 (4.4449) grad_norm 1.0413 (1.2244) [2021-04-15 15:45:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][1010/1251] eta 0:01:08 lr 0.000987 time 0.2722 (0.2829) loss 3.2631 (4.4440) grad_norm 1.2581 (1.2249) [2021-04-15 15:45:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][1020/1251] eta 0:01:05 lr 0.000987 time 0.2842 (0.2829) loss 3.5978 (4.4447) grad_norm 1.0195 (1.2249) [2021-04-15 15:45:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][1030/1251] eta 0:01:02 lr 0.000987 time 0.2623 (0.2829) loss 4.2703 (4.4456) grad_norm 1.3920 (1.2250) [2021-04-15 15:46:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][1040/1251] eta 0:00:59 lr 0.000987 time 0.3028 (0.2829) loss 3.5584 (4.4464) grad_norm 1.5061 (1.2252) [2021-04-15 15:46:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][1050/1251] eta 0:00:56 lr 0.000987 time 0.2584 (0.2829) loss 4.6916 (4.4490) grad_norm 1.0933 (1.2248) [2021-04-15 15:46:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][1060/1251] eta 0:00:54 lr 0.000987 time 0.2701 (0.2828) loss 4.6786 (4.4493) grad_norm 1.2636 (1.2248) [2021-04-15 15:46:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][1070/1251] eta 0:00:51 lr 0.000987 time 0.2666 (0.2827) loss 3.9603 (4.4477) grad_norm 1.0664 (1.2251) [2021-04-15 15:46:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][1080/1251] eta 0:00:48 lr 0.000987 time 0.3121 (0.2826) loss 5.0798 (4.4457) grad_norm 1.2011 (1.2252) [2021-04-15 15:46:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][1090/1251] eta 0:00:45 lr 0.000987 time 0.2661 (0.2826) loss 3.7801 (4.4489) grad_norm 1.6899 (1.2256) [2021-04-15 15:46:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][1100/1251] eta 0:00:42 lr 0.000987 time 0.2805 (0.2825) loss 4.6547 (4.4498) grad_norm 1.0133 (1.2262) [2021-04-15 15:46:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][1110/1251] eta 0:00:39 lr 0.000987 time 0.2565 (0.2825) loss 4.6748 (4.4503) grad_norm 1.4085 (1.2260) [2021-04-15 15:46:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][1120/1251] eta 0:00:36 lr 0.000987 time 0.2760 (0.2824) loss 3.9806 (4.4510) grad_norm 1.2475 (1.2255) [2021-04-15 15:46:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][1130/1251] eta 0:00:34 lr 0.000987 time 0.2833 (0.2823) loss 4.0795 (4.4495) grad_norm 1.0812 (1.2247) [2021-04-15 15:46:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][1140/1251] eta 0:00:31 lr 0.000987 time 0.2586 (0.2823) loss 3.5096 (4.4505) grad_norm 1.0195 (1.2246) [2021-04-15 15:46:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][1150/1251] eta 0:00:28 lr 0.000987 time 0.2624 (0.2822) loss 4.2864 (4.4497) grad_norm 1.0233 (1.2246) [2021-04-15 15:46:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][1160/1251] eta 0:00:25 lr 0.000987 time 0.2754 (0.2822) loss 4.7856 (4.4491) grad_norm 1.2465 (1.2240) [2021-04-15 15:46:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][1170/1251] eta 0:00:22 lr 0.000987 time 0.2843 (0.2821) loss 3.7073 (4.4486) grad_norm 1.0093 (1.2253) [2021-04-15 15:46:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][1180/1251] eta 0:00:20 lr 0.000987 time 0.2575 (0.2820) loss 4.7921 (4.4467) grad_norm 1.2526 (1.2255) [2021-04-15 15:46:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][1190/1251] eta 0:00:17 lr 0.000987 time 0.2946 (0.2820) loss 4.1296 (4.4461) grad_norm 1.2754 (1.2266) [2021-04-15 15:46:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][1200/1251] eta 0:00:14 lr 0.000987 time 0.2814 (0.2819) loss 3.7391 (4.4470) grad_norm 1.0768 (1.2267) [2021-04-15 15:46:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][1210/1251] eta 0:00:11 lr 0.000987 time 0.2657 (0.2819) loss 4.5609 (4.4475) grad_norm 1.1959 (1.2268) [2021-04-15 15:46:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][1220/1251] eta 0:00:08 lr 0.000987 time 0.2727 (0.2819) loss 3.3996 (4.4474) grad_norm 1.2700 (1.2278) [2021-04-15 15:46:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][1230/1251] eta 0:00:05 lr 0.000987 time 0.2540 (0.2818) loss 5.1301 (4.4492) grad_norm 1.0551 (1.2275) [2021-04-15 15:46:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][1240/1251] eta 0:00:03 lr 0.000987 time 0.2488 (0.2817) loss 4.7460 (4.4500) grad_norm 1.4211 (1.2275) [2021-04-15 15:46:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [21/300][1250/1251] eta 0:00:00 lr 0.000987 time 0.2495 (0.2815) loss 4.4802 (4.4520) grad_norm 0.9572 (1.2277) [2021-04-15 15:47:00 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 21 training takes 0:05:54 [2021-04-15 15:47:00 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_21.pth saving...... [2021-04-15 15:47:13 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_21.pth saved !!! [2021-04-15 15:47:14 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.146 (1.146) Loss 1.7400 (1.7400) Acc@1 60.645 (60.645) Acc@5 84.082 (84.082) [2021-04-15 15:47:15 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.228 (0.210) Loss 1.7069 (1.7446) Acc@1 61.035 (60.254) Acc@5 84.473 (83.709) [2021-04-15 15:47:18 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.194 (0.235) Loss 1.8143 (1.7398) Acc@1 59.473 (60.440) Acc@5 81.445 (83.747) [2021-04-15 15:47:20 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.119 (0.235) Loss 1.7423 (1.7375) Acc@1 60.254 (60.522) Acc@5 84.961 (83.742) [2021-04-15 15:47:22 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.219) Loss 1.7777 (1.7412) Acc@1 59.082 (60.552) Acc@5 84.375 (83.787) [2021-04-15 15:47:24 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 60.456 Acc@5 83.888 [2021-04-15 15:47:24 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 60.5% [2021-04-15 15:47:24 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 60.46% [2021-04-15 15:47:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][0/1251] eta 1:18:27 lr 0.000987 time 3.7630 (3.7630) loss 4.7097 (4.7097) grad_norm 1.5915 (1.5915) [2021-04-15 15:47:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][10/1251] eta 0:12:30 lr 0.000987 time 0.2842 (0.6050) loss 5.1604 (4.6324) grad_norm 1.2332 (1.1811) [2021-04-15 15:47:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][20/1251] eta 0:09:10 lr 0.000987 time 0.2764 (0.4474) loss 4.6960 (4.5226) grad_norm 1.2246 (1.1510) [2021-04-15 15:47:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][30/1251] eta 0:07:58 lr 0.000987 time 0.2735 (0.3920) loss 4.6254 (4.4999) grad_norm 1.3552 (1.1851) [2021-04-15 15:47:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][40/1251] eta 0:07:19 lr 0.000987 time 0.2706 (0.3633) loss 3.8861 (4.4858) grad_norm 1.3259 (1.1998) [2021-04-15 15:47:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][50/1251] eta 0:06:55 lr 0.000987 time 0.2622 (0.3461) loss 3.9816 (4.4211) grad_norm 1.4710 (1.2000) [2021-04-15 15:47:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][60/1251] eta 0:06:39 lr 0.000987 time 0.2640 (0.3352) loss 4.3082 (4.4214) grad_norm 1.1347 (1.1995) [2021-04-15 15:47:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][70/1251] eta 0:06:26 lr 0.000987 time 0.2703 (0.3271) loss 4.4810 (4.4362) grad_norm 1.1472 (1.1933) [2021-04-15 15:47:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][80/1251] eta 0:06:16 lr 0.000987 time 0.2828 (0.3211) loss 4.3584 (4.4342) grad_norm 1.2227 (inf) [2021-04-15 15:47:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][90/1251] eta 0:06:07 lr 0.000987 time 0.2777 (0.3164) loss 3.6312 (4.4492) grad_norm 1.0327 (inf) [2021-04-15 15:47:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][100/1251] eta 0:06:00 lr 0.000987 time 0.2786 (0.3130) loss 4.2007 (4.4455) grad_norm 1.2997 (inf) [2021-04-15 15:47:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][110/1251] eta 0:05:53 lr 0.000987 time 0.2758 (0.3096) loss 3.8718 (4.4425) grad_norm 1.8489 (inf) [2021-04-15 15:48:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][120/1251] eta 0:05:47 lr 0.000987 time 0.2816 (0.3069) loss 4.4713 (4.4388) grad_norm 1.2259 (inf) [2021-04-15 15:48:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][130/1251] eta 0:05:42 lr 0.000987 time 0.2568 (0.3052) loss 4.8428 (4.4419) grad_norm 1.2793 (inf) [2021-04-15 15:48:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][140/1251] eta 0:05:36 lr 0.000987 time 0.2743 (0.3032) loss 4.7512 (4.4334) grad_norm 1.0586 (inf) [2021-04-15 15:48:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][150/1251] eta 0:05:32 lr 0.000987 time 0.2667 (0.3017) loss 5.0829 (4.4216) grad_norm 1.1254 (inf) [2021-04-15 15:48:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][160/1251] eta 0:05:27 lr 0.000987 time 0.2561 (0.3004) loss 4.7272 (4.4355) grad_norm 1.0797 (inf) [2021-04-15 15:48:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][170/1251] eta 0:05:23 lr 0.000987 time 0.2755 (0.2993) loss 4.6739 (4.4447) grad_norm 1.1260 (inf) [2021-04-15 15:48:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][180/1251] eta 0:05:19 lr 0.000987 time 0.3000 (0.2980) loss 4.8995 (4.4595) grad_norm 1.2049 (inf) [2021-04-15 15:48:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][190/1251] eta 0:05:15 lr 0.000987 time 0.2738 (0.2969) loss 4.7468 (4.4482) grad_norm 1.2596 (inf) [2021-04-15 15:48:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][200/1251] eta 0:05:11 lr 0.000987 time 0.2697 (0.2959) loss 4.5091 (4.4432) grad_norm 1.2331 (inf) [2021-04-15 15:48:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][210/1251] eta 0:05:07 lr 0.000987 time 0.2926 (0.2951) loss 5.1411 (4.4704) grad_norm 1.1563 (inf) [2021-04-15 15:48:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][220/1251] eta 0:05:03 lr 0.000987 time 0.2865 (0.2945) loss 4.4380 (4.4578) grad_norm 1.3450 (inf) [2021-04-15 15:48:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][230/1251] eta 0:05:00 lr 0.000987 time 0.2665 (0.2939) loss 4.9533 (4.4617) grad_norm 0.9939 (inf) [2021-04-15 15:48:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][240/1251] eta 0:04:56 lr 0.000987 time 0.2908 (0.2931) loss 3.5249 (4.4519) grad_norm 1.3481 (inf) [2021-04-15 15:48:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][250/1251] eta 0:04:52 lr 0.000987 time 0.2775 (0.2925) loss 3.8237 (4.4357) grad_norm 1.3851 (inf) [2021-04-15 15:48:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][260/1251] eta 0:04:49 lr 0.000987 time 0.2845 (0.2921) loss 5.4258 (4.4387) grad_norm 1.1973 (inf) [2021-04-15 15:48:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][270/1251] eta 0:04:45 lr 0.000987 time 0.3022 (0.2914) loss 5.1684 (4.4443) grad_norm 1.4285 (inf) [2021-04-15 15:48:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][280/1251] eta 0:04:42 lr 0.000987 time 0.2802 (0.2909) loss 3.7687 (4.4413) grad_norm 0.9850 (inf) [2021-04-15 15:48:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][290/1251] eta 0:04:39 lr 0.000987 time 0.2944 (0.2906) loss 3.5003 (4.4316) grad_norm 1.0908 (inf) [2021-04-15 15:48:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][300/1251] eta 0:04:35 lr 0.000987 time 0.3011 (0.2902) loss 4.1817 (4.4315) grad_norm 1.1559 (inf) [2021-04-15 15:48:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][310/1251] eta 0:04:32 lr 0.000987 time 0.2662 (0.2897) loss 4.3587 (4.4311) grad_norm 1.0968 (inf) [2021-04-15 15:48:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][320/1251] eta 0:04:29 lr 0.000987 time 0.2841 (0.2892) loss 4.4601 (4.4355) grad_norm 1.2737 (inf) [2021-04-15 15:49:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][330/1251] eta 0:04:25 lr 0.000987 time 0.2651 (0.2888) loss 3.8601 (4.4323) grad_norm 1.1328 (inf) [2021-04-15 15:49:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][340/1251] eta 0:04:22 lr 0.000987 time 0.2839 (0.2886) loss 3.9143 (4.4296) grad_norm 1.0044 (inf) [2021-04-15 15:49:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][350/1251] eta 0:04:19 lr 0.000987 time 0.2841 (0.2882) loss 5.2908 (4.4401) grad_norm 1.1111 (inf) [2021-04-15 15:49:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][360/1251] eta 0:04:16 lr 0.000987 time 0.2570 (0.2878) loss 3.8247 (4.4330) grad_norm 1.5033 (inf) [2021-04-15 15:49:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][370/1251] eta 0:04:13 lr 0.000987 time 0.2994 (0.2877) loss 4.6551 (4.4321) grad_norm 1.1172 (inf) [2021-04-15 15:49:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][380/1251] eta 0:04:10 lr 0.000987 time 0.2798 (0.2874) loss 3.5593 (4.4279) grad_norm 1.1084 (inf) [2021-04-15 15:49:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][390/1251] eta 0:04:07 lr 0.000987 time 0.2906 (0.2872) loss 4.4789 (4.4318) grad_norm 1.4578 (inf) [2021-04-15 15:49:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][400/1251] eta 0:04:04 lr 0.000987 time 0.2714 (0.2874) loss 4.5114 (4.4328) grad_norm 1.0803 (inf) [2021-04-15 15:49:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][410/1251] eta 0:04:01 lr 0.000987 time 0.2754 (0.2871) loss 4.5348 (4.4348) grad_norm 1.1060 (inf) [2021-04-15 15:49:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][420/1251] eta 0:03:58 lr 0.000987 time 0.2715 (0.2870) loss 4.5123 (4.4252) grad_norm 1.1879 (inf) [2021-04-15 15:49:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][430/1251] eta 0:03:55 lr 0.000987 time 0.2805 (0.2868) loss 5.1204 (4.4229) grad_norm 1.2918 (inf) [2021-04-15 15:49:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][440/1251] eta 0:03:52 lr 0.000987 time 0.2810 (0.2865) loss 4.9465 (4.4252) grad_norm 1.0697 (inf) [2021-04-15 15:49:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][450/1251] eta 0:03:49 lr 0.000986 time 0.2769 (0.2863) loss 4.7353 (4.4225) grad_norm 1.0865 (inf) [2021-04-15 15:49:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][460/1251] eta 0:03:46 lr 0.000986 time 0.2790 (0.2862) loss 4.5784 (4.4291) grad_norm 1.1058 (inf) [2021-04-15 15:49:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][470/1251] eta 0:03:43 lr 0.000986 time 0.2548 (0.2859) loss 5.1652 (4.4321) grad_norm 1.2001 (inf) [2021-04-15 15:49:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][480/1251] eta 0:03:40 lr 0.000986 time 0.2789 (0.2857) loss 4.7648 (4.4270) grad_norm 1.2550 (inf) [2021-04-15 15:49:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][490/1251] eta 0:03:37 lr 0.000986 time 0.2911 (0.2856) loss 3.8769 (4.4222) grad_norm 1.1912 (inf) [2021-04-15 15:49:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][500/1251] eta 0:03:34 lr 0.000986 time 0.2650 (0.2854) loss 4.2453 (4.4258) grad_norm 1.4111 (inf) [2021-04-15 15:49:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][510/1251] eta 0:03:31 lr 0.000986 time 0.2822 (0.2854) loss 5.1323 (4.4308) grad_norm 1.2044 (inf) [2021-04-15 15:49:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][520/1251] eta 0:03:28 lr 0.000986 time 0.2627 (0.2852) loss 5.3406 (4.4314) grad_norm 0.8782 (inf) [2021-04-15 15:49:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][530/1251] eta 0:03:25 lr 0.000986 time 0.2828 (0.2849) loss 3.8617 (4.4320) grad_norm 1.0684 (inf) [2021-04-15 15:49:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][540/1251] eta 0:03:22 lr 0.000986 time 0.2712 (0.2851) loss 3.6796 (4.4260) grad_norm 1.3232 (inf) [2021-04-15 15:50:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][550/1251] eta 0:03:19 lr 0.000986 time 0.2856 (0.2849) loss 4.0508 (4.4276) grad_norm 1.0991 (inf) [2021-04-15 15:50:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][560/1251] eta 0:03:16 lr 0.000986 time 0.2706 (0.2848) loss 4.5109 (4.4229) grad_norm 1.4413 (inf) [2021-04-15 15:50:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][570/1251] eta 0:03:13 lr 0.000986 time 0.2667 (0.2847) loss 3.6183 (4.4203) grad_norm 1.1406 (inf) [2021-04-15 15:50:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][580/1251] eta 0:03:10 lr 0.000986 time 0.2749 (0.2845) loss 4.9016 (4.4180) grad_norm 1.1559 (inf) [2021-04-15 15:50:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][590/1251] eta 0:03:07 lr 0.000986 time 0.2988 (0.2844) loss 4.9437 (4.4169) grad_norm 1.1368 (inf) [2021-04-15 15:50:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][600/1251] eta 0:03:05 lr 0.000986 time 0.2838 (0.2842) loss 4.6093 (4.4175) grad_norm 1.3312 (inf) [2021-04-15 15:50:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][610/1251] eta 0:03:02 lr 0.000986 time 0.2567 (0.2841) loss 4.9088 (4.4249) grad_norm 0.8277 (inf) [2021-04-15 15:50:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][620/1251] eta 0:02:59 lr 0.000986 time 0.2779 (0.2841) loss 4.7526 (4.4214) grad_norm 1.1588 (inf) [2021-04-15 15:50:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][630/1251] eta 0:02:56 lr 0.000986 time 0.2587 (0.2839) loss 4.0697 (4.4229) grad_norm 1.3708 (inf) [2021-04-15 15:50:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][640/1251] eta 0:02:53 lr 0.000986 time 0.2854 (0.2838) loss 3.9626 (4.4265) grad_norm 1.0786 (inf) [2021-04-15 15:50:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][650/1251] eta 0:02:50 lr 0.000986 time 0.2579 (0.2837) loss 4.7768 (4.4269) grad_norm 1.3605 (inf) [2021-04-15 15:50:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][660/1251] eta 0:02:47 lr 0.000986 time 0.2678 (0.2836) loss 4.7697 (4.4307) grad_norm 1.2344 (inf) [2021-04-15 15:50:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][670/1251] eta 0:02:44 lr 0.000986 time 0.2576 (0.2835) loss 4.1809 (4.4348) grad_norm 1.3634 (inf) [2021-04-15 15:50:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][680/1251] eta 0:02:41 lr 0.000986 time 0.2630 (0.2834) loss 4.6956 (4.4349) grad_norm 1.2546 (inf) [2021-04-15 15:50:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][690/1251] eta 0:02:38 lr 0.000986 time 0.2736 (0.2834) loss 5.3086 (4.4379) grad_norm 1.2052 (inf) [2021-04-15 15:50:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][700/1251] eta 0:02:36 lr 0.000986 time 0.2662 (0.2833) loss 4.4602 (4.4336) grad_norm 1.4569 (inf) [2021-04-15 15:50:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][710/1251] eta 0:02:33 lr 0.000986 time 0.2564 (0.2832) loss 3.2887 (4.4359) grad_norm 1.1470 (inf) [2021-04-15 15:50:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][720/1251] eta 0:02:30 lr 0.000986 time 0.2859 (0.2831) loss 3.4805 (4.4386) grad_norm 0.9865 (inf) [2021-04-15 15:50:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][730/1251] eta 0:02:27 lr 0.000986 time 0.3008 (0.2831) loss 4.7805 (4.4432) grad_norm 1.0809 (inf) [2021-04-15 15:50:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][740/1251] eta 0:02:24 lr 0.000986 time 0.2782 (0.2830) loss 4.0515 (4.4438) grad_norm 1.1128 (inf) [2021-04-15 15:50:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][750/1251] eta 0:02:21 lr 0.000986 time 0.2649 (0.2830) loss 3.2998 (4.4360) grad_norm 1.3815 (inf) [2021-04-15 15:50:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][760/1251] eta 0:02:18 lr 0.000986 time 0.2559 (0.2828) loss 4.2225 (4.4310) grad_norm 1.2587 (inf) [2021-04-15 15:51:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][770/1251] eta 0:02:16 lr 0.000986 time 0.2728 (0.2828) loss 5.2342 (4.4337) grad_norm 1.7646 (inf) [2021-04-15 15:51:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][780/1251] eta 0:02:13 lr 0.000986 time 0.2715 (0.2827) loss 4.3782 (4.4369) grad_norm 1.3049 (inf) [2021-04-15 15:51:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][790/1251] eta 0:02:10 lr 0.000986 time 0.2822 (0.2827) loss 4.2276 (4.4367) grad_norm 1.3532 (inf) [2021-04-15 15:51:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][800/1251] eta 0:02:07 lr 0.000986 time 0.2592 (0.2826) loss 5.0218 (4.4379) grad_norm 0.9483 (inf) [2021-04-15 15:51:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][810/1251] eta 0:02:04 lr 0.000986 time 0.3007 (0.2825) loss 3.4017 (4.4357) grad_norm 1.4067 (inf) [2021-04-15 15:51:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][820/1251] eta 0:02:01 lr 0.000986 time 0.2901 (0.2825) loss 3.7577 (4.4354) grad_norm 1.1807 (inf) [2021-04-15 15:51:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][830/1251] eta 0:01:58 lr 0.000986 time 0.3114 (0.2824) loss 4.0438 (4.4354) grad_norm 1.1551 (inf) [2021-04-15 15:51:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][840/1251] eta 0:01:56 lr 0.000986 time 0.2746 (0.2824) loss 3.7658 (4.4319) grad_norm 1.0065 (inf) [2021-04-15 15:51:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][850/1251] eta 0:01:53 lr 0.000986 time 0.2773 (0.2823) loss 4.4170 (4.4325) grad_norm 1.0517 (inf) [2021-04-15 15:51:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][860/1251] eta 0:01:50 lr 0.000986 time 0.2778 (0.2823) loss 4.4929 (4.4311) grad_norm 1.3869 (inf) [2021-04-15 15:51:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][870/1251] eta 0:01:47 lr 0.000986 time 0.2934 (0.2822) loss 3.3500 (4.4320) grad_norm 1.0583 (inf) [2021-04-15 15:51:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][880/1251] eta 0:01:44 lr 0.000986 time 0.2821 (0.2821) loss 4.6631 (4.4309) grad_norm 1.6425 (inf) [2021-04-15 15:51:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][890/1251] eta 0:01:41 lr 0.000986 time 0.2867 (0.2821) loss 4.7630 (4.4283) grad_norm 1.0238 (inf) [2021-04-15 15:51:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][900/1251] eta 0:01:38 lr 0.000986 time 0.2663 (0.2820) loss 3.9916 (4.4305) grad_norm 1.3495 (inf) [2021-04-15 15:51:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][910/1251] eta 0:01:36 lr 0.000986 time 0.2823 (0.2820) loss 5.0506 (4.4324) grad_norm 1.0933 (inf) [2021-04-15 15:51:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][920/1251] eta 0:01:33 lr 0.000986 time 0.2993 (0.2819) loss 4.1392 (4.4352) grad_norm 1.1359 (inf) [2021-04-15 15:51:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][930/1251] eta 0:01:30 lr 0.000986 time 0.2616 (0.2819) loss 4.7711 (4.4371) grad_norm 1.2792 (inf) [2021-04-15 15:51:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][940/1251] eta 0:01:27 lr 0.000986 time 0.2954 (0.2819) loss 4.9026 (4.4372) grad_norm 1.1845 (inf) [2021-04-15 15:51:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][950/1251] eta 0:01:24 lr 0.000986 time 0.2805 (0.2818) loss 4.3837 (4.4396) grad_norm 1.0430 (inf) [2021-04-15 15:51:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][960/1251] eta 0:01:21 lr 0.000986 time 0.2674 (0.2817) loss 4.6488 (4.4389) grad_norm 1.1516 (inf) [2021-04-15 15:51:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][970/1251] eta 0:01:19 lr 0.000986 time 0.2860 (0.2817) loss 3.9982 (4.4359) grad_norm 1.0487 (inf) [2021-04-15 15:52:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][980/1251] eta 0:01:16 lr 0.000986 time 0.2632 (0.2817) loss 4.5994 (4.4361) grad_norm 1.0361 (inf) [2021-04-15 15:52:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][990/1251] eta 0:01:13 lr 0.000986 time 0.2581 (0.2817) loss 4.9236 (4.4370) grad_norm 1.0624 (inf) [2021-04-15 15:52:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][1000/1251] eta 0:01:10 lr 0.000986 time 0.2717 (0.2817) loss 4.8569 (4.4344) grad_norm 1.3555 (inf) [2021-04-15 15:52:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][1010/1251] eta 0:01:07 lr 0.000986 time 0.2575 (0.2817) loss 5.1577 (4.4326) grad_norm 1.1465 (inf) [2021-04-15 15:52:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][1020/1251] eta 0:01:05 lr 0.000986 time 0.2934 (0.2817) loss 4.2904 (4.4308) grad_norm 1.9606 (inf) [2021-04-15 15:52:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][1030/1251] eta 0:01:02 lr 0.000986 time 0.2638 (0.2816) loss 4.2401 (4.4308) grad_norm 1.3466 (inf) [2021-04-15 15:52:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][1040/1251] eta 0:00:59 lr 0.000986 time 0.2712 (0.2816) loss 4.2221 (4.4310) grad_norm 1.2152 (inf) [2021-04-15 15:52:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][1050/1251] eta 0:00:56 lr 0.000986 time 0.2762 (0.2815) loss 3.8923 (4.4310) grad_norm 1.2323 (inf) [2021-04-15 15:52:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][1060/1251] eta 0:00:53 lr 0.000986 time 0.2808 (0.2815) loss 3.4779 (4.4283) grad_norm 1.8737 (inf) [2021-04-15 15:52:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][1070/1251] eta 0:00:50 lr 0.000986 time 0.2906 (0.2815) loss 4.4839 (4.4297) grad_norm 1.1657 (inf) [2021-04-15 15:52:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][1080/1251] eta 0:00:48 lr 0.000986 time 0.2693 (0.2814) loss 4.1144 (4.4284) grad_norm 1.5214 (inf) [2021-04-15 15:52:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][1090/1251] eta 0:00:45 lr 0.000986 time 0.2774 (0.2814) loss 4.4365 (4.4272) grad_norm 1.0046 (inf) [2021-04-15 15:52:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][1100/1251] eta 0:00:42 lr 0.000986 time 0.2778 (0.2813) loss 4.4276 (4.4243) grad_norm 1.0559 (inf) [2021-04-15 15:52:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][1110/1251] eta 0:00:39 lr 0.000986 time 0.2803 (0.2813) loss 3.8661 (4.4243) grad_norm 1.2071 (inf) [2021-04-15 15:52:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][1120/1251] eta 0:00:36 lr 0.000986 time 0.3135 (0.2815) loss 4.5171 (4.4243) grad_norm 0.9762 (inf) [2021-04-15 15:52:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][1130/1251] eta 0:00:34 lr 0.000986 time 0.2638 (0.2814) loss 4.2221 (4.4241) grad_norm 1.1835 (inf) [2021-04-15 15:52:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][1140/1251] eta 0:00:31 lr 0.000986 time 0.2810 (0.2814) loss 3.7653 (4.4230) grad_norm 1.2905 (inf) [2021-04-15 15:52:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][1150/1251] eta 0:00:28 lr 0.000986 time 0.2709 (0.2813) loss 4.0578 (4.4217) grad_norm 1.1417 (inf) [2021-04-15 15:52:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][1160/1251] eta 0:00:25 lr 0.000986 time 0.2912 (0.2813) loss 4.5963 (4.4240) grad_norm 1.2284 (inf) [2021-04-15 15:52:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][1170/1251] eta 0:00:22 lr 0.000986 time 0.2911 (0.2814) loss 4.9494 (4.4232) grad_norm 1.0610 (inf) [2021-04-15 15:52:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][1180/1251] eta 0:00:19 lr 0.000986 time 0.2946 (0.2813) loss 3.6465 (4.4223) grad_norm 1.0641 (inf) [2021-04-15 15:52:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][1190/1251] eta 0:00:17 lr 0.000986 time 0.2647 (0.2813) loss 3.5868 (4.4228) grad_norm 1.1796 (inf) [2021-04-15 15:53:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][1200/1251] eta 0:00:14 lr 0.000986 time 0.2535 (0.2812) loss 3.1390 (4.4209) grad_norm 1.3856 (inf) [2021-04-15 15:53:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][1210/1251] eta 0:00:11 lr 0.000986 time 0.2768 (0.2812) loss 4.9759 (4.4209) grad_norm 1.2403 (inf) [2021-04-15 15:53:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][1220/1251] eta 0:00:08 lr 0.000986 time 0.2900 (0.2812) loss 3.1298 (4.4199) grad_norm 1.2871 (inf) [2021-04-15 15:53:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][1230/1251] eta 0:00:05 lr 0.000986 time 0.2418 (0.2811) loss 4.3767 (4.4195) grad_norm 1.1379 (inf) [2021-04-15 15:53:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][1240/1251] eta 0:00:03 lr 0.000986 time 0.2487 (0.2810) loss 4.0998 (4.4180) grad_norm 0.9804 (inf) [2021-04-15 15:53:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [22/300][1250/1251] eta 0:00:00 lr 0.000986 time 0.2488 (0.2808) loss 4.0613 (4.4195) grad_norm 1.4136 (inf) [2021-04-15 15:53:17 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 22 training takes 0:05:53 [2021-04-15 15:53:17 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_22.pth saving...... [2021-04-15 15:53:34 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_22.pth saved !!! [2021-04-15 15:53:35 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.151 (1.151) Loss 1.6494 (1.6494) Acc@1 62.500 (62.500) Acc@5 84.766 (84.766) [2021-04-15 15:53:36 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.125 (0.239) Loss 1.8298 (1.7452) Acc@1 59.375 (60.289) Acc@5 84.180 (83.949) [2021-04-15 15:53:39 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.126 (0.236) Loss 1.7959 (1.7393) Acc@1 61.426 (60.617) Acc@5 83.789 (83.947) [2021-04-15 15:53:41 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.134 (0.237) Loss 1.7446 (1.7259) Acc@1 60.254 (60.934) Acc@5 84.277 (84.230) [2021-04-15 15:53:43 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.075 (0.216) Loss 1.6605 (1.7272) Acc@1 62.305 (60.973) Acc@5 86.426 (84.184) [2021-04-15 15:53:45 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 61.086 Acc@5 84.350 [2021-04-15 15:53:45 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 61.1% [2021-04-15 15:53:45 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 61.09% [2021-04-15 15:53:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][0/1251] eta 1:14:46 lr 0.000986 time 3.5863 (3.5863) loss 4.7527 (4.7527) grad_norm 1.0801 (1.0801) [2021-04-15 15:53:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][10/1251] eta 0:12:32 lr 0.000986 time 0.2941 (0.6063) loss 3.9319 (4.6405) grad_norm 1.1726 (1.1915) [2021-04-15 15:53:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][20/1251] eta 0:09:16 lr 0.000986 time 0.2838 (0.4524) loss 5.3308 (4.6781) grad_norm 0.9630 (1.1588) [2021-04-15 15:53:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][30/1251] eta 0:08:03 lr 0.000986 time 0.2886 (0.3960) loss 5.1428 (4.6462) grad_norm 1.1744 (1.1455) [2021-04-15 15:54:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][40/1251] eta 0:07:28 lr 0.000986 time 0.3184 (0.3700) loss 4.2834 (4.5006) grad_norm 1.3886 (1.2075) [2021-04-15 15:54:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][50/1251] eta 0:07:03 lr 0.000986 time 0.2885 (0.3526) loss 5.1818 (4.5424) grad_norm 1.3310 (1.2090) [2021-04-15 15:54:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][60/1251] eta 0:06:45 lr 0.000986 time 0.2752 (0.3406) loss 3.8345 (4.5203) grad_norm 0.9449 (1.2042) [2021-04-15 15:54:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][70/1251] eta 0:06:34 lr 0.000986 time 0.4006 (0.3342) loss 4.2354 (4.5167) grad_norm 1.2627 (1.2173) [2021-04-15 15:54:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][80/1251] eta 0:06:24 lr 0.000986 time 0.2771 (0.3284) loss 3.5400 (4.4994) grad_norm 1.3338 (1.2081) [2021-04-15 15:54:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][90/1251] eta 0:06:14 lr 0.000986 time 0.2710 (0.3230) loss 4.1623 (4.5022) grad_norm 1.1415 (1.2107) [2021-04-15 15:54:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][100/1251] eta 0:06:06 lr 0.000986 time 0.2752 (0.3188) loss 3.5328 (4.4682) grad_norm 1.4081 (1.2042) [2021-04-15 15:54:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][110/1251] eta 0:05:59 lr 0.000986 time 0.2578 (0.3148) loss 4.7027 (4.4873) grad_norm 1.3289 (1.1988) [2021-04-15 15:54:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][120/1251] eta 0:05:52 lr 0.000986 time 0.2640 (0.3118) loss 4.1190 (4.4865) grad_norm 1.0911 (1.1901) [2021-04-15 15:54:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][130/1251] eta 0:05:47 lr 0.000986 time 0.2922 (0.3096) loss 4.5020 (4.4741) grad_norm 0.9231 (1.1849) [2021-04-15 15:54:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][140/1251] eta 0:05:41 lr 0.000986 time 0.3006 (0.3072) loss 2.9567 (4.4432) grad_norm 1.1486 (1.1748) [2021-04-15 15:54:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][150/1251] eta 0:05:36 lr 0.000986 time 0.2926 (0.3056) loss 3.7609 (4.4288) grad_norm 0.9559 (1.1737) [2021-04-15 15:54:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][160/1251] eta 0:05:31 lr 0.000986 time 0.2755 (0.3038) loss 5.1922 (4.4357) grad_norm 1.4203 (1.1792) [2021-04-15 15:54:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][170/1251] eta 0:05:26 lr 0.000986 time 0.2748 (0.3023) loss 4.7308 (4.4516) grad_norm 1.0852 (1.1786) [2021-04-15 15:54:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][180/1251] eta 0:05:23 lr 0.000986 time 0.4539 (0.3020) loss 4.6732 (4.4391) grad_norm 1.2734 (1.1848) [2021-04-15 15:54:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][190/1251] eta 0:05:18 lr 0.000986 time 0.2822 (0.3003) loss 5.2396 (4.4477) grad_norm 1.0699 (1.1922) [2021-04-15 15:54:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][200/1251] eta 0:05:14 lr 0.000986 time 0.2652 (0.2992) loss 3.2651 (4.4514) grad_norm 1.4306 (1.1898) [2021-04-15 15:54:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][210/1251] eta 0:05:10 lr 0.000986 time 0.2649 (0.2982) loss 3.6517 (4.4518) grad_norm 1.0915 (1.1868) [2021-04-15 15:54:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][220/1251] eta 0:05:06 lr 0.000985 time 0.2704 (0.2972) loss 3.8812 (4.4351) grad_norm 1.2344 (1.1877) [2021-04-15 15:54:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][230/1251] eta 0:05:02 lr 0.000985 time 0.2586 (0.2963) loss 2.9681 (4.4271) grad_norm 1.0698 (1.1888) [2021-04-15 15:54:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][240/1251] eta 0:04:58 lr 0.000985 time 0.2667 (0.2957) loss 5.1370 (4.4382) grad_norm 1.1757 (1.1915) [2021-04-15 15:54:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][250/1251] eta 0:04:55 lr 0.000985 time 0.2942 (0.2952) loss 5.0981 (4.4514) grad_norm 1.1830 (1.1953) [2021-04-15 15:55:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][260/1251] eta 0:04:52 lr 0.000985 time 0.2797 (0.2947) loss 4.5888 (4.4574) grad_norm 1.0195 (1.1960) [2021-04-15 15:55:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][270/1251] eta 0:04:48 lr 0.000985 time 0.2525 (0.2939) loss 5.3268 (4.4451) grad_norm 1.0533 (1.1987) [2021-04-15 15:55:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][280/1251] eta 0:04:44 lr 0.000985 time 0.2697 (0.2935) loss 5.2234 (4.4454) grad_norm 1.0580 (1.1981) [2021-04-15 15:55:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][290/1251] eta 0:04:41 lr 0.000985 time 0.2740 (0.2931) loss 5.2912 (4.4440) grad_norm 1.2468 (1.1977) [2021-04-15 15:55:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][300/1251] eta 0:04:38 lr 0.000985 time 0.2990 (0.2928) loss 3.6103 (4.4493) grad_norm 1.0960 (1.1971) [2021-04-15 15:55:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][310/1251] eta 0:04:35 lr 0.000985 time 0.2874 (0.2924) loss 4.2537 (4.4396) grad_norm 1.1409 (1.1938) [2021-04-15 15:55:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][320/1251] eta 0:04:32 lr 0.000985 time 0.2653 (0.2923) loss 4.5227 (4.4397) grad_norm 1.3698 (1.1939) [2021-04-15 15:55:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][330/1251] eta 0:04:29 lr 0.000985 time 0.2450 (0.2923) loss 4.8826 (4.4422) grad_norm 1.0094 (1.1930) [2021-04-15 15:55:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][340/1251] eta 0:04:26 lr 0.000985 time 0.3047 (0.2921) loss 3.9191 (4.4460) grad_norm 1.0456 (1.1939) [2021-04-15 15:55:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][350/1251] eta 0:04:22 lr 0.000985 time 0.2886 (0.2919) loss 4.5091 (4.4484) grad_norm 1.0062 (1.1925) [2021-04-15 15:55:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][360/1251] eta 0:04:19 lr 0.000985 time 0.2920 (0.2915) loss 4.6077 (4.4526) grad_norm 1.1185 (1.1907) [2021-04-15 15:55:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][370/1251] eta 0:04:16 lr 0.000985 time 0.2798 (0.2911) loss 3.6499 (4.4426) grad_norm 1.1580 (1.1895) [2021-04-15 15:55:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][380/1251] eta 0:04:13 lr 0.000985 time 0.2692 (0.2907) loss 4.8621 (4.4393) grad_norm 1.1171 (1.1874) [2021-04-15 15:55:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][390/1251] eta 0:04:09 lr 0.000985 time 0.2945 (0.2903) loss 3.8916 (4.4386) grad_norm 1.6525 (1.1868) [2021-04-15 15:55:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][400/1251] eta 0:04:06 lr 0.000985 time 0.2933 (0.2901) loss 4.7663 (4.4332) grad_norm 1.3789 (1.1880) [2021-04-15 15:55:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][410/1251] eta 0:04:03 lr 0.000985 time 0.2768 (0.2900) loss 4.0730 (4.4313) grad_norm 1.3164 (1.1867) [2021-04-15 15:55:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][420/1251] eta 0:04:00 lr 0.000985 time 0.2788 (0.2896) loss 3.7508 (4.4229) grad_norm 1.1942 (1.1909) [2021-04-15 15:55:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][430/1251] eta 0:03:57 lr 0.000985 time 0.2894 (0.2894) loss 3.4126 (4.4206) grad_norm 1.1917 (1.1955) [2021-04-15 15:55:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][440/1251] eta 0:03:54 lr 0.000985 time 0.2815 (0.2891) loss 4.3720 (4.4243) grad_norm 1.1496 (1.1935) [2021-04-15 15:55:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][450/1251] eta 0:03:51 lr 0.000985 time 0.2921 (0.2890) loss 4.5955 (4.4233) grad_norm 1.3708 (1.1946) [2021-04-15 15:55:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][460/1251] eta 0:03:48 lr 0.000985 time 0.2606 (0.2887) loss 4.7732 (4.4203) grad_norm 1.2683 (1.1942) [2021-04-15 15:56:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][470/1251] eta 0:03:45 lr 0.000985 time 0.2683 (0.2885) loss 4.3389 (4.4260) grad_norm 1.2480 (1.1920) [2021-04-15 15:56:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][480/1251] eta 0:03:42 lr 0.000985 time 0.2858 (0.2882) loss 3.6427 (4.4212) grad_norm 1.1282 (1.1914) [2021-04-15 15:56:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][490/1251] eta 0:03:39 lr 0.000985 time 0.2804 (0.2881) loss 5.1532 (4.4221) grad_norm 1.5249 (1.1912) [2021-04-15 15:56:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][500/1251] eta 0:03:36 lr 0.000985 time 0.2958 (0.2879) loss 4.5804 (4.4252) grad_norm 0.9900 (1.1911) [2021-04-15 15:56:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][510/1251] eta 0:03:33 lr 0.000985 time 0.2857 (0.2876) loss 4.7708 (4.4239) grad_norm 1.6306 (1.1912) [2021-04-15 15:56:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][520/1251] eta 0:03:30 lr 0.000985 time 0.2703 (0.2874) loss 5.0968 (4.4250) grad_norm 1.3281 (1.1911) [2021-04-15 15:56:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][530/1251] eta 0:03:27 lr 0.000985 time 0.2643 (0.2873) loss 3.0760 (4.4172) grad_norm 1.1082 (1.1906) [2021-04-15 15:56:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][540/1251] eta 0:03:24 lr 0.000985 time 0.2764 (0.2870) loss 4.7888 (4.4201) grad_norm 1.1800 (1.1906) [2021-04-15 15:56:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][550/1251] eta 0:03:21 lr 0.000985 time 0.2761 (0.2869) loss 3.3110 (4.4177) grad_norm 1.0657 (1.1903) [2021-04-15 15:56:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][560/1251] eta 0:03:18 lr 0.000985 time 0.2598 (0.2869) loss 4.9863 (4.4091) grad_norm 1.2320 (1.1903) [2021-04-15 15:56:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][570/1251] eta 0:03:15 lr 0.000985 time 0.2757 (0.2867) loss 5.4711 (4.4116) grad_norm 1.2072 (1.1887) [2021-04-15 15:56:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][580/1251] eta 0:03:12 lr 0.000985 time 0.3170 (0.2867) loss 3.3751 (4.4112) grad_norm 1.2709 (1.1894) [2021-04-15 15:56:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][590/1251] eta 0:03:09 lr 0.000985 time 0.2696 (0.2864) loss 4.3534 (4.4061) grad_norm 1.3750 (1.1911) [2021-04-15 15:56:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][600/1251] eta 0:03:06 lr 0.000985 time 0.2736 (0.2863) loss 4.9732 (4.4059) grad_norm 1.1383 (1.1919) [2021-04-15 15:56:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][610/1251] eta 0:03:03 lr 0.000985 time 0.2564 (0.2861) loss 4.6474 (4.4055) grad_norm 1.2538 (1.1915) [2021-04-15 15:56:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][620/1251] eta 0:03:00 lr 0.000985 time 0.2593 (0.2860) loss 5.2381 (4.4033) grad_norm 1.3394 (1.1918) [2021-04-15 15:56:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][630/1251] eta 0:02:57 lr 0.000985 time 0.2627 (0.2860) loss 3.8447 (4.4087) grad_norm 1.1310 (1.1907) [2021-04-15 15:56:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][640/1251] eta 0:02:54 lr 0.000985 time 0.3092 (0.2858) loss 4.6156 (4.4070) grad_norm 1.2059 (1.1911) [2021-04-15 15:56:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][650/1251] eta 0:02:51 lr 0.000985 time 0.2492 (0.2858) loss 4.2229 (4.4028) grad_norm 1.3825 (1.1910) [2021-04-15 15:56:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][660/1251] eta 0:02:48 lr 0.000985 time 0.2883 (0.2857) loss 4.5159 (4.4024) grad_norm 0.9870 (1.1904) [2021-04-15 15:56:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][670/1251] eta 0:02:45 lr 0.000985 time 0.2630 (0.2855) loss 4.7270 (4.4027) grad_norm 0.9606 (1.1892) [2021-04-15 15:56:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][680/1251] eta 0:02:42 lr 0.000985 time 0.2716 (0.2854) loss 4.7193 (4.4076) grad_norm 1.4094 (1.1915) [2021-04-15 15:57:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][690/1251] eta 0:02:40 lr 0.000985 time 0.2701 (0.2853) loss 4.4241 (4.4086) grad_norm 1.1015 (1.1904) [2021-04-15 15:57:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][700/1251] eta 0:02:37 lr 0.000985 time 0.2952 (0.2851) loss 3.9746 (4.4092) grad_norm 1.2670 (1.1897) [2021-04-15 15:57:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][710/1251] eta 0:02:34 lr 0.000985 time 0.2676 (0.2850) loss 2.8973 (4.4053) grad_norm 1.1587 (1.1893) [2021-04-15 15:57:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][720/1251] eta 0:02:31 lr 0.000985 time 0.2795 (0.2849) loss 4.2993 (4.4042) grad_norm 0.9527 (1.1898) [2021-04-15 15:57:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][730/1251] eta 0:02:28 lr 0.000985 time 0.2730 (0.2848) loss 5.2801 (4.4071) grad_norm 1.2644 (1.1896) [2021-04-15 15:57:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][740/1251] eta 0:02:25 lr 0.000985 time 0.2725 (0.2847) loss 4.1621 (4.4081) grad_norm 1.2978 (1.1892) [2021-04-15 15:57:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][750/1251] eta 0:02:22 lr 0.000985 time 0.2918 (0.2846) loss 3.2323 (4.4078) grad_norm 1.2729 (1.1894) [2021-04-15 15:57:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][760/1251] eta 0:02:19 lr 0.000985 time 0.2777 (0.2846) loss 4.3323 (4.4057) grad_norm 1.3533 (1.1888) [2021-04-15 15:57:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][770/1251] eta 0:02:16 lr 0.000985 time 0.2801 (0.2845) loss 3.9889 (4.4038) grad_norm 1.3674 (1.1889) [2021-04-15 15:57:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][780/1251] eta 0:02:13 lr 0.000985 time 0.2584 (0.2844) loss 4.7364 (4.4058) grad_norm 1.1364 (1.1899) [2021-04-15 15:57:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][790/1251] eta 0:02:11 lr 0.000985 time 0.2988 (0.2844) loss 3.7254 (4.4016) grad_norm 1.0273 (1.1902) [2021-04-15 15:57:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][800/1251] eta 0:02:08 lr 0.000985 time 0.2598 (0.2843) loss 4.3145 (4.4018) grad_norm 1.0806 (1.1895) [2021-04-15 15:57:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][810/1251] eta 0:02:05 lr 0.000985 time 0.2562 (0.2842) loss 4.8308 (4.4025) grad_norm 1.1954 (1.1890) [2021-04-15 15:57:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][820/1251] eta 0:02:02 lr 0.000985 time 0.2748 (0.2841) loss 4.0493 (4.3990) grad_norm 0.9827 (1.1880) [2021-04-15 15:57:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][830/1251] eta 0:01:59 lr 0.000985 time 0.3016 (0.2841) loss 4.8206 (4.4016) grad_norm 1.4855 (1.1879) [2021-04-15 15:57:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][840/1251] eta 0:01:56 lr 0.000985 time 0.2810 (0.2840) loss 4.5221 (4.4018) grad_norm 1.0037 (1.1887) [2021-04-15 15:57:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][850/1251] eta 0:01:53 lr 0.000985 time 0.2684 (0.2839) loss 4.7342 (4.4033) grad_norm 1.0943 (1.1886) [2021-04-15 15:57:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][860/1251] eta 0:01:51 lr 0.000985 time 0.2809 (0.2839) loss 3.8380 (4.4052) grad_norm 1.0150 (1.1877) [2021-04-15 15:57:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][870/1251] eta 0:01:48 lr 0.000985 time 0.2862 (0.2838) loss 5.0861 (4.4047) grad_norm 1.2843 (1.1876) [2021-04-15 15:57:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][880/1251] eta 0:01:45 lr 0.000985 time 0.2719 (0.2837) loss 4.7632 (4.4058) grad_norm 1.1622 (1.1874) [2021-04-15 15:57:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][890/1251] eta 0:01:42 lr 0.000985 time 0.2494 (0.2838) loss 4.1507 (4.4036) grad_norm 1.0431 (1.1874) [2021-04-15 15:58:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][900/1251] eta 0:01:39 lr 0.000985 time 0.4693 (0.2839) loss 4.2139 (4.4042) grad_norm inf (inf) [2021-04-15 15:58:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][910/1251] eta 0:01:36 lr 0.000985 time 0.2422 (0.2838) loss 4.9573 (4.4011) grad_norm 1.0799 (inf) [2021-04-15 15:58:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][920/1251] eta 0:01:33 lr 0.000985 time 0.3015 (0.2837) loss 3.9001 (4.3996) grad_norm 1.2676 (inf) [2021-04-15 15:58:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][930/1251] eta 0:01:31 lr 0.000985 time 0.2600 (0.2836) loss 3.7324 (4.3980) grad_norm 0.8992 (inf) [2021-04-15 15:58:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][940/1251] eta 0:01:28 lr 0.000985 time 0.2935 (0.2836) loss 4.3812 (4.3986) grad_norm 1.2506 (inf) [2021-04-15 15:58:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][950/1251] eta 0:01:25 lr 0.000985 time 0.2850 (0.2835) loss 2.8343 (4.3963) grad_norm 1.2125 (inf) [2021-04-15 15:58:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][960/1251] eta 0:01:22 lr 0.000985 time 0.3041 (0.2835) loss 4.9170 (4.3986) grad_norm 1.3714 (inf) [2021-04-15 15:58:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][970/1251] eta 0:01:19 lr 0.000985 time 0.2975 (0.2834) loss 4.9148 (4.3986) grad_norm 1.1196 (inf) [2021-04-15 15:58:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][980/1251] eta 0:01:16 lr 0.000985 time 0.2567 (0.2833) loss 4.3634 (4.3966) grad_norm 1.2160 (inf) [2021-04-15 15:58:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][990/1251] eta 0:01:13 lr 0.000985 time 0.2567 (0.2832) loss 3.6391 (4.3961) grad_norm 1.2257 (inf) [2021-04-15 15:58:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][1000/1251] eta 0:01:11 lr 0.000985 time 0.3083 (0.2831) loss 4.4578 (4.3983) grad_norm 1.1114 (inf) [2021-04-15 15:58:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][1010/1251] eta 0:01:08 lr 0.000985 time 0.2856 (0.2831) loss 4.3322 (4.3995) grad_norm 1.1088 (inf) [2021-04-15 15:58:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][1020/1251] eta 0:01:05 lr 0.000985 time 0.2815 (0.2830) loss 4.0797 (4.4007) grad_norm 1.1953 (inf) [2021-04-15 15:58:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][1030/1251] eta 0:01:02 lr 0.000985 time 0.2950 (0.2830) loss 4.3378 (4.4000) grad_norm 1.0603 (inf) [2021-04-15 15:58:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][1040/1251] eta 0:00:59 lr 0.000985 time 0.2737 (0.2830) loss 3.7912 (4.4015) grad_norm 1.0832 (inf) [2021-04-15 15:58:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][1050/1251] eta 0:00:56 lr 0.000985 time 0.2658 (0.2829) loss 4.6609 (4.3986) grad_norm 1.3358 (inf) [2021-04-15 15:58:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][1060/1251] eta 0:00:54 lr 0.000985 time 0.2844 (0.2828) loss 5.0823 (4.4016) grad_norm 1.0656 (inf) [2021-04-15 15:58:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][1070/1251] eta 0:00:51 lr 0.000985 time 0.2692 (0.2828) loss 4.1392 (4.3995) grad_norm 1.5086 (inf) [2021-04-15 15:58:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][1080/1251] eta 0:00:48 lr 0.000985 time 0.2761 (0.2827) loss 4.0873 (4.3986) grad_norm 1.0613 (inf) [2021-04-15 15:58:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][1090/1251] eta 0:00:45 lr 0.000985 time 0.2778 (0.2827) loss 5.2030 (4.4012) grad_norm 1.2406 (inf) [2021-04-15 15:58:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][1100/1251] eta 0:00:42 lr 0.000985 time 0.2624 (0.2826) loss 4.6069 (4.3975) grad_norm 1.4214 (inf) [2021-04-15 15:58:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][1110/1251] eta 0:00:39 lr 0.000985 time 0.2968 (0.2826) loss 4.6102 (4.3984) grad_norm 1.0262 (inf) [2021-04-15 15:59:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][1120/1251] eta 0:00:37 lr 0.000985 time 0.2976 (0.2825) loss 4.9345 (4.4000) grad_norm 1.5002 (inf) [2021-04-15 15:59:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][1130/1251] eta 0:00:34 lr 0.000985 time 0.2930 (0.2825) loss 4.5286 (4.3999) grad_norm 1.1397 (inf) [2021-04-15 15:59:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][1140/1251] eta 0:00:31 lr 0.000985 time 0.2949 (0.2824) loss 4.5484 (4.4015) grad_norm 1.6306 (inf) [2021-04-15 15:59:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][1150/1251] eta 0:00:28 lr 0.000985 time 0.2573 (0.2824) loss 4.1547 (4.4014) grad_norm 1.1243 (inf) [2021-04-15 15:59:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][1160/1251] eta 0:00:25 lr 0.000985 time 0.2722 (0.2823) loss 3.2319 (4.4008) grad_norm 0.9191 (inf) [2021-04-15 15:59:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][1170/1251] eta 0:00:22 lr 0.000985 time 0.2786 (0.2823) loss 3.9726 (4.3992) grad_norm 0.9713 (inf) [2021-04-15 15:59:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][1180/1251] eta 0:00:20 lr 0.000985 time 0.2651 (0.2822) loss 4.2673 (4.3964) grad_norm 0.9009 (inf) [2021-04-15 15:59:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][1190/1251] eta 0:00:17 lr 0.000985 time 0.2594 (0.2822) loss 4.9438 (4.3966) grad_norm 1.2590 (inf) [2021-04-15 15:59:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][1200/1251] eta 0:00:14 lr 0.000985 time 0.2679 (0.2822) loss 3.8182 (4.3952) grad_norm 1.2618 (inf) [2021-04-15 15:59:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][1210/1251] eta 0:00:11 lr 0.000984 time 0.2805 (0.2821) loss 4.8639 (4.3970) grad_norm 1.5441 (inf) [2021-04-15 15:59:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][1220/1251] eta 0:00:08 lr 0.000984 time 0.2777 (0.2821) loss 3.5297 (4.3953) grad_norm 1.1252 (inf) [2021-04-15 15:59:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][1230/1251] eta 0:00:05 lr 0.000984 time 0.2591 (0.2820) loss 4.4648 (4.3957) grad_norm 0.9417 (inf) [2021-04-15 15:59:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][1240/1251] eta 0:00:03 lr 0.000984 time 0.2482 (0.2819) loss 3.2727 (4.3942) grad_norm 1.1732 (inf) [2021-04-15 15:59:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [23/300][1250/1251] eta 0:00:00 lr 0.000984 time 0.2493 (0.2817) loss 4.6702 (4.3955) grad_norm 1.0276 (inf) [2021-04-15 15:59:39 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 23 training takes 0:05:54 [2021-04-15 15:59:39 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_23.pth saving...... [2021-04-15 16:00:04 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_23.pth saved !!! [2021-04-15 16:00:05 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.229 (1.229) Loss 1.6866 (1.6866) Acc@1 61.328 (61.328) Acc@5 84.766 (84.766) [2021-04-15 16:00:07 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.117 (0.282) Loss 1.7663 (1.6800) Acc@1 62.012 (62.562) Acc@5 84.375 (85.165) [2021-04-15 16:00:08 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.178 (0.217) Loss 1.7253 (1.6798) Acc@1 60.547 (62.174) Acc@5 85.156 (85.259) [2021-04-15 16:00:11 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.130 (0.237) Loss 1.6617 (1.6714) Acc@1 62.891 (62.383) Acc@5 86.035 (85.320) [2021-04-15 16:00:13 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.155 (0.219) Loss 1.7372 (1.6748) Acc@1 61.621 (62.371) Acc@5 84.277 (85.354) [2021-04-15 16:00:15 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 62.468 Acc@5 85.340 [2021-04-15 16:00:15 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 62.5% [2021-04-15 16:00:15 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 62.47% [2021-04-15 16:00:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][0/1251] eta 1:20:36 lr 0.000984 time 3.8664 (3.8664) loss 3.4544 (3.4544) grad_norm 1.2779 (1.2779) [2021-04-15 16:00:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][10/1251] eta 0:12:37 lr 0.000984 time 0.2738 (0.6106) loss 3.8886 (4.2116) grad_norm 1.1342 (1.2016) [2021-04-15 16:00:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][20/1251] eta 0:09:16 lr 0.000984 time 0.2603 (0.4517) loss 3.0426 (4.1518) grad_norm 1.5772 (1.1967) [2021-04-15 16:00:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][30/1251] eta 0:08:05 lr 0.000984 time 0.2817 (0.3972) loss 4.0927 (4.2618) grad_norm 1.2124 (1.1477) [2021-04-15 16:00:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][40/1251] eta 0:07:25 lr 0.000984 time 0.2625 (0.3680) loss 3.4576 (4.2888) grad_norm 1.1644 (1.1374) [2021-04-15 16:00:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][50/1251] eta 0:07:00 lr 0.000984 time 0.2690 (0.3504) loss 3.2691 (4.3125) grad_norm 0.9496 (1.1500) [2021-04-15 16:00:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][60/1251] eta 0:06:43 lr 0.000984 time 0.2841 (0.3387) loss 3.3797 (4.3065) grad_norm 1.0955 (1.1708) [2021-04-15 16:00:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][70/1251] eta 0:06:29 lr 0.000984 time 0.3004 (0.3301) loss 4.4562 (4.3472) grad_norm 1.1675 (1.1789) [2021-04-15 16:00:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][80/1251] eta 0:06:19 lr 0.000984 time 0.2809 (0.3240) loss 5.2089 (4.3385) grad_norm 1.0515 (1.1741) [2021-04-15 16:00:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][90/1251] eta 0:06:12 lr 0.000984 time 0.2925 (0.3210) loss 5.0295 (4.3680) grad_norm 1.3261 (1.1644) [2021-04-15 16:00:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][100/1251] eta 0:06:06 lr 0.000984 time 0.2696 (0.3188) loss 4.6662 (4.3705) grad_norm 0.9981 (1.1593) [2021-04-15 16:00:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][110/1251] eta 0:06:01 lr 0.000984 time 0.4862 (0.3169) loss 4.1946 (4.3713) grad_norm 1.1577 (1.1605) [2021-04-15 16:00:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][120/1251] eta 0:05:55 lr 0.000984 time 0.2596 (0.3139) loss 4.0111 (4.3702) grad_norm 1.4321 (1.1692) [2021-04-15 16:00:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][130/1251] eta 0:05:49 lr 0.000984 time 0.2700 (0.3114) loss 4.4977 (4.3595) grad_norm 1.0902 (1.1672) [2021-04-15 16:00:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][140/1251] eta 0:05:43 lr 0.000984 time 0.3063 (0.3091) loss 5.2173 (4.3614) grad_norm 0.9928 (1.1630) [2021-04-15 16:01:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][150/1251] eta 0:05:38 lr 0.000984 time 0.2826 (0.3074) loss 3.3742 (4.3320) grad_norm 1.1597 (1.1644) [2021-04-15 16:01:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][160/1251] eta 0:05:33 lr 0.000984 time 0.2647 (0.3055) loss 4.5400 (4.3654) grad_norm 1.0282 (1.1623) [2021-04-15 16:01:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][170/1251] eta 0:05:28 lr 0.000984 time 0.2966 (0.3043) loss 4.8697 (4.3541) grad_norm 1.0492 (1.1605) [2021-04-15 16:01:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][180/1251] eta 0:05:24 lr 0.000984 time 0.3076 (0.3032) loss 4.0442 (4.3579) grad_norm 1.3205 (1.1593) [2021-04-15 16:01:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][190/1251] eta 0:05:20 lr 0.000984 time 0.2649 (0.3017) loss 4.7965 (4.3564) grad_norm 1.1966 (1.1632) [2021-04-15 16:01:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][200/1251] eta 0:05:15 lr 0.000984 time 0.2881 (0.3006) loss 3.9627 (4.3553) grad_norm 1.0786 (1.1638) [2021-04-15 16:01:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][210/1251] eta 0:05:12 lr 0.000984 time 0.3088 (0.2997) loss 3.3482 (4.3554) grad_norm 1.2640 (1.1626) [2021-04-15 16:01:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][220/1251] eta 0:05:08 lr 0.000984 time 0.2670 (0.2991) loss 4.6893 (4.3570) grad_norm 1.2229 (1.1622) [2021-04-15 16:01:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][230/1251] eta 0:05:04 lr 0.000984 time 0.2896 (0.2984) loss 4.3347 (4.3738) grad_norm 1.1703 (1.1626) [2021-04-15 16:01:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][240/1251] eta 0:05:01 lr 0.000984 time 0.2839 (0.2978) loss 4.7142 (4.3887) grad_norm 1.3348 (1.1657) [2021-04-15 16:01:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][250/1251] eta 0:04:57 lr 0.000984 time 0.2746 (0.2969) loss 4.3588 (4.4043) grad_norm 1.3229 (1.1658) [2021-04-15 16:01:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][260/1251] eta 0:04:53 lr 0.000984 time 0.2615 (0.2960) loss 3.1596 (4.4003) grad_norm 1.2813 (1.1646) [2021-04-15 16:01:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][270/1251] eta 0:04:49 lr 0.000984 time 0.2695 (0.2953) loss 3.6080 (4.3920) grad_norm 1.4309 (1.1681) [2021-04-15 16:01:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][280/1251] eta 0:04:45 lr 0.000984 time 0.2636 (0.2945) loss 5.3129 (4.3984) grad_norm 1.1570 (1.1728) [2021-04-15 16:01:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][290/1251] eta 0:04:42 lr 0.000984 time 0.2860 (0.2940) loss 4.6308 (4.4058) grad_norm 1.2083 (1.1721) [2021-04-15 16:01:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][300/1251] eta 0:04:38 lr 0.000984 time 0.2816 (0.2934) loss 3.4333 (4.3984) grad_norm 0.9593 (1.1729) [2021-04-15 16:01:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][310/1251] eta 0:04:35 lr 0.000984 time 0.2969 (0.2930) loss 4.0983 (4.3973) grad_norm 1.0900 (1.1755) [2021-04-15 16:01:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][320/1251] eta 0:04:32 lr 0.000984 time 0.2865 (0.2926) loss 4.4915 (4.3885) grad_norm 1.0462 (1.1744) [2021-04-15 16:01:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][330/1251] eta 0:04:29 lr 0.000984 time 0.2450 (0.2922) loss 3.3669 (4.3808) grad_norm 1.2192 (1.1716) [2021-04-15 16:01:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][340/1251] eta 0:04:25 lr 0.000984 time 0.2749 (0.2919) loss 4.3336 (4.3795) grad_norm 1.1034 (1.1758) [2021-04-15 16:01:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][350/1251] eta 0:04:22 lr 0.000984 time 0.2625 (0.2914) loss 4.6002 (4.3806) grad_norm 0.8415 (1.1749) [2021-04-15 16:02:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][360/1251] eta 0:04:19 lr 0.000984 time 0.2618 (0.2911) loss 5.1844 (4.3845) grad_norm 0.9653 (1.1730) [2021-04-15 16:02:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][370/1251] eta 0:04:16 lr 0.000984 time 0.2998 (0.2909) loss 3.9184 (4.3718) grad_norm 1.1008 (1.1760) [2021-04-15 16:02:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][380/1251] eta 0:04:13 lr 0.000984 time 0.2818 (0.2905) loss 5.2829 (4.3744) grad_norm 1.0938 (1.1829) [2021-04-15 16:02:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][390/1251] eta 0:04:09 lr 0.000984 time 0.2833 (0.2902) loss 4.5929 (4.3727) grad_norm 1.0673 (1.1818) [2021-04-15 16:02:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][400/1251] eta 0:04:06 lr 0.000984 time 0.3033 (0.2900) loss 3.8805 (4.3731) grad_norm 1.1395 (1.1798) [2021-04-15 16:02:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][410/1251] eta 0:04:03 lr 0.000984 time 0.2694 (0.2896) loss 3.8790 (4.3731) grad_norm 1.2729 (1.1785) [2021-04-15 16:02:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][420/1251] eta 0:04:00 lr 0.000984 time 0.2652 (0.2893) loss 3.2039 (4.3743) grad_norm 1.1876 (1.1791) [2021-04-15 16:02:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][430/1251] eta 0:03:57 lr 0.000984 time 0.2836 (0.2890) loss 4.3636 (4.3751) grad_norm 1.0601 (1.1788) [2021-04-15 16:02:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][440/1251] eta 0:03:54 lr 0.000984 time 0.2642 (0.2888) loss 4.5019 (4.3768) grad_norm 1.3440 (1.1782) [2021-04-15 16:02:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][450/1251] eta 0:03:51 lr 0.000984 time 0.2611 (0.2886) loss 4.6058 (4.3776) grad_norm 1.0201 (1.1774) [2021-04-15 16:02:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][460/1251] eta 0:03:47 lr 0.000984 time 0.2678 (0.2882) loss 3.2362 (4.3705) grad_norm 1.0224 (1.1756) [2021-04-15 16:02:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][470/1251] eta 0:03:44 lr 0.000984 time 0.2925 (0.2880) loss 5.4029 (4.3680) grad_norm 1.4294 (1.1739) [2021-04-15 16:02:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][480/1251] eta 0:03:41 lr 0.000984 time 0.2826 (0.2878) loss 5.0567 (4.3663) grad_norm 1.0953 (1.1729) [2021-04-15 16:02:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][490/1251] eta 0:03:38 lr 0.000984 time 0.2852 (0.2877) loss 3.5701 (4.3616) grad_norm 1.3042 (1.1707) [2021-04-15 16:02:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][500/1251] eta 0:03:35 lr 0.000984 time 0.2848 (0.2875) loss 4.5523 (4.3599) grad_norm 1.0318 (1.1688) [2021-04-15 16:02:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][510/1251] eta 0:03:32 lr 0.000984 time 0.2922 (0.2874) loss 4.7042 (4.3598) grad_norm 1.2535 (1.1684) [2021-04-15 16:02:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][520/1251] eta 0:03:30 lr 0.000984 time 0.2703 (0.2873) loss 4.4236 (4.3607) grad_norm 1.4077 (1.1676) [2021-04-15 16:02:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][530/1251] eta 0:03:27 lr 0.000984 time 0.2556 (0.2871) loss 3.4170 (4.3584) grad_norm 1.2130 (1.1678) [2021-04-15 16:02:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][540/1251] eta 0:03:24 lr 0.000984 time 0.2839 (0.2870) loss 4.3566 (4.3552) grad_norm 1.2652 (1.1673) [2021-04-15 16:02:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][550/1251] eta 0:03:21 lr 0.000984 time 0.2718 (0.2868) loss 4.6931 (4.3594) grad_norm 1.0469 (1.1673) [2021-04-15 16:02:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][560/1251] eta 0:03:18 lr 0.000984 time 0.2674 (0.2866) loss 3.4455 (4.3586) grad_norm 1.1456 (1.1665) [2021-04-15 16:02:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][570/1251] eta 0:03:15 lr 0.000984 time 0.2752 (0.2864) loss 3.6449 (4.3570) grad_norm 1.1315 (1.1659) [2021-04-15 16:03:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][580/1251] eta 0:03:12 lr 0.000984 time 0.3138 (0.2865) loss 4.3490 (4.3552) grad_norm 1.0706 (1.1658) [2021-04-15 16:03:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][590/1251] eta 0:03:09 lr 0.000984 time 0.2781 (0.2864) loss 4.9014 (4.3591) grad_norm 1.1412 (1.1661) [2021-04-15 16:03:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][600/1251] eta 0:03:06 lr 0.000984 time 0.2611 (0.2862) loss 3.7738 (4.3549) grad_norm 0.9302 (1.1656) [2021-04-15 16:03:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][610/1251] eta 0:03:03 lr 0.000984 time 0.2949 (0.2863) loss 4.7568 (4.3524) grad_norm 1.2882 (1.1665) [2021-04-15 16:03:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][620/1251] eta 0:03:00 lr 0.000984 time 0.2558 (0.2861) loss 4.4053 (4.3510) grad_norm 1.2484 (1.1689) [2021-04-15 16:03:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][630/1251] eta 0:02:57 lr 0.000984 time 0.2770 (0.2860) loss 5.0196 (4.3473) grad_norm 1.1594 (1.1704) [2021-04-15 16:03:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][640/1251] eta 0:02:54 lr 0.000984 time 0.2977 (0.2859) loss 4.7284 (4.3530) grad_norm 1.1080 (1.1698) [2021-04-15 16:03:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][650/1251] eta 0:02:51 lr 0.000984 time 0.2758 (0.2858) loss 4.2549 (4.3520) grad_norm 0.9889 (1.1695) [2021-04-15 16:03:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][660/1251] eta 0:02:48 lr 0.000984 time 0.2657 (0.2857) loss 3.9292 (4.3520) grad_norm 0.9397 (1.1700) [2021-04-15 16:03:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][670/1251] eta 0:02:46 lr 0.000984 time 0.2823 (0.2858) loss 3.5529 (4.3526) grad_norm 0.9020 (1.1691) [2021-04-15 16:03:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][680/1251] eta 0:02:43 lr 0.000984 time 0.2861 (0.2859) loss 3.8627 (4.3502) grad_norm 1.2508 (1.1683) [2021-04-15 16:03:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][690/1251] eta 0:02:40 lr 0.000984 time 0.2516 (0.2859) loss 4.9173 (4.3523) grad_norm 1.2908 (1.1682) [2021-04-15 16:03:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][700/1251] eta 0:02:37 lr 0.000984 time 0.2698 (0.2858) loss 5.0278 (4.3505) grad_norm 0.9984 (1.1676) [2021-04-15 16:03:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][710/1251] eta 0:02:34 lr 0.000984 time 0.2731 (0.2857) loss 3.8837 (4.3490) grad_norm 1.0187 (1.1666) [2021-04-15 16:03:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][720/1251] eta 0:02:31 lr 0.000984 time 0.2875 (0.2856) loss 4.6965 (4.3499) grad_norm 1.2668 (1.1676) [2021-04-15 16:03:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][730/1251] eta 0:02:28 lr 0.000984 time 0.2859 (0.2855) loss 3.0100 (4.3450) grad_norm 1.2063 (1.1681) [2021-04-15 16:03:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][740/1251] eta 0:02:25 lr 0.000984 time 0.2881 (0.2854) loss 4.5121 (4.3448) grad_norm 1.1669 (1.1677) [2021-04-15 16:03:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][750/1251] eta 0:02:22 lr 0.000984 time 0.2592 (0.2853) loss 3.9267 (4.3457) grad_norm 1.1360 (1.1676) [2021-04-15 16:03:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][760/1251] eta 0:02:20 lr 0.000984 time 0.2709 (0.2852) loss 4.1912 (4.3456) grad_norm 1.0719 (1.1663) [2021-04-15 16:03:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][770/1251] eta 0:02:17 lr 0.000984 time 0.2885 (0.2851) loss 3.8134 (4.3468) grad_norm 0.8868 (1.1644) [2021-04-15 16:03:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][780/1251] eta 0:02:14 lr 0.000984 time 0.2783 (0.2850) loss 3.4586 (4.3482) grad_norm 1.0167 (1.1645) [2021-04-15 16:04:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][790/1251] eta 0:02:11 lr 0.000984 time 0.2701 (0.2849) loss 4.5201 (4.3493) grad_norm 1.1547 (1.1657) [2021-04-15 16:04:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][800/1251] eta 0:02:08 lr 0.000984 time 0.2689 (0.2848) loss 3.7051 (4.3486) grad_norm 1.4705 (1.1659) [2021-04-15 16:04:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][810/1251] eta 0:02:05 lr 0.000984 time 0.2638 (0.2847) loss 5.1444 (4.3521) grad_norm 1.2744 (1.1664) [2021-04-15 16:04:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][820/1251] eta 0:02:02 lr 0.000984 time 0.2706 (0.2846) loss 4.2785 (4.3536) grad_norm 1.0049 (1.1668) [2021-04-15 16:04:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][830/1251] eta 0:01:59 lr 0.000984 time 0.2582 (0.2845) loss 3.4140 (4.3514) grad_norm 1.1949 (1.1679) [2021-04-15 16:04:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][840/1251] eta 0:01:56 lr 0.000984 time 0.2624 (0.2844) loss 4.0575 (4.3464) grad_norm 1.2692 (1.1709) [2021-04-15 16:04:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][850/1251] eta 0:01:54 lr 0.000984 time 0.2783 (0.2843) loss 4.7915 (4.3453) grad_norm 1.0089 (1.1712) [2021-04-15 16:04:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][860/1251] eta 0:01:51 lr 0.000984 time 0.3113 (0.2843) loss 3.9187 (4.3456) grad_norm 1.0707 (1.1711) [2021-04-15 16:04:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][870/1251] eta 0:01:48 lr 0.000984 time 0.2647 (0.2842) loss 4.4039 (4.3436) grad_norm 1.0659 (1.1707) [2021-04-15 16:04:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][880/1251] eta 0:01:45 lr 0.000984 time 0.3004 (0.2842) loss 4.3935 (4.3417) grad_norm 1.1463 (1.1703) [2021-04-15 16:04:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][890/1251] eta 0:01:42 lr 0.000984 time 0.2957 (0.2841) loss 4.5854 (4.3411) grad_norm 1.0274 (1.1691) [2021-04-15 16:04:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][900/1251] eta 0:01:39 lr 0.000984 time 0.2844 (0.2841) loss 5.0480 (4.3406) grad_norm 1.3953 (1.1700) [2021-04-15 16:04:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][910/1251] eta 0:01:36 lr 0.000983 time 0.2570 (0.2839) loss 5.0264 (4.3425) grad_norm 1.0399 (1.1693) [2021-04-15 16:04:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][920/1251] eta 0:01:33 lr 0.000983 time 0.2570 (0.2839) loss 4.5425 (4.3386) grad_norm 1.1473 (1.1689) [2021-04-15 16:04:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][930/1251] eta 0:01:31 lr 0.000983 time 0.2766 (0.2838) loss 4.1554 (4.3351) grad_norm 1.1422 (1.1688) [2021-04-15 16:04:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][940/1251] eta 0:01:28 lr 0.000983 time 0.2878 (0.2838) loss 3.2976 (4.3320) grad_norm 0.9811 (1.1692) [2021-04-15 16:04:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][950/1251] eta 0:01:25 lr 0.000983 time 0.2863 (0.2837) loss 4.8060 (4.3318) grad_norm 1.1295 (1.1687) [2021-04-15 16:04:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][960/1251] eta 0:01:22 lr 0.000983 time 0.2882 (0.2837) loss 4.1407 (4.3316) grad_norm 0.9813 (1.1677) [2021-04-15 16:04:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][970/1251] eta 0:01:19 lr 0.000983 time 0.2714 (0.2836) loss 4.2675 (4.3314) grad_norm 1.2703 (1.1683) [2021-04-15 16:04:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][980/1251] eta 0:01:16 lr 0.000983 time 0.2595 (0.2835) loss 5.0168 (4.3318) grad_norm 1.1009 (1.1680) [2021-04-15 16:04:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][990/1251] eta 0:01:13 lr 0.000983 time 0.2668 (0.2835) loss 4.9063 (4.3313) grad_norm 0.9540 (1.1680) [2021-04-15 16:04:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][1000/1251] eta 0:01:11 lr 0.000983 time 0.2555 (0.2833) loss 4.7240 (4.3321) grad_norm 1.1751 (1.1669) [2021-04-15 16:05:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][1010/1251] eta 0:01:08 lr 0.000983 time 0.3076 (0.2833) loss 3.1742 (4.3323) grad_norm 1.2541 (1.1661) [2021-04-15 16:05:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][1020/1251] eta 0:01:05 lr 0.000983 time 0.2842 (0.2832) loss 5.0982 (4.3311) grad_norm 1.0336 (1.1659) [2021-04-15 16:05:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][1030/1251] eta 0:01:02 lr 0.000983 time 0.2735 (0.2832) loss 4.3932 (4.3324) grad_norm 0.9981 (1.1665) [2021-04-15 16:05:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][1040/1251] eta 0:00:59 lr 0.000983 time 0.2767 (0.2831) loss 4.8372 (4.3339) grad_norm 1.1697 (1.1659) [2021-04-15 16:05:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][1050/1251] eta 0:00:56 lr 0.000983 time 0.2878 (0.2830) loss 4.1514 (4.3329) grad_norm 1.0946 (1.1660) [2021-04-15 16:05:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][1060/1251] eta 0:00:54 lr 0.000983 time 0.2994 (0.2830) loss 4.3082 (4.3333) grad_norm 2.0251 (1.1665) [2021-04-15 16:05:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][1070/1251] eta 0:00:51 lr 0.000983 time 0.2764 (0.2830) loss 4.7738 (4.3345) grad_norm 1.0944 (1.1666) [2021-04-15 16:05:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][1080/1251] eta 0:00:48 lr 0.000983 time 0.2664 (0.2829) loss 4.5889 (4.3330) grad_norm 1.3483 (1.1669) [2021-04-15 16:05:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][1090/1251] eta 0:00:45 lr 0.000983 time 0.3100 (0.2829) loss 4.5773 (4.3342) grad_norm 1.2990 (1.1673) [2021-04-15 16:05:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][1100/1251] eta 0:00:42 lr 0.000983 time 0.2864 (0.2828) loss 4.1555 (4.3334) grad_norm 1.3975 (1.1680) [2021-04-15 16:05:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][1110/1251] eta 0:00:39 lr 0.000983 time 0.2620 (0.2828) loss 3.1578 (4.3294) grad_norm 0.9353 (1.1672) [2021-04-15 16:05:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][1120/1251] eta 0:00:37 lr 0.000983 time 0.2518 (0.2827) loss 4.4334 (4.3296) grad_norm 1.2772 (1.1664) [2021-04-15 16:05:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][1130/1251] eta 0:00:34 lr 0.000983 time 0.2779 (0.2827) loss 3.7165 (4.3293) grad_norm 1.1832 (1.1667) [2021-04-15 16:05:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][1140/1251] eta 0:00:31 lr 0.000983 time 0.2894 (0.2826) loss 3.5238 (4.3273) grad_norm 1.3967 (1.1662) [2021-04-15 16:05:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][1150/1251] eta 0:00:28 lr 0.000983 time 0.2896 (0.2826) loss 5.0704 (4.3291) grad_norm 0.9942 (1.1659) [2021-04-15 16:05:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][1160/1251] eta 0:00:25 lr 0.000983 time 0.2955 (0.2826) loss 5.1042 (4.3269) grad_norm 1.1406 (1.1654) [2021-04-15 16:05:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][1170/1251] eta 0:00:22 lr 0.000983 time 0.2802 (0.2826) loss 3.7559 (4.3272) grad_norm 1.1877 (1.1657) [2021-04-15 16:05:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][1180/1251] eta 0:00:20 lr 0.000983 time 0.3021 (0.2826) loss 3.5782 (4.3238) grad_norm 0.9730 (1.1654) [2021-04-15 16:05:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][1190/1251] eta 0:00:17 lr 0.000983 time 0.2790 (0.2825) loss 3.4261 (4.3238) grad_norm 1.3980 (1.1645) [2021-04-15 16:05:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][1200/1251] eta 0:00:14 lr 0.000983 time 0.2784 (0.2825) loss 3.6299 (4.3218) grad_norm 1.1248 (1.1651) [2021-04-15 16:05:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][1210/1251] eta 0:00:11 lr 0.000983 time 0.2470 (0.2824) loss 4.5457 (4.3237) grad_norm 1.1980 (1.1649) [2021-04-15 16:06:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][1220/1251] eta 0:00:08 lr 0.000983 time 0.2555 (0.2823) loss 4.3815 (4.3240) grad_norm 1.1068 (1.1646) [2021-04-15 16:06:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][1230/1251] eta 0:00:05 lr 0.000983 time 0.2901 (0.2823) loss 4.6449 (4.3239) grad_norm 0.9466 (1.1642) [2021-04-15 16:06:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][1240/1251] eta 0:00:03 lr 0.000983 time 0.3312 (0.2822) loss 3.5062 (4.3235) grad_norm 1.1120 (1.1633) [2021-04-15 16:06:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [24/300][1250/1251] eta 0:00:00 lr 0.000983 time 0.2511 (0.2819) loss 4.8119 (4.3249) grad_norm 1.4154 (1.1628) [2021-04-15 16:06:11 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 24 training takes 0:05:55 [2021-04-15 16:06:11 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_24.pth saving...... [2021-04-15 16:06:28 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_24.pth saved !!! [2021-04-15 16:06:29 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.066 (1.066) Loss 1.5861 (1.5861) Acc@1 62.891 (62.891) Acc@5 87.891 (87.891) [2021-04-15 16:06:31 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.154 (0.263) Loss 1.6114 (1.6524) Acc@1 62.695 (62.340) Acc@5 86.719 (85.574) [2021-04-15 16:06:32 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.184 (0.212) Loss 1.6803 (1.6434) Acc@1 62.793 (62.844) Acc@5 85.840 (85.603) [2021-04-15 16:06:35 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.696 (0.236) Loss 1.5569 (1.6440) Acc@1 64.844 (62.969) Acc@5 86.133 (85.575) [2021-04-15 16:06:36 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.216) Loss 1.6617 (1.6376) Acc@1 62.988 (63.167) Acc@5 84.277 (85.606) [2021-04-15 16:06:39 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 63.186 Acc@5 85.674 [2021-04-15 16:06:39 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 63.2% [2021-04-15 16:06:39 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 63.19% [2021-04-15 16:06:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][0/1251] eta 1:25:37 lr 0.000983 time 4.1069 (4.1069) loss 4.7541 (4.7541) grad_norm 1.1110 (1.1110) [2021-04-15 16:06:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][10/1251] eta 0:13:07 lr 0.000983 time 0.2708 (0.6345) loss 4.6414 (4.6289) grad_norm 1.2182 (1.2034) [2021-04-15 16:06:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][20/1251] eta 0:09:31 lr 0.000983 time 0.3048 (0.4646) loss 3.8028 (4.5122) grad_norm 1.2863 (1.2110) [2021-04-15 16:06:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][30/1251] eta 0:08:14 lr 0.000983 time 0.2925 (0.4053) loss 5.0042 (4.3933) grad_norm 1.0162 (1.1846) [2021-04-15 16:06:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][40/1251] eta 0:07:35 lr 0.000983 time 0.2444 (0.3758) loss 4.9660 (4.4033) grad_norm 1.3460 (1.1940) [2021-04-15 16:06:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][50/1251] eta 0:07:10 lr 0.000983 time 0.2870 (0.3587) loss 3.5626 (4.3466) grad_norm 1.0859 (1.1963) [2021-04-15 16:07:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][60/1251] eta 0:06:53 lr 0.000983 time 0.2693 (0.3469) loss 4.8658 (4.3353) grad_norm 1.4146 (1.1987) [2021-04-15 16:07:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][70/1251] eta 0:06:39 lr 0.000983 time 0.2882 (0.3380) loss 3.5089 (4.3754) grad_norm 1.2290 (1.1929) [2021-04-15 16:07:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][80/1251] eta 0:06:27 lr 0.000983 time 0.2855 (0.3308) loss 4.2273 (4.3931) grad_norm 0.9847 (1.1936) [2021-04-15 16:07:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][90/1251] eta 0:06:16 lr 0.000983 time 0.2598 (0.3246) loss 4.9162 (4.3699) grad_norm 1.0936 (1.1984) [2021-04-15 16:07:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][100/1251] eta 0:06:08 lr 0.000983 time 0.2619 (0.3200) loss 4.0364 (4.3547) grad_norm 1.1250 (1.1919) [2021-04-15 16:07:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][110/1251] eta 0:06:00 lr 0.000983 time 0.2665 (0.3160) loss 4.1663 (4.3654) grad_norm 1.0001 (1.1939) [2021-04-15 16:07:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][120/1251] eta 0:05:54 lr 0.000983 time 0.2580 (0.3130) loss 4.9223 (4.3640) grad_norm 1.1247 (1.1909) [2021-04-15 16:07:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][130/1251] eta 0:05:48 lr 0.000983 time 0.2811 (0.3105) loss 4.7495 (4.3481) grad_norm 0.9671 (1.1911) [2021-04-15 16:07:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][140/1251] eta 0:05:42 lr 0.000983 time 0.2836 (0.3085) loss 4.1556 (4.3410) grad_norm 1.0719 (1.1941) [2021-04-15 16:07:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][150/1251] eta 0:05:37 lr 0.000983 time 0.2843 (0.3066) loss 4.5992 (4.3383) grad_norm 1.3159 (1.1983) [2021-04-15 16:07:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][160/1251] eta 0:05:32 lr 0.000983 time 0.2915 (0.3051) loss 4.7400 (4.3414) grad_norm 0.9589 (1.1920) [2021-04-15 16:07:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][170/1251] eta 0:05:28 lr 0.000983 time 0.2646 (0.3038) loss 4.2819 (4.3527) grad_norm 1.0523 (1.1893) [2021-04-15 16:07:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][180/1251] eta 0:05:24 lr 0.000983 time 0.3210 (0.3027) loss 4.8036 (4.3535) grad_norm 0.9495 (1.1841) [2021-04-15 16:07:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][190/1251] eta 0:05:19 lr 0.000983 time 0.2690 (0.3013) loss 4.8453 (4.3635) grad_norm 1.2448 (1.1835) [2021-04-15 16:07:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][200/1251] eta 0:05:15 lr 0.000983 time 0.3120 (0.3003) loss 4.5591 (4.3573) grad_norm 1.1741 (1.1788) [2021-04-15 16:07:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][210/1251] eta 0:05:11 lr 0.000983 time 0.2910 (0.2995) loss 4.9226 (4.3653) grad_norm 1.2452 (1.1806) [2021-04-15 16:07:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][220/1251] eta 0:05:07 lr 0.000983 time 0.2630 (0.2986) loss 3.3609 (4.3674) grad_norm 1.2158 (1.1791) [2021-04-15 16:07:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][230/1251] eta 0:05:04 lr 0.000983 time 0.2764 (0.2980) loss 5.3333 (4.3671) grad_norm 0.9196 (1.1788) [2021-04-15 16:07:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][240/1251] eta 0:05:00 lr 0.000983 time 0.2992 (0.2974) loss 3.9854 (4.3535) grad_norm 0.9893 (1.1739) [2021-04-15 16:07:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][250/1251] eta 0:04:56 lr 0.000983 time 0.3117 (0.2967) loss 3.2776 (4.3523) grad_norm 1.1598 (1.1733) [2021-04-15 16:07:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][260/1251] eta 0:04:53 lr 0.000983 time 0.2764 (0.2961) loss 5.0820 (4.3642) grad_norm 1.1511 (1.1720) [2021-04-15 16:07:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][270/1251] eta 0:04:49 lr 0.000983 time 0.2711 (0.2956) loss 4.3027 (4.3531) grad_norm 1.3396 (1.1712) [2021-04-15 16:08:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][280/1251] eta 0:04:46 lr 0.000983 time 0.2661 (0.2949) loss 4.4967 (4.3445) grad_norm 1.1993 (1.1716) [2021-04-15 16:08:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][290/1251] eta 0:04:42 lr 0.000983 time 0.2838 (0.2944) loss 4.4014 (4.3417) grad_norm 1.2565 (1.1756) [2021-04-15 16:08:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][300/1251] eta 0:04:39 lr 0.000983 time 0.2638 (0.2938) loss 4.9541 (4.3391) grad_norm 1.1821 (1.1783) [2021-04-15 16:08:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][310/1251] eta 0:04:36 lr 0.000983 time 0.2714 (0.2933) loss 4.7947 (4.3435) grad_norm 1.0534 (1.1759) [2021-04-15 16:08:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][320/1251] eta 0:04:32 lr 0.000983 time 0.2666 (0.2929) loss 3.8049 (4.3422) grad_norm 1.1167 (1.1745) [2021-04-15 16:08:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][330/1251] eta 0:04:29 lr 0.000983 time 0.2758 (0.2925) loss 3.6804 (4.3380) grad_norm 1.0283 (1.1735) [2021-04-15 16:08:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][340/1251] eta 0:04:26 lr 0.000983 time 0.2774 (0.2921) loss 4.1511 (4.3431) grad_norm 0.9922 (1.1716) [2021-04-15 16:08:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][350/1251] eta 0:04:22 lr 0.000983 time 0.2634 (0.2917) loss 3.5973 (4.3356) grad_norm 1.1665 (1.1717) [2021-04-15 16:08:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][360/1251] eta 0:04:19 lr 0.000983 time 0.2583 (0.2915) loss 4.5012 (4.3384) grad_norm 1.3801 (1.1720) [2021-04-15 16:08:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][370/1251] eta 0:04:16 lr 0.000983 time 0.2956 (0.2912) loss 4.7225 (4.3363) grad_norm 1.0788 (1.1737) [2021-04-15 16:08:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][380/1251] eta 0:04:13 lr 0.000983 time 0.2750 (0.2908) loss 4.6805 (4.3329) grad_norm 1.4324 (1.1764) [2021-04-15 16:08:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][390/1251] eta 0:04:10 lr 0.000983 time 0.2594 (0.2907) loss 5.2163 (4.3395) grad_norm 1.0236 (1.1768) [2021-04-15 16:08:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][400/1251] eta 0:04:07 lr 0.000983 time 0.2582 (0.2904) loss 3.3453 (4.3448) grad_norm 1.0080 (1.1766) [2021-04-15 16:08:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][410/1251] eta 0:04:04 lr 0.000983 time 0.2734 (0.2902) loss 4.6659 (4.3496) grad_norm 1.3575 (1.1764) [2021-04-15 16:08:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][420/1251] eta 0:04:00 lr 0.000983 time 0.2942 (0.2900) loss 4.0950 (4.3454) grad_norm 1.1642 (1.1740) [2021-04-15 16:08:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][430/1251] eta 0:03:57 lr 0.000983 time 0.2689 (0.2897) loss 4.8224 (4.3425) grad_norm 1.0229 (1.1735) [2021-04-15 16:08:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][440/1251] eta 0:03:54 lr 0.000983 time 0.2773 (0.2894) loss 4.5617 (4.3410) grad_norm 1.1852 (1.1740) [2021-04-15 16:08:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][450/1251] eta 0:03:51 lr 0.000983 time 0.4518 (0.2895) loss 4.7007 (4.3445) grad_norm 1.3209 (1.1751) [2021-04-15 16:08:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][460/1251] eta 0:03:48 lr 0.000983 time 0.3069 (0.2892) loss 3.6222 (4.3455) grad_norm 0.9713 (1.1742) [2021-04-15 16:08:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][470/1251] eta 0:03:45 lr 0.000983 time 0.2649 (0.2893) loss 3.2009 (4.3406) grad_norm 1.0197 (1.1726) [2021-04-15 16:08:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][480/1251] eta 0:03:43 lr 0.000983 time 0.3018 (0.2893) loss 4.2629 (4.3365) grad_norm 1.2970 (inf) [2021-04-15 16:09:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][490/1251] eta 0:03:39 lr 0.000983 time 0.2757 (0.2891) loss 3.7515 (4.3342) grad_norm 1.0132 (inf) [2021-04-15 16:09:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][500/1251] eta 0:03:36 lr 0.000983 time 0.2698 (0.2889) loss 5.0200 (4.3397) grad_norm 1.0712 (inf) [2021-04-15 16:09:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][510/1251] eta 0:03:33 lr 0.000983 time 0.2961 (0.2888) loss 4.8875 (4.3444) grad_norm 1.3243 (inf) [2021-04-15 16:09:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][520/1251] eta 0:03:30 lr 0.000983 time 0.2626 (0.2885) loss 3.4401 (4.3450) grad_norm 1.2568 (inf) [2021-04-15 16:09:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][530/1251] eta 0:03:27 lr 0.000983 time 0.2921 (0.2883) loss 4.1627 (4.3451) grad_norm 1.0090 (inf) [2021-04-15 16:09:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][540/1251] eta 0:03:24 lr 0.000983 time 0.2824 (0.2880) loss 4.4242 (4.3493) grad_norm 1.0082 (inf) [2021-04-15 16:09:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][550/1251] eta 0:03:21 lr 0.000983 time 0.2991 (0.2879) loss 3.4008 (4.3461) grad_norm 0.9816 (inf) [2021-04-15 16:09:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][560/1251] eta 0:03:18 lr 0.000983 time 0.2757 (0.2877) loss 3.7666 (4.3404) grad_norm 1.3160 (inf) [2021-04-15 16:09:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][570/1251] eta 0:03:16 lr 0.000983 time 0.2768 (0.2878) loss 4.3659 (4.3330) grad_norm 0.9661 (inf) [2021-04-15 16:09:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][580/1251] eta 0:03:12 lr 0.000983 time 0.2421 (0.2876) loss 3.2891 (4.3363) grad_norm 1.1388 (inf) [2021-04-15 16:09:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][590/1251] eta 0:03:10 lr 0.000982 time 0.3023 (0.2875) loss 4.9632 (4.3402) grad_norm 1.1832 (inf) [2021-04-15 16:09:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][600/1251] eta 0:03:07 lr 0.000982 time 0.2995 (0.2873) loss 4.2082 (4.3433) grad_norm 1.0438 (inf) [2021-04-15 16:09:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][610/1251] eta 0:03:04 lr 0.000982 time 0.2828 (0.2871) loss 4.8065 (4.3394) grad_norm 1.1255 (inf) [2021-04-15 16:09:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][620/1251] eta 0:03:01 lr 0.000982 time 0.2564 (0.2869) loss 3.3425 (4.3393) grad_norm 1.0786 (inf) [2021-04-15 16:09:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][630/1251] eta 0:02:58 lr 0.000982 time 0.3093 (0.2868) loss 3.2310 (4.3396) grad_norm 1.0602 (inf) [2021-04-15 16:09:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][640/1251] eta 0:02:55 lr 0.000982 time 0.2759 (0.2866) loss 4.6014 (4.3455) grad_norm 1.0199 (inf) [2021-04-15 16:09:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][650/1251] eta 0:02:52 lr 0.000982 time 0.2540 (0.2863) loss 4.0087 (4.3415) grad_norm 1.1428 (inf) [2021-04-15 16:09:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][660/1251] eta 0:02:49 lr 0.000982 time 0.2819 (0.2862) loss 3.9607 (4.3390) grad_norm 1.2107 (inf) [2021-04-15 16:09:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][670/1251] eta 0:02:46 lr 0.000982 time 0.2744 (0.2861) loss 4.3962 (4.3419) grad_norm 1.3047 (inf) [2021-04-15 16:09:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][680/1251] eta 0:02:43 lr 0.000982 time 0.2644 (0.2858) loss 3.9369 (4.3405) grad_norm 1.2477 (inf) [2021-04-15 16:09:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][690/1251] eta 0:02:40 lr 0.000982 time 0.2912 (0.2857) loss 4.7636 (4.3385) grad_norm 1.4100 (inf) [2021-04-15 16:09:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][700/1251] eta 0:02:37 lr 0.000982 time 0.2781 (0.2856) loss 5.0677 (4.3421) grad_norm 1.2948 (inf) [2021-04-15 16:10:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][710/1251] eta 0:02:34 lr 0.000982 time 0.2701 (0.2856) loss 4.0337 (4.3445) grad_norm 1.2285 (inf) [2021-04-15 16:10:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][720/1251] eta 0:02:31 lr 0.000982 time 0.2939 (0.2856) loss 5.1084 (4.3457) grad_norm 1.4304 (inf) [2021-04-15 16:10:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][730/1251] eta 0:02:28 lr 0.000982 time 0.2743 (0.2855) loss 3.6164 (4.3431) grad_norm 1.0433 (inf) [2021-04-15 16:10:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][740/1251] eta 0:02:25 lr 0.000982 time 0.2695 (0.2853) loss 4.6652 (4.3400) grad_norm 1.0552 (inf) [2021-04-15 16:10:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][750/1251] eta 0:02:22 lr 0.000982 time 0.2746 (0.2852) loss 4.5499 (4.3383) grad_norm 1.3306 (inf) [2021-04-15 16:10:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][760/1251] eta 0:02:20 lr 0.000982 time 0.2878 (0.2852) loss 3.4933 (4.3366) grad_norm 1.4209 (inf) [2021-04-15 16:10:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][770/1251] eta 0:02:17 lr 0.000982 time 0.2709 (0.2851) loss 5.2054 (4.3372) grad_norm 0.9568 (inf) [2021-04-15 16:10:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][780/1251] eta 0:02:14 lr 0.000982 time 0.2641 (0.2850) loss 4.8167 (4.3381) grad_norm 1.1544 (inf) [2021-04-15 16:10:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][790/1251] eta 0:02:11 lr 0.000982 time 0.2917 (0.2849) loss 4.1567 (4.3378) grad_norm 1.0757 (inf) [2021-04-15 16:10:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][800/1251] eta 0:02:08 lr 0.000982 time 0.2918 (0.2848) loss 3.1529 (4.3318) grad_norm 1.2606 (inf) [2021-04-15 16:10:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][810/1251] eta 0:02:05 lr 0.000982 time 0.2613 (0.2847) loss 4.9225 (4.3316) grad_norm 1.3116 (inf) [2021-04-15 16:10:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][820/1251] eta 0:02:02 lr 0.000982 time 0.2898 (0.2846) loss 5.5774 (4.3336) grad_norm 1.2569 (inf) [2021-04-15 16:10:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][830/1251] eta 0:01:59 lr 0.000982 time 0.2742 (0.2845) loss 4.9106 (4.3327) grad_norm 1.1203 (inf) [2021-04-15 16:10:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][840/1251] eta 0:01:56 lr 0.000982 time 0.2874 (0.2845) loss 4.8938 (4.3350) grad_norm 1.0684 (inf) [2021-04-15 16:10:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][850/1251] eta 0:01:54 lr 0.000982 time 0.2675 (0.2844) loss 3.3582 (4.3343) grad_norm 1.3625 (inf) [2021-04-15 16:10:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][860/1251] eta 0:01:51 lr 0.000982 time 0.2620 (0.2843) loss 4.5902 (4.3366) grad_norm 0.9996 (inf) [2021-04-15 16:10:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][870/1251] eta 0:01:48 lr 0.000982 time 0.2816 (0.2843) loss 4.4130 (4.3401) grad_norm 0.8940 (inf) [2021-04-15 16:10:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][880/1251] eta 0:01:45 lr 0.000982 time 0.2564 (0.2842) loss 4.6117 (4.3391) grad_norm 1.0972 (inf) [2021-04-15 16:10:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][890/1251] eta 0:01:42 lr 0.000982 time 0.2887 (0.2841) loss 3.3483 (4.3334) grad_norm 1.2592 (inf) [2021-04-15 16:10:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][900/1251] eta 0:01:39 lr 0.000982 time 0.2651 (0.2840) loss 3.0803 (4.3282) grad_norm 1.2342 (inf) [2021-04-15 16:10:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][910/1251] eta 0:01:36 lr 0.000982 time 0.3109 (0.2839) loss 4.2075 (4.3286) grad_norm 0.9883 (inf) [2021-04-15 16:11:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][920/1251] eta 0:01:33 lr 0.000982 time 0.2704 (0.2838) loss 4.2464 (4.3302) grad_norm 1.1775 (inf) [2021-04-15 16:11:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][930/1251] eta 0:01:31 lr 0.000982 time 0.2854 (0.2838) loss 4.3863 (4.3338) grad_norm 1.0967 (inf) [2021-04-15 16:11:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][940/1251] eta 0:01:28 lr 0.000982 time 0.2910 (0.2838) loss 3.9440 (4.3351) grad_norm 0.9796 (inf) [2021-04-15 16:11:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][950/1251] eta 0:01:25 lr 0.000982 time 0.2656 (0.2838) loss 3.9366 (4.3351) grad_norm 1.1042 (inf) [2021-04-15 16:11:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][960/1251] eta 0:01:22 lr 0.000982 time 0.3031 (0.2837) loss 4.7389 (4.3370) grad_norm 0.9325 (inf) [2021-04-15 16:11:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][970/1251] eta 0:01:19 lr 0.000982 time 0.2505 (0.2836) loss 3.7748 (4.3349) grad_norm 1.3773 (inf) [2021-04-15 16:11:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][980/1251] eta 0:01:16 lr 0.000982 time 0.2785 (0.2836) loss 4.6029 (4.3340) grad_norm 1.2194 (inf) [2021-04-15 16:11:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][990/1251] eta 0:01:13 lr 0.000982 time 0.2741 (0.2835) loss 4.7362 (4.3331) grad_norm 0.9614 (inf) [2021-04-15 16:11:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][1000/1251] eta 0:01:11 lr 0.000982 time 0.2762 (0.2835) loss 4.7867 (4.3347) grad_norm 1.2143 (inf) [2021-04-15 16:11:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][1010/1251] eta 0:01:08 lr 0.000982 time 0.2718 (0.2834) loss 3.1044 (4.3292) grad_norm 1.0747 (inf) [2021-04-15 16:11:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][1020/1251] eta 0:01:05 lr 0.000982 time 0.2596 (0.2833) loss 3.8726 (4.3268) grad_norm 1.2821 (inf) [2021-04-15 16:11:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][1030/1251] eta 0:01:02 lr 0.000982 time 0.3012 (0.2834) loss 4.1484 (4.3269) grad_norm 1.0963 (inf) [2021-04-15 16:11:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][1040/1251] eta 0:00:59 lr 0.000982 time 0.2768 (0.2834) loss 3.3432 (4.3265) grad_norm 1.2684 (inf) [2021-04-15 16:11:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][1050/1251] eta 0:00:56 lr 0.000982 time 0.4054 (0.2835) loss 4.2778 (4.3268) grad_norm 0.9959 (inf) [2021-04-15 16:11:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][1060/1251] eta 0:00:54 lr 0.000982 time 0.2787 (0.2834) loss 4.2714 (4.3240) grad_norm 1.3328 (inf) [2021-04-15 16:11:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][1070/1251] eta 0:00:51 lr 0.000982 time 0.2881 (0.2834) loss 4.5525 (4.3236) grad_norm 0.9853 (inf) [2021-04-15 16:11:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][1080/1251] eta 0:00:48 lr 0.000982 time 0.2670 (0.2834) loss 4.2395 (4.3237) grad_norm 1.0813 (inf) [2021-04-15 16:11:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][1090/1251] eta 0:00:45 lr 0.000982 time 0.2883 (0.2833) loss 3.3226 (4.3205) grad_norm 1.3153 (inf) [2021-04-15 16:11:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][1100/1251] eta 0:00:42 lr 0.000982 time 0.2883 (0.2833) loss 4.7228 (4.3196) grad_norm 1.1369 (inf) [2021-04-15 16:11:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][1110/1251] eta 0:00:39 lr 0.000982 time 0.2572 (0.2832) loss 4.8383 (4.3205) grad_norm 1.1558 (inf) [2021-04-15 16:11:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][1120/1251] eta 0:00:37 lr 0.000982 time 0.2729 (0.2831) loss 4.3433 (4.3228) grad_norm 1.1488 (inf) [2021-04-15 16:11:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][1130/1251] eta 0:00:34 lr 0.000982 time 0.2876 (0.2830) loss 3.4182 (4.3184) grad_norm 1.4156 (inf) [2021-04-15 16:12:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][1140/1251] eta 0:00:31 lr 0.000982 time 0.2880 (0.2830) loss 4.3479 (4.3154) grad_norm 1.0872 (inf) [2021-04-15 16:12:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][1150/1251] eta 0:00:28 lr 0.000982 time 0.2749 (0.2830) loss 3.5059 (4.3130) grad_norm 1.2108 (inf) [2021-04-15 16:12:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][1160/1251] eta 0:00:25 lr 0.000982 time 0.2862 (0.2830) loss 3.6429 (4.3138) grad_norm 1.0319 (inf) [2021-04-15 16:12:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][1170/1251] eta 0:00:22 lr 0.000982 time 0.2575 (0.2829) loss 4.2373 (4.3140) grad_norm 1.0679 (inf) [2021-04-15 16:12:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][1180/1251] eta 0:00:20 lr 0.000982 time 0.2627 (0.2828) loss 3.6184 (4.3129) grad_norm 1.1884 (inf) [2021-04-15 16:12:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][1190/1251] eta 0:00:17 lr 0.000982 time 0.2693 (0.2828) loss 3.6870 (4.3126) grad_norm 1.0366 (inf) [2021-04-15 16:12:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][1200/1251] eta 0:00:14 lr 0.000982 time 0.2688 (0.2827) loss 5.1561 (4.3129) grad_norm 0.9749 (inf) [2021-04-15 16:12:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][1210/1251] eta 0:00:11 lr 0.000982 time 0.2637 (0.2827) loss 4.5897 (4.3129) grad_norm 1.1327 (inf) [2021-04-15 16:12:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][1220/1251] eta 0:00:08 lr 0.000982 time 0.2588 (0.2826) loss 4.8528 (4.3122) grad_norm 1.0342 (inf) [2021-04-15 16:12:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][1230/1251] eta 0:00:05 lr 0.000982 time 0.2613 (0.2825) loss 4.1316 (4.3119) grad_norm 1.0863 (inf) [2021-04-15 16:12:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][1240/1251] eta 0:00:03 lr 0.000982 time 0.2484 (0.2824) loss 4.4021 (4.3130) grad_norm 1.1741 (inf) [2021-04-15 16:12:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [25/300][1250/1251] eta 0:00:00 lr 0.000982 time 0.2515 (0.2822) loss 4.2385 (4.3133) grad_norm 1.3525 (inf) [2021-04-15 16:12:34 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 25 training takes 0:05:54 [2021-04-15 16:12:34 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_25.pth saving...... [2021-04-15 16:12:44 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_25.pth saved !!! [2021-04-15 16:12:45 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.144 (1.144) Loss 1.6100 (1.6100) Acc@1 63.770 (63.770) Acc@5 85.449 (85.449) [2021-04-15 16:12:47 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.135 (0.246) Loss 1.6331 (1.5952) Acc@1 63.379 (63.805) Acc@5 85.156 (86.142) [2021-04-15 16:12:49 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.100 (0.233) Loss 1.6356 (1.5927) Acc@1 63.770 (64.039) Acc@5 85.547 (86.165) [2021-04-15 16:12:52 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.152 (0.236) Loss 1.6110 (1.5947) Acc@1 63.965 (63.977) Acc@5 85.449 (86.038) [2021-04-15 16:12:53 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.089 (0.219) Loss 1.6486 (1.6041) Acc@1 64.648 (63.824) Acc@5 84.766 (85.866) [2021-04-15 16:12:55 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 63.876 Acc@5 85.946 [2021-04-15 16:12:55 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 63.9% [2021-04-15 16:12:55 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 63.88% [2021-04-15 16:12:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][0/1251] eta 1:19:32 lr 0.000982 time 3.8152 (3.8152) loss 4.8158 (4.8158) grad_norm 1.1984 (1.1984) [2021-04-15 16:13:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][10/1251] eta 0:12:48 lr 0.000982 time 0.2878 (0.6192) loss 5.3115 (4.4362) grad_norm 1.0434 (1.1359) [2021-04-15 16:13:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][20/1251] eta 0:09:23 lr 0.000982 time 0.2694 (0.4579) loss 4.4404 (4.5595) grad_norm 0.9480 (1.1456) [2021-04-15 16:13:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][30/1251] eta 0:08:11 lr 0.000982 time 0.2955 (0.4026) loss 4.3042 (4.4279) grad_norm 1.4164 (1.1347) [2021-04-15 16:13:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][40/1251] eta 0:07:32 lr 0.000982 time 0.2737 (0.3734) loss 3.8047 (4.3897) grad_norm 1.1891 (1.1236) [2021-04-15 16:13:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][50/1251] eta 0:07:07 lr 0.000982 time 0.2735 (0.3558) loss 3.8406 (4.3171) grad_norm 1.0844 (1.1299) [2021-04-15 16:13:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][60/1251] eta 0:06:49 lr 0.000982 time 0.2842 (0.3435) loss 4.6375 (4.2829) grad_norm 1.2495 (1.1531) [2021-04-15 16:13:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][70/1251] eta 0:06:37 lr 0.000982 time 0.2804 (0.3362) loss 3.0090 (4.2587) grad_norm 1.0611 (1.1533) [2021-04-15 16:13:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][80/1251] eta 0:06:26 lr 0.000982 time 0.2912 (0.3298) loss 5.0496 (4.3092) grad_norm 1.1120 (1.1425) [2021-04-15 16:13:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][90/1251] eta 0:06:16 lr 0.000982 time 0.2624 (0.3245) loss 4.6431 (4.2995) grad_norm 1.0810 (1.1419) [2021-04-15 16:13:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][100/1251] eta 0:06:08 lr 0.000982 time 0.2993 (0.3201) loss 4.3751 (4.3099) grad_norm 1.0006 (1.1459) [2021-04-15 16:13:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][110/1251] eta 0:06:01 lr 0.000982 time 0.3039 (0.3166) loss 4.5658 (4.3108) grad_norm 0.9988 (1.1434) [2021-04-15 16:13:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][120/1251] eta 0:05:54 lr 0.000982 time 0.2773 (0.3137) loss 4.8016 (4.2801) grad_norm 1.2590 (1.1369) [2021-04-15 16:13:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][130/1251] eta 0:05:49 lr 0.000982 time 0.3048 (0.3115) loss 4.8561 (4.2962) grad_norm 0.9629 (1.1405) [2021-04-15 16:13:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][140/1251] eta 0:05:43 lr 0.000982 time 0.2805 (0.3094) loss 3.7632 (4.2834) grad_norm 0.9459 (1.1343) [2021-04-15 16:13:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][150/1251] eta 0:05:38 lr 0.000982 time 0.2690 (0.3074) loss 4.2531 (4.2798) grad_norm 1.0542 (1.1345) [2021-04-15 16:13:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][160/1251] eta 0:05:34 lr 0.000982 time 0.2585 (0.3063) loss 4.6589 (4.2638) grad_norm 1.1479 (1.1320) [2021-04-15 16:13:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][170/1251] eta 0:05:29 lr 0.000982 time 0.2537 (0.3047) loss 4.1792 (4.2591) grad_norm 1.2288 (1.1298) [2021-04-15 16:13:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][180/1251] eta 0:05:24 lr 0.000982 time 0.2729 (0.3034) loss 4.0562 (4.2545) grad_norm 1.1175 (1.1299) [2021-04-15 16:13:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][190/1251] eta 0:05:20 lr 0.000982 time 0.2724 (0.3020) loss 4.4923 (4.2417) grad_norm 1.2513 (1.1283) [2021-04-15 16:13:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][200/1251] eta 0:05:16 lr 0.000982 time 0.2772 (0.3008) loss 3.5766 (4.2389) grad_norm 1.1381 (1.1247) [2021-04-15 16:13:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][210/1251] eta 0:05:12 lr 0.000982 time 0.2732 (0.2998) loss 4.0072 (4.2419) grad_norm 1.0819 (1.1257) [2021-04-15 16:14:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][220/1251] eta 0:05:08 lr 0.000982 time 0.2783 (0.2989) loss 4.9685 (4.2398) grad_norm 1.1519 (1.1296) [2021-04-15 16:14:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][230/1251] eta 0:05:04 lr 0.000982 time 0.2956 (0.2980) loss 4.8202 (4.2530) grad_norm 1.1942 (1.1333) [2021-04-15 16:14:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][240/1251] eta 0:05:00 lr 0.000981 time 0.2767 (0.2973) loss 5.2412 (4.2636) grad_norm 1.1843 (1.1349) [2021-04-15 16:14:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][250/1251] eta 0:04:57 lr 0.000981 time 0.2680 (0.2971) loss 4.4698 (4.2738) grad_norm 1.1763 (1.1364) [2021-04-15 16:14:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][260/1251] eta 0:04:55 lr 0.000981 time 0.5782 (0.2977) loss 4.2799 (4.2707) grad_norm 0.9285 (1.1341) [2021-04-15 16:14:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][270/1251] eta 0:04:51 lr 0.000981 time 0.2913 (0.2970) loss 3.2396 (4.2717) grad_norm 1.1099 (1.1340) [2021-04-15 16:14:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][280/1251] eta 0:04:47 lr 0.000981 time 0.3099 (0.2963) loss 4.2869 (4.2707) grad_norm 1.1415 (1.1319) [2021-04-15 16:14:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][290/1251] eta 0:04:44 lr 0.000981 time 0.2678 (0.2957) loss 4.6367 (4.2762) grad_norm 1.2997 (1.1323) [2021-04-15 16:14:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][300/1251] eta 0:04:40 lr 0.000981 time 0.2823 (0.2952) loss 4.9119 (4.2711) grad_norm 1.1533 (1.1324) [2021-04-15 16:14:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][310/1251] eta 0:04:37 lr 0.000981 time 0.2892 (0.2948) loss 4.8171 (4.2819) grad_norm 1.7374 (1.1338) [2021-04-15 16:14:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][320/1251] eta 0:04:33 lr 0.000981 time 0.2654 (0.2943) loss 5.2012 (4.2877) grad_norm 1.1659 (1.1348) [2021-04-15 16:14:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][330/1251] eta 0:04:30 lr 0.000981 time 0.2817 (0.2939) loss 3.7275 (4.2922) grad_norm 1.0888 (1.1337) [2021-04-15 16:14:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][340/1251] eta 0:04:27 lr 0.000981 time 0.2789 (0.2935) loss 4.5160 (4.2914) grad_norm 1.2783 (1.1342) [2021-04-15 16:14:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][350/1251] eta 0:04:24 lr 0.000981 time 0.2794 (0.2931) loss 2.9464 (4.2839) grad_norm 0.9874 (1.1329) [2021-04-15 16:14:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][360/1251] eta 0:04:21 lr 0.000981 time 0.2696 (0.2932) loss 4.4719 (4.2831) grad_norm 1.3542 (1.1319) [2021-04-15 16:14:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][370/1251] eta 0:04:17 lr 0.000981 time 0.2798 (0.2928) loss 4.0212 (4.2853) grad_norm 1.4911 (1.1354) [2021-04-15 16:14:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][380/1251] eta 0:04:14 lr 0.000981 time 0.2694 (0.2924) loss 4.4889 (4.2853) grad_norm 0.8689 (1.1353) [2021-04-15 16:14:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][390/1251] eta 0:04:11 lr 0.000981 time 0.2789 (0.2920) loss 4.5961 (4.2921) grad_norm 0.9408 (1.1366) [2021-04-15 16:14:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][400/1251] eta 0:04:08 lr 0.000981 time 0.2613 (0.2917) loss 4.7888 (4.2928) grad_norm 1.3596 (1.1403) [2021-04-15 16:14:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][410/1251] eta 0:04:05 lr 0.000981 time 0.2796 (0.2914) loss 4.7157 (4.2884) grad_norm 0.9983 (1.1418) [2021-04-15 16:14:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][420/1251] eta 0:04:01 lr 0.000981 time 0.2882 (0.2912) loss 5.0890 (4.2865) grad_norm 1.1824 (1.1418) [2021-04-15 16:15:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][430/1251] eta 0:03:58 lr 0.000981 time 0.2780 (0.2908) loss 4.4890 (4.2930) grad_norm 1.1607 (1.1435) [2021-04-15 16:15:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][440/1251] eta 0:03:55 lr 0.000981 time 0.2905 (0.2907) loss 3.4977 (4.2972) grad_norm 1.4019 (1.1423) [2021-04-15 16:15:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][450/1251] eta 0:03:52 lr 0.000981 time 0.2618 (0.2908) loss 3.9135 (4.2867) grad_norm 1.0314 (1.1423) [2021-04-15 16:15:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][460/1251] eta 0:03:49 lr 0.000981 time 0.2539 (0.2905) loss 4.7560 (4.2851) grad_norm 1.5226 (1.1424) [2021-04-15 16:15:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][470/1251] eta 0:03:46 lr 0.000981 time 0.2735 (0.2901) loss 4.9237 (4.2810) grad_norm 1.0927 (1.1427) [2021-04-15 16:15:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][480/1251] eta 0:03:43 lr 0.000981 time 0.2704 (0.2899) loss 5.2124 (4.2832) grad_norm 1.0534 (1.1438) [2021-04-15 16:15:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][490/1251] eta 0:03:40 lr 0.000981 time 0.2658 (0.2896) loss 4.7761 (4.2855) grad_norm 0.9789 (1.1434) [2021-04-15 16:15:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][500/1251] eta 0:03:37 lr 0.000981 time 0.2856 (0.2894) loss 4.7642 (4.2823) grad_norm 1.0780 (1.1425) [2021-04-15 16:15:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][510/1251] eta 0:03:34 lr 0.000981 time 0.3063 (0.2893) loss 4.6717 (4.2806) grad_norm 1.1603 (1.1418) [2021-04-15 16:15:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][520/1251] eta 0:03:31 lr 0.000981 time 0.2721 (0.2891) loss 4.6452 (4.2783) grad_norm 1.4191 (1.1420) [2021-04-15 16:15:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][530/1251] eta 0:03:28 lr 0.000981 time 0.2664 (0.2889) loss 3.9662 (4.2821) grad_norm 1.0279 (1.1432) [2021-04-15 16:15:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][540/1251] eta 0:03:25 lr 0.000981 time 0.2644 (0.2887) loss 3.9660 (4.2828) grad_norm 1.4235 (1.1443) [2021-04-15 16:15:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][550/1251] eta 0:03:22 lr 0.000981 time 0.2643 (0.2885) loss 4.4120 (4.2777) grad_norm 0.9869 (1.1425) [2021-04-15 16:15:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][560/1251] eta 0:03:19 lr 0.000981 time 0.2717 (0.2884) loss 4.4621 (4.2817) grad_norm 1.1313 (1.1412) [2021-04-15 16:15:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][570/1251] eta 0:03:16 lr 0.000981 time 0.3084 (0.2884) loss 4.7620 (4.2827) grad_norm 1.5415 (1.1427) [2021-04-15 16:15:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][580/1251] eta 0:03:13 lr 0.000981 time 0.2751 (0.2883) loss 4.7498 (4.2835) grad_norm 1.1356 (1.1450) [2021-04-15 16:15:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][590/1251] eta 0:03:10 lr 0.000981 time 0.2562 (0.2881) loss 4.5451 (4.2836) grad_norm 0.9811 (1.1466) [2021-04-15 16:15:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][600/1251] eta 0:03:07 lr 0.000981 time 0.2527 (0.2880) loss 4.5739 (4.2819) grad_norm 1.2562 (1.1464) [2021-04-15 16:15:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][610/1251] eta 0:03:04 lr 0.000981 time 0.2798 (0.2879) loss 3.8940 (4.2838) grad_norm 1.1719 (1.1468) [2021-04-15 16:15:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][620/1251] eta 0:03:01 lr 0.000981 time 0.3053 (0.2877) loss 4.1653 (4.2867) grad_norm 1.0913 (1.1467) [2021-04-15 16:15:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][630/1251] eta 0:02:58 lr 0.000981 time 0.3071 (0.2876) loss 4.6929 (4.2870) grad_norm 1.0469 (1.1465) [2021-04-15 16:16:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][640/1251] eta 0:02:55 lr 0.000981 time 0.2815 (0.2875) loss 4.2643 (4.2909) grad_norm 1.0825 (1.1460) [2021-04-15 16:16:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][650/1251] eta 0:02:52 lr 0.000981 time 0.2563 (0.2873) loss 4.7232 (4.2934) grad_norm 1.5247 (1.1457) [2021-04-15 16:16:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][660/1251] eta 0:02:49 lr 0.000981 time 0.2885 (0.2871) loss 4.6432 (4.2922) grad_norm 0.9686 (1.1460) [2021-04-15 16:16:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][670/1251] eta 0:02:46 lr 0.000981 time 0.2875 (0.2870) loss 4.0265 (4.2953) grad_norm 1.4355 (1.1456) [2021-04-15 16:16:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][680/1251] eta 0:02:43 lr 0.000981 time 0.2983 (0.2869) loss 4.7613 (4.2961) grad_norm 0.9535 (1.1447) [2021-04-15 16:16:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][690/1251] eta 0:02:40 lr 0.000981 time 0.2930 (0.2868) loss 4.5669 (4.2966) grad_norm 1.0728 (1.1440) [2021-04-15 16:16:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][700/1251] eta 0:02:37 lr 0.000981 time 0.2635 (0.2867) loss 5.2125 (4.2999) grad_norm 1.1299 (1.1453) [2021-04-15 16:16:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][710/1251] eta 0:02:35 lr 0.000981 time 0.2914 (0.2865) loss 4.7302 (4.3047) grad_norm 0.8960 (1.1434) [2021-04-15 16:16:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][720/1251] eta 0:02:32 lr 0.000981 time 0.3007 (0.2864) loss 4.5783 (4.3039) grad_norm 0.9714 (1.1430) [2021-04-15 16:16:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][730/1251] eta 0:02:29 lr 0.000981 time 0.2636 (0.2864) loss 4.8927 (4.3077) grad_norm 1.1662 (1.1436) [2021-04-15 16:16:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][740/1251] eta 0:02:26 lr 0.000981 time 0.2605 (0.2863) loss 5.1115 (4.3070) grad_norm 1.1803 (1.1427) [2021-04-15 16:16:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][750/1251] eta 0:02:23 lr 0.000981 time 0.2615 (0.2862) loss 4.1471 (4.3042) grad_norm 1.0075 (1.1434) [2021-04-15 16:16:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][760/1251] eta 0:02:20 lr 0.000981 time 0.2809 (0.2861) loss 3.9150 (4.3073) grad_norm 1.1689 (1.1442) [2021-04-15 16:16:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][770/1251] eta 0:02:17 lr 0.000981 time 0.2833 (0.2860) loss 4.8232 (4.3081) grad_norm 1.4264 (1.1452) [2021-04-15 16:16:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][780/1251] eta 0:02:14 lr 0.000981 time 0.2826 (0.2858) loss 3.1429 (4.3057) grad_norm 1.0696 (1.1452) [2021-04-15 16:16:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][790/1251] eta 0:02:11 lr 0.000981 time 0.3051 (0.2858) loss 5.1724 (4.3079) grad_norm 1.2398 (1.1442) [2021-04-15 16:16:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][800/1251] eta 0:02:08 lr 0.000981 time 0.2890 (0.2857) loss 4.1971 (4.3084) grad_norm 1.0176 (1.1440) [2021-04-15 16:16:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][810/1251] eta 0:02:05 lr 0.000981 time 0.2968 (0.2857) loss 4.0813 (4.3063) grad_norm 1.0042 (1.1438) [2021-04-15 16:16:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][820/1251] eta 0:02:03 lr 0.000981 time 0.2732 (0.2856) loss 5.1833 (4.3076) grad_norm 1.0377 (1.1438) [2021-04-15 16:16:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][830/1251] eta 0:02:00 lr 0.000981 time 0.2843 (0.2856) loss 5.2408 (4.3087) grad_norm 1.0114 (1.1434) [2021-04-15 16:16:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][840/1251] eta 0:01:57 lr 0.000981 time 0.2981 (0.2857) loss 5.0392 (4.3107) grad_norm 0.9735 (1.1426) [2021-04-15 16:16:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][850/1251] eta 0:01:54 lr 0.000981 time 0.2687 (0.2855) loss 4.4762 (4.3149) grad_norm 1.0080 (1.1423) [2021-04-15 16:17:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][860/1251] eta 0:01:51 lr 0.000981 time 0.2719 (0.2855) loss 3.3936 (4.3164) grad_norm 1.0618 (1.1416) [2021-04-15 16:17:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][870/1251] eta 0:01:48 lr 0.000981 time 0.2445 (0.2854) loss 3.4774 (4.3159) grad_norm 1.2317 (1.1417) [2021-04-15 16:17:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][880/1251] eta 0:01:45 lr 0.000981 time 0.2785 (0.2853) loss 4.8000 (4.3127) grad_norm 1.1980 (1.1420) [2021-04-15 16:17:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][890/1251] eta 0:01:42 lr 0.000981 time 0.2700 (0.2852) loss 3.9450 (4.3106) grad_norm 1.2233 (1.1428) [2021-04-15 16:17:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][900/1251] eta 0:01:40 lr 0.000981 time 0.2801 (0.2851) loss 3.2241 (4.3115) grad_norm 1.1045 (1.1435) [2021-04-15 16:17:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][910/1251] eta 0:01:37 lr 0.000981 time 0.3078 (0.2850) loss 5.0275 (4.3106) grad_norm 1.2936 (1.1435) [2021-04-15 16:17:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][920/1251] eta 0:01:34 lr 0.000981 time 0.2699 (0.2850) loss 4.2324 (4.3084) grad_norm 1.1765 (1.1439) [2021-04-15 16:17:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][930/1251] eta 0:01:31 lr 0.000981 time 0.2734 (0.2849) loss 3.1615 (4.3046) grad_norm 1.0534 (1.1441) [2021-04-15 16:17:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][940/1251] eta 0:01:28 lr 0.000981 time 0.2888 (0.2851) loss 4.7309 (4.2976) grad_norm 1.2558 (1.1438) [2021-04-15 16:17:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][950/1251] eta 0:01:25 lr 0.000981 time 0.2963 (0.2850) loss 4.6912 (4.2962) grad_norm 0.9136 (1.1436) [2021-04-15 16:17:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][960/1251] eta 0:01:22 lr 0.000981 time 0.2723 (0.2850) loss 3.0364 (4.2944) grad_norm 1.2711 (1.1431) [2021-04-15 16:17:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][970/1251] eta 0:01:20 lr 0.000981 time 0.2610 (0.2850) loss 4.5050 (4.2957) grad_norm 1.1803 (1.1437) [2021-04-15 16:17:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][980/1251] eta 0:01:17 lr 0.000981 time 0.2959 (0.2849) loss 3.3706 (4.2949) grad_norm 1.3444 (1.1437) [2021-04-15 16:17:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][990/1251] eta 0:01:14 lr 0.000981 time 0.2953 (0.2849) loss 4.0096 (4.2957) grad_norm 1.2683 (1.1437) [2021-04-15 16:17:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][1000/1251] eta 0:01:11 lr 0.000981 time 0.2841 (0.2848) loss 5.0324 (4.2941) grad_norm 1.2221 (1.1433) [2021-04-15 16:17:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][1010/1251] eta 0:01:08 lr 0.000981 time 0.2446 (0.2848) loss 2.9780 (4.2917) grad_norm 0.9776 (1.1435) [2021-04-15 16:17:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][1020/1251] eta 0:01:05 lr 0.000981 time 0.2623 (0.2847) loss 4.8135 (4.2910) grad_norm 1.1158 (1.1431) [2021-04-15 16:17:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][1030/1251] eta 0:01:02 lr 0.000981 time 0.2551 (0.2846) loss 4.5609 (4.2939) grad_norm 0.9136 (1.1422) [2021-04-15 16:17:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][1040/1251] eta 0:01:00 lr 0.000981 time 0.2799 (0.2846) loss 4.4558 (4.2950) grad_norm 1.0723 (1.1424) [2021-04-15 16:17:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][1050/1251] eta 0:00:57 lr 0.000981 time 0.2947 (0.2845) loss 3.7432 (4.2949) grad_norm 0.9269 (1.1426) [2021-04-15 16:17:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][1060/1251] eta 0:00:54 lr 0.000981 time 0.2746 (0.2844) loss 3.3177 (4.2923) grad_norm 0.9789 (1.1434) [2021-04-15 16:18:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][1070/1251] eta 0:00:51 lr 0.000981 time 0.2465 (0.2843) loss 3.3990 (4.2919) grad_norm 1.0365 (1.1443) [2021-04-15 16:18:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][1080/1251] eta 0:00:48 lr 0.000981 time 0.2922 (0.2843) loss 4.0135 (4.2929) grad_norm 1.4188 (1.1445) [2021-04-15 16:18:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][1090/1251] eta 0:00:45 lr 0.000981 time 0.2544 (0.2842) loss 3.9852 (4.2935) grad_norm 1.1546 (1.1443) [2021-04-15 16:18:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][1100/1251] eta 0:00:42 lr 0.000981 time 0.2763 (0.2841) loss 4.6248 (4.2930) grad_norm 1.1534 (1.1438) [2021-04-15 16:18:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][1110/1251] eta 0:00:40 lr 0.000981 time 0.2655 (0.2841) loss 3.8554 (4.2919) grad_norm 1.3225 (1.1446) [2021-04-15 16:18:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][1120/1251] eta 0:00:37 lr 0.000980 time 0.2798 (0.2840) loss 4.8019 (4.2912) grad_norm 1.5253 (1.1453) [2021-04-15 16:18:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][1130/1251] eta 0:00:34 lr 0.000980 time 0.2818 (0.2840) loss 5.1069 (4.2913) grad_norm 1.0494 (1.1446) [2021-04-15 16:18:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][1140/1251] eta 0:00:31 lr 0.000980 time 0.2719 (0.2839) loss 3.5254 (4.2905) grad_norm 1.2253 (1.1446) [2021-04-15 16:18:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][1150/1251] eta 0:00:28 lr 0.000980 time 0.2446 (0.2838) loss 4.7891 (4.2919) grad_norm 1.1002 (1.1436) [2021-04-15 16:18:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][1160/1251] eta 0:00:25 lr 0.000980 time 0.2842 (0.2838) loss 4.5065 (4.2922) grad_norm 0.9021 (1.1430) [2021-04-15 16:18:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][1170/1251] eta 0:00:22 lr 0.000980 time 0.2659 (0.2838) loss 3.6492 (4.2896) grad_norm 1.0771 (1.1433) [2021-04-15 16:18:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][1180/1251] eta 0:00:20 lr 0.000980 time 0.2708 (0.2837) loss 3.6451 (4.2914) grad_norm 1.0051 (1.1427) [2021-04-15 16:18:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][1190/1251] eta 0:00:17 lr 0.000980 time 0.2539 (0.2836) loss 5.2502 (4.2913) grad_norm 1.3131 (1.1439) [2021-04-15 16:18:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][1200/1251] eta 0:00:14 lr 0.000980 time 0.2822 (0.2836) loss 4.0181 (4.2913) grad_norm 1.2634 (1.1442) [2021-04-15 16:18:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][1210/1251] eta 0:00:11 lr 0.000980 time 0.3078 (0.2836) loss 3.6239 (4.2915) grad_norm 1.0310 (1.1439) [2021-04-15 16:18:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][1220/1251] eta 0:00:08 lr 0.000980 time 0.2779 (0.2835) loss 4.7026 (4.2915) grad_norm 1.3606 (1.1432) [2021-04-15 16:18:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][1230/1251] eta 0:00:05 lr 0.000980 time 0.2659 (0.2834) loss 4.5068 (4.2908) grad_norm 1.1920 (1.1434) [2021-04-15 16:18:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][1240/1251] eta 0:00:03 lr 0.000980 time 0.2489 (0.2833) loss 4.5056 (4.2920) grad_norm 1.1384 (1.1432) [2021-04-15 16:18:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [26/300][1250/1251] eta 0:00:00 lr 0.000980 time 0.2500 (0.2830) loss 4.7207 (4.2920) grad_norm 1.1245 (1.1429) [2021-04-15 16:18:52 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 26 training takes 0:05:56 [2021-04-15 16:18:52 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_26.pth saving...... [2021-04-15 16:19:05 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_26.pth saved !!! [2021-04-15 16:19:07 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.206 (1.206) Loss 1.5328 (1.5328) Acc@1 66.406 (66.406) Acc@5 86.914 (86.914) [2021-04-15 16:19:08 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.141 (0.243) Loss 1.6358 (1.5687) Acc@1 63.379 (64.782) Acc@5 85.840 (86.373) [2021-04-15 16:19:11 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.182 (0.250) Loss 1.5928 (1.5688) Acc@1 64.941 (64.723) Acc@5 86.133 (86.389) [2021-04-15 16:19:13 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.206 (0.239) Loss 1.5923 (1.5703) Acc@1 64.258 (64.658) Acc@5 86.328 (86.460) [2021-04-15 16:19:15 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.072 (0.224) Loss 1.6740 (1.5799) Acc@1 61.816 (64.401) Acc@5 84.961 (86.423) [2021-04-15 16:19:17 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 64.340 Acc@5 86.490 [2021-04-15 16:19:17 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 64.3% [2021-04-15 16:19:17 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 64.34% [2021-04-15 16:19:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][0/1251] eta 1:21:56 lr 0.000980 time 3.9299 (3.9299) loss 5.0014 (5.0014) grad_norm 1.1241 (1.1241) [2021-04-15 16:19:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][10/1251] eta 0:12:37 lr 0.000980 time 0.2816 (0.6106) loss 4.2233 (4.1645) grad_norm 1.0392 (1.2176) [2021-04-15 16:19:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][20/1251] eta 0:09:18 lr 0.000980 time 0.2671 (0.4533) loss 4.2185 (4.1566) grad_norm 0.9194 (inf) [2021-04-15 16:19:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][30/1251] eta 0:08:13 lr 0.000980 time 0.2728 (0.4043) loss 4.5808 (4.1826) grad_norm 1.0597 (inf) [2021-04-15 16:19:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][40/1251] eta 0:07:39 lr 0.000980 time 0.2474 (0.3793) loss 4.8209 (4.1793) grad_norm 1.0567 (inf) [2021-04-15 16:19:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][50/1251] eta 0:07:12 lr 0.000980 time 0.2640 (0.3602) loss 4.4056 (4.1530) grad_norm 1.3304 (inf) [2021-04-15 16:19:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][60/1251] eta 0:06:54 lr 0.000980 time 0.3080 (0.3477) loss 4.6622 (4.1177) grad_norm 1.5925 (inf) [2021-04-15 16:19:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][70/1251] eta 0:06:39 lr 0.000980 time 0.2739 (0.3384) loss 4.6886 (4.1560) grad_norm 1.1648 (inf) [2021-04-15 16:19:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][80/1251] eta 0:06:27 lr 0.000980 time 0.2673 (0.3307) loss 4.6773 (4.1796) grad_norm 1.2583 (inf) [2021-04-15 16:19:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][90/1251] eta 0:06:17 lr 0.000980 time 0.2852 (0.3255) loss 4.9749 (4.1846) grad_norm 1.1331 (inf) [2021-04-15 16:19:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][100/1251] eta 0:06:10 lr 0.000980 time 0.3056 (0.3217) loss 3.9474 (4.1930) grad_norm 1.0417 (inf) [2021-04-15 16:19:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][110/1251] eta 0:06:02 lr 0.000980 time 0.2777 (0.3175) loss 4.6497 (4.2117) grad_norm 1.0633 (inf) [2021-04-15 16:19:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][120/1251] eta 0:05:55 lr 0.000980 time 0.2732 (0.3144) loss 3.9522 (4.1981) grad_norm 1.4876 (inf) [2021-04-15 16:19:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][130/1251] eta 0:05:49 lr 0.000980 time 0.2821 (0.3120) loss 4.1366 (4.2056) grad_norm 1.2199 (inf) [2021-04-15 16:20:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][140/1251] eta 0:05:44 lr 0.000980 time 0.2621 (0.3105) loss 4.4612 (4.2369) grad_norm 1.1370 (inf) [2021-04-15 16:20:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][150/1251] eta 0:05:39 lr 0.000980 time 0.2674 (0.3083) loss 4.1911 (4.2384) grad_norm 0.9246 (inf) [2021-04-15 16:20:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][160/1251] eta 0:05:34 lr 0.000980 time 0.2998 (0.3064) loss 4.6048 (4.2464) grad_norm 1.0655 (inf) [2021-04-15 16:20:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][170/1251] eta 0:05:29 lr 0.000980 time 0.2742 (0.3048) loss 4.3271 (4.2572) grad_norm 1.1183 (inf) [2021-04-15 16:20:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][180/1251] eta 0:05:25 lr 0.000980 time 0.2966 (0.3038) loss 3.5204 (4.2463) grad_norm 0.9521 (inf) [2021-04-15 16:20:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][190/1251] eta 0:05:20 lr 0.000980 time 0.2630 (0.3024) loss 4.0853 (4.2433) grad_norm 1.0274 (inf) [2021-04-15 16:20:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][200/1251] eta 0:05:16 lr 0.000980 time 0.2722 (0.3010) loss 4.5974 (4.2495) grad_norm 1.2836 (inf) [2021-04-15 16:20:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][210/1251] eta 0:05:12 lr 0.000980 time 0.2771 (0.3003) loss 4.4996 (4.2536) grad_norm 1.0516 (inf) [2021-04-15 16:20:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][220/1251] eta 0:05:08 lr 0.000980 time 0.3049 (0.2994) loss 5.1790 (4.2578) grad_norm 1.2215 (inf) [2021-04-15 16:20:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][230/1251] eta 0:05:04 lr 0.000980 time 0.2742 (0.2984) loss 3.0980 (4.2526) grad_norm 1.0500 (inf) [2021-04-15 16:20:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][240/1251] eta 0:05:00 lr 0.000980 time 0.2776 (0.2976) loss 3.1721 (4.2583) grad_norm 1.1222 (inf) [2021-04-15 16:20:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][250/1251] eta 0:04:57 lr 0.000980 time 0.2846 (0.2968) loss 4.3420 (4.2551) grad_norm 0.9030 (inf) [2021-04-15 16:20:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][260/1251] eta 0:04:53 lr 0.000980 time 0.2650 (0.2961) loss 4.3795 (4.2550) grad_norm 1.1930 (inf) [2021-04-15 16:20:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][270/1251] eta 0:04:49 lr 0.000980 time 0.2794 (0.2956) loss 4.6313 (4.2696) grad_norm 1.0605 (inf) [2021-04-15 16:20:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][280/1251] eta 0:04:46 lr 0.000980 time 0.2916 (0.2950) loss 5.2219 (4.2712) grad_norm 1.1093 (inf) [2021-04-15 16:20:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][290/1251] eta 0:04:43 lr 0.000980 time 0.2809 (0.2945) loss 3.5965 (4.2710) grad_norm 1.3486 (inf) [2021-04-15 16:20:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][300/1251] eta 0:04:39 lr 0.000980 time 0.2651 (0.2942) loss 4.5178 (4.2694) grad_norm 1.2573 (inf) [2021-04-15 16:20:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][310/1251] eta 0:04:36 lr 0.000980 time 0.2974 (0.2938) loss 4.2358 (4.2635) grad_norm 1.1237 (inf) [2021-04-15 16:20:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][320/1251] eta 0:04:33 lr 0.000980 time 0.3086 (0.2934) loss 4.3368 (4.2551) grad_norm 1.0764 (inf) [2021-04-15 16:20:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][330/1251] eta 0:04:30 lr 0.000980 time 0.2963 (0.2932) loss 4.4744 (4.2530) grad_norm 1.0615 (inf) [2021-04-15 16:20:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][340/1251] eta 0:04:26 lr 0.000980 time 0.2893 (0.2929) loss 5.1198 (4.2577) grad_norm 1.2214 (inf) [2021-04-15 16:21:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][350/1251] eta 0:04:23 lr 0.000980 time 0.2815 (0.2924) loss 3.6816 (4.2594) grad_norm 1.0635 (inf) [2021-04-15 16:21:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][360/1251] eta 0:04:20 lr 0.000980 time 0.2532 (0.2920) loss 3.9241 (4.2599) grad_norm 0.9994 (inf) [2021-04-15 16:21:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][370/1251] eta 0:04:16 lr 0.000980 time 0.2774 (0.2917) loss 4.7733 (4.2657) grad_norm 1.1704 (inf) [2021-04-15 16:21:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][380/1251] eta 0:04:13 lr 0.000980 time 0.2938 (0.2912) loss 3.1949 (4.2549) grad_norm 1.2311 (inf) [2021-04-15 16:21:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][390/1251] eta 0:04:10 lr 0.000980 time 0.2451 (0.2909) loss 4.8865 (4.2540) grad_norm 1.0203 (inf) [2021-04-15 16:21:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][400/1251] eta 0:04:07 lr 0.000980 time 0.2783 (0.2906) loss 4.6718 (4.2566) grad_norm 1.0688 (inf) [2021-04-15 16:21:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][410/1251] eta 0:04:04 lr 0.000980 time 0.2805 (0.2903) loss 4.0551 (4.2617) grad_norm 1.0094 (inf) [2021-04-15 16:21:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][420/1251] eta 0:04:01 lr 0.000980 time 0.2783 (0.2900) loss 3.3301 (4.2584) grad_norm 1.0283 (inf) [2021-04-15 16:21:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][430/1251] eta 0:03:57 lr 0.000980 time 0.2838 (0.2897) loss 3.8605 (4.2602) grad_norm 1.1656 (inf) [2021-04-15 16:21:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][440/1251] eta 0:03:54 lr 0.000980 time 0.3033 (0.2896) loss 2.9068 (4.2592) grad_norm 0.9924 (inf) [2021-04-15 16:21:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][450/1251] eta 0:03:51 lr 0.000980 time 0.2998 (0.2895) loss 4.0444 (4.2592) grad_norm 1.2720 (inf) [2021-04-15 16:21:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][460/1251] eta 0:03:48 lr 0.000980 time 0.3022 (0.2895) loss 4.7279 (4.2511) grad_norm 1.2214 (inf) [2021-04-15 16:21:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][470/1251] eta 0:03:45 lr 0.000980 time 0.2593 (0.2892) loss 4.5166 (4.2501) grad_norm 1.0617 (inf) [2021-04-15 16:21:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][480/1251] eta 0:03:42 lr 0.000980 time 0.2790 (0.2890) loss 4.5806 (4.2563) grad_norm 0.9056 (inf) [2021-04-15 16:21:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][490/1251] eta 0:03:39 lr 0.000980 time 0.3045 (0.2888) loss 4.6321 (4.2587) grad_norm 1.1303 (inf) [2021-04-15 16:21:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][500/1251] eta 0:03:36 lr 0.000980 time 0.2641 (0.2886) loss 5.2628 (4.2608) grad_norm 1.0177 (inf) [2021-04-15 16:21:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][510/1251] eta 0:03:33 lr 0.000980 time 0.2678 (0.2884) loss 3.6620 (4.2654) grad_norm 1.2591 (inf) [2021-04-15 16:21:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][520/1251] eta 0:03:30 lr 0.000980 time 0.2452 (0.2881) loss 4.1713 (4.2629) grad_norm 1.1558 (inf) [2021-04-15 16:21:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][530/1251] eta 0:03:27 lr 0.000980 time 0.2784 (0.2879) loss 4.6510 (4.2644) grad_norm 1.1195 (inf) [2021-04-15 16:21:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][540/1251] eta 0:03:24 lr 0.000980 time 0.2843 (0.2879) loss 4.8263 (4.2647) grad_norm 1.0504 (inf) [2021-04-15 16:21:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][550/1251] eta 0:03:21 lr 0.000980 time 0.2786 (0.2877) loss 4.3583 (4.2663) grad_norm 1.1922 (inf) [2021-04-15 16:21:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][560/1251] eta 0:03:18 lr 0.000980 time 0.2867 (0.2875) loss 4.6304 (4.2688) grad_norm 1.0813 (inf) [2021-04-15 16:22:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][570/1251] eta 0:03:15 lr 0.000980 time 0.2862 (0.2873) loss 3.5117 (4.2663) grad_norm 1.1093 (inf) [2021-04-15 16:22:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][580/1251] eta 0:03:12 lr 0.000980 time 0.3134 (0.2873) loss 3.4727 (4.2618) grad_norm 1.2329 (inf) [2021-04-15 16:22:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][590/1251] eta 0:03:09 lr 0.000980 time 0.2789 (0.2871) loss 4.2590 (4.2603) grad_norm 1.1216 (inf) [2021-04-15 16:22:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][600/1251] eta 0:03:06 lr 0.000980 time 0.4473 (0.2872) loss 4.6063 (4.2647) grad_norm 0.9780 (inf) [2021-04-15 16:22:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][610/1251] eta 0:03:03 lr 0.000980 time 0.2688 (0.2870) loss 4.0427 (4.2691) grad_norm 1.1013 (inf) [2021-04-15 16:22:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][620/1251] eta 0:03:01 lr 0.000980 time 0.2805 (0.2870) loss 4.4699 (4.2707) grad_norm 1.4755 (inf) [2021-04-15 16:22:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][630/1251] eta 0:02:58 lr 0.000980 time 0.2537 (0.2868) loss 3.4443 (4.2677) grad_norm 1.3736 (inf) [2021-04-15 16:22:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][640/1251] eta 0:02:55 lr 0.000980 time 0.2984 (0.2868) loss 4.2503 (4.2648) grad_norm 0.9582 (inf) [2021-04-15 16:22:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][650/1251] eta 0:02:52 lr 0.000980 time 0.2886 (0.2866) loss 3.0342 (4.2675) grad_norm 0.9581 (inf) [2021-04-15 16:22:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][660/1251] eta 0:02:49 lr 0.000980 time 0.2791 (0.2864) loss 4.5656 (4.2693) grad_norm 1.2716 (inf) [2021-04-15 16:22:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][670/1251] eta 0:02:46 lr 0.000980 time 0.2745 (0.2863) loss 4.5915 (4.2698) grad_norm 1.0507 (inf) [2021-04-15 16:22:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][680/1251] eta 0:02:43 lr 0.000980 time 0.2759 (0.2862) loss 5.1826 (4.2688) grad_norm 1.2256 (inf) [2021-04-15 16:22:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][690/1251] eta 0:02:40 lr 0.000980 time 0.2799 (0.2860) loss 4.7977 (4.2699) grad_norm 1.2137 (inf) [2021-04-15 16:22:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][700/1251] eta 0:02:37 lr 0.000980 time 0.2601 (0.2859) loss 3.9106 (4.2687) grad_norm 1.0102 (inf) [2021-04-15 16:22:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][710/1251] eta 0:02:34 lr 0.000980 time 0.2637 (0.2860) loss 3.4558 (4.2674) grad_norm 1.1802 (inf) [2021-04-15 16:22:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][720/1251] eta 0:02:31 lr 0.000980 time 0.2528 (0.2859) loss 5.0621 (4.2691) grad_norm 1.1722 (inf) [2021-04-15 16:22:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][730/1251] eta 0:02:28 lr 0.000979 time 0.2805 (0.2858) loss 4.3993 (4.2673) grad_norm 1.1338 (inf) [2021-04-15 16:22:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][740/1251] eta 0:02:25 lr 0.000979 time 0.2845 (0.2857) loss 4.7164 (4.2690) grad_norm 1.1269 (inf) [2021-04-15 16:22:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][750/1251] eta 0:02:23 lr 0.000979 time 0.2772 (0.2856) loss 4.6033 (4.2701) grad_norm 0.9178 (inf) [2021-04-15 16:22:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][760/1251] eta 0:02:20 lr 0.000979 time 0.2711 (0.2854) loss 4.3528 (4.2684) grad_norm 1.2869 (inf) [2021-04-15 16:22:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][770/1251] eta 0:02:17 lr 0.000979 time 0.2888 (0.2853) loss 3.6127 (4.2655) grad_norm 1.1171 (inf) [2021-04-15 16:23:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][780/1251] eta 0:02:14 lr 0.000979 time 0.2672 (0.2853) loss 4.9732 (4.2671) grad_norm 1.0797 (inf) [2021-04-15 16:23:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][790/1251] eta 0:02:11 lr 0.000979 time 0.2828 (0.2852) loss 4.2045 (4.2677) grad_norm 1.2996 (inf) [2021-04-15 16:23:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][800/1251] eta 0:02:08 lr 0.000979 time 0.2574 (0.2851) loss 3.8220 (4.2653) grad_norm 1.1872 (inf) [2021-04-15 16:23:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][810/1251] eta 0:02:05 lr 0.000979 time 0.2756 (0.2850) loss 5.1127 (4.2690) grad_norm 1.3863 (inf) [2021-04-15 16:23:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][820/1251] eta 0:02:02 lr 0.000979 time 0.2886 (0.2849) loss 4.8714 (4.2705) grad_norm 0.9943 (inf) [2021-04-15 16:23:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][830/1251] eta 0:01:59 lr 0.000979 time 0.2720 (0.2848) loss 4.6222 (4.2742) grad_norm 1.1528 (inf) [2021-04-15 16:23:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][840/1251] eta 0:01:56 lr 0.000979 time 0.2675 (0.2847) loss 3.5568 (4.2712) grad_norm 1.0240 (inf) [2021-04-15 16:23:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][850/1251] eta 0:01:54 lr 0.000979 time 0.2752 (0.2845) loss 4.4132 (4.2743) grad_norm 0.9694 (inf) [2021-04-15 16:23:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][860/1251] eta 0:01:51 lr 0.000979 time 0.2759 (0.2845) loss 3.4918 (4.2738) grad_norm 1.2253 (inf) [2021-04-15 16:23:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][870/1251] eta 0:01:48 lr 0.000979 time 0.2968 (0.2844) loss 4.2502 (4.2772) grad_norm 1.1640 (inf) [2021-04-15 16:23:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][880/1251] eta 0:01:45 lr 0.000979 time 0.2774 (0.2843) loss 4.8930 (4.2794) grad_norm 1.0033 (inf) [2021-04-15 16:23:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][890/1251] eta 0:01:42 lr 0.000979 time 0.2744 (0.2843) loss 3.4161 (4.2774) grad_norm 1.3583 (inf) [2021-04-15 16:23:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][900/1251] eta 0:01:39 lr 0.000979 time 0.2982 (0.2842) loss 4.5173 (4.2765) grad_norm 1.0845 (inf) [2021-04-15 16:23:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][910/1251] eta 0:01:36 lr 0.000979 time 0.2763 (0.2841) loss 4.7656 (4.2746) grad_norm 0.9618 (inf) [2021-04-15 16:23:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][920/1251] eta 0:01:34 lr 0.000979 time 0.2826 (0.2841) loss 3.9247 (4.2695) grad_norm 1.2394 (inf) [2021-04-15 16:23:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][930/1251] eta 0:01:31 lr 0.000979 time 0.2852 (0.2841) loss 4.7766 (4.2690) grad_norm 1.3857 (inf) [2021-04-15 16:23:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][940/1251] eta 0:01:28 lr 0.000979 time 0.2648 (0.2840) loss 4.3843 (4.2662) grad_norm 1.0552 (inf) [2021-04-15 16:23:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][950/1251] eta 0:01:25 lr 0.000979 time 0.2674 (0.2840) loss 4.8518 (4.2652) grad_norm 1.0397 (inf) [2021-04-15 16:23:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][960/1251] eta 0:01:22 lr 0.000979 time 0.2613 (0.2839) loss 4.3840 (4.2647) grad_norm 1.1398 (inf) [2021-04-15 16:23:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][970/1251] eta 0:01:19 lr 0.000979 time 0.2682 (0.2838) loss 3.6225 (4.2636) grad_norm 1.1577 (inf) [2021-04-15 16:23:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][980/1251] eta 0:01:16 lr 0.000979 time 0.2861 (0.2838) loss 4.2565 (4.2609) grad_norm 1.3614 (inf) [2021-04-15 16:23:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][990/1251] eta 0:01:14 lr 0.000979 time 0.2744 (0.2838) loss 4.6778 (4.2626) grad_norm 1.1709 (inf) [2021-04-15 16:24:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][1000/1251] eta 0:01:11 lr 0.000979 time 0.2843 (0.2837) loss 4.3169 (4.2606) grad_norm 0.9815 (inf) [2021-04-15 16:24:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][1010/1251] eta 0:01:08 lr 0.000979 time 0.2633 (0.2837) loss 3.3742 (4.2624) grad_norm 1.0703 (inf) [2021-04-15 16:24:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][1020/1251] eta 0:01:05 lr 0.000979 time 0.2975 (0.2836) loss 4.6993 (4.2616) grad_norm 1.0089 (inf) [2021-04-15 16:24:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][1030/1251] eta 0:01:02 lr 0.000979 time 0.2748 (0.2835) loss 3.6433 (4.2611) grad_norm 0.9138 (inf) [2021-04-15 16:24:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][1040/1251] eta 0:00:59 lr 0.000979 time 0.2771 (0.2834) loss 3.5887 (4.2599) grad_norm 1.0285 (inf) [2021-04-15 16:24:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][1050/1251] eta 0:00:56 lr 0.000979 time 0.2655 (0.2833) loss 5.1671 (4.2590) grad_norm 1.2104 (inf) [2021-04-15 16:24:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][1060/1251] eta 0:00:54 lr 0.000979 time 0.3059 (0.2833) loss 4.6485 (4.2602) grad_norm 1.1125 (inf) [2021-04-15 16:24:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][1070/1251] eta 0:00:51 lr 0.000979 time 0.2678 (0.2832) loss 5.1132 (4.2633) grad_norm 0.9129 (inf) [2021-04-15 16:24:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][1080/1251] eta 0:00:48 lr 0.000979 time 0.2691 (0.2832) loss 3.4098 (4.2614) grad_norm 1.1208 (inf) [2021-04-15 16:24:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][1090/1251] eta 0:00:45 lr 0.000979 time 0.2771 (0.2831) loss 4.5692 (4.2593) grad_norm 1.1641 (inf) [2021-04-15 16:24:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][1100/1251] eta 0:00:42 lr 0.000979 time 0.2686 (0.2830) loss 3.8970 (4.2604) grad_norm 0.8634 (inf) [2021-04-15 16:24:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][1110/1251] eta 0:00:39 lr 0.000979 time 0.2757 (0.2830) loss 3.5636 (4.2615) grad_norm 1.1307 (inf) [2021-04-15 16:24:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][1120/1251] eta 0:00:37 lr 0.000979 time 0.2773 (0.2830) loss 3.9966 (4.2618) grad_norm 1.2327 (inf) [2021-04-15 16:24:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][1130/1251] eta 0:00:34 lr 0.000979 time 0.2752 (0.2830) loss 3.7294 (4.2634) grad_norm 1.2309 (inf) [2021-04-15 16:24:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][1140/1251] eta 0:00:31 lr 0.000979 time 0.2866 (0.2829) loss 3.9947 (4.2645) grad_norm 0.9122 (inf) [2021-04-15 16:24:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][1150/1251] eta 0:00:28 lr 0.000979 time 0.2808 (0.2828) loss 3.2241 (4.2650) grad_norm 1.1352 (inf) [2021-04-15 16:24:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][1160/1251] eta 0:00:25 lr 0.000979 time 0.2818 (0.2828) loss 4.8865 (4.2642) grad_norm 0.9748 (inf) [2021-04-15 16:24:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][1170/1251] eta 0:00:22 lr 0.000979 time 0.2794 (0.2827) loss 4.1240 (4.2608) grad_norm 1.4013 (inf) [2021-04-15 16:24:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][1180/1251] eta 0:00:20 lr 0.000979 time 0.2523 (0.2828) loss 4.8857 (4.2615) grad_norm 0.9584 (inf) [2021-04-15 16:24:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][1190/1251] eta 0:00:17 lr 0.000979 time 0.3248 (0.2829) loss 4.9841 (4.2635) grad_norm 1.1090 (inf) [2021-04-15 16:24:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][1200/1251] eta 0:00:14 lr 0.000979 time 0.2796 (0.2828) loss 3.5462 (4.2642) grad_norm 1.0943 (inf) [2021-04-15 16:25:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][1210/1251] eta 0:00:11 lr 0.000979 time 0.2910 (0.2828) loss 4.6298 (4.2642) grad_norm 1.1890 (inf) [2021-04-15 16:25:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][1220/1251] eta 0:00:08 lr 0.000979 time 0.2762 (0.2828) loss 4.7335 (4.2661) grad_norm 1.0207 (inf) [2021-04-15 16:25:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][1230/1251] eta 0:00:05 lr 0.000979 time 0.2931 (0.2827) loss 3.8864 (4.2632) grad_norm 1.2755 (inf) [2021-04-15 16:25:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][1240/1251] eta 0:00:03 lr 0.000979 time 0.2517 (0.2826) loss 3.6608 (4.2636) grad_norm 0.8487 (inf) [2021-04-15 16:25:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [27/300][1250/1251] eta 0:00:00 lr 0.000979 time 0.2479 (0.2824) loss 4.5767 (4.2640) grad_norm 0.9418 (inf) [2021-04-15 16:25:13 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 27 training takes 0:05:55 [2021-04-15 16:25:13 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_27.pth saving...... [2021-04-15 16:25:22 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_27.pth saved !!! [2021-04-15 16:25:23 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.152 (1.152) Loss 1.5562 (1.5562) Acc@1 64.941 (64.941) Acc@5 86.914 (86.914) [2021-04-15 16:25:24 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.094 (0.212) Loss 1.4561 (1.5693) Acc@1 67.578 (64.853) Acc@5 88.184 (87.092) [2021-04-15 16:25:26 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.135 (0.224) Loss 1.6382 (1.5824) Acc@1 63.281 (64.695) Acc@5 85.059 (86.663) [2021-04-15 16:25:29 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.119 (0.226) Loss 1.6670 (1.5872) Acc@1 61.719 (64.652) Acc@5 86.719 (86.627) [2021-04-15 16:25:31 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.222) Loss 1.6275 (1.5836) Acc@1 63.672 (64.648) Acc@5 86.035 (86.664) [2021-04-15 16:25:33 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 64.704 Acc@5 86.652 [2021-04-15 16:25:33 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 64.7% [2021-04-15 16:25:33 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 64.70% [2021-04-15 16:25:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][0/1251] eta 1:12:21 lr 0.000979 time 3.4704 (3.4704) loss 3.1427 (3.1427) grad_norm 1.0982 (1.0982) [2021-04-15 16:25:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][10/1251] eta 0:12:21 lr 0.000979 time 0.2498 (0.5977) loss 2.4869 (3.8451) grad_norm 1.2997 (1.2170) [2021-04-15 16:25:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][20/1251] eta 0:09:12 lr 0.000979 time 0.3014 (0.4484) loss 4.9925 (4.0536) grad_norm 1.1943 (1.1684) [2021-04-15 16:25:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][30/1251] eta 0:08:03 lr 0.000979 time 0.2922 (0.3959) loss 3.7138 (4.0541) grad_norm 1.0706 (1.1497) [2021-04-15 16:25:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][40/1251] eta 0:07:27 lr 0.000979 time 0.2742 (0.3698) loss 4.6196 (4.0835) grad_norm 1.2659 (1.1578) [2021-04-15 16:25:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][50/1251] eta 0:07:04 lr 0.000979 time 0.2782 (0.3531) loss 4.5905 (4.1217) grad_norm 1.4308 (1.1744) [2021-04-15 16:25:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][60/1251] eta 0:06:46 lr 0.000979 time 0.2950 (0.3416) loss 4.2817 (4.1612) grad_norm 1.1735 (1.1780) [2021-04-15 16:25:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][70/1251] eta 0:06:33 lr 0.000979 time 0.2867 (0.3329) loss 3.2440 (4.1560) grad_norm 1.2674 (1.1687) [2021-04-15 16:25:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][80/1251] eta 0:06:23 lr 0.000979 time 0.2896 (0.3271) loss 3.1704 (4.1716) grad_norm 1.1515 (1.1594) [2021-04-15 16:26:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][90/1251] eta 0:06:13 lr 0.000979 time 0.2759 (0.3221) loss 2.9711 (4.1564) grad_norm 0.8605 (1.1609) [2021-04-15 16:26:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][100/1251] eta 0:06:06 lr 0.000979 time 0.2787 (0.3184) loss 3.4580 (4.1573) grad_norm 0.8883 (1.1534) [2021-04-15 16:26:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][110/1251] eta 0:05:59 lr 0.000979 time 0.2879 (0.3153) loss 4.4753 (4.1555) grad_norm 1.0380 (1.1509) [2021-04-15 16:26:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][120/1251] eta 0:05:53 lr 0.000979 time 0.2882 (0.3123) loss 3.8258 (4.1525) grad_norm 0.9588 (1.1520) [2021-04-15 16:26:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][130/1251] eta 0:05:47 lr 0.000979 time 0.2663 (0.3097) loss 5.3392 (4.1646) grad_norm 1.2858 (1.1495) [2021-04-15 16:26:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][140/1251] eta 0:05:42 lr 0.000979 time 0.2714 (0.3081) loss 4.7436 (4.1673) grad_norm 1.2673 (1.1461) [2021-04-15 16:26:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][150/1251] eta 0:05:37 lr 0.000979 time 0.2853 (0.3063) loss 4.7471 (4.1889) grad_norm 1.0549 (1.1443) [2021-04-15 16:26:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][160/1251] eta 0:05:32 lr 0.000979 time 0.2764 (0.3047) loss 4.1080 (4.1825) grad_norm 1.0023 (1.1407) [2021-04-15 16:26:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][170/1251] eta 0:05:27 lr 0.000979 time 0.2781 (0.3032) loss 3.4175 (4.1765) grad_norm 1.3068 (1.1449) [2021-04-15 16:26:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][180/1251] eta 0:05:23 lr 0.000979 time 0.2627 (0.3021) loss 4.3806 (4.2053) grad_norm 1.3199 (1.1481) [2021-04-15 16:26:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][190/1251] eta 0:05:19 lr 0.000979 time 0.2594 (0.3010) loss 3.1089 (4.1922) grad_norm 1.0681 (1.1559) [2021-04-15 16:26:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][200/1251] eta 0:05:15 lr 0.000979 time 0.2676 (0.2998) loss 3.4554 (4.1885) grad_norm 1.1848 (1.1570) [2021-04-15 16:26:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][210/1251] eta 0:05:11 lr 0.000979 time 0.2847 (0.2988) loss 4.6343 (4.1969) grad_norm 0.9514 (1.1567) [2021-04-15 16:26:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][220/1251] eta 0:05:07 lr 0.000979 time 0.2957 (0.2980) loss 4.2329 (4.1895) grad_norm 1.2148 (1.1547) [2021-04-15 16:26:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][230/1251] eta 0:05:03 lr 0.000979 time 0.3033 (0.2973) loss 4.7962 (4.1894) grad_norm 1.0058 (1.1531) [2021-04-15 16:26:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][240/1251] eta 0:05:00 lr 0.000979 time 0.2916 (0.2969) loss 4.3090 (4.2054) grad_norm 1.0474 (1.1471) [2021-04-15 16:26:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][250/1251] eta 0:04:56 lr 0.000979 time 0.2633 (0.2962) loss 4.4371 (4.2141) grad_norm 1.0639 (1.1415) [2021-04-15 16:26:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][260/1251] eta 0:04:53 lr 0.000979 time 0.2768 (0.2958) loss 5.0316 (4.2111) grad_norm 1.0056 (1.1377) [2021-04-15 16:26:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][270/1251] eta 0:04:49 lr 0.000979 time 0.2696 (0.2951) loss 3.8286 (4.2066) grad_norm 1.0779 (1.1389) [2021-04-15 16:26:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][280/1251] eta 0:04:45 lr 0.000979 time 0.2696 (0.2945) loss 4.0899 (4.1951) grad_norm 1.0640 (1.1372) [2021-04-15 16:26:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][290/1251] eta 0:04:42 lr 0.000979 time 0.2799 (0.2941) loss 4.6815 (4.1904) grad_norm 0.9202 (1.1349) [2021-04-15 16:27:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][300/1251] eta 0:04:39 lr 0.000979 time 0.2717 (0.2934) loss 4.7794 (4.1885) grad_norm 1.2310 (1.1379) [2021-04-15 16:27:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][310/1251] eta 0:04:35 lr 0.000979 time 0.2738 (0.2930) loss 4.0629 (4.1910) grad_norm 1.1051 (1.1364) [2021-04-15 16:27:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][320/1251] eta 0:04:32 lr 0.000978 time 0.2801 (0.2931) loss 4.8845 (4.1979) grad_norm 1.0321 (1.1361) [2021-04-15 16:27:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][330/1251] eta 0:04:29 lr 0.000978 time 0.2640 (0.2925) loss 4.9798 (4.2078) grad_norm 1.0581 (1.1358) [2021-04-15 16:27:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][340/1251] eta 0:04:26 lr 0.000978 time 0.3121 (0.2921) loss 4.5094 (4.2004) grad_norm 1.4010 (1.1373) [2021-04-15 16:27:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][350/1251] eta 0:04:22 lr 0.000978 time 0.2938 (0.2918) loss 3.6050 (4.1992) grad_norm 1.4209 (1.1368) [2021-04-15 16:27:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][360/1251] eta 0:04:19 lr 0.000978 time 0.2773 (0.2913) loss 4.2244 (4.1997) grad_norm 1.0944 (1.1370) [2021-04-15 16:27:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][370/1251] eta 0:04:16 lr 0.000978 time 0.2678 (0.2909) loss 4.2537 (4.2075) grad_norm 1.0977 (1.1383) [2021-04-15 16:27:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][380/1251] eta 0:04:13 lr 0.000978 time 0.2756 (0.2908) loss 4.9716 (4.2084) grad_norm 1.3181 (1.1384) [2021-04-15 16:27:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][390/1251] eta 0:04:10 lr 0.000978 time 0.2694 (0.2910) loss 4.5006 (4.2049) grad_norm 0.9458 (1.1407) [2021-04-15 16:27:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][400/1251] eta 0:04:07 lr 0.000978 time 0.2867 (0.2907) loss 4.9237 (4.2093) grad_norm 1.1006 (1.1389) [2021-04-15 16:27:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][410/1251] eta 0:04:04 lr 0.000978 time 0.2735 (0.2904) loss 3.2245 (4.2052) grad_norm 1.2377 (1.1408) [2021-04-15 16:27:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][420/1251] eta 0:04:00 lr 0.000978 time 0.2796 (0.2900) loss 3.3253 (4.2037) grad_norm 1.0782 (1.1408) [2021-04-15 16:27:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][430/1251] eta 0:03:57 lr 0.000978 time 0.2694 (0.2898) loss 4.0501 (4.2077) grad_norm 1.1514 (1.1394) [2021-04-15 16:27:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][440/1251] eta 0:03:54 lr 0.000978 time 0.2617 (0.2896) loss 3.8288 (4.2161) grad_norm 1.2871 (1.1395) [2021-04-15 16:27:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][450/1251] eta 0:03:51 lr 0.000978 time 0.2533 (0.2893) loss 4.1508 (4.2211) grad_norm 1.1515 (1.1410) [2021-04-15 16:27:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][460/1251] eta 0:03:48 lr 0.000978 time 0.2652 (0.2890) loss 4.4948 (4.2148) grad_norm 1.3198 (1.1426) [2021-04-15 16:27:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][470/1251] eta 0:03:45 lr 0.000978 time 0.2723 (0.2888) loss 4.6537 (4.2131) grad_norm 1.0858 (1.1428) [2021-04-15 16:27:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][480/1251] eta 0:03:42 lr 0.000978 time 0.2555 (0.2885) loss 4.9880 (4.2159) grad_norm 1.1407 (1.1420) [2021-04-15 16:27:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][490/1251] eta 0:03:39 lr 0.000978 time 0.2851 (0.2883) loss 5.1883 (4.2220) grad_norm 1.1514 (1.1403) [2021-04-15 16:27:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][500/1251] eta 0:03:36 lr 0.000978 time 0.2755 (0.2882) loss 4.9201 (4.2271) grad_norm 1.0138 (1.1411) [2021-04-15 16:28:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][510/1251] eta 0:03:33 lr 0.000978 time 0.2805 (0.2880) loss 4.3427 (4.2349) grad_norm 1.1820 (1.1419) [2021-04-15 16:28:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][520/1251] eta 0:03:30 lr 0.000978 time 0.2797 (0.2878) loss 4.9499 (4.2350) grad_norm 1.2366 (1.1417) [2021-04-15 16:28:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][530/1251] eta 0:03:27 lr 0.000978 time 0.2862 (0.2876) loss 4.5562 (4.2280) grad_norm 1.1756 (1.1401) [2021-04-15 16:28:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][540/1251] eta 0:03:24 lr 0.000978 time 0.2627 (0.2874) loss 4.7663 (4.2275) grad_norm 1.0406 (1.1403) [2021-04-15 16:28:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][550/1251] eta 0:03:21 lr 0.000978 time 0.2566 (0.2872) loss 4.2473 (4.2228) grad_norm 1.0537 (1.1404) [2021-04-15 16:28:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][560/1251] eta 0:03:18 lr 0.000978 time 0.2666 (0.2871) loss 3.9517 (4.2203) grad_norm 0.9384 (1.1399) [2021-04-15 16:28:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][570/1251] eta 0:03:15 lr 0.000978 time 0.2733 (0.2870) loss 3.6752 (4.2255) grad_norm 0.9590 (1.1383) [2021-04-15 16:28:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][580/1251] eta 0:03:12 lr 0.000978 time 0.2755 (0.2871) loss 4.0452 (4.2300) grad_norm 1.0522 (1.1390) [2021-04-15 16:28:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][590/1251] eta 0:03:09 lr 0.000978 time 0.2730 (0.2870) loss 4.5082 (4.2309) grad_norm 1.0954 (1.1405) [2021-04-15 16:28:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][600/1251] eta 0:03:06 lr 0.000978 time 0.2754 (0.2868) loss 5.0103 (4.2278) grad_norm 1.0237 (1.1401) [2021-04-15 16:28:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][610/1251] eta 0:03:03 lr 0.000978 time 0.2643 (0.2866) loss 4.6773 (4.2312) grad_norm 1.2460 (1.1404) [2021-04-15 16:28:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][620/1251] eta 0:03:00 lr 0.000978 time 0.2959 (0.2864) loss 4.9712 (4.2332) grad_norm 1.0987 (1.1407) [2021-04-15 16:28:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][630/1251] eta 0:02:57 lr 0.000978 time 0.2801 (0.2863) loss 3.8482 (4.2303) grad_norm 1.0927 (1.1402) [2021-04-15 16:28:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][640/1251] eta 0:02:54 lr 0.000978 time 0.2865 (0.2862) loss 4.9574 (4.2314) grad_norm 0.9770 (1.1394) [2021-04-15 16:28:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][650/1251] eta 0:02:51 lr 0.000978 time 0.3004 (0.2862) loss 4.3519 (4.2265) grad_norm 1.2549 (1.1383) [2021-04-15 16:28:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][660/1251] eta 0:02:49 lr 0.000978 time 0.2699 (0.2860) loss 4.2187 (4.2281) grad_norm 1.2081 (1.1388) [2021-04-15 16:28:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][670/1251] eta 0:02:46 lr 0.000978 time 0.2852 (0.2859) loss 3.0303 (4.2236) grad_norm 0.9938 (1.1384) [2021-04-15 16:28:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][680/1251] eta 0:02:43 lr 0.000978 time 0.2620 (0.2858) loss 4.4947 (4.2229) grad_norm 1.0556 (1.1370) [2021-04-15 16:28:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][690/1251] eta 0:02:40 lr 0.000978 time 0.2808 (0.2856) loss 3.4147 (4.2263) grad_norm 0.9552 (1.1373) [2021-04-15 16:28:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][700/1251] eta 0:02:37 lr 0.000978 time 0.2659 (0.2855) loss 4.9701 (4.2280) grad_norm 1.0777 (1.1367) [2021-04-15 16:28:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][710/1251] eta 0:02:34 lr 0.000978 time 0.2735 (0.2854) loss 4.5752 (4.2252) grad_norm 1.0169 (1.1367) [2021-04-15 16:28:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][720/1251] eta 0:02:31 lr 0.000978 time 0.3039 (0.2853) loss 4.0751 (4.2230) grad_norm 1.1028 (1.1381) [2021-04-15 16:29:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][730/1251] eta 0:02:28 lr 0.000978 time 0.2576 (0.2852) loss 3.8454 (4.2219) grad_norm 1.3033 (1.1385) [2021-04-15 16:29:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][740/1251] eta 0:02:25 lr 0.000978 time 0.2736 (0.2851) loss 4.1096 (4.2252) grad_norm 1.2433 (1.1391) [2021-04-15 16:29:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][750/1251] eta 0:02:22 lr 0.000978 time 0.2728 (0.2850) loss 3.7714 (4.2238) grad_norm 1.0870 (1.1400) [2021-04-15 16:29:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][760/1251] eta 0:02:19 lr 0.000978 time 0.2541 (0.2849) loss 4.5500 (4.2208) grad_norm 1.3002 (1.1393) [2021-04-15 16:29:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][770/1251] eta 0:02:17 lr 0.000978 time 0.2686 (0.2849) loss 3.6142 (4.2182) grad_norm 1.1298 (1.1392) [2021-04-15 16:29:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][780/1251] eta 0:02:14 lr 0.000978 time 0.2609 (0.2847) loss 4.8867 (4.2183) grad_norm 1.1340 (1.1389) [2021-04-15 16:29:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][790/1251] eta 0:02:11 lr 0.000978 time 0.2820 (0.2847) loss 3.4380 (4.2148) grad_norm 1.1774 (1.1391) [2021-04-15 16:29:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][800/1251] eta 0:02:08 lr 0.000978 time 0.2745 (0.2846) loss 4.1332 (4.2172) grad_norm 0.9364 (1.1379) [2021-04-15 16:29:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][810/1251] eta 0:02:05 lr 0.000978 time 0.2923 (0.2845) loss 4.2616 (4.2151) grad_norm 1.0004 (1.1375) [2021-04-15 16:29:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][820/1251] eta 0:02:02 lr 0.000978 time 0.2854 (0.2844) loss 4.3633 (4.2172) grad_norm 1.0403 (1.1367) [2021-04-15 16:29:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][830/1251] eta 0:01:59 lr 0.000978 time 0.2913 (0.2843) loss 4.3264 (4.2169) grad_norm 1.1009 (1.1366) [2021-04-15 16:29:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][840/1251] eta 0:01:56 lr 0.000978 time 0.2855 (0.2842) loss 4.3710 (4.2201) grad_norm 0.9466 (1.1366) [2021-04-15 16:29:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][850/1251] eta 0:01:53 lr 0.000978 time 0.2809 (0.2841) loss 4.3337 (4.2175) grad_norm 1.1892 (1.1367) [2021-04-15 16:29:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][860/1251] eta 0:01:51 lr 0.000978 time 0.2701 (0.2840) loss 4.6610 (4.2206) grad_norm 1.1530 (1.1364) [2021-04-15 16:29:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][870/1251] eta 0:01:48 lr 0.000978 time 0.2749 (0.2839) loss 4.7467 (4.2208) grad_norm 1.1081 (1.1355) [2021-04-15 16:29:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][880/1251] eta 0:01:45 lr 0.000978 time 0.2753 (0.2839) loss 3.1675 (4.2197) grad_norm 1.1847 (1.1367) [2021-04-15 16:29:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][890/1251] eta 0:01:42 lr 0.000978 time 0.2656 (0.2839) loss 3.8574 (4.2224) grad_norm 1.2787 (1.1362) [2021-04-15 16:29:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][900/1251] eta 0:01:39 lr 0.000978 time 0.2859 (0.2840) loss 3.4096 (4.2258) grad_norm 1.3154 (1.1359) [2021-04-15 16:29:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][910/1251] eta 0:01:36 lr 0.000978 time 0.3026 (0.2840) loss 4.9018 (4.2275) grad_norm 1.1301 (1.1361) [2021-04-15 16:29:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][920/1251] eta 0:01:33 lr 0.000978 time 0.2530 (0.2838) loss 5.1633 (4.2242) grad_norm 0.9989 (inf) [2021-04-15 16:29:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][930/1251] eta 0:01:31 lr 0.000978 time 0.2546 (0.2838) loss 4.0311 (4.2198) grad_norm 1.0160 (inf) [2021-04-15 16:30:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][940/1251] eta 0:01:28 lr 0.000978 time 0.2795 (0.2837) loss 4.3101 (4.2222) grad_norm 1.2061 (inf) [2021-04-15 16:30:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][950/1251] eta 0:01:25 lr 0.000978 time 0.2889 (0.2836) loss 4.1588 (4.2213) grad_norm 1.0530 (inf) [2021-04-15 16:30:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][960/1251] eta 0:01:22 lr 0.000978 time 0.2767 (0.2837) loss 4.5592 (4.2190) grad_norm 1.0386 (inf) [2021-04-15 16:30:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][970/1251] eta 0:01:19 lr 0.000978 time 0.2706 (0.2839) loss 2.8401 (4.2169) grad_norm 1.1238 (inf) [2021-04-15 16:30:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][980/1251] eta 0:01:16 lr 0.000978 time 0.2708 (0.2838) loss 4.8479 (4.2196) grad_norm 1.1839 (inf) [2021-04-15 16:30:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][990/1251] eta 0:01:14 lr 0.000978 time 0.2870 (0.2837) loss 4.1632 (4.2199) grad_norm 1.2277 (inf) [2021-04-15 16:30:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][1000/1251] eta 0:01:11 lr 0.000978 time 0.2721 (0.2837) loss 4.4424 (4.2211) grad_norm 1.2497 (inf) [2021-04-15 16:30:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][1010/1251] eta 0:01:08 lr 0.000978 time 0.3182 (0.2836) loss 4.9201 (4.2202) grad_norm 1.0589 (inf) [2021-04-15 16:30:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][1020/1251] eta 0:01:05 lr 0.000978 time 0.2815 (0.2836) loss 4.9049 (4.2189) grad_norm 1.0596 (inf) [2021-04-15 16:30:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][1030/1251] eta 0:01:02 lr 0.000978 time 0.2942 (0.2835) loss 4.4852 (4.2197) grad_norm 0.9495 (inf) [2021-04-15 16:30:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][1040/1251] eta 0:00:59 lr 0.000978 time 0.3017 (0.2834) loss 4.5835 (4.2207) grad_norm 1.1437 (inf) [2021-04-15 16:30:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][1050/1251] eta 0:00:56 lr 0.000978 time 0.2942 (0.2834) loss 4.1285 (4.2212) grad_norm 1.4709 (inf) [2021-04-15 16:30:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][1060/1251] eta 0:00:54 lr 0.000978 time 0.2766 (0.2833) loss 4.5266 (4.2207) grad_norm 1.1175 (inf) [2021-04-15 16:30:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][1070/1251] eta 0:00:51 lr 0.000978 time 0.2734 (0.2832) loss 4.5937 (4.2222) grad_norm 1.1922 (inf) [2021-04-15 16:30:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][1080/1251] eta 0:00:48 lr 0.000978 time 0.2560 (0.2831) loss 4.1438 (4.2196) grad_norm 1.3919 (inf) [2021-04-15 16:30:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][1090/1251] eta 0:00:45 lr 0.000978 time 0.3125 (0.2831) loss 4.6224 (4.2210) grad_norm 1.1287 (inf) [2021-04-15 16:30:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][1100/1251] eta 0:00:42 lr 0.000978 time 0.2661 (0.2831) loss 4.6593 (4.2187) grad_norm 1.1006 (inf) [2021-04-15 16:30:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][1110/1251] eta 0:00:39 lr 0.000978 time 0.2675 (0.2830) loss 4.9950 (4.2190) grad_norm 1.2855 (inf) [2021-04-15 16:30:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][1120/1251] eta 0:00:37 lr 0.000978 time 0.2736 (0.2830) loss 4.3962 (4.2162) grad_norm 0.9759 (inf) [2021-04-15 16:30:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][1130/1251] eta 0:00:34 lr 0.000977 time 0.2626 (0.2829) loss 4.4141 (4.2187) grad_norm 1.2283 (inf) [2021-04-15 16:30:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][1140/1251] eta 0:00:31 lr 0.000977 time 0.2516 (0.2829) loss 3.6905 (4.2188) grad_norm 0.9709 (inf) [2021-04-15 16:30:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][1150/1251] eta 0:00:28 lr 0.000977 time 0.2697 (0.2828) loss 4.5022 (4.2203) grad_norm 0.9377 (inf) [2021-04-15 16:31:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][1160/1251] eta 0:00:25 lr 0.000977 time 0.2809 (0.2829) loss 3.9942 (4.2224) grad_norm 1.3411 (inf) [2021-04-15 16:31:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][1170/1251] eta 0:00:22 lr 0.000977 time 0.2585 (0.2828) loss 4.0458 (4.2216) grad_norm 1.2135 (inf) [2021-04-15 16:31:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][1180/1251] eta 0:00:20 lr 0.000977 time 0.2680 (0.2828) loss 4.7531 (4.2215) grad_norm 1.2495 (inf) [2021-04-15 16:31:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][1190/1251] eta 0:00:17 lr 0.000977 time 0.2728 (0.2827) loss 4.7392 (4.2188) grad_norm 1.4078 (inf) [2021-04-15 16:31:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][1200/1251] eta 0:00:14 lr 0.000977 time 0.2841 (0.2827) loss 4.0825 (4.2169) grad_norm 0.8632 (inf) [2021-04-15 16:31:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][1210/1251] eta 0:00:11 lr 0.000977 time 0.2914 (0.2827) loss 4.7583 (4.2181) grad_norm 1.0212 (inf) [2021-04-15 16:31:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][1220/1251] eta 0:00:08 lr 0.000977 time 0.2627 (0.2826) loss 4.4839 (4.2183) grad_norm 1.0466 (inf) [2021-04-15 16:31:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][1230/1251] eta 0:00:05 lr 0.000977 time 0.2439 (0.2825) loss 3.8698 (4.2182) grad_norm 1.0714 (inf) [2021-04-15 16:31:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][1240/1251] eta 0:00:03 lr 0.000977 time 0.2491 (0.2824) loss 3.8658 (4.2153) grad_norm 1.0188 (inf) [2021-04-15 16:31:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [28/300][1250/1251] eta 0:00:00 lr 0.000977 time 0.2495 (0.2821) loss 4.3948 (4.2173) grad_norm 1.2884 (inf) [2021-04-15 16:31:28 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 28 training takes 0:05:55 [2021-04-15 16:31:28 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_28.pth saving...... [2021-04-15 16:31:37 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_28.pth saved !!! [2021-04-15 16:31:38 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.079 (1.079) Loss 1.6000 (1.6000) Acc@1 63.184 (63.184) Acc@5 85.547 (85.547) [2021-04-15 16:31:40 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.100 (0.205) Loss 1.5103 (1.4989) Acc@1 65.527 (65.625) Acc@5 87.012 (86.825) [2021-04-15 16:31:41 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.156 (0.197) Loss 1.4769 (1.5028) Acc@1 65.137 (65.286) Acc@5 88.281 (86.928) [2021-04-15 16:31:44 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.130 (0.224) Loss 1.5205 (1.5093) Acc@1 64.844 (65.080) Acc@5 87.598 (86.990) [2021-04-15 16:31:46 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.284 (0.215) Loss 1.4935 (1.5168) Acc@1 65.918 (64.984) Acc@5 86.914 (86.914) [2021-04-15 16:31:48 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 64.918 Acc@5 86.870 [2021-04-15 16:31:48 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 64.9% [2021-04-15 16:31:48 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 64.92% [2021-04-15 16:31:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][0/1251] eta 1:23:04 lr 0.000977 time 3.9847 (3.9847) loss 4.5781 (4.5781) grad_norm 1.0567 (1.0567) [2021-04-15 16:31:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][10/1251] eta 0:12:52 lr 0.000977 time 0.2851 (0.6222) loss 4.1285 (4.3488) grad_norm 1.1835 (1.2148) [2021-04-15 16:31:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][20/1251] eta 0:09:25 lr 0.000977 time 0.2790 (0.4592) loss 3.4412 (4.1406) grad_norm 1.3896 (1.2645) [2021-04-15 16:32:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][30/1251] eta 0:08:10 lr 0.000977 time 0.2861 (0.4020) loss 3.4572 (4.2192) grad_norm 1.0774 (1.2462) [2021-04-15 16:32:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][40/1251] eta 0:07:30 lr 0.000977 time 0.2878 (0.3722) loss 3.7038 (4.1604) grad_norm 1.0703 (1.2202) [2021-04-15 16:32:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][50/1251] eta 0:07:05 lr 0.000977 time 0.2793 (0.3540) loss 4.7101 (4.1687) grad_norm 1.4648 (1.2195) [2021-04-15 16:32:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][60/1251] eta 0:06:50 lr 0.000977 time 0.2927 (0.3444) loss 3.8927 (4.2189) grad_norm 1.3078 (1.2142) [2021-04-15 16:32:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][70/1251] eta 0:06:35 lr 0.000977 time 0.2922 (0.3349) loss 3.4010 (4.1946) grad_norm 1.1200 (1.2089) [2021-04-15 16:32:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][80/1251] eta 0:06:24 lr 0.000977 time 0.2566 (0.3282) loss 3.4242 (4.2195) grad_norm 1.0757 (1.1993) [2021-04-15 16:32:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][90/1251] eta 0:06:15 lr 0.000977 time 0.3196 (0.3232) loss 4.4054 (4.2365) grad_norm 1.2535 (1.2059) [2021-04-15 16:32:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][100/1251] eta 0:06:07 lr 0.000977 time 0.2840 (0.3193) loss 3.8558 (4.2472) grad_norm 1.0515 (1.1996) [2021-04-15 16:32:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][110/1251] eta 0:06:00 lr 0.000977 time 0.2814 (0.3162) loss 4.1507 (4.2488) grad_norm 0.9894 (1.1924) [2021-04-15 16:32:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][120/1251] eta 0:05:54 lr 0.000977 time 0.2765 (0.3134) loss 4.4202 (4.2315) grad_norm 1.1103 (1.1961) [2021-04-15 16:32:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][130/1251] eta 0:05:48 lr 0.000977 time 0.2823 (0.3110) loss 3.1342 (4.2314) grad_norm 1.0758 (1.1798) [2021-04-15 16:32:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][140/1251] eta 0:05:43 lr 0.000977 time 0.2677 (0.3088) loss 4.6138 (4.2301) grad_norm 0.9831 (1.1738) [2021-04-15 16:32:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][150/1251] eta 0:05:37 lr 0.000977 time 0.2816 (0.3068) loss 4.1624 (4.2301) grad_norm 0.9451 (1.1659) [2021-04-15 16:32:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][160/1251] eta 0:05:33 lr 0.000977 time 0.2914 (0.3053) loss 4.9455 (4.2489) grad_norm 1.0121 (1.1595) [2021-04-15 16:32:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][170/1251] eta 0:05:30 lr 0.000977 time 0.2412 (0.3053) loss 3.2388 (4.2324) grad_norm 1.2203 (1.1530) [2021-04-15 16:32:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][180/1251] eta 0:05:26 lr 0.000977 time 0.2753 (0.3048) loss 4.0876 (4.2022) grad_norm 1.1683 (1.1486) [2021-04-15 16:32:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][190/1251] eta 0:05:21 lr 0.000977 time 0.2899 (0.3033) loss 4.3435 (4.2169) grad_norm 0.8043 (1.1438) [2021-04-15 16:32:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][200/1251] eta 0:05:17 lr 0.000977 time 0.3013 (0.3023) loss 4.5476 (4.2134) grad_norm 1.2168 (1.1397) [2021-04-15 16:32:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][210/1251] eta 0:05:13 lr 0.000977 time 0.2755 (0.3011) loss 3.9454 (4.2066) grad_norm 1.0774 (1.1345) [2021-04-15 16:32:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][220/1251] eta 0:05:09 lr 0.000977 time 0.2619 (0.3000) loss 3.4451 (4.2033) grad_norm 1.0803 (1.1364) [2021-04-15 16:32:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][230/1251] eta 0:05:05 lr 0.000977 time 0.2735 (0.2989) loss 4.4796 (4.2079) grad_norm 1.3821 (1.1378) [2021-04-15 16:33:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][240/1251] eta 0:05:01 lr 0.000977 time 0.2656 (0.2981) loss 4.4287 (4.2159) grad_norm 1.0146 (1.1403) [2021-04-15 16:33:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][250/1251] eta 0:04:57 lr 0.000977 time 0.2649 (0.2974) loss 4.8747 (4.2133) grad_norm 1.1586 (1.1448) [2021-04-15 16:33:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][260/1251] eta 0:04:53 lr 0.000977 time 0.2609 (0.2965) loss 3.7876 (4.2163) grad_norm 0.9341 (1.1426) [2021-04-15 16:33:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][270/1251] eta 0:04:50 lr 0.000977 time 0.2933 (0.2959) loss 3.2279 (4.2084) grad_norm 1.1062 (1.1396) [2021-04-15 16:33:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][280/1251] eta 0:04:46 lr 0.000977 time 0.2618 (0.2954) loss 3.9514 (4.2078) grad_norm 0.9486 (1.1364) [2021-04-15 16:33:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][290/1251] eta 0:04:43 lr 0.000977 time 0.2678 (0.2948) loss 4.5703 (4.2067) grad_norm 1.1397 (1.1363) [2021-04-15 16:33:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][300/1251] eta 0:04:39 lr 0.000977 time 0.2643 (0.2944) loss 3.5664 (4.2068) grad_norm 1.0334 (1.1351) [2021-04-15 16:33:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][310/1251] eta 0:04:36 lr 0.000977 time 0.2665 (0.2938) loss 3.8699 (4.1961) grad_norm 1.3466 (1.1354) [2021-04-15 16:33:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][320/1251] eta 0:04:33 lr 0.000977 time 0.3028 (0.2934) loss 4.8406 (4.1968) grad_norm 1.0373 (1.1351) [2021-04-15 16:33:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][330/1251] eta 0:04:29 lr 0.000977 time 0.2747 (0.2931) loss 4.5290 (4.2013) grad_norm 1.4895 (1.1343) [2021-04-15 16:33:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][340/1251] eta 0:04:26 lr 0.000977 time 0.2616 (0.2927) loss 4.2833 (4.1966) grad_norm 1.0632 (1.1345) [2021-04-15 16:33:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][350/1251] eta 0:04:23 lr 0.000977 time 0.2571 (0.2923) loss 3.0915 (4.1945) grad_norm 1.2113 (1.1351) [2021-04-15 16:33:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][360/1251] eta 0:04:20 lr 0.000977 time 0.2869 (0.2925) loss 5.1393 (4.2003) grad_norm 1.1674 (1.1320) [2021-04-15 16:33:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][370/1251] eta 0:04:17 lr 0.000977 time 0.2808 (0.2920) loss 4.5546 (4.1958) grad_norm 1.1742 (1.1307) [2021-04-15 16:33:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][380/1251] eta 0:04:13 lr 0.000977 time 0.2845 (0.2916) loss 3.6002 (4.1970) grad_norm 1.2300 (1.1284) [2021-04-15 16:33:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][390/1251] eta 0:04:10 lr 0.000977 time 0.3093 (0.2914) loss 4.8117 (4.1960) grad_norm 1.2560 (1.1295) [2021-04-15 16:33:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][400/1251] eta 0:04:07 lr 0.000977 time 0.2610 (0.2910) loss 4.4258 (4.2019) grad_norm 1.0426 (1.1276) [2021-04-15 16:33:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][410/1251] eta 0:04:04 lr 0.000977 time 0.2933 (0.2906) loss 4.5800 (4.2050) grad_norm 1.0931 (1.1264) [2021-04-15 16:33:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][420/1251] eta 0:04:01 lr 0.000977 time 0.2709 (0.2902) loss 3.1705 (4.1995) grad_norm 1.1837 (1.1268) [2021-04-15 16:33:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][430/1251] eta 0:03:57 lr 0.000977 time 0.2489 (0.2898) loss 4.1022 (4.2045) grad_norm 1.0452 (1.1266) [2021-04-15 16:33:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][440/1251] eta 0:03:54 lr 0.000977 time 0.2854 (0.2896) loss 4.2667 (4.2004) grad_norm 1.0317 (1.1250) [2021-04-15 16:33:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][450/1251] eta 0:03:51 lr 0.000977 time 0.2566 (0.2894) loss 4.0726 (4.1978) grad_norm 1.0621 (1.1260) [2021-04-15 16:34:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][460/1251] eta 0:03:48 lr 0.000977 time 0.2463 (0.2890) loss 4.8766 (4.2041) grad_norm 1.3004 (1.1273) [2021-04-15 16:34:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][470/1251] eta 0:03:45 lr 0.000977 time 0.2769 (0.2888) loss 5.0799 (4.1971) grad_norm 1.0556 (1.1279) [2021-04-15 16:34:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][480/1251] eta 0:03:42 lr 0.000977 time 0.2582 (0.2886) loss 3.9094 (4.1965) grad_norm 1.2648 (1.1284) [2021-04-15 16:34:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][490/1251] eta 0:03:39 lr 0.000977 time 0.3002 (0.2883) loss 4.2776 (4.1993) grad_norm 1.1430 (1.1292) [2021-04-15 16:34:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][500/1251] eta 0:03:36 lr 0.000977 time 0.2863 (0.2881) loss 5.1054 (4.2004) grad_norm 1.2981 (1.1293) [2021-04-15 16:34:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][510/1251] eta 0:03:33 lr 0.000977 time 0.2679 (0.2879) loss 3.4797 (4.2009) grad_norm 1.4623 (1.1308) [2021-04-15 16:34:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][520/1251] eta 0:03:30 lr 0.000977 time 0.2889 (0.2877) loss 4.4932 (4.1969) grad_norm 1.0763 (1.1298) [2021-04-15 16:34:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][530/1251] eta 0:03:27 lr 0.000977 time 0.2826 (0.2874) loss 4.3048 (4.1973) grad_norm 1.0636 (1.1294) [2021-04-15 16:34:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][540/1251] eta 0:03:24 lr 0.000977 time 0.3018 (0.2873) loss 2.8844 (4.1927) grad_norm 1.4617 (1.1288) [2021-04-15 16:34:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][550/1251] eta 0:03:21 lr 0.000977 time 0.2816 (0.2871) loss 3.4187 (4.1924) grad_norm 1.1046 (1.1286) [2021-04-15 16:34:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][560/1251] eta 0:03:18 lr 0.000977 time 0.2874 (0.2870) loss 3.4812 (4.1959) grad_norm 1.0939 (1.1276) [2021-04-15 16:34:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][570/1251] eta 0:03:15 lr 0.000977 time 0.2826 (0.2869) loss 4.1438 (4.1966) grad_norm 0.9640 (1.1275) [2021-04-15 16:34:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][580/1251] eta 0:03:12 lr 0.000977 time 0.2878 (0.2867) loss 4.7216 (4.2018) grad_norm 1.0383 (1.1277) [2021-04-15 16:34:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][590/1251] eta 0:03:09 lr 0.000977 time 0.2805 (0.2865) loss 3.0411 (4.2051) grad_norm 1.3650 (1.1276) [2021-04-15 16:34:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][600/1251] eta 0:03:06 lr 0.000977 time 0.2855 (0.2863) loss 3.7207 (4.2040) grad_norm 1.2910 (1.1283) [2021-04-15 16:34:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][610/1251] eta 0:03:03 lr 0.000977 time 0.2625 (0.2862) loss 4.4841 (4.2031) grad_norm 1.0908 (1.1287) [2021-04-15 16:34:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][620/1251] eta 0:03:00 lr 0.000977 time 0.2581 (0.2861) loss 4.2023 (4.2016) grad_norm 0.9374 (1.1289) [2021-04-15 16:34:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][630/1251] eta 0:02:57 lr 0.000977 time 0.2709 (0.2862) loss 4.3408 (4.2048) grad_norm 0.9772 (1.1284) [2021-04-15 16:34:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][640/1251] eta 0:02:54 lr 0.000977 time 0.2825 (0.2861) loss 4.2264 (4.2035) grad_norm 1.2619 (1.1279) [2021-04-15 16:34:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][650/1251] eta 0:02:51 lr 0.000977 time 0.2834 (0.2859) loss 3.2828 (4.2013) grad_norm 1.0308 (1.1270) [2021-04-15 16:34:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][660/1251] eta 0:02:48 lr 0.000977 time 0.2712 (0.2859) loss 3.5803 (4.2070) grad_norm 1.0083 (1.1258) [2021-04-15 16:35:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][670/1251] eta 0:02:46 lr 0.000977 time 0.2949 (0.2858) loss 3.8459 (4.2065) grad_norm 1.0229 (1.1245) [2021-04-15 16:35:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][680/1251] eta 0:02:43 lr 0.000976 time 0.2727 (0.2856) loss 3.5196 (4.2082) grad_norm 1.1314 (1.1246) [2021-04-15 16:35:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][690/1251] eta 0:02:40 lr 0.000976 time 0.2668 (0.2854) loss 3.8186 (4.2098) grad_norm 1.0934 (1.1262) [2021-04-15 16:35:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][700/1251] eta 0:02:37 lr 0.000976 time 0.2859 (0.2854) loss 4.4918 (4.2115) grad_norm 0.9405 (1.1261) [2021-04-15 16:35:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][710/1251] eta 0:02:34 lr 0.000976 time 0.2784 (0.2854) loss 3.0378 (4.2131) grad_norm 0.9893 (1.1255) [2021-04-15 16:35:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][720/1251] eta 0:02:31 lr 0.000976 time 0.2677 (0.2852) loss 4.6950 (4.2150) grad_norm 1.3618 (1.1264) [2021-04-15 16:35:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][730/1251] eta 0:02:28 lr 0.000976 time 0.2562 (0.2851) loss 4.1611 (4.2166) grad_norm 0.9678 (1.1259) [2021-04-15 16:35:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][740/1251] eta 0:02:25 lr 0.000976 time 0.2793 (0.2851) loss 4.7718 (4.2191) grad_norm 0.9373 (1.1250) [2021-04-15 16:35:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][750/1251] eta 0:02:22 lr 0.000976 time 0.2667 (0.2851) loss 3.9083 (4.2212) grad_norm 1.0687 (1.1246) [2021-04-15 16:35:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][760/1251] eta 0:02:20 lr 0.000976 time 0.2664 (0.2851) loss 3.6137 (4.2242) grad_norm 0.8204 (1.1231) [2021-04-15 16:35:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][770/1251] eta 0:02:17 lr 0.000976 time 0.2661 (0.2850) loss 4.6750 (4.2277) grad_norm 1.1389 (1.1228) [2021-04-15 16:35:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][780/1251] eta 0:02:14 lr 0.000976 time 0.2773 (0.2850) loss 3.9166 (4.2229) grad_norm 1.2968 (1.1234) [2021-04-15 16:35:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][790/1251] eta 0:02:11 lr 0.000976 time 0.3039 (0.2850) loss 4.6893 (4.2238) grad_norm 1.0754 (1.1232) [2021-04-15 16:35:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][800/1251] eta 0:02:08 lr 0.000976 time 0.2781 (0.2849) loss 3.2898 (4.2209) grad_norm 0.8552 (1.1221) [2021-04-15 16:35:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][810/1251] eta 0:02:05 lr 0.000976 time 0.2849 (0.2848) loss 4.4286 (4.2247) grad_norm 1.4605 (1.1219) [2021-04-15 16:35:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][820/1251] eta 0:02:02 lr 0.000976 time 0.2781 (0.2847) loss 4.2746 (4.2228) grad_norm 1.0378 (1.1224) [2021-04-15 16:35:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][830/1251] eta 0:01:59 lr 0.000976 time 0.2733 (0.2846) loss 4.3242 (4.2217) grad_norm 0.9463 (1.1215) [2021-04-15 16:35:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][840/1251] eta 0:01:56 lr 0.000976 time 0.2878 (0.2845) loss 2.9465 (4.2193) grad_norm 1.3491 (inf) [2021-04-15 16:35:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][850/1251] eta 0:01:54 lr 0.000976 time 0.2569 (0.2844) loss 4.0739 (4.2192) grad_norm 1.1892 (inf) [2021-04-15 16:35:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][860/1251] eta 0:01:51 lr 0.000976 time 0.2596 (0.2843) loss 4.5785 (4.2183) grad_norm 1.2942 (inf) [2021-04-15 16:35:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][870/1251] eta 0:01:48 lr 0.000976 time 0.2733 (0.2843) loss 4.8150 (4.2199) grad_norm 1.1376 (inf) [2021-04-15 16:35:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][880/1251] eta 0:01:45 lr 0.000976 time 0.2561 (0.2842) loss 3.1106 (4.2194) grad_norm 1.1977 (inf) [2021-04-15 16:36:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][890/1251] eta 0:01:42 lr 0.000976 time 0.2580 (0.2842) loss 4.4827 (4.2220) grad_norm 1.1663 (inf) [2021-04-15 16:36:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][900/1251] eta 0:01:39 lr 0.000976 time 0.2679 (0.2841) loss 3.6166 (4.2228) grad_norm 1.7536 (inf) [2021-04-15 16:36:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][910/1251] eta 0:01:36 lr 0.000976 time 0.2566 (0.2840) loss 3.0980 (4.2219) grad_norm 1.3314 (inf) [2021-04-15 16:36:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][920/1251] eta 0:01:33 lr 0.000976 time 0.2681 (0.2839) loss 3.0079 (4.2206) grad_norm 1.2315 (inf) [2021-04-15 16:36:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][930/1251] eta 0:01:31 lr 0.000976 time 0.4808 (0.2841) loss 4.5087 (4.2175) grad_norm 0.9499 (inf) [2021-04-15 16:36:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][940/1251] eta 0:01:28 lr 0.000976 time 0.2871 (0.2840) loss 3.9223 (4.2158) grad_norm 0.9650 (inf) [2021-04-15 16:36:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][950/1251] eta 0:01:25 lr 0.000976 time 0.2821 (0.2839) loss 3.8150 (4.2134) grad_norm 1.0518 (inf) [2021-04-15 16:36:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][960/1251] eta 0:01:22 lr 0.000976 time 0.2715 (0.2838) loss 4.5933 (4.2162) grad_norm 1.1189 (inf) [2021-04-15 16:36:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][970/1251] eta 0:01:19 lr 0.000976 time 0.2669 (0.2837) loss 3.2613 (4.2164) grad_norm 1.1690 (inf) [2021-04-15 16:36:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][980/1251] eta 0:01:16 lr 0.000976 time 0.2857 (0.2836) loss 4.9364 (4.2174) grad_norm 1.2243 (inf) [2021-04-15 16:36:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][990/1251] eta 0:01:13 lr 0.000976 time 0.2767 (0.2835) loss 2.8285 (4.2200) grad_norm 1.0811 (inf) [2021-04-15 16:36:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][1000/1251] eta 0:01:11 lr 0.000976 time 0.2773 (0.2834) loss 3.7781 (4.2194) grad_norm 1.2434 (inf) [2021-04-15 16:36:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][1010/1251] eta 0:01:08 lr 0.000976 time 0.2819 (0.2834) loss 3.1978 (4.2201) grad_norm 1.1730 (inf) [2021-04-15 16:36:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][1020/1251] eta 0:01:05 lr 0.000976 time 0.2847 (0.2833) loss 3.2214 (4.2204) grad_norm 1.1375 (inf) [2021-04-15 16:36:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][1030/1251] eta 0:01:02 lr 0.000976 time 0.2816 (0.2833) loss 4.2236 (4.2156) grad_norm 1.1686 (inf) [2021-04-15 16:36:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][1040/1251] eta 0:00:59 lr 0.000976 time 0.2560 (0.2832) loss 5.1948 (4.2176) grad_norm 1.3213 (inf) [2021-04-15 16:36:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][1050/1251] eta 0:00:56 lr 0.000976 time 0.3061 (0.2832) loss 3.0993 (4.2154) grad_norm 1.1975 (inf) [2021-04-15 16:36:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][1060/1251] eta 0:00:54 lr 0.000976 time 0.2916 (0.2832) loss 3.1026 (4.2156) grad_norm 1.1045 (inf) [2021-04-15 16:36:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][1070/1251] eta 0:00:51 lr 0.000976 time 0.2817 (0.2831) loss 3.9459 (4.2160) grad_norm 1.3909 (inf) [2021-04-15 16:36:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][1080/1251] eta 0:00:48 lr 0.000976 time 0.2611 (0.2830) loss 3.4013 (4.2146) grad_norm 1.3761 (inf) [2021-04-15 16:36:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][1090/1251] eta 0:00:45 lr 0.000976 time 0.2580 (0.2829) loss 3.8373 (4.2149) grad_norm 1.2918 (inf) [2021-04-15 16:37:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][1100/1251] eta 0:00:42 lr 0.000976 time 0.2613 (0.2828) loss 4.7971 (4.2139) grad_norm 1.3603 (inf) [2021-04-15 16:37:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][1110/1251] eta 0:00:39 lr 0.000976 time 0.2441 (0.2828) loss 3.8985 (4.2118) grad_norm 1.4618 (inf) [2021-04-15 16:37:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][1120/1251] eta 0:00:37 lr 0.000976 time 0.2883 (0.2827) loss 4.8159 (4.2099) grad_norm 1.0476 (inf) [2021-04-15 16:37:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][1130/1251] eta 0:00:34 lr 0.000976 time 0.2800 (0.2827) loss 4.9659 (4.2123) grad_norm 1.1697 (inf) [2021-04-15 16:37:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][1140/1251] eta 0:00:31 lr 0.000976 time 0.2876 (0.2826) loss 4.2311 (4.2115) grad_norm 1.2279 (inf) [2021-04-15 16:37:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][1150/1251] eta 0:00:28 lr 0.000976 time 0.2896 (0.2826) loss 4.3513 (4.2118) grad_norm 1.4660 (inf) [2021-04-15 16:37:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][1160/1251] eta 0:00:25 lr 0.000976 time 0.2722 (0.2825) loss 3.0586 (4.2114) grad_norm 1.1714 (inf) [2021-04-15 16:37:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][1170/1251] eta 0:00:22 lr 0.000976 time 0.2852 (0.2825) loss 3.0264 (4.2114) grad_norm 1.0312 (inf) [2021-04-15 16:37:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][1180/1251] eta 0:00:20 lr 0.000976 time 0.2802 (0.2824) loss 4.9940 (4.2137) grad_norm 1.1187 (inf) [2021-04-15 16:37:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][1190/1251] eta 0:00:17 lr 0.000976 time 0.2752 (0.2824) loss 3.8029 (4.2129) grad_norm 1.1222 (inf) [2021-04-15 16:37:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][1200/1251] eta 0:00:14 lr 0.000976 time 0.2910 (0.2823) loss 4.0973 (4.2102) grad_norm 1.2326 (inf) [2021-04-15 16:37:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][1210/1251] eta 0:00:11 lr 0.000976 time 0.2564 (0.2823) loss 3.9353 (4.2097) grad_norm 0.9813 (inf) [2021-04-15 16:37:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][1220/1251] eta 0:00:08 lr 0.000976 time 0.2577 (0.2822) loss 4.7011 (4.2077) grad_norm 1.0968 (inf) [2021-04-15 16:37:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][1230/1251] eta 0:00:05 lr 0.000976 time 0.2743 (0.2823) loss 3.9867 (4.2069) grad_norm 0.8419 (inf) [2021-04-15 16:37:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][1240/1251] eta 0:00:03 lr 0.000976 time 0.2488 (0.2823) loss 4.4235 (4.2060) grad_norm 1.5142 (inf) [2021-04-15 16:37:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [29/300][1250/1251] eta 0:00:00 lr 0.000976 time 0.2483 (0.2820) loss 4.3385 (4.2042) grad_norm 1.2292 (inf) [2021-04-15 16:37:43 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 29 training takes 0:05:54 [2021-04-15 16:37:43 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_29.pth saving...... [2021-04-15 16:37:57 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_29.pth saved !!! [2021-04-15 16:37:58 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.152 (1.152) Loss 1.5817 (1.5817) Acc@1 66.797 (66.797) Acc@5 85.742 (85.742) [2021-04-15 16:38:00 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.154 (0.262) Loss 1.6492 (1.5379) Acc@1 61.035 (65.439) Acc@5 85.645 (87.172) [2021-04-15 16:38:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.261 (0.237) Loss 1.5337 (1.5548) Acc@1 65.332 (65.309) Acc@5 88.184 (86.970) [2021-04-15 16:38:05 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.258 (0.245) Loss 1.6014 (1.5565) Acc@1 64.844 (65.452) Acc@5 86.133 (87.012) [2021-04-15 16:38:06 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.102 (0.225) Loss 1.4899 (1.5549) Acc@1 67.578 (65.539) Acc@5 87.793 (87.000) [2021-04-15 16:38:08 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 65.580 Acc@5 87.090 [2021-04-15 16:38:08 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 65.6% [2021-04-15 16:38:08 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 65.58% [2021-04-15 16:38:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][0/1251] eta 1:30:07 lr 0.000976 time 4.3228 (4.3228) loss 4.1943 (4.1943) grad_norm 1.4214 (1.4214) [2021-04-15 16:38:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][10/1251] eta 0:13:17 lr 0.000976 time 0.2729 (0.6427) loss 4.3567 (4.4816) grad_norm 0.9707 (1.2096) [2021-04-15 16:38:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][20/1251] eta 0:09:39 lr 0.000976 time 0.2687 (0.4707) loss 4.2269 (4.4086) grad_norm 1.1071 (1.1569) [2021-04-15 16:38:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][30/1251] eta 0:08:19 lr 0.000976 time 0.2568 (0.4088) loss 4.3293 (4.2935) grad_norm 1.1003 (1.1629) [2021-04-15 16:38:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][40/1251] eta 0:07:37 lr 0.000976 time 0.2820 (0.3777) loss 3.4655 (4.2510) grad_norm 1.1905 (1.1617) [2021-04-15 16:38:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][50/1251] eta 0:07:09 lr 0.000976 time 0.2772 (0.3580) loss 4.2885 (4.2880) grad_norm 1.2630 (1.1526) [2021-04-15 16:38:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][60/1251] eta 0:06:51 lr 0.000976 time 0.2884 (0.3453) loss 4.9395 (4.2395) grad_norm 1.0335 (1.1384) [2021-04-15 16:38:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][70/1251] eta 0:06:36 lr 0.000976 time 0.2555 (0.3361) loss 3.0352 (4.1912) grad_norm 0.9068 (1.1150) [2021-04-15 16:38:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][80/1251] eta 0:06:25 lr 0.000976 time 0.2731 (0.3293) loss 4.5067 (4.1721) grad_norm 1.4104 (1.1148) [2021-04-15 16:38:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][90/1251] eta 0:06:15 lr 0.000976 time 0.2511 (0.3231) loss 4.5547 (4.1319) grad_norm 1.3221 (1.1210) [2021-04-15 16:38:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][100/1251] eta 0:06:07 lr 0.000976 time 0.2893 (0.3193) loss 4.3198 (4.1556) grad_norm 1.4129 (1.1231) [2021-04-15 16:38:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][110/1251] eta 0:06:00 lr 0.000976 time 0.2857 (0.3158) loss 4.2916 (4.1444) grad_norm 0.8585 (1.1224) [2021-04-15 16:38:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][120/1251] eta 0:05:54 lr 0.000976 time 0.2850 (0.3131) loss 4.0545 (4.1520) grad_norm 1.1860 (1.1234) [2021-04-15 16:38:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][130/1251] eta 0:05:48 lr 0.000976 time 0.2826 (0.3106) loss 3.8843 (4.1748) grad_norm 0.9794 (1.1222) [2021-04-15 16:38:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][140/1251] eta 0:05:44 lr 0.000976 time 0.4594 (0.3101) loss 4.7084 (4.1816) grad_norm 1.1991 (1.1266) [2021-04-15 16:38:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][150/1251] eta 0:05:38 lr 0.000976 time 0.2663 (0.3076) loss 4.3197 (4.1735) grad_norm 1.0488 (1.1219) [2021-04-15 16:38:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][160/1251] eta 0:05:33 lr 0.000976 time 0.2858 (0.3061) loss 4.8988 (4.1787) grad_norm 1.0258 (1.1213) [2021-04-15 16:39:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][170/1251] eta 0:05:30 lr 0.000976 time 0.2743 (0.3059) loss 4.6187 (4.1678) grad_norm 0.9339 (1.1170) [2021-04-15 16:39:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][180/1251] eta 0:05:26 lr 0.000976 time 0.2951 (0.3046) loss 4.0671 (4.1604) grad_norm 1.2013 (1.1143) [2021-04-15 16:39:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][190/1251] eta 0:05:22 lr 0.000976 time 0.2883 (0.3036) loss 3.6950 (4.1456) grad_norm 1.1236 (1.1139) [2021-04-15 16:39:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][200/1251] eta 0:05:18 lr 0.000976 time 0.2804 (0.3026) loss 4.4989 (4.1516) grad_norm 1.0447 (1.1133) [2021-04-15 16:39:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][210/1251] eta 0:05:13 lr 0.000976 time 0.2793 (0.3016) loss 4.8949 (4.1712) grad_norm 1.2944 (1.1150) [2021-04-15 16:39:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][220/1251] eta 0:05:09 lr 0.000975 time 0.2696 (0.3004) loss 3.6043 (4.1675) grad_norm 1.1515 (1.1157) [2021-04-15 16:39:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][230/1251] eta 0:05:05 lr 0.000975 time 0.2658 (0.2995) loss 4.6044 (4.1636) grad_norm 1.2042 (1.1189) [2021-04-15 16:39:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][240/1251] eta 0:05:01 lr 0.000975 time 0.3002 (0.2986) loss 3.3797 (4.1603) grad_norm 1.1143 (1.1179) [2021-04-15 16:39:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][250/1251] eta 0:04:58 lr 0.000975 time 0.2830 (0.2981) loss 4.4089 (4.1659) grad_norm 0.8523 (1.1164) [2021-04-15 16:39:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][260/1251] eta 0:04:54 lr 0.000975 time 0.2938 (0.2974) loss 3.1245 (4.1630) grad_norm 1.0219 (1.1155) [2021-04-15 16:39:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][270/1251] eta 0:04:51 lr 0.000975 time 0.2649 (0.2966) loss 3.8222 (4.1590) grad_norm 1.0630 (1.1178) [2021-04-15 16:39:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][280/1251] eta 0:04:47 lr 0.000975 time 0.2668 (0.2960) loss 4.2183 (4.1552) grad_norm 1.0471 (1.1158) [2021-04-15 16:39:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][290/1251] eta 0:04:43 lr 0.000975 time 0.2736 (0.2954) loss 4.5487 (4.1600) grad_norm 1.1143 (1.1136) [2021-04-15 16:39:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][300/1251] eta 0:04:40 lr 0.000975 time 0.2718 (0.2947) loss 4.4662 (4.1658) grad_norm 1.1578 (1.1118) [2021-04-15 16:39:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][310/1251] eta 0:04:36 lr 0.000975 time 0.2695 (0.2942) loss 5.0595 (4.1677) grad_norm 1.0110 (1.1100) [2021-04-15 16:39:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][320/1251] eta 0:04:33 lr 0.000975 time 0.2748 (0.2937) loss 4.0818 (4.1736) grad_norm 1.3267 (1.1099) [2021-04-15 16:39:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][330/1251] eta 0:04:29 lr 0.000975 time 0.2857 (0.2931) loss 4.8444 (4.1708) grad_norm 0.9845 (1.1069) [2021-04-15 16:39:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][340/1251] eta 0:04:26 lr 0.000975 time 0.2704 (0.2926) loss 4.8388 (4.1791) grad_norm 1.2957 (1.1056) [2021-04-15 16:39:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][350/1251] eta 0:04:23 lr 0.000975 time 0.2589 (0.2922) loss 3.0248 (4.1788) grad_norm 1.0579 (1.1054) [2021-04-15 16:39:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][360/1251] eta 0:04:20 lr 0.000975 time 0.2549 (0.2918) loss 4.8061 (4.1787) grad_norm 0.9627 (1.1045) [2021-04-15 16:39:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][370/1251] eta 0:04:17 lr 0.000975 time 0.2664 (0.2920) loss 3.7043 (4.1684) grad_norm 1.0263 (1.1038) [2021-04-15 16:39:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][380/1251] eta 0:04:14 lr 0.000975 time 0.2834 (0.2917) loss 4.7247 (4.1669) grad_norm 1.0912 (1.1046) [2021-04-15 16:40:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][390/1251] eta 0:04:10 lr 0.000975 time 0.2725 (0.2912) loss 3.7214 (4.1648) grad_norm 1.0521 (1.1047) [2021-04-15 16:40:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][400/1251] eta 0:04:07 lr 0.000975 time 0.2868 (0.2910) loss 3.1098 (4.1611) grad_norm 1.0962 (1.1030) [2021-04-15 16:40:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][410/1251] eta 0:04:04 lr 0.000975 time 0.2678 (0.2907) loss 4.5387 (4.1720) grad_norm 1.1695 (1.1037) [2021-04-15 16:40:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][420/1251] eta 0:04:01 lr 0.000975 time 0.2750 (0.2906) loss 4.0319 (4.1725) grad_norm 1.3301 (1.1043) [2021-04-15 16:40:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][430/1251] eta 0:03:58 lr 0.000975 time 0.3201 (0.2904) loss 4.6981 (4.1736) grad_norm 0.9416 (1.1041) [2021-04-15 16:40:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][440/1251] eta 0:03:55 lr 0.000975 time 0.2754 (0.2901) loss 4.8504 (4.1744) grad_norm 1.0504 (1.1067) [2021-04-15 16:40:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][450/1251] eta 0:03:52 lr 0.000975 time 0.2474 (0.2898) loss 5.0724 (4.1827) grad_norm 1.0635 (1.1076) [2021-04-15 16:40:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][460/1251] eta 0:03:48 lr 0.000975 time 0.2768 (0.2894) loss 4.4527 (4.1884) grad_norm 1.1837 (1.1091) [2021-04-15 16:40:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][470/1251] eta 0:03:45 lr 0.000975 time 0.2815 (0.2891) loss 5.1112 (4.1930) grad_norm 1.0044 (1.1080) [2021-04-15 16:40:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][480/1251] eta 0:03:42 lr 0.000975 time 0.2891 (0.2889) loss 2.9097 (4.1877) grad_norm 1.2398 (1.1085) [2021-04-15 16:40:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][490/1251] eta 0:03:39 lr 0.000975 time 0.2750 (0.2886) loss 3.0869 (4.1842) grad_norm 1.4437 (1.1125) [2021-04-15 16:40:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][500/1251] eta 0:03:36 lr 0.000975 time 0.2803 (0.2884) loss 3.1410 (4.1823) grad_norm 1.1129 (1.1119) [2021-04-15 16:40:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][510/1251] eta 0:03:33 lr 0.000975 time 0.2770 (0.2883) loss 2.9752 (4.1759) grad_norm 0.8573 (1.1104) [2021-04-15 16:40:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][520/1251] eta 0:03:30 lr 0.000975 time 0.2672 (0.2881) loss 4.2964 (4.1711) grad_norm 1.1104 (1.1098) [2021-04-15 16:40:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][530/1251] eta 0:03:27 lr 0.000975 time 0.2586 (0.2880) loss 4.3140 (4.1727) grad_norm 1.0749 (1.1085) [2021-04-15 16:40:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][540/1251] eta 0:03:24 lr 0.000975 time 0.3004 (0.2878) loss 3.9009 (4.1726) grad_norm 0.9602 (1.1094) [2021-04-15 16:40:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][550/1251] eta 0:03:21 lr 0.000975 time 0.2706 (0.2880) loss 3.0094 (4.1720) grad_norm 1.1403 (1.1102) [2021-04-15 16:40:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][560/1251] eta 0:03:18 lr 0.000975 time 0.2711 (0.2878) loss 3.5720 (4.1704) grad_norm 0.9219 (1.1088) [2021-04-15 16:40:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][570/1251] eta 0:03:16 lr 0.000975 time 0.2490 (0.2878) loss 4.5787 (4.1713) grad_norm 0.9944 (1.1075) [2021-04-15 16:40:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][580/1251] eta 0:03:13 lr 0.000975 time 0.2890 (0.2879) loss 4.4855 (4.1763) grad_norm 0.8577 (1.1057) [2021-04-15 16:40:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][590/1251] eta 0:03:10 lr 0.000975 time 0.2724 (0.2877) loss 3.3681 (4.1750) grad_norm 1.3458 (1.1044) [2021-04-15 16:41:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][600/1251] eta 0:03:07 lr 0.000975 time 0.2776 (0.2876) loss 3.8592 (4.1781) grad_norm 1.2127 (1.1042) [2021-04-15 16:41:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][610/1251] eta 0:03:04 lr 0.000975 time 0.2772 (0.2874) loss 4.5487 (4.1745) grad_norm 1.0735 (1.1042) [2021-04-15 16:41:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][620/1251] eta 0:03:01 lr 0.000975 time 0.2852 (0.2872) loss 3.7812 (4.1714) grad_norm 1.1309 (1.1046) [2021-04-15 16:41:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][630/1251] eta 0:02:58 lr 0.000975 time 0.2595 (0.2870) loss 4.0186 (4.1661) grad_norm 1.0364 (1.1061) [2021-04-15 16:41:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][640/1251] eta 0:02:55 lr 0.000975 time 0.2735 (0.2869) loss 4.0748 (4.1698) grad_norm 0.9753 (1.1055) [2021-04-15 16:41:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][650/1251] eta 0:02:52 lr 0.000975 time 0.2659 (0.2867) loss 5.0684 (4.1734) grad_norm 1.0102 (1.1063) [2021-04-15 16:41:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][660/1251] eta 0:02:49 lr 0.000975 time 0.2843 (0.2866) loss 4.7770 (4.1780) grad_norm 0.9685 (1.1075) [2021-04-15 16:41:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][670/1251] eta 0:02:46 lr 0.000975 time 0.2795 (0.2866) loss 3.9814 (4.1769) grad_norm 1.2637 (1.1088) [2021-04-15 16:41:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][680/1251] eta 0:02:43 lr 0.000975 time 0.2661 (0.2864) loss 3.1637 (4.1776) grad_norm 1.0097 (1.1082) [2021-04-15 16:41:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][690/1251] eta 0:02:40 lr 0.000975 time 0.2612 (0.2862) loss 3.8059 (4.1794) grad_norm 1.0038 (1.1074) [2021-04-15 16:41:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][700/1251] eta 0:02:37 lr 0.000975 time 0.3087 (0.2861) loss 5.1172 (4.1809) grad_norm 1.2799 (1.1067) [2021-04-15 16:41:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][710/1251] eta 0:02:34 lr 0.000975 time 0.3179 (0.2860) loss 4.3687 (4.1827) grad_norm 1.1819 (1.1072) [2021-04-15 16:41:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][720/1251] eta 0:02:31 lr 0.000975 time 0.3172 (0.2861) loss 4.8851 (4.1874) grad_norm 1.1811 (1.1081) [2021-04-15 16:41:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][730/1251] eta 0:02:29 lr 0.000975 time 0.2761 (0.2861) loss 4.4750 (4.1892) grad_norm 1.3163 (1.1087) [2021-04-15 16:41:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][740/1251] eta 0:02:26 lr 0.000975 time 0.2658 (0.2858) loss 2.8969 (4.1852) grad_norm 1.0427 (1.1080) [2021-04-15 16:41:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][750/1251] eta 0:02:23 lr 0.000975 time 0.2593 (0.2857) loss 4.3041 (4.1856) grad_norm 0.9850 (1.1084) [2021-04-15 16:41:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][760/1251] eta 0:02:20 lr 0.000975 time 0.2474 (0.2856) loss 4.7083 (4.1885) grad_norm 1.0258 (1.1087) [2021-04-15 16:41:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][770/1251] eta 0:02:17 lr 0.000975 time 0.2879 (0.2854) loss 4.6041 (4.1910) grad_norm 1.1295 (1.1085) [2021-04-15 16:41:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][780/1251] eta 0:02:14 lr 0.000975 time 0.2826 (0.2853) loss 5.0832 (4.1928) grad_norm 1.2137 (1.1083) [2021-04-15 16:41:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][790/1251] eta 0:02:11 lr 0.000975 time 0.3019 (0.2852) loss 3.6486 (4.1921) grad_norm 1.2588 (1.1088) [2021-04-15 16:41:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][800/1251] eta 0:02:08 lr 0.000975 time 0.2643 (0.2850) loss 4.1792 (4.1939) grad_norm 1.0503 (1.1084) [2021-04-15 16:41:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][810/1251] eta 0:02:05 lr 0.000975 time 0.2531 (0.2851) loss 3.0633 (4.1972) grad_norm 0.9172 (1.1080) [2021-04-15 16:42:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][820/1251] eta 0:02:02 lr 0.000975 time 0.2737 (0.2850) loss 3.7633 (4.1956) grad_norm 1.3885 (1.1080) [2021-04-15 16:42:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][830/1251] eta 0:01:59 lr 0.000975 time 0.2759 (0.2849) loss 4.5703 (4.1991) grad_norm 1.0536 (1.1084) [2021-04-15 16:42:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][840/1251] eta 0:01:57 lr 0.000975 time 0.2880 (0.2848) loss 4.5558 (4.2007) grad_norm 0.9057 (1.1082) [2021-04-15 16:42:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][850/1251] eta 0:01:54 lr 0.000975 time 0.2727 (0.2847) loss 3.2509 (4.1996) grad_norm 1.2037 (1.1086) [2021-04-15 16:42:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][860/1251] eta 0:01:51 lr 0.000975 time 0.3031 (0.2846) loss 4.8877 (4.2025) grad_norm 1.4881 (1.1092) [2021-04-15 16:42:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][870/1251] eta 0:01:48 lr 0.000975 time 0.2752 (0.2845) loss 3.0615 (4.2006) grad_norm 1.0727 (1.1086) [2021-04-15 16:42:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][880/1251] eta 0:01:45 lr 0.000975 time 0.2664 (0.2843) loss 4.0059 (4.2001) grad_norm 1.0774 (1.1081) [2021-04-15 16:42:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][890/1251] eta 0:01:42 lr 0.000975 time 0.2949 (0.2843) loss 3.4577 (4.1975) grad_norm 0.9644 (1.1082) [2021-04-15 16:42:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][900/1251] eta 0:01:39 lr 0.000975 time 0.2743 (0.2842) loss 3.4934 (4.1960) grad_norm 0.9415 (1.1073) [2021-04-15 16:42:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][910/1251] eta 0:01:36 lr 0.000975 time 0.2435 (0.2841) loss 4.7246 (4.1968) grad_norm 1.1560 (1.1074) [2021-04-15 16:42:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][920/1251] eta 0:01:34 lr 0.000975 time 0.3142 (0.2841) loss 4.6951 (4.1966) grad_norm 1.1100 (1.1071) [2021-04-15 16:42:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][930/1251] eta 0:01:31 lr 0.000975 time 0.2766 (0.2840) loss 4.6339 (4.1958) grad_norm 0.9545 (1.1074) [2021-04-15 16:42:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][940/1251] eta 0:01:28 lr 0.000975 time 0.3984 (0.2841) loss 4.1334 (4.1969) grad_norm 1.0694 (1.1077) [2021-04-15 16:42:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][950/1251] eta 0:01:25 lr 0.000975 time 0.2598 (0.2840) loss 3.4669 (4.1937) grad_norm 1.0725 (1.1083) [2021-04-15 16:42:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][960/1251] eta 0:01:22 lr 0.000975 time 0.2735 (0.2840) loss 4.4260 (4.1971) grad_norm 0.9646 (1.1084) [2021-04-15 16:42:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][970/1251] eta 0:01:19 lr 0.000975 time 0.2912 (0.2839) loss 5.0655 (4.1981) grad_norm 0.9527 (1.1077) [2021-04-15 16:42:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][980/1251] eta 0:01:16 lr 0.000975 time 0.2776 (0.2838) loss 4.0048 (4.1969) grad_norm 0.8708 (1.1068) [2021-04-15 16:42:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][990/1251] eta 0:01:14 lr 0.000974 time 0.2775 (0.2837) loss 4.7227 (4.1969) grad_norm 1.1084 (1.1069) [2021-04-15 16:42:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][1000/1251] eta 0:01:11 lr 0.000974 time 0.2847 (0.2836) loss 5.1371 (4.1981) grad_norm 0.9513 (1.1067) [2021-04-15 16:42:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][1010/1251] eta 0:01:08 lr 0.000974 time 0.2975 (0.2836) loss 4.7510 (4.1963) grad_norm 1.2203 (1.1060) [2021-04-15 16:42:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][1020/1251] eta 0:01:05 lr 0.000974 time 0.2902 (0.2835) loss 3.8364 (4.1946) grad_norm 1.4757 (1.1076) [2021-04-15 16:43:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][1030/1251] eta 0:01:02 lr 0.000974 time 0.2855 (0.2835) loss 4.4230 (4.1940) grad_norm 1.4380 (1.1085) [2021-04-15 16:43:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][1040/1251] eta 0:00:59 lr 0.000974 time 0.2941 (0.2834) loss 3.4066 (4.1934) grad_norm 0.9919 (1.1083) [2021-04-15 16:43:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][1050/1251] eta 0:00:56 lr 0.000974 time 0.2579 (0.2833) loss 3.2317 (4.1904) grad_norm 1.1508 (1.1089) [2021-04-15 16:43:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][1060/1251] eta 0:00:54 lr 0.000974 time 0.2566 (0.2833) loss 2.9928 (4.1886) grad_norm 1.1011 (1.1098) [2021-04-15 16:43:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][1070/1251] eta 0:00:51 lr 0.000974 time 0.3164 (0.2832) loss 3.7650 (4.1894) grad_norm 0.9952 (1.1092) [2021-04-15 16:43:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][1080/1251] eta 0:00:48 lr 0.000974 time 0.2492 (0.2832) loss 2.9317 (4.1904) grad_norm 1.4436 (1.1092) [2021-04-15 16:43:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][1090/1251] eta 0:00:45 lr 0.000974 time 0.2509 (0.2832) loss 4.0400 (4.1879) grad_norm 1.0341 (1.1092) [2021-04-15 16:43:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][1100/1251] eta 0:00:42 lr 0.000974 time 0.2743 (0.2831) loss 4.5914 (4.1895) grad_norm 1.0640 (1.1094) [2021-04-15 16:43:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][1110/1251] eta 0:00:39 lr 0.000974 time 0.2743 (0.2831) loss 3.3453 (4.1891) grad_norm 0.8914 (1.1092) [2021-04-15 16:43:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][1120/1251] eta 0:00:37 lr 0.000974 time 0.2691 (0.2830) loss 4.3884 (4.1898) grad_norm 1.2821 (1.1095) [2021-04-15 16:43:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][1130/1251] eta 0:00:34 lr 0.000974 time 0.2680 (0.2831) loss 4.4356 (4.1910) grad_norm 0.8580 (1.1087) [2021-04-15 16:43:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][1140/1251] eta 0:00:31 lr 0.000974 time 0.2717 (0.2832) loss 3.2890 (4.1907) grad_norm 1.2014 (1.1081) [2021-04-15 16:43:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][1150/1251] eta 0:00:28 lr 0.000974 time 0.2610 (0.2832) loss 5.1821 (4.1905) grad_norm 1.2389 (1.1084) [2021-04-15 16:43:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][1160/1251] eta 0:00:25 lr 0.000974 time 0.2942 (0.2832) loss 4.6846 (4.1918) grad_norm 1.0631 (1.1086) [2021-04-15 16:43:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][1170/1251] eta 0:00:22 lr 0.000974 time 0.2618 (0.2831) loss 5.0436 (4.1936) grad_norm 1.1101 (1.1092) [2021-04-15 16:43:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][1180/1251] eta 0:00:20 lr 0.000974 time 0.2436 (0.2831) loss 4.3513 (4.1917) grad_norm 1.2394 (1.1099) [2021-04-15 16:43:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][1190/1251] eta 0:00:17 lr 0.000974 time 0.2687 (0.2831) loss 4.6414 (4.1933) grad_norm 1.2909 (1.1101) [2021-04-15 16:43:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][1200/1251] eta 0:00:14 lr 0.000974 time 0.2594 (0.2830) loss 5.3137 (4.1941) grad_norm 1.2438 (1.1102) [2021-04-15 16:43:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][1210/1251] eta 0:00:11 lr 0.000974 time 0.2894 (0.2829) loss 4.9332 (4.1937) grad_norm 1.1846 (1.1093) [2021-04-15 16:43:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][1220/1251] eta 0:00:08 lr 0.000974 time 0.2928 (0.2829) loss 4.3271 (4.1924) grad_norm 0.9129 (1.1092) [2021-04-15 16:43:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][1230/1251] eta 0:00:05 lr 0.000974 time 0.2912 (0.2829) loss 3.9566 (4.1925) grad_norm 0.8898 (1.1086) [2021-04-15 16:43:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][1240/1251] eta 0:00:03 lr 0.000974 time 0.2490 (0.2828) loss 3.8895 (4.1933) grad_norm 1.0513 (1.1080) [2021-04-15 16:44:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [30/300][1250/1251] eta 0:00:00 lr 0.000974 time 0.2487 (0.2825) loss 4.5565 (4.1920) grad_norm 1.2009 (1.1088) [2021-04-15 16:44:04 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 30 training takes 0:05:55 [2021-04-15 16:44:04 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_30.pth saving...... [2021-04-15 16:44:16 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_30.pth saved !!! [2021-04-15 16:44:17 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.146 (1.146) Loss 1.5378 (1.5378) Acc@1 63.477 (63.477) Acc@5 87.793 (87.793) [2021-04-15 16:44:18 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.155 (0.217) Loss 1.4663 (1.4960) Acc@1 67.188 (65.980) Acc@5 87.793 (87.518) [2021-04-15 16:44:21 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.140 (0.245) Loss 1.5685 (1.5052) Acc@1 63.184 (65.648) Acc@5 87.695 (87.505) [2021-04-15 16:44:23 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.230 (0.236) Loss 1.4595 (1.5032) Acc@1 65.527 (65.893) Acc@5 87.891 (87.412) [2021-04-15 16:44:25 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.215 (0.228) Loss 1.4273 (1.4992) Acc@1 69.434 (66.035) Acc@5 87.598 (87.436) [2021-04-15 16:44:29 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 66.036 Acc@5 87.492 [2021-04-15 16:44:29 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 66.0% [2021-04-15 16:44:29 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 66.04% [2021-04-15 16:44:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][0/1251] eta 1:03:21 lr 0.000974 time 3.0386 (3.0386) loss 4.6250 (4.6250) grad_norm 0.9670 (0.9670) [2021-04-15 16:44:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][10/1251] eta 0:10:51 lr 0.000974 time 0.2880 (0.5254) loss 4.3764 (4.1580) grad_norm 1.1019 (1.1320) [2021-04-15 16:44:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][20/1251] eta 0:08:26 lr 0.000974 time 0.3021 (0.4118) loss 3.4731 (4.3173) grad_norm 1.1160 (1.1539) [2021-04-15 16:44:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][30/1251] eta 0:07:30 lr 0.000974 time 0.2499 (0.3689) loss 4.2575 (4.3118) grad_norm 0.9160 (1.1056) [2021-04-15 16:44:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][40/1251] eta 0:07:02 lr 0.000974 time 0.2864 (0.3488) loss 3.8543 (4.2658) grad_norm 1.4027 (1.1262) [2021-04-15 16:44:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][50/1251] eta 0:06:43 lr 0.000974 time 0.3045 (0.3363) loss 3.3942 (4.2520) grad_norm 1.0563 (1.1271) [2021-04-15 16:44:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][60/1251] eta 0:06:29 lr 0.000974 time 0.2755 (0.3273) loss 4.6886 (4.2593) grad_norm 1.1582 (1.1316) [2021-04-15 16:44:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][70/1251] eta 0:06:18 lr 0.000974 time 0.2835 (0.3206) loss 3.4492 (4.2558) grad_norm 1.2785 (1.1227) [2021-04-15 16:44:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][80/1251] eta 0:06:09 lr 0.000974 time 0.3040 (0.3159) loss 4.3394 (4.2591) grad_norm 0.9717 (1.1205) [2021-04-15 16:44:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][90/1251] eta 0:06:02 lr 0.000974 time 0.2661 (0.3126) loss 3.5228 (4.2486) grad_norm 1.5272 (1.1216) [2021-04-15 16:45:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][100/1251] eta 0:05:56 lr 0.000974 time 0.2702 (0.3095) loss 4.2439 (4.2253) grad_norm 0.9217 (1.1130) [2021-04-15 16:45:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][110/1251] eta 0:05:49 lr 0.000974 time 0.2645 (0.3066) loss 4.5327 (4.2105) grad_norm 1.1900 (1.1119) [2021-04-15 16:45:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][120/1251] eta 0:05:44 lr 0.000974 time 0.2683 (0.3048) loss 4.3218 (4.1946) grad_norm 0.9974 (1.1084) [2021-04-15 16:45:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][130/1251] eta 0:05:40 lr 0.000974 time 0.2587 (0.3042) loss 4.2140 (4.1937) grad_norm 0.9056 (1.1080) [2021-04-15 16:45:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][140/1251] eta 0:05:35 lr 0.000974 time 0.2847 (0.3023) loss 4.4248 (4.2014) grad_norm 1.0977 (1.1142) [2021-04-15 16:45:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][150/1251] eta 0:05:31 lr 0.000974 time 0.2750 (0.3007) loss 4.4932 (4.1925) grad_norm 1.1615 (1.1113) [2021-04-15 16:45:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][160/1251] eta 0:05:26 lr 0.000974 time 0.2830 (0.2994) loss 3.5486 (4.1877) grad_norm 1.1719 (1.1138) [2021-04-15 16:45:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][170/1251] eta 0:05:22 lr 0.000974 time 0.3159 (0.2985) loss 4.8807 (4.1649) grad_norm 1.0007 (1.1138) [2021-04-15 16:45:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][180/1251] eta 0:05:18 lr 0.000974 time 0.3070 (0.2975) loss 3.7830 (4.1692) grad_norm 1.0638 (1.1117) [2021-04-15 16:45:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][190/1251] eta 0:05:14 lr 0.000974 time 0.3061 (0.2965) loss 5.0364 (4.1643) grad_norm 1.2181 (1.1107) [2021-04-15 16:45:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][200/1251] eta 0:05:10 lr 0.000974 time 0.3023 (0.2956) loss 4.4621 (4.1593) grad_norm 1.1143 (1.1102) [2021-04-15 16:45:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][210/1251] eta 0:05:06 lr 0.000974 time 0.2661 (0.2948) loss 4.5858 (4.1680) grad_norm 1.1830 (1.1120) [2021-04-15 16:45:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][220/1251] eta 0:05:03 lr 0.000974 time 0.2672 (0.2941) loss 4.8616 (4.1654) grad_norm 1.0884 (1.1148) [2021-04-15 16:45:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][230/1251] eta 0:04:59 lr 0.000974 time 0.2899 (0.2934) loss 3.9244 (4.1668) grad_norm 1.1671 (1.1171) [2021-04-15 16:45:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][240/1251] eta 0:04:56 lr 0.000974 time 0.2529 (0.2929) loss 4.6101 (4.1555) grad_norm 1.2538 (1.1182) [2021-04-15 16:45:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][250/1251] eta 0:04:52 lr 0.000974 time 0.2693 (0.2924) loss 4.4903 (4.1539) grad_norm 1.0073 (1.1207) [2021-04-15 16:45:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][260/1251] eta 0:04:49 lr 0.000974 time 0.2650 (0.2919) loss 4.5488 (4.1544) grad_norm 0.8750 (1.1204) [2021-04-15 16:45:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][270/1251] eta 0:04:45 lr 0.000974 time 0.2792 (0.2914) loss 4.3086 (4.1645) grad_norm 1.1736 (1.1207) [2021-04-15 16:45:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][280/1251] eta 0:04:42 lr 0.000974 time 0.2715 (0.2909) loss 4.6178 (4.1788) grad_norm 1.0765 (1.1217) [2021-04-15 16:45:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][290/1251] eta 0:04:39 lr 0.000974 time 0.2836 (0.2907) loss 3.4865 (4.1662) grad_norm 1.1490 (1.1211) [2021-04-15 16:45:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][300/1251] eta 0:04:36 lr 0.000974 time 0.2675 (0.2903) loss 4.1320 (4.1713) grad_norm 1.0676 (1.1209) [2021-04-15 16:45:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][310/1251] eta 0:04:32 lr 0.000974 time 0.2999 (0.2901) loss 4.1799 (4.1666) grad_norm 1.0373 (1.1227) [2021-04-15 16:46:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][320/1251] eta 0:04:29 lr 0.000974 time 0.2856 (0.2897) loss 4.8191 (4.1786) grad_norm 1.0381 (1.1228) [2021-04-15 16:46:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][330/1251] eta 0:04:26 lr 0.000974 time 0.2657 (0.2898) loss 2.9333 (4.1719) grad_norm 1.0742 (1.1246) [2021-04-15 16:46:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][340/1251] eta 0:04:24 lr 0.000974 time 0.2600 (0.2899) loss 3.5115 (4.1692) grad_norm 1.1115 (1.1220) [2021-04-15 16:46:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][350/1251] eta 0:04:20 lr 0.000974 time 0.2679 (0.2895) loss 4.3419 (4.1672) grad_norm 1.2011 (1.1214) [2021-04-15 16:46:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][360/1251] eta 0:04:17 lr 0.000974 time 0.2697 (0.2892) loss 3.4994 (4.1709) grad_norm 1.2078 (1.1235) [2021-04-15 16:46:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][370/1251] eta 0:04:14 lr 0.000974 time 0.2673 (0.2892) loss 4.2441 (4.1719) grad_norm 1.2830 (1.1240) [2021-04-15 16:46:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][380/1251] eta 0:04:11 lr 0.000974 time 0.2730 (0.2889) loss 4.7508 (4.1753) grad_norm 1.1502 (1.1241) [2021-04-15 16:46:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][390/1251] eta 0:04:08 lr 0.000974 time 0.2823 (0.2887) loss 4.8226 (4.1730) grad_norm 1.3429 (1.1226) [2021-04-15 16:46:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][400/1251] eta 0:04:05 lr 0.000974 time 0.2982 (0.2885) loss 4.4328 (4.1702) grad_norm 1.0154 (1.1229) [2021-04-15 16:46:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][410/1251] eta 0:04:02 lr 0.000974 time 0.2620 (0.2881) loss 3.9304 (4.1758) grad_norm 0.9865 (1.1212) [2021-04-15 16:46:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][420/1251] eta 0:03:59 lr 0.000974 time 0.2951 (0.2880) loss 3.8898 (4.1775) grad_norm 0.9624 (1.1184) [2021-04-15 16:46:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][430/1251] eta 0:03:56 lr 0.000974 time 0.2838 (0.2878) loss 4.2671 (4.1724) grad_norm 0.9472 (1.1170) [2021-04-15 16:46:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][440/1251] eta 0:03:53 lr 0.000974 time 0.2950 (0.2877) loss 5.0554 (4.1738) grad_norm 1.3487 (1.1150) [2021-04-15 16:46:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][450/1251] eta 0:03:50 lr 0.000974 time 0.2735 (0.2873) loss 4.1967 (4.1734) grad_norm 1.1274 (1.1146) [2021-04-15 16:46:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][460/1251] eta 0:03:47 lr 0.000974 time 0.3297 (0.2872) loss 5.2754 (4.1768) grad_norm 1.3773 (1.1151) [2021-04-15 16:46:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][470/1251] eta 0:03:44 lr 0.000974 time 0.2714 (0.2869) loss 4.3318 (4.1800) grad_norm 1.2733 (1.1155) [2021-04-15 16:46:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][480/1251] eta 0:03:41 lr 0.000974 time 0.2780 (0.2867) loss 4.6750 (4.1785) grad_norm 1.1530 (1.1151) [2021-04-15 16:46:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][490/1251] eta 0:03:38 lr 0.000973 time 0.2660 (0.2865) loss 4.5260 (4.1783) grad_norm 0.9665 (1.1125) [2021-04-15 16:46:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][500/1251] eta 0:03:35 lr 0.000973 time 0.2639 (0.2863) loss 4.5461 (4.1807) grad_norm 1.2536 (1.1115) [2021-04-15 16:46:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][510/1251] eta 0:03:31 lr 0.000973 time 0.2513 (0.2861) loss 5.0278 (4.1834) grad_norm 0.9877 (1.1107) [2021-04-15 16:46:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][520/1251] eta 0:03:28 lr 0.000973 time 0.2734 (0.2859) loss 4.6010 (4.1830) grad_norm 1.3152 (1.1127) [2021-04-15 16:47:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][530/1251] eta 0:03:25 lr 0.000973 time 0.2755 (0.2857) loss 4.6356 (4.1863) grad_norm 1.0974 (1.1148) [2021-04-15 16:47:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][540/1251] eta 0:03:23 lr 0.000973 time 0.3086 (0.2856) loss 3.4615 (4.1815) grad_norm 0.9861 (1.1155) [2021-04-15 16:47:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][550/1251] eta 0:03:20 lr 0.000973 time 0.2826 (0.2856) loss 4.2459 (4.1850) grad_norm 1.0231 (1.1149) [2021-04-15 16:47:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][560/1251] eta 0:03:17 lr 0.000973 time 0.2699 (0.2854) loss 5.0379 (4.1864) grad_norm 1.1614 (1.1158) [2021-04-15 16:47:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][570/1251] eta 0:03:14 lr 0.000973 time 0.2745 (0.2853) loss 4.7794 (4.1924) grad_norm 0.9467 (1.1147) [2021-04-15 16:47:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][580/1251] eta 0:03:11 lr 0.000973 time 0.2732 (0.2854) loss 4.4329 (4.1890) grad_norm 1.0713 (1.1137) [2021-04-15 16:47:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][590/1251] eta 0:03:08 lr 0.000973 time 0.3062 (0.2853) loss 3.9985 (4.1897) grad_norm 1.2005 (1.1126) [2021-04-15 16:47:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][600/1251] eta 0:03:05 lr 0.000973 time 0.2724 (0.2853) loss 4.6007 (4.1955) grad_norm 1.1067 (1.1120) [2021-04-15 16:47:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][610/1251] eta 0:03:02 lr 0.000973 time 0.2748 (0.2851) loss 3.6947 (4.1959) grad_norm 0.9123 (1.1124) [2021-04-15 16:47:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][620/1251] eta 0:02:59 lr 0.000973 time 0.2975 (0.2851) loss 3.3173 (4.1937) grad_norm 1.3293 (1.1155) [2021-04-15 16:47:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][630/1251] eta 0:02:56 lr 0.000973 time 0.2854 (0.2850) loss 4.6914 (4.1925) grad_norm 1.2921 (1.1154) [2021-04-15 16:47:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][640/1251] eta 0:02:54 lr 0.000973 time 0.2784 (0.2848) loss 4.5328 (4.1954) grad_norm 0.9310 (1.1137) [2021-04-15 16:47:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][650/1251] eta 0:02:51 lr 0.000973 time 0.2631 (0.2846) loss 3.1281 (4.1920) grad_norm 0.9979 (1.1140) [2021-04-15 16:47:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][660/1251] eta 0:02:48 lr 0.000973 time 0.2667 (0.2846) loss 4.9938 (4.1943) grad_norm 0.9751 (1.1138) [2021-04-15 16:47:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][670/1251] eta 0:02:45 lr 0.000973 time 0.2815 (0.2845) loss 4.1325 (4.1953) grad_norm 1.1205 (1.1128) [2021-04-15 16:47:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][680/1251] eta 0:02:42 lr 0.000973 time 0.2966 (0.2844) loss 4.1596 (4.1963) grad_norm 1.2063 (1.1133) [2021-04-15 16:47:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][690/1251] eta 0:02:39 lr 0.000973 time 0.2668 (0.2842) loss 4.2532 (4.1965) grad_norm 0.9377 (1.1135) [2021-04-15 16:47:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][700/1251] eta 0:02:36 lr 0.000973 time 0.2473 (0.2843) loss 4.8090 (4.1979) grad_norm 1.1760 (1.1132) [2021-04-15 16:47:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][710/1251] eta 0:02:33 lr 0.000973 time 0.2761 (0.2842) loss 4.5035 (4.2001) grad_norm 1.0595 (1.1131) [2021-04-15 16:47:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][720/1251] eta 0:02:30 lr 0.000973 time 0.2956 (0.2841) loss 4.3046 (4.1972) grad_norm 1.3028 (1.1132) [2021-04-15 16:47:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][730/1251] eta 0:02:28 lr 0.000973 time 0.2529 (0.2841) loss 4.1840 (4.2000) grad_norm 0.9616 (1.1136) [2021-04-15 16:48:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][740/1251] eta 0:02:25 lr 0.000973 time 0.3044 (0.2840) loss 4.5359 (4.2001) grad_norm 1.0131 (1.1131) [2021-04-15 16:48:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][750/1251] eta 0:02:22 lr 0.000973 time 0.2737 (0.2839) loss 4.2713 (4.2008) grad_norm 1.2296 (1.1124) [2021-04-15 16:48:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][760/1251] eta 0:02:19 lr 0.000973 time 0.2759 (0.2838) loss 4.2787 (4.2007) grad_norm 1.0004 (1.1120) [2021-04-15 16:48:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][770/1251] eta 0:02:16 lr 0.000973 time 0.2926 (0.2838) loss 4.6661 (4.2050) grad_norm 1.1523 (1.1120) [2021-04-15 16:48:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][780/1251] eta 0:02:13 lr 0.000973 time 0.2965 (0.2836) loss 3.2886 (4.2015) grad_norm 1.0441 (1.1112) [2021-04-15 16:48:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][790/1251] eta 0:02:10 lr 0.000973 time 0.2925 (0.2835) loss 4.0335 (4.2000) grad_norm 0.9726 (1.1117) [2021-04-15 16:48:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][800/1251] eta 0:02:07 lr 0.000973 time 0.2627 (0.2835) loss 4.7616 (4.2024) grad_norm 1.1184 (1.1109) [2021-04-15 16:48:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][810/1251] eta 0:02:04 lr 0.000973 time 0.2578 (0.2834) loss 3.9328 (4.2019) grad_norm 1.2724 (1.1119) [2021-04-15 16:48:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][820/1251] eta 0:02:02 lr 0.000973 time 0.2964 (0.2834) loss 4.9913 (4.2060) grad_norm 1.0568 (1.1138) [2021-04-15 16:48:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][830/1251] eta 0:01:59 lr 0.000973 time 0.2859 (0.2833) loss 3.7622 (4.2082) grad_norm 1.0280 (1.1146) [2021-04-15 16:48:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][840/1251] eta 0:01:56 lr 0.000973 time 0.2855 (0.2833) loss 4.5750 (4.2054) grad_norm 1.1200 (1.1138) [2021-04-15 16:48:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][850/1251] eta 0:01:53 lr 0.000973 time 0.2844 (0.2833) loss 2.9723 (4.2007) grad_norm 1.1655 (1.1145) [2021-04-15 16:48:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][860/1251] eta 0:01:50 lr 0.000973 time 0.2949 (0.2832) loss 4.2986 (4.2003) grad_norm 1.1172 (1.1148) [2021-04-15 16:48:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][870/1251] eta 0:01:47 lr 0.000973 time 0.2537 (0.2831) loss 4.4074 (4.1967) grad_norm 0.9405 (1.1140) [2021-04-15 16:48:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][880/1251] eta 0:01:45 lr 0.000973 time 0.2717 (0.2830) loss 4.3511 (4.1977) grad_norm 1.7348 (1.1153) [2021-04-15 16:48:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][890/1251] eta 0:01:42 lr 0.000973 time 0.2915 (0.2830) loss 4.0272 (4.1981) grad_norm 1.0734 (1.1158) [2021-04-15 16:48:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][900/1251] eta 0:01:39 lr 0.000973 time 0.2687 (0.2829) loss 4.4945 (4.1979) grad_norm 0.9953 (1.1154) [2021-04-15 16:48:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][910/1251] eta 0:01:36 lr 0.000973 time 0.2664 (0.2830) loss 2.9125 (4.1970) grad_norm 1.0896 (1.1145) [2021-04-15 16:48:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][920/1251] eta 0:01:33 lr 0.000973 time 0.2710 (0.2830) loss 4.6980 (4.1963) grad_norm 1.2317 (1.1146) [2021-04-15 16:48:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][930/1251] eta 0:01:30 lr 0.000973 time 0.2892 (0.2831) loss 2.8566 (4.1902) grad_norm 0.9897 (1.1141) [2021-04-15 16:48:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][940/1251] eta 0:01:28 lr 0.000973 time 0.2941 (0.2832) loss 4.2457 (4.1932) grad_norm 1.0240 (1.1140) [2021-04-15 16:48:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][950/1251] eta 0:01:25 lr 0.000973 time 0.2634 (0.2831) loss 4.6992 (4.1912) grad_norm 0.9211 (1.1146) [2021-04-15 16:49:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][960/1251] eta 0:01:22 lr 0.000973 time 0.2694 (0.2830) loss 3.2119 (4.1887) grad_norm 1.1784 (1.1156) [2021-04-15 16:49:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][970/1251] eta 0:01:19 lr 0.000973 time 0.2568 (0.2829) loss 4.3469 (4.1879) grad_norm 0.9206 (1.1159) [2021-04-15 16:49:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][980/1251] eta 0:01:16 lr 0.000973 time 0.2552 (0.2828) loss 4.5655 (4.1893) grad_norm 1.2181 (1.1164) [2021-04-15 16:49:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][990/1251] eta 0:01:13 lr 0.000973 time 0.2812 (0.2828) loss 3.9722 (4.1899) grad_norm 0.9857 (1.1160) [2021-04-15 16:49:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][1000/1251] eta 0:01:10 lr 0.000973 time 0.2777 (0.2827) loss 4.1736 (4.1873) grad_norm 1.1435 (1.1163) [2021-04-15 16:49:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][1010/1251] eta 0:01:08 lr 0.000973 time 0.2582 (0.2827) loss 4.7498 (4.1845) grad_norm 1.0575 (1.1155) [2021-04-15 16:49:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][1020/1251] eta 0:01:05 lr 0.000973 time 0.2879 (0.2826) loss 4.4157 (4.1836) grad_norm 1.1035 (1.1152) [2021-04-15 16:49:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][1030/1251] eta 0:01:02 lr 0.000973 time 0.2780 (0.2826) loss 3.3946 (4.1835) grad_norm 1.0111 (1.1142) [2021-04-15 16:49:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][1040/1251] eta 0:00:59 lr 0.000973 time 0.2674 (0.2825) loss 4.4597 (4.1827) grad_norm 0.8513 (1.1141) [2021-04-15 16:49:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][1050/1251] eta 0:00:56 lr 0.000973 time 0.2908 (0.2825) loss 5.0745 (4.1832) grad_norm 0.9644 (1.1130) [2021-04-15 16:49:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][1060/1251] eta 0:00:53 lr 0.000973 time 0.2643 (0.2824) loss 4.4072 (4.1815) grad_norm 1.1363 (1.1128) [2021-04-15 16:49:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][1070/1251] eta 0:00:51 lr 0.000973 time 0.3011 (0.2824) loss 4.9717 (4.1813) grad_norm 1.2029 (1.1131) [2021-04-15 16:49:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][1080/1251] eta 0:00:48 lr 0.000973 time 0.2707 (0.2823) loss 4.5114 (4.1795) grad_norm 1.0196 (1.1125) [2021-04-15 16:49:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][1090/1251] eta 0:00:45 lr 0.000973 time 0.2819 (0.2822) loss 3.0295 (4.1783) grad_norm 1.0683 (1.1124) [2021-04-15 16:49:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][1100/1251] eta 0:00:42 lr 0.000973 time 0.2731 (0.2822) loss 4.3006 (4.1771) grad_norm 1.0356 (1.1127) [2021-04-15 16:49:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][1110/1251] eta 0:00:39 lr 0.000973 time 0.2730 (0.2821) loss 3.9453 (4.1744) grad_norm 1.0002 (1.1120) [2021-04-15 16:49:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][1120/1251] eta 0:00:36 lr 0.000973 time 0.2773 (0.2821) loss 3.0488 (4.1722) grad_norm 1.1275 (1.1114) [2021-04-15 16:49:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][1130/1251] eta 0:00:34 lr 0.000973 time 0.3054 (0.2821) loss 4.4842 (4.1728) grad_norm 0.8992 (1.1119) [2021-04-15 16:49:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][1140/1251] eta 0:00:31 lr 0.000973 time 0.2517 (0.2820) loss 3.9466 (4.1738) grad_norm 0.9650 (1.1116) [2021-04-15 16:49:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][1150/1251] eta 0:00:28 lr 0.000973 time 0.3125 (0.2820) loss 4.0016 (4.1728) grad_norm 1.4190 (1.1123) [2021-04-15 16:49:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][1160/1251] eta 0:00:25 lr 0.000973 time 0.2807 (0.2821) loss 3.3241 (4.1719) grad_norm 0.9386 (1.1123) [2021-04-15 16:49:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][1170/1251] eta 0:00:22 lr 0.000973 time 0.2973 (0.2821) loss 5.1400 (4.1734) grad_norm 1.4034 (1.1130) [2021-04-15 16:50:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][1180/1251] eta 0:00:20 lr 0.000973 time 0.2544 (0.2820) loss 3.9225 (4.1760) grad_norm 0.9148 (1.1130) [2021-04-15 16:50:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][1190/1251] eta 0:00:17 lr 0.000973 time 0.2969 (0.2820) loss 3.7687 (4.1753) grad_norm 1.1287 (1.1126) [2021-04-15 16:50:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][1200/1251] eta 0:00:14 lr 0.000973 time 0.2686 (0.2819) loss 4.2292 (4.1768) grad_norm 1.1662 (1.1126) [2021-04-15 16:50:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][1210/1251] eta 0:00:11 lr 0.000973 time 0.2884 (0.2819) loss 4.4860 (4.1774) grad_norm 1.1015 (1.1125) [2021-04-15 16:50:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][1220/1251] eta 0:00:08 lr 0.000973 time 0.2676 (0.2818) loss 4.1062 (4.1777) grad_norm 1.1393 (1.1123) [2021-04-15 16:50:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][1230/1251] eta 0:00:05 lr 0.000972 time 0.2801 (0.2818) loss 4.4139 (4.1786) grad_norm 0.9164 (1.1122) [2021-04-15 16:50:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][1240/1251] eta 0:00:03 lr 0.000972 time 0.2481 (0.2817) loss 4.6120 (4.1779) grad_norm 1.0226 (1.1118) [2021-04-15 16:50:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [31/300][1250/1251] eta 0:00:00 lr 0.000972 time 0.2491 (0.2814) loss 4.6290 (4.1783) grad_norm 1.3257 (1.1127) [2021-04-15 16:50:23 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 31 training takes 0:05:54 [2021-04-15 16:50:23 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_31.pth saving...... [2021-04-15 16:50:42 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_31.pth saved !!! [2021-04-15 16:50:43 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.155 (1.155) Loss 1.5750 (1.5750) Acc@1 64.453 (64.453) Acc@5 86.328 (86.328) [2021-04-15 16:50:45 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.139 (0.242) Loss 1.4462 (1.4846) Acc@1 66.992 (66.486) Acc@5 88.477 (87.811) [2021-04-15 16:50:47 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.119 (0.231) Loss 1.4775 (1.4999) Acc@1 65.820 (66.211) Acc@5 88.965 (87.640) [2021-04-15 16:50:50 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.198 (0.250) Loss 1.5655 (1.4970) Acc@1 65.723 (66.315) Acc@5 86.328 (87.721) [2021-04-15 16:50:51 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.228) Loss 1.5463 (1.4980) Acc@1 64.062 (66.335) Acc@5 86.523 (87.614) [2021-04-15 16:50:55 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 66.280 Acc@5 87.638 [2021-04-15 16:50:55 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 66.3% [2021-04-15 16:50:55 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 66.28% [2021-04-15 16:50:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][0/1251] eta 1:08:47 lr 0.000972 time 3.2993 (3.2993) loss 3.7160 (3.7160) grad_norm 0.9952 (0.9952) [2021-04-15 16:51:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][10/1251] eta 0:11:28 lr 0.000972 time 0.2581 (0.5546) loss 5.1694 (4.2664) grad_norm 1.0440 (1.0393) [2021-04-15 16:51:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][20/1251] eta 0:08:44 lr 0.000972 time 0.2843 (0.4262) loss 4.5082 (4.2821) grad_norm 1.1920 (1.1206) [2021-04-15 16:51:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][30/1251] eta 0:07:44 lr 0.000972 time 0.3004 (0.3803) loss 3.4559 (4.1336) grad_norm 1.2125 (1.0992) [2021-04-15 16:51:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][40/1251] eta 0:07:11 lr 0.000972 time 0.2851 (0.3563) loss 3.4844 (4.0375) grad_norm 1.2169 (1.1108) [2021-04-15 16:51:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][50/1251] eta 0:06:51 lr 0.000972 time 0.2752 (0.3430) loss 4.1980 (3.9697) grad_norm 0.9518 (1.1076) [2021-04-15 16:51:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][60/1251] eta 0:06:36 lr 0.000972 time 0.2620 (0.3333) loss 3.4969 (3.9864) grad_norm 0.9765 (1.0978) [2021-04-15 16:51:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][70/1251] eta 0:06:25 lr 0.000972 time 0.3037 (0.3260) loss 3.9999 (4.0212) grad_norm 1.1615 (1.0938) [2021-04-15 16:51:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][80/1251] eta 0:06:15 lr 0.000972 time 0.3005 (0.3205) loss 4.0849 (4.0497) grad_norm 1.1089 (1.0968) [2021-04-15 16:51:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][90/1251] eta 0:06:07 lr 0.000972 time 0.2630 (0.3161) loss 4.5503 (4.0808) grad_norm 0.9182 (1.1002) [2021-04-15 16:51:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][100/1251] eta 0:06:01 lr 0.000972 time 0.2794 (0.3145) loss 4.0851 (4.1056) grad_norm 1.0233 (1.1006) [2021-04-15 16:51:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][110/1251] eta 0:05:55 lr 0.000972 time 0.2812 (0.3114) loss 3.4112 (4.0996) grad_norm 0.8764 (1.0973) [2021-04-15 16:51:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][120/1251] eta 0:05:51 lr 0.000972 time 0.2820 (0.3106) loss 4.9286 (4.0910) grad_norm 0.9550 (1.0996) [2021-04-15 16:51:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][130/1251] eta 0:05:45 lr 0.000972 time 0.2888 (0.3084) loss 4.0343 (4.0945) grad_norm 1.1484 (1.1010) [2021-04-15 16:51:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][140/1251] eta 0:05:41 lr 0.000972 time 0.2906 (0.3075) loss 4.5386 (4.0913) grad_norm 1.0780 (1.1006) [2021-04-15 16:51:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][150/1251] eta 0:05:36 lr 0.000972 time 0.3053 (0.3060) loss 3.9057 (4.0867) grad_norm 1.1229 (1.1033) [2021-04-15 16:51:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][160/1251] eta 0:05:32 lr 0.000972 time 0.2902 (0.3047) loss 4.5867 (4.0996) grad_norm 1.0950 (1.1019) [2021-04-15 16:51:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][170/1251] eta 0:05:27 lr 0.000972 time 0.2817 (0.3032) loss 3.5987 (4.0921) grad_norm 1.2883 (1.1023) [2021-04-15 16:51:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][180/1251] eta 0:05:23 lr 0.000972 time 0.3026 (0.3022) loss 4.7079 (4.0938) grad_norm 1.2807 (1.1051) [2021-04-15 16:51:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][190/1251] eta 0:05:19 lr 0.000972 time 0.2935 (0.3011) loss 4.2428 (4.0858) grad_norm 1.2046 (1.1086) [2021-04-15 16:51:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][200/1251] eta 0:05:15 lr 0.000972 time 0.2855 (0.3000) loss 4.5708 (4.0760) grad_norm 1.1248 (1.1172) [2021-04-15 16:51:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][210/1251] eta 0:05:10 lr 0.000972 time 0.2829 (0.2988) loss 4.6502 (4.0851) grad_norm 1.1764 (1.1174) [2021-04-15 16:52:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][220/1251] eta 0:05:07 lr 0.000972 time 0.2950 (0.2978) loss 3.2696 (4.0846) grad_norm 1.0073 (1.1131) [2021-04-15 16:52:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][230/1251] eta 0:05:03 lr 0.000972 time 0.2761 (0.2970) loss 4.6258 (4.0762) grad_norm 1.0222 (1.1094) [2021-04-15 16:52:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][240/1251] eta 0:04:59 lr 0.000972 time 0.2870 (0.2963) loss 4.2882 (4.0738) grad_norm 1.0216 (1.1071) [2021-04-15 16:52:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][250/1251] eta 0:04:55 lr 0.000972 time 0.2956 (0.2956) loss 3.1504 (4.0639) grad_norm 1.2026 (1.1082) [2021-04-15 16:52:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][260/1251] eta 0:04:52 lr 0.000972 time 0.2695 (0.2947) loss 4.3543 (4.0756) grad_norm 0.9916 (1.1060) [2021-04-15 16:52:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][270/1251] eta 0:04:48 lr 0.000972 time 0.2912 (0.2941) loss 4.8614 (4.0716) grad_norm 1.2357 (1.1069) [2021-04-15 16:52:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][280/1251] eta 0:04:45 lr 0.000972 time 0.2959 (0.2935) loss 4.7272 (4.0663) grad_norm 1.0372 (1.1104) [2021-04-15 16:52:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][290/1251] eta 0:04:41 lr 0.000972 time 0.2773 (0.2929) loss 4.5224 (4.0615) grad_norm 0.9781 (1.1099) [2021-04-15 16:52:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][300/1251] eta 0:04:38 lr 0.000972 time 0.2935 (0.2924) loss 3.9322 (4.0687) grad_norm 1.0773 (1.1091) [2021-04-15 16:52:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][310/1251] eta 0:04:34 lr 0.000972 time 0.2571 (0.2919) loss 4.8094 (4.0769) grad_norm 0.9960 (1.1075) [2021-04-15 16:52:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][320/1251] eta 0:04:31 lr 0.000972 time 0.2779 (0.2917) loss 4.3553 (4.0807) grad_norm 1.5145 (1.1089) [2021-04-15 16:52:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][330/1251] eta 0:04:28 lr 0.000972 time 0.2699 (0.2913) loss 3.2139 (4.0688) grad_norm 1.1535 (1.1086) [2021-04-15 16:52:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][340/1251] eta 0:04:25 lr 0.000972 time 0.2802 (0.2910) loss 5.0190 (4.0726) grad_norm 1.1098 (1.1131) [2021-04-15 16:52:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][350/1251] eta 0:04:22 lr 0.000972 time 0.4222 (0.2910) loss 4.5529 (4.0624) grad_norm 1.2473 (1.1116) [2021-04-15 16:52:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][360/1251] eta 0:04:18 lr 0.000972 time 0.2623 (0.2905) loss 3.1343 (4.0721) grad_norm 0.9473 (1.1125) [2021-04-15 16:52:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][370/1251] eta 0:04:15 lr 0.000972 time 0.2445 (0.2903) loss 4.5531 (4.0787) grad_norm 0.9260 (1.1117) [2021-04-15 16:52:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][380/1251] eta 0:04:12 lr 0.000972 time 0.2701 (0.2900) loss 4.7872 (4.0844) grad_norm 1.2919 (1.1110) [2021-04-15 16:52:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][390/1251] eta 0:04:09 lr 0.000972 time 0.2750 (0.2897) loss 4.0218 (4.0861) grad_norm 1.0411 (1.1106) [2021-04-15 16:52:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][400/1251] eta 0:04:06 lr 0.000972 time 0.2879 (0.2896) loss 4.6708 (4.0826) grad_norm 1.0252 (1.1093) [2021-04-15 16:52:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][410/1251] eta 0:04:03 lr 0.000972 time 0.2685 (0.2894) loss 4.5432 (4.0867) grad_norm 1.0580 (1.1084) [2021-04-15 16:52:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][420/1251] eta 0:04:00 lr 0.000972 time 0.2807 (0.2892) loss 4.9935 (4.0902) grad_norm 1.0002 (1.1076) [2021-04-15 16:52:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][430/1251] eta 0:03:57 lr 0.000972 time 0.2743 (0.2890) loss 4.4106 (4.0944) grad_norm 1.3259 (1.1062) [2021-04-15 16:53:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][440/1251] eta 0:03:54 lr 0.000972 time 0.3092 (0.2888) loss 3.7036 (4.0935) grad_norm 1.0333 (1.1066) [2021-04-15 16:53:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][450/1251] eta 0:03:51 lr 0.000972 time 0.3143 (0.2886) loss 4.1356 (4.0978) grad_norm 1.0150 (1.1077) [2021-04-15 16:53:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][460/1251] eta 0:03:48 lr 0.000972 time 0.2927 (0.2885) loss 4.3632 (4.1012) grad_norm 1.0504 (1.1080) [2021-04-15 16:53:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][470/1251] eta 0:03:45 lr 0.000972 time 0.2747 (0.2882) loss 3.1700 (4.1014) grad_norm 1.1104 (1.1058) [2021-04-15 16:53:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][480/1251] eta 0:03:42 lr 0.000972 time 0.2967 (0.2880) loss 4.4770 (4.1046) grad_norm 1.0406 (1.1053) [2021-04-15 16:53:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][490/1251] eta 0:03:39 lr 0.000972 time 0.2771 (0.2878) loss 4.7042 (4.1058) grad_norm 0.9580 (1.1038) [2021-04-15 16:53:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][500/1251] eta 0:03:36 lr 0.000972 time 0.2915 (0.2877) loss 3.5077 (4.1132) grad_norm 0.9751 (1.1061) [2021-04-15 16:53:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][510/1251] eta 0:03:33 lr 0.000972 time 0.2815 (0.2876) loss 3.5683 (4.1086) grad_norm 0.8457 (1.1072) [2021-04-15 16:53:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][520/1251] eta 0:03:30 lr 0.000972 time 0.2782 (0.2874) loss 4.3235 (4.1135) grad_norm 0.9994 (1.1061) [2021-04-15 16:53:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][530/1251] eta 0:03:27 lr 0.000972 time 0.2967 (0.2873) loss 3.5508 (4.1048) grad_norm 1.0548 (1.1060) [2021-04-15 16:53:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][540/1251] eta 0:03:24 lr 0.000972 time 0.3049 (0.2873) loss 3.9563 (4.1076) grad_norm 1.1210 (1.1068) [2021-04-15 16:53:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][550/1251] eta 0:03:21 lr 0.000972 time 0.2817 (0.2872) loss 3.9305 (4.1138) grad_norm 1.2624 (1.1073) [2021-04-15 16:53:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][560/1251] eta 0:03:18 lr 0.000972 time 0.2843 (0.2871) loss 4.6795 (4.1196) grad_norm 0.9339 (1.1069) [2021-04-15 16:53:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][570/1251] eta 0:03:15 lr 0.000972 time 0.2866 (0.2872) loss 3.8501 (4.1177) grad_norm 1.4172 (1.1063) [2021-04-15 16:53:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][580/1251] eta 0:03:12 lr 0.000972 time 0.2526 (0.2870) loss 4.6400 (4.1183) grad_norm 1.1449 (1.1087) [2021-04-15 16:53:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][590/1251] eta 0:03:09 lr 0.000972 time 0.2797 (0.2868) loss 4.8880 (4.1129) grad_norm 0.9775 (1.1078) [2021-04-15 16:53:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][600/1251] eta 0:03:06 lr 0.000972 time 0.2765 (0.2867) loss 4.9584 (4.1148) grad_norm 0.9694 (1.1069) [2021-04-15 16:53:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][610/1251] eta 0:03:03 lr 0.000972 time 0.2819 (0.2865) loss 3.9358 (4.1184) grad_norm 0.9428 (1.1067) [2021-04-15 16:53:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][620/1251] eta 0:03:00 lr 0.000972 time 0.2859 (0.2863) loss 4.5140 (4.1161) grad_norm 1.4415 (1.1069) [2021-04-15 16:53:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][630/1251] eta 0:02:57 lr 0.000972 time 0.2577 (0.2861) loss 4.4836 (4.1226) grad_norm 1.0779 (1.1067) [2021-04-15 16:53:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][640/1251] eta 0:02:54 lr 0.000972 time 0.2787 (0.2860) loss 4.0540 (4.1222) grad_norm 1.0152 (1.1067) [2021-04-15 16:54:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][650/1251] eta 0:02:51 lr 0.000972 time 0.2730 (0.2858) loss 4.0723 (4.1227) grad_norm 1.2410 (1.1079) [2021-04-15 16:54:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][660/1251] eta 0:02:48 lr 0.000972 time 0.3045 (0.2857) loss 4.3689 (4.1254) grad_norm 1.0249 (1.1066) [2021-04-15 16:54:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][670/1251] eta 0:02:45 lr 0.000972 time 0.2875 (0.2856) loss 3.9037 (4.1253) grad_norm 0.9810 (1.1056) [2021-04-15 16:54:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][680/1251] eta 0:02:43 lr 0.000972 time 0.2596 (0.2856) loss 4.4740 (4.1260) grad_norm 1.2388 (1.1059) [2021-04-15 16:54:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][690/1251] eta 0:02:40 lr 0.000972 time 0.2791 (0.2858) loss 5.2089 (4.1262) grad_norm 1.2324 (1.1069) [2021-04-15 16:54:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][700/1251] eta 0:02:37 lr 0.000972 time 0.2778 (0.2856) loss 4.5968 (4.1245) grad_norm 0.9172 (1.1063) [2021-04-15 16:54:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][710/1251] eta 0:02:34 lr 0.000971 time 0.2780 (0.2855) loss 4.0428 (4.1268) grad_norm 1.2077 (1.1068) [2021-04-15 16:54:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][720/1251] eta 0:02:31 lr 0.000971 time 0.2684 (0.2855) loss 4.3631 (4.1290) grad_norm 0.9861 (1.1062) [2021-04-15 16:54:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][730/1251] eta 0:02:28 lr 0.000971 time 0.2573 (0.2855) loss 3.1675 (4.1288) grad_norm 1.4355 (1.1070) [2021-04-15 16:54:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][740/1251] eta 0:02:25 lr 0.000971 time 0.2665 (0.2853) loss 4.7644 (4.1350) grad_norm 1.0553 (1.1060) [2021-04-15 16:54:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][750/1251] eta 0:02:22 lr 0.000971 time 0.2880 (0.2853) loss 2.7975 (4.1323) grad_norm 1.1181 (1.1057) [2021-04-15 16:54:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][760/1251] eta 0:02:20 lr 0.000971 time 0.2879 (0.2851) loss 4.1906 (4.1299) grad_norm 1.1820 (1.1071) [2021-04-15 16:54:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][770/1251] eta 0:02:17 lr 0.000971 time 0.2673 (0.2850) loss 3.9151 (4.1258) grad_norm 1.2141 (1.1065) [2021-04-15 16:54:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][780/1251] eta 0:02:14 lr 0.000971 time 0.2650 (0.2849) loss 4.2979 (4.1272) grad_norm 0.9859 (1.1059) [2021-04-15 16:54:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][790/1251] eta 0:02:11 lr 0.000971 time 0.2454 (0.2848) loss 4.2307 (4.1271) grad_norm 1.1443 (1.1052) [2021-04-15 16:54:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][800/1251] eta 0:02:08 lr 0.000971 time 0.2529 (0.2847) loss 4.4813 (4.1284) grad_norm 1.0134 (1.1045) [2021-04-15 16:54:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][810/1251] eta 0:02:05 lr 0.000971 time 0.2817 (0.2848) loss 3.4982 (4.1267) grad_norm 1.0991 (1.1041) [2021-04-15 16:54:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][820/1251] eta 0:02:02 lr 0.000971 time 0.2936 (0.2847) loss 3.0749 (4.1265) grad_norm 0.9923 (1.1038) [2021-04-15 16:54:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][830/1251] eta 0:01:59 lr 0.000971 time 0.2711 (0.2846) loss 3.8616 (4.1262) grad_norm 1.1727 (1.1050) [2021-04-15 16:54:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][840/1251] eta 0:01:56 lr 0.000971 time 0.2812 (0.2846) loss 3.9228 (4.1279) grad_norm 1.1313 (1.1050) [2021-04-15 16:54:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][850/1251] eta 0:01:54 lr 0.000971 time 0.2657 (0.2845) loss 4.3066 (4.1236) grad_norm 1.0954 (1.1053) [2021-04-15 16:55:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][860/1251] eta 0:01:51 lr 0.000971 time 0.2860 (0.2844) loss 4.8565 (4.1236) grad_norm 0.9998 (1.1047) [2021-04-15 16:55:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][870/1251] eta 0:01:48 lr 0.000971 time 0.2807 (0.2843) loss 4.0444 (4.1227) grad_norm 1.0330 (1.1041) [2021-04-15 16:55:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][880/1251] eta 0:01:45 lr 0.000971 time 0.2787 (0.2843) loss 4.8222 (4.1229) grad_norm 1.0583 (1.1037) [2021-04-15 16:55:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][890/1251] eta 0:01:42 lr 0.000971 time 0.2709 (0.2842) loss 4.3342 (4.1238) grad_norm 1.0317 (1.1029) [2021-04-15 16:55:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][900/1251] eta 0:01:39 lr 0.000971 time 0.2815 (0.2841) loss 4.5599 (4.1220) grad_norm 1.0225 (1.1019) [2021-04-15 16:55:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][910/1251] eta 0:01:36 lr 0.000971 time 0.2565 (0.2840) loss 4.7460 (4.1250) grad_norm 1.0817 (1.1008) [2021-04-15 16:55:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][920/1251] eta 0:01:33 lr 0.000971 time 0.2797 (0.2839) loss 4.7384 (4.1311) grad_norm 1.1640 (1.1002) [2021-04-15 16:55:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][930/1251] eta 0:01:31 lr 0.000971 time 0.2726 (0.2840) loss 3.2535 (4.1275) grad_norm 0.9821 (1.0995) [2021-04-15 16:55:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][940/1251] eta 0:01:28 lr 0.000971 time 0.2570 (0.2839) loss 3.2462 (4.1295) grad_norm 1.0416 (1.0992) [2021-04-15 16:55:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][950/1251] eta 0:01:25 lr 0.000971 time 0.2615 (0.2839) loss 4.5023 (4.1285) grad_norm 1.2335 (1.0998) [2021-04-15 16:55:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][960/1251] eta 0:01:22 lr 0.000971 time 0.2549 (0.2838) loss 5.1455 (4.1320) grad_norm 1.0788 (1.1005) [2021-04-15 16:55:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][970/1251] eta 0:01:19 lr 0.000971 time 0.2789 (0.2838) loss 4.7979 (4.1303) grad_norm 1.1239 (1.1009) [2021-04-15 16:55:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][980/1251] eta 0:01:16 lr 0.000971 time 0.2672 (0.2837) loss 5.2259 (4.1308) grad_norm 1.1569 (1.1012) [2021-04-15 16:55:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][990/1251] eta 0:01:14 lr 0.000971 time 0.2737 (0.2836) loss 4.9250 (4.1331) grad_norm 1.5092 (1.1028) [2021-04-15 16:55:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][1000/1251] eta 0:01:11 lr 0.000971 time 0.2728 (0.2836) loss 4.3466 (4.1351) grad_norm 0.9349 (1.1025) [2021-04-15 16:55:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][1010/1251] eta 0:01:08 lr 0.000971 time 0.2889 (0.2835) loss 3.9677 (4.1377) grad_norm 1.0987 (1.1024) [2021-04-15 16:55:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][1020/1251] eta 0:01:05 lr 0.000971 time 0.2617 (0.2834) loss 3.5397 (4.1368) grad_norm 1.5344 (1.1037) [2021-04-15 16:55:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][1030/1251] eta 0:01:02 lr 0.000971 time 0.2876 (0.2834) loss 4.3462 (4.1361) grad_norm 1.0731 (1.1039) [2021-04-15 16:55:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][1040/1251] eta 0:00:59 lr 0.000971 time 0.2956 (0.2834) loss 4.4601 (4.1342) grad_norm 1.0932 (1.1047) [2021-04-15 16:55:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][1050/1251] eta 0:00:56 lr 0.000971 time 0.2960 (0.2833) loss 4.1913 (4.1357) grad_norm 1.1386 (1.1053) [2021-04-15 16:55:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][1060/1251] eta 0:00:54 lr 0.000971 time 0.2600 (0.2832) loss 4.6548 (4.1359) grad_norm 0.9826 (1.1060) [2021-04-15 16:55:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][1070/1251] eta 0:00:51 lr 0.000971 time 0.2964 (0.2832) loss 3.7134 (4.1382) grad_norm 1.1125 (1.1063) [2021-04-15 16:56:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][1080/1251] eta 0:00:48 lr 0.000971 time 0.2603 (0.2832) loss 3.2576 (4.1390) grad_norm 0.9636 (1.1059) [2021-04-15 16:56:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][1090/1251] eta 0:00:45 lr 0.000971 time 0.2761 (0.2831) loss 4.4277 (4.1382) grad_norm 1.0106 (1.1052) [2021-04-15 16:56:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][1100/1251] eta 0:00:42 lr 0.000971 time 0.2703 (0.2831) loss 4.6753 (4.1402) grad_norm 1.0247 (1.1056) [2021-04-15 16:56:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][1110/1251] eta 0:00:39 lr 0.000971 time 0.2616 (0.2831) loss 4.8939 (4.1407) grad_norm 1.1228 (1.1052) [2021-04-15 16:56:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][1120/1251] eta 0:00:37 lr 0.000971 time 0.3009 (0.2830) loss 4.3581 (4.1422) grad_norm 1.0871 (1.1048) [2021-04-15 16:56:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][1130/1251] eta 0:00:34 lr 0.000971 time 0.2906 (0.2830) loss 3.8032 (4.1430) grad_norm 1.0063 (1.1037) [2021-04-15 16:56:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][1140/1251] eta 0:00:31 lr 0.000971 time 0.3083 (0.2829) loss 2.9407 (4.1420) grad_norm 1.1142 (inf) [2021-04-15 16:56:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][1150/1251] eta 0:00:28 lr 0.000971 time 0.2815 (0.2830) loss 4.6888 (4.1437) grad_norm 1.0761 (inf) [2021-04-15 16:56:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][1160/1251] eta 0:00:25 lr 0.000971 time 0.2671 (0.2829) loss 4.5708 (4.1426) grad_norm 1.0725 (inf) [2021-04-15 16:56:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][1170/1251] eta 0:00:22 lr 0.000971 time 0.2821 (0.2829) loss 4.5868 (4.1445) grad_norm 1.0820 (inf) [2021-04-15 16:56:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][1180/1251] eta 0:00:20 lr 0.000971 time 0.2727 (0.2829) loss 4.4261 (4.1473) grad_norm 0.9448 (inf) [2021-04-15 16:56:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][1190/1251] eta 0:00:17 lr 0.000971 time 0.2552 (0.2828) loss 4.4028 (4.1470) grad_norm 1.2151 (inf) [2021-04-15 16:56:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][1200/1251] eta 0:00:14 lr 0.000971 time 0.2686 (0.2827) loss 4.4544 (4.1491) grad_norm 1.0039 (inf) [2021-04-15 16:56:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][1210/1251] eta 0:00:11 lr 0.000971 time 0.2870 (0.2827) loss 4.6685 (4.1478) grad_norm 1.1242 (inf) [2021-04-15 16:56:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][1220/1251] eta 0:00:08 lr 0.000971 time 0.2736 (0.2827) loss 4.3432 (4.1479) grad_norm 0.9732 (inf) [2021-04-15 16:56:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][1230/1251] eta 0:00:05 lr 0.000971 time 0.2704 (0.2827) loss 3.8520 (4.1472) grad_norm 1.0214 (inf) [2021-04-15 16:56:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][1240/1251] eta 0:00:03 lr 0.000971 time 0.2489 (0.2825) loss 4.4976 (4.1490) grad_norm 0.9540 (inf) [2021-04-15 16:56:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [32/300][1250/1251] eta 0:00:00 lr 0.000971 time 0.2481 (0.2823) loss 4.5719 (4.1528) grad_norm 1.1339 (inf) [2021-04-15 16:56:50 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 32 training takes 0:05:55 [2021-04-15 16:56:50 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_32.pth saving...... [2021-04-15 16:57:07 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_32.pth saved !!! [2021-04-15 16:57:08 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.267 (1.267) Loss 1.4875 (1.4875) Acc@1 65.820 (65.820) Acc@5 88.086 (88.086) [2021-04-15 16:57:10 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.114 (0.236) Loss 1.4095 (1.4820) Acc@1 68.652 (66.566) Acc@5 88.672 (88.077) [2021-04-15 16:57:12 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.132 (0.240) Loss 1.4449 (1.4987) Acc@1 67.188 (66.141) Acc@5 87.793 (87.779) [2021-04-15 16:57:15 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.100 (0.245) Loss 1.5027 (1.4892) Acc@1 66.016 (66.485) Acc@5 87.207 (87.856) [2021-04-15 16:57:16 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.156 (0.212) Loss 1.4356 (1.4870) Acc@1 67.871 (66.668) Acc@5 88.379 (87.881) [2021-04-15 16:57:19 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 66.792 Acc@5 88.022 [2021-04-15 16:57:19 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 66.8% [2021-04-15 16:57:19 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 66.79% [2021-04-15 16:57:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][0/1251] eta 1:24:03 lr 0.000971 time 4.0312 (4.0312) loss 4.1054 (4.1054) grad_norm 1.0727 (1.0727) [2021-04-15 16:57:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][10/1251] eta 0:13:02 lr 0.000971 time 0.3036 (0.6304) loss 4.9253 (4.1237) grad_norm 1.3887 (1.1466) [2021-04-15 16:57:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][20/1251] eta 0:09:31 lr 0.000971 time 0.2731 (0.4640) loss 3.7120 (4.0837) grad_norm 1.1750 (1.0900) [2021-04-15 16:57:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][30/1251] eta 0:08:15 lr 0.000971 time 0.2882 (0.4057) loss 4.4562 (3.9973) grad_norm 1.3971 (1.1248) [2021-04-15 16:57:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][40/1251] eta 0:07:33 lr 0.000971 time 0.2765 (0.3746) loss 3.7722 (4.0543) grad_norm 1.3275 (1.1232) [2021-04-15 16:57:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][50/1251] eta 0:07:07 lr 0.000971 time 0.2909 (0.3558) loss 3.6797 (3.9874) grad_norm 0.9669 (1.1089) [2021-04-15 16:57:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][60/1251] eta 0:06:50 lr 0.000971 time 0.2610 (0.3447) loss 4.7227 (4.0604) grad_norm 1.0435 (1.0977) [2021-04-15 16:57:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][70/1251] eta 0:06:35 lr 0.000971 time 0.2869 (0.3350) loss 3.3975 (4.0725) grad_norm 1.0437 (1.0983) [2021-04-15 16:57:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][80/1251] eta 0:06:24 lr 0.000971 time 0.2692 (0.3283) loss 4.6098 (4.0658) grad_norm 1.0063 (1.1076) [2021-04-15 16:57:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][90/1251] eta 0:06:15 lr 0.000971 time 0.2600 (0.3232) loss 4.5742 (4.0852) grad_norm 1.0543 (1.1074) [2021-04-15 16:57:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][100/1251] eta 0:06:08 lr 0.000971 time 0.2960 (0.3198) loss 3.8482 (4.0800) grad_norm 1.1473 (1.0970) [2021-04-15 16:57:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][110/1251] eta 0:06:01 lr 0.000971 time 0.2883 (0.3166) loss 3.8108 (4.1193) grad_norm 1.0554 (1.0928) [2021-04-15 16:57:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][120/1251] eta 0:05:54 lr 0.000971 time 0.2638 (0.3138) loss 4.5765 (4.1271) grad_norm 1.1353 (1.0883) [2021-04-15 16:57:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][130/1251] eta 0:05:50 lr 0.000971 time 0.2922 (0.3127) loss 3.5733 (4.1323) grad_norm 0.9863 (1.0892) [2021-04-15 16:58:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][140/1251] eta 0:05:44 lr 0.000971 time 0.2772 (0.3102) loss 4.6296 (4.1540) grad_norm 1.2292 (1.0951) [2021-04-15 16:58:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][150/1251] eta 0:05:39 lr 0.000971 time 0.2705 (0.3081) loss 3.9726 (4.1465) grad_norm 1.2752 (1.0983) [2021-04-15 16:58:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][160/1251] eta 0:05:34 lr 0.000971 time 0.2811 (0.3063) loss 5.1099 (4.1387) grad_norm 1.0102 (1.0974) [2021-04-15 16:58:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][170/1251] eta 0:05:29 lr 0.000970 time 0.2517 (0.3050) loss 4.1381 (4.1115) grad_norm 1.1472 (1.0975) [2021-04-15 16:58:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][180/1251] eta 0:05:25 lr 0.000970 time 0.2674 (0.3036) loss 3.2625 (4.0880) grad_norm 1.0040 (1.1020) [2021-04-15 16:58:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][190/1251] eta 0:05:20 lr 0.000970 time 0.2764 (0.3024) loss 4.2922 (4.0757) grad_norm 0.9606 (1.1011) [2021-04-15 16:58:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][200/1251] eta 0:05:16 lr 0.000970 time 0.2941 (0.3015) loss 4.9869 (4.0819) grad_norm 1.0917 (1.0966) [2021-04-15 16:58:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][210/1251] eta 0:05:12 lr 0.000970 time 0.2941 (0.3006) loss 3.7592 (4.0764) grad_norm 1.3242 (1.0994) [2021-04-15 16:58:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][220/1251] eta 0:05:08 lr 0.000970 time 0.2867 (0.2996) loss 4.3294 (4.0815) grad_norm 0.9778 (1.1002) [2021-04-15 16:58:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][230/1251] eta 0:05:04 lr 0.000970 time 0.2588 (0.2986) loss 4.6056 (4.0843) grad_norm 1.1276 (1.0990) [2021-04-15 16:58:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][240/1251] eta 0:05:01 lr 0.000970 time 0.2890 (0.2979) loss 2.7481 (4.0919) grad_norm 1.1012 (1.0986) [2021-04-15 16:58:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][250/1251] eta 0:04:57 lr 0.000970 time 0.2676 (0.2973) loss 3.6910 (4.1031) grad_norm 1.0884 (1.1001) [2021-04-15 16:58:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][260/1251] eta 0:04:53 lr 0.000970 time 0.2830 (0.2966) loss 4.4139 (4.1007) grad_norm 1.2340 (1.1000) [2021-04-15 16:58:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][270/1251] eta 0:04:50 lr 0.000970 time 0.2786 (0.2962) loss 3.5694 (4.1069) grad_norm 1.3153 (1.1038) [2021-04-15 16:58:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][280/1251] eta 0:04:47 lr 0.000970 time 0.2780 (0.2956) loss 5.2090 (4.1108) grad_norm 1.0360 (1.1017) [2021-04-15 16:58:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][290/1251] eta 0:04:43 lr 0.000970 time 0.2632 (0.2951) loss 4.6350 (4.1182) grad_norm 1.2158 (1.1043) [2021-04-15 16:58:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][300/1251] eta 0:04:40 lr 0.000970 time 0.2917 (0.2945) loss 4.5611 (4.1205) grad_norm 1.1579 (1.1069) [2021-04-15 16:58:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][310/1251] eta 0:04:36 lr 0.000970 time 0.2705 (0.2942) loss 3.6313 (4.1070) grad_norm 1.0414 (1.1068) [2021-04-15 16:58:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][320/1251] eta 0:04:33 lr 0.000970 time 0.2899 (0.2938) loss 3.3379 (4.1034) grad_norm 1.1417 (1.1050) [2021-04-15 16:58:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][330/1251] eta 0:04:30 lr 0.000970 time 0.2957 (0.2933) loss 4.4864 (4.1055) grad_norm 1.1298 (1.1041) [2021-04-15 16:58:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][340/1251] eta 0:04:26 lr 0.000970 time 0.2824 (0.2929) loss 3.6021 (4.1015) grad_norm 1.3037 (1.1049) [2021-04-15 16:59:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][350/1251] eta 0:04:23 lr 0.000970 time 0.2775 (0.2928) loss 4.9613 (4.1027) grad_norm 1.1976 (1.1061) [2021-04-15 16:59:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][360/1251] eta 0:04:20 lr 0.000970 time 0.2568 (0.2922) loss 3.1042 (4.0869) grad_norm 1.0427 (1.1053) [2021-04-15 16:59:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][370/1251] eta 0:04:17 lr 0.000970 time 0.2770 (0.2918) loss 4.5260 (4.0887) grad_norm 1.0784 (1.1066) [2021-04-15 16:59:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][380/1251] eta 0:04:13 lr 0.000970 time 0.2861 (0.2915) loss 3.4513 (4.0886) grad_norm 1.5511 (1.1102) [2021-04-15 16:59:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][390/1251] eta 0:04:10 lr 0.000970 time 0.2566 (0.2911) loss 4.7157 (4.0919) grad_norm 1.3961 (1.1118) [2021-04-15 16:59:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][400/1251] eta 0:04:07 lr 0.000970 time 0.2776 (0.2908) loss 4.4561 (4.0994) grad_norm 1.3897 (1.1136) [2021-04-15 16:59:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][410/1251] eta 0:04:04 lr 0.000970 time 0.2932 (0.2911) loss 3.6059 (4.0981) grad_norm 1.1180 (1.1131) [2021-04-15 16:59:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][420/1251] eta 0:04:01 lr 0.000970 time 0.3005 (0.2907) loss 4.4866 (4.1024) grad_norm 0.9387 (inf) [2021-04-15 16:59:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][430/1251] eta 0:03:58 lr 0.000970 time 0.2841 (0.2906) loss 3.6183 (4.1011) grad_norm 1.1558 (inf) [2021-04-15 16:59:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][440/1251] eta 0:03:55 lr 0.000970 time 0.2904 (0.2904) loss 4.7353 (4.1042) grad_norm 1.4090 (inf) [2021-04-15 16:59:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][450/1251] eta 0:03:52 lr 0.000970 time 0.2856 (0.2901) loss 4.2575 (4.1039) grad_norm 1.0322 (inf) [2021-04-15 16:59:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][460/1251] eta 0:03:49 lr 0.000970 time 0.3059 (0.2898) loss 4.3056 (4.1021) grad_norm 1.0171 (inf) [2021-04-15 16:59:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][470/1251] eta 0:03:46 lr 0.000970 time 0.3185 (0.2896) loss 4.1888 (4.1030) grad_norm 1.1619 (inf) [2021-04-15 16:59:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][480/1251] eta 0:03:43 lr 0.000970 time 0.2596 (0.2894) loss 3.4595 (4.1040) grad_norm 1.0595 (inf) [2021-04-15 16:59:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][490/1251] eta 0:03:40 lr 0.000970 time 0.3009 (0.2896) loss 4.3551 (4.1105) grad_norm 1.0921 (inf) [2021-04-15 16:59:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][500/1251] eta 0:03:37 lr 0.000970 time 0.2877 (0.2893) loss 4.5328 (4.1137) grad_norm 1.0089 (inf) [2021-04-15 16:59:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][510/1251] eta 0:03:34 lr 0.000970 time 0.2808 (0.2892) loss 4.3003 (4.1120) grad_norm 1.0109 (inf) [2021-04-15 16:59:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][520/1251] eta 0:03:31 lr 0.000970 time 0.2653 (0.2889) loss 3.9278 (4.1119) grad_norm 0.9156 (inf) [2021-04-15 16:59:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][530/1251] eta 0:03:28 lr 0.000970 time 0.2768 (0.2887) loss 3.4866 (4.1062) grad_norm 1.1948 (inf) [2021-04-15 16:59:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][540/1251] eta 0:03:25 lr 0.000970 time 0.2980 (0.2885) loss 4.8044 (4.1064) grad_norm 0.9211 (inf) [2021-04-15 16:59:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][550/1251] eta 0:03:22 lr 0.000970 time 0.2851 (0.2883) loss 4.2836 (4.1038) grad_norm 1.1892 (inf) [2021-04-15 17:00:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][560/1251] eta 0:03:19 lr 0.000970 time 0.2751 (0.2881) loss 4.6124 (4.1029) grad_norm 1.0390 (inf) [2021-04-15 17:00:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][570/1251] eta 0:03:16 lr 0.000970 time 0.2579 (0.2879) loss 3.9242 (4.1044) grad_norm 1.1107 (inf) [2021-04-15 17:00:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][580/1251] eta 0:03:13 lr 0.000970 time 0.2693 (0.2880) loss 4.1704 (4.1022) grad_norm 1.2860 (inf) [2021-04-15 17:00:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][590/1251] eta 0:03:10 lr 0.000970 time 0.2758 (0.2878) loss 4.6209 (4.1070) grad_norm 1.2830 (inf) [2021-04-15 17:00:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][600/1251] eta 0:03:07 lr 0.000970 time 0.2637 (0.2877) loss 4.5935 (4.1029) grad_norm 1.2642 (inf) [2021-04-15 17:00:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][610/1251] eta 0:03:04 lr 0.000970 time 0.2425 (0.2875) loss 3.6050 (4.1023) grad_norm 1.0013 (inf) [2021-04-15 17:00:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][620/1251] eta 0:03:01 lr 0.000970 time 0.2844 (0.2874) loss 4.8228 (4.0996) grad_norm 1.0363 (inf) [2021-04-15 17:00:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][630/1251] eta 0:02:58 lr 0.000970 time 0.2877 (0.2873) loss 4.2600 (4.1002) grad_norm 1.1061 (inf) [2021-04-15 17:00:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][640/1251] eta 0:02:55 lr 0.000970 time 0.2794 (0.2871) loss 4.6490 (4.1020) grad_norm 0.9781 (inf) [2021-04-15 17:00:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][650/1251] eta 0:02:52 lr 0.000970 time 0.2854 (0.2870) loss 2.9235 (4.0986) grad_norm 1.0599 (inf) [2021-04-15 17:00:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][660/1251] eta 0:02:49 lr 0.000970 time 0.2652 (0.2867) loss 4.5879 (4.0989) grad_norm 1.0624 (inf) [2021-04-15 17:00:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][670/1251] eta 0:02:46 lr 0.000970 time 0.2788 (0.2866) loss 4.3104 (4.0970) grad_norm 1.2256 (inf) [2021-04-15 17:00:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][680/1251] eta 0:02:43 lr 0.000970 time 0.2671 (0.2865) loss 4.5486 (4.0999) grad_norm 0.9326 (inf) [2021-04-15 17:00:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][690/1251] eta 0:02:40 lr 0.000970 time 0.2945 (0.2865) loss 4.9736 (4.0991) grad_norm 1.3306 (inf) [2021-04-15 17:00:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][700/1251] eta 0:02:37 lr 0.000970 time 0.3268 (0.2864) loss 4.2518 (4.1004) grad_norm 0.9431 (inf) [2021-04-15 17:00:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][710/1251] eta 0:02:35 lr 0.000970 time 0.2464 (0.2865) loss 4.5507 (4.1039) grad_norm 1.0937 (inf) [2021-04-15 17:00:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][720/1251] eta 0:02:32 lr 0.000970 time 0.2814 (0.2864) loss 4.2681 (4.1061) grad_norm 1.2163 (inf) [2021-04-15 17:00:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][730/1251] eta 0:02:29 lr 0.000970 time 0.2568 (0.2863) loss 3.6021 (4.1034) grad_norm 0.9041 (inf) [2021-04-15 17:00:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][740/1251] eta 0:02:26 lr 0.000970 time 0.2780 (0.2862) loss 4.5424 (4.1054) grad_norm 1.0949 (inf) [2021-04-15 17:00:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][750/1251] eta 0:02:23 lr 0.000970 time 0.2966 (0.2862) loss 4.0682 (4.1097) grad_norm 0.9741 (inf) [2021-04-15 17:00:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][760/1251] eta 0:02:20 lr 0.000970 time 0.2854 (0.2861) loss 5.3464 (4.1118) grad_norm 1.0252 (inf) [2021-04-15 17:00:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][770/1251] eta 0:02:17 lr 0.000970 time 0.2567 (0.2860) loss 2.8865 (4.1072) grad_norm 0.9562 (inf) [2021-04-15 17:01:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][780/1251] eta 0:02:14 lr 0.000970 time 0.2900 (0.2860) loss 2.8491 (4.1073) grad_norm 1.2230 (inf) [2021-04-15 17:01:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][790/1251] eta 0:02:11 lr 0.000970 time 0.2647 (0.2858) loss 4.8768 (4.1096) grad_norm 1.2823 (inf) [2021-04-15 17:01:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][800/1251] eta 0:02:08 lr 0.000970 time 0.2791 (0.2857) loss 4.2438 (4.1122) grad_norm 1.0049 (inf) [2021-04-15 17:01:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][810/1251] eta 0:02:05 lr 0.000970 time 0.2876 (0.2857) loss 4.5155 (4.1114) grad_norm 1.1884 (inf) [2021-04-15 17:01:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][820/1251] eta 0:02:03 lr 0.000970 time 0.2678 (0.2855) loss 3.4136 (4.1083) grad_norm 0.9818 (inf) [2021-04-15 17:01:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][830/1251] eta 0:02:00 lr 0.000970 time 0.2801 (0.2854) loss 4.4334 (4.1103) grad_norm 1.0160 (inf) [2021-04-15 17:01:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][840/1251] eta 0:01:57 lr 0.000970 time 0.3077 (0.2853) loss 3.9405 (4.1095) grad_norm 1.0271 (inf) [2021-04-15 17:01:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][850/1251] eta 0:01:54 lr 0.000970 time 0.2713 (0.2853) loss 4.4496 (4.1100) grad_norm 1.2509 (inf) [2021-04-15 17:01:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][860/1251] eta 0:01:51 lr 0.000970 time 0.2852 (0.2853) loss 4.5103 (4.1095) grad_norm 1.3749 (inf) [2021-04-15 17:01:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][870/1251] eta 0:01:48 lr 0.000970 time 0.2902 (0.2851) loss 3.3135 (4.1085) grad_norm 1.0639 (inf) [2021-04-15 17:01:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][880/1251] eta 0:01:45 lr 0.000969 time 0.2660 (0.2850) loss 4.4383 (4.1062) grad_norm 0.9679 (inf) [2021-04-15 17:01:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][890/1251] eta 0:01:42 lr 0.000969 time 0.2662 (0.2850) loss 3.9141 (4.1022) grad_norm 1.1768 (inf) [2021-04-15 17:01:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][900/1251] eta 0:01:39 lr 0.000969 time 0.2782 (0.2848) loss 4.2258 (4.0985) grad_norm 0.9956 (inf) [2021-04-15 17:01:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][910/1251] eta 0:01:37 lr 0.000969 time 0.2934 (0.2848) loss 4.1853 (4.0967) grad_norm 0.9947 (inf) [2021-04-15 17:01:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][920/1251] eta 0:01:34 lr 0.000969 time 0.2872 (0.2847) loss 2.7610 (4.0962) grad_norm 1.2280 (inf) [2021-04-15 17:01:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][930/1251] eta 0:01:31 lr 0.000969 time 0.2795 (0.2847) loss 5.2334 (4.0991) grad_norm 1.0733 (inf) [2021-04-15 17:01:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][940/1251] eta 0:01:28 lr 0.000969 time 0.3011 (0.2846) loss 3.4836 (4.1019) grad_norm 1.1001 (inf) [2021-04-15 17:01:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][950/1251] eta 0:01:25 lr 0.000969 time 0.2667 (0.2845) loss 4.6238 (4.1037) grad_norm 1.0943 (inf) [2021-04-15 17:01:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][960/1251] eta 0:01:22 lr 0.000969 time 0.2710 (0.2845) loss 4.4707 (4.1041) grad_norm 1.0040 (inf) [2021-04-15 17:01:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][970/1251] eta 0:01:19 lr 0.000969 time 0.2787 (0.2844) loss 4.7856 (4.1046) grad_norm 1.0368 (inf) [2021-04-15 17:01:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][980/1251] eta 0:01:17 lr 0.000969 time 0.2666 (0.2843) loss 3.8994 (4.1012) grad_norm 0.9323 (inf) [2021-04-15 17:02:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][990/1251] eta 0:01:14 lr 0.000969 time 0.2719 (0.2843) loss 4.6754 (4.1020) grad_norm 1.1890 (inf) [2021-04-15 17:02:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][1000/1251] eta 0:01:11 lr 0.000969 time 0.2678 (0.2843) loss 4.3678 (4.1033) grad_norm 0.9708 (inf) [2021-04-15 17:02:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][1010/1251] eta 0:01:08 lr 0.000969 time 0.2746 (0.2843) loss 3.1501 (4.1013) grad_norm 1.1676 (inf) [2021-04-15 17:02:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][1020/1251] eta 0:01:05 lr 0.000969 time 0.2705 (0.2842) loss 2.7641 (4.1022) grad_norm 0.9035 (inf) [2021-04-15 17:02:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][1030/1251] eta 0:01:02 lr 0.000969 time 0.2658 (0.2841) loss 3.8881 (4.1012) grad_norm 1.0637 (inf) [2021-04-15 17:02:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][1040/1251] eta 0:00:59 lr 0.000969 time 0.2843 (0.2841) loss 4.2179 (4.1017) grad_norm 0.9826 (inf) [2021-04-15 17:02:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][1050/1251] eta 0:00:57 lr 0.000969 time 0.2727 (0.2840) loss 3.7355 (4.1027) grad_norm 0.9175 (inf) [2021-04-15 17:02:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][1060/1251] eta 0:00:54 lr 0.000969 time 0.2836 (0.2839) loss 4.0673 (4.1044) grad_norm 0.9904 (inf) [2021-04-15 17:02:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][1070/1251] eta 0:00:51 lr 0.000969 time 0.2649 (0.2841) loss 4.2354 (4.1055) grad_norm 1.0229 (inf) [2021-04-15 17:02:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][1080/1251] eta 0:00:48 lr 0.000969 time 0.2731 (0.2840) loss 4.7916 (4.1082) grad_norm 0.9973 (inf) [2021-04-15 17:02:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][1090/1251] eta 0:00:45 lr 0.000969 time 0.2709 (0.2840) loss 4.4183 (4.1106) grad_norm 1.2581 (inf) [2021-04-15 17:02:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][1100/1251] eta 0:00:42 lr 0.000969 time 0.2771 (0.2839) loss 4.2942 (4.1117) grad_norm 0.9693 (inf) [2021-04-15 17:02:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][1110/1251] eta 0:00:40 lr 0.000969 time 0.2898 (0.2838) loss 4.7151 (4.1132) grad_norm 1.2223 (inf) [2021-04-15 17:02:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][1120/1251] eta 0:00:37 lr 0.000969 time 0.2565 (0.2838) loss 3.5303 (4.1133) grad_norm 1.0422 (inf) [2021-04-15 17:02:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][1130/1251] eta 0:00:34 lr 0.000969 time 0.2810 (0.2837) loss 3.7617 (4.1113) grad_norm 1.1613 (inf) [2021-04-15 17:02:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][1140/1251] eta 0:00:31 lr 0.000969 time 0.2750 (0.2836) loss 4.3840 (4.1115) grad_norm 1.3340 (inf) [2021-04-15 17:02:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][1150/1251] eta 0:00:28 lr 0.000969 time 0.2937 (0.2837) loss 4.1330 (4.1120) grad_norm 1.0081 (inf) [2021-04-15 17:02:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][1160/1251] eta 0:00:25 lr 0.000969 time 0.2615 (0.2836) loss 3.9944 (4.1150) grad_norm 1.4366 (inf) [2021-04-15 17:02:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][1170/1251] eta 0:00:22 lr 0.000969 time 0.2791 (0.2836) loss 4.9665 (4.1145) grad_norm 1.5740 (inf) [2021-04-15 17:02:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][1180/1251] eta 0:00:20 lr 0.000969 time 0.2693 (0.2836) loss 3.5565 (4.1141) grad_norm 1.5990 (inf) [2021-04-15 17:02:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][1190/1251] eta 0:00:17 lr 0.000969 time 0.2690 (0.2835) loss 4.3509 (4.1140) grad_norm 0.9627 (inf) [2021-04-15 17:02:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][1200/1251] eta 0:00:14 lr 0.000969 time 0.2751 (0.2835) loss 5.2915 (4.1158) grad_norm 1.1208 (inf) [2021-04-15 17:03:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][1210/1251] eta 0:00:11 lr 0.000969 time 0.2790 (0.2834) loss 2.9480 (4.1139) grad_norm 1.4815 (inf) [2021-04-15 17:03:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][1220/1251] eta 0:00:08 lr 0.000969 time 0.2759 (0.2833) loss 3.5881 (4.1141) grad_norm 1.1250 (inf) [2021-04-15 17:03:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][1230/1251] eta 0:00:05 lr 0.000969 time 0.2726 (0.2833) loss 3.6619 (4.1138) grad_norm 0.9669 (inf) [2021-04-15 17:03:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][1240/1251] eta 0:00:03 lr 0.000969 time 0.2540 (0.2832) loss 3.8846 (4.1127) grad_norm 0.8816 (inf) [2021-04-15 17:03:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [33/300][1250/1251] eta 0:00:00 lr 0.000969 time 0.2716 (0.2831) loss 3.6757 (4.1150) grad_norm 1.5073 (inf) [2021-04-15 17:03:15 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 33 training takes 0:05:56 [2021-04-15 17:03:15 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_33.pth saving...... [2021-04-15 17:03:24 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_33.pth saved !!! [2021-04-15 17:03:25 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.147 (1.147) Loss 1.5636 (1.5636) Acc@1 64.746 (64.746) Acc@5 87.793 (87.793) [2021-04-15 17:03:26 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.126 (0.220) Loss 1.4479 (1.4580) Acc@1 67.285 (67.294) Acc@5 88.379 (88.379) [2021-04-15 17:03:28 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.110 (0.212) Loss 1.4172 (1.4429) Acc@1 68.262 (67.518) Acc@5 89.453 (88.667) [2021-04-15 17:03:31 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 1.008 (0.238) Loss 1.4781 (1.4604) Acc@1 65.820 (67.065) Acc@5 87.305 (88.335) [2021-04-15 17:03:32 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.211) Loss 1.4477 (1.4577) Acc@1 68.457 (67.195) Acc@5 87.891 (88.293) [2021-04-15 17:03:35 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 67.316 Acc@5 88.296 [2021-04-15 17:03:35 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 67.3% [2021-04-15 17:03:35 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 67.32% [2021-04-15 17:03:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][0/1251] eta 1:06:22 lr 0.000969 time 3.1837 (3.1837) loss 3.9749 (3.9749) grad_norm 1.4908 (1.4908) [2021-04-15 17:03:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][10/1251] eta 0:11:50 lr 0.000969 time 0.3007 (0.5727) loss 4.3156 (4.1487) grad_norm 1.0717 (1.1087) [2021-04-15 17:03:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][20/1251] eta 0:08:52 lr 0.000969 time 0.2986 (0.4326) loss 4.9028 (3.9005) grad_norm 1.0273 (1.0745) [2021-04-15 17:03:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][30/1251] eta 0:07:48 lr 0.000969 time 0.3032 (0.3839) loss 4.5629 (3.8967) grad_norm 1.1266 (1.0489) [2021-04-15 17:03:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][40/1251] eta 0:07:14 lr 0.000969 time 0.2920 (0.3592) loss 4.4207 (3.9112) grad_norm 1.5676 (1.0721) [2021-04-15 17:03:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][50/1251] eta 0:06:52 lr 0.000969 time 0.3032 (0.3435) loss 3.8747 (3.9080) grad_norm 1.3472 (1.1024) [2021-04-15 17:03:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][60/1251] eta 0:06:38 lr 0.000969 time 0.3193 (0.3342) loss 4.2013 (3.9560) grad_norm 0.9530 (1.1102) [2021-04-15 17:03:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][70/1251] eta 0:06:25 lr 0.000969 time 0.2881 (0.3267) loss 3.4915 (3.9966) grad_norm 1.3149 (1.1074) [2021-04-15 17:04:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][80/1251] eta 0:06:16 lr 0.000969 time 0.2836 (0.3213) loss 4.9627 (3.9853) grad_norm 1.1170 (1.1057) [2021-04-15 17:04:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][90/1251] eta 0:06:07 lr 0.000969 time 0.2949 (0.3166) loss 4.6925 (4.0079) grad_norm 1.0620 (1.1014) [2021-04-15 17:04:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][100/1251] eta 0:06:00 lr 0.000969 time 0.3189 (0.3135) loss 4.4554 (4.0428) grad_norm 1.0988 (1.1034) [2021-04-15 17:04:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][110/1251] eta 0:05:54 lr 0.000969 time 0.2579 (0.3110) loss 4.4462 (4.0328) grad_norm 1.0162 (1.0955) [2021-04-15 17:04:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][120/1251] eta 0:05:48 lr 0.000969 time 0.2736 (0.3084) loss 4.5928 (4.0452) grad_norm 1.1334 (1.0973) [2021-04-15 17:04:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][130/1251] eta 0:05:44 lr 0.000969 time 0.4294 (0.3076) loss 3.4208 (4.0494) grad_norm 1.1059 (1.0985) [2021-04-15 17:04:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][140/1251] eta 0:05:39 lr 0.000969 time 0.2944 (0.3057) loss 4.3680 (4.0589) grad_norm 1.1804 (1.1003) [2021-04-15 17:04:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][150/1251] eta 0:05:35 lr 0.000969 time 0.2692 (0.3044) loss 4.3209 (4.0529) grad_norm 1.0706 (1.1002) [2021-04-15 17:04:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][160/1251] eta 0:05:30 lr 0.000969 time 0.2741 (0.3030) loss 4.4552 (4.0627) grad_norm 1.0632 (1.1002) [2021-04-15 17:04:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][170/1251] eta 0:05:26 lr 0.000969 time 0.2923 (0.3022) loss 4.1775 (4.0732) grad_norm 1.2196 (1.0988) [2021-04-15 17:04:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][180/1251] eta 0:05:22 lr 0.000969 time 0.2855 (0.3012) loss 4.1908 (4.0707) grad_norm 1.0170 (1.0974) [2021-04-15 17:04:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][190/1251] eta 0:05:18 lr 0.000969 time 0.2668 (0.3001) loss 4.9287 (4.0831) grad_norm 0.9701 (1.0975) [2021-04-15 17:04:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][200/1251] eta 0:05:14 lr 0.000969 time 0.2799 (0.2991) loss 2.7591 (4.0845) grad_norm 1.5405 (1.0969) [2021-04-15 17:04:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][210/1251] eta 0:05:10 lr 0.000969 time 0.2785 (0.2982) loss 3.4425 (4.0840) grad_norm 1.2232 (1.0989) [2021-04-15 17:04:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][220/1251] eta 0:05:06 lr 0.000969 time 0.3060 (0.2975) loss 4.0182 (4.0882) grad_norm 1.1254 (1.1017) [2021-04-15 17:04:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][230/1251] eta 0:05:03 lr 0.000969 time 0.2849 (0.2970) loss 4.7043 (4.0893) grad_norm 0.9058 (1.1023) [2021-04-15 17:04:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][240/1251] eta 0:04:59 lr 0.000969 time 0.2943 (0.2965) loss 4.4321 (4.1047) grad_norm 1.0297 (1.1010) [2021-04-15 17:04:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][250/1251] eta 0:04:56 lr 0.000969 time 0.2841 (0.2959) loss 3.4148 (4.1035) grad_norm 1.0417 (1.1000) [2021-04-15 17:04:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][260/1251] eta 0:04:52 lr 0.000969 time 0.2908 (0.2953) loss 4.6217 (4.1014) grad_norm 1.0021 (1.0967) [2021-04-15 17:04:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][270/1251] eta 0:04:49 lr 0.000969 time 0.3029 (0.2949) loss 3.6693 (4.0918) grad_norm 1.0383 (1.0961) [2021-04-15 17:04:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][280/1251] eta 0:04:45 lr 0.000969 time 0.2976 (0.2944) loss 4.5545 (4.0968) grad_norm 0.9735 (1.0941) [2021-04-15 17:05:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][290/1251] eta 0:04:42 lr 0.000969 time 0.2902 (0.2945) loss 5.0088 (4.0962) grad_norm 1.1603 (1.0946) [2021-04-15 17:05:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][300/1251] eta 0:04:39 lr 0.000969 time 0.2603 (0.2940) loss 3.5270 (4.0921) grad_norm 1.3086 (1.0964) [2021-04-15 17:05:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][310/1251] eta 0:04:36 lr 0.000969 time 0.2520 (0.2935) loss 3.3654 (4.0917) grad_norm 0.9955 (1.0964) [2021-04-15 17:05:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][320/1251] eta 0:04:33 lr 0.000968 time 0.3015 (0.2934) loss 3.8084 (4.0950) grad_norm 0.9750 (1.0968) [2021-04-15 17:05:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][330/1251] eta 0:04:29 lr 0.000968 time 0.2985 (0.2929) loss 3.7332 (4.1025) grad_norm 0.9657 (1.0940) [2021-04-15 17:05:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][340/1251] eta 0:04:26 lr 0.000968 time 0.2822 (0.2926) loss 4.3090 (4.1039) grad_norm 0.9761 (1.0921) [2021-04-15 17:05:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][350/1251] eta 0:04:23 lr 0.000968 time 0.2782 (0.2923) loss 3.8473 (4.1041) grad_norm 1.0852 (1.0928) [2021-04-15 17:05:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][360/1251] eta 0:04:20 lr 0.000968 time 0.2678 (0.2923) loss 3.7547 (4.0966) grad_norm 1.1405 (1.0928) [2021-04-15 17:05:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][370/1251] eta 0:04:17 lr 0.000968 time 0.2728 (0.2919) loss 4.4472 (4.1051) grad_norm 0.9253 (1.0908) [2021-04-15 17:05:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][380/1251] eta 0:04:14 lr 0.000968 time 0.2737 (0.2918) loss 3.3491 (4.1071) grad_norm 1.2020 (1.0908) [2021-04-15 17:05:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][390/1251] eta 0:04:10 lr 0.000968 time 0.2806 (0.2914) loss 4.9240 (4.1199) grad_norm 1.1075 (1.0905) [2021-04-15 17:05:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][400/1251] eta 0:04:07 lr 0.000968 time 0.2726 (0.2911) loss 3.2151 (4.1238) grad_norm 0.9628 (1.0910) [2021-04-15 17:05:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][410/1251] eta 0:04:04 lr 0.000968 time 0.2856 (0.2908) loss 4.3921 (4.1199) grad_norm 1.3561 (1.0921) [2021-04-15 17:05:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][420/1251] eta 0:04:01 lr 0.000968 time 0.2717 (0.2906) loss 4.3530 (4.1191) grad_norm 1.1810 (1.0933) [2021-04-15 17:05:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][430/1251] eta 0:03:58 lr 0.000968 time 0.2790 (0.2905) loss 4.5865 (4.1239) grad_norm 1.1427 (1.0945) [2021-04-15 17:05:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][440/1251] eta 0:03:55 lr 0.000968 time 0.2807 (0.2903) loss 4.5486 (4.1239) grad_norm 1.0916 (1.0940) [2021-04-15 17:05:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][450/1251] eta 0:03:52 lr 0.000968 time 0.2845 (0.2901) loss 4.3461 (4.1226) grad_norm 1.1600 (1.0933) [2021-04-15 17:05:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][460/1251] eta 0:03:49 lr 0.000968 time 0.2550 (0.2900) loss 2.9451 (4.1188) grad_norm 1.0524 (1.0940) [2021-04-15 17:05:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][470/1251] eta 0:03:46 lr 0.000968 time 0.2743 (0.2897) loss 3.9522 (4.1155) grad_norm 1.1048 (1.0934) [2021-04-15 17:05:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][480/1251] eta 0:03:43 lr 0.000968 time 0.2948 (0.2896) loss 3.6486 (4.1179) grad_norm 1.3714 (1.0938) [2021-04-15 17:05:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][490/1251] eta 0:03:40 lr 0.000968 time 0.2792 (0.2894) loss 4.7520 (4.1150) grad_norm 1.1148 (1.0941) [2021-04-15 17:06:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][500/1251] eta 0:03:37 lr 0.000968 time 0.2665 (0.2892) loss 4.3309 (4.1198) grad_norm 1.0079 (1.0929) [2021-04-15 17:06:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][510/1251] eta 0:03:34 lr 0.000968 time 0.2775 (0.2890) loss 4.6312 (4.1245) grad_norm 1.0401 (1.0941) [2021-04-15 17:06:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][520/1251] eta 0:03:31 lr 0.000968 time 0.2809 (0.2888) loss 3.5170 (4.1184) grad_norm 1.1775 (1.0951) [2021-04-15 17:06:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][530/1251] eta 0:03:28 lr 0.000968 time 0.3139 (0.2886) loss 4.1468 (4.1176) grad_norm 1.3264 (1.0970) [2021-04-15 17:06:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][540/1251] eta 0:03:25 lr 0.000968 time 0.2766 (0.2886) loss 3.9659 (4.1226) grad_norm 1.0525 (1.0953) [2021-04-15 17:06:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][550/1251] eta 0:03:22 lr 0.000968 time 0.2583 (0.2884) loss 4.3860 (4.1237) grad_norm 1.0012 (1.0951) [2021-04-15 17:06:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][560/1251] eta 0:03:19 lr 0.000968 time 0.2960 (0.2883) loss 4.1961 (4.1239) grad_norm 0.9781 (1.0941) [2021-04-15 17:06:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][570/1251] eta 0:03:16 lr 0.000968 time 0.2731 (0.2881) loss 4.7573 (4.1258) grad_norm 1.1777 (1.0956) [2021-04-15 17:06:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][580/1251] eta 0:03:13 lr 0.000968 time 0.2886 (0.2879) loss 4.3213 (4.1254) grad_norm 1.1667 (1.0951) [2021-04-15 17:06:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][590/1251] eta 0:03:10 lr 0.000968 time 0.2806 (0.2878) loss 3.5738 (4.1216) grad_norm 1.0169 (1.0956) [2021-04-15 17:06:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][600/1251] eta 0:03:07 lr 0.000968 time 0.2651 (0.2876) loss 4.1036 (4.1213) grad_norm 0.9962 (1.0946) [2021-04-15 17:06:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][610/1251] eta 0:03:04 lr 0.000968 time 0.2707 (0.2874) loss 3.2045 (4.1191) grad_norm 1.1251 (1.0932) [2021-04-15 17:06:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][620/1251] eta 0:03:01 lr 0.000968 time 0.2816 (0.2873) loss 4.8730 (4.1186) grad_norm 1.0249 (1.0921) [2021-04-15 17:06:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][630/1251] eta 0:02:58 lr 0.000968 time 0.2591 (0.2871) loss 5.1099 (4.1212) grad_norm 1.1303 (1.0923) [2021-04-15 17:06:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][640/1251] eta 0:02:55 lr 0.000968 time 0.2876 (0.2869) loss 4.4443 (4.1224) grad_norm 0.9985 (1.0922) [2021-04-15 17:06:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][650/1251] eta 0:02:52 lr 0.000968 time 0.2846 (0.2868) loss 4.3622 (4.1226) grad_norm 1.0987 (1.0926) [2021-04-15 17:06:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][660/1251] eta 0:02:49 lr 0.000968 time 0.2622 (0.2866) loss 3.9640 (4.1177) grad_norm 0.8420 (1.0920) [2021-04-15 17:06:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][670/1251] eta 0:02:46 lr 0.000968 time 0.2906 (0.2866) loss 3.5378 (4.1176) grad_norm 1.1463 (1.0934) [2021-04-15 17:06:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][680/1251] eta 0:02:43 lr 0.000968 time 0.2659 (0.2865) loss 2.8872 (4.1101) grad_norm 1.1097 (1.0930) [2021-04-15 17:06:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][690/1251] eta 0:02:40 lr 0.000968 time 0.2817 (0.2863) loss 4.1940 (4.1087) grad_norm 1.0408 (1.0934) [2021-04-15 17:06:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][700/1251] eta 0:02:37 lr 0.000968 time 0.2734 (0.2863) loss 4.3905 (4.1086) grad_norm 1.1140 (1.0932) [2021-04-15 17:06:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][710/1251] eta 0:02:34 lr 0.000968 time 0.2661 (0.2863) loss 4.4579 (4.1094) grad_norm 1.1253 (1.0943) [2021-04-15 17:07:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][720/1251] eta 0:02:31 lr 0.000968 time 0.2705 (0.2862) loss 4.6290 (4.1085) grad_norm 0.9371 (1.0945) [2021-04-15 17:07:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][730/1251] eta 0:02:29 lr 0.000968 time 0.2881 (0.2861) loss 4.7903 (4.1076) grad_norm 1.2088 (1.0943) [2021-04-15 17:07:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][740/1251] eta 0:02:26 lr 0.000968 time 0.2805 (0.2860) loss 4.3022 (4.1087) grad_norm 0.9167 (1.0944) [2021-04-15 17:07:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][750/1251] eta 0:02:23 lr 0.000968 time 0.2816 (0.2859) loss 3.7408 (4.1073) grad_norm 1.0937 (1.0939) [2021-04-15 17:07:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][760/1251] eta 0:02:20 lr 0.000968 time 0.2805 (0.2857) loss 3.7100 (4.1052) grad_norm 1.1358 (1.0943) [2021-04-15 17:07:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][770/1251] eta 0:02:17 lr 0.000968 time 0.2717 (0.2857) loss 4.4862 (4.1096) grad_norm 1.1666 (1.0950) [2021-04-15 17:07:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][780/1251] eta 0:02:14 lr 0.000968 time 0.2608 (0.2856) loss 3.5487 (4.1130) grad_norm 0.9655 (1.0943) [2021-04-15 17:07:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][790/1251] eta 0:02:11 lr 0.000968 time 0.2682 (0.2855) loss 2.8196 (4.1108) grad_norm 1.1286 (1.0941) [2021-04-15 17:07:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][800/1251] eta 0:02:08 lr 0.000968 time 0.2461 (0.2854) loss 4.0544 (4.1111) grad_norm 1.2450 (1.0950) [2021-04-15 17:07:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][810/1251] eta 0:02:05 lr 0.000968 time 0.2895 (0.2853) loss 4.2522 (4.1137) grad_norm 1.2033 (1.0948) [2021-04-15 17:07:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][820/1251] eta 0:02:02 lr 0.000968 time 0.2664 (0.2853) loss 4.8211 (4.1134) grad_norm 0.9833 (1.0947) [2021-04-15 17:07:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][830/1251] eta 0:02:00 lr 0.000968 time 0.2657 (0.2851) loss 4.2591 (4.1136) grad_norm 1.3585 (1.0957) [2021-04-15 17:07:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][840/1251] eta 0:01:57 lr 0.000968 time 0.2951 (0.2850) loss 4.3158 (4.1129) grad_norm 0.9738 (1.0952) [2021-04-15 17:07:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][850/1251] eta 0:01:54 lr 0.000968 time 0.2715 (0.2849) loss 4.6905 (4.1111) grad_norm 0.9874 (1.0957) [2021-04-15 17:07:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][860/1251] eta 0:01:51 lr 0.000968 time 0.2705 (0.2848) loss 3.3717 (4.1128) grad_norm 1.6787 (1.0965) [2021-04-15 17:07:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][870/1251] eta 0:01:48 lr 0.000968 time 0.3026 (0.2849) loss 3.8713 (4.1128) grad_norm 1.1490 (1.0963) [2021-04-15 17:07:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][880/1251] eta 0:01:45 lr 0.000968 time 0.2733 (0.2848) loss 4.1510 (4.1151) grad_norm 1.0778 (1.0962) [2021-04-15 17:07:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][890/1251] eta 0:01:42 lr 0.000968 time 0.2855 (0.2848) loss 2.9403 (4.1153) grad_norm 1.1399 (1.0972) [2021-04-15 17:07:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][900/1251] eta 0:01:39 lr 0.000968 time 0.2844 (0.2847) loss 3.9760 (4.1179) grad_norm 1.1893 (1.0980) [2021-04-15 17:07:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][910/1251] eta 0:01:37 lr 0.000968 time 0.2796 (0.2846) loss 4.3951 (4.1180) grad_norm 0.9969 (1.0984) [2021-04-15 17:07:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][920/1251] eta 0:01:34 lr 0.000968 time 0.2622 (0.2846) loss 4.8560 (4.1176) grad_norm 0.9069 (1.0972) [2021-04-15 17:08:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][930/1251] eta 0:01:31 lr 0.000968 time 0.2728 (0.2845) loss 3.6394 (4.1175) grad_norm 0.9235 (1.0965) [2021-04-15 17:08:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][940/1251] eta 0:01:28 lr 0.000968 time 0.2893 (0.2846) loss 3.8834 (4.1176) grad_norm 1.1618 (1.0963) [2021-04-15 17:08:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][950/1251] eta 0:01:25 lr 0.000968 time 0.2944 (0.2845) loss 4.6880 (4.1181) grad_norm 1.5025 (1.0968) [2021-04-15 17:08:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][960/1251] eta 0:01:22 lr 0.000968 time 0.2827 (0.2845) loss 4.5485 (4.1192) grad_norm 1.4644 (1.0979) [2021-04-15 17:08:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][970/1251] eta 0:01:19 lr 0.000968 time 0.2754 (0.2844) loss 4.4319 (4.1235) grad_norm 0.9844 (1.0985) [2021-04-15 17:08:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][980/1251] eta 0:01:17 lr 0.000968 time 0.2764 (0.2844) loss 4.6567 (4.1263) grad_norm 1.0505 (1.0985) [2021-04-15 17:08:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][990/1251] eta 0:01:14 lr 0.000968 time 0.2830 (0.2843) loss 4.1714 (4.1273) grad_norm 1.0368 (1.0980) [2021-04-15 17:08:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][1000/1251] eta 0:01:11 lr 0.000967 time 0.2970 (0.2843) loss 4.6110 (4.1284) grad_norm 1.0784 (1.0976) [2021-04-15 17:08:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][1010/1251] eta 0:01:08 lr 0.000967 time 0.2572 (0.2842) loss 3.3443 (4.1299) grad_norm 1.1131 (1.0972) [2021-04-15 17:08:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][1020/1251] eta 0:01:05 lr 0.000967 time 0.2893 (0.2842) loss 4.7810 (4.1241) grad_norm 1.0616 (1.0971) [2021-04-15 17:08:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][1030/1251] eta 0:01:02 lr 0.000967 time 0.2771 (0.2841) loss 3.8928 (4.1238) grad_norm 1.2253 (1.0966) [2021-04-15 17:08:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][1040/1251] eta 0:00:59 lr 0.000967 time 0.2880 (0.2840) loss 5.0282 (4.1232) grad_norm 1.1190 (1.0970) [2021-04-15 17:08:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][1050/1251] eta 0:00:57 lr 0.000967 time 0.2747 (0.2840) loss 4.5480 (4.1246) grad_norm 1.0018 (1.0966) [2021-04-15 17:08:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][1060/1251] eta 0:00:54 lr 0.000967 time 0.2910 (0.2839) loss 3.5569 (4.1220) grad_norm 1.0256 (1.0962) [2021-04-15 17:08:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][1070/1251] eta 0:00:51 lr 0.000967 time 0.2612 (0.2839) loss 2.9667 (4.1210) grad_norm 0.9693 (1.0954) [2021-04-15 17:08:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][1080/1251] eta 0:00:48 lr 0.000967 time 0.3089 (0.2839) loss 4.2669 (4.1220) grad_norm 1.0706 (1.0945) [2021-04-15 17:08:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][1090/1251] eta 0:00:45 lr 0.000967 time 0.2720 (0.2838) loss 4.2015 (4.1226) grad_norm 1.0178 (1.0940) [2021-04-15 17:08:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][1100/1251] eta 0:00:42 lr 0.000967 time 0.2657 (0.2837) loss 4.8009 (4.1240) grad_norm 0.9809 (1.0935) [2021-04-15 17:08:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][1110/1251] eta 0:00:39 lr 0.000967 time 0.2668 (0.2836) loss 3.4965 (4.1224) grad_norm 1.0050 (1.0931) [2021-04-15 17:08:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][1120/1251] eta 0:00:37 lr 0.000967 time 0.2757 (0.2837) loss 3.2853 (4.1247) grad_norm 0.9714 (1.0928) [2021-04-15 17:08:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][1130/1251] eta 0:00:34 lr 0.000967 time 0.2702 (0.2837) loss 3.0554 (4.1220) grad_norm 1.2348 (1.0925) [2021-04-15 17:08:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][1140/1251] eta 0:00:31 lr 0.000967 time 0.2874 (0.2836) loss 3.2543 (4.1218) grad_norm 1.4105 (1.0928) [2021-04-15 17:09:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][1150/1251] eta 0:00:28 lr 0.000967 time 0.2710 (0.2836) loss 3.7955 (4.1207) grad_norm 1.1075 (1.0932) [2021-04-15 17:09:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][1160/1251] eta 0:00:25 lr 0.000967 time 0.2584 (0.2835) loss 3.1295 (4.1207) grad_norm 1.0413 (1.0935) [2021-04-15 17:09:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][1170/1251] eta 0:00:22 lr 0.000967 time 0.2723 (0.2835) loss 4.4259 (4.1213) grad_norm 1.0896 (1.0930) [2021-04-15 17:09:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][1180/1251] eta 0:00:20 lr 0.000967 time 0.2665 (0.2834) loss 3.7912 (4.1217) grad_norm 1.1323 (1.0932) [2021-04-15 17:09:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][1190/1251] eta 0:00:17 lr 0.000967 time 0.2873 (0.2834) loss 4.3000 (4.1252) grad_norm 1.0629 (1.0934) [2021-04-15 17:09:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][1200/1251] eta 0:00:14 lr 0.000967 time 0.2734 (0.2833) loss 4.8467 (4.1247) grad_norm 0.9028 (1.0932) [2021-04-15 17:09:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][1210/1251] eta 0:00:11 lr 0.000967 time 0.2848 (0.2832) loss 3.5508 (4.1271) grad_norm 1.1727 (1.0934) [2021-04-15 17:09:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][1220/1251] eta 0:00:08 lr 0.000967 time 0.3178 (0.2832) loss 3.0996 (4.1251) grad_norm 0.9394 (1.0924) [2021-04-15 17:09:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][1230/1251] eta 0:00:05 lr 0.000967 time 0.2761 (0.2831) loss 4.3162 (4.1267) grad_norm 0.9717 (1.0922) [2021-04-15 17:09:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][1240/1251] eta 0:00:03 lr 0.000967 time 0.2537 (0.2830) loss 3.5242 (4.1268) grad_norm 1.1909 (1.0922) [2021-04-15 17:09:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [34/300][1250/1251] eta 0:00:00 lr 0.000967 time 0.2498 (0.2828) loss 4.6502 (4.1270) grad_norm 0.9379 (1.0920) [2021-04-15 17:09:31 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 34 training takes 0:05:56 [2021-04-15 17:09:31 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_34.pth saving...... [2021-04-15 17:09:45 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_34.pth saved !!! [2021-04-15 17:09:46 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.191 (1.191) Loss 1.4054 (1.4054) Acc@1 69.043 (69.043) Acc@5 88.574 (88.574) [2021-04-15 17:09:47 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.165 (0.209) Loss 1.3960 (1.4163) Acc@1 69.336 (67.472) Acc@5 88.867 (88.299) [2021-04-15 17:09:50 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.122 (0.242) Loss 1.4504 (1.4226) Acc@1 67.188 (67.318) Acc@5 87.500 (88.142) [2021-04-15 17:09:52 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.103 (0.224) Loss 1.4313 (1.4315) Acc@1 66.895 (67.342) Acc@5 88.184 (88.099) [2021-04-15 17:09:54 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.251 (0.223) Loss 1.4181 (1.4320) Acc@1 66.699 (67.276) Acc@5 87.793 (88.184) [2021-04-15 17:09:58 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 67.312 Acc@5 88.162 [2021-04-15 17:09:58 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 67.3% [2021-04-15 17:09:58 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 67.32% [2021-04-15 17:10:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][0/1251] eta 0:59:49 lr 0.000967 time 2.8691 (2.8691) loss 3.8604 (3.8604) grad_norm 1.1306 (1.1306) [2021-04-15 17:10:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][10/1251] eta 0:10:41 lr 0.000967 time 0.3010 (0.5170) loss 3.0006 (4.0585) grad_norm 1.1599 (1.1044) [2021-04-15 17:10:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][20/1251] eta 0:08:17 lr 0.000967 time 0.2817 (0.4042) loss 3.1206 (4.1287) grad_norm 0.9312 (1.1050) [2021-04-15 17:10:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][30/1251] eta 0:07:26 lr 0.000967 time 0.2750 (0.3658) loss 4.9815 (4.0826) grad_norm 1.1258 (1.1065) [2021-04-15 17:10:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][40/1251] eta 0:06:59 lr 0.000967 time 0.2745 (0.3462) loss 3.1671 (4.0104) grad_norm 0.8966 (1.0838) [2021-04-15 17:10:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][50/1251] eta 0:06:41 lr 0.000967 time 0.2770 (0.3345) loss 3.0458 (3.9952) grad_norm 1.0276 (1.0618) [2021-04-15 17:10:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][60/1251] eta 0:06:29 lr 0.000967 time 0.4146 (0.3274) loss 3.5121 (4.0082) grad_norm 1.0290 (1.0551) [2021-04-15 17:10:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][70/1251] eta 0:06:18 lr 0.000967 time 0.2619 (0.3206) loss 4.1869 (4.0337) grad_norm 1.3069 (1.0636) [2021-04-15 17:10:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][80/1251] eta 0:06:11 lr 0.000967 time 0.3043 (0.3172) loss 4.5806 (4.0430) grad_norm 1.1238 (1.0752) [2021-04-15 17:10:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][90/1251] eta 0:06:03 lr 0.000967 time 0.2844 (0.3133) loss 3.5703 (4.0236) grad_norm 0.8941 (1.0736) [2021-04-15 17:10:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][100/1251] eta 0:05:56 lr 0.000967 time 0.2640 (0.3096) loss 3.9632 (3.9737) grad_norm 1.2325 (1.0795) [2021-04-15 17:10:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][110/1251] eta 0:05:50 lr 0.000967 time 0.2766 (0.3068) loss 4.9579 (3.9863) grad_norm 1.1243 (1.0883) [2021-04-15 17:10:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][120/1251] eta 0:05:44 lr 0.000967 time 0.2988 (0.3046) loss 4.3336 (3.9860) grad_norm 1.0268 (1.0839) [2021-04-15 17:10:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][130/1251] eta 0:05:39 lr 0.000967 time 0.2879 (0.3031) loss 3.6127 (3.9805) grad_norm 0.9990 (1.0847) [2021-04-15 17:10:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][140/1251] eta 0:05:35 lr 0.000967 time 0.2634 (0.3017) loss 4.0867 (3.9774) grad_norm 0.9700 (1.0765) [2021-04-15 17:10:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][150/1251] eta 0:05:31 lr 0.000967 time 0.2940 (0.3013) loss 3.9576 (3.9951) grad_norm 1.0216 (1.0775) [2021-04-15 17:10:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][160/1251] eta 0:05:27 lr 0.000967 time 0.2857 (0.3000) loss 4.1853 (4.0006) grad_norm 1.3908 (1.0852) [2021-04-15 17:10:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][170/1251] eta 0:05:22 lr 0.000967 time 0.2659 (0.2985) loss 3.8962 (3.9954) grad_norm 1.1364 (1.0930) [2021-04-15 17:10:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][180/1251] eta 0:05:18 lr 0.000967 time 0.2564 (0.2974) loss 4.2731 (4.0091) grad_norm 1.1182 (1.0925) [2021-04-15 17:10:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][190/1251] eta 0:05:14 lr 0.000967 time 0.2848 (0.2964) loss 4.2055 (4.0297) grad_norm 0.9680 (1.0887) [2021-04-15 17:10:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][200/1251] eta 0:05:10 lr 0.000967 time 0.2776 (0.2953) loss 4.0206 (4.0511) grad_norm 1.0362 (1.0857) [2021-04-15 17:11:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][210/1251] eta 0:05:06 lr 0.000967 time 0.2731 (0.2945) loss 3.3244 (4.0639) grad_norm 0.9267 (1.0818) [2021-04-15 17:11:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][220/1251] eta 0:05:02 lr 0.000967 time 0.2850 (0.2939) loss 3.2646 (4.0656) grad_norm 1.1783 (1.0850) [2021-04-15 17:11:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][230/1251] eta 0:04:59 lr 0.000967 time 0.2766 (0.2932) loss 3.8973 (4.0716) grad_norm 0.9129 (1.0862) [2021-04-15 17:11:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][240/1251] eta 0:04:55 lr 0.000967 time 0.2475 (0.2928) loss 3.5921 (4.0718) grad_norm 1.1326 (1.0861) [2021-04-15 17:11:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][250/1251] eta 0:04:53 lr 0.000967 time 0.2916 (0.2935) loss 4.1315 (4.0680) grad_norm 1.1193 (1.0882) [2021-04-15 17:11:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][260/1251] eta 0:04:50 lr 0.000967 time 0.2658 (0.2930) loss 4.5360 (4.0667) grad_norm 1.3672 (1.0903) [2021-04-15 17:11:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][270/1251] eta 0:04:47 lr 0.000967 time 0.2646 (0.2928) loss 3.5784 (4.0627) grad_norm 1.2435 (1.0939) [2021-04-15 17:11:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][280/1251] eta 0:04:43 lr 0.000967 time 0.2745 (0.2922) loss 3.9377 (4.0719) grad_norm 1.2000 (1.0954) [2021-04-15 17:11:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][290/1251] eta 0:04:40 lr 0.000967 time 0.2917 (0.2917) loss 3.7165 (4.0703) grad_norm 0.9014 (1.0969) [2021-04-15 17:11:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][300/1251] eta 0:04:36 lr 0.000967 time 0.2612 (0.2912) loss 2.8147 (4.0690) grad_norm 1.0110 (1.1003) [2021-04-15 17:11:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][310/1251] eta 0:04:33 lr 0.000967 time 0.2570 (0.2908) loss 4.3837 (4.0683) grad_norm 1.4343 (1.1026) [2021-04-15 17:11:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][320/1251] eta 0:04:30 lr 0.000967 time 0.2745 (0.2910) loss 4.1494 (4.0705) grad_norm 0.9269 (1.1001) [2021-04-15 17:11:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][330/1251] eta 0:04:27 lr 0.000967 time 0.2773 (0.2906) loss 4.2998 (4.0687) grad_norm 0.9406 (1.0998) [2021-04-15 17:11:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][340/1251] eta 0:04:24 lr 0.000967 time 0.2827 (0.2904) loss 3.5988 (4.0720) grad_norm 0.9469 (1.0971) [2021-04-15 17:11:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][350/1251] eta 0:04:21 lr 0.000967 time 0.2873 (0.2902) loss 4.9029 (4.0756) grad_norm 1.0176 (1.0942) [2021-04-15 17:11:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][360/1251] eta 0:04:18 lr 0.000967 time 0.2749 (0.2899) loss 4.2276 (4.0662) grad_norm 0.9544 (1.0922) [2021-04-15 17:11:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][370/1251] eta 0:04:15 lr 0.000967 time 0.2708 (0.2896) loss 3.9521 (4.0684) grad_norm 1.2250 (1.0952) [2021-04-15 17:11:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][380/1251] eta 0:04:11 lr 0.000967 time 0.2751 (0.2893) loss 4.4564 (4.0687) grad_norm 1.0414 (1.0962) [2021-04-15 17:11:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][390/1251] eta 0:04:08 lr 0.000967 time 0.2696 (0.2891) loss 3.1044 (4.0671) grad_norm 1.1652 (1.0983) [2021-04-15 17:11:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][400/1251] eta 0:04:05 lr 0.000967 time 0.2641 (0.2887) loss 4.5283 (4.0643) grad_norm 0.8714 (1.0980) [2021-04-15 17:11:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][410/1251] eta 0:04:02 lr 0.000967 time 0.2751 (0.2886) loss 4.8311 (4.0641) grad_norm 1.0213 (1.0988) [2021-04-15 17:11:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][420/1251] eta 0:03:59 lr 0.000966 time 0.2886 (0.2885) loss 4.3695 (4.0727) grad_norm 1.0041 (1.0990) [2021-04-15 17:12:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][430/1251] eta 0:03:56 lr 0.000966 time 0.2879 (0.2884) loss 4.6474 (4.0773) grad_norm 1.1360 (1.0976) [2021-04-15 17:12:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][440/1251] eta 0:03:53 lr 0.000966 time 0.2841 (0.2882) loss 4.1692 (4.0673) grad_norm 0.9892 (1.0983) [2021-04-15 17:12:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][450/1251] eta 0:03:50 lr 0.000966 time 0.2921 (0.2880) loss 4.9622 (4.0629) grad_norm 1.0254 (1.0976) [2021-04-15 17:12:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][460/1251] eta 0:03:47 lr 0.000966 time 0.2783 (0.2877) loss 4.4890 (4.0661) grad_norm 0.9619 (1.0959) [2021-04-15 17:12:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][470/1251] eta 0:03:44 lr 0.000966 time 0.2727 (0.2875) loss 3.7167 (4.0666) grad_norm 1.1815 (1.0944) [2021-04-15 17:12:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][480/1251] eta 0:03:41 lr 0.000966 time 0.2898 (0.2874) loss 2.9197 (4.0615) grad_norm 1.1368 (1.0937) [2021-04-15 17:12:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][490/1251] eta 0:03:38 lr 0.000966 time 0.2642 (0.2873) loss 4.0488 (4.0565) grad_norm 0.9683 (1.0928) [2021-04-15 17:12:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][500/1251] eta 0:03:35 lr 0.000966 time 0.2904 (0.2871) loss 4.3908 (4.0544) grad_norm 1.0647 (1.0930) [2021-04-15 17:12:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][510/1251] eta 0:03:32 lr 0.000966 time 0.2610 (0.2870) loss 3.7626 (4.0465) grad_norm 1.1621 (1.0941) [2021-04-15 17:12:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][520/1251] eta 0:03:29 lr 0.000966 time 0.2919 (0.2868) loss 4.4720 (4.0456) grad_norm 1.1098 (1.0940) [2021-04-15 17:12:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][530/1251] eta 0:03:26 lr 0.000966 time 0.2783 (0.2868) loss 4.2032 (4.0441) grad_norm 1.4822 (1.0953) [2021-04-15 17:12:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][540/1251] eta 0:03:23 lr 0.000966 time 0.2745 (0.2868) loss 5.0365 (4.0403) grad_norm 1.0961 (1.0963) [2021-04-15 17:12:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][550/1251] eta 0:03:20 lr 0.000966 time 0.2738 (0.2867) loss 4.1788 (4.0436) grad_norm 0.9476 (1.0965) [2021-04-15 17:12:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][560/1251] eta 0:03:17 lr 0.000966 time 0.2883 (0.2865) loss 4.7654 (4.0439) grad_norm 1.0035 (1.0966) [2021-04-15 17:12:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][570/1251] eta 0:03:14 lr 0.000966 time 0.2544 (0.2863) loss 4.1665 (4.0476) grad_norm 1.2254 (1.0975) [2021-04-15 17:12:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][580/1251] eta 0:03:12 lr 0.000966 time 0.2642 (0.2865) loss 4.9969 (4.0497) grad_norm 1.0070 (1.0975) [2021-04-15 17:12:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][590/1251] eta 0:03:09 lr 0.000966 time 0.2545 (0.2863) loss 4.5532 (4.0516) grad_norm 1.0173 (1.0970) [2021-04-15 17:12:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][600/1251] eta 0:03:06 lr 0.000966 time 0.2625 (0.2863) loss 4.6554 (4.0570) grad_norm 1.0906 (1.0977) [2021-04-15 17:12:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][610/1251] eta 0:03:03 lr 0.000966 time 0.2843 (0.2862) loss 4.4637 (4.0586) grad_norm 1.0445 (1.0968) [2021-04-15 17:12:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][620/1251] eta 0:03:00 lr 0.000966 time 0.2726 (0.2860) loss 4.8481 (4.0559) grad_norm 1.0305 (1.0966) [2021-04-15 17:12:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][630/1251] eta 0:02:57 lr 0.000966 time 0.2836 (0.2859) loss 4.6594 (4.0544) grad_norm 1.0433 (1.0956) [2021-04-15 17:13:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][640/1251] eta 0:02:54 lr 0.000966 time 0.2510 (0.2859) loss 4.3379 (4.0554) grad_norm 1.3958 (1.0948) [2021-04-15 17:13:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][650/1251] eta 0:02:51 lr 0.000966 time 0.2670 (0.2861) loss 4.8163 (4.0534) grad_norm 1.1023 (1.0935) [2021-04-15 17:13:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][660/1251] eta 0:02:49 lr 0.000966 time 0.2727 (0.2860) loss 4.4818 (4.0528) grad_norm 1.0224 (1.0938) [2021-04-15 17:13:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][670/1251] eta 0:02:46 lr 0.000966 time 0.2795 (0.2859) loss 4.4734 (4.0491) grad_norm 1.1941 (1.0926) [2021-04-15 17:13:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][680/1251] eta 0:02:43 lr 0.000966 time 0.2773 (0.2857) loss 4.0996 (4.0468) grad_norm 1.0223 (1.0928) [2021-04-15 17:13:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][690/1251] eta 0:02:40 lr 0.000966 time 0.2776 (0.2857) loss 4.4771 (4.0500) grad_norm 1.0234 (1.0930) [2021-04-15 17:13:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][700/1251] eta 0:02:37 lr 0.000966 time 0.2824 (0.2855) loss 4.9772 (4.0533) grad_norm 1.2830 (1.0928) [2021-04-15 17:13:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][710/1251] eta 0:02:34 lr 0.000966 time 0.2715 (0.2854) loss 4.7386 (4.0537) grad_norm 0.9746 (1.0930) [2021-04-15 17:13:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][720/1251] eta 0:02:31 lr 0.000966 time 0.2676 (0.2853) loss 4.2821 (4.0563) grad_norm 1.2307 (1.0934) [2021-04-15 17:13:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][730/1251] eta 0:02:28 lr 0.000966 time 0.2637 (0.2853) loss 4.5059 (4.0547) grad_norm 1.1850 (1.0953) [2021-04-15 17:13:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][740/1251] eta 0:02:25 lr 0.000966 time 0.2557 (0.2852) loss 4.3416 (4.0578) grad_norm 1.0533 (1.0959) [2021-04-15 17:13:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][750/1251] eta 0:02:22 lr 0.000966 time 0.3051 (0.2852) loss 4.4562 (4.0536) grad_norm 0.9608 (1.0954) [2021-04-15 17:13:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][760/1251] eta 0:02:19 lr 0.000966 time 0.2651 (0.2851) loss 4.1443 (4.0513) grad_norm 1.1948 (1.0956) [2021-04-15 17:13:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][770/1251] eta 0:02:17 lr 0.000966 time 0.3030 (0.2850) loss 4.3378 (4.0510) grad_norm 1.2349 (1.0982) [2021-04-15 17:13:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][780/1251] eta 0:02:14 lr 0.000966 time 0.2858 (0.2849) loss 4.1223 (4.0534) grad_norm 1.2135 (1.1007) [2021-04-15 17:13:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][790/1251] eta 0:02:11 lr 0.000966 time 0.2678 (0.2849) loss 4.2519 (4.0550) grad_norm 1.1575 (1.1006) [2021-04-15 17:13:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][800/1251] eta 0:02:08 lr 0.000966 time 0.2585 (0.2849) loss 4.0367 (4.0564) grad_norm 1.3897 (1.1023) [2021-04-15 17:13:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][810/1251] eta 0:02:05 lr 0.000966 time 0.2659 (0.2847) loss 3.1031 (4.0568) grad_norm 1.0453 (1.1019) [2021-04-15 17:13:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][820/1251] eta 0:02:02 lr 0.000966 time 0.2995 (0.2847) loss 4.9764 (4.0573) grad_norm 1.0555 (1.1020) [2021-04-15 17:13:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][830/1251] eta 0:01:59 lr 0.000966 time 0.2726 (0.2846) loss 4.0473 (4.0616) grad_norm 0.9937 (1.1021) [2021-04-15 17:13:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][840/1251] eta 0:01:56 lr 0.000966 time 0.2570 (0.2844) loss 4.4280 (4.0629) grad_norm 1.6090 (1.1026) [2021-04-15 17:14:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][850/1251] eta 0:01:54 lr 0.000966 time 0.3063 (0.2845) loss 4.0366 (4.0647) grad_norm 1.0633 (1.1026) [2021-04-15 17:14:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][860/1251] eta 0:01:51 lr 0.000966 time 0.2804 (0.2844) loss 3.1392 (4.0608) grad_norm 1.0125 (1.1023) [2021-04-15 17:14:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][870/1251] eta 0:01:48 lr 0.000966 time 0.2713 (0.2843) loss 4.0074 (4.0564) grad_norm 1.1870 (1.1021) [2021-04-15 17:14:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][880/1251] eta 0:01:45 lr 0.000966 time 0.2693 (0.2842) loss 3.7517 (4.0567) grad_norm 1.1900 (1.1020) [2021-04-15 17:14:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][890/1251] eta 0:01:42 lr 0.000966 time 0.2597 (0.2843) loss 3.9929 (4.0537) grad_norm 0.9526 (1.1014) [2021-04-15 17:14:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][900/1251] eta 0:01:39 lr 0.000966 time 0.2610 (0.2843) loss 4.8077 (4.0567) grad_norm 0.9429 (1.1007) [2021-04-15 17:14:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][910/1251] eta 0:01:36 lr 0.000966 time 0.2907 (0.2842) loss 4.3339 (4.0573) grad_norm 0.8753 (1.1002) [2021-04-15 17:14:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][920/1251] eta 0:01:34 lr 0.000966 time 0.2842 (0.2841) loss 4.3350 (4.0586) grad_norm 1.1403 (1.1009) [2021-04-15 17:14:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][930/1251] eta 0:01:31 lr 0.000966 time 0.2857 (0.2840) loss 2.8927 (4.0554) grad_norm 1.0044 (1.1004) [2021-04-15 17:14:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][940/1251] eta 0:01:28 lr 0.000966 time 0.2926 (0.2840) loss 4.5695 (4.0560) grad_norm 1.1801 (1.1009) [2021-04-15 17:14:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][950/1251] eta 0:01:25 lr 0.000966 time 0.2757 (0.2839) loss 4.1425 (4.0566) grad_norm 1.1196 (1.1003) [2021-04-15 17:14:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][960/1251] eta 0:01:22 lr 0.000966 time 0.2892 (0.2839) loss 3.8538 (4.0557) grad_norm 1.0203 (1.0994) [2021-04-15 17:14:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][970/1251] eta 0:01:19 lr 0.000966 time 0.2777 (0.2838) loss 3.1635 (4.0528) grad_norm 0.9605 (1.0996) [2021-04-15 17:14:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][980/1251] eta 0:01:16 lr 0.000966 time 0.2739 (0.2838) loss 3.6567 (4.0520) grad_norm 0.9738 (1.0993) [2021-04-15 17:14:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][990/1251] eta 0:01:14 lr 0.000966 time 0.2911 (0.2837) loss 4.3701 (4.0541) grad_norm 1.0709 (1.0993) [2021-04-15 17:14:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][1000/1251] eta 0:01:11 lr 0.000966 time 0.2819 (0.2836) loss 4.8403 (4.0537) grad_norm 0.8890 (1.0998) [2021-04-15 17:14:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][1010/1251] eta 0:01:08 lr 0.000966 time 0.3093 (0.2835) loss 4.2440 (4.0546) grad_norm 1.3484 (1.0997) [2021-04-15 17:14:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][1020/1251] eta 0:01:05 lr 0.000966 time 0.2628 (0.2835) loss 3.4007 (4.0541) grad_norm 1.1356 (1.1002) [2021-04-15 17:14:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][1030/1251] eta 0:01:02 lr 0.000966 time 0.2747 (0.2834) loss 4.0727 (4.0528) grad_norm 1.1294 (1.1004) [2021-04-15 17:14:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][1040/1251] eta 0:00:59 lr 0.000966 time 0.2744 (0.2834) loss 3.0441 (4.0525) grad_norm 0.9911 (1.1011) [2021-04-15 17:14:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][1050/1251] eta 0:00:56 lr 0.000966 time 0.2911 (0.2833) loss 4.5607 (4.0506) grad_norm 1.2210 (1.1015) [2021-04-15 17:14:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][1060/1251] eta 0:00:54 lr 0.000966 time 0.2822 (0.2833) loss 3.5447 (4.0497) grad_norm 1.2358 (1.1012) [2021-04-15 17:15:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][1070/1251] eta 0:00:51 lr 0.000966 time 0.2617 (0.2833) loss 4.1063 (4.0524) grad_norm 0.9634 (1.1006) [2021-04-15 17:15:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][1080/1251] eta 0:00:48 lr 0.000965 time 0.2644 (0.2832) loss 4.6925 (4.0517) grad_norm 1.0594 (1.1005) [2021-04-15 17:15:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][1090/1251] eta 0:00:45 lr 0.000965 time 0.2596 (0.2832) loss 4.7825 (4.0533) grad_norm 1.1253 (1.1005) [2021-04-15 17:15:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][1100/1251] eta 0:00:42 lr 0.000965 time 0.2697 (0.2831) loss 3.9300 (4.0515) grad_norm 0.9625 (1.0996) [2021-04-15 17:15:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][1110/1251] eta 0:00:39 lr 0.000965 time 0.2989 (0.2830) loss 4.7684 (4.0511) grad_norm 1.0107 (1.0992) [2021-04-15 17:15:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][1120/1251] eta 0:00:37 lr 0.000965 time 0.2698 (0.2830) loss 4.9787 (4.0525) grad_norm 1.0205 (1.0987) [2021-04-15 17:15:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][1130/1251] eta 0:00:34 lr 0.000965 time 0.2786 (0.2830) loss 4.0491 (4.0518) grad_norm 1.2119 (1.0993) [2021-04-15 17:15:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][1140/1251] eta 0:00:31 lr 0.000965 time 0.2772 (0.2830) loss 4.4922 (4.0503) grad_norm 1.3140 (1.0995) [2021-04-15 17:15:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][1150/1251] eta 0:00:28 lr 0.000965 time 0.4276 (0.2830) loss 4.7315 (4.0528) grad_norm 0.9053 (1.0985) [2021-04-15 17:15:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][1160/1251] eta 0:00:25 lr 0.000965 time 0.2664 (0.2830) loss 2.9893 (4.0533) grad_norm 1.0325 (1.0978) [2021-04-15 17:15:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][1170/1251] eta 0:00:22 lr 0.000965 time 0.2552 (0.2829) loss 4.3990 (4.0535) grad_norm 0.8889 (1.0968) [2021-04-15 17:15:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][1180/1251] eta 0:00:20 lr 0.000965 time 0.3176 (0.2830) loss 3.9510 (4.0542) grad_norm 1.0894 (1.0970) [2021-04-15 17:15:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][1190/1251] eta 0:00:17 lr 0.000965 time 0.2739 (0.2829) loss 5.0490 (4.0526) grad_norm 0.9525 (1.0964) [2021-04-15 17:15:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][1200/1251] eta 0:00:14 lr 0.000965 time 0.2867 (0.2830) loss 4.1558 (4.0546) grad_norm 1.1780 (1.0958) [2021-04-15 17:15:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][1210/1251] eta 0:00:11 lr 0.000965 time 0.3060 (0.2829) loss 4.6078 (4.0535) grad_norm 1.0395 (1.0951) [2021-04-15 17:15:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][1220/1251] eta 0:00:08 lr 0.000965 time 0.2781 (0.2831) loss 4.4488 (4.0540) grad_norm 1.0979 (1.0949) [2021-04-15 17:15:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][1230/1251] eta 0:00:05 lr 0.000965 time 0.2774 (0.2831) loss 4.3885 (4.0541) grad_norm 1.1415 (1.0945) [2021-04-15 17:15:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][1240/1251] eta 0:00:03 lr 0.000965 time 0.2493 (0.2830) loss 5.2504 (4.0569) grad_norm 1.0700 (1.0943) [2021-04-15 17:15:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [35/300][1250/1251] eta 0:00:00 lr 0.000965 time 0.2494 (0.2827) loss 4.2888 (4.0572) grad_norm 1.3038 (1.0949) [2021-04-15 17:15:54 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 35 training takes 0:05:55 [2021-04-15 17:15:54 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_35.pth saving...... [2021-04-15 17:16:03 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_35.pth saved !!! [2021-04-15 17:16:04 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.340 (1.340) Loss 1.4119 (1.4119) Acc@1 66.699 (66.699) Acc@5 89.062 (89.062) [2021-04-15 17:16:06 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.142 (0.282) Loss 1.3294 (1.4328) Acc@1 70.605 (67.365) Acc@5 89.453 (88.707) [2021-04-15 17:16:08 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.571 (0.247) Loss 1.5107 (1.4347) Acc@1 66.504 (67.601) Acc@5 86.719 (88.621) [2021-04-15 17:16:10 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.134 (0.251) Loss 1.3915 (1.4325) Acc@1 67.871 (67.682) Acc@5 88.281 (88.533) [2021-04-15 17:16:12 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.222) Loss 1.4730 (1.4329) Acc@1 67.578 (67.676) Acc@5 88.184 (88.572) [2021-04-15 17:16:15 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 67.410 Acc@5 88.372 [2021-04-15 17:16:15 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 67.4% [2021-04-15 17:16:15 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 67.41% [2021-04-15 17:16:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][0/1251] eta 1:13:10 lr 0.000965 time 3.5097 (3.5097) loss 3.3895 (3.3895) grad_norm 1.1526 (1.1526) [2021-04-15 17:16:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][10/1251] eta 0:12:07 lr 0.000965 time 0.3002 (0.5861) loss 4.4789 (4.2413) grad_norm 1.2152 (1.1907) [2021-04-15 17:16:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][20/1251] eta 0:09:01 lr 0.000965 time 0.2629 (0.4402) loss 4.0222 (4.2689) grad_norm 0.9989 (1.1277) [2021-04-15 17:16:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][30/1251] eta 0:07:57 lr 0.000965 time 0.3110 (0.3911) loss 4.3914 (4.2027) grad_norm 1.0042 (1.1172) [2021-04-15 17:16:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][40/1251] eta 0:07:22 lr 0.000965 time 0.2754 (0.3652) loss 4.0391 (4.1381) grad_norm 0.9899 (1.1132) [2021-04-15 17:16:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][50/1251] eta 0:07:00 lr 0.000965 time 0.2719 (0.3501) loss 4.5973 (4.1606) grad_norm 1.5379 (1.1194) [2021-04-15 17:16:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][60/1251] eta 0:06:44 lr 0.000965 time 0.2786 (0.3394) loss 3.2634 (4.1470) grad_norm 1.0808 (1.1162) [2021-04-15 17:16:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][70/1251] eta 0:06:31 lr 0.000965 time 0.2760 (0.3317) loss 4.0448 (4.1763) grad_norm 1.0533 (1.1120) [2021-04-15 17:16:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][80/1251] eta 0:06:20 lr 0.000965 time 0.2767 (0.3254) loss 3.3083 (4.1480) grad_norm 1.1619 (1.1120) [2021-04-15 17:16:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][90/1251] eta 0:06:13 lr 0.000965 time 0.2653 (0.3213) loss 4.4042 (4.1593) grad_norm 0.9183 (1.1024) [2021-04-15 17:16:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][100/1251] eta 0:06:07 lr 0.000965 time 0.2921 (0.3192) loss 4.3148 (4.1630) grad_norm 1.0588 (1.0966) [2021-04-15 17:16:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][110/1251] eta 0:06:00 lr 0.000965 time 0.2837 (0.3162) loss 4.1426 (4.1646) grad_norm 1.0738 (1.0907) [2021-04-15 17:16:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][120/1251] eta 0:05:54 lr 0.000965 time 0.2753 (0.3133) loss 3.8681 (4.1682) grad_norm 1.0838 (1.0816) [2021-04-15 17:16:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][130/1251] eta 0:05:48 lr 0.000965 time 0.2904 (0.3109) loss 3.1920 (4.1733) grad_norm 1.2056 (1.0826) [2021-04-15 17:16:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][140/1251] eta 0:05:42 lr 0.000965 time 0.2441 (0.3085) loss 3.0786 (4.1574) grad_norm 1.2063 (1.0843) [2021-04-15 17:17:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][150/1251] eta 0:05:37 lr 0.000965 time 0.2789 (0.3064) loss 4.3125 (4.1788) grad_norm 1.1253 (1.0879) [2021-04-15 17:17:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][160/1251] eta 0:05:32 lr 0.000965 time 0.3118 (0.3049) loss 4.7001 (4.1818) grad_norm 1.1663 (1.0901) [2021-04-15 17:17:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][170/1251] eta 0:05:28 lr 0.000965 time 0.3119 (0.3035) loss 4.1160 (4.1760) grad_norm 1.0814 (1.0871) [2021-04-15 17:17:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][180/1251] eta 0:05:23 lr 0.000965 time 0.2925 (0.3022) loss 4.5530 (4.1726) grad_norm 0.9490 (1.0853) [2021-04-15 17:17:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][190/1251] eta 0:05:19 lr 0.000965 time 0.2715 (0.3010) loss 4.6193 (4.1736) grad_norm 1.1280 (1.0845) [2021-04-15 17:17:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][200/1251] eta 0:05:15 lr 0.000965 time 0.2688 (0.2998) loss 4.0880 (4.1593) grad_norm 1.5112 (1.0827) [2021-04-15 17:17:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][210/1251] eta 0:05:10 lr 0.000965 time 0.2746 (0.2986) loss 4.0753 (4.1428) grad_norm 1.0827 (1.0849) [2021-04-15 17:17:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][220/1251] eta 0:05:07 lr 0.000965 time 0.2762 (0.2979) loss 4.3430 (4.1402) grad_norm 0.9154 (1.0883) [2021-04-15 17:17:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][230/1251] eta 0:05:03 lr 0.000965 time 0.2949 (0.2972) loss 4.2114 (4.1361) grad_norm 1.1249 (1.0885) [2021-04-15 17:17:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][240/1251] eta 0:04:59 lr 0.000965 time 0.2740 (0.2967) loss 3.9140 (4.1343) grad_norm 1.1869 (1.0920) [2021-04-15 17:17:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][250/1251] eta 0:04:56 lr 0.000965 time 0.2683 (0.2962) loss 3.9502 (4.1345) grad_norm 1.0948 (1.0914) [2021-04-15 17:17:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][260/1251] eta 0:04:52 lr 0.000965 time 0.2835 (0.2955) loss 3.7798 (4.1314) grad_norm 1.2241 (1.0912) [2021-04-15 17:17:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][270/1251] eta 0:04:49 lr 0.000965 time 0.2689 (0.2950) loss 4.4651 (4.1457) grad_norm 1.0971 (1.0902) [2021-04-15 17:17:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][280/1251] eta 0:04:46 lr 0.000965 time 0.2770 (0.2946) loss 4.6302 (4.1521) grad_norm 1.0646 (1.0936) [2021-04-15 17:17:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][290/1251] eta 0:04:42 lr 0.000965 time 0.2626 (0.2940) loss 3.7051 (4.1470) grad_norm 0.8624 (1.0945) [2021-04-15 17:17:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][300/1251] eta 0:04:38 lr 0.000965 time 0.2986 (0.2934) loss 4.2299 (4.1486) grad_norm 0.8953 (1.0931) [2021-04-15 17:17:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][310/1251] eta 0:04:35 lr 0.000965 time 0.2677 (0.2931) loss 4.4705 (4.1474) grad_norm 1.0162 (1.0919) [2021-04-15 17:17:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][320/1251] eta 0:04:32 lr 0.000965 time 0.2613 (0.2927) loss 4.4138 (4.1446) grad_norm 1.0027 (1.0919) [2021-04-15 17:17:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][330/1251] eta 0:04:29 lr 0.000965 time 0.2866 (0.2922) loss 3.6055 (4.1391) grad_norm 1.0532 (1.0925) [2021-04-15 17:17:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][340/1251] eta 0:04:25 lr 0.000965 time 0.2759 (0.2918) loss 4.3738 (4.1312) grad_norm 1.0800 (1.0958) [2021-04-15 17:17:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][350/1251] eta 0:04:22 lr 0.000965 time 0.2651 (0.2915) loss 3.4360 (4.1264) grad_norm 1.1017 (1.0941) [2021-04-15 17:18:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][360/1251] eta 0:04:19 lr 0.000965 time 0.2585 (0.2915) loss 4.2423 (4.1234) grad_norm 1.2699 (1.0960) [2021-04-15 17:18:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][370/1251] eta 0:04:16 lr 0.000965 time 0.2632 (0.2913) loss 3.7390 (4.1099) grad_norm 1.4257 (1.0989) [2021-04-15 17:18:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][380/1251] eta 0:04:13 lr 0.000965 time 0.2621 (0.2910) loss 4.6763 (4.1148) grad_norm 1.4338 (1.1006) [2021-04-15 17:18:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][390/1251] eta 0:04:10 lr 0.000965 time 0.2990 (0.2907) loss 5.1667 (4.1247) grad_norm 1.2961 (1.1001) [2021-04-15 17:18:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][400/1251] eta 0:04:07 lr 0.000965 time 0.2921 (0.2904) loss 3.5238 (4.1260) grad_norm 1.0511 (1.1015) [2021-04-15 17:18:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][410/1251] eta 0:04:04 lr 0.000965 time 0.3030 (0.2902) loss 4.0471 (4.1306) grad_norm 1.1242 (1.0997) [2021-04-15 17:18:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][420/1251] eta 0:04:01 lr 0.000965 time 0.3088 (0.2904) loss 4.3485 (4.1279) grad_norm 1.0373 (1.1001) [2021-04-15 17:18:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][430/1251] eta 0:03:58 lr 0.000965 time 0.2685 (0.2906) loss 3.4364 (4.1272) grad_norm 0.8719 (1.0995) [2021-04-15 17:18:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][440/1251] eta 0:03:55 lr 0.000965 time 0.2739 (0.2903) loss 3.8969 (4.1297) grad_norm 1.2084 (1.0980) [2021-04-15 17:18:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][450/1251] eta 0:03:52 lr 0.000965 time 0.2793 (0.2900) loss 4.8253 (4.1318) grad_norm 1.0238 (1.0952) [2021-04-15 17:18:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][460/1251] eta 0:03:49 lr 0.000965 time 0.2925 (0.2899) loss 3.6715 (4.1327) grad_norm 1.0763 (1.0951) [2021-04-15 17:18:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][470/1251] eta 0:03:46 lr 0.000965 time 0.3043 (0.2896) loss 3.4671 (4.1289) grad_norm 1.0234 (1.0933) [2021-04-15 17:18:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][480/1251] eta 0:03:43 lr 0.000965 time 0.2795 (0.2893) loss 4.6449 (4.1268) grad_norm 1.0671 (1.0925) [2021-04-15 17:18:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][490/1251] eta 0:03:39 lr 0.000964 time 0.2799 (0.2890) loss 3.6433 (4.1203) grad_norm 1.2697 (1.0935) [2021-04-15 17:18:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][500/1251] eta 0:03:36 lr 0.000964 time 0.2905 (0.2889) loss 4.6064 (4.1220) grad_norm 1.1413 (1.0937) [2021-04-15 17:18:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][510/1251] eta 0:03:33 lr 0.000964 time 0.2790 (0.2888) loss 3.9627 (4.1208) grad_norm 1.0521 (1.0943) [2021-04-15 17:18:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][520/1251] eta 0:03:30 lr 0.000964 time 0.2838 (0.2886) loss 3.1180 (4.1230) grad_norm 0.9832 (1.0944) [2021-04-15 17:18:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][530/1251] eta 0:03:27 lr 0.000964 time 0.2976 (0.2884) loss 4.6247 (4.1251) grad_norm 1.2647 (1.0970) [2021-04-15 17:18:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][540/1251] eta 0:03:24 lr 0.000964 time 0.2746 (0.2881) loss 4.6436 (4.1293) grad_norm 1.3365 (1.0972) [2021-04-15 17:18:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][550/1251] eta 0:03:21 lr 0.000964 time 0.2889 (0.2880) loss 3.8998 (4.1281) grad_norm 0.9890 (1.0974) [2021-04-15 17:18:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][560/1251] eta 0:03:18 lr 0.000964 time 0.2790 (0.2879) loss 3.0029 (4.1287) grad_norm 0.8424 (1.0959) [2021-04-15 17:18:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][570/1251] eta 0:03:16 lr 0.000964 time 0.2538 (0.2881) loss 2.9557 (4.1237) grad_norm 1.0181 (1.0959) [2021-04-15 17:19:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][580/1251] eta 0:03:13 lr 0.000964 time 0.3088 (0.2879) loss 4.3741 (4.1252) grad_norm 0.9145 (1.0946) [2021-04-15 17:19:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][590/1251] eta 0:03:10 lr 0.000964 time 0.2865 (0.2878) loss 4.2850 (4.1304) grad_norm 0.9845 (1.0932) [2021-04-15 17:19:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][600/1251] eta 0:03:07 lr 0.000964 time 0.2772 (0.2875) loss 4.9175 (4.1333) grad_norm 1.3244 (1.0922) [2021-04-15 17:19:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][610/1251] eta 0:03:04 lr 0.000964 time 0.2675 (0.2874) loss 4.5035 (4.1358) grad_norm 1.1551 (1.0917) [2021-04-15 17:19:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][620/1251] eta 0:03:01 lr 0.000964 time 0.2701 (0.2873) loss 5.1797 (4.1342) grad_norm 1.1183 (1.0908) [2021-04-15 17:19:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][630/1251] eta 0:02:58 lr 0.000964 time 0.2625 (0.2871) loss 3.7302 (4.1343) grad_norm 0.9101 (1.0909) [2021-04-15 17:19:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][640/1251] eta 0:02:55 lr 0.000964 time 0.2905 (0.2870) loss 3.9705 (4.1335) grad_norm 0.9945 (1.0906) [2021-04-15 17:19:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][650/1251] eta 0:02:52 lr 0.000964 time 0.2796 (0.2869) loss 4.7184 (4.1353) grad_norm 1.0255 (1.0916) [2021-04-15 17:19:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][660/1251] eta 0:02:49 lr 0.000964 time 0.2728 (0.2868) loss 4.3966 (4.1326) grad_norm 1.0250 (1.0910) [2021-04-15 17:19:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][670/1251] eta 0:02:46 lr 0.000964 time 0.2636 (0.2868) loss 3.9887 (4.1320) grad_norm 1.0424 (1.0895) [2021-04-15 17:19:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][680/1251] eta 0:02:43 lr 0.000964 time 0.2745 (0.2869) loss 4.6264 (4.1330) grad_norm 1.1691 (inf) [2021-04-15 17:19:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][690/1251] eta 0:02:40 lr 0.000964 time 0.2748 (0.2867) loss 4.3134 (4.1354) grad_norm 1.0571 (inf) [2021-04-15 17:19:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][700/1251] eta 0:02:37 lr 0.000964 time 0.2700 (0.2866) loss 3.6498 (4.1328) grad_norm 0.9647 (inf) [2021-04-15 17:19:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][710/1251] eta 0:02:35 lr 0.000964 time 0.2689 (0.2865) loss 3.0151 (4.1289) grad_norm 0.9366 (inf) [2021-04-15 17:19:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][720/1251] eta 0:02:32 lr 0.000964 time 0.2745 (0.2864) loss 4.6881 (4.1286) grad_norm 1.0398 (inf) [2021-04-15 17:19:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][730/1251] eta 0:02:29 lr 0.000964 time 0.2824 (0.2863) loss 3.6943 (4.1297) grad_norm 1.1670 (inf) [2021-04-15 17:19:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][740/1251] eta 0:02:26 lr 0.000964 time 0.2658 (0.2862) loss 3.8712 (4.1282) grad_norm 1.1395 (inf) [2021-04-15 17:19:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][750/1251] eta 0:02:23 lr 0.000964 time 0.2778 (0.2861) loss 4.5609 (4.1266) grad_norm 1.1026 (inf) [2021-04-15 17:19:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][760/1251] eta 0:02:20 lr 0.000964 time 0.3010 (0.2860) loss 3.3610 (4.1264) grad_norm 1.1023 (inf) [2021-04-15 17:19:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][770/1251] eta 0:02:17 lr 0.000964 time 0.2943 (0.2859) loss 4.5210 (4.1264) grad_norm 1.0839 (inf) [2021-04-15 17:19:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][780/1251] eta 0:02:14 lr 0.000964 time 0.2747 (0.2858) loss 3.6852 (4.1294) grad_norm 1.0747 (inf) [2021-04-15 17:20:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][790/1251] eta 0:02:11 lr 0.000964 time 0.2502 (0.2857) loss 3.8724 (4.1260) grad_norm 1.1529 (inf) [2021-04-15 17:20:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][800/1251] eta 0:02:08 lr 0.000964 time 0.2762 (0.2856) loss 3.5067 (4.1262) grad_norm 1.1048 (inf) [2021-04-15 17:20:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][810/1251] eta 0:02:05 lr 0.000964 time 0.2801 (0.2855) loss 4.0968 (4.1257) grad_norm 0.9606 (inf) [2021-04-15 17:20:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][820/1251] eta 0:02:03 lr 0.000964 time 0.2896 (0.2855) loss 4.3636 (4.1293) grad_norm 1.1963 (inf) [2021-04-15 17:20:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][830/1251] eta 0:02:00 lr 0.000964 time 0.3230 (0.2855) loss 3.8936 (4.1287) grad_norm 0.9663 (inf) [2021-04-15 17:20:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][840/1251] eta 0:01:57 lr 0.000964 time 0.2720 (0.2855) loss 4.1683 (4.1307) grad_norm 0.9076 (inf) [2021-04-15 17:20:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][850/1251] eta 0:01:54 lr 0.000964 time 0.2540 (0.2853) loss 3.9374 (4.1320) grad_norm 1.0580 (inf) [2021-04-15 17:20:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][860/1251] eta 0:01:51 lr 0.000964 time 0.2851 (0.2853) loss 3.1499 (4.1308) grad_norm 0.9577 (inf) [2021-04-15 17:20:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][870/1251] eta 0:01:48 lr 0.000964 time 0.2927 (0.2853) loss 4.5657 (4.1318) grad_norm 1.0579 (inf) [2021-04-15 17:20:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][880/1251] eta 0:01:45 lr 0.000964 time 0.2817 (0.2852) loss 5.0828 (4.1325) grad_norm 1.1770 (inf) [2021-04-15 17:20:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][890/1251] eta 0:01:42 lr 0.000964 time 0.2751 (0.2852) loss 5.2073 (4.1351) grad_norm 1.1329 (inf) [2021-04-15 17:20:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][900/1251] eta 0:01:40 lr 0.000964 time 0.3135 (0.2851) loss 3.7492 (4.1330) grad_norm 0.8848 (inf) [2021-04-15 17:20:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][910/1251] eta 0:01:37 lr 0.000964 time 0.3084 (0.2851) loss 4.6566 (4.1340) grad_norm 1.1753 (inf) [2021-04-15 17:20:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][920/1251] eta 0:01:34 lr 0.000964 time 0.2717 (0.2850) loss 4.7533 (4.1321) grad_norm 0.9768 (inf) [2021-04-15 17:20:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][930/1251] eta 0:01:31 lr 0.000964 time 0.2641 (0.2850) loss 4.5632 (4.1349) grad_norm 1.3643 (inf) [2021-04-15 17:20:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][940/1251] eta 0:01:28 lr 0.000964 time 0.3026 (0.2850) loss 3.3275 (4.1308) grad_norm 1.0164 (inf) [2021-04-15 17:20:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][950/1251] eta 0:01:25 lr 0.000964 time 0.3089 (0.2850) loss 2.8626 (4.1297) grad_norm 0.9993 (inf) [2021-04-15 17:20:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][960/1251] eta 0:01:22 lr 0.000964 time 0.2974 (0.2849) loss 4.3848 (4.1255) grad_norm 1.2059 (inf) [2021-04-15 17:20:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][970/1251] eta 0:01:20 lr 0.000964 time 0.2605 (0.2848) loss 3.2626 (4.1247) grad_norm 1.0814 (inf) [2021-04-15 17:20:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][980/1251] eta 0:01:17 lr 0.000964 time 0.2898 (0.2847) loss 4.2183 (4.1235) grad_norm 1.2359 (inf) [2021-04-15 17:20:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][990/1251] eta 0:01:14 lr 0.000964 time 0.2825 (0.2847) loss 3.9686 (4.1263) grad_norm 1.1459 (inf) [2021-04-15 17:21:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][1000/1251] eta 0:01:11 lr 0.000964 time 0.3905 (0.2848) loss 4.8638 (4.1278) grad_norm 1.0389 (inf) [2021-04-15 17:21:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][1010/1251] eta 0:01:08 lr 0.000964 time 0.2695 (0.2849) loss 4.0341 (4.1266) grad_norm 0.8905 (inf) [2021-04-15 17:21:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][1020/1251] eta 0:01:05 lr 0.000964 time 0.2687 (0.2848) loss 4.2797 (4.1280) grad_norm 1.0343 (inf) [2021-04-15 17:21:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][1030/1251] eta 0:01:02 lr 0.000964 time 0.2748 (0.2847) loss 4.1142 (4.1260) grad_norm 1.0458 (inf) [2021-04-15 17:21:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][1040/1251] eta 0:01:00 lr 0.000964 time 0.2710 (0.2850) loss 3.8113 (4.1238) grad_norm 1.1932 (inf) [2021-04-15 17:21:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][1050/1251] eta 0:00:57 lr 0.000964 time 0.2860 (0.2849) loss 5.1115 (4.1254) grad_norm 0.9982 (inf) [2021-04-15 17:21:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][1060/1251] eta 0:00:54 lr 0.000964 time 0.2620 (0.2849) loss 4.2590 (4.1257) grad_norm 1.3486 (inf) [2021-04-15 17:21:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][1070/1251] eta 0:00:51 lr 0.000964 time 0.2811 (0.2848) loss 3.4596 (4.1221) grad_norm 0.9888 (inf) [2021-04-15 17:21:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][1080/1251] eta 0:00:48 lr 0.000964 time 0.2773 (0.2847) loss 2.7409 (4.1226) grad_norm 1.0772 (inf) [2021-04-15 17:21:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][1090/1251] eta 0:00:45 lr 0.000964 time 0.2604 (0.2847) loss 3.9852 (4.1230) grad_norm 1.0593 (inf) [2021-04-15 17:21:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][1100/1251] eta 0:00:42 lr 0.000964 time 0.2619 (0.2846) loss 4.8575 (4.1205) grad_norm 0.9986 (inf) [2021-04-15 17:21:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][1110/1251] eta 0:00:40 lr 0.000964 time 0.2726 (0.2845) loss 3.6838 (4.1211) grad_norm 1.0258 (inf) [2021-04-15 17:21:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][1120/1251] eta 0:00:37 lr 0.000964 time 0.2764 (0.2844) loss 3.7742 (4.1234) grad_norm 0.9647 (inf) [2021-04-15 17:21:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][1130/1251] eta 0:00:34 lr 0.000963 time 0.2826 (0.2843) loss 4.6707 (4.1242) grad_norm 1.2689 (inf) [2021-04-15 17:21:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][1140/1251] eta 0:00:31 lr 0.000963 time 0.2613 (0.2843) loss 4.2267 (4.1231) grad_norm 1.1758 (inf) [2021-04-15 17:21:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][1150/1251] eta 0:00:28 lr 0.000963 time 0.2586 (0.2843) loss 3.2010 (4.1196) grad_norm 1.1672 (inf) [2021-04-15 17:21:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][1160/1251] eta 0:00:25 lr 0.000963 time 0.2950 (0.2842) loss 4.5806 (4.1186) grad_norm 1.1447 (inf) [2021-04-15 17:21:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][1170/1251] eta 0:00:23 lr 0.000963 time 0.2826 (0.2842) loss 4.3096 (4.1193) grad_norm 0.9106 (inf) [2021-04-15 17:21:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][1180/1251] eta 0:00:20 lr 0.000963 time 0.2565 (0.2841) loss 4.1184 (4.1184) grad_norm 0.9499 (inf) [2021-04-15 17:21:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][1190/1251] eta 0:00:17 lr 0.000963 time 0.2762 (0.2841) loss 3.8589 (4.1164) grad_norm 1.4756 (inf) [2021-04-15 17:21:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][1200/1251] eta 0:00:14 lr 0.000963 time 0.2723 (0.2840) loss 4.0190 (4.1137) grad_norm 1.1866 (inf) [2021-04-15 17:21:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][1210/1251] eta 0:00:11 lr 0.000963 time 0.2660 (0.2840) loss 4.5831 (4.1137) grad_norm 1.0454 (inf) [2021-04-15 17:22:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][1220/1251] eta 0:00:08 lr 0.000963 time 0.2868 (0.2839) loss 4.8854 (4.1142) grad_norm 0.9389 (inf) [2021-04-15 17:22:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][1230/1251] eta 0:00:05 lr 0.000963 time 0.2838 (0.2839) loss 3.2600 (4.1139) grad_norm 1.1237 (inf) [2021-04-15 17:22:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][1240/1251] eta 0:00:03 lr 0.000963 time 0.4033 (0.2839) loss 4.2443 (4.1121) grad_norm 0.9682 (inf) [2021-04-15 17:22:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [36/300][1250/1251] eta 0:00:00 lr 0.000963 time 0.2496 (0.2836) loss 3.6809 (4.1103) grad_norm 1.1013 (inf) [2021-04-15 17:22:12 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 36 training takes 0:05:56 [2021-04-15 17:22:12 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_36.pth saving...... [2021-04-15 17:22:28 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_36.pth saved !!! [2021-04-15 17:22:30 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.352 (1.352) Loss 1.3441 (1.3441) Acc@1 69.336 (69.336) Acc@5 89.062 (89.062) [2021-04-15 17:22:31 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.098 (0.269) Loss 1.4980 (1.3963) Acc@1 65.918 (68.306) Acc@5 87.695 (88.494) [2021-04-15 17:22:34 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.107 (0.247) Loss 1.4173 (1.4095) Acc@1 67.676 (67.862) Acc@5 88.281 (88.421) [2021-04-15 17:22:35 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.124 (0.226) Loss 1.3916 (1.4010) Acc@1 69.043 (68.041) Acc@5 88.379 (88.540) [2021-04-15 17:22:38 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.257 (0.221) Loss 1.3705 (1.3977) Acc@1 68.652 (68.095) Acc@5 89.746 (88.631) [2021-04-15 17:22:41 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 68.114 Acc@5 88.678 [2021-04-15 17:22:41 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 68.1% [2021-04-15 17:22:41 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 68.11% [2021-04-15 17:22:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][0/1251] eta 1:02:32 lr 0.000963 time 2.9994 (2.9994) loss 4.7881 (4.7881) grad_norm 0.9518 (0.9518) [2021-04-15 17:22:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][10/1251] eta 0:10:51 lr 0.000963 time 0.2524 (0.5247) loss 4.8087 (4.2609) grad_norm 0.9902 (1.0145) [2021-04-15 17:22:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][20/1251] eta 0:08:24 lr 0.000963 time 0.3004 (0.4096) loss 3.1799 (4.3564) grad_norm 1.0658 (1.0580) [2021-04-15 17:22:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][30/1251] eta 0:07:29 lr 0.000963 time 0.2857 (0.3685) loss 3.4863 (4.2615) grad_norm 1.1570 (1.0911) [2021-04-15 17:22:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][40/1251] eta 0:07:00 lr 0.000963 time 0.2734 (0.3469) loss 4.3447 (4.2136) grad_norm 1.2740 (1.1398) [2021-04-15 17:22:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][50/1251] eta 0:06:42 lr 0.000963 time 0.2905 (0.3347) loss 4.1021 (4.1617) grad_norm 1.1170 (1.1374) [2021-04-15 17:23:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][60/1251] eta 0:06:28 lr 0.000963 time 0.2732 (0.3262) loss 4.4758 (4.1516) grad_norm 1.0564 (1.1284) [2021-04-15 17:23:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][70/1251] eta 0:06:17 lr 0.000963 time 0.2624 (0.3194) loss 3.4820 (4.1306) grad_norm 0.9502 (1.1128) [2021-04-15 17:23:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][80/1251] eta 0:06:09 lr 0.000963 time 0.3076 (0.3152) loss 4.6872 (4.1179) grad_norm 1.0098 (1.0987) [2021-04-15 17:23:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][90/1251] eta 0:06:00 lr 0.000963 time 0.2653 (0.3108) loss 4.5265 (4.1102) grad_norm 1.1340 (1.0962) [2021-04-15 17:23:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][100/1251] eta 0:05:54 lr 0.000963 time 0.2753 (0.3077) loss 4.2140 (4.0856) grad_norm 1.0299 (1.0923) [2021-04-15 17:23:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][110/1251] eta 0:05:48 lr 0.000963 time 0.2932 (0.3056) loss 3.5205 (4.0964) grad_norm 1.0068 (1.0944) [2021-04-15 17:23:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][120/1251] eta 0:05:42 lr 0.000963 time 0.2695 (0.3030) loss 3.6192 (4.0849) grad_norm 1.0657 (1.0893) [2021-04-15 17:23:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][130/1251] eta 0:05:37 lr 0.000963 time 0.2726 (0.3012) loss 4.6880 (4.0734) grad_norm 1.3068 (1.0923) [2021-04-15 17:23:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][140/1251] eta 0:05:34 lr 0.000963 time 0.2712 (0.3008) loss 3.9709 (4.0862) grad_norm 1.2948 (1.0984) [2021-04-15 17:23:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][150/1251] eta 0:05:29 lr 0.000963 time 0.2813 (0.2993) loss 3.1171 (4.0837) grad_norm 0.9394 (1.0994) [2021-04-15 17:23:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][160/1251] eta 0:05:24 lr 0.000963 time 0.2602 (0.2977) loss 3.0385 (4.0599) grad_norm 1.0487 (1.0936) [2021-04-15 17:23:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][170/1251] eta 0:05:20 lr 0.000963 time 0.2936 (0.2968) loss 3.0034 (4.0637) grad_norm 0.9862 (1.0939) [2021-04-15 17:23:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][180/1251] eta 0:05:17 lr 0.000963 time 0.2878 (0.2962) loss 4.1469 (4.0706) grad_norm 1.0949 (1.0900) [2021-04-15 17:23:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][190/1251] eta 0:05:13 lr 0.000963 time 0.2572 (0.2952) loss 2.8429 (4.0664) grad_norm 1.0875 (1.0918) [2021-04-15 17:23:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][200/1251] eta 0:05:10 lr 0.000963 time 0.2714 (0.2950) loss 4.0445 (4.0694) grad_norm 1.0666 (1.0957) [2021-04-15 17:23:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][210/1251] eta 0:05:06 lr 0.000963 time 0.3832 (0.2945) loss 3.4887 (4.0731) grad_norm 0.9484 (1.0975) [2021-04-15 17:23:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][220/1251] eta 0:05:03 lr 0.000963 time 0.2547 (0.2940) loss 5.0341 (4.0678) grad_norm 1.3064 (1.0969) [2021-04-15 17:23:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][230/1251] eta 0:04:59 lr 0.000963 time 0.2803 (0.2933) loss 2.6454 (4.0625) grad_norm 1.1709 (1.0987) [2021-04-15 17:23:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][240/1251] eta 0:04:55 lr 0.000963 time 0.2781 (0.2928) loss 4.4996 (4.0646) grad_norm 0.8620 (1.0955) [2021-04-15 17:23:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][250/1251] eta 0:04:52 lr 0.000963 time 0.2686 (0.2924) loss 3.9654 (4.0619) grad_norm 1.0117 (1.0962) [2021-04-15 17:23:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][260/1251] eta 0:04:49 lr 0.000963 time 0.2477 (0.2919) loss 4.0917 (4.0572) grad_norm 1.0012 (1.0933) [2021-04-15 17:24:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][270/1251] eta 0:04:45 lr 0.000963 time 0.3073 (0.2914) loss 3.8830 (4.0564) grad_norm 1.2313 (1.0909) [2021-04-15 17:24:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][280/1251] eta 0:04:42 lr 0.000963 time 0.2449 (0.2908) loss 4.5504 (4.0645) grad_norm 0.9732 (1.0891) [2021-04-15 17:24:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][290/1251] eta 0:04:39 lr 0.000963 time 0.3018 (0.2905) loss 4.5385 (4.0678) grad_norm 0.9075 (1.0873) [2021-04-15 17:24:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][300/1251] eta 0:04:35 lr 0.000963 time 0.2678 (0.2901) loss 3.8393 (4.0678) grad_norm 1.1568 (1.0884) [2021-04-15 17:24:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][310/1251] eta 0:04:32 lr 0.000963 time 0.2846 (0.2899) loss 2.6203 (4.0657) grad_norm 1.1527 (1.0888) [2021-04-15 17:24:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][320/1251] eta 0:04:29 lr 0.000963 time 0.2700 (0.2895) loss 4.8826 (4.0703) grad_norm 1.1685 (1.0905) [2021-04-15 17:24:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][330/1251] eta 0:04:26 lr 0.000963 time 0.2766 (0.2891) loss 4.7772 (4.0812) grad_norm 1.0731 (1.0901) [2021-04-15 17:24:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][340/1251] eta 0:04:23 lr 0.000963 time 0.2825 (0.2888) loss 3.4293 (4.0843) grad_norm 1.0528 (1.0906) [2021-04-15 17:24:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][350/1251] eta 0:04:19 lr 0.000963 time 0.2869 (0.2885) loss 3.1968 (4.0905) grad_norm 1.2240 (1.0898) [2021-04-15 17:24:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][360/1251] eta 0:04:17 lr 0.000963 time 0.2651 (0.2886) loss 3.2399 (4.0912) grad_norm 1.1699 (1.0924) [2021-04-15 17:24:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][370/1251] eta 0:04:13 lr 0.000963 time 0.2804 (0.2883) loss 2.6011 (4.0915) grad_norm 0.9321 (1.0919) [2021-04-15 17:24:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][380/1251] eta 0:04:10 lr 0.000963 time 0.2896 (0.2880) loss 4.0790 (4.0855) grad_norm 0.9657 (1.0917) [2021-04-15 17:24:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][390/1251] eta 0:04:07 lr 0.000963 time 0.2984 (0.2878) loss 3.9409 (4.0778) grad_norm 1.1589 (1.0917) [2021-04-15 17:24:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][400/1251] eta 0:04:04 lr 0.000963 time 0.2440 (0.2875) loss 4.2969 (4.0786) grad_norm 0.9090 (1.0896) [2021-04-15 17:24:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][410/1251] eta 0:04:01 lr 0.000963 time 0.2635 (0.2872) loss 5.0020 (4.0811) grad_norm 1.0071 (1.0887) [2021-04-15 17:24:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][420/1251] eta 0:03:58 lr 0.000963 time 0.3032 (0.2870) loss 5.0738 (4.0749) grad_norm 0.9613 (1.0892) [2021-04-15 17:24:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][430/1251] eta 0:03:55 lr 0.000963 time 0.2882 (0.2867) loss 3.7634 (4.0797) grad_norm 0.9156 (1.0903) [2021-04-15 17:24:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][440/1251] eta 0:03:52 lr 0.000963 time 0.2890 (0.2866) loss 4.1987 (4.0790) grad_norm 1.1268 (1.0905) [2021-04-15 17:24:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][450/1251] eta 0:03:49 lr 0.000963 time 0.4659 (0.2868) loss 3.9553 (4.0821) grad_norm 1.2088 (1.0911) [2021-04-15 17:24:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][460/1251] eta 0:03:46 lr 0.000963 time 0.2447 (0.2865) loss 3.1254 (4.0785) grad_norm 1.1883 (1.0909) [2021-04-15 17:24:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][470/1251] eta 0:03:43 lr 0.000963 time 0.2837 (0.2868) loss 4.2803 (4.0787) grad_norm 1.0959 (1.0902) [2021-04-15 17:24:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][480/1251] eta 0:03:41 lr 0.000963 time 0.2722 (0.2867) loss 4.6061 (4.0828) grad_norm 1.3903 (1.0899) [2021-04-15 17:25:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][490/1251] eta 0:03:38 lr 0.000963 time 0.2783 (0.2865) loss 4.7230 (4.0816) grad_norm 1.3496 (1.0912) [2021-04-15 17:25:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][500/1251] eta 0:03:34 lr 0.000963 time 0.2746 (0.2863) loss 4.4697 (4.0806) grad_norm 1.1125 (1.0925) [2021-04-15 17:25:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][510/1251] eta 0:03:31 lr 0.000963 time 0.2663 (0.2860) loss 3.4461 (4.0766) grad_norm 1.1916 (1.0934) [2021-04-15 17:25:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][520/1251] eta 0:03:28 lr 0.000962 time 0.2870 (0.2859) loss 3.7587 (4.0778) grad_norm 1.0646 (1.0942) [2021-04-15 17:25:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][530/1251] eta 0:03:26 lr 0.000962 time 0.2896 (0.2858) loss 4.8330 (4.0763) grad_norm 1.0445 (1.0946) [2021-04-15 17:25:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][540/1251] eta 0:03:23 lr 0.000962 time 0.3142 (0.2858) loss 4.3646 (4.0721) grad_norm 1.0298 (1.0939) [2021-04-15 17:25:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][550/1251] eta 0:03:20 lr 0.000962 time 0.2928 (0.2856) loss 3.1931 (4.0671) grad_norm 1.0761 (1.0942) [2021-04-15 17:25:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][560/1251] eta 0:03:17 lr 0.000962 time 0.2745 (0.2854) loss 4.4888 (4.0721) grad_norm 1.0164 (1.0937) [2021-04-15 17:25:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][570/1251] eta 0:03:14 lr 0.000962 time 0.2823 (0.2853) loss 3.6878 (4.0781) grad_norm 1.0402 (1.0930) [2021-04-15 17:25:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][580/1251] eta 0:03:11 lr 0.000962 time 0.3107 (0.2853) loss 3.3245 (4.0683) grad_norm 1.0695 (1.0925) [2021-04-15 17:25:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][590/1251] eta 0:03:08 lr 0.000962 time 0.2754 (0.2852) loss 3.8430 (4.0707) grad_norm 1.7156 (1.0942) [2021-04-15 17:25:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][600/1251] eta 0:03:05 lr 0.000962 time 0.2993 (0.2852) loss 3.9339 (4.0680) grad_norm 0.9314 (1.0952) [2021-04-15 17:25:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][610/1251] eta 0:03:02 lr 0.000962 time 0.2842 (0.2851) loss 3.3318 (4.0647) grad_norm 1.0789 (1.0965) [2021-04-15 17:25:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][620/1251] eta 0:02:59 lr 0.000962 time 0.2635 (0.2850) loss 4.0039 (4.0634) grad_norm 0.9543 (1.0965) [2021-04-15 17:25:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][630/1251] eta 0:02:56 lr 0.000962 time 0.2838 (0.2849) loss 4.6743 (4.0700) grad_norm 1.1457 (1.0961) [2021-04-15 17:25:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][640/1251] eta 0:02:54 lr 0.000962 time 0.2679 (0.2849) loss 3.4967 (4.0674) grad_norm 1.0700 (1.0951) [2021-04-15 17:25:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][650/1251] eta 0:02:51 lr 0.000962 time 0.2976 (0.2848) loss 4.7568 (4.0708) grad_norm 1.1002 (1.0936) [2021-04-15 17:25:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][660/1251] eta 0:02:48 lr 0.000962 time 0.2972 (0.2847) loss 4.0679 (4.0706) grad_norm 1.2322 (1.0939) [2021-04-15 17:25:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][670/1251] eta 0:02:45 lr 0.000962 time 0.2729 (0.2846) loss 3.1943 (4.0681) grad_norm 1.0346 (1.0931) [2021-04-15 17:25:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][680/1251] eta 0:02:42 lr 0.000962 time 0.2854 (0.2845) loss 4.6255 (4.0706) grad_norm 1.0816 (1.0929) [2021-04-15 17:25:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][690/1251] eta 0:02:39 lr 0.000962 time 0.2754 (0.2844) loss 3.7156 (4.0690) grad_norm 1.0068 (1.0925) [2021-04-15 17:26:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][700/1251] eta 0:02:36 lr 0.000962 time 0.2892 (0.2844) loss 4.1930 (4.0683) grad_norm 1.0015 (1.0924) [2021-04-15 17:26:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][710/1251] eta 0:02:33 lr 0.000962 time 0.2880 (0.2843) loss 4.2321 (4.0739) grad_norm 1.1880 (1.0916) [2021-04-15 17:26:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][720/1251] eta 0:02:30 lr 0.000962 time 0.2657 (0.2843) loss 3.2356 (4.0684) grad_norm 0.9281 (1.0916) [2021-04-15 17:26:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][730/1251] eta 0:02:28 lr 0.000962 time 0.2802 (0.2842) loss 3.9562 (4.0697) grad_norm 0.9533 (1.0908) [2021-04-15 17:26:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][740/1251] eta 0:02:25 lr 0.000962 time 0.2831 (0.2842) loss 4.1793 (4.0728) grad_norm 1.0200 (1.0902) [2021-04-15 17:26:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][750/1251] eta 0:02:22 lr 0.000962 time 0.2623 (0.2840) loss 3.1282 (4.0736) grad_norm 1.0543 (1.0907) [2021-04-15 17:26:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][760/1251] eta 0:02:19 lr 0.000962 time 0.2749 (0.2840) loss 2.8505 (4.0718) grad_norm 1.1674 (1.0902) [2021-04-15 17:26:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][770/1251] eta 0:02:16 lr 0.000962 time 0.2634 (0.2839) loss 3.7213 (4.0696) grad_norm 1.0193 (1.0907) [2021-04-15 17:26:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][780/1251] eta 0:02:13 lr 0.000962 time 0.2799 (0.2840) loss 3.9589 (4.0681) grad_norm 1.0442 (1.0918) [2021-04-15 17:26:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][790/1251] eta 0:02:10 lr 0.000962 time 0.2751 (0.2841) loss 4.3839 (4.0712) grad_norm 1.2264 (1.0916) [2021-04-15 17:26:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][800/1251] eta 0:02:08 lr 0.000962 time 0.3235 (0.2841) loss 4.3420 (4.0711) grad_norm 1.0387 (1.0915) [2021-04-15 17:26:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][810/1251] eta 0:02:05 lr 0.000962 time 0.2898 (0.2841) loss 4.0183 (4.0686) grad_norm 0.9750 (1.0914) [2021-04-15 17:26:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][820/1251] eta 0:02:02 lr 0.000962 time 0.2484 (0.2840) loss 3.2695 (4.0674) grad_norm 1.0310 (1.0907) [2021-04-15 17:26:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][830/1251] eta 0:01:59 lr 0.000962 time 0.2835 (0.2839) loss 4.2468 (4.0621) grad_norm 1.1792 (1.0913) [2021-04-15 17:26:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][840/1251] eta 0:01:56 lr 0.000962 time 0.2841 (0.2839) loss 4.1866 (4.0646) grad_norm 0.9803 (1.0915) [2021-04-15 17:26:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][850/1251] eta 0:01:53 lr 0.000962 time 0.2642 (0.2837) loss 3.8243 (4.0639) grad_norm 0.9779 (1.0903) [2021-04-15 17:26:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][860/1251] eta 0:01:50 lr 0.000962 time 0.2648 (0.2837) loss 4.2238 (4.0624) grad_norm 1.2616 (1.0903) [2021-04-15 17:26:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][870/1251] eta 0:01:48 lr 0.000962 time 0.2879 (0.2836) loss 3.8142 (4.0594) grad_norm 1.1575 (1.0911) [2021-04-15 17:26:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][880/1251] eta 0:01:45 lr 0.000962 time 0.2993 (0.2836) loss 4.8151 (4.0607) grad_norm 1.0455 (1.0918) [2021-04-15 17:26:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][890/1251] eta 0:01:42 lr 0.000962 time 0.2723 (0.2836) loss 4.2504 (4.0594) grad_norm 1.2280 (1.0934) [2021-04-15 17:26:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][900/1251] eta 0:01:39 lr 0.000962 time 0.2793 (0.2835) loss 4.1976 (4.0592) grad_norm 0.9651 (1.0933) [2021-04-15 17:26:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][910/1251] eta 0:01:36 lr 0.000962 time 0.2768 (0.2834) loss 3.5012 (4.0599) grad_norm 0.9160 (1.0928) [2021-04-15 17:27:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][920/1251] eta 0:01:33 lr 0.000962 time 0.2829 (0.2834) loss 4.3491 (4.0580) grad_norm 1.0855 (1.0927) [2021-04-15 17:27:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][930/1251] eta 0:01:30 lr 0.000962 time 0.2768 (0.2834) loss 3.8283 (4.0572) grad_norm 0.9715 (1.0925) [2021-04-15 17:27:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][940/1251] eta 0:01:28 lr 0.000962 time 0.2918 (0.2835) loss 4.3440 (4.0584) grad_norm 1.0193 (1.0921) [2021-04-15 17:27:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][950/1251] eta 0:01:25 lr 0.000962 time 0.2889 (0.2835) loss 4.4253 (4.0581) grad_norm 0.8970 (1.0923) [2021-04-15 17:27:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][960/1251] eta 0:01:22 lr 0.000962 time 0.3274 (0.2835) loss 4.4217 (4.0550) grad_norm 1.0788 (1.0922) [2021-04-15 17:27:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][970/1251] eta 0:01:19 lr 0.000962 time 0.2984 (0.2834) loss 3.4735 (4.0542) grad_norm 1.4270 (1.0932) [2021-04-15 17:27:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][980/1251] eta 0:01:16 lr 0.000962 time 0.2852 (0.2834) loss 5.0244 (4.0545) grad_norm 0.9226 (1.0930) [2021-04-15 17:27:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][990/1251] eta 0:01:13 lr 0.000962 time 0.2798 (0.2834) loss 4.1304 (4.0570) grad_norm 0.9470 (1.0932) [2021-04-15 17:27:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][1000/1251] eta 0:01:11 lr 0.000962 time 0.2876 (0.2833) loss 4.6222 (4.0564) grad_norm 1.2079 (1.0930) [2021-04-15 17:27:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][1010/1251] eta 0:01:08 lr 0.000962 time 0.2725 (0.2833) loss 4.1423 (4.0563) grad_norm 1.1863 (1.0937) [2021-04-15 17:27:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][1020/1251] eta 0:01:05 lr 0.000962 time 0.2942 (0.2832) loss 4.7977 (4.0566) grad_norm 1.2723 (1.0938) [2021-04-15 17:27:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][1030/1251] eta 0:01:02 lr 0.000962 time 0.2554 (0.2833) loss 4.6880 (4.0574) grad_norm 0.9061 (1.0944) [2021-04-15 17:27:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][1040/1251] eta 0:00:59 lr 0.000962 time 0.2701 (0.2834) loss 4.0436 (4.0596) grad_norm 0.8663 (1.0930) [2021-04-15 17:27:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][1050/1251] eta 0:00:56 lr 0.000962 time 0.2896 (0.2834) loss 4.0925 (4.0580) grad_norm 1.0447 (1.0924) [2021-04-15 17:27:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][1060/1251] eta 0:00:54 lr 0.000962 time 0.2705 (0.2834) loss 4.1581 (4.0602) grad_norm 1.1177 (1.0928) [2021-04-15 17:27:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][1070/1251] eta 0:00:51 lr 0.000962 time 0.2814 (0.2834) loss 3.9590 (4.0600) grad_norm 1.1682 (1.0934) [2021-04-15 17:27:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][1080/1251] eta 0:00:48 lr 0.000962 time 0.2945 (0.2833) loss 4.5136 (4.0614) grad_norm 0.9496 (1.0933) [2021-04-15 17:27:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][1090/1251] eta 0:00:45 lr 0.000962 time 0.2823 (0.2833) loss 4.6527 (4.0629) grad_norm 1.0183 (1.0927) [2021-04-15 17:27:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][1100/1251] eta 0:00:42 lr 0.000962 time 0.2689 (0.2832) loss 5.0569 (4.0652) grad_norm 0.9645 (1.0925) [2021-04-15 17:27:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][1110/1251] eta 0:00:39 lr 0.000962 time 0.2935 (0.2832) loss 4.4028 (4.0641) grad_norm 1.0282 (1.0926) [2021-04-15 17:27:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][1120/1251] eta 0:00:37 lr 0.000962 time 0.2741 (0.2832) loss 3.0926 (4.0647) grad_norm 1.1170 (1.0921) [2021-04-15 17:28:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][1130/1251] eta 0:00:34 lr 0.000962 time 0.2762 (0.2831) loss 3.1918 (4.0655) grad_norm 1.0093 (1.0918) [2021-04-15 17:28:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][1140/1251] eta 0:00:31 lr 0.000962 time 0.2944 (0.2831) loss 4.0679 (4.0661) grad_norm 0.9184 (1.0912) [2021-04-15 17:28:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][1150/1251] eta 0:00:28 lr 0.000961 time 0.2703 (0.2831) loss 3.5646 (4.0649) grad_norm 1.0446 (1.0905) [2021-04-15 17:28:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][1160/1251] eta 0:00:25 lr 0.000961 time 0.2606 (0.2830) loss 3.9809 (4.0662) grad_norm 1.2586 (1.0909) [2021-04-15 17:28:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][1170/1251] eta 0:00:22 lr 0.000961 time 0.2809 (0.2830) loss 4.6066 (4.0674) grad_norm 1.0052 (1.0912) [2021-04-15 17:28:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][1180/1251] eta 0:00:20 lr 0.000961 time 0.2762 (0.2830) loss 4.6010 (4.0690) grad_norm 1.3033 (1.0914) [2021-04-15 17:28:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][1190/1251] eta 0:00:17 lr 0.000961 time 0.2697 (0.2829) loss 3.6671 (4.0695) grad_norm 0.8699 (1.0915) [2021-04-15 17:28:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][1200/1251] eta 0:00:14 lr 0.000961 time 0.2800 (0.2829) loss 4.2329 (4.0707) grad_norm 1.3731 (1.0914) [2021-04-15 17:28:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][1210/1251] eta 0:00:11 lr 0.000961 time 0.2976 (0.2829) loss 4.9943 (4.0726) grad_norm 1.0592 (1.0920) [2021-04-15 17:28:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][1220/1251] eta 0:00:08 lr 0.000961 time 0.2661 (0.2828) loss 3.3052 (4.0709) grad_norm 1.2390 (1.0915) [2021-04-15 17:28:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][1230/1251] eta 0:00:05 lr 0.000961 time 0.2608 (0.2828) loss 3.9027 (4.0731) grad_norm 1.2035 (1.0912) [2021-04-15 17:28:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][1240/1251] eta 0:00:03 lr 0.000961 time 0.2488 (0.2827) loss 4.3994 (4.0722) grad_norm 1.0617 (1.0911) [2021-04-15 17:28:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [37/300][1250/1251] eta 0:00:00 lr 0.000961 time 0.2486 (0.2824) loss 4.1415 (4.0688) grad_norm 0.8733 (1.0914) [2021-04-15 17:28:37 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 37 training takes 0:05:55 [2021-04-15 17:28:37 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_37.pth saving...... [2021-04-15 17:28:53 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_37.pth saved !!! [2021-04-15 17:28:54 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.168 (1.168) Loss 1.3690 (1.3690) Acc@1 69.922 (69.922) Acc@5 88.965 (88.965) [2021-04-15 17:28:55 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.159 (0.239) Loss 1.3871 (1.4098) Acc@1 68.848 (67.720) Acc@5 88.574 (88.752) [2021-04-15 17:28:58 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.124 (0.236) Loss 1.3885 (1.4138) Acc@1 67.871 (67.680) Acc@5 88.867 (88.737) [2021-04-15 17:29:00 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.128 (0.240) Loss 1.4508 (1.4109) Acc@1 65.039 (67.651) Acc@5 88.574 (88.738) [2021-04-15 17:29:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.220) Loss 1.4896 (1.4067) Acc@1 67.188 (67.766) Acc@5 87.305 (88.815) [2021-04-15 17:29:04 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 67.816 Acc@5 88.756 [2021-04-15 17:29:04 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 67.8% [2021-04-15 17:29:04 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 68.11% [2021-04-15 17:29:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][0/1251] eta 1:20:42 lr 0.000961 time 3.8705 (3.8705) loss 3.2609 (3.2609) grad_norm 1.0655 (1.0655) [2021-04-15 17:29:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][10/1251] eta 0:12:56 lr 0.000961 time 0.2675 (0.6261) loss 3.7417 (3.8866) grad_norm 0.9412 (1.0472) [2021-04-15 17:29:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][20/1251] eta 0:09:26 lr 0.000961 time 0.2786 (0.4600) loss 3.6250 (3.9946) grad_norm 1.4774 (1.0616) [2021-04-15 17:29:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][30/1251] eta 0:08:11 lr 0.000961 time 0.2754 (0.4023) loss 4.2378 (4.0093) grad_norm 0.9279 (1.0727) [2021-04-15 17:29:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][40/1251] eta 0:07:31 lr 0.000961 time 0.2758 (0.3726) loss 3.9160 (3.9675) grad_norm 1.1868 (1.0706) [2021-04-15 17:29:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][50/1251] eta 0:07:05 lr 0.000961 time 0.2810 (0.3542) loss 4.2149 (3.9617) grad_norm 1.0115 (1.0759) [2021-04-15 17:29:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][60/1251] eta 0:06:47 lr 0.000961 time 0.2738 (0.3420) loss 3.1959 (3.9831) grad_norm 1.0432 (1.0899) [2021-04-15 17:29:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][70/1251] eta 0:06:33 lr 0.000961 time 0.2777 (0.3333) loss 4.5191 (3.9795) grad_norm 1.1582 (1.0949) [2021-04-15 17:29:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][80/1251] eta 0:06:22 lr 0.000961 time 0.2732 (0.3270) loss 4.5520 (4.0025) grad_norm 1.0829 (1.0854) [2021-04-15 17:29:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][90/1251] eta 0:06:13 lr 0.000961 time 0.2796 (0.3217) loss 4.4775 (4.0510) grad_norm 1.0840 (1.0825) [2021-04-15 17:29:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][100/1251] eta 0:06:05 lr 0.000961 time 0.2691 (0.3176) loss 3.7500 (4.0747) grad_norm 1.0411 (1.0808) [2021-04-15 17:29:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][110/1251] eta 0:05:58 lr 0.000961 time 0.2626 (0.3143) loss 4.3301 (4.0687) grad_norm 1.0175 (1.0786) [2021-04-15 17:29:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][120/1251] eta 0:05:52 lr 0.000961 time 0.2947 (0.3115) loss 4.1656 (4.0364) grad_norm 1.0409 (1.0825) [2021-04-15 17:29:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][130/1251] eta 0:05:46 lr 0.000961 time 0.2635 (0.3092) loss 3.9963 (4.0296) grad_norm 1.3160 (1.0926) [2021-04-15 17:29:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][140/1251] eta 0:05:41 lr 0.000961 time 0.2915 (0.3073) loss 4.7821 (4.0178) grad_norm 1.3600 (1.0960) [2021-04-15 17:29:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][150/1251] eta 0:05:37 lr 0.000961 time 0.2800 (0.3063) loss 2.6948 (4.0050) grad_norm 1.0239 (1.0983) [2021-04-15 17:29:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][160/1251] eta 0:05:32 lr 0.000961 time 0.2705 (0.3045) loss 3.4571 (4.0024) grad_norm 1.1540 (1.0956) [2021-04-15 17:29:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][170/1251] eta 0:05:27 lr 0.000961 time 0.2877 (0.3031) loss 3.6775 (3.9931) grad_norm 1.3741 (1.0942) [2021-04-15 17:29:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][180/1251] eta 0:05:23 lr 0.000961 time 0.2878 (0.3018) loss 4.0217 (3.9843) grad_norm 1.4576 (1.0989) [2021-04-15 17:30:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][190/1251] eta 0:05:18 lr 0.000961 time 0.2778 (0.3003) loss 4.6200 (4.0031) grad_norm 1.0076 (inf) [2021-04-15 17:30:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][200/1251] eta 0:05:15 lr 0.000961 time 0.2984 (0.2997) loss 3.9014 (4.0116) grad_norm 1.0588 (inf) [2021-04-15 17:30:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][210/1251] eta 0:05:10 lr 0.000961 time 0.2765 (0.2987) loss 3.8597 (4.0161) grad_norm 1.1297 (inf) [2021-04-15 17:30:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][220/1251] eta 0:05:07 lr 0.000961 time 0.2831 (0.2979) loss 4.2688 (4.0124) grad_norm 1.1747 (inf) [2021-04-15 17:30:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][230/1251] eta 0:05:03 lr 0.000961 time 0.2840 (0.2971) loss 4.2512 (4.0092) grad_norm 0.9862 (inf) [2021-04-15 17:30:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][240/1251] eta 0:04:59 lr 0.000961 time 0.2799 (0.2964) loss 3.6420 (4.0124) grad_norm 1.0087 (inf) [2021-04-15 17:30:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][250/1251] eta 0:04:56 lr 0.000961 time 0.2797 (0.2964) loss 4.5358 (4.0183) grad_norm 0.9477 (inf) [2021-04-15 17:30:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][260/1251] eta 0:04:53 lr 0.000961 time 0.2981 (0.2958) loss 3.2948 (4.0247) grad_norm 0.8656 (inf) [2021-04-15 17:30:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][270/1251] eta 0:04:49 lr 0.000961 time 0.2696 (0.2951) loss 4.1876 (4.0242) grad_norm 1.0954 (inf) [2021-04-15 17:30:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][280/1251] eta 0:04:46 lr 0.000961 time 0.2850 (0.2946) loss 4.3667 (4.0249) grad_norm 0.9077 (inf) [2021-04-15 17:30:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][290/1251] eta 0:04:42 lr 0.000961 time 0.2640 (0.2940) loss 4.6090 (4.0319) grad_norm 1.3217 (inf) [2021-04-15 17:30:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][300/1251] eta 0:04:39 lr 0.000961 time 0.2931 (0.2935) loss 3.7125 (4.0221) grad_norm 1.0775 (inf) [2021-04-15 17:30:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][310/1251] eta 0:04:35 lr 0.000961 time 0.3065 (0.2931) loss 4.6300 (4.0309) grad_norm 0.8891 (inf) [2021-04-15 17:30:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][320/1251] eta 0:04:32 lr 0.000961 time 0.2958 (0.2924) loss 4.9644 (4.0305) grad_norm 1.1890 (inf) [2021-04-15 17:30:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][330/1251] eta 0:04:28 lr 0.000961 time 0.2768 (0.2919) loss 4.4379 (4.0298) grad_norm 1.0642 (inf) [2021-04-15 17:30:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][340/1251] eta 0:04:25 lr 0.000961 time 0.2971 (0.2915) loss 3.8814 (4.0300) grad_norm 1.2322 (inf) [2021-04-15 17:30:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][350/1251] eta 0:04:22 lr 0.000961 time 0.2988 (0.2912) loss 3.3645 (4.0236) grad_norm 1.0877 (inf) [2021-04-15 17:30:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][360/1251] eta 0:04:19 lr 0.000961 time 0.2840 (0.2909) loss 4.6093 (4.0214) grad_norm 0.9931 (inf) [2021-04-15 17:30:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][370/1251] eta 0:04:15 lr 0.000961 time 0.2775 (0.2906) loss 4.1691 (4.0317) grad_norm 1.0128 (inf) [2021-04-15 17:30:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][380/1251] eta 0:04:12 lr 0.000961 time 0.2667 (0.2903) loss 4.8406 (4.0365) grad_norm 1.4078 (inf) [2021-04-15 17:30:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][390/1251] eta 0:04:09 lr 0.000961 time 0.2725 (0.2900) loss 3.7336 (4.0379) grad_norm 0.9502 (inf) [2021-04-15 17:31:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][400/1251] eta 0:04:06 lr 0.000961 time 0.2723 (0.2897) loss 5.0554 (4.0375) grad_norm 0.9286 (inf) [2021-04-15 17:31:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][410/1251] eta 0:04:03 lr 0.000961 time 0.2938 (0.2895) loss 5.0041 (4.0366) grad_norm 1.0330 (inf) [2021-04-15 17:31:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][420/1251] eta 0:04:00 lr 0.000961 time 0.2459 (0.2892) loss 3.0659 (4.0323) grad_norm 1.1094 (inf) [2021-04-15 17:31:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][430/1251] eta 0:03:57 lr 0.000961 time 0.2582 (0.2889) loss 4.2151 (4.0341) grad_norm 1.1351 (inf) [2021-04-15 17:31:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][440/1251] eta 0:03:54 lr 0.000961 time 0.2694 (0.2888) loss 3.6163 (4.0301) grad_norm 1.0543 (inf) [2021-04-15 17:31:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][450/1251] eta 0:03:51 lr 0.000961 time 0.2919 (0.2886) loss 3.0226 (4.0325) grad_norm 1.1395 (inf) [2021-04-15 17:31:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][460/1251] eta 0:03:47 lr 0.000961 time 0.2461 (0.2882) loss 4.5617 (4.0352) grad_norm 1.1250 (inf) [2021-04-15 17:31:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][470/1251] eta 0:03:44 lr 0.000961 time 0.2967 (0.2881) loss 3.1509 (4.0319) grad_norm 1.0667 (inf) [2021-04-15 17:31:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][480/1251] eta 0:03:41 lr 0.000961 time 0.2761 (0.2878) loss 4.3024 (4.0376) grad_norm 0.9735 (inf) [2021-04-15 17:31:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][490/1251] eta 0:03:38 lr 0.000961 time 0.2782 (0.2877) loss 3.5523 (4.0328) grad_norm 1.0640 (inf) [2021-04-15 17:31:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][500/1251] eta 0:03:35 lr 0.000961 time 0.2813 (0.2874) loss 3.8210 (4.0342) grad_norm 1.0688 (inf) [2021-04-15 17:31:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][510/1251] eta 0:03:32 lr 0.000960 time 0.2584 (0.2872) loss 4.0316 (4.0418) grad_norm 1.3017 (inf) [2021-04-15 17:31:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][520/1251] eta 0:03:29 lr 0.000960 time 0.2583 (0.2871) loss 3.9959 (4.0493) grad_norm 0.9624 (inf) [2021-04-15 17:31:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][530/1251] eta 0:03:26 lr 0.000960 time 0.2735 (0.2869) loss 4.3036 (4.0479) grad_norm 1.0713 (inf) [2021-04-15 17:31:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][540/1251] eta 0:03:23 lr 0.000960 time 0.2857 (0.2868) loss 3.8196 (4.0493) grad_norm 1.1655 (inf) [2021-04-15 17:31:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][550/1251] eta 0:03:21 lr 0.000960 time 0.4232 (0.2869) loss 3.4669 (4.0488) grad_norm 1.2152 (inf) [2021-04-15 17:31:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][560/1251] eta 0:03:18 lr 0.000960 time 0.2877 (0.2867) loss 4.2147 (4.0550) grad_norm 1.0345 (inf) [2021-04-15 17:31:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][570/1251] eta 0:03:15 lr 0.000960 time 0.2480 (0.2868) loss 4.2754 (4.0622) grad_norm 0.9578 (inf) [2021-04-15 17:31:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][580/1251] eta 0:03:12 lr 0.000960 time 0.2763 (0.2868) loss 3.3601 (4.0658) grad_norm 1.0620 (inf) [2021-04-15 17:31:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][590/1251] eta 0:03:09 lr 0.000960 time 0.2758 (0.2868) loss 4.4678 (4.0637) grad_norm 1.0515 (inf) [2021-04-15 17:31:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][600/1251] eta 0:03:06 lr 0.000960 time 0.2809 (0.2867) loss 4.6837 (4.0636) grad_norm 1.0287 (inf) [2021-04-15 17:32:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][610/1251] eta 0:03:03 lr 0.000960 time 0.3681 (0.2870) loss 2.7159 (4.0574) grad_norm 0.9354 (inf) [2021-04-15 17:32:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][620/1251] eta 0:03:01 lr 0.000960 time 0.2683 (0.2869) loss 4.6809 (4.0593) grad_norm 1.0158 (inf) [2021-04-15 17:32:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][630/1251] eta 0:02:58 lr 0.000960 time 0.2741 (0.2868) loss 4.4517 (4.0601) grad_norm 1.1265 (inf) [2021-04-15 17:32:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][640/1251] eta 0:02:55 lr 0.000960 time 0.2989 (0.2868) loss 4.8945 (4.0604) grad_norm 1.1140 (inf) [2021-04-15 17:32:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][650/1251] eta 0:02:52 lr 0.000960 time 0.2616 (0.2867) loss 4.2368 (4.0603) grad_norm 1.1504 (inf) [2021-04-15 17:32:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][660/1251] eta 0:02:49 lr 0.000960 time 0.2761 (0.2865) loss 4.6092 (4.0569) grad_norm 0.9407 (inf) [2021-04-15 17:32:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][670/1251] eta 0:02:46 lr 0.000960 time 0.2856 (0.2865) loss 2.9252 (4.0543) grad_norm 1.1578 (inf) [2021-04-15 17:32:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][680/1251] eta 0:02:43 lr 0.000960 time 0.2711 (0.2863) loss 4.1536 (4.0527) grad_norm 0.9876 (inf) [2021-04-15 17:32:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][690/1251] eta 0:02:40 lr 0.000960 time 0.2824 (0.2862) loss 4.6004 (4.0537) grad_norm 1.2771 (inf) [2021-04-15 17:32:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][700/1251] eta 0:02:37 lr 0.000960 time 0.2887 (0.2861) loss 4.4677 (4.0534) grad_norm 0.9735 (inf) [2021-04-15 17:32:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][710/1251] eta 0:02:34 lr 0.000960 time 0.2993 (0.2860) loss 4.7148 (4.0580) grad_norm 0.9687 (inf) [2021-04-15 17:32:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][720/1251] eta 0:02:31 lr 0.000960 time 0.2933 (0.2861) loss 4.5356 (4.0588) grad_norm 0.9829 (inf) [2021-04-15 17:32:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][730/1251] eta 0:02:29 lr 0.000960 time 0.2994 (0.2860) loss 3.5053 (4.0601) grad_norm 1.0289 (inf) [2021-04-15 17:32:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][740/1251] eta 0:02:26 lr 0.000960 time 0.2894 (0.2860) loss 3.7599 (4.0627) grad_norm 0.9541 (inf) [2021-04-15 17:32:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][750/1251] eta 0:02:23 lr 0.000960 time 0.2832 (0.2858) loss 4.0111 (4.0581) grad_norm 0.9596 (inf) [2021-04-15 17:32:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][760/1251] eta 0:02:20 lr 0.000960 time 0.2843 (0.2857) loss 4.0727 (4.0556) grad_norm 1.0172 (inf) [2021-04-15 17:32:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][770/1251] eta 0:02:17 lr 0.000960 time 0.2993 (0.2857) loss 3.8871 (4.0576) grad_norm 0.9570 (inf) [2021-04-15 17:32:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][780/1251] eta 0:02:14 lr 0.000960 time 0.2766 (0.2856) loss 2.4993 (4.0537) grad_norm 1.0613 (inf) [2021-04-15 17:32:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][790/1251] eta 0:02:11 lr 0.000960 time 0.2965 (0.2856) loss 4.4079 (4.0559) grad_norm 0.9889 (inf) [2021-04-15 17:32:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][800/1251] eta 0:02:08 lr 0.000960 time 0.2576 (0.2854) loss 4.4107 (4.0561) grad_norm 1.1765 (inf) [2021-04-15 17:32:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][810/1251] eta 0:02:05 lr 0.000960 time 0.2886 (0.2855) loss 2.9177 (4.0562) grad_norm 1.1986 (inf) [2021-04-15 17:32:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][820/1251] eta 0:02:03 lr 0.000960 time 0.2542 (0.2854) loss 3.1251 (4.0545) grad_norm 1.0230 (inf) [2021-04-15 17:33:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][830/1251] eta 0:02:00 lr 0.000960 time 0.2541 (0.2855) loss 4.0344 (4.0540) grad_norm 1.1420 (inf) [2021-04-15 17:33:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][840/1251] eta 0:01:57 lr 0.000960 time 0.2920 (0.2855) loss 3.3871 (4.0506) grad_norm 1.0872 (inf) [2021-04-15 17:33:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][850/1251] eta 0:01:54 lr 0.000960 time 0.2899 (0.2854) loss 3.3592 (4.0486) grad_norm 1.3240 (inf) [2021-04-15 17:33:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][860/1251] eta 0:01:51 lr 0.000960 time 0.3121 (0.2853) loss 4.9648 (4.0500) grad_norm 0.9914 (inf) [2021-04-15 17:33:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][870/1251] eta 0:01:48 lr 0.000960 time 0.2912 (0.2853) loss 4.1499 (4.0520) grad_norm 0.9705 (inf) [2021-04-15 17:33:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][880/1251] eta 0:01:45 lr 0.000960 time 0.2617 (0.2852) loss 4.1250 (4.0551) grad_norm 1.1263 (inf) [2021-04-15 17:33:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][890/1251] eta 0:01:42 lr 0.000960 time 0.2917 (0.2851) loss 3.2682 (4.0532) grad_norm 1.5642 (inf) [2021-04-15 17:33:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][900/1251] eta 0:01:40 lr 0.000960 time 0.2746 (0.2851) loss 4.4003 (4.0560) grad_norm 1.2703 (inf) [2021-04-15 17:33:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][910/1251] eta 0:01:37 lr 0.000960 time 0.2783 (0.2850) loss 4.7778 (4.0584) grad_norm 1.0935 (inf) [2021-04-15 17:33:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][920/1251] eta 0:01:34 lr 0.000960 time 0.2883 (0.2849) loss 4.5143 (4.0593) grad_norm 1.0118 (inf) [2021-04-15 17:33:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][930/1251] eta 0:01:31 lr 0.000960 time 0.2848 (0.2849) loss 4.5887 (4.0594) grad_norm 1.0771 (inf) [2021-04-15 17:33:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][940/1251] eta 0:01:28 lr 0.000960 time 0.2881 (0.2848) loss 3.9630 (4.0575) grad_norm 1.0267 (inf) [2021-04-15 17:33:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][950/1251] eta 0:01:25 lr 0.000960 time 0.2676 (0.2848) loss 4.8704 (4.0612) grad_norm 1.1605 (inf) [2021-04-15 17:33:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][960/1251] eta 0:01:22 lr 0.000960 time 0.2831 (0.2847) loss 2.7566 (4.0580) grad_norm 0.9372 (inf) [2021-04-15 17:33:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][970/1251] eta 0:01:20 lr 0.000960 time 0.2918 (0.2847) loss 4.1169 (4.0563) grad_norm 0.9487 (inf) [2021-04-15 17:33:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][980/1251] eta 0:01:17 lr 0.000960 time 0.2665 (0.2846) loss 4.5136 (4.0575) grad_norm 0.9803 (inf) [2021-04-15 17:33:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][990/1251] eta 0:01:14 lr 0.000960 time 0.2819 (0.2845) loss 4.2172 (4.0552) grad_norm 1.6313 (inf) [2021-04-15 17:33:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][1000/1251] eta 0:01:11 lr 0.000960 time 0.2715 (0.2845) loss 3.0926 (4.0554) grad_norm 0.9517 (inf) [2021-04-15 17:33:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][1010/1251] eta 0:01:08 lr 0.000960 time 0.2845 (0.2844) loss 4.6924 (4.0539) grad_norm 1.0134 (inf) [2021-04-15 17:33:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][1020/1251] eta 0:01:05 lr 0.000960 time 0.2883 (0.2843) loss 4.1636 (4.0552) grad_norm 0.9669 (inf) [2021-04-15 17:33:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][1030/1251] eta 0:01:02 lr 0.000960 time 0.2992 (0.2842) loss 4.3816 (4.0568) grad_norm 1.0384 (inf) [2021-04-15 17:34:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][1040/1251] eta 0:00:59 lr 0.000960 time 0.2765 (0.2841) loss 4.4537 (4.0575) grad_norm 1.4179 (inf) [2021-04-15 17:34:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][1050/1251] eta 0:00:57 lr 0.000960 time 0.2907 (0.2840) loss 4.8888 (4.0583) grad_norm 1.2132 (inf) [2021-04-15 17:34:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][1060/1251] eta 0:00:54 lr 0.000960 time 0.3078 (0.2840) loss 4.2465 (4.0597) grad_norm 0.8682 (inf) [2021-04-15 17:34:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][1070/1251] eta 0:00:51 lr 0.000960 time 0.2428 (0.2839) loss 3.7685 (4.0602) grad_norm 1.0127 (inf) [2021-04-15 17:34:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][1080/1251] eta 0:00:48 lr 0.000960 time 0.2875 (0.2839) loss 2.8166 (4.0603) grad_norm 1.0786 (inf) [2021-04-15 17:34:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][1090/1251] eta 0:00:45 lr 0.000960 time 0.2565 (0.2838) loss 3.2633 (4.0605) grad_norm 1.0886 (inf) [2021-04-15 17:34:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][1100/1251] eta 0:00:42 lr 0.000960 time 0.2676 (0.2838) loss 4.8731 (4.0622) grad_norm 0.9894 (inf) [2021-04-15 17:34:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][1110/1251] eta 0:00:40 lr 0.000960 time 0.3106 (0.2837) loss 3.5527 (4.0599) grad_norm 0.9097 (inf) [2021-04-15 17:34:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][1120/1251] eta 0:00:37 lr 0.000960 time 0.2665 (0.2837) loss 3.2625 (4.0573) grad_norm 1.1301 (inf) [2021-04-15 17:34:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][1130/1251] eta 0:00:34 lr 0.000959 time 0.3033 (0.2838) loss 4.4042 (4.0571) grad_norm 1.0557 (inf) [2021-04-15 17:34:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][1140/1251] eta 0:00:31 lr 0.000959 time 0.3040 (0.2838) loss 3.1524 (4.0558) grad_norm 1.1879 (inf) [2021-04-15 17:34:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][1150/1251] eta 0:00:28 lr 0.000959 time 0.2895 (0.2839) loss 4.1283 (4.0553) grad_norm 0.9588 (inf) [2021-04-15 17:34:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][1160/1251] eta 0:00:25 lr 0.000959 time 0.2559 (0.2837) loss 4.7817 (4.0566) grad_norm 1.1283 (inf) [2021-04-15 17:34:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][1170/1251] eta 0:00:22 lr 0.000959 time 0.2471 (0.2838) loss 3.6430 (4.0572) grad_norm 0.9781 (inf) [2021-04-15 17:34:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][1180/1251] eta 0:00:20 lr 0.000959 time 0.2732 (0.2838) loss 4.3847 (4.0593) grad_norm 1.0340 (inf) [2021-04-15 17:34:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][1190/1251] eta 0:00:17 lr 0.000959 time 0.2876 (0.2837) loss 4.9019 (4.0600) grad_norm 1.3167 (inf) [2021-04-15 17:34:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][1200/1251] eta 0:00:14 lr 0.000959 time 0.2717 (0.2837) loss 2.8987 (4.0606) grad_norm 1.0071 (inf) [2021-04-15 17:34:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][1210/1251] eta 0:00:11 lr 0.000959 time 0.2645 (0.2836) loss 2.7637 (4.0589) grad_norm 0.9958 (inf) [2021-04-15 17:34:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][1220/1251] eta 0:00:08 lr 0.000959 time 0.2728 (0.2837) loss 4.8864 (4.0618) grad_norm 0.9278 (inf) [2021-04-15 17:34:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][1230/1251] eta 0:00:05 lr 0.000959 time 0.2553 (0.2836) loss 4.2711 (4.0616) grad_norm 1.1092 (inf) [2021-04-15 17:34:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][1240/1251] eta 0:00:03 lr 0.000959 time 0.2487 (0.2835) loss 4.3973 (4.0613) grad_norm 1.0781 (inf) [2021-04-15 17:34:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [38/300][1250/1251] eta 0:00:00 lr 0.000959 time 0.2488 (0.2832) loss 3.9964 (4.0604) grad_norm 1.0472 (inf) [2021-04-15 17:35:01 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 38 training takes 0:05:56 [2021-04-15 17:35:01 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_38.pth saving...... [2021-04-15 17:35:15 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_38.pth saved !!! [2021-04-15 17:35:16 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.119 (1.119) Loss 1.3796 (1.3796) Acc@1 70.020 (70.020) Acc@5 88.965 (88.965) [2021-04-15 17:35:18 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 1.057 (0.296) Loss 1.4276 (1.4245) Acc@1 68.750 (68.413) Acc@5 89.062 (89.027) [2021-04-15 17:35:20 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.102 (0.227) Loss 1.3751 (1.4259) Acc@1 69.727 (68.150) Acc@5 89.551 (88.974) [2021-04-15 17:35:22 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.150 (0.216) Loss 1.4803 (1.4369) Acc@1 67.188 (67.997) Acc@5 88.184 (88.766) [2021-04-15 17:35:24 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.121 (0.213) Loss 1.4126 (1.4301) Acc@1 67.871 (68.116) Acc@5 89.551 (88.851) [2021-04-15 17:35:28 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 68.150 Acc@5 88.854 [2021-04-15 17:35:28 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 68.2% [2021-04-15 17:35:28 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 68.15% [2021-04-15 17:35:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][0/1251] eta 0:51:21 lr 0.000959 time 2.4632 (2.4632) loss 4.4220 (4.4220) grad_norm 1.1108 (1.1108) [2021-04-15 17:35:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][10/1251] eta 0:09:48 lr 0.000959 time 0.3004 (0.4743) loss 3.2242 (3.9607) grad_norm 1.1115 (1.0179) [2021-04-15 17:35:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][20/1251] eta 0:07:51 lr 0.000959 time 0.2895 (0.3832) loss 3.9810 (3.9880) grad_norm 1.0135 (1.0233) [2021-04-15 17:35:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][30/1251] eta 0:07:11 lr 0.000959 time 0.2503 (0.3535) loss 4.6634 (4.0473) grad_norm 0.9573 (1.0424) [2021-04-15 17:35:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][40/1251] eta 0:06:52 lr 0.000959 time 0.3209 (0.3403) loss 3.8933 (4.0284) grad_norm 1.1416 (1.0514) [2021-04-15 17:35:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][50/1251] eta 0:06:34 lr 0.000959 time 0.2791 (0.3289) loss 5.0832 (4.0752) grad_norm 1.0465 (1.0771) [2021-04-15 17:35:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][60/1251] eta 0:06:22 lr 0.000959 time 0.2995 (0.3208) loss 4.6498 (4.0406) grad_norm 0.9486 (1.0824) [2021-04-15 17:35:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][70/1251] eta 0:06:12 lr 0.000959 time 0.2556 (0.3151) loss 4.1201 (4.0292) grad_norm 1.2488 (1.0850) [2021-04-15 17:35:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][80/1251] eta 0:06:03 lr 0.000959 time 0.2857 (0.3105) loss 3.2937 (4.0547) grad_norm 0.8779 (1.1029) [2021-04-15 17:35:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][90/1251] eta 0:05:56 lr 0.000959 time 0.2956 (0.3073) loss 3.2426 (4.0874) grad_norm 1.0050 (1.1133) [2021-04-15 17:35:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][100/1251] eta 0:05:50 lr 0.000959 time 0.2562 (0.3047) loss 4.4469 (4.0921) grad_norm 1.0224 (1.1083) [2021-04-15 17:36:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][110/1251] eta 0:05:45 lr 0.000959 time 0.2747 (0.3026) loss 4.0177 (4.0999) grad_norm 0.9700 (1.1054) [2021-04-15 17:36:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][120/1251] eta 0:05:40 lr 0.000959 time 0.3081 (0.3008) loss 3.8734 (4.0857) grad_norm 1.0148 (1.0966) [2021-04-15 17:36:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][130/1251] eta 0:05:35 lr 0.000959 time 0.2910 (0.2995) loss 4.1569 (4.1026) grad_norm 0.9813 (1.0936) [2021-04-15 17:36:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][140/1251] eta 0:05:31 lr 0.000959 time 0.2710 (0.2979) loss 3.3025 (4.0919) grad_norm 1.3689 (1.0930) [2021-04-15 17:36:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][150/1251] eta 0:05:26 lr 0.000959 time 0.2619 (0.2969) loss 4.8876 (4.1107) grad_norm 1.0339 (1.1002) [2021-04-15 17:36:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][160/1251] eta 0:05:23 lr 0.000959 time 0.2923 (0.2961) loss 4.4128 (4.0983) grad_norm 1.0460 (1.0984) [2021-04-15 17:36:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][170/1251] eta 0:05:19 lr 0.000959 time 0.2711 (0.2951) loss 3.7106 (4.0993) grad_norm 0.9634 (1.0951) [2021-04-15 17:36:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][180/1251] eta 0:05:15 lr 0.000959 time 0.2833 (0.2941) loss 4.4543 (4.1057) grad_norm 0.9257 (1.0908) [2021-04-15 17:36:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][190/1251] eta 0:05:11 lr 0.000959 time 0.2581 (0.2933) loss 3.6208 (4.1023) grad_norm 1.1653 (1.0942) [2021-04-15 17:36:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][200/1251] eta 0:05:07 lr 0.000959 time 0.2847 (0.2927) loss 3.3648 (4.0816) grad_norm 1.0430 (1.0939) [2021-04-15 17:36:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][210/1251] eta 0:05:03 lr 0.000959 time 0.2866 (0.2919) loss 4.1230 (4.0708) grad_norm 1.0898 (1.0908) [2021-04-15 17:36:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][220/1251] eta 0:05:00 lr 0.000959 time 0.2877 (0.2915) loss 3.9839 (4.0730) grad_norm 1.1900 (1.0881) [2021-04-15 17:36:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][230/1251] eta 0:04:57 lr 0.000959 time 0.2890 (0.2910) loss 4.5552 (4.0890) grad_norm 1.0679 (1.0893) [2021-04-15 17:36:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][240/1251] eta 0:04:53 lr 0.000959 time 0.2634 (0.2904) loss 4.1582 (4.0824) grad_norm 1.0986 (1.0929) [2021-04-15 17:36:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][250/1251] eta 0:04:50 lr 0.000959 time 0.3078 (0.2900) loss 3.5074 (4.0801) grad_norm 1.2105 (1.0946) [2021-04-15 17:36:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][260/1251] eta 0:04:47 lr 0.000959 time 0.2766 (0.2898) loss 3.3346 (4.0746) grad_norm 0.9808 (1.0945) [2021-04-15 17:36:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][270/1251] eta 0:04:43 lr 0.000959 time 0.2774 (0.2894) loss 4.1179 (4.0678) grad_norm 0.9054 (1.0916) [2021-04-15 17:36:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][280/1251] eta 0:04:40 lr 0.000959 time 0.2811 (0.2891) loss 2.9380 (4.0718) grad_norm 1.0112 (1.0912) [2021-04-15 17:36:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][290/1251] eta 0:04:37 lr 0.000959 time 0.3023 (0.2888) loss 3.7772 (4.0719) grad_norm 0.9951 (1.0927) [2021-04-15 17:36:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][300/1251] eta 0:04:34 lr 0.000959 time 0.2810 (0.2883) loss 3.9167 (4.0734) grad_norm 1.0897 (1.0932) [2021-04-15 17:36:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][310/1251] eta 0:04:31 lr 0.000959 time 0.2767 (0.2885) loss 3.3527 (4.0728) grad_norm 0.9999 (1.0933) [2021-04-15 17:37:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][320/1251] eta 0:04:28 lr 0.000959 time 0.2911 (0.2882) loss 4.1240 (4.0757) grad_norm 1.0160 (1.0931) [2021-04-15 17:37:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][330/1251] eta 0:04:25 lr 0.000959 time 0.2896 (0.2883) loss 3.5123 (4.0732) grad_norm 1.0740 (1.0919) [2021-04-15 17:37:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][340/1251] eta 0:04:22 lr 0.000959 time 0.2851 (0.2880) loss 3.5638 (4.0809) grad_norm 1.0142 (1.0918) [2021-04-15 17:37:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][350/1251] eta 0:04:19 lr 0.000959 time 0.2580 (0.2877) loss 3.4539 (4.0792) grad_norm 0.9679 (1.0915) [2021-04-15 17:37:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][360/1251] eta 0:04:16 lr 0.000959 time 0.2954 (0.2875) loss 3.3504 (4.0790) grad_norm 1.3449 (1.0966) [2021-04-15 17:37:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][370/1251] eta 0:04:13 lr 0.000959 time 0.2767 (0.2876) loss 4.7513 (4.0794) grad_norm 1.2094 (1.0986) [2021-04-15 17:37:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][380/1251] eta 0:04:10 lr 0.000959 time 0.2856 (0.2878) loss 4.3597 (4.0716) grad_norm 0.9291 (1.0978) [2021-04-15 17:37:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][390/1251] eta 0:04:07 lr 0.000959 time 0.2534 (0.2875) loss 4.2239 (4.0669) grad_norm 1.0070 (1.0960) [2021-04-15 17:37:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][400/1251] eta 0:04:04 lr 0.000959 time 0.3028 (0.2874) loss 4.4135 (4.0706) grad_norm 1.1393 (1.0954) [2021-04-15 17:37:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][410/1251] eta 0:04:01 lr 0.000959 time 0.2988 (0.2873) loss 4.2761 (4.0737) grad_norm 1.1157 (1.0949) [2021-04-15 17:37:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][420/1251] eta 0:03:58 lr 0.000959 time 0.2977 (0.2871) loss 4.2709 (4.0790) grad_norm 1.0476 (1.0952) [2021-04-15 17:37:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][430/1251] eta 0:03:55 lr 0.000959 time 0.2763 (0.2869) loss 3.9866 (4.0740) grad_norm 1.0609 (1.0944) [2021-04-15 17:37:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][440/1251] eta 0:03:52 lr 0.000959 time 0.2619 (0.2867) loss 4.3023 (4.0742) grad_norm 0.8987 (1.0934) [2021-04-15 17:37:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][450/1251] eta 0:03:49 lr 0.000959 time 0.2889 (0.2865) loss 4.0140 (4.0755) grad_norm 1.1058 (1.0941) [2021-04-15 17:37:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][460/1251] eta 0:03:46 lr 0.000959 time 0.2890 (0.2864) loss 5.0694 (4.0815) grad_norm 1.2092 (1.0937) [2021-04-15 17:37:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][470/1251] eta 0:03:43 lr 0.000959 time 0.2562 (0.2862) loss 3.0359 (4.0814) grad_norm 1.5695 (1.0927) [2021-04-15 17:37:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][480/1251] eta 0:03:40 lr 0.000958 time 0.2864 (0.2860) loss 4.8017 (4.0810) grad_norm 1.0587 (1.0930) [2021-04-15 17:37:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][490/1251] eta 0:03:37 lr 0.000958 time 0.2722 (0.2858) loss 4.1187 (4.0759) grad_norm 1.1522 (1.0931) [2021-04-15 17:37:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][500/1251] eta 0:03:34 lr 0.000958 time 0.2784 (0.2857) loss 3.1765 (4.0706) grad_norm 1.2552 (1.0934) [2021-04-15 17:37:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][510/1251] eta 0:03:31 lr 0.000958 time 0.2680 (0.2856) loss 4.8042 (4.0659) grad_norm 1.0924 (1.0925) [2021-04-15 17:37:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][520/1251] eta 0:03:28 lr 0.000958 time 0.2680 (0.2855) loss 4.2444 (4.0681) grad_norm 0.9661 (1.0920) [2021-04-15 17:38:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][530/1251] eta 0:03:25 lr 0.000958 time 0.2579 (0.2854) loss 4.2778 (4.0660) grad_norm 0.8506 (1.0910) [2021-04-15 17:38:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][540/1251] eta 0:03:22 lr 0.000958 time 0.2852 (0.2854) loss 3.0784 (4.0595) grad_norm 0.8928 (1.0891) [2021-04-15 17:38:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][550/1251] eta 0:03:19 lr 0.000958 time 0.2504 (0.2851) loss 4.1788 (4.0539) grad_norm 1.1594 (1.0883) [2021-04-15 17:38:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][560/1251] eta 0:03:16 lr 0.000958 time 0.2830 (0.2850) loss 3.8353 (4.0553) grad_norm 1.1605 (1.0890) [2021-04-15 17:38:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][570/1251] eta 0:03:14 lr 0.000958 time 0.2860 (0.2849) loss 4.2130 (4.0583) grad_norm 1.0136 (1.0891) [2021-04-15 17:38:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][580/1251] eta 0:03:11 lr 0.000958 time 0.2711 (0.2851) loss 3.0186 (4.0575) grad_norm 1.2900 (1.0885) [2021-04-15 17:38:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][590/1251] eta 0:03:08 lr 0.000958 time 0.3037 (0.2849) loss 5.0142 (4.0569) grad_norm 0.9823 (1.0898) [2021-04-15 17:38:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][600/1251] eta 0:03:05 lr 0.000958 time 0.4547 (0.2850) loss 4.0389 (4.0595) grad_norm 1.3309 (1.0905) [2021-04-15 17:38:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][610/1251] eta 0:03:02 lr 0.000958 time 0.2744 (0.2849) loss 3.0737 (4.0589) grad_norm 0.9351 (1.0906) [2021-04-15 17:38:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][620/1251] eta 0:02:59 lr 0.000958 time 0.2948 (0.2851) loss 4.0934 (4.0581) grad_norm 1.0538 (1.0912) [2021-04-15 17:38:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][630/1251] eta 0:02:56 lr 0.000958 time 0.2710 (0.2849) loss 4.1417 (4.0622) grad_norm 1.0912 (1.0929) [2021-04-15 17:38:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][640/1251] eta 0:02:53 lr 0.000958 time 0.2892 (0.2848) loss 4.7227 (4.0664) grad_norm 0.9543 (1.0927) [2021-04-15 17:38:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][650/1251] eta 0:02:51 lr 0.000958 time 0.2703 (0.2848) loss 4.6841 (4.0638) grad_norm 1.1050 (1.0920) [2021-04-15 17:38:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][660/1251] eta 0:02:48 lr 0.000958 time 0.2846 (0.2847) loss 3.9894 (4.0603) grad_norm 1.2766 (1.0930) [2021-04-15 17:38:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][670/1251] eta 0:02:45 lr 0.000958 time 0.2658 (0.2846) loss 3.7798 (4.0581) grad_norm 0.9578 (1.0935) [2021-04-15 17:38:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][680/1251] eta 0:02:42 lr 0.000958 time 0.2962 (0.2845) loss 3.4761 (4.0568) grad_norm 1.0245 (1.0939) [2021-04-15 17:38:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][690/1251] eta 0:02:39 lr 0.000958 time 0.2955 (0.2845) loss 3.7550 (4.0562) grad_norm 1.0386 (1.0927) [2021-04-15 17:38:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][700/1251] eta 0:02:36 lr 0.000958 time 0.2738 (0.2843) loss 4.5395 (4.0567) grad_norm 1.0956 (1.0924) [2021-04-15 17:38:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][710/1251] eta 0:02:33 lr 0.000958 time 0.2713 (0.2842) loss 3.7233 (4.0585) grad_norm 1.1922 (1.0946) [2021-04-15 17:38:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][720/1251] eta 0:02:30 lr 0.000958 time 0.2675 (0.2841) loss 4.2056 (4.0569) grad_norm 1.0209 (1.0939) [2021-04-15 17:38:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][730/1251] eta 0:02:27 lr 0.000958 time 0.2966 (0.2840) loss 3.6205 (4.0607) grad_norm 1.0909 (1.0939) [2021-04-15 17:38:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][740/1251] eta 0:02:25 lr 0.000958 time 0.2801 (0.2840) loss 3.7793 (4.0619) grad_norm 0.9201 (1.0941) [2021-04-15 17:39:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][750/1251] eta 0:02:22 lr 0.000958 time 0.2894 (0.2840) loss 4.4366 (4.0664) grad_norm 1.1464 (1.0941) [2021-04-15 17:39:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][760/1251] eta 0:02:19 lr 0.000958 time 0.2835 (0.2840) loss 4.3519 (4.0655) grad_norm 0.9841 (1.0931) [2021-04-15 17:39:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][770/1251] eta 0:02:16 lr 0.000958 time 0.2708 (0.2839) loss 4.3365 (4.0682) grad_norm 0.8781 (1.0919) [2021-04-15 17:39:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][780/1251] eta 0:02:13 lr 0.000958 time 0.2808 (0.2838) loss 3.1620 (4.0685) grad_norm 0.9585 (1.0918) [2021-04-15 17:39:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][790/1251] eta 0:02:10 lr 0.000958 time 0.2663 (0.2837) loss 4.3001 (4.0729) grad_norm 1.0051 (1.0916) [2021-04-15 17:39:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][800/1251] eta 0:02:07 lr 0.000958 time 0.2815 (0.2836) loss 4.1074 (4.0749) grad_norm 0.9347 (1.0909) [2021-04-15 17:39:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][810/1251] eta 0:02:05 lr 0.000958 time 0.2732 (0.2836) loss 3.3563 (4.0764) grad_norm 1.4478 (1.0907) [2021-04-15 17:39:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][820/1251] eta 0:02:02 lr 0.000958 time 0.2741 (0.2835) loss 3.4789 (4.0766) grad_norm 1.5447 (1.0919) [2021-04-15 17:39:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][830/1251] eta 0:01:59 lr 0.000958 time 0.2668 (0.2835) loss 3.5134 (4.0766) grad_norm 1.0066 (1.0912) [2021-04-15 17:39:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][840/1251] eta 0:01:56 lr 0.000958 time 0.2574 (0.2834) loss 4.6173 (4.0798) grad_norm 1.1692 (1.0913) [2021-04-15 17:39:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][850/1251] eta 0:01:53 lr 0.000958 time 0.2863 (0.2835) loss 4.0273 (4.0828) grad_norm 0.9881 (1.0908) [2021-04-15 17:39:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][860/1251] eta 0:01:50 lr 0.000958 time 0.2771 (0.2834) loss 3.8367 (4.0824) grad_norm 1.0490 (1.0900) [2021-04-15 17:39:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][870/1251] eta 0:01:47 lr 0.000958 time 0.2784 (0.2834) loss 3.1071 (4.0792) grad_norm 1.1146 (1.0902) [2021-04-15 17:39:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][880/1251] eta 0:01:45 lr 0.000958 time 0.4284 (0.2835) loss 3.9556 (4.0797) grad_norm 1.1279 (1.0898) [2021-04-15 17:39:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][890/1251] eta 0:01:42 lr 0.000958 time 0.2664 (0.2833) loss 4.0779 (4.0808) grad_norm 1.2244 (1.0903) [2021-04-15 17:39:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][900/1251] eta 0:01:39 lr 0.000958 time 0.2824 (0.2834) loss 3.1260 (4.0799) grad_norm 1.0450 (1.0902) [2021-04-15 17:39:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][910/1251] eta 0:01:36 lr 0.000958 time 0.3029 (0.2834) loss 3.6213 (4.0818) grad_norm 1.0280 (1.0897) [2021-04-15 17:39:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][920/1251] eta 0:01:33 lr 0.000958 time 0.2949 (0.2833) loss 3.2958 (4.0780) grad_norm 1.3566 (1.0898) [2021-04-15 17:39:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][930/1251] eta 0:01:30 lr 0.000958 time 0.2931 (0.2833) loss 3.3200 (4.0749) grad_norm 1.2044 (1.0903) [2021-04-15 17:39:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][940/1251] eta 0:01:28 lr 0.000958 time 0.2598 (0.2834) loss 4.5815 (4.0750) grad_norm 1.0367 (1.0908) [2021-04-15 17:39:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][950/1251] eta 0:01:25 lr 0.000958 time 0.4032 (0.2834) loss 4.2774 (4.0742) grad_norm 1.0986 (1.0911) [2021-04-15 17:40:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][960/1251] eta 0:01:22 lr 0.000958 time 0.2648 (0.2834) loss 4.0609 (4.0747) grad_norm 0.9357 (1.0911) [2021-04-15 17:40:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][970/1251] eta 0:01:19 lr 0.000958 time 0.2572 (0.2833) loss 3.4834 (4.0724) grad_norm 0.9657 (1.0920) [2021-04-15 17:40:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][980/1251] eta 0:01:16 lr 0.000958 time 0.2770 (0.2833) loss 3.0770 (4.0713) grad_norm 1.0127 (1.0922) [2021-04-15 17:40:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][990/1251] eta 0:01:13 lr 0.000958 time 0.2768 (0.2832) loss 4.2077 (4.0726) grad_norm 1.2950 (1.0921) [2021-04-15 17:40:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][1000/1251] eta 0:01:11 lr 0.000958 time 0.2857 (0.2832) loss 4.5518 (4.0699) grad_norm 1.1412 (1.0914) [2021-04-15 17:40:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][1010/1251] eta 0:01:08 lr 0.000958 time 0.2862 (0.2832) loss 4.5910 (4.0701) grad_norm 1.0525 (1.0908) [2021-04-15 17:40:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][1020/1251] eta 0:01:05 lr 0.000958 time 0.2811 (0.2831) loss 4.0429 (4.0727) grad_norm 1.1307 (1.0907) [2021-04-15 17:40:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][1030/1251] eta 0:01:02 lr 0.000958 time 0.2729 (0.2831) loss 3.0183 (4.0756) grad_norm 1.1173 (1.0910) [2021-04-15 17:40:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][1040/1251] eta 0:00:59 lr 0.000958 time 0.2689 (0.2830) loss 3.7005 (4.0777) grad_norm 1.4579 (1.0916) [2021-04-15 17:40:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][1050/1251] eta 0:00:56 lr 0.000958 time 0.2661 (0.2830) loss 3.5150 (4.0778) grad_norm 0.9451 (1.0913) [2021-04-15 17:40:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][1060/1251] eta 0:00:54 lr 0.000958 time 0.2954 (0.2830) loss 3.9234 (4.0795) grad_norm 1.0961 (1.0914) [2021-04-15 17:40:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][1070/1251] eta 0:00:51 lr 0.000958 time 0.2915 (0.2830) loss 4.4377 (4.0812) grad_norm 1.3009 (1.0910) [2021-04-15 17:40:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][1080/1251] eta 0:00:48 lr 0.000957 time 0.2953 (0.2829) loss 4.6748 (4.0825) grad_norm 0.9933 (1.0907) [2021-04-15 17:40:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][1090/1251] eta 0:00:45 lr 0.000957 time 0.2871 (0.2829) loss 3.7829 (4.0828) grad_norm 1.4907 (1.0911) [2021-04-15 17:40:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][1100/1251] eta 0:00:42 lr 0.000957 time 0.2891 (0.2828) loss 4.2697 (4.0833) grad_norm 1.0281 (1.0915) [2021-04-15 17:40:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][1110/1251] eta 0:00:39 lr 0.000957 time 0.2608 (0.2827) loss 4.0062 (4.0859) grad_norm 1.0702 (1.0912) [2021-04-15 17:40:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][1120/1251] eta 0:00:37 lr 0.000957 time 0.2739 (0.2827) loss 4.4071 (4.0861) grad_norm 1.1824 (1.0916) [2021-04-15 17:40:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][1130/1251] eta 0:00:34 lr 0.000957 time 0.2809 (0.2827) loss 4.6523 (4.0860) grad_norm 1.0719 (1.0928) [2021-04-15 17:40:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][1140/1251] eta 0:00:31 lr 0.000957 time 0.2648 (0.2827) loss 4.3804 (4.0838) grad_norm 1.3001 (1.0927) [2021-04-15 17:40:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][1150/1251] eta 0:00:28 lr 0.000957 time 0.2873 (0.2826) loss 4.8213 (4.0837) grad_norm 0.9930 (1.0916) [2021-04-15 17:40:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][1160/1251] eta 0:00:25 lr 0.000957 time 0.2792 (0.2827) loss 4.0251 (4.0819) grad_norm 0.9932 (1.0917) [2021-04-15 17:40:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][1170/1251] eta 0:00:22 lr 0.000957 time 0.2724 (0.2827) loss 3.6442 (4.0795) grad_norm 0.8473 (1.0911) [2021-04-15 17:41:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][1180/1251] eta 0:00:20 lr 0.000957 time 0.2667 (0.2827) loss 4.5837 (4.0773) grad_norm 0.8545 (1.0902) [2021-04-15 17:41:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][1190/1251] eta 0:00:17 lr 0.000957 time 0.2726 (0.2827) loss 4.6874 (4.0764) grad_norm 1.2560 (1.0905) [2021-04-15 17:41:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][1200/1251] eta 0:00:14 lr 0.000957 time 0.2734 (0.2827) loss 4.0018 (4.0762) grad_norm 1.1884 (1.0906) [2021-04-15 17:41:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][1210/1251] eta 0:00:11 lr 0.000957 time 0.2874 (0.2827) loss 4.2417 (4.0770) grad_norm 0.9483 (1.0901) [2021-04-15 17:41:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][1220/1251] eta 0:00:08 lr 0.000957 time 0.2647 (0.2826) loss 4.1095 (4.0775) grad_norm 1.0086 (1.0901) [2021-04-15 17:41:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][1230/1251] eta 0:00:05 lr 0.000957 time 0.2750 (0.2826) loss 3.4646 (4.0779) grad_norm 1.1002 (1.0902) [2021-04-15 17:41:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][1240/1251] eta 0:00:03 lr 0.000957 time 0.2492 (0.2825) loss 4.2209 (4.0793) grad_norm 1.1306 (1.0902) [2021-04-15 17:41:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [39/300][1250/1251] eta 0:00:00 lr 0.000957 time 0.2485 (0.2822) loss 3.5934 (4.0770) grad_norm 1.0330 (1.0906) [2021-04-15 17:41:24 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 39 training takes 0:05:55 [2021-04-15 17:41:24 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_39.pth saving...... [2021-04-15 17:41:35 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_39.pth saved !!! [2021-04-15 17:41:36 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.349 (1.349) Loss 1.3836 (1.3836) Acc@1 67.480 (67.480) Acc@5 88.281 (88.281) [2021-04-15 17:41:37 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.161 (0.238) Loss 1.3591 (1.3737) Acc@1 66.992 (68.262) Acc@5 89.551 (88.965) [2021-04-15 17:41:40 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.177 (0.239) Loss 1.3718 (1.3754) Acc@1 67.871 (68.141) Acc@5 88.867 (88.946) [2021-04-15 17:41:42 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.103 (0.240) Loss 1.4249 (1.3801) Acc@1 65.723 (68.025) Acc@5 88.770 (88.918) [2021-04-15 17:41:44 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.223) Loss 1.2734 (1.3710) Acc@1 69.824 (68.309) Acc@5 91.504 (89.065) [2021-04-15 17:41:47 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 68.560 Acc@5 89.210 [2021-04-15 17:41:47 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 68.6% [2021-04-15 17:41:47 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 68.56% [2021-04-15 17:41:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][0/1251] eta 1:23:46 lr 0.000957 time 4.0183 (4.0183) loss 4.7427 (4.7427) grad_norm 1.1717 (1.1717) [2021-04-15 17:41:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][10/1251] eta 0:12:45 lr 0.000957 time 0.2594 (0.6167) loss 4.2838 (4.2580) grad_norm 1.1172 (1.1033) [2021-04-15 17:41:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][20/1251] eta 0:09:20 lr 0.000957 time 0.2915 (0.4550) loss 4.2338 (4.0635) grad_norm 1.2129 (1.1330) [2021-04-15 17:41:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][30/1251] eta 0:08:06 lr 0.000957 time 0.2686 (0.3988) loss 4.3020 (3.9530) grad_norm 1.3843 (1.1511) [2021-04-15 17:42:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][40/1251] eta 0:07:28 lr 0.000957 time 0.3225 (0.3706) loss 3.5851 (3.9804) grad_norm 1.0049 (1.1413) [2021-04-15 17:42:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][50/1251] eta 0:07:03 lr 0.000957 time 0.2772 (0.3527) loss 3.1349 (3.9238) grad_norm 1.0844 (1.1247) [2021-04-15 17:42:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][60/1251] eta 0:06:45 lr 0.000957 time 0.2852 (0.3408) loss 4.2388 (3.9839) grad_norm 1.0001 (1.1064) [2021-04-15 17:42:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][70/1251] eta 0:06:34 lr 0.000957 time 0.2596 (0.3344) loss 3.1373 (3.9989) grad_norm 1.0134 (1.1007) [2021-04-15 17:42:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][80/1251] eta 0:06:23 lr 0.000957 time 0.2668 (0.3277) loss 2.9915 (3.9981) grad_norm 1.2944 (1.1222) [2021-04-15 17:42:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][90/1251] eta 0:06:14 lr 0.000957 time 0.2566 (0.3224) loss 4.7250 (4.0254) grad_norm 1.0470 (1.1243) [2021-04-15 17:42:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][100/1251] eta 0:06:08 lr 0.000957 time 0.4416 (0.3200) loss 5.1365 (4.0352) grad_norm 1.1135 (1.1217) [2021-04-15 17:42:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][110/1251] eta 0:06:01 lr 0.000957 time 0.2615 (0.3164) loss 4.5832 (4.0315) grad_norm 0.9934 (1.1183) [2021-04-15 17:42:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][120/1251] eta 0:05:54 lr 0.000957 time 0.2835 (0.3134) loss 3.6334 (4.0344) grad_norm 1.0117 (1.1133) [2021-04-15 17:42:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][130/1251] eta 0:05:48 lr 0.000957 time 0.3157 (0.3108) loss 4.2479 (4.0303) grad_norm 1.0308 (1.1093) [2021-04-15 17:42:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][140/1251] eta 0:05:44 lr 0.000957 time 0.2745 (0.3101) loss 4.4948 (4.0297) grad_norm 1.0634 (1.1074) [2021-04-15 17:42:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][150/1251] eta 0:05:40 lr 0.000957 time 0.2837 (0.3095) loss 4.1683 (4.0503) grad_norm 1.0037 (1.1082) [2021-04-15 17:42:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][160/1251] eta 0:05:35 lr 0.000957 time 0.3002 (0.3076) loss 4.0509 (4.0654) grad_norm 1.3315 (1.1140) [2021-04-15 17:42:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][170/1251] eta 0:05:30 lr 0.000957 time 0.2706 (0.3061) loss 3.1253 (4.0557) grad_norm 1.0764 (1.1141) [2021-04-15 17:42:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][180/1251] eta 0:05:26 lr 0.000957 time 0.3161 (0.3047) loss 4.4155 (4.0525) grad_norm 1.0025 (1.1148) [2021-04-15 17:42:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][190/1251] eta 0:05:21 lr 0.000957 time 0.2678 (0.3033) loss 3.6141 (4.0522) grad_norm 1.4429 (1.1123) [2021-04-15 17:42:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][200/1251] eta 0:05:17 lr 0.000957 time 0.2515 (0.3020) loss 3.9730 (4.0565) grad_norm 1.0211 (1.1100) [2021-04-15 17:42:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][210/1251] eta 0:05:13 lr 0.000957 time 0.2892 (0.3011) loss 4.1816 (4.0513) grad_norm 0.9964 (1.1048) [2021-04-15 17:42:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][220/1251] eta 0:05:09 lr 0.000957 time 0.2757 (0.3004) loss 3.8248 (4.0530) grad_norm 1.1828 (1.1062) [2021-04-15 17:42:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][230/1251] eta 0:05:05 lr 0.000957 time 0.3029 (0.2997) loss 4.0189 (4.0590) grad_norm 1.0409 (1.1074) [2021-04-15 17:42:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][240/1251] eta 0:05:02 lr 0.000957 time 0.2435 (0.2990) loss 4.2481 (4.0601) grad_norm 1.0680 (1.1071) [2021-04-15 17:43:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][250/1251] eta 0:04:58 lr 0.000957 time 0.2704 (0.2986) loss 4.3048 (4.0681) grad_norm 1.2514 (1.1084) [2021-04-15 17:43:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][260/1251] eta 0:04:55 lr 0.000957 time 0.2686 (0.2978) loss 4.4567 (4.0661) grad_norm 0.9125 (1.1085) [2021-04-15 17:43:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][270/1251] eta 0:04:51 lr 0.000957 time 0.2935 (0.2971) loss 2.7799 (4.0431) grad_norm 0.8673 (1.1064) [2021-04-15 17:43:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][280/1251] eta 0:04:47 lr 0.000957 time 0.2511 (0.2961) loss 4.1132 (4.0401) grad_norm 1.1947 (1.1055) [2021-04-15 17:43:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][290/1251] eta 0:04:44 lr 0.000957 time 0.2652 (0.2956) loss 4.1535 (4.0338) grad_norm 0.9163 (1.1050) [2021-04-15 17:43:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][300/1251] eta 0:04:40 lr 0.000957 time 0.2965 (0.2952) loss 4.8245 (4.0426) grad_norm 1.2361 (1.1058) [2021-04-15 17:43:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][310/1251] eta 0:04:37 lr 0.000957 time 0.3205 (0.2947) loss 3.2615 (4.0457) grad_norm 1.2527 (1.1060) [2021-04-15 17:43:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][320/1251] eta 0:04:33 lr 0.000957 time 0.2884 (0.2943) loss 2.7017 (4.0502) grad_norm 1.0009 (1.1048) [2021-04-15 17:43:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][330/1251] eta 0:04:30 lr 0.000957 time 0.2791 (0.2938) loss 3.1737 (4.0444) grad_norm 1.1110 (1.1041) [2021-04-15 17:43:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][340/1251] eta 0:04:27 lr 0.000957 time 0.2732 (0.2934) loss 4.2853 (4.0399) grad_norm 1.1058 (1.1030) [2021-04-15 17:43:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][350/1251] eta 0:04:24 lr 0.000957 time 0.3001 (0.2932) loss 4.8488 (4.0379) grad_norm 0.8899 (1.1019) [2021-04-15 17:43:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][360/1251] eta 0:04:21 lr 0.000957 time 0.2726 (0.2931) loss 4.1280 (4.0335) grad_norm 0.9926 (1.1022) [2021-04-15 17:43:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][370/1251] eta 0:04:17 lr 0.000957 time 0.2669 (0.2928) loss 4.5937 (4.0377) grad_norm 1.0362 (1.1024) [2021-04-15 17:43:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][380/1251] eta 0:04:15 lr 0.000957 time 0.2907 (0.2930) loss 4.1695 (4.0401) grad_norm 1.0668 (1.1021) [2021-04-15 17:43:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][390/1251] eta 0:04:12 lr 0.000957 time 0.2627 (0.2930) loss 3.7722 (4.0374) grad_norm 0.9083 (1.1011) [2021-04-15 17:43:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][400/1251] eta 0:04:09 lr 0.000957 time 0.2959 (0.2927) loss 4.7962 (4.0403) grad_norm 1.0729 (1.1000) [2021-04-15 17:43:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][410/1251] eta 0:04:05 lr 0.000957 time 0.2586 (0.2924) loss 2.8805 (4.0406) grad_norm 1.1426 (1.0987) [2021-04-15 17:43:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][420/1251] eta 0:04:02 lr 0.000956 time 0.2701 (0.2923) loss 4.3590 (4.0380) grad_norm 0.9740 (1.0988) [2021-04-15 17:43:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][430/1251] eta 0:03:59 lr 0.000956 time 0.2766 (0.2922) loss 4.2670 (4.0437) grad_norm 1.1039 (1.0987) [2021-04-15 17:43:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][440/1251] eta 0:03:56 lr 0.000956 time 0.2681 (0.2918) loss 4.7287 (4.0463) grad_norm 1.2219 (1.0987) [2021-04-15 17:43:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][450/1251] eta 0:03:53 lr 0.000956 time 0.2796 (0.2916) loss 4.4777 (4.0486) grad_norm 0.9037 (1.0969) [2021-04-15 17:44:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][460/1251] eta 0:03:50 lr 0.000956 time 0.2918 (0.2915) loss 3.9569 (4.0428) grad_norm 0.9557 (1.0970) [2021-04-15 17:44:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][470/1251] eta 0:03:47 lr 0.000956 time 0.2627 (0.2912) loss 4.8945 (4.0435) grad_norm 1.2669 (1.0966) [2021-04-15 17:44:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][480/1251] eta 0:03:44 lr 0.000956 time 0.2915 (0.2910) loss 3.5474 (4.0431) grad_norm 1.0223 (1.0942) [2021-04-15 17:44:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][490/1251] eta 0:03:41 lr 0.000956 time 0.2706 (0.2908) loss 4.5594 (4.0479) grad_norm 1.1611 (1.0937) [2021-04-15 17:44:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][500/1251] eta 0:03:38 lr 0.000956 time 0.2779 (0.2906) loss 4.6012 (4.0528) grad_norm 0.9800 (1.0927) [2021-04-15 17:44:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][510/1251] eta 0:03:35 lr 0.000956 time 0.2630 (0.2904) loss 3.2609 (4.0540) grad_norm 0.9505 (1.0914) [2021-04-15 17:44:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][520/1251] eta 0:03:32 lr 0.000956 time 0.2757 (0.2903) loss 4.0830 (4.0577) grad_norm 1.2947 (1.0902) [2021-04-15 17:44:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][530/1251] eta 0:03:29 lr 0.000956 time 0.2676 (0.2900) loss 4.5634 (4.0566) grad_norm 1.1101 (1.0895) [2021-04-15 17:44:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][540/1251] eta 0:03:25 lr 0.000956 time 0.2628 (0.2897) loss 4.1304 (4.0572) grad_norm 1.1275 (1.0900) [2021-04-15 17:44:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][550/1251] eta 0:03:22 lr 0.000956 time 0.2731 (0.2895) loss 3.8904 (4.0559) grad_norm 1.0426 (1.0897) [2021-04-15 17:44:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][560/1251] eta 0:03:20 lr 0.000956 time 0.2771 (0.2896) loss 2.8664 (4.0571) grad_norm 0.9192 (1.0901) [2021-04-15 17:44:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][570/1251] eta 0:03:17 lr 0.000956 time 0.2980 (0.2893) loss 3.8144 (4.0573) grad_norm 1.1744 (1.0919) [2021-04-15 17:44:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][580/1251] eta 0:03:14 lr 0.000956 time 0.2764 (0.2892) loss 3.2413 (4.0496) grad_norm 1.1648 (1.0923) [2021-04-15 17:44:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][590/1251] eta 0:03:11 lr 0.000956 time 0.2843 (0.2890) loss 4.4350 (4.0524) grad_norm 1.0029 (1.0915) [2021-04-15 17:44:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][600/1251] eta 0:03:08 lr 0.000956 time 0.2660 (0.2889) loss 4.3753 (4.0542) grad_norm 1.0581 (1.0917) [2021-04-15 17:44:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][610/1251] eta 0:03:05 lr 0.000956 time 0.2953 (0.2888) loss 4.0504 (4.0542) grad_norm 1.0008 (1.0933) [2021-04-15 17:44:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][620/1251] eta 0:03:02 lr 0.000956 time 0.2606 (0.2887) loss 2.9586 (4.0526) grad_norm 1.0352 (1.0953) [2021-04-15 17:44:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][630/1251] eta 0:02:59 lr 0.000956 time 0.2916 (0.2885) loss 3.9782 (4.0535) grad_norm 1.2212 (1.0945) [2021-04-15 17:44:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][640/1251] eta 0:02:56 lr 0.000956 time 0.2716 (0.2887) loss 3.9860 (4.0538) grad_norm 1.0736 (1.0940) [2021-04-15 17:44:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][650/1251] eta 0:02:53 lr 0.000956 time 0.2624 (0.2885) loss 4.2441 (4.0566) grad_norm 1.4471 (1.0960) [2021-04-15 17:44:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][660/1251] eta 0:02:50 lr 0.000956 time 0.2989 (0.2885) loss 3.8732 (4.0581) grad_norm 1.2562 (1.0970) [2021-04-15 17:45:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][670/1251] eta 0:02:47 lr 0.000956 time 0.2732 (0.2883) loss 3.4692 (4.0595) grad_norm 1.0344 (1.0973) [2021-04-15 17:45:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][680/1251] eta 0:02:44 lr 0.000956 time 0.2945 (0.2884) loss 2.8854 (4.0600) grad_norm 1.1515 (1.0979) [2021-04-15 17:45:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][690/1251] eta 0:02:41 lr 0.000956 time 0.2922 (0.2883) loss 2.9854 (4.0562) grad_norm 0.9638 (1.0969) [2021-04-15 17:45:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][700/1251] eta 0:02:38 lr 0.000956 time 0.2597 (0.2881) loss 2.9903 (4.0535) grad_norm 0.9176 (1.0966) [2021-04-15 17:45:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][710/1251] eta 0:02:35 lr 0.000956 time 0.2762 (0.2880) loss 4.4872 (4.0487) grad_norm 1.0374 (1.0964) [2021-04-15 17:45:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][720/1251] eta 0:02:32 lr 0.000956 time 0.2939 (0.2881) loss 4.3213 (4.0493) grad_norm 1.1313 (1.0968) [2021-04-15 17:45:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][730/1251] eta 0:02:30 lr 0.000956 time 0.3016 (0.2881) loss 3.9055 (4.0492) grad_norm 1.1432 (1.0968) [2021-04-15 17:45:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][740/1251] eta 0:02:27 lr 0.000956 time 0.2808 (0.2880) loss 4.3192 (4.0478) grad_norm 1.1956 (1.0969) [2021-04-15 17:45:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][750/1251] eta 0:02:24 lr 0.000956 time 0.3111 (0.2879) loss 3.5454 (4.0474) grad_norm 1.0220 (1.0967) [2021-04-15 17:45:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][760/1251] eta 0:02:21 lr 0.000956 time 0.2580 (0.2877) loss 2.6874 (4.0465) grad_norm 1.0474 (1.0953) [2021-04-15 17:45:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][770/1251] eta 0:02:18 lr 0.000956 time 0.2949 (0.2876) loss 4.5291 (4.0450) grad_norm 1.0080 (1.0942) [2021-04-15 17:45:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][780/1251] eta 0:02:15 lr 0.000956 time 0.2981 (0.2875) loss 4.1109 (4.0435) grad_norm 1.3217 (inf) [2021-04-15 17:45:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][790/1251] eta 0:02:12 lr 0.000956 time 0.2629 (0.2874) loss 4.5999 (4.0458) grad_norm 1.1913 (inf) [2021-04-15 17:45:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][800/1251] eta 0:02:09 lr 0.000956 time 0.2869 (0.2873) loss 4.4127 (4.0451) grad_norm 0.9978 (inf) [2021-04-15 17:45:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][810/1251] eta 0:02:06 lr 0.000956 time 0.2969 (0.2872) loss 4.3090 (4.0461) grad_norm 0.9867 (inf) [2021-04-15 17:45:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][820/1251] eta 0:02:03 lr 0.000956 time 0.2794 (0.2871) loss 3.5579 (4.0445) grad_norm 1.0791 (inf) [2021-04-15 17:45:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][830/1251] eta 0:02:00 lr 0.000956 time 0.2718 (0.2870) loss 4.9297 (4.0457) grad_norm 0.9779 (inf) [2021-04-15 17:45:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][840/1251] eta 0:01:57 lr 0.000956 time 0.2632 (0.2869) loss 4.5787 (4.0465) grad_norm 1.0054 (inf) [2021-04-15 17:45:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][850/1251] eta 0:01:55 lr 0.000956 time 0.2566 (0.2869) loss 3.3006 (4.0490) grad_norm 0.9182 (inf) [2021-04-15 17:45:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][860/1251] eta 0:01:52 lr 0.000956 time 0.2790 (0.2868) loss 3.4097 (4.0486) grad_norm 1.0321 (inf) [2021-04-15 17:45:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][870/1251] eta 0:01:49 lr 0.000956 time 0.2740 (0.2866) loss 4.3068 (4.0508) grad_norm 1.0670 (inf) [2021-04-15 17:45:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][880/1251] eta 0:01:46 lr 0.000956 time 0.2925 (0.2866) loss 4.4429 (4.0526) grad_norm 1.1171 (inf) [2021-04-15 17:46:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][890/1251] eta 0:01:43 lr 0.000956 time 0.2894 (0.2866) loss 4.2024 (4.0540) grad_norm 1.2574 (inf) [2021-04-15 17:46:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][900/1251] eta 0:01:40 lr 0.000956 time 0.2733 (0.2865) loss 4.3045 (4.0551) grad_norm 1.2113 (inf) [2021-04-15 17:46:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][910/1251] eta 0:01:37 lr 0.000956 time 0.2478 (0.2864) loss 5.0867 (4.0566) grad_norm 1.2339 (inf) [2021-04-15 17:46:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][920/1251] eta 0:01:34 lr 0.000956 time 0.2849 (0.2863) loss 3.9832 (4.0574) grad_norm 1.1257 (inf) [2021-04-15 17:46:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][930/1251] eta 0:01:31 lr 0.000956 time 0.2742 (0.2864) loss 4.1136 (4.0564) grad_norm 1.1630 (inf) [2021-04-15 17:46:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][940/1251] eta 0:01:29 lr 0.000956 time 0.2934 (0.2864) loss 4.1596 (4.0590) grad_norm 0.9069 (inf) [2021-04-15 17:46:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][950/1251] eta 0:01:26 lr 0.000956 time 0.2748 (0.2864) loss 3.7392 (4.0569) grad_norm 0.9607 (inf) [2021-04-15 17:46:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][960/1251] eta 0:01:23 lr 0.000956 time 0.2656 (0.2864) loss 4.9449 (4.0548) grad_norm 1.0646 (inf) [2021-04-15 17:46:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][970/1251] eta 0:01:20 lr 0.000956 time 0.2898 (0.2864) loss 4.4647 (4.0558) grad_norm 1.3060 (inf) [2021-04-15 17:46:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][980/1251] eta 0:01:17 lr 0.000956 time 0.2829 (0.2864) loss 4.4749 (4.0563) grad_norm 1.1515 (inf) [2021-04-15 17:46:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][990/1251] eta 0:01:14 lr 0.000956 time 0.2801 (0.2863) loss 4.7755 (4.0571) grad_norm 1.0254 (inf) [2021-04-15 17:46:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][1000/1251] eta 0:01:11 lr 0.000956 time 0.2798 (0.2862) loss 3.5330 (4.0564) grad_norm 1.2153 (inf) [2021-04-15 17:46:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][1010/1251] eta 0:01:08 lr 0.000955 time 0.2848 (0.2862) loss 3.1934 (4.0570) grad_norm 0.9898 (inf) [2021-04-15 17:46:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][1020/1251] eta 0:01:06 lr 0.000955 time 0.2751 (0.2861) loss 4.2235 (4.0551) grad_norm 0.9437 (inf) [2021-04-15 17:46:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][1030/1251] eta 0:01:03 lr 0.000955 time 0.3208 (0.2861) loss 4.5275 (4.0563) grad_norm 0.9279 (inf) [2021-04-15 17:46:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][1040/1251] eta 0:01:00 lr 0.000955 time 0.2809 (0.2859) loss 4.6029 (4.0572) grad_norm 1.1077 (inf) [2021-04-15 17:46:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][1050/1251] eta 0:00:57 lr 0.000955 time 0.2941 (0.2859) loss 2.8071 (4.0573) grad_norm 1.2161 (inf) [2021-04-15 17:46:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][1060/1251] eta 0:00:54 lr 0.000955 time 0.2684 (0.2858) loss 4.6016 (4.0613) grad_norm 0.9368 (inf) [2021-04-15 17:46:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][1070/1251] eta 0:00:51 lr 0.000955 time 0.2924 (0.2857) loss 3.0444 (4.0626) grad_norm 1.0148 (inf) [2021-04-15 17:46:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][1080/1251] eta 0:00:48 lr 0.000955 time 0.2966 (0.2856) loss 3.3144 (4.0621) grad_norm 1.0174 (inf) [2021-04-15 17:46:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][1090/1251] eta 0:00:45 lr 0.000955 time 0.2912 (0.2856) loss 3.2416 (4.0600) grad_norm 0.9921 (inf) [2021-04-15 17:47:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][1100/1251] eta 0:00:43 lr 0.000955 time 0.2898 (0.2855) loss 3.4584 (4.0600) grad_norm 0.9673 (inf) [2021-04-15 17:47:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][1110/1251] eta 0:00:40 lr 0.000955 time 0.2697 (0.2855) loss 3.8981 (4.0603) grad_norm 1.0976 (inf) [2021-04-15 17:47:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][1120/1251] eta 0:00:37 lr 0.000955 time 0.3447 (0.2854) loss 3.7802 (4.0597) grad_norm 1.0646 (inf) [2021-04-15 17:47:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][1130/1251] eta 0:00:34 lr 0.000955 time 0.2604 (0.2853) loss 4.2555 (4.0608) grad_norm 0.9200 (inf) [2021-04-15 17:47:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][1140/1251] eta 0:00:31 lr 0.000955 time 0.2568 (0.2853) loss 3.6815 (4.0596) grad_norm 1.0907 (inf) [2021-04-15 17:47:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][1150/1251] eta 0:00:28 lr 0.000955 time 0.2695 (0.2853) loss 3.6591 (4.0611) grad_norm 1.1173 (inf) [2021-04-15 17:47:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][1160/1251] eta 0:00:25 lr 0.000955 time 0.2582 (0.2852) loss 4.2111 (4.0611) grad_norm 0.8819 (inf) [2021-04-15 17:47:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][1170/1251] eta 0:00:23 lr 0.000955 time 0.2934 (0.2852) loss 4.4351 (4.0638) grad_norm 0.8485 (inf) [2021-04-15 17:47:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][1180/1251] eta 0:00:20 lr 0.000955 time 0.2806 (0.2851) loss 3.5396 (4.0604) grad_norm 1.0071 (inf) [2021-04-15 17:47:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][1190/1251] eta 0:00:17 lr 0.000955 time 0.2887 (0.2850) loss 3.3657 (4.0618) grad_norm 1.0359 (inf) [2021-04-15 17:47:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][1200/1251] eta 0:00:14 lr 0.000955 time 0.2841 (0.2850) loss 5.0380 (4.0627) grad_norm 1.0773 (inf) [2021-04-15 17:47:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][1210/1251] eta 0:00:11 lr 0.000955 time 0.3117 (0.2850) loss 4.5888 (4.0610) grad_norm 1.0819 (inf) [2021-04-15 17:47:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][1220/1251] eta 0:00:08 lr 0.000955 time 0.2700 (0.2851) loss 4.0650 (4.0602) grad_norm 1.0233 (inf) [2021-04-15 17:47:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][1230/1251] eta 0:00:05 lr 0.000955 time 0.2598 (0.2850) loss 4.1874 (4.0591) grad_norm 0.9027 (inf) [2021-04-15 17:47:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][1240/1251] eta 0:00:03 lr 0.000955 time 0.2495 (0.2849) loss 3.8933 (4.0571) grad_norm 1.0616 (inf) [2021-04-15 17:47:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [40/300][1250/1251] eta 0:00:00 lr 0.000955 time 0.2488 (0.2846) loss 4.3914 (4.0567) grad_norm 1.1302 (inf) [2021-04-15 17:47:45 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 40 training takes 0:05:58 [2021-04-15 17:47:45 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_40.pth saving...... [2021-04-15 17:47:54 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_40.pth saved !!! [2021-04-15 17:47:55 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.109 (1.109) Loss 1.3088 (1.3088) Acc@1 69.629 (69.629) Acc@5 90.332 (90.332) [2021-04-15 17:47:57 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.132 (0.221) Loss 1.3673 (1.3547) Acc@1 68.066 (68.439) Acc@5 89.648 (89.560) [2021-04-15 17:47:59 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.170 (0.223) Loss 1.3564 (1.3513) Acc@1 67.285 (68.727) Acc@5 89.355 (89.500) [2021-04-15 17:48:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.112 (0.244) Loss 1.3388 (1.3504) Acc@1 68.945 (68.961) Acc@5 90.039 (89.418) [2021-04-15 17:48:03 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.082 (0.211) Loss 1.3768 (1.3585) Acc@1 67.969 (68.843) Acc@5 89.844 (89.348) [2021-04-15 17:48:06 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 68.740 Acc@5 89.240 [2021-04-15 17:48:06 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 68.7% [2021-04-15 17:48:06 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 68.74% [2021-04-15 17:48:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][0/1251] eta 1:41:25 lr 0.000955 time 4.8642 (4.8642) loss 3.9121 (3.9121) grad_norm 1.0343 (1.0343) [2021-04-15 17:48:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][10/1251] eta 0:14:21 lr 0.000955 time 0.2602 (0.6943) loss 3.7408 (3.8532) grad_norm 1.4307 (1.0823) [2021-04-15 17:48:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][20/1251] eta 0:10:12 lr 0.000955 time 0.2888 (0.4978) loss 3.9776 (3.9498) grad_norm 1.2293 (1.1423) [2021-04-15 17:48:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][30/1251] eta 0:08:42 lr 0.000955 time 0.2837 (0.4279) loss 4.5837 (3.8715) grad_norm 0.9582 (1.1502) [2021-04-15 17:48:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][40/1251] eta 0:07:54 lr 0.000955 time 0.2736 (0.3917) loss 4.9505 (3.9807) grad_norm 0.9925 (1.1334) [2021-04-15 17:48:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][50/1251] eta 0:07:25 lr 0.000955 time 0.3109 (0.3706) loss 4.6046 (3.9996) grad_norm 0.9593 (1.1183) [2021-04-15 17:48:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][60/1251] eta 0:07:04 lr 0.000955 time 0.2624 (0.3560) loss 4.0266 (4.0080) grad_norm 1.0158 (1.1130) [2021-04-15 17:48:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][70/1251] eta 0:06:47 lr 0.000955 time 0.2668 (0.3455) loss 4.6491 (4.0095) grad_norm 1.4741 (1.1176) [2021-04-15 17:48:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][80/1251] eta 0:06:35 lr 0.000955 time 0.2814 (0.3377) loss 4.1591 (4.0051) grad_norm 1.4391 (1.1408) [2021-04-15 17:48:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][90/1251] eta 0:06:24 lr 0.000955 time 0.2996 (0.3313) loss 3.4870 (4.0030) grad_norm 1.1358 (1.1365) [2021-04-15 17:48:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][100/1251] eta 0:06:15 lr 0.000955 time 0.2666 (0.3260) loss 4.9295 (3.9942) grad_norm 1.0524 (1.1393) [2021-04-15 17:48:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][110/1251] eta 0:06:06 lr 0.000955 time 0.2457 (0.3216) loss 4.5352 (4.0136) grad_norm 1.0960 (1.1335) [2021-04-15 17:48:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][120/1251] eta 0:06:00 lr 0.000955 time 0.2710 (0.3184) loss 3.4652 (4.0223) grad_norm 0.9620 (1.1340) [2021-04-15 17:48:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][130/1251] eta 0:05:53 lr 0.000955 time 0.2852 (0.3152) loss 4.0255 (4.0097) grad_norm 0.9153 (1.1312) [2021-04-15 17:48:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][140/1251] eta 0:05:49 lr 0.000955 time 0.2729 (0.3144) loss 3.6914 (4.0022) grad_norm 1.0349 (1.1315) [2021-04-15 17:48:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][150/1251] eta 0:05:43 lr 0.000955 time 0.3072 (0.3122) loss 3.0356 (3.9993) grad_norm 1.5689 (1.1354) [2021-04-15 17:48:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][160/1251] eta 0:05:38 lr 0.000955 time 0.2797 (0.3099) loss 2.7452 (4.0053) grad_norm 1.1978 (1.1310) [2021-04-15 17:48:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][170/1251] eta 0:05:33 lr 0.000955 time 0.2523 (0.3089) loss 4.2275 (3.9903) grad_norm 1.0417 (1.1250) [2021-04-15 17:49:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][180/1251] eta 0:05:30 lr 0.000955 time 0.2687 (0.3081) loss 4.3338 (3.9765) grad_norm 1.0091 (1.1191) [2021-04-15 17:49:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][190/1251] eta 0:05:25 lr 0.000955 time 0.2737 (0.3066) loss 4.0666 (3.9932) grad_norm 0.9461 (1.1133) [2021-04-15 17:49:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][200/1251] eta 0:05:20 lr 0.000955 time 0.2511 (0.3051) loss 4.4562 (4.0046) grad_norm 0.9922 (1.1108) [2021-04-15 17:49:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][210/1251] eta 0:05:16 lr 0.000955 time 0.2689 (0.3037) loss 4.0511 (4.0188) grad_norm 1.1069 (1.1069) [2021-04-15 17:49:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][220/1251] eta 0:05:12 lr 0.000955 time 0.2796 (0.3027) loss 4.2264 (4.0154) grad_norm 1.0751 (1.1050) [2021-04-15 17:49:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][230/1251] eta 0:05:07 lr 0.000955 time 0.2887 (0.3016) loss 4.6789 (4.0129) grad_norm 1.2263 (1.1062) [2021-04-15 17:49:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][240/1251] eta 0:05:04 lr 0.000955 time 0.2936 (0.3008) loss 4.3481 (4.0109) grad_norm 1.0166 (1.1063) [2021-04-15 17:49:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][250/1251] eta 0:05:00 lr 0.000955 time 0.2846 (0.3001) loss 4.1280 (4.0207) grad_norm 1.0939 (1.1057) [2021-04-15 17:49:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][260/1251] eta 0:04:56 lr 0.000955 time 0.2648 (0.2992) loss 4.5871 (4.0218) grad_norm 0.9640 (1.1042) [2021-04-15 17:49:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][270/1251] eta 0:04:52 lr 0.000955 time 0.3040 (0.2985) loss 4.4021 (4.0216) grad_norm 0.9721 (1.1020) [2021-04-15 17:49:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][280/1251] eta 0:04:49 lr 0.000955 time 0.2948 (0.2980) loss 3.8905 (4.0127) grad_norm 1.3011 (1.1055) [2021-04-15 17:49:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][290/1251] eta 0:04:46 lr 0.000955 time 0.2695 (0.2979) loss 4.4111 (4.0220) grad_norm 1.0138 (1.1065) [2021-04-15 17:49:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][300/1251] eta 0:04:42 lr 0.000955 time 0.2918 (0.2971) loss 4.9073 (4.0330) grad_norm 0.9630 (1.1071) [2021-04-15 17:49:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][310/1251] eta 0:04:39 lr 0.000955 time 0.2688 (0.2965) loss 4.8229 (4.0357) grad_norm 0.9599 (1.1085) [2021-04-15 17:49:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][320/1251] eta 0:04:35 lr 0.000955 time 0.2755 (0.2959) loss 4.2571 (4.0371) grad_norm 1.2078 (1.1067) [2021-04-15 17:49:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][330/1251] eta 0:04:31 lr 0.000955 time 0.2773 (0.2952) loss 4.7293 (4.0479) grad_norm 1.2146 (1.1053) [2021-04-15 17:49:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][340/1251] eta 0:04:28 lr 0.000954 time 0.2647 (0.2947) loss 3.1430 (4.0526) grad_norm 1.0401 (1.1047) [2021-04-15 17:49:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][350/1251] eta 0:04:24 lr 0.000954 time 0.2695 (0.2941) loss 3.3177 (4.0477) grad_norm 0.8938 (1.1033) [2021-04-15 17:49:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][360/1251] eta 0:04:21 lr 0.000954 time 0.3102 (0.2937) loss 3.8643 (4.0437) grad_norm 1.4870 (1.1035) [2021-04-15 17:49:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][370/1251] eta 0:04:18 lr 0.000954 time 0.2679 (0.2932) loss 4.0924 (4.0360) grad_norm 1.0199 (1.1037) [2021-04-15 17:49:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][380/1251] eta 0:04:14 lr 0.000954 time 0.2914 (0.2928) loss 3.1798 (4.0305) grad_norm 1.0696 (1.1025) [2021-04-15 17:50:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][390/1251] eta 0:04:11 lr 0.000954 time 0.2825 (0.2924) loss 3.3766 (4.0323) grad_norm 0.9951 (1.1001) [2021-04-15 17:50:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][400/1251] eta 0:04:08 lr 0.000954 time 0.2683 (0.2924) loss 4.2778 (4.0281) grad_norm 1.1965 (1.0992) [2021-04-15 17:50:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][410/1251] eta 0:04:05 lr 0.000954 time 0.2721 (0.2920) loss 4.5198 (4.0303) grad_norm 1.0929 (1.0988) [2021-04-15 17:50:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][420/1251] eta 0:04:02 lr 0.000954 time 0.2847 (0.2918) loss 4.4019 (4.0319) grad_norm 0.8914 (1.0968) [2021-04-15 17:50:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][430/1251] eta 0:03:59 lr 0.000954 time 0.2540 (0.2915) loss 4.5641 (4.0283) grad_norm 1.3671 (1.0977) [2021-04-15 17:50:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][440/1251] eta 0:03:56 lr 0.000954 time 0.2823 (0.2911) loss 3.9319 (4.0245) grad_norm 1.0587 (1.0981) [2021-04-15 17:50:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][450/1251] eta 0:03:53 lr 0.000954 time 0.2589 (0.2910) loss 3.9016 (4.0231) grad_norm 0.8504 (1.0963) [2021-04-15 17:50:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][460/1251] eta 0:03:49 lr 0.000954 time 0.3122 (0.2908) loss 4.5774 (4.0150) grad_norm 0.8727 (1.0939) [2021-04-15 17:50:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][470/1251] eta 0:03:46 lr 0.000954 time 0.2803 (0.2904) loss 4.6965 (4.0167) grad_norm 0.9668 (1.0920) [2021-04-15 17:50:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][480/1251] eta 0:03:43 lr 0.000954 time 0.2863 (0.2902) loss 4.2107 (4.0151) grad_norm 1.1452 (1.0913) [2021-04-15 17:50:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][490/1251] eta 0:03:40 lr 0.000954 time 0.2865 (0.2901) loss 4.4319 (4.0136) grad_norm 1.0995 (1.0917) [2021-04-15 17:50:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][500/1251] eta 0:03:37 lr 0.000954 time 0.2686 (0.2899) loss 4.6846 (4.0204) grad_norm 1.0572 (1.0921) [2021-04-15 17:50:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][510/1251] eta 0:03:34 lr 0.000954 time 0.2824 (0.2896) loss 4.3354 (4.0187) grad_norm 1.0323 (1.0914) [2021-04-15 17:50:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][520/1251] eta 0:03:31 lr 0.000954 time 0.2702 (0.2896) loss 4.0029 (4.0208) grad_norm 1.0098 (1.0907) [2021-04-15 17:50:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][530/1251] eta 0:03:28 lr 0.000954 time 0.2603 (0.2895) loss 4.3240 (4.0239) grad_norm 0.9484 (1.0894) [2021-04-15 17:50:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][540/1251] eta 0:03:25 lr 0.000954 time 0.2645 (0.2896) loss 4.1977 (4.0245) grad_norm 0.9778 (1.0882) [2021-04-15 17:50:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][550/1251] eta 0:03:22 lr 0.000954 time 0.2704 (0.2894) loss 4.0598 (4.0221) grad_norm 1.0428 (1.0880) [2021-04-15 17:50:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][560/1251] eta 0:03:19 lr 0.000954 time 0.2735 (0.2892) loss 4.0094 (4.0253) grad_norm 1.0651 (1.0881) [2021-04-15 17:50:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][570/1251] eta 0:03:16 lr 0.000954 time 0.2695 (0.2890) loss 3.4707 (4.0255) grad_norm 1.1325 (1.0895) [2021-04-15 17:50:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][580/1251] eta 0:03:13 lr 0.000954 time 0.2753 (0.2889) loss 4.0246 (4.0260) grad_norm 1.0127 (1.0882) [2021-04-15 17:50:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][590/1251] eta 0:03:10 lr 0.000954 time 0.2772 (0.2889) loss 4.1316 (4.0302) grad_norm 1.3125 (1.0880) [2021-04-15 17:50:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][600/1251] eta 0:03:07 lr 0.000954 time 0.2690 (0.2887) loss 4.4278 (4.0319) grad_norm 0.9406 (1.0880) [2021-04-15 17:51:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][610/1251] eta 0:03:04 lr 0.000954 time 0.2683 (0.2886) loss 3.0624 (4.0314) grad_norm 0.9764 (1.0884) [2021-04-15 17:51:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][620/1251] eta 0:03:01 lr 0.000954 time 0.2632 (0.2883) loss 2.3960 (4.0233) grad_norm 1.0827 (1.0882) [2021-04-15 17:51:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][630/1251] eta 0:02:58 lr 0.000954 time 0.2552 (0.2881) loss 4.0914 (4.0220) grad_norm 1.1109 (1.0885) [2021-04-15 17:51:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][640/1251] eta 0:02:55 lr 0.000954 time 0.2714 (0.2880) loss 4.4424 (4.0249) grad_norm 0.9157 (1.0878) [2021-04-15 17:51:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][650/1251] eta 0:02:53 lr 0.000954 time 0.2956 (0.2879) loss 4.2849 (4.0236) grad_norm 1.1473 (1.0880) [2021-04-15 17:51:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][660/1251] eta 0:02:50 lr 0.000954 time 0.2747 (0.2877) loss 4.4609 (4.0228) grad_norm 1.0009 (1.0879) [2021-04-15 17:51:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][670/1251] eta 0:02:47 lr 0.000954 time 0.2688 (0.2875) loss 4.2188 (4.0222) grad_norm 1.1212 (1.0877) [2021-04-15 17:51:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][680/1251] eta 0:02:44 lr 0.000954 time 0.2843 (0.2874) loss 3.6479 (4.0239) grad_norm 1.1393 (1.0879) [2021-04-15 17:51:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][690/1251] eta 0:02:41 lr 0.000954 time 0.2740 (0.2873) loss 4.0016 (4.0222) grad_norm 0.8626 (1.0870) [2021-04-15 17:51:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][700/1251] eta 0:02:38 lr 0.000954 time 0.2808 (0.2872) loss 4.4230 (4.0204) grad_norm 1.1388 (1.0866) [2021-04-15 17:51:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][710/1251] eta 0:02:35 lr 0.000954 time 0.2828 (0.2871) loss 4.3860 (4.0217) grad_norm 0.9833 (1.0874) [2021-04-15 17:51:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][720/1251] eta 0:02:32 lr 0.000954 time 0.2631 (0.2871) loss 4.3796 (4.0256) grad_norm 1.1630 (1.0875) [2021-04-15 17:51:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][730/1251] eta 0:02:29 lr 0.000954 time 0.2732 (0.2869) loss 3.7683 (4.0230) grad_norm 1.5962 (1.0882) [2021-04-15 17:51:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][740/1251] eta 0:02:26 lr 0.000954 time 0.2759 (0.2869) loss 4.6624 (4.0255) grad_norm 1.0391 (1.0907) [2021-04-15 17:51:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][750/1251] eta 0:02:23 lr 0.000954 time 0.2828 (0.2868) loss 3.4548 (4.0255) grad_norm 0.9604 (1.0903) [2021-04-15 17:51:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][760/1251] eta 0:02:20 lr 0.000954 time 0.2793 (0.2869) loss 3.9696 (4.0291) grad_norm 1.1967 (1.0903) [2021-04-15 17:51:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][770/1251] eta 0:02:17 lr 0.000954 time 0.2975 (0.2869) loss 3.6906 (4.0307) grad_norm 0.9702 (1.0896) [2021-04-15 17:51:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][780/1251] eta 0:02:15 lr 0.000954 time 0.2699 (0.2868) loss 4.7744 (4.0328) grad_norm 1.0346 (1.0903) [2021-04-15 17:51:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][790/1251] eta 0:02:12 lr 0.000954 time 0.2776 (0.2867) loss 4.5652 (4.0350) grad_norm 1.0395 (1.0898) [2021-04-15 17:51:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][800/1251] eta 0:02:09 lr 0.000954 time 0.2734 (0.2867) loss 5.0153 (4.0380) grad_norm 1.1351 (1.0905) [2021-04-15 17:51:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][810/1251] eta 0:02:06 lr 0.000954 time 0.3016 (0.2866) loss 4.1969 (4.0390) grad_norm 1.3594 (1.0909) [2021-04-15 17:52:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][820/1251] eta 0:02:03 lr 0.000954 time 0.2775 (0.2865) loss 4.5857 (4.0398) grad_norm 1.0948 (1.0907) [2021-04-15 17:52:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][830/1251] eta 0:02:00 lr 0.000954 time 0.2633 (0.2864) loss 4.0353 (4.0421) grad_norm 1.1358 (1.0909) [2021-04-15 17:52:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][840/1251] eta 0:01:57 lr 0.000954 time 0.2589 (0.2863) loss 2.8981 (4.0387) grad_norm 0.9723 (1.0897) [2021-04-15 17:52:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][850/1251] eta 0:01:54 lr 0.000954 time 0.2733 (0.2861) loss 4.2364 (4.0397) grad_norm 1.2232 (1.0900) [2021-04-15 17:52:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][860/1251] eta 0:01:51 lr 0.000954 time 0.3025 (0.2861) loss 3.9975 (4.0418) grad_norm 1.2653 (1.0895) [2021-04-15 17:52:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][870/1251] eta 0:01:48 lr 0.000954 time 0.2891 (0.2859) loss 3.4406 (4.0400) grad_norm 1.2729 (1.0901) [2021-04-15 17:52:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][880/1251] eta 0:01:46 lr 0.000954 time 0.2852 (0.2858) loss 4.3205 (4.0379) grad_norm 1.0884 (1.0910) [2021-04-15 17:52:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][890/1251] eta 0:01:43 lr 0.000954 time 0.2885 (0.2858) loss 4.4343 (4.0348) grad_norm 0.9971 (1.0903) [2021-04-15 17:52:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][900/1251] eta 0:01:40 lr 0.000954 time 0.2772 (0.2856) loss 4.3537 (4.0335) grad_norm 1.0115 (1.0896) [2021-04-15 17:52:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][910/1251] eta 0:01:37 lr 0.000953 time 0.2768 (0.2856) loss 3.6113 (4.0291) grad_norm 0.9456 (1.0889) [2021-04-15 17:52:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][920/1251] eta 0:01:34 lr 0.000953 time 0.2582 (0.2855) loss 4.1414 (4.0288) grad_norm 1.0602 (1.0900) [2021-04-15 17:52:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][930/1251] eta 0:01:31 lr 0.000953 time 0.2483 (0.2854) loss 3.3935 (4.0283) grad_norm 1.0723 (1.0898) [2021-04-15 17:52:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][940/1251] eta 0:01:28 lr 0.000953 time 0.2516 (0.2853) loss 4.0974 (4.0302) grad_norm 0.8494 (1.0891) [2021-04-15 17:52:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][950/1251] eta 0:01:25 lr 0.000953 time 0.2749 (0.2852) loss 4.4848 (4.0347) grad_norm 1.0831 (1.0883) [2021-04-15 17:52:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][960/1251] eta 0:01:22 lr 0.000953 time 0.2532 (0.2851) loss 3.2398 (4.0362) grad_norm 1.2054 (1.0876) [2021-04-15 17:52:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][970/1251] eta 0:01:20 lr 0.000953 time 0.2551 (0.2851) loss 4.5434 (4.0361) grad_norm 1.1342 (1.0883) [2021-04-15 17:52:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][980/1251] eta 0:01:17 lr 0.000953 time 0.2608 (0.2851) loss 3.2088 (4.0384) grad_norm 1.0881 (1.0881) [2021-04-15 17:52:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][990/1251] eta 0:01:14 lr 0.000953 time 0.2685 (0.2850) loss 4.2733 (4.0403) grad_norm 1.3658 (1.0881) [2021-04-15 17:52:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][1000/1251] eta 0:01:11 lr 0.000953 time 0.2689 (0.2850) loss 4.4281 (4.0410) grad_norm 0.8872 (1.0889) [2021-04-15 17:52:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][1010/1251] eta 0:01:08 lr 0.000953 time 0.2792 (0.2849) loss 4.1260 (4.0413) grad_norm 1.0711 (1.0892) [2021-04-15 17:52:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][1020/1251] eta 0:01:05 lr 0.000953 time 0.2945 (0.2849) loss 3.4305 (4.0419) grad_norm 0.9917 (1.0895) [2021-04-15 17:52:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][1030/1251] eta 0:01:02 lr 0.000953 time 0.3103 (0.2849) loss 4.3483 (4.0416) grad_norm 1.0483 (1.0890) [2021-04-15 17:53:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][1040/1251] eta 0:01:00 lr 0.000953 time 0.2691 (0.2848) loss 3.5342 (4.0421) grad_norm 1.7357 (1.0906) [2021-04-15 17:53:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][1050/1251] eta 0:00:57 lr 0.000953 time 0.2736 (0.2847) loss 3.0290 (4.0417) grad_norm 1.0265 (1.0905) [2021-04-15 17:53:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][1060/1251] eta 0:00:54 lr 0.000953 time 0.2815 (0.2846) loss 4.1715 (4.0425) grad_norm 1.1656 (1.0909) [2021-04-15 17:53:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][1070/1251] eta 0:00:51 lr 0.000953 time 0.2645 (0.2845) loss 4.2453 (4.0446) grad_norm 1.2293 (1.0916) [2021-04-15 17:53:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][1080/1251] eta 0:00:48 lr 0.000953 time 0.2563 (0.2845) loss 4.7478 (4.0447) grad_norm 1.0458 (1.0911) [2021-04-15 17:53:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][1090/1251] eta 0:00:45 lr 0.000953 time 0.2732 (0.2845) loss 4.2014 (4.0473) grad_norm 0.9350 (1.0907) [2021-04-15 17:53:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][1100/1251] eta 0:00:42 lr 0.000953 time 0.2845 (0.2845) loss 4.3585 (4.0475) grad_norm 1.1094 (1.0905) [2021-04-15 17:53:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][1110/1251] eta 0:00:40 lr 0.000953 time 0.2937 (0.2845) loss 3.6794 (4.0473) grad_norm 0.9462 (1.0898) [2021-04-15 17:53:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][1120/1251] eta 0:00:37 lr 0.000953 time 0.2717 (0.2846) loss 4.3652 (4.0449) grad_norm 1.1528 (1.0897) [2021-04-15 17:53:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][1130/1251] eta 0:00:34 lr 0.000953 time 0.2737 (0.2845) loss 4.3160 (4.0432) grad_norm 1.2120 (1.0906) [2021-04-15 17:53:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][1140/1251] eta 0:00:31 lr 0.000953 time 0.2708 (0.2845) loss 3.1496 (4.0423) grad_norm 1.1905 (1.0910) [2021-04-15 17:53:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][1150/1251] eta 0:00:28 lr 0.000953 time 0.2586 (0.2846) loss 2.4730 (4.0414) grad_norm 1.0179 (1.0916) [2021-04-15 17:53:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][1160/1251] eta 0:00:25 lr 0.000953 time 0.2684 (0.2845) loss 3.8641 (4.0399) grad_norm 1.3359 (1.0915) [2021-04-15 17:53:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][1170/1251] eta 0:00:23 lr 0.000953 time 0.2807 (0.2845) loss 4.3092 (4.0390) grad_norm 1.0440 (1.0918) [2021-04-15 17:53:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][1180/1251] eta 0:00:20 lr 0.000953 time 0.2647 (0.2844) loss 3.8960 (4.0390) grad_norm 1.1719 (1.0915) [2021-04-15 17:53:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][1190/1251] eta 0:00:17 lr 0.000953 time 0.2925 (0.2844) loss 4.1903 (4.0388) grad_norm 0.9804 (1.0912) [2021-04-15 17:53:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][1200/1251] eta 0:00:14 lr 0.000953 time 0.2839 (0.2843) loss 4.1969 (4.0362) grad_norm 1.2524 (1.0920) [2021-04-15 17:53:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][1210/1251] eta 0:00:11 lr 0.000953 time 0.2574 (0.2844) loss 4.0770 (4.0363) grad_norm 0.9695 (1.0911) [2021-04-15 17:53:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][1220/1251] eta 0:00:08 lr 0.000953 time 0.2926 (0.2843) loss 3.9389 (4.0391) grad_norm 1.0777 (1.0907) [2021-04-15 17:53:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][1230/1251] eta 0:00:05 lr 0.000953 time 0.2873 (0.2843) loss 3.6667 (4.0359) grad_norm 1.4195 (1.0910) [2021-04-15 17:53:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][1240/1251] eta 0:00:03 lr 0.000953 time 0.2598 (0.2842) loss 3.6994 (4.0341) grad_norm 0.9936 (1.0915) [2021-04-15 17:54:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [41/300][1250/1251] eta 0:00:00 lr 0.000953 time 0.2504 (0.2839) loss 4.4743 (4.0343) grad_norm 1.0257 (1.0913) [2021-04-15 17:54:03 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 41 training takes 0:05:57 [2021-04-15 17:54:03 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_41.pth saving...... [2021-04-15 17:54:13 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_41.pth saved !!! [2021-04-15 17:54:14 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.173 (1.173) Loss 1.3259 (1.3259) Acc@1 69.141 (69.141) Acc@5 89.355 (89.355) [2021-04-15 17:54:15 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.109 (0.214) Loss 1.2884 (1.3521) Acc@1 70.020 (68.324) Acc@5 89.648 (89.071) [2021-04-15 17:54:17 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.149 (0.218) Loss 1.3175 (1.3387) Acc@1 70.117 (68.797) Acc@5 89.355 (89.165) [2021-04-15 17:54:20 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.106 (0.239) Loss 1.2627 (1.3330) Acc@1 70.898 (68.889) Acc@5 90.625 (89.315) [2021-04-15 17:54:22 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.144 (0.221) Loss 1.4264 (1.3379) Acc@1 68.066 (68.862) Acc@5 87.695 (89.239) [2021-04-15 17:54:25 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 68.838 Acc@5 89.236 [2021-04-15 17:54:25 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 68.8% [2021-04-15 17:54:25 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 68.84% [2021-04-15 17:54:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][0/1251] eta 1:22:59 lr 0.000953 time 3.9801 (3.9801) loss 4.0029 (4.0029) grad_norm 0.9327 (0.9327) [2021-04-15 17:54:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][10/1251] eta 0:12:52 lr 0.000953 time 0.2950 (0.6223) loss 2.8092 (3.7479) grad_norm 0.9934 (1.0099) [2021-04-15 17:54:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][20/1251] eta 0:09:29 lr 0.000953 time 0.2816 (0.4629) loss 3.9194 (3.8527) grad_norm 1.1660 (1.0483) [2021-04-15 17:54:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][30/1251] eta 0:08:12 lr 0.000953 time 0.2679 (0.4034) loss 4.3839 (4.0340) grad_norm 1.0411 (1.0778) [2021-04-15 17:54:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][40/1251] eta 0:07:33 lr 0.000953 time 0.2788 (0.3742) loss 4.6247 (4.0561) grad_norm 1.1148 (1.0681) [2021-04-15 17:54:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][50/1251] eta 0:07:08 lr 0.000953 time 0.2895 (0.3567) loss 3.5150 (4.0268) grad_norm 1.0250 (1.0631) [2021-04-15 17:54:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][60/1251] eta 0:06:50 lr 0.000953 time 0.2807 (0.3444) loss 3.1017 (4.0376) grad_norm 0.9807 (1.0661) [2021-04-15 17:54:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][70/1251] eta 0:06:36 lr 0.000953 time 0.2681 (0.3357) loss 3.8593 (4.0553) grad_norm 1.1638 (1.0936) [2021-04-15 17:54:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][80/1251] eta 0:06:26 lr 0.000953 time 0.2806 (0.3297) loss 4.5932 (4.0768) grad_norm 1.0592 (1.0938) [2021-04-15 17:54:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][90/1251] eta 0:06:16 lr 0.000953 time 0.2797 (0.3246) loss 3.7854 (4.0976) grad_norm 1.0003 (1.0913) [2021-04-15 17:54:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][100/1251] eta 0:06:08 lr 0.000953 time 0.2803 (0.3202) loss 4.1340 (4.1041) grad_norm 0.9684 (1.0938) [2021-04-15 17:55:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][110/1251] eta 0:06:01 lr 0.000953 time 0.2689 (0.3168) loss 4.7306 (4.0828) grad_norm 0.9873 (1.0868) [2021-04-15 17:55:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][120/1251] eta 0:05:55 lr 0.000953 time 0.2787 (0.3141) loss 4.0850 (4.0926) grad_norm 1.1552 (1.0892) [2021-04-15 17:55:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][130/1251] eta 0:05:49 lr 0.000953 time 0.3091 (0.3119) loss 3.4688 (4.0680) grad_norm 1.1393 (1.0919) [2021-04-15 17:55:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][140/1251] eta 0:05:44 lr 0.000953 time 0.2980 (0.3098) loss 4.2232 (4.0559) grad_norm 1.1683 (1.0960) [2021-04-15 17:55:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][150/1251] eta 0:05:38 lr 0.000953 time 0.2695 (0.3078) loss 4.5251 (4.0591) grad_norm 0.9536 (1.0952) [2021-04-15 17:55:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][160/1251] eta 0:05:33 lr 0.000953 time 0.2856 (0.3060) loss 3.7743 (4.0615) grad_norm 0.9345 (1.0968) [2021-04-15 17:55:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][170/1251] eta 0:05:29 lr 0.000953 time 0.2722 (0.3045) loss 4.6110 (4.0500) grad_norm 1.1920 (1.0990) [2021-04-15 17:55:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][180/1251] eta 0:05:25 lr 0.000953 time 0.4716 (0.3042) loss 4.9120 (4.0236) grad_norm 0.9876 (1.1006) [2021-04-15 17:55:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][190/1251] eta 0:05:21 lr 0.000953 time 0.2719 (0.3026) loss 4.2062 (4.0300) grad_norm 0.9726 (1.0969) [2021-04-15 17:55:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][200/1251] eta 0:05:16 lr 0.000953 time 0.2629 (0.3016) loss 4.2915 (4.0308) grad_norm 0.9310 (1.0944) [2021-04-15 17:55:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][210/1251] eta 0:05:12 lr 0.000953 time 0.2938 (0.3005) loss 4.2886 (4.0255) grad_norm 1.8944 (1.0972) [2021-04-15 17:55:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][220/1251] eta 0:05:08 lr 0.000953 time 0.2880 (0.2995) loss 3.4612 (4.0173) grad_norm 1.0122 (1.0979) [2021-04-15 17:55:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][230/1251] eta 0:05:05 lr 0.000952 time 0.3006 (0.2988) loss 4.0221 (4.0231) grad_norm 1.0683 (1.0972) [2021-04-15 17:55:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][240/1251] eta 0:05:01 lr 0.000952 time 0.2912 (0.2979) loss 3.2319 (4.0176) grad_norm 0.8205 (1.0945) [2021-04-15 17:55:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][250/1251] eta 0:04:57 lr 0.000952 time 0.2564 (0.2971) loss 3.6952 (4.0182) grad_norm 1.1032 (1.0952) [2021-04-15 17:55:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][260/1251] eta 0:04:53 lr 0.000952 time 0.2865 (0.2964) loss 4.1733 (4.0229) grad_norm 1.1889 (1.0985) [2021-04-15 17:55:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][270/1251] eta 0:04:50 lr 0.000952 time 0.2874 (0.2958) loss 3.8073 (4.0193) grad_norm 1.1682 (1.1016) [2021-04-15 17:55:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][280/1251] eta 0:04:46 lr 0.000952 time 0.2700 (0.2953) loss 4.3318 (4.0157) grad_norm 0.9863 (1.1044) [2021-04-15 17:55:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][290/1251] eta 0:04:43 lr 0.000952 time 0.2735 (0.2950) loss 4.1796 (4.0269) grad_norm 0.9827 (1.1042) [2021-04-15 17:55:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][300/1251] eta 0:04:40 lr 0.000952 time 0.2845 (0.2950) loss 3.5137 (4.0310) grad_norm 1.1914 (1.1035) [2021-04-15 17:55:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][310/1251] eta 0:04:37 lr 0.000952 time 0.2590 (0.2944) loss 3.7674 (4.0181) grad_norm 1.0241 (1.1022) [2021-04-15 17:55:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][320/1251] eta 0:04:34 lr 0.000952 time 0.2764 (0.2944) loss 4.4722 (4.0282) grad_norm 0.9524 (1.1018) [2021-04-15 17:56:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][330/1251] eta 0:04:30 lr 0.000952 time 0.2991 (0.2941) loss 2.7742 (4.0198) grad_norm 1.2130 (1.1017) [2021-04-15 17:56:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][340/1251] eta 0:04:27 lr 0.000952 time 0.2820 (0.2937) loss 4.3104 (4.0163) grad_norm 1.2978 (1.1040) [2021-04-15 17:56:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][350/1251] eta 0:04:24 lr 0.000952 time 0.2713 (0.2934) loss 2.7993 (4.0186) grad_norm 1.2922 (1.1053) [2021-04-15 17:56:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][360/1251] eta 0:04:21 lr 0.000952 time 0.2792 (0.2934) loss 4.2009 (4.0220) grad_norm 0.9036 (1.1044) [2021-04-15 17:56:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][370/1251] eta 0:04:18 lr 0.000952 time 0.2663 (0.2931) loss 3.1966 (4.0189) grad_norm 1.1724 (1.1054) [2021-04-15 17:56:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][380/1251] eta 0:04:14 lr 0.000952 time 0.2635 (0.2927) loss 4.1788 (4.0201) grad_norm 1.0969 (1.1052) [2021-04-15 17:56:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][390/1251] eta 0:04:11 lr 0.000952 time 0.2812 (0.2924) loss 2.9425 (4.0240) grad_norm 1.0511 (1.1040) [2021-04-15 17:56:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][400/1251] eta 0:04:08 lr 0.000952 time 0.2995 (0.2921) loss 3.7397 (4.0115) grad_norm 0.9239 (1.1026) [2021-04-15 17:56:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][410/1251] eta 0:04:05 lr 0.000952 time 0.3007 (0.2918) loss 4.1094 (4.0203) grad_norm 0.9799 (1.1011) [2021-04-15 17:56:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][420/1251] eta 0:04:02 lr 0.000952 time 0.2850 (0.2916) loss 3.9422 (4.0074) grad_norm 1.0102 (1.1009) [2021-04-15 17:56:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][430/1251] eta 0:03:59 lr 0.000952 time 0.2710 (0.2913) loss 4.3761 (4.0026) grad_norm 0.9219 (1.1001) [2021-04-15 17:56:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][440/1251] eta 0:03:56 lr 0.000952 time 0.2680 (0.2910) loss 4.7514 (3.9989) grad_norm 1.1402 (1.1001) [2021-04-15 17:56:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][450/1251] eta 0:03:52 lr 0.000952 time 0.2766 (0.2907) loss 4.0256 (3.9893) grad_norm 1.3239 (1.1014) [2021-04-15 17:56:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][460/1251] eta 0:03:49 lr 0.000952 time 0.2757 (0.2905) loss 4.6537 (3.9926) grad_norm 1.0477 (1.1018) [2021-04-15 17:56:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][470/1251] eta 0:03:46 lr 0.000952 time 0.2789 (0.2903) loss 3.2690 (3.9914) grad_norm 1.4379 (1.1026) [2021-04-15 17:56:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][480/1251] eta 0:03:43 lr 0.000952 time 0.2778 (0.2901) loss 3.2836 (3.9918) grad_norm 1.2514 (1.1043) [2021-04-15 17:56:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][490/1251] eta 0:03:40 lr 0.000952 time 0.2678 (0.2898) loss 4.6087 (3.9898) grad_norm 0.8902 (1.1034) [2021-04-15 17:56:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][500/1251] eta 0:03:37 lr 0.000952 time 0.2686 (0.2896) loss 3.6804 (3.9948) grad_norm 1.0242 (1.1034) [2021-04-15 17:56:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][510/1251] eta 0:03:34 lr 0.000952 time 0.2493 (0.2895) loss 5.0539 (3.9977) grad_norm 1.1540 (1.1029) [2021-04-15 17:56:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][520/1251] eta 0:03:31 lr 0.000952 time 0.2797 (0.2893) loss 4.4069 (3.9983) grad_norm 1.0938 (1.1020) [2021-04-15 17:56:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][530/1251] eta 0:03:28 lr 0.000952 time 0.2815 (0.2892) loss 4.6325 (4.0055) grad_norm 1.0911 (1.1015) [2021-04-15 17:57:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][540/1251] eta 0:03:25 lr 0.000952 time 0.2906 (0.2891) loss 4.4936 (4.0043) grad_norm 1.0256 (1.1005) [2021-04-15 17:57:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][550/1251] eta 0:03:22 lr 0.000952 time 0.2648 (0.2888) loss 4.1396 (4.0087) grad_norm 0.9931 (1.1005) [2021-04-15 17:57:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][560/1251] eta 0:03:19 lr 0.000952 time 0.2826 (0.2886) loss 3.1520 (4.0065) grad_norm 0.9360 (1.0990) [2021-04-15 17:57:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][570/1251] eta 0:03:16 lr 0.000952 time 0.2620 (0.2886) loss 3.3688 (3.9984) grad_norm 1.0341 (1.0996) [2021-04-15 17:57:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][580/1251] eta 0:03:13 lr 0.000952 time 0.2730 (0.2888) loss 4.2984 (4.0032) grad_norm 0.9193 (1.0990) [2021-04-15 17:57:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][590/1251] eta 0:03:10 lr 0.000952 time 0.2773 (0.2887) loss 3.9892 (4.0071) grad_norm 1.1878 (1.0979) [2021-04-15 17:57:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][600/1251] eta 0:03:07 lr 0.000952 time 0.2923 (0.2886) loss 4.3198 (4.0129) grad_norm 1.0801 (1.0975) [2021-04-15 17:57:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][610/1251] eta 0:03:04 lr 0.000952 time 0.3002 (0.2885) loss 4.5726 (4.0128) grad_norm 1.1984 (1.0978) [2021-04-15 17:57:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][620/1251] eta 0:03:01 lr 0.000952 time 0.2688 (0.2883) loss 4.7953 (4.0124) grad_norm 1.2304 (1.0990) [2021-04-15 17:57:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][630/1251] eta 0:02:58 lr 0.000952 time 0.2678 (0.2881) loss 4.4649 (4.0145) grad_norm 0.9461 (1.0985) [2021-04-15 17:57:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][640/1251] eta 0:02:55 lr 0.000952 time 0.2604 (0.2880) loss 2.8318 (4.0140) grad_norm 0.9619 (1.0983) [2021-04-15 17:57:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][650/1251] eta 0:02:53 lr 0.000952 time 0.2658 (0.2879) loss 3.9731 (4.0110) grad_norm 1.5659 (1.0999) [2021-04-15 17:57:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][660/1251] eta 0:02:50 lr 0.000952 time 0.3107 (0.2878) loss 3.9058 (4.0097) grad_norm 1.0222 (1.1007) [2021-04-15 17:57:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][670/1251] eta 0:02:47 lr 0.000952 time 0.2904 (0.2878) loss 3.9169 (4.0114) grad_norm 1.1744 (1.1010) [2021-04-15 17:57:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][680/1251] eta 0:02:44 lr 0.000952 time 0.2602 (0.2877) loss 3.9507 (4.0120) grad_norm 1.2892 (inf) [2021-04-15 17:57:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][690/1251] eta 0:02:41 lr 0.000952 time 0.2864 (0.2875) loss 4.1355 (4.0109) grad_norm 1.1825 (inf) [2021-04-15 17:57:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][700/1251] eta 0:02:38 lr 0.000952 time 0.2801 (0.2874) loss 3.9442 (4.0099) grad_norm 1.3405 (inf) [2021-04-15 17:57:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][710/1251] eta 0:02:35 lr 0.000952 time 0.3046 (0.2874) loss 4.3114 (4.0136) grad_norm 1.0510 (inf) [2021-04-15 17:57:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][720/1251] eta 0:02:32 lr 0.000952 time 0.2744 (0.2872) loss 3.6855 (4.0133) grad_norm 1.1118 (inf) [2021-04-15 17:57:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][730/1251] eta 0:02:29 lr 0.000952 time 0.2915 (0.2872) loss 3.4315 (4.0125) grad_norm 1.2218 (inf) [2021-04-15 17:57:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][740/1251] eta 0:02:26 lr 0.000952 time 0.2835 (0.2871) loss 3.9662 (4.0119) grad_norm 1.0059 (inf) [2021-04-15 17:58:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][750/1251] eta 0:02:23 lr 0.000952 time 0.2784 (0.2870) loss 3.9418 (4.0124) grad_norm 1.2564 (inf) [2021-04-15 17:58:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][760/1251] eta 0:02:20 lr 0.000952 time 0.2487 (0.2870) loss 3.7572 (4.0129) grad_norm 1.2267 (inf) [2021-04-15 17:58:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][770/1251] eta 0:02:18 lr 0.000952 time 0.2863 (0.2869) loss 5.0018 (4.0118) grad_norm 1.0305 (inf) [2021-04-15 17:58:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][780/1251] eta 0:02:15 lr 0.000952 time 0.2778 (0.2867) loss 4.3461 (4.0134) grad_norm 1.0181 (inf) [2021-04-15 17:58:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][790/1251] eta 0:02:12 lr 0.000951 time 0.2683 (0.2867) loss 3.8280 (4.0123) grad_norm 1.1247 (inf) [2021-04-15 17:58:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][800/1251] eta 0:02:09 lr 0.000951 time 0.2878 (0.2867) loss 4.1151 (4.0125) grad_norm 1.0891 (inf) [2021-04-15 17:58:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][810/1251] eta 0:02:06 lr 0.000951 time 0.2837 (0.2866) loss 3.6477 (4.0121) grad_norm 1.0424 (inf) [2021-04-15 17:58:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][820/1251] eta 0:02:03 lr 0.000951 time 0.2796 (0.2865) loss 3.7755 (4.0089) grad_norm 1.0101 (inf) [2021-04-15 17:58:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][830/1251] eta 0:02:00 lr 0.000951 time 0.2684 (0.2863) loss 2.9510 (4.0078) grad_norm 1.1418 (inf) [2021-04-15 17:58:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][840/1251] eta 0:01:57 lr 0.000951 time 0.2859 (0.2862) loss 4.6995 (4.0054) grad_norm 0.9704 (inf) [2021-04-15 17:58:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][850/1251] eta 0:01:54 lr 0.000951 time 0.2657 (0.2861) loss 4.8218 (4.0038) grad_norm 1.1639 (inf) [2021-04-15 17:58:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][860/1251] eta 0:01:51 lr 0.000951 time 0.2727 (0.2860) loss 3.4338 (4.0040) grad_norm 1.2619 (inf) [2021-04-15 17:58:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][870/1251] eta 0:01:48 lr 0.000951 time 0.2599 (0.2859) loss 4.0608 (4.0009) grad_norm 1.2774 (inf) [2021-04-15 17:58:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][880/1251] eta 0:01:46 lr 0.000951 time 0.2837 (0.2861) loss 3.4889 (4.0003) grad_norm 1.0543 (inf) [2021-04-15 17:58:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][890/1251] eta 0:01:43 lr 0.000951 time 0.2659 (0.2861) loss 4.7014 (4.0033) grad_norm 0.9837 (inf) [2021-04-15 17:58:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][900/1251] eta 0:01:40 lr 0.000951 time 0.2885 (0.2861) loss 3.0392 (4.0025) grad_norm 1.0452 (inf) [2021-04-15 17:58:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][910/1251] eta 0:01:37 lr 0.000951 time 0.2729 (0.2860) loss 3.9031 (4.0039) grad_norm 1.1420 (inf) [2021-04-15 17:58:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][920/1251] eta 0:01:34 lr 0.000951 time 0.3132 (0.2860) loss 3.6570 (4.0074) grad_norm 0.8857 (inf) [2021-04-15 17:58:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][930/1251] eta 0:01:31 lr 0.000951 time 0.2815 (0.2860) loss 4.2440 (4.0125) grad_norm 1.0745 (inf) [2021-04-15 17:58:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][940/1251] eta 0:01:28 lr 0.000951 time 0.2638 (0.2860) loss 4.1562 (4.0135) grad_norm 1.0138 (inf) [2021-04-15 17:58:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][950/1251] eta 0:01:26 lr 0.000951 time 0.2836 (0.2860) loss 3.8512 (4.0135) grad_norm 0.9938 (inf) [2021-04-15 17:58:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][960/1251] eta 0:01:23 lr 0.000951 time 0.2877 (0.2859) loss 4.2208 (4.0126) grad_norm 0.9479 (inf) [2021-04-15 17:59:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][970/1251] eta 0:01:20 lr 0.000951 time 0.2851 (0.2858) loss 3.8026 (4.0113) grad_norm 1.0000 (inf) [2021-04-15 17:59:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][980/1251] eta 0:01:17 lr 0.000951 time 0.2939 (0.2858) loss 3.4233 (4.0125) grad_norm 1.0701 (inf) [2021-04-15 17:59:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][990/1251] eta 0:01:14 lr 0.000951 time 0.2625 (0.2857) loss 4.8439 (4.0149) grad_norm 1.1231 (inf) [2021-04-15 17:59:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][1000/1251] eta 0:01:11 lr 0.000951 time 0.2943 (0.2857) loss 4.9193 (4.0167) grad_norm 1.2664 (inf) [2021-04-15 17:59:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][1010/1251] eta 0:01:08 lr 0.000951 time 0.2642 (0.2856) loss 3.2662 (4.0151) grad_norm 0.9717 (inf) [2021-04-15 17:59:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][1020/1251] eta 0:01:05 lr 0.000951 time 0.2697 (0.2855) loss 3.3207 (4.0140) grad_norm 0.9878 (inf) [2021-04-15 17:59:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][1030/1251] eta 0:01:03 lr 0.000951 time 0.2717 (0.2854) loss 4.5186 (4.0135) grad_norm 1.0345 (inf) [2021-04-15 17:59:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][1040/1251] eta 0:01:00 lr 0.000951 time 0.2888 (0.2854) loss 3.6932 (4.0132) grad_norm 1.2660 (inf) [2021-04-15 17:59:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][1050/1251] eta 0:00:57 lr 0.000951 time 0.2904 (0.2854) loss 4.0055 (4.0120) grad_norm 1.0368 (inf) [2021-04-15 17:59:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][1060/1251] eta 0:00:54 lr 0.000951 time 0.2522 (0.2853) loss 4.5687 (4.0150) grad_norm 0.9505 (inf) [2021-04-15 17:59:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][1070/1251] eta 0:00:51 lr 0.000951 time 0.2783 (0.2852) loss 2.9943 (4.0136) grad_norm 1.2124 (inf) [2021-04-15 17:59:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][1080/1251] eta 0:00:48 lr 0.000951 time 0.2517 (0.2852) loss 3.2632 (4.0161) grad_norm 1.1703 (inf) [2021-04-15 17:59:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][1090/1251] eta 0:00:45 lr 0.000951 time 0.2804 (0.2851) loss 4.1449 (4.0179) grad_norm 1.0428 (inf) [2021-04-15 17:59:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][1100/1251] eta 0:00:43 lr 0.000951 time 0.2826 (0.2851) loss 3.1048 (4.0194) grad_norm 0.9707 (inf) [2021-04-15 17:59:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][1110/1251] eta 0:00:40 lr 0.000951 time 0.2755 (0.2850) loss 4.4048 (4.0173) grad_norm 1.1437 (inf) [2021-04-15 17:59:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][1120/1251] eta 0:00:37 lr 0.000951 time 0.2939 (0.2850) loss 4.3224 (4.0163) grad_norm 1.1274 (inf) [2021-04-15 17:59:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][1130/1251] eta 0:00:34 lr 0.000951 time 0.3026 (0.2849) loss 3.9339 (4.0165) grad_norm 1.2499 (inf) [2021-04-15 17:59:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][1140/1251] eta 0:00:31 lr 0.000951 time 0.2528 (0.2850) loss 2.9514 (4.0173) grad_norm 1.0190 (inf) [2021-04-15 17:59:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][1150/1251] eta 0:00:28 lr 0.000951 time 0.2686 (0.2849) loss 3.6087 (4.0178) grad_norm 1.3484 (inf) [2021-04-15 17:59:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][1160/1251] eta 0:00:25 lr 0.000951 time 0.2820 (0.2850) loss 3.8743 (4.0176) grad_norm 0.9627 (inf) [2021-04-15 17:59:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][1170/1251] eta 0:00:23 lr 0.000951 time 0.3160 (0.2850) loss 4.9325 (4.0195) grad_norm 1.0720 (inf) [2021-04-15 18:00:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][1180/1251] eta 0:00:20 lr 0.000951 time 0.2930 (0.2850) loss 4.0378 (4.0209) grad_norm 1.0212 (inf) [2021-04-15 18:00:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][1190/1251] eta 0:00:17 lr 0.000951 time 0.2774 (0.2848) loss 4.4564 (4.0228) grad_norm 1.2526 (inf) [2021-04-15 18:00:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][1200/1251] eta 0:00:14 lr 0.000951 time 0.2770 (0.2848) loss 3.1542 (4.0230) grad_norm 1.0934 (inf) [2021-04-15 18:00:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][1210/1251] eta 0:00:11 lr 0.000951 time 0.2747 (0.2848) loss 3.3609 (4.0207) grad_norm 1.4291 (inf) [2021-04-15 18:00:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][1220/1251] eta 0:00:08 lr 0.000951 time 0.2839 (0.2847) loss 4.3913 (4.0210) grad_norm 1.0818 (inf) [2021-04-15 18:00:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][1230/1251] eta 0:00:05 lr 0.000951 time 0.2841 (0.2846) loss 4.4964 (4.0206) grad_norm 1.4365 (inf) [2021-04-15 18:00:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][1240/1251] eta 0:00:03 lr 0.000951 time 0.2541 (0.2845) loss 5.0243 (4.0198) grad_norm 1.0126 (inf) [2021-04-15 18:00:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [42/300][1250/1251] eta 0:00:00 lr 0.000951 time 0.2485 (0.2842) loss 2.8922 (4.0176) grad_norm 1.2153 (inf) [2021-04-15 18:00:22 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 42 training takes 0:05:57 [2021-04-15 18:00:22 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_42.pth saving...... [2021-04-15 18:00:35 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_42.pth saved !!! [2021-04-15 18:00:36 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.128 (1.128) Loss 1.3866 (1.3866) Acc@1 68.945 (68.945) Acc@5 88.574 (88.574) [2021-04-15 18:00:38 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.191 (0.254) Loss 1.4867 (1.3749) Acc@1 67.578 (69.025) Acc@5 86.621 (89.178) [2021-04-15 18:00:40 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.137 (0.235) Loss 1.3381 (1.3694) Acc@1 70.996 (69.076) Acc@5 90.820 (89.504) [2021-04-15 18:00:42 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.126 (0.218) Loss 1.4266 (1.3792) Acc@1 67.676 (68.854) Acc@5 88.965 (89.327) [2021-04-15 18:00:44 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.216) Loss 1.4722 (1.3753) Acc@1 67.871 (68.798) Acc@5 87.891 (89.425) [2021-04-15 18:00:47 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 68.878 Acc@5 89.368 [2021-04-15 18:00:47 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 68.9% [2021-04-15 18:00:47 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 68.88% [2021-04-15 18:00:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][0/1251] eta 1:15:29 lr 0.000951 time 3.6208 (3.6208) loss 4.0248 (4.0248) grad_norm 1.4516 (1.4516) [2021-04-15 18:00:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][10/1251] eta 0:12:27 lr 0.000951 time 0.2994 (0.6026) loss 4.4035 (4.1594) grad_norm 1.1335 (1.0932) [2021-04-15 18:00:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][20/1251] eta 0:09:13 lr 0.000951 time 0.3173 (0.4496) loss 2.2716 (3.9165) grad_norm 1.0336 (1.1061) [2021-04-15 18:00:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][30/1251] eta 0:08:04 lr 0.000951 time 0.3088 (0.3967) loss 3.6746 (3.9283) grad_norm 1.0784 (1.0922) [2021-04-15 18:01:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][40/1251] eta 0:07:26 lr 0.000951 time 0.2584 (0.3690) loss 3.5693 (3.9560) grad_norm 1.0195 (1.0876) [2021-04-15 18:01:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][50/1251] eta 0:07:04 lr 0.000951 time 0.2941 (0.3536) loss 2.8710 (3.9004) grad_norm 0.9388 (1.0772) [2021-04-15 18:01:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][60/1251] eta 0:06:47 lr 0.000951 time 0.2764 (0.3419) loss 4.3354 (3.9548) grad_norm 1.0686 (1.0997) [2021-04-15 18:01:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][70/1251] eta 0:06:34 lr 0.000951 time 0.3144 (0.3339) loss 4.5641 (3.9649) grad_norm 0.9300 (1.0922) [2021-04-15 18:01:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][80/1251] eta 0:06:25 lr 0.000951 time 0.2876 (0.3290) loss 2.9314 (3.9438) grad_norm 0.8537 (1.0793) [2021-04-15 18:01:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][90/1251] eta 0:06:17 lr 0.000950 time 0.2847 (0.3248) loss 4.1151 (3.9801) grad_norm 1.0429 (1.0760) [2021-04-15 18:01:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][100/1251] eta 0:06:10 lr 0.000950 time 0.2584 (0.3223) loss 4.8542 (4.0160) grad_norm 1.0464 (1.0843) [2021-04-15 18:01:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][110/1251] eta 0:06:02 lr 0.000950 time 0.2602 (0.3181) loss 3.5045 (4.0486) grad_norm 1.3516 (1.0863) [2021-04-15 18:01:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][120/1251] eta 0:05:56 lr 0.000950 time 0.2822 (0.3154) loss 4.0379 (4.0479) grad_norm 1.0660 (1.0814) [2021-04-15 18:01:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][130/1251] eta 0:05:50 lr 0.000950 time 0.2580 (0.3128) loss 3.0831 (4.0548) grad_norm 0.9878 (1.0788) [2021-04-15 18:01:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][140/1251] eta 0:05:45 lr 0.000950 time 0.2772 (0.3108) loss 4.2879 (4.0692) grad_norm 1.4186 (1.0802) [2021-04-15 18:01:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][150/1251] eta 0:05:40 lr 0.000950 time 0.2711 (0.3096) loss 4.1452 (4.0686) grad_norm 1.1438 (1.0828) [2021-04-15 18:01:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][160/1251] eta 0:05:35 lr 0.000950 time 0.3078 (0.3079) loss 4.3199 (4.0789) grad_norm 1.1574 (1.0809) [2021-04-15 18:01:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][170/1251] eta 0:05:31 lr 0.000950 time 0.2979 (0.3064) loss 3.5246 (4.0826) grad_norm 1.2998 (1.0802) [2021-04-15 18:01:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][180/1251] eta 0:05:26 lr 0.000950 time 0.2766 (0.3048) loss 4.8828 (4.0914) grad_norm 1.1655 (1.0794) [2021-04-15 18:01:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][190/1251] eta 0:05:21 lr 0.000950 time 0.2885 (0.3034) loss 3.4719 (4.0873) grad_norm 1.0429 (1.0789) [2021-04-15 18:01:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][200/1251] eta 0:05:17 lr 0.000950 time 0.2878 (0.3024) loss 3.8218 (4.0840) grad_norm 1.0301 (1.0766) [2021-04-15 18:01:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][210/1251] eta 0:05:13 lr 0.000950 time 0.2629 (0.3011) loss 3.8431 (4.0807) grad_norm 1.4045 (1.0866) [2021-04-15 18:01:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][220/1251] eta 0:05:09 lr 0.000950 time 0.2595 (0.3002) loss 4.8673 (4.0929) grad_norm 1.1757 (1.0881) [2021-04-15 18:01:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][230/1251] eta 0:05:05 lr 0.000950 time 0.2900 (0.2993) loss 4.4839 (4.1004) grad_norm 1.0473 (1.0860) [2021-04-15 18:01:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][240/1251] eta 0:05:01 lr 0.000950 time 0.2776 (0.2984) loss 4.2636 (4.0956) grad_norm 1.2528 (1.0840) [2021-04-15 18:02:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][250/1251] eta 0:04:58 lr 0.000950 time 0.2928 (0.2977) loss 3.6477 (4.0853) grad_norm 1.1398 (1.0830) [2021-04-15 18:02:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][260/1251] eta 0:04:54 lr 0.000950 time 0.2774 (0.2970) loss 2.8759 (4.0682) grad_norm 1.0795 (1.0827) [2021-04-15 18:02:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][270/1251] eta 0:04:50 lr 0.000950 time 0.2759 (0.2964) loss 4.3938 (4.0671) grad_norm 1.5831 (1.0819) [2021-04-15 18:02:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][280/1251] eta 0:04:47 lr 0.000950 time 0.2778 (0.2956) loss 4.3178 (4.0702) grad_norm 1.0228 (1.0811) [2021-04-15 18:02:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][290/1251] eta 0:04:43 lr 0.000950 time 0.3031 (0.2951) loss 3.5640 (4.0648) grad_norm 0.9093 (1.0810) [2021-04-15 18:02:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][300/1251] eta 0:04:40 lr 0.000950 time 0.3048 (0.2947) loss 3.1357 (4.0635) grad_norm 1.1131 (1.0815) [2021-04-15 18:02:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][310/1251] eta 0:04:36 lr 0.000950 time 0.2762 (0.2942) loss 3.5698 (4.0610) grad_norm 1.0714 (1.0819) [2021-04-15 18:02:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][320/1251] eta 0:04:33 lr 0.000950 time 0.2677 (0.2939) loss 4.1700 (4.0568) grad_norm 1.0416 (1.0822) [2021-04-15 18:02:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][330/1251] eta 0:04:30 lr 0.000950 time 0.2651 (0.2934) loss 3.5293 (4.0414) grad_norm 0.9692 (1.0798) [2021-04-15 18:02:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][340/1251] eta 0:04:27 lr 0.000950 time 0.2954 (0.2936) loss 4.7291 (4.0421) grad_norm 1.0715 (1.0791) [2021-04-15 18:02:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][350/1251] eta 0:04:24 lr 0.000950 time 0.2740 (0.2931) loss 3.1937 (4.0316) grad_norm 0.9545 (1.0779) [2021-04-15 18:02:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][360/1251] eta 0:04:20 lr 0.000950 time 0.2775 (0.2928) loss 4.4466 (4.0356) grad_norm 0.9466 (1.0772) [2021-04-15 18:02:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][370/1251] eta 0:04:17 lr 0.000950 time 0.2806 (0.2928) loss 4.2312 (4.0222) grad_norm 1.0287 (1.0760) [2021-04-15 18:02:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][380/1251] eta 0:04:14 lr 0.000950 time 0.2815 (0.2924) loss 4.0047 (4.0144) grad_norm 1.1673 (1.0764) [2021-04-15 18:02:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][390/1251] eta 0:04:11 lr 0.000950 time 0.3092 (0.2921) loss 3.1677 (4.0177) grad_norm 1.2037 (1.0765) [2021-04-15 18:02:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][400/1251] eta 0:04:08 lr 0.000950 time 0.2955 (0.2919) loss 2.9990 (4.0194) grad_norm 1.0079 (1.0784) [2021-04-15 18:02:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][410/1251] eta 0:04:05 lr 0.000950 time 0.2864 (0.2916) loss 4.2536 (4.0237) grad_norm 0.9995 (1.0787) [2021-04-15 18:02:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][420/1251] eta 0:04:02 lr 0.000950 time 0.2806 (0.2912) loss 3.9725 (4.0189) grad_norm 1.2013 (1.0789) [2021-04-15 18:02:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][430/1251] eta 0:03:58 lr 0.000950 time 0.2749 (0.2910) loss 4.2453 (4.0177) grad_norm 1.0689 (1.0818) [2021-04-15 18:02:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][440/1251] eta 0:03:55 lr 0.000950 time 0.2701 (0.2907) loss 4.6050 (4.0252) grad_norm 1.0347 (1.0807) [2021-04-15 18:02:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][450/1251] eta 0:03:52 lr 0.000950 time 0.2659 (0.2904) loss 4.0625 (4.0254) grad_norm 0.9755 (1.0801) [2021-04-15 18:03:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][460/1251] eta 0:03:49 lr 0.000950 time 0.2751 (0.2902) loss 4.1987 (4.0276) grad_norm 1.0544 (1.0798) [2021-04-15 18:03:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][470/1251] eta 0:03:46 lr 0.000950 time 0.2879 (0.2899) loss 4.5091 (4.0294) grad_norm 1.0896 (1.0802) [2021-04-15 18:03:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][480/1251] eta 0:03:43 lr 0.000950 time 0.2503 (0.2896) loss 4.4524 (4.0280) grad_norm 1.0732 (1.0799) [2021-04-15 18:03:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][490/1251] eta 0:03:40 lr 0.000950 time 0.2829 (0.2894) loss 4.7586 (4.0299) grad_norm 0.9710 (1.0800) [2021-04-15 18:03:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][500/1251] eta 0:03:37 lr 0.000950 time 0.3148 (0.2894) loss 3.9224 (4.0279) grad_norm 0.9598 (1.0782) [2021-04-15 18:03:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][510/1251] eta 0:03:34 lr 0.000950 time 0.2811 (0.2892) loss 4.8429 (4.0224) grad_norm 1.0829 (1.0786) [2021-04-15 18:03:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][520/1251] eta 0:03:31 lr 0.000950 time 0.2820 (0.2889) loss 3.8935 (4.0231) grad_norm 1.0805 (1.0782) [2021-04-15 18:03:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][530/1251] eta 0:03:28 lr 0.000950 time 0.2977 (0.2888) loss 3.8681 (4.0206) grad_norm 1.1496 (1.0780) [2021-04-15 18:03:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][540/1251] eta 0:03:25 lr 0.000950 time 0.2728 (0.2886) loss 3.3884 (4.0217) grad_norm 0.9434 (1.0788) [2021-04-15 18:03:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][550/1251] eta 0:03:22 lr 0.000950 time 0.3011 (0.2885) loss 3.9383 (4.0248) grad_norm 1.1985 (1.0778) [2021-04-15 18:03:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][560/1251] eta 0:03:19 lr 0.000950 time 0.2596 (0.2883) loss 3.8977 (4.0263) grad_norm 0.8870 (1.0784) [2021-04-15 18:03:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][570/1251] eta 0:03:16 lr 0.000950 time 0.2956 (0.2882) loss 4.4909 (4.0295) grad_norm 1.0431 (1.0789) [2021-04-15 18:03:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][580/1251] eta 0:03:13 lr 0.000950 time 0.2685 (0.2882) loss 4.1190 (4.0351) grad_norm 1.0053 (1.0793) [2021-04-15 18:03:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][590/1251] eta 0:03:10 lr 0.000950 time 0.2864 (0.2882) loss 4.2083 (4.0337) grad_norm 0.9471 (1.0793) [2021-04-15 18:03:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][600/1251] eta 0:03:07 lr 0.000950 time 0.2804 (0.2881) loss 4.2900 (4.0383) grad_norm 1.2019 (1.0793) [2021-04-15 18:03:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][610/1251] eta 0:03:04 lr 0.000950 time 0.2645 (0.2878) loss 3.9670 (4.0373) grad_norm 1.1920 (1.0798) [2021-04-15 18:03:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][620/1251] eta 0:03:01 lr 0.000950 time 0.2601 (0.2877) loss 4.5084 (4.0312) grad_norm 1.2876 (1.0805) [2021-04-15 18:03:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][630/1251] eta 0:02:58 lr 0.000950 time 0.2660 (0.2875) loss 4.2656 (4.0327) grad_norm 0.9951 (1.0806) [2021-04-15 18:03:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][640/1251] eta 0:02:55 lr 0.000949 time 0.2658 (0.2873) loss 3.2357 (4.0268) grad_norm 0.9734 (1.0801) [2021-04-15 18:03:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][650/1251] eta 0:02:52 lr 0.000949 time 0.3039 (0.2872) loss 3.3959 (4.0239) grad_norm 1.0206 (1.0811) [2021-04-15 18:03:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][660/1251] eta 0:02:49 lr 0.000949 time 0.2922 (0.2871) loss 2.9002 (4.0225) grad_norm 1.1114 (1.0813) [2021-04-15 18:04:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][670/1251] eta 0:02:46 lr 0.000949 time 0.2637 (0.2872) loss 3.5896 (4.0251) grad_norm 1.0718 (1.0810) [2021-04-15 18:04:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][680/1251] eta 0:02:43 lr 0.000949 time 0.2517 (0.2872) loss 4.4297 (4.0229) grad_norm 0.9807 (1.0825) [2021-04-15 18:04:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][690/1251] eta 0:02:41 lr 0.000949 time 0.2808 (0.2871) loss 2.8654 (4.0173) grad_norm 1.2175 (1.0819) [2021-04-15 18:04:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][700/1251] eta 0:02:38 lr 0.000949 time 0.2549 (0.2869) loss 4.1614 (4.0200) grad_norm 1.0666 (1.0818) [2021-04-15 18:04:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][710/1251] eta 0:02:35 lr 0.000949 time 0.2835 (0.2868) loss 4.0766 (4.0175) grad_norm 1.1475 (1.0825) [2021-04-15 18:04:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][720/1251] eta 0:02:32 lr 0.000949 time 0.2794 (0.2868) loss 4.5993 (4.0189) grad_norm 1.0970 (1.0822) [2021-04-15 18:04:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][730/1251] eta 0:02:29 lr 0.000949 time 0.2853 (0.2868) loss 3.5083 (4.0239) grad_norm 1.0108 (1.0821) [2021-04-15 18:04:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][740/1251] eta 0:02:26 lr 0.000949 time 0.2648 (0.2866) loss 3.3883 (4.0161) grad_norm 1.0788 (1.0821) [2021-04-15 18:04:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][750/1251] eta 0:02:23 lr 0.000949 time 0.3045 (0.2865) loss 4.1538 (4.0156) grad_norm 1.1660 (1.0828) [2021-04-15 18:04:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][760/1251] eta 0:02:20 lr 0.000949 time 0.2743 (0.2864) loss 3.8028 (4.0174) grad_norm 1.1925 (1.0837) [2021-04-15 18:04:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][770/1251] eta 0:02:17 lr 0.000949 time 0.2791 (0.2863) loss 4.0606 (4.0193) grad_norm 1.0595 (1.0830) [2021-04-15 18:04:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][780/1251] eta 0:02:14 lr 0.000949 time 0.2694 (0.2862) loss 4.0005 (4.0173) grad_norm 1.2826 (1.0847) [2021-04-15 18:04:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][790/1251] eta 0:02:11 lr 0.000949 time 0.2732 (0.2861) loss 3.8377 (4.0158) grad_norm 0.9578 (1.0851) [2021-04-15 18:04:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][800/1251] eta 0:02:08 lr 0.000949 time 0.2633 (0.2860) loss 4.1795 (4.0170) grad_norm 1.0308 (1.0844) [2021-04-15 18:04:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][810/1251] eta 0:02:06 lr 0.000949 time 0.2666 (0.2859) loss 3.8602 (4.0168) grad_norm 1.1163 (1.0838) [2021-04-15 18:04:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][820/1251] eta 0:02:03 lr 0.000949 time 0.2628 (0.2859) loss 3.3118 (4.0166) grad_norm 1.0311 (1.0836) [2021-04-15 18:04:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][830/1251] eta 0:02:00 lr 0.000949 time 0.2629 (0.2858) loss 4.2497 (4.0139) grad_norm 1.1140 (1.0833) [2021-04-15 18:04:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][840/1251] eta 0:01:57 lr 0.000949 time 0.2587 (0.2857) loss 3.2079 (4.0129) grad_norm 1.2338 (1.0836) [2021-04-15 18:04:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][850/1251] eta 0:01:54 lr 0.000949 time 0.2991 (0.2856) loss 3.1189 (4.0130) grad_norm 1.0467 (1.0841) [2021-04-15 18:04:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][860/1251] eta 0:01:51 lr 0.000949 time 0.2835 (0.2856) loss 4.5421 (4.0086) grad_norm 1.1262 (1.0835) [2021-04-15 18:04:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][870/1251] eta 0:01:48 lr 0.000949 time 0.2726 (0.2855) loss 4.2349 (4.0037) grad_norm 0.9545 (1.0828) [2021-04-15 18:04:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][880/1251] eta 0:01:45 lr 0.000949 time 0.2825 (0.2854) loss 4.2145 (4.0056) grad_norm 0.9668 (1.0814) [2021-04-15 18:05:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][890/1251] eta 0:01:43 lr 0.000949 time 0.2576 (0.2854) loss 4.9045 (4.0038) grad_norm 1.1416 (1.0812) [2021-04-15 18:05:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][900/1251] eta 0:01:40 lr 0.000949 time 0.2726 (0.2853) loss 4.4077 (4.0049) grad_norm 0.9682 (1.0814) [2021-04-15 18:05:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][910/1251] eta 0:01:37 lr 0.000949 time 0.2623 (0.2853) loss 4.2533 (4.0054) grad_norm 1.2651 (1.0818) [2021-04-15 18:05:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][920/1251] eta 0:01:34 lr 0.000949 time 0.2940 (0.2854) loss 4.3555 (4.0067) grad_norm 1.4287 (1.0821) [2021-04-15 18:05:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][930/1251] eta 0:01:31 lr 0.000949 time 0.2884 (0.2854) loss 4.2296 (4.0091) grad_norm 0.9899 (1.0828) [2021-04-15 18:05:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][940/1251] eta 0:01:28 lr 0.000949 time 0.2563 (0.2854) loss 3.7704 (4.0071) grad_norm 1.0892 (1.0828) [2021-04-15 18:05:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][950/1251] eta 0:01:25 lr 0.000949 time 0.2940 (0.2854) loss 3.1267 (4.0032) grad_norm 1.1756 (1.0842) [2021-04-15 18:05:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][960/1251] eta 0:01:23 lr 0.000949 time 0.3052 (0.2853) loss 4.1206 (4.0018) grad_norm 1.4309 (1.0848) [2021-04-15 18:05:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][970/1251] eta 0:01:20 lr 0.000949 time 0.2715 (0.2852) loss 4.5002 (4.0006) grad_norm 1.0495 (1.0859) [2021-04-15 18:05:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][980/1251] eta 0:01:17 lr 0.000949 time 0.2651 (0.2852) loss 2.6974 (3.9981) grad_norm 0.9330 (1.0864) [2021-04-15 18:05:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][990/1251] eta 0:01:14 lr 0.000949 time 0.2682 (0.2851) loss 3.7115 (3.9986) grad_norm 0.8825 (1.0866) [2021-04-15 18:05:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][1000/1251] eta 0:01:11 lr 0.000949 time 0.2714 (0.2850) loss 3.7423 (3.9974) grad_norm 0.9933 (1.0866) [2021-04-15 18:05:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][1010/1251] eta 0:01:08 lr 0.000949 time 0.2908 (0.2850) loss 4.2766 (3.9984) grad_norm 1.0591 (1.0867) [2021-04-15 18:05:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][1020/1251] eta 0:01:05 lr 0.000949 time 0.3002 (0.2849) loss 4.6814 (4.0018) grad_norm 1.5737 (1.0878) [2021-04-15 18:05:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][1030/1251] eta 0:01:02 lr 0.000949 time 0.3316 (0.2849) loss 4.2425 (4.0002) grad_norm 1.0964 (1.0882) [2021-04-15 18:05:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][1040/1251] eta 0:01:00 lr 0.000949 time 0.2655 (0.2849) loss 4.1860 (4.0008) grad_norm 1.7118 (1.0899) [2021-04-15 18:05:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][1050/1251] eta 0:00:57 lr 0.000949 time 0.2755 (0.2849) loss 4.1847 (3.9992) grad_norm 1.1064 (1.0908) [2021-04-15 18:05:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][1060/1251] eta 0:00:54 lr 0.000949 time 0.2675 (0.2850) loss 4.7311 (3.9980) grad_norm 1.1670 (1.0911) [2021-04-15 18:05:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][1070/1251] eta 0:00:51 lr 0.000949 time 0.2922 (0.2849) loss 3.0173 (3.9964) grad_norm 1.2558 (1.0915) [2021-04-15 18:05:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][1080/1251] eta 0:00:48 lr 0.000949 time 0.2768 (0.2848) loss 5.0959 (3.9965) grad_norm 1.0973 (1.0916) [2021-04-15 18:05:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][1090/1251] eta 0:00:45 lr 0.000949 time 0.2835 (0.2848) loss 4.2766 (3.9956) grad_norm 1.1235 (1.0912) [2021-04-15 18:06:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][1100/1251] eta 0:00:42 lr 0.000949 time 0.2791 (0.2847) loss 4.0331 (3.9945) grad_norm 1.1626 (1.0917) [2021-04-15 18:06:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][1110/1251] eta 0:00:40 lr 0.000949 time 0.2561 (0.2846) loss 4.5655 (3.9943) grad_norm 1.1944 (1.0916) [2021-04-15 18:06:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][1120/1251] eta 0:00:37 lr 0.000949 time 0.2719 (0.2845) loss 4.1683 (3.9935) grad_norm 0.9211 (1.0909) [2021-04-15 18:06:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][1130/1251] eta 0:00:34 lr 0.000949 time 0.2711 (0.2845) loss 4.0400 (3.9953) grad_norm 1.0256 (1.0908) [2021-04-15 18:06:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][1140/1251] eta 0:00:31 lr 0.000949 time 0.2718 (0.2844) loss 4.2609 (3.9958) grad_norm 1.2040 (1.0911) [2021-04-15 18:06:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][1150/1251] eta 0:00:28 lr 0.000949 time 0.2900 (0.2843) loss 3.5624 (3.9971) grad_norm 1.2218 (1.0921) [2021-04-15 18:06:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][1160/1251] eta 0:00:25 lr 0.000949 time 0.2891 (0.2844) loss 3.8099 (3.9961) grad_norm 1.0550 (1.0930) [2021-04-15 18:06:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][1170/1251] eta 0:00:23 lr 0.000949 time 0.2651 (0.2843) loss 4.9514 (3.9961) grad_norm 1.2335 (1.0932) [2021-04-15 18:06:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][1180/1251] eta 0:00:20 lr 0.000949 time 0.2649 (0.2843) loss 4.0727 (3.9966) grad_norm 0.9788 (1.0934) [2021-04-15 18:06:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][1190/1251] eta 0:00:17 lr 0.000948 time 0.2916 (0.2843) loss 3.5515 (3.9971) grad_norm 1.3513 (1.0936) [2021-04-15 18:06:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][1200/1251] eta 0:00:14 lr 0.000948 time 0.2597 (0.2842) loss 4.0382 (3.9962) grad_norm 1.1803 (1.0942) [2021-04-15 18:06:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][1210/1251] eta 0:00:11 lr 0.000948 time 0.2732 (0.2841) loss 3.4132 (3.9959) grad_norm 0.9697 (1.0941) [2021-04-15 18:06:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][1220/1251] eta 0:00:08 lr 0.000948 time 0.2793 (0.2841) loss 4.1037 (3.9976) grad_norm 0.9190 (1.0937) [2021-04-15 18:06:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][1230/1251] eta 0:00:05 lr 0.000948 time 0.3109 (0.2841) loss 4.2882 (3.9970) grad_norm 0.9336 (1.0932) [2021-04-15 18:06:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][1240/1251] eta 0:00:03 lr 0.000948 time 0.2660 (0.2840) loss 4.2890 (3.9984) grad_norm 1.3417 (1.0928) [2021-04-15 18:06:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [43/300][1250/1251] eta 0:00:00 lr 0.000948 time 0.2485 (0.2838) loss 4.1861 (4.0012) grad_norm 1.2552 (1.0924) [2021-04-15 18:06:44 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 43 training takes 0:05:57 [2021-04-15 18:06:44 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_43.pth saving...... [2021-04-15 18:06:59 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_43.pth saved !!! [2021-04-15 18:07:00 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.312 (1.312) Loss 1.3876 (1.3876) Acc@1 68.555 (68.555) Acc@5 89.160 (89.160) [2021-04-15 18:07:01 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.128 (0.240) Loss 1.3761 (1.3675) Acc@1 66.309 (68.670) Acc@5 89.160 (89.426) [2021-04-15 18:07:04 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.292 (0.239) Loss 1.3130 (1.3485) Acc@1 70.020 (69.043) Acc@5 89.355 (89.518) [2021-04-15 18:07:06 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.108 (0.227) Loss 1.3430 (1.3458) Acc@1 69.238 (69.097) Acc@5 89.062 (89.526) [2021-04-15 18:07:08 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.218) Loss 1.3276 (1.3473) Acc@1 69.336 (68.995) Acc@5 90.137 (89.570) [2021-04-15 18:07:11 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 69.016 Acc@5 89.460 [2021-04-15 18:07:11 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 69.0% [2021-04-15 18:07:11 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 69.02% [2021-04-15 18:07:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][0/1251] eta 1:41:44 lr 0.000948 time 4.8794 (4.8794) loss 4.3629 (4.3629) grad_norm 1.0500 (1.0500) [2021-04-15 18:07:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][10/1251] eta 0:14:19 lr 0.000948 time 0.2761 (0.6922) loss 4.4155 (4.1342) grad_norm 0.8923 (0.9964) [2021-04-15 18:07:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][20/1251] eta 0:10:11 lr 0.000948 time 0.2729 (0.4969) loss 4.5060 (4.0047) grad_norm 1.0115 (1.0401) [2021-04-15 18:07:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][30/1251] eta 0:08:43 lr 0.000948 time 0.3159 (0.4283) loss 3.7064 (3.9904) grad_norm 0.9996 (1.0318) [2021-04-15 18:07:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][40/1251] eta 0:07:54 lr 0.000948 time 0.2768 (0.3921) loss 4.3947 (3.9311) grad_norm 0.9783 (1.0626) [2021-04-15 18:07:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][50/1251] eta 0:07:24 lr 0.000948 time 0.2612 (0.3703) loss 2.8940 (3.9195) grad_norm 1.0885 (1.0705) [2021-04-15 18:07:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][60/1251] eta 0:07:03 lr 0.000948 time 0.2965 (0.3553) loss 4.3603 (3.8718) grad_norm 0.8806 (1.0717) [2021-04-15 18:07:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][70/1251] eta 0:06:48 lr 0.000948 time 0.2857 (0.3455) loss 2.9816 (3.8608) grad_norm 0.9567 (1.0644) [2021-04-15 18:07:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][80/1251] eta 0:06:34 lr 0.000948 time 0.2593 (0.3369) loss 4.2300 (3.8840) grad_norm 1.1835 (1.0654) [2021-04-15 18:07:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][90/1251] eta 0:06:24 lr 0.000948 time 0.2726 (0.3309) loss 4.1718 (3.8977) grad_norm 1.1474 (1.0677) [2021-04-15 18:07:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][100/1251] eta 0:06:15 lr 0.000948 time 0.2807 (0.3260) loss 4.4161 (3.8958) grad_norm 0.8795 (1.0672) [2021-04-15 18:07:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][110/1251] eta 0:06:07 lr 0.000948 time 0.2938 (0.3222) loss 3.2402 (3.8748) grad_norm 1.0232 (1.0663) [2021-04-15 18:07:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][120/1251] eta 0:06:02 lr 0.000948 time 0.2780 (0.3205) loss 4.3951 (3.9036) grad_norm 0.9720 (1.0642) [2021-04-15 18:07:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][130/1251] eta 0:05:55 lr 0.000948 time 0.2541 (0.3172) loss 4.7444 (3.9205) grad_norm 1.7159 (1.0685) [2021-04-15 18:07:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][140/1251] eta 0:05:50 lr 0.000948 time 0.2984 (0.3159) loss 4.5429 (3.9048) grad_norm 1.4041 (1.0701) [2021-04-15 18:07:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][150/1251] eta 0:05:45 lr 0.000948 time 0.3123 (0.3137) loss 3.7262 (3.8997) grad_norm 0.9793 (1.0669) [2021-04-15 18:08:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][160/1251] eta 0:05:39 lr 0.000948 time 0.2766 (0.3113) loss 4.3361 (3.9272) grad_norm 1.0848 (1.0733) [2021-04-15 18:08:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][170/1251] eta 0:05:34 lr 0.000948 time 0.2867 (0.3097) loss 3.5655 (3.9290) grad_norm 1.1024 (1.0784) [2021-04-15 18:08:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][180/1251] eta 0:05:29 lr 0.000948 time 0.2831 (0.3081) loss 3.3631 (3.9326) grad_norm 1.0710 (1.0812) [2021-04-15 18:08:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][190/1251] eta 0:05:25 lr 0.000948 time 0.2870 (0.3069) loss 4.5599 (3.9263) grad_norm 0.8732 (1.0817) [2021-04-15 18:08:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][200/1251] eta 0:05:20 lr 0.000948 time 0.2669 (0.3054) loss 2.9509 (3.9163) grad_norm 1.3382 (1.0880) [2021-04-15 18:08:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][210/1251] eta 0:05:16 lr 0.000948 time 0.2860 (0.3041) loss 3.4869 (3.9234) grad_norm 1.0845 (1.0864) [2021-04-15 18:08:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][220/1251] eta 0:05:12 lr 0.000948 time 0.2758 (0.3032) loss 3.0669 (3.9378) grad_norm 1.1081 (1.0867) [2021-04-15 18:08:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][230/1251] eta 0:05:08 lr 0.000948 time 0.2853 (0.3021) loss 3.8728 (3.9420) grad_norm 1.2328 (inf) [2021-04-15 18:08:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][240/1251] eta 0:05:04 lr 0.000948 time 0.2639 (0.3011) loss 4.0771 (3.9496) grad_norm 1.1208 (inf) [2021-04-15 18:08:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][250/1251] eta 0:05:00 lr 0.000948 time 0.2853 (0.3002) loss 4.0513 (3.9461) grad_norm 1.1711 (inf) [2021-04-15 18:08:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][260/1251] eta 0:04:56 lr 0.000948 time 0.2875 (0.2994) loss 4.4891 (3.9395) grad_norm 1.0341 (inf) [2021-04-15 18:08:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][270/1251] eta 0:04:53 lr 0.000948 time 0.2920 (0.2987) loss 3.3874 (3.9382) grad_norm 1.1794 (inf) [2021-04-15 18:08:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][280/1251] eta 0:04:49 lr 0.000948 time 0.2718 (0.2980) loss 3.0501 (3.9347) grad_norm 1.1316 (inf) [2021-04-15 18:08:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][290/1251] eta 0:04:45 lr 0.000948 time 0.2795 (0.2974) loss 4.6378 (3.9322) grad_norm 0.8840 (inf) [2021-04-15 18:08:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][300/1251] eta 0:04:42 lr 0.000948 time 0.2835 (0.2967) loss 4.0259 (3.9368) grad_norm 1.0852 (inf) [2021-04-15 18:08:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][310/1251] eta 0:04:38 lr 0.000948 time 0.2808 (0.2962) loss 4.4587 (3.9474) grad_norm 1.3812 (inf) [2021-04-15 18:08:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][320/1251] eta 0:04:35 lr 0.000948 time 0.2564 (0.2956) loss 4.8374 (3.9518) grad_norm 1.0179 (inf) [2021-04-15 18:08:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][330/1251] eta 0:04:31 lr 0.000948 time 0.2756 (0.2952) loss 4.0678 (3.9570) grad_norm 0.9714 (inf) [2021-04-15 18:08:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][340/1251] eta 0:04:28 lr 0.000948 time 0.2911 (0.2947) loss 3.5943 (3.9547) grad_norm 1.3809 (inf) [2021-04-15 18:08:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][350/1251] eta 0:04:24 lr 0.000948 time 0.2632 (0.2940) loss 4.1681 (3.9533) grad_norm 0.9666 (inf) [2021-04-15 18:08:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][360/1251] eta 0:04:21 lr 0.000948 time 0.2752 (0.2940) loss 4.8116 (3.9569) grad_norm 1.2658 (inf) [2021-04-15 18:08:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][370/1251] eta 0:04:18 lr 0.000948 time 0.2856 (0.2936) loss 5.1786 (3.9513) grad_norm 1.3351 (inf) [2021-04-15 18:09:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][380/1251] eta 0:04:15 lr 0.000948 time 0.2810 (0.2932) loss 3.9109 (3.9599) grad_norm 1.1643 (inf) [2021-04-15 18:09:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][390/1251] eta 0:04:12 lr 0.000948 time 0.2748 (0.2930) loss 4.5771 (3.9696) grad_norm 0.9179 (inf) [2021-04-15 18:09:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][400/1251] eta 0:04:09 lr 0.000948 time 0.3005 (0.2927) loss 3.8537 (3.9682) grad_norm 0.8829 (inf) [2021-04-15 18:09:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][410/1251] eta 0:04:05 lr 0.000948 time 0.2640 (0.2924) loss 4.5181 (3.9695) grad_norm 1.2495 (inf) [2021-04-15 18:09:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][420/1251] eta 0:04:02 lr 0.000948 time 0.2888 (0.2923) loss 4.4387 (3.9651) grad_norm 1.1449 (inf) [2021-04-15 18:09:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][430/1251] eta 0:03:59 lr 0.000948 time 0.3037 (0.2920) loss 4.3774 (3.9665) grad_norm 0.9237 (inf) [2021-04-15 18:09:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][440/1251] eta 0:03:56 lr 0.000948 time 0.2776 (0.2921) loss 3.1912 (3.9626) grad_norm 0.9309 (inf) [2021-04-15 18:09:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][450/1251] eta 0:03:53 lr 0.000948 time 0.4524 (0.2921) loss 2.8532 (3.9611) grad_norm 0.9131 (inf) [2021-04-15 18:09:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][460/1251] eta 0:03:50 lr 0.000948 time 0.2885 (0.2919) loss 3.8955 (3.9618) grad_norm 0.9723 (inf) [2021-04-15 18:09:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][470/1251] eta 0:03:47 lr 0.000948 time 0.2715 (0.2916) loss 4.7545 (3.9633) grad_norm 1.2116 (inf) [2021-04-15 18:09:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][480/1251] eta 0:03:44 lr 0.000947 time 0.2632 (0.2916) loss 4.1048 (3.9698) grad_norm 0.9697 (inf) [2021-04-15 18:09:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][490/1251] eta 0:03:41 lr 0.000947 time 0.2778 (0.2914) loss 3.8503 (3.9725) grad_norm 1.0430 (inf) [2021-04-15 18:09:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][500/1251] eta 0:03:38 lr 0.000947 time 0.2694 (0.2911) loss 3.2347 (3.9710) grad_norm 0.9092 (inf) [2021-04-15 18:09:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][510/1251] eta 0:03:35 lr 0.000947 time 0.2751 (0.2909) loss 4.5273 (3.9712) grad_norm 1.1199 (inf) [2021-04-15 18:09:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][520/1251] eta 0:03:32 lr 0.000947 time 0.2842 (0.2906) loss 4.8505 (3.9690) grad_norm 0.9619 (inf) [2021-04-15 18:09:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][530/1251] eta 0:03:29 lr 0.000947 time 0.2744 (0.2903) loss 3.8301 (3.9682) grad_norm 0.9078 (inf) [2021-04-15 18:09:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][540/1251] eta 0:03:26 lr 0.000947 time 0.2630 (0.2899) loss 4.2731 (3.9649) grad_norm 1.1367 (inf) [2021-04-15 18:09:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][550/1251] eta 0:03:23 lr 0.000947 time 0.2793 (0.2898) loss 2.8900 (3.9664) grad_norm 1.0515 (inf) [2021-04-15 18:09:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][560/1251] eta 0:03:20 lr 0.000947 time 0.2980 (0.2896) loss 4.0638 (3.9643) grad_norm 1.0157 (inf) [2021-04-15 18:09:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][570/1251] eta 0:03:17 lr 0.000947 time 0.2698 (0.2896) loss 4.3382 (3.9648) grad_norm 0.8840 (inf) [2021-04-15 18:09:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][580/1251] eta 0:03:14 lr 0.000947 time 0.2591 (0.2895) loss 3.4604 (3.9609) grad_norm 0.9279 (inf) [2021-04-15 18:10:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][590/1251] eta 0:03:11 lr 0.000947 time 0.2732 (0.2893) loss 3.3937 (3.9636) grad_norm 0.9550 (inf) [2021-04-15 18:10:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][600/1251] eta 0:03:08 lr 0.000947 time 0.2846 (0.2892) loss 4.2079 (3.9644) grad_norm 1.0323 (inf) [2021-04-15 18:10:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][610/1251] eta 0:03:05 lr 0.000947 time 0.2962 (0.2890) loss 4.7730 (3.9678) grad_norm 1.0599 (inf) [2021-04-15 18:10:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][620/1251] eta 0:03:02 lr 0.000947 time 0.2822 (0.2889) loss 4.3405 (3.9661) grad_norm 0.9288 (inf) [2021-04-15 18:10:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][630/1251] eta 0:02:59 lr 0.000947 time 0.2978 (0.2886) loss 4.4356 (3.9656) grad_norm 1.0259 (inf) [2021-04-15 18:10:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][640/1251] eta 0:02:56 lr 0.000947 time 0.2894 (0.2884) loss 3.0348 (3.9658) grad_norm 1.3505 (inf) [2021-04-15 18:10:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][650/1251] eta 0:02:53 lr 0.000947 time 0.2859 (0.2883) loss 4.6421 (3.9653) grad_norm 1.0146 (inf) [2021-04-15 18:10:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][660/1251] eta 0:02:50 lr 0.000947 time 0.2800 (0.2881) loss 3.7385 (3.9704) grad_norm 1.1947 (inf) [2021-04-15 18:10:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][670/1251] eta 0:02:47 lr 0.000947 time 0.3078 (0.2881) loss 3.9195 (3.9680) grad_norm 1.2020 (inf) [2021-04-15 18:10:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][680/1251] eta 0:02:44 lr 0.000947 time 0.2750 (0.2879) loss 4.6429 (3.9720) grad_norm 1.0004 (inf) [2021-04-15 18:10:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][690/1251] eta 0:02:41 lr 0.000947 time 0.2644 (0.2878) loss 3.4133 (3.9747) grad_norm 0.9575 (inf) [2021-04-15 18:10:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][700/1251] eta 0:02:38 lr 0.000947 time 0.2663 (0.2878) loss 4.6883 (3.9773) grad_norm 1.0355 (inf) [2021-04-15 18:10:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][710/1251] eta 0:02:35 lr 0.000947 time 0.2507 (0.2876) loss 4.5029 (3.9814) grad_norm 1.0631 (inf) [2021-04-15 18:10:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][720/1251] eta 0:02:32 lr 0.000947 time 0.2740 (0.2877) loss 3.6594 (3.9814) grad_norm 1.0750 (inf) [2021-04-15 18:10:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][730/1251] eta 0:02:29 lr 0.000947 time 0.2964 (0.2876) loss 2.5636 (3.9804) grad_norm 1.2232 (inf) [2021-04-15 18:10:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][740/1251] eta 0:02:26 lr 0.000947 time 0.2724 (0.2875) loss 3.9000 (3.9779) grad_norm 1.2013 (inf) [2021-04-15 18:10:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][750/1251] eta 0:02:23 lr 0.000947 time 0.2689 (0.2873) loss 4.5323 (3.9760) grad_norm 1.1063 (inf) [2021-04-15 18:10:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][760/1251] eta 0:02:21 lr 0.000947 time 0.2883 (0.2873) loss 3.2194 (3.9764) grad_norm 1.2185 (inf) [2021-04-15 18:10:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][770/1251] eta 0:02:18 lr 0.000947 time 0.2906 (0.2872) loss 4.1985 (3.9744) grad_norm 1.0181 (inf) [2021-04-15 18:10:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][780/1251] eta 0:02:15 lr 0.000947 time 0.2650 (0.2871) loss 4.6038 (3.9764) grad_norm 1.3086 (inf) [2021-04-15 18:10:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][790/1251] eta 0:02:12 lr 0.000947 time 0.2625 (0.2870) loss 3.9924 (3.9747) grad_norm 1.3231 (inf) [2021-04-15 18:11:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][800/1251] eta 0:02:09 lr 0.000947 time 0.3013 (0.2871) loss 4.6949 (3.9774) grad_norm 1.1466 (inf) [2021-04-15 18:11:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][810/1251] eta 0:02:06 lr 0.000947 time 0.2709 (0.2869) loss 4.4484 (3.9769) grad_norm 1.1570 (inf) [2021-04-15 18:11:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][820/1251] eta 0:02:03 lr 0.000947 time 0.2682 (0.2868) loss 4.7448 (3.9775) grad_norm 1.0615 (inf) [2021-04-15 18:11:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][830/1251] eta 0:02:00 lr 0.000947 time 0.2841 (0.2868) loss 2.9203 (3.9753) grad_norm 0.9891 (inf) [2021-04-15 18:11:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][840/1251] eta 0:01:57 lr 0.000947 time 0.2684 (0.2867) loss 4.8297 (3.9753) grad_norm 1.1348 (inf) [2021-04-15 18:11:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][850/1251] eta 0:01:54 lr 0.000947 time 0.2612 (0.2865) loss 4.6107 (3.9730) grad_norm 1.0553 (inf) [2021-04-15 18:11:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][860/1251] eta 0:01:52 lr 0.000947 time 0.2869 (0.2865) loss 3.3267 (3.9747) grad_norm 0.9966 (inf) [2021-04-15 18:11:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][870/1251] eta 0:01:49 lr 0.000947 time 0.2938 (0.2864) loss 4.0731 (3.9739) grad_norm 0.9547 (inf) [2021-04-15 18:11:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][880/1251] eta 0:01:46 lr 0.000947 time 0.2801 (0.2863) loss 2.9287 (3.9741) grad_norm 0.9578 (inf) [2021-04-15 18:11:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][890/1251] eta 0:01:43 lr 0.000947 time 0.2582 (0.2862) loss 4.7540 (3.9754) grad_norm 1.0719 (inf) [2021-04-15 18:11:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][900/1251] eta 0:01:40 lr 0.000947 time 0.2875 (0.2862) loss 4.4397 (3.9782) grad_norm 1.0429 (inf) [2021-04-15 18:11:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][910/1251] eta 0:01:37 lr 0.000947 time 0.2803 (0.2861) loss 3.8425 (3.9785) grad_norm 0.9789 (inf) [2021-04-15 18:11:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][920/1251] eta 0:01:34 lr 0.000947 time 0.2737 (0.2860) loss 4.1878 (3.9752) grad_norm 1.0725 (inf) [2021-04-15 18:11:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][930/1251] eta 0:01:31 lr 0.000947 time 0.2895 (0.2860) loss 4.0396 (3.9714) grad_norm 1.0207 (inf) [2021-04-15 18:11:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][940/1251] eta 0:01:28 lr 0.000947 time 0.2764 (0.2860) loss 3.7253 (3.9729) grad_norm 1.0991 (inf) [2021-04-15 18:11:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][950/1251] eta 0:01:26 lr 0.000947 time 0.2876 (0.2859) loss 3.3784 (3.9721) grad_norm 0.9867 (inf) [2021-04-15 18:11:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][960/1251] eta 0:01:23 lr 0.000947 time 0.3021 (0.2859) loss 4.5126 (3.9745) grad_norm 0.9303 (inf) [2021-04-15 18:11:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][970/1251] eta 0:01:20 lr 0.000947 time 0.2956 (0.2858) loss 4.7696 (3.9726) grad_norm 0.9855 (inf) [2021-04-15 18:11:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][980/1251] eta 0:01:17 lr 0.000947 time 0.2669 (0.2857) loss 4.5506 (3.9726) grad_norm 1.0824 (inf) [2021-04-15 18:11:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][990/1251] eta 0:01:14 lr 0.000947 time 0.3140 (0.2856) loss 5.0002 (3.9737) grad_norm 1.0437 (inf) [2021-04-15 18:11:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][1000/1251] eta 0:01:11 lr 0.000947 time 0.2814 (0.2856) loss 4.8050 (3.9754) grad_norm 1.2195 (inf) [2021-04-15 18:11:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][1010/1251] eta 0:01:08 lr 0.000947 time 0.2944 (0.2855) loss 4.4648 (3.9780) grad_norm 0.9346 (inf) [2021-04-15 18:12:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][1020/1251] eta 0:01:05 lr 0.000946 time 0.3070 (0.2855) loss 3.4326 (3.9769) grad_norm 1.0496 (inf) [2021-04-15 18:12:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][1030/1251] eta 0:01:03 lr 0.000946 time 0.2803 (0.2856) loss 4.8004 (3.9793) grad_norm 1.0435 (inf) [2021-04-15 18:12:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][1040/1251] eta 0:01:00 lr 0.000946 time 0.2639 (0.2855) loss 3.0593 (3.9792) grad_norm 1.0087 (inf) [2021-04-15 18:12:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][1050/1251] eta 0:00:57 lr 0.000946 time 0.4064 (0.2856) loss 5.0151 (3.9829) grad_norm 1.0699 (inf) [2021-04-15 18:12:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][1060/1251] eta 0:00:54 lr 0.000946 time 0.2736 (0.2855) loss 4.4417 (3.9801) grad_norm 1.5006 (inf) [2021-04-15 18:12:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][1070/1251] eta 0:00:51 lr 0.000946 time 0.2658 (0.2854) loss 3.8707 (3.9823) grad_norm 1.0545 (inf) [2021-04-15 18:12:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][1080/1251] eta 0:00:48 lr 0.000946 time 0.2699 (0.2854) loss 4.6001 (3.9827) grad_norm 1.0975 (inf) [2021-04-15 18:12:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][1090/1251] eta 0:00:45 lr 0.000946 time 0.2734 (0.2853) loss 3.7121 (3.9812) grad_norm 1.0416 (inf) [2021-04-15 18:12:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][1100/1251] eta 0:00:43 lr 0.000946 time 0.3037 (0.2853) loss 3.8034 (3.9789) grad_norm 1.0141 (inf) [2021-04-15 18:12:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][1110/1251] eta 0:00:40 lr 0.000946 time 0.2859 (0.2852) loss 4.3851 (3.9805) grad_norm 1.2060 (inf) [2021-04-15 18:12:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][1120/1251] eta 0:00:37 lr 0.000946 time 0.2806 (0.2851) loss 3.1063 (3.9811) grad_norm 1.4967 (inf) [2021-04-15 18:12:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][1130/1251] eta 0:00:34 lr 0.000946 time 0.2570 (0.2850) loss 4.3405 (3.9808) grad_norm 1.0114 (inf) [2021-04-15 18:12:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][1140/1251] eta 0:00:31 lr 0.000946 time 0.2774 (0.2850) loss 4.4953 (3.9812) grad_norm 1.2339 (inf) [2021-04-15 18:12:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][1150/1251] eta 0:00:28 lr 0.000946 time 0.2585 (0.2850) loss 3.4917 (3.9807) grad_norm 0.9790 (inf) [2021-04-15 18:12:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][1160/1251] eta 0:00:25 lr 0.000946 time 0.2661 (0.2850) loss 4.8641 (3.9806) grad_norm 1.0232 (inf) [2021-04-15 18:12:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][1170/1251] eta 0:00:23 lr 0.000946 time 0.2765 (0.2849) loss 3.1427 (3.9796) grad_norm 0.9286 (inf) [2021-04-15 18:12:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][1180/1251] eta 0:00:20 lr 0.000946 time 0.2889 (0.2849) loss 4.1137 (3.9790) grad_norm 0.9030 (inf) [2021-04-15 18:12:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][1190/1251] eta 0:00:17 lr 0.000946 time 0.2523 (0.2849) loss 3.1418 (3.9786) grad_norm 1.1877 (inf) [2021-04-15 18:12:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][1200/1251] eta 0:00:14 lr 0.000946 time 0.2707 (0.2849) loss 4.2889 (3.9773) grad_norm 1.1562 (inf) [2021-04-15 18:12:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][1210/1251] eta 0:00:11 lr 0.000946 time 0.2662 (0.2849) loss 2.7771 (3.9771) grad_norm 1.1950 (inf) [2021-04-15 18:12:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][1220/1251] eta 0:00:08 lr 0.000946 time 0.2952 (0.2848) loss 3.4056 (3.9748) grad_norm 0.9404 (inf) [2021-04-15 18:13:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][1230/1251] eta 0:00:05 lr 0.000946 time 0.2523 (0.2847) loss 4.5773 (3.9752) grad_norm 1.1356 (inf) [2021-04-15 18:13:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][1240/1251] eta 0:00:03 lr 0.000946 time 0.2489 (0.2846) loss 4.9446 (3.9749) grad_norm 1.2057 (inf) [2021-04-15 18:13:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [44/300][1250/1251] eta 0:00:00 lr 0.000946 time 0.2488 (0.2844) loss 4.2800 (3.9754) grad_norm 1.2593 (inf) [2021-04-15 18:13:09 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 44 training takes 0:05:58 [2021-04-15 18:13:09 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_44.pth saving...... [2021-04-15 18:13:20 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_44.pth saved !!! [2021-04-15 18:13:22 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.311 (1.311) Loss 1.4512 (1.4512) Acc@1 66.699 (66.699) Acc@5 88.574 (88.574) [2021-04-15 18:13:23 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.138 (0.231) Loss 1.4087 (1.3755) Acc@1 68.164 (68.342) Acc@5 87.793 (89.444) [2021-04-15 18:13:25 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.115 (0.244) Loss 1.3895 (1.3544) Acc@1 66.992 (69.057) Acc@5 89.844 (89.732) [2021-04-15 18:13:28 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.144 (0.247) Loss 1.3629 (1.3644) Acc@1 68.066 (68.926) Acc@5 89.746 (89.604) [2021-04-15 18:13:29 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.217) Loss 1.3107 (1.3650) Acc@1 70.215 (69.052) Acc@5 90.332 (89.529) [2021-04-15 18:13:32 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 68.898 Acc@5 89.456 [2021-04-15 18:13:32 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 68.9% [2021-04-15 18:13:32 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 69.02% [2021-04-15 18:13:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][0/1251] eta 1:21:32 lr 0.000946 time 3.9109 (3.9109) loss 4.5143 (4.5143) grad_norm 1.0752 (1.0752) [2021-04-15 18:13:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][10/1251] eta 0:12:34 lr 0.000946 time 0.2796 (0.6081) loss 4.5072 (4.0330) grad_norm 1.0510 (1.0874) [2021-04-15 18:13:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][20/1251] eta 0:09:18 lr 0.000946 time 0.3012 (0.4534) loss 4.0966 (3.9870) grad_norm 1.1830 (1.1400) [2021-04-15 18:13:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][30/1251] eta 0:08:05 lr 0.000946 time 0.2944 (0.3978) loss 3.8485 (3.9523) grad_norm 0.9314 (1.1627) [2021-04-15 18:13:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][40/1251] eta 0:07:26 lr 0.000946 time 0.2938 (0.3689) loss 3.3112 (3.9535) grad_norm 1.0099 (1.1460) [2021-04-15 18:13:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][50/1251] eta 0:07:02 lr 0.000946 time 0.2964 (0.3515) loss 4.2405 (4.0043) grad_norm 1.1270 (1.1230) [2021-04-15 18:13:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][60/1251] eta 0:06:44 lr 0.000946 time 0.2635 (0.3394) loss 4.9504 (4.0426) grad_norm 1.2195 (1.1225) [2021-04-15 18:13:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][70/1251] eta 0:06:31 lr 0.000946 time 0.2731 (0.3314) loss 3.3928 (3.9703) grad_norm 1.0171 (1.1171) [2021-04-15 18:13:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][80/1251] eta 0:06:21 lr 0.000946 time 0.2856 (0.3254) loss 4.3058 (3.9947) grad_norm 0.9335 (1.0982) [2021-04-15 18:14:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][90/1251] eta 0:06:11 lr 0.000946 time 0.2632 (0.3203) loss 4.0161 (3.9846) grad_norm 1.0658 (1.0943) [2021-04-15 18:14:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][100/1251] eta 0:06:04 lr 0.000946 time 0.2848 (0.3166) loss 3.0173 (3.9767) grad_norm 1.0860 (1.0913) [2021-04-15 18:14:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][110/1251] eta 0:05:57 lr 0.000946 time 0.2892 (0.3134) loss 4.2930 (3.9995) grad_norm 1.2680 (1.0982) [2021-04-15 18:14:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][120/1251] eta 0:05:51 lr 0.000946 time 0.2820 (0.3104) loss 3.0977 (4.0259) grad_norm 0.9650 (1.0948) [2021-04-15 18:14:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][130/1251] eta 0:05:45 lr 0.000946 time 0.2731 (0.3079) loss 4.0927 (4.0268) grad_norm 1.0232 (1.0951) [2021-04-15 18:14:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][140/1251] eta 0:05:39 lr 0.000946 time 0.2877 (0.3058) loss 4.2239 (4.0276) grad_norm 1.0755 (1.0992) [2021-04-15 18:14:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][150/1251] eta 0:05:34 lr 0.000946 time 0.2716 (0.3040) loss 3.2944 (4.0090) grad_norm 1.1479 (1.0987) [2021-04-15 18:14:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][160/1251] eta 0:05:30 lr 0.000946 time 0.2807 (0.3031) loss 3.9275 (4.0201) grad_norm 1.0630 (1.0960) [2021-04-15 18:14:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][170/1251] eta 0:05:26 lr 0.000946 time 0.2989 (0.3021) loss 4.5581 (4.0182) grad_norm 1.0466 (1.0920) [2021-04-15 18:14:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][180/1251] eta 0:05:22 lr 0.000946 time 0.2700 (0.3007) loss 3.2843 (4.0158) grad_norm 1.1229 (1.0913) [2021-04-15 18:14:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][190/1251] eta 0:05:17 lr 0.000946 time 0.2856 (0.2997) loss 3.5693 (4.0321) grad_norm 0.9373 (1.0886) [2021-04-15 18:14:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][200/1251] eta 0:05:14 lr 0.000946 time 0.2712 (0.2989) loss 4.8942 (4.0302) grad_norm 1.1130 (1.0870) [2021-04-15 18:14:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][210/1251] eta 0:05:10 lr 0.000946 time 0.2727 (0.2982) loss 2.6100 (4.0171) grad_norm 0.8847 (1.0863) [2021-04-15 18:14:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][220/1251] eta 0:05:06 lr 0.000946 time 0.2635 (0.2973) loss 2.8517 (3.9962) grad_norm 1.2620 (1.0875) [2021-04-15 18:14:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][230/1251] eta 0:05:03 lr 0.000946 time 0.2564 (0.2974) loss 2.7388 (3.9938) grad_norm 0.9746 (1.0891) [2021-04-15 18:14:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][240/1251] eta 0:05:00 lr 0.000946 time 0.2994 (0.2974) loss 3.3967 (3.9910) grad_norm 1.1825 (1.0898) [2021-04-15 18:14:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][250/1251] eta 0:04:57 lr 0.000946 time 0.2867 (0.2971) loss 3.1237 (3.9894) grad_norm 0.9793 (1.0870) [2021-04-15 18:14:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][260/1251] eta 0:04:54 lr 0.000946 time 0.4417 (0.2972) loss 4.1947 (3.9830) grad_norm 0.9113 (1.0862) [2021-04-15 18:14:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][270/1251] eta 0:04:50 lr 0.000946 time 0.3082 (0.2965) loss 4.1551 (3.9945) grad_norm 0.9801 (1.0834) [2021-04-15 18:14:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][280/1251] eta 0:04:47 lr 0.000946 time 0.2723 (0.2962) loss 4.2235 (3.9852) grad_norm 1.0707 (1.0842) [2021-04-15 18:14:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][290/1251] eta 0:04:44 lr 0.000946 time 0.2686 (0.2957) loss 4.1613 (3.9829) grad_norm 1.0974 (1.0852) [2021-04-15 18:15:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][300/1251] eta 0:04:40 lr 0.000945 time 0.2939 (0.2951) loss 3.7383 (3.9765) grad_norm 1.3406 (1.0846) [2021-04-15 18:15:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][310/1251] eta 0:04:37 lr 0.000945 time 0.2851 (0.2947) loss 3.1738 (3.9759) grad_norm 1.0170 (1.0854) [2021-04-15 18:15:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][320/1251] eta 0:04:33 lr 0.000945 time 0.2653 (0.2942) loss 3.0826 (3.9723) grad_norm 0.9370 (1.0849) [2021-04-15 18:15:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][330/1251] eta 0:04:30 lr 0.000945 time 0.2664 (0.2939) loss 2.7264 (3.9793) grad_norm 1.1300 (1.0850) [2021-04-15 18:15:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][340/1251] eta 0:04:27 lr 0.000945 time 0.2755 (0.2935) loss 3.9956 (3.9772) grad_norm 0.9034 (1.0854) [2021-04-15 18:15:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][350/1251] eta 0:04:24 lr 0.000945 time 0.2849 (0.2930) loss 3.6495 (3.9835) grad_norm 1.0702 (1.0836) [2021-04-15 18:15:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][360/1251] eta 0:04:21 lr 0.000945 time 0.2737 (0.2931) loss 4.0151 (3.9861) grad_norm 1.1213 (1.0825) [2021-04-15 18:15:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][370/1251] eta 0:04:18 lr 0.000945 time 0.2871 (0.2929) loss 3.0835 (3.9808) grad_norm 1.5001 (1.0845) [2021-04-15 18:15:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][380/1251] eta 0:04:14 lr 0.000945 time 0.2673 (0.2924) loss 3.1744 (3.9842) grad_norm 1.0940 (1.0843) [2021-04-15 18:15:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][390/1251] eta 0:04:11 lr 0.000945 time 0.2774 (0.2921) loss 4.8265 (3.9867) grad_norm 1.0547 (1.0834) [2021-04-15 18:15:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][400/1251] eta 0:04:08 lr 0.000945 time 0.2680 (0.2918) loss 4.1924 (3.9874) grad_norm 1.0177 (1.0830) [2021-04-15 18:15:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][410/1251] eta 0:04:05 lr 0.000945 time 0.3063 (0.2920) loss 3.3143 (3.9788) grad_norm 1.2361 (1.0837) [2021-04-15 18:15:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][420/1251] eta 0:04:02 lr 0.000945 time 0.2847 (0.2919) loss 4.5335 (3.9739) grad_norm 1.1795 (1.0837) [2021-04-15 18:15:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][430/1251] eta 0:03:59 lr 0.000945 time 0.2507 (0.2915) loss 3.7159 (3.9748) grad_norm 1.1040 (1.0845) [2021-04-15 18:15:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][440/1251] eta 0:03:56 lr 0.000945 time 0.2818 (0.2912) loss 4.0437 (3.9749) grad_norm 1.0485 (1.0847) [2021-04-15 18:15:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][450/1251] eta 0:03:53 lr 0.000945 time 0.2966 (0.2910) loss 3.1183 (3.9769) grad_norm 1.8477 (1.0873) [2021-04-15 18:15:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][460/1251] eta 0:03:50 lr 0.000945 time 0.2831 (0.2908) loss 3.9188 (3.9752) grad_norm 1.0009 (1.0897) [2021-04-15 18:15:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][470/1251] eta 0:03:47 lr 0.000945 time 0.3039 (0.2907) loss 3.4966 (3.9787) grad_norm 1.3454 (1.0914) [2021-04-15 18:15:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][480/1251] eta 0:03:43 lr 0.000945 time 0.2589 (0.2905) loss 3.7551 (3.9773) grad_norm 1.0519 (1.0901) [2021-04-15 18:15:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][490/1251] eta 0:03:41 lr 0.000945 time 0.2811 (0.2906) loss 4.5126 (3.9762) grad_norm 0.9380 (1.0892) [2021-04-15 18:15:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][500/1251] eta 0:03:38 lr 0.000945 time 0.2828 (0.2904) loss 3.0584 (3.9751) grad_norm 0.8867 (1.0892) [2021-04-15 18:16:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][510/1251] eta 0:03:34 lr 0.000945 time 0.2731 (0.2901) loss 3.5258 (3.9755) grad_norm 1.2137 (1.0906) [2021-04-15 18:16:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][520/1251] eta 0:03:31 lr 0.000945 time 0.2963 (0.2900) loss 2.9404 (3.9752) grad_norm 1.0659 (1.0919) [2021-04-15 18:16:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][530/1251] eta 0:03:28 lr 0.000945 time 0.2831 (0.2897) loss 3.9454 (3.9746) grad_norm 1.0871 (1.0925) [2021-04-15 18:16:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][540/1251] eta 0:03:25 lr 0.000945 time 0.2931 (0.2897) loss 4.3361 (3.9707) grad_norm 1.0740 (1.0919) [2021-04-15 18:16:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][550/1251] eta 0:03:22 lr 0.000945 time 0.2885 (0.2896) loss 3.3859 (3.9715) grad_norm 1.3932 (1.0918) [2021-04-15 18:16:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][560/1251] eta 0:03:19 lr 0.000945 time 0.3024 (0.2894) loss 4.5991 (3.9726) grad_norm 1.0815 (1.0919) [2021-04-15 18:16:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][570/1251] eta 0:03:16 lr 0.000945 time 0.2649 (0.2892) loss 4.2066 (3.9774) grad_norm 1.1805 (1.0936) [2021-04-15 18:16:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][580/1251] eta 0:03:14 lr 0.000945 time 0.2763 (0.2893) loss 4.2019 (3.9823) grad_norm 1.3687 (1.0945) [2021-04-15 18:16:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][590/1251] eta 0:03:11 lr 0.000945 time 0.3022 (0.2893) loss 4.6744 (3.9855) grad_norm 1.1156 (1.0948) [2021-04-15 18:16:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][600/1251] eta 0:03:08 lr 0.000945 time 0.2830 (0.2891) loss 3.1033 (3.9883) grad_norm 0.9282 (1.0955) [2021-04-15 18:16:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][610/1251] eta 0:03:05 lr 0.000945 time 0.2925 (0.2889) loss 3.8715 (3.9882) grad_norm 0.9042 (1.0951) [2021-04-15 18:16:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][620/1251] eta 0:03:02 lr 0.000945 time 0.2871 (0.2888) loss 4.5773 (3.9899) grad_norm 1.2832 (1.0954) [2021-04-15 18:16:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][630/1251] eta 0:02:59 lr 0.000945 time 0.2798 (0.2887) loss 3.8163 (3.9926) grad_norm 0.9715 (1.0956) [2021-04-15 18:16:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][640/1251] eta 0:02:56 lr 0.000945 time 0.2870 (0.2885) loss 3.7712 (3.9925) grad_norm 0.9602 (1.0956) [2021-04-15 18:16:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][650/1251] eta 0:02:53 lr 0.000945 time 0.2749 (0.2883) loss 2.8376 (3.9884) grad_norm 1.0554 (1.0954) [2021-04-15 18:16:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][660/1251] eta 0:02:50 lr 0.000945 time 0.2683 (0.2882) loss 4.6393 (3.9934) grad_norm 0.8817 (1.0943) [2021-04-15 18:16:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][670/1251] eta 0:02:47 lr 0.000945 time 0.2829 (0.2880) loss 3.7831 (3.9904) grad_norm 1.1831 (1.0944) [2021-04-15 18:16:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][680/1251] eta 0:02:44 lr 0.000945 time 0.2611 (0.2879) loss 3.7320 (3.9938) grad_norm 1.2356 (1.0960) [2021-04-15 18:16:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][690/1251] eta 0:02:41 lr 0.000945 time 0.2510 (0.2878) loss 4.3597 (3.9915) grad_norm 1.1523 (1.0967) [2021-04-15 18:16:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][700/1251] eta 0:02:38 lr 0.000945 time 0.2907 (0.2878) loss 3.4873 (3.9900) grad_norm 1.5942 (1.0980) [2021-04-15 18:16:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][710/1251] eta 0:02:35 lr 0.000945 time 0.2855 (0.2877) loss 3.8992 (3.9929) grad_norm 1.0527 (1.0976) [2021-04-15 18:16:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][720/1251] eta 0:02:32 lr 0.000945 time 0.3023 (0.2876) loss 3.6645 (3.9901) grad_norm 0.9203 (1.0963) [2021-04-15 18:17:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][730/1251] eta 0:02:29 lr 0.000945 time 0.2634 (0.2877) loss 3.8564 (3.9948) grad_norm 1.1024 (1.0972) [2021-04-15 18:17:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][740/1251] eta 0:02:26 lr 0.000945 time 0.3278 (0.2876) loss 4.6449 (3.9958) grad_norm 1.3069 (1.0997) [2021-04-15 18:17:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][750/1251] eta 0:02:23 lr 0.000945 time 0.2641 (0.2873) loss 3.6129 (3.9965) grad_norm 0.9175 (1.0994) [2021-04-15 18:17:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][760/1251] eta 0:02:21 lr 0.000945 time 0.2858 (0.2873) loss 3.2707 (3.9982) grad_norm 1.0319 (1.0989) [2021-04-15 18:17:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][770/1251] eta 0:02:18 lr 0.000945 time 0.2849 (0.2872) loss 4.6604 (3.9958) grad_norm 1.0258 (1.0994) [2021-04-15 18:17:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][780/1251] eta 0:02:15 lr 0.000945 time 0.2854 (0.2872) loss 3.8796 (3.9894) grad_norm 1.0379 (1.0985) [2021-04-15 18:17:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][790/1251] eta 0:02:12 lr 0.000945 time 0.2948 (0.2871) loss 3.9704 (3.9867) grad_norm 1.1962 (1.0977) [2021-04-15 18:17:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][800/1251] eta 0:02:09 lr 0.000945 time 0.2999 (0.2870) loss 3.9805 (3.9852) grad_norm 0.9829 (1.0979) [2021-04-15 18:17:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][810/1251] eta 0:02:06 lr 0.000945 time 0.2489 (0.2870) loss 3.3847 (3.9845) grad_norm 1.2013 (1.0976) [2021-04-15 18:17:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][820/1251] eta 0:02:03 lr 0.000944 time 0.2615 (0.2871) loss 4.2356 (3.9833) grad_norm 1.4511 (1.0976) [2021-04-15 18:17:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][830/1251] eta 0:02:00 lr 0.000944 time 0.2884 (0.2871) loss 3.9711 (3.9850) grad_norm 0.9566 (1.0977) [2021-04-15 18:17:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][840/1251] eta 0:01:58 lr 0.000944 time 0.2991 (0.2872) loss 3.9026 (3.9849) grad_norm 1.0632 (1.0979) [2021-04-15 18:17:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][850/1251] eta 0:01:55 lr 0.000944 time 0.2679 (0.2870) loss 2.9141 (3.9818) grad_norm 1.1287 (1.0972) [2021-04-15 18:17:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][860/1251] eta 0:01:52 lr 0.000944 time 0.2890 (0.2869) loss 4.6865 (3.9838) grad_norm 1.0773 (1.0963) [2021-04-15 18:17:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][870/1251] eta 0:01:49 lr 0.000944 time 0.2645 (0.2868) loss 2.9665 (3.9804) grad_norm 1.0472 (1.0955) [2021-04-15 18:17:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][880/1251] eta 0:01:46 lr 0.000944 time 0.2848 (0.2867) loss 3.5289 (3.9801) grad_norm 1.3716 (1.0958) [2021-04-15 18:17:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][890/1251] eta 0:01:43 lr 0.000944 time 0.2649 (0.2866) loss 4.6255 (3.9780) grad_norm 1.0931 (1.0955) [2021-04-15 18:17:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][900/1251] eta 0:01:40 lr 0.000944 time 0.2649 (0.2866) loss 3.7482 (3.9822) grad_norm 1.1373 (1.0951) [2021-04-15 18:17:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][910/1251] eta 0:01:37 lr 0.000944 time 0.2823 (0.2865) loss 4.1632 (3.9830) grad_norm 1.1824 (1.0947) [2021-04-15 18:17:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][920/1251] eta 0:01:34 lr 0.000944 time 0.2741 (0.2865) loss 3.7242 (3.9813) grad_norm 0.9985 (1.0944) [2021-04-15 18:17:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][930/1251] eta 0:01:31 lr 0.000944 time 0.2767 (0.2864) loss 3.3275 (3.9791) grad_norm 1.1888 (1.0935) [2021-04-15 18:18:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][940/1251] eta 0:01:29 lr 0.000944 time 0.2724 (0.2864) loss 3.6138 (3.9782) grad_norm 1.2961 (1.0931) [2021-04-15 18:18:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][950/1251] eta 0:01:26 lr 0.000944 time 0.2533 (0.2862) loss 3.8963 (3.9806) grad_norm 0.9834 (1.0926) [2021-04-15 18:18:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][960/1251] eta 0:01:23 lr 0.000944 time 0.2661 (0.2862) loss 4.4898 (3.9788) grad_norm 0.9531 (1.0922) [2021-04-15 18:18:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][970/1251] eta 0:01:20 lr 0.000944 time 0.2839 (0.2861) loss 4.2245 (3.9779) grad_norm 0.9516 (1.0913) [2021-04-15 18:18:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][980/1251] eta 0:01:17 lr 0.000944 time 0.2664 (0.2860) loss 4.7495 (3.9786) grad_norm 1.1183 (1.0910) [2021-04-15 18:18:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][990/1251] eta 0:01:14 lr 0.000944 time 0.2728 (0.2859) loss 4.2594 (3.9839) grad_norm 1.0260 (1.0902) [2021-04-15 18:18:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][1000/1251] eta 0:01:11 lr 0.000944 time 0.2638 (0.2858) loss 4.8092 (3.9859) grad_norm 0.9389 (1.0898) [2021-04-15 18:18:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][1010/1251] eta 0:01:08 lr 0.000944 time 0.2755 (0.2857) loss 4.5412 (3.9844) grad_norm 1.3170 (1.0902) [2021-04-15 18:18:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][1020/1251] eta 0:01:05 lr 0.000944 time 0.2794 (0.2856) loss 3.5192 (3.9807) grad_norm 1.1722 (1.0906) [2021-04-15 18:18:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][1030/1251] eta 0:01:03 lr 0.000944 time 0.2609 (0.2856) loss 3.8041 (3.9803) grad_norm 1.1023 (1.0911) [2021-04-15 18:18:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][1040/1251] eta 0:01:00 lr 0.000944 time 0.3242 (0.2856) loss 3.8655 (3.9801) grad_norm 0.9015 (1.0920) [2021-04-15 18:18:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][1050/1251] eta 0:00:57 lr 0.000944 time 0.2784 (0.2855) loss 4.4055 (3.9810) grad_norm 1.4470 (1.0929) [2021-04-15 18:18:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][1060/1251] eta 0:00:54 lr 0.000944 time 0.2932 (0.2855) loss 4.0864 (3.9813) grad_norm 1.0468 (1.0935) [2021-04-15 18:18:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][1070/1251] eta 0:00:51 lr 0.000944 time 0.2631 (0.2855) loss 4.1375 (3.9808) grad_norm 1.1913 (1.0936) [2021-04-15 18:18:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][1080/1251] eta 0:00:48 lr 0.000944 time 0.3105 (0.2855) loss 3.8471 (3.9797) grad_norm 0.9753 (1.0928) [2021-04-15 18:18:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][1090/1251] eta 0:00:45 lr 0.000944 time 0.2520 (0.2855) loss 3.6207 (3.9801) grad_norm 0.9602 (1.0925) [2021-04-15 18:18:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][1100/1251] eta 0:00:43 lr 0.000944 time 0.2784 (0.2854) loss 4.0499 (3.9780) grad_norm 0.9533 (1.0921) [2021-04-15 18:18:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][1110/1251] eta 0:00:40 lr 0.000944 time 0.3124 (0.2854) loss 3.7739 (3.9782) grad_norm 0.9790 (1.0919) [2021-04-15 18:18:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][1120/1251] eta 0:00:37 lr 0.000944 time 0.2868 (0.2854) loss 3.1950 (3.9770) grad_norm 0.9434 (1.0910) [2021-04-15 18:18:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][1130/1251] eta 0:00:34 lr 0.000944 time 0.2901 (0.2854) loss 3.1252 (3.9769) grad_norm 1.1417 (1.0910) [2021-04-15 18:18:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][1140/1251] eta 0:00:31 lr 0.000944 time 0.2890 (0.2854) loss 3.9899 (3.9761) grad_norm 0.9167 (1.0906) [2021-04-15 18:19:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][1150/1251] eta 0:00:28 lr 0.000944 time 0.2805 (0.2855) loss 3.8695 (3.9759) grad_norm 1.0686 (1.0909) [2021-04-15 18:19:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][1160/1251] eta 0:00:25 lr 0.000944 time 0.2804 (0.2854) loss 4.7000 (3.9772) grad_norm 1.2576 (1.0912) [2021-04-15 18:19:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][1170/1251] eta 0:00:23 lr 0.000944 time 0.2663 (0.2853) loss 3.5131 (3.9769) grad_norm 1.0734 (1.0916) [2021-04-15 18:19:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][1180/1251] eta 0:00:20 lr 0.000944 time 0.2913 (0.2853) loss 4.2881 (3.9776) grad_norm 1.3317 (1.0919) [2021-04-15 18:19:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][1190/1251] eta 0:00:17 lr 0.000944 time 0.3022 (0.2853) loss 3.0188 (3.9771) grad_norm 1.2478 (1.0927) [2021-04-15 18:19:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][1200/1251] eta 0:00:14 lr 0.000944 time 0.2853 (0.2852) loss 4.1181 (3.9768) grad_norm 1.2534 (1.0927) [2021-04-15 18:19:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][1210/1251] eta 0:00:11 lr 0.000944 time 0.2861 (0.2851) loss 3.8427 (3.9780) grad_norm 0.9749 (1.0925) [2021-04-15 18:19:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][1220/1251] eta 0:00:08 lr 0.000944 time 0.2817 (0.2851) loss 3.0544 (3.9792) grad_norm 0.8963 (1.0925) [2021-04-15 18:19:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][1230/1251] eta 0:00:05 lr 0.000944 time 0.2697 (0.2850) loss 3.1039 (3.9786) grad_norm 1.0315 (1.0920) [2021-04-15 18:19:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][1240/1251] eta 0:00:03 lr 0.000944 time 0.2492 (0.2849) loss 3.5475 (3.9766) grad_norm 1.0463 (1.0910) [2021-04-15 18:19:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [45/300][1250/1251] eta 0:00:00 lr 0.000944 time 0.2493 (0.2846) loss 3.7703 (3.9768) grad_norm 1.3014 (1.0911) [2021-04-15 18:19:30 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 45 training takes 0:05:58 [2021-04-15 18:19:30 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_45.pth saving...... [2021-04-15 18:19:41 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_45.pth saved !!! [2021-04-15 18:19:43 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.196 (1.196) Loss 1.3228 (1.3228) Acc@1 67.578 (67.578) Acc@5 88.770 (88.770) [2021-04-15 18:19:44 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.136 (0.226) Loss 1.3178 (1.2879) Acc@1 69.531 (69.558) Acc@5 88.965 (89.933) [2021-04-15 18:19:46 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.289 (0.224) Loss 1.2141 (1.3065) Acc@1 71.680 (69.127) Acc@5 91.309 (89.639) [2021-04-15 18:19:49 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.123 (0.235) Loss 1.3290 (1.3104) Acc@1 67.188 (69.241) Acc@5 89.453 (89.655) [2021-04-15 18:19:50 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.248 (0.220) Loss 1.2715 (1.3079) Acc@1 68.457 (69.195) Acc@5 90.039 (89.691) [2021-04-15 18:19:53 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 69.284 Acc@5 89.732 [2021-04-15 18:19:53 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 69.3% [2021-04-15 18:19:53 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 69.28% [2021-04-15 18:19:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][0/1251] eta 1:26:15 lr 0.000944 time 4.1369 (4.1369) loss 2.7936 (2.7936) grad_norm 1.1606 (1.1606) [2021-04-15 18:20:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][10/1251] eta 0:13:03 lr 0.000944 time 0.2941 (0.6312) loss 4.5774 (4.0295) grad_norm 1.3439 (1.2026) [2021-04-15 18:20:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][20/1251] eta 0:09:36 lr 0.000944 time 0.2900 (0.4681) loss 3.7212 (3.9855) grad_norm 1.0018 (1.1272) [2021-04-15 18:20:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][30/1251] eta 0:08:21 lr 0.000944 time 0.2676 (0.4104) loss 4.4149 (3.8969) grad_norm 1.0988 (1.1034) [2021-04-15 18:20:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][40/1251] eta 0:07:45 lr 0.000944 time 0.2659 (0.3842) loss 2.7129 (3.8418) grad_norm 0.9767 (1.0943) [2021-04-15 18:20:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][50/1251] eta 0:07:17 lr 0.000944 time 0.2601 (0.3640) loss 4.1006 (3.8275) grad_norm 1.2530 (1.0903) [2021-04-15 18:20:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][60/1251] eta 0:06:57 lr 0.000944 time 0.2888 (0.3507) loss 4.2781 (3.8172) grad_norm 1.1251 (1.0856) [2021-04-15 18:20:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][70/1251] eta 0:06:42 lr 0.000944 time 0.2752 (0.3410) loss 4.2315 (3.8065) grad_norm 1.1759 (1.0889) [2021-04-15 18:20:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][80/1251] eta 0:06:30 lr 0.000944 time 0.2749 (0.3335) loss 4.7419 (3.7888) grad_norm 1.2493 (1.0952) [2021-04-15 18:20:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][90/1251] eta 0:06:20 lr 0.000943 time 0.3000 (0.3276) loss 4.3577 (3.8273) grad_norm 0.9314 (1.1100) [2021-04-15 18:20:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][100/1251] eta 0:06:11 lr 0.000943 time 0.2718 (0.3227) loss 4.2285 (3.8759) grad_norm 1.2642 (1.1183) [2021-04-15 18:20:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][110/1251] eta 0:06:03 lr 0.000943 time 0.2808 (0.3187) loss 4.4965 (3.9092) grad_norm 1.0013 (1.1136) [2021-04-15 18:20:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][120/1251] eta 0:05:56 lr 0.000943 time 0.3141 (0.3152) loss 4.4366 (3.9357) grad_norm 1.0558 (1.1089) [2021-04-15 18:20:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][130/1251] eta 0:05:50 lr 0.000943 time 0.2923 (0.3130) loss 4.9871 (3.9489) grad_norm 1.0914 (1.1090) [2021-04-15 18:20:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][140/1251] eta 0:05:45 lr 0.000943 time 0.2656 (0.3106) loss 2.9677 (3.9324) grad_norm 1.0922 (1.1128) [2021-04-15 18:20:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][150/1251] eta 0:05:40 lr 0.000943 time 0.2905 (0.3096) loss 3.2789 (3.9306) grad_norm 1.0764 (1.1074) [2021-04-15 18:20:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][160/1251] eta 0:05:35 lr 0.000943 time 0.2645 (0.3077) loss 2.3546 (3.9172) grad_norm 0.8934 (1.1074) [2021-04-15 18:20:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][170/1251] eta 0:05:30 lr 0.000943 time 0.2717 (0.3062) loss 4.6891 (3.9238) grad_norm 1.0334 (1.1060) [2021-04-15 18:20:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][180/1251] eta 0:05:26 lr 0.000943 time 0.2505 (0.3050) loss 2.5127 (3.9431) grad_norm 1.0685 (1.1082) [2021-04-15 18:20:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][190/1251] eta 0:05:22 lr 0.000943 time 0.2750 (0.3038) loss 3.5808 (3.9378) grad_norm 1.1417 (1.1092) [2021-04-15 18:20:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][200/1251] eta 0:05:18 lr 0.000943 time 0.2959 (0.3027) loss 4.2931 (3.9521) grad_norm 0.9885 (1.1076) [2021-04-15 18:20:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][210/1251] eta 0:05:14 lr 0.000943 time 0.2706 (0.3017) loss 4.3368 (3.9556) grad_norm 1.2680 (1.1077) [2021-04-15 18:21:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][220/1251] eta 0:05:09 lr 0.000943 time 0.2653 (0.3007) loss 3.6976 (3.9517) grad_norm 1.1750 (1.1086) [2021-04-15 18:21:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][230/1251] eta 0:05:06 lr 0.000943 time 0.2996 (0.2998) loss 2.6874 (3.9421) grad_norm 1.3396 (1.1109) [2021-04-15 18:21:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][240/1251] eta 0:05:02 lr 0.000943 time 0.2849 (0.2990) loss 4.2044 (3.9401) grad_norm 1.4467 (1.1148) [2021-04-15 18:21:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][250/1251] eta 0:04:58 lr 0.000943 time 0.2895 (0.2982) loss 4.3130 (3.9421) grad_norm 0.9529 (1.1163) [2021-04-15 18:21:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][260/1251] eta 0:04:54 lr 0.000943 time 0.2616 (0.2975) loss 3.7365 (3.9408) grad_norm 1.2635 (1.1181) [2021-04-15 18:21:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][270/1251] eta 0:04:51 lr 0.000943 time 0.2967 (0.2969) loss 3.0683 (3.9521) grad_norm 0.9447 (1.1183) [2021-04-15 18:21:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][280/1251] eta 0:04:47 lr 0.000943 time 0.2703 (0.2962) loss 3.9496 (3.9500) grad_norm 1.3274 (1.1175) [2021-04-15 18:21:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][290/1251] eta 0:04:44 lr 0.000943 time 0.2725 (0.2961) loss 3.7525 (3.9550) grad_norm 1.2337 (1.1186) [2021-04-15 18:21:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][300/1251] eta 0:04:41 lr 0.000943 time 0.2496 (0.2956) loss 3.7613 (3.9621) grad_norm 1.1792 (1.1163) [2021-04-15 18:21:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][310/1251] eta 0:04:37 lr 0.000943 time 0.2792 (0.2953) loss 4.4679 (3.9672) grad_norm 1.0313 (1.1141) [2021-04-15 18:21:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][320/1251] eta 0:04:34 lr 0.000943 time 0.2890 (0.2951) loss 4.8137 (3.9613) grad_norm 0.9239 (1.1116) [2021-04-15 18:21:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][330/1251] eta 0:04:31 lr 0.000943 time 0.2958 (0.2946) loss 4.3818 (3.9704) grad_norm 1.0855 (1.1101) [2021-04-15 18:21:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][340/1251] eta 0:04:28 lr 0.000943 time 0.3023 (0.2942) loss 4.3087 (3.9697) grad_norm 0.9912 (1.1090) [2021-04-15 18:21:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][350/1251] eta 0:04:24 lr 0.000943 time 0.2639 (0.2937) loss 3.4102 (3.9632) grad_norm 1.0445 (1.1095) [2021-04-15 18:21:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][360/1251] eta 0:04:21 lr 0.000943 time 0.2606 (0.2938) loss 3.3960 (3.9604) grad_norm 1.0794 (1.1092) [2021-04-15 18:21:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][370/1251] eta 0:04:18 lr 0.000943 time 0.2952 (0.2935) loss 2.6442 (3.9520) grad_norm 0.9007 (1.1069) [2021-04-15 18:21:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][380/1251] eta 0:04:15 lr 0.000943 time 0.2858 (0.2932) loss 3.5083 (3.9524) grad_norm 0.9722 (1.1048) [2021-04-15 18:21:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][390/1251] eta 0:04:12 lr 0.000943 time 0.2674 (0.2927) loss 3.7700 (3.9503) grad_norm 1.1463 (1.1041) [2021-04-15 18:21:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][400/1251] eta 0:04:08 lr 0.000943 time 0.2747 (0.2924) loss 3.8550 (3.9641) grad_norm 1.0603 (1.1035) [2021-04-15 18:21:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][410/1251] eta 0:04:05 lr 0.000943 time 0.2686 (0.2920) loss 4.3188 (3.9621) grad_norm 1.3683 (1.1036) [2021-04-15 18:21:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][420/1251] eta 0:04:02 lr 0.000943 time 0.2660 (0.2916) loss 4.4173 (3.9628) grad_norm 1.0952 (1.1033) [2021-04-15 18:21:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][430/1251] eta 0:03:59 lr 0.000943 time 0.2861 (0.2914) loss 4.9272 (3.9647) grad_norm 0.9619 (1.1021) [2021-04-15 18:22:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][440/1251] eta 0:03:56 lr 0.000943 time 0.2794 (0.2911) loss 3.6692 (3.9631) grad_norm 1.0037 (1.1002) [2021-04-15 18:22:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][450/1251] eta 0:03:52 lr 0.000943 time 0.2681 (0.2908) loss 4.5598 (3.9669) grad_norm 0.9172 (1.0977) [2021-04-15 18:22:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][460/1251] eta 0:03:49 lr 0.000943 time 0.2857 (0.2906) loss 4.7968 (3.9664) grad_norm 0.9331 (1.0979) [2021-04-15 18:22:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][470/1251] eta 0:03:46 lr 0.000943 time 0.2756 (0.2903) loss 4.7074 (3.9657) grad_norm 1.0208 (1.0967) [2021-04-15 18:22:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][480/1251] eta 0:03:43 lr 0.000943 time 0.3090 (0.2902) loss 3.4992 (3.9713) grad_norm 0.9646 (1.0969) [2021-04-15 18:22:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][490/1251] eta 0:03:40 lr 0.000943 time 0.3241 (0.2900) loss 3.2431 (3.9703) grad_norm 1.2438 (1.1009) [2021-04-15 18:22:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][500/1251] eta 0:03:37 lr 0.000943 time 0.3079 (0.2900) loss 3.5820 (3.9706) grad_norm 0.9779 (1.1022) [2021-04-15 18:22:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][510/1251] eta 0:03:34 lr 0.000943 time 0.2802 (0.2898) loss 3.3568 (3.9661) grad_norm 1.0520 (1.1012) [2021-04-15 18:22:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][520/1251] eta 0:03:31 lr 0.000943 time 0.3012 (0.2897) loss 4.8537 (3.9624) grad_norm 1.1208 (1.1007) [2021-04-15 18:22:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][530/1251] eta 0:03:28 lr 0.000943 time 0.2794 (0.2895) loss 3.0286 (3.9638) grad_norm 1.2598 (1.1023) [2021-04-15 18:22:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][540/1251] eta 0:03:26 lr 0.000943 time 0.2824 (0.2898) loss 4.2269 (3.9677) grad_norm 1.1719 (1.1045) [2021-04-15 18:22:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][550/1251] eta 0:03:22 lr 0.000943 time 0.2795 (0.2896) loss 2.8803 (3.9622) grad_norm 1.0324 (1.1075) [2021-04-15 18:22:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][560/1251] eta 0:03:20 lr 0.000943 time 0.3052 (0.2895) loss 2.9397 (3.9590) grad_norm 1.0265 (1.1078) [2021-04-15 18:22:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][570/1251] eta 0:03:16 lr 0.000943 time 0.2858 (0.2893) loss 3.3224 (3.9600) grad_norm 1.0681 (1.1082) [2021-04-15 18:22:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][580/1251] eta 0:03:13 lr 0.000943 time 0.2658 (0.2890) loss 4.2235 (3.9629) grad_norm 1.0027 (1.1097) [2021-04-15 18:22:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][590/1251] eta 0:03:11 lr 0.000943 time 0.2849 (0.2891) loss 3.1704 (3.9619) grad_norm 1.0677 (1.1096) [2021-04-15 18:22:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][600/1251] eta 0:03:08 lr 0.000943 time 0.2688 (0.2889) loss 3.9321 (3.9611) grad_norm 1.1346 (1.1108) [2021-04-15 18:22:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][610/1251] eta 0:03:05 lr 0.000942 time 0.2980 (0.2891) loss 3.3836 (3.9602) grad_norm 1.0453 (1.1124) [2021-04-15 18:22:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][620/1251] eta 0:03:02 lr 0.000942 time 0.2831 (0.2891) loss 4.9436 (3.9686) grad_norm 1.2607 (1.1119) [2021-04-15 18:22:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][630/1251] eta 0:02:59 lr 0.000942 time 0.2746 (0.2889) loss 4.7101 (3.9705) grad_norm 1.1963 (1.1109) [2021-04-15 18:22:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][640/1251] eta 0:02:56 lr 0.000942 time 0.2718 (0.2887) loss 4.8046 (3.9715) grad_norm 1.0674 (1.1110) [2021-04-15 18:23:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][650/1251] eta 0:02:53 lr 0.000942 time 0.2700 (0.2886) loss 2.8954 (3.9714) grad_norm 0.9652 (1.1098) [2021-04-15 18:23:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][660/1251] eta 0:02:50 lr 0.000942 time 0.2964 (0.2885) loss 4.5535 (3.9719) grad_norm 0.9428 (1.1092) [2021-04-15 18:23:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][670/1251] eta 0:02:47 lr 0.000942 time 0.3155 (0.2884) loss 2.5496 (3.9724) grad_norm 1.1368 (1.1091) [2021-04-15 18:23:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][680/1251] eta 0:02:44 lr 0.000942 time 0.2841 (0.2883) loss 4.3951 (3.9732) grad_norm 1.2905 (1.1095) [2021-04-15 18:23:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][690/1251] eta 0:02:41 lr 0.000942 time 0.2810 (0.2883) loss 3.3608 (3.9741) grad_norm 1.2550 (1.1094) [2021-04-15 18:23:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][700/1251] eta 0:02:38 lr 0.000942 time 0.2935 (0.2881) loss 4.3163 (3.9766) grad_norm 1.0219 (1.1083) [2021-04-15 18:23:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][710/1251] eta 0:02:35 lr 0.000942 time 0.2961 (0.2880) loss 3.9149 (3.9753) grad_norm 1.1748 (1.1084) [2021-04-15 18:23:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][720/1251] eta 0:02:33 lr 0.000942 time 0.2620 (0.2881) loss 4.4809 (3.9766) grad_norm 1.0457 (1.1074) [2021-04-15 18:23:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][730/1251] eta 0:02:29 lr 0.000942 time 0.2795 (0.2879) loss 3.2678 (3.9755) grad_norm 1.1159 (1.1069) [2021-04-15 18:23:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][740/1251] eta 0:02:27 lr 0.000942 time 0.3048 (0.2879) loss 4.4279 (3.9812) grad_norm 1.0401 (1.1065) [2021-04-15 18:23:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][750/1251] eta 0:02:24 lr 0.000942 time 0.2590 (0.2877) loss 4.7032 (3.9820) grad_norm 0.8640 (1.1059) [2021-04-15 18:23:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][760/1251] eta 0:02:21 lr 0.000942 time 0.2548 (0.2876) loss 4.0383 (3.9775) grad_norm 1.1717 (1.1058) [2021-04-15 18:23:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][770/1251] eta 0:02:18 lr 0.000942 time 0.2649 (0.2875) loss 2.5794 (3.9719) grad_norm 1.1736 (1.1059) [2021-04-15 18:23:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][780/1251] eta 0:02:15 lr 0.000942 time 0.2638 (0.2875) loss 2.4985 (3.9701) grad_norm 1.0533 (1.1055) [2021-04-15 18:23:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][790/1251] eta 0:02:12 lr 0.000942 time 0.2888 (0.2874) loss 4.0145 (3.9714) grad_norm 1.0432 (1.1049) [2021-04-15 18:23:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][800/1251] eta 0:02:09 lr 0.000942 time 0.2718 (0.2873) loss 4.0800 (3.9768) grad_norm 1.0429 (1.1047) [2021-04-15 18:23:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][810/1251] eta 0:02:06 lr 0.000942 time 0.2766 (0.2872) loss 4.2047 (3.9782) grad_norm 1.1892 (1.1047) [2021-04-15 18:23:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][820/1251] eta 0:02:03 lr 0.000942 time 0.2773 (0.2871) loss 3.8867 (3.9788) grad_norm 0.9915 (1.1032) [2021-04-15 18:23:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][830/1251] eta 0:02:00 lr 0.000942 time 0.2796 (0.2870) loss 3.7481 (3.9778) grad_norm 1.1071 (1.1034) [2021-04-15 18:23:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][840/1251] eta 0:01:57 lr 0.000942 time 0.2579 (0.2869) loss 3.8071 (3.9798) grad_norm 1.0144 (1.1027) [2021-04-15 18:23:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][850/1251] eta 0:01:54 lr 0.000942 time 0.2629 (0.2868) loss 3.2069 (3.9808) grad_norm 0.9265 (1.1024) [2021-04-15 18:24:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][860/1251] eta 0:01:52 lr 0.000942 time 0.2689 (0.2867) loss 3.0148 (3.9817) grad_norm 1.2383 (1.1029) [2021-04-15 18:24:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][870/1251] eta 0:01:49 lr 0.000942 time 0.2909 (0.2868) loss 4.3431 (3.9829) grad_norm 1.6869 (1.1045) [2021-04-15 18:24:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][880/1251] eta 0:01:46 lr 0.000942 time 0.2684 (0.2867) loss 4.4372 (3.9835) grad_norm 0.9637 (1.1044) [2021-04-15 18:24:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][890/1251] eta 0:01:43 lr 0.000942 time 0.2742 (0.2867) loss 4.3683 (3.9831) grad_norm 1.0078 (1.1035) [2021-04-15 18:24:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][900/1251] eta 0:01:40 lr 0.000942 time 0.2848 (0.2866) loss 4.0919 (3.9840) grad_norm 1.1071 (1.1033) [2021-04-15 18:24:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][910/1251] eta 0:01:37 lr 0.000942 time 0.2616 (0.2866) loss 4.8165 (3.9859) grad_norm 1.0129 (1.1026) [2021-04-15 18:24:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][920/1251] eta 0:01:34 lr 0.000942 time 0.2786 (0.2865) loss 2.8401 (3.9857) grad_norm 0.9582 (1.1022) [2021-04-15 18:24:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][930/1251] eta 0:01:31 lr 0.000942 time 0.2606 (0.2864) loss 4.7784 (3.9860) grad_norm 1.1619 (1.1023) [2021-04-15 18:24:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][940/1251] eta 0:01:29 lr 0.000942 time 0.2676 (0.2865) loss 4.4110 (3.9856) grad_norm 1.0756 (1.1028) [2021-04-15 18:24:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][950/1251] eta 0:01:26 lr 0.000942 time 0.2728 (0.2864) loss 3.5666 (3.9877) grad_norm 1.1382 (1.1033) [2021-04-15 18:24:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][960/1251] eta 0:01:23 lr 0.000942 time 0.3004 (0.2863) loss 4.5080 (3.9891) grad_norm 1.2907 (1.1036) [2021-04-15 18:24:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][970/1251] eta 0:01:20 lr 0.000942 time 0.2643 (0.2862) loss 2.8831 (3.9882) grad_norm 0.9705 (1.1035) [2021-04-15 18:24:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][980/1251] eta 0:01:17 lr 0.000942 time 0.2504 (0.2862) loss 4.0103 (3.9894) grad_norm 1.4040 (1.1033) [2021-04-15 18:24:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][990/1251] eta 0:01:14 lr 0.000942 time 0.2751 (0.2861) loss 3.9777 (3.9890) grad_norm 0.9392 (1.1031) [2021-04-15 18:24:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][1000/1251] eta 0:01:11 lr 0.000942 time 0.2702 (0.2860) loss 3.9795 (3.9872) grad_norm 1.1224 (1.1026) [2021-04-15 18:24:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][1010/1251] eta 0:01:08 lr 0.000942 time 0.2904 (0.2860) loss 2.8265 (3.9860) grad_norm 0.9472 (1.1026) [2021-04-15 18:24:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][1020/1251] eta 0:01:06 lr 0.000942 time 0.2852 (0.2859) loss 3.5012 (3.9864) grad_norm 1.1947 (1.1038) [2021-04-15 18:24:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][1030/1251] eta 0:01:03 lr 0.000942 time 0.2583 (0.2858) loss 4.5390 (3.9866) grad_norm 1.1792 (1.1037) [2021-04-15 18:24:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][1040/1251] eta 0:01:00 lr 0.000942 time 0.2814 (0.2857) loss 3.7094 (3.9877) grad_norm 1.2702 (1.1041) [2021-04-15 18:24:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][1050/1251] eta 0:00:57 lr 0.000942 time 0.2615 (0.2856) loss 4.3310 (3.9885) grad_norm 1.0705 (1.1041) [2021-04-15 18:24:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][1060/1251] eta 0:00:54 lr 0.000942 time 0.2717 (0.2856) loss 3.4667 (3.9871) grad_norm 1.1173 (1.1036) [2021-04-15 18:24:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][1070/1251] eta 0:00:51 lr 0.000942 time 0.2734 (0.2855) loss 3.7331 (3.9857) grad_norm 1.1436 (1.1030) [2021-04-15 18:25:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][1080/1251] eta 0:00:48 lr 0.000942 time 0.2723 (0.2854) loss 3.7768 (3.9867) grad_norm 0.9079 (1.1024) [2021-04-15 18:25:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][1090/1251] eta 0:00:45 lr 0.000942 time 0.2784 (0.2854) loss 3.2070 (3.9879) grad_norm 0.9991 (1.1025) [2021-04-15 18:25:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][1100/1251] eta 0:00:43 lr 0.000942 time 0.2694 (0.2853) loss 2.7248 (3.9869) grad_norm 1.1966 (1.1028) [2021-04-15 18:25:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][1110/1251] eta 0:00:40 lr 0.000942 time 0.2850 (0.2853) loss 4.8484 (3.9851) grad_norm 1.0146 (1.1026) [2021-04-15 18:25:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][1120/1251] eta 0:00:37 lr 0.000942 time 0.2729 (0.2852) loss 2.9657 (3.9852) grad_norm 0.9237 (1.1021) [2021-04-15 18:25:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][1130/1251] eta 0:00:34 lr 0.000941 time 0.2626 (0.2852) loss 4.3343 (3.9854) grad_norm 1.0976 (1.1013) [2021-04-15 18:25:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][1140/1251] eta 0:00:31 lr 0.000941 time 0.2804 (0.2851) loss 3.7928 (3.9860) grad_norm 1.1293 (1.1012) [2021-04-15 18:25:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][1150/1251] eta 0:00:28 lr 0.000941 time 0.2976 (0.2850) loss 4.1598 (3.9847) grad_norm 1.0210 (1.1013) [2021-04-15 18:25:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][1160/1251] eta 0:00:25 lr 0.000941 time 0.2718 (0.2850) loss 4.4574 (3.9887) grad_norm 0.9884 (1.1012) [2021-04-15 18:25:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][1170/1251] eta 0:00:23 lr 0.000941 time 0.2788 (0.2850) loss 2.8244 (3.9891) grad_norm 1.0016 (1.1008) [2021-04-15 18:25:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][1180/1251] eta 0:00:20 lr 0.000941 time 0.2765 (0.2851) loss 2.8456 (3.9865) grad_norm 1.2909 (1.1008) [2021-04-15 18:25:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][1190/1251] eta 0:00:17 lr 0.000941 time 0.2848 (0.2851) loss 3.0350 (3.9869) grad_norm 1.2432 (1.1008) [2021-04-15 18:25:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][1200/1251] eta 0:00:14 lr 0.000941 time 0.2934 (0.2850) loss 4.5324 (3.9881) grad_norm 1.2343 (1.1010) [2021-04-15 18:25:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][1210/1251] eta 0:00:11 lr 0.000941 time 0.2711 (0.2850) loss 4.4486 (3.9908) grad_norm 1.0465 (1.1012) [2021-04-15 18:25:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][1220/1251] eta 0:00:08 lr 0.000941 time 0.2655 (0.2849) loss 4.3027 (3.9898) grad_norm 0.9810 (1.1009) [2021-04-15 18:25:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][1230/1251] eta 0:00:05 lr 0.000941 time 0.2842 (0.2849) loss 4.0554 (3.9904) grad_norm 1.2317 (1.1005) [2021-04-15 18:25:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][1240/1251] eta 0:00:03 lr 0.000941 time 0.2518 (0.2847) loss 4.3990 (3.9883) grad_norm 1.0242 (1.1004) [2021-04-15 18:25:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [46/300][1250/1251] eta 0:00:00 lr 0.000941 time 0.2488 (0.2844) loss 3.6765 (3.9865) grad_norm 1.0428 (1.1001) [2021-04-15 18:25:52 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 46 training takes 0:05:58 [2021-04-15 18:25:52 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_46.pth saving...... [2021-04-15 18:26:08 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_46.pth saved !!! [2021-04-15 18:26:09 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.091 (1.091) Loss 1.3974 (1.3974) Acc@1 67.480 (67.480) Acc@5 88.965 (88.965) [2021-04-15 18:26:10 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.126 (0.245) Loss 1.3400 (1.3276) Acc@1 69.141 (69.407) Acc@5 90.625 (90.110) [2021-04-15 18:26:12 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.384 (0.221) Loss 1.3273 (1.3572) Acc@1 70.703 (69.127) Acc@5 90.137 (89.462) [2021-04-15 18:26:15 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.120 (0.241) Loss 1.3421 (1.3473) Acc@1 69.629 (69.336) Acc@5 90.137 (89.570) [2021-04-15 18:26:17 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.217) Loss 1.3385 (1.3444) Acc@1 69.043 (69.410) Acc@5 89.453 (89.570) [2021-04-15 18:26:20 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 69.494 Acc@5 89.628 [2021-04-15 18:26:20 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 69.5% [2021-04-15 18:26:20 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 69.49% [2021-04-15 18:26:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][0/1251] eta 1:28:39 lr 0.000941 time 4.2521 (4.2521) loss 3.5790 (3.5790) grad_norm 0.9468 (0.9468) [2021-04-15 18:26:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][10/1251] eta 0:13:19 lr 0.000941 time 0.2932 (0.6444) loss 3.9273 (3.8032) grad_norm 1.2881 (1.0471) [2021-04-15 18:26:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][20/1251] eta 0:09:41 lr 0.000941 time 0.2865 (0.4726) loss 4.3177 (3.9150) grad_norm 1.1185 (1.0961) [2021-04-15 18:26:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][30/1251] eta 0:08:22 lr 0.000941 time 0.2833 (0.4113) loss 4.1590 (3.8923) grad_norm 1.1749 (1.1033) [2021-04-15 18:26:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][40/1251] eta 0:07:41 lr 0.000941 time 0.2717 (0.3811) loss 4.7627 (3.9516) grad_norm 0.9075 (1.0900) [2021-04-15 18:26:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][50/1251] eta 0:07:15 lr 0.000941 time 0.2797 (0.3625) loss 4.3330 (3.9663) grad_norm 1.0246 (1.0956) [2021-04-15 18:26:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][60/1251] eta 0:07:00 lr 0.000941 time 0.4694 (0.3528) loss 4.4266 (3.9909) grad_norm 1.1471 (1.0834) [2021-04-15 18:26:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][70/1251] eta 0:06:44 lr 0.000941 time 0.2718 (0.3423) loss 3.7231 (3.9669) grad_norm 1.2905 (inf) [2021-04-15 18:26:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][80/1251] eta 0:06:32 lr 0.000941 time 0.2797 (0.3349) loss 4.4905 (3.9642) grad_norm 1.1525 (inf) [2021-04-15 18:26:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][90/1251] eta 0:06:24 lr 0.000941 time 0.2948 (0.3315) loss 3.8841 (3.9424) grad_norm 0.9934 (inf) [2021-04-15 18:26:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][100/1251] eta 0:06:16 lr 0.000941 time 0.2627 (0.3270) loss 3.9579 (3.9539) grad_norm 0.9464 (inf) [2021-04-15 18:26:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][110/1251] eta 0:06:08 lr 0.000941 time 0.2996 (0.3231) loss 4.0022 (3.9809) grad_norm 1.2447 (inf) [2021-04-15 18:26:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][120/1251] eta 0:06:01 lr 0.000941 time 0.3024 (0.3199) loss 2.9943 (3.9900) grad_norm 1.0734 (inf) [2021-04-15 18:27:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][130/1251] eta 0:05:55 lr 0.000941 time 0.2782 (0.3167) loss 2.9559 (3.9891) grad_norm 0.9530 (inf) [2021-04-15 18:27:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][140/1251] eta 0:05:50 lr 0.000941 time 0.2827 (0.3155) loss 4.5422 (4.0000) grad_norm 1.2566 (inf) [2021-04-15 18:27:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][150/1251] eta 0:05:45 lr 0.000941 time 0.2836 (0.3142) loss 4.4946 (3.9976) grad_norm 1.1596 (inf) [2021-04-15 18:27:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][160/1251] eta 0:05:40 lr 0.000941 time 0.2951 (0.3124) loss 3.7435 (3.9796) grad_norm 1.1172 (inf) [2021-04-15 18:27:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][170/1251] eta 0:05:35 lr 0.000941 time 0.2781 (0.3105) loss 4.7156 (3.9855) grad_norm 0.9379 (inf) [2021-04-15 18:27:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][180/1251] eta 0:05:30 lr 0.000941 time 0.2857 (0.3087) loss 4.4572 (3.9770) grad_norm 0.9677 (inf) [2021-04-15 18:27:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][190/1251] eta 0:05:25 lr 0.000941 time 0.2629 (0.3072) loss 4.0013 (3.9783) grad_norm 1.2033 (inf) [2021-04-15 18:27:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][200/1251] eta 0:05:21 lr 0.000941 time 0.2862 (0.3059) loss 3.3831 (3.9816) grad_norm 1.2145 (inf) [2021-04-15 18:27:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][210/1251] eta 0:05:17 lr 0.000941 time 0.2694 (0.3045) loss 4.0501 (3.9663) grad_norm 1.2634 (inf) [2021-04-15 18:27:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][220/1251] eta 0:05:12 lr 0.000941 time 0.2849 (0.3032) loss 4.1399 (3.9615) grad_norm 1.1216 (inf) [2021-04-15 18:27:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][230/1251] eta 0:05:08 lr 0.000941 time 0.2794 (0.3022) loss 4.5272 (3.9682) grad_norm 0.8991 (inf) [2021-04-15 18:27:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][240/1251] eta 0:05:04 lr 0.000941 time 0.2629 (0.3012) loss 3.8525 (3.9788) grad_norm 1.0387 (inf) [2021-04-15 18:27:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][250/1251] eta 0:05:00 lr 0.000941 time 0.3128 (0.3003) loss 3.8719 (3.9830) grad_norm 1.1828 (inf) [2021-04-15 18:27:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][260/1251] eta 0:04:56 lr 0.000941 time 0.2861 (0.2995) loss 3.6742 (3.9975) grad_norm 1.0482 (inf) [2021-04-15 18:27:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][270/1251] eta 0:04:53 lr 0.000941 time 0.2756 (0.2988) loss 4.7381 (4.0040) grad_norm 1.0550 (inf) [2021-04-15 18:27:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][280/1251] eta 0:04:49 lr 0.000941 time 0.2678 (0.2982) loss 3.7780 (4.0089) grad_norm 0.9197 (inf) [2021-04-15 18:27:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][290/1251] eta 0:04:45 lr 0.000941 time 0.2809 (0.2974) loss 4.2370 (4.0056) grad_norm 1.0784 (inf) [2021-04-15 18:27:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][300/1251] eta 0:04:42 lr 0.000941 time 0.2770 (0.2967) loss 4.4531 (4.0038) grad_norm 1.2604 (inf) [2021-04-15 18:27:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][310/1251] eta 0:04:38 lr 0.000941 time 0.2702 (0.2961) loss 4.2173 (4.0024) grad_norm 0.9979 (inf) [2021-04-15 18:27:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][320/1251] eta 0:04:35 lr 0.000941 time 0.2680 (0.2959) loss 4.0567 (3.9971) grad_norm 0.9130 (inf) [2021-04-15 18:27:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][330/1251] eta 0:04:32 lr 0.000941 time 0.2963 (0.2953) loss 4.1362 (3.9896) grad_norm 1.0722 (inf) [2021-04-15 18:28:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][340/1251] eta 0:04:28 lr 0.000941 time 0.2924 (0.2949) loss 4.7571 (3.9916) grad_norm 1.0788 (inf) [2021-04-15 18:28:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][350/1251] eta 0:04:25 lr 0.000941 time 0.2979 (0.2943) loss 5.2331 (3.9962) grad_norm 0.9785 (inf) [2021-04-15 18:28:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][360/1251] eta 0:04:21 lr 0.000941 time 0.3013 (0.2940) loss 3.8020 (3.9962) grad_norm 1.0656 (inf) [2021-04-15 18:28:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][370/1251] eta 0:04:18 lr 0.000941 time 0.2665 (0.2938) loss 3.5090 (3.9897) grad_norm 1.1417 (inf) [2021-04-15 18:28:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][380/1251] eta 0:04:15 lr 0.000940 time 0.2788 (0.2936) loss 3.8013 (3.9863) grad_norm 1.0775 (inf) [2021-04-15 18:28:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][390/1251] eta 0:04:12 lr 0.000940 time 0.2641 (0.2936) loss 4.2944 (3.9834) grad_norm 1.0202 (inf) [2021-04-15 18:28:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][400/1251] eta 0:04:09 lr 0.000940 time 0.2887 (0.2932) loss 3.2117 (3.9801) grad_norm 1.1181 (inf) [2021-04-15 18:28:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][410/1251] eta 0:04:06 lr 0.000940 time 0.2709 (0.2927) loss 3.8711 (3.9761) grad_norm 0.9886 (inf) [2021-04-15 18:28:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][420/1251] eta 0:04:03 lr 0.000940 time 0.2755 (0.2925) loss 4.2045 (3.9816) grad_norm 1.0839 (inf) [2021-04-15 18:28:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][430/1251] eta 0:03:59 lr 0.000940 time 0.2718 (0.2922) loss 4.7497 (3.9884) grad_norm 1.1643 (inf) [2021-04-15 18:28:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][440/1251] eta 0:03:56 lr 0.000940 time 0.2877 (0.2921) loss 4.6264 (3.9927) grad_norm 1.1820 (inf) [2021-04-15 18:28:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][450/1251] eta 0:03:53 lr 0.000940 time 0.2754 (0.2917) loss 4.5620 (3.9943) grad_norm 1.1468 (inf) [2021-04-15 18:28:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][460/1251] eta 0:03:50 lr 0.000940 time 0.2931 (0.2914) loss 4.3414 (3.9947) grad_norm 0.9824 (inf) [2021-04-15 18:28:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][470/1251] eta 0:03:47 lr 0.000940 time 0.2763 (0.2911) loss 3.7115 (3.9959) grad_norm 1.2668 (inf) [2021-04-15 18:28:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][480/1251] eta 0:03:44 lr 0.000940 time 0.2795 (0.2908) loss 4.3706 (3.9964) grad_norm 1.1378 (inf) [2021-04-15 18:28:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][490/1251] eta 0:03:41 lr 0.000940 time 0.2598 (0.2906) loss 4.2631 (3.9992) grad_norm 1.0590 (inf) [2021-04-15 18:28:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][500/1251] eta 0:03:38 lr 0.000940 time 0.2721 (0.2903) loss 3.5899 (3.9997) grad_norm 1.0824 (inf) [2021-04-15 18:28:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][510/1251] eta 0:03:34 lr 0.000940 time 0.2920 (0.2901) loss 3.6445 (3.9925) grad_norm 1.2181 (inf) [2021-04-15 18:28:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][520/1251] eta 0:03:31 lr 0.000940 time 0.3048 (0.2899) loss 4.4019 (3.9903) grad_norm 1.1375 (inf) [2021-04-15 18:28:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][530/1251] eta 0:03:28 lr 0.000940 time 0.2890 (0.2897) loss 5.0373 (3.9860) grad_norm 1.0399 (inf) [2021-04-15 18:28:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][540/1251] eta 0:03:25 lr 0.000940 time 0.2905 (0.2896) loss 4.1733 (3.9896) grad_norm 1.3071 (inf) [2021-04-15 18:29:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][550/1251] eta 0:03:22 lr 0.000940 time 0.3093 (0.2894) loss 3.0646 (3.9889) grad_norm 1.0896 (inf) [2021-04-15 18:29:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][560/1251] eta 0:03:19 lr 0.000940 time 0.2574 (0.2892) loss 4.2633 (3.9874) grad_norm 1.0937 (inf) [2021-04-15 18:29:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][570/1251] eta 0:03:16 lr 0.000940 time 0.3017 (0.2891) loss 4.4875 (3.9879) grad_norm 1.0188 (inf) [2021-04-15 18:29:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][580/1251] eta 0:03:14 lr 0.000940 time 0.2741 (0.2892) loss 2.7625 (3.9835) grad_norm 1.0395 (inf) [2021-04-15 18:29:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][590/1251] eta 0:03:10 lr 0.000940 time 0.2615 (0.2889) loss 3.2948 (3.9849) grad_norm 1.1784 (inf) [2021-04-15 18:29:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][600/1251] eta 0:03:07 lr 0.000940 time 0.2753 (0.2887) loss 3.0103 (3.9803) grad_norm 1.1722 (inf) [2021-04-15 18:29:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][610/1251] eta 0:03:04 lr 0.000940 time 0.2748 (0.2885) loss 4.4975 (3.9841) grad_norm 1.0897 (inf) [2021-04-15 18:29:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][620/1251] eta 0:03:02 lr 0.000940 time 0.2583 (0.2884) loss 3.9669 (3.9817) grad_norm 1.0476 (inf) [2021-04-15 18:29:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][630/1251] eta 0:02:58 lr 0.000940 time 0.2915 (0.2882) loss 4.7697 (3.9799) grad_norm 1.0850 (inf) [2021-04-15 18:29:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][640/1251] eta 0:02:56 lr 0.000940 time 0.2797 (0.2883) loss 4.0980 (3.9824) grad_norm 1.0160 (inf) [2021-04-15 18:29:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][650/1251] eta 0:02:53 lr 0.000940 time 0.2708 (0.2881) loss 2.8838 (3.9774) grad_norm 0.9674 (inf) [2021-04-15 18:29:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][660/1251] eta 0:02:50 lr 0.000940 time 0.2716 (0.2880) loss 3.7043 (3.9732) grad_norm 0.9762 (inf) [2021-04-15 18:29:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][670/1251] eta 0:02:47 lr 0.000940 time 0.2915 (0.2881) loss 4.3434 (3.9718) grad_norm 0.9652 (inf) [2021-04-15 18:29:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][680/1251] eta 0:02:44 lr 0.000940 time 0.2474 (0.2879) loss 4.1594 (3.9705) grad_norm 1.2023 (inf) [2021-04-15 18:29:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][690/1251] eta 0:02:41 lr 0.000940 time 0.2839 (0.2878) loss 4.2654 (3.9731) grad_norm 0.9434 (inf) [2021-04-15 18:29:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][700/1251] eta 0:02:38 lr 0.000940 time 0.2690 (0.2875) loss 3.9350 (3.9745) grad_norm 1.0725 (inf) [2021-04-15 18:29:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][710/1251] eta 0:02:35 lr 0.000940 time 0.4344 (0.2876) loss 2.9960 (3.9724) grad_norm 1.0256 (inf) [2021-04-15 18:29:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][720/1251] eta 0:02:32 lr 0.000940 time 0.2672 (0.2874) loss 4.4233 (3.9752) grad_norm 1.0068 (inf) [2021-04-15 18:29:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][730/1251] eta 0:02:29 lr 0.000940 time 0.2753 (0.2876) loss 4.3198 (3.9743) grad_norm 1.0639 (inf) [2021-04-15 18:29:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][740/1251] eta 0:02:26 lr 0.000940 time 0.2808 (0.2874) loss 3.9060 (3.9732) grad_norm 0.9661 (inf) [2021-04-15 18:29:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][750/1251] eta 0:02:23 lr 0.000940 time 0.2575 (0.2873) loss 4.6012 (3.9735) grad_norm 0.9725 (inf) [2021-04-15 18:29:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][760/1251] eta 0:02:20 lr 0.000940 time 0.2717 (0.2872) loss 3.6397 (3.9711) grad_norm 0.9965 (inf) [2021-04-15 18:30:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][770/1251] eta 0:02:18 lr 0.000940 time 0.2793 (0.2871) loss 3.9376 (3.9694) grad_norm 0.8385 (inf) [2021-04-15 18:30:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][780/1251] eta 0:02:15 lr 0.000940 time 0.2775 (0.2870) loss 4.0344 (3.9709) grad_norm 1.1803 (inf) [2021-04-15 18:30:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][790/1251] eta 0:02:12 lr 0.000940 time 0.2878 (0.2869) loss 3.7402 (3.9702) grad_norm 0.8968 (inf) [2021-04-15 18:30:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][800/1251] eta 0:02:09 lr 0.000940 time 0.2724 (0.2867) loss 3.9256 (3.9657) grad_norm 1.1534 (inf) [2021-04-15 18:30:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][810/1251] eta 0:02:06 lr 0.000940 time 0.2793 (0.2866) loss 3.7444 (3.9650) grad_norm 0.9531 (inf) [2021-04-15 18:30:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][820/1251] eta 0:02:03 lr 0.000940 time 0.2693 (0.2865) loss 3.9288 (3.9638) grad_norm 0.9719 (inf) [2021-04-15 18:30:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][830/1251] eta 0:02:00 lr 0.000940 time 0.2606 (0.2864) loss 3.9899 (3.9673) grad_norm 0.9984 (inf) [2021-04-15 18:30:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][840/1251] eta 0:01:57 lr 0.000940 time 0.2658 (0.2863) loss 3.1088 (3.9690) grad_norm 0.9738 (inf) [2021-04-15 18:30:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][850/1251] eta 0:01:54 lr 0.000940 time 0.3016 (0.2862) loss 4.3297 (3.9685) grad_norm 1.0035 (inf) [2021-04-15 18:30:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][860/1251] eta 0:01:51 lr 0.000940 time 0.2915 (0.2861) loss 3.8861 (3.9649) grad_norm 0.9992 (inf) [2021-04-15 18:30:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][870/1251] eta 0:01:48 lr 0.000940 time 0.2674 (0.2860) loss 4.7630 (3.9662) grad_norm 1.0018 (inf) [2021-04-15 18:30:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][880/1251] eta 0:01:46 lr 0.000940 time 0.2631 (0.2859) loss 4.6688 (3.9629) grad_norm 1.3126 (inf) [2021-04-15 18:30:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][890/1251] eta 0:01:43 lr 0.000939 time 0.2681 (0.2858) loss 3.7313 (3.9633) grad_norm 1.0406 (inf) [2021-04-15 18:30:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][900/1251] eta 0:01:40 lr 0.000939 time 0.2624 (0.2859) loss 4.1138 (3.9620) grad_norm 1.0434 (inf) [2021-04-15 18:30:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][910/1251] eta 0:01:37 lr 0.000939 time 0.2885 (0.2859) loss 3.1569 (3.9595) grad_norm 1.0886 (inf) [2021-04-15 18:30:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][920/1251] eta 0:01:34 lr 0.000939 time 0.2811 (0.2858) loss 3.8625 (3.9579) grad_norm 1.0470 (inf) [2021-04-15 18:30:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][930/1251] eta 0:01:31 lr 0.000939 time 0.2776 (0.2857) loss 4.6216 (3.9570) grad_norm 1.2792 (inf) [2021-04-15 18:30:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][940/1251] eta 0:01:28 lr 0.000939 time 0.3575 (0.2858) loss 4.2297 (3.9571) grad_norm 1.0923 (inf) [2021-04-15 18:30:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][950/1251] eta 0:01:26 lr 0.000939 time 0.2683 (0.2857) loss 4.5182 (3.9585) grad_norm 1.0434 (inf) [2021-04-15 18:30:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][960/1251] eta 0:01:23 lr 0.000939 time 0.2777 (0.2857) loss 4.5921 (3.9581) grad_norm 0.9860 (inf) [2021-04-15 18:30:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][970/1251] eta 0:01:20 lr 0.000939 time 0.2929 (0.2858) loss 4.3679 (3.9571) grad_norm 1.0175 (inf) [2021-04-15 18:31:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][980/1251] eta 0:01:17 lr 0.000939 time 0.2971 (0.2858) loss 4.4671 (3.9576) grad_norm 1.2796 (inf) [2021-04-15 18:31:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][990/1251] eta 0:01:14 lr 0.000939 time 0.2668 (0.2857) loss 3.9525 (3.9585) grad_norm 1.2320 (inf) [2021-04-15 18:31:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][1000/1251] eta 0:01:11 lr 0.000939 time 0.2932 (0.2857) loss 4.2926 (3.9595) grad_norm 1.2224 (inf) [2021-04-15 18:31:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][1010/1251] eta 0:01:08 lr 0.000939 time 0.2724 (0.2856) loss 4.1687 (3.9596) grad_norm 0.9258 (inf) [2021-04-15 18:31:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][1020/1251] eta 0:01:05 lr 0.000939 time 0.3101 (0.2855) loss 3.7828 (3.9612) grad_norm 0.9563 (inf) [2021-04-15 18:31:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][1030/1251] eta 0:01:03 lr 0.000939 time 0.2584 (0.2855) loss 4.5598 (3.9612) grad_norm 1.0232 (inf) [2021-04-15 18:31:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][1040/1251] eta 0:01:00 lr 0.000939 time 0.2784 (0.2855) loss 3.7355 (3.9634) grad_norm 1.0224 (inf) [2021-04-15 18:31:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][1050/1251] eta 0:00:57 lr 0.000939 time 0.2700 (0.2854) loss 2.9537 (3.9641) grad_norm 1.1356 (inf) [2021-04-15 18:31:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][1060/1251] eta 0:00:54 lr 0.000939 time 0.3069 (0.2854) loss 4.6192 (3.9664) grad_norm 1.0252 (inf) [2021-04-15 18:31:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][1070/1251] eta 0:00:51 lr 0.000939 time 0.2946 (0.2854) loss 3.4616 (3.9685) grad_norm 1.0635 (inf) [2021-04-15 18:31:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][1080/1251] eta 0:00:48 lr 0.000939 time 0.2882 (0.2854) loss 4.7050 (3.9705) grad_norm 1.0148 (inf) [2021-04-15 18:31:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][1090/1251] eta 0:00:45 lr 0.000939 time 0.2665 (0.2853) loss 3.2003 (3.9705) grad_norm 0.9586 (inf) [2021-04-15 18:31:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][1100/1251] eta 0:00:43 lr 0.000939 time 0.2845 (0.2853) loss 4.1345 (3.9726) grad_norm 1.1282 (inf) [2021-04-15 18:31:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][1110/1251] eta 0:00:40 lr 0.000939 time 0.2920 (0.2852) loss 2.7530 (3.9753) grad_norm 0.9645 (inf) [2021-04-15 18:31:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][1120/1251] eta 0:00:37 lr 0.000939 time 0.2647 (0.2852) loss 4.4976 (3.9771) grad_norm 1.3741 (inf) [2021-04-15 18:31:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][1130/1251] eta 0:00:34 lr 0.000939 time 0.2843 (0.2851) loss 4.5517 (3.9773) grad_norm 1.0213 (inf) [2021-04-15 18:31:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][1140/1251] eta 0:00:31 lr 0.000939 time 0.2801 (0.2850) loss 3.1448 (3.9771) grad_norm 1.0632 (inf) [2021-04-15 18:31:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][1150/1251] eta 0:00:28 lr 0.000939 time 0.2711 (0.2850) loss 3.9850 (3.9764) grad_norm 1.3915 (inf) [2021-04-15 18:31:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][1160/1251] eta 0:00:25 lr 0.000939 time 0.2976 (0.2850) loss 4.3996 (3.9758) grad_norm 0.9983 (inf) [2021-04-15 18:31:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][1170/1251] eta 0:00:23 lr 0.000939 time 0.2773 (0.2849) loss 4.4216 (3.9758) grad_norm 1.2734 (inf) [2021-04-15 18:31:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][1180/1251] eta 0:00:20 lr 0.000939 time 0.2843 (0.2849) loss 3.5118 (3.9762) grad_norm 1.0315 (inf) [2021-04-15 18:32:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][1190/1251] eta 0:00:17 lr 0.000939 time 0.2832 (0.2848) loss 3.6756 (3.9754) grad_norm 0.9891 (inf) [2021-04-15 18:32:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][1200/1251] eta 0:00:14 lr 0.000939 time 0.2837 (0.2848) loss 4.6465 (3.9741) grad_norm 0.9704 (inf) [2021-04-15 18:32:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][1210/1251] eta 0:00:11 lr 0.000939 time 0.2719 (0.2847) loss 4.0559 (3.9717) grad_norm 0.9673 (inf) [2021-04-15 18:32:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][1220/1251] eta 0:00:08 lr 0.000939 time 0.2843 (0.2848) loss 4.1732 (3.9717) grad_norm 1.0185 (inf) [2021-04-15 18:32:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][1230/1251] eta 0:00:05 lr 0.000939 time 0.2861 (0.2847) loss 4.3127 (3.9701) grad_norm 1.2051 (inf) [2021-04-15 18:32:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][1240/1251] eta 0:00:03 lr 0.000939 time 0.3612 (0.2847) loss 4.1412 (3.9709) grad_norm 0.9158 (inf) [2021-04-15 18:32:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [47/300][1250/1251] eta 0:00:00 lr 0.000939 time 0.2617 (0.2845) loss 4.6682 (3.9731) grad_norm 1.3227 (inf) [2021-04-15 18:32:19 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 47 training takes 0:05:58 [2021-04-15 18:32:19 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_47.pth saving...... [2021-04-15 18:32:32 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_47.pth saved !!! [2021-04-15 18:32:33 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.078 (1.078) Loss 1.2904 (1.2904) Acc@1 70.801 (70.801) Acc@5 89.844 (89.844) [2021-04-15 18:32:35 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.107 (0.215) Loss 1.3504 (1.3397) Acc@1 69.922 (69.993) Acc@5 89.258 (89.915) [2021-04-15 18:32:37 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.475 (0.251) Loss 1.3324 (1.3444) Acc@1 69.043 (69.550) Acc@5 89.648 (89.783) [2021-04-15 18:32:39 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.124 (0.226) Loss 1.3795 (1.3517) Acc@1 69.922 (69.509) Acc@5 88.867 (89.727) [2021-04-15 18:32:41 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.076 (0.215) Loss 1.3107 (1.3490) Acc@1 69.531 (69.641) Acc@5 89.844 (89.694) [2021-04-15 18:32:44 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 69.774 Acc@5 89.776 [2021-04-15 18:32:44 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 69.8% [2021-04-15 18:32:44 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 69.77% [2021-04-15 18:32:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][0/1251] eta 1:19:33 lr 0.000939 time 3.8156 (3.8156) loss 4.4773 (4.4773) grad_norm 1.0278 (1.0278) [2021-04-15 18:32:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][10/1251] eta 0:12:25 lr 0.000939 time 0.2862 (0.6008) loss 3.6285 (4.0719) grad_norm 1.1153 (1.0975) [2021-04-15 18:32:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][20/1251] eta 0:09:13 lr 0.000939 time 0.3231 (0.4494) loss 4.5556 (4.0715) grad_norm 1.1447 (1.1047) [2021-04-15 18:32:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][30/1251] eta 0:08:00 lr 0.000939 time 0.2603 (0.3932) loss 4.3436 (3.9865) grad_norm 0.9404 (1.1084) [2021-04-15 18:32:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][40/1251] eta 0:07:29 lr 0.000939 time 0.4874 (0.3711) loss 4.1223 (3.9539) grad_norm 0.9901 (1.0917) [2021-04-15 18:33:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][50/1251] eta 0:07:03 lr 0.000939 time 0.2583 (0.3530) loss 3.3824 (3.9482) grad_norm 1.0993 (1.1000) [2021-04-15 18:33:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][60/1251] eta 0:06:46 lr 0.000939 time 0.2809 (0.3411) loss 4.2474 (3.9035) grad_norm 1.0514 (1.1052) [2021-04-15 18:33:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][70/1251] eta 0:06:33 lr 0.000939 time 0.2851 (0.3330) loss 4.2313 (3.9208) grad_norm 1.4208 (1.1391) [2021-04-15 18:33:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][80/1251] eta 0:06:21 lr 0.000939 time 0.2739 (0.3260) loss 4.2147 (3.9410) grad_norm 1.0883 (1.1449) [2021-04-15 18:33:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][90/1251] eta 0:06:12 lr 0.000939 time 0.2886 (0.3210) loss 4.3039 (3.9476) grad_norm 1.0669 (1.1436) [2021-04-15 18:33:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][100/1251] eta 0:06:05 lr 0.000939 time 0.2625 (0.3173) loss 3.7570 (3.9635) grad_norm 0.9585 (1.1410) [2021-04-15 18:33:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][110/1251] eta 0:05:58 lr 0.000939 time 0.2796 (0.3140) loss 4.2506 (3.9650) grad_norm 1.3219 (1.1416) [2021-04-15 18:33:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][120/1251] eta 0:05:51 lr 0.000939 time 0.2728 (0.3108) loss 4.4944 (3.9728) grad_norm 1.0210 (1.1348) [2021-04-15 18:33:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][130/1251] eta 0:05:46 lr 0.000939 time 0.2897 (0.3089) loss 4.1940 (3.9781) grad_norm 1.2782 (1.1277) [2021-04-15 18:33:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][140/1251] eta 0:05:41 lr 0.000938 time 0.2583 (0.3072) loss 3.2614 (3.9557) grad_norm 1.1071 (1.1254) [2021-04-15 18:33:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][150/1251] eta 0:05:36 lr 0.000938 time 0.3571 (0.3059) loss 3.6857 (3.9503) grad_norm 1.1080 (1.1265) [2021-04-15 18:33:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][160/1251] eta 0:05:31 lr 0.000938 time 0.2851 (0.3042) loss 4.3581 (3.9445) grad_norm 1.0654 (1.1266) [2021-04-15 18:33:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][170/1251] eta 0:05:28 lr 0.000938 time 0.3021 (0.3037) loss 4.3766 (3.9425) grad_norm 1.1165 (1.1378) [2021-04-15 18:33:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][180/1251] eta 0:05:24 lr 0.000938 time 0.3016 (0.3032) loss 4.1637 (3.9342) grad_norm 1.1363 (1.1429) [2021-04-15 18:33:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][190/1251] eta 0:05:20 lr 0.000938 time 0.2750 (0.3017) loss 4.2882 (3.9278) grad_norm 1.1538 (1.1416) [2021-04-15 18:33:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][200/1251] eta 0:05:15 lr 0.000938 time 0.2918 (0.3005) loss 3.1685 (3.9237) grad_norm 1.1596 (1.1417) [2021-04-15 18:33:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][210/1251] eta 0:05:11 lr 0.000938 time 0.2769 (0.2997) loss 4.5366 (3.9210) grad_norm 1.1870 (1.1435) [2021-04-15 18:33:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][220/1251] eta 0:05:07 lr 0.000938 time 0.2951 (0.2987) loss 4.2985 (3.9058) grad_norm 1.3850 (1.1442) [2021-04-15 18:33:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][230/1251] eta 0:05:04 lr 0.000938 time 0.2753 (0.2980) loss 4.5895 (3.9170) grad_norm 1.2586 (1.1444) [2021-04-15 18:33:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][240/1251] eta 0:05:00 lr 0.000938 time 0.2974 (0.2974) loss 3.4763 (3.9104) grad_norm 1.1969 (1.1437) [2021-04-15 18:33:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][250/1251] eta 0:04:57 lr 0.000938 time 0.2760 (0.2967) loss 4.3667 (3.9103) grad_norm 1.0587 (1.1414) [2021-04-15 18:34:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][260/1251] eta 0:04:53 lr 0.000938 time 0.2812 (0.2961) loss 2.6334 (3.9067) grad_norm 0.9701 (1.1375) [2021-04-15 18:34:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][270/1251] eta 0:04:49 lr 0.000938 time 0.2735 (0.2954) loss 3.8345 (3.9078) grad_norm 1.1705 (1.1376) [2021-04-15 18:34:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][280/1251] eta 0:04:46 lr 0.000938 time 0.3031 (0.2950) loss 4.0273 (3.9211) grad_norm 0.9603 (1.1363) [2021-04-15 18:34:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][290/1251] eta 0:04:42 lr 0.000938 time 0.2708 (0.2944) loss 3.6487 (3.9268) grad_norm 0.9805 (1.1370) [2021-04-15 18:34:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][300/1251] eta 0:04:39 lr 0.000938 time 0.2713 (0.2940) loss 4.3087 (3.9347) grad_norm 1.2761 (1.1365) [2021-04-15 18:34:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][310/1251] eta 0:04:36 lr 0.000938 time 0.2778 (0.2937) loss 4.0547 (3.9353) grad_norm 1.0628 (1.1347) [2021-04-15 18:34:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][320/1251] eta 0:04:33 lr 0.000938 time 0.3133 (0.2934) loss 3.2482 (3.9326) grad_norm 1.0760 (1.1310) [2021-04-15 18:34:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][330/1251] eta 0:04:29 lr 0.000938 time 0.2749 (0.2928) loss 2.8029 (3.9271) grad_norm 1.0443 (1.1298) [2021-04-15 18:34:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][340/1251] eta 0:04:26 lr 0.000938 time 0.2821 (0.2923) loss 3.9141 (3.9352) grad_norm 1.0852 (1.1275) [2021-04-15 18:34:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][350/1251] eta 0:04:23 lr 0.000938 time 0.2800 (0.2919) loss 3.1643 (3.9306) grad_norm 1.0657 (1.1244) [2021-04-15 18:34:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][360/1251] eta 0:04:20 lr 0.000938 time 0.2503 (0.2921) loss 2.8410 (3.9210) grad_norm 0.9773 (1.1217) [2021-04-15 18:34:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][370/1251] eta 0:04:17 lr 0.000938 time 0.2745 (0.2919) loss 3.0401 (3.9220) grad_norm 1.3217 (1.1204) [2021-04-15 18:34:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][380/1251] eta 0:04:13 lr 0.000938 time 0.2607 (0.2915) loss 4.4017 (3.9206) grad_norm 1.0117 (1.1212) [2021-04-15 18:34:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][390/1251] eta 0:04:10 lr 0.000938 time 0.2774 (0.2912) loss 4.6091 (3.9230) grad_norm 1.2019 (1.1222) [2021-04-15 18:34:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][400/1251] eta 0:04:07 lr 0.000938 time 0.2660 (0.2909) loss 3.6717 (3.9185) grad_norm 1.1009 (1.1203) [2021-04-15 18:34:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][410/1251] eta 0:04:04 lr 0.000938 time 0.2591 (0.2906) loss 4.1859 (3.9210) grad_norm 0.9441 (1.1202) [2021-04-15 18:34:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][420/1251] eta 0:04:01 lr 0.000938 time 0.2700 (0.2907) loss 3.5583 (3.9154) grad_norm 1.4116 (1.1202) [2021-04-15 18:34:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][430/1251] eta 0:03:58 lr 0.000938 time 0.2877 (0.2905) loss 3.4993 (3.9179) grad_norm 1.0463 (1.1197) [2021-04-15 18:34:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][440/1251] eta 0:03:55 lr 0.000938 time 0.2512 (0.2902) loss 3.2764 (3.9136) grad_norm 1.1330 (1.1189) [2021-04-15 18:34:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][450/1251] eta 0:03:52 lr 0.000938 time 0.4202 (0.2903) loss 4.4722 (3.9175) grad_norm 1.1256 (1.1190) [2021-04-15 18:34:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][460/1251] eta 0:03:49 lr 0.000938 time 0.2810 (0.2900) loss 4.1007 (3.9190) grad_norm 1.2826 (1.1204) [2021-04-15 18:35:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][470/1251] eta 0:03:46 lr 0.000938 time 0.2803 (0.2899) loss 2.8808 (3.9109) grad_norm 0.9312 (1.1212) [2021-04-15 18:35:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][480/1251] eta 0:03:43 lr 0.000938 time 0.2594 (0.2897) loss 2.9475 (3.9128) grad_norm 0.8625 (1.1200) [2021-04-15 18:35:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][490/1251] eta 0:03:40 lr 0.000938 time 0.2630 (0.2895) loss 3.0308 (3.9080) grad_norm 1.0247 (1.1196) [2021-04-15 18:35:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][500/1251] eta 0:03:37 lr 0.000938 time 0.2643 (0.2894) loss 4.3539 (3.9109) grad_norm 1.0482 (1.1181) [2021-04-15 18:35:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][510/1251] eta 0:03:34 lr 0.000938 time 0.2841 (0.2892) loss 3.9018 (3.9133) grad_norm 1.0028 (1.1168) [2021-04-15 18:35:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][520/1251] eta 0:03:31 lr 0.000938 time 0.2909 (0.2891) loss 3.0889 (3.9125) grad_norm 1.1216 (1.1152) [2021-04-15 18:35:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][530/1251] eta 0:03:28 lr 0.000938 time 0.2769 (0.2889) loss 3.6794 (3.9149) grad_norm 1.0731 (1.1157) [2021-04-15 18:35:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][540/1251] eta 0:03:25 lr 0.000938 time 0.2847 (0.2887) loss 4.2491 (3.9102) grad_norm 0.9352 (1.1140) [2021-04-15 18:35:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][550/1251] eta 0:03:22 lr 0.000938 time 0.2469 (0.2885) loss 4.9079 (3.9091) grad_norm 1.0136 (1.1128) [2021-04-15 18:35:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][560/1251] eta 0:03:19 lr 0.000938 time 0.2878 (0.2884) loss 4.0396 (3.9019) grad_norm 1.1958 (1.1124) [2021-04-15 18:35:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][570/1251] eta 0:03:16 lr 0.000938 time 0.3010 (0.2886) loss 3.5092 (3.9021) grad_norm 0.9989 (1.1119) [2021-04-15 18:35:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][580/1251] eta 0:03:13 lr 0.000938 time 0.3078 (0.2884) loss 3.0485 (3.9025) grad_norm 1.0436 (1.1115) [2021-04-15 18:35:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][590/1251] eta 0:03:10 lr 0.000938 time 0.2821 (0.2883) loss 4.2101 (3.9057) grad_norm 1.1773 (1.1110) [2021-04-15 18:35:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][600/1251] eta 0:03:07 lr 0.000938 time 0.2812 (0.2881) loss 3.4428 (3.9071) grad_norm 0.9093 (1.1099) [2021-04-15 18:35:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][610/1251] eta 0:03:04 lr 0.000938 time 0.3015 (0.2881) loss 4.2416 (3.9076) grad_norm 1.0999 (1.1097) [2021-04-15 18:35:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][620/1251] eta 0:03:01 lr 0.000938 time 0.2712 (0.2883) loss 3.9248 (3.9095) grad_norm 1.2435 (1.1095) [2021-04-15 18:35:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][630/1251] eta 0:02:59 lr 0.000938 time 0.2766 (0.2883) loss 3.2401 (3.9184) grad_norm 0.8955 (1.1103) [2021-04-15 18:35:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][640/1251] eta 0:02:56 lr 0.000937 time 0.2779 (0.2881) loss 4.1460 (3.9135) grad_norm 0.9342 (1.1105) [2021-04-15 18:35:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][650/1251] eta 0:02:53 lr 0.000937 time 0.3047 (0.2880) loss 3.7433 (3.9115) grad_norm 1.0072 (1.1103) [2021-04-15 18:35:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][660/1251] eta 0:02:50 lr 0.000937 time 0.2874 (0.2878) loss 3.3754 (3.9126) grad_norm 1.0405 (1.1104) [2021-04-15 18:35:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][670/1251] eta 0:02:47 lr 0.000937 time 0.2784 (0.2877) loss 4.8677 (3.9170) grad_norm 1.1141 (1.1094) [2021-04-15 18:36:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][680/1251] eta 0:02:44 lr 0.000937 time 0.2796 (0.2875) loss 3.2494 (3.9216) grad_norm 1.1617 (1.1096) [2021-04-15 18:36:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][690/1251] eta 0:02:41 lr 0.000937 time 0.2747 (0.2874) loss 4.2379 (3.9195) grad_norm 1.2293 (1.1102) [2021-04-15 18:36:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][700/1251] eta 0:02:38 lr 0.000937 time 0.2743 (0.2873) loss 4.5145 (3.9228) grad_norm 1.3483 (1.1106) [2021-04-15 18:36:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][710/1251] eta 0:02:35 lr 0.000937 time 0.2680 (0.2872) loss 3.4921 (3.9257) grad_norm 0.9937 (1.1103) [2021-04-15 18:36:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][720/1251] eta 0:02:32 lr 0.000937 time 0.2715 (0.2871) loss 3.7495 (3.9259) grad_norm 0.8460 (1.1091) [2021-04-15 18:36:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][730/1251] eta 0:02:29 lr 0.000937 time 0.3098 (0.2872) loss 4.4461 (3.9268) grad_norm 1.1324 (1.1082) [2021-04-15 18:36:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][740/1251] eta 0:02:26 lr 0.000937 time 0.4093 (0.2873) loss 3.2511 (3.9264) grad_norm 1.1949 (1.1086) [2021-04-15 18:36:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][750/1251] eta 0:02:23 lr 0.000937 time 0.2601 (0.2872) loss 4.5994 (3.9298) grad_norm 1.2571 (1.1100) [2021-04-15 18:36:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][760/1251] eta 0:02:20 lr 0.000937 time 0.2665 (0.2872) loss 4.1649 (3.9316) grad_norm 1.2512 (1.1102) [2021-04-15 18:36:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][770/1251] eta 0:02:18 lr 0.000937 time 0.2684 (0.2871) loss 4.7794 (3.9326) grad_norm 0.9950 (1.1091) [2021-04-15 18:36:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][780/1251] eta 0:02:15 lr 0.000937 time 0.2777 (0.2870) loss 3.3755 (3.9331) grad_norm 0.9510 (1.1083) [2021-04-15 18:36:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][790/1251] eta 0:02:12 lr 0.000937 time 0.2741 (0.2869) loss 2.8482 (3.9360) grad_norm 0.9373 (1.1077) [2021-04-15 18:36:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][800/1251] eta 0:02:09 lr 0.000937 time 0.2750 (0.2868) loss 2.9319 (3.9340) grad_norm 1.0680 (1.1071) [2021-04-15 18:36:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][810/1251] eta 0:02:06 lr 0.000937 time 0.2669 (0.2866) loss 4.7598 (3.9329) grad_norm 1.1264 (1.1060) [2021-04-15 18:36:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][820/1251] eta 0:02:03 lr 0.000937 time 0.2635 (0.2865) loss 4.2652 (3.9324) grad_norm 0.9054 (1.1057) [2021-04-15 18:36:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][830/1251] eta 0:02:00 lr 0.000937 time 0.2667 (0.2864) loss 3.7388 (3.9319) grad_norm 1.1741 (1.1055) [2021-04-15 18:36:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][840/1251] eta 0:01:57 lr 0.000937 time 0.2636 (0.2864) loss 3.4258 (3.9347) grad_norm 1.2228 (1.1066) [2021-04-15 18:36:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][850/1251] eta 0:01:54 lr 0.000937 time 0.2770 (0.2863) loss 3.7883 (3.9337) grad_norm 1.3866 (1.1079) [2021-04-15 18:36:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][860/1251] eta 0:01:51 lr 0.000937 time 0.2680 (0.2862) loss 4.4134 (3.9347) grad_norm 1.3770 (1.1089) [2021-04-15 18:36:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][870/1251] eta 0:01:49 lr 0.000937 time 0.3030 (0.2861) loss 3.6278 (3.9349) grad_norm 1.4269 (1.1090) [2021-04-15 18:36:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][880/1251] eta 0:01:46 lr 0.000937 time 0.2742 (0.2861) loss 3.6479 (3.9332) grad_norm 0.9073 (1.1091) [2021-04-15 18:36:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][890/1251] eta 0:01:43 lr 0.000937 time 0.2764 (0.2860) loss 4.7380 (3.9339) grad_norm 0.9678 (1.1084) [2021-04-15 18:37:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][900/1251] eta 0:01:40 lr 0.000937 time 0.2603 (0.2859) loss 2.9225 (3.9358) grad_norm 0.9639 (1.1079) [2021-04-15 18:37:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][910/1251] eta 0:01:37 lr 0.000937 time 0.2824 (0.2858) loss 4.1855 (3.9367) grad_norm 1.2793 (1.1083) [2021-04-15 18:37:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][920/1251] eta 0:01:34 lr 0.000937 time 0.2892 (0.2858) loss 3.8198 (3.9404) grad_norm 1.0502 (1.1079) [2021-04-15 18:37:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][930/1251] eta 0:01:31 lr 0.000937 time 0.2607 (0.2858) loss 4.3524 (3.9393) grad_norm 1.0087 (1.1073) [2021-04-15 18:37:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][940/1251] eta 0:01:28 lr 0.000937 time 0.2564 (0.2857) loss 4.4803 (3.9402) grad_norm 1.1579 (1.1074) [2021-04-15 18:37:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][950/1251] eta 0:01:25 lr 0.000937 time 0.2762 (0.2857) loss 4.6437 (3.9425) grad_norm 1.2789 (1.1072) [2021-04-15 18:37:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][960/1251] eta 0:01:23 lr 0.000937 time 0.2844 (0.2856) loss 3.3424 (3.9416) grad_norm 1.2351 (1.1074) [2021-04-15 18:37:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][970/1251] eta 0:01:20 lr 0.000937 time 0.2863 (0.2857) loss 4.4693 (3.9428) grad_norm 1.2970 (1.1073) [2021-04-15 18:37:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][980/1251] eta 0:01:17 lr 0.000937 time 0.2786 (0.2858) loss 3.4478 (3.9435) grad_norm 0.9906 (1.1073) [2021-04-15 18:37:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][990/1251] eta 0:01:14 lr 0.000937 time 0.2714 (0.2859) loss 4.2076 (3.9403) grad_norm 0.9167 (1.1070) [2021-04-15 18:37:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][1000/1251] eta 0:01:11 lr 0.000937 time 0.2975 (0.2857) loss 3.9294 (3.9401) grad_norm 1.1167 (1.1072) [2021-04-15 18:37:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][1010/1251] eta 0:01:08 lr 0.000937 time 0.2833 (0.2856) loss 3.1402 (3.9380) grad_norm 0.9911 (1.1073) [2021-04-15 18:37:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][1020/1251] eta 0:01:05 lr 0.000937 time 0.2791 (0.2856) loss 4.0627 (3.9394) grad_norm 1.1768 (1.1064) [2021-04-15 18:37:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][1030/1251] eta 0:01:03 lr 0.000937 time 0.2900 (0.2856) loss 3.7531 (3.9409) grad_norm 0.9695 (1.1052) [2021-04-15 18:37:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][1040/1251] eta 0:01:00 lr 0.000937 time 0.3059 (0.2856) loss 4.5561 (3.9442) grad_norm 1.2054 (1.1049) [2021-04-15 18:37:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][1050/1251] eta 0:00:57 lr 0.000937 time 0.2649 (0.2855) loss 3.9874 (3.9452) grad_norm 1.0280 (1.1041) [2021-04-15 18:37:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][1060/1251] eta 0:00:54 lr 0.000937 time 0.2700 (0.2855) loss 3.8998 (3.9467) grad_norm 1.6177 (1.1041) [2021-04-15 18:37:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][1070/1251] eta 0:00:51 lr 0.000937 time 0.2660 (0.2854) loss 3.0732 (3.9448) grad_norm 1.1554 (1.1038) [2021-04-15 18:37:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][1080/1251] eta 0:00:48 lr 0.000937 time 0.2804 (0.2853) loss 4.4851 (3.9463) grad_norm 1.0321 (1.1031) [2021-04-15 18:37:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][1090/1251] eta 0:00:45 lr 0.000937 time 0.2731 (0.2853) loss 3.3762 (3.9458) grad_norm 0.9467 (1.1032) [2021-04-15 18:37:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][1100/1251] eta 0:00:43 lr 0.000937 time 0.2951 (0.2852) loss 4.2455 (3.9438) grad_norm 1.0045 (1.1026) [2021-04-15 18:38:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][1110/1251] eta 0:00:40 lr 0.000937 time 0.2680 (0.2852) loss 4.7895 (3.9450) grad_norm 0.8659 (1.1028) [2021-04-15 18:38:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][1120/1251] eta 0:00:37 lr 0.000937 time 0.2737 (0.2852) loss 4.4713 (3.9462) grad_norm 1.1510 (1.1030) [2021-04-15 18:38:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][1130/1251] eta 0:00:34 lr 0.000936 time 0.2860 (0.2851) loss 3.1111 (3.9463) grad_norm 1.1405 (1.1027) [2021-04-15 18:38:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][1140/1251] eta 0:00:31 lr 0.000936 time 0.2711 (0.2850) loss 3.4306 (3.9453) grad_norm 0.8605 (1.1027) [2021-04-15 18:38:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][1150/1251] eta 0:00:28 lr 0.000936 time 0.3087 (0.2851) loss 2.9464 (3.9459) grad_norm 0.9500 (1.1017) [2021-04-15 18:38:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][1160/1251] eta 0:00:25 lr 0.000936 time 0.2662 (0.2850) loss 4.5551 (3.9468) grad_norm 1.2102 (1.1015) [2021-04-15 18:38:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][1170/1251] eta 0:00:23 lr 0.000936 time 0.2583 (0.2850) loss 3.8264 (3.9480) grad_norm 1.1332 (1.1016) [2021-04-15 18:38:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][1180/1251] eta 0:00:20 lr 0.000936 time 0.2424 (0.2849) loss 3.9149 (3.9475) grad_norm 1.1102 (1.1018) [2021-04-15 18:38:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][1190/1251] eta 0:00:17 lr 0.000936 time 0.2667 (0.2849) loss 4.6797 (3.9471) grad_norm 1.1457 (1.1014) [2021-04-15 18:38:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][1200/1251] eta 0:00:14 lr 0.000936 time 0.3058 (0.2849) loss 4.4928 (3.9487) grad_norm 1.0896 (1.1011) [2021-04-15 18:38:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][1210/1251] eta 0:00:11 lr 0.000936 time 0.2788 (0.2849) loss 2.9493 (3.9472) grad_norm 1.5190 (1.1017) [2021-04-15 18:38:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][1220/1251] eta 0:00:08 lr 0.000936 time 0.2938 (0.2849) loss 4.9028 (3.9512) grad_norm 1.0351 (1.1016) [2021-04-15 18:38:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][1230/1251] eta 0:00:05 lr 0.000936 time 0.2787 (0.2848) loss 3.7297 (3.9514) grad_norm 1.4190 (1.1017) [2021-04-15 18:38:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][1240/1251] eta 0:00:03 lr 0.000936 time 0.2488 (0.2847) loss 3.7759 (3.9517) grad_norm 1.0467 (1.1022) [2021-04-15 18:38:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [48/300][1250/1251] eta 0:00:00 lr 0.000936 time 0.2489 (0.2844) loss 4.3954 (3.9514) grad_norm 1.1450 (1.1024) [2021-04-15 18:38:42 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 48 training takes 0:05:58 [2021-04-15 18:38:42 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_48.pth saving...... [2021-04-15 18:38:58 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_48.pth saved !!! [2021-04-15 18:38:59 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.186 (1.186) Loss 1.2849 (1.2849) Acc@1 70.215 (70.215) Acc@5 89.844 (89.844) [2021-04-15 18:39:01 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.112 (0.234) Loss 1.3115 (1.2792) Acc@1 69.336 (70.126) Acc@5 89.844 (90.243) [2021-04-15 18:39:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.174 (0.213) Loss 1.3149 (1.2765) Acc@1 69.824 (70.308) Acc@5 89.160 (90.383) [2021-04-15 18:39:05 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.133 (0.239) Loss 1.3065 (1.2785) Acc@1 68.848 (70.130) Acc@5 89.746 (90.313) [2021-04-15 18:39:07 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.224) Loss 1.2764 (1.2836) Acc@1 71.582 (70.070) Acc@5 89.941 (90.220) [2021-04-15 18:39:13 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 70.098 Acc@5 90.164 [2021-04-15 18:39:13 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 70.1% [2021-04-15 18:39:13 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 70.10% [2021-04-15 18:39:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][0/1251] eta 0:43:03 lr 0.000936 time 2.0648 (2.0648) loss 4.2219 (4.2219) grad_norm 0.9127 (0.9127) [2021-04-15 18:39:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][10/1251] eta 0:09:07 lr 0.000936 time 0.2932 (0.4415) loss 4.8956 (4.0876) grad_norm 1.0918 (1.0351) [2021-04-15 18:39:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][20/1251] eta 0:07:30 lr 0.000936 time 0.2469 (0.3661) loss 3.7686 (3.9342) grad_norm 1.1956 (1.0737) [2021-04-15 18:39:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][30/1251] eta 0:06:53 lr 0.000936 time 0.2587 (0.3383) loss 3.8638 (3.7329) grad_norm 1.3755 (1.1352) [2021-04-15 18:39:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][40/1251] eta 0:06:33 lr 0.000936 time 0.2638 (0.3246) loss 4.1463 (3.8415) grad_norm 1.0564 (1.1542) [2021-04-15 18:39:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][50/1251] eta 0:06:18 lr 0.000936 time 0.2556 (0.3154) loss 4.5738 (3.8794) grad_norm 1.1600 (1.1362) [2021-04-15 18:39:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][60/1251] eta 0:06:09 lr 0.000936 time 0.3097 (0.3099) loss 4.3932 (3.8829) grad_norm 0.9448 (1.1302) [2021-04-15 18:39:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][70/1251] eta 0:06:01 lr 0.000936 time 0.2759 (0.3060) loss 3.6518 (3.8274) grad_norm 1.2338 (1.1330) [2021-04-15 18:39:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][80/1251] eta 0:05:54 lr 0.000936 time 0.2824 (0.3030) loss 3.8389 (3.8190) grad_norm 1.1106 (1.1247) [2021-04-15 18:39:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][90/1251] eta 0:05:49 lr 0.000936 time 0.2666 (0.3007) loss 2.6125 (3.8202) grad_norm 1.1327 (1.1221) [2021-04-15 18:39:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][100/1251] eta 0:05:43 lr 0.000936 time 0.2890 (0.2987) loss 4.5202 (3.8619) grad_norm 1.0365 (1.1157) [2021-04-15 18:39:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][110/1251] eta 0:05:39 lr 0.000936 time 0.3123 (0.2971) loss 5.3453 (3.8932) grad_norm 0.9164 (1.1137) [2021-04-15 18:39:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][120/1251] eta 0:05:33 lr 0.000936 time 0.2593 (0.2952) loss 2.3681 (3.9002) grad_norm 0.9620 (1.1119) [2021-04-15 18:39:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][130/1251] eta 0:05:30 lr 0.000936 time 0.3028 (0.2945) loss 2.4329 (3.8767) grad_norm 1.0657 (1.1077) [2021-04-15 18:39:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][140/1251] eta 0:05:27 lr 0.000936 time 0.2805 (0.2946) loss 4.3349 (3.8803) grad_norm 0.9909 (1.1097) [2021-04-15 18:39:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][150/1251] eta 0:05:23 lr 0.000936 time 0.2547 (0.2934) loss 4.4335 (3.8751) grad_norm 1.0859 (1.1061) [2021-04-15 18:40:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][160/1251] eta 0:05:19 lr 0.000936 time 0.2656 (0.2927) loss 2.7170 (3.8629) grad_norm 1.0620 (1.1064) [2021-04-15 18:40:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][170/1251] eta 0:05:15 lr 0.000936 time 0.2694 (0.2921) loss 3.4561 (3.8576) grad_norm 0.9907 (1.1045) [2021-04-15 18:40:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][180/1251] eta 0:05:12 lr 0.000936 time 0.2667 (0.2914) loss 4.5919 (3.8682) grad_norm 0.8635 (1.0996) [2021-04-15 18:40:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][190/1251] eta 0:05:08 lr 0.000936 time 0.2652 (0.2906) loss 3.6029 (3.8584) grad_norm 1.0713 (1.0974) [2021-04-15 18:40:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][200/1251] eta 0:05:05 lr 0.000936 time 0.2905 (0.2907) loss 4.2544 (3.8763) grad_norm 1.0744 (1.1006) [2021-04-15 18:40:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][210/1251] eta 0:05:01 lr 0.000936 time 0.2558 (0.2900) loss 4.0746 (3.8836) grad_norm 1.2726 (1.1001) [2021-04-15 18:40:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][220/1251] eta 0:04:58 lr 0.000936 time 0.2835 (0.2899) loss 2.8955 (3.8829) grad_norm 1.1029 (1.1007) [2021-04-15 18:40:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][230/1251] eta 0:04:55 lr 0.000936 time 0.2717 (0.2895) loss 4.3620 (3.8877) grad_norm 1.1171 (1.1004) [2021-04-15 18:40:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][240/1251] eta 0:04:52 lr 0.000936 time 0.2738 (0.2893) loss 4.1534 (3.8861) grad_norm 1.3520 (1.1019) [2021-04-15 18:40:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][250/1251] eta 0:04:49 lr 0.000936 time 0.2630 (0.2893) loss 4.6596 (3.8849) grad_norm 1.0477 (1.1029) [2021-04-15 18:40:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][260/1251] eta 0:04:46 lr 0.000936 time 0.2797 (0.2892) loss 3.7088 (3.8876) grad_norm 1.0544 (1.1047) [2021-04-15 18:40:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][270/1251] eta 0:04:43 lr 0.000936 time 0.2724 (0.2889) loss 4.2964 (3.8998) grad_norm 0.9658 (1.1074) [2021-04-15 18:40:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][280/1251] eta 0:04:40 lr 0.000936 time 0.2960 (0.2887) loss 3.3922 (3.8936) grad_norm 0.9839 (1.1058) [2021-04-15 18:40:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][290/1251] eta 0:04:37 lr 0.000936 time 0.2794 (0.2884) loss 4.7030 (3.9084) grad_norm 1.2393 (1.1043) [2021-04-15 18:40:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][300/1251] eta 0:04:34 lr 0.000936 time 0.2916 (0.2881) loss 3.9348 (3.9055) grad_norm 1.1155 (1.1058) [2021-04-15 18:40:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][310/1251] eta 0:04:30 lr 0.000936 time 0.2778 (0.2876) loss 4.2478 (3.9098) grad_norm 1.2112 (1.1053) [2021-04-15 18:40:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][320/1251] eta 0:04:27 lr 0.000936 time 0.2929 (0.2875) loss 4.1142 (3.9096) grad_norm 1.0649 (1.1049) [2021-04-15 18:40:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][330/1251] eta 0:04:24 lr 0.000936 time 0.2922 (0.2873) loss 3.6255 (3.9138) grad_norm 0.9907 (1.1026) [2021-04-15 18:40:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][340/1251] eta 0:04:21 lr 0.000936 time 0.2833 (0.2873) loss 3.9373 (3.9180) grad_norm 1.1551 (1.1011) [2021-04-15 18:40:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][350/1251] eta 0:04:18 lr 0.000936 time 0.2822 (0.2872) loss 4.4843 (3.9160) grad_norm 1.0543 (1.0988) [2021-04-15 18:40:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][360/1251] eta 0:04:16 lr 0.000936 time 0.2745 (0.2879) loss 2.9942 (3.9095) grad_norm 1.0843 (1.0989) [2021-04-15 18:40:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][370/1251] eta 0:04:13 lr 0.000935 time 0.2972 (0.2876) loss 4.5691 (3.9154) grad_norm 0.9300 (1.0989) [2021-04-15 18:41:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][380/1251] eta 0:04:10 lr 0.000935 time 0.2918 (0.2874) loss 4.8488 (3.9192) grad_norm 0.9005 (1.1006) [2021-04-15 18:41:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][390/1251] eta 0:04:07 lr 0.000935 time 0.2748 (0.2871) loss 4.5426 (3.9122) grad_norm 1.2406 (1.1028) [2021-04-15 18:41:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][400/1251] eta 0:04:04 lr 0.000935 time 0.2876 (0.2870) loss 3.4627 (3.9099) grad_norm 1.0852 (1.1019) [2021-04-15 18:41:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][410/1251] eta 0:04:01 lr 0.000935 time 0.2541 (0.2868) loss 3.1492 (3.9038) grad_norm 1.1804 (1.1012) [2021-04-15 18:41:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][420/1251] eta 0:03:58 lr 0.000935 time 0.2786 (0.2866) loss 4.3428 (3.9058) grad_norm 1.0457 (1.1029) [2021-04-15 18:41:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][430/1251] eta 0:03:55 lr 0.000935 time 0.3023 (0.2864) loss 3.8727 (3.9037) grad_norm 1.0371 (1.1013) [2021-04-15 18:41:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][440/1251] eta 0:03:52 lr 0.000935 time 0.2770 (0.2862) loss 2.9837 (3.9046) grad_norm 1.1820 (1.1006) [2021-04-15 18:41:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][450/1251] eta 0:03:49 lr 0.000935 time 0.2879 (0.2861) loss 3.7091 (3.9098) grad_norm 1.0204 (1.0993) [2021-04-15 18:41:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][460/1251] eta 0:03:46 lr 0.000935 time 0.2869 (0.2860) loss 2.6396 (3.9035) grad_norm 1.2898 (1.0997) [2021-04-15 18:41:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][470/1251] eta 0:03:43 lr 0.000935 time 0.2862 (0.2858) loss 3.5819 (3.9061) grad_norm 1.1698 (1.0983) [2021-04-15 18:41:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][480/1251] eta 0:03:40 lr 0.000935 time 0.2693 (0.2857) loss 4.1806 (3.9099) grad_norm 0.8796 (1.0979) [2021-04-15 18:41:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][490/1251] eta 0:03:37 lr 0.000935 time 0.3005 (0.2856) loss 4.4383 (3.9086) grad_norm 1.0238 (1.0960) [2021-04-15 18:41:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][500/1251] eta 0:03:34 lr 0.000935 time 0.3037 (0.2856) loss 4.2754 (3.9120) grad_norm 1.0345 (1.0970) [2021-04-15 18:41:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][510/1251] eta 0:03:31 lr 0.000935 time 0.2780 (0.2854) loss 4.2465 (3.9152) grad_norm 0.9370 (1.0951) [2021-04-15 18:41:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][520/1251] eta 0:03:28 lr 0.000935 time 0.2807 (0.2852) loss 4.7428 (3.9167) grad_norm 0.9246 (1.0938) [2021-04-15 18:41:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][530/1251] eta 0:03:25 lr 0.000935 time 0.2908 (0.2851) loss 4.6281 (3.9179) grad_norm 0.9946 (1.0936) [2021-04-15 18:41:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][540/1251] eta 0:03:22 lr 0.000935 time 0.2761 (0.2851) loss 3.1824 (3.9176) grad_norm 1.1736 (1.0927) [2021-04-15 18:41:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][550/1251] eta 0:03:19 lr 0.000935 time 0.3049 (0.2853) loss 4.4651 (3.9233) grad_norm 1.2989 (1.0931) [2021-04-15 18:41:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][560/1251] eta 0:03:16 lr 0.000935 time 0.2729 (0.2851) loss 3.9656 (3.9246) grad_norm 1.2566 (1.0946) [2021-04-15 18:41:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][570/1251] eta 0:03:14 lr 0.000935 time 0.2938 (0.2851) loss 4.4575 (3.9264) grad_norm 1.1372 (1.0956) [2021-04-15 18:41:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][580/1251] eta 0:03:11 lr 0.000935 time 0.2656 (0.2852) loss 4.2482 (3.9283) grad_norm 1.0655 (1.0957) [2021-04-15 18:42:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][590/1251] eta 0:03:08 lr 0.000935 time 0.2704 (0.2850) loss 3.0868 (3.9282) grad_norm 1.1030 (1.0955) [2021-04-15 18:42:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][600/1251] eta 0:03:05 lr 0.000935 time 0.2770 (0.2849) loss 3.5175 (3.9242) grad_norm 1.2124 (1.0948) [2021-04-15 18:42:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][610/1251] eta 0:03:02 lr 0.000935 time 0.2578 (0.2848) loss 4.5506 (3.9240) grad_norm 1.0967 (1.0945) [2021-04-15 18:42:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][620/1251] eta 0:02:59 lr 0.000935 time 0.2914 (0.2848) loss 4.6892 (3.9227) grad_norm 1.1488 (1.0940) [2021-04-15 18:42:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][630/1251] eta 0:02:56 lr 0.000935 time 0.2526 (0.2847) loss 3.0306 (3.9215) grad_norm 1.1595 (1.0949) [2021-04-15 18:42:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][640/1251] eta 0:02:53 lr 0.000935 time 0.2801 (0.2847) loss 3.9348 (3.9235) grad_norm 1.0880 (1.0967) [2021-04-15 18:42:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][650/1251] eta 0:02:51 lr 0.000935 time 0.2689 (0.2846) loss 3.7173 (3.9293) grad_norm 0.8843 (1.0967) [2021-04-15 18:42:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][660/1251] eta 0:02:48 lr 0.000935 time 0.2707 (0.2846) loss 4.2216 (3.9289) grad_norm 1.0751 (1.0960) [2021-04-15 18:42:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][670/1251] eta 0:02:45 lr 0.000935 time 0.2656 (0.2845) loss 3.7346 (3.9299) grad_norm 1.1277 (1.0962) [2021-04-15 18:42:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][680/1251] eta 0:02:42 lr 0.000935 time 0.2661 (0.2844) loss 3.8350 (3.9322) grad_norm 1.0386 (1.0957) [2021-04-15 18:42:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][690/1251] eta 0:02:39 lr 0.000935 time 0.2864 (0.2843) loss 4.4173 (3.9332) grad_norm 1.1760 (1.0948) [2021-04-15 18:42:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][700/1251] eta 0:02:36 lr 0.000935 time 0.3346 (0.2843) loss 3.8286 (3.9340) grad_norm 1.2606 (1.0950) [2021-04-15 18:42:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][710/1251] eta 0:02:33 lr 0.000935 time 0.2562 (0.2842) loss 4.4516 (3.9373) grad_norm 1.1451 (1.0963) [2021-04-15 18:42:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][720/1251] eta 0:02:30 lr 0.000935 time 0.2795 (0.2843) loss 4.1025 (3.9393) grad_norm 1.0386 (1.0964) [2021-04-15 18:42:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][730/1251] eta 0:02:28 lr 0.000935 time 0.2753 (0.2842) loss 3.7478 (3.9397) grad_norm 1.2590 (1.0962) [2021-04-15 18:42:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][740/1251] eta 0:02:25 lr 0.000935 time 0.2445 (0.2841) loss 4.4589 (3.9423) grad_norm 1.2609 (1.0953) [2021-04-15 18:42:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][750/1251] eta 0:02:22 lr 0.000935 time 0.2750 (0.2841) loss 3.8380 (3.9448) grad_norm 1.1584 (1.0952) [2021-04-15 18:42:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][760/1251] eta 0:02:19 lr 0.000935 time 0.2471 (0.2840) loss 3.3804 (3.9496) grad_norm 0.9649 (1.0951) [2021-04-15 18:42:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][770/1251] eta 0:02:16 lr 0.000935 time 0.2932 (0.2839) loss 3.9088 (3.9461) grad_norm 0.9506 (1.0948) [2021-04-15 18:42:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][780/1251] eta 0:02:13 lr 0.000935 time 0.2688 (0.2840) loss 4.2732 (3.9476) grad_norm 1.0590 (1.0954) [2021-04-15 18:42:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][790/1251] eta 0:02:10 lr 0.000935 time 0.2946 (0.2840) loss 4.3654 (3.9468) grad_norm 1.2273 (1.0956) [2021-04-15 18:43:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][800/1251] eta 0:02:08 lr 0.000935 time 0.2799 (0.2838) loss 4.1790 (3.9483) grad_norm 1.3449 (1.0961) [2021-04-15 18:43:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][810/1251] eta 0:02:05 lr 0.000935 time 0.2647 (0.2838) loss 4.3061 (3.9492) grad_norm 1.2862 (1.0971) [2021-04-15 18:43:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][820/1251] eta 0:02:02 lr 0.000935 time 0.2794 (0.2838) loss 3.2786 (3.9474) grad_norm 1.0291 (1.0976) [2021-04-15 18:43:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][830/1251] eta 0:01:59 lr 0.000935 time 0.2534 (0.2839) loss 4.6947 (3.9468) grad_norm 1.0904 (1.0974) [2021-04-15 18:43:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][840/1251] eta 0:01:56 lr 0.000935 time 0.2759 (0.2837) loss 3.9748 (3.9487) grad_norm 0.9590 (inf) [2021-04-15 18:43:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][850/1251] eta 0:01:53 lr 0.000935 time 0.2804 (0.2837) loss 4.9841 (3.9519) grad_norm 0.9448 (inf) [2021-04-15 18:43:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][860/1251] eta 0:01:50 lr 0.000934 time 0.2800 (0.2837) loss 3.9309 (3.9505) grad_norm 1.0866 (inf) [2021-04-15 18:43:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][870/1251] eta 0:01:48 lr 0.000934 time 0.2533 (0.2837) loss 4.0002 (3.9517) grad_norm 0.9931 (inf) [2021-04-15 18:43:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][880/1251] eta 0:01:45 lr 0.000934 time 0.2550 (0.2836) loss 4.2527 (3.9521) grad_norm 1.3329 (inf) [2021-04-15 18:43:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][890/1251] eta 0:01:42 lr 0.000934 time 0.2767 (0.2835) loss 4.8505 (3.9539) grad_norm 1.0212 (inf) [2021-04-15 18:43:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][900/1251] eta 0:01:39 lr 0.000934 time 0.2655 (0.2835) loss 4.3523 (3.9556) grad_norm 0.9634 (inf) [2021-04-15 18:43:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][910/1251] eta 0:01:36 lr 0.000934 time 0.2835 (0.2835) loss 3.7307 (3.9568) grad_norm 1.1264 (inf) [2021-04-15 18:43:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][920/1251] eta 0:01:33 lr 0.000934 time 0.2845 (0.2834) loss 3.6849 (3.9594) grad_norm 1.1975 (inf) [2021-04-15 18:43:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][930/1251] eta 0:01:30 lr 0.000934 time 0.2733 (0.2834) loss 2.9721 (3.9590) grad_norm 1.1182 (inf) [2021-04-15 18:43:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][940/1251] eta 0:01:28 lr 0.000934 time 0.2929 (0.2834) loss 4.8818 (3.9584) grad_norm 1.2268 (inf) [2021-04-15 18:43:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][950/1251] eta 0:01:25 lr 0.000934 time 0.2709 (0.2834) loss 4.3018 (3.9580) grad_norm 0.9938 (inf) [2021-04-15 18:43:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][960/1251] eta 0:01:22 lr 0.000934 time 0.2657 (0.2834) loss 3.5622 (3.9559) grad_norm 0.9686 (inf) [2021-04-15 18:43:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][970/1251] eta 0:01:19 lr 0.000934 time 0.2905 (0.2834) loss 2.8442 (3.9536) grad_norm 0.9491 (inf) [2021-04-15 18:43:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][980/1251] eta 0:01:16 lr 0.000934 time 0.2752 (0.2834) loss 3.4505 (3.9548) grad_norm 0.9931 (inf) [2021-04-15 18:43:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][990/1251] eta 0:01:13 lr 0.000934 time 0.2975 (0.2833) loss 4.6904 (3.9572) grad_norm 1.2602 (inf) [2021-04-15 18:43:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][1000/1251] eta 0:01:11 lr 0.000934 time 0.2781 (0.2833) loss 4.3182 (3.9545) grad_norm 1.1520 (inf) [2021-04-15 18:43:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][1010/1251] eta 0:01:08 lr 0.000934 time 0.2827 (0.2832) loss 3.9015 (3.9542) grad_norm 1.2101 (inf) [2021-04-15 18:44:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][1020/1251] eta 0:01:05 lr 0.000934 time 0.2797 (0.2832) loss 4.0592 (3.9564) grad_norm 1.1902 (inf) [2021-04-15 18:44:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][1030/1251] eta 0:01:02 lr 0.000934 time 0.2666 (0.2832) loss 4.0068 (3.9559) grad_norm 1.0325 (inf) [2021-04-15 18:44:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][1040/1251] eta 0:00:59 lr 0.000934 time 0.2754 (0.2831) loss 2.8729 (3.9549) grad_norm 1.1063 (inf) [2021-04-15 18:44:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][1050/1251] eta 0:00:56 lr 0.000934 time 0.2946 (0.2830) loss 2.5530 (3.9517) grad_norm 1.2404 (inf) [2021-04-15 18:44:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][1060/1251] eta 0:00:54 lr 0.000934 time 0.3000 (0.2830) loss 4.6916 (3.9522) grad_norm 0.9910 (inf) [2021-04-15 18:44:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][1070/1251] eta 0:00:51 lr 0.000934 time 0.2848 (0.2830) loss 2.9061 (3.9548) grad_norm 1.0307 (inf) [2021-04-15 18:44:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][1080/1251] eta 0:00:48 lr 0.000934 time 0.2563 (0.2829) loss 3.7555 (3.9542) grad_norm 1.0846 (inf) [2021-04-15 18:44:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][1090/1251] eta 0:00:45 lr 0.000934 time 0.2872 (0.2829) loss 3.3272 (3.9519) grad_norm 1.3860 (inf) [2021-04-15 18:44:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][1100/1251] eta 0:00:42 lr 0.000934 time 0.2976 (0.2828) loss 3.8225 (3.9536) grad_norm 1.1392 (inf) [2021-04-15 18:44:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][1110/1251] eta 0:00:39 lr 0.000934 time 0.2654 (0.2827) loss 3.7796 (3.9524) grad_norm 1.0029 (inf) [2021-04-15 18:44:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][1120/1251] eta 0:00:37 lr 0.000934 time 0.2942 (0.2827) loss 4.3665 (3.9542) grad_norm 1.1265 (inf) [2021-04-15 18:44:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][1130/1251] eta 0:00:34 lr 0.000934 time 0.3075 (0.2828) loss 4.0015 (3.9544) grad_norm 1.2443 (inf) [2021-04-15 18:44:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][1140/1251] eta 0:00:31 lr 0.000934 time 0.2755 (0.2829) loss 2.8335 (3.9546) grad_norm 1.0648 (inf) [2021-04-15 18:44:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][1150/1251] eta 0:00:28 lr 0.000934 time 0.2747 (0.2828) loss 3.9303 (3.9563) grad_norm 1.1447 (inf) [2021-04-15 18:44:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][1160/1251] eta 0:00:25 lr 0.000934 time 0.3012 (0.2829) loss 4.2580 (3.9571) grad_norm 0.9038 (inf) [2021-04-15 18:44:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][1170/1251] eta 0:00:22 lr 0.000934 time 0.2793 (0.2829) loss 3.3899 (3.9554) grad_norm 1.0253 (inf) [2021-04-15 18:44:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][1180/1251] eta 0:00:20 lr 0.000934 time 0.2702 (0.2829) loss 4.6010 (3.9561) grad_norm 0.9888 (inf) [2021-04-15 18:44:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][1190/1251] eta 0:00:17 lr 0.000934 time 0.2861 (0.2829) loss 2.2693 (3.9561) grad_norm 0.9007 (inf) [2021-04-15 18:44:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][1200/1251] eta 0:00:14 lr 0.000934 time 0.2778 (0.2828) loss 4.6787 (3.9528) grad_norm 1.0411 (inf) [2021-04-15 18:44:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][1210/1251] eta 0:00:11 lr 0.000934 time 0.2772 (0.2828) loss 3.6446 (3.9545) grad_norm 1.3377 (inf) [2021-04-15 18:44:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][1220/1251] eta 0:00:08 lr 0.000934 time 0.2850 (0.2828) loss 3.5160 (3.9554) grad_norm 1.0080 (inf) [2021-04-15 18:45:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][1230/1251] eta 0:00:05 lr 0.000934 time 0.2894 (0.2828) loss 3.8462 (3.9532) grad_norm 1.0182 (inf) [2021-04-15 18:45:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][1240/1251] eta 0:00:03 lr 0.000934 time 0.2483 (0.2826) loss 3.4528 (3.9510) grad_norm 1.0805 (inf) [2021-04-15 18:45:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [49/300][1250/1251] eta 0:00:00 lr 0.000934 time 0.2485 (0.2824) loss 3.9354 (3.9511) grad_norm 1.2762 (inf) [2021-04-15 18:45:08 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 49 training takes 0:05:55 [2021-04-15 18:45:08 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_49.pth saving...... [2021-04-15 18:45:30 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_49.pth saved !!! [2021-04-15 18:45:31 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.119 (1.119) Loss 1.3281 (1.3281) Acc@1 68.457 (68.457) Acc@5 89.746 (89.746) [2021-04-15 18:45:33 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.124 (0.249) Loss 1.3173 (1.2877) Acc@1 68.066 (69.753) Acc@5 90.527 (89.817) [2021-04-15 18:45:34 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.152 (0.203) Loss 1.2739 (1.2805) Acc@1 69.531 (70.010) Acc@5 89.941 (89.923) [2021-04-15 18:45:37 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.174 (0.240) Loss 1.3360 (1.2824) Acc@1 69.043 (70.035) Acc@5 89.062 (89.976) [2021-04-15 18:45:39 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.220) Loss 1.2843 (1.2767) Acc@1 70.508 (70.270) Acc@5 89.551 (89.991) [2021-04-15 18:45:42 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 70.266 Acc@5 90.064 [2021-04-15 18:45:42 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 70.3% [2021-04-15 18:45:42 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 70.27% [2021-04-15 18:45:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][0/1251] eta 1:40:10 lr 0.000934 time 4.8049 (4.8049) loss 4.5776 (4.5776) grad_norm 1.1230 (1.1230) [2021-04-15 18:45:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][10/1251] eta 0:14:12 lr 0.000934 time 0.2790 (0.6868) loss 3.3675 (3.9177) grad_norm 0.8536 (1.0407) [2021-04-15 18:45:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][20/1251] eta 0:10:06 lr 0.000934 time 0.3009 (0.4927) loss 4.2661 (3.8457) grad_norm 1.2451 (1.0560) [2021-04-15 18:45:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][30/1251] eta 0:08:37 lr 0.000934 time 0.2910 (0.4241) loss 3.8381 (3.8666) grad_norm 0.9204 (1.0469) [2021-04-15 18:45:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][40/1251] eta 0:07:52 lr 0.000934 time 0.2536 (0.3902) loss 4.5125 (3.9284) grad_norm 1.0467 (1.0468) [2021-04-15 18:46:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][50/1251] eta 0:07:22 lr 0.000934 time 0.2745 (0.3687) loss 2.9001 (3.8953) grad_norm 1.0928 (1.0675) [2021-04-15 18:46:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][60/1251] eta 0:07:02 lr 0.000934 time 0.2904 (0.3544) loss 4.1288 (3.8931) grad_norm 1.0032 (1.0930) [2021-04-15 18:46:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][70/1251] eta 0:06:46 lr 0.000934 time 0.2508 (0.3439) loss 4.5465 (3.9170) grad_norm 1.0675 (1.1090) [2021-04-15 18:46:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][80/1251] eta 0:06:33 lr 0.000934 time 0.2667 (0.3360) loss 3.7450 (3.9134) grad_norm 1.2636 (1.1132) [2021-04-15 18:46:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][90/1251] eta 0:06:24 lr 0.000933 time 0.2795 (0.3310) loss 3.7592 (3.9075) grad_norm 1.1010 (1.1066) [2021-04-15 18:46:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][100/1251] eta 0:06:14 lr 0.000933 time 0.2759 (0.3256) loss 2.8443 (3.9002) grad_norm 1.0165 (1.1109) [2021-04-15 18:46:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][110/1251] eta 0:06:08 lr 0.000933 time 0.4565 (0.3227) loss 3.3423 (3.8920) grad_norm 0.9954 (1.1041) [2021-04-15 18:46:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][120/1251] eta 0:06:00 lr 0.000933 time 0.2819 (0.3188) loss 4.3834 (3.9171) grad_norm 1.0120 (1.1086) [2021-04-15 18:46:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][130/1251] eta 0:05:53 lr 0.000933 time 0.2862 (0.3157) loss 3.9419 (3.9237) grad_norm 0.9390 (1.1051) [2021-04-15 18:46:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][140/1251] eta 0:05:48 lr 0.000933 time 0.2600 (0.3138) loss 2.8576 (3.9367) grad_norm 1.1264 (1.1061) [2021-04-15 18:46:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][150/1251] eta 0:05:43 lr 0.000933 time 0.2584 (0.3121) loss 4.0081 (3.9408) grad_norm 1.1021 (1.1091) [2021-04-15 18:46:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][160/1251] eta 0:05:38 lr 0.000933 time 0.2523 (0.3100) loss 3.1820 (3.9237) grad_norm 0.9858 (1.1059) [2021-04-15 18:46:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][170/1251] eta 0:05:33 lr 0.000933 time 0.3021 (0.3085) loss 4.2889 (3.9297) grad_norm 1.2054 (1.1102) [2021-04-15 18:46:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][180/1251] eta 0:05:29 lr 0.000933 time 0.2946 (0.3072) loss 4.4843 (3.9236) grad_norm 1.2213 (1.1134) [2021-04-15 18:46:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][190/1251] eta 0:05:24 lr 0.000933 time 0.2701 (0.3058) loss 3.9272 (3.9397) grad_norm 1.1735 (1.1153) [2021-04-15 18:46:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][200/1251] eta 0:05:19 lr 0.000933 time 0.2851 (0.3042) loss 2.8466 (3.9343) grad_norm 1.1720 (1.1158) [2021-04-15 18:46:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][210/1251] eta 0:05:15 lr 0.000933 time 0.2894 (0.3030) loss 3.8767 (3.9373) grad_norm 1.3143 (1.1162) [2021-04-15 18:46:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][220/1251] eta 0:05:11 lr 0.000933 time 0.2736 (0.3020) loss 3.9473 (3.9444) grad_norm 1.0632 (1.1184) [2021-04-15 18:46:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][230/1251] eta 0:05:07 lr 0.000933 time 0.2699 (0.3012) loss 3.5227 (3.9260) grad_norm 0.8998 (1.1158) [2021-04-15 18:46:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][240/1251] eta 0:05:03 lr 0.000933 time 0.2705 (0.3003) loss 4.4993 (3.9178) grad_norm 1.3393 (1.1170) [2021-04-15 18:46:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][250/1251] eta 0:04:59 lr 0.000933 time 0.3149 (0.2997) loss 3.8791 (3.9248) grad_norm 1.0999 (1.1194) [2021-04-15 18:47:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][260/1251] eta 0:04:56 lr 0.000933 time 0.2713 (0.2990) loss 2.9722 (3.9257) grad_norm 1.0917 (1.1194) [2021-04-15 18:47:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][270/1251] eta 0:04:52 lr 0.000933 time 0.2702 (0.2982) loss 3.7486 (3.9387) grad_norm 1.1639 (1.1171) [2021-04-15 18:47:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][280/1251] eta 0:04:49 lr 0.000933 time 0.2846 (0.2977) loss 4.5825 (3.9332) grad_norm 1.3082 (1.1193) [2021-04-15 18:47:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][290/1251] eta 0:04:45 lr 0.000933 time 0.2686 (0.2971) loss 4.1216 (3.9362) grad_norm 1.1489 (1.1176) [2021-04-15 18:47:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][300/1251] eta 0:04:41 lr 0.000933 time 0.2605 (0.2964) loss 4.2881 (3.9360) grad_norm 1.2881 (1.1195) [2021-04-15 18:47:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][310/1251] eta 0:04:38 lr 0.000933 time 0.2699 (0.2958) loss 3.9596 (3.9301) grad_norm 0.9035 (1.1179) [2021-04-15 18:47:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][320/1251] eta 0:04:34 lr 0.000933 time 0.2877 (0.2954) loss 3.0664 (3.9304) grad_norm 1.0445 (1.1188) [2021-04-15 18:47:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][330/1251] eta 0:04:31 lr 0.000933 time 0.2874 (0.2951) loss 3.5171 (3.9429) grad_norm 1.2910 (1.1180) [2021-04-15 18:47:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][340/1251] eta 0:04:28 lr 0.000933 time 0.2788 (0.2952) loss 4.3137 (3.9510) grad_norm 0.9779 (1.1157) [2021-04-15 18:47:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][350/1251] eta 0:04:26 lr 0.000933 time 0.3129 (0.2952) loss 3.2207 (3.9473) grad_norm 1.2706 (1.1168) [2021-04-15 18:47:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][360/1251] eta 0:04:22 lr 0.000933 time 0.2608 (0.2947) loss 4.2765 (3.9476) grad_norm 0.9332 (1.1159) [2021-04-15 18:47:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][370/1251] eta 0:04:19 lr 0.000933 time 0.2577 (0.2947) loss 4.7230 (3.9505) grad_norm 1.5503 (1.1179) [2021-04-15 18:47:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][380/1251] eta 0:04:16 lr 0.000933 time 0.2790 (0.2943) loss 4.7091 (3.9573) grad_norm 1.3448 (1.1220) [2021-04-15 18:47:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][390/1251] eta 0:04:12 lr 0.000933 time 0.2731 (0.2938) loss 4.4773 (3.9601) grad_norm 1.0217 (1.1215) [2021-04-15 18:47:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][400/1251] eta 0:04:09 lr 0.000933 time 0.2799 (0.2934) loss 4.4908 (3.9621) grad_norm 1.2121 (1.1203) [2021-04-15 18:47:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][410/1251] eta 0:04:06 lr 0.000933 time 0.2822 (0.2933) loss 4.7200 (3.9672) grad_norm 1.2875 (1.1229) [2021-04-15 18:47:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][420/1251] eta 0:04:03 lr 0.000933 time 0.2759 (0.2928) loss 3.9316 (3.9690) grad_norm 1.0823 (1.1214) [2021-04-15 18:47:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][430/1251] eta 0:04:00 lr 0.000933 time 0.2879 (0.2926) loss 4.3050 (3.9644) grad_norm 0.9708 (1.1209) [2021-04-15 18:47:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][440/1251] eta 0:03:56 lr 0.000933 time 0.2685 (0.2922) loss 4.5606 (3.9654) grad_norm 1.1174 (1.1192) [2021-04-15 18:47:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][450/1251] eta 0:03:53 lr 0.000933 time 0.2943 (0.2920) loss 3.6817 (3.9667) grad_norm 0.9126 (1.1180) [2021-04-15 18:47:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][460/1251] eta 0:03:50 lr 0.000933 time 0.2954 (0.2917) loss 4.5165 (3.9697) grad_norm 1.0133 (1.1165) [2021-04-15 18:47:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][470/1251] eta 0:03:47 lr 0.000933 time 0.2702 (0.2914) loss 4.1684 (3.9712) grad_norm 1.1273 (1.1151) [2021-04-15 18:48:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][480/1251] eta 0:03:44 lr 0.000933 time 0.3239 (0.2913) loss 4.8327 (3.9677) grad_norm 0.9920 (1.1154) [2021-04-15 18:48:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][490/1251] eta 0:03:41 lr 0.000933 time 0.2843 (0.2910) loss 3.5081 (3.9664) grad_norm 0.9381 (1.1153) [2021-04-15 18:48:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][500/1251] eta 0:03:38 lr 0.000933 time 0.2897 (0.2908) loss 3.7839 (3.9624) grad_norm 1.0131 (1.1126) [2021-04-15 18:48:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][510/1251] eta 0:03:35 lr 0.000933 time 0.2736 (0.2905) loss 4.3937 (3.9611) grad_norm 1.1544 (1.1132) [2021-04-15 18:48:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][520/1251] eta 0:03:32 lr 0.000933 time 0.2733 (0.2904) loss 4.4246 (3.9641) grad_norm 1.1438 (1.1134) [2021-04-15 18:48:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][530/1251] eta 0:03:29 lr 0.000933 time 0.2838 (0.2903) loss 3.4341 (3.9630) grad_norm 1.1756 (1.1119) [2021-04-15 18:48:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][540/1251] eta 0:03:26 lr 0.000933 time 0.2713 (0.2901) loss 4.8836 (3.9629) grad_norm 1.1386 (1.1110) [2021-04-15 18:48:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][550/1251] eta 0:03:23 lr 0.000933 time 0.4674 (0.2903) loss 3.0819 (3.9618) grad_norm 1.1976 (1.1101) [2021-04-15 18:48:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][560/1251] eta 0:03:20 lr 0.000933 time 0.2854 (0.2901) loss 3.9945 (3.9621) grad_norm 1.0061 (1.1093) [2021-04-15 18:48:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][570/1251] eta 0:03:17 lr 0.000932 time 0.2693 (0.2903) loss 3.0431 (3.9671) grad_norm 1.1040 (1.1085) [2021-04-15 18:48:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][580/1251] eta 0:03:14 lr 0.000932 time 0.2870 (0.2903) loss 3.7636 (3.9712) grad_norm 1.2655 (1.1079) [2021-04-15 18:48:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][590/1251] eta 0:03:11 lr 0.000932 time 0.2997 (0.2901) loss 4.4833 (3.9721) grad_norm 1.3866 (1.1091) [2021-04-15 18:48:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][600/1251] eta 0:03:08 lr 0.000932 time 0.2814 (0.2900) loss 3.5155 (3.9746) grad_norm 1.0938 (1.1107) [2021-04-15 18:48:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][610/1251] eta 0:03:05 lr 0.000932 time 0.2864 (0.2898) loss 3.7237 (3.9751) grad_norm 1.1006 (1.1114) [2021-04-15 18:48:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][620/1251] eta 0:03:02 lr 0.000932 time 0.2548 (0.2898) loss 3.7136 (3.9739) grad_norm 1.0835 (1.1107) [2021-04-15 18:48:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][630/1251] eta 0:02:59 lr 0.000932 time 0.2955 (0.2898) loss 2.9904 (3.9688) grad_norm 1.0536 (1.1105) [2021-04-15 18:48:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][640/1251] eta 0:02:57 lr 0.000932 time 0.2857 (0.2897) loss 3.5166 (3.9651) grad_norm 0.9635 (1.1101) [2021-04-15 18:48:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][650/1251] eta 0:02:54 lr 0.000932 time 0.2896 (0.2897) loss 4.1876 (3.9661) grad_norm 1.3894 (1.1115) [2021-04-15 18:48:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][660/1251] eta 0:02:51 lr 0.000932 time 0.2835 (0.2896) loss 3.8353 (3.9630) grad_norm 0.8881 (1.1106) [2021-04-15 18:48:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][670/1251] eta 0:02:48 lr 0.000932 time 0.2604 (0.2894) loss 4.0357 (3.9584) grad_norm 0.9420 (1.1095) [2021-04-15 18:48:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][680/1251] eta 0:02:45 lr 0.000932 time 0.2884 (0.2892) loss 4.2939 (3.9609) grad_norm 0.9430 (1.1079) [2021-04-15 18:49:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][690/1251] eta 0:02:42 lr 0.000932 time 0.2982 (0.2892) loss 4.2525 (3.9580) grad_norm 1.0608 (1.1069) [2021-04-15 18:49:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][700/1251] eta 0:02:39 lr 0.000932 time 0.2947 (0.2891) loss 4.1040 (3.9556) grad_norm 1.0492 (1.1058) [2021-04-15 18:49:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][710/1251] eta 0:02:36 lr 0.000932 time 0.2861 (0.2889) loss 3.8603 (3.9585) grad_norm 0.9968 (1.1056) [2021-04-15 18:49:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][720/1251] eta 0:02:33 lr 0.000932 time 0.2402 (0.2890) loss 4.1875 (3.9585) grad_norm 1.1558 (1.1055) [2021-04-15 18:49:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][730/1251] eta 0:02:30 lr 0.000932 time 0.2807 (0.2888) loss 4.7822 (3.9578) grad_norm 1.1271 (1.1081) [2021-04-15 18:49:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][740/1251] eta 0:02:27 lr 0.000932 time 0.2828 (0.2887) loss 4.0034 (3.9548) grad_norm 0.9827 (1.1079) [2021-04-15 18:49:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][750/1251] eta 0:02:24 lr 0.000932 time 0.2764 (0.2885) loss 2.7871 (3.9489) grad_norm 1.1120 (1.1074) [2021-04-15 18:49:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][760/1251] eta 0:02:21 lr 0.000932 time 0.2912 (0.2883) loss 4.4879 (3.9534) grad_norm 1.2373 (1.1065) [2021-04-15 18:49:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][770/1251] eta 0:02:18 lr 0.000932 time 0.2923 (0.2882) loss 2.7244 (3.9509) grad_norm 1.0344 (1.1053) [2021-04-15 18:49:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][780/1251] eta 0:02:15 lr 0.000932 time 0.2646 (0.2880) loss 5.0016 (3.9523) grad_norm 1.1287 (1.1058) [2021-04-15 18:49:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][790/1251] eta 0:02:12 lr 0.000932 time 0.2862 (0.2879) loss 3.1433 (3.9505) grad_norm 1.1371 (1.1060) [2021-04-15 18:49:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][800/1251] eta 0:02:09 lr 0.000932 time 0.2591 (0.2878) loss 4.3569 (3.9536) grad_norm 1.2869 (1.1062) [2021-04-15 18:49:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][810/1251] eta 0:02:06 lr 0.000932 time 0.2759 (0.2877) loss 3.9438 (3.9543) grad_norm 1.0631 (1.1075) [2021-04-15 18:49:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][820/1251] eta 0:02:03 lr 0.000932 time 0.3151 (0.2876) loss 4.7251 (3.9564) grad_norm 1.2213 (1.1066) [2021-04-15 18:49:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][830/1251] eta 0:02:01 lr 0.000932 time 0.2729 (0.2875) loss 3.4448 (3.9544) grad_norm 1.2282 (1.1078) [2021-04-15 18:49:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][840/1251] eta 0:01:58 lr 0.000932 time 0.2799 (0.2875) loss 4.6743 (3.9560) grad_norm 1.1823 (1.1074) [2021-04-15 18:49:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][850/1251] eta 0:01:55 lr 0.000932 time 0.3006 (0.2874) loss 4.3099 (3.9555) grad_norm 1.1822 (1.1069) [2021-04-15 18:49:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][860/1251] eta 0:01:52 lr 0.000932 time 0.2619 (0.2872) loss 4.5178 (3.9588) grad_norm 0.9214 (1.1069) [2021-04-15 18:49:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][870/1251] eta 0:01:49 lr 0.000932 time 0.2820 (0.2872) loss 4.7076 (3.9604) grad_norm 0.9651 (1.1070) [2021-04-15 18:49:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][880/1251] eta 0:01:46 lr 0.000932 time 0.2839 (0.2871) loss 4.5637 (3.9594) grad_norm 1.0408 (1.1070) [2021-04-15 18:49:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][890/1251] eta 0:01:43 lr 0.000932 time 0.2992 (0.2871) loss 3.8862 (3.9578) grad_norm 1.0234 (1.1067) [2021-04-15 18:50:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][900/1251] eta 0:01:40 lr 0.000932 time 0.2727 (0.2871) loss 3.3831 (3.9581) grad_norm 0.9747 (1.1067) [2021-04-15 18:50:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][910/1251] eta 0:01:37 lr 0.000932 time 0.2778 (0.2870) loss 3.7252 (3.9577) grad_norm 1.1623 (1.1071) [2021-04-15 18:50:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][920/1251] eta 0:01:35 lr 0.000932 time 0.3023 (0.2870) loss 4.3017 (3.9577) grad_norm 1.0286 (1.1074) [2021-04-15 18:50:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][930/1251] eta 0:01:32 lr 0.000932 time 0.2824 (0.2870) loss 4.1126 (3.9588) grad_norm 1.0076 (1.1071) [2021-04-15 18:50:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][940/1251] eta 0:01:29 lr 0.000932 time 0.2799 (0.2872) loss 3.6207 (3.9611) grad_norm 1.0432 (1.1073) [2021-04-15 18:50:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][950/1251] eta 0:01:26 lr 0.000932 time 0.2694 (0.2871) loss 4.4858 (3.9630) grad_norm 0.9309 (1.1071) [2021-04-15 18:50:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][960/1251] eta 0:01:23 lr 0.000932 time 0.2857 (0.2870) loss 4.0772 (3.9626) grad_norm 0.9666 (1.1078) [2021-04-15 18:50:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][970/1251] eta 0:01:20 lr 0.000932 time 0.2938 (0.2870) loss 4.3003 (3.9619) grad_norm 1.0193 (1.1077) [2021-04-15 18:50:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][980/1251] eta 0:01:17 lr 0.000932 time 0.2817 (0.2869) loss 3.8910 (3.9603) grad_norm 1.0041 (1.1080) [2021-04-15 18:50:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][990/1251] eta 0:01:14 lr 0.000932 time 0.2737 (0.2868) loss 4.4093 (3.9601) grad_norm 1.0036 (1.1075) [2021-04-15 18:50:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][1000/1251] eta 0:01:11 lr 0.000932 time 0.3104 (0.2867) loss 4.0151 (3.9612) grad_norm 1.2288 (1.1085) [2021-04-15 18:50:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][1010/1251] eta 0:01:09 lr 0.000932 time 0.2675 (0.2866) loss 3.4840 (3.9614) grad_norm 1.3237 (1.1084) [2021-04-15 18:50:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][1020/1251] eta 0:01:06 lr 0.000932 time 0.2638 (0.2865) loss 4.1135 (3.9629) grad_norm 1.3697 (1.1087) [2021-04-15 18:50:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][1030/1251] eta 0:01:03 lr 0.000932 time 0.2852 (0.2864) loss 3.7393 (3.9593) grad_norm 0.9948 (1.1091) [2021-04-15 18:50:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][1040/1251] eta 0:01:00 lr 0.000932 time 0.2835 (0.2864) loss 3.4957 (3.9568) grad_norm 0.9349 (1.1083) [2021-04-15 18:50:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][1050/1251] eta 0:00:57 lr 0.000931 time 0.2925 (0.2863) loss 3.8031 (3.9580) grad_norm 0.9698 (1.1073) [2021-04-15 18:50:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][1060/1251] eta 0:00:54 lr 0.000931 time 0.2551 (0.2863) loss 3.8680 (3.9561) grad_norm 1.4327 (1.1078) [2021-04-15 18:50:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][1070/1251] eta 0:00:51 lr 0.000931 time 0.2729 (0.2863) loss 4.0733 (3.9561) grad_norm 1.2069 (1.1073) [2021-04-15 18:50:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][1080/1251] eta 0:00:48 lr 0.000931 time 0.2635 (0.2862) loss 3.0063 (3.9537) grad_norm 0.9778 (1.1073) [2021-04-15 18:50:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][1090/1251] eta 0:00:46 lr 0.000931 time 0.2901 (0.2862) loss 3.6815 (3.9526) grad_norm 1.2939 (1.1067) [2021-04-15 18:50:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][1100/1251] eta 0:00:43 lr 0.000931 time 0.2779 (0.2861) loss 4.8987 (3.9549) grad_norm 1.1477 (1.1069) [2021-04-15 18:51:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][1110/1251] eta 0:00:40 lr 0.000931 time 0.2727 (0.2860) loss 4.0535 (3.9552) grad_norm 1.0068 (1.1068) [2021-04-15 18:51:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][1120/1251] eta 0:00:37 lr 0.000931 time 0.2966 (0.2860) loss 2.8862 (3.9527) grad_norm 1.3638 (1.1072) [2021-04-15 18:51:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][1130/1251] eta 0:00:34 lr 0.000931 time 0.2861 (0.2860) loss 3.4310 (3.9528) grad_norm 0.8989 (1.1072) [2021-04-15 18:51:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][1140/1251] eta 0:00:31 lr 0.000931 time 0.4161 (0.2861) loss 3.2325 (3.9527) grad_norm 1.1059 (1.1067) [2021-04-15 18:51:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][1150/1251] eta 0:00:28 lr 0.000931 time 0.2851 (0.2861) loss 4.5346 (3.9512) grad_norm 1.0158 (1.1062) [2021-04-15 18:51:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][1160/1251] eta 0:00:26 lr 0.000931 time 0.2428 (0.2861) loss 4.3971 (3.9528) grad_norm 1.0598 (1.1060) [2021-04-15 18:51:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][1170/1251] eta 0:00:23 lr 0.000931 time 0.2814 (0.2860) loss 4.5117 (3.9544) grad_norm 0.9890 (1.1051) [2021-04-15 18:51:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][1180/1251] eta 0:00:20 lr 0.000931 time 0.2827 (0.2860) loss 3.2237 (3.9531) grad_norm 1.0087 (1.1051) [2021-04-15 18:51:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][1190/1251] eta 0:00:17 lr 0.000931 time 0.4355 (0.2860) loss 4.0300 (3.9558) grad_norm 1.1283 (1.1056) [2021-04-15 18:51:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][1200/1251] eta 0:00:14 lr 0.000931 time 0.2573 (0.2860) loss 2.9900 (3.9523) grad_norm 0.8975 (1.1048) [2021-04-15 18:51:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][1210/1251] eta 0:00:11 lr 0.000931 time 0.2778 (0.2859) loss 4.3942 (3.9528) grad_norm 1.1319 (1.1052) [2021-04-15 18:51:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][1220/1251] eta 0:00:08 lr 0.000931 time 0.2705 (0.2858) loss 3.8053 (3.9531) grad_norm 1.3346 (1.1055) [2021-04-15 18:51:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][1230/1251] eta 0:00:06 lr 0.000931 time 0.2682 (0.2858) loss 3.1068 (3.9539) grad_norm 1.0627 (1.1056) [2021-04-15 18:51:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][1240/1251] eta 0:00:03 lr 0.000931 time 0.2485 (0.2856) loss 3.2761 (3.9539) grad_norm 0.9987 (1.1048) [2021-04-15 18:51:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [50/300][1250/1251] eta 0:00:00 lr 0.000931 time 0.2487 (0.2853) loss 3.6904 (3.9543) grad_norm 1.1626 (1.1049) [2021-04-15 18:51:41 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 50 training takes 0:05:58 [2021-04-15 18:51:41 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_50.pth saving...... [2021-04-15 18:51:54 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_50.pth saved !!! [2021-04-15 18:51:56 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.200 (1.200) Loss 1.2862 (1.2862) Acc@1 70.605 (70.605) Acc@5 90.430 (90.430) [2021-04-15 18:51:57 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.418 (0.254) Loss 1.2754 (1.3130) Acc@1 71.094 (69.798) Acc@5 90.430 (90.048) [2021-04-15 18:51:59 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.132 (0.233) Loss 1.2634 (1.3129) Acc@1 70.898 (69.820) Acc@5 90.332 (90.053) [2021-04-15 18:52:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.136 (0.230) Loss 1.2802 (1.3017) Acc@1 71.680 (70.168) Acc@5 89.648 (90.219) [2021-04-15 18:52:04 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.227) Loss 1.2839 (1.3024) Acc@1 71.484 (70.315) Acc@5 89.941 (90.118) [2021-04-15 18:52:07 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 70.364 Acc@5 90.090 [2021-04-15 18:52:07 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 70.4% [2021-04-15 18:52:07 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 70.36% [2021-04-15 18:52:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][0/1251] eta 2:10:44 lr 0.000931 time 6.2705 (6.2705) loss 3.9285 (3.9285) grad_norm 0.9164 (0.9164) [2021-04-15 18:52:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][10/1251] eta 0:16:59 lr 0.000931 time 0.2845 (0.8212) loss 4.1249 (3.9023) grad_norm 1.0777 (1.0458) [2021-04-15 18:52:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][20/1251] eta 0:11:35 lr 0.000931 time 0.2858 (0.5647) loss 4.3425 (3.8841) grad_norm 1.2654 (1.0613) [2021-04-15 18:52:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][30/1251] eta 0:09:38 lr 0.000931 time 0.2931 (0.4734) loss 3.6826 (3.9002) grad_norm 1.2278 (1.0658) [2021-04-15 18:52:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][40/1251] eta 0:08:37 lr 0.000931 time 0.2705 (0.4271) loss 3.7618 (3.9632) grad_norm 1.4389 (1.0620) [2021-04-15 18:52:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][50/1251] eta 0:07:58 lr 0.000931 time 0.2836 (0.3984) loss 4.0559 (4.0259) grad_norm 0.9643 (1.0553) [2021-04-15 18:52:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][60/1251] eta 0:07:31 lr 0.000931 time 0.2756 (0.3792) loss 3.9581 (4.0235) grad_norm 1.1442 (1.0653) [2021-04-15 18:52:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][70/1251] eta 0:07:12 lr 0.000931 time 0.2909 (0.3662) loss 4.2449 (4.0013) grad_norm 1.1035 (1.0685) [2021-04-15 18:52:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][80/1251] eta 0:06:57 lr 0.000931 time 0.3023 (0.3563) loss 3.1779 (3.9928) grad_norm 0.9642 (1.0733) [2021-04-15 18:52:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][90/1251] eta 0:06:43 lr 0.000931 time 0.2728 (0.3479) loss 4.0994 (3.9932) grad_norm 1.0027 (1.0838) [2021-04-15 18:52:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][100/1251] eta 0:06:33 lr 0.000931 time 0.2951 (0.3419) loss 3.9076 (3.9940) grad_norm 1.1103 (1.0826) [2021-04-15 18:52:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][110/1251] eta 0:06:23 lr 0.000931 time 0.2701 (0.3361) loss 4.1317 (3.9902) grad_norm 1.2117 (1.0927) [2021-04-15 18:52:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][120/1251] eta 0:06:16 lr 0.000931 time 0.2819 (0.3331) loss 3.3889 (3.9738) grad_norm 1.3779 (1.0943) [2021-04-15 18:52:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][130/1251] eta 0:06:09 lr 0.000931 time 0.2839 (0.3299) loss 4.3451 (3.9811) grad_norm 1.1522 (1.0912) [2021-04-15 18:52:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][140/1251] eta 0:06:03 lr 0.000931 time 0.2940 (0.3272) loss 4.0757 (3.9818) grad_norm 1.3414 (1.0984) [2021-04-15 18:52:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][150/1251] eta 0:05:56 lr 0.000931 time 0.2881 (0.3241) loss 3.7737 (3.9902) grad_norm 1.0977 (1.1004) [2021-04-15 18:52:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][160/1251] eta 0:05:50 lr 0.000931 time 0.2770 (0.3216) loss 3.4784 (3.9722) grad_norm 1.4218 (1.1043) [2021-04-15 18:53:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][170/1251] eta 0:05:45 lr 0.000931 time 0.3093 (0.3193) loss 4.4173 (3.9784) grad_norm 1.1252 (1.1028) [2021-04-15 18:53:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][180/1251] eta 0:05:39 lr 0.000931 time 0.2718 (0.3169) loss 3.7952 (3.9778) grad_norm 1.1535 (1.1020) [2021-04-15 18:53:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][190/1251] eta 0:05:34 lr 0.000931 time 0.2886 (0.3152) loss 4.4101 (3.9775) grad_norm 1.0200 (1.1062) [2021-04-15 18:53:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][200/1251] eta 0:05:29 lr 0.000931 time 0.2772 (0.3136) loss 4.5022 (3.9841) grad_norm 1.1634 (1.1024) [2021-04-15 18:53:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][210/1251] eta 0:05:24 lr 0.000931 time 0.3001 (0.3121) loss 4.1011 (3.9726) grad_norm 1.0854 (1.1015) [2021-04-15 18:53:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][220/1251] eta 0:05:20 lr 0.000931 time 0.2549 (0.3105) loss 4.0707 (3.9660) grad_norm 1.1122 (1.1021) [2021-04-15 18:53:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][230/1251] eta 0:05:15 lr 0.000931 time 0.2871 (0.3093) loss 3.9670 (3.9706) grad_norm 1.0727 (1.0998) [2021-04-15 18:53:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][240/1251] eta 0:05:11 lr 0.000931 time 0.2573 (0.3081) loss 3.7162 (3.9651) grad_norm 1.3406 (1.1026) [2021-04-15 18:53:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][250/1251] eta 0:05:07 lr 0.000931 time 0.2672 (0.3070) loss 3.8797 (3.9725) grad_norm 0.9928 (1.1064) [2021-04-15 18:53:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][260/1251] eta 0:05:03 lr 0.000931 time 0.2694 (0.3059) loss 4.7284 (3.9796) grad_norm 1.0949 (1.1093) [2021-04-15 18:53:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][270/1251] eta 0:04:58 lr 0.000930 time 0.2731 (0.3048) loss 3.6817 (3.9877) grad_norm 1.0816 (1.1053) [2021-04-15 18:53:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][280/1251] eta 0:04:55 lr 0.000930 time 0.2834 (0.3040) loss 4.2753 (3.9917) grad_norm 1.2781 (1.1053) [2021-04-15 18:53:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][290/1251] eta 0:04:51 lr 0.000930 time 0.2782 (0.3031) loss 3.8176 (3.9789) grad_norm 1.0961 (1.1076) [2021-04-15 18:53:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][300/1251] eta 0:04:47 lr 0.000930 time 0.2786 (0.3024) loss 4.4678 (3.9709) grad_norm 1.1653 (1.1076) [2021-04-15 18:53:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][310/1251] eta 0:04:43 lr 0.000930 time 0.2784 (0.3016) loss 2.8978 (3.9712) grad_norm 0.9563 (1.1116) [2021-04-15 18:53:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][320/1251] eta 0:04:40 lr 0.000930 time 0.2938 (0.3008) loss 3.4326 (3.9550) grad_norm 1.0582 (1.1095) [2021-04-15 18:53:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][330/1251] eta 0:04:36 lr 0.000930 time 0.2751 (0.3002) loss 3.4397 (3.9548) grad_norm 0.9032 (1.1070) [2021-04-15 18:53:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][340/1251] eta 0:04:33 lr 0.000930 time 0.2726 (0.3002) loss 2.7254 (3.9501) grad_norm 1.0857 (inf) [2021-04-15 18:53:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][350/1251] eta 0:04:29 lr 0.000930 time 0.2878 (0.2996) loss 4.1603 (3.9494) grad_norm 1.0503 (inf) [2021-04-15 18:53:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][360/1251] eta 0:04:26 lr 0.000930 time 0.2785 (0.2991) loss 3.5489 (3.9526) grad_norm 1.1767 (inf) [2021-04-15 18:53:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][370/1251] eta 0:04:23 lr 0.000930 time 0.2430 (0.2989) loss 3.8304 (3.9525) grad_norm 1.0530 (inf) [2021-04-15 18:54:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][380/1251] eta 0:04:19 lr 0.000930 time 0.2696 (0.2984) loss 4.4083 (3.9531) grad_norm 1.1709 (inf) [2021-04-15 18:54:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][390/1251] eta 0:04:16 lr 0.000930 time 0.2543 (0.2984) loss 4.4639 (3.9569) grad_norm 1.2001 (inf) [2021-04-15 18:54:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][400/1251] eta 0:04:13 lr 0.000930 time 0.2791 (0.2979) loss 4.6723 (3.9593) grad_norm 1.0084 (inf) [2021-04-15 18:54:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][410/1251] eta 0:04:10 lr 0.000930 time 0.2805 (0.2975) loss 4.0288 (3.9575) grad_norm 1.0415 (inf) [2021-04-15 18:54:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][420/1251] eta 0:04:06 lr 0.000930 time 0.2993 (0.2971) loss 4.3900 (3.9570) grad_norm 1.1299 (inf) [2021-04-15 18:54:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][430/1251] eta 0:04:03 lr 0.000930 time 0.2717 (0.2967) loss 4.4574 (3.9582) grad_norm 1.0124 (inf) [2021-04-15 18:54:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][440/1251] eta 0:04:00 lr 0.000930 time 0.2825 (0.2962) loss 4.0940 (3.9495) grad_norm 1.0881 (inf) [2021-04-15 18:54:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][450/1251] eta 0:03:57 lr 0.000930 time 0.2800 (0.2959) loss 4.8640 (3.9562) grad_norm 0.9094 (inf) [2021-04-15 18:54:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][460/1251] eta 0:03:53 lr 0.000930 time 0.2774 (0.2955) loss 4.6751 (3.9520) grad_norm 0.8817 (inf) [2021-04-15 18:54:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][470/1251] eta 0:03:50 lr 0.000930 time 0.3197 (0.2954) loss 4.2971 (3.9511) grad_norm 0.9449 (inf) [2021-04-15 18:54:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][480/1251] eta 0:03:47 lr 0.000930 time 0.2682 (0.2951) loss 4.4024 (3.9450) grad_norm 1.1627 (inf) [2021-04-15 18:54:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][490/1251] eta 0:03:44 lr 0.000930 time 0.2612 (0.2947) loss 3.6979 (3.9426) grad_norm 0.9752 (inf) [2021-04-15 18:54:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][500/1251] eta 0:03:41 lr 0.000930 time 0.2766 (0.2945) loss 3.7976 (3.9326) grad_norm 1.5462 (inf) [2021-04-15 18:54:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][510/1251] eta 0:03:38 lr 0.000930 time 0.2923 (0.2942) loss 4.4162 (3.9371) grad_norm 1.1030 (inf) [2021-04-15 18:54:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][520/1251] eta 0:03:34 lr 0.000930 time 0.2851 (0.2938) loss 4.3274 (3.9360) grad_norm 1.3200 (inf) [2021-04-15 18:54:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][530/1251] eta 0:03:31 lr 0.000930 time 0.2648 (0.2935) loss 4.2812 (3.9336) grad_norm 1.1140 (inf) [2021-04-15 18:54:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][540/1251] eta 0:03:28 lr 0.000930 time 0.2773 (0.2932) loss 4.1771 (3.9294) grad_norm 0.9938 (inf) [2021-04-15 18:54:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][550/1251] eta 0:03:25 lr 0.000930 time 0.2665 (0.2929) loss 4.4991 (3.9274) grad_norm 1.2678 (inf) [2021-04-15 18:54:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][560/1251] eta 0:03:22 lr 0.000930 time 0.2704 (0.2927) loss 4.2726 (3.9321) grad_norm 1.0872 (inf) [2021-04-15 18:54:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][570/1251] eta 0:03:19 lr 0.000930 time 0.2736 (0.2924) loss 3.4376 (3.9311) grad_norm 1.0054 (inf) [2021-04-15 18:54:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][580/1251] eta 0:03:16 lr 0.000930 time 0.2537 (0.2924) loss 4.3676 (3.9332) grad_norm 1.0808 (inf) [2021-04-15 18:55:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][590/1251] eta 0:03:13 lr 0.000930 time 0.2838 (0.2922) loss 4.2392 (3.9301) grad_norm 0.9749 (inf) [2021-04-15 18:55:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][600/1251] eta 0:03:10 lr 0.000930 time 0.2769 (0.2919) loss 4.4968 (3.9270) grad_norm 1.1140 (inf) [2021-04-15 18:55:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][610/1251] eta 0:03:06 lr 0.000930 time 0.2817 (0.2917) loss 4.0379 (3.9240) grad_norm 1.2068 (inf) [2021-04-15 18:55:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][620/1251] eta 0:03:03 lr 0.000930 time 0.3087 (0.2914) loss 4.4574 (3.9242) grad_norm 1.0910 (inf) [2021-04-15 18:55:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][630/1251] eta 0:03:00 lr 0.000930 time 0.2953 (0.2912) loss 4.3076 (3.9225) grad_norm 1.0334 (inf) [2021-04-15 18:55:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][640/1251] eta 0:02:57 lr 0.000930 time 0.2713 (0.2910) loss 3.7601 (3.9254) grad_norm 0.9669 (inf) [2021-04-15 18:55:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][650/1251] eta 0:02:54 lr 0.000930 time 0.2755 (0.2908) loss 3.7194 (3.9276) grad_norm 1.4487 (inf) [2021-04-15 18:55:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][660/1251] eta 0:02:51 lr 0.000930 time 0.2716 (0.2907) loss 4.1963 (3.9283) grad_norm 1.1182 (inf) [2021-04-15 18:55:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][670/1251] eta 0:02:48 lr 0.000930 time 0.2816 (0.2905) loss 3.1622 (3.9283) grad_norm 0.9966 (inf) [2021-04-15 18:55:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][680/1251] eta 0:02:45 lr 0.000930 time 0.2771 (0.2904) loss 4.2641 (3.9268) grad_norm 0.8754 (inf) [2021-04-15 18:55:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][690/1251] eta 0:02:42 lr 0.000930 time 0.2567 (0.2904) loss 4.1832 (3.9312) grad_norm 1.1163 (inf) [2021-04-15 18:55:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][700/1251] eta 0:02:39 lr 0.000930 time 0.2915 (0.2903) loss 4.1197 (3.9274) grad_norm 1.0616 (inf) [2021-04-15 18:55:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][710/1251] eta 0:02:37 lr 0.000930 time 0.2514 (0.2903) loss 3.6986 (3.9272) grad_norm 1.0873 (inf) [2021-04-15 18:55:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][720/1251] eta 0:02:34 lr 0.000930 time 0.2885 (0.2902) loss 3.2226 (3.9242) grad_norm 1.3271 (inf) [2021-04-15 18:55:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][730/1251] eta 0:02:31 lr 0.000930 time 0.2880 (0.2901) loss 2.5849 (3.9256) grad_norm 1.0525 (inf) [2021-04-15 18:55:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][740/1251] eta 0:02:28 lr 0.000929 time 0.2847 (0.2902) loss 4.0701 (3.9254) grad_norm 1.0079 (inf) [2021-04-15 18:55:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][750/1251] eta 0:02:25 lr 0.000929 time 0.2988 (0.2901) loss 3.7936 (3.9269) grad_norm 1.1065 (inf) [2021-04-15 18:55:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][760/1251] eta 0:02:22 lr 0.000929 time 0.2768 (0.2900) loss 4.4186 (3.9248) grad_norm 1.1976 (inf) [2021-04-15 18:55:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][770/1251] eta 0:02:19 lr 0.000929 time 0.2829 (0.2899) loss 3.4506 (3.9184) grad_norm 1.1034 (inf) [2021-04-15 18:55:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][780/1251] eta 0:02:16 lr 0.000929 time 0.2981 (0.2897) loss 4.5572 (3.9199) grad_norm 1.0259 (inf) [2021-04-15 18:55:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][790/1251] eta 0:02:13 lr 0.000929 time 0.2907 (0.2896) loss 4.1425 (3.9226) grad_norm 1.0855 (inf) [2021-04-15 18:55:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][800/1251] eta 0:02:10 lr 0.000929 time 0.2871 (0.2895) loss 3.4422 (3.9233) grad_norm 1.0568 (inf) [2021-04-15 18:56:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][810/1251] eta 0:02:07 lr 0.000929 time 0.2946 (0.2895) loss 4.1991 (3.9194) grad_norm 1.1908 (inf) [2021-04-15 18:56:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][820/1251] eta 0:02:04 lr 0.000929 time 0.2841 (0.2893) loss 2.6810 (3.9179) grad_norm 1.0508 (inf) [2021-04-15 18:56:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][830/1251] eta 0:02:01 lr 0.000929 time 0.2874 (0.2892) loss 2.9113 (3.9148) grad_norm 1.1097 (inf) [2021-04-15 18:56:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][840/1251] eta 0:01:58 lr 0.000929 time 0.2889 (0.2891) loss 4.5678 (3.9139) grad_norm 1.1147 (inf) [2021-04-15 18:56:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][850/1251] eta 0:01:55 lr 0.000929 time 0.2790 (0.2890) loss 4.4401 (3.9152) grad_norm 1.3885 (inf) [2021-04-15 18:56:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][860/1251] eta 0:01:52 lr 0.000929 time 0.2854 (0.2888) loss 4.6646 (3.9160) grad_norm 1.0471 (inf) [2021-04-15 18:56:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][870/1251] eta 0:01:49 lr 0.000929 time 0.2677 (0.2887) loss 3.5674 (3.9147) grad_norm 0.9405 (inf) [2021-04-15 18:56:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][880/1251] eta 0:01:47 lr 0.000929 time 0.2846 (0.2886) loss 3.9307 (3.9144) grad_norm 0.9698 (inf) [2021-04-15 18:56:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][890/1251] eta 0:01:44 lr 0.000929 time 0.2924 (0.2886) loss 3.8502 (3.9156) grad_norm 1.2145 (inf) [2021-04-15 18:56:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][900/1251] eta 0:01:41 lr 0.000929 time 0.2453 (0.2885) loss 3.5998 (3.9146) grad_norm 0.9816 (inf) [2021-04-15 18:56:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][910/1251] eta 0:01:38 lr 0.000929 time 0.2716 (0.2884) loss 4.0457 (3.9195) grad_norm 0.9843 (inf) [2021-04-15 18:56:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][920/1251] eta 0:01:35 lr 0.000929 time 0.2802 (0.2885) loss 4.0494 (3.9198) grad_norm 1.1185 (inf) [2021-04-15 18:56:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][930/1251] eta 0:01:32 lr 0.000929 time 0.2694 (0.2885) loss 4.8932 (3.9215) grad_norm 1.0131 (inf) [2021-04-15 18:56:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][940/1251] eta 0:01:29 lr 0.000929 time 0.2781 (0.2885) loss 2.9355 (3.9219) grad_norm 0.8464 (inf) [2021-04-15 18:56:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][950/1251] eta 0:01:26 lr 0.000929 time 0.2690 (0.2884) loss 4.2223 (3.9211) grad_norm 1.2993 (inf) [2021-04-15 18:56:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][960/1251] eta 0:01:23 lr 0.000929 time 0.2751 (0.2883) loss 4.1234 (3.9218) grad_norm 1.0011 (inf) [2021-04-15 18:56:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][970/1251] eta 0:01:21 lr 0.000929 time 0.2748 (0.2884) loss 4.5401 (3.9244) grad_norm 1.2016 (inf) [2021-04-15 18:56:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][980/1251] eta 0:01:18 lr 0.000929 time 0.2527 (0.2883) loss 4.4605 (3.9244) grad_norm 0.9959 (inf) [2021-04-15 18:56:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][990/1251] eta 0:01:15 lr 0.000929 time 0.2675 (0.2883) loss 4.4080 (3.9257) grad_norm 0.9974 (inf) [2021-04-15 18:56:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][1000/1251] eta 0:01:12 lr 0.000929 time 0.2914 (0.2882) loss 2.7775 (3.9248) grad_norm 1.1543 (inf) [2021-04-15 18:56:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][1010/1251] eta 0:01:09 lr 0.000929 time 0.2893 (0.2881) loss 4.4026 (3.9246) grad_norm 0.9181 (inf) [2021-04-15 18:57:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][1020/1251] eta 0:01:06 lr 0.000929 time 0.3108 (0.2881) loss 3.2234 (3.9249) grad_norm 1.0713 (inf) [2021-04-15 18:57:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][1030/1251] eta 0:01:03 lr 0.000929 time 0.2690 (0.2880) loss 4.9115 (3.9251) grad_norm 1.2866 (inf) [2021-04-15 18:57:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][1040/1251] eta 0:01:00 lr 0.000929 time 0.2785 (0.2879) loss 3.9974 (3.9246) grad_norm 1.3480 (inf) [2021-04-15 18:57:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][1050/1251] eta 0:00:57 lr 0.000929 time 0.2879 (0.2877) loss 3.3774 (3.9229) grad_norm 1.1163 (inf) [2021-04-15 18:57:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][1060/1251] eta 0:00:54 lr 0.000929 time 0.2909 (0.2877) loss 3.1858 (3.9237) grad_norm 0.9107 (inf) [2021-04-15 18:57:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][1070/1251] eta 0:00:52 lr 0.000929 time 0.3180 (0.2876) loss 3.7240 (3.9237) grad_norm 1.0406 (inf) [2021-04-15 18:57:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][1080/1251] eta 0:00:49 lr 0.000929 time 0.2858 (0.2875) loss 4.4970 (3.9236) grad_norm 1.1763 (inf) [2021-04-15 18:57:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][1090/1251] eta 0:00:46 lr 0.000929 time 0.2767 (0.2874) loss 4.1594 (3.9249) grad_norm 1.1568 (inf) [2021-04-15 18:57:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][1100/1251] eta 0:00:43 lr 0.000929 time 0.2764 (0.2873) loss 3.7275 (3.9240) grad_norm 1.0394 (inf) [2021-04-15 18:57:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][1110/1251] eta 0:00:40 lr 0.000929 time 0.2801 (0.2873) loss 5.1528 (3.9271) grad_norm 0.9815 (inf) [2021-04-15 18:57:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][1120/1251] eta 0:00:37 lr 0.000929 time 0.2886 (0.2873) loss 3.2638 (3.9264) grad_norm 1.0520 (inf) [2021-04-15 18:57:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][1130/1251] eta 0:00:34 lr 0.000929 time 0.2556 (0.2871) loss 3.0699 (3.9269) grad_norm 1.1715 (inf) [2021-04-15 18:57:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][1140/1251] eta 0:00:31 lr 0.000929 time 0.2797 (0.2871) loss 3.7560 (3.9279) grad_norm 1.2481 (inf) [2021-04-15 18:57:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][1150/1251] eta 0:00:28 lr 0.000929 time 0.3227 (0.2870) loss 4.3839 (3.9283) grad_norm 1.1851 (inf) [2021-04-15 18:57:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][1160/1251] eta 0:00:26 lr 0.000929 time 0.2774 (0.2871) loss 4.9486 (3.9326) grad_norm 1.1964 (inf) [2021-04-15 18:57:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][1170/1251] eta 0:00:23 lr 0.000929 time 0.2744 (0.2870) loss 4.5927 (3.9312) grad_norm 0.9726 (inf) [2021-04-15 18:57:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][1180/1251] eta 0:00:20 lr 0.000929 time 0.2530 (0.2869) loss 3.4141 (3.9315) grad_norm 1.4576 (inf) [2021-04-15 18:57:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][1190/1251] eta 0:00:17 lr 0.000929 time 0.3025 (0.2869) loss 4.3496 (3.9327) grad_norm 1.3401 (inf) [2021-04-15 18:57:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][1200/1251] eta 0:00:14 lr 0.000929 time 0.2785 (0.2868) loss 4.3676 (3.9321) grad_norm 1.0597 (inf) [2021-04-15 18:57:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][1210/1251] eta 0:00:11 lr 0.000928 time 0.2823 (0.2867) loss 4.2633 (3.9305) grad_norm 1.0159 (inf) [2021-04-15 18:57:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][1220/1251] eta 0:00:08 lr 0.000928 time 0.2932 (0.2866) loss 4.6020 (3.9320) grad_norm 1.2593 (inf) [2021-04-15 18:58:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][1230/1251] eta 0:00:06 lr 0.000928 time 0.3013 (0.2866) loss 3.3045 (3.9333) grad_norm 1.0360 (inf) [2021-04-15 18:58:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][1240/1251] eta 0:00:03 lr 0.000928 time 0.2490 (0.2864) loss 4.6186 (3.9334) grad_norm 0.9367 (inf) [2021-04-15 18:58:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [51/300][1250/1251] eta 0:00:00 lr 0.000928 time 0.2487 (0.2861) loss 3.8695 (3.9351) grad_norm 1.2022 (inf) [2021-04-15 18:58:07 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 51 training takes 0:06:00 [2021-04-15 18:58:07 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_51.pth saving...... [2021-04-15 18:58:17 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_51.pth saved !!! [2021-04-15 18:58:18 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.223 (1.223) Loss 1.2494 (1.2494) Acc@1 70.605 (70.605) Acc@5 91.504 (91.504) [2021-04-15 18:58:19 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.187 (0.238) Loss 1.3864 (1.2847) Acc@1 67.969 (70.188) Acc@5 88.574 (90.119) [2021-04-15 18:58:21 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.109 (0.223) Loss 1.2834 (1.2794) Acc@1 68.457 (70.247) Acc@5 90.820 (90.299) [2021-04-15 18:58:24 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.209 (0.248) Loss 1.2494 (1.2860) Acc@1 69.531 (70.180) Acc@5 89.648 (90.175) [2021-04-15 18:58:26 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.079 (0.222) Loss 1.3828 (1.2877) Acc@1 69.336 (70.222) Acc@5 88.086 (90.134) [2021-04-15 18:58:28 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 70.340 Acc@5 90.148 [2021-04-15 18:58:28 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 70.3% [2021-04-15 18:58:28 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 70.36% [2021-04-15 18:58:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][0/1251] eta 1:27:36 lr 0.000928 time 4.2015 (4.2015) loss 4.1055 (4.1055) grad_norm 1.2347 (1.2347) [2021-04-15 18:58:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][10/1251] eta 0:13:10 lr 0.000928 time 0.2959 (0.6373) loss 4.7965 (4.2026) grad_norm 1.1826 (1.1343) [2021-04-15 18:58:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][20/1251] eta 0:09:36 lr 0.000928 time 0.2946 (0.4683) loss 3.7667 (3.8448) grad_norm 1.1837 (1.1023) [2021-04-15 18:58:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][30/1251] eta 0:08:18 lr 0.000928 time 0.3021 (0.4080) loss 4.3919 (3.9339) grad_norm 1.0359 (1.1085) [2021-04-15 18:58:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][40/1251] eta 0:07:38 lr 0.000928 time 0.2714 (0.3783) loss 3.4808 (3.8949) grad_norm 1.0486 (1.1212) [2021-04-15 18:58:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][50/1251] eta 0:07:10 lr 0.000928 time 0.2765 (0.3583) loss 4.0769 (3.9510) grad_norm 1.0366 (1.1045) [2021-04-15 18:58:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][60/1251] eta 0:06:52 lr 0.000928 time 0.2932 (0.3462) loss 3.4704 (3.9273) grad_norm 0.9842 (1.0967) [2021-04-15 18:58:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][70/1251] eta 0:06:37 lr 0.000928 time 0.2718 (0.3369) loss 2.5356 (3.9120) grad_norm 1.0226 (1.1027) [2021-04-15 18:58:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][80/1251] eta 0:06:26 lr 0.000928 time 0.2861 (0.3299) loss 3.8382 (3.9548) grad_norm 1.2912 (1.1067) [2021-04-15 18:58:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][90/1251] eta 0:06:16 lr 0.000928 time 0.2801 (0.3247) loss 3.8107 (3.9438) grad_norm 1.0631 (1.1027) [2021-04-15 18:59:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][100/1251] eta 0:06:07 lr 0.000928 time 0.2647 (0.3196) loss 4.5659 (3.9327) grad_norm 0.9396 (1.1016) [2021-04-15 18:59:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][110/1251] eta 0:06:00 lr 0.000928 time 0.2889 (0.3156) loss 4.0129 (3.9332) grad_norm 1.2597 (1.1066) [2021-04-15 18:59:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][120/1251] eta 0:05:57 lr 0.000928 time 0.3979 (0.3157) loss 4.3073 (3.9385) grad_norm 1.0024 (1.1084) [2021-04-15 18:59:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][130/1251] eta 0:05:50 lr 0.000928 time 0.2620 (0.3126) loss 2.9967 (3.9423) grad_norm 1.1655 (1.1119) [2021-04-15 18:59:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][140/1251] eta 0:05:46 lr 0.000928 time 0.2449 (0.3114) loss 4.1408 (3.9432) grad_norm 1.1201 (1.1114) [2021-04-15 18:59:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][150/1251] eta 0:05:40 lr 0.000928 time 0.2584 (0.3093) loss 4.1756 (3.9323) grad_norm 0.9381 (1.1075) [2021-04-15 18:59:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][160/1251] eta 0:05:35 lr 0.000928 time 0.2637 (0.3075) loss 4.3932 (3.9245) grad_norm 0.9641 (1.1065) [2021-04-15 18:59:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][170/1251] eta 0:05:30 lr 0.000928 time 0.2817 (0.3058) loss 4.5109 (3.9380) grad_norm 0.9832 (1.1049) [2021-04-15 18:59:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][180/1251] eta 0:05:26 lr 0.000928 time 0.2809 (0.3050) loss 3.6362 (3.9359) grad_norm 1.1468 (1.1027) [2021-04-15 18:59:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][190/1251] eta 0:05:21 lr 0.000928 time 0.2490 (0.3034) loss 3.5400 (3.9291) grad_norm 1.0817 (1.0995) [2021-04-15 18:59:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][200/1251] eta 0:05:17 lr 0.000928 time 0.2625 (0.3025) loss 4.3470 (3.9272) grad_norm 1.2189 (1.0992) [2021-04-15 18:59:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][210/1251] eta 0:05:13 lr 0.000928 time 0.2570 (0.3015) loss 4.2040 (3.9247) grad_norm 1.0275 (1.1019) [2021-04-15 18:59:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][220/1251] eta 0:05:09 lr 0.000928 time 0.2763 (0.3006) loss 4.8185 (3.9250) grad_norm 1.0426 (1.1033) [2021-04-15 18:59:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][230/1251] eta 0:05:05 lr 0.000928 time 0.2584 (0.2997) loss 3.2499 (3.9210) grad_norm 0.9509 (1.1041) [2021-04-15 18:59:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][240/1251] eta 0:05:02 lr 0.000928 time 0.2744 (0.2987) loss 3.9175 (3.9175) grad_norm 1.0939 (1.0992) [2021-04-15 18:59:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][250/1251] eta 0:04:58 lr 0.000928 time 0.2732 (0.2978) loss 3.0011 (3.9219) grad_norm 1.0022 (1.0976) [2021-04-15 18:59:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][260/1251] eta 0:04:54 lr 0.000928 time 0.2879 (0.2973) loss 4.4515 (3.9224) grad_norm 1.0517 (1.0991) [2021-04-15 18:59:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][270/1251] eta 0:04:51 lr 0.000928 time 0.2981 (0.2969) loss 2.5253 (3.9185) grad_norm 1.3341 (1.0993) [2021-04-15 18:59:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][280/1251] eta 0:04:47 lr 0.000928 time 0.2604 (0.2963) loss 4.4351 (3.9253) grad_norm 1.2648 (1.1018) [2021-04-15 18:59:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][290/1251] eta 0:04:44 lr 0.000928 time 0.2695 (0.2960) loss 4.4518 (3.9320) grad_norm 1.4462 (1.1052) [2021-04-15 18:59:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][300/1251] eta 0:04:41 lr 0.000928 time 0.2855 (0.2955) loss 3.3638 (3.9301) grad_norm 1.0776 (1.1058) [2021-04-15 19:00:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][310/1251] eta 0:04:37 lr 0.000928 time 0.2431 (0.2949) loss 3.8954 (3.9346) grad_norm 1.1539 (1.1089) [2021-04-15 19:00:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][320/1251] eta 0:04:34 lr 0.000928 time 0.2884 (0.2945) loss 4.1237 (3.9323) grad_norm 1.6260 (1.1129) [2021-04-15 19:00:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][330/1251] eta 0:04:30 lr 0.000928 time 0.2476 (0.2940) loss 3.1134 (3.9263) grad_norm 0.9503 (1.1116) [2021-04-15 19:00:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][340/1251] eta 0:04:27 lr 0.000928 time 0.2793 (0.2936) loss 4.1396 (3.9239) grad_norm 1.0046 (1.1120) [2021-04-15 19:00:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][350/1251] eta 0:04:24 lr 0.000928 time 0.2924 (0.2932) loss 4.5371 (3.9304) grad_norm 1.2822 (1.1123) [2021-04-15 19:00:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][360/1251] eta 0:04:21 lr 0.000928 time 0.2554 (0.2937) loss 4.0173 (3.9336) grad_norm 1.2049 (1.1147) [2021-04-15 19:00:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][370/1251] eta 0:04:18 lr 0.000928 time 0.2699 (0.2931) loss 4.3633 (3.9269) grad_norm 1.1544 (1.1144) [2021-04-15 19:00:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][380/1251] eta 0:04:15 lr 0.000928 time 0.2869 (0.2928) loss 4.2346 (3.9239) grad_norm 1.1509 (1.1135) [2021-04-15 19:00:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][390/1251] eta 0:04:11 lr 0.000928 time 0.2708 (0.2925) loss 4.1285 (3.9247) grad_norm 1.0867 (1.1129) [2021-04-15 19:00:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][400/1251] eta 0:04:08 lr 0.000928 time 0.2543 (0.2921) loss 3.8124 (3.9169) grad_norm 0.8293 (1.1097) [2021-04-15 19:00:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][410/1251] eta 0:04:05 lr 0.000928 time 0.2955 (0.2918) loss 3.5284 (3.9165) grad_norm 1.0890 (1.1089) [2021-04-15 19:00:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][420/1251] eta 0:04:02 lr 0.000928 time 0.2560 (0.2915) loss 3.4163 (3.9103) grad_norm 1.2932 (1.1091) [2021-04-15 19:00:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][430/1251] eta 0:03:59 lr 0.000927 time 0.2747 (0.2912) loss 4.5000 (3.9127) grad_norm 1.0280 (1.1093) [2021-04-15 19:00:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][440/1251] eta 0:03:55 lr 0.000927 time 0.3018 (0.2909) loss 2.9057 (3.9112) grad_norm 1.0076 (1.1084) [2021-04-15 19:00:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][450/1251] eta 0:03:52 lr 0.000927 time 0.2595 (0.2905) loss 3.8058 (3.9083) grad_norm 1.2525 (1.1083) [2021-04-15 19:00:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][460/1251] eta 0:03:49 lr 0.000927 time 0.2733 (0.2904) loss 3.2941 (3.9006) grad_norm 0.9714 (1.1061) [2021-04-15 19:00:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][470/1251] eta 0:03:46 lr 0.000927 time 0.2893 (0.2902) loss 2.7493 (3.8972) grad_norm 1.0317 (1.1053) [2021-04-15 19:00:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][480/1251] eta 0:03:43 lr 0.000927 time 0.2829 (0.2900) loss 4.5140 (3.9055) grad_norm 1.0711 (1.1053) [2021-04-15 19:00:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][490/1251] eta 0:03:40 lr 0.000927 time 0.2473 (0.2902) loss 3.7323 (3.9112) grad_norm 1.1040 (1.1056) [2021-04-15 19:00:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][500/1251] eta 0:03:37 lr 0.000927 time 0.2722 (0.2900) loss 3.2206 (3.9154) grad_norm 1.0946 (1.1057) [2021-04-15 19:00:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][510/1251] eta 0:03:34 lr 0.000927 time 0.2708 (0.2898) loss 2.8212 (3.9234) grad_norm 1.0996 (1.1047) [2021-04-15 19:00:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][520/1251] eta 0:03:31 lr 0.000927 time 0.2895 (0.2899) loss 4.5648 (3.9230) grad_norm 1.2722 (1.1063) [2021-04-15 19:01:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][530/1251] eta 0:03:28 lr 0.000927 time 0.2877 (0.2897) loss 4.2777 (3.9288) grad_norm 0.9986 (1.1059) [2021-04-15 19:01:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][540/1251] eta 0:03:26 lr 0.000927 time 0.2660 (0.2898) loss 4.0519 (3.9333) grad_norm 1.2741 (1.1066) [2021-04-15 19:01:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][550/1251] eta 0:03:23 lr 0.000927 time 0.2642 (0.2896) loss 4.2336 (3.9341) grad_norm 1.1023 (1.1079) [2021-04-15 19:01:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][560/1251] eta 0:03:20 lr 0.000927 time 0.2817 (0.2894) loss 4.0290 (3.9329) grad_norm 1.0114 (1.1068) [2021-04-15 19:01:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][570/1251] eta 0:03:16 lr 0.000927 time 0.2751 (0.2893) loss 4.4588 (3.9329) grad_norm 0.9525 (1.1061) [2021-04-15 19:01:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][580/1251] eta 0:03:14 lr 0.000927 time 0.2763 (0.2893) loss 4.4283 (3.9318) grad_norm 1.1236 (1.1065) [2021-04-15 19:01:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][590/1251] eta 0:03:11 lr 0.000927 time 0.2768 (0.2892) loss 4.1154 (3.9299) grad_norm 1.0749 (1.1070) [2021-04-15 19:01:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][600/1251] eta 0:03:08 lr 0.000927 time 0.2821 (0.2890) loss 4.7941 (3.9339) grad_norm 1.1094 (1.1072) [2021-04-15 19:01:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][610/1251] eta 0:03:05 lr 0.000927 time 0.2730 (0.2888) loss 3.0921 (3.9307) grad_norm 1.0243 (1.1064) [2021-04-15 19:01:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][620/1251] eta 0:03:02 lr 0.000927 time 0.3059 (0.2887) loss 3.8033 (3.9295) grad_norm 0.9252 (1.1061) [2021-04-15 19:01:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][630/1251] eta 0:02:59 lr 0.000927 time 0.3031 (0.2887) loss 3.9472 (3.9236) grad_norm 1.3166 (1.1085) [2021-04-15 19:01:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][640/1251] eta 0:02:56 lr 0.000927 time 0.2894 (0.2885) loss 4.1277 (3.9273) grad_norm 0.9694 (1.1075) [2021-04-15 19:01:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][650/1251] eta 0:02:53 lr 0.000927 time 0.2660 (0.2884) loss 4.3078 (3.9302) grad_norm 1.0930 (1.1080) [2021-04-15 19:01:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][660/1251] eta 0:02:50 lr 0.000927 time 0.2917 (0.2883) loss 3.7034 (3.9333) grad_norm 1.2879 (1.1086) [2021-04-15 19:01:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][670/1251] eta 0:02:47 lr 0.000927 time 0.2943 (0.2882) loss 4.1613 (3.9341) grad_norm 1.1890 (1.1083) [2021-04-15 19:01:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][680/1251] eta 0:02:44 lr 0.000927 time 0.2820 (0.2880) loss 4.3749 (3.9375) grad_norm 1.0027 (1.1073) [2021-04-15 19:01:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][690/1251] eta 0:02:41 lr 0.000927 time 0.2632 (0.2880) loss 4.5932 (3.9356) grad_norm 1.0569 (1.1063) [2021-04-15 19:01:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][700/1251] eta 0:02:38 lr 0.000927 time 0.2544 (0.2878) loss 2.7381 (3.9316) grad_norm 1.1682 (1.1058) [2021-04-15 19:01:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][710/1251] eta 0:02:35 lr 0.000927 time 0.2630 (0.2877) loss 4.5474 (3.9322) grad_norm 1.1773 (1.1062) [2021-04-15 19:01:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][720/1251] eta 0:02:32 lr 0.000927 time 0.2879 (0.2877) loss 4.8289 (3.9371) grad_norm 1.0739 (1.1065) [2021-04-15 19:01:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][730/1251] eta 0:02:29 lr 0.000927 time 0.2734 (0.2877) loss 4.0219 (3.9391) grad_norm 0.9658 (1.1059) [2021-04-15 19:02:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][740/1251] eta 0:02:26 lr 0.000927 time 0.2574 (0.2876) loss 4.1267 (3.9428) grad_norm 1.1000 (1.1050) [2021-04-15 19:02:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][750/1251] eta 0:02:24 lr 0.000927 time 0.2862 (0.2874) loss 3.8988 (3.9438) grad_norm 1.1397 (1.1057) [2021-04-15 19:02:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][760/1251] eta 0:02:21 lr 0.000927 time 0.3116 (0.2875) loss 4.2947 (3.9448) grad_norm 0.9393 (1.1059) [2021-04-15 19:02:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][770/1251] eta 0:02:18 lr 0.000927 time 0.2859 (0.2874) loss 4.3248 (3.9428) grad_norm 1.1297 (1.1062) [2021-04-15 19:02:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][780/1251] eta 0:02:15 lr 0.000927 time 0.2880 (0.2874) loss 3.5352 (3.9387) grad_norm 1.1775 (1.1080) [2021-04-15 19:02:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][790/1251] eta 0:02:12 lr 0.000927 time 0.2747 (0.2873) loss 4.3072 (3.9385) grad_norm 1.3305 (1.1088) [2021-04-15 19:02:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][800/1251] eta 0:02:09 lr 0.000927 time 0.2863 (0.2872) loss 3.8357 (3.9404) grad_norm 1.1690 (1.1086) [2021-04-15 19:02:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][810/1251] eta 0:02:06 lr 0.000927 time 0.2700 (0.2871) loss 4.6430 (3.9385) grad_norm 0.9945 (1.1083) [2021-04-15 19:02:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][820/1251] eta 0:02:03 lr 0.000927 time 0.2635 (0.2870) loss 4.4963 (3.9404) grad_norm 0.9688 (1.1073) [2021-04-15 19:02:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][830/1251] eta 0:02:00 lr 0.000927 time 0.2897 (0.2870) loss 2.7775 (3.9405) grad_norm 1.1441 (1.1065) [2021-04-15 19:02:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][840/1251] eta 0:01:57 lr 0.000927 time 0.2991 (0.2869) loss 3.9923 (3.9349) grad_norm 1.2772 (1.1077) [2021-04-15 19:02:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][850/1251] eta 0:01:54 lr 0.000927 time 0.2769 (0.2867) loss 3.9538 (3.9364) grad_norm 0.9956 (1.1090) [2021-04-15 19:02:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][860/1251] eta 0:01:52 lr 0.000927 time 0.2772 (0.2867) loss 3.4976 (3.9354) grad_norm 1.0490 (1.1092) [2021-04-15 19:02:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][870/1251] eta 0:01:49 lr 0.000927 time 0.2515 (0.2866) loss 3.1951 (3.9376) grad_norm 0.8987 (1.1088) [2021-04-15 19:02:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][880/1251] eta 0:01:46 lr 0.000927 time 0.2781 (0.2866) loss 3.8049 (3.9377) grad_norm 1.2383 (1.1091) [2021-04-15 19:02:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][890/1251] eta 0:01:43 lr 0.000926 time 0.2960 (0.2865) loss 3.6927 (3.9385) grad_norm 1.1044 (1.1086) [2021-04-15 19:02:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][900/1251] eta 0:01:40 lr 0.000926 time 0.2615 (0.2864) loss 4.0069 (3.9395) grad_norm 0.9205 (1.1084) [2021-04-15 19:02:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][910/1251] eta 0:01:37 lr 0.000926 time 0.2733 (0.2862) loss 4.2809 (3.9351) grad_norm 1.0579 (1.1074) [2021-04-15 19:02:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][920/1251] eta 0:01:34 lr 0.000926 time 0.3045 (0.2862) loss 3.6198 (3.9345) grad_norm 1.0618 (1.1072) [2021-04-15 19:02:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][930/1251] eta 0:01:31 lr 0.000926 time 0.2594 (0.2863) loss 4.2719 (3.9354) grad_norm 1.2569 (1.1071) [2021-04-15 19:02:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][940/1251] eta 0:01:29 lr 0.000926 time 0.2731 (0.2863) loss 4.7194 (3.9384) grad_norm 1.1132 (1.1067) [2021-04-15 19:03:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][950/1251] eta 0:01:26 lr 0.000926 time 0.2902 (0.2863) loss 3.8978 (3.9391) grad_norm 1.0715 (1.1072) [2021-04-15 19:03:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][960/1251] eta 0:01:23 lr 0.000926 time 0.2769 (0.2862) loss 3.7909 (3.9388) grad_norm 1.0008 (1.1070) [2021-04-15 19:03:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][970/1251] eta 0:01:20 lr 0.000926 time 0.2645 (0.2861) loss 4.7489 (3.9398) grad_norm 1.0593 (1.1063) [2021-04-15 19:03:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][980/1251] eta 0:01:17 lr 0.000926 time 0.2629 (0.2861) loss 4.2927 (3.9401) grad_norm 1.2678 (1.1066) [2021-04-15 19:03:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][990/1251] eta 0:01:14 lr 0.000926 time 0.2896 (0.2860) loss 3.6675 (3.9417) grad_norm 1.1156 (1.1072) [2021-04-15 19:03:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][1000/1251] eta 0:01:11 lr 0.000926 time 0.3033 (0.2860) loss 4.1370 (3.9426) grad_norm 1.3695 (1.1076) [2021-04-15 19:03:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][1010/1251] eta 0:01:08 lr 0.000926 time 0.2968 (0.2859) loss 3.9940 (3.9404) grad_norm 1.2099 (1.1074) [2021-04-15 19:03:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][1020/1251] eta 0:01:06 lr 0.000926 time 0.2784 (0.2858) loss 4.6265 (3.9421) grad_norm 1.0650 (1.1077) [2021-04-15 19:03:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][1030/1251] eta 0:01:03 lr 0.000926 time 0.3097 (0.2858) loss 3.4259 (3.9413) grad_norm 1.1117 (1.1074) [2021-04-15 19:03:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][1040/1251] eta 0:01:00 lr 0.000926 time 0.2577 (0.2857) loss 4.4178 (3.9417) grad_norm 1.0457 (1.1070) [2021-04-15 19:03:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][1050/1251] eta 0:00:57 lr 0.000926 time 0.2601 (0.2856) loss 4.5051 (3.9396) grad_norm 1.2603 (1.1077) [2021-04-15 19:03:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][1060/1251] eta 0:00:54 lr 0.000926 time 0.3041 (0.2856) loss 3.2482 (3.9394) grad_norm 1.0197 (1.1097) [2021-04-15 19:03:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][1070/1251] eta 0:00:51 lr 0.000926 time 0.2497 (0.2856) loss 3.7916 (3.9429) grad_norm 1.2300 (1.1093) [2021-04-15 19:03:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][1080/1251] eta 0:00:48 lr 0.000926 time 0.2709 (0.2855) loss 4.9172 (3.9444) grad_norm 1.0686 (1.1095) [2021-04-15 19:03:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][1090/1251] eta 0:00:45 lr 0.000926 time 0.2830 (0.2856) loss 3.0791 (3.9423) grad_norm 1.0678 (1.1095) [2021-04-15 19:03:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][1100/1251] eta 0:00:43 lr 0.000926 time 0.2686 (0.2855) loss 4.6620 (3.9424) grad_norm 1.0395 (1.1089) [2021-04-15 19:03:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][1110/1251] eta 0:00:40 lr 0.000926 time 0.2752 (0.2855) loss 4.6038 (3.9433) grad_norm 1.1832 (1.1083) [2021-04-15 19:03:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][1120/1251] eta 0:00:37 lr 0.000926 time 0.2535 (0.2855) loss 4.1263 (3.9427) grad_norm 1.0768 (1.1090) [2021-04-15 19:03:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][1130/1251] eta 0:00:34 lr 0.000926 time 0.2804 (0.2855) loss 4.1822 (3.9450) grad_norm 1.0375 (1.1092) [2021-04-15 19:03:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][1140/1251] eta 0:00:31 lr 0.000926 time 0.2554 (0.2855) loss 3.6417 (3.9465) grad_norm 1.0016 (1.1082) [2021-04-15 19:03:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][1150/1251] eta 0:00:28 lr 0.000926 time 0.2751 (0.2855) loss 4.6934 (3.9483) grad_norm 1.0492 (1.1078) [2021-04-15 19:04:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][1160/1251] eta 0:00:25 lr 0.000926 time 0.2656 (0.2854) loss 4.7104 (3.9482) grad_norm 1.1863 (1.1074) [2021-04-15 19:04:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][1170/1251] eta 0:00:23 lr 0.000926 time 0.2663 (0.2854) loss 3.7886 (3.9455) grad_norm 1.0298 (1.1071) [2021-04-15 19:04:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][1180/1251] eta 0:00:20 lr 0.000926 time 0.2724 (0.2853) loss 4.7096 (3.9475) grad_norm 1.0567 (1.1074) [2021-04-15 19:04:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][1190/1251] eta 0:00:17 lr 0.000926 time 0.2874 (0.2853) loss 4.0217 (3.9503) grad_norm 1.0254 (1.1073) [2021-04-15 19:04:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][1200/1251] eta 0:00:14 lr 0.000926 time 0.2656 (0.2852) loss 4.7297 (3.9504) grad_norm 1.1377 (1.1075) [2021-04-15 19:04:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][1210/1251] eta 0:00:11 lr 0.000926 time 0.2747 (0.2851) loss 3.4257 (3.9497) grad_norm 1.0657 (1.1072) [2021-04-15 19:04:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][1220/1251] eta 0:00:08 lr 0.000926 time 0.2652 (0.2850) loss 3.7980 (3.9494) grad_norm 0.9901 (1.1065) [2021-04-15 19:04:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][1230/1251] eta 0:00:05 lr 0.000926 time 0.2861 (0.2850) loss 4.6488 (3.9478) grad_norm 1.1723 (1.1061) [2021-04-15 19:04:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][1240/1251] eta 0:00:03 lr 0.000926 time 0.2479 (0.2849) loss 3.9911 (3.9491) grad_norm 0.9719 (1.1063) [2021-04-15 19:04:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [52/300][1250/1251] eta 0:00:00 lr 0.000926 time 0.2494 (0.2846) loss 4.3302 (3.9506) grad_norm 1.1956 (1.1065) [2021-04-15 19:04:27 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 52 training takes 0:05:58 [2021-04-15 19:04:27 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_52.pth saving...... [2021-04-15 19:04:42 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_52.pth saved !!! [2021-04-15 19:04:43 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.238 (1.238) Loss 1.2553 (1.2553) Acc@1 72.559 (72.559) Acc@5 91.113 (91.113) [2021-04-15 19:04:45 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.106 (0.213) Loss 1.3133 (1.2932) Acc@1 70.703 (70.463) Acc@5 90.137 (90.225) [2021-04-15 19:04:47 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.128 (0.250) Loss 1.1943 (1.2958) Acc@1 70.801 (70.317) Acc@5 92.090 (90.141) [2021-04-15 19:04:50 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.332 (0.257) Loss 1.3652 (1.2994) Acc@1 69.336 (70.369) Acc@5 88.965 (90.099) [2021-04-15 19:04:51 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.223) Loss 1.3679 (1.2968) Acc@1 69.922 (70.389) Acc@5 89.746 (90.180) [2021-04-15 19:04:54 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 70.424 Acc@5 90.178 [2021-04-15 19:04:54 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 70.4% [2021-04-15 19:04:54 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 70.42% [2021-04-15 19:04:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][0/1251] eta 1:32:24 lr 0.000926 time 4.4324 (4.4324) loss 4.0055 (4.0055) grad_norm 1.1245 (1.1245) [2021-04-15 19:05:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][10/1251] eta 0:13:32 lr 0.000926 time 0.2874 (0.6549) loss 3.0315 (3.8719) grad_norm 0.9078 (1.0785) [2021-04-15 19:05:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][20/1251] eta 0:09:46 lr 0.000926 time 0.2657 (0.4762) loss 3.9850 (3.9210) grad_norm 1.2376 (1.0968) [2021-04-15 19:05:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][30/1251] eta 0:08:24 lr 0.000926 time 0.2646 (0.4128) loss 3.9392 (3.9103) grad_norm 1.0434 (1.0976) [2021-04-15 19:05:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][40/1251] eta 0:07:42 lr 0.000926 time 0.3045 (0.3815) loss 4.0101 (3.8751) grad_norm 1.0501 (1.0818) [2021-04-15 19:05:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][50/1251] eta 0:07:13 lr 0.000926 time 0.2927 (0.3613) loss 4.0654 (3.8649) grad_norm 1.1688 (1.0792) [2021-04-15 19:05:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][60/1251] eta 0:06:54 lr 0.000926 time 0.2707 (0.3480) loss 4.4589 (3.9013) grad_norm 1.3760 (1.0813) [2021-04-15 19:05:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][70/1251] eta 0:06:40 lr 0.000926 time 0.2981 (0.3389) loss 4.3641 (3.9252) grad_norm 1.3131 (1.0859) [2021-04-15 19:05:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][80/1251] eta 0:06:28 lr 0.000926 time 0.2779 (0.3315) loss 3.9109 (3.8941) grad_norm 1.0996 (1.1096) [2021-04-15 19:05:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][90/1251] eta 0:06:18 lr 0.000926 time 0.2870 (0.3257) loss 2.5795 (3.8772) grad_norm 1.0833 (1.1186) [2021-04-15 19:05:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][100/1251] eta 0:06:09 lr 0.000925 time 0.2779 (0.3211) loss 2.9224 (3.8741) grad_norm 0.9886 (1.1186) [2021-04-15 19:05:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][110/1251] eta 0:06:02 lr 0.000925 time 0.2864 (0.3174) loss 4.2979 (3.8981) grad_norm 1.1738 (1.1160) [2021-04-15 19:05:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][120/1251] eta 0:05:55 lr 0.000925 time 0.2847 (0.3144) loss 4.0211 (3.9040) grad_norm 0.9618 (1.1132) [2021-04-15 19:05:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][130/1251] eta 0:05:50 lr 0.000925 time 0.2449 (0.3124) loss 3.4663 (3.9087) grad_norm 0.9352 (1.1081) [2021-04-15 19:05:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][140/1251] eta 0:05:46 lr 0.000925 time 0.2711 (0.3116) loss 4.3840 (3.9119) grad_norm 1.0759 (1.1088) [2021-04-15 19:05:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][150/1251] eta 0:05:42 lr 0.000925 time 0.2887 (0.3108) loss 3.8902 (3.9134) grad_norm 1.0209 (1.1092) [2021-04-15 19:05:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][160/1251] eta 0:05:37 lr 0.000925 time 0.2704 (0.3094) loss 3.7649 (3.9164) grad_norm 1.2488 (1.1134) [2021-04-15 19:05:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][170/1251] eta 0:05:32 lr 0.000925 time 0.2663 (0.3079) loss 4.5258 (3.9234) grad_norm 1.0878 (1.1136) [2021-04-15 19:05:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][180/1251] eta 0:05:28 lr 0.000925 time 0.2718 (0.3065) loss 4.2326 (3.9428) grad_norm 0.9940 (1.1085) [2021-04-15 19:05:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][190/1251] eta 0:05:23 lr 0.000925 time 0.2935 (0.3052) loss 3.2627 (3.9395) grad_norm 1.0704 (1.1082) [2021-04-15 19:05:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][200/1251] eta 0:05:19 lr 0.000925 time 0.3037 (0.3039) loss 4.4174 (3.9382) grad_norm 1.0216 (1.1080) [2021-04-15 19:05:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][210/1251] eta 0:05:15 lr 0.000925 time 0.2852 (0.3027) loss 4.0881 (3.9500) grad_norm 1.1106 (1.1116) [2021-04-15 19:06:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][220/1251] eta 0:05:11 lr 0.000925 time 0.2948 (0.3017) loss 4.1770 (3.9600) grad_norm 1.3223 (1.1108) [2021-04-15 19:06:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][230/1251] eta 0:05:06 lr 0.000925 time 0.3047 (0.3007) loss 4.3096 (3.9569) grad_norm 0.9540 (1.1094) [2021-04-15 19:06:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][240/1251] eta 0:05:03 lr 0.000925 time 0.2761 (0.2997) loss 3.6999 (3.9597) grad_norm 1.0799 (1.1147) [2021-04-15 19:06:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][250/1251] eta 0:04:59 lr 0.000925 time 0.2918 (0.2991) loss 4.7442 (3.9651) grad_norm 1.3511 (1.1150) [2021-04-15 19:06:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][260/1251] eta 0:04:55 lr 0.000925 time 0.2839 (0.2982) loss 4.2974 (3.9734) grad_norm 1.0783 (1.1135) [2021-04-15 19:06:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][270/1251] eta 0:04:51 lr 0.000925 time 0.2606 (0.2976) loss 3.7189 (3.9620) grad_norm 1.2289 (1.1186) [2021-04-15 19:06:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][280/1251] eta 0:04:48 lr 0.000925 time 0.2671 (0.2970) loss 4.3802 (3.9532) grad_norm 1.2760 (1.1226) [2021-04-15 19:06:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][290/1251] eta 0:04:45 lr 0.000925 time 0.2701 (0.2967) loss 3.7385 (3.9585) grad_norm 1.1304 (1.1237) [2021-04-15 19:06:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][300/1251] eta 0:04:42 lr 0.000925 time 0.2716 (0.2967) loss 4.1520 (3.9558) grad_norm 1.1409 (1.1238) [2021-04-15 19:06:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][310/1251] eta 0:04:38 lr 0.000925 time 0.2844 (0.2962) loss 4.7247 (3.9555) grad_norm 0.9563 (1.1240) [2021-04-15 19:06:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][320/1251] eta 0:04:35 lr 0.000925 time 0.2654 (0.2961) loss 3.0885 (3.9502) grad_norm 0.9878 (1.1231) [2021-04-15 19:06:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][330/1251] eta 0:04:32 lr 0.000925 time 0.2750 (0.2955) loss 4.5298 (3.9579) grad_norm 1.1341 (1.1228) [2021-04-15 19:06:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][340/1251] eta 0:04:28 lr 0.000925 time 0.3035 (0.2951) loss 4.6623 (3.9603) grad_norm 0.9575 (1.1208) [2021-04-15 19:06:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][350/1251] eta 0:04:25 lr 0.000925 time 0.2733 (0.2950) loss 2.8154 (3.9577) grad_norm 1.2250 (1.1189) [2021-04-15 19:06:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][360/1251] eta 0:04:22 lr 0.000925 time 0.2602 (0.2946) loss 4.1872 (3.9612) grad_norm 1.2767 (1.1191) [2021-04-15 19:06:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][370/1251] eta 0:04:19 lr 0.000925 time 0.2659 (0.2940) loss 3.1039 (3.9623) grad_norm 1.0885 (1.1173) [2021-04-15 19:06:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][380/1251] eta 0:04:15 lr 0.000925 time 0.2986 (0.2937) loss 4.2708 (3.9625) grad_norm 1.0901 (1.1157) [2021-04-15 19:06:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][390/1251] eta 0:04:12 lr 0.000925 time 0.2727 (0.2932) loss 4.4962 (3.9575) grad_norm 1.0099 (1.1156) [2021-04-15 19:06:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][400/1251] eta 0:04:09 lr 0.000925 time 0.2749 (0.2928) loss 4.6354 (3.9580) grad_norm 0.8848 (1.1142) [2021-04-15 19:06:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][410/1251] eta 0:04:06 lr 0.000925 time 0.2631 (0.2926) loss 4.3473 (3.9565) grad_norm 1.1971 (1.1132) [2021-04-15 19:06:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][420/1251] eta 0:04:02 lr 0.000925 time 0.2947 (0.2923) loss 5.0905 (3.9579) grad_norm 1.2127 (1.1124) [2021-04-15 19:07:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][430/1251] eta 0:03:59 lr 0.000925 time 0.2438 (0.2919) loss 2.5309 (3.9507) grad_norm 1.1056 (1.1126) [2021-04-15 19:07:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][440/1251] eta 0:03:56 lr 0.000925 time 0.2856 (0.2918) loss 4.3243 (3.9458) grad_norm 1.2128 (1.1111) [2021-04-15 19:07:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][450/1251] eta 0:03:53 lr 0.000925 time 0.2852 (0.2913) loss 4.1910 (3.9494) grad_norm 1.1224 (1.1107) [2021-04-15 19:07:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][460/1251] eta 0:03:50 lr 0.000925 time 0.3020 (0.2911) loss 3.3540 (3.9496) grad_norm 1.1499 (1.1113) [2021-04-15 19:07:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][470/1251] eta 0:03:47 lr 0.000925 time 0.2567 (0.2908) loss 3.2200 (3.9532) grad_norm 1.1393 (1.1111) [2021-04-15 19:07:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][480/1251] eta 0:03:44 lr 0.000925 time 0.2796 (0.2906) loss 3.7415 (3.9565) grad_norm 1.0286 (1.1118) [2021-04-15 19:07:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][490/1251] eta 0:03:40 lr 0.000925 time 0.3129 (0.2903) loss 4.3247 (3.9539) grad_norm 0.8799 (1.1104) [2021-04-15 19:07:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][500/1251] eta 0:03:37 lr 0.000925 time 0.2713 (0.2900) loss 4.4937 (3.9575) grad_norm 1.2258 (1.1109) [2021-04-15 19:07:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][510/1251] eta 0:03:34 lr 0.000925 time 0.2502 (0.2897) loss 3.9650 (3.9571) grad_norm inf (inf) [2021-04-15 19:07:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][520/1251] eta 0:03:31 lr 0.000925 time 0.2633 (0.2899) loss 4.3597 (3.9564) grad_norm 1.1226 (inf) [2021-04-15 19:07:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][530/1251] eta 0:03:28 lr 0.000925 time 0.2601 (0.2897) loss 4.8597 (3.9556) grad_norm 1.2651 (inf) [2021-04-15 19:07:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][540/1251] eta 0:03:25 lr 0.000925 time 0.2562 (0.2897) loss 3.3782 (3.9522) grad_norm 1.0186 (inf) [2021-04-15 19:07:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][550/1251] eta 0:03:22 lr 0.000924 time 0.2708 (0.2895) loss 3.4148 (3.9508) grad_norm 1.0144 (inf) [2021-04-15 19:07:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][560/1251] eta 0:03:19 lr 0.000924 time 0.2634 (0.2892) loss 2.5402 (3.9409) grad_norm 1.0699 (inf) [2021-04-15 19:07:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][570/1251] eta 0:03:16 lr 0.000924 time 0.2665 (0.2891) loss 3.9429 (3.9410) grad_norm 1.4815 (inf) [2021-04-15 19:07:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][580/1251] eta 0:03:14 lr 0.000924 time 0.2716 (0.2894) loss 4.6499 (3.9385) grad_norm 1.1518 (inf) [2021-04-15 19:07:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][590/1251] eta 0:03:11 lr 0.000924 time 0.2996 (0.2892) loss 3.9456 (3.9393) grad_norm 1.0338 (inf) [2021-04-15 19:07:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][600/1251] eta 0:03:08 lr 0.000924 time 0.2822 (0.2892) loss 4.4749 (3.9371) grad_norm 1.3011 (inf) [2021-04-15 19:07:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][610/1251] eta 0:03:05 lr 0.000924 time 0.2795 (0.2890) loss 3.9227 (3.9345) grad_norm 1.0515 (inf) [2021-04-15 19:07:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][620/1251] eta 0:03:02 lr 0.000924 time 0.2903 (0.2889) loss 4.5096 (3.9387) grad_norm 1.1876 (inf) [2021-04-15 19:07:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][630/1251] eta 0:02:59 lr 0.000924 time 0.3064 (0.2888) loss 4.1970 (3.9399) grad_norm 0.9132 (inf) [2021-04-15 19:07:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][640/1251] eta 0:02:56 lr 0.000924 time 0.2785 (0.2887) loss 3.1706 (3.9343) grad_norm 0.9729 (inf) [2021-04-15 19:08:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][650/1251] eta 0:02:53 lr 0.000924 time 0.2497 (0.2885) loss 4.7182 (3.9380) grad_norm 0.9600 (inf) [2021-04-15 19:08:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][660/1251] eta 0:02:50 lr 0.000924 time 0.2724 (0.2883) loss 3.8391 (3.9317) grad_norm 1.0674 (inf) [2021-04-15 19:08:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][670/1251] eta 0:02:47 lr 0.000924 time 0.2931 (0.2882) loss 3.6264 (3.9328) grad_norm 0.8962 (inf) [2021-04-15 19:08:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][680/1251] eta 0:02:44 lr 0.000924 time 0.2835 (0.2881) loss 3.0488 (3.9346) grad_norm 1.1210 (inf) [2021-04-15 19:08:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][690/1251] eta 0:02:41 lr 0.000924 time 0.2748 (0.2880) loss 3.5014 (3.9362) grad_norm 1.2185 (inf) [2021-04-15 19:08:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][700/1251] eta 0:02:38 lr 0.000924 time 0.2621 (0.2879) loss 4.9432 (3.9413) grad_norm 1.3889 (inf) [2021-04-15 19:08:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][710/1251] eta 0:02:35 lr 0.000924 time 0.2844 (0.2880) loss 4.1948 (3.9422) grad_norm 0.9526 (inf) [2021-04-15 19:08:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][720/1251] eta 0:02:33 lr 0.000924 time 0.2617 (0.2883) loss 4.0389 (3.9401) grad_norm 1.0215 (inf) [2021-04-15 19:08:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][730/1251] eta 0:02:30 lr 0.000924 time 0.2982 (0.2881) loss 4.7148 (3.9391) grad_norm 1.2127 (inf) [2021-04-15 19:08:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][740/1251] eta 0:02:27 lr 0.000924 time 0.2667 (0.2880) loss 3.6070 (3.9424) grad_norm 1.3311 (inf) [2021-04-15 19:08:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][750/1251] eta 0:02:24 lr 0.000924 time 0.2941 (0.2879) loss 5.0967 (3.9480) grad_norm 1.2071 (inf) [2021-04-15 19:08:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][760/1251] eta 0:02:21 lr 0.000924 time 0.2611 (0.2877) loss 2.5368 (3.9430) grad_norm 1.2052 (inf) [2021-04-15 19:08:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][770/1251] eta 0:02:18 lr 0.000924 time 0.2897 (0.2876) loss 4.1036 (3.9402) grad_norm 1.0474 (inf) [2021-04-15 19:08:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][780/1251] eta 0:02:15 lr 0.000924 time 0.2813 (0.2874) loss 3.5551 (3.9350) grad_norm 0.9723 (inf) [2021-04-15 19:08:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][790/1251] eta 0:02:12 lr 0.000924 time 0.2776 (0.2873) loss 2.9618 (3.9342) grad_norm 1.0812 (inf) [2021-04-15 19:08:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][800/1251] eta 0:02:09 lr 0.000924 time 0.2976 (0.2872) loss 3.6276 (3.9377) grad_norm 1.1916 (inf) [2021-04-15 19:08:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][810/1251] eta 0:02:06 lr 0.000924 time 0.2848 (0.2871) loss 2.9112 (3.9355) grad_norm 1.2536 (inf) [2021-04-15 19:08:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][820/1251] eta 0:02:03 lr 0.000924 time 0.2806 (0.2870) loss 3.8838 (3.9337) grad_norm 1.3843 (inf) [2021-04-15 19:08:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][830/1251] eta 0:02:00 lr 0.000924 time 0.2621 (0.2869) loss 4.6035 (3.9359) grad_norm 1.0533 (inf) [2021-04-15 19:08:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][840/1251] eta 0:01:57 lr 0.000924 time 0.2694 (0.2868) loss 3.3572 (3.9378) grad_norm 1.1986 (inf) [2021-04-15 19:08:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][850/1251] eta 0:01:54 lr 0.000924 time 0.2992 (0.2867) loss 4.3089 (3.9379) grad_norm 1.0352 (inf) [2021-04-15 19:09:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][860/1251] eta 0:01:52 lr 0.000924 time 0.2643 (0.2866) loss 4.0400 (3.9411) grad_norm 1.0583 (inf) [2021-04-15 19:09:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][870/1251] eta 0:01:49 lr 0.000924 time 0.2741 (0.2866) loss 2.8007 (3.9421) grad_norm 1.1354 (inf) [2021-04-15 19:09:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][880/1251] eta 0:01:46 lr 0.000924 time 0.2861 (0.2867) loss 5.0917 (3.9420) grad_norm 1.0736 (inf) [2021-04-15 19:09:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][890/1251] eta 0:01:43 lr 0.000924 time 0.2762 (0.2868) loss 4.1151 (3.9401) grad_norm 1.0335 (inf) [2021-04-15 19:09:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][900/1251] eta 0:01:40 lr 0.000924 time 0.2851 (0.2867) loss 3.7775 (3.9447) grad_norm 1.0313 (inf) [2021-04-15 19:09:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][910/1251] eta 0:01:37 lr 0.000924 time 0.2933 (0.2866) loss 4.3898 (3.9471) grad_norm 1.2434 (inf) [2021-04-15 19:09:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][920/1251] eta 0:01:34 lr 0.000924 time 0.2981 (0.2865) loss 2.6687 (3.9447) grad_norm 1.1154 (inf) [2021-04-15 19:09:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][930/1251] eta 0:01:31 lr 0.000924 time 0.2841 (0.2865) loss 3.8726 (3.9428) grad_norm 0.9489 (inf) [2021-04-15 19:09:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][940/1251] eta 0:01:29 lr 0.000924 time 0.2767 (0.2864) loss 4.2090 (3.9419) grad_norm 1.1205 (inf) [2021-04-15 19:09:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][950/1251] eta 0:01:26 lr 0.000924 time 0.2671 (0.2863) loss 4.6237 (3.9431) grad_norm 1.1480 (inf) [2021-04-15 19:09:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][960/1251] eta 0:01:23 lr 0.000924 time 0.2903 (0.2862) loss 3.5457 (3.9422) grad_norm 1.3191 (inf) [2021-04-15 19:09:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][970/1251] eta 0:01:20 lr 0.000924 time 0.2838 (0.2861) loss 2.8369 (3.9438) grad_norm 0.9046 (inf) [2021-04-15 19:09:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][980/1251] eta 0:01:17 lr 0.000924 time 0.2776 (0.2860) loss 3.0264 (3.9420) grad_norm 1.0761 (inf) [2021-04-15 19:09:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][990/1251] eta 0:01:14 lr 0.000924 time 0.2882 (0.2859) loss 3.2567 (3.9403) grad_norm 1.1772 (inf) [2021-04-15 19:09:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][1000/1251] eta 0:01:11 lr 0.000923 time 0.2759 (0.2858) loss 4.5526 (3.9409) grad_norm 1.0822 (inf) [2021-04-15 19:09:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][1010/1251] eta 0:01:08 lr 0.000923 time 0.2990 (0.2858) loss 3.0383 (3.9374) grad_norm 1.0096 (inf) [2021-04-15 19:09:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][1020/1251] eta 0:01:05 lr 0.000923 time 0.2816 (0.2857) loss 4.4301 (3.9387) grad_norm 1.4215 (inf) [2021-04-15 19:09:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][1030/1251] eta 0:01:03 lr 0.000923 time 0.2526 (0.2855) loss 3.3146 (3.9386) grad_norm 1.1395 (inf) [2021-04-15 19:09:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][1040/1251] eta 0:01:00 lr 0.000923 time 0.2900 (0.2855) loss 3.8472 (3.9384) grad_norm 0.9838 (inf) [2021-04-15 19:09:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][1050/1251] eta 0:00:57 lr 0.000923 time 0.2654 (0.2854) loss 4.6454 (3.9393) grad_norm 1.1205 (inf) [2021-04-15 19:09:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][1060/1251] eta 0:00:54 lr 0.000923 time 0.2637 (0.2853) loss 3.5647 (3.9382) grad_norm 1.1826 (inf) [2021-04-15 19:10:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][1070/1251] eta 0:00:51 lr 0.000923 time 0.2602 (0.2853) loss 3.6363 (3.9378) grad_norm 1.1458 (inf) [2021-04-15 19:10:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][1080/1251] eta 0:00:48 lr 0.000923 time 0.2439 (0.2852) loss 3.7189 (3.9389) grad_norm 1.1145 (inf) [2021-04-15 19:10:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][1090/1251] eta 0:00:45 lr 0.000923 time 0.2623 (0.2853) loss 4.9580 (3.9409) grad_norm 1.0971 (inf) [2021-04-15 19:10:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][1100/1251] eta 0:00:43 lr 0.000923 time 0.2853 (0.2853) loss 2.5955 (3.9378) grad_norm 1.1719 (inf) [2021-04-15 19:10:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][1110/1251] eta 0:00:40 lr 0.000923 time 0.2767 (0.2852) loss 3.7323 (3.9399) grad_norm 1.3342 (inf) [2021-04-15 19:10:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][1120/1251] eta 0:00:37 lr 0.000923 time 0.2678 (0.2853) loss 4.2040 (3.9377) grad_norm 0.9555 (inf) [2021-04-15 19:10:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][1130/1251] eta 0:00:34 lr 0.000923 time 0.3028 (0.2853) loss 4.3490 (3.9371) grad_norm 1.0623 (inf) [2021-04-15 19:10:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][1140/1251] eta 0:00:31 lr 0.000923 time 0.2852 (0.2853) loss 3.9101 (3.9383) grad_norm 1.4724 (inf) [2021-04-15 19:10:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][1150/1251] eta 0:00:28 lr 0.000923 time 0.2653 (0.2853) loss 4.8904 (3.9405) grad_norm 1.1490 (inf) [2021-04-15 19:10:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][1160/1251] eta 0:00:25 lr 0.000923 time 0.3056 (0.2854) loss 3.5753 (3.9388) grad_norm 1.2441 (inf) [2021-04-15 19:10:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][1170/1251] eta 0:00:23 lr 0.000923 time 0.2797 (0.2853) loss 4.5915 (3.9393) grad_norm 1.2174 (inf) [2021-04-15 19:10:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][1180/1251] eta 0:00:20 lr 0.000923 time 0.2711 (0.2852) loss 4.3068 (3.9367) grad_norm 0.8493 (inf) [2021-04-15 19:10:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][1190/1251] eta 0:00:17 lr 0.000923 time 0.2649 (0.2852) loss 4.1170 (3.9373) grad_norm 1.1604 (inf) [2021-04-15 19:10:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][1200/1251] eta 0:00:14 lr 0.000923 time 0.2958 (0.2851) loss 3.8238 (3.9370) grad_norm 1.2057 (inf) [2021-04-15 19:10:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][1210/1251] eta 0:00:11 lr 0.000923 time 0.2773 (0.2851) loss 4.0179 (3.9382) grad_norm 0.9961 (inf) [2021-04-15 19:10:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][1220/1251] eta 0:00:08 lr 0.000923 time 0.2710 (0.2851) loss 2.9749 (3.9389) grad_norm 1.0725 (inf) [2021-04-15 19:10:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][1230/1251] eta 0:00:05 lr 0.000923 time 0.2655 (0.2850) loss 3.8391 (3.9392) grad_norm 0.9506 (inf) [2021-04-15 19:10:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][1240/1251] eta 0:00:03 lr 0.000923 time 0.2492 (0.2849) loss 3.8561 (3.9395) grad_norm 1.0525 (inf) [2021-04-15 19:10:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [53/300][1250/1251] eta 0:00:00 lr 0.000923 time 0.2485 (0.2846) loss 4.7227 (3.9370) grad_norm 0.9909 (inf) [2021-04-15 19:10:52 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 53 training takes 0:05:58 [2021-04-15 19:10:52 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_53.pth saving...... [2021-04-15 19:11:05 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_53.pth saved !!! [2021-04-15 19:11:07 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.259 (1.259) Loss 1.3345 (1.3345) Acc@1 69.043 (69.043) Acc@5 89.355 (89.355) [2021-04-15 19:11:08 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.113 (0.231) Loss 1.3185 (1.2920) Acc@1 70.312 (70.064) Acc@5 89.844 (90.439) [2021-04-15 19:11:10 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.238 (0.234) Loss 1.1793 (1.2919) Acc@1 72.363 (70.215) Acc@5 91.211 (90.383) [2021-04-15 19:11:13 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.114 (0.245) Loss 1.2774 (1.2954) Acc@1 70.996 (70.243) Acc@5 90.137 (90.222) [2021-04-15 19:11:15 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.228) Loss 1.3746 (1.2917) Acc@1 69.141 (70.203) Acc@5 89.941 (90.356) [2021-04-15 19:11:17 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 70.316 Acc@5 90.372 [2021-04-15 19:11:17 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 70.3% [2021-04-15 19:11:17 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 70.42% [2021-04-15 19:11:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][0/1251] eta 1:28:25 lr 0.000923 time 4.2412 (4.2412) loss 3.9147 (3.9147) grad_norm 1.0840 (1.0840) [2021-04-15 19:11:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][10/1251] eta 0:13:23 lr 0.000923 time 0.3122 (0.6474) loss 4.7084 (4.0106) grad_norm 1.1317 (1.1541) [2021-04-15 19:11:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][20/1251] eta 0:09:40 lr 0.000923 time 0.2621 (0.4718) loss 4.3814 (4.0829) grad_norm 1.1403 (1.1199) [2021-04-15 19:11:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][30/1251] eta 0:08:22 lr 0.000923 time 0.3097 (0.4114) loss 3.7079 (4.0206) grad_norm 1.1301 (1.1089) [2021-04-15 19:11:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][40/1251] eta 0:07:37 lr 0.000923 time 0.2765 (0.3779) loss 2.9410 (3.9650) grad_norm 1.0156 (1.1076) [2021-04-15 19:11:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][50/1251] eta 0:07:09 lr 0.000923 time 0.2578 (0.3578) loss 3.8753 (3.9334) grad_norm 1.0990 (1.1038) [2021-04-15 19:11:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][60/1251] eta 0:06:55 lr 0.000923 time 0.4225 (0.3488) loss 4.0302 (3.8949) grad_norm 1.2887 (1.1193) [2021-04-15 19:11:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][70/1251] eta 0:06:40 lr 0.000923 time 0.3052 (0.3388) loss 3.9339 (3.8992) grad_norm 0.9715 (1.1169) [2021-04-15 19:11:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][80/1251] eta 0:06:29 lr 0.000923 time 0.2679 (0.3327) loss 3.5831 (3.8760) grad_norm 1.0309 (1.1110) [2021-04-15 19:11:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][90/1251] eta 0:06:21 lr 0.000923 time 0.2646 (0.3286) loss 3.9694 (3.9088) grad_norm 0.9576 (1.1059) [2021-04-15 19:11:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][100/1251] eta 0:06:14 lr 0.000923 time 0.2548 (0.3255) loss 3.4567 (3.9329) grad_norm 1.0779 (1.0992) [2021-04-15 19:11:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][110/1251] eta 0:06:06 lr 0.000923 time 0.2541 (0.3215) loss 3.8192 (3.9600) grad_norm 1.0966 (1.0974) [2021-04-15 19:11:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][120/1251] eta 0:06:00 lr 0.000923 time 0.2755 (0.3184) loss 4.6094 (3.9667) grad_norm 0.9398 (1.0954) [2021-04-15 19:11:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][130/1251] eta 0:05:55 lr 0.000923 time 0.2689 (0.3168) loss 4.0912 (3.9909) grad_norm 1.1361 (1.0943) [2021-04-15 19:12:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][140/1251] eta 0:05:48 lr 0.000923 time 0.2598 (0.3141) loss 3.8558 (3.9899) grad_norm 1.3971 (1.1016) [2021-04-15 19:12:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][150/1251] eta 0:05:43 lr 0.000923 time 0.3031 (0.3121) loss 4.1212 (3.9881) grad_norm 1.3868 (1.1125) [2021-04-15 19:12:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][160/1251] eta 0:05:38 lr 0.000923 time 0.2956 (0.3102) loss 4.9480 (3.9812) grad_norm 1.1515 (1.1188) [2021-04-15 19:12:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][170/1251] eta 0:05:33 lr 0.000923 time 0.3020 (0.3085) loss 4.2678 (3.9686) grad_norm 1.0897 (1.1242) [2021-04-15 19:12:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][180/1251] eta 0:05:28 lr 0.000923 time 0.2593 (0.3067) loss 4.1558 (3.9618) grad_norm 1.1355 (1.1265) [2021-04-15 19:12:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][190/1251] eta 0:05:23 lr 0.000923 time 0.2970 (0.3053) loss 4.2623 (3.9728) grad_norm 1.0198 (1.1284) [2021-04-15 19:12:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][200/1251] eta 0:05:19 lr 0.000922 time 0.3185 (0.3041) loss 2.6744 (3.9692) grad_norm 0.9852 (1.1297) [2021-04-15 19:12:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][210/1251] eta 0:05:15 lr 0.000922 time 0.2845 (0.3029) loss 3.4026 (3.9594) grad_norm 1.3284 (1.1275) [2021-04-15 19:12:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][220/1251] eta 0:05:11 lr 0.000922 time 0.2699 (0.3018) loss 4.2319 (3.9705) grad_norm 0.8964 (1.1262) [2021-04-15 19:12:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][230/1251] eta 0:05:07 lr 0.000922 time 0.2687 (0.3009) loss 3.9486 (3.9692) grad_norm 1.1064 (1.1258) [2021-04-15 19:12:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][240/1251] eta 0:05:03 lr 0.000922 time 0.3083 (0.3001) loss 3.6815 (3.9680) grad_norm 1.1752 (1.1244) [2021-04-15 19:12:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][250/1251] eta 0:04:59 lr 0.000922 time 0.2618 (0.2991) loss 3.2529 (3.9508) grad_norm 0.9252 (1.1216) [2021-04-15 19:12:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][260/1251] eta 0:04:55 lr 0.000922 time 0.2750 (0.2983) loss 4.3372 (3.9597) grad_norm 1.0681 (1.1209) [2021-04-15 19:12:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][270/1251] eta 0:04:52 lr 0.000922 time 0.2891 (0.2977) loss 4.2630 (3.9656) grad_norm 1.1826 (1.1205) [2021-04-15 19:12:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][280/1251] eta 0:04:48 lr 0.000922 time 0.2757 (0.2971) loss 3.4870 (3.9656) grad_norm 1.2787 (1.1257) [2021-04-15 19:12:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][290/1251] eta 0:04:44 lr 0.000922 time 0.2787 (0.2964) loss 3.7578 (3.9539) grad_norm 0.9571 (1.1254) [2021-04-15 19:12:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][300/1251] eta 0:04:41 lr 0.000922 time 0.2511 (0.2964) loss 4.6726 (3.9578) grad_norm 1.3017 (1.1262) [2021-04-15 19:12:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][310/1251] eta 0:04:38 lr 0.000922 time 0.2679 (0.2960) loss 4.4180 (3.9574) grad_norm 1.4217 (1.1260) [2021-04-15 19:12:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][320/1251] eta 0:04:35 lr 0.000922 time 0.2820 (0.2959) loss 4.2645 (3.9623) grad_norm 1.3577 (1.1249) [2021-04-15 19:12:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][330/1251] eta 0:04:32 lr 0.000922 time 0.2725 (0.2954) loss 2.3198 (3.9597) grad_norm 1.2337 (1.1270) [2021-04-15 19:12:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][340/1251] eta 0:04:28 lr 0.000922 time 0.2893 (0.2950) loss 4.7625 (3.9599) grad_norm 1.0185 (1.1293) [2021-04-15 19:13:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][350/1251] eta 0:04:26 lr 0.000922 time 0.2598 (0.2954) loss 4.1961 (3.9646) grad_norm 0.9105 (1.1277) [2021-04-15 19:13:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][360/1251] eta 0:04:22 lr 0.000922 time 0.2729 (0.2952) loss 2.6983 (3.9530) grad_norm 1.3703 (1.1291) [2021-04-15 19:13:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][370/1251] eta 0:04:19 lr 0.000922 time 0.2533 (0.2946) loss 2.6365 (3.9533) grad_norm 1.0496 (1.1271) [2021-04-15 19:13:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][380/1251] eta 0:04:16 lr 0.000922 time 0.2741 (0.2943) loss 3.5496 (3.9593) grad_norm 1.3298 (1.1275) [2021-04-15 19:13:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][390/1251] eta 0:04:13 lr 0.000922 time 0.2572 (0.2940) loss 3.3493 (3.9629) grad_norm 0.9562 (1.1256) [2021-04-15 19:13:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][400/1251] eta 0:04:09 lr 0.000922 time 0.2688 (0.2936) loss 4.3412 (3.9638) grad_norm 1.2690 (1.1249) [2021-04-15 19:13:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][410/1251] eta 0:04:06 lr 0.000922 time 0.2677 (0.2933) loss 3.6327 (3.9647) grad_norm 0.9817 (1.1235) [2021-04-15 19:13:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][420/1251] eta 0:04:03 lr 0.000922 time 0.2782 (0.2931) loss 3.0591 (3.9650) grad_norm 1.1023 (1.1237) [2021-04-15 19:13:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][430/1251] eta 0:04:00 lr 0.000922 time 0.3057 (0.2928) loss 4.3072 (3.9682) grad_norm 1.0145 (1.1235) [2021-04-15 19:13:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][440/1251] eta 0:03:57 lr 0.000922 time 0.2888 (0.2925) loss 3.1044 (3.9668) grad_norm 1.0270 (1.1235) [2021-04-15 19:13:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][450/1251] eta 0:03:54 lr 0.000922 time 0.2895 (0.2922) loss 3.6316 (3.9621) grad_norm 1.1382 (1.1238) [2021-04-15 19:13:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][460/1251] eta 0:03:50 lr 0.000922 time 0.2777 (0.2919) loss 3.2656 (3.9564) grad_norm 1.0815 (1.1228) [2021-04-15 19:13:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][470/1251] eta 0:03:47 lr 0.000922 time 0.2922 (0.2915) loss 3.5003 (3.9602) grad_norm 0.9558 (1.1235) [2021-04-15 19:13:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][480/1251] eta 0:03:44 lr 0.000922 time 0.2556 (0.2913) loss 3.8130 (3.9649) grad_norm 0.9759 (1.1225) [2021-04-15 19:13:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][490/1251] eta 0:03:41 lr 0.000922 time 0.2699 (0.2910) loss 4.0460 (3.9703) grad_norm 0.9307 (1.1231) [2021-04-15 19:13:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][500/1251] eta 0:03:38 lr 0.000922 time 0.2575 (0.2909) loss 4.1866 (3.9640) grad_norm 1.0748 (1.1230) [2021-04-15 19:13:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][510/1251] eta 0:03:35 lr 0.000922 time 0.2759 (0.2908) loss 3.9557 (3.9661) grad_norm 1.0532 (1.1226) [2021-04-15 19:13:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][520/1251] eta 0:03:32 lr 0.000922 time 0.2983 (0.2905) loss 4.6276 (3.9706) grad_norm 1.1821 (1.1223) [2021-04-15 19:13:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][530/1251] eta 0:03:29 lr 0.000922 time 0.2894 (0.2903) loss 3.4103 (3.9663) grad_norm 1.0703 (1.1229) [2021-04-15 19:13:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][540/1251] eta 0:03:26 lr 0.000922 time 0.2708 (0.2901) loss 4.5019 (3.9634) grad_norm 1.0298 (1.1242) [2021-04-15 19:13:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][550/1251] eta 0:03:23 lr 0.000922 time 0.2920 (0.2899) loss 4.0133 (3.9634) grad_norm 1.0589 (1.1250) [2021-04-15 19:14:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][560/1251] eta 0:03:20 lr 0.000922 time 0.2801 (0.2897) loss 2.6220 (3.9633) grad_norm 1.0217 (1.1239) [2021-04-15 19:14:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][570/1251] eta 0:03:17 lr 0.000922 time 0.2472 (0.2894) loss 4.0718 (3.9661) grad_norm 1.3816 (1.1239) [2021-04-15 19:14:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][580/1251] eta 0:03:14 lr 0.000922 time 0.2882 (0.2896) loss 4.4175 (3.9683) grad_norm 1.1607 (1.1259) [2021-04-15 19:14:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][590/1251] eta 0:03:11 lr 0.000922 time 0.2556 (0.2894) loss 3.7804 (3.9700) grad_norm 1.0735 (1.1271) [2021-04-15 19:14:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][600/1251] eta 0:03:08 lr 0.000922 time 0.3011 (0.2892) loss 3.9446 (3.9733) grad_norm 1.1757 (1.1274) [2021-04-15 19:14:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][610/1251] eta 0:03:05 lr 0.000922 time 0.2545 (0.2890) loss 3.7784 (3.9773) grad_norm 1.1277 (1.1263) [2021-04-15 19:14:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][620/1251] eta 0:03:02 lr 0.000922 time 0.2730 (0.2888) loss 4.0194 (3.9742) grad_norm 1.0549 (1.1247) [2021-04-15 19:14:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][630/1251] eta 0:02:59 lr 0.000922 time 0.2582 (0.2887) loss 4.3963 (3.9741) grad_norm 0.9183 (1.1249) [2021-04-15 19:14:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][640/1251] eta 0:02:56 lr 0.000922 time 0.2781 (0.2889) loss 3.2208 (3.9684) grad_norm 1.0204 (1.1249) [2021-04-15 19:14:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][650/1251] eta 0:02:53 lr 0.000921 time 0.2593 (0.2889) loss 3.9659 (3.9645) grad_norm 1.0140 (1.1242) [2021-04-15 19:14:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][660/1251] eta 0:02:50 lr 0.000921 time 0.2892 (0.2887) loss 3.5735 (3.9620) grad_norm 1.1828 (1.1265) [2021-04-15 19:14:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][670/1251] eta 0:02:47 lr 0.000921 time 0.2822 (0.2887) loss 4.1637 (3.9664) grad_norm 0.9664 (1.1267) [2021-04-15 19:14:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][680/1251] eta 0:02:44 lr 0.000921 time 0.2851 (0.2887) loss 2.3559 (3.9611) grad_norm 1.0303 (1.1269) [2021-04-15 19:14:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][690/1251] eta 0:02:41 lr 0.000921 time 0.3012 (0.2886) loss 3.1024 (3.9583) grad_norm 1.2700 (1.1282) [2021-04-15 19:14:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][700/1251] eta 0:02:38 lr 0.000921 time 0.2474 (0.2885) loss 3.3086 (3.9556) grad_norm 1.0941 (1.1285) [2021-04-15 19:14:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][710/1251] eta 0:02:36 lr 0.000921 time 0.2559 (0.2884) loss 4.6520 (3.9571) grad_norm 1.0619 (1.1281) [2021-04-15 19:14:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][720/1251] eta 0:02:33 lr 0.000921 time 0.2591 (0.2882) loss 4.0431 (3.9545) grad_norm 0.9766 (1.1277) [2021-04-15 19:14:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][730/1251] eta 0:02:30 lr 0.000921 time 0.3063 (0.2882) loss 3.0535 (3.9561) grad_norm 1.1078 (1.1266) [2021-04-15 19:14:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][740/1251] eta 0:02:27 lr 0.000921 time 0.2886 (0.2881) loss 3.7622 (3.9526) grad_norm 1.0429 (1.1252) [2021-04-15 19:14:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][750/1251] eta 0:02:24 lr 0.000921 time 0.2749 (0.2880) loss 3.7185 (3.9515) grad_norm 1.2699 (1.1258) [2021-04-15 19:14:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][760/1251] eta 0:02:21 lr 0.000921 time 0.2674 (0.2879) loss 2.8436 (3.9502) grad_norm 1.0705 (1.1248) [2021-04-15 19:14:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][770/1251] eta 0:02:18 lr 0.000921 time 0.2615 (0.2878) loss 2.9121 (3.9486) grad_norm 1.1411 (1.1243) [2021-04-15 19:15:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][780/1251] eta 0:02:15 lr 0.000921 time 0.2704 (0.2877) loss 4.5535 (3.9473) grad_norm 1.3713 (1.1245) [2021-04-15 19:15:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][790/1251] eta 0:02:12 lr 0.000921 time 0.2443 (0.2876) loss 3.5631 (3.9454) grad_norm 1.2673 (1.1251) [2021-04-15 19:15:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][800/1251] eta 0:02:09 lr 0.000921 time 0.2860 (0.2875) loss 4.8544 (3.9479) grad_norm 1.3092 (1.1251) [2021-04-15 19:15:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][810/1251] eta 0:02:06 lr 0.000921 time 0.2565 (0.2873) loss 3.7781 (3.9463) grad_norm 1.0039 (1.1240) [2021-04-15 19:15:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][820/1251] eta 0:02:03 lr 0.000921 time 0.2695 (0.2872) loss 4.3973 (3.9468) grad_norm 1.2214 (1.1241) [2021-04-15 19:15:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][830/1251] eta 0:02:00 lr 0.000921 time 0.2686 (0.2871) loss 3.0121 (3.9438) grad_norm 1.2303 (1.1241) [2021-04-15 19:15:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][840/1251] eta 0:01:57 lr 0.000921 time 0.2741 (0.2869) loss 3.2634 (3.9422) grad_norm 1.2984 (1.1244) [2021-04-15 19:15:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][850/1251] eta 0:01:55 lr 0.000921 time 0.3054 (0.2869) loss 2.6716 (3.9402) grad_norm 1.2293 (1.1244) [2021-04-15 19:15:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][860/1251] eta 0:01:52 lr 0.000921 time 0.2773 (0.2867) loss 3.9082 (3.9412) grad_norm 0.9919 (1.1229) [2021-04-15 19:15:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][870/1251] eta 0:01:49 lr 0.000921 time 0.2928 (0.2866) loss 4.1658 (3.9411) grad_norm 1.3130 (1.1224) [2021-04-15 19:15:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][880/1251] eta 0:01:46 lr 0.000921 time 0.2702 (0.2866) loss 3.9589 (3.9396) grad_norm 0.9587 (1.1228) [2021-04-15 19:15:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][890/1251] eta 0:01:43 lr 0.000921 time 0.2452 (0.2866) loss 4.1093 (3.9399) grad_norm 1.0078 (1.1229) [2021-04-15 19:15:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][900/1251] eta 0:01:40 lr 0.000921 time 0.2732 (0.2866) loss 3.8401 (3.9427) grad_norm 1.3255 (1.1232) [2021-04-15 19:15:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][910/1251] eta 0:01:37 lr 0.000921 time 0.2994 (0.2865) loss 4.3301 (3.9420) grad_norm 1.1377 (1.1230) [2021-04-15 19:15:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][920/1251] eta 0:01:34 lr 0.000921 time 0.2806 (0.2864) loss 3.6387 (3.9386) grad_norm 1.0421 (1.1230) [2021-04-15 19:15:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][930/1251] eta 0:01:31 lr 0.000921 time 0.2699 (0.2864) loss 4.5461 (3.9375) grad_norm 1.0879 (1.1224) [2021-04-15 19:15:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][940/1251] eta 0:01:29 lr 0.000921 time 0.2722 (0.2864) loss 4.2753 (3.9361) grad_norm 1.1825 (1.1224) [2021-04-15 19:15:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][950/1251] eta 0:01:26 lr 0.000921 time 0.2681 (0.2864) loss 4.4307 (3.9389) grad_norm 1.1512 (1.1230) [2021-04-15 19:15:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][960/1251] eta 0:01:23 lr 0.000921 time 0.2782 (0.2864) loss 3.7057 (3.9376) grad_norm 1.2556 (1.1229) [2021-04-15 19:15:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][970/1251] eta 0:01:20 lr 0.000921 time 0.2660 (0.2863) loss 4.3168 (3.9360) grad_norm 1.2552 (1.1222) [2021-04-15 19:15:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][980/1251] eta 0:01:17 lr 0.000921 time 0.2829 (0.2862) loss 3.0553 (3.9370) grad_norm 1.1103 (1.1242) [2021-04-15 19:16:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][990/1251] eta 0:01:14 lr 0.000921 time 0.2939 (0.2861) loss 4.8279 (3.9370) grad_norm 0.9226 (1.1243) [2021-04-15 19:16:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][1000/1251] eta 0:01:11 lr 0.000921 time 0.2710 (0.2860) loss 3.8563 (3.9375) grad_norm 1.2898 (1.1245) [2021-04-15 19:16:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][1010/1251] eta 0:01:08 lr 0.000921 time 0.2835 (0.2860) loss 3.5594 (3.9366) grad_norm 0.9451 (1.1245) [2021-04-15 19:16:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][1020/1251] eta 0:01:06 lr 0.000921 time 0.2826 (0.2860) loss 3.2332 (3.9367) grad_norm 1.3032 (1.1248) [2021-04-15 19:16:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][1030/1251] eta 0:01:03 lr 0.000921 time 0.2946 (0.2860) loss 4.1432 (3.9390) grad_norm 1.0307 (1.1249) [2021-04-15 19:16:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][1040/1251] eta 0:01:00 lr 0.000921 time 0.2581 (0.2859) loss 3.5971 (3.9380) grad_norm 1.1576 (1.1246) [2021-04-15 19:16:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][1050/1251] eta 0:00:57 lr 0.000921 time 0.2781 (0.2859) loss 3.4477 (3.9383) grad_norm 1.2239 (1.1246) [2021-04-15 19:16:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][1060/1251] eta 0:00:54 lr 0.000921 time 0.2757 (0.2858) loss 3.3888 (3.9370) grad_norm 1.2614 (1.1241) [2021-04-15 19:16:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][1070/1251] eta 0:00:51 lr 0.000921 time 0.2993 (0.2857) loss 4.2427 (3.9364) grad_norm 1.1146 (1.1234) [2021-04-15 19:16:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][1080/1251] eta 0:00:48 lr 0.000921 time 0.2762 (0.2856) loss 4.8723 (3.9395) grad_norm 1.3365 (1.1239) [2021-04-15 19:16:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][1090/1251] eta 0:00:45 lr 0.000921 time 0.2676 (0.2856) loss 3.9534 (3.9414) grad_norm 0.9523 (1.1236) [2021-04-15 19:16:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][1100/1251] eta 0:00:43 lr 0.000920 time 0.2755 (0.2856) loss 3.9866 (3.9412) grad_norm 1.0466 (1.1233) [2021-04-15 19:16:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][1110/1251] eta 0:00:40 lr 0.000920 time 0.3244 (0.2855) loss 4.3394 (3.9433) grad_norm 1.0774 (1.1227) [2021-04-15 19:16:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][1120/1251] eta 0:00:37 lr 0.000920 time 0.2762 (0.2855) loss 3.8770 (3.9439) grad_norm 1.1323 (1.1227) [2021-04-15 19:16:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][1130/1251] eta 0:00:34 lr 0.000920 time 0.2738 (0.2854) loss 3.0935 (3.9449) grad_norm 0.9625 (1.1231) [2021-04-15 19:16:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][1140/1251] eta 0:00:31 lr 0.000920 time 0.2808 (0.2854) loss 4.7356 (3.9457) grad_norm 1.0810 (1.1228) [2021-04-15 19:16:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][1150/1251] eta 0:00:28 lr 0.000920 time 0.4408 (0.2854) loss 4.2855 (3.9486) grad_norm 1.0824 (1.1232) [2021-04-15 19:16:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][1160/1251] eta 0:00:25 lr 0.000920 time 0.2803 (0.2853) loss 4.2732 (3.9513) grad_norm 1.0685 (1.1227) [2021-04-15 19:16:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][1170/1251] eta 0:00:23 lr 0.000920 time 0.2858 (0.2852) loss 4.2820 (3.9499) grad_norm 1.0079 (1.1219) [2021-04-15 19:16:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][1180/1251] eta 0:00:20 lr 0.000920 time 0.2907 (0.2852) loss 2.7592 (3.9479) grad_norm 1.2075 (1.1218) [2021-04-15 19:16:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][1190/1251] eta 0:00:17 lr 0.000920 time 0.2845 (0.2851) loss 3.9765 (3.9501) grad_norm 0.9464 (1.1216) [2021-04-15 19:17:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][1200/1251] eta 0:00:14 lr 0.000920 time 0.2605 (0.2850) loss 4.0933 (3.9517) grad_norm 1.0876 (1.1214) [2021-04-15 19:17:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][1210/1251] eta 0:00:11 lr 0.000920 time 0.2666 (0.2849) loss 4.2391 (3.9494) grad_norm 1.0547 (1.1210) [2021-04-15 19:17:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][1220/1251] eta 0:00:08 lr 0.000920 time 0.3056 (0.2850) loss 3.9387 (3.9470) grad_norm 1.0639 (1.1207) [2021-04-15 19:17:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][1230/1251] eta 0:00:05 lr 0.000920 time 0.2937 (0.2850) loss 4.0888 (3.9464) grad_norm 1.0644 (1.1210) [2021-04-15 19:17:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][1240/1251] eta 0:00:03 lr 0.000920 time 0.2504 (0.2849) loss 3.1734 (3.9468) grad_norm 1.1834 (1.1212) [2021-04-15 19:17:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [54/300][1250/1251] eta 0:00:00 lr 0.000920 time 0.2498 (0.2846) loss 2.8102 (3.9442) grad_norm 0.9303 (1.1205) [2021-04-15 19:17:16 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 54 training takes 0:05:58 [2021-04-15 19:17:16 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_54.pth saving...... [2021-04-15 19:17:29 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_54.pth saved !!! [2021-04-15 19:17:31 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.156 (1.156) Loss 1.3674 (1.3674) Acc@1 69.629 (69.629) Acc@5 88.965 (88.965) [2021-04-15 19:17:32 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.163 (0.238) Loss 1.2534 (1.2995) Acc@1 72.168 (69.798) Acc@5 90.234 (90.190) [2021-04-15 19:17:35 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.114 (0.245) Loss 1.2127 (1.2869) Acc@1 71.484 (70.238) Acc@5 90.723 (90.309) [2021-04-15 19:17:37 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.124 (0.231) Loss 1.2908 (1.2808) Acc@1 69.629 (70.297) Acc@5 89.551 (90.329) [2021-04-15 19:17:39 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.632 (0.225) Loss 1.3060 (1.2818) Acc@1 70.117 (70.312) Acc@5 89.453 (90.292) [2021-04-15 19:17:41 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 70.500 Acc@5 90.380 [2021-04-15 19:17:41 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 70.5% [2021-04-15 19:17:41 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 70.50% [2021-04-15 19:17:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][0/1251] eta 1:21:49 lr 0.000920 time 3.9246 (3.9246) loss 2.6533 (2.6533) grad_norm 1.2753 (1.2753) [2021-04-15 19:17:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][10/1251] eta 0:12:51 lr 0.000920 time 0.3083 (0.6216) loss 3.8115 (3.8568) grad_norm 1.0210 (1.1124) [2021-04-15 19:17:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][20/1251] eta 0:09:22 lr 0.000920 time 0.2918 (0.4569) loss 4.0386 (3.9698) grad_norm 1.2119 (1.1266) [2021-04-15 19:17:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][30/1251] eta 0:08:07 lr 0.000920 time 0.2834 (0.3995) loss 3.4246 (4.0587) grad_norm 1.0437 (1.1369) [2021-04-15 19:17:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][40/1251] eta 0:07:27 lr 0.000920 time 0.2959 (0.3698) loss 4.7476 (4.0153) grad_norm 1.2705 (1.1602) [2021-04-15 19:17:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][50/1251] eta 0:07:02 lr 0.000920 time 0.2535 (0.3515) loss 2.9087 (3.9927) grad_norm 0.9727 (1.1516) [2021-04-15 19:18:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][60/1251] eta 0:06:44 lr 0.000920 time 0.2580 (0.3394) loss 2.8293 (3.9812) grad_norm 1.1310 (1.1443) [2021-04-15 19:18:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][70/1251] eta 0:06:30 lr 0.000920 time 0.2618 (0.3307) loss 3.9203 (3.9252) grad_norm 1.0670 (1.1381) [2021-04-15 19:18:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][80/1251] eta 0:06:21 lr 0.000920 time 0.3002 (0.3259) loss 4.3945 (3.9356) grad_norm 1.1154 (1.1290) [2021-04-15 19:18:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][90/1251] eta 0:06:13 lr 0.000920 time 0.3045 (0.3217) loss 4.2538 (3.8991) grad_norm 1.1852 (1.1237) [2021-04-15 19:18:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][100/1251] eta 0:06:05 lr 0.000920 time 0.2958 (0.3173) loss 4.0807 (3.8908) grad_norm 1.0726 (1.1261) [2021-04-15 19:18:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][110/1251] eta 0:05:58 lr 0.000920 time 0.2658 (0.3141) loss 2.4992 (3.8794) grad_norm 0.9701 (1.1277) [2021-04-15 19:18:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][120/1251] eta 0:05:51 lr 0.000920 time 0.2518 (0.3107) loss 4.7956 (3.8889) grad_norm 1.1188 (1.1296) [2021-04-15 19:18:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][130/1251] eta 0:05:46 lr 0.000920 time 0.2874 (0.3091) loss 3.3904 (3.8865) grad_norm 1.1511 (1.1230) [2021-04-15 19:18:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][140/1251] eta 0:05:41 lr 0.000920 time 0.2811 (0.3078) loss 4.1905 (3.8873) grad_norm 1.2654 (1.1220) [2021-04-15 19:18:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][150/1251] eta 0:05:36 lr 0.000920 time 0.2472 (0.3058) loss 4.1319 (3.8942) grad_norm 1.1407 (1.1229) [2021-04-15 19:18:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][160/1251] eta 0:05:31 lr 0.000920 time 0.2743 (0.3042) loss 4.1714 (3.8920) grad_norm 1.1083 (1.1219) [2021-04-15 19:18:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][170/1251] eta 0:05:27 lr 0.000920 time 0.2923 (0.3030) loss 4.3006 (3.9015) grad_norm 1.0766 (1.1229) [2021-04-15 19:18:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][180/1251] eta 0:05:22 lr 0.000920 time 0.2571 (0.3015) loss 3.6168 (3.8962) grad_norm 1.1090 (1.1266) [2021-04-15 19:18:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][190/1251] eta 0:05:18 lr 0.000920 time 0.2785 (0.3005) loss 3.7899 (3.8971) grad_norm 0.9647 (1.1269) [2021-04-15 19:18:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][200/1251] eta 0:05:14 lr 0.000920 time 0.2661 (0.2995) loss 4.0234 (3.8978) grad_norm 1.2545 (1.1269) [2021-04-15 19:18:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][210/1251] eta 0:05:10 lr 0.000920 time 0.2921 (0.2987) loss 3.1091 (3.8988) grad_norm 1.0862 (1.1274) [2021-04-15 19:18:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][220/1251] eta 0:05:06 lr 0.000920 time 0.2587 (0.2977) loss 4.2808 (3.8999) grad_norm 1.4606 (1.1281) [2021-04-15 19:18:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][230/1251] eta 0:05:03 lr 0.000920 time 0.2727 (0.2973) loss 4.1612 (3.9034) grad_norm 1.2422 (1.1296) [2021-04-15 19:18:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][240/1251] eta 0:04:59 lr 0.000920 time 0.2895 (0.2965) loss 4.7911 (3.9075) grad_norm 1.3946 (1.1298) [2021-04-15 19:18:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][250/1251] eta 0:04:55 lr 0.000920 time 0.2448 (0.2956) loss 4.4850 (3.9185) grad_norm 0.9504 (1.1251) [2021-04-15 19:18:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][260/1251] eta 0:04:52 lr 0.000920 time 0.2648 (0.2949) loss 3.9175 (3.9126) grad_norm 1.1124 (inf) [2021-04-15 19:19:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][270/1251] eta 0:04:48 lr 0.000920 time 0.2765 (0.2943) loss 3.3536 (3.9056) grad_norm 0.9858 (inf) [2021-04-15 19:19:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][280/1251] eta 0:04:45 lr 0.000920 time 0.2880 (0.2937) loss 3.5940 (3.9016) grad_norm 1.0638 (inf) [2021-04-15 19:19:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][290/1251] eta 0:04:41 lr 0.000919 time 0.2637 (0.2932) loss 4.4557 (3.9038) grad_norm 1.0908 (inf) [2021-04-15 19:19:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][300/1251] eta 0:04:38 lr 0.000919 time 0.2658 (0.2927) loss 4.1804 (3.9063) grad_norm 1.0589 (inf) [2021-04-15 19:19:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][310/1251] eta 0:04:35 lr 0.000919 time 0.2963 (0.2923) loss 4.9355 (3.9186) grad_norm 1.1541 (inf) [2021-04-15 19:19:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][320/1251] eta 0:04:31 lr 0.000919 time 0.2828 (0.2918) loss 4.3229 (3.9064) grad_norm 1.2759 (inf) [2021-04-15 19:19:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][330/1251] eta 0:04:28 lr 0.000919 time 0.2839 (0.2914) loss 4.3242 (3.9078) grad_norm 0.9151 (inf) [2021-04-15 19:19:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][340/1251] eta 0:04:25 lr 0.000919 time 0.2775 (0.2911) loss 4.0303 (3.9150) grad_norm 0.9640 (inf) [2021-04-15 19:19:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][350/1251] eta 0:04:22 lr 0.000919 time 0.4444 (0.2912) loss 3.1858 (3.9122) grad_norm 1.0619 (inf) [2021-04-15 19:19:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][360/1251] eta 0:04:19 lr 0.000919 time 0.2684 (0.2911) loss 4.7190 (3.9242) grad_norm 1.0924 (inf) [2021-04-15 19:19:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][370/1251] eta 0:04:16 lr 0.000919 time 0.2978 (0.2913) loss 3.8925 (3.9280) grad_norm 0.9680 (inf) [2021-04-15 19:19:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][380/1251] eta 0:04:13 lr 0.000919 time 0.2738 (0.2910) loss 4.1926 (3.9344) grad_norm 1.0100 (inf) [2021-04-15 19:19:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][390/1251] eta 0:04:10 lr 0.000919 time 0.3086 (0.2907) loss 4.3894 (3.9391) grad_norm 1.4501 (inf) [2021-04-15 19:19:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][400/1251] eta 0:04:07 lr 0.000919 time 0.2570 (0.2904) loss 4.5166 (3.9471) grad_norm 1.2629 (inf) [2021-04-15 19:19:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][410/1251] eta 0:04:03 lr 0.000919 time 0.2817 (0.2901) loss 4.4406 (3.9489) grad_norm 0.8751 (inf) [2021-04-15 19:19:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][420/1251] eta 0:04:01 lr 0.000919 time 0.2694 (0.2901) loss 4.0743 (3.9548) grad_norm 0.9617 (inf) [2021-04-15 19:19:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][430/1251] eta 0:03:58 lr 0.000919 time 0.2876 (0.2901) loss 4.3828 (3.9579) grad_norm 0.9773 (inf) [2021-04-15 19:19:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][440/1251] eta 0:03:55 lr 0.000919 time 0.2999 (0.2900) loss 2.9531 (3.9602) grad_norm 1.0546 (inf) [2021-04-15 19:19:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][450/1251] eta 0:03:52 lr 0.000919 time 0.2747 (0.2897) loss 3.1536 (3.9504) grad_norm 0.9678 (inf) [2021-04-15 19:19:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][460/1251] eta 0:03:48 lr 0.000919 time 0.2962 (0.2895) loss 3.0852 (3.9541) grad_norm 1.1205 (inf) [2021-04-15 19:19:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][470/1251] eta 0:03:46 lr 0.000919 time 0.2577 (0.2896) loss 2.6467 (3.9514) grad_norm 1.1720 (inf) [2021-04-15 19:20:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][480/1251] eta 0:03:43 lr 0.000919 time 0.2607 (0.2893) loss 3.2425 (3.9499) grad_norm 1.0403 (inf) [2021-04-15 19:20:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][490/1251] eta 0:03:39 lr 0.000919 time 0.2795 (0.2891) loss 4.4164 (3.9466) grad_norm 1.2909 (inf) [2021-04-15 19:20:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][500/1251] eta 0:03:36 lr 0.000919 time 0.2925 (0.2888) loss 4.4391 (3.9538) grad_norm 1.1178 (inf) [2021-04-15 19:20:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][510/1251] eta 0:03:33 lr 0.000919 time 0.2920 (0.2887) loss 4.1689 (3.9558) grad_norm 1.0160 (inf) [2021-04-15 19:20:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][520/1251] eta 0:03:30 lr 0.000919 time 0.2759 (0.2885) loss 4.6049 (3.9580) grad_norm 0.9978 (inf) [2021-04-15 19:20:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][530/1251] eta 0:03:27 lr 0.000919 time 0.2837 (0.2883) loss 3.8574 (3.9587) grad_norm 1.3024 (inf) [2021-04-15 19:20:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][540/1251] eta 0:03:24 lr 0.000919 time 0.2555 (0.2880) loss 3.2184 (3.9597) grad_norm 0.9775 (inf) [2021-04-15 19:20:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][550/1251] eta 0:03:21 lr 0.000919 time 0.2734 (0.2879) loss 4.2174 (3.9582) grad_norm 1.1640 (inf) [2021-04-15 19:20:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][560/1251] eta 0:03:18 lr 0.000919 time 0.2750 (0.2877) loss 4.1112 (3.9523) grad_norm 0.9779 (inf) [2021-04-15 19:20:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][570/1251] eta 0:03:16 lr 0.000919 time 0.2856 (0.2879) loss 4.3456 (3.9511) grad_norm 0.9326 (inf) [2021-04-15 19:20:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][580/1251] eta 0:03:12 lr 0.000919 time 0.2651 (0.2876) loss 4.0614 (3.9455) grad_norm 1.2400 (inf) [2021-04-15 19:20:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][590/1251] eta 0:03:10 lr 0.000919 time 0.3003 (0.2875) loss 3.2455 (3.9367) grad_norm 1.2961 (inf) [2021-04-15 19:20:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][600/1251] eta 0:03:07 lr 0.000919 time 0.2565 (0.2874) loss 3.7176 (3.9401) grad_norm 1.2158 (inf) [2021-04-15 19:20:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][610/1251] eta 0:03:04 lr 0.000919 time 0.2859 (0.2872) loss 3.5038 (3.9374) grad_norm 1.3661 (inf) [2021-04-15 19:20:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][620/1251] eta 0:03:01 lr 0.000919 time 0.2630 (0.2871) loss 3.9535 (3.9332) grad_norm 1.1156 (inf) [2021-04-15 19:20:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][630/1251] eta 0:02:58 lr 0.000919 time 0.3088 (0.2870) loss 4.6966 (3.9346) grad_norm 1.0934 (inf) [2021-04-15 19:20:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][640/1251] eta 0:02:55 lr 0.000919 time 0.2551 (0.2869) loss 4.3955 (3.9349) grad_norm 1.1142 (inf) [2021-04-15 19:20:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][650/1251] eta 0:02:52 lr 0.000919 time 0.3450 (0.2869) loss 4.1277 (3.9386) grad_norm 1.0201 (inf) [2021-04-15 19:20:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][660/1251] eta 0:02:49 lr 0.000919 time 0.2556 (0.2867) loss 3.8148 (3.9408) grad_norm 1.0729 (inf) [2021-04-15 19:20:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][670/1251] eta 0:02:46 lr 0.000919 time 0.2913 (0.2868) loss 4.6197 (3.9391) grad_norm 0.9208 (inf) [2021-04-15 19:20:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][680/1251] eta 0:02:43 lr 0.000919 time 0.2886 (0.2867) loss 2.9967 (3.9438) grad_norm 1.1416 (inf) [2021-04-15 19:20:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][690/1251] eta 0:02:40 lr 0.000919 time 0.2765 (0.2869) loss 4.2990 (3.9467) grad_norm 0.9439 (inf) [2021-04-15 19:21:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][700/1251] eta 0:02:38 lr 0.000919 time 0.2902 (0.2869) loss 3.8181 (3.9488) grad_norm 1.0470 (inf) [2021-04-15 19:21:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][710/1251] eta 0:02:35 lr 0.000919 time 0.3851 (0.2871) loss 3.3854 (3.9454) grad_norm 1.1192 (inf) [2021-04-15 19:21:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][720/1251] eta 0:02:32 lr 0.000919 time 0.2483 (0.2870) loss 4.2458 (3.9434) grad_norm 0.9958 (inf) [2021-04-15 19:21:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][730/1251] eta 0:02:29 lr 0.000918 time 0.2831 (0.2869) loss 4.0899 (3.9461) grad_norm 1.0977 (inf) [2021-04-15 19:21:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][740/1251] eta 0:02:26 lr 0.000918 time 0.2601 (0.2868) loss 4.0818 (3.9401) grad_norm 1.1972 (inf) [2021-04-15 19:21:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][750/1251] eta 0:02:23 lr 0.000918 time 0.2669 (0.2866) loss 3.7804 (3.9370) grad_norm 1.7280 (inf) [2021-04-15 19:21:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][760/1251] eta 0:02:20 lr 0.000918 time 0.2809 (0.2865) loss 3.1636 (3.9360) grad_norm 0.9199 (inf) [2021-04-15 19:21:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][770/1251] eta 0:02:17 lr 0.000918 time 0.2749 (0.2864) loss 3.9880 (3.9377) grad_norm 0.9970 (inf) [2021-04-15 19:21:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][780/1251] eta 0:02:14 lr 0.000918 time 0.2834 (0.2863) loss 4.3935 (3.9355) grad_norm 1.1718 (inf) [2021-04-15 19:21:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][790/1251] eta 0:02:11 lr 0.000918 time 0.2667 (0.2861) loss 4.4172 (3.9400) grad_norm 1.1056 (inf) [2021-04-15 19:21:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][800/1251] eta 0:02:08 lr 0.000918 time 0.2591 (0.2860) loss 3.4240 (3.9396) grad_norm 1.1567 (inf) [2021-04-15 19:21:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][810/1251] eta 0:02:06 lr 0.000918 time 0.2893 (0.2859) loss 4.7228 (3.9424) grad_norm 1.0870 (inf) [2021-04-15 19:21:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][820/1251] eta 0:02:03 lr 0.000918 time 0.3098 (0.2859) loss 4.2648 (3.9442) grad_norm 0.9462 (inf) [2021-04-15 19:21:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][830/1251] eta 0:02:00 lr 0.000918 time 0.2706 (0.2858) loss 4.2432 (3.9458) grad_norm 1.1773 (inf) [2021-04-15 19:21:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][840/1251] eta 0:01:57 lr 0.000918 time 0.3026 (0.2857) loss 5.0457 (3.9483) grad_norm 1.1775 (inf) [2021-04-15 19:21:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][850/1251] eta 0:01:54 lr 0.000918 time 0.2867 (0.2856) loss 2.9607 (3.9489) grad_norm 1.2935 (inf) [2021-04-15 19:21:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][860/1251] eta 0:01:51 lr 0.000918 time 0.3088 (0.2855) loss 3.0213 (3.9525) grad_norm 1.2841 (inf) [2021-04-15 19:21:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][870/1251] eta 0:01:48 lr 0.000918 time 0.2802 (0.2855) loss 4.6489 (3.9525) grad_norm 1.3512 (inf) [2021-04-15 19:21:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][880/1251] eta 0:01:45 lr 0.000918 time 0.2753 (0.2854) loss 4.2066 (3.9555) grad_norm 1.1014 (inf) [2021-04-15 19:21:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][890/1251] eta 0:01:42 lr 0.000918 time 0.2696 (0.2853) loss 3.9468 (3.9545) grad_norm 1.2647 (inf) [2021-04-15 19:21:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][900/1251] eta 0:01:40 lr 0.000918 time 0.2837 (0.2852) loss 3.9581 (3.9545) grad_norm 1.2120 (inf) [2021-04-15 19:22:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][910/1251] eta 0:01:37 lr 0.000918 time 0.2857 (0.2852) loss 3.7691 (3.9566) grad_norm 1.2726 (inf) [2021-04-15 19:22:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][920/1251] eta 0:01:34 lr 0.000918 time 0.2807 (0.2852) loss 3.6947 (3.9553) grad_norm 0.9987 (inf) [2021-04-15 19:22:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][930/1251] eta 0:01:31 lr 0.000918 time 0.2802 (0.2853) loss 4.4602 (3.9551) grad_norm 1.0049 (inf) [2021-04-15 19:22:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][940/1251] eta 0:01:28 lr 0.000918 time 0.2806 (0.2856) loss 4.1442 (3.9557) grad_norm 0.9525 (inf) [2021-04-15 19:22:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][950/1251] eta 0:01:25 lr 0.000918 time 0.2611 (0.2855) loss 4.3519 (3.9518) grad_norm 0.9680 (inf) [2021-04-15 19:22:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][960/1251] eta 0:01:23 lr 0.000918 time 0.2500 (0.2853) loss 4.0752 (3.9492) grad_norm 1.1118 (inf) [2021-04-15 19:22:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][970/1251] eta 0:01:20 lr 0.000918 time 0.2691 (0.2853) loss 4.3748 (3.9487) grad_norm 1.1239 (inf) [2021-04-15 19:22:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][980/1251] eta 0:01:17 lr 0.000918 time 0.2894 (0.2852) loss 3.4803 (3.9498) grad_norm 1.1479 (inf) [2021-04-15 19:22:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][990/1251] eta 0:01:14 lr 0.000918 time 0.2730 (0.2852) loss 4.1014 (3.9503) grad_norm 1.1389 (inf) [2021-04-15 19:22:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][1000/1251] eta 0:01:11 lr 0.000918 time 0.2757 (0.2852) loss 4.0144 (3.9543) grad_norm 1.6032 (inf) [2021-04-15 19:22:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][1010/1251] eta 0:01:08 lr 0.000918 time 0.2873 (0.2852) loss 3.1028 (3.9548) grad_norm 1.0841 (inf) [2021-04-15 19:22:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][1020/1251] eta 0:01:05 lr 0.000918 time 0.2746 (0.2851) loss 4.0605 (3.9561) grad_norm 0.9471 (inf) [2021-04-15 19:22:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][1030/1251] eta 0:01:02 lr 0.000918 time 0.2752 (0.2851) loss 4.7264 (3.9553) grad_norm 1.1696 (inf) [2021-04-15 19:22:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][1040/1251] eta 0:01:00 lr 0.000918 time 0.2765 (0.2852) loss 4.3326 (3.9545) grad_norm 1.0205 (inf) [2021-04-15 19:22:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][1050/1251] eta 0:00:57 lr 0.000918 time 0.2676 (0.2851) loss 3.8413 (3.9556) grad_norm 0.9309 (inf) [2021-04-15 19:22:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][1060/1251] eta 0:00:54 lr 0.000918 time 0.3122 (0.2850) loss 3.9341 (3.9549) grad_norm 1.3356 (inf) [2021-04-15 19:22:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][1070/1251] eta 0:00:51 lr 0.000918 time 0.2697 (0.2849) loss 5.0060 (3.9547) grad_norm 1.0334 (inf) [2021-04-15 19:22:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][1080/1251] eta 0:00:48 lr 0.000918 time 0.2907 (0.2849) loss 4.3292 (3.9550) grad_norm 1.1039 (inf) [2021-04-15 19:22:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][1090/1251] eta 0:00:45 lr 0.000918 time 0.2943 (0.2849) loss 2.8311 (3.9508) grad_norm 1.0261 (inf) [2021-04-15 19:22:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][1100/1251] eta 0:00:43 lr 0.000918 time 0.2622 (0.2848) loss 4.1528 (3.9494) grad_norm 0.9504 (inf) [2021-04-15 19:22:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][1110/1251] eta 0:00:40 lr 0.000918 time 0.2793 (0.2848) loss 3.5087 (3.9483) grad_norm 1.0599 (inf) [2021-04-15 19:23:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][1120/1251] eta 0:00:37 lr 0.000918 time 0.3190 (0.2848) loss 4.0583 (3.9491) grad_norm 1.0772 (inf) [2021-04-15 19:23:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][1130/1251] eta 0:00:34 lr 0.000918 time 0.3035 (0.2848) loss 4.4929 (3.9498) grad_norm 0.9601 (inf) [2021-04-15 19:23:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][1140/1251] eta 0:00:31 lr 0.000918 time 0.3055 (0.2848) loss 4.2308 (3.9479) grad_norm 1.1673 (inf) [2021-04-15 19:23:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][1150/1251] eta 0:00:28 lr 0.000918 time 0.2726 (0.2848) loss 3.9115 (3.9485) grad_norm 1.0106 (inf) [2021-04-15 19:23:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][1160/1251] eta 0:00:25 lr 0.000918 time 0.2884 (0.2848) loss 4.6698 (3.9494) grad_norm 1.1899 (inf) [2021-04-15 19:23:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][1170/1251] eta 0:00:23 lr 0.000917 time 0.2771 (0.2848) loss 3.0917 (3.9462) grad_norm 1.0460 (inf) [2021-04-15 19:23:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][1180/1251] eta 0:00:20 lr 0.000917 time 0.2656 (0.2847) loss 2.5311 (3.9432) grad_norm 1.0387 (inf) [2021-04-15 19:23:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][1190/1251] eta 0:00:17 lr 0.000917 time 0.2873 (0.2846) loss 3.3436 (3.9387) grad_norm 1.4134 (inf) [2021-04-15 19:23:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][1200/1251] eta 0:00:14 lr 0.000917 time 0.2709 (0.2846) loss 4.7162 (3.9380) grad_norm 1.2072 (inf) [2021-04-15 19:23:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][1210/1251] eta 0:00:11 lr 0.000917 time 0.2847 (0.2845) loss 4.8112 (3.9359) grad_norm 1.0598 (inf) [2021-04-15 19:23:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][1220/1251] eta 0:00:08 lr 0.000917 time 0.2675 (0.2845) loss 4.1101 (3.9364) grad_norm 1.2226 (inf) [2021-04-15 19:23:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][1230/1251] eta 0:00:05 lr 0.000917 time 0.2957 (0.2845) loss 2.8528 (3.9362) grad_norm 1.3151 (inf) [2021-04-15 19:23:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][1240/1251] eta 0:00:03 lr 0.000917 time 0.2635 (0.2844) loss 4.1848 (3.9352) grad_norm 1.1908 (inf) [2021-04-15 19:23:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [55/300][1250/1251] eta 0:00:00 lr 0.000917 time 0.2488 (0.2841) loss 4.1873 (3.9340) grad_norm 0.9755 (inf) [2021-04-15 19:23:39 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 55 training takes 0:05:57 [2021-04-15 19:23:39 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_55.pth saving...... [2021-04-15 19:23:49 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_55.pth saved !!! [2021-04-15 19:23:51 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.084 (1.084) Loss 1.2922 (1.2922) Acc@1 71.484 (71.484) Acc@5 90.625 (90.625) [2021-04-15 19:23:53 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.184 (0.292) Loss 1.3936 (1.3121) Acc@1 69.336 (70.890) Acc@5 88.184 (90.225) [2021-04-15 19:23:54 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.424 (0.241) Loss 1.2823 (1.3184) Acc@1 70.215 (70.587) Acc@5 90.723 (90.118) [2021-04-15 19:23:57 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.116 (0.232) Loss 1.3505 (1.3188) Acc@1 68.945 (70.407) Acc@5 89.844 (90.118) [2021-04-15 19:23:59 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.075 (0.223) Loss 1.3606 (1.3137) Acc@1 70.215 (70.555) Acc@5 89.746 (90.225) [2021-04-15 19:24:01 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 70.632 Acc@5 90.330 [2021-04-15 19:24:01 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 70.6% [2021-04-15 19:24:01 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 70.63% [2021-04-15 19:24:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][0/1251] eta 1:36:23 lr 0.000917 time 4.6228 (4.6228) loss 4.2422 (4.2422) grad_norm 0.9491 (0.9491) [2021-04-15 19:24:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][10/1251] eta 0:14:02 lr 0.000917 time 0.2547 (0.6792) loss 2.8845 (3.9537) grad_norm 1.1492 (1.0955) [2021-04-15 19:24:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][20/1251] eta 0:10:03 lr 0.000917 time 0.2801 (0.4905) loss 3.9109 (3.9394) grad_norm 1.0789 (1.0840) [2021-04-15 19:24:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][30/1251] eta 0:08:39 lr 0.000917 time 0.3097 (0.4258) loss 4.0270 (3.7236) grad_norm 1.1842 (1.1036) [2021-04-15 19:24:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][40/1251] eta 0:07:52 lr 0.000917 time 0.2663 (0.3901) loss 4.5902 (3.7907) grad_norm 1.1474 (1.0967) [2021-04-15 19:24:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][50/1251] eta 0:07:23 lr 0.000917 time 0.2678 (0.3695) loss 4.2557 (3.8261) grad_norm 1.1930 (1.0974) [2021-04-15 19:24:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][60/1251] eta 0:07:03 lr 0.000917 time 0.3168 (0.3556) loss 3.6689 (3.8410) grad_norm 1.3683 (1.1051) [2021-04-15 19:24:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][70/1251] eta 0:06:47 lr 0.000917 time 0.2561 (0.3454) loss 4.1553 (3.8754) grad_norm 1.2467 (1.1103) [2021-04-15 19:24:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][80/1251] eta 0:06:34 lr 0.000917 time 0.2606 (0.3372) loss 4.0259 (3.8783) grad_norm 0.9934 (1.1051) [2021-04-15 19:24:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][90/1251] eta 0:06:24 lr 0.000917 time 0.2625 (0.3310) loss 5.0810 (3.8754) grad_norm 1.0955 (1.1036) [2021-04-15 19:24:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][100/1251] eta 0:06:15 lr 0.000917 time 0.2846 (0.3260) loss 3.9191 (3.8958) grad_norm 0.9988 (1.1040) [2021-04-15 19:24:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][110/1251] eta 0:06:08 lr 0.000917 time 0.2606 (0.3226) loss 4.2730 (3.9044) grad_norm 1.0170 (1.1016) [2021-04-15 19:24:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][120/1251] eta 0:06:00 lr 0.000917 time 0.2792 (0.3190) loss 4.1332 (3.9056) grad_norm 1.4960 (1.1038) [2021-04-15 19:24:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][130/1251] eta 0:05:55 lr 0.000917 time 0.2790 (0.3169) loss 4.0320 (3.9167) grad_norm 1.2399 (1.1102) [2021-04-15 19:24:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][140/1251] eta 0:05:51 lr 0.000917 time 0.2784 (0.3166) loss 3.2665 (3.9237) grad_norm 1.1194 (1.1106) [2021-04-15 19:24:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][150/1251] eta 0:05:46 lr 0.000917 time 0.3377 (0.3147) loss 4.1720 (3.9314) grad_norm 0.8848 (1.1077) [2021-04-15 19:24:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][160/1251] eta 0:05:41 lr 0.000917 time 0.2701 (0.3128) loss 3.0424 (3.9230) grad_norm 1.1950 (1.1071) [2021-04-15 19:24:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][170/1251] eta 0:05:36 lr 0.000917 time 0.2848 (0.3112) loss 4.0403 (3.9181) grad_norm 0.9428 (1.1077) [2021-04-15 19:24:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][180/1251] eta 0:05:31 lr 0.000917 time 0.2453 (0.3094) loss 3.8324 (3.9241) grad_norm 0.9251 (1.1077) [2021-04-15 19:25:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][190/1251] eta 0:05:26 lr 0.000917 time 0.2780 (0.3080) loss 3.1082 (3.9008) grad_norm 0.9873 (1.1073) [2021-04-15 19:25:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][200/1251] eta 0:05:23 lr 0.000917 time 0.2962 (0.3075) loss 4.3065 (3.8964) grad_norm 1.0061 (1.1060) [2021-04-15 19:25:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][210/1251] eta 0:05:18 lr 0.000917 time 0.2865 (0.3061) loss 3.8057 (3.9045) grad_norm 1.1461 (1.1062) [2021-04-15 19:25:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][220/1251] eta 0:05:14 lr 0.000917 time 0.2712 (0.3054) loss 4.4999 (3.9031) grad_norm 1.0629 (1.1099) [2021-04-15 19:25:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][230/1251] eta 0:05:11 lr 0.000917 time 0.2936 (0.3049) loss 4.2009 (3.9074) grad_norm 1.0593 (1.1095) [2021-04-15 19:25:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][240/1251] eta 0:05:07 lr 0.000917 time 0.2778 (0.3040) loss 3.3771 (3.9178) grad_norm 0.9514 (1.1084) [2021-04-15 19:25:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][250/1251] eta 0:05:04 lr 0.000917 time 0.2828 (0.3038) loss 4.6520 (3.9271) grad_norm 1.0571 (1.1117) [2021-04-15 19:25:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][260/1251] eta 0:05:00 lr 0.000917 time 0.2855 (0.3031) loss 4.3519 (3.9167) grad_norm 1.1616 (1.1131) [2021-04-15 19:25:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][270/1251] eta 0:04:56 lr 0.000917 time 0.2866 (0.3024) loss 3.6684 (3.9233) grad_norm 1.1174 (1.1141) [2021-04-15 19:25:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][280/1251] eta 0:04:53 lr 0.000917 time 0.2988 (0.3019) loss 3.7484 (3.9113) grad_norm 1.0050 (1.1126) [2021-04-15 19:25:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][290/1251] eta 0:04:49 lr 0.000917 time 0.2775 (0.3012) loss 3.9068 (3.9137) grad_norm 0.9975 (1.1109) [2021-04-15 19:25:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][300/1251] eta 0:04:45 lr 0.000917 time 0.2852 (0.3005) loss 4.6053 (3.9188) grad_norm 0.9694 (1.1088) [2021-04-15 19:25:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][310/1251] eta 0:04:42 lr 0.000917 time 0.2983 (0.2999) loss 2.8176 (3.9152) grad_norm 1.0277 (1.1071) [2021-04-15 19:25:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][320/1251] eta 0:04:38 lr 0.000917 time 0.2865 (0.2993) loss 4.2606 (3.9276) grad_norm 0.9729 (1.1088) [2021-04-15 19:25:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][330/1251] eta 0:04:35 lr 0.000917 time 0.2804 (0.2989) loss 2.8685 (3.9316) grad_norm 1.0326 (1.1086) [2021-04-15 19:25:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][340/1251] eta 0:04:31 lr 0.000917 time 0.2695 (0.2983) loss 3.7961 (3.9275) grad_norm 1.1452 (1.1064) [2021-04-15 19:25:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][350/1251] eta 0:04:28 lr 0.000916 time 0.2839 (0.2978) loss 4.3438 (3.9381) grad_norm 1.0738 (1.1051) [2021-04-15 19:25:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][360/1251] eta 0:04:25 lr 0.000916 time 0.2694 (0.2976) loss 3.5571 (3.9308) grad_norm 1.0952 (1.1026) [2021-04-15 19:25:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][370/1251] eta 0:04:21 lr 0.000916 time 0.3122 (0.2973) loss 4.1159 (3.9269) grad_norm 1.1153 (1.1010) [2021-04-15 19:25:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][380/1251] eta 0:04:18 lr 0.000916 time 0.2960 (0.2970) loss 4.3206 (3.9309) grad_norm 1.4262 (1.1035) [2021-04-15 19:25:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][390/1251] eta 0:04:15 lr 0.000916 time 0.2972 (0.2966) loss 4.5402 (3.9333) grad_norm 1.0196 (1.1052) [2021-04-15 19:26:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][400/1251] eta 0:04:12 lr 0.000916 time 0.2795 (0.2965) loss 4.3778 (3.9284) grad_norm 0.9390 (1.1053) [2021-04-15 19:26:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][410/1251] eta 0:04:09 lr 0.000916 time 0.2899 (0.2963) loss 4.3486 (3.9201) grad_norm 0.9941 (1.1056) [2021-04-15 19:26:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][420/1251] eta 0:04:06 lr 0.000916 time 0.3020 (0.2960) loss 3.5048 (3.9172) grad_norm 1.1221 (1.1051) [2021-04-15 19:26:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][430/1251] eta 0:04:03 lr 0.000916 time 0.2927 (0.2961) loss 2.7021 (3.9139) grad_norm 1.1495 (1.1042) [2021-04-15 19:26:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][440/1251] eta 0:04:00 lr 0.000916 time 0.2768 (0.2961) loss 4.3399 (3.9117) grad_norm 1.3678 (1.1043) [2021-04-15 19:26:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][450/1251] eta 0:03:57 lr 0.000916 time 0.4343 (0.2962) loss 3.8459 (3.9154) grad_norm 0.8894 (1.1040) [2021-04-15 19:26:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][460/1251] eta 0:03:54 lr 0.000916 time 0.2867 (0.2959) loss 4.1300 (3.9181) grad_norm 0.9295 (1.1023) [2021-04-15 19:26:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][470/1251] eta 0:03:50 lr 0.000916 time 0.2943 (0.2956) loss 2.6632 (3.9130) grad_norm 1.0019 (1.1015) [2021-04-15 19:26:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][480/1251] eta 0:03:47 lr 0.000916 time 0.2857 (0.2954) loss 3.9234 (3.9082) grad_norm 1.0320 (1.1019) [2021-04-15 19:26:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][490/1251] eta 0:03:44 lr 0.000916 time 0.2769 (0.2952) loss 4.2294 (3.9083) grad_norm 1.1539 (1.1014) [2021-04-15 19:26:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][500/1251] eta 0:03:41 lr 0.000916 time 0.2775 (0.2950) loss 2.6868 (3.9086) grad_norm 1.0855 (1.1009) [2021-04-15 19:26:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][510/1251] eta 0:03:38 lr 0.000916 time 0.2679 (0.2947) loss 4.0702 (3.9095) grad_norm 1.1340 (1.0996) [2021-04-15 19:26:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][520/1251] eta 0:03:35 lr 0.000916 time 0.2880 (0.2946) loss 3.7540 (3.9091) grad_norm 1.1016 (1.1014) [2021-04-15 19:26:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][530/1251] eta 0:03:32 lr 0.000916 time 0.2803 (0.2943) loss 4.2348 (3.9077) grad_norm 1.1800 (1.1018) [2021-04-15 19:26:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][540/1251] eta 0:03:29 lr 0.000916 time 0.2744 (0.2941) loss 2.9767 (3.9040) grad_norm 1.1490 (1.1012) [2021-04-15 19:26:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][550/1251] eta 0:03:25 lr 0.000916 time 0.2573 (0.2938) loss 3.1168 (3.9035) grad_norm 1.0516 (1.1014) [2021-04-15 19:26:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][560/1251] eta 0:03:22 lr 0.000916 time 0.2678 (0.2936) loss 4.5457 (3.9030) grad_norm 1.1351 (1.1012) [2021-04-15 19:26:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][570/1251] eta 0:03:19 lr 0.000916 time 0.2478 (0.2936) loss 3.9199 (3.9032) grad_norm 0.9168 (1.1003) [2021-04-15 19:26:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][580/1251] eta 0:03:16 lr 0.000916 time 0.3101 (0.2934) loss 4.0703 (3.9029) grad_norm 0.8995 (1.1001) [2021-04-15 19:26:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][590/1251] eta 0:03:13 lr 0.000916 time 0.2919 (0.2932) loss 3.7880 (3.8999) grad_norm 1.1329 (1.1003) [2021-04-15 19:26:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][600/1251] eta 0:03:10 lr 0.000916 time 0.2708 (0.2930) loss 4.2275 (3.9060) grad_norm 1.1770 (1.1002) [2021-04-15 19:27:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][610/1251] eta 0:03:07 lr 0.000916 time 0.2987 (0.2929) loss 2.7487 (3.9097) grad_norm 1.1048 (1.0995) [2021-04-15 19:27:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][620/1251] eta 0:03:04 lr 0.000916 time 0.2962 (0.2927) loss 4.0391 (3.9118) grad_norm 1.1854 (1.1011) [2021-04-15 19:27:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][630/1251] eta 0:03:01 lr 0.000916 time 0.2694 (0.2925) loss 4.3206 (3.9123) grad_norm 0.8944 (1.1004) [2021-04-15 19:27:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][640/1251] eta 0:02:58 lr 0.000916 time 0.3079 (0.2923) loss 3.6022 (3.9161) grad_norm 1.1343 (1.0995) [2021-04-15 19:27:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][650/1251] eta 0:02:55 lr 0.000916 time 0.2940 (0.2922) loss 3.9094 (3.9171) grad_norm 1.1658 (1.0991) [2021-04-15 19:27:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][660/1251] eta 0:02:52 lr 0.000916 time 0.2773 (0.2920) loss 4.5921 (3.9163) grad_norm 1.3248 (1.0989) [2021-04-15 19:27:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][670/1251] eta 0:02:49 lr 0.000916 time 0.2749 (0.2918) loss 4.3259 (3.9189) grad_norm 1.0058 (1.0985) [2021-04-15 19:27:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][680/1251] eta 0:02:46 lr 0.000916 time 0.2805 (0.2917) loss 4.3698 (3.9176) grad_norm 1.0285 (1.0989) [2021-04-15 19:27:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][690/1251] eta 0:02:43 lr 0.000916 time 0.2594 (0.2916) loss 4.4607 (3.9176) grad_norm 1.0432 (1.0993) [2021-04-15 19:27:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][700/1251] eta 0:02:40 lr 0.000916 time 0.2728 (0.2913) loss 4.2177 (3.9160) grad_norm 1.1783 (1.0994) [2021-04-15 19:27:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][710/1251] eta 0:02:37 lr 0.000916 time 0.2767 (0.2914) loss 3.7220 (3.9175) grad_norm 1.1751 (1.0991) [2021-04-15 19:27:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][720/1251] eta 0:02:34 lr 0.000916 time 0.2708 (0.2916) loss 4.6224 (3.9179) grad_norm 1.0613 (1.1001) [2021-04-15 19:27:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][730/1251] eta 0:02:31 lr 0.000916 time 0.2952 (0.2915) loss 3.6808 (3.9172) grad_norm 1.0288 (1.0987) [2021-04-15 19:27:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][740/1251] eta 0:02:28 lr 0.000916 time 0.2807 (0.2913) loss 2.5782 (3.9161) grad_norm 0.9214 (1.0982) [2021-04-15 19:27:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][750/1251] eta 0:02:25 lr 0.000916 time 0.2789 (0.2912) loss 4.1762 (3.9188) grad_norm 1.0403 (1.0988) [2021-04-15 19:27:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][760/1251] eta 0:02:22 lr 0.000916 time 0.2810 (0.2910) loss 4.0004 (3.9142) grad_norm 1.1458 (1.0992) [2021-04-15 19:27:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][770/1251] eta 0:02:19 lr 0.000916 time 0.3074 (0.2910) loss 4.2987 (3.9144) grad_norm 1.1141 (1.0991) [2021-04-15 19:27:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][780/1251] eta 0:02:17 lr 0.000915 time 0.2788 (0.2910) loss 4.1502 (3.9124) grad_norm 1.1859 (1.0997) [2021-04-15 19:27:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][790/1251] eta 0:02:14 lr 0.000915 time 0.2671 (0.2910) loss 3.9829 (3.9150) grad_norm 1.3077 (1.1010) [2021-04-15 19:27:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][800/1251] eta 0:02:11 lr 0.000915 time 0.4171 (0.2910) loss 3.0002 (3.9127) grad_norm 1.1066 (1.1014) [2021-04-15 19:27:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][810/1251] eta 0:02:08 lr 0.000915 time 0.2630 (0.2909) loss 3.7001 (3.9139) grad_norm 1.1371 (1.1022) [2021-04-15 19:28:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][820/1251] eta 0:02:05 lr 0.000915 time 0.2681 (0.2907) loss 3.9314 (3.9140) grad_norm 0.9803 (1.1019) [2021-04-15 19:28:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][830/1251] eta 0:02:02 lr 0.000915 time 0.2818 (0.2907) loss 3.8977 (3.9167) grad_norm 1.0509 (1.1019) [2021-04-15 19:28:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][840/1251] eta 0:01:59 lr 0.000915 time 0.2775 (0.2905) loss 3.5315 (3.9198) grad_norm 1.4360 (1.1020) [2021-04-15 19:28:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][850/1251] eta 0:01:56 lr 0.000915 time 0.2604 (0.2904) loss 4.1223 (3.9223) grad_norm 0.9666 (1.1019) [2021-04-15 19:28:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][860/1251] eta 0:01:53 lr 0.000915 time 0.2661 (0.2903) loss 3.9905 (3.9242) grad_norm 1.1421 (1.1026) [2021-04-15 19:28:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][870/1251] eta 0:01:50 lr 0.000915 time 0.2647 (0.2901) loss 4.4947 (3.9269) grad_norm 1.0872 (1.1025) [2021-04-15 19:28:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][880/1251] eta 0:01:47 lr 0.000915 time 0.2626 (0.2900) loss 3.8566 (3.9272) grad_norm 1.2851 (1.1033) [2021-04-15 19:28:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][890/1251] eta 0:01:44 lr 0.000915 time 0.2770 (0.2899) loss 4.0768 (3.9269) grad_norm 1.2030 (1.1037) [2021-04-15 19:28:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][900/1251] eta 0:01:41 lr 0.000915 time 0.2958 (0.2898) loss 4.0859 (3.9262) grad_norm 1.0054 (1.1040) [2021-04-15 19:28:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][910/1251] eta 0:01:38 lr 0.000915 time 0.3104 (0.2897) loss 3.0956 (3.9240) grad_norm 1.2723 (1.1043) [2021-04-15 19:28:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][920/1251] eta 0:01:35 lr 0.000915 time 0.2917 (0.2896) loss 4.2229 (3.9231) grad_norm 1.2796 (1.1045) [2021-04-15 19:28:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][930/1251] eta 0:01:32 lr 0.000915 time 0.3063 (0.2895) loss 3.9207 (3.9253) grad_norm 0.9631 (1.1040) [2021-04-15 19:28:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][940/1251] eta 0:01:30 lr 0.000915 time 0.3003 (0.2896) loss 4.7198 (3.9292) grad_norm 1.2859 (1.1034) [2021-04-15 19:28:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][950/1251] eta 0:01:27 lr 0.000915 time 0.2883 (0.2895) loss 4.6584 (3.9292) grad_norm 0.9978 (1.1027) [2021-04-15 19:28:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][960/1251] eta 0:01:24 lr 0.000915 time 0.2807 (0.2894) loss 4.1372 (3.9319) grad_norm 1.0382 (1.1025) [2021-04-15 19:28:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][970/1251] eta 0:01:21 lr 0.000915 time 0.2575 (0.2892) loss 4.0141 (3.9310) grad_norm 1.3729 (1.1036) [2021-04-15 19:28:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][980/1251] eta 0:01:18 lr 0.000915 time 0.2914 (0.2892) loss 2.9777 (3.9296) grad_norm 1.1401 (1.1034) [2021-04-15 19:28:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][990/1251] eta 0:01:15 lr 0.000915 time 0.2695 (0.2891) loss 3.5424 (3.9278) grad_norm 1.3948 (1.1039) [2021-04-15 19:28:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][1000/1251] eta 0:01:12 lr 0.000915 time 0.2963 (0.2890) loss 3.5843 (3.9271) grad_norm 0.8795 (1.1039) [2021-04-15 19:28:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][1010/1251] eta 0:01:09 lr 0.000915 time 0.2711 (0.2890) loss 4.7056 (3.9285) grad_norm 1.1410 (1.1040) [2021-04-15 19:28:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][1020/1251] eta 0:01:06 lr 0.000915 time 0.2815 (0.2890) loss 4.7169 (3.9305) grad_norm 1.2982 (1.1050) [2021-04-15 19:28:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][1030/1251] eta 0:01:03 lr 0.000915 time 0.2884 (0.2891) loss 4.0849 (3.9288) grad_norm 1.3426 (1.1061) [2021-04-15 19:29:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][1040/1251] eta 0:01:00 lr 0.000915 time 0.2628 (0.2890) loss 4.6093 (3.9298) grad_norm 1.0386 (1.1066) [2021-04-15 19:29:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][1050/1251] eta 0:00:58 lr 0.000915 time 0.2744 (0.2889) loss 2.7664 (3.9266) grad_norm 1.1526 (1.1070) [2021-04-15 19:29:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][1060/1251] eta 0:00:55 lr 0.000915 time 0.2807 (0.2888) loss 3.1333 (3.9240) grad_norm 1.1798 (1.1074) [2021-04-15 19:29:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][1070/1251] eta 0:00:52 lr 0.000915 time 0.2955 (0.2887) loss 3.5205 (3.9239) grad_norm 1.0256 (1.1071) [2021-04-15 19:29:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][1080/1251] eta 0:00:49 lr 0.000915 time 0.2759 (0.2886) loss 4.6583 (3.9271) grad_norm 1.0124 (1.1066) [2021-04-15 19:29:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][1090/1251] eta 0:00:46 lr 0.000915 time 0.2424 (0.2886) loss 4.1495 (3.9271) grad_norm 1.1534 (1.1065) [2021-04-15 19:29:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][1100/1251] eta 0:00:43 lr 0.000915 time 0.2723 (0.2886) loss 3.5764 (3.9299) grad_norm 1.2336 (1.1063) [2021-04-15 19:29:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][1110/1251] eta 0:00:40 lr 0.000915 time 0.3023 (0.2885) loss 3.1316 (3.9290) grad_norm 1.2932 (1.1059) [2021-04-15 19:29:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][1120/1251] eta 0:00:37 lr 0.000915 time 0.2738 (0.2885) loss 3.6867 (3.9318) grad_norm 1.1140 (1.1057) [2021-04-15 19:29:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][1130/1251] eta 0:00:34 lr 0.000915 time 0.2717 (0.2884) loss 4.3194 (3.9322) grad_norm 0.9242 (1.1048) [2021-04-15 19:29:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][1140/1251] eta 0:00:32 lr 0.000915 time 0.2747 (0.2883) loss 4.1989 (3.9319) grad_norm 1.1022 (1.1045) [2021-04-15 19:29:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][1150/1251] eta 0:00:29 lr 0.000915 time 0.2814 (0.2884) loss 3.7364 (3.9340) grad_norm 1.1734 (1.1033) [2021-04-15 19:29:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][1160/1251] eta 0:00:26 lr 0.000915 time 0.2921 (0.2883) loss 4.3541 (3.9346) grad_norm 1.1109 (1.1035) [2021-04-15 19:29:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][1170/1251] eta 0:00:23 lr 0.000915 time 0.2839 (0.2882) loss 4.0307 (3.9357) grad_norm 0.9635 (1.1041) [2021-04-15 19:29:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][1180/1251] eta 0:00:20 lr 0.000915 time 0.2918 (0.2882) loss 4.5067 (3.9377) grad_norm 1.2557 (1.1038) [2021-04-15 19:29:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][1190/1251] eta 0:00:17 lr 0.000915 time 0.2812 (0.2881) loss 4.1510 (3.9362) grad_norm 1.2588 (1.1043) [2021-04-15 19:29:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][1200/1251] eta 0:00:14 lr 0.000915 time 0.3045 (0.2881) loss 3.2997 (3.9345) grad_norm 1.0642 (1.1051) [2021-04-15 19:29:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][1210/1251] eta 0:00:11 lr 0.000915 time 0.2734 (0.2880) loss 2.9326 (3.9339) grad_norm 0.9621 (1.1047) [2021-04-15 19:29:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][1220/1251] eta 0:00:08 lr 0.000914 time 0.2770 (0.2879) loss 4.5528 (3.9334) grad_norm 0.9315 (1.1046) [2021-04-15 19:29:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][1230/1251] eta 0:00:06 lr 0.000914 time 0.2697 (0.2879) loss 3.6457 (3.9331) grad_norm 1.0744 (1.1045) [2021-04-15 19:29:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][1240/1251] eta 0:00:03 lr 0.000914 time 0.2478 (0.2877) loss 3.7904 (3.9338) grad_norm 1.2194 (1.1043) [2021-04-15 19:30:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [56/300][1250/1251] eta 0:00:00 lr 0.000914 time 0.2484 (0.2874) loss 4.3280 (3.9342) grad_norm 1.0926 (1.1041) [2021-04-15 19:30:03 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 56 training takes 0:06:01 [2021-04-15 19:30:03 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_56.pth saving...... [2021-04-15 19:30:20 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_56.pth saved !!! [2021-04-15 19:30:21 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.230 (1.230) Loss 1.2653 (1.2653) Acc@1 72.266 (72.266) Acc@5 90.137 (90.137) [2021-04-15 19:30:23 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.148 (0.266) Loss 1.3396 (1.3001) Acc@1 70.605 (70.605) Acc@5 89.160 (90.270) [2021-04-15 19:30:25 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.096 (0.239) Loss 1.2321 (1.2929) Acc@1 70.410 (70.550) Acc@5 91.895 (90.392) [2021-04-15 19:30:28 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.281 (0.256) Loss 1.3279 (1.2965) Acc@1 70.605 (70.435) Acc@5 90.234 (90.515) [2021-04-15 19:30:29 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.216) Loss 1.2627 (1.2883) Acc@1 70.312 (70.529) Acc@5 90.918 (90.623) [2021-04-15 19:30:31 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 70.668 Acc@5 90.580 [2021-04-15 19:30:31 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 70.7% [2021-04-15 19:30:31 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 70.67% [2021-04-15 19:30:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][0/1251] eta 2:44:43 lr 0.000914 time 7.9002 (7.9002) loss 3.9230 (3.9230) grad_norm 1.3315 (1.3315) [2021-04-15 19:30:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][10/1251] eta 0:20:16 lr 0.000914 time 0.4491 (0.9806) loss 4.0936 (4.0377) grad_norm 1.1337 (1.2358) [2021-04-15 19:30:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][20/1251] eta 0:13:15 lr 0.000914 time 0.2825 (0.6465) loss 4.5447 (4.0414) grad_norm 1.2485 (1.1967) [2021-04-15 19:30:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][30/1251] eta 0:10:48 lr 0.000914 time 0.2893 (0.5315) loss 4.8523 (4.0304) grad_norm 1.0142 (1.1684) [2021-04-15 19:30:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][40/1251] eta 0:09:30 lr 0.000914 time 0.2900 (0.4708) loss 3.9640 (4.0834) grad_norm 1.0635 (1.1353) [2021-04-15 19:30:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][50/1251] eta 0:08:41 lr 0.000914 time 0.2920 (0.4342) loss 3.3756 (4.0506) grad_norm 1.1171 (1.1403) [2021-04-15 19:30:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][60/1251] eta 0:08:06 lr 0.000914 time 0.2921 (0.4082) loss 4.0960 (4.0178) grad_norm 1.1007 (1.1238) [2021-04-15 19:30:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][70/1251] eta 0:07:40 lr 0.000914 time 0.2938 (0.3898) loss 4.5944 (4.0131) grad_norm 1.1775 (1.1382) [2021-04-15 19:31:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][80/1251] eta 0:07:21 lr 0.000914 time 0.2924 (0.3767) loss 3.6702 (4.0050) grad_norm 1.0354 (1.1520) [2021-04-15 19:31:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][90/1251] eta 0:07:05 lr 0.000914 time 0.2772 (0.3666) loss 4.0152 (3.9806) grad_norm 1.1924 (1.1462) [2021-04-15 19:31:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][100/1251] eta 0:06:51 lr 0.000914 time 0.2902 (0.3575) loss 4.5421 (3.9957) grad_norm 1.2285 (1.1463) [2021-04-15 19:31:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][110/1251] eta 0:06:39 lr 0.000914 time 0.2577 (0.3500) loss 4.2842 (3.9619) grad_norm 0.9212 (1.1372) [2021-04-15 19:31:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][120/1251] eta 0:06:28 lr 0.000914 time 0.2575 (0.3437) loss 3.0726 (3.9200) grad_norm 0.9831 (1.1297) [2021-04-15 19:31:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][130/1251] eta 0:06:20 lr 0.000914 time 0.2638 (0.3392) loss 2.5156 (3.8966) grad_norm 1.1344 (1.1304) [2021-04-15 19:31:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][140/1251] eta 0:06:13 lr 0.000914 time 0.2967 (0.3360) loss 2.5049 (3.8793) grad_norm 1.1891 (1.1320) [2021-04-15 19:31:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][150/1251] eta 0:06:05 lr 0.000914 time 0.2784 (0.3323) loss 4.1185 (3.8782) grad_norm 0.9916 (1.1316) [2021-04-15 19:31:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][160/1251] eta 0:05:58 lr 0.000914 time 0.2634 (0.3288) loss 4.9687 (3.8936) grad_norm 0.9756 (1.1294) [2021-04-15 19:31:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][170/1251] eta 0:05:52 lr 0.000914 time 0.2784 (0.3264) loss 3.7850 (3.8834) grad_norm 0.9898 (1.1268) [2021-04-15 19:31:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][180/1251] eta 0:05:46 lr 0.000914 time 0.2794 (0.3236) loss 3.8567 (3.9069) grad_norm 1.0048 (1.1236) [2021-04-15 19:31:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][190/1251] eta 0:05:40 lr 0.000914 time 0.2808 (0.3214) loss 4.3201 (3.9036) grad_norm 0.9757 (1.1239) [2021-04-15 19:31:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][200/1251] eta 0:05:35 lr 0.000914 time 0.2754 (0.3196) loss 2.8769 (3.8981) grad_norm 1.0801 (1.1219) [2021-04-15 19:31:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][210/1251] eta 0:05:30 lr 0.000914 time 0.2781 (0.3176) loss 3.8023 (3.9014) grad_norm 1.0152 (1.1225) [2021-04-15 19:31:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][220/1251] eta 0:05:26 lr 0.000914 time 0.2746 (0.3163) loss 3.5066 (3.9117) grad_norm 1.0275 (1.1216) [2021-04-15 19:31:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][230/1251] eta 0:05:21 lr 0.000914 time 0.2822 (0.3151) loss 3.3564 (3.9023) grad_norm 1.0715 (1.1211) [2021-04-15 19:31:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][240/1251] eta 0:05:17 lr 0.000914 time 0.2973 (0.3136) loss 3.4695 (3.8890) grad_norm 1.0368 (1.1204) [2021-04-15 19:31:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][250/1251] eta 0:05:13 lr 0.000914 time 0.2694 (0.3136) loss 4.1047 (3.8891) grad_norm 0.9449 (1.1178) [2021-04-15 19:31:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][260/1251] eta 0:05:09 lr 0.000914 time 0.2788 (0.3124) loss 3.2492 (3.8830) grad_norm 0.9915 (1.1146) [2021-04-15 19:31:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][270/1251] eta 0:05:05 lr 0.000914 time 0.2876 (0.3111) loss 3.7196 (3.8876) grad_norm 0.9106 (1.1136) [2021-04-15 19:31:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][280/1251] eta 0:05:00 lr 0.000914 time 0.2596 (0.3099) loss 4.2687 (3.8794) grad_norm 1.0013 (1.1144) [2021-04-15 19:32:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][290/1251] eta 0:04:56 lr 0.000914 time 0.2921 (0.3089) loss 4.5248 (3.8774) grad_norm 0.9664 (1.1125) [2021-04-15 19:32:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][300/1251] eta 0:04:53 lr 0.000914 time 0.2683 (0.3083) loss 4.1411 (3.8716) grad_norm 1.0155 (1.1127) [2021-04-15 19:32:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][310/1251] eta 0:04:49 lr 0.000914 time 0.2809 (0.3075) loss 4.4315 (3.8712) grad_norm 1.0736 (1.1133) [2021-04-15 19:32:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][320/1251] eta 0:04:45 lr 0.000914 time 0.2737 (0.3064) loss 4.3301 (3.8725) grad_norm 1.1901 (1.1150) [2021-04-15 19:32:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][330/1251] eta 0:04:41 lr 0.000914 time 0.3037 (0.3057) loss 3.6791 (3.8736) grad_norm 1.3924 (1.1170) [2021-04-15 19:32:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][340/1251] eta 0:04:37 lr 0.000914 time 0.2787 (0.3049) loss 4.5853 (3.8814) grad_norm 1.0417 (1.1151) [2021-04-15 19:32:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][350/1251] eta 0:04:33 lr 0.000914 time 0.2698 (0.3041) loss 3.9993 (3.8841) grad_norm 0.9964 (1.1117) [2021-04-15 19:32:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][360/1251] eta 0:04:30 lr 0.000914 time 0.3008 (0.3037) loss 4.5131 (3.8772) grad_norm 1.1682 (1.1101) [2021-04-15 19:32:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][370/1251] eta 0:04:26 lr 0.000914 time 0.2499 (0.3029) loss 4.4989 (3.8712) grad_norm 1.2922 (1.1111) [2021-04-15 19:32:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][380/1251] eta 0:04:23 lr 0.000914 time 0.2676 (0.3022) loss 4.1290 (3.8730) grad_norm 1.0532 (1.1132) [2021-04-15 19:32:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][390/1251] eta 0:04:19 lr 0.000913 time 0.2800 (0.3016) loss 3.7377 (3.8756) grad_norm 1.3847 (1.1190) [2021-04-15 19:32:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][400/1251] eta 0:04:16 lr 0.000913 time 0.2469 (0.3010) loss 3.7578 (3.8830) grad_norm 1.0079 (1.1177) [2021-04-15 19:32:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][410/1251] eta 0:04:12 lr 0.000913 time 0.2789 (0.3004) loss 4.5095 (3.8788) grad_norm 0.9494 (1.1158) [2021-04-15 19:32:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][420/1251] eta 0:04:09 lr 0.000913 time 0.2815 (0.2998) loss 3.1464 (3.8718) grad_norm 0.9703 (1.1166) [2021-04-15 19:32:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][430/1251] eta 0:04:05 lr 0.000913 time 0.2717 (0.2994) loss 3.8153 (3.8683) grad_norm 1.2053 (1.1179) [2021-04-15 19:32:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][440/1251] eta 0:04:02 lr 0.000913 time 0.2593 (0.2988) loss 3.7535 (3.8712) grad_norm 1.2913 (1.1191) [2021-04-15 19:32:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][450/1251] eta 0:03:58 lr 0.000913 time 0.2663 (0.2983) loss 4.1630 (3.8699) grad_norm 0.9698 (1.1191) [2021-04-15 19:32:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][460/1251] eta 0:03:55 lr 0.000913 time 0.2973 (0.2979) loss 4.4634 (3.8712) grad_norm 0.9927 (1.1188) [2021-04-15 19:32:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][470/1251] eta 0:03:52 lr 0.000913 time 0.2818 (0.2974) loss 3.4395 (3.8712) grad_norm 1.0630 (1.1175) [2021-04-15 19:32:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][480/1251] eta 0:03:48 lr 0.000913 time 0.2627 (0.2969) loss 3.9824 (3.8726) grad_norm 1.4001 (1.1176) [2021-04-15 19:32:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][490/1251] eta 0:03:45 lr 0.000913 time 0.2833 (0.2965) loss 4.1865 (3.8794) grad_norm 1.1154 (1.1175) [2021-04-15 19:32:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][500/1251] eta 0:03:42 lr 0.000913 time 0.2637 (0.2961) loss 3.5928 (3.8811) grad_norm 1.2413 (1.1187) [2021-04-15 19:33:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][510/1251] eta 0:03:39 lr 0.000913 time 0.2827 (0.2957) loss 2.9310 (3.8796) grad_norm 1.3501 (1.1186) [2021-04-15 19:33:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][520/1251] eta 0:03:36 lr 0.000913 time 0.2669 (0.2957) loss 4.1853 (3.8854) grad_norm 1.0227 (1.1187) [2021-04-15 19:33:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][530/1251] eta 0:03:32 lr 0.000913 time 0.2637 (0.2954) loss 4.3442 (3.8818) grad_norm 1.0886 (1.1177) [2021-04-15 19:33:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][540/1251] eta 0:03:29 lr 0.000913 time 0.2574 (0.2952) loss 3.6401 (3.8854) grad_norm 1.0746 (1.1165) [2021-04-15 19:33:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][550/1251] eta 0:03:26 lr 0.000913 time 0.4674 (0.2952) loss 2.7742 (3.8834) grad_norm 1.2825 (1.1163) [2021-04-15 19:33:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][560/1251] eta 0:03:23 lr 0.000913 time 0.2893 (0.2949) loss 4.2733 (3.8856) grad_norm 0.9667 (1.1159) [2021-04-15 19:33:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][570/1251] eta 0:03:20 lr 0.000913 time 0.2432 (0.2948) loss 4.6440 (3.8847) grad_norm 0.9954 (1.1153) [2021-04-15 19:33:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][580/1251] eta 0:03:17 lr 0.000913 time 0.2839 (0.2946) loss 4.3481 (3.8901) grad_norm 1.1277 (1.1166) [2021-04-15 19:33:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][590/1251] eta 0:03:14 lr 0.000913 time 0.2924 (0.2945) loss 3.9856 (3.8872) grad_norm 1.0848 (1.1178) [2021-04-15 19:33:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][600/1251] eta 0:03:11 lr 0.000913 time 0.4335 (0.2944) loss 3.8026 (3.8859) grad_norm 1.5452 (1.1203) [2021-04-15 19:33:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][610/1251] eta 0:03:08 lr 0.000913 time 0.2785 (0.2943) loss 4.3097 (3.8873) grad_norm 1.0861 (1.1209) [2021-04-15 19:33:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][620/1251] eta 0:03:05 lr 0.000913 time 0.2893 (0.2940) loss 4.6806 (3.8936) grad_norm 1.1172 (1.1209) [2021-04-15 19:33:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][630/1251] eta 0:03:02 lr 0.000913 time 0.2552 (0.2937) loss 3.8972 (3.8886) grad_norm 1.1392 (inf) [2021-04-15 19:33:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][640/1251] eta 0:02:59 lr 0.000913 time 0.2833 (0.2935) loss 3.9838 (3.8854) grad_norm 1.0646 (inf) [2021-04-15 19:33:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][650/1251] eta 0:02:56 lr 0.000913 time 0.2837 (0.2932) loss 4.4057 (3.8876) grad_norm 1.0468 (inf) [2021-04-15 19:33:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][660/1251] eta 0:02:53 lr 0.000913 time 0.2960 (0.2931) loss 4.4297 (3.8841) grad_norm 1.2614 (inf) [2021-04-15 19:33:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][670/1251] eta 0:02:50 lr 0.000913 time 0.2594 (0.2928) loss 2.6747 (3.8813) grad_norm 1.1984 (inf) [2021-04-15 19:33:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][680/1251] eta 0:02:47 lr 0.000913 time 0.2736 (0.2926) loss 4.2990 (3.8846) grad_norm 1.1419 (inf) [2021-04-15 19:33:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][690/1251] eta 0:02:43 lr 0.000913 time 0.3081 (0.2923) loss 4.0297 (3.8858) grad_norm 1.2671 (inf) [2021-04-15 19:33:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][700/1251] eta 0:02:40 lr 0.000913 time 0.2878 (0.2921) loss 3.8714 (3.8850) grad_norm 1.1573 (inf) [2021-04-15 19:33:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][710/1251] eta 0:02:38 lr 0.000913 time 0.2594 (0.2921) loss 2.9961 (3.8834) grad_norm 1.0484 (inf) [2021-04-15 19:34:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][720/1251] eta 0:02:34 lr 0.000913 time 0.3008 (0.2919) loss 2.9269 (3.8810) grad_norm 1.0416 (inf) [2021-04-15 19:34:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][730/1251] eta 0:02:31 lr 0.000913 time 0.2748 (0.2917) loss 3.6106 (3.8779) grad_norm 1.1549 (inf) [2021-04-15 19:34:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][740/1251] eta 0:02:28 lr 0.000913 time 0.2911 (0.2915) loss 3.2770 (3.8808) grad_norm 1.0775 (inf) [2021-04-15 19:34:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][750/1251] eta 0:02:25 lr 0.000913 time 0.2785 (0.2913) loss 3.4222 (3.8814) grad_norm 1.1740 (inf) [2021-04-15 19:34:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][760/1251] eta 0:02:22 lr 0.000913 time 0.2852 (0.2912) loss 4.1423 (3.8827) grad_norm 1.2180 (inf) [2021-04-15 19:34:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][770/1251] eta 0:02:19 lr 0.000913 time 0.2560 (0.2910) loss 4.0218 (3.8814) grad_norm 1.3587 (inf) [2021-04-15 19:34:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][780/1251] eta 0:02:16 lr 0.000913 time 0.2993 (0.2909) loss 3.2247 (3.8769) grad_norm 0.9579 (inf) [2021-04-15 19:34:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][790/1251] eta 0:02:14 lr 0.000913 time 0.2831 (0.2908) loss 3.0902 (3.8760) grad_norm 1.2585 (inf) [2021-04-15 19:34:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][800/1251] eta 0:02:11 lr 0.000913 time 0.2656 (0.2906) loss 3.7674 (3.8765) grad_norm 1.4775 (inf) [2021-04-15 19:34:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][810/1251] eta 0:02:08 lr 0.000913 time 0.2955 (0.2906) loss 4.2602 (3.8747) grad_norm 1.2419 (inf) [2021-04-15 19:34:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][820/1251] eta 0:02:05 lr 0.000912 time 0.2691 (0.2905) loss 4.3920 (3.8775) grad_norm 0.9650 (inf) [2021-04-15 19:34:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][830/1251] eta 0:02:02 lr 0.000912 time 0.2541 (0.2904) loss 3.8216 (3.8790) grad_norm 1.0011 (inf) [2021-04-15 19:34:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][840/1251] eta 0:01:59 lr 0.000912 time 0.2699 (0.2903) loss 3.3022 (3.8758) grad_norm 1.2630 (inf) [2021-04-15 19:34:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][850/1251] eta 0:01:56 lr 0.000912 time 0.2733 (0.2901) loss 3.6957 (3.8752) grad_norm 1.0620 (inf) [2021-04-15 19:34:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][860/1251] eta 0:01:53 lr 0.000912 time 0.2609 (0.2899) loss 3.7135 (3.8751) grad_norm 1.0193 (inf) [2021-04-15 19:34:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][870/1251] eta 0:01:50 lr 0.000912 time 0.2945 (0.2898) loss 4.3132 (3.8769) grad_norm 1.2174 (inf) [2021-04-15 19:34:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][880/1251] eta 0:01:47 lr 0.000912 time 0.2765 (0.2897) loss 3.9028 (3.8782) grad_norm 1.1625 (inf) [2021-04-15 19:34:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][890/1251] eta 0:01:44 lr 0.000912 time 0.2958 (0.2896) loss 3.5429 (3.8772) grad_norm 0.9353 (inf) [2021-04-15 19:34:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][900/1251] eta 0:01:41 lr 0.000912 time 0.2498 (0.2895) loss 4.8018 (3.8764) grad_norm 1.1132 (inf) [2021-04-15 19:34:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][910/1251] eta 0:01:38 lr 0.000912 time 0.2942 (0.2894) loss 3.6948 (3.8775) grad_norm 1.1693 (inf) [2021-04-15 19:34:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][920/1251] eta 0:01:35 lr 0.000912 time 0.2930 (0.2892) loss 4.6651 (3.8742) grad_norm 1.2442 (inf) [2021-04-15 19:35:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][930/1251] eta 0:01:32 lr 0.000912 time 0.2786 (0.2891) loss 3.8054 (3.8751) grad_norm 1.2449 (inf) [2021-04-15 19:35:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][940/1251] eta 0:01:29 lr 0.000912 time 0.2883 (0.2891) loss 4.2173 (3.8731) grad_norm 0.9719 (inf) [2021-04-15 19:35:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][950/1251] eta 0:01:26 lr 0.000912 time 0.2635 (0.2890) loss 4.9035 (3.8766) grad_norm 1.1528 (inf) [2021-04-15 19:35:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][960/1251] eta 0:01:24 lr 0.000912 time 0.2716 (0.2889) loss 4.1086 (3.8769) grad_norm 1.1465 (inf) [2021-04-15 19:35:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][970/1251] eta 0:01:21 lr 0.000912 time 0.2699 (0.2887) loss 4.2434 (3.8775) grad_norm 1.1051 (inf) [2021-04-15 19:35:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][980/1251] eta 0:01:18 lr 0.000912 time 0.2758 (0.2886) loss 3.8540 (3.8784) grad_norm 1.3086 (inf) [2021-04-15 19:35:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][990/1251] eta 0:01:15 lr 0.000912 time 0.2877 (0.2885) loss 4.4757 (3.8816) grad_norm 1.1276 (inf) [2021-04-15 19:35:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][1000/1251] eta 0:01:12 lr 0.000912 time 0.2867 (0.2884) loss 3.0923 (3.8806) grad_norm 0.9790 (inf) [2021-04-15 19:35:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][1010/1251] eta 0:01:09 lr 0.000912 time 0.2564 (0.2883) loss 4.8561 (3.8832) grad_norm 1.0784 (inf) [2021-04-15 19:35:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][1020/1251] eta 0:01:06 lr 0.000912 time 0.2765 (0.2881) loss 4.0434 (3.8825) grad_norm 1.3405 (inf) [2021-04-15 19:35:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][1030/1251] eta 0:01:03 lr 0.000912 time 0.2884 (0.2880) loss 3.6829 (3.8823) grad_norm 1.0427 (inf) [2021-04-15 19:35:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][1040/1251] eta 0:01:00 lr 0.000912 time 0.2885 (0.2879) loss 3.3544 (3.8831) grad_norm 1.2708 (inf) [2021-04-15 19:35:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][1050/1251] eta 0:00:57 lr 0.000912 time 0.2859 (0.2879) loss 4.1439 (3.8858) grad_norm 0.9561 (inf) [2021-04-15 19:35:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][1060/1251] eta 0:00:54 lr 0.000912 time 0.2857 (0.2878) loss 4.8084 (3.8869) grad_norm 1.0537 (inf) [2021-04-15 19:35:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][1070/1251] eta 0:00:52 lr 0.000912 time 0.2448 (0.2877) loss 3.3664 (3.8874) grad_norm 0.9360 (inf) [2021-04-15 19:35:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][1080/1251] eta 0:00:49 lr 0.000912 time 0.2818 (0.2876) loss 4.6473 (3.8856) grad_norm 1.2051 (inf) [2021-04-15 19:35:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][1090/1251] eta 0:00:46 lr 0.000912 time 0.2613 (0.2877) loss 4.1435 (3.8858) grad_norm 1.2105 (inf) [2021-04-15 19:35:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][1100/1251] eta 0:00:43 lr 0.000912 time 0.2624 (0.2875) loss 4.0720 (3.8835) grad_norm 1.0171 (inf) [2021-04-15 19:35:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][1110/1251] eta 0:00:40 lr 0.000912 time 0.2617 (0.2875) loss 4.0656 (3.8841) grad_norm 0.9857 (inf) [2021-04-15 19:35:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][1120/1251] eta 0:00:37 lr 0.000912 time 0.2671 (0.2875) loss 2.9618 (3.8805) grad_norm 1.0756 (inf) [2021-04-15 19:35:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][1130/1251] eta 0:00:34 lr 0.000912 time 0.2709 (0.2875) loss 3.3004 (3.8792) grad_norm 1.0041 (inf) [2021-04-15 19:35:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][1140/1251] eta 0:00:31 lr 0.000912 time 0.2809 (0.2874) loss 4.3741 (3.8804) grad_norm 0.9793 (inf) [2021-04-15 19:36:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][1150/1251] eta 0:00:29 lr 0.000912 time 0.2703 (0.2875) loss 4.1333 (3.8821) grad_norm 1.0714 (inf) [2021-04-15 19:36:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][1160/1251] eta 0:00:26 lr 0.000912 time 0.2693 (0.2874) loss 4.5598 (3.8836) grad_norm 1.0223 (inf) [2021-04-15 19:36:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][1170/1251] eta 0:00:23 lr 0.000912 time 0.3065 (0.2874) loss 3.5001 (3.8841) grad_norm 1.1235 (inf) [2021-04-15 19:36:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][1180/1251] eta 0:00:20 lr 0.000912 time 0.2550 (0.2874) loss 3.7537 (3.8839) grad_norm 1.1207 (inf) [2021-04-15 19:36:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][1190/1251] eta 0:00:17 lr 0.000912 time 0.2673 (0.2873) loss 3.9903 (3.8845) grad_norm 1.0760 (inf) [2021-04-15 19:36:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][1200/1251] eta 0:00:14 lr 0.000912 time 0.2855 (0.2873) loss 4.5449 (3.8855) grad_norm 1.0725 (inf) [2021-04-15 19:36:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][1210/1251] eta 0:00:11 lr 0.000912 time 0.2738 (0.2872) loss 3.3681 (3.8847) grad_norm 1.1354 (inf) [2021-04-15 19:36:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][1220/1251] eta 0:00:08 lr 0.000912 time 0.2734 (0.2871) loss 4.4079 (3.8854) grad_norm 1.3794 (inf) [2021-04-15 19:36:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][1230/1251] eta 0:00:06 lr 0.000912 time 0.2821 (0.2870) loss 3.3343 (3.8840) grad_norm 1.0834 (inf) [2021-04-15 19:36:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][1240/1251] eta 0:00:03 lr 0.000911 time 0.2482 (0.2868) loss 2.8276 (3.8841) grad_norm 1.1980 (inf) [2021-04-15 19:36:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [57/300][1250/1251] eta 0:00:00 lr 0.000911 time 0.2484 (0.2865) loss 2.6431 (3.8833) grad_norm 1.1186 (inf) [2021-04-15 19:36:32 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 57 training takes 0:06:01 [2021-04-15 19:36:32 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_57.pth saving...... [2021-04-15 19:36:53 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_57.pth saved !!! [2021-04-15 19:36:54 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.195 (1.195) Loss 1.2908 (1.2908) Acc@1 69.238 (69.238) Acc@5 90.430 (90.430) [2021-04-15 19:36:56 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.105 (0.287) Loss 1.2098 (1.2617) Acc@1 72.949 (70.961) Acc@5 91.602 (90.705) [2021-04-15 19:36:57 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.101 (0.223) Loss 1.2646 (1.2633) Acc@1 71.387 (70.778) Acc@5 90.332 (90.537) [2021-04-15 19:37:00 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.136 (0.219) Loss 1.2769 (1.2773) Acc@1 70.215 (70.580) Acc@5 90.625 (90.360) [2021-04-15 19:37:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.219) Loss 1.3424 (1.2756) Acc@1 69.434 (70.644) Acc@5 88.184 (90.334) [2021-04-15 19:37:06 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 70.600 Acc@5 90.408 [2021-04-15 19:37:06 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 70.6% [2021-04-15 19:37:06 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 70.67% [2021-04-15 19:37:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][0/1251] eta 1:20:17 lr 0.000911 time 3.8507 (3.8507) loss 3.7564 (3.7564) grad_norm 1.2159 (1.2159) [2021-04-15 19:37:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][10/1251] eta 0:12:31 lr 0.000911 time 0.2875 (0.6052) loss 4.6112 (3.7319) grad_norm 1.1237 (1.1185) [2021-04-15 19:37:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][20/1251] eta 0:09:23 lr 0.000911 time 0.2953 (0.4579) loss 2.8871 (3.8148) grad_norm 1.0495 (1.1881) [2021-04-15 19:37:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][30/1251] eta 0:08:10 lr 0.000911 time 0.2939 (0.4019) loss 4.3759 (3.9360) grad_norm 1.0414 (1.1576) [2021-04-15 19:37:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][40/1251] eta 0:07:33 lr 0.000911 time 0.2915 (0.3746) loss 4.3664 (3.9297) grad_norm 0.9369 (1.1467) [2021-04-15 19:37:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][50/1251] eta 0:07:07 lr 0.000911 time 0.2759 (0.3561) loss 3.2760 (3.9558) grad_norm 0.9819 (1.1575) [2021-04-15 19:37:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][60/1251] eta 0:06:49 lr 0.000911 time 0.2706 (0.3435) loss 3.4840 (3.9575) grad_norm 1.1223 (1.1466) [2021-04-15 19:37:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][70/1251] eta 0:06:34 lr 0.000911 time 0.2812 (0.3344) loss 3.6519 (3.9353) grad_norm 1.3299 (1.1408) [2021-04-15 19:37:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][80/1251] eta 0:06:24 lr 0.000911 time 0.2615 (0.3281) loss 3.3577 (3.9334) grad_norm 1.1360 (1.1315) [2021-04-15 19:37:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][90/1251] eta 0:06:15 lr 0.000911 time 0.2922 (0.3238) loss 4.4532 (3.9562) grad_norm 1.1164 (1.1202) [2021-04-15 19:37:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][100/1251] eta 0:06:07 lr 0.000911 time 0.2681 (0.3193) loss 4.1437 (3.9536) grad_norm 1.1149 (1.1201) [2021-04-15 19:37:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][110/1251] eta 0:06:00 lr 0.000911 time 0.2787 (0.3156) loss 3.7105 (3.9590) grad_norm 1.0380 (1.1228) [2021-04-15 19:37:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][120/1251] eta 0:05:53 lr 0.000911 time 0.2769 (0.3127) loss 4.7144 (3.9598) grad_norm 1.1229 (1.1235) [2021-04-15 19:37:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][130/1251] eta 0:05:47 lr 0.000911 time 0.2593 (0.3101) loss 4.2331 (3.9412) grad_norm 1.0821 (1.1256) [2021-04-15 19:37:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][140/1251] eta 0:05:42 lr 0.000911 time 0.2774 (0.3080) loss 4.1556 (3.9269) grad_norm 1.0279 (1.1307) [2021-04-15 19:37:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][150/1251] eta 0:05:37 lr 0.000911 time 0.2842 (0.3068) loss 3.9425 (3.9422) grad_norm 1.1881 (1.1365) [2021-04-15 19:37:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][160/1251] eta 0:05:33 lr 0.000911 time 0.2616 (0.3053) loss 4.6222 (3.9443) grad_norm 0.9561 (1.1339) [2021-04-15 19:37:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][170/1251] eta 0:05:28 lr 0.000911 time 0.2671 (0.3039) loss 3.8234 (3.9379) grad_norm 1.5131 (1.1370) [2021-04-15 19:38:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][180/1251] eta 0:05:24 lr 0.000911 time 0.2830 (0.3028) loss 4.2239 (3.9243) grad_norm 1.2519 (1.1378) [2021-04-15 19:38:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][190/1251] eta 0:05:19 lr 0.000911 time 0.2828 (0.3015) loss 4.2421 (3.9209) grad_norm 1.0681 (1.1389) [2021-04-15 19:38:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][200/1251] eta 0:05:15 lr 0.000911 time 0.2782 (0.3004) loss 4.1274 (3.9239) grad_norm 1.0612 (1.1369) [2021-04-15 19:38:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][210/1251] eta 0:05:11 lr 0.000911 time 0.2920 (0.2996) loss 4.0133 (3.9088) grad_norm 0.9905 (1.1345) [2021-04-15 19:38:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][220/1251] eta 0:05:08 lr 0.000911 time 0.3058 (0.2989) loss 4.3747 (3.9036) grad_norm 1.1616 (1.1389) [2021-04-15 19:38:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][230/1251] eta 0:05:04 lr 0.000911 time 0.2736 (0.2979) loss 4.3387 (3.9003) grad_norm 1.0237 (1.1370) [2021-04-15 19:38:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][240/1251] eta 0:05:00 lr 0.000911 time 0.2883 (0.2973) loss 3.3610 (3.9123) grad_norm 1.0125 (1.1349) [2021-04-15 19:38:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][250/1251] eta 0:04:56 lr 0.000911 time 0.2918 (0.2964) loss 3.6989 (3.9176) grad_norm 1.2315 (1.1330) [2021-04-15 19:38:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][260/1251] eta 0:04:52 lr 0.000911 time 0.2776 (0.2957) loss 3.0695 (3.9133) grad_norm 1.2680 (1.1315) [2021-04-15 19:38:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][270/1251] eta 0:04:49 lr 0.000911 time 0.2843 (0.2950) loss 3.5025 (3.9245) grad_norm 0.9985 (1.1278) [2021-04-15 19:38:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][280/1251] eta 0:04:45 lr 0.000911 time 0.2635 (0.2945) loss 4.1705 (3.9259) grad_norm 1.1542 (1.1264) [2021-04-15 19:38:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][290/1251] eta 0:04:42 lr 0.000911 time 0.2704 (0.2940) loss 3.6536 (3.9315) grad_norm 1.1244 (1.1239) [2021-04-15 19:38:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][300/1251] eta 0:04:39 lr 0.000911 time 0.2781 (0.2940) loss 3.7265 (3.9418) grad_norm 1.0412 (1.1238) [2021-04-15 19:38:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][310/1251] eta 0:04:36 lr 0.000911 time 0.2717 (0.2940) loss 3.9996 (3.9272) grad_norm 1.6461 (1.1235) [2021-04-15 19:38:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][320/1251] eta 0:04:33 lr 0.000911 time 0.2859 (0.2936) loss 3.7945 (3.9269) grad_norm 1.1477 (1.1266) [2021-04-15 19:38:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][330/1251] eta 0:04:30 lr 0.000911 time 0.2621 (0.2936) loss 4.2202 (3.9162) grad_norm 1.1938 (1.1260) [2021-04-15 19:38:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][340/1251] eta 0:04:27 lr 0.000911 time 0.2848 (0.2936) loss 4.1074 (3.9134) grad_norm 1.1101 (1.1236) [2021-04-15 19:38:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][350/1251] eta 0:04:24 lr 0.000911 time 0.2786 (0.2935) loss 3.4818 (3.9141) grad_norm 1.1028 (1.1225) [2021-04-15 19:38:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][360/1251] eta 0:04:21 lr 0.000911 time 0.2609 (0.2934) loss 3.1420 (3.9100) grad_norm 1.0321 (1.1206) [2021-04-15 19:38:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][370/1251] eta 0:04:18 lr 0.000911 time 0.2700 (0.2934) loss 3.5604 (3.9054) grad_norm 1.1754 (1.1189) [2021-04-15 19:38:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][380/1251] eta 0:04:15 lr 0.000911 time 0.2661 (0.2934) loss 3.8226 (3.9053) grad_norm 1.4353 (1.1190) [2021-04-15 19:39:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][390/1251] eta 0:04:12 lr 0.000911 time 0.2722 (0.2931) loss 4.0271 (3.9077) grad_norm 1.2895 (1.1226) [2021-04-15 19:39:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][400/1251] eta 0:04:09 lr 0.000911 time 0.2682 (0.2927) loss 2.8510 (3.9124) grad_norm 1.2089 (1.1215) [2021-04-15 19:39:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][410/1251] eta 0:04:05 lr 0.000910 time 0.2901 (0.2924) loss 3.9226 (3.9180) grad_norm 1.1228 (1.1212) [2021-04-15 19:39:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][420/1251] eta 0:04:02 lr 0.000910 time 0.2721 (0.2921) loss 4.1494 (3.9162) grad_norm 1.1463 (1.1207) [2021-04-15 19:39:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][430/1251] eta 0:03:59 lr 0.000910 time 0.2769 (0.2919) loss 3.9992 (3.9174) grad_norm 1.0055 (1.1200) [2021-04-15 19:39:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][440/1251] eta 0:03:56 lr 0.000910 time 0.2807 (0.2916) loss 4.3169 (3.9189) grad_norm 0.9157 (1.1181) [2021-04-15 19:39:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][450/1251] eta 0:03:53 lr 0.000910 time 0.2932 (0.2913) loss 4.8467 (3.9224) grad_norm 1.0922 (1.1174) [2021-04-15 19:39:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][460/1251] eta 0:03:50 lr 0.000910 time 0.2616 (0.2911) loss 3.6806 (3.9218) grad_norm 1.4044 (1.1185) [2021-04-15 19:39:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][470/1251] eta 0:03:47 lr 0.000910 time 0.2903 (0.2908) loss 3.6468 (3.9254) grad_norm 0.9226 (1.1188) [2021-04-15 19:39:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][480/1251] eta 0:03:44 lr 0.000910 time 0.2619 (0.2906) loss 4.7913 (3.9277) grad_norm 1.1002 (1.1176) [2021-04-15 19:39:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][490/1251] eta 0:03:40 lr 0.000910 time 0.2687 (0.2904) loss 4.7299 (3.9271) grad_norm 1.0478 (1.1183) [2021-04-15 19:39:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][500/1251] eta 0:03:37 lr 0.000910 time 0.2858 (0.2903) loss 3.9011 (3.9223) grad_norm 1.0885 (1.1200) [2021-04-15 19:39:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][510/1251] eta 0:03:34 lr 0.000910 time 0.2749 (0.2900) loss 4.1398 (3.9224) grad_norm 0.9496 (1.1204) [2021-04-15 19:39:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][520/1251] eta 0:03:31 lr 0.000910 time 0.2689 (0.2898) loss 3.4411 (3.9213) grad_norm 1.2952 (1.1213) [2021-04-15 19:39:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][530/1251] eta 0:03:28 lr 0.000910 time 0.2776 (0.2897) loss 3.7865 (3.9203) grad_norm 1.0496 (1.1220) [2021-04-15 19:39:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][540/1251] eta 0:03:25 lr 0.000910 time 0.2763 (0.2894) loss 3.7396 (3.9200) grad_norm 0.9792 (1.1208) [2021-04-15 19:39:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][550/1251] eta 0:03:22 lr 0.000910 time 0.2687 (0.2892) loss 4.0096 (3.9189) grad_norm 1.1061 (1.1186) [2021-04-15 19:39:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][560/1251] eta 0:03:19 lr 0.000910 time 0.2934 (0.2889) loss 3.5765 (3.9094) grad_norm 1.0645 (1.1179) [2021-04-15 19:39:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][570/1251] eta 0:03:16 lr 0.000910 time 0.2672 (0.2890) loss 3.2805 (3.9086) grad_norm 1.0747 (1.1183) [2021-04-15 19:39:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][580/1251] eta 0:03:13 lr 0.000910 time 0.2877 (0.2888) loss 5.0395 (3.9041) grad_norm 0.9470 (1.1169) [2021-04-15 19:39:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][590/1251] eta 0:03:10 lr 0.000910 time 0.2440 (0.2889) loss 4.1161 (3.9079) grad_norm 1.0699 (1.1167) [2021-04-15 19:39:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][600/1251] eta 0:03:07 lr 0.000910 time 0.2427 (0.2887) loss 3.6417 (3.9110) grad_norm 1.1455 (1.1167) [2021-04-15 19:40:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][610/1251] eta 0:03:04 lr 0.000910 time 0.2631 (0.2885) loss 4.4009 (3.9036) grad_norm 1.0589 (1.1163) [2021-04-15 19:40:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][620/1251] eta 0:03:02 lr 0.000910 time 0.2430 (0.2885) loss 4.8135 (3.8977) grad_norm 1.0657 (1.1184) [2021-04-15 19:40:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][630/1251] eta 0:02:59 lr 0.000910 time 0.2660 (0.2883) loss 3.8661 (3.8955) grad_norm 1.1169 (1.1181) [2021-04-15 19:40:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][640/1251] eta 0:02:56 lr 0.000910 time 0.2989 (0.2882) loss 4.0551 (3.8961) grad_norm 1.3582 (1.1189) [2021-04-15 19:40:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][650/1251] eta 0:02:53 lr 0.000910 time 0.2871 (0.2880) loss 3.5697 (3.8937) grad_norm 1.3252 (1.1196) [2021-04-15 19:40:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][660/1251] eta 0:02:50 lr 0.000910 time 0.2473 (0.2878) loss 4.4631 (3.8894) grad_norm 1.0841 (1.1193) [2021-04-15 19:40:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][670/1251] eta 0:02:47 lr 0.000910 time 0.2809 (0.2880) loss 3.1830 (3.8883) grad_norm 1.1391 (1.1200) [2021-04-15 19:40:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][680/1251] eta 0:02:44 lr 0.000910 time 0.2681 (0.2878) loss 4.2843 (3.8916) grad_norm 1.0940 (1.1199) [2021-04-15 19:40:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][690/1251] eta 0:02:41 lr 0.000910 time 0.2869 (0.2877) loss 2.5933 (3.8909) grad_norm 1.0116 (1.1196) [2021-04-15 19:40:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][700/1251] eta 0:02:38 lr 0.000910 time 0.2882 (0.2875) loss 4.4703 (3.8885) grad_norm 1.0672 (1.1210) [2021-04-15 19:40:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][710/1251] eta 0:02:35 lr 0.000910 time 0.3146 (0.2873) loss 4.3618 (3.8937) grad_norm 0.9836 (1.1215) [2021-04-15 19:40:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][720/1251] eta 0:02:32 lr 0.000910 time 0.2871 (0.2874) loss 4.0884 (3.8919) grad_norm 1.1925 (1.1211) [2021-04-15 19:40:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][730/1251] eta 0:02:29 lr 0.000910 time 0.2941 (0.2874) loss 3.8338 (3.8922) grad_norm 1.0557 (1.1213) [2021-04-15 19:40:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][740/1251] eta 0:02:26 lr 0.000910 time 0.2756 (0.2872) loss 3.2156 (3.8899) grad_norm 1.0646 (1.1207) [2021-04-15 19:40:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][750/1251] eta 0:02:23 lr 0.000910 time 0.2934 (0.2871) loss 4.2120 (3.8897) grad_norm 1.2068 (1.1215) [2021-04-15 19:40:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][760/1251] eta 0:02:20 lr 0.000910 time 0.2745 (0.2870) loss 4.1850 (3.8913) grad_norm 1.2348 (1.1217) [2021-04-15 19:40:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][770/1251] eta 0:02:18 lr 0.000910 time 0.2747 (0.2870) loss 4.1793 (3.8940) grad_norm 1.1498 (1.1215) [2021-04-15 19:40:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][780/1251] eta 0:02:15 lr 0.000910 time 0.2987 (0.2868) loss 4.6690 (3.8968) grad_norm 1.0820 (1.1221) [2021-04-15 19:40:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][790/1251] eta 0:02:12 lr 0.000910 time 0.2976 (0.2868) loss 4.2488 (3.8985) grad_norm 1.8356 (1.1242) [2021-04-15 19:40:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][800/1251] eta 0:02:09 lr 0.000910 time 0.2631 (0.2866) loss 4.9044 (3.9004) grad_norm 1.0383 (1.1254) [2021-04-15 19:40:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][810/1251] eta 0:02:06 lr 0.000910 time 0.3078 (0.2866) loss 3.5354 (3.9016) grad_norm 0.9863 (1.1254) [2021-04-15 19:41:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][820/1251] eta 0:02:03 lr 0.000910 time 0.2966 (0.2865) loss 3.4452 (3.8990) grad_norm 1.0797 (1.1247) [2021-04-15 19:41:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][830/1251] eta 0:02:00 lr 0.000909 time 0.2530 (0.2864) loss 4.1249 (3.8995) grad_norm 1.3773 (1.1260) [2021-04-15 19:41:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][840/1251] eta 0:01:57 lr 0.000909 time 0.2895 (0.2863) loss 3.7315 (3.9006) grad_norm 1.0426 (1.1266) [2021-04-15 19:41:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][850/1251] eta 0:01:54 lr 0.000909 time 0.2953 (0.2863) loss 4.4387 (3.8984) grad_norm 1.1271 (1.1267) [2021-04-15 19:41:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][860/1251] eta 0:01:51 lr 0.000909 time 0.2858 (0.2861) loss 3.7522 (3.8966) grad_norm 1.1028 (1.1268) [2021-04-15 19:41:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][870/1251] eta 0:01:48 lr 0.000909 time 0.2701 (0.2860) loss 3.6173 (3.8954) grad_norm 1.2893 (1.1266) [2021-04-15 19:41:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][880/1251] eta 0:01:46 lr 0.000909 time 0.4309 (0.2863) loss 4.6204 (3.8950) grad_norm 1.2931 (1.1268) [2021-04-15 19:41:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][890/1251] eta 0:01:43 lr 0.000909 time 0.2689 (0.2861) loss 4.4461 (3.8955) grad_norm 1.1018 (1.1275) [2021-04-15 19:41:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][900/1251] eta 0:01:40 lr 0.000909 time 0.4843 (0.2863) loss 4.4379 (3.8954) grad_norm 0.9754 (1.1274) [2021-04-15 19:41:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][910/1251] eta 0:01:37 lr 0.000909 time 0.2827 (0.2862) loss 3.5766 (3.8913) grad_norm 1.1173 (1.1271) [2021-04-15 19:41:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][920/1251] eta 0:01:34 lr 0.000909 time 0.2834 (0.2863) loss 4.7177 (3.8903) grad_norm 1.0595 (1.1271) [2021-04-15 19:41:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][930/1251] eta 0:01:31 lr 0.000909 time 0.2573 (0.2863) loss 4.0209 (3.8907) grad_norm 0.9997 (1.1267) [2021-04-15 19:41:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][940/1251] eta 0:01:29 lr 0.000909 time 0.2846 (0.2865) loss 3.2318 (3.8877) grad_norm 1.1116 (1.1275) [2021-04-15 19:41:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][950/1251] eta 0:01:26 lr 0.000909 time 0.2706 (0.2864) loss 4.2862 (3.8853) grad_norm 1.1078 (1.1286) [2021-04-15 19:41:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][960/1251] eta 0:01:23 lr 0.000909 time 0.2818 (0.2865) loss 4.2257 (3.8847) grad_norm 0.9764 (1.1282) [2021-04-15 19:41:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][970/1251] eta 0:01:20 lr 0.000909 time 0.2682 (0.2863) loss 4.0491 (3.8826) grad_norm 1.2116 (1.1276) [2021-04-15 19:41:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][980/1251] eta 0:01:17 lr 0.000909 time 0.2526 (0.2862) loss 4.0039 (3.8833) grad_norm 1.1754 (1.1276) [2021-04-15 19:41:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][990/1251] eta 0:01:14 lr 0.000909 time 0.2830 (0.2862) loss 4.4405 (3.8829) grad_norm 1.0462 (1.1277) [2021-04-15 19:41:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][1000/1251] eta 0:01:11 lr 0.000909 time 0.2433 (0.2861) loss 3.3269 (3.8827) grad_norm 1.0802 (1.1282) [2021-04-15 19:41:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][1010/1251] eta 0:01:08 lr 0.000909 time 0.2715 (0.2860) loss 3.7406 (3.8826) grad_norm 1.1440 (1.1278) [2021-04-15 19:41:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][1020/1251] eta 0:01:06 lr 0.000909 time 0.2873 (0.2859) loss 4.4058 (3.8822) grad_norm 1.0149 (1.1281) [2021-04-15 19:42:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][1030/1251] eta 0:01:03 lr 0.000909 time 0.2749 (0.2858) loss 3.1409 (3.8817) grad_norm 1.1118 (1.1283) [2021-04-15 19:42:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][1040/1251] eta 0:01:00 lr 0.000909 time 0.2821 (0.2858) loss 4.1882 (3.8828) grad_norm 1.4516 (1.1286) [2021-04-15 19:42:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][1050/1251] eta 0:00:57 lr 0.000909 time 0.2743 (0.2857) loss 4.0766 (3.8811) grad_norm 1.1246 (1.1289) [2021-04-15 19:42:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][1060/1251] eta 0:00:54 lr 0.000909 time 0.2734 (0.2856) loss 4.2667 (3.8811) grad_norm 1.1052 (1.1298) [2021-04-15 19:42:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][1070/1251] eta 0:00:51 lr 0.000909 time 0.2690 (0.2856) loss 3.4973 (3.8813) grad_norm 1.2987 (1.1297) [2021-04-15 19:42:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][1080/1251] eta 0:00:48 lr 0.000909 time 0.2861 (0.2855) loss 4.4166 (3.8831) grad_norm 1.1511 (1.1299) [2021-04-15 19:42:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][1090/1251] eta 0:00:45 lr 0.000909 time 0.2689 (0.2854) loss 3.8624 (3.8845) grad_norm 1.1699 (1.1295) [2021-04-15 19:42:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][1100/1251] eta 0:00:43 lr 0.000909 time 0.2675 (0.2854) loss 4.4402 (3.8838) grad_norm 0.9873 (1.1289) [2021-04-15 19:42:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][1110/1251] eta 0:00:40 lr 0.000909 time 0.2656 (0.2853) loss 4.2378 (3.8862) grad_norm 1.1384 (1.1287) [2021-04-15 19:42:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][1120/1251] eta 0:00:37 lr 0.000909 time 0.2589 (0.2852) loss 3.8695 (3.8872) grad_norm 0.9510 (1.1288) [2021-04-15 19:42:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][1130/1251] eta 0:00:34 lr 0.000909 time 0.2666 (0.2851) loss 3.5921 (3.8860) grad_norm 1.3273 (1.1286) [2021-04-15 19:42:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][1140/1251] eta 0:00:31 lr 0.000909 time 0.2797 (0.2850) loss 4.3072 (3.8865) grad_norm 0.8990 (1.1290) [2021-04-15 19:42:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][1150/1251] eta 0:00:28 lr 0.000909 time 0.2976 (0.2851) loss 4.4653 (3.8883) grad_norm 1.0493 (1.1289) [2021-04-15 19:42:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][1160/1251] eta 0:00:25 lr 0.000909 time 0.2780 (0.2850) loss 4.3195 (3.8867) grad_norm 1.3139 (1.1293) [2021-04-15 19:42:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][1170/1251] eta 0:00:23 lr 0.000909 time 0.2609 (0.2850) loss 3.9998 (3.8866) grad_norm 0.9423 (1.1287) [2021-04-15 19:42:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][1180/1251] eta 0:00:20 lr 0.000909 time 0.2654 (0.2850) loss 3.0290 (3.8837) grad_norm 0.9790 (1.1284) [2021-04-15 19:42:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][1190/1251] eta 0:00:17 lr 0.000909 time 0.4313 (0.2851) loss 3.9268 (3.8849) grad_norm 1.0813 (1.1289) [2021-04-15 19:42:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][1200/1251] eta 0:00:14 lr 0.000909 time 0.2830 (0.2851) loss 2.7580 (3.8835) grad_norm 1.0638 (1.1289) [2021-04-15 19:42:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][1210/1251] eta 0:00:11 lr 0.000909 time 0.2605 (0.2851) loss 3.9876 (3.8837) grad_norm 1.0853 (1.1291) [2021-04-15 19:42:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][1220/1251] eta 0:00:08 lr 0.000909 time 0.2774 (0.2850) loss 2.7860 (3.8820) grad_norm 1.0522 (1.1292) [2021-04-15 19:42:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][1230/1251] eta 0:00:05 lr 0.000909 time 0.2567 (0.2849) loss 2.7854 (3.8827) grad_norm 1.0578 (1.1286) [2021-04-15 19:42:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][1240/1251] eta 0:00:03 lr 0.000909 time 0.2482 (0.2849) loss 3.8746 (3.8835) grad_norm 1.0078 (1.1279) [2021-04-15 19:43:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [58/300][1250/1251] eta 0:00:00 lr 0.000908 time 0.2478 (0.2846) loss 4.3945 (3.8820) grad_norm 1.0078 (1.1274) [2021-04-15 19:43:04 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 58 training takes 0:05:58 [2021-04-15 19:43:04 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_58.pth saving...... [2021-04-15 19:43:13 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_58.pth saved !!! [2021-04-15 19:43:14 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.164 (1.164) Loss 1.1770 (1.1770) Acc@1 74.121 (74.121) Acc@5 91.992 (91.992) [2021-04-15 19:43:16 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.156 (0.264) Loss 1.2555 (1.2649) Acc@1 71.973 (71.298) Acc@5 91.309 (90.607) [2021-04-15 19:43:18 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.179 (0.248) Loss 1.3223 (1.2813) Acc@1 68.945 (70.991) Acc@5 90.234 (90.574) [2021-04-15 19:43:20 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.101 (0.234) Loss 1.3276 (1.2821) Acc@1 70.410 (70.977) Acc@5 89.746 (90.578) [2021-04-15 19:43:22 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.216 (0.224) Loss 1.2000 (1.2756) Acc@1 73.730 (71.058) Acc@5 90.625 (90.615) [2021-04-15 19:43:25 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 70.980 Acc@5 90.564 [2021-04-15 19:43:25 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 71.0% [2021-04-15 19:43:25 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 70.98% [2021-04-15 19:43:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][0/1251] eta 1:38:16 lr 0.000908 time 4.7137 (4.7137) loss 4.1650 (4.1650) grad_norm 0.9852 (0.9852) [2021-04-15 19:43:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][10/1251] eta 0:14:08 lr 0.000908 time 0.2921 (0.6835) loss 4.2245 (3.9368) grad_norm 1.0186 (1.0414) [2021-04-15 19:43:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][20/1251] eta 0:10:03 lr 0.000908 time 0.2854 (0.4902) loss 3.8261 (3.9686) grad_norm 0.9347 (1.0493) [2021-04-15 19:43:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][30/1251] eta 0:08:33 lr 0.000908 time 0.3021 (0.4209) loss 4.1312 (4.0612) grad_norm 1.1322 (1.0643) [2021-04-15 19:43:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][40/1251] eta 0:07:48 lr 0.000908 time 0.3029 (0.3865) loss 4.0768 (3.9641) grad_norm 1.1012 (1.0627) [2021-04-15 19:43:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][50/1251] eta 0:07:18 lr 0.000908 time 0.2807 (0.3651) loss 3.3994 (3.9902) grad_norm 1.0240 (1.0681) [2021-04-15 19:43:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][60/1251] eta 0:06:58 lr 0.000908 time 0.2760 (0.3512) loss 4.4417 (3.9664) grad_norm 1.0468 (1.0781) [2021-04-15 19:43:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][70/1251] eta 0:06:45 lr 0.000908 time 0.2728 (0.3432) loss 4.2031 (3.9392) grad_norm 1.0895 (1.0751) [2021-04-15 19:43:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][80/1251] eta 0:06:34 lr 0.000908 time 0.2835 (0.3369) loss 4.0762 (3.9508) grad_norm 1.0804 (1.0842) [2021-04-15 19:43:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][90/1251] eta 0:06:23 lr 0.000908 time 0.2734 (0.3304) loss 3.2019 (3.9123) grad_norm 1.0249 (1.0837) [2021-04-15 19:43:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][100/1251] eta 0:06:16 lr 0.000908 time 0.2645 (0.3270) loss 3.6419 (3.9228) grad_norm 0.9428 (1.0798) [2021-04-15 19:44:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][110/1251] eta 0:06:07 lr 0.000908 time 0.2654 (0.3225) loss 4.3668 (3.9070) grad_norm 1.1966 (1.0831) [2021-04-15 19:44:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][120/1251] eta 0:06:02 lr 0.000908 time 0.2835 (0.3203) loss 3.8966 (3.9029) grad_norm 1.0947 (1.0907) [2021-04-15 19:44:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][130/1251] eta 0:05:55 lr 0.000908 time 0.3093 (0.3170) loss 4.6692 (3.9213) grad_norm 1.0573 (1.0877) [2021-04-15 19:44:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][140/1251] eta 0:05:51 lr 0.000908 time 0.2745 (0.3163) loss 2.9355 (3.8958) grad_norm 1.1598 (1.0872) [2021-04-15 19:44:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][150/1251] eta 0:05:46 lr 0.000908 time 0.3074 (0.3147) loss 4.4184 (3.9085) grad_norm 1.0905 (1.0862) [2021-04-15 19:44:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][160/1251] eta 0:05:40 lr 0.000908 time 0.2648 (0.3125) loss 4.8912 (3.9228) grad_norm 1.1583 (1.0875) [2021-04-15 19:44:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][170/1251] eta 0:05:36 lr 0.000908 time 0.2864 (0.3113) loss 4.2461 (3.9149) grad_norm 1.2131 (1.0889) [2021-04-15 19:44:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][180/1251] eta 0:05:31 lr 0.000908 time 0.2797 (0.3094) loss 4.3674 (3.9151) grad_norm 0.9708 (1.0901) [2021-04-15 19:44:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][190/1251] eta 0:05:26 lr 0.000908 time 0.2794 (0.3077) loss 3.6942 (3.9222) grad_norm 1.0198 (1.0881) [2021-04-15 19:44:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][200/1251] eta 0:05:22 lr 0.000908 time 0.2601 (0.3064) loss 2.9755 (3.9044) grad_norm 1.0335 (1.0876) [2021-04-15 19:44:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][210/1251] eta 0:05:17 lr 0.000908 time 0.2850 (0.3052) loss 3.4973 (3.9089) grad_norm 1.0560 (1.0868) [2021-04-15 19:44:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][220/1251] eta 0:05:13 lr 0.000908 time 0.2908 (0.3039) loss 3.6046 (3.8960) grad_norm 1.1860 (1.0895) [2021-04-15 19:44:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][230/1251] eta 0:05:09 lr 0.000908 time 0.3241 (0.3029) loss 4.4806 (3.8953) grad_norm 1.2787 (1.0953) [2021-04-15 19:44:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][240/1251] eta 0:05:05 lr 0.000908 time 0.2938 (0.3019) loss 4.4186 (3.8955) grad_norm 1.1210 (1.0970) [2021-04-15 19:44:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][250/1251] eta 0:05:01 lr 0.000908 time 0.2816 (0.3010) loss 3.6974 (3.8866) grad_norm 1.1335 (1.0956) [2021-04-15 19:44:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][260/1251] eta 0:04:57 lr 0.000908 time 0.2702 (0.3002) loss 4.3073 (3.8900) grad_norm 1.1825 (1.0966) [2021-04-15 19:44:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][270/1251] eta 0:04:53 lr 0.000908 time 0.2893 (0.2994) loss 3.7857 (3.8928) grad_norm 1.3332 (1.0988) [2021-04-15 19:44:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][280/1251] eta 0:04:50 lr 0.000908 time 0.2763 (0.2988) loss 4.1582 (3.8933) grad_norm 1.1648 (1.1024) [2021-04-15 19:44:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][290/1251] eta 0:04:46 lr 0.000908 time 0.2832 (0.2981) loss 4.4794 (3.9037) grad_norm 0.9694 (1.1005) [2021-04-15 19:44:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][300/1251] eta 0:04:42 lr 0.000908 time 0.2745 (0.2975) loss 4.0600 (3.9058) grad_norm 1.0655 (1.1001) [2021-04-15 19:44:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][310/1251] eta 0:04:39 lr 0.000908 time 0.2802 (0.2969) loss 4.0537 (3.9057) grad_norm 1.0435 (1.1008) [2021-04-15 19:45:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][320/1251] eta 0:04:35 lr 0.000908 time 0.2657 (0.2962) loss 3.5088 (3.9041) grad_norm 1.0938 (1.1009) [2021-04-15 19:45:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][330/1251] eta 0:04:32 lr 0.000908 time 0.2786 (0.2956) loss 4.3519 (3.9094) grad_norm 1.5204 (1.1014) [2021-04-15 19:45:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][340/1251] eta 0:04:28 lr 0.000908 time 0.2931 (0.2952) loss 3.4263 (3.9051) grad_norm 1.1088 (1.1025) [2021-04-15 19:45:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][350/1251] eta 0:04:25 lr 0.000908 time 0.2747 (0.2949) loss 3.9328 (3.9067) grad_norm 1.1824 (1.1027) [2021-04-15 19:45:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][360/1251] eta 0:04:22 lr 0.000908 time 0.2686 (0.2944) loss 3.6257 (3.9004) grad_norm 1.0662 (1.1028) [2021-04-15 19:45:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][370/1251] eta 0:04:19 lr 0.000908 time 0.2456 (0.2944) loss 4.6902 (3.9068) grad_norm 1.0051 (1.1022) [2021-04-15 19:45:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][380/1251] eta 0:04:15 lr 0.000908 time 0.2537 (0.2939) loss 3.8794 (3.9103) grad_norm 1.1332 (1.1021) [2021-04-15 19:45:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][390/1251] eta 0:04:13 lr 0.000908 time 0.2752 (0.2939) loss 4.0493 (3.9088) grad_norm 1.1069 (1.1026) [2021-04-15 19:45:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][400/1251] eta 0:04:09 lr 0.000908 time 0.2561 (0.2935) loss 3.9488 (3.9087) grad_norm 1.1073 (1.1019) [2021-04-15 19:45:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][410/1251] eta 0:04:06 lr 0.000908 time 0.2833 (0.2930) loss 4.7247 (3.9081) grad_norm 1.0884 (1.1016) [2021-04-15 19:45:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][420/1251] eta 0:04:03 lr 0.000907 time 0.3009 (0.2927) loss 3.7179 (3.9078) grad_norm 1.2524 (1.1019) [2021-04-15 19:45:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][430/1251] eta 0:04:00 lr 0.000907 time 0.2721 (0.2925) loss 4.8210 (3.9086) grad_norm 1.1787 (1.1033) [2021-04-15 19:45:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][440/1251] eta 0:03:57 lr 0.000907 time 0.2923 (0.2924) loss 4.9391 (3.9144) grad_norm 1.0795 (1.1023) [2021-04-15 19:45:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][450/1251] eta 0:03:53 lr 0.000907 time 0.2758 (0.2921) loss 3.1403 (3.9147) grad_norm 1.0350 (1.1022) [2021-04-15 19:45:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][460/1251] eta 0:03:50 lr 0.000907 time 0.2808 (0.2919) loss 4.5585 (3.9078) grad_norm 1.0293 (1.1035) [2021-04-15 19:45:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][470/1251] eta 0:03:47 lr 0.000907 time 0.2831 (0.2917) loss 4.1400 (3.9139) grad_norm 1.0836 (1.1051) [2021-04-15 19:45:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][480/1251] eta 0:03:44 lr 0.000907 time 0.2741 (0.2914) loss 3.8379 (3.9153) grad_norm 0.9612 (1.1046) [2021-04-15 19:45:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][490/1251] eta 0:03:41 lr 0.000907 time 0.2880 (0.2912) loss 3.2428 (3.9170) grad_norm 0.9911 (1.1027) [2021-04-15 19:45:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][500/1251] eta 0:03:38 lr 0.000907 time 0.2639 (0.2910) loss 2.8785 (3.9107) grad_norm 1.1947 (1.1020) [2021-04-15 19:45:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][510/1251] eta 0:03:35 lr 0.000907 time 0.2726 (0.2909) loss 3.4353 (3.9112) grad_norm 1.2075 (1.1028) [2021-04-15 19:45:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][520/1251] eta 0:03:32 lr 0.000907 time 0.2928 (0.2908) loss 4.3844 (3.9162) grad_norm 1.1476 (1.1031) [2021-04-15 19:45:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][530/1251] eta 0:03:29 lr 0.000907 time 0.3175 (0.2907) loss 3.6674 (3.9127) grad_norm 1.1275 (1.1030) [2021-04-15 19:46:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][540/1251] eta 0:03:26 lr 0.000907 time 0.2739 (0.2905) loss 4.6623 (3.9154) grad_norm 1.5994 (1.1068) [2021-04-15 19:46:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][550/1251] eta 0:03:23 lr 0.000907 time 0.2785 (0.2903) loss 3.2753 (3.9143) grad_norm 1.3046 (1.1073) [2021-04-15 19:46:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][560/1251] eta 0:03:20 lr 0.000907 time 0.2868 (0.2900) loss 3.7247 (3.9147) grad_norm 1.1572 (1.1082) [2021-04-15 19:46:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][570/1251] eta 0:03:17 lr 0.000907 time 0.2762 (0.2898) loss 3.7749 (3.9133) grad_norm 1.2286 (1.1082) [2021-04-15 19:46:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][580/1251] eta 0:03:14 lr 0.000907 time 0.2818 (0.2898) loss 4.4213 (3.9131) grad_norm 1.1696 (1.1089) [2021-04-15 19:46:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][590/1251] eta 0:03:11 lr 0.000907 time 0.2888 (0.2897) loss 4.8325 (3.9159) grad_norm 1.1153 (1.1076) [2021-04-15 19:46:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][600/1251] eta 0:03:08 lr 0.000907 time 0.2564 (0.2895) loss 4.0653 (3.9138) grad_norm 1.0618 (1.1074) [2021-04-15 19:46:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][610/1251] eta 0:03:05 lr 0.000907 time 0.2714 (0.2894) loss 2.6446 (3.9099) grad_norm 1.3244 (1.1085) [2021-04-15 19:46:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][620/1251] eta 0:03:02 lr 0.000907 time 0.2854 (0.2892) loss 3.4621 (3.9075) grad_norm 0.9438 (1.1090) [2021-04-15 19:46:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][630/1251] eta 0:02:59 lr 0.000907 time 0.2955 (0.2890) loss 4.3583 (3.9071) grad_norm 1.0307 (1.1098) [2021-04-15 19:46:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][640/1251] eta 0:02:56 lr 0.000907 time 0.2445 (0.2890) loss 2.9502 (3.9055) grad_norm 1.3170 (1.1116) [2021-04-15 19:46:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][650/1251] eta 0:02:53 lr 0.000907 time 0.3767 (0.2891) loss 4.2998 (3.9065) grad_norm 1.1336 (1.1136) [2021-04-15 19:46:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][660/1251] eta 0:02:50 lr 0.000907 time 0.3055 (0.2888) loss 3.8377 (3.9103) grad_norm 1.4734 (1.1145) [2021-04-15 19:46:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][670/1251] eta 0:02:47 lr 0.000907 time 0.2877 (0.2887) loss 3.2890 (3.9058) grad_norm 0.9713 (1.1136) [2021-04-15 19:46:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][680/1251] eta 0:02:44 lr 0.000907 time 0.2844 (0.2887) loss 4.0406 (3.9015) grad_norm 1.2338 (1.1137) [2021-04-15 19:46:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][690/1251] eta 0:02:42 lr 0.000907 time 0.2676 (0.2888) loss 3.4119 (3.9000) grad_norm 1.0220 (1.1131) [2021-04-15 19:46:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][700/1251] eta 0:02:39 lr 0.000907 time 0.2846 (0.2886) loss 2.7096 (3.8983) grad_norm 1.1870 (1.1121) [2021-04-15 19:46:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][710/1251] eta 0:02:36 lr 0.000907 time 0.4333 (0.2887) loss 4.1243 (3.9008) grad_norm 0.9484 (1.1120) [2021-04-15 19:46:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][720/1251] eta 0:02:33 lr 0.000907 time 0.2687 (0.2886) loss 3.3657 (3.8971) grad_norm 1.0786 (1.1123) [2021-04-15 19:46:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][730/1251] eta 0:02:30 lr 0.000907 time 0.2870 (0.2886) loss 4.1956 (3.8953) grad_norm 0.9301 (1.1127) [2021-04-15 19:46:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][740/1251] eta 0:02:27 lr 0.000907 time 0.3846 (0.2886) loss 4.4664 (3.8952) grad_norm 1.1775 (1.1128) [2021-04-15 19:47:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][750/1251] eta 0:02:24 lr 0.000907 time 0.2888 (0.2885) loss 2.8343 (3.8967) grad_norm 0.9900 (1.1126) [2021-04-15 19:47:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][760/1251] eta 0:02:21 lr 0.000907 time 0.2457 (0.2884) loss 3.9738 (3.8934) grad_norm 1.0043 (1.1132) [2021-04-15 19:47:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][770/1251] eta 0:02:18 lr 0.000907 time 0.2894 (0.2883) loss 3.3839 (3.8950) grad_norm 0.9698 (1.1129) [2021-04-15 19:47:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][780/1251] eta 0:02:15 lr 0.000907 time 0.2788 (0.2882) loss 2.5261 (3.8908) grad_norm 1.2676 (1.1127) [2021-04-15 19:47:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][790/1251] eta 0:02:12 lr 0.000907 time 0.2557 (0.2881) loss 4.1315 (3.8882) grad_norm 1.1797 (1.1144) [2021-04-15 19:47:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][800/1251] eta 0:02:09 lr 0.000907 time 0.2495 (0.2879) loss 3.7283 (3.8889) grad_norm 1.0951 (1.1152) [2021-04-15 19:47:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][810/1251] eta 0:02:06 lr 0.000907 time 0.2811 (0.2879) loss 4.2732 (3.8872) grad_norm 0.9668 (1.1155) [2021-04-15 19:47:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][820/1251] eta 0:02:04 lr 0.000907 time 0.2610 (0.2877) loss 4.6133 (3.8917) grad_norm 1.1018 (1.1155) [2021-04-15 19:47:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][830/1251] eta 0:02:01 lr 0.000906 time 0.2681 (0.2876) loss 4.6057 (3.8929) grad_norm 1.0704 (1.1158) [2021-04-15 19:47:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][840/1251] eta 0:01:58 lr 0.000906 time 0.2723 (0.2875) loss 2.7354 (3.8917) grad_norm 1.0819 (1.1157) [2021-04-15 19:47:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][850/1251] eta 0:01:55 lr 0.000906 time 0.3105 (0.2874) loss 4.1493 (3.8917) grad_norm 1.1799 (1.1158) [2021-04-15 19:47:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][860/1251] eta 0:01:52 lr 0.000906 time 0.2716 (0.2873) loss 3.5133 (3.8888) grad_norm 1.1377 (1.1160) [2021-04-15 19:47:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][870/1251] eta 0:01:49 lr 0.000906 time 0.2761 (0.2871) loss 4.1390 (3.8913) grad_norm 1.0961 (1.1163) [2021-04-15 19:47:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][880/1251] eta 0:01:46 lr 0.000906 time 0.2682 (0.2870) loss 3.7490 (3.8903) grad_norm 0.9908 (1.1162) [2021-04-15 19:47:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][890/1251] eta 0:01:43 lr 0.000906 time 0.2677 (0.2869) loss 3.9894 (3.8943) grad_norm 0.8951 (1.1156) [2021-04-15 19:47:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][900/1251] eta 0:01:40 lr 0.000906 time 0.2807 (0.2868) loss 3.7428 (3.8937) grad_norm 1.2405 (1.1155) [2021-04-15 19:47:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][910/1251] eta 0:01:37 lr 0.000906 time 0.2763 (0.2867) loss 3.6776 (3.8954) grad_norm 1.1058 (1.1153) [2021-04-15 19:47:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][920/1251] eta 0:01:34 lr 0.000906 time 0.2734 (0.2868) loss 3.9559 (3.8958) grad_norm 1.0413 (1.1148) [2021-04-15 19:47:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][930/1251] eta 0:01:32 lr 0.000906 time 0.2814 (0.2867) loss 4.4683 (3.8978) grad_norm 1.1336 (1.1146) [2021-04-15 19:47:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][940/1251] eta 0:01:29 lr 0.000906 time 0.4371 (0.2868) loss 4.2272 (3.8984) grad_norm 1.2311 (1.1150) [2021-04-15 19:47:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][950/1251] eta 0:01:26 lr 0.000906 time 0.2720 (0.2867) loss 3.8781 (3.8996) grad_norm 1.2810 (1.1162) [2021-04-15 19:48:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][960/1251] eta 0:01:23 lr 0.000906 time 0.2599 (0.2865) loss 4.0289 (3.8991) grad_norm 1.3138 (1.1170) [2021-04-15 19:48:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][970/1251] eta 0:01:20 lr 0.000906 time 0.2611 (0.2866) loss 4.3734 (3.9008) grad_norm 1.1088 (1.1162) [2021-04-15 19:48:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][980/1251] eta 0:01:17 lr 0.000906 time 0.2816 (0.2865) loss 4.7453 (3.9000) grad_norm 1.1898 (1.1163) [2021-04-15 19:48:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][990/1251] eta 0:01:14 lr 0.000906 time 0.2926 (0.2865) loss 4.3342 (3.8996) grad_norm 1.1342 (1.1174) [2021-04-15 19:48:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][1000/1251] eta 0:01:11 lr 0.000906 time 0.2608 (0.2864) loss 3.9469 (3.8993) grad_norm 0.9685 (1.1176) [2021-04-15 19:48:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][1010/1251] eta 0:01:08 lr 0.000906 time 0.2756 (0.2863) loss 4.1251 (3.9001) grad_norm 1.0555 (1.1170) [2021-04-15 19:48:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][1020/1251] eta 0:01:06 lr 0.000906 time 0.2874 (0.2863) loss 3.9385 (3.9017) grad_norm 1.1300 (1.1175) [2021-04-15 19:48:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][1030/1251] eta 0:01:03 lr 0.000906 time 0.2507 (0.2862) loss 2.9437 (3.9013) grad_norm 1.1662 (1.1176) [2021-04-15 19:48:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][1040/1251] eta 0:01:00 lr 0.000906 time 0.3069 (0.2862) loss 4.2979 (3.8999) grad_norm 1.0149 (1.1168) [2021-04-15 19:48:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][1050/1251] eta 0:00:57 lr 0.000906 time 0.2949 (0.2862) loss 3.8212 (3.8989) grad_norm 1.1707 (1.1163) [2021-04-15 19:48:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][1060/1251] eta 0:00:54 lr 0.000906 time 0.2918 (0.2861) loss 4.4504 (3.8997) grad_norm 1.2606 (1.1162) [2021-04-15 19:48:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][1070/1251] eta 0:00:51 lr 0.000906 time 0.2946 (0.2861) loss 4.4524 (3.9009) grad_norm 1.0988 (1.1174) [2021-04-15 19:48:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][1080/1251] eta 0:00:48 lr 0.000906 time 0.2803 (0.2860) loss 3.9871 (3.8975) grad_norm 0.9292 (1.1170) [2021-04-15 19:48:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][1090/1251] eta 0:00:46 lr 0.000906 time 0.2871 (0.2859) loss 4.1587 (3.8964) grad_norm 1.0340 (1.1174) [2021-04-15 19:48:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][1100/1251] eta 0:00:43 lr 0.000906 time 0.2795 (0.2859) loss 4.3970 (3.8974) grad_norm 1.1493 (1.1178) [2021-04-15 19:48:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][1110/1251] eta 0:00:40 lr 0.000906 time 0.2732 (0.2858) loss 3.3000 (3.8953) grad_norm 0.9566 (1.1179) [2021-04-15 19:48:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][1120/1251] eta 0:00:37 lr 0.000906 time 0.2713 (0.2857) loss 3.9656 (3.8932) grad_norm 0.9839 (1.1185) [2021-04-15 19:48:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][1130/1251] eta 0:00:34 lr 0.000906 time 0.2735 (0.2857) loss 3.7796 (3.8939) grad_norm 1.1805 (1.1180) [2021-04-15 19:48:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][1140/1251] eta 0:00:31 lr 0.000906 time 0.2695 (0.2856) loss 3.8041 (3.8941) grad_norm 1.0775 (1.1173) [2021-04-15 19:48:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][1150/1251] eta 0:00:28 lr 0.000906 time 0.2787 (0.2856) loss 4.0179 (3.8958) grad_norm 1.0068 (1.1173) [2021-04-15 19:48:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][1160/1251] eta 0:00:25 lr 0.000906 time 0.2734 (0.2857) loss 3.3449 (3.8945) grad_norm 1.4376 (1.1180) [2021-04-15 19:48:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][1170/1251] eta 0:00:23 lr 0.000906 time 0.2691 (0.2856) loss 4.5584 (3.8949) grad_norm 1.2430 (1.1189) [2021-04-15 19:49:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][1180/1251] eta 0:00:20 lr 0.000906 time 0.2707 (0.2856) loss 3.7462 (3.8947) grad_norm 1.0622 (1.1183) [2021-04-15 19:49:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][1190/1251] eta 0:00:17 lr 0.000906 time 0.2608 (0.2855) loss 3.7594 (3.8951) grad_norm 1.1056 (1.1176) [2021-04-15 19:49:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][1200/1251] eta 0:00:14 lr 0.000906 time 0.2548 (0.2854) loss 3.9253 (3.8972) grad_norm 1.3041 (1.1171) [2021-04-15 19:49:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][1210/1251] eta 0:00:11 lr 0.000906 time 0.2741 (0.2854) loss 4.1805 (3.8969) grad_norm 1.0013 (1.1167) [2021-04-15 19:49:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][1220/1251] eta 0:00:08 lr 0.000906 time 0.2559 (0.2854) loss 4.1956 (3.8958) grad_norm 1.0507 (1.1166) [2021-04-15 19:49:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][1230/1251] eta 0:00:05 lr 0.000906 time 0.2717 (0.2855) loss 3.8219 (3.8973) grad_norm 1.0659 (1.1168) [2021-04-15 19:49:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][1240/1251] eta 0:00:03 lr 0.000905 time 0.2483 (0.2853) loss 3.1258 (3.8988) grad_norm 1.0869 (1.1164) [2021-04-15 19:49:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [59/300][1250/1251] eta 0:00:00 lr 0.000905 time 0.2603 (0.2850) loss 3.6038 (3.8985) grad_norm 0.9701 (1.1158) [2021-04-15 19:49:25 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 59 training takes 0:05:59 [2021-04-15 19:49:25 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_59.pth saving...... [2021-04-15 19:49:41 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_59.pth saved !!! [2021-04-15 19:49:42 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.279 (1.279) Loss 1.3207 (1.3207) Acc@1 69.922 (69.922) Acc@5 90.137 (90.137) [2021-04-15 19:49:43 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.112 (0.238) Loss 1.1159 (1.2500) Acc@1 73.145 (70.872) Acc@5 92.383 (90.767) [2021-04-15 19:49:46 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.289 (0.262) Loss 1.2031 (1.2628) Acc@1 70.703 (70.452) Acc@5 91.992 (90.569) [2021-04-15 19:49:48 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.136 (0.234) Loss 1.1788 (1.2487) Acc@1 73.047 (70.839) Acc@5 91.895 (90.738) [2021-04-15 19:49:50 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.221) Loss 1.2326 (1.2467) Acc@1 72.070 (71.010) Acc@5 90.430 (90.770) [2021-04-15 19:49:53 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 71.060 Acc@5 90.720 [2021-04-15 19:49:53 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 71.1% [2021-04-15 19:49:53 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 71.06% [2021-04-15 19:49:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][0/1251] eta 1:35:01 lr 0.000905 time 4.5574 (4.5574) loss 4.7863 (4.7863) grad_norm 1.2815 (1.2815) [2021-04-15 19:50:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][10/1251] eta 0:13:49 lr 0.000905 time 0.2945 (0.6685) loss 4.0009 (3.9416) grad_norm 1.2046 (1.2122) [2021-04-15 19:50:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][20/1251] eta 0:09:56 lr 0.000905 time 0.2697 (0.4848) loss 4.7627 (4.0801) grad_norm 1.0272 (1.1932) [2021-04-15 19:50:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][30/1251] eta 0:08:32 lr 0.000905 time 0.2799 (0.4193) loss 4.2738 (4.1307) grad_norm 1.0719 (1.1632) [2021-04-15 19:50:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][40/1251] eta 0:07:47 lr 0.000905 time 0.2633 (0.3863) loss 4.3543 (4.1004) grad_norm 0.9708 (1.1455) [2021-04-15 19:50:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][50/1251] eta 0:07:20 lr 0.000905 time 0.2726 (0.3669) loss 3.4568 (4.0140) grad_norm 0.9644 (1.1490) [2021-04-15 19:50:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][60/1251] eta 0:06:59 lr 0.000905 time 0.2886 (0.3525) loss 4.4220 (4.0202) grad_norm 1.1977 (1.1396) [2021-04-15 19:50:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][70/1251] eta 0:06:44 lr 0.000905 time 0.2823 (0.3424) loss 3.3419 (4.0120) grad_norm 0.9969 (1.1368) [2021-04-15 19:50:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][80/1251] eta 0:06:31 lr 0.000905 time 0.2656 (0.3342) loss 4.6966 (4.0033) grad_norm 1.1151 (1.1404) [2021-04-15 19:50:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][90/1251] eta 0:06:21 lr 0.000905 time 0.2579 (0.3284) loss 4.4067 (4.0417) grad_norm 1.2296 (1.1383) [2021-04-15 19:50:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][100/1251] eta 0:06:12 lr 0.000905 time 0.2651 (0.3238) loss 3.9766 (4.0176) grad_norm 1.1942 (1.1398) [2021-04-15 19:50:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][110/1251] eta 0:06:05 lr 0.000905 time 0.2700 (0.3201) loss 4.4723 (4.0334) grad_norm 1.1281 (1.1319) [2021-04-15 19:50:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][120/1251] eta 0:05:58 lr 0.000905 time 0.2769 (0.3169) loss 4.3180 (4.0128) grad_norm 0.9232 (1.1236) [2021-04-15 19:50:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][130/1251] eta 0:05:52 lr 0.000905 time 0.2939 (0.3146) loss 3.8847 (4.0189) grad_norm 1.0927 (1.1184) [2021-04-15 19:50:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][140/1251] eta 0:05:47 lr 0.000905 time 0.2857 (0.3127) loss 3.4602 (3.9909) grad_norm 1.1766 (1.1182) [2021-04-15 19:50:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][150/1251] eta 0:05:44 lr 0.000905 time 0.4172 (0.3128) loss 3.2127 (3.9725) grad_norm 0.8940 (1.1174) [2021-04-15 19:50:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][160/1251] eta 0:05:39 lr 0.000905 time 0.2625 (0.3108) loss 3.1972 (3.9544) grad_norm 0.9308 (1.1150) [2021-04-15 19:50:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][170/1251] eta 0:05:34 lr 0.000905 time 0.2922 (0.3095) loss 4.3808 (3.9418) grad_norm 1.1699 (1.1121) [2021-04-15 19:50:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][180/1251] eta 0:05:30 lr 0.000905 time 0.2804 (0.3088) loss 4.5916 (3.9398) grad_norm 1.2358 (1.1161) [2021-04-15 19:50:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][190/1251] eta 0:05:26 lr 0.000905 time 0.3183 (0.3077) loss 3.8631 (3.9525) grad_norm 1.4132 (1.1183) [2021-04-15 19:50:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][200/1251] eta 0:05:22 lr 0.000905 time 0.2808 (0.3064) loss 4.2995 (3.9440) grad_norm 1.1005 (1.1194) [2021-04-15 19:50:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][210/1251] eta 0:05:17 lr 0.000905 time 0.2744 (0.3052) loss 3.4701 (3.9470) grad_norm 1.1029 (1.1172) [2021-04-15 19:51:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][220/1251] eta 0:05:13 lr 0.000905 time 0.2694 (0.3040) loss 4.6273 (3.9443) grad_norm 1.0328 (1.1173) [2021-04-15 19:51:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][230/1251] eta 0:05:09 lr 0.000905 time 0.2700 (0.3035) loss 4.2502 (3.9509) grad_norm 1.1245 (1.1154) [2021-04-15 19:51:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][240/1251] eta 0:05:05 lr 0.000905 time 0.2581 (0.3025) loss 3.9708 (3.9322) grad_norm 1.1162 (1.1196) [2021-04-15 19:51:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][250/1251] eta 0:05:01 lr 0.000905 time 0.2791 (0.3016) loss 4.3860 (3.9267) grad_norm 1.0683 (1.1205) [2021-04-15 19:51:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][260/1251] eta 0:04:58 lr 0.000905 time 0.2829 (0.3009) loss 3.8724 (3.9315) grad_norm 1.0656 (1.1194) [2021-04-15 19:51:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][270/1251] eta 0:04:54 lr 0.000905 time 0.3014 (0.3002) loss 3.1565 (3.9270) grad_norm 1.0655 (1.1223) [2021-04-15 19:51:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][280/1251] eta 0:04:50 lr 0.000905 time 0.2916 (0.2995) loss 4.3070 (3.9203) grad_norm 1.2471 (1.1226) [2021-04-15 19:51:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][290/1251] eta 0:04:47 lr 0.000905 time 0.2996 (0.2990) loss 3.2692 (3.9103) grad_norm 1.2967 (1.1255) [2021-04-15 19:51:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][300/1251] eta 0:04:43 lr 0.000905 time 0.2740 (0.2984) loss 3.3528 (3.9107) grad_norm 0.9785 (1.1255) [2021-04-15 19:51:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][310/1251] eta 0:04:40 lr 0.000905 time 0.2807 (0.2977) loss 4.7316 (3.9115) grad_norm 1.1330 (1.1251) [2021-04-15 19:51:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][320/1251] eta 0:04:36 lr 0.000905 time 0.2810 (0.2973) loss 4.6396 (3.9122) grad_norm 1.1331 (1.1257) [2021-04-15 19:51:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][330/1251] eta 0:04:33 lr 0.000905 time 0.2753 (0.2968) loss 3.4310 (3.9080) grad_norm 1.0057 (1.1251) [2021-04-15 19:51:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][340/1251] eta 0:04:29 lr 0.000905 time 0.2569 (0.2963) loss 3.7809 (3.9067) grad_norm 1.1136 (1.1242) [2021-04-15 19:51:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][350/1251] eta 0:04:26 lr 0.000905 time 0.2714 (0.2958) loss 2.9835 (3.9076) grad_norm 1.1865 (1.1260) [2021-04-15 19:51:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][360/1251] eta 0:04:23 lr 0.000905 time 0.2748 (0.2960) loss 3.2369 (3.9068) grad_norm 1.0571 (1.1268) [2021-04-15 19:51:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][370/1251] eta 0:04:20 lr 0.000905 time 0.2674 (0.2955) loss 4.1828 (3.9059) grad_norm 1.1316 (1.1253) [2021-04-15 19:51:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][380/1251] eta 0:04:17 lr 0.000905 time 0.2782 (0.2951) loss 3.6778 (3.9076) grad_norm 1.2158 (1.1260) [2021-04-15 19:51:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][390/1251] eta 0:04:13 lr 0.000905 time 0.2754 (0.2949) loss 4.0390 (3.9074) grad_norm 1.1586 (1.1259) [2021-04-15 19:51:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][400/1251] eta 0:04:11 lr 0.000904 time 0.2777 (0.2951) loss 3.8872 (3.9038) grad_norm 1.0016 (1.1253) [2021-04-15 19:51:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][410/1251] eta 0:04:07 lr 0.000904 time 0.2891 (0.2947) loss 4.2783 (3.9004) grad_norm 0.9327 (1.1250) [2021-04-15 19:51:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][420/1251] eta 0:04:04 lr 0.000904 time 0.3119 (0.2945) loss 4.7886 (3.9010) grad_norm 1.0454 (1.1227) [2021-04-15 19:52:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][430/1251] eta 0:04:01 lr 0.000904 time 0.2764 (0.2946) loss 3.8993 (3.9014) grad_norm 1.3024 (1.1233) [2021-04-15 19:52:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][440/1251] eta 0:03:58 lr 0.000904 time 0.2612 (0.2943) loss 2.7337 (3.8954) grad_norm 1.2705 (1.1235) [2021-04-15 19:52:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][450/1251] eta 0:03:55 lr 0.000904 time 0.2679 (0.2940) loss 3.8630 (3.8947) grad_norm 0.9218 (1.1225) [2021-04-15 19:52:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][460/1251] eta 0:03:52 lr 0.000904 time 0.2956 (0.2937) loss 3.0525 (3.8883) grad_norm 1.0726 (1.1216) [2021-04-15 19:52:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][470/1251] eta 0:03:49 lr 0.000904 time 0.2754 (0.2937) loss 3.6511 (3.8901) grad_norm 1.1301 (1.1226) [2021-04-15 19:52:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][480/1251] eta 0:03:46 lr 0.000904 time 0.2930 (0.2935) loss 4.5960 (3.8916) grad_norm 1.1890 (1.1212) [2021-04-15 19:52:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][490/1251] eta 0:03:43 lr 0.000904 time 0.2719 (0.2932) loss 3.4173 (3.8944) grad_norm 0.9413 (1.1200) [2021-04-15 19:52:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][500/1251] eta 0:03:39 lr 0.000904 time 0.2932 (0.2929) loss 2.8977 (3.8973) grad_norm 1.1202 (1.1207) [2021-04-15 19:52:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][510/1251] eta 0:03:36 lr 0.000904 time 0.2950 (0.2926) loss 4.1430 (3.8989) grad_norm 0.9617 (1.1219) [2021-04-15 19:52:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][520/1251] eta 0:03:33 lr 0.000904 time 0.2649 (0.2927) loss 5.0242 (3.8962) grad_norm 1.2124 (1.1231) [2021-04-15 19:52:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][530/1251] eta 0:03:30 lr 0.000904 time 0.2750 (0.2926) loss 4.3293 (3.8995) grad_norm 0.9909 (1.1233) [2021-04-15 19:52:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][540/1251] eta 0:03:27 lr 0.000904 time 0.2701 (0.2925) loss 4.4568 (3.9024) grad_norm 0.9329 (1.1250) [2021-04-15 19:52:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][550/1251] eta 0:03:24 lr 0.000904 time 0.2849 (0.2922) loss 3.8187 (3.9054) grad_norm 0.9802 (1.1254) [2021-04-15 19:52:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][560/1251] eta 0:03:21 lr 0.000904 time 0.2682 (0.2920) loss 3.8233 (3.9055) grad_norm 1.1818 (1.1245) [2021-04-15 19:52:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][570/1251] eta 0:03:19 lr 0.000904 time 0.2686 (0.2923) loss 2.7907 (3.9037) grad_norm 1.0475 (1.1233) [2021-04-15 19:52:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][580/1251] eta 0:03:16 lr 0.000904 time 0.2613 (0.2923) loss 3.9169 (3.8986) grad_norm 1.2842 (1.1237) [2021-04-15 19:52:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][590/1251] eta 0:03:13 lr 0.000904 time 0.2708 (0.2922) loss 3.3616 (3.9012) grad_norm 1.0773 (1.1234) [2021-04-15 19:52:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][600/1251] eta 0:03:10 lr 0.000904 time 0.2589 (0.2920) loss 4.4456 (3.8973) grad_norm 0.9992 (1.1236) [2021-04-15 19:52:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][610/1251] eta 0:03:07 lr 0.000904 time 0.2696 (0.2918) loss 2.7290 (3.8998) grad_norm 1.2987 (1.1246) [2021-04-15 19:52:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][620/1251] eta 0:03:04 lr 0.000904 time 0.2805 (0.2917) loss 3.7512 (3.9018) grad_norm 1.0058 (1.1245) [2021-04-15 19:52:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][630/1251] eta 0:03:01 lr 0.000904 time 0.2879 (0.2915) loss 3.3564 (3.9008) grad_norm 1.0079 (1.1248) [2021-04-15 19:52:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][640/1251] eta 0:02:58 lr 0.000904 time 0.2847 (0.2914) loss 4.6900 (3.9001) grad_norm 1.1199 (1.1242) [2021-04-15 19:53:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][650/1251] eta 0:02:55 lr 0.000904 time 0.2966 (0.2913) loss 3.9636 (3.9005) grad_norm 1.2560 (1.1256) [2021-04-15 19:53:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][660/1251] eta 0:02:52 lr 0.000904 time 0.2856 (0.2912) loss 3.9106 (3.8980) grad_norm 1.0375 (1.1252) [2021-04-15 19:53:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][670/1251] eta 0:02:49 lr 0.000904 time 0.2629 (0.2910) loss 4.2936 (3.8980) grad_norm 1.0662 (1.1239) [2021-04-15 19:53:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][680/1251] eta 0:02:46 lr 0.000904 time 0.2840 (0.2910) loss 3.2203 (3.8967) grad_norm 1.2431 (1.1234) [2021-04-15 19:53:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][690/1251] eta 0:02:43 lr 0.000904 time 0.2861 (0.2909) loss 4.2811 (3.9002) grad_norm 1.0690 (1.1226) [2021-04-15 19:53:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][700/1251] eta 0:02:40 lr 0.000904 time 0.2917 (0.2908) loss 4.0013 (3.8990) grad_norm 1.0834 (1.1246) [2021-04-15 19:53:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][710/1251] eta 0:02:37 lr 0.000904 time 0.2810 (0.2907) loss 3.8645 (3.8954) grad_norm 1.0924 (1.1252) [2021-04-15 19:53:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][720/1251] eta 0:02:34 lr 0.000904 time 0.2819 (0.2907) loss 2.6290 (3.8959) grad_norm 0.8950 (1.1241) [2021-04-15 19:53:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][730/1251] eta 0:02:31 lr 0.000904 time 0.2889 (0.2908) loss 4.2112 (3.8928) grad_norm 1.0641 (1.1236) [2021-04-15 19:53:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][740/1251] eta 0:02:28 lr 0.000904 time 0.2971 (0.2906) loss 3.3196 (3.8896) grad_norm 1.0271 (1.1229) [2021-04-15 19:53:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][750/1251] eta 0:02:25 lr 0.000904 time 0.2487 (0.2906) loss 4.4794 (3.8899) grad_norm 1.1689 (1.1234) [2021-04-15 19:53:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][760/1251] eta 0:02:22 lr 0.000904 time 0.2688 (0.2906) loss 3.2682 (3.8915) grad_norm 1.2421 (1.1244) [2021-04-15 19:53:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][770/1251] eta 0:02:19 lr 0.000904 time 0.3020 (0.2906) loss 3.7891 (3.8966) grad_norm 1.1831 (1.1247) [2021-04-15 19:53:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][780/1251] eta 0:02:16 lr 0.000904 time 0.3189 (0.2908) loss 3.5120 (3.8935) grad_norm 1.2560 (1.1253) [2021-04-15 19:53:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][790/1251] eta 0:02:14 lr 0.000904 time 0.2539 (0.2907) loss 3.9364 (3.8919) grad_norm 1.1409 (1.1259) [2021-04-15 19:53:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][800/1251] eta 0:02:11 lr 0.000904 time 0.2533 (0.2907) loss 4.2979 (3.8923) grad_norm 1.0722 (1.1253) [2021-04-15 19:53:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][810/1251] eta 0:02:08 lr 0.000903 time 0.2727 (0.2908) loss 3.1092 (3.8927) grad_norm 1.1034 (1.1250) [2021-04-15 19:53:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][820/1251] eta 0:02:05 lr 0.000903 time 0.3019 (0.2907) loss 3.5266 (3.8925) grad_norm 1.3919 (1.1250) [2021-04-15 19:53:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][830/1251] eta 0:02:02 lr 0.000903 time 0.2757 (0.2906) loss 3.5115 (3.8947) grad_norm 1.0533 (1.1243) [2021-04-15 19:53:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][840/1251] eta 0:01:59 lr 0.000903 time 0.2925 (0.2905) loss 2.9368 (3.8916) grad_norm 1.0270 (1.1244) [2021-04-15 19:54:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][850/1251] eta 0:01:56 lr 0.000903 time 0.2865 (0.2905) loss 4.2322 (3.8924) grad_norm 0.9646 (1.1242) [2021-04-15 19:54:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][860/1251] eta 0:01:53 lr 0.000903 time 0.3115 (0.2905) loss 4.0615 (3.8924) grad_norm 1.2900 (1.1244) [2021-04-15 19:54:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][870/1251] eta 0:01:50 lr 0.000903 time 0.3081 (0.2904) loss 3.7340 (3.8919) grad_norm 1.1260 (1.1241) [2021-04-15 19:54:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][880/1251] eta 0:01:47 lr 0.000903 time 0.2960 (0.2903) loss 2.7560 (3.8899) grad_norm 1.0805 (1.1252) [2021-04-15 19:54:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][890/1251] eta 0:01:44 lr 0.000903 time 0.2868 (0.2902) loss 2.9390 (3.8873) grad_norm 1.0875 (inf) [2021-04-15 19:54:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][900/1251] eta 0:01:41 lr 0.000903 time 0.2705 (0.2901) loss 3.5126 (3.8854) grad_norm 1.3362 (inf) [2021-04-15 19:54:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][910/1251] eta 0:01:38 lr 0.000903 time 0.2842 (0.2901) loss 4.1309 (3.8887) grad_norm 1.0908 (inf) [2021-04-15 19:54:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][920/1251] eta 0:01:35 lr 0.000903 time 0.2581 (0.2900) loss 3.4884 (3.8863) grad_norm 1.1096 (inf) [2021-04-15 19:54:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][930/1251] eta 0:01:33 lr 0.000903 time 0.2880 (0.2901) loss 3.0397 (3.8840) grad_norm 1.0764 (inf) [2021-04-15 19:54:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][940/1251] eta 0:01:30 lr 0.000903 time 0.3062 (0.2900) loss 3.2499 (3.8821) grad_norm 1.0408 (inf) [2021-04-15 19:54:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][950/1251] eta 0:01:27 lr 0.000903 time 0.2955 (0.2900) loss 4.5740 (3.8801) grad_norm 1.0000 (inf) [2021-04-15 19:54:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][960/1251] eta 0:01:24 lr 0.000903 time 0.2654 (0.2899) loss 4.1788 (3.8796) grad_norm 1.1627 (inf) [2021-04-15 19:54:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][970/1251] eta 0:01:21 lr 0.000903 time 0.3604 (0.2899) loss 3.3091 (3.8816) grad_norm 1.0108 (inf) [2021-04-15 19:54:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][980/1251] eta 0:01:18 lr 0.000903 time 0.3049 (0.2899) loss 4.2260 (3.8810) grad_norm 1.0172 (inf) [2021-04-15 19:54:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][990/1251] eta 0:01:15 lr 0.000903 time 0.2703 (0.2898) loss 3.8813 (3.8816) grad_norm 1.2617 (inf) [2021-04-15 19:54:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][1000/1251] eta 0:01:12 lr 0.000903 time 0.2868 (0.2898) loss 4.6767 (3.8804) grad_norm 0.9685 (inf) [2021-04-15 19:54:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][1010/1251] eta 0:01:09 lr 0.000903 time 0.2755 (0.2899) loss 4.0212 (3.8820) grad_norm 1.2124 (inf) [2021-04-15 19:54:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][1020/1251] eta 0:01:06 lr 0.000903 time 0.2939 (0.2898) loss 4.1868 (3.8859) grad_norm 1.1330 (inf) [2021-04-15 19:54:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][1030/1251] eta 0:01:04 lr 0.000903 time 0.2606 (0.2898) loss 4.0051 (3.8878) grad_norm 1.1812 (inf) [2021-04-15 19:54:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][1040/1251] eta 0:01:01 lr 0.000903 time 0.2843 (0.2899) loss 3.5737 (3.8859) grad_norm 1.0943 (inf) [2021-04-15 19:54:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][1050/1251] eta 0:00:58 lr 0.000903 time 0.2763 (0.2898) loss 3.6845 (3.8868) grad_norm 1.0812 (inf) [2021-04-15 19:55:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][1060/1251] eta 0:00:55 lr 0.000903 time 0.2488 (0.2897) loss 3.9432 (3.8854) grad_norm 1.3602 (inf) [2021-04-15 19:55:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][1070/1251] eta 0:00:52 lr 0.000903 time 0.2696 (0.2896) loss 3.3179 (3.8837) grad_norm 0.9612 (inf) [2021-04-15 19:55:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][1080/1251] eta 0:00:49 lr 0.000903 time 0.2770 (0.2895) loss 4.3806 (3.8829) grad_norm 1.1734 (inf) [2021-04-15 19:55:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][1090/1251] eta 0:00:46 lr 0.000903 time 0.2885 (0.2897) loss 4.1444 (3.8836) grad_norm 1.0579 (inf) [2021-04-15 19:55:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][1100/1251] eta 0:00:43 lr 0.000903 time 0.3181 (0.2896) loss 2.9133 (3.8829) grad_norm 1.0683 (inf) [2021-04-15 19:55:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][1110/1251] eta 0:00:40 lr 0.000903 time 0.2862 (0.2896) loss 3.2570 (3.8815) grad_norm 1.1982 (inf) [2021-04-15 19:55:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][1120/1251] eta 0:00:37 lr 0.000903 time 0.2668 (0.2895) loss 3.7615 (3.8838) grad_norm 1.1203 (inf) [2021-04-15 19:55:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][1130/1251] eta 0:00:35 lr 0.000903 time 0.3014 (0.2896) loss 2.7561 (3.8827) grad_norm 1.3420 (inf) [2021-04-15 19:55:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][1140/1251] eta 0:00:32 lr 0.000903 time 0.2650 (0.2896) loss 3.6378 (3.8815) grad_norm 1.1990 (inf) [2021-04-15 19:55:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][1150/1251] eta 0:00:29 lr 0.000903 time 0.2766 (0.2896) loss 4.0866 (3.8813) grad_norm 1.1093 (inf) [2021-04-15 19:55:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][1160/1251] eta 0:00:26 lr 0.000903 time 0.2856 (0.2897) loss 2.8302 (3.8818) grad_norm 1.4480 (inf) [2021-04-15 19:55:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][1170/1251] eta 0:00:23 lr 0.000903 time 0.2527 (0.2896) loss 2.7942 (3.8790) grad_norm 1.3811 (inf) [2021-04-15 19:55:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][1180/1251] eta 0:00:20 lr 0.000903 time 0.2828 (0.2895) loss 3.7888 (3.8784) grad_norm 0.9867 (inf) [2021-04-15 19:55:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][1190/1251] eta 0:00:17 lr 0.000903 time 0.3036 (0.2895) loss 4.2865 (3.8774) grad_norm 0.9638 (inf) [2021-04-15 19:55:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][1200/1251] eta 0:00:14 lr 0.000903 time 0.2998 (0.2894) loss 4.5424 (3.8775) grad_norm 1.0482 (inf) [2021-04-15 19:55:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][1210/1251] eta 0:00:11 lr 0.000902 time 0.3261 (0.2894) loss 4.3005 (3.8768) grad_norm 1.3398 (inf) [2021-04-15 19:55:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][1220/1251] eta 0:00:08 lr 0.000902 time 0.3272 (0.2894) loss 3.9035 (3.8783) grad_norm 1.0363 (inf) [2021-04-15 19:55:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][1230/1251] eta 0:00:06 lr 0.000902 time 0.2921 (0.2894) loss 3.4750 (3.8774) grad_norm 1.2785 (inf) [2021-04-15 19:55:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][1240/1251] eta 0:00:03 lr 0.000902 time 0.2482 (0.2892) loss 3.7870 (3.8795) grad_norm 1.0496 (inf) [2021-04-15 19:55:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [60/300][1250/1251] eta 0:00:00 lr 0.000902 time 0.2488 (0.2889) loss 4.3441 (3.8797) grad_norm 0.9270 (inf) [2021-04-15 19:55:56 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 60 training takes 0:06:03 [2021-04-15 19:55:56 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_60.pth saving...... [2021-04-15 19:56:13 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_60.pth saved !!! [2021-04-15 19:56:14 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.132 (1.132) Loss 1.2110 (1.2110) Acc@1 72.070 (72.070) Acc@5 90.234 (90.234) [2021-04-15 19:56:15 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.144 (0.241) Loss 1.2691 (1.2237) Acc@1 71.094 (71.493) Acc@5 89.453 (90.767) [2021-04-15 19:56:18 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.111 (0.252) Loss 1.1838 (1.2104) Acc@1 70.996 (71.568) Acc@5 92.090 (91.076) [2021-04-15 19:56:19 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.161 (0.215) Loss 1.1228 (1.2192) Acc@1 72.656 (71.412) Acc@5 91.992 (90.934) [2021-04-15 19:56:22 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.111 (0.218) Loss 1.3321 (1.2247) Acc@1 68.652 (71.341) Acc@5 89.258 (90.808) [2021-04-15 19:56:24 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 71.272 Acc@5 90.772 [2021-04-15 19:56:24 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 71.3% [2021-04-15 19:56:24 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 71.27% [2021-04-15 19:56:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][0/1251] eta 1:44:59 lr 0.000902 time 5.0353 (5.0353) loss 3.4981 (3.4981) grad_norm 1.0173 (1.0173) [2021-04-15 19:56:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][10/1251] eta 0:14:38 lr 0.000902 time 0.2533 (0.7083) loss 4.0721 (3.6080) grad_norm 1.0607 (1.0719) [2021-04-15 19:56:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][20/1251] eta 0:10:33 lr 0.000902 time 0.2966 (0.5149) loss 4.1903 (3.6655) grad_norm 1.0357 (1.0864) [2021-04-15 19:56:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][30/1251] eta 0:08:57 lr 0.000902 time 0.2824 (0.4403) loss 3.3268 (3.7177) grad_norm 1.0047 (1.1017) [2021-04-15 19:56:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][40/1251] eta 0:08:04 lr 0.000902 time 0.2977 (0.4000) loss 3.7570 (3.8076) grad_norm 1.1611 (1.1125) [2021-04-15 19:56:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][50/1251] eta 0:07:32 lr 0.000902 time 0.2946 (0.3771) loss 3.9115 (3.8284) grad_norm 1.3806 (1.1162) [2021-04-15 19:56:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][60/1251] eta 0:07:10 lr 0.000902 time 0.2837 (0.3612) loss 4.4913 (3.9001) grad_norm 1.0139 (1.1169) [2021-04-15 19:56:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][70/1251] eta 0:06:53 lr 0.000902 time 0.2806 (0.3499) loss 4.2532 (3.9036) grad_norm 0.9778 (1.1054) [2021-04-15 19:56:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][80/1251] eta 0:06:39 lr 0.000902 time 0.2482 (0.3409) loss 2.9199 (3.8976) grad_norm 1.1367 (1.1041) [2021-04-15 19:56:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][90/1251] eta 0:06:28 lr 0.000902 time 0.2619 (0.3342) loss 4.1889 (3.9359) grad_norm 1.0157 (1.1149) [2021-04-15 19:56:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][100/1251] eta 0:06:18 lr 0.000902 time 0.2809 (0.3288) loss 4.1818 (3.9292) grad_norm 1.0159 (1.1116) [2021-04-15 19:57:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][110/1251] eta 0:06:10 lr 0.000902 time 0.2838 (0.3244) loss 4.2578 (3.9091) grad_norm 0.9364 (1.1107) [2021-04-15 19:57:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][120/1251] eta 0:06:02 lr 0.000902 time 0.2897 (0.3207) loss 3.8281 (3.9034) grad_norm 1.4645 (1.1182) [2021-04-15 19:57:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][130/1251] eta 0:05:56 lr 0.000902 time 0.3111 (0.3178) loss 3.6980 (3.8978) grad_norm 1.1026 (1.1255) [2021-04-15 19:57:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][140/1251] eta 0:05:52 lr 0.000902 time 0.3091 (0.3172) loss 5.2125 (3.9252) grad_norm 1.2871 (1.1281) [2021-04-15 19:57:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][150/1251] eta 0:05:46 lr 0.000902 time 0.3116 (0.3148) loss 3.9589 (3.9279) grad_norm 1.0156 (1.1242) [2021-04-15 19:57:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][160/1251] eta 0:05:41 lr 0.000902 time 0.2825 (0.3129) loss 3.7998 (3.9217) grad_norm 1.0807 (1.1228) [2021-04-15 19:57:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][170/1251] eta 0:05:36 lr 0.000902 time 0.3110 (0.3111) loss 4.1283 (3.9105) grad_norm 1.1062 (1.1232) [2021-04-15 19:57:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][180/1251] eta 0:05:32 lr 0.000902 time 0.4591 (0.3105) loss 4.3283 (3.9021) grad_norm 1.0138 (1.1222) [2021-04-15 19:57:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][190/1251] eta 0:05:27 lr 0.000902 time 0.2689 (0.3088) loss 3.0719 (3.8944) grad_norm 1.1320 (1.1188) [2021-04-15 19:57:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][200/1251] eta 0:05:23 lr 0.000902 time 0.2488 (0.3075) loss 4.2317 (3.9052) grad_norm 1.0740 (1.1152) [2021-04-15 19:57:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][210/1251] eta 0:05:19 lr 0.000902 time 0.3114 (0.3065) loss 3.0371 (3.8959) grad_norm 1.0908 (1.1161) [2021-04-15 19:57:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][220/1251] eta 0:05:15 lr 0.000902 time 0.2684 (0.3061) loss 3.8748 (3.9028) grad_norm 0.9981 (1.1144) [2021-04-15 19:57:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][230/1251] eta 0:05:11 lr 0.000902 time 0.2627 (0.3052) loss 4.2573 (3.9050) grad_norm 1.2524 (1.1128) [2021-04-15 19:57:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][240/1251] eta 0:05:07 lr 0.000902 time 0.2474 (0.3042) loss 4.2620 (3.9082) grad_norm 1.2528 (1.1121) [2021-04-15 19:57:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][250/1251] eta 0:05:04 lr 0.000902 time 0.2746 (0.3039) loss 3.5456 (3.9121) grad_norm 0.9290 (1.1115) [2021-04-15 19:57:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][260/1251] eta 0:05:00 lr 0.000902 time 0.2643 (0.3029) loss 2.8616 (3.9122) grad_norm 1.0252 (1.1122) [2021-04-15 19:57:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][270/1251] eta 0:04:56 lr 0.000902 time 0.2759 (0.3021) loss 2.7411 (3.8984) grad_norm 0.9400 (1.1139) [2021-04-15 19:57:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][280/1251] eta 0:04:52 lr 0.000902 time 0.2681 (0.3014) loss 2.9720 (3.8936) grad_norm 1.0429 (1.1147) [2021-04-15 19:57:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][290/1251] eta 0:04:48 lr 0.000902 time 0.2829 (0.3006) loss 3.8730 (3.8947) grad_norm 1.2198 (1.1161) [2021-04-15 19:57:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][300/1251] eta 0:04:45 lr 0.000902 time 0.2653 (0.3003) loss 4.3506 (3.8991) grad_norm 1.7364 (1.1185) [2021-04-15 19:57:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][310/1251] eta 0:04:42 lr 0.000902 time 0.2989 (0.2999) loss 2.5457 (3.8870) grad_norm 1.1352 (1.1197) [2021-04-15 19:58:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][320/1251] eta 0:04:38 lr 0.000902 time 0.2733 (0.2994) loss 3.2480 (3.8818) grad_norm 1.0257 (1.1203) [2021-04-15 19:58:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][330/1251] eta 0:04:35 lr 0.000902 time 0.2793 (0.2992) loss 4.3403 (3.8702) grad_norm 1.0974 (1.1227) [2021-04-15 19:58:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][340/1251] eta 0:04:32 lr 0.000902 time 0.2683 (0.2992) loss 4.3393 (3.8716) grad_norm 1.2502 (1.1244) [2021-04-15 19:58:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][350/1251] eta 0:04:29 lr 0.000902 time 0.2618 (0.2990) loss 3.5892 (3.8722) grad_norm 1.2913 (1.1241) [2021-04-15 19:58:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][360/1251] eta 0:04:26 lr 0.000902 time 0.2769 (0.2989) loss 4.2574 (3.8700) grad_norm 1.0806 (1.1232) [2021-04-15 19:58:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][370/1251] eta 0:04:23 lr 0.000901 time 0.2791 (0.2988) loss 3.1991 (3.8647) grad_norm 1.0672 (1.1232) [2021-04-15 19:58:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][380/1251] eta 0:04:19 lr 0.000901 time 0.3105 (0.2984) loss 4.2058 (3.8672) grad_norm 1.0895 (1.1230) [2021-04-15 19:58:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][390/1251] eta 0:04:16 lr 0.000901 time 0.2874 (0.2980) loss 3.9436 (3.8656) grad_norm 1.3664 (1.1247) [2021-04-15 19:58:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][400/1251] eta 0:04:13 lr 0.000901 time 0.2722 (0.2975) loss 3.6987 (3.8713) grad_norm 1.0742 (1.1249) [2021-04-15 19:58:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][410/1251] eta 0:04:09 lr 0.000901 time 0.2700 (0.2971) loss 4.2650 (3.8741) grad_norm 1.0825 (1.1244) [2021-04-15 19:58:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][420/1251] eta 0:04:06 lr 0.000901 time 0.3016 (0.2966) loss 4.6523 (3.8766) grad_norm 1.0738 (1.1245) [2021-04-15 19:58:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][430/1251] eta 0:04:03 lr 0.000901 time 0.2788 (0.2962) loss 4.9221 (3.8760) grad_norm 0.9412 (1.1259) [2021-04-15 19:58:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][440/1251] eta 0:03:59 lr 0.000901 time 0.2943 (0.2958) loss 4.0820 (3.8774) grad_norm 1.0224 (1.1253) [2021-04-15 19:58:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][450/1251] eta 0:03:56 lr 0.000901 time 0.3279 (0.2956) loss 3.9570 (3.8744) grad_norm 1.1319 (1.1243) [2021-04-15 19:58:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][460/1251] eta 0:03:53 lr 0.000901 time 0.3004 (0.2954) loss 3.7761 (3.8734) grad_norm 0.9369 (1.1230) [2021-04-15 19:58:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][470/1251] eta 0:03:50 lr 0.000901 time 0.3003 (0.2952) loss 4.7813 (3.8765) grad_norm 0.9836 (1.1229) [2021-04-15 19:58:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][480/1251] eta 0:03:47 lr 0.000901 time 0.2461 (0.2950) loss 3.8356 (3.8726) grad_norm 1.3520 (1.1234) [2021-04-15 19:58:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][490/1251] eta 0:03:44 lr 0.000901 time 0.2963 (0.2949) loss 3.5753 (3.8749) grad_norm 1.1042 (1.1225) [2021-04-15 19:58:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][500/1251] eta 0:03:41 lr 0.000901 time 0.2699 (0.2945) loss 4.3171 (3.8817) grad_norm 0.9455 (1.1211) [2021-04-15 19:58:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][510/1251] eta 0:03:38 lr 0.000901 time 0.2692 (0.2943) loss 4.0831 (3.8845) grad_norm 1.0520 (1.1207) [2021-04-15 19:58:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][520/1251] eta 0:03:34 lr 0.000901 time 0.2773 (0.2939) loss 3.5897 (3.8901) grad_norm 1.0316 (1.1202) [2021-04-15 19:59:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][530/1251] eta 0:03:31 lr 0.000901 time 0.2590 (0.2938) loss 4.0320 (3.8921) grad_norm 1.3153 (1.1210) [2021-04-15 19:59:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][540/1251] eta 0:03:28 lr 0.000901 time 0.2760 (0.2936) loss 3.7630 (3.8879) grad_norm 1.0767 (1.1206) [2021-04-15 19:59:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][550/1251] eta 0:03:25 lr 0.000901 time 0.3054 (0.2934) loss 4.4310 (3.8873) grad_norm 1.1712 (1.1200) [2021-04-15 19:59:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][560/1251] eta 0:03:22 lr 0.000901 time 0.2601 (0.2931) loss 4.5628 (3.8913) grad_norm 1.0832 (1.1186) [2021-04-15 19:59:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][570/1251] eta 0:03:19 lr 0.000901 time 0.2763 (0.2932) loss 4.0923 (3.8921) grad_norm 1.0668 (1.1199) [2021-04-15 19:59:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][580/1251] eta 0:03:16 lr 0.000901 time 0.2752 (0.2934) loss 4.0946 (3.8908) grad_norm 1.0291 (1.1202) [2021-04-15 19:59:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][590/1251] eta 0:03:14 lr 0.000901 time 0.2602 (0.2935) loss 4.3478 (3.8908) grad_norm 1.0391 (1.1208) [2021-04-15 19:59:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][600/1251] eta 0:03:10 lr 0.000901 time 0.2686 (0.2933) loss 3.3881 (3.8868) grad_norm 0.9901 (1.1200) [2021-04-15 19:59:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][610/1251] eta 0:03:07 lr 0.000901 time 0.2893 (0.2931) loss 4.1936 (3.8829) grad_norm 1.0832 (1.1190) [2021-04-15 19:59:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][620/1251] eta 0:03:04 lr 0.000901 time 0.2509 (0.2929) loss 2.3154 (3.8825) grad_norm 1.1712 (1.1207) [2021-04-15 19:59:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][630/1251] eta 0:03:01 lr 0.000901 time 0.2866 (0.2928) loss 4.2580 (3.8776) grad_norm 1.3639 (1.1209) [2021-04-15 19:59:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][640/1251] eta 0:02:58 lr 0.000901 time 0.3446 (0.2927) loss 3.8801 (3.8805) grad_norm 1.1128 (1.1228) [2021-04-15 19:59:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][650/1251] eta 0:02:55 lr 0.000901 time 0.2486 (0.2924) loss 3.9983 (3.8779) grad_norm 0.9619 (1.1237) [2021-04-15 19:59:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][660/1251] eta 0:02:52 lr 0.000901 time 0.3011 (0.2924) loss 4.3262 (3.8827) grad_norm 1.1356 (1.1244) [2021-04-15 19:59:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][670/1251] eta 0:02:49 lr 0.000901 time 0.2663 (0.2922) loss 3.4806 (3.8793) grad_norm 1.3689 (1.1262) [2021-04-15 19:59:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][680/1251] eta 0:02:46 lr 0.000901 time 0.2624 (0.2921) loss 4.3147 (3.8809) grad_norm 0.9745 (1.1253) [2021-04-15 19:59:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][690/1251] eta 0:02:43 lr 0.000901 time 0.2702 (0.2919) loss 3.8050 (3.8814) grad_norm 1.2353 (1.1264) [2021-04-15 19:59:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][700/1251] eta 0:02:40 lr 0.000901 time 0.2796 (0.2918) loss 3.9448 (3.8785) grad_norm 0.9950 (1.1265) [2021-04-15 19:59:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][710/1251] eta 0:02:37 lr 0.000901 time 0.2747 (0.2917) loss 4.4732 (3.8775) grad_norm 1.0201 (1.1255) [2021-04-15 19:59:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][720/1251] eta 0:02:34 lr 0.000901 time 0.2652 (0.2918) loss 3.5976 (3.8775) grad_norm 1.0963 (1.1246) [2021-04-15 19:59:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][730/1251] eta 0:02:31 lr 0.000901 time 0.2888 (0.2917) loss 4.7677 (3.8785) grad_norm 0.9206 (1.1240) [2021-04-15 20:00:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][740/1251] eta 0:02:28 lr 0.000901 time 0.3046 (0.2916) loss 3.9204 (3.8767) grad_norm 1.1659 (1.1247) [2021-04-15 20:00:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][750/1251] eta 0:02:25 lr 0.000901 time 0.2648 (0.2914) loss 4.2644 (3.8713) grad_norm 1.0840 (1.1249) [2021-04-15 20:00:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][760/1251] eta 0:02:23 lr 0.000901 time 0.2877 (0.2915) loss 3.0744 (3.8735) grad_norm 1.0063 (1.1252) [2021-04-15 20:00:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][770/1251] eta 0:02:20 lr 0.000900 time 0.2761 (0.2913) loss 4.1344 (3.8762) grad_norm 1.2553 (1.1269) [2021-04-15 20:00:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][780/1251] eta 0:02:17 lr 0.000900 time 0.2584 (0.2912) loss 4.3009 (3.8775) grad_norm 1.0844 (1.1272) [2021-04-15 20:00:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][790/1251] eta 0:02:14 lr 0.000900 time 0.2907 (0.2913) loss 4.2490 (3.8798) grad_norm 1.0510 (1.1276) [2021-04-15 20:00:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][800/1251] eta 0:02:11 lr 0.000900 time 0.2906 (0.2911) loss 4.4274 (3.8773) grad_norm 0.9762 (1.1280) [2021-04-15 20:00:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][810/1251] eta 0:02:08 lr 0.000900 time 0.2739 (0.2909) loss 3.5999 (3.8753) grad_norm 0.9728 (1.1281) [2021-04-15 20:00:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][820/1251] eta 0:02:05 lr 0.000900 time 0.3001 (0.2908) loss 3.8207 (3.8764) grad_norm 0.9564 (1.1274) [2021-04-15 20:00:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][830/1251] eta 0:02:02 lr 0.000900 time 0.2832 (0.2908) loss 3.0599 (3.8732) grad_norm 1.1371 (1.1266) [2021-04-15 20:00:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][840/1251] eta 0:01:59 lr 0.000900 time 0.2878 (0.2907) loss 4.7194 (3.8745) grad_norm 1.1251 (1.1260) [2021-04-15 20:00:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][850/1251] eta 0:01:56 lr 0.000900 time 0.2960 (0.2907) loss 4.5104 (3.8698) grad_norm 1.2221 (1.1251) [2021-04-15 20:00:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][860/1251] eta 0:01:53 lr 0.000900 time 0.2911 (0.2906) loss 4.5067 (3.8669) grad_norm 1.0225 (1.1241) [2021-04-15 20:00:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][870/1251] eta 0:01:50 lr 0.000900 time 0.2957 (0.2905) loss 3.4105 (3.8647) grad_norm 1.3170 (1.1240) [2021-04-15 20:00:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][880/1251] eta 0:01:47 lr 0.000900 time 0.2812 (0.2905) loss 4.1290 (3.8624) grad_norm 0.9768 (1.1229) [2021-04-15 20:00:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][890/1251] eta 0:01:44 lr 0.000900 time 0.2638 (0.2904) loss 4.8939 (3.8695) grad_norm 1.0367 (1.1219) [2021-04-15 20:00:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][900/1251] eta 0:01:41 lr 0.000900 time 0.4321 (0.2905) loss 4.4510 (3.8710) grad_norm 1.1533 (1.1211) [2021-04-15 20:00:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][910/1251] eta 0:01:39 lr 0.000900 time 0.3165 (0.2904) loss 4.3566 (3.8695) grad_norm 1.0182 (1.1203) [2021-04-15 20:00:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][920/1251] eta 0:01:36 lr 0.000900 time 0.4202 (0.2907) loss 4.1905 (3.8727) grad_norm 1.1548 (1.1200) [2021-04-15 20:00:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][930/1251] eta 0:01:33 lr 0.000900 time 0.2844 (0.2905) loss 2.9595 (3.8725) grad_norm 1.2163 (1.1198) [2021-04-15 20:00:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][940/1251] eta 0:01:30 lr 0.000900 time 0.2877 (0.2908) loss 2.8760 (3.8728) grad_norm 1.4224 (1.1204) [2021-04-15 20:01:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][950/1251] eta 0:01:27 lr 0.000900 time 0.3065 (0.2907) loss 4.1943 (3.8743) grad_norm 0.9939 (1.1202) [2021-04-15 20:01:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][960/1251] eta 0:01:24 lr 0.000900 time 0.2797 (0.2906) loss 4.1201 (3.8772) grad_norm 1.0238 (1.1198) [2021-04-15 20:01:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][970/1251] eta 0:01:21 lr 0.000900 time 0.2599 (0.2905) loss 3.2246 (3.8768) grad_norm 0.9810 (1.1194) [2021-04-15 20:01:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][980/1251] eta 0:01:18 lr 0.000900 time 0.2936 (0.2905) loss 4.4006 (3.8751) grad_norm 1.0157 (1.1193) [2021-04-15 20:01:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][990/1251] eta 0:01:15 lr 0.000900 time 0.2826 (0.2904) loss 2.9905 (3.8753) grad_norm 0.9527 (1.1187) [2021-04-15 20:01:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][1000/1251] eta 0:01:12 lr 0.000900 time 0.2711 (0.2903) loss 3.0684 (3.8732) grad_norm 1.0275 (1.1187) [2021-04-15 20:01:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][1010/1251] eta 0:01:09 lr 0.000900 time 0.3032 (0.2902) loss 3.9455 (3.8717) grad_norm 1.1657 (1.1185) [2021-04-15 20:01:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][1020/1251] eta 0:01:07 lr 0.000900 time 0.2722 (0.2902) loss 3.0054 (3.8693) grad_norm 1.1368 (1.1186) [2021-04-15 20:01:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][1030/1251] eta 0:01:04 lr 0.000900 time 0.2818 (0.2901) loss 3.4170 (3.8678) grad_norm 1.1323 (1.1188) [2021-04-15 20:01:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][1040/1251] eta 0:01:01 lr 0.000900 time 0.2853 (0.2901) loss 4.2296 (3.8680) grad_norm 1.0184 (1.1186) [2021-04-15 20:01:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][1050/1251] eta 0:00:58 lr 0.000900 time 0.2868 (0.2900) loss 4.1297 (3.8680) grad_norm 1.3788 (1.1186) [2021-04-15 20:01:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][1060/1251] eta 0:00:55 lr 0.000900 time 0.2919 (0.2899) loss 3.6882 (3.8660) grad_norm 1.1520 (1.1182) [2021-04-15 20:01:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][1070/1251] eta 0:00:52 lr 0.000900 time 0.2736 (0.2899) loss 4.4032 (3.8669) grad_norm 1.1162 (1.1200) [2021-04-15 20:01:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][1080/1251] eta 0:00:49 lr 0.000900 time 0.2656 (0.2897) loss 4.8273 (3.8645) grad_norm 1.1949 (1.1196) [2021-04-15 20:01:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][1090/1251] eta 0:00:46 lr 0.000900 time 0.2634 (0.2897) loss 3.9056 (3.8629) grad_norm 1.0241 (1.1192) [2021-04-15 20:01:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][1100/1251] eta 0:00:43 lr 0.000900 time 0.2824 (0.2896) loss 2.4452 (3.8628) grad_norm 1.0808 (1.1190) [2021-04-15 20:01:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][1110/1251] eta 0:00:40 lr 0.000900 time 0.2961 (0.2895) loss 4.0536 (3.8630) grad_norm 1.2958 (1.1189) [2021-04-15 20:01:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][1120/1251] eta 0:00:37 lr 0.000900 time 0.2914 (0.2895) loss 3.7620 (3.8630) grad_norm 1.0708 (1.1196) [2021-04-15 20:01:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][1130/1251] eta 0:00:35 lr 0.000900 time 0.2787 (0.2895) loss 4.6253 (3.8664) grad_norm 1.0174 (1.1195) [2021-04-15 20:01:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][1140/1251] eta 0:00:32 lr 0.000900 time 0.2672 (0.2895) loss 3.6443 (3.8664) grad_norm 1.1333 (1.1192) [2021-04-15 20:01:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][1150/1251] eta 0:00:29 lr 0.000900 time 0.3004 (0.2894) loss 4.4003 (3.8669) grad_norm 0.9791 (1.1189) [2021-04-15 20:02:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][1160/1251] eta 0:00:26 lr 0.000900 time 0.2825 (0.2895) loss 4.2068 (3.8678) grad_norm 1.0196 (1.1182) [2021-04-15 20:02:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][1170/1251] eta 0:00:23 lr 0.000899 time 0.2819 (0.2896) loss 4.1623 (3.8685) grad_norm 0.9446 (1.1191) [2021-04-15 20:02:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][1180/1251] eta 0:00:20 lr 0.000899 time 0.3111 (0.2895) loss 4.3234 (3.8674) grad_norm 1.1437 (1.1196) [2021-04-15 20:02:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][1190/1251] eta 0:00:17 lr 0.000899 time 0.2990 (0.2895) loss 3.4503 (3.8660) grad_norm 1.1154 (1.1194) [2021-04-15 20:02:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][1200/1251] eta 0:00:14 lr 0.000899 time 0.2847 (0.2894) loss 4.1163 (3.8658) grad_norm 1.2180 (1.1195) [2021-04-15 20:02:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][1210/1251] eta 0:00:11 lr 0.000899 time 0.3043 (0.2894) loss 4.5808 (3.8670) grad_norm 0.9556 (1.1192) [2021-04-15 20:02:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][1220/1251] eta 0:00:08 lr 0.000899 time 0.2884 (0.2894) loss 3.4527 (3.8672) grad_norm 0.9821 (1.1187) [2021-04-15 20:02:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][1230/1251] eta 0:00:06 lr 0.000899 time 0.2853 (0.2893) loss 4.2997 (3.8688) grad_norm 0.9694 (1.1183) [2021-04-15 20:02:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][1240/1251] eta 0:00:03 lr 0.000899 time 0.2483 (0.2892) loss 3.3128 (3.8696) grad_norm 0.9539 (1.1180) [2021-04-15 20:02:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [61/300][1250/1251] eta 0:00:00 lr 0.000899 time 0.2482 (0.2889) loss 4.3786 (3.8712) grad_norm 1.1557 (1.1189) [2021-04-15 20:02:28 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 61 training takes 0:06:03 [2021-04-15 20:02:28 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_61.pth saving...... [2021-04-15 20:02:35 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_61.pth saved !!! [2021-04-15 20:02:37 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.276 (1.276) Loss 1.2907 (1.2907) Acc@1 68.945 (68.945) Acc@5 89.551 (89.551) [2021-04-15 20:02:38 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.115 (0.252) Loss 1.1832 (1.2318) Acc@1 73.730 (71.325) Acc@5 90.234 (90.891) [2021-04-15 20:02:40 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.126 (0.245) Loss 1.1564 (1.2343) Acc@1 73.047 (71.275) Acc@5 91.699 (91.085) [2021-04-15 20:02:43 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.158 (0.248) Loss 1.2498 (1.2358) Acc@1 72.754 (71.384) Acc@5 90.332 (91.025) [2021-04-15 20:02:44 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.398 (0.225) Loss 1.2858 (1.2421) Acc@1 70.312 (71.282) Acc@5 91.113 (90.949) [2021-04-15 20:02:47 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 71.366 Acc@5 90.890 [2021-04-15 20:02:47 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 71.4% [2021-04-15 20:02:47 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 71.37% [2021-04-15 20:02:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][0/1251] eta 1:16:31 lr 0.000899 time 3.6706 (3.6706) loss 4.2693 (4.2693) grad_norm 1.3090 (1.3090) [2021-04-15 20:02:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][10/1251] eta 0:12:13 lr 0.000899 time 0.2842 (0.5914) loss 4.7988 (3.9444) grad_norm 1.0793 (1.1933) [2021-04-15 20:02:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][20/1251] eta 0:09:15 lr 0.000899 time 0.2937 (0.4513) loss 3.8615 (3.7796) grad_norm 1.0731 (1.1814) [2021-04-15 20:03:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][30/1251] eta 0:08:09 lr 0.000899 time 0.2996 (0.4007) loss 3.8967 (3.7985) grad_norm 1.0949 (1.1676) [2021-04-15 20:03:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][40/1251] eta 0:07:29 lr 0.000899 time 0.2868 (0.3713) loss 4.0304 (3.7651) grad_norm 1.2774 (1.1622) [2021-04-15 20:03:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][50/1251] eta 0:07:04 lr 0.000899 time 0.3002 (0.3533) loss 3.7848 (3.7707) grad_norm 1.1758 (1.1827) [2021-04-15 20:03:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][60/1251] eta 0:06:47 lr 0.000899 time 0.3224 (0.3421) loss 4.6196 (3.8046) grad_norm 0.9333 (1.1754) [2021-04-15 20:03:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][70/1251] eta 0:06:34 lr 0.000899 time 0.2910 (0.3341) loss 4.4243 (3.8185) grad_norm 0.9611 (1.1581) [2021-04-15 20:03:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][80/1251] eta 0:06:23 lr 0.000899 time 0.2821 (0.3278) loss 3.6520 (3.8154) grad_norm 1.1110 (1.1512) [2021-04-15 20:03:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][90/1251] eta 0:06:17 lr 0.000899 time 0.3162 (0.3251) loss 4.5672 (3.8512) grad_norm 1.0825 (1.1465) [2021-04-15 20:03:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][100/1251] eta 0:06:10 lr 0.000899 time 0.2735 (0.3216) loss 3.5249 (3.8274) grad_norm 1.0636 (1.1376) [2021-04-15 20:03:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][110/1251] eta 0:06:03 lr 0.000899 time 0.2860 (0.3185) loss 4.4363 (3.8172) grad_norm 1.0366 (1.1337) [2021-04-15 20:03:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][120/1251] eta 0:05:57 lr 0.000899 time 0.2821 (0.3163) loss 4.1415 (3.7887) grad_norm 1.1684 (1.1295) [2021-04-15 20:03:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][130/1251] eta 0:05:53 lr 0.000899 time 0.4395 (0.3150) loss 3.8234 (3.8165) grad_norm 1.0082 (1.1280) [2021-04-15 20:03:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][140/1251] eta 0:05:48 lr 0.000899 time 0.2871 (0.3134) loss 3.2815 (3.8124) grad_norm 1.0323 (1.1242) [2021-04-15 20:03:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][150/1251] eta 0:05:43 lr 0.000899 time 0.2966 (0.3122) loss 3.9416 (3.8094) grad_norm 1.0120 (1.1219) [2021-04-15 20:03:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][160/1251] eta 0:05:38 lr 0.000899 time 0.2829 (0.3105) loss 4.5511 (3.8385) grad_norm 1.0464 (1.1215) [2021-04-15 20:03:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][170/1251] eta 0:05:33 lr 0.000899 time 0.2788 (0.3088) loss 4.2050 (3.8530) grad_norm 1.0885 (1.1265) [2021-04-15 20:03:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][180/1251] eta 0:05:28 lr 0.000899 time 0.2581 (0.3072) loss 4.3609 (3.8455) grad_norm 1.2688 (1.1313) [2021-04-15 20:03:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][190/1251] eta 0:05:24 lr 0.000899 time 0.2612 (0.3058) loss 4.0424 (3.8523) grad_norm 1.1318 (1.1319) [2021-04-15 20:03:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][200/1251] eta 0:05:20 lr 0.000899 time 0.3130 (0.3047) loss 4.1515 (3.8447) grad_norm 1.0461 (1.1276) [2021-04-15 20:03:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][210/1251] eta 0:05:16 lr 0.000899 time 0.2859 (0.3037) loss 3.1152 (3.8355) grad_norm 1.0772 (1.1258) [2021-04-15 20:03:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][220/1251] eta 0:05:12 lr 0.000899 time 0.2797 (0.3029) loss 3.6394 (3.8336) grad_norm 1.1891 (1.1241) [2021-04-15 20:03:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][230/1251] eta 0:05:08 lr 0.000899 time 0.2824 (0.3020) loss 3.4788 (3.8299) grad_norm 1.1936 (1.1252) [2021-04-15 20:04:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][240/1251] eta 0:05:04 lr 0.000899 time 0.2952 (0.3013) loss 3.9526 (3.8328) grad_norm 0.9942 (1.1240) [2021-04-15 20:04:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][250/1251] eta 0:05:00 lr 0.000899 time 0.3076 (0.3005) loss 3.7949 (3.8405) grad_norm 1.1412 (1.1224) [2021-04-15 20:04:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][260/1251] eta 0:04:56 lr 0.000899 time 0.2830 (0.2997) loss 4.0679 (3.8290) grad_norm 1.0330 (1.1216) [2021-04-15 20:04:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][270/1251] eta 0:04:53 lr 0.000899 time 0.2712 (0.2990) loss 3.8662 (3.8266) grad_norm 1.1192 (1.1225) [2021-04-15 20:04:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][280/1251] eta 0:04:49 lr 0.000899 time 0.2696 (0.2984) loss 4.6134 (3.8426) grad_norm 0.9790 (1.1235) [2021-04-15 20:04:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][290/1251] eta 0:04:46 lr 0.000899 time 0.2994 (0.2979) loss 3.1043 (3.8409) grad_norm 1.1675 (1.1224) [2021-04-15 20:04:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][300/1251] eta 0:04:42 lr 0.000899 time 0.2708 (0.2974) loss 4.0980 (3.8392) grad_norm 1.1482 (1.1199) [2021-04-15 20:04:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][310/1251] eta 0:04:39 lr 0.000899 time 0.2933 (0.2968) loss 3.2605 (3.8346) grad_norm 1.0402 (1.1184) [2021-04-15 20:04:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][320/1251] eta 0:04:35 lr 0.000898 time 0.2673 (0.2962) loss 2.9766 (3.8443) grad_norm 1.2041 (1.1171) [2021-04-15 20:04:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][330/1251] eta 0:04:32 lr 0.000898 time 0.2685 (0.2957) loss 4.0900 (3.8515) grad_norm 1.0931 (1.1185) [2021-04-15 20:04:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][340/1251] eta 0:04:29 lr 0.000898 time 0.2942 (0.2957) loss 3.3433 (3.8484) grad_norm 1.1002 (1.1164) [2021-04-15 20:04:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][350/1251] eta 0:04:26 lr 0.000898 time 0.2927 (0.2955) loss 3.2751 (3.8442) grad_norm 1.5208 (1.1173) [2021-04-15 20:04:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][360/1251] eta 0:04:23 lr 0.000898 time 0.3039 (0.2956) loss 3.5173 (3.8465) grad_norm 1.2408 (1.1163) [2021-04-15 20:04:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][370/1251] eta 0:04:20 lr 0.000898 time 0.2908 (0.2958) loss 3.6873 (3.8446) grad_norm 1.1732 (1.1175) [2021-04-15 20:04:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][380/1251] eta 0:04:17 lr 0.000898 time 0.2790 (0.2954) loss 3.1015 (3.8493) grad_norm 1.0431 (1.1181) [2021-04-15 20:04:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][390/1251] eta 0:04:14 lr 0.000898 time 0.2838 (0.2952) loss 3.6462 (3.8415) grad_norm 1.0179 (1.1192) [2021-04-15 20:04:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][400/1251] eta 0:04:10 lr 0.000898 time 0.2805 (0.2948) loss 4.3286 (3.8441) grad_norm 1.0142 (1.1187) [2021-04-15 20:04:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][410/1251] eta 0:04:07 lr 0.000898 time 0.2945 (0.2945) loss 4.4656 (3.8494) grad_norm 1.1004 (1.1160) [2021-04-15 20:04:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][420/1251] eta 0:04:04 lr 0.000898 time 0.2869 (0.2941) loss 4.3213 (3.8577) grad_norm 1.0318 (1.1170) [2021-04-15 20:04:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][430/1251] eta 0:04:01 lr 0.000898 time 0.2706 (0.2938) loss 3.7027 (3.8597) grad_norm 1.3459 (1.1176) [2021-04-15 20:04:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][440/1251] eta 0:03:58 lr 0.000898 time 0.2862 (0.2935) loss 4.2041 (3.8603) grad_norm 1.1077 (1.1185) [2021-04-15 20:05:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][450/1251] eta 0:03:54 lr 0.000898 time 0.3227 (0.2932) loss 4.1621 (3.8563) grad_norm 0.9259 (inf) [2021-04-15 20:05:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][460/1251] eta 0:03:51 lr 0.000898 time 0.2988 (0.2930) loss 3.7952 (3.8521) grad_norm 1.1223 (inf) [2021-04-15 20:05:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][470/1251] eta 0:03:48 lr 0.000898 time 0.2809 (0.2927) loss 3.3960 (3.8538) grad_norm 1.2364 (inf) [2021-04-15 20:05:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][480/1251] eta 0:03:45 lr 0.000898 time 0.2962 (0.2925) loss 2.5651 (3.8457) grad_norm 1.1049 (inf) [2021-04-15 20:05:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][490/1251] eta 0:03:42 lr 0.000898 time 0.2538 (0.2923) loss 3.9582 (3.8434) grad_norm 1.3535 (inf) [2021-04-15 20:05:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][500/1251] eta 0:03:39 lr 0.000898 time 0.3209 (0.2920) loss 3.8184 (3.8417) grad_norm 1.1826 (inf) [2021-04-15 20:05:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][510/1251] eta 0:03:36 lr 0.000898 time 0.2963 (0.2918) loss 4.6076 (3.8475) grad_norm 1.1890 (inf) [2021-04-15 20:05:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][520/1251] eta 0:03:33 lr 0.000898 time 0.2748 (0.2916) loss 4.5900 (3.8465) grad_norm 1.2209 (inf) [2021-04-15 20:05:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][530/1251] eta 0:03:30 lr 0.000898 time 0.2534 (0.2913) loss 3.1899 (3.8468) grad_norm 1.0345 (inf) [2021-04-15 20:05:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][540/1251] eta 0:03:27 lr 0.000898 time 0.2962 (0.2911) loss 3.2648 (3.8457) grad_norm 1.2907 (inf) [2021-04-15 20:05:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][550/1251] eta 0:03:23 lr 0.000898 time 0.2559 (0.2909) loss 4.2642 (3.8509) grad_norm 1.3708 (inf) [2021-04-15 20:05:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][560/1251] eta 0:03:20 lr 0.000898 time 0.2866 (0.2908) loss 4.1370 (3.8561) grad_norm 1.1137 (inf) [2021-04-15 20:05:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][570/1251] eta 0:03:17 lr 0.000898 time 0.2868 (0.2906) loss 3.6361 (3.8606) grad_norm 1.0175 (inf) [2021-04-15 20:05:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][580/1251] eta 0:03:15 lr 0.000898 time 0.2892 (0.2911) loss 3.1624 (3.8602) grad_norm 1.2512 (inf) [2021-04-15 20:05:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][590/1251] eta 0:03:12 lr 0.000898 time 0.2745 (0.2910) loss 4.2645 (3.8635) grad_norm 1.4307 (inf) [2021-04-15 20:05:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][600/1251] eta 0:03:09 lr 0.000898 time 0.3861 (0.2911) loss 3.4509 (3.8645) grad_norm 0.9788 (inf) [2021-04-15 20:05:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][610/1251] eta 0:03:06 lr 0.000898 time 0.2868 (0.2910) loss 3.8475 (3.8649) grad_norm 1.3525 (inf) [2021-04-15 20:05:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][620/1251] eta 0:03:03 lr 0.000898 time 0.2809 (0.2908) loss 3.3964 (3.8654) grad_norm 1.1037 (inf) [2021-04-15 20:05:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][630/1251] eta 0:03:00 lr 0.000898 time 0.2941 (0.2906) loss 3.1782 (3.8660) grad_norm 1.2844 (inf) [2021-04-15 20:05:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][640/1251] eta 0:02:57 lr 0.000898 time 0.3058 (0.2906) loss 4.4099 (3.8654) grad_norm 1.1597 (inf) [2021-04-15 20:05:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][650/1251] eta 0:02:54 lr 0.000898 time 0.2940 (0.2904) loss 3.3771 (3.8662) grad_norm 0.9834 (inf) [2021-04-15 20:05:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][660/1251] eta 0:02:51 lr 0.000898 time 0.3026 (0.2903) loss 3.7451 (3.8634) grad_norm 1.1765 (inf) [2021-04-15 20:06:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][670/1251] eta 0:02:48 lr 0.000898 time 0.2555 (0.2904) loss 3.9938 (3.8597) grad_norm 1.1767 (inf) [2021-04-15 20:06:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][680/1251] eta 0:02:45 lr 0.000898 time 0.2668 (0.2904) loss 3.8127 (3.8565) grad_norm 1.2707 (inf) [2021-04-15 20:06:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][690/1251] eta 0:02:42 lr 0.000898 time 0.2908 (0.2904) loss 4.3338 (3.8568) grad_norm 1.1681 (inf) [2021-04-15 20:06:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][700/1251] eta 0:02:39 lr 0.000898 time 0.2780 (0.2902) loss 4.6090 (3.8588) grad_norm 1.0488 (inf) [2021-04-15 20:06:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][710/1251] eta 0:02:37 lr 0.000897 time 0.2677 (0.2902) loss 3.1829 (3.8608) grad_norm 1.0010 (inf) [2021-04-15 20:06:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][720/1251] eta 0:02:34 lr 0.000897 time 0.2629 (0.2905) loss 4.0001 (3.8660) grad_norm 1.0256 (inf) [2021-04-15 20:06:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][730/1251] eta 0:02:31 lr 0.000897 time 0.2895 (0.2903) loss 4.7116 (3.8693) grad_norm 0.9511 (inf) [2021-04-15 20:06:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][740/1251] eta 0:02:28 lr 0.000897 time 0.2779 (0.2901) loss 3.7594 (3.8676) grad_norm 1.1447 (inf) [2021-04-15 20:06:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][750/1251] eta 0:02:25 lr 0.000897 time 0.2833 (0.2900) loss 3.9114 (3.8637) grad_norm 1.0736 (inf) [2021-04-15 20:06:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][760/1251] eta 0:02:22 lr 0.000897 time 0.2725 (0.2898) loss 3.7483 (3.8689) grad_norm 1.4259 (inf) [2021-04-15 20:06:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][770/1251] eta 0:02:19 lr 0.000897 time 0.2698 (0.2898) loss 3.2787 (3.8679) grad_norm 1.3492 (inf) [2021-04-15 20:06:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][780/1251] eta 0:02:16 lr 0.000897 time 0.2674 (0.2897) loss 4.5246 (3.8675) grad_norm 1.1848 (inf) [2021-04-15 20:06:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][790/1251] eta 0:02:13 lr 0.000897 time 0.2931 (0.2896) loss 3.7972 (3.8654) grad_norm 1.2062 (inf) [2021-04-15 20:06:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][800/1251] eta 0:02:10 lr 0.000897 time 0.2657 (0.2895) loss 4.2087 (3.8687) grad_norm 1.1760 (inf) [2021-04-15 20:06:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][810/1251] eta 0:02:07 lr 0.000897 time 0.2953 (0.2894) loss 4.1337 (3.8681) grad_norm 1.1644 (inf) [2021-04-15 20:06:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][820/1251] eta 0:02:04 lr 0.000897 time 0.2808 (0.2893) loss 3.5974 (3.8673) grad_norm 1.0098 (inf) [2021-04-15 20:06:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][830/1251] eta 0:02:01 lr 0.000897 time 0.2653 (0.2893) loss 4.7618 (3.8708) grad_norm 1.3280 (inf) [2021-04-15 20:06:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][840/1251] eta 0:01:58 lr 0.000897 time 0.3106 (0.2892) loss 4.4937 (3.8709) grad_norm 1.3935 (inf) [2021-04-15 20:06:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][850/1251] eta 0:01:55 lr 0.000897 time 0.3057 (0.2891) loss 3.3226 (3.8714) grad_norm 1.1097 (inf) [2021-04-15 20:06:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][860/1251] eta 0:01:52 lr 0.000897 time 0.2803 (0.2889) loss 2.6442 (3.8666) grad_norm 1.0440 (inf) [2021-04-15 20:06:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][870/1251] eta 0:01:50 lr 0.000897 time 0.2574 (0.2888) loss 3.8846 (3.8646) grad_norm 0.9225 (inf) [2021-04-15 20:07:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][880/1251] eta 0:01:47 lr 0.000897 time 0.2807 (0.2887) loss 4.4476 (3.8646) grad_norm 1.0762 (inf) [2021-04-15 20:07:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][890/1251] eta 0:01:44 lr 0.000897 time 0.2880 (0.2886) loss 2.6151 (3.8617) grad_norm 1.0038 (inf) [2021-04-15 20:07:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][900/1251] eta 0:01:41 lr 0.000897 time 0.2605 (0.2885) loss 2.4213 (3.8559) grad_norm 0.8844 (inf) [2021-04-15 20:07:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][910/1251] eta 0:01:38 lr 0.000897 time 0.2992 (0.2884) loss 3.4302 (3.8565) grad_norm 1.2134 (inf) [2021-04-15 20:07:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][920/1251] eta 0:01:35 lr 0.000897 time 0.2829 (0.2885) loss 4.0826 (3.8579) grad_norm 1.1021 (inf) [2021-04-15 20:07:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][930/1251] eta 0:01:32 lr 0.000897 time 0.2914 (0.2885) loss 3.5978 (3.8559) grad_norm 1.0865 (inf) [2021-04-15 20:07:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][940/1251] eta 0:01:29 lr 0.000897 time 0.4419 (0.2887) loss 4.4734 (3.8513) grad_norm 1.4733 (inf) [2021-04-15 20:07:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][950/1251] eta 0:01:26 lr 0.000897 time 0.2854 (0.2886) loss 4.1244 (3.8550) grad_norm 1.1647 (inf) [2021-04-15 20:07:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][960/1251] eta 0:01:23 lr 0.000897 time 0.2785 (0.2885) loss 5.0458 (3.8566) grad_norm 1.1895 (inf) [2021-04-15 20:07:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][970/1251] eta 0:01:21 lr 0.000897 time 0.2873 (0.2884) loss 3.8700 (3.8538) grad_norm 1.4444 (inf) [2021-04-15 20:07:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][980/1251] eta 0:01:18 lr 0.000897 time 0.2658 (0.2884) loss 3.6297 (3.8523) grad_norm 1.1917 (inf) [2021-04-15 20:07:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][990/1251] eta 0:01:15 lr 0.000897 time 0.2793 (0.2883) loss 2.8040 (3.8503) grad_norm 1.1523 (inf) [2021-04-15 20:07:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][1000/1251] eta 0:01:12 lr 0.000897 time 0.2644 (0.2882) loss 4.4581 (3.8504) grad_norm 1.3188 (inf) [2021-04-15 20:07:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][1010/1251] eta 0:01:09 lr 0.000897 time 0.2591 (0.2881) loss 4.0526 (3.8539) grad_norm 1.2308 (inf) [2021-04-15 20:07:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][1020/1251] eta 0:01:06 lr 0.000897 time 0.2876 (0.2880) loss 4.5784 (3.8543) grad_norm 1.4392 (inf) [2021-04-15 20:07:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][1030/1251] eta 0:01:03 lr 0.000897 time 0.2978 (0.2879) loss 3.9942 (3.8547) grad_norm 1.1282 (inf) [2021-04-15 20:07:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][1040/1251] eta 0:01:00 lr 0.000897 time 0.2991 (0.2878) loss 3.6321 (3.8544) grad_norm 1.1376 (inf) [2021-04-15 20:07:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][1050/1251] eta 0:00:57 lr 0.000897 time 0.2605 (0.2877) loss 4.3307 (3.8513) grad_norm 1.0223 (inf) [2021-04-15 20:07:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][1060/1251] eta 0:00:54 lr 0.000897 time 0.2563 (0.2877) loss 4.6510 (3.8520) grad_norm 1.2634 (inf) [2021-04-15 20:07:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][1070/1251] eta 0:00:52 lr 0.000897 time 0.2943 (0.2876) loss 3.8692 (3.8541) grad_norm 1.5132 (inf) [2021-04-15 20:07:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][1080/1251] eta 0:00:49 lr 0.000897 time 0.2976 (0.2876) loss 4.6841 (3.8543) grad_norm 1.0102 (inf) [2021-04-15 20:08:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][1090/1251] eta 0:00:46 lr 0.000897 time 0.3058 (0.2876) loss 4.1896 (3.8534) grad_norm 0.9467 (inf) [2021-04-15 20:08:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][1100/1251] eta 0:00:43 lr 0.000897 time 0.3095 (0.2875) loss 4.3705 (3.8554) grad_norm 0.9582 (inf) [2021-04-15 20:08:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][1110/1251] eta 0:00:40 lr 0.000896 time 0.2813 (0.2874) loss 2.7162 (3.8551) grad_norm 1.2900 (inf) [2021-04-15 20:08:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][1120/1251] eta 0:00:37 lr 0.000896 time 0.2668 (0.2873) loss 3.5710 (3.8531) grad_norm 1.0806 (inf) [2021-04-15 20:08:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][1130/1251] eta 0:00:34 lr 0.000896 time 0.2569 (0.2872) loss 4.1173 (3.8529) grad_norm 1.1024 (inf) [2021-04-15 20:08:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][1140/1251] eta 0:00:31 lr 0.000896 time 0.2850 (0.2871) loss 3.4210 (3.8519) grad_norm 1.0392 (inf) [2021-04-15 20:08:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][1150/1251] eta 0:00:28 lr 0.000896 time 0.2929 (0.2871) loss 3.5925 (3.8536) grad_norm 1.4384 (inf) [2021-04-15 20:08:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][1160/1251] eta 0:00:26 lr 0.000896 time 0.2863 (0.2872) loss 4.3528 (3.8540) grad_norm 0.9499 (inf) [2021-04-15 20:08:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][1170/1251] eta 0:00:23 lr 0.000896 time 0.2721 (0.2872) loss 4.3624 (3.8557) grad_norm 1.1989 (inf) [2021-04-15 20:08:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][1180/1251] eta 0:00:20 lr 0.000896 time 0.2698 (0.2872) loss 4.2735 (3.8581) grad_norm 0.9473 (inf) [2021-04-15 20:08:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][1190/1251] eta 0:00:17 lr 0.000896 time 0.2760 (0.2872) loss 2.7614 (3.8559) grad_norm 1.4301 (inf) [2021-04-15 20:08:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][1200/1251] eta 0:00:14 lr 0.000896 time 0.2632 (0.2872) loss 3.7687 (3.8592) grad_norm 1.0995 (inf) [2021-04-15 20:08:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][1210/1251] eta 0:00:11 lr 0.000896 time 0.2822 (0.2872) loss 3.9369 (3.8580) grad_norm 1.2463 (inf) [2021-04-15 20:08:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][1220/1251] eta 0:00:08 lr 0.000896 time 0.2575 (0.2871) loss 4.3832 (3.8598) grad_norm 0.9662 (inf) [2021-04-15 20:08:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][1230/1251] eta 0:00:06 lr 0.000896 time 0.2883 (0.2870) loss 3.8178 (3.8596) grad_norm 1.2374 (inf) [2021-04-15 20:08:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][1240/1251] eta 0:00:03 lr 0.000896 time 0.2481 (0.2870) loss 3.2478 (3.8593) grad_norm 1.0721 (inf) [2021-04-15 20:08:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [62/300][1250/1251] eta 0:00:00 lr 0.000896 time 0.2489 (0.2867) loss 3.5534 (3.8618) grad_norm 1.0493 (inf) [2021-04-15 20:08:49 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 62 training takes 0:06:01 [2021-04-15 20:08:49 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_62.pth saving...... [2021-04-15 20:08:57 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_62.pth saved !!! [2021-04-15 20:08:58 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.137 (1.137) Loss 1.3167 (1.3167) Acc@1 70.703 (70.703) Acc@5 90.430 (90.430) [2021-04-15 20:09:00 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.768 (0.257) Loss 1.2543 (1.2447) Acc@1 69.824 (71.289) Acc@5 90.625 (90.918) [2021-04-15 20:09:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.120 (0.253) Loss 1.2946 (1.2551) Acc@1 70.508 (71.159) Acc@5 90.723 (90.760) [2021-04-15 20:09:04 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.155 (0.229) Loss 1.2177 (1.2519) Acc@1 71.582 (71.229) Acc@5 92.285 (90.764) [2021-04-15 20:09:06 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.219) Loss 1.3490 (1.2493) Acc@1 69.043 (71.234) Acc@5 89.746 (90.820) [2021-04-15 20:09:09 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 71.194 Acc@5 90.770 [2021-04-15 20:09:09 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 71.2% [2021-04-15 20:09:09 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 71.37% [2021-04-15 20:09:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][0/1251] eta 1:29:23 lr 0.000896 time 4.2877 (4.2877) loss 4.1330 (4.1330) grad_norm 1.0860 (1.0860) [2021-04-15 20:09:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][10/1251] eta 0:13:11 lr 0.000896 time 0.2660 (0.6379) loss 3.6223 (4.1449) grad_norm 1.0788 (1.0285) [2021-04-15 20:09:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][20/1251] eta 0:09:35 lr 0.000896 time 0.2861 (0.4673) loss 3.6606 (3.9951) grad_norm 1.1771 (1.0835) [2021-04-15 20:09:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][30/1251] eta 0:08:17 lr 0.000896 time 0.2840 (0.4077) loss 3.5122 (3.9147) grad_norm 1.0895 (1.0670) [2021-04-15 20:09:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][40/1251] eta 0:07:35 lr 0.000896 time 0.3061 (0.3764) loss 4.6311 (3.9464) grad_norm 1.3529 (1.0799) [2021-04-15 20:09:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][50/1251] eta 0:07:08 lr 0.000896 time 0.2598 (0.3567) loss 4.2510 (3.9440) grad_norm 0.9734 (1.0893) [2021-04-15 20:09:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][60/1251] eta 0:06:49 lr 0.000896 time 0.2638 (0.3441) loss 3.8718 (3.9098) grad_norm 1.1628 (1.0930) [2021-04-15 20:09:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][70/1251] eta 0:06:35 lr 0.000896 time 0.2812 (0.3351) loss 4.2956 (3.9170) grad_norm 1.0547 (1.0951) [2021-04-15 20:09:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][80/1251] eta 0:06:24 lr 0.000896 time 0.2906 (0.3285) loss 3.2512 (3.9251) grad_norm 1.4089 (1.1105) [2021-04-15 20:09:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][90/1251] eta 0:06:15 lr 0.000896 time 0.2854 (0.3235) loss 3.8750 (3.9437) grad_norm 0.9988 (1.1063) [2021-04-15 20:09:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][100/1251] eta 0:06:06 lr 0.000896 time 0.2595 (0.3188) loss 3.1352 (3.9257) grad_norm 1.1643 (1.1112) [2021-04-15 20:09:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][110/1251] eta 0:05:59 lr 0.000896 time 0.2795 (0.3153) loss 3.4135 (3.9342) grad_norm 1.0677 (1.1138) [2021-04-15 20:09:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][120/1251] eta 0:05:54 lr 0.000896 time 0.2787 (0.3138) loss 2.5406 (3.9231) grad_norm 1.0096 (1.1135) [2021-04-15 20:09:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][130/1251] eta 0:05:48 lr 0.000896 time 0.2945 (0.3110) loss 2.7458 (3.9175) grad_norm 1.0540 (1.1137) [2021-04-15 20:09:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][140/1251] eta 0:05:45 lr 0.000896 time 0.2752 (0.3108) loss 4.0336 (3.9183) grad_norm 1.0444 (1.1111) [2021-04-15 20:09:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][150/1251] eta 0:05:40 lr 0.000896 time 0.4117 (0.3096) loss 3.3724 (3.8967) grad_norm 1.1607 (1.1119) [2021-04-15 20:09:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][160/1251] eta 0:05:35 lr 0.000896 time 0.2716 (0.3074) loss 3.7429 (3.9125) grad_norm 1.2217 (1.1151) [2021-04-15 20:10:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][170/1251] eta 0:05:30 lr 0.000896 time 0.2771 (0.3058) loss 3.5537 (3.9238) grad_norm 1.0096 (1.1177) [2021-04-15 20:10:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][180/1251] eta 0:05:26 lr 0.000896 time 0.2871 (0.3046) loss 4.0113 (3.9099) grad_norm 1.0892 (1.1163) [2021-04-15 20:10:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][190/1251] eta 0:05:21 lr 0.000896 time 0.2762 (0.3032) loss 3.4079 (3.9099) grad_norm 1.0732 (1.1193) [2021-04-15 20:10:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][200/1251] eta 0:05:17 lr 0.000896 time 0.2896 (0.3023) loss 4.2724 (3.9068) grad_norm 1.1926 (1.1140) [2021-04-15 20:10:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][210/1251] eta 0:05:13 lr 0.000896 time 0.2953 (0.3013) loss 3.9436 (3.8894) grad_norm 1.1557 (1.1196) [2021-04-15 20:10:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][220/1251] eta 0:05:09 lr 0.000896 time 0.2722 (0.3002) loss 2.6059 (3.8808) grad_norm 0.9278 (1.1243) [2021-04-15 20:10:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][230/1251] eta 0:05:05 lr 0.000896 time 0.2645 (0.2995) loss 3.5469 (3.8735) grad_norm 1.0325 (1.1295) [2021-04-15 20:10:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][240/1251] eta 0:05:02 lr 0.000896 time 0.2892 (0.2988) loss 3.9344 (3.8683) grad_norm 1.0969 (1.1289) [2021-04-15 20:10:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][250/1251] eta 0:04:58 lr 0.000895 time 0.2508 (0.2979) loss 3.0851 (3.8773) grad_norm 1.0924 (1.1286) [2021-04-15 20:10:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][260/1251] eta 0:04:54 lr 0.000895 time 0.2847 (0.2974) loss 3.2198 (3.8733) grad_norm 1.2113 (1.1307) [2021-04-15 20:10:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][270/1251] eta 0:04:51 lr 0.000895 time 0.3007 (0.2968) loss 4.5324 (3.8751) grad_norm 1.1264 (1.1310) [2021-04-15 20:10:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][280/1251] eta 0:04:47 lr 0.000895 time 0.2679 (0.2963) loss 4.1496 (3.8775) grad_norm 1.1246 (1.1306) [2021-04-15 20:10:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][290/1251] eta 0:04:44 lr 0.000895 time 0.3048 (0.2958) loss 4.3331 (3.8728) grad_norm 1.1406 (1.1294) [2021-04-15 20:10:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][300/1251] eta 0:04:40 lr 0.000895 time 0.2682 (0.2952) loss 4.1065 (3.8786) grad_norm 1.1643 (1.1269) [2021-04-15 20:10:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][310/1251] eta 0:04:37 lr 0.000895 time 0.2949 (0.2948) loss 2.5803 (3.8782) grad_norm 0.9440 (1.1253) [2021-04-15 20:10:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][320/1251] eta 0:04:33 lr 0.000895 time 0.2836 (0.2943) loss 3.5512 (3.8801) grad_norm 1.0845 (1.1220) [2021-04-15 20:10:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][330/1251] eta 0:04:30 lr 0.000895 time 0.2803 (0.2938) loss 4.1285 (3.8710) grad_norm 1.0670 (1.1247) [2021-04-15 20:10:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][340/1251] eta 0:04:27 lr 0.000895 time 0.2825 (0.2934) loss 3.6436 (3.8767) grad_norm 1.1281 (1.1259) [2021-04-15 20:10:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][350/1251] eta 0:04:24 lr 0.000895 time 0.2899 (0.2931) loss 3.2010 (3.8760) grad_norm 1.1289 (1.1248) [2021-04-15 20:10:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][360/1251] eta 0:04:21 lr 0.000895 time 0.2671 (0.2931) loss 4.2678 (3.8776) grad_norm 0.9563 (1.1228) [2021-04-15 20:10:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][370/1251] eta 0:04:18 lr 0.000895 time 0.2881 (0.2933) loss 4.3975 (3.8730) grad_norm 1.0751 (1.1214) [2021-04-15 20:11:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][380/1251] eta 0:04:15 lr 0.000895 time 0.2613 (0.2933) loss 3.8811 (3.8713) grad_norm 1.1508 (1.1205) [2021-04-15 20:11:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][390/1251] eta 0:04:12 lr 0.000895 time 0.2783 (0.2929) loss 2.7640 (3.8784) grad_norm 1.2512 (1.1211) [2021-04-15 20:11:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][400/1251] eta 0:04:08 lr 0.000895 time 0.2789 (0.2926) loss 4.2023 (3.8777) grad_norm 1.3059 (1.1233) [2021-04-15 20:11:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][410/1251] eta 0:04:05 lr 0.000895 time 0.2971 (0.2921) loss 3.1178 (3.8759) grad_norm 1.0316 (1.1230) [2021-04-15 20:11:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][420/1251] eta 0:04:02 lr 0.000895 time 0.3001 (0.2918) loss 3.2042 (3.8781) grad_norm 1.1585 (1.1242) [2021-04-15 20:11:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][430/1251] eta 0:03:59 lr 0.000895 time 0.2911 (0.2914) loss 4.0290 (3.8718) grad_norm 1.1082 (1.1251) [2021-04-15 20:11:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][440/1251] eta 0:03:56 lr 0.000895 time 0.3030 (0.2915) loss 3.4871 (3.8707) grad_norm 1.1176 (1.1259) [2021-04-15 20:11:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][450/1251] eta 0:03:53 lr 0.000895 time 0.2828 (0.2912) loss 3.5526 (3.8681) grad_norm 0.9870 (1.1255) [2021-04-15 20:11:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][460/1251] eta 0:03:50 lr 0.000895 time 0.2587 (0.2909) loss 4.3565 (3.8649) grad_norm 0.9893 (1.1238) [2021-04-15 20:11:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][470/1251] eta 0:03:47 lr 0.000895 time 0.2655 (0.2907) loss 3.9148 (3.8697) grad_norm 1.1162 (1.1229) [2021-04-15 20:11:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][480/1251] eta 0:03:44 lr 0.000895 time 0.2948 (0.2909) loss 4.3522 (3.8724) grad_norm 1.1640 (1.1236) [2021-04-15 20:11:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][490/1251] eta 0:03:41 lr 0.000895 time 0.2941 (0.2911) loss 4.2926 (3.8775) grad_norm 1.0748 (1.1248) [2021-04-15 20:11:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][500/1251] eta 0:03:38 lr 0.000895 time 0.2733 (0.2908) loss 4.2819 (3.8848) grad_norm 1.3858 (1.1249) [2021-04-15 20:11:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][510/1251] eta 0:03:35 lr 0.000895 time 0.3036 (0.2906) loss 3.8732 (3.8853) grad_norm 1.0964 (1.1249) [2021-04-15 20:11:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][520/1251] eta 0:03:32 lr 0.000895 time 0.2756 (0.2904) loss 3.0784 (3.8786) grad_norm 1.1520 (1.1257) [2021-04-15 20:11:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][530/1251] eta 0:03:29 lr 0.000895 time 0.2478 (0.2901) loss 4.5405 (3.8772) grad_norm 1.0845 (1.1262) [2021-04-15 20:11:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][540/1251] eta 0:03:26 lr 0.000895 time 0.2804 (0.2899) loss 4.0293 (3.8850) grad_norm 1.0053 (1.1266) [2021-04-15 20:11:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][550/1251] eta 0:03:23 lr 0.000895 time 0.2913 (0.2898) loss 2.8851 (3.8860) grad_norm 1.1429 (1.1269) [2021-04-15 20:11:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][560/1251] eta 0:03:20 lr 0.000895 time 0.2950 (0.2895) loss 4.6587 (3.8844) grad_norm 1.1927 (1.1269) [2021-04-15 20:11:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][570/1251] eta 0:03:17 lr 0.000895 time 0.2780 (0.2893) loss 4.2288 (3.8881) grad_norm 1.2972 (1.1271) [2021-04-15 20:11:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][580/1251] eta 0:03:14 lr 0.000895 time 0.2863 (0.2898) loss 3.4598 (3.8851) grad_norm 0.9911 (1.1278) [2021-04-15 20:12:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][590/1251] eta 0:03:11 lr 0.000895 time 0.3005 (0.2896) loss 3.5188 (3.8872) grad_norm 0.9756 (1.1281) [2021-04-15 20:12:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][600/1251] eta 0:03:08 lr 0.000895 time 0.2937 (0.2895) loss 4.1446 (3.8900) grad_norm 1.0714 (1.1285) [2021-04-15 20:12:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][610/1251] eta 0:03:05 lr 0.000895 time 0.2858 (0.2894) loss 3.7680 (3.8895) grad_norm 1.0195 (1.1275) [2021-04-15 20:12:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][620/1251] eta 0:03:02 lr 0.000895 time 0.2650 (0.2892) loss 4.1914 (3.8905) grad_norm 0.9457 (1.1278) [2021-04-15 20:12:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][630/1251] eta 0:02:59 lr 0.000895 time 0.2910 (0.2891) loss 3.4145 (3.8904) grad_norm 1.1710 (1.1284) [2021-04-15 20:12:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][640/1251] eta 0:02:56 lr 0.000894 time 0.2715 (0.2890) loss 4.3488 (3.8917) grad_norm 1.0173 (1.1279) [2021-04-15 20:12:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][650/1251] eta 0:02:53 lr 0.000894 time 0.2762 (0.2888) loss 3.3740 (3.8933) grad_norm 1.0777 (1.1273) [2021-04-15 20:12:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][660/1251] eta 0:02:50 lr 0.000894 time 0.2551 (0.2887) loss 4.3866 (3.8899) grad_norm 1.0172 (1.1262) [2021-04-15 20:12:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][670/1251] eta 0:02:47 lr 0.000894 time 0.2801 (0.2886) loss 3.4510 (3.8897) grad_norm 1.0451 (1.1252) [2021-04-15 20:12:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][680/1251] eta 0:02:44 lr 0.000894 time 0.2912 (0.2886) loss 3.7961 (3.8883) grad_norm 1.0105 (1.1255) [2021-04-15 20:12:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][690/1251] eta 0:02:41 lr 0.000894 time 0.2647 (0.2886) loss 4.4143 (3.8892) grad_norm 1.0832 (1.1250) [2021-04-15 20:12:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][700/1251] eta 0:02:38 lr 0.000894 time 0.2876 (0.2885) loss 3.8291 (3.8907) grad_norm 1.1038 (1.1251) [2021-04-15 20:12:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][710/1251] eta 0:02:36 lr 0.000894 time 0.2677 (0.2884) loss 4.0023 (3.8940) grad_norm 1.3702 (1.1247) [2021-04-15 20:12:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][720/1251] eta 0:02:33 lr 0.000894 time 0.2817 (0.2885) loss 2.6144 (3.8865) grad_norm 1.2588 (1.1257) [2021-04-15 20:12:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][730/1251] eta 0:02:30 lr 0.000894 time 0.2798 (0.2886) loss 4.3678 (3.8842) grad_norm 1.0616 (1.1262) [2021-04-15 20:12:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][740/1251] eta 0:02:27 lr 0.000894 time 0.2950 (0.2885) loss 4.0328 (3.8839) grad_norm 1.2547 (1.1261) [2021-04-15 20:12:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][750/1251] eta 0:02:24 lr 0.000894 time 0.2663 (0.2883) loss 4.4542 (3.8851) grad_norm 1.1054 (1.1276) [2021-04-15 20:12:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][760/1251] eta 0:02:21 lr 0.000894 time 0.2952 (0.2883) loss 4.3567 (3.8864) grad_norm 1.0440 (1.1306) [2021-04-15 20:12:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][770/1251] eta 0:02:18 lr 0.000894 time 0.2772 (0.2882) loss 3.1686 (3.8864) grad_norm 1.0903 (1.1305) [2021-04-15 20:12:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][780/1251] eta 0:02:15 lr 0.000894 time 0.2854 (0.2880) loss 4.6285 (3.8878) grad_norm 1.0778 (1.1311) [2021-04-15 20:12:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][790/1251] eta 0:02:12 lr 0.000894 time 0.2715 (0.2879) loss 4.4007 (3.8909) grad_norm 1.1989 (1.1312) [2021-04-15 20:12:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][800/1251] eta 0:02:09 lr 0.000894 time 0.2733 (0.2878) loss 4.0574 (3.8915) grad_norm 1.1220 (1.1318) [2021-04-15 20:13:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][810/1251] eta 0:02:06 lr 0.000894 time 0.2889 (0.2876) loss 2.9922 (3.8860) grad_norm 1.0350 (1.1322) [2021-04-15 20:13:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][820/1251] eta 0:02:03 lr 0.000894 time 0.2760 (0.2875) loss 4.2531 (3.8864) grad_norm 1.1800 (1.1328) [2021-04-15 20:13:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][830/1251] eta 0:02:01 lr 0.000894 time 0.2483 (0.2874) loss 4.2409 (3.8876) grad_norm 1.1902 (1.1341) [2021-04-15 20:13:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][840/1251] eta 0:01:58 lr 0.000894 time 0.2607 (0.2874) loss 3.6711 (3.8881) grad_norm 1.0623 (1.1340) [2021-04-15 20:13:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][850/1251] eta 0:01:55 lr 0.000894 time 0.2721 (0.2873) loss 3.7731 (3.8855) grad_norm 1.3106 (1.1338) [2021-04-15 20:13:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][860/1251] eta 0:01:52 lr 0.000894 time 0.2583 (0.2871) loss 4.0598 (3.8869) grad_norm 0.9960 (1.1328) [2021-04-15 20:13:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][870/1251] eta 0:01:49 lr 0.000894 time 0.2898 (0.2871) loss 3.8878 (3.8868) grad_norm 1.1149 (1.1330) [2021-04-15 20:13:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][880/1251] eta 0:01:46 lr 0.000894 time 0.2562 (0.2870) loss 2.8540 (3.8873) grad_norm 1.2866 (1.1328) [2021-04-15 20:13:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][890/1251] eta 0:01:43 lr 0.000894 time 0.2930 (0.2870) loss 4.4237 (3.8849) grad_norm 1.1218 (1.1320) [2021-04-15 20:13:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][900/1251] eta 0:01:40 lr 0.000894 time 0.2810 (0.2869) loss 4.3419 (3.8848) grad_norm 1.2409 (1.1323) [2021-04-15 20:13:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][910/1251] eta 0:01:37 lr 0.000894 time 0.2694 (0.2868) loss 2.8719 (3.8848) grad_norm 1.2731 (1.1331) [2021-04-15 20:13:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][920/1251] eta 0:01:34 lr 0.000894 time 0.3104 (0.2867) loss 4.4675 (3.8835) grad_norm 1.1831 (1.1335) [2021-04-15 20:13:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][930/1251] eta 0:01:32 lr 0.000894 time 0.3032 (0.2867) loss 3.8918 (3.8856) grad_norm 1.1297 (1.1328) [2021-04-15 20:13:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][940/1251] eta 0:01:29 lr 0.000894 time 0.3618 (0.2869) loss 3.8615 (3.8878) grad_norm 1.0454 (1.1321) [2021-04-15 20:13:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][950/1251] eta 0:01:26 lr 0.000894 time 0.2971 (0.2868) loss 4.0367 (3.8867) grad_norm 1.0331 (1.1326) [2021-04-15 20:13:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][960/1251] eta 0:01:23 lr 0.000894 time 0.2699 (0.2868) loss 3.2596 (3.8863) grad_norm 1.1731 (1.1330) [2021-04-15 20:13:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][970/1251] eta 0:01:20 lr 0.000894 time 0.2653 (0.2868) loss 4.0671 (3.8851) grad_norm 1.3473 (1.1333) [2021-04-15 20:13:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][980/1251] eta 0:01:17 lr 0.000894 time 0.2890 (0.2867) loss 3.8195 (3.8847) grad_norm 1.1152 (1.1338) [2021-04-15 20:13:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][990/1251] eta 0:01:14 lr 0.000894 time 0.2641 (0.2866) loss 3.9144 (3.8874) grad_norm 1.1652 (1.1330) [2021-04-15 20:13:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][1000/1251] eta 0:01:11 lr 0.000894 time 0.2691 (0.2866) loss 3.3774 (3.8863) grad_norm 1.1821 (1.1333) [2021-04-15 20:13:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][1010/1251] eta 0:01:09 lr 0.000894 time 0.3066 (0.2865) loss 4.2305 (3.8870) grad_norm 1.1174 (1.1331) [2021-04-15 20:14:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][1020/1251] eta 0:01:06 lr 0.000894 time 0.2702 (0.2866) loss 4.1325 (3.8846) grad_norm 1.0263 (1.1331) [2021-04-15 20:14:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][1030/1251] eta 0:01:03 lr 0.000893 time 0.2886 (0.2865) loss 4.9111 (3.8833) grad_norm 1.2372 (1.1332) [2021-04-15 20:14:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][1040/1251] eta 0:01:00 lr 0.000893 time 0.2665 (0.2864) loss 3.8598 (3.8818) grad_norm 0.9479 (1.1332) [2021-04-15 20:14:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][1050/1251] eta 0:00:57 lr 0.000893 time 0.4225 (0.2865) loss 3.4793 (3.8792) grad_norm 1.4953 (1.1339) [2021-04-15 20:14:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][1060/1251] eta 0:00:54 lr 0.000893 time 0.2615 (0.2864) loss 3.4340 (3.8796) grad_norm 0.9565 (1.1335) [2021-04-15 20:14:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][1070/1251] eta 0:00:51 lr 0.000893 time 0.2752 (0.2864) loss 4.4460 (3.8805) grad_norm 0.9940 (1.1327) [2021-04-15 20:14:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][1080/1251] eta 0:00:48 lr 0.000893 time 0.2735 (0.2864) loss 3.8857 (3.8798) grad_norm 1.0676 (1.1327) [2021-04-15 20:14:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][1090/1251] eta 0:00:46 lr 0.000893 time 0.2594 (0.2863) loss 3.1815 (3.8796) grad_norm 1.0716 (1.1322) [2021-04-15 20:14:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][1100/1251] eta 0:00:43 lr 0.000893 time 0.2780 (0.2863) loss 4.2959 (3.8815) grad_norm 1.2757 (1.1319) [2021-04-15 20:14:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][1110/1251] eta 0:00:40 lr 0.000893 time 0.2769 (0.2863) loss 3.4585 (3.8807) grad_norm 1.0412 (1.1320) [2021-04-15 20:14:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][1120/1251] eta 0:00:37 lr 0.000893 time 0.2997 (0.2863) loss 3.3770 (3.8770) grad_norm 1.2664 (1.1322) [2021-04-15 20:14:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][1130/1251] eta 0:00:34 lr 0.000893 time 0.2654 (0.2862) loss 4.0728 (3.8772) grad_norm 1.1126 (1.1326) [2021-04-15 20:14:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][1140/1251] eta 0:00:31 lr 0.000893 time 0.2647 (0.2861) loss 3.9092 (3.8754) grad_norm 0.9544 (1.1324) [2021-04-15 20:14:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][1150/1251] eta 0:00:28 lr 0.000893 time 0.2751 (0.2861) loss 3.9632 (3.8748) grad_norm 1.0418 (1.1316) [2021-04-15 20:14:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][1160/1251] eta 0:00:26 lr 0.000893 time 0.2888 (0.2862) loss 4.2160 (3.8743) grad_norm 1.2763 (1.1315) [2021-04-15 20:14:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][1170/1251] eta 0:00:23 lr 0.000893 time 0.2572 (0.2861) loss 4.4529 (3.8757) grad_norm 1.0046 (1.1322) [2021-04-15 20:14:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][1180/1251] eta 0:00:20 lr 0.000893 time 0.2697 (0.2861) loss 4.4389 (3.8771) grad_norm 1.0771 (1.1320) [2021-04-15 20:14:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][1190/1251] eta 0:00:17 lr 0.000893 time 0.2732 (0.2860) loss 4.0717 (3.8790) grad_norm 1.0948 (1.1315) [2021-04-15 20:14:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][1200/1251] eta 0:00:14 lr 0.000893 time 0.2699 (0.2860) loss 3.3815 (3.8784) grad_norm 1.0357 (1.1307) [2021-04-15 20:14:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][1210/1251] eta 0:00:11 lr 0.000893 time 0.2686 (0.2859) loss 3.2446 (3.8775) grad_norm 1.1449 (1.1301) [2021-04-15 20:14:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][1220/1251] eta 0:00:08 lr 0.000893 time 0.3044 (0.2860) loss 4.4146 (3.8781) grad_norm 0.9543 (1.1301) [2021-04-15 20:15:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][1230/1251] eta 0:00:06 lr 0.000893 time 0.2639 (0.2859) loss 4.3278 (3.8794) grad_norm 1.2746 (1.1304) [2021-04-15 20:15:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][1240/1251] eta 0:00:03 lr 0.000893 time 0.2487 (0.2858) loss 4.1258 (3.8763) grad_norm 1.1981 (1.1305) [2021-04-15 20:15:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [63/300][1250/1251] eta 0:00:00 lr 0.000893 time 0.2486 (0.2855) loss 3.7558 (3.8770) grad_norm 1.1693 (1.1313) [2021-04-15 20:15:08 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 63 training takes 0:05:59 [2021-04-15 20:15:08 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_63.pth saving...... [2021-04-15 20:15:25 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_63.pth saved !!! [2021-04-15 20:15:26 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.202 (1.202) Loss 1.2390 (1.2390) Acc@1 70.605 (70.605) Acc@5 90.234 (90.234) [2021-04-15 20:15:28 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.647 (0.254) Loss 1.2830 (1.2321) Acc@1 69.727 (71.129) Acc@5 90.625 (90.314) [2021-04-15 20:15:30 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.114 (0.237) Loss 1.1082 (1.2184) Acc@1 73.633 (71.489) Acc@5 91.992 (90.509) [2021-04-15 20:15:33 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.374 (0.247) Loss 1.2964 (1.2234) Acc@1 69.336 (71.349) Acc@5 89.941 (90.653) [2021-04-15 20:15:34 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.221) Loss 1.2461 (1.2254) Acc@1 70.215 (71.334) Acc@5 90.527 (90.642) [2021-04-15 20:15:37 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 71.348 Acc@5 90.614 [2021-04-15 20:15:37 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 71.3% [2021-04-15 20:15:37 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 71.37% [2021-04-15 20:15:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][0/1251] eta 1:14:06 lr 0.000893 time 3.5542 (3.5542) loss 3.8875 (3.8875) grad_norm 1.2565 (1.2565) [2021-04-15 20:15:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][10/1251] eta 0:12:01 lr 0.000893 time 0.2549 (0.5810) loss 4.4954 (4.0824) grad_norm 1.0386 (1.1497) [2021-04-15 20:15:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][20/1251] eta 0:08:59 lr 0.000893 time 0.2799 (0.4379) loss 2.9366 (3.9265) grad_norm 0.9673 (1.1267) [2021-04-15 20:15:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][30/1251] eta 0:07:53 lr 0.000893 time 0.2557 (0.3876) loss 4.2824 (3.8971) grad_norm 1.2504 (1.1210) [2021-04-15 20:15:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][40/1251] eta 0:07:18 lr 0.000893 time 0.2734 (0.3619) loss 4.2686 (3.8723) grad_norm 1.3037 (1.1276) [2021-04-15 20:15:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][50/1251] eta 0:06:57 lr 0.000893 time 0.2926 (0.3474) loss 2.6190 (3.8952) grad_norm 1.0435 (1.1202) [2021-04-15 20:15:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][60/1251] eta 0:06:39 lr 0.000893 time 0.2987 (0.3357) loss 3.9832 (3.8456) grad_norm 1.3499 (1.1190) [2021-04-15 20:16:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][70/1251] eta 0:06:27 lr 0.000893 time 0.2598 (0.3280) loss 4.5791 (3.8847) grad_norm 1.1361 (1.1214) [2021-04-15 20:16:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][80/1251] eta 0:06:17 lr 0.000893 time 0.2770 (0.3222) loss 3.3036 (3.8832) grad_norm 0.9986 (1.1138) [2021-04-15 20:16:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][90/1251] eta 0:06:08 lr 0.000893 time 0.2683 (0.3177) loss 3.9555 (3.8678) grad_norm 1.5289 (1.1203) [2021-04-15 20:16:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][100/1251] eta 0:06:01 lr 0.000893 time 0.2901 (0.3143) loss 3.3470 (3.8429) grad_norm 1.0179 (1.1226) [2021-04-15 20:16:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][110/1251] eta 0:05:54 lr 0.000893 time 0.2827 (0.3111) loss 4.3729 (3.8434) grad_norm 1.1018 (1.1214) [2021-04-15 20:16:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][120/1251] eta 0:05:48 lr 0.000893 time 0.2725 (0.3085) loss 4.0001 (3.8437) grad_norm 1.0602 (1.1279) [2021-04-15 20:16:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][130/1251] eta 0:05:43 lr 0.000893 time 0.2905 (0.3063) loss 3.9615 (3.8308) grad_norm 1.1090 (1.1304) [2021-04-15 20:16:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][140/1251] eta 0:05:39 lr 0.000893 time 0.2742 (0.3053) loss 4.0959 (3.8294) grad_norm 1.0961 (1.1298) [2021-04-15 20:16:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][150/1251] eta 0:05:35 lr 0.000893 time 0.4656 (0.3045) loss 4.0956 (3.8228) grad_norm 0.9533 (1.1316) [2021-04-15 20:16:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][160/1251] eta 0:05:31 lr 0.000893 time 0.2723 (0.3035) loss 3.7548 (3.8285) grad_norm 1.0459 (1.1302) [2021-04-15 20:16:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][170/1251] eta 0:05:27 lr 0.000892 time 0.2820 (0.3028) loss 3.8655 (3.8106) grad_norm 1.1675 (1.1288) [2021-04-15 20:16:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][180/1251] eta 0:05:22 lr 0.000892 time 0.2744 (0.3014) loss 2.8625 (3.8107) grad_norm 1.2238 (1.1302) [2021-04-15 20:16:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][190/1251] eta 0:05:18 lr 0.000892 time 0.2646 (0.3003) loss 3.2285 (3.8107) grad_norm 1.4579 (1.1363) [2021-04-15 20:16:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][200/1251] eta 0:05:14 lr 0.000892 time 0.2905 (0.2994) loss 4.1065 (3.8043) grad_norm 1.2111 (1.1366) [2021-04-15 20:16:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][210/1251] eta 0:05:10 lr 0.000892 time 0.2736 (0.2984) loss 2.7386 (3.7903) grad_norm 1.1829 (1.1388) [2021-04-15 20:16:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][220/1251] eta 0:05:06 lr 0.000892 time 0.2879 (0.2976) loss 4.4278 (3.7946) grad_norm 1.0990 (1.1372) [2021-04-15 20:16:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][230/1251] eta 0:05:03 lr 0.000892 time 0.2813 (0.2969) loss 3.9397 (3.8007) grad_norm 1.1428 (inf) [2021-04-15 20:16:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][240/1251] eta 0:04:59 lr 0.000892 time 0.3061 (0.2963) loss 4.2215 (3.8118) grad_norm 1.2188 (inf) [2021-04-15 20:16:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][250/1251] eta 0:04:55 lr 0.000892 time 0.2476 (0.2955) loss 4.2255 (3.8086) grad_norm 1.3658 (inf) [2021-04-15 20:16:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][260/1251] eta 0:04:52 lr 0.000892 time 0.4301 (0.2954) loss 3.8354 (3.8092) grad_norm 1.0765 (inf) [2021-04-15 20:16:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][270/1251] eta 0:04:49 lr 0.000892 time 0.2673 (0.2947) loss 4.5982 (3.8115) grad_norm 1.2410 (inf) [2021-04-15 20:17:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][280/1251] eta 0:04:45 lr 0.000892 time 0.2897 (0.2941) loss 3.7728 (3.8144) grad_norm 1.0817 (inf) [2021-04-15 20:17:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][290/1251] eta 0:04:42 lr 0.000892 time 0.3454 (0.2942) loss 4.4066 (3.8196) grad_norm 0.9792 (inf) [2021-04-15 20:17:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][300/1251] eta 0:04:39 lr 0.000892 time 0.2715 (0.2936) loss 3.9045 (3.8165) grad_norm 1.2836 (inf) [2021-04-15 20:17:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][310/1251] eta 0:04:36 lr 0.000892 time 0.3199 (0.2934) loss 3.5434 (3.8198) grad_norm 1.2954 (inf) [2021-04-15 20:17:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][320/1251] eta 0:04:32 lr 0.000892 time 0.2700 (0.2929) loss 2.6667 (3.8157) grad_norm 1.0572 (inf) [2021-04-15 20:17:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][330/1251] eta 0:04:29 lr 0.000892 time 0.2929 (0.2926) loss 4.0411 (3.8225) grad_norm 1.2175 (inf) [2021-04-15 20:17:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][340/1251] eta 0:04:26 lr 0.000892 time 0.2601 (0.2923) loss 3.4691 (3.8211) grad_norm 0.9656 (inf) [2021-04-15 20:17:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][350/1251] eta 0:04:23 lr 0.000892 time 0.2712 (0.2919) loss 3.7619 (3.8200) grad_norm 1.2312 (inf) [2021-04-15 20:17:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][360/1251] eta 0:04:20 lr 0.000892 time 0.2671 (0.2924) loss 4.5690 (3.8250) grad_norm 1.1977 (inf) [2021-04-15 20:17:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][370/1251] eta 0:04:17 lr 0.000892 time 0.2702 (0.2921) loss 3.2483 (3.8284) grad_norm 1.1473 (inf) [2021-04-15 20:17:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][380/1251] eta 0:04:14 lr 0.000892 time 0.2784 (0.2917) loss 4.5692 (3.8324) grad_norm 1.1732 (inf) [2021-04-15 20:17:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][390/1251] eta 0:04:10 lr 0.000892 time 0.2725 (0.2915) loss 3.5034 (3.8293) grad_norm 1.0756 (inf) [2021-04-15 20:17:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][400/1251] eta 0:04:07 lr 0.000892 time 0.3041 (0.2912) loss 3.8974 (3.8368) grad_norm 1.4848 (inf) [2021-04-15 20:17:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][410/1251] eta 0:04:04 lr 0.000892 time 0.2749 (0.2910) loss 3.8717 (3.8347) grad_norm 1.1140 (inf) [2021-04-15 20:17:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][420/1251] eta 0:04:01 lr 0.000892 time 0.2655 (0.2908) loss 4.2354 (3.8384) grad_norm 1.0679 (inf) [2021-04-15 20:17:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][430/1251] eta 0:03:58 lr 0.000892 time 0.2962 (0.2906) loss 4.4695 (3.8366) grad_norm 1.0145 (inf) [2021-04-15 20:17:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][440/1251] eta 0:03:55 lr 0.000892 time 0.2734 (0.2903) loss 4.2482 (3.8441) grad_norm 1.0570 (inf) [2021-04-15 20:17:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][450/1251] eta 0:03:52 lr 0.000892 time 0.3012 (0.2900) loss 4.1636 (3.8462) grad_norm 1.2778 (inf) [2021-04-15 20:17:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][460/1251] eta 0:03:49 lr 0.000892 time 0.2745 (0.2898) loss 3.0099 (3.8461) grad_norm 1.0723 (inf) [2021-04-15 20:17:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][470/1251] eta 0:03:46 lr 0.000892 time 0.2625 (0.2895) loss 3.8439 (3.8416) grad_norm 1.0468 (inf) [2021-04-15 20:17:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][480/1251] eta 0:03:43 lr 0.000892 time 0.2759 (0.2893) loss 4.6373 (3.8399) grad_norm 1.2294 (inf) [2021-04-15 20:17:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][490/1251] eta 0:03:40 lr 0.000892 time 0.2670 (0.2894) loss 4.8146 (3.8441) grad_norm 1.0947 (inf) [2021-04-15 20:18:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][500/1251] eta 0:03:37 lr 0.000892 time 0.2978 (0.2892) loss 3.4317 (3.8405) grad_norm 1.1809 (inf) [2021-04-15 20:18:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][510/1251] eta 0:03:34 lr 0.000892 time 0.3275 (0.2891) loss 2.7216 (3.8401) grad_norm 0.9403 (inf) [2021-04-15 20:18:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][520/1251] eta 0:03:31 lr 0.000892 time 0.2693 (0.2889) loss 4.7370 (3.8456) grad_norm 1.1791 (inf) [2021-04-15 20:18:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][530/1251] eta 0:03:28 lr 0.000892 time 0.2855 (0.2887) loss 2.5805 (3.8437) grad_norm 1.1987 (inf) [2021-04-15 20:18:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][540/1251] eta 0:03:25 lr 0.000892 time 0.2654 (0.2886) loss 4.1269 (3.8433) grad_norm 1.0877 (inf) [2021-04-15 20:18:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][550/1251] eta 0:03:22 lr 0.000892 time 0.2707 (0.2884) loss 4.0270 (3.8387) grad_norm 1.1648 (inf) [2021-04-15 20:18:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][560/1251] eta 0:03:19 lr 0.000891 time 0.3148 (0.2883) loss 3.8762 (3.8365) grad_norm 1.2323 (inf) [2021-04-15 20:18:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][570/1251] eta 0:03:16 lr 0.000891 time 0.2974 (0.2884) loss 4.2642 (3.8424) grad_norm 1.1963 (inf) [2021-04-15 20:18:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][580/1251] eta 0:03:13 lr 0.000891 time 0.2999 (0.2885) loss 3.4397 (3.8420) grad_norm 1.1724 (inf) [2021-04-15 20:18:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][590/1251] eta 0:03:10 lr 0.000891 time 0.2770 (0.2883) loss 3.7737 (3.8420) grad_norm 1.2670 (inf) [2021-04-15 20:18:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][600/1251] eta 0:03:07 lr 0.000891 time 0.2807 (0.2882) loss 4.1186 (3.8432) grad_norm 1.0392 (inf) [2021-04-15 20:18:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][610/1251] eta 0:03:04 lr 0.000891 time 0.3039 (0.2881) loss 4.7085 (3.8453) grad_norm 0.9716 (inf) [2021-04-15 20:18:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][620/1251] eta 0:03:01 lr 0.000891 time 0.3044 (0.2880) loss 4.5821 (3.8491) grad_norm 1.1919 (inf) [2021-04-15 20:18:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][630/1251] eta 0:02:58 lr 0.000891 time 0.2883 (0.2879) loss 4.1684 (3.8520) grad_norm 1.1691 (inf) [2021-04-15 20:18:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][640/1251] eta 0:02:55 lr 0.000891 time 0.2853 (0.2877) loss 3.1868 (3.8542) grad_norm 1.0924 (inf) [2021-04-15 20:18:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][650/1251] eta 0:02:52 lr 0.000891 time 0.2953 (0.2876) loss 3.9579 (3.8583) grad_norm 1.0226 (inf) [2021-04-15 20:18:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][660/1251] eta 0:02:49 lr 0.000891 time 0.2707 (0.2875) loss 2.9549 (3.8552) grad_norm 1.2570 (inf) [2021-04-15 20:18:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][670/1251] eta 0:02:46 lr 0.000891 time 0.2884 (0.2874) loss 3.8472 (3.8552) grad_norm 0.9948 (inf) [2021-04-15 20:18:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][680/1251] eta 0:02:43 lr 0.000891 time 0.2782 (0.2872) loss 4.0176 (3.8537) grad_norm 1.1547 (inf) [2021-04-15 20:18:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][690/1251] eta 0:02:41 lr 0.000891 time 0.2550 (0.2871) loss 4.7856 (3.8514) grad_norm 1.0881 (inf) [2021-04-15 20:18:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][700/1251] eta 0:02:38 lr 0.000891 time 0.2768 (0.2870) loss 4.3782 (3.8533) grad_norm 1.1498 (inf) [2021-04-15 20:19:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][710/1251] eta 0:02:35 lr 0.000891 time 0.2570 (0.2868) loss 2.9437 (3.8520) grad_norm 1.0476 (inf) [2021-04-15 20:19:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][720/1251] eta 0:02:32 lr 0.000891 time 0.2816 (0.2869) loss 3.8360 (3.8526) grad_norm 1.2231 (inf) [2021-04-15 20:19:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][730/1251] eta 0:02:29 lr 0.000891 time 0.2654 (0.2872) loss 3.8536 (3.8528) grad_norm 1.0373 (inf) [2021-04-15 20:19:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][740/1251] eta 0:02:26 lr 0.000891 time 0.2948 (0.2873) loss 3.7221 (3.8529) grad_norm 1.0334 (inf) [2021-04-15 20:19:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][750/1251] eta 0:02:23 lr 0.000891 time 0.2805 (0.2871) loss 3.7550 (3.8540) grad_norm 1.1970 (inf) [2021-04-15 20:19:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][760/1251] eta 0:02:20 lr 0.000891 time 0.2843 (0.2870) loss 3.8891 (3.8557) grad_norm 1.0160 (inf) [2021-04-15 20:19:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][770/1251] eta 0:02:17 lr 0.000891 time 0.2666 (0.2869) loss 3.8462 (3.8525) grad_norm 1.0071 (inf) [2021-04-15 20:19:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][780/1251] eta 0:02:15 lr 0.000891 time 0.2906 (0.2868) loss 3.0086 (3.8500) grad_norm 1.1648 (inf) [2021-04-15 20:19:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][790/1251] eta 0:02:12 lr 0.000891 time 0.2711 (0.2868) loss 4.1925 (3.8515) grad_norm 1.3331 (inf) [2021-04-15 20:19:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][800/1251] eta 0:02:09 lr 0.000891 time 0.2686 (0.2867) loss 3.9335 (3.8494) grad_norm 1.1291 (inf) [2021-04-15 20:19:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][810/1251] eta 0:02:06 lr 0.000891 time 0.2847 (0.2868) loss 4.4754 (3.8480) grad_norm 1.0724 (inf) [2021-04-15 20:19:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][820/1251] eta 0:02:03 lr 0.000891 time 0.2610 (0.2868) loss 4.1468 (3.8460) grad_norm 1.1326 (inf) [2021-04-15 20:19:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][830/1251] eta 0:02:00 lr 0.000891 time 0.3050 (0.2868) loss 4.2688 (3.8454) grad_norm 1.1328 (inf) [2021-04-15 20:19:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][840/1251] eta 0:01:57 lr 0.000891 time 0.2697 (0.2868) loss 4.1858 (3.8451) grad_norm 1.0461 (inf) [2021-04-15 20:19:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][850/1251] eta 0:01:54 lr 0.000891 time 0.2624 (0.2868) loss 3.9517 (3.8436) grad_norm 1.2910 (inf) [2021-04-15 20:19:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][860/1251] eta 0:01:52 lr 0.000891 time 0.2776 (0.2867) loss 4.7954 (3.8425) grad_norm 1.1813 (inf) [2021-04-15 20:19:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][870/1251] eta 0:01:49 lr 0.000891 time 0.2736 (0.2867) loss 3.8095 (3.8427) grad_norm 1.0708 (inf) [2021-04-15 20:19:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][880/1251] eta 0:01:46 lr 0.000891 time 0.3087 (0.2867) loss 3.8991 (3.8463) grad_norm 1.1194 (inf) [2021-04-15 20:19:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][890/1251] eta 0:01:43 lr 0.000891 time 0.2679 (0.2866) loss 4.0287 (3.8477) grad_norm 1.1670 (inf) [2021-04-15 20:19:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][900/1251] eta 0:01:40 lr 0.000891 time 0.2715 (0.2865) loss 2.7163 (3.8465) grad_norm 1.0989 (inf) [2021-04-15 20:19:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][910/1251] eta 0:01:37 lr 0.000891 time 0.2877 (0.2865) loss 4.2383 (3.8469) grad_norm 1.0159 (inf) [2021-04-15 20:20:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][920/1251] eta 0:01:34 lr 0.000891 time 0.2826 (0.2864) loss 4.2397 (3.8455) grad_norm 1.3145 (inf) [2021-04-15 20:20:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][930/1251] eta 0:01:31 lr 0.000891 time 0.2798 (0.2865) loss 4.2032 (3.8453) grad_norm 1.3991 (inf) [2021-04-15 20:20:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][940/1251] eta 0:01:29 lr 0.000890 time 0.2865 (0.2866) loss 2.6255 (3.8428) grad_norm 1.3644 (inf) [2021-04-15 20:20:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][950/1251] eta 0:01:26 lr 0.000890 time 0.2987 (0.2865) loss 3.7631 (3.8449) grad_norm 1.1306 (inf) [2021-04-15 20:20:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][960/1251] eta 0:01:23 lr 0.000890 time 0.2745 (0.2865) loss 3.8629 (3.8409) grad_norm 0.9833 (inf) [2021-04-15 20:20:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][970/1251] eta 0:01:20 lr 0.000890 time 0.2732 (0.2864) loss 4.2577 (3.8409) grad_norm 1.0215 (inf) [2021-04-15 20:20:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][980/1251] eta 0:01:17 lr 0.000890 time 0.3170 (0.2864) loss 4.5041 (3.8379) grad_norm 1.4298 (inf) [2021-04-15 20:20:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][990/1251] eta 0:01:14 lr 0.000890 time 0.2977 (0.2863) loss 4.8172 (3.8394) grad_norm 1.1229 (inf) [2021-04-15 20:20:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][1000/1251] eta 0:01:11 lr 0.000890 time 0.2675 (0.2862) loss 4.3187 (3.8395) grad_norm 1.0149 (inf) [2021-04-15 20:20:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][1010/1251] eta 0:01:08 lr 0.000890 time 0.3201 (0.2862) loss 4.0153 (3.8389) grad_norm 1.2313 (inf) [2021-04-15 20:20:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][1020/1251] eta 0:01:06 lr 0.000890 time 0.2947 (0.2862) loss 3.6414 (3.8389) grad_norm 1.2501 (inf) [2021-04-15 20:20:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][1030/1251] eta 0:01:03 lr 0.000890 time 0.2827 (0.2862) loss 4.8936 (3.8398) grad_norm 0.9221 (inf) [2021-04-15 20:20:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][1040/1251] eta 0:01:00 lr 0.000890 time 0.2956 (0.2861) loss 4.5980 (3.8410) grad_norm 0.9715 (inf) [2021-04-15 20:20:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][1050/1251] eta 0:00:57 lr 0.000890 time 0.2924 (0.2861) loss 4.3691 (3.8407) grad_norm 1.0648 (inf) [2021-04-15 20:20:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][1060/1251] eta 0:00:54 lr 0.000890 time 0.2538 (0.2860) loss 4.0607 (3.8409) grad_norm 0.9795 (inf) [2021-04-15 20:20:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][1070/1251] eta 0:00:51 lr 0.000890 time 0.2895 (0.2861) loss 4.2650 (3.8434) grad_norm 1.4045 (inf) [2021-04-15 20:20:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][1080/1251] eta 0:00:48 lr 0.000890 time 0.2806 (0.2861) loss 3.1723 (3.8381) grad_norm 1.1543 (inf) [2021-04-15 20:20:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][1090/1251] eta 0:00:46 lr 0.000890 time 0.3000 (0.2861) loss 3.3606 (3.8384) grad_norm 1.1770 (inf) [2021-04-15 20:20:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][1100/1251] eta 0:00:43 lr 0.000890 time 0.2755 (0.2860) loss 3.8252 (3.8391) grad_norm 1.1066 (inf) [2021-04-15 20:20:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][1110/1251] eta 0:00:40 lr 0.000890 time 0.2737 (0.2859) loss 4.0545 (3.8399) grad_norm 1.0911 (inf) [2021-04-15 20:20:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][1120/1251] eta 0:00:37 lr 0.000890 time 0.2865 (0.2858) loss 4.1803 (3.8376) grad_norm 1.1300 (inf) [2021-04-15 20:21:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][1130/1251] eta 0:00:34 lr 0.000890 time 0.2718 (0.2858) loss 3.0965 (3.8368) grad_norm 0.9715 (inf) [2021-04-15 20:21:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][1140/1251] eta 0:00:31 lr 0.000890 time 0.2536 (0.2858) loss 4.1165 (3.8382) grad_norm 1.2550 (inf) [2021-04-15 20:21:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][1150/1251] eta 0:00:28 lr 0.000890 time 0.2655 (0.2859) loss 4.3352 (3.8412) grad_norm 1.0828 (inf) [2021-04-15 20:21:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][1160/1251] eta 0:00:26 lr 0.000890 time 0.2907 (0.2859) loss 2.8937 (3.8427) grad_norm 1.0956 (inf) [2021-04-15 20:21:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][1170/1251] eta 0:00:23 lr 0.000890 time 0.2752 (0.2858) loss 4.3763 (3.8428) grad_norm 1.2159 (inf) [2021-04-15 20:21:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][1180/1251] eta 0:00:20 lr 0.000890 time 0.2878 (0.2858) loss 3.6952 (3.8413) grad_norm 1.1366 (inf) [2021-04-15 20:21:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][1190/1251] eta 0:00:17 lr 0.000890 time 0.2920 (0.2857) loss 4.4194 (3.8414) grad_norm 1.1481 (inf) [2021-04-15 20:21:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][1200/1251] eta 0:00:14 lr 0.000890 time 0.2801 (0.2857) loss 4.7166 (3.8402) grad_norm 1.3884 (inf) [2021-04-15 20:21:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][1210/1251] eta 0:00:11 lr 0.000890 time 0.2953 (0.2857) loss 3.9812 (3.8411) grad_norm 1.0064 (inf) [2021-04-15 20:21:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][1220/1251] eta 0:00:08 lr 0.000890 time 0.2628 (0.2856) loss 3.5877 (3.8363) grad_norm 1.0384 (inf) [2021-04-15 20:21:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][1230/1251] eta 0:00:05 lr 0.000890 time 0.2626 (0.2855) loss 3.5024 (3.8364) grad_norm 1.2339 (inf) [2021-04-15 20:21:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][1240/1251] eta 0:00:03 lr 0.000890 time 0.2379 (0.2854) loss 3.9369 (3.8384) grad_norm 1.1225 (inf) [2021-04-15 20:21:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [64/300][1250/1251] eta 0:00:00 lr 0.000890 time 0.2489 (0.2851) loss 4.5632 (3.8398) grad_norm 1.1469 (inf) [2021-04-15 20:21:36 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 64 training takes 0:05:58 [2021-04-15 20:21:36 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_64.pth saving...... [2021-04-15 20:21:45 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_64.pth saved !!! [2021-04-15 20:21:46 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.325 (1.325) Loss 1.2534 (1.2534) Acc@1 69.922 (69.922) Acc@5 90.332 (90.332) [2021-04-15 20:21:48 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.170 (0.240) Loss 1.2386 (1.2247) Acc@1 72.754 (71.671) Acc@5 90.234 (90.732) [2021-04-15 20:21:50 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.416 (0.247) Loss 1.2780 (1.2172) Acc@1 71.484 (71.763) Acc@5 89.160 (90.783) [2021-04-15 20:21:52 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.299 (0.238) Loss 1.2530 (1.2162) Acc@1 72.363 (71.727) Acc@5 90.039 (90.729) [2021-04-15 20:21:54 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.198 (0.224) Loss 1.2271 (1.2192) Acc@1 72.754 (71.639) Acc@5 91.602 (90.732) [2021-04-15 20:21:58 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 71.588 Acc@5 90.780 [2021-04-15 20:21:58 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 71.6% [2021-04-15 20:21:58 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 71.59% [2021-04-15 20:22:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][0/1251] eta 1:24:39 lr 0.000890 time 4.0607 (4.0607) loss 3.3023 (3.3023) grad_norm 1.1631 (1.1631) [2021-04-15 20:22:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][10/1251] eta 0:12:53 lr 0.000890 time 0.2778 (0.6230) loss 3.7929 (3.6721) grad_norm 1.2083 (1.1547) [2021-04-15 20:22:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][20/1251] eta 0:09:37 lr 0.000890 time 0.2797 (0.4691) loss 2.7540 (3.6694) grad_norm 1.0777 (1.1511) [2021-04-15 20:22:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][30/1251] eta 0:08:19 lr 0.000890 time 0.3063 (0.4093) loss 2.7599 (3.7543) grad_norm 1.1067 (1.1425) [2021-04-15 20:22:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][40/1251] eta 0:07:40 lr 0.000890 time 0.2875 (0.3802) loss 3.0580 (3.7292) grad_norm 1.2155 (1.1478) [2021-04-15 20:22:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][50/1251] eta 0:07:14 lr 0.000890 time 0.2924 (0.3615) loss 4.7078 (3.7385) grad_norm 1.0973 (1.1466) [2021-04-15 20:22:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][60/1251] eta 0:06:55 lr 0.000890 time 0.2879 (0.3487) loss 4.1418 (3.8096) grad_norm 1.1494 (1.1551) [2021-04-15 20:22:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][70/1251] eta 0:06:40 lr 0.000890 time 0.2790 (0.3391) loss 2.9145 (3.7853) grad_norm 1.0666 (1.1604) [2021-04-15 20:22:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][80/1251] eta 0:06:31 lr 0.000889 time 0.2735 (0.3339) loss 4.2554 (3.7780) grad_norm 1.3225 (1.1681) [2021-04-15 20:22:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][90/1251] eta 0:06:22 lr 0.000889 time 0.2888 (0.3293) loss 4.5066 (3.7972) grad_norm 1.5714 (1.1781) [2021-04-15 20:22:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][100/1251] eta 0:06:14 lr 0.000889 time 0.2938 (0.3250) loss 3.6486 (3.7768) grad_norm 1.1581 (1.1835) [2021-04-15 20:22:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][110/1251] eta 0:06:06 lr 0.000889 time 0.2746 (0.3210) loss 4.6169 (3.7834) grad_norm 1.1942 (1.1814) [2021-04-15 20:22:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][120/1251] eta 0:05:59 lr 0.000889 time 0.2747 (0.3181) loss 3.8341 (3.8112) grad_norm 1.1025 (1.1719) [2021-04-15 20:22:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][130/1251] eta 0:05:53 lr 0.000889 time 0.2879 (0.3151) loss 3.8102 (3.7979) grad_norm 1.1250 (1.1660) [2021-04-15 20:22:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][140/1251] eta 0:05:49 lr 0.000889 time 0.2872 (0.3144) loss 4.5163 (3.8265) grad_norm 1.2478 (1.1667) [2021-04-15 20:22:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][150/1251] eta 0:05:44 lr 0.000889 time 0.2729 (0.3132) loss 3.6496 (3.8324) grad_norm 1.3075 (1.1705) [2021-04-15 20:22:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][160/1251] eta 0:05:39 lr 0.000889 time 0.2772 (0.3116) loss 4.7478 (3.8343) grad_norm 1.5562 (1.1730) [2021-04-15 20:22:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][170/1251] eta 0:05:34 lr 0.000889 time 0.2895 (0.3099) loss 4.0354 (3.8434) grad_norm 1.0894 (1.1674) [2021-04-15 20:22:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][180/1251] eta 0:05:30 lr 0.000889 time 0.2892 (0.3082) loss 4.1731 (3.8369) grad_norm 1.0109 (1.1626) [2021-04-15 20:22:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][190/1251] eta 0:05:25 lr 0.000889 time 0.3081 (0.3070) loss 4.4787 (3.8438) grad_norm 1.1867 (1.1585) [2021-04-15 20:22:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][200/1251] eta 0:05:21 lr 0.000889 time 0.3081 (0.3057) loss 3.9173 (3.8318) grad_norm 1.0570 (1.1565) [2021-04-15 20:23:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][210/1251] eta 0:05:17 lr 0.000889 time 0.2897 (0.3048) loss 4.7736 (3.8292) grad_norm 1.1995 (1.1561) [2021-04-15 20:23:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][220/1251] eta 0:05:13 lr 0.000889 time 0.2423 (0.3038) loss 3.4226 (3.8275) grad_norm 1.2870 (1.1547) [2021-04-15 20:23:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][230/1251] eta 0:05:09 lr 0.000889 time 0.2978 (0.3029) loss 4.2823 (3.8301) grad_norm 0.9978 (1.1523) [2021-04-15 20:23:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][240/1251] eta 0:05:05 lr 0.000889 time 0.2670 (0.3023) loss 4.1744 (3.8448) grad_norm 0.9646 (1.1494) [2021-04-15 20:23:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][250/1251] eta 0:05:01 lr 0.000889 time 0.2597 (0.3015) loss 4.0144 (3.8381) grad_norm 1.1256 (1.1485) [2021-04-15 20:23:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][260/1251] eta 0:04:58 lr 0.000889 time 0.3105 (0.3009) loss 4.4891 (3.8408) grad_norm 1.2758 (1.1462) [2021-04-15 20:23:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][270/1251] eta 0:04:54 lr 0.000889 time 0.2880 (0.3002) loss 4.2734 (3.8456) grad_norm 1.2794 (1.1467) [2021-04-15 20:23:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][280/1251] eta 0:04:50 lr 0.000889 time 0.2853 (0.2996) loss 4.8943 (3.8483) grad_norm 1.6410 (1.1504) [2021-04-15 20:23:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][290/1251] eta 0:04:47 lr 0.000889 time 0.2987 (0.2995) loss 3.0339 (3.8414) grad_norm 1.4488 (1.1565) [2021-04-15 20:23:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][300/1251] eta 0:04:44 lr 0.000889 time 0.2606 (0.2989) loss 3.1922 (3.8330) grad_norm 1.2685 (1.1576) [2021-04-15 20:23:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][310/1251] eta 0:04:40 lr 0.000889 time 0.2669 (0.2983) loss 3.3263 (3.8432) grad_norm 1.2881 (1.1633) [2021-04-15 20:23:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][320/1251] eta 0:04:37 lr 0.000889 time 0.3171 (0.2979) loss 4.2048 (3.8508) grad_norm 1.2472 (1.1652) [2021-04-15 20:23:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][330/1251] eta 0:04:33 lr 0.000889 time 0.2466 (0.2974) loss 3.0864 (3.8562) grad_norm 1.0449 (1.1662) [2021-04-15 20:23:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][340/1251] eta 0:04:31 lr 0.000889 time 0.2813 (0.2975) loss 4.0798 (3.8559) grad_norm 1.1192 (1.1649) [2021-04-15 20:23:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][350/1251] eta 0:04:27 lr 0.000889 time 0.2602 (0.2974) loss 3.0627 (3.8666) grad_norm 1.1680 (1.1644) [2021-04-15 20:23:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][360/1251] eta 0:04:24 lr 0.000889 time 0.2526 (0.2974) loss 2.9202 (3.8678) grad_norm 1.0711 (1.1627) [2021-04-15 20:23:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][370/1251] eta 0:04:21 lr 0.000889 time 0.2701 (0.2970) loss 3.4540 (3.8628) grad_norm 1.0565 (1.1609) [2021-04-15 20:23:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][380/1251] eta 0:04:18 lr 0.000889 time 0.2528 (0.2965) loss 3.7853 (3.8653) grad_norm 1.0646 (1.1596) [2021-04-15 20:23:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][390/1251] eta 0:04:14 lr 0.000889 time 0.2768 (0.2962) loss 4.0213 (3.8654) grad_norm 1.1659 (1.1590) [2021-04-15 20:23:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][400/1251] eta 0:04:11 lr 0.000889 time 0.2927 (0.2958) loss 3.9850 (3.8650) grad_norm 1.1900 (1.1574) [2021-04-15 20:23:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][410/1251] eta 0:04:08 lr 0.000889 time 0.2813 (0.2956) loss 2.9016 (3.8629) grad_norm 1.1039 (1.1557) [2021-04-15 20:24:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][420/1251] eta 0:04:05 lr 0.000889 time 0.2954 (0.2953) loss 4.2070 (3.8694) grad_norm 1.2252 (1.1553) [2021-04-15 20:24:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][430/1251] eta 0:04:02 lr 0.000889 time 0.2646 (0.2951) loss 4.3483 (3.8784) grad_norm 1.0332 (1.1523) [2021-04-15 20:24:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][440/1251] eta 0:03:58 lr 0.000889 time 0.2710 (0.2947) loss 3.9952 (3.8730) grad_norm 0.9366 (1.1514) [2021-04-15 20:24:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][450/1251] eta 0:03:55 lr 0.000889 time 0.2584 (0.2944) loss 3.2424 (3.8720) grad_norm 1.1842 (1.1494) [2021-04-15 20:24:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][460/1251] eta 0:03:52 lr 0.000888 time 0.2865 (0.2942) loss 4.8448 (3.8741) grad_norm 1.1235 (1.1481) [2021-04-15 20:24:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][470/1251] eta 0:03:49 lr 0.000888 time 0.3123 (0.2942) loss 4.1084 (3.8757) grad_norm 0.9521 (1.1464) [2021-04-15 20:24:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][480/1251] eta 0:03:46 lr 0.000888 time 0.2736 (0.2940) loss 3.8807 (3.8716) grad_norm 1.2051 (1.1448) [2021-04-15 20:24:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][490/1251] eta 0:03:43 lr 0.000888 time 0.2994 (0.2939) loss 4.1619 (3.8668) grad_norm 1.1334 (1.1447) [2021-04-15 20:24:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][500/1251] eta 0:03:40 lr 0.000888 time 0.2793 (0.2937) loss 4.5292 (3.8725) grad_norm 1.1913 (1.1447) [2021-04-15 20:24:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][510/1251] eta 0:03:37 lr 0.000888 time 0.2817 (0.2934) loss 3.9158 (3.8690) grad_norm 0.9938 (1.1444) [2021-04-15 20:24:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][520/1251] eta 0:03:34 lr 0.000888 time 0.2786 (0.2931) loss 4.0607 (3.8630) grad_norm 1.1393 (1.1458) [2021-04-15 20:24:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][530/1251] eta 0:03:31 lr 0.000888 time 0.3080 (0.2930) loss 3.6356 (3.8656) grad_norm 1.1431 (1.1456) [2021-04-15 20:24:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][540/1251] eta 0:03:28 lr 0.000888 time 0.2916 (0.2930) loss 3.6409 (3.8670) grad_norm 1.0507 (1.1464) [2021-04-15 20:24:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][550/1251] eta 0:03:25 lr 0.000888 time 0.2594 (0.2928) loss 4.0556 (3.8710) grad_norm 1.1088 (1.1458) [2021-04-15 20:24:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][560/1251] eta 0:03:22 lr 0.000888 time 0.2667 (0.2926) loss 2.7675 (3.8679) grad_norm 1.0375 (1.1444) [2021-04-15 20:24:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][570/1251] eta 0:03:19 lr 0.000888 time 0.2684 (0.2926) loss 2.3143 (3.8664) grad_norm 1.1531 (1.1443) [2021-04-15 20:24:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][580/1251] eta 0:03:16 lr 0.000888 time 0.3010 (0.2927) loss 4.1852 (3.8676) grad_norm 1.0395 (1.1435) [2021-04-15 20:24:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][590/1251] eta 0:03:13 lr 0.000888 time 0.2865 (0.2929) loss 3.0926 (3.8666) grad_norm 1.0861 (1.1420) [2021-04-15 20:24:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][600/1251] eta 0:03:10 lr 0.000888 time 0.2800 (0.2926) loss 3.4858 (3.8628) grad_norm 1.1699 (1.1425) [2021-04-15 20:24:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][610/1251] eta 0:03:07 lr 0.000888 time 0.2997 (0.2925) loss 4.2940 (3.8565) grad_norm 1.3197 (1.1413) [2021-04-15 20:24:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][620/1251] eta 0:03:04 lr 0.000888 time 0.2647 (0.2925) loss 4.0818 (3.8539) grad_norm 0.9899 (1.1405) [2021-04-15 20:25:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][630/1251] eta 0:03:01 lr 0.000888 time 0.2828 (0.2923) loss 3.8171 (3.8534) grad_norm 1.0701 (1.1401) [2021-04-15 20:25:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][640/1251] eta 0:02:58 lr 0.000888 time 0.2884 (0.2922) loss 3.0389 (3.8533) grad_norm 1.1771 (1.1399) [2021-04-15 20:25:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][650/1251] eta 0:02:55 lr 0.000888 time 0.4541 (0.2923) loss 3.1021 (3.8513) grad_norm 1.0587 (1.1393) [2021-04-15 20:25:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][660/1251] eta 0:02:52 lr 0.000888 time 0.2822 (0.2921) loss 3.9748 (3.8528) grad_norm 1.0693 (1.1385) [2021-04-15 20:25:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][670/1251] eta 0:02:49 lr 0.000888 time 0.2810 (0.2919) loss 4.2026 (3.8526) grad_norm 0.9863 (1.1382) [2021-04-15 20:25:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][680/1251] eta 0:02:46 lr 0.000888 time 0.3178 (0.2919) loss 3.3852 (3.8527) grad_norm 1.1787 (1.1388) [2021-04-15 20:25:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][690/1251] eta 0:02:43 lr 0.000888 time 0.2745 (0.2916) loss 3.2276 (3.8544) grad_norm 1.0981 (1.1390) [2021-04-15 20:25:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][700/1251] eta 0:02:40 lr 0.000888 time 0.2828 (0.2915) loss 3.3874 (3.8534) grad_norm 1.0020 (1.1387) [2021-04-15 20:25:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][710/1251] eta 0:02:37 lr 0.000888 time 0.2952 (0.2915) loss 4.0601 (3.8546) grad_norm 1.0399 (1.1385) [2021-04-15 20:25:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][720/1251] eta 0:02:34 lr 0.000888 time 0.2540 (0.2915) loss 4.6028 (3.8592) grad_norm 0.9724 (1.1384) [2021-04-15 20:25:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][730/1251] eta 0:02:31 lr 0.000888 time 0.3052 (0.2916) loss 3.4540 (3.8554) grad_norm 1.2618 (1.1385) [2021-04-15 20:25:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][740/1251] eta 0:02:28 lr 0.000888 time 0.2973 (0.2915) loss 3.3491 (3.8548) grad_norm 1.2038 (1.1387) [2021-04-15 20:25:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][750/1251] eta 0:02:25 lr 0.000888 time 0.2796 (0.2913) loss 2.6340 (3.8531) grad_norm 1.3978 (1.1400) [2021-04-15 20:25:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][760/1251] eta 0:02:22 lr 0.000888 time 0.2713 (0.2912) loss 3.1415 (3.8555) grad_norm 1.1217 (1.1405) [2021-04-15 20:25:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][770/1251] eta 0:02:20 lr 0.000888 time 0.2569 (0.2911) loss 4.3614 (3.8549) grad_norm 1.0379 (1.1422) [2021-04-15 20:25:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][780/1251] eta 0:02:17 lr 0.000888 time 0.2820 (0.2910) loss 3.6482 (3.8564) grad_norm 1.0438 (1.1423) [2021-04-15 20:25:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][790/1251] eta 0:02:14 lr 0.000888 time 0.2849 (0.2908) loss 3.8252 (3.8615) grad_norm 1.1052 (1.1426) [2021-04-15 20:25:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][800/1251] eta 0:02:11 lr 0.000888 time 0.2810 (0.2907) loss 2.8592 (3.8613) grad_norm 1.2268 (1.1421) [2021-04-15 20:25:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][810/1251] eta 0:02:08 lr 0.000888 time 0.2910 (0.2906) loss 3.7835 (3.8608) grad_norm 1.0722 (1.1422) [2021-04-15 20:25:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][820/1251] eta 0:02:05 lr 0.000888 time 0.2833 (0.2905) loss 4.0189 (3.8637) grad_norm 1.1024 (1.1418) [2021-04-15 20:25:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][830/1251] eta 0:02:02 lr 0.000888 time 0.2636 (0.2903) loss 3.3645 (3.8643) grad_norm 1.2438 (1.1426) [2021-04-15 20:26:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][840/1251] eta 0:01:59 lr 0.000887 time 0.2672 (0.2902) loss 3.0511 (3.8630) grad_norm 1.2843 (1.1429) [2021-04-15 20:26:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][850/1251] eta 0:01:56 lr 0.000887 time 0.2844 (0.2901) loss 3.6607 (3.8627) grad_norm 1.1744 (1.1433) [2021-04-15 20:26:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][860/1251] eta 0:01:53 lr 0.000887 time 0.2886 (0.2900) loss 2.5202 (3.8596) grad_norm 1.3016 (1.1430) [2021-04-15 20:26:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][870/1251] eta 0:01:50 lr 0.000887 time 0.3035 (0.2901) loss 4.0322 (3.8606) grad_norm 1.1686 (1.1425) [2021-04-15 20:26:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][880/1251] eta 0:01:47 lr 0.000887 time 0.2858 (0.2900) loss 3.8644 (3.8616) grad_norm 1.1758 (1.1432) [2021-04-15 20:26:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][890/1251] eta 0:01:44 lr 0.000887 time 0.2710 (0.2899) loss 4.1961 (3.8624) grad_norm 1.1288 (1.1432) [2021-04-15 20:26:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][900/1251] eta 0:01:41 lr 0.000887 time 0.2717 (0.2898) loss 4.2400 (3.8631) grad_norm 1.1249 (1.1431) [2021-04-15 20:26:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][910/1251] eta 0:01:38 lr 0.000887 time 0.2989 (0.2897) loss 3.9944 (3.8623) grad_norm 1.0549 (1.1430) [2021-04-15 20:26:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][920/1251] eta 0:01:35 lr 0.000887 time 0.2889 (0.2897) loss 3.6585 (3.8603) grad_norm 0.9435 (1.1427) [2021-04-15 20:26:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][930/1251] eta 0:01:33 lr 0.000887 time 0.2701 (0.2897) loss 3.5476 (3.8598) grad_norm 1.0747 (1.1426) [2021-04-15 20:26:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][940/1251] eta 0:01:30 lr 0.000887 time 0.2712 (0.2900) loss 4.3333 (3.8596) grad_norm 1.3899 (1.1428) [2021-04-15 20:26:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][950/1251] eta 0:01:27 lr 0.000887 time 0.2755 (0.2899) loss 4.6216 (3.8622) grad_norm 1.1260 (1.1432) [2021-04-15 20:26:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][960/1251] eta 0:01:24 lr 0.000887 time 0.3069 (0.2898) loss 4.2441 (3.8645) grad_norm 1.4408 (1.1437) [2021-04-15 20:26:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][970/1251] eta 0:01:21 lr 0.000887 time 0.2805 (0.2898) loss 4.3943 (3.8630) grad_norm 1.0852 (1.1447) [2021-04-15 20:26:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][980/1251] eta 0:01:18 lr 0.000887 time 0.2466 (0.2898) loss 3.6420 (3.8613) grad_norm 1.0883 (1.1440) [2021-04-15 20:26:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][990/1251] eta 0:01:15 lr 0.000887 time 0.2751 (0.2897) loss 3.9304 (3.8634) grad_norm 1.0065 (1.1434) [2021-04-15 20:26:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][1000/1251] eta 0:01:12 lr 0.000887 time 0.2959 (0.2897) loss 3.1400 (3.8627) grad_norm 1.1746 (1.1429) [2021-04-15 20:26:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][1010/1251] eta 0:01:09 lr 0.000887 time 0.2726 (0.2896) loss 3.7506 (3.8632) grad_norm 1.0073 (1.1424) [2021-04-15 20:26:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][1020/1251] eta 0:01:06 lr 0.000887 time 0.2603 (0.2894) loss 4.4568 (3.8654) grad_norm 0.9404 (1.1420) [2021-04-15 20:26:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][1030/1251] eta 0:01:03 lr 0.000887 time 0.2782 (0.2894) loss 3.8520 (3.8646) grad_norm 0.9887 (1.1425) [2021-04-15 20:26:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][1040/1251] eta 0:01:01 lr 0.000887 time 0.2703 (0.2893) loss 3.4887 (3.8646) grad_norm 1.2172 (1.1418) [2021-04-15 20:27:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][1050/1251] eta 0:00:58 lr 0.000887 time 0.2771 (0.2893) loss 3.8730 (3.8617) grad_norm 1.0916 (1.1417) [2021-04-15 20:27:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][1060/1251] eta 0:00:55 lr 0.000887 time 0.2784 (0.2892) loss 3.6853 (3.8606) grad_norm 1.3215 (1.1421) [2021-04-15 20:27:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][1070/1251] eta 0:00:52 lr 0.000887 time 0.2642 (0.2891) loss 3.4396 (3.8599) grad_norm 1.0469 (1.1425) [2021-04-15 20:27:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][1080/1251] eta 0:00:49 lr 0.000887 time 0.2653 (0.2890) loss 4.5844 (3.8596) grad_norm 0.9913 (1.1423) [2021-04-15 20:27:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][1090/1251] eta 0:00:46 lr 0.000887 time 0.2879 (0.2889) loss 3.9032 (3.8604) grad_norm 1.1599 (1.1424) [2021-04-15 20:27:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][1100/1251] eta 0:00:43 lr 0.000887 time 0.2958 (0.2888) loss 3.8850 (3.8596) grad_norm 1.0757 (1.1424) [2021-04-15 20:27:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][1110/1251] eta 0:00:40 lr 0.000887 time 0.2746 (0.2887) loss 4.0518 (3.8596) grad_norm 1.4685 (1.1422) [2021-04-15 20:27:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][1120/1251] eta 0:00:37 lr 0.000887 time 0.2696 (0.2887) loss 3.5054 (3.8639) grad_norm 1.1743 (1.1419) [2021-04-15 20:27:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][1130/1251] eta 0:00:34 lr 0.000887 time 0.2852 (0.2887) loss 3.9145 (3.8645) grad_norm 1.2607 (1.1423) [2021-04-15 20:27:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][1140/1251] eta 0:00:32 lr 0.000887 time 0.2645 (0.2887) loss 3.2897 (3.8641) grad_norm 1.3416 (1.1434) [2021-04-15 20:27:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][1150/1251] eta 0:00:29 lr 0.000887 time 0.2748 (0.2887) loss 4.3766 (3.8651) grad_norm 1.2991 (1.1436) [2021-04-15 20:27:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][1160/1251] eta 0:00:26 lr 0.000887 time 0.2960 (0.2887) loss 3.4977 (3.8639) grad_norm 1.0863 (1.1442) [2021-04-15 20:27:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][1170/1251] eta 0:00:23 lr 0.000887 time 0.2929 (0.2887) loss 4.6674 (3.8646) grad_norm 1.2228 (1.1447) [2021-04-15 20:27:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][1180/1251] eta 0:00:20 lr 0.000887 time 0.3020 (0.2887) loss 2.7545 (3.8648) grad_norm 1.2377 (1.1448) [2021-04-15 20:27:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][1190/1251] eta 0:00:17 lr 0.000887 time 0.2889 (0.2888) loss 2.9412 (3.8651) grad_norm 1.0631 (1.1453) [2021-04-15 20:27:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][1200/1251] eta 0:00:14 lr 0.000887 time 0.2601 (0.2887) loss 3.5814 (3.8653) grad_norm 1.3582 (1.1462) [2021-04-15 20:27:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][1210/1251] eta 0:00:11 lr 0.000887 time 0.2581 (0.2886) loss 3.3034 (3.8661) grad_norm 1.2217 (1.1461) [2021-04-15 20:27:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][1220/1251] eta 0:00:08 lr 0.000886 time 0.2730 (0.2886) loss 4.0333 (3.8667) grad_norm 1.0595 (1.1457) [2021-04-15 20:27:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][1230/1251] eta 0:00:06 lr 0.000886 time 0.2826 (0.2887) loss 4.3296 (3.8656) grad_norm 1.0479 (1.1452) [2021-04-15 20:27:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][1240/1251] eta 0:00:03 lr 0.000886 time 0.2589 (0.2885) loss 3.9270 (3.8678) grad_norm 1.1175 (1.1449) [2021-04-15 20:27:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [65/300][1250/1251] eta 0:00:00 lr 0.000886 time 0.2663 (0.2882) loss 4.3863 (3.8675) grad_norm 0.9631 (1.1444) [2021-04-15 20:28:01 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 65 training takes 0:06:03 [2021-04-15 20:28:01 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_65.pth saving...... [2021-04-15 20:28:16 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_65.pth saved !!! [2021-04-15 20:28:18 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.192 (1.192) Loss 1.2370 (1.2370) Acc@1 70.508 (70.508) Acc@5 90.527 (90.527) [2021-04-15 20:28:19 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.096 (0.225) Loss 1.2688 (1.2015) Acc@1 70.605 (71.760) Acc@5 89.648 (91.140) [2021-04-15 20:28:22 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.262 (0.269) Loss 1.2568 (1.2153) Acc@1 72.363 (71.601) Acc@5 90.234 (90.913) [2021-04-15 20:28:24 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.152 (0.238) Loss 1.1525 (1.2090) Acc@1 73.438 (71.645) Acc@5 92.188 (90.997) [2021-04-15 20:28:26 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.298 (0.232) Loss 1.2171 (1.2111) Acc@1 70.703 (71.613) Acc@5 90.820 (90.942) [2021-04-15 20:28:29 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 71.500 Acc@5 90.872 [2021-04-15 20:28:29 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 71.5% [2021-04-15 20:28:29 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 71.59% [2021-04-15 20:28:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][0/1251] eta 1:50:34 lr 0.000886 time 5.3037 (5.3037) loss 2.2108 (2.2108) grad_norm 1.1423 (1.1423) [2021-04-15 20:28:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][10/1251] eta 0:15:14 lr 0.000886 time 0.2811 (0.7371) loss 4.0102 (3.8136) grad_norm 1.1264 (1.1226) [2021-04-15 20:28:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][20/1251] eta 0:10:41 lr 0.000886 time 0.2932 (0.5215) loss 4.2636 (3.8480) grad_norm 1.1823 (1.1315) [2021-04-15 20:28:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][30/1251] eta 0:09:01 lr 0.000886 time 0.2945 (0.4438) loss 4.3342 (3.8803) grad_norm 1.0309 (1.1285) [2021-04-15 20:28:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][40/1251] eta 0:08:09 lr 0.000886 time 0.2628 (0.4046) loss 3.7825 (3.8327) grad_norm 1.0969 (1.1187) [2021-04-15 20:28:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][50/1251] eta 0:07:38 lr 0.000886 time 0.2809 (0.3819) loss 3.6436 (3.8742) grad_norm 1.0363 (1.1337) [2021-04-15 20:28:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][60/1251] eta 0:07:19 lr 0.000886 time 0.4660 (0.3687) loss 4.0061 (3.8270) grad_norm 1.1694 (1.1365) [2021-04-15 20:28:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][70/1251] eta 0:07:01 lr 0.000886 time 0.2826 (0.3568) loss 4.4064 (3.8699) grad_norm 0.9948 (1.1265) [2021-04-15 20:28:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][80/1251] eta 0:06:47 lr 0.000886 time 0.2707 (0.3479) loss 4.3429 (3.8922) grad_norm 1.0263 (1.1295) [2021-04-15 20:29:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][90/1251] eta 0:06:36 lr 0.000886 time 0.3038 (0.3412) loss 4.5226 (3.8798) grad_norm 1.0171 (1.1305) [2021-04-15 20:29:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][100/1251] eta 0:06:26 lr 0.000886 time 0.3285 (0.3360) loss 3.5258 (3.9002) grad_norm 1.3944 (1.1386) [2021-04-15 20:29:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][110/1251] eta 0:06:17 lr 0.000886 time 0.2852 (0.3311) loss 4.0087 (3.8819) grad_norm 1.3160 (1.1412) [2021-04-15 20:29:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][120/1251] eta 0:06:11 lr 0.000886 time 0.2971 (0.3286) loss 3.7838 (3.8807) grad_norm 1.3482 (1.1363) [2021-04-15 20:29:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][130/1251] eta 0:06:05 lr 0.000886 time 0.2803 (0.3260) loss 4.7375 (3.8865) grad_norm 1.2138 (1.1339) [2021-04-15 20:29:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][140/1251] eta 0:05:59 lr 0.000886 time 0.2917 (0.3233) loss 3.2324 (3.8559) grad_norm 1.0693 (1.1339) [2021-04-15 20:29:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][150/1251] eta 0:05:54 lr 0.000886 time 0.2690 (0.3218) loss 4.7789 (3.8408) grad_norm 1.2203 (1.1385) [2021-04-15 20:29:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][160/1251] eta 0:05:48 lr 0.000886 time 0.2612 (0.3191) loss 3.1989 (3.8564) grad_norm 1.2566 (1.1427) [2021-04-15 20:29:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][170/1251] eta 0:05:42 lr 0.000886 time 0.2826 (0.3170) loss 4.3295 (3.8501) grad_norm 1.3096 (1.1437) [2021-04-15 20:29:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][180/1251] eta 0:05:37 lr 0.000886 time 0.2795 (0.3150) loss 4.6358 (3.8519) grad_norm 1.2898 (1.1441) [2021-04-15 20:29:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][190/1251] eta 0:05:32 lr 0.000886 time 0.2476 (0.3130) loss 3.2278 (3.8530) grad_norm 1.2049 (1.1435) [2021-04-15 20:29:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][200/1251] eta 0:05:27 lr 0.000886 time 0.2835 (0.3113) loss 4.6025 (3.8549) grad_norm 1.1091 (1.1418) [2021-04-15 20:29:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][210/1251] eta 0:05:22 lr 0.000886 time 0.2925 (0.3101) loss 3.9256 (3.8621) grad_norm 1.1998 (1.1395) [2021-04-15 20:29:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][220/1251] eta 0:05:18 lr 0.000886 time 0.2829 (0.3088) loss 4.3034 (3.8630) grad_norm 1.1459 (1.1360) [2021-04-15 20:29:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][230/1251] eta 0:05:14 lr 0.000886 time 0.2742 (0.3076) loss 3.0323 (3.8592) grad_norm 1.1660 (1.1356) [2021-04-15 20:29:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][240/1251] eta 0:05:09 lr 0.000886 time 0.2751 (0.3064) loss 4.6047 (3.8551) grad_norm 1.0575 (1.1320) [2021-04-15 20:29:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][250/1251] eta 0:05:05 lr 0.000886 time 0.2845 (0.3056) loss 2.5054 (3.8498) grad_norm 1.2931 (1.1322) [2021-04-15 20:29:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][260/1251] eta 0:05:01 lr 0.000886 time 0.2679 (0.3047) loss 4.6645 (3.8553) grad_norm 1.1642 (1.1306) [2021-04-15 20:29:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][270/1251] eta 0:04:58 lr 0.000886 time 0.2784 (0.3038) loss 4.2287 (3.8554) grad_norm 1.3390 (1.1320) [2021-04-15 20:29:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][280/1251] eta 0:04:54 lr 0.000886 time 0.2580 (0.3031) loss 3.9774 (3.8518) grad_norm 1.2635 (1.1332) [2021-04-15 20:29:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][290/1251] eta 0:04:50 lr 0.000886 time 0.2942 (0.3025) loss 2.8838 (3.8481) grad_norm 1.5587 (1.1364) [2021-04-15 20:29:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][300/1251] eta 0:04:47 lr 0.000886 time 0.2959 (0.3020) loss 3.8546 (3.8585) grad_norm 1.1264 (1.1389) [2021-04-15 20:30:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][310/1251] eta 0:04:43 lr 0.000886 time 0.2684 (0.3014) loss 3.3460 (3.8572) grad_norm 1.0397 (1.1387) [2021-04-15 20:30:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][320/1251] eta 0:04:40 lr 0.000886 time 0.2838 (0.3012) loss 4.8807 (3.8514) grad_norm 1.0469 (1.1371) [2021-04-15 20:30:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][330/1251] eta 0:04:36 lr 0.000886 time 0.2796 (0.3007) loss 3.7769 (3.8439) grad_norm 1.1263 (1.1356) [2021-04-15 20:30:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][340/1251] eta 0:04:33 lr 0.000886 time 0.2804 (0.3005) loss 4.5732 (3.8465) grad_norm 1.2307 (1.1351) [2021-04-15 20:30:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][350/1251] eta 0:04:30 lr 0.000885 time 0.3136 (0.3004) loss 3.1787 (3.8473) grad_norm 1.0462 (1.1352) [2021-04-15 20:30:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][360/1251] eta 0:04:27 lr 0.000885 time 0.2896 (0.2998) loss 3.1119 (3.8399) grad_norm 1.2914 (1.1366) [2021-04-15 20:30:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][370/1251] eta 0:04:24 lr 0.000885 time 0.2660 (0.2997) loss 4.9817 (3.8475) grad_norm 1.1302 (1.1360) [2021-04-15 20:30:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][380/1251] eta 0:04:20 lr 0.000885 time 0.3079 (0.2992) loss 4.1219 (3.8531) grad_norm 1.1702 (1.1362) [2021-04-15 20:30:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][390/1251] eta 0:04:17 lr 0.000885 time 0.2890 (0.2991) loss 5.0189 (3.8464) grad_norm 1.1405 (1.1356) [2021-04-15 20:30:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][400/1251] eta 0:04:14 lr 0.000885 time 0.2606 (0.2986) loss 3.6619 (3.8423) grad_norm 1.4496 (1.1358) [2021-04-15 20:30:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][410/1251] eta 0:04:10 lr 0.000885 time 0.2810 (0.2983) loss 3.4184 (3.8404) grad_norm 1.0479 (1.1395) [2021-04-15 20:30:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][420/1251] eta 0:04:07 lr 0.000885 time 0.2603 (0.2979) loss 3.4555 (3.8411) grad_norm 1.1515 (1.1396) [2021-04-15 20:30:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][430/1251] eta 0:04:04 lr 0.000885 time 0.2968 (0.2978) loss 4.0533 (3.8437) grad_norm 0.9222 (1.1382) [2021-04-15 20:30:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][440/1251] eta 0:04:01 lr 0.000885 time 0.3080 (0.2975) loss 3.6936 (3.8426) grad_norm 1.1280 (1.1384) [2021-04-15 20:30:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][450/1251] eta 0:03:58 lr 0.000885 time 0.3046 (0.2972) loss 4.3756 (3.8440) grad_norm 0.9934 (1.1383) [2021-04-15 20:30:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][460/1251] eta 0:03:54 lr 0.000885 time 0.2630 (0.2968) loss 2.5025 (3.8388) grad_norm 1.1208 (1.1404) [2021-04-15 20:30:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][470/1251] eta 0:03:51 lr 0.000885 time 0.2882 (0.2964) loss 3.7904 (3.8379) grad_norm 1.0636 (1.1395) [2021-04-15 20:30:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][480/1251] eta 0:03:48 lr 0.000885 time 0.2852 (0.2962) loss 3.0386 (3.8401) grad_norm 1.0224 (1.1378) [2021-04-15 20:30:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][490/1251] eta 0:03:45 lr 0.000885 time 0.2678 (0.2958) loss 2.9038 (3.8396) grad_norm 1.0284 (1.1387) [2021-04-15 20:30:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][500/1251] eta 0:03:42 lr 0.000885 time 0.2874 (0.2956) loss 4.3526 (3.8452) grad_norm 1.0964 (1.1392) [2021-04-15 20:30:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][510/1251] eta 0:03:38 lr 0.000885 time 0.2590 (0.2953) loss 4.3056 (3.8479) grad_norm 1.0982 (1.1387) [2021-04-15 20:31:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][520/1251] eta 0:03:35 lr 0.000885 time 0.2811 (0.2952) loss 4.2530 (3.8479) grad_norm 1.0054 (1.1376) [2021-04-15 20:31:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][530/1251] eta 0:03:32 lr 0.000885 time 0.2982 (0.2950) loss 4.7419 (3.8518) grad_norm 1.0755 (1.1375) [2021-04-15 20:31:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][540/1251] eta 0:03:29 lr 0.000885 time 0.3089 (0.2948) loss 4.0459 (3.8540) grad_norm 1.1377 (1.1378) [2021-04-15 20:31:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][550/1251] eta 0:03:26 lr 0.000885 time 0.3005 (0.2946) loss 2.8680 (3.8513) grad_norm 1.1333 (1.1402) [2021-04-15 20:31:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][560/1251] eta 0:03:23 lr 0.000885 time 0.3044 (0.2944) loss 3.1333 (3.8504) grad_norm 1.2584 (1.1403) [2021-04-15 20:31:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][570/1251] eta 0:03:20 lr 0.000885 time 0.2452 (0.2947) loss 3.6570 (3.8494) grad_norm 1.1748 (1.1419) [2021-04-15 20:31:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][580/1251] eta 0:03:17 lr 0.000885 time 0.2823 (0.2948) loss 3.5793 (3.8486) grad_norm 1.3617 (1.1431) [2021-04-15 20:31:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][590/1251] eta 0:03:14 lr 0.000885 time 0.2764 (0.2945) loss 2.8601 (3.8494) grad_norm 1.1634 (1.1435) [2021-04-15 20:31:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][600/1251] eta 0:03:11 lr 0.000885 time 0.2776 (0.2943) loss 3.8761 (3.8470) grad_norm 1.1321 (1.1435) [2021-04-15 20:31:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][610/1251] eta 0:03:08 lr 0.000885 time 0.2622 (0.2941) loss 3.6857 (3.8509) grad_norm 0.9567 (1.1431) [2021-04-15 20:31:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][620/1251] eta 0:03:05 lr 0.000885 time 0.2831 (0.2939) loss 3.8155 (3.8526) grad_norm 1.1292 (1.1439) [2021-04-15 20:31:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][630/1251] eta 0:03:02 lr 0.000885 time 0.2799 (0.2937) loss 2.6005 (3.8481) grad_norm 1.2624 (1.1437) [2021-04-15 20:31:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][640/1251] eta 0:02:59 lr 0.000885 time 0.2739 (0.2937) loss 4.0554 (3.8483) grad_norm 1.0126 (1.1443) [2021-04-15 20:31:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][650/1251] eta 0:02:56 lr 0.000885 time 0.2775 (0.2935) loss 3.6441 (3.8492) grad_norm 1.2584 (1.1447) [2021-04-15 20:31:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][660/1251] eta 0:02:53 lr 0.000885 time 0.2578 (0.2934) loss 4.2652 (3.8502) grad_norm 1.4975 (1.1455) [2021-04-15 20:31:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][670/1251] eta 0:02:50 lr 0.000885 time 0.2768 (0.2932) loss 3.8350 (3.8501) grad_norm 1.1035 (1.1456) [2021-04-15 20:31:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][680/1251] eta 0:02:47 lr 0.000885 time 0.2692 (0.2931) loss 4.5925 (3.8501) grad_norm 1.1887 (1.1454) [2021-04-15 20:31:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][690/1251] eta 0:02:44 lr 0.000885 time 0.2787 (0.2931) loss 3.7969 (3.8496) grad_norm 1.2050 (1.1456) [2021-04-15 20:31:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][700/1251] eta 0:02:41 lr 0.000885 time 0.2838 (0.2929) loss 2.9564 (3.8478) grad_norm 1.0782 (1.1449) [2021-04-15 20:31:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][710/1251] eta 0:02:38 lr 0.000885 time 0.2836 (0.2929) loss 3.3138 (3.8467) grad_norm 1.1272 (1.1443) [2021-04-15 20:32:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][720/1251] eta 0:02:35 lr 0.000884 time 0.2878 (0.2927) loss 3.8899 (3.8478) grad_norm 0.9546 (1.1453) [2021-04-15 20:32:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][730/1251] eta 0:02:32 lr 0.000884 time 0.2749 (0.2926) loss 3.8942 (3.8437) grad_norm 1.1575 (1.1453) [2021-04-15 20:32:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][740/1251] eta 0:02:29 lr 0.000884 time 0.2622 (0.2925) loss 4.2940 (3.8427) grad_norm 1.0459 (1.1442) [2021-04-15 20:32:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][750/1251] eta 0:02:26 lr 0.000884 time 0.2701 (0.2924) loss 4.1724 (3.8434) grad_norm 1.1039 (1.1436) [2021-04-15 20:32:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][760/1251] eta 0:02:23 lr 0.000884 time 0.2939 (0.2923) loss 2.6252 (3.8433) grad_norm 1.3046 (1.1449) [2021-04-15 20:32:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][770/1251] eta 0:02:20 lr 0.000884 time 0.2724 (0.2921) loss 4.4524 (3.8416) grad_norm 1.2333 (1.1460) [2021-04-15 20:32:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][780/1251] eta 0:02:17 lr 0.000884 time 0.2754 (0.2920) loss 3.5672 (3.8472) grad_norm 1.0510 (1.1462) [2021-04-15 20:32:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][790/1251] eta 0:02:14 lr 0.000884 time 0.2962 (0.2919) loss 3.8254 (3.8434) grad_norm 1.0637 (1.1455) [2021-04-15 20:32:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][800/1251] eta 0:02:11 lr 0.000884 time 0.2804 (0.2919) loss 3.7239 (3.8477) grad_norm 0.9834 (1.1452) [2021-04-15 20:32:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][810/1251] eta 0:02:08 lr 0.000884 time 0.2659 (0.2919) loss 4.0144 (3.8526) grad_norm 1.1986 (1.1450) [2021-04-15 20:32:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][820/1251] eta 0:02:05 lr 0.000884 time 0.2668 (0.2917) loss 2.8121 (3.8520) grad_norm 1.0446 (1.1453) [2021-04-15 20:32:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][830/1251] eta 0:02:02 lr 0.000884 time 0.3115 (0.2916) loss 3.1514 (3.8526) grad_norm 1.1587 (1.1464) [2021-04-15 20:32:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][840/1251] eta 0:01:59 lr 0.000884 time 0.3079 (0.2915) loss 3.8875 (3.8509) grad_norm 1.0384 (1.1460) [2021-04-15 20:32:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][850/1251] eta 0:01:56 lr 0.000884 time 0.2678 (0.2913) loss 4.1113 (3.8492) grad_norm 1.2531 (1.1467) [2021-04-15 20:32:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][860/1251] eta 0:01:53 lr 0.000884 time 0.2623 (0.2911) loss 2.7843 (3.8484) grad_norm 1.1034 (1.1464) [2021-04-15 20:32:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][870/1251] eta 0:01:50 lr 0.000884 time 0.2900 (0.2910) loss 3.8091 (3.8442) grad_norm 1.0904 (1.1472) [2021-04-15 20:32:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][880/1251] eta 0:01:47 lr 0.000884 time 0.3209 (0.2910) loss 4.1402 (3.8439) grad_norm 1.1617 (1.1466) [2021-04-15 20:32:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][890/1251] eta 0:01:45 lr 0.000884 time 0.2657 (0.2909) loss 4.4364 (3.8448) grad_norm 1.0791 (1.1469) [2021-04-15 20:32:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][900/1251] eta 0:01:42 lr 0.000884 time 0.2719 (0.2909) loss 3.7216 (3.8437) grad_norm 1.0951 (1.1476) [2021-04-15 20:32:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][910/1251] eta 0:01:39 lr 0.000884 time 0.3075 (0.2908) loss 4.7110 (3.8453) grad_norm 1.3067 (1.1478) [2021-04-15 20:32:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][920/1251] eta 0:01:36 lr 0.000884 time 0.2716 (0.2907) loss 2.9248 (3.8424) grad_norm 1.1391 (1.1476) [2021-04-15 20:32:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][930/1251] eta 0:01:33 lr 0.000884 time 0.2789 (0.2908) loss 4.1435 (3.8428) grad_norm 1.1289 (1.1479) [2021-04-15 20:33:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][940/1251] eta 0:01:30 lr 0.000884 time 0.4026 (0.2908) loss 4.5590 (3.8432) grad_norm 1.3464 (1.1484) [2021-04-15 20:33:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][950/1251] eta 0:01:27 lr 0.000884 time 0.2916 (0.2906) loss 3.0336 (3.8447) grad_norm 0.9985 (1.1484) [2021-04-15 20:33:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][960/1251] eta 0:01:24 lr 0.000884 time 0.2720 (0.2905) loss 3.7733 (3.8433) grad_norm 1.0498 (1.1483) [2021-04-15 20:33:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][970/1251] eta 0:01:21 lr 0.000884 time 0.3068 (0.2906) loss 4.6419 (3.8417) grad_norm 1.2085 (1.1479) [2021-04-15 20:33:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][980/1251] eta 0:01:18 lr 0.000884 time 0.2794 (0.2905) loss 3.2898 (3.8405) grad_norm 1.0820 (1.1477) [2021-04-15 20:33:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][990/1251] eta 0:01:15 lr 0.000884 time 0.2636 (0.2904) loss 4.2810 (3.8420) grad_norm 1.0371 (1.1489) [2021-04-15 20:33:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][1000/1251] eta 0:01:12 lr 0.000884 time 0.2702 (0.2903) loss 3.2265 (3.8390) grad_norm 1.1651 (1.1497) [2021-04-15 20:33:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][1010/1251] eta 0:01:09 lr 0.000884 time 0.2839 (0.2904) loss 2.6992 (3.8392) grad_norm 1.0115 (1.1503) [2021-04-15 20:33:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][1020/1251] eta 0:01:07 lr 0.000884 time 0.3095 (0.2903) loss 3.9752 (3.8386) grad_norm 1.0926 (1.1499) [2021-04-15 20:33:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][1030/1251] eta 0:01:04 lr 0.000884 time 0.3090 (0.2902) loss 3.2079 (3.8357) grad_norm 1.0802 (1.1498) [2021-04-15 20:33:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][1040/1251] eta 0:01:01 lr 0.000884 time 0.2839 (0.2902) loss 3.3806 (3.8354) grad_norm 1.0889 (1.1499) [2021-04-15 20:33:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][1050/1251] eta 0:00:58 lr 0.000884 time 0.2871 (0.2901) loss 4.3896 (3.8362) grad_norm 1.2448 (1.1498) [2021-04-15 20:33:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][1060/1251] eta 0:00:55 lr 0.000884 time 0.2724 (0.2900) loss 3.6797 (3.8371) grad_norm 1.1390 (1.1499) [2021-04-15 20:33:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][1070/1251] eta 0:00:52 lr 0.000884 time 0.2994 (0.2899) loss 3.4958 (3.8377) grad_norm 1.3282 (1.1498) [2021-04-15 20:33:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][1080/1251] eta 0:00:49 lr 0.000884 time 0.2698 (0.2898) loss 4.3907 (3.8389) grad_norm 1.0900 (1.1497) [2021-04-15 20:33:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][1090/1251] eta 0:00:46 lr 0.000884 time 0.2959 (0.2898) loss 3.5000 (3.8395) grad_norm 0.9758 (1.1492) [2021-04-15 20:33:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][1100/1251] eta 0:00:43 lr 0.000883 time 0.2807 (0.2897) loss 3.7227 (3.8409) grad_norm 1.3407 (1.1494) [2021-04-15 20:33:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][1110/1251] eta 0:00:40 lr 0.000883 time 0.2833 (0.2896) loss 4.7668 (3.8400) grad_norm 0.9669 (1.1495) [2021-04-15 20:33:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][1120/1251] eta 0:00:37 lr 0.000883 time 0.3069 (0.2895) loss 3.4640 (3.8415) grad_norm 1.1330 (1.1490) [2021-04-15 20:33:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][1130/1251] eta 0:00:35 lr 0.000883 time 0.2787 (0.2894) loss 4.3627 (3.8437) grad_norm 1.0703 (1.1494) [2021-04-15 20:33:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][1140/1251] eta 0:00:32 lr 0.000883 time 0.2446 (0.2895) loss 4.0478 (3.8426) grad_norm 1.1742 (1.1499) [2021-04-15 20:34:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][1150/1251] eta 0:00:29 lr 0.000883 time 0.3067 (0.2896) loss 3.8375 (3.8435) grad_norm 1.0508 (1.1501) [2021-04-15 20:34:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][1160/1251] eta 0:00:26 lr 0.000883 time 0.2838 (0.2895) loss 3.0097 (3.8416) grad_norm 1.3924 (1.1504) [2021-04-15 20:34:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][1170/1251] eta 0:00:23 lr 0.000883 time 0.2867 (0.2895) loss 3.0993 (3.8396) grad_norm 1.1872 (1.1501) [2021-04-15 20:34:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][1180/1251] eta 0:00:20 lr 0.000883 time 0.2719 (0.2894) loss 4.3049 (3.8422) grad_norm 0.9952 (1.1499) [2021-04-15 20:34:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][1190/1251] eta 0:00:17 lr 0.000883 time 0.2878 (0.2894) loss 4.0205 (3.8413) grad_norm 1.1298 (1.1494) [2021-04-15 20:34:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][1200/1251] eta 0:00:14 lr 0.000883 time 0.2667 (0.2893) loss 3.4170 (3.8407) grad_norm 1.1486 (1.1489) [2021-04-15 20:34:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][1210/1251] eta 0:00:11 lr 0.000883 time 0.2836 (0.2892) loss 3.3741 (3.8404) grad_norm 1.1512 (1.1483) [2021-04-15 20:34:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][1220/1251] eta 0:00:08 lr 0.000883 time 0.2992 (0.2893) loss 4.2446 (3.8406) grad_norm 1.0206 (1.1476) [2021-04-15 20:34:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][1230/1251] eta 0:00:06 lr 0.000883 time 0.2778 (0.2892) loss 3.2905 (3.8418) grad_norm 1.1802 (1.1476) [2021-04-15 20:34:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][1240/1251] eta 0:00:03 lr 0.000883 time 0.2485 (0.2890) loss 4.3007 (3.8406) grad_norm 1.3109 (1.1483) [2021-04-15 20:34:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [66/300][1250/1251] eta 0:00:00 lr 0.000883 time 0.2487 (0.2887) loss 4.1108 (3.8416) grad_norm 1.1549 (1.1489) [2021-04-15 20:34:33 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 66 training takes 0:06:04 [2021-04-15 20:34:33 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_66.pth saving...... [2021-04-15 20:34:45 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_66.pth saved !!! [2021-04-15 20:34:46 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.124 (1.124) Loss 1.3283 (1.3283) Acc@1 68.750 (68.750) Acc@5 89.941 (89.941) [2021-04-15 20:34:48 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.119 (0.255) Loss 1.1695 (1.2435) Acc@1 72.168 (71.076) Acc@5 92.188 (90.874) [2021-04-15 20:34:51 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.298 (0.264) Loss 1.1991 (1.2412) Acc@1 73.242 (71.247) Acc@5 91.113 (90.895) [2021-04-15 20:34:52 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.131 (0.218) Loss 1.2313 (1.2355) Acc@1 70.215 (71.264) Acc@5 91.309 (90.962) [2021-04-15 20:34:54 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.342 (0.221) Loss 1.2480 (1.2313) Acc@1 71.582 (71.499) Acc@5 90.234 (91.037) [2021-04-15 20:34:57 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 71.476 Acc@5 91.034 [2021-04-15 20:34:57 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 71.5% [2021-04-15 20:34:57 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 71.59% [2021-04-15 20:35:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][0/1251] eta 1:46:26 lr 0.000883 time 5.1054 (5.1054) loss 3.9741 (3.9741) grad_norm 1.0310 (1.0310) [2021-04-15 20:35:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][10/1251] eta 0:14:46 lr 0.000883 time 0.2616 (0.7142) loss 4.3809 (4.0237) grad_norm 1.2319 (1.0919) [2021-04-15 20:35:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][20/1251] eta 0:10:27 lr 0.000883 time 0.2842 (0.5096) loss 3.9960 (3.9463) grad_norm 1.1464 (1.1977) [2021-04-15 20:35:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][30/1251] eta 0:08:52 lr 0.000883 time 0.2728 (0.4359) loss 4.0053 (3.9009) grad_norm 1.1845 (1.1806) [2021-04-15 20:35:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][40/1251] eta 0:08:09 lr 0.000883 time 0.4413 (0.4040) loss 3.7169 (3.9524) grad_norm 1.3123 (1.1847) [2021-04-15 20:35:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][50/1251] eta 0:07:38 lr 0.000883 time 0.2946 (0.3816) loss 4.4966 (3.8974) grad_norm 1.2470 (1.1956) [2021-04-15 20:35:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][60/1251] eta 0:07:15 lr 0.000883 time 0.2958 (0.3656) loss 4.4221 (3.9477) grad_norm 1.1058 (1.1816) [2021-04-15 20:35:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][70/1251] eta 0:06:57 lr 0.000883 time 0.2598 (0.3532) loss 2.8118 (3.9228) grad_norm 1.2999 (1.1820) [2021-04-15 20:35:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][80/1251] eta 0:06:43 lr 0.000883 time 0.3009 (0.3444) loss 2.9831 (3.8801) grad_norm 1.0244 (1.1762) [2021-04-15 20:35:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][90/1251] eta 0:06:31 lr 0.000883 time 0.2741 (0.3372) loss 3.4260 (3.8754) grad_norm 1.0044 (1.1706) [2021-04-15 20:35:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][100/1251] eta 0:06:21 lr 0.000883 time 0.2866 (0.3314) loss 4.3268 (3.8439) grad_norm 1.1164 (1.1724) [2021-04-15 20:35:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][110/1251] eta 0:06:13 lr 0.000883 time 0.2516 (0.3271) loss 4.3877 (3.8621) grad_norm 1.0584 (1.1630) [2021-04-15 20:35:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][120/1251] eta 0:06:08 lr 0.000883 time 0.2731 (0.3254) loss 3.7855 (3.8710) grad_norm 1.2280 (1.1636) [2021-04-15 20:35:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][130/1251] eta 0:06:02 lr 0.000883 time 0.2969 (0.3233) loss 4.5192 (3.8856) grad_norm 1.0885 (1.1594) [2021-04-15 20:35:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][140/1251] eta 0:05:56 lr 0.000883 time 0.2954 (0.3205) loss 3.6982 (3.8995) grad_norm 0.9839 (1.1561) [2021-04-15 20:35:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][150/1251] eta 0:05:51 lr 0.000883 time 0.5125 (0.3197) loss 4.4841 (3.9126) grad_norm 1.1281 (1.1558) [2021-04-15 20:35:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][160/1251] eta 0:05:45 lr 0.000883 time 0.2857 (0.3171) loss 4.4612 (3.9241) grad_norm 1.0186 (1.1534) [2021-04-15 20:35:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][170/1251] eta 0:05:40 lr 0.000883 time 0.3086 (0.3154) loss 3.9229 (3.9016) grad_norm 1.1875 (1.1523) [2021-04-15 20:35:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][180/1251] eta 0:05:37 lr 0.000883 time 0.2913 (0.3147) loss 4.2924 (3.9079) grad_norm 1.1683 (1.1496) [2021-04-15 20:35:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][190/1251] eta 0:05:32 lr 0.000883 time 0.2918 (0.3131) loss 4.5929 (3.8951) grad_norm 0.9798 (1.1456) [2021-04-15 20:36:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][200/1251] eta 0:05:27 lr 0.000883 time 0.2623 (0.3118) loss 3.7594 (3.8958) grad_norm 1.0190 (1.1465) [2021-04-15 20:36:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][210/1251] eta 0:05:23 lr 0.000883 time 0.2723 (0.3106) loss 3.2168 (3.8938) grad_norm 0.9305 (1.1480) [2021-04-15 20:36:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][220/1251] eta 0:05:19 lr 0.000882 time 0.2861 (0.3099) loss 3.5224 (3.9017) grad_norm 1.1346 (1.1453) [2021-04-15 20:36:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][230/1251] eta 0:05:15 lr 0.000882 time 0.2918 (0.3088) loss 4.5374 (3.9068) grad_norm 1.1313 (1.1427) [2021-04-15 20:36:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][240/1251] eta 0:05:10 lr 0.000882 time 0.2805 (0.3074) loss 4.3444 (3.8962) grad_norm 1.1301 (1.1387) [2021-04-15 20:36:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][250/1251] eta 0:05:06 lr 0.000882 time 0.2804 (0.3064) loss 4.0465 (3.8817) grad_norm 1.1361 (1.1377) [2021-04-15 20:36:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][260/1251] eta 0:05:02 lr 0.000882 time 0.2913 (0.3056) loss 4.0618 (3.8772) grad_norm 1.2368 (1.1402) [2021-04-15 20:36:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][270/1251] eta 0:04:59 lr 0.000882 time 0.2852 (0.3049) loss 3.0044 (3.8798) grad_norm 1.4297 (1.1409) [2021-04-15 20:36:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][280/1251] eta 0:04:55 lr 0.000882 time 0.3198 (0.3043) loss 4.0612 (3.8771) grad_norm 1.1407 (1.1390) [2021-04-15 20:36:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][290/1251] eta 0:04:51 lr 0.000882 time 0.3070 (0.3033) loss 4.4733 (3.8635) grad_norm 1.0549 (1.1374) [2021-04-15 20:36:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][300/1251] eta 0:04:47 lr 0.000882 time 0.2834 (0.3027) loss 4.5868 (3.8659) grad_norm 1.0371 (1.1374) [2021-04-15 20:36:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][310/1251] eta 0:04:44 lr 0.000882 time 0.2777 (0.3019) loss 4.5305 (3.8639) grad_norm 1.1448 (1.1374) [2021-04-15 20:36:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][320/1251] eta 0:04:40 lr 0.000882 time 0.3067 (0.3014) loss 5.1299 (3.8622) grad_norm 1.2245 (1.1374) [2021-04-15 20:36:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][330/1251] eta 0:04:37 lr 0.000882 time 0.2725 (0.3010) loss 4.0171 (3.8617) grad_norm 1.1244 (1.1367) [2021-04-15 20:36:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][340/1251] eta 0:04:33 lr 0.000882 time 0.2671 (0.3004) loss 4.6089 (3.8624) grad_norm 1.0702 (1.1378) [2021-04-15 20:36:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][350/1251] eta 0:04:30 lr 0.000882 time 0.3176 (0.3006) loss 4.7891 (3.8661) grad_norm 1.1616 (1.1374) [2021-04-15 20:36:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][360/1251] eta 0:04:27 lr 0.000882 time 0.2665 (0.3003) loss 4.0180 (3.8644) grad_norm 1.0117 (1.1349) [2021-04-15 20:36:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][370/1251] eta 0:04:24 lr 0.000882 time 0.3134 (0.3000) loss 3.5489 (3.8676) grad_norm 1.1489 (1.1357) [2021-04-15 20:36:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][380/1251] eta 0:04:20 lr 0.000882 time 0.2946 (0.2995) loss 2.9532 (3.8702) grad_norm 1.6891 (1.1369) [2021-04-15 20:36:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][390/1251] eta 0:04:17 lr 0.000882 time 0.2943 (0.2991) loss 3.2527 (3.8757) grad_norm 1.4011 (1.1362) [2021-04-15 20:36:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][400/1251] eta 0:04:14 lr 0.000882 time 0.2722 (0.2986) loss 4.1126 (3.8806) grad_norm 1.0109 (1.1354) [2021-04-15 20:37:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][410/1251] eta 0:04:10 lr 0.000882 time 0.2973 (0.2984) loss 4.0395 (3.8835) grad_norm 1.0426 (1.1353) [2021-04-15 20:37:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][420/1251] eta 0:04:07 lr 0.000882 time 0.2972 (0.2983) loss 2.6459 (3.8791) grad_norm 1.1147 (1.1354) [2021-04-15 20:37:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][430/1251] eta 0:04:04 lr 0.000882 time 0.2822 (0.2980) loss 4.3453 (3.8805) grad_norm 1.0783 (1.1362) [2021-04-15 20:37:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][440/1251] eta 0:04:01 lr 0.000882 time 0.2873 (0.2977) loss 4.1630 (3.8824) grad_norm 1.0964 (1.1368) [2021-04-15 20:37:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][450/1251] eta 0:03:58 lr 0.000882 time 0.2812 (0.2973) loss 4.3564 (3.8839) grad_norm 1.0760 (1.1362) [2021-04-15 20:37:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][460/1251] eta 0:03:54 lr 0.000882 time 0.2928 (0.2970) loss 3.0708 (3.8860) grad_norm 1.0821 (1.1352) [2021-04-15 20:37:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][470/1251] eta 0:03:51 lr 0.000882 time 0.2687 (0.2967) loss 3.8995 (3.8856) grad_norm 0.9480 (1.1342) [2021-04-15 20:37:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][480/1251] eta 0:03:48 lr 0.000882 time 0.2766 (0.2963) loss 3.9854 (3.8886) grad_norm 1.0677 (inf) [2021-04-15 20:37:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][490/1251] eta 0:03:45 lr 0.000882 time 0.2875 (0.2963) loss 3.0416 (3.8909) grad_norm 1.1361 (inf) [2021-04-15 20:37:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][500/1251] eta 0:03:42 lr 0.000882 time 0.2576 (0.2962) loss 4.6592 (3.8912) grad_norm 1.3118 (inf) [2021-04-15 20:37:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][510/1251] eta 0:03:39 lr 0.000882 time 0.2625 (0.2960) loss 3.0004 (3.8903) grad_norm 0.9245 (inf) [2021-04-15 20:37:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][520/1251] eta 0:03:36 lr 0.000882 time 0.3013 (0.2959) loss 4.1145 (3.8936) grad_norm 1.1955 (inf) [2021-04-15 20:37:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][530/1251] eta 0:03:33 lr 0.000882 time 0.2664 (0.2956) loss 3.4981 (3.8889) grad_norm 1.3544 (inf) [2021-04-15 20:37:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][540/1251] eta 0:03:29 lr 0.000882 time 0.3048 (0.2953) loss 3.1461 (3.8840) grad_norm 1.4045 (inf) [2021-04-15 20:37:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][550/1251] eta 0:03:26 lr 0.000882 time 0.2677 (0.2950) loss 3.9876 (3.8813) grad_norm 1.2986 (inf) [2021-04-15 20:37:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][560/1251] eta 0:03:23 lr 0.000882 time 0.2645 (0.2948) loss 3.6068 (3.8840) grad_norm 1.1767 (inf) [2021-04-15 20:37:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][570/1251] eta 0:03:20 lr 0.000882 time 0.2724 (0.2949) loss 3.2355 (3.8790) grad_norm 1.0603 (inf) [2021-04-15 20:37:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][580/1251] eta 0:03:17 lr 0.000882 time 0.2971 (0.2946) loss 4.4033 (3.8792) grad_norm 1.1671 (inf) [2021-04-15 20:37:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][590/1251] eta 0:03:14 lr 0.000881 time 0.2929 (0.2944) loss 4.3953 (3.8778) grad_norm 0.9640 (inf) [2021-04-15 20:37:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][600/1251] eta 0:03:11 lr 0.000881 time 0.2553 (0.2942) loss 3.8270 (3.8733) grad_norm 1.2658 (inf) [2021-04-15 20:37:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][610/1251] eta 0:03:08 lr 0.000881 time 0.2959 (0.2940) loss 4.4052 (3.8708) grad_norm 1.0224 (inf) [2021-04-15 20:38:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][620/1251] eta 0:03:05 lr 0.000881 time 0.2687 (0.2941) loss 4.2577 (3.8732) grad_norm 1.0840 (inf) [2021-04-15 20:38:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][630/1251] eta 0:03:02 lr 0.000881 time 0.2904 (0.2939) loss 3.2186 (3.8729) grad_norm 1.0547 (inf) [2021-04-15 20:38:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][640/1251] eta 0:02:59 lr 0.000881 time 0.3062 (0.2938) loss 2.5545 (3.8705) grad_norm 1.2854 (inf) [2021-04-15 20:38:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][650/1251] eta 0:02:56 lr 0.000881 time 0.2852 (0.2936) loss 4.3521 (3.8705) grad_norm 1.3637 (inf) [2021-04-15 20:38:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][660/1251] eta 0:02:53 lr 0.000881 time 0.2722 (0.2934) loss 3.5630 (3.8700) grad_norm 1.0643 (inf) [2021-04-15 20:38:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][670/1251] eta 0:02:50 lr 0.000881 time 0.3083 (0.2933) loss 4.1857 (3.8690) grad_norm 1.3074 (inf) [2021-04-15 20:38:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][680/1251] eta 0:02:47 lr 0.000881 time 0.2970 (0.2932) loss 3.1443 (3.8647) grad_norm 1.2421 (inf) [2021-04-15 20:38:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][690/1251] eta 0:02:44 lr 0.000881 time 0.2683 (0.2932) loss 3.9238 (3.8646) grad_norm 1.1226 (inf) [2021-04-15 20:38:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][700/1251] eta 0:02:41 lr 0.000881 time 0.2605 (0.2931) loss 4.2657 (3.8642) grad_norm 1.0776 (inf) [2021-04-15 20:38:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][710/1251] eta 0:02:38 lr 0.000881 time 0.2750 (0.2929) loss 3.0776 (3.8660) grad_norm 1.3207 (inf) [2021-04-15 20:38:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][720/1251] eta 0:02:35 lr 0.000881 time 0.2680 (0.2928) loss 4.3486 (3.8684) grad_norm 1.2434 (inf) [2021-04-15 20:38:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][730/1251] eta 0:02:32 lr 0.000881 time 0.2744 (0.2928) loss 4.3709 (3.8619) grad_norm 1.0619 (inf) [2021-04-15 20:38:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][740/1251] eta 0:02:29 lr 0.000881 time 0.2810 (0.2926) loss 3.4495 (3.8646) grad_norm 1.1265 (inf) [2021-04-15 20:38:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][750/1251] eta 0:02:26 lr 0.000881 time 0.2858 (0.2925) loss 2.9157 (3.8651) grad_norm 1.4634 (inf) [2021-04-15 20:38:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][760/1251] eta 0:02:23 lr 0.000881 time 0.2972 (0.2927) loss 2.7721 (3.8640) grad_norm 0.9884 (inf) [2021-04-15 20:38:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][770/1251] eta 0:02:20 lr 0.000881 time 0.2883 (0.2926) loss 3.3732 (3.8590) grad_norm 1.3285 (inf) [2021-04-15 20:38:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][780/1251] eta 0:02:17 lr 0.000881 time 0.2686 (0.2924) loss 3.0722 (3.8542) grad_norm 1.0603 (inf) [2021-04-15 20:38:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][790/1251] eta 0:02:14 lr 0.000881 time 0.2598 (0.2923) loss 4.7462 (3.8536) grad_norm 1.2075 (inf) [2021-04-15 20:38:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][800/1251] eta 0:02:11 lr 0.000881 time 0.3119 (0.2921) loss 4.4559 (3.8557) grad_norm 1.0383 (inf) [2021-04-15 20:38:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][810/1251] eta 0:02:08 lr 0.000881 time 0.2691 (0.2919) loss 3.3314 (3.8573) grad_norm 1.2520 (inf) [2021-04-15 20:38:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][820/1251] eta 0:02:05 lr 0.000881 time 0.2898 (0.2919) loss 3.9762 (3.8580) grad_norm 0.9265 (inf) [2021-04-15 20:39:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][830/1251] eta 0:02:02 lr 0.000881 time 0.2880 (0.2918) loss 4.1000 (3.8609) grad_norm 1.4444 (inf) [2021-04-15 20:39:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][840/1251] eta 0:01:59 lr 0.000881 time 0.2678 (0.2917) loss 2.9985 (3.8590) grad_norm 1.0539 (inf) [2021-04-15 20:39:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][850/1251] eta 0:01:56 lr 0.000881 time 0.2972 (0.2916) loss 4.9192 (3.8607) grad_norm 1.3928 (inf) [2021-04-15 20:39:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][860/1251] eta 0:01:53 lr 0.000881 time 0.3006 (0.2915) loss 4.7139 (3.8644) grad_norm 0.9905 (inf) [2021-04-15 20:39:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][870/1251] eta 0:01:50 lr 0.000881 time 0.2787 (0.2913) loss 4.2049 (3.8653) grad_norm 1.0112 (inf) [2021-04-15 20:39:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][880/1251] eta 0:01:48 lr 0.000881 time 0.2686 (0.2912) loss 2.9958 (3.8640) grad_norm 0.9785 (inf) [2021-04-15 20:39:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][890/1251] eta 0:01:45 lr 0.000881 time 0.2602 (0.2911) loss 4.3034 (3.8659) grad_norm 1.2324 (inf) [2021-04-15 20:39:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][900/1251] eta 0:01:42 lr 0.000881 time 0.2898 (0.2910) loss 4.1903 (3.8646) grad_norm 1.0910 (inf) [2021-04-15 20:39:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][910/1251] eta 0:01:39 lr 0.000881 time 0.2650 (0.2909) loss 3.4777 (3.8637) grad_norm 1.1782 (inf) [2021-04-15 20:39:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][920/1251] eta 0:01:36 lr 0.000881 time 0.2839 (0.2908) loss 4.3909 (3.8625) grad_norm 1.4890 (inf) [2021-04-15 20:39:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][930/1251] eta 0:01:33 lr 0.000881 time 0.2840 (0.2910) loss 3.4395 (3.8591) grad_norm 1.0303 (inf) [2021-04-15 20:39:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][940/1251] eta 0:01:30 lr 0.000881 time 0.2746 (0.2910) loss 3.9928 (3.8627) grad_norm 1.0538 (inf) [2021-04-15 20:39:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][950/1251] eta 0:01:27 lr 0.000881 time 0.2481 (0.2909) loss 4.3531 (3.8632) grad_norm 1.0279 (inf) [2021-04-15 20:39:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][960/1251] eta 0:01:24 lr 0.000880 time 0.2835 (0.2908) loss 3.6889 (3.8636) grad_norm 1.1833 (inf) [2021-04-15 20:39:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][970/1251] eta 0:01:21 lr 0.000880 time 0.2764 (0.2907) loss 4.3171 (3.8622) grad_norm 1.0722 (inf) [2021-04-15 20:39:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][980/1251] eta 0:01:18 lr 0.000880 time 0.3008 (0.2906) loss 3.0480 (3.8628) grad_norm 1.2878 (inf) [2021-04-15 20:39:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][990/1251] eta 0:01:15 lr 0.000880 time 0.2740 (0.2906) loss 4.5029 (3.8631) grad_norm 0.9819 (inf) [2021-04-15 20:39:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][1000/1251] eta 0:01:12 lr 0.000880 time 0.2803 (0.2905) loss 3.9096 (3.8640) grad_norm 1.3418 (inf) [2021-04-15 20:39:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][1010/1251] eta 0:01:09 lr 0.000880 time 0.2912 (0.2904) loss 4.2856 (3.8626) grad_norm 1.2244 (inf) [2021-04-15 20:39:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][1020/1251] eta 0:01:07 lr 0.000880 time 0.2674 (0.2903) loss 4.0828 (3.8624) grad_norm 1.8990 (inf) [2021-04-15 20:39:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][1030/1251] eta 0:01:04 lr 0.000880 time 0.2611 (0.2902) loss 4.4878 (3.8648) grad_norm 1.2780 (inf) [2021-04-15 20:39:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][1040/1251] eta 0:01:01 lr 0.000880 time 0.2785 (0.2902) loss 3.8041 (3.8651) grad_norm 1.0713 (inf) [2021-04-15 20:40:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][1050/1251] eta 0:00:58 lr 0.000880 time 0.2650 (0.2901) loss 2.6909 (3.8670) grad_norm 1.2579 (inf) [2021-04-15 20:40:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][1060/1251] eta 0:00:55 lr 0.000880 time 0.2686 (0.2900) loss 3.5415 (3.8682) grad_norm 1.2621 (inf) [2021-04-15 20:40:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][1070/1251] eta 0:00:52 lr 0.000880 time 0.2962 (0.2900) loss 3.1728 (3.8651) grad_norm 0.9280 (inf) [2021-04-15 20:40:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][1080/1251] eta 0:00:49 lr 0.000880 time 0.3091 (0.2899) loss 3.4652 (3.8661) grad_norm 1.4016 (inf) [2021-04-15 20:40:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][1090/1251] eta 0:00:46 lr 0.000880 time 0.2777 (0.2899) loss 3.3872 (3.8662) grad_norm 1.1005 (inf) [2021-04-15 20:40:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][1100/1251] eta 0:00:43 lr 0.000880 time 0.3048 (0.2898) loss 4.2266 (3.8669) grad_norm 1.2425 (inf) [2021-04-15 20:40:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][1110/1251] eta 0:00:40 lr 0.000880 time 0.2772 (0.2897) loss 3.3137 (3.8691) grad_norm 1.1227 (inf) [2021-04-15 20:40:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][1120/1251] eta 0:00:37 lr 0.000880 time 0.2930 (0.2896) loss 3.2914 (3.8683) grad_norm 1.1100 (inf) [2021-04-15 20:40:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][1130/1251] eta 0:00:35 lr 0.000880 time 0.2599 (0.2895) loss 2.6345 (3.8681) grad_norm 1.3123 (inf) [2021-04-15 20:40:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][1140/1251] eta 0:00:32 lr 0.000880 time 0.2786 (0.2895) loss 3.5985 (3.8672) grad_norm 1.2020 (inf) [2021-04-15 20:40:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][1150/1251] eta 0:00:29 lr 0.000880 time 0.2817 (0.2895) loss 3.9296 (3.8685) grad_norm 1.0006 (inf) [2021-04-15 20:40:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][1160/1251] eta 0:00:26 lr 0.000880 time 0.2832 (0.2895) loss 3.4039 (3.8665) grad_norm 1.1242 (inf) [2021-04-15 20:40:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][1170/1251] eta 0:00:23 lr 0.000880 time 0.2664 (0.2894) loss 2.7921 (3.8632) grad_norm 1.1620 (inf) [2021-04-15 20:40:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][1180/1251] eta 0:00:20 lr 0.000880 time 0.2690 (0.2893) loss 4.6251 (3.8664) grad_norm 1.1002 (inf) [2021-04-15 20:40:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][1190/1251] eta 0:00:17 lr 0.000880 time 0.2597 (0.2893) loss 4.3947 (3.8688) grad_norm 1.1614 (inf) [2021-04-15 20:40:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][1200/1251] eta 0:00:14 lr 0.000880 time 0.3296 (0.2894) loss 3.8478 (3.8678) grad_norm 0.9395 (inf) [2021-04-15 20:40:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][1210/1251] eta 0:00:11 lr 0.000880 time 0.2793 (0.2893) loss 4.3686 (3.8679) grad_norm 0.8989 (inf) [2021-04-15 20:40:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][1220/1251] eta 0:00:08 lr 0.000880 time 0.2673 (0.2893) loss 4.1273 (3.8672) grad_norm 1.1369 (inf) [2021-04-15 20:40:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][1230/1251] eta 0:00:06 lr 0.000880 time 0.2958 (0.2892) loss 3.5243 (3.8656) grad_norm 1.1016 (inf) [2021-04-15 20:40:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][1240/1251] eta 0:00:03 lr 0.000880 time 0.2486 (0.2890) loss 4.2013 (3.8662) grad_norm 0.9836 (inf) [2021-04-15 20:40:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [67/300][1250/1251] eta 0:00:00 lr 0.000880 time 0.2492 (0.2887) loss 4.4412 (3.8658) grad_norm 1.1354 (inf) [2021-04-15 20:41:01 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 67 training takes 0:06:04 [2021-04-15 20:41:01 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_67.pth saving...... [2021-04-15 20:41:13 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_67.pth saved !!! [2021-04-15 20:41:14 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.289 (1.289) Loss 1.3428 (1.3428) Acc@1 68.164 (68.164) Acc@5 89.551 (89.551) [2021-04-15 20:41:15 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.100 (0.215) Loss 1.2569 (1.2353) Acc@1 71.387 (71.227) Acc@5 90.527 (91.175) [2021-04-15 20:41:19 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.120 (0.289) Loss 1.2518 (1.2293) Acc@1 71.484 (71.517) Acc@5 90.625 (91.118) [2021-04-15 20:41:21 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.124 (0.254) Loss 1.2804 (1.2371) Acc@1 70.215 (71.390) Acc@5 90.918 (90.953) [2021-04-15 20:41:22 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.236) Loss 1.2754 (1.2315) Acc@1 72.070 (71.732) Acc@5 90.332 (91.049) [2021-04-15 20:41:26 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 71.672 Acc@5 91.074 [2021-04-15 20:41:26 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 71.7% [2021-04-15 20:41:26 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 71.67% [2021-04-15 20:41:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][0/1251] eta 2:23:17 lr 0.000880 time 6.8727 (6.8727) loss 4.1952 (4.1952) grad_norm 1.2016 (1.2016) [2021-04-15 20:41:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][10/1251] eta 0:18:18 lr 0.000880 time 0.3984 (0.8852) loss 2.8262 (3.4869) grad_norm 1.1627 (1.1040) [2021-04-15 20:41:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][20/1251] eta 0:12:13 lr 0.000880 time 0.2888 (0.5961) loss 2.6985 (3.4972) grad_norm 1.1748 (1.1508) [2021-04-15 20:41:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][30/1251] eta 0:10:06 lr 0.000880 time 0.2786 (0.4963) loss 3.9562 (3.6041) grad_norm 0.9500 (1.1612) [2021-04-15 20:41:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][40/1251] eta 0:08:58 lr 0.000880 time 0.3018 (0.4449) loss 4.1801 (3.6342) grad_norm 1.1599 (1.1657) [2021-04-15 20:41:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][50/1251] eta 0:08:17 lr 0.000880 time 0.2977 (0.4142) loss 4.4610 (3.7091) grad_norm 1.0535 (1.1634) [2021-04-15 20:41:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][60/1251] eta 0:07:48 lr 0.000880 time 0.3061 (0.3930) loss 4.0884 (3.7593) grad_norm 1.4624 (1.1720) [2021-04-15 20:41:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][70/1251] eta 0:07:25 lr 0.000880 time 0.3057 (0.3776) loss 4.9943 (3.8030) grad_norm 1.2051 (1.1785) [2021-04-15 20:41:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][80/1251] eta 0:07:08 lr 0.000879 time 0.2936 (0.3661) loss 4.5079 (3.7979) grad_norm 1.1479 (1.1728) [2021-04-15 20:41:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][90/1251] eta 0:06:54 lr 0.000879 time 0.2847 (0.3571) loss 3.8076 (3.8123) grad_norm 1.3705 (1.1785) [2021-04-15 20:42:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][100/1251] eta 0:06:43 lr 0.000879 time 0.3062 (0.3505) loss 3.4941 (3.8347) grad_norm 1.0545 (1.1731) [2021-04-15 20:42:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][110/1251] eta 0:06:33 lr 0.000879 time 0.2841 (0.3446) loss 3.2828 (3.8242) grad_norm 1.1990 (1.1730) [2021-04-15 20:42:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][120/1251] eta 0:06:24 lr 0.000879 time 0.2861 (0.3398) loss 2.7608 (3.8042) grad_norm 1.1994 (1.1691) [2021-04-15 20:42:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][130/1251] eta 0:06:17 lr 0.000879 time 0.2825 (0.3369) loss 4.5854 (3.8213) grad_norm 1.1627 (1.1722) [2021-04-15 20:42:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][140/1251] eta 0:06:09 lr 0.000879 time 0.2982 (0.3330) loss 2.3562 (3.8162) grad_norm 1.3557 (1.1674) [2021-04-15 20:42:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][150/1251] eta 0:06:03 lr 0.000879 time 0.2662 (0.3305) loss 4.1097 (3.7943) grad_norm 0.9980 (1.1664) [2021-04-15 20:42:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][160/1251] eta 0:05:57 lr 0.000879 time 0.3062 (0.3276) loss 3.7667 (3.8033) grad_norm 1.3040 (1.1657) [2021-04-15 20:42:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][170/1251] eta 0:05:51 lr 0.000879 time 0.2813 (0.3252) loss 4.1363 (3.7934) grad_norm 1.4057 (1.1691) [2021-04-15 20:42:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][180/1251] eta 0:05:46 lr 0.000879 time 0.2959 (0.3233) loss 4.3002 (3.7952) grad_norm 1.2537 (1.1667) [2021-04-15 20:42:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][190/1251] eta 0:05:40 lr 0.000879 time 0.2821 (0.3212) loss 4.1837 (3.7892) grad_norm 1.1406 (1.1663) [2021-04-15 20:42:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][200/1251] eta 0:05:36 lr 0.000879 time 0.4194 (0.3201) loss 4.7774 (3.7870) grad_norm 1.1140 (1.1674) [2021-04-15 20:42:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][210/1251] eta 0:05:31 lr 0.000879 time 0.2979 (0.3183) loss 3.9029 (3.7857) grad_norm 1.4223 (1.1697) [2021-04-15 20:42:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][220/1251] eta 0:05:26 lr 0.000879 time 0.2658 (0.3167) loss 3.8956 (3.8035) grad_norm 1.1683 (1.1659) [2021-04-15 20:42:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][230/1251] eta 0:05:21 lr 0.000879 time 0.2921 (0.3152) loss 3.9175 (3.8033) grad_norm 0.9824 (1.1686) [2021-04-15 20:42:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][240/1251] eta 0:05:17 lr 0.000879 time 0.2754 (0.3139) loss 3.9580 (3.8070) grad_norm 1.2270 (1.1692) [2021-04-15 20:42:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][250/1251] eta 0:05:13 lr 0.000879 time 0.2745 (0.3127) loss 4.4507 (3.8061) grad_norm 1.0326 (1.1698) [2021-04-15 20:42:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][260/1251] eta 0:05:08 lr 0.000879 time 0.3191 (0.3116) loss 3.5390 (3.8130) grad_norm 0.9989 (1.1692) [2021-04-15 20:42:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][270/1251] eta 0:05:04 lr 0.000879 time 0.2904 (0.3104) loss 3.7823 (3.8137) grad_norm 1.1990 (1.1689) [2021-04-15 20:42:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][280/1251] eta 0:05:00 lr 0.000879 time 0.2695 (0.3093) loss 3.5161 (3.8167) grad_norm 1.3159 (1.1651) [2021-04-15 20:42:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][290/1251] eta 0:04:56 lr 0.000879 time 0.2721 (0.3089) loss 3.3680 (3.8096) grad_norm 1.3545 (1.1653) [2021-04-15 20:42:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][300/1251] eta 0:04:53 lr 0.000879 time 0.2763 (0.3084) loss 2.6384 (3.8093) grad_norm 1.6274 (1.1692) [2021-04-15 20:43:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][310/1251] eta 0:04:49 lr 0.000879 time 0.2922 (0.3075) loss 4.3778 (3.8101) grad_norm 1.0927 (1.1702) [2021-04-15 20:43:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][320/1251] eta 0:04:45 lr 0.000879 time 0.2583 (0.3067) loss 3.3095 (3.7946) grad_norm 1.2188 (1.1697) [2021-04-15 20:43:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][330/1251] eta 0:04:42 lr 0.000879 time 0.2869 (0.3062) loss 4.0132 (3.8032) grad_norm 1.4047 (1.1709) [2021-04-15 20:43:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][340/1251] eta 0:04:38 lr 0.000879 time 0.2740 (0.3054) loss 4.0855 (3.7979) grad_norm 1.1101 (1.1711) [2021-04-15 20:43:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][350/1251] eta 0:04:35 lr 0.000879 time 0.2868 (0.3053) loss 3.3353 (3.7962) grad_norm 0.9763 (1.1674) [2021-04-15 20:43:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][360/1251] eta 0:04:32 lr 0.000879 time 0.2829 (0.3055) loss 4.3232 (3.8002) grad_norm 1.0707 (1.1647) [2021-04-15 20:43:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][370/1251] eta 0:04:28 lr 0.000879 time 0.2997 (0.3048) loss 4.1578 (3.8020) grad_norm 0.9537 (1.1622) [2021-04-15 20:43:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][380/1251] eta 0:04:24 lr 0.000879 time 0.2831 (0.3042) loss 4.5716 (3.8044) grad_norm 1.5541 (1.1622) [2021-04-15 20:43:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][390/1251] eta 0:04:21 lr 0.000879 time 0.3122 (0.3037) loss 4.3399 (3.8033) grad_norm 1.1569 (1.1624) [2021-04-15 20:43:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][400/1251] eta 0:04:17 lr 0.000879 time 0.2972 (0.3030) loss 4.1822 (3.8078) grad_norm 1.0927 (1.1638) [2021-04-15 20:43:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][410/1251] eta 0:04:14 lr 0.000879 time 0.2617 (0.3025) loss 4.0559 (3.8093) grad_norm 1.2861 (1.1637) [2021-04-15 20:43:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][420/1251] eta 0:04:10 lr 0.000879 time 0.2832 (0.3020) loss 3.8384 (3.8142) grad_norm 1.1527 (1.1631) [2021-04-15 20:43:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][430/1251] eta 0:04:07 lr 0.000879 time 0.2577 (0.3015) loss 3.4329 (3.8130) grad_norm 1.0670 (1.1640) [2021-04-15 20:43:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][440/1251] eta 0:04:04 lr 0.000879 time 0.2690 (0.3012) loss 3.9420 (3.8124) grad_norm 1.1603 (1.1632) [2021-04-15 20:43:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][450/1251] eta 0:04:00 lr 0.000878 time 0.2733 (0.3007) loss 3.6836 (3.8173) grad_norm 1.0648 (1.1634) [2021-04-15 20:43:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][460/1251] eta 0:03:57 lr 0.000878 time 0.2824 (0.3004) loss 3.9286 (3.8186) grad_norm 1.4025 (1.1645) [2021-04-15 20:43:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][470/1251] eta 0:03:54 lr 0.000878 time 0.2909 (0.3000) loss 3.8319 (3.8230) grad_norm 1.0836 (1.1649) [2021-04-15 20:43:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][480/1251] eta 0:03:51 lr 0.000878 time 0.2810 (0.2998) loss 4.3994 (3.8221) grad_norm 1.1129 (1.1635) [2021-04-15 20:43:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][490/1251] eta 0:03:48 lr 0.000878 time 0.2730 (0.2998) loss 3.1444 (3.8247) grad_norm 1.0516 (1.1631) [2021-04-15 20:43:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][500/1251] eta 0:03:44 lr 0.000878 time 0.2587 (0.2994) loss 4.1077 (3.8258) grad_norm 1.1240 (1.1644) [2021-04-15 20:43:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][510/1251] eta 0:03:41 lr 0.000878 time 0.2912 (0.2991) loss 4.0022 (3.8295) grad_norm 1.4843 (1.1654) [2021-04-15 20:44:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][520/1251] eta 0:03:38 lr 0.000878 time 0.2731 (0.2988) loss 4.4185 (3.8319) grad_norm 1.2455 (1.1657) [2021-04-15 20:44:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][530/1251] eta 0:03:35 lr 0.000878 time 0.2928 (0.2986) loss 4.5482 (3.8374) grad_norm 1.2560 (1.1632) [2021-04-15 20:44:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][540/1251] eta 0:03:32 lr 0.000878 time 0.2937 (0.2982) loss 3.4821 (3.8376) grad_norm 1.1150 (1.1627) [2021-04-15 20:44:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][550/1251] eta 0:03:29 lr 0.000878 time 0.2636 (0.2982) loss 4.8022 (3.8459) grad_norm 1.1918 (1.1625) [2021-04-15 20:44:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][560/1251] eta 0:03:25 lr 0.000878 time 0.2641 (0.2979) loss 3.6527 (3.8465) grad_norm 1.0637 (1.1633) [2021-04-15 20:44:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][570/1251] eta 0:03:22 lr 0.000878 time 0.3036 (0.2980) loss 4.3566 (3.8438) grad_norm 0.9994 (1.1631) [2021-04-15 20:44:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][580/1251] eta 0:03:19 lr 0.000878 time 0.3100 (0.2978) loss 4.3801 (3.8442) grad_norm 1.2011 (1.1646) [2021-04-15 20:44:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][590/1251] eta 0:03:16 lr 0.000878 time 0.2672 (0.2978) loss 3.8311 (3.8437) grad_norm 1.0539 (1.1633) [2021-04-15 20:44:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][600/1251] eta 0:03:13 lr 0.000878 time 0.2794 (0.2975) loss 4.1347 (3.8402) grad_norm 1.1532 (1.1623) [2021-04-15 20:44:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][610/1251] eta 0:03:10 lr 0.000878 time 0.2928 (0.2972) loss 3.9135 (3.8394) grad_norm 1.2428 (1.1622) [2021-04-15 20:44:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][620/1251] eta 0:03:07 lr 0.000878 time 0.2859 (0.2970) loss 4.3658 (3.8415) grad_norm 1.0953 (1.1621) [2021-04-15 20:44:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][630/1251] eta 0:03:04 lr 0.000878 time 0.2584 (0.2967) loss 4.5063 (3.8426) grad_norm 0.9881 (1.1613) [2021-04-15 20:44:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][640/1251] eta 0:03:01 lr 0.000878 time 0.2827 (0.2965) loss 2.7602 (3.8426) grad_norm 1.2038 (1.1605) [2021-04-15 20:44:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][650/1251] eta 0:02:58 lr 0.000878 time 0.2659 (0.2962) loss 4.7674 (3.8455) grad_norm 1.1693 (1.1609) [2021-04-15 20:44:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][660/1251] eta 0:02:54 lr 0.000878 time 0.2453 (0.2960) loss 2.4942 (3.8399) grad_norm 1.0300 (1.1604) [2021-04-15 20:44:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][670/1251] eta 0:02:51 lr 0.000878 time 0.2892 (0.2959) loss 4.2999 (3.8434) grad_norm 1.1056 (1.1603) [2021-04-15 20:44:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][680/1251] eta 0:02:48 lr 0.000878 time 0.2644 (0.2956) loss 3.5201 (3.8426) grad_norm 1.2509 (1.1601) [2021-04-15 20:44:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][690/1251] eta 0:02:45 lr 0.000878 time 0.3148 (0.2955) loss 3.3519 (3.8430) grad_norm 1.1039 (1.1603) [2021-04-15 20:44:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][700/1251] eta 0:02:42 lr 0.000878 time 0.2665 (0.2953) loss 3.1339 (3.8439) grad_norm 1.1312 (1.1606) [2021-04-15 20:44:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][710/1251] eta 0:02:39 lr 0.000878 time 0.2894 (0.2953) loss 3.9789 (3.8428) grad_norm 1.1466 (1.1598) [2021-04-15 20:44:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][720/1251] eta 0:02:36 lr 0.000878 time 0.3035 (0.2954) loss 2.6211 (3.8428) grad_norm 1.2704 (1.1612) [2021-04-15 20:45:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][730/1251] eta 0:02:33 lr 0.000878 time 0.2784 (0.2952) loss 3.6424 (3.8390) grad_norm 1.0029 (1.1613) [2021-04-15 20:45:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][740/1251] eta 0:02:30 lr 0.000878 time 0.2913 (0.2950) loss 3.9410 (3.8400) grad_norm 1.0599 (1.1605) [2021-04-15 20:45:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][750/1251] eta 0:02:27 lr 0.000878 time 0.2947 (0.2947) loss 3.7068 (3.8386) grad_norm 1.0267 (1.1593) [2021-04-15 20:45:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][760/1251] eta 0:02:24 lr 0.000878 time 0.2541 (0.2945) loss 4.2312 (3.8383) grad_norm 1.0984 (1.1592) [2021-04-15 20:45:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][770/1251] eta 0:02:21 lr 0.000878 time 0.2831 (0.2944) loss 4.1845 (3.8409) grad_norm 1.0776 (1.1594) [2021-04-15 20:45:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][780/1251] eta 0:02:18 lr 0.000878 time 0.2967 (0.2945) loss 3.8096 (3.8361) grad_norm 1.2555 (1.1603) [2021-04-15 20:45:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][790/1251] eta 0:02:15 lr 0.000878 time 0.2579 (0.2943) loss 2.8621 (3.8340) grad_norm 1.4716 (1.1616) [2021-04-15 20:45:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][800/1251] eta 0:02:12 lr 0.000878 time 0.3066 (0.2942) loss 3.2347 (3.8336) grad_norm 1.0075 (1.1616) [2021-04-15 20:45:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][810/1251] eta 0:02:09 lr 0.000878 time 0.2841 (0.2940) loss 2.7787 (3.8307) grad_norm 1.2123 (1.1610) [2021-04-15 20:45:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][820/1251] eta 0:02:06 lr 0.000877 time 0.2554 (0.2938) loss 4.8218 (3.8320) grad_norm 0.9436 (1.1600) [2021-04-15 20:45:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][830/1251] eta 0:02:03 lr 0.000877 time 0.2776 (0.2937) loss 2.8335 (3.8312) grad_norm 1.0573 (1.1596) [2021-04-15 20:45:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][840/1251] eta 0:02:00 lr 0.000877 time 0.2932 (0.2936) loss 3.3942 (3.8282) grad_norm 1.0508 (1.1592) [2021-04-15 20:45:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][850/1251] eta 0:01:57 lr 0.000877 time 0.2930 (0.2934) loss 4.4952 (3.8297) grad_norm 1.1286 (1.1587) [2021-04-15 20:45:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][860/1251] eta 0:01:54 lr 0.000877 time 0.2556 (0.2933) loss 3.2503 (3.8292) grad_norm 1.1279 (1.1582) [2021-04-15 20:45:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][870/1251] eta 0:01:51 lr 0.000877 time 0.2936 (0.2933) loss 4.0098 (3.8317) grad_norm 1.2580 (1.1583) [2021-04-15 20:45:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][880/1251] eta 0:01:48 lr 0.000877 time 0.2698 (0.2933) loss 4.0229 (3.8345) grad_norm 0.9971 (1.1586) [2021-04-15 20:45:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][890/1251] eta 0:01:45 lr 0.000877 time 0.2767 (0.2932) loss 3.1955 (3.8343) grad_norm 1.3630 (1.1593) [2021-04-15 20:45:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][900/1251] eta 0:01:42 lr 0.000877 time 0.2821 (0.2931) loss 4.2995 (3.8318) grad_norm 1.0632 (1.1589) [2021-04-15 20:45:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][910/1251] eta 0:01:39 lr 0.000877 time 0.2904 (0.2930) loss 3.0725 (3.8323) grad_norm 1.0714 (1.1585) [2021-04-15 20:45:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][920/1251] eta 0:01:37 lr 0.000877 time 0.4594 (0.2931) loss 3.4089 (3.8333) grad_norm 1.1531 (1.1580) [2021-04-15 20:45:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][930/1251] eta 0:01:34 lr 0.000877 time 0.2548 (0.2931) loss 4.1968 (3.8320) grad_norm 1.3252 (1.1580) [2021-04-15 20:46:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][940/1251] eta 0:01:31 lr 0.000877 time 0.2668 (0.2931) loss 2.9185 (3.8288) grad_norm 1.0538 (1.1577) [2021-04-15 20:46:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][950/1251] eta 0:01:28 lr 0.000877 time 0.3144 (0.2931) loss 2.9061 (3.8252) grad_norm 1.0175 (1.1579) [2021-04-15 20:46:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][960/1251] eta 0:01:25 lr 0.000877 time 0.2916 (0.2930) loss 3.2106 (3.8237) grad_norm 1.0988 (1.1587) [2021-04-15 20:46:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][970/1251] eta 0:01:22 lr 0.000877 time 0.2915 (0.2929) loss 3.3708 (3.8239) grad_norm 1.2090 (1.1595) [2021-04-15 20:46:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][980/1251] eta 0:01:19 lr 0.000877 time 0.2950 (0.2928) loss 3.9525 (3.8236) grad_norm 0.9779 (1.1592) [2021-04-15 20:46:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][990/1251] eta 0:01:16 lr 0.000877 time 0.2737 (0.2927) loss 4.0758 (3.8260) grad_norm 1.1216 (1.1592) [2021-04-15 20:46:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][1000/1251] eta 0:01:13 lr 0.000877 time 0.2511 (0.2925) loss 2.9311 (3.8252) grad_norm 1.0470 (1.1587) [2021-04-15 20:46:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][1010/1251] eta 0:01:10 lr 0.000877 time 0.2755 (0.2924) loss 4.1710 (3.8249) grad_norm 1.0502 (1.1577) [2021-04-15 20:46:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][1020/1251] eta 0:01:07 lr 0.000877 time 0.2944 (0.2924) loss 4.1005 (3.8217) grad_norm 1.3194 (1.1576) [2021-04-15 20:46:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][1030/1251] eta 0:01:04 lr 0.000877 time 0.2914 (0.2923) loss 3.7586 (3.8214) grad_norm 1.2632 (1.1576) [2021-04-15 20:46:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][1040/1251] eta 0:01:01 lr 0.000877 time 0.2887 (0.2922) loss 4.1775 (3.8236) grad_norm 1.1016 (1.1579) [2021-04-15 20:46:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][1050/1251] eta 0:00:58 lr 0.000877 time 0.2740 (0.2921) loss 3.5904 (3.8231) grad_norm 1.0140 (1.1573) [2021-04-15 20:46:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][1060/1251] eta 0:00:55 lr 0.000877 time 0.2847 (0.2920) loss 4.1483 (3.8246) grad_norm 1.1944 (1.1581) [2021-04-15 20:46:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][1070/1251] eta 0:00:52 lr 0.000877 time 0.2722 (0.2920) loss 4.1325 (3.8255) grad_norm 1.2569 (1.1575) [2021-04-15 20:46:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][1080/1251] eta 0:00:49 lr 0.000877 time 0.3077 (0.2919) loss 3.8168 (3.8236) grad_norm 1.2210 (1.1573) [2021-04-15 20:46:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][1090/1251] eta 0:00:46 lr 0.000877 time 0.2670 (0.2918) loss 4.1975 (3.8260) grad_norm 1.1427 (1.1578) [2021-04-15 20:46:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][1100/1251] eta 0:00:44 lr 0.000877 time 0.2702 (0.2917) loss 3.6700 (3.8236) grad_norm 1.2753 (1.1577) [2021-04-15 20:46:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][1110/1251] eta 0:00:41 lr 0.000877 time 0.2954 (0.2916) loss 2.6704 (3.8241) grad_norm 0.9931 (1.1566) [2021-04-15 20:46:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][1120/1251] eta 0:00:38 lr 0.000877 time 0.2678 (0.2915) loss 4.4872 (3.8259) grad_norm 1.0829 (1.1561) [2021-04-15 20:46:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][1130/1251] eta 0:00:35 lr 0.000877 time 0.2556 (0.2916) loss 4.2679 (3.8262) grad_norm 1.2045 (1.1567) [2021-04-15 20:46:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][1140/1251] eta 0:00:32 lr 0.000877 time 0.2662 (0.2916) loss 3.8729 (3.8266) grad_norm 1.2155 (1.1566) [2021-04-15 20:47:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][1150/1251] eta 0:00:29 lr 0.000877 time 0.2780 (0.2916) loss 3.3457 (3.8246) grad_norm 1.0785 (1.1563) [2021-04-15 20:47:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][1160/1251] eta 0:00:26 lr 0.000877 time 0.2430 (0.2914) loss 2.6474 (3.8239) grad_norm 1.1341 (1.1557) [2021-04-15 20:47:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][1170/1251] eta 0:00:23 lr 0.000877 time 0.2596 (0.2915) loss 3.1507 (3.8248) grad_norm 1.1440 (1.1556) [2021-04-15 20:47:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][1180/1251] eta 0:00:20 lr 0.000876 time 0.2714 (0.2913) loss 4.2978 (3.8237) grad_norm 1.0625 (1.1552) [2021-04-15 20:47:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][1190/1251] eta 0:00:17 lr 0.000876 time 0.2848 (0.2913) loss 4.4944 (3.8230) grad_norm 1.1555 (1.1546) [2021-04-15 20:47:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][1200/1251] eta 0:00:14 lr 0.000876 time 0.2671 (0.2912) loss 3.8226 (3.8241) grad_norm 1.2054 (1.1551) [2021-04-15 20:47:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][1210/1251] eta 0:00:11 lr 0.000876 time 0.2897 (0.2911) loss 4.4979 (3.8232) grad_norm 1.0654 (1.1546) [2021-04-15 20:47:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][1220/1251] eta 0:00:09 lr 0.000876 time 0.3070 (0.2911) loss 3.5182 (3.8218) grad_norm 1.0876 (1.1537) [2021-04-15 20:47:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][1230/1251] eta 0:00:06 lr 0.000876 time 0.3012 (0.2910) loss 4.1074 (3.8195) grad_norm 1.2695 (1.1538) [2021-04-15 20:47:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][1240/1251] eta 0:00:03 lr 0.000876 time 0.2481 (0.2908) loss 4.5092 (3.8198) grad_norm 1.0812 (1.1538) [2021-04-15 20:47:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [68/300][1250/1251] eta 0:00:00 lr 0.000876 time 0.2481 (0.2905) loss 3.6562 (3.8175) grad_norm 1.0128 (1.1533) [2021-04-15 20:47:32 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 68 training takes 0:06:05 [2021-04-15 20:47:32 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_68.pth saving...... [2021-04-15 20:47:42 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_68.pth saved !!! [2021-04-15 20:47:43 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.168 (1.168) Loss 1.1427 (1.1427) Acc@1 73.145 (73.145) Acc@5 91.992 (91.992) [2021-04-15 20:47:45 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.633 (0.281) Loss 1.0652 (1.1771) Acc@1 75.000 (72.399) Acc@5 92.676 (91.486) [2021-04-15 20:47:47 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.185 (0.222) Loss 1.1950 (1.1927) Acc@1 71.094 (72.103) Acc@5 91.602 (91.197) [2021-04-15 20:47:50 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.192 (0.240) Loss 1.1602 (1.1887) Acc@1 72.559 (71.992) Acc@5 91.016 (91.192) [2021-04-15 20:47:51 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.223) Loss 1.1572 (1.1956) Acc@1 71.680 (71.892) Acc@5 91.504 (91.156) [2021-04-15 20:47:54 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 71.910 Acc@5 91.152 [2021-04-15 20:47:54 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 71.9% [2021-04-15 20:47:54 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 71.91% [2021-04-15 20:47:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][0/1251] eta 1:34:17 lr 0.000876 time 4.5222 (4.5222) loss 2.4046 (2.4046) grad_norm 1.0859 (1.0859) [2021-04-15 20:48:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][10/1251] eta 0:13:43 lr 0.000876 time 0.2904 (0.6633) loss 3.5151 (3.9167) grad_norm 1.2449 (1.1883) [2021-04-15 20:48:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][20/1251] eta 0:09:50 lr 0.000876 time 0.2748 (0.4793) loss 4.0723 (3.9950) grad_norm 1.0310 (1.1510) [2021-04-15 20:48:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][30/1251] eta 0:08:27 lr 0.000876 time 0.3089 (0.4154) loss 4.4172 (3.9281) grad_norm 1.0927 (1.1426) [2021-04-15 20:48:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][40/1251] eta 0:07:42 lr 0.000876 time 0.2686 (0.3822) loss 3.6603 (3.9022) grad_norm 1.2590 (1.1594) [2021-04-15 20:48:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][50/1251] eta 0:07:15 lr 0.000876 time 0.3075 (0.3626) loss 4.0170 (3.8698) grad_norm 1.1707 (1.1587) [2021-04-15 20:48:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][60/1251] eta 0:06:58 lr 0.000876 time 0.4008 (0.3514) loss 4.4962 (3.9066) grad_norm 1.3085 (1.1624) [2021-04-15 20:48:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][70/1251] eta 0:06:43 lr 0.000876 time 0.2780 (0.3419) loss 3.9237 (3.9205) grad_norm 1.1022 (1.1571) [2021-04-15 20:48:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][80/1251] eta 0:06:32 lr 0.000876 time 0.2980 (0.3348) loss 3.8639 (3.9007) grad_norm 1.0631 (1.1575) [2021-04-15 20:48:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][90/1251] eta 0:06:23 lr 0.000876 time 0.2950 (0.3301) loss 4.0267 (3.9226) grad_norm 1.0960 (1.1545) [2021-04-15 20:48:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][100/1251] eta 0:06:14 lr 0.000876 time 0.3001 (0.3252) loss 4.4905 (3.9135) grad_norm 1.1969 (1.1577) [2021-04-15 20:48:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][110/1251] eta 0:06:08 lr 0.000876 time 0.4597 (0.3229) loss 4.6848 (3.9248) grad_norm 0.9885 (1.1574) [2021-04-15 20:48:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][120/1251] eta 0:06:02 lr 0.000876 time 0.2535 (0.3201) loss 3.7037 (3.9252) grad_norm 1.0321 (1.1552) [2021-04-15 20:48:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][130/1251] eta 0:05:57 lr 0.000876 time 0.4757 (0.3188) loss 3.6453 (3.9194) grad_norm 0.9685 (1.1493) [2021-04-15 20:48:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][140/1251] eta 0:05:51 lr 0.000876 time 0.2925 (0.3161) loss 4.5466 (3.9153) grad_norm 1.3658 (1.1499) [2021-04-15 20:48:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][150/1251] eta 0:05:46 lr 0.000876 time 0.2559 (0.3144) loss 4.5912 (3.9295) grad_norm 1.2479 (1.1525) [2021-04-15 20:48:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][160/1251] eta 0:05:41 lr 0.000876 time 0.2916 (0.3127) loss 4.8397 (3.9281) grad_norm 1.0722 (1.1493) [2021-04-15 20:48:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][170/1251] eta 0:05:36 lr 0.000876 time 0.2753 (0.3109) loss 3.0585 (3.9088) grad_norm 1.2559 (1.1491) [2021-04-15 20:48:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][180/1251] eta 0:05:30 lr 0.000876 time 0.2650 (0.3088) loss 4.2487 (3.9189) grad_norm 1.0838 (1.1464) [2021-04-15 20:48:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][190/1251] eta 0:05:26 lr 0.000876 time 0.3132 (0.3074) loss 4.4008 (3.9070) grad_norm 1.4999 (1.1507) [2021-04-15 20:48:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][200/1251] eta 0:05:21 lr 0.000876 time 0.2866 (0.3058) loss 3.8188 (3.9085) grad_norm 1.3000 (1.1525) [2021-04-15 20:48:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][210/1251] eta 0:05:17 lr 0.000876 time 0.2958 (0.3048) loss 3.9755 (3.9031) grad_norm 1.3647 (1.1557) [2021-04-15 20:49:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][220/1251] eta 0:05:13 lr 0.000876 time 0.2951 (0.3038) loss 4.2083 (3.8830) grad_norm 1.1746 (1.1546) [2021-04-15 20:49:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][230/1251] eta 0:05:09 lr 0.000876 time 0.2873 (0.3029) loss 2.9315 (3.8786) grad_norm 1.0279 (1.1539) [2021-04-15 20:49:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][240/1251] eta 0:05:05 lr 0.000876 time 0.2470 (0.3019) loss 3.1990 (3.8755) grad_norm 1.0278 (1.1547) [2021-04-15 20:49:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][250/1251] eta 0:05:01 lr 0.000876 time 0.3382 (0.3014) loss 4.1267 (3.8861) grad_norm 1.2292 (1.1527) [2021-04-15 20:49:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][260/1251] eta 0:04:57 lr 0.000876 time 0.2805 (0.3005) loss 4.6758 (3.8900) grad_norm 1.2923 (1.1544) [2021-04-15 20:49:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][270/1251] eta 0:04:54 lr 0.000876 time 0.2793 (0.2998) loss 4.0538 (3.8884) grad_norm 1.1028 (1.1551) [2021-04-15 20:49:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][280/1251] eta 0:04:50 lr 0.000876 time 0.2656 (0.2990) loss 4.4024 (3.8852) grad_norm 1.1180 (inf) [2021-04-15 20:49:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][290/1251] eta 0:04:47 lr 0.000876 time 0.2795 (0.2988) loss 3.5697 (3.8797) grad_norm 1.2301 (inf) [2021-04-15 20:49:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][300/1251] eta 0:04:43 lr 0.000875 time 0.2686 (0.2981) loss 3.5791 (3.8672) grad_norm 1.1594 (inf) [2021-04-15 20:49:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][310/1251] eta 0:04:39 lr 0.000875 time 0.2771 (0.2975) loss 2.8479 (3.8619) grad_norm 1.0919 (inf) [2021-04-15 20:49:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][320/1251] eta 0:04:36 lr 0.000875 time 0.3075 (0.2970) loss 3.7619 (3.8614) grad_norm 1.0075 (inf) [2021-04-15 20:49:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][330/1251] eta 0:04:33 lr 0.000875 time 0.2792 (0.2965) loss 3.6167 (3.8645) grad_norm 1.0642 (inf) [2021-04-15 20:49:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][340/1251] eta 0:04:30 lr 0.000875 time 0.2682 (0.2965) loss 2.6853 (3.8652) grad_norm 1.1716 (inf) [2021-04-15 20:49:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][350/1251] eta 0:04:27 lr 0.000875 time 0.2750 (0.2965) loss 4.1612 (3.8664) grad_norm 1.1350 (inf) [2021-04-15 20:49:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][360/1251] eta 0:04:24 lr 0.000875 time 0.2851 (0.2964) loss 3.9664 (3.8621) grad_norm 1.0579 (inf) [2021-04-15 20:49:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][370/1251] eta 0:04:21 lr 0.000875 time 0.2582 (0.2971) loss 4.5291 (3.8588) grad_norm 1.0375 (inf) [2021-04-15 20:49:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][380/1251] eta 0:04:18 lr 0.000875 time 0.3294 (0.2967) loss 4.5974 (3.8546) grad_norm 1.2218 (inf) [2021-04-15 20:49:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][390/1251] eta 0:04:15 lr 0.000875 time 0.2499 (0.2965) loss 3.8343 (3.8559) grad_norm 1.2027 (inf) [2021-04-15 20:49:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][400/1251] eta 0:04:12 lr 0.000875 time 0.2496 (0.2961) loss 3.9749 (3.8562) grad_norm 1.0823 (inf) [2021-04-15 20:49:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][410/1251] eta 0:04:08 lr 0.000875 time 0.2731 (0.2958) loss 3.9867 (3.8555) grad_norm 1.2953 (inf) [2021-04-15 20:49:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][420/1251] eta 0:04:05 lr 0.000875 time 0.2975 (0.2955) loss 2.8364 (3.8511) grad_norm 1.3187 (inf) [2021-04-15 20:50:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][430/1251] eta 0:04:02 lr 0.000875 time 0.2995 (0.2953) loss 4.0381 (3.8536) grad_norm 1.2185 (inf) [2021-04-15 20:50:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][440/1251] eta 0:03:59 lr 0.000875 time 0.2851 (0.2949) loss 3.6280 (3.8460) grad_norm 0.9681 (inf) [2021-04-15 20:50:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][450/1251] eta 0:03:55 lr 0.000875 time 0.2808 (0.2944) loss 3.9589 (3.8480) grad_norm 1.2259 (inf) [2021-04-15 20:50:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][460/1251] eta 0:03:52 lr 0.000875 time 0.2855 (0.2941) loss 4.0792 (3.8448) grad_norm 1.1532 (inf) [2021-04-15 20:50:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][470/1251] eta 0:03:49 lr 0.000875 time 0.2591 (0.2938) loss 3.9545 (3.8401) grad_norm 1.2151 (inf) [2021-04-15 20:50:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][480/1251] eta 0:03:46 lr 0.000875 time 0.2940 (0.2936) loss 4.2237 (3.8466) grad_norm 1.1185 (inf) [2021-04-15 20:50:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][490/1251] eta 0:03:43 lr 0.000875 time 0.2533 (0.2934) loss 3.3230 (3.8442) grad_norm 1.2263 (inf) [2021-04-15 20:50:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][500/1251] eta 0:03:40 lr 0.000875 time 0.2908 (0.2932) loss 2.6826 (3.8378) grad_norm 1.1080 (inf) [2021-04-15 20:50:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][510/1251] eta 0:03:37 lr 0.000875 time 0.2988 (0.2931) loss 4.1082 (3.8377) grad_norm 1.1127 (inf) [2021-04-15 20:50:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][520/1251] eta 0:03:34 lr 0.000875 time 0.2575 (0.2930) loss 3.2855 (3.8378) grad_norm 1.1566 (inf) [2021-04-15 20:50:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][530/1251] eta 0:03:31 lr 0.000875 time 0.2741 (0.2927) loss 4.1584 (3.8339) grad_norm 1.1151 (inf) [2021-04-15 20:50:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][540/1251] eta 0:03:28 lr 0.000875 time 0.2780 (0.2927) loss 3.1788 (3.8368) grad_norm 1.3088 (inf) [2021-04-15 20:50:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][550/1251] eta 0:03:24 lr 0.000875 time 0.2677 (0.2924) loss 4.3574 (3.8436) grad_norm 1.1191 (inf) [2021-04-15 20:50:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][560/1251] eta 0:03:21 lr 0.000875 time 0.2904 (0.2923) loss 4.0747 (3.8447) grad_norm 1.1150 (inf) [2021-04-15 20:50:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][570/1251] eta 0:03:18 lr 0.000875 time 0.3074 (0.2921) loss 4.1532 (3.8393) grad_norm 1.1934 (inf) [2021-04-15 20:50:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][580/1251] eta 0:03:16 lr 0.000875 time 0.2734 (0.2925) loss 3.4699 (3.8458) grad_norm 1.1267 (inf) [2021-04-15 20:50:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][590/1251] eta 0:03:13 lr 0.000875 time 0.2785 (0.2925) loss 4.0491 (3.8461) grad_norm 1.1982 (inf) [2021-04-15 20:50:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][600/1251] eta 0:03:10 lr 0.000875 time 0.2942 (0.2923) loss 4.0473 (3.8462) grad_norm 1.0846 (inf) [2021-04-15 20:50:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][610/1251] eta 0:03:07 lr 0.000875 time 0.3000 (0.2921) loss 4.2306 (3.8430) grad_norm 1.0821 (inf) [2021-04-15 20:50:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][620/1251] eta 0:03:04 lr 0.000875 time 0.2638 (0.2919) loss 4.2844 (3.8478) grad_norm 1.0665 (inf) [2021-04-15 20:50:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][630/1251] eta 0:03:01 lr 0.000875 time 0.2829 (0.2917) loss 4.0729 (3.8543) grad_norm 1.3692 (inf) [2021-04-15 20:51:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][640/1251] eta 0:02:58 lr 0.000875 time 0.2829 (0.2918) loss 4.0384 (3.8544) grad_norm 1.0716 (inf) [2021-04-15 20:51:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][650/1251] eta 0:02:55 lr 0.000875 time 0.2842 (0.2916) loss 3.1011 (3.8546) grad_norm 1.1669 (inf) [2021-04-15 20:51:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][660/1251] eta 0:02:52 lr 0.000874 time 0.2716 (0.2914) loss 2.6811 (3.8544) grad_norm 0.9618 (inf) [2021-04-15 20:51:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][670/1251] eta 0:02:49 lr 0.000874 time 0.2885 (0.2914) loss 3.5197 (3.8570) grad_norm 0.9977 (inf) [2021-04-15 20:51:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][680/1251] eta 0:02:46 lr 0.000874 time 0.2753 (0.2913) loss 4.2380 (3.8602) grad_norm 1.1012 (inf) [2021-04-15 20:51:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][690/1251] eta 0:02:43 lr 0.000874 time 0.2938 (0.2914) loss 4.3177 (3.8594) grad_norm 1.0486 (inf) [2021-04-15 20:51:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][700/1251] eta 0:02:40 lr 0.000874 time 0.2760 (0.2911) loss 4.0899 (3.8598) grad_norm 1.1023 (inf) [2021-04-15 20:51:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][710/1251] eta 0:02:37 lr 0.000874 time 0.2696 (0.2913) loss 3.2122 (3.8617) grad_norm 1.2060 (inf) [2021-04-15 20:51:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][720/1251] eta 0:02:34 lr 0.000874 time 0.2597 (0.2913) loss 4.1148 (3.8654) grad_norm 0.9962 (inf) [2021-04-15 20:51:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][730/1251] eta 0:02:31 lr 0.000874 time 0.2632 (0.2911) loss 3.9783 (3.8628) grad_norm 1.0691 (inf) [2021-04-15 20:51:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][740/1251] eta 0:02:28 lr 0.000874 time 0.2498 (0.2909) loss 2.7840 (3.8618) grad_norm 1.3004 (inf) [2021-04-15 20:51:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][750/1251] eta 0:02:25 lr 0.000874 time 0.2684 (0.2907) loss 2.9393 (3.8653) grad_norm 1.0646 (inf) [2021-04-15 20:51:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][760/1251] eta 0:02:22 lr 0.000874 time 0.2798 (0.2906) loss 3.8384 (3.8663) grad_norm 1.3639 (inf) [2021-04-15 20:51:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][770/1251] eta 0:02:19 lr 0.000874 time 0.2877 (0.2905) loss 3.8137 (3.8674) grad_norm 1.0572 (inf) [2021-04-15 20:51:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][780/1251] eta 0:02:16 lr 0.000874 time 0.2671 (0.2903) loss 2.8956 (3.8685) grad_norm 1.2510 (inf) [2021-04-15 20:51:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][790/1251] eta 0:02:13 lr 0.000874 time 0.2634 (0.2902) loss 3.9669 (3.8696) grad_norm 1.0292 (inf) [2021-04-15 20:51:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][800/1251] eta 0:02:10 lr 0.000874 time 0.2773 (0.2900) loss 4.4177 (3.8702) grad_norm 1.4273 (inf) [2021-04-15 20:51:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][810/1251] eta 0:02:07 lr 0.000874 time 0.2942 (0.2899) loss 4.2355 (3.8702) grad_norm 1.1154 (inf) [2021-04-15 20:51:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][820/1251] eta 0:02:04 lr 0.000874 time 0.2781 (0.2898) loss 4.5121 (3.8749) grad_norm 1.1461 (inf) [2021-04-15 20:51:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][830/1251] eta 0:02:01 lr 0.000874 time 0.2789 (0.2897) loss 3.7551 (3.8766) grad_norm 1.1248 (inf) [2021-04-15 20:51:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][840/1251] eta 0:01:59 lr 0.000874 time 0.2972 (0.2896) loss 4.1292 (3.8752) grad_norm 1.4024 (inf) [2021-04-15 20:52:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][850/1251] eta 0:01:56 lr 0.000874 time 0.2746 (0.2896) loss 2.8947 (3.8779) grad_norm 1.6206 (inf) [2021-04-15 20:52:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][860/1251] eta 0:01:53 lr 0.000874 time 0.2615 (0.2895) loss 4.1820 (3.8795) grad_norm 1.2264 (inf) [2021-04-15 20:52:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][870/1251] eta 0:01:50 lr 0.000874 time 0.2884 (0.2895) loss 3.8221 (3.8773) grad_norm 1.1839 (inf) [2021-04-15 20:52:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][880/1251] eta 0:01:47 lr 0.000874 time 0.2889 (0.2894) loss 3.8924 (3.8800) grad_norm 1.0479 (inf) [2021-04-15 20:52:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][890/1251] eta 0:01:44 lr 0.000874 time 0.2942 (0.2893) loss 4.5159 (3.8839) grad_norm 1.1289 (inf) [2021-04-15 20:52:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][900/1251] eta 0:01:41 lr 0.000874 time 0.2697 (0.2891) loss 4.1260 (3.8812) grad_norm 1.1017 (inf) [2021-04-15 20:52:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][910/1251] eta 0:01:38 lr 0.000874 time 0.2875 (0.2890) loss 3.1311 (3.8826) grad_norm 1.1500 (inf) [2021-04-15 20:52:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][920/1251] eta 0:01:35 lr 0.000874 time 0.2674 (0.2890) loss 4.1837 (3.8817) grad_norm 1.0667 (inf) [2021-04-15 20:52:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][930/1251] eta 0:01:32 lr 0.000874 time 0.2745 (0.2890) loss 3.4071 (3.8781) grad_norm 1.0621 (inf) [2021-04-15 20:52:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][940/1251] eta 0:01:29 lr 0.000874 time 0.2656 (0.2891) loss 4.1632 (3.8779) grad_norm 1.2254 (inf) [2021-04-15 20:52:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][950/1251] eta 0:01:27 lr 0.000874 time 0.2506 (0.2891) loss 4.0146 (3.8742) grad_norm 1.0924 (inf) [2021-04-15 20:52:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][960/1251] eta 0:01:24 lr 0.000874 time 0.2699 (0.2891) loss 3.3955 (3.8734) grad_norm 0.9868 (inf) [2021-04-15 20:52:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][970/1251] eta 0:01:21 lr 0.000874 time 0.2854 (0.2890) loss 4.1372 (3.8731) grad_norm 1.1895 (inf) [2021-04-15 20:52:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][980/1251] eta 0:01:18 lr 0.000874 time 0.2749 (0.2889) loss 3.7255 (3.8737) grad_norm 1.1834 (inf) [2021-04-15 20:52:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][990/1251] eta 0:01:15 lr 0.000874 time 0.2911 (0.2889) loss 4.4685 (3.8762) grad_norm 1.1293 (inf) [2021-04-15 20:52:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][1000/1251] eta 0:01:12 lr 0.000874 time 0.2874 (0.2888) loss 3.9137 (3.8764) grad_norm 1.2472 (inf) [2021-04-15 20:52:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][1010/1251] eta 0:01:09 lr 0.000874 time 0.2991 (0.2887) loss 4.0617 (3.8797) grad_norm 1.5186 (inf) [2021-04-15 20:52:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][1020/1251] eta 0:01:06 lr 0.000873 time 0.2927 (0.2886) loss 3.5609 (3.8797) grad_norm 1.0699 (inf) [2021-04-15 20:52:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][1030/1251] eta 0:01:03 lr 0.000873 time 0.2753 (0.2885) loss 3.8155 (3.8784) grad_norm 1.2748 (inf) [2021-04-15 20:52:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][1040/1251] eta 0:01:00 lr 0.000873 time 0.2815 (0.2884) loss 3.5296 (3.8800) grad_norm 1.3701 (inf) [2021-04-15 20:52:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][1050/1251] eta 0:00:57 lr 0.000873 time 0.2651 (0.2884) loss 4.2203 (3.8780) grad_norm 1.1970 (inf) [2021-04-15 20:53:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][1060/1251] eta 0:00:55 lr 0.000873 time 0.2922 (0.2883) loss 4.0445 (3.8765) grad_norm 1.1628 (inf) [2021-04-15 20:53:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][1070/1251] eta 0:00:52 lr 0.000873 time 0.2910 (0.2882) loss 4.4351 (3.8780) grad_norm 1.1178 (inf) [2021-04-15 20:53:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][1080/1251] eta 0:00:49 lr 0.000873 time 0.2565 (0.2881) loss 4.0015 (3.8771) grad_norm 1.1188 (inf) [2021-04-15 20:53:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][1090/1251] eta 0:00:46 lr 0.000873 time 0.2575 (0.2881) loss 4.4712 (3.8771) grad_norm 0.9369 (inf) [2021-04-15 20:53:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][1100/1251] eta 0:00:43 lr 0.000873 time 0.3071 (0.2881) loss 4.5201 (3.8779) grad_norm 1.0342 (inf) [2021-04-15 20:53:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][1110/1251] eta 0:00:40 lr 0.000873 time 0.2905 (0.2880) loss 3.6654 (3.8776) grad_norm 1.3453 (inf) [2021-04-15 20:53:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][1120/1251] eta 0:00:37 lr 0.000873 time 0.2594 (0.2880) loss 2.9841 (3.8780) grad_norm 1.2040 (inf) [2021-04-15 20:53:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][1130/1251] eta 0:00:34 lr 0.000873 time 0.2867 (0.2879) loss 3.5521 (3.8763) grad_norm 0.9992 (inf) [2021-04-15 20:53:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][1140/1251] eta 0:00:31 lr 0.000873 time 0.2977 (0.2879) loss 4.4037 (3.8756) grad_norm 1.1450 (inf) [2021-04-15 20:53:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][1150/1251] eta 0:00:29 lr 0.000873 time 0.2651 (0.2878) loss 4.4708 (3.8742) grad_norm 1.1419 (inf) [2021-04-15 20:53:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][1160/1251] eta 0:00:26 lr 0.000873 time 0.3320 (0.2880) loss 3.8859 (3.8745) grad_norm 1.1386 (inf) [2021-04-15 20:53:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][1170/1251] eta 0:00:23 lr 0.000873 time 0.3222 (0.2879) loss 4.1988 (3.8779) grad_norm 0.9679 (inf) [2021-04-15 20:53:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][1180/1251] eta 0:00:20 lr 0.000873 time 0.2757 (0.2878) loss 2.9166 (3.8764) grad_norm 1.2293 (inf) [2021-04-15 20:53:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][1190/1251] eta 0:00:17 lr 0.000873 time 0.2949 (0.2877) loss 3.5742 (3.8757) grad_norm 1.0929 (inf) [2021-04-15 20:53:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][1200/1251] eta 0:00:14 lr 0.000873 time 0.2786 (0.2876) loss 4.2821 (3.8768) grad_norm 1.6144 (inf) [2021-04-15 20:53:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][1210/1251] eta 0:00:11 lr 0.000873 time 0.2597 (0.2875) loss 4.0605 (3.8753) grad_norm 1.0440 (inf) [2021-04-15 20:53:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][1220/1251] eta 0:00:08 lr 0.000873 time 0.2777 (0.2875) loss 4.0450 (3.8733) grad_norm 1.0352 (inf) [2021-04-15 20:53:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][1230/1251] eta 0:00:06 lr 0.000873 time 0.2804 (0.2875) loss 3.4530 (3.8728) grad_norm 1.1484 (inf) [2021-04-15 20:53:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][1240/1251] eta 0:00:03 lr 0.000873 time 0.2483 (0.2874) loss 4.0416 (3.8713) grad_norm 1.0716 (inf) [2021-04-15 20:53:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [69/300][1250/1251] eta 0:00:00 lr 0.000873 time 0.2485 (0.2871) loss 3.3706 (3.8708) grad_norm 1.0163 (inf) [2021-04-15 20:53:55 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 69 training takes 0:06:01 [2021-04-15 20:53:55 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_69.pth saving...... [2021-04-15 20:54:04 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_69.pth saved !!! [2021-04-15 20:54:06 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.280 (1.280) Loss 1.2948 (1.2948) Acc@1 69.531 (69.531) Acc@5 90.332 (90.332) [2021-04-15 20:54:08 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.424 (0.297) Loss 1.2143 (1.2276) Acc@1 72.559 (71.742) Acc@5 91.699 (91.335) [2021-04-15 20:54:09 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.115 (0.245) Loss 1.1207 (1.2243) Acc@1 74.414 (71.866) Acc@5 93.164 (91.509) [2021-04-15 20:54:12 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.151 (0.238) Loss 1.3021 (1.2336) Acc@1 69.824 (71.692) Acc@5 89.160 (91.384) [2021-04-15 20:54:14 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.278 (0.226) Loss 1.2466 (1.2366) Acc@1 71.387 (71.644) Acc@5 91.504 (91.254) [2021-04-15 20:54:16 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 71.614 Acc@5 91.124 [2021-04-15 20:54:16 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 71.6% [2021-04-15 20:54:16 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 71.91% [2021-04-15 20:54:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][0/1251] eta 1:36:19 lr 0.000873 time 4.6203 (4.6203) loss 4.1328 (4.1328) grad_norm 1.0516 (1.0516) [2021-04-15 20:54:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][10/1251] eta 0:13:51 lr 0.000873 time 0.2747 (0.6702) loss 4.4914 (3.7260) grad_norm 1.1874 (1.1178) [2021-04-15 20:54:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][20/1251] eta 0:09:55 lr 0.000873 time 0.2879 (0.4841) loss 4.3669 (3.8200) grad_norm 0.9580 (1.1140) [2021-04-15 20:54:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][30/1251] eta 0:08:32 lr 0.000873 time 0.3052 (0.4194) loss 4.1426 (3.8869) grad_norm 0.9528 (1.1241) [2021-04-15 20:54:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][40/1251] eta 0:07:47 lr 0.000873 time 0.2962 (0.3859) loss 4.2441 (3.8378) grad_norm 0.9938 (1.1067) [2021-04-15 20:54:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][50/1251] eta 0:07:18 lr 0.000873 time 0.2947 (0.3650) loss 3.9142 (3.7872) grad_norm 1.1855 (1.1060) [2021-04-15 20:54:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][60/1251] eta 0:07:01 lr 0.000873 time 0.4150 (0.3539) loss 3.9327 (3.7592) grad_norm 1.0601 (1.1020) [2021-04-15 20:54:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][70/1251] eta 0:06:44 lr 0.000873 time 0.2828 (0.3428) loss 4.2141 (3.7455) grad_norm 1.1648 (1.1091) [2021-04-15 20:54:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][80/1251] eta 0:06:32 lr 0.000873 time 0.2749 (0.3354) loss 4.7619 (3.7766) grad_norm 1.1875 (1.1223) [2021-04-15 20:54:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][90/1251] eta 0:06:22 lr 0.000873 time 0.2884 (0.3296) loss 3.9676 (3.7909) grad_norm 1.2760 (1.1267) [2021-04-15 20:54:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][100/1251] eta 0:06:13 lr 0.000873 time 0.2911 (0.3249) loss 3.9695 (3.7666) grad_norm 1.0870 (1.1250) [2021-04-15 20:54:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][110/1251] eta 0:06:06 lr 0.000873 time 0.3028 (0.3212) loss 4.3834 (3.7669) grad_norm 1.0721 (1.1269) [2021-04-15 20:54:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][120/1251] eta 0:06:01 lr 0.000873 time 0.3110 (0.3192) loss 3.1895 (3.7866) grad_norm 1.0756 (1.1296) [2021-04-15 20:54:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][130/1251] eta 0:05:55 lr 0.000872 time 0.3001 (0.3174) loss 3.8405 (3.7905) grad_norm 1.1019 (1.1250) [2021-04-15 20:55:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][140/1251] eta 0:05:51 lr 0.000872 time 0.2812 (0.3163) loss 4.2005 (3.8062) grad_norm 1.8554 (1.1332) [2021-04-15 20:55:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][150/1251] eta 0:05:46 lr 0.000872 time 0.2891 (0.3148) loss 3.4182 (3.8025) grad_norm 1.2529 (1.1367) [2021-04-15 20:55:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][160/1251] eta 0:05:40 lr 0.000872 time 0.2773 (0.3125) loss 4.3996 (3.8223) grad_norm 1.2084 (1.1338) [2021-04-15 20:55:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][170/1251] eta 0:05:35 lr 0.000872 time 0.2621 (0.3108) loss 4.1577 (3.8228) grad_norm 0.9772 (1.1310) [2021-04-15 20:55:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][180/1251] eta 0:05:31 lr 0.000872 time 0.2775 (0.3093) loss 3.8075 (3.8258) grad_norm 1.0371 (1.1293) [2021-04-15 20:55:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][190/1251] eta 0:05:26 lr 0.000872 time 0.2944 (0.3079) loss 3.1748 (3.8239) grad_norm 1.1136 (1.1285) [2021-04-15 20:55:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][200/1251] eta 0:05:21 lr 0.000872 time 0.2693 (0.3064) loss 4.0292 (3.8309) grad_norm 1.1913 (1.1268) [2021-04-15 20:55:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][210/1251] eta 0:05:17 lr 0.000872 time 0.2702 (0.3052) loss 3.1452 (3.8283) grad_norm 1.0650 (1.1283) [2021-04-15 20:55:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][220/1251] eta 0:05:13 lr 0.000872 time 0.2668 (0.3041) loss 3.1710 (3.8314) grad_norm 1.0196 (1.1253) [2021-04-15 20:55:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][230/1251] eta 0:05:09 lr 0.000872 time 0.3064 (0.3034) loss 3.6428 (3.8261) grad_norm 1.0023 (1.1227) [2021-04-15 20:55:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][240/1251] eta 0:05:05 lr 0.000872 time 0.2540 (0.3024) loss 3.8974 (3.8077) grad_norm 1.0294 (1.1266) [2021-04-15 20:55:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][250/1251] eta 0:05:01 lr 0.000872 time 0.2628 (0.3015) loss 4.4893 (3.8245) grad_norm 1.2634 (1.1315) [2021-04-15 20:55:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][260/1251] eta 0:04:57 lr 0.000872 time 0.2950 (0.3006) loss 3.1156 (3.8344) grad_norm 1.0656 (1.1305) [2021-04-15 20:55:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][270/1251] eta 0:04:54 lr 0.000872 time 0.2847 (0.2998) loss 3.6442 (3.8239) grad_norm 1.1005 (1.1350) [2021-04-15 20:55:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][280/1251] eta 0:04:50 lr 0.000872 time 0.2802 (0.2993) loss 4.4248 (3.8182) grad_norm 1.0471 (1.1325) [2021-04-15 20:55:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][290/1251] eta 0:04:47 lr 0.000872 time 0.3232 (0.2987) loss 3.2734 (3.8054) grad_norm 1.0944 (1.1332) [2021-04-15 20:55:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][300/1251] eta 0:04:43 lr 0.000872 time 0.2825 (0.2984) loss 3.3269 (3.8021) grad_norm 1.1938 (1.1334) [2021-04-15 20:55:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][310/1251] eta 0:04:40 lr 0.000872 time 0.2729 (0.2979) loss 4.0049 (3.8061) grad_norm 1.0473 (1.1365) [2021-04-15 20:55:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][320/1251] eta 0:04:37 lr 0.000872 time 0.2739 (0.2979) loss 4.0339 (3.8006) grad_norm 0.9991 (1.1395) [2021-04-15 20:55:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][330/1251] eta 0:04:33 lr 0.000872 time 0.2838 (0.2973) loss 3.7433 (3.8030) grad_norm 1.0290 (1.1392) [2021-04-15 20:55:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][340/1251] eta 0:04:30 lr 0.000872 time 0.2587 (0.2970) loss 3.6763 (3.7977) grad_norm 1.0599 (1.1377) [2021-04-15 20:56:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][350/1251] eta 0:04:27 lr 0.000872 time 0.4377 (0.2971) loss 3.7299 (3.7984) grad_norm 1.0200 (1.1362) [2021-04-15 20:56:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][360/1251] eta 0:04:24 lr 0.000872 time 0.2914 (0.2969) loss 4.3019 (3.8009) grad_norm 0.9995 (1.1363) [2021-04-15 20:56:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][370/1251] eta 0:04:21 lr 0.000872 time 0.2629 (0.2964) loss 4.0688 (3.8023) grad_norm 1.3546 (1.1391) [2021-04-15 20:56:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][380/1251] eta 0:04:17 lr 0.000872 time 0.2628 (0.2959) loss 3.4107 (3.8066) grad_norm 1.0987 (1.1394) [2021-04-15 20:56:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][390/1251] eta 0:04:14 lr 0.000872 time 0.2459 (0.2954) loss 3.1362 (3.8011) grad_norm 1.2964 (1.1388) [2021-04-15 20:56:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][400/1251] eta 0:04:11 lr 0.000872 time 0.2887 (0.2950) loss 2.7127 (3.8000) grad_norm 1.1207 (1.1375) [2021-04-15 20:56:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][410/1251] eta 0:04:07 lr 0.000872 time 0.2825 (0.2947) loss 3.5724 (3.8029) grad_norm 1.1696 (1.1368) [2021-04-15 20:56:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][420/1251] eta 0:04:04 lr 0.000872 time 0.2629 (0.2945) loss 4.1260 (3.8075) grad_norm 1.0742 (1.1369) [2021-04-15 20:56:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][430/1251] eta 0:04:01 lr 0.000872 time 0.2841 (0.2942) loss 3.8070 (3.8068) grad_norm 1.0542 (1.1354) [2021-04-15 20:56:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][440/1251] eta 0:03:58 lr 0.000872 time 0.2568 (0.2941) loss 4.5896 (3.8034) grad_norm 1.1229 (1.1351) [2021-04-15 20:56:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][450/1251] eta 0:03:55 lr 0.000872 time 0.2672 (0.2938) loss 4.5286 (3.8017) grad_norm 1.2147 (1.1352) [2021-04-15 20:56:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][460/1251] eta 0:03:52 lr 0.000872 time 0.3021 (0.2935) loss 3.6858 (3.8031) grad_norm 1.5305 (1.1360) [2021-04-15 20:56:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][470/1251] eta 0:03:49 lr 0.000872 time 0.2600 (0.2932) loss 3.0975 (3.8002) grad_norm 1.2838 (1.1378) [2021-04-15 20:56:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][480/1251] eta 0:03:45 lr 0.000872 time 0.2809 (0.2930) loss 4.5383 (3.8023) grad_norm 1.1999 (1.1376) [2021-04-15 20:56:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][490/1251] eta 0:03:42 lr 0.000871 time 0.2592 (0.2928) loss 2.8213 (3.8014) grad_norm 1.0312 (1.1382) [2021-04-15 20:56:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][500/1251] eta 0:03:39 lr 0.000871 time 0.3264 (0.2928) loss 3.2520 (3.8007) grad_norm 1.1793 (1.1387) [2021-04-15 20:56:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][510/1251] eta 0:03:36 lr 0.000871 time 0.3080 (0.2925) loss 4.2967 (3.8027) grad_norm 1.2442 (1.1388) [2021-04-15 20:56:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][520/1251] eta 0:03:33 lr 0.000871 time 0.2456 (0.2921) loss 4.0207 (3.8051) grad_norm 1.2628 (1.1374) [2021-04-15 20:56:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][530/1251] eta 0:03:30 lr 0.000871 time 0.2792 (0.2920) loss 4.2303 (3.8069) grad_norm 1.1592 (1.1372) [2021-04-15 20:56:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][540/1251] eta 0:03:27 lr 0.000871 time 0.2735 (0.2917) loss 4.0791 (3.8098) grad_norm 1.1269 (1.1392) [2021-04-15 20:56:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][550/1251] eta 0:03:24 lr 0.000871 time 0.2700 (0.2915) loss 3.9334 (3.8122) grad_norm 1.4493 (1.1392) [2021-04-15 20:57:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][560/1251] eta 0:03:21 lr 0.000871 time 0.2943 (0.2913) loss 3.9171 (3.8125) grad_norm 1.0061 (1.1395) [2021-04-15 20:57:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][570/1251] eta 0:03:18 lr 0.000871 time 0.2964 (0.2912) loss 4.3777 (3.8129) grad_norm 1.1148 (1.1391) [2021-04-15 20:57:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][580/1251] eta 0:03:15 lr 0.000871 time 0.2999 (0.2914) loss 4.4678 (3.8142) grad_norm 1.2601 (1.1386) [2021-04-15 20:57:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][590/1251] eta 0:03:12 lr 0.000871 time 0.2795 (0.2912) loss 2.6462 (3.8127) grad_norm 1.3362 (1.1386) [2021-04-15 20:57:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][600/1251] eta 0:03:09 lr 0.000871 time 0.2989 (0.2910) loss 3.2687 (3.8097) grad_norm 1.3989 (1.1393) [2021-04-15 20:57:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][610/1251] eta 0:03:06 lr 0.000871 time 0.2785 (0.2909) loss 4.9881 (3.8100) grad_norm 1.1936 (1.1396) [2021-04-15 20:57:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][620/1251] eta 0:03:03 lr 0.000871 time 0.2662 (0.2908) loss 3.7221 (3.8154) grad_norm 1.5438 (1.1407) [2021-04-15 20:57:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][630/1251] eta 0:03:00 lr 0.000871 time 0.2779 (0.2906) loss 4.0348 (3.8204) grad_norm 1.1119 (1.1418) [2021-04-15 20:57:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][640/1251] eta 0:02:57 lr 0.000871 time 0.2646 (0.2907) loss 4.0790 (3.8234) grad_norm 1.2683 (1.1429) [2021-04-15 20:57:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][650/1251] eta 0:02:54 lr 0.000871 time 0.2680 (0.2908) loss 3.6456 (3.8237) grad_norm 1.1099 (1.1434) [2021-04-15 20:57:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][660/1251] eta 0:02:51 lr 0.000871 time 0.2890 (0.2906) loss 3.8194 (3.8254) grad_norm 1.1557 (1.1433) [2021-04-15 20:57:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][670/1251] eta 0:02:48 lr 0.000871 time 0.2628 (0.2905) loss 2.9060 (3.8204) grad_norm 1.3348 (1.1437) [2021-04-15 20:57:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][680/1251] eta 0:02:45 lr 0.000871 time 0.3085 (0.2903) loss 4.0694 (3.8227) grad_norm 1.0697 (1.1437) [2021-04-15 20:57:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][690/1251] eta 0:02:42 lr 0.000871 time 0.2488 (0.2903) loss 4.2653 (3.8219) grad_norm 1.3492 (1.1442) [2021-04-15 20:57:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][700/1251] eta 0:02:40 lr 0.000871 time 0.4193 (0.2904) loss 4.5251 (3.8193) grad_norm 1.2821 (1.1460) [2021-04-15 20:57:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][710/1251] eta 0:02:37 lr 0.000871 time 0.2527 (0.2904) loss 3.8689 (3.8179) grad_norm 1.1704 (1.1470) [2021-04-15 20:57:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][720/1251] eta 0:02:34 lr 0.000871 time 0.2698 (0.2904) loss 3.4087 (3.8172) grad_norm 1.3850 (1.1474) [2021-04-15 20:57:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][730/1251] eta 0:02:31 lr 0.000871 time 0.2740 (0.2905) loss 4.2163 (3.8197) grad_norm 1.3857 (1.1471) [2021-04-15 20:57:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][740/1251] eta 0:02:28 lr 0.000871 time 0.2742 (0.2904) loss 4.0998 (3.8210) grad_norm 0.9936 (1.1464) [2021-04-15 20:57:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][750/1251] eta 0:02:25 lr 0.000871 time 0.2761 (0.2903) loss 3.4659 (3.8202) grad_norm 1.1189 (1.1461) [2021-04-15 20:57:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][760/1251] eta 0:02:22 lr 0.000871 time 0.2670 (0.2901) loss 4.3361 (3.8200) grad_norm 1.1186 (1.1476) [2021-04-15 20:58:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][770/1251] eta 0:02:19 lr 0.000871 time 0.2891 (0.2900) loss 3.6811 (3.8160) grad_norm 1.0448 (1.1477) [2021-04-15 20:58:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][780/1251] eta 0:02:16 lr 0.000871 time 0.2791 (0.2899) loss 2.8232 (3.8139) grad_norm 1.0425 (1.1467) [2021-04-15 20:58:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][790/1251] eta 0:02:13 lr 0.000871 time 0.2861 (0.2897) loss 2.9712 (3.8130) grad_norm 1.1564 (1.1490) [2021-04-15 20:58:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][800/1251] eta 0:02:10 lr 0.000871 time 0.2917 (0.2895) loss 3.0006 (3.8116) grad_norm 1.2991 (1.1506) [2021-04-15 20:58:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][810/1251] eta 0:02:07 lr 0.000871 time 0.2711 (0.2895) loss 3.7457 (3.8125) grad_norm 1.1784 (1.1518) [2021-04-15 20:58:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][820/1251] eta 0:02:04 lr 0.000871 time 0.2789 (0.2894) loss 4.1115 (3.8109) grad_norm 1.1515 (1.1523) [2021-04-15 20:58:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][830/1251] eta 0:02:01 lr 0.000871 time 0.2777 (0.2893) loss 4.1622 (3.8110) grad_norm 1.3894 (1.1517) [2021-04-15 20:58:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][840/1251] eta 0:01:58 lr 0.000871 time 0.2591 (0.2891) loss 3.8219 (3.8107) grad_norm 1.1376 (1.1523) [2021-04-15 20:58:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][850/1251] eta 0:01:55 lr 0.000870 time 0.2755 (0.2891) loss 3.0297 (3.8089) grad_norm 1.0736 (1.1527) [2021-04-15 20:58:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][860/1251] eta 0:01:52 lr 0.000870 time 0.2678 (0.2889) loss 3.6052 (3.8087) grad_norm 0.9972 (1.1526) [2021-04-15 20:58:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][870/1251] eta 0:01:50 lr 0.000870 time 0.2824 (0.2888) loss 3.4093 (3.8088) grad_norm 1.1097 (1.1529) [2021-04-15 20:58:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][880/1251] eta 0:01:47 lr 0.000870 time 0.3084 (0.2889) loss 3.9728 (3.8128) grad_norm 1.1161 (1.1535) [2021-04-15 20:58:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][890/1251] eta 0:01:44 lr 0.000870 time 0.2578 (0.2889) loss 3.3345 (3.8133) grad_norm 1.1450 (1.1533) [2021-04-15 20:58:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][900/1251] eta 0:01:41 lr 0.000870 time 0.3212 (0.2888) loss 4.4866 (3.8144) grad_norm 1.0393 (1.1522) [2021-04-15 20:58:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][910/1251] eta 0:01:38 lr 0.000870 time 0.2777 (0.2887) loss 3.6901 (3.8166) grad_norm 1.1371 (1.1524) [2021-04-15 20:58:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][920/1251] eta 0:01:35 lr 0.000870 time 0.2816 (0.2886) loss 3.7912 (3.8148) grad_norm 1.4110 (1.1516) [2021-04-15 20:58:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][930/1251] eta 0:01:32 lr 0.000870 time 0.2746 (0.2888) loss 4.2266 (3.8186) grad_norm 1.0572 (1.1514) [2021-04-15 20:58:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][940/1251] eta 0:01:29 lr 0.000870 time 0.3184 (0.2887) loss 4.4021 (3.8219) grad_norm 1.1616 (1.1520) [2021-04-15 20:58:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][950/1251] eta 0:01:26 lr 0.000870 time 0.2894 (0.2886) loss 4.6747 (3.8231) grad_norm 1.1019 (1.1516) [2021-04-15 20:58:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][960/1251] eta 0:01:23 lr 0.000870 time 0.2802 (0.2886) loss 4.3502 (3.8261) grad_norm 1.2076 (1.1516) [2021-04-15 20:58:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][970/1251] eta 0:01:21 lr 0.000870 time 0.2736 (0.2885) loss 3.2290 (3.8248) grad_norm 0.9991 (1.1510) [2021-04-15 20:58:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][980/1251] eta 0:01:18 lr 0.000870 time 0.2678 (0.2884) loss 2.9295 (3.8227) grad_norm 1.0751 (1.1505) [2021-04-15 20:59:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][990/1251] eta 0:01:15 lr 0.000870 time 0.2678 (0.2885) loss 4.0634 (3.8248) grad_norm 1.0261 (1.1508) [2021-04-15 20:59:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][1000/1251] eta 0:01:12 lr 0.000870 time 0.2626 (0.2884) loss 4.0853 (3.8240) grad_norm 1.1886 (1.1511) [2021-04-15 20:59:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][1010/1251] eta 0:01:09 lr 0.000870 time 0.2907 (0.2884) loss 2.4576 (3.8209) grad_norm 1.1782 (1.1516) [2021-04-15 20:59:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][1020/1251] eta 0:01:06 lr 0.000870 time 0.2948 (0.2884) loss 2.8170 (3.8215) grad_norm 1.0954 (1.1513) [2021-04-15 20:59:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][1030/1251] eta 0:01:03 lr 0.000870 time 0.2639 (0.2883) loss 3.9502 (3.8229) grad_norm 1.1917 (1.1514) [2021-04-15 20:59:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][1040/1251] eta 0:01:00 lr 0.000870 time 0.2754 (0.2882) loss 3.8375 (3.8234) grad_norm 1.3989 (1.1514) [2021-04-15 20:59:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][1050/1251] eta 0:00:57 lr 0.000870 time 0.2701 (0.2881) loss 3.0892 (3.8217) grad_norm 1.2377 (1.1516) [2021-04-15 20:59:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][1060/1251] eta 0:00:55 lr 0.000870 time 0.2544 (0.2882) loss 3.7601 (3.8223) grad_norm 1.3364 (1.1515) [2021-04-15 20:59:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][1070/1251] eta 0:00:52 lr 0.000870 time 0.2888 (0.2881) loss 2.9786 (3.8172) grad_norm 1.2006 (1.1517) [2021-04-15 20:59:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][1080/1251] eta 0:00:49 lr 0.000870 time 0.2555 (0.2880) loss 2.9977 (3.8172) grad_norm 1.2086 (1.1518) [2021-04-15 20:59:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][1090/1251] eta 0:00:46 lr 0.000870 time 0.2853 (0.2880) loss 4.0062 (3.8186) grad_norm 1.3525 (1.1523) [2021-04-15 20:59:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][1100/1251] eta 0:00:43 lr 0.000870 time 0.2665 (0.2878) loss 3.9800 (3.8173) grad_norm 1.1124 (1.1527) [2021-04-15 20:59:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][1110/1251] eta 0:00:40 lr 0.000870 time 0.2891 (0.2878) loss 3.5962 (3.8185) grad_norm 1.0088 (1.1532) [2021-04-15 20:59:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][1120/1251] eta 0:00:37 lr 0.000870 time 0.2611 (0.2877) loss 4.2888 (3.8193) grad_norm 1.2674 (1.1539) [2021-04-15 20:59:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][1130/1251] eta 0:00:34 lr 0.000870 time 0.2925 (0.2876) loss 4.0149 (3.8207) grad_norm 1.0885 (1.1537) [2021-04-15 20:59:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][1140/1251] eta 0:00:31 lr 0.000870 time 0.2567 (0.2875) loss 4.5411 (3.8216) grad_norm 1.1481 (1.1540) [2021-04-15 20:59:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][1150/1251] eta 0:00:29 lr 0.000870 time 0.2951 (0.2875) loss 4.3413 (3.8215) grad_norm 1.1365 (1.1539) [2021-04-15 20:59:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][1160/1251] eta 0:00:26 lr 0.000870 time 0.2817 (0.2876) loss 2.7618 (3.8200) grad_norm 1.0186 (1.1536) [2021-04-15 20:59:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][1170/1251] eta 0:00:23 lr 0.000870 time 0.2868 (0.2875) loss 4.5089 (3.8199) grad_norm 1.2884 (inf) [2021-04-15 20:59:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][1180/1251] eta 0:00:20 lr 0.000870 time 0.2862 (0.2874) loss 2.9615 (3.8177) grad_norm 1.0881 (inf) [2021-04-15 20:59:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][1190/1251] eta 0:00:17 lr 0.000870 time 0.3181 (0.2874) loss 2.7827 (3.8163) grad_norm 1.0221 (inf) [2021-04-15 21:00:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][1200/1251] eta 0:00:14 lr 0.000870 time 0.2868 (0.2874) loss 2.7966 (3.8126) grad_norm 1.3681 (inf) [2021-04-15 21:00:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][1210/1251] eta 0:00:11 lr 0.000869 time 0.2804 (0.2873) loss 3.8308 (3.8139) grad_norm 0.9565 (inf) [2021-04-15 21:00:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][1220/1251] eta 0:00:08 lr 0.000869 time 0.2792 (0.2874) loss 4.1611 (3.8120) grad_norm 1.1297 (inf) [2021-04-15 21:00:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][1230/1251] eta 0:00:06 lr 0.000869 time 0.2640 (0.2873) loss 4.0354 (3.8114) grad_norm 1.1153 (inf) [2021-04-15 21:00:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][1240/1251] eta 0:00:03 lr 0.000869 time 0.2485 (0.2871) loss 4.0151 (3.8110) grad_norm 0.9454 (inf) [2021-04-15 21:00:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [70/300][1250/1251] eta 0:00:00 lr 0.000869 time 0.2485 (0.2868) loss 4.1971 (3.8118) grad_norm 1.0531 (inf) [2021-04-15 21:00:18 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 70 training takes 0:06:01 [2021-04-15 21:00:18 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_70.pth saving...... [2021-04-15 21:00:31 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_70.pth saved !!! [2021-04-15 21:00:32 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.202 (1.202) Loss 1.1835 (1.1835) Acc@1 71.680 (71.680) Acc@5 92.090 (92.090) [2021-04-15 21:00:34 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.168 (0.237) Loss 1.2377 (1.1688) Acc@1 72.754 (72.852) Acc@5 90.039 (91.557) [2021-04-15 21:00:36 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.101 (0.257) Loss 1.2790 (1.1901) Acc@1 69.141 (72.210) Acc@5 90.430 (91.183) [2021-04-15 21:00:38 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.112 (0.230) Loss 1.2464 (1.2014) Acc@1 69.141 (71.777) Acc@5 89.746 (91.028) [2021-04-15 21:00:40 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.223) Loss 1.1829 (1.2038) Acc@1 72.168 (71.739) Acc@5 91.309 (90.951) [2021-04-15 21:00:43 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 71.762 Acc@5 91.034 [2021-04-15 21:00:43 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 71.8% [2021-04-15 21:00:43 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 71.91% [2021-04-15 21:00:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][0/1251] eta 1:59:12 lr 0.000869 time 5.7173 (5.7173) loss 3.5043 (3.5043) grad_norm 1.0518 (1.0518) [2021-04-15 21:00:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][10/1251] eta 0:15:51 lr 0.000869 time 0.2692 (0.7667) loss 2.5249 (3.5404) grad_norm 1.3041 (1.1685) [2021-04-15 21:00:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][20/1251] eta 0:11:01 lr 0.000869 time 0.2915 (0.5370) loss 4.5738 (3.8337) grad_norm 0.9627 (1.1424) [2021-04-15 21:00:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][30/1251] eta 0:09:14 lr 0.000869 time 0.2561 (0.4542) loss 3.7322 (3.8208) grad_norm 1.3279 (1.1372) [2021-04-15 21:01:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][40/1251] eta 0:08:20 lr 0.000869 time 0.2913 (0.4130) loss 3.6962 (3.7553) grad_norm 1.0780 (1.1586) [2021-04-15 21:01:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][50/1251] eta 0:07:44 lr 0.000869 time 0.2813 (0.3865) loss 3.1768 (3.6907) grad_norm 1.0526 (1.1611) [2021-04-15 21:01:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][60/1251] eta 0:07:20 lr 0.000869 time 0.2984 (0.3700) loss 3.9040 (3.7232) grad_norm 1.2136 (1.1581) [2021-04-15 21:01:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][70/1251] eta 0:07:02 lr 0.000869 time 0.2798 (0.3580) loss 4.3221 (3.7260) grad_norm 1.3388 (1.1567) [2021-04-15 21:01:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][80/1251] eta 0:06:49 lr 0.000869 time 0.2910 (0.3501) loss 4.8100 (3.7210) grad_norm 1.0600 (1.1575) [2021-04-15 21:01:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][90/1251] eta 0:06:40 lr 0.000869 time 0.2550 (0.3448) loss 4.3616 (3.7123) grad_norm 1.1194 (1.1579) [2021-04-15 21:01:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][100/1251] eta 0:06:32 lr 0.000869 time 0.3011 (0.3409) loss 3.1771 (3.7490) grad_norm 1.1419 (1.1507) [2021-04-15 21:01:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][110/1251] eta 0:06:22 lr 0.000869 time 0.2789 (0.3354) loss 4.2563 (3.7726) grad_norm 1.0151 (1.1488) [2021-04-15 21:01:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][120/1251] eta 0:06:14 lr 0.000869 time 0.2452 (0.3307) loss 2.5446 (3.7535) grad_norm 1.1509 (1.1428) [2021-04-15 21:01:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][130/1251] eta 0:06:07 lr 0.000869 time 0.2692 (0.3279) loss 3.2373 (3.7622) grad_norm 1.1321 (1.1407) [2021-04-15 21:01:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][140/1251] eta 0:06:02 lr 0.000869 time 0.2831 (0.3259) loss 4.4524 (3.7733) grad_norm 1.0756 (1.1320) [2021-04-15 21:01:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][150/1251] eta 0:05:55 lr 0.000869 time 0.3056 (0.3230) loss 3.3034 (3.7927) grad_norm 1.1257 (1.1304) [2021-04-15 21:01:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][160/1251] eta 0:05:49 lr 0.000869 time 0.2720 (0.3205) loss 4.0503 (3.7756) grad_norm 1.0358 (1.1309) [2021-04-15 21:01:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][170/1251] eta 0:05:43 lr 0.000869 time 0.2690 (0.3182) loss 3.8082 (3.7702) grad_norm 1.2676 (1.1301) [2021-04-15 21:01:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][180/1251] eta 0:05:38 lr 0.000869 time 0.2660 (0.3164) loss 3.3641 (3.7760) grad_norm 1.3776 (1.1301) [2021-04-15 21:01:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][190/1251] eta 0:05:34 lr 0.000869 time 0.2595 (0.3153) loss 4.0347 (3.7841) grad_norm 1.0524 (1.1282) [2021-04-15 21:01:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][200/1251] eta 0:05:29 lr 0.000869 time 0.2901 (0.3135) loss 2.7862 (3.7711) grad_norm 1.1281 (1.1264) [2021-04-15 21:01:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][210/1251] eta 0:05:24 lr 0.000869 time 0.2779 (0.3118) loss 3.0511 (3.7811) grad_norm 1.1068 (1.1277) [2021-04-15 21:01:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][220/1251] eta 0:05:20 lr 0.000869 time 0.2799 (0.3110) loss 4.5009 (3.7831) grad_norm 1.0996 (1.1262) [2021-04-15 21:01:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][230/1251] eta 0:05:15 lr 0.000869 time 0.2519 (0.3094) loss 4.1233 (3.7745) grad_norm 0.9896 (1.1253) [2021-04-15 21:01:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][240/1251] eta 0:05:11 lr 0.000869 time 0.2859 (0.3082) loss 3.5259 (3.7834) grad_norm 1.1827 (1.1252) [2021-04-15 21:02:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][250/1251] eta 0:05:07 lr 0.000869 time 0.2988 (0.3071) loss 3.7757 (3.7799) grad_norm 1.0857 (1.1269) [2021-04-15 21:02:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][260/1251] eta 0:05:03 lr 0.000869 time 0.2478 (0.3060) loss 4.1390 (3.7782) grad_norm 1.1125 (1.1284) [2021-04-15 21:02:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][270/1251] eta 0:04:59 lr 0.000869 time 0.2753 (0.3051) loss 3.1395 (3.7739) grad_norm 1.0433 (1.1297) [2021-04-15 21:02:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][280/1251] eta 0:04:55 lr 0.000869 time 0.2987 (0.3043) loss 4.1267 (3.7863) grad_norm 1.2719 (1.1305) [2021-04-15 21:02:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][290/1251] eta 0:04:51 lr 0.000869 time 0.2880 (0.3036) loss 4.0606 (3.7964) grad_norm 1.3497 (1.1321) [2021-04-15 21:02:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][300/1251] eta 0:04:47 lr 0.000869 time 0.2676 (0.3027) loss 4.2503 (3.7940) grad_norm 1.0604 (1.1315) [2021-04-15 21:02:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][310/1251] eta 0:04:44 lr 0.000868 time 0.2645 (0.3021) loss 2.7108 (3.7921) grad_norm 1.0188 (1.1310) [2021-04-15 21:02:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][320/1251] eta 0:04:40 lr 0.000868 time 0.2987 (0.3016) loss 3.8495 (3.7969) grad_norm 1.2021 (1.1322) [2021-04-15 21:02:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][330/1251] eta 0:04:37 lr 0.000868 time 0.2974 (0.3009) loss 4.3946 (3.8052) grad_norm 1.0922 (1.1333) [2021-04-15 21:02:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][340/1251] eta 0:04:33 lr 0.000868 time 0.2725 (0.3002) loss 4.0170 (3.8113) grad_norm 1.2439 (1.1326) [2021-04-15 21:02:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][350/1251] eta 0:04:30 lr 0.000868 time 0.2818 (0.2998) loss 3.9459 (3.8127) grad_norm 0.9842 (1.1323) [2021-04-15 21:02:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][360/1251] eta 0:04:27 lr 0.000868 time 0.2557 (0.3001) loss 2.5455 (3.8152) grad_norm 1.0776 (1.1325) [2021-04-15 21:02:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][370/1251] eta 0:04:23 lr 0.000868 time 0.2889 (0.2996) loss 3.8560 (3.8151) grad_norm 1.0216 (1.1327) [2021-04-15 21:02:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][380/1251] eta 0:04:20 lr 0.000868 time 0.2872 (0.2991) loss 3.5747 (3.8153) grad_norm 1.1682 (1.1325) [2021-04-15 21:02:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][390/1251] eta 0:04:16 lr 0.000868 time 0.2733 (0.2985) loss 3.4919 (3.8146) grad_norm 1.0771 (1.1330) [2021-04-15 21:02:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][400/1251] eta 0:04:13 lr 0.000868 time 0.2704 (0.2981) loss 4.5014 (3.8196) grad_norm 1.1286 (1.1331) [2021-04-15 21:02:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][410/1251] eta 0:04:10 lr 0.000868 time 0.3124 (0.2977) loss 4.7096 (3.8149) grad_norm 1.0890 (1.1329) [2021-04-15 21:02:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][420/1251] eta 0:04:07 lr 0.000868 time 0.2971 (0.2975) loss 4.1308 (3.8106) grad_norm 0.9627 (1.1333) [2021-04-15 21:02:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][430/1251] eta 0:04:03 lr 0.000868 time 0.2700 (0.2970) loss 3.7602 (3.8142) grad_norm 1.3646 (1.1345) [2021-04-15 21:02:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][440/1251] eta 0:04:00 lr 0.000868 time 0.2611 (0.2966) loss 3.7710 (3.8101) grad_norm 1.0245 (1.1360) [2021-04-15 21:02:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][450/1251] eta 0:03:57 lr 0.000868 time 0.2746 (0.2962) loss 3.0602 (3.8134) grad_norm 1.0363 (1.1353) [2021-04-15 21:02:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][460/1251] eta 0:03:54 lr 0.000868 time 0.2628 (0.2959) loss 4.4201 (3.8158) grad_norm 1.2259 (1.1361) [2021-04-15 21:03:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][470/1251] eta 0:03:50 lr 0.000868 time 0.2978 (0.2956) loss 4.3409 (3.8154) grad_norm 1.2367 (1.1371) [2021-04-15 21:03:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][480/1251] eta 0:03:47 lr 0.000868 time 0.2752 (0.2952) loss 4.0622 (3.8218) grad_norm 1.0324 (1.1381) [2021-04-15 21:03:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][490/1251] eta 0:03:44 lr 0.000868 time 0.2661 (0.2951) loss 4.3312 (3.8192) grad_norm 1.2789 (1.1385) [2021-04-15 21:03:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][500/1251] eta 0:03:41 lr 0.000868 time 0.2772 (0.2947) loss 4.1370 (3.8192) grad_norm 1.3898 (1.1402) [2021-04-15 21:03:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][510/1251] eta 0:03:38 lr 0.000868 time 0.2696 (0.2944) loss 4.0229 (3.8162) grad_norm 1.2747 (1.1425) [2021-04-15 21:03:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][520/1251] eta 0:03:34 lr 0.000868 time 0.2749 (0.2941) loss 3.2579 (3.8154) grad_norm 1.0469 (1.1428) [2021-04-15 21:03:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][530/1251] eta 0:03:31 lr 0.000868 time 0.2804 (0.2938) loss 2.9158 (3.8139) grad_norm 1.0910 (1.1424) [2021-04-15 21:03:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][540/1251] eta 0:03:28 lr 0.000868 time 0.2705 (0.2937) loss 4.4756 (3.8190) grad_norm 1.3057 (1.1431) [2021-04-15 21:03:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][550/1251] eta 0:03:25 lr 0.000868 time 0.2802 (0.2935) loss 3.8295 (3.8244) grad_norm 1.3272 (1.1434) [2021-04-15 21:03:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][560/1251] eta 0:03:22 lr 0.000868 time 0.3049 (0.2932) loss 2.5702 (3.8250) grad_norm 1.2981 (1.1453) [2021-04-15 21:03:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][570/1251] eta 0:03:19 lr 0.000868 time 0.4871 (0.2936) loss 3.6419 (3.8230) grad_norm 1.0779 (1.1448) [2021-04-15 21:03:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][580/1251] eta 0:03:16 lr 0.000868 time 0.2672 (0.2933) loss 4.4344 (3.8256) grad_norm 1.0486 (1.1448) [2021-04-15 21:03:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][590/1251] eta 0:03:13 lr 0.000868 time 0.2520 (0.2931) loss 4.0174 (3.8218) grad_norm 1.1332 (1.1446) [2021-04-15 21:03:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][600/1251] eta 0:03:10 lr 0.000868 time 0.2782 (0.2928) loss 3.8859 (3.8227) grad_norm 1.0837 (1.1426) [2021-04-15 21:03:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][610/1251] eta 0:03:07 lr 0.000868 time 0.2883 (0.2926) loss 3.3697 (3.8263) grad_norm 1.0795 (1.1417) [2021-04-15 21:03:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][620/1251] eta 0:03:04 lr 0.000868 time 0.2695 (0.2924) loss 3.9288 (3.8282) grad_norm 0.9767 (1.1407) [2021-04-15 21:03:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][630/1251] eta 0:03:01 lr 0.000868 time 0.2916 (0.2923) loss 4.1436 (3.8297) grad_norm 1.1603 (1.1411) [2021-04-15 21:03:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][640/1251] eta 0:02:58 lr 0.000868 time 0.2780 (0.2920) loss 4.2058 (3.8288) grad_norm 1.0802 (1.1414) [2021-04-15 21:03:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][650/1251] eta 0:02:55 lr 0.000868 time 0.4058 (0.2920) loss 3.6293 (3.8324) grad_norm 1.2545 (1.1422) [2021-04-15 21:03:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][660/1251] eta 0:02:52 lr 0.000868 time 0.2783 (0.2918) loss 4.0837 (3.8330) grad_norm 1.1483 (1.1419) [2021-04-15 21:03:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][670/1251] eta 0:02:49 lr 0.000867 time 0.2749 (0.2917) loss 4.1769 (3.8348) grad_norm 1.2150 (1.1426) [2021-04-15 21:04:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][680/1251] eta 0:02:46 lr 0.000867 time 0.2990 (0.2918) loss 3.6923 (3.8367) grad_norm 1.2084 (1.1435) [2021-04-15 21:04:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][690/1251] eta 0:02:43 lr 0.000867 time 0.2771 (0.2916) loss 3.8738 (3.8355) grad_norm 1.0336 (1.1441) [2021-04-15 21:04:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][700/1251] eta 0:02:40 lr 0.000867 time 0.2734 (0.2914) loss 3.2416 (3.8322) grad_norm 1.1925 (1.1447) [2021-04-15 21:04:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][710/1251] eta 0:02:37 lr 0.000867 time 0.2968 (0.2914) loss 3.7905 (3.8338) grad_norm 1.1452 (1.1446) [2021-04-15 21:04:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][720/1251] eta 0:02:34 lr 0.000867 time 0.2867 (0.2915) loss 4.3108 (3.8307) grad_norm 1.3202 (1.1449) [2021-04-15 21:04:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][730/1251] eta 0:02:31 lr 0.000867 time 0.2986 (0.2913) loss 2.5772 (3.8328) grad_norm 1.1295 (1.1460) [2021-04-15 21:04:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][740/1251] eta 0:02:28 lr 0.000867 time 0.2892 (0.2912) loss 3.9873 (3.8253) grad_norm 1.3583 (1.1465) [2021-04-15 21:04:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][750/1251] eta 0:02:25 lr 0.000867 time 0.2535 (0.2910) loss 4.5131 (3.8247) grad_norm 1.0529 (1.1457) [2021-04-15 21:04:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][760/1251] eta 0:02:22 lr 0.000867 time 0.2718 (0.2908) loss 3.7756 (3.8235) grad_norm 1.1026 (1.1458) [2021-04-15 21:04:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][770/1251] eta 0:02:19 lr 0.000867 time 0.2895 (0.2908) loss 4.6080 (3.8246) grad_norm 1.0876 (1.1459) [2021-04-15 21:04:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][780/1251] eta 0:02:16 lr 0.000867 time 0.2757 (0.2907) loss 3.7180 (3.8219) grad_norm 1.1392 (1.1452) [2021-04-15 21:04:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][790/1251] eta 0:02:14 lr 0.000867 time 0.2786 (0.2907) loss 4.0225 (3.8200) grad_norm 1.0691 (1.1453) [2021-04-15 21:04:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][800/1251] eta 0:02:11 lr 0.000867 time 0.2545 (0.2906) loss 3.9534 (3.8229) grad_norm 1.0244 (1.1455) [2021-04-15 21:04:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][810/1251] eta 0:02:08 lr 0.000867 time 0.2837 (0.2904) loss 3.5248 (3.8256) grad_norm 1.4992 (1.1458) [2021-04-15 21:04:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][820/1251] eta 0:02:05 lr 0.000867 time 0.2819 (0.2903) loss 3.5868 (3.8233) grad_norm 1.0523 (1.1451) [2021-04-15 21:04:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][830/1251] eta 0:02:02 lr 0.000867 time 0.2735 (0.2902) loss 4.2192 (3.8213) grad_norm 1.3552 (1.1449) [2021-04-15 21:04:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][840/1251] eta 0:01:59 lr 0.000867 time 0.3052 (0.2900) loss 4.0286 (3.8198) grad_norm 1.0352 (1.1447) [2021-04-15 21:04:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][850/1251] eta 0:01:56 lr 0.000867 time 0.2935 (0.2899) loss 4.0450 (3.8194) grad_norm 1.0199 (1.1447) [2021-04-15 21:04:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][860/1251] eta 0:01:53 lr 0.000867 time 0.2723 (0.2897) loss 3.8478 (3.8246) grad_norm 0.9776 (1.1440) [2021-04-15 21:04:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][870/1251] eta 0:01:50 lr 0.000867 time 0.2868 (0.2896) loss 3.9721 (3.8239) grad_norm 1.1756 (1.1436) [2021-04-15 21:04:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][880/1251] eta 0:01:47 lr 0.000867 time 0.2607 (0.2895) loss 3.4318 (3.8210) grad_norm 1.4407 (1.1446) [2021-04-15 21:05:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][890/1251] eta 0:01:44 lr 0.000867 time 0.3002 (0.2894) loss 3.4597 (3.8249) grad_norm 1.2241 (1.1467) [2021-04-15 21:05:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][900/1251] eta 0:01:41 lr 0.000867 time 0.2574 (0.2893) loss 4.3537 (3.8276) grad_norm 1.0976 (1.1464) [2021-04-15 21:05:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][910/1251] eta 0:01:38 lr 0.000867 time 0.2860 (0.2892) loss 4.2887 (3.8311) grad_norm 1.3525 (1.1456) [2021-04-15 21:05:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][920/1251] eta 0:01:35 lr 0.000867 time 0.2683 (0.2891) loss 3.4977 (3.8308) grad_norm 1.1368 (1.1455) [2021-04-15 21:05:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][930/1251] eta 0:01:32 lr 0.000867 time 0.3076 (0.2892) loss 4.1141 (3.8310) grad_norm 1.6282 (1.1455) [2021-04-15 21:05:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][940/1251] eta 0:01:29 lr 0.000867 time 0.2665 (0.2892) loss 3.8610 (3.8295) grad_norm 1.2808 (1.1466) [2021-04-15 21:05:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][950/1251] eta 0:01:27 lr 0.000867 time 0.3029 (0.2891) loss 4.1239 (3.8298) grad_norm 1.2070 (1.1461) [2021-04-15 21:05:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][960/1251] eta 0:01:24 lr 0.000867 time 0.2713 (0.2890) loss 4.5765 (3.8312) grad_norm 1.1721 (1.1464) [2021-04-15 21:05:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][970/1251] eta 0:01:21 lr 0.000867 time 0.2931 (0.2890) loss 3.7362 (3.8332) grad_norm 1.1789 (1.1458) [2021-04-15 21:05:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][980/1251] eta 0:01:18 lr 0.000867 time 0.2992 (0.2888) loss 3.3886 (3.8344) grad_norm 1.1264 (1.1458) [2021-04-15 21:05:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][990/1251] eta 0:01:15 lr 0.000867 time 0.2865 (0.2889) loss 4.2271 (3.8371) grad_norm 1.2871 (1.1464) [2021-04-15 21:05:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][1000/1251] eta 0:01:12 lr 0.000867 time 0.2616 (0.2888) loss 4.3247 (3.8393) grad_norm 1.1859 (1.1464) [2021-04-15 21:05:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][1010/1251] eta 0:01:09 lr 0.000867 time 0.2704 (0.2887) loss 4.0542 (3.8380) grad_norm 1.1353 (1.1465) [2021-04-15 21:05:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][1020/1251] eta 0:01:06 lr 0.000866 time 0.2851 (0.2886) loss 3.8597 (3.8350) grad_norm 1.1507 (1.1467) [2021-04-15 21:05:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][1030/1251] eta 0:01:03 lr 0.000866 time 0.2952 (0.2886) loss 4.1069 (3.8323) grad_norm 1.1121 (1.1466) [2021-04-15 21:05:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][1040/1251] eta 0:01:00 lr 0.000866 time 0.2705 (0.2885) loss 3.9338 (3.8308) grad_norm 1.1552 (1.1467) [2021-04-15 21:05:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][1050/1251] eta 0:00:57 lr 0.000866 time 0.3215 (0.2884) loss 3.1260 (3.8280) grad_norm 1.2314 (1.1470) [2021-04-15 21:05:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][1060/1251] eta 0:00:55 lr 0.000866 time 0.2844 (0.2883) loss 3.9689 (3.8271) grad_norm 1.0811 (1.1476) [2021-04-15 21:05:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][1070/1251] eta 0:00:52 lr 0.000866 time 0.3105 (0.2884) loss 1.9878 (3.8281) grad_norm 1.0953 (1.1479) [2021-04-15 21:05:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][1080/1251] eta 0:00:49 lr 0.000866 time 0.2850 (0.2883) loss 3.8160 (3.8282) grad_norm 1.2739 (1.1481) [2021-04-15 21:05:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][1090/1251] eta 0:00:46 lr 0.000866 time 0.2865 (0.2882) loss 3.4669 (3.8267) grad_norm 1.0205 (1.1481) [2021-04-15 21:06:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][1100/1251] eta 0:00:43 lr 0.000866 time 0.2728 (0.2882) loss 3.1761 (3.8272) grad_norm 1.0756 (1.1476) [2021-04-15 21:06:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][1110/1251] eta 0:00:40 lr 0.000866 time 0.2623 (0.2881) loss 3.3643 (3.8268) grad_norm 1.1743 (1.1476) [2021-04-15 21:06:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][1120/1251] eta 0:00:37 lr 0.000866 time 0.2684 (0.2881) loss 3.2019 (3.8266) grad_norm 1.0453 (1.1473) [2021-04-15 21:06:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][1130/1251] eta 0:00:34 lr 0.000866 time 0.2534 (0.2880) loss 3.9396 (3.8270) grad_norm 1.1181 (1.1478) [2021-04-15 21:06:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][1140/1251] eta 0:00:31 lr 0.000866 time 0.2474 (0.2879) loss 2.4743 (3.8237) grad_norm 1.0567 (1.1476) [2021-04-15 21:06:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][1150/1251] eta 0:00:29 lr 0.000866 time 0.2789 (0.2881) loss 3.8372 (3.8227) grad_norm 1.0826 (1.1475) [2021-04-15 21:06:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][1160/1251] eta 0:00:26 lr 0.000866 time 0.2672 (0.2880) loss 4.2340 (3.8229) grad_norm 1.1679 (1.1482) [2021-04-15 21:06:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][1170/1251] eta 0:00:23 lr 0.000866 time 0.2847 (0.2879) loss 3.5624 (3.8230) grad_norm 1.0535 (1.1487) [2021-04-15 21:06:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][1180/1251] eta 0:00:20 lr 0.000866 time 0.2693 (0.2878) loss 3.8873 (3.8232) grad_norm 1.0944 (1.1483) [2021-04-15 21:06:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][1190/1251] eta 0:00:17 lr 0.000866 time 0.3056 (0.2877) loss 4.4480 (3.8257) grad_norm 1.1826 (1.1481) [2021-04-15 21:06:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][1200/1251] eta 0:00:14 lr 0.000866 time 0.2817 (0.2877) loss 2.8824 (3.8245) grad_norm 1.0293 (1.1482) [2021-04-15 21:06:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][1210/1251] eta 0:00:11 lr 0.000866 time 0.2773 (0.2876) loss 4.2368 (3.8249) grad_norm 1.2018 (1.1487) [2021-04-15 21:06:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][1220/1251] eta 0:00:08 lr 0.000866 time 0.2625 (0.2875) loss 4.2734 (3.8267) grad_norm 1.2558 (1.1488) [2021-04-15 21:06:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][1230/1251] eta 0:00:06 lr 0.000866 time 0.2575 (0.2875) loss 3.8555 (3.8240) grad_norm 1.2186 (1.1488) [2021-04-15 21:06:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][1240/1251] eta 0:00:03 lr 0.000866 time 0.4243 (0.2875) loss 3.9325 (3.8235) grad_norm 1.0851 (1.1486) [2021-04-15 21:06:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [71/300][1250/1251] eta 0:00:00 lr 0.000866 time 0.2492 (0.2872) loss 4.1220 (3.8232) grad_norm 0.9755 (1.1483) [2021-04-15 21:06:44 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 71 training takes 0:06:01 [2021-04-15 21:06:44 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_71.pth saving...... [2021-04-15 21:06:58 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_71.pth saved !!! [2021-04-15 21:07:00 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.089 (1.089) Loss 1.1969 (1.1969) Acc@1 72.949 (72.949) Acc@5 90.625 (90.625) [2021-04-15 21:07:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.182 (0.283) Loss 1.1468 (1.2204) Acc@1 73.926 (71.884) Acc@5 91.895 (91.246) [2021-04-15 21:07:04 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 1.050 (0.252) Loss 1.2398 (1.2125) Acc@1 73.438 (72.163) Acc@5 90.332 (91.295) [2021-04-15 21:07:06 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.164 (0.248) Loss 1.2338 (1.2208) Acc@1 72.656 (71.998) Acc@5 90.625 (91.224) [2021-04-15 21:07:08 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.141 (0.222) Loss 1.1741 (1.2248) Acc@1 72.559 (71.877) Acc@5 91.602 (91.123) [2021-04-15 21:07:11 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 71.992 Acc@5 91.170 [2021-04-15 21:07:11 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 72.0% [2021-04-15 21:07:11 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 71.99% [2021-04-15 21:07:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][0/1251] eta 1:08:18 lr 0.000866 time 3.2758 (3.2758) loss 2.8441 (2.8441) grad_norm 1.0545 (1.0545) [2021-04-15 21:07:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][10/1251] eta 0:11:19 lr 0.000866 time 0.2754 (0.5472) loss 4.2520 (4.0999) grad_norm 1.2505 (1.0971) [2021-04-15 21:07:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][20/1251] eta 0:08:46 lr 0.000866 time 0.2855 (0.4280) loss 3.5383 (3.7092) grad_norm 1.3625 (1.1637) [2021-04-15 21:07:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][30/1251] eta 0:07:42 lr 0.000866 time 0.2629 (0.3789) loss 2.8515 (3.6948) grad_norm 1.3408 (1.1859) [2021-04-15 21:07:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][40/1251] eta 0:07:08 lr 0.000866 time 0.2835 (0.3538) loss 4.2912 (3.7090) grad_norm 1.4365 (1.1863) [2021-04-15 21:07:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][50/1251] eta 0:06:48 lr 0.000866 time 0.2908 (0.3401) loss 4.1967 (3.6962) grad_norm 0.9991 (1.1742) [2021-04-15 21:07:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][60/1251] eta 0:06:32 lr 0.000866 time 0.2527 (0.3293) loss 2.7694 (3.7067) grad_norm 1.1526 (1.1616) [2021-04-15 21:07:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][70/1251] eta 0:06:20 lr 0.000866 time 0.3040 (0.3223) loss 3.8935 (3.7063) grad_norm 1.3614 (1.1635) [2021-04-15 21:07:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][80/1251] eta 0:06:11 lr 0.000866 time 0.2647 (0.3174) loss 4.2311 (3.7339) grad_norm 1.1242 (1.1655) [2021-04-15 21:07:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][90/1251] eta 0:06:02 lr 0.000866 time 0.2616 (0.3126) loss 4.2903 (3.7671) grad_norm 1.1947 (1.1608) [2021-04-15 21:07:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][100/1251] eta 0:05:56 lr 0.000866 time 0.2428 (0.3095) loss 3.8887 (3.7741) grad_norm 1.1839 (1.1609) [2021-04-15 21:07:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][110/1251] eta 0:05:50 lr 0.000866 time 0.2626 (0.3072) loss 2.5704 (3.7908) grad_norm 1.0630 (1.1531) [2021-04-15 21:07:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][120/1251] eta 0:05:44 lr 0.000865 time 0.2963 (0.3050) loss 3.9861 (3.7645) grad_norm 1.2291 (1.1494) [2021-04-15 21:07:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][130/1251] eta 0:05:40 lr 0.000865 time 0.2755 (0.3033) loss 4.1458 (3.7763) grad_norm 1.1468 (1.1471) [2021-04-15 21:07:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][140/1251] eta 0:05:35 lr 0.000865 time 0.2628 (0.3024) loss 3.5178 (3.8087) grad_norm 1.1055 (1.1497) [2021-04-15 21:07:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][150/1251] eta 0:05:32 lr 0.000865 time 0.2704 (0.3019) loss 3.7668 (3.8175) grad_norm 1.1062 (1.1480) [2021-04-15 21:08:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][160/1251] eta 0:05:27 lr 0.000865 time 0.2691 (0.3006) loss 3.8215 (3.8129) grad_norm 1.1564 (1.1466) [2021-04-15 21:08:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][170/1251] eta 0:05:23 lr 0.000865 time 0.2882 (0.2994) loss 4.2094 (3.8161) grad_norm 1.2084 (1.1480) [2021-04-15 21:08:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][180/1251] eta 0:05:19 lr 0.000865 time 0.2952 (0.2986) loss 2.7276 (3.8184) grad_norm 0.9594 (1.1505) [2021-04-15 21:08:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][190/1251] eta 0:05:15 lr 0.000865 time 0.2771 (0.2975) loss 2.9969 (3.8150) grad_norm 1.1764 (1.1524) [2021-04-15 21:08:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][200/1251] eta 0:05:12 lr 0.000865 time 0.2803 (0.2972) loss 3.1591 (3.8132) grad_norm 1.0157 (1.1527) [2021-04-15 21:08:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][210/1251] eta 0:05:08 lr 0.000865 time 0.2986 (0.2964) loss 4.6489 (3.7939) grad_norm 1.1260 (1.1527) [2021-04-15 21:08:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][220/1251] eta 0:05:04 lr 0.000865 time 0.2910 (0.2956) loss 3.7052 (3.7857) grad_norm 1.2154 (1.1499) [2021-04-15 21:08:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][230/1251] eta 0:05:01 lr 0.000865 time 0.2814 (0.2948) loss 4.1953 (3.7845) grad_norm 1.2569 (1.1529) [2021-04-15 21:08:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][240/1251] eta 0:04:57 lr 0.000865 time 0.2693 (0.2944) loss 4.3295 (3.7839) grad_norm 1.3003 (1.1554) [2021-04-15 21:08:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][250/1251] eta 0:04:53 lr 0.000865 time 0.2790 (0.2937) loss 4.2549 (3.7839) grad_norm 1.0784 (1.1573) [2021-04-15 21:08:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][260/1251] eta 0:04:50 lr 0.000865 time 0.2640 (0.2932) loss 4.0194 (3.7822) grad_norm 1.1084 (1.1568) [2021-04-15 21:08:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][270/1251] eta 0:04:47 lr 0.000865 time 0.2917 (0.2926) loss 3.7428 (3.7828) grad_norm 0.9096 (1.1575) [2021-04-15 21:08:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][280/1251] eta 0:04:43 lr 0.000865 time 0.3151 (0.2923) loss 4.1186 (3.7822) grad_norm 1.1949 (1.1596) [2021-04-15 21:08:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][290/1251] eta 0:04:40 lr 0.000865 time 0.2905 (0.2923) loss 2.6832 (3.7812) grad_norm 1.0066 (1.1586) [2021-04-15 21:08:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][300/1251] eta 0:04:37 lr 0.000865 time 0.2701 (0.2918) loss 2.8299 (3.7815) grad_norm 0.9935 (1.1579) [2021-04-15 21:08:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][310/1251] eta 0:04:34 lr 0.000865 time 0.2676 (0.2915) loss 4.4959 (3.7863) grad_norm 1.2670 (1.1583) [2021-04-15 21:08:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][320/1251] eta 0:04:31 lr 0.000865 time 0.2676 (0.2913) loss 4.8019 (3.7959) grad_norm 1.3674 (1.1602) [2021-04-15 21:08:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][330/1251] eta 0:04:27 lr 0.000865 time 0.2612 (0.2909) loss 4.1566 (3.7963) grad_norm 1.1624 (1.1589) [2021-04-15 21:08:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][340/1251] eta 0:04:24 lr 0.000865 time 0.2980 (0.2906) loss 2.6310 (3.7867) grad_norm 1.1319 (1.1570) [2021-04-15 21:08:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][350/1251] eta 0:04:21 lr 0.000865 time 0.3070 (0.2907) loss 4.6368 (3.7885) grad_norm 1.1133 (1.1542) [2021-04-15 21:08:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][360/1251] eta 0:04:19 lr 0.000865 time 0.2824 (0.2909) loss 4.3515 (3.7917) grad_norm 1.2952 (1.1527) [2021-04-15 21:08:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][370/1251] eta 0:04:16 lr 0.000865 time 0.2645 (0.2910) loss 2.5766 (3.7928) grad_norm 1.2744 (1.1528) [2021-04-15 21:09:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][380/1251] eta 0:04:13 lr 0.000865 time 0.2779 (0.2905) loss 4.3387 (3.7931) grad_norm 1.2695 (1.1504) [2021-04-15 21:09:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][390/1251] eta 0:04:09 lr 0.000865 time 0.2427 (0.2903) loss 4.1953 (3.8000) grad_norm 1.0418 (1.1486) [2021-04-15 21:09:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][400/1251] eta 0:04:06 lr 0.000865 time 0.3034 (0.2901) loss 4.3357 (3.8078) grad_norm 1.0875 (1.1478) [2021-04-15 21:09:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][410/1251] eta 0:04:03 lr 0.000865 time 0.2719 (0.2898) loss 3.9285 (3.8049) grad_norm 1.3533 (1.1477) [2021-04-15 21:09:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][420/1251] eta 0:04:00 lr 0.000865 time 0.2560 (0.2896) loss 4.8339 (3.8054) grad_norm 1.2775 (1.1485) [2021-04-15 21:09:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][430/1251] eta 0:03:57 lr 0.000865 time 0.2572 (0.2896) loss 3.1966 (3.8050) grad_norm 1.1324 (1.1489) [2021-04-15 21:09:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][440/1251] eta 0:03:54 lr 0.000865 time 0.2555 (0.2894) loss 3.8730 (3.8064) grad_norm 1.1640 (1.1492) [2021-04-15 21:09:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][450/1251] eta 0:03:51 lr 0.000865 time 0.2868 (0.2895) loss 4.0759 (3.8156) grad_norm 1.3093 (1.1504) [2021-04-15 21:09:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][460/1251] eta 0:03:48 lr 0.000865 time 0.2787 (0.2893) loss 4.0710 (3.8135) grad_norm 1.2933 (1.1565) [2021-04-15 21:09:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][470/1251] eta 0:03:46 lr 0.000865 time 0.3024 (0.2894) loss 4.1133 (3.8181) grad_norm 1.0285 (1.1578) [2021-04-15 21:09:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][480/1251] eta 0:03:43 lr 0.000864 time 0.2981 (0.2893) loss 3.9463 (3.8147) grad_norm 1.0647 (1.1572) [2021-04-15 21:09:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][490/1251] eta 0:03:39 lr 0.000864 time 0.2723 (0.2890) loss 4.0535 (3.8099) grad_norm 1.0072 (1.1586) [2021-04-15 21:09:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][500/1251] eta 0:03:36 lr 0.000864 time 0.2800 (0.2889) loss 4.0551 (3.8061) grad_norm 1.1783 (1.1588) [2021-04-15 21:09:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][510/1251] eta 0:03:33 lr 0.000864 time 0.2901 (0.2888) loss 2.7189 (3.8087) grad_norm 1.2448 (1.1581) [2021-04-15 21:09:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][520/1251] eta 0:03:30 lr 0.000864 time 0.2822 (0.2886) loss 3.9705 (3.8105) grad_norm 1.1816 (1.1568) [2021-04-15 21:09:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][530/1251] eta 0:03:27 lr 0.000864 time 0.2811 (0.2884) loss 4.3079 (3.8098) grad_norm 1.3960 (1.1574) [2021-04-15 21:09:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][540/1251] eta 0:03:25 lr 0.000864 time 0.2700 (0.2884) loss 4.4758 (3.8117) grad_norm 1.0294 (1.1577) [2021-04-15 21:09:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][550/1251] eta 0:03:22 lr 0.000864 time 0.2461 (0.2883) loss 4.1244 (3.8106) grad_norm 0.9995 (1.1566) [2021-04-15 21:09:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][560/1251] eta 0:03:19 lr 0.000864 time 0.2648 (0.2881) loss 2.6489 (3.8076) grad_norm 1.0834 (1.1563) [2021-04-15 21:09:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][570/1251] eta 0:03:16 lr 0.000864 time 0.2701 (0.2881) loss 4.6382 (3.8069) grad_norm 1.2363 (1.1578) [2021-04-15 21:09:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][580/1251] eta 0:03:13 lr 0.000864 time 0.2651 (0.2882) loss 3.6322 (3.8008) grad_norm 1.0991 (1.1576) [2021-04-15 21:10:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][590/1251] eta 0:03:10 lr 0.000864 time 0.2714 (0.2881) loss 4.1235 (3.8023) grad_norm 1.0886 (1.1584) [2021-04-15 21:10:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][600/1251] eta 0:03:07 lr 0.000864 time 0.3037 (0.2880) loss 2.7691 (3.7991) grad_norm 1.1875 (1.1574) [2021-04-15 21:10:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][610/1251] eta 0:03:04 lr 0.000864 time 0.2723 (0.2878) loss 3.0005 (3.7985) grad_norm 1.1755 (1.1581) [2021-04-15 21:10:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][620/1251] eta 0:03:01 lr 0.000864 time 0.2881 (0.2877) loss 3.1071 (3.7951) grad_norm 1.3827 (1.1611) [2021-04-15 21:10:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][630/1251] eta 0:02:58 lr 0.000864 time 0.3026 (0.2876) loss 2.6427 (3.7979) grad_norm 1.1692 (1.1624) [2021-04-15 21:10:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][640/1251] eta 0:02:55 lr 0.000864 time 0.2881 (0.2874) loss 4.3314 (3.7975) grad_norm 1.1043 (1.1640) [2021-04-15 21:10:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][650/1251] eta 0:02:52 lr 0.000864 time 0.2700 (0.2872) loss 2.9298 (3.8001) grad_norm 1.1683 (1.1650) [2021-04-15 21:10:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][660/1251] eta 0:02:49 lr 0.000864 time 0.2873 (0.2871) loss 4.9511 (3.8031) grad_norm 1.0890 (1.1641) [2021-04-15 21:10:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][670/1251] eta 0:02:46 lr 0.000864 time 0.3052 (0.2872) loss 4.1174 (3.8034) grad_norm 1.0357 (1.1638) [2021-04-15 21:10:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][680/1251] eta 0:02:43 lr 0.000864 time 0.2669 (0.2870) loss 3.2447 (3.8047) grad_norm 1.0960 (1.1629) [2021-04-15 21:10:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][690/1251] eta 0:02:40 lr 0.000864 time 0.2695 (0.2869) loss 4.2698 (3.8056) grad_norm 1.1176 (1.1629) [2021-04-15 21:10:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][700/1251] eta 0:02:38 lr 0.000864 time 0.2716 (0.2868) loss 4.6800 (3.8089) grad_norm 1.2459 (1.1635) [2021-04-15 21:10:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][710/1251] eta 0:02:35 lr 0.000864 time 0.2947 (0.2867) loss 3.0350 (3.8104) grad_norm 1.0923 (1.1632) [2021-04-15 21:10:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][720/1251] eta 0:02:32 lr 0.000864 time 0.2478 (0.2869) loss 3.2760 (3.8102) grad_norm 1.2464 (1.1635) [2021-04-15 21:10:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][730/1251] eta 0:02:29 lr 0.000864 time 0.2889 (0.2868) loss 3.7952 (3.8120) grad_norm 1.3688 (1.1637) [2021-04-15 21:10:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][740/1251] eta 0:02:26 lr 0.000864 time 0.2896 (0.2867) loss 2.9623 (3.8104) grad_norm 1.1606 (1.1634) [2021-04-15 21:10:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][750/1251] eta 0:02:23 lr 0.000864 time 0.2553 (0.2865) loss 2.9985 (3.8076) grad_norm 1.1484 (1.1625) [2021-04-15 21:10:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][760/1251] eta 0:02:20 lr 0.000864 time 0.2797 (0.2864) loss 3.2791 (3.8071) grad_norm 1.1070 (1.1627) [2021-04-15 21:10:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][770/1251] eta 0:02:17 lr 0.000864 time 0.2608 (0.2863) loss 4.5584 (3.8083) grad_norm 1.1597 (1.1633) [2021-04-15 21:10:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][780/1251] eta 0:02:14 lr 0.000864 time 0.2903 (0.2863) loss 4.0666 (3.8123) grad_norm 1.0832 (1.1624) [2021-04-15 21:10:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][790/1251] eta 0:02:11 lr 0.000864 time 0.2966 (0.2863) loss 4.6010 (3.8151) grad_norm 1.0916 (1.1617) [2021-04-15 21:11:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][800/1251] eta 0:02:09 lr 0.000864 time 0.2921 (0.2862) loss 3.5876 (3.8137) grad_norm 1.1170 (1.1615) [2021-04-15 21:11:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][810/1251] eta 0:02:06 lr 0.000864 time 0.2908 (0.2861) loss 3.8859 (3.8164) grad_norm 1.0090 (1.1604) [2021-04-15 21:11:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][820/1251] eta 0:02:03 lr 0.000864 time 0.2946 (0.2860) loss 3.8823 (3.8151) grad_norm 1.1400 (1.1605) [2021-04-15 21:11:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][830/1251] eta 0:02:00 lr 0.000863 time 0.2865 (0.2859) loss 3.4373 (3.8133) grad_norm 1.2299 (1.1604) [2021-04-15 21:11:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][840/1251] eta 0:01:57 lr 0.000863 time 0.2531 (0.2858) loss 4.2607 (3.8130) grad_norm 1.1692 (1.1604) [2021-04-15 21:11:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][850/1251] eta 0:01:54 lr 0.000863 time 0.2714 (0.2857) loss 3.2348 (3.8100) grad_norm 1.6575 (1.1610) [2021-04-15 21:11:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][860/1251] eta 0:01:51 lr 0.000863 time 0.2775 (0.2856) loss 4.1779 (3.8092) grad_norm 1.0989 (1.1602) [2021-04-15 21:11:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][870/1251] eta 0:01:48 lr 0.000863 time 0.2696 (0.2857) loss 2.7958 (3.8072) grad_norm 1.1966 (1.1601) [2021-04-15 21:11:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][880/1251] eta 0:01:45 lr 0.000863 time 0.2542 (0.2856) loss 2.8872 (3.8073) grad_norm 1.2492 (1.1610) [2021-04-15 21:11:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][890/1251] eta 0:01:43 lr 0.000863 time 0.2979 (0.2857) loss 3.7416 (3.8059) grad_norm 1.0645 (1.1617) [2021-04-15 21:11:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][900/1251] eta 0:01:40 lr 0.000863 time 0.2900 (0.2856) loss 4.3147 (3.8069) grad_norm 1.2175 (1.1618) [2021-04-15 21:11:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][910/1251] eta 0:01:37 lr 0.000863 time 0.2897 (0.2855) loss 3.7703 (3.8099) grad_norm 1.0398 (1.1610) [2021-04-15 21:11:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][920/1251] eta 0:01:34 lr 0.000863 time 0.3078 (0.2856) loss 3.2966 (3.8124) grad_norm 1.0220 (1.1619) [2021-04-15 21:11:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][930/1251] eta 0:01:31 lr 0.000863 time 0.2813 (0.2855) loss 3.2180 (3.8142) grad_norm 0.9462 (1.1617) [2021-04-15 21:11:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][940/1251] eta 0:01:28 lr 0.000863 time 0.2652 (0.2856) loss 4.5622 (3.8148) grad_norm 1.0759 (1.1615) [2021-04-15 21:11:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][950/1251] eta 0:01:25 lr 0.000863 time 0.2849 (0.2856) loss 4.6966 (3.8162) grad_norm 1.2610 (1.1616) [2021-04-15 21:11:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][960/1251] eta 0:01:23 lr 0.000863 time 0.2688 (0.2855) loss 2.8149 (3.8186) grad_norm 1.0969 (1.1610) [2021-04-15 21:11:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][970/1251] eta 0:01:20 lr 0.000863 time 0.2748 (0.2855) loss 4.1252 (3.8205) grad_norm 1.1237 (1.1609) [2021-04-15 21:11:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][980/1251] eta 0:01:17 lr 0.000863 time 0.2663 (0.2853) loss 4.4322 (3.8231) grad_norm 1.0297 (1.1602) [2021-04-15 21:11:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][990/1251] eta 0:01:14 lr 0.000863 time 0.2784 (0.2853) loss 3.7453 (3.8218) grad_norm 1.0553 (1.1597) [2021-04-15 21:11:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][1000/1251] eta 0:01:11 lr 0.000863 time 0.2646 (0.2852) loss 4.0849 (3.8220) grad_norm 1.3814 (1.1604) [2021-04-15 21:12:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][1010/1251] eta 0:01:08 lr 0.000863 time 0.2867 (0.2852) loss 3.3326 (3.8232) grad_norm 1.0752 (1.1600) [2021-04-15 21:12:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][1020/1251] eta 0:01:05 lr 0.000863 time 0.2547 (0.2851) loss 3.0066 (3.8213) grad_norm 1.8146 (1.1605) [2021-04-15 21:12:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][1030/1251] eta 0:01:03 lr 0.000863 time 0.2476 (0.2852) loss 2.4357 (3.8208) grad_norm 1.0604 (1.1609) [2021-04-15 21:12:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][1040/1251] eta 0:01:00 lr 0.000863 time 0.2849 (0.2853) loss 3.8924 (3.8218) grad_norm 1.1605 (1.1601) [2021-04-15 21:12:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][1050/1251] eta 0:00:57 lr 0.000863 time 0.2757 (0.2852) loss 3.7967 (3.8205) grad_norm 1.1412 (1.1602) [2021-04-15 21:12:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][1060/1251] eta 0:00:54 lr 0.000863 time 0.2742 (0.2851) loss 2.6116 (3.8189) grad_norm 1.2216 (1.1601) [2021-04-15 21:12:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][1070/1251] eta 0:00:51 lr 0.000863 time 0.2737 (0.2851) loss 3.7093 (3.8194) grad_norm 1.4105 (1.1602) [2021-04-15 21:12:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][1080/1251] eta 0:00:48 lr 0.000863 time 0.2892 (0.2850) loss 4.2503 (3.8209) grad_norm 1.0591 (1.1613) [2021-04-15 21:12:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][1090/1251] eta 0:00:45 lr 0.000863 time 0.2717 (0.2849) loss 3.2875 (3.8213) grad_norm 1.2876 (1.1614) [2021-04-15 21:12:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][1100/1251] eta 0:00:43 lr 0.000863 time 0.2875 (0.2849) loss 4.1918 (3.8222) grad_norm 1.1062 (1.1623) [2021-04-15 21:12:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][1110/1251] eta 0:00:40 lr 0.000863 time 0.2775 (0.2848) loss 4.7101 (3.8241) grad_norm 1.2241 (1.1633) [2021-04-15 21:12:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][1120/1251] eta 0:00:37 lr 0.000863 time 0.3049 (0.2849) loss 3.0647 (3.8211) grad_norm 1.2021 (1.1644) [2021-04-15 21:12:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][1130/1251] eta 0:00:34 lr 0.000863 time 0.2743 (0.2848) loss 3.6809 (3.8195) grad_norm 1.2270 (1.1651) [2021-04-15 21:12:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][1140/1251] eta 0:00:31 lr 0.000863 time 0.2429 (0.2849) loss 3.9619 (3.8205) grad_norm 1.4099 (1.1659) [2021-04-15 21:12:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][1150/1251] eta 0:00:28 lr 0.000863 time 0.2721 (0.2848) loss 4.6255 (3.8220) grad_norm 1.0356 (1.1665) [2021-04-15 21:12:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][1160/1251] eta 0:00:25 lr 0.000863 time 0.2676 (0.2848) loss 4.1575 (3.8220) grad_norm 1.0650 (1.1661) [2021-04-15 21:12:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][1170/1251] eta 0:00:23 lr 0.000863 time 0.2866 (0.2849) loss 3.8098 (3.8201) grad_norm 1.1084 (1.1661) [2021-04-15 21:12:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][1180/1251] eta 0:00:20 lr 0.000862 time 0.2679 (0.2848) loss 3.1657 (3.8184) grad_norm 1.4077 (1.1666) [2021-04-15 21:12:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][1190/1251] eta 0:00:17 lr 0.000862 time 0.2668 (0.2847) loss 4.2471 (3.8176) grad_norm 1.1544 (1.1662) [2021-04-15 21:12:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][1200/1251] eta 0:00:14 lr 0.000862 time 0.2601 (0.2847) loss 4.4754 (3.8178) grad_norm 1.1611 (1.1657) [2021-04-15 21:12:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][1210/1251] eta 0:00:11 lr 0.000862 time 0.2901 (0.2846) loss 3.7009 (3.8147) grad_norm 1.2090 (1.1655) [2021-04-15 21:12:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][1220/1251] eta 0:00:08 lr 0.000862 time 0.2725 (0.2846) loss 3.7284 (3.8122) grad_norm 1.2138 (1.1656) [2021-04-15 21:13:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][1230/1251] eta 0:00:05 lr 0.000862 time 0.2811 (0.2845) loss 4.1797 (3.8124) grad_norm 1.1338 (1.1656) [2021-04-15 21:13:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][1240/1251] eta 0:00:03 lr 0.000862 time 0.2494 (0.2844) loss 4.7643 (3.8133) grad_norm 1.1591 (1.1660) [2021-04-15 21:13:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [72/300][1250/1251] eta 0:00:00 lr 0.000862 time 0.2757 (0.2841) loss 3.6832 (3.8126) grad_norm 1.2222 (1.1663) [2021-04-15 21:13:09 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 72 training takes 0:05:57 [2021-04-15 21:13:09 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_72.pth saving...... [2021-04-15 21:13:17 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_72.pth saved !!! [2021-04-15 21:13:19 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.171 (1.171) Loss 1.2302 (1.2302) Acc@1 71.289 (71.289) Acc@5 91.504 (91.504) [2021-04-15 21:13:20 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.108 (0.219) Loss 1.0769 (1.1759) Acc@1 74.707 (72.514) Acc@5 92.676 (91.602) [2021-04-15 21:13:22 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.106 (0.226) Loss 1.2619 (1.1917) Acc@1 70.020 (71.894) Acc@5 90.430 (91.476) [2021-04-15 21:13:25 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.121 (0.247) Loss 1.1890 (1.1897) Acc@1 72.363 (72.133) Acc@5 92.090 (91.397) [2021-04-15 21:13:27 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.224) Loss 1.1167 (1.1906) Acc@1 73.242 (72.073) Acc@5 93.164 (91.330) [2021-04-15 21:13:30 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 72.126 Acc@5 91.376 [2021-04-15 21:13:30 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 72.1% [2021-04-15 21:13:30 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 72.13% [2021-04-15 21:13:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][0/1251] eta 1:29:21 lr 0.000862 time 4.2861 (4.2861) loss 4.0607 (4.0607) grad_norm 1.1099 (1.1099) [2021-04-15 21:13:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][10/1251] eta 0:13:13 lr 0.000862 time 0.2785 (0.6396) loss 4.0690 (3.5033) grad_norm 1.1675 (1.1426) [2021-04-15 21:13:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][20/1251] eta 0:09:36 lr 0.000862 time 0.2660 (0.4686) loss 3.6082 (3.6527) grad_norm 1.2830 (1.1571) [2021-04-15 21:13:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][30/1251] eta 0:08:18 lr 0.000862 time 0.2764 (0.4081) loss 2.9743 (3.6338) grad_norm 1.2017 (1.1736) [2021-04-15 21:13:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][40/1251] eta 0:07:35 lr 0.000862 time 0.2780 (0.3764) loss 3.0263 (3.6520) grad_norm 1.3049 (1.1893) [2021-04-15 21:13:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][50/1251] eta 0:07:10 lr 0.000862 time 0.2638 (0.3581) loss 3.7253 (3.6596) grad_norm 1.0292 (1.1832) [2021-04-15 21:13:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][60/1251] eta 0:06:55 lr 0.000862 time 0.4491 (0.3488) loss 2.6596 (3.6370) grad_norm 1.1603 (1.1705) [2021-04-15 21:13:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][70/1251] eta 0:06:40 lr 0.000862 time 0.3061 (0.3395) loss 3.6926 (3.6579) grad_norm 1.1365 (1.1717) [2021-04-15 21:13:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][80/1251] eta 0:06:30 lr 0.000862 time 0.2434 (0.3331) loss 4.4242 (3.7070) grad_norm 1.1991 (1.1760) [2021-04-15 21:14:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][90/1251] eta 0:06:21 lr 0.000862 time 0.2862 (0.3289) loss 4.5683 (3.7717) grad_norm 1.4279 (1.2034) [2021-04-15 21:14:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][100/1251] eta 0:06:13 lr 0.000862 time 0.2885 (0.3241) loss 4.8424 (3.7739) grad_norm 1.2582 (1.2040) [2021-04-15 21:14:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][110/1251] eta 0:06:05 lr 0.000862 time 0.2753 (0.3201) loss 4.1910 (3.7783) grad_norm 1.0594 (1.1892) [2021-04-15 21:14:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][120/1251] eta 0:05:58 lr 0.000862 time 0.2700 (0.3166) loss 3.0333 (3.7721) grad_norm 1.1903 (1.1815) [2021-04-15 21:14:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][130/1251] eta 0:05:52 lr 0.000862 time 0.2697 (0.3149) loss 3.3629 (3.7719) grad_norm 1.0363 (1.1768) [2021-04-15 21:14:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][140/1251] eta 0:05:46 lr 0.000862 time 0.2768 (0.3122) loss 4.4540 (3.7782) grad_norm 1.0034 (1.1769) [2021-04-15 21:14:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][150/1251] eta 0:05:42 lr 0.000862 time 0.2941 (0.3111) loss 3.9248 (3.7900) grad_norm 1.1394 (1.1744) [2021-04-15 21:14:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][160/1251] eta 0:05:37 lr 0.000862 time 0.2653 (0.3098) loss 4.2635 (3.8006) grad_norm 1.1590 (1.1770) [2021-04-15 21:14:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][170/1251] eta 0:05:32 lr 0.000862 time 0.2954 (0.3078) loss 4.2189 (3.8118) grad_norm 1.1573 (1.1761) [2021-04-15 21:14:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][180/1251] eta 0:05:28 lr 0.000862 time 0.2936 (0.3064) loss 3.1412 (3.8077) grad_norm 0.9981 (1.1706) [2021-04-15 21:14:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][190/1251] eta 0:05:23 lr 0.000862 time 0.2988 (0.3054) loss 3.4982 (3.8083) grad_norm 1.5607 (1.1723) [2021-04-15 21:14:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][200/1251] eta 0:05:19 lr 0.000862 time 0.2962 (0.3043) loss 3.3143 (3.8062) grad_norm 0.9696 (1.1714) [2021-04-15 21:14:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][210/1251] eta 0:05:15 lr 0.000862 time 0.2764 (0.3035) loss 4.2169 (3.8063) grad_norm 1.1892 (1.1694) [2021-04-15 21:14:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][220/1251] eta 0:05:12 lr 0.000862 time 0.2975 (0.3030) loss 3.6816 (3.8006) grad_norm 1.2068 (1.1674) [2021-04-15 21:14:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][230/1251] eta 0:05:08 lr 0.000862 time 0.2851 (0.3021) loss 4.6231 (3.7912) grad_norm 1.1741 (1.1664) [2021-04-15 21:14:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][240/1251] eta 0:05:04 lr 0.000862 time 0.2566 (0.3015) loss 3.0185 (3.7900) grad_norm 1.1262 (1.1643) [2021-04-15 21:14:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][250/1251] eta 0:05:01 lr 0.000862 time 0.2778 (0.3013) loss 4.3904 (3.7831) grad_norm 1.1608 (1.1674) [2021-04-15 21:14:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][260/1251] eta 0:04:57 lr 0.000862 time 0.2774 (0.3005) loss 4.3952 (3.7933) grad_norm 1.0704 (1.1692) [2021-04-15 21:14:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][270/1251] eta 0:04:53 lr 0.000861 time 0.2938 (0.2997) loss 3.2983 (3.7949) grad_norm 1.1939 (1.1698) [2021-04-15 21:14:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][280/1251] eta 0:04:50 lr 0.000861 time 0.3004 (0.2990) loss 4.3038 (3.8001) grad_norm 1.3596 (1.1710) [2021-04-15 21:14:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][290/1251] eta 0:04:46 lr 0.000861 time 0.2736 (0.2982) loss 2.6983 (3.8030) grad_norm 1.1719 (1.1708) [2021-04-15 21:15:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][300/1251] eta 0:04:43 lr 0.000861 time 0.2741 (0.2978) loss 4.0265 (3.8049) grad_norm 1.1752 (1.1711) [2021-04-15 21:15:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][310/1251] eta 0:04:39 lr 0.000861 time 0.2935 (0.2973) loss 3.3203 (3.8033) grad_norm 1.2195 (1.1721) [2021-04-15 21:15:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][320/1251] eta 0:04:36 lr 0.000861 time 0.3202 (0.2973) loss 3.7256 (3.8046) grad_norm 1.1216 (1.1702) [2021-04-15 21:15:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][330/1251] eta 0:04:33 lr 0.000861 time 0.2908 (0.2968) loss 3.8603 (3.8092) grad_norm 1.1350 (1.1723) [2021-04-15 21:15:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][340/1251] eta 0:04:30 lr 0.000861 time 0.2692 (0.2968) loss 3.0651 (3.8151) grad_norm 1.0321 (1.1715) [2021-04-15 21:15:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][350/1251] eta 0:04:27 lr 0.000861 time 0.2721 (0.2967) loss 4.0033 (3.8182) grad_norm 1.0083 (1.1695) [2021-04-15 21:15:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][360/1251] eta 0:04:23 lr 0.000861 time 0.2800 (0.2962) loss 2.8246 (3.8255) grad_norm 1.2473 (1.1698) [2021-04-15 21:15:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][370/1251] eta 0:04:20 lr 0.000861 time 0.2935 (0.2961) loss 4.1992 (3.8199) grad_norm 1.0858 (1.1688) [2021-04-15 21:15:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][380/1251] eta 0:04:17 lr 0.000861 time 0.2832 (0.2956) loss 4.0829 (3.8168) grad_norm 1.4174 (1.1705) [2021-04-15 21:15:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][390/1251] eta 0:04:14 lr 0.000861 time 0.3365 (0.2953) loss 4.3986 (3.8179) grad_norm 1.2395 (1.1718) [2021-04-15 21:15:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][400/1251] eta 0:04:10 lr 0.000861 time 0.3081 (0.2949) loss 3.8437 (3.8188) grad_norm 1.2577 (1.1712) [2021-04-15 21:15:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][410/1251] eta 0:04:07 lr 0.000861 time 0.2856 (0.2946) loss 4.2988 (3.8230) grad_norm 1.2898 (1.1697) [2021-04-15 21:15:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][420/1251] eta 0:04:04 lr 0.000861 time 0.2709 (0.2942) loss 2.8077 (3.8251) grad_norm 1.0471 (1.1682) [2021-04-15 21:15:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][430/1251] eta 0:04:01 lr 0.000861 time 0.2663 (0.2939) loss 4.1212 (3.8326) grad_norm 1.0261 (1.1684) [2021-04-15 21:15:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][440/1251] eta 0:03:58 lr 0.000861 time 0.2680 (0.2935) loss 4.0445 (3.8364) grad_norm 1.2076 (1.1690) [2021-04-15 21:15:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][450/1251] eta 0:03:54 lr 0.000861 time 0.2753 (0.2932) loss 3.2083 (3.8345) grad_norm 1.1644 (1.1694) [2021-04-15 21:15:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][460/1251] eta 0:03:51 lr 0.000861 time 0.2602 (0.2929) loss 4.5066 (3.8294) grad_norm 1.2933 (1.1700) [2021-04-15 21:15:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][470/1251] eta 0:03:48 lr 0.000861 time 0.2606 (0.2926) loss 3.8812 (3.8301) grad_norm 1.0911 (1.1699) [2021-04-15 21:15:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][480/1251] eta 0:03:45 lr 0.000861 time 0.2785 (0.2925) loss 4.6675 (3.8345) grad_norm 1.1088 (1.1691) [2021-04-15 21:15:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][490/1251] eta 0:03:42 lr 0.000861 time 0.2598 (0.2922) loss 3.3799 (3.8311) grad_norm 1.1605 (1.1678) [2021-04-15 21:15:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][500/1251] eta 0:03:39 lr 0.000861 time 0.3048 (0.2920) loss 2.3881 (3.8246) grad_norm 1.1510 (1.1659) [2021-04-15 21:15:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][510/1251] eta 0:03:36 lr 0.000861 time 0.3064 (0.2918) loss 4.3997 (3.8297) grad_norm 1.0266 (1.1647) [2021-04-15 21:16:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][520/1251] eta 0:03:33 lr 0.000861 time 0.2775 (0.2915) loss 4.5532 (3.8303) grad_norm 1.1691 (1.1641) [2021-04-15 21:16:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][530/1251] eta 0:03:30 lr 0.000861 time 0.2865 (0.2914) loss 2.7678 (3.8302) grad_norm 1.1771 (1.1651) [2021-04-15 21:16:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][540/1251] eta 0:03:27 lr 0.000861 time 0.3001 (0.2913) loss 4.4952 (3.8359) grad_norm 1.0735 (1.1644) [2021-04-15 21:16:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][550/1251] eta 0:03:24 lr 0.000861 time 0.4398 (0.2913) loss 3.7980 (3.8330) grad_norm 1.1489 (1.1641) [2021-04-15 21:16:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][560/1251] eta 0:03:21 lr 0.000861 time 0.2817 (0.2911) loss 4.1128 (3.8335) grad_norm 1.1308 (1.1648) [2021-04-15 21:16:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][570/1251] eta 0:03:18 lr 0.000861 time 0.2639 (0.2911) loss 4.3955 (3.8375) grad_norm 1.3906 (1.1652) [2021-04-15 21:16:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][580/1251] eta 0:03:15 lr 0.000861 time 0.3047 (0.2912) loss 4.3795 (3.8437) grad_norm 1.0517 (1.1633) [2021-04-15 21:16:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][590/1251] eta 0:03:12 lr 0.000861 time 0.2682 (0.2909) loss 3.1166 (3.8376) grad_norm 1.1282 (1.1629) [2021-04-15 21:16:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][600/1251] eta 0:03:09 lr 0.000861 time 0.2957 (0.2908) loss 4.0550 (3.8383) grad_norm 0.9793 (1.1629) [2021-04-15 21:16:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][610/1251] eta 0:03:06 lr 0.000861 time 0.2698 (0.2907) loss 4.6320 (3.8388) grad_norm 1.1078 (1.1617) [2021-04-15 21:16:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][620/1251] eta 0:03:03 lr 0.000860 time 0.2773 (0.2905) loss 3.8579 (3.8405) grad_norm 1.0937 (1.1613) [2021-04-15 21:16:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][630/1251] eta 0:03:00 lr 0.000860 time 0.2787 (0.2904) loss 4.4126 (3.8375) grad_norm 1.2127 (1.1621) [2021-04-15 21:16:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][640/1251] eta 0:02:57 lr 0.000860 time 0.2552 (0.2905) loss 4.1730 (3.8408) grad_norm 1.0873 (1.1610) [2021-04-15 21:16:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][650/1251] eta 0:02:54 lr 0.000860 time 0.2806 (0.2903) loss 3.1572 (3.8441) grad_norm 1.4337 (1.1616) [2021-04-15 21:16:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][660/1251] eta 0:02:51 lr 0.000860 time 0.2760 (0.2902) loss 3.9606 (3.8475) grad_norm 1.1702 (1.1614) [2021-04-15 21:16:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][670/1251] eta 0:02:48 lr 0.000860 time 0.2927 (0.2902) loss 3.8264 (3.8489) grad_norm 1.2265 (1.1612) [2021-04-15 21:16:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][680/1251] eta 0:02:45 lr 0.000860 time 0.2907 (0.2900) loss 3.8888 (3.8462) grad_norm 1.0594 (1.1608) [2021-04-15 21:16:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][690/1251] eta 0:02:42 lr 0.000860 time 0.2698 (0.2899) loss 3.0937 (3.8430) grad_norm 1.2243 (1.1607) [2021-04-15 21:16:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][700/1251] eta 0:02:39 lr 0.000860 time 0.2957 (0.2898) loss 3.9613 (3.8442) grad_norm 1.1790 (1.1616) [2021-04-15 21:16:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][710/1251] eta 0:02:36 lr 0.000860 time 0.2688 (0.2898) loss 4.1723 (3.8469) grad_norm 1.1919 (1.1619) [2021-04-15 21:16:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][720/1251] eta 0:02:33 lr 0.000860 time 0.2984 (0.2896) loss 2.5831 (3.8492) grad_norm 1.1601 (1.1610) [2021-04-15 21:17:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][730/1251] eta 0:02:31 lr 0.000860 time 0.3969 (0.2899) loss 4.4245 (3.8502) grad_norm 0.9822 (1.1611) [2021-04-15 21:17:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][740/1251] eta 0:02:28 lr 0.000860 time 0.2755 (0.2897) loss 3.3254 (3.8486) grad_norm 1.2180 (1.1608) [2021-04-15 21:17:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][750/1251] eta 0:02:25 lr 0.000860 time 0.2710 (0.2895) loss 3.7873 (3.8487) grad_norm 1.1286 (1.1617) [2021-04-15 21:17:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][760/1251] eta 0:02:22 lr 0.000860 time 0.2670 (0.2894) loss 4.1700 (3.8455) grad_norm 1.3226 (1.1619) [2021-04-15 21:17:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][770/1251] eta 0:02:19 lr 0.000860 time 0.2734 (0.2893) loss 3.7018 (3.8448) grad_norm 1.1515 (1.1617) [2021-04-15 21:17:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][780/1251] eta 0:02:16 lr 0.000860 time 0.2446 (0.2892) loss 4.4008 (3.8437) grad_norm 1.2447 (1.1623) [2021-04-15 21:17:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][790/1251] eta 0:02:13 lr 0.000860 time 0.3005 (0.2892) loss 4.2996 (3.8435) grad_norm 1.1851 (1.1618) [2021-04-15 21:17:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][800/1251] eta 0:02:10 lr 0.000860 time 0.2675 (0.2890) loss 2.5969 (3.8438) grad_norm 1.0590 (1.1617) [2021-04-15 21:17:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][810/1251] eta 0:02:07 lr 0.000860 time 0.2470 (0.2888) loss 2.8154 (3.8439) grad_norm 1.0564 (1.1610) [2021-04-15 21:17:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][820/1251] eta 0:02:04 lr 0.000860 time 0.2680 (0.2888) loss 3.3400 (3.8432) grad_norm 1.4232 (1.1607) [2021-04-15 21:17:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][830/1251] eta 0:02:01 lr 0.000860 time 0.2461 (0.2889) loss 3.8961 (3.8433) grad_norm 1.1215 (1.1605) [2021-04-15 21:17:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][840/1251] eta 0:01:58 lr 0.000860 time 0.3010 (0.2888) loss 3.3978 (3.8459) grad_norm 1.0475 (1.1606) [2021-04-15 21:17:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][850/1251] eta 0:01:55 lr 0.000860 time 0.2686 (0.2887) loss 3.8868 (3.8435) grad_norm 1.0246 (1.1608) [2021-04-15 21:17:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][860/1251] eta 0:01:52 lr 0.000860 time 0.2705 (0.2886) loss 3.8253 (3.8428) grad_norm 1.0403 (1.1606) [2021-04-15 21:17:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][870/1251] eta 0:01:49 lr 0.000860 time 0.2542 (0.2885) loss 4.1869 (3.8428) grad_norm 1.0891 (1.1600) [2021-04-15 21:17:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][880/1251] eta 0:01:46 lr 0.000860 time 0.3076 (0.2884) loss 3.5760 (3.8433) grad_norm 1.0834 (1.1596) [2021-04-15 21:17:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][890/1251] eta 0:01:44 lr 0.000860 time 0.2436 (0.2885) loss 4.3672 (3.8419) grad_norm 1.0125 (1.1585) [2021-04-15 21:17:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][900/1251] eta 0:01:41 lr 0.000860 time 0.2777 (0.2884) loss 3.3737 (3.8431) grad_norm 1.0695 (1.1580) [2021-04-15 21:17:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][910/1251] eta 0:01:38 lr 0.000860 time 0.2902 (0.2884) loss 3.4198 (3.8398) grad_norm 1.1062 (1.1582) [2021-04-15 21:17:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][920/1251] eta 0:01:35 lr 0.000860 time 0.2862 (0.2884) loss 3.8893 (3.8407) grad_norm 1.1822 (1.1587) [2021-04-15 21:17:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][930/1251] eta 0:01:32 lr 0.000860 time 0.2577 (0.2885) loss 4.0897 (3.8420) grad_norm 1.0866 (1.1588) [2021-04-15 21:18:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][940/1251] eta 0:01:29 lr 0.000860 time 0.4099 (0.2885) loss 4.2397 (3.8432) grad_norm 1.1640 (1.1587) [2021-04-15 21:18:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][950/1251] eta 0:01:26 lr 0.000860 time 0.2894 (0.2884) loss 3.5067 (3.8416) grad_norm 1.1015 (1.1597) [2021-04-15 21:18:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][960/1251] eta 0:01:23 lr 0.000860 time 0.2646 (0.2883) loss 4.6669 (3.8398) grad_norm 1.3657 (1.1604) [2021-04-15 21:18:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][970/1251] eta 0:01:20 lr 0.000859 time 0.2857 (0.2882) loss 3.9004 (3.8408) grad_norm 1.0023 (1.1598) [2021-04-15 21:18:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][980/1251] eta 0:01:18 lr 0.000859 time 0.2889 (0.2881) loss 4.2225 (3.8418) grad_norm 1.2559 (1.1598) [2021-04-15 21:18:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][990/1251] eta 0:01:15 lr 0.000859 time 0.2719 (0.2880) loss 4.2056 (3.8403) grad_norm 1.4413 (1.1601) [2021-04-15 21:18:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][1000/1251] eta 0:01:12 lr 0.000859 time 0.2796 (0.2879) loss 3.8725 (3.8376) grad_norm 1.2462 (1.1598) [2021-04-15 21:18:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][1010/1251] eta 0:01:09 lr 0.000859 time 0.2535 (0.2879) loss 4.0888 (3.8369) grad_norm 1.0683 (1.1599) [2021-04-15 21:18:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][1020/1251] eta 0:01:06 lr 0.000859 time 0.2705 (0.2878) loss 3.0179 (3.8376) grad_norm 1.2196 (1.1597) [2021-04-15 21:18:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][1030/1251] eta 0:01:03 lr 0.000859 time 0.2581 (0.2877) loss 3.8860 (3.8367) grad_norm inf (inf) [2021-04-15 21:18:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][1040/1251] eta 0:01:00 lr 0.000859 time 0.2830 (0.2876) loss 3.6715 (3.8359) grad_norm 1.3436 (inf) [2021-04-15 21:18:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][1050/1251] eta 0:00:57 lr 0.000859 time 0.2714 (0.2876) loss 4.9609 (3.8385) grad_norm 1.2446 (inf) [2021-04-15 21:18:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][1060/1251] eta 0:00:54 lr 0.000859 time 0.2851 (0.2875) loss 4.0421 (3.8379) grad_norm 1.3030 (inf) [2021-04-15 21:18:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][1070/1251] eta 0:00:52 lr 0.000859 time 0.2723 (0.2874) loss 3.9014 (3.8388) grad_norm 1.0619 (inf) [2021-04-15 21:18:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][1080/1251] eta 0:00:49 lr 0.000859 time 0.2788 (0.2873) loss 3.5080 (3.8395) grad_norm 1.0177 (inf) [2021-04-15 21:18:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][1090/1251] eta 0:00:46 lr 0.000859 time 0.2707 (0.2872) loss 3.3848 (3.8395) grad_norm 1.1725 (inf) [2021-04-15 21:18:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][1100/1251] eta 0:00:43 lr 0.000859 time 0.2683 (0.2872) loss 3.9667 (3.8378) grad_norm 1.1045 (inf) [2021-04-15 21:18:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][1110/1251] eta 0:00:40 lr 0.000859 time 0.3043 (0.2871) loss 4.0381 (3.8379) grad_norm 1.1681 (inf) [2021-04-15 21:18:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][1120/1251] eta 0:00:37 lr 0.000859 time 0.2889 (0.2871) loss 4.4402 (3.8395) grad_norm 1.2871 (inf) [2021-04-15 21:18:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][1130/1251] eta 0:00:34 lr 0.000859 time 0.2569 (0.2871) loss 4.1171 (3.8373) grad_norm 1.1473 (inf) [2021-04-15 21:18:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][1140/1251] eta 0:00:31 lr 0.000859 time 0.2782 (0.2872) loss 4.0832 (3.8355) grad_norm 1.0395 (inf) [2021-04-15 21:19:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][1150/1251] eta 0:00:29 lr 0.000859 time 0.4385 (0.2872) loss 4.4878 (3.8342) grad_norm 1.1022 (inf) [2021-04-15 21:19:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][1160/1251] eta 0:00:26 lr 0.000859 time 0.2728 (0.2871) loss 4.7976 (3.8313) grad_norm 1.1136 (inf) [2021-04-15 21:19:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][1170/1251] eta 0:00:23 lr 0.000859 time 0.2716 (0.2870) loss 4.0835 (3.8323) grad_norm 1.4491 (inf) [2021-04-15 21:19:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][1180/1251] eta 0:00:20 lr 0.000859 time 0.2793 (0.2870) loss 4.3283 (3.8326) grad_norm 0.9719 (inf) [2021-04-15 21:19:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][1190/1251] eta 0:00:17 lr 0.000859 time 0.2950 (0.2869) loss 3.9107 (3.8326) grad_norm 1.1392 (inf) [2021-04-15 21:19:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][1200/1251] eta 0:00:14 lr 0.000859 time 0.2718 (0.2868) loss 4.1486 (3.8311) grad_norm 1.2253 (inf) [2021-04-15 21:19:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][1210/1251] eta 0:00:11 lr 0.000859 time 0.2664 (0.2868) loss 2.7233 (3.8315) grad_norm 0.9809 (inf) [2021-04-15 21:19:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][1220/1251] eta 0:00:08 lr 0.000859 time 0.2686 (0.2868) loss 2.7043 (3.8304) grad_norm 1.0889 (inf) [2021-04-15 21:19:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][1230/1251] eta 0:00:06 lr 0.000859 time 0.2807 (0.2867) loss 4.3512 (3.8301) grad_norm 1.1760 (inf) [2021-04-15 21:19:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][1240/1251] eta 0:00:03 lr 0.000859 time 0.3372 (0.2867) loss 4.1853 (3.8317) grad_norm 1.0680 (inf) [2021-04-15 21:19:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [73/300][1250/1251] eta 0:00:00 lr 0.000859 time 0.2478 (0.2864) loss 3.0187 (3.8310) grad_norm 0.9763 (inf) [2021-04-15 21:19:31 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 73 training takes 0:06:01 [2021-04-15 21:19:31 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_73.pth saving...... [2021-04-15 21:19:43 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_73.pth saved !!! [2021-04-15 21:19:44 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.111 (1.111) Loss 1.1392 (1.1392) Acc@1 72.461 (72.461) Acc@5 92.090 (92.090) [2021-04-15 21:19:45 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.127 (0.217) Loss 1.2474 (1.1880) Acc@1 70.410 (71.662) Acc@5 91.016 (91.690) [2021-04-15 21:19:48 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.148 (0.219) Loss 1.1737 (1.1847) Acc@1 72.559 (72.256) Acc@5 92.480 (91.695) [2021-04-15 21:19:51 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.219 (0.257) Loss 1.1564 (1.1923) Acc@1 73.828 (72.190) Acc@5 91.602 (91.450) [2021-04-15 21:19:52 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.220) Loss 1.2255 (1.2046) Acc@1 71.875 (71.944) Acc@5 90.137 (91.240) [2021-04-15 21:19:56 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 71.966 Acc@5 91.340 [2021-04-15 21:19:56 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 72.0% [2021-04-15 21:19:56 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 72.13% [2021-04-15 21:20:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][0/1251] eta 1:15:38 lr 0.000859 time 3.6276 (3.6276) loss 4.0742 (4.0742) grad_norm 1.0517 (1.0517) [2021-04-15 21:20:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][10/1251] eta 0:12:04 lr 0.000859 time 0.2869 (0.5839) loss 3.7386 (3.8456) grad_norm 1.1236 (1.1502) [2021-04-15 21:20:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][20/1251] eta 0:09:05 lr 0.000859 time 0.2802 (0.4433) loss 4.4213 (3.7163) grad_norm 1.2244 (1.1595) [2021-04-15 21:20:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][30/1251] eta 0:08:03 lr 0.000859 time 0.2790 (0.3962) loss 4.6363 (3.8105) grad_norm 1.1640 (1.1435) [2021-04-15 21:20:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][40/1251] eta 0:07:23 lr 0.000859 time 0.2732 (0.3666) loss 4.1352 (3.8442) grad_norm 1.1803 (1.1453) [2021-04-15 21:20:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][50/1251] eta 0:07:01 lr 0.000859 time 0.2747 (0.3509) loss 3.6296 (3.7987) grad_norm 1.2268 (1.1619) [2021-04-15 21:20:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][60/1251] eta 0:06:45 lr 0.000858 time 0.3093 (0.3401) loss 2.8350 (3.8115) grad_norm 1.1080 (1.1584) [2021-04-15 21:20:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][70/1251] eta 0:06:31 lr 0.000858 time 0.2819 (0.3315) loss 4.0798 (3.7720) grad_norm 1.3081 (1.1732) [2021-04-15 21:20:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][80/1251] eta 0:06:20 lr 0.000858 time 0.2755 (0.3253) loss 4.7298 (3.7518) grad_norm 1.0398 (1.1657) [2021-04-15 21:20:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][90/1251] eta 0:06:14 lr 0.000858 time 0.2737 (0.3222) loss 2.7473 (3.7458) grad_norm 1.1422 (1.1639) [2021-04-15 21:20:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][100/1251] eta 0:06:06 lr 0.000858 time 0.3074 (0.3186) loss 3.6618 (3.7583) grad_norm 1.3823 (1.1696) [2021-04-15 21:20:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][110/1251] eta 0:05:59 lr 0.000858 time 0.2776 (0.3151) loss 4.0242 (3.7888) grad_norm 0.9180 (1.1665) [2021-04-15 21:20:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][120/1251] eta 0:05:54 lr 0.000858 time 0.2762 (0.3133) loss 4.2492 (3.7817) grad_norm 1.0437 (1.1619) [2021-04-15 21:20:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][130/1251] eta 0:05:49 lr 0.000858 time 0.2956 (0.3121) loss 4.2718 (3.7659) grad_norm 1.0104 (1.1587) [2021-04-15 21:20:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][140/1251] eta 0:05:44 lr 0.000858 time 0.2910 (0.3100) loss 2.9728 (3.7589) grad_norm 1.2555 (1.1640) [2021-04-15 21:20:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][150/1251] eta 0:05:40 lr 0.000858 time 0.4089 (0.3089) loss 4.3266 (3.7608) grad_norm 1.1108 (1.1618) [2021-04-15 21:20:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][160/1251] eta 0:05:34 lr 0.000858 time 0.2770 (0.3069) loss 3.6535 (3.7742) grad_norm 1.0824 (1.1622) [2021-04-15 21:20:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][170/1251] eta 0:05:30 lr 0.000858 time 0.2929 (0.3057) loss 4.1957 (3.7784) grad_norm 1.1925 (1.1668) [2021-04-15 21:20:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][180/1251] eta 0:05:26 lr 0.000858 time 0.2794 (0.3048) loss 4.0871 (3.7849) grad_norm 1.2948 (1.1735) [2021-04-15 21:20:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][190/1251] eta 0:05:22 lr 0.000858 time 0.2765 (0.3035) loss 4.2737 (3.7930) grad_norm 1.1811 (1.1758) [2021-04-15 21:20:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][200/1251] eta 0:05:18 lr 0.000858 time 0.2662 (0.3026) loss 4.4357 (3.7943) grad_norm 1.0863 (1.1797) [2021-04-15 21:21:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][210/1251] eta 0:05:14 lr 0.000858 time 0.2704 (0.3018) loss 4.2090 (3.8009) grad_norm 1.1536 (1.1779) [2021-04-15 21:21:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][220/1251] eta 0:05:10 lr 0.000858 time 0.2737 (0.3008) loss 3.5022 (3.8102) grad_norm 1.0916 (1.1756) [2021-04-15 21:21:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][230/1251] eta 0:05:06 lr 0.000858 time 0.2541 (0.2997) loss 3.4224 (3.8059) grad_norm 1.1723 (1.1760) [2021-04-15 21:21:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][240/1251] eta 0:05:02 lr 0.000858 time 0.2851 (0.2990) loss 3.5112 (3.8066) grad_norm 1.3531 (1.1810) [2021-04-15 21:21:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][250/1251] eta 0:04:58 lr 0.000858 time 0.2758 (0.2984) loss 4.0276 (3.8016) grad_norm 1.3752 (1.1843) [2021-04-15 21:21:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][260/1251] eta 0:04:55 lr 0.000858 time 0.2682 (0.2978) loss 3.5396 (3.7957) grad_norm 1.1389 (1.1844) [2021-04-15 21:21:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][270/1251] eta 0:04:51 lr 0.000858 time 0.2752 (0.2972) loss 4.2065 (3.8023) grad_norm 1.0773 (1.1808) [2021-04-15 21:21:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][280/1251] eta 0:04:47 lr 0.000858 time 0.2766 (0.2965) loss 3.0985 (3.7973) grad_norm 1.1348 (1.1794) [2021-04-15 21:21:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][290/1251] eta 0:04:44 lr 0.000858 time 0.3169 (0.2960) loss 2.5259 (3.7939) grad_norm 1.0416 (1.1774) [2021-04-15 21:21:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][300/1251] eta 0:04:41 lr 0.000858 time 0.2697 (0.2955) loss 3.9646 (3.8021) grad_norm 1.2051 (1.1795) [2021-04-15 21:21:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][310/1251] eta 0:04:38 lr 0.000858 time 0.2829 (0.2958) loss 4.1434 (3.8162) grad_norm 1.0722 (1.1769) [2021-04-15 21:21:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][320/1251] eta 0:04:34 lr 0.000858 time 0.2647 (0.2954) loss 3.9988 (3.8196) grad_norm 1.2345 (1.1751) [2021-04-15 21:21:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][330/1251] eta 0:04:31 lr 0.000858 time 0.2784 (0.2951) loss 4.3511 (3.8185) grad_norm 1.3826 (1.1754) [2021-04-15 21:21:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][340/1251] eta 0:04:28 lr 0.000858 time 0.2621 (0.2951) loss 2.5628 (3.8156) grad_norm 1.1807 (1.1768) [2021-04-15 21:21:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][350/1251] eta 0:04:25 lr 0.000858 time 0.2682 (0.2951) loss 3.4353 (3.8152) grad_norm 1.1109 (1.1747) [2021-04-15 21:21:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][360/1251] eta 0:04:22 lr 0.000858 time 0.2613 (0.2950) loss 4.0156 (3.8177) grad_norm 1.0703 (1.1730) [2021-04-15 21:21:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][370/1251] eta 0:04:19 lr 0.000858 time 0.2682 (0.2946) loss 3.8324 (3.8108) grad_norm 1.1419 (1.1735) [2021-04-15 21:21:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][380/1251] eta 0:04:16 lr 0.000858 time 0.2709 (0.2941) loss 4.0969 (3.8141) grad_norm 1.5556 (1.1750) [2021-04-15 21:21:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][390/1251] eta 0:04:13 lr 0.000858 time 0.2724 (0.2939) loss 3.5213 (3.8168) grad_norm 1.1530 (1.1749) [2021-04-15 21:21:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][400/1251] eta 0:04:09 lr 0.000858 time 0.2921 (0.2934) loss 3.2303 (3.8052) grad_norm 1.1598 (1.1744) [2021-04-15 21:21:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][410/1251] eta 0:04:06 lr 0.000857 time 0.3095 (0.2932) loss 4.3439 (3.7957) grad_norm 1.5444 (1.1753) [2021-04-15 21:22:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][420/1251] eta 0:04:03 lr 0.000857 time 0.2742 (0.2931) loss 3.1287 (3.7973) grad_norm 1.1090 (1.1758) [2021-04-15 21:22:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][430/1251] eta 0:04:00 lr 0.000857 time 0.3064 (0.2928) loss 3.2337 (3.7916) grad_norm 1.4577 (1.1750) [2021-04-15 21:22:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][440/1251] eta 0:03:57 lr 0.000857 time 0.2960 (0.2925) loss 3.7792 (3.7943) grad_norm 1.2577 (1.1769) [2021-04-15 21:22:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][450/1251] eta 0:03:54 lr 0.000857 time 0.2883 (0.2928) loss 3.8523 (3.7888) grad_norm 1.2048 (1.1770) [2021-04-15 21:22:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][460/1251] eta 0:03:51 lr 0.000857 time 0.2914 (0.2924) loss 4.2613 (3.7963) grad_norm 1.4185 (1.1767) [2021-04-15 21:22:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][470/1251] eta 0:03:48 lr 0.000857 time 0.2561 (0.2921) loss 3.8957 (3.8002) grad_norm 1.0762 (1.1756) [2021-04-15 21:22:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][480/1251] eta 0:03:45 lr 0.000857 time 0.2949 (0.2918) loss 4.3375 (3.8003) grad_norm 1.2175 (1.1749) [2021-04-15 21:22:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][490/1251] eta 0:03:41 lr 0.000857 time 0.2564 (0.2915) loss 4.3161 (3.8014) grad_norm 1.4575 (1.1762) [2021-04-15 21:22:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][500/1251] eta 0:03:38 lr 0.000857 time 0.2951 (0.2914) loss 3.7908 (3.8019) grad_norm 1.0835 (1.1755) [2021-04-15 21:22:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][510/1251] eta 0:03:35 lr 0.000857 time 0.2712 (0.2911) loss 4.7010 (3.8050) grad_norm 1.3942 (1.1764) [2021-04-15 21:22:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][520/1251] eta 0:03:32 lr 0.000857 time 0.2993 (0.2909) loss 3.9394 (3.8070) grad_norm 1.1365 (1.1756) [2021-04-15 21:22:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][530/1251] eta 0:03:29 lr 0.000857 time 0.2972 (0.2906) loss 4.0828 (3.8109) grad_norm 1.0901 (1.1744) [2021-04-15 21:22:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][540/1251] eta 0:03:26 lr 0.000857 time 0.2876 (0.2904) loss 3.8504 (3.8104) grad_norm 1.3122 (1.1737) [2021-04-15 21:22:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][550/1251] eta 0:03:23 lr 0.000857 time 0.2868 (0.2902) loss 4.6845 (3.8178) grad_norm 0.9981 (1.1728) [2021-04-15 21:22:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][560/1251] eta 0:03:20 lr 0.000857 time 0.2767 (0.2900) loss 3.2059 (3.8167) grad_norm 1.1439 (1.1720) [2021-04-15 21:22:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][570/1251] eta 0:03:17 lr 0.000857 time 0.2668 (0.2901) loss 4.4335 (3.8151) grad_norm 1.1365 (1.1714) [2021-04-15 21:22:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][580/1251] eta 0:03:14 lr 0.000857 time 0.2852 (0.2900) loss 3.3750 (3.8191) grad_norm 1.3027 (1.1702) [2021-04-15 21:22:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][590/1251] eta 0:03:11 lr 0.000857 time 0.3001 (0.2901) loss 2.9292 (3.8219) grad_norm 1.0737 (1.1707) [2021-04-15 21:22:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][600/1251] eta 0:03:08 lr 0.000857 time 0.4792 (0.2902) loss 4.4504 (3.8196) grad_norm 1.2907 (1.1713) [2021-04-15 21:22:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][610/1251] eta 0:03:05 lr 0.000857 time 0.2917 (0.2900) loss 4.4325 (3.8165) grad_norm 1.0925 (1.1727) [2021-04-15 21:22:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][620/1251] eta 0:03:02 lr 0.000857 time 0.2437 (0.2899) loss 3.2906 (3.8180) grad_norm 1.1066 (1.1712) [2021-04-15 21:22:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][630/1251] eta 0:03:00 lr 0.000857 time 0.2910 (0.2899) loss 3.6183 (3.8170) grad_norm 0.9548 (1.1700) [2021-04-15 21:23:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][640/1251] eta 0:02:57 lr 0.000857 time 0.2638 (0.2898) loss 3.0605 (3.8143) grad_norm 1.1167 (1.1693) [2021-04-15 21:23:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][650/1251] eta 0:02:54 lr 0.000857 time 0.3059 (0.2897) loss 3.3165 (3.8139) grad_norm 1.2256 (1.1689) [2021-04-15 21:23:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][660/1251] eta 0:02:51 lr 0.000857 time 0.3013 (0.2896) loss 4.4385 (3.8191) grad_norm 1.0914 (1.1685) [2021-04-15 21:23:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][670/1251] eta 0:02:48 lr 0.000857 time 0.2819 (0.2896) loss 3.7038 (3.8200) grad_norm 0.9670 (1.1677) [2021-04-15 21:23:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][680/1251] eta 0:02:45 lr 0.000857 time 0.3248 (0.2895) loss 4.0822 (3.8179) grad_norm 0.9831 (1.1673) [2021-04-15 21:23:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][690/1251] eta 0:02:42 lr 0.000857 time 0.2849 (0.2895) loss 3.5401 (3.8174) grad_norm 1.0918 (1.1672) [2021-04-15 21:23:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][700/1251] eta 0:02:39 lr 0.000857 time 0.2645 (0.2894) loss 4.4939 (3.8168) grad_norm 1.0768 (1.1672) [2021-04-15 21:23:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][710/1251] eta 0:02:36 lr 0.000857 time 0.2896 (0.2894) loss 4.2344 (3.8142) grad_norm 1.1175 (1.1675) [2021-04-15 21:23:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][720/1251] eta 0:02:33 lr 0.000857 time 0.2683 (0.2893) loss 3.3985 (3.8101) grad_norm 1.0807 (1.1669) [2021-04-15 21:23:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][730/1251] eta 0:02:30 lr 0.000857 time 0.3160 (0.2893) loss 2.9357 (3.8121) grad_norm 1.3176 (1.1668) [2021-04-15 21:23:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][740/1251] eta 0:02:27 lr 0.000857 time 0.2588 (0.2892) loss 4.0578 (3.8078) grad_norm 1.3786 (1.1683) [2021-04-15 21:23:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][750/1251] eta 0:02:24 lr 0.000856 time 0.3057 (0.2891) loss 3.9630 (3.8099) grad_norm 1.0658 (1.1697) [2021-04-15 21:23:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][760/1251] eta 0:02:21 lr 0.000856 time 0.3194 (0.2890) loss 4.3698 (3.8106) grad_norm 1.2178 (1.1712) [2021-04-15 21:23:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][770/1251] eta 0:02:18 lr 0.000856 time 0.2914 (0.2889) loss 4.1818 (3.8113) grad_norm 1.1135 (1.1708) [2021-04-15 21:23:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][780/1251] eta 0:02:16 lr 0.000856 time 0.2731 (0.2888) loss 4.0322 (3.8114) grad_norm 1.1820 (1.1713) [2021-04-15 21:23:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][790/1251] eta 0:02:13 lr 0.000856 time 0.2683 (0.2887) loss 4.3536 (3.8133) grad_norm 1.0585 (1.1714) [2021-04-15 21:23:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][800/1251] eta 0:02:10 lr 0.000856 time 0.2751 (0.2887) loss 2.9059 (3.8161) grad_norm 0.9885 (1.1712) [2021-04-15 21:23:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][810/1251] eta 0:02:07 lr 0.000856 time 0.2648 (0.2886) loss 2.5944 (3.8110) grad_norm 1.2045 (1.1717) [2021-04-15 21:23:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][820/1251] eta 0:02:04 lr 0.000856 time 0.2839 (0.2886) loss 3.7849 (3.8063) grad_norm 1.1690 (1.1719) [2021-04-15 21:23:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][830/1251] eta 0:02:01 lr 0.000856 time 0.2881 (0.2885) loss 4.3825 (3.8125) grad_norm 1.0785 (1.1719) [2021-04-15 21:23:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][840/1251] eta 0:01:58 lr 0.000856 time 0.2828 (0.2885) loss 4.1182 (3.8116) grad_norm 1.2138 (1.1711) [2021-04-15 21:24:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][850/1251] eta 0:01:55 lr 0.000856 time 0.2771 (0.2884) loss 3.9401 (3.8117) grad_norm 1.1071 (1.1708) [2021-04-15 21:24:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][860/1251] eta 0:01:52 lr 0.000856 time 0.2699 (0.2884) loss 4.0214 (3.8135) grad_norm 1.0460 (1.1708) [2021-04-15 21:24:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][870/1251] eta 0:01:49 lr 0.000856 time 0.2608 (0.2883) loss 2.4759 (3.8131) grad_norm 1.2462 (1.1715) [2021-04-15 21:24:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][880/1251] eta 0:01:46 lr 0.000856 time 0.2816 (0.2882) loss 3.9928 (3.8161) grad_norm 1.1348 (1.1720) [2021-04-15 21:24:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][890/1251] eta 0:01:44 lr 0.000856 time 0.2557 (0.2882) loss 3.6906 (3.8154) grad_norm 1.1979 (1.1714) [2021-04-15 21:24:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][900/1251] eta 0:01:41 lr 0.000856 time 0.2820 (0.2881) loss 3.9654 (3.8156) grad_norm 1.1799 (1.1707) [2021-04-15 21:24:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][910/1251] eta 0:01:38 lr 0.000856 time 0.2662 (0.2881) loss 4.0084 (3.8176) grad_norm 1.0849 (1.1701) [2021-04-15 21:24:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][920/1251] eta 0:01:35 lr 0.000856 time 0.2767 (0.2882) loss 4.1644 (3.8194) grad_norm 1.3074 (1.1711) [2021-04-15 21:24:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][930/1251] eta 0:01:32 lr 0.000856 time 0.2656 (0.2883) loss 4.2035 (3.8232) grad_norm 1.3216 (1.1712) [2021-04-15 21:24:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][940/1251] eta 0:01:29 lr 0.000856 time 0.2928 (0.2883) loss 4.0006 (3.8238) grad_norm 1.1646 (1.1710) [2021-04-15 21:24:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][950/1251] eta 0:01:26 lr 0.000856 time 0.2780 (0.2882) loss 3.9229 (3.8235) grad_norm 1.0331 (1.1704) [2021-04-15 21:24:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][960/1251] eta 0:01:23 lr 0.000856 time 0.2862 (0.2882) loss 4.2543 (3.8228) grad_norm 1.0089 (1.1691) [2021-04-15 21:24:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][970/1251] eta 0:01:20 lr 0.000856 time 0.2585 (0.2882) loss 3.5100 (3.8211) grad_norm 0.9968 (1.1689) [2021-04-15 21:24:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][980/1251] eta 0:01:18 lr 0.000856 time 0.3167 (0.2882) loss 3.8747 (3.8230) grad_norm 1.3632 (1.1697) [2021-04-15 21:24:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][990/1251] eta 0:01:15 lr 0.000856 time 0.2975 (0.2882) loss 4.0995 (3.8243) grad_norm 1.3459 (1.1694) [2021-04-15 21:24:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][1000/1251] eta 0:01:12 lr 0.000856 time 0.2824 (0.2881) loss 3.0220 (3.8254) grad_norm 1.3116 (1.1697) [2021-04-15 21:24:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][1010/1251] eta 0:01:09 lr 0.000856 time 0.2696 (0.2880) loss 3.0315 (3.8243) grad_norm 1.1908 (1.1699) [2021-04-15 21:24:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][1020/1251] eta 0:01:06 lr 0.000856 time 0.2756 (0.2879) loss 4.3375 (3.8233) grad_norm 1.1390 (1.1697) [2021-04-15 21:24:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][1030/1251] eta 0:01:03 lr 0.000856 time 0.2830 (0.2879) loss 2.7433 (3.8229) grad_norm 1.3839 (1.1693) [2021-04-15 21:24:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][1040/1251] eta 0:01:00 lr 0.000856 time 0.3034 (0.2878) loss 4.3577 (3.8237) grad_norm 1.0502 (1.1689) [2021-04-15 21:24:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][1050/1251] eta 0:00:57 lr 0.000856 time 0.2883 (0.2878) loss 3.3642 (3.8256) grad_norm 1.0094 (1.1687) [2021-04-15 21:25:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][1060/1251] eta 0:00:54 lr 0.000856 time 0.2476 (0.2878) loss 4.2589 (3.8262) grad_norm 1.4691 (1.1694) [2021-04-15 21:25:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][1070/1251] eta 0:00:52 lr 0.000856 time 0.2917 (0.2877) loss 3.9273 (3.8236) grad_norm 1.3345 (1.1688) [2021-04-15 21:25:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][1080/1251] eta 0:00:49 lr 0.000856 time 0.2694 (0.2877) loss 4.0211 (3.8249) grad_norm 0.9890 (1.1684) [2021-04-15 21:25:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][1090/1251] eta 0:00:46 lr 0.000855 time 0.2869 (0.2876) loss 4.4189 (3.8255) grad_norm 1.1794 (1.1683) [2021-04-15 21:25:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][1100/1251] eta 0:00:43 lr 0.000855 time 0.3033 (0.2876) loss 4.4236 (3.8283) grad_norm 1.2119 (1.1684) [2021-04-15 21:25:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][1110/1251] eta 0:00:40 lr 0.000855 time 0.2924 (0.2875) loss 3.2397 (3.8267) grad_norm 1.3432 (1.1689) [2021-04-15 21:25:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][1120/1251] eta 0:00:37 lr 0.000855 time 0.2844 (0.2875) loss 3.0252 (3.8272) grad_norm 1.2907 (1.1692) [2021-04-15 21:25:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][1130/1251] eta 0:00:34 lr 0.000855 time 0.2588 (0.2874) loss 4.3600 (3.8280) grad_norm 1.1847 (1.1690) [2021-04-15 21:25:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][1140/1251] eta 0:00:31 lr 0.000855 time 0.3011 (0.2874) loss 4.2110 (3.8277) grad_norm 1.0543 (1.1691) [2021-04-15 21:25:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][1150/1251] eta 0:00:29 lr 0.000855 time 0.2819 (0.2874) loss 4.1468 (3.8297) grad_norm 1.1130 (1.1696) [2021-04-15 21:25:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][1160/1251] eta 0:00:26 lr 0.000855 time 0.2925 (0.2875) loss 3.3486 (3.8268) grad_norm 1.1672 (1.1697) [2021-04-15 21:25:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][1170/1251] eta 0:00:23 lr 0.000855 time 0.2764 (0.2877) loss 4.6011 (3.8270) grad_norm 1.2780 (1.1702) [2021-04-15 21:25:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][1180/1251] eta 0:00:20 lr 0.000855 time 0.2910 (0.2877) loss 4.3985 (3.8274) grad_norm 1.0492 (1.1695) [2021-04-15 21:25:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][1190/1251] eta 0:00:17 lr 0.000855 time 0.2621 (0.2875) loss 3.5348 (3.8259) grad_norm 1.0819 (1.1693) [2021-04-15 21:25:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][1200/1251] eta 0:00:14 lr 0.000855 time 0.2668 (0.2874) loss 4.3835 (3.8249) grad_norm 1.1058 (1.1691) [2021-04-15 21:25:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][1210/1251] eta 0:00:11 lr 0.000855 time 0.2613 (0.2874) loss 4.0863 (3.8251) grad_norm 1.1375 (1.1688) [2021-04-15 21:25:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][1220/1251] eta 0:00:08 lr 0.000855 time 0.2719 (0.2873) loss 4.1135 (3.8257) grad_norm 1.0578 (1.1685) [2021-04-15 21:25:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][1230/1251] eta 0:00:06 lr 0.000855 time 0.3225 (0.2873) loss 2.7359 (3.8245) grad_norm 1.2133 (1.1680) [2021-04-15 21:25:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][1240/1251] eta 0:00:03 lr 0.000855 time 0.3185 (0.2872) loss 4.0081 (3.8256) grad_norm 1.1667 (1.1677) [2021-04-15 21:25:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [74/300][1250/1251] eta 0:00:00 lr 0.000855 time 0.2489 (0.2869) loss 4.7122 (3.8254) grad_norm 1.0995 (1.1677) [2021-04-15 21:25:58 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 74 training takes 0:06:01 [2021-04-15 21:25:58 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_74.pth saving...... [2021-04-15 21:26:07 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_74.pth saved !!! [2021-04-15 21:26:08 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.191 (1.191) Loss 1.2234 (1.2234) Acc@1 72.266 (72.266) Acc@5 90.820 (90.820) [2021-04-15 21:26:09 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.120 (0.234) Loss 1.1135 (1.2053) Acc@1 73.242 (71.928) Acc@5 93.262 (91.371) [2021-04-15 21:26:12 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.107 (0.249) Loss 1.1548 (1.1963) Acc@1 74.219 (72.373) Acc@5 91.309 (91.392) [2021-04-15 21:26:14 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.127 (0.228) Loss 1.1689 (1.1961) Acc@1 72.754 (72.162) Acc@5 92.285 (91.564) [2021-04-15 21:26:16 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.217 (0.219) Loss 1.2503 (1.2037) Acc@1 71.680 (72.042) Acc@5 90.234 (91.506) [2021-04-15 21:26:19 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 72.126 Acc@5 91.540 [2021-04-15 21:26:19 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 72.1% [2021-04-15 21:26:19 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 72.13% [2021-04-15 21:26:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][0/1251] eta 1:22:38 lr 0.000855 time 3.9635 (3.9635) loss 3.6430 (3.6430) grad_norm 1.0947 (1.0947) [2021-04-15 21:26:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][10/1251] eta 0:12:40 lr 0.000855 time 0.2807 (0.6128) loss 4.2441 (3.7096) grad_norm 1.0854 (1.2103) [2021-04-15 21:26:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][20/1251] eta 0:09:17 lr 0.000855 time 0.2580 (0.4527) loss 3.2481 (3.6065) grad_norm 1.3702 (1.1810) [2021-04-15 21:26:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][30/1251] eta 0:08:06 lr 0.000855 time 0.2857 (0.3987) loss 3.9254 (3.7397) grad_norm 1.0484 (1.1590) [2021-04-15 21:26:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][40/1251] eta 0:07:26 lr 0.000855 time 0.2796 (0.3687) loss 3.5636 (3.7336) grad_norm 1.2080 (1.1663) [2021-04-15 21:26:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][50/1251] eta 0:07:00 lr 0.000855 time 0.2663 (0.3503) loss 4.1846 (3.8095) grad_norm 1.1363 (1.1623) [2021-04-15 21:26:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][60/1251] eta 0:06:43 lr 0.000855 time 0.2804 (0.3392) loss 3.9064 (3.8124) grad_norm 1.0611 (1.1574) [2021-04-15 21:26:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][70/1251] eta 0:06:30 lr 0.000855 time 0.3120 (0.3307) loss 4.7952 (3.8314) grad_norm 1.1005 (1.1484) [2021-04-15 21:26:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][80/1251] eta 0:06:20 lr 0.000855 time 0.2624 (0.3250) loss 3.8506 (3.8399) grad_norm 1.5181 (1.1596) [2021-04-15 21:26:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][90/1251] eta 0:06:10 lr 0.000855 time 0.2613 (0.3194) loss 4.2164 (3.8227) grad_norm 1.0917 (1.1591) [2021-04-15 21:26:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][100/1251] eta 0:06:02 lr 0.000855 time 0.2899 (0.3153) loss 4.4942 (3.8158) grad_norm 1.1288 (1.1597) [2021-04-15 21:26:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][110/1251] eta 0:05:55 lr 0.000855 time 0.2836 (0.3117) loss 4.5362 (3.8148) grad_norm 1.3365 (1.1620) [2021-04-15 21:26:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][120/1251] eta 0:05:50 lr 0.000855 time 0.2686 (0.3101) loss 3.8309 (3.8099) grad_norm 1.5845 (1.1690) [2021-04-15 21:27:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][130/1251] eta 0:05:45 lr 0.000855 time 0.2711 (0.3086) loss 3.5904 (3.8042) grad_norm 1.2122 (1.1746) [2021-04-15 21:27:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][140/1251] eta 0:05:41 lr 0.000855 time 0.2704 (0.3070) loss 4.6982 (3.8179) grad_norm 0.9801 (1.1743) [2021-04-15 21:27:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][150/1251] eta 0:05:36 lr 0.000855 time 0.2930 (0.3054) loss 3.6996 (3.8054) grad_norm 1.0069 (1.1685) [2021-04-15 21:27:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][160/1251] eta 0:05:31 lr 0.000855 time 0.2866 (0.3036) loss 3.2644 (3.8072) grad_norm 1.3177 (1.1711) [2021-04-15 21:27:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][170/1251] eta 0:05:26 lr 0.000855 time 0.2588 (0.3018) loss 3.6865 (3.8297) grad_norm 1.2745 (1.1733) [2021-04-15 21:27:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][180/1251] eta 0:05:22 lr 0.000854 time 0.3045 (0.3009) loss 3.0650 (3.8122) grad_norm 1.0984 (1.1712) [2021-04-15 21:27:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][190/1251] eta 0:05:17 lr 0.000854 time 0.2651 (0.2994) loss 3.6367 (3.8130) grad_norm 1.1241 (1.1669) [2021-04-15 21:27:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][200/1251] eta 0:05:14 lr 0.000854 time 0.2873 (0.2991) loss 3.8460 (3.8095) grad_norm 1.5365 (1.1752) [2021-04-15 21:27:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][210/1251] eta 0:05:10 lr 0.000854 time 0.2701 (0.2979) loss 4.0429 (3.8025) grad_norm 1.1356 (1.1753) [2021-04-15 21:27:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][220/1251] eta 0:05:06 lr 0.000854 time 0.2921 (0.2969) loss 3.9941 (3.8072) grad_norm 1.1298 (1.1729) [2021-04-15 21:27:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][230/1251] eta 0:05:02 lr 0.000854 time 0.2701 (0.2961) loss 3.2950 (3.8037) grad_norm 1.2863 (1.1719) [2021-04-15 21:27:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][240/1251] eta 0:04:59 lr 0.000854 time 0.2599 (0.2958) loss 4.3058 (3.7992) grad_norm 1.1397 (1.1717) [2021-04-15 21:27:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][250/1251] eta 0:04:55 lr 0.000854 time 0.2825 (0.2950) loss 3.8992 (3.8057) grad_norm 1.1141 (1.1735) [2021-04-15 21:27:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][260/1251] eta 0:04:51 lr 0.000854 time 0.3012 (0.2943) loss 3.4004 (3.8032) grad_norm 1.0909 (1.1749) [2021-04-15 21:27:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][270/1251] eta 0:04:48 lr 0.000854 time 0.2565 (0.2937) loss 4.5966 (3.8095) grad_norm 1.1093 (1.1761) [2021-04-15 21:27:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][280/1251] eta 0:04:44 lr 0.000854 time 0.2836 (0.2932) loss 2.7424 (3.8133) grad_norm 0.9278 (1.1744) [2021-04-15 21:27:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][290/1251] eta 0:04:41 lr 0.000854 time 0.2445 (0.2924) loss 4.4896 (3.8133) grad_norm 1.2157 (1.1732) [2021-04-15 21:27:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][300/1251] eta 0:04:37 lr 0.000854 time 0.2905 (0.2919) loss 4.1103 (3.8070) grad_norm 1.0879 (1.1723) [2021-04-15 21:27:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][310/1251] eta 0:04:34 lr 0.000854 time 0.2902 (0.2916) loss 3.7103 (3.8100) grad_norm 1.0487 (1.1711) [2021-04-15 21:27:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][320/1251] eta 0:04:31 lr 0.000854 time 0.2933 (0.2911) loss 3.4855 (3.8094) grad_norm 1.2301 (1.1711) [2021-04-15 21:27:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][330/1251] eta 0:04:27 lr 0.000854 time 0.2695 (0.2908) loss 3.7623 (3.8010) grad_norm 1.0843 (1.1740) [2021-04-15 21:27:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][340/1251] eta 0:04:24 lr 0.000854 time 0.2811 (0.2904) loss 4.1081 (3.8074) grad_norm 1.0303 (1.1735) [2021-04-15 21:28:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][350/1251] eta 0:04:21 lr 0.000854 time 0.4214 (0.2905) loss 2.7886 (3.8102) grad_norm 1.2741 (1.1729) [2021-04-15 21:28:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][360/1251] eta 0:04:19 lr 0.000854 time 0.2988 (0.2907) loss 4.5869 (3.8128) grad_norm 1.3476 (1.1754) [2021-04-15 21:28:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][370/1251] eta 0:04:16 lr 0.000854 time 0.2999 (0.2907) loss 3.8750 (3.8151) grad_norm 1.0904 (1.1773) [2021-04-15 21:28:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][380/1251] eta 0:04:13 lr 0.000854 time 0.2711 (0.2909) loss 3.0597 (3.8196) grad_norm 1.4923 (1.1789) [2021-04-15 21:28:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][390/1251] eta 0:04:10 lr 0.000854 time 0.2689 (0.2905) loss 3.7759 (3.8170) grad_norm 1.2051 (1.1786) [2021-04-15 21:28:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][400/1251] eta 0:04:06 lr 0.000854 time 0.3010 (0.2902) loss 4.4213 (3.8199) grad_norm 1.2348 (1.1769) [2021-04-15 21:28:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][410/1251] eta 0:04:03 lr 0.000854 time 0.2581 (0.2898) loss 3.1658 (3.8230) grad_norm 1.2615 (1.1784) [2021-04-15 21:28:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][420/1251] eta 0:04:00 lr 0.000854 time 0.2929 (0.2895) loss 4.5346 (3.8202) grad_norm 1.1974 (1.1798) [2021-04-15 21:28:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][430/1251] eta 0:03:57 lr 0.000854 time 0.2800 (0.2892) loss 4.4758 (3.8188) grad_norm 1.2004 (1.1794) [2021-04-15 21:28:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][440/1251] eta 0:03:54 lr 0.000854 time 0.2890 (0.2889) loss 4.0902 (3.8195) grad_norm 1.3090 (1.1793) [2021-04-15 21:28:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][450/1251] eta 0:03:51 lr 0.000854 time 0.2656 (0.2888) loss 4.2488 (3.8218) grad_norm 1.3085 (1.1778) [2021-04-15 21:28:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][460/1251] eta 0:03:48 lr 0.000854 time 0.2814 (0.2885) loss 4.5436 (3.8287) grad_norm 1.1339 (1.1768) [2021-04-15 21:28:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][470/1251] eta 0:03:45 lr 0.000854 time 0.2951 (0.2882) loss 3.1978 (3.8278) grad_norm 1.0726 (1.1763) [2021-04-15 21:28:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][480/1251] eta 0:03:42 lr 0.000854 time 0.2726 (0.2880) loss 4.2873 (3.8299) grad_norm 1.2334 (1.1751) [2021-04-15 21:28:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][490/1251] eta 0:03:39 lr 0.000854 time 0.2450 (0.2878) loss 3.5346 (3.8267) grad_norm 1.6848 (1.1774) [2021-04-15 21:28:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][500/1251] eta 0:03:35 lr 0.000854 time 0.2676 (0.2876) loss 3.9149 (3.8256) grad_norm 1.0103 (1.1767) [2021-04-15 21:28:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][510/1251] eta 0:03:32 lr 0.000854 time 0.3059 (0.2874) loss 3.2101 (3.8269) grad_norm 1.2546 (1.1772) [2021-04-15 21:28:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][520/1251] eta 0:03:30 lr 0.000853 time 0.2839 (0.2875) loss 4.6232 (3.8290) grad_norm 1.0658 (1.1770) [2021-04-15 21:28:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][530/1251] eta 0:03:27 lr 0.000853 time 0.2841 (0.2873) loss 3.6663 (3.8327) grad_norm 1.2556 (1.1773) [2021-04-15 21:28:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][540/1251] eta 0:03:24 lr 0.000853 time 0.2602 (0.2871) loss 3.7408 (3.8324) grad_norm 1.0674 (1.1773) [2021-04-15 21:28:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][550/1251] eta 0:03:21 lr 0.000853 time 0.2482 (0.2870) loss 4.1249 (3.8369) grad_norm 1.1902 (1.1769) [2021-04-15 21:29:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][560/1251] eta 0:03:18 lr 0.000853 time 0.2758 (0.2869) loss 4.3049 (3.8394) grad_norm 1.2920 (1.1781) [2021-04-15 21:29:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][570/1251] eta 0:03:15 lr 0.000853 time 0.2959 (0.2869) loss 4.2000 (3.8400) grad_norm 1.1767 (1.1786) [2021-04-15 21:29:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][580/1251] eta 0:03:12 lr 0.000853 time 0.2749 (0.2867) loss 3.6047 (3.8416) grad_norm 1.1331 (1.1784) [2021-04-15 21:29:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][590/1251] eta 0:03:09 lr 0.000853 time 0.2639 (0.2865) loss 3.0846 (3.8416) grad_norm 1.1876 (1.1790) [2021-04-15 21:29:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][600/1251] eta 0:03:06 lr 0.000853 time 0.2675 (0.2864) loss 3.7321 (3.8419) grad_norm 1.0913 (1.1795) [2021-04-15 21:29:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][610/1251] eta 0:03:03 lr 0.000853 time 0.2635 (0.2861) loss 4.4590 (3.8450) grad_norm 1.3953 (1.1799) [2021-04-15 21:29:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][620/1251] eta 0:03:00 lr 0.000853 time 0.2779 (0.2860) loss 2.9613 (3.8467) grad_norm 1.0894 (1.1788) [2021-04-15 21:29:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][630/1251] eta 0:02:57 lr 0.000853 time 0.2782 (0.2859) loss 3.6429 (3.8463) grad_norm 1.1593 (1.1783) [2021-04-15 21:29:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][640/1251] eta 0:02:54 lr 0.000853 time 0.2922 (0.2857) loss 4.3343 (3.8464) grad_norm 1.1538 (1.1774) [2021-04-15 21:29:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][650/1251] eta 0:02:51 lr 0.000853 time 0.2921 (0.2856) loss 3.6489 (3.8473) grad_norm 1.0953 (1.1767) [2021-04-15 21:29:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][660/1251] eta 0:02:48 lr 0.000853 time 0.2730 (0.2854) loss 4.3593 (3.8460) grad_norm 1.1125 (1.1758) [2021-04-15 21:29:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][670/1251] eta 0:02:45 lr 0.000853 time 0.2708 (0.2853) loss 4.3538 (3.8446) grad_norm 1.1525 (1.1751) [2021-04-15 21:29:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][680/1251] eta 0:02:42 lr 0.000853 time 0.3006 (0.2852) loss 4.2420 (3.8487) grad_norm 1.0793 (1.1766) [2021-04-15 21:29:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][690/1251] eta 0:02:40 lr 0.000853 time 0.2661 (0.2852) loss 4.0273 (3.8469) grad_norm 1.3855 (1.1760) [2021-04-15 21:29:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][700/1251] eta 0:02:37 lr 0.000853 time 0.2599 (0.2851) loss 4.1038 (3.8479) grad_norm 1.0519 (1.1748) [2021-04-15 21:29:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][710/1251] eta 0:02:34 lr 0.000853 time 0.2712 (0.2852) loss 3.5483 (3.8461) grad_norm 1.0209 (1.1744) [2021-04-15 21:29:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][720/1251] eta 0:02:31 lr 0.000853 time 0.2851 (0.2851) loss 3.8922 (3.8441) grad_norm 1.4620 (1.1761) [2021-04-15 21:29:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][730/1251] eta 0:02:28 lr 0.000853 time 0.2611 (0.2850) loss 3.0082 (3.8429) grad_norm 1.2273 (1.1774) [2021-04-15 21:29:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][740/1251] eta 0:02:25 lr 0.000853 time 0.2532 (0.2848) loss 3.8899 (3.8456) grad_norm 0.9241 (1.1780) [2021-04-15 21:29:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][750/1251] eta 0:02:22 lr 0.000853 time 0.2671 (0.2848) loss 4.2545 (3.8414) grad_norm 1.4548 (1.1780) [2021-04-15 21:29:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][760/1251] eta 0:02:19 lr 0.000853 time 0.2794 (0.2847) loss 3.9579 (3.8366) grad_norm 1.0177 (1.1774) [2021-04-15 21:29:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][770/1251] eta 0:02:16 lr 0.000853 time 0.2639 (0.2846) loss 3.5907 (3.8354) grad_norm 1.0171 (1.1768) [2021-04-15 21:30:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][780/1251] eta 0:02:14 lr 0.000853 time 0.2757 (0.2847) loss 3.5504 (3.8353) grad_norm 0.9601 (1.1759) [2021-04-15 21:30:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][790/1251] eta 0:02:11 lr 0.000853 time 0.2910 (0.2847) loss 3.8540 (3.8370) grad_norm 1.2093 (1.1760) [2021-04-15 21:30:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][800/1251] eta 0:02:08 lr 0.000853 time 0.2850 (0.2846) loss 2.8954 (3.8357) grad_norm 1.2194 (1.1772) [2021-04-15 21:30:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][810/1251] eta 0:02:05 lr 0.000853 time 0.2923 (0.2845) loss 3.7249 (3.8347) grad_norm 1.0388 (1.1765) [2021-04-15 21:30:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][820/1251] eta 0:02:02 lr 0.000853 time 0.2540 (0.2845) loss 3.6539 (3.8355) grad_norm 1.0624 (1.1773) [2021-04-15 21:30:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][830/1251] eta 0:01:59 lr 0.000853 time 0.2863 (0.2844) loss 4.3142 (3.8321) grad_norm 1.0250 (1.1773) [2021-04-15 21:30:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][840/1251] eta 0:01:56 lr 0.000853 time 0.2720 (0.2843) loss 3.7671 (3.8311) grad_norm 1.0076 (1.1774) [2021-04-15 21:30:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][850/1251] eta 0:01:53 lr 0.000853 time 0.2597 (0.2841) loss 2.6442 (3.8309) grad_norm 1.0712 (1.1772) [2021-04-15 21:30:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][860/1251] eta 0:01:51 lr 0.000852 time 0.2567 (0.2840) loss 4.0197 (3.8318) grad_norm 1.0047 (1.1775) [2021-04-15 21:30:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][870/1251] eta 0:01:48 lr 0.000852 time 0.2925 (0.2839) loss 3.7428 (3.8275) grad_norm 0.9791 (1.1765) [2021-04-15 21:30:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][880/1251] eta 0:01:45 lr 0.000852 time 0.2657 (0.2839) loss 2.8044 (3.8274) grad_norm 1.1121 (1.1761) [2021-04-15 21:30:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][890/1251] eta 0:01:42 lr 0.000852 time 0.2638 (0.2838) loss 3.2712 (3.8276) grad_norm 1.1188 (1.1760) [2021-04-15 21:30:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][900/1251] eta 0:01:39 lr 0.000852 time 0.2478 (0.2837) loss 2.7617 (3.8262) grad_norm 1.2046 (1.1755) [2021-04-15 21:30:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][910/1251] eta 0:01:36 lr 0.000852 time 0.2819 (0.2836) loss 4.0057 (3.8252) grad_norm 1.1053 (1.1753) [2021-04-15 21:30:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][920/1251] eta 0:01:33 lr 0.000852 time 0.2860 (0.2836) loss 4.0291 (3.8225) grad_norm 1.1093 (1.1746) [2021-04-15 21:30:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][930/1251] eta 0:01:31 lr 0.000852 time 0.2860 (0.2836) loss 3.8597 (3.8201) grad_norm 1.1372 (1.1748) [2021-04-15 21:30:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][940/1251] eta 0:01:28 lr 0.000852 time 0.4514 (0.2838) loss 3.9442 (3.8206) grad_norm 1.1817 (1.1759) [2021-04-15 21:30:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][950/1251] eta 0:01:25 lr 0.000852 time 0.2800 (0.2838) loss 3.9615 (3.8207) grad_norm 1.1839 (1.1753) [2021-04-15 21:30:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][960/1251] eta 0:01:22 lr 0.000852 time 0.3052 (0.2839) loss 4.0876 (3.8190) grad_norm 1.1295 (1.1755) [2021-04-15 21:30:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][970/1251] eta 0:01:19 lr 0.000852 time 0.2969 (0.2838) loss 3.8014 (3.8195) grad_norm 1.2179 (1.1752) [2021-04-15 21:30:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][980/1251] eta 0:01:16 lr 0.000852 time 0.2741 (0.2837) loss 3.8410 (3.8174) grad_norm 1.3751 (1.1758) [2021-04-15 21:31:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][990/1251] eta 0:01:14 lr 0.000852 time 0.2805 (0.2836) loss 3.9053 (3.8179) grad_norm 1.2996 (1.1760) [2021-04-15 21:31:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][1000/1251] eta 0:01:11 lr 0.000852 time 0.3000 (0.2836) loss 3.0453 (3.8158) grad_norm 1.6353 (1.1765) [2021-04-15 21:31:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][1010/1251] eta 0:01:08 lr 0.000852 time 0.2992 (0.2836) loss 4.3624 (3.8174) grad_norm 1.2393 (1.1762) [2021-04-15 21:31:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][1020/1251] eta 0:01:05 lr 0.000852 time 0.2682 (0.2835) loss 3.0062 (3.8179) grad_norm 1.1778 (1.1767) [2021-04-15 21:31:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][1030/1251] eta 0:01:02 lr 0.000852 time 0.2509 (0.2836) loss 4.5958 (3.8210) grad_norm 1.1515 (1.1764) [2021-04-15 21:31:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][1040/1251] eta 0:00:59 lr 0.000852 time 0.2796 (0.2835) loss 4.0253 (3.8203) grad_norm 1.1012 (1.1761) [2021-04-15 21:31:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][1050/1251] eta 0:00:56 lr 0.000852 time 0.3006 (0.2835) loss 4.6039 (3.8213) grad_norm 1.0730 (1.1760) [2021-04-15 21:31:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][1060/1251] eta 0:00:54 lr 0.000852 time 0.2753 (0.2834) loss 4.4242 (3.8231) grad_norm 1.3151 (1.1760) [2021-04-15 21:31:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][1070/1251] eta 0:00:51 lr 0.000852 time 0.3018 (0.2833) loss 3.6084 (3.8236) grad_norm 1.2954 (1.1761) [2021-04-15 21:31:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][1080/1251] eta 0:00:48 lr 0.000852 time 0.2868 (0.2833) loss 4.5731 (3.8248) grad_norm 1.0997 (1.1762) [2021-04-15 21:31:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][1090/1251] eta 0:00:45 lr 0.000852 time 0.2411 (0.2834) loss 3.7470 (3.8236) grad_norm 1.2208 (1.1755) [2021-04-15 21:31:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][1100/1251] eta 0:00:42 lr 0.000852 time 0.2701 (0.2833) loss 2.7324 (3.8213) grad_norm 1.1521 (1.1759) [2021-04-15 21:31:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][1110/1251] eta 0:00:39 lr 0.000852 time 0.2803 (0.2833) loss 4.4261 (3.8216) grad_norm 1.1712 (1.1761) [2021-04-15 21:31:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][1120/1251] eta 0:00:37 lr 0.000852 time 0.2861 (0.2832) loss 3.8914 (3.8211) grad_norm 0.9858 (1.1766) [2021-04-15 21:31:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][1130/1251] eta 0:00:34 lr 0.000852 time 0.2584 (0.2831) loss 2.6058 (3.8204) grad_norm 1.1605 (1.1772) [2021-04-15 21:31:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][1140/1251] eta 0:00:31 lr 0.000852 time 0.2781 (0.2832) loss 4.1240 (3.8220) grad_norm 1.1023 (1.1769) [2021-04-15 21:31:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][1150/1251] eta 0:00:28 lr 0.000852 time 0.2846 (0.2831) loss 4.4699 (3.8190) grad_norm 0.9734 (1.1763) [2021-04-15 21:31:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][1160/1251] eta 0:00:25 lr 0.000852 time 0.2758 (0.2830) loss 4.0790 (3.8208) grad_norm 1.1059 (1.1758) [2021-04-15 21:31:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][1170/1251] eta 0:00:22 lr 0.000852 time 0.2737 (0.2829) loss 3.6367 (3.8214) grad_norm 1.3345 (1.1757) [2021-04-15 21:31:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][1180/1251] eta 0:00:20 lr 0.000852 time 0.2784 (0.2829) loss 3.7111 (3.8215) grad_norm 1.1538 (1.1764) [2021-04-15 21:31:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][1190/1251] eta 0:00:17 lr 0.000852 time 0.3053 (0.2828) loss 4.5611 (3.8209) grad_norm 1.0371 (1.1767) [2021-04-15 21:31:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][1200/1251] eta 0:00:14 lr 0.000851 time 0.2859 (0.2828) loss 4.2929 (3.8217) grad_norm 1.0517 (1.1764) [2021-04-15 21:32:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][1210/1251] eta 0:00:11 lr 0.000851 time 0.2688 (0.2827) loss 4.4970 (3.8204) grad_norm 1.2377 (1.1766) [2021-04-15 21:32:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][1220/1251] eta 0:00:08 lr 0.000851 time 0.2990 (0.2827) loss 4.3909 (3.8223) grad_norm 1.1626 (1.1766) [2021-04-15 21:32:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][1230/1251] eta 0:00:05 lr 0.000851 time 0.2564 (0.2826) loss 3.9396 (3.8211) grad_norm 1.1299 (1.1768) [2021-04-15 21:32:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][1240/1251] eta 0:00:03 lr 0.000851 time 0.2600 (0.2825) loss 3.1541 (3.8200) grad_norm 1.4172 (1.1772) [2021-04-15 21:32:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [75/300][1250/1251] eta 0:00:00 lr 0.000851 time 0.2502 (0.2822) loss 4.0226 (3.8204) grad_norm 1.2010 (1.1770) [2021-04-15 21:32:14 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 75 training takes 0:05:55 [2021-04-15 21:32:14 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_75.pth saving...... [2021-04-15 21:32:29 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_75.pth saved !!! [2021-04-15 21:32:30 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.197 (1.197) Loss 1.1318 (1.1318) Acc@1 73.633 (73.633) Acc@5 91.797 (91.797) [2021-04-15 21:32:32 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.126 (0.250) Loss 1.1377 (1.2043) Acc@1 73.535 (72.079) Acc@5 91.406 (90.936) [2021-04-15 21:32:33 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.130 (0.215) Loss 1.2653 (1.2000) Acc@1 72.168 (72.173) Acc@5 90.234 (91.216) [2021-04-15 21:32:36 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.100 (0.237) Loss 1.0806 (1.2002) Acc@1 74.902 (72.140) Acc@5 92.578 (91.195) [2021-04-15 21:32:38 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.089 (0.213) Loss 1.2062 (1.1951) Acc@1 71.387 (72.173) Acc@5 90.625 (91.309) [2021-04-15 21:32:41 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 72.192 Acc@5 91.240 [2021-04-15 21:32:41 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 72.2% [2021-04-15 21:32:41 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 72.19% [2021-04-15 21:32:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][0/1251] eta 1:20:02 lr 0.000851 time 3.8388 (3.8388) loss 4.1822 (4.1822) grad_norm 1.2990 (1.2990) [2021-04-15 21:32:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][10/1251] eta 0:12:23 lr 0.000851 time 0.2734 (0.5987) loss 4.5167 (3.7701) grad_norm 1.0724 (1.1450) [2021-04-15 21:32:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][20/1251] eta 0:09:18 lr 0.000851 time 0.2824 (0.4533) loss 2.8007 (3.6801) grad_norm 1.0542 (1.1675) [2021-04-15 21:32:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][30/1251] eta 0:08:04 lr 0.000851 time 0.2816 (0.3966) loss 3.8937 (3.6784) grad_norm 1.3906 (1.1945) [2021-04-15 21:32:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][40/1251] eta 0:07:24 lr 0.000851 time 0.2443 (0.3674) loss 3.7001 (3.7342) grad_norm 1.3435 (1.2097) [2021-04-15 21:32:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][50/1251] eta 0:06:59 lr 0.000851 time 0.2733 (0.3497) loss 3.3221 (3.7063) grad_norm 1.1596 (1.2040) [2021-04-15 21:33:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][60/1251] eta 0:06:43 lr 0.000851 time 0.2724 (0.3390) loss 3.9499 (3.7013) grad_norm 1.3803 (1.2119) [2021-04-15 21:33:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][70/1251] eta 0:06:29 lr 0.000851 time 0.2664 (0.3300) loss 2.9616 (3.6968) grad_norm 1.1772 (1.2156) [2021-04-15 21:33:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][80/1251] eta 0:06:19 lr 0.000851 time 0.2720 (0.3237) loss 4.1045 (3.6747) grad_norm 1.1205 (1.2081) [2021-04-15 21:33:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][90/1251] eta 0:06:10 lr 0.000851 time 0.2810 (0.3189) loss 4.0232 (3.7200) grad_norm 1.2803 (1.2066) [2021-04-15 21:33:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][100/1251] eta 0:06:01 lr 0.000851 time 0.2453 (0.3144) loss 4.3845 (3.7433) grad_norm 1.3106 (1.1986) [2021-04-15 21:33:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][110/1251] eta 0:05:55 lr 0.000851 time 0.2782 (0.3115) loss 3.9409 (3.7142) grad_norm 1.4142 (1.1970) [2021-04-15 21:33:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][120/1251] eta 0:05:49 lr 0.000851 time 0.2815 (0.3092) loss 4.3610 (3.7308) grad_norm 1.4288 (1.2057) [2021-04-15 21:33:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][130/1251] eta 0:05:44 lr 0.000851 time 0.2523 (0.3075) loss 4.0938 (3.7406) grad_norm 1.1630 (1.2067) [2021-04-15 21:33:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][140/1251] eta 0:05:40 lr 0.000851 time 0.2670 (0.3062) loss 4.0659 (3.7533) grad_norm 1.0179 (1.2114) [2021-04-15 21:33:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][150/1251] eta 0:05:36 lr 0.000851 time 0.4562 (0.3053) loss 4.5899 (3.7475) grad_norm 1.1022 (1.2080) [2021-04-15 21:33:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][160/1251] eta 0:05:31 lr 0.000851 time 0.2605 (0.3043) loss 4.3858 (3.7339) grad_norm 1.2574 (1.2057) [2021-04-15 21:33:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][170/1251] eta 0:05:28 lr 0.000851 time 0.2720 (0.3036) loss 4.2990 (3.7389) grad_norm 1.1252 (1.2067) [2021-04-15 21:33:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][180/1251] eta 0:05:23 lr 0.000851 time 0.2821 (0.3023) loss 3.9327 (3.7407) grad_norm 1.0115 (1.2067) [2021-04-15 21:33:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][190/1251] eta 0:05:19 lr 0.000851 time 0.2623 (0.3009) loss 3.0405 (3.7388) grad_norm 1.0982 (1.2029) [2021-04-15 21:33:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][200/1251] eta 0:05:14 lr 0.000851 time 0.2614 (0.2996) loss 3.6516 (3.7461) grad_norm 0.9815 (1.2006) [2021-04-15 21:33:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][210/1251] eta 0:05:10 lr 0.000851 time 0.2975 (0.2984) loss 4.2637 (3.7492) grad_norm 1.4910 (1.1994) [2021-04-15 21:33:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][220/1251] eta 0:05:06 lr 0.000851 time 0.2653 (0.2971) loss 4.1869 (3.7428) grad_norm 1.1608 (1.1977) [2021-04-15 21:33:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][230/1251] eta 0:05:02 lr 0.000851 time 0.2546 (0.2960) loss 4.0764 (3.7656) grad_norm 1.0930 (1.1929) [2021-04-15 21:33:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][240/1251] eta 0:04:59 lr 0.000851 time 0.2915 (0.2958) loss 3.8902 (3.7696) grad_norm 1.0068 (1.1870) [2021-04-15 21:33:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][250/1251] eta 0:04:55 lr 0.000851 time 0.2784 (0.2951) loss 4.0653 (3.7649) grad_norm 1.3089 (1.1870) [2021-04-15 21:33:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][260/1251] eta 0:04:51 lr 0.000851 time 0.2714 (0.2943) loss 3.3963 (3.7598) grad_norm 1.1351 (1.1846) [2021-04-15 21:34:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][270/1251] eta 0:04:48 lr 0.000851 time 0.2528 (0.2936) loss 4.1117 (3.7618) grad_norm 1.5425 (1.1849) [2021-04-15 21:34:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][280/1251] eta 0:04:44 lr 0.000851 time 0.2420 (0.2931) loss 3.3848 (3.7684) grad_norm 1.3016 (1.1874) [2021-04-15 21:34:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][290/1251] eta 0:04:41 lr 0.000850 time 0.2800 (0.2925) loss 4.5506 (3.7732) grad_norm 0.9918 (1.1908) [2021-04-15 21:34:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][300/1251] eta 0:04:38 lr 0.000850 time 0.2845 (0.2926) loss 4.3405 (3.7681) grad_norm 1.2309 (1.1930) [2021-04-15 21:34:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][310/1251] eta 0:04:34 lr 0.000850 time 0.2819 (0.2918) loss 2.7305 (3.7698) grad_norm 1.2705 (1.1937) [2021-04-15 21:34:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][320/1251] eta 0:04:31 lr 0.000850 time 0.2820 (0.2913) loss 4.1204 (3.7757) grad_norm 1.4460 (1.1966) [2021-04-15 21:34:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][330/1251] eta 0:04:27 lr 0.000850 time 0.2824 (0.2908) loss 3.3916 (3.7757) grad_norm 1.4689 (1.1993) [2021-04-15 21:34:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][340/1251] eta 0:04:24 lr 0.000850 time 0.2845 (0.2904) loss 4.6119 (3.7731) grad_norm 1.1284 (1.1988) [2021-04-15 21:34:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][350/1251] eta 0:04:21 lr 0.000850 time 0.2762 (0.2902) loss 3.4628 (3.7725) grad_norm 1.1593 (1.1972) [2021-04-15 21:34:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][360/1251] eta 0:04:18 lr 0.000850 time 0.2820 (0.2899) loss 3.8380 (3.7733) grad_norm 1.3981 (1.1963) [2021-04-15 21:34:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][370/1251] eta 0:04:15 lr 0.000850 time 0.2488 (0.2895) loss 4.2369 (3.7810) grad_norm 1.3159 (1.1946) [2021-04-15 21:34:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][380/1251] eta 0:04:11 lr 0.000850 time 0.2962 (0.2892) loss 3.0389 (3.7848) grad_norm 1.1079 (1.1939) [2021-04-15 21:34:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][390/1251] eta 0:04:08 lr 0.000850 time 0.2730 (0.2888) loss 3.1421 (3.7816) grad_norm 1.1394 (1.1916) [2021-04-15 21:34:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][400/1251] eta 0:04:05 lr 0.000850 time 0.2605 (0.2885) loss 4.5482 (3.7832) grad_norm 1.1494 (1.1904) [2021-04-15 21:34:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][410/1251] eta 0:04:02 lr 0.000850 time 0.2733 (0.2881) loss 3.3165 (3.7820) grad_norm 0.9804 (1.1876) [2021-04-15 21:34:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][420/1251] eta 0:03:59 lr 0.000850 time 0.2559 (0.2878) loss 3.3258 (3.7818) grad_norm 1.1795 (1.1865) [2021-04-15 21:34:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][430/1251] eta 0:03:56 lr 0.000850 time 0.2654 (0.2875) loss 3.9312 (3.7821) grad_norm 1.6571 (1.1886) [2021-04-15 21:34:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][440/1251] eta 0:03:52 lr 0.000850 time 0.2808 (0.2873) loss 3.6669 (3.7826) grad_norm 1.2153 (1.1891) [2021-04-15 21:34:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][450/1251] eta 0:03:49 lr 0.000850 time 0.2708 (0.2870) loss 4.0098 (3.7833) grad_norm 0.9977 (1.1882) [2021-04-15 21:34:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][460/1251] eta 0:03:46 lr 0.000850 time 0.2894 (0.2869) loss 4.2660 (3.7861) grad_norm 1.0388 (1.1873) [2021-04-15 21:34:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][470/1251] eta 0:03:43 lr 0.000850 time 0.2634 (0.2867) loss 4.2924 (3.7884) grad_norm 1.0856 (1.1863) [2021-04-15 21:34:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][480/1251] eta 0:03:40 lr 0.000850 time 0.2772 (0.2864) loss 2.5962 (3.7871) grad_norm 1.2291 (1.1843) [2021-04-15 21:35:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][490/1251] eta 0:03:37 lr 0.000850 time 0.2807 (0.2861) loss 3.7324 (3.7901) grad_norm 1.1573 (1.1845) [2021-04-15 21:35:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][500/1251] eta 0:03:34 lr 0.000850 time 0.2771 (0.2859) loss 2.6655 (3.7956) grad_norm 1.0828 (1.1846) [2021-04-15 21:35:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][510/1251] eta 0:03:31 lr 0.000850 time 0.2744 (0.2857) loss 4.1771 (3.7978) grad_norm 1.2170 (1.1838) [2021-04-15 21:35:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][520/1251] eta 0:03:28 lr 0.000850 time 0.2488 (0.2857) loss 4.7268 (3.8041) grad_norm 1.4318 (1.1842) [2021-04-15 21:35:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][530/1251] eta 0:03:25 lr 0.000850 time 0.2638 (0.2855) loss 4.1954 (3.8060) grad_norm 1.1423 (1.1836) [2021-04-15 21:35:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][540/1251] eta 0:03:22 lr 0.000850 time 0.2855 (0.2854) loss 3.3674 (3.8037) grad_norm 1.1005 (1.1834) [2021-04-15 21:35:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][550/1251] eta 0:03:19 lr 0.000850 time 0.2839 (0.2852) loss 4.4499 (3.8062) grad_norm 1.0444 (1.1839) [2021-04-15 21:35:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][560/1251] eta 0:03:17 lr 0.000850 time 0.2922 (0.2852) loss 4.6565 (3.8075) grad_norm 0.9776 (1.1826) [2021-04-15 21:35:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][570/1251] eta 0:03:14 lr 0.000850 time 0.2441 (0.2854) loss 3.4297 (3.8081) grad_norm 1.1518 (1.1827) [2021-04-15 21:35:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][580/1251] eta 0:03:11 lr 0.000850 time 0.2485 (0.2852) loss 3.8769 (3.8113) grad_norm 1.2537 (1.1831) [2021-04-15 21:35:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][590/1251] eta 0:03:08 lr 0.000850 time 0.2859 (0.2853) loss 4.2035 (3.8121) grad_norm 1.1577 (1.1837) [2021-04-15 21:35:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][600/1251] eta 0:03:05 lr 0.000850 time 0.2680 (0.2851) loss 3.3837 (3.8121) grad_norm 1.1059 (1.1828) [2021-04-15 21:35:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][610/1251] eta 0:03:02 lr 0.000850 time 0.2655 (0.2850) loss 4.1917 (3.8136) grad_norm 1.2647 (1.1831) [2021-04-15 21:35:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][620/1251] eta 0:02:59 lr 0.000849 time 0.3063 (0.2849) loss 4.0020 (3.8116) grad_norm 1.5024 (1.1839) [2021-04-15 21:35:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][630/1251] eta 0:02:56 lr 0.000849 time 0.2571 (0.2847) loss 4.5663 (3.8130) grad_norm 1.0364 (1.1819) [2021-04-15 21:35:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][640/1251] eta 0:02:53 lr 0.000849 time 0.2635 (0.2847) loss 3.2583 (3.8106) grad_norm 1.1815 (1.1806) [2021-04-15 21:35:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][650/1251] eta 0:02:51 lr 0.000849 time 0.2632 (0.2846) loss 2.8690 (3.8105) grad_norm 0.9984 (1.1803) [2021-04-15 21:35:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][660/1251] eta 0:02:48 lr 0.000849 time 0.2822 (0.2846) loss 2.7608 (3.8095) grad_norm 1.1693 (1.1799) [2021-04-15 21:35:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][670/1251] eta 0:02:45 lr 0.000849 time 0.2699 (0.2845) loss 4.7177 (3.8066) grad_norm 1.1627 (1.1793) [2021-04-15 21:35:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][680/1251] eta 0:02:42 lr 0.000849 time 0.2705 (0.2845) loss 4.1596 (3.8069) grad_norm 1.2786 (1.1802) [2021-04-15 21:35:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][690/1251] eta 0:02:39 lr 0.000849 time 0.2971 (0.2844) loss 4.3737 (3.8055) grad_norm 1.0157 (1.1796) [2021-04-15 21:36:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][700/1251] eta 0:02:36 lr 0.000849 time 0.2623 (0.2843) loss 3.0705 (3.8033) grad_norm 1.0167 (1.1784) [2021-04-15 21:36:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][710/1251] eta 0:02:33 lr 0.000849 time 0.2804 (0.2844) loss 3.7466 (3.8037) grad_norm 1.2342 (1.1779) [2021-04-15 21:36:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][720/1251] eta 0:02:31 lr 0.000849 time 0.2723 (0.2845) loss 3.6118 (3.8055) grad_norm 1.1333 (1.1787) [2021-04-15 21:36:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][730/1251] eta 0:02:28 lr 0.000849 time 0.4178 (0.2848) loss 4.1495 (3.8048) grad_norm 1.4809 (1.1796) [2021-04-15 21:36:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][740/1251] eta 0:02:25 lr 0.000849 time 0.2591 (0.2848) loss 4.2898 (3.8049) grad_norm 1.0503 (1.1800) [2021-04-15 21:36:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][750/1251] eta 0:02:22 lr 0.000849 time 0.2742 (0.2848) loss 4.3353 (3.8038) grad_norm 1.2813 (1.1795) [2021-04-15 21:36:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][760/1251] eta 0:02:19 lr 0.000849 time 0.2878 (0.2847) loss 3.9493 (3.8087) grad_norm 1.1611 (1.1793) [2021-04-15 21:36:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][770/1251] eta 0:02:16 lr 0.000849 time 0.2809 (0.2846) loss 4.0966 (3.8100) grad_norm 1.4628 (1.1803) [2021-04-15 21:36:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][780/1251] eta 0:02:14 lr 0.000849 time 0.2717 (0.2845) loss 3.5405 (3.8111) grad_norm 1.0858 (1.1805) [2021-04-15 21:36:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][790/1251] eta 0:02:11 lr 0.000849 time 0.2817 (0.2844) loss 3.4866 (3.8140) grad_norm 1.1642 (1.1810) [2021-04-15 21:36:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][800/1251] eta 0:02:08 lr 0.000849 time 0.2909 (0.2845) loss 4.0226 (3.8122) grad_norm 1.0315 (1.1814) [2021-04-15 21:36:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][810/1251] eta 0:02:05 lr 0.000849 time 0.2630 (0.2844) loss 4.2240 (3.8119) grad_norm 1.0343 (1.1815) [2021-04-15 21:36:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][820/1251] eta 0:02:02 lr 0.000849 time 0.2790 (0.2845) loss 3.1458 (3.8085) grad_norm 1.1546 (1.1822) [2021-04-15 21:36:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][830/1251] eta 0:01:59 lr 0.000849 time 0.2888 (0.2844) loss 3.6950 (3.8096) grad_norm 0.9845 (1.1818) [2021-04-15 21:36:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][840/1251] eta 0:01:56 lr 0.000849 time 0.2623 (0.2844) loss 4.0028 (3.8108) grad_norm 1.1361 (1.1810) [2021-04-15 21:36:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][850/1251] eta 0:01:54 lr 0.000849 time 0.2618 (0.2843) loss 2.9364 (3.8119) grad_norm 1.3307 (1.1810) [2021-04-15 21:36:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][860/1251] eta 0:01:51 lr 0.000849 time 0.2892 (0.2843) loss 4.3121 (3.8162) grad_norm 1.1270 (1.1803) [2021-04-15 21:36:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][870/1251] eta 0:01:48 lr 0.000849 time 0.3140 (0.2843) loss 3.4819 (3.8174) grad_norm 1.0582 (1.1801) [2021-04-15 21:36:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][880/1251] eta 0:01:45 lr 0.000849 time 0.2719 (0.2844) loss 4.6702 (3.8179) grad_norm 1.1165 (1.1806) [2021-04-15 21:36:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][890/1251] eta 0:01:42 lr 0.000849 time 0.2804 (0.2844) loss 4.4420 (3.8203) grad_norm 1.5669 (1.1811) [2021-04-15 21:36:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][900/1251] eta 0:01:39 lr 0.000849 time 0.2716 (0.2844) loss 3.9625 (3.8227) grad_norm 1.0905 (1.1806) [2021-04-15 21:37:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][910/1251] eta 0:01:36 lr 0.000849 time 0.2786 (0.2843) loss 2.7201 (3.8222) grad_norm 1.1055 (1.1799) [2021-04-15 21:37:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][920/1251] eta 0:01:34 lr 0.000849 time 0.2699 (0.2843) loss 4.5822 (3.8268) grad_norm 1.1685 (1.1795) [2021-04-15 21:37:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][930/1251] eta 0:01:31 lr 0.000849 time 0.2675 (0.2843) loss 3.5401 (3.8267) grad_norm 1.0476 (1.1791) [2021-04-15 21:37:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][940/1251] eta 0:01:28 lr 0.000849 time 0.2836 (0.2843) loss 2.4048 (3.8272) grad_norm 1.0995 (1.1794) [2021-04-15 21:37:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][950/1251] eta 0:01:25 lr 0.000849 time 0.2603 (0.2842) loss 2.9118 (3.8255) grad_norm 1.2414 (1.1792) [2021-04-15 21:37:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][960/1251] eta 0:01:22 lr 0.000848 time 0.2638 (0.2842) loss 3.1738 (3.8271) grad_norm 1.3120 (1.1791) [2021-04-15 21:37:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][970/1251] eta 0:01:19 lr 0.000848 time 0.2774 (0.2842) loss 3.0079 (3.8259) grad_norm 1.2134 (1.1781) [2021-04-15 21:37:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][980/1251] eta 0:01:17 lr 0.000848 time 0.2621 (0.2842) loss 4.3426 (3.8243) grad_norm 1.3724 (1.1789) [2021-04-15 21:37:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][990/1251] eta 0:01:14 lr 0.000848 time 0.2719 (0.2841) loss 4.7100 (3.8242) grad_norm 1.0427 (1.1786) [2021-04-15 21:37:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][1000/1251] eta 0:01:11 lr 0.000848 time 0.2700 (0.2841) loss 4.2709 (3.8229) grad_norm 1.3179 (1.1791) [2021-04-15 21:37:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][1010/1251] eta 0:01:08 lr 0.000848 time 0.2906 (0.2841) loss 4.1143 (3.8243) grad_norm 1.0144 (1.1787) [2021-04-15 21:37:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][1020/1251] eta 0:01:05 lr 0.000848 time 0.2993 (0.2840) loss 4.3630 (3.8257) grad_norm 1.1936 (1.1782) [2021-04-15 21:37:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][1030/1251] eta 0:01:02 lr 0.000848 time 0.2513 (0.2840) loss 4.2521 (3.8267) grad_norm 1.1111 (1.1781) [2021-04-15 21:37:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][1040/1251] eta 0:00:59 lr 0.000848 time 0.2927 (0.2840) loss 4.8618 (3.8292) grad_norm 1.1687 (1.1784) [2021-04-15 21:37:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][1050/1251] eta 0:00:57 lr 0.000848 time 0.2849 (0.2839) loss 2.7675 (3.8290) grad_norm 1.0241 (1.1781) [2021-04-15 21:37:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][1060/1251] eta 0:00:54 lr 0.000848 time 0.3305 (0.2840) loss 3.1503 (3.8277) grad_norm 1.1488 (1.1787) [2021-04-15 21:37:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][1070/1251] eta 0:00:51 lr 0.000848 time 0.3037 (0.2839) loss 4.5292 (3.8293) grad_norm 1.0698 (1.1788) [2021-04-15 21:37:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][1080/1251] eta 0:00:48 lr 0.000848 time 0.2755 (0.2839) loss 4.0512 (3.8307) grad_norm 0.9922 (1.1791) [2021-04-15 21:37:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][1090/1251] eta 0:00:45 lr 0.000848 time 0.2937 (0.2840) loss 3.7079 (3.8325) grad_norm 1.1336 (1.1785) [2021-04-15 21:37:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][1100/1251] eta 0:00:42 lr 0.000848 time 0.2885 (0.2840) loss 3.4237 (3.8338) grad_norm 1.3275 (1.1784) [2021-04-15 21:37:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][1110/1251] eta 0:00:40 lr 0.000848 time 0.2638 (0.2839) loss 4.5240 (3.8358) grad_norm 1.0662 (1.1782) [2021-04-15 21:37:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][1120/1251] eta 0:00:37 lr 0.000848 time 0.2639 (0.2839) loss 3.4002 (3.8352) grad_norm 1.0903 (1.1782) [2021-04-15 21:38:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][1130/1251] eta 0:00:34 lr 0.000848 time 0.2666 (0.2840) loss 3.6065 (3.8291) grad_norm 1.1549 (1.1784) [2021-04-15 21:38:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][1140/1251] eta 0:00:31 lr 0.000848 time 0.2621 (0.2840) loss 4.2384 (3.8304) grad_norm 1.0419 (1.1776) [2021-04-15 21:38:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][1150/1251] eta 0:00:28 lr 0.000848 time 0.2873 (0.2841) loss 3.9112 (3.8298) grad_norm 1.1774 (1.1777) [2021-04-15 21:38:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][1160/1251] eta 0:00:25 lr 0.000848 time 0.2919 (0.2841) loss 4.1171 (3.8308) grad_norm 1.1619 (1.1773) [2021-04-15 21:38:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][1170/1251] eta 0:00:23 lr 0.000848 time 0.2565 (0.2841) loss 2.9073 (3.8314) grad_norm 1.1378 (1.1777) [2021-04-15 21:38:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][1180/1251] eta 0:00:20 lr 0.000848 time 0.2895 (0.2841) loss 4.2858 (3.8313) grad_norm 1.1060 (1.1775) [2021-04-15 21:38:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][1190/1251] eta 0:00:17 lr 0.000848 time 0.3045 (0.2841) loss 3.0956 (3.8313) grad_norm 1.3593 (1.1777) [2021-04-15 21:38:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][1200/1251] eta 0:00:14 lr 0.000848 time 0.2760 (0.2841) loss 2.5044 (3.8296) grad_norm 1.0802 (1.1774) [2021-04-15 21:38:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][1210/1251] eta 0:00:11 lr 0.000848 time 0.3002 (0.2840) loss 3.3673 (3.8288) grad_norm 1.1827 (1.1775) [2021-04-15 21:38:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][1220/1251] eta 0:00:08 lr 0.000848 time 0.2828 (0.2840) loss 2.8358 (3.8285) grad_norm 0.9331 (1.1771) [2021-04-15 21:38:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][1230/1251] eta 0:00:05 lr 0.000848 time 0.2705 (0.2841) loss 3.8998 (3.8261) grad_norm 1.0534 (1.1773) [2021-04-15 21:38:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][1240/1251] eta 0:00:03 lr 0.000848 time 0.2493 (0.2840) loss 4.1048 (3.8254) grad_norm 1.1244 (1.1770) [2021-04-15 21:38:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [76/300][1250/1251] eta 0:00:00 lr 0.000848 time 0.2488 (0.2837) loss 3.7800 (3.8241) grad_norm 1.1314 (1.1769) [2021-04-15 21:38:39 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 76 training takes 0:05:57 [2021-04-15 21:38:39 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_76.pth saving...... [2021-04-15 21:39:04 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_76.pth saved !!! [2021-04-15 21:39:05 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.161 (1.161) Loss 1.2228 (1.2228) Acc@1 71.289 (71.289) Acc@5 91.016 (91.016) [2021-04-15 21:39:07 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.152 (0.266) Loss 1.1576 (1.1655) Acc@1 73.242 (72.869) Acc@5 91.113 (91.708) [2021-04-15 21:39:09 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.128 (0.235) Loss 1.2406 (1.1744) Acc@1 70.508 (72.800) Acc@5 90.625 (91.583) [2021-04-15 21:39:11 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.132 (0.241) Loss 1.1770 (1.1815) Acc@1 73.926 (72.555) Acc@5 91.016 (91.441) [2021-04-15 21:39:13 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.067 (0.224) Loss 1.1748 (1.1835) Acc@1 72.559 (72.401) Acc@5 91.504 (91.373) [2021-04-15 21:39:17 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 72.314 Acc@5 91.270 [2021-04-15 21:39:17 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 72.3% [2021-04-15 21:39:17 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 72.31% [2021-04-15 21:39:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][0/1251] eta 2:19:16 lr 0.000848 time 6.6802 (6.6802) loss 3.9382 (3.9382) grad_norm 1.5408 (1.5408) [2021-04-15 21:39:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][10/1251] eta 0:17:45 lr 0.000848 time 0.2969 (0.8585) loss 2.2064 (3.7128) grad_norm 1.1774 (1.2074) [2021-04-15 21:39:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][20/1251] eta 0:12:08 lr 0.000848 time 0.2536 (0.5917) loss 2.8813 (3.6269) grad_norm 1.2836 (1.2048) [2021-04-15 21:39:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][30/1251] eta 0:09:58 lr 0.000848 time 0.2803 (0.4905) loss 2.6539 (3.7572) grad_norm inf (inf) [2021-04-15 21:39:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][40/1251] eta 0:08:51 lr 0.000847 time 0.3002 (0.4390) loss 4.1388 (3.7540) grad_norm 1.2546 (inf) [2021-04-15 21:39:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][50/1251] eta 0:08:10 lr 0.000847 time 0.2937 (0.4087) loss 3.3529 (3.7828) grad_norm 1.0921 (inf) [2021-04-15 21:39:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][60/1251] eta 0:07:41 lr 0.000847 time 0.2543 (0.3876) loss 4.2162 (3.7944) grad_norm 1.3993 (inf) [2021-04-15 21:39:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][70/1251] eta 0:07:20 lr 0.000847 time 0.2886 (0.3730) loss 3.8033 (3.7906) grad_norm 1.1579 (inf) [2021-04-15 21:39:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][80/1251] eta 0:07:06 lr 0.000847 time 0.2666 (0.3639) loss 2.8997 (3.7610) grad_norm 1.2393 (inf) [2021-04-15 21:39:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][90/1251] eta 0:06:53 lr 0.000847 time 0.2608 (0.3564) loss 4.1697 (3.7688) grad_norm 1.1342 (inf) [2021-04-15 21:39:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][100/1251] eta 0:06:42 lr 0.000847 time 0.2969 (0.3494) loss 4.2257 (3.7433) grad_norm 1.1114 (inf) [2021-04-15 21:39:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][110/1251] eta 0:06:31 lr 0.000847 time 0.2816 (0.3431) loss 3.6391 (3.7465) grad_norm 1.2350 (inf) [2021-04-15 21:39:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][120/1251] eta 0:06:22 lr 0.000847 time 0.2892 (0.3382) loss 3.3436 (3.7568) grad_norm 1.1843 (inf) [2021-04-15 21:40:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][130/1251] eta 0:06:15 lr 0.000847 time 0.2909 (0.3349) loss 3.4033 (3.7380) grad_norm 1.1019 (inf) [2021-04-15 21:40:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][140/1251] eta 0:06:07 lr 0.000847 time 0.2939 (0.3312) loss 4.2068 (3.7330) grad_norm 1.0317 (inf) [2021-04-15 21:40:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][150/1251] eta 0:06:01 lr 0.000847 time 0.2972 (0.3283) loss 4.3865 (3.7321) grad_norm 1.4871 (inf) [2021-04-15 21:40:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][160/1251] eta 0:05:55 lr 0.000847 time 0.3094 (0.3257) loss 4.0976 (3.7166) grad_norm 1.4049 (inf) [2021-04-15 21:40:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][170/1251] eta 0:05:49 lr 0.000847 time 0.3073 (0.3236) loss 3.0348 (3.7055) grad_norm 1.2825 (inf) [2021-04-15 21:40:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][180/1251] eta 0:05:44 lr 0.000847 time 0.2802 (0.3212) loss 4.2971 (3.7261) grad_norm 1.2501 (inf) [2021-04-15 21:40:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][190/1251] eta 0:05:38 lr 0.000847 time 0.2861 (0.3193) loss 4.0501 (3.7181) grad_norm 0.9937 (inf) [2021-04-15 21:40:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][200/1251] eta 0:05:34 lr 0.000847 time 0.2849 (0.3179) loss 4.4119 (3.7304) grad_norm 1.1311 (inf) [2021-04-15 21:40:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][210/1251] eta 0:05:29 lr 0.000847 time 0.2762 (0.3166) loss 3.8425 (3.7381) grad_norm 1.0940 (inf) [2021-04-15 21:40:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][220/1251] eta 0:05:25 lr 0.000847 time 0.2961 (0.3156) loss 3.1402 (3.7313) grad_norm 1.1539 (inf) [2021-04-15 21:40:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][230/1251] eta 0:05:20 lr 0.000847 time 0.2756 (0.3144) loss 4.2405 (3.7321) grad_norm 1.2539 (inf) [2021-04-15 21:40:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][240/1251] eta 0:05:16 lr 0.000847 time 0.2787 (0.3133) loss 3.3407 (3.7407) grad_norm 1.1822 (inf) [2021-04-15 21:40:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][250/1251] eta 0:05:12 lr 0.000847 time 0.2974 (0.3121) loss 3.9524 (3.7401) grad_norm 1.1085 (inf) [2021-04-15 21:40:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][260/1251] eta 0:05:08 lr 0.000847 time 0.2706 (0.3112) loss 3.8805 (3.7386) grad_norm 1.1733 (inf) [2021-04-15 21:40:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][270/1251] eta 0:05:04 lr 0.000847 time 0.2879 (0.3103) loss 2.8802 (3.7333) grad_norm 1.1603 (inf) [2021-04-15 21:40:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][280/1251] eta 0:05:00 lr 0.000847 time 0.3034 (0.3098) loss 3.3567 (3.7332) grad_norm 1.0765 (inf) [2021-04-15 21:40:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][290/1251] eta 0:04:56 lr 0.000847 time 0.2841 (0.3090) loss 3.8943 (3.7348) grad_norm 1.0909 (inf) [2021-04-15 21:40:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][300/1251] eta 0:04:53 lr 0.000847 time 0.2888 (0.3090) loss 4.4516 (3.7387) grad_norm 1.1132 (inf) [2021-04-15 21:40:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][310/1251] eta 0:04:50 lr 0.000847 time 0.2581 (0.3087) loss 2.6603 (3.7440) grad_norm 1.0232 (inf) [2021-04-15 21:40:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][320/1251] eta 0:04:46 lr 0.000847 time 0.2765 (0.3079) loss 3.6389 (3.7446) grad_norm 1.0531 (inf) [2021-04-15 21:40:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][330/1251] eta 0:04:42 lr 0.000847 time 0.2703 (0.3072) loss 3.1766 (3.7449) grad_norm 1.3952 (inf) [2021-04-15 21:41:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][340/1251] eta 0:04:39 lr 0.000847 time 0.3014 (0.3070) loss 3.8562 (3.7435) grad_norm 1.1795 (inf) [2021-04-15 21:41:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][350/1251] eta 0:04:36 lr 0.000847 time 0.2654 (0.3063) loss 4.2962 (3.7426) grad_norm 1.0961 (inf) [2021-04-15 21:41:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][360/1251] eta 0:04:32 lr 0.000847 time 0.3087 (0.3058) loss 3.9571 (3.7463) grad_norm 1.4191 (inf) [2021-04-15 21:41:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][370/1251] eta 0:04:29 lr 0.000847 time 0.3188 (0.3056) loss 3.9764 (3.7547) grad_norm 1.1780 (inf) [2021-04-15 21:41:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][380/1251] eta 0:04:25 lr 0.000846 time 0.2812 (0.3051) loss 3.7742 (3.7607) grad_norm 1.1882 (inf) [2021-04-15 21:41:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][390/1251] eta 0:04:22 lr 0.000846 time 0.2883 (0.3047) loss 3.5947 (3.7625) grad_norm 0.9697 (inf) [2021-04-15 21:41:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][400/1251] eta 0:04:18 lr 0.000846 time 0.2851 (0.3043) loss 3.9883 (3.7642) grad_norm 1.2157 (inf) [2021-04-15 21:41:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][410/1251] eta 0:04:15 lr 0.000846 time 0.2921 (0.3039) loss 3.7888 (3.7616) grad_norm 1.3085 (inf) [2021-04-15 21:41:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][420/1251] eta 0:04:12 lr 0.000846 time 0.2981 (0.3035) loss 4.1855 (3.7648) grad_norm 1.3152 (inf) [2021-04-15 21:41:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][430/1251] eta 0:04:08 lr 0.000846 time 0.2930 (0.3032) loss 4.0844 (3.7650) grad_norm 1.4102 (inf) [2021-04-15 21:41:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][440/1251] eta 0:04:05 lr 0.000846 time 0.2717 (0.3028) loss 3.3097 (3.7591) grad_norm 1.1398 (inf) [2021-04-15 21:41:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][450/1251] eta 0:04:02 lr 0.000846 time 0.3036 (0.3024) loss 3.0592 (3.7540) grad_norm 1.0050 (inf) [2021-04-15 21:41:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][460/1251] eta 0:03:59 lr 0.000846 time 0.2567 (0.3022) loss 3.5888 (3.7558) grad_norm 1.2545 (inf) [2021-04-15 21:41:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][470/1251] eta 0:03:55 lr 0.000846 time 0.2707 (0.3021) loss 4.7148 (3.7598) grad_norm 1.1216 (inf) [2021-04-15 21:41:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][480/1251] eta 0:03:52 lr 0.000846 time 0.3135 (0.3019) loss 3.9039 (3.7602) grad_norm 1.0891 (inf) [2021-04-15 21:41:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][490/1251] eta 0:03:49 lr 0.000846 time 0.2948 (0.3017) loss 4.0366 (3.7614) grad_norm 1.1744 (inf) [2021-04-15 21:41:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][500/1251] eta 0:03:46 lr 0.000846 time 0.2947 (0.3015) loss 3.7222 (3.7651) grad_norm 0.9887 (inf) [2021-04-15 21:41:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][510/1251] eta 0:03:43 lr 0.000846 time 0.3094 (0.3012) loss 3.8922 (3.7629) grad_norm 1.1900 (inf) [2021-04-15 21:41:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][520/1251] eta 0:03:40 lr 0.000846 time 0.3099 (0.3010) loss 4.2904 (3.7626) grad_norm 1.0721 (inf) [2021-04-15 21:41:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][530/1251] eta 0:03:36 lr 0.000846 time 0.2621 (0.3006) loss 4.1239 (3.7633) grad_norm 1.1927 (inf) [2021-04-15 21:41:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][540/1251] eta 0:03:33 lr 0.000846 time 0.2661 (0.3008) loss 3.2091 (3.7662) grad_norm 1.1648 (inf) [2021-04-15 21:42:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][550/1251] eta 0:03:30 lr 0.000846 time 0.2615 (0.3004) loss 4.0768 (3.7671) grad_norm 1.1742 (inf) [2021-04-15 21:42:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][560/1251] eta 0:03:27 lr 0.000846 time 0.2650 (0.3001) loss 4.6672 (3.7671) grad_norm 1.3201 (inf) [2021-04-15 21:42:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][570/1251] eta 0:03:24 lr 0.000846 time 0.3056 (0.3004) loss 4.3970 (3.7732) grad_norm 1.0734 (inf) [2021-04-15 21:42:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][580/1251] eta 0:03:21 lr 0.000846 time 0.2816 (0.3002) loss 3.1237 (3.7742) grad_norm 1.1894 (inf) [2021-04-15 21:42:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][590/1251] eta 0:03:18 lr 0.000846 time 0.2791 (0.3002) loss 4.1348 (3.7758) grad_norm 1.2172 (inf) [2021-04-15 21:42:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][600/1251] eta 0:03:15 lr 0.000846 time 0.2699 (0.3000) loss 3.2104 (3.7794) grad_norm 1.4773 (inf) [2021-04-15 21:42:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][610/1251] eta 0:03:12 lr 0.000846 time 0.2863 (0.2998) loss 3.8931 (3.7806) grad_norm 1.1840 (inf) [2021-04-15 21:42:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][620/1251] eta 0:03:09 lr 0.000846 time 0.3120 (0.2997) loss 2.5289 (3.7819) grad_norm 1.1275 (inf) [2021-04-15 21:42:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][630/1251] eta 0:03:05 lr 0.000846 time 0.2669 (0.2994) loss 4.1889 (3.7845) grad_norm 1.0646 (inf) [2021-04-15 21:42:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][640/1251] eta 0:03:02 lr 0.000846 time 0.3277 (0.2993) loss 3.0654 (3.7792) grad_norm 1.2146 (inf) [2021-04-15 21:42:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][650/1251] eta 0:02:59 lr 0.000846 time 0.4647 (0.2994) loss 4.2221 (3.7788) grad_norm 1.3073 (inf) [2021-04-15 21:42:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][660/1251] eta 0:02:56 lr 0.000846 time 0.2996 (0.2991) loss 3.6330 (3.7830) grad_norm 1.1323 (inf) [2021-04-15 21:42:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][670/1251] eta 0:02:53 lr 0.000846 time 0.2846 (0.2991) loss 2.7589 (3.7818) grad_norm 1.3178 (inf) [2021-04-15 21:42:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][680/1251] eta 0:02:50 lr 0.000846 time 0.2792 (0.2989) loss 4.1670 (3.7832) grad_norm 1.1594 (inf) [2021-04-15 21:42:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][690/1251] eta 0:02:47 lr 0.000846 time 0.2785 (0.2987) loss 4.0786 (3.7803) grad_norm 1.0624 (inf) [2021-04-15 21:42:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][700/1251] eta 0:02:44 lr 0.000846 time 0.4071 (0.2987) loss 3.9362 (3.7787) grad_norm 1.3034 (inf) [2021-04-15 21:42:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][710/1251] eta 0:02:41 lr 0.000845 time 0.2582 (0.2986) loss 3.6597 (3.7779) grad_norm 1.1097 (inf) [2021-04-15 21:42:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][720/1251] eta 0:02:38 lr 0.000845 time 0.2951 (0.2985) loss 3.9347 (3.7772) grad_norm 1.2107 (inf) [2021-04-15 21:42:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][730/1251] eta 0:02:35 lr 0.000845 time 0.2904 (0.2984) loss 3.5578 (3.7764) grad_norm 1.3688 (inf) [2021-04-15 21:42:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][740/1251] eta 0:02:32 lr 0.000845 time 0.3141 (0.2983) loss 3.4426 (3.7792) grad_norm 1.3656 (inf) [2021-04-15 21:43:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][750/1251] eta 0:02:29 lr 0.000845 time 0.2772 (0.2981) loss 4.7311 (3.7790) grad_norm 1.4599 (inf) [2021-04-15 21:43:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][760/1251] eta 0:02:26 lr 0.000845 time 0.2682 (0.2980) loss 4.0207 (3.7822) grad_norm 1.3383 (inf) [2021-04-15 21:43:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][770/1251] eta 0:02:23 lr 0.000845 time 0.2880 (0.2978) loss 3.8840 (3.7835) grad_norm 1.1044 (inf) [2021-04-15 21:43:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][780/1251] eta 0:02:20 lr 0.000845 time 0.3079 (0.2977) loss 4.4222 (3.7811) grad_norm 1.0131 (inf) [2021-04-15 21:43:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][790/1251] eta 0:02:17 lr 0.000845 time 0.2754 (0.2976) loss 3.0185 (3.7782) grad_norm 1.0469 (inf) [2021-04-15 21:43:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][800/1251] eta 0:02:14 lr 0.000845 time 0.3088 (0.2975) loss 2.9110 (3.7780) grad_norm 1.1633 (inf) [2021-04-15 21:43:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][810/1251] eta 0:02:11 lr 0.000845 time 0.2966 (0.2973) loss 3.8066 (3.7796) grad_norm 1.0992 (inf) [2021-04-15 21:43:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][820/1251] eta 0:02:08 lr 0.000845 time 0.2845 (0.2972) loss 3.7376 (3.7805) grad_norm 1.3493 (inf) [2021-04-15 21:43:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][830/1251] eta 0:02:05 lr 0.000845 time 0.3105 (0.2971) loss 3.0296 (3.7778) grad_norm 0.9760 (inf) [2021-04-15 21:43:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][840/1251] eta 0:02:02 lr 0.000845 time 0.3068 (0.2971) loss 3.5941 (3.7761) grad_norm 1.1204 (inf) [2021-04-15 21:43:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][850/1251] eta 0:01:59 lr 0.000845 time 0.2612 (0.2970) loss 4.1359 (3.7769) grad_norm 0.9121 (inf) [2021-04-15 21:43:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][860/1251] eta 0:01:56 lr 0.000845 time 0.2821 (0.2969) loss 4.0789 (3.7768) grad_norm 1.1214 (inf) [2021-04-15 21:43:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][870/1251] eta 0:01:53 lr 0.000845 time 0.2933 (0.2968) loss 2.9389 (3.7707) grad_norm 1.2703 (inf) [2021-04-15 21:43:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][880/1251] eta 0:01:50 lr 0.000845 time 0.4575 (0.2971) loss 4.1197 (3.7705) grad_norm 1.1380 (inf) [2021-04-15 21:43:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][890/1251] eta 0:01:47 lr 0.000845 time 0.2868 (0.2969) loss 4.6577 (3.7718) grad_norm 1.1379 (inf) [2021-04-15 21:43:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][900/1251] eta 0:01:44 lr 0.000845 time 0.2748 (0.2968) loss 3.4396 (3.7703) grad_norm 1.2312 (inf) [2021-04-15 21:43:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][910/1251] eta 0:01:41 lr 0.000845 time 0.2943 (0.2966) loss 3.9401 (3.7703) grad_norm 0.9717 (inf) [2021-04-15 21:43:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][920/1251] eta 0:01:38 lr 0.000845 time 0.2914 (0.2967) loss 4.3157 (3.7709) grad_norm 1.4471 (inf) [2021-04-15 21:43:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][930/1251] eta 0:01:35 lr 0.000845 time 0.2589 (0.2965) loss 3.4901 (3.7705) grad_norm 1.0548 (inf) [2021-04-15 21:43:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][940/1251] eta 0:01:32 lr 0.000845 time 0.4355 (0.2966) loss 3.9519 (3.7692) grad_norm 1.1732 (inf) [2021-04-15 21:43:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][950/1251] eta 0:01:29 lr 0.000845 time 0.3100 (0.2965) loss 3.8324 (3.7707) grad_norm 1.1674 (inf) [2021-04-15 21:44:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][960/1251] eta 0:01:26 lr 0.000845 time 0.2971 (0.2964) loss 3.0974 (3.7695) grad_norm 1.3256 (inf) [2021-04-15 21:44:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][970/1251] eta 0:01:23 lr 0.000845 time 0.3142 (0.2963) loss 3.7589 (3.7725) grad_norm 1.1531 (inf) [2021-04-15 21:44:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][980/1251] eta 0:01:20 lr 0.000845 time 0.2841 (0.2963) loss 3.3861 (3.7708) grad_norm 1.2386 (inf) [2021-04-15 21:44:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][990/1251] eta 0:01:17 lr 0.000845 time 0.3311 (0.2962) loss 4.4883 (3.7724) grad_norm 1.2324 (inf) [2021-04-15 21:44:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][1000/1251] eta 0:01:14 lr 0.000845 time 0.2731 (0.2961) loss 3.8800 (3.7715) grad_norm 1.1023 (inf) [2021-04-15 21:44:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][1010/1251] eta 0:01:11 lr 0.000845 time 0.3086 (0.2961) loss 4.5066 (3.7716) grad_norm 1.0828 (inf) [2021-04-15 21:44:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][1020/1251] eta 0:01:08 lr 0.000845 time 0.2926 (0.2960) loss 3.9327 (3.7741) grad_norm 1.1457 (inf) [2021-04-15 21:44:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][1030/1251] eta 0:01:05 lr 0.000845 time 0.2910 (0.2959) loss 4.8475 (3.7783) grad_norm 1.0552 (inf) [2021-04-15 21:44:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][1040/1251] eta 0:01:02 lr 0.000844 time 0.2885 (0.2958) loss 4.6142 (3.7783) grad_norm 1.0197 (inf) [2021-04-15 21:44:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][1050/1251] eta 0:00:59 lr 0.000844 time 0.2948 (0.2958) loss 4.3238 (3.7773) grad_norm 1.1727 (inf) [2021-04-15 21:44:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][1060/1251] eta 0:00:56 lr 0.000844 time 0.2701 (0.2957) loss 2.8248 (3.7780) grad_norm 1.1720 (inf) [2021-04-15 21:44:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][1070/1251] eta 0:00:53 lr 0.000844 time 0.2790 (0.2957) loss 4.1083 (3.7781) grad_norm 1.0954 (inf) [2021-04-15 21:44:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][1080/1251] eta 0:00:50 lr 0.000844 time 0.2991 (0.2956) loss 3.5787 (3.7776) grad_norm 1.0322 (inf) [2021-04-15 21:44:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][1090/1251] eta 0:00:47 lr 0.000844 time 0.2912 (0.2956) loss 4.2173 (3.7792) grad_norm 1.1195 (inf) [2021-04-15 21:44:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][1100/1251] eta 0:00:44 lr 0.000844 time 0.2859 (0.2955) loss 4.4331 (3.7810) grad_norm 1.1271 (inf) [2021-04-15 21:44:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][1110/1251] eta 0:00:41 lr 0.000844 time 0.2752 (0.2954) loss 3.4416 (3.7825) grad_norm 1.1920 (inf) [2021-04-15 21:44:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][1120/1251] eta 0:00:38 lr 0.000844 time 0.3044 (0.2955) loss 3.8331 (3.7821) grad_norm 1.1225 (inf) [2021-04-15 21:44:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][1130/1251] eta 0:00:35 lr 0.000844 time 0.2691 (0.2954) loss 3.8850 (3.7823) grad_norm 1.1427 (inf) [2021-04-15 21:44:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][1140/1251] eta 0:00:32 lr 0.000844 time 0.2651 (0.2955) loss 4.2442 (3.7839) grad_norm 1.1225 (inf) [2021-04-15 21:44:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][1150/1251] eta 0:00:29 lr 0.000844 time 0.2888 (0.2954) loss 4.3685 (3.7817) grad_norm 1.0172 (inf) [2021-04-15 21:45:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][1160/1251] eta 0:00:26 lr 0.000844 time 0.2896 (0.2954) loss 3.2787 (3.7822) grad_norm 1.5388 (inf) [2021-04-15 21:45:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][1170/1251] eta 0:00:23 lr 0.000844 time 0.2748 (0.2955) loss 4.1120 (3.7834) grad_norm 1.0549 (inf) [2021-04-15 21:45:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][1180/1251] eta 0:00:20 lr 0.000844 time 0.2734 (0.2954) loss 2.6337 (3.7829) grad_norm 1.0519 (inf) [2021-04-15 21:45:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][1190/1251] eta 0:00:18 lr 0.000844 time 0.2780 (0.2953) loss 3.4840 (3.7847) grad_norm 1.1301 (inf) [2021-04-15 21:45:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][1200/1251] eta 0:00:15 lr 0.000844 time 0.2758 (0.2952) loss 4.5127 (3.7851) grad_norm 1.1462 (inf) [2021-04-15 21:45:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][1210/1251] eta 0:00:12 lr 0.000844 time 0.2824 (0.2952) loss 4.9161 (3.7854) grad_norm 1.1314 (inf) [2021-04-15 21:45:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][1220/1251] eta 0:00:09 lr 0.000844 time 0.2774 (0.2951) loss 4.1129 (3.7849) grad_norm 1.1930 (inf) [2021-04-15 21:45:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][1230/1251] eta 0:00:06 lr 0.000844 time 0.2769 (0.2951) loss 3.3731 (3.7833) grad_norm 1.2752 (inf) [2021-04-15 21:45:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][1240/1251] eta 0:00:03 lr 0.000844 time 0.2500 (0.2950) loss 4.5109 (3.7808) grad_norm 1.2080 (inf) [2021-04-15 21:45:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [77/300][1250/1251] eta 0:00:00 lr 0.000844 time 0.2493 (0.2946) loss 4.0832 (3.7823) grad_norm 1.1633 (inf) [2021-04-15 21:45:27 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 77 training takes 0:06:10 [2021-04-15 21:45:27 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_77.pth saving...... [2021-04-15 21:45:43 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_77.pth saved !!! [2021-04-15 21:45:44 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.388 (1.388) Loss 1.2637 (1.2637) Acc@1 70.898 (70.898) Acc@5 90.137 (90.137) [2021-04-15 21:45:46 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.138 (0.243) Loss 1.2084 (1.2076) Acc@1 71.973 (72.044) Acc@5 91.602 (91.566) [2021-04-15 21:45:48 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.133 (0.259) Loss 1.2145 (1.2059) Acc@1 73.340 (72.456) Acc@5 91.309 (91.416) [2021-04-15 21:45:51 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.522 (0.267) Loss 1.1303 (1.2032) Acc@1 73.438 (72.647) Acc@5 92.480 (91.419) [2021-04-15 21:45:53 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.244) Loss 1.1855 (1.2020) Acc@1 73.438 (72.563) Acc@5 91.113 (91.399) [2021-04-15 21:45:56 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 72.534 Acc@5 91.372 [2021-04-15 21:45:56 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 72.5% [2021-04-15 21:45:56 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 72.53% [2021-04-15 21:46:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][0/1251] eta 1:46:16 lr 0.000844 time 5.0972 (5.0972) loss 4.2379 (4.2379) grad_norm 1.0271 (1.0271) [2021-04-15 21:46:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][10/1251] eta 0:15:08 lr 0.000844 time 0.2816 (0.7318) loss 2.6363 (3.5549) grad_norm 1.2181 (1.1234) [2021-04-15 21:46:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][20/1251] eta 0:10:41 lr 0.000844 time 0.2933 (0.5212) loss 3.1304 (3.6483) grad_norm 1.1993 (1.1378) [2021-04-15 21:46:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][30/1251] eta 0:09:05 lr 0.000844 time 0.3162 (0.4468) loss 2.7980 (3.6923) grad_norm 1.3370 (1.1497) [2021-04-15 21:46:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][40/1251] eta 0:08:14 lr 0.000844 time 0.2813 (0.4087) loss 3.1239 (3.7181) grad_norm 1.1375 (1.1562) [2021-04-15 21:46:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][50/1251] eta 0:07:44 lr 0.000844 time 0.3013 (0.3869) loss 3.7785 (3.7363) grad_norm 1.1437 (1.1731) [2021-04-15 21:46:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][60/1251] eta 0:07:21 lr 0.000844 time 0.2659 (0.3710) loss 3.4107 (3.7567) grad_norm 1.1478 (1.1663) [2021-04-15 21:46:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][70/1251] eta 0:07:07 lr 0.000844 time 0.3129 (0.3620) loss 3.2311 (3.7326) grad_norm 1.1557 (1.1841) [2021-04-15 21:46:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][80/1251] eta 0:06:55 lr 0.000844 time 0.2841 (0.3545) loss 2.6762 (3.6939) grad_norm 1.0285 (1.1892) [2021-04-15 21:46:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][90/1251] eta 0:06:45 lr 0.000844 time 0.2927 (0.3491) loss 3.7055 (3.7285) grad_norm 1.2153 (1.1861) [2021-04-15 21:46:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][100/1251] eta 0:06:34 lr 0.000844 time 0.2841 (0.3432) loss 4.1654 (3.7396) grad_norm 1.2803 (1.1852) [2021-04-15 21:46:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][110/1251] eta 0:06:26 lr 0.000844 time 0.2979 (0.3386) loss 4.4834 (3.7581) grad_norm 1.1937 (1.1841) [2021-04-15 21:46:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][120/1251] eta 0:06:19 lr 0.000843 time 0.2608 (0.3354) loss 2.7268 (3.7464) grad_norm 1.0704 (1.1833) [2021-04-15 21:46:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][130/1251] eta 0:06:12 lr 0.000843 time 0.3731 (0.3323) loss 3.0397 (3.7485) grad_norm 1.2793 (1.1828) [2021-04-15 21:46:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][140/1251] eta 0:06:05 lr 0.000843 time 0.2993 (0.3289) loss 3.9082 (3.7751) grad_norm 1.3336 (1.1831) [2021-04-15 21:46:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][150/1251] eta 0:06:00 lr 0.000843 time 0.4607 (0.3276) loss 3.7845 (3.7589) grad_norm 1.1793 (1.1826) [2021-04-15 21:46:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][160/1251] eta 0:05:54 lr 0.000843 time 0.2929 (0.3248) loss 3.6994 (3.7312) grad_norm 1.2865 (1.1834) [2021-04-15 21:46:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][170/1251] eta 0:05:49 lr 0.000843 time 0.2815 (0.3233) loss 4.4644 (3.7373) grad_norm 1.3107 (1.1829) [2021-04-15 21:46:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][180/1251] eta 0:05:44 lr 0.000843 time 0.2864 (0.3214) loss 3.5889 (3.7442) grad_norm 1.2229 (1.1859) [2021-04-15 21:46:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][190/1251] eta 0:05:39 lr 0.000843 time 0.2931 (0.3197) loss 4.1466 (3.7406) grad_norm 1.3660 (1.1886) [2021-04-15 21:47:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][200/1251] eta 0:05:34 lr 0.000843 time 0.2819 (0.3184) loss 3.5194 (3.7441) grad_norm 1.1003 (1.1869) [2021-04-15 21:47:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][210/1251] eta 0:05:29 lr 0.000843 time 0.2760 (0.3170) loss 2.7740 (3.7459) grad_norm 1.3491 (1.1851) [2021-04-15 21:47:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][220/1251] eta 0:05:25 lr 0.000843 time 0.3009 (0.3157) loss 4.5459 (3.7480) grad_norm 1.3146 (1.1837) [2021-04-15 21:47:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][230/1251] eta 0:05:21 lr 0.000843 time 0.2874 (0.3145) loss 3.7543 (3.7568) grad_norm 1.1775 (1.1828) [2021-04-15 21:47:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][240/1251] eta 0:05:16 lr 0.000843 time 0.2533 (0.3133) loss 4.3100 (3.7508) grad_norm 1.1181 (1.1829) [2021-04-15 21:47:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][250/1251] eta 0:05:12 lr 0.000843 time 0.2911 (0.3124) loss 2.5692 (3.7502) grad_norm 1.1763 (1.1816) [2021-04-15 21:47:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][260/1251] eta 0:05:08 lr 0.000843 time 0.3138 (0.3117) loss 3.4322 (3.7549) grad_norm 1.2974 (1.1857) [2021-04-15 21:47:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][270/1251] eta 0:05:05 lr 0.000843 time 0.2780 (0.3110) loss 2.7830 (3.7529) grad_norm 1.1296 (1.1898) [2021-04-15 21:47:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][280/1251] eta 0:05:01 lr 0.000843 time 0.2843 (0.3102) loss 4.2705 (3.7475) grad_norm 0.9491 (1.1882) [2021-04-15 21:47:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][290/1251] eta 0:04:57 lr 0.000843 time 0.2720 (0.3094) loss 3.3785 (3.7406) grad_norm 1.0787 (1.1862) [2021-04-15 21:47:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][300/1251] eta 0:04:53 lr 0.000843 time 0.3228 (0.3088) loss 2.9682 (3.7442) grad_norm 1.3102 (1.1872) [2021-04-15 21:47:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][310/1251] eta 0:04:50 lr 0.000843 time 0.3193 (0.3083) loss 4.5227 (3.7464) grad_norm 1.1513 (1.1876) [2021-04-15 21:47:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][320/1251] eta 0:04:46 lr 0.000843 time 0.3199 (0.3082) loss 4.4746 (3.7501) grad_norm 1.0654 (1.1856) [2021-04-15 21:47:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][330/1251] eta 0:04:43 lr 0.000843 time 0.3018 (0.3074) loss 3.4033 (3.7489) grad_norm 1.0582 (1.1860) [2021-04-15 21:47:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][340/1251] eta 0:04:40 lr 0.000843 time 0.2885 (0.3077) loss 3.4345 (3.7517) grad_norm 1.0299 (1.1850) [2021-04-15 21:47:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][350/1251] eta 0:04:36 lr 0.000843 time 0.3389 (0.3071) loss 3.9305 (3.7568) grad_norm 1.0514 (1.1828) [2021-04-15 21:47:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][360/1251] eta 0:04:33 lr 0.000843 time 0.2935 (0.3066) loss 4.1972 (3.7608) grad_norm 1.0725 (1.1804) [2021-04-15 21:47:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][370/1251] eta 0:04:30 lr 0.000843 time 0.2714 (0.3066) loss 3.8147 (3.7612) grad_norm 1.2759 (1.1788) [2021-04-15 21:47:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][380/1251] eta 0:04:26 lr 0.000843 time 0.3121 (0.3061) loss 4.4029 (3.7669) grad_norm 0.9682 (1.1791) [2021-04-15 21:47:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][390/1251] eta 0:04:23 lr 0.000843 time 0.3053 (0.3057) loss 3.7281 (3.7693) grad_norm 1.0828 (1.1773) [2021-04-15 21:47:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][400/1251] eta 0:04:19 lr 0.000843 time 0.2828 (0.3052) loss 3.0668 (3.7703) grad_norm 1.1253 (1.1772) [2021-04-15 21:48:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][410/1251] eta 0:04:16 lr 0.000843 time 0.2812 (0.3047) loss 3.7769 (3.7757) grad_norm 1.1147 (1.1758) [2021-04-15 21:48:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][420/1251] eta 0:04:12 lr 0.000843 time 0.2863 (0.3044) loss 3.4738 (3.7828) grad_norm 1.4245 (1.1765) [2021-04-15 21:48:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][430/1251] eta 0:04:09 lr 0.000843 time 0.2929 (0.3044) loss 4.3046 (3.7882) grad_norm 1.0772 (1.1781) [2021-04-15 21:48:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][440/1251] eta 0:04:06 lr 0.000843 time 0.3057 (0.3043) loss 4.2387 (3.7982) grad_norm 1.1930 (1.1780) [2021-04-15 21:48:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][450/1251] eta 0:04:03 lr 0.000842 time 0.2680 (0.3038) loss 3.6155 (3.7924) grad_norm 1.0472 (1.1785) [2021-04-15 21:48:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][460/1251] eta 0:04:00 lr 0.000842 time 0.2803 (0.3035) loss 2.5966 (3.7912) grad_norm 1.0859 (1.1783) [2021-04-15 21:48:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][470/1251] eta 0:03:56 lr 0.000842 time 0.2752 (0.3033) loss 3.4031 (3.7910) grad_norm 1.0888 (1.1765) [2021-04-15 21:48:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][480/1251] eta 0:03:53 lr 0.000842 time 0.2929 (0.3030) loss 2.8344 (3.7881) grad_norm 1.1450 (1.1754) [2021-04-15 21:48:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][490/1251] eta 0:03:50 lr 0.000842 time 0.2777 (0.3026) loss 4.4504 (3.7894) grad_norm 1.1983 (1.1759) [2021-04-15 21:48:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][500/1251] eta 0:03:47 lr 0.000842 time 0.2966 (0.3025) loss 3.8831 (3.7946) grad_norm 1.1022 (1.1755) [2021-04-15 21:48:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][510/1251] eta 0:03:43 lr 0.000842 time 0.2854 (0.3022) loss 4.1818 (3.7949) grad_norm 1.0995 (1.1748) [2021-04-15 21:48:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][520/1251] eta 0:03:40 lr 0.000842 time 0.3107 (0.3020) loss 4.4597 (3.7953) grad_norm 1.1018 (1.1742) [2021-04-15 21:48:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][530/1251] eta 0:03:37 lr 0.000842 time 0.2851 (0.3018) loss 3.7116 (3.7996) grad_norm 1.4390 (1.1763) [2021-04-15 21:48:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][540/1251] eta 0:03:34 lr 0.000842 time 0.3088 (0.3015) loss 4.2677 (3.8009) grad_norm 1.1530 (1.1756) [2021-04-15 21:48:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][550/1251] eta 0:03:31 lr 0.000842 time 0.2721 (0.3015) loss 4.2119 (3.8045) grad_norm 1.1124 (1.1750) [2021-04-15 21:48:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][560/1251] eta 0:03:28 lr 0.000842 time 0.2747 (0.3011) loss 3.8782 (3.8061) grad_norm 1.2393 (1.1741) [2021-04-15 21:48:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][570/1251] eta 0:03:25 lr 0.000842 time 0.3666 (0.3011) loss 3.6698 (3.8029) grad_norm 1.2688 (1.1745) [2021-04-15 21:48:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][580/1251] eta 0:03:21 lr 0.000842 time 0.3269 (0.3009) loss 3.4995 (3.8020) grad_norm 1.1175 (1.1749) [2021-04-15 21:48:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][590/1251] eta 0:03:18 lr 0.000842 time 0.2866 (0.3006) loss 2.9519 (3.7955) grad_norm 1.6489 (1.1767) [2021-04-15 21:48:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][600/1251] eta 0:03:15 lr 0.000842 time 0.2901 (0.3003) loss 3.2140 (3.7920) grad_norm 1.5236 (1.1773) [2021-04-15 21:48:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][610/1251] eta 0:03:12 lr 0.000842 time 0.3081 (0.3002) loss 3.7357 (3.7963) grad_norm 1.3292 (1.1774) [2021-04-15 21:49:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][620/1251] eta 0:03:09 lr 0.000842 time 0.2910 (0.3000) loss 4.1680 (3.8036) grad_norm 1.0383 (1.1775) [2021-04-15 21:49:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][630/1251] eta 0:03:06 lr 0.000842 time 0.2958 (0.2999) loss 3.6902 (3.8035) grad_norm 1.3167 (1.1779) [2021-04-15 21:49:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][640/1251] eta 0:03:03 lr 0.000842 time 0.2447 (0.3000) loss 3.9721 (3.8061) grad_norm 1.2924 (1.1783) [2021-04-15 21:49:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][650/1251] eta 0:03:00 lr 0.000842 time 0.4826 (0.3000) loss 4.2674 (3.8089) grad_norm 1.1950 (1.1795) [2021-04-15 21:49:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][660/1251] eta 0:02:57 lr 0.000842 time 0.2965 (0.2998) loss 2.7552 (3.8049) grad_norm 1.1329 (1.1802) [2021-04-15 21:49:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][670/1251] eta 0:02:54 lr 0.000842 time 0.3023 (0.2998) loss 3.2142 (3.8081) grad_norm 1.0485 (1.1800) [2021-04-15 21:49:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][680/1251] eta 0:02:51 lr 0.000842 time 0.2919 (0.2997) loss 3.9008 (3.8076) grad_norm 1.3775 (1.1795) [2021-04-15 21:49:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][690/1251] eta 0:02:48 lr 0.000842 time 0.4399 (0.2997) loss 4.0984 (3.8049) grad_norm 1.2918 (1.1801) [2021-04-15 21:49:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][700/1251] eta 0:02:44 lr 0.000842 time 0.2654 (0.2994) loss 4.3048 (3.8086) grad_norm 1.0275 (1.1809) [2021-04-15 21:49:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][710/1251] eta 0:02:41 lr 0.000842 time 0.2829 (0.2993) loss 3.8652 (3.8078) grad_norm 1.2397 (1.1818) [2021-04-15 21:49:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][720/1251] eta 0:02:38 lr 0.000842 time 0.2859 (0.2991) loss 3.9190 (3.8046) grad_norm 1.3381 (1.1817) [2021-04-15 21:49:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][730/1251] eta 0:02:35 lr 0.000842 time 0.2822 (0.2992) loss 3.6566 (3.8024) grad_norm 1.1149 (1.1810) [2021-04-15 21:49:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][740/1251] eta 0:02:32 lr 0.000842 time 0.2971 (0.2990) loss 3.3430 (3.8006) grad_norm 1.0705 (1.1800) [2021-04-15 21:49:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][750/1251] eta 0:02:29 lr 0.000842 time 0.3207 (0.2988) loss 3.9918 (3.8037) grad_norm 1.3506 (1.1805) [2021-04-15 21:49:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][760/1251] eta 0:02:26 lr 0.000842 time 0.2890 (0.2986) loss 3.8262 (3.8031) grad_norm 1.1448 (1.1804) [2021-04-15 21:49:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][770/1251] eta 0:02:23 lr 0.000842 time 0.2767 (0.2984) loss 4.2956 (3.8023) grad_norm 1.3129 (1.1816) [2021-04-15 21:49:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][780/1251] eta 0:02:20 lr 0.000841 time 0.2773 (0.2982) loss 3.5908 (3.7990) grad_norm 1.0699 (1.1804) [2021-04-15 21:49:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][790/1251] eta 0:02:17 lr 0.000841 time 0.2948 (0.2980) loss 4.2516 (3.7988) grad_norm 1.1232 (inf) [2021-04-15 21:49:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][800/1251] eta 0:02:14 lr 0.000841 time 0.2825 (0.2979) loss 4.7647 (3.7961) grad_norm 1.0094 (inf) [2021-04-15 21:49:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][810/1251] eta 0:02:11 lr 0.000841 time 0.2678 (0.2977) loss 3.9613 (3.7985) grad_norm 1.3271 (inf) [2021-04-15 21:50:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][820/1251] eta 0:02:08 lr 0.000841 time 0.2949 (0.2976) loss 4.7141 (3.7974) grad_norm 1.2280 (inf) [2021-04-15 21:50:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][830/1251] eta 0:02:05 lr 0.000841 time 0.2973 (0.2975) loss 3.9535 (3.7935) grad_norm 1.1152 (inf) [2021-04-15 21:50:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][840/1251] eta 0:02:02 lr 0.000841 time 0.2840 (0.2973) loss 4.2889 (3.7938) grad_norm 1.1312 (inf) [2021-04-15 21:50:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][850/1251] eta 0:01:59 lr 0.000841 time 0.2786 (0.2971) loss 3.5547 (3.7904) grad_norm 1.2534 (inf) [2021-04-15 21:50:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][860/1251] eta 0:01:56 lr 0.000841 time 0.2848 (0.2969) loss 3.4016 (3.7887) grad_norm 1.0504 (inf) [2021-04-15 21:50:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][870/1251] eta 0:01:53 lr 0.000841 time 0.2751 (0.2968) loss 4.5957 (3.7905) grad_norm 1.3903 (inf) [2021-04-15 21:50:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][880/1251] eta 0:01:50 lr 0.000841 time 0.2826 (0.2965) loss 3.6673 (3.7917) grad_norm 1.1018 (inf) [2021-04-15 21:50:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][890/1251] eta 0:01:47 lr 0.000841 time 0.2654 (0.2966) loss 4.3768 (3.7887) grad_norm 1.2926 (inf) [2021-04-15 21:50:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][900/1251] eta 0:01:44 lr 0.000841 time 0.3063 (0.2964) loss 3.5129 (3.7882) grad_norm 1.0249 (inf) [2021-04-15 21:50:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][910/1251] eta 0:01:41 lr 0.000841 time 0.2736 (0.2963) loss 4.8928 (3.7879) grad_norm 1.1342 (inf) [2021-04-15 21:50:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][920/1251] eta 0:01:38 lr 0.000841 time 0.2619 (0.2964) loss 4.3555 (3.7855) grad_norm 1.3073 (inf) [2021-04-15 21:50:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][930/1251] eta 0:01:35 lr 0.000841 time 0.2688 (0.2963) loss 4.3657 (3.7899) grad_norm 1.0971 (inf) [2021-04-15 21:50:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][940/1251] eta 0:01:32 lr 0.000841 time 0.4381 (0.2964) loss 4.3699 (3.7900) grad_norm 1.0573 (inf) [2021-04-15 21:50:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][950/1251] eta 0:01:29 lr 0.000841 time 0.2624 (0.2963) loss 3.1579 (3.7888) grad_norm 1.1893 (inf) [2021-04-15 21:50:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][960/1251] eta 0:01:26 lr 0.000841 time 0.2772 (0.2961) loss 3.3426 (3.7862) grad_norm 1.1610 (inf) [2021-04-15 21:50:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][970/1251] eta 0:01:23 lr 0.000841 time 0.2894 (0.2961) loss 4.1187 (3.7865) grad_norm 1.0736 (inf) [2021-04-15 21:50:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][980/1251] eta 0:01:20 lr 0.000841 time 0.2669 (0.2959) loss 3.8169 (3.7823) grad_norm 1.0283 (inf) [2021-04-15 21:50:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][990/1251] eta 0:01:17 lr 0.000841 time 0.3015 (0.2958) loss 4.7017 (3.7830) grad_norm 1.2259 (inf) [2021-04-15 21:50:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][1000/1251] eta 0:01:14 lr 0.000841 time 0.2864 (0.2957) loss 4.4857 (3.7886) grad_norm 1.0655 (inf) [2021-04-15 21:50:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][1010/1251] eta 0:01:11 lr 0.000841 time 0.2888 (0.2957) loss 3.6707 (3.7870) grad_norm 1.0464 (inf) [2021-04-15 21:50:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][1020/1251] eta 0:01:08 lr 0.000841 time 0.2697 (0.2956) loss 3.7371 (3.7872) grad_norm 1.2211 (inf) [2021-04-15 21:51:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][1030/1251] eta 0:01:05 lr 0.000841 time 0.3218 (0.2955) loss 3.5393 (3.7867) grad_norm 1.1340 (inf) [2021-04-15 21:51:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][1040/1251] eta 0:01:02 lr 0.000841 time 0.2627 (0.2954) loss 3.6488 (3.7885) grad_norm 0.9417 (inf) [2021-04-15 21:51:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][1050/1251] eta 0:00:59 lr 0.000841 time 0.2795 (0.2952) loss 2.7411 (3.7870) grad_norm 1.0609 (inf) [2021-04-15 21:51:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][1060/1251] eta 0:00:56 lr 0.000841 time 0.2733 (0.2952) loss 3.9661 (3.7860) grad_norm 1.0044 (inf) [2021-04-15 21:51:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][1070/1251] eta 0:00:53 lr 0.000841 time 0.3130 (0.2951) loss 3.0343 (3.7852) grad_norm 0.9897 (inf) [2021-04-15 21:51:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][1080/1251] eta 0:00:50 lr 0.000841 time 0.2629 (0.2950) loss 3.8436 (3.7852) grad_norm 1.1465 (inf) [2021-04-15 21:51:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][1090/1251] eta 0:00:47 lr 0.000841 time 0.2992 (0.2949) loss 4.6321 (3.7859) grad_norm 1.0322 (inf) [2021-04-15 21:51:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][1100/1251] eta 0:00:44 lr 0.000841 time 0.2648 (0.2947) loss 3.6604 (3.7835) grad_norm 1.2035 (inf) [2021-04-15 21:51:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][1110/1251] eta 0:00:41 lr 0.000840 time 0.2770 (0.2946) loss 4.2236 (3.7849) grad_norm 1.4210 (inf) [2021-04-15 21:51:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][1120/1251] eta 0:00:38 lr 0.000840 time 0.2715 (0.2945) loss 3.9288 (3.7866) grad_norm 1.2086 (inf) [2021-04-15 21:51:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][1130/1251] eta 0:00:35 lr 0.000840 time 0.2949 (0.2944) loss 3.8288 (3.7869) grad_norm 1.2303 (inf) [2021-04-15 21:51:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][1140/1251] eta 0:00:32 lr 0.000840 time 0.3051 (0.2944) loss 3.9784 (3.7886) grad_norm 1.1144 (inf) [2021-04-15 21:51:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][1150/1251] eta 0:00:29 lr 0.000840 time 0.2872 (0.2944) loss 4.1831 (3.7887) grad_norm 1.0294 (inf) [2021-04-15 21:51:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][1160/1251] eta 0:00:26 lr 0.000840 time 0.2968 (0.2943) loss 4.0528 (3.7902) grad_norm 1.0222 (inf) [2021-04-15 21:51:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][1170/1251] eta 0:00:23 lr 0.000840 time 0.3035 (0.2942) loss 4.2206 (3.7915) grad_norm 1.0768 (inf) [2021-04-15 21:51:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][1180/1251] eta 0:00:20 lr 0.000840 time 0.3037 (0.2941) loss 4.2898 (3.7911) grad_norm 1.5925 (inf) [2021-04-15 21:51:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][1190/1251] eta 0:00:17 lr 0.000840 time 0.2710 (0.2940) loss 4.7455 (3.7927) grad_norm 1.1804 (inf) [2021-04-15 21:51:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][1200/1251] eta 0:00:14 lr 0.000840 time 0.2872 (0.2939) loss 3.0529 (3.7918) grad_norm 1.0897 (inf) [2021-04-15 21:51:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][1210/1251] eta 0:00:12 lr 0.000840 time 0.2722 (0.2938) loss 4.2042 (3.7938) grad_norm 1.1630 (inf) [2021-04-15 21:51:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][1220/1251] eta 0:00:09 lr 0.000840 time 0.2674 (0.2938) loss 2.5792 (3.7905) grad_norm 1.2889 (inf) [2021-04-15 21:51:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][1230/1251] eta 0:00:06 lr 0.000840 time 0.2743 (0.2938) loss 3.7601 (3.7900) grad_norm 1.2842 (inf) [2021-04-15 21:52:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][1240/1251] eta 0:00:03 lr 0.000840 time 0.2489 (0.2937) loss 2.6944 (3.7900) grad_norm 1.0647 (inf) [2021-04-15 21:52:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [78/300][1250/1251] eta 0:00:00 lr 0.000840 time 0.2482 (0.2934) loss 2.5450 (3.7881) grad_norm 1.0164 (inf) [2021-04-15 21:52:05 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 78 training takes 0:06:09 [2021-04-15 21:52:05 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_78.pth saving...... [2021-04-15 21:52:17 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_78.pth saved !!! [2021-04-15 21:52:19 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.288 (1.288) Loss 1.2436 (1.2436) Acc@1 71.484 (71.484) Acc@5 90.137 (90.137) [2021-04-15 21:52:20 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.183 (0.242) Loss 1.0811 (1.1889) Acc@1 74.219 (72.488) Acc@5 92.871 (91.451) [2021-04-15 21:52:22 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.147 (0.218) Loss 1.1707 (1.1845) Acc@1 72.461 (72.628) Acc@5 92.188 (91.453) [2021-04-15 21:52:25 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.097 (0.237) Loss 1.2190 (1.1929) Acc@1 71.582 (72.499) Acc@5 91.016 (91.384) [2021-04-15 21:52:27 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.228) Loss 1.2040 (1.1975) Acc@1 71.973 (72.273) Acc@5 91.016 (91.316) [2021-04-15 21:52:30 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 72.406 Acc@5 91.342 [2021-04-15 21:52:30 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 72.4% [2021-04-15 21:52:30 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 72.53% [2021-04-15 21:52:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][0/1251] eta 1:18:02 lr 0.000840 time 3.7428 (3.7428) loss 3.6754 (3.6754) grad_norm 1.2514 (1.2514) [2021-04-15 21:52:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][10/1251] eta 0:12:24 lr 0.000840 time 0.2835 (0.5996) loss 4.5789 (3.9741) grad_norm 1.1003 (1.1475) [2021-04-15 21:52:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][20/1251] eta 0:09:08 lr 0.000840 time 0.2690 (0.4455) loss 3.9066 (3.7968) grad_norm 1.0868 (1.1433) [2021-04-15 21:52:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][30/1251] eta 0:07:59 lr 0.000840 time 0.2734 (0.3924) loss 2.6931 (3.7957) grad_norm 1.0070 (1.1459) [2021-04-15 21:52:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][40/1251] eta 0:07:22 lr 0.000840 time 0.2616 (0.3657) loss 3.4224 (3.7221) grad_norm 1.1219 (1.1413) [2021-04-15 21:52:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][50/1251] eta 0:06:58 lr 0.000840 time 0.2682 (0.3484) loss 3.6378 (3.7400) grad_norm 1.2269 (1.1405) [2021-04-15 21:52:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][60/1251] eta 0:06:44 lr 0.000840 time 0.3064 (0.3393) loss 4.3768 (3.7759) grad_norm 1.0547 (1.1462) [2021-04-15 21:52:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][70/1251] eta 0:06:30 lr 0.000840 time 0.2831 (0.3304) loss 4.2559 (3.7937) grad_norm 1.1649 (1.1404) [2021-04-15 21:52:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][80/1251] eta 0:06:19 lr 0.000840 time 0.2764 (0.3244) loss 3.9830 (3.7612) grad_norm 1.9314 (1.1582) [2021-04-15 21:53:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][90/1251] eta 0:06:13 lr 0.000840 time 0.2806 (0.3213) loss 4.2911 (3.7460) grad_norm 1.3944 (1.1865) [2021-04-15 21:53:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][100/1251] eta 0:06:05 lr 0.000840 time 0.3078 (0.3179) loss 4.2661 (3.7513) grad_norm 1.0890 (1.1831) [2021-04-15 21:53:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][110/1251] eta 0:05:59 lr 0.000840 time 0.2702 (0.3150) loss 2.7034 (3.7158) grad_norm 1.1618 (1.1783) [2021-04-15 21:53:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][120/1251] eta 0:05:54 lr 0.000840 time 0.2832 (0.3138) loss 4.0773 (3.6997) grad_norm 1.1113 (1.1722) [2021-04-15 21:53:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][130/1251] eta 0:05:50 lr 0.000840 time 0.3048 (0.3129) loss 3.6700 (3.7041) grad_norm 1.2400 (1.1668) [2021-04-15 21:53:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][140/1251] eta 0:05:45 lr 0.000840 time 0.3001 (0.3108) loss 3.3968 (3.7122) grad_norm 0.9974 (1.1617) [2021-04-15 21:53:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][150/1251] eta 0:05:41 lr 0.000840 time 0.4709 (0.3100) loss 3.9406 (3.7165) grad_norm 1.2646 (1.1622) [2021-04-15 21:53:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][160/1251] eta 0:05:36 lr 0.000840 time 0.2997 (0.3087) loss 4.2126 (3.7308) grad_norm 1.1841 (1.1592) [2021-04-15 21:53:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][170/1251] eta 0:05:31 lr 0.000840 time 0.2571 (0.3070) loss 4.3612 (3.7386) grad_norm 1.1666 (1.1635) [2021-04-15 21:53:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][180/1251] eta 0:05:27 lr 0.000840 time 0.2846 (0.3055) loss 4.3834 (3.7465) grad_norm 1.1505 (1.1661) [2021-04-15 21:53:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][190/1251] eta 0:05:22 lr 0.000839 time 0.2825 (0.3043) loss 3.9952 (3.7498) grad_norm 1.0429 (1.1680) [2021-04-15 21:53:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][200/1251] eta 0:05:18 lr 0.000839 time 0.2717 (0.3032) loss 2.9362 (3.7530) grad_norm 1.0719 (1.1648) [2021-04-15 21:53:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][210/1251] eta 0:05:14 lr 0.000839 time 0.2695 (0.3025) loss 4.4221 (3.7516) grad_norm 1.3221 (1.1645) [2021-04-15 21:53:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][220/1251] eta 0:05:11 lr 0.000839 time 0.2808 (0.3026) loss 4.8023 (3.7561) grad_norm 1.1188 (1.1658) [2021-04-15 21:53:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][230/1251] eta 0:05:08 lr 0.000839 time 0.2787 (0.3025) loss 4.0751 (3.7524) grad_norm 1.0822 (1.1653) [2021-04-15 21:53:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][240/1251] eta 0:05:05 lr 0.000839 time 0.2691 (0.3017) loss 3.8303 (3.7654) grad_norm 1.0881 (1.1635) [2021-04-15 21:53:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][250/1251] eta 0:05:01 lr 0.000839 time 0.2800 (0.3010) loss 3.6459 (3.7587) grad_norm 1.4303 (1.1647) [2021-04-15 21:53:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][260/1251] eta 0:04:57 lr 0.000839 time 0.2836 (0.3004) loss 4.2875 (3.7648) grad_norm 1.3237 (1.1697) [2021-04-15 21:53:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][270/1251] eta 0:04:54 lr 0.000839 time 0.2967 (0.2998) loss 3.9425 (3.7766) grad_norm 1.1040 (1.1704) [2021-04-15 21:53:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][280/1251] eta 0:04:50 lr 0.000839 time 0.2943 (0.2993) loss 4.1312 (3.7748) grad_norm 1.0151 (1.1712) [2021-04-15 21:53:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][290/1251] eta 0:04:47 lr 0.000839 time 0.2864 (0.2989) loss 4.0388 (3.7679) grad_norm 1.1466 (1.1705) [2021-04-15 21:54:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][300/1251] eta 0:04:43 lr 0.000839 time 0.2652 (0.2984) loss 3.6093 (3.7619) grad_norm 1.1711 (1.1696) [2021-04-15 21:54:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][310/1251] eta 0:04:40 lr 0.000839 time 0.2806 (0.2979) loss 3.9909 (3.7608) grad_norm 1.0956 (1.1687) [2021-04-15 21:54:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][320/1251] eta 0:04:36 lr 0.000839 time 0.2447 (0.2973) loss 3.7018 (3.7575) grad_norm 1.2300 (1.1688) [2021-04-15 21:54:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][330/1251] eta 0:04:33 lr 0.000839 time 0.2676 (0.2969) loss 4.1612 (3.7646) grad_norm 1.0576 (1.1732) [2021-04-15 21:54:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][340/1251] eta 0:04:30 lr 0.000839 time 0.3133 (0.2965) loss 3.6627 (3.7616) grad_norm 1.1684 (1.1734) [2021-04-15 21:54:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][350/1251] eta 0:04:26 lr 0.000839 time 0.3043 (0.2963) loss 2.8260 (3.7637) grad_norm 1.3620 (1.1772) [2021-04-15 21:54:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][360/1251] eta 0:04:23 lr 0.000839 time 0.2483 (0.2960) loss 3.7047 (3.7594) grad_norm 1.1176 (1.1763) [2021-04-15 21:54:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][370/1251] eta 0:04:20 lr 0.000839 time 0.2836 (0.2962) loss 4.3140 (3.7591) grad_norm 1.4383 (1.1764) [2021-04-15 21:54:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][380/1251] eta 0:04:17 lr 0.000839 time 0.3172 (0.2959) loss 4.0045 (3.7596) grad_norm 1.3721 (1.1781) [2021-04-15 21:54:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][390/1251] eta 0:04:15 lr 0.000839 time 0.2517 (0.2962) loss 3.7820 (3.7549) grad_norm 1.2672 (1.1773) [2021-04-15 21:54:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][400/1251] eta 0:04:12 lr 0.000839 time 0.2692 (0.2963) loss 3.6037 (3.7561) grad_norm 1.3532 (1.1776) [2021-04-15 21:54:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][410/1251] eta 0:04:08 lr 0.000839 time 0.3044 (0.2959) loss 3.7943 (3.7614) grad_norm 1.2385 (1.1785) [2021-04-15 21:54:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][420/1251] eta 0:04:05 lr 0.000839 time 0.2768 (0.2956) loss 3.3123 (3.7583) grad_norm 1.3396 (1.1796) [2021-04-15 21:54:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][430/1251] eta 0:04:02 lr 0.000839 time 0.2760 (0.2957) loss 2.8372 (3.7579) grad_norm 1.0745 (1.1801) [2021-04-15 21:54:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][440/1251] eta 0:03:59 lr 0.000839 time 0.2931 (0.2959) loss 3.7156 (3.7561) grad_norm 1.2047 (1.1805) [2021-04-15 21:54:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][450/1251] eta 0:03:56 lr 0.000839 time 0.2923 (0.2957) loss 3.1989 (3.7596) grad_norm 1.2252 (1.1805) [2021-04-15 21:54:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][460/1251] eta 0:03:53 lr 0.000839 time 0.2865 (0.2955) loss 4.4046 (3.7608) grad_norm 1.1329 (1.1814) [2021-04-15 21:54:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][470/1251] eta 0:03:50 lr 0.000839 time 0.3141 (0.2953) loss 3.9264 (3.7603) grad_norm 1.3340 (1.1813) [2021-04-15 21:54:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][480/1251] eta 0:03:47 lr 0.000839 time 0.2945 (0.2950) loss 3.4108 (3.7571) grad_norm 1.0540 (1.1825) [2021-04-15 21:54:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][490/1251] eta 0:03:44 lr 0.000839 time 0.2968 (0.2948) loss 3.9689 (3.7563) grad_norm 1.2580 (1.1825) [2021-04-15 21:54:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][500/1251] eta 0:03:41 lr 0.000839 time 0.2634 (0.2945) loss 3.2192 (3.7575) grad_norm 1.0470 (1.1809) [2021-04-15 21:55:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][510/1251] eta 0:03:38 lr 0.000838 time 0.3169 (0.2943) loss 2.9537 (3.7597) grad_norm 1.3619 (1.1795) [2021-04-15 21:55:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][520/1251] eta 0:03:34 lr 0.000838 time 0.2874 (0.2941) loss 4.2844 (3.7627) grad_norm 1.1403 (1.1817) [2021-04-15 21:55:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][530/1251] eta 0:03:31 lr 0.000838 time 0.2804 (0.2939) loss 4.0605 (3.7599) grad_norm 1.0771 (1.1811) [2021-04-15 21:55:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][540/1251] eta 0:03:28 lr 0.000838 time 0.2683 (0.2940) loss 3.1758 (3.7631) grad_norm 1.0626 (1.1806) [2021-04-15 21:55:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][550/1251] eta 0:03:25 lr 0.000838 time 0.2915 (0.2938) loss 4.4036 (3.7683) grad_norm 1.2309 (1.1811) [2021-04-15 21:55:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][560/1251] eta 0:03:22 lr 0.000838 time 0.2753 (0.2936) loss 4.3133 (3.7677) grad_norm 1.1989 (1.1816) [2021-04-15 21:55:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][570/1251] eta 0:03:19 lr 0.000838 time 0.2822 (0.2936) loss 3.6246 (3.7671) grad_norm 1.3979 (1.1823) [2021-04-15 21:55:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][580/1251] eta 0:03:16 lr 0.000838 time 0.2977 (0.2935) loss 4.0843 (3.7706) grad_norm 1.2520 (1.1837) [2021-04-15 21:55:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][590/1251] eta 0:03:13 lr 0.000838 time 0.2985 (0.2933) loss 4.4350 (3.7744) grad_norm 1.1293 (1.1842) [2021-04-15 21:55:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][600/1251] eta 0:03:10 lr 0.000838 time 0.2660 (0.2932) loss 4.4549 (3.7718) grad_norm 1.2530 (1.1853) [2021-04-15 21:55:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][610/1251] eta 0:03:07 lr 0.000838 time 0.2741 (0.2930) loss 4.3660 (3.7711) grad_norm 1.4149 (1.1853) [2021-04-15 21:55:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][620/1251] eta 0:03:04 lr 0.000838 time 0.2562 (0.2928) loss 3.5759 (3.7725) grad_norm 1.0767 (1.1844) [2021-04-15 21:55:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][630/1251] eta 0:03:01 lr 0.000838 time 0.2708 (0.2927) loss 4.5156 (3.7733) grad_norm 1.0599 (1.1829) [2021-04-15 21:55:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][640/1251] eta 0:02:58 lr 0.000838 time 0.2596 (0.2926) loss 3.1121 (3.7700) grad_norm 1.0694 (1.1843) [2021-04-15 21:55:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][650/1251] eta 0:02:55 lr 0.000838 time 0.2902 (0.2924) loss 2.5104 (3.7669) grad_norm 1.2728 (1.1837) [2021-04-15 21:55:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][660/1251] eta 0:02:52 lr 0.000838 time 0.2867 (0.2922) loss 3.9824 (3.7656) grad_norm 1.3308 (1.1832) [2021-04-15 21:55:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][670/1251] eta 0:02:49 lr 0.000838 time 0.2865 (0.2924) loss 3.8520 (3.7667) grad_norm 1.2527 (1.1822) [2021-04-15 21:55:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][680/1251] eta 0:02:46 lr 0.000838 time 0.2909 (0.2922) loss 3.6508 (3.7690) grad_norm 1.0510 (1.1820) [2021-04-15 21:55:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][690/1251] eta 0:02:44 lr 0.000838 time 0.2937 (0.2924) loss 4.1932 (3.7693) grad_norm 1.2672 (1.1812) [2021-04-15 21:55:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][700/1251] eta 0:02:41 lr 0.000838 time 0.2967 (0.2924) loss 3.5731 (3.7663) grad_norm 1.3997 (1.1822) [2021-04-15 21:55:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][710/1251] eta 0:02:38 lr 0.000838 time 0.2963 (0.2924) loss 2.2827 (3.7641) grad_norm 1.5166 (1.1827) [2021-04-15 21:56:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][720/1251] eta 0:02:35 lr 0.000838 time 0.2668 (0.2923) loss 4.1429 (3.7649) grad_norm 1.0628 (1.1822) [2021-04-15 21:56:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][730/1251] eta 0:02:32 lr 0.000838 time 0.2586 (0.2924) loss 4.1548 (3.7635) grad_norm 1.3545 (1.1835) [2021-04-15 21:56:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][740/1251] eta 0:02:29 lr 0.000838 time 0.2920 (0.2923) loss 3.2654 (3.7641) grad_norm 1.1872 (1.1830) [2021-04-15 21:56:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][750/1251] eta 0:02:26 lr 0.000838 time 0.2781 (0.2921) loss 3.8579 (3.7650) grad_norm 1.3732 (1.1825) [2021-04-15 21:56:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][760/1251] eta 0:02:23 lr 0.000838 time 0.3077 (0.2920) loss 4.3854 (3.7640) grad_norm 1.2644 (1.1841) [2021-04-15 21:56:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][770/1251] eta 0:02:20 lr 0.000838 time 0.2613 (0.2918) loss 3.7969 (3.7666) grad_norm 1.3109 (1.1852) [2021-04-15 21:56:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][780/1251] eta 0:02:17 lr 0.000838 time 0.2693 (0.2917) loss 4.7235 (3.7685) grad_norm 1.0712 (1.1850) [2021-04-15 21:56:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][790/1251] eta 0:02:14 lr 0.000838 time 0.2707 (0.2917) loss 3.9959 (3.7704) grad_norm 1.2379 (1.1854) [2021-04-15 21:56:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][800/1251] eta 0:02:11 lr 0.000838 time 0.4239 (0.2918) loss 4.5655 (3.7727) grad_norm 1.1972 (1.1859) [2021-04-15 21:56:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][810/1251] eta 0:02:08 lr 0.000838 time 0.2754 (0.2916) loss 3.8186 (3.7710) grad_norm 0.9713 (1.1852) [2021-04-15 21:56:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][820/1251] eta 0:02:05 lr 0.000838 time 0.2890 (0.2915) loss 3.8098 (3.7738) grad_norm 1.4170 (1.1853) [2021-04-15 21:56:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][830/1251] eta 0:02:02 lr 0.000838 time 0.2737 (0.2914) loss 4.2024 (3.7763) grad_norm 1.2653 (1.1852) [2021-04-15 21:56:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][840/1251] eta 0:01:59 lr 0.000837 time 0.2605 (0.2913) loss 4.6984 (3.7788) grad_norm 1.2348 (1.1852) [2021-04-15 21:56:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][850/1251] eta 0:01:56 lr 0.000837 time 0.2884 (0.2913) loss 4.1453 (3.7782) grad_norm 1.1583 (1.1852) [2021-04-15 21:56:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][860/1251] eta 0:01:53 lr 0.000837 time 0.3074 (0.2912) loss 3.4924 (3.7811) grad_norm 1.0269 (1.1852) [2021-04-15 21:56:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][870/1251] eta 0:01:50 lr 0.000837 time 0.2974 (0.2911) loss 4.0285 (3.7855) grad_norm 1.0181 (1.1841) [2021-04-15 21:56:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][880/1251] eta 0:01:47 lr 0.000837 time 0.2603 (0.2909) loss 4.2460 (3.7837) grad_norm 1.1728 (1.1839) [2021-04-15 21:56:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][890/1251] eta 0:01:44 lr 0.000837 time 0.2759 (0.2908) loss 3.8799 (3.7855) grad_norm 1.0512 (1.1839) [2021-04-15 21:56:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][900/1251] eta 0:01:42 lr 0.000837 time 0.2670 (0.2906) loss 3.8664 (3.7851) grad_norm 1.2951 (1.1845) [2021-04-15 21:56:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][910/1251] eta 0:01:39 lr 0.000837 time 0.2671 (0.2906) loss 4.0451 (3.7871) grad_norm 1.1466 (1.1852) [2021-04-15 21:56:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][920/1251] eta 0:01:36 lr 0.000837 time 0.2766 (0.2905) loss 3.2430 (3.7886) grad_norm 1.2227 (1.1856) [2021-04-15 21:57:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][930/1251] eta 0:01:33 lr 0.000837 time 0.2805 (0.2903) loss 3.9419 (3.7909) grad_norm 1.1358 (1.1862) [2021-04-15 21:57:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][940/1251] eta 0:01:30 lr 0.000837 time 0.3181 (0.2902) loss 4.7621 (3.7931) grad_norm 1.3920 (1.1863) [2021-04-15 21:57:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][950/1251] eta 0:01:27 lr 0.000837 time 0.2861 (0.2902) loss 3.7456 (3.7951) grad_norm 1.1893 (1.1875) [2021-04-15 21:57:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][960/1251] eta 0:01:24 lr 0.000837 time 0.2900 (0.2901) loss 3.6685 (3.7928) grad_norm 0.9986 (1.1871) [2021-04-15 21:57:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][970/1251] eta 0:01:21 lr 0.000837 time 0.2724 (0.2900) loss 2.3404 (3.7908) grad_norm 1.7554 (1.1871) [2021-04-15 21:57:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][980/1251] eta 0:01:18 lr 0.000837 time 0.2728 (0.2900) loss 4.3303 (3.7926) grad_norm 1.1606 (1.1874) [2021-04-15 21:57:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][990/1251] eta 0:01:15 lr 0.000837 time 0.2856 (0.2900) loss 2.7496 (3.7904) grad_norm 1.2378 (1.1873) [2021-04-15 21:57:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][1000/1251] eta 0:01:12 lr 0.000837 time 0.2559 (0.2898) loss 3.5822 (3.7888) grad_norm 1.1727 (1.1866) [2021-04-15 21:57:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][1010/1251] eta 0:01:09 lr 0.000837 time 0.2727 (0.2898) loss 4.1639 (3.7908) grad_norm 1.0389 (1.1863) [2021-04-15 21:57:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][1020/1251] eta 0:01:06 lr 0.000837 time 0.2755 (0.2898) loss 3.8466 (3.7894) grad_norm 1.0388 (1.1861) [2021-04-15 21:57:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][1030/1251] eta 0:01:04 lr 0.000837 time 0.2795 (0.2897) loss 3.8312 (3.7882) grad_norm 1.0989 (1.1863) [2021-04-15 21:57:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][1040/1251] eta 0:01:01 lr 0.000837 time 0.2931 (0.2896) loss 4.1576 (3.7879) grad_norm 0.9865 (1.1857) [2021-04-15 21:57:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][1050/1251] eta 0:00:58 lr 0.000837 time 0.3094 (0.2895) loss 2.8937 (3.7870) grad_norm 1.4944 (1.1863) [2021-04-15 21:57:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][1060/1251] eta 0:00:55 lr 0.000837 time 0.2641 (0.2894) loss 4.1147 (3.7873) grad_norm 1.0945 (1.1868) [2021-04-15 21:57:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][1070/1251] eta 0:00:52 lr 0.000837 time 0.2963 (0.2893) loss 4.2356 (3.7878) grad_norm 0.9799 (1.1875) [2021-04-15 21:57:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][1080/1251] eta 0:00:49 lr 0.000837 time 0.2690 (0.2893) loss 2.7319 (3.7878) grad_norm 1.1085 (1.1876) [2021-04-15 21:57:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][1090/1251] eta 0:00:46 lr 0.000837 time 0.3145 (0.2892) loss 4.2827 (3.7912) grad_norm 1.1062 (1.1880) [2021-04-15 21:57:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][1100/1251] eta 0:00:43 lr 0.000837 time 0.2856 (0.2891) loss 3.3303 (3.7892) grad_norm 1.2051 (1.1876) [2021-04-15 21:57:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][1110/1251] eta 0:00:40 lr 0.000837 time 0.2771 (0.2890) loss 4.1795 (3.7905) grad_norm 1.1996 (1.1873) [2021-04-15 21:57:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][1120/1251] eta 0:00:37 lr 0.000837 time 0.2782 (0.2890) loss 4.3169 (3.7923) grad_norm 1.4247 (1.1877) [2021-04-15 21:57:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][1130/1251] eta 0:00:34 lr 0.000837 time 0.2965 (0.2889) loss 3.2754 (3.7914) grad_norm 1.6204 (1.1882) [2021-04-15 21:58:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][1140/1251] eta 0:00:32 lr 0.000837 time 0.2872 (0.2889) loss 4.5179 (3.7919) grad_norm 1.2469 (1.1880) [2021-04-15 21:58:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][1150/1251] eta 0:00:29 lr 0.000837 time 0.2767 (0.2888) loss 3.4380 (3.7916) grad_norm 1.2036 (1.1876) [2021-04-15 21:58:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][1160/1251] eta 0:00:26 lr 0.000836 time 0.2846 (0.2887) loss 3.9152 (3.7889) grad_norm 1.1751 (1.1872) [2021-04-15 21:58:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][1170/1251] eta 0:00:23 lr 0.000836 time 0.2766 (0.2887) loss 4.3468 (3.7898) grad_norm 1.3076 (1.1867) [2021-04-15 21:58:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][1180/1251] eta 0:00:20 lr 0.000836 time 0.2881 (0.2886) loss 2.6925 (3.7893) grad_norm 1.3194 (1.1866) [2021-04-15 21:58:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][1190/1251] eta 0:00:17 lr 0.000836 time 0.3105 (0.2885) loss 4.0721 (3.7876) grad_norm 1.1121 (1.1863) [2021-04-15 21:58:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][1200/1251] eta 0:00:14 lr 0.000836 time 0.2796 (0.2884) loss 4.1895 (3.7895) grad_norm 1.4100 (1.1864) [2021-04-15 21:58:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][1210/1251] eta 0:00:11 lr 0.000836 time 0.2682 (0.2883) loss 2.7239 (3.7862) grad_norm 1.0700 (1.1858) [2021-04-15 21:58:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][1220/1251] eta 0:00:08 lr 0.000836 time 0.2764 (0.2883) loss 3.1462 (3.7863) grad_norm 1.1608 (1.1855) [2021-04-15 21:58:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][1230/1251] eta 0:00:06 lr 0.000836 time 0.2889 (0.2882) loss 4.2742 (3.7882) grad_norm 1.0338 (1.1847) [2021-04-15 21:58:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][1240/1251] eta 0:00:03 lr 0.000836 time 0.3707 (0.2882) loss 2.6984 (3.7866) grad_norm 0.9580 (1.1841) [2021-04-15 21:58:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [79/300][1250/1251] eta 0:00:00 lr 0.000836 time 0.2488 (0.2879) loss 3.3470 (3.7852) grad_norm 1.0902 (1.1844) [2021-04-15 21:58:34 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 79 training takes 0:06:04 [2021-04-15 21:58:34 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_79.pth saving...... [2021-04-15 21:58:49 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_79.pth saved !!! [2021-04-15 21:58:51 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.171 (1.171) Loss 1.1721 (1.1721) Acc@1 73.340 (73.340) Acc@5 91.406 (91.406) [2021-04-15 21:58:52 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.562 (0.273) Loss 1.1573 (1.1568) Acc@1 72.949 (72.949) Acc@5 92.773 (92.063) [2021-04-15 21:58:54 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.092 (0.235) Loss 1.1780 (1.1692) Acc@1 74.414 (72.884) Acc@5 91.406 (91.811) [2021-04-15 21:58:56 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.164 (0.230) Loss 1.1704 (1.1769) Acc@1 72.754 (72.659) Acc@5 92.383 (91.617) [2021-04-15 21:58:58 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.218) Loss 1.2206 (1.1868) Acc@1 72.461 (72.482) Acc@5 91.016 (91.473) [2021-04-15 21:59:02 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 72.470 Acc@5 91.476 [2021-04-15 21:59:02 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 72.5% [2021-04-15 21:59:02 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 72.53% [2021-04-15 21:59:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][0/1251] eta 1:44:08 lr 0.000836 time 4.9950 (4.9950) loss 3.0519 (3.0519) grad_norm 1.1354 (1.1354) [2021-04-15 21:59:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][10/1251] eta 0:14:50 lr 0.000836 time 0.4357 (0.7177) loss 4.2468 (3.7832) grad_norm 1.2192 (1.1661) [2021-04-15 21:59:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][20/1251] eta 0:10:28 lr 0.000836 time 0.2832 (0.5105) loss 3.4949 (3.7263) grad_norm 1.1862 (1.1922) [2021-04-15 21:59:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][30/1251] eta 0:08:53 lr 0.000836 time 0.2944 (0.4369) loss 3.6643 (3.7659) grad_norm 0.9866 (1.1996) [2021-04-15 21:59:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][40/1251] eta 0:08:03 lr 0.000836 time 0.2679 (0.3990) loss 3.6288 (3.7489) grad_norm 1.0648 (1.2135) [2021-04-15 21:59:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][50/1251] eta 0:07:32 lr 0.000836 time 0.2945 (0.3766) loss 3.2188 (3.7913) grad_norm 1.0823 (1.1932) [2021-04-15 21:59:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][60/1251] eta 0:07:09 lr 0.000836 time 0.2619 (0.3605) loss 4.6861 (3.7719) grad_norm 1.1667 (1.1998) [2021-04-15 21:59:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][70/1251] eta 0:06:52 lr 0.000836 time 0.3035 (0.3494) loss 3.7641 (3.7747) grad_norm 1.0290 (1.1977) [2021-04-15 21:59:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][80/1251] eta 0:06:39 lr 0.000836 time 0.2799 (0.3412) loss 2.3129 (3.7329) grad_norm 1.2160 (1.1939) [2021-04-15 21:59:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][90/1251] eta 0:06:29 lr 0.000836 time 0.2651 (0.3352) loss 2.2275 (3.7113) grad_norm 1.1891 (1.1900) [2021-04-15 21:59:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][100/1251] eta 0:06:19 lr 0.000836 time 0.2554 (0.3299) loss 4.4161 (3.7011) grad_norm 1.2525 (1.1881) [2021-04-15 21:59:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][110/1251] eta 0:06:11 lr 0.000836 time 0.2805 (0.3259) loss 3.7600 (3.7190) grad_norm 1.4980 (1.2013) [2021-04-15 21:59:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][120/1251] eta 0:06:04 lr 0.000836 time 0.2932 (0.3225) loss 3.2623 (3.7125) grad_norm 1.2635 (1.2074) [2021-04-15 21:59:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][130/1251] eta 0:05:57 lr 0.000836 time 0.2862 (0.3191) loss 3.4007 (3.7093) grad_norm 1.3130 (1.2036) [2021-04-15 21:59:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][140/1251] eta 0:05:50 lr 0.000836 time 0.2666 (0.3159) loss 2.6361 (3.7261) grad_norm 1.1981 (1.2039) [2021-04-15 21:59:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][150/1251] eta 0:05:46 lr 0.000836 time 0.2799 (0.3152) loss 4.3279 (3.7401) grad_norm 1.2857 (1.1999) [2021-04-15 21:59:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][160/1251] eta 0:05:42 lr 0.000836 time 0.2556 (0.3135) loss 4.3849 (3.7471) grad_norm 1.2721 (1.2001) [2021-04-15 21:59:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][170/1251] eta 0:05:36 lr 0.000836 time 0.2608 (0.3113) loss 4.7445 (3.7513) grad_norm 1.0582 (1.1986) [2021-04-15 21:59:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][180/1251] eta 0:05:32 lr 0.000836 time 0.4503 (0.3103) loss 4.5323 (3.7355) grad_norm 1.5216 (1.2055) [2021-04-15 22:00:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][190/1251] eta 0:05:27 lr 0.000836 time 0.2653 (0.3086) loss 4.4160 (3.7293) grad_norm 1.0916 (1.1999) [2021-04-15 22:00:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][200/1251] eta 0:05:22 lr 0.000836 time 0.2594 (0.3072) loss 4.2083 (3.7363) grad_norm 1.0553 (1.1969) [2021-04-15 22:00:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][210/1251] eta 0:05:18 lr 0.000836 time 0.2904 (0.3060) loss 4.2928 (3.7352) grad_norm 1.0471 (1.1942) [2021-04-15 22:00:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][220/1251] eta 0:05:15 lr 0.000836 time 0.2792 (0.3055) loss 3.3723 (3.7289) grad_norm 1.1937 (1.1903) [2021-04-15 22:00:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][230/1251] eta 0:05:11 lr 0.000836 time 0.3084 (0.3049) loss 4.1917 (3.7285) grad_norm 1.2297 (1.1901) [2021-04-15 22:00:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][240/1251] eta 0:05:07 lr 0.000835 time 0.2693 (0.3038) loss 3.7899 (3.7332) grad_norm 1.1415 (1.1907) [2021-04-15 22:00:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][250/1251] eta 0:05:03 lr 0.000835 time 0.2669 (0.3027) loss 4.8130 (3.7378) grad_norm 1.1099 (1.1906) [2021-04-15 22:00:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][260/1251] eta 0:04:59 lr 0.000835 time 0.3054 (0.3018) loss 3.2122 (3.7381) grad_norm 1.1909 (1.1917) [2021-04-15 22:00:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][270/1251] eta 0:04:55 lr 0.000835 time 0.2859 (0.3011) loss 3.8180 (3.7346) grad_norm 1.1304 (1.1923) [2021-04-15 22:00:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][280/1251] eta 0:04:51 lr 0.000835 time 0.2735 (0.3003) loss 4.3668 (3.7362) grad_norm 1.1651 (1.1922) [2021-04-15 22:00:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][290/1251] eta 0:04:47 lr 0.000835 time 0.2682 (0.2995) loss 4.0278 (3.7395) grad_norm 1.0762 (inf) [2021-04-15 22:00:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][300/1251] eta 0:04:44 lr 0.000835 time 0.2530 (0.2987) loss 3.9280 (3.7391) grad_norm 1.0319 (inf) [2021-04-15 22:00:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][310/1251] eta 0:04:40 lr 0.000835 time 0.2719 (0.2981) loss 4.1851 (3.7473) grad_norm 1.3875 (inf) [2021-04-15 22:00:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][320/1251] eta 0:04:37 lr 0.000835 time 0.3098 (0.2979) loss 3.8483 (3.7536) grad_norm 1.1467 (inf) [2021-04-15 22:00:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][330/1251] eta 0:04:33 lr 0.000835 time 0.2878 (0.2974) loss 3.3387 (3.7484) grad_norm 1.3921 (inf) [2021-04-15 22:00:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][340/1251] eta 0:04:31 lr 0.000835 time 0.2529 (0.2976) loss 3.8979 (3.7477) grad_norm 1.1739 (inf) [2021-04-15 22:00:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][350/1251] eta 0:04:27 lr 0.000835 time 0.2800 (0.2973) loss 3.0731 (3.7593) grad_norm 1.0314 (inf) [2021-04-15 22:00:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][360/1251] eta 0:04:24 lr 0.000835 time 0.2531 (0.2967) loss 2.7176 (3.7604) grad_norm 1.0426 (inf) [2021-04-15 22:00:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][370/1251] eta 0:04:21 lr 0.000835 time 0.2849 (0.2965) loss 3.1744 (3.7608) grad_norm 1.1699 (inf) [2021-04-15 22:00:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][380/1251] eta 0:04:17 lr 0.000835 time 0.2633 (0.2961) loss 3.7066 (3.7611) grad_norm 1.0366 (inf) [2021-04-15 22:00:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][390/1251] eta 0:04:14 lr 0.000835 time 0.2788 (0.2957) loss 3.8556 (3.7630) grad_norm 1.4063 (inf) [2021-04-15 22:01:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][400/1251] eta 0:04:11 lr 0.000835 time 0.2727 (0.2953) loss 3.0394 (3.7634) grad_norm 1.1844 (inf) [2021-04-15 22:01:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][410/1251] eta 0:04:08 lr 0.000835 time 0.2675 (0.2950) loss 3.7870 (3.7623) grad_norm 1.6127 (inf) [2021-04-15 22:01:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][420/1251] eta 0:04:04 lr 0.000835 time 0.2690 (0.2945) loss 3.5121 (3.7614) grad_norm 1.2832 (inf) [2021-04-15 22:01:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][430/1251] eta 0:04:01 lr 0.000835 time 0.3065 (0.2942) loss 4.3936 (3.7627) grad_norm 1.1435 (inf) [2021-04-15 22:01:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][440/1251] eta 0:03:58 lr 0.000835 time 0.2954 (0.2938) loss 3.3698 (3.7546) grad_norm 1.1593 (inf) [2021-04-15 22:01:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][450/1251] eta 0:03:55 lr 0.000835 time 0.2409 (0.2939) loss 4.1961 (3.7537) grad_norm 1.1163 (inf) [2021-04-15 22:01:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][460/1251] eta 0:03:52 lr 0.000835 time 0.2819 (0.2937) loss 2.7476 (3.7537) grad_norm 1.0790 (inf) [2021-04-15 22:01:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][470/1251] eta 0:03:49 lr 0.000835 time 0.2521 (0.2934) loss 3.9683 (3.7531) grad_norm 1.1528 (inf) [2021-04-15 22:01:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][480/1251] eta 0:03:45 lr 0.000835 time 0.3033 (0.2931) loss 4.3419 (3.7524) grad_norm 1.2017 (inf) [2021-04-15 22:01:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][490/1251] eta 0:03:43 lr 0.000835 time 0.2759 (0.2932) loss 4.0674 (3.7562) grad_norm 1.4011 (inf) [2021-04-15 22:01:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][500/1251] eta 0:03:40 lr 0.000835 time 0.2977 (0.2930) loss 2.6563 (3.7556) grad_norm 1.1106 (inf) [2021-04-15 22:01:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][510/1251] eta 0:03:36 lr 0.000835 time 0.2947 (0.2928) loss 4.1107 (3.7600) grad_norm 1.2004 (inf) [2021-04-15 22:01:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][520/1251] eta 0:03:33 lr 0.000835 time 0.2773 (0.2926) loss 3.1907 (3.7554) grad_norm 1.2334 (inf) [2021-04-15 22:01:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][530/1251] eta 0:03:30 lr 0.000835 time 0.2811 (0.2923) loss 4.2426 (3.7615) grad_norm 1.0856 (inf) [2021-04-15 22:01:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][540/1251] eta 0:03:27 lr 0.000835 time 0.2718 (0.2920) loss 3.2328 (3.7643) grad_norm 1.3103 (inf) [2021-04-15 22:01:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][550/1251] eta 0:03:24 lr 0.000835 time 0.2746 (0.2917) loss 4.1712 (3.7672) grad_norm 1.5192 (inf) [2021-04-15 22:01:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][560/1251] eta 0:03:21 lr 0.000834 time 0.2846 (0.2916) loss 2.7192 (3.7663) grad_norm 1.0779 (inf) [2021-04-15 22:01:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][570/1251] eta 0:03:18 lr 0.000834 time 0.2870 (0.2916) loss 3.3806 (3.7641) grad_norm 1.3509 (inf) [2021-04-15 22:01:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][580/1251] eta 0:03:15 lr 0.000834 time 0.2702 (0.2914) loss 4.4932 (3.7637) grad_norm 1.4515 (inf) [2021-04-15 22:01:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][590/1251] eta 0:03:12 lr 0.000834 time 0.2826 (0.2914) loss 3.0090 (3.7694) grad_norm 1.1421 (inf) [2021-04-15 22:01:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][600/1251] eta 0:03:09 lr 0.000834 time 0.2974 (0.2912) loss 4.2148 (3.7692) grad_norm 1.1557 (inf) [2021-04-15 22:02:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][610/1251] eta 0:03:06 lr 0.000834 time 0.2643 (0.2910) loss 2.8225 (3.7645) grad_norm 1.1326 (inf) [2021-04-15 22:02:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][620/1251] eta 0:03:03 lr 0.000834 time 0.3015 (0.2909) loss 4.3867 (3.7683) grad_norm 1.2435 (inf) [2021-04-15 22:02:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][630/1251] eta 0:03:00 lr 0.000834 time 0.2714 (0.2907) loss 2.9359 (3.7643) grad_norm 1.3243 (inf) [2021-04-15 22:02:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][640/1251] eta 0:02:57 lr 0.000834 time 0.2816 (0.2906) loss 3.9111 (3.7667) grad_norm 1.3017 (inf) [2021-04-15 22:02:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][650/1251] eta 0:02:54 lr 0.000834 time 0.2583 (0.2904) loss 4.0888 (3.7664) grad_norm 1.0968 (inf) [2021-04-15 22:02:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][660/1251] eta 0:02:51 lr 0.000834 time 0.2776 (0.2902) loss 4.3213 (3.7696) grad_norm 1.5285 (inf) [2021-04-15 22:02:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][670/1251] eta 0:02:48 lr 0.000834 time 0.2611 (0.2900) loss 4.8130 (3.7716) grad_norm 1.0192 (inf) [2021-04-15 22:02:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][680/1251] eta 0:02:45 lr 0.000834 time 0.2748 (0.2899) loss 3.3704 (3.7725) grad_norm 1.1214 (inf) [2021-04-15 22:02:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][690/1251] eta 0:02:42 lr 0.000834 time 0.2659 (0.2897) loss 4.3419 (3.7752) grad_norm 1.1889 (inf) [2021-04-15 22:02:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][700/1251] eta 0:02:39 lr 0.000834 time 0.2974 (0.2896) loss 4.3449 (3.7749) grad_norm 1.0999 (inf) [2021-04-15 22:02:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][710/1251] eta 0:02:36 lr 0.000834 time 0.2869 (0.2894) loss 2.7590 (3.7777) grad_norm 1.6071 (inf) [2021-04-15 22:02:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][720/1251] eta 0:02:33 lr 0.000834 time 0.2577 (0.2893) loss 3.9048 (3.7791) grad_norm 1.0722 (inf) [2021-04-15 22:02:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][730/1251] eta 0:02:30 lr 0.000834 time 0.2848 (0.2893) loss 4.0128 (3.7801) grad_norm 1.0206 (inf) [2021-04-15 22:02:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][740/1251] eta 0:02:27 lr 0.000834 time 0.2809 (0.2892) loss 3.8463 (3.7819) grad_norm 1.1897 (inf) [2021-04-15 22:02:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][750/1251] eta 0:02:24 lr 0.000834 time 0.2785 (0.2892) loss 3.0163 (3.7840) grad_norm 1.1141 (inf) [2021-04-15 22:02:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][760/1251] eta 0:02:22 lr 0.000834 time 0.2852 (0.2893) loss 3.5054 (3.7874) grad_norm 0.9866 (inf) [2021-04-15 22:02:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][770/1251] eta 0:02:19 lr 0.000834 time 0.3015 (0.2892) loss 4.2971 (3.7904) grad_norm 1.1088 (inf) [2021-04-15 22:02:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][780/1251] eta 0:02:16 lr 0.000834 time 0.2829 (0.2890) loss 4.6914 (3.7905) grad_norm 1.1585 (inf) [2021-04-15 22:02:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][790/1251] eta 0:02:13 lr 0.000834 time 0.2739 (0.2891) loss 3.7557 (3.7916) grad_norm 1.1580 (inf) [2021-04-15 22:02:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][800/1251] eta 0:02:10 lr 0.000834 time 0.3997 (0.2892) loss 3.4564 (3.7895) grad_norm 1.5327 (inf) [2021-04-15 22:02:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][810/1251] eta 0:02:07 lr 0.000834 time 0.2575 (0.2890) loss 3.9952 (3.7936) grad_norm 1.1083 (inf) [2021-04-15 22:02:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][820/1251] eta 0:02:04 lr 0.000834 time 0.2840 (0.2888) loss 2.4520 (3.7895) grad_norm 1.0037 (inf) [2021-04-15 22:03:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][830/1251] eta 0:02:01 lr 0.000834 time 0.2744 (0.2887) loss 3.8016 (3.7914) grad_norm 1.2200 (inf) [2021-04-15 22:03:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][840/1251] eta 0:01:58 lr 0.000834 time 0.2900 (0.2886) loss 3.0144 (3.7881) grad_norm 1.2359 (inf) [2021-04-15 22:03:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][850/1251] eta 0:01:55 lr 0.000834 time 0.2601 (0.2884) loss 3.8869 (3.7873) grad_norm 1.1733 (inf) [2021-04-15 22:03:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][860/1251] eta 0:01:52 lr 0.000834 time 0.2562 (0.2883) loss 3.5450 (3.7900) grad_norm 1.0528 (inf) [2021-04-15 22:03:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][870/1251] eta 0:01:49 lr 0.000834 time 0.2440 (0.2881) loss 3.7042 (3.7873) grad_norm 1.1891 (inf) [2021-04-15 22:03:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][880/1251] eta 0:01:46 lr 0.000834 time 0.2790 (0.2880) loss 4.8735 (3.7887) grad_norm 1.1554 (inf) [2021-04-15 22:03:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][890/1251] eta 0:01:43 lr 0.000833 time 0.2773 (0.2881) loss 3.5342 (3.7913) grad_norm 1.1794 (inf) [2021-04-15 22:03:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][900/1251] eta 0:01:41 lr 0.000833 time 0.2885 (0.2879) loss 3.3239 (3.7905) grad_norm 1.3468 (inf) [2021-04-15 22:03:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][910/1251] eta 0:01:38 lr 0.000833 time 0.2676 (0.2878) loss 4.3908 (3.7901) grad_norm 1.0404 (inf) [2021-04-15 22:03:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][920/1251] eta 0:01:35 lr 0.000833 time 0.2712 (0.2878) loss 3.0431 (3.7916) grad_norm 1.2088 (inf) [2021-04-15 22:03:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][930/1251] eta 0:01:32 lr 0.000833 time 0.2735 (0.2877) loss 3.6915 (3.7894) grad_norm 0.9947 (inf) [2021-04-15 22:03:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][940/1251] eta 0:01:29 lr 0.000833 time 0.2645 (0.2877) loss 3.5990 (3.7883) grad_norm 1.2145 (inf) [2021-04-15 22:03:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][950/1251] eta 0:01:26 lr 0.000833 time 0.2809 (0.2876) loss 4.3266 (3.7895) grad_norm 1.1509 (inf) [2021-04-15 22:03:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][960/1251] eta 0:01:23 lr 0.000833 time 0.2595 (0.2875) loss 2.7485 (3.7877) grad_norm 1.2064 (inf) [2021-04-15 22:03:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][970/1251] eta 0:01:20 lr 0.000833 time 0.2696 (0.2873) loss 3.1480 (3.7874) grad_norm 1.1800 (inf) [2021-04-15 22:03:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][980/1251] eta 0:01:17 lr 0.000833 time 0.2636 (0.2872) loss 4.1067 (3.7893) grad_norm 1.3666 (inf) [2021-04-15 22:03:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][990/1251] eta 0:01:14 lr 0.000833 time 0.2836 (0.2872) loss 3.4504 (3.7897) grad_norm 1.1505 (inf) [2021-04-15 22:03:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][1000/1251] eta 0:01:12 lr 0.000833 time 0.2877 (0.2871) loss 2.6472 (3.7903) grad_norm 1.1641 (inf) [2021-04-15 22:03:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][1010/1251] eta 0:01:09 lr 0.000833 time 0.2903 (0.2870) loss 2.4611 (3.7876) grad_norm 1.1284 (inf) [2021-04-15 22:03:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][1020/1251] eta 0:01:06 lr 0.000833 time 0.2840 (0.2869) loss 4.4412 (3.7882) grad_norm 1.1767 (inf) [2021-04-15 22:03:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][1030/1251] eta 0:01:03 lr 0.000833 time 0.2869 (0.2869) loss 4.4982 (3.7876) grad_norm 1.0615 (inf) [2021-04-15 22:04:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][1040/1251] eta 0:01:00 lr 0.000833 time 0.2678 (0.2868) loss 4.9814 (3.7877) grad_norm 1.0963 (inf) [2021-04-15 22:04:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][1050/1251] eta 0:00:57 lr 0.000833 time 0.2664 (0.2867) loss 4.1067 (3.7863) grad_norm 1.0086 (inf) [2021-04-15 22:04:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][1060/1251] eta 0:00:54 lr 0.000833 time 0.2544 (0.2866) loss 3.3122 (3.7865) grad_norm 1.4854 (inf) [2021-04-15 22:04:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][1070/1251] eta 0:00:51 lr 0.000833 time 0.2756 (0.2867) loss 3.8027 (3.7861) grad_norm 1.4321 (inf) [2021-04-15 22:04:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][1080/1251] eta 0:00:49 lr 0.000833 time 0.2724 (0.2866) loss 2.8744 (3.7837) grad_norm 1.0995 (inf) [2021-04-15 22:04:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][1090/1251] eta 0:00:46 lr 0.000833 time 0.2612 (0.2865) loss 4.0314 (3.7836) grad_norm 1.1131 (inf) [2021-04-15 22:04:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][1100/1251] eta 0:00:43 lr 0.000833 time 0.2793 (0.2864) loss 3.6729 (3.7855) grad_norm 1.1591 (inf) [2021-04-15 22:04:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][1110/1251] eta 0:00:40 lr 0.000833 time 0.2462 (0.2864) loss 3.1375 (3.7857) grad_norm 1.0213 (inf) [2021-04-15 22:04:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][1120/1251] eta 0:00:37 lr 0.000833 time 0.2753 (0.2863) loss 4.1779 (3.7884) grad_norm 1.2578 (inf) [2021-04-15 22:04:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][1130/1251] eta 0:00:34 lr 0.000833 time 0.2595 (0.2862) loss 4.4930 (3.7887) grad_norm 1.1646 (inf) [2021-04-15 22:04:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][1140/1251] eta 0:00:31 lr 0.000833 time 0.2642 (0.2863) loss 3.3254 (3.7896) grad_norm 1.2383 (inf) [2021-04-15 22:04:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][1150/1251] eta 0:00:28 lr 0.000833 time 0.2531 (0.2862) loss 2.8025 (3.7889) grad_norm 1.1822 (inf) [2021-04-15 22:04:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][1160/1251] eta 0:00:26 lr 0.000833 time 0.2702 (0.2862) loss 3.5089 (3.7900) grad_norm 2.0180 (inf) [2021-04-15 22:04:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][1170/1251] eta 0:00:23 lr 0.000833 time 0.2525 (0.2862) loss 3.5668 (3.7886) grad_norm 1.1835 (inf) [2021-04-15 22:04:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][1180/1251] eta 0:00:20 lr 0.000833 time 0.2884 (0.2861) loss 3.6624 (3.7900) grad_norm 1.1149 (inf) [2021-04-15 22:04:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][1190/1251] eta 0:00:17 lr 0.000833 time 0.3140 (0.2860) loss 4.5199 (3.7912) grad_norm 1.3451 (inf) [2021-04-15 22:04:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][1200/1251] eta 0:00:14 lr 0.000833 time 0.2690 (0.2860) loss 4.7264 (3.7938) grad_norm 1.2151 (inf) [2021-04-15 22:04:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][1210/1251] eta 0:00:11 lr 0.000832 time 0.2926 (0.2859) loss 5.0557 (3.7921) grad_norm 1.0754 (inf) [2021-04-15 22:04:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][1220/1251] eta 0:00:08 lr 0.000832 time 0.2697 (0.2858) loss 3.8989 (3.7925) grad_norm 1.3492 (inf) [2021-04-15 22:04:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][1230/1251] eta 0:00:06 lr 0.000832 time 0.3081 (0.2858) loss 3.9167 (3.7931) grad_norm 1.1981 (inf) [2021-04-15 22:04:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][1240/1251] eta 0:00:03 lr 0.000832 time 0.2481 (0.2856) loss 3.3781 (3.7936) grad_norm 1.3329 (inf) [2021-04-15 22:04:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [80/300][1250/1251] eta 0:00:00 lr 0.000832 time 0.2561 (0.2854) loss 4.0198 (3.7934) grad_norm 1.1046 (inf) [2021-04-15 22:05:05 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 80 training takes 0:06:02 [2021-04-15 22:05:05 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_80.pth saving...... [2021-04-15 22:05:14 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_80.pth saved !!! [2021-04-15 22:05:16 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.152 (1.152) Loss 1.2344 (1.2344) Acc@1 72.266 (72.266) Acc@5 90.039 (90.039) [2021-04-15 22:05:17 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.124 (0.280) Loss 1.2159 (1.1915) Acc@1 73.926 (72.275) Acc@5 90.527 (91.415) [2021-04-15 22:05:19 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.141 (0.228) Loss 1.1940 (1.1900) Acc@1 72.559 (72.247) Acc@5 91.211 (91.485) [2021-04-15 22:05:21 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.104 (0.217) Loss 1.1651 (1.1763) Acc@1 72.363 (72.502) Acc@5 90.918 (91.602) [2021-04-15 22:05:23 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.221) Loss 1.1578 (1.1725) Acc@1 70.508 (72.597) Acc@5 92.188 (91.587) [2021-04-15 22:05:29 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 72.544 Acc@5 91.462 [2021-04-15 22:05:29 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 72.5% [2021-04-15 22:05:29 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 72.54% [2021-04-15 22:05:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][0/1251] eta 0:57:21 lr 0.000832 time 2.7513 (2.7513) loss 3.2683 (3.2683) grad_norm 1.0126 (1.0126) [2021-04-15 22:05:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][10/1251] eta 0:10:35 lr 0.000832 time 0.4454 (0.5124) loss 3.3207 (3.9526) grad_norm 1.5236 (1.2026) [2021-04-15 22:05:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][20/1251] eta 0:08:11 lr 0.000832 time 0.2816 (0.3994) loss 3.2989 (3.8386) grad_norm 1.1087 (1.1814) [2021-04-15 22:05:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][30/1251] eta 0:07:21 lr 0.000832 time 0.2847 (0.3612) loss 3.9771 (3.8553) grad_norm 1.3132 (1.1801) [2021-04-15 22:05:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][40/1251] eta 0:06:53 lr 0.000832 time 0.2481 (0.3418) loss 4.1192 (3.8646) grad_norm 1.2432 (1.1717) [2021-04-15 22:05:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][50/1251] eta 0:06:37 lr 0.000832 time 0.2847 (0.3311) loss 4.7555 (3.9401) grad_norm 1.4103 (1.1818) [2021-04-15 22:05:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][60/1251] eta 0:06:25 lr 0.000832 time 0.3055 (0.3238) loss 3.9749 (3.8975) grad_norm 1.1294 (1.1770) [2021-04-15 22:05:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][70/1251] eta 0:06:15 lr 0.000832 time 0.2813 (0.3176) loss 4.3852 (3.8783) grad_norm 1.1315 (1.1730) [2021-04-15 22:05:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][80/1251] eta 0:06:07 lr 0.000832 time 0.2962 (0.3135) loss 4.2619 (3.8659) grad_norm 1.4700 (1.1723) [2021-04-15 22:05:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][90/1251] eta 0:06:00 lr 0.000832 time 0.3232 (0.3105) loss 2.7614 (3.8535) grad_norm 1.4455 (1.1717) [2021-04-15 22:06:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][100/1251] eta 0:05:55 lr 0.000832 time 0.2513 (0.3088) loss 3.6275 (3.8240) grad_norm 1.4308 (1.1749) [2021-04-15 22:06:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][110/1251] eta 0:05:49 lr 0.000832 time 0.2849 (0.3063) loss 4.3815 (3.8295) grad_norm 1.1246 (1.1764) [2021-04-15 22:06:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][120/1251] eta 0:05:45 lr 0.000832 time 0.2728 (0.3053) loss 3.1743 (3.8210) grad_norm 1.0625 (1.1734) [2021-04-15 22:06:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][130/1251] eta 0:05:40 lr 0.000832 time 0.2760 (0.3038) loss 4.3790 (3.8327) grad_norm 1.1750 (1.1720) [2021-04-15 22:06:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][140/1251] eta 0:05:36 lr 0.000832 time 0.3237 (0.3029) loss 4.1465 (3.8316) grad_norm 1.1163 (1.1736) [2021-04-15 22:06:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][150/1251] eta 0:05:32 lr 0.000832 time 0.3034 (0.3017) loss 4.1469 (3.8312) grad_norm 1.2771 (1.1754) [2021-04-15 22:06:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][160/1251] eta 0:05:27 lr 0.000832 time 0.2815 (0.3003) loss 2.5394 (3.8219) grad_norm 1.1707 (1.1781) [2021-04-15 22:06:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][170/1251] eta 0:05:23 lr 0.000832 time 0.2752 (0.2993) loss 2.9377 (3.8132) grad_norm 1.0883 (1.1734) [2021-04-15 22:06:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][180/1251] eta 0:05:19 lr 0.000832 time 0.2996 (0.2985) loss 2.5862 (3.7962) grad_norm 1.4328 (1.1747) [2021-04-15 22:06:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][190/1251] eta 0:05:15 lr 0.000832 time 0.3066 (0.2975) loss 3.6187 (3.8020) grad_norm 1.4268 (1.1827) [2021-04-15 22:06:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][200/1251] eta 0:05:11 lr 0.000832 time 0.2820 (0.2967) loss 3.4644 (3.8143) grad_norm 1.6397 (1.1859) [2021-04-15 22:06:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][210/1251] eta 0:05:08 lr 0.000832 time 0.2948 (0.2961) loss 3.6962 (3.8080) grad_norm 1.3615 (1.1862) [2021-04-15 22:06:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][220/1251] eta 0:05:04 lr 0.000832 time 0.2858 (0.2957) loss 2.7118 (3.7996) grad_norm 0.9780 (1.1848) [2021-04-15 22:06:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][230/1251] eta 0:05:01 lr 0.000832 time 0.2890 (0.2951) loss 2.7409 (3.7968) grad_norm 1.2270 (1.1864) [2021-04-15 22:06:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][240/1251] eta 0:04:58 lr 0.000832 time 0.2601 (0.2953) loss 4.2428 (3.7978) grad_norm 1.0026 (1.1897) [2021-04-15 22:06:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][250/1251] eta 0:04:55 lr 0.000832 time 0.2749 (0.2949) loss 2.8650 (3.7831) grad_norm 1.1856 (1.1901) [2021-04-15 22:06:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][260/1251] eta 0:04:51 lr 0.000832 time 0.2906 (0.2942) loss 3.6379 (3.7820) grad_norm 1.2019 (1.1885) [2021-04-15 22:06:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][270/1251] eta 0:04:47 lr 0.000832 time 0.2781 (0.2936) loss 3.3727 (3.7889) grad_norm 1.1583 (1.1905) [2021-04-15 22:06:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][280/1251] eta 0:04:44 lr 0.000831 time 0.2985 (0.2931) loss 4.1004 (3.7947) grad_norm 1.1133 (1.1889) [2021-04-15 22:06:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][290/1251] eta 0:04:41 lr 0.000831 time 0.2645 (0.2931) loss 3.4129 (3.8023) grad_norm 1.3088 (1.1921) [2021-04-15 22:06:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][300/1251] eta 0:04:38 lr 0.000831 time 0.2425 (0.2926) loss 3.9651 (3.8020) grad_norm 1.0165 (1.1920) [2021-04-15 22:07:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][310/1251] eta 0:04:35 lr 0.000831 time 0.2800 (0.2923) loss 4.0396 (3.7980) grad_norm 1.3139 (1.1947) [2021-04-15 22:07:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][320/1251] eta 0:04:31 lr 0.000831 time 0.2979 (0.2920) loss 3.5548 (3.7978) grad_norm 1.1200 (1.2000) [2021-04-15 22:07:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][330/1251] eta 0:04:28 lr 0.000831 time 0.2806 (0.2919) loss 4.5294 (3.7999) grad_norm 1.2064 (1.2005) [2021-04-15 22:07:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][340/1251] eta 0:04:25 lr 0.000831 time 0.2700 (0.2919) loss 4.0104 (3.7997) grad_norm 1.2788 (1.2003) [2021-04-15 22:07:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][350/1251] eta 0:04:22 lr 0.000831 time 0.2843 (0.2916) loss 3.1004 (3.7906) grad_norm 1.1265 (1.2004) [2021-04-15 22:07:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][360/1251] eta 0:04:19 lr 0.000831 time 0.3054 (0.2913) loss 4.7941 (3.7992) grad_norm 1.1707 (1.1987) [2021-04-15 22:07:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][370/1251] eta 0:04:16 lr 0.000831 time 0.2815 (0.2913) loss 4.1174 (3.8032) grad_norm 1.4200 (1.1978) [2021-04-15 22:07:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][380/1251] eta 0:04:13 lr 0.000831 time 0.2625 (0.2912) loss 3.7102 (3.8034) grad_norm 1.2662 (1.1992) [2021-04-15 22:07:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][390/1251] eta 0:04:10 lr 0.000831 time 0.2450 (0.2908) loss 4.3097 (3.7993) grad_norm 1.1588 (1.1991) [2021-04-15 22:07:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][400/1251] eta 0:04:07 lr 0.000831 time 0.2834 (0.2905) loss 4.0777 (3.8008) grad_norm 1.1955 (1.2017) [2021-04-15 22:07:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][410/1251] eta 0:04:04 lr 0.000831 time 0.2992 (0.2902) loss 3.5461 (3.7943) grad_norm 1.2676 (1.2029) [2021-04-15 22:07:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][420/1251] eta 0:04:00 lr 0.000831 time 0.2822 (0.2899) loss 3.2371 (3.7932) grad_norm 1.0061 (1.2024) [2021-04-15 22:07:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][430/1251] eta 0:03:57 lr 0.000831 time 0.2802 (0.2896) loss 3.4755 (3.7922) grad_norm 1.1601 (1.2009) [2021-04-15 22:07:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][440/1251] eta 0:03:54 lr 0.000831 time 0.2849 (0.2894) loss 2.7732 (3.7915) grad_norm 1.4489 (1.2011) [2021-04-15 22:07:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][450/1251] eta 0:03:51 lr 0.000831 time 0.2727 (0.2892) loss 3.1528 (3.7893) grad_norm 1.1735 (1.1995) [2021-04-15 22:07:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][460/1251] eta 0:03:48 lr 0.000831 time 0.2807 (0.2890) loss 4.3308 (3.7877) grad_norm 1.2488 (1.1978) [2021-04-15 22:07:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][470/1251] eta 0:03:45 lr 0.000831 time 0.2806 (0.2888) loss 3.1851 (3.7850) grad_norm 1.1709 (1.1970) [2021-04-15 22:07:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][480/1251] eta 0:03:42 lr 0.000831 time 0.2764 (0.2886) loss 3.8892 (3.7878) grad_norm 1.0782 (1.1955) [2021-04-15 22:07:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][490/1251] eta 0:03:39 lr 0.000831 time 0.2892 (0.2884) loss 3.8820 (3.7802) grad_norm 1.1570 (1.1961) [2021-04-15 22:07:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][500/1251] eta 0:03:36 lr 0.000831 time 0.2809 (0.2881) loss 4.1004 (3.7780) grad_norm 1.2923 (1.1964) [2021-04-15 22:07:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][510/1251] eta 0:03:33 lr 0.000831 time 0.2603 (0.2879) loss 3.6723 (3.7810) grad_norm 1.4240 (1.1963) [2021-04-15 22:07:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][520/1251] eta 0:03:30 lr 0.000831 time 0.2883 (0.2878) loss 3.4131 (3.7839) grad_norm 1.1231 (1.1968) [2021-04-15 22:08:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][530/1251] eta 0:03:27 lr 0.000831 time 0.2437 (0.2876) loss 3.4207 (3.7841) grad_norm 1.1249 (1.1961) [2021-04-15 22:08:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][540/1251] eta 0:03:24 lr 0.000831 time 0.2967 (0.2875) loss 2.8091 (3.7797) grad_norm 1.1496 (1.1955) [2021-04-15 22:08:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][550/1251] eta 0:03:21 lr 0.000831 time 0.2707 (0.2874) loss 4.5739 (3.7858) grad_norm 1.1553 (1.1949) [2021-04-15 22:08:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][560/1251] eta 0:03:18 lr 0.000831 time 0.2812 (0.2874) loss 4.0001 (3.7859) grad_norm 1.2146 (1.1932) [2021-04-15 22:08:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][570/1251] eta 0:03:15 lr 0.000831 time 0.2702 (0.2873) loss 2.9543 (3.7868) grad_norm 1.1456 (1.1925) [2021-04-15 22:08:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][580/1251] eta 0:03:12 lr 0.000831 time 0.2575 (0.2875) loss 3.6789 (3.7862) grad_norm 1.1388 (1.1924) [2021-04-15 22:08:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][590/1251] eta 0:03:10 lr 0.000831 time 0.2884 (0.2877) loss 3.1428 (3.7871) grad_norm 1.1388 (1.1929) [2021-04-15 22:08:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][600/1251] eta 0:03:07 lr 0.000830 time 0.2954 (0.2875) loss 3.6426 (3.7890) grad_norm 1.2608 (1.1944) [2021-04-15 22:08:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][610/1251] eta 0:03:04 lr 0.000830 time 0.2742 (0.2875) loss 2.9009 (3.7886) grad_norm 1.0655 (1.1937) [2021-04-15 22:08:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][620/1251] eta 0:03:01 lr 0.000830 time 0.3006 (0.2873) loss 3.8619 (3.7867) grad_norm 1.1795 (1.1942) [2021-04-15 22:08:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][630/1251] eta 0:02:58 lr 0.000830 time 0.3152 (0.2873) loss 2.6929 (3.7879) grad_norm 1.0750 (1.1943) [2021-04-15 22:08:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][640/1251] eta 0:02:55 lr 0.000830 time 0.2764 (0.2871) loss 4.2366 (3.7918) grad_norm 1.0849 (1.1942) [2021-04-15 22:08:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][650/1251] eta 0:02:52 lr 0.000830 time 0.2494 (0.2870) loss 4.2433 (3.7946) grad_norm 1.1794 (1.1927) [2021-04-15 22:08:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][660/1251] eta 0:02:49 lr 0.000830 time 0.2587 (0.2870) loss 3.0384 (3.7975) grad_norm 1.0341 (1.1933) [2021-04-15 22:08:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][670/1251] eta 0:02:46 lr 0.000830 time 0.2828 (0.2869) loss 2.8776 (3.7984) grad_norm 1.2113 (1.1933) [2021-04-15 22:08:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][680/1251] eta 0:02:43 lr 0.000830 time 0.3045 (0.2870) loss 2.5794 (3.7952) grad_norm 1.5507 (1.1928) [2021-04-15 22:08:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][690/1251] eta 0:02:41 lr 0.000830 time 0.2699 (0.2871) loss 4.0328 (3.7958) grad_norm 1.1406 (1.1931) [2021-04-15 22:08:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][700/1251] eta 0:02:38 lr 0.000830 time 0.2851 (0.2869) loss 3.6409 (3.7992) grad_norm 1.1142 (1.1936) [2021-04-15 22:08:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][710/1251] eta 0:02:35 lr 0.000830 time 0.2796 (0.2868) loss 3.7369 (3.8001) grad_norm 1.1314 (1.1943) [2021-04-15 22:08:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][720/1251] eta 0:02:32 lr 0.000830 time 0.2748 (0.2867) loss 4.3334 (3.8038) grad_norm 1.6677 (1.1970) [2021-04-15 22:08:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][730/1251] eta 0:02:29 lr 0.000830 time 0.2903 (0.2866) loss 3.2700 (3.8001) grad_norm 1.5941 (1.1986) [2021-04-15 22:09:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][740/1251] eta 0:02:26 lr 0.000830 time 0.2802 (0.2865) loss 3.8056 (3.7979) grad_norm 1.2171 (1.1987) [2021-04-15 22:09:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][750/1251] eta 0:02:23 lr 0.000830 time 0.2598 (0.2864) loss 3.4327 (3.8004) grad_norm 1.0836 (1.1993) [2021-04-15 22:09:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][760/1251] eta 0:02:20 lr 0.000830 time 0.2915 (0.2864) loss 3.4604 (3.7968) grad_norm 1.2333 (1.1991) [2021-04-15 22:09:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][770/1251] eta 0:02:17 lr 0.000830 time 0.2849 (0.2863) loss 4.6741 (3.7969) grad_norm 1.4757 (1.1991) [2021-04-15 22:09:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][780/1251] eta 0:02:14 lr 0.000830 time 0.2797 (0.2862) loss 3.5394 (3.7973) grad_norm 1.1377 (1.1998) [2021-04-15 22:09:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][790/1251] eta 0:02:11 lr 0.000830 time 0.3035 (0.2861) loss 3.4718 (3.7957) grad_norm 1.3653 (1.2007) [2021-04-15 22:09:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][800/1251] eta 0:02:09 lr 0.000830 time 0.2858 (0.2861) loss 3.6571 (3.7925) grad_norm 1.0548 (1.2003) [2021-04-15 22:09:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][810/1251] eta 0:02:06 lr 0.000830 time 0.2790 (0.2860) loss 3.9174 (3.7919) grad_norm 1.1022 (1.2006) [2021-04-15 22:09:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][820/1251] eta 0:02:03 lr 0.000830 time 0.2606 (0.2861) loss 3.9123 (3.7901) grad_norm 1.1827 (1.2006) [2021-04-15 22:09:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][830/1251] eta 0:02:00 lr 0.000830 time 0.2903 (0.2860) loss 3.3168 (3.7915) grad_norm 1.4179 (1.2016) [2021-04-15 22:09:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][840/1251] eta 0:01:57 lr 0.000830 time 0.2790 (0.2859) loss 4.1928 (3.7924) grad_norm 1.1345 (1.2020) [2021-04-15 22:09:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][850/1251] eta 0:01:54 lr 0.000830 time 0.3001 (0.2858) loss 4.0769 (3.7930) grad_norm 1.0389 (1.2014) [2021-04-15 22:09:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][860/1251] eta 0:01:51 lr 0.000830 time 0.2789 (0.2857) loss 3.0558 (3.7933) grad_norm 1.2620 (1.2008) [2021-04-15 22:09:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][870/1251] eta 0:01:48 lr 0.000830 time 0.2836 (0.2857) loss 3.4612 (3.7923) grad_norm 1.8394 (1.2016) [2021-04-15 22:09:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][880/1251] eta 0:01:45 lr 0.000830 time 0.2729 (0.2856) loss 4.3812 (3.7920) grad_norm 1.1963 (1.2026) [2021-04-15 22:09:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][890/1251] eta 0:01:43 lr 0.000830 time 0.2858 (0.2856) loss 4.5270 (3.7937) grad_norm 1.1650 (1.2022) [2021-04-15 22:09:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][900/1251] eta 0:01:40 lr 0.000830 time 0.2865 (0.2856) loss 3.2565 (3.7971) grad_norm 1.2263 (1.2016) [2021-04-15 22:09:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][910/1251] eta 0:01:37 lr 0.000830 time 0.2540 (0.2855) loss 4.0339 (3.7986) grad_norm 1.2962 (1.2016) [2021-04-15 22:09:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][920/1251] eta 0:01:34 lr 0.000829 time 0.2724 (0.2856) loss 4.3305 (3.8003) grad_norm 1.1993 (1.2006) [2021-04-15 22:09:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][930/1251] eta 0:01:31 lr 0.000829 time 0.2862 (0.2855) loss 4.2216 (3.7975) grad_norm 1.0484 (1.2004) [2021-04-15 22:09:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][940/1251] eta 0:01:28 lr 0.000829 time 0.2886 (0.2855) loss 2.9859 (3.7981) grad_norm 1.1197 (1.2008) [2021-04-15 22:10:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][950/1251] eta 0:01:25 lr 0.000829 time 0.2609 (0.2854) loss 3.0591 (3.7959) grad_norm 1.1520 (1.2004) [2021-04-15 22:10:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][960/1251] eta 0:01:23 lr 0.000829 time 0.2871 (0.2854) loss 3.6656 (3.7948) grad_norm 1.0841 (1.2001) [2021-04-15 22:10:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][970/1251] eta 0:01:20 lr 0.000829 time 0.2586 (0.2854) loss 3.9915 (3.7961) grad_norm 1.0795 (1.1991) [2021-04-15 22:10:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][980/1251] eta 0:01:17 lr 0.000829 time 0.2679 (0.2853) loss 4.5899 (3.7942) grad_norm 1.0046 (1.1992) [2021-04-15 22:10:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][990/1251] eta 0:01:14 lr 0.000829 time 0.2908 (0.2852) loss 2.7492 (3.7921) grad_norm 1.1324 (1.1991) [2021-04-15 22:10:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][1000/1251] eta 0:01:11 lr 0.000829 time 0.2642 (0.2852) loss 2.8110 (3.7946) grad_norm 1.2899 (1.1994) [2021-04-15 22:10:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][1010/1251] eta 0:01:08 lr 0.000829 time 0.2779 (0.2851) loss 4.6714 (3.7973) grad_norm 1.0739 (1.1993) [2021-04-15 22:10:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][1020/1251] eta 0:01:05 lr 0.000829 time 0.2728 (0.2851) loss 3.4793 (3.7972) grad_norm 1.3319 (1.1996) [2021-04-15 22:10:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][1030/1251] eta 0:01:03 lr 0.000829 time 0.2848 (0.2851) loss 4.3876 (3.8010) grad_norm 1.0942 (1.1998) [2021-04-15 22:10:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][1040/1251] eta 0:01:00 lr 0.000829 time 0.2602 (0.2850) loss 4.2901 (3.7988) grad_norm 1.2141 (1.1995) [2021-04-15 22:10:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][1050/1251] eta 0:00:57 lr 0.000829 time 0.2876 (0.2850) loss 3.6140 (3.7990) grad_norm 1.3965 (1.2001) [2021-04-15 22:10:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][1060/1251] eta 0:00:54 lr 0.000829 time 0.2842 (0.2849) loss 3.9674 (3.8009) grad_norm 1.1317 (1.2012) [2021-04-15 22:10:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][1070/1251] eta 0:00:51 lr 0.000829 time 0.2856 (0.2848) loss 4.0019 (3.8001) grad_norm 1.0834 (1.2012) [2021-04-15 22:10:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][1080/1251] eta 0:00:48 lr 0.000829 time 0.2567 (0.2848) loss 3.9335 (3.7993) grad_norm 1.3180 (1.2012) [2021-04-15 22:10:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][1090/1251] eta 0:00:45 lr 0.000829 time 0.2435 (0.2847) loss 4.2478 (3.7989) grad_norm inf (inf) [2021-04-15 22:10:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][1100/1251] eta 0:00:42 lr 0.000829 time 0.2860 (0.2846) loss 4.1465 (3.7996) grad_norm 1.2692 (inf) [2021-04-15 22:10:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][1110/1251] eta 0:00:40 lr 0.000829 time 0.3032 (0.2846) loss 4.2189 (3.8013) grad_norm 1.2197 (inf) [2021-04-15 22:10:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][1120/1251] eta 0:00:37 lr 0.000829 time 0.2776 (0.2845) loss 4.0486 (3.8013) grad_norm 1.0792 (inf) [2021-04-15 22:10:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][1130/1251] eta 0:00:34 lr 0.000829 time 0.2866 (0.2845) loss 3.9777 (3.8011) grad_norm 1.0965 (inf) [2021-04-15 22:10:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][1140/1251] eta 0:00:31 lr 0.000829 time 0.2793 (0.2844) loss 2.5638 (3.7978) grad_norm 1.1758 (inf) [2021-04-15 22:10:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][1150/1251] eta 0:00:28 lr 0.000829 time 0.2795 (0.2845) loss 4.1065 (3.7965) grad_norm 0.9927 (inf) [2021-04-15 22:11:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][1160/1251] eta 0:00:25 lr 0.000829 time 0.3197 (0.2846) loss 3.0275 (3.7950) grad_norm 1.0800 (inf) [2021-04-15 22:11:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][1170/1251] eta 0:00:23 lr 0.000829 time 0.2663 (0.2847) loss 4.1751 (3.7945) grad_norm 1.0603 (inf) [2021-04-15 22:11:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][1180/1251] eta 0:00:20 lr 0.000829 time 0.3068 (0.2847) loss 4.0757 (3.7957) grad_norm 1.5000 (inf) [2021-04-15 22:11:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][1190/1251] eta 0:00:17 lr 0.000829 time 0.2654 (0.2847) loss 4.7136 (3.7976) grad_norm 1.4708 (inf) [2021-04-15 22:11:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][1200/1251] eta 0:00:14 lr 0.000829 time 0.3079 (0.2847) loss 3.8047 (3.7973) grad_norm 1.1166 (inf) [2021-04-15 22:11:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][1210/1251] eta 0:00:11 lr 0.000829 time 0.2839 (0.2847) loss 4.0111 (3.7978) grad_norm 1.2371 (inf) [2021-04-15 22:11:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][1220/1251] eta 0:00:08 lr 0.000829 time 0.2496 (0.2846) loss 4.2500 (3.7987) grad_norm 1.2114 (inf) [2021-04-15 22:11:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][1230/1251] eta 0:00:05 lr 0.000829 time 0.3012 (0.2845) loss 4.0644 (3.7976) grad_norm 1.1019 (inf) [2021-04-15 22:11:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][1240/1251] eta 0:00:03 lr 0.000828 time 0.2482 (0.2844) loss 2.4496 (3.7981) grad_norm 1.1849 (inf) [2021-04-15 22:11:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [81/300][1250/1251] eta 0:00:00 lr 0.000828 time 0.2477 (0.2841) loss 3.9912 (3.8002) grad_norm 1.4527 (inf) [2021-04-15 22:11:28 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 81 training takes 0:05:58 [2021-04-15 22:11:28 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_81.pth saving...... [2021-04-15 22:11:51 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_81.pth saved !!! [2021-04-15 22:11:52 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.135 (1.135) Loss 1.1212 (1.1212) Acc@1 74.414 (74.414) Acc@5 91.992 (91.992) [2021-04-15 22:11:53 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.172 (0.205) Loss 1.2673 (1.1817) Acc@1 69.922 (72.710) Acc@5 91.016 (91.699) [2021-04-15 22:11:56 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.140 (0.216) Loss 1.0688 (1.1835) Acc@1 75.293 (72.749) Acc@5 92.383 (91.639) [2021-04-15 22:11:58 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.111 (0.233) Loss 1.1507 (1.1796) Acc@1 74.023 (72.782) Acc@5 91.406 (91.661) [2021-04-15 22:12:00 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.441 (0.221) Loss 1.1078 (1.1810) Acc@1 73.730 (72.711) Acc@5 93.262 (91.585) [2021-04-15 22:12:03 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 72.652 Acc@5 91.568 [2021-04-15 22:12:03 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 72.7% [2021-04-15 22:12:03 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 72.65% [2021-04-15 22:12:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][0/1251] eta 1:59:46 lr 0.000828 time 5.7443 (5.7443) loss 3.1737 (3.1737) grad_norm 1.3578 (1.3578) [2021-04-15 22:12:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][10/1251] eta 0:16:03 lr 0.000828 time 0.2944 (0.7766) loss 3.4052 (3.8666) grad_norm 1.3363 (1.1960) [2021-04-15 22:12:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][20/1251] eta 0:11:10 lr 0.000828 time 0.2425 (0.5444) loss 4.0344 (3.7473) grad_norm 1.1764 (1.2036) [2021-04-15 22:12:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][30/1251] eta 0:09:20 lr 0.000828 time 0.2758 (0.4594) loss 4.1146 (3.7913) grad_norm 1.0689 (1.2119) [2021-04-15 22:12:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][40/1251] eta 0:08:21 lr 0.000828 time 0.2754 (0.4145) loss 4.0950 (3.7538) grad_norm 1.1648 (1.2143) [2021-04-15 22:12:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][50/1251] eta 0:07:45 lr 0.000828 time 0.2881 (0.3876) loss 4.8255 (3.7461) grad_norm 1.1552 (1.2343) [2021-04-15 22:12:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][60/1251] eta 0:07:21 lr 0.000828 time 0.3401 (0.3709) loss 3.9040 (3.7816) grad_norm 1.3910 (1.2279) [2021-04-15 22:12:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][70/1251] eta 0:07:01 lr 0.000828 time 0.2714 (0.3569) loss 4.0789 (3.8031) grad_norm 1.1893 (1.2172) [2021-04-15 22:12:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][80/1251] eta 0:06:48 lr 0.000828 time 0.2691 (0.3489) loss 3.4048 (3.7859) grad_norm 1.0677 (1.2073) [2021-04-15 22:12:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][90/1251] eta 0:06:36 lr 0.000828 time 0.2818 (0.3413) loss 3.8099 (3.7566) grad_norm 1.4259 (1.2083) [2021-04-15 22:12:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][100/1251] eta 0:06:25 lr 0.000828 time 0.2808 (0.3350) loss 2.6324 (3.7272) grad_norm 1.1941 (1.2072) [2021-04-15 22:12:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][110/1251] eta 0:06:16 lr 0.000828 time 0.2974 (0.3303) loss 4.3795 (3.7440) grad_norm 1.0924 (1.2059) [2021-04-15 22:12:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][120/1251] eta 0:06:10 lr 0.000828 time 0.2975 (0.3274) loss 3.8284 (3.7439) grad_norm 1.1432 (1.2030) [2021-04-15 22:12:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][130/1251] eta 0:06:03 lr 0.000828 time 0.2882 (0.3239) loss 2.6571 (3.7300) grad_norm 1.1478 (1.1989) [2021-04-15 22:12:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][140/1251] eta 0:05:57 lr 0.000828 time 0.2564 (0.3216) loss 3.9290 (3.7215) grad_norm 0.9853 (1.1947) [2021-04-15 22:12:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][150/1251] eta 0:05:52 lr 0.000828 time 0.4309 (0.3199) loss 3.0033 (3.7255) grad_norm 1.1533 (1.1935) [2021-04-15 22:12:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][160/1251] eta 0:05:46 lr 0.000828 time 0.3046 (0.3176) loss 3.9647 (3.7233) grad_norm 1.4119 (1.1939) [2021-04-15 22:12:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][170/1251] eta 0:05:40 lr 0.000828 time 0.2773 (0.3154) loss 4.0680 (3.7113) grad_norm 1.3963 (1.1941) [2021-04-15 22:13:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][180/1251] eta 0:05:35 lr 0.000828 time 0.2852 (0.3136) loss 3.6735 (3.7234) grad_norm 1.1033 (1.1975) [2021-04-15 22:13:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][190/1251] eta 0:05:30 lr 0.000828 time 0.2886 (0.3118) loss 4.1263 (3.7189) grad_norm 1.0396 (1.1965) [2021-04-15 22:13:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][200/1251] eta 0:05:26 lr 0.000828 time 0.2941 (0.3103) loss 4.5009 (3.7400) grad_norm 1.1274 (1.1993) [2021-04-15 22:13:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][210/1251] eta 0:05:21 lr 0.000828 time 0.2630 (0.3088) loss 2.4034 (3.7334) grad_norm 1.2633 (1.2033) [2021-04-15 22:13:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][220/1251] eta 0:05:17 lr 0.000828 time 0.2813 (0.3075) loss 4.1999 (3.7335) grad_norm 1.3018 (1.2088) [2021-04-15 22:13:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][230/1251] eta 0:05:12 lr 0.000828 time 0.2803 (0.3063) loss 3.6489 (3.7436) grad_norm 1.3756 (1.2081) [2021-04-15 22:13:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][240/1251] eta 0:05:08 lr 0.000828 time 0.2641 (0.3053) loss 3.8194 (3.7379) grad_norm 1.1113 (1.2058) [2021-04-15 22:13:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][250/1251] eta 0:05:04 lr 0.000828 time 0.2896 (0.3041) loss 4.2598 (3.7373) grad_norm 1.4313 (1.2088) [2021-04-15 22:13:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][260/1251] eta 0:05:00 lr 0.000828 time 0.2884 (0.3030) loss 2.9816 (3.7317) grad_norm 1.0960 (1.2106) [2021-04-15 22:13:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][270/1251] eta 0:04:56 lr 0.000828 time 0.2690 (0.3022) loss 4.4822 (3.7403) grad_norm 1.0237 (1.2105) [2021-04-15 22:13:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][280/1251] eta 0:04:52 lr 0.000828 time 0.2804 (0.3017) loss 4.4174 (3.7554) grad_norm 1.3460 (1.2149) [2021-04-15 22:13:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][290/1251] eta 0:04:49 lr 0.000828 time 0.2830 (0.3009) loss 3.5354 (3.7535) grad_norm 1.0719 (1.2177) [2021-04-15 22:13:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][300/1251] eta 0:04:45 lr 0.000828 time 0.3115 (0.3002) loss 4.1876 (3.7485) grad_norm 1.1967 (1.2176) [2021-04-15 22:13:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][310/1251] eta 0:04:41 lr 0.000827 time 0.2739 (0.2996) loss 4.0769 (3.7517) grad_norm 1.1261 (1.2151) [2021-04-15 22:13:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][320/1251] eta 0:04:38 lr 0.000827 time 0.3098 (0.2990) loss 4.0015 (3.7481) grad_norm 0.9894 (1.2133) [2021-04-15 22:13:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][330/1251] eta 0:04:34 lr 0.000827 time 0.2954 (0.2984) loss 3.1924 (3.7484) grad_norm 1.0938 (1.2122) [2021-04-15 22:13:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][340/1251] eta 0:04:31 lr 0.000827 time 0.2901 (0.2980) loss 4.5085 (3.7499) grad_norm 1.2885 (1.2117) [2021-04-15 22:13:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][350/1251] eta 0:04:28 lr 0.000827 time 0.2957 (0.2977) loss 4.2958 (3.7532) grad_norm 1.1735 (1.2101) [2021-04-15 22:13:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][360/1251] eta 0:04:25 lr 0.000827 time 0.2772 (0.2976) loss 4.1308 (3.7541) grad_norm 1.2193 (1.2100) [2021-04-15 22:13:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][370/1251] eta 0:04:22 lr 0.000827 time 0.2889 (0.2976) loss 3.2558 (3.7531) grad_norm 1.1663 (1.2091) [2021-04-15 22:13:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][380/1251] eta 0:04:18 lr 0.000827 time 0.2469 (0.2971) loss 3.3293 (3.7504) grad_norm 1.1508 (1.2088) [2021-04-15 22:13:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][390/1251] eta 0:04:15 lr 0.000827 time 0.2867 (0.2966) loss 3.9567 (3.7546) grad_norm 1.1830 (1.2080) [2021-04-15 22:14:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][400/1251] eta 0:04:12 lr 0.000827 time 0.2871 (0.2962) loss 3.4212 (3.7513) grad_norm 1.3050 (1.2088) [2021-04-15 22:14:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][410/1251] eta 0:04:08 lr 0.000827 time 0.2882 (0.2958) loss 4.1593 (3.7573) grad_norm 1.0929 (1.2090) [2021-04-15 22:14:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][420/1251] eta 0:04:05 lr 0.000827 time 0.3143 (0.2955) loss 4.5902 (3.7620) grad_norm 1.2004 (1.2087) [2021-04-15 22:14:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][430/1251] eta 0:04:02 lr 0.000827 time 0.2816 (0.2951) loss 3.0369 (3.7607) grad_norm 1.0237 (1.2064) [2021-04-15 22:14:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][440/1251] eta 0:03:58 lr 0.000827 time 0.2581 (0.2946) loss 3.8511 (3.7600) grad_norm 1.0243 (1.2054) [2021-04-15 22:14:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][450/1251] eta 0:03:55 lr 0.000827 time 0.3053 (0.2942) loss 4.4700 (3.7605) grad_norm 1.2055 (1.2044) [2021-04-15 22:14:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][460/1251] eta 0:03:52 lr 0.000827 time 0.2436 (0.2938) loss 3.4744 (3.7624) grad_norm 1.1118 (1.2028) [2021-04-15 22:14:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][470/1251] eta 0:03:49 lr 0.000827 time 0.2836 (0.2937) loss 3.7974 (3.7680) grad_norm 1.0279 (1.2011) [2021-04-15 22:14:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][480/1251] eta 0:03:46 lr 0.000827 time 0.2976 (0.2934) loss 4.4650 (3.7687) grad_norm 1.1712 (1.2035) [2021-04-15 22:14:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][490/1251] eta 0:03:43 lr 0.000827 time 0.2800 (0.2935) loss 3.2296 (3.7705) grad_norm 1.3302 (1.2057) [2021-04-15 22:14:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][500/1251] eta 0:03:40 lr 0.000827 time 0.2658 (0.2932) loss 3.9289 (3.7709) grad_norm 1.7043 (1.2059) [2021-04-15 22:14:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][510/1251] eta 0:03:37 lr 0.000827 time 0.2692 (0.2929) loss 4.6991 (3.7758) grad_norm 1.1488 (1.2055) [2021-04-15 22:14:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][520/1251] eta 0:03:33 lr 0.000827 time 0.2910 (0.2926) loss 3.8365 (3.7785) grad_norm 1.0913 (1.2055) [2021-04-15 22:14:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][530/1251] eta 0:03:30 lr 0.000827 time 0.2442 (0.2923) loss 4.3088 (3.7782) grad_norm 1.2394 (1.2038) [2021-04-15 22:14:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][540/1251] eta 0:03:27 lr 0.000827 time 0.2754 (0.2921) loss 3.7728 (3.7780) grad_norm 1.0691 (1.2049) [2021-04-15 22:14:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][550/1251] eta 0:03:24 lr 0.000827 time 0.2650 (0.2918) loss 3.9449 (3.7807) grad_norm 1.3211 (1.2041) [2021-04-15 22:14:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][560/1251] eta 0:03:21 lr 0.000827 time 0.2668 (0.2916) loss 2.3828 (3.7744) grad_norm 1.1203 (1.2047) [2021-04-15 22:14:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][570/1251] eta 0:03:18 lr 0.000827 time 0.2932 (0.2914) loss 4.2054 (3.7776) grad_norm 1.1865 (1.2034) [2021-04-15 22:14:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][580/1251] eta 0:03:15 lr 0.000827 time 0.2889 (0.2912) loss 3.9099 (3.7770) grad_norm 1.0945 (1.2025) [2021-04-15 22:14:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][590/1251] eta 0:03:12 lr 0.000827 time 0.2915 (0.2911) loss 4.1955 (3.7753) grad_norm 1.5076 (1.2033) [2021-04-15 22:14:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][600/1251] eta 0:03:09 lr 0.000827 time 0.2668 (0.2909) loss 4.0615 (3.7775) grad_norm 1.2151 (1.2016) [2021-04-15 22:15:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][610/1251] eta 0:03:06 lr 0.000827 time 0.2770 (0.2907) loss 3.7231 (3.7757) grad_norm 1.0265 (1.2003) [2021-04-15 22:15:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][620/1251] eta 0:03:03 lr 0.000826 time 0.2895 (0.2905) loss 4.0692 (3.7713) grad_norm 1.2159 (1.2009) [2021-04-15 22:15:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][630/1251] eta 0:03:00 lr 0.000826 time 0.2762 (0.2903) loss 3.9540 (3.7710) grad_norm 1.1359 (1.2014) [2021-04-15 22:15:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][640/1251] eta 0:02:57 lr 0.000826 time 0.3009 (0.2903) loss 3.5747 (3.7721) grad_norm 1.3648 (1.2004) [2021-04-15 22:15:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][650/1251] eta 0:02:54 lr 0.000826 time 0.2460 (0.2903) loss 2.5246 (3.7725) grad_norm 1.1015 (1.2000) [2021-04-15 22:15:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][660/1251] eta 0:02:51 lr 0.000826 time 0.2962 (0.2901) loss 3.9615 (3.7752) grad_norm 1.2399 (1.2008) [2021-04-15 22:15:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][670/1251] eta 0:02:48 lr 0.000826 time 0.3091 (0.2900) loss 3.7768 (3.7756) grad_norm 1.0499 (1.2018) [2021-04-15 22:15:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][680/1251] eta 0:02:45 lr 0.000826 time 0.2695 (0.2898) loss 4.2136 (3.7789) grad_norm 1.3558 (1.2027) [2021-04-15 22:15:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][690/1251] eta 0:02:42 lr 0.000826 time 0.2459 (0.2898) loss 3.8865 (3.7803) grad_norm 1.1852 (1.2036) [2021-04-15 22:15:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][700/1251] eta 0:02:39 lr 0.000826 time 0.2825 (0.2897) loss 3.7452 (3.7751) grad_norm 1.2796 (1.2044) [2021-04-15 22:15:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][710/1251] eta 0:02:36 lr 0.000826 time 0.2812 (0.2896) loss 4.3239 (3.7791) grad_norm 1.3768 (1.2050) [2021-04-15 22:15:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][720/1251] eta 0:02:33 lr 0.000826 time 0.2717 (0.2897) loss 4.0633 (3.7763) grad_norm 1.1671 (1.2048) [2021-04-15 22:15:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][730/1251] eta 0:02:30 lr 0.000826 time 0.2857 (0.2897) loss 4.2467 (3.7775) grad_norm 1.1293 (1.2054) [2021-04-15 22:15:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][740/1251] eta 0:02:28 lr 0.000826 time 0.3009 (0.2897) loss 3.7953 (3.7770) grad_norm 1.0567 (1.2046) [2021-04-15 22:15:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][750/1251] eta 0:02:25 lr 0.000826 time 0.2912 (0.2897) loss 3.9653 (3.7760) grad_norm 1.3020 (1.2043) [2021-04-15 22:15:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][760/1251] eta 0:02:22 lr 0.000826 time 0.3004 (0.2896) loss 3.9694 (3.7764) grad_norm 1.2345 (1.2041) [2021-04-15 22:15:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][770/1251] eta 0:02:19 lr 0.000826 time 0.2884 (0.2894) loss 3.9524 (3.7775) grad_norm 1.1442 (1.2047) [2021-04-15 22:15:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][780/1251] eta 0:02:16 lr 0.000826 time 0.2973 (0.2892) loss 4.8971 (3.7779) grad_norm 1.0869 (1.2047) [2021-04-15 22:15:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][790/1251] eta 0:02:13 lr 0.000826 time 0.2751 (0.2891) loss 3.0894 (3.7801) grad_norm 1.1032 (1.2058) [2021-04-15 22:15:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][800/1251] eta 0:02:10 lr 0.000826 time 0.2775 (0.2890) loss 2.6495 (3.7781) grad_norm 1.1754 (1.2060) [2021-04-15 22:15:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][810/1251] eta 0:02:07 lr 0.000826 time 0.2885 (0.2888) loss 4.5030 (3.7788) grad_norm 1.2934 (1.2057) [2021-04-15 22:16:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][820/1251] eta 0:02:04 lr 0.000826 time 0.2647 (0.2887) loss 4.3148 (3.7819) grad_norm 1.4219 (1.2064) [2021-04-15 22:16:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][830/1251] eta 0:02:01 lr 0.000826 time 0.2664 (0.2887) loss 2.8705 (3.7840) grad_norm 1.3590 (1.2061) [2021-04-15 22:16:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][840/1251] eta 0:01:58 lr 0.000826 time 0.2861 (0.2886) loss 4.2541 (3.7850) grad_norm 1.2021 (1.2053) [2021-04-15 22:16:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][850/1251] eta 0:01:55 lr 0.000826 time 0.2583 (0.2884) loss 3.0721 (3.7836) grad_norm 1.1382 (1.2046) [2021-04-15 22:16:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][860/1251] eta 0:01:52 lr 0.000826 time 0.2685 (0.2883) loss 4.8054 (3.7895) grad_norm 1.1636 (1.2044) [2021-04-15 22:16:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][870/1251] eta 0:01:49 lr 0.000826 time 0.2758 (0.2885) loss 3.4694 (3.7932) grad_norm 1.2471 (1.2037) [2021-04-15 22:16:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][880/1251] eta 0:01:46 lr 0.000826 time 0.2615 (0.2883) loss 4.0259 (3.7918) grad_norm 1.0510 (1.2032) [2021-04-15 22:16:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][890/1251] eta 0:01:44 lr 0.000826 time 0.2910 (0.2883) loss 4.1549 (3.7916) grad_norm 1.2122 (1.2033) [2021-04-15 22:16:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][900/1251] eta 0:01:41 lr 0.000826 time 0.2570 (0.2882) loss 2.8788 (3.7952) grad_norm 1.1363 (1.2025) [2021-04-15 22:16:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][910/1251] eta 0:01:38 lr 0.000826 time 0.3042 (0.2881) loss 2.7025 (3.7957) grad_norm 1.1897 (1.2021) [2021-04-15 22:16:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][920/1251] eta 0:01:35 lr 0.000826 time 0.2544 (0.2880) loss 3.7467 (3.7931) grad_norm 1.0992 (1.2023) [2021-04-15 22:16:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][930/1251] eta 0:01:32 lr 0.000826 time 0.2710 (0.2879) loss 2.7025 (3.7927) grad_norm 1.1431 (1.2017) [2021-04-15 22:16:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][940/1251] eta 0:01:29 lr 0.000825 time 0.2782 (0.2880) loss 3.7591 (3.7936) grad_norm 1.2003 (1.2015) [2021-04-15 22:16:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][950/1251] eta 0:01:26 lr 0.000825 time 0.3019 (0.2879) loss 3.5595 (3.7894) grad_norm 1.0476 (1.2012) [2021-04-15 22:16:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][960/1251] eta 0:01:23 lr 0.000825 time 0.2839 (0.2878) loss 3.2949 (3.7897) grad_norm 1.2412 (1.2021) [2021-04-15 22:16:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][970/1251] eta 0:01:20 lr 0.000825 time 0.2660 (0.2877) loss 3.4597 (3.7880) grad_norm 1.2887 (1.2018) [2021-04-15 22:16:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][980/1251] eta 0:01:17 lr 0.000825 time 0.2737 (0.2876) loss 4.3844 (3.7883) grad_norm 1.3520 (1.2018) [2021-04-15 22:16:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][990/1251] eta 0:01:15 lr 0.000825 time 0.2934 (0.2875) loss 3.9673 (3.7914) grad_norm 1.5054 (1.2029) [2021-04-15 22:16:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][1000/1251] eta 0:01:12 lr 0.000825 time 0.2837 (0.2875) loss 3.3816 (3.7907) grad_norm 1.2148 (1.2029) [2021-04-15 22:16:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][1010/1251] eta 0:01:09 lr 0.000825 time 0.2877 (0.2874) loss 4.4890 (3.7907) grad_norm 1.2815 (1.2028) [2021-04-15 22:16:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][1020/1251] eta 0:01:06 lr 0.000825 time 0.2734 (0.2873) loss 3.0503 (3.7907) grad_norm 1.0721 (1.2025) [2021-04-15 22:17:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][1030/1251] eta 0:01:03 lr 0.000825 time 0.2431 (0.2872) loss 3.9834 (3.7953) grad_norm 1.1917 (1.2028) [2021-04-15 22:17:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][1040/1251] eta 0:01:00 lr 0.000825 time 0.2975 (0.2872) loss 3.7581 (3.7951) grad_norm 1.1289 (1.2022) [2021-04-15 22:17:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][1050/1251] eta 0:00:57 lr 0.000825 time 0.2576 (0.2871) loss 2.8784 (3.7942) grad_norm 1.3108 (1.2025) [2021-04-15 22:17:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][1060/1251] eta 0:00:54 lr 0.000825 time 0.2930 (0.2870) loss 3.3248 (3.7935) grad_norm 1.1950 (1.2025) [2021-04-15 22:17:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][1070/1251] eta 0:00:51 lr 0.000825 time 0.2828 (0.2870) loss 3.3924 (3.7928) grad_norm 1.2198 (1.2033) [2021-04-15 22:17:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][1080/1251] eta 0:00:49 lr 0.000825 time 0.2780 (0.2869) loss 3.8082 (3.7931) grad_norm 1.0154 (1.2033) [2021-04-15 22:17:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][1090/1251] eta 0:00:46 lr 0.000825 time 0.2683 (0.2868) loss 4.0210 (3.7930) grad_norm 1.1155 (1.2032) [2021-04-15 22:17:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][1100/1251] eta 0:00:43 lr 0.000825 time 0.2859 (0.2867) loss 3.8990 (3.7952) grad_norm 1.2398 (1.2038) [2021-04-15 22:17:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][1110/1251] eta 0:00:40 lr 0.000825 time 0.2660 (0.2866) loss 3.7715 (3.7958) grad_norm 1.1154 (1.2038) [2021-04-15 22:17:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][1120/1251] eta 0:00:37 lr 0.000825 time 0.2529 (0.2866) loss 4.4612 (3.7964) grad_norm 1.3500 (1.2051) [2021-04-15 22:17:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][1130/1251] eta 0:00:34 lr 0.000825 time 0.3184 (0.2865) loss 3.5329 (3.7974) grad_norm 1.1343 (1.2047) [2021-04-15 22:17:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][1140/1251] eta 0:00:31 lr 0.000825 time 0.2736 (0.2864) loss 3.3871 (3.7970) grad_norm 1.0232 (1.2041) [2021-04-15 22:17:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][1150/1251] eta 0:00:28 lr 0.000825 time 0.2886 (0.2864) loss 4.1035 (3.7970) grad_norm 1.1699 (1.2037) [2021-04-15 22:17:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][1160/1251] eta 0:00:26 lr 0.000825 time 0.2955 (0.2863) loss 4.5906 (3.7992) grad_norm 1.1389 (1.2032) [2021-04-15 22:17:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][1170/1251] eta 0:00:23 lr 0.000825 time 0.2904 (0.2863) loss 3.7265 (3.7972) grad_norm 1.1693 (1.2023) [2021-04-15 22:17:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][1180/1251] eta 0:00:20 lr 0.000825 time 0.3052 (0.2863) loss 3.7521 (3.7968) grad_norm 1.3232 (1.2025) [2021-04-15 22:17:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][1190/1251] eta 0:00:17 lr 0.000825 time 0.2542 (0.2862) loss 4.1983 (3.7980) grad_norm 1.2294 (1.2016) [2021-04-15 22:17:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][1200/1251] eta 0:00:14 lr 0.000825 time 0.2630 (0.2861) loss 4.4865 (3.7952) grad_norm 1.2548 (1.2017) [2021-04-15 22:17:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][1210/1251] eta 0:00:11 lr 0.000825 time 0.2907 (0.2860) loss 2.7781 (3.7930) grad_norm 1.1839 (1.2012) [2021-04-15 22:17:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][1220/1251] eta 0:00:08 lr 0.000825 time 0.2976 (0.2860) loss 4.0787 (3.7941) grad_norm 1.0698 (1.2013) [2021-04-15 22:17:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][1230/1251] eta 0:00:06 lr 0.000825 time 0.2520 (0.2860) loss 3.0615 (3.7937) grad_norm 1.1182 (1.2014) [2021-04-15 22:17:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][1240/1251] eta 0:00:03 lr 0.000825 time 0.2484 (0.2859) loss 3.5804 (3.7921) grad_norm 1.2281 (1.2012) [2021-04-15 22:18:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [82/300][1250/1251] eta 0:00:00 lr 0.000825 time 0.2475 (0.2856) loss 4.3082 (3.7911) grad_norm 1.2050 (1.2010) [2021-04-15 22:18:04 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 82 training takes 0:06:00 [2021-04-15 22:18:04 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_82.pth saving...... [2021-04-15 22:18:22 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_82.pth saved !!! [2021-04-15 22:18:23 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.127 (1.127) Loss 1.1114 (1.1114) Acc@1 74.414 (74.414) Acc@5 93.066 (93.066) [2021-04-15 22:18:24 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.179 (0.235) Loss 1.1923 (1.1642) Acc@1 71.875 (73.233) Acc@5 90.625 (91.388) [2021-04-15 22:18:26 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.124 (0.213) Loss 1.2563 (1.1626) Acc@1 71.289 (73.270) Acc@5 89.941 (91.392) [2021-04-15 22:18:29 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.097 (0.226) Loss 1.2305 (1.1784) Acc@1 70.703 (72.754) Acc@5 91.211 (91.224) [2021-04-15 22:18:30 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.218) Loss 1.1287 (1.1728) Acc@1 72.754 (72.775) Acc@5 92.285 (91.366) [2021-04-15 22:18:37 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 72.734 Acc@5 91.410 [2021-04-15 22:18:37 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 72.7% [2021-04-15 22:18:37 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 72.73% [2021-04-15 22:18:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][0/1251] eta 1:59:48 lr 0.000825 time 5.7461 (5.7461) loss 4.5701 (4.5701) grad_norm 1.1235 (1.1235) [2021-04-15 22:18:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][10/1251] eta 0:15:52 lr 0.000824 time 0.2588 (0.7673) loss 4.1389 (3.8260) grad_norm 1.1881 (1.2032) [2021-04-15 22:18:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][20/1251] eta 0:10:58 lr 0.000824 time 0.2589 (0.5349) loss 3.3443 (3.6592) grad_norm 1.1811 (1.1921) [2021-04-15 22:18:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][30/1251] eta 0:09:16 lr 0.000824 time 0.2967 (0.4560) loss 4.4805 (3.7444) grad_norm 1.2194 (1.1936) [2021-04-15 22:18:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][40/1251] eta 0:08:18 lr 0.000824 time 0.2596 (0.4114) loss 3.3712 (3.8291) grad_norm 1.7600 (1.2048) [2021-04-15 22:18:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][50/1251] eta 0:07:42 lr 0.000824 time 0.2591 (0.3849) loss 4.3293 (3.7750) grad_norm 1.1830 (1.2113) [2021-04-15 22:18:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][60/1251] eta 0:07:18 lr 0.000824 time 0.2954 (0.3679) loss 3.7811 (3.7591) grad_norm 1.4964 (1.2342) [2021-04-15 22:19:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][70/1251] eta 0:06:59 lr 0.000824 time 0.3063 (0.3551) loss 3.6303 (3.7893) grad_norm 1.5550 (1.2439) [2021-04-15 22:19:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][80/1251] eta 0:06:45 lr 0.000824 time 0.2848 (0.3466) loss 4.5240 (3.8210) grad_norm 1.3238 (1.2411) [2021-04-15 22:19:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][90/1251] eta 0:06:33 lr 0.000824 time 0.2944 (0.3388) loss 4.4309 (3.8326) grad_norm 1.1116 (1.2337) [2021-04-15 22:19:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][100/1251] eta 0:06:23 lr 0.000824 time 0.3088 (0.3331) loss 2.9180 (3.8300) grad_norm 1.0736 (1.2300) [2021-04-15 22:19:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][110/1251] eta 0:06:14 lr 0.000824 time 0.2650 (0.3282) loss 4.1216 (3.8509) grad_norm 1.4788 (1.2367) [2021-04-15 22:19:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][120/1251] eta 0:06:06 lr 0.000824 time 0.2645 (0.3238) loss 3.7567 (3.8473) grad_norm 1.3288 (1.2443) [2021-04-15 22:19:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][130/1251] eta 0:05:58 lr 0.000824 time 0.2722 (0.3202) loss 4.2445 (3.8207) grad_norm 1.1318 (1.2413) [2021-04-15 22:19:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][140/1251] eta 0:05:53 lr 0.000824 time 0.2831 (0.3185) loss 4.1771 (3.8093) grad_norm 1.0781 (1.2367) [2021-04-15 22:19:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][150/1251] eta 0:05:49 lr 0.000824 time 0.4297 (0.3174) loss 3.2441 (3.8105) grad_norm 1.1607 (1.2324) [2021-04-15 22:19:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][160/1251] eta 0:05:44 lr 0.000824 time 0.3003 (0.3157) loss 3.5509 (3.8205) grad_norm 1.3850 (1.2316) [2021-04-15 22:19:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][170/1251] eta 0:05:38 lr 0.000824 time 0.2849 (0.3134) loss 4.7501 (3.8266) grad_norm 1.4229 (1.2355) [2021-04-15 22:19:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][180/1251] eta 0:05:33 lr 0.000824 time 0.2584 (0.3114) loss 3.5400 (3.8226) grad_norm 1.2153 (1.2380) [2021-04-15 22:19:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][190/1251] eta 0:05:28 lr 0.000824 time 0.2659 (0.3099) loss 4.2578 (3.8209) grad_norm 1.2791 (1.2384) [2021-04-15 22:19:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][200/1251] eta 0:05:24 lr 0.000824 time 0.2926 (0.3084) loss 2.5190 (3.8000) grad_norm 1.1854 (1.2362) [2021-04-15 22:19:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][210/1251] eta 0:05:19 lr 0.000824 time 0.2792 (0.3068) loss 2.9182 (3.7966) grad_norm 1.5249 (1.2441) [2021-04-15 22:19:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][220/1251] eta 0:05:14 lr 0.000824 time 0.2769 (0.3055) loss 4.4551 (3.7932) grad_norm 1.2758 (1.2441) [2021-04-15 22:19:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][230/1251] eta 0:05:10 lr 0.000824 time 0.3032 (0.3044) loss 3.2855 (3.7921) grad_norm 1.1212 (1.2397) [2021-04-15 22:19:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][240/1251] eta 0:05:06 lr 0.000824 time 0.2884 (0.3032) loss 3.5255 (3.7790) grad_norm 1.1790 (1.2358) [2021-04-15 22:19:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][250/1251] eta 0:05:03 lr 0.000824 time 0.2680 (0.3028) loss 3.6166 (3.7843) grad_norm 1.2027 (1.2336) [2021-04-15 22:19:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][260/1251] eta 0:04:58 lr 0.000824 time 0.2859 (0.3017) loss 3.5388 (3.7835) grad_norm 1.1126 (1.2316) [2021-04-15 22:19:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][270/1251] eta 0:04:54 lr 0.000824 time 0.2993 (0.3007) loss 4.5156 (3.7848) grad_norm 1.1309 (1.2285) [2021-04-15 22:20:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][280/1251] eta 0:04:51 lr 0.000824 time 0.2723 (0.2998) loss 4.2325 (3.7824) grad_norm 1.2467 (1.2252) [2021-04-15 22:20:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][290/1251] eta 0:04:47 lr 0.000824 time 0.2785 (0.2994) loss 3.5873 (3.7870) grad_norm 1.1470 (1.2241) [2021-04-15 22:20:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][300/1251] eta 0:04:43 lr 0.000824 time 0.2544 (0.2985) loss 4.1793 (3.7896) grad_norm 1.2328 (1.2231) [2021-04-15 22:20:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][310/1251] eta 0:04:40 lr 0.000824 time 0.2967 (0.2980) loss 4.1367 (3.7802) grad_norm 1.2931 (1.2245) [2021-04-15 22:20:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][320/1251] eta 0:04:36 lr 0.000823 time 0.2865 (0.2973) loss 4.7143 (3.7832) grad_norm 1.2299 (1.2230) [2021-04-15 22:20:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][330/1251] eta 0:04:33 lr 0.000823 time 0.2683 (0.2966) loss 4.5539 (3.7783) grad_norm 1.2351 (1.2213) [2021-04-15 22:20:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][340/1251] eta 0:04:29 lr 0.000823 time 0.2691 (0.2960) loss 3.9022 (3.7843) grad_norm 1.4818 (1.2251) [2021-04-15 22:20:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][350/1251] eta 0:04:26 lr 0.000823 time 0.2820 (0.2955) loss 4.0890 (3.7788) grad_norm 1.2059 (1.2253) [2021-04-15 22:20:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][360/1251] eta 0:04:22 lr 0.000823 time 0.2799 (0.2950) loss 3.1736 (3.7785) grad_norm 1.2166 (1.2256) [2021-04-15 22:20:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][370/1251] eta 0:04:19 lr 0.000823 time 0.2977 (0.2949) loss 4.4372 (3.7914) grad_norm 1.3664 (1.2251) [2021-04-15 22:20:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][380/1251] eta 0:04:16 lr 0.000823 time 0.3022 (0.2945) loss 4.2756 (3.7975) grad_norm 1.1401 (1.2241) [2021-04-15 22:20:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][390/1251] eta 0:04:13 lr 0.000823 time 0.2658 (0.2939) loss 3.7558 (3.7970) grad_norm 1.1394 (1.2262) [2021-04-15 22:20:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][400/1251] eta 0:04:09 lr 0.000823 time 0.2879 (0.2935) loss 4.4056 (3.7999) grad_norm 1.1537 (1.2252) [2021-04-15 22:20:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][410/1251] eta 0:04:06 lr 0.000823 time 0.2726 (0.2930) loss 3.6506 (3.7984) grad_norm 1.3713 (1.2251) [2021-04-15 22:20:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][420/1251] eta 0:04:03 lr 0.000823 time 0.2507 (0.2931) loss 4.0386 (3.7970) grad_norm 1.1419 (1.2250) [2021-04-15 22:20:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][430/1251] eta 0:04:00 lr 0.000823 time 0.2774 (0.2932) loss 3.8703 (3.7956) grad_norm 1.0395 (1.2234) [2021-04-15 22:20:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][440/1251] eta 0:03:57 lr 0.000823 time 0.2731 (0.2928) loss 4.1124 (3.7873) grad_norm 1.2069 (1.2229) [2021-04-15 22:20:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][450/1251] eta 0:03:54 lr 0.000823 time 0.2743 (0.2925) loss 4.1499 (3.7856) grad_norm 1.1207 (1.2221) [2021-04-15 22:20:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][460/1251] eta 0:03:51 lr 0.000823 time 0.2794 (0.2921) loss 4.1639 (3.7899) grad_norm 1.3636 (1.2210) [2021-04-15 22:20:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][470/1251] eta 0:03:47 lr 0.000823 time 0.2501 (0.2918) loss 3.9224 (3.7898) grad_norm 1.2577 (1.2224) [2021-04-15 22:20:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][480/1251] eta 0:03:44 lr 0.000823 time 0.2727 (0.2915) loss 3.5237 (3.7895) grad_norm 1.0741 (1.2212) [2021-04-15 22:21:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][490/1251] eta 0:03:41 lr 0.000823 time 0.2789 (0.2915) loss 4.2856 (3.7908) grad_norm 1.2550 (1.2251) [2021-04-15 22:21:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][500/1251] eta 0:03:38 lr 0.000823 time 0.2892 (0.2915) loss 4.0500 (3.7948) grad_norm 1.0247 (1.2248) [2021-04-15 22:21:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][510/1251] eta 0:03:35 lr 0.000823 time 0.2890 (0.2912) loss 4.4596 (3.7937) grad_norm 1.0951 (1.2228) [2021-04-15 22:21:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][520/1251] eta 0:03:32 lr 0.000823 time 0.3115 (0.2911) loss 3.8585 (3.7958) grad_norm 1.5052 (1.2236) [2021-04-15 22:21:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][530/1251] eta 0:03:29 lr 0.000823 time 0.2755 (0.2909) loss 3.4073 (3.7980) grad_norm 1.2191 (1.2232) [2021-04-15 22:21:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][540/1251] eta 0:03:26 lr 0.000823 time 0.2806 (0.2911) loss 3.4467 (3.7938) grad_norm 1.2381 (1.2215) [2021-04-15 22:21:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][550/1251] eta 0:03:23 lr 0.000823 time 0.2718 (0.2907) loss 2.4726 (3.7891) grad_norm 1.5704 (1.2211) [2021-04-15 22:21:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][560/1251] eta 0:03:20 lr 0.000823 time 0.2883 (0.2905) loss 4.1130 (3.7937) grad_norm 1.8306 (1.2235) [2021-04-15 22:21:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][570/1251] eta 0:03:17 lr 0.000823 time 0.2507 (0.2904) loss 3.8425 (3.7943) grad_norm 1.0899 (1.2231) [2021-04-15 22:21:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][580/1251] eta 0:03:14 lr 0.000823 time 0.2715 (0.2902) loss 4.1095 (3.7953) grad_norm 1.2271 (1.2231) [2021-04-15 22:21:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][590/1251] eta 0:03:11 lr 0.000823 time 0.2802 (0.2899) loss 3.2877 (3.7977) grad_norm 1.1915 (1.2233) [2021-04-15 22:21:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][600/1251] eta 0:03:08 lr 0.000823 time 0.2760 (0.2896) loss 3.4611 (3.7947) grad_norm 1.1105 (1.2223) [2021-04-15 22:21:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][610/1251] eta 0:03:05 lr 0.000823 time 0.2842 (0.2894) loss 3.8097 (3.7909) grad_norm 1.0677 (1.2221) [2021-04-15 22:21:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][620/1251] eta 0:03:02 lr 0.000823 time 0.2705 (0.2892) loss 3.6065 (3.7909) grad_norm 1.4290 (1.2229) [2021-04-15 22:21:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][630/1251] eta 0:02:59 lr 0.000823 time 0.2699 (0.2890) loss 3.7790 (3.7945) grad_norm 1.3543 (1.2227) [2021-04-15 22:21:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][640/1251] eta 0:02:56 lr 0.000822 time 0.2810 (0.2888) loss 3.0149 (3.7940) grad_norm 1.4018 (1.2223) [2021-04-15 22:21:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][650/1251] eta 0:02:53 lr 0.000822 time 0.2560 (0.2886) loss 4.6723 (3.7971) grad_norm 1.2414 (1.2219) [2021-04-15 22:21:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][660/1251] eta 0:02:50 lr 0.000822 time 0.2818 (0.2884) loss 4.1181 (3.7969) grad_norm 1.2526 (1.2229) [2021-04-15 22:21:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][670/1251] eta 0:02:47 lr 0.000822 time 0.2834 (0.2883) loss 3.6544 (3.7956) grad_norm 1.1325 (1.2220) [2021-04-15 22:21:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][680/1251] eta 0:02:44 lr 0.000822 time 0.2806 (0.2882) loss 4.1110 (3.7967) grad_norm 1.4544 (1.2226) [2021-04-15 22:21:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][690/1251] eta 0:02:41 lr 0.000822 time 0.2802 (0.2880) loss 3.6596 (3.7979) grad_norm 1.2559 (1.2236) [2021-04-15 22:21:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][700/1251] eta 0:02:38 lr 0.000822 time 0.3050 (0.2879) loss 4.3122 (3.7992) grad_norm 1.0811 (1.2229) [2021-04-15 22:22:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][710/1251] eta 0:02:35 lr 0.000822 time 0.2950 (0.2877) loss 4.1116 (3.7999) grad_norm 1.1476 (1.2233) [2021-04-15 22:22:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][720/1251] eta 0:02:32 lr 0.000822 time 0.2530 (0.2880) loss 4.5696 (3.7974) grad_norm 1.2888 (1.2248) [2021-04-15 22:22:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][730/1251] eta 0:02:30 lr 0.000822 time 0.2690 (0.2885) loss 3.3077 (3.7948) grad_norm 1.2153 (1.2246) [2021-04-15 22:22:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][740/1251] eta 0:02:27 lr 0.000822 time 0.2882 (0.2883) loss 4.3650 (3.7925) grad_norm 1.1805 (1.2237) [2021-04-15 22:22:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][750/1251] eta 0:02:24 lr 0.000822 time 0.2577 (0.2882) loss 3.8767 (3.7946) grad_norm 1.1227 (1.2235) [2021-04-15 22:22:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][760/1251] eta 0:02:21 lr 0.000822 time 0.2935 (0.2880) loss 4.2444 (3.7940) grad_norm 1.0055 (1.2231) [2021-04-15 22:22:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][770/1251] eta 0:02:18 lr 0.000822 time 0.2693 (0.2879) loss 3.8214 (3.7930) grad_norm 1.2631 (1.2225) [2021-04-15 22:22:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][780/1251] eta 0:02:15 lr 0.000822 time 0.2675 (0.2877) loss 4.2554 (3.7970) grad_norm 1.2774 (1.2223) [2021-04-15 22:22:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][790/1251] eta 0:02:12 lr 0.000822 time 0.2754 (0.2875) loss 4.1532 (3.7999) grad_norm 0.9887 (1.2223) [2021-04-15 22:22:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][800/1251] eta 0:02:09 lr 0.000822 time 0.2752 (0.2874) loss 2.6800 (3.8008) grad_norm 0.9745 (1.2214) [2021-04-15 22:22:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][810/1251] eta 0:02:06 lr 0.000822 time 0.2780 (0.2873) loss 3.7221 (3.8005) grad_norm 1.2641 (1.2208) [2021-04-15 22:22:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][820/1251] eta 0:02:03 lr 0.000822 time 0.2998 (0.2872) loss 3.7633 (3.7980) grad_norm 1.2272 (1.2215) [2021-04-15 22:22:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][830/1251] eta 0:02:00 lr 0.000822 time 0.2575 (0.2871) loss 4.5254 (3.7968) grad_norm 1.2064 (1.2216) [2021-04-15 22:22:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][840/1251] eta 0:01:57 lr 0.000822 time 0.2502 (0.2870) loss 3.2391 (3.7958) grad_norm 1.5280 (1.2214) [2021-04-15 22:22:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][850/1251] eta 0:01:55 lr 0.000822 time 0.2901 (0.2868) loss 4.0085 (3.7944) grad_norm 1.1080 (1.2212) [2021-04-15 22:22:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][860/1251] eta 0:01:52 lr 0.000822 time 0.2649 (0.2866) loss 2.8757 (3.7930) grad_norm 1.2810 (1.2207) [2021-04-15 22:22:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][870/1251] eta 0:01:49 lr 0.000822 time 0.2984 (0.2867) loss 4.2420 (3.7901) grad_norm 1.0724 (1.2199) [2021-04-15 22:22:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][880/1251] eta 0:01:46 lr 0.000822 time 0.2969 (0.2865) loss 2.9395 (3.7896) grad_norm 1.1928 (1.2199) [2021-04-15 22:22:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][890/1251] eta 0:01:43 lr 0.000822 time 0.2815 (0.2864) loss 4.6800 (3.7909) grad_norm 1.2640 (1.2193) [2021-04-15 22:22:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][900/1251] eta 0:01:40 lr 0.000822 time 0.2767 (0.2863) loss 3.5102 (3.7894) grad_norm 1.1629 (1.2194) [2021-04-15 22:22:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][910/1251] eta 0:01:37 lr 0.000822 time 0.2809 (0.2862) loss 4.3922 (3.7914) grad_norm 1.0734 (1.2186) [2021-04-15 22:23:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][920/1251] eta 0:01:34 lr 0.000822 time 0.2475 (0.2860) loss 3.6234 (3.7901) grad_norm 1.1188 (1.2183) [2021-04-15 22:23:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][930/1251] eta 0:01:31 lr 0.000822 time 0.2608 (0.2859) loss 3.5501 (3.7891) grad_norm 1.1924 (1.2181) [2021-04-15 22:23:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][940/1251] eta 0:01:28 lr 0.000822 time 0.3946 (0.2859) loss 3.5515 (3.7899) grad_norm 1.2113 (1.2177) [2021-04-15 22:23:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][950/1251] eta 0:01:26 lr 0.000821 time 0.2879 (0.2858) loss 3.5909 (3.7881) grad_norm 1.1655 (1.2169) [2021-04-15 22:23:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][960/1251] eta 0:01:23 lr 0.000821 time 0.2812 (0.2858) loss 3.7048 (3.7867) grad_norm 1.1089 (1.2162) [2021-04-15 22:23:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][970/1251] eta 0:01:20 lr 0.000821 time 0.2781 (0.2857) loss 4.0640 (3.7852) grad_norm 1.3137 (1.2168) [2021-04-15 22:23:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][980/1251] eta 0:01:17 lr 0.000821 time 0.2751 (0.2857) loss 3.8865 (3.7816) grad_norm 1.1984 (1.2163) [2021-04-15 22:23:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][990/1251] eta 0:01:14 lr 0.000821 time 0.2720 (0.2856) loss 2.7387 (3.7829) grad_norm 1.2696 (1.2167) [2021-04-15 22:23:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][1000/1251] eta 0:01:11 lr 0.000821 time 0.2744 (0.2855) loss 4.6279 (3.7800) grad_norm 1.2975 (1.2168) [2021-04-15 22:23:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][1010/1251] eta 0:01:08 lr 0.000821 time 0.2851 (0.2855) loss 3.6832 (3.7789) grad_norm 1.1425 (1.2161) [2021-04-15 22:23:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][1020/1251] eta 0:01:05 lr 0.000821 time 0.3050 (0.2855) loss 3.1755 (3.7785) grad_norm 1.2261 (1.2162) [2021-04-15 22:23:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][1030/1251] eta 0:01:03 lr 0.000821 time 0.2525 (0.2853) loss 3.5780 (3.7803) grad_norm 1.1077 (1.2163) [2021-04-15 22:23:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][1040/1251] eta 0:01:00 lr 0.000821 time 0.2753 (0.2853) loss 2.6224 (3.7792) grad_norm 0.9784 (1.2151) [2021-04-15 22:23:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][1050/1251] eta 0:00:57 lr 0.000821 time 0.3020 (0.2852) loss 4.2174 (3.7788) grad_norm 1.2868 (1.2155) [2021-04-15 22:23:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][1060/1251] eta 0:00:54 lr 0.000821 time 0.2663 (0.2851) loss 4.5094 (3.7805) grad_norm 1.2448 (1.2151) [2021-04-15 22:23:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][1070/1251] eta 0:00:51 lr 0.000821 time 0.2510 (0.2850) loss 4.3743 (3.7786) grad_norm 1.2059 (1.2145) [2021-04-15 22:23:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][1080/1251] eta 0:00:48 lr 0.000821 time 0.2670 (0.2851) loss 3.8154 (3.7767) grad_norm 0.9451 (1.2146) [2021-04-15 22:23:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][1090/1251] eta 0:00:45 lr 0.000821 time 0.2959 (0.2850) loss 4.6339 (3.7771) grad_norm 1.1093 (1.2147) [2021-04-15 22:23:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][1100/1251] eta 0:00:43 lr 0.000821 time 0.2824 (0.2849) loss 3.0195 (3.7748) grad_norm 1.1246 (1.2142) [2021-04-15 22:23:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][1110/1251] eta 0:00:40 lr 0.000821 time 0.2640 (0.2848) loss 4.1910 (3.7754) grad_norm 1.2464 (1.2132) [2021-04-15 22:23:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][1120/1251] eta 0:00:37 lr 0.000821 time 0.2819 (0.2848) loss 3.5235 (3.7759) grad_norm 1.2048 (1.2131) [2021-04-15 22:23:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][1130/1251] eta 0:00:34 lr 0.000821 time 0.2808 (0.2847) loss 3.7574 (3.7756) grad_norm 1.2132 (1.2125) [2021-04-15 22:24:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][1140/1251] eta 0:00:31 lr 0.000821 time 0.2724 (0.2847) loss 4.4813 (3.7779) grad_norm 1.1780 (1.2122) [2021-04-15 22:24:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][1150/1251] eta 0:00:28 lr 0.000821 time 0.2676 (0.2846) loss 3.7154 (3.7764) grad_norm 1.2008 (1.2119) [2021-04-15 22:24:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][1160/1251] eta 0:00:25 lr 0.000821 time 0.2907 (0.2846) loss 3.9406 (3.7736) grad_norm 1.1468 (1.2113) [2021-04-15 22:24:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][1170/1251] eta 0:00:23 lr 0.000821 time 0.2848 (0.2845) loss 3.5403 (3.7721) grad_norm 1.3603 (1.2110) [2021-04-15 22:24:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][1180/1251] eta 0:00:20 lr 0.000821 time 0.2681 (0.2844) loss 3.1773 (3.7723) grad_norm 1.1251 (1.2110) [2021-04-15 22:24:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][1190/1251] eta 0:00:17 lr 0.000821 time 0.2780 (0.2843) loss 4.4668 (3.7738) grad_norm 1.0543 (1.2106) [2021-04-15 22:24:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][1200/1251] eta 0:00:14 lr 0.000821 time 0.2567 (0.2843) loss 2.5397 (3.7723) grad_norm 1.2128 (1.2102) [2021-04-15 22:24:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][1210/1251] eta 0:00:11 lr 0.000821 time 0.2929 (0.2842) loss 4.1264 (3.7704) grad_norm 1.2319 (1.2103) [2021-04-15 22:24:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][1220/1251] eta 0:00:08 lr 0.000821 time 0.2773 (0.2841) loss 3.0533 (3.7676) grad_norm 1.1895 (1.2097) [2021-04-15 22:24:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][1230/1251] eta 0:00:05 lr 0.000821 time 0.2692 (0.2840) loss 4.5174 (3.7710) grad_norm 1.4173 (1.2092) [2021-04-15 22:24:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][1240/1251] eta 0:00:03 lr 0.000821 time 0.2486 (0.2839) loss 4.3109 (3.7728) grad_norm 1.0689 (1.2088) [2021-04-15 22:24:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [83/300][1250/1251] eta 0:00:00 lr 0.000821 time 0.2490 (0.2836) loss 2.8848 (3.7746) grad_norm 0.9902 (1.2079) [2021-04-15 22:24:34 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 83 training takes 0:05:57 [2021-04-15 22:24:34 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_83.pth saving...... [2021-04-15 22:24:50 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_83.pth saved !!! [2021-04-15 22:24:52 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.202 (1.202) Loss 1.2150 (1.2150) Acc@1 72.266 (72.266) Acc@5 90.527 (90.527) [2021-04-15 22:24:54 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.437 (0.299) Loss 1.1278 (1.1752) Acc@1 73.145 (72.203) Acc@5 91.797 (91.184) [2021-04-15 22:24:55 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.123 (0.222) Loss 1.1515 (1.1691) Acc@1 73.633 (72.480) Acc@5 91.309 (91.383) [2021-04-15 22:24:57 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.138 (0.227) Loss 1.1492 (1.1639) Acc@1 73.242 (72.527) Acc@5 91.113 (91.428) [2021-04-15 22:24:59 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.142 (0.219) Loss 1.1583 (1.1701) Acc@1 72.656 (72.389) Acc@5 91.211 (91.409) [2021-04-15 22:25:03 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 72.460 Acc@5 91.452 [2021-04-15 22:25:03 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 72.5% [2021-04-15 22:25:03 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 72.73% [2021-04-15 22:25:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][0/1251] eta 2:22:52 lr 0.000821 time 6.8522 (6.8522) loss 3.4567 (3.4567) grad_norm 1.3880 (1.3880) [2021-04-15 22:25:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][10/1251] eta 0:18:02 lr 0.000820 time 0.2928 (0.8723) loss 3.9972 (3.8869) grad_norm 1.5102 (1.1996) [2021-04-15 22:25:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][20/1251] eta 0:12:06 lr 0.000820 time 0.3040 (0.5899) loss 4.5392 (3.9252) grad_norm 1.0841 (1.2134) [2021-04-15 22:25:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][30/1251] eta 0:10:00 lr 0.000820 time 0.2938 (0.4916) loss 3.2299 (3.8264) grad_norm 1.2387 (1.2238) [2021-04-15 22:25:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][40/1251] eta 0:08:55 lr 0.000820 time 0.2905 (0.4424) loss 3.7997 (3.7545) grad_norm 1.1349 (1.2429) [2021-04-15 22:25:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][50/1251] eta 0:08:12 lr 0.000820 time 0.2975 (0.4100) loss 4.0752 (3.7564) grad_norm 1.2034 (1.2490) [2021-04-15 22:25:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][60/1251] eta 0:07:43 lr 0.000820 time 0.2814 (0.3889) loss 3.7753 (3.7836) grad_norm 1.2456 (1.2439) [2021-04-15 22:25:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][70/1251] eta 0:07:19 lr 0.000820 time 0.2822 (0.3724) loss 4.1454 (3.7876) grad_norm 1.2260 (1.2336) [2021-04-15 22:25:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][80/1251] eta 0:07:04 lr 0.000820 time 0.2726 (0.3624) loss 4.7148 (3.7767) grad_norm 1.0871 (1.2313) [2021-04-15 22:25:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][90/1251] eta 0:06:49 lr 0.000820 time 0.2707 (0.3529) loss 4.1177 (3.7563) grad_norm 1.2694 (1.2249) [2021-04-15 22:25:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][100/1251] eta 0:06:38 lr 0.000820 time 0.2885 (0.3463) loss 3.8061 (3.7754) grad_norm 1.1095 (1.2243) [2021-04-15 22:25:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][110/1251] eta 0:06:27 lr 0.000820 time 0.2706 (0.3397) loss 4.0490 (3.7854) grad_norm 1.1511 (1.2208) [2021-04-15 22:25:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][120/1251] eta 0:06:18 lr 0.000820 time 0.3118 (0.3346) loss 3.4985 (3.7732) grad_norm 1.0757 (1.2091) [2021-04-15 22:25:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][130/1251] eta 0:06:10 lr 0.000820 time 0.2560 (0.3305) loss 4.3904 (3.8039) grad_norm 1.1515 (1.2079) [2021-04-15 22:25:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][140/1251] eta 0:06:03 lr 0.000820 time 0.2675 (0.3268) loss 4.3144 (3.7894) grad_norm 1.2505 (1.2047) [2021-04-15 22:25:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][150/1251] eta 0:05:57 lr 0.000820 time 0.4089 (0.3243) loss 3.5076 (3.7848) grad_norm 1.4297 (1.2060) [2021-04-15 22:25:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][160/1251] eta 0:05:52 lr 0.000820 time 0.2902 (0.3227) loss 3.3053 (3.7805) grad_norm 1.1698 (1.2069) [2021-04-15 22:25:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][170/1251] eta 0:05:46 lr 0.000820 time 0.2959 (0.3201) loss 3.0854 (3.7787) grad_norm 1.1744 (1.2090) [2021-04-15 22:26:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][180/1251] eta 0:05:40 lr 0.000820 time 0.2733 (0.3177) loss 2.5145 (3.7534) grad_norm 1.1232 (1.2075) [2021-04-15 22:26:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][190/1251] eta 0:05:34 lr 0.000820 time 0.2784 (0.3155) loss 3.6762 (3.7454) grad_norm 1.0110 (1.2055) [2021-04-15 22:26:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][200/1251] eta 0:05:30 lr 0.000820 time 0.2596 (0.3145) loss 4.1515 (3.7503) grad_norm 1.1944 (1.2059) [2021-04-15 22:26:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][210/1251] eta 0:05:25 lr 0.000820 time 0.3048 (0.3130) loss 2.5494 (3.7494) grad_norm 1.1761 (1.2084) [2021-04-15 22:26:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][220/1251] eta 0:05:21 lr 0.000820 time 0.2993 (0.3120) loss 4.6095 (3.7434) grad_norm 1.2061 (1.2107) [2021-04-15 22:26:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][230/1251] eta 0:05:16 lr 0.000820 time 0.2498 (0.3103) loss 2.6474 (3.7484) grad_norm 1.0832 (1.2096) [2021-04-15 22:26:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][240/1251] eta 0:05:12 lr 0.000820 time 0.2997 (0.3089) loss 4.1432 (3.7374) grad_norm 1.4000 (1.2104) [2021-04-15 22:26:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][250/1251] eta 0:05:08 lr 0.000820 time 0.2929 (0.3078) loss 3.4349 (3.7280) grad_norm 1.0611 (1.2089) [2021-04-15 22:26:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][260/1251] eta 0:05:04 lr 0.000820 time 0.3121 (0.3068) loss 3.7932 (3.7358) grad_norm 1.5068 (1.2122) [2021-04-15 22:26:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][270/1251] eta 0:05:00 lr 0.000820 time 0.2766 (0.3060) loss 4.5025 (3.7292) grad_norm 1.1644 (1.2127) [2021-04-15 22:26:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][280/1251] eta 0:04:56 lr 0.000820 time 0.2754 (0.3050) loss 4.2559 (3.7282) grad_norm 1.4137 (1.2152) [2021-04-15 22:26:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][290/1251] eta 0:04:52 lr 0.000820 time 0.2876 (0.3045) loss 4.5074 (3.7344) grad_norm 1.3548 (1.2177) [2021-04-15 22:26:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][300/1251] eta 0:04:49 lr 0.000820 time 0.2618 (0.3040) loss 4.0785 (3.7352) grad_norm 1.4308 (1.2184) [2021-04-15 22:26:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][310/1251] eta 0:04:45 lr 0.000820 time 0.2821 (0.3031) loss 2.8915 (3.7353) grad_norm 1.2000 (1.2201) [2021-04-15 22:26:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][320/1251] eta 0:04:41 lr 0.000820 time 0.2926 (0.3024) loss 4.4096 (3.7393) grad_norm 1.2206 (1.2186) [2021-04-15 22:26:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][330/1251] eta 0:04:38 lr 0.000819 time 0.3068 (0.3022) loss 2.8496 (3.7301) grad_norm 1.3280 (1.2192) [2021-04-15 22:26:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][340/1251] eta 0:04:34 lr 0.000819 time 0.2688 (0.3015) loss 4.5328 (3.7333) grad_norm 0.9606 (1.2178) [2021-04-15 22:26:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][350/1251] eta 0:04:31 lr 0.000819 time 0.2713 (0.3012) loss 3.8301 (3.7333) grad_norm 1.1721 (1.2155) [2021-04-15 22:26:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][360/1251] eta 0:04:27 lr 0.000819 time 0.2637 (0.3004) loss 4.4213 (3.7344) grad_norm 1.1227 (1.2163) [2021-04-15 22:26:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][370/1251] eta 0:04:24 lr 0.000819 time 0.2853 (0.2997) loss 3.1952 (3.7295) grad_norm 1.0813 (1.2154) [2021-04-15 22:26:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][380/1251] eta 0:04:20 lr 0.000819 time 0.2823 (0.2992) loss 3.1597 (3.7187) grad_norm 1.1226 (1.2127) [2021-04-15 22:26:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][390/1251] eta 0:04:17 lr 0.000819 time 0.2735 (0.2987) loss 4.1439 (3.7228) grad_norm 1.1011 (1.2113) [2021-04-15 22:27:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][400/1251] eta 0:04:13 lr 0.000819 time 0.2757 (0.2981) loss 4.3077 (3.7265) grad_norm 1.0949 (1.2106) [2021-04-15 22:27:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][410/1251] eta 0:04:10 lr 0.000819 time 0.2615 (0.2979) loss 3.8048 (3.7284) grad_norm 1.1307 (1.2118) [2021-04-15 22:27:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][420/1251] eta 0:04:07 lr 0.000819 time 0.3013 (0.2974) loss 3.8029 (3.7233) grad_norm 1.0949 (1.2105) [2021-04-15 22:27:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][430/1251] eta 0:04:03 lr 0.000819 time 0.2871 (0.2969) loss 3.0532 (3.7258) grad_norm 1.0500 (1.2089) [2021-04-15 22:27:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][440/1251] eta 0:04:00 lr 0.000819 time 0.2724 (0.2964) loss 4.1184 (3.7259) grad_norm 1.0935 (1.2083) [2021-04-15 22:27:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][450/1251] eta 0:03:56 lr 0.000819 time 0.2798 (0.2959) loss 3.1875 (3.7238) grad_norm 1.1360 (inf) [2021-04-15 22:27:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][460/1251] eta 0:03:53 lr 0.000819 time 0.2903 (0.2955) loss 3.8181 (3.7193) grad_norm 1.0659 (inf) [2021-04-15 22:27:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][470/1251] eta 0:03:50 lr 0.000819 time 0.2923 (0.2951) loss 3.6978 (3.7255) grad_norm 1.1998 (inf) [2021-04-15 22:27:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][480/1251] eta 0:03:47 lr 0.000819 time 0.2911 (0.2948) loss 3.3642 (3.7252) grad_norm 1.2071 (inf) [2021-04-15 22:27:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][490/1251] eta 0:03:43 lr 0.000819 time 0.2460 (0.2943) loss 3.8019 (3.7215) grad_norm 1.1163 (inf) [2021-04-15 22:27:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][500/1251] eta 0:03:40 lr 0.000819 time 0.2771 (0.2940) loss 4.0990 (3.7232) grad_norm 1.3158 (inf) [2021-04-15 22:27:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][510/1251] eta 0:03:37 lr 0.000819 time 0.2842 (0.2937) loss 4.3397 (3.7291) grad_norm 1.1371 (inf) [2021-04-15 22:27:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][520/1251] eta 0:03:34 lr 0.000819 time 0.2916 (0.2933) loss 4.4597 (3.7318) grad_norm 1.2858 (inf) [2021-04-15 22:27:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][530/1251] eta 0:03:31 lr 0.000819 time 0.2624 (0.2930) loss 4.4830 (3.7310) grad_norm 1.2420 (inf) [2021-04-15 22:27:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][540/1251] eta 0:03:28 lr 0.000819 time 0.2703 (0.2930) loss 3.8245 (3.7283) grad_norm 1.2703 (inf) [2021-04-15 22:27:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][550/1251] eta 0:03:25 lr 0.000819 time 0.2939 (0.2928) loss 4.3113 (3.7311) grad_norm 1.6942 (inf) [2021-04-15 22:27:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][560/1251] eta 0:03:22 lr 0.000819 time 0.2744 (0.2924) loss 3.5943 (3.7327) grad_norm 1.1105 (inf) [2021-04-15 22:27:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][570/1251] eta 0:03:19 lr 0.000819 time 0.2818 (0.2924) loss 4.1171 (3.7371) grad_norm 1.2947 (inf) [2021-04-15 22:27:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][580/1251] eta 0:03:16 lr 0.000819 time 0.2759 (0.2926) loss 3.6957 (3.7369) grad_norm 1.0137 (inf) [2021-04-15 22:27:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][590/1251] eta 0:03:13 lr 0.000819 time 0.2737 (0.2923) loss 4.3957 (3.7447) grad_norm 1.1245 (inf) [2021-04-15 22:27:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][600/1251] eta 0:03:10 lr 0.000819 time 0.2848 (0.2920) loss 3.9968 (3.7446) grad_norm 1.4258 (inf) [2021-04-15 22:28:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][610/1251] eta 0:03:06 lr 0.000819 time 0.2748 (0.2917) loss 4.4057 (3.7489) grad_norm 1.1729 (inf) [2021-04-15 22:28:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][620/1251] eta 0:03:04 lr 0.000819 time 0.2681 (0.2917) loss 4.3646 (3.7551) grad_norm 1.2087 (inf) [2021-04-15 22:28:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][630/1251] eta 0:03:01 lr 0.000819 time 0.2608 (0.2915) loss 3.2013 (3.7522) grad_norm 1.0789 (inf) [2021-04-15 22:28:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][640/1251] eta 0:02:57 lr 0.000818 time 0.2581 (0.2912) loss 3.7724 (3.7520) grad_norm 1.0967 (inf) [2021-04-15 22:28:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][650/1251] eta 0:02:55 lr 0.000818 time 0.3715 (0.2912) loss 4.1495 (3.7504) grad_norm 1.0303 (inf) [2021-04-15 22:28:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][660/1251] eta 0:02:51 lr 0.000818 time 0.2991 (0.2909) loss 3.4861 (3.7547) grad_norm 1.1955 (inf) [2021-04-15 22:28:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][670/1251] eta 0:02:48 lr 0.000818 time 0.2974 (0.2907) loss 3.9733 (3.7585) grad_norm 1.6513 (inf) [2021-04-15 22:28:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][680/1251] eta 0:02:45 lr 0.000818 time 0.2762 (0.2906) loss 3.7587 (3.7570) grad_norm 1.3996 (inf) [2021-04-15 22:28:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][690/1251] eta 0:02:42 lr 0.000818 time 0.2740 (0.2903) loss 3.2823 (3.7531) grad_norm 1.4611 (inf) [2021-04-15 22:28:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][700/1251] eta 0:02:39 lr 0.000818 time 0.2748 (0.2901) loss 3.8191 (3.7482) grad_norm 1.4146 (inf) [2021-04-15 22:28:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][710/1251] eta 0:02:36 lr 0.000818 time 0.2705 (0.2899) loss 4.5337 (3.7512) grad_norm 1.2494 (inf) [2021-04-15 22:28:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][720/1251] eta 0:02:33 lr 0.000818 time 0.2464 (0.2897) loss 3.4062 (3.7524) grad_norm 0.9914 (inf) [2021-04-15 22:28:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][730/1251] eta 0:02:30 lr 0.000818 time 0.2692 (0.2897) loss 2.9766 (3.7526) grad_norm 1.1290 (inf) [2021-04-15 22:28:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][740/1251] eta 0:02:27 lr 0.000818 time 0.2840 (0.2896) loss 3.2878 (3.7482) grad_norm 1.0593 (inf) [2021-04-15 22:28:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][750/1251] eta 0:02:24 lr 0.000818 time 0.2932 (0.2894) loss 3.2853 (3.7481) grad_norm 1.0911 (inf) [2021-04-15 22:28:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][760/1251] eta 0:02:21 lr 0.000818 time 0.2727 (0.2892) loss 4.4827 (3.7469) grad_norm 1.0871 (inf) [2021-04-15 22:28:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][770/1251] eta 0:02:18 lr 0.000818 time 0.2547 (0.2890) loss 3.8417 (3.7455) grad_norm 1.2092 (inf) [2021-04-15 22:28:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][780/1251] eta 0:02:16 lr 0.000818 time 0.2543 (0.2889) loss 4.1660 (3.7451) grad_norm 1.2821 (inf) [2021-04-15 22:28:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][790/1251] eta 0:02:13 lr 0.000818 time 0.2529 (0.2889) loss 2.8766 (3.7439) grad_norm 1.1292 (inf) [2021-04-15 22:28:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][800/1251] eta 0:02:10 lr 0.000818 time 0.2796 (0.2888) loss 4.1236 (3.7494) grad_norm 1.1773 (inf) [2021-04-15 22:28:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][810/1251] eta 0:02:07 lr 0.000818 time 0.2740 (0.2886) loss 2.8535 (3.7477) grad_norm 1.1025 (inf) [2021-04-15 22:28:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][820/1251] eta 0:02:04 lr 0.000818 time 0.2809 (0.2884) loss 3.4963 (3.7470) grad_norm 1.2413 (inf) [2021-04-15 22:29:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][830/1251] eta 0:02:01 lr 0.000818 time 0.2689 (0.2883) loss 4.5321 (3.7471) grad_norm 1.3514 (inf) [2021-04-15 22:29:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][840/1251] eta 0:01:58 lr 0.000818 time 0.2795 (0.2881) loss 4.1005 (3.7473) grad_norm 1.1854 (inf) [2021-04-15 22:29:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][850/1251] eta 0:01:55 lr 0.000818 time 0.2763 (0.2880) loss 3.5595 (3.7465) grad_norm 1.2622 (inf) [2021-04-15 22:29:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][860/1251] eta 0:01:52 lr 0.000818 time 0.2667 (0.2878) loss 3.6535 (3.7495) grad_norm 1.1703 (inf) [2021-04-15 22:29:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][870/1251] eta 0:01:49 lr 0.000818 time 0.2660 (0.2878) loss 4.0574 (3.7534) grad_norm 1.1930 (inf) [2021-04-15 22:29:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][880/1251] eta 0:01:46 lr 0.000818 time 0.2943 (0.2878) loss 3.1641 (3.7553) grad_norm 1.1990 (inf) [2021-04-15 22:29:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][890/1251] eta 0:01:43 lr 0.000818 time 0.2831 (0.2877) loss 3.3278 (3.7534) grad_norm 1.1782 (inf) [2021-04-15 22:29:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][900/1251] eta 0:01:40 lr 0.000818 time 0.3810 (0.2877) loss 3.2439 (3.7536) grad_norm 1.2327 (inf) [2021-04-15 22:29:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][910/1251] eta 0:01:38 lr 0.000818 time 0.2596 (0.2876) loss 3.2460 (3.7548) grad_norm 1.3845 (inf) [2021-04-15 22:29:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][920/1251] eta 0:01:35 lr 0.000818 time 0.2930 (0.2874) loss 4.3280 (3.7564) grad_norm 1.3912 (inf) [2021-04-15 22:29:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][930/1251] eta 0:01:32 lr 0.000818 time 0.2681 (0.2875) loss 4.1026 (3.7579) grad_norm 1.1384 (inf) [2021-04-15 22:29:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][940/1251] eta 0:01:29 lr 0.000818 time 0.2694 (0.2874) loss 4.6911 (3.7610) grad_norm 1.1383 (inf) [2021-04-15 22:29:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][950/1251] eta 0:01:26 lr 0.000817 time 0.2717 (0.2872) loss 2.9781 (3.7613) grad_norm 1.1800 (inf) [2021-04-15 22:29:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][960/1251] eta 0:01:23 lr 0.000817 time 0.2548 (0.2870) loss 3.0478 (3.7597) grad_norm 1.1512 (inf) [2021-04-15 22:29:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][970/1251] eta 0:01:20 lr 0.000817 time 0.2769 (0.2869) loss 4.0158 (3.7619) grad_norm 1.1458 (inf) [2021-04-15 22:29:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][980/1251] eta 0:01:17 lr 0.000817 time 0.2820 (0.2868) loss 3.7117 (3.7613) grad_norm 1.1586 (inf) [2021-04-15 22:29:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][990/1251] eta 0:01:14 lr 0.000817 time 0.2650 (0.2866) loss 4.0715 (3.7630) grad_norm 1.1307 (inf) [2021-04-15 22:29:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][1000/1251] eta 0:01:11 lr 0.000817 time 0.2657 (0.2866) loss 3.8993 (3.7614) grad_norm 1.5427 (inf) [2021-04-15 22:29:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][1010/1251] eta 0:01:09 lr 0.000817 time 0.2800 (0.2864) loss 4.0562 (3.7616) grad_norm 1.1242 (inf) [2021-04-15 22:29:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][1020/1251] eta 0:01:06 lr 0.000817 time 0.2781 (0.2863) loss 3.7205 (3.7631) grad_norm 1.2271 (inf) [2021-04-15 22:29:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][1030/1251] eta 0:01:03 lr 0.000817 time 0.2697 (0.2862) loss 4.2074 (3.7622) grad_norm 1.2325 (inf) [2021-04-15 22:30:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][1040/1251] eta 0:01:00 lr 0.000817 time 0.2664 (0.2861) loss 3.1846 (3.7638) grad_norm 1.2009 (inf) [2021-04-15 22:30:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][1050/1251] eta 0:00:57 lr 0.000817 time 0.2832 (0.2861) loss 4.2817 (3.7665) grad_norm 1.1685 (inf) [2021-04-15 22:30:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][1060/1251] eta 0:00:54 lr 0.000817 time 0.2947 (0.2861) loss 2.6718 (3.7650) grad_norm 1.1267 (inf) [2021-04-15 22:30:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][1070/1251] eta 0:00:51 lr 0.000817 time 0.2786 (0.2860) loss 4.0284 (3.7646) grad_norm 1.5629 (inf) [2021-04-15 22:30:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][1080/1251] eta 0:00:48 lr 0.000817 time 0.2813 (0.2859) loss 2.6160 (3.7639) grad_norm 1.1224 (inf) [2021-04-15 22:30:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][1090/1251] eta 0:00:46 lr 0.000817 time 0.2874 (0.2858) loss 4.1199 (3.7644) grad_norm 1.1719 (inf) [2021-04-15 22:30:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][1100/1251] eta 0:00:43 lr 0.000817 time 0.2880 (0.2857) loss 3.8701 (3.7650) grad_norm 1.1302 (inf) [2021-04-15 22:30:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][1110/1251] eta 0:00:40 lr 0.000817 time 0.2736 (0.2855) loss 3.7033 (3.7674) grad_norm 1.1924 (inf) [2021-04-15 22:30:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][1120/1251] eta 0:00:37 lr 0.000817 time 0.2755 (0.2855) loss 2.9480 (3.7634) grad_norm 1.1060 (inf) [2021-04-15 22:30:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][1130/1251] eta 0:00:34 lr 0.000817 time 0.2649 (0.2854) loss 4.7412 (3.7667) grad_norm 1.1987 (inf) [2021-04-15 22:30:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][1140/1251] eta 0:00:31 lr 0.000817 time 0.2808 (0.2855) loss 3.8625 (3.7663) grad_norm 1.1002 (inf) [2021-04-15 22:30:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][1150/1251] eta 0:00:28 lr 0.000817 time 0.2794 (0.2854) loss 2.9448 (3.7653) grad_norm 1.4048 (inf) [2021-04-15 22:30:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][1160/1251] eta 0:00:25 lr 0.000817 time 0.2668 (0.2854) loss 3.7572 (3.7649) grad_norm 1.1035 (inf) [2021-04-15 22:30:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][1170/1251] eta 0:00:23 lr 0.000817 time 0.2786 (0.2854) loss 3.9488 (3.7662) grad_norm 1.2811 (inf) [2021-04-15 22:30:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][1180/1251] eta 0:00:20 lr 0.000817 time 0.2979 (0.2853) loss 4.0595 (3.7682) grad_norm 1.2044 (inf) [2021-04-15 22:30:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][1190/1251] eta 0:00:17 lr 0.000817 time 0.2631 (0.2853) loss 3.4328 (3.7644) grad_norm 1.0838 (inf) [2021-04-15 22:30:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][1200/1251] eta 0:00:14 lr 0.000817 time 0.2691 (0.2852) loss 2.8249 (3.7648) grad_norm 1.4450 (inf) [2021-04-15 22:30:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][1210/1251] eta 0:00:11 lr 0.000817 time 0.2950 (0.2851) loss 4.4786 (3.7672) grad_norm 1.1022 (inf) [2021-04-15 22:30:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][1220/1251] eta 0:00:08 lr 0.000817 time 0.2845 (0.2850) loss 2.9155 (3.7672) grad_norm 1.2410 (inf) [2021-04-15 22:30:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][1230/1251] eta 0:00:05 lr 0.000817 time 0.2430 (0.2850) loss 3.6438 (3.7669) grad_norm 1.4431 (inf) [2021-04-15 22:30:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][1240/1251] eta 0:00:03 lr 0.000817 time 0.2595 (0.2849) loss 3.7051 (3.7678) grad_norm 1.4221 (inf) [2021-04-15 22:30:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [84/300][1250/1251] eta 0:00:00 lr 0.000817 time 0.2483 (0.2846) loss 3.6950 (3.7671) grad_norm 1.1621 (inf) [2021-04-15 22:31:01 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 84 training takes 0:05:58 [2021-04-15 22:31:01 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_84.pth saving...... [2021-04-15 22:31:25 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_84.pth saved !!! [2021-04-15 22:31:26 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.177 (1.177) Loss 1.2189 (1.2189) Acc@1 71.094 (71.094) Acc@5 90.820 (90.820) [2021-04-15 22:31:27 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.132 (0.199) Loss 1.1453 (1.1614) Acc@1 73.633 (72.843) Acc@5 92.188 (91.779) [2021-04-15 22:31:30 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.512 (0.229) Loss 1.1670 (1.1519) Acc@1 71.875 (73.005) Acc@5 91.016 (91.839) [2021-04-15 22:31:32 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.612 (0.237) Loss 1.2148 (1.1597) Acc@1 70.605 (72.933) Acc@5 90.820 (91.712) [2021-04-15 22:31:34 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.220) Loss 1.2596 (1.1622) Acc@1 70.801 (72.925) Acc@5 90.332 (91.518) [2021-04-15 22:31:37 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 72.882 Acc@5 91.596 [2021-04-15 22:31:37 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 72.9% [2021-04-15 22:31:37 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 72.88% [2021-04-15 22:31:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][0/1251] eta 1:51:03 lr 0.000817 time 5.3269 (5.3269) loss 3.9488 (3.9488) grad_norm 1.0872 (1.0872) [2021-04-15 22:31:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][10/1251] eta 0:15:29 lr 0.000816 time 0.4242 (0.7494) loss 3.4048 (3.9358) grad_norm 1.1661 (1.1984) [2021-04-15 22:31:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][20/1251] eta 0:10:44 lr 0.000816 time 0.2896 (0.5234) loss 3.7294 (3.8055) grad_norm 1.2272 (1.2196) [2021-04-15 22:31:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][30/1251] eta 0:09:00 lr 0.000816 time 0.2578 (0.4425) loss 3.5659 (3.7845) grad_norm 1.3452 (1.2068) [2021-04-15 22:31:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][40/1251] eta 0:08:07 lr 0.000816 time 0.2746 (0.4025) loss 3.9495 (3.8214) grad_norm 0.9877 (1.2017) [2021-04-15 22:31:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][50/1251] eta 0:07:34 lr 0.000816 time 0.2962 (0.3783) loss 3.6829 (3.7516) grad_norm 1.3460 (1.1993) [2021-04-15 22:31:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][60/1251] eta 0:07:10 lr 0.000816 time 0.2674 (0.3615) loss 4.0424 (3.7942) grad_norm 1.0630 (1.1962) [2021-04-15 22:32:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][70/1251] eta 0:06:53 lr 0.000816 time 0.2582 (0.3500) loss 2.8377 (3.7705) grad_norm 1.4072 (1.2042) [2021-04-15 22:32:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][80/1251] eta 0:06:40 lr 0.000816 time 0.2725 (0.3422) loss 2.7133 (3.7535) grad_norm 1.1551 (1.2107) [2021-04-15 22:32:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][90/1251] eta 0:06:30 lr 0.000816 time 0.2689 (0.3362) loss 3.9101 (3.7477) grad_norm 1.2557 (1.2049) [2021-04-15 22:32:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][100/1251] eta 0:06:20 lr 0.000816 time 0.2919 (0.3304) loss 3.0223 (3.7400) grad_norm 1.1026 (1.2041) [2021-04-15 22:32:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][110/1251] eta 0:06:12 lr 0.000816 time 0.4081 (0.3265) loss 4.0201 (3.7452) grad_norm 1.0433 (1.2072) [2021-04-15 22:32:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][120/1251] eta 0:06:04 lr 0.000816 time 0.2961 (0.3222) loss 3.2906 (3.7317) grad_norm 1.1265 (1.2074) [2021-04-15 22:32:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][130/1251] eta 0:05:58 lr 0.000816 time 0.2505 (0.3200) loss 4.4588 (3.7526) grad_norm 1.1333 (1.2104) [2021-04-15 22:32:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][140/1251] eta 0:05:52 lr 0.000816 time 0.2866 (0.3173) loss 4.1698 (3.7394) grad_norm 1.2672 (1.2054) [2021-04-15 22:32:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][150/1251] eta 0:05:46 lr 0.000816 time 0.2882 (0.3147) loss 3.5675 (3.7437) grad_norm 1.2417 (1.2094) [2021-04-15 22:32:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][160/1251] eta 0:05:40 lr 0.000816 time 0.2424 (0.3124) loss 3.9730 (3.7484) grad_norm 1.3326 (1.2195) [2021-04-15 22:32:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][170/1251] eta 0:05:36 lr 0.000816 time 0.2888 (0.3109) loss 3.0825 (3.7359) grad_norm 1.2709 (1.2272) [2021-04-15 22:32:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][180/1251] eta 0:05:31 lr 0.000816 time 0.2836 (0.3092) loss 3.9918 (3.7429) grad_norm 1.0677 (1.2357) [2021-04-15 22:32:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][190/1251] eta 0:05:26 lr 0.000816 time 0.2891 (0.3074) loss 4.3755 (3.7513) grad_norm 1.0108 (1.2294) [2021-04-15 22:32:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][200/1251] eta 0:05:21 lr 0.000816 time 0.2805 (0.3058) loss 3.1466 (3.7576) grad_norm 1.1360 (1.2280) [2021-04-15 22:32:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][210/1251] eta 0:05:16 lr 0.000816 time 0.2782 (0.3044) loss 4.0048 (3.7548) grad_norm 1.0982 (1.2234) [2021-04-15 22:32:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][220/1251] eta 0:05:12 lr 0.000816 time 0.2507 (0.3029) loss 4.5417 (3.7608) grad_norm 1.0712 (1.2229) [2021-04-15 22:32:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][230/1251] eta 0:05:08 lr 0.000816 time 0.2672 (0.3018) loss 4.1008 (3.7640) grad_norm 1.0551 (1.2212) [2021-04-15 22:32:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][240/1251] eta 0:05:04 lr 0.000816 time 0.2546 (0.3007) loss 4.1733 (3.7676) grad_norm 1.1209 (1.2197) [2021-04-15 22:32:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][250/1251] eta 0:04:59 lr 0.000816 time 0.2636 (0.2996) loss 3.9526 (3.7668) grad_norm 1.3222 (1.2195) [2021-04-15 22:32:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][260/1251] eta 0:04:55 lr 0.000816 time 0.2510 (0.2986) loss 3.9316 (3.7667) grad_norm 1.1867 (1.2196) [2021-04-15 22:32:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][270/1251] eta 0:04:51 lr 0.000816 time 0.2462 (0.2976) loss 2.4678 (3.7621) grad_norm 1.3188 (1.2207) [2021-04-15 22:33:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][280/1251] eta 0:04:48 lr 0.000816 time 0.2865 (0.2968) loss 4.1294 (3.7636) grad_norm 1.0914 (1.2194) [2021-04-15 22:33:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][290/1251] eta 0:04:44 lr 0.000816 time 0.2828 (0.2960) loss 3.9006 (3.7586) grad_norm 1.3506 (1.2188) [2021-04-15 22:33:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][300/1251] eta 0:04:40 lr 0.000816 time 0.2923 (0.2953) loss 4.1478 (3.7580) grad_norm 1.2537 (1.2156) [2021-04-15 22:33:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][310/1251] eta 0:04:37 lr 0.000816 time 0.2453 (0.2946) loss 3.6467 (3.7669) grad_norm 1.1436 (1.2148) [2021-04-15 22:33:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][320/1251] eta 0:04:34 lr 0.000815 time 0.2727 (0.2953) loss 3.5569 (3.7657) grad_norm 1.1552 (1.2140) [2021-04-15 22:33:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][330/1251] eta 0:04:31 lr 0.000815 time 0.2611 (0.2945) loss 4.4280 (3.7663) grad_norm 1.2236 (1.2144) [2021-04-15 22:33:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][340/1251] eta 0:04:27 lr 0.000815 time 0.2770 (0.2940) loss 3.6822 (3.7614) grad_norm 1.2191 (1.2153) [2021-04-15 22:33:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][350/1251] eta 0:04:24 lr 0.000815 time 0.2759 (0.2939) loss 3.8173 (3.7567) grad_norm 1.3068 (1.2190) [2021-04-15 22:33:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][360/1251] eta 0:04:21 lr 0.000815 time 0.2736 (0.2938) loss 3.5174 (3.7585) grad_norm 1.0800 (1.2178) [2021-04-15 22:33:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][370/1251] eta 0:04:18 lr 0.000815 time 0.2588 (0.2935) loss 2.6325 (3.7524) grad_norm 1.0928 (1.2161) [2021-04-15 22:33:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][380/1251] eta 0:04:15 lr 0.000815 time 0.2621 (0.2931) loss 3.7375 (3.7517) grad_norm 1.2962 (1.2166) [2021-04-15 22:33:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][390/1251] eta 0:04:12 lr 0.000815 time 0.2825 (0.2931) loss 4.4031 (3.7579) grad_norm 1.0653 (1.2156) [2021-04-15 22:33:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][400/1251] eta 0:04:09 lr 0.000815 time 0.2903 (0.2926) loss 3.9061 (3.7457) grad_norm 1.2169 (1.2159) [2021-04-15 22:33:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][410/1251] eta 0:04:05 lr 0.000815 time 0.2648 (0.2922) loss 4.0800 (3.7547) grad_norm 1.2909 (1.2141) [2021-04-15 22:33:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][420/1251] eta 0:04:02 lr 0.000815 time 0.2803 (0.2918) loss 4.8045 (3.7619) grad_norm 1.1730 (1.2156) [2021-04-15 22:33:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][430/1251] eta 0:03:59 lr 0.000815 time 0.2619 (0.2917) loss 3.8794 (3.7587) grad_norm 1.2789 (1.2163) [2021-04-15 22:33:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][440/1251] eta 0:03:56 lr 0.000815 time 0.2428 (0.2915) loss 3.7608 (3.7565) grad_norm 1.0409 (1.2152) [2021-04-15 22:33:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][450/1251] eta 0:03:53 lr 0.000815 time 0.2686 (0.2914) loss 2.7120 (3.7506) grad_norm 1.1332 (1.2135) [2021-04-15 22:33:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][460/1251] eta 0:03:50 lr 0.000815 time 0.2657 (0.2911) loss 3.6029 (3.7510) grad_norm 1.2713 (1.2158) [2021-04-15 22:33:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][470/1251] eta 0:03:47 lr 0.000815 time 0.2649 (0.2908) loss 4.3236 (3.7535) grad_norm 0.9369 (1.2141) [2021-04-15 22:33:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][480/1251] eta 0:03:43 lr 0.000815 time 0.2648 (0.2903) loss 4.0509 (3.7520) grad_norm 1.1363 (1.2133) [2021-04-15 22:34:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][490/1251] eta 0:03:40 lr 0.000815 time 0.2718 (0.2901) loss 3.1971 (3.7459) grad_norm 1.0765 (1.2123) [2021-04-15 22:34:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][500/1251] eta 0:03:37 lr 0.000815 time 0.2435 (0.2896) loss 3.5456 (3.7496) grad_norm 1.2478 (1.2138) [2021-04-15 22:34:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][510/1251] eta 0:03:34 lr 0.000815 time 0.2750 (0.2895) loss 3.3116 (3.7453) grad_norm 1.2307 (1.2130) [2021-04-15 22:34:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][520/1251] eta 0:03:31 lr 0.000815 time 0.2806 (0.2893) loss 4.1809 (3.7449) grad_norm 1.0993 (1.2136) [2021-04-15 22:34:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][530/1251] eta 0:03:28 lr 0.000815 time 0.3119 (0.2892) loss 2.6390 (3.7425) grad_norm 2.1134 (1.2166) [2021-04-15 22:34:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][540/1251] eta 0:03:25 lr 0.000815 time 0.3019 (0.2889) loss 3.8368 (3.7463) grad_norm 1.3581 (1.2188) [2021-04-15 22:34:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][550/1251] eta 0:03:22 lr 0.000815 time 0.4211 (0.2889) loss 2.9481 (3.7425) grad_norm 1.1971 (1.2217) [2021-04-15 22:34:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][560/1251] eta 0:03:19 lr 0.000815 time 0.2835 (0.2886) loss 3.9998 (3.7412) grad_norm 1.2416 (1.2219) [2021-04-15 22:34:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][570/1251] eta 0:03:16 lr 0.000815 time 0.2836 (0.2887) loss 3.6784 (3.7372) grad_norm 1.1621 (1.2218) [2021-04-15 22:34:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][580/1251] eta 0:03:13 lr 0.000815 time 0.2672 (0.2884) loss 4.2357 (3.7409) grad_norm 0.9901 (1.2201) [2021-04-15 22:34:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][590/1251] eta 0:03:10 lr 0.000815 time 0.2749 (0.2885) loss 3.9520 (3.7418) grad_norm 1.2857 (1.2202) [2021-04-15 22:34:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][600/1251] eta 0:03:07 lr 0.000815 time 0.2692 (0.2883) loss 4.0141 (3.7416) grad_norm 1.0726 (1.2194) [2021-04-15 22:34:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][610/1251] eta 0:03:04 lr 0.000815 time 0.2568 (0.2881) loss 4.3120 (3.7397) grad_norm 1.1356 (1.2188) [2021-04-15 22:34:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][620/1251] eta 0:03:01 lr 0.000815 time 0.2570 (0.2880) loss 3.6457 (3.7450) grad_norm 1.2597 (1.2194) [2021-04-15 22:34:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][630/1251] eta 0:02:58 lr 0.000814 time 0.2756 (0.2879) loss 4.0123 (3.7487) grad_norm 1.0477 (1.2178) [2021-04-15 22:34:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][640/1251] eta 0:02:55 lr 0.000814 time 0.3058 (0.2880) loss 3.0047 (3.7457) grad_norm 1.0383 (1.2176) [2021-04-15 22:34:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][650/1251] eta 0:02:53 lr 0.000814 time 0.2749 (0.2881) loss 3.9863 (3.7483) grad_norm 1.1208 (1.2161) [2021-04-15 22:34:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][660/1251] eta 0:02:50 lr 0.000814 time 0.2825 (0.2879) loss 4.0736 (3.7420) grad_norm 0.9947 (1.2154) [2021-04-15 22:34:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][670/1251] eta 0:02:47 lr 0.000814 time 0.2810 (0.2880) loss 4.2115 (3.7438) grad_norm 1.4075 (1.2144) [2021-04-15 22:34:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][680/1251] eta 0:02:44 lr 0.000814 time 0.2687 (0.2878) loss 4.3803 (3.7433) grad_norm 1.1563 (1.2139) [2021-04-15 22:34:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][690/1251] eta 0:02:41 lr 0.000814 time 0.2630 (0.2878) loss 3.8857 (3.7416) grad_norm 1.1170 (1.2143) [2021-04-15 22:34:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][700/1251] eta 0:02:38 lr 0.000814 time 0.3981 (0.2878) loss 3.4919 (3.7467) grad_norm 1.2131 (1.2139) [2021-04-15 22:35:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][710/1251] eta 0:02:35 lr 0.000814 time 0.2828 (0.2876) loss 4.1824 (3.7471) grad_norm 1.3947 (1.2140) [2021-04-15 22:35:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][720/1251] eta 0:02:32 lr 0.000814 time 0.2723 (0.2874) loss 3.1569 (3.7478) grad_norm 1.3380 (1.2148) [2021-04-15 22:35:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][730/1251] eta 0:02:29 lr 0.000814 time 0.2735 (0.2874) loss 2.6189 (3.7456) grad_norm 1.3739 (1.2149) [2021-04-15 22:35:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][740/1251] eta 0:02:26 lr 0.000814 time 0.2830 (0.2873) loss 3.3304 (3.7464) grad_norm 1.2509 (1.2154) [2021-04-15 22:35:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][750/1251] eta 0:02:23 lr 0.000814 time 0.3086 (0.2872) loss 4.1286 (3.7464) grad_norm 1.6153 (1.2165) [2021-04-15 22:35:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][760/1251] eta 0:02:20 lr 0.000814 time 0.2984 (0.2872) loss 2.8302 (3.7448) grad_norm 1.1283 (1.2158) [2021-04-15 22:35:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][770/1251] eta 0:02:18 lr 0.000814 time 0.2549 (0.2870) loss 4.3299 (3.7430) grad_norm 1.0175 (1.2149) [2021-04-15 22:35:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][780/1251] eta 0:02:15 lr 0.000814 time 0.2854 (0.2869) loss 3.0850 (3.7427) grad_norm 1.0990 (1.2149) [2021-04-15 22:35:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][790/1251] eta 0:02:12 lr 0.000814 time 0.2700 (0.2867) loss 4.2708 (3.7465) grad_norm 1.3268 (1.2151) [2021-04-15 22:35:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][800/1251] eta 0:02:09 lr 0.000814 time 0.2722 (0.2866) loss 4.3568 (3.7479) grad_norm 0.9871 (1.2150) [2021-04-15 22:35:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][810/1251] eta 0:02:06 lr 0.000814 time 0.2740 (0.2865) loss 4.2199 (3.7515) grad_norm 1.2297 (1.2150) [2021-04-15 22:35:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][820/1251] eta 0:02:03 lr 0.000814 time 0.2989 (0.2864) loss 3.9874 (3.7542) grad_norm 1.1338 (1.2147) [2021-04-15 22:35:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][830/1251] eta 0:02:00 lr 0.000814 time 0.2845 (0.2863) loss 4.4734 (3.7555) grad_norm 1.3217 (1.2144) [2021-04-15 22:35:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][840/1251] eta 0:01:57 lr 0.000814 time 0.3138 (0.2862) loss 4.4089 (3.7553) grad_norm 1.1713 (1.2147) [2021-04-15 22:35:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][850/1251] eta 0:01:54 lr 0.000814 time 0.2676 (0.2860) loss 3.0656 (3.7547) grad_norm 1.0482 (1.2148) [2021-04-15 22:35:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][860/1251] eta 0:01:51 lr 0.000814 time 0.2521 (0.2859) loss 3.7915 (3.7566) grad_norm 1.0171 (1.2151) [2021-04-15 22:35:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][870/1251] eta 0:01:48 lr 0.000814 time 0.3119 (0.2858) loss 2.7964 (3.7562) grad_norm 1.3727 (1.2145) [2021-04-15 22:35:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][880/1251] eta 0:01:46 lr 0.000814 time 0.2680 (0.2857) loss 4.1202 (3.7579) grad_norm 1.3093 (1.2137) [2021-04-15 22:35:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][890/1251] eta 0:01:43 lr 0.000814 time 0.2682 (0.2858) loss 3.8787 (3.7583) grad_norm 1.2660 (1.2140) [2021-04-15 22:35:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][900/1251] eta 0:01:40 lr 0.000814 time 0.2826 (0.2857) loss 4.2445 (3.7604) grad_norm 1.2011 (1.2135) [2021-04-15 22:35:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][910/1251] eta 0:01:37 lr 0.000814 time 0.2672 (0.2856) loss 3.5151 (3.7627) grad_norm 1.3571 (1.2142) [2021-04-15 22:36:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][920/1251] eta 0:01:34 lr 0.000814 time 0.3087 (0.2857) loss 4.3447 (3.7627) grad_norm 1.3825 (1.2143) [2021-04-15 22:36:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][930/1251] eta 0:01:31 lr 0.000814 time 0.2616 (0.2859) loss 4.4177 (3.7645) grad_norm 1.2857 (1.2139) [2021-04-15 22:36:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][940/1251] eta 0:01:28 lr 0.000813 time 0.2669 (0.2860) loss 2.6600 (3.7645) grad_norm 1.2039 (1.2134) [2021-04-15 22:36:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][950/1251] eta 0:01:26 lr 0.000813 time 0.3095 (0.2858) loss 4.3022 (3.7630) grad_norm 1.1287 (1.2130) [2021-04-15 22:36:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][960/1251] eta 0:01:23 lr 0.000813 time 0.2693 (0.2857) loss 3.4279 (3.7631) grad_norm 1.0752 (1.2129) [2021-04-15 22:36:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][970/1251] eta 0:01:20 lr 0.000813 time 0.2859 (0.2858) loss 4.1669 (3.7639) grad_norm 1.2891 (1.2120) [2021-04-15 22:36:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][980/1251] eta 0:01:17 lr 0.000813 time 0.2812 (0.2857) loss 4.1671 (3.7620) grad_norm 1.2609 (1.2126) [2021-04-15 22:36:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][990/1251] eta 0:01:14 lr 0.000813 time 0.2712 (0.2856) loss 4.1698 (3.7632) grad_norm 1.3157 (1.2125) [2021-04-15 22:36:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][1000/1251] eta 0:01:11 lr 0.000813 time 0.2521 (0.2855) loss 4.8811 (3.7624) grad_norm 1.1406 (1.2123) [2021-04-15 22:36:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][1010/1251] eta 0:01:08 lr 0.000813 time 0.2547 (0.2856) loss 3.0112 (3.7611) grad_norm 1.1783 (1.2118) [2021-04-15 22:36:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][1020/1251] eta 0:01:05 lr 0.000813 time 0.2642 (0.2855) loss 4.8158 (3.7655) grad_norm 1.2482 (1.2119) [2021-04-15 22:36:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][1030/1251] eta 0:01:03 lr 0.000813 time 0.2800 (0.2854) loss 3.7867 (3.7699) grad_norm 1.3892 (1.2117) [2021-04-15 22:36:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][1040/1251] eta 0:01:00 lr 0.000813 time 0.2716 (0.2853) loss 3.4729 (3.7674) grad_norm 1.1461 (1.2110) [2021-04-15 22:36:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][1050/1251] eta 0:00:57 lr 0.000813 time 0.2979 (0.2853) loss 3.7280 (3.7687) grad_norm 1.3799 (1.2109) [2021-04-15 22:36:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][1060/1251] eta 0:00:54 lr 0.000813 time 0.2705 (0.2852) loss 3.9637 (3.7707) grad_norm 1.4635 (1.2119) [2021-04-15 22:36:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][1070/1251] eta 0:00:51 lr 0.000813 time 0.2679 (0.2852) loss 3.2971 (3.7700) grad_norm 1.1158 (1.2133) [2021-04-15 22:36:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][1080/1251] eta 0:00:48 lr 0.000813 time 0.2762 (0.2851) loss 4.4874 (3.7719) grad_norm 1.2606 (1.2144) [2021-04-15 22:36:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][1090/1251] eta 0:00:45 lr 0.000813 time 0.2954 (0.2850) loss 3.6351 (3.7722) grad_norm 1.4479 (1.2151) [2021-04-15 22:36:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][1100/1251] eta 0:00:43 lr 0.000813 time 0.3268 (0.2849) loss 4.5325 (3.7725) grad_norm 1.2694 (1.2161) [2021-04-15 22:36:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][1110/1251] eta 0:00:40 lr 0.000813 time 0.2969 (0.2849) loss 3.9342 (3.7722) grad_norm 1.3102 (1.2164) [2021-04-15 22:36:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][1120/1251] eta 0:00:37 lr 0.000813 time 0.2703 (0.2848) loss 4.3093 (3.7740) grad_norm 1.1463 (1.2155) [2021-04-15 22:36:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][1130/1251] eta 0:00:34 lr 0.000813 time 0.2647 (0.2849) loss 4.6616 (3.7764) grad_norm 1.1630 (1.2161) [2021-04-15 22:37:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][1140/1251] eta 0:00:31 lr 0.000813 time 0.2610 (0.2850) loss 3.8945 (3.7752) grad_norm 1.1252 (1.2161) [2021-04-15 22:37:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][1150/1251] eta 0:00:28 lr 0.000813 time 0.2970 (0.2849) loss 4.1222 (3.7756) grad_norm 1.5333 (1.2166) [2021-04-15 22:37:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][1160/1251] eta 0:00:25 lr 0.000813 time 0.2671 (0.2848) loss 3.0866 (3.7769) grad_norm 1.1297 (1.2163) [2021-04-15 22:37:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][1170/1251] eta 0:00:23 lr 0.000813 time 0.2798 (0.2850) loss 3.6163 (3.7737) grad_norm 1.3209 (1.2168) [2021-04-15 22:37:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][1180/1251] eta 0:00:20 lr 0.000813 time 0.2750 (0.2849) loss 3.6153 (3.7733) grad_norm 1.2720 (1.2168) [2021-04-15 22:37:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][1190/1251] eta 0:00:17 lr 0.000813 time 0.2691 (0.2848) loss 3.0400 (3.7722) grad_norm 1.4070 (1.2171) [2021-04-15 22:37:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][1200/1251] eta 0:00:14 lr 0.000813 time 0.2683 (0.2847) loss 4.1522 (3.7720) grad_norm 1.1143 (1.2170) [2021-04-15 22:37:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][1210/1251] eta 0:00:11 lr 0.000813 time 0.2775 (0.2846) loss 4.3850 (3.7712) grad_norm 1.5690 (1.2173) [2021-04-15 22:37:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][1220/1251] eta 0:00:08 lr 0.000813 time 0.2580 (0.2846) loss 4.2704 (3.7705) grad_norm 1.2388 (1.2176) [2021-04-15 22:37:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][1230/1251] eta 0:00:05 lr 0.000813 time 0.2878 (0.2846) loss 4.0100 (3.7711) grad_norm 1.2059 (1.2174) [2021-04-15 22:37:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][1240/1251] eta 0:00:03 lr 0.000813 time 0.2498 (0.2845) loss 3.8824 (3.7696) grad_norm 1.1049 (1.2170) [2021-04-15 22:37:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [85/300][1250/1251] eta 0:00:00 lr 0.000812 time 0.2488 (0.2842) loss 3.9688 (3.7682) grad_norm 1.1135 (1.2164) [2021-04-15 22:37:36 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 85 training takes 0:05:58 [2021-04-15 22:37:36 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_85.pth saving...... [2021-04-15 22:38:00 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_85.pth saved !!! [2021-04-15 22:38:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.303 (1.303) Loss 1.1969 (1.1969) Acc@1 72.168 (72.168) Acc@5 91.016 (91.016) [2021-04-15 22:38:03 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.160 (0.247) Loss 1.0945 (1.1537) Acc@1 74.512 (72.701) Acc@5 91.406 (91.522) [2021-04-15 22:38:05 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.179 (0.238) Loss 1.1184 (1.1472) Acc@1 72.949 (72.926) Acc@5 91.699 (91.699) [2021-04-15 22:38:07 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.153 (0.224) Loss 1.2318 (1.1490) Acc@1 71.582 (73.019) Acc@5 90.234 (91.709) [2021-04-15 22:38:09 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.479 (0.222) Loss 1.1514 (1.1498) Acc@1 72.656 (73.059) Acc@5 90.234 (91.685) [2021-04-15 22:38:13 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 72.956 Acc@5 91.682 [2021-04-15 22:38:13 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 73.0% [2021-04-15 22:38:13 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 72.96% [2021-04-15 22:38:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][0/1251] eta 2:18:45 lr 0.000812 time 6.6553 (6.6553) loss 3.1560 (3.1560) grad_norm 1.2193 (1.2193) [2021-04-15 22:38:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][10/1251] eta 0:17:33 lr 0.000812 time 0.2661 (0.8486) loss 4.6473 (3.9983) grad_norm 1.2443 (1.2239) [2021-04-15 22:38:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][20/1251] eta 0:11:50 lr 0.000812 time 0.2864 (0.5772) loss 3.8071 (3.9254) grad_norm 1.2325 (1.2156) [2021-04-15 22:38:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][30/1251] eta 0:09:45 lr 0.000812 time 0.2532 (0.4795) loss 4.0039 (3.8251) grad_norm 1.2147 (1.1890) [2021-04-15 22:38:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][40/1251] eta 0:08:41 lr 0.000812 time 0.2673 (0.4307) loss 3.8215 (3.8226) grad_norm 1.2578 (1.1675) [2021-04-15 22:38:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][50/1251] eta 0:07:59 lr 0.000812 time 0.2672 (0.3996) loss 3.5188 (3.8117) grad_norm 1.1012 (1.1664) [2021-04-15 22:38:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][60/1251] eta 0:07:35 lr 0.000812 time 0.2705 (0.3822) loss 3.6874 (3.8055) grad_norm 1.1232 (1.1608) [2021-04-15 22:38:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][70/1251] eta 0:07:14 lr 0.000812 time 0.2891 (0.3676) loss 3.7530 (3.8211) grad_norm 1.4160 (1.1601) [2021-04-15 22:38:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][80/1251] eta 0:06:58 lr 0.000812 time 0.2642 (0.3571) loss 4.0408 (3.7829) grad_norm 1.2875 (1.1707) [2021-04-15 22:38:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][90/1251] eta 0:06:46 lr 0.000812 time 0.2785 (0.3503) loss 4.2705 (3.7927) grad_norm 1.1667 (1.1869) [2021-04-15 22:38:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][100/1251] eta 0:06:35 lr 0.000812 time 0.2925 (0.3436) loss 3.9718 (3.7708) grad_norm 1.3180 (1.1908) [2021-04-15 22:38:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][110/1251] eta 0:06:25 lr 0.000812 time 0.2953 (0.3383) loss 4.0925 (3.7848) grad_norm 1.2420 (1.1917) [2021-04-15 22:38:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][120/1251] eta 0:06:16 lr 0.000812 time 0.2927 (0.3332) loss 4.0830 (3.7869) grad_norm 1.4929 (1.2002) [2021-04-15 22:38:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][130/1251] eta 0:06:10 lr 0.000812 time 0.2885 (0.3303) loss 3.4822 (3.7663) grad_norm 1.0967 (1.1994) [2021-04-15 22:39:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][140/1251] eta 0:06:05 lr 0.000812 time 0.2671 (0.3286) loss 4.7343 (3.7758) grad_norm 1.1091 (1.2026) [2021-04-15 22:39:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][150/1251] eta 0:05:58 lr 0.000812 time 0.2793 (0.3252) loss 4.5118 (3.7801) grad_norm 1.0408 (1.2041) [2021-04-15 22:39:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][160/1251] eta 0:05:51 lr 0.000812 time 0.2926 (0.3222) loss 2.7829 (3.7636) grad_norm 1.1864 (1.2065) [2021-04-15 22:39:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][170/1251] eta 0:05:45 lr 0.000812 time 0.2749 (0.3199) loss 3.2966 (3.7675) grad_norm 1.0458 (1.2043) [2021-04-15 22:39:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][180/1251] eta 0:05:41 lr 0.000812 time 0.2775 (0.3185) loss 3.6513 (3.7781) grad_norm 1.2689 (1.2008) [2021-04-15 22:39:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][190/1251] eta 0:05:35 lr 0.000812 time 0.2410 (0.3161) loss 3.4974 (3.7710) grad_norm 1.1439 (1.2051) [2021-04-15 22:39:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][200/1251] eta 0:05:30 lr 0.000812 time 0.2695 (0.3144) loss 4.3444 (3.7519) grad_norm 1.0855 (1.2061) [2021-04-15 22:39:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][210/1251] eta 0:05:25 lr 0.000812 time 0.2813 (0.3126) loss 4.0524 (3.7591) grad_norm 1.3032 (1.2070) [2021-04-15 22:39:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][220/1251] eta 0:05:21 lr 0.000812 time 0.2594 (0.3116) loss 2.8788 (3.7541) grad_norm 1.1430 (1.2070) [2021-04-15 22:39:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][230/1251] eta 0:05:16 lr 0.000812 time 0.2689 (0.3101) loss 3.5927 (3.7503) grad_norm 1.3450 (1.2109) [2021-04-15 22:39:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][240/1251] eta 0:05:12 lr 0.000812 time 0.2718 (0.3087) loss 4.0437 (3.7609) grad_norm 1.3795 (1.2104) [2021-04-15 22:39:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][250/1251] eta 0:05:07 lr 0.000812 time 0.2672 (0.3073) loss 3.5290 (3.7404) grad_norm 1.1731 (1.2083) [2021-04-15 22:39:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][260/1251] eta 0:05:03 lr 0.000812 time 0.2848 (0.3062) loss 3.3814 (3.7443) grad_norm 1.1629 (1.2082) [2021-04-15 22:39:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][270/1251] eta 0:04:59 lr 0.000812 time 0.2805 (0.3052) loss 4.6697 (3.7483) grad_norm 1.2799 (1.2062) [2021-04-15 22:39:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][280/1251] eta 0:04:55 lr 0.000812 time 0.2727 (0.3043) loss 3.1384 (3.7355) grad_norm 1.0644 (1.2066) [2021-04-15 22:39:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][290/1251] eta 0:04:51 lr 0.000812 time 0.2688 (0.3034) loss 3.6697 (3.7403) grad_norm 1.2443 (1.2072) [2021-04-15 22:39:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][300/1251] eta 0:04:47 lr 0.000811 time 0.2618 (0.3025) loss 3.9507 (3.7347) grad_norm 1.0799 (1.2050) [2021-04-15 22:39:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][310/1251] eta 0:04:43 lr 0.000811 time 0.2685 (0.3016) loss 3.3468 (3.7350) grad_norm 1.2484 (1.2039) [2021-04-15 22:39:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][320/1251] eta 0:04:39 lr 0.000811 time 0.2564 (0.3006) loss 4.3571 (3.7412) grad_norm 1.0667 (1.2005) [2021-04-15 22:39:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][330/1251] eta 0:04:36 lr 0.000811 time 0.2897 (0.3000) loss 3.9972 (3.7374) grad_norm 1.1291 (1.1990) [2021-04-15 22:39:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][340/1251] eta 0:04:33 lr 0.000811 time 0.2620 (0.2998) loss 2.9346 (3.7307) grad_norm 1.1625 (1.1980) [2021-04-15 22:39:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][350/1251] eta 0:04:29 lr 0.000811 time 0.2684 (0.2991) loss 3.5960 (3.7146) grad_norm 1.4729 (1.1984) [2021-04-15 22:40:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][360/1251] eta 0:04:25 lr 0.000811 time 0.2673 (0.2984) loss 2.7686 (3.7117) grad_norm 1.2085 (1.1978) [2021-04-15 22:40:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][370/1251] eta 0:04:22 lr 0.000811 time 0.2547 (0.2982) loss 4.2517 (3.7161) grad_norm 1.2294 (1.1989) [2021-04-15 22:40:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][380/1251] eta 0:04:19 lr 0.000811 time 0.2829 (0.2978) loss 3.2202 (3.7149) grad_norm 0.9980 (1.1986) [2021-04-15 22:40:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][390/1251] eta 0:04:15 lr 0.000811 time 0.2659 (0.2972) loss 4.3358 (3.7188) grad_norm 0.9224 (1.1968) [2021-04-15 22:40:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][400/1251] eta 0:04:12 lr 0.000811 time 0.2782 (0.2965) loss 3.5014 (3.7149) grad_norm 1.0424 (1.1958) [2021-04-15 22:40:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][410/1251] eta 0:04:08 lr 0.000811 time 0.2808 (0.2960) loss 3.1393 (3.7107) grad_norm 1.2350 (1.1966) [2021-04-15 22:40:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][420/1251] eta 0:04:05 lr 0.000811 time 0.2820 (0.2955) loss 2.4800 (3.7086) grad_norm 1.1354 (1.2000) [2021-04-15 22:40:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][430/1251] eta 0:04:02 lr 0.000811 time 0.2477 (0.2954) loss 4.2789 (3.7071) grad_norm 1.2212 (1.2017) [2021-04-15 22:40:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][440/1251] eta 0:03:59 lr 0.000811 time 0.2841 (0.2953) loss 3.4738 (3.7069) grad_norm 1.1137 (1.2002) [2021-04-15 22:40:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][450/1251] eta 0:03:56 lr 0.000811 time 0.2703 (0.2948) loss 3.2398 (3.7090) grad_norm 1.1474 (1.2001) [2021-04-15 22:40:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][460/1251] eta 0:03:52 lr 0.000811 time 0.2774 (0.2943) loss 4.3074 (3.7151) grad_norm 1.1427 (1.2003) [2021-04-15 22:40:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][470/1251] eta 0:03:49 lr 0.000811 time 0.2709 (0.2940) loss 4.2125 (3.7170) grad_norm 1.1330 (1.2007) [2021-04-15 22:40:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][480/1251] eta 0:03:46 lr 0.000811 time 0.2626 (0.2937) loss 2.5441 (3.7184) grad_norm 1.2066 (1.2023) [2021-04-15 22:40:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][490/1251] eta 0:03:43 lr 0.000811 time 0.2942 (0.2934) loss 4.1714 (3.7221) grad_norm 1.0786 (1.2013) [2021-04-15 22:40:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][500/1251] eta 0:03:39 lr 0.000811 time 0.2727 (0.2929) loss 3.4672 (3.7189) grad_norm 1.2414 (1.2014) [2021-04-15 22:40:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][510/1251] eta 0:03:36 lr 0.000811 time 0.2739 (0.2926) loss 2.9612 (3.7163) grad_norm 1.2494 (1.2010) [2021-04-15 22:40:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][520/1251] eta 0:03:33 lr 0.000811 time 0.2713 (0.2924) loss 4.3766 (3.7202) grad_norm 1.4314 (1.2010) [2021-04-15 22:40:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][530/1251] eta 0:03:30 lr 0.000811 time 0.2877 (0.2921) loss 3.4498 (3.7164) grad_norm 1.5385 (1.2019) [2021-04-15 22:40:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][540/1251] eta 0:03:27 lr 0.000811 time 0.2701 (0.2918) loss 4.1036 (3.7162) grad_norm 1.0564 (1.2029) [2021-04-15 22:40:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][550/1251] eta 0:03:24 lr 0.000811 time 0.2820 (0.2916) loss 3.3232 (3.7159) grad_norm 1.3476 (1.2036) [2021-04-15 22:40:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][560/1251] eta 0:03:21 lr 0.000811 time 0.2777 (0.2914) loss 3.4332 (3.7127) grad_norm 1.2401 (1.2044) [2021-04-15 22:41:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][570/1251] eta 0:03:18 lr 0.000811 time 0.2552 (0.2914) loss 3.4886 (3.7139) grad_norm 1.2724 (1.2052) [2021-04-15 22:41:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][580/1251] eta 0:03:15 lr 0.000811 time 0.2771 (0.2912) loss 3.9630 (3.7097) grad_norm 1.5295 (1.2063) [2021-04-15 22:41:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][590/1251] eta 0:03:12 lr 0.000811 time 0.2749 (0.2908) loss 3.6945 (3.7136) grad_norm 1.1044 (1.2055) [2021-04-15 22:41:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][600/1251] eta 0:03:09 lr 0.000811 time 0.2964 (0.2906) loss 3.0015 (3.7118) grad_norm 1.3685 (1.2053) [2021-04-15 22:41:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][610/1251] eta 0:03:06 lr 0.000810 time 0.2753 (0.2903) loss 3.4917 (3.7189) grad_norm 1.1378 (1.2054) [2021-04-15 22:41:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][620/1251] eta 0:03:03 lr 0.000810 time 0.2689 (0.2901) loss 4.6520 (3.7209) grad_norm 1.3879 (1.2051) [2021-04-15 22:41:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][630/1251] eta 0:03:00 lr 0.000810 time 0.2623 (0.2901) loss 4.1211 (3.7233) grad_norm 1.0546 (1.2050) [2021-04-15 22:41:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][640/1251] eta 0:02:57 lr 0.000810 time 0.2686 (0.2898) loss 2.6472 (3.7219) grad_norm 1.0950 (1.2048) [2021-04-15 22:41:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][650/1251] eta 0:02:54 lr 0.000810 time 0.2937 (0.2896) loss 3.8728 (3.7233) grad_norm 1.1278 (1.2047) [2021-04-15 22:41:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][660/1251] eta 0:02:50 lr 0.000810 time 0.2695 (0.2893) loss 3.9370 (3.7257) grad_norm 1.1451 (1.2044) [2021-04-15 22:41:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][670/1251] eta 0:02:48 lr 0.000810 time 0.2700 (0.2894) loss 4.1986 (3.7288) grad_norm 1.2258 (1.2054) [2021-04-15 22:41:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][680/1251] eta 0:02:45 lr 0.000810 time 0.3084 (0.2893) loss 4.0257 (3.7307) grad_norm 1.3908 (1.2054) [2021-04-15 22:41:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][690/1251] eta 0:02:42 lr 0.000810 time 0.2898 (0.2891) loss 3.7318 (3.7326) grad_norm 1.0568 (1.2056) [2021-04-15 22:41:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][700/1251] eta 0:02:39 lr 0.000810 time 0.2605 (0.2889) loss 3.7208 (3.7328) grad_norm 1.3159 (1.2072) [2021-04-15 22:41:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][710/1251] eta 0:02:36 lr 0.000810 time 0.2742 (0.2890) loss 4.2368 (3.7301) grad_norm 1.1934 (1.2069) [2021-04-15 22:41:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][720/1251] eta 0:02:33 lr 0.000810 time 0.2757 (0.2891) loss 3.9673 (3.7274) grad_norm 1.2998 (1.2082) [2021-04-15 22:41:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][730/1251] eta 0:02:30 lr 0.000810 time 0.2739 (0.2889) loss 4.0880 (3.7275) grad_norm 1.1588 (1.2074) [2021-04-15 22:41:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][740/1251] eta 0:02:27 lr 0.000810 time 0.2796 (0.2888) loss 3.5925 (3.7291) grad_norm 1.1369 (1.2073) [2021-04-15 22:41:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][750/1251] eta 0:02:24 lr 0.000810 time 0.2731 (0.2887) loss 4.6940 (3.7274) grad_norm 1.2937 (1.2084) [2021-04-15 22:41:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][760/1251] eta 0:02:21 lr 0.000810 time 0.2639 (0.2886) loss 4.3131 (3.7298) grad_norm 1.1438 (1.2087) [2021-04-15 22:41:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][770/1251] eta 0:02:18 lr 0.000810 time 0.2779 (0.2884) loss 3.8935 (3.7328) grad_norm 1.1816 (1.2082) [2021-04-15 22:41:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][780/1251] eta 0:02:15 lr 0.000810 time 0.2626 (0.2883) loss 4.0478 (3.7362) grad_norm 1.3544 (1.2080) [2021-04-15 22:42:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][790/1251] eta 0:02:12 lr 0.000810 time 0.2480 (0.2883) loss 3.8413 (3.7388) grad_norm 1.1261 (1.2078) [2021-04-15 22:42:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][800/1251] eta 0:02:09 lr 0.000810 time 0.2734 (0.2881) loss 4.8860 (3.7441) grad_norm 1.1795 (1.2079) [2021-04-15 22:42:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][810/1251] eta 0:02:06 lr 0.000810 time 0.2879 (0.2880) loss 3.2660 (3.7409) grad_norm 1.4021 (1.2078) [2021-04-15 22:42:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][820/1251] eta 0:02:04 lr 0.000810 time 0.2669 (0.2878) loss 3.2024 (3.7404) grad_norm 1.0341 (nan) [2021-04-15 22:42:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][830/1251] eta 0:02:01 lr 0.000810 time 0.2527 (0.2876) loss 4.4649 (3.7454) grad_norm 1.1825 (nan) [2021-04-15 22:42:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][840/1251] eta 0:01:58 lr 0.000810 time 0.2686 (0.2875) loss 3.9433 (3.7459) grad_norm 1.1398 (nan) [2021-04-15 22:42:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][850/1251] eta 0:01:55 lr 0.000810 time 0.2744 (0.2873) loss 3.3431 (3.7450) grad_norm 1.1284 (nan) [2021-04-15 22:42:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][860/1251] eta 0:01:52 lr 0.000810 time 0.2681 (0.2872) loss 3.5015 (3.7445) grad_norm 1.2189 (nan) [2021-04-15 22:42:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][870/1251] eta 0:01:49 lr 0.000810 time 0.2906 (0.2871) loss 2.5959 (3.7411) grad_norm 1.1091 (nan) [2021-04-15 22:42:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][880/1251] eta 0:01:46 lr 0.000810 time 0.2666 (0.2870) loss 3.7960 (3.7401) grad_norm 1.1876 (nan) [2021-04-15 22:42:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][890/1251] eta 0:01:43 lr 0.000810 time 0.2494 (0.2868) loss 3.2191 (3.7374) grad_norm 1.0592 (nan) [2021-04-15 22:42:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][900/1251] eta 0:01:40 lr 0.000810 time 0.2924 (0.2867) loss 2.9983 (3.7396) grad_norm 1.0983 (nan) [2021-04-15 22:42:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][910/1251] eta 0:01:37 lr 0.000810 time 0.2639 (0.2865) loss 3.6809 (3.7395) grad_norm 1.1536 (nan) [2021-04-15 22:42:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][920/1251] eta 0:01:34 lr 0.000809 time 0.2908 (0.2865) loss 3.2760 (3.7371) grad_norm 1.5686 (nan) [2021-04-15 22:42:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][930/1251] eta 0:01:31 lr 0.000809 time 0.2663 (0.2864) loss 3.9724 (3.7379) grad_norm 1.1504 (nan) [2021-04-15 22:42:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][940/1251] eta 0:01:29 lr 0.000809 time 0.2659 (0.2864) loss 4.3344 (3.7382) grad_norm 1.1789 (nan) [2021-04-15 22:42:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][950/1251] eta 0:01:26 lr 0.000809 time 0.2819 (0.2864) loss 2.8253 (3.7414) grad_norm 1.2699 (nan) [2021-04-15 22:42:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][960/1251] eta 0:01:23 lr 0.000809 time 0.2978 (0.2863) loss 3.8321 (3.7436) grad_norm 1.1285 (nan) [2021-04-15 22:42:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][970/1251] eta 0:01:20 lr 0.000809 time 0.2830 (0.2862) loss 4.2177 (3.7434) grad_norm 1.2061 (nan) [2021-04-15 22:42:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][980/1251] eta 0:01:17 lr 0.000809 time 0.2689 (0.2860) loss 4.0996 (3.7448) grad_norm 1.1185 (nan) [2021-04-15 22:42:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][990/1251] eta 0:01:14 lr 0.000809 time 0.2565 (0.2859) loss 3.7250 (3.7473) grad_norm 1.2541 (nan) [2021-04-15 22:42:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][1000/1251] eta 0:01:11 lr 0.000809 time 0.2736 (0.2858) loss 4.1380 (3.7464) grad_norm 1.0486 (nan) [2021-04-15 22:43:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][1010/1251] eta 0:01:08 lr 0.000809 time 0.3088 (0.2859) loss 4.1029 (3.7485) grad_norm 1.1540 (nan) [2021-04-15 22:43:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][1020/1251] eta 0:01:06 lr 0.000809 time 0.2736 (0.2859) loss 3.2078 (3.7490) grad_norm 1.3511 (nan) [2021-04-15 22:43:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][1030/1251] eta 0:01:03 lr 0.000809 time 0.2730 (0.2859) loss 3.3705 (3.7463) grad_norm 1.2805 (nan) [2021-04-15 22:43:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][1040/1251] eta 0:01:00 lr 0.000809 time 0.2904 (0.2857) loss 2.6281 (3.7449) grad_norm 1.1810 (nan) [2021-04-15 22:43:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][1050/1251] eta 0:00:57 lr 0.000809 time 0.2701 (0.2856) loss 3.8422 (3.7436) grad_norm 1.1073 (nan) [2021-04-15 22:43:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][1060/1251] eta 0:00:54 lr 0.000809 time 0.2560 (0.2856) loss 3.0525 (3.7440) grad_norm 1.0322 (nan) [2021-04-15 22:43:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][1070/1251] eta 0:00:51 lr 0.000809 time 0.2820 (0.2855) loss 4.5041 (3.7448) grad_norm 1.3404 (nan) [2021-04-15 22:43:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][1080/1251] eta 0:00:48 lr 0.000809 time 0.3115 (0.2855) loss 3.1107 (3.7454) grad_norm 1.1438 (nan) [2021-04-15 22:43:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][1090/1251] eta 0:00:45 lr 0.000809 time 0.2715 (0.2853) loss 4.2337 (3.7476) grad_norm 1.1274 (nan) [2021-04-15 22:43:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][1100/1251] eta 0:00:43 lr 0.000809 time 0.2767 (0.2853) loss 3.3499 (3.7454) grad_norm 1.1633 (nan) [2021-04-15 22:43:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][1110/1251] eta 0:00:40 lr 0.000809 time 0.2777 (0.2852) loss 3.7234 (3.7428) grad_norm 1.0346 (nan) [2021-04-15 22:43:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][1120/1251] eta 0:00:37 lr 0.000809 time 0.2676 (0.2851) loss 4.4041 (3.7441) grad_norm 1.0860 (nan) [2021-04-15 22:43:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][1130/1251] eta 0:00:34 lr 0.000809 time 0.2655 (0.2850) loss 4.0975 (3.7467) grad_norm 1.4261 (nan) [2021-04-15 22:43:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][1140/1251] eta 0:00:31 lr 0.000809 time 0.2748 (0.2850) loss 3.1181 (3.7466) grad_norm 1.4519 (nan) [2021-04-15 22:43:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][1150/1251] eta 0:00:28 lr 0.000809 time 0.3003 (0.2851) loss 4.3656 (3.7488) grad_norm 1.2345 (nan) [2021-04-15 22:43:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][1160/1251] eta 0:00:25 lr 0.000809 time 0.2570 (0.2850) loss 2.8021 (3.7493) grad_norm 1.0771 (nan) [2021-04-15 22:43:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][1170/1251] eta 0:00:23 lr 0.000809 time 0.2632 (0.2849) loss 3.9587 (3.7487) grad_norm 1.1489 (nan) [2021-04-15 22:43:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][1180/1251] eta 0:00:20 lr 0.000809 time 0.2645 (0.2849) loss 3.6980 (3.7466) grad_norm 1.1848 (nan) [2021-04-15 22:43:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][1190/1251] eta 0:00:17 lr 0.000809 time 0.2813 (0.2848) loss 3.1616 (3.7435) grad_norm 1.0953 (nan) [2021-04-15 22:43:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][1200/1251] eta 0:00:14 lr 0.000809 time 0.3190 (0.2848) loss 3.8388 (3.7437) grad_norm 1.2135 (nan) [2021-04-15 22:43:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][1210/1251] eta 0:00:11 lr 0.000809 time 0.2761 (0.2847) loss 4.0808 (3.7437) grad_norm 1.1788 (nan) [2021-04-15 22:44:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][1220/1251] eta 0:00:08 lr 0.000808 time 0.2634 (0.2847) loss 3.9456 (3.7428) grad_norm 1.5953 (nan) [2021-04-15 22:44:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][1230/1251] eta 0:00:05 lr 0.000808 time 0.2814 (0.2846) loss 4.0968 (3.7422) grad_norm 1.3323 (nan) [2021-04-15 22:44:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][1240/1251] eta 0:00:03 lr 0.000808 time 0.3251 (0.2845) loss 4.0727 (3.7420) grad_norm 1.1227 (nan) [2021-04-15 22:44:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [86/300][1250/1251] eta 0:00:00 lr 0.000808 time 0.2486 (0.2842) loss 2.2261 (3.7421) grad_norm 1.3441 (nan) [2021-04-15 22:44:12 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 86 training takes 0:05:58 [2021-04-15 22:44:12 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_86.pth saving...... [2021-04-15 22:44:36 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_86.pth saved !!! [2021-04-15 22:44:38 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.209 (1.209) Loss 1.1927 (1.1927) Acc@1 70.508 (70.508) Acc@5 91.699 (91.699) [2021-04-15 22:44:39 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.092 (0.204) Loss 1.1505 (1.1718) Acc@1 73.633 (72.514) Acc@5 92.090 (91.841) [2021-04-15 22:44:41 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.173 (0.227) Loss 1.1114 (1.1737) Acc@1 74.023 (72.638) Acc@5 92.480 (91.769) [2021-04-15 22:44:44 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.106 (0.240) Loss 1.1649 (1.1747) Acc@1 73.633 (72.779) Acc@5 92.090 (91.724) [2021-04-15 22:44:46 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.223) Loss 1.1365 (1.1745) Acc@1 73.340 (72.775) Acc@5 91.602 (91.697) [2021-04-15 22:44:50 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 72.618 Acc@5 91.646 [2021-04-15 22:44:50 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 72.6% [2021-04-15 22:44:50 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 72.96% [2021-04-15 22:44:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][0/1251] eta 1:46:17 lr 0.000808 time 5.0980 (5.0980) loss 3.6629 (3.6629) grad_norm 1.1236 (1.1236) [2021-04-15 22:44:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][10/1251] eta 0:14:47 lr 0.000808 time 0.2803 (0.7152) loss 3.7211 (3.8295) grad_norm 1.1592 (1.2418) [2021-04-15 22:45:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][20/1251] eta 0:10:28 lr 0.000808 time 0.2725 (0.5109) loss 4.7558 (3.8093) grad_norm 1.0749 (1.2178) [2021-04-15 22:45:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][30/1251] eta 0:08:54 lr 0.000808 time 0.2919 (0.4374) loss 4.2500 (3.7975) grad_norm 0.9640 (1.1978) [2021-04-15 22:45:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][40/1251] eta 0:08:02 lr 0.000808 time 0.2755 (0.3983) loss 3.8471 (3.8312) grad_norm 1.3350 (1.2038) [2021-04-15 22:45:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][50/1251] eta 0:07:29 lr 0.000808 time 0.2822 (0.3743) loss 3.2247 (3.8337) grad_norm 1.1863 (1.2142) [2021-04-15 22:45:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][60/1251] eta 0:07:06 lr 0.000808 time 0.2605 (0.3580) loss 4.3768 (3.8281) grad_norm 1.1143 (1.2071) [2021-04-15 22:45:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][70/1251] eta 0:06:48 lr 0.000808 time 0.2605 (0.3462) loss 3.6907 (3.7988) grad_norm 1.1582 (1.2009) [2021-04-15 22:45:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][80/1251] eta 0:06:34 lr 0.000808 time 0.2647 (0.3370) loss 3.8905 (3.7614) grad_norm 1.7056 (1.2040) [2021-04-15 22:45:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][90/1251] eta 0:06:24 lr 0.000808 time 0.3012 (0.3308) loss 3.2411 (3.7046) grad_norm 1.1968 (1.2081) [2021-04-15 22:45:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][100/1251] eta 0:06:13 lr 0.000808 time 0.2481 (0.3249) loss 3.1617 (3.7130) grad_norm 1.4406 (1.2138) [2021-04-15 22:45:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][110/1251] eta 0:06:05 lr 0.000808 time 0.2484 (0.3201) loss 4.1577 (3.7396) grad_norm 1.1817 (1.2097) [2021-04-15 22:45:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][120/1251] eta 0:05:59 lr 0.000808 time 0.2815 (0.3180) loss 3.8012 (3.7442) grad_norm 1.1998 (1.2046) [2021-04-15 22:45:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][130/1251] eta 0:05:52 lr 0.000808 time 0.2425 (0.3146) loss 4.4369 (3.7318) grad_norm 1.2485 (1.2075) [2021-04-15 22:45:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][140/1251] eta 0:05:47 lr 0.000808 time 0.2822 (0.3131) loss 3.5371 (3.7405) grad_norm 1.3642 (1.2101) [2021-04-15 22:45:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][150/1251] eta 0:05:42 lr 0.000808 time 0.2715 (0.3109) loss 3.3249 (3.7153) grad_norm 1.0455 (1.2023) [2021-04-15 22:45:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][160/1251] eta 0:05:36 lr 0.000808 time 0.2910 (0.3087) loss 3.7425 (3.7269) grad_norm 1.1706 (1.2037) [2021-04-15 22:45:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][170/1251] eta 0:05:31 lr 0.000808 time 0.2652 (0.3070) loss 3.7973 (3.7214) grad_norm 1.2442 (1.2072) [2021-04-15 22:45:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][180/1251] eta 0:05:27 lr 0.000808 time 0.2956 (0.3054) loss 3.1652 (3.7291) grad_norm 1.1544 (1.2034) [2021-04-15 22:45:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][190/1251] eta 0:05:22 lr 0.000808 time 0.2613 (0.3041) loss 3.5807 (3.7251) grad_norm 1.2207 (1.2053) [2021-04-15 22:45:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][200/1251] eta 0:05:18 lr 0.000808 time 0.2855 (0.3029) loss 4.2531 (3.7247) grad_norm 1.2001 (1.2021) [2021-04-15 22:45:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][210/1251] eta 0:05:14 lr 0.000808 time 0.2982 (0.3017) loss 3.9693 (3.7330) grad_norm 1.0215 (1.2028) [2021-04-15 22:45:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][220/1251] eta 0:05:10 lr 0.000808 time 0.2682 (0.3009) loss 3.3230 (3.7272) grad_norm 1.0994 (1.2009) [2021-04-15 22:46:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][230/1251] eta 0:05:07 lr 0.000808 time 0.2884 (0.3009) loss 4.2485 (3.7192) grad_norm 1.3418 (1.2010) [2021-04-15 22:46:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][240/1251] eta 0:05:03 lr 0.000808 time 0.2762 (0.2998) loss 2.3333 (3.7063) grad_norm 1.0866 (1.1997) [2021-04-15 22:46:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][250/1251] eta 0:04:59 lr 0.000808 time 0.2671 (0.2989) loss 3.6558 (3.7084) grad_norm 1.0901 (1.1972) [2021-04-15 22:46:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][260/1251] eta 0:04:55 lr 0.000808 time 0.3037 (0.2981) loss 3.8270 (3.7092) grad_norm 1.3985 (1.1971) [2021-04-15 22:46:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][270/1251] eta 0:04:51 lr 0.000808 time 0.2406 (0.2974) loss 4.0666 (3.6975) grad_norm 1.1780 (1.1994) [2021-04-15 22:46:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][280/1251] eta 0:04:48 lr 0.000807 time 0.2929 (0.2967) loss 3.2478 (3.7064) grad_norm 1.1603 (1.1974) [2021-04-15 22:46:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][290/1251] eta 0:04:44 lr 0.000807 time 0.2827 (0.2960) loss 3.5836 (3.6986) grad_norm 1.4087 (1.2025) [2021-04-15 22:46:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][300/1251] eta 0:04:40 lr 0.000807 time 0.2821 (0.2954) loss 4.3182 (3.7123) grad_norm 1.2353 (1.2043) [2021-04-15 22:46:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][310/1251] eta 0:04:37 lr 0.000807 time 0.2623 (0.2949) loss 3.4155 (3.7131) grad_norm 1.3696 (1.2033) [2021-04-15 22:46:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][320/1251] eta 0:04:34 lr 0.000807 time 0.2778 (0.2945) loss 3.6829 (3.7144) grad_norm 1.1768 (1.2025) [2021-04-15 22:46:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][330/1251] eta 0:04:30 lr 0.000807 time 0.2559 (0.2939) loss 4.1782 (3.7166) grad_norm 1.6838 (1.2041) [2021-04-15 22:46:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][340/1251] eta 0:04:27 lr 0.000807 time 0.2758 (0.2935) loss 4.2482 (3.7283) grad_norm 1.0996 (1.2058) [2021-04-15 22:46:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][350/1251] eta 0:04:24 lr 0.000807 time 0.2702 (0.2930) loss 4.2091 (3.7330) grad_norm 1.4744 (1.2071) [2021-04-15 22:46:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][360/1251] eta 0:04:20 lr 0.000807 time 0.2660 (0.2929) loss 4.0212 (3.7369) grad_norm 1.0781 (1.2066) [2021-04-15 22:46:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][370/1251] eta 0:04:18 lr 0.000807 time 0.2768 (0.2931) loss 4.0748 (3.7361) grad_norm 1.1679 (1.2087) [2021-04-15 22:46:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][380/1251] eta 0:04:14 lr 0.000807 time 0.2729 (0.2926) loss 3.4985 (3.7381) grad_norm 1.2572 (1.2080) [2021-04-15 22:46:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][390/1251] eta 0:04:11 lr 0.000807 time 0.2748 (0.2923) loss 4.1035 (3.7308) grad_norm 1.2494 (1.2073) [2021-04-15 22:46:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][400/1251] eta 0:04:08 lr 0.000807 time 0.2807 (0.2920) loss 4.4741 (3.7336) grad_norm 1.2593 (1.2069) [2021-04-15 22:46:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][410/1251] eta 0:04:05 lr 0.000807 time 0.2690 (0.2915) loss 2.5968 (3.7360) grad_norm 1.2458 (1.2078) [2021-04-15 22:46:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][420/1251] eta 0:04:01 lr 0.000807 time 0.2510 (0.2911) loss 4.0306 (3.7402) grad_norm 1.3287 (1.2084) [2021-04-15 22:46:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][430/1251] eta 0:03:58 lr 0.000807 time 0.2679 (0.2909) loss 4.3121 (3.7398) grad_norm 1.2958 (1.2091) [2021-04-15 22:46:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][440/1251] eta 0:03:55 lr 0.000807 time 0.2750 (0.2907) loss 4.0589 (3.7386) grad_norm 0.9869 (1.2083) [2021-04-15 22:47:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][450/1251] eta 0:03:52 lr 0.000807 time 0.3771 (0.2906) loss 2.8673 (3.7314) grad_norm 1.2167 (1.2084) [2021-04-15 22:47:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][460/1251] eta 0:03:49 lr 0.000807 time 0.2689 (0.2902) loss 3.6675 (3.7362) grad_norm 1.0407 (1.2090) [2021-04-15 22:47:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][470/1251] eta 0:03:46 lr 0.000807 time 0.3157 (0.2900) loss 4.3554 (3.7370) grad_norm 1.4609 (1.2112) [2021-04-15 22:47:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][480/1251] eta 0:03:43 lr 0.000807 time 0.2539 (0.2897) loss 3.9777 (3.7414) grad_norm 1.1714 (1.2127) [2021-04-15 22:47:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][490/1251] eta 0:03:40 lr 0.000807 time 0.2632 (0.2895) loss 2.7276 (3.7391) grad_norm 1.1473 (1.2125) [2021-04-15 22:47:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][500/1251] eta 0:03:37 lr 0.000807 time 0.2603 (0.2893) loss 3.1391 (3.7347) grad_norm 1.6362 (1.2155) [2021-04-15 22:47:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][510/1251] eta 0:03:34 lr 0.000807 time 0.2765 (0.2892) loss 4.2339 (3.7320) grad_norm 1.2535 (1.2160) [2021-04-15 22:47:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][520/1251] eta 0:03:31 lr 0.000807 time 0.2658 (0.2892) loss 3.9960 (3.7301) grad_norm 1.0991 (1.2149) [2021-04-15 22:47:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][530/1251] eta 0:03:28 lr 0.000807 time 0.2590 (0.2891) loss 4.2957 (3.7314) grad_norm 1.1150 (1.2136) [2021-04-15 22:47:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][540/1251] eta 0:03:25 lr 0.000807 time 0.2507 (0.2891) loss 4.2360 (3.7311) grad_norm 1.0796 (1.2126) [2021-04-15 22:47:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][550/1251] eta 0:03:22 lr 0.000807 time 0.2694 (0.2891) loss 2.8749 (3.7267) grad_norm 1.1535 (1.2140) [2021-04-15 22:47:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][560/1251] eta 0:03:19 lr 0.000807 time 0.2859 (0.2890) loss 3.7995 (3.7298) grad_norm 1.3025 (1.2154) [2021-04-15 22:47:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][570/1251] eta 0:03:16 lr 0.000807 time 0.2766 (0.2887) loss 3.0689 (3.7245) grad_norm 1.2293 (1.2160) [2021-04-15 22:47:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][580/1251] eta 0:03:13 lr 0.000806 time 0.2707 (0.2887) loss 3.8484 (3.7262) grad_norm 1.2328 (1.2164) [2021-04-15 22:47:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][590/1251] eta 0:03:10 lr 0.000806 time 0.2891 (0.2888) loss 3.0177 (3.7238) grad_norm 1.2034 (1.2152) [2021-04-15 22:47:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][600/1251] eta 0:03:07 lr 0.000806 time 0.2660 (0.2885) loss 3.8163 (3.7190) grad_norm 1.1835 (1.2150) [2021-04-15 22:47:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][610/1251] eta 0:03:04 lr 0.000806 time 0.2874 (0.2884) loss 4.3930 (3.7202) grad_norm 1.1188 (1.2143) [2021-04-15 22:47:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][620/1251] eta 0:03:01 lr 0.000806 time 0.2696 (0.2882) loss 4.3375 (3.7219) grad_norm 1.1985 (1.2135) [2021-04-15 22:47:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][630/1251] eta 0:02:58 lr 0.000806 time 0.2741 (0.2881) loss 2.8465 (3.7263) grad_norm 1.2092 (1.2126) [2021-04-15 22:47:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][640/1251] eta 0:02:55 lr 0.000806 time 0.2742 (0.2879) loss 3.8797 (3.7308) grad_norm 1.1460 (1.2136) [2021-04-15 22:47:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][650/1251] eta 0:02:52 lr 0.000806 time 0.2877 (0.2878) loss 4.2876 (3.7343) grad_norm 1.1998 (1.2130) [2021-04-15 22:48:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][660/1251] eta 0:02:49 lr 0.000806 time 0.2815 (0.2876) loss 2.7523 (3.7343) grad_norm 1.4351 (1.2135) [2021-04-15 22:48:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][670/1251] eta 0:02:47 lr 0.000806 time 0.3128 (0.2875) loss 4.4653 (3.7327) grad_norm 1.0335 (1.2142) [2021-04-15 22:48:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][680/1251] eta 0:02:44 lr 0.000806 time 0.2818 (0.2873) loss 4.0432 (3.7304) grad_norm 1.1785 (1.2136) [2021-04-15 22:48:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][690/1251] eta 0:02:41 lr 0.000806 time 0.2913 (0.2873) loss 4.2892 (3.7334) grad_norm 1.1835 (1.2142) [2021-04-15 22:48:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][700/1251] eta 0:02:38 lr 0.000806 time 0.2516 (0.2871) loss 4.2757 (3.7314) grad_norm 1.1180 (1.2146) [2021-04-15 22:48:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][710/1251] eta 0:02:35 lr 0.000806 time 0.2763 (0.2869) loss 4.1101 (3.7336) grad_norm 1.1962 (1.2158) [2021-04-15 22:48:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][720/1251] eta 0:02:32 lr 0.000806 time 0.2672 (0.2871) loss 3.8391 (3.7364) grad_norm 1.1666 (1.2144) [2021-04-15 22:48:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][730/1251] eta 0:02:29 lr 0.000806 time 0.2770 (0.2869) loss 4.3197 (3.7337) grad_norm 1.1577 (1.2134) [2021-04-15 22:48:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][740/1251] eta 0:02:26 lr 0.000806 time 0.2775 (0.2867) loss 2.6472 (3.7326) grad_norm 1.4813 (1.2132) [2021-04-15 22:48:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][750/1251] eta 0:02:23 lr 0.000806 time 0.2959 (0.2866) loss 4.0031 (3.7316) grad_norm 1.4893 (1.2148) [2021-04-15 22:48:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][760/1251] eta 0:02:20 lr 0.000806 time 0.2849 (0.2864) loss 3.9144 (3.7324) grad_norm 1.2026 (1.2154) [2021-04-15 22:48:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][770/1251] eta 0:02:17 lr 0.000806 time 0.2587 (0.2862) loss 3.0401 (3.7308) grad_norm 1.2395 (1.2156) [2021-04-15 22:48:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][780/1251] eta 0:02:14 lr 0.000806 time 0.2704 (0.2862) loss 3.3226 (3.7326) grad_norm 1.3777 (1.2171) [2021-04-15 22:48:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][790/1251] eta 0:02:11 lr 0.000806 time 0.2499 (0.2862) loss 4.0030 (3.7379) grad_norm inf (inf) [2021-04-15 22:48:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][800/1251] eta 0:02:09 lr 0.000806 time 0.3831 (0.2863) loss 4.1670 (3.7391) grad_norm 1.1991 (inf) [2021-04-15 22:48:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][810/1251] eta 0:02:06 lr 0.000806 time 0.2678 (0.2861) loss 3.7302 (3.7416) grad_norm 1.2520 (inf) [2021-04-15 22:48:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][820/1251] eta 0:02:03 lr 0.000806 time 0.2632 (0.2860) loss 2.5864 (3.7438) grad_norm 1.1613 (inf) [2021-04-15 22:48:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][830/1251] eta 0:02:00 lr 0.000806 time 0.2501 (0.2859) loss 3.9672 (3.7459) grad_norm 1.2490 (inf) [2021-04-15 22:48:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][840/1251] eta 0:01:57 lr 0.000806 time 0.2846 (0.2858) loss 3.8070 (3.7427) grad_norm 1.0498 (inf) [2021-04-15 22:48:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][850/1251] eta 0:01:54 lr 0.000806 time 0.2704 (0.2857) loss 2.7398 (3.7445) grad_norm 1.1135 (inf) [2021-04-15 22:48:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][860/1251] eta 0:01:51 lr 0.000806 time 0.2595 (0.2856) loss 4.8734 (3.7465) grad_norm 1.0673 (inf) [2021-04-15 22:48:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][870/1251] eta 0:01:48 lr 0.000806 time 0.2879 (0.2855) loss 4.3505 (3.7461) grad_norm 1.1188 (inf) [2021-04-15 22:49:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][880/1251] eta 0:01:45 lr 0.000805 time 0.2732 (0.2854) loss 3.6517 (3.7462) grad_norm 1.1248 (inf) [2021-04-15 22:49:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][890/1251] eta 0:01:42 lr 0.000805 time 0.2525 (0.2853) loss 3.2832 (3.7470) grad_norm 1.0817 (inf) [2021-04-15 22:49:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][900/1251] eta 0:01:40 lr 0.000805 time 0.2685 (0.2853) loss 4.1972 (3.7467) grad_norm 1.0888 (inf) [2021-04-15 22:49:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][910/1251] eta 0:01:37 lr 0.000805 time 0.2940 (0.2852) loss 4.7960 (3.7466) grad_norm 1.0667 (inf) [2021-04-15 22:49:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][920/1251] eta 0:01:34 lr 0.000805 time 0.2569 (0.2850) loss 3.9835 (3.7478) grad_norm 1.3080 (inf) [2021-04-15 22:49:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][930/1251] eta 0:01:31 lr 0.000805 time 0.2655 (0.2850) loss 2.6010 (3.7449) grad_norm 1.2470 (inf) [2021-04-15 22:49:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][940/1251] eta 0:01:28 lr 0.000805 time 0.4107 (0.2851) loss 3.0718 (3.7416) grad_norm 1.3606 (inf) [2021-04-15 22:49:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][950/1251] eta 0:01:25 lr 0.000805 time 0.2582 (0.2851) loss 3.2736 (3.7391) grad_norm 1.2356 (inf) [2021-04-15 22:49:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][960/1251] eta 0:01:22 lr 0.000805 time 0.2814 (0.2851) loss 2.6652 (3.7388) grad_norm 1.4237 (inf) [2021-04-15 22:49:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][970/1251] eta 0:01:20 lr 0.000805 time 0.2874 (0.2849) loss 4.5994 (3.7394) grad_norm 1.1859 (inf) [2021-04-15 22:49:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][980/1251] eta 0:01:17 lr 0.000805 time 0.2876 (0.2848) loss 4.1108 (3.7406) grad_norm 1.2295 (inf) [2021-04-15 22:49:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][990/1251] eta 0:01:14 lr 0.000805 time 0.2664 (0.2848) loss 4.4621 (3.7399) grad_norm 1.1575 (inf) [2021-04-15 22:49:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][1000/1251] eta 0:01:11 lr 0.000805 time 0.2989 (0.2847) loss 3.7888 (3.7401) grad_norm 1.1277 (inf) [2021-04-15 22:49:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][1010/1251] eta 0:01:08 lr 0.000805 time 0.2840 (0.2846) loss 4.3815 (3.7418) grad_norm 1.3198 (inf) [2021-04-15 22:49:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][1020/1251] eta 0:01:05 lr 0.000805 time 0.2975 (0.2845) loss 3.9515 (3.7430) grad_norm 1.3071 (inf) [2021-04-15 22:49:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][1030/1251] eta 0:01:02 lr 0.000805 time 0.2781 (0.2846) loss 3.0826 (3.7400) grad_norm 1.0821 (inf) [2021-04-15 22:49:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][1040/1251] eta 0:01:00 lr 0.000805 time 0.2899 (0.2846) loss 4.5327 (3.7408) grad_norm 1.1310 (inf) [2021-04-15 22:49:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][1050/1251] eta 0:00:57 lr 0.000805 time 0.2838 (0.2845) loss 2.6942 (3.7377) grad_norm 1.3477 (inf) [2021-04-15 22:49:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][1060/1251] eta 0:00:54 lr 0.000805 time 0.2757 (0.2844) loss 4.3290 (3.7396) grad_norm 1.2796 (inf) [2021-04-15 22:49:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][1070/1251] eta 0:00:51 lr 0.000805 time 0.2756 (0.2843) loss 4.5075 (3.7431) grad_norm 1.3082 (inf) [2021-04-15 22:49:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][1080/1251] eta 0:00:48 lr 0.000805 time 0.2807 (0.2843) loss 4.4002 (3.7459) grad_norm 1.1333 (inf) [2021-04-15 22:50:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][1090/1251] eta 0:00:45 lr 0.000805 time 0.2601 (0.2844) loss 4.1383 (3.7472) grad_norm 1.2639 (inf) [2021-04-15 22:50:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][1100/1251] eta 0:00:42 lr 0.000805 time 0.2575 (0.2843) loss 4.3629 (3.7474) grad_norm 1.2623 (inf) [2021-04-15 22:50:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][1110/1251] eta 0:00:40 lr 0.000805 time 0.2577 (0.2843) loss 2.7662 (3.7468) grad_norm 1.1573 (inf) [2021-04-15 22:50:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][1120/1251] eta 0:00:37 lr 0.000805 time 0.2865 (0.2843) loss 4.3583 (3.7434) grad_norm 1.0850 (inf) [2021-04-15 22:50:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][1130/1251] eta 0:00:34 lr 0.000805 time 0.2680 (0.2843) loss 4.1930 (3.7462) grad_norm 1.2160 (inf) [2021-04-15 22:50:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][1140/1251] eta 0:00:31 lr 0.000805 time 0.2880 (0.2843) loss 2.9036 (3.7450) grad_norm 1.5782 (inf) [2021-04-15 22:50:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][1150/1251] eta 0:00:28 lr 0.000805 time 0.2693 (0.2842) loss 2.7870 (3.7422) grad_norm 1.2457 (inf) [2021-04-15 22:50:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][1160/1251] eta 0:00:25 lr 0.000805 time 0.2715 (0.2842) loss 3.6748 (3.7430) grad_norm 1.1590 (inf) [2021-04-15 22:50:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][1170/1251] eta 0:00:23 lr 0.000805 time 0.2604 (0.2842) loss 3.5668 (3.7421) grad_norm 1.2676 (inf) [2021-04-15 22:50:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][1180/1251] eta 0:00:20 lr 0.000805 time 0.2811 (0.2841) loss 3.7552 (3.7415) grad_norm 1.3631 (inf) [2021-04-15 22:50:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][1190/1251] eta 0:00:17 lr 0.000804 time 0.2733 (0.2841) loss 4.3800 (3.7435) grad_norm 1.1042 (inf) [2021-04-15 22:50:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][1200/1251] eta 0:00:14 lr 0.000804 time 0.2621 (0.2841) loss 3.0719 (3.7440) grad_norm 1.1747 (inf) [2021-04-15 22:50:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][1210/1251] eta 0:00:11 lr 0.000804 time 0.2626 (0.2840) loss 4.0770 (3.7435) grad_norm 1.1680 (inf) [2021-04-15 22:50:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][1220/1251] eta 0:00:08 lr 0.000804 time 0.2737 (0.2839) loss 3.8612 (3.7434) grad_norm 1.3327 (inf) [2021-04-15 22:50:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][1230/1251] eta 0:00:05 lr 0.000804 time 0.2789 (0.2839) loss 3.5751 (3.7457) grad_norm 1.2101 (inf) [2021-04-15 22:50:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][1240/1251] eta 0:00:03 lr 0.000804 time 0.2484 (0.2838) loss 4.2042 (3.7460) grad_norm 1.3348 (inf) [2021-04-15 22:50:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [87/300][1250/1251] eta 0:00:00 lr 0.000804 time 0.2488 (0.2835) loss 4.2041 (3.7457) grad_norm 1.2659 (inf) [2021-04-15 22:50:48 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 87 training takes 0:05:57 [2021-04-15 22:50:48 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_87.pth saving...... [2021-04-15 22:50:58 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_87.pth saved !!! [2021-04-15 22:51:00 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.081 (1.081) Loss 1.1669 (1.1669) Acc@1 71.387 (71.387) Acc@5 91.016 (91.016) [2021-04-15 22:51:01 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.307 (0.253) Loss 1.1561 (1.1684) Acc@1 72.852 (72.479) Acc@5 91.309 (91.646) [2021-04-15 22:51:03 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.135 (0.236) Loss 1.1520 (1.1662) Acc@1 73.242 (72.786) Acc@5 91.406 (91.541) [2021-04-15 22:51:06 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.085 (0.239) Loss 1.1052 (1.1549) Acc@1 73.340 (73.009) Acc@5 93.457 (91.712) [2021-04-15 22:51:07 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.219) Loss 1.0979 (1.1557) Acc@1 73.730 (72.999) Acc@5 92.871 (91.721) [2021-04-15 22:51:12 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 73.030 Acc@5 91.702 [2021-04-15 22:51:12 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 73.0% [2021-04-15 22:51:12 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 73.03% [2021-04-15 22:51:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][0/1251] eta 0:56:35 lr 0.000804 time 2.7142 (2.7142) loss 3.6531 (3.6531) grad_norm 1.1038 (1.1038) [2021-04-15 22:51:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][10/1251] eta 0:10:54 lr 0.000804 time 0.5104 (0.5272) loss 2.7181 (3.4666) grad_norm 1.1122 (1.2609) [2021-04-15 22:51:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][20/1251] eta 0:08:22 lr 0.000804 time 0.2810 (0.4078) loss 3.9320 (3.5822) grad_norm 1.3028 (1.2630) [2021-04-15 22:51:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][30/1251] eta 0:07:26 lr 0.000804 time 0.2798 (0.3657) loss 2.7386 (3.4690) grad_norm 1.2222 (1.2280) [2021-04-15 22:51:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][40/1251] eta 0:06:56 lr 0.000804 time 0.2810 (0.3435) loss 3.4799 (3.4824) grad_norm 1.2098 (1.2090) [2021-04-15 22:51:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][50/1251] eta 0:06:36 lr 0.000804 time 0.2748 (0.3301) loss 3.7741 (3.5329) grad_norm 1.0034 (1.1926) [2021-04-15 22:51:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][60/1251] eta 0:06:23 lr 0.000804 time 0.2893 (0.3219) loss 3.6200 (3.5584) grad_norm 1.2550 (1.1867) [2021-04-15 22:51:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][70/1251] eta 0:06:11 lr 0.000804 time 0.2770 (0.3147) loss 3.7620 (3.5686) grad_norm 1.1500 (1.1956) [2021-04-15 22:51:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][80/1251] eta 0:06:02 lr 0.000804 time 0.2599 (0.3099) loss 3.1315 (3.5858) grad_norm 1.2691 (1.2052) [2021-04-15 22:51:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][90/1251] eta 0:05:55 lr 0.000804 time 0.2679 (0.3063) loss 4.2894 (3.5957) grad_norm 1.1016 (1.2104) [2021-04-15 22:51:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][100/1251] eta 0:05:49 lr 0.000804 time 0.2813 (0.3034) loss 3.1788 (3.5990) grad_norm 1.0832 (1.2175) [2021-04-15 22:51:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][110/1251] eta 0:05:43 lr 0.000804 time 0.3007 (0.3012) loss 4.3719 (3.6095) grad_norm 1.4232 (1.2165) [2021-04-15 22:51:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][120/1251] eta 0:05:38 lr 0.000804 time 0.2764 (0.2993) loss 4.4050 (3.6217) grad_norm 1.1864 (1.2144) [2021-04-15 22:51:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][130/1251] eta 0:05:34 lr 0.000804 time 0.2865 (0.2988) loss 4.4410 (3.6538) grad_norm 1.2640 (1.2147) [2021-04-15 22:51:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][140/1251] eta 0:05:31 lr 0.000804 time 0.2736 (0.2980) loss 3.3574 (3.6541) grad_norm 1.2601 (1.2226) [2021-04-15 22:51:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][150/1251] eta 0:05:27 lr 0.000804 time 0.2615 (0.2974) loss 3.8873 (3.6801) grad_norm 1.5111 (1.2236) [2021-04-15 22:52:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][160/1251] eta 0:05:23 lr 0.000804 time 0.2785 (0.2963) loss 2.2425 (3.6800) grad_norm 1.4685 (1.2233) [2021-04-15 22:52:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][170/1251] eta 0:05:19 lr 0.000804 time 0.2982 (0.2951) loss 4.1143 (3.6781) grad_norm 1.2030 (1.2198) [2021-04-15 22:52:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][180/1251] eta 0:05:14 lr 0.000804 time 0.2639 (0.2941) loss 4.2616 (3.6689) grad_norm 1.1785 (1.2159) [2021-04-15 22:52:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][190/1251] eta 0:05:10 lr 0.000804 time 0.2755 (0.2930) loss 4.4115 (3.6562) grad_norm 1.2793 (1.2138) [2021-04-15 22:52:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][200/1251] eta 0:05:07 lr 0.000804 time 0.2720 (0.2922) loss 3.3017 (3.6718) grad_norm 1.3811 (1.2139) [2021-04-15 22:52:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][210/1251] eta 0:05:03 lr 0.000804 time 0.2635 (0.2914) loss 4.6027 (3.6890) grad_norm 1.2065 (1.2234) [2021-04-15 22:52:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][220/1251] eta 0:04:59 lr 0.000804 time 0.2835 (0.2907) loss 3.9334 (3.7025) grad_norm 1.3337 (1.2216) [2021-04-15 22:52:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][230/1251] eta 0:04:56 lr 0.000804 time 0.2862 (0.2900) loss 3.1611 (3.6981) grad_norm 1.0938 (1.2182) [2021-04-15 22:52:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][240/1251] eta 0:04:52 lr 0.000803 time 0.2927 (0.2894) loss 3.0021 (3.6941) grad_norm 1.0431 (1.2167) [2021-04-15 22:52:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][250/1251] eta 0:04:49 lr 0.000803 time 0.2727 (0.2895) loss 4.1035 (3.7112) grad_norm 1.1191 (1.2174) [2021-04-15 22:52:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][260/1251] eta 0:04:46 lr 0.000803 time 0.2785 (0.2888) loss 4.0687 (3.7089) grad_norm 1.1076 (1.2171) [2021-04-15 22:52:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][270/1251] eta 0:04:42 lr 0.000803 time 0.2856 (0.2882) loss 4.3582 (3.7145) grad_norm 0.9981 (1.2140) [2021-04-15 22:52:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][280/1251] eta 0:04:39 lr 0.000803 time 0.2713 (0.2879) loss 3.8596 (3.7215) grad_norm 1.2330 (1.2132) [2021-04-15 22:52:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][290/1251] eta 0:04:36 lr 0.000803 time 0.2751 (0.2877) loss 3.4885 (3.7187) grad_norm 1.2977 (1.2149) [2021-04-15 22:52:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][300/1251] eta 0:04:33 lr 0.000803 time 0.2424 (0.2879) loss 3.7667 (3.7179) grad_norm 1.2797 (1.2146) [2021-04-15 22:52:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][310/1251] eta 0:04:30 lr 0.000803 time 0.2707 (0.2875) loss 3.4792 (3.7238) grad_norm 1.4185 (1.2184) [2021-04-15 22:52:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][320/1251] eta 0:04:27 lr 0.000803 time 0.2795 (0.2876) loss 3.9197 (3.7140) grad_norm 1.3160 (1.2184) [2021-04-15 22:52:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][330/1251] eta 0:04:24 lr 0.000803 time 0.2696 (0.2872) loss 4.3561 (3.7215) grad_norm 1.1140 (1.2190) [2021-04-15 22:52:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][340/1251] eta 0:04:21 lr 0.000803 time 0.2904 (0.2873) loss 4.0459 (3.7253) grad_norm 1.1710 (1.2176) [2021-04-15 22:52:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][350/1251] eta 0:04:18 lr 0.000803 time 0.2679 (0.2869) loss 3.6775 (3.7198) grad_norm 1.1857 (1.2165) [2021-04-15 22:52:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][360/1251] eta 0:04:15 lr 0.000803 time 0.2755 (0.2867) loss 4.4261 (3.7234) grad_norm 1.4455 (1.2188) [2021-04-15 22:52:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][370/1251] eta 0:04:12 lr 0.000803 time 0.2432 (0.2870) loss 3.5178 (3.7219) grad_norm 1.1852 (1.2169) [2021-04-15 22:53:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][380/1251] eta 0:04:09 lr 0.000803 time 0.2843 (0.2868) loss 3.6806 (3.7237) grad_norm 1.4611 (1.2229) [2021-04-15 22:53:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][390/1251] eta 0:04:06 lr 0.000803 time 0.2944 (0.2865) loss 3.9710 (3.7246) grad_norm 1.3093 (1.2235) [2021-04-15 22:53:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][400/1251] eta 0:04:03 lr 0.000803 time 0.2652 (0.2861) loss 3.5326 (3.7232) grad_norm 1.1564 (1.2255) [2021-04-15 22:53:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][410/1251] eta 0:04:00 lr 0.000803 time 0.2806 (0.2861) loss 3.4261 (3.7230) grad_norm 1.2892 (1.2261) [2021-04-15 22:53:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][420/1251] eta 0:03:57 lr 0.000803 time 0.2687 (0.2859) loss 2.7473 (3.7140) grad_norm 1.3276 (1.2260) [2021-04-15 22:53:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][430/1251] eta 0:03:54 lr 0.000803 time 0.2670 (0.2855) loss 2.8812 (3.7186) grad_norm 1.0596 (1.2239) [2021-04-15 22:53:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][440/1251] eta 0:03:51 lr 0.000803 time 0.3011 (0.2854) loss 3.1504 (3.7222) grad_norm 1.1860 (1.2252) [2021-04-15 22:53:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][450/1251] eta 0:03:48 lr 0.000803 time 0.2597 (0.2852) loss 3.7646 (3.7141) grad_norm 1.2483 (1.2238) [2021-04-15 22:53:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][460/1251] eta 0:03:45 lr 0.000803 time 0.2757 (0.2851) loss 4.3529 (3.7153) grad_norm 1.1698 (1.2247) [2021-04-15 22:53:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][470/1251] eta 0:03:42 lr 0.000803 time 0.2874 (0.2849) loss 2.8189 (3.7142) grad_norm 1.1471 (1.2236) [2021-04-15 22:53:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][480/1251] eta 0:03:39 lr 0.000803 time 0.2440 (0.2847) loss 2.4513 (3.7093) grad_norm 1.2161 (1.2235) [2021-04-15 22:53:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][490/1251] eta 0:03:36 lr 0.000803 time 0.2577 (0.2847) loss 3.8688 (3.7137) grad_norm 1.4349 (1.2237) [2021-04-15 22:53:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][500/1251] eta 0:03:33 lr 0.000803 time 0.3323 (0.2846) loss 4.3877 (3.7104) grad_norm 1.4837 (1.2242) [2021-04-15 22:53:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][510/1251] eta 0:03:30 lr 0.000803 time 0.2589 (0.2844) loss 4.1160 (3.7140) grad_norm 1.0676 (1.2240) [2021-04-15 22:53:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][520/1251] eta 0:03:28 lr 0.000803 time 0.2714 (0.2846) loss 4.3130 (3.7108) grad_norm 1.1984 (1.2230) [2021-04-15 22:53:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][530/1251] eta 0:03:25 lr 0.000803 time 0.2764 (0.2845) loss 2.4910 (3.7068) grad_norm 1.1175 (1.2216) [2021-04-15 22:53:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][540/1251] eta 0:03:22 lr 0.000802 time 0.2918 (0.2846) loss 4.0932 (3.7006) grad_norm 1.2252 (1.2207) [2021-04-15 22:53:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][550/1251] eta 0:03:19 lr 0.000802 time 0.2500 (0.2844) loss 3.1045 (3.7001) grad_norm 1.6218 (1.2216) [2021-04-15 22:53:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][560/1251] eta 0:03:16 lr 0.000802 time 0.2858 (0.2844) loss 4.0544 (3.6968) grad_norm 1.4046 (1.2227) [2021-04-15 22:53:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][570/1251] eta 0:03:13 lr 0.000802 time 0.2992 (0.2843) loss 4.5369 (3.7013) grad_norm 1.2831 (1.2238) [2021-04-15 22:53:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][580/1251] eta 0:03:10 lr 0.000802 time 0.2771 (0.2841) loss 4.6600 (3.7041) grad_norm 1.3186 (1.2251) [2021-04-15 22:54:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][590/1251] eta 0:03:07 lr 0.000802 time 0.2746 (0.2842) loss 3.5616 (3.7012) grad_norm 1.2008 (1.2277) [2021-04-15 22:54:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][600/1251] eta 0:03:04 lr 0.000802 time 0.2775 (0.2840) loss 3.8057 (3.7012) grad_norm 1.1992 (1.2276) [2021-04-15 22:54:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][610/1251] eta 0:03:01 lr 0.000802 time 0.2688 (0.2839) loss 4.3684 (3.6985) grad_norm 1.1216 (1.2273) [2021-04-15 22:54:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][620/1251] eta 0:02:59 lr 0.000802 time 0.2764 (0.2838) loss 3.4299 (3.7018) grad_norm 1.9285 (1.2295) [2021-04-15 22:54:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][630/1251] eta 0:02:56 lr 0.000802 time 0.2993 (0.2837) loss 4.2102 (3.6997) grad_norm 1.1667 (1.2287) [2021-04-15 22:54:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][640/1251] eta 0:02:53 lr 0.000802 time 0.2647 (0.2836) loss 4.7227 (3.6997) grad_norm 1.1055 (1.2276) [2021-04-15 22:54:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][650/1251] eta 0:02:50 lr 0.000802 time 0.2641 (0.2835) loss 4.2872 (3.7001) grad_norm 1.1473 (1.2279) [2021-04-15 22:54:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][660/1251] eta 0:02:47 lr 0.000802 time 0.2995 (0.2834) loss 3.9436 (3.7035) grad_norm 1.1432 (1.2274) [2021-04-15 22:54:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][670/1251] eta 0:02:44 lr 0.000802 time 0.2519 (0.2833) loss 4.3014 (3.7051) grad_norm 1.2497 (1.2267) [2021-04-15 22:54:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][680/1251] eta 0:02:41 lr 0.000802 time 0.2703 (0.2832) loss 3.0616 (3.7066) grad_norm 1.2536 (1.2264) [2021-04-15 22:54:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][690/1251] eta 0:02:38 lr 0.000802 time 0.2834 (0.2833) loss 4.1244 (3.7111) grad_norm 1.0759 (1.2264) [2021-04-15 22:54:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][700/1251] eta 0:02:36 lr 0.000802 time 0.2455 (0.2833) loss 3.7874 (3.7083) grad_norm 1.0369 (1.2257) [2021-04-15 22:54:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][710/1251] eta 0:02:33 lr 0.000802 time 0.2725 (0.2834) loss 3.5717 (3.7059) grad_norm 1.3961 (1.2258) [2021-04-15 22:54:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][720/1251] eta 0:02:30 lr 0.000802 time 0.2712 (0.2833) loss 3.7645 (3.7070) grad_norm 1.4380 (1.2271) [2021-04-15 22:54:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][730/1251] eta 0:02:27 lr 0.000802 time 0.2746 (0.2833) loss 3.3838 (3.7049) grad_norm 1.2090 (1.2261) [2021-04-15 22:54:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][740/1251] eta 0:02:24 lr 0.000802 time 0.2750 (0.2832) loss 3.9154 (3.7059) grad_norm 1.2998 (1.2264) [2021-04-15 22:54:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][750/1251] eta 0:02:21 lr 0.000802 time 0.2895 (0.2831) loss 2.8764 (3.7084) grad_norm 1.1564 (1.2273) [2021-04-15 22:54:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][760/1251] eta 0:02:18 lr 0.000802 time 0.2713 (0.2830) loss 4.1468 (3.7111) grad_norm 1.1511 (1.2267) [2021-04-15 22:54:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][770/1251] eta 0:02:16 lr 0.000802 time 0.2868 (0.2829) loss 3.9021 (3.7083) grad_norm 1.2821 (1.2266) [2021-04-15 22:54:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][780/1251] eta 0:02:13 lr 0.000802 time 0.2546 (0.2828) loss 4.1775 (3.7091) grad_norm 1.3183 (1.2270) [2021-04-15 22:54:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][790/1251] eta 0:02:10 lr 0.000802 time 0.2651 (0.2828) loss 3.6462 (3.7067) grad_norm 1.1513 (1.2282) [2021-04-15 22:54:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][800/1251] eta 0:02:07 lr 0.000802 time 0.2823 (0.2827) loss 3.9056 (3.7077) grad_norm 1.4858 (1.2281) [2021-04-15 22:55:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][810/1251] eta 0:02:04 lr 0.000802 time 0.2697 (0.2825) loss 3.9958 (3.7071) grad_norm 1.1616 (1.2291) [2021-04-15 22:55:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][820/1251] eta 0:02:01 lr 0.000802 time 0.2450 (0.2824) loss 3.8464 (3.7065) grad_norm 1.2819 (1.2296) [2021-04-15 22:55:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][830/1251] eta 0:01:58 lr 0.000802 time 0.2731 (0.2825) loss 4.1760 (3.7098) grad_norm 1.2795 (1.2296) [2021-04-15 22:55:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][840/1251] eta 0:01:56 lr 0.000801 time 0.2999 (0.2825) loss 3.8546 (3.7100) grad_norm 1.0878 (1.2299) [2021-04-15 22:55:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][850/1251] eta 0:01:53 lr 0.000801 time 0.2819 (0.2824) loss 3.7285 (3.7129) grad_norm 1.1612 (1.2297) [2021-04-15 22:55:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][860/1251] eta 0:01:50 lr 0.000801 time 0.2501 (0.2824) loss 3.4857 (3.7132) grad_norm 1.2529 (1.2296) [2021-04-15 22:55:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][870/1251] eta 0:01:47 lr 0.000801 time 0.2832 (0.2824) loss 4.3022 (3.7116) grad_norm 1.1437 (1.2279) [2021-04-15 22:55:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][880/1251] eta 0:01:44 lr 0.000801 time 0.3119 (0.2824) loss 3.3664 (3.7131) grad_norm 1.2312 (1.2284) [2021-04-15 22:55:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][890/1251] eta 0:01:41 lr 0.000801 time 0.2666 (0.2825) loss 4.4912 (3.7139) grad_norm 1.5219 (1.2297) [2021-04-15 22:55:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][900/1251] eta 0:01:39 lr 0.000801 time 0.2761 (0.2824) loss 4.6673 (3.7153) grad_norm 1.3427 (1.2290) [2021-04-15 22:55:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][910/1251] eta 0:01:36 lr 0.000801 time 0.2792 (0.2824) loss 3.9902 (3.7155) grad_norm 1.1295 (1.2286) [2021-04-15 22:55:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][920/1251] eta 0:01:33 lr 0.000801 time 0.2760 (0.2825) loss 3.9578 (3.7148) grad_norm 1.0545 (1.2285) [2021-04-15 22:55:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][930/1251] eta 0:01:30 lr 0.000801 time 0.2780 (0.2824) loss 3.7124 (3.7160) grad_norm 1.0723 (1.2283) [2021-04-15 22:55:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][940/1251] eta 0:01:27 lr 0.000801 time 0.4172 (0.2826) loss 4.2082 (3.7149) grad_norm 1.2136 (1.2278) [2021-04-15 22:55:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][950/1251] eta 0:01:25 lr 0.000801 time 0.2524 (0.2826) loss 3.4638 (3.7166) grad_norm 1.2296 (1.2277) [2021-04-15 22:55:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][960/1251] eta 0:01:22 lr 0.000801 time 0.2664 (0.2825) loss 2.6178 (3.7147) grad_norm 1.2794 (1.2279) [2021-04-15 22:55:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][970/1251] eta 0:01:19 lr 0.000801 time 0.2736 (0.2824) loss 3.6061 (3.7124) grad_norm 1.2465 (1.2279) [2021-04-15 22:55:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][980/1251] eta 0:01:16 lr 0.000801 time 0.2715 (0.2824) loss 3.1342 (3.7119) grad_norm 1.2223 (1.2270) [2021-04-15 22:55:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][990/1251] eta 0:01:13 lr 0.000801 time 0.2614 (0.2823) loss 3.3658 (3.7144) grad_norm 1.1282 (1.2262) [2021-04-15 22:55:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][1000/1251] eta 0:01:10 lr 0.000801 time 0.2752 (0.2822) loss 4.1672 (3.7181) grad_norm 1.2538 (1.2261) [2021-04-15 22:55:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][1010/1251] eta 0:01:08 lr 0.000801 time 0.2689 (0.2822) loss 4.0908 (3.7177) grad_norm 1.2623 (1.2259) [2021-04-15 22:56:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][1020/1251] eta 0:01:05 lr 0.000801 time 0.2796 (0.2822) loss 3.4994 (3.7173) grad_norm 1.3645 (1.2255) [2021-04-15 22:56:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][1030/1251] eta 0:01:02 lr 0.000801 time 0.2664 (0.2821) loss 4.2744 (3.7192) grad_norm 1.1175 (1.2252) [2021-04-15 22:56:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][1040/1251] eta 0:00:59 lr 0.000801 time 0.2651 (0.2821) loss 3.1404 (3.7186) grad_norm 0.9932 (1.2246) [2021-04-15 22:56:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][1050/1251] eta 0:00:56 lr 0.000801 time 0.2897 (0.2820) loss 4.2803 (3.7166) grad_norm 1.3421 (1.2245) [2021-04-15 22:56:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][1060/1251] eta 0:00:53 lr 0.000801 time 0.2746 (0.2820) loss 3.6730 (3.7175) grad_norm 1.3900 (1.2247) [2021-04-15 22:56:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][1070/1251] eta 0:00:51 lr 0.000801 time 0.2755 (0.2821) loss 3.1681 (3.7175) grad_norm 1.5216 (1.2259) [2021-04-15 22:56:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][1080/1251] eta 0:00:48 lr 0.000801 time 0.2870 (0.2821) loss 3.6425 (3.7178) grad_norm 1.0702 (1.2254) [2021-04-15 22:56:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][1090/1251] eta 0:00:45 lr 0.000801 time 0.2711 (0.2821) loss 3.8932 (3.7187) grad_norm 1.2981 (1.2260) [2021-04-15 22:56:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][1100/1251] eta 0:00:42 lr 0.000801 time 0.2643 (0.2821) loss 3.8567 (3.7148) grad_norm 1.2904 (1.2259) [2021-04-15 22:56:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][1110/1251] eta 0:00:39 lr 0.000801 time 0.2974 (0.2820) loss 4.0367 (3.7134) grad_norm 1.2615 (1.2251) [2021-04-15 22:56:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][1120/1251] eta 0:00:36 lr 0.000801 time 0.2670 (0.2821) loss 4.4728 (3.7147) grad_norm 1.2799 (1.2253) [2021-04-15 22:56:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][1130/1251] eta 0:00:34 lr 0.000801 time 0.2786 (0.2820) loss 4.5629 (3.7153) grad_norm 1.1283 (1.2257) [2021-04-15 22:56:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][1140/1251] eta 0:00:31 lr 0.000801 time 0.2546 (0.2819) loss 2.4781 (3.7174) grad_norm 1.5582 (1.2259) [2021-04-15 22:56:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][1150/1251] eta 0:00:28 lr 0.000800 time 0.2562 (0.2819) loss 3.0994 (3.7176) grad_norm 1.2759 (1.2258) [2021-04-15 22:56:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][1160/1251] eta 0:00:25 lr 0.000800 time 0.3148 (0.2819) loss 2.7497 (3.7155) grad_norm 1.2777 (1.2257) [2021-04-15 22:56:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][1170/1251] eta 0:00:22 lr 0.000800 time 0.2461 (0.2820) loss 3.1299 (3.7157) grad_norm 1.1689 (1.2253) [2021-04-15 22:56:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][1180/1251] eta 0:00:20 lr 0.000800 time 0.2884 (0.2819) loss 3.7688 (3.7162) grad_norm 1.1075 (1.2249) [2021-04-15 22:56:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][1190/1251] eta 0:00:17 lr 0.000800 time 0.2878 (0.2818) loss 2.8870 (3.7148) grad_norm 1.1046 (1.2246) [2021-04-15 22:56:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][1200/1251] eta 0:00:14 lr 0.000800 time 0.2684 (0.2818) loss 2.9857 (3.7150) grad_norm 1.6149 (1.2248) [2021-04-15 22:56:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][1210/1251] eta 0:00:11 lr 0.000800 time 0.2935 (0.2818) loss 2.7593 (3.7149) grad_norm 1.2028 (1.2251) [2021-04-15 22:56:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][1220/1251] eta 0:00:08 lr 0.000800 time 0.2736 (0.2817) loss 4.6603 (3.7134) grad_norm 1.3325 (1.2253) [2021-04-15 22:56:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][1230/1251] eta 0:00:05 lr 0.000800 time 0.2626 (0.2816) loss 3.7620 (3.7134) grad_norm 1.1739 (1.2254) [2021-04-15 22:57:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][1240/1251] eta 0:00:03 lr 0.000800 time 0.2485 (0.2815) loss 3.5707 (3.7132) grad_norm 1.2044 (1.2253) [2021-04-15 22:57:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [88/300][1250/1251] eta 0:00:00 lr 0.000800 time 0.2480 (0.2813) loss 3.9323 (3.7133) grad_norm 1.7201 (1.2256) [2021-04-15 22:57:06 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 88 training takes 0:05:54 [2021-04-15 22:57:06 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_88.pth saving...... [2021-04-15 22:57:16 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_88.pth saved !!! [2021-04-15 22:57:18 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.341 (1.341) Loss 1.1602 (1.1602) Acc@1 72.168 (72.168) Acc@5 91.309 (91.309) [2021-04-15 22:57:19 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.138 (0.264) Loss 1.1127 (1.1401) Acc@1 75.000 (73.295) Acc@5 92.090 (91.939) [2021-04-15 22:57:21 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.114 (0.233) Loss 1.1407 (1.1462) Acc@1 72.559 (73.368) Acc@5 91.992 (91.709) [2021-04-15 22:57:23 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.125 (0.227) Loss 1.1428 (1.1435) Acc@1 72.559 (73.412) Acc@5 92.383 (91.879) [2021-04-15 22:57:26 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.281 (0.224) Loss 1.2197 (1.1460) Acc@1 71.875 (73.428) Acc@5 90.918 (91.856) [2021-04-15 22:57:30 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 73.366 Acc@5 91.852 [2021-04-15 22:57:30 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 73.4% [2021-04-15 22:57:30 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 73.37% [2021-04-15 22:57:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][0/1251] eta 0:55:34 lr 0.000800 time 2.6658 (2.6658) loss 3.5838 (3.5838) grad_norm 1.5032 (1.5032) [2021-04-15 22:57:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][10/1251] eta 0:10:32 lr 0.000800 time 0.2915 (0.5096) loss 3.8709 (3.6965) grad_norm 1.1079 (1.3334) [2021-04-15 22:57:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][20/1251] eta 0:08:10 lr 0.000800 time 0.2839 (0.3981) loss 3.3121 (3.5441) grad_norm 1.4074 (1.2808) [2021-04-15 22:57:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][30/1251] eta 0:07:19 lr 0.000800 time 0.2942 (0.3600) loss 4.7983 (3.5504) grad_norm 1.1865 (1.2836) [2021-04-15 22:57:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][40/1251] eta 0:06:54 lr 0.000800 time 0.2789 (0.3421) loss 4.6268 (3.6039) grad_norm 1.0795 (1.2503) [2021-04-15 22:57:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][50/1251] eta 0:06:34 lr 0.000800 time 0.2750 (0.3284) loss 3.5012 (3.6351) grad_norm 1.0978 (1.2437) [2021-04-15 22:57:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][60/1251] eta 0:06:21 lr 0.000800 time 0.2712 (0.3203) loss 3.8456 (3.6652) grad_norm 1.1655 (1.2496) [2021-04-15 22:57:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][70/1251] eta 0:06:12 lr 0.000800 time 0.3007 (0.3150) loss 3.9060 (3.6899) grad_norm 1.1487 (1.2477) [2021-04-15 22:57:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][80/1251] eta 0:06:03 lr 0.000800 time 0.2737 (0.3107) loss 3.9226 (3.6923) grad_norm 1.2222 (1.2485) [2021-04-15 22:57:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][90/1251] eta 0:05:57 lr 0.000800 time 0.2839 (0.3080) loss 4.3234 (3.7098) grad_norm 1.2289 (1.2432) [2021-04-15 22:58:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][100/1251] eta 0:05:52 lr 0.000800 time 0.2785 (0.3064) loss 3.6804 (3.7357) grad_norm 1.0980 (1.2321) [2021-04-15 22:58:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][110/1251] eta 0:05:46 lr 0.000800 time 0.2705 (0.3035) loss 2.8242 (3.7180) grad_norm 1.1614 (1.2251) [2021-04-15 22:58:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][120/1251] eta 0:05:42 lr 0.000800 time 0.2753 (0.3024) loss 3.1442 (3.7257) grad_norm 1.2587 (1.2286) [2021-04-15 22:58:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][130/1251] eta 0:05:37 lr 0.000800 time 0.2742 (0.3007) loss 3.9006 (3.7312) grad_norm 1.0634 (1.2312) [2021-04-15 22:58:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][140/1251] eta 0:05:32 lr 0.000800 time 0.2619 (0.2996) loss 4.2217 (3.7432) grad_norm 1.0734 (1.2303) [2021-04-15 22:58:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][150/1251] eta 0:05:29 lr 0.000800 time 0.4325 (0.2989) loss 4.1582 (3.7521) grad_norm 1.1034 (1.2324) [2021-04-15 22:58:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][160/1251] eta 0:05:24 lr 0.000800 time 0.2769 (0.2971) loss 4.4299 (3.7516) grad_norm 1.4090 (1.2274) [2021-04-15 22:58:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][170/1251] eta 0:05:19 lr 0.000800 time 0.3030 (0.2960) loss 4.2143 (3.7468) grad_norm 1.2454 (1.2317) [2021-04-15 22:58:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][180/1251] eta 0:05:15 lr 0.000800 time 0.2889 (0.2948) loss 4.4112 (3.7484) grad_norm 1.1805 (1.2272) [2021-04-15 22:58:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][190/1251] eta 0:05:11 lr 0.000799 time 0.2800 (0.2938) loss 3.6204 (3.7529) grad_norm 1.4216 (1.2282) [2021-04-15 22:58:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][200/1251] eta 0:05:07 lr 0.000799 time 0.2661 (0.2929) loss 2.6006 (3.7481) grad_norm 1.1848 (1.2291) [2021-04-15 22:58:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][210/1251] eta 0:05:04 lr 0.000799 time 0.3120 (0.2923) loss 2.5623 (3.7377) grad_norm 1.1465 (1.2259) [2021-04-15 22:58:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][220/1251] eta 0:05:00 lr 0.000799 time 0.2818 (0.2916) loss 2.7261 (3.7304) grad_norm 1.3759 (1.2261) [2021-04-15 22:58:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][230/1251] eta 0:04:57 lr 0.000799 time 0.2756 (0.2913) loss 3.7469 (3.7299) grad_norm 1.1433 (1.2240) [2021-04-15 22:58:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][240/1251] eta 0:04:54 lr 0.000799 time 0.2783 (0.2909) loss 3.2732 (3.7254) grad_norm 1.1961 (1.2239) [2021-04-15 22:58:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][250/1251] eta 0:04:50 lr 0.000799 time 0.2524 (0.2903) loss 3.1054 (3.7131) grad_norm 1.5058 (1.2259) [2021-04-15 22:58:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][260/1251] eta 0:04:47 lr 0.000799 time 0.2810 (0.2898) loss 4.0228 (3.7190) grad_norm 1.1588 (1.2260) [2021-04-15 22:58:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][270/1251] eta 0:04:43 lr 0.000799 time 0.2776 (0.2893) loss 4.4865 (3.7161) grad_norm 1.1191 (1.2263) [2021-04-15 22:58:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][280/1251] eta 0:04:40 lr 0.000799 time 0.2689 (0.2888) loss 3.9815 (3.7144) grad_norm 1.1347 (1.2231) [2021-04-15 22:58:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][290/1251] eta 0:04:37 lr 0.000799 time 0.2729 (0.2887) loss 4.1012 (3.7110) grad_norm 1.1651 (1.2219) [2021-04-15 22:58:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][300/1251] eta 0:04:34 lr 0.000799 time 0.2727 (0.2887) loss 4.5240 (3.7058) grad_norm 1.2604 (1.2200) [2021-04-15 22:59:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][310/1251] eta 0:04:31 lr 0.000799 time 0.2808 (0.2885) loss 3.9195 (3.7038) grad_norm 1.1492 (1.2193) [2021-04-15 22:59:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][320/1251] eta 0:04:28 lr 0.000799 time 0.2665 (0.2886) loss 4.0992 (3.7094) grad_norm 1.2671 (1.2192) [2021-04-15 22:59:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][330/1251] eta 0:04:25 lr 0.000799 time 0.2944 (0.2884) loss 3.8868 (3.7079) grad_norm 1.2030 (1.2237) [2021-04-15 22:59:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][340/1251] eta 0:04:22 lr 0.000799 time 0.2872 (0.2880) loss 3.7462 (3.7052) grad_norm 1.2860 (1.2233) [2021-04-15 22:59:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][350/1251] eta 0:04:19 lr 0.000799 time 0.2860 (0.2877) loss 4.0396 (3.7131) grad_norm 1.2466 (1.2228) [2021-04-15 22:59:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][360/1251] eta 0:04:16 lr 0.000799 time 0.2770 (0.2875) loss 4.4574 (3.7126) grad_norm 1.2970 (1.2256) [2021-04-15 22:59:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][370/1251] eta 0:04:13 lr 0.000799 time 0.2666 (0.2876) loss 4.6672 (3.7259) grad_norm 1.0951 (1.2253) [2021-04-15 22:59:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][380/1251] eta 0:04:10 lr 0.000799 time 0.2746 (0.2876) loss 4.2221 (3.7295) grad_norm 1.0093 (1.2237) [2021-04-15 22:59:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][390/1251] eta 0:04:07 lr 0.000799 time 0.2509 (0.2872) loss 3.8035 (3.7401) grad_norm 1.0503 (1.2239) [2021-04-15 22:59:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][400/1251] eta 0:04:04 lr 0.000799 time 0.2748 (0.2870) loss 3.4971 (3.7387) grad_norm 1.3541 (1.2238) [2021-04-15 22:59:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][410/1251] eta 0:04:01 lr 0.000799 time 0.2620 (0.2867) loss 3.4234 (3.7455) grad_norm 1.1711 (1.2249) [2021-04-15 22:59:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][420/1251] eta 0:03:58 lr 0.000799 time 0.2880 (0.2865) loss 3.7402 (3.7424) grad_norm 1.1226 (1.2266) [2021-04-15 22:59:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][430/1251] eta 0:03:54 lr 0.000799 time 0.2691 (0.2862) loss 3.0057 (3.7428) grad_norm 1.2789 (1.2256) [2021-04-15 22:59:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][440/1251] eta 0:03:51 lr 0.000799 time 0.2839 (0.2860) loss 4.3010 (3.7467) grad_norm 1.4897 (1.2256) [2021-04-15 22:59:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][450/1251] eta 0:03:48 lr 0.000799 time 0.3150 (0.2858) loss 4.0307 (3.7574) grad_norm 1.2194 (1.2255) [2021-04-15 22:59:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][460/1251] eta 0:03:45 lr 0.000799 time 0.2607 (0.2856) loss 3.9076 (3.7558) grad_norm 1.2397 (1.2254) [2021-04-15 22:59:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][470/1251] eta 0:03:42 lr 0.000799 time 0.2635 (0.2854) loss 4.0779 (3.7546) grad_norm 1.6229 (1.2245) [2021-04-15 22:59:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][480/1251] eta 0:03:40 lr 0.000799 time 0.2929 (0.2854) loss 3.6336 (3.7542) grad_norm 1.6149 (1.2251) [2021-04-15 22:59:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][490/1251] eta 0:03:37 lr 0.000798 time 0.2674 (0.2852) loss 4.5555 (3.7519) grad_norm 1.1053 (1.2245) [2021-04-15 22:59:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][500/1251] eta 0:03:34 lr 0.000798 time 0.2566 (0.2851) loss 3.2869 (3.7486) grad_norm 1.2686 (1.2250) [2021-04-15 22:59:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][510/1251] eta 0:03:31 lr 0.000798 time 0.2648 (0.2849) loss 3.7676 (3.7474) grad_norm 1.1276 (1.2250) [2021-04-15 22:59:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][520/1251] eta 0:03:28 lr 0.000798 time 0.2743 (0.2848) loss 4.2347 (3.7505) grad_norm 1.2746 (1.2247) [2021-04-15 23:00:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][530/1251] eta 0:03:25 lr 0.000798 time 0.2851 (0.2846) loss 3.8225 (3.7538) grad_norm 1.2228 (1.2253) [2021-04-15 23:00:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][540/1251] eta 0:03:22 lr 0.000798 time 0.2685 (0.2845) loss 3.7795 (3.7588) grad_norm 1.1104 (1.2253) [2021-04-15 23:00:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][550/1251] eta 0:03:19 lr 0.000798 time 0.2736 (0.2844) loss 3.7516 (3.7612) grad_norm 1.0755 (1.2246) [2021-04-15 23:00:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][560/1251] eta 0:03:16 lr 0.000798 time 0.2890 (0.2842) loss 4.1749 (3.7635) grad_norm 1.2577 (1.2252) [2021-04-15 23:00:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][570/1251] eta 0:03:13 lr 0.000798 time 0.2906 (0.2843) loss 3.2364 (3.7639) grad_norm 1.2292 (1.2266) [2021-04-15 23:00:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][580/1251] eta 0:03:10 lr 0.000798 time 0.2567 (0.2844) loss 4.5225 (3.7672) grad_norm 1.1643 (1.2253) [2021-04-15 23:00:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][590/1251] eta 0:03:07 lr 0.000798 time 0.2754 (0.2842) loss 4.3795 (3.7647) grad_norm 1.1721 (1.2252) [2021-04-15 23:00:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][600/1251] eta 0:03:04 lr 0.000798 time 0.2738 (0.2841) loss 2.8589 (3.7626) grad_norm 1.1046 (1.2265) [2021-04-15 23:00:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][610/1251] eta 0:03:02 lr 0.000798 time 0.2814 (0.2840) loss 3.7814 (3.7637) grad_norm 1.2607 (1.2282) [2021-04-15 23:00:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][620/1251] eta 0:02:59 lr 0.000798 time 0.2881 (0.2840) loss 3.2741 (3.7616) grad_norm 0.9537 (1.2280) [2021-04-15 23:00:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][630/1251] eta 0:02:56 lr 0.000798 time 0.2651 (0.2839) loss 3.8921 (3.7587) grad_norm 1.3217 (1.2270) [2021-04-15 23:00:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][640/1251] eta 0:02:53 lr 0.000798 time 0.2643 (0.2837) loss 3.9010 (3.7548) grad_norm 1.1287 (1.2283) [2021-04-15 23:00:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][650/1251] eta 0:02:50 lr 0.000798 time 0.2599 (0.2837) loss 4.5444 (3.7531) grad_norm 1.1925 (1.2288) [2021-04-15 23:00:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][660/1251] eta 0:02:47 lr 0.000798 time 0.2767 (0.2835) loss 4.1858 (3.7529) grad_norm 1.2586 (1.2283) [2021-04-15 23:00:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][670/1251] eta 0:02:44 lr 0.000798 time 0.3111 (0.2836) loss 3.0260 (3.7492) grad_norm 1.4334 (1.2286) [2021-04-15 23:00:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][680/1251] eta 0:02:42 lr 0.000798 time 0.2908 (0.2837) loss 3.9931 (3.7481) grad_norm 1.5049 (1.2291) [2021-04-15 23:00:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][690/1251] eta 0:02:39 lr 0.000798 time 0.2926 (0.2838) loss 4.0802 (3.7527) grad_norm 1.1753 (1.2303) [2021-04-15 23:00:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][700/1251] eta 0:02:36 lr 0.000798 time 0.2675 (0.2837) loss 4.1220 (3.7519) grad_norm 1.2387 (1.2300) [2021-04-15 23:00:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][710/1251] eta 0:02:33 lr 0.000798 time 0.2638 (0.2836) loss 4.1974 (3.7529) grad_norm 1.0882 (1.2297) [2021-04-15 23:00:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][720/1251] eta 0:02:30 lr 0.000798 time 0.2647 (0.2835) loss 3.3559 (3.7513) grad_norm 1.2251 (1.2309) [2021-04-15 23:00:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][730/1251] eta 0:02:27 lr 0.000798 time 0.2707 (0.2836) loss 3.9921 (3.7510) grad_norm 1.1608 (1.2308) [2021-04-15 23:01:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][740/1251] eta 0:02:24 lr 0.000798 time 0.3072 (0.2835) loss 3.7858 (3.7477) grad_norm 1.1637 (1.2304) [2021-04-15 23:01:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][750/1251] eta 0:02:21 lr 0.000798 time 0.2948 (0.2834) loss 3.2524 (3.7451) grad_norm 1.1821 (1.2286) [2021-04-15 23:01:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][760/1251] eta 0:02:19 lr 0.000798 time 0.2907 (0.2833) loss 2.8723 (3.7457) grad_norm 1.0432 (1.2282) [2021-04-15 23:01:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][770/1251] eta 0:02:16 lr 0.000798 time 0.2645 (0.2831) loss 3.8585 (3.7474) grad_norm 0.9845 (1.2284) [2021-04-15 23:01:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][780/1251] eta 0:02:13 lr 0.000798 time 0.2846 (0.2831) loss 3.4554 (3.7499) grad_norm 1.2051 (1.2286) [2021-04-15 23:01:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][790/1251] eta 0:02:10 lr 0.000797 time 0.2836 (0.2830) loss 3.2754 (3.7469) grad_norm 1.2185 (1.2297) [2021-04-15 23:01:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][800/1251] eta 0:02:07 lr 0.000797 time 0.2650 (0.2829) loss 4.4257 (3.7468) grad_norm 1.1982 (1.2297) [2021-04-15 23:01:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][810/1251] eta 0:02:04 lr 0.000797 time 0.2671 (0.2828) loss 3.9779 (3.7463) grad_norm 1.2757 (1.2299) [2021-04-15 23:01:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][820/1251] eta 0:02:01 lr 0.000797 time 0.2486 (0.2828) loss 2.5763 (3.7440) grad_norm 1.0658 (1.2300) [2021-04-15 23:01:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][830/1251] eta 0:01:58 lr 0.000797 time 0.2714 (0.2826) loss 3.1900 (3.7417) grad_norm 1.2126 (1.2301) [2021-04-15 23:01:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][840/1251] eta 0:01:56 lr 0.000797 time 0.2696 (0.2826) loss 3.7083 (3.7415) grad_norm 1.1377 (1.2295) [2021-04-15 23:01:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][850/1251] eta 0:01:53 lr 0.000797 time 0.2696 (0.2825) loss 3.3128 (3.7418) grad_norm 1.2080 (1.2289) [2021-04-15 23:01:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][860/1251] eta 0:01:50 lr 0.000797 time 0.2760 (0.2824) loss 4.1390 (3.7436) grad_norm 1.1859 (1.2283) [2021-04-15 23:01:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][870/1251] eta 0:01:47 lr 0.000797 time 0.2725 (0.2824) loss 4.0619 (3.7443) grad_norm 1.6777 (1.2287) [2021-04-15 23:01:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][880/1251] eta 0:01:44 lr 0.000797 time 0.2892 (0.2825) loss 4.0680 (3.7484) grad_norm 1.1572 (1.2293) [2021-04-15 23:01:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][890/1251] eta 0:01:42 lr 0.000797 time 0.2489 (0.2826) loss 4.1591 (3.7489) grad_norm 1.0309 (1.2293) [2021-04-15 23:01:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][900/1251] eta 0:01:39 lr 0.000797 time 0.3374 (0.2826) loss 4.4141 (3.7515) grad_norm 1.1538 (1.2284) [2021-04-15 23:01:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][910/1251] eta 0:01:36 lr 0.000797 time 0.2679 (0.2825) loss 4.3859 (3.7553) grad_norm 1.1757 (1.2283) [2021-04-15 23:01:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][920/1251] eta 0:01:33 lr 0.000797 time 0.2739 (0.2824) loss 4.0386 (3.7581) grad_norm 1.3611 (1.2284) [2021-04-15 23:01:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][930/1251] eta 0:01:30 lr 0.000797 time 0.2676 (0.2823) loss 4.4620 (3.7565) grad_norm 1.0979 (1.2279) [2021-04-15 23:01:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][940/1251] eta 0:01:27 lr 0.000797 time 0.2566 (0.2824) loss 4.1843 (3.7562) grad_norm 1.0979 (1.2280) [2021-04-15 23:01:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][950/1251] eta 0:01:24 lr 0.000797 time 0.3044 (0.2824) loss 3.3766 (3.7555) grad_norm 1.6244 (1.2290) [2021-04-15 23:02:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][960/1251] eta 0:01:22 lr 0.000797 time 0.2567 (0.2824) loss 3.9676 (3.7556) grad_norm 1.0636 (1.2296) [2021-04-15 23:02:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][970/1251] eta 0:01:19 lr 0.000797 time 0.2790 (0.2824) loss 3.4059 (3.7545) grad_norm 1.1587 (1.2296) [2021-04-15 23:02:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][980/1251] eta 0:01:16 lr 0.000797 time 0.2866 (0.2823) loss 3.7526 (3.7536) grad_norm 1.2679 (1.2299) [2021-04-15 23:02:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][990/1251] eta 0:01:13 lr 0.000797 time 0.2753 (0.2823) loss 3.7397 (3.7534) grad_norm 1.2930 (1.2296) [2021-04-15 23:02:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][1000/1251] eta 0:01:10 lr 0.000797 time 0.2780 (0.2822) loss 3.8967 (3.7534) grad_norm 1.1072 (1.2296) [2021-04-15 23:02:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][1010/1251] eta 0:01:07 lr 0.000797 time 0.2876 (0.2821) loss 3.2649 (3.7508) grad_norm 1.2475 (1.2297) [2021-04-15 23:02:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][1020/1251] eta 0:01:05 lr 0.000797 time 0.2892 (0.2821) loss 3.7762 (3.7525) grad_norm 1.5897 (1.2300) [2021-04-15 23:02:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][1030/1251] eta 0:01:02 lr 0.000797 time 0.2880 (0.2820) loss 4.7454 (3.7508) grad_norm 1.3239 (1.2305) [2021-04-15 23:02:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][1040/1251] eta 0:00:59 lr 0.000797 time 0.2901 (0.2820) loss 3.8975 (3.7491) grad_norm 1.0034 (1.2310) [2021-04-15 23:02:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][1050/1251] eta 0:00:56 lr 0.000797 time 0.2548 (0.2819) loss 3.4417 (3.7508) grad_norm 1.1784 (1.2307) [2021-04-15 23:02:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][1060/1251] eta 0:00:53 lr 0.000797 time 0.2705 (0.2819) loss 3.0524 (3.7523) grad_norm 1.6135 (1.2308) [2021-04-15 23:02:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][1070/1251] eta 0:00:51 lr 0.000797 time 0.2573 (0.2819) loss 4.2386 (3.7541) grad_norm 1.1202 (1.2305) [2021-04-15 23:02:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][1080/1251] eta 0:00:48 lr 0.000797 time 0.2837 (0.2818) loss 2.8004 (3.7551) grad_norm 1.2178 (1.2301) [2021-04-15 23:02:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][1090/1251] eta 0:00:45 lr 0.000796 time 0.2724 (0.2818) loss 3.7782 (3.7547) grad_norm 1.1822 (1.2298) [2021-04-15 23:02:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][1100/1251] eta 0:00:42 lr 0.000796 time 0.2573 (0.2817) loss 4.0272 (3.7566) grad_norm 1.1740 (1.2295) [2021-04-15 23:02:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][1110/1251] eta 0:00:39 lr 0.000796 time 0.2681 (0.2817) loss 4.3527 (3.7573) grad_norm 1.7912 (1.2297) [2021-04-15 23:02:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][1120/1251] eta 0:00:36 lr 0.000796 time 0.2535 (0.2816) loss 4.4910 (3.7562) grad_norm 1.0955 (1.2298) [2021-04-15 23:02:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][1130/1251] eta 0:00:34 lr 0.000796 time 0.2834 (0.2816) loss 4.5551 (3.7571) grad_norm 1.1213 (1.2293) [2021-04-15 23:02:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][1140/1251] eta 0:00:31 lr 0.000796 time 0.2821 (0.2815) loss 3.9392 (3.7580) grad_norm 1.2201 (1.2293) [2021-04-15 23:02:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][1150/1251] eta 0:00:28 lr 0.000796 time 0.2776 (0.2815) loss 3.8923 (3.7571) grad_norm 1.1264 (1.2303) [2021-04-15 23:02:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][1160/1251] eta 0:00:25 lr 0.000796 time 0.2670 (0.2817) loss 3.6646 (3.7568) grad_norm 1.0952 (1.2295) [2021-04-15 23:03:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][1170/1251] eta 0:00:22 lr 0.000796 time 0.2805 (0.2816) loss 4.0815 (3.7578) grad_norm 1.0015 (1.2291) [2021-04-15 23:03:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][1180/1251] eta 0:00:19 lr 0.000796 time 0.2860 (0.2815) loss 3.0212 (3.7561) grad_norm 1.2591 (1.2288) [2021-04-15 23:03:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][1190/1251] eta 0:00:17 lr 0.000796 time 0.4333 (0.2816) loss 3.4634 (3.7557) grad_norm 1.1372 (1.2288) [2021-04-15 23:03:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][1200/1251] eta 0:00:14 lr 0.000796 time 0.2583 (0.2816) loss 2.8311 (3.7551) grad_norm 1.3202 (1.2291) [2021-04-15 23:03:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][1210/1251] eta 0:00:11 lr 0.000796 time 0.2781 (0.2815) loss 4.3773 (3.7569) grad_norm 1.3490 (1.2298) [2021-04-15 23:03:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][1220/1251] eta 0:00:08 lr 0.000796 time 0.2601 (0.2815) loss 3.4705 (3.7590) grad_norm 1.1701 (1.2302) [2021-04-15 23:03:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][1230/1251] eta 0:00:05 lr 0.000796 time 0.2690 (0.2815) loss 4.2689 (3.7589) grad_norm 1.2732 (1.2301) [2021-04-15 23:03:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][1240/1251] eta 0:00:03 lr 0.000796 time 0.2475 (0.2814) loss 4.0489 (3.7607) grad_norm 1.3691 (1.2299) [2021-04-15 23:03:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [89/300][1250/1251] eta 0:00:00 lr 0.000796 time 0.2483 (0.2812) loss 3.9835 (3.7604) grad_norm 1.1812 (1.2296) [2021-04-15 23:03:27 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 89 training takes 0:05:56 [2021-04-15 23:03:27 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_89.pth saving...... [2021-04-15 23:03:43 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_89.pth saved !!! [2021-04-15 23:03:44 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.163 (1.163) Loss 1.1699 (1.1699) Acc@1 72.363 (72.363) Acc@5 91.406 (91.406) [2021-04-15 23:03:46 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.130 (0.276) Loss 1.0840 (1.1378) Acc@1 75.684 (73.331) Acc@5 92.578 (92.143) [2021-04-15 23:03:48 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.201 (0.224) Loss 1.1255 (1.1471) Acc@1 74.316 (73.214) Acc@5 92.578 (91.895) [2021-04-15 23:03:50 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.112 (0.232) Loss 1.1756 (1.1509) Acc@1 73.047 (73.113) Acc@5 91.406 (91.753) [2021-04-15 23:03:52 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.326 (0.211) Loss 1.2084 (1.1522) Acc@1 71.094 (73.123) Acc@5 91.016 (91.704) [2021-04-15 23:03:55 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 73.062 Acc@5 91.722 [2021-04-15 23:03:55 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 73.1% [2021-04-15 23:03:55 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 73.37% [2021-04-15 23:03:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][0/1251] eta 1:23:04 lr 0.000796 time 3.9842 (3.9842) loss 4.4237 (4.4237) grad_norm 1.1166 (1.1166) [2021-04-15 23:04:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][10/1251] eta 0:12:39 lr 0.000796 time 0.2800 (0.6119) loss 4.1970 (3.8279) grad_norm 1.1116 (1.2028) [2021-04-15 23:04:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][20/1251] eta 0:09:18 lr 0.000796 time 0.2897 (0.4534) loss 3.6197 (3.7094) grad_norm 1.0900 (1.1754) [2021-04-15 23:04:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][30/1251] eta 0:08:02 lr 0.000796 time 0.2788 (0.3952) loss 3.3413 (3.7559) grad_norm 1.2962 (1.1828) [2021-04-15 23:04:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][40/1251] eta 0:07:23 lr 0.000796 time 0.2828 (0.3665) loss 3.8653 (3.7958) grad_norm 1.3322 (1.1906) [2021-04-15 23:04:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][50/1251] eta 0:06:58 lr 0.000796 time 0.2920 (0.3482) loss 4.0977 (3.8168) grad_norm 1.3751 (1.2057) [2021-04-15 23:04:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][60/1251] eta 0:06:41 lr 0.000796 time 0.2664 (0.3368) loss 3.7162 (3.8301) grad_norm 1.3096 (1.2143) [2021-04-15 23:04:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][70/1251] eta 0:06:27 lr 0.000796 time 0.2581 (0.3283) loss 3.6156 (3.7765) grad_norm 1.1091 (1.2101) [2021-04-15 23:04:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][80/1251] eta 0:06:17 lr 0.000796 time 0.2779 (0.3226) loss 3.8687 (3.7405) grad_norm 1.1275 (1.2039) [2021-04-15 23:04:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][90/1251] eta 0:06:08 lr 0.000796 time 0.2794 (0.3177) loss 4.3904 (3.7570) grad_norm 1.3154 (1.1995) [2021-04-15 23:04:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][100/1251] eta 0:06:03 lr 0.000796 time 0.2531 (0.3155) loss 3.9443 (3.7609) grad_norm 1.1410 (1.2033) [2021-04-15 23:04:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][110/1251] eta 0:05:56 lr 0.000796 time 0.3726 (0.3127) loss 3.5451 (3.7604) grad_norm 1.1378 (1.2069) [2021-04-15 23:04:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][120/1251] eta 0:05:50 lr 0.000796 time 0.2693 (0.3096) loss 4.0140 (3.7559) grad_norm 1.1102 (1.2094) [2021-04-15 23:04:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][130/1251] eta 0:05:44 lr 0.000796 time 0.2999 (0.3070) loss 2.8337 (3.7458) grad_norm 1.1029 (1.2081) [2021-04-15 23:04:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][140/1251] eta 0:05:40 lr 0.000795 time 0.2682 (0.3064) loss 3.9480 (3.7676) grad_norm 1.2426 (1.2121) [2021-04-15 23:04:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][150/1251] eta 0:05:36 lr 0.000795 time 0.2586 (0.3055) loss 3.2288 (3.7806) grad_norm 1.2413 (1.2154) [2021-04-15 23:04:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][160/1251] eta 0:05:31 lr 0.000795 time 0.2719 (0.3039) loss 4.1914 (3.7872) grad_norm 1.2547 (1.2198) [2021-04-15 23:04:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][170/1251] eta 0:05:26 lr 0.000795 time 0.2677 (0.3020) loss 4.1432 (3.7998) grad_norm 1.3329 (1.2236) [2021-04-15 23:04:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][180/1251] eta 0:05:22 lr 0.000795 time 0.2898 (0.3007) loss 3.3564 (3.7923) grad_norm 0.9985 (1.2224) [2021-04-15 23:04:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][190/1251] eta 0:05:17 lr 0.000795 time 0.2792 (0.2993) loss 4.3210 (3.7941) grad_norm 1.1350 (1.2222) [2021-04-15 23:04:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][200/1251] eta 0:05:13 lr 0.000795 time 0.2777 (0.2983) loss 3.9955 (3.8010) grad_norm 1.0541 (1.2171) [2021-04-15 23:04:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][210/1251] eta 0:05:09 lr 0.000795 time 0.2611 (0.2972) loss 2.4793 (3.7870) grad_norm 1.1365 (1.2191) [2021-04-15 23:05:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][220/1251] eta 0:05:05 lr 0.000795 time 0.2805 (0.2962) loss 2.7787 (3.7851) grad_norm 1.3570 (1.2215) [2021-04-15 23:05:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][230/1251] eta 0:05:01 lr 0.000795 time 0.2580 (0.2951) loss 2.2684 (3.7723) grad_norm 1.3367 (1.2195) [2021-04-15 23:05:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][240/1251] eta 0:04:57 lr 0.000795 time 0.2689 (0.2944) loss 4.6045 (3.7761) grad_norm 1.1634 (1.2217) [2021-04-15 23:05:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][250/1251] eta 0:04:53 lr 0.000795 time 0.2717 (0.2935) loss 4.0338 (3.7744) grad_norm 1.1379 (1.2217) [2021-04-15 23:05:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][260/1251] eta 0:04:50 lr 0.000795 time 0.2526 (0.2928) loss 3.4864 (3.7766) grad_norm 1.3109 (1.2219) [2021-04-15 23:05:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][270/1251] eta 0:04:46 lr 0.000795 time 0.2821 (0.2922) loss 3.5588 (3.7724) grad_norm 1.1388 (1.2214) [2021-04-15 23:05:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][280/1251] eta 0:04:43 lr 0.000795 time 0.2726 (0.2916) loss 4.0867 (3.7808) grad_norm 1.1611 (1.2261) [2021-04-15 23:05:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][290/1251] eta 0:04:39 lr 0.000795 time 0.2714 (0.2911) loss 4.6218 (3.7953) grad_norm 1.1988 (1.2285) [2021-04-15 23:05:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][300/1251] eta 0:04:36 lr 0.000795 time 0.2509 (0.2906) loss 3.6797 (3.7963) grad_norm 1.1563 (1.2272) [2021-04-15 23:05:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][310/1251] eta 0:04:33 lr 0.000795 time 0.2863 (0.2902) loss 4.0761 (3.7853) grad_norm 1.2437 (1.2248) [2021-04-15 23:05:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][320/1251] eta 0:04:29 lr 0.000795 time 0.2730 (0.2897) loss 4.5695 (3.7890) grad_norm 1.0942 (1.2261) [2021-04-15 23:05:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][330/1251] eta 0:04:26 lr 0.000795 time 0.2690 (0.2893) loss 4.4417 (3.7809) grad_norm 1.6956 (1.2295) [2021-04-15 23:05:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][340/1251] eta 0:04:23 lr 0.000795 time 0.2797 (0.2890) loss 3.9700 (3.7847) grad_norm 1.4014 (1.2310) [2021-04-15 23:05:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][350/1251] eta 0:04:20 lr 0.000795 time 0.2826 (0.2889) loss 3.9855 (3.7851) grad_norm 1.1460 (1.2330) [2021-04-15 23:05:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][360/1251] eta 0:04:17 lr 0.000795 time 0.2595 (0.2894) loss 3.1458 (3.7878) grad_norm 1.1406 (1.2301) [2021-04-15 23:05:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][370/1251] eta 0:04:14 lr 0.000795 time 0.2449 (0.2891) loss 4.0858 (3.7850) grad_norm 1.1401 (1.2308) [2021-04-15 23:05:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][380/1251] eta 0:04:11 lr 0.000795 time 0.2544 (0.2887) loss 4.3055 (3.7759) grad_norm 1.3315 (1.2309) [2021-04-15 23:05:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][390/1251] eta 0:04:08 lr 0.000795 time 0.2762 (0.2890) loss 3.8485 (3.7761) grad_norm 1.4729 (1.2313) [2021-04-15 23:05:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][400/1251] eta 0:04:05 lr 0.000795 time 0.2748 (0.2885) loss 4.1152 (3.7750) grad_norm 1.3907 (1.2331) [2021-04-15 23:05:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][410/1251] eta 0:04:02 lr 0.000795 time 0.2906 (0.2883) loss 4.1707 (3.7777) grad_norm 1.2512 (1.2347) [2021-04-15 23:05:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][420/1251] eta 0:03:59 lr 0.000795 time 0.2449 (0.2879) loss 4.0464 (3.7664) grad_norm 1.2914 (1.2343) [2021-04-15 23:05:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][430/1251] eta 0:03:56 lr 0.000795 time 0.2922 (0.2878) loss 4.0236 (3.7693) grad_norm 1.1889 (1.2334) [2021-04-15 23:06:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][440/1251] eta 0:03:53 lr 0.000794 time 0.2853 (0.2879) loss 4.0990 (3.7672) grad_norm 1.1363 (1.2339) [2021-04-15 23:06:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][450/1251] eta 0:03:50 lr 0.000794 time 0.3117 (0.2877) loss 3.0599 (3.7687) grad_norm 1.4693 (1.2359) [2021-04-15 23:06:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][460/1251] eta 0:03:47 lr 0.000794 time 0.2833 (0.2875) loss 2.3590 (3.7699) grad_norm 1.1826 (1.2355) [2021-04-15 23:06:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][470/1251] eta 0:03:44 lr 0.000794 time 0.2642 (0.2872) loss 4.2512 (3.7715) grad_norm 1.1614 (1.2348) [2021-04-15 23:06:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][480/1251] eta 0:03:41 lr 0.000794 time 0.2691 (0.2873) loss 3.2407 (3.7715) grad_norm 1.0240 (1.2339) [2021-04-15 23:06:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][490/1251] eta 0:03:38 lr 0.000794 time 0.2845 (0.2872) loss 3.9808 (3.7710) grad_norm 1.4952 (1.2350) [2021-04-15 23:06:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][500/1251] eta 0:03:35 lr 0.000794 time 0.2600 (0.2869) loss 3.4149 (3.7719) grad_norm 1.1643 (1.2380) [2021-04-15 23:06:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][510/1251] eta 0:03:32 lr 0.000794 time 0.2564 (0.2867) loss 3.8391 (3.7677) grad_norm 1.2281 (1.2376) [2021-04-15 23:06:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][520/1251] eta 0:03:29 lr 0.000794 time 0.2624 (0.2864) loss 3.2236 (3.7713) grad_norm 1.1582 (1.2357) [2021-04-15 23:06:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][530/1251] eta 0:03:26 lr 0.000794 time 0.2847 (0.2862) loss 4.7493 (3.7779) grad_norm 1.0114 (1.2353) [2021-04-15 23:06:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][540/1251] eta 0:03:23 lr 0.000794 time 0.2540 (0.2859) loss 3.8151 (3.7807) grad_norm 1.1654 (1.2361) [2021-04-15 23:06:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][550/1251] eta 0:03:20 lr 0.000794 time 0.2746 (0.2857) loss 3.3879 (3.7786) grad_norm 1.3099 (1.2371) [2021-04-15 23:06:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][560/1251] eta 0:03:17 lr 0.000794 time 0.2998 (0.2855) loss 3.8018 (3.7748) grad_norm 1.3736 (1.2372) [2021-04-15 23:06:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][570/1251] eta 0:03:14 lr 0.000794 time 0.2792 (0.2855) loss 3.7648 (3.7782) grad_norm 1.2991 (1.2362) [2021-04-15 23:06:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][580/1251] eta 0:03:11 lr 0.000794 time 0.2559 (0.2855) loss 3.7395 (3.7817) grad_norm 1.2957 (1.2366) [2021-04-15 23:06:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][590/1251] eta 0:03:08 lr 0.000794 time 0.2902 (0.2854) loss 4.8384 (3.7872) grad_norm 1.4566 (1.2382) [2021-04-15 23:06:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][600/1251] eta 0:03:05 lr 0.000794 time 0.2706 (0.2853) loss 3.4266 (3.7872) grad_norm 1.2178 (1.2379) [2021-04-15 23:06:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][610/1251] eta 0:03:02 lr 0.000794 time 0.2915 (0.2851) loss 3.8927 (3.7896) grad_norm 1.0890 (1.2384) [2021-04-15 23:06:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][620/1251] eta 0:02:59 lr 0.000794 time 0.2881 (0.2850) loss 3.7699 (3.7885) grad_norm 1.2078 (1.2380) [2021-04-15 23:06:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][630/1251] eta 0:02:56 lr 0.000794 time 0.2693 (0.2848) loss 3.4945 (3.7902) grad_norm 1.5327 (1.2374) [2021-04-15 23:06:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][640/1251] eta 0:02:53 lr 0.000794 time 0.2882 (0.2847) loss 3.9504 (3.7903) grad_norm 1.2457 (1.2373) [2021-04-15 23:07:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][650/1251] eta 0:02:51 lr 0.000794 time 0.2729 (0.2848) loss 4.0070 (3.7913) grad_norm 1.0840 (1.2363) [2021-04-15 23:07:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][660/1251] eta 0:02:48 lr 0.000794 time 0.2629 (0.2847) loss 4.0781 (3.7924) grad_norm 1.3880 (1.2373) [2021-04-15 23:07:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][670/1251] eta 0:02:45 lr 0.000794 time 0.2730 (0.2845) loss 4.4323 (3.7946) grad_norm 1.2126 (1.2368) [2021-04-15 23:07:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][680/1251] eta 0:02:42 lr 0.000794 time 0.2689 (0.2845) loss 3.3836 (3.7935) grad_norm 1.1969 (1.2359) [2021-04-15 23:07:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][690/1251] eta 0:02:39 lr 0.000794 time 0.2572 (0.2845) loss 4.1188 (3.7933) grad_norm 1.2186 (1.2364) [2021-04-15 23:07:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][700/1251] eta 0:02:36 lr 0.000794 time 0.2751 (0.2843) loss 3.6702 (3.7927) grad_norm 1.3261 (1.2376) [2021-04-15 23:07:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][710/1251] eta 0:02:33 lr 0.000794 time 0.2770 (0.2842) loss 2.9931 (3.7896) grad_norm 1.2528 (1.2383) [2021-04-15 23:07:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][720/1251] eta 0:02:30 lr 0.000794 time 0.2695 (0.2842) loss 3.8196 (3.7880) grad_norm 1.2298 (1.2393) [2021-04-15 23:07:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][730/1251] eta 0:02:28 lr 0.000794 time 0.2489 (0.2843) loss 4.1872 (3.7870) grad_norm 1.2599 (1.2390) [2021-04-15 23:07:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][740/1251] eta 0:02:25 lr 0.000793 time 0.2852 (0.2842) loss 2.2787 (3.7848) grad_norm 1.2321 (1.2391) [2021-04-15 23:07:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][750/1251] eta 0:02:22 lr 0.000793 time 0.2677 (0.2840) loss 4.3096 (3.7790) grad_norm 1.8363 (1.2390) [2021-04-15 23:07:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][760/1251] eta 0:02:19 lr 0.000793 time 0.2927 (0.2840) loss 3.7298 (3.7763) grad_norm 1.3234 (1.2395) [2021-04-15 23:07:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][770/1251] eta 0:02:16 lr 0.000793 time 0.2674 (0.2839) loss 2.9367 (3.7780) grad_norm 1.3235 (1.2397) [2021-04-15 23:07:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][780/1251] eta 0:02:13 lr 0.000793 time 0.2680 (0.2838) loss 3.7893 (3.7802) grad_norm 1.2138 (1.2413) [2021-04-15 23:07:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][790/1251] eta 0:02:10 lr 0.000793 time 0.2872 (0.2838) loss 3.8200 (3.7827) grad_norm 1.1030 (1.2424) [2021-04-15 23:07:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][800/1251] eta 0:02:07 lr 0.000793 time 0.2793 (0.2838) loss 2.7593 (3.7826) grad_norm 1.0820 (1.2421) [2021-04-15 23:07:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][810/1251] eta 0:02:05 lr 0.000793 time 0.2868 (0.2837) loss 3.3058 (3.7804) grad_norm 1.1282 (1.2420) [2021-04-15 23:07:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][820/1251] eta 0:02:02 lr 0.000793 time 0.2537 (0.2836) loss 3.4245 (3.7783) grad_norm 1.2810 (1.2417) [2021-04-15 23:07:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][830/1251] eta 0:01:59 lr 0.000793 time 0.2559 (0.2836) loss 3.4378 (3.7792) grad_norm 1.1400 (1.2416) [2021-04-15 23:07:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][840/1251] eta 0:01:56 lr 0.000793 time 0.2614 (0.2834) loss 3.5961 (3.7779) grad_norm 1.3535 (1.2413) [2021-04-15 23:07:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][850/1251] eta 0:01:53 lr 0.000793 time 0.2609 (0.2834) loss 4.0051 (3.7776) grad_norm 1.3796 (1.2415) [2021-04-15 23:07:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][860/1251] eta 0:01:50 lr 0.000793 time 0.2484 (0.2833) loss 4.2213 (3.7753) grad_norm 1.1827 (1.2415) [2021-04-15 23:08:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][870/1251] eta 0:01:47 lr 0.000793 time 0.2730 (0.2833) loss 2.4120 (3.7765) grad_norm 1.1464 (1.2421) [2021-04-15 23:08:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][880/1251] eta 0:01:45 lr 0.000793 time 0.2482 (0.2832) loss 4.0958 (3.7783) grad_norm 1.1765 (1.2412) [2021-04-15 23:08:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][890/1251] eta 0:01:42 lr 0.000793 time 0.2682 (0.2831) loss 3.3615 (3.7797) grad_norm 1.2670 (1.2417) [2021-04-15 23:08:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][900/1251] eta 0:01:39 lr 0.000793 time 0.2812 (0.2831) loss 3.8366 (3.7746) grad_norm 1.7616 (1.2428) [2021-04-15 23:08:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][910/1251] eta 0:01:36 lr 0.000793 time 0.2750 (0.2830) loss 3.6499 (3.7701) grad_norm 1.1590 (1.2439) [2021-04-15 23:08:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][920/1251] eta 0:01:33 lr 0.000793 time 0.2578 (0.2829) loss 4.4824 (3.7692) grad_norm 1.2604 (1.2437) [2021-04-15 23:08:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][930/1251] eta 0:01:30 lr 0.000793 time 0.4263 (0.2830) loss 3.9607 (3.7692) grad_norm 1.3200 (1.2440) [2021-04-15 23:08:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][940/1251] eta 0:01:28 lr 0.000793 time 0.2813 (0.2830) loss 4.1831 (3.7680) grad_norm 1.1713 (1.2434) [2021-04-15 23:08:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][950/1251] eta 0:01:25 lr 0.000793 time 0.2661 (0.2830) loss 4.2241 (3.7705) grad_norm 1.0995 (1.2435) [2021-04-15 23:08:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][960/1251] eta 0:01:22 lr 0.000793 time 0.2980 (0.2829) loss 4.2526 (3.7736) grad_norm 1.1918 (1.2436) [2021-04-15 23:08:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][970/1251] eta 0:01:19 lr 0.000793 time 0.2733 (0.2831) loss 3.3075 (3.7737) grad_norm 1.1712 (1.2433) [2021-04-15 23:08:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][980/1251] eta 0:01:16 lr 0.000793 time 0.2773 (0.2830) loss 3.2447 (3.7717) grad_norm 1.4059 (1.2435) [2021-04-15 23:08:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][990/1251] eta 0:01:13 lr 0.000793 time 0.3166 (0.2829) loss 2.9070 (3.7715) grad_norm 1.1990 (1.2435) [2021-04-15 23:08:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][1000/1251] eta 0:01:11 lr 0.000793 time 0.2809 (0.2829) loss 4.6024 (3.7740) grad_norm 1.3688 (1.2437) [2021-04-15 23:08:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][1010/1251] eta 0:01:08 lr 0.000793 time 0.2771 (0.2828) loss 3.9221 (3.7739) grad_norm 1.5444 (1.2450) [2021-04-15 23:08:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][1020/1251] eta 0:01:05 lr 0.000793 time 0.2790 (0.2828) loss 3.9321 (3.7750) grad_norm 1.1430 (1.2454) [2021-04-15 23:08:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][1030/1251] eta 0:01:02 lr 0.000792 time 0.2920 (0.2828) loss 4.0334 (3.7767) grad_norm 1.1194 (1.2453) [2021-04-15 23:08:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][1040/1251] eta 0:00:59 lr 0.000792 time 0.2845 (0.2827) loss 3.3228 (3.7775) grad_norm 1.1481 (1.2446) [2021-04-15 23:08:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][1050/1251] eta 0:00:56 lr 0.000792 time 0.4541 (0.2828) loss 4.6587 (3.7803) grad_norm 1.1186 (1.2440) [2021-04-15 23:08:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][1060/1251] eta 0:00:53 lr 0.000792 time 0.2650 (0.2827) loss 4.4123 (3.7816) grad_norm 1.4089 (1.2436) [2021-04-15 23:08:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][1070/1251] eta 0:00:51 lr 0.000792 time 0.2646 (0.2828) loss 4.4629 (3.7826) grad_norm 1.2212 (1.2437) [2021-04-15 23:09:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][1080/1251] eta 0:00:48 lr 0.000792 time 0.2883 (0.2827) loss 3.8117 (3.7796) grad_norm 1.0702 (1.2433) [2021-04-15 23:09:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][1090/1251] eta 0:00:45 lr 0.000792 time 0.2732 (0.2826) loss 3.4688 (3.7764) grad_norm 1.1524 (1.2429) [2021-04-15 23:09:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][1100/1251] eta 0:00:42 lr 0.000792 time 0.2816 (0.2826) loss 4.1233 (3.7771) grad_norm 1.0943 (1.2433) [2021-04-15 23:09:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][1110/1251] eta 0:00:39 lr 0.000792 time 0.2804 (0.2826) loss 3.8008 (3.7772) grad_norm 1.2896 (1.2434) [2021-04-15 23:09:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][1120/1251] eta 0:00:37 lr 0.000792 time 0.2851 (0.2825) loss 3.8969 (3.7741) grad_norm 1.1227 (1.2435) [2021-04-15 23:09:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][1130/1251] eta 0:00:34 lr 0.000792 time 0.2656 (0.2824) loss 3.3680 (3.7755) grad_norm 1.3228 (1.2430) [2021-04-15 23:09:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][1140/1251] eta 0:00:31 lr 0.000792 time 0.2487 (0.2825) loss 3.2255 (3.7733) grad_norm 1.1990 (1.2426) [2021-04-15 23:09:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][1150/1251] eta 0:00:28 lr 0.000792 time 0.2442 (0.2825) loss 3.4919 (3.7722) grad_norm 1.1212 (1.2425) [2021-04-15 23:09:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][1160/1251] eta 0:00:25 lr 0.000792 time 0.3012 (0.2825) loss 3.8763 (3.7731) grad_norm 1.4457 (1.2425) [2021-04-15 23:09:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][1170/1251] eta 0:00:22 lr 0.000792 time 0.2744 (0.2824) loss 4.1295 (3.7715) grad_norm 1.2676 (1.2424) [2021-04-15 23:09:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][1180/1251] eta 0:00:20 lr 0.000792 time 0.2891 (0.2824) loss 3.2799 (3.7728) grad_norm 1.2160 (1.2428) [2021-04-15 23:09:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][1190/1251] eta 0:00:17 lr 0.000792 time 0.2540 (0.2823) loss 3.9967 (3.7736) grad_norm 1.3712 (1.2428) [2021-04-15 23:09:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][1200/1251] eta 0:00:14 lr 0.000792 time 0.2651 (0.2823) loss 4.1696 (3.7719) grad_norm 1.3329 (1.2427) [2021-04-15 23:09:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][1210/1251] eta 0:00:11 lr 0.000792 time 0.2800 (0.2822) loss 4.5331 (3.7699) grad_norm 1.4620 (1.2427) [2021-04-15 23:09:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][1220/1251] eta 0:00:08 lr 0.000792 time 0.2835 (0.2821) loss 4.3998 (3.7693) grad_norm 1.3200 (1.2423) [2021-04-15 23:09:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][1230/1251] eta 0:00:05 lr 0.000792 time 0.2652 (0.2822) loss 3.5842 (3.7696) grad_norm 1.0080 (1.2420) [2021-04-15 23:09:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][1240/1251] eta 0:00:03 lr 0.000792 time 0.2460 (0.2821) loss 4.1737 (3.7702) grad_norm 1.1632 (1.2417) [2021-04-15 23:09:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [90/300][1250/1251] eta 0:00:00 lr 0.000792 time 0.2483 (0.2818) loss 4.5381 (3.7706) grad_norm 1.1448 (1.2415) [2021-04-15 23:09:50 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 90 training takes 0:05:55 [2021-04-15 23:09:50 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_90.pth saving...... [2021-04-15 23:10:02 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_90.pth saved !!! [2021-04-15 23:10:03 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.159 (1.159) Loss 1.1822 (1.1822) Acc@1 74.512 (74.512) Acc@5 90.820 (90.820) [2021-04-15 23:10:04 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.216 (0.212) Loss 1.2292 (1.1838) Acc@1 72.168 (73.056) Acc@5 91.211 (91.788) [2021-04-15 23:10:07 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.109 (0.258) Loss 1.1699 (1.1822) Acc@1 73.926 (73.168) Acc@5 90.625 (91.625) [2021-04-15 23:10:09 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.105 (0.246) Loss 1.2125 (1.1870) Acc@1 71.289 (72.993) Acc@5 92.188 (91.658) [2021-04-15 23:10:10 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.146 (0.213) Loss 1.2369 (1.1889) Acc@1 70.605 (72.852) Acc@5 91.895 (91.744) [2021-04-15 23:10:15 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 72.936 Acc@5 91.736 [2021-04-15 23:10:15 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 72.9% [2021-04-15 23:10:15 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 73.37% [2021-04-15 23:10:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][0/1251] eta 0:55:22 lr 0.000792 time 2.6561 (2.6561) loss 3.9179 (3.9179) grad_norm 1.1625 (1.1625) [2021-04-15 23:10:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][10/1251] eta 0:10:17 lr 0.000792 time 0.2942 (0.4977) loss 3.6493 (3.7466) grad_norm 1.1608 (1.1924) [2021-04-15 23:10:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][20/1251] eta 0:08:04 lr 0.000792 time 0.2722 (0.3933) loss 3.3617 (3.7183) grad_norm 1.4929 (1.2391) [2021-04-15 23:10:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][30/1251] eta 0:07:15 lr 0.000792 time 0.3022 (0.3569) loss 3.5203 (3.7872) grad_norm 1.0818 (1.2092) [2021-04-15 23:10:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][40/1251] eta 0:06:48 lr 0.000792 time 0.2796 (0.3374) loss 3.6469 (3.8062) grad_norm 1.1434 (1.2128) [2021-04-15 23:10:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][50/1251] eta 0:06:31 lr 0.000792 time 0.2823 (0.3257) loss 3.4913 (3.8098) grad_norm 1.6272 (1.2273) [2021-04-15 23:10:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][60/1251] eta 0:06:17 lr 0.000792 time 0.2708 (0.3173) loss 4.2843 (3.7876) grad_norm 1.3261 (1.2253) [2021-04-15 23:10:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][70/1251] eta 0:06:08 lr 0.000792 time 0.2564 (0.3119) loss 4.3108 (3.7756) grad_norm 1.2414 (1.2282) [2021-04-15 23:10:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][80/1251] eta 0:06:00 lr 0.000791 time 0.2691 (0.3076) loss 3.0239 (3.7686) grad_norm 1.1207 (1.2271) [2021-04-15 23:10:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][90/1251] eta 0:05:53 lr 0.000791 time 0.2874 (0.3048) loss 4.2077 (3.7910) grad_norm 1.0523 (1.2177) [2021-04-15 23:10:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][100/1251] eta 0:05:47 lr 0.000791 time 0.2798 (0.3021) loss 4.0282 (3.7874) grad_norm 1.3873 (1.2198) [2021-04-15 23:10:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][110/1251] eta 0:05:41 lr 0.000791 time 0.2960 (0.2997) loss 3.0383 (3.7975) grad_norm 1.2026 (1.2335) [2021-04-15 23:10:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][120/1251] eta 0:05:36 lr 0.000791 time 0.2454 (0.2978) loss 3.8419 (3.7982) grad_norm inf (inf) [2021-04-15 23:10:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][130/1251] eta 0:05:32 lr 0.000791 time 0.2856 (0.2963) loss 4.4375 (3.8106) grad_norm 1.1686 (inf) [2021-04-15 23:10:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][140/1251] eta 0:05:28 lr 0.000791 time 0.4099 (0.2960) loss 3.7104 (3.8098) grad_norm 1.2772 (inf) [2021-04-15 23:11:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][150/1251] eta 0:05:25 lr 0.000791 time 0.2739 (0.2959) loss 3.7526 (3.7966) grad_norm 1.3879 (inf) [2021-04-15 23:11:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][160/1251] eta 0:05:22 lr 0.000791 time 0.2810 (0.2957) loss 4.4314 (3.7980) grad_norm 1.2252 (inf) [2021-04-15 23:11:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][170/1251] eta 0:05:19 lr 0.000791 time 0.2825 (0.2953) loss 3.5468 (3.7847) grad_norm 1.1530 (inf) [2021-04-15 23:11:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][180/1251] eta 0:05:15 lr 0.000791 time 0.2692 (0.2941) loss 3.9570 (3.7614) grad_norm 1.3013 (inf) [2021-04-15 23:11:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][190/1251] eta 0:05:11 lr 0.000791 time 0.2845 (0.2932) loss 4.6034 (3.7730) grad_norm 1.5701 (inf) [2021-04-15 23:11:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][200/1251] eta 0:05:07 lr 0.000791 time 0.2743 (0.2922) loss 3.2226 (3.7787) grad_norm 1.2929 (inf) [2021-04-15 23:11:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][210/1251] eta 0:05:03 lr 0.000791 time 0.2990 (0.2915) loss 3.5867 (3.7719) grad_norm 1.5211 (inf) [2021-04-15 23:11:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][220/1251] eta 0:05:00 lr 0.000791 time 0.2642 (0.2914) loss 4.2291 (3.7688) grad_norm 1.3195 (inf) [2021-04-15 23:11:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][230/1251] eta 0:04:56 lr 0.000791 time 0.2678 (0.2908) loss 3.9455 (3.7650) grad_norm 1.2782 (inf) [2021-04-15 23:11:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][240/1251] eta 0:04:53 lr 0.000791 time 0.2556 (0.2902) loss 3.7736 (3.7707) grad_norm 1.0671 (inf) [2021-04-15 23:11:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][250/1251] eta 0:04:50 lr 0.000791 time 0.2844 (0.2898) loss 2.6218 (3.7467) grad_norm 1.2153 (inf) [2021-04-15 23:11:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][260/1251] eta 0:04:47 lr 0.000791 time 0.4372 (0.2899) loss 3.8799 (3.7461) grad_norm 1.2206 (inf) [2021-04-15 23:11:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][270/1251] eta 0:04:43 lr 0.000791 time 0.2499 (0.2893) loss 3.7489 (3.7492) grad_norm 1.0893 (inf) [2021-04-15 23:11:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][280/1251] eta 0:04:40 lr 0.000791 time 0.2643 (0.2889) loss 3.4786 (3.7470) grad_norm 1.3830 (inf) [2021-04-15 23:11:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][290/1251] eta 0:04:37 lr 0.000791 time 0.2787 (0.2890) loss 2.9542 (3.7451) grad_norm 1.3141 (inf) [2021-04-15 23:11:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][300/1251] eta 0:04:34 lr 0.000791 time 0.2939 (0.2887) loss 3.4956 (3.7399) grad_norm 1.3774 (inf) [2021-04-15 23:11:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][310/1251] eta 0:04:31 lr 0.000791 time 0.2758 (0.2884) loss 2.7303 (3.7371) grad_norm 1.1564 (inf) [2021-04-15 23:11:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][320/1251] eta 0:04:28 lr 0.000791 time 0.2717 (0.2880) loss 3.0762 (3.7444) grad_norm 1.1694 (inf) [2021-04-15 23:11:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][330/1251] eta 0:04:24 lr 0.000791 time 0.2701 (0.2876) loss 3.0093 (3.7388) grad_norm 0.9869 (inf) [2021-04-15 23:11:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][340/1251] eta 0:04:21 lr 0.000791 time 0.2684 (0.2873) loss 3.9541 (3.7358) grad_norm 1.4062 (inf) [2021-04-15 23:11:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][350/1251] eta 0:04:19 lr 0.000791 time 0.3101 (0.2875) loss 3.8407 (3.7412) grad_norm 1.3544 (inf) [2021-04-15 23:11:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][360/1251] eta 0:04:16 lr 0.000791 time 0.2892 (0.2874) loss 3.5878 (3.7417) grad_norm 1.4007 (inf) [2021-04-15 23:12:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][370/1251] eta 0:04:13 lr 0.000790 time 0.2599 (0.2872) loss 2.5545 (3.7271) grad_norm 1.3112 (inf) [2021-04-15 23:12:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][380/1251] eta 0:04:10 lr 0.000790 time 0.2891 (0.2871) loss 3.9312 (3.7287) grad_norm 1.0772 (inf) [2021-04-15 23:12:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][390/1251] eta 0:04:06 lr 0.000790 time 0.2783 (0.2868) loss 2.6496 (3.7176) grad_norm 1.2823 (inf) [2021-04-15 23:12:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][400/1251] eta 0:04:03 lr 0.000790 time 0.2821 (0.2865) loss 3.0637 (3.7189) grad_norm 1.2192 (inf) [2021-04-15 23:12:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][410/1251] eta 0:04:00 lr 0.000790 time 0.2810 (0.2863) loss 4.2689 (3.7217) grad_norm 1.1789 (inf) [2021-04-15 23:12:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][420/1251] eta 0:03:57 lr 0.000790 time 0.2708 (0.2860) loss 3.7848 (3.7238) grad_norm 1.1387 (inf) [2021-04-15 23:12:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][430/1251] eta 0:03:54 lr 0.000790 time 0.2755 (0.2862) loss 3.9059 (3.7264) grad_norm 1.0758 (inf) [2021-04-15 23:12:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][440/1251] eta 0:03:51 lr 0.000790 time 0.2678 (0.2860) loss 4.0046 (3.7289) grad_norm 1.3420 (inf) [2021-04-15 23:12:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][450/1251] eta 0:03:48 lr 0.000790 time 0.2846 (0.2857) loss 3.6972 (3.7256) grad_norm 1.1527 (inf) [2021-04-15 23:12:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][460/1251] eta 0:03:45 lr 0.000790 time 0.2649 (0.2854) loss 3.2159 (3.7259) grad_norm 1.6836 (inf) [2021-04-15 23:12:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][470/1251] eta 0:03:43 lr 0.000790 time 0.2516 (0.2856) loss 2.5234 (3.7250) grad_norm 1.1679 (inf) [2021-04-15 23:12:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][480/1251] eta 0:03:40 lr 0.000790 time 0.2690 (0.2860) loss 3.3804 (3.7236) grad_norm 1.1597 (inf) [2021-04-15 23:12:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][490/1251] eta 0:03:37 lr 0.000790 time 0.2719 (0.2858) loss 4.0740 (3.7245) grad_norm 0.9877 (inf) [2021-04-15 23:12:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][500/1251] eta 0:03:34 lr 0.000790 time 0.2663 (0.2855) loss 2.9761 (3.7229) grad_norm 1.0558 (inf) [2021-04-15 23:12:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][510/1251] eta 0:03:31 lr 0.000790 time 0.2879 (0.2854) loss 3.9526 (3.7218) grad_norm 1.4348 (inf) [2021-04-15 23:12:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][520/1251] eta 0:03:28 lr 0.000790 time 0.2652 (0.2852) loss 4.0164 (3.7187) grad_norm 1.1512 (inf) [2021-04-15 23:12:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][530/1251] eta 0:03:25 lr 0.000790 time 0.2705 (0.2851) loss 3.4939 (3.7197) grad_norm 1.2217 (inf) [2021-04-15 23:12:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][540/1251] eta 0:03:22 lr 0.000790 time 0.2787 (0.2853) loss 3.6385 (3.7214) grad_norm 1.0628 (inf) [2021-04-15 23:12:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][550/1251] eta 0:03:19 lr 0.000790 time 0.3007 (0.2852) loss 4.2945 (3.7266) grad_norm 1.2951 (inf) [2021-04-15 23:12:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][560/1251] eta 0:03:16 lr 0.000790 time 0.2704 (0.2850) loss 3.6637 (3.7278) grad_norm 1.3673 (inf) [2021-04-15 23:12:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][570/1251] eta 0:03:14 lr 0.000790 time 0.2764 (0.2850) loss 3.6234 (3.7230) grad_norm 1.3842 (inf) [2021-04-15 23:13:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][580/1251] eta 0:03:11 lr 0.000790 time 0.2841 (0.2849) loss 4.1083 (3.7232) grad_norm 1.2374 (inf) [2021-04-15 23:13:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][590/1251] eta 0:03:08 lr 0.000790 time 0.2772 (0.2847) loss 3.6026 (3.7249) grad_norm 1.0516 (inf) [2021-04-15 23:13:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][600/1251] eta 0:03:05 lr 0.000790 time 0.2720 (0.2845) loss 3.8493 (3.7250) grad_norm 1.1470 (inf) [2021-04-15 23:13:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][610/1251] eta 0:03:02 lr 0.000790 time 0.2814 (0.2844) loss 4.3438 (3.7238) grad_norm 1.0745 (inf) [2021-04-15 23:13:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][620/1251] eta 0:02:59 lr 0.000790 time 0.2680 (0.2842) loss 4.5151 (3.7226) grad_norm 1.2998 (inf) [2021-04-15 23:13:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][630/1251] eta 0:02:56 lr 0.000790 time 0.2847 (0.2840) loss 4.1056 (3.7283) grad_norm 1.3326 (inf) [2021-04-15 23:13:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][640/1251] eta 0:02:53 lr 0.000790 time 0.2683 (0.2839) loss 3.0764 (3.7245) grad_norm 1.2632 (inf) [2021-04-15 23:13:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][650/1251] eta 0:02:50 lr 0.000790 time 0.2696 (0.2838) loss 4.2225 (3.7282) grad_norm 1.3549 (inf) [2021-04-15 23:13:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][660/1251] eta 0:02:47 lr 0.000790 time 0.3094 (0.2837) loss 4.1453 (3.7316) grad_norm 1.5092 (inf) [2021-04-15 23:13:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][670/1251] eta 0:02:44 lr 0.000789 time 0.2693 (0.2836) loss 3.8235 (3.7314) grad_norm 1.2903 (inf) [2021-04-15 23:13:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][680/1251] eta 0:02:41 lr 0.000789 time 0.2834 (0.2835) loss 4.0551 (3.7369) grad_norm 1.1635 (inf) [2021-04-15 23:13:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][690/1251] eta 0:02:38 lr 0.000789 time 0.2505 (0.2834) loss 3.9802 (3.7336) grad_norm 1.3664 (inf) [2021-04-15 23:13:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][700/1251] eta 0:02:36 lr 0.000789 time 0.2728 (0.2833) loss 4.5366 (3.7377) grad_norm 1.3586 (inf) [2021-04-15 23:13:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][710/1251] eta 0:02:33 lr 0.000789 time 0.2843 (0.2832) loss 4.1017 (3.7406) grad_norm 1.1736 (inf) [2021-04-15 23:13:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][720/1251] eta 0:02:30 lr 0.000789 time 0.2485 (0.2834) loss 2.4605 (3.7414) grad_norm 1.4865 (inf) [2021-04-15 23:13:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][730/1251] eta 0:02:27 lr 0.000789 time 0.4311 (0.2835) loss 4.4183 (3.7444) grad_norm 1.2563 (inf) [2021-04-15 23:13:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][740/1251] eta 0:02:24 lr 0.000789 time 0.4209 (0.2836) loss 4.4186 (3.7442) grad_norm 1.2672 (inf) [2021-04-15 23:13:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][750/1251] eta 0:02:21 lr 0.000789 time 0.2777 (0.2834) loss 3.7907 (3.7466) grad_norm 1.1062 (inf) [2021-04-15 23:13:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][760/1251] eta 0:02:19 lr 0.000789 time 0.2692 (0.2834) loss 3.6134 (3.7494) grad_norm 1.4260 (inf) [2021-04-15 23:13:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][770/1251] eta 0:02:16 lr 0.000789 time 0.2776 (0.2832) loss 3.2941 (3.7491) grad_norm 1.1374 (inf) [2021-04-15 23:13:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][780/1251] eta 0:02:13 lr 0.000789 time 0.2740 (0.2831) loss 2.8633 (3.7472) grad_norm 1.4948 (inf) [2021-04-15 23:13:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][790/1251] eta 0:02:10 lr 0.000789 time 0.2581 (0.2831) loss 2.4598 (3.7423) grad_norm 1.0618 (inf) [2021-04-15 23:14:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][800/1251] eta 0:02:07 lr 0.000789 time 0.2969 (0.2830) loss 3.6815 (3.7435) grad_norm 1.2079 (inf) [2021-04-15 23:14:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][810/1251] eta 0:02:04 lr 0.000789 time 0.2799 (0.2829) loss 4.3963 (3.7474) grad_norm 1.0786 (inf) [2021-04-15 23:14:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][820/1251] eta 0:02:01 lr 0.000789 time 0.2615 (0.2828) loss 3.2493 (3.7441) grad_norm 1.2072 (inf) [2021-04-15 23:14:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][830/1251] eta 0:01:59 lr 0.000789 time 0.2818 (0.2828) loss 3.8171 (3.7461) grad_norm 1.1010 (inf) [2021-04-15 23:14:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][840/1251] eta 0:01:56 lr 0.000789 time 0.2489 (0.2828) loss 4.3568 (3.7458) grad_norm 1.3431 (inf) [2021-04-15 23:14:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][850/1251] eta 0:01:53 lr 0.000789 time 0.2910 (0.2827) loss 4.0124 (3.7453) grad_norm 1.2533 (inf) [2021-04-15 23:14:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][860/1251] eta 0:01:50 lr 0.000789 time 0.2648 (0.2826) loss 2.8663 (3.7426) grad_norm 0.9672 (inf) [2021-04-15 23:14:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][870/1251] eta 0:01:47 lr 0.000789 time 0.2696 (0.2826) loss 3.8519 (3.7420) grad_norm 1.1806 (inf) [2021-04-15 23:14:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][880/1251] eta 0:01:44 lr 0.000789 time 0.2788 (0.2825) loss 3.9801 (3.7392) grad_norm 1.0511 (inf) [2021-04-15 23:14:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][890/1251] eta 0:01:41 lr 0.000789 time 0.2613 (0.2825) loss 3.3771 (3.7411) grad_norm 1.1297 (inf) [2021-04-15 23:14:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][900/1251] eta 0:01:39 lr 0.000789 time 0.2779 (0.2824) loss 4.1001 (3.7431) grad_norm 1.2333 (inf) [2021-04-15 23:14:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][910/1251] eta 0:01:36 lr 0.000789 time 0.2705 (0.2824) loss 2.9265 (3.7392) grad_norm 1.2227 (inf) [2021-04-15 23:14:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][920/1251] eta 0:01:33 lr 0.000789 time 0.2882 (0.2823) loss 4.0642 (3.7364) grad_norm 1.1955 (inf) [2021-04-15 23:14:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][930/1251] eta 0:01:30 lr 0.000789 time 0.2841 (0.2824) loss 3.8527 (3.7366) grad_norm 1.2203 (inf) [2021-04-15 23:14:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][940/1251] eta 0:01:27 lr 0.000789 time 0.2560 (0.2824) loss 4.9465 (3.7320) grad_norm 1.5612 (inf) [2021-04-15 23:14:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][950/1251] eta 0:01:24 lr 0.000789 time 0.2960 (0.2823) loss 3.7735 (3.7304) grad_norm 1.2475 (inf) [2021-04-15 23:14:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][960/1251] eta 0:01:22 lr 0.000788 time 0.2850 (0.2822) loss 4.2979 (3.7314) grad_norm 1.1350 (inf) [2021-04-15 23:14:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][970/1251] eta 0:01:19 lr 0.000788 time 0.2796 (0.2822) loss 4.1645 (3.7331) grad_norm 1.3319 (inf) [2021-04-15 23:14:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][980/1251] eta 0:01:16 lr 0.000788 time 0.2804 (0.2821) loss 3.7528 (3.7336) grad_norm 1.5395 (inf) [2021-04-15 23:14:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][990/1251] eta 0:01:13 lr 0.000788 time 0.2773 (0.2821) loss 3.7175 (3.7334) grad_norm 1.0714 (inf) [2021-04-15 23:14:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][1000/1251] eta 0:01:10 lr 0.000788 time 0.3830 (0.2821) loss 4.2139 (3.7336) grad_norm 1.4017 (inf) [2021-04-15 23:15:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][1010/1251] eta 0:01:07 lr 0.000788 time 0.2908 (0.2821) loss 3.2286 (3.7334) grad_norm 1.1338 (inf) [2021-04-15 23:15:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][1020/1251] eta 0:01:05 lr 0.000788 time 0.2896 (0.2821) loss 3.1612 (3.7324) grad_norm 1.0725 (inf) [2021-04-15 23:15:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][1030/1251] eta 0:01:02 lr 0.000788 time 0.2785 (0.2820) loss 3.9020 (3.7309) grad_norm 1.2171 (inf) [2021-04-15 23:15:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][1040/1251] eta 0:00:59 lr 0.000788 time 0.2860 (0.2821) loss 4.1261 (3.7325) grad_norm 1.0865 (inf) [2021-04-15 23:15:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][1050/1251] eta 0:00:56 lr 0.000788 time 0.4477 (0.2822) loss 4.0454 (3.7361) grad_norm 1.1627 (inf) [2021-04-15 23:15:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][1060/1251] eta 0:00:53 lr 0.000788 time 0.3035 (0.2821) loss 4.2555 (3.7363) grad_norm 1.4529 (inf) [2021-04-15 23:15:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][1070/1251] eta 0:00:51 lr 0.000788 time 0.2577 (0.2821) loss 3.8134 (3.7380) grad_norm 1.2322 (inf) [2021-04-15 23:15:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][1080/1251] eta 0:00:48 lr 0.000788 time 0.2552 (0.2820) loss 3.8362 (3.7394) grad_norm 1.1504 (inf) [2021-04-15 23:15:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][1090/1251] eta 0:00:45 lr 0.000788 time 0.2674 (0.2820) loss 4.0426 (3.7407) grad_norm 1.1476 (inf) [2021-04-15 23:15:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][1100/1251] eta 0:00:42 lr 0.000788 time 0.2763 (0.2819) loss 3.7527 (3.7398) grad_norm 1.2373 (inf) [2021-04-15 23:15:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][1110/1251] eta 0:00:39 lr 0.000788 time 0.2851 (0.2819) loss 4.1179 (3.7401) grad_norm 1.0695 (inf) [2021-04-15 23:15:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][1120/1251] eta 0:00:36 lr 0.000788 time 0.2911 (0.2819) loss 2.9771 (3.7401) grad_norm 1.4800 (inf) [2021-04-15 23:15:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][1130/1251] eta 0:00:34 lr 0.000788 time 0.2567 (0.2819) loss 2.5773 (3.7363) grad_norm 1.2572 (inf) [2021-04-15 23:15:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][1140/1251] eta 0:00:31 lr 0.000788 time 0.2600 (0.2818) loss 3.3440 (3.7354) grad_norm 1.3003 (inf) [2021-04-15 23:15:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][1150/1251] eta 0:00:28 lr 0.000788 time 0.2876 (0.2819) loss 2.7373 (3.7342) grad_norm 1.5147 (inf) [2021-04-15 23:15:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][1160/1251] eta 0:00:25 lr 0.000788 time 0.2934 (0.2818) loss 3.7884 (3.7348) grad_norm 1.0498 (inf) [2021-04-15 23:15:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][1170/1251] eta 0:00:22 lr 0.000788 time 0.2511 (0.2818) loss 3.6319 (3.7343) grad_norm 1.3212 (inf) [2021-04-15 23:15:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][1180/1251] eta 0:00:20 lr 0.000788 time 0.2545 (0.2817) loss 3.3517 (3.7324) grad_norm 1.2587 (inf) [2021-04-15 23:15:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][1190/1251] eta 0:00:17 lr 0.000788 time 0.2701 (0.2817) loss 4.2177 (3.7328) grad_norm 1.3162 (inf) [2021-04-15 23:15:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][1200/1251] eta 0:00:14 lr 0.000788 time 0.2863 (0.2817) loss 4.1173 (3.7335) grad_norm 1.4133 (inf) [2021-04-15 23:15:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][1210/1251] eta 0:00:11 lr 0.000788 time 0.2943 (0.2817) loss 4.2814 (3.7342) grad_norm 1.3028 (inf) [2021-04-15 23:15:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][1220/1251] eta 0:00:08 lr 0.000788 time 0.2722 (0.2816) loss 3.4501 (3.7330) grad_norm 1.3442 (inf) [2021-04-15 23:16:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][1230/1251] eta 0:00:05 lr 0.000788 time 0.2791 (0.2816) loss 4.2609 (3.7338) grad_norm 1.3740 (inf) [2021-04-15 23:16:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][1240/1251] eta 0:00:03 lr 0.000788 time 0.2518 (0.2815) loss 3.4944 (3.7330) grad_norm 1.1111 (inf) [2021-04-15 23:16:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [91/300][1250/1251] eta 0:00:00 lr 0.000788 time 0.2486 (0.2813) loss 3.4138 (3.7312) grad_norm 1.2782 (inf) [2021-04-15 23:16:10 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 91 training takes 0:05:54 [2021-04-15 23:16:10 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_91.pth saving...... [2021-04-15 23:16:19 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_91.pth saved !!! [2021-04-15 23:16:20 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.086 (1.086) Loss 1.1438 (1.1438) Acc@1 72.461 (72.461) Acc@5 91.895 (91.895) [2021-04-15 23:16:21 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.136 (0.225) Loss 1.0810 (1.1408) Acc@1 74.414 (73.295) Acc@5 92.480 (91.983) [2021-04-15 23:16:23 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.130 (0.195) Loss 1.1168 (1.1572) Acc@1 73.730 (72.996) Acc@5 93.066 (91.662) [2021-04-15 23:16:26 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.106 (0.226) Loss 1.1564 (1.1598) Acc@1 73.438 (72.952) Acc@5 91.211 (91.617) [2021-04-15 23:16:28 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.206 (0.217) Loss 1.1384 (1.1551) Acc@1 74.902 (73.025) Acc@5 91.602 (91.694) [2021-04-15 23:16:31 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 73.040 Acc@5 91.736 [2021-04-15 23:16:31 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 73.0% [2021-04-15 23:16:31 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 73.37% [2021-04-15 23:16:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][0/1251] eta 1:05:42 lr 0.000788 time 3.1514 (3.1514) loss 2.9911 (2.9911) grad_norm 1.3263 (1.3263) [2021-04-15 23:16:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][10/1251] eta 0:11:13 lr 0.000787 time 0.2922 (0.5426) loss 3.6299 (3.7718) grad_norm 1.2398 (1.2187) [2021-04-15 23:16:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][20/1251] eta 0:08:40 lr 0.000787 time 0.2768 (0.4230) loss 2.5822 (3.6108) grad_norm 1.3987 (1.2258) [2021-04-15 23:16:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][30/1251] eta 0:07:39 lr 0.000787 time 0.2753 (0.3762) loss 4.2776 (3.5652) grad_norm 1.5711 (1.2865) [2021-04-15 23:16:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][40/1251] eta 0:07:09 lr 0.000787 time 0.2559 (0.3549) loss 3.7683 (3.6632) grad_norm 1.1489 (1.2623) [2021-04-15 23:16:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][50/1251] eta 0:06:48 lr 0.000787 time 0.2770 (0.3399) loss 3.1494 (3.6571) grad_norm 1.1234 (1.2484) [2021-04-15 23:16:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][60/1251] eta 0:06:34 lr 0.000787 time 0.3607 (0.3313) loss 2.6330 (3.6488) grad_norm 1.5523 (1.2568) [2021-04-15 23:16:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][70/1251] eta 0:06:21 lr 0.000787 time 0.2840 (0.3231) loss 4.0707 (3.7098) grad_norm 1.1389 (1.2511) [2021-04-15 23:16:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][80/1251] eta 0:06:12 lr 0.000787 time 0.2684 (0.3178) loss 3.5402 (3.7241) grad_norm 1.1596 (1.2532) [2021-04-15 23:16:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][90/1251] eta 0:06:03 lr 0.000787 time 0.2903 (0.3130) loss 4.1625 (3.7447) grad_norm 1.1872 (1.2479) [2021-04-15 23:17:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][100/1251] eta 0:05:55 lr 0.000787 time 0.2621 (0.3092) loss 3.3770 (3.7696) grad_norm 1.4076 (1.2508) [2021-04-15 23:17:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][110/1251] eta 0:05:49 lr 0.000787 time 0.3051 (0.3064) loss 3.9703 (3.7637) grad_norm 1.1892 (1.2492) [2021-04-15 23:17:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][120/1251] eta 0:05:43 lr 0.000787 time 0.2914 (0.3037) loss 3.4559 (3.7608) grad_norm 1.7213 (1.2456) [2021-04-15 23:17:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][130/1251] eta 0:05:39 lr 0.000787 time 0.2788 (0.3027) loss 4.3659 (3.7703) grad_norm 1.2784 (1.2452) [2021-04-15 23:17:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][140/1251] eta 0:05:34 lr 0.000787 time 0.2871 (0.3009) loss 4.4164 (3.7862) grad_norm 1.3470 (1.2446) [2021-04-15 23:17:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][150/1251] eta 0:05:30 lr 0.000787 time 0.2882 (0.3001) loss 3.2250 (3.7842) grad_norm 1.1672 (1.2478) [2021-04-15 23:17:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][160/1251] eta 0:05:25 lr 0.000787 time 0.2795 (0.2987) loss 4.1665 (3.7871) grad_norm 1.2356 (1.2474) [2021-04-15 23:17:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][170/1251] eta 0:05:21 lr 0.000787 time 0.2898 (0.2974) loss 4.4318 (3.7880) grad_norm 1.1800 (1.2463) [2021-04-15 23:17:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][180/1251] eta 0:05:17 lr 0.000787 time 0.2716 (0.2962) loss 2.6330 (3.7844) grad_norm 1.1671 (1.2470) [2021-04-15 23:17:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][190/1251] eta 0:05:13 lr 0.000787 time 0.2858 (0.2952) loss 4.2473 (3.7895) grad_norm 1.2087 (1.2527) [2021-04-15 23:17:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][200/1251] eta 0:05:09 lr 0.000787 time 0.2755 (0.2944) loss 2.4576 (3.7862) grad_norm 1.2502 (1.2532) [2021-04-15 23:17:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][210/1251] eta 0:05:06 lr 0.000787 time 0.4481 (0.2941) loss 3.6121 (3.7859) grad_norm 1.3629 (1.2512) [2021-04-15 23:17:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][220/1251] eta 0:05:03 lr 0.000787 time 0.2761 (0.2939) loss 3.7131 (3.7725) grad_norm 1.4718 (1.2608) [2021-04-15 23:17:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][230/1251] eta 0:04:59 lr 0.000787 time 0.2729 (0.2930) loss 4.8133 (3.7778) grad_norm 1.1189 (1.2572) [2021-04-15 23:17:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][240/1251] eta 0:04:55 lr 0.000787 time 0.3066 (0.2924) loss 3.8438 (3.7794) grad_norm 1.2630 (1.2543) [2021-04-15 23:17:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][250/1251] eta 0:04:52 lr 0.000787 time 0.2903 (0.2923) loss 3.6296 (3.7787) grad_norm 1.2092 (1.2569) [2021-04-15 23:17:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][260/1251] eta 0:04:48 lr 0.000787 time 0.2492 (0.2915) loss 4.3243 (3.7880) grad_norm 1.2491 (1.2561) [2021-04-15 23:17:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][270/1251] eta 0:04:45 lr 0.000787 time 0.3074 (0.2911) loss 3.9492 (3.7893) grad_norm 1.2135 (1.2575) [2021-04-15 23:17:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][280/1251] eta 0:04:42 lr 0.000787 time 0.2616 (0.2905) loss 3.2898 (3.7923) grad_norm 1.2629 (1.2579) [2021-04-15 23:17:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][290/1251] eta 0:04:38 lr 0.000787 time 0.2739 (0.2900) loss 3.7860 (3.7818) grad_norm 1.2454 (1.2570) [2021-04-15 23:17:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][300/1251] eta 0:04:35 lr 0.000786 time 0.2729 (0.2895) loss 3.9585 (3.7862) grad_norm 1.1408 (1.2549) [2021-04-15 23:18:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][310/1251] eta 0:04:31 lr 0.000786 time 0.2753 (0.2890) loss 3.9766 (3.7845) grad_norm 1.3162 (1.2535) [2021-04-15 23:18:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][320/1251] eta 0:04:29 lr 0.000786 time 0.2571 (0.2890) loss 3.6555 (3.7858) grad_norm 1.3828 (1.2532) [2021-04-15 23:18:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][330/1251] eta 0:04:25 lr 0.000786 time 0.2893 (0.2886) loss 3.5345 (3.7779) grad_norm 1.1800 (1.2529) [2021-04-15 23:18:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][340/1251] eta 0:04:22 lr 0.000786 time 0.2610 (0.2882) loss 4.1861 (3.7908) grad_norm 1.1114 (1.2524) [2021-04-15 23:18:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][350/1251] eta 0:04:19 lr 0.000786 time 0.2919 (0.2879) loss 4.0017 (3.7850) grad_norm 1.1768 (1.2503) [2021-04-15 23:18:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][360/1251] eta 0:04:16 lr 0.000786 time 0.2702 (0.2879) loss 3.2197 (3.7850) grad_norm 1.4680 (1.2493) [2021-04-15 23:18:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][370/1251] eta 0:04:13 lr 0.000786 time 0.2674 (0.2876) loss 3.8010 (3.7842) grad_norm 1.3077 (1.2541) [2021-04-15 23:18:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][380/1251] eta 0:04:10 lr 0.000786 time 0.2886 (0.2873) loss 3.9746 (3.7890) grad_norm 1.0499 (1.2511) [2021-04-15 23:18:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][390/1251] eta 0:04:07 lr 0.000786 time 0.2777 (0.2870) loss 4.0496 (3.7878) grad_norm 1.0902 (1.2506) [2021-04-15 23:18:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][400/1251] eta 0:04:03 lr 0.000786 time 0.2698 (0.2866) loss 3.4001 (3.7847) grad_norm 1.4288 (1.2505) [2021-04-15 23:18:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][410/1251] eta 0:04:00 lr 0.000786 time 0.2694 (0.2864) loss 4.4572 (3.7871) grad_norm 1.1700 (1.2489) [2021-04-15 23:18:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][420/1251] eta 0:03:57 lr 0.000786 time 0.2646 (0.2862) loss 4.1752 (3.7904) grad_norm 1.1278 (1.2498) [2021-04-15 23:18:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][430/1251] eta 0:03:54 lr 0.000786 time 0.3053 (0.2862) loss 3.4965 (3.7931) grad_norm 1.3408 (1.2515) [2021-04-15 23:18:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][440/1251] eta 0:03:51 lr 0.000786 time 0.2434 (0.2859) loss 3.8143 (3.7897) grad_norm 1.1387 (1.2491) [2021-04-15 23:18:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][450/1251] eta 0:03:48 lr 0.000786 time 0.2857 (0.2858) loss 4.1040 (3.7933) grad_norm 1.3425 (1.2468) [2021-04-15 23:18:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][460/1251] eta 0:03:45 lr 0.000786 time 0.2532 (0.2856) loss 3.8575 (3.7956) grad_norm 1.1127 (1.2449) [2021-04-15 23:18:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][470/1251] eta 0:03:42 lr 0.000786 time 0.2750 (0.2854) loss 4.0648 (3.7943) grad_norm 1.1658 (1.2442) [2021-04-15 23:18:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][480/1251] eta 0:03:39 lr 0.000786 time 0.2871 (0.2852) loss 4.1216 (3.7925) grad_norm 1.0418 (1.2427) [2021-04-15 23:18:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][490/1251] eta 0:03:36 lr 0.000786 time 0.3057 (0.2850) loss 2.7652 (3.7938) grad_norm 1.1331 (1.2425) [2021-04-15 23:18:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][500/1251] eta 0:03:34 lr 0.000786 time 0.2638 (0.2850) loss 2.8564 (3.7954) grad_norm 1.5167 (1.2430) [2021-04-15 23:18:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][510/1251] eta 0:03:31 lr 0.000786 time 0.2889 (0.2849) loss 3.8359 (3.7977) grad_norm 1.1053 (1.2428) [2021-04-15 23:18:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][520/1251] eta 0:03:28 lr 0.000786 time 0.2941 (0.2849) loss 2.7140 (3.7925) grad_norm 0.9662 (1.2433) [2021-04-15 23:19:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][530/1251] eta 0:03:25 lr 0.000786 time 0.2851 (0.2847) loss 3.7697 (3.7875) grad_norm 1.3767 (1.2433) [2021-04-15 23:19:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][540/1251] eta 0:03:22 lr 0.000786 time 0.2932 (0.2847) loss 4.7051 (3.7848) grad_norm 1.2884 (1.2442) [2021-04-15 23:19:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][550/1251] eta 0:03:19 lr 0.000786 time 0.2973 (0.2845) loss 4.5570 (3.7887) grad_norm 1.2063 (1.2437) [2021-04-15 23:19:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][560/1251] eta 0:03:16 lr 0.000786 time 0.2907 (0.2843) loss 3.7303 (3.7892) grad_norm 1.2544 (1.2432) [2021-04-15 23:19:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][570/1251] eta 0:03:13 lr 0.000786 time 0.2963 (0.2842) loss 3.1244 (3.7836) grad_norm 1.1192 (1.2426) [2021-04-15 23:19:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][580/1251] eta 0:03:10 lr 0.000786 time 0.2690 (0.2845) loss 3.8158 (3.7843) grad_norm 1.0484 (1.2419) [2021-04-15 23:19:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][590/1251] eta 0:03:08 lr 0.000785 time 0.2669 (0.2846) loss 3.8542 (3.7827) grad_norm 1.4244 (1.2412) [2021-04-15 23:19:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][600/1251] eta 0:03:05 lr 0.000785 time 0.2860 (0.2846) loss 3.8466 (3.7886) grad_norm 1.1489 (1.2401) [2021-04-15 23:19:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][610/1251] eta 0:03:02 lr 0.000785 time 0.2969 (0.2845) loss 4.4580 (3.7892) grad_norm 1.2155 (1.2380) [2021-04-15 23:19:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][620/1251] eta 0:02:59 lr 0.000785 time 0.2847 (0.2847) loss 3.9998 (3.7869) grad_norm 1.1681 (1.2383) [2021-04-15 23:19:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][630/1251] eta 0:02:56 lr 0.000785 time 0.2704 (0.2845) loss 2.5646 (3.7862) grad_norm 1.2054 (1.2375) [2021-04-15 23:19:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][640/1251] eta 0:02:53 lr 0.000785 time 0.2701 (0.2846) loss 3.5281 (3.7843) grad_norm 1.1595 (1.2381) [2021-04-15 23:19:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][650/1251] eta 0:02:50 lr 0.000785 time 0.2915 (0.2844) loss 3.8694 (3.7821) grad_norm 1.1898 (1.2373) [2021-04-15 23:19:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][660/1251] eta 0:02:47 lr 0.000785 time 0.2886 (0.2843) loss 3.7497 (3.7804) grad_norm 1.7332 (1.2376) [2021-04-15 23:19:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][670/1251] eta 0:02:45 lr 0.000785 time 0.2826 (0.2841) loss 2.2707 (3.7793) grad_norm 1.2849 (1.2388) [2021-04-15 23:19:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][680/1251] eta 0:02:42 lr 0.000785 time 0.2729 (0.2841) loss 4.1462 (3.7777) grad_norm 1.1740 (1.2389) [2021-04-15 23:19:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][690/1251] eta 0:02:39 lr 0.000785 time 0.2857 (0.2840) loss 4.0054 (3.7793) grad_norm 1.0171 (1.2382) [2021-04-15 23:19:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][700/1251] eta 0:02:36 lr 0.000785 time 0.4433 (0.2841) loss 4.1530 (3.7802) grad_norm 1.6562 (1.2387) [2021-04-15 23:19:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][710/1251] eta 0:02:33 lr 0.000785 time 0.2825 (0.2839) loss 4.1785 (3.7818) grad_norm 1.2606 (1.2396) [2021-04-15 23:19:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][720/1251] eta 0:02:30 lr 0.000785 time 0.2683 (0.2838) loss 3.8372 (3.7781) grad_norm 1.1279 (1.2379) [2021-04-15 23:19:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][730/1251] eta 0:02:27 lr 0.000785 time 0.2715 (0.2838) loss 3.6592 (3.7786) grad_norm 1.1407 (1.2380) [2021-04-15 23:20:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][740/1251] eta 0:02:24 lr 0.000785 time 0.2937 (0.2837) loss 4.4842 (3.7818) grad_norm 1.3482 (1.2381) [2021-04-15 23:20:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][750/1251] eta 0:02:22 lr 0.000785 time 0.2895 (0.2836) loss 4.1959 (3.7811) grad_norm 1.2134 (1.2380) [2021-04-15 23:20:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][760/1251] eta 0:02:19 lr 0.000785 time 0.2790 (0.2834) loss 3.5348 (3.7784) grad_norm 1.2442 (1.2379) [2021-04-15 23:20:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][770/1251] eta 0:02:16 lr 0.000785 time 0.2707 (0.2834) loss 3.9067 (3.7782) grad_norm 1.4888 (1.2385) [2021-04-15 23:20:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][780/1251] eta 0:02:13 lr 0.000785 time 0.2760 (0.2834) loss 4.3839 (3.7794) grad_norm 1.3732 (1.2394) [2021-04-15 23:20:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][790/1251] eta 0:02:10 lr 0.000785 time 0.2728 (0.2835) loss 3.5964 (3.7799) grad_norm 1.3459 (1.2392) [2021-04-15 23:20:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][800/1251] eta 0:02:07 lr 0.000785 time 0.2513 (0.2834) loss 3.0030 (3.7784) grad_norm 1.2540 (1.2397) [2021-04-15 23:20:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][810/1251] eta 0:02:04 lr 0.000785 time 0.2772 (0.2833) loss 3.6624 (3.7784) grad_norm 1.1726 (1.2407) [2021-04-15 23:20:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][820/1251] eta 0:02:02 lr 0.000785 time 0.2593 (0.2832) loss 4.4194 (3.7788) grad_norm 1.0355 (1.2410) [2021-04-15 23:20:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][830/1251] eta 0:01:59 lr 0.000785 time 0.2955 (0.2833) loss 2.7077 (3.7777) grad_norm 1.1227 (1.2408) [2021-04-15 23:20:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][840/1251] eta 0:01:56 lr 0.000785 time 0.2930 (0.2831) loss 4.7791 (3.7786) grad_norm 1.1101 (1.2408) [2021-04-15 23:20:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][850/1251] eta 0:01:53 lr 0.000785 time 0.2797 (0.2831) loss 4.2668 (3.7783) grad_norm 1.3809 (1.2400) [2021-04-15 23:20:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][860/1251] eta 0:01:50 lr 0.000785 time 0.2871 (0.2830) loss 2.8863 (3.7794) grad_norm 1.2658 (1.2396) [2021-04-15 23:20:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][870/1251] eta 0:01:47 lr 0.000785 time 0.2898 (0.2829) loss 3.9648 (3.7788) grad_norm 1.4319 (1.2404) [2021-04-15 23:20:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][880/1251] eta 0:01:44 lr 0.000785 time 0.2711 (0.2828) loss 3.7378 (3.7777) grad_norm 1.2391 (1.2400) [2021-04-15 23:20:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][890/1251] eta 0:01:42 lr 0.000784 time 0.2916 (0.2829) loss 3.6772 (3.7775) grad_norm 1.3679 (1.2401) [2021-04-15 23:20:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][900/1251] eta 0:01:39 lr 0.000784 time 0.2840 (0.2829) loss 2.3224 (3.7773) grad_norm 1.7380 (1.2412) [2021-04-15 23:20:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][910/1251] eta 0:01:36 lr 0.000784 time 0.3084 (0.2828) loss 3.1923 (3.7775) grad_norm 1.2484 (1.2417) [2021-04-15 23:20:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][920/1251] eta 0:01:33 lr 0.000784 time 0.2870 (0.2828) loss 3.4020 (3.7775) grad_norm 1.4368 (1.2424) [2021-04-15 23:20:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][930/1251] eta 0:01:30 lr 0.000784 time 0.2727 (0.2827) loss 4.4823 (3.7777) grad_norm 1.2764 (1.2423) [2021-04-15 23:20:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][940/1251] eta 0:01:27 lr 0.000784 time 0.2717 (0.2828) loss 3.6513 (3.7759) grad_norm 1.2460 (1.2421) [2021-04-15 23:21:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][950/1251] eta 0:01:25 lr 0.000784 time 0.2423 (0.2826) loss 4.0098 (3.7774) grad_norm 1.1564 (1.2418) [2021-04-15 23:21:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][960/1251] eta 0:01:22 lr 0.000784 time 0.2861 (0.2825) loss 3.6731 (3.7773) grad_norm 1.2587 (1.2418) [2021-04-15 23:21:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][970/1251] eta 0:01:19 lr 0.000784 time 0.2878 (0.2825) loss 4.2484 (3.7793) grad_norm 1.3828 (1.2416) [2021-04-15 23:21:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][980/1251] eta 0:01:16 lr 0.000784 time 0.2908 (0.2825) loss 3.8409 (3.7793) grad_norm 1.0991 (1.2410) [2021-04-15 23:21:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][990/1251] eta 0:01:13 lr 0.000784 time 0.2694 (0.2824) loss 3.1473 (3.7788) grad_norm 1.0905 (1.2409) [2021-04-15 23:21:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][1000/1251] eta 0:01:10 lr 0.000784 time 0.2982 (0.2823) loss 3.2087 (3.7773) grad_norm 1.2213 (1.2409) [2021-04-15 23:21:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][1010/1251] eta 0:01:08 lr 0.000784 time 0.2797 (0.2822) loss 2.8210 (3.7775) grad_norm 1.4079 (1.2412) [2021-04-15 23:21:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][1020/1251] eta 0:01:05 lr 0.000784 time 0.2679 (0.2822) loss 4.2399 (3.7786) grad_norm 1.3015 (1.2417) [2021-04-15 23:21:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][1030/1251] eta 0:01:02 lr 0.000784 time 0.2604 (0.2822) loss 4.2096 (3.7811) grad_norm 1.0943 (1.2422) [2021-04-15 23:21:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][1040/1251] eta 0:00:59 lr 0.000784 time 0.2893 (0.2821) loss 2.8608 (3.7791) grad_norm 1.1410 (1.2425) [2021-04-15 23:21:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][1050/1251] eta 0:00:56 lr 0.000784 time 0.2809 (0.2820) loss 4.3400 (3.7787) grad_norm 1.3415 (1.2417) [2021-04-15 23:21:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][1060/1251] eta 0:00:53 lr 0.000784 time 0.2861 (0.2820) loss 4.0881 (3.7777) grad_norm 1.0987 (1.2412) [2021-04-15 23:21:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][1070/1251] eta 0:00:51 lr 0.000784 time 0.2626 (0.2821) loss 3.3709 (3.7810) grad_norm 1.0306 (1.2405) [2021-04-15 23:21:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][1080/1251] eta 0:00:48 lr 0.000784 time 0.2797 (0.2821) loss 2.6765 (3.7801) grad_norm 1.1962 (1.2411) [2021-04-15 23:21:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][1090/1251] eta 0:00:45 lr 0.000784 time 0.2746 (0.2820) loss 4.3420 (3.7808) grad_norm 1.5905 (1.2416) [2021-04-15 23:21:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][1100/1251] eta 0:00:42 lr 0.000784 time 0.2843 (0.2820) loss 4.7314 (3.7790) grad_norm 1.1982 (1.2411) [2021-04-15 23:21:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][1110/1251] eta 0:00:39 lr 0.000784 time 0.2764 (0.2820) loss 4.1926 (3.7797) grad_norm 1.3924 (1.2419) [2021-04-15 23:21:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][1120/1251] eta 0:00:36 lr 0.000784 time 0.2729 (0.2819) loss 3.9094 (3.7792) grad_norm 1.1133 (1.2422) [2021-04-15 23:21:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][1130/1251] eta 0:00:34 lr 0.000784 time 0.2676 (0.2818) loss 3.0619 (3.7768) grad_norm 1.1120 (1.2423) [2021-04-15 23:21:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][1140/1251] eta 0:00:31 lr 0.000784 time 0.3007 (0.2818) loss 4.1669 (3.7753) grad_norm 1.4403 (1.2423) [2021-04-15 23:21:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][1150/1251] eta 0:00:28 lr 0.000784 time 0.4321 (0.2819) loss 3.5587 (3.7748) grad_norm 1.1075 (1.2423) [2021-04-15 23:21:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][1160/1251] eta 0:00:25 lr 0.000784 time 0.2801 (0.2821) loss 4.2007 (3.7742) grad_norm 1.0890 (1.2420) [2021-04-15 23:22:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][1170/1251] eta 0:00:22 lr 0.000784 time 0.2835 (0.2821) loss 3.7818 (3.7702) grad_norm 1.1433 (1.2418) [2021-04-15 23:22:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][1180/1251] eta 0:00:20 lr 0.000783 time 0.2712 (0.2820) loss 2.2989 (3.7689) grad_norm 1.2104 (1.2416) [2021-04-15 23:22:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][1190/1251] eta 0:00:17 lr 0.000783 time 0.2799 (0.2820) loss 3.3228 (3.7667) grad_norm 1.3315 (1.2426) [2021-04-15 23:22:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][1200/1251] eta 0:00:14 lr 0.000783 time 0.2794 (0.2820) loss 3.6482 (3.7652) grad_norm 1.2434 (1.2427) [2021-04-15 23:22:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][1210/1251] eta 0:00:11 lr 0.000783 time 0.2946 (0.2819) loss 4.3643 (3.7674) grad_norm 1.6589 (1.2429) [2021-04-15 23:22:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][1220/1251] eta 0:00:08 lr 0.000783 time 0.2620 (0.2820) loss 3.9867 (3.7680) grad_norm 1.0586 (1.2427) [2021-04-15 23:22:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][1230/1251] eta 0:00:05 lr 0.000783 time 0.2802 (0.2819) loss 4.1837 (3.7690) grad_norm 1.1276 (1.2424) [2021-04-15 23:22:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][1240/1251] eta 0:00:03 lr 0.000783 time 0.2483 (0.2818) loss 3.7890 (3.7668) grad_norm 1.1625 (1.2416) [2021-04-15 23:22:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [92/300][1250/1251] eta 0:00:00 lr 0.000783 time 0.2478 (0.2815) loss 3.4879 (3.7677) grad_norm 1.0499 (1.2411) [2021-04-15 23:22:25 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 92 training takes 0:05:54 [2021-04-15 23:22:26 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_92.pth saving...... [2021-04-15 23:22:36 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_92.pth saved !!! [2021-04-15 23:22:37 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.190 (1.190) Loss 1.1122 (1.1122) Acc@1 74.121 (74.121) Acc@5 92.578 (92.578) [2021-04-15 23:22:39 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.120 (0.257) Loss 1.1029 (1.1253) Acc@1 74.512 (73.517) Acc@5 91.992 (92.072) [2021-04-15 23:22:41 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.288 (0.237) Loss 1.1799 (1.1320) Acc@1 71.289 (73.419) Acc@5 91.602 (91.843) [2021-04-15 23:22:43 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.110 (0.229) Loss 1.1099 (1.1250) Acc@1 74.414 (73.428) Acc@5 91.797 (91.923) [2021-04-15 23:22:45 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.094 (0.218) Loss 1.1381 (1.1270) Acc@1 74.414 (73.426) Acc@5 91.699 (91.968) [2021-04-15 23:22:48 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 73.182 Acc@5 91.890 [2021-04-15 23:22:48 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 73.2% [2021-04-15 23:22:48 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 73.37% [2021-04-15 23:22:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][0/1251] eta 2:03:27 lr 0.000783 time 5.9210 (5.9210) loss 4.1949 (4.1949) grad_norm 1.1961 (1.1961) [2021-04-15 23:22:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][10/1251] eta 0:16:34 lr 0.000783 time 0.2441 (0.8016) loss 4.6722 (4.1322) grad_norm 1.1271 (1.2075) [2021-04-15 23:23:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][20/1251] eta 0:11:30 lr 0.000783 time 0.2844 (0.5609) loss 3.6141 (4.0698) grad_norm 1.2409 (1.2349) [2021-04-15 23:23:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][30/1251] eta 0:09:37 lr 0.000783 time 0.2689 (0.4733) loss 4.0495 (3.9252) grad_norm 1.0843 (1.2097) [2021-04-15 23:23:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][40/1251] eta 0:08:38 lr 0.000783 time 0.2785 (0.4284) loss 4.0137 (3.9051) grad_norm 1.5448 (1.2173) [2021-04-15 23:23:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][50/1251] eta 0:07:58 lr 0.000783 time 0.2756 (0.3983) loss 2.8757 (3.8167) grad_norm 1.1515 (1.2221) [2021-04-15 23:23:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][60/1251] eta 0:07:31 lr 0.000783 time 0.2741 (0.3792) loss 3.4823 (3.7231) grad_norm 1.1778 (1.2251) [2021-04-15 23:23:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][70/1251] eta 0:07:11 lr 0.000783 time 0.2629 (0.3650) loss 3.5284 (3.6877) grad_norm 1.0110 (1.2289) [2021-04-15 23:23:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][80/1251] eta 0:06:54 lr 0.000783 time 0.2846 (0.3541) loss 4.1395 (3.7189) grad_norm 1.3031 (1.2396) [2021-04-15 23:23:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][90/1251] eta 0:06:43 lr 0.000783 time 0.2777 (0.3479) loss 2.3103 (3.7235) grad_norm 1.1827 (1.2447) [2021-04-15 23:23:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][100/1251] eta 0:06:33 lr 0.000783 time 0.2831 (0.3418) loss 4.3571 (3.7609) grad_norm 1.1151 (1.2335) [2021-04-15 23:23:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][110/1251] eta 0:06:23 lr 0.000783 time 0.2695 (0.3360) loss 4.2374 (3.7486) grad_norm 1.2065 (1.2291) [2021-04-15 23:23:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][120/1251] eta 0:06:14 lr 0.000783 time 0.2743 (0.3310) loss 2.7773 (3.7553) grad_norm 1.4095 (1.2336) [2021-04-15 23:23:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][130/1251] eta 0:06:05 lr 0.000783 time 0.2698 (0.3264) loss 4.3681 (3.7684) grad_norm 1.0924 (1.2341) [2021-04-15 23:23:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][140/1251] eta 0:05:59 lr 0.000783 time 0.2937 (0.3237) loss 3.2801 (3.7649) grad_norm 1.2541 (1.2312) [2021-04-15 23:23:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][150/1251] eta 0:05:53 lr 0.000783 time 0.3048 (0.3211) loss 3.0525 (3.7568) grad_norm 1.2194 (1.2333) [2021-04-15 23:23:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][160/1251] eta 0:05:47 lr 0.000783 time 0.2551 (0.3187) loss 3.9433 (3.7720) grad_norm 1.3543 (1.2381) [2021-04-15 23:23:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][170/1251] eta 0:05:41 lr 0.000783 time 0.2584 (0.3162) loss 3.4074 (3.7565) grad_norm 1.1620 (1.2371) [2021-04-15 23:23:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][180/1251] eta 0:05:36 lr 0.000783 time 0.2692 (0.3138) loss 4.0403 (3.7554) grad_norm 1.0898 (1.2355) [2021-04-15 23:23:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][190/1251] eta 0:05:31 lr 0.000783 time 0.2513 (0.3121) loss 2.5958 (3.7168) grad_norm 1.3006 (1.2379) [2021-04-15 23:23:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][200/1251] eta 0:05:26 lr 0.000783 time 0.2643 (0.3104) loss 4.0913 (3.7253) grad_norm 1.2347 (1.2371) [2021-04-15 23:23:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][210/1251] eta 0:05:21 lr 0.000783 time 0.2507 (0.3088) loss 3.4823 (3.7199) grad_norm 1.2463 (1.2378) [2021-04-15 23:23:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][220/1251] eta 0:05:16 lr 0.000782 time 0.2440 (0.3071) loss 2.9293 (3.7243) grad_norm 1.4237 (1.2431) [2021-04-15 23:23:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][230/1251] eta 0:05:12 lr 0.000782 time 0.2772 (0.3060) loss 3.9849 (3.7281) grad_norm 1.1718 (1.2454) [2021-04-15 23:24:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][240/1251] eta 0:05:08 lr 0.000782 time 0.2757 (0.3047) loss 3.6009 (3.7248) grad_norm 1.0740 (1.2438) [2021-04-15 23:24:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][250/1251] eta 0:05:04 lr 0.000782 time 0.2831 (0.3038) loss 3.6095 (3.7173) grad_norm 1.1879 (1.2431) [2021-04-15 23:24:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][260/1251] eta 0:04:59 lr 0.000782 time 0.2788 (0.3026) loss 3.5092 (3.7026) grad_norm 1.1463 (1.2397) [2021-04-15 23:24:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][270/1251] eta 0:04:56 lr 0.000782 time 0.2583 (0.3019) loss 4.1503 (3.7114) grad_norm 1.3667 (1.2416) [2021-04-15 23:24:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][280/1251] eta 0:04:52 lr 0.000782 time 0.2834 (0.3012) loss 3.9032 (3.7123) grad_norm 1.5600 (1.2457) [2021-04-15 23:24:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][290/1251] eta 0:04:48 lr 0.000782 time 0.2521 (0.3002) loss 4.0993 (3.7110) grad_norm 1.1940 (1.2466) [2021-04-15 23:24:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][300/1251] eta 0:04:44 lr 0.000782 time 0.2880 (0.2994) loss 2.5732 (3.7134) grad_norm 1.4750 (1.2480) [2021-04-15 23:24:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][310/1251] eta 0:04:41 lr 0.000782 time 0.2869 (0.2987) loss 3.6085 (3.7140) grad_norm 1.4221 (1.2515) [2021-04-15 23:24:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][320/1251] eta 0:04:37 lr 0.000782 time 0.2911 (0.2981) loss 4.6923 (3.7102) grad_norm 1.1826 (1.2504) [2021-04-15 23:24:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][330/1251] eta 0:04:33 lr 0.000782 time 0.2647 (0.2974) loss 3.5859 (3.7015) grad_norm 1.2046 (1.2502) [2021-04-15 23:24:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][340/1251] eta 0:04:30 lr 0.000782 time 0.2923 (0.2970) loss 3.8737 (3.7006) grad_norm 1.8176 (1.2540) [2021-04-15 23:24:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][350/1251] eta 0:04:27 lr 0.000782 time 0.4224 (0.2969) loss 2.5934 (3.7042) grad_norm 1.3149 (1.2557) [2021-04-15 23:24:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][360/1251] eta 0:04:24 lr 0.000782 time 0.2759 (0.2967) loss 3.0055 (3.7084) grad_norm 1.1471 (1.2555) [2021-04-15 23:24:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][370/1251] eta 0:04:21 lr 0.000782 time 0.2843 (0.2967) loss 3.8756 (3.7142) grad_norm 1.4105 (1.2567) [2021-04-15 23:24:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][380/1251] eta 0:04:18 lr 0.000782 time 0.2982 (0.2964) loss 3.5129 (3.7126) grad_norm 1.1146 (1.2547) [2021-04-15 23:24:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][390/1251] eta 0:04:15 lr 0.000782 time 0.2782 (0.2965) loss 3.0365 (3.7099) grad_norm 1.0216 (1.2548) [2021-04-15 23:24:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][400/1251] eta 0:04:11 lr 0.000782 time 0.2794 (0.2960) loss 4.5022 (3.7081) grad_norm 1.1761 (1.2530) [2021-04-15 23:24:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][410/1251] eta 0:04:08 lr 0.000782 time 0.2767 (0.2956) loss 4.1637 (3.7089) grad_norm 1.1698 (1.2509) [2021-04-15 23:24:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][420/1251] eta 0:04:05 lr 0.000782 time 0.3083 (0.2956) loss 4.1877 (3.7154) grad_norm 1.1624 (1.2543) [2021-04-15 23:24:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][430/1251] eta 0:04:02 lr 0.000782 time 0.2570 (0.2953) loss 4.3444 (3.7138) grad_norm 1.1240 (1.2538) [2021-04-15 23:24:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][440/1251] eta 0:03:59 lr 0.000782 time 0.2879 (0.2950) loss 4.0692 (3.7132) grad_norm 1.7827 (1.2534) [2021-04-15 23:25:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][450/1251] eta 0:03:55 lr 0.000782 time 0.2760 (0.2946) loss 4.1317 (3.7115) grad_norm 1.1518 (1.2522) [2021-04-15 23:25:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][460/1251] eta 0:03:52 lr 0.000782 time 0.2851 (0.2944) loss 3.5409 (3.7105) grad_norm 1.2788 (1.2519) [2021-04-15 23:25:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][470/1251] eta 0:03:49 lr 0.000782 time 0.2625 (0.2941) loss 3.1746 (3.7065) grad_norm 1.1795 (1.2508) [2021-04-15 23:25:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][480/1251] eta 0:03:46 lr 0.000782 time 0.2806 (0.2936) loss 4.4146 (3.7074) grad_norm 1.3331 (1.2510) [2021-04-15 23:25:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][490/1251] eta 0:03:43 lr 0.000782 time 0.2676 (0.2933) loss 2.7135 (3.7104) grad_norm 1.0774 (1.2508) [2021-04-15 23:25:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][500/1251] eta 0:03:40 lr 0.000782 time 0.3018 (0.2930) loss 4.7825 (3.7171) grad_norm 1.1390 (1.2502) [2021-04-15 23:25:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][510/1251] eta 0:03:36 lr 0.000781 time 0.2748 (0.2927) loss 3.4609 (3.7178) grad_norm 1.2394 (1.2493) [2021-04-15 23:25:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][520/1251] eta 0:03:33 lr 0.000781 time 0.2727 (0.2924) loss 4.3485 (3.7165) grad_norm 1.1736 (1.2480) [2021-04-15 23:25:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][530/1251] eta 0:03:30 lr 0.000781 time 0.2626 (0.2921) loss 4.2675 (3.7146) grad_norm 1.1648 (1.2474) [2021-04-15 23:25:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][540/1251] eta 0:03:27 lr 0.000781 time 0.2663 (0.2918) loss 4.0278 (3.7192) grad_norm 1.0822 (1.2480) [2021-04-15 23:25:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][550/1251] eta 0:03:24 lr 0.000781 time 0.2923 (0.2915) loss 3.4231 (3.7126) grad_norm 1.2038 (1.2467) [2021-04-15 23:25:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][560/1251] eta 0:03:21 lr 0.000781 time 0.2591 (0.2913) loss 2.2996 (3.7120) grad_norm 1.7550 (1.2476) [2021-04-15 23:25:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][570/1251] eta 0:03:18 lr 0.000781 time 0.3615 (0.2915) loss 4.0614 (3.7094) grad_norm 1.0733 (1.2478) [2021-04-15 23:25:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][580/1251] eta 0:03:15 lr 0.000781 time 0.2658 (0.2912) loss 3.0450 (3.7119) grad_norm 1.4236 (1.2461) [2021-04-15 23:25:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][590/1251] eta 0:03:12 lr 0.000781 time 0.2562 (0.2913) loss 3.3092 (3.7106) grad_norm 1.2035 (1.2449) [2021-04-15 23:25:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][600/1251] eta 0:03:09 lr 0.000781 time 0.3772 (0.2913) loss 2.4249 (3.7032) grad_norm 1.2531 (1.2468) [2021-04-15 23:25:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][610/1251] eta 0:03:06 lr 0.000781 time 0.2803 (0.2911) loss 2.8665 (3.7013) grad_norm 1.1743 (1.2462) [2021-04-15 23:25:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][620/1251] eta 0:03:03 lr 0.000781 time 0.2775 (0.2911) loss 4.4491 (3.7039) grad_norm 1.2752 (1.2464) [2021-04-15 23:25:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][630/1251] eta 0:03:00 lr 0.000781 time 0.2558 (0.2910) loss 4.0607 (3.6997) grad_norm 1.1981 (1.2458) [2021-04-15 23:25:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][640/1251] eta 0:02:57 lr 0.000781 time 0.2658 (0.2908) loss 3.3154 (3.6955) grad_norm 1.2686 (1.2454) [2021-04-15 23:25:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][650/1251] eta 0:02:54 lr 0.000781 time 0.2808 (0.2905) loss 3.9295 (3.6983) grad_norm 1.3489 (1.2451) [2021-04-15 23:26:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][660/1251] eta 0:02:51 lr 0.000781 time 0.2672 (0.2902) loss 2.3327 (3.6943) grad_norm 1.1556 (1.2458) [2021-04-15 23:26:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][670/1251] eta 0:02:48 lr 0.000781 time 0.2850 (0.2902) loss 4.3979 (3.6969) grad_norm 1.3677 (1.2454) [2021-04-15 23:26:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][680/1251] eta 0:02:45 lr 0.000781 time 0.2871 (0.2902) loss 3.9451 (3.7015) grad_norm 1.5988 (1.2462) [2021-04-15 23:26:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][690/1251] eta 0:02:42 lr 0.000781 time 0.2879 (0.2901) loss 4.5237 (3.7033) grad_norm 1.1678 (1.2467) [2021-04-15 23:26:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][700/1251] eta 0:02:39 lr 0.000781 time 0.3065 (0.2900) loss 4.2307 (3.7048) grad_norm 1.2442 (1.2461) [2021-04-15 23:26:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][710/1251] eta 0:02:36 lr 0.000781 time 0.3082 (0.2901) loss 4.0397 (3.7073) grad_norm 1.4541 (1.2457) [2021-04-15 23:26:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][720/1251] eta 0:02:33 lr 0.000781 time 0.3035 (0.2899) loss 3.6019 (3.7072) grad_norm 1.1524 (1.2452) [2021-04-15 23:26:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][730/1251] eta 0:02:30 lr 0.000781 time 0.2417 (0.2898) loss 3.5562 (3.7078) grad_norm 1.2754 (1.2451) [2021-04-15 23:26:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][740/1251] eta 0:02:28 lr 0.000781 time 0.2780 (0.2896) loss 4.3256 (3.7061) grad_norm 1.1434 (1.2447) [2021-04-15 23:26:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][750/1251] eta 0:02:25 lr 0.000781 time 0.2795 (0.2894) loss 3.9376 (3.7105) grad_norm 1.1927 (1.2449) [2021-04-15 23:26:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][760/1251] eta 0:02:22 lr 0.000781 time 0.2963 (0.2894) loss 3.5071 (3.7111) grad_norm 1.1489 (1.2441) [2021-04-15 23:26:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][770/1251] eta 0:02:19 lr 0.000781 time 0.3014 (0.2892) loss 4.4801 (3.7122) grad_norm 1.3495 (1.2437) [2021-04-15 23:26:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][780/1251] eta 0:02:16 lr 0.000781 time 0.2730 (0.2890) loss 4.5098 (3.7134) grad_norm 1.2271 (1.2445) [2021-04-15 23:26:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][790/1251] eta 0:02:13 lr 0.000781 time 0.2451 (0.2888) loss 3.9943 (3.7146) grad_norm 1.5872 (1.2443) [2021-04-15 23:26:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][800/1251] eta 0:02:10 lr 0.000780 time 0.2411 (0.2887) loss 2.7577 (3.7166) grad_norm 1.1474 (1.2453) [2021-04-15 23:26:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][810/1251] eta 0:02:07 lr 0.000780 time 0.2415 (0.2886) loss 2.6548 (3.7147) grad_norm 1.1847 (1.2461) [2021-04-15 23:26:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][820/1251] eta 0:02:04 lr 0.000780 time 0.2854 (0.2884) loss 2.6302 (3.7145) grad_norm 1.3687 (1.2468) [2021-04-15 23:26:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][830/1251] eta 0:02:01 lr 0.000780 time 0.2633 (0.2883) loss 4.4502 (3.7171) grad_norm 1.4322 (1.2475) [2021-04-15 23:26:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][840/1251] eta 0:01:58 lr 0.000780 time 0.2796 (0.2882) loss 2.8250 (3.7125) grad_norm 1.2709 (inf) [2021-04-15 23:26:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][850/1251] eta 0:01:55 lr 0.000780 time 0.2722 (0.2880) loss 3.8072 (3.7126) grad_norm 1.1534 (inf) [2021-04-15 23:26:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][860/1251] eta 0:01:52 lr 0.000780 time 0.2615 (0.2879) loss 3.8590 (3.7182) grad_norm 1.5729 (inf) [2021-04-15 23:26:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][870/1251] eta 0:01:49 lr 0.000780 time 0.2709 (0.2877) loss 4.5764 (3.7197) grad_norm 1.2008 (inf) [2021-04-15 23:27:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][880/1251] eta 0:01:46 lr 0.000780 time 0.2832 (0.2876) loss 3.1830 (3.7209) grad_norm 1.1580 (inf) [2021-04-15 23:27:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][890/1251] eta 0:01:43 lr 0.000780 time 0.2812 (0.2875) loss 3.4933 (3.7234) grad_norm 1.1987 (inf) [2021-04-15 23:27:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][900/1251] eta 0:01:40 lr 0.000780 time 0.2826 (0.2873) loss 3.7201 (3.7240) grad_norm 1.3125 (inf) [2021-04-15 23:27:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][910/1251] eta 0:01:37 lr 0.000780 time 0.2748 (0.2872) loss 4.1719 (3.7243) grad_norm 1.2257 (inf) [2021-04-15 23:27:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][920/1251] eta 0:01:35 lr 0.000780 time 0.2853 (0.2872) loss 3.8116 (3.7251) grad_norm 1.2691 (inf) [2021-04-15 23:27:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][930/1251] eta 0:01:32 lr 0.000780 time 0.2788 (0.2872) loss 3.2226 (3.7246) grad_norm 1.2064 (inf) [2021-04-15 23:27:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][940/1251] eta 0:01:29 lr 0.000780 time 0.3723 (0.2874) loss 4.3189 (3.7264) grad_norm 1.3129 (inf) [2021-04-15 23:27:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][950/1251] eta 0:01:26 lr 0.000780 time 0.2700 (0.2873) loss 3.9088 (3.7269) grad_norm 1.0500 (inf) [2021-04-15 23:27:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][960/1251] eta 0:01:23 lr 0.000780 time 0.2718 (0.2873) loss 3.0133 (3.7272) grad_norm 1.1896 (inf) [2021-04-15 23:27:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][970/1251] eta 0:01:20 lr 0.000780 time 0.2838 (0.2874) loss 3.9068 (3.7294) grad_norm 1.3180 (inf) [2021-04-15 23:27:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][980/1251] eta 0:01:17 lr 0.000780 time 0.2774 (0.2873) loss 2.9886 (3.7292) grad_norm 1.3946 (inf) [2021-04-15 23:27:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][990/1251] eta 0:01:15 lr 0.000780 time 0.2544 (0.2874) loss 3.4651 (3.7313) grad_norm 1.0623 (inf) [2021-04-15 23:27:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][1000/1251] eta 0:01:12 lr 0.000780 time 0.2836 (0.2873) loss 2.6769 (3.7293) grad_norm 1.1208 (inf) [2021-04-15 23:27:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][1010/1251] eta 0:01:09 lr 0.000780 time 0.2819 (0.2872) loss 4.1183 (3.7283) grad_norm 1.3463 (inf) [2021-04-15 23:27:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][1020/1251] eta 0:01:06 lr 0.000780 time 0.2782 (0.2871) loss 2.5514 (3.7278) grad_norm 1.1140 (inf) [2021-04-15 23:27:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][1030/1251] eta 0:01:03 lr 0.000780 time 0.2501 (0.2870) loss 3.9731 (3.7307) grad_norm 1.3456 (inf) [2021-04-15 23:27:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][1040/1251] eta 0:01:00 lr 0.000780 time 0.2649 (0.2869) loss 3.6272 (3.7296) grad_norm 1.5594 (inf) [2021-04-15 23:27:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][1050/1251] eta 0:00:57 lr 0.000780 time 0.2607 (0.2868) loss 3.9411 (3.7334) grad_norm 1.2971 (inf) [2021-04-15 23:27:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][1060/1251] eta 0:00:54 lr 0.000780 time 0.2743 (0.2868) loss 3.0514 (3.7349) grad_norm 1.3936 (inf) [2021-04-15 23:27:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][1070/1251] eta 0:00:51 lr 0.000780 time 0.2892 (0.2867) loss 3.9120 (3.7347) grad_norm 1.1264 (inf) [2021-04-15 23:27:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][1080/1251] eta 0:00:49 lr 0.000780 time 0.2985 (0.2867) loss 3.8895 (3.7327) grad_norm 1.1958 (inf) [2021-04-15 23:28:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][1090/1251] eta 0:00:46 lr 0.000779 time 0.2821 (0.2866) loss 3.7339 (3.7325) grad_norm 1.0515 (inf) [2021-04-15 23:28:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][1100/1251] eta 0:00:43 lr 0.000779 time 0.2484 (0.2865) loss 2.6019 (3.7313) grad_norm 1.1052 (inf) [2021-04-15 23:28:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][1110/1251] eta 0:00:40 lr 0.000779 time 0.2735 (0.2864) loss 3.9188 (3.7333) grad_norm 1.0825 (inf) [2021-04-15 23:28:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][1120/1251] eta 0:00:37 lr 0.000779 time 0.2671 (0.2864) loss 3.9253 (3.7337) grad_norm 1.1819 (inf) [2021-04-15 23:28:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][1130/1251] eta 0:00:34 lr 0.000779 time 0.2974 (0.2863) loss 2.6694 (3.7338) grad_norm 1.2811 (inf) [2021-04-15 23:28:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][1140/1251] eta 0:00:31 lr 0.000779 time 0.2590 (0.2861) loss 3.7792 (3.7319) grad_norm 1.2878 (inf) [2021-04-15 23:28:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][1150/1251] eta 0:00:28 lr 0.000779 time 0.2582 (0.2862) loss 3.9521 (3.7323) grad_norm 1.3445 (inf) [2021-04-15 23:28:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][1160/1251] eta 0:00:26 lr 0.000779 time 0.3041 (0.2862) loss 4.0836 (3.7310) grad_norm 1.1775 (inf) [2021-04-15 23:28:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][1170/1251] eta 0:00:23 lr 0.000779 time 0.2486 (0.2863) loss 4.0343 (3.7322) grad_norm 1.4659 (inf) [2021-04-15 23:28:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][1180/1251] eta 0:00:20 lr 0.000779 time 0.2528 (0.2863) loss 3.0210 (3.7325) grad_norm 1.2687 (inf) [2021-04-15 23:28:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][1190/1251] eta 0:00:17 lr 0.000779 time 0.2862 (0.2862) loss 3.1061 (3.7308) grad_norm 1.4056 (inf) [2021-04-15 23:28:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][1200/1251] eta 0:00:14 lr 0.000779 time 0.2949 (0.2861) loss 4.6494 (3.7331) grad_norm 1.0563 (inf) [2021-04-15 23:28:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][1210/1251] eta 0:00:11 lr 0.000779 time 0.2947 (0.2860) loss 3.3968 (3.7338) grad_norm 1.3500 (inf) [2021-04-15 23:28:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][1220/1251] eta 0:00:08 lr 0.000779 time 0.2696 (0.2859) loss 4.0005 (3.7339) grad_norm 1.4377 (inf) [2021-04-15 23:28:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][1230/1251] eta 0:00:06 lr 0.000779 time 0.2837 (0.2858) loss 3.5702 (3.7354) grad_norm 1.1805 (inf) [2021-04-15 23:28:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][1240/1251] eta 0:00:03 lr 0.000779 time 0.3659 (0.2858) loss 4.2736 (3.7345) grad_norm 1.2498 (inf) [2021-04-15 23:28:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [93/300][1250/1251] eta 0:00:00 lr 0.000779 time 0.2574 (0.2855) loss 4.0193 (3.7338) grad_norm 1.0932 (inf) [2021-04-15 23:28:48 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 93 training takes 0:05:59 [2021-04-15 23:28:48 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_93.pth saving...... [2021-04-15 23:28:56 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_93.pth saved !!! [2021-04-15 23:28:57 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.123 (1.123) Loss 1.1482 (1.1482) Acc@1 71.680 (71.680) Acc@5 91.797 (91.797) [2021-04-15 23:28:58 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.110 (0.223) Loss 1.1622 (1.1076) Acc@1 70.898 (73.615) Acc@5 92.188 (92.267) [2021-04-15 23:29:01 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.713 (0.255) Loss 1.0899 (1.1184) Acc@1 74.414 (73.475) Acc@5 91.504 (92.150) [2021-04-15 23:29:03 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.144 (0.231) Loss 1.1492 (1.1309) Acc@1 74.023 (73.400) Acc@5 90.918 (91.917) [2021-04-15 23:29:05 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.215) Loss 1.1688 (1.1354) Acc@1 71.680 (73.259) Acc@5 91.504 (91.878) [2021-04-15 23:29:08 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 73.238 Acc@5 91.856 [2021-04-15 23:29:08 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 73.2% [2021-04-15 23:29:08 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 73.37% [2021-04-15 23:29:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][0/1251] eta 1:35:30 lr 0.000779 time 4.5808 (4.5808) loss 3.1945 (3.1945) grad_norm 1.1418 (1.1418) [2021-04-15 23:29:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][10/1251] eta 0:13:50 lr 0.000779 time 0.2974 (0.6690) loss 2.5392 (3.7252) grad_norm 1.2366 (1.2197) [2021-04-15 23:29:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][20/1251] eta 0:09:52 lr 0.000779 time 0.2794 (0.4814) loss 3.6891 (3.6642) grad_norm 1.1622 (1.2848) [2021-04-15 23:29:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][30/1251] eta 0:08:27 lr 0.000779 time 0.2447 (0.4153) loss 4.4938 (3.7390) grad_norm 1.2050 (1.2711) [2021-04-15 23:29:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][40/1251] eta 0:07:42 lr 0.000779 time 0.2806 (0.3817) loss 4.1555 (3.6875) grad_norm 1.2545 (1.2924) [2021-04-15 23:29:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][50/1251] eta 0:07:14 lr 0.000779 time 0.2777 (0.3614) loss 2.8129 (3.6702) grad_norm 1.3547 (1.2855) [2021-04-15 23:29:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][60/1251] eta 0:06:53 lr 0.000779 time 0.2631 (0.3472) loss 3.9641 (3.6506) grad_norm 1.3460 (1.2797) [2021-04-15 23:29:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][70/1251] eta 0:06:38 lr 0.000779 time 0.2689 (0.3377) loss 4.2333 (3.6267) grad_norm 1.0755 (1.2682) [2021-04-15 23:29:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][80/1251] eta 0:06:26 lr 0.000779 time 0.2680 (0.3301) loss 3.9185 (3.6354) grad_norm 1.3876 (1.2704) [2021-04-15 23:29:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][90/1251] eta 0:06:16 lr 0.000779 time 0.2570 (0.3244) loss 4.1996 (3.6407) grad_norm 1.1827 (1.2636) [2021-04-15 23:29:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][100/1251] eta 0:06:07 lr 0.000779 time 0.2948 (0.3196) loss 3.1908 (3.6429) grad_norm 1.6015 (1.2700) [2021-04-15 23:29:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][110/1251] eta 0:06:00 lr 0.000779 time 0.2686 (0.3157) loss 3.9495 (3.6508) grad_norm 1.2820 (1.2677) [2021-04-15 23:29:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][120/1251] eta 0:05:52 lr 0.000779 time 0.2672 (0.3120) loss 3.4024 (3.6525) grad_norm 1.3664 (1.2675) [2021-04-15 23:29:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][130/1251] eta 0:05:47 lr 0.000778 time 0.2632 (0.3099) loss 4.8938 (3.6812) grad_norm 1.1777 (1.2677) [2021-04-15 23:29:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][140/1251] eta 0:05:43 lr 0.000778 time 0.2474 (0.3094) loss 2.9636 (3.6773) grad_norm 1.0283 (1.2670) [2021-04-15 23:29:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][150/1251] eta 0:05:38 lr 0.000778 time 0.3870 (0.3078) loss 4.0204 (3.6849) grad_norm 1.4745 (1.2712) [2021-04-15 23:29:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][160/1251] eta 0:05:33 lr 0.000778 time 0.2826 (0.3056) loss 3.7566 (3.6998) grad_norm 1.1977 (1.2714) [2021-04-15 23:30:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][170/1251] eta 0:05:29 lr 0.000778 time 0.2997 (0.3049) loss 3.0155 (3.6999) grad_norm 1.2132 (1.2655) [2021-04-15 23:30:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][180/1251] eta 0:05:24 lr 0.000778 time 0.2609 (0.3034) loss 3.8891 (3.7156) grad_norm 1.2068 (1.2629) [2021-04-15 23:30:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][190/1251] eta 0:05:20 lr 0.000778 time 0.2802 (0.3021) loss 4.4287 (3.7143) grad_norm 1.1516 (1.2597) [2021-04-15 23:30:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][200/1251] eta 0:05:16 lr 0.000778 time 0.2672 (0.3012) loss 3.3438 (3.7124) grad_norm 1.6076 (1.2642) [2021-04-15 23:30:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][210/1251] eta 0:05:12 lr 0.000778 time 0.2888 (0.2999) loss 4.2391 (3.7135) grad_norm 1.6013 (1.2634) [2021-04-15 23:30:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][220/1251] eta 0:05:08 lr 0.000778 time 0.2739 (0.2990) loss 2.8804 (3.7151) grad_norm 1.4414 (1.2645) [2021-04-15 23:30:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][230/1251] eta 0:05:04 lr 0.000778 time 0.2683 (0.2981) loss 3.2638 (3.7131) grad_norm 1.2094 (1.2667) [2021-04-15 23:30:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][240/1251] eta 0:05:00 lr 0.000778 time 0.2896 (0.2971) loss 3.7091 (3.7084) grad_norm 1.2188 (1.2661) [2021-04-15 23:30:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][250/1251] eta 0:04:56 lr 0.000778 time 0.2602 (0.2962) loss 2.4998 (3.7020) grad_norm 1.0597 (1.2629) [2021-04-15 23:30:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][260/1251] eta 0:04:52 lr 0.000778 time 0.2846 (0.2955) loss 4.0089 (3.6981) grad_norm 1.2228 (1.2652) [2021-04-15 23:30:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][270/1251] eta 0:04:49 lr 0.000778 time 0.2809 (0.2950) loss 3.4451 (3.7076) grad_norm 1.2386 (1.2606) [2021-04-15 23:30:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][280/1251] eta 0:04:45 lr 0.000778 time 0.2733 (0.2944) loss 3.7569 (3.7080) grad_norm 1.1911 (1.2568) [2021-04-15 23:30:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][290/1251] eta 0:04:42 lr 0.000778 time 0.2753 (0.2937) loss 4.2518 (3.7076) grad_norm 1.1165 (1.2545) [2021-04-15 23:30:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][300/1251] eta 0:04:38 lr 0.000778 time 0.2510 (0.2931) loss 3.9493 (3.7178) grad_norm 1.1987 (1.2522) [2021-04-15 23:30:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][310/1251] eta 0:04:35 lr 0.000778 time 0.2494 (0.2926) loss 2.6623 (3.7229) grad_norm 1.4001 (1.2520) [2021-04-15 23:30:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][320/1251] eta 0:04:32 lr 0.000778 time 0.2912 (0.2922) loss 4.6633 (3.7174) grad_norm 1.3391 (1.2505) [2021-04-15 23:30:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][330/1251] eta 0:04:28 lr 0.000778 time 0.2659 (0.2916) loss 4.0784 (3.7210) grad_norm 1.3644 (1.2559) [2021-04-15 23:30:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][340/1251] eta 0:04:25 lr 0.000778 time 0.2645 (0.2911) loss 3.3493 (3.7194) grad_norm 1.1090 (1.2531) [2021-04-15 23:30:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][350/1251] eta 0:04:22 lr 0.000778 time 0.2928 (0.2909) loss 3.7787 (3.7170) grad_norm 1.1854 (1.2522) [2021-04-15 23:30:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][360/1251] eta 0:04:19 lr 0.000778 time 0.2900 (0.2909) loss 4.6938 (3.7204) grad_norm 1.3707 (1.2505) [2021-04-15 23:30:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][370/1251] eta 0:04:16 lr 0.000778 time 0.2816 (0.2911) loss 4.3214 (3.7222) grad_norm 1.0581 (1.2523) [2021-04-15 23:30:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][380/1251] eta 0:04:13 lr 0.000778 time 0.2733 (0.2913) loss 3.9437 (3.7175) grad_norm 1.1192 (1.2523) [2021-04-15 23:31:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][390/1251] eta 0:04:11 lr 0.000778 time 0.2973 (0.2915) loss 3.4615 (3.7216) grad_norm 1.1685 (1.2547) [2021-04-15 23:31:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][400/1251] eta 0:04:07 lr 0.000778 time 0.2680 (0.2912) loss 2.8318 (3.7111) grad_norm 1.2218 (1.2546) [2021-04-15 23:31:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][410/1251] eta 0:04:04 lr 0.000778 time 0.2820 (0.2907) loss 4.3669 (3.7176) grad_norm 1.3912 (1.2535) [2021-04-15 23:31:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][420/1251] eta 0:04:01 lr 0.000777 time 0.3008 (0.2904) loss 4.4190 (3.7216) grad_norm 1.2276 (1.2520) [2021-04-15 23:31:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][430/1251] eta 0:03:58 lr 0.000777 time 0.2681 (0.2900) loss 4.1972 (3.7287) grad_norm 1.1049 (1.2508) [2021-04-15 23:31:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][440/1251] eta 0:03:54 lr 0.000777 time 0.2721 (0.2897) loss 3.1670 (3.7288) grad_norm 1.2601 (1.2511) [2021-04-15 23:31:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][450/1251] eta 0:03:52 lr 0.000777 time 0.4204 (0.2897) loss 4.1549 (3.7230) grad_norm 1.1870 (1.2500) [2021-04-15 23:31:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][460/1251] eta 0:03:48 lr 0.000777 time 0.2551 (0.2894) loss 4.2633 (3.7265) grad_norm 1.5917 (1.2507) [2021-04-15 23:31:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][470/1251] eta 0:03:45 lr 0.000777 time 0.2830 (0.2891) loss 2.7901 (3.7263) grad_norm 1.2831 (1.2505) [2021-04-15 23:31:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][480/1251] eta 0:03:42 lr 0.000777 time 0.2851 (0.2890) loss 2.8146 (3.7301) grad_norm 1.4542 (1.2510) [2021-04-15 23:31:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][490/1251] eta 0:03:39 lr 0.000777 time 0.2955 (0.2888) loss 4.1755 (3.7298) grad_norm 1.0413 (1.2510) [2021-04-15 23:31:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][500/1251] eta 0:03:36 lr 0.000777 time 0.2765 (0.2884) loss 3.4126 (3.7280) grad_norm 1.3103 (1.2536) [2021-04-15 23:31:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][510/1251] eta 0:03:33 lr 0.000777 time 0.2657 (0.2882) loss 3.4689 (3.7253) grad_norm 1.1685 (1.2527) [2021-04-15 23:31:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][520/1251] eta 0:03:30 lr 0.000777 time 0.2732 (0.2879) loss 3.4000 (3.7248) grad_norm 1.3138 (1.2527) [2021-04-15 23:31:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][530/1251] eta 0:03:27 lr 0.000777 time 0.2607 (0.2877) loss 2.7643 (3.7262) grad_norm 1.5470 (1.2533) [2021-04-15 23:31:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][540/1251] eta 0:03:24 lr 0.000777 time 0.2772 (0.2874) loss 4.3251 (3.7291) grad_norm 1.4876 (1.2539) [2021-04-15 23:31:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][550/1251] eta 0:03:21 lr 0.000777 time 0.2744 (0.2872) loss 3.9870 (3.7350) grad_norm 1.2004 (1.2528) [2021-04-15 23:31:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][560/1251] eta 0:03:18 lr 0.000777 time 0.2765 (0.2870) loss 3.9274 (3.7394) grad_norm 1.1270 (1.2522) [2021-04-15 23:31:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][570/1251] eta 0:03:15 lr 0.000777 time 0.2731 (0.2868) loss 3.9463 (3.7385) grad_norm 1.0125 (1.2499) [2021-04-15 23:31:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][580/1251] eta 0:03:12 lr 0.000777 time 0.2551 (0.2868) loss 3.8878 (3.7374) grad_norm 1.3231 (1.2493) [2021-04-15 23:31:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][590/1251] eta 0:03:09 lr 0.000777 time 0.2701 (0.2867) loss 4.3789 (3.7422) grad_norm 1.1717 (1.2491) [2021-04-15 23:32:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][600/1251] eta 0:03:06 lr 0.000777 time 0.2650 (0.2865) loss 3.7396 (3.7439) grad_norm 1.4028 (1.2499) [2021-04-15 23:32:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][610/1251] eta 0:03:03 lr 0.000777 time 0.2785 (0.2864) loss 4.3047 (3.7451) grad_norm 1.0804 (1.2484) [2021-04-15 23:32:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][620/1251] eta 0:03:00 lr 0.000777 time 0.2857 (0.2862) loss 4.9398 (3.7483) grad_norm 1.2527 (1.2482) [2021-04-15 23:32:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][630/1251] eta 0:02:57 lr 0.000777 time 0.2750 (0.2860) loss 4.1285 (3.7539) grad_norm 1.4744 (1.2474) [2021-04-15 23:32:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][640/1251] eta 0:02:54 lr 0.000777 time 0.2608 (0.2858) loss 3.6156 (3.7529) grad_norm 1.1326 (1.2482) [2021-04-15 23:32:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][650/1251] eta 0:02:51 lr 0.000777 time 0.2837 (0.2857) loss 3.1734 (3.7473) grad_norm 1.2978 (1.2481) [2021-04-15 23:32:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][660/1251] eta 0:02:48 lr 0.000777 time 0.2768 (0.2855) loss 3.0700 (3.7439) grad_norm 1.0933 (1.2475) [2021-04-15 23:32:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][670/1251] eta 0:02:45 lr 0.000777 time 0.2530 (0.2853) loss 4.1247 (3.7457) grad_norm 1.4205 (1.2475) [2021-04-15 23:32:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][680/1251] eta 0:02:42 lr 0.000777 time 0.2813 (0.2852) loss 3.4722 (3.7466) grad_norm 1.6001 (1.2476) [2021-04-15 23:32:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][690/1251] eta 0:02:40 lr 0.000777 time 0.2808 (0.2852) loss 4.1947 (3.7482) grad_norm 1.1084 (1.2485) [2021-04-15 23:32:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][700/1251] eta 0:02:37 lr 0.000777 time 0.2685 (0.2853) loss 4.0005 (3.7470) grad_norm 1.2361 (1.2489) [2021-04-15 23:32:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][710/1251] eta 0:02:34 lr 0.000776 time 0.3044 (0.2856) loss 3.1989 (3.7446) grad_norm 1.1336 (1.2484) [2021-04-15 23:32:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][720/1251] eta 0:02:31 lr 0.000776 time 0.2779 (0.2858) loss 3.0383 (3.7462) grad_norm 1.3581 (1.2490) [2021-04-15 23:32:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][730/1251] eta 0:02:28 lr 0.000776 time 0.3143 (0.2857) loss 3.9817 (3.7459) grad_norm 1.3317 (1.2487) [2021-04-15 23:32:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][740/1251] eta 0:02:26 lr 0.000776 time 0.2568 (0.2857) loss 4.5694 (3.7482) grad_norm 1.2284 (1.2484) [2021-04-15 23:32:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][750/1251] eta 0:02:23 lr 0.000776 time 0.2645 (0.2856) loss 2.9961 (3.7482) grad_norm 1.2456 (1.2478) [2021-04-15 23:32:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][760/1251] eta 0:02:20 lr 0.000776 time 0.2763 (0.2855) loss 4.3994 (3.7526) grad_norm 1.0984 (1.2478) [2021-04-15 23:32:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][770/1251] eta 0:02:17 lr 0.000776 time 0.2899 (0.2853) loss 3.6826 (3.7455) grad_norm 1.2362 (1.2488) [2021-04-15 23:32:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][780/1251] eta 0:02:14 lr 0.000776 time 0.2702 (0.2853) loss 3.8862 (3.7435) grad_norm 1.3493 (1.2487) [2021-04-15 23:32:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][790/1251] eta 0:02:11 lr 0.000776 time 0.2710 (0.2852) loss 4.5438 (3.7451) grad_norm 1.2774 (1.2488) [2021-04-15 23:32:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][800/1251] eta 0:02:08 lr 0.000776 time 0.3057 (0.2851) loss 3.9624 (3.7423) grad_norm 1.3014 (1.2490) [2021-04-15 23:32:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][810/1251] eta 0:02:05 lr 0.000776 time 0.2483 (0.2849) loss 4.1254 (3.7425) grad_norm 1.3507 (1.2488) [2021-04-15 23:33:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][820/1251] eta 0:02:02 lr 0.000776 time 0.2747 (0.2848) loss 4.0283 (3.7441) grad_norm 1.1768 (1.2487) [2021-04-15 23:33:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][830/1251] eta 0:01:59 lr 0.000776 time 0.2745 (0.2847) loss 3.9696 (3.7426) grad_norm 1.3032 (1.2478) [2021-04-15 23:33:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][840/1251] eta 0:01:56 lr 0.000776 time 0.2797 (0.2846) loss 4.0784 (3.7440) grad_norm 1.3276 (1.2481) [2021-04-15 23:33:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][850/1251] eta 0:01:54 lr 0.000776 time 0.3153 (0.2845) loss 3.4999 (3.7421) grad_norm 1.1305 (1.2476) [2021-04-15 23:33:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][860/1251] eta 0:01:51 lr 0.000776 time 0.3001 (0.2844) loss 4.2769 (3.7410) grad_norm 1.2677 (1.2467) [2021-04-15 23:33:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][870/1251] eta 0:01:48 lr 0.000776 time 0.2854 (0.2843) loss 3.7192 (3.7395) grad_norm 1.1397 (1.2463) [2021-04-15 23:33:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][880/1251] eta 0:01:45 lr 0.000776 time 0.2707 (0.2842) loss 3.2502 (3.7363) grad_norm 1.1035 (1.2468) [2021-04-15 23:33:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][890/1251] eta 0:01:42 lr 0.000776 time 0.3129 (0.2841) loss 4.3453 (3.7354) grad_norm 1.1540 (1.2460) [2021-04-15 23:33:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][900/1251] eta 0:01:39 lr 0.000776 time 0.2634 (0.2840) loss 3.2855 (3.7360) grad_norm 1.1243 (1.2462) [2021-04-15 23:33:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][910/1251] eta 0:01:36 lr 0.000776 time 0.2925 (0.2840) loss 3.6781 (3.7377) grad_norm 1.2603 (1.2462) [2021-04-15 23:33:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][920/1251] eta 0:01:33 lr 0.000776 time 0.2687 (0.2839) loss 3.9863 (3.7389) grad_norm 1.2446 (1.2454) [2021-04-15 23:33:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][930/1251] eta 0:01:31 lr 0.000776 time 0.3020 (0.2838) loss 2.3210 (3.7377) grad_norm 1.2418 (1.2460) [2021-04-15 23:33:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][940/1251] eta 0:01:28 lr 0.000776 time 0.4513 (0.2840) loss 2.8683 (3.7364) grad_norm 1.2687 (1.2465) [2021-04-15 23:33:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][950/1251] eta 0:01:25 lr 0.000776 time 0.4104 (0.2841) loss 4.4998 (3.7377) grad_norm 1.1119 (1.2479) [2021-04-15 23:33:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][960/1251] eta 0:01:22 lr 0.000776 time 0.2976 (0.2842) loss 3.2298 (3.7384) grad_norm 1.2109 (1.2476) [2021-04-15 23:33:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][970/1251] eta 0:01:19 lr 0.000776 time 0.2515 (0.2843) loss 3.8031 (3.7363) grad_norm 1.2546 (1.2471) [2021-04-15 23:33:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][980/1251] eta 0:01:17 lr 0.000776 time 0.2799 (0.2842) loss 3.9081 (3.7370) grad_norm 1.2054 (1.2461) [2021-04-15 23:33:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][990/1251] eta 0:01:14 lr 0.000776 time 0.2884 (0.2841) loss 4.0652 (3.7390) grad_norm 1.2006 (1.2462) [2021-04-15 23:33:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][1000/1251] eta 0:01:11 lr 0.000775 time 0.2782 (0.2840) loss 3.3128 (3.7396) grad_norm 1.1689 (1.2453) [2021-04-15 23:33:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][1010/1251] eta 0:01:08 lr 0.000775 time 0.2843 (0.2839) loss 3.7038 (3.7419) grad_norm 1.4496 (1.2451) [2021-04-15 23:33:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][1020/1251] eta 0:01:05 lr 0.000775 time 0.2696 (0.2838) loss 3.8760 (3.7395) grad_norm 1.1381 (1.2456) [2021-04-15 23:34:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][1030/1251] eta 0:01:02 lr 0.000775 time 0.2863 (0.2840) loss 3.4542 (3.7392) grad_norm 1.7682 (1.2471) [2021-04-15 23:34:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][1040/1251] eta 0:00:59 lr 0.000775 time 0.2796 (0.2840) loss 3.2956 (3.7370) grad_norm 1.2581 (1.2480) [2021-04-15 23:34:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][1050/1251] eta 0:00:57 lr 0.000775 time 0.4148 (0.2840) loss 4.1117 (3.7389) grad_norm 1.0627 (1.2481) [2021-04-15 23:34:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][1060/1251] eta 0:00:54 lr 0.000775 time 0.2495 (0.2839) loss 4.0009 (3.7410) grad_norm 1.3889 (1.2484) [2021-04-15 23:34:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][1070/1251] eta 0:00:51 lr 0.000775 time 0.2820 (0.2838) loss 3.4306 (3.7393) grad_norm 1.1799 (1.2485) [2021-04-15 23:34:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][1080/1251] eta 0:00:48 lr 0.000775 time 0.2710 (0.2837) loss 4.0431 (3.7412) grad_norm 1.2327 (1.2488) [2021-04-15 23:34:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][1090/1251] eta 0:00:45 lr 0.000775 time 0.2665 (0.2836) loss 4.1348 (3.7418) grad_norm 1.5867 (1.2489) [2021-04-15 23:34:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][1100/1251] eta 0:00:42 lr 0.000775 time 0.2657 (0.2835) loss 4.4728 (3.7410) grad_norm 1.1121 (1.2491) [2021-04-15 23:34:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][1110/1251] eta 0:00:39 lr 0.000775 time 0.2705 (0.2835) loss 3.9603 (3.7385) grad_norm 1.4232 (1.2492) [2021-04-15 23:34:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][1120/1251] eta 0:00:37 lr 0.000775 time 0.2799 (0.2834) loss 4.1153 (3.7376) grad_norm 1.2218 (1.2487) [2021-04-15 23:34:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][1130/1251] eta 0:00:34 lr 0.000775 time 0.2966 (0.2834) loss 3.9729 (3.7375) grad_norm 1.2018 (1.2479) [2021-04-15 23:34:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][1140/1251] eta 0:00:31 lr 0.000775 time 0.3059 (0.2834) loss 3.9329 (3.7388) grad_norm 1.0882 (1.2476) [2021-04-15 23:34:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][1150/1251] eta 0:00:28 lr 0.000775 time 0.2763 (0.2833) loss 4.0893 (3.7407) grad_norm 1.1296 (1.2479) [2021-04-15 23:34:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][1160/1251] eta 0:00:25 lr 0.000775 time 0.3030 (0.2834) loss 4.0327 (3.7408) grad_norm 1.2245 (1.2480) [2021-04-15 23:34:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][1170/1251] eta 0:00:22 lr 0.000775 time 0.2531 (0.2833) loss 3.1308 (3.7416) grad_norm 1.5339 (1.2481) [2021-04-15 23:34:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][1180/1251] eta 0:00:20 lr 0.000775 time 0.2866 (0.2833) loss 2.7460 (3.7400) grad_norm 1.1699 (1.2482) [2021-04-15 23:34:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][1190/1251] eta 0:00:17 lr 0.000775 time 0.2870 (0.2832) loss 4.0133 (3.7391) grad_norm 1.2903 (1.2486) [2021-04-15 23:34:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][1200/1251] eta 0:00:14 lr 0.000775 time 0.2857 (0.2831) loss 3.0114 (3.7386) grad_norm 1.2312 (1.2485) [2021-04-15 23:34:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][1210/1251] eta 0:00:11 lr 0.000775 time 0.3081 (0.2831) loss 2.8090 (3.7375) grad_norm 1.1331 (1.2483) [2021-04-15 23:34:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][1220/1251] eta 0:00:08 lr 0.000775 time 0.2679 (0.2830) loss 3.9799 (3.7379) grad_norm 1.0694 (1.2486) [2021-04-15 23:34:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][1230/1251] eta 0:00:05 lr 0.000775 time 0.2882 (0.2830) loss 2.8841 (3.7356) grad_norm 1.3257 (1.2488) [2021-04-15 23:34:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][1240/1251] eta 0:00:03 lr 0.000775 time 0.2484 (0.2828) loss 3.9532 (3.7357) grad_norm 1.2881 (1.2489) [2021-04-15 23:35:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [94/300][1250/1251] eta 0:00:00 lr 0.000775 time 0.2485 (0.2826) loss 4.3963 (3.7362) grad_norm 1.3958 (1.2488) [2021-04-15 23:35:04 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 94 training takes 0:05:56 [2021-04-15 23:35:04 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_94.pth saving...... [2021-04-15 23:35:22 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_94.pth saved !!! [2021-04-15 23:35:23 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.324 (1.324) Loss 1.1446 (1.1446) Acc@1 73.340 (73.340) Acc@5 91.211 (91.211) [2021-04-15 23:35:25 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.130 (0.237) Loss 1.1300 (1.1112) Acc@1 73.730 (73.801) Acc@5 92.969 (92.241) [2021-04-15 23:35:27 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.580 (0.238) Loss 1.0957 (1.1182) Acc@1 74.805 (73.554) Acc@5 92.578 (92.122) [2021-04-15 23:35:30 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.096 (0.251) Loss 1.1560 (1.1221) Acc@1 72.559 (73.510) Acc@5 91.309 (92.162) [2021-04-15 23:35:31 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.219) Loss 1.1228 (1.1181) Acc@1 72.949 (73.490) Acc@5 92.090 (92.192) [2021-04-15 23:35:35 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 73.388 Acc@5 92.110 [2021-04-15 23:35:35 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 73.4% [2021-04-15 23:35:35 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 73.39% [2021-04-15 23:35:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][0/1251] eta 1:15:35 lr 0.000775 time 3.6258 (3.6258) loss 3.0023 (3.0023) grad_norm 1.2619 (1.2619) [2021-04-15 23:35:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][10/1251] eta 0:12:08 lr 0.000775 time 0.2883 (0.5871) loss 4.7468 (3.6906) grad_norm 1.2429 (1.1846) [2021-04-15 23:35:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][20/1251] eta 0:09:02 lr 0.000775 time 0.3086 (0.4406) loss 3.2908 (3.5881) grad_norm 1.2567 (1.1770) [2021-04-15 23:35:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][30/1251] eta 0:07:52 lr 0.000774 time 0.2765 (0.3873) loss 2.9132 (3.6330) grad_norm 1.1366 (1.2171) [2021-04-15 23:35:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][40/1251] eta 0:07:16 lr 0.000774 time 0.2996 (0.3607) loss 4.2940 (3.6673) grad_norm 1.2738 (1.2271) [2021-04-15 23:35:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][50/1251] eta 0:06:53 lr 0.000774 time 0.2975 (0.3441) loss 3.8837 (3.6365) grad_norm 1.4348 (1.2331) [2021-04-15 23:35:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][60/1251] eta 0:06:36 lr 0.000774 time 0.2624 (0.3328) loss 2.1524 (3.6365) grad_norm 1.3255 (1.2442) [2021-04-15 23:35:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][70/1251] eta 0:06:24 lr 0.000774 time 0.2683 (0.3254) loss 3.9465 (3.6124) grad_norm 1.1576 (1.2432) [2021-04-15 23:36:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][80/1251] eta 0:06:13 lr 0.000774 time 0.2501 (0.3193) loss 3.6557 (3.6364) grad_norm 1.0486 (1.2377) [2021-04-15 23:36:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][90/1251] eta 0:06:05 lr 0.000774 time 0.2827 (0.3150) loss 4.6253 (3.6886) grad_norm 1.1951 (1.2247) [2021-04-15 23:36:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][100/1251] eta 0:05:58 lr 0.000774 time 0.2435 (0.3112) loss 3.8536 (3.7146) grad_norm 1.2358 (1.2176) [2021-04-15 23:36:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][110/1251] eta 0:05:52 lr 0.000774 time 0.2848 (0.3086) loss 3.9787 (3.7145) grad_norm 1.0723 (1.2123) [2021-04-15 23:36:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][120/1251] eta 0:05:46 lr 0.000774 time 0.2685 (0.3063) loss 3.8669 (3.7290) grad_norm 1.4040 (1.2151) [2021-04-15 23:36:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][130/1251] eta 0:05:40 lr 0.000774 time 0.2742 (0.3039) loss 3.7830 (3.7382) grad_norm 1.4354 (1.2209) [2021-04-15 23:36:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][140/1251] eta 0:05:36 lr 0.000774 time 0.2430 (0.3028) loss 3.6966 (3.7251) grad_norm 1.1326 (1.2275) [2021-04-15 23:36:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][150/1251] eta 0:05:35 lr 0.000774 time 0.2815 (0.3044) loss 3.8231 (3.7175) grad_norm 1.3058 (1.2332) [2021-04-15 23:36:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][160/1251] eta 0:05:29 lr 0.000774 time 0.2747 (0.3023) loss 3.4019 (3.7286) grad_norm 1.4103 (1.2400) [2021-04-15 23:36:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][170/1251] eta 0:05:27 lr 0.000774 time 0.3847 (0.3027) loss 4.2123 (3.7479) grad_norm 1.4854 (1.2452) [2021-04-15 23:36:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][180/1251] eta 0:05:24 lr 0.000774 time 0.2852 (0.3032) loss 3.9727 (3.7531) grad_norm 1.4145 (1.2446) [2021-04-15 23:36:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][190/1251] eta 0:05:20 lr 0.000774 time 0.2656 (0.3021) loss 2.7546 (3.7429) grad_norm 1.0906 (1.2413) [2021-04-15 23:36:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][200/1251] eta 0:05:16 lr 0.000774 time 0.2955 (0.3009) loss 3.9972 (3.7519) grad_norm 1.1659 (1.2379) [2021-04-15 23:36:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][210/1251] eta 0:05:11 lr 0.000774 time 0.2844 (0.2996) loss 3.4241 (3.7486) grad_norm 1.2068 (1.2364) [2021-04-15 23:36:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][220/1251] eta 0:05:07 lr 0.000774 time 0.2653 (0.2986) loss 3.0133 (3.7344) grad_norm 1.2954 (1.2327) [2021-04-15 23:36:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][230/1251] eta 0:05:03 lr 0.000774 time 0.2969 (0.2976) loss 3.8357 (3.7423) grad_norm 1.0950 (1.2329) [2021-04-15 23:36:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][240/1251] eta 0:05:00 lr 0.000774 time 0.2688 (0.2967) loss 4.0787 (3.7621) grad_norm 1.5807 (1.2366) [2021-04-15 23:36:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][250/1251] eta 0:04:56 lr 0.000774 time 0.2731 (0.2967) loss 3.6515 (3.7663) grad_norm 1.2723 (1.2379) [2021-04-15 23:36:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][260/1251] eta 0:04:53 lr 0.000774 time 0.2543 (0.2957) loss 3.0041 (3.7552) grad_norm 1.0962 (1.2376) [2021-04-15 23:36:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][270/1251] eta 0:04:49 lr 0.000774 time 0.2524 (0.2953) loss 4.1329 (3.7585) grad_norm 1.1758 (1.2362) [2021-04-15 23:36:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][280/1251] eta 0:04:45 lr 0.000774 time 0.2711 (0.2944) loss 4.5029 (3.7653) grad_norm 1.1521 (1.2365) [2021-04-15 23:37:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][290/1251] eta 0:04:42 lr 0.000774 time 0.2952 (0.2938) loss 3.4025 (3.7636) grad_norm 1.1823 (1.2368) [2021-04-15 23:37:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][300/1251] eta 0:04:38 lr 0.000774 time 0.2984 (0.2933) loss 3.8161 (3.7622) grad_norm 1.1848 (1.2367) [2021-04-15 23:37:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][310/1251] eta 0:04:35 lr 0.000774 time 0.2944 (0.2929) loss 4.1154 (3.7592) grad_norm 1.1172 (1.2387) [2021-04-15 23:37:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][320/1251] eta 0:04:32 lr 0.000773 time 0.2699 (0.2926) loss 4.2097 (3.7561) grad_norm 1.1997 (1.2396) [2021-04-15 23:37:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][330/1251] eta 0:04:29 lr 0.000773 time 0.2552 (0.2923) loss 4.1778 (3.7631) grad_norm 1.6398 (1.2411) [2021-04-15 23:37:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][340/1251] eta 0:04:25 lr 0.000773 time 0.2989 (0.2919) loss 3.1422 (3.7517) grad_norm 1.2310 (1.2416) [2021-04-15 23:37:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][350/1251] eta 0:04:22 lr 0.000773 time 0.2551 (0.2916) loss 4.2718 (3.7456) grad_norm 1.2739 (1.2393) [2021-04-15 23:37:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][360/1251] eta 0:04:19 lr 0.000773 time 0.2526 (0.2915) loss 4.0363 (3.7495) grad_norm 1.1276 (1.2380) [2021-04-15 23:37:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][370/1251] eta 0:04:16 lr 0.000773 time 0.2928 (0.2914) loss 4.3940 (3.7415) grad_norm 1.3884 (1.2384) [2021-04-15 23:37:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][380/1251] eta 0:04:13 lr 0.000773 time 0.2739 (0.2911) loss 4.0070 (3.7434) grad_norm 1.3868 (1.2405) [2021-04-15 23:37:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][390/1251] eta 0:04:10 lr 0.000773 time 0.2669 (0.2908) loss 3.8986 (3.7358) grad_norm 1.3226 (1.2409) [2021-04-15 23:37:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][400/1251] eta 0:04:07 lr 0.000773 time 0.2914 (0.2906) loss 4.0550 (3.7307) grad_norm 1.3632 (1.2411) [2021-04-15 23:37:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][410/1251] eta 0:04:04 lr 0.000773 time 0.2864 (0.2904) loss 3.8976 (3.7269) grad_norm 1.4469 (1.2434) [2021-04-15 23:37:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][420/1251] eta 0:04:01 lr 0.000773 time 0.2894 (0.2900) loss 4.2572 (3.7256) grad_norm 1.1739 (1.2417) [2021-04-15 23:37:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][430/1251] eta 0:03:57 lr 0.000773 time 0.2758 (0.2898) loss 3.5215 (3.7220) grad_norm 1.1865 (1.2429) [2021-04-15 23:37:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][440/1251] eta 0:03:54 lr 0.000773 time 0.2991 (0.2896) loss 4.0156 (3.7283) grad_norm 1.3018 (1.2443) [2021-04-15 23:37:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][450/1251] eta 0:03:51 lr 0.000773 time 0.2672 (0.2892) loss 3.4646 (3.7310) grad_norm 1.2415 (1.2453) [2021-04-15 23:37:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][460/1251] eta 0:03:48 lr 0.000773 time 0.2754 (0.2890) loss 3.7421 (3.7319) grad_norm 1.0824 (1.2444) [2021-04-15 23:37:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][470/1251] eta 0:03:45 lr 0.000773 time 0.2470 (0.2887) loss 4.5739 (3.7286) grad_norm 1.2006 (1.2431) [2021-04-15 23:37:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][480/1251] eta 0:03:42 lr 0.000773 time 0.3004 (0.2884) loss 4.4162 (3.7300) grad_norm 1.1932 (1.2435) [2021-04-15 23:37:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][490/1251] eta 0:03:39 lr 0.000773 time 0.2731 (0.2882) loss 4.4540 (3.7281) grad_norm 1.1265 (1.2421) [2021-04-15 23:37:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][500/1251] eta 0:03:36 lr 0.000773 time 0.2517 (0.2880) loss 3.8980 (3.7323) grad_norm 1.2148 (1.2427) [2021-04-15 23:38:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][510/1251] eta 0:03:33 lr 0.000773 time 0.2821 (0.2878) loss 3.6208 (3.7343) grad_norm 1.2839 (1.2439) [2021-04-15 23:38:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][520/1251] eta 0:03:30 lr 0.000773 time 0.2787 (0.2877) loss 2.7907 (3.7297) grad_norm 1.0787 (1.2451) [2021-04-15 23:38:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][530/1251] eta 0:03:27 lr 0.000773 time 0.2612 (0.2875) loss 4.2615 (3.7326) grad_norm 1.3477 (1.2448) [2021-04-15 23:38:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][540/1251] eta 0:03:24 lr 0.000773 time 0.2835 (0.2875) loss 4.3140 (3.7255) grad_norm 1.1205 (1.2459) [2021-04-15 23:38:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][550/1251] eta 0:03:21 lr 0.000773 time 0.4142 (0.2876) loss 4.0473 (3.7253) grad_norm 1.0966 (1.2469) [2021-04-15 23:38:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][560/1251] eta 0:03:18 lr 0.000773 time 0.2570 (0.2873) loss 4.0681 (3.7231) grad_norm 1.4693 (1.2480) [2021-04-15 23:38:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][570/1251] eta 0:03:15 lr 0.000773 time 0.2593 (0.2877) loss 3.4878 (3.7270) grad_norm 1.1690 (1.2465) [2021-04-15 23:38:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][580/1251] eta 0:03:12 lr 0.000773 time 0.2734 (0.2875) loss 2.5631 (3.7222) grad_norm 1.4038 (1.2474) [2021-04-15 23:38:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][590/1251] eta 0:03:09 lr 0.000773 time 0.2595 (0.2873) loss 3.8978 (3.7292) grad_norm 1.5109 (1.2485) [2021-04-15 23:38:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][600/1251] eta 0:03:06 lr 0.000773 time 0.2603 (0.2872) loss 2.8149 (3.7246) grad_norm 1.1745 (1.2488) [2021-04-15 23:38:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][610/1251] eta 0:03:03 lr 0.000772 time 0.2722 (0.2870) loss 3.4309 (3.7229) grad_norm 1.4304 (1.2481) [2021-04-15 23:38:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][620/1251] eta 0:03:01 lr 0.000772 time 0.2713 (0.2869) loss 3.2245 (3.7210) grad_norm 1.3967 (1.2479) [2021-04-15 23:38:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][630/1251] eta 0:02:58 lr 0.000772 time 0.2836 (0.2867) loss 2.6571 (3.7204) grad_norm 1.1200 (1.2479) [2021-04-15 23:38:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][640/1251] eta 0:02:55 lr 0.000772 time 0.2758 (0.2866) loss 3.2021 (3.7231) grad_norm 1.4226 (1.2482) [2021-04-15 23:38:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][650/1251] eta 0:02:52 lr 0.000772 time 0.2665 (0.2865) loss 3.2630 (3.7267) grad_norm 1.1234 (1.2493) [2021-04-15 23:38:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][660/1251] eta 0:02:49 lr 0.000772 time 0.2669 (0.2862) loss 3.6274 (3.7284) grad_norm 1.3167 (1.2495) [2021-04-15 23:38:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][670/1251] eta 0:02:46 lr 0.000772 time 0.3025 (0.2863) loss 3.7472 (3.7261) grad_norm 1.2387 (1.2498) [2021-04-15 23:38:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][680/1251] eta 0:02:43 lr 0.000772 time 0.2673 (0.2861) loss 4.0562 (3.7255) grad_norm 1.1388 (1.2486) [2021-04-15 23:38:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][690/1251] eta 0:02:40 lr 0.000772 time 0.3770 (0.2862) loss 3.9066 (3.7246) grad_norm 1.1314 (1.2474) [2021-04-15 23:38:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][700/1251] eta 0:02:37 lr 0.000772 time 0.2945 (0.2862) loss 2.7789 (3.7240) grad_norm 1.3188 (1.2476) [2021-04-15 23:38:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][710/1251] eta 0:02:34 lr 0.000772 time 0.2594 (0.2861) loss 4.2505 (3.7273) grad_norm 1.1637 (1.2474) [2021-04-15 23:39:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][720/1251] eta 0:02:32 lr 0.000772 time 0.2759 (0.2863) loss 2.6761 (3.7241) grad_norm 1.2051 (1.2478) [2021-04-15 23:39:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][730/1251] eta 0:02:29 lr 0.000772 time 0.2662 (0.2864) loss 4.1604 (3.7230) grad_norm 1.3175 (1.2472) [2021-04-15 23:39:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][740/1251] eta 0:02:26 lr 0.000772 time 0.2575 (0.2865) loss 4.4708 (3.7220) grad_norm 1.2555 (1.2477) [2021-04-15 23:39:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][750/1251] eta 0:02:23 lr 0.000772 time 0.2600 (0.2863) loss 3.5581 (3.7220) grad_norm 1.3422 (1.2482) [2021-04-15 23:39:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][760/1251] eta 0:02:20 lr 0.000772 time 0.2683 (0.2864) loss 4.3899 (3.7236) grad_norm 1.8353 (1.2487) [2021-04-15 23:39:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][770/1251] eta 0:02:17 lr 0.000772 time 0.2899 (0.2863) loss 4.5435 (3.7250) grad_norm 1.3890 (1.2492) [2021-04-15 23:39:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][780/1251] eta 0:02:14 lr 0.000772 time 0.2691 (0.2861) loss 3.8534 (3.7281) grad_norm 1.2188 (1.2495) [2021-04-15 23:39:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][790/1251] eta 0:02:11 lr 0.000772 time 0.2986 (0.2860) loss 3.4261 (3.7252) grad_norm 1.1714 (1.2489) [2021-04-15 23:39:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][800/1251] eta 0:02:08 lr 0.000772 time 0.2565 (0.2860) loss 3.5943 (3.7251) grad_norm 1.1978 (1.2482) [2021-04-15 23:39:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][810/1251] eta 0:02:06 lr 0.000772 time 0.3078 (0.2858) loss 3.5599 (3.7244) grad_norm 1.3046 (1.2488) [2021-04-15 23:39:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][820/1251] eta 0:02:03 lr 0.000772 time 0.2915 (0.2857) loss 3.0556 (3.7215) grad_norm 1.3064 (1.2485) [2021-04-15 23:39:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][830/1251] eta 0:02:00 lr 0.000772 time 0.2683 (0.2859) loss 2.6807 (3.7180) grad_norm 1.2981 (1.2490) [2021-04-15 23:39:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][840/1251] eta 0:01:57 lr 0.000772 time 0.2586 (0.2857) loss 2.8570 (3.7173) grad_norm 1.3308 (1.2487) [2021-04-15 23:39:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][850/1251] eta 0:01:54 lr 0.000772 time 0.2625 (0.2856) loss 4.0120 (3.7180) grad_norm 1.2072 (1.2496) [2021-04-15 23:39:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][860/1251] eta 0:01:51 lr 0.000772 time 0.2568 (0.2855) loss 3.5749 (3.7153) grad_norm 1.2337 (1.2499) [2021-04-15 23:39:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][870/1251] eta 0:01:48 lr 0.000772 time 0.2815 (0.2855) loss 4.2136 (3.7136) grad_norm 1.2796 (1.2511) [2021-04-15 23:39:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][880/1251] eta 0:01:45 lr 0.000772 time 0.2551 (0.2854) loss 2.5999 (3.7141) grad_norm 1.0063 (1.2514) [2021-04-15 23:39:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][890/1251] eta 0:01:43 lr 0.000771 time 0.2805 (0.2854) loss 4.1275 (3.7150) grad_norm 1.3187 (1.2530) [2021-04-15 23:39:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][900/1251] eta 0:01:40 lr 0.000771 time 0.2572 (0.2852) loss 3.8491 (3.7148) grad_norm inf (inf) [2021-04-15 23:39:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][910/1251] eta 0:01:37 lr 0.000771 time 0.2758 (0.2851) loss 3.7659 (3.7148) grad_norm 1.3951 (inf) [2021-04-15 23:39:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][920/1251] eta 0:01:34 lr 0.000771 time 0.2964 (0.2850) loss 3.0394 (3.7162) grad_norm 1.2468 (inf) [2021-04-15 23:40:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][930/1251] eta 0:01:31 lr 0.000771 time 0.4042 (0.2851) loss 4.2134 (3.7184) grad_norm 1.1770 (inf) [2021-04-15 23:40:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][940/1251] eta 0:01:28 lr 0.000771 time 0.2672 (0.2849) loss 3.5954 (3.7191) grad_norm 1.1873 (inf) [2021-04-15 23:40:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][950/1251] eta 0:01:25 lr 0.000771 time 0.2677 (0.2849) loss 4.2549 (3.7178) grad_norm 1.1542 (inf) [2021-04-15 23:40:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][960/1251] eta 0:01:22 lr 0.000771 time 0.2680 (0.2848) loss 4.3527 (3.7190) grad_norm 1.0599 (inf) [2021-04-15 23:40:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][970/1251] eta 0:01:20 lr 0.000771 time 0.2714 (0.2847) loss 3.7422 (3.7172) grad_norm 1.2485 (inf) [2021-04-15 23:40:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][980/1251] eta 0:01:17 lr 0.000771 time 0.2725 (0.2847) loss 4.1425 (3.7201) grad_norm 1.2301 (inf) [2021-04-15 23:40:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][990/1251] eta 0:01:14 lr 0.000771 time 0.2950 (0.2847) loss 4.2447 (3.7228) grad_norm 1.2388 (inf) [2021-04-15 23:40:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][1000/1251] eta 0:01:11 lr 0.000771 time 0.2788 (0.2846) loss 2.7981 (3.7189) grad_norm 1.3152 (inf) [2021-04-15 23:40:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][1010/1251] eta 0:01:08 lr 0.000771 time 0.2814 (0.2846) loss 4.1588 (3.7173) grad_norm 1.1306 (inf) [2021-04-15 23:40:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][1020/1251] eta 0:01:05 lr 0.000771 time 0.2704 (0.2846) loss 3.3251 (3.7193) grad_norm 1.1288 (inf) [2021-04-15 23:40:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][1030/1251] eta 0:01:02 lr 0.000771 time 0.3016 (0.2846) loss 4.1505 (3.7194) grad_norm 1.3186 (inf) [2021-04-15 23:40:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][1040/1251] eta 0:01:00 lr 0.000771 time 0.2750 (0.2845) loss 2.3324 (3.7201) grad_norm 1.1694 (inf) [2021-04-15 23:40:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][1050/1251] eta 0:00:57 lr 0.000771 time 0.2792 (0.2844) loss 3.7729 (3.7176) grad_norm 1.1955 (inf) [2021-04-15 23:40:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][1060/1251] eta 0:00:54 lr 0.000771 time 0.2652 (0.2843) loss 3.8839 (3.7180) grad_norm 1.2586 (inf) [2021-04-15 23:40:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][1070/1251] eta 0:00:51 lr 0.000771 time 0.2551 (0.2842) loss 4.5302 (3.7203) grad_norm 1.1348 (inf) [2021-04-15 23:40:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][1080/1251] eta 0:00:48 lr 0.000771 time 0.3097 (0.2842) loss 3.9270 (3.7224) grad_norm 1.2277 (inf) [2021-04-15 23:40:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][1090/1251] eta 0:00:45 lr 0.000771 time 0.2856 (0.2841) loss 3.0472 (3.7230) grad_norm 1.2713 (inf) [2021-04-15 23:40:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][1100/1251] eta 0:00:42 lr 0.000771 time 0.2766 (0.2841) loss 2.7162 (3.7227) grad_norm 1.0833 (inf) [2021-04-15 23:40:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][1110/1251] eta 0:00:40 lr 0.000771 time 0.2969 (0.2840) loss 2.9126 (3.7191) grad_norm 1.3541 (inf) [2021-04-15 23:40:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][1120/1251] eta 0:00:37 lr 0.000771 time 0.2758 (0.2841) loss 4.2675 (3.7186) grad_norm 1.1613 (inf) [2021-04-15 23:40:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][1130/1251] eta 0:00:34 lr 0.000771 time 0.2623 (0.2841) loss 4.1330 (3.7194) grad_norm 1.1483 (inf) [2021-04-15 23:40:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][1140/1251] eta 0:00:31 lr 0.000771 time 0.2947 (0.2842) loss 3.7485 (3.7211) grad_norm 1.2575 (inf) [2021-04-15 23:41:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][1150/1251] eta 0:00:28 lr 0.000771 time 0.2819 (0.2842) loss 3.4474 (3.7198) grad_norm 1.2038 (inf) [2021-04-15 23:41:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][1160/1251] eta 0:00:25 lr 0.000771 time 0.3014 (0.2841) loss 3.5832 (3.7196) grad_norm 1.4226 (inf) [2021-04-15 23:41:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][1170/1251] eta 0:00:23 lr 0.000771 time 0.2921 (0.2840) loss 4.4048 (3.7214) grad_norm 1.1239 (inf) [2021-04-15 23:41:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][1180/1251] eta 0:00:20 lr 0.000770 time 0.2547 (0.2839) loss 3.9773 (3.7221) grad_norm 1.4366 (inf) [2021-04-15 23:41:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][1190/1251] eta 0:00:17 lr 0.000770 time 0.2948 (0.2838) loss 4.4425 (3.7228) grad_norm 1.1843 (inf) [2021-04-15 23:41:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][1200/1251] eta 0:00:14 lr 0.000770 time 0.2763 (0.2838) loss 4.2137 (3.7236) grad_norm 1.4437 (inf) [2021-04-15 23:41:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][1210/1251] eta 0:00:11 lr 0.000770 time 0.2513 (0.2838) loss 4.0537 (3.7243) grad_norm 1.4401 (inf) [2021-04-15 23:41:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][1220/1251] eta 0:00:08 lr 0.000770 time 0.2787 (0.2837) loss 4.5494 (3.7250) grad_norm 1.4429 (inf) [2021-04-15 23:41:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][1230/1251] eta 0:00:05 lr 0.000770 time 0.2682 (0.2837) loss 4.0281 (3.7264) grad_norm 1.6165 (inf) [2021-04-15 23:41:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][1240/1251] eta 0:00:03 lr 0.000770 time 0.2483 (0.2836) loss 4.3461 (3.7303) grad_norm 1.2658 (inf) [2021-04-15 23:41:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [95/300][1250/1251] eta 0:00:00 lr 0.000770 time 0.2492 (0.2833) loss 3.5137 (3.7318) grad_norm 1.1286 (inf) [2021-04-15 23:41:32 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 95 training takes 0:05:56 [2021-04-15 23:41:32 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_95.pth saving...... [2021-04-15 23:41:51 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_95.pth saved !!! [2021-04-15 23:41:52 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.315 (1.315) Loss 1.2045 (1.2045) Acc@1 71.387 (71.387) Acc@5 91.406 (91.406) [2021-04-15 23:41:53 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.164 (0.228) Loss 1.0890 (1.1490) Acc@1 73.730 (73.127) Acc@5 93.652 (92.303) [2021-04-15 23:41:56 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.112 (0.234) Loss 1.0983 (1.1424) Acc@1 74.512 (73.396) Acc@5 92.480 (92.160) [2021-04-15 23:41:58 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.085 (0.241) Loss 1.2362 (1.1453) Acc@1 71.680 (73.293) Acc@5 91.309 (92.188) [2021-04-15 23:42:00 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.093 (0.216) Loss 1.1135 (1.1449) Acc@1 76.465 (73.516) Acc@5 92.285 (92.114) [2021-04-15 23:42:04 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 73.532 Acc@5 92.042 [2021-04-15 23:42:04 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 73.5% [2021-04-15 23:42:04 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 73.53% [2021-04-15 23:42:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][0/1251] eta 1:20:10 lr 0.000770 time 3.8454 (3.8454) loss 3.2090 (3.2090) grad_norm 1.5723 (1.5723) [2021-04-15 23:42:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][10/1251] eta 0:12:30 lr 0.000770 time 0.2690 (0.6044) loss 3.8475 (3.5073) grad_norm 1.1277 (1.3920) [2021-04-15 23:42:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][20/1251] eta 0:09:13 lr 0.000770 time 0.3050 (0.4500) loss 3.9435 (3.6439) grad_norm 1.2171 (1.3292) [2021-04-15 23:42:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][30/1251] eta 0:08:07 lr 0.000770 time 0.2717 (0.3989) loss 4.3393 (3.5981) grad_norm 1.3210 (1.2824) [2021-04-15 23:42:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][40/1251] eta 0:07:32 lr 0.000770 time 0.2785 (0.3737) loss 3.9318 (3.5625) grad_norm 1.2360 (1.2594) [2021-04-15 23:42:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][50/1251] eta 0:07:09 lr 0.000770 time 0.2795 (0.3572) loss 4.1353 (3.6197) grad_norm 1.3979 (1.2840) [2021-04-15 23:42:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][60/1251] eta 0:06:48 lr 0.000770 time 0.2483 (0.3432) loss 3.0949 (3.6279) grad_norm 1.1371 (1.2805) [2021-04-15 23:42:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][70/1251] eta 0:06:35 lr 0.000770 time 0.2746 (0.3346) loss 3.8881 (3.6658) grad_norm 1.0614 (1.2678) [2021-04-15 23:42:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][80/1251] eta 0:06:23 lr 0.000770 time 0.2869 (0.3277) loss 3.9601 (3.7078) grad_norm 1.0575 (1.2634) [2021-04-15 23:42:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][90/1251] eta 0:06:14 lr 0.000770 time 0.2978 (0.3222) loss 4.3649 (3.7161) grad_norm 1.2979 (1.2756) [2021-04-15 23:42:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][100/1251] eta 0:06:05 lr 0.000770 time 0.2428 (0.3174) loss 3.8071 (3.7097) grad_norm 1.2085 (1.2735) [2021-04-15 23:42:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][110/1251] eta 0:05:58 lr 0.000770 time 0.2868 (0.3145) loss 2.3860 (3.6558) grad_norm 1.1419 (1.2682) [2021-04-15 23:42:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][120/1251] eta 0:05:51 lr 0.000770 time 0.2722 (0.3112) loss 3.0738 (3.6704) grad_norm 1.1588 (1.2593) [2021-04-15 23:42:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][130/1251] eta 0:05:45 lr 0.000770 time 0.2791 (0.3086) loss 4.0233 (3.6796) grad_norm 1.4517 (1.2621) [2021-04-15 23:42:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][140/1251] eta 0:05:41 lr 0.000770 time 0.4563 (0.3077) loss 4.0571 (3.6701) grad_norm 1.0566 (1.2603) [2021-04-15 23:42:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][150/1251] eta 0:05:36 lr 0.000770 time 0.2881 (0.3053) loss 3.9708 (3.6930) grad_norm 1.1690 (1.2623) [2021-04-15 23:42:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][160/1251] eta 0:05:30 lr 0.000770 time 0.2959 (0.3033) loss 3.9308 (3.7005) grad_norm 1.4708 (1.2656) [2021-04-15 23:42:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][170/1251] eta 0:05:26 lr 0.000770 time 0.2906 (0.3019) loss 4.2141 (3.7153) grad_norm 1.0966 (1.2608) [2021-04-15 23:42:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][180/1251] eta 0:05:21 lr 0.000770 time 0.2975 (0.3005) loss 4.1925 (3.7126) grad_norm 1.1255 (1.2599) [2021-04-15 23:43:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][190/1251] eta 0:05:17 lr 0.000770 time 0.2560 (0.2991) loss 4.2624 (3.7148) grad_norm 1.2538 (1.2587) [2021-04-15 23:43:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][200/1251] eta 0:05:13 lr 0.000770 time 0.2994 (0.2982) loss 3.1830 (3.7100) grad_norm 1.1989 (1.2585) [2021-04-15 23:43:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][210/1251] eta 0:05:09 lr 0.000769 time 0.2784 (0.2971) loss 4.1766 (3.7123) grad_norm 1.2488 (1.2574) [2021-04-15 23:43:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][220/1251] eta 0:05:05 lr 0.000769 time 0.2935 (0.2962) loss 4.5426 (3.7082) grad_norm 1.2626 (1.2575) [2021-04-15 23:43:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][230/1251] eta 0:05:01 lr 0.000769 time 0.2632 (0.2956) loss 3.9944 (3.6983) grad_norm 1.3333 (1.2589) [2021-04-15 23:43:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][240/1251] eta 0:04:58 lr 0.000769 time 0.2625 (0.2951) loss 3.4917 (3.7042) grad_norm 1.2194 (1.2598) [2021-04-15 23:43:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][250/1251] eta 0:04:55 lr 0.000769 time 0.2717 (0.2947) loss 3.8873 (3.7038) grad_norm 1.2512 (1.2616) [2021-04-15 23:43:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][260/1251] eta 0:04:51 lr 0.000769 time 0.2621 (0.2947) loss 3.7923 (3.7062) grad_norm 1.2336 (1.2606) [2021-04-15 23:43:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][270/1251] eta 0:04:48 lr 0.000769 time 0.2551 (0.2938) loss 4.0592 (3.7129) grad_norm 1.0827 (1.2567) [2021-04-15 23:43:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][280/1251] eta 0:04:44 lr 0.000769 time 0.2755 (0.2932) loss 3.9800 (3.7048) grad_norm 1.3799 (1.2614) [2021-04-15 23:43:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][290/1251] eta 0:04:41 lr 0.000769 time 0.2711 (0.2928) loss 3.9308 (3.7061) grad_norm 1.1654 (1.2610) [2021-04-15 23:43:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][300/1251] eta 0:04:37 lr 0.000769 time 0.2661 (0.2922) loss 3.7785 (3.7089) grad_norm 1.1640 (1.2616) [2021-04-15 23:43:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][310/1251] eta 0:04:34 lr 0.000769 time 0.3042 (0.2922) loss 3.5918 (3.7138) grad_norm 1.1682 (1.2628) [2021-04-15 23:43:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][320/1251] eta 0:04:32 lr 0.000769 time 0.2838 (0.2924) loss 4.1107 (3.7111) grad_norm 1.5885 (1.2628) [2021-04-15 23:43:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][330/1251] eta 0:04:28 lr 0.000769 time 0.3044 (0.2919) loss 3.7609 (3.7141) grad_norm 1.2614 (1.2636) [2021-04-15 23:43:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][340/1251] eta 0:04:25 lr 0.000769 time 0.2722 (0.2919) loss 3.2844 (3.7090) grad_norm 1.1544 (1.2636) [2021-04-15 23:43:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][350/1251] eta 0:04:22 lr 0.000769 time 0.2794 (0.2918) loss 3.2398 (3.7066) grad_norm 1.1593 (1.2614) [2021-04-15 23:43:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][360/1251] eta 0:04:19 lr 0.000769 time 0.2727 (0.2918) loss 3.7502 (3.7119) grad_norm 1.1289 (1.2619) [2021-04-15 23:43:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][370/1251] eta 0:04:16 lr 0.000769 time 0.2783 (0.2916) loss 3.8851 (3.7076) grad_norm 1.1069 (1.2603) [2021-04-15 23:43:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][380/1251] eta 0:04:13 lr 0.000769 time 0.2767 (0.2914) loss 2.5932 (3.7048) grad_norm 1.3120 (1.2593) [2021-04-15 23:43:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][390/1251] eta 0:04:10 lr 0.000769 time 0.2978 (0.2912) loss 3.1069 (3.7074) grad_norm 1.2939 (1.2598) [2021-04-15 23:44:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][400/1251] eta 0:04:07 lr 0.000769 time 0.2941 (0.2909) loss 4.5581 (3.7170) grad_norm 1.3023 (1.2601) [2021-04-15 23:44:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][410/1251] eta 0:04:04 lr 0.000769 time 0.2809 (0.2907) loss 4.1062 (3.7110) grad_norm 1.1899 (1.2590) [2021-04-15 23:44:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][420/1251] eta 0:04:01 lr 0.000769 time 0.2774 (0.2905) loss 4.4653 (3.7216) grad_norm 1.3918 (1.2583) [2021-04-15 23:44:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][430/1251] eta 0:03:58 lr 0.000769 time 0.2599 (0.2902) loss 3.6492 (3.7187) grad_norm 1.3005 (1.2590) [2021-04-15 23:44:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][440/1251] eta 0:03:55 lr 0.000769 time 0.2652 (0.2899) loss 3.6673 (3.7196) grad_norm 1.1056 (1.2594) [2021-04-15 23:44:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][450/1251] eta 0:03:52 lr 0.000769 time 0.2536 (0.2897) loss 2.9519 (3.7207) grad_norm 1.6276 (1.2595) [2021-04-15 23:44:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][460/1251] eta 0:03:48 lr 0.000769 time 0.2735 (0.2894) loss 4.0321 (3.7253) grad_norm 1.2991 (1.2637) [2021-04-15 23:44:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][470/1251] eta 0:03:45 lr 0.000769 time 0.2855 (0.2891) loss 3.2090 (3.7280) grad_norm 1.4007 (1.2666) [2021-04-15 23:44:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][480/1251] eta 0:03:42 lr 0.000769 time 0.2876 (0.2889) loss 4.2858 (3.7238) grad_norm 1.1932 (1.2660) [2021-04-15 23:44:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][490/1251] eta 0:03:39 lr 0.000769 time 0.2546 (0.2886) loss 4.0048 (3.7251) grad_norm 1.0970 (1.2639) [2021-04-15 23:44:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][500/1251] eta 0:03:36 lr 0.000768 time 0.3065 (0.2886) loss 4.1383 (3.7325) grad_norm 1.1114 (1.2621) [2021-04-15 23:44:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][510/1251] eta 0:03:33 lr 0.000768 time 0.2811 (0.2883) loss 3.9476 (3.7249) grad_norm 1.3511 (1.2616) [2021-04-15 23:44:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][520/1251] eta 0:03:30 lr 0.000768 time 0.2905 (0.2881) loss 2.9099 (3.7212) grad_norm 1.2993 (1.2607) [2021-04-15 23:44:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][530/1251] eta 0:03:27 lr 0.000768 time 0.2812 (0.2880) loss 4.0880 (3.7196) grad_norm 1.1876 (1.2587) [2021-04-15 23:44:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][540/1251] eta 0:03:24 lr 0.000768 time 0.2792 (0.2878) loss 4.2731 (3.7202) grad_norm 1.4246 (1.2602) [2021-04-15 23:44:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][550/1251] eta 0:03:21 lr 0.000768 time 0.2759 (0.2879) loss 3.8921 (3.7218) grad_norm 1.1302 (1.2591) [2021-04-15 23:44:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][560/1251] eta 0:03:18 lr 0.000768 time 0.2705 (0.2877) loss 2.4921 (3.7210) grad_norm 1.3545 (1.2573) [2021-04-15 23:44:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][570/1251] eta 0:03:16 lr 0.000768 time 0.2717 (0.2879) loss 3.6959 (3.7239) grad_norm 1.1180 (1.2596) [2021-04-15 23:44:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][580/1251] eta 0:03:13 lr 0.000768 time 0.2878 (0.2883) loss 3.8808 (3.7279) grad_norm 1.2795 (1.2599) [2021-04-15 23:44:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][590/1251] eta 0:03:10 lr 0.000768 time 0.2443 (0.2881) loss 4.0272 (3.7265) grad_norm 1.4599 (1.2584) [2021-04-15 23:44:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][600/1251] eta 0:03:07 lr 0.000768 time 0.4372 (0.2881) loss 3.9796 (3.7305) grad_norm 1.2073 (1.2583) [2021-04-15 23:45:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][610/1251] eta 0:03:04 lr 0.000768 time 0.2530 (0.2879) loss 3.6147 (3.7317) grad_norm 1.1160 (1.2571) [2021-04-15 23:45:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][620/1251] eta 0:03:01 lr 0.000768 time 0.2794 (0.2878) loss 2.5522 (3.7326) grad_norm 1.2921 (1.2570) [2021-04-15 23:45:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][630/1251] eta 0:02:58 lr 0.000768 time 0.2498 (0.2876) loss 4.2369 (3.7376) grad_norm 1.3000 (1.2579) [2021-04-15 23:45:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][640/1251] eta 0:02:55 lr 0.000768 time 0.2977 (0.2875) loss 3.9595 (3.7409) grad_norm 1.3259 (1.2579) [2021-04-15 23:45:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][650/1251] eta 0:02:52 lr 0.000768 time 0.2731 (0.2873) loss 3.3827 (3.7407) grad_norm 1.1708 (1.2570) [2021-04-15 23:45:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][660/1251] eta 0:02:49 lr 0.000768 time 0.2891 (0.2871) loss 4.1150 (3.7429) grad_norm 1.1903 (1.2565) [2021-04-15 23:45:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][670/1251] eta 0:02:46 lr 0.000768 time 0.2734 (0.2870) loss 4.1546 (3.7432) grad_norm 1.0721 (1.2560) [2021-04-15 23:45:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][680/1251] eta 0:02:43 lr 0.000768 time 0.2869 (0.2869) loss 4.0676 (3.7462) grad_norm 1.5439 (1.2562) [2021-04-15 23:45:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][690/1251] eta 0:02:40 lr 0.000768 time 0.2542 (0.2867) loss 4.2197 (3.7431) grad_norm 1.1066 (1.2559) [2021-04-15 23:45:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][700/1251] eta 0:02:37 lr 0.000768 time 0.3194 (0.2866) loss 4.3924 (3.7432) grad_norm 1.2398 (1.2570) [2021-04-15 23:45:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][710/1251] eta 0:02:34 lr 0.000768 time 0.2707 (0.2864) loss 4.3080 (3.7418) grad_norm 1.2927 (1.2583) [2021-04-15 23:45:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][720/1251] eta 0:02:32 lr 0.000768 time 0.2603 (0.2865) loss 3.7133 (3.7411) grad_norm 1.5062 (1.2594) [2021-04-15 23:45:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][730/1251] eta 0:02:29 lr 0.000768 time 0.2799 (0.2864) loss 4.1881 (3.7421) grad_norm 1.2978 (1.2597) [2021-04-15 23:45:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][740/1251] eta 0:02:26 lr 0.000768 time 0.2724 (0.2863) loss 3.6684 (3.7392) grad_norm 1.2067 (1.2597) [2021-04-15 23:45:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][750/1251] eta 0:02:23 lr 0.000768 time 0.2851 (0.2862) loss 3.6246 (3.7380) grad_norm 1.4129 (1.2605) [2021-04-15 23:45:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][760/1251] eta 0:02:20 lr 0.000768 time 0.2885 (0.2862) loss 4.0321 (3.7389) grad_norm 1.2546 (1.2601) [2021-04-15 23:45:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][770/1251] eta 0:02:17 lr 0.000768 time 0.2751 (0.2862) loss 4.1524 (3.7340) grad_norm 1.2679 (1.2596) [2021-04-15 23:45:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][780/1251] eta 0:02:14 lr 0.000767 time 0.2632 (0.2861) loss 3.3268 (3.7298) grad_norm 1.1261 (1.2580) [2021-04-15 23:45:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][790/1251] eta 0:02:11 lr 0.000767 time 0.2705 (0.2860) loss 4.0517 (3.7314) grad_norm 1.2213 (1.2576) [2021-04-15 23:45:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][800/1251] eta 0:02:08 lr 0.000767 time 0.3063 (0.2860) loss 4.5158 (3.7326) grad_norm 1.3593 (1.2571) [2021-04-15 23:45:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][810/1251] eta 0:02:06 lr 0.000767 time 0.2953 (0.2859) loss 4.2547 (3.7351) grad_norm 1.2819 (1.2581) [2021-04-15 23:45:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][820/1251] eta 0:02:03 lr 0.000767 time 0.2760 (0.2858) loss 3.6811 (3.7370) grad_norm 1.1853 (1.2574) [2021-04-15 23:46:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][830/1251] eta 0:02:00 lr 0.000767 time 0.3045 (0.2857) loss 3.7660 (3.7365) grad_norm 1.7263 (1.2572) [2021-04-15 23:46:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][840/1251] eta 0:01:57 lr 0.000767 time 0.2487 (0.2856) loss 3.9810 (3.7377) grad_norm 1.1430 (1.2567) [2021-04-15 23:46:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][850/1251] eta 0:01:54 lr 0.000767 time 0.2648 (0.2856) loss 3.8478 (3.7389) grad_norm 1.2617 (1.2572) [2021-04-15 23:46:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][860/1251] eta 0:01:51 lr 0.000767 time 0.2555 (0.2856) loss 4.7403 (3.7372) grad_norm 1.2790 (1.2572) [2021-04-15 23:46:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][870/1251] eta 0:01:48 lr 0.000767 time 0.2757 (0.2855) loss 3.0453 (3.7374) grad_norm 1.3137 (1.2576) [2021-04-15 23:46:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][880/1251] eta 0:01:45 lr 0.000767 time 0.3994 (0.2855) loss 3.7631 (3.7342) grad_norm 1.3842 (1.2575) [2021-04-15 23:46:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][890/1251] eta 0:01:43 lr 0.000767 time 0.2583 (0.2855) loss 3.9092 (3.7352) grad_norm 1.2756 (1.2572) [2021-04-15 23:46:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][900/1251] eta 0:01:40 lr 0.000767 time 0.2767 (0.2855) loss 4.1250 (3.7372) grad_norm 1.2117 (1.2582) [2021-04-15 23:46:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][910/1251] eta 0:01:37 lr 0.000767 time 0.2789 (0.2854) loss 3.6391 (3.7336) grad_norm 1.5368 (1.2579) [2021-04-15 23:46:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][920/1251] eta 0:01:34 lr 0.000767 time 0.2872 (0.2855) loss 3.4916 (3.7330) grad_norm 1.0879 (1.2573) [2021-04-15 23:46:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][930/1251] eta 0:01:31 lr 0.000767 time 0.2839 (0.2856) loss 3.5328 (3.7332) grad_norm 1.1744 (1.2565) [2021-04-15 23:46:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][940/1251] eta 0:01:28 lr 0.000767 time 0.2622 (0.2856) loss 3.7487 (3.7319) grad_norm 1.1938 (1.2570) [2021-04-15 23:46:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][950/1251] eta 0:01:25 lr 0.000767 time 0.2625 (0.2855) loss 4.1952 (3.7323) grad_norm 1.2928 (1.2573) [2021-04-15 23:46:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][960/1251] eta 0:01:23 lr 0.000767 time 0.2661 (0.2854) loss 3.8234 (3.7328) grad_norm 1.4626 (1.2580) [2021-04-15 23:46:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][970/1251] eta 0:01:20 lr 0.000767 time 0.2683 (0.2854) loss 2.5832 (3.7337) grad_norm 1.3487 (1.2589) [2021-04-15 23:46:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][980/1251] eta 0:01:17 lr 0.000767 time 0.2751 (0.2853) loss 3.5557 (3.7336) grad_norm 1.1081 (1.2588) [2021-04-15 23:46:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][990/1251] eta 0:01:14 lr 0.000767 time 0.2796 (0.2853) loss 3.9387 (3.7326) grad_norm 1.3712 (1.2588) [2021-04-15 23:46:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][1000/1251] eta 0:01:11 lr 0.000767 time 0.2832 (0.2851) loss 4.1304 (3.7323) grad_norm 1.0640 (1.2584) [2021-04-15 23:46:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][1010/1251] eta 0:01:08 lr 0.000767 time 0.2869 (0.2851) loss 3.4505 (3.7325) grad_norm 1.2803 (1.2585) [2021-04-15 23:46:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][1020/1251] eta 0:01:05 lr 0.000767 time 0.2774 (0.2850) loss 2.8603 (3.7337) grad_norm 1.3951 (1.2585) [2021-04-15 23:46:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][1030/1251] eta 0:01:02 lr 0.000767 time 0.3157 (0.2849) loss 3.9670 (3.7355) grad_norm 1.4335 (1.2599) [2021-04-15 23:47:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][1040/1251] eta 0:01:00 lr 0.000767 time 0.2760 (0.2848) loss 3.3984 (3.7346) grad_norm 1.5874 (1.2605) [2021-04-15 23:47:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][1050/1251] eta 0:00:57 lr 0.000767 time 0.2720 (0.2847) loss 4.5646 (3.7362) grad_norm 1.2071 (1.2598) [2021-04-15 23:47:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][1060/1251] eta 0:00:54 lr 0.000767 time 0.2876 (0.2847) loss 3.8934 (3.7371) grad_norm 1.1904 (1.2591) [2021-04-15 23:47:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][1070/1251] eta 0:00:51 lr 0.000766 time 0.2510 (0.2847) loss 2.3368 (3.7310) grad_norm 1.5700 (1.2595) [2021-04-15 23:47:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][1080/1251] eta 0:00:48 lr 0.000766 time 0.2987 (0.2846) loss 4.0732 (3.7314) grad_norm 1.0637 (1.2590) [2021-04-15 23:47:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][1090/1251] eta 0:00:45 lr 0.000766 time 0.3007 (0.2846) loss 4.2158 (3.7317) grad_norm 1.4802 (1.2592) [2021-04-15 23:47:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][1100/1251] eta 0:00:42 lr 0.000766 time 0.2831 (0.2845) loss 3.5700 (3.7309) grad_norm 1.3312 (1.2593) [2021-04-15 23:47:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][1110/1251] eta 0:00:40 lr 0.000766 time 0.2700 (0.2844) loss 4.5617 (3.7285) grad_norm 1.2903 (1.2604) [2021-04-15 23:47:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][1120/1251] eta 0:00:37 lr 0.000766 time 0.3030 (0.2844) loss 2.8365 (3.7266) grad_norm 1.2910 (1.2605) [2021-04-15 23:47:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][1130/1251] eta 0:00:34 lr 0.000766 time 0.2814 (0.2844) loss 3.8910 (3.7272) grad_norm 1.3271 (1.2608) [2021-04-15 23:47:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][1140/1251] eta 0:00:31 lr 0.000766 time 0.2811 (0.2845) loss 3.9616 (3.7248) grad_norm 0.9804 (1.2611) [2021-04-15 23:47:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][1150/1251] eta 0:00:28 lr 0.000766 time 0.2773 (0.2846) loss 4.7479 (3.7254) grad_norm 1.3857 (1.2616) [2021-04-15 23:47:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][1160/1251] eta 0:00:25 lr 0.000766 time 0.2788 (0.2847) loss 2.7469 (3.7241) grad_norm 1.1151 (1.2614) [2021-04-15 23:47:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][1170/1251] eta 0:00:23 lr 0.000766 time 0.2656 (0.2846) loss 4.3820 (3.7246) grad_norm 1.2613 (1.2610) [2021-04-15 23:47:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][1180/1251] eta 0:00:20 lr 0.000766 time 0.2616 (0.2847) loss 3.5198 (3.7242) grad_norm 1.4278 (1.2611) [2021-04-15 23:47:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][1190/1251] eta 0:00:17 lr 0.000766 time 0.3793 (0.2846) loss 4.1464 (3.7245) grad_norm 1.0803 (1.2610) [2021-04-15 23:47:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][1200/1251] eta 0:00:14 lr 0.000766 time 0.2941 (0.2846) loss 3.0458 (3.7253) grad_norm 1.2991 (1.2610) [2021-04-15 23:47:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][1210/1251] eta 0:00:11 lr 0.000766 time 0.2750 (0.2845) loss 2.8239 (3.7243) grad_norm 1.1840 (1.2614) [2021-04-15 23:47:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][1220/1251] eta 0:00:08 lr 0.000766 time 0.3033 (0.2845) loss 3.6152 (3.7250) grad_norm 1.1857 (1.2614) [2021-04-15 23:47:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][1230/1251] eta 0:00:05 lr 0.000766 time 0.2920 (0.2844) loss 4.5064 (3.7263) grad_norm 1.3532 (1.2619) [2021-04-15 23:47:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][1240/1251] eta 0:00:03 lr 0.000766 time 0.2491 (0.2843) loss 3.6414 (3.7258) grad_norm 1.0887 (1.2619) [2021-04-15 23:47:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [96/300][1250/1251] eta 0:00:00 lr 0.000766 time 0.2484 (0.2840) loss 4.0606 (3.7275) grad_norm 1.3534 (1.2618) [2021-04-15 23:48:02 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 96 training takes 0:05:58 [2021-04-15 23:48:02 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_96.pth saving...... [2021-04-15 23:48:12 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_96.pth saved !!! [2021-04-15 23:48:13 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.213 (1.213) Loss 1.0745 (1.0745) Acc@1 75.195 (75.195) Acc@5 93.262 (93.262) [2021-04-15 23:48:14 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.172 (0.238) Loss 1.1855 (1.1466) Acc@1 73.926 (73.846) Acc@5 91.602 (91.797) [2021-04-15 23:48:17 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 1.062 (0.275) Loss 1.0232 (1.1439) Acc@1 75.977 (73.554) Acc@5 93.359 (91.885) [2021-04-15 23:48:19 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.121 (0.240) Loss 1.1406 (1.1366) Acc@1 73.242 (73.507) Acc@5 92.480 (92.027) [2021-04-15 23:48:20 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.216) Loss 1.1458 (1.1386) Acc@1 74.023 (73.495) Acc@5 90.430 (91.873) [2021-04-15 23:48:24 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 73.548 Acc@5 91.894 [2021-04-15 23:48:24 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 73.5% [2021-04-15 23:48:24 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 73.55% [2021-04-15 23:48:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][0/1251] eta 1:20:56 lr 0.000766 time 3.8821 (3.8821) loss 2.3541 (2.3541) grad_norm 1.1050 (1.1050) [2021-04-15 23:48:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][10/1251] eta 0:12:28 lr 0.000766 time 0.2662 (0.6032) loss 3.7803 (3.7306) grad_norm 1.0672 (1.1828) [2021-04-15 23:48:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][20/1251] eta 0:09:12 lr 0.000766 time 0.2774 (0.4491) loss 3.8299 (3.6452) grad_norm 1.2285 (1.1781) [2021-04-15 23:48:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][30/1251] eta 0:08:00 lr 0.000766 time 0.2729 (0.3932) loss 3.6082 (3.6484) grad_norm 1.3093 (1.1792) [2021-04-15 23:48:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][40/1251] eta 0:07:22 lr 0.000766 time 0.2930 (0.3656) loss 4.1882 (3.7221) grad_norm 1.1465 (1.2031) [2021-04-15 23:48:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][50/1251] eta 0:06:57 lr 0.000766 time 0.2456 (0.3473) loss 3.8519 (3.7399) grad_norm 1.1198 (1.2317) [2021-04-15 23:48:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][60/1251] eta 0:06:39 lr 0.000766 time 0.3005 (0.3356) loss 4.2522 (3.7256) grad_norm 1.1513 (1.2409) [2021-04-15 23:48:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][70/1251] eta 0:06:28 lr 0.000766 time 0.2778 (0.3289) loss 3.5986 (3.7310) grad_norm 1.2560 (1.2381) [2021-04-15 23:48:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][80/1251] eta 0:06:17 lr 0.000766 time 0.2766 (0.3222) loss 2.9465 (3.6914) grad_norm 1.3402 (1.2357) [2021-04-15 23:48:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][90/1251] eta 0:06:10 lr 0.000766 time 0.2894 (0.3188) loss 4.3560 (3.6952) grad_norm 1.5224 (1.2411) [2021-04-15 23:48:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][100/1251] eta 0:06:02 lr 0.000765 time 0.2715 (0.3154) loss 4.1592 (3.7045) grad_norm 1.1668 (1.2407) [2021-04-15 23:48:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][110/1251] eta 0:05:56 lr 0.000765 time 0.2909 (0.3124) loss 3.7495 (3.7063) grad_norm 1.2122 (1.2419) [2021-04-15 23:49:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][120/1251] eta 0:05:51 lr 0.000765 time 0.2679 (0.3107) loss 4.2472 (3.7073) grad_norm 1.4416 (1.2427) [2021-04-15 23:49:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][130/1251] eta 0:05:45 lr 0.000765 time 0.2736 (0.3086) loss 4.2498 (3.7093) grad_norm 1.2748 (1.2402) [2021-04-15 23:49:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][140/1251] eta 0:05:40 lr 0.000765 time 0.2891 (0.3068) loss 3.2857 (3.7207) grad_norm 1.1467 (1.2462) [2021-04-15 23:49:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][150/1251] eta 0:05:37 lr 0.000765 time 0.2722 (0.3062) loss 3.8267 (3.7237) grad_norm 1.1845 (1.2473) [2021-04-15 23:49:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][160/1251] eta 0:05:31 lr 0.000765 time 0.2738 (0.3041) loss 4.1997 (3.7332) grad_norm 1.2506 (1.2526) [2021-04-15 23:49:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][170/1251] eta 0:05:26 lr 0.000765 time 0.2833 (0.3023) loss 4.0390 (3.7197) grad_norm 1.1687 (1.2506) [2021-04-15 23:49:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][180/1251] eta 0:05:22 lr 0.000765 time 0.2787 (0.3010) loss 3.3257 (3.7298) grad_norm 1.2232 (1.2558) [2021-04-15 23:49:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][190/1251] eta 0:05:18 lr 0.000765 time 0.2782 (0.3002) loss 3.7503 (3.7446) grad_norm 1.3522 (1.2576) [2021-04-15 23:49:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][200/1251] eta 0:05:14 lr 0.000765 time 0.2794 (0.2990) loss 2.4559 (3.7417) grad_norm 1.1367 (1.2534) [2021-04-15 23:49:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][210/1251] eta 0:05:10 lr 0.000765 time 0.2820 (0.2983) loss 3.2838 (3.7344) grad_norm 1.3403 (1.2513) [2021-04-15 23:49:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][220/1251] eta 0:05:06 lr 0.000765 time 0.2881 (0.2974) loss 3.4788 (3.7237) grad_norm 1.1293 (1.2536) [2021-04-15 23:49:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][230/1251] eta 0:05:02 lr 0.000765 time 0.2699 (0.2966) loss 4.5141 (3.7275) grad_norm 1.3428 (1.2541) [2021-04-15 23:49:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][240/1251] eta 0:04:58 lr 0.000765 time 0.2737 (0.2957) loss 3.0301 (3.7303) grad_norm 1.3448 (1.2530) [2021-04-15 23:49:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][250/1251] eta 0:04:55 lr 0.000765 time 0.2757 (0.2950) loss 2.7999 (3.7243) grad_norm 1.2948 (1.2539) [2021-04-15 23:49:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][260/1251] eta 0:04:51 lr 0.000765 time 0.2453 (0.2941) loss 4.4085 (3.7366) grad_norm 1.1443 (1.2533) [2021-04-15 23:49:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][270/1251] eta 0:04:47 lr 0.000765 time 0.2566 (0.2934) loss 4.0870 (3.7405) grad_norm 1.5051 (1.2570) [2021-04-15 23:49:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][280/1251] eta 0:04:44 lr 0.000765 time 0.2951 (0.2929) loss 3.5166 (3.7414) grad_norm 1.1841 (1.2552) [2021-04-15 23:49:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][290/1251] eta 0:04:40 lr 0.000765 time 0.3115 (0.2923) loss 4.0386 (3.7507) grad_norm 1.1965 (1.2541) [2021-04-15 23:49:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][300/1251] eta 0:04:37 lr 0.000765 time 0.2688 (0.2918) loss 3.9850 (3.7410) grad_norm 1.2199 (1.2546) [2021-04-15 23:49:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][310/1251] eta 0:04:34 lr 0.000765 time 0.2863 (0.2914) loss 2.8866 (3.7421) grad_norm 1.1648 (1.2566) [2021-04-15 23:49:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][320/1251] eta 0:04:30 lr 0.000765 time 0.2705 (0.2908) loss 2.8442 (3.7383) grad_norm 1.1071 (1.2549) [2021-04-15 23:50:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][330/1251] eta 0:04:27 lr 0.000765 time 0.2912 (0.2909) loss 2.8315 (3.7388) grad_norm 1.0755 (1.2526) [2021-04-15 23:50:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][340/1251] eta 0:04:24 lr 0.000765 time 0.2639 (0.2908) loss 4.5487 (3.7419) grad_norm 1.2449 (1.2530) [2021-04-15 23:50:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][350/1251] eta 0:04:22 lr 0.000765 time 0.2685 (0.2908) loss 4.2826 (3.7390) grad_norm 1.3966 (1.2510) [2021-04-15 23:50:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][360/1251] eta 0:04:19 lr 0.000765 time 0.2451 (0.2914) loss 3.1623 (3.7424) grad_norm 1.0828 (1.2483) [2021-04-15 23:50:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][370/1251] eta 0:04:16 lr 0.000765 time 0.3082 (0.2910) loss 3.6019 (3.7402) grad_norm 1.2219 (1.2501) [2021-04-15 23:50:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][380/1251] eta 0:04:13 lr 0.000765 time 0.3161 (0.2910) loss 4.2122 (3.7374) grad_norm 1.3835 (1.2513) [2021-04-15 23:50:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][390/1251] eta 0:04:10 lr 0.000764 time 0.2638 (0.2909) loss 4.2271 (3.7370) grad_norm 1.5084 (1.2544) [2021-04-15 23:50:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][400/1251] eta 0:04:07 lr 0.000764 time 0.2665 (0.2905) loss 2.5176 (3.7345) grad_norm 1.4992 (1.2550) [2021-04-15 23:50:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][410/1251] eta 0:04:04 lr 0.000764 time 0.2897 (0.2903) loss 3.8845 (3.7367) grad_norm 1.0614 (1.2535) [2021-04-15 23:50:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][420/1251] eta 0:04:00 lr 0.000764 time 0.2676 (0.2899) loss 4.1577 (3.7351) grad_norm 1.2034 (1.2525) [2021-04-15 23:50:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][430/1251] eta 0:03:57 lr 0.000764 time 0.2675 (0.2895) loss 2.5992 (3.7274) grad_norm 1.3463 (1.2522) [2021-04-15 23:50:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][440/1251] eta 0:03:54 lr 0.000764 time 0.2952 (0.2892) loss 4.0518 (3.7268) grad_norm 1.4924 (1.2521) [2021-04-15 23:50:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][450/1251] eta 0:03:51 lr 0.000764 time 0.2853 (0.2891) loss 3.9205 (3.7283) grad_norm 1.1290 (1.2495) [2021-04-15 23:50:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][460/1251] eta 0:03:48 lr 0.000764 time 0.2743 (0.2887) loss 3.8863 (3.7279) grad_norm 1.3355 (1.2486) [2021-04-15 23:50:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][470/1251] eta 0:03:45 lr 0.000764 time 0.2722 (0.2887) loss 3.2899 (3.7202) grad_norm 1.2223 (1.2482) [2021-04-15 23:50:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][480/1251] eta 0:03:42 lr 0.000764 time 0.2918 (0.2885) loss 4.3271 (3.7165) grad_norm 1.1952 (1.2479) [2021-04-15 23:50:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][490/1251] eta 0:03:39 lr 0.000764 time 0.2899 (0.2882) loss 3.0880 (3.7188) grad_norm 1.1565 (1.2465) [2021-04-15 23:50:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][500/1251] eta 0:03:36 lr 0.000764 time 0.2766 (0.2879) loss 4.0100 (3.7174) grad_norm 1.1483 (1.2464) [2021-04-15 23:50:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][510/1251] eta 0:03:33 lr 0.000764 time 0.2666 (0.2877) loss 4.3816 (3.7147) grad_norm 1.2358 (1.2473) [2021-04-15 23:50:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][520/1251] eta 0:03:30 lr 0.000764 time 0.2723 (0.2874) loss 4.7431 (3.7132) grad_norm 1.2505 (1.2468) [2021-04-15 23:50:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][530/1251] eta 0:03:27 lr 0.000764 time 0.2518 (0.2872) loss 3.9818 (3.7171) grad_norm 1.1796 (1.2460) [2021-04-15 23:50:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][540/1251] eta 0:03:23 lr 0.000764 time 0.2902 (0.2869) loss 4.1683 (3.7177) grad_norm 1.2381 (1.2453) [2021-04-15 23:51:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][550/1251] eta 0:03:21 lr 0.000764 time 0.2788 (0.2869) loss 3.0932 (3.7186) grad_norm 1.3141 (1.2452) [2021-04-15 23:51:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][560/1251] eta 0:03:18 lr 0.000764 time 0.2476 (0.2867) loss 3.5954 (3.7147) grad_norm 1.1349 (1.2448) [2021-04-15 23:51:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][570/1251] eta 0:03:15 lr 0.000764 time 0.2755 (0.2865) loss 4.2217 (3.7115) grad_norm 1.2515 (1.2444) [2021-04-15 23:51:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][580/1251] eta 0:03:12 lr 0.000764 time 0.2673 (0.2866) loss 4.1188 (3.7159) grad_norm 1.3241 (1.2444) [2021-04-15 23:51:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][590/1251] eta 0:03:09 lr 0.000764 time 0.2761 (0.2864) loss 2.9600 (3.7138) grad_norm 1.2851 (1.2447) [2021-04-15 23:51:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][600/1251] eta 0:03:06 lr 0.000764 time 0.2791 (0.2862) loss 4.4283 (3.7120) grad_norm 1.2860 (1.2436) [2021-04-15 23:51:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][610/1251] eta 0:03:03 lr 0.000764 time 0.2765 (0.2861) loss 4.2200 (3.7156) grad_norm 1.2012 (1.2435) [2021-04-15 23:51:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][620/1251] eta 0:03:00 lr 0.000764 time 0.2647 (0.2860) loss 3.1368 (3.7177) grad_norm 1.2442 (1.2438) [2021-04-15 23:51:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][630/1251] eta 0:02:57 lr 0.000764 time 0.2795 (0.2859) loss 3.9966 (3.7204) grad_norm 1.0761 (1.2431) [2021-04-15 23:51:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][640/1251] eta 0:02:54 lr 0.000764 time 0.2444 (0.2860) loss 3.8694 (3.7179) grad_norm 1.3084 (1.2445) [2021-04-15 23:51:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][650/1251] eta 0:02:51 lr 0.000764 time 0.2795 (0.2858) loss 4.6333 (3.7195) grad_norm 1.2996 (1.2451) [2021-04-15 23:51:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][660/1251] eta 0:02:48 lr 0.000764 time 0.2652 (0.2858) loss 3.1566 (3.7203) grad_norm 1.1653 (1.2453) [2021-04-15 23:51:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][670/1251] eta 0:02:46 lr 0.000763 time 0.2999 (0.2858) loss 2.8378 (3.7152) grad_norm 1.2690 (1.2449) [2021-04-15 23:51:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][680/1251] eta 0:02:43 lr 0.000763 time 0.2781 (0.2857) loss 2.5680 (3.7149) grad_norm 1.3821 (1.2443) [2021-04-15 23:51:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][690/1251] eta 0:02:40 lr 0.000763 time 0.2744 (0.2858) loss 4.6702 (3.7170) grad_norm 1.0194 (1.2448) [2021-04-15 23:51:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][700/1251] eta 0:02:37 lr 0.000763 time 0.3930 (0.2858) loss 4.5126 (3.7197) grad_norm 1.1152 (1.2447) [2021-04-15 23:51:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][710/1251] eta 0:02:34 lr 0.000763 time 0.3195 (0.2857) loss 3.0712 (3.7169) grad_norm 1.4714 (1.2437) [2021-04-15 23:51:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][720/1251] eta 0:02:31 lr 0.000763 time 0.3020 (0.2855) loss 3.9573 (3.7178) grad_norm 1.1394 (1.2433) [2021-04-15 23:51:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][730/1251] eta 0:02:28 lr 0.000763 time 0.2859 (0.2855) loss 4.2097 (3.7163) grad_norm 1.2070 (1.2423) [2021-04-15 23:51:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][740/1251] eta 0:02:25 lr 0.000763 time 0.2576 (0.2854) loss 4.0162 (3.7153) grad_norm 1.1203 (inf) [2021-04-15 23:51:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][750/1251] eta 0:02:22 lr 0.000763 time 0.2877 (0.2852) loss 4.0843 (3.7170) grad_norm 1.0364 (inf) [2021-04-15 23:52:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][760/1251] eta 0:02:20 lr 0.000763 time 0.2665 (0.2853) loss 3.4148 (3.7179) grad_norm 1.1203 (inf) [2021-04-15 23:52:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][770/1251] eta 0:02:17 lr 0.000763 time 0.2626 (0.2852) loss 2.9267 (3.7156) grad_norm 1.1793 (inf) [2021-04-15 23:52:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][780/1251] eta 0:02:14 lr 0.000763 time 0.2702 (0.2852) loss 3.9367 (3.7130) grad_norm 1.1568 (inf) [2021-04-15 23:52:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][790/1251] eta 0:02:11 lr 0.000763 time 0.3125 (0.2852) loss 4.1145 (3.7162) grad_norm 1.3257 (inf) [2021-04-15 23:52:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][800/1251] eta 0:02:08 lr 0.000763 time 0.2592 (0.2851) loss 3.1889 (3.7176) grad_norm 1.5629 (inf) [2021-04-15 23:52:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][810/1251] eta 0:02:05 lr 0.000763 time 0.2542 (0.2850) loss 4.2213 (3.7204) grad_norm 1.0252 (inf) [2021-04-15 23:52:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][820/1251] eta 0:02:02 lr 0.000763 time 0.2917 (0.2850) loss 4.7527 (3.7236) grad_norm 1.2182 (inf) [2021-04-15 23:52:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][830/1251] eta 0:01:59 lr 0.000763 time 0.2899 (0.2849) loss 3.2648 (3.7242) grad_norm 1.2578 (inf) [2021-04-15 23:52:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][840/1251] eta 0:01:57 lr 0.000763 time 0.2913 (0.2848) loss 3.2226 (3.7275) grad_norm 1.1881 (inf) [2021-04-15 23:52:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][850/1251] eta 0:01:54 lr 0.000763 time 0.2815 (0.2847) loss 4.7188 (3.7302) grad_norm 1.1816 (inf) [2021-04-15 23:52:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][860/1251] eta 0:01:51 lr 0.000763 time 0.2678 (0.2845) loss 4.4853 (3.7262) grad_norm 1.1797 (inf) [2021-04-15 23:52:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][870/1251] eta 0:01:48 lr 0.000763 time 0.2486 (0.2845) loss 3.1332 (3.7226) grad_norm 1.2786 (inf) [2021-04-15 23:52:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][880/1251] eta 0:01:45 lr 0.000763 time 0.2935 (0.2844) loss 3.8987 (3.7228) grad_norm 1.6613 (inf) [2021-04-15 23:52:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][890/1251] eta 0:01:42 lr 0.000763 time 0.2881 (0.2843) loss 3.8714 (3.7235) grad_norm 1.1539 (inf) [2021-04-15 23:52:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][900/1251] eta 0:01:39 lr 0.000763 time 0.2613 (0.2842) loss 4.1972 (3.7246) grad_norm 1.3690 (inf) [2021-04-15 23:52:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][910/1251] eta 0:01:36 lr 0.000763 time 0.2662 (0.2842) loss 4.0993 (3.7241) grad_norm 1.5477 (inf) [2021-04-15 23:52:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][920/1251] eta 0:01:34 lr 0.000763 time 0.2538 (0.2843) loss 3.9354 (3.7246) grad_norm 1.2727 (inf) [2021-04-15 23:52:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][930/1251] eta 0:01:31 lr 0.000763 time 0.2757 (0.2844) loss 4.3967 (3.7240) grad_norm 1.2361 (inf) [2021-04-15 23:52:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][940/1251] eta 0:01:28 lr 0.000763 time 0.2689 (0.2846) loss 3.3643 (3.7228) grad_norm 1.3361 (inf) [2021-04-15 23:52:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][950/1251] eta 0:01:25 lr 0.000762 time 0.2839 (0.2845) loss 3.8624 (3.7240) grad_norm 1.3189 (inf) [2021-04-15 23:52:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][960/1251] eta 0:01:22 lr 0.000762 time 0.2764 (0.2845) loss 3.3838 (3.7261) grad_norm 1.1238 (inf) [2021-04-15 23:53:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][970/1251] eta 0:01:19 lr 0.000762 time 0.2939 (0.2846) loss 4.1421 (3.7222) grad_norm 1.2386 (inf) [2021-04-15 23:53:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][980/1251] eta 0:01:17 lr 0.000762 time 0.2789 (0.2845) loss 3.6101 (3.7243) grad_norm 1.1777 (inf) [2021-04-15 23:53:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][990/1251] eta 0:01:14 lr 0.000762 time 0.2661 (0.2845) loss 4.0663 (3.7259) grad_norm 1.4587 (inf) [2021-04-15 23:53:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][1000/1251] eta 0:01:11 lr 0.000762 time 0.2768 (0.2844) loss 3.6091 (3.7254) grad_norm 1.1084 (inf) [2021-04-15 23:53:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][1010/1251] eta 0:01:08 lr 0.000762 time 0.2451 (0.2843) loss 2.4509 (3.7257) grad_norm 1.2994 (inf) [2021-04-15 23:53:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][1020/1251] eta 0:01:05 lr 0.000762 time 0.2779 (0.2843) loss 3.8153 (3.7243) grad_norm 1.3024 (inf) [2021-04-15 23:53:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][1030/1251] eta 0:01:02 lr 0.000762 time 0.2778 (0.2843) loss 3.4468 (3.7231) grad_norm 1.1209 (inf) [2021-04-15 23:53:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][1040/1251] eta 0:00:59 lr 0.000762 time 0.2806 (0.2841) loss 2.3029 (3.7224) grad_norm 1.2393 (inf) [2021-04-15 23:53:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][1050/1251] eta 0:00:57 lr 0.000762 time 0.2801 (0.2840) loss 4.8233 (3.7222) grad_norm 1.1244 (inf) [2021-04-15 23:53:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][1060/1251] eta 0:00:54 lr 0.000762 time 0.2728 (0.2839) loss 4.0371 (3.7236) grad_norm 1.1913 (inf) [2021-04-15 23:53:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][1070/1251] eta 0:00:51 lr 0.000762 time 0.2902 (0.2839) loss 3.7978 (3.7237) grad_norm 1.3564 (inf) [2021-04-15 23:53:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][1080/1251] eta 0:00:48 lr 0.000762 time 0.2654 (0.2838) loss 4.1297 (3.7222) grad_norm 1.3867 (inf) [2021-04-15 23:53:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][1090/1251] eta 0:00:45 lr 0.000762 time 0.2962 (0.2838) loss 3.9477 (3.7239) grad_norm 1.1842 (inf) [2021-04-15 23:53:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][1100/1251] eta 0:00:42 lr 0.000762 time 0.2922 (0.2837) loss 3.9817 (3.7237) grad_norm 1.3580 (inf) [2021-04-15 23:53:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][1110/1251] eta 0:00:39 lr 0.000762 time 0.2732 (0.2836) loss 3.6861 (3.7237) grad_norm 1.4591 (inf) [2021-04-15 23:53:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][1120/1251] eta 0:00:37 lr 0.000762 time 0.2766 (0.2836) loss 4.0807 (3.7223) grad_norm 1.3558 (inf) [2021-04-15 23:53:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][1130/1251] eta 0:00:34 lr 0.000762 time 0.2689 (0.2834) loss 3.9887 (3.7223) grad_norm 1.1540 (inf) [2021-04-15 23:53:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][1140/1251] eta 0:00:31 lr 0.000762 time 0.2891 (0.2834) loss 3.1412 (3.7208) grad_norm 1.3239 (inf) [2021-04-15 23:53:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][1150/1251] eta 0:00:28 lr 0.000762 time 0.2539 (0.2834) loss 2.8436 (3.7216) grad_norm 1.2765 (inf) [2021-04-15 23:53:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][1160/1251] eta 0:00:25 lr 0.000762 time 0.2818 (0.2834) loss 3.8922 (3.7197) grad_norm 1.3875 (inf) [2021-04-15 23:53:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][1170/1251] eta 0:00:22 lr 0.000762 time 0.3181 (0.2833) loss 2.6938 (3.7191) grad_norm 1.1669 (inf) [2021-04-15 23:53:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][1180/1251] eta 0:00:20 lr 0.000762 time 0.2650 (0.2833) loss 2.9523 (3.7184) grad_norm 1.1791 (inf) [2021-04-15 23:54:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][1190/1251] eta 0:00:17 lr 0.000762 time 0.2702 (0.2832) loss 3.7229 (3.7170) grad_norm 1.3754 (inf) [2021-04-15 23:54:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][1200/1251] eta 0:00:14 lr 0.000762 time 0.3220 (0.2832) loss 4.5256 (3.7183) grad_norm 1.2067 (inf) [2021-04-15 23:54:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][1210/1251] eta 0:00:11 lr 0.000762 time 0.2938 (0.2832) loss 4.1526 (3.7200) grad_norm 1.1180 (inf) [2021-04-15 23:54:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][1220/1251] eta 0:00:08 lr 0.000762 time 0.2635 (0.2832) loss 3.4485 (3.7196) grad_norm 1.4950 (inf) [2021-04-15 23:54:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][1230/1251] eta 0:00:05 lr 0.000761 time 0.3119 (0.2832) loss 4.0547 (3.7209) grad_norm 1.2032 (inf) [2021-04-15 23:54:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][1240/1251] eta 0:00:03 lr 0.000761 time 0.3272 (0.2831) loss 2.9113 (3.7215) grad_norm 1.0694 (inf) [2021-04-15 23:54:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [97/300][1250/1251] eta 0:00:00 lr 0.000761 time 0.2483 (0.2828) loss 4.4746 (3.7220) grad_norm 1.0771 (inf) [2021-04-15 23:54:21 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 97 training takes 0:05:56 [2021-04-15 23:54:21 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_97.pth saving...... [2021-04-15 23:54:36 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_97.pth saved !!! [2021-04-15 23:54:37 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.110 (1.110) Loss 1.2524 (1.2524) Acc@1 72.266 (72.266) Acc@5 90.527 (90.527) [2021-04-15 23:54:38 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.173 (0.232) Loss 1.1636 (1.1566) Acc@1 74.219 (73.677) Acc@5 91.211 (91.726) [2021-04-15 23:54:41 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.428 (0.238) Loss 1.1212 (1.1480) Acc@1 74.316 (73.870) Acc@5 92.480 (92.011) [2021-04-15 23:54:43 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.125 (0.234) Loss 1.1451 (1.1448) Acc@1 74.023 (73.942) Acc@5 92.188 (92.090) [2021-04-15 23:54:45 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.213) Loss 1.1428 (1.1487) Acc@1 72.852 (73.728) Acc@5 92.480 (92.021) [2021-04-15 23:54:48 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 73.692 Acc@5 92.074 [2021-04-15 23:54:48 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 73.7% [2021-04-15 23:54:48 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 73.69% [2021-04-15 23:54:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][0/1251] eta 2:33:07 lr 0.000761 time 7.3442 (7.3442) loss 3.1789 (3.1789) grad_norm 1.2967 (1.2967) [2021-04-15 23:54:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][10/1251] eta 0:18:51 lr 0.000761 time 0.2971 (0.9116) loss 3.5936 (3.5658) grad_norm 1.1572 (1.2421) [2021-04-15 23:55:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][20/1251] eta 0:12:29 lr 0.000761 time 0.2858 (0.6092) loss 4.4560 (3.6219) grad_norm 1.2581 (1.2552) [2021-04-15 23:55:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][30/1251] eta 0:10:13 lr 0.000761 time 0.2892 (0.5021) loss 3.8085 (3.6836) grad_norm 1.1530 (1.2379) [2021-04-15 23:55:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][40/1251] eta 0:09:01 lr 0.000761 time 0.2755 (0.4468) loss 3.3282 (3.7013) grad_norm 1.0991 (1.2355) [2021-04-15 23:55:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][50/1251] eta 0:08:17 lr 0.000761 time 0.2829 (0.4141) loss 4.2015 (3.7444) grad_norm 1.1802 (1.2325) [2021-04-15 23:55:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][60/1251] eta 0:07:46 lr 0.000761 time 0.2783 (0.3920) loss 3.5289 (3.7672) grad_norm 1.3421 (1.2340) [2021-04-15 23:55:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][70/1251] eta 0:07:24 lr 0.000761 time 0.3232 (0.3763) loss 3.8916 (3.7446) grad_norm 1.3598 (1.2291) [2021-04-15 23:55:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][80/1251] eta 0:07:06 lr 0.000761 time 0.2773 (0.3638) loss 3.7997 (3.7578) grad_norm 1.2984 (1.2239) [2021-04-15 23:55:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][90/1251] eta 0:06:51 lr 0.000761 time 0.2924 (0.3540) loss 3.5496 (3.7712) grad_norm 1.2385 (1.2186) [2021-04-15 23:55:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][100/1251] eta 0:06:41 lr 0.000761 time 0.3146 (0.3484) loss 3.4753 (3.7944) grad_norm 1.3803 (1.2239) [2021-04-15 23:55:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][110/1251] eta 0:06:30 lr 0.000761 time 0.3161 (0.3420) loss 3.2984 (3.7799) grad_norm 1.3250 (1.2401) [2021-04-15 23:55:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][120/1251] eta 0:06:21 lr 0.000761 time 0.2600 (0.3369) loss 4.4474 (3.8009) grad_norm 1.3756 (1.2479) [2021-04-15 23:55:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][130/1251] eta 0:06:14 lr 0.000761 time 0.2608 (0.3337) loss 2.7986 (3.7776) grad_norm 1.2798 (1.2500) [2021-04-15 23:55:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][140/1251] eta 0:06:07 lr 0.000761 time 0.2767 (0.3311) loss 4.6718 (3.7876) grad_norm 1.2996 (1.2521) [2021-04-15 23:55:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][150/1251] eta 0:06:01 lr 0.000761 time 0.2581 (0.3286) loss 3.8063 (3.7743) grad_norm 1.1569 (1.2496) [2021-04-15 23:55:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][160/1251] eta 0:05:55 lr 0.000761 time 0.2759 (0.3262) loss 3.8782 (3.7742) grad_norm 1.1308 (1.2537) [2021-04-15 23:55:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][170/1251] eta 0:05:50 lr 0.000761 time 0.2777 (0.3239) loss 4.0736 (3.7636) grad_norm 1.0117 (1.2508) [2021-04-15 23:55:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][180/1251] eta 0:05:44 lr 0.000761 time 0.2694 (0.3219) loss 4.0571 (3.7545) grad_norm 1.3766 (1.2510) [2021-04-15 23:55:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][190/1251] eta 0:05:39 lr 0.000761 time 0.2718 (0.3197) loss 3.3976 (3.7451) grad_norm 1.2045 (1.2508) [2021-04-15 23:55:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][200/1251] eta 0:05:33 lr 0.000761 time 0.2806 (0.3175) loss 3.4877 (3.7470) grad_norm 1.1786 (1.2514) [2021-04-15 23:55:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][210/1251] eta 0:05:28 lr 0.000761 time 0.3001 (0.3156) loss 4.7656 (3.7458) grad_norm 1.1172 (1.2524) [2021-04-15 23:55:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][220/1251] eta 0:05:23 lr 0.000761 time 0.2885 (0.3136) loss 4.3322 (3.7539) grad_norm 1.1790 (1.2501) [2021-04-15 23:56:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][230/1251] eta 0:05:18 lr 0.000761 time 0.2728 (0.3121) loss 3.7702 (3.7508) grad_norm 1.4541 (1.2508) [2021-04-15 23:56:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][240/1251] eta 0:05:13 lr 0.000761 time 0.2410 (0.3105) loss 2.8358 (3.7534) grad_norm 1.1655 (1.2524) [2021-04-15 23:56:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][250/1251] eta 0:05:09 lr 0.000761 time 0.2696 (0.3092) loss 2.8548 (3.7366) grad_norm 1.2654 (1.2533) [2021-04-15 23:56:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][260/1251] eta 0:05:05 lr 0.000761 time 0.2925 (0.3079) loss 2.5456 (3.7277) grad_norm 1.4740 (1.2552) [2021-04-15 23:56:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][270/1251] eta 0:05:00 lr 0.000760 time 0.2826 (0.3067) loss 2.7217 (3.7204) grad_norm 1.3741 (1.2551) [2021-04-15 23:56:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][280/1251] eta 0:04:56 lr 0.000760 time 0.2949 (0.3056) loss 3.1399 (3.7188) grad_norm 1.0675 (1.2525) [2021-04-15 23:56:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][290/1251] eta 0:04:52 lr 0.000760 time 0.2790 (0.3048) loss 4.4214 (3.7175) grad_norm 1.1690 (1.2528) [2021-04-15 23:56:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][300/1251] eta 0:04:49 lr 0.000760 time 0.2691 (0.3040) loss 4.4061 (3.7265) grad_norm 1.0797 (1.2525) [2021-04-15 23:56:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][310/1251] eta 0:04:45 lr 0.000760 time 0.2721 (0.3030) loss 2.7075 (3.7121) grad_norm 1.1711 (1.2483) [2021-04-15 23:56:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][320/1251] eta 0:04:41 lr 0.000760 time 0.2836 (0.3022) loss 3.0537 (3.7090) grad_norm 1.1805 (1.2468) [2021-04-15 23:56:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][330/1251] eta 0:04:37 lr 0.000760 time 0.2479 (0.3014) loss 4.4367 (3.7030) grad_norm 1.0588 (1.2479) [2021-04-15 23:56:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][340/1251] eta 0:04:34 lr 0.000760 time 0.2960 (0.3008) loss 3.8935 (3.7082) grad_norm 1.4259 (1.2476) [2021-04-15 23:56:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][350/1251] eta 0:04:30 lr 0.000760 time 0.3901 (0.3005) loss 3.7973 (3.7081) grad_norm 1.0931 (1.2458) [2021-04-15 23:56:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][360/1251] eta 0:04:27 lr 0.000760 time 0.3020 (0.3000) loss 3.5766 (3.7081) grad_norm 1.2505 (1.2435) [2021-04-15 23:56:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][370/1251] eta 0:04:23 lr 0.000760 time 0.2589 (0.2993) loss 3.2114 (3.7046) grad_norm 1.2435 (1.2419) [2021-04-15 23:56:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][380/1251] eta 0:04:20 lr 0.000760 time 0.2612 (0.2987) loss 4.4251 (3.7022) grad_norm 1.4434 (1.2411) [2021-04-15 23:56:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][390/1251] eta 0:04:16 lr 0.000760 time 0.2758 (0.2982) loss 3.7136 (3.7037) grad_norm 1.1825 (1.2416) [2021-04-15 23:56:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][400/1251] eta 0:04:13 lr 0.000760 time 0.2679 (0.2980) loss 3.3191 (3.6980) grad_norm 1.2674 (1.2463) [2021-04-15 23:56:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][410/1251] eta 0:04:10 lr 0.000760 time 0.2773 (0.2974) loss 3.6553 (3.7058) grad_norm 1.2822 (1.2449) [2021-04-15 23:56:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][420/1251] eta 0:04:06 lr 0.000760 time 0.2560 (0.2968) loss 3.5885 (3.7051) grad_norm 1.1151 (1.2443) [2021-04-15 23:56:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][430/1251] eta 0:04:03 lr 0.000760 time 0.2851 (0.2964) loss 3.9923 (3.7031) grad_norm 1.1466 (1.2437) [2021-04-15 23:56:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][440/1251] eta 0:03:59 lr 0.000760 time 0.2678 (0.2959) loss 3.3331 (3.7066) grad_norm 1.1343 (1.2423) [2021-04-15 23:57:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][450/1251] eta 0:03:56 lr 0.000760 time 0.3931 (0.2957) loss 4.3711 (3.7012) grad_norm 1.1538 (1.2428) [2021-04-15 23:57:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][460/1251] eta 0:03:53 lr 0.000760 time 0.2760 (0.2953) loss 3.7069 (3.6943) grad_norm 1.2976 (1.2449) [2021-04-15 23:57:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][470/1251] eta 0:03:50 lr 0.000760 time 0.2896 (0.2950) loss 3.9728 (3.6962) grad_norm 1.0621 (1.2447) [2021-04-15 23:57:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][480/1251] eta 0:03:47 lr 0.000760 time 0.2743 (0.2945) loss 3.5679 (3.7004) grad_norm 1.2848 (1.2441) [2021-04-15 23:57:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][490/1251] eta 0:03:44 lr 0.000760 time 0.2723 (0.2945) loss 4.0232 (3.7027) grad_norm 1.4780 (1.2439) [2021-04-15 23:57:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][500/1251] eta 0:03:40 lr 0.000760 time 0.2876 (0.2941) loss 3.1570 (3.6981) grad_norm 1.5555 (1.2476) [2021-04-15 23:57:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][510/1251] eta 0:03:37 lr 0.000760 time 0.2803 (0.2936) loss 3.8091 (3.7003) grad_norm 1.3995 (1.2486) [2021-04-15 23:57:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][520/1251] eta 0:03:34 lr 0.000760 time 0.2797 (0.2935) loss 4.5800 (3.7008) grad_norm 1.1218 (1.2492) [2021-04-15 23:57:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][530/1251] eta 0:03:31 lr 0.000760 time 0.2646 (0.2931) loss 3.2550 (3.6938) grad_norm 1.1171 (1.2490) [2021-04-15 23:57:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][540/1251] eta 0:03:28 lr 0.000760 time 0.2693 (0.2928) loss 4.1152 (3.6943) grad_norm 1.3419 (1.2481) [2021-04-15 23:57:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][550/1251] eta 0:03:25 lr 0.000759 time 0.2469 (0.2926) loss 3.8338 (3.6970) grad_norm 1.2551 (1.2478) [2021-04-15 23:57:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][560/1251] eta 0:03:21 lr 0.000759 time 0.2708 (0.2922) loss 3.7839 (3.6987) grad_norm 1.2039 (1.2465) [2021-04-15 23:57:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][570/1251] eta 0:03:19 lr 0.000759 time 0.4205 (0.2925) loss 3.2160 (3.7027) grad_norm 1.1969 (1.2456) [2021-04-15 23:57:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][580/1251] eta 0:03:16 lr 0.000759 time 0.3024 (0.2924) loss 2.7346 (3.7009) grad_norm 1.5530 (1.2459) [2021-04-15 23:57:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][590/1251] eta 0:03:13 lr 0.000759 time 0.2558 (0.2921) loss 4.1252 (3.6971) grad_norm 1.2776 (1.2454) [2021-04-15 23:57:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][600/1251] eta 0:03:09 lr 0.000759 time 0.2721 (0.2918) loss 2.7525 (3.6948) grad_norm 1.2082 (1.2453) [2021-04-15 23:57:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][610/1251] eta 0:03:06 lr 0.000759 time 0.2744 (0.2916) loss 4.1025 (3.6983) grad_norm 1.4399 (1.2473) [2021-04-15 23:57:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][620/1251] eta 0:03:03 lr 0.000759 time 0.2743 (0.2914) loss 4.2294 (3.7006) grad_norm 1.3313 (1.2483) [2021-04-15 23:57:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][630/1251] eta 0:03:00 lr 0.000759 time 0.2579 (0.2911) loss 3.5838 (3.6985) grad_norm 1.1423 (1.2496) [2021-04-15 23:57:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][640/1251] eta 0:02:57 lr 0.000759 time 0.2685 (0.2909) loss 4.0613 (3.7014) grad_norm 1.2908 (1.2507) [2021-04-15 23:57:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][650/1251] eta 0:02:54 lr 0.000759 time 0.2932 (0.2907) loss 4.7986 (3.7027) grad_norm 1.1620 (1.2514) [2021-04-15 23:58:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][660/1251] eta 0:02:51 lr 0.000759 time 0.2681 (0.2905) loss 4.3449 (3.7034) grad_norm 1.3893 (1.2513) [2021-04-15 23:58:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][670/1251] eta 0:02:48 lr 0.000759 time 0.2646 (0.2903) loss 2.7081 (3.7027) grad_norm 1.0848 (1.2510) [2021-04-15 23:58:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][680/1251] eta 0:02:45 lr 0.000759 time 0.2652 (0.2903) loss 2.5157 (3.7011) grad_norm 1.3164 (1.2525) [2021-04-15 23:58:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][690/1251] eta 0:02:42 lr 0.000759 time 0.2636 (0.2902) loss 3.9572 (3.7004) grad_norm 1.1933 (1.2526) [2021-04-15 23:58:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][700/1251] eta 0:02:39 lr 0.000759 time 0.4062 (0.2902) loss 2.5540 (3.6984) grad_norm 1.2229 (1.2518) [2021-04-15 23:58:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][710/1251] eta 0:02:36 lr 0.000759 time 0.2583 (0.2899) loss 3.7858 (3.7005) grad_norm 1.1846 (1.2508) [2021-04-15 23:58:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][720/1251] eta 0:02:33 lr 0.000759 time 0.2975 (0.2900) loss 3.9274 (3.6996) grad_norm 1.0508 (1.2503) [2021-04-15 23:58:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][730/1251] eta 0:02:31 lr 0.000759 time 0.4126 (0.2900) loss 3.4307 (3.6960) grad_norm 1.0302 (1.2489) [2021-04-15 23:58:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][740/1251] eta 0:02:28 lr 0.000759 time 0.2753 (0.2899) loss 3.1540 (3.6956) grad_norm 1.0690 (1.2484) [2021-04-15 23:58:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][750/1251] eta 0:02:25 lr 0.000759 time 0.2609 (0.2897) loss 4.1148 (3.6982) grad_norm 1.1519 (1.2486) [2021-04-15 23:58:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][760/1251] eta 0:02:22 lr 0.000759 time 0.2674 (0.2896) loss 4.4485 (3.6962) grad_norm 1.2415 (1.2482) [2021-04-15 23:58:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][770/1251] eta 0:02:19 lr 0.000759 time 0.2889 (0.2895) loss 2.6529 (3.6928) grad_norm 1.0804 (1.2483) [2021-04-15 23:58:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][780/1251] eta 0:02:16 lr 0.000759 time 0.2637 (0.2893) loss 5.1192 (3.6980) grad_norm 1.4571 (1.2505) [2021-04-15 23:58:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][790/1251] eta 0:02:13 lr 0.000759 time 0.2724 (0.2891) loss 3.9710 (3.6986) grad_norm 1.5188 (1.2508) [2021-04-15 23:58:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][800/1251] eta 0:02:10 lr 0.000759 time 0.2848 (0.2890) loss 3.5774 (3.6968) grad_norm 1.2512 (1.2513) [2021-04-15 23:58:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][810/1251] eta 0:02:07 lr 0.000759 time 0.2957 (0.2888) loss 3.3885 (3.6921) grad_norm 1.4618 (1.2520) [2021-04-15 23:58:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][820/1251] eta 0:02:04 lr 0.000759 time 0.2693 (0.2886) loss 3.6871 (3.6915) grad_norm 1.4413 (1.2519) [2021-04-15 23:58:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][830/1251] eta 0:02:01 lr 0.000758 time 0.2754 (0.2886) loss 3.9042 (3.6908) grad_norm 1.2047 (1.2521) [2021-04-15 23:58:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][840/1251] eta 0:01:58 lr 0.000758 time 0.2770 (0.2884) loss 4.0254 (3.6908) grad_norm 1.2143 (1.2529) [2021-04-15 23:58:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][850/1251] eta 0:01:55 lr 0.000758 time 0.2891 (0.2883) loss 3.9082 (3.6911) grad_norm 1.2449 (1.2535) [2021-04-15 23:58:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][860/1251] eta 0:01:52 lr 0.000758 time 0.2813 (0.2883) loss 4.4928 (3.6910) grad_norm 1.1671 (1.2529) [2021-04-15 23:58:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][870/1251] eta 0:01:49 lr 0.000758 time 0.2834 (0.2882) loss 3.9827 (3.6940) grad_norm 1.2781 (1.2531) [2021-04-15 23:59:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][880/1251] eta 0:01:46 lr 0.000758 time 0.2815 (0.2881) loss 2.9353 (3.6953) grad_norm 1.3598 (1.2530) [2021-04-15 23:59:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][890/1251] eta 0:01:43 lr 0.000758 time 0.2623 (0.2879) loss 3.4302 (3.6960) grad_norm 1.1839 (1.2529) [2021-04-15 23:59:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][900/1251] eta 0:01:41 lr 0.000758 time 0.2864 (0.2878) loss 2.9705 (3.6948) grad_norm 1.3410 (1.2532) [2021-04-15 23:59:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][910/1251] eta 0:01:38 lr 0.000758 time 0.2606 (0.2876) loss 3.6202 (3.6965) grad_norm 1.2065 (1.2530) [2021-04-15 23:59:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][920/1251] eta 0:01:35 lr 0.000758 time 0.2428 (0.2875) loss 4.1557 (3.6971) grad_norm 1.2413 (1.2533) [2021-04-15 23:59:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][930/1251] eta 0:01:32 lr 0.000758 time 0.2614 (0.2874) loss 4.2895 (3.6957) grad_norm 1.3628 (1.2539) [2021-04-15 23:59:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][940/1251] eta 0:01:29 lr 0.000758 time 0.2768 (0.2873) loss 4.0833 (3.6986) grad_norm 1.2455 (1.2537) [2021-04-15 23:59:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][950/1251] eta 0:01:26 lr 0.000758 time 0.2787 (0.2872) loss 4.1188 (3.7004) grad_norm 1.1509 (1.2532) [2021-04-15 23:59:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][960/1251] eta 0:01:23 lr 0.000758 time 0.2709 (0.2870) loss 4.2092 (3.7012) grad_norm 1.2233 (1.2525) [2021-04-15 23:59:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][970/1251] eta 0:01:20 lr 0.000758 time 0.2657 (0.2869) loss 3.0738 (3.7030) grad_norm 1.4012 (1.2538) [2021-04-15 23:59:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][980/1251] eta 0:01:17 lr 0.000758 time 0.2597 (0.2869) loss 3.4427 (3.7037) grad_norm 1.3114 (1.2539) [2021-04-15 23:59:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][990/1251] eta 0:01:14 lr 0.000758 time 0.2829 (0.2868) loss 2.9072 (3.7040) grad_norm 1.3277 (1.2544) [2021-04-15 23:59:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][1000/1251] eta 0:01:11 lr 0.000758 time 0.2755 (0.2868) loss 4.3885 (3.7046) grad_norm 1.1592 (1.2546) [2021-04-15 23:59:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][1010/1251] eta 0:01:09 lr 0.000758 time 0.2615 (0.2866) loss 4.0411 (3.7028) grad_norm 1.3494 (1.2555) [2021-04-15 23:59:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][1020/1251] eta 0:01:06 lr 0.000758 time 0.2933 (0.2866) loss 3.8615 (3.7028) grad_norm 1.1354 (1.2561) [2021-04-15 23:59:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][1030/1251] eta 0:01:03 lr 0.000758 time 0.2668 (0.2866) loss 4.0305 (3.7003) grad_norm 1.2130 (1.2560) [2021-04-15 23:59:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][1040/1251] eta 0:01:00 lr 0.000758 time 0.2724 (0.2865) loss 3.7426 (3.6986) grad_norm 1.4257 (1.2565) [2021-04-15 23:59:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][1050/1251] eta 0:00:57 lr 0.000758 time 0.2749 (0.2864) loss 3.9348 (3.6987) grad_norm 1.2650 (1.2561) [2021-04-15 23:59:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][1060/1251] eta 0:00:54 lr 0.000758 time 0.2772 (0.2863) loss 3.7216 (3.6992) grad_norm 1.2051 (1.2564) [2021-04-15 23:59:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][1070/1251] eta 0:00:51 lr 0.000758 time 0.2475 (0.2863) loss 3.9456 (3.7012) grad_norm 1.0805 (1.2557) [2021-04-15 23:59:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][1080/1251] eta 0:00:48 lr 0.000758 time 0.2795 (0.2862) loss 4.0936 (3.6994) grad_norm 1.2625 (1.2553) [2021-04-16 00:00:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][1090/1251] eta 0:00:46 lr 0.000758 time 0.2818 (0.2863) loss 3.4204 (3.6995) grad_norm 1.7239 (1.2561) [2021-04-16 00:00:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][1100/1251] eta 0:00:43 lr 0.000758 time 0.2925 (0.2862) loss 3.6515 (3.7010) grad_norm 1.3236 (1.2559) [2021-04-16 00:00:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][1110/1251] eta 0:00:40 lr 0.000757 time 0.2972 (0.2862) loss 3.4680 (3.7006) grad_norm 1.2005 (1.2563) [2021-04-16 00:00:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][1120/1251] eta 0:00:37 lr 0.000757 time 0.2785 (0.2861) loss 3.9952 (3.6994) grad_norm 1.2334 (1.2571) [2021-04-16 00:00:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][1130/1251] eta 0:00:34 lr 0.000757 time 0.2610 (0.2860) loss 4.2587 (3.7003) grad_norm 1.2476 (1.2568) [2021-04-16 00:00:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][1140/1251] eta 0:00:31 lr 0.000757 time 0.2760 (0.2858) loss 4.0572 (3.7001) grad_norm 1.2454 (1.2566) [2021-04-16 00:00:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][1150/1251] eta 0:00:28 lr 0.000757 time 0.2643 (0.2860) loss 3.3005 (3.7020) grad_norm 1.3394 (1.2563) [2021-04-16 00:00:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][1160/1251] eta 0:00:26 lr 0.000757 time 0.2868 (0.2859) loss 4.3832 (3.7027) grad_norm 1.0708 (1.2554) [2021-04-16 00:00:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][1170/1251] eta 0:00:23 lr 0.000757 time 0.2622 (0.2858) loss 3.6602 (3.7051) grad_norm 1.4530 (1.2551) [2021-04-16 00:00:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][1180/1251] eta 0:00:20 lr 0.000757 time 0.2913 (0.2857) loss 2.5341 (3.7044) grad_norm 1.4568 (1.2556) [2021-04-16 00:00:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][1190/1251] eta 0:00:17 lr 0.000757 time 0.2516 (0.2857) loss 3.9114 (3.7058) grad_norm 1.2447 (1.2556) [2021-04-16 00:00:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][1200/1251] eta 0:00:14 lr 0.000757 time 0.2843 (0.2856) loss 2.9355 (3.7054) grad_norm 1.0698 (1.2551) [2021-04-16 00:00:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][1210/1251] eta 0:00:11 lr 0.000757 time 0.2918 (0.2855) loss 4.3953 (3.7061) grad_norm 1.1103 (1.2556) [2021-04-16 00:00:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][1220/1251] eta 0:00:08 lr 0.000757 time 0.2653 (0.2855) loss 4.3596 (3.7065) grad_norm 1.0704 (1.2552) [2021-04-16 00:00:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][1230/1251] eta 0:00:05 lr 0.000757 time 0.2694 (0.2853) loss 3.3947 (3.7074) grad_norm 1.2162 (1.2552) [2021-04-16 00:00:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][1240/1251] eta 0:00:03 lr 0.000757 time 0.2488 (0.2852) loss 3.6659 (3.7079) grad_norm 1.1194 (1.2551) [2021-04-16 00:00:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [98/300][1250/1251] eta 0:00:00 lr 0.000757 time 0.2487 (0.2849) loss 4.1229 (3.7070) grad_norm 1.8377 (1.2563) [2021-04-16 00:00:48 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 98 training takes 0:05:59 [2021-04-16 00:00:48 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_98.pth saving...... [2021-04-16 00:01:01 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_98.pth saved !!! [2021-04-16 00:01:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.176 (1.176) Loss 1.0547 (1.0547) Acc@1 76.367 (76.367) Acc@5 93.848 (93.848) [2021-04-16 00:01:03 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.165 (0.218) Loss 1.1245 (1.1317) Acc@1 73.633 (73.739) Acc@5 92.773 (92.205) [2021-04-16 00:01:05 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.432 (0.217) Loss 1.0995 (1.1367) Acc@1 74.219 (73.531) Acc@5 92.188 (92.039) [2021-04-16 00:01:08 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.095 (0.218) Loss 1.1780 (1.1295) Acc@1 71.777 (73.526) Acc@5 91.699 (92.197) [2021-04-16 00:01:10 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.216) Loss 1.1772 (1.1260) Acc@1 72.949 (73.702) Acc@5 91.113 (92.166) [2021-04-16 00:01:17 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 73.614 Acc@5 92.150 [2021-04-16 00:01:17 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 73.6% [2021-04-16 00:01:17 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 73.69% [2021-04-16 00:01:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][0/1251] eta 1:27:53 lr 0.000757 time 4.2153 (4.2153) loss 3.7363 (3.7363) grad_norm 1.4422 (1.4422) [2021-04-16 00:01:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][10/1251] eta 0:13:07 lr 0.000757 time 0.2734 (0.6345) loss 2.7654 (3.5881) grad_norm 1.1952 (1.1674) [2021-04-16 00:01:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][20/1251] eta 0:09:33 lr 0.000757 time 0.2900 (0.4661) loss 4.2216 (3.7681) grad_norm 1.1859 (1.2195) [2021-04-16 00:01:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][30/1251] eta 0:08:15 lr 0.000757 time 0.2917 (0.4058) loss 3.6289 (3.7295) grad_norm 1.2301 (1.2529) [2021-04-16 00:01:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][40/1251] eta 0:07:34 lr 0.000757 time 0.2792 (0.3752) loss 4.1063 (3.7418) grad_norm 1.5512 (1.2549) [2021-04-16 00:01:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][50/1251] eta 0:07:07 lr 0.000757 time 0.2872 (0.3562) loss 3.3353 (3.7217) grad_norm 1.2245 (1.2493) [2021-04-16 00:01:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][60/1251] eta 0:06:48 lr 0.000757 time 0.2699 (0.3428) loss 3.9238 (3.7490) grad_norm 1.1641 (1.2467) [2021-04-16 00:01:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][70/1251] eta 0:06:34 lr 0.000757 time 0.2731 (0.3338) loss 3.4529 (3.7164) grad_norm 1.1108 (1.2376) [2021-04-16 00:01:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][80/1251] eta 0:06:22 lr 0.000757 time 0.2933 (0.3265) loss 2.6461 (3.7168) grad_norm 1.2907 (1.2299) [2021-04-16 00:01:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][90/1251] eta 0:06:14 lr 0.000757 time 0.3054 (0.3223) loss 4.1166 (3.7183) grad_norm 1.1309 (1.2305) [2021-04-16 00:01:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][100/1251] eta 0:06:05 lr 0.000757 time 0.3171 (0.3180) loss 2.3675 (3.7065) grad_norm 1.3201 (1.2337) [2021-04-16 00:01:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][110/1251] eta 0:05:58 lr 0.000757 time 0.2667 (0.3139) loss 3.4723 (3.7017) grad_norm 1.2240 (1.2322) [2021-04-16 00:01:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][120/1251] eta 0:05:51 lr 0.000757 time 0.2939 (0.3112) loss 4.4167 (3.7003) grad_norm 1.1348 (1.2272) [2021-04-16 00:01:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][130/1251] eta 0:05:46 lr 0.000757 time 0.2453 (0.3092) loss 3.9732 (3.7201) grad_norm 1.4287 (1.2333) [2021-04-16 00:02:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][140/1251] eta 0:05:40 lr 0.000756 time 0.2738 (0.3068) loss 4.1334 (3.7268) grad_norm 1.2858 (1.2302) [2021-04-16 00:02:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][150/1251] eta 0:05:35 lr 0.000756 time 0.2627 (0.3044) loss 2.9791 (3.7246) grad_norm 1.3498 (1.2353) [2021-04-16 00:02:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][160/1251] eta 0:05:30 lr 0.000756 time 0.2506 (0.3026) loss 4.0322 (3.7226) grad_norm 1.6279 (1.2435) [2021-04-16 00:02:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][170/1251] eta 0:05:25 lr 0.000756 time 0.2886 (0.3008) loss 3.9026 (3.7128) grad_norm 1.3092 (1.2445) [2021-04-16 00:02:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][180/1251] eta 0:05:21 lr 0.000756 time 0.4450 (0.3004) loss 4.0507 (3.7100) grad_norm 1.1859 (1.2434) [2021-04-16 00:02:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][190/1251] eta 0:05:17 lr 0.000756 time 0.2673 (0.2988) loss 3.9735 (3.7096) grad_norm 1.0122 (1.2432) [2021-04-16 00:02:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][200/1251] eta 0:05:12 lr 0.000756 time 0.2538 (0.2977) loss 4.1634 (3.7076) grad_norm 1.1111 (1.2439) [2021-04-16 00:02:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][210/1251] eta 0:05:08 lr 0.000756 time 0.2785 (0.2968) loss 4.4784 (3.6927) grad_norm 1.3875 (1.2458) [2021-04-16 00:02:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][220/1251] eta 0:05:05 lr 0.000756 time 0.2709 (0.2958) loss 4.1613 (3.6926) grad_norm 1.0526 (1.2455) [2021-04-16 00:02:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][230/1251] eta 0:05:01 lr 0.000756 time 0.2941 (0.2956) loss 2.6126 (3.6754) grad_norm 1.2483 (1.2479) [2021-04-16 00:02:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][240/1251] eta 0:04:57 lr 0.000756 time 0.2652 (0.2947) loss 3.6935 (3.6792) grad_norm 1.1270 (1.2480) [2021-04-16 00:02:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][250/1251] eta 0:04:54 lr 0.000756 time 0.2731 (0.2945) loss 3.4162 (3.6724) grad_norm 1.2794 (1.2469) [2021-04-16 00:02:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][260/1251] eta 0:04:50 lr 0.000756 time 0.2875 (0.2936) loss 4.4135 (3.6854) grad_norm 1.2899 (inf) [2021-04-16 00:02:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][270/1251] eta 0:04:47 lr 0.000756 time 0.2865 (0.2930) loss 4.0610 (3.6878) grad_norm 1.2943 (inf) [2021-04-16 00:02:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][280/1251] eta 0:04:44 lr 0.000756 time 0.3043 (0.2926) loss 2.5898 (3.6768) grad_norm 1.4350 (inf) [2021-04-16 00:02:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][290/1251] eta 0:04:41 lr 0.000756 time 0.2707 (0.2926) loss 4.1362 (3.6854) grad_norm 1.2031 (inf) [2021-04-16 00:02:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][300/1251] eta 0:04:38 lr 0.000756 time 0.2872 (0.2926) loss 2.5030 (3.6825) grad_norm 1.1929 (inf) [2021-04-16 00:02:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][310/1251] eta 0:04:34 lr 0.000756 time 0.2892 (0.2922) loss 4.5431 (3.6961) grad_norm 1.1095 (inf) [2021-04-16 00:02:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][320/1251] eta 0:04:31 lr 0.000756 time 0.2692 (0.2916) loss 2.4837 (3.6914) grad_norm 1.1369 (inf) [2021-04-16 00:02:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][330/1251] eta 0:04:28 lr 0.000756 time 0.2698 (0.2912) loss 4.0451 (3.6966) grad_norm 1.1551 (inf) [2021-04-16 00:02:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][340/1251] eta 0:04:25 lr 0.000756 time 0.2846 (0.2910) loss 3.6822 (3.6953) grad_norm 1.4114 (inf) [2021-04-16 00:02:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][350/1251] eta 0:04:21 lr 0.000756 time 0.2659 (0.2906) loss 2.6314 (3.6945) grad_norm 1.0278 (inf) [2021-04-16 00:03:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][360/1251] eta 0:04:18 lr 0.000756 time 0.2685 (0.2905) loss 4.4567 (3.7012) grad_norm 1.1629 (inf) [2021-04-16 00:03:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][370/1251] eta 0:04:15 lr 0.000756 time 0.2580 (0.2904) loss 3.8846 (3.6997) grad_norm 1.1705 (inf) [2021-04-16 00:03:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][380/1251] eta 0:04:12 lr 0.000756 time 0.2783 (0.2900) loss 4.0670 (3.7007) grad_norm 1.2879 (inf) [2021-04-16 00:03:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][390/1251] eta 0:04:09 lr 0.000756 time 0.2750 (0.2897) loss 3.3535 (3.7035) grad_norm 1.3080 (inf) [2021-04-16 00:03:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][400/1251] eta 0:04:06 lr 0.000756 time 0.2637 (0.2894) loss 2.5391 (3.7041) grad_norm 1.1663 (inf) [2021-04-16 00:03:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][410/1251] eta 0:04:03 lr 0.000756 time 0.2902 (0.2891) loss 2.5922 (3.6956) grad_norm 1.2270 (inf) [2021-04-16 00:03:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][420/1251] eta 0:03:59 lr 0.000755 time 0.2576 (0.2888) loss 4.1246 (3.7018) grad_norm 1.3393 (inf) [2021-04-16 00:03:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][430/1251] eta 0:03:56 lr 0.000755 time 0.2733 (0.2885) loss 4.1215 (3.7023) grad_norm 1.1643 (inf) [2021-04-16 00:03:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][440/1251] eta 0:03:53 lr 0.000755 time 0.2893 (0.2882) loss 3.9638 (3.7033) grad_norm 1.3496 (inf) [2021-04-16 00:03:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][450/1251] eta 0:03:50 lr 0.000755 time 0.3035 (0.2880) loss 3.7810 (3.7037) grad_norm 1.3969 (inf) [2021-04-16 00:03:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][460/1251] eta 0:03:47 lr 0.000755 time 0.2671 (0.2876) loss 3.7867 (3.7053) grad_norm 1.3711 (inf) [2021-04-16 00:03:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][470/1251] eta 0:03:44 lr 0.000755 time 0.2676 (0.2875) loss 3.7127 (3.7114) grad_norm 1.2396 (inf) [2021-04-16 00:03:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][480/1251] eta 0:03:41 lr 0.000755 time 0.2589 (0.2874) loss 4.1356 (3.7162) grad_norm 1.2132 (inf) [2021-04-16 00:03:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][490/1251] eta 0:03:38 lr 0.000755 time 0.3048 (0.2875) loss 4.1983 (3.7131) grad_norm 1.2206 (inf) [2021-04-16 00:03:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][500/1251] eta 0:03:35 lr 0.000755 time 0.2545 (0.2873) loss 3.6151 (3.7113) grad_norm 1.1930 (inf) [2021-04-16 00:03:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][510/1251] eta 0:03:32 lr 0.000755 time 0.2567 (0.2870) loss 3.3245 (3.7090) grad_norm 1.1185 (inf) [2021-04-16 00:03:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][520/1251] eta 0:03:29 lr 0.000755 time 0.2725 (0.2868) loss 3.4829 (3.7075) grad_norm 1.6451 (inf) [2021-04-16 00:03:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][530/1251] eta 0:03:26 lr 0.000755 time 0.2710 (0.2866) loss 3.6640 (3.7094) grad_norm 1.2360 (inf) [2021-04-16 00:03:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][540/1251] eta 0:03:23 lr 0.000755 time 0.2613 (0.2864) loss 2.2888 (3.7101) grad_norm 1.1514 (inf) [2021-04-16 00:03:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][550/1251] eta 0:03:20 lr 0.000755 time 0.2811 (0.2863) loss 3.8325 (3.7080) grad_norm 1.2610 (inf) [2021-04-16 00:03:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][560/1251] eta 0:03:17 lr 0.000755 time 0.2825 (0.2861) loss 4.0150 (3.7066) grad_norm 1.3513 (inf) [2021-04-16 00:04:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][570/1251] eta 0:03:15 lr 0.000755 time 0.2902 (0.2866) loss 3.5146 (3.7013) grad_norm 1.1148 (inf) [2021-04-16 00:04:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][580/1251] eta 0:03:12 lr 0.000755 time 0.2723 (0.2871) loss 2.8563 (3.7033) grad_norm 1.1038 (inf) [2021-04-16 00:04:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][590/1251] eta 0:03:09 lr 0.000755 time 0.2806 (0.2868) loss 3.3317 (3.7009) grad_norm 1.2566 (inf) [2021-04-16 00:04:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][600/1251] eta 0:03:06 lr 0.000755 time 0.2917 (0.2866) loss 3.9013 (3.7020) grad_norm 1.0909 (inf) [2021-04-16 00:04:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][610/1251] eta 0:03:03 lr 0.000755 time 0.2719 (0.2864) loss 3.0572 (3.7015) grad_norm 1.3814 (inf) [2021-04-16 00:04:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][620/1251] eta 0:03:00 lr 0.000755 time 0.2711 (0.2863) loss 4.1278 (3.7016) grad_norm 1.1525 (inf) [2021-04-16 00:04:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][630/1251] eta 0:02:57 lr 0.000755 time 0.2725 (0.2862) loss 2.8536 (3.7014) grad_norm 1.2206 (inf) [2021-04-16 00:04:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][640/1251] eta 0:02:54 lr 0.000755 time 0.2890 (0.2860) loss 3.6588 (3.7032) grad_norm 1.2443 (inf) [2021-04-16 00:04:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][650/1251] eta 0:02:51 lr 0.000755 time 0.2596 (0.2859) loss 4.4059 (3.7013) grad_norm 1.3192 (inf) [2021-04-16 00:04:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][660/1251] eta 0:02:48 lr 0.000755 time 0.2661 (0.2857) loss 4.1230 (3.7029) grad_norm 1.1413 (inf) [2021-04-16 00:04:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][670/1251] eta 0:02:45 lr 0.000755 time 0.2909 (0.2856) loss 3.7551 (3.7065) grad_norm 1.3323 (inf) [2021-04-16 00:04:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][680/1251] eta 0:02:42 lr 0.000755 time 0.2674 (0.2854) loss 3.5660 (3.7057) grad_norm 1.3490 (inf) [2021-04-16 00:04:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][690/1251] eta 0:02:40 lr 0.000755 time 0.2719 (0.2854) loss 3.4301 (3.7069) grad_norm 1.3130 (inf) [2021-04-16 00:04:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][700/1251] eta 0:02:37 lr 0.000754 time 0.2464 (0.2852) loss 3.1108 (3.7074) grad_norm 1.5013 (inf) [2021-04-16 00:04:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][710/1251] eta 0:02:34 lr 0.000754 time 0.2929 (0.2854) loss 4.0441 (3.7081) grad_norm 1.3124 (inf) [2021-04-16 00:04:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][720/1251] eta 0:02:31 lr 0.000754 time 0.2826 (0.2853) loss 3.6393 (3.7058) grad_norm 1.4433 (inf) [2021-04-16 00:04:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][730/1251] eta 0:02:28 lr 0.000754 time 0.2688 (0.2851) loss 4.6030 (3.7086) grad_norm 1.3038 (inf) [2021-04-16 00:04:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][740/1251] eta 0:02:25 lr 0.000754 time 0.2622 (0.2849) loss 4.0684 (3.7095) grad_norm 1.2911 (inf) [2021-04-16 00:04:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][750/1251] eta 0:02:22 lr 0.000754 time 0.2654 (0.2849) loss 4.3381 (3.7107) grad_norm 1.1858 (inf) [2021-04-16 00:04:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][760/1251] eta 0:02:19 lr 0.000754 time 0.2737 (0.2849) loss 3.3737 (3.7101) grad_norm 1.3967 (inf) [2021-04-16 00:04:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][770/1251] eta 0:02:16 lr 0.000754 time 0.2594 (0.2848) loss 4.2326 (3.7071) grad_norm 1.2304 (inf) [2021-04-16 00:05:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][780/1251] eta 0:02:14 lr 0.000754 time 0.3082 (0.2847) loss 2.9263 (3.7074) grad_norm 1.2343 (inf) [2021-04-16 00:05:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][790/1251] eta 0:02:11 lr 0.000754 time 0.2832 (0.2849) loss 4.1584 (3.7099) grad_norm 1.3912 (inf) [2021-04-16 00:05:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][800/1251] eta 0:02:08 lr 0.000754 time 0.2815 (0.2849) loss 4.0038 (3.7086) grad_norm 1.1646 (inf) [2021-04-16 00:05:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][810/1251] eta 0:02:05 lr 0.000754 time 0.2955 (0.2848) loss 3.1030 (3.7099) grad_norm 1.1650 (inf) [2021-04-16 00:05:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][820/1251] eta 0:02:02 lr 0.000754 time 0.2618 (0.2847) loss 4.4088 (3.7128) grad_norm 1.2381 (inf) [2021-04-16 00:05:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][830/1251] eta 0:01:59 lr 0.000754 time 0.2782 (0.2848) loss 3.6657 (3.7137) grad_norm 1.3407 (inf) [2021-04-16 00:05:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][840/1251] eta 0:01:56 lr 0.000754 time 0.2709 (0.2846) loss 4.1377 (3.7115) grad_norm 1.3741 (inf) [2021-04-16 00:05:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][850/1251] eta 0:01:54 lr 0.000754 time 0.2846 (0.2844) loss 4.0647 (3.7113) grad_norm 1.3555 (inf) [2021-04-16 00:05:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][860/1251] eta 0:01:51 lr 0.000754 time 0.2565 (0.2843) loss 4.1325 (3.7123) grad_norm 1.2036 (inf) [2021-04-16 00:05:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][870/1251] eta 0:01:48 lr 0.000754 time 0.2634 (0.2843) loss 3.3435 (3.7124) grad_norm 1.3662 (inf) [2021-04-16 00:05:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][880/1251] eta 0:01:45 lr 0.000754 time 0.3155 (0.2844) loss 3.6492 (3.7130) grad_norm 1.3560 (inf) [2021-04-16 00:05:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][890/1251] eta 0:01:42 lr 0.000754 time 0.2731 (0.2843) loss 3.5044 (3.7130) grad_norm 1.1485 (inf) [2021-04-16 00:05:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][900/1251] eta 0:01:39 lr 0.000754 time 0.2632 (0.2842) loss 4.2582 (3.7139) grad_norm 1.3103 (inf) [2021-04-16 00:05:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][910/1251] eta 0:01:36 lr 0.000754 time 0.2705 (0.2841) loss 3.4969 (3.7139) grad_norm 1.3555 (inf) [2021-04-16 00:05:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][920/1251] eta 0:01:34 lr 0.000754 time 0.2447 (0.2840) loss 4.0033 (3.7160) grad_norm 1.3479 (inf) [2021-04-16 00:05:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][930/1251] eta 0:01:31 lr 0.000754 time 0.3300 (0.2841) loss 4.2139 (3.7173) grad_norm 1.2626 (inf) [2021-04-16 00:05:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][940/1251] eta 0:01:28 lr 0.000754 time 0.3068 (0.2841) loss 3.7120 (3.7180) grad_norm 1.1776 (inf) [2021-04-16 00:05:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][950/1251] eta 0:01:25 lr 0.000754 time 0.2479 (0.2842) loss 4.2983 (3.7171) grad_norm 1.2557 (inf) [2021-04-16 00:05:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][960/1251] eta 0:01:22 lr 0.000754 time 0.2641 (0.2841) loss 4.3187 (3.7177) grad_norm 1.3052 (inf) [2021-04-16 00:05:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][970/1251] eta 0:01:19 lr 0.000754 time 0.2635 (0.2839) loss 3.9304 (3.7155) grad_norm 1.2384 (inf) [2021-04-16 00:05:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][980/1251] eta 0:01:16 lr 0.000753 time 0.2761 (0.2838) loss 4.3208 (3.7161) grad_norm 1.2149 (inf) [2021-04-16 00:05:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][990/1251] eta 0:01:14 lr 0.000753 time 0.2745 (0.2838) loss 4.3199 (3.7171) grad_norm 1.3348 (inf) [2021-04-16 00:06:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][1000/1251] eta 0:01:11 lr 0.000753 time 0.2448 (0.2836) loss 2.9490 (3.7166) grad_norm 1.3309 (inf) [2021-04-16 00:06:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][1010/1251] eta 0:01:08 lr 0.000753 time 0.2586 (0.2836) loss 3.7466 (3.7155) grad_norm 1.1222 (inf) [2021-04-16 00:06:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][1020/1251] eta 0:01:05 lr 0.000753 time 0.2623 (0.2835) loss 4.0945 (3.7202) grad_norm 1.2431 (inf) [2021-04-16 00:06:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][1030/1251] eta 0:01:02 lr 0.000753 time 0.2860 (0.2834) loss 3.0035 (3.7204) grad_norm 1.2379 (inf) [2021-04-16 00:06:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][1040/1251] eta 0:00:59 lr 0.000753 time 0.2688 (0.2833) loss 4.5790 (3.7202) grad_norm 1.1068 (inf) [2021-04-16 00:06:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][1050/1251] eta 0:00:56 lr 0.000753 time 0.2817 (0.2832) loss 3.9707 (3.7187) grad_norm 1.1679 (inf) [2021-04-16 00:06:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][1060/1251] eta 0:00:54 lr 0.000753 time 0.2900 (0.2832) loss 3.7358 (3.7169) grad_norm 1.1461 (inf) [2021-04-16 00:06:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][1070/1251] eta 0:00:51 lr 0.000753 time 0.2738 (0.2833) loss 4.1737 (3.7173) grad_norm 1.2487 (inf) [2021-04-16 00:06:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][1080/1251] eta 0:00:48 lr 0.000753 time 0.2612 (0.2832) loss 2.6285 (3.7156) grad_norm 1.2456 (inf) [2021-04-16 00:06:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][1090/1251] eta 0:00:45 lr 0.000753 time 0.3063 (0.2832) loss 2.4256 (3.7140) grad_norm 1.2063 (inf) [2021-04-16 00:06:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][1100/1251] eta 0:00:42 lr 0.000753 time 0.2884 (0.2831) loss 2.9664 (3.7158) grad_norm 1.4675 (inf) [2021-04-16 00:06:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][1110/1251] eta 0:00:39 lr 0.000753 time 0.2864 (0.2831) loss 4.0877 (3.7161) grad_norm 1.0529 (inf) [2021-04-16 00:06:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][1120/1251] eta 0:00:37 lr 0.000753 time 0.3055 (0.2830) loss 1.9900 (3.7151) grad_norm 1.2822 (inf) [2021-04-16 00:06:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][1130/1251] eta 0:00:34 lr 0.000753 time 0.2663 (0.2829) loss 4.2466 (3.7143) grad_norm 1.4061 (inf) [2021-04-16 00:06:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][1140/1251] eta 0:00:31 lr 0.000753 time 0.2680 (0.2829) loss 4.0303 (3.7149) grad_norm 1.1767 (inf) [2021-04-16 00:06:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][1150/1251] eta 0:00:28 lr 0.000753 time 0.2694 (0.2830) loss 3.5852 (3.7143) grad_norm 1.2055 (inf) [2021-04-16 00:06:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][1160/1251] eta 0:00:25 lr 0.000753 time 0.3022 (0.2831) loss 3.9166 (3.7144) grad_norm 1.3645 (inf) [2021-04-16 00:06:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][1170/1251] eta 0:00:22 lr 0.000753 time 0.2993 (0.2831) loss 3.6169 (3.7172) grad_norm 1.2807 (inf) [2021-04-16 00:06:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][1180/1251] eta 0:00:20 lr 0.000753 time 0.2654 (0.2831) loss 4.2245 (3.7178) grad_norm 1.3049 (inf) [2021-04-16 00:06:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][1190/1251] eta 0:00:17 lr 0.000753 time 0.2631 (0.2830) loss 4.0896 (3.7156) grad_norm 1.2906 (inf) [2021-04-16 00:06:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][1200/1251] eta 0:00:14 lr 0.000753 time 0.2691 (0.2830) loss 4.4170 (3.7149) grad_norm 1.3850 (inf) [2021-04-16 00:07:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][1210/1251] eta 0:00:11 lr 0.000753 time 0.2670 (0.2829) loss 4.4788 (3.7154) grad_norm 1.2236 (inf) [2021-04-16 00:07:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][1220/1251] eta 0:00:08 lr 0.000753 time 0.2746 (0.2828) loss 4.1060 (3.7149) grad_norm 1.2314 (inf) [2021-04-16 00:07:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][1230/1251] eta 0:00:05 lr 0.000753 time 0.2905 (0.2828) loss 3.1790 (3.7148) grad_norm 1.2648 (inf) [2021-04-16 00:07:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][1240/1251] eta 0:00:03 lr 0.000753 time 0.2481 (0.2827) loss 4.3356 (3.7166) grad_norm 1.3099 (inf) [2021-04-16 00:07:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [99/300][1250/1251] eta 0:00:00 lr 0.000753 time 0.2534 (0.2824) loss 4.0804 (3.7175) grad_norm 1.4505 (inf) [2021-04-16 00:07:13 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 99 training takes 0:05:56 [2021-04-16 00:07:13 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_99.pth saving...... [2021-04-16 00:07:26 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_99.pth saved !!! [2021-04-16 00:07:27 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.158 (1.158) Loss 1.2757 (1.2757) Acc@1 70.312 (70.312) Acc@5 90.137 (90.137) [2021-04-16 00:07:28 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.107 (0.238) Loss 1.1779 (1.1505) Acc@1 73.730 (73.393) Acc@5 91.699 (91.966) [2021-04-16 00:07:30 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.205 (0.228) Loss 1.0633 (1.1455) Acc@1 75.098 (73.796) Acc@5 92.871 (91.964) [2021-04-16 00:07:33 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.094 (0.232) Loss 1.1568 (1.1380) Acc@1 73.535 (73.938) Acc@5 91.895 (92.137) [2021-04-16 00:07:35 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.675 (0.220) Loss 1.1612 (1.1414) Acc@1 73.438 (73.850) Acc@5 92.773 (92.116) [2021-04-16 00:07:39 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 73.844 Acc@5 92.034 [2021-04-16 00:07:39 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 73.8% [2021-04-16 00:07:39 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 73.84% [2021-04-16 00:07:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][0/1251] eta 1:07:31 lr 0.000753 time 3.2385 (3.2385) loss 2.5087 (2.5087) grad_norm 1.0800 (1.0800) [2021-04-16 00:07:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][10/1251] eta 0:11:26 lr 0.000752 time 0.3001 (0.5529) loss 4.1796 (3.4785) grad_norm 1.3848 (1.2537) [2021-04-16 00:07:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][20/1251] eta 0:08:40 lr 0.000752 time 0.2804 (0.4229) loss 3.3659 (3.6507) grad_norm 1.4489 (1.2849) [2021-04-16 00:07:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][30/1251] eta 0:07:39 lr 0.000752 time 0.2846 (0.3767) loss 3.4468 (3.6905) grad_norm 1.1059 (1.2809) [2021-04-16 00:07:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][40/1251] eta 0:07:10 lr 0.000752 time 0.2598 (0.3553) loss 4.2016 (3.6956) grad_norm 1.0917 (1.2676) [2021-04-16 00:07:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][50/1251] eta 0:06:48 lr 0.000752 time 0.2629 (0.3405) loss 4.1449 (3.6968) grad_norm 1.2711 (1.2659) [2021-04-16 00:08:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][60/1251] eta 0:06:38 lr 0.000752 time 0.4326 (0.3344) loss 4.5220 (3.7090) grad_norm 1.7156 (1.2879) [2021-04-16 00:08:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][70/1251] eta 0:06:24 lr 0.000752 time 0.2763 (0.3259) loss 4.2953 (3.7107) grad_norm 1.2571 (1.2833) [2021-04-16 00:08:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][80/1251] eta 0:06:16 lr 0.000752 time 0.2871 (0.3214) loss 3.9234 (3.7186) grad_norm 1.2588 (1.2743) [2021-04-16 00:08:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][90/1251] eta 0:06:08 lr 0.000752 time 0.2923 (0.3177) loss 3.4128 (3.7177) grad_norm 1.2849 (1.2683) [2021-04-16 00:08:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][100/1251] eta 0:06:01 lr 0.000752 time 0.2755 (0.3137) loss 2.5442 (3.7019) grad_norm 1.6427 (1.2713) [2021-04-16 00:08:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][110/1251] eta 0:05:53 lr 0.000752 time 0.2661 (0.3103) loss 3.9155 (3.7166) grad_norm 1.3648 (1.2702) [2021-04-16 00:08:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][120/1251] eta 0:05:47 lr 0.000752 time 0.2973 (0.3073) loss 3.6560 (3.7101) grad_norm 1.2350 (1.2657) [2021-04-16 00:08:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][130/1251] eta 0:05:41 lr 0.000752 time 0.2749 (0.3050) loss 3.4684 (3.7075) grad_norm 1.2372 (1.2643) [2021-04-16 00:08:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][140/1251] eta 0:05:36 lr 0.000752 time 0.2884 (0.3029) loss 2.6935 (3.6763) grad_norm 1.4166 (1.2728) [2021-04-16 00:08:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][150/1251] eta 0:05:32 lr 0.000752 time 0.2664 (0.3017) loss 3.7341 (3.6703) grad_norm 1.2287 (1.2732) [2021-04-16 00:08:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][160/1251] eta 0:05:28 lr 0.000752 time 0.2967 (0.3011) loss 4.1875 (3.6665) grad_norm 1.3951 (1.2735) [2021-04-16 00:08:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][170/1251] eta 0:05:23 lr 0.000752 time 0.2579 (0.2996) loss 4.2955 (3.6754) grad_norm 1.0306 (1.2700) [2021-04-16 00:08:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][180/1251] eta 0:05:19 lr 0.000752 time 0.2905 (0.2984) loss 3.9287 (3.6754) grad_norm 1.2438 (1.2663) [2021-04-16 00:08:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][190/1251] eta 0:05:15 lr 0.000752 time 0.2878 (0.2971) loss 4.2691 (3.6668) grad_norm 1.2528 (1.2650) [2021-04-16 00:08:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][200/1251] eta 0:05:10 lr 0.000752 time 0.2605 (0.2959) loss 4.0352 (3.6773) grad_norm 1.2449 (1.2652) [2021-04-16 00:08:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][210/1251] eta 0:05:07 lr 0.000752 time 0.2559 (0.2949) loss 3.7442 (3.6645) grad_norm 1.0334 (1.2625) [2021-04-16 00:08:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][220/1251] eta 0:05:02 lr 0.000752 time 0.2625 (0.2939) loss 2.3856 (3.6647) grad_norm 1.1832 (1.2617) [2021-04-16 00:08:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][230/1251] eta 0:04:59 lr 0.000752 time 0.2675 (0.2931) loss 3.2980 (3.6634) grad_norm 1.2627 (1.2586) [2021-04-16 00:08:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][240/1251] eta 0:04:55 lr 0.000752 time 0.2605 (0.2927) loss 3.9387 (3.6641) grad_norm 1.0831 (1.2646) [2021-04-16 00:08:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][250/1251] eta 0:04:52 lr 0.000752 time 0.2813 (0.2920) loss 3.7632 (3.6603) grad_norm 1.2520 (1.2670) [2021-04-16 00:08:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][260/1251] eta 0:04:48 lr 0.000752 time 0.2762 (0.2912) loss 3.8911 (3.6661) grad_norm 1.1559 (1.2679) [2021-04-16 00:08:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][270/1251] eta 0:04:45 lr 0.000752 time 0.2900 (0.2906) loss 3.1516 (3.6713) grad_norm 1.2677 (1.2671) [2021-04-16 00:09:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][280/1251] eta 0:04:41 lr 0.000751 time 0.2734 (0.2900) loss 3.6691 (3.6820) grad_norm 1.1008 (1.2651) [2021-04-16 00:09:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][290/1251] eta 0:04:38 lr 0.000751 time 0.2993 (0.2900) loss 3.1330 (3.6862) grad_norm 1.1768 (1.2671) [2021-04-16 00:09:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][300/1251] eta 0:04:35 lr 0.000751 time 0.2632 (0.2895) loss 4.2335 (3.6943) grad_norm 1.3330 (1.2699) [2021-04-16 00:09:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][310/1251] eta 0:04:32 lr 0.000751 time 0.2723 (0.2892) loss 4.1977 (3.6904) grad_norm 1.2561 (1.2703) [2021-04-16 00:09:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][320/1251] eta 0:04:28 lr 0.000751 time 0.2600 (0.2887) loss 3.6464 (3.6947) grad_norm 1.1441 (1.2698) [2021-04-16 00:09:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][330/1251] eta 0:04:25 lr 0.000751 time 0.2880 (0.2884) loss 4.5843 (3.6994) grad_norm 1.1399 (1.2691) [2021-04-16 00:09:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][340/1251] eta 0:04:22 lr 0.000751 time 0.2868 (0.2887) loss 3.3076 (3.7014) grad_norm 1.6676 (1.2702) [2021-04-16 00:09:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][350/1251] eta 0:04:19 lr 0.000751 time 0.2636 (0.2882) loss 4.4215 (3.6994) grad_norm 1.4526 (1.2721) [2021-04-16 00:09:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][360/1251] eta 0:04:17 lr 0.000751 time 0.2689 (0.2887) loss 2.4976 (3.6969) grad_norm 1.1037 (1.2710) [2021-04-16 00:09:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][370/1251] eta 0:04:14 lr 0.000751 time 0.2722 (0.2886) loss 4.2708 (3.6948) grad_norm 1.2434 (1.2691) [2021-04-16 00:09:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][380/1251] eta 0:04:11 lr 0.000751 time 0.2716 (0.2884) loss 4.2499 (3.7003) grad_norm 1.3176 (1.2668) [2021-04-16 00:09:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][390/1251] eta 0:04:08 lr 0.000751 time 0.2981 (0.2881) loss 4.1218 (3.7005) grad_norm 1.1422 (1.2649) [2021-04-16 00:09:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][400/1251] eta 0:04:04 lr 0.000751 time 0.2772 (0.2878) loss 3.7102 (3.6970) grad_norm 1.3198 (1.2657) [2021-04-16 00:09:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][410/1251] eta 0:04:01 lr 0.000751 time 0.2892 (0.2874) loss 3.4724 (3.7001) grad_norm 1.1900 (1.2636) [2021-04-16 00:09:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][420/1251] eta 0:03:58 lr 0.000751 time 0.2748 (0.2872) loss 3.2754 (3.7053) grad_norm 1.1907 (1.2649) [2021-04-16 00:09:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][430/1251] eta 0:03:55 lr 0.000751 time 0.2611 (0.2870) loss 3.3207 (3.7004) grad_norm 1.2240 (1.2651) [2021-04-16 00:09:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][440/1251] eta 0:03:52 lr 0.000751 time 0.2733 (0.2868) loss 3.5838 (3.6982) grad_norm 1.4306 (1.2664) [2021-04-16 00:09:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][450/1251] eta 0:03:49 lr 0.000751 time 0.2511 (0.2867) loss 3.8840 (3.7008) grad_norm 1.2888 (1.2676) [2021-04-16 00:09:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][460/1251] eta 0:03:46 lr 0.000751 time 0.2647 (0.2864) loss 2.9314 (3.6947) grad_norm 1.2863 (1.2674) [2021-04-16 00:09:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][470/1251] eta 0:03:43 lr 0.000751 time 0.2949 (0.2861) loss 2.7240 (3.6925) grad_norm 1.1926 (1.2670) [2021-04-16 00:09:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][480/1251] eta 0:03:40 lr 0.000751 time 0.2803 (0.2859) loss 3.7581 (3.6866) grad_norm 1.4468 (1.2669) [2021-04-16 00:09:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][490/1251] eta 0:03:37 lr 0.000751 time 0.2491 (0.2856) loss 3.8901 (3.6887) grad_norm 1.0956 (1.2648) [2021-04-16 00:10:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][500/1251] eta 0:03:34 lr 0.000751 time 0.2840 (0.2855) loss 3.7201 (3.6883) grad_norm 1.2101 (1.2635) [2021-04-16 00:10:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][510/1251] eta 0:03:31 lr 0.000751 time 0.2722 (0.2852) loss 4.5509 (3.6891) grad_norm 1.1074 (1.2625) [2021-04-16 00:10:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][520/1251] eta 0:03:28 lr 0.000751 time 0.2601 (0.2851) loss 4.2411 (3.6909) grad_norm 1.1893 (1.2624) [2021-04-16 00:10:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][530/1251] eta 0:03:25 lr 0.000751 time 0.2817 (0.2849) loss 2.7313 (3.6884) grad_norm 1.3242 (1.2626) [2021-04-16 00:10:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][540/1251] eta 0:03:22 lr 0.000751 time 0.2662 (0.2848) loss 2.7456 (3.6867) grad_norm 1.2643 (1.2626) [2021-04-16 00:10:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][550/1251] eta 0:03:19 lr 0.000751 time 0.2461 (0.2846) loss 4.0702 (3.6875) grad_norm 1.1690 (1.2618) [2021-04-16 00:10:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][560/1251] eta 0:03:16 lr 0.000750 time 0.2841 (0.2845) loss 3.4950 (3.6867) grad_norm 1.2430 (1.2607) [2021-04-16 00:10:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][570/1251] eta 0:03:13 lr 0.000750 time 0.2742 (0.2844) loss 4.3900 (3.6905) grad_norm 1.4480 (1.2621) [2021-04-16 00:10:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][580/1251] eta 0:03:10 lr 0.000750 time 0.2883 (0.2846) loss 4.7258 (3.6974) grad_norm 1.1708 (1.2611) [2021-04-16 00:10:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][590/1251] eta 0:03:08 lr 0.000750 time 0.2702 (0.2844) loss 4.1456 (3.6974) grad_norm 1.7202 (1.2619) [2021-04-16 00:10:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][600/1251] eta 0:03:05 lr 0.000750 time 0.2978 (0.2843) loss 3.0681 (3.6924) grad_norm 1.3070 (1.2624) [2021-04-16 00:10:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][610/1251] eta 0:03:02 lr 0.000750 time 0.2721 (0.2841) loss 2.8248 (3.6924) grad_norm 1.3997 (1.2625) [2021-04-16 00:10:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][620/1251] eta 0:02:59 lr 0.000750 time 0.2627 (0.2840) loss 4.6375 (3.6976) grad_norm 1.4762 (1.2636) [2021-04-16 00:10:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][630/1251] eta 0:02:56 lr 0.000750 time 0.2967 (0.2840) loss 3.6075 (3.6966) grad_norm 1.1852 (1.2625) [2021-04-16 00:10:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][640/1251] eta 0:02:53 lr 0.000750 time 0.2609 (0.2840) loss 3.7708 (3.7016) grad_norm 1.5304 (1.2632) [2021-04-16 00:10:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][650/1251] eta 0:02:50 lr 0.000750 time 0.2781 (0.2841) loss 3.5119 (3.6981) grad_norm 1.3711 (1.2629) [2021-04-16 00:10:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][660/1251] eta 0:02:47 lr 0.000750 time 0.2768 (0.2839) loss 3.8265 (3.6986) grad_norm 1.1307 (1.2631) [2021-04-16 00:10:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][670/1251] eta 0:02:44 lr 0.000750 time 0.2516 (0.2839) loss 3.1554 (3.6974) grad_norm 1.1728 (1.2633) [2021-04-16 00:10:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][680/1251] eta 0:02:42 lr 0.000750 time 0.2569 (0.2838) loss 4.2124 (3.6986) grad_norm 1.2006 (1.2635) [2021-04-16 00:10:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][690/1251] eta 0:02:39 lr 0.000750 time 0.3003 (0.2836) loss 3.9834 (3.7033) grad_norm 1.1634 (1.2638) [2021-04-16 00:10:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][700/1251] eta 0:02:36 lr 0.000750 time 0.2726 (0.2835) loss 4.1163 (3.7005) grad_norm 1.2197 (1.2641) [2021-04-16 00:11:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][710/1251] eta 0:02:33 lr 0.000750 time 0.2747 (0.2835) loss 3.8894 (3.7009) grad_norm 1.6285 (1.2650) [2021-04-16 00:11:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][720/1251] eta 0:02:30 lr 0.000750 time 0.2682 (0.2836) loss 3.9250 (3.7004) grad_norm 1.1560 (1.2659) [2021-04-16 00:11:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][730/1251] eta 0:02:27 lr 0.000750 time 0.2581 (0.2834) loss 3.3008 (3.6972) grad_norm 1.3536 (1.2667) [2021-04-16 00:11:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][740/1251] eta 0:02:24 lr 0.000750 time 0.2670 (0.2835) loss 4.1114 (3.6984) grad_norm 1.1592 (1.2665) [2021-04-16 00:11:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][750/1251] eta 0:02:22 lr 0.000750 time 0.2575 (0.2834) loss 4.1637 (3.7001) grad_norm 1.7475 (1.2678) [2021-04-16 00:11:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][760/1251] eta 0:02:19 lr 0.000750 time 0.2678 (0.2833) loss 4.0251 (3.7018) grad_norm 1.2526 (1.2674) [2021-04-16 00:11:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][770/1251] eta 0:02:16 lr 0.000750 time 0.2749 (0.2832) loss 2.7341 (3.6979) grad_norm 1.3511 (1.2677) [2021-04-16 00:11:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][780/1251] eta 0:02:13 lr 0.000750 time 0.2744 (0.2831) loss 3.1582 (3.6983) grad_norm 1.2036 (1.2674) [2021-04-16 00:11:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][790/1251] eta 0:02:10 lr 0.000750 time 0.2814 (0.2830) loss 4.5765 (3.6976) grad_norm 1.1230 (1.2666) [2021-04-16 00:11:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][800/1251] eta 0:02:07 lr 0.000750 time 0.2723 (0.2830) loss 3.8066 (3.6981) grad_norm 1.1227 (1.2667) [2021-04-16 00:11:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][810/1251] eta 0:02:04 lr 0.000750 time 0.2594 (0.2830) loss 4.4619 (3.6955) grad_norm 1.2425 (1.2672) [2021-04-16 00:11:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][820/1251] eta 0:02:01 lr 0.000750 time 0.3094 (0.2829) loss 2.9292 (3.6960) grad_norm 1.1746 (1.2667) [2021-04-16 00:11:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][830/1251] eta 0:01:59 lr 0.000750 time 0.2855 (0.2828) loss 3.7674 (3.6975) grad_norm 1.3909 (1.2666) [2021-04-16 00:11:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][840/1251] eta 0:01:56 lr 0.000749 time 0.2760 (0.2827) loss 3.6396 (3.6983) grad_norm 1.2756 (1.2667) [2021-04-16 00:11:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][850/1251] eta 0:01:53 lr 0.000749 time 0.2737 (0.2827) loss 3.7752 (3.6942) grad_norm 1.5403 (1.2669) [2021-04-16 00:11:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][860/1251] eta 0:01:50 lr 0.000749 time 0.2772 (0.2827) loss 4.0334 (3.6937) grad_norm 1.2220 (1.2666) [2021-04-16 00:11:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][870/1251] eta 0:01:47 lr 0.000749 time 0.2618 (0.2827) loss 3.2082 (3.6926) grad_norm 1.4017 (1.2654) [2021-04-16 00:11:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][880/1251] eta 0:01:44 lr 0.000749 time 0.2818 (0.2826) loss 3.3624 (3.6921) grad_norm 1.2035 (1.2651) [2021-04-16 00:11:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][890/1251] eta 0:01:41 lr 0.000749 time 0.2667 (0.2825) loss 4.2549 (3.6950) grad_norm 1.2883 (1.2656) [2021-04-16 00:11:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][900/1251] eta 0:01:39 lr 0.000749 time 0.2471 (0.2824) loss 3.4122 (3.6970) grad_norm 1.2718 (1.2652) [2021-04-16 00:11:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][910/1251] eta 0:01:36 lr 0.000749 time 0.2751 (0.2824) loss 2.8917 (3.6971) grad_norm 1.2017 (1.2652) [2021-04-16 00:11:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][920/1251] eta 0:01:33 lr 0.000749 time 0.2562 (0.2825) loss 3.9390 (3.6990) grad_norm 1.4294 (1.2650) [2021-04-16 00:12:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][930/1251] eta 0:01:30 lr 0.000749 time 0.2801 (0.2825) loss 2.6988 (3.6992) grad_norm 1.3892 (1.2656) [2021-04-16 00:12:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][940/1251] eta 0:01:27 lr 0.000749 time 0.2724 (0.2827) loss 3.9731 (3.6998) grad_norm 1.5853 (1.2662) [2021-04-16 00:12:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][950/1251] eta 0:01:25 lr 0.000749 time 0.2773 (0.2827) loss 3.5228 (3.6984) grad_norm 1.8106 (1.2665) [2021-04-16 00:12:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][960/1251] eta 0:01:22 lr 0.000749 time 0.2631 (0.2826) loss 3.0896 (3.6958) grad_norm 1.4796 (1.2665) [2021-04-16 00:12:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][970/1251] eta 0:01:19 lr 0.000749 time 0.2683 (0.2826) loss 3.7876 (3.6974) grad_norm 1.2008 (1.2667) [2021-04-16 00:12:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][980/1251] eta 0:01:16 lr 0.000749 time 0.2506 (0.2825) loss 2.6400 (3.6983) grad_norm 1.1843 (1.2667) [2021-04-16 00:12:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][990/1251] eta 0:01:13 lr 0.000749 time 0.3140 (0.2824) loss 3.9900 (3.7008) grad_norm 1.0436 (1.2665) [2021-04-16 00:12:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][1000/1251] eta 0:01:10 lr 0.000749 time 0.2640 (0.2824) loss 3.1768 (3.6971) grad_norm 1.2601 (1.2668) [2021-04-16 00:12:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][1010/1251] eta 0:01:08 lr 0.000749 time 0.2873 (0.2823) loss 4.0332 (3.6977) grad_norm 1.4464 (1.2679) [2021-04-16 00:12:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][1020/1251] eta 0:01:05 lr 0.000749 time 0.2887 (0.2822) loss 3.8006 (3.6992) grad_norm 1.1870 (1.2677) [2021-04-16 00:12:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][1030/1251] eta 0:01:02 lr 0.000749 time 0.2832 (0.2821) loss 4.4683 (3.6995) grad_norm 1.1956 (1.2685) [2021-04-16 00:12:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][1040/1251] eta 0:00:59 lr 0.000749 time 0.2832 (0.2821) loss 3.8312 (3.6983) grad_norm 1.2151 (1.2677) [2021-04-16 00:12:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][1050/1251] eta 0:00:56 lr 0.000749 time 0.2742 (0.2820) loss 4.1570 (3.6988) grad_norm 1.4336 (1.2674) [2021-04-16 00:12:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][1060/1251] eta 0:00:53 lr 0.000749 time 0.2627 (0.2819) loss 3.2305 (3.7013) grad_norm 1.1527 (1.2676) [2021-04-16 00:12:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][1070/1251] eta 0:00:51 lr 0.000749 time 0.2921 (0.2819) loss 2.4712 (3.7002) grad_norm 1.1133 (1.2679) [2021-04-16 00:12:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][1080/1251] eta 0:00:48 lr 0.000749 time 0.2512 (0.2818) loss 3.9667 (3.7003) grad_norm 1.2714 (1.2680) [2021-04-16 00:12:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][1090/1251] eta 0:00:45 lr 0.000749 time 0.3056 (0.2817) loss 4.2724 (3.6994) grad_norm 1.3345 (1.2685) [2021-04-16 00:12:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][1100/1251] eta 0:00:42 lr 0.000749 time 0.2750 (0.2817) loss 4.0998 (3.6985) grad_norm 1.0605 (1.2683) [2021-04-16 00:12:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][1110/1251] eta 0:00:39 lr 0.000749 time 0.2663 (0.2816) loss 4.3028 (3.6990) grad_norm 1.2078 (1.2678) [2021-04-16 00:12:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][1120/1251] eta 0:00:36 lr 0.000748 time 0.2799 (0.2816) loss 4.3196 (3.7015) grad_norm 1.4010 (1.2682) [2021-04-16 00:12:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][1130/1251] eta 0:00:34 lr 0.000748 time 0.2953 (0.2816) loss 3.8518 (3.7032) grad_norm 1.4257 (1.2688) [2021-04-16 00:13:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][1140/1251] eta 0:00:31 lr 0.000748 time 0.2524 (0.2815) loss 3.0556 (3.7023) grad_norm 1.2071 (1.2688) [2021-04-16 00:13:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][1150/1251] eta 0:00:28 lr 0.000748 time 0.2694 (0.2815) loss 4.4230 (3.7047) grad_norm 1.2257 (1.2683) [2021-04-16 00:13:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][1160/1251] eta 0:00:25 lr 0.000748 time 0.2697 (0.2816) loss 3.7723 (3.7024) grad_norm 1.4240 (1.2681) [2021-04-16 00:13:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][1170/1251] eta 0:00:22 lr 0.000748 time 0.2910 (0.2815) loss 3.5583 (3.7027) grad_norm 1.4700 (1.2679) [2021-04-16 00:13:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][1180/1251] eta 0:00:19 lr 0.000748 time 0.2792 (0.2815) loss 2.7787 (3.7019) grad_norm 1.2127 (1.2675) [2021-04-16 00:13:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][1190/1251] eta 0:00:17 lr 0.000748 time 0.3914 (0.2815) loss 4.0756 (3.7037) grad_norm 1.3582 (1.2676) [2021-04-16 00:13:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][1200/1251] eta 0:00:14 lr 0.000748 time 0.2672 (0.2815) loss 4.0570 (3.7026) grad_norm 1.4142 (1.2672) [2021-04-16 00:13:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][1210/1251] eta 0:00:11 lr 0.000748 time 0.2650 (0.2814) loss 2.7299 (3.7022) grad_norm 1.2208 (1.2667) [2021-04-16 00:13:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][1220/1251] eta 0:00:08 lr 0.000748 time 0.2840 (0.2815) loss 3.6097 (3.7037) grad_norm 1.3012 (1.2674) [2021-04-16 00:13:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][1230/1251] eta 0:00:05 lr 0.000748 time 0.2820 (0.2815) loss 4.2648 (3.7023) grad_norm 1.2971 (1.2671) [2021-04-16 00:13:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][1240/1251] eta 0:00:03 lr 0.000748 time 0.2483 (0.2814) loss 4.1171 (3.7044) grad_norm 1.1259 (1.2674) [2021-04-16 00:13:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [100/300][1250/1251] eta 0:00:00 lr 0.000748 time 0.2484 (0.2812) loss 4.0679 (3.7044) grad_norm 1.2834 (1.2675) [2021-04-16 00:13:33 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 100 training takes 0:05:54 [2021-04-16 00:13:33 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_100.pth saving...... [2021-04-16 00:13:45 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_100.pth saved !!! [2021-04-16 00:13:46 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.227 (1.227) Loss 1.0977 (1.0977) Acc@1 75.293 (75.293) Acc@5 92.383 (92.383) [2021-04-16 00:13:48 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.425 (0.248) Loss 1.1134 (1.1513) Acc@1 74.414 (73.349) Acc@5 91.992 (91.850) [2021-04-16 00:13:50 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.132 (0.253) Loss 1.0954 (1.1370) Acc@1 74.512 (73.605) Acc@5 92.480 (92.094) [2021-04-16 00:13:52 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.148 (0.220) Loss 1.1540 (1.1366) Acc@1 72.949 (73.611) Acc@5 91.797 (92.039) [2021-04-16 00:13:54 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.215) Loss 1.0450 (1.1328) Acc@1 75.781 (73.764) Acc@5 92.871 (92.083) [2021-04-16 00:13:57 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 73.686 Acc@5 92.034 [2021-04-16 00:13:57 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 73.7% [2021-04-16 00:13:57 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 73.84% [2021-04-16 00:14:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][0/1251] eta 1:49:34 lr 0.000748 time 5.2552 (5.2552) loss 4.1285 (4.1285) grad_norm 1.1078 (1.1078) [2021-04-16 00:14:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][10/1251] eta 0:15:00 lr 0.000748 time 0.2529 (0.7258) loss 3.4672 (3.8009) grad_norm 1.0202 (1.2774) [2021-04-16 00:14:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][20/1251] eta 0:10:27 lr 0.000748 time 0.2712 (0.5095) loss 3.0735 (3.7574) grad_norm 1.3902 (inf) [2021-04-16 00:14:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][30/1251] eta 0:08:51 lr 0.000748 time 0.2564 (0.4353) loss 4.2148 (3.8356) grad_norm 1.2142 (inf) [2021-04-16 00:14:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][40/1251] eta 0:08:03 lr 0.000748 time 0.2837 (0.3989) loss 4.5100 (3.7990) grad_norm 1.3629 (inf) [2021-04-16 00:14:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][50/1251] eta 0:07:29 lr 0.000748 time 0.2588 (0.3744) loss 2.7656 (3.8026) grad_norm 1.2414 (inf) [2021-04-16 00:14:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][60/1251] eta 0:07:06 lr 0.000748 time 0.2972 (0.3579) loss 3.6811 (3.7332) grad_norm 1.1153 (inf) [2021-04-16 00:14:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][70/1251] eta 0:06:50 lr 0.000748 time 0.3091 (0.3474) loss 3.8460 (3.7256) grad_norm 1.1629 (inf) [2021-04-16 00:14:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][80/1251] eta 0:06:38 lr 0.000748 time 0.3004 (0.3400) loss 3.6062 (3.7215) grad_norm 1.6217 (inf) [2021-04-16 00:14:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][90/1251] eta 0:06:27 lr 0.000748 time 0.2760 (0.3334) loss 4.4236 (3.7411) grad_norm 1.3078 (inf) [2021-04-16 00:14:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][100/1251] eta 0:06:17 lr 0.000748 time 0.2891 (0.3279) loss 4.3776 (3.7417) grad_norm 1.2533 (inf) [2021-04-16 00:14:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][110/1251] eta 0:06:09 lr 0.000748 time 0.2957 (0.3240) loss 4.0785 (3.7531) grad_norm 1.2011 (inf) [2021-04-16 00:14:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][120/1251] eta 0:06:03 lr 0.000748 time 0.2599 (0.3213) loss 3.6238 (3.7436) grad_norm 1.0999 (inf) [2021-04-16 00:14:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][130/1251] eta 0:05:57 lr 0.000748 time 0.3091 (0.3185) loss 3.7321 (3.7410) grad_norm 1.2664 (inf) [2021-04-16 00:14:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][140/1251] eta 0:05:52 lr 0.000747 time 0.2932 (0.3169) loss 3.5992 (3.7465) grad_norm 1.4536 (inf) [2021-04-16 00:14:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][150/1251] eta 0:05:47 lr 0.000747 time 0.3142 (0.3157) loss 3.0781 (3.7197) grad_norm 1.3260 (inf) [2021-04-16 00:14:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][160/1251] eta 0:05:41 lr 0.000747 time 0.2809 (0.3132) loss 3.7054 (3.7105) grad_norm 1.2260 (inf) [2021-04-16 00:14:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][170/1251] eta 0:05:36 lr 0.000747 time 0.2775 (0.3111) loss 4.5652 (3.7131) grad_norm 1.2710 (inf) [2021-04-16 00:14:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][180/1251] eta 0:05:31 lr 0.000747 time 0.2867 (0.3094) loss 4.4418 (3.7300) grad_norm 1.2853 (inf) [2021-04-16 00:14:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][190/1251] eta 0:05:26 lr 0.000747 time 0.2942 (0.3077) loss 3.5634 (3.7296) grad_norm 1.6496 (inf) [2021-04-16 00:14:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][200/1251] eta 0:05:22 lr 0.000747 time 0.3006 (0.3064) loss 4.0545 (3.7316) grad_norm 1.2379 (inf) [2021-04-16 00:15:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][210/1251] eta 0:05:17 lr 0.000747 time 0.2900 (0.3051) loss 3.6947 (3.7345) grad_norm 1.4313 (inf) [2021-04-16 00:15:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][220/1251] eta 0:05:13 lr 0.000747 time 0.2673 (0.3037) loss 3.4168 (3.7347) grad_norm 1.2910 (inf) [2021-04-16 00:15:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][230/1251] eta 0:05:08 lr 0.000747 time 0.2675 (0.3024) loss 4.1650 (3.7397) grad_norm 1.3281 (inf) [2021-04-16 00:15:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][240/1251] eta 0:05:04 lr 0.000747 time 0.2940 (0.3013) loss 4.5927 (3.7451) grad_norm 1.1600 (inf) [2021-04-16 00:15:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][250/1251] eta 0:05:00 lr 0.000747 time 0.3088 (0.3006) loss 3.9824 (3.7484) grad_norm 1.2451 (inf) [2021-04-16 00:15:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][260/1251] eta 0:04:56 lr 0.000747 time 0.2674 (0.2997) loss 4.2579 (3.7573) grad_norm 1.0927 (inf) [2021-04-16 00:15:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][270/1251] eta 0:04:53 lr 0.000747 time 0.2939 (0.2989) loss 4.3457 (3.7588) grad_norm 1.2259 (inf) [2021-04-16 00:15:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][280/1251] eta 0:04:49 lr 0.000747 time 0.2867 (0.2984) loss 3.7918 (3.7654) grad_norm 1.1225 (inf) [2021-04-16 00:15:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][290/1251] eta 0:04:46 lr 0.000747 time 0.3033 (0.2977) loss 4.2107 (3.7707) grad_norm 1.4475 (inf) [2021-04-16 00:15:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][300/1251] eta 0:04:42 lr 0.000747 time 0.3061 (0.2973) loss 3.8226 (3.7716) grad_norm 1.2560 (inf) [2021-04-16 00:15:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][310/1251] eta 0:04:39 lr 0.000747 time 0.2944 (0.2968) loss 3.2856 (3.7676) grad_norm 1.1932 (inf) [2021-04-16 00:15:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][320/1251] eta 0:04:36 lr 0.000747 time 0.2631 (0.2966) loss 2.5948 (3.7624) grad_norm 1.3047 (inf) [2021-04-16 00:15:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][330/1251] eta 0:04:32 lr 0.000747 time 0.2713 (0.2960) loss 3.2511 (3.7529) grad_norm 1.1434 (inf) [2021-04-16 00:15:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][340/1251] eta 0:04:29 lr 0.000747 time 0.2496 (0.2953) loss 3.7714 (3.7538) grad_norm 1.4196 (inf) [2021-04-16 00:15:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][350/1251] eta 0:04:25 lr 0.000747 time 0.2883 (0.2949) loss 3.7771 (3.7538) grad_norm 1.2806 (inf) [2021-04-16 00:15:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][360/1251] eta 0:04:22 lr 0.000747 time 0.2770 (0.2946) loss 4.1240 (3.7468) grad_norm 1.0797 (inf) [2021-04-16 00:15:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][370/1251] eta 0:04:18 lr 0.000747 time 0.2566 (0.2940) loss 2.8923 (3.7487) grad_norm 1.2358 (inf) [2021-04-16 00:15:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][380/1251] eta 0:04:15 lr 0.000747 time 0.2643 (0.2937) loss 3.3834 (3.7450) grad_norm 1.2447 (inf) [2021-04-16 00:15:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][390/1251] eta 0:04:13 lr 0.000747 time 0.2861 (0.2941) loss 3.1497 (3.7458) grad_norm 1.2666 (inf) [2021-04-16 00:15:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][400/1251] eta 0:04:09 lr 0.000747 time 0.2597 (0.2937) loss 3.9786 (3.7429) grad_norm 1.2441 (inf) [2021-04-16 00:15:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][410/1251] eta 0:04:06 lr 0.000747 time 0.2624 (0.2933) loss 3.3415 (3.7476) grad_norm 1.4556 (inf) [2021-04-16 00:16:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][420/1251] eta 0:04:03 lr 0.000746 time 0.2653 (0.2932) loss 2.7229 (3.7396) grad_norm 1.1818 (inf) [2021-04-16 00:16:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][430/1251] eta 0:04:00 lr 0.000746 time 0.2559 (0.2931) loss 3.2104 (3.7376) grad_norm 1.1198 (inf) [2021-04-16 00:16:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][440/1251] eta 0:03:57 lr 0.000746 time 0.2771 (0.2930) loss 4.1041 (3.7351) grad_norm 1.2085 (inf) [2021-04-16 00:16:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][450/1251] eta 0:03:54 lr 0.000746 time 0.2757 (0.2927) loss 3.9009 (3.7345) grad_norm 1.3110 (inf) [2021-04-16 00:16:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][460/1251] eta 0:03:51 lr 0.000746 time 0.2887 (0.2923) loss 3.5386 (3.7282) grad_norm 1.1826 (inf) [2021-04-16 00:16:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][470/1251] eta 0:03:48 lr 0.000746 time 0.2836 (0.2921) loss 4.0927 (3.7248) grad_norm 1.3738 (inf) [2021-04-16 00:16:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][480/1251] eta 0:03:44 lr 0.000746 time 0.2775 (0.2917) loss 3.6256 (3.7243) grad_norm 1.2358 (inf) [2021-04-16 00:16:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][490/1251] eta 0:03:41 lr 0.000746 time 0.2856 (0.2914) loss 4.1374 (3.7230) grad_norm 1.4335 (inf) [2021-04-16 00:16:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][500/1251] eta 0:03:38 lr 0.000746 time 0.3015 (0.2911) loss 3.6628 (3.7217) grad_norm 1.1207 (inf) [2021-04-16 00:16:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][510/1251] eta 0:03:35 lr 0.000746 time 0.3014 (0.2910) loss 4.0340 (3.7240) grad_norm 1.2195 (inf) [2021-04-16 00:16:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][520/1251] eta 0:03:32 lr 0.000746 time 0.2624 (0.2906) loss 2.9983 (3.7245) grad_norm 1.2576 (inf) [2021-04-16 00:16:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][530/1251] eta 0:03:29 lr 0.000746 time 0.2671 (0.2903) loss 3.6460 (3.7243) grad_norm 1.0767 (inf) [2021-04-16 00:16:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][540/1251] eta 0:03:26 lr 0.000746 time 0.2887 (0.2900) loss 4.0363 (3.7233) grad_norm 1.2018 (inf) [2021-04-16 00:16:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][550/1251] eta 0:03:23 lr 0.000746 time 0.2912 (0.2899) loss 3.0999 (3.7205) grad_norm 1.2199 (inf) [2021-04-16 00:16:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][560/1251] eta 0:03:20 lr 0.000746 time 0.2641 (0.2897) loss 3.9462 (3.7213) grad_norm 1.2325 (inf) [2021-04-16 00:16:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][570/1251] eta 0:03:17 lr 0.000746 time 0.2645 (0.2897) loss 3.7218 (3.7252) grad_norm 1.2625 (inf) [2021-04-16 00:16:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][580/1251] eta 0:03:14 lr 0.000746 time 0.2836 (0.2896) loss 3.8577 (3.7225) grad_norm 1.7060 (inf) [2021-04-16 00:16:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][590/1251] eta 0:03:11 lr 0.000746 time 0.2843 (0.2894) loss 4.3113 (3.7182) grad_norm 1.2674 (inf) [2021-04-16 00:16:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][600/1251] eta 0:03:08 lr 0.000746 time 0.2659 (0.2892) loss 4.2103 (3.7168) grad_norm 1.4410 (inf) [2021-04-16 00:16:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][610/1251] eta 0:03:05 lr 0.000746 time 0.2762 (0.2889) loss 4.2088 (3.7209) grad_norm 1.3300 (inf) [2021-04-16 00:16:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][620/1251] eta 0:03:02 lr 0.000746 time 0.2674 (0.2887) loss 4.2051 (3.7224) grad_norm 1.1848 (inf) [2021-04-16 00:16:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][630/1251] eta 0:02:59 lr 0.000746 time 0.2686 (0.2884) loss 3.8183 (3.7231) grad_norm 1.1533 (inf) [2021-04-16 00:17:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][640/1251] eta 0:02:56 lr 0.000746 time 0.2668 (0.2882) loss 3.3727 (3.7282) grad_norm 1.2011 (inf) [2021-04-16 00:17:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][650/1251] eta 0:02:53 lr 0.000746 time 0.2711 (0.2883) loss 4.1075 (3.7289) grad_norm 1.0877 (inf) [2021-04-16 00:17:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][660/1251] eta 0:02:50 lr 0.000746 time 0.2814 (0.2881) loss 4.4934 (3.7278) grad_norm 1.2681 (inf) [2021-04-16 00:17:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][670/1251] eta 0:02:47 lr 0.000746 time 0.2952 (0.2880) loss 3.7218 (3.7286) grad_norm 1.3765 (inf) [2021-04-16 00:17:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][680/1251] eta 0:02:44 lr 0.000746 time 0.2872 (0.2879) loss 3.1174 (3.7273) grad_norm 1.2226 (inf) [2021-04-16 00:17:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][690/1251] eta 0:02:41 lr 0.000746 time 0.2622 (0.2878) loss 2.3783 (3.7264) grad_norm 1.2720 (inf) [2021-04-16 00:17:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][700/1251] eta 0:02:38 lr 0.000745 time 0.2776 (0.2877) loss 2.9520 (3.7243) grad_norm 1.4757 (inf) [2021-04-16 00:17:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][710/1251] eta 0:02:35 lr 0.000745 time 0.2805 (0.2877) loss 3.5735 (3.7199) grad_norm 1.2579 (inf) [2021-04-16 00:17:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][720/1251] eta 0:02:32 lr 0.000745 time 0.2424 (0.2879) loss 3.8466 (3.7242) grad_norm 1.1432 (inf) [2021-04-16 00:17:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][730/1251] eta 0:02:30 lr 0.000745 time 0.2823 (0.2880) loss 2.5801 (3.7234) grad_norm 1.3062 (inf) [2021-04-16 00:17:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][740/1251] eta 0:02:27 lr 0.000745 time 0.2769 (0.2878) loss 4.1012 (3.7230) grad_norm 1.1839 (inf) [2021-04-16 00:17:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][750/1251] eta 0:02:24 lr 0.000745 time 0.3288 (0.2878) loss 2.5448 (3.7224) grad_norm 1.2888 (inf) [2021-04-16 00:17:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][760/1251] eta 0:02:21 lr 0.000745 time 0.2742 (0.2876) loss 3.9264 (3.7190) grad_norm 1.4113 (inf) [2021-04-16 00:17:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][770/1251] eta 0:02:18 lr 0.000745 time 0.2955 (0.2875) loss 4.3473 (3.7204) grad_norm 1.3739 (inf) [2021-04-16 00:17:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][780/1251] eta 0:02:15 lr 0.000745 time 0.2899 (0.2875) loss 2.8505 (3.7182) grad_norm 1.0801 (inf) [2021-04-16 00:17:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][790/1251] eta 0:02:12 lr 0.000745 time 0.2645 (0.2874) loss 3.9574 (3.7183) grad_norm 1.3629 (inf) [2021-04-16 00:17:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][800/1251] eta 0:02:09 lr 0.000745 time 0.3133 (0.2872) loss 4.2865 (3.7203) grad_norm 1.3425 (inf) [2021-04-16 00:17:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][810/1251] eta 0:02:06 lr 0.000745 time 0.2703 (0.2871) loss 3.2281 (3.7227) grad_norm 1.0855 (inf) [2021-04-16 00:17:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][820/1251] eta 0:02:03 lr 0.000745 time 0.2968 (0.2870) loss 4.5924 (3.7267) grad_norm 1.2479 (inf) [2021-04-16 00:17:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][830/1251] eta 0:02:00 lr 0.000745 time 0.2902 (0.2868) loss 3.6244 (3.7263) grad_norm 1.1900 (inf) [2021-04-16 00:17:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][840/1251] eta 0:01:57 lr 0.000745 time 0.2755 (0.2867) loss 4.6908 (3.7261) grad_norm 1.3327 (inf) [2021-04-16 00:18:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][850/1251] eta 0:01:54 lr 0.000745 time 0.2754 (0.2865) loss 4.1239 (3.7256) grad_norm 1.1783 (inf) [2021-04-16 00:18:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][860/1251] eta 0:01:51 lr 0.000745 time 0.3036 (0.2864) loss 3.0300 (3.7274) grad_norm 1.3140 (inf) [2021-04-16 00:18:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][870/1251] eta 0:01:49 lr 0.000745 time 0.2783 (0.2863) loss 2.9075 (3.7256) grad_norm 1.1920 (inf) [2021-04-16 00:18:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][880/1251] eta 0:01:46 lr 0.000745 time 0.2859 (0.2862) loss 3.2313 (3.7271) grad_norm 1.4527 (inf) [2021-04-16 00:18:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][890/1251] eta 0:01:43 lr 0.000745 time 0.3039 (0.2863) loss 4.3163 (3.7250) grad_norm 1.3161 (inf) [2021-04-16 00:18:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][900/1251] eta 0:01:40 lr 0.000745 time 0.2925 (0.2861) loss 4.5819 (3.7241) grad_norm 1.2068 (inf) [2021-04-16 00:18:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][910/1251] eta 0:01:37 lr 0.000745 time 0.2775 (0.2860) loss 4.4134 (3.7260) grad_norm 1.2616 (inf) [2021-04-16 00:18:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][920/1251] eta 0:01:34 lr 0.000745 time 0.2678 (0.2858) loss 4.1477 (3.7258) grad_norm 1.3135 (inf) [2021-04-16 00:18:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][930/1251] eta 0:01:31 lr 0.000745 time 0.2818 (0.2858) loss 3.8501 (3.7270) grad_norm 1.1736 (inf) [2021-04-16 00:18:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][940/1251] eta 0:01:28 lr 0.000745 time 0.2843 (0.2858) loss 3.3495 (3.7265) grad_norm 1.0953 (inf) [2021-04-16 00:18:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][950/1251] eta 0:01:26 lr 0.000745 time 0.2856 (0.2858) loss 4.0769 (3.7260) grad_norm 1.2263 (inf) [2021-04-16 00:18:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][960/1251] eta 0:01:23 lr 0.000745 time 0.2769 (0.2856) loss 3.7908 (3.7267) grad_norm 1.4641 (inf) [2021-04-16 00:18:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][970/1251] eta 0:01:20 lr 0.000744 time 0.2835 (0.2857) loss 3.5254 (3.7273) grad_norm 1.3764 (inf) [2021-04-16 00:18:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][980/1251] eta 0:01:17 lr 0.000744 time 0.2409 (0.2856) loss 3.7259 (3.7249) grad_norm 1.1184 (inf) [2021-04-16 00:18:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][990/1251] eta 0:01:14 lr 0.000744 time 0.2872 (0.2856) loss 3.0561 (3.7220) grad_norm 1.1311 (inf) [2021-04-16 00:18:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][1000/1251] eta 0:01:11 lr 0.000744 time 0.3206 (0.2856) loss 3.5234 (3.7174) grad_norm 1.3139 (inf) [2021-04-16 00:18:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][1010/1251] eta 0:01:08 lr 0.000744 time 0.2694 (0.2857) loss 3.2748 (3.7145) grad_norm 1.3937 (inf) [2021-04-16 00:18:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][1020/1251] eta 0:01:06 lr 0.000744 time 0.2978 (0.2858) loss 2.8486 (3.7126) grad_norm 1.1378 (inf) [2021-04-16 00:18:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][1030/1251] eta 0:01:03 lr 0.000744 time 0.2827 (0.2857) loss 4.0641 (3.7117) grad_norm 1.4070 (inf) [2021-04-16 00:18:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][1040/1251] eta 0:01:00 lr 0.000744 time 0.2894 (0.2856) loss 3.9915 (3.7111) grad_norm 1.2692 (inf) [2021-04-16 00:18:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][1050/1251] eta 0:00:57 lr 0.000744 time 0.2568 (0.2855) loss 3.2231 (3.7113) grad_norm 1.3384 (inf) [2021-04-16 00:19:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][1060/1251] eta 0:00:54 lr 0.000744 time 0.2976 (0.2855) loss 4.5280 (3.7121) grad_norm 1.1463 (inf) [2021-04-16 00:19:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][1070/1251] eta 0:00:51 lr 0.000744 time 0.2919 (0.2854) loss 3.4039 (3.7128) grad_norm 1.3243 (inf) [2021-04-16 00:19:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][1080/1251] eta 0:00:48 lr 0.000744 time 0.2688 (0.2853) loss 4.2233 (3.7124) grad_norm 1.5493 (inf) [2021-04-16 00:19:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][1090/1251] eta 0:00:45 lr 0.000744 time 0.2586 (0.2852) loss 3.4826 (3.7111) grad_norm 1.4901 (inf) [2021-04-16 00:19:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][1100/1251] eta 0:00:43 lr 0.000744 time 0.2652 (0.2851) loss 3.7953 (3.7143) grad_norm 1.2963 (inf) [2021-04-16 00:19:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][1110/1251] eta 0:00:40 lr 0.000744 time 0.2691 (0.2850) loss 3.6222 (3.7131) grad_norm 1.1842 (inf) [2021-04-16 00:19:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][1120/1251] eta 0:00:37 lr 0.000744 time 0.2829 (0.2850) loss 4.1899 (3.7161) grad_norm 1.3227 (inf) [2021-04-16 00:19:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][1130/1251] eta 0:00:34 lr 0.000744 time 0.2670 (0.2849) loss 4.2621 (3.7162) grad_norm 1.3035 (inf) [2021-04-16 00:19:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][1140/1251] eta 0:00:31 lr 0.000744 time 0.2706 (0.2848) loss 3.3053 (3.7165) grad_norm 1.2660 (inf) [2021-04-16 00:19:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][1150/1251] eta 0:00:28 lr 0.000744 time 0.2729 (0.2848) loss 4.4659 (3.7186) grad_norm 1.1932 (inf) [2021-04-16 00:19:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][1160/1251] eta 0:00:25 lr 0.000744 time 0.2664 (0.2848) loss 3.9545 (3.7174) grad_norm 1.0966 (inf) [2021-04-16 00:19:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][1170/1251] eta 0:00:23 lr 0.000744 time 0.2844 (0.2848) loss 4.2835 (3.7165) grad_norm 1.2674 (inf) [2021-04-16 00:19:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][1180/1251] eta 0:00:20 lr 0.000744 time 0.2937 (0.2848) loss 3.6679 (3.7118) grad_norm 1.0510 (inf) [2021-04-16 00:19:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][1190/1251] eta 0:00:17 lr 0.000744 time 0.2968 (0.2847) loss 3.9559 (3.7126) grad_norm 1.1299 (inf) [2021-04-16 00:19:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][1200/1251] eta 0:00:14 lr 0.000744 time 0.2531 (0.2846) loss 4.4347 (3.7105) grad_norm 1.2414 (inf) [2021-04-16 00:19:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][1210/1251] eta 0:00:11 lr 0.000744 time 0.2883 (0.2846) loss 3.7788 (3.7090) grad_norm 1.5322 (inf) [2021-04-16 00:19:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][1220/1251] eta 0:00:08 lr 0.000744 time 0.3043 (0.2845) loss 3.7553 (3.7109) grad_norm 1.4322 (inf) [2021-04-16 00:19:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][1230/1251] eta 0:00:05 lr 0.000744 time 0.2676 (0.2846) loss 4.1270 (3.7125) grad_norm 1.2805 (inf) [2021-04-16 00:19:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][1240/1251] eta 0:00:03 lr 0.000744 time 0.2526 (0.2845) loss 4.1867 (3.7140) grad_norm 1.3126 (inf) [2021-04-16 00:19:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [101/300][1250/1251] eta 0:00:00 lr 0.000743 time 0.2484 (0.2842) loss 4.0864 (3.7129) grad_norm 1.3866 (inf) [2021-04-16 00:19:56 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 101 training takes 0:05:58 [2021-04-16 00:19:56 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_101.pth saving...... [2021-04-16 00:20:10 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_101.pth saved !!! [2021-04-16 00:20:11 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.163 (1.163) Loss 1.1374 (1.1374) Acc@1 73.926 (73.926) Acc@5 91.699 (91.699) [2021-04-16 00:20:13 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.344 (0.272) Loss 1.0827 (1.1521) Acc@1 73.535 (73.242) Acc@5 92.773 (91.912) [2021-04-16 00:20:14 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.096 (0.210) Loss 1.1144 (1.1419) Acc@1 72.461 (73.517) Acc@5 92.871 (91.964) [2021-04-16 00:20:17 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.304 (0.227) Loss 1.2512 (1.1495) Acc@1 70.215 (73.362) Acc@5 89.746 (91.888) [2021-04-16 00:20:19 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.222) Loss 1.1462 (1.1430) Acc@1 71.973 (73.466) Acc@5 92.285 (91.978) [2021-04-16 00:20:22 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 73.552 Acc@5 91.966 [2021-04-16 00:20:22 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 73.6% [2021-04-16 00:20:22 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 73.84% [2021-04-16 00:20:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][0/1251] eta 1:16:34 lr 0.000743 time 3.6726 (3.6726) loss 3.6518 (3.6518) grad_norm 1.1243 (1.1243) [2021-04-16 00:20:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][10/1251] eta 0:12:07 lr 0.000743 time 0.2944 (0.5865) loss 2.8480 (3.7165) grad_norm 1.3460 (1.2734) [2021-04-16 00:20:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][20/1251] eta 0:08:59 lr 0.000743 time 0.2651 (0.4382) loss 3.2440 (3.6319) grad_norm 1.2509 (1.2365) [2021-04-16 00:20:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][30/1251] eta 0:07:50 lr 0.000743 time 0.2687 (0.3851) loss 3.3486 (3.6354) grad_norm 1.3014 (1.2376) [2021-04-16 00:20:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][40/1251] eta 0:07:15 lr 0.000743 time 0.2959 (0.3593) loss 3.6442 (3.6267) grad_norm 1.2022 (1.2380) [2021-04-16 00:20:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][50/1251] eta 0:06:52 lr 0.000743 time 0.2910 (0.3432) loss 3.9127 (3.6290) grad_norm 1.5403 (1.2489) [2021-04-16 00:20:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][60/1251] eta 0:06:36 lr 0.000743 time 0.2919 (0.3326) loss 3.7732 (3.6539) grad_norm 1.2207 (1.2619) [2021-04-16 00:20:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][70/1251] eta 0:06:23 lr 0.000743 time 0.2887 (0.3245) loss 4.4801 (3.6920) grad_norm 1.2195 (1.2651) [2021-04-16 00:20:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][80/1251] eta 0:06:13 lr 0.000743 time 0.2664 (0.3186) loss 4.0545 (3.6967) grad_norm 1.3260 (1.2804) [2021-04-16 00:20:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][90/1251] eta 0:06:04 lr 0.000743 time 0.2878 (0.3142) loss 2.8189 (3.6801) grad_norm 1.2241 (1.2940) [2021-04-16 00:20:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][100/1251] eta 0:05:59 lr 0.000743 time 0.2413 (0.3123) loss 3.9940 (3.7076) grad_norm 1.1645 (1.2875) [2021-04-16 00:20:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][110/1251] eta 0:05:52 lr 0.000743 time 0.2755 (0.3092) loss 2.9200 (3.6850) grad_norm 1.2431 (1.2853) [2021-04-16 00:20:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][120/1251] eta 0:05:46 lr 0.000743 time 0.2863 (0.3065) loss 2.9851 (3.6513) grad_norm 1.1411 (1.2774) [2021-04-16 00:21:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][130/1251] eta 0:05:41 lr 0.000743 time 0.2658 (0.3043) loss 2.3049 (3.6306) grad_norm 1.2734 (1.2786) [2021-04-16 00:21:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][140/1251] eta 0:05:36 lr 0.000743 time 0.2757 (0.3030) loss 3.7766 (3.6339) grad_norm 1.1794 (1.2790) [2021-04-16 00:21:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][150/1251] eta 0:05:31 lr 0.000743 time 0.2695 (0.3013) loss 4.1494 (3.6222) grad_norm 1.3684 (1.2778) [2021-04-16 00:21:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][160/1251] eta 0:05:28 lr 0.000743 time 0.2756 (0.3007) loss 3.5332 (3.6339) grad_norm 1.1679 (1.2758) [2021-04-16 00:21:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][170/1251] eta 0:05:23 lr 0.000743 time 0.2728 (0.2991) loss 2.7881 (3.6280) grad_norm 1.1794 (1.2733) [2021-04-16 00:21:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][180/1251] eta 0:05:19 lr 0.000743 time 0.2713 (0.2986) loss 3.3630 (3.6200) grad_norm 1.1572 (1.2701) [2021-04-16 00:21:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][190/1251] eta 0:05:15 lr 0.000743 time 0.2798 (0.2975) loss 3.3248 (3.6121) grad_norm 1.3994 (1.2688) [2021-04-16 00:21:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][200/1251] eta 0:05:12 lr 0.000743 time 0.2590 (0.2969) loss 2.9638 (3.6016) grad_norm 1.6136 (1.2709) [2021-04-16 00:21:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][210/1251] eta 0:05:07 lr 0.000743 time 0.2865 (0.2958) loss 3.0110 (3.6030) grad_norm 1.1040 (1.2720) [2021-04-16 00:21:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][220/1251] eta 0:05:04 lr 0.000743 time 0.2732 (0.2955) loss 4.4358 (3.6075) grad_norm 1.5172 (1.2777) [2021-04-16 00:21:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][230/1251] eta 0:05:01 lr 0.000743 time 0.2638 (0.2950) loss 2.5638 (3.6082) grad_norm 1.6092 (1.2804) [2021-04-16 00:21:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][240/1251] eta 0:04:57 lr 0.000743 time 0.2771 (0.2942) loss 3.6578 (3.6106) grad_norm 1.1837 (1.2801) [2021-04-16 00:21:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][250/1251] eta 0:04:53 lr 0.000743 time 0.2998 (0.2936) loss 3.1961 (3.6157) grad_norm 1.1876 (1.2831) [2021-04-16 00:21:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][260/1251] eta 0:04:50 lr 0.000743 time 0.2658 (0.2927) loss 3.6710 (3.6032) grad_norm 1.3737 (inf) [2021-04-16 00:21:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][270/1251] eta 0:04:46 lr 0.000742 time 0.2678 (0.2922) loss 4.1857 (3.6053) grad_norm 1.1057 (inf) [2021-04-16 00:21:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][280/1251] eta 0:04:43 lr 0.000742 time 0.2662 (0.2917) loss 4.2814 (3.6129) grad_norm 1.3311 (inf) [2021-04-16 00:21:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][290/1251] eta 0:04:39 lr 0.000742 time 0.2704 (0.2911) loss 4.2092 (3.6239) grad_norm 1.1557 (inf) [2021-04-16 00:21:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][300/1251] eta 0:04:36 lr 0.000742 time 0.2755 (0.2908) loss 3.3190 (3.6226) grad_norm 1.4559 (inf) [2021-04-16 00:21:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][310/1251] eta 0:04:33 lr 0.000742 time 0.2901 (0.2906) loss 3.0454 (3.6191) grad_norm 1.0828 (inf) [2021-04-16 00:21:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][320/1251] eta 0:04:30 lr 0.000742 time 0.2689 (0.2902) loss 4.2849 (3.6253) grad_norm 1.1307 (inf) [2021-04-16 00:21:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][330/1251] eta 0:04:26 lr 0.000742 time 0.2640 (0.2897) loss 4.4869 (3.6228) grad_norm 1.5542 (inf) [2021-04-16 00:22:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][340/1251] eta 0:04:23 lr 0.000742 time 0.2664 (0.2895) loss 4.3886 (3.6247) grad_norm 1.4980 (inf) [2021-04-16 00:22:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][350/1251] eta 0:04:20 lr 0.000742 time 0.2844 (0.2890) loss 3.0995 (3.6238) grad_norm 1.5245 (inf) [2021-04-16 00:22:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][360/1251] eta 0:04:17 lr 0.000742 time 0.2857 (0.2894) loss 3.8721 (3.6301) grad_norm 1.2507 (inf) [2021-04-16 00:22:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][370/1251] eta 0:04:14 lr 0.000742 time 0.2797 (0.2891) loss 3.4051 (3.6344) grad_norm 1.2687 (inf) [2021-04-16 00:22:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][380/1251] eta 0:04:11 lr 0.000742 time 0.2679 (0.2888) loss 4.3511 (3.6432) grad_norm 1.0958 (inf) [2021-04-16 00:22:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][390/1251] eta 0:04:08 lr 0.000742 time 0.3083 (0.2885) loss 3.8849 (3.6479) grad_norm 1.1130 (inf) [2021-04-16 00:22:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][400/1251] eta 0:04:05 lr 0.000742 time 0.2699 (0.2880) loss 2.5882 (3.6386) grad_norm 1.4547 (inf) [2021-04-16 00:22:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][410/1251] eta 0:04:02 lr 0.000742 time 0.2803 (0.2878) loss 4.0329 (3.6353) grad_norm 1.1203 (inf) [2021-04-16 00:22:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][420/1251] eta 0:03:58 lr 0.000742 time 0.2768 (0.2876) loss 3.8088 (3.6370) grad_norm 1.4906 (inf) [2021-04-16 00:22:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][430/1251] eta 0:03:56 lr 0.000742 time 0.2700 (0.2876) loss 2.8671 (3.6390) grad_norm 1.4567 (inf) [2021-04-16 00:22:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][440/1251] eta 0:03:53 lr 0.000742 time 0.2668 (0.2873) loss 4.2528 (3.6393) grad_norm 1.5873 (inf) [2021-04-16 00:22:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][450/1251] eta 0:03:49 lr 0.000742 time 0.2731 (0.2871) loss 3.9075 (3.6468) grad_norm 1.2686 (inf) [2021-04-16 00:22:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][460/1251] eta 0:03:46 lr 0.000742 time 0.2774 (0.2868) loss 3.4694 (3.6440) grad_norm 1.1659 (inf) [2021-04-16 00:22:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][470/1251] eta 0:03:43 lr 0.000742 time 0.2691 (0.2865) loss 4.4903 (3.6464) grad_norm 1.3529 (inf) [2021-04-16 00:22:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][480/1251] eta 0:03:40 lr 0.000742 time 0.2741 (0.2863) loss 3.9512 (3.6541) grad_norm 1.1155 (inf) [2021-04-16 00:22:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][490/1251] eta 0:03:37 lr 0.000742 time 0.2873 (0.2863) loss 2.8593 (3.6549) grad_norm 1.2537 (inf) [2021-04-16 00:22:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][500/1251] eta 0:03:34 lr 0.000742 time 0.2730 (0.2860) loss 4.1365 (3.6578) grad_norm 1.0714 (inf) [2021-04-16 00:22:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][510/1251] eta 0:03:31 lr 0.000742 time 0.2613 (0.2858) loss 4.2007 (3.6618) grad_norm 1.4627 (inf) [2021-04-16 00:22:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][520/1251] eta 0:03:28 lr 0.000742 time 0.2787 (0.2857) loss 4.6122 (3.6611) grad_norm 1.6209 (inf) [2021-04-16 00:22:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][530/1251] eta 0:03:25 lr 0.000742 time 0.2920 (0.2855) loss 3.7282 (3.6612) grad_norm 1.1099 (inf) [2021-04-16 00:22:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][540/1251] eta 0:03:22 lr 0.000742 time 0.2734 (0.2853) loss 3.5222 (3.6616) grad_norm 1.6379 (inf) [2021-04-16 00:22:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][550/1251] eta 0:03:19 lr 0.000741 time 0.2688 (0.2851) loss 3.2907 (3.6568) grad_norm 1.2068 (inf) [2021-04-16 00:23:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][560/1251] eta 0:03:16 lr 0.000741 time 0.2711 (0.2849) loss 3.8798 (3.6593) grad_norm 1.5536 (inf) [2021-04-16 00:23:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][570/1251] eta 0:03:14 lr 0.000741 time 0.2946 (0.2853) loss 4.2019 (3.6613) grad_norm 1.4370 (inf) [2021-04-16 00:23:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][580/1251] eta 0:03:11 lr 0.000741 time 0.2853 (0.2856) loss 3.9032 (3.6672) grad_norm 1.3431 (inf) [2021-04-16 00:23:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][590/1251] eta 0:03:08 lr 0.000741 time 0.2514 (0.2854) loss 3.4175 (3.6682) grad_norm 1.2441 (inf) [2021-04-16 00:23:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][600/1251] eta 0:03:05 lr 0.000741 time 0.2838 (0.2852) loss 3.1667 (3.6674) grad_norm 1.1370 (inf) [2021-04-16 00:23:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][610/1251] eta 0:03:02 lr 0.000741 time 0.2650 (0.2850) loss 3.5141 (3.6656) grad_norm 1.2765 (inf) [2021-04-16 00:23:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][620/1251] eta 0:02:59 lr 0.000741 time 0.2887 (0.2849) loss 2.8366 (3.6663) grad_norm 1.3817 (inf) [2021-04-16 00:23:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][630/1251] eta 0:02:56 lr 0.000741 time 0.2606 (0.2848) loss 3.1843 (3.6685) grad_norm 1.3104 (inf) [2021-04-16 00:23:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][640/1251] eta 0:02:53 lr 0.000741 time 0.2674 (0.2846) loss 3.5883 (3.6700) grad_norm 1.9317 (inf) [2021-04-16 00:23:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][650/1251] eta 0:02:50 lr 0.000741 time 0.2689 (0.2844) loss 3.8685 (3.6748) grad_norm 1.2664 (inf) [2021-04-16 00:23:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][660/1251] eta 0:02:48 lr 0.000741 time 0.2740 (0.2844) loss 3.0032 (3.6738) grad_norm 1.2995 (inf) [2021-04-16 00:23:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][670/1251] eta 0:02:45 lr 0.000741 time 0.2653 (0.2842) loss 3.6479 (3.6725) grad_norm 1.0830 (inf) [2021-04-16 00:23:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][680/1251] eta 0:02:42 lr 0.000741 time 0.2629 (0.2843) loss 3.9987 (3.6713) grad_norm 1.2872 (inf) [2021-04-16 00:23:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][690/1251] eta 0:02:39 lr 0.000741 time 0.2654 (0.2842) loss 2.6832 (3.6688) grad_norm 1.1022 (inf) [2021-04-16 00:23:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][700/1251] eta 0:02:36 lr 0.000741 time 0.2934 (0.2840) loss 2.8895 (3.6700) grad_norm 1.2654 (inf) [2021-04-16 00:23:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][710/1251] eta 0:02:33 lr 0.000741 time 0.2882 (0.2839) loss 4.1048 (3.6719) grad_norm 1.2465 (inf) [2021-04-16 00:23:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][720/1251] eta 0:02:30 lr 0.000741 time 0.2773 (0.2838) loss 3.9297 (3.6729) grad_norm 1.1699 (inf) [2021-04-16 00:23:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][730/1251] eta 0:02:27 lr 0.000741 time 0.2587 (0.2839) loss 3.8518 (3.6755) grad_norm 1.2600 (inf) [2021-04-16 00:23:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][740/1251] eta 0:02:25 lr 0.000741 time 0.2637 (0.2838) loss 2.9276 (3.6730) grad_norm 1.1477 (inf) [2021-04-16 00:23:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][750/1251] eta 0:02:22 lr 0.000741 time 0.2714 (0.2837) loss 3.5813 (3.6712) grad_norm 1.4647 (inf) [2021-04-16 00:23:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][760/1251] eta 0:02:19 lr 0.000741 time 0.2639 (0.2837) loss 3.6568 (3.6730) grad_norm 1.1443 (inf) [2021-04-16 00:24:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][770/1251] eta 0:02:16 lr 0.000741 time 0.2661 (0.2837) loss 3.8203 (3.6745) grad_norm 1.1765 (inf) [2021-04-16 00:24:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][780/1251] eta 0:02:13 lr 0.000741 time 0.2922 (0.2836) loss 3.7969 (3.6773) grad_norm 1.3281 (inf) [2021-04-16 00:24:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][790/1251] eta 0:02:10 lr 0.000741 time 0.2733 (0.2837) loss 2.9263 (3.6726) grad_norm 1.1961 (inf) [2021-04-16 00:24:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][800/1251] eta 0:02:08 lr 0.000741 time 0.4301 (0.2838) loss 3.5566 (3.6745) grad_norm 1.1596 (inf) [2021-04-16 00:24:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][810/1251] eta 0:02:05 lr 0.000741 time 0.2634 (0.2837) loss 3.9979 (3.6707) grad_norm 1.4365 (inf) [2021-04-16 00:24:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][820/1251] eta 0:02:02 lr 0.000740 time 0.2645 (0.2837) loss 2.8985 (3.6688) grad_norm 1.2934 (inf) [2021-04-16 00:24:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][830/1251] eta 0:01:59 lr 0.000740 time 0.2773 (0.2835) loss 3.4604 (3.6689) grad_norm 1.4218 (inf) [2021-04-16 00:24:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][840/1251] eta 0:01:56 lr 0.000740 time 0.2702 (0.2835) loss 4.1384 (3.6692) grad_norm 1.6109 (inf) [2021-04-16 00:24:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][850/1251] eta 0:01:53 lr 0.000740 time 0.2647 (0.2834) loss 2.9948 (3.6689) grad_norm 1.1698 (inf) [2021-04-16 00:24:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][860/1251] eta 0:01:50 lr 0.000740 time 0.2628 (0.2833) loss 4.0691 (3.6688) grad_norm 1.4651 (inf) [2021-04-16 00:24:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][870/1251] eta 0:01:47 lr 0.000740 time 0.2604 (0.2832) loss 3.8901 (3.6701) grad_norm 1.0443 (inf) [2021-04-16 00:24:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][880/1251] eta 0:01:45 lr 0.000740 time 0.2908 (0.2831) loss 2.8239 (3.6686) grad_norm 1.3074 (inf) [2021-04-16 00:24:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][890/1251] eta 0:01:42 lr 0.000740 time 0.2953 (0.2831) loss 4.5907 (3.6659) grad_norm 1.4319 (inf) [2021-04-16 00:24:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][900/1251] eta 0:01:39 lr 0.000740 time 0.2774 (0.2831) loss 3.4150 (3.6668) grad_norm 1.6863 (inf) [2021-04-16 00:24:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][910/1251] eta 0:01:36 lr 0.000740 time 0.2688 (0.2830) loss 3.8539 (3.6686) grad_norm 1.3779 (inf) [2021-04-16 00:24:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][920/1251] eta 0:01:33 lr 0.000740 time 0.2575 (0.2830) loss 2.9373 (3.6663) grad_norm 1.3359 (inf) [2021-04-16 00:24:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][930/1251] eta 0:01:30 lr 0.000740 time 0.2650 (0.2830) loss 3.7453 (3.6647) grad_norm 1.2984 (inf) [2021-04-16 00:24:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][940/1251] eta 0:01:28 lr 0.000740 time 0.2805 (0.2830) loss 4.5342 (3.6670) grad_norm 1.2368 (inf) [2021-04-16 00:24:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][950/1251] eta 0:01:25 lr 0.000740 time 0.2848 (0.2829) loss 3.9633 (3.6698) grad_norm 1.4004 (inf) [2021-04-16 00:24:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][960/1251] eta 0:01:22 lr 0.000740 time 0.2658 (0.2828) loss 2.7657 (3.6718) grad_norm 1.3477 (inf) [2021-04-16 00:24:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][970/1251] eta 0:01:19 lr 0.000740 time 0.2686 (0.2827) loss 2.8419 (3.6689) grad_norm 1.1474 (inf) [2021-04-16 00:25:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][980/1251] eta 0:01:16 lr 0.000740 time 0.2755 (0.2827) loss 3.5371 (3.6705) grad_norm 1.1524 (inf) [2021-04-16 00:25:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][990/1251] eta 0:01:13 lr 0.000740 time 0.2917 (0.2827) loss 3.6657 (3.6711) grad_norm 1.5567 (inf) [2021-04-16 00:25:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][1000/1251] eta 0:01:10 lr 0.000740 time 0.2839 (0.2826) loss 2.9539 (3.6709) grad_norm 1.0973 (inf) [2021-04-16 00:25:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][1010/1251] eta 0:01:08 lr 0.000740 time 0.2686 (0.2826) loss 3.7903 (3.6734) grad_norm 1.0762 (inf) [2021-04-16 00:25:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][1020/1251] eta 0:01:05 lr 0.000740 time 0.2893 (0.2826) loss 3.7325 (3.6756) grad_norm 1.1976 (inf) [2021-04-16 00:25:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][1030/1251] eta 0:01:02 lr 0.000740 time 0.2723 (0.2826) loss 3.9600 (3.6753) grad_norm 1.5529 (inf) [2021-04-16 00:25:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][1040/1251] eta 0:00:59 lr 0.000740 time 0.2762 (0.2827) loss 4.2814 (3.6772) grad_norm 1.3660 (inf) [2021-04-16 00:25:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][1050/1251] eta 0:00:56 lr 0.000740 time 0.2523 (0.2827) loss 3.7130 (3.6783) grad_norm 1.6432 (inf) [2021-04-16 00:25:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][1060/1251] eta 0:00:53 lr 0.000740 time 0.2892 (0.2827) loss 4.2051 (3.6779) grad_norm 1.5613 (inf) [2021-04-16 00:25:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][1070/1251] eta 0:00:51 lr 0.000740 time 0.2846 (0.2827) loss 3.9145 (3.6769) grad_norm 1.2906 (inf) [2021-04-16 00:25:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][1080/1251] eta 0:00:48 lr 0.000740 time 0.3220 (0.2827) loss 3.7665 (3.6771) grad_norm 1.2428 (inf) [2021-04-16 00:25:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][1090/1251] eta 0:00:45 lr 0.000740 time 0.3004 (0.2826) loss 3.6372 (3.6775) grad_norm 1.6335 (inf) [2021-04-16 00:25:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][1100/1251] eta 0:00:42 lr 0.000739 time 0.2844 (0.2825) loss 4.3629 (3.6812) grad_norm 1.2144 (inf) [2021-04-16 00:25:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][1110/1251] eta 0:00:39 lr 0.000739 time 0.2933 (0.2825) loss 4.0666 (3.6812) grad_norm 1.2477 (inf) [2021-04-16 00:25:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][1120/1251] eta 0:00:37 lr 0.000739 time 0.2809 (0.2825) loss 4.6557 (3.6833) grad_norm 1.7914 (inf) [2021-04-16 00:25:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][1130/1251] eta 0:00:34 lr 0.000739 time 0.2690 (0.2824) loss 2.7813 (3.6838) grad_norm 1.5142 (inf) [2021-04-16 00:25:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][1140/1251] eta 0:00:31 lr 0.000739 time 0.2674 (0.2825) loss 4.1719 (3.6828) grad_norm 1.1131 (inf) [2021-04-16 00:25:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][1150/1251] eta 0:00:28 lr 0.000739 time 0.2698 (0.2826) loss 3.8457 (3.6832) grad_norm 1.0793 (inf) [2021-04-16 00:25:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][1160/1251] eta 0:00:25 lr 0.000739 time 0.2666 (0.2826) loss 4.3943 (3.6824) grad_norm 1.1831 (inf) [2021-04-16 00:25:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][1170/1251] eta 0:00:22 lr 0.000739 time 0.2718 (0.2825) loss 3.7165 (3.6848) grad_norm 1.2380 (inf) [2021-04-16 00:25:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][1180/1251] eta 0:00:20 lr 0.000739 time 0.2552 (0.2824) loss 3.7085 (3.6826) grad_norm 1.3162 (inf) [2021-04-16 00:25:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][1190/1251] eta 0:00:17 lr 0.000739 time 0.2833 (0.2824) loss 4.0316 (3.6817) grad_norm 1.3446 (inf) [2021-04-16 00:26:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][1200/1251] eta 0:00:14 lr 0.000739 time 0.2715 (0.2824) loss 3.8172 (3.6825) grad_norm 1.1284 (inf) [2021-04-16 00:26:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][1210/1251] eta 0:00:11 lr 0.000739 time 0.2432 (0.2823) loss 3.9432 (3.6816) grad_norm 1.5185 (inf) [2021-04-16 00:26:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][1220/1251] eta 0:00:08 lr 0.000739 time 0.2787 (0.2823) loss 3.8738 (3.6836) grad_norm 1.6384 (inf) [2021-04-16 00:26:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][1230/1251] eta 0:00:05 lr 0.000739 time 0.2909 (0.2822) loss 3.0644 (3.6828) grad_norm 1.2348 (inf) [2021-04-16 00:26:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][1240/1251] eta 0:00:03 lr 0.000739 time 0.2487 (0.2821) loss 2.4196 (3.6814) grad_norm 1.1859 (inf) [2021-04-16 00:26:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [102/300][1250/1251] eta 0:00:00 lr 0.000739 time 0.2467 (0.2819) loss 3.0873 (3.6821) grad_norm 1.3998 (inf) [2021-04-16 00:26:18 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 102 training takes 0:05:55 [2021-04-16 00:26:18 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_102.pth saving...... [2021-04-16 00:26:29 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_102.pth saved !!! [2021-04-16 00:26:30 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.283 (1.283) Loss 1.0570 (1.0570) Acc@1 75.879 (75.879) Acc@5 93.164 (93.164) [2021-04-16 00:26:32 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.194 (0.223) Loss 1.1936 (1.1232) Acc@1 72.461 (74.272) Acc@5 91.992 (92.108) [2021-04-16 00:26:35 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.115 (0.260) Loss 1.0883 (1.1225) Acc@1 74.609 (74.200) Acc@5 92.578 (92.174) [2021-04-16 00:26:36 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.157 (0.236) Loss 1.1531 (1.1256) Acc@1 73.242 (74.008) Acc@5 90.625 (91.995) [2021-04-16 00:26:38 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.218) Loss 1.0957 (1.1192) Acc@1 73.438 (74.000) Acc@5 92.383 (92.137) [2021-04-16 00:26:41 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 74.164 Acc@5 92.202 [2021-04-16 00:26:41 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 74.2% [2021-04-16 00:26:41 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 74.16% [2021-04-16 00:26:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][0/1251] eta 1:25:36 lr 0.000739 time 4.1059 (4.1059) loss 4.1029 (4.1029) grad_norm 1.2245 (1.2245) [2021-04-16 00:26:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][10/1251] eta 0:13:24 lr 0.000739 time 0.4933 (0.6484) loss 3.9271 (3.3546) grad_norm 1.2477 (1.2920) [2021-04-16 00:26:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][20/1251] eta 0:09:43 lr 0.000739 time 0.2838 (0.4743) loss 3.9661 (3.4869) grad_norm 1.4742 (1.3383) [2021-04-16 00:26:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][30/1251] eta 0:08:23 lr 0.000739 time 0.2755 (0.4121) loss 3.9356 (3.4695) grad_norm 1.0611 (1.3120) [2021-04-16 00:26:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][40/1251] eta 0:07:41 lr 0.000739 time 0.3054 (0.3808) loss 4.6077 (3.6044) grad_norm 1.1945 (1.2910) [2021-04-16 00:27:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][50/1251] eta 0:07:12 lr 0.000739 time 0.2562 (0.3600) loss 2.6010 (3.6593) grad_norm 1.2656 (1.2813) [2021-04-16 00:27:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][60/1251] eta 0:06:53 lr 0.000739 time 0.2972 (0.3469) loss 4.0465 (3.6666) grad_norm 1.2845 (1.2828) [2021-04-16 00:27:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][70/1251] eta 0:06:38 lr 0.000739 time 0.2772 (0.3375) loss 2.6089 (3.6261) grad_norm 1.2996 (1.2778) [2021-04-16 00:27:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][80/1251] eta 0:06:25 lr 0.000739 time 0.2677 (0.3296) loss 4.4389 (3.6050) grad_norm 1.2312 (1.2756) [2021-04-16 00:27:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][90/1251] eta 0:06:15 lr 0.000739 time 0.2688 (0.3237) loss 4.0803 (3.6465) grad_norm 1.4357 (1.2916) [2021-04-16 00:27:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][100/1251] eta 0:06:07 lr 0.000739 time 0.2681 (0.3194) loss 4.3887 (3.6475) grad_norm 1.1663 (1.3000) [2021-04-16 00:27:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][110/1251] eta 0:06:00 lr 0.000739 time 0.2956 (0.3159) loss 3.6653 (3.6630) grad_norm 1.2680 (1.2967) [2021-04-16 00:27:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][120/1251] eta 0:05:54 lr 0.000738 time 0.2909 (0.3132) loss 2.8701 (3.6628) grad_norm 1.4208 (1.3063) [2021-04-16 00:27:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][130/1251] eta 0:05:48 lr 0.000738 time 0.2986 (0.3111) loss 4.1231 (3.6486) grad_norm 1.3925 (1.3139) [2021-04-16 00:27:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][140/1251] eta 0:05:43 lr 0.000738 time 0.2713 (0.3096) loss 4.0716 (3.6517) grad_norm 1.3723 (1.3151) [2021-04-16 00:27:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][150/1251] eta 0:05:38 lr 0.000738 time 0.2682 (0.3076) loss 4.3721 (3.6599) grad_norm 1.1125 (1.3129) [2021-04-16 00:27:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][160/1251] eta 0:05:34 lr 0.000738 time 0.2734 (0.3063) loss 3.4199 (3.6647) grad_norm 1.2930 (1.3118) [2021-04-16 00:27:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][170/1251] eta 0:05:29 lr 0.000738 time 0.2734 (0.3047) loss 3.2909 (3.6502) grad_norm 1.1089 (1.3053) [2021-04-16 00:27:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][180/1251] eta 0:05:24 lr 0.000738 time 0.2735 (0.3033) loss 4.2358 (3.6531) grad_norm 1.6979 (1.3044) [2021-04-16 00:27:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][190/1251] eta 0:05:20 lr 0.000738 time 0.2687 (0.3020) loss 4.1458 (3.6540) grad_norm 1.2122 (1.3032) [2021-04-16 00:27:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][200/1251] eta 0:05:16 lr 0.000738 time 0.2446 (0.3008) loss 4.4734 (3.6616) grad_norm 1.3592 (1.3009) [2021-04-16 00:27:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][210/1251] eta 0:05:12 lr 0.000738 time 0.2836 (0.3000) loss 4.0311 (3.6672) grad_norm 1.2855 (1.3016) [2021-04-16 00:27:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][220/1251] eta 0:05:09 lr 0.000738 time 0.2710 (0.2997) loss 3.6825 (3.6690) grad_norm 1.4118 (1.3000) [2021-04-16 00:27:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][230/1251] eta 0:05:04 lr 0.000738 time 0.2690 (0.2987) loss 2.9176 (3.6720) grad_norm 1.5194 (1.3011) [2021-04-16 00:27:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][240/1251] eta 0:05:01 lr 0.000738 time 0.2931 (0.2981) loss 2.8719 (3.6630) grad_norm 1.2200 (1.2990) [2021-04-16 00:27:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][250/1251] eta 0:04:57 lr 0.000738 time 0.2542 (0.2972) loss 3.8802 (3.6680) grad_norm 1.2508 (1.2978) [2021-04-16 00:27:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][260/1251] eta 0:04:53 lr 0.000738 time 0.2743 (0.2964) loss 2.9658 (3.6693) grad_norm 1.2135 (1.3028) [2021-04-16 00:28:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][270/1251] eta 0:04:50 lr 0.000738 time 0.2940 (0.2957) loss 4.0909 (3.6702) grad_norm 1.5856 (1.3044) [2021-04-16 00:28:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][280/1251] eta 0:04:46 lr 0.000738 time 0.2766 (0.2951) loss 2.9025 (3.6737) grad_norm 1.3075 (1.3079) [2021-04-16 00:28:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][290/1251] eta 0:04:43 lr 0.000738 time 0.3042 (0.2950) loss 4.4189 (3.6681) grad_norm 1.2169 (1.3064) [2021-04-16 00:28:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][300/1251] eta 0:04:39 lr 0.000738 time 0.2543 (0.2944) loss 4.3549 (3.6639) grad_norm 1.1354 (1.3038) [2021-04-16 00:28:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][310/1251] eta 0:04:36 lr 0.000738 time 0.2728 (0.2940) loss 3.8730 (3.6641) grad_norm 1.4550 (1.3023) [2021-04-16 00:28:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][320/1251] eta 0:04:33 lr 0.000738 time 0.3157 (0.2935) loss 4.4664 (3.6569) grad_norm 1.2068 (1.3006) [2021-04-16 00:28:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][330/1251] eta 0:04:29 lr 0.000738 time 0.3113 (0.2931) loss 4.6763 (3.6606) grad_norm 1.3734 (1.3000) [2021-04-16 00:28:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][340/1251] eta 0:04:27 lr 0.000738 time 0.3041 (0.2933) loss 4.0774 (3.6576) grad_norm 1.2391 (1.2962) [2021-04-16 00:28:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][350/1251] eta 0:04:23 lr 0.000738 time 0.2852 (0.2927) loss 2.6368 (3.6549) grad_norm 1.1275 (1.2958) [2021-04-16 00:28:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][360/1251] eta 0:04:21 lr 0.000738 time 0.2903 (0.2933) loss 4.3900 (3.6529) grad_norm 1.1239 (1.2964) [2021-04-16 00:28:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][370/1251] eta 0:04:17 lr 0.000738 time 0.2480 (0.2927) loss 3.6388 (3.6598) grad_norm 1.2473 (1.2957) [2021-04-16 00:28:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][380/1251] eta 0:04:14 lr 0.000738 time 0.2961 (0.2922) loss 4.2865 (3.6675) grad_norm 1.2837 (1.2946) [2021-04-16 00:28:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][390/1251] eta 0:04:11 lr 0.000737 time 0.2686 (0.2918) loss 4.4348 (3.6627) grad_norm 1.2837 (1.2936) [2021-04-16 00:28:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][400/1251] eta 0:04:07 lr 0.000737 time 0.2900 (0.2914) loss 3.4948 (3.6561) grad_norm 1.3563 (1.2950) [2021-04-16 00:28:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][410/1251] eta 0:04:04 lr 0.000737 time 0.3094 (0.2911) loss 4.7302 (3.6604) grad_norm 1.4082 (1.2953) [2021-04-16 00:28:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][420/1251] eta 0:04:01 lr 0.000737 time 0.2938 (0.2908) loss 3.9530 (3.6645) grad_norm 1.3952 (1.2959) [2021-04-16 00:28:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][430/1251] eta 0:03:58 lr 0.000737 time 0.2925 (0.2905) loss 3.7370 (3.6661) grad_norm 1.1839 (1.2935) [2021-04-16 00:28:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][440/1251] eta 0:03:55 lr 0.000737 time 0.2579 (0.2902) loss 4.0638 (3.6679) grad_norm 1.1737 (1.2910) [2021-04-16 00:28:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][450/1251] eta 0:03:52 lr 0.000737 time 0.2831 (0.2899) loss 4.1081 (3.6742) grad_norm 1.2032 (1.2921) [2021-04-16 00:28:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][460/1251] eta 0:03:48 lr 0.000737 time 0.2751 (0.2895) loss 3.0948 (3.6672) grad_norm 1.2467 (1.2927) [2021-04-16 00:28:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][470/1251] eta 0:03:46 lr 0.000737 time 0.2951 (0.2896) loss 3.3782 (3.6703) grad_norm 1.2659 (1.2937) [2021-04-16 00:29:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][480/1251] eta 0:03:43 lr 0.000737 time 0.2755 (0.2894) loss 4.0099 (3.6715) grad_norm 1.4119 (1.2924) [2021-04-16 00:29:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][490/1251] eta 0:03:40 lr 0.000737 time 0.2793 (0.2891) loss 3.6779 (3.6733) grad_norm 1.2430 (1.2916) [2021-04-16 00:29:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][500/1251] eta 0:03:36 lr 0.000737 time 0.2698 (0.2888) loss 2.8917 (3.6712) grad_norm 1.2632 (1.2908) [2021-04-16 00:29:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][510/1251] eta 0:03:33 lr 0.000737 time 0.2942 (0.2887) loss 3.1247 (3.6677) grad_norm 1.2749 (1.2914) [2021-04-16 00:29:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][520/1251] eta 0:03:30 lr 0.000737 time 0.3116 (0.2886) loss 3.6010 (3.6663) grad_norm 1.3015 (1.2921) [2021-04-16 00:29:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][530/1251] eta 0:03:27 lr 0.000737 time 0.2725 (0.2884) loss 3.6810 (3.6686) grad_norm 1.4625 (1.2919) [2021-04-16 00:29:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][540/1251] eta 0:03:25 lr 0.000737 time 0.2934 (0.2885) loss 3.5310 (3.6695) grad_norm 1.3601 (1.2918) [2021-04-16 00:29:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][550/1251] eta 0:03:22 lr 0.000737 time 0.4195 (0.2885) loss 2.8279 (3.6714) grad_norm 1.2518 (1.2918) [2021-04-16 00:29:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][560/1251] eta 0:03:19 lr 0.000737 time 0.2759 (0.2885) loss 3.5033 (3.6713) grad_norm 1.1886 (1.2912) [2021-04-16 00:29:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][570/1251] eta 0:03:16 lr 0.000737 time 0.2764 (0.2882) loss 2.9167 (3.6684) grad_norm 1.1281 (1.2892) [2021-04-16 00:29:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][580/1251] eta 0:03:13 lr 0.000737 time 0.2702 (0.2882) loss 3.8946 (3.6738) grad_norm 1.4694 (1.2887) [2021-04-16 00:29:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][590/1251] eta 0:03:10 lr 0.000737 time 0.2764 (0.2885) loss 3.6393 (3.6788) grad_norm 1.2483 (1.2913) [2021-04-16 00:29:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][600/1251] eta 0:03:07 lr 0.000737 time 0.2764 (0.2883) loss 4.1873 (3.6786) grad_norm 1.1351 (1.2907) [2021-04-16 00:29:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][610/1251] eta 0:03:04 lr 0.000737 time 0.2941 (0.2881) loss 3.2251 (3.6774) grad_norm 1.4006 (1.2918) [2021-04-16 00:29:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][620/1251] eta 0:03:01 lr 0.000737 time 0.2891 (0.2881) loss 3.6209 (3.6745) grad_norm 1.5685 (1.2917) [2021-04-16 00:29:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][630/1251] eta 0:02:58 lr 0.000737 time 0.2807 (0.2880) loss 3.1022 (3.6727) grad_norm 1.2247 (1.2915) [2021-04-16 00:29:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][640/1251] eta 0:02:55 lr 0.000737 time 0.2604 (0.2877) loss 3.7310 (3.6723) grad_norm 1.3107 (1.2911) [2021-04-16 00:29:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][650/1251] eta 0:02:52 lr 0.000737 time 0.2967 (0.2876) loss 4.0405 (3.6718) grad_norm 1.1523 (1.2900) [2021-04-16 00:29:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][660/1251] eta 0:02:49 lr 0.000736 time 0.2703 (0.2875) loss 3.2166 (3.6671) grad_norm 1.2471 (1.2911) [2021-04-16 00:29:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][670/1251] eta 0:02:46 lr 0.000736 time 0.2583 (0.2874) loss 3.3718 (3.6692) grad_norm 1.3462 (1.2911) [2021-04-16 00:29:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][680/1251] eta 0:02:43 lr 0.000736 time 0.3017 (0.2872) loss 3.4920 (3.6698) grad_norm 1.2107 (1.2897) [2021-04-16 00:30:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][690/1251] eta 0:02:41 lr 0.000736 time 0.2912 (0.2871) loss 2.1292 (3.6700) grad_norm 1.2176 (1.2889) [2021-04-16 00:30:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][700/1251] eta 0:02:38 lr 0.000736 time 0.2759 (0.2870) loss 4.3518 (3.6732) grad_norm 1.1812 (1.2877) [2021-04-16 00:30:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][710/1251] eta 0:02:35 lr 0.000736 time 0.2893 (0.2869) loss 3.2646 (3.6706) grad_norm 1.2982 (1.2874) [2021-04-16 00:30:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][720/1251] eta 0:02:32 lr 0.000736 time 0.2750 (0.2869) loss 3.4185 (3.6695) grad_norm 1.2279 (1.2867) [2021-04-16 00:30:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][730/1251] eta 0:02:29 lr 0.000736 time 0.4222 (0.2869) loss 4.1227 (3.6724) grad_norm 1.1622 (1.2865) [2021-04-16 00:30:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][740/1251] eta 0:02:26 lr 0.000736 time 0.2772 (0.2867) loss 2.8793 (3.6675) grad_norm 1.5942 (1.2871) [2021-04-16 00:30:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][750/1251] eta 0:02:23 lr 0.000736 time 0.2936 (0.2866) loss 2.6469 (3.6673) grad_norm 1.1232 (1.2872) [2021-04-16 00:30:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][760/1251] eta 0:02:20 lr 0.000736 time 0.2931 (0.2865) loss 2.4904 (3.6680) grad_norm 1.3925 (1.2878) [2021-04-16 00:30:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][770/1251] eta 0:02:17 lr 0.000736 time 0.2654 (0.2864) loss 4.0418 (3.6696) grad_norm 1.0163 (1.2877) [2021-04-16 00:30:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][780/1251] eta 0:02:14 lr 0.000736 time 0.2800 (0.2862) loss 3.7058 (3.6758) grad_norm 1.3638 (1.2878) [2021-04-16 00:30:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][790/1251] eta 0:02:11 lr 0.000736 time 0.2613 (0.2863) loss 4.4626 (3.6752) grad_norm 1.5282 (1.2870) [2021-04-16 00:30:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][800/1251] eta 0:02:09 lr 0.000736 time 0.3257 (0.2862) loss 2.7898 (3.6749) grad_norm 1.5618 (1.2873) [2021-04-16 00:30:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][810/1251] eta 0:02:06 lr 0.000736 time 0.2853 (0.2861) loss 3.7123 (3.6773) grad_norm 1.2565 (1.2870) [2021-04-16 00:30:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][820/1251] eta 0:02:03 lr 0.000736 time 0.2517 (0.2860) loss 4.1927 (3.6795) grad_norm 1.2134 (1.2868) [2021-04-16 00:30:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][830/1251] eta 0:02:00 lr 0.000736 time 0.3040 (0.2860) loss 4.4634 (3.6804) grad_norm 1.2312 (1.2862) [2021-04-16 00:30:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][840/1251] eta 0:01:57 lr 0.000736 time 0.2587 (0.2859) loss 2.6424 (3.6792) grad_norm 1.0548 (1.2850) [2021-04-16 00:30:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][850/1251] eta 0:01:54 lr 0.000736 time 0.2539 (0.2858) loss 3.8231 (3.6785) grad_norm 1.3566 (1.2841) [2021-04-16 00:30:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][860/1251] eta 0:01:51 lr 0.000736 time 0.2807 (0.2857) loss 4.4249 (3.6807) grad_norm 1.1328 (1.2837) [2021-04-16 00:30:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][870/1251] eta 0:01:48 lr 0.000736 time 0.2700 (0.2856) loss 3.2289 (3.6819) grad_norm 1.0573 (1.2839) [2021-04-16 00:30:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][880/1251] eta 0:01:45 lr 0.000736 time 0.2588 (0.2855) loss 2.7461 (3.6835) grad_norm 1.1353 (1.2835) [2021-04-16 00:30:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][890/1251] eta 0:01:43 lr 0.000736 time 0.2565 (0.2854) loss 4.0395 (3.6824) grad_norm 1.2590 (1.2828) [2021-04-16 00:30:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][900/1251] eta 0:01:40 lr 0.000736 time 0.2657 (0.2854) loss 3.2145 (3.6848) grad_norm 1.2600 (1.2830) [2021-04-16 00:31:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][910/1251] eta 0:01:37 lr 0.000736 time 0.2635 (0.2853) loss 4.2883 (3.6882) grad_norm 1.1273 (1.2828) [2021-04-16 00:31:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][920/1251] eta 0:01:34 lr 0.000736 time 0.2612 (0.2854) loss 4.3695 (3.6879) grad_norm 1.1991 (1.2820) [2021-04-16 00:31:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][930/1251] eta 0:01:31 lr 0.000736 time 0.2651 (0.2854) loss 4.6904 (3.6892) grad_norm 1.4388 (1.2814) [2021-04-16 00:31:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][940/1251] eta 0:01:28 lr 0.000735 time 0.2745 (0.2855) loss 4.1379 (3.6886) grad_norm 1.6980 (1.2823) [2021-04-16 00:31:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][950/1251] eta 0:01:25 lr 0.000735 time 0.2943 (0.2854) loss 4.3164 (3.6896) grad_norm 1.2356 (1.2828) [2021-04-16 00:31:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][960/1251] eta 0:01:23 lr 0.000735 time 0.2685 (0.2854) loss 4.6642 (3.6914) grad_norm 1.2907 (1.2830) [2021-04-16 00:31:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][970/1251] eta 0:01:20 lr 0.000735 time 0.2506 (0.2853) loss 3.1446 (3.6919) grad_norm 1.2958 (1.2823) [2021-04-16 00:31:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][980/1251] eta 0:01:17 lr 0.000735 time 0.2790 (0.2853) loss 4.2302 (3.6950) grad_norm 1.1482 (1.2814) [2021-04-16 00:31:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][990/1251] eta 0:01:14 lr 0.000735 time 0.2584 (0.2852) loss 4.2762 (3.6977) grad_norm 1.1018 (1.2823) [2021-04-16 00:31:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][1000/1251] eta 0:01:11 lr 0.000735 time 0.2780 (0.2851) loss 3.9927 (3.6985) grad_norm 1.3195 (1.2831) [2021-04-16 00:31:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][1010/1251] eta 0:01:08 lr 0.000735 time 0.2841 (0.2851) loss 4.4900 (3.6986) grad_norm 1.6518 (1.2833) [2021-04-16 00:31:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][1020/1251] eta 0:01:05 lr 0.000735 time 0.2879 (0.2850) loss 3.1010 (3.6999) grad_norm 1.5158 (1.2840) [2021-04-16 00:31:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][1030/1251] eta 0:01:02 lr 0.000735 time 0.2655 (0.2849) loss 4.1290 (3.7010) grad_norm 1.1439 (1.2842) [2021-04-16 00:31:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][1040/1251] eta 0:01:00 lr 0.000735 time 0.2758 (0.2850) loss 3.8477 (3.7019) grad_norm 1.3229 (1.2847) [2021-04-16 00:31:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][1050/1251] eta 0:00:57 lr 0.000735 time 0.2635 (0.2850) loss 3.7343 (3.7017) grad_norm 1.3600 (1.2851) [2021-04-16 00:31:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][1060/1251] eta 0:00:54 lr 0.000735 time 0.3214 (0.2850) loss 3.8639 (3.7037) grad_norm 1.2335 (1.2852) [2021-04-16 00:31:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][1070/1251] eta 0:00:51 lr 0.000735 time 0.2707 (0.2850) loss 3.5387 (3.7031) grad_norm 1.2626 (1.2862) [2021-04-16 00:31:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][1080/1251] eta 0:00:48 lr 0.000735 time 0.2681 (0.2848) loss 3.6729 (3.7056) grad_norm 1.3935 (1.2866) [2021-04-16 00:31:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][1090/1251] eta 0:00:45 lr 0.000735 time 0.3043 (0.2848) loss 4.2891 (3.7067) grad_norm 1.3288 (1.2870) [2021-04-16 00:31:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][1100/1251] eta 0:00:42 lr 0.000735 time 0.2780 (0.2847) loss 3.0059 (3.7056) grad_norm 1.2730 (1.2870) [2021-04-16 00:31:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][1110/1251] eta 0:00:40 lr 0.000735 time 0.2758 (0.2847) loss 4.3394 (3.7055) grad_norm 1.1763 (1.2871) [2021-04-16 00:32:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][1120/1251] eta 0:00:37 lr 0.000735 time 0.2857 (0.2847) loss 2.8116 (3.7028) grad_norm 1.5653 (1.2866) [2021-04-16 00:32:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][1130/1251] eta 0:00:34 lr 0.000735 time 0.2594 (0.2848) loss 3.4346 (3.7024) grad_norm 1.1211 (1.2868) [2021-04-16 00:32:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][1140/1251] eta 0:00:31 lr 0.000735 time 0.3060 (0.2847) loss 4.4373 (3.7029) grad_norm 1.2298 (1.2872) [2021-04-16 00:32:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][1150/1251] eta 0:00:28 lr 0.000735 time 0.2764 (0.2846) loss 4.1619 (3.7045) grad_norm 1.1567 (1.2864) [2021-04-16 00:32:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][1160/1251] eta 0:00:25 lr 0.000735 time 0.2523 (0.2847) loss 4.1505 (3.7064) grad_norm 1.3182 (1.2872) [2021-04-16 00:32:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][1170/1251] eta 0:00:23 lr 0.000735 time 0.2678 (0.2847) loss 4.1756 (3.7064) grad_norm 1.1325 (1.2870) [2021-04-16 00:32:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][1180/1251] eta 0:00:20 lr 0.000735 time 0.2696 (0.2847) loss 3.5247 (3.7056) grad_norm 1.1730 (1.2870) [2021-04-16 00:32:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][1190/1251] eta 0:00:17 lr 0.000735 time 0.2662 (0.2847) loss 4.3765 (3.7038) grad_norm 1.2053 (1.2871) [2021-04-16 00:32:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][1200/1251] eta 0:00:14 lr 0.000735 time 0.2856 (0.2846) loss 3.9915 (3.7036) grad_norm 1.1724 (1.2872) [2021-04-16 00:32:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][1210/1251] eta 0:00:11 lr 0.000734 time 0.2965 (0.2845) loss 2.8813 (3.7018) grad_norm 1.2336 (1.2875) [2021-04-16 00:32:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][1220/1251] eta 0:00:08 lr 0.000734 time 0.2733 (0.2845) loss 3.6564 (3.7025) grad_norm 1.1488 (1.2872) [2021-04-16 00:32:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][1230/1251] eta 0:00:05 lr 0.000734 time 0.2728 (0.2844) loss 2.8409 (3.7010) grad_norm 1.1836 (1.2870) [2021-04-16 00:32:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][1240/1251] eta 0:00:03 lr 0.000734 time 0.2481 (0.2843) loss 3.4599 (3.7007) grad_norm 1.1699 (1.2868) [2021-04-16 00:32:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [103/300][1250/1251] eta 0:00:00 lr 0.000734 time 0.2482 (0.2840) loss 4.0176 (3.7013) grad_norm 1.2451 (1.2861) [2021-04-16 00:32:39 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 103 training takes 0:05:57 [2021-04-16 00:32:39 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_103.pth saving...... [2021-04-16 00:32:46 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_103.pth saved !!! [2021-04-16 00:32:47 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.122 (1.122) Loss 1.1674 (1.1674) Acc@1 73.730 (73.730) Acc@5 91.504 (91.504) [2021-04-16 00:32:49 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.111 (0.239) Loss 1.1160 (1.1313) Acc@1 73.828 (73.988) Acc@5 92.383 (92.116) [2021-04-16 00:32:51 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.146 (0.204) Loss 1.1437 (1.1334) Acc@1 72.168 (73.670) Acc@5 91.699 (92.118) [2021-04-16 00:32:53 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.363 (0.227) Loss 1.1550 (1.1282) Acc@1 73.535 (73.705) Acc@5 91.309 (92.210) [2021-04-16 00:32:55 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.210) Loss 1.1149 (1.1284) Acc@1 74.023 (73.647) Acc@5 92.480 (92.173) [2021-04-16 00:32:58 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 73.636 Acc@5 92.194 [2021-04-16 00:32:58 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 73.6% [2021-04-16 00:32:58 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 74.16% [2021-04-16 00:33:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][0/1251] eta 1:12:26 lr 0.000734 time 3.4743 (3.4743) loss 3.6349 (3.6349) grad_norm 1.2402 (1.2402) [2021-04-16 00:33:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][10/1251] eta 0:12:13 lr 0.000734 time 0.4029 (0.5908) loss 4.4395 (3.5807) grad_norm 1.2445 (1.2765) [2021-04-16 00:33:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][20/1251] eta 0:09:02 lr 0.000734 time 0.2696 (0.4407) loss 3.3449 (3.6398) grad_norm 1.2263 (1.2808) [2021-04-16 00:33:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][30/1251] eta 0:07:56 lr 0.000734 time 0.2860 (0.3905) loss 3.8853 (3.6610) grad_norm 1.1445 (1.2736) [2021-04-16 00:33:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][40/1251] eta 0:07:18 lr 0.000734 time 0.2718 (0.3622) loss 3.1146 (3.6415) grad_norm 1.4070 (1.2837) [2021-04-16 00:33:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][50/1251] eta 0:06:55 lr 0.000734 time 0.2958 (0.3461) loss 3.9559 (3.6660) grad_norm 1.4399 (1.2807) [2021-04-16 00:33:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][60/1251] eta 0:06:40 lr 0.000734 time 0.3061 (0.3359) loss 4.0848 (3.6505) grad_norm 1.6872 (1.3031) [2021-04-16 00:33:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][70/1251] eta 0:06:28 lr 0.000734 time 0.2914 (0.3293) loss 4.2427 (3.6444) grad_norm 1.3321 (1.3003) [2021-04-16 00:33:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][80/1251] eta 0:06:19 lr 0.000734 time 0.2839 (0.3242) loss 3.2142 (3.6353) grad_norm 1.4569 (1.3030) [2021-04-16 00:33:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][90/1251] eta 0:06:11 lr 0.000734 time 0.2607 (0.3203) loss 2.8737 (3.6172) grad_norm 1.4312 (1.3105) [2021-04-16 00:33:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][100/1251] eta 0:06:03 lr 0.000734 time 0.2540 (0.3158) loss 3.5856 (3.6247) grad_norm 1.2595 (1.3144) [2021-04-16 00:33:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][110/1251] eta 0:05:56 lr 0.000734 time 0.2705 (0.3125) loss 3.1872 (3.6349) grad_norm 1.2566 (1.3152) [2021-04-16 00:33:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][120/1251] eta 0:05:51 lr 0.000734 time 0.2793 (0.3105) loss 3.2613 (3.6318) grad_norm 1.1042 (1.3095) [2021-04-16 00:33:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][130/1251] eta 0:05:45 lr 0.000734 time 0.2990 (0.3086) loss 3.0640 (3.6508) grad_norm 1.5109 (1.3049) [2021-04-16 00:33:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][140/1251] eta 0:05:42 lr 0.000734 time 0.2978 (0.3086) loss 2.8192 (3.6383) grad_norm 1.3063 (1.3022) [2021-04-16 00:33:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][150/1251] eta 0:05:38 lr 0.000734 time 0.2717 (0.3079) loss 3.6199 (3.6416) grad_norm 1.3280 (1.2980) [2021-04-16 00:33:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][160/1251] eta 0:05:34 lr 0.000734 time 0.2749 (0.3064) loss 3.8051 (3.6614) grad_norm 1.3883 (1.2944) [2021-04-16 00:33:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][170/1251] eta 0:05:29 lr 0.000734 time 0.2744 (0.3051) loss 3.7862 (3.6590) grad_norm 1.3111 (1.2959) [2021-04-16 00:33:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][180/1251] eta 0:05:24 lr 0.000734 time 0.2727 (0.3033) loss 3.8572 (3.6529) grad_norm 1.4460 (1.2966) [2021-04-16 00:33:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][190/1251] eta 0:05:20 lr 0.000734 time 0.2472 (0.3021) loss 3.6875 (3.6625) grad_norm 1.4859 (1.2976) [2021-04-16 00:33:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][200/1251] eta 0:05:16 lr 0.000734 time 0.2802 (0.3008) loss 3.9426 (3.6582) grad_norm 1.1819 (1.2963) [2021-04-16 00:34:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][210/1251] eta 0:05:11 lr 0.000734 time 0.2566 (0.2994) loss 3.7014 (3.6656) grad_norm 1.3112 (1.3009) [2021-04-16 00:34:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][220/1251] eta 0:05:08 lr 0.000734 time 0.2843 (0.2988) loss 3.8095 (3.6686) grad_norm 1.5244 (1.3026) [2021-04-16 00:34:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][230/1251] eta 0:05:04 lr 0.000733 time 0.2636 (0.2978) loss 4.5463 (3.6729) grad_norm 1.4621 (1.2992) [2021-04-16 00:34:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][240/1251] eta 0:05:00 lr 0.000733 time 0.2519 (0.2970) loss 3.7799 (3.6631) grad_norm 1.2924 (1.2994) [2021-04-16 00:34:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][250/1251] eta 0:04:57 lr 0.000733 time 0.2918 (0.2969) loss 3.7865 (3.6683) grad_norm 1.3791 (1.3017) [2021-04-16 00:34:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][260/1251] eta 0:04:53 lr 0.000733 time 0.2705 (0.2962) loss 3.9801 (3.6814) grad_norm 1.3517 (1.3000) [2021-04-16 00:34:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][270/1251] eta 0:04:50 lr 0.000733 time 0.2911 (0.2960) loss 2.6499 (3.6700) grad_norm 1.3917 (1.3009) [2021-04-16 00:34:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][280/1251] eta 0:04:47 lr 0.000733 time 0.2895 (0.2956) loss 3.2119 (3.6644) grad_norm 1.1315 (1.2995) [2021-04-16 00:34:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][290/1251] eta 0:04:43 lr 0.000733 time 0.2775 (0.2950) loss 4.2786 (3.6710) grad_norm 1.4062 (1.3017) [2021-04-16 00:34:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][300/1251] eta 0:04:40 lr 0.000733 time 0.2894 (0.2946) loss 4.4962 (3.6828) grad_norm 1.0772 (1.3013) [2021-04-16 00:34:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][310/1251] eta 0:04:36 lr 0.000733 time 0.2695 (0.2943) loss 4.1698 (3.6873) grad_norm 1.3241 (1.2996) [2021-04-16 00:34:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][320/1251] eta 0:04:33 lr 0.000733 time 0.2888 (0.2941) loss 3.6086 (3.6893) grad_norm 1.2579 (1.3029) [2021-04-16 00:34:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][330/1251] eta 0:04:30 lr 0.000733 time 0.2929 (0.2936) loss 3.8821 (3.6873) grad_norm 1.3706 (1.3033) [2021-04-16 00:34:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][340/1251] eta 0:04:27 lr 0.000733 time 0.2813 (0.2937) loss 3.0591 (3.6850) grad_norm 1.3026 (1.3038) [2021-04-16 00:34:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][350/1251] eta 0:04:24 lr 0.000733 time 0.2690 (0.2934) loss 3.7526 (3.6906) grad_norm 1.2595 (1.3060) [2021-04-16 00:34:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][360/1251] eta 0:04:21 lr 0.000733 time 0.2905 (0.2932) loss 4.0498 (3.6940) grad_norm 1.1225 (1.3058) [2021-04-16 00:34:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][370/1251] eta 0:04:18 lr 0.000733 time 0.2558 (0.2934) loss 3.5710 (3.6873) grad_norm 1.2688 (1.3080) [2021-04-16 00:34:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][380/1251] eta 0:04:15 lr 0.000733 time 0.2703 (0.2931) loss 3.0915 (3.6844) grad_norm 1.2959 (1.3090) [2021-04-16 00:34:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][390/1251] eta 0:04:11 lr 0.000733 time 0.2734 (0.2926) loss 3.6220 (3.6741) grad_norm 1.1372 (1.3100) [2021-04-16 00:34:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][400/1251] eta 0:04:08 lr 0.000733 time 0.2977 (0.2924) loss 4.0356 (3.6750) grad_norm 1.1765 (1.3070) [2021-04-16 00:34:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][410/1251] eta 0:04:05 lr 0.000733 time 0.2920 (0.2920) loss 4.2478 (3.6724) grad_norm 1.2760 (1.3051) [2021-04-16 00:35:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][420/1251] eta 0:04:02 lr 0.000733 time 0.2718 (0.2917) loss 3.9120 (3.6732) grad_norm 1.1682 (1.3046) [2021-04-16 00:35:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][430/1251] eta 0:03:59 lr 0.000733 time 0.2691 (0.2915) loss 3.9552 (3.6723) grad_norm 1.1900 (1.3058) [2021-04-16 00:35:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][440/1251] eta 0:03:56 lr 0.000733 time 0.2811 (0.2912) loss 3.1179 (3.6704) grad_norm 1.2199 (1.3040) [2021-04-16 00:35:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][450/1251] eta 0:03:53 lr 0.000733 time 0.3048 (0.2910) loss 4.3022 (3.6684) grad_norm 1.4114 (1.3027) [2021-04-16 00:35:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][460/1251] eta 0:03:50 lr 0.000733 time 0.3013 (0.2909) loss 4.5060 (3.6714) grad_norm 1.2697 (1.3009) [2021-04-16 00:35:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][470/1251] eta 0:03:47 lr 0.000733 time 0.3068 (0.2908) loss 3.6421 (3.6742) grad_norm 1.5217 (1.3006) [2021-04-16 00:35:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][480/1251] eta 0:03:44 lr 0.000733 time 0.2728 (0.2906) loss 3.7782 (3.6692) grad_norm 1.2016 (1.3002) [2021-04-16 00:35:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][490/1251] eta 0:03:41 lr 0.000733 time 0.2902 (0.2905) loss 2.5862 (3.6690) grad_norm 1.3004 (1.2990) [2021-04-16 00:35:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][500/1251] eta 0:03:37 lr 0.000732 time 0.2745 (0.2901) loss 4.2873 (3.6666) grad_norm 1.3278 (1.2992) [2021-04-16 00:35:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][510/1251] eta 0:03:34 lr 0.000732 time 0.2765 (0.2898) loss 3.5121 (3.6607) grad_norm 0.9911 (1.2974) [2021-04-16 00:35:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][520/1251] eta 0:03:31 lr 0.000732 time 0.2769 (0.2896) loss 2.4407 (3.6588) grad_norm 1.2595 (1.2975) [2021-04-16 00:35:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][530/1251] eta 0:03:28 lr 0.000732 time 0.2687 (0.2893) loss 3.5033 (3.6623) grad_norm 1.3620 (1.2994) [2021-04-16 00:35:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][540/1251] eta 0:03:25 lr 0.000732 time 0.3053 (0.2894) loss 3.7399 (3.6670) grad_norm 1.3152 (1.2998) [2021-04-16 00:35:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][550/1251] eta 0:03:22 lr 0.000732 time 0.2645 (0.2891) loss 3.5346 (3.6627) grad_norm 1.2752 (1.2990) [2021-04-16 00:35:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][560/1251] eta 0:03:19 lr 0.000732 time 0.2678 (0.2889) loss 4.2354 (3.6625) grad_norm 1.2946 (1.2973) [2021-04-16 00:35:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][570/1251] eta 0:03:16 lr 0.000732 time 0.2785 (0.2888) loss 3.9298 (3.6622) grad_norm 1.1494 (1.2963) [2021-04-16 00:35:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][580/1251] eta 0:03:13 lr 0.000732 time 0.2878 (0.2891) loss 3.2814 (3.6604) grad_norm 1.1365 (1.2964) [2021-04-16 00:35:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][590/1251] eta 0:03:11 lr 0.000732 time 0.2673 (0.2892) loss 3.4282 (3.6610) grad_norm 1.4493 (1.2955) [2021-04-16 00:35:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][600/1251] eta 0:03:08 lr 0.000732 time 0.2910 (0.2891) loss 4.2261 (3.6643) grad_norm 1.1869 (1.2951) [2021-04-16 00:35:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][610/1251] eta 0:03:05 lr 0.000732 time 0.2887 (0.2890) loss 3.9389 (3.6653) grad_norm 1.3396 (1.2958) [2021-04-16 00:35:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][620/1251] eta 0:03:02 lr 0.000732 time 0.2733 (0.2890) loss 3.3537 (3.6635) grad_norm 1.2125 (1.2961) [2021-04-16 00:36:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][630/1251] eta 0:02:59 lr 0.000732 time 0.2596 (0.2888) loss 3.6810 (3.6605) grad_norm 1.2861 (1.2955) [2021-04-16 00:36:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][640/1251] eta 0:02:56 lr 0.000732 time 0.2562 (0.2888) loss 3.0337 (3.6584) grad_norm 1.3017 (1.2950) [2021-04-16 00:36:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][650/1251] eta 0:02:53 lr 0.000732 time 0.2701 (0.2888) loss 2.9991 (3.6621) grad_norm 1.1587 (1.2951) [2021-04-16 00:36:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][660/1251] eta 0:02:50 lr 0.000732 time 0.2766 (0.2886) loss 4.4031 (3.6627) grad_norm 1.4418 (1.2950) [2021-04-16 00:36:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][670/1251] eta 0:02:47 lr 0.000732 time 0.2684 (0.2884) loss 3.4517 (3.6645) grad_norm 1.3321 (1.2946) [2021-04-16 00:36:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][680/1251] eta 0:02:44 lr 0.000732 time 0.2726 (0.2882) loss 3.6565 (3.6700) grad_norm 1.2614 (1.2953) [2021-04-16 00:36:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][690/1251] eta 0:02:41 lr 0.000732 time 0.2874 (0.2883) loss 3.0602 (3.6718) grad_norm 1.2738 (1.2956) [2021-04-16 00:36:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][700/1251] eta 0:02:38 lr 0.000732 time 0.2829 (0.2882) loss 3.8855 (3.6742) grad_norm 1.2519 (1.2961) [2021-04-16 00:36:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][710/1251] eta 0:02:35 lr 0.000732 time 0.2847 (0.2881) loss 3.6002 (3.6757) grad_norm 1.1849 (1.2966) [2021-04-16 00:36:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][720/1251] eta 0:02:33 lr 0.000732 time 0.3044 (0.2883) loss 4.6096 (3.6789) grad_norm 1.2981 (1.2970) [2021-04-16 00:36:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][730/1251] eta 0:02:30 lr 0.000732 time 0.2720 (0.2883) loss 4.3327 (3.6806) grad_norm 1.2667 (1.2969) [2021-04-16 00:36:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][740/1251] eta 0:02:27 lr 0.000732 time 0.2637 (0.2882) loss 3.9676 (3.6805) grad_norm 1.0353 (1.2972) [2021-04-16 00:36:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][750/1251] eta 0:02:24 lr 0.000732 time 0.2480 (0.2881) loss 3.1937 (3.6765) grad_norm 1.1709 (1.2974) [2021-04-16 00:36:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][760/1251] eta 0:02:21 lr 0.000732 time 0.2540 (0.2880) loss 4.1146 (3.6798) grad_norm 1.3575 (1.2972) [2021-04-16 00:36:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][770/1251] eta 0:02:18 lr 0.000731 time 0.2809 (0.2879) loss 3.6796 (3.6824) grad_norm 1.2138 (1.2975) [2021-04-16 00:36:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][780/1251] eta 0:02:15 lr 0.000731 time 0.2755 (0.2878) loss 4.4319 (3.6845) grad_norm 1.4510 (1.2973) [2021-04-16 00:36:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][790/1251] eta 0:02:12 lr 0.000731 time 0.2510 (0.2876) loss 4.1954 (3.6849) grad_norm 1.1951 (1.2972) [2021-04-16 00:36:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][800/1251] eta 0:02:09 lr 0.000731 time 0.3022 (0.2876) loss 3.8700 (3.6856) grad_norm 1.2398 (1.2968) [2021-04-16 00:36:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][810/1251] eta 0:02:06 lr 0.000731 time 0.2731 (0.2876) loss 2.9373 (3.6889) grad_norm 1.0155 (1.2956) [2021-04-16 00:36:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][820/1251] eta 0:02:03 lr 0.000731 time 0.2780 (0.2876) loss 3.6874 (3.6877) grad_norm 1.2879 (1.2957) [2021-04-16 00:36:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][830/1251] eta 0:02:01 lr 0.000731 time 0.2990 (0.2876) loss 3.8521 (3.6899) grad_norm 1.4969 (1.2944) [2021-04-16 00:37:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][840/1251] eta 0:01:58 lr 0.000731 time 0.3090 (0.2875) loss 2.8171 (3.6855) grad_norm 1.0703 (1.2955) [2021-04-16 00:37:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][850/1251] eta 0:01:55 lr 0.000731 time 0.2720 (0.2874) loss 3.8972 (3.6881) grad_norm 1.2212 (1.2960) [2021-04-16 00:37:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][860/1251] eta 0:01:52 lr 0.000731 time 0.2749 (0.2873) loss 2.7728 (3.6856) grad_norm 1.3224 (1.2958) [2021-04-16 00:37:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][870/1251] eta 0:01:49 lr 0.000731 time 0.2806 (0.2871) loss 3.8432 (3.6898) grad_norm 1.1513 (1.2958) [2021-04-16 00:37:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][880/1251] eta 0:01:46 lr 0.000731 time 0.2707 (0.2870) loss 3.7774 (3.6882) grad_norm 1.2090 (1.2963) [2021-04-16 00:37:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][890/1251] eta 0:01:43 lr 0.000731 time 0.2947 (0.2870) loss 4.1185 (3.6888) grad_norm 1.3504 (1.2962) [2021-04-16 00:37:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][900/1251] eta 0:01:40 lr 0.000731 time 0.2868 (0.2870) loss 3.6254 (3.6912) grad_norm 1.2954 (1.2962) [2021-04-16 00:37:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][910/1251] eta 0:01:37 lr 0.000731 time 0.2518 (0.2870) loss 4.0554 (3.6927) grad_norm 1.1638 (1.2957) [2021-04-16 00:37:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][920/1251] eta 0:01:34 lr 0.000731 time 0.2872 (0.2870) loss 4.5467 (3.6966) grad_norm 1.2921 (1.2957) [2021-04-16 00:37:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][930/1251] eta 0:01:32 lr 0.000731 time 0.2713 (0.2870) loss 4.1595 (3.6971) grad_norm 1.1278 (1.2955) [2021-04-16 00:37:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][940/1251] eta 0:01:29 lr 0.000731 time 0.2713 (0.2871) loss 3.8404 (3.6984) grad_norm 1.4057 (1.2950) [2021-04-16 00:37:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][950/1251] eta 0:01:26 lr 0.000731 time 0.2892 (0.2870) loss 2.9038 (3.6979) grad_norm 1.1352 (1.2952) [2021-04-16 00:37:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][960/1251] eta 0:01:23 lr 0.000731 time 0.2726 (0.2869) loss 3.5980 (3.6971) grad_norm 1.4152 (1.2948) [2021-04-16 00:37:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][970/1251] eta 0:01:20 lr 0.000731 time 0.3195 (0.2869) loss 3.0239 (3.6949) grad_norm 1.3707 (1.2943) [2021-04-16 00:37:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][980/1251] eta 0:01:17 lr 0.000731 time 0.2707 (0.2868) loss 3.6680 (3.6960) grad_norm 1.4533 (inf) [2021-04-16 00:37:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][990/1251] eta 0:01:14 lr 0.000731 time 0.2805 (0.2867) loss 4.5300 (3.6972) grad_norm 1.1696 (inf) [2021-04-16 00:37:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][1000/1251] eta 0:01:11 lr 0.000731 time 0.2626 (0.2866) loss 3.2831 (3.6948) grad_norm 1.2205 (inf) [2021-04-16 00:37:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][1010/1251] eta 0:01:09 lr 0.000731 time 0.2830 (0.2868) loss 2.5237 (3.6969) grad_norm 1.1411 (inf) [2021-04-16 00:37:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][1020/1251] eta 0:01:06 lr 0.000731 time 0.2622 (0.2868) loss 4.2507 (3.6989) grad_norm 1.2754 (inf) [2021-04-16 00:37:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][1030/1251] eta 0:01:03 lr 0.000731 time 0.2554 (0.2867) loss 3.4917 (3.6976) grad_norm 1.2654 (inf) [2021-04-16 00:37:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][1040/1251] eta 0:01:00 lr 0.000731 time 0.2718 (0.2866) loss 3.7430 (3.6963) grad_norm 1.3085 (inf) [2021-04-16 00:37:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][1050/1251] eta 0:00:57 lr 0.000730 time 0.2617 (0.2866) loss 3.4305 (3.6960) grad_norm 1.2152 (inf) [2021-04-16 00:38:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][1060/1251] eta 0:00:54 lr 0.000730 time 0.2862 (0.2865) loss 3.1345 (3.6960) grad_norm 1.3229 (inf) [2021-04-16 00:38:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][1070/1251] eta 0:00:51 lr 0.000730 time 0.2971 (0.2865) loss 4.1442 (3.6969) grad_norm 1.1270 (inf) [2021-04-16 00:38:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][1080/1251] eta 0:00:48 lr 0.000730 time 0.2966 (0.2864) loss 4.1236 (3.6959) grad_norm 1.5148 (inf) [2021-04-16 00:38:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][1090/1251] eta 0:00:46 lr 0.000730 time 0.2864 (0.2863) loss 3.9088 (3.6964) grad_norm 1.4727 (inf) [2021-04-16 00:38:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][1100/1251] eta 0:00:43 lr 0.000730 time 0.2969 (0.2862) loss 4.3283 (3.6980) grad_norm 1.3449 (inf) [2021-04-16 00:38:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][1110/1251] eta 0:00:40 lr 0.000730 time 0.2942 (0.2862) loss 3.6774 (3.6987) grad_norm 1.4226 (inf) [2021-04-16 00:38:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][1120/1251] eta 0:00:37 lr 0.000730 time 0.3012 (0.2862) loss 2.7049 (3.6972) grad_norm 1.2612 (inf) [2021-04-16 00:38:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][1130/1251] eta 0:00:34 lr 0.000730 time 0.3091 (0.2861) loss 3.9960 (3.6968) grad_norm 1.1943 (inf) [2021-04-16 00:38:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][1140/1251] eta 0:00:31 lr 0.000730 time 0.2821 (0.2861) loss 2.7849 (3.6963) grad_norm 1.1323 (inf) [2021-04-16 00:38:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][1150/1251] eta 0:00:28 lr 0.000730 time 0.2609 (0.2860) loss 3.5928 (3.6934) grad_norm 1.4557 (inf) [2021-04-16 00:38:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][1160/1251] eta 0:00:26 lr 0.000730 time 0.3015 (0.2860) loss 4.1711 (3.6941) grad_norm 1.1285 (inf) [2021-04-16 00:38:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][1170/1251] eta 0:00:23 lr 0.000730 time 0.2694 (0.2860) loss 3.4039 (3.6947) grad_norm 1.1520 (inf) [2021-04-16 00:38:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][1180/1251] eta 0:00:20 lr 0.000730 time 0.3063 (0.2860) loss 3.9404 (3.6957) grad_norm 1.1063 (inf) [2021-04-16 00:38:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][1190/1251] eta 0:00:17 lr 0.000730 time 0.2728 (0.2859) loss 4.1339 (3.6940) grad_norm 1.0901 (inf) [2021-04-16 00:38:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][1200/1251] eta 0:00:14 lr 0.000730 time 0.2660 (0.2859) loss 3.8077 (3.6926) grad_norm 1.5370 (inf) [2021-04-16 00:38:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][1210/1251] eta 0:00:11 lr 0.000730 time 0.2771 (0.2858) loss 3.7817 (3.6927) grad_norm 1.5022 (inf) [2021-04-16 00:38:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][1220/1251] eta 0:00:08 lr 0.000730 time 0.2752 (0.2857) loss 4.0585 (3.6920) grad_norm 1.3484 (inf) [2021-04-16 00:38:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][1230/1251] eta 0:00:05 lr 0.000730 time 0.3079 (0.2857) loss 4.4164 (3.6903) grad_norm 1.3392 (inf) [2021-04-16 00:38:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][1240/1251] eta 0:00:03 lr 0.000730 time 0.2484 (0.2856) loss 3.9254 (3.6884) grad_norm 1.2829 (inf) [2021-04-16 00:38:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [104/300][1250/1251] eta 0:00:00 lr 0.000730 time 0.2486 (0.2853) loss 3.5002 (3.6885) grad_norm 1.2491 (inf) [2021-04-16 00:38:58 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 104 training takes 0:05:59 [2021-04-16 00:38:58 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_104.pth saving...... [2021-04-16 00:39:07 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_104.pth saved !!! [2021-04-16 00:39:08 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.245 (1.245) Loss 1.1045 (1.1045) Acc@1 73.926 (73.926) Acc@5 91.797 (91.797) [2021-04-16 00:39:09 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.138 (0.217) Loss 1.0305 (1.1462) Acc@1 74.805 (73.375) Acc@5 94.043 (91.779) [2021-04-16 00:39:13 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 1.966 (0.292) Loss 1.0997 (1.1292) Acc@1 75.586 (73.861) Acc@5 92.676 (92.081) [2021-04-16 00:39:14 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.116 (0.245) Loss 1.1658 (1.1275) Acc@1 73.730 (73.894) Acc@5 92.188 (92.061) [2021-04-16 00:39:16 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.226) Loss 1.0592 (1.1275) Acc@1 75.488 (73.821) Acc@5 93.066 (92.133) [2021-04-16 00:39:19 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 73.780 Acc@5 92.178 [2021-04-16 00:39:19 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 73.8% [2021-04-16 00:39:19 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 74.16% [2021-04-16 00:39:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][0/1251] eta 1:50:02 lr 0.000730 time 5.2781 (5.2781) loss 4.0842 (4.0842) grad_norm 1.4749 (1.4749) [2021-04-16 00:39:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][10/1251] eta 0:15:03 lr 0.000730 time 0.2784 (0.7282) loss 3.3375 (3.8430) grad_norm 1.2120 (1.2950) [2021-04-16 00:39:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][20/1251] eta 0:10:36 lr 0.000730 time 0.2998 (0.5167) loss 4.5593 (3.8670) grad_norm 1.3463 (1.2758) [2021-04-16 00:39:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][30/1251] eta 0:09:02 lr 0.000730 time 0.2920 (0.4446) loss 4.2969 (3.7795) grad_norm 1.5916 (1.2668) [2021-04-16 00:39:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][40/1251] eta 0:08:09 lr 0.000730 time 0.2429 (0.4041) loss 4.1232 (3.7627) grad_norm 1.1354 (1.2669) [2021-04-16 00:39:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][50/1251] eta 0:07:35 lr 0.000730 time 0.2766 (0.3795) loss 3.7410 (3.7812) grad_norm 1.1586 (1.2608) [2021-04-16 00:39:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][60/1251] eta 0:07:13 lr 0.000730 time 0.2861 (0.3642) loss 2.8469 (3.7745) grad_norm 1.2899 (1.2510) [2021-04-16 00:39:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][70/1251] eta 0:06:55 lr 0.000729 time 0.2902 (0.3518) loss 3.8795 (3.7899) grad_norm 1.1194 (1.2707) [2021-04-16 00:39:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][80/1251] eta 0:06:40 lr 0.000729 time 0.2716 (0.3422) loss 4.0581 (3.7655) grad_norm 1.4229 (1.2784) [2021-04-16 00:39:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][90/1251] eta 0:06:28 lr 0.000729 time 0.2963 (0.3347) loss 4.6159 (3.7513) grad_norm 1.2130 (1.2842) [2021-04-16 00:39:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][100/1251] eta 0:06:19 lr 0.000729 time 0.2747 (0.3296) loss 4.4381 (3.7544) grad_norm 1.3491 (1.2809) [2021-04-16 00:39:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][110/1251] eta 0:06:10 lr 0.000729 time 0.2560 (0.3251) loss 4.2235 (3.7146) grad_norm 1.4831 (1.2818) [2021-04-16 00:39:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][120/1251] eta 0:06:05 lr 0.000729 time 0.3039 (0.3230) loss 2.4501 (3.7219) grad_norm 1.2720 (1.2848) [2021-04-16 00:40:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][130/1251] eta 0:05:59 lr 0.000729 time 0.2729 (0.3203) loss 3.4054 (3.7049) grad_norm 2.0218 (1.2895) [2021-04-16 00:40:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][140/1251] eta 0:05:54 lr 0.000729 time 0.2759 (0.3187) loss 3.6961 (3.6865) grad_norm 1.6768 (1.2986) [2021-04-16 00:40:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][150/1251] eta 0:05:48 lr 0.000729 time 0.3133 (0.3162) loss 3.3453 (3.6609) grad_norm 1.2637 (1.3022) [2021-04-16 00:40:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][160/1251] eta 0:05:42 lr 0.000729 time 0.2725 (0.3135) loss 3.2241 (3.6657) grad_norm 1.2911 (1.3009) [2021-04-16 00:40:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][170/1251] eta 0:05:36 lr 0.000729 time 0.2849 (0.3115) loss 2.5402 (3.6789) grad_norm 1.3835 (1.3002) [2021-04-16 00:40:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][180/1251] eta 0:05:32 lr 0.000729 time 0.2752 (0.3101) loss 3.4036 (3.6626) grad_norm 1.3421 (1.3029) [2021-04-16 00:40:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][190/1251] eta 0:05:27 lr 0.000729 time 0.2598 (0.3082) loss 4.3690 (3.6453) grad_norm 1.6583 (1.3010) [2021-04-16 00:40:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][200/1251] eta 0:05:22 lr 0.000729 time 0.2734 (0.3068) loss 3.9428 (3.6453) grad_norm 1.2016 (1.2997) [2021-04-16 00:40:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][210/1251] eta 0:05:18 lr 0.000729 time 0.2854 (0.3056) loss 4.0209 (3.6469) grad_norm 1.2014 (1.2999) [2021-04-16 00:40:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][220/1251] eta 0:05:14 lr 0.000729 time 0.2758 (0.3048) loss 4.4942 (3.6566) grad_norm 1.2075 (1.3003) [2021-04-16 00:40:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][230/1251] eta 0:05:10 lr 0.000729 time 0.2755 (0.3038) loss 3.7193 (3.6560) grad_norm 1.3038 (1.2993) [2021-04-16 00:40:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][240/1251] eta 0:05:05 lr 0.000729 time 0.2531 (0.3026) loss 4.1155 (3.6535) grad_norm 1.2832 (1.3013) [2021-04-16 00:40:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][250/1251] eta 0:05:02 lr 0.000729 time 0.2786 (0.3018) loss 4.0841 (3.6566) grad_norm 1.1116 (1.3024) [2021-04-16 00:40:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][260/1251] eta 0:04:58 lr 0.000729 time 0.2895 (0.3011) loss 4.0689 (3.6572) grad_norm 1.4605 (1.3071) [2021-04-16 00:40:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][270/1251] eta 0:04:54 lr 0.000729 time 0.2469 (0.3001) loss 4.1417 (3.6570) grad_norm 1.0882 (1.3078) [2021-04-16 00:40:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][280/1251] eta 0:04:50 lr 0.000729 time 0.2740 (0.2996) loss 3.3611 (3.6506) grad_norm 1.1588 (1.3061) [2021-04-16 00:40:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][290/1251] eta 0:04:47 lr 0.000729 time 0.2542 (0.2988) loss 4.1226 (3.6608) grad_norm 1.1391 (1.3057) [2021-04-16 00:40:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][300/1251] eta 0:04:43 lr 0.000729 time 0.2973 (0.2982) loss 4.0144 (3.6639) grad_norm 1.1326 (1.3041) [2021-04-16 00:40:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][310/1251] eta 0:04:40 lr 0.000729 time 0.3637 (0.2977) loss 4.2829 (3.6766) grad_norm 1.0847 (1.3028) [2021-04-16 00:40:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][320/1251] eta 0:04:36 lr 0.000729 time 0.2794 (0.2972) loss 4.0705 (3.6862) grad_norm 1.1222 (1.3025) [2021-04-16 00:40:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][330/1251] eta 0:04:33 lr 0.000729 time 0.2793 (0.2966) loss 4.0158 (3.6909) grad_norm 1.3298 (1.3033) [2021-04-16 00:41:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][340/1251] eta 0:04:29 lr 0.000728 time 0.2889 (0.2962) loss 2.8011 (3.6869) grad_norm 1.1748 (1.3045) [2021-04-16 00:41:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][350/1251] eta 0:04:26 lr 0.000728 time 0.2800 (0.2957) loss 3.8328 (3.6826) grad_norm 1.1706 (1.3057) [2021-04-16 00:41:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][360/1251] eta 0:04:23 lr 0.000728 time 0.2585 (0.2954) loss 4.2524 (3.6817) grad_norm 1.3398 (1.3087) [2021-04-16 00:41:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][370/1251] eta 0:04:20 lr 0.000728 time 0.2708 (0.2953) loss 3.0247 (3.6751) grad_norm 1.2101 (1.3086) [2021-04-16 00:41:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][380/1251] eta 0:04:17 lr 0.000728 time 0.2695 (0.2951) loss 3.5085 (3.6818) grad_norm 1.2878 (1.3105) [2021-04-16 00:41:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][390/1251] eta 0:04:13 lr 0.000728 time 0.3099 (0.2946) loss 3.4930 (3.6807) grad_norm 1.5555 (1.3122) [2021-04-16 00:41:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][400/1251] eta 0:04:10 lr 0.000728 time 0.2773 (0.2940) loss 2.5838 (3.6789) grad_norm 1.3796 (1.3137) [2021-04-16 00:41:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][410/1251] eta 0:04:07 lr 0.000728 time 0.2927 (0.2937) loss 4.1190 (3.6842) grad_norm 1.1293 (1.3113) [2021-04-16 00:41:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][420/1251] eta 0:04:03 lr 0.000728 time 0.2618 (0.2934) loss 4.0413 (3.6827) grad_norm 1.5274 (1.3118) [2021-04-16 00:41:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][430/1251] eta 0:04:00 lr 0.000728 time 0.2873 (0.2933) loss 4.0921 (3.6785) grad_norm 1.7432 (1.3121) [2021-04-16 00:41:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][440/1251] eta 0:03:57 lr 0.000728 time 0.2922 (0.2930) loss 3.6950 (3.6835) grad_norm 1.1936 (1.3116) [2021-04-16 00:41:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][450/1251] eta 0:03:54 lr 0.000728 time 0.2766 (0.2928) loss 3.4061 (3.6873) grad_norm 1.1877 (1.3094) [2021-04-16 00:41:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][460/1251] eta 0:03:51 lr 0.000728 time 0.2839 (0.2925) loss 4.3127 (3.6896) grad_norm 1.2251 (1.3085) [2021-04-16 00:41:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][470/1251] eta 0:03:48 lr 0.000728 time 0.2766 (0.2921) loss 3.6418 (3.6899) grad_norm 1.1710 (1.3076) [2021-04-16 00:41:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][480/1251] eta 0:03:44 lr 0.000728 time 0.2739 (0.2917) loss 2.4335 (3.6913) grad_norm 1.3288 (1.3067) [2021-04-16 00:41:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][490/1251] eta 0:03:41 lr 0.000728 time 0.2543 (0.2916) loss 3.1695 (3.6892) grad_norm 1.3911 (1.3084) [2021-04-16 00:41:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][500/1251] eta 0:03:38 lr 0.000728 time 0.2795 (0.2913) loss 3.8992 (3.6838) grad_norm 1.2544 (1.3077) [2021-04-16 00:41:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][510/1251] eta 0:03:35 lr 0.000728 time 0.2820 (0.2910) loss 3.0289 (3.6862) grad_norm 1.2273 (1.3075) [2021-04-16 00:41:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][520/1251] eta 0:03:32 lr 0.000728 time 0.2823 (0.2908) loss 4.2228 (3.6878) grad_norm 1.5145 (1.3085) [2021-04-16 00:41:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][530/1251] eta 0:03:29 lr 0.000728 time 0.2642 (0.2906) loss 4.0488 (3.6876) grad_norm 1.0877 (1.3080) [2021-04-16 00:41:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][540/1251] eta 0:03:26 lr 0.000728 time 0.2840 (0.2903) loss 3.5318 (3.6862) grad_norm 1.5580 (1.3089) [2021-04-16 00:41:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][550/1251] eta 0:03:23 lr 0.000728 time 0.2867 (0.2902) loss 4.5040 (3.6794) grad_norm 1.2203 (1.3091) [2021-04-16 00:42:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][560/1251] eta 0:03:20 lr 0.000728 time 0.2883 (0.2901) loss 4.0872 (3.6750) grad_norm 1.1500 (1.3075) [2021-04-16 00:42:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][570/1251] eta 0:03:17 lr 0.000728 time 0.2842 (0.2903) loss 3.9999 (3.6733) grad_norm 1.1890 (1.3070) [2021-04-16 00:42:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][580/1251] eta 0:03:14 lr 0.000728 time 0.2487 (0.2902) loss 2.6445 (3.6729) grad_norm 1.2468 (1.3067) [2021-04-16 00:42:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][590/1251] eta 0:03:11 lr 0.000728 time 0.2877 (0.2901) loss 3.3453 (3.6724) grad_norm 1.1406 (1.3053) [2021-04-16 00:42:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][600/1251] eta 0:03:08 lr 0.000728 time 0.5216 (0.2902) loss 3.5358 (3.6747) grad_norm 1.0891 (1.3038) [2021-04-16 00:42:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][610/1251] eta 0:03:05 lr 0.000727 time 0.3017 (0.2901) loss 3.9854 (3.6774) grad_norm 1.1724 (1.3042) [2021-04-16 00:42:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][620/1251] eta 0:03:02 lr 0.000727 time 0.2874 (0.2900) loss 3.7609 (3.6813) grad_norm 1.2818 (1.3049) [2021-04-16 00:42:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][630/1251] eta 0:03:00 lr 0.000727 time 0.2589 (0.2899) loss 3.5648 (3.6778) grad_norm 1.3346 (1.3047) [2021-04-16 00:42:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][640/1251] eta 0:02:57 lr 0.000727 time 0.2661 (0.2897) loss 2.6381 (3.6740) grad_norm 1.1735 (1.3042) [2021-04-16 00:42:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][650/1251] eta 0:02:53 lr 0.000727 time 0.2698 (0.2895) loss 3.4856 (3.6773) grad_norm 1.2746 (1.3050) [2021-04-16 00:42:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][660/1251] eta 0:02:50 lr 0.000727 time 0.2649 (0.2893) loss 3.9468 (3.6787) grad_norm 1.2698 (1.3047) [2021-04-16 00:42:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][670/1251] eta 0:02:47 lr 0.000727 time 0.2766 (0.2892) loss 2.9580 (3.6780) grad_norm 1.1289 (1.3046) [2021-04-16 00:42:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][680/1251] eta 0:02:45 lr 0.000727 time 0.2802 (0.2890) loss 2.3947 (3.6767) grad_norm 1.3088 (1.3054) [2021-04-16 00:42:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][690/1251] eta 0:02:42 lr 0.000727 time 0.2677 (0.2890) loss 4.1800 (3.6751) grad_norm 1.2140 (1.3052) [2021-04-16 00:42:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][700/1251] eta 0:02:39 lr 0.000727 time 0.2867 (0.2889) loss 4.0877 (3.6750) grad_norm 1.1341 (1.3047) [2021-04-16 00:42:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][710/1251] eta 0:02:36 lr 0.000727 time 0.2629 (0.2888) loss 4.1550 (3.6771) grad_norm 1.2042 (1.3046) [2021-04-16 00:42:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][720/1251] eta 0:02:33 lr 0.000727 time 0.2699 (0.2888) loss 3.7119 (3.6756) grad_norm 1.1915 (1.3022) [2021-04-16 00:42:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][730/1251] eta 0:02:30 lr 0.000727 time 0.2756 (0.2886) loss 4.2649 (3.6756) grad_norm 1.1789 (1.3007) [2021-04-16 00:42:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][740/1251] eta 0:02:27 lr 0.000727 time 0.2646 (0.2886) loss 4.2413 (3.6774) grad_norm 1.3099 (1.2997) [2021-04-16 00:42:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][750/1251] eta 0:02:24 lr 0.000727 time 0.2728 (0.2884) loss 4.0751 (3.6778) grad_norm 1.3784 (1.2993) [2021-04-16 00:42:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][760/1251] eta 0:02:21 lr 0.000727 time 0.2526 (0.2882) loss 3.5371 (3.6741) grad_norm 1.2706 (1.3008) [2021-04-16 00:43:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][770/1251] eta 0:02:18 lr 0.000727 time 0.2687 (0.2881) loss 2.7712 (3.6727) grad_norm 1.3254 (1.3016) [2021-04-16 00:43:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][780/1251] eta 0:02:15 lr 0.000727 time 0.3103 (0.2879) loss 4.2189 (3.6727) grad_norm 1.3441 (1.3027) [2021-04-16 00:43:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][790/1251] eta 0:02:12 lr 0.000727 time 0.2691 (0.2877) loss 4.3587 (3.6714) grad_norm 1.3702 (1.3028) [2021-04-16 00:43:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][800/1251] eta 0:02:09 lr 0.000727 time 0.2949 (0.2876) loss 4.3661 (3.6731) grad_norm 1.4918 (1.3027) [2021-04-16 00:43:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][810/1251] eta 0:02:06 lr 0.000727 time 0.2885 (0.2875) loss 3.9446 (3.6743) grad_norm 0.9765 (1.3014) [2021-04-16 00:43:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][820/1251] eta 0:02:03 lr 0.000727 time 0.2746 (0.2873) loss 2.9536 (3.6771) grad_norm 1.0623 (1.3012) [2021-04-16 00:43:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][830/1251] eta 0:02:00 lr 0.000727 time 0.2791 (0.2872) loss 4.4969 (3.6775) grad_norm 1.2590 (1.3003) [2021-04-16 00:43:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][840/1251] eta 0:01:57 lr 0.000727 time 0.2746 (0.2870) loss 4.0013 (3.6796) grad_norm 1.8749 (1.3010) [2021-04-16 00:43:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][850/1251] eta 0:01:55 lr 0.000727 time 0.2858 (0.2869) loss 3.3871 (3.6795) grad_norm 1.1987 (1.3002) [2021-04-16 00:43:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][860/1251] eta 0:01:52 lr 0.000727 time 0.2424 (0.2867) loss 3.1564 (3.6761) grad_norm 1.2096 (1.3008) [2021-04-16 00:43:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][870/1251] eta 0:01:49 lr 0.000727 time 0.2750 (0.2866) loss 3.4609 (3.6754) grad_norm 1.3478 (1.3009) [2021-04-16 00:43:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][880/1251] eta 0:01:46 lr 0.000726 time 0.2644 (0.2865) loss 4.0303 (3.6777) grad_norm 1.3433 (1.3010) [2021-04-16 00:43:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][890/1251] eta 0:01:43 lr 0.000726 time 0.2645 (0.2864) loss 2.9062 (3.6767) grad_norm 1.4368 (1.3022) [2021-04-16 00:43:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][900/1251] eta 0:01:40 lr 0.000726 time 0.2851 (0.2863) loss 3.9101 (3.6790) grad_norm 1.2978 (1.3043) [2021-04-16 00:43:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][910/1251] eta 0:01:37 lr 0.000726 time 0.2812 (0.2862) loss 3.6747 (3.6794) grad_norm 1.3388 (1.3053) [2021-04-16 00:43:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][920/1251] eta 0:01:34 lr 0.000726 time 0.2707 (0.2861) loss 4.1255 (3.6790) grad_norm 1.2277 (1.3054) [2021-04-16 00:43:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][930/1251] eta 0:01:31 lr 0.000726 time 0.2483 (0.2861) loss 3.6528 (3.6803) grad_norm 1.0869 (1.3045) [2021-04-16 00:43:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][940/1251] eta 0:01:29 lr 0.000726 time 0.2722 (0.2862) loss 4.2322 (3.6813) grad_norm 1.2043 (1.3037) [2021-04-16 00:43:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][950/1251] eta 0:01:26 lr 0.000726 time 0.2760 (0.2861) loss 3.5485 (3.6833) grad_norm 1.3190 (1.3030) [2021-04-16 00:43:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][960/1251] eta 0:01:23 lr 0.000726 time 0.2729 (0.2860) loss 2.4088 (3.6809) grad_norm 1.2678 (1.3026) [2021-04-16 00:43:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][970/1251] eta 0:01:20 lr 0.000726 time 0.2744 (0.2859) loss 4.0015 (3.6823) grad_norm 1.1946 (1.3019) [2021-04-16 00:43:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][980/1251] eta 0:01:17 lr 0.000726 time 0.2717 (0.2858) loss 2.9774 (3.6809) grad_norm 1.2034 (1.3011) [2021-04-16 00:44:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][990/1251] eta 0:01:14 lr 0.000726 time 0.2779 (0.2857) loss 2.9520 (3.6818) grad_norm 1.3684 (1.3014) [2021-04-16 00:44:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][1000/1251] eta 0:01:11 lr 0.000726 time 0.3859 (0.2857) loss 2.9931 (3.6810) grad_norm 1.2050 (1.3007) [2021-04-16 00:44:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][1010/1251] eta 0:01:08 lr 0.000726 time 0.2689 (0.2856) loss 3.6186 (3.6801) grad_norm 1.0560 (1.3006) [2021-04-16 00:44:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][1020/1251] eta 0:01:05 lr 0.000726 time 0.2653 (0.2855) loss 4.0824 (3.6808) grad_norm 1.6057 (1.3008) [2021-04-16 00:44:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][1030/1251] eta 0:01:03 lr 0.000726 time 0.2583 (0.2855) loss 2.7593 (3.6780) grad_norm 1.3857 (1.3007) [2021-04-16 00:44:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][1040/1251] eta 0:01:00 lr 0.000726 time 0.2725 (0.2854) loss 3.9381 (3.6794) grad_norm 1.3534 (1.3006) [2021-04-16 00:44:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][1050/1251] eta 0:00:57 lr 0.000726 time 0.2669 (0.2853) loss 3.6196 (3.6790) grad_norm 1.2136 (1.2999) [2021-04-16 00:44:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][1060/1251] eta 0:00:54 lr 0.000726 time 0.2741 (0.2853) loss 2.5952 (3.6789) grad_norm 1.4327 (1.2993) [2021-04-16 00:44:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][1070/1251] eta 0:00:51 lr 0.000726 time 0.2746 (0.2854) loss 4.2382 (3.6799) grad_norm 1.3272 (1.2996) [2021-04-16 00:44:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][1080/1251] eta 0:00:48 lr 0.000726 time 0.2690 (0.2853) loss 2.6169 (3.6803) grad_norm 1.1598 (1.2997) [2021-04-16 00:44:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][1090/1251] eta 0:00:45 lr 0.000726 time 0.2791 (0.2852) loss 4.6554 (3.6834) grad_norm 1.1701 (1.2994) [2021-04-16 00:44:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][1100/1251] eta 0:00:43 lr 0.000726 time 0.3045 (0.2852) loss 3.9101 (3.6816) grad_norm 1.1724 (1.2988) [2021-04-16 00:44:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][1110/1251] eta 0:00:40 lr 0.000726 time 0.2889 (0.2852) loss 2.6134 (3.6812) grad_norm 1.0847 (1.2982) [2021-04-16 00:44:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][1120/1251] eta 0:00:37 lr 0.000726 time 0.2722 (0.2851) loss 3.4850 (3.6806) grad_norm 1.2777 (1.2979) [2021-04-16 00:44:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][1130/1251] eta 0:00:34 lr 0.000726 time 0.2724 (0.2850) loss 3.2508 (3.6759) grad_norm 1.3303 (1.2974) [2021-04-16 00:44:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][1140/1251] eta 0:00:31 lr 0.000726 time 0.2456 (0.2850) loss 3.6904 (3.6746) grad_norm 1.3548 (1.2975) [2021-04-16 00:44:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][1150/1251] eta 0:00:28 lr 0.000725 time 0.2756 (0.2850) loss 3.1135 (3.6756) grad_norm 1.0507 (1.2965) [2021-04-16 00:44:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][1160/1251] eta 0:00:25 lr 0.000725 time 0.2796 (0.2850) loss 4.2720 (3.6761) grad_norm 1.6361 (1.2969) [2021-04-16 00:44:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][1170/1251] eta 0:00:23 lr 0.000725 time 0.2896 (0.2849) loss 3.3650 (3.6761) grad_norm 1.2761 (1.2963) [2021-04-16 00:44:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][1180/1251] eta 0:00:20 lr 0.000725 time 0.2597 (0.2850) loss 4.1272 (3.6760) grad_norm 1.3412 (1.2962) [2021-04-16 00:44:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][1190/1251] eta 0:00:17 lr 0.000725 time 0.3087 (0.2850) loss 4.1597 (3.6758) grad_norm 1.4495 (1.2958) [2021-04-16 00:45:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][1200/1251] eta 0:00:14 lr 0.000725 time 0.2935 (0.2849) loss 2.9273 (3.6757) grad_norm 1.2150 (1.2954) [2021-04-16 00:45:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][1210/1251] eta 0:00:11 lr 0.000725 time 0.2750 (0.2849) loss 3.8842 (3.6773) grad_norm 1.1345 (1.2949) [2021-04-16 00:45:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][1220/1251] eta 0:00:08 lr 0.000725 time 0.2676 (0.2849) loss 4.1341 (3.6765) grad_norm 1.2826 (1.2947) [2021-04-16 00:45:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][1230/1251] eta 0:00:05 lr 0.000725 time 0.3070 (0.2848) loss 3.5177 (3.6739) grad_norm 1.4770 (1.2945) [2021-04-16 00:45:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][1240/1251] eta 0:00:03 lr 0.000725 time 0.2485 (0.2847) loss 3.9771 (3.6738) grad_norm 1.2798 (1.2942) [2021-04-16 00:45:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [105/300][1250/1251] eta 0:00:00 lr 0.000725 time 0.2484 (0.2844) loss 3.1346 (3.6745) grad_norm 1.2488 (1.2936) [2021-04-16 00:45:17 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 105 training takes 0:05:58 [2021-04-16 00:45:17 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_105.pth saving...... [2021-04-16 00:45:32 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_105.pth saved !!! [2021-04-16 00:45:34 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.136 (1.136) Loss 1.1711 (1.1711) Acc@1 71.289 (71.289) Acc@5 91.797 (91.797) [2021-04-16 00:45:35 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.118 (0.244) Loss 1.1166 (1.1371) Acc@1 74.805 (73.651) Acc@5 91.797 (92.037) [2021-04-16 00:45:37 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.121 (0.213) Loss 1.0951 (1.1284) Acc@1 75.195 (73.726) Acc@5 93.066 (92.174) [2021-04-16 00:45:39 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.374 (0.228) Loss 1.0832 (1.1223) Acc@1 73.828 (74.017) Acc@5 92.871 (92.150) [2021-04-16 00:45:41 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.215) Loss 1.1271 (1.1222) Acc@1 74.512 (74.081) Acc@5 91.699 (92.130) [2021-04-16 00:45:45 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 74.088 Acc@5 92.126 [2021-04-16 00:45:45 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 74.1% [2021-04-16 00:45:45 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 74.16% [2021-04-16 00:45:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][0/1251] eta 1:32:12 lr 0.000725 time 4.4225 (4.4225) loss 3.3808 (3.3808) grad_norm 1.2147 (1.2147) [2021-04-16 00:45:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][10/1251] eta 0:13:29 lr 0.000725 time 0.2659 (0.6519) loss 3.8981 (3.5786) grad_norm 1.2216 (1.2412) [2021-04-16 00:45:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][20/1251] eta 0:09:44 lr 0.000725 time 0.2657 (0.4748) loss 2.8224 (3.5937) grad_norm 1.0747 (1.1967) [2021-04-16 00:45:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][30/1251] eta 0:08:25 lr 0.000725 time 0.2870 (0.4143) loss 4.2396 (3.6853) grad_norm 1.1158 (1.1872) [2021-04-16 00:46:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][40/1251] eta 0:07:43 lr 0.000725 time 0.2792 (0.3825) loss 3.4386 (3.6751) grad_norm 1.2671 (1.1907) [2021-04-16 00:46:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][50/1251] eta 0:07:14 lr 0.000725 time 0.2792 (0.3619) loss 4.2381 (3.7178) grad_norm 1.1967 (1.2032) [2021-04-16 00:46:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][60/1251] eta 0:06:54 lr 0.000725 time 0.2955 (0.3479) loss 3.6013 (3.7780) grad_norm 1.1991 (1.2082) [2021-04-16 00:46:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][70/1251] eta 0:06:39 lr 0.000725 time 0.2911 (0.3384) loss 3.1333 (3.7643) grad_norm 1.1836 (1.2191) [2021-04-16 00:46:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][80/1251] eta 0:06:28 lr 0.000725 time 0.2965 (0.3317) loss 3.8683 (3.7774) grad_norm 1.4323 (1.2308) [2021-04-16 00:46:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][90/1251] eta 0:06:17 lr 0.000725 time 0.2790 (0.3254) loss 4.0681 (3.7599) grad_norm 1.3767 (1.2321) [2021-04-16 00:46:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][100/1251] eta 0:06:09 lr 0.000725 time 0.2730 (0.3210) loss 4.0326 (3.7484) grad_norm 1.3253 (1.2397) [2021-04-16 00:46:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][110/1251] eta 0:06:02 lr 0.000725 time 0.2764 (0.3175) loss 3.6573 (3.7432) grad_norm 1.2317 (1.2504) [2021-04-16 00:46:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][120/1251] eta 0:05:55 lr 0.000725 time 0.2746 (0.3145) loss 3.0838 (3.7409) grad_norm 1.0825 (1.2520) [2021-04-16 00:46:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][130/1251] eta 0:05:49 lr 0.000725 time 0.2932 (0.3116) loss 3.8779 (3.7290) grad_norm 1.1887 (1.2524) [2021-04-16 00:46:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][140/1251] eta 0:05:45 lr 0.000725 time 0.2701 (0.3109) loss 4.0985 (3.7361) grad_norm 1.3025 (1.2521) [2021-04-16 00:46:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][150/1251] eta 0:05:39 lr 0.000725 time 0.2921 (0.3085) loss 3.4316 (3.7126) grad_norm 1.2212 (1.2529) [2021-04-16 00:46:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][160/1251] eta 0:05:34 lr 0.000725 time 0.2792 (0.3065) loss 2.9194 (3.7085) grad_norm 1.3419 (1.2632) [2021-04-16 00:46:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][170/1251] eta 0:05:29 lr 0.000724 time 0.2920 (0.3048) loss 3.9461 (3.7135) grad_norm 1.2554 (1.2641) [2021-04-16 00:46:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][180/1251] eta 0:05:24 lr 0.000724 time 0.2774 (0.3030) loss 3.9671 (3.7235) grad_norm 1.0869 (1.2659) [2021-04-16 00:46:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][190/1251] eta 0:05:19 lr 0.000724 time 0.2874 (0.3016) loss 2.5075 (3.7003) grad_norm 1.1932 (1.2631) [2021-04-16 00:46:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][200/1251] eta 0:05:15 lr 0.000724 time 0.2594 (0.3003) loss 3.8979 (3.7078) grad_norm 1.3816 (1.2664) [2021-04-16 00:46:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][210/1251] eta 0:05:12 lr 0.000724 time 0.4208 (0.2999) loss 4.4164 (3.7116) grad_norm 1.5013 (1.2684) [2021-04-16 00:46:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][220/1251] eta 0:05:07 lr 0.000724 time 0.2719 (0.2987) loss 3.1660 (3.6892) grad_norm 1.4284 (1.2702) [2021-04-16 00:46:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][230/1251] eta 0:05:04 lr 0.000724 time 0.2729 (0.2978) loss 4.4944 (3.6821) grad_norm 1.9507 (1.2747) [2021-04-16 00:46:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][240/1251] eta 0:05:00 lr 0.000724 time 0.2734 (0.2968) loss 4.4220 (3.6904) grad_norm 1.1719 (1.2763) [2021-04-16 00:47:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][250/1251] eta 0:04:56 lr 0.000724 time 0.2539 (0.2962) loss 2.9731 (3.6913) grad_norm 1.3739 (1.2765) [2021-04-16 00:47:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][260/1251] eta 0:04:53 lr 0.000724 time 0.2925 (0.2957) loss 4.5698 (3.6932) grad_norm 1.5017 (1.2805) [2021-04-16 00:47:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][270/1251] eta 0:04:49 lr 0.000724 time 0.2520 (0.2951) loss 4.3551 (3.6985) grad_norm 1.2679 (1.2783) [2021-04-16 00:47:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][280/1251] eta 0:04:45 lr 0.000724 time 0.2667 (0.2945) loss 4.3611 (3.7004) grad_norm 1.3411 (1.2820) [2021-04-16 00:47:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][290/1251] eta 0:04:43 lr 0.000724 time 0.2465 (0.2945) loss 3.7071 (3.6976) grad_norm 1.2216 (1.2834) [2021-04-16 00:47:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][300/1251] eta 0:04:39 lr 0.000724 time 0.2895 (0.2941) loss 4.0794 (3.6960) grad_norm 1.2848 (1.2817) [2021-04-16 00:47:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][310/1251] eta 0:04:36 lr 0.000724 time 0.2790 (0.2937) loss 4.0087 (3.7041) grad_norm 1.6276 (1.2801) [2021-04-16 00:47:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][320/1251] eta 0:04:33 lr 0.000724 time 0.2741 (0.2934) loss 3.5865 (3.7066) grad_norm 1.1798 (1.2807) [2021-04-16 00:47:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][330/1251] eta 0:04:29 lr 0.000724 time 0.2416 (0.2930) loss 3.0836 (3.7041) grad_norm 1.1713 (1.2804) [2021-04-16 00:47:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][340/1251] eta 0:04:26 lr 0.000724 time 0.2783 (0.2924) loss 3.2804 (3.7032) grad_norm 1.1902 (1.2805) [2021-04-16 00:47:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][350/1251] eta 0:04:23 lr 0.000724 time 0.2589 (0.2926) loss 3.9449 (3.7059) grad_norm 1.3708 (1.2821) [2021-04-16 00:47:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][360/1251] eta 0:04:20 lr 0.000724 time 0.2760 (0.2928) loss 3.2724 (3.7093) grad_norm 1.1269 (1.2809) [2021-04-16 00:47:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][370/1251] eta 0:04:17 lr 0.000724 time 0.2817 (0.2928) loss 3.2427 (3.7105) grad_norm 1.1910 (1.2805) [2021-04-16 00:47:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][380/1251] eta 0:04:15 lr 0.000724 time 0.2674 (0.2928) loss 4.4235 (3.7178) grad_norm 1.5269 (1.2816) [2021-04-16 00:47:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][390/1251] eta 0:04:11 lr 0.000724 time 0.2828 (0.2925) loss 4.1103 (3.7189) grad_norm 1.1209 (1.2809) [2021-04-16 00:47:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][400/1251] eta 0:04:08 lr 0.000724 time 0.2870 (0.2922) loss 2.8519 (3.7205) grad_norm 1.0519 (1.2826) [2021-04-16 00:47:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][410/1251] eta 0:04:05 lr 0.000724 time 0.2770 (0.2919) loss 2.3911 (3.7144) grad_norm 1.3964 (1.2821) [2021-04-16 00:47:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][420/1251] eta 0:04:02 lr 0.000724 time 0.2854 (0.2915) loss 3.8465 (3.7108) grad_norm 1.1939 (1.2828) [2021-04-16 00:47:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][430/1251] eta 0:03:59 lr 0.000723 time 0.2869 (0.2912) loss 3.3901 (3.7104) grad_norm 1.1243 (1.2825) [2021-04-16 00:47:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][440/1251] eta 0:03:55 lr 0.000723 time 0.2824 (0.2908) loss 3.2390 (3.7095) grad_norm 1.2054 (1.2825) [2021-04-16 00:47:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][450/1251] eta 0:03:52 lr 0.000723 time 0.2759 (0.2906) loss 3.4793 (3.7028) grad_norm 1.2439 (1.2819) [2021-04-16 00:47:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][460/1251] eta 0:03:49 lr 0.000723 time 0.2835 (0.2903) loss 4.0731 (3.7078) grad_norm 1.1394 (1.2855) [2021-04-16 00:48:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][470/1251] eta 0:03:46 lr 0.000723 time 0.2868 (0.2900) loss 3.5760 (3.7090) grad_norm 1.2906 (1.2880) [2021-04-16 00:48:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][480/1251] eta 0:03:43 lr 0.000723 time 0.2977 (0.2898) loss 3.5153 (3.7083) grad_norm 1.2690 (1.2896) [2021-04-16 00:48:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][490/1251] eta 0:03:40 lr 0.000723 time 0.2799 (0.2895) loss 3.7022 (3.7061) grad_norm 1.3138 (1.2914) [2021-04-16 00:48:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][500/1251] eta 0:03:37 lr 0.000723 time 0.2905 (0.2894) loss 3.4015 (3.7109) grad_norm 1.4108 (1.2917) [2021-04-16 00:48:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][510/1251] eta 0:03:34 lr 0.000723 time 0.2627 (0.2891) loss 2.5777 (3.7109) grad_norm 1.4477 (1.2911) [2021-04-16 00:48:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][520/1251] eta 0:03:31 lr 0.000723 time 0.2838 (0.2892) loss 3.7252 (3.7126) grad_norm 1.4750 (1.2913) [2021-04-16 00:48:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][530/1251] eta 0:03:28 lr 0.000723 time 0.2834 (0.2889) loss 2.5550 (3.7094) grad_norm 1.7779 (1.2919) [2021-04-16 00:48:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][540/1251] eta 0:03:25 lr 0.000723 time 0.2655 (0.2889) loss 2.6769 (3.7134) grad_norm 1.2770 (1.2920) [2021-04-16 00:48:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][550/1251] eta 0:03:22 lr 0.000723 time 0.2749 (0.2888) loss 3.7826 (3.7147) grad_norm 1.4801 (1.2939) [2021-04-16 00:48:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][560/1251] eta 0:03:19 lr 0.000723 time 0.2876 (0.2887) loss 3.8184 (3.7115) grad_norm 1.2789 (1.2935) [2021-04-16 00:48:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][570/1251] eta 0:03:16 lr 0.000723 time 0.2693 (0.2884) loss 4.4501 (3.7097) grad_norm 1.5962 (1.2937) [2021-04-16 00:48:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][580/1251] eta 0:03:13 lr 0.000723 time 0.2708 (0.2882) loss 4.3512 (3.7092) grad_norm 1.6398 (1.2960) [2021-04-16 00:48:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][590/1251] eta 0:03:10 lr 0.000723 time 0.2874 (0.2880) loss 4.4158 (3.7106) grad_norm 1.5788 (1.2965) [2021-04-16 00:48:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][600/1251] eta 0:03:07 lr 0.000723 time 0.2981 (0.2878) loss 3.9312 (3.7113) grad_norm 1.5015 (1.2962) [2021-04-16 00:48:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][610/1251] eta 0:03:04 lr 0.000723 time 0.3078 (0.2877) loss 3.9271 (3.7127) grad_norm 1.2853 (1.2959) [2021-04-16 00:48:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][620/1251] eta 0:03:01 lr 0.000723 time 0.2922 (0.2876) loss 4.2512 (3.7082) grad_norm 1.1758 (1.2948) [2021-04-16 00:48:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][630/1251] eta 0:02:58 lr 0.000723 time 0.3079 (0.2876) loss 3.5353 (3.7076) grad_norm 1.3962 (1.2957) [2021-04-16 00:48:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][640/1251] eta 0:02:55 lr 0.000723 time 0.3066 (0.2876) loss 3.6652 (3.7101) grad_norm 1.4646 (1.2966) [2021-04-16 00:48:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][650/1251] eta 0:02:52 lr 0.000723 time 0.2690 (0.2876) loss 4.2437 (3.7125) grad_norm 1.3587 (1.2972) [2021-04-16 00:48:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][660/1251] eta 0:02:49 lr 0.000723 time 0.2935 (0.2875) loss 4.0152 (3.7114) grad_norm 1.2587 (1.2963) [2021-04-16 00:48:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][670/1251] eta 0:02:46 lr 0.000723 time 0.2900 (0.2873) loss 4.2630 (3.7146) grad_norm 1.0924 (1.2960) [2021-04-16 00:49:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][680/1251] eta 0:02:43 lr 0.000723 time 0.3027 (0.2872) loss 2.8718 (3.7143) grad_norm 1.6814 (1.2956) [2021-04-16 00:49:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][690/1251] eta 0:02:41 lr 0.000723 time 0.2667 (0.2870) loss 2.8404 (3.7152) grad_norm 1.3153 (1.2949) [2021-04-16 00:49:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][700/1251] eta 0:02:38 lr 0.000722 time 0.2968 (0.2870) loss 4.6195 (3.7170) grad_norm 1.5970 (1.2964) [2021-04-16 00:49:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][710/1251] eta 0:02:35 lr 0.000722 time 0.2505 (0.2868) loss 3.6798 (3.7188) grad_norm 1.5646 (1.2979) [2021-04-16 00:49:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][720/1251] eta 0:02:32 lr 0.000722 time 0.2752 (0.2869) loss 4.1284 (3.7169) grad_norm 1.4449 (1.2985) [2021-04-16 00:49:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][730/1251] eta 0:02:29 lr 0.000722 time 0.2881 (0.2869) loss 4.0944 (3.7141) grad_norm 1.3326 (1.2989) [2021-04-16 00:49:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][740/1251] eta 0:02:26 lr 0.000722 time 0.2742 (0.2868) loss 4.2552 (3.7180) grad_norm 1.3200 (1.2992) [2021-04-16 00:49:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][750/1251] eta 0:02:23 lr 0.000722 time 0.2690 (0.2867) loss 3.2032 (3.7171) grad_norm 1.2572 (1.2989) [2021-04-16 00:49:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][760/1251] eta 0:02:20 lr 0.000722 time 0.2889 (0.2866) loss 3.7542 (3.7175) grad_norm 1.2958 (1.2983) [2021-04-16 00:49:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][770/1251] eta 0:02:17 lr 0.000722 time 0.2705 (0.2865) loss 4.0350 (3.7216) grad_norm 1.2426 (1.2980) [2021-04-16 00:49:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][780/1251] eta 0:02:14 lr 0.000722 time 0.3045 (0.2865) loss 3.5656 (3.7238) grad_norm 1.1736 (1.2974) [2021-04-16 00:49:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][790/1251] eta 0:02:12 lr 0.000722 time 0.2965 (0.2866) loss 3.4842 (3.7207) grad_norm 1.4723 (1.2979) [2021-04-16 00:49:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][800/1251] eta 0:02:09 lr 0.000722 time 0.2619 (0.2864) loss 4.1873 (3.7224) grad_norm 1.3702 (1.2990) [2021-04-16 00:49:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][810/1251] eta 0:02:06 lr 0.000722 time 0.2765 (0.2864) loss 3.8818 (3.7247) grad_norm 1.3529 (1.2990) [2021-04-16 00:49:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][820/1251] eta 0:02:03 lr 0.000722 time 0.2654 (0.2863) loss 3.0760 (3.7243) grad_norm 1.5232 (1.3003) [2021-04-16 00:49:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][830/1251] eta 0:02:00 lr 0.000722 time 0.2937 (0.2863) loss 2.6459 (3.7254) grad_norm 1.2160 (1.3010) [2021-04-16 00:49:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][840/1251] eta 0:01:57 lr 0.000722 time 0.2583 (0.2861) loss 4.0708 (3.7246) grad_norm 1.2071 (1.3011) [2021-04-16 00:49:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][850/1251] eta 0:01:54 lr 0.000722 time 0.2724 (0.2861) loss 3.6630 (3.7288) grad_norm 1.2365 (1.3025) [2021-04-16 00:49:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][860/1251] eta 0:01:51 lr 0.000722 time 0.2895 (0.2860) loss 3.8969 (3.7294) grad_norm 1.3014 (1.3019) [2021-04-16 00:49:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][870/1251] eta 0:01:48 lr 0.000722 time 0.2734 (0.2860) loss 3.1695 (3.7292) grad_norm 1.2339 (1.3012) [2021-04-16 00:49:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][880/1251] eta 0:01:46 lr 0.000722 time 0.2810 (0.2859) loss 3.7097 (3.7270) grad_norm 1.1815 (1.3002) [2021-04-16 00:50:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][890/1251] eta 0:01:43 lr 0.000722 time 0.2669 (0.2858) loss 3.6724 (3.7278) grad_norm 1.1065 (1.3012) [2021-04-16 00:50:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][900/1251] eta 0:01:40 lr 0.000722 time 0.2439 (0.2857) loss 4.0224 (3.7273) grad_norm 1.3577 (1.3013) [2021-04-16 00:50:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][910/1251] eta 0:01:37 lr 0.000722 time 0.2724 (0.2856) loss 3.0567 (3.7263) grad_norm 1.3059 (1.3011) [2021-04-16 00:50:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][920/1251] eta 0:01:34 lr 0.000722 time 0.2783 (0.2856) loss 3.9100 (3.7270) grad_norm 1.3334 (1.3009) [2021-04-16 00:50:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][930/1251] eta 0:01:31 lr 0.000722 time 0.2739 (0.2857) loss 3.7694 (3.7246) grad_norm 1.1605 (1.3006) [2021-04-16 00:50:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][940/1251] eta 0:01:28 lr 0.000722 time 0.3930 (0.2858) loss 3.8711 (3.7234) grad_norm 1.0898 (1.2998) [2021-04-16 00:50:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][950/1251] eta 0:01:26 lr 0.000722 time 0.2801 (0.2857) loss 2.9145 (3.7246) grad_norm 1.2186 (1.2995) [2021-04-16 00:50:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][960/1251] eta 0:01:23 lr 0.000722 time 0.2604 (0.2857) loss 3.9348 (3.7261) grad_norm 1.3848 (1.2996) [2021-04-16 00:50:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][970/1251] eta 0:01:20 lr 0.000721 time 0.2962 (0.2856) loss 4.0178 (3.7277) grad_norm 1.4176 (1.2996) [2021-04-16 00:50:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][980/1251] eta 0:01:17 lr 0.000721 time 0.2631 (0.2855) loss 2.6698 (3.7260) grad_norm 1.1883 (1.2995) [2021-04-16 00:50:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][990/1251] eta 0:01:14 lr 0.000721 time 0.2768 (0.2856) loss 4.1532 (3.7286) grad_norm 1.2176 (1.2990) [2021-04-16 00:50:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][1000/1251] eta 0:01:11 lr 0.000721 time 0.2867 (0.2856) loss 4.0475 (3.7278) grad_norm 1.2573 (1.2985) [2021-04-16 00:50:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][1010/1251] eta 0:01:08 lr 0.000721 time 0.2681 (0.2854) loss 3.3627 (3.7299) grad_norm 1.2443 (1.2976) [2021-04-16 00:50:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][1020/1251] eta 0:01:05 lr 0.000721 time 0.2742 (0.2853) loss 4.4421 (3.7316) grad_norm 2.0110 (1.2978) [2021-04-16 00:50:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][1030/1251] eta 0:01:03 lr 0.000721 time 0.2497 (0.2852) loss 3.8848 (3.7320) grad_norm 1.3516 (1.2982) [2021-04-16 00:50:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][1040/1251] eta 0:01:00 lr 0.000721 time 0.2598 (0.2851) loss 3.6799 (3.7300) grad_norm 1.5005 (1.2990) [2021-04-16 00:50:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][1050/1251] eta 0:00:57 lr 0.000721 time 0.2695 (0.2850) loss 4.2645 (3.7302) grad_norm 1.1729 (1.2986) [2021-04-16 00:50:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][1060/1251] eta 0:00:54 lr 0.000721 time 0.2657 (0.2849) loss 4.3193 (3.7302) grad_norm 1.2495 (1.2991) [2021-04-16 00:50:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][1070/1251] eta 0:00:51 lr 0.000721 time 0.2897 (0.2849) loss 3.9479 (3.7315) grad_norm 1.0914 (1.2988) [2021-04-16 00:50:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][1080/1251] eta 0:00:48 lr 0.000721 time 0.2768 (0.2847) loss 4.1924 (3.7297) grad_norm 1.1654 (1.2981) [2021-04-16 00:50:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][1090/1251] eta 0:00:45 lr 0.000721 time 0.2465 (0.2848) loss 2.5601 (3.7287) grad_norm 1.3311 (1.2985) [2021-04-16 00:50:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][1100/1251] eta 0:00:42 lr 0.000721 time 0.2825 (0.2848) loss 3.4322 (3.7294) grad_norm 1.5157 (1.2986) [2021-04-16 00:51:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][1110/1251] eta 0:00:40 lr 0.000721 time 0.2477 (0.2846) loss 3.9644 (3.7298) grad_norm 1.3825 (1.2986) [2021-04-16 00:51:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][1120/1251] eta 0:00:37 lr 0.000721 time 0.2706 (0.2846) loss 4.2378 (3.7294) grad_norm 1.2538 (1.2984) [2021-04-16 00:51:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][1130/1251] eta 0:00:34 lr 0.000721 time 0.2921 (0.2845) loss 3.3693 (3.7277) grad_norm 1.3426 (1.2978) [2021-04-16 00:51:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][1140/1251] eta 0:00:31 lr 0.000721 time 0.2652 (0.2844) loss 3.6491 (3.7279) grad_norm 1.3035 (1.2969) [2021-04-16 00:51:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][1150/1251] eta 0:00:28 lr 0.000721 time 0.2842 (0.2843) loss 3.5057 (3.7252) grad_norm 1.1830 (1.2961) [2021-04-16 00:51:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][1160/1251] eta 0:00:25 lr 0.000721 time 0.2780 (0.2843) loss 3.4484 (3.7242) grad_norm 1.2780 (1.2953) [2021-04-16 00:51:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][1170/1251] eta 0:00:23 lr 0.000721 time 0.2956 (0.2842) loss 4.2780 (3.7263) grad_norm 1.1728 (1.2952) [2021-04-16 00:51:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][1180/1251] eta 0:00:20 lr 0.000721 time 0.2564 (0.2842) loss 4.2463 (3.7250) grad_norm 1.4066 (1.2947) [2021-04-16 00:51:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][1190/1251] eta 0:00:17 lr 0.000721 time 0.2634 (0.2842) loss 3.9927 (3.7232) grad_norm 1.1428 (1.2949) [2021-04-16 00:51:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][1200/1251] eta 0:00:14 lr 0.000721 time 0.2869 (0.2842) loss 2.7337 (3.7194) grad_norm 1.2105 (1.2951) [2021-04-16 00:51:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][1210/1251] eta 0:00:11 lr 0.000721 time 0.3109 (0.2841) loss 3.1508 (3.7192) grad_norm 1.4266 (1.2947) [2021-04-16 00:51:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][1220/1251] eta 0:00:08 lr 0.000721 time 0.2792 (0.2841) loss 3.8829 (3.7196) grad_norm 1.2294 (1.2950) [2021-04-16 00:51:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][1230/1251] eta 0:00:05 lr 0.000721 time 0.2756 (0.2840) loss 3.7662 (3.7204) grad_norm 1.2816 (1.2957) [2021-04-16 00:51:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][1240/1251] eta 0:00:03 lr 0.000720 time 0.2481 (0.2839) loss 3.8025 (3.7176) grad_norm 1.2161 (1.2948) [2021-04-16 00:51:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [106/300][1250/1251] eta 0:00:00 lr 0.000720 time 0.2506 (0.2836) loss 4.4604 (3.7170) grad_norm 1.3316 (1.2950) [2021-04-16 00:51:43 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 106 training takes 0:05:57 [2021-04-16 00:51:43 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_106.pth saving...... [2021-04-16 00:51:56 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_106.pth saved !!! [2021-04-16 00:51:57 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.076 (1.076) Loss 1.1178 (1.1178) Acc@1 73.926 (73.926) Acc@5 92.285 (92.285) [2021-04-16 00:51:59 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.120 (0.246) Loss 1.1406 (1.1377) Acc@1 73.828 (73.553) Acc@5 91.602 (92.347) [2021-04-16 00:52:00 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.142 (0.211) Loss 1.1696 (1.1437) Acc@1 73.730 (73.289) Acc@5 90.527 (92.104) [2021-04-16 00:52:03 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.129 (0.241) Loss 1.1700 (1.1409) Acc@1 73.535 (73.409) Acc@5 91.211 (92.118) [2021-04-16 00:52:05 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.395 (0.217) Loss 1.0670 (1.1363) Acc@1 75.781 (73.514) Acc@5 93.457 (92.130) [2021-04-16 00:52:08 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 73.606 Acc@5 92.170 [2021-04-16 00:52:08 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 73.6% [2021-04-16 00:52:08 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 74.16% [2021-04-16 00:52:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][0/1251] eta 1:28:44 lr 0.000720 time 4.2563 (4.2563) loss 3.1540 (3.1540) grad_norm 1.3107 (1.3107) [2021-04-16 00:52:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][10/1251] eta 0:13:24 lr 0.000720 time 0.3097 (0.6480) loss 3.5514 (3.4655) grad_norm 1.0749 (1.2523) [2021-04-16 00:52:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][20/1251] eta 0:09:36 lr 0.000720 time 0.2776 (0.4684) loss 4.0490 (3.5496) grad_norm 1.4106 (1.2755) [2021-04-16 00:52:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][30/1251] eta 0:08:16 lr 0.000720 time 0.3057 (0.4063) loss 4.0122 (3.5377) grad_norm 1.2688 (1.2787) [2021-04-16 00:52:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][40/1251] eta 0:07:33 lr 0.000720 time 0.2459 (0.3741) loss 4.4761 (3.6110) grad_norm 1.3230 (1.2857) [2021-04-16 00:52:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][50/1251] eta 0:07:07 lr 0.000720 time 0.3047 (0.3556) loss 3.8779 (3.6206) grad_norm 1.2797 (1.2934) [2021-04-16 00:52:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][60/1251] eta 0:06:47 lr 0.000720 time 0.2618 (0.3421) loss 3.8354 (3.5919) grad_norm 1.3729 (1.2969) [2021-04-16 00:52:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][70/1251] eta 0:06:33 lr 0.000720 time 0.3018 (0.3333) loss 4.3956 (3.6179) grad_norm 1.2847 (1.3032) [2021-04-16 00:52:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][80/1251] eta 0:06:25 lr 0.000720 time 0.2608 (0.3291) loss 3.4707 (3.6152) grad_norm 1.3996 (1.3114) [2021-04-16 00:52:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][90/1251] eta 0:06:14 lr 0.000720 time 0.2778 (0.3229) loss 3.6307 (3.5998) grad_norm 1.1650 (1.3027) [2021-04-16 00:52:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][100/1251] eta 0:06:06 lr 0.000720 time 0.2849 (0.3180) loss 3.7924 (3.5978) grad_norm 1.5787 (1.2937) [2021-04-16 00:52:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][110/1251] eta 0:05:58 lr 0.000720 time 0.2576 (0.3141) loss 4.3874 (3.6192) grad_norm 1.1534 (1.2903) [2021-04-16 00:52:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][120/1251] eta 0:05:51 lr 0.000720 time 0.2541 (0.3110) loss 4.1949 (3.6510) grad_norm 1.1484 (1.2917) [2021-04-16 00:52:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][130/1251] eta 0:05:46 lr 0.000720 time 0.2653 (0.3094) loss 3.9564 (3.6688) grad_norm 1.0967 (1.2854) [2021-04-16 00:52:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][140/1251] eta 0:05:42 lr 0.000720 time 0.2781 (0.3087) loss 4.1329 (3.6866) grad_norm 1.2330 (1.2823) [2021-04-16 00:52:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][150/1251] eta 0:05:38 lr 0.000720 time 0.2638 (0.3076) loss 3.4982 (3.6851) grad_norm 1.5842 (1.2931) [2021-04-16 00:52:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][160/1251] eta 0:05:33 lr 0.000720 time 0.2704 (0.3059) loss 4.1369 (3.6996) grad_norm 1.3670 (1.2999) [2021-04-16 00:53:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][170/1251] eta 0:05:29 lr 0.000720 time 0.2816 (0.3049) loss 3.2265 (3.7118) grad_norm 1.1727 (1.3039) [2021-04-16 00:53:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][180/1251] eta 0:05:25 lr 0.000720 time 0.2985 (0.3036) loss 3.6683 (3.7131) grad_norm 1.2770 (1.3014) [2021-04-16 00:53:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][190/1251] eta 0:05:20 lr 0.000720 time 0.2926 (0.3021) loss 4.3098 (3.7077) grad_norm 1.3812 (1.3016) [2021-04-16 00:53:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][200/1251] eta 0:05:16 lr 0.000720 time 0.2561 (0.3007) loss 3.1172 (3.6867) grad_norm 1.2336 (1.2996) [2021-04-16 00:53:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][210/1251] eta 0:05:11 lr 0.000720 time 0.2820 (0.2996) loss 4.1247 (3.6914) grad_norm 1.4208 (1.3021) [2021-04-16 00:53:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][220/1251] eta 0:05:07 lr 0.000720 time 0.2540 (0.2982) loss 3.6272 (3.6864) grad_norm 1.2226 (1.3017) [2021-04-16 00:53:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][230/1251] eta 0:05:03 lr 0.000720 time 0.2893 (0.2974) loss 4.1084 (3.6945) grad_norm 1.1530 (1.3010) [2021-04-16 00:53:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][240/1251] eta 0:04:59 lr 0.000720 time 0.2718 (0.2967) loss 3.4258 (3.6950) grad_norm 1.1688 (1.2996) [2021-04-16 00:53:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][250/1251] eta 0:04:56 lr 0.000720 time 0.3118 (0.2962) loss 2.9003 (3.6988) grad_norm 1.1625 (1.2999) [2021-04-16 00:53:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][260/1251] eta 0:04:52 lr 0.000719 time 0.2815 (0.2954) loss 2.8928 (3.7003) grad_norm 1.2114 (1.2990) [2021-04-16 00:53:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][270/1251] eta 0:04:49 lr 0.000719 time 0.2550 (0.2948) loss 3.8654 (3.6999) grad_norm 1.2104 (1.2960) [2021-04-16 00:53:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][280/1251] eta 0:04:45 lr 0.000719 time 0.2631 (0.2941) loss 3.7057 (3.6921) grad_norm 1.2862 (1.2958) [2021-04-16 00:53:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][290/1251] eta 0:04:42 lr 0.000719 time 0.2814 (0.2936) loss 2.5838 (3.6936) grad_norm 1.1436 (1.2950) [2021-04-16 00:53:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][300/1251] eta 0:04:39 lr 0.000719 time 0.2678 (0.2937) loss 3.2322 (3.6913) grad_norm 1.2747 (1.2972) [2021-04-16 00:53:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][310/1251] eta 0:04:35 lr 0.000719 time 0.2852 (0.2930) loss 3.8669 (3.6926) grad_norm 1.1701 (1.2975) [2021-04-16 00:53:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][320/1251] eta 0:04:32 lr 0.000719 time 0.2640 (0.2928) loss 3.2860 (3.6967) grad_norm 1.2844 (1.2964) [2021-04-16 00:53:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][330/1251] eta 0:04:29 lr 0.000719 time 0.2724 (0.2922) loss 4.1054 (3.6991) grad_norm 1.5369 (1.2974) [2021-04-16 00:53:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][340/1251] eta 0:04:26 lr 0.000719 time 0.3007 (0.2920) loss 4.0596 (3.6985) grad_norm 1.2675 (1.2979) [2021-04-16 00:53:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][350/1251] eta 0:04:22 lr 0.000719 time 0.2876 (0.2917) loss 3.9402 (3.7041) grad_norm 1.2963 (1.2961) [2021-04-16 00:53:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][360/1251] eta 0:04:19 lr 0.000719 time 0.3063 (0.2912) loss 3.5810 (3.7062) grad_norm 1.4145 (1.2955) [2021-04-16 00:53:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][370/1251] eta 0:04:16 lr 0.000719 time 0.2831 (0.2908) loss 3.0976 (3.7122) grad_norm 1.4473 (1.2946) [2021-04-16 00:53:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][380/1251] eta 0:04:12 lr 0.000719 time 0.2698 (0.2905) loss 3.2381 (3.7055) grad_norm 1.0936 (1.2938) [2021-04-16 00:54:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][390/1251] eta 0:04:09 lr 0.000719 time 0.2801 (0.2902) loss 3.1828 (3.7021) grad_norm 1.2698 (1.2928) [2021-04-16 00:54:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][400/1251] eta 0:04:06 lr 0.000719 time 0.2757 (0.2901) loss 3.5934 (3.7076) grad_norm 1.3494 (1.2917) [2021-04-16 00:54:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][410/1251] eta 0:04:03 lr 0.000719 time 0.2569 (0.2899) loss 2.7503 (3.7074) grad_norm 1.3460 (1.2917) [2021-04-16 00:54:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][420/1251] eta 0:04:00 lr 0.000719 time 0.2905 (0.2897) loss 2.5176 (3.7086) grad_norm 1.5627 (1.2925) [2021-04-16 00:54:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][430/1251] eta 0:03:57 lr 0.000719 time 0.2926 (0.2894) loss 3.8509 (3.7120) grad_norm 1.3248 (1.2927) [2021-04-16 00:54:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][440/1251] eta 0:03:54 lr 0.000719 time 0.2636 (0.2892) loss 4.2470 (3.7145) grad_norm 1.2994 (1.2925) [2021-04-16 00:54:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][450/1251] eta 0:03:51 lr 0.000719 time 0.2723 (0.2888) loss 3.1991 (3.7132) grad_norm 1.2722 (1.2927) [2021-04-16 00:54:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][460/1251] eta 0:03:48 lr 0.000719 time 0.2816 (0.2885) loss 4.3266 (3.7089) grad_norm 1.5104 (1.2941) [2021-04-16 00:54:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][470/1251] eta 0:03:45 lr 0.000719 time 0.2724 (0.2883) loss 3.5578 (3.7140) grad_norm 1.4793 (1.2952) [2021-04-16 00:54:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][480/1251] eta 0:03:42 lr 0.000719 time 0.2420 (0.2880) loss 4.2412 (3.7155) grad_norm 1.2366 (1.2956) [2021-04-16 00:54:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][490/1251] eta 0:03:38 lr 0.000719 time 0.2635 (0.2878) loss 3.7499 (3.7192) grad_norm 1.3381 (1.2958) [2021-04-16 00:54:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][500/1251] eta 0:03:36 lr 0.000719 time 0.2633 (0.2876) loss 2.2793 (3.7196) grad_norm 1.2259 (1.2963) [2021-04-16 00:54:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][510/1251] eta 0:03:32 lr 0.000719 time 0.2744 (0.2874) loss 4.4279 (3.7155) grad_norm 1.2013 (1.2961) [2021-04-16 00:54:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][520/1251] eta 0:03:29 lr 0.000718 time 0.2686 (0.2872) loss 4.2223 (3.7139) grad_norm 1.1671 (1.2962) [2021-04-16 00:54:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][530/1251] eta 0:03:26 lr 0.000718 time 0.2847 (0.2871) loss 4.2514 (3.7176) grad_norm 1.3537 (1.2963) [2021-04-16 00:54:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][540/1251] eta 0:03:24 lr 0.000718 time 0.2574 (0.2871) loss 3.1982 (3.7208) grad_norm 1.1643 (1.2985) [2021-04-16 00:54:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][550/1251] eta 0:03:21 lr 0.000718 time 0.2574 (0.2868) loss 4.4083 (3.7201) grad_norm 1.1007 (1.2981) [2021-04-16 00:54:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][560/1251] eta 0:03:18 lr 0.000718 time 0.2762 (0.2867) loss 4.1311 (3.7246) grad_norm 1.1628 (1.2970) [2021-04-16 00:54:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][570/1251] eta 0:03:15 lr 0.000718 time 0.2886 (0.2866) loss 3.0167 (3.7172) grad_norm 1.2291 (1.2954) [2021-04-16 00:54:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][580/1251] eta 0:03:12 lr 0.000718 time 0.2987 (0.2868) loss 4.0939 (3.7138) grad_norm 1.4954 (1.2948) [2021-04-16 00:54:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][590/1251] eta 0:03:09 lr 0.000718 time 0.2739 (0.2868) loss 4.0567 (3.7122) grad_norm 1.1548 (1.2950) [2021-04-16 00:55:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][600/1251] eta 0:03:06 lr 0.000718 time 0.2678 (0.2866) loss 3.9691 (3.7132) grad_norm 1.2545 (1.2949) [2021-04-16 00:55:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][610/1251] eta 0:03:03 lr 0.000718 time 0.2761 (0.2864) loss 2.4875 (3.7160) grad_norm 1.3398 (1.2947) [2021-04-16 00:55:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][620/1251] eta 0:03:00 lr 0.000718 time 0.2540 (0.2863) loss 4.0818 (3.7163) grad_norm 1.4322 (1.2958) [2021-04-16 00:55:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][630/1251] eta 0:02:57 lr 0.000718 time 0.2668 (0.2861) loss 4.6738 (3.7131) grad_norm 1.2170 (1.2967) [2021-04-16 00:55:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][640/1251] eta 0:02:54 lr 0.000718 time 0.3009 (0.2860) loss 3.9229 (3.7166) grad_norm 1.2182 (1.2968) [2021-04-16 00:55:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][650/1251] eta 0:02:51 lr 0.000718 time 0.2984 (0.2861) loss 3.9041 (3.7181) grad_norm 1.2023 (1.2964) [2021-04-16 00:55:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][660/1251] eta 0:02:49 lr 0.000718 time 0.2676 (0.2860) loss 3.7898 (3.7212) grad_norm 1.1897 (1.2963) [2021-04-16 00:55:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][670/1251] eta 0:02:46 lr 0.000718 time 0.2562 (0.2859) loss 3.7951 (3.7203) grad_norm 1.4824 (1.2957) [2021-04-16 00:55:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][680/1251] eta 0:02:43 lr 0.000718 time 0.2514 (0.2857) loss 3.5053 (3.7224) grad_norm 1.3896 (1.2956) [2021-04-16 00:55:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][690/1251] eta 0:02:40 lr 0.000718 time 0.2893 (0.2856) loss 4.2081 (3.7202) grad_norm 1.0389 (1.2954) [2021-04-16 00:55:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][700/1251] eta 0:02:37 lr 0.000718 time 0.2842 (0.2856) loss 3.8829 (3.7163) grad_norm 1.2229 (1.2965) [2021-04-16 00:55:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][710/1251] eta 0:02:34 lr 0.000718 time 0.2737 (0.2854) loss 4.4367 (3.7198) grad_norm 1.3987 (1.2957) [2021-04-16 00:55:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][720/1251] eta 0:02:31 lr 0.000718 time 0.2659 (0.2857) loss 3.6061 (3.7210) grad_norm 1.2346 (1.2969) [2021-04-16 00:55:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][730/1251] eta 0:02:28 lr 0.000718 time 0.2802 (0.2855) loss 2.7490 (3.7201) grad_norm 1.6364 (1.2979) [2021-04-16 00:55:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][740/1251] eta 0:02:25 lr 0.000718 time 0.2688 (0.2855) loss 4.5455 (3.7226) grad_norm 1.7561 (1.2983) [2021-04-16 00:55:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][750/1251] eta 0:02:22 lr 0.000718 time 0.2614 (0.2854) loss 4.5997 (3.7235) grad_norm 1.2205 (1.2992) [2021-04-16 00:55:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][760/1251] eta 0:02:20 lr 0.000718 time 0.2577 (0.2853) loss 4.2639 (3.7201) grad_norm 1.2440 (1.2984) [2021-04-16 00:55:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][770/1251] eta 0:02:17 lr 0.000718 time 0.3011 (0.2852) loss 3.3949 (3.7192) grad_norm 1.1024 (1.2976) [2021-04-16 00:55:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][780/1251] eta 0:02:14 lr 0.000718 time 0.2710 (0.2851) loss 2.9576 (3.7162) grad_norm 1.3127 (1.2977) [2021-04-16 00:55:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][790/1251] eta 0:02:11 lr 0.000717 time 0.2861 (0.2851) loss 3.6405 (3.7161) grad_norm 1.3482 (1.2972) [2021-04-16 00:55:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][800/1251] eta 0:02:08 lr 0.000717 time 0.3063 (0.2850) loss 3.9833 (3.7154) grad_norm 1.3363 (1.2966) [2021-04-16 00:55:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][810/1251] eta 0:02:05 lr 0.000717 time 0.2597 (0.2849) loss 3.1319 (3.7138) grad_norm 1.1924 (1.2964) [2021-04-16 00:56:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][820/1251] eta 0:02:02 lr 0.000717 time 0.2847 (0.2848) loss 3.8991 (3.7180) grad_norm 1.1132 (1.2958) [2021-04-16 00:56:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][830/1251] eta 0:01:59 lr 0.000717 time 0.2686 (0.2847) loss 3.7633 (3.7170) grad_norm 1.2477 (1.2966) [2021-04-16 00:56:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][840/1251] eta 0:01:56 lr 0.000717 time 0.2675 (0.2846) loss 4.1400 (3.7148) grad_norm 1.1562 (1.2965) [2021-04-16 00:56:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][850/1251] eta 0:01:54 lr 0.000717 time 0.2874 (0.2845) loss 3.9619 (3.7135) grad_norm 1.0905 (1.2962) [2021-04-16 00:56:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][860/1251] eta 0:01:51 lr 0.000717 time 0.3170 (0.2844) loss 2.7632 (3.7141) grad_norm 1.1980 (1.2959) [2021-04-16 00:56:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][870/1251] eta 0:01:48 lr 0.000717 time 0.2976 (0.2843) loss 3.9199 (3.7146) grad_norm 1.3498 (1.2959) [2021-04-16 00:56:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][880/1251] eta 0:01:45 lr 0.000717 time 0.2859 (0.2844) loss 3.5507 (3.7129) grad_norm 1.4773 (1.2960) [2021-04-16 00:56:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][890/1251] eta 0:01:42 lr 0.000717 time 0.2574 (0.2844) loss 3.8576 (3.7162) grad_norm 1.1528 (1.2963) [2021-04-16 00:56:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][900/1251] eta 0:01:39 lr 0.000717 time 0.2816 (0.2843) loss 3.9781 (3.7153) grad_norm 1.3236 (1.2959) [2021-04-16 00:56:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][910/1251] eta 0:01:36 lr 0.000717 time 0.2700 (0.2842) loss 2.7791 (3.7144) grad_norm 1.3029 (1.2969) [2021-04-16 00:56:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][920/1251] eta 0:01:34 lr 0.000717 time 0.2658 (0.2841) loss 3.3407 (3.7145) grad_norm 1.0470 (1.2974) [2021-04-16 00:56:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][930/1251] eta 0:01:31 lr 0.000717 time 0.2629 (0.2840) loss 4.3696 (3.7136) grad_norm 1.1910 (1.2975) [2021-04-16 00:56:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][940/1251] eta 0:01:28 lr 0.000717 time 0.2635 (0.2840) loss 3.3838 (3.7142) grad_norm 1.3323 (1.2977) [2021-04-16 00:56:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][950/1251] eta 0:01:25 lr 0.000717 time 0.2690 (0.2840) loss 3.0273 (3.7149) grad_norm 1.2425 (1.2979) [2021-04-16 00:56:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][960/1251] eta 0:01:22 lr 0.000717 time 0.2520 (0.2839) loss 3.7945 (3.7120) grad_norm 1.1941 (1.2983) [2021-04-16 00:56:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][970/1251] eta 0:01:19 lr 0.000717 time 0.2696 (0.2838) loss 4.2385 (3.7118) grad_norm 1.5313 (1.2988) [2021-04-16 00:56:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][980/1251] eta 0:01:16 lr 0.000717 time 0.2748 (0.2837) loss 4.0975 (3.7116) grad_norm 1.6996 (1.2989) [2021-04-16 00:56:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][990/1251] eta 0:01:14 lr 0.000717 time 0.2970 (0.2836) loss 4.1904 (3.7109) grad_norm 1.2935 (1.2990) [2021-04-16 00:56:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][1000/1251] eta 0:01:11 lr 0.000717 time 0.3015 (0.2835) loss 4.3785 (3.7124) grad_norm 1.3183 (1.2986) [2021-04-16 00:56:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][1010/1251] eta 0:01:08 lr 0.000717 time 0.2841 (0.2834) loss 4.5168 (3.7117) grad_norm 1.5571 (1.2997) [2021-04-16 00:56:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][1020/1251] eta 0:01:05 lr 0.000717 time 0.2758 (0.2833) loss 3.6699 (3.7107) grad_norm 1.1899 (1.3001) [2021-04-16 00:57:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][1030/1251] eta 0:01:02 lr 0.000717 time 0.3010 (0.2832) loss 3.2252 (3.7108) grad_norm 1.0909 (1.3001) [2021-04-16 00:57:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][1040/1251] eta 0:00:59 lr 0.000717 time 0.2677 (0.2831) loss 3.9634 (3.7110) grad_norm 1.4709 (1.2996) [2021-04-16 00:57:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][1050/1251] eta 0:00:56 lr 0.000717 time 0.2613 (0.2830) loss 3.7524 (3.7111) grad_norm 1.2796 (1.2995) [2021-04-16 00:57:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][1060/1251] eta 0:00:54 lr 0.000716 time 0.2698 (0.2829) loss 3.8393 (3.7072) grad_norm 1.2454 (1.2990) [2021-04-16 00:57:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][1070/1251] eta 0:00:51 lr 0.000716 time 0.2726 (0.2829) loss 3.7132 (3.7082) grad_norm 1.3065 (1.2992) [2021-04-16 00:57:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][1080/1251] eta 0:00:48 lr 0.000716 time 0.2682 (0.2829) loss 3.6255 (3.7070) grad_norm 1.2081 (1.3005) [2021-04-16 00:57:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][1090/1251] eta 0:00:45 lr 0.000716 time 0.2786 (0.2828) loss 4.1493 (3.7073) grad_norm 1.1864 (1.3003) [2021-04-16 00:57:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][1100/1251] eta 0:00:42 lr 0.000716 time 0.2716 (0.2827) loss 3.8461 (3.7084) grad_norm 1.1632 (1.3004) [2021-04-16 00:57:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][1110/1251] eta 0:00:39 lr 0.000716 time 0.2655 (0.2827) loss 3.8045 (3.7094) grad_norm 1.3569 (1.3009) [2021-04-16 00:57:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][1120/1251] eta 0:00:37 lr 0.000716 time 0.2699 (0.2827) loss 4.4365 (3.7067) grad_norm 1.1359 (1.2999) [2021-04-16 00:57:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][1130/1251] eta 0:00:34 lr 0.000716 time 0.2667 (0.2827) loss 4.0203 (3.7070) grad_norm 1.1863 (1.2998) [2021-04-16 00:57:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][1140/1251] eta 0:00:31 lr 0.000716 time 0.2706 (0.2827) loss 4.5213 (3.7083) grad_norm 1.7079 (1.3003) [2021-04-16 00:57:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][1150/1251] eta 0:00:28 lr 0.000716 time 0.2845 (0.2828) loss 3.8437 (3.7067) grad_norm 1.4505 (1.3009) [2021-04-16 00:57:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][1160/1251] eta 0:00:25 lr 0.000716 time 0.3061 (0.2828) loss 4.1254 (3.7092) grad_norm 1.5607 (1.3012) [2021-04-16 00:57:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][1170/1251] eta 0:00:22 lr 0.000716 time 0.3147 (0.2829) loss 4.2369 (3.7075) grad_norm 1.2096 (1.3008) [2021-04-16 00:57:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][1180/1251] eta 0:00:20 lr 0.000716 time 0.2757 (0.2829) loss 3.9394 (3.7059) grad_norm 1.1995 (1.3003) [2021-04-16 00:57:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][1190/1251] eta 0:00:17 lr 0.000716 time 0.2964 (0.2828) loss 3.0707 (3.7039) grad_norm 1.1185 (1.3003) [2021-04-16 00:57:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][1200/1251] eta 0:00:14 lr 0.000716 time 0.2898 (0.2829) loss 3.1449 (3.7032) grad_norm 1.4655 (1.3002) [2021-04-16 00:57:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][1210/1251] eta 0:00:11 lr 0.000716 time 0.2762 (0.2829) loss 3.9040 (3.7054) grad_norm 1.8136 (1.3019) [2021-04-16 00:57:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][1220/1251] eta 0:00:08 lr 0.000716 time 0.2569 (0.2829) loss 4.3048 (3.7039) grad_norm 1.1868 (1.3022) [2021-04-16 00:57:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][1230/1251] eta 0:00:05 lr 0.000716 time 0.2651 (0.2829) loss 4.2276 (3.7043) grad_norm 1.2158 (1.3021) [2021-04-16 00:57:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][1240/1251] eta 0:00:03 lr 0.000716 time 0.2496 (0.2828) loss 3.8331 (3.7038) grad_norm 1.4574 (1.3022) [2021-04-16 00:58:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [107/300][1250/1251] eta 0:00:00 lr 0.000716 time 0.2483 (0.2825) loss 3.6671 (3.7035) grad_norm 1.3749 (1.3026) [2021-04-16 00:58:05 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 107 training takes 0:05:56 [2021-04-16 00:58:05 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_107.pth saving...... [2021-04-16 00:58:21 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_107.pth saved !!! [2021-04-16 00:58:22 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.065 (1.065) Loss 1.1044 (1.1044) Acc@1 73.730 (73.730) Acc@5 92.969 (92.969) [2021-04-16 00:58:24 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.324 (0.283) Loss 1.1500 (1.1124) Acc@1 73.438 (73.952) Acc@5 91.504 (92.072) [2021-04-16 00:58:26 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.134 (0.220) Loss 1.0920 (1.1127) Acc@1 74.609 (73.916) Acc@5 92.090 (92.081) [2021-04-16 00:58:28 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.147 (0.214) Loss 1.0567 (1.1211) Acc@1 75.684 (73.787) Acc@5 93.262 (92.071) [2021-04-16 00:58:30 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.092 (0.221) Loss 1.1513 (1.1186) Acc@1 72.754 (73.916) Acc@5 91.797 (92.104) [2021-04-16 00:58:34 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 73.998 Acc@5 92.132 [2021-04-16 00:58:34 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 74.0% [2021-04-16 00:58:34 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 74.16% [2021-04-16 00:58:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][0/1251] eta 1:46:23 lr 0.000716 time 5.1029 (5.1029) loss 3.7629 (3.7629) grad_norm 1.4448 (1.4448) [2021-04-16 00:58:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][10/1251] eta 0:14:41 lr 0.000716 time 0.2897 (0.7106) loss 3.8187 (3.6595) grad_norm 1.2600 (1.3651) [2021-04-16 00:58:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][20/1251] eta 0:10:20 lr 0.000716 time 0.2730 (0.5042) loss 3.7277 (3.8102) grad_norm 1.2978 (1.3532) [2021-04-16 00:58:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][30/1251] eta 0:08:47 lr 0.000716 time 0.2671 (0.4323) loss 3.9492 (3.7705) grad_norm 1.3075 (1.3311) [2021-04-16 00:58:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][40/1251] eta 0:08:01 lr 0.000716 time 0.2830 (0.3973) loss 4.1322 (3.7759) grad_norm 1.2790 (1.3092) [2021-04-16 00:58:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][50/1251] eta 0:07:30 lr 0.000716 time 0.2640 (0.3748) loss 3.3159 (3.7069) grad_norm 1.4648 (1.3150) [2021-04-16 00:58:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][60/1251] eta 0:07:07 lr 0.000716 time 0.3054 (0.3591) loss 4.1873 (3.6958) grad_norm 1.6085 (1.3061) [2021-04-16 00:58:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][70/1251] eta 0:06:51 lr 0.000715 time 0.2893 (0.3487) loss 2.7329 (3.7108) grad_norm 1.1377 (1.2983) [2021-04-16 00:59:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][80/1251] eta 0:06:39 lr 0.000715 time 0.2949 (0.3415) loss 4.4717 (3.7080) grad_norm 1.2304 (1.2926) [2021-04-16 00:59:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][90/1251] eta 0:06:29 lr 0.000715 time 0.2511 (0.3353) loss 4.0852 (3.7248) grad_norm 1.3329 (1.2911) [2021-04-16 00:59:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][100/1251] eta 0:06:21 lr 0.000715 time 0.2702 (0.3313) loss 4.1181 (3.7512) grad_norm 1.2429 (1.2921) [2021-04-16 00:59:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][110/1251] eta 0:06:12 lr 0.000715 time 0.2705 (0.3262) loss 4.1355 (3.7506) grad_norm 1.2617 (1.2888) [2021-04-16 00:59:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][120/1251] eta 0:06:04 lr 0.000715 time 0.2724 (0.3220) loss 3.1569 (3.7446) grad_norm 1.3601 (1.2935) [2021-04-16 00:59:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][130/1251] eta 0:05:57 lr 0.000715 time 0.2704 (0.3185) loss 4.1864 (3.7297) grad_norm 1.1635 (1.2970) [2021-04-16 00:59:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][140/1251] eta 0:05:50 lr 0.000715 time 0.2710 (0.3157) loss 3.5805 (3.7149) grad_norm 1.2398 (1.2988) [2021-04-16 00:59:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][150/1251] eta 0:05:44 lr 0.000715 time 0.2876 (0.3133) loss 3.7662 (3.7272) grad_norm 1.2766 (inf) [2021-04-16 00:59:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][160/1251] eta 0:05:39 lr 0.000715 time 0.2749 (0.3111) loss 3.6912 (3.7304) grad_norm 1.2712 (inf) [2021-04-16 00:59:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][170/1251] eta 0:05:34 lr 0.000715 time 0.2816 (0.3093) loss 3.9884 (3.7194) grad_norm 1.3618 (inf) [2021-04-16 00:59:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][180/1251] eta 0:05:29 lr 0.000715 time 0.2692 (0.3073) loss 4.3195 (3.7389) grad_norm 1.3036 (inf) [2021-04-16 00:59:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][190/1251] eta 0:05:24 lr 0.000715 time 0.2889 (0.3061) loss 2.8817 (3.7270) grad_norm 1.4214 (inf) [2021-04-16 00:59:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][200/1251] eta 0:05:20 lr 0.000715 time 0.2717 (0.3046) loss 3.6705 (3.7078) grad_norm 1.2406 (inf) [2021-04-16 00:59:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][210/1251] eta 0:05:15 lr 0.000715 time 0.2606 (0.3035) loss 3.5320 (3.7042) grad_norm 1.5421 (inf) [2021-04-16 00:59:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][220/1251] eta 0:05:12 lr 0.000715 time 0.2878 (0.3026) loss 2.8048 (3.6948) grad_norm 1.5186 (inf) [2021-04-16 00:59:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][230/1251] eta 0:05:07 lr 0.000715 time 0.2666 (0.3016) loss 3.6671 (3.6999) grad_norm 1.7151 (inf) [2021-04-16 00:59:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][240/1251] eta 0:05:03 lr 0.000715 time 0.2868 (0.3006) loss 2.8217 (3.6879) grad_norm 1.5141 (inf) [2021-04-16 00:59:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][250/1251] eta 0:05:00 lr 0.000715 time 0.2822 (0.2997) loss 4.3631 (3.6928) grad_norm 1.2364 (inf) [2021-04-16 00:59:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][260/1251] eta 0:04:56 lr 0.000715 time 0.2814 (0.2991) loss 4.3529 (3.6992) grad_norm 1.4508 (inf) [2021-04-16 00:59:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][270/1251] eta 0:04:52 lr 0.000715 time 0.2976 (0.2986) loss 3.3242 (3.7020) grad_norm 1.1972 (inf) [2021-04-16 00:59:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][280/1251] eta 0:04:48 lr 0.000715 time 0.2685 (0.2976) loss 3.0802 (3.7035) grad_norm 1.1491 (inf) [2021-04-16 01:00:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][290/1251] eta 0:04:45 lr 0.000715 time 0.2565 (0.2970) loss 3.8087 (3.6993) grad_norm 1.3756 (inf) [2021-04-16 01:00:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][300/1251] eta 0:04:41 lr 0.000715 time 0.2703 (0.2963) loss 2.5927 (3.7047) grad_norm 1.3257 (inf) [2021-04-16 01:00:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][310/1251] eta 0:04:38 lr 0.000715 time 0.2626 (0.2957) loss 4.1828 (3.7089) grad_norm 1.3262 (inf) [2021-04-16 01:00:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][320/1251] eta 0:04:35 lr 0.000715 time 0.2735 (0.2956) loss 4.1061 (3.7052) grad_norm 1.3959 (inf) [2021-04-16 01:00:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][330/1251] eta 0:04:31 lr 0.000715 time 0.2988 (0.2951) loss 4.1846 (3.6974) grad_norm 1.3031 (inf) [2021-04-16 01:00:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][340/1251] eta 0:04:28 lr 0.000714 time 0.2643 (0.2947) loss 3.9752 (3.6935) grad_norm 1.2977 (inf) [2021-04-16 01:00:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][350/1251] eta 0:04:25 lr 0.000714 time 0.2708 (0.2943) loss 4.1636 (3.6999) grad_norm 1.3263 (inf) [2021-04-16 01:00:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][360/1251] eta 0:04:22 lr 0.000714 time 0.2701 (0.2942) loss 4.0173 (3.7078) grad_norm 1.2959 (inf) [2021-04-16 01:00:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][370/1251] eta 0:04:18 lr 0.000714 time 0.2685 (0.2938) loss 2.7823 (3.7011) grad_norm 1.2888 (inf) [2021-04-16 01:00:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][380/1251] eta 0:04:15 lr 0.000714 time 0.2700 (0.2936) loss 3.1549 (3.6930) grad_norm 1.3800 (inf) [2021-04-16 01:00:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][390/1251] eta 0:04:12 lr 0.000714 time 0.2633 (0.2932) loss 3.9494 (3.6902) grad_norm 1.2261 (inf) [2021-04-16 01:00:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][400/1251] eta 0:04:09 lr 0.000714 time 0.2675 (0.2927) loss 2.7393 (3.6937) grad_norm 1.3690 (inf) [2021-04-16 01:00:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][410/1251] eta 0:04:05 lr 0.000714 time 0.2595 (0.2924) loss 3.9270 (3.7003) grad_norm 1.5284 (inf) [2021-04-16 01:00:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][420/1251] eta 0:04:02 lr 0.000714 time 0.2660 (0.2920) loss 2.7470 (3.6948) grad_norm 1.2952 (inf) [2021-04-16 01:00:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][430/1251] eta 0:03:59 lr 0.000714 time 0.2485 (0.2920) loss 3.0530 (3.6962) grad_norm 1.2102 (inf) [2021-04-16 01:00:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][440/1251] eta 0:03:56 lr 0.000714 time 0.2780 (0.2916) loss 3.0803 (3.6957) grad_norm 1.1734 (inf) [2021-04-16 01:00:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][450/1251] eta 0:03:53 lr 0.000714 time 0.2653 (0.2912) loss 3.2710 (3.6981) grad_norm 1.3359 (inf) [2021-04-16 01:00:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][460/1251] eta 0:03:50 lr 0.000714 time 0.2879 (0.2909) loss 3.9503 (3.7041) grad_norm 1.3664 (inf) [2021-04-16 01:00:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][470/1251] eta 0:03:47 lr 0.000714 time 0.2790 (0.2909) loss 3.9757 (3.7010) grad_norm 1.2938 (inf) [2021-04-16 01:00:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][480/1251] eta 0:03:44 lr 0.000714 time 0.2581 (0.2907) loss 4.5081 (3.7013) grad_norm 1.3131 (inf) [2021-04-16 01:00:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][490/1251] eta 0:03:40 lr 0.000714 time 0.2646 (0.2903) loss 2.5499 (3.7024) grad_norm 1.2716 (inf) [2021-04-16 01:01:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][500/1251] eta 0:03:37 lr 0.000714 time 0.2584 (0.2901) loss 3.8971 (3.7080) grad_norm 1.3887 (inf) [2021-04-16 01:01:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][510/1251] eta 0:03:34 lr 0.000714 time 0.2935 (0.2898) loss 4.3099 (3.7085) grad_norm 1.3085 (inf) [2021-04-16 01:01:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][520/1251] eta 0:03:31 lr 0.000714 time 0.2806 (0.2898) loss 4.7058 (3.7085) grad_norm 1.3365 (inf) [2021-04-16 01:01:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][530/1251] eta 0:03:28 lr 0.000714 time 0.2866 (0.2896) loss 3.3124 (3.7010) grad_norm 1.2758 (inf) [2021-04-16 01:01:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][540/1251] eta 0:03:25 lr 0.000714 time 0.2842 (0.2894) loss 4.1051 (3.6996) grad_norm 1.4632 (inf) [2021-04-16 01:01:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][550/1251] eta 0:03:22 lr 0.000714 time 0.2792 (0.2893) loss 3.6766 (3.6972) grad_norm 1.2167 (inf) [2021-04-16 01:01:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][560/1251] eta 0:03:19 lr 0.000714 time 0.2688 (0.2892) loss 3.4920 (3.6998) grad_norm 1.5166 (inf) [2021-04-16 01:01:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][570/1251] eta 0:03:17 lr 0.000714 time 0.2611 (0.2895) loss 3.9294 (3.7023) grad_norm 1.3722 (inf) [2021-04-16 01:01:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][580/1251] eta 0:03:14 lr 0.000714 time 0.2764 (0.2896) loss 4.3808 (3.7016) grad_norm 1.3042 (inf) [2021-04-16 01:01:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][590/1251] eta 0:03:11 lr 0.000714 time 0.3016 (0.2895) loss 3.8700 (3.7018) grad_norm 1.4475 (inf) [2021-04-16 01:01:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][600/1251] eta 0:03:08 lr 0.000714 time 0.2935 (0.2893) loss 3.6859 (3.7046) grad_norm 1.1692 (inf) [2021-04-16 01:01:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][610/1251] eta 0:03:05 lr 0.000713 time 0.2920 (0.2891) loss 4.3523 (3.7022) grad_norm 1.1995 (inf) [2021-04-16 01:01:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][620/1251] eta 0:03:02 lr 0.000713 time 0.2845 (0.2890) loss 4.2551 (3.7013) grad_norm 1.3363 (inf) [2021-04-16 01:01:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][630/1251] eta 0:02:59 lr 0.000713 time 0.2663 (0.2888) loss 4.0263 (3.7013) grad_norm 1.2843 (inf) [2021-04-16 01:01:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][640/1251] eta 0:02:56 lr 0.000713 time 0.2556 (0.2887) loss 4.0287 (3.7018) grad_norm 1.3587 (inf) [2021-04-16 01:01:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][650/1251] eta 0:02:53 lr 0.000713 time 0.3713 (0.2887) loss 4.3222 (3.7023) grad_norm 1.2084 (inf) [2021-04-16 01:01:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][660/1251] eta 0:02:50 lr 0.000713 time 0.2422 (0.2885) loss 4.0879 (3.7010) grad_norm 1.1434 (inf) [2021-04-16 01:01:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][670/1251] eta 0:02:47 lr 0.000713 time 0.2661 (0.2885) loss 4.2959 (3.7028) grad_norm 1.3305 (inf) [2021-04-16 01:01:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][680/1251] eta 0:02:44 lr 0.000713 time 0.2543 (0.2885) loss 3.9585 (3.7016) grad_norm 1.1806 (inf) [2021-04-16 01:01:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][690/1251] eta 0:02:41 lr 0.000713 time 0.2870 (0.2884) loss 4.3654 (3.7036) grad_norm 1.1459 (inf) [2021-04-16 01:01:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][700/1251] eta 0:02:38 lr 0.000713 time 0.2690 (0.2882) loss 4.4208 (3.6999) grad_norm 1.3164 (inf) [2021-04-16 01:01:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][710/1251] eta 0:02:35 lr 0.000713 time 0.2859 (0.2881) loss 3.4234 (3.6992) grad_norm 1.5147 (inf) [2021-04-16 01:02:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][720/1251] eta 0:02:32 lr 0.000713 time 0.2915 (0.2879) loss 3.6398 (3.6995) grad_norm 1.3492 (inf) [2021-04-16 01:02:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][730/1251] eta 0:02:29 lr 0.000713 time 0.3132 (0.2877) loss 4.0733 (3.7024) grad_norm 1.5797 (inf) [2021-04-16 01:02:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][740/1251] eta 0:02:26 lr 0.000713 time 0.2720 (0.2876) loss 4.3048 (3.7012) grad_norm 1.2057 (inf) [2021-04-16 01:02:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][750/1251] eta 0:02:24 lr 0.000713 time 0.2793 (0.2875) loss 4.2133 (3.7016) grad_norm 1.2477 (inf) [2021-04-16 01:02:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][760/1251] eta 0:02:21 lr 0.000713 time 0.2676 (0.2874) loss 3.7310 (3.7040) grad_norm 1.4374 (inf) [2021-04-16 01:02:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][770/1251] eta 0:02:18 lr 0.000713 time 0.2698 (0.2872) loss 3.8457 (3.7054) grad_norm 1.0629 (inf) [2021-04-16 01:02:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][780/1251] eta 0:02:15 lr 0.000713 time 0.2873 (0.2871) loss 2.5098 (3.7045) grad_norm 1.4217 (inf) [2021-04-16 01:02:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][790/1251] eta 0:02:12 lr 0.000713 time 0.2752 (0.2871) loss 4.0587 (3.7050) grad_norm 1.6579 (inf) [2021-04-16 01:02:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][800/1251] eta 0:02:09 lr 0.000713 time 0.2841 (0.2870) loss 2.9274 (3.7056) grad_norm 1.2376 (inf) [2021-04-16 01:02:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][810/1251] eta 0:02:06 lr 0.000713 time 0.2948 (0.2869) loss 3.8089 (3.7035) grad_norm 1.1916 (inf) [2021-04-16 01:02:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][820/1251] eta 0:02:03 lr 0.000713 time 0.3107 (0.2868) loss 4.3971 (3.7027) grad_norm 1.3056 (inf) [2021-04-16 01:02:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][830/1251] eta 0:02:00 lr 0.000713 time 0.2770 (0.2867) loss 4.4655 (3.7036) grad_norm 1.3673 (inf) [2021-04-16 01:02:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][840/1251] eta 0:01:57 lr 0.000713 time 0.2540 (0.2866) loss 2.8997 (3.7030) grad_norm 1.3422 (inf) [2021-04-16 01:02:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][850/1251] eta 0:01:54 lr 0.000713 time 0.2575 (0.2865) loss 4.5682 (3.7038) grad_norm 1.3580 (inf) [2021-04-16 01:02:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][860/1251] eta 0:01:51 lr 0.000713 time 0.2658 (0.2864) loss 3.5566 (3.7039) grad_norm 1.1299 (inf) [2021-04-16 01:02:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][870/1251] eta 0:01:49 lr 0.000712 time 0.2604 (0.2862) loss 3.6757 (3.7065) grad_norm 1.1132 (inf) [2021-04-16 01:02:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][880/1251] eta 0:01:46 lr 0.000712 time 0.3115 (0.2861) loss 3.5054 (3.7016) grad_norm 1.2421 (inf) [2021-04-16 01:02:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][890/1251] eta 0:01:43 lr 0.000712 time 0.2883 (0.2862) loss 2.5800 (3.7018) grad_norm 1.3448 (inf) [2021-04-16 01:02:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][900/1251] eta 0:01:40 lr 0.000712 time 0.2594 (0.2860) loss 2.6140 (3.7005) grad_norm 1.4186 (inf) [2021-04-16 01:02:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][910/1251] eta 0:01:37 lr 0.000712 time 0.2843 (0.2860) loss 3.8268 (3.6993) grad_norm 1.2436 (inf) [2021-04-16 01:02:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][920/1251] eta 0:01:34 lr 0.000712 time 0.2541 (0.2860) loss 3.6194 (3.6989) grad_norm 1.1108 (inf) [2021-04-16 01:03:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][930/1251] eta 0:01:31 lr 0.000712 time 0.2741 (0.2859) loss 3.5895 (3.6988) grad_norm 1.1890 (inf) [2021-04-16 01:03:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][940/1251] eta 0:01:28 lr 0.000712 time 0.2762 (0.2859) loss 4.1155 (3.7000) grad_norm 1.6454 (inf) [2021-04-16 01:03:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][950/1251] eta 0:01:26 lr 0.000712 time 0.4267 (0.2860) loss 3.6758 (3.6984) grad_norm 1.4743 (inf) [2021-04-16 01:03:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][960/1251] eta 0:01:23 lr 0.000712 time 0.2862 (0.2859) loss 3.1714 (3.6963) grad_norm 1.3029 (inf) [2021-04-16 01:03:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][970/1251] eta 0:01:20 lr 0.000712 time 0.2932 (0.2859) loss 3.1007 (3.6962) grad_norm 1.0484 (inf) [2021-04-16 01:03:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][980/1251] eta 0:01:17 lr 0.000712 time 0.2866 (0.2859) loss 3.9626 (3.6988) grad_norm 1.3021 (inf) [2021-04-16 01:03:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][990/1251] eta 0:01:14 lr 0.000712 time 0.2427 (0.2857) loss 3.1848 (3.6982) grad_norm 1.2994 (inf) [2021-04-16 01:03:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][1000/1251] eta 0:01:11 lr 0.000712 time 0.2768 (0.2856) loss 4.6179 (3.6958) grad_norm 1.4002 (inf) [2021-04-16 01:03:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][1010/1251] eta 0:01:08 lr 0.000712 time 0.2712 (0.2856) loss 3.4195 (3.6934) grad_norm 1.4083 (inf) [2021-04-16 01:03:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][1020/1251] eta 0:01:05 lr 0.000712 time 0.3010 (0.2856) loss 3.6666 (3.6936) grad_norm 1.3718 (inf) [2021-04-16 01:03:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][1030/1251] eta 0:01:03 lr 0.000712 time 0.2625 (0.2855) loss 4.4522 (3.6904) grad_norm 1.2761 (inf) [2021-04-16 01:03:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][1040/1251] eta 0:01:00 lr 0.000712 time 0.2762 (0.2855) loss 3.9984 (3.6945) grad_norm 1.2291 (inf) [2021-04-16 01:03:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][1050/1251] eta 0:00:57 lr 0.000712 time 0.2597 (0.2853) loss 3.2179 (3.6917) grad_norm 1.2087 (inf) [2021-04-16 01:03:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][1060/1251] eta 0:00:54 lr 0.000712 time 0.2692 (0.2853) loss 3.7705 (3.6928) grad_norm 1.2874 (inf) [2021-04-16 01:03:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][1070/1251] eta 0:00:51 lr 0.000712 time 0.2879 (0.2853) loss 3.2987 (3.6926) grad_norm 1.3730 (inf) [2021-04-16 01:03:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][1080/1251] eta 0:00:48 lr 0.000712 time 0.2887 (0.2852) loss 3.9055 (3.6932) grad_norm 1.5748 (inf) [2021-04-16 01:03:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][1090/1251] eta 0:00:45 lr 0.000712 time 0.2712 (0.2852) loss 3.1544 (3.6927) grad_norm 1.1574 (inf) [2021-04-16 01:03:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][1100/1251] eta 0:00:43 lr 0.000712 time 0.2813 (0.2852) loss 3.3937 (3.6948) grad_norm 1.1826 (inf) [2021-04-16 01:03:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][1110/1251] eta 0:00:40 lr 0.000712 time 0.2654 (0.2850) loss 2.8929 (3.6929) grad_norm 1.2630 (inf) [2021-04-16 01:03:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][1120/1251] eta 0:00:37 lr 0.000712 time 0.2860 (0.2850) loss 4.0078 (3.6908) grad_norm 1.2161 (inf) [2021-04-16 01:03:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][1130/1251] eta 0:00:34 lr 0.000712 time 0.2849 (0.2849) loss 3.7300 (3.6921) grad_norm 1.2682 (inf) [2021-04-16 01:03:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][1140/1251] eta 0:00:31 lr 0.000711 time 0.2662 (0.2850) loss 3.5512 (3.6909) grad_norm 1.2244 (inf) [2021-04-16 01:04:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][1150/1251] eta 0:00:28 lr 0.000711 time 0.3002 (0.2851) loss 2.9686 (3.6912) grad_norm 1.3971 (inf) [2021-04-16 01:04:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][1160/1251] eta 0:00:25 lr 0.000711 time 0.2738 (0.2851) loss 3.9239 (3.6919) grad_norm 1.2425 (inf) [2021-04-16 01:04:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][1170/1251] eta 0:00:23 lr 0.000711 time 0.2576 (0.2851) loss 4.2907 (3.6948) grad_norm 1.4367 (inf) [2021-04-16 01:04:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][1180/1251] eta 0:00:20 lr 0.000711 time 0.2899 (0.2851) loss 3.3618 (3.6943) grad_norm 1.3525 (inf) [2021-04-16 01:04:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][1190/1251] eta 0:00:17 lr 0.000711 time 0.2796 (0.2851) loss 4.3954 (3.6945) grad_norm 1.3183 (inf) [2021-04-16 01:04:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][1200/1251] eta 0:00:14 lr 0.000711 time 0.2704 (0.2851) loss 3.7002 (3.6931) grad_norm 1.2065 (inf) [2021-04-16 01:04:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][1210/1251] eta 0:00:11 lr 0.000711 time 0.2858 (0.2851) loss 4.2108 (3.6933) grad_norm 1.3771 (inf) [2021-04-16 01:04:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][1220/1251] eta 0:00:08 lr 0.000711 time 0.2473 (0.2850) loss 4.1275 (3.6954) grad_norm 1.1447 (inf) [2021-04-16 01:04:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][1230/1251] eta 0:00:05 lr 0.000711 time 0.2618 (0.2850) loss 4.2118 (3.6960) grad_norm 1.3436 (inf) [2021-04-16 01:04:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][1240/1251] eta 0:00:03 lr 0.000711 time 0.3248 (0.2849) loss 3.5330 (3.6968) grad_norm 1.4146 (inf) [2021-04-16 01:04:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [108/300][1250/1251] eta 0:00:00 lr 0.000711 time 0.2535 (0.2846) loss 3.8592 (3.6976) grad_norm 1.3737 (inf) [2021-04-16 01:04:34 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 108 training takes 0:05:59 [2021-04-16 01:04:34 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_108.pth saving...... [2021-04-16 01:04:43 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_108.pth saved !!! [2021-04-16 01:04:44 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.228 (1.228) Loss 1.0792 (1.0792) Acc@1 73.438 (73.438) Acc@5 93.359 (93.359) [2021-04-16 01:04:46 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.104 (0.231) Loss 1.0707 (1.0965) Acc@1 72.363 (73.358) Acc@5 92.773 (92.480) [2021-04-16 01:04:48 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.109 (0.239) Loss 1.1251 (1.1019) Acc@1 74.609 (73.633) Acc@5 92.090 (92.211) [2021-04-16 01:04:51 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.652 (0.242) Loss 1.0735 (1.1020) Acc@1 74.707 (73.598) Acc@5 91.211 (92.247) [2021-04-16 01:04:52 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.223) Loss 1.1551 (1.0992) Acc@1 71.875 (73.761) Acc@5 91.211 (92.268) [2021-04-16 01:04:56 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 73.880 Acc@5 92.284 [2021-04-16 01:04:56 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 73.9% [2021-04-16 01:04:56 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 74.16% [2021-04-16 01:05:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][0/1251] eta 1:07:04 lr 0.000711 time 3.2170 (3.2170) loss 3.3601 (3.3601) grad_norm 1.1905 (1.1905) [2021-04-16 01:05:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][10/1251] eta 0:11:18 lr 0.000711 time 0.2807 (0.5470) loss 3.8512 (3.6967) grad_norm 1.2911 (1.3010) [2021-04-16 01:05:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][20/1251] eta 0:08:37 lr 0.000711 time 0.2718 (0.4200) loss 3.9602 (3.6010) grad_norm 1.3497 (1.3347) [2021-04-16 01:05:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][30/1251] eta 0:07:36 lr 0.000711 time 0.2904 (0.3736) loss 3.8361 (3.6344) grad_norm 1.0722 (1.3063) [2021-04-16 01:05:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][40/1251] eta 0:07:03 lr 0.000711 time 0.2820 (0.3498) loss 3.3404 (3.5944) grad_norm 1.5284 (1.3102) [2021-04-16 01:05:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][50/1251] eta 0:06:43 lr 0.000711 time 0.2891 (0.3361) loss 4.2463 (3.6417) grad_norm 1.2558 (1.3097) [2021-04-16 01:05:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][60/1251] eta 0:06:28 lr 0.000711 time 0.2830 (0.3261) loss 3.9420 (3.6292) grad_norm 1.1947 (1.3204) [2021-04-16 01:05:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][70/1251] eta 0:06:17 lr 0.000711 time 0.2808 (0.3193) loss 2.7067 (3.6108) grad_norm 1.1557 (1.3143) [2021-04-16 01:05:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][80/1251] eta 0:06:07 lr 0.000711 time 0.2965 (0.3138) loss 3.2335 (3.6158) grad_norm 1.5178 (1.3085) [2021-04-16 01:05:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][90/1251] eta 0:06:01 lr 0.000711 time 0.2776 (0.3112) loss 4.0714 (3.6160) grad_norm 1.2586 (1.3045) [2021-04-16 01:05:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][100/1251] eta 0:05:54 lr 0.000711 time 0.2777 (0.3079) loss 4.4481 (3.6188) grad_norm 1.2453 (1.3030) [2021-04-16 01:05:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][110/1251] eta 0:05:48 lr 0.000711 time 0.2675 (0.3054) loss 2.6007 (3.6036) grad_norm 1.3952 (1.3065) [2021-04-16 01:05:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][120/1251] eta 0:05:44 lr 0.000711 time 0.2672 (0.3044) loss 4.4405 (3.6088) grad_norm 1.3267 (1.3010) [2021-04-16 01:05:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][130/1251] eta 0:05:38 lr 0.000711 time 0.2615 (0.3019) loss 4.0061 (3.6448) grad_norm 1.0923 (1.3053) [2021-04-16 01:05:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][140/1251] eta 0:05:33 lr 0.000711 time 0.2864 (0.3002) loss 3.7314 (3.6590) grad_norm 1.2756 (1.3047) [2021-04-16 01:05:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][150/1251] eta 0:05:31 lr 0.000710 time 0.2680 (0.3008) loss 3.7166 (3.6712) grad_norm 1.3743 (1.3057) [2021-04-16 01:05:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][160/1251] eta 0:05:26 lr 0.000710 time 0.2913 (0.2994) loss 3.3470 (3.6801) grad_norm 1.0724 (1.3079) [2021-04-16 01:05:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][170/1251] eta 0:05:22 lr 0.000710 time 0.2587 (0.2982) loss 3.9774 (3.6879) grad_norm 1.2973 (1.3153) [2021-04-16 01:05:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][180/1251] eta 0:05:18 lr 0.000710 time 0.2706 (0.2974) loss 4.7232 (3.6792) grad_norm 1.2754 (1.3162) [2021-04-16 01:05:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][190/1251] eta 0:05:14 lr 0.000710 time 0.2801 (0.2962) loss 3.4239 (3.6825) grad_norm 1.4372 (1.3122) [2021-04-16 01:05:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][200/1251] eta 0:05:10 lr 0.000710 time 0.2862 (0.2951) loss 2.6682 (3.6721) grad_norm 1.2033 (1.3097) [2021-04-16 01:05:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][210/1251] eta 0:05:06 lr 0.000710 time 0.2753 (0.2943) loss 2.4552 (3.6861) grad_norm 1.1019 (1.3082) [2021-04-16 01:06:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][220/1251] eta 0:05:03 lr 0.000710 time 0.3000 (0.2941) loss 4.1055 (3.6926) grad_norm 1.1623 (1.3079) [2021-04-16 01:06:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][230/1251] eta 0:04:59 lr 0.000710 time 0.3174 (0.2934) loss 4.3184 (3.6797) grad_norm 1.2054 (1.3102) [2021-04-16 01:06:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][240/1251] eta 0:04:55 lr 0.000710 time 0.2515 (0.2924) loss 4.3476 (3.6701) grad_norm 1.2206 (1.3065) [2021-04-16 01:06:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][250/1251] eta 0:04:52 lr 0.000710 time 0.2873 (0.2919) loss 3.7183 (3.6717) grad_norm 1.7604 (1.3089) [2021-04-16 01:06:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][260/1251] eta 0:04:48 lr 0.000710 time 0.2539 (0.2912) loss 3.6774 (3.6696) grad_norm 1.2456 (1.3073) [2021-04-16 01:06:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][270/1251] eta 0:04:45 lr 0.000710 time 0.2647 (0.2908) loss 2.6862 (3.6531) grad_norm 1.2424 (1.3050) [2021-04-16 01:06:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][280/1251] eta 0:04:41 lr 0.000710 time 0.2709 (0.2902) loss 4.0077 (3.6640) grad_norm 1.1769 (1.3057) [2021-04-16 01:06:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][290/1251] eta 0:04:39 lr 0.000710 time 0.2917 (0.2905) loss 2.5432 (3.6611) grad_norm 1.0378 (1.3029) [2021-04-16 01:06:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][300/1251] eta 0:04:35 lr 0.000710 time 0.2828 (0.2900) loss 2.9915 (3.6655) grad_norm 1.1657 (1.2999) [2021-04-16 01:06:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][310/1251] eta 0:04:32 lr 0.000710 time 0.2715 (0.2895) loss 3.9479 (3.6725) grad_norm 1.3862 (1.3006) [2021-04-16 01:06:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][320/1251] eta 0:04:29 lr 0.000710 time 0.2873 (0.2893) loss 3.4773 (3.6766) grad_norm 1.1824 (1.3008) [2021-04-16 01:06:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][330/1251] eta 0:04:26 lr 0.000710 time 0.2987 (0.2889) loss 4.1184 (3.6846) grad_norm 1.2488 (1.3014) [2021-04-16 01:06:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][340/1251] eta 0:04:23 lr 0.000710 time 0.2622 (0.2887) loss 4.4823 (3.6901) grad_norm 1.2672 (1.3007) [2021-04-16 01:06:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][350/1251] eta 0:04:20 lr 0.000710 time 0.2550 (0.2887) loss 3.9782 (3.6937) grad_norm 1.2732 (1.2992) [2021-04-16 01:06:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][360/1251] eta 0:04:17 lr 0.000710 time 0.2779 (0.2887) loss 3.6915 (3.6963) grad_norm 1.4285 (1.2968) [2021-04-16 01:06:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][370/1251] eta 0:04:14 lr 0.000710 time 0.2931 (0.2886) loss 4.1440 (3.6999) grad_norm 1.3691 (1.2972) [2021-04-16 01:06:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][380/1251] eta 0:04:11 lr 0.000710 time 0.2888 (0.2883) loss 3.9575 (3.7045) grad_norm 1.3034 (1.2983) [2021-04-16 01:06:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][390/1251] eta 0:04:08 lr 0.000710 time 0.3037 (0.2882) loss 4.3185 (3.7029) grad_norm 1.2734 (1.3006) [2021-04-16 01:06:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][400/1251] eta 0:04:04 lr 0.000710 time 0.2721 (0.2878) loss 3.8639 (3.7076) grad_norm 1.3139 (1.3017) [2021-04-16 01:06:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][410/1251] eta 0:04:01 lr 0.000710 time 0.2701 (0.2874) loss 3.0068 (3.6963) grad_norm 1.2053 (1.3000) [2021-04-16 01:06:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][420/1251] eta 0:03:58 lr 0.000709 time 0.2918 (0.2873) loss 3.6508 (3.6902) grad_norm 1.2506 (1.2993) [2021-04-16 01:07:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][430/1251] eta 0:03:55 lr 0.000709 time 0.2768 (0.2872) loss 2.8708 (3.6925) grad_norm 1.1401 (1.2989) [2021-04-16 01:07:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][440/1251] eta 0:03:52 lr 0.000709 time 0.2538 (0.2869) loss 4.0364 (3.6993) grad_norm 1.2328 (1.3001) [2021-04-16 01:07:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][450/1251] eta 0:03:49 lr 0.000709 time 0.2631 (0.2869) loss 3.8083 (3.7021) grad_norm 1.3246 (1.3032) [2021-04-16 01:07:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][460/1251] eta 0:03:46 lr 0.000709 time 0.2771 (0.2868) loss 4.2233 (3.6994) grad_norm 1.1718 (1.3027) [2021-04-16 01:07:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][470/1251] eta 0:03:44 lr 0.000709 time 0.2743 (0.2869) loss 4.4683 (3.6943) grad_norm 1.2790 (1.3036) [2021-04-16 01:07:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][480/1251] eta 0:03:41 lr 0.000709 time 0.2767 (0.2867) loss 2.4412 (3.6859) grad_norm 1.2943 (1.3045) [2021-04-16 01:07:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][490/1251] eta 0:03:38 lr 0.000709 time 0.2674 (0.2866) loss 3.8708 (3.6815) grad_norm 1.1668 (1.3062) [2021-04-16 01:07:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][500/1251] eta 0:03:35 lr 0.000709 time 0.2951 (0.2864) loss 3.9242 (3.6847) grad_norm 1.2998 (1.3068) [2021-04-16 01:07:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][510/1251] eta 0:03:32 lr 0.000709 time 0.2652 (0.2863) loss 3.1942 (3.6881) grad_norm 1.4377 (1.3059) [2021-04-16 01:07:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][520/1251] eta 0:03:29 lr 0.000709 time 0.2801 (0.2861) loss 4.2566 (3.6929) grad_norm 1.3743 (1.3059) [2021-04-16 01:07:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][530/1251] eta 0:03:26 lr 0.000709 time 0.2760 (0.2860) loss 3.9493 (3.6900) grad_norm 1.1549 (1.3045) [2021-04-16 01:07:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][540/1251] eta 0:03:23 lr 0.000709 time 0.2745 (0.2859) loss 3.4611 (3.6930) grad_norm 1.2785 (1.3032) [2021-04-16 01:07:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][550/1251] eta 0:03:20 lr 0.000709 time 0.2822 (0.2858) loss 3.3790 (3.6883) grad_norm 1.3184 (1.3031) [2021-04-16 01:07:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][560/1251] eta 0:03:17 lr 0.000709 time 0.2590 (0.2856) loss 2.5665 (3.6873) grad_norm 1.0826 (1.3032) [2021-04-16 01:07:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][570/1251] eta 0:03:14 lr 0.000709 time 0.2783 (0.2855) loss 4.4025 (3.6845) grad_norm 1.1390 (1.3022) [2021-04-16 01:07:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][580/1251] eta 0:03:11 lr 0.000709 time 0.2871 (0.2853) loss 4.5354 (3.6864) grad_norm 1.1556 (1.3007) [2021-04-16 01:07:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][590/1251] eta 0:03:08 lr 0.000709 time 0.2595 (0.2852) loss 3.2073 (3.6816) grad_norm 1.2059 (1.3011) [2021-04-16 01:07:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][600/1251] eta 0:03:05 lr 0.000709 time 0.2890 (0.2851) loss 3.1335 (3.6817) grad_norm 1.0800 (1.2999) [2021-04-16 01:07:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][610/1251] eta 0:03:02 lr 0.000709 time 0.2451 (0.2849) loss 3.6916 (3.6864) grad_norm 1.2319 (1.2995) [2021-04-16 01:07:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][620/1251] eta 0:02:59 lr 0.000709 time 0.2777 (0.2848) loss 3.8833 (3.6861) grad_norm 1.1919 (1.2988) [2021-04-16 01:07:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][630/1251] eta 0:02:56 lr 0.000709 time 0.2617 (0.2846) loss 2.6789 (3.6824) grad_norm 1.2454 (1.2996) [2021-04-16 01:07:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][640/1251] eta 0:02:53 lr 0.000709 time 0.2618 (0.2845) loss 2.1792 (3.6782) grad_norm 1.2740 (1.3005) [2021-04-16 01:08:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][650/1251] eta 0:02:50 lr 0.000709 time 0.2467 (0.2844) loss 4.1377 (3.6804) grad_norm 1.2991 (1.3011) [2021-04-16 01:08:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][660/1251] eta 0:02:48 lr 0.000709 time 0.2878 (0.2843) loss 4.4313 (3.6812) grad_norm 1.1906 (1.3014) [2021-04-16 01:08:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][670/1251] eta 0:02:45 lr 0.000709 time 0.2446 (0.2845) loss 3.7271 (3.6858) grad_norm 1.2982 (1.3012) [2021-04-16 01:08:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][680/1251] eta 0:02:42 lr 0.000708 time 0.2651 (0.2844) loss 4.6291 (3.6830) grad_norm 1.5149 (1.3015) [2021-04-16 01:08:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][690/1251] eta 0:02:39 lr 0.000708 time 0.2726 (0.2844) loss 4.2621 (3.6893) grad_norm 1.3055 (1.3016) [2021-04-16 01:08:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][700/1251] eta 0:02:36 lr 0.000708 time 0.2789 (0.2843) loss 3.4829 (3.6910) grad_norm 1.3994 (1.3042) [2021-04-16 01:08:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][710/1251] eta 0:02:33 lr 0.000708 time 0.2680 (0.2843) loss 3.6234 (3.6947) grad_norm 1.2847 (1.3042) [2021-04-16 01:08:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][720/1251] eta 0:02:30 lr 0.000708 time 0.3071 (0.2842) loss 3.2725 (3.6974) grad_norm 1.3371 (1.3056) [2021-04-16 01:08:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][730/1251] eta 0:02:28 lr 0.000708 time 0.2556 (0.2844) loss 4.3434 (3.6998) grad_norm 1.2651 (1.3059) [2021-04-16 01:08:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][740/1251] eta 0:02:25 lr 0.000708 time 0.2692 (0.2842) loss 3.7449 (3.6973) grad_norm 1.4369 (1.3066) [2021-04-16 01:08:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][750/1251] eta 0:02:22 lr 0.000708 time 0.2777 (0.2841) loss 4.0563 (3.6991) grad_norm 1.3555 (1.3061) [2021-04-16 01:08:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][760/1251] eta 0:02:19 lr 0.000708 time 0.2650 (0.2840) loss 2.5629 (3.6999) grad_norm 1.2666 (1.3059) [2021-04-16 01:08:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][770/1251] eta 0:02:16 lr 0.000708 time 0.2785 (0.2839) loss 4.1510 (3.7004) grad_norm 1.7811 (1.3065) [2021-04-16 01:08:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][780/1251] eta 0:02:13 lr 0.000708 time 0.2704 (0.2838) loss 3.0426 (3.7021) grad_norm 1.3441 (1.3067) [2021-04-16 01:08:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][790/1251] eta 0:02:10 lr 0.000708 time 0.2753 (0.2839) loss 2.7765 (3.7023) grad_norm 1.2304 (1.3067) [2021-04-16 01:08:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][800/1251] eta 0:02:07 lr 0.000708 time 0.2719 (0.2837) loss 3.6093 (3.7019) grad_norm 1.2208 (1.3056) [2021-04-16 01:08:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][810/1251] eta 0:02:05 lr 0.000708 time 0.2584 (0.2837) loss 2.7228 (3.7011) grad_norm 1.3717 (1.3061) [2021-04-16 01:08:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][820/1251] eta 0:02:02 lr 0.000708 time 0.2618 (0.2837) loss 3.8744 (3.7022) grad_norm 1.1189 (1.3065) [2021-04-16 01:08:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][830/1251] eta 0:01:59 lr 0.000708 time 0.2715 (0.2835) loss 4.0189 (3.6997) grad_norm 1.3220 (1.3087) [2021-04-16 01:08:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][840/1251] eta 0:01:56 lr 0.000708 time 0.2707 (0.2835) loss 4.3749 (3.6964) grad_norm 1.3983 (1.3086) [2021-04-16 01:08:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][850/1251] eta 0:01:53 lr 0.000708 time 0.2648 (0.2834) loss 3.7893 (3.6943) grad_norm 1.2707 (1.3082) [2021-04-16 01:09:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][860/1251] eta 0:01:50 lr 0.000708 time 0.2553 (0.2833) loss 3.4181 (3.6951) grad_norm 1.4797 (1.3083) [2021-04-16 01:09:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][870/1251] eta 0:01:47 lr 0.000708 time 0.2643 (0.2833) loss 3.9062 (3.6946) grad_norm 1.4579 (1.3093) [2021-04-16 01:09:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][880/1251] eta 0:01:45 lr 0.000708 time 0.2571 (0.2833) loss 4.4090 (3.6978) grad_norm 1.1714 (1.3096) [2021-04-16 01:09:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][890/1251] eta 0:01:42 lr 0.000708 time 0.2644 (0.2832) loss 3.0055 (3.6936) grad_norm 1.4731 (1.3099) [2021-04-16 01:09:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][900/1251] eta 0:01:39 lr 0.000708 time 0.2815 (0.2832) loss 4.4531 (3.6987) grad_norm 1.3196 (1.3104) [2021-04-16 01:09:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][910/1251] eta 0:01:36 lr 0.000708 time 0.2475 (0.2831) loss 2.9591 (3.6981) grad_norm 1.2302 (1.3104) [2021-04-16 01:09:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][920/1251] eta 0:01:33 lr 0.000708 time 0.2732 (0.2832) loss 3.8995 (3.6968) grad_norm 1.2242 (1.3094) [2021-04-16 01:09:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][930/1251] eta 0:01:30 lr 0.000708 time 0.2407 (0.2834) loss 4.1135 (3.6954) grad_norm 1.3267 (1.3089) [2021-04-16 01:09:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][940/1251] eta 0:01:28 lr 0.000708 time 0.2748 (0.2834) loss 3.3764 (3.6963) grad_norm 1.1358 (1.3083) [2021-04-16 01:09:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][950/1251] eta 0:01:25 lr 0.000707 time 0.2612 (0.2833) loss 2.7323 (3.6960) grad_norm 1.4346 (1.3083) [2021-04-16 01:09:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][960/1251] eta 0:01:22 lr 0.000707 time 0.2660 (0.2832) loss 4.4222 (3.6966) grad_norm 1.2118 (1.3080) [2021-04-16 01:09:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][970/1251] eta 0:01:19 lr 0.000707 time 0.2747 (0.2831) loss 2.8554 (3.6953) grad_norm 1.1014 (1.3071) [2021-04-16 01:09:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][980/1251] eta 0:01:16 lr 0.000707 time 0.3040 (0.2831) loss 4.3219 (3.6962) grad_norm 1.0935 (1.3071) [2021-04-16 01:09:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][990/1251] eta 0:01:13 lr 0.000707 time 0.2805 (0.2830) loss 3.9144 (3.6958) grad_norm 1.3977 (1.3064) [2021-04-16 01:09:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][1000/1251] eta 0:01:11 lr 0.000707 time 0.2752 (0.2829) loss 4.0307 (3.6971) grad_norm 1.1160 (1.3054) [2021-04-16 01:09:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][1010/1251] eta 0:01:08 lr 0.000707 time 0.2825 (0.2830) loss 4.3808 (3.6966) grad_norm 1.4497 (1.3062) [2021-04-16 01:09:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][1020/1251] eta 0:01:05 lr 0.000707 time 0.2846 (0.2829) loss 4.5160 (3.6946) grad_norm 1.3256 (1.3066) [2021-04-16 01:09:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][1030/1251] eta 0:01:02 lr 0.000707 time 0.2834 (0.2830) loss 4.0305 (3.6926) grad_norm 1.3524 (1.3065) [2021-04-16 01:09:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][1040/1251] eta 0:00:59 lr 0.000707 time 0.2476 (0.2831) loss 4.0048 (3.6926) grad_norm 1.3110 (1.3068) [2021-04-16 01:09:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][1050/1251] eta 0:00:56 lr 0.000707 time 0.2905 (0.2830) loss 3.9067 (3.6954) grad_norm 1.3884 (1.3070) [2021-04-16 01:09:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][1060/1251] eta 0:00:54 lr 0.000707 time 0.2462 (0.2829) loss 4.6987 (3.6966) grad_norm 1.2360 (1.3066) [2021-04-16 01:09:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][1070/1251] eta 0:00:51 lr 0.000707 time 0.2517 (0.2830) loss 3.2944 (3.6968) grad_norm 1.3200 (1.3065) [2021-04-16 01:10:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][1080/1251] eta 0:00:48 lr 0.000707 time 0.2712 (0.2829) loss 2.5983 (3.6960) grad_norm 1.1959 (1.3061) [2021-04-16 01:10:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][1090/1251] eta 0:00:45 lr 0.000707 time 0.2662 (0.2829) loss 3.6769 (3.6937) grad_norm 1.2787 (1.3058) [2021-04-16 01:10:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][1100/1251] eta 0:00:42 lr 0.000707 time 0.2870 (0.2828) loss 4.0526 (3.6945) grad_norm 1.5458 (1.3065) [2021-04-16 01:10:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][1110/1251] eta 0:00:39 lr 0.000707 time 0.2941 (0.2828) loss 3.9488 (3.6945) grad_norm 1.2837 (1.3065) [2021-04-16 01:10:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][1120/1251] eta 0:00:37 lr 0.000707 time 0.2634 (0.2827) loss 3.2415 (3.6933) grad_norm 1.2242 (1.3061) [2021-04-16 01:10:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][1130/1251] eta 0:00:34 lr 0.000707 time 0.2795 (0.2827) loss 3.7590 (3.6939) grad_norm 1.2758 (1.3065) [2021-04-16 01:10:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][1140/1251] eta 0:00:31 lr 0.000707 time 0.2989 (0.2826) loss 3.8797 (3.6930) grad_norm 1.4162 (1.3066) [2021-04-16 01:10:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][1150/1251] eta 0:00:28 lr 0.000707 time 0.2645 (0.2826) loss 3.1273 (3.6914) grad_norm 1.2971 (1.3067) [2021-04-16 01:10:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][1160/1251] eta 0:00:25 lr 0.000707 time 0.3250 (0.2826) loss 3.1384 (3.6915) grad_norm 1.2801 (1.3065) [2021-04-16 01:10:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][1170/1251] eta 0:00:22 lr 0.000707 time 0.2656 (0.2825) loss 2.8377 (3.6899) grad_norm 1.2998 (inf) [2021-04-16 01:10:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][1180/1251] eta 0:00:20 lr 0.000707 time 0.2866 (0.2825) loss 4.0646 (3.6893) grad_norm 1.2305 (inf) [2021-04-16 01:10:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][1190/1251] eta 0:00:17 lr 0.000707 time 0.2981 (0.2826) loss 2.8357 (3.6863) grad_norm 1.2950 (inf) [2021-04-16 01:10:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][1200/1251] eta 0:00:14 lr 0.000707 time 0.2665 (0.2825) loss 2.7993 (3.6880) grad_norm 1.4114 (inf) [2021-04-16 01:10:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][1210/1251] eta 0:00:11 lr 0.000706 time 0.2786 (0.2824) loss 2.7933 (3.6893) grad_norm 1.3190 (inf) [2021-04-16 01:10:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][1220/1251] eta 0:00:08 lr 0.000706 time 0.2643 (0.2824) loss 4.2308 (3.6875) grad_norm 1.4108 (inf) [2021-04-16 01:10:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][1230/1251] eta 0:00:05 lr 0.000706 time 0.2632 (0.2823) loss 4.4096 (3.6854) grad_norm 1.0499 (inf) [2021-04-16 01:10:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][1240/1251] eta 0:00:03 lr 0.000706 time 0.3503 (0.2823) loss 3.8727 (3.6851) grad_norm 1.2194 (inf) [2021-04-16 01:10:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [109/300][1250/1251] eta 0:00:00 lr 0.000706 time 0.2480 (0.2820) loss 3.8857 (3.6851) grad_norm 1.5332 (inf) [2021-04-16 01:10:53 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 109 training takes 0:05:56 [2021-04-16 01:10:53 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_109.pth saving...... [2021-04-16 01:11:06 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_109.pth saved !!! [2021-04-16 01:11:08 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.176 (1.176) Loss 1.0225 (1.0225) Acc@1 75.488 (75.488) Acc@5 91.797 (91.797) [2021-04-16 01:11:09 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.162 (0.268) Loss 1.1126 (1.0778) Acc@1 74.512 (74.547) Acc@5 92.090 (92.365) [2021-04-16 01:11:12 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.321 (0.249) Loss 1.0882 (1.0982) Acc@1 75.195 (74.023) Acc@5 92.285 (92.276) [2021-04-16 01:11:13 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.349 (0.220) Loss 1.1010 (1.0945) Acc@1 75.977 (74.342) Acc@5 92.480 (92.436) [2021-04-16 01:11:15 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.215) Loss 1.1973 (1.0983) Acc@1 71.680 (74.324) Acc@5 90.723 (92.392) [2021-04-16 01:11:19 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 74.250 Acc@5 92.448 [2021-04-16 01:11:19 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 74.3% [2021-04-16 01:11:19 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 74.25% [2021-04-16 01:11:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][0/1251] eta 1:38:48 lr 0.000706 time 4.7389 (4.7389) loss 3.9798 (3.9798) grad_norm 1.2109 (1.2109) [2021-04-16 01:11:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][10/1251] eta 0:14:30 lr 0.000706 time 0.4377 (0.7017) loss 3.5658 (3.3667) grad_norm 1.4757 (1.3416) [2021-04-16 01:11:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][20/1251] eta 0:10:17 lr 0.000706 time 0.2933 (0.5020) loss 4.1238 (3.4808) grad_norm 1.3341 (1.3541) [2021-04-16 01:11:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][30/1251] eta 0:08:44 lr 0.000706 time 0.2866 (0.4298) loss 3.8067 (3.4909) grad_norm 1.3962 (1.3594) [2021-04-16 01:11:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][40/1251] eta 0:07:57 lr 0.000706 time 0.2797 (0.3941) loss 4.1485 (3.5392) grad_norm 1.1528 (1.3255) [2021-04-16 01:11:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][50/1251] eta 0:07:27 lr 0.000706 time 0.2749 (0.3723) loss 3.8379 (3.5967) grad_norm 1.1219 (1.3096) [2021-04-16 01:11:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][60/1251] eta 0:07:05 lr 0.000706 time 0.2787 (0.3576) loss 3.9952 (3.6461) grad_norm 1.1628 (1.3116) [2021-04-16 01:11:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][70/1251] eta 0:06:51 lr 0.000706 time 0.2774 (0.3482) loss 2.7846 (3.6360) grad_norm 1.5834 (1.3113) [2021-04-16 01:11:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][80/1251] eta 0:06:38 lr 0.000706 time 0.2988 (0.3406) loss 4.5231 (3.6320) grad_norm 1.3387 (1.3004) [2021-04-16 01:11:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][90/1251] eta 0:06:28 lr 0.000706 time 0.3031 (0.3347) loss 4.6601 (3.6856) grad_norm 1.3087 (1.3008) [2021-04-16 01:11:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][100/1251] eta 0:06:19 lr 0.000706 time 0.2835 (0.3299) loss 2.5270 (3.6795) grad_norm 1.3490 (1.3056) [2021-04-16 01:11:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][110/1251] eta 0:06:11 lr 0.000706 time 0.2842 (0.3255) loss 3.9020 (3.6741) grad_norm 1.1969 (1.3045) [2021-04-16 01:11:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][120/1251] eta 0:06:06 lr 0.000706 time 0.2848 (0.3240) loss 4.0309 (3.6854) grad_norm 1.9622 (1.3099) [2021-04-16 01:12:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][130/1251] eta 0:05:59 lr 0.000706 time 0.2614 (0.3203) loss 3.0787 (3.6607) grad_norm 1.2476 (1.3123) [2021-04-16 01:12:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][140/1251] eta 0:05:55 lr 0.000706 time 0.2918 (0.3196) loss 3.7967 (3.6592) grad_norm 1.3128 (1.3128) [2021-04-16 01:12:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][150/1251] eta 0:05:48 lr 0.000706 time 0.2567 (0.3165) loss 3.7088 (3.6667) grad_norm 1.3762 (1.3209) [2021-04-16 01:12:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][160/1251] eta 0:05:42 lr 0.000706 time 0.2682 (0.3142) loss 4.0256 (3.6474) grad_norm 1.1736 (1.3208) [2021-04-16 01:12:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][170/1251] eta 0:05:37 lr 0.000706 time 0.2589 (0.3120) loss 3.8032 (3.6468) grad_norm 1.2076 (1.3210) [2021-04-16 01:12:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][180/1251] eta 0:05:32 lr 0.000706 time 0.2942 (0.3104) loss 3.5335 (3.6487) grad_norm 1.3252 (1.3190) [2021-04-16 01:12:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][190/1251] eta 0:05:27 lr 0.000706 time 0.2905 (0.3085) loss 4.5229 (3.6478) grad_norm 1.3447 (1.3166) [2021-04-16 01:12:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][200/1251] eta 0:05:22 lr 0.000706 time 0.2472 (0.3069) loss 3.1249 (3.6430) grad_norm 1.1946 (1.3117) [2021-04-16 01:12:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][210/1251] eta 0:05:17 lr 0.000706 time 0.2699 (0.3053) loss 3.4145 (3.6447) grad_norm 1.3427 (1.3121) [2021-04-16 01:12:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][220/1251] eta 0:05:14 lr 0.000706 time 0.3047 (0.3050) loss 4.0330 (3.6551) grad_norm 1.2242 (1.3112) [2021-04-16 01:12:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][230/1251] eta 0:05:10 lr 0.000705 time 0.2841 (0.3041) loss 2.8720 (3.6513) grad_norm 1.5605 (1.3130) [2021-04-16 01:12:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][240/1251] eta 0:05:07 lr 0.000705 time 0.2733 (0.3037) loss 3.6102 (3.6575) grad_norm 1.2220 (1.3124) [2021-04-16 01:12:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][250/1251] eta 0:05:03 lr 0.000705 time 0.2743 (0.3031) loss 4.3005 (3.6628) grad_norm 1.3584 (1.3133) [2021-04-16 01:12:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][260/1251] eta 0:04:59 lr 0.000705 time 0.2751 (0.3025) loss 3.3232 (3.6565) grad_norm 1.2875 (1.3165) [2021-04-16 01:12:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][270/1251] eta 0:04:56 lr 0.000705 time 0.2691 (0.3019) loss 4.0773 (3.6661) grad_norm 1.3626 (1.3160) [2021-04-16 01:12:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][280/1251] eta 0:04:52 lr 0.000705 time 0.2858 (0.3009) loss 4.0154 (3.6713) grad_norm 1.1085 (1.3169) [2021-04-16 01:12:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][290/1251] eta 0:04:48 lr 0.000705 time 0.2793 (0.3007) loss 3.3285 (3.6737) grad_norm 1.2803 (1.3156) [2021-04-16 01:12:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][300/1251] eta 0:04:45 lr 0.000705 time 0.2608 (0.2998) loss 4.3238 (3.6876) grad_norm 1.3044 (1.3148) [2021-04-16 01:12:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][310/1251] eta 0:04:41 lr 0.000705 time 0.2408 (0.2990) loss 3.7554 (3.6897) grad_norm 1.3438 (1.3122) [2021-04-16 01:12:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][320/1251] eta 0:04:37 lr 0.000705 time 0.2578 (0.2982) loss 3.5464 (3.6908) grad_norm 1.3514 (1.3097) [2021-04-16 01:12:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][330/1251] eta 0:04:34 lr 0.000705 time 0.2659 (0.2975) loss 4.0609 (3.6894) grad_norm 1.4888 (1.3089) [2021-04-16 01:13:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][340/1251] eta 0:04:30 lr 0.000705 time 0.2598 (0.2970) loss 4.5424 (3.7018) grad_norm 1.4874 (1.3109) [2021-04-16 01:13:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][350/1251] eta 0:04:27 lr 0.000705 time 0.2722 (0.2964) loss 3.0413 (3.6999) grad_norm 1.3305 (1.3095) [2021-04-16 01:13:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][360/1251] eta 0:04:23 lr 0.000705 time 0.2959 (0.2960) loss 3.0186 (3.6993) grad_norm 1.1943 (1.3086) [2021-04-16 01:13:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][370/1251] eta 0:04:20 lr 0.000705 time 0.2819 (0.2955) loss 3.5619 (3.6999) grad_norm 1.3660 (1.3091) [2021-04-16 01:13:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][380/1251] eta 0:04:17 lr 0.000705 time 0.2878 (0.2952) loss 2.8174 (3.6966) grad_norm 1.3040 (1.3099) [2021-04-16 01:13:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][390/1251] eta 0:04:13 lr 0.000705 time 0.2757 (0.2949) loss 2.6668 (3.6942) grad_norm 1.3816 (1.3105) [2021-04-16 01:13:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][400/1251] eta 0:04:10 lr 0.000705 time 0.2733 (0.2944) loss 4.1225 (3.6890) grad_norm 1.0551 (1.3090) [2021-04-16 01:13:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][410/1251] eta 0:04:07 lr 0.000705 time 0.2462 (0.2941) loss 3.8606 (3.6894) grad_norm 1.1560 (1.3076) [2021-04-16 01:13:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][420/1251] eta 0:04:04 lr 0.000705 time 0.2708 (0.2939) loss 2.7469 (3.6788) grad_norm 1.2970 (1.3061) [2021-04-16 01:13:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][430/1251] eta 0:04:00 lr 0.000705 time 0.2758 (0.2935) loss 3.5785 (3.6776) grad_norm 1.1960 (1.3070) [2021-04-16 01:13:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][440/1251] eta 0:03:58 lr 0.000705 time 0.2631 (0.2938) loss 4.3024 (3.6802) grad_norm 1.4760 (1.3079) [2021-04-16 01:13:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][450/1251] eta 0:03:55 lr 0.000705 time 0.3114 (0.2937) loss 3.5303 (3.6732) grad_norm 1.2173 (1.3082) [2021-04-16 01:13:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][460/1251] eta 0:03:51 lr 0.000705 time 0.2818 (0.2933) loss 4.4327 (3.6763) grad_norm 1.3013 (1.3099) [2021-04-16 01:13:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][470/1251] eta 0:03:48 lr 0.000705 time 0.2745 (0.2929) loss 2.4271 (3.6765) grad_norm 1.2948 (1.3090) [2021-04-16 01:13:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][480/1251] eta 0:03:45 lr 0.000705 time 0.2724 (0.2927) loss 3.7799 (3.6777) grad_norm 1.2156 (1.3074) [2021-04-16 01:13:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][490/1251] eta 0:03:42 lr 0.000704 time 0.2913 (0.2927) loss 3.8028 (3.6749) grad_norm 1.3242 (1.3079) [2021-04-16 01:13:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][500/1251] eta 0:03:39 lr 0.000704 time 0.3033 (0.2926) loss 2.4247 (3.6716) grad_norm 1.2511 (1.3078) [2021-04-16 01:13:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][510/1251] eta 0:03:36 lr 0.000704 time 0.2777 (0.2925) loss 3.6267 (3.6684) grad_norm 1.3818 (1.3094) [2021-04-16 01:13:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][520/1251] eta 0:03:33 lr 0.000704 time 0.2942 (0.2922) loss 3.9291 (3.6718) grad_norm 1.2435 (1.3096) [2021-04-16 01:13:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][530/1251] eta 0:03:30 lr 0.000704 time 0.2744 (0.2919) loss 2.5603 (3.6702) grad_norm 1.1929 (1.3087) [2021-04-16 01:13:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][540/1251] eta 0:03:27 lr 0.000704 time 0.3038 (0.2918) loss 4.5944 (3.6774) grad_norm 1.6464 (1.3097) [2021-04-16 01:14:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][550/1251] eta 0:03:24 lr 0.000704 time 0.2552 (0.2916) loss 3.5259 (3.6809) grad_norm 1.3487 (1.3108) [2021-04-16 01:14:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][560/1251] eta 0:03:21 lr 0.000704 time 0.2724 (0.2914) loss 4.5068 (3.6821) grad_norm 1.1938 (1.3104) [2021-04-16 01:14:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][570/1251] eta 0:03:18 lr 0.000704 time 0.2823 (0.2914) loss 4.1593 (3.6852) grad_norm 1.3556 (1.3103) [2021-04-16 01:14:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][580/1251] eta 0:03:15 lr 0.000704 time 0.2562 (0.2914) loss 3.3402 (3.6899) grad_norm 1.1568 (1.3088) [2021-04-16 01:14:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][590/1251] eta 0:03:12 lr 0.000704 time 0.2678 (0.2914) loss 4.0539 (3.6885) grad_norm 1.5754 (1.3094) [2021-04-16 01:14:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][600/1251] eta 0:03:09 lr 0.000704 time 0.2985 (0.2911) loss 2.9030 (3.6827) grad_norm 1.3183 (1.3110) [2021-04-16 01:14:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][610/1251] eta 0:03:06 lr 0.000704 time 0.2817 (0.2909) loss 2.1640 (3.6822) grad_norm 1.3501 (1.3115) [2021-04-16 01:14:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][620/1251] eta 0:03:03 lr 0.000704 time 0.2773 (0.2907) loss 3.8140 (3.6816) grad_norm 1.1255 (1.3117) [2021-04-16 01:14:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][630/1251] eta 0:03:00 lr 0.000704 time 0.2751 (0.2905) loss 3.5853 (3.6804) grad_norm 1.5273 (1.3116) [2021-04-16 01:14:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][640/1251] eta 0:02:57 lr 0.000704 time 0.2703 (0.2906) loss 3.8220 (3.6808) grad_norm 1.3918 (1.3112) [2021-04-16 01:14:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][650/1251] eta 0:02:54 lr 0.000704 time 0.2863 (0.2904) loss 4.5837 (3.6781) grad_norm 1.2404 (1.3105) [2021-04-16 01:14:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][660/1251] eta 0:02:51 lr 0.000704 time 0.2662 (0.2902) loss 3.3061 (3.6829) grad_norm 1.0482 (1.3100) [2021-04-16 01:14:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][670/1251] eta 0:02:48 lr 0.000704 time 0.2807 (0.2901) loss 4.0098 (3.6818) grad_norm 1.4118 (1.3100) [2021-04-16 01:14:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][680/1251] eta 0:02:45 lr 0.000704 time 0.2681 (0.2899) loss 2.7657 (3.6759) grad_norm 1.1779 (1.3094) [2021-04-16 01:14:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][690/1251] eta 0:02:42 lr 0.000704 time 0.2733 (0.2899) loss 4.2693 (3.6729) grad_norm 1.3132 (1.3096) [2021-04-16 01:14:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][700/1251] eta 0:02:39 lr 0.000704 time 0.2747 (0.2898) loss 3.3094 (3.6757) grad_norm 1.4087 (1.3104) [2021-04-16 01:14:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][710/1251] eta 0:02:36 lr 0.000704 time 0.2895 (0.2897) loss 4.2852 (3.6751) grad_norm 1.1038 (1.3102) [2021-04-16 01:14:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][720/1251] eta 0:02:33 lr 0.000704 time 0.2543 (0.2898) loss 3.9277 (3.6719) grad_norm 1.0883 (1.3089) [2021-04-16 01:14:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][730/1251] eta 0:02:30 lr 0.000704 time 0.2909 (0.2896) loss 3.8746 (3.6694) grad_norm 1.2055 (1.3079) [2021-04-16 01:14:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][740/1251] eta 0:02:27 lr 0.000704 time 0.2828 (0.2895) loss 3.8034 (3.6686) grad_norm 1.1830 (1.3067) [2021-04-16 01:14:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][750/1251] eta 0:02:25 lr 0.000703 time 0.2522 (0.2894) loss 3.7559 (3.6688) grad_norm 1.3529 (1.3064) [2021-04-16 01:14:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][760/1251] eta 0:02:22 lr 0.000703 time 0.2769 (0.2894) loss 4.2339 (3.6669) grad_norm 1.1812 (1.3051) [2021-04-16 01:15:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][770/1251] eta 0:02:19 lr 0.000703 time 0.2820 (0.2894) loss 4.3572 (3.6689) grad_norm 1.2393 (1.3043) [2021-04-16 01:15:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][780/1251] eta 0:02:16 lr 0.000703 time 0.2773 (0.2893) loss 4.4451 (3.6716) grad_norm 1.3387 (1.3039) [2021-04-16 01:15:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][790/1251] eta 0:02:13 lr 0.000703 time 0.3176 (0.2895) loss 3.8386 (3.6723) grad_norm 1.2750 (1.3047) [2021-04-16 01:15:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][800/1251] eta 0:02:10 lr 0.000703 time 0.2568 (0.2893) loss 3.2385 (3.6727) grad_norm 1.1759 (1.3053) [2021-04-16 01:15:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][810/1251] eta 0:02:07 lr 0.000703 time 0.2865 (0.2892) loss 3.5439 (3.6720) grad_norm 1.2992 (1.3069) [2021-04-16 01:15:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][820/1251] eta 0:02:04 lr 0.000703 time 0.2725 (0.2891) loss 3.6852 (3.6702) grad_norm 1.1856 (1.3079) [2021-04-16 01:15:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][830/1251] eta 0:02:01 lr 0.000703 time 0.2767 (0.2891) loss 4.4192 (3.6730) grad_norm 1.4520 (1.3084) [2021-04-16 01:15:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][840/1251] eta 0:01:58 lr 0.000703 time 0.2824 (0.2891) loss 4.0566 (3.6743) grad_norm 1.4363 (1.3090) [2021-04-16 01:15:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][850/1251] eta 0:01:55 lr 0.000703 time 0.2924 (0.2889) loss 3.9485 (3.6753) grad_norm 1.2022 (1.3091) [2021-04-16 01:15:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][860/1251] eta 0:01:52 lr 0.000703 time 0.2586 (0.2888) loss 3.5272 (3.6781) grad_norm 1.1711 (1.3089) [2021-04-16 01:15:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][870/1251] eta 0:01:50 lr 0.000703 time 0.2841 (0.2888) loss 2.3662 (3.6751) grad_norm 1.2505 (1.3086) [2021-04-16 01:15:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][880/1251] eta 0:01:47 lr 0.000703 time 0.3081 (0.2887) loss 4.4908 (3.6748) grad_norm 1.1531 (1.3074) [2021-04-16 01:15:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][890/1251] eta 0:01:44 lr 0.000703 time 0.2800 (0.2886) loss 3.1703 (3.6707) grad_norm 1.2400 (1.3068) [2021-04-16 01:15:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][900/1251] eta 0:01:41 lr 0.000703 time 0.2789 (0.2884) loss 3.7372 (3.6707) grad_norm 1.3175 (1.3058) [2021-04-16 01:15:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][910/1251] eta 0:01:38 lr 0.000703 time 0.3066 (0.2884) loss 3.5983 (3.6714) grad_norm 1.2045 (1.3055) [2021-04-16 01:15:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][920/1251] eta 0:01:35 lr 0.000703 time 0.2765 (0.2883) loss 3.5471 (3.6723) grad_norm 1.6652 (1.3059) [2021-04-16 01:15:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][930/1251] eta 0:01:32 lr 0.000703 time 0.2752 (0.2883) loss 3.9696 (3.6745) grad_norm 1.3442 (1.3062) [2021-04-16 01:15:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][940/1251] eta 0:01:29 lr 0.000703 time 0.2644 (0.2881) loss 4.1087 (3.6775) grad_norm 1.3512 (1.3059) [2021-04-16 01:15:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][950/1251] eta 0:01:26 lr 0.000703 time 0.2689 (0.2880) loss 3.4615 (3.6767) grad_norm 1.3242 (1.3067) [2021-04-16 01:15:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][960/1251] eta 0:01:23 lr 0.000703 time 0.2818 (0.2880) loss 4.0729 (3.6767) grad_norm 1.2123 (1.3074) [2021-04-16 01:15:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][970/1251] eta 0:01:20 lr 0.000703 time 0.2682 (0.2879) loss 3.5060 (3.6757) grad_norm 1.5753 (1.3085) [2021-04-16 01:16:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][980/1251] eta 0:01:18 lr 0.000703 time 0.3066 (0.2879) loss 2.7558 (3.6732) grad_norm 1.4451 (1.3095) [2021-04-16 01:16:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][990/1251] eta 0:01:15 lr 0.000703 time 0.2904 (0.2878) loss 3.8242 (3.6717) grad_norm 1.3871 (1.3108) [2021-04-16 01:16:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][1000/1251] eta 0:01:12 lr 0.000703 time 0.3103 (0.2878) loss 3.1939 (3.6693) grad_norm 1.4293 (1.3107) [2021-04-16 01:16:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][1010/1251] eta 0:01:09 lr 0.000703 time 0.2691 (0.2877) loss 4.0851 (3.6703) grad_norm 1.1366 (1.3108) [2021-04-16 01:16:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][1020/1251] eta 0:01:06 lr 0.000702 time 0.2696 (0.2878) loss 3.7771 (3.6698) grad_norm 1.1954 (1.3105) [2021-04-16 01:16:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][1030/1251] eta 0:01:03 lr 0.000702 time 0.2661 (0.2877) loss 3.8393 (3.6702) grad_norm 1.1110 (1.3103) [2021-04-16 01:16:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][1040/1251] eta 0:01:00 lr 0.000702 time 0.2682 (0.2876) loss 4.4499 (3.6686) grad_norm 1.1546 (1.3096) [2021-04-16 01:16:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][1050/1251] eta 0:00:57 lr 0.000702 time 0.2569 (0.2875) loss 3.3972 (3.6687) grad_norm 1.3424 (1.3091) [2021-04-16 01:16:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][1060/1251] eta 0:00:54 lr 0.000702 time 0.2671 (0.2874) loss 4.4129 (3.6699) grad_norm 1.2385 (1.3087) [2021-04-16 01:16:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][1070/1251] eta 0:00:52 lr 0.000702 time 0.2661 (0.2874) loss 3.7595 (3.6699) grad_norm 1.2542 (1.3092) [2021-04-16 01:16:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][1080/1251] eta 0:00:49 lr 0.000702 time 0.3203 (0.2873) loss 3.3235 (3.6696) grad_norm 1.3309 (1.3098) [2021-04-16 01:16:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][1090/1251] eta 0:00:46 lr 0.000702 time 0.2745 (0.2872) loss 4.3242 (3.6712) grad_norm 1.2323 (1.3092) [2021-04-16 01:16:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][1100/1251] eta 0:00:43 lr 0.000702 time 0.2881 (0.2871) loss 4.0818 (3.6736) grad_norm 1.4324 (1.3094) [2021-04-16 01:16:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][1110/1251] eta 0:00:40 lr 0.000702 time 0.3076 (0.2872) loss 2.3125 (3.6717) grad_norm 1.1122 (1.3089) [2021-04-16 01:16:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][1120/1251] eta 0:00:37 lr 0.000702 time 0.2744 (0.2870) loss 4.6645 (3.6713) grad_norm 1.2411 (1.3092) [2021-04-16 01:16:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][1130/1251] eta 0:00:34 lr 0.000702 time 0.2751 (0.2870) loss 2.9508 (3.6726) grad_norm 1.2783 (1.3088) [2021-04-16 01:16:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][1140/1251] eta 0:00:31 lr 0.000702 time 0.2746 (0.2869) loss 4.2981 (3.6754) grad_norm 1.1886 (1.3085) [2021-04-16 01:16:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][1150/1251] eta 0:00:28 lr 0.000702 time 0.2485 (0.2869) loss 3.6778 (3.6734) grad_norm 1.8158 (1.3086) [2021-04-16 01:16:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][1160/1251] eta 0:00:26 lr 0.000702 time 0.2452 (0.2868) loss 4.2453 (3.6712) grad_norm 1.2550 (1.3085) [2021-04-16 01:16:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][1170/1251] eta 0:00:23 lr 0.000702 time 0.2762 (0.2868) loss 4.1073 (3.6726) grad_norm 1.6701 (1.3091) [2021-04-16 01:16:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][1180/1251] eta 0:00:20 lr 0.000702 time 0.2714 (0.2868) loss 2.4938 (3.6718) grad_norm 1.3903 (1.3091) [2021-04-16 01:17:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][1190/1251] eta 0:00:17 lr 0.000702 time 0.2930 (0.2868) loss 3.7374 (3.6709) grad_norm 1.2304 (1.3087) [2021-04-16 01:17:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][1200/1251] eta 0:00:14 lr 0.000702 time 0.3077 (0.2867) loss 4.0048 (3.6734) grad_norm 1.3066 (1.3088) [2021-04-16 01:17:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][1210/1251] eta 0:00:11 lr 0.000702 time 0.2817 (0.2866) loss 4.3203 (3.6745) grad_norm 1.2227 (1.3086) [2021-04-16 01:17:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][1220/1251] eta 0:00:08 lr 0.000702 time 0.2669 (0.2866) loss 3.4872 (3.6750) grad_norm 1.2132 (1.3089) [2021-04-16 01:17:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][1230/1251] eta 0:00:06 lr 0.000702 time 0.2716 (0.2866) loss 3.2072 (3.6754) grad_norm 1.4162 (1.3094) [2021-04-16 01:17:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][1240/1251] eta 0:00:03 lr 0.000702 time 0.2494 (0.2864) loss 4.7784 (3.6748) grad_norm 1.3028 (1.3092) [2021-04-16 01:17:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [110/300][1250/1251] eta 0:00:00 lr 0.000702 time 0.2487 (0.2862) loss 4.4273 (3.6749) grad_norm 1.1472 (1.3093) [2021-04-16 01:17:19 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 110 training takes 0:06:00 [2021-04-16 01:17:19 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_110.pth saving...... [2021-04-16 01:17:38 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_110.pth saved !!! [2021-04-16 01:17:39 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.067 (1.067) Loss 1.1158 (1.1158) Acc@1 75.000 (75.000) Acc@5 90.527 (90.527) [2021-04-16 01:17:40 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.121 (0.222) Loss 1.0443 (1.0926) Acc@1 75.000 (74.316) Acc@5 92.578 (92.063) [2021-04-16 01:17:43 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.124 (0.230) Loss 1.0950 (1.0873) Acc@1 73.438 (74.312) Acc@5 92.676 (92.029) [2021-04-16 01:17:45 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.095 (0.224) Loss 1.0546 (1.0915) Acc@1 75.684 (74.134) Acc@5 91.992 (92.011) [2021-04-16 01:17:47 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.447 (0.223) Loss 1.1643 (1.0883) Acc@1 70.996 (74.216) Acc@5 91.309 (92.168) [2021-04-16 01:17:51 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 74.096 Acc@5 92.170 [2021-04-16 01:17:51 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 74.1% [2021-04-16 01:17:51 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 74.25% [2021-04-16 01:17:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][0/1251] eta 1:34:46 lr 0.000702 time 4.5459 (4.5459) loss 4.5034 (4.5034) grad_norm 1.3871 (1.3871) [2021-04-16 01:17:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][10/1251] eta 0:13:39 lr 0.000702 time 0.2846 (0.6602) loss 4.0874 (3.7001) grad_norm 1.6799 (1.3508) [2021-04-16 01:18:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][20/1251] eta 0:09:53 lr 0.000702 time 0.2599 (0.4823) loss 3.6374 (3.6078) grad_norm 1.3183 (1.3367) [2021-04-16 01:18:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][30/1251] eta 0:08:34 lr 0.000701 time 0.3041 (0.4217) loss 4.0679 (3.7187) grad_norm 1.3721 (1.2975) [2021-04-16 01:18:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][40/1251] eta 0:07:47 lr 0.000701 time 0.2889 (0.3862) loss 2.9568 (3.6032) grad_norm 1.5934 (1.3134) [2021-04-16 01:18:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][50/1251] eta 0:07:19 lr 0.000701 time 0.2894 (0.3662) loss 3.2099 (3.5445) grad_norm 1.4707 (1.3151) [2021-04-16 01:18:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][60/1251] eta 0:06:59 lr 0.000701 time 0.2917 (0.3526) loss 2.6664 (3.5705) grad_norm 1.2746 (1.3143) [2021-04-16 01:18:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][70/1251] eta 0:06:44 lr 0.000701 time 0.2775 (0.3429) loss 2.6227 (3.5709) grad_norm 1.1979 (1.3196) [2021-04-16 01:18:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][80/1251] eta 0:06:33 lr 0.000701 time 0.2549 (0.3362) loss 4.0272 (3.5584) grad_norm 1.2269 (1.3282) [2021-04-16 01:18:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][90/1251] eta 0:06:23 lr 0.000701 time 0.2892 (0.3307) loss 4.2044 (3.5825) grad_norm 1.4455 (1.3296) [2021-04-16 01:18:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][100/1251] eta 0:06:15 lr 0.000701 time 0.2756 (0.3263) loss 4.4977 (3.5768) grad_norm 1.6450 (1.3352) [2021-04-16 01:18:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][110/1251] eta 0:06:07 lr 0.000701 time 0.3070 (0.3222) loss 3.3771 (3.5931) grad_norm 1.4874 (1.3359) [2021-04-16 01:18:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][120/1251] eta 0:06:00 lr 0.000701 time 0.2826 (0.3187) loss 3.9069 (3.6215) grad_norm 1.3563 (1.3355) [2021-04-16 01:18:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][130/1251] eta 0:05:53 lr 0.000701 time 0.2784 (0.3157) loss 3.8779 (3.6305) grad_norm 1.3909 (1.3305) [2021-04-16 01:18:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][140/1251] eta 0:05:47 lr 0.000701 time 0.2700 (0.3129) loss 3.3677 (3.6229) grad_norm 1.3051 (1.3248) [2021-04-16 01:18:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][150/1251] eta 0:05:41 lr 0.000701 time 0.2505 (0.3103) loss 4.0393 (3.6272) grad_norm 1.3476 (1.3220) [2021-04-16 01:18:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][160/1251] eta 0:05:36 lr 0.000701 time 0.2977 (0.3083) loss 3.6905 (3.6245) grad_norm 1.3186 (1.3189) [2021-04-16 01:18:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][170/1251] eta 0:05:31 lr 0.000701 time 0.2704 (0.3067) loss 3.7765 (3.6385) grad_norm 1.4676 (1.3210) [2021-04-16 01:18:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][180/1251] eta 0:05:26 lr 0.000701 time 0.2838 (0.3047) loss 3.8704 (3.6509) grad_norm 1.3540 (1.3193) [2021-04-16 01:18:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][190/1251] eta 0:05:21 lr 0.000701 time 0.2741 (0.3032) loss 3.8324 (3.6447) grad_norm 1.4350 (1.3182) [2021-04-16 01:18:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][200/1251] eta 0:05:17 lr 0.000701 time 0.2783 (0.3017) loss 2.7739 (3.6486) grad_norm 1.4638 (1.3213) [2021-04-16 01:18:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][210/1251] eta 0:05:13 lr 0.000701 time 0.2766 (0.3008) loss 4.0544 (3.6703) grad_norm 1.0393 (1.3189) [2021-04-16 01:18:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][220/1251] eta 0:05:08 lr 0.000701 time 0.2757 (0.2996) loss 3.1848 (3.6520) grad_norm 1.4074 (1.3162) [2021-04-16 01:19:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][230/1251] eta 0:05:05 lr 0.000701 time 0.2840 (0.2996) loss 2.9628 (3.6494) grad_norm 1.2657 (1.3142) [2021-04-16 01:19:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][240/1251] eta 0:05:01 lr 0.000701 time 0.2819 (0.2986) loss 3.9867 (3.6486) grad_norm 1.3722 (1.3160) [2021-04-16 01:19:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][250/1251] eta 0:04:58 lr 0.000701 time 0.2610 (0.2977) loss 3.9107 (3.6435) grad_norm 1.2915 (1.3161) [2021-04-16 01:19:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][260/1251] eta 0:04:54 lr 0.000701 time 0.2915 (0.2973) loss 5.1098 (3.6593) grad_norm 1.4375 (1.3205) [2021-04-16 01:19:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][270/1251] eta 0:04:51 lr 0.000701 time 0.2867 (0.2968) loss 3.6312 (3.6642) grad_norm 1.8574 (1.3257) [2021-04-16 01:19:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][280/1251] eta 0:04:47 lr 0.000701 time 0.2542 (0.2960) loss 3.9969 (3.6640) grad_norm 1.2391 (1.3258) [2021-04-16 01:19:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][290/1251] eta 0:04:44 lr 0.000700 time 0.2946 (0.2958) loss 3.5853 (3.6689) grad_norm 1.2652 (1.3272) [2021-04-16 01:19:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][300/1251] eta 0:04:40 lr 0.000700 time 0.2467 (0.2952) loss 2.9236 (3.6690) grad_norm 1.1517 (1.3257) [2021-04-16 01:19:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][310/1251] eta 0:04:37 lr 0.000700 time 0.2702 (0.2947) loss 4.6386 (3.6718) grad_norm 1.5388 (1.3272) [2021-04-16 01:19:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][320/1251] eta 0:04:33 lr 0.000700 time 0.2716 (0.2939) loss 2.7536 (3.6763) grad_norm 1.2087 (1.3297) [2021-04-16 01:19:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][330/1251] eta 0:04:30 lr 0.000700 time 0.2490 (0.2933) loss 3.7693 (3.6750) grad_norm 1.1903 (1.3275) [2021-04-16 01:19:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][340/1251] eta 0:04:26 lr 0.000700 time 0.2973 (0.2930) loss 3.2448 (3.6752) grad_norm 1.2735 (1.3274) [2021-04-16 01:19:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][350/1251] eta 0:04:24 lr 0.000700 time 0.2755 (0.2931) loss 4.6067 (3.6751) grad_norm 1.7570 (1.3258) [2021-04-16 01:19:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][360/1251] eta 0:04:20 lr 0.000700 time 0.2644 (0.2926) loss 3.8055 (3.6750) grad_norm 1.3085 (1.3257) [2021-04-16 01:19:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][370/1251] eta 0:04:17 lr 0.000700 time 0.2678 (0.2924) loss 3.4835 (3.6815) grad_norm 1.2241 (1.3239) [2021-04-16 01:19:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][380/1251] eta 0:04:14 lr 0.000700 time 0.2820 (0.2919) loss 3.9169 (3.6852) grad_norm 1.3058 (1.3233) [2021-04-16 01:19:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][390/1251] eta 0:04:10 lr 0.000700 time 0.2750 (0.2915) loss 3.9931 (3.6836) grad_norm 1.2266 (1.3233) [2021-04-16 01:19:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][400/1251] eta 0:04:07 lr 0.000700 time 0.3080 (0.2913) loss 2.8686 (3.6799) grad_norm 1.4763 (1.3249) [2021-04-16 01:19:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][410/1251] eta 0:04:04 lr 0.000700 time 0.3080 (0.2909) loss 3.3501 (3.6805) grad_norm 1.3243 (1.3249) [2021-04-16 01:19:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][420/1251] eta 0:04:01 lr 0.000700 time 0.2945 (0.2906) loss 3.1007 (3.6784) grad_norm 1.6001 (1.3267) [2021-04-16 01:19:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][430/1251] eta 0:03:58 lr 0.000700 time 0.2929 (0.2906) loss 4.1613 (3.6789) grad_norm 1.4344 (1.3276) [2021-04-16 01:19:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][440/1251] eta 0:03:55 lr 0.000700 time 0.2553 (0.2903) loss 3.7737 (3.6821) grad_norm 1.3243 (1.3283) [2021-04-16 01:20:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][450/1251] eta 0:03:52 lr 0.000700 time 0.3098 (0.2903) loss 4.4516 (3.6841) grad_norm 1.3452 (1.3271) [2021-04-16 01:20:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][460/1251] eta 0:03:49 lr 0.000700 time 0.2757 (0.2900) loss 3.2876 (3.6804) grad_norm 1.2821 (1.3262) [2021-04-16 01:20:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][470/1251] eta 0:03:46 lr 0.000700 time 0.2829 (0.2896) loss 3.9668 (3.6721) grad_norm 1.0649 (1.3250) [2021-04-16 01:20:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][480/1251] eta 0:03:43 lr 0.000700 time 0.2604 (0.2894) loss 3.7060 (3.6714) grad_norm 1.3038 (1.3252) [2021-04-16 01:20:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][490/1251] eta 0:03:40 lr 0.000700 time 0.2757 (0.2892) loss 3.9080 (3.6681) grad_norm 1.2390 (1.3258) [2021-04-16 01:20:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][500/1251] eta 0:03:37 lr 0.000700 time 0.2642 (0.2890) loss 3.2201 (3.6674) grad_norm 1.1728 (1.3250) [2021-04-16 01:20:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][510/1251] eta 0:03:34 lr 0.000700 time 0.2796 (0.2889) loss 3.7322 (3.6703) grad_norm 1.2859 (1.3247) [2021-04-16 01:20:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][520/1251] eta 0:03:31 lr 0.000700 time 0.2758 (0.2888) loss 3.3645 (3.6674) grad_norm 1.2344 (1.3257) [2021-04-16 01:20:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][530/1251] eta 0:03:28 lr 0.000700 time 0.2575 (0.2885) loss 3.9580 (3.6708) grad_norm 1.5286 (1.3257) [2021-04-16 01:20:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][540/1251] eta 0:03:24 lr 0.000700 time 0.2990 (0.2883) loss 4.0093 (3.6719) grad_norm 1.5218 (1.3251) [2021-04-16 01:20:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][550/1251] eta 0:03:21 lr 0.000699 time 0.2924 (0.2881) loss 3.6728 (3.6781) grad_norm 1.3222 (1.3251) [2021-04-16 01:20:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][560/1251] eta 0:03:18 lr 0.000699 time 0.2711 (0.2878) loss 3.8043 (3.6839) grad_norm 1.4427 (1.3252) [2021-04-16 01:20:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][570/1251] eta 0:03:15 lr 0.000699 time 0.2768 (0.2877) loss 3.9970 (3.6858) grad_norm 1.1896 (1.3257) [2021-04-16 01:20:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][580/1251] eta 0:03:13 lr 0.000699 time 0.2905 (0.2877) loss 3.5481 (3.6822) grad_norm 1.1720 (1.3260) [2021-04-16 01:20:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][590/1251] eta 0:03:10 lr 0.000699 time 0.2932 (0.2879) loss 4.3500 (3.6848) grad_norm 1.1665 (1.3266) [2021-04-16 01:20:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][600/1251] eta 0:03:07 lr 0.000699 time 0.4471 (0.2880) loss 3.6246 (3.6851) grad_norm 1.4538 (1.3272) [2021-04-16 01:20:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][610/1251] eta 0:03:04 lr 0.000699 time 0.2755 (0.2878) loss 4.3970 (3.6889) grad_norm 1.4784 (1.3264) [2021-04-16 01:20:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][620/1251] eta 0:03:01 lr 0.000699 time 0.3102 (0.2877) loss 4.2980 (3.6916) grad_norm 1.4091 (1.3251) [2021-04-16 01:20:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][630/1251] eta 0:02:58 lr 0.000699 time 0.2721 (0.2875) loss 4.4684 (3.6931) grad_norm 1.2803 (1.3246) [2021-04-16 01:20:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][640/1251] eta 0:02:55 lr 0.000699 time 0.2770 (0.2874) loss 3.7725 (3.6932) grad_norm 1.3081 (1.3240) [2021-04-16 01:20:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][650/1251] eta 0:02:52 lr 0.000699 time 0.3666 (0.2873) loss 3.5428 (3.6979) grad_norm 1.3074 (1.3258) [2021-04-16 01:21:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][660/1251] eta 0:02:49 lr 0.000699 time 0.2718 (0.2872) loss 3.0277 (3.6983) grad_norm 1.4313 (1.3256) [2021-04-16 01:21:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][670/1251] eta 0:02:46 lr 0.000699 time 0.2664 (0.2870) loss 3.1963 (3.7007) grad_norm 1.3521 (1.3251) [2021-04-16 01:21:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][680/1251] eta 0:02:43 lr 0.000699 time 0.2849 (0.2869) loss 2.9789 (3.6970) grad_norm 1.3654 (1.3254) [2021-04-16 01:21:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][690/1251] eta 0:02:40 lr 0.000699 time 0.2508 (0.2868) loss 3.8824 (3.6945) grad_norm 1.2404 (1.3249) [2021-04-16 01:21:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][700/1251] eta 0:02:38 lr 0.000699 time 0.3113 (0.2868) loss 3.8538 (3.6946) grad_norm 1.4222 (1.3249) [2021-04-16 01:21:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][710/1251] eta 0:02:35 lr 0.000699 time 0.2709 (0.2866) loss 2.9079 (3.6967) grad_norm 1.4385 (1.3263) [2021-04-16 01:21:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][720/1251] eta 0:02:32 lr 0.000699 time 0.2655 (0.2864) loss 3.0224 (3.6981) grad_norm 1.4277 (inf) [2021-04-16 01:21:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][730/1251] eta 0:02:29 lr 0.000699 time 0.2771 (0.2863) loss 2.7027 (3.6967) grad_norm 1.4772 (inf) [2021-04-16 01:21:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][740/1251] eta 0:02:26 lr 0.000699 time 0.2885 (0.2865) loss 4.7020 (3.6959) grad_norm 1.1992 (inf) [2021-04-16 01:21:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][750/1251] eta 0:02:23 lr 0.000699 time 0.2983 (0.2863) loss 3.0738 (3.6971) grad_norm 1.3119 (inf) [2021-04-16 01:21:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][760/1251] eta 0:02:20 lr 0.000699 time 0.2790 (0.2862) loss 5.0185 (3.7012) grad_norm 1.4622 (inf) [2021-04-16 01:21:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][770/1251] eta 0:02:17 lr 0.000699 time 0.3042 (0.2861) loss 3.0590 (3.7013) grad_norm 1.1909 (inf) [2021-04-16 01:21:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][780/1251] eta 0:02:14 lr 0.000699 time 0.2779 (0.2859) loss 4.0287 (3.7031) grad_norm 1.3314 (inf) [2021-04-16 01:21:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][790/1251] eta 0:02:11 lr 0.000699 time 0.2718 (0.2859) loss 4.3527 (3.7050) grad_norm 1.2145 (inf) [2021-04-16 01:21:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][800/1251] eta 0:02:08 lr 0.000699 time 0.2980 (0.2859) loss 4.2096 (3.7050) grad_norm 1.2485 (inf) [2021-04-16 01:21:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][810/1251] eta 0:02:06 lr 0.000699 time 0.3049 (0.2861) loss 4.0861 (3.7042) grad_norm 1.4733 (inf) [2021-04-16 01:21:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][820/1251] eta 0:02:03 lr 0.000698 time 0.2861 (0.2860) loss 2.7156 (3.7024) grad_norm 1.2607 (inf) [2021-04-16 01:21:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][830/1251] eta 0:02:00 lr 0.000698 time 0.2790 (0.2859) loss 3.4316 (3.7028) grad_norm 1.1898 (inf) [2021-04-16 01:21:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][840/1251] eta 0:01:57 lr 0.000698 time 0.2955 (0.2857) loss 2.8607 (3.7060) grad_norm 1.3227 (inf) [2021-04-16 01:21:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][850/1251] eta 0:01:54 lr 0.000698 time 0.2789 (0.2857) loss 4.1944 (3.7037) grad_norm 1.1937 (inf) [2021-04-16 01:21:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][860/1251] eta 0:01:51 lr 0.000698 time 0.2905 (0.2857) loss 3.9938 (3.7037) grad_norm 1.2933 (inf) [2021-04-16 01:22:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][870/1251] eta 0:01:48 lr 0.000698 time 0.2983 (0.2858) loss 3.6971 (3.7031) grad_norm 1.1461 (inf) [2021-04-16 01:22:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][880/1251] eta 0:01:45 lr 0.000698 time 0.2612 (0.2856) loss 3.1971 (3.7041) grad_norm 1.6352 (inf) [2021-04-16 01:22:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][890/1251] eta 0:01:43 lr 0.000698 time 0.2762 (0.2855) loss 3.5028 (3.7046) grad_norm 1.3842 (inf) [2021-04-16 01:22:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][900/1251] eta 0:01:40 lr 0.000698 time 0.2518 (0.2854) loss 4.5981 (3.7082) grad_norm 1.2640 (inf) [2021-04-16 01:22:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][910/1251] eta 0:01:37 lr 0.000698 time 0.2651 (0.2853) loss 4.1767 (3.7039) grad_norm 1.3297 (inf) [2021-04-16 01:22:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][920/1251] eta 0:01:34 lr 0.000698 time 0.2675 (0.2852) loss 2.6702 (3.7032) grad_norm 1.2965 (inf) [2021-04-16 01:22:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][930/1251] eta 0:01:31 lr 0.000698 time 0.2619 (0.2853) loss 3.8905 (3.7025) grad_norm 1.0904 (inf) [2021-04-16 01:22:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][940/1251] eta 0:01:28 lr 0.000698 time 0.2981 (0.2854) loss 2.9666 (3.7026) grad_norm 1.1493 (inf) [2021-04-16 01:22:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][950/1251] eta 0:01:25 lr 0.000698 time 0.2714 (0.2853) loss 3.0672 (3.7016) grad_norm 1.2638 (inf) [2021-04-16 01:22:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][960/1251] eta 0:01:23 lr 0.000698 time 0.2882 (0.2853) loss 2.9854 (3.7035) grad_norm 1.6130 (inf) [2021-04-16 01:22:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][970/1251] eta 0:01:20 lr 0.000698 time 0.2818 (0.2852) loss 4.0150 (3.7030) grad_norm 1.2012 (inf) [2021-04-16 01:22:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][980/1251] eta 0:01:17 lr 0.000698 time 0.2637 (0.2851) loss 4.3991 (3.7050) grad_norm 1.1442 (inf) [2021-04-16 01:22:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][990/1251] eta 0:01:14 lr 0.000698 time 0.2891 (0.2850) loss 3.6132 (3.7051) grad_norm 1.3795 (inf) [2021-04-16 01:22:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][1000/1251] eta 0:01:11 lr 0.000698 time 0.2510 (0.2849) loss 3.7934 (3.7048) grad_norm 1.2191 (inf) [2021-04-16 01:22:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][1010/1251] eta 0:01:08 lr 0.000698 time 0.2657 (0.2849) loss 4.0417 (3.7053) grad_norm 1.4122 (inf) [2021-04-16 01:22:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][1020/1251] eta 0:01:05 lr 0.000698 time 0.2757 (0.2849) loss 3.6570 (3.7059) grad_norm 1.1306 (inf) [2021-04-16 01:22:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][1030/1251] eta 0:01:02 lr 0.000698 time 0.2592 (0.2848) loss 2.4678 (3.7062) grad_norm 1.4527 (inf) [2021-04-16 01:22:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][1040/1251] eta 0:01:00 lr 0.000698 time 0.2726 (0.2847) loss 3.7676 (3.7027) grad_norm 1.3529 (inf) [2021-04-16 01:22:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][1050/1251] eta 0:00:57 lr 0.000698 time 0.2813 (0.2845) loss 3.5467 (3.7004) grad_norm 1.3558 (inf) [2021-04-16 01:22:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][1060/1251] eta 0:00:54 lr 0.000698 time 0.2958 (0.2845) loss 3.9066 (3.7010) grad_norm 1.4708 (inf) [2021-04-16 01:22:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][1070/1251] eta 0:00:51 lr 0.000698 time 0.2807 (0.2845) loss 4.4426 (3.7024) grad_norm 1.5167 (inf) [2021-04-16 01:22:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][1080/1251] eta 0:00:48 lr 0.000697 time 0.2745 (0.2845) loss 4.5008 (3.7055) grad_norm 1.3477 (inf) [2021-04-16 01:23:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][1090/1251] eta 0:00:45 lr 0.000697 time 0.2919 (0.2845) loss 4.3337 (3.7042) grad_norm 1.5961 (inf) [2021-04-16 01:23:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][1100/1251] eta 0:00:42 lr 0.000697 time 0.3008 (0.2845) loss 3.3897 (3.7034) grad_norm 1.1000 (inf) [2021-04-16 01:23:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][1110/1251] eta 0:00:40 lr 0.000697 time 0.2814 (0.2844) loss 4.0673 (3.7054) grad_norm 1.1388 (inf) [2021-04-16 01:23:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][1120/1251] eta 0:00:37 lr 0.000697 time 0.2757 (0.2844) loss 3.9518 (3.7039) grad_norm 1.3469 (inf) [2021-04-16 01:23:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][1130/1251] eta 0:00:34 lr 0.000697 time 0.2693 (0.2843) loss 3.4464 (3.7024) grad_norm 1.5542 (inf) [2021-04-16 01:23:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][1140/1251] eta 0:00:31 lr 0.000697 time 0.2851 (0.2844) loss 3.7542 (3.7034) grad_norm 1.3923 (inf) [2021-04-16 01:23:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][1150/1251] eta 0:00:28 lr 0.000697 time 0.4155 (0.2844) loss 3.5557 (3.7032) grad_norm 1.1584 (inf) [2021-04-16 01:23:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][1160/1251] eta 0:00:25 lr 0.000697 time 0.2920 (0.2843) loss 3.3213 (3.7029) grad_norm 1.3757 (inf) [2021-04-16 01:23:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][1170/1251] eta 0:00:23 lr 0.000697 time 0.2732 (0.2844) loss 3.1146 (3.7010) grad_norm 1.3540 (inf) [2021-04-16 01:23:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][1180/1251] eta 0:00:20 lr 0.000697 time 0.2448 (0.2845) loss 4.2589 (3.7019) grad_norm 1.2783 (inf) [2021-04-16 01:23:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][1190/1251] eta 0:00:17 lr 0.000697 time 0.2776 (0.2844) loss 3.4841 (3.7021) grad_norm 1.5132 (inf) [2021-04-16 01:23:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][1200/1251] eta 0:00:14 lr 0.000697 time 0.3083 (0.2843) loss 4.3719 (3.7046) grad_norm 1.1469 (inf) [2021-04-16 01:23:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][1210/1251] eta 0:00:11 lr 0.000697 time 0.3025 (0.2843) loss 4.4518 (3.7066) grad_norm 1.3996 (inf) [2021-04-16 01:23:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][1220/1251] eta 0:00:08 lr 0.000697 time 0.3127 (0.2842) loss 3.0485 (3.7050) grad_norm 1.2889 (inf) [2021-04-16 01:23:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][1230/1251] eta 0:00:05 lr 0.000697 time 0.2695 (0.2843) loss 3.4749 (3.7055) grad_norm 1.2580 (inf) [2021-04-16 01:23:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][1240/1251] eta 0:00:03 lr 0.000697 time 0.2622 (0.2841) loss 3.1952 (3.7035) grad_norm 1.2088 (inf) [2021-04-16 01:23:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [111/300][1250/1251] eta 0:00:00 lr 0.000697 time 0.2484 (0.2838) loss 3.8808 (3.7033) grad_norm 1.3602 (inf) [2021-04-16 01:23:51 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 111 training takes 0:05:59 [2021-04-16 01:23:51 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_111.pth saving...... [2021-04-16 01:24:02 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_111.pth saved !!! [2021-04-16 01:24:03 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.130 (1.130) Loss 1.1444 (1.1444) Acc@1 73.438 (73.438) Acc@5 92.285 (92.285) [2021-04-16 01:24:05 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.187 (0.232) Loss 1.1509 (1.1472) Acc@1 72.949 (73.544) Acc@5 91.699 (92.125) [2021-04-16 01:24:07 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.442 (0.222) Loss 1.1230 (1.1324) Acc@1 74.316 (73.907) Acc@5 92.871 (92.332) [2021-04-16 01:24:09 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.138 (0.227) Loss 1.0476 (1.1215) Acc@1 75.977 (74.219) Acc@5 93.945 (92.455) [2021-04-16 01:24:11 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.214) Loss 1.1472 (1.1232) Acc@1 75.000 (74.245) Acc@5 91.602 (92.433) [2021-04-16 01:24:17 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 74.266 Acc@5 92.428 [2021-04-16 01:24:17 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 74.3% [2021-04-16 01:24:17 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 74.27% [2021-04-16 01:24:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][0/1251] eta 2:09:04 lr 0.000697 time 6.1907 (6.1907) loss 3.1642 (3.1642) grad_norm 1.4816 (1.4816) [2021-04-16 01:24:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][10/1251] eta 0:16:44 lr 0.000697 time 0.2885 (0.8095) loss 3.7497 (3.5814) grad_norm 1.3250 (1.4545) [2021-04-16 01:24:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][20/1251] eta 0:11:39 lr 0.000697 time 0.2906 (0.5681) loss 4.1504 (3.8771) grad_norm 1.1428 (1.4122) [2021-04-16 01:24:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][30/1251] eta 0:09:41 lr 0.000697 time 0.2930 (0.4760) loss 3.9719 (3.8422) grad_norm 1.3130 (1.3782) [2021-04-16 01:24:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][40/1251] eta 0:08:37 lr 0.000697 time 0.2543 (0.4275) loss 3.1081 (3.7495) grad_norm 1.5007 (1.3597) [2021-04-16 01:24:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][50/1251] eta 0:08:00 lr 0.000697 time 0.2972 (0.3999) loss 3.4402 (3.7068) grad_norm 1.4527 (1.3515) [2021-04-16 01:24:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][60/1251] eta 0:07:35 lr 0.000697 time 0.3838 (0.3821) loss 2.3897 (3.6938) grad_norm 1.2656 (1.3500) [2021-04-16 01:24:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][70/1251] eta 0:07:13 lr 0.000697 time 0.2618 (0.3673) loss 4.5149 (3.6732) grad_norm 1.1229 (1.3395) [2021-04-16 01:24:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][80/1251] eta 0:06:57 lr 0.000697 time 0.3145 (0.3569) loss 3.5544 (3.6710) grad_norm 1.4017 (1.3424) [2021-04-16 01:24:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][90/1251] eta 0:06:43 lr 0.000696 time 0.2999 (0.3479) loss 3.6372 (3.6627) grad_norm 1.2633 (1.3407) [2021-04-16 01:24:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][100/1251] eta 0:06:32 lr 0.000696 time 0.2848 (0.3410) loss 4.2099 (3.6501) grad_norm 1.0376 (1.3397) [2021-04-16 01:24:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][110/1251] eta 0:06:22 lr 0.000696 time 0.3041 (0.3352) loss 3.7864 (3.6722) grad_norm 1.3156 (1.3438) [2021-04-16 01:24:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][120/1251] eta 0:06:14 lr 0.000696 time 0.3019 (0.3308) loss 2.6692 (3.6872) grad_norm 1.5977 (1.3450) [2021-04-16 01:25:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][130/1251] eta 0:06:06 lr 0.000696 time 0.2793 (0.3269) loss 4.0107 (3.7123) grad_norm 1.2480 (1.3513) [2021-04-16 01:25:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][140/1251] eta 0:06:01 lr 0.000696 time 0.2776 (0.3257) loss 2.2500 (3.7076) grad_norm 1.0520 (1.3442) [2021-04-16 01:25:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][150/1251] eta 0:05:55 lr 0.000696 time 0.2991 (0.3230) loss 3.9956 (3.7104) grad_norm 1.3334 (1.3474) [2021-04-16 01:25:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][160/1251] eta 0:05:49 lr 0.000696 time 0.2660 (0.3199) loss 3.9390 (3.7025) grad_norm 1.4379 (1.3464) [2021-04-16 01:25:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][170/1251] eta 0:05:43 lr 0.000696 time 0.2759 (0.3174) loss 3.8183 (3.7052) grad_norm 1.2639 (1.3436) [2021-04-16 01:25:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][180/1251] eta 0:05:37 lr 0.000696 time 0.2899 (0.3152) loss 2.7072 (3.7136) grad_norm 1.2519 (1.3390) [2021-04-16 01:25:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][190/1251] eta 0:05:32 lr 0.000696 time 0.2584 (0.3131) loss 3.6936 (3.7071) grad_norm 1.3112 (1.3386) [2021-04-16 01:25:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][200/1251] eta 0:05:27 lr 0.000696 time 0.2787 (0.3115) loss 3.3681 (3.7053) grad_norm 1.2518 (1.3348) [2021-04-16 01:25:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][210/1251] eta 0:05:22 lr 0.000696 time 0.2772 (0.3101) loss 3.6046 (3.7022) grad_norm 1.2484 (1.3319) [2021-04-16 01:25:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][220/1251] eta 0:05:18 lr 0.000696 time 0.2904 (0.3089) loss 2.9520 (3.6825) grad_norm 1.3427 (1.3290) [2021-04-16 01:25:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][230/1251] eta 0:05:14 lr 0.000696 time 0.3151 (0.3083) loss 3.8185 (3.6880) grad_norm 1.2013 (1.3239) [2021-04-16 01:25:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][240/1251] eta 0:05:10 lr 0.000696 time 0.2534 (0.3069) loss 2.7711 (3.6952) grad_norm 1.3218 (1.3221) [2021-04-16 01:25:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][250/1251] eta 0:05:06 lr 0.000696 time 0.2891 (0.3058) loss 3.3223 (3.6935) grad_norm 1.2617 (1.3207) [2021-04-16 01:25:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][260/1251] eta 0:05:01 lr 0.000696 time 0.2719 (0.3047) loss 3.4173 (3.6834) grad_norm 1.2319 (1.3198) [2021-04-16 01:25:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][270/1251] eta 0:04:57 lr 0.000696 time 0.2469 (0.3038) loss 4.3290 (3.6877) grad_norm 1.4258 (1.3258) [2021-04-16 01:25:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][280/1251] eta 0:04:54 lr 0.000696 time 0.3157 (0.3029) loss 3.9597 (3.6866) grad_norm 1.2088 (1.3232) [2021-04-16 01:25:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][290/1251] eta 0:04:50 lr 0.000696 time 0.2801 (0.3025) loss 4.8772 (3.6845) grad_norm 1.4208 (1.3287) [2021-04-16 01:25:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][300/1251] eta 0:04:47 lr 0.000696 time 0.2576 (0.3021) loss 4.3616 (3.6891) grad_norm 1.3818 (1.3262) [2021-04-16 01:25:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][310/1251] eta 0:04:43 lr 0.000696 time 0.2662 (0.3014) loss 2.7883 (3.6980) grad_norm 1.3031 (1.3242) [2021-04-16 01:25:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][320/1251] eta 0:04:39 lr 0.000696 time 0.2639 (0.3006) loss 3.9923 (3.6944) grad_norm 1.3018 (1.3226) [2021-04-16 01:25:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][330/1251] eta 0:04:36 lr 0.000696 time 0.2552 (0.3001) loss 2.5941 (3.6950) grad_norm 1.2573 (1.3209) [2021-04-16 01:25:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][340/1251] eta 0:04:32 lr 0.000696 time 0.2647 (0.2995) loss 4.3095 (3.6936) grad_norm 1.3490 (1.3183) [2021-04-16 01:26:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][350/1251] eta 0:04:29 lr 0.000695 time 0.2889 (0.2993) loss 2.7524 (3.6848) grad_norm 1.3518 (1.3183) [2021-04-16 01:26:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][360/1251] eta 0:04:26 lr 0.000695 time 0.2715 (0.2991) loss 3.9882 (3.6848) grad_norm 1.3503 (1.3174) [2021-04-16 01:26:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][370/1251] eta 0:04:23 lr 0.000695 time 0.2840 (0.2992) loss 2.3818 (3.6778) grad_norm 1.2844 (1.3164) [2021-04-16 01:26:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][380/1251] eta 0:04:20 lr 0.000695 time 0.2900 (0.2992) loss 3.8150 (3.6812) grad_norm 1.5025 (1.3169) [2021-04-16 01:26:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][390/1251] eta 0:04:17 lr 0.000695 time 0.2734 (0.2987) loss 3.8353 (3.6859) grad_norm 1.4909 (1.3203) [2021-04-16 01:26:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][400/1251] eta 0:04:13 lr 0.000695 time 0.2633 (0.2983) loss 3.8903 (3.6905) grad_norm 1.4433 (1.3194) [2021-04-16 01:26:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][410/1251] eta 0:04:10 lr 0.000695 time 0.3091 (0.2978) loss 2.8917 (3.6897) grad_norm 1.3491 (1.3200) [2021-04-16 01:26:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][420/1251] eta 0:04:07 lr 0.000695 time 0.2597 (0.2973) loss 4.1063 (3.6860) grad_norm 1.4951 (1.3217) [2021-04-16 01:26:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][430/1251] eta 0:04:04 lr 0.000695 time 0.3059 (0.2973) loss 4.0924 (3.6801) grad_norm 1.6122 (1.3206) [2021-04-16 01:26:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][440/1251] eta 0:04:00 lr 0.000695 time 0.2738 (0.2968) loss 3.8884 (3.6792) grad_norm 1.6376 (1.3206) [2021-04-16 01:26:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][450/1251] eta 0:03:57 lr 0.000695 time 0.2882 (0.2964) loss 4.0209 (3.6796) grad_norm 1.2519 (1.3209) [2021-04-16 01:26:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][460/1251] eta 0:03:54 lr 0.000695 time 0.2979 (0.2960) loss 4.5167 (3.6811) grad_norm 1.4366 (1.3239) [2021-04-16 01:26:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][470/1251] eta 0:03:50 lr 0.000695 time 0.2835 (0.2955) loss 4.6376 (3.6839) grad_norm 1.3453 (1.3252) [2021-04-16 01:26:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][480/1251] eta 0:03:47 lr 0.000695 time 0.2484 (0.2951) loss 3.4044 (3.6843) grad_norm 1.2967 (1.3248) [2021-04-16 01:26:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][490/1251] eta 0:03:44 lr 0.000695 time 0.2984 (0.2948) loss 3.8945 (3.6834) grad_norm 1.3940 (1.3227) [2021-04-16 01:26:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][500/1251] eta 0:03:41 lr 0.000695 time 0.2784 (0.2944) loss 3.7011 (3.6830) grad_norm 1.5324 (1.3222) [2021-04-16 01:26:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][510/1251] eta 0:03:37 lr 0.000695 time 0.2650 (0.2941) loss 4.0160 (3.6797) grad_norm 1.4918 (1.3235) [2021-04-16 01:26:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][520/1251] eta 0:03:34 lr 0.000695 time 0.2794 (0.2937) loss 3.5657 (3.6799) grad_norm 1.6745 (1.3247) [2021-04-16 01:26:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][530/1251] eta 0:03:31 lr 0.000695 time 0.2600 (0.2933) loss 3.0466 (3.6815) grad_norm 1.2121 (1.3236) [2021-04-16 01:26:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][540/1251] eta 0:03:28 lr 0.000695 time 0.2852 (0.2931) loss 3.9507 (3.6776) grad_norm 1.2885 (1.3241) [2021-04-16 01:26:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][550/1251] eta 0:03:25 lr 0.000695 time 0.3001 (0.2930) loss 4.0169 (3.6763) grad_norm 1.2246 (1.3239) [2021-04-16 01:27:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][560/1251] eta 0:03:22 lr 0.000695 time 0.2920 (0.2927) loss 3.8379 (3.6812) grad_norm 1.1321 (1.3232) [2021-04-16 01:27:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][570/1251] eta 0:03:19 lr 0.000695 time 0.2867 (0.2925) loss 3.8442 (3.6777) grad_norm 1.2746 (1.3221) [2021-04-16 01:27:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][580/1251] eta 0:03:16 lr 0.000695 time 0.2759 (0.2921) loss 3.0894 (3.6764) grad_norm 1.1550 (1.3216) [2021-04-16 01:27:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][590/1251] eta 0:03:13 lr 0.000695 time 0.4166 (0.2926) loss 3.1006 (3.6703) grad_norm 1.3242 (1.3224) [2021-04-16 01:27:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][600/1251] eta 0:03:10 lr 0.000695 time 0.2681 (0.2924) loss 4.3888 (3.6757) grad_norm 1.1868 (1.3217) [2021-04-16 01:27:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][610/1251] eta 0:03:07 lr 0.000694 time 0.2605 (0.2921) loss 3.5692 (3.6800) grad_norm 1.2313 (1.3224) [2021-04-16 01:27:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][620/1251] eta 0:03:04 lr 0.000694 time 0.2787 (0.2920) loss 4.1937 (3.6805) grad_norm 1.3854 (1.3225) [2021-04-16 01:27:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][630/1251] eta 0:03:01 lr 0.000694 time 0.2733 (0.2917) loss 3.9809 (3.6809) grad_norm 1.2291 (1.3226) [2021-04-16 01:27:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][640/1251] eta 0:02:58 lr 0.000694 time 0.2645 (0.2916) loss 3.8583 (3.6835) grad_norm 1.3435 (1.3225) [2021-04-16 01:27:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][650/1251] eta 0:02:55 lr 0.000694 time 0.2751 (0.2915) loss 3.6242 (3.6813) grad_norm 1.5745 (1.3236) [2021-04-16 01:27:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][660/1251] eta 0:02:52 lr 0.000694 time 0.2822 (0.2912) loss 3.3793 (3.6829) grad_norm 1.2195 (1.3244) [2021-04-16 01:27:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][670/1251] eta 0:02:49 lr 0.000694 time 0.2739 (0.2910) loss 4.9218 (3.6880) grad_norm 1.5432 (1.3257) [2021-04-16 01:27:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][680/1251] eta 0:02:46 lr 0.000694 time 0.2806 (0.2907) loss 3.7894 (3.6862) grad_norm 1.3067 (1.3265) [2021-04-16 01:27:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][690/1251] eta 0:02:42 lr 0.000694 time 0.2836 (0.2905) loss 3.4752 (3.6852) grad_norm 1.3389 (1.3262) [2021-04-16 01:27:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][700/1251] eta 0:02:39 lr 0.000694 time 0.2574 (0.2904) loss 3.6013 (3.6795) grad_norm 1.1778 (1.3254) [2021-04-16 01:27:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][710/1251] eta 0:02:37 lr 0.000694 time 0.2801 (0.2904) loss 2.9301 (3.6750) grad_norm 1.3227 (1.3243) [2021-04-16 01:27:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][720/1251] eta 0:02:34 lr 0.000694 time 0.2796 (0.2905) loss 3.2926 (3.6746) grad_norm 1.4059 (1.3257) [2021-04-16 01:27:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][730/1251] eta 0:02:31 lr 0.000694 time 0.2837 (0.2903) loss 2.7190 (3.6733) grad_norm 1.3228 (1.3268) [2021-04-16 01:27:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][740/1251] eta 0:02:28 lr 0.000694 time 0.2784 (0.2901) loss 4.3008 (3.6741) grad_norm 1.4421 (1.3285) [2021-04-16 01:27:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][750/1251] eta 0:02:25 lr 0.000694 time 0.2895 (0.2900) loss 2.4093 (3.6716) grad_norm 1.5433 (1.3288) [2021-04-16 01:27:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][760/1251] eta 0:02:22 lr 0.000694 time 0.2744 (0.2897) loss 2.9486 (3.6740) grad_norm 1.0674 (1.3290) [2021-04-16 01:28:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][770/1251] eta 0:02:19 lr 0.000694 time 0.2624 (0.2895) loss 3.6977 (3.6751) grad_norm 1.2199 (1.3292) [2021-04-16 01:28:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][780/1251] eta 0:02:16 lr 0.000694 time 0.2713 (0.2893) loss 3.6649 (3.6751) grad_norm 1.2217 (1.3291) [2021-04-16 01:28:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][790/1251] eta 0:02:13 lr 0.000694 time 0.2703 (0.2891) loss 3.0059 (3.6744) grad_norm 1.3292 (1.3298) [2021-04-16 01:28:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][800/1251] eta 0:02:10 lr 0.000694 time 0.2584 (0.2889) loss 4.0124 (3.6741) grad_norm 1.2427 (1.3299) [2021-04-16 01:28:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][810/1251] eta 0:02:07 lr 0.000694 time 0.2827 (0.2889) loss 3.9716 (3.6755) grad_norm 1.3533 (1.3315) [2021-04-16 01:28:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][820/1251] eta 0:02:04 lr 0.000694 time 0.2976 (0.2887) loss 2.8142 (3.6757) grad_norm 1.0965 (1.3306) [2021-04-16 01:28:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][830/1251] eta 0:02:01 lr 0.000694 time 0.2915 (0.2886) loss 4.2612 (3.6766) grad_norm 1.2145 (1.3295) [2021-04-16 01:28:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][840/1251] eta 0:01:58 lr 0.000694 time 0.2775 (0.2885) loss 3.8968 (3.6773) grad_norm 1.1082 (1.3288) [2021-04-16 01:28:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][850/1251] eta 0:01:55 lr 0.000694 time 0.2819 (0.2885) loss 3.7887 (3.6771) grad_norm 1.2488 (1.3285) [2021-04-16 01:28:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][860/1251] eta 0:01:52 lr 0.000694 time 0.2809 (0.2884) loss 3.3519 (3.6776) grad_norm 1.2298 (1.3281) [2021-04-16 01:28:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][870/1251] eta 0:01:49 lr 0.000693 time 0.2750 (0.2884) loss 4.4182 (3.6810) grad_norm 1.1637 (1.3283) [2021-04-16 01:28:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][880/1251] eta 0:01:46 lr 0.000693 time 0.2566 (0.2883) loss 3.8459 (3.6796) grad_norm 1.2254 (1.3280) [2021-04-16 01:28:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][890/1251] eta 0:01:44 lr 0.000693 time 0.2981 (0.2881) loss 3.3456 (3.6796) grad_norm 1.6936 (1.3285) [2021-04-16 01:28:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][900/1251] eta 0:01:41 lr 0.000693 time 0.2729 (0.2880) loss 4.3674 (3.6796) grad_norm 1.3427 (1.3279) [2021-04-16 01:28:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][910/1251] eta 0:01:38 lr 0.000693 time 0.2833 (0.2879) loss 3.8118 (3.6821) grad_norm 1.3789 (1.3275) [2021-04-16 01:28:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][920/1251] eta 0:01:35 lr 0.000693 time 0.3020 (0.2878) loss 2.4649 (3.6768) grad_norm 1.3615 (1.3275) [2021-04-16 01:28:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][930/1251] eta 0:01:32 lr 0.000693 time 0.2582 (0.2877) loss 4.6958 (3.6766) grad_norm 1.1634 (1.3274) [2021-04-16 01:28:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][940/1251] eta 0:01:29 lr 0.000693 time 0.4056 (0.2880) loss 4.3442 (3.6727) grad_norm 1.3380 (1.3273) [2021-04-16 01:28:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][950/1251] eta 0:01:26 lr 0.000693 time 0.2734 (0.2879) loss 3.5897 (3.6704) grad_norm 1.2372 (1.3274) [2021-04-16 01:28:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][960/1251] eta 0:01:23 lr 0.000693 time 0.2601 (0.2878) loss 3.6757 (3.6723) grad_norm 1.1794 (1.3284) [2021-04-16 01:28:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][970/1251] eta 0:01:20 lr 0.000693 time 0.2794 (0.2878) loss 2.4399 (3.6705) grad_norm 1.2526 (1.3285) [2021-04-16 01:28:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][980/1251] eta 0:01:17 lr 0.000693 time 0.2787 (0.2877) loss 3.5983 (3.6707) grad_norm 1.0608 (1.3282) [2021-04-16 01:29:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][990/1251] eta 0:01:15 lr 0.000693 time 0.2700 (0.2875) loss 4.0369 (3.6690) grad_norm 1.2303 (1.3271) [2021-04-16 01:29:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][1000/1251] eta 0:01:12 lr 0.000693 time 0.2558 (0.2874) loss 2.2848 (3.6671) grad_norm 1.1799 (1.3262) [2021-04-16 01:29:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][1010/1251] eta 0:01:09 lr 0.000693 time 0.2824 (0.2874) loss 3.9130 (3.6669) grad_norm 1.1624 (1.3253) [2021-04-16 01:29:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][1020/1251] eta 0:01:06 lr 0.000693 time 0.2885 (0.2873) loss 4.2132 (3.6689) grad_norm 1.4922 (1.3254) [2021-04-16 01:29:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][1030/1251] eta 0:01:03 lr 0.000693 time 0.2991 (0.2872) loss 4.3254 (3.6704) grad_norm 1.3657 (1.3254) [2021-04-16 01:29:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][1040/1251] eta 0:01:00 lr 0.000693 time 0.2700 (0.2871) loss 3.3217 (3.6670) grad_norm 1.7418 (1.3266) [2021-04-16 01:29:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][1050/1251] eta 0:00:57 lr 0.000693 time 0.2722 (0.2870) loss 4.0193 (3.6684) grad_norm 1.3594 (1.3271) [2021-04-16 01:29:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][1060/1251] eta 0:00:54 lr 0.000693 time 0.2802 (0.2869) loss 3.3095 (3.6661) grad_norm 1.1598 (1.3269) [2021-04-16 01:29:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][1070/1251] eta 0:00:51 lr 0.000693 time 0.2790 (0.2869) loss 2.7760 (3.6647) grad_norm 1.3730 (1.3268) [2021-04-16 01:29:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][1080/1251] eta 0:00:49 lr 0.000693 time 0.2974 (0.2868) loss 3.5780 (3.6656) grad_norm 1.2103 (1.3264) [2021-04-16 01:29:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][1090/1251] eta 0:00:46 lr 0.000693 time 0.2920 (0.2867) loss 3.0772 (3.6648) grad_norm 1.1942 (1.3260) [2021-04-16 01:29:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][1100/1251] eta 0:00:43 lr 0.000693 time 0.2768 (0.2866) loss 3.8805 (3.6651) grad_norm 1.1857 (1.3262) [2021-04-16 01:29:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][1110/1251] eta 0:00:40 lr 0.000693 time 0.2821 (0.2865) loss 3.0158 (3.6625) grad_norm 1.3261 (1.3263) [2021-04-16 01:29:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][1120/1251] eta 0:00:37 lr 0.000693 time 0.2776 (0.2865) loss 4.4914 (3.6630) grad_norm 1.3812 (1.3256) [2021-04-16 01:29:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][1130/1251] eta 0:00:34 lr 0.000692 time 0.2801 (0.2864) loss 3.4672 (3.6650) grad_norm 1.1919 (1.3257) [2021-04-16 01:29:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][1140/1251] eta 0:00:31 lr 0.000692 time 0.2750 (0.2864) loss 4.4480 (3.6660) grad_norm 1.3839 (1.3260) [2021-04-16 01:29:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][1150/1251] eta 0:00:28 lr 0.000692 time 0.2902 (0.2863) loss 3.1620 (3.6650) grad_norm 1.2952 (1.3261) [2021-04-16 01:29:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][1160/1251] eta 0:00:26 lr 0.000692 time 0.2783 (0.2863) loss 3.0754 (3.6640) grad_norm 1.2622 (1.3254) [2021-04-16 01:29:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][1170/1251] eta 0:00:23 lr 0.000692 time 0.2630 (0.2865) loss 4.3036 (3.6651) grad_norm 1.2603 (1.3251) [2021-04-16 01:29:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][1180/1251] eta 0:00:20 lr 0.000692 time 0.2772 (0.2864) loss 4.0694 (3.6676) grad_norm 1.3247 (1.3248) [2021-04-16 01:29:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][1190/1251] eta 0:00:17 lr 0.000692 time 0.2735 (0.2864) loss 4.1965 (3.6705) grad_norm 1.3579 (1.3253) [2021-04-16 01:30:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][1200/1251] eta 0:00:14 lr 0.000692 time 0.2793 (0.2864) loss 3.2456 (3.6697) grad_norm 1.6358 (1.3264) [2021-04-16 01:30:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][1210/1251] eta 0:00:11 lr 0.000692 time 0.2552 (0.2863) loss 2.7431 (3.6678) grad_norm 1.3230 (1.3269) [2021-04-16 01:30:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][1220/1251] eta 0:00:08 lr 0.000692 time 0.2941 (0.2863) loss 3.1894 (3.6677) grad_norm 1.4733 (1.3269) [2021-04-16 01:30:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][1230/1251] eta 0:00:06 lr 0.000692 time 0.2745 (0.2862) loss 3.2546 (3.6675) grad_norm 1.3014 (1.3271) [2021-04-16 01:30:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][1240/1251] eta 0:00:03 lr 0.000692 time 0.2479 (0.2861) loss 4.1710 (3.6660) grad_norm 1.2493 (1.3277) [2021-04-16 01:30:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [112/300][1250/1251] eta 0:00:00 lr 0.000692 time 0.2483 (0.2858) loss 4.3131 (3.6648) grad_norm 1.2867 (1.3276) [2021-04-16 01:30:18 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 112 training takes 0:06:00 [2021-04-16 01:30:18 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_112.pth saving...... [2021-04-16 01:30:35 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_112.pth saved !!! [2021-04-16 01:30:36 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.161 (1.161) Loss 1.1718 (1.1718) Acc@1 72.070 (72.070) Acc@5 91.992 (91.992) [2021-04-16 01:30:37 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.351 (0.223) Loss 1.1292 (1.0863) Acc@1 73.730 (74.556) Acc@5 91.992 (92.418) [2021-04-16 01:30:39 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.154 (0.213) Loss 1.1248 (1.0808) Acc@1 73.926 (74.554) Acc@5 91.992 (92.550) [2021-04-16 01:30:42 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.123 (0.229) Loss 1.1056 (1.0860) Acc@1 73.242 (74.505) Acc@5 92.676 (92.496) [2021-04-16 01:30:43 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.212) Loss 1.0671 (1.0912) Acc@1 74.609 (74.428) Acc@5 92.578 (92.433) [2021-04-16 01:30:49 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 74.322 Acc@5 92.356 [2021-04-16 01:30:49 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 74.3% [2021-04-16 01:30:49 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 74.32% [2021-04-16 01:30:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][0/1251] eta 1:02:56 lr 0.000692 time 3.0186 (3.0186) loss 4.1296 (4.1296) grad_norm 1.1849 (1.1849) [2021-04-16 01:30:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][10/1251] eta 0:10:54 lr 0.000692 time 0.2796 (0.5273) loss 4.3097 (3.4549) grad_norm 1.4049 (1.3157) [2021-04-16 01:30:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][20/1251] eta 0:08:32 lr 0.000692 time 0.3107 (0.4163) loss 4.2463 (3.6269) grad_norm 1.3436 (1.2991) [2021-04-16 01:31:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][30/1251] eta 0:07:30 lr 0.000692 time 0.2735 (0.3692) loss 4.4215 (3.6475) grad_norm 1.1762 (1.3357) [2021-04-16 01:31:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][40/1251] eta 0:06:59 lr 0.000692 time 0.2677 (0.3465) loss 3.4162 (3.6763) grad_norm 1.2265 (1.3336) [2021-04-16 01:31:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][50/1251] eta 0:06:40 lr 0.000692 time 0.2994 (0.3335) loss 4.1464 (3.6759) grad_norm 1.4420 (1.3285) [2021-04-16 01:31:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][60/1251] eta 0:06:27 lr 0.000692 time 0.2809 (0.3251) loss 2.7721 (3.6742) grad_norm 1.6063 (1.3281) [2021-04-16 01:31:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][70/1251] eta 0:06:17 lr 0.000692 time 0.3061 (0.3196) loss 2.9972 (3.6522) grad_norm 1.4198 (1.3281) [2021-04-16 01:31:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][80/1251] eta 0:06:09 lr 0.000692 time 0.2847 (0.3151) loss 3.7672 (3.6630) grad_norm 1.2626 (1.3197) [2021-04-16 01:31:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][90/1251] eta 0:06:02 lr 0.000692 time 0.2634 (0.3122) loss 4.0782 (3.6891) grad_norm 1.3414 (1.3163) [2021-04-16 01:31:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][100/1251] eta 0:05:55 lr 0.000692 time 0.2624 (0.3087) loss 4.0585 (3.6910) grad_norm 1.2999 (1.3150) [2021-04-16 01:31:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][110/1251] eta 0:05:49 lr 0.000692 time 0.2708 (0.3063) loss 3.2656 (3.6867) grad_norm 1.5908 (1.3152) [2021-04-16 01:31:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][120/1251] eta 0:05:43 lr 0.000692 time 0.2874 (0.3039) loss 3.3392 (3.6635) grad_norm 1.2623 (1.3198) [2021-04-16 01:31:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][130/1251] eta 0:05:39 lr 0.000692 time 0.2607 (0.3026) loss 4.0650 (3.6802) grad_norm 1.4688 (1.3193) [2021-04-16 01:31:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][140/1251] eta 0:05:36 lr 0.000691 time 0.2789 (0.3027) loss 3.9082 (3.6718) grad_norm 1.2227 (1.3170) [2021-04-16 01:31:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][150/1251] eta 0:05:32 lr 0.000691 time 0.4466 (0.3023) loss 3.8895 (3.6543) grad_norm 1.4627 (1.3184) [2021-04-16 01:31:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][160/1251] eta 0:05:27 lr 0.000691 time 0.2756 (0.3004) loss 3.3078 (3.6553) grad_norm 1.3791 (1.3216) [2021-04-16 01:31:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][170/1251] eta 0:05:23 lr 0.000691 time 0.2488 (0.2996) loss 3.9419 (3.6502) grad_norm 1.2423 (1.3201) [2021-04-16 01:31:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][180/1251] eta 0:05:19 lr 0.000691 time 0.3026 (0.2985) loss 4.4226 (3.6592) grad_norm 1.4686 (1.3200) [2021-04-16 01:31:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][190/1251] eta 0:05:15 lr 0.000691 time 0.2982 (0.2974) loss 3.9333 (3.6665) grad_norm 1.2356 (inf) [2021-04-16 01:31:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][200/1251] eta 0:05:11 lr 0.000691 time 0.2798 (0.2964) loss 4.3307 (3.6662) grad_norm 1.3116 (inf) [2021-04-16 01:31:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][210/1251] eta 0:05:07 lr 0.000691 time 0.2866 (0.2955) loss 3.8432 (3.6616) grad_norm 1.5168 (inf) [2021-04-16 01:31:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][220/1251] eta 0:05:04 lr 0.000691 time 0.2683 (0.2953) loss 3.5618 (3.6718) grad_norm 1.1181 (inf) [2021-04-16 01:31:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][230/1251] eta 0:05:00 lr 0.000691 time 0.2850 (0.2946) loss 3.6837 (3.6813) grad_norm 1.4689 (inf) [2021-04-16 01:32:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][240/1251] eta 0:04:57 lr 0.000691 time 0.2914 (0.2939) loss 3.9001 (3.6790) grad_norm 1.2963 (inf) [2021-04-16 01:32:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][250/1251] eta 0:04:53 lr 0.000691 time 0.2805 (0.2932) loss 3.4992 (3.6864) grad_norm 1.1374 (inf) [2021-04-16 01:32:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][260/1251] eta 0:04:49 lr 0.000691 time 0.2648 (0.2925) loss 3.7307 (3.6792) grad_norm 1.4500 (inf) [2021-04-16 01:32:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][270/1251] eta 0:04:46 lr 0.000691 time 0.2860 (0.2919) loss 3.8226 (3.6722) grad_norm 1.2506 (inf) [2021-04-16 01:32:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][280/1251] eta 0:04:43 lr 0.000691 time 0.3014 (0.2916) loss 3.7933 (3.6634) grad_norm 1.2090 (inf) [2021-04-16 01:32:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][290/1251] eta 0:04:39 lr 0.000691 time 0.2963 (0.2913) loss 3.9688 (3.6706) grad_norm 1.2105 (inf) [2021-04-16 01:32:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][300/1251] eta 0:04:36 lr 0.000691 time 0.2908 (0.2910) loss 4.0310 (3.6776) grad_norm 1.2609 (inf) [2021-04-16 01:32:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][310/1251] eta 0:04:33 lr 0.000691 time 0.2941 (0.2907) loss 4.1368 (3.6710) grad_norm 1.2142 (inf) [2021-04-16 01:32:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][320/1251] eta 0:04:30 lr 0.000691 time 0.2846 (0.2909) loss 2.4651 (3.6643) grad_norm 1.1730 (inf) [2021-04-16 01:32:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][330/1251] eta 0:04:27 lr 0.000691 time 0.2534 (0.2904) loss 4.0307 (3.6654) grad_norm 1.3390 (inf) [2021-04-16 01:32:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][340/1251] eta 0:04:24 lr 0.000691 time 0.2738 (0.2900) loss 4.0758 (3.6653) grad_norm 1.2521 (inf) [2021-04-16 01:32:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][350/1251] eta 0:04:20 lr 0.000691 time 0.2888 (0.2896) loss 3.5977 (3.6721) grad_norm 1.4479 (inf) [2021-04-16 01:32:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][360/1251] eta 0:04:17 lr 0.000691 time 0.3060 (0.2894) loss 3.2857 (3.6712) grad_norm 1.1973 (inf) [2021-04-16 01:32:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][370/1251] eta 0:04:15 lr 0.000691 time 0.2571 (0.2895) loss 3.7935 (3.6628) grad_norm 1.2317 (inf) [2021-04-16 01:32:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][380/1251] eta 0:04:11 lr 0.000691 time 0.2718 (0.2892) loss 3.9177 (3.6693) grad_norm 1.4077 (inf) [2021-04-16 01:32:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][390/1251] eta 0:04:08 lr 0.000691 time 0.2760 (0.2890) loss 3.8400 (3.6700) grad_norm 1.6738 (inf) [2021-04-16 01:32:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][400/1251] eta 0:04:05 lr 0.000690 time 0.2740 (0.2887) loss 3.2164 (3.6724) grad_norm 1.3697 (inf) [2021-04-16 01:32:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][410/1251] eta 0:04:02 lr 0.000690 time 0.2746 (0.2885) loss 3.8715 (3.6777) grad_norm 1.6814 (inf) [2021-04-16 01:32:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][420/1251] eta 0:03:59 lr 0.000690 time 0.2887 (0.2886) loss 3.6031 (3.6752) grad_norm 1.2494 (inf) [2021-04-16 01:32:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][430/1251] eta 0:03:56 lr 0.000690 time 0.2927 (0.2883) loss 4.0661 (3.6733) grad_norm 1.3993 (inf) [2021-04-16 01:32:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][440/1251] eta 0:03:53 lr 0.000690 time 0.2789 (0.2881) loss 4.2264 (3.6806) grad_norm 1.2732 (inf) [2021-04-16 01:32:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][450/1251] eta 0:03:50 lr 0.000690 time 0.2650 (0.2880) loss 4.2128 (3.6823) grad_norm 1.3902 (inf) [2021-04-16 01:33:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][460/1251] eta 0:03:47 lr 0.000690 time 0.2715 (0.2877) loss 2.8559 (3.6790) grad_norm 1.2953 (inf) [2021-04-16 01:33:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][470/1251] eta 0:03:44 lr 0.000690 time 0.2610 (0.2874) loss 2.9151 (3.6724) grad_norm 1.3108 (inf) [2021-04-16 01:33:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][480/1251] eta 0:03:41 lr 0.000690 time 0.2928 (0.2873) loss 2.9072 (3.6727) grad_norm 1.2191 (inf) [2021-04-16 01:33:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][490/1251] eta 0:03:38 lr 0.000690 time 0.2523 (0.2870) loss 2.7557 (3.6653) grad_norm 1.2942 (inf) [2021-04-16 01:33:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][500/1251] eta 0:03:35 lr 0.000690 time 0.2685 (0.2869) loss 4.1575 (3.6648) grad_norm 1.4100 (inf) [2021-04-16 01:33:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][510/1251] eta 0:03:32 lr 0.000690 time 0.2527 (0.2866) loss 3.9567 (3.6641) grad_norm 1.3430 (inf) [2021-04-16 01:33:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][520/1251] eta 0:03:29 lr 0.000690 time 0.2897 (0.2867) loss 3.5601 (3.6650) grad_norm 1.3944 (inf) [2021-04-16 01:33:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][530/1251] eta 0:03:26 lr 0.000690 time 0.2870 (0.2866) loss 4.0318 (3.6666) grad_norm 1.5107 (inf) [2021-04-16 01:33:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][540/1251] eta 0:03:23 lr 0.000690 time 0.2842 (0.2866) loss 3.4333 (3.6624) grad_norm 1.2154 (inf) [2021-04-16 01:33:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][550/1251] eta 0:03:20 lr 0.000690 time 0.2726 (0.2865) loss 2.6715 (3.6550) grad_norm 1.1921 (inf) [2021-04-16 01:33:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][560/1251] eta 0:03:17 lr 0.000690 time 0.2651 (0.2864) loss 4.4758 (3.6538) grad_norm 1.3748 (inf) [2021-04-16 01:33:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][570/1251] eta 0:03:15 lr 0.000690 time 0.2641 (0.2867) loss 3.0218 (3.6567) grad_norm 1.2449 (inf) [2021-04-16 01:33:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][580/1251] eta 0:03:12 lr 0.000690 time 0.2675 (0.2866) loss 3.2739 (3.6560) grad_norm 1.2404 (inf) [2021-04-16 01:33:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][590/1251] eta 0:03:09 lr 0.000690 time 0.2977 (0.2868) loss 4.3022 (3.6551) grad_norm 1.3435 (inf) [2021-04-16 01:33:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][600/1251] eta 0:03:06 lr 0.000690 time 0.3147 (0.2866) loss 3.8534 (3.6579) grad_norm 1.5200 (inf) [2021-04-16 01:33:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][610/1251] eta 0:03:03 lr 0.000690 time 0.2481 (0.2865) loss 3.4520 (3.6535) grad_norm 1.2713 (inf) [2021-04-16 01:33:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][620/1251] eta 0:03:00 lr 0.000690 time 0.2598 (0.2862) loss 2.6220 (3.6498) grad_norm 1.4113 (inf) [2021-04-16 01:33:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][630/1251] eta 0:02:57 lr 0.000690 time 0.2585 (0.2860) loss 3.7551 (3.6488) grad_norm 1.2197 (inf) [2021-04-16 01:33:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][640/1251] eta 0:02:54 lr 0.000690 time 0.2659 (0.2860) loss 2.2865 (3.6471) grad_norm 1.4653 (inf) [2021-04-16 01:33:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][650/1251] eta 0:02:51 lr 0.000690 time 0.2597 (0.2860) loss 3.5556 (3.6436) grad_norm 1.3428 (inf) [2021-04-16 01:33:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][660/1251] eta 0:02:48 lr 0.000689 time 0.2574 (0.2858) loss 3.3519 (3.6486) grad_norm 1.2078 (inf) [2021-04-16 01:34:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][670/1251] eta 0:02:46 lr 0.000689 time 0.2726 (0.2857) loss 3.3456 (3.6465) grad_norm 1.3778 (inf) [2021-04-16 01:34:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][680/1251] eta 0:02:43 lr 0.000689 time 0.2989 (0.2856) loss 4.2190 (3.6478) grad_norm 1.4346 (inf) [2021-04-16 01:34:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][690/1251] eta 0:02:40 lr 0.000689 time 0.2678 (0.2854) loss 3.6765 (3.6488) grad_norm 1.1651 (inf) [2021-04-16 01:34:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][700/1251] eta 0:02:37 lr 0.000689 time 0.2893 (0.2852) loss 2.9540 (3.6505) grad_norm 1.2809 (inf) [2021-04-16 01:34:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][710/1251] eta 0:02:34 lr 0.000689 time 0.2801 (0.2853) loss 2.6585 (3.6502) grad_norm 1.3615 (inf) [2021-04-16 01:34:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][720/1251] eta 0:02:31 lr 0.000689 time 0.2852 (0.2855) loss 3.2398 (3.6511) grad_norm 1.1446 (inf) [2021-04-16 01:34:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][730/1251] eta 0:02:28 lr 0.000689 time 0.2555 (0.2856) loss 4.3578 (3.6515) grad_norm 1.4264 (inf) [2021-04-16 01:34:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][740/1251] eta 0:02:25 lr 0.000689 time 0.2788 (0.2855) loss 3.2617 (3.6541) grad_norm 1.2886 (inf) [2021-04-16 01:34:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][750/1251] eta 0:02:22 lr 0.000689 time 0.2942 (0.2854) loss 2.8128 (3.6522) grad_norm 1.2087 (inf) [2021-04-16 01:34:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][760/1251] eta 0:02:20 lr 0.000689 time 0.2835 (0.2852) loss 3.1950 (3.6481) grad_norm 1.1961 (inf) [2021-04-16 01:34:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][770/1251] eta 0:02:17 lr 0.000689 time 0.2922 (0.2852) loss 3.9421 (3.6474) grad_norm 1.5172 (inf) [2021-04-16 01:34:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][780/1251] eta 0:02:14 lr 0.000689 time 0.2883 (0.2852) loss 4.0098 (3.6469) grad_norm 1.3256 (inf) [2021-04-16 01:34:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][790/1251] eta 0:02:11 lr 0.000689 time 0.2787 (0.2853) loss 4.5516 (3.6505) grad_norm 1.4441 (inf) [2021-04-16 01:34:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][800/1251] eta 0:02:08 lr 0.000689 time 0.2938 (0.2852) loss 3.3340 (3.6494) grad_norm 1.5067 (inf) [2021-04-16 01:34:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][810/1251] eta 0:02:05 lr 0.000689 time 0.2613 (0.2851) loss 3.8624 (3.6443) grad_norm 1.2022 (inf) [2021-04-16 01:34:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][820/1251] eta 0:02:02 lr 0.000689 time 0.2581 (0.2850) loss 2.7479 (3.6420) grad_norm 1.3802 (inf) [2021-04-16 01:34:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][830/1251] eta 0:01:59 lr 0.000689 time 0.2980 (0.2849) loss 3.7442 (3.6392) grad_norm 1.3158 (inf) [2021-04-16 01:34:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][840/1251] eta 0:01:57 lr 0.000689 time 0.2962 (0.2849) loss 2.5328 (3.6378) grad_norm 1.2790 (inf) [2021-04-16 01:34:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][850/1251] eta 0:01:54 lr 0.000689 time 0.2557 (0.2848) loss 4.1755 (3.6385) grad_norm 1.1524 (inf) [2021-04-16 01:34:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][860/1251] eta 0:01:51 lr 0.000689 time 0.2442 (0.2846) loss 4.4540 (3.6403) grad_norm 1.1872 (inf) [2021-04-16 01:34:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][870/1251] eta 0:01:48 lr 0.000689 time 0.2803 (0.2845) loss 3.8763 (3.6398) grad_norm 1.3308 (inf) [2021-04-16 01:35:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][880/1251] eta 0:01:45 lr 0.000689 time 0.2974 (0.2845) loss 4.4521 (3.6434) grad_norm 1.2004 (inf) [2021-04-16 01:35:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][890/1251] eta 0:01:42 lr 0.000689 time 0.2935 (0.2845) loss 3.5923 (3.6444) grad_norm 1.4940 (inf) [2021-04-16 01:35:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][900/1251] eta 0:01:39 lr 0.000689 time 0.2883 (0.2844) loss 3.1448 (3.6443) grad_norm 1.0967 (inf) [2021-04-16 01:35:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][910/1251] eta 0:01:36 lr 0.000689 time 0.3116 (0.2844) loss 4.1335 (3.6452) grad_norm 1.2221 (inf) [2021-04-16 01:35:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][920/1251] eta 0:01:34 lr 0.000688 time 0.2619 (0.2842) loss 4.1803 (3.6467) grad_norm 1.3362 (inf) [2021-04-16 01:35:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][930/1251] eta 0:01:31 lr 0.000688 time 0.2671 (0.2841) loss 3.8632 (3.6481) grad_norm 1.3904 (inf) [2021-04-16 01:35:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][940/1251] eta 0:01:28 lr 0.000688 time 0.4045 (0.2842) loss 2.6156 (3.6491) grad_norm 1.2326 (inf) [2021-04-16 01:35:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][950/1251] eta 0:01:25 lr 0.000688 time 0.2888 (0.2842) loss 4.3046 (3.6536) grad_norm 1.3408 (inf) [2021-04-16 01:35:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][960/1251] eta 0:01:22 lr 0.000688 time 0.2874 (0.2841) loss 3.6894 (3.6546) grad_norm 1.1953 (inf) [2021-04-16 01:35:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][970/1251] eta 0:01:19 lr 0.000688 time 0.2752 (0.2840) loss 3.6751 (3.6521) grad_norm 1.3177 (inf) [2021-04-16 01:35:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][980/1251] eta 0:01:16 lr 0.000688 time 0.2735 (0.2840) loss 3.0000 (3.6533) grad_norm 1.5189 (inf) [2021-04-16 01:35:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][990/1251] eta 0:01:14 lr 0.000688 time 0.2748 (0.2840) loss 4.1807 (3.6522) grad_norm 1.4351 (inf) [2021-04-16 01:35:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][1000/1251] eta 0:01:11 lr 0.000688 time 0.2685 (0.2839) loss 4.0629 (3.6522) grad_norm 1.2521 (inf) [2021-04-16 01:35:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][1010/1251] eta 0:01:08 lr 0.000688 time 0.2593 (0.2838) loss 3.1336 (3.6525) grad_norm 1.2567 (inf) [2021-04-16 01:35:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][1020/1251] eta 0:01:05 lr 0.000688 time 0.2689 (0.2838) loss 4.0708 (3.6529) grad_norm 1.4519 (inf) [2021-04-16 01:35:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][1030/1251] eta 0:01:02 lr 0.000688 time 0.2537 (0.2837) loss 3.3231 (3.6531) grad_norm 1.2104 (inf) [2021-04-16 01:35:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][1040/1251] eta 0:00:59 lr 0.000688 time 0.2977 (0.2837) loss 3.9138 (3.6547) grad_norm 1.1504 (inf) [2021-04-16 01:35:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][1050/1251] eta 0:00:57 lr 0.000688 time 0.3109 (0.2836) loss 3.7217 (3.6548) grad_norm 1.2456 (inf) [2021-04-16 01:35:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][1060/1251] eta 0:00:54 lr 0.000688 time 0.2782 (0.2835) loss 3.9065 (3.6551) grad_norm 1.5065 (inf) [2021-04-16 01:35:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][1070/1251] eta 0:00:51 lr 0.000688 time 0.2792 (0.2834) loss 3.1948 (3.6546) grad_norm 1.5050 (inf) [2021-04-16 01:35:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][1080/1251] eta 0:00:48 lr 0.000688 time 0.2908 (0.2834) loss 4.4340 (3.6581) grad_norm 1.2507 (inf) [2021-04-16 01:35:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][1090/1251] eta 0:00:45 lr 0.000688 time 0.2834 (0.2834) loss 3.9356 (3.6565) grad_norm 1.1188 (inf) [2021-04-16 01:36:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][1100/1251] eta 0:00:42 lr 0.000688 time 0.2569 (0.2833) loss 4.2702 (3.6564) grad_norm 1.3009 (inf) [2021-04-16 01:36:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][1110/1251] eta 0:00:39 lr 0.000688 time 0.2859 (0.2832) loss 4.2489 (3.6566) grad_norm 1.4118 (inf) [2021-04-16 01:36:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][1120/1251] eta 0:00:37 lr 0.000688 time 0.2551 (0.2833) loss 3.6489 (3.6575) grad_norm 1.4231 (inf) [2021-04-16 01:36:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][1130/1251] eta 0:00:34 lr 0.000688 time 0.2520 (0.2833) loss 2.6253 (3.6554) grad_norm 1.1358 (inf) [2021-04-16 01:36:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][1140/1251] eta 0:00:31 lr 0.000688 time 0.2937 (0.2832) loss 3.2099 (3.6558) grad_norm 1.3681 (inf) [2021-04-16 01:36:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][1150/1251] eta 0:00:28 lr 0.000688 time 0.2662 (0.2833) loss 4.3353 (3.6572) grad_norm 1.4693 (inf) [2021-04-16 01:36:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][1160/1251] eta 0:00:25 lr 0.000688 time 0.3069 (0.2832) loss 4.3052 (3.6558) grad_norm 1.2620 (inf) [2021-04-16 01:36:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][1170/1251] eta 0:00:22 lr 0.000688 time 0.2927 (0.2834) loss 4.3665 (3.6583) grad_norm 1.2065 (inf) [2021-04-16 01:36:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][1180/1251] eta 0:00:20 lr 0.000687 time 0.2938 (0.2833) loss 2.7076 (3.6594) grad_norm 1.4740 (inf) [2021-04-16 01:36:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][1190/1251] eta 0:00:17 lr 0.000687 time 0.2600 (0.2833) loss 3.8876 (3.6580) grad_norm 1.4789 (inf) [2021-04-16 01:36:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][1200/1251] eta 0:00:14 lr 0.000687 time 0.2619 (0.2833) loss 2.8252 (3.6602) grad_norm 1.3879 (inf) [2021-04-16 01:36:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][1210/1251] eta 0:00:11 lr 0.000687 time 0.2878 (0.2832) loss 3.8149 (3.6581) grad_norm 1.3029 (inf) [2021-04-16 01:36:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][1220/1251] eta 0:00:08 lr 0.000687 time 0.2918 (0.2832) loss 2.9191 (3.6572) grad_norm 1.1654 (inf) [2021-04-16 01:36:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][1230/1251] eta 0:00:05 lr 0.000687 time 0.2874 (0.2832) loss 3.0070 (3.6566) grad_norm 1.3797 (inf) [2021-04-16 01:36:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][1240/1251] eta 0:00:03 lr 0.000687 time 0.2482 (0.2831) loss 4.5450 (3.6544) grad_norm 1.3303 (inf) [2021-04-16 01:36:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [113/300][1250/1251] eta 0:00:00 lr 0.000687 time 0.2503 (0.2829) loss 3.7471 (3.6565) grad_norm 1.3596 (inf) [2021-04-16 01:36:46 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 113 training takes 0:05:56 [2021-04-16 01:36:46 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_113.pth saving...... [2021-04-16 01:36:59 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_113.pth saved !!! [2021-04-16 01:37:00 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.123 (1.123) Loss 1.0620 (1.0620) Acc@1 75.977 (75.977) Acc@5 91.895 (91.895) [2021-04-16 01:37:01 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.117 (0.229) Loss 1.1123 (1.0725) Acc@1 73.242 (74.805) Acc@5 92.090 (92.640) [2021-04-16 01:37:04 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.309 (0.247) Loss 1.0683 (1.0698) Acc@1 75.684 (75.060) Acc@5 92.285 (92.578) [2021-04-16 01:37:06 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.212 (0.241) Loss 1.0762 (1.0735) Acc@1 73.926 (74.729) Acc@5 92.578 (92.562) [2021-04-16 01:37:08 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.217) Loss 1.1847 (1.0801) Acc@1 73.340 (74.562) Acc@5 91.211 (92.476) [2021-04-16 01:37:13 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 74.452 Acc@5 92.438 [2021-04-16 01:37:13 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 74.5% [2021-04-16 01:37:13 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 74.45% [2021-04-16 01:37:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][0/1251] eta 0:56:34 lr 0.000687 time 2.7135 (2.7135) loss 3.1155 (3.1155) grad_norm 1.3415 (1.3415) [2021-04-16 01:37:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][10/1251] eta 0:10:24 lr 0.000687 time 0.3016 (0.5031) loss 3.3964 (3.3732) grad_norm 1.4230 (1.3168) [2021-04-16 01:37:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][20/1251] eta 0:08:19 lr 0.000687 time 0.2945 (0.4056) loss 3.9567 (3.6239) grad_norm 1.2637 (1.3249) [2021-04-16 01:37:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][30/1251] eta 0:07:26 lr 0.000687 time 0.2547 (0.3659) loss 3.4910 (3.6716) grad_norm 1.2275 (1.3465) [2021-04-16 01:37:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][40/1251] eta 0:06:57 lr 0.000687 time 0.2583 (0.3446) loss 3.3088 (3.6847) grad_norm 1.2695 (1.3549) [2021-04-16 01:37:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][50/1251] eta 0:06:38 lr 0.000687 time 0.2907 (0.3315) loss 3.6337 (3.6810) grad_norm 1.2461 (1.3489) [2021-04-16 01:37:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][60/1251] eta 0:06:25 lr 0.000687 time 0.2719 (0.3238) loss 3.6224 (3.6295) grad_norm 1.4018 (1.3516) [2021-04-16 01:37:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][70/1251] eta 0:06:15 lr 0.000687 time 0.2989 (0.3177) loss 2.8315 (3.6316) grad_norm 1.4458 (1.3611) [2021-04-16 01:37:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][80/1251] eta 0:06:07 lr 0.000687 time 0.2852 (0.3136) loss 3.1189 (3.6441) grad_norm 1.3165 (1.3700) [2021-04-16 01:37:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][90/1251] eta 0:06:01 lr 0.000687 time 0.2917 (0.3110) loss 3.8429 (3.6604) grad_norm 1.3357 (1.3655) [2021-04-16 01:37:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][100/1251] eta 0:05:54 lr 0.000687 time 0.2805 (0.3079) loss 2.7103 (3.6058) grad_norm 1.2823 (1.3577) [2021-04-16 01:37:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][110/1251] eta 0:05:48 lr 0.000687 time 0.2816 (0.3056) loss 3.5142 (3.6249) grad_norm 1.5136 (1.3563) [2021-04-16 01:37:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][120/1251] eta 0:05:43 lr 0.000687 time 0.2882 (0.3034) loss 2.6806 (3.6233) grad_norm 1.1923 (1.3528) [2021-04-16 01:37:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][130/1251] eta 0:05:37 lr 0.000687 time 0.2524 (0.3014) loss 2.5646 (3.6012) grad_norm 1.2966 (1.3557) [2021-04-16 01:37:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][140/1251] eta 0:05:33 lr 0.000687 time 0.2717 (0.3001) loss 3.1187 (3.5943) grad_norm 1.1844 (1.3600) [2021-04-16 01:37:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][150/1251] eta 0:05:29 lr 0.000687 time 0.4001 (0.2996) loss 3.6887 (3.5874) grad_norm 1.4641 (1.3593) [2021-04-16 01:38:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][160/1251] eta 0:05:26 lr 0.000687 time 0.2991 (0.2993) loss 4.5088 (3.6098) grad_norm 1.2299 (1.3561) [2021-04-16 01:38:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][170/1251] eta 0:05:22 lr 0.000687 time 0.2860 (0.2979) loss 3.2539 (3.6139) grad_norm 1.3983 (1.3470) [2021-04-16 01:38:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][180/1251] eta 0:05:18 lr 0.000687 time 0.2738 (0.2972) loss 3.5891 (3.6075) grad_norm 1.2954 (1.3438) [2021-04-16 01:38:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][190/1251] eta 0:05:14 lr 0.000686 time 0.2631 (0.2961) loss 3.5456 (3.6169) grad_norm 1.1456 (1.3399) [2021-04-16 01:38:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][200/1251] eta 0:05:11 lr 0.000686 time 0.2540 (0.2960) loss 2.8621 (3.5985) grad_norm 1.1954 (1.3423) [2021-04-16 01:38:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][210/1251] eta 0:05:07 lr 0.000686 time 0.2826 (0.2956) loss 3.6698 (3.6011) grad_norm 1.4782 (1.3415) [2021-04-16 01:38:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][220/1251] eta 0:05:04 lr 0.000686 time 0.2957 (0.2950) loss 3.6885 (3.5982) grad_norm 1.2748 (1.3369) [2021-04-16 01:38:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][230/1251] eta 0:05:00 lr 0.000686 time 0.3161 (0.2945) loss 3.1153 (3.5927) grad_norm 1.3810 (1.3355) [2021-04-16 01:38:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][240/1251] eta 0:04:56 lr 0.000686 time 0.2890 (0.2937) loss 4.5115 (3.6141) grad_norm 1.3725 (1.3331) [2021-04-16 01:38:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][250/1251] eta 0:04:53 lr 0.000686 time 0.2943 (0.2929) loss 3.8407 (3.6160) grad_norm 1.3065 (1.3318) [2021-04-16 01:38:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][260/1251] eta 0:04:49 lr 0.000686 time 0.2740 (0.2923) loss 3.6410 (3.6254) grad_norm 1.4017 (1.3368) [2021-04-16 01:38:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][270/1251] eta 0:04:46 lr 0.000686 time 0.2955 (0.2919) loss 3.9321 (3.6240) grad_norm 1.2503 (1.3366) [2021-04-16 01:38:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][280/1251] eta 0:04:43 lr 0.000686 time 0.2975 (0.2917) loss 3.3064 (3.6264) grad_norm 1.1002 (1.3339) [2021-04-16 01:38:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][290/1251] eta 0:04:39 lr 0.000686 time 0.2673 (0.2912) loss 4.2654 (3.6238) grad_norm 1.2173 (1.3337) [2021-04-16 01:38:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][300/1251] eta 0:04:36 lr 0.000686 time 0.2714 (0.2907) loss 2.1695 (3.6128) grad_norm 1.2061 (1.3329) [2021-04-16 01:38:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][310/1251] eta 0:04:33 lr 0.000686 time 0.2832 (0.2901) loss 4.1229 (3.6203) grad_norm 1.2792 (1.3315) [2021-04-16 01:38:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][320/1251] eta 0:04:30 lr 0.000686 time 0.2997 (0.2903) loss 3.5732 (3.6201) grad_norm 1.1080 (1.3321) [2021-04-16 01:38:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][330/1251] eta 0:04:27 lr 0.000686 time 0.2780 (0.2899) loss 3.9546 (3.6263) grad_norm 1.4987 (1.3320) [2021-04-16 01:38:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][340/1251] eta 0:04:24 lr 0.000686 time 0.2864 (0.2898) loss 3.9971 (3.6332) grad_norm 1.5397 (1.3319) [2021-04-16 01:38:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][350/1251] eta 0:04:21 lr 0.000686 time 0.2824 (0.2899) loss 3.0279 (3.6396) grad_norm 1.0987 (1.3299) [2021-04-16 01:38:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][360/1251] eta 0:04:18 lr 0.000686 time 0.2576 (0.2899) loss 3.8101 (3.6336) grad_norm 1.3057 (1.3276) [2021-04-16 01:39:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][370/1251] eta 0:04:15 lr 0.000686 time 0.2708 (0.2901) loss 3.9612 (3.6369) grad_norm 1.1821 (1.3271) [2021-04-16 01:39:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][380/1251] eta 0:04:12 lr 0.000686 time 0.2673 (0.2896) loss 4.0565 (3.6415) grad_norm 1.6073 (1.3259) [2021-04-16 01:39:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][390/1251] eta 0:04:09 lr 0.000686 time 0.2898 (0.2893) loss 4.3776 (3.6461) grad_norm 1.4428 (1.3264) [2021-04-16 01:39:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][400/1251] eta 0:04:05 lr 0.000686 time 0.2555 (0.2889) loss 3.1914 (3.6493) grad_norm 1.1263 (1.3272) [2021-04-16 01:39:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][410/1251] eta 0:04:02 lr 0.000686 time 0.2707 (0.2886) loss 3.7299 (3.6484) grad_norm 1.4125 (1.3259) [2021-04-16 01:39:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][420/1251] eta 0:03:59 lr 0.000686 time 0.2689 (0.2884) loss 3.4281 (3.6473) grad_norm 1.2379 (1.3285) [2021-04-16 01:39:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][430/1251] eta 0:03:56 lr 0.000686 time 0.2600 (0.2884) loss 4.1315 (3.6513) grad_norm 1.4018 (1.3291) [2021-04-16 01:39:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][440/1251] eta 0:03:53 lr 0.000686 time 0.2764 (0.2884) loss 3.8185 (3.6495) grad_norm 1.5189 (1.3298) [2021-04-16 01:39:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][450/1251] eta 0:03:50 lr 0.000685 time 0.2751 (0.2881) loss 2.9176 (3.6527) grad_norm 1.2797 (1.3306) [2021-04-16 01:39:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][460/1251] eta 0:03:47 lr 0.000685 time 0.2429 (0.2881) loss 4.0445 (3.6582) grad_norm 1.3682 (1.3292) [2021-04-16 01:39:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][470/1251] eta 0:03:44 lr 0.000685 time 0.2940 (0.2879) loss 2.4306 (3.6519) grad_norm 1.3200 (1.3298) [2021-04-16 01:39:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][480/1251] eta 0:03:41 lr 0.000685 time 0.2898 (0.2878) loss 3.5790 (3.6486) grad_norm 1.2295 (1.3304) [2021-04-16 01:39:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][490/1251] eta 0:03:38 lr 0.000685 time 0.2973 (0.2876) loss 2.7158 (3.6523) grad_norm 1.4315 (1.3312) [2021-04-16 01:39:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][500/1251] eta 0:03:35 lr 0.000685 time 0.2855 (0.2875) loss 3.2771 (3.6521) grad_norm 1.3416 (1.3326) [2021-04-16 01:39:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][510/1251] eta 0:03:32 lr 0.000685 time 0.2868 (0.2872) loss 2.8036 (3.6471) grad_norm 1.1995 (1.3322) [2021-04-16 01:39:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][520/1251] eta 0:03:29 lr 0.000685 time 0.2663 (0.2872) loss 2.6749 (3.6462) grad_norm 1.3468 (1.3346) [2021-04-16 01:39:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][530/1251] eta 0:03:26 lr 0.000685 time 0.2638 (0.2871) loss 4.2556 (3.6405) grad_norm 1.2396 (1.3349) [2021-04-16 01:39:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][540/1251] eta 0:03:24 lr 0.000685 time 0.2785 (0.2869) loss 4.0132 (3.6385) grad_norm 1.1692 (1.3348) [2021-04-16 01:39:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][550/1251] eta 0:03:20 lr 0.000685 time 0.2634 (0.2867) loss 2.5144 (3.6415) grad_norm 1.2745 (1.3344) [2021-04-16 01:39:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][560/1251] eta 0:03:17 lr 0.000685 time 0.2825 (0.2865) loss 4.1166 (3.6424) grad_norm 1.4375 (1.3345) [2021-04-16 01:39:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][570/1251] eta 0:03:15 lr 0.000685 time 0.2473 (0.2868) loss 4.0963 (3.6456) grad_norm 1.5362 (1.3342) [2021-04-16 01:40:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][580/1251] eta 0:03:12 lr 0.000685 time 0.2829 (0.2868) loss 2.7797 (3.6459) grad_norm 1.2675 (1.3345) [2021-04-16 01:40:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][590/1251] eta 0:03:09 lr 0.000685 time 0.2706 (0.2868) loss 3.6509 (3.6452) grad_norm 1.1506 (1.3343) [2021-04-16 01:40:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][600/1251] eta 0:03:06 lr 0.000685 time 0.2627 (0.2867) loss 4.4521 (3.6466) grad_norm 1.2249 (1.3332) [2021-04-16 01:40:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][610/1251] eta 0:03:03 lr 0.000685 time 0.2556 (0.2865) loss 3.7874 (3.6475) grad_norm 1.3175 (1.3340) [2021-04-16 01:40:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][620/1251] eta 0:03:00 lr 0.000685 time 0.2912 (0.2865) loss 3.8500 (3.6432) grad_norm 1.4012 (1.3345) [2021-04-16 01:40:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][630/1251] eta 0:02:57 lr 0.000685 time 0.2882 (0.2863) loss 4.5641 (3.6429) grad_norm 1.4912 (1.3366) [2021-04-16 01:40:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][640/1251] eta 0:02:54 lr 0.000685 time 0.2712 (0.2862) loss 4.7584 (3.6467) grad_norm 1.2923 (1.3370) [2021-04-16 01:40:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][650/1251] eta 0:02:52 lr 0.000685 time 0.2958 (0.2863) loss 4.2054 (3.6513) grad_norm 1.4528 (1.3355) [2021-04-16 01:40:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][660/1251] eta 0:02:49 lr 0.000685 time 0.2863 (0.2862) loss 4.2747 (3.6509) grad_norm 1.5132 (1.3364) [2021-04-16 01:40:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][670/1251] eta 0:02:46 lr 0.000685 time 0.2677 (0.2861) loss 3.6997 (3.6494) grad_norm 1.2042 (1.3359) [2021-04-16 01:40:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][680/1251] eta 0:02:43 lr 0.000685 time 0.2844 (0.2860) loss 4.0882 (3.6478) grad_norm 1.3256 (1.3344) [2021-04-16 01:40:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][690/1251] eta 0:02:40 lr 0.000685 time 0.2877 (0.2860) loss 3.2470 (3.6459) grad_norm 1.4202 (1.3342) [2021-04-16 01:40:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][700/1251] eta 0:02:37 lr 0.000685 time 0.2720 (0.2859) loss 3.5426 (3.6472) grad_norm 1.1495 (1.3335) [2021-04-16 01:40:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][710/1251] eta 0:02:34 lr 0.000684 time 0.3016 (0.2858) loss 3.7211 (3.6433) grad_norm 1.6365 (1.3353) [2021-04-16 01:40:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][720/1251] eta 0:02:31 lr 0.000684 time 0.2652 (0.2857) loss 3.0978 (3.6424) grad_norm 1.2707 (1.3358) [2021-04-16 01:40:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][730/1251] eta 0:02:28 lr 0.000684 time 0.2760 (0.2859) loss 4.7057 (3.6413) grad_norm 1.4339 (1.3343) [2021-04-16 01:40:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][740/1251] eta 0:02:26 lr 0.000684 time 0.2656 (0.2857) loss 3.3785 (3.6431) grad_norm 1.3659 (1.3340) [2021-04-16 01:40:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][750/1251] eta 0:02:23 lr 0.000684 time 0.2743 (0.2857) loss 4.3004 (3.6476) grad_norm 1.3718 (1.3343) [2021-04-16 01:40:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][760/1251] eta 0:02:20 lr 0.000684 time 0.2683 (0.2858) loss 3.8094 (3.6469) grad_norm 1.1913 (1.3332) [2021-04-16 01:40:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][770/1251] eta 0:02:17 lr 0.000684 time 0.2859 (0.2856) loss 4.0996 (3.6433) grad_norm 1.4004 (1.3340) [2021-04-16 01:40:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][780/1251] eta 0:02:14 lr 0.000684 time 0.3259 (0.2858) loss 4.0375 (3.6443) grad_norm 1.3383 (1.3336) [2021-04-16 01:40:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][790/1251] eta 0:02:11 lr 0.000684 time 0.2890 (0.2857) loss 4.1925 (3.6489) grad_norm 1.2373 (1.3337) [2021-04-16 01:41:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][800/1251] eta 0:02:08 lr 0.000684 time 0.2439 (0.2856) loss 4.0449 (3.6483) grad_norm 1.2300 (1.3331) [2021-04-16 01:41:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][810/1251] eta 0:02:05 lr 0.000684 time 0.2685 (0.2857) loss 3.4628 (3.6532) grad_norm 1.6779 (1.3329) [2021-04-16 01:41:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][820/1251] eta 0:02:03 lr 0.000684 time 0.2668 (0.2856) loss 4.3207 (3.6534) grad_norm 1.2546 (1.3316) [2021-04-16 01:41:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][830/1251] eta 0:02:00 lr 0.000684 time 0.3387 (0.2856) loss 3.8389 (3.6562) grad_norm 1.3018 (1.3325) [2021-04-16 01:41:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][840/1251] eta 0:01:57 lr 0.000684 time 0.2963 (0.2855) loss 3.1575 (3.6542) grad_norm 1.2581 (1.3327) [2021-04-16 01:41:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][850/1251] eta 0:01:54 lr 0.000684 time 0.2901 (0.2854) loss 3.4962 (3.6565) grad_norm 1.2257 (1.3330) [2021-04-16 01:41:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][860/1251] eta 0:01:51 lr 0.000684 time 0.2748 (0.2854) loss 3.9297 (3.6599) grad_norm 1.2616 (1.3320) [2021-04-16 01:41:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][870/1251] eta 0:01:48 lr 0.000684 time 0.2982 (0.2853) loss 2.6794 (3.6588) grad_norm 1.4632 (1.3312) [2021-04-16 01:41:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][880/1251] eta 0:01:45 lr 0.000684 time 0.2439 (0.2852) loss 4.5290 (3.6604) grad_norm 1.4276 (1.3314) [2021-04-16 01:41:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][890/1251] eta 0:01:42 lr 0.000684 time 0.2709 (0.2853) loss 3.7099 (3.6586) grad_norm 1.4829 (1.3314) [2021-04-16 01:41:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][900/1251] eta 0:01:40 lr 0.000684 time 0.2678 (0.2852) loss 3.1414 (3.6608) grad_norm 1.3015 (1.3308) [2021-04-16 01:41:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][910/1251] eta 0:01:37 lr 0.000684 time 0.2700 (0.2852) loss 2.8773 (3.6586) grad_norm 1.1825 (1.3306) [2021-04-16 01:41:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][920/1251] eta 0:01:34 lr 0.000684 time 0.2970 (0.2852) loss 3.9751 (3.6598) grad_norm 1.2046 (1.3300) [2021-04-16 01:41:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][930/1251] eta 0:01:31 lr 0.000684 time 0.2819 (0.2851) loss 4.2243 (3.6576) grad_norm 1.1227 (1.3297) [2021-04-16 01:41:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][940/1251] eta 0:01:28 lr 0.000684 time 0.2846 (0.2853) loss 3.9090 (3.6601) grad_norm 1.4643 (1.3293) [2021-04-16 01:41:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][950/1251] eta 0:01:25 lr 0.000684 time 0.2814 (0.2853) loss 3.8393 (3.6654) grad_norm 1.2097 (1.3294) [2021-04-16 01:41:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][960/1251] eta 0:01:23 lr 0.000684 time 0.3035 (0.2852) loss 2.6932 (3.6630) grad_norm 1.2967 (1.3302) [2021-04-16 01:41:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][970/1251] eta 0:01:20 lr 0.000683 time 0.2975 (0.2851) loss 3.7923 (3.6655) grad_norm 1.1138 (1.3304) [2021-04-16 01:41:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][980/1251] eta 0:01:17 lr 0.000683 time 0.2799 (0.2851) loss 4.2934 (3.6656) grad_norm 1.3809 (1.3304) [2021-04-16 01:41:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][990/1251] eta 0:01:14 lr 0.000683 time 0.2819 (0.2851) loss 3.3922 (3.6670) grad_norm 1.3107 (1.3317) [2021-04-16 01:41:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][1000/1251] eta 0:01:11 lr 0.000683 time 0.2779 (0.2850) loss 2.8857 (3.6685) grad_norm 1.3207 (1.3316) [2021-04-16 01:42:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][1010/1251] eta 0:01:08 lr 0.000683 time 0.2746 (0.2851) loss 4.0164 (3.6700) grad_norm 1.3320 (1.3312) [2021-04-16 01:42:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][1020/1251] eta 0:01:05 lr 0.000683 time 0.2842 (0.2851) loss 3.6599 (3.6695) grad_norm 1.4501 (1.3308) [2021-04-16 01:42:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][1030/1251] eta 0:01:02 lr 0.000683 time 0.2762 (0.2849) loss 3.5112 (3.6693) grad_norm 1.2763 (1.3305) [2021-04-16 01:42:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][1040/1251] eta 0:01:00 lr 0.000683 time 0.2597 (0.2849) loss 3.0043 (3.6689) grad_norm 1.2221 (1.3304) [2021-04-16 01:42:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][1050/1251] eta 0:00:57 lr 0.000683 time 0.3004 (0.2849) loss 2.4612 (3.6661) grad_norm 1.3414 (1.3303) [2021-04-16 01:42:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][1060/1251] eta 0:00:54 lr 0.000683 time 0.2868 (0.2848) loss 3.6428 (3.6644) grad_norm 1.3038 (1.3300) [2021-04-16 01:42:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][1070/1251] eta 0:00:51 lr 0.000683 time 0.2739 (0.2847) loss 4.0063 (3.6638) grad_norm 1.2230 (1.3295) [2021-04-16 01:42:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][1080/1251] eta 0:00:48 lr 0.000683 time 0.2997 (0.2846) loss 4.6327 (3.6653) grad_norm 1.4418 (1.3292) [2021-04-16 01:42:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][1090/1251] eta 0:00:45 lr 0.000683 time 0.2974 (0.2847) loss 3.8301 (3.6674) grad_norm 1.2631 (1.3287) [2021-04-16 01:42:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][1100/1251] eta 0:00:42 lr 0.000683 time 0.2833 (0.2846) loss 3.1561 (3.6671) grad_norm 1.2422 (1.3289) [2021-04-16 01:42:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][1110/1251] eta 0:00:40 lr 0.000683 time 0.3126 (0.2847) loss 4.7058 (3.6659) grad_norm 1.3512 (1.3289) [2021-04-16 01:42:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][1120/1251] eta 0:00:37 lr 0.000683 time 0.2510 (0.2846) loss 2.0882 (3.6648) grad_norm 1.2318 (1.3287) [2021-04-16 01:42:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][1130/1251] eta 0:00:34 lr 0.000683 time 0.2708 (0.2846) loss 4.5750 (3.6669) grad_norm 1.2707 (1.3288) [2021-04-16 01:42:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][1140/1251] eta 0:00:31 lr 0.000683 time 0.2710 (0.2846) loss 3.8655 (3.6672) grad_norm 1.3475 (1.3292) [2021-04-16 01:42:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][1150/1251] eta 0:00:28 lr 0.000683 time 0.2567 (0.2848) loss 4.4619 (3.6668) grad_norm 1.5214 (1.3293) [2021-04-16 01:42:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][1160/1251] eta 0:00:25 lr 0.000683 time 0.2831 (0.2848) loss 3.7755 (3.6670) grad_norm 1.3639 (1.3295) [2021-04-16 01:42:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][1170/1251] eta 0:00:23 lr 0.000683 time 0.3033 (0.2848) loss 3.3742 (3.6663) grad_norm 1.6875 (1.3302) [2021-04-16 01:42:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][1180/1251] eta 0:00:20 lr 0.000683 time 0.2565 (0.2848) loss 3.8844 (3.6672) grad_norm 1.4853 (1.3304) [2021-04-16 01:42:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][1190/1251] eta 0:00:17 lr 0.000683 time 0.2743 (0.2847) loss 4.3648 (3.6680) grad_norm 1.2305 (1.3303) [2021-04-16 01:42:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][1200/1251] eta 0:00:14 lr 0.000683 time 0.2839 (0.2847) loss 3.9569 (3.6700) grad_norm 1.4390 (1.3306) [2021-04-16 01:42:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][1210/1251] eta 0:00:11 lr 0.000683 time 0.3010 (0.2847) loss 4.2981 (3.6713) grad_norm 1.5151 (1.3312) [2021-04-16 01:43:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][1220/1251] eta 0:00:08 lr 0.000683 time 0.2578 (0.2846) loss 3.8043 (3.6720) grad_norm 1.5623 (1.3316) [2021-04-16 01:43:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][1230/1251] eta 0:00:05 lr 0.000682 time 0.2816 (0.2846) loss 3.9317 (3.6737) grad_norm 1.3493 (1.3315) [2021-04-16 01:43:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][1240/1251] eta 0:00:03 lr 0.000682 time 0.2486 (0.2844) loss 3.4341 (3.6743) grad_norm 1.3772 (1.3312) [2021-04-16 01:43:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [114/300][1250/1251] eta 0:00:00 lr 0.000682 time 0.2477 (0.2841) loss 4.3378 (3.6733) grad_norm 1.3632 (1.3310) [2021-04-16 01:43:12 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 114 training takes 0:05:58 [2021-04-16 01:43:12 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_114.pth saving...... [2021-04-16 01:43:23 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_114.pth saved !!! [2021-04-16 01:43:24 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.148 (1.148) Loss 1.1385 (1.1385) Acc@1 72.363 (72.363) Acc@5 91.992 (91.992) [2021-04-16 01:43:25 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.133 (0.266) Loss 1.1849 (1.0793) Acc@1 72.852 (74.760) Acc@5 90.918 (92.729) [2021-04-16 01:43:28 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.125 (0.250) Loss 1.1044 (1.0775) Acc@1 75.000 (74.716) Acc@5 91.895 (92.639) [2021-04-16 01:43:29 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.113 (0.214) Loss 1.1803 (1.0904) Acc@1 71.191 (74.383) Acc@5 91.309 (92.528) [2021-04-16 01:43:32 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.223 (0.223) Loss 1.1472 (1.0914) Acc@1 73.047 (74.428) Acc@5 92.773 (92.526) [2021-04-16 01:43:37 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 74.360 Acc@5 92.464 [2021-04-16 01:43:37 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 74.4% [2021-04-16 01:43:37 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 74.45% [2021-04-16 01:43:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][0/1251] eta 1:20:33 lr 0.000682 time 3.8637 (3.8637) loss 3.1107 (3.1107) grad_norm 1.5789 (1.5789) [2021-04-16 01:43:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][10/1251] eta 0:12:29 lr 0.000682 time 0.2793 (0.6041) loss 4.3701 (3.6687) grad_norm 1.5329 (1.3684) [2021-04-16 01:43:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][20/1251] eta 0:09:16 lr 0.000682 time 0.3091 (0.4521) loss 3.3788 (3.6745) grad_norm 1.4248 (1.3652) [2021-04-16 01:43:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][30/1251] eta 0:08:03 lr 0.000682 time 0.2965 (0.3962) loss 4.2599 (3.7159) grad_norm 1.3222 (1.3592) [2021-04-16 01:43:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][40/1251] eta 0:07:26 lr 0.000682 time 0.2700 (0.3688) loss 4.4278 (3.6678) grad_norm 1.2868 (1.3479) [2021-04-16 01:43:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][50/1251] eta 0:07:03 lr 0.000682 time 0.2852 (0.3529) loss 4.3564 (3.7203) grad_norm 1.3256 (1.3518) [2021-04-16 01:43:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][60/1251] eta 0:06:44 lr 0.000682 time 0.2868 (0.3399) loss 3.3615 (3.6694) grad_norm 1.2952 (1.3487) [2021-04-16 01:44:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][70/1251] eta 0:06:30 lr 0.000682 time 0.2595 (0.3311) loss 3.2438 (3.6592) grad_norm 1.1606 (1.3433) [2021-04-16 01:44:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][80/1251] eta 0:06:22 lr 0.000682 time 0.2884 (0.3263) loss 2.1764 (3.6181) grad_norm 1.2765 (1.3434) [2021-04-16 01:44:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][90/1251] eta 0:06:13 lr 0.000682 time 0.2990 (0.3214) loss 4.2285 (3.6519) grad_norm 1.2510 (1.3395) [2021-04-16 01:44:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][100/1251] eta 0:06:06 lr 0.000682 time 0.3265 (0.3184) loss 3.0275 (3.6523) grad_norm 1.5011 (1.3470) [2021-04-16 01:44:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][110/1251] eta 0:05:59 lr 0.000682 time 0.2643 (0.3149) loss 4.1935 (3.6145) grad_norm 1.4864 (1.3410) [2021-04-16 01:44:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][120/1251] eta 0:05:52 lr 0.000682 time 0.2741 (0.3119) loss 3.7292 (3.6042) grad_norm 1.5892 (1.3450) [2021-04-16 01:44:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][130/1251] eta 0:05:48 lr 0.000682 time 0.2830 (0.3106) loss 3.7325 (3.5907) grad_norm 1.7610 (1.3536) [2021-04-16 01:44:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][140/1251] eta 0:05:43 lr 0.000682 time 0.2896 (0.3088) loss 4.1011 (3.5667) grad_norm 1.1452 (1.3515) [2021-04-16 01:44:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][150/1251] eta 0:05:37 lr 0.000682 time 0.2979 (0.3067) loss 4.4989 (3.5925) grad_norm 1.1931 (1.3480) [2021-04-16 01:44:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][160/1251] eta 0:05:33 lr 0.000682 time 0.2936 (0.3058) loss 2.7246 (3.5818) grad_norm 1.3840 (1.3475) [2021-04-16 01:44:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][170/1251] eta 0:05:28 lr 0.000682 time 0.2828 (0.3043) loss 3.6557 (3.5876) grad_norm 1.4000 (1.3538) [2021-04-16 01:44:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][180/1251] eta 0:05:24 lr 0.000682 time 0.2952 (0.3032) loss 4.0241 (3.5966) grad_norm 1.4433 (1.3527) [2021-04-16 01:44:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][190/1251] eta 0:05:20 lr 0.000682 time 0.2729 (0.3018) loss 4.2495 (3.6069) grad_norm 1.4146 (1.3539) [2021-04-16 01:44:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][200/1251] eta 0:05:15 lr 0.000682 time 0.2693 (0.3007) loss 3.2458 (3.5992) grad_norm 1.2323 (1.3539) [2021-04-16 01:44:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][210/1251] eta 0:05:11 lr 0.000682 time 0.2618 (0.2996) loss 4.3236 (3.6105) grad_norm 1.2875 (1.3541) [2021-04-16 01:44:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][220/1251] eta 0:05:08 lr 0.000682 time 0.2738 (0.2990) loss 3.6748 (3.6192) grad_norm 1.1827 (1.3522) [2021-04-16 01:44:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][230/1251] eta 0:05:04 lr 0.000682 time 0.2925 (0.2986) loss 3.7733 (3.6192) grad_norm 1.2088 (1.3488) [2021-04-16 01:44:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][240/1251] eta 0:05:01 lr 0.000681 time 0.2718 (0.2980) loss 4.0168 (3.6346) grad_norm 1.4537 (1.3473) [2021-04-16 01:44:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][250/1251] eta 0:04:57 lr 0.000681 time 0.2713 (0.2970) loss 3.8766 (3.6355) grad_norm 1.2828 (1.3514) [2021-04-16 01:44:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][260/1251] eta 0:04:53 lr 0.000681 time 0.2899 (0.2964) loss 4.2820 (3.6370) grad_norm 1.1939 (1.3494) [2021-04-16 01:44:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][270/1251] eta 0:04:50 lr 0.000681 time 0.2732 (0.2957) loss 3.3450 (3.6303) grad_norm 1.3420 (1.3508) [2021-04-16 01:45:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][280/1251] eta 0:04:46 lr 0.000681 time 0.2631 (0.2953) loss 3.9706 (3.6298) grad_norm 1.1870 (1.3501) [2021-04-16 01:45:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][290/1251] eta 0:04:43 lr 0.000681 time 0.2879 (0.2951) loss 4.2171 (3.6361) grad_norm 1.7773 (1.3518) [2021-04-16 01:45:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][300/1251] eta 0:04:39 lr 0.000681 time 0.2746 (0.2944) loss 3.3654 (3.6281) grad_norm 1.2672 (1.3512) [2021-04-16 01:45:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][310/1251] eta 0:04:37 lr 0.000681 time 0.2697 (0.2944) loss 3.3891 (3.6316) grad_norm 1.2531 (1.3527) [2021-04-16 01:45:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][320/1251] eta 0:04:33 lr 0.000681 time 0.2912 (0.2939) loss 3.9112 (3.6368) grad_norm 1.5089 (1.3545) [2021-04-16 01:45:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][330/1251] eta 0:04:30 lr 0.000681 time 0.2784 (0.2935) loss 4.8482 (3.6434) grad_norm 1.3041 (1.3528) [2021-04-16 01:45:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][340/1251] eta 0:04:27 lr 0.000681 time 0.2800 (0.2934) loss 4.1528 (3.6480) grad_norm 1.1761 (1.3490) [2021-04-16 01:45:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][350/1251] eta 0:04:24 lr 0.000681 time 0.2597 (0.2932) loss 3.2646 (3.6488) grad_norm 1.1141 (1.3475) [2021-04-16 01:45:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][360/1251] eta 0:04:21 lr 0.000681 time 0.2947 (0.2932) loss 3.7123 (3.6490) grad_norm 1.1596 (1.3488) [2021-04-16 01:45:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][370/1251] eta 0:04:18 lr 0.000681 time 0.2606 (0.2931) loss 3.5386 (3.6447) grad_norm 1.1961 (1.3487) [2021-04-16 01:45:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][380/1251] eta 0:04:14 lr 0.000681 time 0.2730 (0.2926) loss 4.1168 (3.6392) grad_norm 1.4885 (1.3487) [2021-04-16 01:45:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][390/1251] eta 0:04:11 lr 0.000681 time 0.2851 (0.2923) loss 3.5609 (3.6380) grad_norm 1.2373 (1.3483) [2021-04-16 01:45:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][400/1251] eta 0:04:08 lr 0.000681 time 0.2854 (0.2920) loss 3.7040 (3.6381) grad_norm 1.2082 (1.3460) [2021-04-16 01:45:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][410/1251] eta 0:04:05 lr 0.000681 time 0.2745 (0.2918) loss 4.6124 (3.6394) grad_norm 1.1657 (1.3444) [2021-04-16 01:45:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][420/1251] eta 0:04:02 lr 0.000681 time 0.3093 (0.2917) loss 3.9190 (3.6389) grad_norm 1.6204 (1.3451) [2021-04-16 01:45:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][430/1251] eta 0:03:59 lr 0.000681 time 0.2814 (0.2913) loss 3.5185 (3.6395) grad_norm 1.1162 (1.3444) [2021-04-16 01:45:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][440/1251] eta 0:03:55 lr 0.000681 time 0.2690 (0.2910) loss 3.8664 (3.6407) grad_norm 1.4077 (1.3435) [2021-04-16 01:45:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][450/1251] eta 0:03:52 lr 0.000681 time 0.2508 (0.2907) loss 3.3123 (3.6386) grad_norm 1.2823 (1.3438) [2021-04-16 01:45:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][460/1251] eta 0:03:49 lr 0.000681 time 0.2609 (0.2904) loss 3.7987 (3.6431) grad_norm 1.6483 (1.3447) [2021-04-16 01:45:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][470/1251] eta 0:03:46 lr 0.000681 time 0.2599 (0.2903) loss 3.4907 (3.6476) grad_norm 1.1264 (1.3480) [2021-04-16 01:45:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][480/1251] eta 0:03:43 lr 0.000681 time 0.2866 (0.2901) loss 3.9841 (3.6443) grad_norm 1.4113 (1.3486) [2021-04-16 01:45:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][490/1251] eta 0:03:40 lr 0.000680 time 0.3041 (0.2900) loss 3.7304 (3.6547) grad_norm 1.3043 (1.3476) [2021-04-16 01:46:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][500/1251] eta 0:03:37 lr 0.000680 time 0.2946 (0.2898) loss 3.6220 (3.6572) grad_norm 1.4429 (1.3469) [2021-04-16 01:46:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][510/1251] eta 0:03:34 lr 0.000680 time 0.2929 (0.2897) loss 4.4355 (3.6575) grad_norm 1.3544 (1.3467) [2021-04-16 01:46:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][520/1251] eta 0:03:31 lr 0.000680 time 0.2837 (0.2894) loss 4.2426 (3.6586) grad_norm 1.0895 (1.3476) [2021-04-16 01:46:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][530/1251] eta 0:03:28 lr 0.000680 time 0.2680 (0.2893) loss 3.5353 (3.6616) grad_norm 1.4263 (1.3461) [2021-04-16 01:46:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][540/1251] eta 0:03:25 lr 0.000680 time 0.2886 (0.2891) loss 3.0761 (3.6653) grad_norm 1.0806 (1.3458) [2021-04-16 01:46:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][550/1251] eta 0:03:22 lr 0.000680 time 0.4296 (0.2891) loss 4.2901 (3.6659) grad_norm 1.2915 (1.3452) [2021-04-16 01:46:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][560/1251] eta 0:03:19 lr 0.000680 time 0.2834 (0.2889) loss 3.1819 (3.6675) grad_norm 1.3339 (1.3465) [2021-04-16 01:46:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][570/1251] eta 0:03:16 lr 0.000680 time 0.2731 (0.2889) loss 4.3436 (3.6705) grad_norm 1.1498 (1.3457) [2021-04-16 01:46:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][580/1251] eta 0:03:13 lr 0.000680 time 0.2768 (0.2891) loss 2.8345 (3.6733) grad_norm 1.3946 (1.3474) [2021-04-16 01:46:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][590/1251] eta 0:03:11 lr 0.000680 time 0.2515 (0.2890) loss 4.1298 (3.6744) grad_norm 1.1963 (1.3482) [2021-04-16 01:46:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][600/1251] eta 0:03:08 lr 0.000680 time 0.2842 (0.2890) loss 3.4980 (3.6731) grad_norm 1.2105 (1.3464) [2021-04-16 01:46:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][610/1251] eta 0:03:05 lr 0.000680 time 0.2874 (0.2889) loss 2.7655 (3.6659) grad_norm 1.4538 (1.3464) [2021-04-16 01:46:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][620/1251] eta 0:03:02 lr 0.000680 time 0.2833 (0.2886) loss 4.3056 (3.6657) grad_norm 1.4763 (1.3470) [2021-04-16 01:46:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][630/1251] eta 0:02:59 lr 0.000680 time 0.2778 (0.2885) loss 3.4094 (3.6641) grad_norm 1.3362 (1.3472) [2021-04-16 01:46:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][640/1251] eta 0:02:56 lr 0.000680 time 0.2965 (0.2883) loss 3.7482 (3.6629) grad_norm 1.2140 (1.3459) [2021-04-16 01:46:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][650/1251] eta 0:02:53 lr 0.000680 time 0.2651 (0.2881) loss 2.8525 (3.6633) grad_norm 1.2924 (1.3444) [2021-04-16 01:46:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][660/1251] eta 0:02:50 lr 0.000680 time 0.2806 (0.2879) loss 4.4719 (3.6686) grad_norm 1.3753 (1.3437) [2021-04-16 01:46:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][670/1251] eta 0:02:47 lr 0.000680 time 0.2955 (0.2879) loss 4.2203 (3.6720) grad_norm 1.2577 (1.3433) [2021-04-16 01:46:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][680/1251] eta 0:02:44 lr 0.000680 time 0.2769 (0.2878) loss 3.8746 (3.6706) grad_norm 1.3965 (1.3438) [2021-04-16 01:46:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][690/1251] eta 0:02:41 lr 0.000680 time 0.2714 (0.2877) loss 3.3437 (3.6695) grad_norm 1.3160 (1.3450) [2021-04-16 01:46:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][700/1251] eta 0:02:38 lr 0.000680 time 0.2769 (0.2875) loss 4.0116 (3.6731) grad_norm 1.3059 (1.3457) [2021-04-16 01:47:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][710/1251] eta 0:02:35 lr 0.000680 time 0.2992 (0.2877) loss 4.0898 (3.6720) grad_norm 1.2608 (1.3447) [2021-04-16 01:47:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][720/1251] eta 0:02:32 lr 0.000680 time 0.2914 (0.2876) loss 3.6848 (3.6678) grad_norm 1.2540 (1.3443) [2021-04-16 01:47:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][730/1251] eta 0:02:29 lr 0.000680 time 0.2571 (0.2877) loss 2.4702 (3.6677) grad_norm 1.1782 (1.3437) [2021-04-16 01:47:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][740/1251] eta 0:02:26 lr 0.000680 time 0.2770 (0.2875) loss 3.2086 (3.6657) grad_norm 1.4753 (1.3450) [2021-04-16 01:47:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][750/1251] eta 0:02:23 lr 0.000679 time 0.2971 (0.2874) loss 4.0849 (3.6637) grad_norm 1.2763 (1.3448) [2021-04-16 01:47:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][760/1251] eta 0:02:21 lr 0.000679 time 0.2791 (0.2873) loss 2.6298 (3.6614) grad_norm 1.3452 (1.3448) [2021-04-16 01:47:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][770/1251] eta 0:02:18 lr 0.000679 time 0.3005 (0.2872) loss 4.4549 (3.6630) grad_norm 1.3861 (1.3447) [2021-04-16 01:47:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][780/1251] eta 0:02:15 lr 0.000679 time 0.2808 (0.2872) loss 4.1070 (3.6604) grad_norm 1.4299 (1.3448) [2021-04-16 01:47:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][790/1251] eta 0:02:12 lr 0.000679 time 0.2956 (0.2873) loss 4.2006 (3.6612) grad_norm 1.4439 (1.3457) [2021-04-16 01:47:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][800/1251] eta 0:02:09 lr 0.000679 time 0.3689 (0.2873) loss 3.4867 (3.6586) grad_norm 1.3029 (1.3460) [2021-04-16 01:47:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][810/1251] eta 0:02:06 lr 0.000679 time 0.3015 (0.2871) loss 3.4321 (3.6561) grad_norm 1.2947 (1.3460) [2021-04-16 01:47:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][820/1251] eta 0:02:03 lr 0.000679 time 0.2756 (0.2871) loss 2.8801 (3.6554) grad_norm 1.4485 (1.3459) [2021-04-16 01:47:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][830/1251] eta 0:02:00 lr 0.000679 time 0.2863 (0.2871) loss 3.7685 (3.6503) grad_norm 1.4499 (1.3452) [2021-04-16 01:47:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][840/1251] eta 0:01:57 lr 0.000679 time 0.2831 (0.2870) loss 4.6490 (3.6522) grad_norm 1.2479 (1.3451) [2021-04-16 01:47:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][850/1251] eta 0:01:55 lr 0.000679 time 0.2504 (0.2868) loss 3.4638 (3.6511) grad_norm 1.5121 (1.3445) [2021-04-16 01:47:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][860/1251] eta 0:01:52 lr 0.000679 time 0.2847 (0.2868) loss 3.8531 (3.6503) grad_norm 1.3359 (1.3437) [2021-04-16 01:47:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][870/1251] eta 0:01:49 lr 0.000679 time 0.2734 (0.2868) loss 2.2222 (3.6477) grad_norm 1.2995 (1.3440) [2021-04-16 01:47:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][880/1251] eta 0:01:46 lr 0.000679 time 0.4084 (0.2869) loss 3.5424 (3.6486) grad_norm 1.2387 (1.3434) [2021-04-16 01:47:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][890/1251] eta 0:01:43 lr 0.000679 time 0.2747 (0.2868) loss 3.8888 (3.6479) grad_norm 1.3134 (1.3423) [2021-04-16 01:47:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][900/1251] eta 0:01:40 lr 0.000679 time 0.2863 (0.2867) loss 4.3941 (3.6511) grad_norm 1.4602 (1.3421) [2021-04-16 01:47:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][910/1251] eta 0:01:37 lr 0.000679 time 0.2845 (0.2865) loss 2.2660 (3.6497) grad_norm 1.2035 (1.3430) [2021-04-16 01:48:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][920/1251] eta 0:01:34 lr 0.000679 time 0.2895 (0.2866) loss 3.0520 (3.6487) grad_norm 1.2151 (1.3425) [2021-04-16 01:48:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][930/1251] eta 0:01:31 lr 0.000679 time 0.2846 (0.2866) loss 3.7238 (3.6512) grad_norm 1.3675 (1.3420) [2021-04-16 01:48:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][940/1251] eta 0:01:29 lr 0.000679 time 0.4400 (0.2867) loss 4.0805 (3.6549) grad_norm 1.1616 (1.3415) [2021-04-16 01:48:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][950/1251] eta 0:01:26 lr 0.000679 time 0.2818 (0.2865) loss 3.2359 (3.6549) grad_norm 1.2266 (1.3404) [2021-04-16 01:48:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][960/1251] eta 0:01:23 lr 0.000679 time 0.2752 (0.2864) loss 3.2058 (3.6547) grad_norm 1.3877 (1.3403) [2021-04-16 01:48:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][970/1251] eta 0:01:20 lr 0.000679 time 0.2708 (0.2863) loss 3.4395 (3.6562) grad_norm 1.4485 (1.3397) [2021-04-16 01:48:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][980/1251] eta 0:01:17 lr 0.000679 time 0.3004 (0.2863) loss 2.4802 (3.6557) grad_norm 1.2266 (1.3408) [2021-04-16 01:48:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][990/1251] eta 0:01:14 lr 0.000679 time 0.2602 (0.2862) loss 2.5780 (3.6528) grad_norm 1.5463 (1.3414) [2021-04-16 01:48:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][1000/1251] eta 0:01:11 lr 0.000679 time 0.3011 (0.2861) loss 3.6262 (3.6532) grad_norm 1.1370 (1.3404) [2021-04-16 01:48:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][1010/1251] eta 0:01:08 lr 0.000678 time 0.2712 (0.2861) loss 4.2622 (3.6514) grad_norm 1.2049 (1.3397) [2021-04-16 01:48:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][1020/1251] eta 0:01:06 lr 0.000678 time 0.2772 (0.2861) loss 2.9778 (3.6505) grad_norm 1.2897 (1.3393) [2021-04-16 01:48:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][1030/1251] eta 0:01:03 lr 0.000678 time 0.2923 (0.2861) loss 3.3326 (3.6484) grad_norm 1.2490 (1.3386) [2021-04-16 01:48:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][1040/1251] eta 0:01:00 lr 0.000678 time 0.2457 (0.2860) loss 3.8067 (3.6494) grad_norm 1.3969 (1.3390) [2021-04-16 01:48:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][1050/1251] eta 0:00:57 lr 0.000678 time 0.2958 (0.2860) loss 3.7659 (3.6469) grad_norm 1.4823 (1.3393) [2021-04-16 01:48:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][1060/1251] eta 0:00:54 lr 0.000678 time 0.2959 (0.2859) loss 3.6738 (3.6490) grad_norm 1.2405 (1.3391) [2021-04-16 01:48:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][1070/1251] eta 0:00:51 lr 0.000678 time 0.2944 (0.2858) loss 4.0033 (3.6513) grad_norm 1.2878 (1.3393) [2021-04-16 01:48:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][1080/1251] eta 0:00:48 lr 0.000678 time 0.2775 (0.2858) loss 3.7758 (3.6535) grad_norm 1.3978 (1.3390) [2021-04-16 01:48:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][1090/1251] eta 0:00:45 lr 0.000678 time 0.2777 (0.2857) loss 3.3621 (3.6570) grad_norm 1.4992 (1.3398) [2021-04-16 01:48:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][1100/1251] eta 0:00:43 lr 0.000678 time 0.2893 (0.2856) loss 3.1258 (3.6561) grad_norm 1.5410 (1.3400) [2021-04-16 01:48:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][1110/1251] eta 0:00:40 lr 0.000678 time 0.2696 (0.2855) loss 4.4204 (3.6573) grad_norm 1.3798 (1.3395) [2021-04-16 01:48:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][1120/1251] eta 0:00:37 lr 0.000678 time 0.2824 (0.2855) loss 2.5659 (3.6561) grad_norm 1.2529 (1.3390) [2021-04-16 01:49:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][1130/1251] eta 0:00:34 lr 0.000678 time 0.2683 (0.2855) loss 3.9138 (3.6582) grad_norm 1.3142 (1.3391) [2021-04-16 01:49:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][1140/1251] eta 0:00:31 lr 0.000678 time 0.3174 (0.2855) loss 3.6336 (3.6603) grad_norm 1.3313 (1.3388) [2021-04-16 01:49:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][1150/1251] eta 0:00:28 lr 0.000678 time 0.2809 (0.2855) loss 3.1600 (3.6594) grad_norm 1.3336 (1.3385) [2021-04-16 01:49:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][1160/1251] eta 0:00:25 lr 0.000678 time 0.2818 (0.2854) loss 3.0262 (3.6581) grad_norm 1.3439 (1.3387) [2021-04-16 01:49:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][1170/1251] eta 0:00:23 lr 0.000678 time 0.2770 (0.2854) loss 2.3518 (3.6572) grad_norm 1.3857 (1.3380) [2021-04-16 01:49:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][1180/1251] eta 0:00:20 lr 0.000678 time 0.2901 (0.2853) loss 4.2872 (3.6571) grad_norm 1.1960 (1.3386) [2021-04-16 01:49:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][1190/1251] eta 0:00:17 lr 0.000678 time 0.2841 (0.2853) loss 3.0597 (3.6554) grad_norm 1.1987 (1.3382) [2021-04-16 01:49:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][1200/1251] eta 0:00:14 lr 0.000678 time 0.2865 (0.2852) loss 2.9618 (3.6557) grad_norm 1.5020 (1.3386) [2021-04-16 01:49:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][1210/1251] eta 0:00:11 lr 0.000678 time 0.3033 (0.2852) loss 4.4104 (3.6579) grad_norm 1.3381 (1.3386) [2021-04-16 01:49:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][1220/1251] eta 0:00:08 lr 0.000678 time 0.2715 (0.2851) loss 2.6791 (3.6567) grad_norm 1.3753 (1.3390) [2021-04-16 01:49:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][1230/1251] eta 0:00:05 lr 0.000678 time 0.2774 (0.2850) loss 3.2172 (3.6553) grad_norm 1.2373 (1.3389) [2021-04-16 01:49:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][1240/1251] eta 0:00:03 lr 0.000678 time 0.2487 (0.2849) loss 3.8204 (3.6555) grad_norm 1.2278 (1.3388) [2021-04-16 01:49:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [115/300][1250/1251] eta 0:00:00 lr 0.000678 time 0.2508 (0.2846) loss 3.6577 (3.6558) grad_norm 1.2979 (1.3389) [2021-04-16 01:49:36 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 115 training takes 0:05:59 [2021-04-16 01:49:36 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_115.pth saving...... [2021-04-16 01:49:49 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_115.pth saved !!! [2021-04-16 01:49:50 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.159 (1.159) Loss 0.9614 (0.9614) Acc@1 77.832 (77.832) Acc@5 94.922 (94.922) [2021-04-16 01:49:51 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.113 (0.218) Loss 1.0900 (1.0835) Acc@1 74.414 (74.316) Acc@5 92.969 (92.649) [2021-04-16 01:49:53 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.160 (0.207) Loss 1.0958 (1.0859) Acc@1 73.340 (74.488) Acc@5 91.992 (92.532) [2021-04-16 01:49:56 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.118 (0.235) Loss 1.0935 (1.0829) Acc@1 73.926 (74.553) Acc@5 91.992 (92.490) [2021-04-16 01:49:58 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.071 (0.220) Loss 1.1676 (1.0857) Acc@1 72.070 (74.528) Acc@5 91.699 (92.450) [2021-04-16 01:50:02 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 74.476 Acc@5 92.426 [2021-04-16 01:50:02 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 74.5% [2021-04-16 01:50:02 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 74.48% [2021-04-16 01:50:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][0/1251] eta 2:28:13 lr 0.000678 time 7.1094 (7.1094) loss 3.1349 (3.1349) grad_norm 1.5840 (1.5840) [2021-04-16 01:50:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][10/1251] eta 0:18:41 lr 0.000678 time 0.3968 (0.9037) loss 4.4375 (3.5988) grad_norm 1.2868 (1.3928) [2021-04-16 01:50:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][20/1251] eta 0:12:24 lr 0.000677 time 0.2848 (0.6044) loss 3.7331 (3.4909) grad_norm 1.3005 (1.3375) [2021-04-16 01:50:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][30/1251] eta 0:10:09 lr 0.000677 time 0.2748 (0.4993) loss 3.4462 (3.4530) grad_norm 1.2993 (1.3133) [2021-04-16 01:50:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][40/1251] eta 0:08:58 lr 0.000677 time 0.2690 (0.4447) loss 3.1276 (3.5571) grad_norm 1.5126 (1.3345) [2021-04-16 01:50:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][50/1251] eta 0:08:14 lr 0.000677 time 0.2639 (0.4115) loss 4.0834 (3.5832) grad_norm 1.0936 (1.3409) [2021-04-16 01:50:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][60/1251] eta 0:07:43 lr 0.000677 time 0.2572 (0.3894) loss 3.8580 (3.6335) grad_norm 1.3672 (1.3366) [2021-04-16 01:50:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][70/1251] eta 0:07:22 lr 0.000677 time 0.2712 (0.3750) loss 2.2012 (3.6618) grad_norm 1.1753 (1.3359) [2021-04-16 01:50:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][80/1251] eta 0:07:06 lr 0.000677 time 0.2779 (0.3640) loss 3.4877 (3.6986) grad_norm 1.3512 (1.3334) [2021-04-16 01:50:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][90/1251] eta 0:06:51 lr 0.000677 time 0.2642 (0.3545) loss 3.2002 (3.6847) grad_norm 1.5089 (1.3393) [2021-04-16 01:50:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][100/1251] eta 0:06:39 lr 0.000677 time 0.2766 (0.3471) loss 4.3723 (3.6939) grad_norm 1.5559 (1.3430) [2021-04-16 01:50:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][110/1251] eta 0:06:28 lr 0.000677 time 0.2835 (0.3408) loss 4.2363 (3.6905) grad_norm 1.3100 (1.3404) [2021-04-16 01:50:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][120/1251] eta 0:06:21 lr 0.000677 time 0.2913 (0.3371) loss 2.4813 (3.7066) grad_norm 1.5327 (1.3382) [2021-04-16 01:50:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][130/1251] eta 0:06:13 lr 0.000677 time 0.2797 (0.3329) loss 3.5038 (3.7052) grad_norm 1.2627 (1.3391) [2021-04-16 01:50:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][140/1251] eta 0:06:05 lr 0.000677 time 0.3018 (0.3293) loss 3.3586 (3.6936) grad_norm 1.2757 (1.3405) [2021-04-16 01:50:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][150/1251] eta 0:05:59 lr 0.000677 time 0.3577 (0.3261) loss 3.9435 (3.6923) grad_norm 1.2525 (1.3387) [2021-04-16 01:50:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][160/1251] eta 0:05:52 lr 0.000677 time 0.2717 (0.3230) loss 3.3935 (3.6944) grad_norm 1.3256 (1.3416) [2021-04-16 01:50:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][170/1251] eta 0:05:46 lr 0.000677 time 0.2573 (0.3202) loss 4.0637 (3.6875) grad_norm 1.1408 (1.3382) [2021-04-16 01:51:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][180/1251] eta 0:05:40 lr 0.000677 time 0.2862 (0.3179) loss 3.0847 (3.6864) grad_norm 1.3984 (1.3373) [2021-04-16 01:51:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][190/1251] eta 0:05:34 lr 0.000677 time 0.2969 (0.3156) loss 3.7664 (3.6830) grad_norm 1.2074 (1.3386) [2021-04-16 01:51:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][200/1251] eta 0:05:29 lr 0.000677 time 0.2694 (0.3135) loss 3.8396 (3.6785) grad_norm 1.5386 (1.3416) [2021-04-16 01:51:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][210/1251] eta 0:05:24 lr 0.000677 time 0.2632 (0.3118) loss 3.0855 (3.6713) grad_norm 1.1538 (1.3410) [2021-04-16 01:51:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][220/1251] eta 0:05:19 lr 0.000677 time 0.2945 (0.3101) loss 3.4702 (3.6728) grad_norm 1.5138 (1.3432) [2021-04-16 01:51:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][230/1251] eta 0:05:15 lr 0.000677 time 0.2703 (0.3086) loss 2.9394 (3.6804) grad_norm 1.1874 (1.3452) [2021-04-16 01:51:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][240/1251] eta 0:05:10 lr 0.000677 time 0.2638 (0.3071) loss 3.7259 (3.6664) grad_norm 1.1074 (1.3448) [2021-04-16 01:51:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][250/1251] eta 0:05:06 lr 0.000677 time 0.2680 (0.3059) loss 2.9019 (3.6568) grad_norm 1.3593 (1.3420) [2021-04-16 01:51:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][260/1251] eta 0:05:02 lr 0.000677 time 0.3156 (0.3048) loss 3.8430 (3.6557) grad_norm 1.2824 (1.3372) [2021-04-16 01:51:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][270/1251] eta 0:04:57 lr 0.000676 time 0.2821 (0.3037) loss 3.4958 (3.6532) grad_norm 1.4981 (1.3368) [2021-04-16 01:51:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][280/1251] eta 0:04:53 lr 0.000676 time 0.2627 (0.3026) loss 3.5998 (3.6588) grad_norm 1.3575 (1.3374) [2021-04-16 01:51:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][290/1251] eta 0:04:49 lr 0.000676 time 0.2779 (0.3017) loss 2.9128 (3.6519) grad_norm 1.3309 (1.3359) [2021-04-16 01:51:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][300/1251] eta 0:04:46 lr 0.000676 time 0.2933 (0.3008) loss 3.6341 (3.6485) grad_norm 1.2166 (1.3347) [2021-04-16 01:51:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][310/1251] eta 0:04:42 lr 0.000676 time 0.2723 (0.2998) loss 4.4333 (3.6490) grad_norm 1.8115 (1.3368) [2021-04-16 01:51:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][320/1251] eta 0:04:38 lr 0.000676 time 0.2790 (0.2991) loss 3.0242 (3.6470) grad_norm 1.3758 (1.3358) [2021-04-16 01:51:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][330/1251] eta 0:04:34 lr 0.000676 time 0.2654 (0.2984) loss 3.9031 (3.6560) grad_norm 1.1108 (1.3350) [2021-04-16 01:51:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][340/1251] eta 0:04:31 lr 0.000676 time 0.2708 (0.2983) loss 4.1001 (3.6593) grad_norm 1.2284 (1.3344) [2021-04-16 01:51:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][350/1251] eta 0:04:28 lr 0.000676 time 0.4216 (0.2979) loss 3.6493 (3.6584) grad_norm 1.1734 (1.3331) [2021-04-16 01:51:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][360/1251] eta 0:04:24 lr 0.000676 time 0.2962 (0.2974) loss 3.0840 (3.6506) grad_norm 1.3391 (1.3334) [2021-04-16 01:51:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][370/1251] eta 0:04:21 lr 0.000676 time 0.2879 (0.2968) loss 3.7383 (3.6512) grad_norm 1.3556 (1.3315) [2021-04-16 01:51:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][380/1251] eta 0:04:18 lr 0.000676 time 0.2757 (0.2963) loss 3.6501 (3.6506) grad_norm 1.6556 (1.3320) [2021-04-16 01:51:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][390/1251] eta 0:04:14 lr 0.000676 time 0.2565 (0.2958) loss 2.8011 (3.6458) grad_norm 1.3887 (1.3373) [2021-04-16 01:52:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][400/1251] eta 0:04:11 lr 0.000676 time 0.2569 (0.2953) loss 3.2454 (3.6518) grad_norm 1.2401 (1.3364) [2021-04-16 01:52:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][410/1251] eta 0:04:07 lr 0.000676 time 0.2476 (0.2948) loss 3.9739 (3.6533) grad_norm 1.2671 (1.3361) [2021-04-16 01:52:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][420/1251] eta 0:04:04 lr 0.000676 time 0.2838 (0.2944) loss 4.3174 (3.6492) grad_norm 1.2184 (1.3351) [2021-04-16 01:52:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][430/1251] eta 0:04:01 lr 0.000676 time 0.2962 (0.2939) loss 3.2186 (3.6541) grad_norm 1.2459 (1.3340) [2021-04-16 01:52:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][440/1251] eta 0:03:58 lr 0.000676 time 0.2712 (0.2935) loss 4.1945 (3.6557) grad_norm 1.1023 (1.3327) [2021-04-16 01:52:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][450/1251] eta 0:03:54 lr 0.000676 time 0.2567 (0.2930) loss 2.9933 (3.6534) grad_norm 1.2981 (1.3323) [2021-04-16 01:52:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][460/1251] eta 0:03:51 lr 0.000676 time 0.2913 (0.2927) loss 4.1841 (3.6602) grad_norm 1.4356 (1.3335) [2021-04-16 01:52:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][470/1251] eta 0:03:48 lr 0.000676 time 0.2584 (0.2927) loss 3.5758 (3.6552) grad_norm 1.5990 (1.3343) [2021-04-16 01:52:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][480/1251] eta 0:03:45 lr 0.000676 time 0.3032 (0.2924) loss 4.4650 (3.6580) grad_norm 1.4088 (1.3344) [2021-04-16 01:52:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][490/1251] eta 0:03:42 lr 0.000676 time 0.2781 (0.2920) loss 3.9201 (3.6569) grad_norm 1.4504 (1.3365) [2021-04-16 01:52:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][500/1251] eta 0:03:38 lr 0.000676 time 0.2509 (0.2916) loss 2.8592 (3.6555) grad_norm 1.1820 (1.3351) [2021-04-16 01:52:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][510/1251] eta 0:03:35 lr 0.000676 time 0.2888 (0.2913) loss 3.7864 (3.6582) grad_norm 1.1586 (1.3350) [2021-04-16 01:52:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][520/1251] eta 0:03:32 lr 0.000676 time 0.2577 (0.2909) loss 2.4082 (3.6538) grad_norm 1.2507 (1.3345) [2021-04-16 01:52:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][530/1251] eta 0:03:29 lr 0.000675 time 0.2856 (0.2906) loss 2.4524 (3.6544) grad_norm 1.5573 (1.3342) [2021-04-16 01:52:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][540/1251] eta 0:03:26 lr 0.000675 time 0.2626 (0.2903) loss 3.9598 (3.6548) grad_norm 1.2028 (1.3325) [2021-04-16 01:52:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][550/1251] eta 0:03:23 lr 0.000675 time 0.2576 (0.2900) loss 3.9139 (3.6576) grad_norm 1.1824 (1.3328) [2021-04-16 01:52:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][560/1251] eta 0:03:20 lr 0.000675 time 0.2869 (0.2898) loss 3.0020 (3.6615) grad_norm 1.3146 (1.3347) [2021-04-16 01:52:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][570/1251] eta 0:03:17 lr 0.000675 time 0.2436 (0.2897) loss 3.4862 (3.6658) grad_norm 1.2425 (nan) [2021-04-16 01:52:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][580/1251] eta 0:03:14 lr 0.000675 time 0.2630 (0.2897) loss 4.4935 (3.6635) grad_norm 1.2075 (nan) [2021-04-16 01:52:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][590/1251] eta 0:03:11 lr 0.000675 time 0.2665 (0.2898) loss 4.1896 (3.6676) grad_norm 1.1670 (nan) [2021-04-16 01:52:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][600/1251] eta 0:03:08 lr 0.000675 time 0.2856 (0.2896) loss 3.3507 (3.6660) grad_norm 1.8330 (nan) [2021-04-16 01:52:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][610/1251] eta 0:03:05 lr 0.000675 time 0.3117 (0.2894) loss 3.9313 (3.6656) grad_norm 1.4320 (nan) [2021-04-16 01:53:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][620/1251] eta 0:03:02 lr 0.000675 time 0.3094 (0.2892) loss 3.7610 (3.6657) grad_norm 1.3072 (nan) [2021-04-16 01:53:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][630/1251] eta 0:02:59 lr 0.000675 time 0.2571 (0.2890) loss 3.2698 (3.6655) grad_norm 1.4378 (nan) [2021-04-16 01:53:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][640/1251] eta 0:02:56 lr 0.000675 time 0.2706 (0.2890) loss 4.1136 (3.6641) grad_norm 1.6135 (nan) [2021-04-16 01:53:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][650/1251] eta 0:02:53 lr 0.000675 time 0.4151 (0.2890) loss 4.0571 (3.6621) grad_norm 1.7245 (nan) [2021-04-16 01:53:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][660/1251] eta 0:02:50 lr 0.000675 time 0.2784 (0.2888) loss 3.4584 (3.6595) grad_norm 1.5222 (nan) [2021-04-16 01:53:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][670/1251] eta 0:02:47 lr 0.000675 time 0.2713 (0.2887) loss 3.6529 (3.6552) grad_norm 1.3839 (nan) [2021-04-16 01:53:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][680/1251] eta 0:02:44 lr 0.000675 time 0.2803 (0.2886) loss 3.1748 (3.6535) grad_norm 1.3407 (nan) [2021-04-16 01:53:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][690/1251] eta 0:02:41 lr 0.000675 time 0.2648 (0.2885) loss 4.1150 (3.6555) grad_norm 1.4151 (nan) [2021-04-16 01:53:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][700/1251] eta 0:02:39 lr 0.000675 time 0.2860 (0.2886) loss 4.1378 (3.6578) grad_norm 1.2078 (nan) [2021-04-16 01:53:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][710/1251] eta 0:02:36 lr 0.000675 time 0.2906 (0.2884) loss 3.9684 (3.6603) grad_norm 1.4320 (nan) [2021-04-16 01:53:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][720/1251] eta 0:02:33 lr 0.000675 time 0.3058 (0.2882) loss 4.1411 (3.6630) grad_norm 1.4763 (nan) [2021-04-16 01:53:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][730/1251] eta 0:02:30 lr 0.000675 time 0.2777 (0.2881) loss 4.4331 (3.6644) grad_norm 1.7197 (nan) [2021-04-16 01:53:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][740/1251] eta 0:02:27 lr 0.000675 time 0.2695 (0.2879) loss 2.5980 (3.6662) grad_norm 1.7838 (nan) [2021-04-16 01:53:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][750/1251] eta 0:02:24 lr 0.000675 time 0.2736 (0.2877) loss 3.6290 (3.6646) grad_norm 1.4055 (nan) [2021-04-16 01:53:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][760/1251] eta 0:02:21 lr 0.000675 time 0.2766 (0.2876) loss 3.7065 (3.6661) grad_norm 1.6244 (nan) [2021-04-16 01:53:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][770/1251] eta 0:02:18 lr 0.000675 time 0.2538 (0.2874) loss 3.0753 (3.6676) grad_norm 1.4522 (nan) [2021-04-16 01:53:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][780/1251] eta 0:02:15 lr 0.000675 time 0.2971 (0.2873) loss 3.8005 (3.6705) grad_norm 1.2993 (nan) [2021-04-16 01:53:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][790/1251] eta 0:02:12 lr 0.000674 time 0.2971 (0.2871) loss 4.3429 (3.6720) grad_norm 1.3267 (nan) [2021-04-16 01:53:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][800/1251] eta 0:02:09 lr 0.000674 time 0.2727 (0.2869) loss 4.1321 (3.6672) grad_norm 1.2758 (nan) [2021-04-16 01:53:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][810/1251] eta 0:02:06 lr 0.000674 time 0.2642 (0.2867) loss 3.8044 (3.6655) grad_norm 1.1327 (nan) [2021-04-16 01:53:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][820/1251] eta 0:02:03 lr 0.000674 time 0.2804 (0.2866) loss 2.8725 (3.6658) grad_norm 1.3097 (nan) [2021-04-16 01:54:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][830/1251] eta 0:02:00 lr 0.000674 time 0.2799 (0.2865) loss 3.0884 (3.6636) grad_norm 1.3277 (nan) [2021-04-16 01:54:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][840/1251] eta 0:01:57 lr 0.000674 time 0.2660 (0.2864) loss 2.8709 (3.6618) grad_norm 1.6162 (nan) [2021-04-16 01:54:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][850/1251] eta 0:01:54 lr 0.000674 time 0.2703 (0.2863) loss 3.8969 (3.6621) grad_norm 1.2458 (nan) [2021-04-16 01:54:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][860/1251] eta 0:01:51 lr 0.000674 time 0.2624 (0.2862) loss 3.6944 (3.6647) grad_norm 1.4856 (nan) [2021-04-16 01:54:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][870/1251] eta 0:01:49 lr 0.000674 time 0.2703 (0.2861) loss 4.3570 (3.6651) grad_norm 1.1071 (nan) [2021-04-16 01:54:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][880/1251] eta 0:01:46 lr 0.000674 time 0.2986 (0.2861) loss 4.5294 (3.6640) grad_norm 1.2398 (nan) [2021-04-16 01:54:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][890/1251] eta 0:01:43 lr 0.000674 time 0.2574 (0.2860) loss 4.0332 (3.6650) grad_norm 1.2840 (nan) [2021-04-16 01:54:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][900/1251] eta 0:01:40 lr 0.000674 time 0.2602 (0.2858) loss 3.3289 (3.6634) grad_norm 1.3467 (nan) [2021-04-16 01:54:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][910/1251] eta 0:01:37 lr 0.000674 time 0.2609 (0.2857) loss 2.8377 (3.6631) grad_norm 1.2267 (nan) [2021-04-16 01:54:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][920/1251] eta 0:01:34 lr 0.000674 time 0.2799 (0.2857) loss 3.6550 (3.6630) grad_norm 1.3666 (nan) [2021-04-16 01:54:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][930/1251] eta 0:01:31 lr 0.000674 time 0.2428 (0.2857) loss 2.2145 (3.6602) grad_norm 1.3375 (nan) [2021-04-16 01:54:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][940/1251] eta 0:01:28 lr 0.000674 time 0.2809 (0.2857) loss 3.5717 (3.6635) grad_norm 1.3363 (nan) [2021-04-16 01:54:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][950/1251] eta 0:01:25 lr 0.000674 time 0.2931 (0.2856) loss 2.9019 (3.6629) grad_norm 1.6632 (nan) [2021-04-16 01:54:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][960/1251] eta 0:01:23 lr 0.000674 time 0.2785 (0.2855) loss 3.0213 (3.6617) grad_norm 1.3366 (nan) [2021-04-16 01:54:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][970/1251] eta 0:01:20 lr 0.000674 time 0.2789 (0.2853) loss 3.4579 (3.6633) grad_norm 1.6526 (nan) [2021-04-16 01:54:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][980/1251] eta 0:01:17 lr 0.000674 time 0.2863 (0.2852) loss 4.2470 (3.6639) grad_norm 1.1990 (nan) [2021-04-16 01:54:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][990/1251] eta 0:01:14 lr 0.000674 time 0.2990 (0.2852) loss 3.7082 (3.6611) grad_norm 1.1937 (nan) [2021-04-16 01:54:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][1000/1251] eta 0:01:11 lr 0.000674 time 0.2640 (0.2851) loss 3.2998 (3.6607) grad_norm 1.1734 (nan) [2021-04-16 01:54:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][1010/1251] eta 0:01:08 lr 0.000674 time 0.2864 (0.2850) loss 3.8383 (3.6615) grad_norm 1.4252 (nan) [2021-04-16 01:54:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][1020/1251] eta 0:01:05 lr 0.000674 time 0.2589 (0.2849) loss 4.5069 (3.6634) grad_norm 1.3312 (nan) [2021-04-16 01:54:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][1030/1251] eta 0:01:02 lr 0.000674 time 0.2887 (0.2848) loss 2.9785 (3.6591) grad_norm 1.3684 (nan) [2021-04-16 01:54:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][1040/1251] eta 0:01:00 lr 0.000673 time 0.2788 (0.2849) loss 2.7100 (3.6578) grad_norm 1.4678 (nan) [2021-04-16 01:55:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][1050/1251] eta 0:00:57 lr 0.000673 time 0.2738 (0.2848) loss 3.9094 (3.6593) grad_norm 1.2803 (nan) [2021-04-16 01:55:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][1060/1251] eta 0:00:54 lr 0.000673 time 0.2802 (0.2847) loss 3.9116 (3.6590) grad_norm 1.2856 (nan) [2021-04-16 01:55:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][1070/1251] eta 0:00:51 lr 0.000673 time 0.2680 (0.2845) loss 4.0721 (3.6573) grad_norm 1.6662 (nan) [2021-04-16 01:55:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][1080/1251] eta 0:00:48 lr 0.000673 time 0.2647 (0.2845) loss 3.8185 (3.6583) grad_norm 1.2333 (nan) [2021-04-16 01:55:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][1090/1251] eta 0:00:45 lr 0.000673 time 0.2724 (0.2844) loss 3.9444 (3.6573) grad_norm 1.3153 (nan) [2021-04-16 01:55:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][1100/1251] eta 0:00:42 lr 0.000673 time 0.2723 (0.2843) loss 3.6317 (3.6580) grad_norm 1.9887 (nan) [2021-04-16 01:55:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][1110/1251] eta 0:00:40 lr 0.000673 time 0.2779 (0.2843) loss 3.7947 (3.6573) grad_norm 1.1516 (nan) [2021-04-16 01:55:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][1120/1251] eta 0:00:37 lr 0.000673 time 0.2831 (0.2843) loss 3.6223 (3.6589) grad_norm 1.3899 (nan) [2021-04-16 01:55:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][1130/1251] eta 0:00:34 lr 0.000673 time 0.2570 (0.2842) loss 4.0748 (3.6615) grad_norm 1.8498 (nan) [2021-04-16 01:55:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][1140/1251] eta 0:00:31 lr 0.000673 time 0.2836 (0.2841) loss 4.0110 (3.6606) grad_norm 1.2640 (nan) [2021-04-16 01:55:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][1150/1251] eta 0:00:28 lr 0.000673 time 0.2980 (0.2842) loss 4.2128 (3.6594) grad_norm 1.4170 (nan) [2021-04-16 01:55:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][1160/1251] eta 0:00:25 lr 0.000673 time 0.2682 (0.2842) loss 3.6511 (3.6595) grad_norm 1.3844 (nan) [2021-04-16 01:55:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][1170/1251] eta 0:00:23 lr 0.000673 time 0.2559 (0.2843) loss 4.0186 (3.6615) grad_norm 1.2279 (nan) [2021-04-16 01:55:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][1180/1251] eta 0:00:20 lr 0.000673 time 0.2832 (0.2843) loss 2.4010 (3.6595) grad_norm 1.2026 (nan) [2021-04-16 01:55:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][1190/1251] eta 0:00:17 lr 0.000673 time 0.2430 (0.2842) loss 3.8386 (3.6615) grad_norm 1.5879 (nan) [2021-04-16 01:55:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][1200/1251] eta 0:00:14 lr 0.000673 time 0.2681 (0.2841) loss 3.7364 (3.6636) grad_norm 1.3779 (nan) [2021-04-16 01:55:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][1210/1251] eta 0:00:11 lr 0.000673 time 0.2919 (0.2840) loss 3.7987 (3.6652) grad_norm 1.1610 (nan) [2021-04-16 01:55:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][1220/1251] eta 0:00:08 lr 0.000673 time 0.2773 (0.2841) loss 3.6165 (3.6631) grad_norm 1.6061 (nan) [2021-04-16 01:55:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][1230/1251] eta 0:00:05 lr 0.000673 time 0.2546 (0.2841) loss 3.4650 (3.6635) grad_norm 1.1774 (nan) [2021-04-16 01:55:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][1240/1251] eta 0:00:03 lr 0.000673 time 0.2590 (0.2840) loss 2.8362 (3.6629) grad_norm 1.3434 (nan) [2021-04-16 01:55:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [116/300][1250/1251] eta 0:00:00 lr 0.000673 time 0.2391 (0.2838) loss 4.1316 (3.6640) grad_norm 1.4631 (nan) [2021-04-16 01:56:00 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 116 training takes 0:05:57 [2021-04-16 01:56:00 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_116.pth saving...... [2021-04-16 01:56:16 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_116.pth saved !!! [2021-04-16 01:56:17 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.143 (1.143) Loss 1.1200 (1.1200) Acc@1 74.121 (74.121) Acc@5 92.480 (92.480) [2021-04-16 01:56:19 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.943 (0.272) Loss 1.0980 (1.1147) Acc@1 75.488 (74.467) Acc@5 92.383 (91.992) [2021-04-16 01:56:22 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 1.165 (0.265) Loss 1.1285 (1.1048) Acc@1 72.656 (74.395) Acc@5 92.578 (92.434) [2021-04-16 01:56:23 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.156 (0.224) Loss 1.0897 (1.0990) Acc@1 74.414 (74.543) Acc@5 93.066 (92.465) [2021-04-16 01:56:25 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.380 (0.211) Loss 1.1616 (1.0974) Acc@1 74.414 (74.521) Acc@5 91.699 (92.519) [2021-04-16 01:56:28 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 74.446 Acc@5 92.504 [2021-04-16 01:56:28 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 74.4% [2021-04-16 01:56:28 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 74.48% [2021-04-16 01:56:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][0/1251] eta 1:34:18 lr 0.000673 time 4.5231 (4.5231) loss 3.7280 (3.7280) grad_norm 1.6679 (1.6679) [2021-04-16 01:56:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][10/1251] eta 0:13:41 lr 0.000673 time 0.2946 (0.6620) loss 2.5454 (3.4756) grad_norm 1.3805 (1.4756) [2021-04-16 01:56:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][20/1251] eta 0:09:52 lr 0.000673 time 0.2957 (0.4815) loss 4.2001 (3.5571) grad_norm 1.3591 (1.4116) [2021-04-16 01:56:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][30/1251] eta 0:08:26 lr 0.000673 time 0.2844 (0.4149) loss 4.5064 (3.5354) grad_norm 1.3484 (1.4215) [2021-04-16 01:56:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][40/1251] eta 0:07:42 lr 0.000673 time 0.2925 (0.3821) loss 3.6728 (3.6079) grad_norm 1.4439 (1.4012) [2021-04-16 01:56:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][50/1251] eta 0:07:13 lr 0.000672 time 0.2689 (0.3611) loss 4.8677 (3.6454) grad_norm 1.3245 (1.3813) [2021-04-16 01:56:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][60/1251] eta 0:06:55 lr 0.000672 time 0.2482 (0.3490) loss 3.1898 (3.6139) grad_norm 1.2130 (1.3759) [2021-04-16 01:56:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][70/1251] eta 0:06:40 lr 0.000672 time 0.2841 (0.3392) loss 3.8977 (3.5498) grad_norm 1.2857 (1.3717) [2021-04-16 01:56:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][80/1251] eta 0:06:27 lr 0.000672 time 0.2752 (0.3310) loss 4.4778 (3.5785) grad_norm 1.8014 (1.3836) [2021-04-16 01:56:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][90/1251] eta 0:06:17 lr 0.000672 time 0.2896 (0.3256) loss 4.4737 (3.5818) grad_norm 1.4798 (1.3807) [2021-04-16 01:57:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][100/1251] eta 0:06:09 lr 0.000672 time 0.2788 (0.3207) loss 4.2447 (3.5962) grad_norm 1.5241 (1.3798) [2021-04-16 01:57:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][110/1251] eta 0:06:01 lr 0.000672 time 0.2528 (0.3165) loss 4.3601 (3.6059) grad_norm 1.3399 (1.3740) [2021-04-16 01:57:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][120/1251] eta 0:05:55 lr 0.000672 time 0.2785 (0.3146) loss 4.2660 (3.5726) grad_norm 1.3506 (1.3705) [2021-04-16 01:57:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][130/1251] eta 0:05:50 lr 0.000672 time 0.2919 (0.3125) loss 3.7903 (3.5864) grad_norm 1.1469 (1.3626) [2021-04-16 01:57:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][140/1251] eta 0:05:44 lr 0.000672 time 0.2926 (0.3100) loss 4.0617 (3.5824) grad_norm 1.4974 (1.3604) [2021-04-16 01:57:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][150/1251] eta 0:05:38 lr 0.000672 time 0.3036 (0.3078) loss 2.4518 (3.5858) grad_norm 1.1906 (1.3542) [2021-04-16 01:57:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][160/1251] eta 0:05:33 lr 0.000672 time 0.2877 (0.3060) loss 4.4810 (3.5836) grad_norm 1.2978 (1.3537) [2021-04-16 01:57:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][170/1251] eta 0:05:28 lr 0.000672 time 0.2456 (0.3043) loss 2.1847 (3.5850) grad_norm 1.3911 (1.3538) [2021-04-16 01:57:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][180/1251] eta 0:05:24 lr 0.000672 time 0.2730 (0.3028) loss 3.6485 (3.5899) grad_norm 1.4927 (1.3615) [2021-04-16 01:57:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][190/1251] eta 0:05:19 lr 0.000672 time 0.2863 (0.3016) loss 2.3558 (3.5760) grad_norm 1.3019 (1.3603) [2021-04-16 01:57:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][200/1251] eta 0:05:15 lr 0.000672 time 0.2975 (0.3004) loss 4.1386 (3.5873) grad_norm 1.2458 (1.3620) [2021-04-16 01:57:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][210/1251] eta 0:05:11 lr 0.000672 time 0.2804 (0.2991) loss 2.6955 (3.5836) grad_norm 1.1845 (1.3586) [2021-04-16 01:57:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][220/1251] eta 0:05:07 lr 0.000672 time 0.3034 (0.2983) loss 4.2190 (3.5878) grad_norm 1.1042 (1.3522) [2021-04-16 01:57:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][230/1251] eta 0:05:03 lr 0.000672 time 0.2771 (0.2973) loss 3.4764 (3.5986) grad_norm 1.2633 (1.3512) [2021-04-16 01:57:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][240/1251] eta 0:04:59 lr 0.000672 time 0.2767 (0.2967) loss 3.3528 (3.6052) grad_norm 1.4635 (1.3528) [2021-04-16 01:57:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][250/1251] eta 0:04:56 lr 0.000672 time 0.2768 (0.2965) loss 4.3396 (3.6043) grad_norm 1.3024 (1.3546) [2021-04-16 01:57:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][260/1251] eta 0:04:53 lr 0.000672 time 0.2564 (0.2958) loss 4.1179 (3.6078) grad_norm 1.1719 (1.3533) [2021-04-16 01:57:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][270/1251] eta 0:04:49 lr 0.000672 time 0.2787 (0.2951) loss 3.5948 (3.6172) grad_norm 1.1446 (1.3563) [2021-04-16 01:57:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][280/1251] eta 0:04:46 lr 0.000672 time 0.2863 (0.2946) loss 4.3632 (3.6264) grad_norm 1.4282 (1.3559) [2021-04-16 01:57:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][290/1251] eta 0:04:42 lr 0.000672 time 0.2719 (0.2939) loss 3.2839 (3.6338) grad_norm 1.2614 (1.3532) [2021-04-16 01:57:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][300/1251] eta 0:04:38 lr 0.000672 time 0.2636 (0.2933) loss 3.6918 (3.6438) grad_norm 1.2837 (1.3536) [2021-04-16 01:58:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][310/1251] eta 0:04:35 lr 0.000671 time 0.2957 (0.2929) loss 3.4481 (3.6501) grad_norm 1.4885 (1.3545) [2021-04-16 01:58:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][320/1251] eta 0:04:32 lr 0.000671 time 0.2827 (0.2924) loss 4.4046 (3.6516) grad_norm 1.4657 (1.3541) [2021-04-16 01:58:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][330/1251] eta 0:04:28 lr 0.000671 time 0.2804 (0.2919) loss 4.3473 (3.6571) grad_norm 1.3222 (1.3519) [2021-04-16 01:58:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][340/1251] eta 0:04:25 lr 0.000671 time 0.2798 (0.2914) loss 3.4731 (3.6578) grad_norm 1.2801 (1.3519) [2021-04-16 01:58:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][350/1251] eta 0:04:22 lr 0.000671 time 0.3977 (0.2918) loss 3.0877 (3.6582) grad_norm 1.4376 (1.3531) [2021-04-16 01:58:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][360/1251] eta 0:04:19 lr 0.000671 time 0.2688 (0.2912) loss 4.1100 (3.6594) grad_norm 1.0987 (1.3499) [2021-04-16 01:58:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][370/1251] eta 0:04:16 lr 0.000671 time 0.2645 (0.2913) loss 3.2575 (3.6547) grad_norm 1.2949 (1.3509) [2021-04-16 01:58:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][380/1251] eta 0:04:13 lr 0.000671 time 0.2589 (0.2908) loss 3.5564 (3.6539) grad_norm 1.3492 (1.3515) [2021-04-16 01:58:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][390/1251] eta 0:04:10 lr 0.000671 time 0.2900 (0.2904) loss 3.5245 (3.6525) grad_norm 1.3468 (1.3513) [2021-04-16 01:58:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][400/1251] eta 0:04:07 lr 0.000671 time 0.2617 (0.2905) loss 4.0410 (3.6640) grad_norm 1.1527 (1.3491) [2021-04-16 01:58:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][410/1251] eta 0:04:04 lr 0.000671 time 0.2790 (0.2902) loss 3.8252 (3.6599) grad_norm 1.2254 (1.3490) [2021-04-16 01:58:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][420/1251] eta 0:04:00 lr 0.000671 time 0.2700 (0.2897) loss 2.9620 (3.6618) grad_norm 1.4119 (1.3479) [2021-04-16 01:58:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][430/1251] eta 0:03:57 lr 0.000671 time 0.2701 (0.2897) loss 3.8646 (3.6685) grad_norm 1.3439 (1.3485) [2021-04-16 01:58:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][440/1251] eta 0:03:54 lr 0.000671 time 0.2689 (0.2895) loss 2.8802 (3.6621) grad_norm 1.2320 (1.3489) [2021-04-16 01:58:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][450/1251] eta 0:03:51 lr 0.000671 time 0.2695 (0.2891) loss 3.3058 (3.6599) grad_norm 1.2194 (1.3488) [2021-04-16 01:58:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][460/1251] eta 0:03:48 lr 0.000671 time 0.2962 (0.2887) loss 3.8315 (3.6602) grad_norm 1.1301 (1.3464) [2021-04-16 01:58:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][470/1251] eta 0:03:45 lr 0.000671 time 0.3000 (0.2885) loss 4.5281 (3.6556) grad_norm 1.4038 (1.3457) [2021-04-16 01:58:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][480/1251] eta 0:03:42 lr 0.000671 time 0.2783 (0.2883) loss 2.7986 (3.6522) grad_norm 1.3089 (1.3445) [2021-04-16 01:58:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][490/1251] eta 0:03:39 lr 0.000671 time 0.2809 (0.2883) loss 3.1121 (3.6467) grad_norm 1.3019 (1.3424) [2021-04-16 01:58:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][500/1251] eta 0:03:36 lr 0.000671 time 0.2716 (0.2880) loss 3.1343 (3.6508) grad_norm 1.3754 (1.3426) [2021-04-16 01:58:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][510/1251] eta 0:03:33 lr 0.000671 time 0.2513 (0.2877) loss 3.7416 (3.6431) grad_norm 1.1801 (1.3411) [2021-04-16 01:58:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][520/1251] eta 0:03:30 lr 0.000671 time 0.2916 (0.2878) loss 3.6962 (3.6471) grad_norm 1.1236 (1.3412) [2021-04-16 01:59:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][530/1251] eta 0:03:27 lr 0.000671 time 0.2740 (0.2875) loss 4.0647 (3.6436) grad_norm 1.1769 (1.3411) [2021-04-16 01:59:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][540/1251] eta 0:03:24 lr 0.000671 time 0.2837 (0.2873) loss 3.4837 (3.6460) grad_norm 1.2667 (1.3397) [2021-04-16 01:59:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][550/1251] eta 0:03:21 lr 0.000671 time 0.2831 (0.2871) loss 3.8539 (3.6464) grad_norm 1.3972 (1.3388) [2021-04-16 01:59:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][560/1251] eta 0:03:18 lr 0.000670 time 0.2808 (0.2868) loss 3.3312 (3.6444) grad_norm 1.3809 (1.3399) [2021-04-16 01:59:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][570/1251] eta 0:03:15 lr 0.000670 time 0.2702 (0.2872) loss 3.6839 (3.6459) grad_norm 1.1876 (1.3389) [2021-04-16 01:59:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][580/1251] eta 0:03:12 lr 0.000670 time 0.2563 (0.2869) loss 3.4321 (3.6476) grad_norm 1.4241 (1.3379) [2021-04-16 01:59:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][590/1251] eta 0:03:09 lr 0.000670 time 0.2528 (0.2866) loss 3.9932 (3.6480) grad_norm 1.5537 (1.3387) [2021-04-16 01:59:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][600/1251] eta 0:03:06 lr 0.000670 time 0.2685 (0.2864) loss 3.2543 (3.6483) grad_norm 1.3960 (1.3403) [2021-04-16 01:59:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][610/1251] eta 0:03:03 lr 0.000670 time 0.2586 (0.2862) loss 3.8468 (3.6506) grad_norm 1.4396 (1.3428) [2021-04-16 01:59:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][620/1251] eta 0:03:00 lr 0.000670 time 0.2771 (0.2861) loss 3.6052 (3.6566) grad_norm 1.2018 (1.3426) [2021-04-16 01:59:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][630/1251] eta 0:02:57 lr 0.000670 time 0.2695 (0.2858) loss 3.9135 (3.6542) grad_norm 1.4549 (1.3421) [2021-04-16 01:59:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][640/1251] eta 0:02:54 lr 0.000670 time 0.2744 (0.2857) loss 2.8698 (3.6512) grad_norm 1.3596 (1.3411) [2021-04-16 01:59:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][650/1251] eta 0:02:51 lr 0.000670 time 0.2834 (0.2855) loss 3.1535 (3.6489) grad_norm 1.5153 (1.3407) [2021-04-16 01:59:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][660/1251] eta 0:02:48 lr 0.000670 time 0.2855 (0.2853) loss 3.4920 (3.6511) grad_norm 1.3125 (1.3413) [2021-04-16 01:59:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][670/1251] eta 0:02:45 lr 0.000670 time 0.2822 (0.2852) loss 3.6045 (3.6511) grad_norm 1.3851 (1.3422) [2021-04-16 01:59:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][680/1251] eta 0:02:42 lr 0.000670 time 0.3017 (0.2850) loss 4.4560 (3.6525) grad_norm 1.3123 (1.3421) [2021-04-16 01:59:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][690/1251] eta 0:02:39 lr 0.000670 time 0.2909 (0.2850) loss 4.2446 (3.6515) grad_norm 1.0968 (1.3405) [2021-04-16 01:59:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][700/1251] eta 0:02:37 lr 0.000670 time 0.2682 (0.2850) loss 4.0816 (3.6517) grad_norm 1.4939 (1.3396) [2021-04-16 01:59:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][710/1251] eta 0:02:34 lr 0.000670 time 0.2986 (0.2850) loss 4.5354 (3.6503) grad_norm 1.4404 (1.3393) [2021-04-16 01:59:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][720/1251] eta 0:02:31 lr 0.000670 time 0.2807 (0.2850) loss 2.9251 (3.6454) grad_norm 1.4788 (1.3390) [2021-04-16 01:59:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][730/1251] eta 0:02:28 lr 0.000670 time 0.2703 (0.2848) loss 4.1128 (3.6477) grad_norm 1.6302 (1.3398) [2021-04-16 01:59:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][740/1251] eta 0:02:25 lr 0.000670 time 0.2783 (0.2847) loss 3.8541 (3.6476) grad_norm 1.3584 (1.3406) [2021-04-16 02:00:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][750/1251] eta 0:02:22 lr 0.000670 time 0.2798 (0.2846) loss 3.2678 (3.6500) grad_norm 1.3306 (1.3413) [2021-04-16 02:00:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][760/1251] eta 0:02:19 lr 0.000670 time 0.3034 (0.2845) loss 3.8334 (3.6509) grad_norm 1.2058 (1.3408) [2021-04-16 02:00:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][770/1251] eta 0:02:16 lr 0.000670 time 0.2886 (0.2844) loss 4.1392 (3.6541) grad_norm 1.2634 (1.3410) [2021-04-16 02:00:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][780/1251] eta 0:02:13 lr 0.000670 time 0.2842 (0.2844) loss 3.8817 (3.6548) grad_norm 1.4702 (1.3406) [2021-04-16 02:00:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][790/1251] eta 0:02:11 lr 0.000670 time 0.2544 (0.2844) loss 2.5885 (3.6569) grad_norm 1.1154 (1.3399) [2021-04-16 02:00:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][800/1251] eta 0:02:08 lr 0.000670 time 0.2546 (0.2843) loss 4.0905 (3.6562) grad_norm 1.1731 (1.3399) [2021-04-16 02:00:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][810/1251] eta 0:02:05 lr 0.000670 time 0.2646 (0.2842) loss 3.9321 (3.6548) grad_norm 1.3859 (1.3402) [2021-04-16 02:00:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][820/1251] eta 0:02:02 lr 0.000669 time 0.2731 (0.2841) loss 4.5335 (3.6554) grad_norm 1.1712 (1.3400) [2021-04-16 02:00:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][830/1251] eta 0:01:59 lr 0.000669 time 0.2644 (0.2841) loss 3.6844 (3.6553) grad_norm 1.1941 (1.3391) [2021-04-16 02:00:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][840/1251] eta 0:01:56 lr 0.000669 time 0.3223 (0.2841) loss 4.1592 (3.6568) grad_norm 1.2175 (1.3387) [2021-04-16 02:00:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][850/1251] eta 0:01:53 lr 0.000669 time 0.2828 (0.2840) loss 3.0008 (3.6573) grad_norm 1.2116 (1.3380) [2021-04-16 02:00:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][860/1251] eta 0:01:51 lr 0.000669 time 0.2785 (0.2839) loss 2.9578 (3.6525) grad_norm 1.3695 (1.3373) [2021-04-16 02:00:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][870/1251] eta 0:01:48 lr 0.000669 time 0.2431 (0.2839) loss 3.5892 (3.6507) grad_norm 1.3779 (1.3381) [2021-04-16 02:00:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][880/1251] eta 0:01:45 lr 0.000669 time 0.3232 (0.2838) loss 3.5141 (3.6520) grad_norm 1.1676 (1.3380) [2021-04-16 02:00:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][890/1251] eta 0:01:42 lr 0.000669 time 0.2914 (0.2838) loss 3.5671 (3.6533) grad_norm 1.3624 (1.3373) [2021-04-16 02:00:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][900/1251] eta 0:01:39 lr 0.000669 time 0.2651 (0.2838) loss 4.5860 (3.6573) grad_norm 1.4240 (1.3380) [2021-04-16 02:00:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][910/1251] eta 0:01:36 lr 0.000669 time 0.2860 (0.2837) loss 2.7639 (3.6575) grad_norm 1.4003 (1.3392) [2021-04-16 02:00:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][920/1251] eta 0:01:33 lr 0.000669 time 0.2481 (0.2837) loss 4.2941 (3.6608) grad_norm 1.2887 (1.3399) [2021-04-16 02:00:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][930/1251] eta 0:01:31 lr 0.000669 time 0.2977 (0.2838) loss 3.6401 (3.6576) grad_norm 1.1923 (1.3392) [2021-04-16 02:00:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][940/1251] eta 0:01:28 lr 0.000669 time 0.2735 (0.2838) loss 4.4274 (3.6620) grad_norm 1.1683 (1.3386) [2021-04-16 02:00:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][950/1251] eta 0:01:25 lr 0.000669 time 0.2635 (0.2838) loss 3.8229 (3.6638) grad_norm 1.4616 (1.3390) [2021-04-16 02:01:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][960/1251] eta 0:01:22 lr 0.000669 time 0.2973 (0.2838) loss 3.8860 (3.6622) grad_norm 1.4821 (1.3394) [2021-04-16 02:01:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][970/1251] eta 0:01:19 lr 0.000669 time 0.3054 (0.2838) loss 4.3500 (3.6627) grad_norm 1.1558 (1.3388) [2021-04-16 02:01:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][980/1251] eta 0:01:16 lr 0.000669 time 0.2711 (0.2838) loss 3.5430 (3.6609) grad_norm 1.3864 (1.3385) [2021-04-16 02:01:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][990/1251] eta 0:01:14 lr 0.000669 time 0.2830 (0.2837) loss 3.4772 (3.6594) grad_norm 1.1945 (1.3387) [2021-04-16 02:01:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][1000/1251] eta 0:01:11 lr 0.000669 time 0.2549 (0.2837) loss 3.6491 (3.6573) grad_norm 1.3509 (1.3387) [2021-04-16 02:01:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][1010/1251] eta 0:01:08 lr 0.000669 time 0.2944 (0.2838) loss 3.1519 (3.6573) grad_norm 1.3820 (1.3383) [2021-04-16 02:01:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][1020/1251] eta 0:01:05 lr 0.000669 time 0.3084 (0.2838) loss 2.9936 (3.6585) grad_norm 1.2327 (1.3387) [2021-04-16 02:01:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][1030/1251] eta 0:01:02 lr 0.000669 time 0.2747 (0.2837) loss 4.2688 (3.6587) grad_norm 1.3046 (1.3399) [2021-04-16 02:01:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][1040/1251] eta 0:00:59 lr 0.000669 time 0.2584 (0.2836) loss 3.6881 (3.6582) grad_norm 1.4308 (1.3412) [2021-04-16 02:01:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][1050/1251] eta 0:00:57 lr 0.000669 time 0.2876 (0.2836) loss 4.3197 (3.6577) grad_norm 1.4255 (1.3420) [2021-04-16 02:01:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][1060/1251] eta 0:00:54 lr 0.000669 time 0.2930 (0.2836) loss 3.9204 (3.6570) grad_norm 1.3202 (1.3418) [2021-04-16 02:01:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][1070/1251] eta 0:00:51 lr 0.000668 time 0.2774 (0.2836) loss 4.4901 (3.6586) grad_norm 1.1063 (1.3411) [2021-04-16 02:01:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][1080/1251] eta 0:00:48 lr 0.000668 time 0.2865 (0.2835) loss 3.8879 (3.6594) grad_norm 1.3532 (1.3411) [2021-04-16 02:01:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][1090/1251] eta 0:00:45 lr 0.000668 time 0.2555 (0.2834) loss 3.6878 (3.6583) grad_norm 1.3109 (1.3411) [2021-04-16 02:01:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][1100/1251] eta 0:00:42 lr 0.000668 time 0.2917 (0.2835) loss 4.3116 (3.6572) grad_norm 1.3654 (1.3413) [2021-04-16 02:01:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][1110/1251] eta 0:00:39 lr 0.000668 time 0.2506 (0.2834) loss 2.7862 (3.6584) grad_norm 1.2120 (1.3410) [2021-04-16 02:01:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][1120/1251] eta 0:00:37 lr 0.000668 time 0.2830 (0.2834) loss 4.1355 (3.6572) grad_norm 1.5172 (1.3412) [2021-04-16 02:01:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][1130/1251] eta 0:00:34 lr 0.000668 time 0.2597 (0.2833) loss 3.8577 (3.6577) grad_norm 1.3640 (1.3416) [2021-04-16 02:01:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][1140/1251] eta 0:00:31 lr 0.000668 time 0.2901 (0.2834) loss 2.6550 (3.6563) grad_norm 1.6841 (1.3422) [2021-04-16 02:01:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][1150/1251] eta 0:00:28 lr 0.000668 time 0.2519 (0.2833) loss 3.2475 (3.6539) grad_norm 1.3075 (1.3416) [2021-04-16 02:01:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][1160/1251] eta 0:00:25 lr 0.000668 time 0.3037 (0.2833) loss 3.2951 (3.6548) grad_norm 1.2161 (1.3419) [2021-04-16 02:02:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][1170/1251] eta 0:00:22 lr 0.000668 time 0.2486 (0.2833) loss 3.5653 (3.6567) grad_norm 1.2901 (1.3425) [2021-04-16 02:02:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][1180/1251] eta 0:00:20 lr 0.000668 time 0.2635 (0.2833) loss 2.9489 (3.6574) grad_norm 1.3670 (1.3421) [2021-04-16 02:02:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][1190/1251] eta 0:00:17 lr 0.000668 time 0.2752 (0.2833) loss 2.6404 (3.6569) grad_norm 1.8360 (1.3425) [2021-04-16 02:02:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][1200/1251] eta 0:00:14 lr 0.000668 time 0.2907 (0.2832) loss 3.1295 (3.6558) grad_norm 1.5017 (1.3433) [2021-04-16 02:02:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][1210/1251] eta 0:00:11 lr 0.000668 time 0.2896 (0.2832) loss 3.2315 (3.6531) grad_norm 1.4323 (1.3430) [2021-04-16 02:02:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][1220/1251] eta 0:00:08 lr 0.000668 time 0.2796 (0.2832) loss 3.8976 (3.6539) grad_norm 1.3756 (1.3438) [2021-04-16 02:02:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][1230/1251] eta 0:00:05 lr 0.000668 time 0.2788 (0.2831) loss 3.2785 (3.6526) grad_norm 1.4204 (1.3440) [2021-04-16 02:02:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][1240/1251] eta 0:00:03 lr 0.000668 time 0.2521 (0.2830) loss 3.9797 (3.6522) grad_norm 1.4273 (1.3437) [2021-04-16 02:02:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [117/300][1250/1251] eta 0:00:00 lr 0.000668 time 0.2485 (0.2828) loss 3.7498 (3.6536) grad_norm 1.2870 (1.3435) [2021-04-16 02:02:25 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 117 training takes 0:05:56 [2021-04-16 02:02:25 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_117.pth saving...... [2021-04-16 02:02:37 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_117.pth saved !!! [2021-04-16 02:02:38 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.092 (1.092) Loss 1.1203 (1.1203) Acc@1 74.414 (74.414) Acc@5 91.309 (91.309) [2021-04-16 02:02:40 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.145 (0.219) Loss 1.1180 (1.0891) Acc@1 72.949 (74.192) Acc@5 92.090 (92.276) [2021-04-16 02:02:42 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.113 (0.238) Loss 1.0672 (1.0749) Acc@1 75.098 (74.628) Acc@5 93.262 (92.732) [2021-04-16 02:02:44 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.098 (0.233) Loss 1.1221 (1.0826) Acc@1 73.535 (74.534) Acc@5 92.480 (92.632) [2021-04-16 02:02:46 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.212) Loss 0.9975 (1.0817) Acc@1 75.684 (74.538) Acc@5 92.578 (92.647) [2021-04-16 02:02:51 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 74.488 Acc@5 92.616 [2021-04-16 02:02:51 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 74.5% [2021-04-16 02:02:51 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 74.49% [2021-04-16 02:02:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][0/1251] eta 1:15:45 lr 0.000668 time 3.6335 (3.6335) loss 3.6964 (3.6964) grad_norm 1.3416 (1.3416) [2021-04-16 02:02:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][10/1251] eta 0:11:59 lr 0.000668 time 0.2631 (0.5797) loss 3.0769 (3.5349) grad_norm 1.4792 (1.3489) [2021-04-16 02:03:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][20/1251] eta 0:08:55 lr 0.000668 time 0.2898 (0.4353) loss 3.5218 (3.6339) grad_norm 1.2867 (1.3397) [2021-04-16 02:03:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][30/1251] eta 0:07:52 lr 0.000668 time 0.2930 (0.3873) loss 4.3895 (3.6696) grad_norm 1.5249 (1.3401) [2021-04-16 02:03:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][40/1251] eta 0:07:15 lr 0.000668 time 0.2618 (0.3595) loss 2.6256 (3.6011) grad_norm 1.3832 (1.3363) [2021-04-16 02:03:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][50/1251] eta 0:06:51 lr 0.000668 time 0.2714 (0.3430) loss 4.5650 (3.5763) grad_norm 1.2703 (1.3381) [2021-04-16 02:03:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][60/1251] eta 0:06:36 lr 0.000668 time 0.2892 (0.3333) loss 4.5262 (3.6456) grad_norm 1.5165 (1.3277) [2021-04-16 02:03:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][70/1251] eta 0:06:24 lr 0.000668 time 0.2999 (0.3256) loss 3.8153 (3.6653) grad_norm 1.5857 (1.3303) [2021-04-16 02:03:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][80/1251] eta 0:06:13 lr 0.000667 time 0.2843 (0.3193) loss 3.6387 (3.6659) grad_norm 1.2633 (1.3313) [2021-04-16 02:03:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][90/1251] eta 0:06:05 lr 0.000667 time 0.2901 (0.3147) loss 4.2798 (3.6831) grad_norm 1.5639 (1.3342) [2021-04-16 02:03:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][100/1251] eta 0:05:57 lr 0.000667 time 0.2666 (0.3105) loss 3.7403 (3.6481) grad_norm 1.3499 (1.3382) [2021-04-16 02:03:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][110/1251] eta 0:05:50 lr 0.000667 time 0.2511 (0.3070) loss 3.4379 (3.6270) grad_norm 1.3557 (inf) [2021-04-16 02:03:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][120/1251] eta 0:05:45 lr 0.000667 time 0.2610 (0.3052) loss 3.7104 (3.6393) grad_norm 1.2575 (inf) [2021-04-16 02:03:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][130/1251] eta 0:05:41 lr 0.000667 time 0.2848 (0.3051) loss 2.6584 (3.6366) grad_norm 1.4036 (inf) [2021-04-16 02:03:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][140/1251] eta 0:05:37 lr 0.000667 time 0.2603 (0.3042) loss 4.1965 (3.6380) grad_norm 1.3376 (inf) [2021-04-16 02:03:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][150/1251] eta 0:05:32 lr 0.000667 time 0.2648 (0.3021) loss 4.3017 (3.6368) grad_norm 1.4866 (inf) [2021-04-16 02:03:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][160/1251] eta 0:05:27 lr 0.000667 time 0.2826 (0.3002) loss 3.7241 (3.6499) grad_norm 1.2322 (nan) [2021-04-16 02:03:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][170/1251] eta 0:05:22 lr 0.000667 time 0.2748 (0.2987) loss 3.1054 (3.6466) grad_norm 1.3950 (nan) [2021-04-16 02:03:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][180/1251] eta 0:05:19 lr 0.000667 time 0.4223 (0.2982) loss 4.0459 (3.6438) grad_norm 1.3104 (nan) [2021-04-16 02:03:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][190/1251] eta 0:05:15 lr 0.000667 time 0.2837 (0.2971) loss 2.7801 (3.6419) grad_norm 1.2344 (nan) [2021-04-16 02:03:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][200/1251] eta 0:05:11 lr 0.000667 time 0.4283 (0.2967) loss 3.8785 (3.6513) grad_norm 1.4071 (nan) [2021-04-16 02:03:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][210/1251] eta 0:05:07 lr 0.000667 time 0.2977 (0.2958) loss 3.5678 (3.6544) grad_norm 1.2175 (nan) [2021-04-16 02:03:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][220/1251] eta 0:05:03 lr 0.000667 time 0.2660 (0.2948) loss 3.8374 (3.6505) grad_norm 1.2913 (nan) [2021-04-16 02:03:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][230/1251] eta 0:05:00 lr 0.000667 time 0.2867 (0.2938) loss 3.3730 (3.6469) grad_norm 1.2143 (nan) [2021-04-16 02:04:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][240/1251] eta 0:04:56 lr 0.000667 time 0.2743 (0.2931) loss 4.3594 (3.6457) grad_norm 1.4764 (nan) [2021-04-16 02:04:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][250/1251] eta 0:04:52 lr 0.000667 time 0.2830 (0.2925) loss 2.8311 (3.6529) grad_norm 1.4307 (nan) [2021-04-16 02:04:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][260/1251] eta 0:04:49 lr 0.000667 time 0.3006 (0.2920) loss 3.9597 (3.6535) grad_norm 1.4102 (nan) [2021-04-16 02:04:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][270/1251] eta 0:04:46 lr 0.000667 time 0.3111 (0.2916) loss 3.6494 (3.6500) grad_norm 1.3304 (nan) [2021-04-16 02:04:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][280/1251] eta 0:04:42 lr 0.000667 time 0.2785 (0.2911) loss 4.2955 (3.6505) grad_norm 1.2633 (nan) [2021-04-16 02:04:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][290/1251] eta 0:04:39 lr 0.000667 time 0.2942 (0.2910) loss 2.5193 (3.6450) grad_norm 1.4412 (nan) [2021-04-16 02:04:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][300/1251] eta 0:04:36 lr 0.000667 time 0.2829 (0.2907) loss 3.3496 (3.6404) grad_norm 1.3461 (nan) [2021-04-16 02:04:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][310/1251] eta 0:04:33 lr 0.000667 time 0.2900 (0.2903) loss 2.9732 (3.6374) grad_norm 1.6547 (nan) [2021-04-16 02:04:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][320/1251] eta 0:04:29 lr 0.000667 time 0.2635 (0.2898) loss 4.5596 (3.6295) grad_norm 1.2576 (nan) [2021-04-16 02:04:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][330/1251] eta 0:04:26 lr 0.000666 time 0.2641 (0.2894) loss 3.7556 (3.6334) grad_norm 1.3723 (nan) [2021-04-16 02:04:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][340/1251] eta 0:04:23 lr 0.000666 time 0.2685 (0.2895) loss 2.6356 (3.6315) grad_norm 1.3323 (nan) [2021-04-16 02:04:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][350/1251] eta 0:04:20 lr 0.000666 time 0.2952 (0.2891) loss 2.3793 (3.6312) grad_norm 1.3121 (nan) [2021-04-16 02:04:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][360/1251] eta 0:04:17 lr 0.000666 time 0.2687 (0.2890) loss 3.8398 (3.6336) grad_norm 1.2565 (nan) [2021-04-16 02:04:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][370/1251] eta 0:04:14 lr 0.000666 time 0.2742 (0.2888) loss 3.9167 (3.6313) grad_norm 1.1905 (nan) [2021-04-16 02:04:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][380/1251] eta 0:04:11 lr 0.000666 time 0.2451 (0.2884) loss 4.2261 (3.6339) grad_norm 1.2811 (nan) [2021-04-16 02:04:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][390/1251] eta 0:04:08 lr 0.000666 time 0.2800 (0.2881) loss 4.0215 (3.6357) grad_norm 1.1510 (nan) [2021-04-16 02:04:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][400/1251] eta 0:04:04 lr 0.000666 time 0.2975 (0.2879) loss 4.3069 (3.6360) grad_norm 1.1570 (nan) [2021-04-16 02:04:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][410/1251] eta 0:04:01 lr 0.000666 time 0.2561 (0.2874) loss 3.6701 (3.6421) grad_norm 1.5768 (nan) [2021-04-16 02:04:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][420/1251] eta 0:03:58 lr 0.000666 time 0.2954 (0.2871) loss 3.7906 (3.6421) grad_norm 1.6875 (nan) [2021-04-16 02:04:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][430/1251] eta 0:03:55 lr 0.000666 time 0.2717 (0.2869) loss 3.7355 (3.6418) grad_norm 1.2780 (nan) [2021-04-16 02:04:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][440/1251] eta 0:03:52 lr 0.000666 time 0.2641 (0.2866) loss 4.5125 (3.6483) grad_norm 1.3296 (nan) [2021-04-16 02:05:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][450/1251] eta 0:03:49 lr 0.000666 time 0.2750 (0.2869) loss 3.6932 (3.6406) grad_norm 1.4634 (nan) [2021-04-16 02:05:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][460/1251] eta 0:03:47 lr 0.000666 time 0.2629 (0.2871) loss 3.2513 (3.6344) grad_norm 1.2673 (nan) [2021-04-16 02:05:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][470/1251] eta 0:03:44 lr 0.000666 time 0.2604 (0.2870) loss 4.0943 (3.6361) grad_norm 1.2649 (nan) [2021-04-16 02:05:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][480/1251] eta 0:03:41 lr 0.000666 time 0.2708 (0.2868) loss 3.9294 (3.6370) grad_norm 1.2907 (nan) [2021-04-16 02:05:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][490/1251] eta 0:03:38 lr 0.000666 time 0.2794 (0.2870) loss 4.5511 (3.6361) grad_norm 1.2913 (nan) [2021-04-16 02:05:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][500/1251] eta 0:03:35 lr 0.000666 time 0.2694 (0.2867) loss 3.6282 (3.6352) grad_norm 1.3731 (nan) [2021-04-16 02:05:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][510/1251] eta 0:03:32 lr 0.000666 time 0.2764 (0.2867) loss 3.8312 (3.6369) grad_norm 1.3124 (nan) [2021-04-16 02:05:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][520/1251] eta 0:03:29 lr 0.000666 time 0.2911 (0.2868) loss 4.0646 (3.6377) grad_norm 1.4718 (nan) [2021-04-16 02:05:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][530/1251] eta 0:03:26 lr 0.000666 time 0.2914 (0.2867) loss 3.8976 (3.6338) grad_norm 1.2753 (nan) [2021-04-16 02:05:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][540/1251] eta 0:03:23 lr 0.000666 time 0.2610 (0.2866) loss 2.6460 (3.6301) grad_norm 1.2119 (nan) [2021-04-16 02:05:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][550/1251] eta 0:03:20 lr 0.000666 time 0.2757 (0.2866) loss 3.4425 (3.6278) grad_norm 1.5024 (nan) [2021-04-16 02:05:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][560/1251] eta 0:03:17 lr 0.000666 time 0.3092 (0.2864) loss 3.0821 (3.6280) grad_norm 1.3044 (nan) [2021-04-16 02:05:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][570/1251] eta 0:03:14 lr 0.000666 time 0.2498 (0.2862) loss 4.0874 (3.6324) grad_norm 1.2539 (nan) [2021-04-16 02:05:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][580/1251] eta 0:03:11 lr 0.000666 time 0.2928 (0.2861) loss 4.4907 (3.6372) grad_norm 1.2306 (nan) [2021-04-16 02:05:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][590/1251] eta 0:03:08 lr 0.000665 time 0.2828 (0.2858) loss 4.0753 (3.6389) grad_norm 1.2748 (nan) [2021-04-16 02:05:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][600/1251] eta 0:03:05 lr 0.000665 time 0.2561 (0.2857) loss 2.6432 (3.6335) grad_norm 1.1777 (nan) [2021-04-16 02:05:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][610/1251] eta 0:03:03 lr 0.000665 time 0.2704 (0.2857) loss 4.0838 (3.6354) grad_norm 1.1250 (nan) [2021-04-16 02:05:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][620/1251] eta 0:03:00 lr 0.000665 time 0.2681 (0.2856) loss 3.6793 (3.6350) grad_norm 1.1751 (nan) [2021-04-16 02:05:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][630/1251] eta 0:02:57 lr 0.000665 time 0.2561 (0.2854) loss 3.7490 (3.6359) grad_norm 1.5951 (nan) [2021-04-16 02:05:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][640/1251] eta 0:02:54 lr 0.000665 time 0.2874 (0.2853) loss 3.9514 (3.6392) grad_norm 1.4395 (nan) [2021-04-16 02:05:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][650/1251] eta 0:02:51 lr 0.000665 time 0.2823 (0.2851) loss 2.7610 (3.6426) grad_norm 1.2514 (nan) [2021-04-16 02:05:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][660/1251] eta 0:02:48 lr 0.000665 time 0.2974 (0.2851) loss 3.4273 (3.6440) grad_norm 1.4075 (nan) [2021-04-16 02:06:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][670/1251] eta 0:02:45 lr 0.000665 time 0.2778 (0.2849) loss 3.5210 (3.6433) grad_norm 1.4903 (nan) [2021-04-16 02:06:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][680/1251] eta 0:02:42 lr 0.000665 time 0.2761 (0.2848) loss 4.4257 (3.6424) grad_norm 1.3011 (nan) [2021-04-16 02:06:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][690/1251] eta 0:02:39 lr 0.000665 time 0.2823 (0.2847) loss 4.1122 (3.6436) grad_norm 1.2424 (nan) [2021-04-16 02:06:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][700/1251] eta 0:02:36 lr 0.000665 time 0.2820 (0.2846) loss 2.6035 (3.6425) grad_norm 1.4194 (nan) [2021-04-16 02:06:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][710/1251] eta 0:02:34 lr 0.000665 time 0.2758 (0.2848) loss 3.7328 (3.6423) grad_norm 1.3299 (nan) [2021-04-16 02:06:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][720/1251] eta 0:02:31 lr 0.000665 time 0.2979 (0.2850) loss 3.7237 (3.6430) grad_norm 1.2799 (nan) [2021-04-16 02:06:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][730/1251] eta 0:02:28 lr 0.000665 time 0.2644 (0.2849) loss 3.8711 (3.6433) grad_norm 1.4920 (nan) [2021-04-16 02:06:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][740/1251] eta 0:02:25 lr 0.000665 time 0.3112 (0.2849) loss 4.3210 (3.6438) grad_norm 1.1949 (nan) [2021-04-16 02:06:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][750/1251] eta 0:02:22 lr 0.000665 time 0.2563 (0.2848) loss 4.1461 (3.6448) grad_norm 1.4343 (nan) [2021-04-16 02:06:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][760/1251] eta 0:02:19 lr 0.000665 time 0.2721 (0.2849) loss 4.2144 (3.6396) grad_norm 1.5307 (nan) [2021-04-16 02:06:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][770/1251] eta 0:02:17 lr 0.000665 time 0.2959 (0.2849) loss 4.1040 (3.6403) grad_norm 1.2187 (nan) [2021-04-16 02:06:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][780/1251] eta 0:02:14 lr 0.000665 time 0.2548 (0.2850) loss 3.4812 (3.6406) grad_norm 1.1694 (nan) [2021-04-16 02:06:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][790/1251] eta 0:02:11 lr 0.000665 time 0.2764 (0.2848) loss 4.1746 (3.6382) grad_norm 1.6136 (nan) [2021-04-16 02:06:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][800/1251] eta 0:02:08 lr 0.000665 time 0.3053 (0.2848) loss 3.8941 (3.6340) grad_norm 1.2890 (nan) [2021-04-16 02:06:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][810/1251] eta 0:02:05 lr 0.000665 time 0.2617 (0.2847) loss 3.4515 (3.6340) grad_norm 1.4127 (nan) [2021-04-16 02:06:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][820/1251] eta 0:02:02 lr 0.000665 time 0.2761 (0.2847) loss 3.9528 (3.6352) grad_norm 1.4700 (nan) [2021-04-16 02:06:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][830/1251] eta 0:01:59 lr 0.000665 time 0.2775 (0.2846) loss 3.7968 (3.6370) grad_norm 1.8992 (nan) [2021-04-16 02:06:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][840/1251] eta 0:01:56 lr 0.000664 time 0.2764 (0.2845) loss 4.1007 (3.6416) grad_norm 1.3430 (nan) [2021-04-16 02:06:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][850/1251] eta 0:01:54 lr 0.000664 time 0.2740 (0.2844) loss 4.0092 (3.6406) grad_norm 1.4460 (nan) [2021-04-16 02:06:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][860/1251] eta 0:01:51 lr 0.000664 time 0.2905 (0.2843) loss 4.7130 (3.6391) grad_norm 1.4680 (nan) [2021-04-16 02:06:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][870/1251] eta 0:01:48 lr 0.000664 time 0.2587 (0.2844) loss 2.6523 (3.6410) grad_norm 1.4240 (nan) [2021-04-16 02:07:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][880/1251] eta 0:01:45 lr 0.000664 time 0.3233 (0.2844) loss 3.1080 (3.6402) grad_norm 1.3939 (nan) [2021-04-16 02:07:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][890/1251] eta 0:01:42 lr 0.000664 time 0.2917 (0.2844) loss 2.6540 (3.6386) grad_norm 1.3410 (nan) [2021-04-16 02:07:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][900/1251] eta 0:01:39 lr 0.000664 time 0.2927 (0.2843) loss 3.7343 (3.6364) grad_norm 1.3093 (nan) [2021-04-16 02:07:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][910/1251] eta 0:01:36 lr 0.000664 time 0.2832 (0.2842) loss 2.7174 (3.6362) grad_norm 1.4492 (nan) [2021-04-16 02:07:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][920/1251] eta 0:01:34 lr 0.000664 time 0.2913 (0.2842) loss 4.0874 (3.6372) grad_norm 1.4472 (nan) [2021-04-16 02:07:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][930/1251] eta 0:01:31 lr 0.000664 time 0.2768 (0.2841) loss 4.4895 (3.6414) grad_norm 1.2657 (nan) [2021-04-16 02:07:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][940/1251] eta 0:01:28 lr 0.000664 time 0.2800 (0.2841) loss 3.7584 (3.6411) grad_norm 1.1289 (nan) [2021-04-16 02:07:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][950/1251] eta 0:01:25 lr 0.000664 time 0.2793 (0.2840) loss 4.0279 (3.6450) grad_norm 1.3193 (nan) [2021-04-16 02:07:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][960/1251] eta 0:01:22 lr 0.000664 time 0.2855 (0.2839) loss 4.1244 (3.6473) grad_norm 1.2093 (nan) [2021-04-16 02:07:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][970/1251] eta 0:01:19 lr 0.000664 time 0.2503 (0.2838) loss 4.2304 (3.6497) grad_norm 1.7166 (nan) [2021-04-16 02:07:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][980/1251] eta 0:01:16 lr 0.000664 time 0.2813 (0.2837) loss 3.2807 (3.6492) grad_norm 1.2611 (nan) [2021-04-16 02:07:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][990/1251] eta 0:01:14 lr 0.000664 time 0.3017 (0.2836) loss 3.7783 (3.6526) grad_norm 1.3181 (nan) [2021-04-16 02:07:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][1000/1251] eta 0:01:11 lr 0.000664 time 0.2966 (0.2836) loss 3.9199 (3.6521) grad_norm 1.4093 (nan) [2021-04-16 02:07:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][1010/1251] eta 0:01:08 lr 0.000664 time 0.2580 (0.2836) loss 3.8279 (3.6535) grad_norm 1.4179 (nan) [2021-04-16 02:07:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][1020/1251] eta 0:01:05 lr 0.000664 time 0.2569 (0.2835) loss 3.7147 (3.6536) grad_norm 1.3953 (nan) [2021-04-16 02:07:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][1030/1251] eta 0:01:02 lr 0.000664 time 0.2932 (0.2835) loss 3.6188 (3.6522) grad_norm 1.3672 (nan) [2021-04-16 02:07:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][1040/1251] eta 0:00:59 lr 0.000664 time 0.2871 (0.2835) loss 2.9749 (3.6532) grad_norm 1.6681 (nan) [2021-04-16 02:07:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][1050/1251] eta 0:00:56 lr 0.000664 time 0.2804 (0.2834) loss 4.7201 (3.6533) grad_norm 1.3228 (nan) [2021-04-16 02:07:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][1060/1251] eta 0:00:54 lr 0.000664 time 0.2893 (0.2833) loss 3.6300 (3.6525) grad_norm 1.4966 (nan) [2021-04-16 02:07:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][1070/1251] eta 0:00:51 lr 0.000664 time 0.2647 (0.2833) loss 4.0528 (3.6504) grad_norm 1.2595 (nan) [2021-04-16 02:07:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][1080/1251] eta 0:00:48 lr 0.000664 time 0.2613 (0.2833) loss 3.0178 (3.6493) grad_norm 1.3859 (nan) [2021-04-16 02:08:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][1090/1251] eta 0:00:45 lr 0.000664 time 0.2583 (0.2834) loss 3.8368 (3.6497) grad_norm 1.1291 (nan) [2021-04-16 02:08:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][1100/1251] eta 0:00:42 lr 0.000663 time 0.2648 (0.2833) loss 4.1714 (3.6512) grad_norm 1.3764 (nan) [2021-04-16 02:08:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][1110/1251] eta 0:00:39 lr 0.000663 time 0.2934 (0.2832) loss 3.7913 (3.6522) grad_norm 1.3163 (nan) [2021-04-16 02:08:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][1120/1251] eta 0:00:37 lr 0.000663 time 0.2687 (0.2833) loss 3.1897 (3.6521) grad_norm 1.3372 (nan) [2021-04-16 02:08:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][1130/1251] eta 0:00:34 lr 0.000663 time 0.2843 (0.2833) loss 4.4414 (3.6541) grad_norm 1.2146 (nan) [2021-04-16 02:08:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][1140/1251] eta 0:00:31 lr 0.000663 time 0.2695 (0.2832) loss 4.0941 (3.6539) grad_norm 1.2871 (nan) [2021-04-16 02:08:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][1150/1251] eta 0:00:28 lr 0.000663 time 0.2821 (0.2831) loss 3.2973 (3.6540) grad_norm 1.4055 (nan) [2021-04-16 02:08:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][1160/1251] eta 0:00:25 lr 0.000663 time 0.2742 (0.2831) loss 2.9579 (3.6563) grad_norm 1.2456 (nan) [2021-04-16 02:08:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][1170/1251] eta 0:00:22 lr 0.000663 time 0.2942 (0.2830) loss 4.0727 (3.6576) grad_norm 1.3139 (nan) [2021-04-16 02:08:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][1180/1251] eta 0:00:20 lr 0.000663 time 0.2543 (0.2831) loss 3.1881 (3.6565) grad_norm 1.4427 (nan) [2021-04-16 02:08:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][1190/1251] eta 0:00:17 lr 0.000663 time 0.2470 (0.2830) loss 4.2581 (3.6564) grad_norm 1.4861 (nan) [2021-04-16 02:08:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][1200/1251] eta 0:00:14 lr 0.000663 time 0.2933 (0.2829) loss 4.2878 (3.6566) grad_norm 1.3474 (nan) [2021-04-16 02:08:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][1210/1251] eta 0:00:11 lr 0.000663 time 0.2634 (0.2829) loss 3.7479 (3.6577) grad_norm 1.4011 (nan) [2021-04-16 02:08:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][1220/1251] eta 0:00:08 lr 0.000663 time 0.2747 (0.2829) loss 3.7818 (3.6567) grad_norm 1.5949 (nan) [2021-04-16 02:08:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][1230/1251] eta 0:00:05 lr 0.000663 time 0.2826 (0.2828) loss 3.9571 (3.6586) grad_norm 1.1557 (nan) [2021-04-16 02:08:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][1240/1251] eta 0:00:03 lr 0.000663 time 0.2484 (0.2827) loss 4.4834 (3.6590) grad_norm 1.1262 (nan) [2021-04-16 02:08:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [118/300][1250/1251] eta 0:00:00 lr 0.000663 time 0.2483 (0.2824) loss 3.5631 (3.6580) grad_norm 1.2046 (nan) [2021-04-16 02:08:46 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 118 training takes 0:05:55 [2021-04-16 02:08:46 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_118.pth saving...... [2021-04-16 02:08:59 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_118.pth saved !!! [2021-04-16 02:09:00 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.167 (1.167) Loss 1.0224 (1.0224) Acc@1 75.293 (75.293) Acc@5 93.848 (93.848) [2021-04-16 02:09:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.160 (0.253) Loss 1.1443 (1.0744) Acc@1 74.512 (74.538) Acc@5 91.797 (92.623) [2021-04-16 02:09:04 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.140 (0.228) Loss 1.0422 (1.0816) Acc@1 76.270 (74.340) Acc@5 92.188 (92.448) [2021-04-16 02:09:06 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.102 (0.235) Loss 1.0516 (1.0903) Acc@1 74.219 (74.121) Acc@5 93.457 (92.436) [2021-04-16 02:09:08 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.083 (0.218) Loss 1.1090 (1.0833) Acc@1 73.145 (74.262) Acc@5 92.090 (92.507) [2021-04-16 02:09:11 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 74.372 Acc@5 92.528 [2021-04-16 02:09:11 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 74.4% [2021-04-16 02:09:11 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 74.49% [2021-04-16 02:09:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][0/1251] eta 1:58:09 lr 0.000663 time 5.6670 (5.6670) loss 3.8271 (3.8271) grad_norm 1.3207 (1.3207) [2021-04-16 02:09:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][10/1251] eta 0:15:48 lr 0.000663 time 0.2460 (0.7641) loss 2.9082 (3.5291) grad_norm 1.3486 (1.3550) [2021-04-16 02:09:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][20/1251] eta 0:10:55 lr 0.000663 time 0.2774 (0.5323) loss 3.5608 (3.4802) grad_norm 1.3421 (1.3740) [2021-04-16 02:09:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][30/1251] eta 0:09:08 lr 0.000663 time 0.2519 (0.4491) loss 3.6247 (3.5448) grad_norm 1.1872 (1.3958) [2021-04-16 02:09:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][40/1251] eta 0:08:15 lr 0.000663 time 0.2539 (0.4089) loss 3.3905 (3.5798) grad_norm 1.3511 (1.4012) [2021-04-16 02:09:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][50/1251] eta 0:07:42 lr 0.000663 time 0.2647 (0.3851) loss 3.8601 (3.5322) grad_norm 1.3279 (1.3802) [2021-04-16 02:09:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][60/1251] eta 0:07:18 lr 0.000663 time 0.2500 (0.3678) loss 3.4391 (3.5708) grad_norm 1.1387 (1.3767) [2021-04-16 02:09:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][70/1251] eta 0:07:00 lr 0.000663 time 0.2782 (0.3560) loss 2.9198 (3.5470) grad_norm 1.0845 (1.3734) [2021-04-16 02:09:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][80/1251] eta 0:06:46 lr 0.000663 time 0.2628 (0.3468) loss 3.6864 (3.5557) grad_norm 1.6171 (1.3782) [2021-04-16 02:09:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][90/1251] eta 0:06:33 lr 0.000663 time 0.2596 (0.3390) loss 4.2966 (3.5661) grad_norm 1.4351 (1.3756) [2021-04-16 02:09:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][100/1251] eta 0:06:24 lr 0.000662 time 0.2879 (0.3340) loss 3.8598 (3.5924) grad_norm 1.4776 (1.3740) [2021-04-16 02:09:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][110/1251] eta 0:06:15 lr 0.000662 time 0.2583 (0.3292) loss 2.6042 (3.5999) grad_norm 1.8359 (1.3762) [2021-04-16 02:09:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][120/1251] eta 0:06:08 lr 0.000662 time 0.2697 (0.3261) loss 2.7293 (3.5915) grad_norm 1.5149 (1.3715) [2021-04-16 02:09:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][130/1251] eta 0:06:01 lr 0.000662 time 0.3013 (0.3225) loss 3.8612 (3.6168) grad_norm 1.7251 (1.3706) [2021-04-16 02:09:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][140/1251] eta 0:05:55 lr 0.000662 time 0.2867 (0.3200) loss 2.3792 (3.6110) grad_norm 1.3643 (1.3692) [2021-04-16 02:09:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][150/1251] eta 0:05:49 lr 0.000662 time 0.2809 (0.3173) loss 3.2594 (3.6098) grad_norm 1.7394 (1.3691) [2021-04-16 02:10:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][160/1251] eta 0:05:43 lr 0.000662 time 0.3042 (0.3148) loss 3.8536 (3.6278) grad_norm 1.4410 (1.3678) [2021-04-16 02:10:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][170/1251] eta 0:05:38 lr 0.000662 time 0.2557 (0.3130) loss 4.1781 (3.6344) grad_norm 1.3013 (1.3661) [2021-04-16 02:10:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][180/1251] eta 0:05:33 lr 0.000662 time 0.3051 (0.3110) loss 3.4959 (3.6339) grad_norm 1.4332 (1.3652) [2021-04-16 02:10:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][190/1251] eta 0:05:27 lr 0.000662 time 0.2721 (0.3091) loss 4.0295 (3.6307) grad_norm 1.2793 (1.3645) [2021-04-16 02:10:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][200/1251] eta 0:05:22 lr 0.000662 time 0.2732 (0.3073) loss 3.6248 (3.6318) grad_norm 1.2263 (1.3602) [2021-04-16 02:10:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][210/1251] eta 0:05:18 lr 0.000662 time 0.2593 (0.3061) loss 3.3510 (3.6291) grad_norm 1.3856 (1.3584) [2021-04-16 02:10:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][220/1251] eta 0:05:14 lr 0.000662 time 0.2570 (0.3047) loss 4.0970 (3.6267) grad_norm 1.4322 (1.3568) [2021-04-16 02:10:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][230/1251] eta 0:05:10 lr 0.000662 time 0.3059 (0.3038) loss 4.7984 (3.6408) grad_norm 1.3703 (1.3581) [2021-04-16 02:10:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][240/1251] eta 0:05:06 lr 0.000662 time 0.2814 (0.3029) loss 2.9436 (3.6327) grad_norm 1.5937 (1.3581) [2021-04-16 02:10:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][250/1251] eta 0:05:02 lr 0.000662 time 0.2914 (0.3020) loss 3.4301 (3.6328) grad_norm 1.5911 (1.3604) [2021-04-16 02:10:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][260/1251] eta 0:04:58 lr 0.000662 time 0.2927 (0.3012) loss 2.9341 (3.6211) grad_norm 1.6803 (1.3624) [2021-04-16 02:10:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][270/1251] eta 0:04:54 lr 0.000662 time 0.2608 (0.3003) loss 4.1838 (3.6281) grad_norm 1.3677 (1.3618) [2021-04-16 02:10:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][280/1251] eta 0:04:51 lr 0.000662 time 0.2828 (0.2997) loss 3.2571 (3.6298) grad_norm 1.6571 (1.3651) [2021-04-16 02:10:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][290/1251] eta 0:04:47 lr 0.000662 time 0.2904 (0.2996) loss 4.0857 (3.6343) grad_norm 1.3110 (1.3640) [2021-04-16 02:10:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][300/1251] eta 0:04:44 lr 0.000662 time 0.2704 (0.2994) loss 3.8812 (3.6388) grad_norm 1.3199 (1.3634) [2021-04-16 02:10:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][310/1251] eta 0:04:40 lr 0.000662 time 0.2814 (0.2986) loss 4.5382 (3.6355) grad_norm 1.4053 (1.3612) [2021-04-16 02:10:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][320/1251] eta 0:04:37 lr 0.000662 time 0.2814 (0.2981) loss 3.5753 (3.6225) grad_norm 1.2356 (1.3588) [2021-04-16 02:10:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][330/1251] eta 0:04:34 lr 0.000662 time 0.2899 (0.2979) loss 4.2763 (3.6217) grad_norm 1.2806 (1.3574) [2021-04-16 02:10:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][340/1251] eta 0:04:30 lr 0.000662 time 0.2967 (0.2973) loss 3.9139 (3.6224) grad_norm 1.2376 (1.3564) [2021-04-16 02:10:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][350/1251] eta 0:04:27 lr 0.000662 time 0.2856 (0.2967) loss 2.9607 (3.6134) grad_norm 1.3799 (1.3553) [2021-04-16 02:10:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][360/1251] eta 0:04:23 lr 0.000661 time 0.2516 (0.2962) loss 4.0896 (3.6117) grad_norm 1.2414 (1.3562) [2021-04-16 02:11:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][370/1251] eta 0:04:20 lr 0.000661 time 0.2699 (0.2957) loss 3.9051 (3.6224) grad_norm 1.4840 (1.3555) [2021-04-16 02:11:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][380/1251] eta 0:04:17 lr 0.000661 time 0.2960 (0.2957) loss 2.9101 (3.6318) grad_norm 1.3966 (1.3562) [2021-04-16 02:11:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][390/1251] eta 0:04:14 lr 0.000661 time 0.2851 (0.2954) loss 4.3397 (3.6342) grad_norm 1.2389 (1.3541) [2021-04-16 02:11:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][400/1251] eta 0:04:10 lr 0.000661 time 0.2584 (0.2949) loss 3.4244 (3.6262) grad_norm 1.2663 (1.3540) [2021-04-16 02:11:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][410/1251] eta 0:04:07 lr 0.000661 time 0.2875 (0.2946) loss 4.1268 (3.6230) grad_norm 1.3178 (1.3531) [2021-04-16 02:11:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][420/1251] eta 0:04:04 lr 0.000661 time 0.2779 (0.2941) loss 3.8655 (3.6210) grad_norm 1.2608 (1.3532) [2021-04-16 02:11:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][430/1251] eta 0:04:01 lr 0.000661 time 0.2744 (0.2937) loss 3.7344 (3.6157) grad_norm 1.3040 (1.3524) [2021-04-16 02:11:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][440/1251] eta 0:03:57 lr 0.000661 time 0.2973 (0.2933) loss 2.5495 (3.6113) grad_norm 1.3826 (1.3510) [2021-04-16 02:11:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][450/1251] eta 0:03:54 lr 0.000661 time 0.2700 (0.2929) loss 3.9246 (3.6082) grad_norm 1.6723 (1.3527) [2021-04-16 02:11:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][460/1251] eta 0:03:51 lr 0.000661 time 0.2853 (0.2926) loss 3.0066 (3.6066) grad_norm 1.1993 (1.3520) [2021-04-16 02:11:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][470/1251] eta 0:03:48 lr 0.000661 time 0.2899 (0.2923) loss 3.4884 (3.6025) grad_norm 1.2932 (1.3523) [2021-04-16 02:11:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][480/1251] eta 0:03:45 lr 0.000661 time 0.2983 (0.2921) loss 4.1523 (3.6069) grad_norm 1.2627 (1.3513) [2021-04-16 02:11:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][490/1251] eta 0:03:42 lr 0.000661 time 0.2824 (0.2918) loss 4.3508 (3.6140) grad_norm 1.9373 (1.3511) [2021-04-16 02:11:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][500/1251] eta 0:03:38 lr 0.000661 time 0.2848 (0.2916) loss 4.0057 (3.6211) grad_norm 1.3469 (1.3508) [2021-04-16 02:11:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][510/1251] eta 0:03:35 lr 0.000661 time 0.2658 (0.2913) loss 4.6594 (3.6278) grad_norm 1.3189 (1.3502) [2021-04-16 02:11:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][520/1251] eta 0:03:32 lr 0.000661 time 0.2775 (0.2911) loss 4.7055 (3.6302) grad_norm 1.4450 (1.3494) [2021-04-16 02:11:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][530/1251] eta 0:03:29 lr 0.000661 time 0.3108 (0.2909) loss 3.0852 (3.6309) grad_norm 1.3564 (1.3512) [2021-04-16 02:11:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][540/1251] eta 0:03:26 lr 0.000661 time 0.2716 (0.2908) loss 3.7988 (3.6263) grad_norm 1.3037 (1.3523) [2021-04-16 02:11:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][550/1251] eta 0:03:23 lr 0.000661 time 0.2833 (0.2906) loss 4.2484 (3.6335) grad_norm 1.3185 (1.3533) [2021-04-16 02:11:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][560/1251] eta 0:03:20 lr 0.000661 time 0.2772 (0.2903) loss 2.4231 (3.6301) grad_norm 1.5346 (1.3551) [2021-04-16 02:11:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][570/1251] eta 0:03:17 lr 0.000661 time 0.2888 (0.2902) loss 4.2612 (3.6306) grad_norm 1.3637 (1.3558) [2021-04-16 02:12:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][580/1251] eta 0:03:14 lr 0.000661 time 0.2905 (0.2906) loss 3.5150 (3.6274) grad_norm 1.4769 (1.3554) [2021-04-16 02:12:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][590/1251] eta 0:03:11 lr 0.000661 time 0.2817 (0.2904) loss 3.7862 (3.6246) grad_norm 1.2324 (1.3547) [2021-04-16 02:12:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][600/1251] eta 0:03:08 lr 0.000661 time 0.2820 (0.2902) loss 3.9813 (3.6236) grad_norm 1.4402 (1.3557) [2021-04-16 02:12:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][610/1251] eta 0:03:05 lr 0.000660 time 0.2664 (0.2900) loss 3.6473 (3.6293) grad_norm 1.4421 (1.3571) [2021-04-16 02:12:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][620/1251] eta 0:03:02 lr 0.000660 time 0.2672 (0.2899) loss 3.9923 (3.6337) grad_norm 1.2774 (1.3572) [2021-04-16 02:12:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][630/1251] eta 0:02:59 lr 0.000660 time 0.2744 (0.2897) loss 2.5263 (3.6317) grad_norm 1.5436 (1.3574) [2021-04-16 02:12:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][640/1251] eta 0:02:56 lr 0.000660 time 0.2952 (0.2896) loss 3.6081 (3.6274) grad_norm 1.3064 (1.3576) [2021-04-16 02:12:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][650/1251] eta 0:02:54 lr 0.000660 time 0.2614 (0.2896) loss 2.5827 (3.6233) grad_norm 1.4296 (1.3564) [2021-04-16 02:12:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][660/1251] eta 0:02:51 lr 0.000660 time 0.3089 (0.2894) loss 4.3366 (3.6185) grad_norm 1.4003 (1.3564) [2021-04-16 02:12:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][670/1251] eta 0:02:48 lr 0.000660 time 0.2595 (0.2893) loss 3.0338 (3.6193) grad_norm 1.2816 (1.3562) [2021-04-16 02:12:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][680/1251] eta 0:02:45 lr 0.000660 time 0.2777 (0.2891) loss 3.4327 (3.6219) grad_norm 1.2865 (1.3575) [2021-04-16 02:12:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][690/1251] eta 0:02:42 lr 0.000660 time 0.2701 (0.2891) loss 3.4901 (3.6211) grad_norm 1.3265 (1.3569) [2021-04-16 02:12:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][700/1251] eta 0:02:39 lr 0.000660 time 0.2711 (0.2890) loss 3.7107 (3.6175) grad_norm 1.2762 (1.3571) [2021-04-16 02:12:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][710/1251] eta 0:02:36 lr 0.000660 time 0.2429 (0.2890) loss 3.4454 (3.6179) grad_norm 1.4973 (1.3581) [2021-04-16 02:12:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][720/1251] eta 0:02:33 lr 0.000660 time 0.2690 (0.2889) loss 3.3623 (3.6151) grad_norm 1.2742 (1.3579) [2021-04-16 02:12:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][730/1251] eta 0:02:30 lr 0.000660 time 0.3107 (0.2887) loss 3.6196 (3.6198) grad_norm 1.3121 (1.3594) [2021-04-16 02:12:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][740/1251] eta 0:02:27 lr 0.000660 time 0.2666 (0.2886) loss 2.7972 (3.6184) grad_norm 1.2648 (1.3597) [2021-04-16 02:12:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][750/1251] eta 0:02:24 lr 0.000660 time 0.2831 (0.2885) loss 3.3845 (3.6224) grad_norm 1.2134 (1.3590) [2021-04-16 02:12:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][760/1251] eta 0:02:21 lr 0.000660 time 0.2839 (0.2883) loss 2.5070 (3.6201) grad_norm 1.3213 (1.3604) [2021-04-16 02:12:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][770/1251] eta 0:02:18 lr 0.000660 time 0.2516 (0.2881) loss 3.5830 (3.6192) grad_norm 1.2254 (1.3599) [2021-04-16 02:12:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][780/1251] eta 0:02:15 lr 0.000660 time 0.3001 (0.2880) loss 4.1640 (3.6211) grad_norm 1.2817 (1.3589) [2021-04-16 02:12:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][790/1251] eta 0:02:12 lr 0.000660 time 0.2673 (0.2877) loss 3.8637 (3.6218) grad_norm 1.3398 (1.3584) [2021-04-16 02:13:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][800/1251] eta 0:02:09 lr 0.000660 time 0.2637 (0.2877) loss 4.3729 (3.6246) grad_norm 1.3864 (1.3591) [2021-04-16 02:13:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][810/1251] eta 0:02:06 lr 0.000660 time 0.2938 (0.2876) loss 3.9636 (3.6237) grad_norm 1.5313 (1.3595) [2021-04-16 02:13:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][820/1251] eta 0:02:03 lr 0.000660 time 0.2931 (0.2875) loss 4.1004 (3.6259) grad_norm 1.3337 (1.3603) [2021-04-16 02:13:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][830/1251] eta 0:02:00 lr 0.000660 time 0.2569 (0.2873) loss 4.4986 (3.6257) grad_norm 1.5418 (1.3604) [2021-04-16 02:13:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][840/1251] eta 0:01:58 lr 0.000660 time 0.2575 (0.2872) loss 4.2858 (3.6267) grad_norm 1.2720 (1.3608) [2021-04-16 02:13:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][850/1251] eta 0:01:55 lr 0.000660 time 0.2678 (0.2871) loss 3.7364 (3.6255) grad_norm 1.2863 (1.3604) [2021-04-16 02:13:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][860/1251] eta 0:01:52 lr 0.000660 time 0.2892 (0.2870) loss 3.4074 (3.6233) grad_norm 1.4120 (1.3603) [2021-04-16 02:13:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][870/1251] eta 0:01:49 lr 0.000659 time 0.2655 (0.2870) loss 3.5951 (3.6222) grad_norm 1.2704 (1.3600) [2021-04-16 02:13:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][880/1251] eta 0:01:46 lr 0.000659 time 0.2977 (0.2870) loss 4.1151 (3.6241) grad_norm 1.2741 (1.3600) [2021-04-16 02:13:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][890/1251] eta 0:01:43 lr 0.000659 time 0.2884 (0.2870) loss 4.2270 (3.6224) grad_norm 1.1660 (1.3599) [2021-04-16 02:13:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][900/1251] eta 0:01:40 lr 0.000659 time 0.4035 (0.2871) loss 3.7577 (3.6228) grad_norm 1.2192 (1.3597) [2021-04-16 02:13:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][910/1251] eta 0:01:37 lr 0.000659 time 0.2887 (0.2870) loss 4.3945 (3.6241) grad_norm 1.1639 (1.3594) [2021-04-16 02:13:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][920/1251] eta 0:01:34 lr 0.000659 time 0.2750 (0.2868) loss 2.6399 (3.6219) grad_norm 1.5956 (1.3589) [2021-04-16 02:13:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][930/1251] eta 0:01:32 lr 0.000659 time 0.2709 (0.2867) loss 3.2074 (3.6193) grad_norm 1.1671 (1.3587) [2021-04-16 02:13:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][940/1251] eta 0:01:29 lr 0.000659 time 0.2642 (0.2866) loss 2.8219 (3.6184) grad_norm 1.1465 (1.3577) [2021-04-16 02:13:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][950/1251] eta 0:01:26 lr 0.000659 time 0.2701 (0.2866) loss 2.6361 (3.6183) grad_norm 1.4069 (1.3572) [2021-04-16 02:13:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][960/1251] eta 0:01:23 lr 0.000659 time 0.2518 (0.2866) loss 4.1180 (3.6187) grad_norm 1.4090 (1.3573) [2021-04-16 02:13:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][970/1251] eta 0:01:20 lr 0.000659 time 0.2967 (0.2865) loss 2.9876 (3.6171) grad_norm 1.4911 (1.3583) [2021-04-16 02:13:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][980/1251] eta 0:01:17 lr 0.000659 time 0.2725 (0.2865) loss 3.3286 (3.6171) grad_norm 1.4836 (1.3600) [2021-04-16 02:13:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][990/1251] eta 0:01:14 lr 0.000659 time 0.2869 (0.2864) loss 4.6830 (3.6183) grad_norm 1.2561 (1.3599) [2021-04-16 02:13:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][1000/1251] eta 0:01:11 lr 0.000659 time 0.2751 (0.2863) loss 3.4682 (3.6188) grad_norm 1.4483 (1.3604) [2021-04-16 02:14:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][1010/1251] eta 0:01:08 lr 0.000659 time 0.2645 (0.2862) loss 3.3318 (3.6198) grad_norm 1.5249 (1.3606) [2021-04-16 02:14:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][1020/1251] eta 0:01:06 lr 0.000659 time 0.2861 (0.2861) loss 4.4556 (3.6182) grad_norm 1.2625 (1.3614) [2021-04-16 02:14:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][1030/1251] eta 0:01:03 lr 0.000659 time 0.2789 (0.2860) loss 3.9356 (3.6208) grad_norm 1.4078 (1.3624) [2021-04-16 02:14:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][1040/1251] eta 0:01:00 lr 0.000659 time 0.2907 (0.2858) loss 4.0039 (3.6244) grad_norm 1.2388 (1.3618) [2021-04-16 02:14:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][1050/1251] eta 0:00:57 lr 0.000659 time 0.2508 (0.2858) loss 3.8530 (3.6224) grad_norm 1.2899 (1.3618) [2021-04-16 02:14:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][1060/1251] eta 0:00:54 lr 0.000659 time 0.2832 (0.2857) loss 3.6044 (3.6221) grad_norm 1.7745 (1.3622) [2021-04-16 02:14:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][1070/1251] eta 0:00:51 lr 0.000659 time 0.2896 (0.2856) loss 4.3561 (3.6235) grad_norm 1.2339 (1.3621) [2021-04-16 02:14:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][1080/1251] eta 0:00:48 lr 0.000659 time 0.2760 (0.2855) loss 3.6313 (3.6227) grad_norm 1.2202 (1.3620) [2021-04-16 02:14:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][1090/1251] eta 0:00:45 lr 0.000659 time 0.2664 (0.2855) loss 3.3041 (3.6218) grad_norm 1.3063 (1.3614) [2021-04-16 02:14:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][1100/1251] eta 0:00:43 lr 0.000659 time 0.2835 (0.2854) loss 4.5418 (3.6211) grad_norm 1.3387 (1.3608) [2021-04-16 02:14:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][1110/1251] eta 0:00:40 lr 0.000659 time 0.2801 (0.2854) loss 3.8162 (3.6190) grad_norm 1.2332 (1.3604) [2021-04-16 02:14:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][1120/1251] eta 0:00:37 lr 0.000658 time 0.2556 (0.2854) loss 3.9848 (3.6205) grad_norm 1.4469 (1.3592) [2021-04-16 02:14:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][1130/1251] eta 0:00:34 lr 0.000658 time 0.2945 (0.2854) loss 2.8407 (3.6209) grad_norm 1.4934 (1.3590) [2021-04-16 02:14:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][1140/1251] eta 0:00:31 lr 0.000658 time 0.3206 (0.2853) loss 3.4195 (3.6219) grad_norm 1.4532 (1.3594) [2021-04-16 02:14:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][1150/1251] eta 0:00:28 lr 0.000658 time 0.2633 (0.2854) loss 3.9673 (3.6251) grad_norm 1.1537 (1.3601) [2021-04-16 02:14:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][1160/1251] eta 0:00:25 lr 0.000658 time 0.2741 (0.2854) loss 4.6536 (3.6237) grad_norm 1.1937 (1.3597) [2021-04-16 02:14:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][1170/1251] eta 0:00:23 lr 0.000658 time 0.2677 (0.2854) loss 4.1044 (3.6256) grad_norm 1.3769 (1.3590) [2021-04-16 02:14:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][1180/1251] eta 0:00:20 lr 0.000658 time 0.2794 (0.2853) loss 2.8090 (3.6244) grad_norm 1.0761 (1.3585) [2021-04-16 02:14:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][1190/1251] eta 0:00:17 lr 0.000658 time 0.2668 (0.2853) loss 4.0448 (3.6265) grad_norm 1.2343 (1.3582) [2021-04-16 02:14:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][1200/1251] eta 0:00:14 lr 0.000658 time 0.3257 (0.2852) loss 2.3729 (3.6263) grad_norm 1.5344 (1.3584) [2021-04-16 02:14:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][1210/1251] eta 0:00:11 lr 0.000658 time 0.2947 (0.2851) loss 4.5197 (3.6286) grad_norm 1.2376 (1.3580) [2021-04-16 02:14:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][1220/1251] eta 0:00:08 lr 0.000658 time 0.2866 (0.2851) loss 3.9068 (3.6286) grad_norm 1.1701 (1.3584) [2021-04-16 02:15:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][1230/1251] eta 0:00:05 lr 0.000658 time 0.2918 (0.2851) loss 3.5284 (3.6278) grad_norm 1.2497 (1.3581) [2021-04-16 02:15:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][1240/1251] eta 0:00:03 lr 0.000658 time 0.2608 (0.2850) loss 4.6247 (3.6278) grad_norm 1.4697 (1.3587) [2021-04-16 02:15:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [119/300][1250/1251] eta 0:00:00 lr 0.000658 time 0.2588 (0.2847) loss 4.1272 (3.6272) grad_norm 1.2061 (1.3588) [2021-04-16 02:15:11 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 119 training takes 0:06:00 [2021-04-16 02:15:11 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_119.pth saving...... [2021-04-16 02:15:25 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_119.pth saved !!! [2021-04-16 02:15:26 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.138 (1.138) Loss 1.0435 (1.0435) Acc@1 75.684 (75.684) Acc@5 92.480 (92.480) [2021-04-16 02:15:27 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.120 (0.225) Loss 1.0338 (1.0635) Acc@1 76.074 (75.195) Acc@5 93.164 (92.836) [2021-04-16 02:15:29 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.175 (0.214) Loss 1.0638 (1.0654) Acc@1 74.316 (75.070) Acc@5 92.188 (92.653) [2021-04-16 02:15:32 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.107 (0.218) Loss 1.0978 (1.0645) Acc@1 74.121 (75.003) Acc@5 92.480 (92.613) [2021-04-16 02:15:34 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.221 (0.214) Loss 1.0710 (1.0668) Acc@1 75.684 (74.890) Acc@5 92.383 (92.626) [2021-04-16 02:15:38 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 74.732 Acc@5 92.626 [2021-04-16 02:15:38 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 74.7% [2021-04-16 02:15:38 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 74.73% [2021-04-16 02:15:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][0/1251] eta 1:34:35 lr 0.000658 time 4.5366 (4.5366) loss 3.8754 (3.8754) grad_norm 1.1947 (1.1947) [2021-04-16 02:15:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][10/1251] eta 0:13:41 lr 0.000658 time 0.2929 (0.6616) loss 4.0185 (3.8659) grad_norm 1.4940 (1.3905) [2021-04-16 02:15:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][20/1251] eta 0:09:49 lr 0.000658 time 0.2963 (0.4790) loss 3.2779 (3.7572) grad_norm 1.3468 (1.3433) [2021-04-16 02:15:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][30/1251] eta 0:08:24 lr 0.000658 time 0.2842 (0.4131) loss 3.4944 (3.7244) grad_norm 1.4289 (1.3628) [2021-04-16 02:15:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][40/1251] eta 0:07:43 lr 0.000658 time 0.2975 (0.3827) loss 3.9892 (3.7633) grad_norm 1.2943 (1.3801) [2021-04-16 02:15:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][50/1251] eta 0:07:15 lr 0.000658 time 0.2555 (0.3622) loss 4.2664 (3.6737) grad_norm 1.6216 (1.3749) [2021-04-16 02:15:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][60/1251] eta 0:06:55 lr 0.000658 time 0.2751 (0.3485) loss 3.9223 (3.6130) grad_norm 1.2720 (1.3557) [2021-04-16 02:16:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][70/1251] eta 0:06:39 lr 0.000658 time 0.2820 (0.3379) loss 4.4348 (3.6274) grad_norm 1.2804 (1.3581) [2021-04-16 02:16:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][80/1251] eta 0:06:28 lr 0.000658 time 0.2802 (0.3322) loss 3.4239 (3.5859) grad_norm 1.1621 (1.3563) [2021-04-16 02:16:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][90/1251] eta 0:06:19 lr 0.000658 time 0.2708 (0.3273) loss 3.1870 (3.5604) grad_norm 1.3044 (1.3645) [2021-04-16 02:16:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][100/1251] eta 0:06:13 lr 0.000658 time 0.2814 (0.3245) loss 4.2028 (3.5650) grad_norm 1.3277 (1.3669) [2021-04-16 02:16:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][110/1251] eta 0:06:05 lr 0.000658 time 0.2709 (0.3208) loss 3.0742 (3.5438) grad_norm 1.3991 (1.3604) [2021-04-16 02:16:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][120/1251] eta 0:05:59 lr 0.000657 time 0.2746 (0.3174) loss 4.0594 (3.5430) grad_norm 1.2292 (1.3617) [2021-04-16 02:16:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][130/1251] eta 0:05:52 lr 0.000657 time 0.2877 (0.3147) loss 2.7144 (3.5629) grad_norm 1.3457 (1.3634) [2021-04-16 02:16:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][140/1251] eta 0:05:46 lr 0.000657 time 0.2783 (0.3117) loss 3.9482 (3.5605) grad_norm 1.3433 (1.3602) [2021-04-16 02:16:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][150/1251] eta 0:05:40 lr 0.000657 time 0.2928 (0.3096) loss 4.3139 (3.5741) grad_norm 1.1702 (1.3620) [2021-04-16 02:16:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][160/1251] eta 0:05:35 lr 0.000657 time 0.2454 (0.3077) loss 3.6616 (3.5673) grad_norm 1.2477 (1.3577) [2021-04-16 02:16:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][170/1251] eta 0:05:31 lr 0.000657 time 0.2904 (0.3066) loss 3.6632 (3.5796) grad_norm 1.7847 (1.3604) [2021-04-16 02:16:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][180/1251] eta 0:05:26 lr 0.000657 time 0.2740 (0.3052) loss 4.3806 (3.5831) grad_norm 1.2952 (1.3597) [2021-04-16 02:16:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][190/1251] eta 0:05:22 lr 0.000657 time 0.2462 (0.3037) loss 3.2682 (3.5731) grad_norm 1.2873 (1.3575) [2021-04-16 02:16:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][200/1251] eta 0:05:17 lr 0.000657 time 0.2865 (0.3024) loss 3.7725 (3.5663) grad_norm 1.5805 (1.3551) [2021-04-16 02:16:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][210/1251] eta 0:05:13 lr 0.000657 time 0.2629 (0.3014) loss 3.5894 (3.5839) grad_norm 1.1128 (1.3512) [2021-04-16 02:16:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][220/1251] eta 0:05:09 lr 0.000657 time 0.2545 (0.3002) loss 3.9290 (3.5809) grad_norm 1.3384 (1.3491) [2021-04-16 02:16:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][230/1251] eta 0:05:05 lr 0.000657 time 0.2734 (0.2992) loss 3.7525 (3.5733) grad_norm 1.2508 (1.3470) [2021-04-16 02:16:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][240/1251] eta 0:05:01 lr 0.000657 time 0.2933 (0.2983) loss 4.3761 (3.5889) grad_norm 1.5032 (1.3471) [2021-04-16 02:16:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][250/1251] eta 0:04:57 lr 0.000657 time 0.2678 (0.2973) loss 3.6078 (3.5863) grad_norm 1.5523 (1.3497) [2021-04-16 02:16:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][260/1251] eta 0:04:53 lr 0.000657 time 0.2897 (0.2965) loss 3.7176 (3.5735) grad_norm 1.3409 (1.3517) [2021-04-16 02:16:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][270/1251] eta 0:04:50 lr 0.000657 time 0.3085 (0.2957) loss 3.4930 (3.5808) grad_norm 1.3143 (1.3489) [2021-04-16 02:17:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][280/1251] eta 0:04:46 lr 0.000657 time 0.2744 (0.2950) loss 3.0928 (3.5802) grad_norm 1.3525 (1.3498) [2021-04-16 02:17:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][290/1251] eta 0:04:42 lr 0.000657 time 0.2738 (0.2943) loss 2.7484 (3.5891) grad_norm 1.2599 (1.3501) [2021-04-16 02:17:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][300/1251] eta 0:04:39 lr 0.000657 time 0.2763 (0.2936) loss 3.5827 (3.5970) grad_norm 1.3167 (1.3506) [2021-04-16 02:17:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][310/1251] eta 0:04:35 lr 0.000657 time 0.2963 (0.2931) loss 2.8645 (3.5984) grad_norm 1.3636 (1.3484) [2021-04-16 02:17:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][320/1251] eta 0:04:32 lr 0.000657 time 0.2638 (0.2931) loss 3.8922 (3.5962) grad_norm 1.5402 (1.3491) [2021-04-16 02:17:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][330/1251] eta 0:04:29 lr 0.000657 time 0.2687 (0.2926) loss 4.2166 (3.5987) grad_norm 1.3634 (1.3478) [2021-04-16 02:17:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][340/1251] eta 0:04:26 lr 0.000657 time 0.2848 (0.2922) loss 2.6366 (3.5994) grad_norm 1.3941 (1.3496) [2021-04-16 02:17:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][350/1251] eta 0:04:23 lr 0.000657 time 0.4522 (0.2929) loss 3.5777 (3.5879) grad_norm 1.3153 (1.3496) [2021-04-16 02:17:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][360/1251] eta 0:04:20 lr 0.000657 time 0.2501 (0.2927) loss 4.4324 (3.5862) grad_norm 1.2137 (1.3475) [2021-04-16 02:17:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][370/1251] eta 0:04:17 lr 0.000657 time 0.3209 (0.2923) loss 4.3261 (3.5870) grad_norm 1.3585 (1.3479) [2021-04-16 02:17:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][380/1251] eta 0:04:14 lr 0.000656 time 0.2628 (0.2918) loss 4.1720 (3.5893) grad_norm 1.3229 (1.3474) [2021-04-16 02:17:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][390/1251] eta 0:04:10 lr 0.000656 time 0.2890 (0.2914) loss 3.4427 (3.6004) grad_norm 1.5439 (1.3478) [2021-04-16 02:17:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][400/1251] eta 0:04:07 lr 0.000656 time 0.2848 (0.2910) loss 3.3908 (3.5951) grad_norm 1.2181 (1.3476) [2021-04-16 02:17:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][410/1251] eta 0:04:04 lr 0.000656 time 0.2576 (0.2906) loss 4.2649 (3.6034) grad_norm 1.4529 (1.3478) [2021-04-16 02:17:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][420/1251] eta 0:04:01 lr 0.000656 time 0.3196 (0.2904) loss 3.4202 (3.6016) grad_norm 1.3541 (1.3488) [2021-04-16 02:17:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][430/1251] eta 0:03:58 lr 0.000656 time 0.2645 (0.2903) loss 2.3745 (3.5965) grad_norm 1.3544 (1.3481) [2021-04-16 02:17:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][440/1251] eta 0:03:55 lr 0.000656 time 0.2579 (0.2900) loss 4.5587 (3.6040) grad_norm 1.4465 (1.3476) [2021-04-16 02:17:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][450/1251] eta 0:03:52 lr 0.000656 time 0.2636 (0.2898) loss 4.1417 (3.6013) grad_norm 1.3547 (1.3491) [2021-04-16 02:17:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][460/1251] eta 0:03:48 lr 0.000656 time 0.2428 (0.2895) loss 3.8037 (3.6006) grad_norm 1.5827 (1.3514) [2021-04-16 02:17:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][470/1251] eta 0:03:45 lr 0.000656 time 0.2896 (0.2891) loss 4.3476 (3.6028) grad_norm 1.3726 (1.3530) [2021-04-16 02:17:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][480/1251] eta 0:03:42 lr 0.000656 time 0.2822 (0.2889) loss 3.9452 (3.6082) grad_norm 1.3141 (1.3533) [2021-04-16 02:18:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][490/1251] eta 0:03:39 lr 0.000656 time 0.2733 (0.2890) loss 3.1849 (3.6049) grad_norm 1.2779 (1.3543) [2021-04-16 02:18:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][500/1251] eta 0:03:36 lr 0.000656 time 0.2845 (0.2887) loss 3.2926 (3.6056) grad_norm 1.1785 (1.3536) [2021-04-16 02:18:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][510/1251] eta 0:03:33 lr 0.000656 time 0.3067 (0.2885) loss 2.9955 (3.6026) grad_norm 1.4256 (1.3556) [2021-04-16 02:18:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][520/1251] eta 0:03:30 lr 0.000656 time 0.2874 (0.2883) loss 3.8775 (3.6056) grad_norm 1.2542 (1.3560) [2021-04-16 02:18:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][530/1251] eta 0:03:27 lr 0.000656 time 0.2926 (0.2881) loss 2.5244 (3.6060) grad_norm 1.4137 (1.3567) [2021-04-16 02:18:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][540/1251] eta 0:03:24 lr 0.000656 time 0.2639 (0.2878) loss 2.2191 (3.6043) grad_norm 1.4256 (1.3577) [2021-04-16 02:18:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][550/1251] eta 0:03:21 lr 0.000656 time 0.2787 (0.2876) loss 3.8144 (3.6036) grad_norm 1.1651 (1.3569) [2021-04-16 02:18:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][560/1251] eta 0:03:18 lr 0.000656 time 0.3006 (0.2874) loss 4.4214 (3.6031) grad_norm 1.6404 (1.3570) [2021-04-16 02:18:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][570/1251] eta 0:03:15 lr 0.000656 time 0.2699 (0.2875) loss 3.1082 (3.6029) grad_norm 1.2620 (1.3574) [2021-04-16 02:18:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][580/1251] eta 0:03:12 lr 0.000656 time 0.2748 (0.2873) loss 4.2135 (3.6032) grad_norm 1.2174 (1.3574) [2021-04-16 02:18:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][590/1251] eta 0:03:09 lr 0.000656 time 0.2777 (0.2872) loss 3.9778 (3.6007) grad_norm 1.1436 (1.3588) [2021-04-16 02:18:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][600/1251] eta 0:03:06 lr 0.000656 time 0.2925 (0.2871) loss 3.4205 (3.5982) grad_norm 1.3671 (1.3598) [2021-04-16 02:18:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][610/1251] eta 0:03:03 lr 0.000656 time 0.2744 (0.2869) loss 3.2973 (3.5985) grad_norm 1.1399 (1.3590) [2021-04-16 02:18:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][620/1251] eta 0:03:00 lr 0.000656 time 0.2516 (0.2866) loss 4.0548 (3.5992) grad_norm 1.3378 (1.3595) [2021-04-16 02:18:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][630/1251] eta 0:02:57 lr 0.000655 time 0.3002 (0.2865) loss 4.1837 (3.6004) grad_norm 1.3017 (1.3608) [2021-04-16 02:18:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][640/1251] eta 0:02:54 lr 0.000655 time 0.2877 (0.2864) loss 3.1672 (3.5999) grad_norm 1.2180 (1.3607) [2021-04-16 02:18:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][650/1251] eta 0:02:52 lr 0.000655 time 0.2819 (0.2864) loss 4.8036 (3.6005) grad_norm 1.1719 (1.3600) [2021-04-16 02:18:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][660/1251] eta 0:02:49 lr 0.000655 time 0.2903 (0.2863) loss 3.9555 (3.6031) grad_norm 1.8506 (1.3611) [2021-04-16 02:18:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][670/1251] eta 0:02:46 lr 0.000655 time 0.2776 (0.2864) loss 4.2178 (3.6041) grad_norm 1.1163 (1.3613) [2021-04-16 02:18:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][680/1251] eta 0:02:43 lr 0.000655 time 0.2792 (0.2863) loss 3.5363 (3.6054) grad_norm 1.3741 (1.3617) [2021-04-16 02:18:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][690/1251] eta 0:02:40 lr 0.000655 time 0.2780 (0.2862) loss 3.9828 (3.6062) grad_norm 1.2718 (1.3615) [2021-04-16 02:18:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][700/1251] eta 0:02:37 lr 0.000655 time 0.2985 (0.2861) loss 4.1095 (3.6056) grad_norm 1.3034 (1.3609) [2021-04-16 02:19:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][710/1251] eta 0:02:34 lr 0.000655 time 0.2826 (0.2860) loss 3.6792 (3.6062) grad_norm 1.1754 (1.3609) [2021-04-16 02:19:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][720/1251] eta 0:02:31 lr 0.000655 time 0.2894 (0.2859) loss 4.3795 (3.6108) grad_norm 1.3176 (1.3610) [2021-04-16 02:19:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][730/1251] eta 0:02:28 lr 0.000655 time 0.2760 (0.2857) loss 3.0054 (3.6142) grad_norm 1.3191 (1.3611) [2021-04-16 02:19:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][740/1251] eta 0:02:26 lr 0.000655 time 0.4187 (0.2858) loss 3.6560 (3.6134) grad_norm 1.2537 (1.3618) [2021-04-16 02:19:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][750/1251] eta 0:02:23 lr 0.000655 time 0.2825 (0.2857) loss 4.3936 (3.6172) grad_norm 1.3356 (1.3615) [2021-04-16 02:19:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][760/1251] eta 0:02:20 lr 0.000655 time 0.2642 (0.2856) loss 4.0418 (3.6161) grad_norm 1.2737 (1.3605) [2021-04-16 02:19:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][770/1251] eta 0:02:17 lr 0.000655 time 0.2849 (0.2855) loss 2.6770 (3.6151) grad_norm 1.3949 (1.3612) [2021-04-16 02:19:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][780/1251] eta 0:02:14 lr 0.000655 time 0.2907 (0.2855) loss 3.5120 (3.6118) grad_norm 1.2751 (1.3616) [2021-04-16 02:19:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][790/1251] eta 0:02:11 lr 0.000655 time 0.2792 (0.2854) loss 3.1160 (3.6108) grad_norm 1.4289 (1.3612) [2021-04-16 02:19:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][800/1251] eta 0:02:08 lr 0.000655 time 0.2861 (0.2853) loss 3.7201 (3.6147) grad_norm 1.3417 (1.3622) [2021-04-16 02:19:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][810/1251] eta 0:02:05 lr 0.000655 time 0.2971 (0.2851) loss 2.5183 (3.6168) grad_norm 1.2582 (1.3619) [2021-04-16 02:19:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][820/1251] eta 0:02:02 lr 0.000655 time 0.2678 (0.2851) loss 4.0673 (3.6200) grad_norm 1.2811 (1.3618) [2021-04-16 02:19:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][830/1251] eta 0:02:00 lr 0.000655 time 0.3118 (0.2851) loss 4.5090 (3.6208) grad_norm 1.3133 (1.3614) [2021-04-16 02:19:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][840/1251] eta 0:01:57 lr 0.000655 time 0.2638 (0.2850) loss 3.6194 (3.6216) grad_norm 1.3157 (1.3613) [2021-04-16 02:19:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][850/1251] eta 0:01:54 lr 0.000655 time 0.2756 (0.2849) loss 2.5041 (3.6199) grad_norm 1.4181 (1.3614) [2021-04-16 02:19:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][860/1251] eta 0:01:51 lr 0.000655 time 0.2810 (0.2848) loss 4.5637 (3.6216) grad_norm 1.5326 (1.3609) [2021-04-16 02:19:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][870/1251] eta 0:01:48 lr 0.000655 time 0.2747 (0.2848) loss 3.6038 (3.6194) grad_norm 1.6716 (1.3633) [2021-04-16 02:19:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][880/1251] eta 0:01:45 lr 0.000654 time 0.2812 (0.2846) loss 3.3757 (3.6208) grad_norm 1.3188 (1.3628) [2021-04-16 02:19:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][890/1251] eta 0:01:42 lr 0.000654 time 0.2718 (0.2848) loss 4.2833 (3.6240) grad_norm 1.6187 (1.3623) [2021-04-16 02:19:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][900/1251] eta 0:01:39 lr 0.000654 time 0.2872 (0.2847) loss 3.1288 (3.6198) grad_norm 1.3339 (1.3620) [2021-04-16 02:19:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][910/1251] eta 0:01:37 lr 0.000654 time 0.2860 (0.2846) loss 4.0702 (3.6220) grad_norm 1.3376 (1.3620) [2021-04-16 02:20:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][920/1251] eta 0:01:34 lr 0.000654 time 0.2889 (0.2845) loss 3.2510 (3.6228) grad_norm 1.5605 (1.3618) [2021-04-16 02:20:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][930/1251] eta 0:01:31 lr 0.000654 time 0.2754 (0.2848) loss 3.1193 (3.6209) grad_norm 1.1806 (1.3621) [2021-04-16 02:20:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][940/1251] eta 0:01:28 lr 0.000654 time 0.2816 (0.2848) loss 4.0549 (3.6174) grad_norm 1.5216 (1.3617) [2021-04-16 02:20:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][950/1251] eta 0:01:25 lr 0.000654 time 0.2548 (0.2847) loss 3.5936 (3.6209) grad_norm 1.3971 (1.3618) [2021-04-16 02:20:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][960/1251] eta 0:01:22 lr 0.000654 time 0.2784 (0.2846) loss 3.6218 (3.6212) grad_norm 1.3967 (1.3610) [2021-04-16 02:20:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][970/1251] eta 0:01:19 lr 0.000654 time 0.2818 (0.2846) loss 3.3419 (3.6177) grad_norm 1.2944 (1.3600) [2021-04-16 02:20:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][980/1251] eta 0:01:17 lr 0.000654 time 0.3105 (0.2846) loss 3.5014 (3.6184) grad_norm 1.4121 (1.3598) [2021-04-16 02:20:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][990/1251] eta 0:01:14 lr 0.000654 time 0.2455 (0.2845) loss 4.5763 (3.6192) grad_norm 1.4917 (1.3596) [2021-04-16 02:20:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][1000/1251] eta 0:01:11 lr 0.000654 time 0.2733 (0.2845) loss 3.6227 (3.6185) grad_norm 1.3218 (1.3593) [2021-04-16 02:20:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][1010/1251] eta 0:01:08 lr 0.000654 time 0.2487 (0.2844) loss 3.6602 (3.6172) grad_norm 1.4855 (1.3591) [2021-04-16 02:20:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][1020/1251] eta 0:01:05 lr 0.000654 time 0.2820 (0.2844) loss 4.1993 (3.6180) grad_norm 1.3461 (1.3589) [2021-04-16 02:20:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][1030/1251] eta 0:01:02 lr 0.000654 time 0.2491 (0.2843) loss 3.6649 (3.6186) grad_norm 1.2707 (1.3585) [2021-04-16 02:20:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][1040/1251] eta 0:00:59 lr 0.000654 time 0.2818 (0.2842) loss 4.2115 (3.6191) grad_norm 1.4401 (1.3586) [2021-04-16 02:20:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][1050/1251] eta 0:00:57 lr 0.000654 time 0.2960 (0.2842) loss 3.7828 (3.6219) grad_norm 1.4995 (1.3592) [2021-04-16 02:20:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][1060/1251] eta 0:00:54 lr 0.000654 time 0.2850 (0.2841) loss 4.3264 (3.6213) grad_norm 1.3056 (1.3591) [2021-04-16 02:20:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][1070/1251] eta 0:00:51 lr 0.000654 time 0.2572 (0.2841) loss 3.2891 (3.6235) grad_norm 1.5289 (1.3585) [2021-04-16 02:20:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][1080/1251] eta 0:00:48 lr 0.000654 time 0.2721 (0.2841) loss 4.3702 (3.6241) grad_norm 1.2186 (1.3577) [2021-04-16 02:20:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][1090/1251] eta 0:00:45 lr 0.000654 time 0.2675 (0.2840) loss 4.1773 (3.6259) grad_norm 1.3481 (1.3596) [2021-04-16 02:20:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][1100/1251] eta 0:00:42 lr 0.000654 time 0.2761 (0.2840) loss 2.2185 (3.6242) grad_norm 1.3024 (1.3608) [2021-04-16 02:20:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][1110/1251] eta 0:00:40 lr 0.000654 time 0.2743 (0.2839) loss 2.7441 (3.6260) grad_norm 1.2983 (1.3612) [2021-04-16 02:20:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][1120/1251] eta 0:00:37 lr 0.000654 time 0.2792 (0.2839) loss 4.5768 (3.6261) grad_norm 1.7371 (1.3618) [2021-04-16 02:20:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][1130/1251] eta 0:00:34 lr 0.000654 time 0.3008 (0.2839) loss 4.3558 (3.6262) grad_norm 1.5312 (1.3630) [2021-04-16 02:21:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][1140/1251] eta 0:00:31 lr 0.000653 time 0.2519 (0.2838) loss 3.9256 (3.6285) grad_norm 1.2806 (1.3628) [2021-04-16 02:21:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][1150/1251] eta 0:00:28 lr 0.000653 time 0.2546 (0.2838) loss 4.0035 (3.6309) grad_norm 1.5220 (1.3634) [2021-04-16 02:21:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][1160/1251] eta 0:00:25 lr 0.000653 time 0.2537 (0.2838) loss 3.5836 (3.6310) grad_norm 1.2730 (1.3637) [2021-04-16 02:21:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][1170/1251] eta 0:00:22 lr 0.000653 time 0.2940 (0.2838) loss 4.1332 (3.6327) grad_norm 1.1636 (1.3638) [2021-04-16 02:21:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][1180/1251] eta 0:00:20 lr 0.000653 time 0.3130 (0.2839) loss 4.4519 (3.6331) grad_norm 1.2624 (1.3637) [2021-04-16 02:21:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][1190/1251] eta 0:00:17 lr 0.000653 time 0.2772 (0.2838) loss 3.7494 (3.6336) grad_norm 1.8996 (1.3644) [2021-04-16 02:21:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][1200/1251] eta 0:00:14 lr 0.000653 time 0.3302 (0.2838) loss 4.0498 (3.6343) grad_norm 1.1593 (1.3641) [2021-04-16 02:21:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][1210/1251] eta 0:00:11 lr 0.000653 time 0.2472 (0.2837) loss 3.5186 (3.6379) grad_norm 1.2345 (1.3643) [2021-04-16 02:21:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][1220/1251] eta 0:00:08 lr 0.000653 time 0.2627 (0.2836) loss 4.4737 (3.6370) grad_norm 1.6235 (1.3642) [2021-04-16 02:21:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][1230/1251] eta 0:00:05 lr 0.000653 time 0.2658 (0.2836) loss 3.6239 (3.6385) grad_norm 1.5491 (1.3637) [2021-04-16 02:21:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][1240/1251] eta 0:00:03 lr 0.000653 time 0.2485 (0.2835) loss 3.3636 (3.6384) grad_norm 1.2477 (nan) [2021-04-16 02:21:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [120/300][1250/1251] eta 0:00:00 lr 0.000653 time 0.2483 (0.2832) loss 3.2879 (3.6362) grad_norm 1.4785 (nan) [2021-04-16 02:21:34 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 120 training takes 0:05:56 [2021-04-16 02:21:34 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_120.pth saving...... [2021-04-16 02:21:45 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_120.pth saved !!! [2021-04-16 02:21:46 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.055 (1.055) Loss 1.1111 (1.1111) Acc@1 73.340 (73.340) Acc@5 92.871 (92.871) [2021-04-16 02:21:48 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.137 (0.223) Loss 1.0566 (1.0877) Acc@1 75.488 (74.938) Acc@5 92.383 (92.658) [2021-04-16 02:21:50 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.102 (0.249) Loss 1.1684 (1.0976) Acc@1 74.609 (74.633) Acc@5 91.602 (92.560) [2021-04-16 02:21:52 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.066 (0.223) Loss 1.0844 (1.0989) Acc@1 74.609 (74.603) Acc@5 92.285 (92.518) [2021-04-16 02:21:54 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.213) Loss 1.0546 (1.0995) Acc@1 77.148 (74.552) Acc@5 92.090 (92.526) [2021-04-16 02:22:01 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 74.582 Acc@5 92.534 [2021-04-16 02:22:01 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 74.6% [2021-04-16 02:22:01 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 74.73% [2021-04-16 02:22:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][0/1251] eta 0:57:37 lr 0.000653 time 2.7639 (2.7639) loss 3.3353 (3.3353) grad_norm 1.4771 (1.4771) [2021-04-16 02:22:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][10/1251] eta 0:10:25 lr 0.000653 time 0.2580 (0.5043) loss 3.4390 (3.5383) grad_norm 1.4596 (1.3765) [2021-04-16 02:22:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][20/1251] eta 0:08:06 lr 0.000653 time 0.2751 (0.3952) loss 3.6209 (3.5337) grad_norm 1.2259 (1.3410) [2021-04-16 02:22:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][30/1251] eta 0:07:15 lr 0.000653 time 0.2766 (0.3569) loss 3.5370 (3.5318) grad_norm 1.3152 (1.3214) [2021-04-16 02:22:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][40/1251] eta 0:06:49 lr 0.000653 time 0.3096 (0.3378) loss 4.2490 (3.5763) grad_norm 1.4540 (1.3445) [2021-04-16 02:22:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][50/1251] eta 0:06:30 lr 0.000653 time 0.2827 (0.3248) loss 4.3670 (3.5976) grad_norm 1.3749 (1.3583) [2021-04-16 02:22:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][60/1251] eta 0:06:16 lr 0.000653 time 0.2580 (0.3165) loss 3.9007 (3.6328) grad_norm 1.2942 (1.3520) [2021-04-16 02:22:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][70/1251] eta 0:06:07 lr 0.000653 time 0.2638 (0.3113) loss 3.1039 (3.6287) grad_norm 1.3214 (1.3559) [2021-04-16 02:22:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][80/1251] eta 0:05:59 lr 0.000653 time 0.2691 (0.3069) loss 3.1974 (3.5662) grad_norm 1.4386 (1.3535) [2021-04-16 02:22:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][90/1251] eta 0:05:54 lr 0.000653 time 0.2791 (0.3055) loss 4.0020 (3.5493) grad_norm 1.3470 (1.3460) [2021-04-16 02:22:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][100/1251] eta 0:05:50 lr 0.000653 time 0.2699 (0.3046) loss 3.7016 (3.5625) grad_norm 1.2102 (1.3390) [2021-04-16 02:22:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][110/1251] eta 0:05:44 lr 0.000653 time 0.3003 (0.3019) loss 3.6637 (3.5468) grad_norm 1.2740 (1.3328) [2021-04-16 02:22:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][120/1251] eta 0:05:38 lr 0.000653 time 0.2803 (0.2994) loss 3.0984 (3.5536) grad_norm 1.6319 (1.3323) [2021-04-16 02:22:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][130/1251] eta 0:05:36 lr 0.000653 time 0.2673 (0.2999) loss 3.8871 (3.5485) grad_norm 1.3689 (1.3446) [2021-04-16 02:22:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][140/1251] eta 0:05:32 lr 0.000652 time 0.2879 (0.2993) loss 4.0377 (3.5321) grad_norm 1.3628 (1.3577) [2021-04-16 02:22:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][150/1251] eta 0:05:27 lr 0.000652 time 0.2532 (0.2978) loss 2.9993 (3.5043) grad_norm 1.2458 (1.3528) [2021-04-16 02:22:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][160/1251] eta 0:05:24 lr 0.000652 time 0.2730 (0.2976) loss 3.7568 (3.5227) grad_norm 1.1646 (1.3486) [2021-04-16 02:22:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][170/1251] eta 0:05:20 lr 0.000652 time 0.2560 (0.2965) loss 3.8327 (3.5377) grad_norm 1.4918 (1.3488) [2021-04-16 02:22:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][180/1251] eta 0:05:16 lr 0.000652 time 0.3123 (0.2955) loss 4.1127 (3.5377) grad_norm 1.2421 (1.3508) [2021-04-16 02:22:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][190/1251] eta 0:05:12 lr 0.000652 time 0.2666 (0.2946) loss 4.3823 (3.5646) grad_norm 1.4120 (1.3484) [2021-04-16 02:23:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][200/1251] eta 0:05:09 lr 0.000652 time 0.2959 (0.2940) loss 3.8361 (3.5666) grad_norm 1.3679 (1.3478) [2021-04-16 02:23:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][210/1251] eta 0:05:05 lr 0.000652 time 0.2776 (0.2931) loss 3.6405 (3.5810) grad_norm 1.3327 (1.3467) [2021-04-16 02:23:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][220/1251] eta 0:05:02 lr 0.000652 time 0.2718 (0.2931) loss 3.1858 (3.5848) grad_norm 1.3643 (1.3452) [2021-04-16 02:23:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][230/1251] eta 0:04:58 lr 0.000652 time 0.2607 (0.2923) loss 3.8937 (3.5919) grad_norm 1.2009 (1.3473) [2021-04-16 02:23:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][240/1251] eta 0:04:55 lr 0.000652 time 0.2897 (0.2918) loss 3.5323 (3.6020) grad_norm 1.3119 (1.3501) [2021-04-16 02:23:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][250/1251] eta 0:04:51 lr 0.000652 time 0.2917 (0.2911) loss 3.5965 (3.6105) grad_norm 1.2582 (1.3483) [2021-04-16 02:23:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][260/1251] eta 0:04:47 lr 0.000652 time 0.2493 (0.2903) loss 3.8764 (3.6208) grad_norm 1.1503 (1.3465) [2021-04-16 02:23:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][270/1251] eta 0:04:44 lr 0.000652 time 0.2957 (0.2898) loss 3.8870 (3.6211) grad_norm 1.3502 (1.3454) [2021-04-16 02:23:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][280/1251] eta 0:04:40 lr 0.000652 time 0.2671 (0.2892) loss 4.2298 (3.6295) grad_norm 1.2029 (1.3432) [2021-04-16 02:23:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][290/1251] eta 0:04:37 lr 0.000652 time 0.2586 (0.2889) loss 4.0231 (3.6302) grad_norm 1.4137 (1.3423) [2021-04-16 02:23:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][300/1251] eta 0:04:34 lr 0.000652 time 0.2647 (0.2885) loss 3.4071 (3.6283) grad_norm 1.4678 (1.3439) [2021-04-16 02:23:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][310/1251] eta 0:04:31 lr 0.000652 time 0.2812 (0.2883) loss 4.3899 (3.6312) grad_norm 1.4821 (1.3439) [2021-04-16 02:23:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][320/1251] eta 0:04:27 lr 0.000652 time 0.2709 (0.2878) loss 3.5073 (3.6400) grad_norm 1.2680 (1.3451) [2021-04-16 02:23:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][330/1251] eta 0:04:24 lr 0.000652 time 0.2750 (0.2875) loss 4.0412 (3.6393) grad_norm 1.3335 (1.3447) [2021-04-16 02:23:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][340/1251] eta 0:04:21 lr 0.000652 time 0.2831 (0.2875) loss 3.9139 (3.6436) grad_norm 1.2367 (1.3427) [2021-04-16 02:23:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][350/1251] eta 0:04:19 lr 0.000652 time 0.2740 (0.2875) loss 3.6377 (3.6432) grad_norm 1.5432 (1.3450) [2021-04-16 02:23:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][360/1251] eta 0:04:15 lr 0.000652 time 0.2752 (0.2871) loss 3.9608 (3.6419) grad_norm 1.1690 (1.3457) [2021-04-16 02:23:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][370/1251] eta 0:04:12 lr 0.000652 time 0.2537 (0.2867) loss 4.1139 (3.6413) grad_norm 1.6482 (1.3468) [2021-04-16 02:23:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][380/1251] eta 0:04:09 lr 0.000652 time 0.2855 (0.2865) loss 2.5806 (3.6412) grad_norm 1.2575 (1.3486) [2021-04-16 02:23:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][390/1251] eta 0:04:06 lr 0.000651 time 0.2510 (0.2861) loss 3.5327 (3.6398) grad_norm 1.6044 (1.3500) [2021-04-16 02:23:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][400/1251] eta 0:04:03 lr 0.000651 time 0.2843 (0.2860) loss 3.9906 (3.6418) grad_norm 1.1914 (1.3494) [2021-04-16 02:23:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][410/1251] eta 0:04:00 lr 0.000651 time 0.2661 (0.2858) loss 3.7336 (3.6370) grad_norm 1.1549 (1.3482) [2021-04-16 02:24:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][420/1251] eta 0:03:57 lr 0.000651 time 0.2668 (0.2855) loss 3.8085 (3.6407) grad_norm 1.3887 (1.3471) [2021-04-16 02:24:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][430/1251] eta 0:03:54 lr 0.000651 time 0.2825 (0.2855) loss 4.3189 (3.6420) grad_norm 1.4043 (1.3473) [2021-04-16 02:24:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][440/1251] eta 0:03:51 lr 0.000651 time 0.2846 (0.2856) loss 3.9536 (3.6413) grad_norm 1.2548 (1.3492) [2021-04-16 02:24:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][450/1251] eta 0:03:48 lr 0.000651 time 0.2817 (0.2854) loss 3.9487 (3.6501) grad_norm 1.2368 (1.3476) [2021-04-16 02:24:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][460/1251] eta 0:03:45 lr 0.000651 time 0.2784 (0.2853) loss 4.4510 (3.6542) grad_norm 1.1835 (1.3477) [2021-04-16 02:24:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][470/1251] eta 0:03:42 lr 0.000651 time 0.2495 (0.2853) loss 2.9168 (3.6470) grad_norm 1.4355 (1.3472) [2021-04-16 02:24:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][480/1251] eta 0:03:39 lr 0.000651 time 0.2970 (0.2851) loss 3.7521 (3.6450) grad_norm 1.2264 (1.3488) [2021-04-16 02:24:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][490/1251] eta 0:03:36 lr 0.000651 time 0.2439 (0.2849) loss 4.1779 (3.6462) grad_norm 1.3787 (1.3489) [2021-04-16 02:24:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][500/1251] eta 0:03:33 lr 0.000651 time 0.2546 (0.2847) loss 3.9444 (3.6540) grad_norm 1.2753 (1.3489) [2021-04-16 02:24:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][510/1251] eta 0:03:30 lr 0.000651 time 0.2525 (0.2845) loss 4.1750 (3.6619) grad_norm 1.3296 (1.3484) [2021-04-16 02:24:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][520/1251] eta 0:03:27 lr 0.000651 time 0.2872 (0.2843) loss 3.1326 (3.6592) grad_norm 1.6477 (1.3483) [2021-04-16 02:24:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][530/1251] eta 0:03:24 lr 0.000651 time 0.2557 (0.2842) loss 4.1609 (3.6625) grad_norm 1.5183 (1.3497) [2021-04-16 02:24:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][540/1251] eta 0:03:22 lr 0.000651 time 0.2737 (0.2842) loss 4.1829 (3.6639) grad_norm 1.3397 (1.3508) [2021-04-16 02:24:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][550/1251] eta 0:03:19 lr 0.000651 time 0.2954 (0.2841) loss 3.0383 (3.6586) grad_norm 1.2007 (1.3492) [2021-04-16 02:24:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][560/1251] eta 0:03:16 lr 0.000651 time 0.3009 (0.2839) loss 4.1329 (3.6572) grad_norm 1.2639 (1.3490) [2021-04-16 02:24:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][570/1251] eta 0:03:13 lr 0.000651 time 0.2585 (0.2837) loss 4.1222 (3.6498) grad_norm 1.7437 (1.3501) [2021-04-16 02:24:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][580/1251] eta 0:03:10 lr 0.000651 time 0.2703 (0.2839) loss 4.1163 (3.6480) grad_norm 1.1916 (1.3502) [2021-04-16 02:24:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][590/1251] eta 0:03:07 lr 0.000651 time 0.2815 (0.2838) loss 4.3503 (3.6481) grad_norm 1.2438 (1.3482) [2021-04-16 02:24:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][600/1251] eta 0:03:04 lr 0.000651 time 0.3130 (0.2838) loss 3.2183 (3.6478) grad_norm 1.3350 (1.3477) [2021-04-16 02:24:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][610/1251] eta 0:03:01 lr 0.000651 time 0.2815 (0.2836) loss 4.2347 (3.6469) grad_norm 1.5125 (1.3476) [2021-04-16 02:24:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][620/1251] eta 0:02:58 lr 0.000651 time 0.2691 (0.2835) loss 3.6845 (3.6500) grad_norm 1.6866 (1.3475) [2021-04-16 02:24:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][630/1251] eta 0:02:55 lr 0.000651 time 0.2639 (0.2833) loss 4.1140 (3.6526) grad_norm 1.3595 (1.3493) [2021-04-16 02:25:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][640/1251] eta 0:02:53 lr 0.000650 time 0.2992 (0.2833) loss 4.0233 (3.6558) grad_norm 1.3574 (1.3504) [2021-04-16 02:25:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][650/1251] eta 0:02:50 lr 0.000650 time 0.2716 (0.2832) loss 3.4925 (3.6575) grad_norm 1.5509 (1.3507) [2021-04-16 02:25:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][660/1251] eta 0:02:47 lr 0.000650 time 0.2910 (0.2831) loss 3.6065 (3.6578) grad_norm 1.3340 (1.3502) [2021-04-16 02:25:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][670/1251] eta 0:02:44 lr 0.000650 time 0.2843 (0.2831) loss 2.8234 (3.6578) grad_norm 1.5584 (1.3505) [2021-04-16 02:25:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][680/1251] eta 0:02:41 lr 0.000650 time 0.2622 (0.2832) loss 3.8638 (3.6599) grad_norm 1.4351 (1.3511) [2021-04-16 02:25:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][690/1251] eta 0:02:38 lr 0.000650 time 0.2783 (0.2831) loss 2.9105 (3.6607) grad_norm 1.6605 (1.3527) [2021-04-16 02:25:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][700/1251] eta 0:02:36 lr 0.000650 time 0.2865 (0.2832) loss 2.8013 (3.6572) grad_norm 1.4992 (1.3526) [2021-04-16 02:25:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][710/1251] eta 0:02:33 lr 0.000650 time 0.2683 (0.2832) loss 2.6927 (3.6565) grad_norm 1.3289 (1.3536) [2021-04-16 02:25:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][720/1251] eta 0:02:30 lr 0.000650 time 0.2558 (0.2832) loss 2.5340 (3.6545) grad_norm 1.2080 (1.3530) [2021-04-16 02:25:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][730/1251] eta 0:02:27 lr 0.000650 time 0.2461 (0.2833) loss 4.3631 (3.6571) grad_norm 1.3638 (1.3532) [2021-04-16 02:25:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][740/1251] eta 0:02:24 lr 0.000650 time 0.2714 (0.2833) loss 4.3177 (3.6610) grad_norm 1.3687 (1.3544) [2021-04-16 02:25:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][750/1251] eta 0:02:21 lr 0.000650 time 0.2447 (0.2832) loss 3.6566 (3.6640) grad_norm 1.1392 (1.3548) [2021-04-16 02:25:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][760/1251] eta 0:02:19 lr 0.000650 time 0.2713 (0.2832) loss 2.9065 (3.6623) grad_norm 1.4280 (1.3536) [2021-04-16 02:25:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][770/1251] eta 0:02:16 lr 0.000650 time 0.2682 (0.2831) loss 4.2260 (3.6618) grad_norm 1.5603 (1.3541) [2021-04-16 02:25:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][780/1251] eta 0:02:13 lr 0.000650 time 0.2634 (0.2830) loss 4.1389 (3.6622) grad_norm 1.3520 (1.3544) [2021-04-16 02:25:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][790/1251] eta 0:02:10 lr 0.000650 time 0.2790 (0.2831) loss 4.0871 (3.6614) grad_norm 1.4954 (1.3560) [2021-04-16 02:25:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][800/1251] eta 0:02:07 lr 0.000650 time 0.2742 (0.2830) loss 3.8554 (3.6609) grad_norm 1.2508 (1.3551) [2021-04-16 02:25:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][810/1251] eta 0:02:04 lr 0.000650 time 0.2568 (0.2829) loss 3.5509 (3.6619) grad_norm 1.2893 (1.3549) [2021-04-16 02:25:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][820/1251] eta 0:02:01 lr 0.000650 time 0.2894 (0.2828) loss 3.8295 (3.6624) grad_norm 1.1919 (1.3543) [2021-04-16 02:25:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][830/1251] eta 0:01:59 lr 0.000650 time 0.2825 (0.2828) loss 4.3932 (3.6642) grad_norm 1.4017 (1.3547) [2021-04-16 02:25:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][840/1251] eta 0:01:56 lr 0.000650 time 0.2642 (0.2828) loss 3.9232 (3.6649) grad_norm 1.2371 (1.3540) [2021-04-16 02:26:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][850/1251] eta 0:01:53 lr 0.000650 time 0.2869 (0.2827) loss 3.8041 (3.6630) grad_norm 1.3302 (1.3539) [2021-04-16 02:26:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][860/1251] eta 0:01:50 lr 0.000650 time 0.2631 (0.2826) loss 4.0451 (3.6621) grad_norm 1.4231 (1.3535) [2021-04-16 02:26:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][870/1251] eta 0:01:47 lr 0.000650 time 0.2833 (0.2827) loss 3.6370 (3.6628) grad_norm 1.3036 (1.3533) [2021-04-16 02:26:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][880/1251] eta 0:01:44 lr 0.000650 time 0.2699 (0.2826) loss 3.8920 (3.6642) grad_norm 1.2568 (1.3534) [2021-04-16 02:26:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][890/1251] eta 0:01:42 lr 0.000650 time 0.2857 (0.2826) loss 2.3671 (3.6665) grad_norm 1.4703 (1.3535) [2021-04-16 02:26:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][900/1251] eta 0:01:39 lr 0.000649 time 0.2771 (0.2825) loss 3.7161 (3.6654) grad_norm 1.2097 (1.3533) [2021-04-16 02:26:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][910/1251] eta 0:01:36 lr 0.000649 time 0.2753 (0.2825) loss 4.0423 (3.6695) grad_norm 1.4703 (1.3532) [2021-04-16 02:26:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][920/1251] eta 0:01:33 lr 0.000649 time 0.2980 (0.2824) loss 3.3434 (3.6659) grad_norm 1.6100 (1.3540) [2021-04-16 02:26:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][930/1251] eta 0:01:30 lr 0.000649 time 0.2943 (0.2824) loss 2.1958 (3.6648) grad_norm 1.1770 (1.3541) [2021-04-16 02:26:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][940/1251] eta 0:01:27 lr 0.000649 time 0.2739 (0.2823) loss 3.7621 (3.6653) grad_norm 1.6386 (1.3546) [2021-04-16 02:26:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][950/1251] eta 0:01:24 lr 0.000649 time 0.2593 (0.2822) loss 4.4667 (3.6616) grad_norm 1.4111 (1.3548) [2021-04-16 02:26:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][960/1251] eta 0:01:22 lr 0.000649 time 0.2655 (0.2822) loss 4.5415 (3.6605) grad_norm 1.1739 (1.3550) [2021-04-16 02:26:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][970/1251] eta 0:01:19 lr 0.000649 time 0.2583 (0.2821) loss 4.1681 (3.6612) grad_norm 1.2550 (1.3546) [2021-04-16 02:26:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][980/1251] eta 0:01:16 lr 0.000649 time 0.2593 (0.2820) loss 4.0373 (3.6614) grad_norm 1.4064 (1.3538) [2021-04-16 02:26:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][990/1251] eta 0:01:13 lr 0.000649 time 0.2611 (0.2820) loss 3.9639 (3.6630) grad_norm 1.3275 (1.3534) [2021-04-16 02:26:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][1000/1251] eta 0:01:10 lr 0.000649 time 0.2566 (0.2819) loss 4.0048 (3.6635) grad_norm 1.3727 (1.3532) [2021-04-16 02:26:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][1010/1251] eta 0:01:07 lr 0.000649 time 0.2763 (0.2819) loss 3.9038 (3.6628) grad_norm 1.6710 (1.3543) [2021-04-16 02:26:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][1020/1251] eta 0:01:05 lr 0.000649 time 0.2709 (0.2820) loss 3.8291 (3.6624) grad_norm 1.4353 (1.3545) [2021-04-16 02:26:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][1030/1251] eta 0:01:02 lr 0.000649 time 0.2812 (0.2819) loss 2.8650 (3.6631) grad_norm 1.3148 (1.3543) [2021-04-16 02:26:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][1040/1251] eta 0:00:59 lr 0.000649 time 0.2634 (0.2821) loss 3.9725 (3.6647) grad_norm 1.3667 (1.3539) [2021-04-16 02:26:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][1050/1251] eta 0:00:56 lr 0.000649 time 0.2571 (0.2822) loss 2.8468 (3.6618) grad_norm 1.4151 (1.3541) [2021-04-16 02:27:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][1060/1251] eta 0:00:53 lr 0.000649 time 0.2959 (0.2822) loss 4.1320 (3.6619) grad_norm 1.2949 (1.3543) [2021-04-16 02:27:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][1070/1251] eta 0:00:51 lr 0.000649 time 0.2842 (0.2822) loss 3.1299 (3.6626) grad_norm 1.4053 (1.3547) [2021-04-16 02:27:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][1080/1251] eta 0:00:48 lr 0.000649 time 0.2999 (0.2821) loss 4.0153 (3.6641) grad_norm 1.3776 (1.3547) [2021-04-16 02:27:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][1090/1251] eta 0:00:45 lr 0.000649 time 0.2789 (0.2821) loss 3.0009 (3.6626) grad_norm 1.3987 (1.3551) [2021-04-16 02:27:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][1100/1251] eta 0:00:42 lr 0.000649 time 0.2515 (0.2820) loss 4.3636 (3.6624) grad_norm 1.3029 (1.3546) [2021-04-16 02:27:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][1110/1251] eta 0:00:39 lr 0.000649 time 0.3052 (0.2820) loss 3.1322 (3.6623) grad_norm 1.1677 (1.3544) [2021-04-16 02:27:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][1120/1251] eta 0:00:36 lr 0.000649 time 0.2719 (0.2820) loss 2.8623 (3.6592) grad_norm 1.3269 (1.3543) [2021-04-16 02:27:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][1130/1251] eta 0:00:34 lr 0.000649 time 0.2686 (0.2819) loss 4.3262 (3.6612) grad_norm 1.4722 (1.3541) [2021-04-16 02:27:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][1140/1251] eta 0:00:31 lr 0.000649 time 0.2802 (0.2818) loss 2.7514 (3.6586) grad_norm 1.4166 (1.3538) [2021-04-16 02:27:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][1150/1251] eta 0:00:28 lr 0.000648 time 0.2847 (0.2818) loss 4.0463 (3.6584) grad_norm 1.2387 (1.3541) [2021-04-16 02:27:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][1160/1251] eta 0:00:25 lr 0.000648 time 0.2688 (0.2819) loss 3.3521 (3.6581) grad_norm 1.2540 (1.3538) [2021-04-16 02:27:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][1170/1251] eta 0:00:22 lr 0.000648 time 0.2888 (0.2819) loss 3.8434 (3.6573) grad_norm 1.2639 (1.3534) [2021-04-16 02:27:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][1180/1251] eta 0:00:20 lr 0.000648 time 0.2916 (0.2818) loss 3.6497 (3.6578) grad_norm 1.6057 (1.3539) [2021-04-16 02:27:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][1190/1251] eta 0:00:17 lr 0.000648 time 0.3122 (0.2818) loss 4.0664 (3.6579) grad_norm 1.1742 (1.3541) [2021-04-16 02:27:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][1200/1251] eta 0:00:14 lr 0.000648 time 0.2737 (0.2817) loss 2.4500 (3.6559) grad_norm 1.1912 (1.3533) [2021-04-16 02:27:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][1210/1251] eta 0:00:11 lr 0.000648 time 0.2919 (0.2817) loss 3.6515 (3.6570) grad_norm 1.2875 (1.3541) [2021-04-16 02:27:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][1220/1251] eta 0:00:08 lr 0.000648 time 0.2822 (0.2816) loss 3.6795 (3.6582) grad_norm 1.5242 (1.3546) [2021-04-16 02:27:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][1230/1251] eta 0:00:05 lr 0.000648 time 0.2751 (0.2816) loss 3.4738 (3.6580) grad_norm 1.2170 (1.3554) [2021-04-16 02:27:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][1240/1251] eta 0:00:03 lr 0.000648 time 0.3924 (0.2815) loss 4.1076 (3.6557) grad_norm 1.2191 (1.3548) [2021-04-16 02:27:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [121/300][1250/1251] eta 0:00:00 lr 0.000648 time 0.2479 (0.2813) loss 3.6755 (3.6555) grad_norm 1.2378 (1.3545) [2021-04-16 02:27:56 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 121 training takes 0:05:55 [2021-04-16 02:27:56 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_121.pth saving...... [2021-04-16 02:28:05 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_121.pth saved !!! [2021-04-16 02:28:06 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.389 (1.389) Loss 1.1586 (1.1586) Acc@1 73.438 (73.438) Acc@5 90.918 (90.918) [2021-04-16 02:28:08 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.351 (0.235) Loss 1.0902 (1.1051) Acc@1 75.293 (73.739) Acc@5 92.969 (92.418) [2021-04-16 02:28:10 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.104 (0.226) Loss 1.0462 (1.0943) Acc@1 76.465 (74.461) Acc@5 92.773 (92.573) [2021-04-16 02:28:12 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.134 (0.240) Loss 1.1018 (1.0885) Acc@1 73.535 (74.540) Acc@5 92.676 (92.641) [2021-04-16 02:28:14 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.079 (0.227) Loss 1.0721 (1.0828) Acc@1 74.219 (74.698) Acc@5 92.871 (92.704) [2021-04-16 02:28:21 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 74.626 Acc@5 92.608 [2021-04-16 02:28:21 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 74.6% [2021-04-16 02:28:21 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 74.73% [2021-04-16 02:28:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][0/1251] eta 1:07:42 lr 0.000648 time 3.2471 (3.2471) loss 3.1380 (3.1380) grad_norm 1.2183 (1.2183) [2021-04-16 02:28:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][10/1251] eta 0:11:32 lr 0.000648 time 0.4133 (0.5579) loss 3.6768 (3.4732) grad_norm 1.4012 (1.3480) [2021-04-16 02:28:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][20/1251] eta 0:08:42 lr 0.000648 time 0.2693 (0.4249) loss 4.1932 (3.4709) grad_norm 1.4770 (1.3807) [2021-04-16 02:28:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][30/1251] eta 0:07:39 lr 0.000648 time 0.2633 (0.3763) loss 2.4428 (3.4784) grad_norm 1.2653 (1.3884) [2021-04-16 02:28:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][40/1251] eta 0:07:07 lr 0.000648 time 0.2846 (0.3529) loss 3.7750 (3.5251) grad_norm 1.6300 (1.4074) [2021-04-16 02:28:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][50/1251] eta 0:06:45 lr 0.000648 time 0.2922 (0.3372) loss 4.1732 (3.5299) grad_norm 1.2467 (1.3898) [2021-04-16 02:28:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][60/1251] eta 0:06:33 lr 0.000648 time 0.4401 (0.3305) loss 4.3337 (3.5374) grad_norm 1.1842 (1.3877) [2021-04-16 02:28:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][70/1251] eta 0:06:21 lr 0.000648 time 0.2866 (0.3234) loss 3.7033 (3.5614) grad_norm 1.3169 (1.3818) [2021-04-16 02:28:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][80/1251] eta 0:06:12 lr 0.000648 time 0.2590 (0.3180) loss 4.3790 (3.5822) grad_norm 1.4080 (1.3847) [2021-04-16 02:28:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][90/1251] eta 0:06:03 lr 0.000648 time 0.2709 (0.3134) loss 2.3790 (3.5413) grad_norm 1.1845 (1.3973) [2021-04-16 02:28:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][100/1251] eta 0:05:58 lr 0.000648 time 0.2646 (0.3111) loss 3.7018 (3.5709) grad_norm 1.3212 (1.3944) [2021-04-16 02:28:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][110/1251] eta 0:05:51 lr 0.000648 time 0.2965 (0.3083) loss 3.6212 (3.5560) grad_norm 1.4503 (1.3974) [2021-04-16 02:28:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][120/1251] eta 0:05:45 lr 0.000648 time 0.2715 (0.3053) loss 3.2562 (3.5557) grad_norm 1.3441 (1.3965) [2021-04-16 02:29:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][130/1251] eta 0:05:40 lr 0.000648 time 0.2784 (0.3042) loss 3.0529 (3.5564) grad_norm 1.3430 (1.3839) [2021-04-16 02:29:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][140/1251] eta 0:05:35 lr 0.000648 time 0.2990 (0.3023) loss 3.8814 (3.5494) grad_norm 1.2981 (1.3796) [2021-04-16 02:29:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][150/1251] eta 0:05:32 lr 0.000647 time 0.2943 (0.3016) loss 3.7415 (3.5469) grad_norm 1.4083 (1.3751) [2021-04-16 02:29:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][160/1251] eta 0:05:27 lr 0.000647 time 0.2697 (0.3005) loss 2.6435 (3.5413) grad_norm 1.3231 (1.3700) [2021-04-16 02:29:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][170/1251] eta 0:05:23 lr 0.000647 time 0.3131 (0.2996) loss 4.1694 (3.5442) grad_norm 1.4228 (1.3748) [2021-04-16 02:29:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][180/1251] eta 0:05:20 lr 0.000647 time 0.2548 (0.2989) loss 3.8877 (3.5680) grad_norm 1.2991 (1.3716) [2021-04-16 02:29:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][190/1251] eta 0:05:15 lr 0.000647 time 0.2656 (0.2977) loss 3.6312 (3.5509) grad_norm 1.2485 (1.3703) [2021-04-16 02:29:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][200/1251] eta 0:05:12 lr 0.000647 time 0.3036 (0.2971) loss 2.5602 (3.5392) grad_norm 1.3318 (1.3694) [2021-04-16 02:29:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][210/1251] eta 0:05:08 lr 0.000647 time 0.2767 (0.2964) loss 4.1167 (3.5541) grad_norm 1.2994 (1.3683) [2021-04-16 02:29:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][220/1251] eta 0:05:05 lr 0.000647 time 0.2772 (0.2959) loss 3.8080 (3.5496) grad_norm 1.4943 (1.3710) [2021-04-16 02:29:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][230/1251] eta 0:05:01 lr 0.000647 time 0.2794 (0.2950) loss 3.4952 (3.5434) grad_norm 1.1877 (1.3707) [2021-04-16 02:29:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][240/1251] eta 0:04:57 lr 0.000647 time 0.2700 (0.2943) loss 3.7436 (3.5611) grad_norm 1.4741 (1.3719) [2021-04-16 02:29:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][250/1251] eta 0:04:54 lr 0.000647 time 0.2714 (0.2941) loss 3.2462 (3.5656) grad_norm 1.2083 (1.3706) [2021-04-16 02:29:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][260/1251] eta 0:04:50 lr 0.000647 time 0.2808 (0.2935) loss 2.6825 (3.5632) grad_norm 1.2241 (1.3670) [2021-04-16 02:29:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][270/1251] eta 0:04:47 lr 0.000647 time 0.2918 (0.2928) loss 4.1311 (3.5625) grad_norm 1.4572 (1.3663) [2021-04-16 02:29:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][280/1251] eta 0:04:43 lr 0.000647 time 0.2752 (0.2922) loss 3.9733 (3.5623) grad_norm 1.5663 (1.3705) [2021-04-16 02:29:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][290/1251] eta 0:04:40 lr 0.000647 time 0.2985 (0.2918) loss 4.0071 (3.5784) grad_norm 1.3417 (1.3697) [2021-04-16 02:29:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][300/1251] eta 0:04:37 lr 0.000647 time 0.2700 (0.2915) loss 4.0230 (3.5787) grad_norm 1.5042 (1.3694) [2021-04-16 02:29:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][310/1251] eta 0:04:34 lr 0.000647 time 0.2650 (0.2912) loss 3.8546 (3.5855) grad_norm 1.3356 (1.3682) [2021-04-16 02:29:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][320/1251] eta 0:04:31 lr 0.000647 time 0.2986 (0.2911) loss 3.8725 (3.5833) grad_norm 1.2077 (1.3666) [2021-04-16 02:29:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][330/1251] eta 0:04:28 lr 0.000647 time 0.2779 (0.2911) loss 4.3143 (3.5918) grad_norm 1.4501 (1.3667) [2021-04-16 02:30:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][340/1251] eta 0:04:24 lr 0.000647 time 0.2758 (0.2905) loss 3.7330 (3.5890) grad_norm 1.4766 (1.3648) [2021-04-16 02:30:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][350/1251] eta 0:04:22 lr 0.000647 time 0.2993 (0.2908) loss 3.7436 (3.5918) grad_norm 1.8116 (1.3679) [2021-04-16 02:30:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][360/1251] eta 0:04:18 lr 0.000647 time 0.2670 (0.2903) loss 4.0148 (3.5884) grad_norm 1.5200 (1.3689) [2021-04-16 02:30:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][370/1251] eta 0:04:15 lr 0.000647 time 0.2528 (0.2899) loss 3.6461 (3.5899) grad_norm 1.2272 (1.3679) [2021-04-16 02:30:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][380/1251] eta 0:04:12 lr 0.000647 time 0.2633 (0.2895) loss 4.0947 (3.5947) grad_norm 1.3653 (1.3658) [2021-04-16 02:30:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][390/1251] eta 0:04:09 lr 0.000647 time 0.2881 (0.2892) loss 3.1424 (3.5947) grad_norm 1.1519 (1.3652) [2021-04-16 02:30:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][400/1251] eta 0:04:05 lr 0.000646 time 0.3042 (0.2890) loss 4.5190 (3.5970) grad_norm 1.2829 (1.3647) [2021-04-16 02:30:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][410/1251] eta 0:04:02 lr 0.000646 time 0.2669 (0.2886) loss 4.2278 (3.5927) grad_norm 1.3852 (1.3649) [2021-04-16 02:30:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][420/1251] eta 0:03:59 lr 0.000646 time 0.2998 (0.2884) loss 3.4104 (3.5952) grad_norm 1.4300 (1.3669) [2021-04-16 02:30:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][430/1251] eta 0:03:56 lr 0.000646 time 0.2584 (0.2881) loss 3.6436 (3.5971) grad_norm 1.2236 (1.3673) [2021-04-16 02:30:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][440/1251] eta 0:03:53 lr 0.000646 time 0.2778 (0.2883) loss 2.9167 (3.6002) grad_norm 1.4189 (1.3688) [2021-04-16 02:30:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][450/1251] eta 0:03:51 lr 0.000646 time 0.2669 (0.2884) loss 3.3231 (3.6002) grad_norm 1.3697 (1.3684) [2021-04-16 02:30:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][460/1251] eta 0:03:47 lr 0.000646 time 0.2688 (0.2881) loss 3.6315 (3.5994) grad_norm 1.5312 (1.3674) [2021-04-16 02:30:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][470/1251] eta 0:03:45 lr 0.000646 time 0.2966 (0.2881) loss 3.5529 (3.5974) grad_norm 1.3136 (1.3666) [2021-04-16 02:30:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][480/1251] eta 0:03:41 lr 0.000646 time 0.2582 (0.2878) loss 3.9531 (3.6016) grad_norm 1.5068 (1.3683) [2021-04-16 02:30:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][490/1251] eta 0:03:38 lr 0.000646 time 0.2557 (0.2876) loss 3.0324 (3.6020) grad_norm 1.4568 (1.3681) [2021-04-16 02:30:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][500/1251] eta 0:03:35 lr 0.000646 time 0.2612 (0.2874) loss 2.6685 (3.6028) grad_norm 1.5276 (1.3682) [2021-04-16 02:30:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][510/1251] eta 0:03:32 lr 0.000646 time 0.2883 (0.2872) loss 3.1690 (3.6044) grad_norm 1.4445 (1.3678) [2021-04-16 02:30:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][520/1251] eta 0:03:29 lr 0.000646 time 0.2679 (0.2869) loss 3.0080 (3.6055) grad_norm 1.4188 (1.3668) [2021-04-16 02:30:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][530/1251] eta 0:03:26 lr 0.000646 time 0.2760 (0.2868) loss 2.8531 (3.6008) grad_norm 1.3494 (1.3669) [2021-04-16 02:30:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][540/1251] eta 0:03:23 lr 0.000646 time 0.2822 (0.2868) loss 3.0144 (3.6055) grad_norm 1.5820 (1.3698) [2021-04-16 02:30:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][550/1251] eta 0:03:20 lr 0.000646 time 0.2828 (0.2866) loss 3.8422 (3.6052) grad_norm 1.2793 (1.3701) [2021-04-16 02:31:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][560/1251] eta 0:03:17 lr 0.000646 time 0.2732 (0.2864) loss 4.0246 (3.6058) grad_norm 1.2721 (1.3690) [2021-04-16 02:31:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][570/1251] eta 0:03:15 lr 0.000646 time 0.4305 (0.2867) loss 3.8666 (3.6107) grad_norm 1.2849 (1.3693) [2021-04-16 02:31:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][580/1251] eta 0:03:12 lr 0.000646 time 0.3088 (0.2866) loss 2.5204 (3.6101) grad_norm 1.5516 (1.3701) [2021-04-16 02:31:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][590/1251] eta 0:03:09 lr 0.000646 time 0.2780 (0.2868) loss 2.9375 (3.6120) grad_norm 1.4637 (1.3699) [2021-04-16 02:31:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][600/1251] eta 0:03:06 lr 0.000646 time 0.2711 (0.2866) loss 2.9466 (3.6077) grad_norm 1.4202 (1.3704) [2021-04-16 02:31:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][610/1251] eta 0:03:03 lr 0.000646 time 0.2653 (0.2864) loss 4.2218 (3.6040) grad_norm 1.2480 (1.3694) [2021-04-16 02:31:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][620/1251] eta 0:03:00 lr 0.000646 time 0.2728 (0.2862) loss 3.8069 (3.6044) grad_norm 1.3915 (1.3694) [2021-04-16 02:31:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][630/1251] eta 0:02:57 lr 0.000646 time 0.3016 (0.2860) loss 4.4930 (3.6072) grad_norm 1.2829 (1.3698) [2021-04-16 02:31:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][640/1251] eta 0:02:54 lr 0.000646 time 0.2945 (0.2861) loss 3.6486 (3.6111) grad_norm 1.4722 (1.3703) [2021-04-16 02:31:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][650/1251] eta 0:02:51 lr 0.000645 time 0.2561 (0.2859) loss 2.8119 (3.6155) grad_norm 1.4673 (1.3689) [2021-04-16 02:31:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][660/1251] eta 0:02:48 lr 0.000645 time 0.2859 (0.2857) loss 3.6062 (3.6136) grad_norm 1.3830 (1.3690) [2021-04-16 02:31:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][670/1251] eta 0:02:45 lr 0.000645 time 0.2822 (0.2856) loss 4.0877 (3.6124) grad_norm 1.3892 (1.3693) [2021-04-16 02:31:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][680/1251] eta 0:02:43 lr 0.000645 time 0.2826 (0.2855) loss 3.9468 (3.6106) grad_norm 1.3243 (1.3694) [2021-04-16 02:31:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][690/1251] eta 0:02:40 lr 0.000645 time 0.3007 (0.2853) loss 3.5689 (3.6070) grad_norm 1.3605 (1.3691) [2021-04-16 02:31:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][700/1251] eta 0:02:37 lr 0.000645 time 0.2674 (0.2852) loss 3.9288 (3.6083) grad_norm 1.3982 (1.3685) [2021-04-16 02:31:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][710/1251] eta 0:02:34 lr 0.000645 time 0.2630 (0.2853) loss 2.6607 (3.6064) grad_norm 1.1745 (1.3678) [2021-04-16 02:31:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][720/1251] eta 0:02:31 lr 0.000645 time 0.2558 (0.2852) loss 3.6283 (3.6038) grad_norm 1.4029 (1.3673) [2021-04-16 02:31:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][730/1251] eta 0:02:28 lr 0.000645 time 0.2964 (0.2851) loss 4.2248 (3.6021) grad_norm 1.3392 (1.3657) [2021-04-16 02:31:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][740/1251] eta 0:02:25 lr 0.000645 time 0.2895 (0.2850) loss 4.1821 (3.6038) grad_norm 1.1258 (1.3657) [2021-04-16 02:31:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][750/1251] eta 0:02:22 lr 0.000645 time 0.2619 (0.2849) loss 2.6863 (3.6003) grad_norm 1.6629 (1.3663) [2021-04-16 02:31:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][760/1251] eta 0:02:19 lr 0.000645 time 0.2897 (0.2849) loss 2.4574 (3.5996) grad_norm 1.1814 (1.3662) [2021-04-16 02:32:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][770/1251] eta 0:02:16 lr 0.000645 time 0.3008 (0.2847) loss 2.2879 (3.5982) grad_norm 1.1858 (1.3654) [2021-04-16 02:32:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][780/1251] eta 0:02:14 lr 0.000645 time 0.2465 (0.2846) loss 3.5722 (3.5971) grad_norm 1.3519 (1.3654) [2021-04-16 02:32:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][790/1251] eta 0:02:11 lr 0.000645 time 0.2789 (0.2847) loss 4.0412 (3.6017) grad_norm 1.4765 (1.3655) [2021-04-16 02:32:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][800/1251] eta 0:02:08 lr 0.000645 time 0.2913 (0.2846) loss 3.8036 (3.6031) grad_norm 1.2947 (1.3653) [2021-04-16 02:32:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][810/1251] eta 0:02:05 lr 0.000645 time 0.2477 (0.2845) loss 3.9706 (3.6037) grad_norm 1.6770 (1.3652) [2021-04-16 02:32:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][820/1251] eta 0:02:02 lr 0.000645 time 0.2748 (0.2844) loss 3.6521 (3.6074) grad_norm 1.2296 (1.3644) [2021-04-16 02:32:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][830/1251] eta 0:01:59 lr 0.000645 time 0.2791 (0.2845) loss 3.2590 (3.6079) grad_norm 1.1484 (1.3649) [2021-04-16 02:32:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][840/1251] eta 0:01:56 lr 0.000645 time 0.2677 (0.2844) loss 3.6982 (3.6073) grad_norm 1.2717 (1.3656) [2021-04-16 02:32:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][850/1251] eta 0:01:54 lr 0.000645 time 0.2961 (0.2844) loss 4.2561 (3.6085) grad_norm 1.4563 (1.3662) [2021-04-16 02:32:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][860/1251] eta 0:01:51 lr 0.000645 time 0.2646 (0.2843) loss 4.0602 (3.6074) grad_norm 1.1857 (1.3653) [2021-04-16 02:32:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][870/1251] eta 0:01:48 lr 0.000645 time 0.3102 (0.2841) loss 3.5456 (3.6078) grad_norm 1.4553 (1.3660) [2021-04-16 02:32:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][880/1251] eta 0:01:45 lr 0.000645 time 0.2596 (0.2840) loss 4.1403 (3.6085) grad_norm 1.3836 (1.3662) [2021-04-16 02:32:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][890/1251] eta 0:01:42 lr 0.000645 time 0.2704 (0.2839) loss 3.7816 (3.6089) grad_norm 1.2355 (1.3658) [2021-04-16 02:32:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][900/1251] eta 0:01:39 lr 0.000644 time 0.2594 (0.2838) loss 3.1530 (3.6064) grad_norm 1.2133 (1.3646) [2021-04-16 02:32:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][910/1251] eta 0:01:36 lr 0.000644 time 0.2679 (0.2838) loss 4.1449 (3.6103) grad_norm 1.6571 (1.3652) [2021-04-16 02:32:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][920/1251] eta 0:01:33 lr 0.000644 time 0.2711 (0.2837) loss 4.3704 (3.6132) grad_norm 1.3752 (1.3645) [2021-04-16 02:32:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][930/1251] eta 0:01:31 lr 0.000644 time 0.2551 (0.2837) loss 2.5019 (3.6150) grad_norm 1.2512 (1.3641) [2021-04-16 02:32:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][940/1251] eta 0:01:28 lr 0.000644 time 0.2917 (0.2837) loss 2.9376 (3.6153) grad_norm 1.3441 (1.3635) [2021-04-16 02:32:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][950/1251] eta 0:01:25 lr 0.000644 time 0.2902 (0.2836) loss 3.1150 (3.6155) grad_norm 1.2655 (1.3639) [2021-04-16 02:32:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][960/1251] eta 0:01:22 lr 0.000644 time 0.2952 (0.2835) loss 2.6860 (3.6175) grad_norm 1.3281 (1.3635) [2021-04-16 02:32:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][970/1251] eta 0:01:19 lr 0.000644 time 0.2845 (0.2835) loss 4.1806 (3.6184) grad_norm 1.3817 (1.3627) [2021-04-16 02:32:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][980/1251] eta 0:01:16 lr 0.000644 time 0.2769 (0.2834) loss 3.7764 (3.6187) grad_norm 1.2217 (1.3622) [2021-04-16 02:33:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][990/1251] eta 0:01:13 lr 0.000644 time 0.2845 (0.2834) loss 3.9296 (3.6173) grad_norm 1.3945 (1.3623) [2021-04-16 02:33:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][1000/1251] eta 0:01:11 lr 0.000644 time 0.2653 (0.2833) loss 3.7461 (3.6168) grad_norm 1.3091 (1.3631) [2021-04-16 02:33:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][1010/1251] eta 0:01:08 lr 0.000644 time 0.2686 (0.2833) loss 3.6578 (3.6179) grad_norm 1.1575 (1.3636) [2021-04-16 02:33:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][1020/1251] eta 0:01:05 lr 0.000644 time 0.2689 (0.2833) loss 4.1575 (3.6205) grad_norm 1.3977 (1.3640) [2021-04-16 02:33:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][1030/1251] eta 0:01:02 lr 0.000644 time 0.2892 (0.2833) loss 3.5960 (3.6203) grad_norm 1.2533 (1.3636) [2021-04-16 02:33:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][1040/1251] eta 0:00:59 lr 0.000644 time 0.2766 (0.2834) loss 3.8778 (3.6215) grad_norm 1.3244 (1.3637) [2021-04-16 02:33:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][1050/1251] eta 0:00:56 lr 0.000644 time 0.2658 (0.2834) loss 3.1775 (3.6203) grad_norm 1.2918 (1.3634) [2021-04-16 02:33:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][1060/1251] eta 0:00:54 lr 0.000644 time 0.2661 (0.2833) loss 3.8416 (3.6213) grad_norm 1.3671 (1.3639) [2021-04-16 02:33:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][1070/1251] eta 0:00:51 lr 0.000644 time 0.2779 (0.2832) loss 3.9166 (3.6221) grad_norm 1.2087 (1.3637) [2021-04-16 02:33:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][1080/1251] eta 0:00:48 lr 0.000644 time 0.2848 (0.2832) loss 3.6077 (3.6248) grad_norm 1.3690 (1.3639) [2021-04-16 02:33:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][1090/1251] eta 0:00:45 lr 0.000644 time 0.2609 (0.2832) loss 3.8662 (3.6247) grad_norm 1.2491 (1.3639) [2021-04-16 02:33:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][1100/1251] eta 0:00:42 lr 0.000644 time 0.2501 (0.2832) loss 4.3975 (3.6219) grad_norm 1.2776 (1.3638) [2021-04-16 02:33:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][1110/1251] eta 0:00:39 lr 0.000644 time 0.2859 (0.2831) loss 4.2281 (3.6211) grad_norm 1.3016 (1.3630) [2021-04-16 02:33:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][1120/1251] eta 0:00:37 lr 0.000644 time 0.2951 (0.2831) loss 3.4414 (3.6199) grad_norm 1.3155 (1.3623) [2021-04-16 02:33:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][1130/1251] eta 0:00:34 lr 0.000644 time 0.2680 (0.2830) loss 4.1733 (3.6189) grad_norm 1.4188 (1.3621) [2021-04-16 02:33:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][1140/1251] eta 0:00:31 lr 0.000644 time 0.2855 (0.2831) loss 4.3168 (3.6203) grad_norm 1.3020 (1.3620) [2021-04-16 02:33:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][1150/1251] eta 0:00:28 lr 0.000644 time 0.2866 (0.2831) loss 3.5021 (3.6196) grad_norm 1.2984 (1.3620) [2021-04-16 02:33:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][1160/1251] eta 0:00:25 lr 0.000643 time 0.3017 (0.2831) loss 4.2534 (3.6210) grad_norm 1.3955 (1.3616) [2021-04-16 02:33:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][1170/1251] eta 0:00:22 lr 0.000643 time 0.2498 (0.2831) loss 3.9429 (3.6194) grad_norm 1.3077 (1.3623) [2021-04-16 02:33:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][1180/1251] eta 0:00:20 lr 0.000643 time 0.2697 (0.2831) loss 3.7268 (3.6211) grad_norm 1.2481 (1.3619) [2021-04-16 02:33:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][1190/1251] eta 0:00:17 lr 0.000643 time 0.2640 (0.2830) loss 2.3191 (3.6192) grad_norm 1.4867 (1.3620) [2021-04-16 02:34:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][1200/1251] eta 0:00:14 lr 0.000643 time 0.3120 (0.2831) loss 3.8735 (3.6198) grad_norm 1.3236 (1.3618) [2021-04-16 02:34:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][1210/1251] eta 0:00:11 lr 0.000643 time 0.2849 (0.2830) loss 2.5646 (3.6168) grad_norm 1.2704 (1.3617) [2021-04-16 02:34:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][1220/1251] eta 0:00:08 lr 0.000643 time 0.2610 (0.2830) loss 4.3717 (3.6165) grad_norm 1.4795 (1.3616) [2021-04-16 02:34:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][1230/1251] eta 0:00:05 lr 0.000643 time 0.2711 (0.2829) loss 3.9934 (3.6153) grad_norm 1.3246 (1.3617) [2021-04-16 02:34:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][1240/1251] eta 0:00:03 lr 0.000643 time 0.2479 (0.2828) loss 4.6168 (3.6160) grad_norm 1.2129 (1.3610) [2021-04-16 02:34:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [122/300][1250/1251] eta 0:00:00 lr 0.000643 time 0.2485 (0.2825) loss 3.0901 (3.6155) grad_norm 1.4140 (1.3608) [2021-04-16 02:34:18 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 122 training takes 0:05:57 [2021-04-16 02:34:18 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_122.pth saving...... [2021-04-16 02:34:29 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_122.pth saved !!! [2021-04-16 02:34:30 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.100 (1.100) Loss 1.0713 (1.0713) Acc@1 76.074 (76.074) Acc@5 91.211 (91.211) [2021-04-16 02:34:31 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.080 (0.196) Loss 1.0532 (1.0876) Acc@1 76.758 (74.672) Acc@5 92.578 (92.125) [2021-04-16 02:34:33 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.129 (0.207) Loss 1.0431 (1.0795) Acc@1 75.684 (74.702) Acc@5 92.383 (92.499) [2021-04-16 02:34:36 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.154 (0.224) Loss 1.0836 (1.0765) Acc@1 75.098 (74.694) Acc@5 93.164 (92.622) [2021-04-16 02:34:38 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.198 (0.219) Loss 1.0246 (1.0772) Acc@1 75.391 (74.667) Acc@5 93.359 (92.585) [2021-04-16 02:34:43 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 74.822 Acc@5 92.616 [2021-04-16 02:34:43 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 74.8% [2021-04-16 02:34:43 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 74.82% [2021-04-16 02:34:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][0/1251] eta 1:02:52 lr 0.000643 time 3.0159 (3.0159) loss 3.8177 (3.8177) grad_norm 1.2409 (1.2409) [2021-04-16 02:34:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][10/1251] eta 0:11:26 lr 0.000643 time 0.4081 (0.5528) loss 2.6700 (3.6256) grad_norm 1.3633 (1.3470) [2021-04-16 02:34:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][20/1251] eta 0:08:41 lr 0.000643 time 0.2694 (0.4234) loss 2.9611 (3.6565) grad_norm 1.3741 (1.3574) [2021-04-16 02:34:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][30/1251] eta 0:07:42 lr 0.000643 time 0.2608 (0.3786) loss 3.6702 (3.6803) grad_norm 1.4091 (1.3352) [2021-04-16 02:34:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][40/1251] eta 0:07:09 lr 0.000643 time 0.2854 (0.3548) loss 3.8133 (3.6949) grad_norm 1.2697 (1.3416) [2021-04-16 02:35:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][50/1251] eta 0:06:48 lr 0.000643 time 0.2757 (0.3401) loss 2.8808 (3.6889) grad_norm 1.1832 (1.3415) [2021-04-16 02:35:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][60/1251] eta 0:06:33 lr 0.000643 time 0.2874 (0.3302) loss 3.4194 (3.6790) grad_norm 1.4894 (1.3540) [2021-04-16 02:35:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][70/1251] eta 0:06:20 lr 0.000643 time 0.2814 (0.3224) loss 3.4471 (3.6515) grad_norm 1.2893 (1.3518) [2021-04-16 02:35:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][80/1251] eta 0:06:11 lr 0.000643 time 0.3178 (0.3174) loss 4.2922 (3.6455) grad_norm 1.3610 (1.3465) [2021-04-16 02:35:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][90/1251] eta 0:06:02 lr 0.000643 time 0.2753 (0.3127) loss 2.3009 (3.6241) grad_norm 1.3809 (1.3532) [2021-04-16 02:35:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][100/1251] eta 0:05:56 lr 0.000643 time 0.3138 (0.3094) loss 3.9460 (3.6103) grad_norm 1.3270 (1.3543) [2021-04-16 02:35:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][110/1251] eta 0:05:49 lr 0.000643 time 0.2427 (0.3065) loss 3.5679 (3.6071) grad_norm 1.3902 (1.3511) [2021-04-16 02:35:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][120/1251] eta 0:05:43 lr 0.000643 time 0.2935 (0.3041) loss 2.6883 (3.5984) grad_norm 1.3196 (1.3544) [2021-04-16 02:35:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][130/1251] eta 0:05:39 lr 0.000643 time 0.2984 (0.3032) loss 3.8995 (3.6319) grad_norm 1.2293 (1.3519) [2021-04-16 02:35:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][140/1251] eta 0:05:34 lr 0.000643 time 0.2834 (0.3011) loss 2.8354 (3.6295) grad_norm 1.4308 (1.3549) [2021-04-16 02:35:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][150/1251] eta 0:05:29 lr 0.000643 time 0.2967 (0.2996) loss 2.3412 (3.6364) grad_norm 1.2428 (1.3540) [2021-04-16 02:35:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][160/1251] eta 0:05:25 lr 0.000642 time 0.2436 (0.2983) loss 2.6533 (3.6258) grad_norm 1.1844 (1.3588) [2021-04-16 02:35:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][170/1251] eta 0:05:21 lr 0.000642 time 0.2800 (0.2970) loss 4.1212 (3.6283) grad_norm 1.1197 (1.3548) [2021-04-16 02:35:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][180/1251] eta 0:05:16 lr 0.000642 time 0.2761 (0.2958) loss 4.0105 (3.6286) grad_norm 1.3832 (1.3581) [2021-04-16 02:35:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][190/1251] eta 0:05:12 lr 0.000642 time 0.2822 (0.2947) loss 3.9334 (3.6357) grad_norm 1.2411 (1.3630) [2021-04-16 02:35:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][200/1251] eta 0:05:08 lr 0.000642 time 0.2668 (0.2937) loss 4.1835 (3.6404) grad_norm 1.3170 (1.3656) [2021-04-16 02:35:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][210/1251] eta 0:05:05 lr 0.000642 time 0.3016 (0.2933) loss 3.9207 (3.6358) grad_norm 1.3238 (1.3628) [2021-04-16 02:35:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][220/1251] eta 0:05:01 lr 0.000642 time 0.2797 (0.2925) loss 3.7351 (3.6442) grad_norm 1.2462 (1.3585) [2021-04-16 02:35:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][230/1251] eta 0:04:57 lr 0.000642 time 0.2964 (0.2919) loss 4.2120 (3.6430) grad_norm 1.2835 (1.3573) [2021-04-16 02:35:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][240/1251] eta 0:04:54 lr 0.000642 time 0.2858 (0.2918) loss 4.3211 (3.6431) grad_norm 1.3693 (1.3555) [2021-04-16 02:35:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][250/1251] eta 0:04:51 lr 0.000642 time 0.2759 (0.2910) loss 3.0982 (3.6477) grad_norm 1.1327 (1.3573) [2021-04-16 02:35:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][260/1251] eta 0:04:48 lr 0.000642 time 0.2961 (0.2907) loss 3.5467 (3.6470) grad_norm 1.3078 (1.3569) [2021-04-16 02:36:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][270/1251] eta 0:04:44 lr 0.000642 time 0.2874 (0.2902) loss 4.0440 (3.6562) grad_norm 1.2520 (1.3579) [2021-04-16 02:36:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][280/1251] eta 0:04:41 lr 0.000642 time 0.2894 (0.2897) loss 3.7594 (3.6512) grad_norm 1.2510 (1.3570) [2021-04-16 02:36:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][290/1251] eta 0:04:37 lr 0.000642 time 0.2567 (0.2891) loss 3.7506 (3.6544) grad_norm 1.1902 (1.3572) [2021-04-16 02:36:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][300/1251] eta 0:04:34 lr 0.000642 time 0.2650 (0.2886) loss 3.8801 (3.6557) grad_norm 1.2764 (1.3579) [2021-04-16 02:36:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][310/1251] eta 0:04:31 lr 0.000642 time 0.2932 (0.2883) loss 2.8139 (3.6565) grad_norm 1.4129 (1.3589) [2021-04-16 02:36:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][320/1251] eta 0:04:28 lr 0.000642 time 0.2567 (0.2883) loss 4.0184 (3.6581) grad_norm 1.3801 (1.3619) [2021-04-16 02:36:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][330/1251] eta 0:04:25 lr 0.000642 time 0.2677 (0.2882) loss 3.9915 (3.6637) grad_norm 1.4816 (1.3616) [2021-04-16 02:36:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][340/1251] eta 0:04:22 lr 0.000642 time 0.3094 (0.2879) loss 3.6795 (3.6619) grad_norm 1.2921 (1.3619) [2021-04-16 02:36:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][350/1251] eta 0:04:19 lr 0.000642 time 0.2865 (0.2880) loss 4.1924 (3.6656) grad_norm 1.5832 (1.3659) [2021-04-16 02:36:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][360/1251] eta 0:04:16 lr 0.000642 time 0.2770 (0.2877) loss 4.2535 (3.6699) grad_norm 1.2716 (1.3710) [2021-04-16 02:36:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][370/1251] eta 0:04:13 lr 0.000642 time 0.2558 (0.2880) loss 3.1328 (3.6550) grad_norm 1.2837 (1.3731) [2021-04-16 02:36:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][380/1251] eta 0:04:10 lr 0.000642 time 0.2708 (0.2877) loss 3.2631 (3.6490) grad_norm 1.3518 (1.3721) [2021-04-16 02:36:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][390/1251] eta 0:04:07 lr 0.000642 time 0.2697 (0.2875) loss 3.7876 (3.6437) grad_norm 1.2698 (1.3718) [2021-04-16 02:36:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][400/1251] eta 0:04:04 lr 0.000642 time 0.2759 (0.2873) loss 3.7961 (3.6438) grad_norm 1.2305 (1.3738) [2021-04-16 02:36:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][410/1251] eta 0:04:01 lr 0.000641 time 0.2766 (0.2871) loss 3.8186 (3.6509) grad_norm 1.2761 (1.3715) [2021-04-16 02:36:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][420/1251] eta 0:03:58 lr 0.000641 time 0.2670 (0.2871) loss 3.7038 (3.6467) grad_norm 1.2090 (1.3690) [2021-04-16 02:36:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][430/1251] eta 0:03:55 lr 0.000641 time 0.2770 (0.2870) loss 3.5365 (3.6460) grad_norm 1.4600 (1.3698) [2021-04-16 02:36:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][440/1251] eta 0:03:52 lr 0.000641 time 0.2830 (0.2870) loss 3.5553 (3.6512) grad_norm 1.1721 (1.3684) [2021-04-16 02:36:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][450/1251] eta 0:03:49 lr 0.000641 time 0.2662 (0.2870) loss 4.0159 (3.6545) grad_norm 1.3414 (1.3674) [2021-04-16 02:36:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][460/1251] eta 0:03:46 lr 0.000641 time 0.2717 (0.2868) loss 2.9392 (3.6540) grad_norm 1.3240 (1.3677) [2021-04-16 02:36:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][470/1251] eta 0:03:43 lr 0.000641 time 0.2762 (0.2866) loss 3.2826 (3.6459) grad_norm 1.1952 (1.3656) [2021-04-16 02:37:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][480/1251] eta 0:03:40 lr 0.000641 time 0.2809 (0.2864) loss 3.0146 (3.6447) grad_norm 1.2462 (1.3644) [2021-04-16 02:37:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][490/1251] eta 0:03:37 lr 0.000641 time 0.2712 (0.2862) loss 3.7664 (3.6483) grad_norm 1.6394 (1.3647) [2021-04-16 02:37:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][500/1251] eta 0:03:34 lr 0.000641 time 0.2832 (0.2860) loss 2.6314 (3.6458) grad_norm 1.3659 (1.3643) [2021-04-16 02:37:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][510/1251] eta 0:03:31 lr 0.000641 time 0.2696 (0.2858) loss 3.7055 (3.6448) grad_norm 1.3858 (1.3630) [2021-04-16 02:37:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][520/1251] eta 0:03:28 lr 0.000641 time 0.2814 (0.2856) loss 3.7919 (3.6447) grad_norm 1.2295 (1.3626) [2021-04-16 02:37:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][530/1251] eta 0:03:25 lr 0.000641 time 0.2910 (0.2853) loss 4.1175 (3.6496) grad_norm 1.3115 (1.3633) [2021-04-16 02:37:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][540/1251] eta 0:03:22 lr 0.000641 time 0.2863 (0.2852) loss 4.0781 (3.6504) grad_norm 1.3746 (1.3621) [2021-04-16 02:37:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][550/1251] eta 0:03:20 lr 0.000641 time 0.2939 (0.2853) loss 4.1891 (3.6501) grad_norm 1.1341 (1.3617) [2021-04-16 02:37:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][560/1251] eta 0:03:16 lr 0.000641 time 0.2743 (0.2851) loss 3.6860 (3.6500) grad_norm 1.4783 (1.3621) [2021-04-16 02:37:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][570/1251] eta 0:03:13 lr 0.000641 time 0.2689 (0.2849) loss 4.0567 (3.6576) grad_norm 1.4048 (1.3618) [2021-04-16 02:37:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][580/1251] eta 0:03:11 lr 0.000641 time 0.3007 (0.2850) loss 3.0959 (3.6585) grad_norm 1.2835 (1.3613) [2021-04-16 02:37:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][590/1251] eta 0:03:08 lr 0.000641 time 0.2753 (0.2850) loss 3.9403 (3.6536) grad_norm 1.2133 (1.3624) [2021-04-16 02:37:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][600/1251] eta 0:03:05 lr 0.000641 time 0.3348 (0.2850) loss 2.7840 (3.6535) grad_norm 1.4239 (1.3632) [2021-04-16 02:37:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][610/1251] eta 0:03:02 lr 0.000641 time 0.2712 (0.2849) loss 3.5706 (3.6582) grad_norm 1.3641 (1.3618) [2021-04-16 02:37:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][620/1251] eta 0:02:59 lr 0.000641 time 0.3057 (0.2849) loss 4.1522 (3.6604) grad_norm 1.2736 (1.3620) [2021-04-16 02:37:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][630/1251] eta 0:02:56 lr 0.000641 time 0.2775 (0.2848) loss 4.5132 (3.6566) grad_norm 1.1889 (1.3608) [2021-04-16 02:37:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][640/1251] eta 0:02:53 lr 0.000641 time 0.2662 (0.2847) loss 3.8864 (3.6580) grad_norm 1.1967 (1.3600) [2021-04-16 02:37:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][650/1251] eta 0:02:51 lr 0.000641 time 0.2879 (0.2846) loss 3.7654 (3.6613) grad_norm 1.1518 (1.3591) [2021-04-16 02:37:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][660/1251] eta 0:02:48 lr 0.000640 time 0.2793 (0.2845) loss 4.1629 (3.6654) grad_norm 1.4587 (1.3590) [2021-04-16 02:37:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][670/1251] eta 0:02:45 lr 0.000640 time 0.2947 (0.2845) loss 3.8458 (3.6665) grad_norm 1.4765 (1.3586) [2021-04-16 02:37:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][680/1251] eta 0:02:42 lr 0.000640 time 0.2781 (0.2844) loss 3.5207 (3.6629) grad_norm 1.6192 (1.3605) [2021-04-16 02:38:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][690/1251] eta 0:02:39 lr 0.000640 time 0.2773 (0.2842) loss 3.8889 (3.6581) grad_norm 1.4505 (1.3609) [2021-04-16 02:38:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][700/1251] eta 0:02:36 lr 0.000640 time 0.2902 (0.2842) loss 3.3398 (3.6577) grad_norm 1.3510 (1.3618) [2021-04-16 02:38:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][710/1251] eta 0:02:33 lr 0.000640 time 0.2833 (0.2843) loss 2.6832 (3.6591) grad_norm 1.1217 (1.3615) [2021-04-16 02:38:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][720/1251] eta 0:02:30 lr 0.000640 time 0.2773 (0.2842) loss 3.9518 (3.6604) grad_norm 1.3291 (1.3612) [2021-04-16 02:38:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][730/1251] eta 0:02:28 lr 0.000640 time 0.2970 (0.2841) loss 3.7679 (3.6602) grad_norm 1.4263 (1.3618) [2021-04-16 02:38:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][740/1251] eta 0:02:25 lr 0.000640 time 0.2829 (0.2840) loss 3.6981 (3.6608) grad_norm 1.3699 (1.3631) [2021-04-16 02:38:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][750/1251] eta 0:02:22 lr 0.000640 time 0.2668 (0.2839) loss 4.1513 (3.6587) grad_norm 1.3600 (1.3628) [2021-04-16 02:38:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][760/1251] eta 0:02:19 lr 0.000640 time 0.2672 (0.2837) loss 4.2919 (3.6565) grad_norm 1.4214 (1.3626) [2021-04-16 02:38:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][770/1251] eta 0:02:16 lr 0.000640 time 0.3160 (0.2837) loss 3.4194 (3.6549) grad_norm 1.4972 (1.3630) [2021-04-16 02:38:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][780/1251] eta 0:02:13 lr 0.000640 time 0.2867 (0.2836) loss 3.9820 (3.6527) grad_norm 1.2616 (1.3628) [2021-04-16 02:38:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][790/1251] eta 0:02:10 lr 0.000640 time 0.2608 (0.2834) loss 2.8000 (3.6538) grad_norm 1.4206 (1.3625) [2021-04-16 02:38:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][800/1251] eta 0:02:07 lr 0.000640 time 0.2671 (0.2833) loss 3.7297 (3.6568) grad_norm 1.4031 (1.3636) [2021-04-16 02:38:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][810/1251] eta 0:02:04 lr 0.000640 time 0.2591 (0.2832) loss 3.6683 (3.6582) grad_norm 1.4174 (1.3642) [2021-04-16 02:38:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][820/1251] eta 0:02:02 lr 0.000640 time 0.2691 (0.2832) loss 3.1391 (3.6589) grad_norm 1.3282 (1.3637) [2021-04-16 02:38:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][830/1251] eta 0:01:59 lr 0.000640 time 0.2668 (0.2831) loss 4.6361 (3.6586) grad_norm 1.2134 (1.3644) [2021-04-16 02:38:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][840/1251] eta 0:01:56 lr 0.000640 time 0.2562 (0.2830) loss 3.4160 (3.6616) grad_norm 1.2820 (1.3647) [2021-04-16 02:38:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][850/1251] eta 0:01:53 lr 0.000640 time 0.2568 (0.2831) loss 4.1123 (3.6627) grad_norm 1.5931 (1.3660) [2021-04-16 02:38:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][860/1251] eta 0:01:50 lr 0.000640 time 0.2432 (0.2830) loss 3.9580 (3.6638) grad_norm 1.3603 (1.3661) [2021-04-16 02:38:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][870/1251] eta 0:01:47 lr 0.000640 time 0.2859 (0.2830) loss 3.9290 (3.6645) grad_norm 1.2654 (1.3656) [2021-04-16 02:38:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][880/1251] eta 0:01:44 lr 0.000640 time 0.2776 (0.2829) loss 3.5563 (3.6649) grad_norm 1.3921 (1.3648) [2021-04-16 02:38:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][890/1251] eta 0:01:42 lr 0.000640 time 0.2725 (0.2828) loss 4.2024 (3.6653) grad_norm 1.6231 (1.3644) [2021-04-16 02:38:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][900/1251] eta 0:01:39 lr 0.000640 time 0.3982 (0.2830) loss 3.0779 (3.6630) grad_norm 1.3592 (1.3640) [2021-04-16 02:39:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][910/1251] eta 0:01:36 lr 0.000639 time 0.2872 (0.2830) loss 4.0793 (3.6625) grad_norm 1.2939 (1.3645) [2021-04-16 02:39:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][920/1251] eta 0:01:33 lr 0.000639 time 0.2668 (0.2829) loss 2.9425 (3.6598) grad_norm 1.3848 (1.3649) [2021-04-16 02:39:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][930/1251] eta 0:01:30 lr 0.000639 time 0.2871 (0.2830) loss 3.9989 (3.6595) grad_norm 1.4365 (1.3659) [2021-04-16 02:39:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][940/1251] eta 0:01:28 lr 0.000639 time 0.2980 (0.2830) loss 3.8134 (3.6614) grad_norm 1.5147 (1.3652) [2021-04-16 02:39:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][950/1251] eta 0:01:25 lr 0.000639 time 0.2891 (0.2832) loss 4.0183 (3.6632) grad_norm 1.4315 (1.3662) [2021-04-16 02:39:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][960/1251] eta 0:01:22 lr 0.000639 time 0.2578 (0.2831) loss 4.1832 (3.6618) grad_norm 1.2672 (1.3657) [2021-04-16 02:39:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][970/1251] eta 0:01:19 lr 0.000639 time 0.3387 (0.2830) loss 3.2433 (3.6623) grad_norm 1.1726 (1.3653) [2021-04-16 02:39:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][980/1251] eta 0:01:16 lr 0.000639 time 0.2897 (0.2830) loss 3.5966 (3.6608) grad_norm 1.2733 (1.3643) [2021-04-16 02:39:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][990/1251] eta 0:01:13 lr 0.000639 time 0.2746 (0.2830) loss 4.5986 (3.6633) grad_norm 1.3389 (1.3639) [2021-04-16 02:39:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][1000/1251] eta 0:01:11 lr 0.000639 time 0.2742 (0.2830) loss 3.7907 (3.6660) grad_norm 1.6177 (1.3638) [2021-04-16 02:39:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][1010/1251] eta 0:01:08 lr 0.000639 time 0.2900 (0.2829) loss 3.5617 (3.6659) grad_norm 1.3623 (1.3632) [2021-04-16 02:39:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][1020/1251] eta 0:01:05 lr 0.000639 time 0.2708 (0.2828) loss 4.4036 (3.6673) grad_norm 1.4861 (1.3643) [2021-04-16 02:39:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][1030/1251] eta 0:01:02 lr 0.000639 time 0.2549 (0.2827) loss 3.0503 (3.6645) grad_norm 1.4255 (1.3644) [2021-04-16 02:39:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][1040/1251] eta 0:00:59 lr 0.000639 time 0.2662 (0.2827) loss 3.8771 (3.6624) grad_norm 1.2547 (1.3643) [2021-04-16 02:39:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][1050/1251] eta 0:00:56 lr 0.000639 time 0.2728 (0.2826) loss 3.3632 (3.6622) grad_norm 1.2571 (1.3644) [2021-04-16 02:39:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][1060/1251] eta 0:00:53 lr 0.000639 time 0.2903 (0.2825) loss 3.3041 (3.6598) grad_norm 1.2831 (1.3634) [2021-04-16 02:39:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][1070/1251] eta 0:00:51 lr 0.000639 time 0.2936 (0.2825) loss 4.1928 (3.6604) grad_norm 1.7033 (1.3642) [2021-04-16 02:39:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][1080/1251] eta 0:00:48 lr 0.000639 time 0.2939 (0.2824) loss 4.0231 (3.6623) grad_norm 1.3498 (1.3647) [2021-04-16 02:39:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][1090/1251] eta 0:00:45 lr 0.000639 time 0.2561 (0.2824) loss 4.0604 (3.6642) grad_norm 1.4120 (1.3653) [2021-04-16 02:39:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][1100/1251] eta 0:00:42 lr 0.000639 time 0.2813 (0.2824) loss 3.1716 (3.6638) grad_norm 1.5222 (1.3650) [2021-04-16 02:39:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][1110/1251] eta 0:00:39 lr 0.000639 time 0.3063 (0.2823) loss 3.9568 (3.6630) grad_norm 1.5030 (1.3658) [2021-04-16 02:40:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][1120/1251] eta 0:00:36 lr 0.000639 time 0.2719 (0.2823) loss 4.2481 (3.6645) grad_norm 1.3030 (1.3661) [2021-04-16 02:40:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][1130/1251] eta 0:00:34 lr 0.000639 time 0.2644 (0.2822) loss 3.3408 (3.6604) grad_norm 1.2458 (1.3655) [2021-04-16 02:40:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][1140/1251] eta 0:00:31 lr 0.000639 time 0.3001 (0.2821) loss 3.7493 (3.6588) grad_norm 1.4374 (1.3654) [2021-04-16 02:40:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][1150/1251] eta 0:00:28 lr 0.000639 time 0.2956 (0.2821) loss 4.0494 (3.6609) grad_norm 1.4954 (1.3664) [2021-04-16 02:40:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][1160/1251] eta 0:00:25 lr 0.000638 time 0.2825 (0.2821) loss 3.9193 (3.6623) grad_norm 1.2802 (1.3657) [2021-04-16 02:40:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][1170/1251] eta 0:00:22 lr 0.000638 time 0.2741 (0.2821) loss 4.3638 (3.6645) grad_norm 1.5389 (1.3656) [2021-04-16 02:40:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][1180/1251] eta 0:00:20 lr 0.000638 time 0.2965 (0.2821) loss 4.3115 (3.6650) grad_norm 1.1046 (1.3655) [2021-04-16 02:40:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][1190/1251] eta 0:00:17 lr 0.000638 time 0.2844 (0.2821) loss 2.5495 (3.6644) grad_norm 1.3799 (1.3657) [2021-04-16 02:40:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][1200/1251] eta 0:00:14 lr 0.000638 time 0.2787 (0.2821) loss 3.0430 (3.6646) grad_norm 1.5511 (1.3659) [2021-04-16 02:40:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][1210/1251] eta 0:00:11 lr 0.000638 time 0.2797 (0.2820) loss 3.9356 (3.6631) grad_norm 1.3478 (1.3660) [2021-04-16 02:40:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][1220/1251] eta 0:00:08 lr 0.000638 time 0.2718 (0.2820) loss 3.5077 (3.6627) grad_norm 1.5493 (1.3661) [2021-04-16 02:40:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][1230/1251] eta 0:00:05 lr 0.000638 time 0.2797 (0.2819) loss 3.7331 (3.6641) grad_norm 1.3711 (1.3660) [2021-04-16 02:40:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][1240/1251] eta 0:00:03 lr 0.000638 time 0.2478 (0.2818) loss 3.1101 (3.6639) grad_norm 1.1819 (1.3653) [2021-04-16 02:40:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [123/300][1250/1251] eta 0:00:00 lr 0.000638 time 0.2481 (0.2816) loss 4.1785 (3.6660) grad_norm 1.4152 (1.3645) [2021-04-16 02:40:38 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 123 training takes 0:05:55 [2021-04-16 02:40:38 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_123.pth saving...... [2021-04-16 02:40:58 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_123.pth saved !!! [2021-04-16 02:40:59 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.179 (1.179) Loss 1.0817 (1.0817) Acc@1 76.172 (76.172) Acc@5 92.383 (92.383) [2021-04-16 02:41:01 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.115 (0.215) Loss 1.1206 (1.0977) Acc@1 74.316 (74.547) Acc@5 92.676 (92.480) [2021-04-16 02:41:03 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.128 (0.241) Loss 1.0032 (1.0854) Acc@1 76.855 (74.637) Acc@5 93.457 (92.606) [2021-04-16 02:41:05 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.126 (0.214) Loss 1.0781 (1.0800) Acc@1 76.758 (74.773) Acc@5 91.992 (92.673) [2021-04-16 02:41:07 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.206) Loss 1.0619 (1.0778) Acc@1 75.586 (74.809) Acc@5 92.676 (92.721) [2021-04-16 02:41:11 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 74.854 Acc@5 92.726 [2021-04-16 02:41:11 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 74.9% [2021-04-16 02:41:11 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 74.85% [2021-04-16 02:41:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][0/1251] eta 2:19:53 lr 0.000638 time 6.7095 (6.7095) loss 3.3875 (3.3875) grad_norm 1.3911 (1.3911) [2021-04-16 02:41:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][10/1251] eta 0:17:51 lr 0.000638 time 0.2779 (0.8636) loss 4.0063 (3.5263) grad_norm 1.3907 (1.4169) [2021-04-16 02:41:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][20/1251] eta 0:12:09 lr 0.000638 time 0.2455 (0.5927) loss 3.2365 (3.4748) grad_norm 1.7297 (1.4325) [2021-04-16 02:41:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][30/1251] eta 0:10:05 lr 0.000638 time 0.2617 (0.4962) loss 2.7400 (3.4772) grad_norm 1.4653 (1.4090) [2021-04-16 02:41:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][40/1251] eta 0:08:59 lr 0.000638 time 0.4049 (0.4453) loss 4.0326 (3.5339) grad_norm 1.7170 (1.4165) [2021-04-16 02:41:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][50/1251] eta 0:08:14 lr 0.000638 time 0.2662 (0.4116) loss 3.9066 (3.5549) grad_norm 1.2988 (1.4034) [2021-04-16 02:41:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][60/1251] eta 0:07:46 lr 0.000638 time 0.2886 (0.3914) loss 4.3237 (3.6048) grad_norm 1.4704 (1.3972) [2021-04-16 02:41:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][70/1251] eta 0:07:24 lr 0.000638 time 0.3167 (0.3764) loss 2.5777 (3.5885) grad_norm 1.3679 (1.3970) [2021-04-16 02:41:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][80/1251] eta 0:07:07 lr 0.000638 time 0.2729 (0.3647) loss 3.8174 (3.5852) grad_norm 1.4230 (1.4027) [2021-04-16 02:41:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][90/1251] eta 0:06:52 lr 0.000638 time 0.2649 (0.3557) loss 3.6802 (3.5582) grad_norm 1.2424 (1.3951) [2021-04-16 02:41:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][100/1251] eta 0:06:40 lr 0.000638 time 0.2961 (0.3483) loss 3.5906 (3.5536) grad_norm 1.3849 (1.3963) [2021-04-16 02:41:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][110/1251] eta 0:06:30 lr 0.000638 time 0.3370 (0.3420) loss 2.9804 (3.5413) grad_norm 1.4607 (1.3909) [2021-04-16 02:41:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][120/1251] eta 0:06:20 lr 0.000638 time 0.2983 (0.3367) loss 2.5846 (3.5459) grad_norm 1.6722 (1.3879) [2021-04-16 02:41:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][130/1251] eta 0:06:13 lr 0.000638 time 0.2729 (0.3334) loss 4.7446 (3.5489) grad_norm 1.4806 (1.3876) [2021-04-16 02:41:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][140/1251] eta 0:06:07 lr 0.000638 time 0.2747 (0.3306) loss 2.0765 (3.5207) grad_norm 1.2664 (1.3860) [2021-04-16 02:42:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][150/1251] eta 0:06:00 lr 0.000638 time 0.3030 (0.3272) loss 4.0473 (3.5169) grad_norm 1.2625 (1.3785) [2021-04-16 02:42:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][160/1251] eta 0:05:54 lr 0.000637 time 0.2658 (0.3249) loss 4.1210 (3.4966) grad_norm 1.2891 (1.3755) [2021-04-16 02:42:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][170/1251] eta 0:05:48 lr 0.000637 time 0.2547 (0.3222) loss 4.0043 (3.5036) grad_norm 1.4856 (1.3744) [2021-04-16 02:42:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][180/1251] eta 0:05:42 lr 0.000637 time 0.2768 (0.3199) loss 2.9542 (3.5063) grad_norm 1.3760 (1.3752) [2021-04-16 02:42:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][190/1251] eta 0:05:36 lr 0.000637 time 0.2701 (0.3175) loss 3.7362 (3.5029) grad_norm 1.3079 (1.3732) [2021-04-16 02:42:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][200/1251] eta 0:05:32 lr 0.000637 time 0.2797 (0.3161) loss 3.4435 (3.5142) grad_norm 1.3604 (1.3754) [2021-04-16 02:42:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][210/1251] eta 0:05:27 lr 0.000637 time 0.2758 (0.3143) loss 3.9133 (3.5322) grad_norm 1.4997 (1.3793) [2021-04-16 02:42:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][220/1251] eta 0:05:22 lr 0.000637 time 0.2904 (0.3127) loss 3.9383 (3.5513) grad_norm 1.4902 (1.3804) [2021-04-16 02:42:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][230/1251] eta 0:05:18 lr 0.000637 time 0.3194 (0.3116) loss 3.9885 (3.5559) grad_norm 1.3990 (1.3802) [2021-04-16 02:42:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][240/1251] eta 0:05:14 lr 0.000637 time 0.2790 (0.3106) loss 4.0500 (3.5642) grad_norm 1.4811 (1.3795) [2021-04-16 02:42:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][250/1251] eta 0:05:09 lr 0.000637 time 0.2823 (0.3093) loss 2.7979 (3.5640) grad_norm 1.1827 (1.3770) [2021-04-16 02:42:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][260/1251] eta 0:05:05 lr 0.000637 time 0.2633 (0.3082) loss 3.6105 (3.5636) grad_norm 1.1654 (1.3707) [2021-04-16 02:42:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][270/1251] eta 0:05:01 lr 0.000637 time 0.2693 (0.3071) loss 3.4031 (3.5566) grad_norm 1.2576 (1.3707) [2021-04-16 02:42:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][280/1251] eta 0:04:57 lr 0.000637 time 0.2814 (0.3062) loss 3.6286 (3.5614) grad_norm 1.4018 (1.3718) [2021-04-16 02:42:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][290/1251] eta 0:04:53 lr 0.000637 time 0.2706 (0.3051) loss 4.2365 (3.5698) grad_norm 1.1443 (inf) [2021-04-16 02:42:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][300/1251] eta 0:04:49 lr 0.000637 time 0.2531 (0.3041) loss 3.8115 (3.5595) grad_norm 1.5451 (inf) [2021-04-16 02:42:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][310/1251] eta 0:04:45 lr 0.000637 time 0.2797 (0.3033) loss 3.1256 (3.5622) grad_norm 1.3163 (inf) [2021-04-16 02:42:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][320/1251] eta 0:04:41 lr 0.000637 time 0.3047 (0.3027) loss 2.6931 (3.5592) grad_norm 1.3463 (inf) [2021-04-16 02:42:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][330/1251] eta 0:04:38 lr 0.000637 time 0.3169 (0.3022) loss 4.0917 (3.5630) grad_norm 1.4307 (inf) [2021-04-16 02:42:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][340/1251] eta 0:04:34 lr 0.000637 time 0.2858 (0.3018) loss 3.9127 (3.5680) grad_norm 1.3509 (inf) [2021-04-16 02:42:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][350/1251] eta 0:04:31 lr 0.000637 time 0.4168 (0.3014) loss 4.1215 (3.5627) grad_norm 1.2642 (inf) [2021-04-16 02:42:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][360/1251] eta 0:04:27 lr 0.000637 time 0.2885 (0.3007) loss 3.9586 (3.5683) grad_norm 1.3945 (inf) [2021-04-16 02:43:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][370/1251] eta 0:04:24 lr 0.000637 time 0.2846 (0.3006) loss 2.4036 (3.5730) grad_norm 1.2662 (inf) [2021-04-16 02:43:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][380/1251] eta 0:04:21 lr 0.000637 time 0.2730 (0.3005) loss 3.5075 (3.5740) grad_norm 1.2020 (inf) [2021-04-16 02:43:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][390/1251] eta 0:04:18 lr 0.000637 time 0.3071 (0.3000) loss 3.5541 (3.5731) grad_norm 1.3514 (inf) [2021-04-16 02:43:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][400/1251] eta 0:04:14 lr 0.000637 time 0.2919 (0.2994) loss 2.4796 (3.5760) grad_norm 1.3527 (inf) [2021-04-16 02:43:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][410/1251] eta 0:04:11 lr 0.000636 time 0.2667 (0.2990) loss 3.8357 (3.5814) grad_norm 1.6513 (inf) [2021-04-16 02:43:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][420/1251] eta 0:04:08 lr 0.000636 time 0.2810 (0.2986) loss 3.7666 (3.5862) grad_norm 1.3771 (inf) [2021-04-16 02:43:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][430/1251] eta 0:04:04 lr 0.000636 time 0.2848 (0.2982) loss 2.9232 (3.5828) grad_norm 1.2515 (inf) [2021-04-16 02:43:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][440/1251] eta 0:04:01 lr 0.000636 time 0.2763 (0.2978) loss 2.4353 (3.5785) grad_norm 1.5027 (inf) [2021-04-16 02:43:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][450/1251] eta 0:03:58 lr 0.000636 time 0.2609 (0.2973) loss 4.1682 (3.5814) grad_norm 1.6654 (inf) [2021-04-16 02:43:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][460/1251] eta 0:03:54 lr 0.000636 time 0.2721 (0.2968) loss 4.4667 (3.5811) grad_norm 1.1931 (inf) [2021-04-16 02:43:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][470/1251] eta 0:03:51 lr 0.000636 time 0.2825 (0.2962) loss 2.8003 (3.5832) grad_norm 1.2680 (inf) [2021-04-16 02:43:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][480/1251] eta 0:03:48 lr 0.000636 time 0.3053 (0.2959) loss 3.9539 (3.5873) grad_norm 1.2008 (inf) [2021-04-16 02:43:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][490/1251] eta 0:03:44 lr 0.000636 time 0.2764 (0.2954) loss 4.5076 (3.5898) grad_norm 1.2217 (inf) [2021-04-16 02:43:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][500/1251] eta 0:03:41 lr 0.000636 time 0.2753 (0.2951) loss 4.2116 (3.5949) grad_norm 1.3280 (inf) [2021-04-16 02:43:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][510/1251] eta 0:03:38 lr 0.000636 time 0.2849 (0.2949) loss 3.9448 (3.5993) grad_norm 1.3623 (inf) [2021-04-16 02:43:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][520/1251] eta 0:03:35 lr 0.000636 time 0.2715 (0.2948) loss 2.8329 (3.6000) grad_norm 1.4010 (inf) [2021-04-16 02:43:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][530/1251] eta 0:03:32 lr 0.000636 time 0.2793 (0.2946) loss 4.1677 (3.5983) grad_norm 1.3634 (inf) [2021-04-16 02:43:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][540/1251] eta 0:03:29 lr 0.000636 time 0.2610 (0.2942) loss 3.6854 (3.5960) grad_norm 1.4133 (inf) [2021-04-16 02:43:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][550/1251] eta 0:03:26 lr 0.000636 time 0.2775 (0.2939) loss 4.8586 (3.6011) grad_norm 1.4234 (inf) [2021-04-16 02:43:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][560/1251] eta 0:03:22 lr 0.000636 time 0.3457 (0.2937) loss 4.4785 (3.6023) grad_norm 1.2331 (inf) [2021-04-16 02:43:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][570/1251] eta 0:03:19 lr 0.000636 time 0.2635 (0.2934) loss 4.2743 (3.6003) grad_norm 1.4991 (inf) [2021-04-16 02:44:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][580/1251] eta 0:03:16 lr 0.000636 time 0.2796 (0.2932) loss 2.7907 (3.6008) grad_norm 1.4017 (inf) [2021-04-16 02:44:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][590/1251] eta 0:03:13 lr 0.000636 time 0.2727 (0.2932) loss 3.2322 (3.6036) grad_norm 1.3684 (inf) [2021-04-16 02:44:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][600/1251] eta 0:03:10 lr 0.000636 time 0.2799 (0.2929) loss 3.5257 (3.5991) grad_norm 1.7046 (inf) [2021-04-16 02:44:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][610/1251] eta 0:03:07 lr 0.000636 time 0.3057 (0.2929) loss 3.8819 (3.6032) grad_norm 1.5803 (inf) [2021-04-16 02:44:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][620/1251] eta 0:03:04 lr 0.000636 time 0.2427 (0.2928) loss 4.2213 (3.6060) grad_norm 1.4924 (inf) [2021-04-16 02:44:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][630/1251] eta 0:03:01 lr 0.000636 time 0.2747 (0.2926) loss 4.3981 (3.6067) grad_norm 1.1948 (inf) [2021-04-16 02:44:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][640/1251] eta 0:02:58 lr 0.000636 time 0.2792 (0.2923) loss 3.4843 (3.6073) grad_norm 1.3104 (inf) [2021-04-16 02:44:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][650/1251] eta 0:02:55 lr 0.000636 time 0.2805 (0.2920) loss 4.0295 (3.6092) grad_norm 1.5758 (inf) [2021-04-16 02:44:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][660/1251] eta 0:02:52 lr 0.000635 time 0.2762 (0.2918) loss 3.6264 (3.6063) grad_norm 1.4988 (inf) [2021-04-16 02:44:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][670/1251] eta 0:02:49 lr 0.000635 time 0.2829 (0.2917) loss 4.1726 (3.6130) grad_norm 1.3334 (inf) [2021-04-16 02:44:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][680/1251] eta 0:02:46 lr 0.000635 time 0.2913 (0.2915) loss 3.0740 (3.6108) grad_norm 1.1464 (inf) [2021-04-16 02:44:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][690/1251] eta 0:02:43 lr 0.000635 time 0.2658 (0.2915) loss 4.1170 (3.6136) grad_norm 1.5429 (inf) [2021-04-16 02:44:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][700/1251] eta 0:02:40 lr 0.000635 time 0.3611 (0.2915) loss 3.9995 (3.6138) grad_norm 1.4437 (inf) [2021-04-16 02:44:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][710/1251] eta 0:02:37 lr 0.000635 time 0.2732 (0.2912) loss 3.3733 (3.6121) grad_norm 1.2581 (inf) [2021-04-16 02:44:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][720/1251] eta 0:02:34 lr 0.000635 time 0.2899 (0.2912) loss 3.6938 (3.6108) grad_norm 1.2683 (inf) [2021-04-16 02:44:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][730/1251] eta 0:02:31 lr 0.000635 time 0.2556 (0.2912) loss 3.8414 (3.6114) grad_norm 1.1452 (inf) [2021-04-16 02:44:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][740/1251] eta 0:02:28 lr 0.000635 time 0.2787 (0.2912) loss 4.1678 (3.6104) grad_norm 1.2478 (inf) [2021-04-16 02:44:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][750/1251] eta 0:02:25 lr 0.000635 time 0.2924 (0.2911) loss 4.0108 (3.6146) grad_norm 1.3824 (inf) [2021-04-16 02:44:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][760/1251] eta 0:02:22 lr 0.000635 time 0.2775 (0.2910) loss 3.7602 (3.6113) grad_norm 1.7020 (inf) [2021-04-16 02:44:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][770/1251] eta 0:02:19 lr 0.000635 time 0.2724 (0.2908) loss 3.3786 (3.6135) grad_norm 2.1461 (inf) [2021-04-16 02:44:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][780/1251] eta 0:02:16 lr 0.000635 time 0.2671 (0.2908) loss 4.3081 (3.6119) grad_norm 1.3548 (inf) [2021-04-16 02:45:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][790/1251] eta 0:02:13 lr 0.000635 time 0.2853 (0.2906) loss 2.9899 (3.6098) grad_norm 1.3096 (inf) [2021-04-16 02:45:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][800/1251] eta 0:02:11 lr 0.000635 time 0.2788 (0.2906) loss 4.4224 (3.6128) grad_norm 1.3611 (inf) [2021-04-16 02:45:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][810/1251] eta 0:02:08 lr 0.000635 time 0.2884 (0.2904) loss 2.9156 (3.6155) grad_norm 1.3521 (inf) [2021-04-16 02:45:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][820/1251] eta 0:02:05 lr 0.000635 time 0.2848 (0.2903) loss 3.7464 (3.6151) grad_norm 1.4053 (inf) [2021-04-16 02:45:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][830/1251] eta 0:02:02 lr 0.000635 time 0.2797 (0.2901) loss 4.2104 (3.6149) grad_norm 1.4324 (inf) [2021-04-16 02:45:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][840/1251] eta 0:01:59 lr 0.000635 time 0.2527 (0.2900) loss 4.5047 (3.6146) grad_norm 1.4664 (inf) [2021-04-16 02:45:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][850/1251] eta 0:01:56 lr 0.000635 time 0.2867 (0.2899) loss 3.4370 (3.6116) grad_norm 1.3805 (inf) [2021-04-16 02:45:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][860/1251] eta 0:01:53 lr 0.000635 time 0.3179 (0.2898) loss 4.0032 (3.6140) grad_norm 1.3740 (inf) [2021-04-16 02:45:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][870/1251] eta 0:01:50 lr 0.000635 time 0.2587 (0.2897) loss 2.6288 (3.6141) grad_norm 1.2277 (inf) [2021-04-16 02:45:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][880/1251] eta 0:01:47 lr 0.000635 time 0.2707 (0.2895) loss 4.2812 (3.6117) grad_norm 1.2559 (inf) [2021-04-16 02:45:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][890/1251] eta 0:01:44 lr 0.000635 time 0.2784 (0.2894) loss 3.8054 (3.6142) grad_norm 1.4970 (inf) [2021-04-16 02:45:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][900/1251] eta 0:01:41 lr 0.000635 time 0.2726 (0.2893) loss 3.4141 (3.6111) grad_norm 1.1209 (inf) [2021-04-16 02:45:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][910/1251] eta 0:01:38 lr 0.000634 time 0.2793 (0.2892) loss 3.8812 (3.6133) grad_norm 1.4691 (inf) [2021-04-16 02:45:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][920/1251] eta 0:01:35 lr 0.000634 time 0.2778 (0.2890) loss 3.4474 (3.6160) grad_norm 1.2423 (inf) [2021-04-16 02:45:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][930/1251] eta 0:01:32 lr 0.000634 time 0.2723 (0.2890) loss 3.8905 (3.6193) grad_norm 1.2005 (inf) [2021-04-16 02:45:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][940/1251] eta 0:01:29 lr 0.000634 time 0.2713 (0.2891) loss 3.7570 (3.6204) grad_norm 1.4890 (inf) [2021-04-16 02:45:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][950/1251] eta 0:01:26 lr 0.000634 time 0.2931 (0.2890) loss 3.0303 (3.6163) grad_norm 1.6006 (inf) [2021-04-16 02:45:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][960/1251] eta 0:01:24 lr 0.000634 time 0.2593 (0.2891) loss 3.7471 (3.6150) grad_norm 1.1814 (inf) [2021-04-16 02:45:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][970/1251] eta 0:01:21 lr 0.000634 time 0.2785 (0.2889) loss 2.6958 (3.6150) grad_norm 1.6605 (inf) [2021-04-16 02:45:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][980/1251] eta 0:01:18 lr 0.000634 time 0.3045 (0.2888) loss 2.3393 (3.6165) grad_norm 1.5197 (inf) [2021-04-16 02:45:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][990/1251] eta 0:01:15 lr 0.000634 time 0.2647 (0.2887) loss 4.1424 (3.6189) grad_norm 1.4020 (inf) [2021-04-16 02:46:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][1000/1251] eta 0:01:12 lr 0.000634 time 0.2839 (0.2886) loss 2.9560 (3.6182) grad_norm 1.3183 (inf) [2021-04-16 02:46:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][1010/1251] eta 0:01:09 lr 0.000634 time 0.2749 (0.2885) loss 3.4879 (3.6159) grad_norm 1.6852 (inf) [2021-04-16 02:46:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][1020/1251] eta 0:01:06 lr 0.000634 time 0.2803 (0.2884) loss 3.6979 (3.6139) grad_norm 1.5565 (inf) [2021-04-16 02:46:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][1030/1251] eta 0:01:03 lr 0.000634 time 0.2899 (0.2884) loss 2.9631 (3.6113) grad_norm 1.6762 (inf) [2021-04-16 02:46:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][1040/1251] eta 0:01:00 lr 0.000634 time 0.2641 (0.2883) loss 3.6519 (3.6118) grad_norm 1.2973 (inf) [2021-04-16 02:46:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][1050/1251] eta 0:00:57 lr 0.000634 time 0.2564 (0.2882) loss 3.4770 (3.6107) grad_norm 1.2331 (inf) [2021-04-16 02:46:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][1060/1251] eta 0:00:55 lr 0.000634 time 0.2707 (0.2881) loss 3.3203 (3.6103) grad_norm 1.4737 (inf) [2021-04-16 02:46:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][1070/1251] eta 0:00:52 lr 0.000634 time 0.2971 (0.2881) loss 3.5801 (3.6121) grad_norm 1.2911 (inf) [2021-04-16 02:46:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][1080/1251] eta 0:00:49 lr 0.000634 time 0.3048 (0.2880) loss 3.4308 (3.6134) grad_norm 1.4688 (inf) [2021-04-16 02:46:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][1090/1251] eta 0:00:46 lr 0.000634 time 0.2789 (0.2880) loss 3.6176 (3.6160) grad_norm 1.1846 (inf) [2021-04-16 02:46:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][1100/1251] eta 0:00:43 lr 0.000634 time 0.2430 (0.2879) loss 3.9734 (3.6155) grad_norm 1.3637 (inf) [2021-04-16 02:46:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][1110/1251] eta 0:00:40 lr 0.000634 time 0.2868 (0.2879) loss 3.8792 (3.6173) grad_norm 1.2817 (inf) [2021-04-16 02:46:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][1120/1251] eta 0:00:37 lr 0.000634 time 0.2767 (0.2878) loss 3.7168 (3.6167) grad_norm 1.2065 (inf) [2021-04-16 02:46:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][1130/1251] eta 0:00:34 lr 0.000634 time 0.3133 (0.2877) loss 3.8826 (3.6167) grad_norm 1.2709 (inf) [2021-04-16 02:46:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][1140/1251] eta 0:00:31 lr 0.000634 time 0.2875 (0.2877) loss 3.0492 (3.6173) grad_norm 1.2661 (inf) [2021-04-16 02:46:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][1150/1251] eta 0:00:29 lr 0.000634 time 0.2699 (0.2875) loss 4.4287 (3.6168) grad_norm 1.2979 (inf) [2021-04-16 02:46:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][1160/1251] eta 0:00:26 lr 0.000633 time 0.2796 (0.2875) loss 3.6272 (3.6166) grad_norm 1.1955 (inf) [2021-04-16 02:46:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][1170/1251] eta 0:00:23 lr 0.000633 time 0.2770 (0.2875) loss 3.2451 (3.6154) grad_norm 1.4794 (inf) [2021-04-16 02:46:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][1180/1251] eta 0:00:20 lr 0.000633 time 0.2667 (0.2875) loss 2.7512 (3.6160) grad_norm 1.3822 (inf) [2021-04-16 02:46:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][1190/1251] eta 0:00:17 lr 0.000633 time 0.2738 (0.2874) loss 3.6610 (3.6155) grad_norm 1.2646 (inf) [2021-04-16 02:46:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][1200/1251] eta 0:00:14 lr 0.000633 time 0.2553 (0.2874) loss 3.1164 (3.6153) grad_norm 1.5161 (inf) [2021-04-16 02:46:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][1210/1251] eta 0:00:11 lr 0.000633 time 0.2625 (0.2873) loss 3.6012 (3.6157) grad_norm 1.3746 (inf) [2021-04-16 02:47:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][1220/1251] eta 0:00:08 lr 0.000633 time 0.2866 (0.2872) loss 3.2020 (3.6160) grad_norm 1.5754 (inf) [2021-04-16 02:47:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][1230/1251] eta 0:00:06 lr 0.000633 time 0.2649 (0.2872) loss 3.8486 (3.6144) grad_norm 1.7892 (inf) [2021-04-16 02:47:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][1240/1251] eta 0:00:03 lr 0.000633 time 0.2483 (0.2870) loss 4.0928 (3.6157) grad_norm 1.3452 (inf) [2021-04-16 02:47:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [124/300][1250/1251] eta 0:00:00 lr 0.000633 time 0.2484 (0.2867) loss 3.1674 (3.6149) grad_norm 1.4666 (inf) [2021-04-16 02:47:13 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 124 training takes 0:06:02 [2021-04-16 02:47:13 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_124.pth saving...... [2021-04-16 02:47:31 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_124.pth saved !!! [2021-04-16 02:47:32 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.073 (1.073) Loss 1.0531 (1.0531) Acc@1 75.488 (75.488) Acc@5 92.773 (92.773) [2021-04-16 02:47:34 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.112 (0.277) Loss 1.0328 (1.0564) Acc@1 73.535 (74.920) Acc@5 93.750 (92.623) [2021-04-16 02:47:36 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.431 (0.230) Loss 1.1500 (1.0581) Acc@1 73.145 (75.005) Acc@5 91.504 (92.708) [2021-04-16 02:47:38 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.116 (0.237) Loss 1.0975 (1.0677) Acc@1 73.047 (74.792) Acc@5 92.773 (92.685) [2021-04-16 02:47:40 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.205 (0.219) Loss 1.0847 (1.0728) Acc@1 73.438 (74.709) Acc@5 93.164 (92.566) [2021-04-16 02:47:44 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 74.706 Acc@5 92.588 [2021-04-16 02:47:44 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 74.7% [2021-04-16 02:47:44 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 74.85% [2021-04-16 02:47:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][0/1251] eta 2:17:22 lr 0.000633 time 6.5885 (6.5885) loss 4.1888 (4.1888) grad_norm 1.3591 (1.3591) [2021-04-16 02:47:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][10/1251] eta 0:17:31 lr 0.000633 time 0.2728 (0.8470) loss 3.2131 (3.7453) grad_norm 1.2100 (1.4456) [2021-04-16 02:47:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][20/1251] eta 0:11:48 lr 0.000633 time 0.2944 (0.5753) loss 3.0211 (3.7006) grad_norm 1.2693 (1.3838) [2021-04-16 02:47:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][30/1251] eta 0:09:49 lr 0.000633 time 0.3134 (0.4828) loss 3.6606 (3.6610) grad_norm 1.3382 (1.3859) [2021-04-16 02:48:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][40/1251] eta 0:08:44 lr 0.000633 time 0.2908 (0.4331) loss 2.3947 (3.5999) grad_norm 1.3794 (1.3951) [2021-04-16 02:48:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][50/1251] eta 0:08:03 lr 0.000633 time 0.2952 (0.4026) loss 3.1772 (3.6157) grad_norm 1.4687 (1.4030) [2021-04-16 02:48:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][60/1251] eta 0:07:35 lr 0.000633 time 0.2562 (0.3821) loss 2.4524 (3.6173) grad_norm 1.3065 (1.4030) [2021-04-16 02:48:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][70/1251] eta 0:07:13 lr 0.000633 time 0.2725 (0.3672) loss 4.2106 (3.6116) grad_norm 1.3727 (1.4075) [2021-04-16 02:48:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][80/1251] eta 0:06:56 lr 0.000633 time 0.2520 (0.3560) loss 3.2879 (3.6108) grad_norm 1.2737 (1.4046) [2021-04-16 02:48:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][90/1251] eta 0:06:43 lr 0.000633 time 0.2968 (0.3475) loss 3.5644 (3.5865) grad_norm 1.2338 (1.3970) [2021-04-16 02:48:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][100/1251] eta 0:06:31 lr 0.000633 time 0.2899 (0.3404) loss 3.5907 (3.5953) grad_norm 1.4558 (1.4026) [2021-04-16 02:48:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][110/1251] eta 0:06:21 lr 0.000633 time 0.2725 (0.3346) loss 2.5012 (3.5858) grad_norm 1.3735 (1.4013) [2021-04-16 02:48:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][120/1251] eta 0:06:13 lr 0.000633 time 0.2843 (0.3299) loss 4.5560 (3.6081) grad_norm 1.7415 (1.3994) [2021-04-16 02:48:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][130/1251] eta 0:06:07 lr 0.000633 time 0.2996 (0.3274) loss 3.8396 (3.6242) grad_norm 1.3564 (1.3927) [2021-04-16 02:48:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][140/1251] eta 0:06:01 lr 0.000633 time 0.2721 (0.3254) loss 4.2389 (3.6401) grad_norm 1.4761 (1.3931) [2021-04-16 02:48:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][150/1251] eta 0:05:55 lr 0.000633 time 0.2545 (0.3231) loss 3.7112 (3.6272) grad_norm 1.5743 (1.3958) [2021-04-16 02:48:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][160/1251] eta 0:05:49 lr 0.000632 time 0.2573 (0.3200) loss 2.8342 (3.6229) grad_norm 1.2685 (1.3999) [2021-04-16 02:48:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][170/1251] eta 0:05:45 lr 0.000632 time 0.2689 (0.3194) loss 2.5109 (3.6165) grad_norm 1.4179 (1.4041) [2021-04-16 02:48:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][180/1251] eta 0:05:41 lr 0.000632 time 0.2669 (0.3187) loss 3.7759 (3.6309) grad_norm 1.3944 (1.4051) [2021-04-16 02:48:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][190/1251] eta 0:05:36 lr 0.000632 time 0.2700 (0.3167) loss 2.6975 (3.6217) grad_norm 1.2704 (1.4011) [2021-04-16 02:48:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][200/1251] eta 0:05:30 lr 0.000632 time 0.2670 (0.3147) loss 4.1930 (3.6279) grad_norm 1.3610 (1.3973) [2021-04-16 02:48:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][210/1251] eta 0:05:25 lr 0.000632 time 0.2809 (0.3128) loss 3.8342 (3.6295) grad_norm 1.2988 (1.3956) [2021-04-16 02:48:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][220/1251] eta 0:05:20 lr 0.000632 time 0.2889 (0.3111) loss 4.3598 (3.6314) grad_norm 1.1990 (1.3965) [2021-04-16 02:48:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][230/1251] eta 0:05:15 lr 0.000632 time 0.2620 (0.3095) loss 3.9335 (3.6211) grad_norm 1.2220 (1.3933) [2021-04-16 02:48:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][240/1251] eta 0:05:11 lr 0.000632 time 0.2583 (0.3083) loss 3.3393 (3.6235) grad_norm 1.3061 (1.3907) [2021-04-16 02:49:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][250/1251] eta 0:05:07 lr 0.000632 time 0.2893 (0.3076) loss 2.7057 (3.6121) grad_norm 1.2063 (1.3884) [2021-04-16 02:49:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][260/1251] eta 0:05:03 lr 0.000632 time 0.2906 (0.3067) loss 2.4657 (3.6146) grad_norm 1.2510 (1.3858) [2021-04-16 02:49:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][270/1251] eta 0:04:59 lr 0.000632 time 0.2959 (0.3058) loss 3.8313 (3.6216) grad_norm 1.6184 (1.3857) [2021-04-16 02:49:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][280/1251] eta 0:04:55 lr 0.000632 time 0.2698 (0.3047) loss 3.3917 (3.6156) grad_norm 1.2109 (1.3817) [2021-04-16 02:49:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][290/1251] eta 0:04:52 lr 0.000632 time 0.3038 (0.3041) loss 4.2031 (3.6130) grad_norm 1.5613 (1.3824) [2021-04-16 02:49:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][300/1251] eta 0:04:48 lr 0.000632 time 0.2879 (0.3036) loss 3.1020 (3.6082) grad_norm 1.4563 (1.3832) [2021-04-16 02:49:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][310/1251] eta 0:04:45 lr 0.000632 time 0.3186 (0.3031) loss 4.0836 (3.6108) grad_norm 1.5014 (1.3840) [2021-04-16 02:49:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][320/1251] eta 0:04:41 lr 0.000632 time 0.2565 (0.3022) loss 3.5897 (3.6046) grad_norm 1.4194 (1.3840) [2021-04-16 02:49:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][330/1251] eta 0:04:37 lr 0.000632 time 0.2704 (0.3015) loss 3.5205 (3.6083) grad_norm 1.4408 (1.3811) [2021-04-16 02:49:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][340/1251] eta 0:04:34 lr 0.000632 time 0.2759 (0.3009) loss 4.0689 (3.6074) grad_norm 1.5291 (1.3814) [2021-04-16 02:49:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][350/1251] eta 0:04:30 lr 0.000632 time 0.2849 (0.3002) loss 4.3946 (3.6036) grad_norm 1.3354 (1.3815) [2021-04-16 02:49:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][360/1251] eta 0:04:27 lr 0.000632 time 0.2801 (0.3000) loss 3.5903 (3.6019) grad_norm 1.2964 (1.3837) [2021-04-16 02:49:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][370/1251] eta 0:04:24 lr 0.000632 time 0.2732 (0.2997) loss 3.6674 (3.5985) grad_norm 1.5682 (1.3842) [2021-04-16 02:49:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][380/1251] eta 0:04:20 lr 0.000632 time 0.2695 (0.2996) loss 3.4966 (3.5982) grad_norm 1.2954 (1.3855) [2021-04-16 02:49:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][390/1251] eta 0:04:17 lr 0.000632 time 0.2903 (0.2992) loss 4.4531 (3.6118) grad_norm 1.3481 (1.3842) [2021-04-16 02:49:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][400/1251] eta 0:04:14 lr 0.000632 time 0.2761 (0.2986) loss 3.4960 (3.6170) grad_norm 1.4233 (1.3857) [2021-04-16 02:49:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][410/1251] eta 0:04:10 lr 0.000631 time 0.2796 (0.2980) loss 2.7944 (3.6212) grad_norm 1.5062 (1.3855) [2021-04-16 02:49:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][420/1251] eta 0:04:07 lr 0.000631 time 0.2880 (0.2976) loss 3.5791 (3.6224) grad_norm 1.2291 (1.3845) [2021-04-16 02:49:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][430/1251] eta 0:04:03 lr 0.000631 time 0.2845 (0.2971) loss 3.8107 (3.6264) grad_norm 1.5377 (1.3854) [2021-04-16 02:49:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][440/1251] eta 0:04:00 lr 0.000631 time 0.2892 (0.2967) loss 3.0637 (3.6269) grad_norm 1.4310 (1.3856) [2021-04-16 02:49:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][450/1251] eta 0:03:57 lr 0.000631 time 0.2689 (0.2963) loss 2.7052 (3.6237) grad_norm 1.4609 (1.3850) [2021-04-16 02:50:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][460/1251] eta 0:03:54 lr 0.000631 time 0.3051 (0.2958) loss 3.6637 (3.6284) grad_norm 1.3801 (1.3862) [2021-04-16 02:50:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][470/1251] eta 0:03:50 lr 0.000631 time 0.2768 (0.2954) loss 3.7803 (3.6250) grad_norm 1.2385 (1.3864) [2021-04-16 02:50:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][480/1251] eta 0:03:47 lr 0.000631 time 0.2945 (0.2951) loss 3.1980 (3.6228) grad_norm 1.5413 (1.3868) [2021-04-16 02:50:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][490/1251] eta 0:03:44 lr 0.000631 time 0.3083 (0.2947) loss 3.5894 (3.6210) grad_norm 1.3225 (1.3877) [2021-04-16 02:50:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][500/1251] eta 0:03:41 lr 0.000631 time 0.3014 (0.2944) loss 3.9509 (3.6246) grad_norm 1.7926 (1.3883) [2021-04-16 02:50:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][510/1251] eta 0:03:38 lr 0.000631 time 0.2588 (0.2943) loss 2.6035 (3.6254) grad_norm 1.3617 (1.3884) [2021-04-16 02:50:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][520/1251] eta 0:03:34 lr 0.000631 time 0.2820 (0.2940) loss 2.6406 (3.6236) grad_norm 1.4084 (1.3886) [2021-04-16 02:50:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][530/1251] eta 0:03:31 lr 0.000631 time 0.2929 (0.2937) loss 4.1923 (3.6280) grad_norm 1.2150 (1.3887) [2021-04-16 02:50:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][540/1251] eta 0:03:28 lr 0.000631 time 0.2658 (0.2936) loss 3.1815 (3.6308) grad_norm 1.4921 (1.3886) [2021-04-16 02:50:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][550/1251] eta 0:03:25 lr 0.000631 time 0.4219 (0.2935) loss 3.5826 (3.6318) grad_norm 1.3401 (1.3882) [2021-04-16 02:50:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][560/1251] eta 0:03:22 lr 0.000631 time 0.2909 (0.2932) loss 4.2095 (3.6287) grad_norm 1.4588 (1.3879) [2021-04-16 02:50:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][570/1251] eta 0:03:19 lr 0.000631 time 0.3800 (0.2934) loss 3.1415 (3.6282) grad_norm 1.1930 (1.3871) [2021-04-16 02:50:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][580/1251] eta 0:03:16 lr 0.000631 time 0.2676 (0.2931) loss 4.6488 (3.6286) grad_norm 1.2949 (1.3862) [2021-04-16 02:50:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][590/1251] eta 0:03:13 lr 0.000631 time 0.2885 (0.2928) loss 4.0442 (3.6265) grad_norm 1.3959 (inf) [2021-04-16 02:50:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][600/1251] eta 0:03:10 lr 0.000631 time 0.2697 (0.2926) loss 4.5035 (3.6286) grad_norm 1.4460 (inf) [2021-04-16 02:50:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][610/1251] eta 0:03:07 lr 0.000631 time 0.2732 (0.2923) loss 2.7204 (3.6271) grad_norm 1.7584 (inf) [2021-04-16 02:50:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][620/1251] eta 0:03:04 lr 0.000631 time 0.2539 (0.2921) loss 4.4901 (3.6287) grad_norm 1.9725 (inf) [2021-04-16 02:50:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][630/1251] eta 0:03:01 lr 0.000631 time 0.2686 (0.2919) loss 3.9326 (3.6306) grad_norm 1.5409 (inf) [2021-04-16 02:50:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][640/1251] eta 0:02:58 lr 0.000631 time 0.2565 (0.2916) loss 3.6009 (3.6293) grad_norm 1.3055 (inf) [2021-04-16 02:50:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][650/1251] eta 0:02:55 lr 0.000631 time 0.2887 (0.2914) loss 4.0537 (3.6315) grad_norm 1.6043 (inf) [2021-04-16 02:50:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][660/1251] eta 0:02:52 lr 0.000630 time 0.2746 (0.2912) loss 3.5808 (3.6285) grad_norm 1.3026 (inf) [2021-04-16 02:51:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][670/1251] eta 0:02:49 lr 0.000630 time 0.2751 (0.2910) loss 4.0521 (3.6294) grad_norm 1.1821 (inf) [2021-04-16 02:51:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][680/1251] eta 0:02:46 lr 0.000630 time 0.2791 (0.2908) loss 2.9659 (3.6257) grad_norm 1.5281 (inf) [2021-04-16 02:51:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][690/1251] eta 0:02:43 lr 0.000630 time 0.2689 (0.2907) loss 3.9040 (3.6260) grad_norm 1.2912 (inf) [2021-04-16 02:51:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][700/1251] eta 0:02:40 lr 0.000630 time 0.2833 (0.2905) loss 4.0798 (3.6233) grad_norm 1.6651 (inf) [2021-04-16 02:51:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][710/1251] eta 0:02:37 lr 0.000630 time 0.2745 (0.2904) loss 3.8996 (3.6247) grad_norm 1.3880 (inf) [2021-04-16 02:51:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][720/1251] eta 0:02:34 lr 0.000630 time 0.2656 (0.2903) loss 3.3617 (3.6285) grad_norm 1.3033 (inf) [2021-04-16 02:51:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][730/1251] eta 0:02:31 lr 0.000630 time 0.2473 (0.2903) loss 4.4058 (3.6268) grad_norm 1.5088 (inf) [2021-04-16 02:51:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][740/1251] eta 0:02:28 lr 0.000630 time 0.2650 (0.2904) loss 3.3270 (3.6224) grad_norm 1.4524 (inf) [2021-04-16 02:51:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][750/1251] eta 0:02:25 lr 0.000630 time 0.2582 (0.2902) loss 4.0061 (3.6224) grad_norm 1.3325 (inf) [2021-04-16 02:51:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][760/1251] eta 0:02:22 lr 0.000630 time 0.2737 (0.2900) loss 3.2594 (3.6197) grad_norm 1.3265 (inf) [2021-04-16 02:51:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][770/1251] eta 0:02:19 lr 0.000630 time 0.2753 (0.2899) loss 3.5600 (3.6202) grad_norm 1.3368 (inf) [2021-04-16 02:51:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][780/1251] eta 0:02:16 lr 0.000630 time 0.2607 (0.2897) loss 3.3045 (3.6190) grad_norm 1.4718 (inf) [2021-04-16 02:51:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][790/1251] eta 0:02:13 lr 0.000630 time 0.2796 (0.2895) loss 4.0650 (3.6224) grad_norm 1.4427 (inf) [2021-04-16 02:51:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][800/1251] eta 0:02:10 lr 0.000630 time 0.3018 (0.2893) loss 2.8384 (3.6204) grad_norm 1.4511 (inf) [2021-04-16 02:51:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][810/1251] eta 0:02:07 lr 0.000630 time 0.2834 (0.2892) loss 2.7153 (3.6191) grad_norm 1.3009 (inf) [2021-04-16 02:51:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][820/1251] eta 0:02:04 lr 0.000630 time 0.2686 (0.2889) loss 4.3026 (3.6202) grad_norm 1.3896 (inf) [2021-04-16 02:51:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][830/1251] eta 0:02:01 lr 0.000630 time 0.2740 (0.2889) loss 4.2328 (3.6209) grad_norm 1.4682 (inf) [2021-04-16 02:51:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][840/1251] eta 0:01:58 lr 0.000630 time 0.2938 (0.2887) loss 3.8175 (3.6210) grad_norm 1.3047 (inf) [2021-04-16 02:51:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][850/1251] eta 0:01:55 lr 0.000630 time 0.2979 (0.2886) loss 3.7004 (3.6210) grad_norm 1.5845 (inf) [2021-04-16 02:51:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][860/1251] eta 0:01:52 lr 0.000630 time 0.3418 (0.2885) loss 2.6490 (3.6217) grad_norm 1.2627 (inf) [2021-04-16 02:51:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][870/1251] eta 0:01:49 lr 0.000630 time 0.2603 (0.2884) loss 2.4675 (3.6211) grad_norm 1.4212 (inf) [2021-04-16 02:51:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][880/1251] eta 0:01:46 lr 0.000630 time 0.2582 (0.2884) loss 2.5269 (3.6196) grad_norm 1.3619 (inf) [2021-04-16 02:52:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][890/1251] eta 0:01:44 lr 0.000630 time 0.2737 (0.2883) loss 4.1462 (3.6167) grad_norm 1.7473 (inf) [2021-04-16 02:52:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][900/1251] eta 0:01:41 lr 0.000630 time 0.2972 (0.2882) loss 4.2176 (3.6157) grad_norm 1.5730 (inf) [2021-04-16 02:52:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][910/1251] eta 0:01:38 lr 0.000629 time 0.2694 (0.2880) loss 3.8818 (3.6151) grad_norm 1.4552 (inf) [2021-04-16 02:52:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][920/1251] eta 0:01:35 lr 0.000629 time 0.2726 (0.2878) loss 3.7503 (3.6165) grad_norm 1.6253 (inf) [2021-04-16 02:52:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][930/1251] eta 0:01:32 lr 0.000629 time 0.2615 (0.2879) loss 4.1387 (3.6163) grad_norm 1.1910 (inf) [2021-04-16 02:52:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][940/1251] eta 0:01:29 lr 0.000629 time 0.3903 (0.2879) loss 3.9331 (3.6135) grad_norm 1.1801 (inf) [2021-04-16 02:52:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][950/1251] eta 0:01:26 lr 0.000629 time 0.2701 (0.2878) loss 2.8451 (3.6152) grad_norm 1.3079 (inf) [2021-04-16 02:52:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][960/1251] eta 0:01:23 lr 0.000629 time 0.2640 (0.2878) loss 4.1164 (3.6145) grad_norm 1.4204 (inf) [2021-04-16 02:52:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][970/1251] eta 0:01:20 lr 0.000629 time 0.2943 (0.2878) loss 3.7975 (3.6170) grad_norm 1.1849 (inf) [2021-04-16 02:52:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][980/1251] eta 0:01:17 lr 0.000629 time 0.2647 (0.2877) loss 3.5929 (3.6157) grad_norm 1.3541 (inf) [2021-04-16 02:52:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][990/1251] eta 0:01:15 lr 0.000629 time 0.2712 (0.2875) loss 2.9472 (3.6142) grad_norm 1.2869 (inf) [2021-04-16 02:52:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][1000/1251] eta 0:01:12 lr 0.000629 time 0.2749 (0.2875) loss 4.0858 (3.6118) grad_norm 1.5076 (inf) [2021-04-16 02:52:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][1010/1251] eta 0:01:09 lr 0.000629 time 0.2944 (0.2873) loss 3.7056 (3.6104) grad_norm 1.3002 (inf) [2021-04-16 02:52:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][1020/1251] eta 0:01:06 lr 0.000629 time 0.2837 (0.2873) loss 2.9347 (3.6086) grad_norm 1.5395 (inf) [2021-04-16 02:52:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][1030/1251] eta 0:01:03 lr 0.000629 time 0.2880 (0.2872) loss 3.4142 (3.6091) grad_norm 1.3507 (inf) [2021-04-16 02:52:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][1040/1251] eta 0:01:00 lr 0.000629 time 0.2774 (0.2871) loss 3.3489 (3.6085) grad_norm 1.2923 (inf) [2021-04-16 02:52:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][1050/1251] eta 0:00:57 lr 0.000629 time 0.2874 (0.2870) loss 4.5657 (3.6083) grad_norm 1.5724 (inf) [2021-04-16 02:52:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][1060/1251] eta 0:00:54 lr 0.000629 time 0.2650 (0.2869) loss 4.4242 (3.6094) grad_norm 1.2141 (inf) [2021-04-16 02:52:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][1070/1251] eta 0:00:51 lr 0.000629 time 0.2604 (0.2868) loss 3.5157 (3.6088) grad_norm 1.2892 (inf) [2021-04-16 02:52:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][1080/1251] eta 0:00:49 lr 0.000629 time 0.2821 (0.2868) loss 3.3265 (3.6070) grad_norm 1.1817 (inf) [2021-04-16 02:52:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][1090/1251] eta 0:00:46 lr 0.000629 time 0.2624 (0.2866) loss 3.9807 (3.6085) grad_norm 1.4192 (inf) [2021-04-16 02:53:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][1100/1251] eta 0:00:43 lr 0.000629 time 0.2938 (0.2866) loss 4.3994 (3.6046) grad_norm 1.3233 (inf) [2021-04-16 02:53:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][1110/1251] eta 0:00:40 lr 0.000629 time 0.2754 (0.2866) loss 3.6774 (3.6036) grad_norm 1.3501 (inf) [2021-04-16 02:53:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][1120/1251] eta 0:00:37 lr 0.000629 time 0.2925 (0.2866) loss 3.8438 (3.6038) grad_norm 1.2193 (inf) [2021-04-16 02:53:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][1130/1251] eta 0:00:34 lr 0.000629 time 0.2643 (0.2866) loss 4.0571 (3.6049) grad_norm 1.4053 (inf) [2021-04-16 02:53:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][1140/1251] eta 0:00:31 lr 0.000629 time 0.3031 (0.2865) loss 3.5566 (3.6065) grad_norm 1.5469 (inf) [2021-04-16 02:53:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][1150/1251] eta 0:00:28 lr 0.000629 time 0.2625 (0.2867) loss 3.3070 (3.6065) grad_norm 1.3890 (inf) [2021-04-16 02:53:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][1160/1251] eta 0:00:26 lr 0.000628 time 0.2689 (0.2866) loss 3.7578 (3.6061) grad_norm 1.2955 (inf) [2021-04-16 02:53:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][1170/1251] eta 0:00:23 lr 0.000628 time 0.2897 (0.2865) loss 3.0086 (3.6043) grad_norm 1.4484 (inf) [2021-04-16 02:53:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][1180/1251] eta 0:00:20 lr 0.000628 time 0.2993 (0.2865) loss 3.5342 (3.6053) grad_norm 1.7715 (inf) [2021-04-16 02:53:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][1190/1251] eta 0:00:17 lr 0.000628 time 0.2825 (0.2864) loss 3.9316 (3.6043) grad_norm 1.4126 (inf) [2021-04-16 02:53:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][1200/1251] eta 0:00:14 lr 0.000628 time 0.3017 (0.2863) loss 3.8089 (3.6039) grad_norm 1.2813 (inf) [2021-04-16 02:53:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][1210/1251] eta 0:00:11 lr 0.000628 time 0.2893 (0.2862) loss 4.4815 (3.6055) grad_norm 1.4962 (inf) [2021-04-16 02:53:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][1220/1251] eta 0:00:08 lr 0.000628 time 0.2834 (0.2862) loss 2.7980 (3.6027) grad_norm 1.5947 (inf) [2021-04-16 02:53:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][1230/1251] eta 0:00:06 lr 0.000628 time 0.2760 (0.2862) loss 3.4621 (3.6046) grad_norm 1.4582 (inf) [2021-04-16 02:53:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][1240/1251] eta 0:00:03 lr 0.000628 time 0.2502 (0.2860) loss 4.3598 (3.6040) grad_norm 1.3518 (inf) [2021-04-16 02:53:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [125/300][1250/1251] eta 0:00:00 lr 0.000628 time 0.2477 (0.2857) loss 3.7161 (3.6030) grad_norm 1.1918 (inf) [2021-04-16 02:53:45 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 125 training takes 0:06:00 [2021-04-16 02:53:45 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_125.pth saving...... [2021-04-16 02:53:55 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_125.pth saved !!! [2021-04-16 02:53:56 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.102 (1.102) Loss 1.0451 (1.0451) Acc@1 74.902 (74.902) Acc@5 94.336 (94.336) [2021-04-16 02:53:57 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.133 (0.201) Loss 1.0750 (1.0626) Acc@1 73.828 (74.441) Acc@5 92.871 (93.004) [2021-04-16 02:54:00 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.156 (0.225) Loss 1.0312 (1.0628) Acc@1 75.879 (74.786) Acc@5 92.969 (92.890) [2021-04-16 02:54:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.092 (0.235) Loss 0.9894 (1.0614) Acc@1 76.465 (74.978) Acc@5 93.555 (92.802) [2021-04-16 02:54:04 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.225) Loss 1.0084 (1.0645) Acc@1 76.172 (74.812) Acc@5 93.652 (92.778) [2021-04-16 02:54:10 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 74.740 Acc@5 92.704 [2021-04-16 02:54:10 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 74.7% [2021-04-16 02:54:10 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 74.85% [2021-04-16 02:54:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][0/1251] eta 2:17:12 lr 0.000628 time 6.5809 (6.5809) loss 3.6604 (3.6604) grad_norm 1.2129 (1.2129) [2021-04-16 02:54:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][10/1251] eta 0:17:28 lr 0.000628 time 0.2789 (0.8446) loss 3.7594 (3.5445) grad_norm 1.1250 (1.4700) [2021-04-16 02:54:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][20/1251] eta 0:11:47 lr 0.000628 time 0.2655 (0.5747) loss 3.9649 (3.5514) grad_norm 1.4689 (1.4215) [2021-04-16 02:54:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][30/1251] eta 0:09:43 lr 0.000628 time 0.2620 (0.4779) loss 2.7275 (3.5200) grad_norm 1.1835 (1.4011) [2021-04-16 02:54:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][40/1251] eta 0:08:38 lr 0.000628 time 0.2863 (0.4283) loss 4.3730 (3.5369) grad_norm 1.2653 (1.3989) [2021-04-16 02:54:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][50/1251] eta 0:07:58 lr 0.000628 time 0.3004 (0.3983) loss 4.1501 (3.5612) grad_norm 1.5500 (1.4094) [2021-04-16 02:54:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][60/1251] eta 0:07:30 lr 0.000628 time 0.2794 (0.3781) loss 4.0731 (3.6050) grad_norm 1.3588 (1.4037) [2021-04-16 02:54:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][70/1251] eta 0:07:10 lr 0.000628 time 0.2743 (0.3647) loss 3.7692 (3.6030) grad_norm 1.2965 (1.4026) [2021-04-16 02:54:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][80/1251] eta 0:06:54 lr 0.000628 time 0.2624 (0.3539) loss 3.7351 (3.5924) grad_norm 1.4006 (1.4028) [2021-04-16 02:54:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][90/1251] eta 0:06:42 lr 0.000628 time 0.2441 (0.3463) loss 3.2373 (3.5857) grad_norm 1.6372 (1.4126) [2021-04-16 02:54:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][100/1251] eta 0:06:31 lr 0.000628 time 0.2799 (0.3397) loss 3.6409 (3.5987) grad_norm 1.6017 (1.4128) [2021-04-16 02:54:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][110/1251] eta 0:06:22 lr 0.000628 time 0.5060 (0.3355) loss 3.8486 (3.5990) grad_norm 1.2565 (1.4091) [2021-04-16 02:54:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][120/1251] eta 0:06:13 lr 0.000628 time 0.2673 (0.3302) loss 3.1309 (3.6122) grad_norm 1.3134 (1.4118) [2021-04-16 02:54:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][130/1251] eta 0:06:05 lr 0.000628 time 0.2603 (0.3257) loss 3.3126 (3.5851) grad_norm 1.2108 (1.4129) [2021-04-16 02:54:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][140/1251] eta 0:05:58 lr 0.000628 time 0.2703 (0.3226) loss 3.3928 (3.5718) grad_norm 1.2271 (1.4152) [2021-04-16 02:54:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][150/1251] eta 0:05:53 lr 0.000627 time 0.4556 (0.3214) loss 3.5301 (3.5744) grad_norm 1.7210 (1.4186) [2021-04-16 02:55:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][160/1251] eta 0:05:47 lr 0.000627 time 0.2837 (0.3184) loss 4.1549 (3.5768) grad_norm 1.4919 (1.4221) [2021-04-16 02:55:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][170/1251] eta 0:05:41 lr 0.000627 time 0.2761 (0.3162) loss 4.3973 (3.5861) grad_norm 1.4571 (1.4210) [2021-04-16 02:55:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][180/1251] eta 0:05:36 lr 0.000627 time 0.2595 (0.3139) loss 4.4032 (3.6078) grad_norm 1.5911 (1.4228) [2021-04-16 02:55:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][190/1251] eta 0:05:31 lr 0.000627 time 0.2708 (0.3121) loss 2.4659 (3.6048) grad_norm 1.1810 (1.4204) [2021-04-16 02:55:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][200/1251] eta 0:05:26 lr 0.000627 time 0.2761 (0.3103) loss 3.9354 (3.6105) grad_norm 1.1097 (1.4144) [2021-04-16 02:55:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][210/1251] eta 0:05:21 lr 0.000627 time 0.2870 (0.3087) loss 3.6579 (3.6109) grad_norm 1.7508 (1.4153) [2021-04-16 02:55:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][220/1251] eta 0:05:16 lr 0.000627 time 0.2606 (0.3072) loss 2.9031 (3.6079) grad_norm 1.3993 (1.4169) [2021-04-16 02:55:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][230/1251] eta 0:05:12 lr 0.000627 time 0.2947 (0.3060) loss 3.0270 (3.6017) grad_norm 1.7251 (1.4159) [2021-04-16 02:55:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][240/1251] eta 0:05:08 lr 0.000627 time 0.2953 (0.3048) loss 4.3848 (3.6033) grad_norm 1.3243 (1.4158) [2021-04-16 02:55:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][250/1251] eta 0:05:04 lr 0.000627 time 0.2831 (0.3039) loss 4.8722 (3.6154) grad_norm 1.4497 (1.4144) [2021-04-16 02:55:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][260/1251] eta 0:05:00 lr 0.000627 time 0.2850 (0.3029) loss 3.2649 (3.6280) grad_norm 1.4041 (1.4134) [2021-04-16 02:55:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][270/1251] eta 0:04:56 lr 0.000627 time 0.2591 (0.3018) loss 3.1276 (3.6240) grad_norm 1.3856 (1.4144) [2021-04-16 02:55:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][280/1251] eta 0:04:52 lr 0.000627 time 0.2760 (0.3010) loss 3.7935 (3.6136) grad_norm 1.4762 (1.4117) [2021-04-16 02:55:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][290/1251] eta 0:04:48 lr 0.000627 time 0.2976 (0.3003) loss 3.4320 (3.6140) grad_norm 1.4004 (1.4094) [2021-04-16 02:55:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][300/1251] eta 0:04:44 lr 0.000627 time 0.2618 (0.2994) loss 2.5388 (3.6025) grad_norm 1.3395 (1.4083) [2021-04-16 02:55:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][310/1251] eta 0:04:40 lr 0.000627 time 0.2867 (0.2986) loss 3.9145 (3.6030) grad_norm 1.2254 (1.4073) [2021-04-16 02:55:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][320/1251] eta 0:04:37 lr 0.000627 time 0.3097 (0.2984) loss 4.3584 (3.6098) grad_norm 1.6781 (1.4074) [2021-04-16 02:55:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][330/1251] eta 0:04:34 lr 0.000627 time 0.2520 (0.2980) loss 4.0182 (3.6123) grad_norm 1.8327 (1.4095) [2021-04-16 02:55:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][340/1251] eta 0:04:31 lr 0.000627 time 0.2536 (0.2976) loss 3.9206 (3.6015) grad_norm 1.3475 (1.4132) [2021-04-16 02:55:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][350/1251] eta 0:04:27 lr 0.000627 time 0.2829 (0.2971) loss 3.8549 (3.5979) grad_norm 1.3319 (1.4147) [2021-04-16 02:55:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][360/1251] eta 0:04:24 lr 0.000627 time 0.2663 (0.2970) loss 3.8793 (3.5907) grad_norm 1.2752 (1.4124) [2021-04-16 02:56:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][370/1251] eta 0:04:21 lr 0.000627 time 0.2706 (0.2967) loss 3.8747 (3.5878) grad_norm 1.7152 (1.4139) [2021-04-16 02:56:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][380/1251] eta 0:04:18 lr 0.000627 time 0.3096 (0.2963) loss 3.8355 (3.5871) grad_norm 1.4675 (1.4130) [2021-04-16 02:56:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][390/1251] eta 0:04:14 lr 0.000627 time 0.2825 (0.2960) loss 3.7781 (3.5918) grad_norm 1.4598 (1.4125) [2021-04-16 02:56:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][400/1251] eta 0:04:11 lr 0.000626 time 0.2775 (0.2954) loss 2.6486 (3.5949) grad_norm 1.4784 (1.4119) [2021-04-16 02:56:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][410/1251] eta 0:04:08 lr 0.000626 time 0.2881 (0.2950) loss 4.3448 (3.6102) grad_norm 1.0926 (1.4103) [2021-04-16 02:56:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][420/1251] eta 0:04:04 lr 0.000626 time 0.2952 (0.2945) loss 4.4086 (3.6159) grad_norm 1.3809 (1.4088) [2021-04-16 02:56:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][430/1251] eta 0:04:01 lr 0.000626 time 0.2878 (0.2941) loss 3.3222 (3.6121) grad_norm 1.4651 (1.4094) [2021-04-16 02:56:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][440/1251] eta 0:03:58 lr 0.000626 time 0.3035 (0.2937) loss 3.7720 (3.6176) grad_norm 1.3662 (1.4089) [2021-04-16 02:56:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][450/1251] eta 0:03:54 lr 0.000626 time 0.2750 (0.2933) loss 2.5908 (3.6125) grad_norm 1.8069 (1.4093) [2021-04-16 02:56:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][460/1251] eta 0:03:51 lr 0.000626 time 0.2681 (0.2930) loss 4.3281 (3.6147) grad_norm 1.7814 (1.4094) [2021-04-16 02:56:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][470/1251] eta 0:03:48 lr 0.000626 time 0.2675 (0.2926) loss 3.7980 (3.6176) grad_norm 1.2287 (1.4093) [2021-04-16 02:56:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][480/1251] eta 0:03:45 lr 0.000626 time 0.2873 (0.2924) loss 4.3875 (3.6186) grad_norm 1.4151 (1.4085) [2021-04-16 02:56:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][490/1251] eta 0:03:42 lr 0.000626 time 0.2560 (0.2920) loss 2.9045 (3.6171) grad_norm 1.5281 (1.4070) [2021-04-16 02:56:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][500/1251] eta 0:03:39 lr 0.000626 time 0.2705 (0.2917) loss 3.5466 (3.6161) grad_norm 1.1723 (1.4076) [2021-04-16 02:56:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][510/1251] eta 0:03:35 lr 0.000626 time 0.2599 (0.2914) loss 2.9662 (3.6153) grad_norm 1.5050 (1.4076) [2021-04-16 02:56:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][520/1251] eta 0:03:32 lr 0.000626 time 0.2550 (0.2910) loss 3.7901 (3.6160) grad_norm 1.5240 (1.4082) [2021-04-16 02:56:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][530/1251] eta 0:03:29 lr 0.000626 time 0.2793 (0.2908) loss 3.9966 (3.6149) grad_norm 1.4136 (1.4075) [2021-04-16 02:56:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][540/1251] eta 0:03:26 lr 0.000626 time 0.2897 (0.2908) loss 3.8021 (3.6162) grad_norm 1.7669 (1.4083) [2021-04-16 02:56:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][550/1251] eta 0:03:23 lr 0.000626 time 0.2787 (0.2906) loss 3.4258 (3.6142) grad_norm 1.3723 (1.4086) [2021-04-16 02:56:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][560/1251] eta 0:03:20 lr 0.000626 time 0.2882 (0.2903) loss 3.3305 (3.6140) grad_norm 1.8813 (1.4089) [2021-04-16 02:56:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][570/1251] eta 0:03:17 lr 0.000626 time 0.2650 (0.2902) loss 3.8328 (3.6164) grad_norm 1.2750 (1.4088) [2021-04-16 02:56:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][580/1251] eta 0:03:14 lr 0.000626 time 0.2505 (0.2900) loss 3.7593 (3.6161) grad_norm 1.3367 (1.4072) [2021-04-16 02:57:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][590/1251] eta 0:03:11 lr 0.000626 time 0.2701 (0.2897) loss 4.2095 (3.6180) grad_norm 1.4522 (1.4071) [2021-04-16 02:57:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][600/1251] eta 0:03:08 lr 0.000626 time 0.2814 (0.2896) loss 3.8672 (3.6185) grad_norm 1.2496 (1.4065) [2021-04-16 02:57:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][610/1251] eta 0:03:05 lr 0.000626 time 0.2737 (0.2892) loss 4.4089 (3.6201) grad_norm 1.2734 (1.4069) [2021-04-16 02:57:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][620/1251] eta 0:03:02 lr 0.000626 time 0.2685 (0.2890) loss 3.5962 (3.6199) grad_norm 1.2144 (1.4074) [2021-04-16 02:57:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][630/1251] eta 0:02:59 lr 0.000626 time 0.2781 (0.2888) loss 3.4070 (3.6164) grad_norm 1.2401 (1.4054) [2021-04-16 02:57:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][640/1251] eta 0:02:56 lr 0.000626 time 0.2599 (0.2886) loss 2.7430 (3.6193) grad_norm 1.3414 (1.4041) [2021-04-16 02:57:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][650/1251] eta 0:02:53 lr 0.000625 time 0.2724 (0.2884) loss 3.9643 (3.6172) grad_norm 1.1472 (1.4033) [2021-04-16 02:57:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][660/1251] eta 0:02:50 lr 0.000625 time 0.2961 (0.2882) loss 3.7896 (3.6162) grad_norm 1.4397 (1.4014) [2021-04-16 02:57:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][670/1251] eta 0:02:47 lr 0.000625 time 0.2647 (0.2881) loss 4.0738 (3.6189) grad_norm 1.4129 (1.4008) [2021-04-16 02:57:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][680/1251] eta 0:02:44 lr 0.000625 time 0.2595 (0.2880) loss 3.6724 (3.6231) grad_norm 1.3013 (1.4012) [2021-04-16 02:57:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][690/1251] eta 0:02:41 lr 0.000625 time 0.2762 (0.2880) loss 3.8063 (3.6216) grad_norm 1.1378 (1.4001) [2021-04-16 02:57:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][700/1251] eta 0:02:38 lr 0.000625 time 0.2850 (0.2879) loss 3.8048 (3.6219) grad_norm 1.4877 (1.3996) [2021-04-16 02:57:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][710/1251] eta 0:02:35 lr 0.000625 time 0.2649 (0.2877) loss 2.6089 (3.6196) grad_norm 1.5119 (1.3985) [2021-04-16 02:57:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][720/1251] eta 0:02:32 lr 0.000625 time 0.2930 (0.2876) loss 2.9582 (3.6199) grad_norm 1.3230 (1.3971) [2021-04-16 02:57:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][730/1251] eta 0:02:29 lr 0.000625 time 0.2965 (0.2877) loss 4.1145 (3.6175) grad_norm 1.4395 (1.3969) [2021-04-16 02:57:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][740/1251] eta 0:02:26 lr 0.000625 time 0.2859 (0.2875) loss 3.8721 (3.6170) grad_norm 1.3215 (1.3967) [2021-04-16 02:57:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][750/1251] eta 0:02:23 lr 0.000625 time 0.2726 (0.2874) loss 4.1830 (3.6196) grad_norm 1.3750 (1.3963) [2021-04-16 02:57:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][760/1251] eta 0:02:21 lr 0.000625 time 0.2876 (0.2873) loss 3.8872 (3.6229) grad_norm 1.2926 (1.3970) [2021-04-16 02:57:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][770/1251] eta 0:02:18 lr 0.000625 time 0.3065 (0.2872) loss 4.4725 (3.6232) grad_norm 1.4021 (1.3971) [2021-04-16 02:57:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][780/1251] eta 0:02:15 lr 0.000625 time 0.3103 (0.2871) loss 3.2915 (3.6232) grad_norm 1.4011 (1.3960) [2021-04-16 02:57:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][790/1251] eta 0:02:12 lr 0.000625 time 0.2669 (0.2869) loss 4.2685 (3.6208) grad_norm 1.3941 (1.3956) [2021-04-16 02:58:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][800/1251] eta 0:02:09 lr 0.000625 time 0.2789 (0.2867) loss 2.9722 (3.6192) grad_norm 1.2618 (1.3952) [2021-04-16 02:58:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][810/1251] eta 0:02:06 lr 0.000625 time 0.3037 (0.2866) loss 4.2528 (3.6211) grad_norm 1.3774 (1.3954) [2021-04-16 02:58:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][820/1251] eta 0:02:03 lr 0.000625 time 0.2615 (0.2865) loss 2.3377 (3.6197) grad_norm 1.9288 (1.3963) [2021-04-16 02:58:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][830/1251] eta 0:02:00 lr 0.000625 time 0.2865 (0.2864) loss 3.9140 (3.6230) grad_norm 1.6490 (1.3988) [2021-04-16 02:58:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][840/1251] eta 0:01:57 lr 0.000625 time 0.2606 (0.2863) loss 4.8139 (3.6220) grad_norm 1.4428 (1.3990) [2021-04-16 02:58:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][850/1251] eta 0:01:54 lr 0.000625 time 0.2639 (0.2862) loss 3.8341 (3.6235) grad_norm 1.3386 (1.3984) [2021-04-16 02:58:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][860/1251] eta 0:01:51 lr 0.000625 time 0.2750 (0.2861) loss 3.6418 (3.6224) grad_norm 1.3724 (1.3985) [2021-04-16 02:58:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][870/1251] eta 0:01:48 lr 0.000625 time 0.2648 (0.2859) loss 4.6694 (3.6231) grad_norm 1.5305 (1.3984) [2021-04-16 02:58:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][880/1251] eta 0:01:46 lr 0.000625 time 0.2669 (0.2858) loss 3.8709 (3.6222) grad_norm 1.3507 (1.3976) [2021-04-16 02:58:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][890/1251] eta 0:01:43 lr 0.000625 time 0.2660 (0.2859) loss 3.4244 (3.6217) grad_norm 1.2600 (1.3966) [2021-04-16 02:58:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][900/1251] eta 0:01:40 lr 0.000624 time 0.4483 (0.2859) loss 4.1730 (3.6198) grad_norm 1.4393 (1.3961) [2021-04-16 02:58:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][910/1251] eta 0:01:37 lr 0.000624 time 0.2841 (0.2858) loss 4.0590 (3.6213) grad_norm 1.3458 (1.3974) [2021-04-16 02:58:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][920/1251] eta 0:01:34 lr 0.000624 time 0.2613 (0.2858) loss 4.3650 (3.6246) grad_norm 1.4795 (1.3976) [2021-04-16 02:58:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][930/1251] eta 0:01:31 lr 0.000624 time 0.2844 (0.2856) loss 3.2725 (3.6233) grad_norm 1.1922 (1.3974) [2021-04-16 02:58:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][940/1251] eta 0:01:28 lr 0.000624 time 0.2935 (0.2856) loss 3.5253 (3.6235) grad_norm 1.2850 (1.3966) [2021-04-16 02:58:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][950/1251] eta 0:01:25 lr 0.000624 time 0.2772 (0.2856) loss 3.0924 (3.6235) grad_norm 1.2984 (1.3956) [2021-04-16 02:58:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][960/1251] eta 0:01:23 lr 0.000624 time 0.2779 (0.2855) loss 4.1097 (3.6240) grad_norm 1.3335 (1.3954) [2021-04-16 02:58:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][970/1251] eta 0:01:20 lr 0.000624 time 0.2917 (0.2855) loss 3.6896 (3.6259) grad_norm 1.4279 (1.3957) [2021-04-16 02:58:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][980/1251] eta 0:01:17 lr 0.000624 time 0.2818 (0.2854) loss 4.4558 (3.6298) grad_norm 1.5913 (1.3963) [2021-04-16 02:58:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][990/1251] eta 0:01:14 lr 0.000624 time 0.2618 (0.2853) loss 3.7162 (3.6291) grad_norm 1.3967 (1.3961) [2021-04-16 02:58:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][1000/1251] eta 0:01:11 lr 0.000624 time 0.3005 (0.2852) loss 3.2004 (3.6317) grad_norm 1.6416 (1.3966) [2021-04-16 02:58:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][1010/1251] eta 0:01:08 lr 0.000624 time 0.2894 (0.2851) loss 3.9587 (3.6315) grad_norm 1.3876 (1.3962) [2021-04-16 02:59:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][1020/1251] eta 0:01:05 lr 0.000624 time 0.3100 (0.2850) loss 4.0100 (3.6332) grad_norm 1.5676 (1.3965) [2021-04-16 02:59:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][1030/1251] eta 0:01:02 lr 0.000624 time 0.2600 (0.2849) loss 4.3607 (3.6326) grad_norm 1.6260 (1.3971) [2021-04-16 02:59:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][1040/1251] eta 0:01:00 lr 0.000624 time 0.2712 (0.2848) loss 4.1002 (3.6351) grad_norm 1.3729 (1.3967) [2021-04-16 02:59:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][1050/1251] eta 0:00:57 lr 0.000624 time 0.2750 (0.2848) loss 3.3970 (3.6356) grad_norm 1.3360 (1.3969) [2021-04-16 02:59:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][1060/1251] eta 0:00:54 lr 0.000624 time 0.2919 (0.2847) loss 4.1431 (3.6380) grad_norm 1.5143 (1.3975) [2021-04-16 02:59:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][1070/1251] eta 0:00:51 lr 0.000624 time 0.2651 (0.2845) loss 3.7165 (3.6335) grad_norm 1.2209 (1.3969) [2021-04-16 02:59:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][1080/1251] eta 0:00:48 lr 0.000624 time 0.3161 (0.2845) loss 2.9756 (3.6350) grad_norm 1.4716 (1.3964) [2021-04-16 02:59:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][1090/1251] eta 0:00:45 lr 0.000624 time 0.3015 (0.2845) loss 4.1480 (3.6365) grad_norm 1.4214 (1.3959) [2021-04-16 02:59:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][1100/1251] eta 0:00:42 lr 0.000624 time 0.2974 (0.2844) loss 2.8859 (3.6334) grad_norm 1.3620 (1.3953) [2021-04-16 02:59:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][1110/1251] eta 0:00:40 lr 0.000624 time 0.2766 (0.2845) loss 2.6366 (3.6307) grad_norm 1.1664 (1.3946) [2021-04-16 02:59:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][1120/1251] eta 0:00:37 lr 0.000624 time 0.2878 (0.2847) loss 3.6573 (3.6304) grad_norm 1.1350 (1.3933) [2021-04-16 02:59:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][1130/1251] eta 0:00:34 lr 0.000624 time 0.2617 (0.2846) loss 3.8818 (3.6328) grad_norm 1.3265 (1.3930) [2021-04-16 02:59:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][1140/1251] eta 0:00:31 lr 0.000624 time 0.2882 (0.2846) loss 3.2183 (3.6309) grad_norm 1.5376 (1.3932) [2021-04-16 02:59:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][1150/1251] eta 0:00:28 lr 0.000623 time 0.2566 (0.2845) loss 4.1337 (3.6301) grad_norm 1.4278 (1.3930) [2021-04-16 02:59:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][1160/1251] eta 0:00:25 lr 0.000623 time 0.2766 (0.2845) loss 2.5880 (3.6307) grad_norm 1.6215 (1.3933) [2021-04-16 02:59:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][1170/1251] eta 0:00:23 lr 0.000623 time 0.2703 (0.2844) loss 3.9123 (3.6315) grad_norm 1.5453 (1.3930) [2021-04-16 02:59:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][1180/1251] eta 0:00:20 lr 0.000623 time 0.2934 (0.2843) loss 4.0129 (3.6324) grad_norm 1.8321 (1.3941) [2021-04-16 02:59:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][1190/1251] eta 0:00:17 lr 0.000623 time 0.2809 (0.2843) loss 2.6376 (3.6338) grad_norm 1.3497 (1.3942) [2021-04-16 02:59:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][1200/1251] eta 0:00:14 lr 0.000623 time 0.2857 (0.2842) loss 4.1114 (3.6363) grad_norm 1.3220 (1.3935) [2021-04-16 02:59:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][1210/1251] eta 0:00:11 lr 0.000623 time 0.2739 (0.2842) loss 3.1634 (3.6356) grad_norm 1.2772 (1.3931) [2021-04-16 02:59:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][1220/1251] eta 0:00:08 lr 0.000623 time 0.2918 (0.2841) loss 3.7838 (3.6364) grad_norm 1.2315 (1.3925) [2021-04-16 03:00:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][1230/1251] eta 0:00:05 lr 0.000623 time 0.2810 (0.2841) loss 3.6783 (3.6368) grad_norm 1.1591 (1.3925) [2021-04-16 03:00:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][1240/1251] eta 0:00:03 lr 0.000623 time 0.2486 (0.2840) loss 2.7597 (3.6353) grad_norm 1.4053 (1.3925) [2021-04-16 03:00:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [126/300][1250/1251] eta 0:00:00 lr 0.000623 time 0.2486 (0.2837) loss 2.8562 (3.6343) grad_norm 1.4396 (1.3924) [2021-04-16 03:00:08 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 126 training takes 0:05:57 [2021-04-16 03:00:08 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_126.pth saving...... [2021-04-16 03:00:22 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_126.pth saved !!! [2021-04-16 03:00:23 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.171 (1.171) Loss 1.1017 (1.1017) Acc@1 74.414 (74.414) Acc@5 92.090 (92.090) [2021-04-16 03:00:25 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.158 (0.219) Loss 1.1409 (1.0986) Acc@1 74.512 (74.503) Acc@5 91.406 (92.551) [2021-04-16 03:00:28 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 1.364 (0.266) Loss 1.0511 (1.0892) Acc@1 77.637 (74.661) Acc@5 92.969 (92.741) [2021-04-16 03:00:29 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.120 (0.225) Loss 1.0518 (1.0921) Acc@1 76.660 (74.726) Acc@5 92.578 (92.710) [2021-04-16 03:00:31 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.219) Loss 1.1471 (1.0899) Acc@1 73.438 (74.774) Acc@5 91.992 (92.702) [2021-04-16 03:00:35 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 74.770 Acc@5 92.638 [2021-04-16 03:00:35 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 74.8% [2021-04-16 03:00:35 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 74.85% [2021-04-16 03:00:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][0/1251] eta 2:08:58 lr 0.000623 time 6.1862 (6.1862) loss 4.1640 (4.1640) grad_norm 1.2332 (1.2332) [2021-04-16 03:00:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][10/1251] eta 0:16:48 lr 0.000623 time 0.3335 (0.8129) loss 3.3209 (3.7168) grad_norm 1.4374 (1.4311) [2021-04-16 03:00:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][20/1251] eta 0:11:28 lr 0.000623 time 0.2661 (0.5594) loss 3.8211 (3.6341) grad_norm 1.2796 (1.3974) [2021-04-16 03:00:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][30/1251] eta 0:09:34 lr 0.000623 time 0.2671 (0.4704) loss 2.5692 (3.5529) grad_norm 1.2511 (1.3949) [2021-04-16 03:00:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][40/1251] eta 0:08:32 lr 0.000623 time 0.2836 (0.4233) loss 4.0292 (3.5600) grad_norm 1.1407 (1.4105) [2021-04-16 03:00:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][50/1251] eta 0:07:53 lr 0.000623 time 0.2788 (0.3941) loss 3.6406 (3.5444) grad_norm 1.3437 (1.4146) [2021-04-16 03:00:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][60/1251] eta 0:07:26 lr 0.000623 time 0.2990 (0.3752) loss 3.8439 (3.5535) grad_norm 1.4628 (1.4158) [2021-04-16 03:01:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][70/1251] eta 0:07:07 lr 0.000623 time 0.2841 (0.3618) loss 4.1848 (3.5766) grad_norm 1.6083 (1.4100) [2021-04-16 03:01:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][80/1251] eta 0:06:51 lr 0.000623 time 0.2810 (0.3517) loss 3.5961 (3.5618) grad_norm 1.2795 (1.4113) [2021-04-16 03:01:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][90/1251] eta 0:06:40 lr 0.000623 time 0.2589 (0.3453) loss 2.2336 (3.5609) grad_norm 1.3340 (1.4078) [2021-04-16 03:01:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][100/1251] eta 0:06:32 lr 0.000623 time 0.2688 (0.3411) loss 4.0680 (3.5741) grad_norm 1.4444 (1.4007) [2021-04-16 03:01:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][110/1251] eta 0:06:22 lr 0.000623 time 0.2857 (0.3354) loss 3.9460 (3.5728) grad_norm 1.2780 (1.3952) [2021-04-16 03:01:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][120/1251] eta 0:06:14 lr 0.000623 time 0.2535 (0.3311) loss 2.7325 (3.5594) grad_norm 1.1764 (1.3909) [2021-04-16 03:01:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][130/1251] eta 0:06:06 lr 0.000623 time 0.2711 (0.3271) loss 3.5498 (3.5474) grad_norm 1.4384 (1.3895) [2021-04-16 03:01:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][140/1251] eta 0:06:00 lr 0.000623 time 0.2787 (0.3247) loss 3.8871 (3.5580) grad_norm 1.2301 (1.3865) [2021-04-16 03:01:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][150/1251] eta 0:05:54 lr 0.000622 time 0.2764 (0.3224) loss 3.4768 (3.5749) grad_norm 1.3963 (1.3833) [2021-04-16 03:01:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][160/1251] eta 0:05:49 lr 0.000622 time 0.2739 (0.3205) loss 2.3905 (3.5675) grad_norm 1.3967 (1.3830) [2021-04-16 03:01:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][170/1251] eta 0:05:44 lr 0.000622 time 0.2457 (0.3183) loss 3.6714 (3.5789) grad_norm 1.4571 (1.3862) [2021-04-16 03:01:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][180/1251] eta 0:05:38 lr 0.000622 time 0.2629 (0.3161) loss 2.7629 (3.5606) grad_norm 1.1922 (1.3842) [2021-04-16 03:01:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][190/1251] eta 0:05:33 lr 0.000622 time 0.2862 (0.3140) loss 3.6831 (3.5677) grad_norm 1.3015 (1.3884) [2021-04-16 03:01:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][200/1251] eta 0:05:28 lr 0.000622 time 0.2766 (0.3121) loss 4.0015 (3.5690) grad_norm 1.2372 (1.3870) [2021-04-16 03:01:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][210/1251] eta 0:05:23 lr 0.000622 time 0.2740 (0.3104) loss 3.4683 (3.5687) grad_norm 1.3474 (1.3894) [2021-04-16 03:01:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][220/1251] eta 0:05:18 lr 0.000622 time 0.2716 (0.3092) loss 4.1396 (3.5686) grad_norm 1.4547 (1.3902) [2021-04-16 03:01:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][230/1251] eta 0:05:14 lr 0.000622 time 0.2725 (0.3079) loss 4.4282 (3.5600) grad_norm 1.3001 (1.3922) [2021-04-16 03:01:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][240/1251] eta 0:05:10 lr 0.000622 time 0.2840 (0.3066) loss 4.0312 (3.5649) grad_norm 1.5051 (1.3954) [2021-04-16 03:01:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][250/1251] eta 0:05:05 lr 0.000622 time 0.2930 (0.3056) loss 3.0950 (3.5595) grad_norm 1.2916 (1.3979) [2021-04-16 03:01:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][260/1251] eta 0:05:01 lr 0.000622 time 0.2888 (0.3047) loss 3.9601 (3.5557) grad_norm 1.3116 (1.3955) [2021-04-16 03:01:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][270/1251] eta 0:04:57 lr 0.000622 time 0.2888 (0.3037) loss 4.2333 (3.5676) grad_norm 1.1934 (1.3950) [2021-04-16 03:02:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][280/1251] eta 0:04:54 lr 0.000622 time 0.2573 (0.3031) loss 2.9372 (3.5649) grad_norm 1.6125 (1.3939) [2021-04-16 03:02:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][290/1251] eta 0:04:50 lr 0.000622 time 0.2728 (0.3020) loss 4.0900 (3.5702) grad_norm 1.4602 (1.3925) [2021-04-16 03:02:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][300/1251] eta 0:04:46 lr 0.000622 time 0.2609 (0.3012) loss 3.7344 (3.5709) grad_norm 1.1175 (1.3901) [2021-04-16 03:02:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][310/1251] eta 0:04:42 lr 0.000622 time 0.2668 (0.3005) loss 4.5252 (3.5815) grad_norm 1.4548 (1.3900) [2021-04-16 03:02:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][320/1251] eta 0:04:39 lr 0.000622 time 0.2680 (0.3001) loss 3.2147 (3.5761) grad_norm 1.2872 (1.3918) [2021-04-16 03:02:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][330/1251] eta 0:04:35 lr 0.000622 time 0.2654 (0.2996) loss 3.5044 (3.5782) grad_norm 1.3172 (1.3913) [2021-04-16 03:02:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][340/1251] eta 0:04:32 lr 0.000622 time 0.2792 (0.2992) loss 4.7652 (3.5803) grad_norm 1.1981 (1.3885) [2021-04-16 03:02:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][350/1251] eta 0:04:28 lr 0.000622 time 0.2708 (0.2985) loss 4.0610 (3.5819) grad_norm 1.3642 (1.3870) [2021-04-16 03:02:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][360/1251] eta 0:04:25 lr 0.000622 time 0.2670 (0.2980) loss 3.7254 (3.5856) grad_norm 1.4345 (1.3859) [2021-04-16 03:02:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][370/1251] eta 0:04:22 lr 0.000622 time 0.2717 (0.2975) loss 4.1474 (3.5866) grad_norm 1.4316 (1.3848) [2021-04-16 03:02:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][380/1251] eta 0:04:18 lr 0.000622 time 0.3089 (0.2970) loss 4.2475 (3.5926) grad_norm 1.7761 (1.3853) [2021-04-16 03:02:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][390/1251] eta 0:04:15 lr 0.000622 time 0.2928 (0.2964) loss 3.7213 (3.5979) grad_norm 1.3146 (1.3847) [2021-04-16 03:02:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][400/1251] eta 0:04:11 lr 0.000621 time 0.2715 (0.2959) loss 3.6029 (3.6021) grad_norm 1.2780 (1.3863) [2021-04-16 03:02:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][410/1251] eta 0:04:08 lr 0.000621 time 0.2962 (0.2956) loss 2.5376 (3.5932) grad_norm 1.1298 (1.3882) [2021-04-16 03:02:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][420/1251] eta 0:04:05 lr 0.000621 time 0.2953 (0.2951) loss 3.4810 (3.5969) grad_norm 1.2462 (1.3908) [2021-04-16 03:02:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][430/1251] eta 0:04:01 lr 0.000621 time 0.2651 (0.2946) loss 3.6244 (3.6010) grad_norm 1.3772 (1.3903) [2021-04-16 03:02:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][440/1251] eta 0:03:58 lr 0.000621 time 0.2846 (0.2947) loss 4.0302 (3.6067) grad_norm 1.6760 (1.3906) [2021-04-16 03:02:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][450/1251] eta 0:03:55 lr 0.000621 time 0.2879 (0.2944) loss 3.9964 (3.6091) grad_norm 1.6177 (1.3924) [2021-04-16 03:02:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][460/1251] eta 0:03:52 lr 0.000621 time 0.2657 (0.2940) loss 2.8000 (3.6074) grad_norm 1.3918 (1.3937) [2021-04-16 03:02:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][470/1251] eta 0:03:49 lr 0.000621 time 0.2838 (0.2938) loss 4.1986 (3.6078) grad_norm 1.3877 (1.3938) [2021-04-16 03:02:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][480/1251] eta 0:03:46 lr 0.000621 time 0.2907 (0.2935) loss 2.4396 (3.6018) grad_norm 1.3831 (1.3951) [2021-04-16 03:02:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][490/1251] eta 0:03:43 lr 0.000621 time 0.2866 (0.2932) loss 3.9679 (3.6065) grad_norm 1.2905 (1.3946) [2021-04-16 03:03:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][500/1251] eta 0:03:39 lr 0.000621 time 0.2618 (0.2928) loss 3.7173 (3.6083) grad_norm 1.3245 (1.3941) [2021-04-16 03:03:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][510/1251] eta 0:03:36 lr 0.000621 time 0.2869 (0.2924) loss 3.6304 (3.6136) grad_norm 1.3171 (1.3952) [2021-04-16 03:03:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][520/1251] eta 0:03:33 lr 0.000621 time 0.2523 (0.2920) loss 2.5305 (3.6111) grad_norm 1.4209 (1.3959) [2021-04-16 03:03:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][530/1251] eta 0:03:30 lr 0.000621 time 0.2646 (0.2917) loss 2.8466 (3.6157) grad_norm 1.4994 (1.3939) [2021-04-16 03:03:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][540/1251] eta 0:03:27 lr 0.000621 time 0.2562 (0.2914) loss 4.1674 (3.6171) grad_norm 1.3379 (1.3921) [2021-04-16 03:03:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][550/1251] eta 0:03:24 lr 0.000621 time 0.2757 (0.2912) loss 4.1186 (3.6178) grad_norm 1.2703 (1.3937) [2021-04-16 03:03:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][560/1251] eta 0:03:21 lr 0.000621 time 0.3090 (0.2909) loss 4.0809 (3.6136) grad_norm 1.3833 (1.3943) [2021-04-16 03:03:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][570/1251] eta 0:03:18 lr 0.000621 time 0.2700 (0.2912) loss 4.0234 (3.6145) grad_norm 1.3455 (1.3939) [2021-04-16 03:03:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][580/1251] eta 0:03:15 lr 0.000621 time 0.2923 (0.2912) loss 3.3767 (3.6118) grad_norm 1.3561 (1.3932) [2021-04-16 03:03:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][590/1251] eta 0:03:12 lr 0.000621 time 0.2916 (0.2911) loss 3.9549 (3.6104) grad_norm 1.2917 (1.3930) [2021-04-16 03:03:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][600/1251] eta 0:03:09 lr 0.000621 time 0.2697 (0.2908) loss 3.9515 (3.6126) grad_norm 1.3888 (1.3934) [2021-04-16 03:03:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][610/1251] eta 0:03:06 lr 0.000621 time 0.2920 (0.2906) loss 3.4470 (3.6154) grad_norm 1.2908 (1.3937) [2021-04-16 03:03:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][620/1251] eta 0:03:03 lr 0.000621 time 0.2742 (0.2904) loss 3.1001 (3.6074) grad_norm 1.2825 (1.3942) [2021-04-16 03:03:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][630/1251] eta 0:03:00 lr 0.000621 time 0.3041 (0.2902) loss 3.5450 (3.6054) grad_norm 1.5195 (1.3944) [2021-04-16 03:03:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][640/1251] eta 0:02:57 lr 0.000620 time 0.3072 (0.2900) loss 4.3606 (3.6068) grad_norm 1.6491 (1.3940) [2021-04-16 03:03:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][650/1251] eta 0:02:54 lr 0.000620 time 0.2881 (0.2898) loss 3.7653 (3.6067) grad_norm 1.6296 (1.3948) [2021-04-16 03:03:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][660/1251] eta 0:02:51 lr 0.000620 time 0.2619 (0.2895) loss 4.2521 (3.6083) grad_norm 1.8674 (1.3953) [2021-04-16 03:03:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][670/1251] eta 0:02:48 lr 0.000620 time 0.2725 (0.2895) loss 2.8198 (3.6064) grad_norm 1.3424 (1.3964) [2021-04-16 03:03:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][680/1251] eta 0:02:45 lr 0.000620 time 0.2825 (0.2895) loss 2.5592 (3.6043) grad_norm 1.5216 (1.3984) [2021-04-16 03:03:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][690/1251] eta 0:02:42 lr 0.000620 time 0.2981 (0.2895) loss 3.6799 (3.6069) grad_norm 1.3830 (1.3980) [2021-04-16 03:03:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][700/1251] eta 0:02:39 lr 0.000620 time 0.2838 (0.2893) loss 3.5067 (3.6099) grad_norm 1.3617 (1.3975) [2021-04-16 03:04:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][710/1251] eta 0:02:36 lr 0.000620 time 0.2672 (0.2892) loss 3.3383 (3.6100) grad_norm 1.4440 (1.3975) [2021-04-16 03:04:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][720/1251] eta 0:02:33 lr 0.000620 time 0.2774 (0.2892) loss 3.5430 (3.6106) grad_norm 1.3619 (1.3970) [2021-04-16 03:04:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][730/1251] eta 0:02:30 lr 0.000620 time 0.2688 (0.2890) loss 3.1084 (3.6115) grad_norm 1.3286 (1.3969) [2021-04-16 03:04:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][740/1251] eta 0:02:27 lr 0.000620 time 0.2821 (0.2889) loss 3.4540 (3.6089) grad_norm 1.3006 (1.3967) [2021-04-16 03:04:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][750/1251] eta 0:02:24 lr 0.000620 time 0.2620 (0.2887) loss 4.3096 (3.6074) grad_norm 1.5412 (1.3976) [2021-04-16 03:04:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][760/1251] eta 0:02:21 lr 0.000620 time 0.2677 (0.2885) loss 4.4593 (3.6097) grad_norm 1.4551 (1.3992) [2021-04-16 03:04:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][770/1251] eta 0:02:18 lr 0.000620 time 0.2774 (0.2884) loss 4.3400 (3.6125) grad_norm 1.6718 (1.3992) [2021-04-16 03:04:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][780/1251] eta 0:02:15 lr 0.000620 time 0.2748 (0.2883) loss 3.6485 (3.6111) grad_norm 1.2156 (1.3984) [2021-04-16 03:04:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][790/1251] eta 0:02:12 lr 0.000620 time 0.2796 (0.2882) loss 3.4938 (3.6110) grad_norm 1.5844 (1.3985) [2021-04-16 03:04:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][800/1251] eta 0:02:09 lr 0.000620 time 0.2846 (0.2881) loss 3.4666 (3.6104) grad_norm 1.4671 (1.3993) [2021-04-16 03:04:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][810/1251] eta 0:02:06 lr 0.000620 time 0.2748 (0.2879) loss 3.9570 (3.6144) grad_norm 1.5911 (1.3993) [2021-04-16 03:04:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][820/1251] eta 0:02:04 lr 0.000620 time 0.2876 (0.2878) loss 2.8069 (3.6167) grad_norm 1.4810 (1.3996) [2021-04-16 03:04:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][830/1251] eta 0:02:01 lr 0.000620 time 0.2838 (0.2877) loss 3.6656 (3.6170) grad_norm 1.2851 (1.3999) [2021-04-16 03:04:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][840/1251] eta 0:01:58 lr 0.000620 time 0.2890 (0.2877) loss 4.1616 (3.6175) grad_norm 1.4721 (1.3998) [2021-04-16 03:04:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][850/1251] eta 0:01:55 lr 0.000620 time 0.2580 (0.2875) loss 3.5942 (3.6172) grad_norm 1.4140 (1.3992) [2021-04-16 03:04:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][860/1251] eta 0:01:52 lr 0.000620 time 0.2755 (0.2873) loss 3.4246 (3.6161) grad_norm 1.2484 (1.3982) [2021-04-16 03:04:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][870/1251] eta 0:01:49 lr 0.000620 time 0.2806 (0.2872) loss 3.2879 (3.6161) grad_norm 1.3065 (1.3975) [2021-04-16 03:04:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][880/1251] eta 0:01:46 lr 0.000620 time 0.2658 (0.2871) loss 3.3116 (3.6130) grad_norm 1.3099 (1.3973) [2021-04-16 03:04:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][890/1251] eta 0:01:43 lr 0.000619 time 0.2681 (0.2871) loss 3.8375 (3.6120) grad_norm 1.3551 (1.3977) [2021-04-16 03:04:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][900/1251] eta 0:01:40 lr 0.000619 time 0.3881 (0.2872) loss 3.5719 (3.6129) grad_norm 1.8192 (1.3979) [2021-04-16 03:04:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][910/1251] eta 0:01:37 lr 0.000619 time 0.2721 (0.2871) loss 3.5983 (3.6127) grad_norm 1.4058 (1.3992) [2021-04-16 03:04:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][920/1251] eta 0:01:35 lr 0.000619 time 0.2589 (0.2870) loss 3.7210 (3.6143) grad_norm 1.2002 (1.3995) [2021-04-16 03:05:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][930/1251] eta 0:01:32 lr 0.000619 time 0.2866 (0.2869) loss 4.3175 (3.6138) grad_norm 1.5911 (1.3999) [2021-04-16 03:05:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][940/1251] eta 0:01:29 lr 0.000619 time 0.2760 (0.2867) loss 3.0898 (3.6130) grad_norm 1.4770 (1.4022) [2021-04-16 03:05:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][950/1251] eta 0:01:26 lr 0.000619 time 0.2815 (0.2866) loss 2.3714 (3.6121) grad_norm 1.4436 (1.4024) [2021-04-16 03:05:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][960/1251] eta 0:01:23 lr 0.000619 time 0.2487 (0.2865) loss 3.5929 (3.6119) grad_norm 1.4065 (1.4018) [2021-04-16 03:05:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][970/1251] eta 0:01:20 lr 0.000619 time 0.2706 (0.2864) loss 3.2737 (3.6133) grad_norm 1.2028 (1.4016) [2021-04-16 03:05:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][980/1251] eta 0:01:17 lr 0.000619 time 0.2938 (0.2863) loss 3.8276 (3.6130) grad_norm 1.2004 (1.4016) [2021-04-16 03:05:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][990/1251] eta 0:01:14 lr 0.000619 time 0.3012 (0.2862) loss 3.1482 (3.6166) grad_norm 1.7647 (1.4025) [2021-04-16 03:05:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][1000/1251] eta 0:01:11 lr 0.000619 time 0.2681 (0.2861) loss 4.0676 (3.6186) grad_norm 1.4286 (1.4028) [2021-04-16 03:05:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][1010/1251] eta 0:01:08 lr 0.000619 time 0.2784 (0.2860) loss 4.0723 (3.6158) grad_norm 1.6056 (1.4036) [2021-04-16 03:05:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][1020/1251] eta 0:01:06 lr 0.000619 time 0.2713 (0.2860) loss 3.9539 (3.6183) grad_norm 1.1924 (1.4039) [2021-04-16 03:05:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][1030/1251] eta 0:01:03 lr 0.000619 time 0.2817 (0.2859) loss 2.8487 (3.6163) grad_norm 1.7454 (1.4038) [2021-04-16 03:05:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][1040/1251] eta 0:01:00 lr 0.000619 time 0.2651 (0.2858) loss 2.6952 (3.6134) grad_norm 1.3471 (1.4042) [2021-04-16 03:05:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][1050/1251] eta 0:00:57 lr 0.000619 time 0.2653 (0.2858) loss 2.8131 (3.6106) grad_norm 1.4139 (1.4040) [2021-04-16 03:05:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][1060/1251] eta 0:00:54 lr 0.000619 time 0.2545 (0.2857) loss 3.7556 (3.6105) grad_norm 1.1733 (1.4035) [2021-04-16 03:05:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][1070/1251] eta 0:00:51 lr 0.000619 time 0.2852 (0.2856) loss 4.1526 (3.6120) grad_norm 1.6367 (1.4042) [2021-04-16 03:05:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][1080/1251] eta 0:00:48 lr 0.000619 time 0.2844 (0.2855) loss 2.8327 (3.6142) grad_norm 1.1623 (1.4039) [2021-04-16 03:05:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][1090/1251] eta 0:00:45 lr 0.000619 time 0.2715 (0.2854) loss 3.8301 (3.6120) grad_norm 1.2843 (1.4034) [2021-04-16 03:05:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][1100/1251] eta 0:00:43 lr 0.000619 time 0.2538 (0.2854) loss 3.7051 (3.6095) grad_norm 1.2351 (1.4031) [2021-04-16 03:05:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][1110/1251] eta 0:00:40 lr 0.000619 time 0.2846 (0.2853) loss 3.0616 (3.6048) grad_norm 1.2710 (1.4024) [2021-04-16 03:05:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][1120/1251] eta 0:00:37 lr 0.000619 time 0.2880 (0.2852) loss 2.9503 (3.6040) grad_norm 1.2730 (1.4025) [2021-04-16 03:05:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][1130/1251] eta 0:00:34 lr 0.000619 time 0.2616 (0.2852) loss 3.8529 (3.6049) grad_norm 1.2653 (1.4018) [2021-04-16 03:06:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][1140/1251] eta 0:00:31 lr 0.000618 time 0.2713 (0.2852) loss 2.9285 (3.6030) grad_norm 1.2879 (1.4012) [2021-04-16 03:06:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][1150/1251] eta 0:00:28 lr 0.000618 time 0.2788 (0.2852) loss 3.4138 (3.6003) grad_norm 1.4156 (1.4016) [2021-04-16 03:06:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][1160/1251] eta 0:00:25 lr 0.000618 time 0.2566 (0.2852) loss 3.7023 (3.6009) grad_norm 1.5310 (1.4019) [2021-04-16 03:06:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][1170/1251] eta 0:00:23 lr 0.000618 time 0.2714 (0.2851) loss 3.9896 (3.6012) grad_norm 1.5903 (1.4018) [2021-04-16 03:06:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][1180/1251] eta 0:00:20 lr 0.000618 time 0.2907 (0.2851) loss 3.6005 (3.6009) grad_norm 1.4414 (1.4014) [2021-04-16 03:06:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][1190/1251] eta 0:00:17 lr 0.000618 time 0.2669 (0.2850) loss 3.4039 (3.6000) grad_norm 1.4484 (1.4017) [2021-04-16 03:06:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][1200/1251] eta 0:00:14 lr 0.000618 time 0.2479 (0.2849) loss 3.7187 (3.6023) grad_norm 1.4160 (1.4016) [2021-04-16 03:06:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][1210/1251] eta 0:00:11 lr 0.000618 time 0.2825 (0.2848) loss 3.5803 (3.6033) grad_norm 1.4195 (1.4018) [2021-04-16 03:06:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][1220/1251] eta 0:00:08 lr 0.000618 time 0.3027 (0.2848) loss 2.9259 (3.6003) grad_norm 1.2240 (1.4011) [2021-04-16 03:06:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][1230/1251] eta 0:00:05 lr 0.000618 time 0.2598 (0.2847) loss 4.2481 (3.6013) grad_norm 1.4810 (1.4005) [2021-04-16 03:06:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][1240/1251] eta 0:00:03 lr 0.000618 time 0.2510 (0.2845) loss 3.7317 (3.6018) grad_norm 1.5184 (1.4000) [2021-04-16 03:06:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [127/300][1250/1251] eta 0:00:00 lr 0.000618 time 0.2480 (0.2843) loss 4.0502 (3.6033) grad_norm 1.4951 (1.3996) [2021-04-16 03:06:33 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 127 training takes 0:05:58 [2021-04-16 03:06:33 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_127.pth saving...... [2021-04-16 03:06:45 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_127.pth saved !!! [2021-04-16 03:06:46 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.157 (1.157) Loss 1.0140 (1.0140) Acc@1 75.098 (75.098) Acc@5 92.676 (92.676) [2021-04-16 03:06:48 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.528 (0.235) Loss 1.1401 (1.0841) Acc@1 72.656 (74.467) Acc@5 90.820 (92.383) [2021-04-16 03:06:50 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.121 (0.241) Loss 1.0526 (1.0626) Acc@1 75.000 (75.177) Acc@5 92.383 (92.666) [2021-04-16 03:06:52 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.122 (0.225) Loss 1.1122 (1.0664) Acc@1 73.535 (75.003) Acc@5 92.480 (92.584) [2021-04-16 03:06:54 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.215) Loss 1.0089 (1.0676) Acc@1 76.465 (74.981) Acc@5 93.262 (92.595) [2021-04-16 03:06:59 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 74.880 Acc@5 92.636 [2021-04-16 03:06:59 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 74.9% [2021-04-16 03:06:59 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 74.88% [2021-04-16 03:07:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][0/1251] eta 1:12:22 lr 0.000618 time 3.4710 (3.4710) loss 2.6968 (2.6968) grad_norm 1.5296 (1.5296) [2021-04-16 03:07:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][10/1251] eta 0:11:42 lr 0.000618 time 0.2862 (0.5658) loss 4.0125 (3.7312) grad_norm 1.5453 (1.4750) [2021-04-16 03:07:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][20/1251] eta 0:08:46 lr 0.000618 time 0.2758 (0.4278) loss 4.1054 (3.4682) grad_norm 1.3621 (1.4180) [2021-04-16 03:07:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][30/1251] eta 0:07:41 lr 0.000618 time 0.2488 (0.3782) loss 3.3630 (3.4898) grad_norm 1.7024 (1.4170) [2021-04-16 03:07:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][40/1251] eta 0:07:09 lr 0.000618 time 0.2749 (0.3543) loss 3.7375 (3.4831) grad_norm 1.4971 (1.4485) [2021-04-16 03:07:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][50/1251] eta 0:06:46 lr 0.000618 time 0.2635 (0.3382) loss 3.7647 (3.4571) grad_norm 1.2783 (1.4332) [2021-04-16 03:07:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][60/1251] eta 0:06:31 lr 0.000618 time 0.2636 (0.3285) loss 3.2816 (3.4842) grad_norm 1.6550 (1.4160) [2021-04-16 03:07:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][70/1251] eta 0:06:20 lr 0.000618 time 0.2992 (0.3218) loss 4.3623 (3.5053) grad_norm 1.3788 (1.4210) [2021-04-16 03:07:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][80/1251] eta 0:06:11 lr 0.000618 time 0.2747 (0.3169) loss 3.7772 (3.5045) grad_norm 1.6081 (1.4221) [2021-04-16 03:07:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][90/1251] eta 0:06:04 lr 0.000618 time 0.2979 (0.3141) loss 4.2410 (3.5246) grad_norm 1.5744 (1.4386) [2021-04-16 03:07:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][100/1251] eta 0:05:58 lr 0.000618 time 0.2618 (0.3117) loss 3.7775 (3.5345) grad_norm 1.3234 (1.4332) [2021-04-16 03:07:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][110/1251] eta 0:05:52 lr 0.000618 time 0.2755 (0.3088) loss 2.5500 (3.5379) grad_norm 1.3949 (1.4247) [2021-04-16 03:07:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][120/1251] eta 0:05:46 lr 0.000618 time 0.2916 (0.3067) loss 4.1727 (3.5620) grad_norm 1.5074 (1.4220) [2021-04-16 03:07:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][130/1251] eta 0:05:41 lr 0.000618 time 0.2891 (0.3045) loss 3.0316 (3.5927) grad_norm 1.3101 (1.4189) [2021-04-16 03:07:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][140/1251] eta 0:05:36 lr 0.000617 time 0.2835 (0.3027) loss 3.9404 (3.5879) grad_norm 1.2462 (1.4207) [2021-04-16 03:07:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][150/1251] eta 0:05:31 lr 0.000617 time 0.2850 (0.3008) loss 2.6742 (3.5966) grad_norm 1.4338 (1.4178) [2021-04-16 03:07:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][160/1251] eta 0:05:26 lr 0.000617 time 0.2528 (0.2989) loss 3.4969 (3.5849) grad_norm 1.5178 (1.4175) [2021-04-16 03:07:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][170/1251] eta 0:05:22 lr 0.000617 time 0.2711 (0.2979) loss 2.3493 (3.5839) grad_norm 1.3329 (1.4140) [2021-04-16 03:07:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][180/1251] eta 0:05:17 lr 0.000617 time 0.3013 (0.2967) loss 3.3576 (3.5863) grad_norm 1.3354 (1.4084) [2021-04-16 03:07:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][190/1251] eta 0:05:13 lr 0.000617 time 0.2878 (0.2955) loss 3.3008 (3.5844) grad_norm 1.5880 (1.4120) [2021-04-16 03:07:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][200/1251] eta 0:05:09 lr 0.000617 time 0.2833 (0.2946) loss 4.3707 (3.5856) grad_norm 1.4055 (1.4099) [2021-04-16 03:08:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][210/1251] eta 0:05:05 lr 0.000617 time 0.2906 (0.2938) loss 3.6858 (3.5880) grad_norm 1.3946 (1.4085) [2021-04-16 03:08:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][220/1251] eta 0:05:02 lr 0.000617 time 0.2731 (0.2931) loss 4.0353 (3.5912) grad_norm 1.9167 (1.4171) [2021-04-16 03:08:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][230/1251] eta 0:04:58 lr 0.000617 time 0.2659 (0.2925) loss 4.1297 (3.5912) grad_norm 1.3657 (1.4185) [2021-04-16 03:08:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][240/1251] eta 0:04:55 lr 0.000617 time 0.2445 (0.2925) loss 3.4788 (3.5859) grad_norm 1.3336 (1.4145) [2021-04-16 03:08:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][250/1251] eta 0:04:52 lr 0.000617 time 0.2902 (0.2918) loss 2.6561 (3.5801) grad_norm 1.3659 (1.4123) [2021-04-16 03:08:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][260/1251] eta 0:04:48 lr 0.000617 time 0.2767 (0.2911) loss 3.5640 (3.5847) grad_norm 1.4173 (1.4122) [2021-04-16 03:08:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][270/1251] eta 0:04:45 lr 0.000617 time 0.2910 (0.2907) loss 3.0583 (3.5801) grad_norm 1.3426 (1.4147) [2021-04-16 03:08:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][280/1251] eta 0:04:42 lr 0.000617 time 0.2754 (0.2905) loss 3.7256 (3.5819) grad_norm 1.3524 (1.4148) [2021-04-16 03:08:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][290/1251] eta 0:04:38 lr 0.000617 time 0.2796 (0.2901) loss 3.8364 (3.5861) grad_norm 1.5385 (1.4166) [2021-04-16 03:08:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][300/1251] eta 0:04:35 lr 0.000617 time 0.2985 (0.2898) loss 4.1628 (3.5931) grad_norm 1.4133 (1.4200) [2021-04-16 03:08:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][310/1251] eta 0:04:32 lr 0.000617 time 0.2525 (0.2894) loss 4.1922 (3.5898) grad_norm 1.2908 (1.4217) [2021-04-16 03:08:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][320/1251] eta 0:04:29 lr 0.000617 time 0.2989 (0.2892) loss 4.3038 (3.5908) grad_norm 1.1867 (1.4235) [2021-04-16 03:08:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][330/1251] eta 0:04:26 lr 0.000617 time 0.2891 (0.2889) loss 4.2134 (3.5966) grad_norm 1.3701 (1.4218) [2021-04-16 03:08:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][340/1251] eta 0:04:23 lr 0.000617 time 0.2769 (0.2888) loss 3.3889 (3.5988) grad_norm 1.4481 (1.4209) [2021-04-16 03:08:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][350/1251] eta 0:04:20 lr 0.000617 time 0.4557 (0.2894) loss 3.1435 (3.6020) grad_norm 1.2416 (1.4249) [2021-04-16 03:08:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][360/1251] eta 0:04:17 lr 0.000617 time 0.2752 (0.2889) loss 3.5517 (3.6057) grad_norm 1.3223 (1.4231) [2021-04-16 03:08:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][370/1251] eta 0:04:14 lr 0.000617 time 0.2689 (0.2886) loss 2.7417 (3.6074) grad_norm 1.2120 (1.4210) [2021-04-16 03:08:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][380/1251] eta 0:04:11 lr 0.000617 time 0.2860 (0.2885) loss 3.6769 (3.6087) grad_norm 1.3096 (1.4194) [2021-04-16 03:08:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][390/1251] eta 0:04:08 lr 0.000616 time 0.3229 (0.2883) loss 3.1376 (3.6179) grad_norm 1.2530 (nan) [2021-04-16 03:08:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][400/1251] eta 0:04:05 lr 0.000616 time 0.2974 (0.2879) loss 4.3713 (3.6236) grad_norm 1.3926 (nan) [2021-04-16 03:08:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][410/1251] eta 0:04:01 lr 0.000616 time 0.2742 (0.2877) loss 3.9193 (3.6220) grad_norm 1.5022 (nan) [2021-04-16 03:09:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][420/1251] eta 0:03:58 lr 0.000616 time 0.2683 (0.2875) loss 3.8236 (3.6184) grad_norm 1.1284 (nan) [2021-04-16 03:09:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][430/1251] eta 0:03:55 lr 0.000616 time 0.2862 (0.2873) loss 3.6386 (3.6187) grad_norm 1.3547 (nan) [2021-04-16 03:09:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][440/1251] eta 0:03:52 lr 0.000616 time 0.2881 (0.2871) loss 2.6880 (3.6171) grad_norm 1.5968 (nan) [2021-04-16 03:09:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][450/1251] eta 0:03:49 lr 0.000616 time 0.2631 (0.2867) loss 3.8372 (3.6187) grad_norm 1.3431 (nan) [2021-04-16 03:09:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][460/1251] eta 0:03:46 lr 0.000616 time 0.2850 (0.2868) loss 3.7333 (3.6187) grad_norm 1.2475 (nan) [2021-04-16 03:09:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][470/1251] eta 0:03:44 lr 0.000616 time 0.2685 (0.2869) loss 4.1063 (3.6177) grad_norm 1.2899 (nan) [2021-04-16 03:09:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][480/1251] eta 0:03:41 lr 0.000616 time 0.2441 (0.2870) loss 3.4889 (3.6159) grad_norm 1.4615 (nan) [2021-04-16 03:09:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][490/1251] eta 0:03:38 lr 0.000616 time 0.2420 (0.2867) loss 3.7540 (3.6163) grad_norm 1.3559 (nan) [2021-04-16 03:09:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][500/1251] eta 0:03:35 lr 0.000616 time 0.2702 (0.2866) loss 4.2535 (3.6175) grad_norm 1.2587 (nan) [2021-04-16 03:09:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][510/1251] eta 0:03:32 lr 0.000616 time 0.2860 (0.2864) loss 2.6178 (3.6098) grad_norm 1.4134 (nan) [2021-04-16 03:09:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][520/1251] eta 0:03:29 lr 0.000616 time 0.2941 (0.2864) loss 3.8928 (3.6123) grad_norm 1.3052 (nan) [2021-04-16 03:09:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][530/1251] eta 0:03:26 lr 0.000616 time 0.2716 (0.2861) loss 3.9533 (3.6137) grad_norm 1.2203 (nan) [2021-04-16 03:09:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][540/1251] eta 0:03:23 lr 0.000616 time 0.2926 (0.2860) loss 3.0019 (3.6130) grad_norm 1.3026 (nan) [2021-04-16 03:09:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][550/1251] eta 0:03:20 lr 0.000616 time 0.2825 (0.2858) loss 3.3764 (3.6157) grad_norm 1.3638 (nan) [2021-04-16 03:09:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][560/1251] eta 0:03:17 lr 0.000616 time 0.2920 (0.2856) loss 3.5981 (3.6175) grad_norm 1.2549 (nan) [2021-04-16 03:09:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][570/1251] eta 0:03:14 lr 0.000616 time 0.2790 (0.2856) loss 4.3965 (3.6122) grad_norm 1.3031 (nan) [2021-04-16 03:09:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][580/1251] eta 0:03:11 lr 0.000616 time 0.2733 (0.2857) loss 3.1272 (3.6131) grad_norm 1.4104 (nan) [2021-04-16 03:09:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][590/1251] eta 0:03:08 lr 0.000616 time 0.2873 (0.2856) loss 3.7582 (3.6180) grad_norm 1.2888 (nan) [2021-04-16 03:09:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][600/1251] eta 0:03:05 lr 0.000616 time 0.2706 (0.2855) loss 4.0875 (3.6162) grad_norm 1.6019 (nan) [2021-04-16 03:09:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][610/1251] eta 0:03:02 lr 0.000616 time 0.2772 (0.2854) loss 3.5209 (3.6122) grad_norm 1.2250 (nan) [2021-04-16 03:09:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][620/1251] eta 0:02:59 lr 0.000616 time 0.2608 (0.2853) loss 4.1420 (3.6115) grad_norm 1.3399 (nan) [2021-04-16 03:09:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][630/1251] eta 0:02:57 lr 0.000615 time 0.2660 (0.2851) loss 2.9520 (3.6112) grad_norm 1.5811 (nan) [2021-04-16 03:10:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][640/1251] eta 0:02:54 lr 0.000615 time 0.2637 (0.2849) loss 4.5355 (3.6090) grad_norm 1.3131 (nan) [2021-04-16 03:10:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][650/1251] eta 0:02:51 lr 0.000615 time 0.2433 (0.2848) loss 3.9918 (3.6095) grad_norm 1.1593 (nan) [2021-04-16 03:10:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][660/1251] eta 0:02:48 lr 0.000615 time 0.2560 (0.2847) loss 3.9149 (3.6078) grad_norm 1.6703 (nan) [2021-04-16 03:10:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][670/1251] eta 0:02:45 lr 0.000615 time 0.2852 (0.2846) loss 3.0946 (3.6082) grad_norm 1.6077 (nan) [2021-04-16 03:10:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][680/1251] eta 0:02:42 lr 0.000615 time 0.2543 (0.2846) loss 2.3618 (3.6048) grad_norm 1.6317 (nan) [2021-04-16 03:10:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][690/1251] eta 0:02:39 lr 0.000615 time 0.2818 (0.2847) loss 3.4485 (3.6046) grad_norm 1.5385 (nan) [2021-04-16 03:10:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][700/1251] eta 0:02:36 lr 0.000615 time 0.2543 (0.2846) loss 4.2819 (3.6067) grad_norm 1.3537 (nan) [2021-04-16 03:10:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][710/1251] eta 0:02:33 lr 0.000615 time 0.3061 (0.2846) loss 4.0185 (3.6059) grad_norm 1.3029 (nan) [2021-04-16 03:10:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][720/1251] eta 0:02:31 lr 0.000615 time 0.2859 (0.2847) loss 4.2871 (3.6073) grad_norm 1.3253 (nan) [2021-04-16 03:10:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][730/1251] eta 0:02:28 lr 0.000615 time 0.2907 (0.2846) loss 3.3851 (3.6065) grad_norm 1.4590 (nan) [2021-04-16 03:10:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][740/1251] eta 0:02:25 lr 0.000615 time 0.2443 (0.2845) loss 3.8648 (3.6084) grad_norm 1.6021 (nan) [2021-04-16 03:10:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][750/1251] eta 0:02:22 lr 0.000615 time 0.2670 (0.2843) loss 4.1179 (3.6072) grad_norm 1.2179 (nan) [2021-04-16 03:10:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][760/1251] eta 0:02:19 lr 0.000615 time 0.2926 (0.2842) loss 3.1059 (3.6072) grad_norm 1.2191 (nan) [2021-04-16 03:10:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][770/1251] eta 0:02:16 lr 0.000615 time 0.2866 (0.2841) loss 4.0663 (3.6123) grad_norm 1.2390 (nan) [2021-04-16 03:10:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][780/1251] eta 0:02:13 lr 0.000615 time 0.3062 (0.2840) loss 2.3973 (3.6102) grad_norm 1.2655 (nan) [2021-04-16 03:10:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][790/1251] eta 0:02:10 lr 0.000615 time 0.2813 (0.2840) loss 4.2669 (3.6110) grad_norm 1.6260 (nan) [2021-04-16 03:10:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][800/1251] eta 0:02:08 lr 0.000615 time 0.2661 (0.2839) loss 3.4300 (3.6130) grad_norm 1.6326 (nan) [2021-04-16 03:10:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][810/1251] eta 0:02:05 lr 0.000615 time 0.2914 (0.2838) loss 3.9799 (3.6131) grad_norm 1.3595 (nan) [2021-04-16 03:10:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][820/1251] eta 0:02:02 lr 0.000615 time 0.3102 (0.2840) loss 3.3226 (3.6140) grad_norm 1.3596 (nan) [2021-04-16 03:10:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][830/1251] eta 0:01:59 lr 0.000615 time 0.2829 (0.2839) loss 4.1789 (3.6158) grad_norm 1.1614 (nan) [2021-04-16 03:10:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][840/1251] eta 0:01:56 lr 0.000615 time 0.2773 (0.2839) loss 3.6416 (3.6145) grad_norm 1.2303 (nan) [2021-04-16 03:11:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][850/1251] eta 0:01:53 lr 0.000615 time 0.2782 (0.2839) loss 2.7930 (3.6110) grad_norm 1.3860 (nan) [2021-04-16 03:11:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][860/1251] eta 0:01:50 lr 0.000615 time 0.2737 (0.2839) loss 3.0873 (3.6114) grad_norm 1.3010 (nan) [2021-04-16 03:11:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][870/1251] eta 0:01:48 lr 0.000615 time 0.2670 (0.2838) loss 3.7003 (3.6116) grad_norm 1.9405 (nan) [2021-04-16 03:11:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][880/1251] eta 0:01:45 lr 0.000614 time 0.2651 (0.2838) loss 3.5254 (3.6089) grad_norm 1.4993 (nan) [2021-04-16 03:11:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][890/1251] eta 0:01:42 lr 0.000614 time 0.2479 (0.2836) loss 3.1792 (3.6071) grad_norm 1.1960 (nan) [2021-04-16 03:11:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][900/1251] eta 0:01:39 lr 0.000614 time 0.2947 (0.2835) loss 3.9082 (3.6095) grad_norm 1.1826 (nan) [2021-04-16 03:11:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][910/1251] eta 0:01:36 lr 0.000614 time 0.2952 (0.2835) loss 3.7924 (3.6075) grad_norm 1.7195 (nan) [2021-04-16 03:11:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][920/1251] eta 0:01:33 lr 0.000614 time 0.2652 (0.2835) loss 2.9505 (3.6090) grad_norm 1.4953 (nan) [2021-04-16 03:11:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][930/1251] eta 0:01:31 lr 0.000614 time 0.2643 (0.2837) loss 3.4414 (3.6092) grad_norm 1.4876 (nan) [2021-04-16 03:11:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][940/1251] eta 0:01:28 lr 0.000614 time 0.2753 (0.2837) loss 4.1145 (3.6079) grad_norm 1.2773 (nan) [2021-04-16 03:11:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][950/1251] eta 0:01:25 lr 0.000614 time 0.2738 (0.2836) loss 2.9908 (3.6099) grad_norm 1.4042 (nan) [2021-04-16 03:11:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][960/1251] eta 0:01:22 lr 0.000614 time 0.2866 (0.2835) loss 3.6578 (3.6103) grad_norm 1.3268 (nan) [2021-04-16 03:11:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][970/1251] eta 0:01:19 lr 0.000614 time 0.2821 (0.2835) loss 3.1705 (3.6079) grad_norm 1.4357 (nan) [2021-04-16 03:11:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][980/1251] eta 0:01:16 lr 0.000614 time 0.2786 (0.2835) loss 2.7516 (3.6066) grad_norm 1.4073 (nan) [2021-04-16 03:11:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][990/1251] eta 0:01:13 lr 0.000614 time 0.2563 (0.2833) loss 4.1116 (3.6089) grad_norm 1.2573 (nan) [2021-04-16 03:11:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][1000/1251] eta 0:01:11 lr 0.000614 time 0.2679 (0.2833) loss 3.8078 (3.6066) grad_norm 1.2102 (nan) [2021-04-16 03:11:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][1010/1251] eta 0:01:08 lr 0.000614 time 0.2562 (0.2833) loss 4.4818 (3.6075) grad_norm 1.6667 (nan) [2021-04-16 03:11:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][1020/1251] eta 0:01:05 lr 0.000614 time 0.2462 (0.2833) loss 4.1371 (3.6067) grad_norm 1.5300 (nan) [2021-04-16 03:11:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][1030/1251] eta 0:01:02 lr 0.000614 time 0.2842 (0.2833) loss 3.7798 (3.6073) grad_norm 1.5538 (nan) [2021-04-16 03:11:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][1040/1251] eta 0:00:59 lr 0.000614 time 0.2647 (0.2834) loss 3.9279 (3.6059) grad_norm 1.3983 (nan) [2021-04-16 03:11:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][1050/1251] eta 0:00:56 lr 0.000614 time 0.4073 (0.2834) loss 3.1687 (3.6035) grad_norm 1.4445 (nan) [2021-04-16 03:11:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][1060/1251] eta 0:00:54 lr 0.000614 time 0.2666 (0.2833) loss 4.1692 (3.6031) grad_norm 1.1960 (nan) [2021-04-16 03:12:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][1070/1251] eta 0:00:51 lr 0.000614 time 0.3006 (0.2833) loss 4.0712 (3.6020) grad_norm 1.3657 (nan) [2021-04-16 03:12:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][1080/1251] eta 0:00:48 lr 0.000614 time 0.3122 (0.2833) loss 4.1632 (3.6033) grad_norm 1.2705 (nan) [2021-04-16 03:12:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][1090/1251] eta 0:00:45 lr 0.000614 time 0.2890 (0.2833) loss 2.4545 (3.6050) grad_norm 1.2381 (nan) [2021-04-16 03:12:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][1100/1251] eta 0:00:42 lr 0.000614 time 0.2834 (0.2833) loss 4.0515 (3.6061) grad_norm 1.3484 (nan) [2021-04-16 03:12:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][1110/1251] eta 0:00:39 lr 0.000614 time 0.2888 (0.2832) loss 2.9618 (3.6029) grad_norm 1.3682 (nan) [2021-04-16 03:12:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][1120/1251] eta 0:00:37 lr 0.000614 time 0.2676 (0.2831) loss 3.0641 (3.6042) grad_norm 1.4118 (nan) [2021-04-16 03:12:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][1130/1251] eta 0:00:34 lr 0.000613 time 0.2857 (0.2830) loss 3.7187 (3.6037) grad_norm 1.5720 (nan) [2021-04-16 03:12:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][1140/1251] eta 0:00:31 lr 0.000613 time 0.2680 (0.2829) loss 4.0833 (3.6046) grad_norm 1.3025 (nan) [2021-04-16 03:12:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][1150/1251] eta 0:00:28 lr 0.000613 time 0.2909 (0.2829) loss 3.0638 (3.6039) grad_norm 1.3794 (nan) [2021-04-16 03:12:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][1160/1251] eta 0:00:25 lr 0.000613 time 0.2782 (0.2830) loss 3.7404 (3.6058) grad_norm 1.3362 (nan) [2021-04-16 03:12:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][1170/1251] eta 0:00:22 lr 0.000613 time 0.2923 (0.2829) loss 4.2781 (3.6049) grad_norm 1.3926 (nan) [2021-04-16 03:12:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][1180/1251] eta 0:00:20 lr 0.000613 time 0.2675 (0.2829) loss 3.5703 (3.6045) grad_norm 1.2152 (nan) [2021-04-16 03:12:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][1190/1251] eta 0:00:17 lr 0.000613 time 0.2515 (0.2829) loss 4.0718 (3.6047) grad_norm 1.4142 (nan) [2021-04-16 03:12:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][1200/1251] eta 0:00:14 lr 0.000613 time 0.2888 (0.2828) loss 3.3245 (3.6071) grad_norm 1.3063 (nan) [2021-04-16 03:12:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][1210/1251] eta 0:00:11 lr 0.000613 time 0.2892 (0.2828) loss 3.5606 (3.6059) grad_norm 1.3389 (nan) [2021-04-16 03:12:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][1220/1251] eta 0:00:08 lr 0.000613 time 0.2853 (0.2828) loss 4.1887 (3.6056) grad_norm 1.2712 (nan) [2021-04-16 03:12:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][1230/1251] eta 0:00:05 lr 0.000613 time 0.2796 (0.2828) loss 2.5207 (3.6061) grad_norm 1.4894 (nan) [2021-04-16 03:12:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][1240/1251] eta 0:00:03 lr 0.000613 time 0.2483 (0.2826) loss 3.7783 (3.6063) grad_norm 1.4453 (nan) [2021-04-16 03:12:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [128/300][1250/1251] eta 0:00:00 lr 0.000613 time 0.2481 (0.2824) loss 4.1365 (3.6058) grad_norm 1.3204 (nan) [2021-04-16 03:12:56 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 128 training takes 0:05:57 [2021-04-16 03:12:56 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_128.pth saving...... [2021-04-16 03:13:11 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_128.pth saved !!! [2021-04-16 03:13:13 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.187 (1.187) Loss 1.1397 (1.1397) Acc@1 74.316 (74.316) Acc@5 91.895 (91.895) [2021-04-16 03:13:14 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.124 (0.226) Loss 1.0573 (1.0869) Acc@1 74.609 (74.734) Acc@5 92.188 (92.321) [2021-04-16 03:13:16 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.127 (0.213) Loss 1.0827 (1.0781) Acc@1 75.293 (74.837) Acc@5 92.285 (92.527) [2021-04-16 03:13:18 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.209 (0.226) Loss 1.1317 (1.0741) Acc@1 73.145 (74.852) Acc@5 92.090 (92.673) [2021-04-16 03:13:21 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.293 (0.224) Loss 1.0232 (1.0690) Acc@1 75.977 (74.862) Acc@5 93.262 (92.804) [2021-04-16 03:13:28 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 74.890 Acc@5 92.806 [2021-04-16 03:13:28 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 74.9% [2021-04-16 03:13:28 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 74.89% [2021-04-16 03:13:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][0/1251] eta 1:18:46 lr 0.000613 time 3.7781 (3.7781) loss 4.0321 (4.0321) grad_norm 1.2847 (1.2847) [2021-04-16 03:13:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][10/1251] eta 0:12:12 lr 0.000613 time 0.3017 (0.5903) loss 4.1431 (3.7324) grad_norm 1.4378 (1.3941) [2021-04-16 03:13:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][20/1251] eta 0:09:13 lr 0.000613 time 0.2743 (0.4499) loss 4.2015 (3.7985) grad_norm 1.2664 (1.3993) [2021-04-16 03:13:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][30/1251] eta 0:08:01 lr 0.000613 time 0.2549 (0.3944) loss 3.0567 (3.6287) grad_norm 1.1672 (1.3688) [2021-04-16 03:13:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][40/1251] eta 0:07:22 lr 0.000613 time 0.2695 (0.3652) loss 3.3466 (3.5697) grad_norm 1.4990 (1.3650) [2021-04-16 03:13:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][50/1251] eta 0:06:56 lr 0.000613 time 0.2776 (0.3468) loss 4.5639 (3.5607) grad_norm 1.2169 (1.3538) [2021-04-16 03:13:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][60/1251] eta 0:06:40 lr 0.000613 time 0.3078 (0.3360) loss 3.9774 (3.5835) grad_norm 1.3494 (1.3670) [2021-04-16 03:13:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][70/1251] eta 0:06:27 lr 0.000613 time 0.2769 (0.3277) loss 2.6213 (3.5908) grad_norm 1.4676 (1.3769) [2021-04-16 03:13:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][80/1251] eta 0:06:16 lr 0.000613 time 0.2543 (0.3215) loss 2.9151 (3.5845) grad_norm 1.4193 (1.3955) [2021-04-16 03:13:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][90/1251] eta 0:06:08 lr 0.000613 time 0.2683 (0.3175) loss 3.6851 (3.5906) grad_norm 1.4167 (1.4125) [2021-04-16 03:14:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][100/1251] eta 0:06:01 lr 0.000613 time 0.2868 (0.3137) loss 4.3549 (3.6163) grad_norm 1.4917 (1.4195) [2021-04-16 03:14:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][110/1251] eta 0:05:53 lr 0.000613 time 0.2894 (0.3100) loss 3.6616 (3.5995) grad_norm 1.1634 (1.4094) [2021-04-16 03:14:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][120/1251] eta 0:05:48 lr 0.000612 time 0.2755 (0.3084) loss 3.6281 (3.6160) grad_norm 1.4944 (1.4118) [2021-04-16 03:14:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][130/1251] eta 0:05:46 lr 0.000612 time 0.4493 (0.3094) loss 3.2948 (3.6203) grad_norm 1.3598 (1.4074) [2021-04-16 03:14:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][140/1251] eta 0:05:40 lr 0.000612 time 0.2688 (0.3068) loss 3.6047 (3.6096) grad_norm 1.2761 (1.4017) [2021-04-16 03:14:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][150/1251] eta 0:05:35 lr 0.000612 time 0.2696 (0.3048) loss 3.5966 (3.6199) grad_norm 1.3703 (1.4049) [2021-04-16 03:14:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][160/1251] eta 0:05:30 lr 0.000612 time 0.2740 (0.3031) loss 3.9759 (3.6290) grad_norm 1.1357 (1.4092) [2021-04-16 03:14:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][170/1251] eta 0:05:26 lr 0.000612 time 0.2906 (0.3016) loss 3.4671 (3.6247) grad_norm 1.2838 (1.4092) [2021-04-16 03:14:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][180/1251] eta 0:05:22 lr 0.000612 time 0.2792 (0.3007) loss 3.4846 (3.6255) grad_norm 1.2994 (1.4033) [2021-04-16 03:14:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][190/1251] eta 0:05:17 lr 0.000612 time 0.2615 (0.2992) loss 3.4125 (3.6284) grad_norm 1.5565 (1.4026) [2021-04-16 03:14:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][200/1251] eta 0:05:13 lr 0.000612 time 0.2897 (0.2981) loss 4.0785 (3.6316) grad_norm 1.3106 (1.4021) [2021-04-16 03:14:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][210/1251] eta 0:05:09 lr 0.000612 time 0.2875 (0.2971) loss 2.8318 (3.6167) grad_norm 1.4925 (1.4013) [2021-04-16 03:14:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][220/1251] eta 0:05:05 lr 0.000612 time 0.2893 (0.2963) loss 3.9666 (3.6112) grad_norm 1.2788 (1.4032) [2021-04-16 03:14:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][230/1251] eta 0:05:01 lr 0.000612 time 0.2958 (0.2953) loss 3.8431 (3.6155) grad_norm 1.3428 (1.3996) [2021-04-16 03:14:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][240/1251] eta 0:04:58 lr 0.000612 time 0.2618 (0.2953) loss 2.6218 (3.6113) grad_norm 1.4778 (1.3986) [2021-04-16 03:14:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][250/1251] eta 0:04:55 lr 0.000612 time 0.2696 (0.2950) loss 3.5274 (3.6129) grad_norm 1.3825 (1.3985) [2021-04-16 03:14:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][260/1251] eta 0:04:51 lr 0.000612 time 0.4044 (0.2947) loss 3.8946 (3.6192) grad_norm 1.6182 (1.3995) [2021-04-16 03:14:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][270/1251] eta 0:04:48 lr 0.000612 time 0.2723 (0.2942) loss 3.8318 (3.6251) grad_norm 1.2344 (1.4009) [2021-04-16 03:14:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][280/1251] eta 0:04:45 lr 0.000612 time 0.2939 (0.2935) loss 4.0586 (3.6208) grad_norm 1.4262 (1.4007) [2021-04-16 03:14:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][290/1251] eta 0:04:41 lr 0.000612 time 0.2882 (0.2929) loss 4.1278 (3.6263) grad_norm 1.3929 (1.4009) [2021-04-16 03:14:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][300/1251] eta 0:04:38 lr 0.000612 time 0.2588 (0.2926) loss 3.3046 (3.6187) grad_norm 1.2557 (1.4016) [2021-04-16 03:14:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][310/1251] eta 0:04:34 lr 0.000612 time 0.2741 (0.2922) loss 4.3853 (3.6179) grad_norm 1.4148 (1.4005) [2021-04-16 03:15:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][320/1251] eta 0:04:31 lr 0.000612 time 0.2757 (0.2917) loss 3.8752 (3.6167) grad_norm 1.3920 (1.4004) [2021-04-16 03:15:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][330/1251] eta 0:04:28 lr 0.000612 time 0.2775 (0.2913) loss 3.8992 (3.6155) grad_norm 1.8497 (1.4016) [2021-04-16 03:15:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][340/1251] eta 0:04:25 lr 0.000612 time 0.2699 (0.2909) loss 2.9783 (3.6039) grad_norm 1.3870 (1.4052) [2021-04-16 03:15:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][350/1251] eta 0:04:21 lr 0.000612 time 0.2876 (0.2906) loss 4.0444 (3.6080) grad_norm 1.5597 (1.4106) [2021-04-16 03:15:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][360/1251] eta 0:04:19 lr 0.000612 time 0.2898 (0.2910) loss 4.2348 (3.6094) grad_norm 1.2760 (1.4103) [2021-04-16 03:15:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][370/1251] eta 0:04:16 lr 0.000611 time 0.2789 (0.2907) loss 3.8293 (3.6104) grad_norm 1.2891 (1.4089) [2021-04-16 03:15:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][380/1251] eta 0:04:12 lr 0.000611 time 0.2738 (0.2902) loss 3.7073 (3.6121) grad_norm 1.3490 (1.4072) [2021-04-16 03:15:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][390/1251] eta 0:04:09 lr 0.000611 time 0.2709 (0.2898) loss 4.0402 (3.6118) grad_norm 1.6517 (1.4077) [2021-04-16 03:15:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][400/1251] eta 0:04:06 lr 0.000611 time 0.2700 (0.2894) loss 3.9306 (3.6172) grad_norm 1.3203 (1.4084) [2021-04-16 03:15:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][410/1251] eta 0:04:03 lr 0.000611 time 0.2872 (0.2891) loss 2.3774 (3.6122) grad_norm 1.4174 (1.4094) [2021-04-16 03:15:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][420/1251] eta 0:04:00 lr 0.000611 time 0.2771 (0.2888) loss 2.5432 (3.6098) grad_norm 1.9354 (1.4105) [2021-04-16 03:15:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][430/1251] eta 0:03:56 lr 0.000611 time 0.2742 (0.2885) loss 3.1331 (3.6098) grad_norm 1.4603 (1.4098) [2021-04-16 03:15:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][440/1251] eta 0:03:53 lr 0.000611 time 0.2543 (0.2883) loss 3.8130 (3.6038) grad_norm 1.2440 (1.4087) [2021-04-16 03:15:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][450/1251] eta 0:03:50 lr 0.000611 time 0.2904 (0.2881) loss 2.8446 (3.6022) grad_norm 1.2823 (1.4069) [2021-04-16 03:15:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][460/1251] eta 0:03:47 lr 0.000611 time 0.2693 (0.2878) loss 3.7865 (3.6066) grad_norm 1.2571 (1.4048) [2021-04-16 03:15:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][470/1251] eta 0:03:44 lr 0.000611 time 0.2597 (0.2881) loss 4.0327 (3.6122) grad_norm 1.4157 (1.4048) [2021-04-16 03:15:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][480/1251] eta 0:03:42 lr 0.000611 time 0.3053 (0.2883) loss 3.6824 (3.6054) grad_norm 1.2116 (1.4043) [2021-04-16 03:15:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][490/1251] eta 0:03:39 lr 0.000611 time 0.2796 (0.2883) loss 4.2947 (3.6050) grad_norm 1.4907 (1.4043) [2021-04-16 03:15:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][500/1251] eta 0:03:36 lr 0.000611 time 0.2578 (0.2880) loss 2.4357 (3.5982) grad_norm 1.3395 (1.4053) [2021-04-16 03:15:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][510/1251] eta 0:03:33 lr 0.000611 time 0.2796 (0.2878) loss 3.7621 (3.5981) grad_norm 1.5634 (1.4061) [2021-04-16 03:15:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][520/1251] eta 0:03:30 lr 0.000611 time 0.2679 (0.2876) loss 4.1816 (3.5982) grad_norm 1.4196 (1.4061) [2021-04-16 03:16:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][530/1251] eta 0:03:27 lr 0.000611 time 0.2901 (0.2874) loss 4.0047 (3.5978) grad_norm 1.5363 (1.4063) [2021-04-16 03:16:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][540/1251] eta 0:03:24 lr 0.000611 time 0.2690 (0.2872) loss 3.4424 (3.5936) grad_norm 1.4236 (1.4066) [2021-04-16 03:16:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][550/1251] eta 0:03:21 lr 0.000611 time 0.2842 (0.2870) loss 3.2617 (3.5906) grad_norm 1.3640 (1.4067) [2021-04-16 03:16:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][560/1251] eta 0:03:18 lr 0.000611 time 0.2667 (0.2869) loss 4.2282 (3.5895) grad_norm 1.3369 (1.4063) [2021-04-16 03:16:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][570/1251] eta 0:03:15 lr 0.000611 time 0.2741 (0.2867) loss 3.6258 (3.5841) grad_norm 1.2004 (1.4047) [2021-04-16 03:16:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][580/1251] eta 0:03:12 lr 0.000611 time 0.2860 (0.2865) loss 2.5460 (3.5850) grad_norm 1.3645 (1.4037) [2021-04-16 03:16:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][590/1251] eta 0:03:09 lr 0.000611 time 0.2863 (0.2866) loss 2.5910 (3.5790) grad_norm 1.2108 (1.4030) [2021-04-16 03:16:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][600/1251] eta 0:03:06 lr 0.000611 time 0.2969 (0.2864) loss 3.8889 (3.5818) grad_norm 1.8604 (1.4054) [2021-04-16 03:16:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][610/1251] eta 0:03:03 lr 0.000611 time 0.2678 (0.2862) loss 3.6497 (3.5798) grad_norm 1.7895 (1.4064) [2021-04-16 03:16:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][620/1251] eta 0:03:00 lr 0.000610 time 0.2592 (0.2862) loss 3.7813 (3.5869) grad_norm 1.3305 (1.4064) [2021-04-16 03:16:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][630/1251] eta 0:02:57 lr 0.000610 time 0.2593 (0.2860) loss 3.8600 (3.5889) grad_norm 1.4177 (1.4068) [2021-04-16 03:16:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][640/1251] eta 0:02:54 lr 0.000610 time 0.2557 (0.2859) loss 4.3041 (3.5939) grad_norm 1.3977 (1.4078) [2021-04-16 03:16:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][650/1251] eta 0:02:51 lr 0.000610 time 0.2450 (0.2858) loss 2.4978 (3.5931) grad_norm 1.3788 (1.4076) [2021-04-16 03:16:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][660/1251] eta 0:02:48 lr 0.000610 time 0.2774 (0.2856) loss 3.6477 (3.5935) grad_norm 1.3747 (1.4086) [2021-04-16 03:16:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][670/1251] eta 0:02:45 lr 0.000610 time 0.2846 (0.2855) loss 3.7751 (3.5943) grad_norm 1.5421 (1.4088) [2021-04-16 03:16:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][680/1251] eta 0:02:42 lr 0.000610 time 0.2640 (0.2854) loss 3.7673 (3.5955) grad_norm 1.3416 (1.4081) [2021-04-16 03:16:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][690/1251] eta 0:02:40 lr 0.000610 time 0.2531 (0.2854) loss 4.1121 (3.5976) grad_norm 1.3681 (1.4077) [2021-04-16 03:16:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][700/1251] eta 0:02:37 lr 0.000610 time 0.4247 (0.2855) loss 2.9590 (3.5999) grad_norm 1.3786 (1.4073) [2021-04-16 03:16:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][710/1251] eta 0:02:34 lr 0.000610 time 0.2751 (0.2856) loss 3.7647 (3.6025) grad_norm 1.7082 (1.4088) [2021-04-16 03:16:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][720/1251] eta 0:02:31 lr 0.000610 time 0.2784 (0.2855) loss 3.9086 (3.6021) grad_norm 1.4154 (1.4090) [2021-04-16 03:16:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][730/1251] eta 0:02:28 lr 0.000610 time 0.2570 (0.2854) loss 3.0744 (3.6020) grad_norm 1.4022 (1.4095) [2021-04-16 03:17:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][740/1251] eta 0:02:25 lr 0.000610 time 0.2635 (0.2852) loss 2.6294 (3.5986) grad_norm 1.6530 (1.4112) [2021-04-16 03:17:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][750/1251] eta 0:02:22 lr 0.000610 time 0.2762 (0.2851) loss 3.4534 (3.6032) grad_norm 1.3100 (1.4108) [2021-04-16 03:17:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][760/1251] eta 0:02:19 lr 0.000610 time 0.2834 (0.2850) loss 3.9657 (3.6019) grad_norm 1.4028 (1.4108) [2021-04-16 03:17:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][770/1251] eta 0:02:17 lr 0.000610 time 0.2641 (0.2848) loss 3.6608 (3.6043) grad_norm 1.2944 (1.4096) [2021-04-16 03:17:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][780/1251] eta 0:02:14 lr 0.000610 time 0.2738 (0.2847) loss 3.7309 (3.5997) grad_norm 1.5095 (1.4109) [2021-04-16 03:17:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][790/1251] eta 0:02:11 lr 0.000610 time 0.2892 (0.2847) loss 3.7931 (3.6025) grad_norm 1.6005 (1.4115) [2021-04-16 03:17:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][800/1251] eta 0:02:08 lr 0.000610 time 0.2583 (0.2846) loss 4.5037 (3.6048) grad_norm 1.2763 (1.4111) [2021-04-16 03:17:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][810/1251] eta 0:02:05 lr 0.000610 time 0.2548 (0.2845) loss 2.7771 (3.6035) grad_norm 1.2277 (1.4103) [2021-04-16 03:17:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][820/1251] eta 0:02:02 lr 0.000610 time 0.2524 (0.2846) loss 3.3553 (3.6003) grad_norm 1.4085 (1.4106) [2021-04-16 03:17:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][830/1251] eta 0:01:59 lr 0.000610 time 0.2743 (0.2846) loss 3.9929 (3.6024) grad_norm 1.4190 (1.4111) [2021-04-16 03:17:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][840/1251] eta 0:01:56 lr 0.000610 time 0.2836 (0.2846) loss 4.1037 (3.6062) grad_norm 1.2843 (1.4110) [2021-04-16 03:17:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][850/1251] eta 0:01:54 lr 0.000610 time 0.2844 (0.2845) loss 3.7192 (3.6066) grad_norm 1.3561 (1.4111) [2021-04-16 03:17:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][860/1251] eta 0:01:51 lr 0.000610 time 0.2584 (0.2844) loss 3.3344 (3.6095) grad_norm 1.5112 (1.4125) [2021-04-16 03:17:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][870/1251] eta 0:01:48 lr 0.000609 time 0.2977 (0.2843) loss 3.8061 (3.6088) grad_norm 1.3909 (1.4133) [2021-04-16 03:17:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][880/1251] eta 0:01:45 lr 0.000609 time 0.2971 (0.2845) loss 3.7043 (3.6088) grad_norm 1.2041 (1.4150) [2021-04-16 03:17:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][890/1251] eta 0:01:42 lr 0.000609 time 0.2737 (0.2844) loss 3.6616 (3.6088) grad_norm 1.2596 (1.4154) [2021-04-16 03:17:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][900/1251] eta 0:01:39 lr 0.000609 time 0.2725 (0.2842) loss 3.6501 (3.6103) grad_norm 1.4398 (1.4152) [2021-04-16 03:17:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][910/1251] eta 0:01:36 lr 0.000609 time 0.2447 (0.2841) loss 3.7338 (3.6083) grad_norm 1.2674 (1.4142) [2021-04-16 03:17:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][920/1251] eta 0:01:34 lr 0.000609 time 0.2609 (0.2841) loss 3.9771 (3.6089) grad_norm 1.3775 (1.4141) [2021-04-16 03:17:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][930/1251] eta 0:01:31 lr 0.000609 time 0.3191 (0.2842) loss 3.9675 (3.6115) grad_norm 1.3821 (1.4138) [2021-04-16 03:17:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][940/1251] eta 0:01:28 lr 0.000609 time 0.2669 (0.2841) loss 3.6948 (3.6112) grad_norm 1.1286 (1.4125) [2021-04-16 03:17:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][950/1251] eta 0:01:25 lr 0.000609 time 0.2699 (0.2839) loss 3.8418 (3.6112) grad_norm 1.3317 (1.4123) [2021-04-16 03:18:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][960/1251] eta 0:01:22 lr 0.000609 time 0.2870 (0.2838) loss 3.9642 (3.6120) grad_norm 1.2281 (1.4121) [2021-04-16 03:18:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][970/1251] eta 0:01:19 lr 0.000609 time 0.2787 (0.2838) loss 3.5063 (3.6097) grad_norm 1.3752 (1.4126) [2021-04-16 03:18:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][980/1251] eta 0:01:16 lr 0.000609 time 0.3124 (0.2837) loss 3.5920 (3.6106) grad_norm 1.3956 (1.4119) [2021-04-16 03:18:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][990/1251] eta 0:01:14 lr 0.000609 time 0.2767 (0.2836) loss 3.8031 (3.6101) grad_norm 1.3169 (1.4117) [2021-04-16 03:18:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][1000/1251] eta 0:01:11 lr 0.000609 time 0.2605 (0.2835) loss 3.1637 (3.6060) grad_norm 1.2717 (1.4126) [2021-04-16 03:18:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][1010/1251] eta 0:01:08 lr 0.000609 time 0.2817 (0.2834) loss 3.3115 (3.6029) grad_norm 1.6198 (1.4127) [2021-04-16 03:18:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][1020/1251] eta 0:01:05 lr 0.000609 time 0.2678 (0.2834) loss 4.0060 (3.6005) grad_norm 1.2032 (1.4128) [2021-04-16 03:18:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][1030/1251] eta 0:01:02 lr 0.000609 time 0.2727 (0.2833) loss 3.4353 (3.5998) grad_norm 1.3012 (1.4123) [2021-04-16 03:18:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][1040/1251] eta 0:00:59 lr 0.000609 time 0.2535 (0.2834) loss 3.9052 (3.5994) grad_norm 1.4461 (1.4118) [2021-04-16 03:18:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][1050/1251] eta 0:00:56 lr 0.000609 time 0.4083 (0.2835) loss 3.8930 (3.6004) grad_norm 1.1422 (1.4121) [2021-04-16 03:18:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][1060/1251] eta 0:00:54 lr 0.000609 time 0.2703 (0.2835) loss 3.5850 (3.6015) grad_norm 1.3391 (1.4119) [2021-04-16 03:18:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][1070/1251] eta 0:00:51 lr 0.000609 time 0.2718 (0.2835) loss 3.6789 (3.6009) grad_norm 1.3239 (1.4122) [2021-04-16 03:18:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][1080/1251] eta 0:00:48 lr 0.000609 time 0.2918 (0.2834) loss 4.1373 (3.5998) grad_norm 1.4947 (1.4118) [2021-04-16 03:18:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][1090/1251] eta 0:00:45 lr 0.000609 time 0.2665 (0.2833) loss 3.3494 (3.5969) grad_norm 1.3209 (1.4112) [2021-04-16 03:18:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][1100/1251] eta 0:00:42 lr 0.000609 time 0.2481 (0.2833) loss 3.9623 (3.5997) grad_norm 1.4393 (1.4112) [2021-04-16 03:18:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][1110/1251] eta 0:00:39 lr 0.000608 time 0.2517 (0.2832) loss 4.2949 (3.6025) grad_norm 1.2405 (1.4115) [2021-04-16 03:18:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][1120/1251] eta 0:00:37 lr 0.000608 time 0.2931 (0.2832) loss 3.9340 (3.6036) grad_norm 1.4440 (1.4118) [2021-04-16 03:18:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][1130/1251] eta 0:00:34 lr 0.000608 time 0.2748 (0.2831) loss 3.9632 (3.6033) grad_norm 1.2496 (1.4112) [2021-04-16 03:18:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][1140/1251] eta 0:00:31 lr 0.000608 time 0.2459 (0.2830) loss 3.8530 (3.6008) grad_norm 1.3556 (1.4106) [2021-04-16 03:18:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][1150/1251] eta 0:00:28 lr 0.000608 time 0.2810 (0.2829) loss 3.4550 (3.6036) grad_norm 1.3619 (1.4108) [2021-04-16 03:18:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][1160/1251] eta 0:00:25 lr 0.000608 time 0.2634 (0.2829) loss 3.9991 (3.6030) grad_norm 1.5647 (1.4107) [2021-04-16 03:19:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][1170/1251] eta 0:00:22 lr 0.000608 time 0.2709 (0.2830) loss 3.9934 (3.6024) grad_norm 1.2147 (1.4100) [2021-04-16 03:19:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][1180/1251] eta 0:00:20 lr 0.000608 time 0.2903 (0.2830) loss 3.4247 (3.6011) grad_norm 1.4437 (1.4098) [2021-04-16 03:19:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][1190/1251] eta 0:00:17 lr 0.000608 time 0.2976 (0.2829) loss 4.0130 (3.5994) grad_norm 1.2298 (1.4098) [2021-04-16 03:19:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][1200/1251] eta 0:00:14 lr 0.000608 time 0.2765 (0.2829) loss 3.6239 (3.6015) grad_norm 1.4383 (1.4092) [2021-04-16 03:19:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][1210/1251] eta 0:00:11 lr 0.000608 time 0.3043 (0.2828) loss 4.3792 (3.5971) grad_norm 1.4922 (1.4093) [2021-04-16 03:19:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][1220/1251] eta 0:00:08 lr 0.000608 time 0.2783 (0.2828) loss 4.1276 (3.5954) grad_norm 1.5094 (1.4096) [2021-04-16 03:19:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][1230/1251] eta 0:00:05 lr 0.000608 time 0.2613 (0.2827) loss 4.4884 (3.5958) grad_norm 1.6605 (1.4100) [2021-04-16 03:19:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][1240/1251] eta 0:00:03 lr 0.000608 time 0.2482 (0.2826) loss 4.1209 (3.5970) grad_norm 1.1973 (1.4101) [2021-04-16 03:19:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [129/300][1250/1251] eta 0:00:00 lr 0.000608 time 0.2487 (0.2823) loss 4.1610 (3.5956) grad_norm 1.4357 (1.4100) [2021-04-16 03:19:24 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 129 training takes 0:05:55 [2021-04-16 03:19:24 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_129.pth saving...... [2021-04-16 03:19:37 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_129.pth saved !!! [2021-04-16 03:19:38 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.116 (1.116) Loss 1.0368 (1.0368) Acc@1 77.051 (77.051) Acc@5 93.359 (93.359) [2021-04-16 03:19:40 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.166 (0.253) Loss 1.1445 (1.0850) Acc@1 73.438 (75.000) Acc@5 92.871 (92.667) [2021-04-16 03:19:42 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.111 (0.207) Loss 1.1101 (1.0931) Acc@1 73.828 (74.693) Acc@5 93.262 (92.708) [2021-04-16 03:19:44 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.407 (0.225) Loss 1.0495 (1.0936) Acc@1 74.902 (74.868) Acc@5 93.359 (92.704) [2021-04-16 03:19:46 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.215) Loss 1.0589 (1.0905) Acc@1 75.391 (74.974) Acc@5 93.262 (92.762) [2021-04-16 03:19:49 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 74.838 Acc@5 92.708 [2021-04-16 03:19:49 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 74.8% [2021-04-16 03:19:49 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 74.89% [2021-04-16 03:19:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][0/1251] eta 2:05:30 lr 0.000608 time 6.0193 (6.0193) loss 4.7010 (4.7010) grad_norm 1.7156 (1.7156) [2021-04-16 03:19:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][10/1251] eta 0:16:27 lr 0.000608 time 0.2764 (0.7957) loss 3.4422 (3.7872) grad_norm 1.4922 (1.5252) [2021-04-16 03:20:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][20/1251] eta 0:11:26 lr 0.000608 time 0.2807 (0.5574) loss 3.5543 (3.7232) grad_norm 1.3435 (1.4585) [2021-04-16 03:20:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][30/1251] eta 0:09:31 lr 0.000608 time 0.2630 (0.4683) loss 3.5186 (3.6281) grad_norm 1.1887 (1.4310) [2021-04-16 03:20:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][40/1251] eta 0:08:34 lr 0.000608 time 0.2987 (0.4250) loss 3.6449 (3.5915) grad_norm 1.3870 (inf) [2021-04-16 03:20:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][50/1251] eta 0:07:54 lr 0.000608 time 0.3099 (0.3954) loss 3.1377 (3.5666) grad_norm 1.3598 (inf) [2021-04-16 03:20:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][60/1251] eta 0:07:28 lr 0.000608 time 0.2724 (0.3767) loss 4.0383 (3.6074) grad_norm 1.5510 (inf) [2021-04-16 03:20:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][70/1251] eta 0:07:08 lr 0.000608 time 0.2967 (0.3627) loss 3.6183 (3.6301) grad_norm 1.2317 (inf) [2021-04-16 03:20:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][80/1251] eta 0:06:51 lr 0.000608 time 0.2748 (0.3518) loss 2.3748 (3.5944) grad_norm 1.5407 (inf) [2021-04-16 03:20:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][90/1251] eta 0:06:39 lr 0.000608 time 0.2836 (0.3442) loss 3.2739 (3.5694) grad_norm 1.3233 (inf) [2021-04-16 03:20:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][100/1251] eta 0:06:28 lr 0.000608 time 0.2888 (0.3375) loss 2.3767 (3.5806) grad_norm 1.4317 (inf) [2021-04-16 03:20:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][110/1251] eta 0:06:19 lr 0.000607 time 0.2851 (0.3322) loss 3.8130 (3.6030) grad_norm 1.4447 (inf) [2021-04-16 03:20:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][120/1251] eta 0:06:10 lr 0.000607 time 0.2856 (0.3273) loss 4.0434 (3.5900) grad_norm 1.3615 (inf) [2021-04-16 03:20:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][130/1251] eta 0:06:02 lr 0.000607 time 0.2814 (0.3232) loss 3.3702 (3.5961) grad_norm 1.6362 (inf) [2021-04-16 03:20:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][140/1251] eta 0:05:57 lr 0.000607 time 0.2735 (0.3215) loss 3.7089 (3.5677) grad_norm 1.2966 (inf) [2021-04-16 03:20:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][150/1251] eta 0:05:50 lr 0.000607 time 0.2926 (0.3185) loss 3.8422 (3.5725) grad_norm 1.2683 (inf) [2021-04-16 03:20:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][160/1251] eta 0:05:45 lr 0.000607 time 0.2806 (0.3168) loss 4.0496 (3.5714) grad_norm 1.3737 (inf) [2021-04-16 03:20:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][170/1251] eta 0:05:39 lr 0.000607 time 0.2933 (0.3145) loss 3.6502 (3.5810) grad_norm 1.5793 (inf) [2021-04-16 03:20:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][180/1251] eta 0:05:34 lr 0.000607 time 0.2678 (0.3123) loss 3.7254 (3.5848) grad_norm 1.3442 (inf) [2021-04-16 03:20:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][190/1251] eta 0:05:29 lr 0.000607 time 0.2718 (0.3104) loss 4.0368 (3.5720) grad_norm 1.1998 (inf) [2021-04-16 03:20:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][200/1251] eta 0:05:24 lr 0.000607 time 0.2475 (0.3085) loss 3.6812 (3.5647) grad_norm 1.3823 (inf) [2021-04-16 03:20:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][210/1251] eta 0:05:19 lr 0.000607 time 0.2996 (0.3071) loss 4.3554 (3.5808) grad_norm 1.2853 (inf) [2021-04-16 03:20:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][220/1251] eta 0:05:14 lr 0.000607 time 0.2667 (0.3054) loss 3.1331 (3.5857) grad_norm 1.3062 (inf) [2021-04-16 03:21:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][230/1251] eta 0:05:10 lr 0.000607 time 0.2786 (0.3041) loss 4.2107 (3.5995) grad_norm 1.2903 (inf) [2021-04-16 03:21:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][240/1251] eta 0:05:06 lr 0.000607 time 0.2651 (0.3033) loss 3.6262 (3.6016) grad_norm 1.5366 (inf) [2021-04-16 03:21:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][250/1251] eta 0:05:03 lr 0.000607 time 0.2644 (0.3028) loss 4.0208 (3.5931) grad_norm 1.2703 (inf) [2021-04-16 03:21:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][260/1251] eta 0:04:59 lr 0.000607 time 0.4053 (0.3022) loss 4.2632 (3.5965) grad_norm 1.2428 (inf) [2021-04-16 03:21:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][270/1251] eta 0:04:55 lr 0.000607 time 0.2735 (0.3015) loss 3.4735 (3.5972) grad_norm 1.4702 (inf) [2021-04-16 03:21:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][280/1251] eta 0:04:52 lr 0.000607 time 0.2916 (0.3007) loss 3.1082 (3.5853) grad_norm 1.6871 (inf) [2021-04-16 03:21:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][290/1251] eta 0:04:48 lr 0.000607 time 0.2780 (0.3004) loss 3.3244 (3.5888) grad_norm 1.1193 (inf) [2021-04-16 03:21:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][300/1251] eta 0:04:44 lr 0.000607 time 0.2626 (0.2997) loss 4.1558 (3.5916) grad_norm 1.4373 (inf) [2021-04-16 03:21:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][310/1251] eta 0:04:41 lr 0.000607 time 0.2907 (0.2991) loss 3.4343 (3.5929) grad_norm 1.2540 (inf) [2021-04-16 03:21:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][320/1251] eta 0:04:37 lr 0.000607 time 0.2539 (0.2983) loss 3.3888 (3.5889) grad_norm 1.3378 (inf) [2021-04-16 03:21:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][330/1251] eta 0:04:34 lr 0.000607 time 0.2591 (0.2976) loss 4.0206 (3.5970) grad_norm 1.7335 (inf) [2021-04-16 03:21:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][340/1251] eta 0:04:30 lr 0.000607 time 0.3240 (0.2971) loss 3.4623 (3.5974) grad_norm 1.3953 (inf) [2021-04-16 03:21:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][350/1251] eta 0:04:27 lr 0.000606 time 0.2828 (0.2965) loss 3.4636 (3.5967) grad_norm 1.3046 (inf) [2021-04-16 03:21:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][360/1251] eta 0:04:23 lr 0.000606 time 0.3012 (0.2959) loss 4.1118 (3.6032) grad_norm 1.8360 (inf) [2021-04-16 03:21:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][370/1251] eta 0:04:20 lr 0.000606 time 0.2663 (0.2957) loss 3.9943 (3.5961) grad_norm 1.3337 (inf) [2021-04-16 03:21:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][380/1251] eta 0:04:17 lr 0.000606 time 0.2706 (0.2953) loss 3.3501 (3.5881) grad_norm 1.2821 (inf) [2021-04-16 03:21:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][390/1251] eta 0:04:13 lr 0.000606 time 0.2829 (0.2948) loss 3.5767 (3.5839) grad_norm 1.3814 (inf) [2021-04-16 03:21:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][400/1251] eta 0:04:10 lr 0.000606 time 0.2794 (0.2944) loss 2.4311 (3.5798) grad_norm 1.2189 (inf) [2021-04-16 03:21:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][410/1251] eta 0:04:07 lr 0.000606 time 0.2702 (0.2940) loss 4.4729 (3.5761) grad_norm 1.5768 (inf) [2021-04-16 03:21:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][420/1251] eta 0:04:03 lr 0.000606 time 0.2602 (0.2936) loss 3.9870 (3.5732) grad_norm 1.6020 (inf) [2021-04-16 03:21:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][430/1251] eta 0:04:00 lr 0.000606 time 0.3028 (0.2932) loss 3.8774 (3.5750) grad_norm 1.4070 (inf) [2021-04-16 03:21:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][440/1251] eta 0:03:57 lr 0.000606 time 0.2757 (0.2928) loss 3.7763 (3.5752) grad_norm 1.6399 (inf) [2021-04-16 03:22:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][450/1251] eta 0:03:54 lr 0.000606 time 0.2939 (0.2925) loss 4.1131 (3.5787) grad_norm 1.7479 (inf) [2021-04-16 03:22:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][460/1251] eta 0:03:51 lr 0.000606 time 0.2829 (0.2921) loss 2.8783 (3.5732) grad_norm 1.4586 (inf) [2021-04-16 03:22:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][470/1251] eta 0:03:47 lr 0.000606 time 0.2681 (0.2917) loss 3.4942 (3.5722) grad_norm 1.4698 (inf) [2021-04-16 03:22:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][480/1251] eta 0:03:44 lr 0.000606 time 0.2741 (0.2914) loss 3.6472 (3.5736) grad_norm 1.3287 (inf) [2021-04-16 03:22:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][490/1251] eta 0:03:41 lr 0.000606 time 0.2673 (0.2914) loss 3.6635 (3.5728) grad_norm 1.3830 (inf) [2021-04-16 03:22:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][500/1251] eta 0:03:38 lr 0.000606 time 0.2923 (0.2911) loss 2.6069 (3.5767) grad_norm 1.3743 (inf) [2021-04-16 03:22:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][510/1251] eta 0:03:35 lr 0.000606 time 0.2500 (0.2907) loss 3.8749 (3.5754) grad_norm 1.2097 (inf) [2021-04-16 03:22:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][520/1251] eta 0:03:32 lr 0.000606 time 0.3053 (0.2905) loss 3.8397 (3.5824) grad_norm 1.4464 (inf) [2021-04-16 03:22:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][530/1251] eta 0:03:29 lr 0.000606 time 0.2833 (0.2903) loss 3.6370 (3.5796) grad_norm 1.3424 (inf) [2021-04-16 03:22:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][540/1251] eta 0:03:26 lr 0.000606 time 0.2659 (0.2904) loss 3.6057 (3.5807) grad_norm 1.5349 (inf) [2021-04-16 03:22:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][550/1251] eta 0:03:23 lr 0.000606 time 0.2769 (0.2902) loss 3.6872 (3.5800) grad_norm 1.3297 (inf) [2021-04-16 03:22:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][560/1251] eta 0:03:20 lr 0.000606 time 0.2656 (0.2899) loss 3.5022 (3.5768) grad_norm 1.3700 (inf) [2021-04-16 03:22:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][570/1251] eta 0:03:17 lr 0.000606 time 0.4228 (0.2899) loss 3.5861 (3.5760) grad_norm 1.2882 (inf) [2021-04-16 03:22:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][580/1251] eta 0:03:14 lr 0.000606 time 0.2545 (0.2898) loss 3.0652 (3.5747) grad_norm 1.5687 (inf) [2021-04-16 03:22:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][590/1251] eta 0:03:11 lr 0.000606 time 0.4259 (0.2899) loss 3.2165 (3.5750) grad_norm 1.6138 (inf) [2021-04-16 03:22:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][600/1251] eta 0:03:08 lr 0.000605 time 0.2486 (0.2896) loss 3.6657 (3.5726) grad_norm 1.5050 (inf) [2021-04-16 03:22:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][610/1251] eta 0:03:05 lr 0.000605 time 0.2873 (0.2894) loss 2.3509 (3.5694) grad_norm 1.5139 (inf) [2021-04-16 03:22:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][620/1251] eta 0:03:02 lr 0.000605 time 0.2710 (0.2893) loss 3.2934 (3.5694) grad_norm 1.2683 (inf) [2021-04-16 03:22:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][630/1251] eta 0:02:59 lr 0.000605 time 0.2600 (0.2892) loss 3.4515 (3.5693) grad_norm 1.3567 (inf) [2021-04-16 03:22:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][640/1251] eta 0:02:56 lr 0.000605 time 0.2686 (0.2891) loss 3.6703 (3.5724) grad_norm 1.2352 (inf) [2021-04-16 03:22:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][650/1251] eta 0:02:53 lr 0.000605 time 0.2930 (0.2889) loss 3.8049 (3.5683) grad_norm 1.6000 (inf) [2021-04-16 03:23:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][660/1251] eta 0:02:50 lr 0.000605 time 0.2573 (0.2887) loss 3.0337 (3.5699) grad_norm 1.4184 (inf) [2021-04-16 03:23:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][670/1251] eta 0:02:47 lr 0.000605 time 0.2718 (0.2888) loss 3.8963 (3.5704) grad_norm 1.1692 (inf) [2021-04-16 03:23:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][680/1251] eta 0:02:44 lr 0.000605 time 0.2887 (0.2886) loss 4.4424 (3.5757) grad_norm 1.4063 (inf) [2021-04-16 03:23:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][690/1251] eta 0:02:41 lr 0.000605 time 0.2908 (0.2884) loss 4.2133 (3.5772) grad_norm 1.8515 (inf) [2021-04-16 03:23:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][700/1251] eta 0:02:38 lr 0.000605 time 0.2890 (0.2883) loss 3.7342 (3.5812) grad_norm 1.3047 (inf) [2021-04-16 03:23:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][710/1251] eta 0:02:35 lr 0.000605 time 0.2744 (0.2881) loss 2.5926 (3.5836) grad_norm 1.6126 (inf) [2021-04-16 03:23:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][720/1251] eta 0:02:32 lr 0.000605 time 0.2709 (0.2881) loss 3.9406 (3.5812) grad_norm 1.3418 (inf) [2021-04-16 03:23:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][730/1251] eta 0:02:30 lr 0.000605 time 0.3652 (0.2880) loss 4.1758 (3.5829) grad_norm 1.3479 (inf) [2021-04-16 03:23:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][740/1251] eta 0:02:27 lr 0.000605 time 0.2672 (0.2878) loss 3.9075 (3.5854) grad_norm 1.9128 (inf) [2021-04-16 03:23:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][750/1251] eta 0:02:24 lr 0.000605 time 0.2814 (0.2876) loss 3.5993 (3.5876) grad_norm 1.5789 (inf) [2021-04-16 03:23:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][760/1251] eta 0:02:21 lr 0.000605 time 0.3037 (0.2876) loss 3.9362 (3.5898) grad_norm 1.3126 (inf) [2021-04-16 03:23:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][770/1251] eta 0:02:18 lr 0.000605 time 0.2912 (0.2876) loss 3.6485 (3.5916) grad_norm 1.1709 (inf) [2021-04-16 03:23:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][780/1251] eta 0:02:15 lr 0.000605 time 0.2823 (0.2875) loss 3.0776 (3.5927) grad_norm 1.3817 (inf) [2021-04-16 03:23:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][790/1251] eta 0:02:12 lr 0.000605 time 0.2659 (0.2874) loss 3.4267 (3.5922) grad_norm 1.4524 (inf) [2021-04-16 03:23:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][800/1251] eta 0:02:09 lr 0.000605 time 0.2768 (0.2873) loss 4.1044 (3.5921) grad_norm 1.5355 (inf) [2021-04-16 03:23:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][810/1251] eta 0:02:06 lr 0.000605 time 0.2680 (0.2871) loss 3.2181 (3.5889) grad_norm 1.2868 (inf) [2021-04-16 03:23:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][820/1251] eta 0:02:03 lr 0.000605 time 0.2774 (0.2870) loss 3.3306 (3.5899) grad_norm 1.4970 (inf) [2021-04-16 03:23:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][830/1251] eta 0:02:00 lr 0.000605 time 0.2743 (0.2870) loss 3.8611 (3.5923) grad_norm 1.2134 (inf) [2021-04-16 03:23:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][840/1251] eta 0:01:57 lr 0.000605 time 0.2419 (0.2870) loss 3.1778 (3.5885) grad_norm 1.5249 (inf) [2021-04-16 03:23:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][850/1251] eta 0:01:55 lr 0.000604 time 0.2949 (0.2869) loss 3.4606 (3.5864) grad_norm 1.1839 (inf) [2021-04-16 03:23:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][860/1251] eta 0:01:52 lr 0.000604 time 0.2869 (0.2869) loss 4.2465 (3.5871) grad_norm 1.3396 (inf) [2021-04-16 03:23:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][870/1251] eta 0:01:49 lr 0.000604 time 0.2873 (0.2868) loss 3.3169 (3.5853) grad_norm 1.4045 (inf) [2021-04-16 03:24:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][880/1251] eta 0:01:46 lr 0.000604 time 0.2743 (0.2867) loss 3.5397 (3.5829) grad_norm 1.3221 (inf) [2021-04-16 03:24:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][890/1251] eta 0:01:43 lr 0.000604 time 0.2723 (0.2865) loss 3.5663 (3.5822) grad_norm 1.4635 (inf) [2021-04-16 03:24:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][900/1251] eta 0:01:40 lr 0.000604 time 0.2619 (0.2864) loss 4.2569 (3.5807) grad_norm 1.3246 (inf) [2021-04-16 03:24:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][910/1251] eta 0:01:37 lr 0.000604 time 0.2783 (0.2863) loss 3.5014 (3.5819) grad_norm 1.4258 (inf) [2021-04-16 03:24:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][920/1251] eta 0:01:34 lr 0.000604 time 0.2642 (0.2862) loss 3.4881 (3.5806) grad_norm 1.2627 (inf) [2021-04-16 03:24:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][930/1251] eta 0:01:31 lr 0.000604 time 0.2608 (0.2861) loss 3.7188 (3.5789) grad_norm 1.7304 (inf) [2021-04-16 03:24:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][940/1251] eta 0:01:28 lr 0.000604 time 0.2837 (0.2860) loss 4.1358 (3.5800) grad_norm 1.1886 (inf) [2021-04-16 03:24:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][950/1251] eta 0:01:26 lr 0.000604 time 0.2778 (0.2860) loss 4.1326 (3.5815) grad_norm 1.4651 (inf) [2021-04-16 03:24:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][960/1251] eta 0:01:23 lr 0.000604 time 0.2640 (0.2859) loss 3.9710 (3.5812) grad_norm 1.2341 (inf) [2021-04-16 03:24:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][970/1251] eta 0:01:20 lr 0.000604 time 0.2925 (0.2858) loss 3.0382 (3.5818) grad_norm 1.5345 (inf) [2021-04-16 03:24:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][980/1251] eta 0:01:17 lr 0.000604 time 0.2574 (0.2857) loss 3.7819 (3.5834) grad_norm 1.3035 (inf) [2021-04-16 03:24:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][990/1251] eta 0:01:14 lr 0.000604 time 0.2930 (0.2856) loss 3.6435 (3.5834) grad_norm 1.5641 (inf) [2021-04-16 03:24:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][1000/1251] eta 0:01:11 lr 0.000604 time 0.2418 (0.2855) loss 3.4204 (3.5827) grad_norm 1.3473 (inf) [2021-04-16 03:24:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][1010/1251] eta 0:01:08 lr 0.000604 time 0.3081 (0.2854) loss 3.8738 (3.5852) grad_norm 1.3297 (inf) [2021-04-16 03:24:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][1020/1251] eta 0:01:05 lr 0.000604 time 0.2961 (0.2853) loss 3.1474 (3.5855) grad_norm 1.2963 (inf) [2021-04-16 03:24:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][1030/1251] eta 0:01:03 lr 0.000604 time 0.2682 (0.2852) loss 3.9008 (3.5843) grad_norm 1.3130 (inf) [2021-04-16 03:24:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][1040/1251] eta 0:01:00 lr 0.000604 time 0.2637 (0.2851) loss 3.1522 (3.5866) grad_norm 1.6215 (inf) [2021-04-16 03:24:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][1050/1251] eta 0:00:57 lr 0.000604 time 0.2648 (0.2850) loss 3.9076 (3.5844) grad_norm 1.3568 (inf) [2021-04-16 03:24:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][1060/1251] eta 0:00:54 lr 0.000604 time 0.2997 (0.2850) loss 3.2530 (3.5846) grad_norm 1.5387 (inf) [2021-04-16 03:24:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][1070/1251] eta 0:00:51 lr 0.000604 time 0.2872 (0.2850) loss 4.5578 (3.5835) grad_norm 1.3823 (inf) [2021-04-16 03:24:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][1080/1251] eta 0:00:48 lr 0.000604 time 0.2777 (0.2849) loss 3.0670 (3.5830) grad_norm 1.4934 (inf) [2021-04-16 03:25:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][1090/1251] eta 0:00:45 lr 0.000603 time 0.2793 (0.2849) loss 4.1951 (3.5829) grad_norm 1.5229 (inf) [2021-04-16 03:25:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][1100/1251] eta 0:00:43 lr 0.000603 time 0.2915 (0.2848) loss 2.9959 (3.5811) grad_norm 1.3948 (inf) [2021-04-16 03:25:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][1110/1251] eta 0:00:40 lr 0.000603 time 0.2807 (0.2847) loss 4.4651 (3.5789) grad_norm 1.2522 (inf) [2021-04-16 03:25:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][1120/1251] eta 0:00:37 lr 0.000603 time 0.2656 (0.2848) loss 3.7501 (3.5786) grad_norm 1.2564 (inf) [2021-04-16 03:25:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][1130/1251] eta 0:00:34 lr 0.000603 time 0.2609 (0.2847) loss 4.2339 (3.5827) grad_norm 1.4618 (inf) [2021-04-16 03:25:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][1140/1251] eta 0:00:31 lr 0.000603 time 0.2654 (0.2846) loss 4.1236 (3.5823) grad_norm 1.5510 (inf) [2021-04-16 03:25:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][1150/1251] eta 0:00:28 lr 0.000603 time 0.4116 (0.2848) loss 3.4387 (3.5823) grad_norm 1.6136 (inf) [2021-04-16 03:25:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][1160/1251] eta 0:00:25 lr 0.000603 time 0.2821 (0.2847) loss 3.8420 (3.5822) grad_norm 1.5674 (inf) [2021-04-16 03:25:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][1170/1251] eta 0:00:23 lr 0.000603 time 0.2616 (0.2848) loss 2.7527 (3.5814) grad_norm 1.4636 (inf) [2021-04-16 03:25:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][1180/1251] eta 0:00:20 lr 0.000603 time 0.2743 (0.2847) loss 2.9594 (3.5824) grad_norm 1.2972 (inf) [2021-04-16 03:25:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][1190/1251] eta 0:00:17 lr 0.000603 time 0.2708 (0.2847) loss 2.8941 (3.5807) grad_norm 1.4886 (inf) [2021-04-16 03:25:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][1200/1251] eta 0:00:14 lr 0.000603 time 0.2922 (0.2846) loss 3.7343 (3.5782) grad_norm 1.2542 (inf) [2021-04-16 03:25:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][1210/1251] eta 0:00:11 lr 0.000603 time 0.2998 (0.2846) loss 3.0835 (3.5769) grad_norm 1.2873 (inf) [2021-04-16 03:25:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][1220/1251] eta 0:00:08 lr 0.000603 time 0.2855 (0.2845) loss 4.2620 (3.5794) grad_norm 1.3945 (inf) [2021-04-16 03:25:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][1230/1251] eta 0:00:05 lr 0.000603 time 0.2843 (0.2844) loss 3.8495 (3.5798) grad_norm 1.2487 (inf) [2021-04-16 03:25:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][1240/1251] eta 0:00:03 lr 0.000603 time 0.2556 (0.2844) loss 2.9918 (3.5784) grad_norm 1.6213 (inf) [2021-04-16 03:25:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [130/300][1250/1251] eta 0:00:00 lr 0.000603 time 0.2767 (0.2841) loss 3.6582 (3.5783) grad_norm 1.3694 (inf) [2021-04-16 03:25:48 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 130 training takes 0:05:58 [2021-04-16 03:25:48 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_130.pth saving...... [2021-04-16 03:25:56 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_130.pth saved !!! [2021-04-16 03:25:57 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.138 (1.138) Loss 1.0726 (1.0726) Acc@1 74.316 (74.316) Acc@5 92.285 (92.285) [2021-04-16 03:25:59 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.116 (0.228) Loss 1.0756 (1.0685) Acc@1 75.098 (74.956) Acc@5 93.359 (92.543) [2021-04-16 03:26:01 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.181 (0.208) Loss 1.0613 (1.0659) Acc@1 75.391 (74.991) Acc@5 92.969 (92.597) [2021-04-16 03:26:03 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.458 (0.232) Loss 1.0121 (1.0654) Acc@1 76.660 (75.060) Acc@5 93.555 (92.666) [2021-04-16 03:26:05 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.238 (0.219) Loss 0.9995 (1.0647) Acc@1 75.879 (74.988) Acc@5 94.336 (92.695) [2021-04-16 03:26:08 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 74.950 Acc@5 92.708 [2021-04-16 03:26:08 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 75.0% [2021-04-16 03:26:08 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 74.95% [2021-04-16 03:26:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][0/1251] eta 2:02:12 lr 0.000603 time 5.8617 (5.8617) loss 3.5918 (3.5918) grad_norm 1.3192 (1.3192) [2021-04-16 03:26:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][10/1251] eta 0:16:09 lr 0.000603 time 0.2747 (0.7814) loss 3.6131 (3.5679) grad_norm 1.2308 (1.3056) [2021-04-16 03:26:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][20/1251] eta 0:11:08 lr 0.000603 time 0.2839 (0.5427) loss 3.5504 (3.4463) grad_norm 1.4522 (1.3305) [2021-04-16 03:26:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][30/1251] eta 0:09:25 lr 0.000603 time 0.2706 (0.4631) loss 4.0671 (3.5889) grad_norm 1.4032 (1.3535) [2021-04-16 03:26:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][40/1251] eta 0:08:28 lr 0.000603 time 0.2698 (0.4197) loss 4.3798 (3.6130) grad_norm 1.4263 (1.3695) [2021-04-16 03:26:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][50/1251] eta 0:07:51 lr 0.000603 time 0.2721 (0.3924) loss 3.1639 (3.6044) grad_norm 1.5457 (1.3947) [2021-04-16 03:26:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][60/1251] eta 0:07:26 lr 0.000603 time 0.2922 (0.3748) loss 3.7880 (3.5679) grad_norm 1.3725 (1.3938) [2021-04-16 03:26:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][70/1251] eta 0:07:07 lr 0.000603 time 0.3085 (0.3616) loss 3.7499 (3.6151) grad_norm 1.3295 (1.4027) [2021-04-16 03:26:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][80/1251] eta 0:06:52 lr 0.000603 time 0.2598 (0.3522) loss 2.8546 (3.5861) grad_norm 1.4637 (1.4086) [2021-04-16 03:26:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][90/1251] eta 0:06:39 lr 0.000602 time 0.2796 (0.3440) loss 3.0740 (3.5745) grad_norm 1.2569 (1.4230) [2021-04-16 03:26:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][100/1251] eta 0:06:29 lr 0.000602 time 0.2804 (0.3381) loss 4.1361 (3.5909) grad_norm 1.2840 (1.4259) [2021-04-16 03:26:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][110/1251] eta 0:06:19 lr 0.000602 time 0.2835 (0.3326) loss 2.7687 (3.5476) grad_norm 1.3951 (1.4245) [2021-04-16 03:26:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][120/1251] eta 0:06:11 lr 0.000602 time 0.2728 (0.3288) loss 4.4102 (3.5522) grad_norm 1.1656 (1.4154) [2021-04-16 03:26:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][130/1251] eta 0:06:04 lr 0.000602 time 0.2786 (0.3252) loss 4.2599 (3.5553) grad_norm 1.3389 (1.4100) [2021-04-16 03:26:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][140/1251] eta 0:05:57 lr 0.000602 time 0.2660 (0.3215) loss 3.8508 (3.5665) grad_norm 1.4936 (1.4104) [2021-04-16 03:26:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][150/1251] eta 0:05:51 lr 0.000602 time 0.2856 (0.3189) loss 4.4489 (3.5429) grad_norm 1.3510 (1.4187) [2021-04-16 03:26:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][160/1251] eta 0:05:45 lr 0.000602 time 0.2557 (0.3170) loss 4.7390 (3.5615) grad_norm 1.4807 (1.4190) [2021-04-16 03:27:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][170/1251] eta 0:05:40 lr 0.000602 time 0.2801 (0.3146) loss 4.1508 (3.5684) grad_norm 1.4413 (1.4165) [2021-04-16 03:27:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][180/1251] eta 0:05:34 lr 0.000602 time 0.2740 (0.3128) loss 3.6637 (3.5773) grad_norm 1.2856 (1.4175) [2021-04-16 03:27:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][190/1251] eta 0:05:29 lr 0.000602 time 0.3037 (0.3109) loss 3.8487 (3.5790) grad_norm 1.3212 (1.4167) [2021-04-16 03:27:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][200/1251] eta 0:05:25 lr 0.000602 time 0.2754 (0.3096) loss 3.0552 (3.5727) grad_norm 1.3310 (1.4160) [2021-04-16 03:27:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][210/1251] eta 0:05:20 lr 0.000602 time 0.2938 (0.3082) loss 3.3806 (3.5679) grad_norm 1.6384 (1.4154) [2021-04-16 03:27:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][220/1251] eta 0:05:16 lr 0.000602 time 0.2784 (0.3067) loss 3.7530 (3.5774) grad_norm 1.4382 (1.4159) [2021-04-16 03:27:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][230/1251] eta 0:05:11 lr 0.000602 time 0.2736 (0.3053) loss 3.9484 (3.5703) grad_norm 1.4277 (1.4135) [2021-04-16 03:27:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][240/1251] eta 0:05:07 lr 0.000602 time 0.2767 (0.3039) loss 4.1934 (3.5725) grad_norm 1.9812 (1.4161) [2021-04-16 03:27:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][250/1251] eta 0:05:03 lr 0.000602 time 0.2968 (0.3030) loss 3.4783 (3.5693) grad_norm 1.4743 (1.4157) [2021-04-16 03:27:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][260/1251] eta 0:04:59 lr 0.000602 time 0.2586 (0.3019) loss 3.8474 (3.5600) grad_norm 1.4730 (1.4138) [2021-04-16 03:27:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][270/1251] eta 0:04:55 lr 0.000602 time 0.2745 (0.3009) loss 3.9581 (3.5706) grad_norm 1.3169 (1.4135) [2021-04-16 03:27:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][280/1251] eta 0:04:51 lr 0.000602 time 0.2970 (0.3004) loss 4.3176 (3.5715) grad_norm 1.3401 (1.4121) [2021-04-16 03:27:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][290/1251] eta 0:04:48 lr 0.000602 time 0.2575 (0.2998) loss 3.7501 (3.5708) grad_norm 1.3939 (1.4127) [2021-04-16 03:27:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][300/1251] eta 0:04:44 lr 0.000602 time 0.2664 (0.2992) loss 3.8395 (3.5652) grad_norm 1.4439 (1.4118) [2021-04-16 03:27:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][310/1251] eta 0:04:41 lr 0.000602 time 0.2916 (0.2987) loss 2.4522 (3.5676) grad_norm 1.4366 (1.4127) [2021-04-16 03:27:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][320/1251] eta 0:04:37 lr 0.000602 time 0.2611 (0.2984) loss 3.6991 (3.5758) grad_norm 1.5424 (1.4142) [2021-04-16 03:27:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][330/1251] eta 0:04:34 lr 0.000601 time 0.2758 (0.2976) loss 4.0367 (3.5870) grad_norm 1.3638 (1.4163) [2021-04-16 03:27:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][340/1251] eta 0:04:30 lr 0.000601 time 0.2612 (0.2969) loss 4.5522 (3.5912) grad_norm 1.5917 (1.4190) [2021-04-16 03:27:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][350/1251] eta 0:04:26 lr 0.000601 time 0.2454 (0.2962) loss 2.7326 (3.5818) grad_norm 1.5731 (1.4177) [2021-04-16 03:27:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][360/1251] eta 0:04:23 lr 0.000601 time 0.2611 (0.2960) loss 3.9879 (3.5811) grad_norm 1.3733 (1.4186) [2021-04-16 03:27:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][370/1251] eta 0:04:20 lr 0.000601 time 0.2913 (0.2960) loss 2.3961 (3.5833) grad_norm 1.5680 (1.4204) [2021-04-16 03:28:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][380/1251] eta 0:04:17 lr 0.000601 time 0.2674 (0.2955) loss 3.1229 (3.5757) grad_norm 1.2482 (1.4220) [2021-04-16 03:28:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][390/1251] eta 0:04:14 lr 0.000601 time 0.2720 (0.2956) loss 3.0313 (3.5790) grad_norm 1.5895 (1.4233) [2021-04-16 03:28:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][400/1251] eta 0:04:11 lr 0.000601 time 0.3126 (0.2953) loss 2.7786 (3.5692) grad_norm 1.4710 (1.4227) [2021-04-16 03:28:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][410/1251] eta 0:04:08 lr 0.000601 time 0.2980 (0.2950) loss 3.7329 (3.5634) grad_norm 1.3216 (1.4217) [2021-04-16 03:28:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][420/1251] eta 0:04:04 lr 0.000601 time 0.2831 (0.2946) loss 4.2026 (3.5636) grad_norm 1.5353 (1.4211) [2021-04-16 03:28:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][430/1251] eta 0:04:01 lr 0.000601 time 0.3142 (0.2942) loss 3.9718 (3.5644) grad_norm 1.3643 (1.4193) [2021-04-16 03:28:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][440/1251] eta 0:03:58 lr 0.000601 time 0.2891 (0.2942) loss 3.0004 (3.5658) grad_norm 1.2819 (1.4196) [2021-04-16 03:28:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][450/1251] eta 0:03:55 lr 0.000601 time 0.2843 (0.2939) loss 3.4624 (3.5628) grad_norm 1.8164 (1.4206) [2021-04-16 03:28:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][460/1251] eta 0:03:52 lr 0.000601 time 0.2717 (0.2936) loss 4.0471 (3.5689) grad_norm 1.4288 (1.4195) [2021-04-16 03:28:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][470/1251] eta 0:03:49 lr 0.000601 time 0.3167 (0.2933) loss 3.6632 (3.5709) grad_norm 1.4266 (1.4190) [2021-04-16 03:28:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][480/1251] eta 0:03:46 lr 0.000601 time 0.3007 (0.2932) loss 3.8384 (3.5760) grad_norm 1.4739 (1.4192) [2021-04-16 03:28:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][490/1251] eta 0:03:42 lr 0.000601 time 0.2770 (0.2929) loss 2.3605 (3.5720) grad_norm 1.4853 (1.4182) [2021-04-16 03:28:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][500/1251] eta 0:03:39 lr 0.000601 time 0.2803 (0.2927) loss 3.6755 (3.5774) grad_norm 1.3968 (1.4165) [2021-04-16 03:28:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][510/1251] eta 0:03:36 lr 0.000601 time 0.2854 (0.2924) loss 2.9634 (3.5769) grad_norm 1.6251 (1.4167) [2021-04-16 03:28:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][520/1251] eta 0:03:33 lr 0.000601 time 0.3079 (0.2922) loss 3.4950 (3.5775) grad_norm 1.3801 (1.4163) [2021-04-16 03:28:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][530/1251] eta 0:03:30 lr 0.000601 time 0.2739 (0.2919) loss 3.7128 (3.5740) grad_norm 1.3853 (1.4174) [2021-04-16 03:28:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][540/1251] eta 0:03:27 lr 0.000601 time 0.2510 (0.2916) loss 2.6092 (3.5692) grad_norm 1.3714 (1.4164) [2021-04-16 03:28:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][550/1251] eta 0:03:24 lr 0.000601 time 0.2817 (0.2914) loss 2.4275 (3.5681) grad_norm 1.8417 (1.4197) [2021-04-16 03:28:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][560/1251] eta 0:03:21 lr 0.000601 time 0.2637 (0.2912) loss 3.2987 (3.5696) grad_norm 1.4357 (1.4205) [2021-04-16 03:28:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][570/1251] eta 0:03:18 lr 0.000601 time 0.2664 (0.2910) loss 4.0896 (3.5719) grad_norm 1.3921 (1.4209) [2021-04-16 03:28:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][580/1251] eta 0:03:15 lr 0.000600 time 0.2706 (0.2913) loss 3.3204 (3.5760) grad_norm 1.2160 (1.4197) [2021-04-16 03:29:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][590/1251] eta 0:03:12 lr 0.000600 time 0.3009 (0.2910) loss 4.0044 (3.5770) grad_norm 1.5229 (1.4199) [2021-04-16 03:29:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][600/1251] eta 0:03:09 lr 0.000600 time 0.4384 (0.2910) loss 3.2335 (3.5807) grad_norm 1.4437 (1.4200) [2021-04-16 03:29:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][610/1251] eta 0:03:06 lr 0.000600 time 0.2674 (0.2908) loss 3.3069 (3.5794) grad_norm 1.2645 (1.4188) [2021-04-16 03:29:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][620/1251] eta 0:03:03 lr 0.000600 time 0.2550 (0.2907) loss 2.4878 (3.5808) grad_norm 1.3671 (1.4196) [2021-04-16 03:29:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][630/1251] eta 0:03:00 lr 0.000600 time 0.2945 (0.2905) loss 2.5419 (3.5777) grad_norm 1.9493 (1.4210) [2021-04-16 03:29:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][640/1251] eta 0:02:57 lr 0.000600 time 0.2774 (0.2903) loss 3.8499 (3.5789) grad_norm 1.2711 (1.4219) [2021-04-16 03:29:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][650/1251] eta 0:02:54 lr 0.000600 time 0.2700 (0.2902) loss 3.9774 (3.5810) grad_norm 1.4136 (1.4217) [2021-04-16 03:29:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][660/1251] eta 0:02:51 lr 0.000600 time 0.2690 (0.2900) loss 3.8154 (3.5808) grad_norm 1.3074 (1.4210) [2021-04-16 03:29:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][670/1251] eta 0:02:48 lr 0.000600 time 0.2909 (0.2898) loss 3.5790 (3.5785) grad_norm 1.2397 (1.4213) [2021-04-16 03:29:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][680/1251] eta 0:02:45 lr 0.000600 time 0.3257 (0.2897) loss 3.4360 (3.5834) grad_norm 1.4079 (1.4199) [2021-04-16 03:29:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][690/1251] eta 0:02:42 lr 0.000600 time 0.2583 (0.2895) loss 3.5575 (3.5810) grad_norm 1.3883 (1.4194) [2021-04-16 03:29:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][700/1251] eta 0:02:39 lr 0.000600 time 0.2756 (0.2893) loss 3.2977 (3.5856) grad_norm 1.3410 (1.4194) [2021-04-16 03:29:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][710/1251] eta 0:02:36 lr 0.000600 time 0.2810 (0.2892) loss 3.9393 (3.5889) grad_norm 1.2309 (1.4196) [2021-04-16 03:29:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][720/1251] eta 0:02:33 lr 0.000600 time 0.2778 (0.2890) loss 3.5618 (3.5891) grad_norm 1.5880 (1.4191) [2021-04-16 03:29:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][730/1251] eta 0:02:30 lr 0.000600 time 0.2779 (0.2890) loss 4.3203 (3.5904) grad_norm 1.3245 (1.4188) [2021-04-16 03:29:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][740/1251] eta 0:02:27 lr 0.000600 time 0.2631 (0.2889) loss 3.8440 (3.5944) grad_norm 1.3805 (1.4197) [2021-04-16 03:29:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][750/1251] eta 0:02:24 lr 0.000600 time 0.2750 (0.2888) loss 4.6310 (3.5922) grad_norm 1.6380 (1.4195) [2021-04-16 03:29:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][760/1251] eta 0:02:21 lr 0.000600 time 0.2905 (0.2887) loss 2.6592 (3.5904) grad_norm 1.3150 (1.4199) [2021-04-16 03:29:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][770/1251] eta 0:02:18 lr 0.000600 time 0.2924 (0.2885) loss 4.0356 (3.5889) grad_norm 1.4158 (1.4189) [2021-04-16 03:29:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][780/1251] eta 0:02:15 lr 0.000600 time 0.2773 (0.2885) loss 3.7213 (3.5883) grad_norm 1.3814 (1.4183) [2021-04-16 03:29:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][790/1251] eta 0:02:12 lr 0.000600 time 0.2732 (0.2883) loss 3.7380 (3.5891) grad_norm 1.3314 (1.4181) [2021-04-16 03:29:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][800/1251] eta 0:02:09 lr 0.000600 time 0.2896 (0.2882) loss 3.9480 (3.5878) grad_norm 1.3504 (1.4176) [2021-04-16 03:30:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][810/1251] eta 0:02:07 lr 0.000600 time 0.2585 (0.2881) loss 3.4812 (3.5915) grad_norm 1.3718 (1.4169) [2021-04-16 03:30:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][820/1251] eta 0:02:04 lr 0.000600 time 0.2675 (0.2879) loss 2.1482 (3.5887) grad_norm 1.4384 (1.4162) [2021-04-16 03:30:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][830/1251] eta 0:02:01 lr 0.000599 time 0.2600 (0.2878) loss 3.8301 (3.5869) grad_norm 1.5983 (1.4160) [2021-04-16 03:30:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][840/1251] eta 0:01:58 lr 0.000599 time 0.2791 (0.2877) loss 4.3567 (3.5881) grad_norm 1.3955 (1.4150) [2021-04-16 03:30:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][850/1251] eta 0:01:55 lr 0.000599 time 0.2740 (0.2875) loss 3.4699 (3.5894) grad_norm 1.5113 (1.4154) [2021-04-16 03:30:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][860/1251] eta 0:01:52 lr 0.000599 time 0.2686 (0.2874) loss 3.9439 (3.5884) grad_norm 1.6620 (1.4157) [2021-04-16 03:30:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][870/1251] eta 0:01:49 lr 0.000599 time 0.2817 (0.2874) loss 3.7088 (3.5858) grad_norm 1.5306 (1.4163) [2021-04-16 03:30:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][880/1251] eta 0:01:46 lr 0.000599 time 0.2882 (0.2872) loss 3.5427 (3.5879) grad_norm 1.4640 (1.4163) [2021-04-16 03:30:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][890/1251] eta 0:01:43 lr 0.000599 time 0.2873 (0.2873) loss 4.7429 (3.5926) grad_norm 1.4955 (1.4167) [2021-04-16 03:30:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][900/1251] eta 0:01:40 lr 0.000599 time 0.2661 (0.2872) loss 3.8987 (3.5943) grad_norm 1.4269 (1.4173) [2021-04-16 03:30:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][910/1251] eta 0:01:37 lr 0.000599 time 0.2791 (0.2871) loss 3.4797 (3.5937) grad_norm 1.3267 (1.4175) [2021-04-16 03:30:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][920/1251] eta 0:01:35 lr 0.000599 time 0.2709 (0.2871) loss 3.5877 (3.5962) grad_norm 1.4228 (1.4197) [2021-04-16 03:30:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][930/1251] eta 0:01:32 lr 0.000599 time 0.2629 (0.2870) loss 3.3239 (3.5973) grad_norm 1.4370 (1.4202) [2021-04-16 03:30:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][940/1251] eta 0:01:29 lr 0.000599 time 0.2744 (0.2871) loss 4.7490 (3.5998) grad_norm 1.3680 (1.4202) [2021-04-16 03:30:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][950/1251] eta 0:01:26 lr 0.000599 time 0.2758 (0.2872) loss 3.1187 (3.5975) grad_norm 1.2396 (1.4193) [2021-04-16 03:30:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][960/1251] eta 0:01:23 lr 0.000599 time 0.2945 (0.2870) loss 4.2070 (3.5980) grad_norm 1.5126 (1.4192) [2021-04-16 03:30:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][970/1251] eta 0:01:20 lr 0.000599 time 0.2749 (0.2870) loss 3.6864 (3.5982) grad_norm 1.5919 (1.4205) [2021-04-16 03:30:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][980/1251] eta 0:01:17 lr 0.000599 time 0.2743 (0.2869) loss 3.3816 (3.5954) grad_norm 1.3007 (1.4201) [2021-04-16 03:30:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][990/1251] eta 0:01:14 lr 0.000599 time 0.2927 (0.2868) loss 4.1075 (3.5970) grad_norm 1.4424 (1.4197) [2021-04-16 03:30:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][1000/1251] eta 0:01:11 lr 0.000599 time 0.3057 (0.2867) loss 3.7071 (3.5976) grad_norm 1.7012 (1.4203) [2021-04-16 03:30:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][1010/1251] eta 0:01:09 lr 0.000599 time 0.2782 (0.2866) loss 2.4478 (3.5976) grad_norm 1.3385 (1.4216) [2021-04-16 03:31:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][1020/1251] eta 0:01:06 lr 0.000599 time 0.3011 (0.2867) loss 2.9264 (3.5967) grad_norm 1.3410 (1.4225) [2021-04-16 03:31:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][1030/1251] eta 0:01:03 lr 0.000599 time 0.3173 (0.2867) loss 3.7196 (3.5972) grad_norm 1.4507 (1.4222) [2021-04-16 03:31:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][1040/1251] eta 0:01:00 lr 0.000599 time 0.2678 (0.2866) loss 2.7404 (3.5953) grad_norm 1.2863 (1.4222) [2021-04-16 03:31:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][1050/1251] eta 0:00:57 lr 0.000599 time 0.2901 (0.2866) loss 2.9910 (3.5964) grad_norm 1.2189 (1.4221) [2021-04-16 03:31:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][1060/1251] eta 0:00:54 lr 0.000599 time 0.2661 (0.2865) loss 3.9766 (3.5989) grad_norm 1.3615 (1.4227) [2021-04-16 03:31:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][1070/1251] eta 0:00:51 lr 0.000598 time 0.2931 (0.2864) loss 2.3985 (3.5974) grad_norm 1.2502 (1.4231) [2021-04-16 03:31:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][1080/1251] eta 0:00:48 lr 0.000598 time 0.2634 (0.2863) loss 3.6535 (3.5989) grad_norm 1.2213 (1.4231) [2021-04-16 03:31:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][1090/1251] eta 0:00:46 lr 0.000598 time 0.2753 (0.2862) loss 3.9701 (3.6003) grad_norm 1.4055 (1.4226) [2021-04-16 03:31:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][1100/1251] eta 0:00:43 lr 0.000598 time 0.2803 (0.2861) loss 4.1326 (3.6030) grad_norm 1.8554 (1.4226) [2021-04-16 03:31:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][1110/1251] eta 0:00:40 lr 0.000598 time 0.2922 (0.2861) loss 3.2668 (3.6043) grad_norm 1.4474 (1.4231) [2021-04-16 03:31:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][1120/1251] eta 0:00:37 lr 0.000598 time 0.2731 (0.2860) loss 3.2905 (3.6057) grad_norm 1.3841 (1.4227) [2021-04-16 03:31:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][1130/1251] eta 0:00:34 lr 0.000598 time 0.2831 (0.2860) loss 3.7312 (3.6046) grad_norm 1.5997 (1.4232) [2021-04-16 03:31:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][1140/1251] eta 0:00:31 lr 0.000598 time 0.2753 (0.2859) loss 2.8922 (3.6041) grad_norm 1.2507 (1.4229) [2021-04-16 03:31:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][1150/1251] eta 0:00:28 lr 0.000598 time 0.2684 (0.2858) loss 3.1672 (3.6017) grad_norm 1.3816 (1.4231) [2021-04-16 03:31:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][1160/1251] eta 0:00:26 lr 0.000598 time 0.2422 (0.2860) loss 3.4411 (3.6013) grad_norm 1.7101 (1.4237) [2021-04-16 03:31:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][1170/1251] eta 0:00:23 lr 0.000598 time 0.2578 (0.2859) loss 2.6653 (3.6001) grad_norm 1.5903 (1.4236) [2021-04-16 03:31:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][1180/1251] eta 0:00:20 lr 0.000598 time 0.2853 (0.2860) loss 3.5383 (3.5987) grad_norm 1.3709 (1.4239) [2021-04-16 03:31:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][1190/1251] eta 0:00:17 lr 0.000598 time 0.2801 (0.2860) loss 3.6191 (3.5987) grad_norm 1.5154 (1.4245) [2021-04-16 03:31:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][1200/1251] eta 0:00:14 lr 0.000598 time 0.2653 (0.2859) loss 3.5915 (3.5984) grad_norm 1.5959 (1.4245) [2021-04-16 03:31:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][1210/1251] eta 0:00:11 lr 0.000598 time 0.3028 (0.2858) loss 3.8240 (3.5981) grad_norm 1.2885 (1.4243) [2021-04-16 03:31:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][1220/1251] eta 0:00:08 lr 0.000598 time 0.2648 (0.2857) loss 3.6702 (3.5999) grad_norm 1.2879 (1.4245) [2021-04-16 03:32:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][1230/1251] eta 0:00:05 lr 0.000598 time 0.2623 (0.2857) loss 3.4022 (3.6003) grad_norm 1.4523 (1.4243) [2021-04-16 03:32:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][1240/1251] eta 0:00:03 lr 0.000598 time 0.2484 (0.2855) loss 2.5247 (3.5998) grad_norm 1.7724 (1.4251) [2021-04-16 03:32:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [131/300][1250/1251] eta 0:00:00 lr 0.000598 time 0.2484 (0.2853) loss 2.9740 (3.6000) grad_norm 1.3846 (1.4244) [2021-04-16 03:32:09 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 131 training takes 0:06:00 [2021-04-16 03:32:09 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_131.pth saving...... [2021-04-16 03:32:23 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_131.pth saved !!! [2021-04-16 03:32:24 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.142 (1.142) Loss 1.0378 (1.0378) Acc@1 76.270 (76.270) Acc@5 92.773 (92.773) [2021-04-16 03:32:25 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.223 (0.216) Loss 1.1427 (1.0746) Acc@1 71.973 (74.805) Acc@5 92.285 (92.738) [2021-04-16 03:32:28 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.112 (0.234) Loss 1.0971 (1.0854) Acc@1 74.219 (74.693) Acc@5 92.090 (92.639) [2021-04-16 03:32:30 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.116 (0.232) Loss 1.0893 (1.0818) Acc@1 75.195 (74.839) Acc@5 91.797 (92.660) [2021-04-16 03:32:32 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.351 (0.218) Loss 1.1295 (1.0766) Acc@1 73.535 (74.871) Acc@5 91.895 (92.738) [2021-04-16 03:32:41 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 74.978 Acc@5 92.814 [2021-04-16 03:32:41 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 75.0% [2021-04-16 03:32:41 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 74.98% [2021-04-16 03:32:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][0/1251] eta 0:43:35 lr 0.000598 time 2.0905 (2.0905) loss 3.7765 (3.7765) grad_norm 1.3658 (1.3658) [2021-04-16 03:32:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][10/1251] eta 0:09:05 lr 0.000598 time 0.2823 (0.4392) loss 3.0692 (3.5260) grad_norm 1.5841 (1.4434) [2021-04-16 03:32:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][20/1251] eta 0:07:24 lr 0.000598 time 0.2515 (0.3612) loss 3.3376 (3.5879) grad_norm 1.3355 (1.4498) [2021-04-16 03:32:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][30/1251] eta 0:06:52 lr 0.000598 time 0.2973 (0.3382) loss 3.7332 (3.7120) grad_norm 1.5977 (1.4533) [2021-04-16 03:32:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][40/1251] eta 0:06:32 lr 0.000598 time 0.2556 (0.3242) loss 2.5154 (3.6480) grad_norm 1.2603 (1.4687) [2021-04-16 03:32:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][50/1251] eta 0:06:17 lr 0.000598 time 0.2730 (0.3142) loss 2.9386 (3.6664) grad_norm 1.1796 (1.4474) [2021-04-16 03:33:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][60/1251] eta 0:06:09 lr 0.000598 time 0.4161 (0.3106) loss 3.0101 (3.6278) grad_norm 1.3266 (1.4388) [2021-04-16 03:33:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][70/1251] eta 0:06:01 lr 0.000597 time 0.2672 (0.3059) loss 3.9546 (3.6106) grad_norm 1.2202 (1.4258) [2021-04-16 03:33:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][80/1251] eta 0:05:53 lr 0.000597 time 0.2882 (0.3022) loss 4.1496 (3.6165) grad_norm 1.4765 (1.4215) [2021-04-16 03:33:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][90/1251] eta 0:05:49 lr 0.000597 time 0.2826 (0.3011) loss 3.1558 (3.6239) grad_norm 1.3992 (1.4273) [2021-04-16 03:33:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][100/1251] eta 0:05:43 lr 0.000597 time 0.2983 (0.2986) loss 4.3272 (3.6345) grad_norm 1.4759 (1.4293) [2021-04-16 03:33:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][110/1251] eta 0:05:38 lr 0.000597 time 0.2834 (0.2962) loss 3.8817 (3.6298) grad_norm 1.4592 (1.4323) [2021-04-16 03:33:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][120/1251] eta 0:05:33 lr 0.000597 time 0.2992 (0.2948) loss 2.9227 (3.6144) grad_norm 1.3232 (1.4331) [2021-04-16 03:33:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][130/1251] eta 0:05:29 lr 0.000597 time 0.2872 (0.2938) loss 2.5342 (3.6088) grad_norm 1.4923 (1.4325) [2021-04-16 03:33:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][140/1251] eta 0:05:26 lr 0.000597 time 0.2598 (0.2936) loss 3.7920 (3.6068) grad_norm 1.3992 (1.4289) [2021-04-16 03:33:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][150/1251] eta 0:05:22 lr 0.000597 time 0.2793 (0.2929) loss 3.7870 (3.6057) grad_norm 1.5691 (1.4265) [2021-04-16 03:33:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][160/1251] eta 0:05:19 lr 0.000597 time 0.2697 (0.2925) loss 3.2016 (3.6013) grad_norm 1.3793 (1.4295) [2021-04-16 03:33:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][170/1251] eta 0:05:15 lr 0.000597 time 0.3010 (0.2920) loss 4.3489 (3.6113) grad_norm 1.6867 (1.4300) [2021-04-16 03:33:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][180/1251] eta 0:05:12 lr 0.000597 time 0.3098 (0.2917) loss 4.3146 (3.6142) grad_norm 1.1850 (1.4295) [2021-04-16 03:33:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][190/1251] eta 0:05:08 lr 0.000597 time 0.2749 (0.2909) loss 3.9099 (3.6217) grad_norm 1.4437 (1.4309) [2021-04-16 03:33:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][200/1251] eta 0:05:04 lr 0.000597 time 0.2973 (0.2902) loss 3.6258 (3.6218) grad_norm 1.6873 (1.4311) [2021-04-16 03:33:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][210/1251] eta 0:05:01 lr 0.000597 time 0.2912 (0.2896) loss 4.1930 (3.6100) grad_norm 1.4614 (1.4327) [2021-04-16 03:33:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][220/1251] eta 0:04:58 lr 0.000597 time 0.3004 (0.2892) loss 2.8404 (3.6107) grad_norm 1.2946 (1.4283) [2021-04-16 03:33:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][230/1251] eta 0:04:55 lr 0.000597 time 0.2549 (0.2892) loss 2.3321 (3.6054) grad_norm 1.6890 (1.4279) [2021-04-16 03:33:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][240/1251] eta 0:04:52 lr 0.000597 time 0.3012 (0.2889) loss 2.8454 (3.5940) grad_norm 1.4827 (1.4267) [2021-04-16 03:33:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][250/1251] eta 0:04:48 lr 0.000597 time 0.2735 (0.2885) loss 2.8165 (3.5969) grad_norm 1.5592 (1.4281) [2021-04-16 03:33:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][260/1251] eta 0:04:45 lr 0.000597 time 0.2675 (0.2879) loss 3.3657 (3.5915) grad_norm 1.2974 (1.4288) [2021-04-16 03:33:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][270/1251] eta 0:04:42 lr 0.000597 time 0.3051 (0.2875) loss 3.4271 (3.5856) grad_norm 1.3168 (1.4267) [2021-04-16 03:34:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][280/1251] eta 0:04:38 lr 0.000597 time 0.2733 (0.2871) loss 4.1509 (3.5770) grad_norm 1.3676 (1.4255) [2021-04-16 03:34:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][290/1251] eta 0:04:35 lr 0.000597 time 0.2791 (0.2867) loss 4.3399 (3.5794) grad_norm 1.4959 (1.4239) [2021-04-16 03:34:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][300/1251] eta 0:04:32 lr 0.000597 time 0.2880 (0.2864) loss 2.5304 (3.5751) grad_norm 1.6142 (1.4264) [2021-04-16 03:34:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][310/1251] eta 0:04:29 lr 0.000596 time 0.2930 (0.2861) loss 3.7795 (3.5805) grad_norm 1.6031 (1.4294) [2021-04-16 03:34:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][320/1251] eta 0:04:25 lr 0.000596 time 0.2651 (0.2857) loss 3.7139 (3.5800) grad_norm 1.3574 (1.4274) [2021-04-16 03:34:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][330/1251] eta 0:04:22 lr 0.000596 time 0.2645 (0.2854) loss 3.5388 (3.5853) grad_norm 1.6627 (1.4290) [2021-04-16 03:34:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][340/1251] eta 0:04:19 lr 0.000596 time 0.2841 (0.2852) loss 3.9546 (3.5855) grad_norm 1.6765 (1.4305) [2021-04-16 03:34:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][350/1251] eta 0:04:16 lr 0.000596 time 0.3574 (0.2852) loss 3.9976 (3.5810) grad_norm 1.4781 (1.4304) [2021-04-16 03:34:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][360/1251] eta 0:04:13 lr 0.000596 time 0.2725 (0.2851) loss 3.4029 (3.5794) grad_norm 1.2644 (1.4299) [2021-04-16 03:34:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][370/1251] eta 0:04:11 lr 0.000596 time 0.2706 (0.2852) loss 3.3378 (3.5818) grad_norm 1.2696 (1.4276) [2021-04-16 03:34:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][380/1251] eta 0:04:08 lr 0.000596 time 0.2465 (0.2854) loss 3.1266 (3.5774) grad_norm 1.3686 (1.4272) [2021-04-16 03:34:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][390/1251] eta 0:04:05 lr 0.000596 time 0.2952 (0.2856) loss 2.5149 (3.5740) grad_norm 1.7083 (1.4295) [2021-04-16 03:34:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][400/1251] eta 0:04:02 lr 0.000596 time 0.2501 (0.2853) loss 2.4007 (3.5708) grad_norm 1.1605 (1.4310) [2021-04-16 03:34:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][410/1251] eta 0:03:59 lr 0.000596 time 0.2907 (0.2852) loss 3.1049 (3.5716) grad_norm 1.5900 (1.4304) [2021-04-16 03:34:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][420/1251] eta 0:03:56 lr 0.000596 time 0.2733 (0.2849) loss 3.8143 (3.5724) grad_norm 1.3840 (1.4306) [2021-04-16 03:34:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][430/1251] eta 0:03:53 lr 0.000596 time 0.2584 (0.2849) loss 2.7949 (3.5738) grad_norm 1.4757 (1.4296) [2021-04-16 03:34:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][440/1251] eta 0:03:50 lr 0.000596 time 0.2696 (0.2848) loss 3.9337 (3.5792) grad_norm 1.3655 (1.4284) [2021-04-16 03:34:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][450/1251] eta 0:03:47 lr 0.000596 time 0.2827 (0.2846) loss 4.1228 (3.5831) grad_norm 1.4503 (1.4279) [2021-04-16 03:34:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][460/1251] eta 0:03:45 lr 0.000596 time 0.2718 (0.2846) loss 3.7315 (3.5858) grad_norm 1.4115 (1.4285) [2021-04-16 03:34:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][470/1251] eta 0:03:42 lr 0.000596 time 0.2665 (0.2843) loss 3.7936 (3.5934) grad_norm 1.3127 (1.4270) [2021-04-16 03:34:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][480/1251] eta 0:03:39 lr 0.000596 time 0.2810 (0.2842) loss 4.1132 (3.5986) grad_norm 1.4363 (1.4260) [2021-04-16 03:35:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][490/1251] eta 0:03:36 lr 0.000596 time 0.2816 (0.2841) loss 3.5431 (3.5953) grad_norm 1.3839 (1.4250) [2021-04-16 03:35:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][500/1251] eta 0:03:33 lr 0.000596 time 0.2946 (0.2842) loss 3.5977 (3.5931) grad_norm 1.3554 (1.4260) [2021-04-16 03:35:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][510/1251] eta 0:03:30 lr 0.000596 time 0.2781 (0.2840) loss 3.8315 (3.5852) grad_norm 1.4665 (1.4252) [2021-04-16 03:35:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][520/1251] eta 0:03:27 lr 0.000596 time 0.2945 (0.2838) loss 2.5649 (3.5841) grad_norm 1.5839 (1.4250) [2021-04-16 03:35:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][530/1251] eta 0:03:24 lr 0.000596 time 0.2902 (0.2836) loss 3.6540 (3.5871) grad_norm 1.4849 (1.4257) [2021-04-16 03:35:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][540/1251] eta 0:03:21 lr 0.000596 time 0.2723 (0.2835) loss 3.7562 (3.5906) grad_norm 1.6848 (1.4280) [2021-04-16 03:35:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][550/1251] eta 0:03:18 lr 0.000596 time 0.2765 (0.2834) loss 3.1303 (3.5874) grad_norm 1.4169 (1.4295) [2021-04-16 03:35:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][560/1251] eta 0:03:15 lr 0.000595 time 0.2616 (0.2832) loss 4.0954 (3.5875) grad_norm 1.7810 (1.4304) [2021-04-16 03:35:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][570/1251] eta 0:03:12 lr 0.000595 time 0.2625 (0.2831) loss 3.7735 (3.5864) grad_norm 1.2418 (1.4294) [2021-04-16 03:35:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][580/1251] eta 0:03:10 lr 0.000595 time 0.2665 (0.2832) loss 3.1496 (3.5849) grad_norm 1.3365 (1.4284) [2021-04-16 03:35:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][590/1251] eta 0:03:07 lr 0.000595 time 0.2691 (0.2831) loss 3.8894 (3.5855) grad_norm 1.3658 (1.4288) [2021-04-16 03:35:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][600/1251] eta 0:03:04 lr 0.000595 time 0.2995 (0.2830) loss 3.6844 (3.5871) grad_norm 1.3671 (1.4286) [2021-04-16 03:35:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][610/1251] eta 0:03:01 lr 0.000595 time 0.2678 (0.2828) loss 3.1801 (3.5854) grad_norm 1.3824 (1.4281) [2021-04-16 03:35:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][620/1251] eta 0:02:58 lr 0.000595 time 0.2743 (0.2827) loss 3.5252 (3.5866) grad_norm 1.4107 (1.4298) [2021-04-16 03:35:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][630/1251] eta 0:02:55 lr 0.000595 time 0.2690 (0.2827) loss 3.9201 (3.5892) grad_norm 1.3806 (1.4301) [2021-04-16 03:35:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][640/1251] eta 0:02:52 lr 0.000595 time 0.2714 (0.2828) loss 3.9327 (3.5892) grad_norm 1.5134 (1.4302) [2021-04-16 03:35:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][650/1251] eta 0:02:49 lr 0.000595 time 0.2871 (0.2828) loss 3.5551 (3.5898) grad_norm 1.4443 (1.4304) [2021-04-16 03:35:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][660/1251] eta 0:02:47 lr 0.000595 time 0.2668 (0.2827) loss 4.1214 (3.5894) grad_norm 1.2492 (1.4302) [2021-04-16 03:35:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][670/1251] eta 0:02:44 lr 0.000595 time 0.2653 (0.2828) loss 4.1837 (3.5894) grad_norm 1.4622 (1.4301) [2021-04-16 03:35:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][680/1251] eta 0:02:41 lr 0.000595 time 0.2875 (0.2827) loss 3.5975 (3.5858) grad_norm 1.3678 (1.4298) [2021-04-16 03:35:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][690/1251] eta 0:02:38 lr 0.000595 time 0.3031 (0.2826) loss 3.3224 (3.5855) grad_norm 1.6159 (1.4313) [2021-04-16 03:35:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][700/1251] eta 0:02:35 lr 0.000595 time 0.2689 (0.2826) loss 3.7439 (3.5848) grad_norm 1.3540 (1.4313) [2021-04-16 03:36:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][710/1251] eta 0:02:32 lr 0.000595 time 0.2794 (0.2825) loss 3.5157 (3.5859) grad_norm 1.5590 (1.4312) [2021-04-16 03:36:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][720/1251] eta 0:02:29 lr 0.000595 time 0.2623 (0.2825) loss 3.7643 (3.5854) grad_norm 1.6249 (1.4299) [2021-04-16 03:36:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][730/1251] eta 0:02:27 lr 0.000595 time 0.2781 (0.2827) loss 4.0383 (3.5823) grad_norm 1.6517 (1.4300) [2021-04-16 03:36:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][740/1251] eta 0:02:24 lr 0.000595 time 0.2732 (0.2826) loss 4.2365 (3.5852) grad_norm 1.3377 (1.4295) [2021-04-16 03:36:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][750/1251] eta 0:02:21 lr 0.000595 time 0.2867 (0.2825) loss 3.8375 (3.5830) grad_norm 1.6237 (1.4295) [2021-04-16 03:36:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][760/1251] eta 0:02:18 lr 0.000595 time 0.2860 (0.2826) loss 3.8712 (3.5851) grad_norm 1.4825 (1.4303) [2021-04-16 03:36:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][770/1251] eta 0:02:15 lr 0.000595 time 0.2427 (0.2824) loss 3.3912 (3.5873) grad_norm 1.7094 (1.4318) [2021-04-16 03:36:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][780/1251] eta 0:02:13 lr 0.000595 time 0.2532 (0.2824) loss 3.5924 (3.5883) grad_norm 1.2530 (1.4316) [2021-04-16 03:36:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][790/1251] eta 0:02:10 lr 0.000595 time 0.2968 (0.2823) loss 3.8458 (3.5905) grad_norm 1.5870 (1.4312) [2021-04-16 03:36:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][800/1251] eta 0:02:07 lr 0.000594 time 0.2670 (0.2823) loss 4.0885 (3.5914) grad_norm 1.4148 (1.4310) [2021-04-16 03:36:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][810/1251] eta 0:02:04 lr 0.000594 time 0.2550 (0.2823) loss 3.7571 (3.5899) grad_norm 1.1756 (1.4300) [2021-04-16 03:36:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][820/1251] eta 0:02:01 lr 0.000594 time 0.2953 (0.2823) loss 4.1214 (3.5907) grad_norm 1.3114 (1.4285) [2021-04-16 03:36:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][830/1251] eta 0:01:58 lr 0.000594 time 0.2898 (0.2822) loss 3.5966 (3.5898) grad_norm 1.1386 (1.4275) [2021-04-16 03:36:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][840/1251] eta 0:01:55 lr 0.000594 time 0.2825 (0.2821) loss 3.0812 (3.5917) grad_norm 1.3533 (1.4264) [2021-04-16 03:36:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][850/1251] eta 0:01:53 lr 0.000594 time 0.2692 (0.2820) loss 3.8848 (3.5932) grad_norm 1.4524 (1.4268) [2021-04-16 03:36:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][860/1251] eta 0:01:50 lr 0.000594 time 0.2932 (0.2821) loss 2.7028 (3.5957) grad_norm 1.4071 (1.4277) [2021-04-16 03:36:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][870/1251] eta 0:01:47 lr 0.000594 time 0.3047 (0.2820) loss 2.6249 (3.5961) grad_norm 1.2440 (1.4280) [2021-04-16 03:36:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][880/1251] eta 0:01:44 lr 0.000594 time 0.2857 (0.2819) loss 2.7928 (3.5934) grad_norm 1.7517 (1.4289) [2021-04-16 03:36:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][890/1251] eta 0:01:41 lr 0.000594 time 0.2601 (0.2819) loss 4.0018 (3.5908) grad_norm 1.4613 (1.4304) [2021-04-16 03:36:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][900/1251] eta 0:01:38 lr 0.000594 time 0.2675 (0.2818) loss 3.8186 (3.5897) grad_norm 1.3232 (1.4303) [2021-04-16 03:36:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][910/1251] eta 0:01:36 lr 0.000594 time 0.2976 (0.2818) loss 4.2611 (3.5894) grad_norm 1.4632 (1.4295) [2021-04-16 03:37:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][920/1251] eta 0:01:33 lr 0.000594 time 0.2602 (0.2817) loss 4.2697 (3.5923) grad_norm 1.4313 (1.4285) [2021-04-16 03:37:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][930/1251] eta 0:01:30 lr 0.000594 time 0.2764 (0.2818) loss 4.1144 (3.5900) grad_norm 1.7390 (1.4286) [2021-04-16 03:37:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][940/1251] eta 0:01:27 lr 0.000594 time 0.2541 (0.2819) loss 2.6405 (3.5884) grad_norm 1.4789 (1.4290) [2021-04-16 03:37:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][950/1251] eta 0:01:24 lr 0.000594 time 0.2611 (0.2818) loss 4.0025 (3.5903) grad_norm 1.5575 (1.4288) [2021-04-16 03:37:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][960/1251] eta 0:01:22 lr 0.000594 time 0.2730 (0.2819) loss 4.5431 (3.5908) grad_norm 1.5244 (1.4294) [2021-04-16 03:37:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][970/1251] eta 0:01:19 lr 0.000594 time 0.2645 (0.2820) loss 4.2283 (3.5928) grad_norm 1.4591 (1.4288) [2021-04-16 03:37:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][980/1251] eta 0:01:16 lr 0.000594 time 0.2873 (0.2819) loss 3.5116 (3.5952) grad_norm 1.5030 (1.4292) [2021-04-16 03:37:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][990/1251] eta 0:01:13 lr 0.000594 time 0.2767 (0.2819) loss 3.9814 (3.5955) grad_norm 1.3824 (1.4286) [2021-04-16 03:37:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][1000/1251] eta 0:01:10 lr 0.000594 time 0.2879 (0.2818) loss 2.9591 (3.5964) grad_norm 1.5433 (1.4291) [2021-04-16 03:37:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][1010/1251] eta 0:01:07 lr 0.000594 time 0.2725 (0.2818) loss 3.0030 (3.5968) grad_norm 1.6857 (1.4298) [2021-04-16 03:37:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][1020/1251] eta 0:01:05 lr 0.000594 time 0.2674 (0.2818) loss 2.5059 (3.5980) grad_norm 1.1755 (1.4296) [2021-04-16 03:37:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][1030/1251] eta 0:01:02 lr 0.000594 time 0.2767 (0.2818) loss 4.2131 (3.5992) grad_norm 1.5413 (1.4297) [2021-04-16 03:37:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][1040/1251] eta 0:00:59 lr 0.000594 time 0.2620 (0.2817) loss 3.6562 (3.5986) grad_norm 1.7979 (1.4303) [2021-04-16 03:37:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][1050/1251] eta 0:00:56 lr 0.000593 time 0.2771 (0.2817) loss 4.1923 (3.5992) grad_norm 1.3722 (1.4300) [2021-04-16 03:37:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][1060/1251] eta 0:00:53 lr 0.000593 time 0.2764 (0.2816) loss 2.6798 (3.5968) grad_norm 1.4725 (1.4299) [2021-04-16 03:37:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][1070/1251] eta 0:00:50 lr 0.000593 time 0.2785 (0.2815) loss 3.0720 (3.5946) grad_norm 1.4284 (1.4297) [2021-04-16 03:37:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][1080/1251] eta 0:00:48 lr 0.000593 time 0.2709 (0.2814) loss 2.7756 (3.5943) grad_norm 1.2104 (1.4286) [2021-04-16 03:37:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][1090/1251] eta 0:00:45 lr 0.000593 time 0.2782 (0.2814) loss 3.6204 (3.5960) grad_norm 1.2514 (1.4284) [2021-04-16 03:37:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][1100/1251] eta 0:00:42 lr 0.000593 time 0.2711 (0.2814) loss 4.2895 (3.5984) grad_norm 1.3745 (1.4286) [2021-04-16 03:37:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][1110/1251] eta 0:00:39 lr 0.000593 time 0.2663 (0.2814) loss 3.5757 (3.5989) grad_norm 1.4277 (1.4286) [2021-04-16 03:37:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][1120/1251] eta 0:00:36 lr 0.000593 time 0.2924 (0.2814) loss 4.4902 (3.5995) grad_norm 1.5562 (1.4289) [2021-04-16 03:37:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][1130/1251] eta 0:00:34 lr 0.000593 time 0.2452 (0.2814) loss 3.0371 (3.5969) grad_norm 1.3651 (1.4298) [2021-04-16 03:38:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][1140/1251] eta 0:00:31 lr 0.000593 time 0.2687 (0.2813) loss 3.5354 (3.5970) grad_norm 1.5175 (1.4297) [2021-04-16 03:38:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][1150/1251] eta 0:00:28 lr 0.000593 time 0.2985 (0.2813) loss 3.4476 (3.5970) grad_norm 1.6300 (1.4297) [2021-04-16 03:38:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][1160/1251] eta 0:00:25 lr 0.000593 time 0.2607 (0.2814) loss 3.9234 (3.5966) grad_norm 1.3115 (1.4299) [2021-04-16 03:38:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][1170/1251] eta 0:00:22 lr 0.000593 time 0.2732 (0.2814) loss 3.7236 (3.5974) grad_norm 1.3241 (1.4298) [2021-04-16 03:38:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][1180/1251] eta 0:00:19 lr 0.000593 time 0.2617 (0.2814) loss 3.0987 (3.5962) grad_norm 1.3982 (1.4292) [2021-04-16 03:38:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][1190/1251] eta 0:00:17 lr 0.000593 time 0.2798 (0.2814) loss 4.2880 (3.5972) grad_norm 1.4908 (1.4294) [2021-04-16 03:38:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][1200/1251] eta 0:00:14 lr 0.000593 time 0.2647 (0.2814) loss 4.0362 (3.5966) grad_norm 1.4383 (1.4300) [2021-04-16 03:38:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][1210/1251] eta 0:00:11 lr 0.000593 time 0.2793 (0.2814) loss 4.3388 (3.5990) grad_norm 1.5470 (1.4304) [2021-04-16 03:38:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][1220/1251] eta 0:00:08 lr 0.000593 time 0.2658 (0.2814) loss 4.2913 (3.6032) grad_norm 1.5377 (1.4317) [2021-04-16 03:38:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][1230/1251] eta 0:00:05 lr 0.000593 time 0.2552 (0.2813) loss 3.9939 (3.6047) grad_norm 1.5800 (1.4323) [2021-04-16 03:38:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][1240/1251] eta 0:00:03 lr 0.000593 time 0.3298 (0.2813) loss 3.5154 (3.6052) grad_norm 1.2624 (1.4317) [2021-04-16 03:38:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [132/300][1250/1251] eta 0:00:00 lr 0.000593 time 0.2486 (0.2811) loss 3.8643 (3.6085) grad_norm 1.4671 (1.4318) [2021-04-16 03:38:35 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 132 training takes 0:05:54 [2021-04-16 03:38:35 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_132.pth saving...... [2021-04-16 03:38:45 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_132.pth saved !!! [2021-04-16 03:38:46 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.061 (1.061) Loss 1.0274 (1.0274) Acc@1 75.098 (75.098) Acc@5 93.262 (93.262) [2021-04-16 03:38:48 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.092 (0.215) Loss 1.0695 (1.0279) Acc@1 74.316 (75.808) Acc@5 93.848 (93.253) [2021-04-16 03:38:50 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.140 (0.239) Loss 1.0313 (1.0404) Acc@1 74.805 (75.623) Acc@5 93.652 (93.201) [2021-04-16 03:38:52 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.159 (0.225) Loss 1.1205 (1.0598) Acc@1 75.195 (75.224) Acc@5 90.918 (92.969) [2021-04-16 03:38:54 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.215) Loss 1.0261 (1.0607) Acc@1 76.562 (75.202) Acc@5 93.848 (92.852) [2021-04-16 03:39:00 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 75.212 Acc@5 92.900 [2021-04-16 03:39:00 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 75.2% [2021-04-16 03:39:00 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 75.21% [2021-04-16 03:39:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][0/1251] eta 1:05:08 lr 0.000593 time 3.1246 (3.1246) loss 2.5967 (2.5967) grad_norm 1.5878 (1.5878) [2021-04-16 03:39:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][10/1251] eta 0:10:59 lr 0.000593 time 0.2723 (0.5312) loss 3.6033 (3.5245) grad_norm 1.5382 (1.4343) [2021-04-16 03:39:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][20/1251] eta 0:08:39 lr 0.000593 time 0.2816 (0.4218) loss 4.4488 (3.5590) grad_norm 1.4424 (1.4459) [2021-04-16 03:39:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][30/1251] eta 0:07:39 lr 0.000593 time 0.2955 (0.3763) loss 3.7331 (3.6146) grad_norm 1.3064 (1.4347) [2021-04-16 03:39:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][40/1251] eta 0:07:06 lr 0.000592 time 0.2691 (0.3519) loss 4.4285 (3.6119) grad_norm 1.9752 (1.4471) [2021-04-16 03:39:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][50/1251] eta 0:06:45 lr 0.000592 time 0.2935 (0.3374) loss 4.2719 (3.6813) grad_norm 1.6651 (1.4482) [2021-04-16 03:39:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][60/1251] eta 0:06:31 lr 0.000592 time 0.3213 (0.3285) loss 3.7466 (3.6550) grad_norm 1.7022 (1.4538) [2021-04-16 03:39:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][70/1251] eta 0:06:18 lr 0.000592 time 0.2490 (0.3207) loss 2.8524 (3.5720) grad_norm 1.4384 (1.4727) [2021-04-16 03:39:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][80/1251] eta 0:06:09 lr 0.000592 time 0.2716 (0.3158) loss 4.0228 (3.5878) grad_norm 1.6318 (1.4685) [2021-04-16 03:39:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][90/1251] eta 0:06:01 lr 0.000592 time 0.2525 (0.3116) loss 4.2303 (3.5710) grad_norm 1.4397 (1.4607) [2021-04-16 03:39:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][100/1251] eta 0:05:54 lr 0.000592 time 0.2835 (0.3083) loss 3.9347 (3.5776) grad_norm 1.4835 (1.4615) [2021-04-16 03:39:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][110/1251] eta 0:05:49 lr 0.000592 time 0.2908 (0.3061) loss 3.6300 (3.5953) grad_norm 1.2635 (1.4645) [2021-04-16 03:39:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][120/1251] eta 0:05:43 lr 0.000592 time 0.2655 (0.3038) loss 3.8785 (3.5888) grad_norm 1.3058 (1.4589) [2021-04-16 03:39:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][130/1251] eta 0:05:39 lr 0.000592 time 0.2797 (0.3030) loss 2.7795 (3.5709) grad_norm 1.3501 (1.4507) [2021-04-16 03:39:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][140/1251] eta 0:05:36 lr 0.000592 time 0.2786 (0.3030) loss 3.8874 (3.5639) grad_norm 1.4116 (1.4490) [2021-04-16 03:39:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][150/1251] eta 0:05:32 lr 0.000592 time 0.2927 (0.3019) loss 3.8751 (3.5826) grad_norm 1.3070 (1.4475) [2021-04-16 03:39:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][160/1251] eta 0:05:27 lr 0.000592 time 0.3190 (0.3005) loss 3.4178 (3.5897) grad_norm 1.4552 (1.4467) [2021-04-16 03:39:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][170/1251] eta 0:05:24 lr 0.000592 time 0.2895 (0.3004) loss 3.5111 (3.6022) grad_norm 1.3609 (1.4428) [2021-04-16 03:39:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][180/1251] eta 0:05:21 lr 0.000592 time 0.2807 (0.3001) loss 3.4772 (3.6195) grad_norm 1.6386 (1.4418) [2021-04-16 03:39:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][190/1251] eta 0:05:16 lr 0.000592 time 0.2740 (0.2988) loss 4.2708 (3.6138) grad_norm 1.4709 (1.4433) [2021-04-16 03:40:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][200/1251] eta 0:05:13 lr 0.000592 time 0.2810 (0.2978) loss 3.2135 (3.6116) grad_norm 1.6137 (1.4456) [2021-04-16 03:40:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][210/1251] eta 0:05:09 lr 0.000592 time 0.3040 (0.2970) loss 4.0649 (3.6024) grad_norm 1.3040 (1.4456) [2021-04-16 03:40:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][220/1251] eta 0:05:05 lr 0.000592 time 0.2879 (0.2965) loss 3.5228 (3.6012) grad_norm 1.5195 (1.4444) [2021-04-16 03:40:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][230/1251] eta 0:05:02 lr 0.000592 time 0.2660 (0.2958) loss 3.6295 (3.5981) grad_norm 1.1805 (1.4429) [2021-04-16 03:40:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][240/1251] eta 0:04:58 lr 0.000592 time 0.2884 (0.2950) loss 2.4167 (3.5924) grad_norm 1.6206 (1.4428) [2021-04-16 03:40:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][250/1251] eta 0:04:54 lr 0.000592 time 0.2764 (0.2943) loss 3.1297 (3.5902) grad_norm 1.3344 (1.4410) [2021-04-16 03:40:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][260/1251] eta 0:04:50 lr 0.000592 time 0.2706 (0.2936) loss 4.0926 (3.5813) grad_norm 1.3351 (nan) [2021-04-16 03:40:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][270/1251] eta 0:04:47 lr 0.000592 time 0.2942 (0.2934) loss 3.4215 (3.5716) grad_norm 1.5551 (nan) [2021-04-16 03:40:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][280/1251] eta 0:04:44 lr 0.000592 time 0.2605 (0.2928) loss 2.5817 (3.5758) grad_norm 1.4774 (nan) [2021-04-16 03:40:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][290/1251] eta 0:04:41 lr 0.000591 time 0.2614 (0.2924) loss 3.8315 (3.5724) grad_norm 1.4512 (nan) [2021-04-16 03:40:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][300/1251] eta 0:04:37 lr 0.000591 time 0.2720 (0.2920) loss 3.5914 (3.5647) grad_norm 1.2819 (nan) [2021-04-16 03:40:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][310/1251] eta 0:04:34 lr 0.000591 time 0.2812 (0.2916) loss 3.7702 (3.5593) grad_norm 1.3942 (nan) [2021-04-16 03:40:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][320/1251] eta 0:04:30 lr 0.000591 time 0.2615 (0.2911) loss 3.5421 (3.5502) grad_norm 1.4708 (nan) [2021-04-16 03:40:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][330/1251] eta 0:04:27 lr 0.000591 time 0.2925 (0.2906) loss 4.3000 (3.5613) grad_norm 1.3079 (nan) [2021-04-16 03:40:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][340/1251] eta 0:04:24 lr 0.000591 time 0.2649 (0.2904) loss 2.6993 (3.5618) grad_norm 1.7819 (nan) [2021-04-16 03:40:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][350/1251] eta 0:04:21 lr 0.000591 time 0.2558 (0.2900) loss 4.0671 (3.5668) grad_norm 2.1238 (nan) [2021-04-16 03:40:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][360/1251] eta 0:04:18 lr 0.000591 time 0.2741 (0.2897) loss 3.2003 (3.5701) grad_norm 1.5951 (nan) [2021-04-16 03:40:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][370/1251] eta 0:04:15 lr 0.000591 time 0.2971 (0.2900) loss 4.0924 (3.5743) grad_norm 1.4217 (nan) [2021-04-16 03:40:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][380/1251] eta 0:04:12 lr 0.000591 time 0.2693 (0.2898) loss 3.8307 (3.5719) grad_norm 1.5407 (nan) [2021-04-16 03:40:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][390/1251] eta 0:04:09 lr 0.000591 time 0.2833 (0.2896) loss 3.6434 (3.5722) grad_norm 1.4823 (nan) [2021-04-16 03:40:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][400/1251] eta 0:04:06 lr 0.000591 time 0.2651 (0.2893) loss 2.8665 (3.5771) grad_norm 1.3101 (nan) [2021-04-16 03:40:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][410/1251] eta 0:04:03 lr 0.000591 time 0.2803 (0.2890) loss 4.1012 (3.5706) grad_norm 1.4676 (nan) [2021-04-16 03:41:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][420/1251] eta 0:04:00 lr 0.000591 time 0.2925 (0.2890) loss 3.6136 (3.5737) grad_norm 1.3602 (nan) [2021-04-16 03:41:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][430/1251] eta 0:03:57 lr 0.000591 time 0.2872 (0.2889) loss 3.7859 (3.5797) grad_norm 1.3853 (nan) [2021-04-16 03:41:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][440/1251] eta 0:03:54 lr 0.000591 time 0.2696 (0.2887) loss 3.7596 (3.5830) grad_norm 1.2238 (nan) [2021-04-16 03:41:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][450/1251] eta 0:03:51 lr 0.000591 time 0.2950 (0.2885) loss 2.7965 (3.5779) grad_norm 1.3477 (nan) [2021-04-16 03:41:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][460/1251] eta 0:03:48 lr 0.000591 time 0.2876 (0.2883) loss 3.8889 (3.5699) grad_norm 1.2710 (nan) [2021-04-16 03:41:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][470/1251] eta 0:03:44 lr 0.000591 time 0.2912 (0.2880) loss 3.1181 (3.5689) grad_norm 1.2832 (nan) [2021-04-16 03:41:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][480/1251] eta 0:03:41 lr 0.000591 time 0.2806 (0.2878) loss 3.7081 (3.5725) grad_norm 1.4155 (nan) [2021-04-16 03:41:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][490/1251] eta 0:03:38 lr 0.000591 time 0.3004 (0.2877) loss 3.9519 (3.5755) grad_norm 1.4781 (nan) [2021-04-16 03:41:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][500/1251] eta 0:03:35 lr 0.000591 time 0.3095 (0.2874) loss 3.1873 (3.5762) grad_norm 1.1150 (nan) [2021-04-16 03:41:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][510/1251] eta 0:03:32 lr 0.000591 time 0.2928 (0.2873) loss 3.4314 (3.5733) grad_norm 1.4452 (nan) [2021-04-16 03:41:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][520/1251] eta 0:03:29 lr 0.000591 time 0.2847 (0.2871) loss 3.9353 (3.5773) grad_norm 1.5041 (nan) [2021-04-16 03:41:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][530/1251] eta 0:03:26 lr 0.000590 time 0.2805 (0.2870) loss 3.8858 (3.5766) grad_norm 1.2896 (nan) [2021-04-16 03:41:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][540/1251] eta 0:03:23 lr 0.000590 time 0.2823 (0.2869) loss 4.1151 (3.5764) grad_norm 1.4799 (nan) [2021-04-16 03:41:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][550/1251] eta 0:03:21 lr 0.000590 time 0.2651 (0.2873) loss 4.1289 (3.5701) grad_norm 1.3822 (nan) [2021-04-16 03:41:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][560/1251] eta 0:03:18 lr 0.000590 time 0.2913 (0.2871) loss 4.5417 (3.5734) grad_norm 1.2557 (nan) [2021-04-16 03:41:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][570/1251] eta 0:03:15 lr 0.000590 time 0.2438 (0.2873) loss 3.4334 (3.5777) grad_norm 1.5933 (nan) [2021-04-16 03:41:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][580/1251] eta 0:03:12 lr 0.000590 time 0.2920 (0.2874) loss 3.4988 (3.5796) grad_norm 1.3942 (nan) [2021-04-16 03:41:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][590/1251] eta 0:03:09 lr 0.000590 time 0.2707 (0.2873) loss 4.2276 (3.5814) grad_norm 1.4396 (nan) [2021-04-16 03:41:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][600/1251] eta 0:03:06 lr 0.000590 time 0.2973 (0.2872) loss 3.9683 (3.5829) grad_norm 1.3050 (nan) [2021-04-16 03:41:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][610/1251] eta 0:03:04 lr 0.000590 time 0.2722 (0.2871) loss 2.5689 (3.5763) grad_norm 1.2582 (nan) [2021-04-16 03:41:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][620/1251] eta 0:03:01 lr 0.000590 time 0.2826 (0.2869) loss 3.4437 (3.5811) grad_norm 1.3841 (nan) [2021-04-16 03:42:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][630/1251] eta 0:02:58 lr 0.000590 time 0.2757 (0.2869) loss 4.4721 (3.5817) grad_norm 1.3355 (nan) [2021-04-16 03:42:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][640/1251] eta 0:02:55 lr 0.000590 time 0.2818 (0.2866) loss 3.8669 (3.5888) grad_norm 1.2488 (nan) [2021-04-16 03:42:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][650/1251] eta 0:02:52 lr 0.000590 time 0.2763 (0.2864) loss 4.1262 (3.5880) grad_norm 1.4324 (nan) [2021-04-16 03:42:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][660/1251] eta 0:02:49 lr 0.000590 time 0.2779 (0.2863) loss 3.5063 (3.5893) grad_norm 1.4127 (nan) [2021-04-16 03:42:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][670/1251] eta 0:02:46 lr 0.000590 time 0.2694 (0.2863) loss 3.8797 (3.5878) grad_norm 1.2623 (nan) [2021-04-16 03:42:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][680/1251] eta 0:02:43 lr 0.000590 time 0.2638 (0.2861) loss 3.8069 (3.5850) grad_norm 1.3308 (nan) [2021-04-16 03:42:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][690/1251] eta 0:02:40 lr 0.000590 time 0.2715 (0.2860) loss 4.0672 (3.5843) grad_norm 1.9270 (nan) [2021-04-16 03:42:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][700/1251] eta 0:02:37 lr 0.000590 time 0.2537 (0.2860) loss 3.2331 (3.5854) grad_norm 1.3298 (nan) [2021-04-16 03:42:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][710/1251] eta 0:02:34 lr 0.000590 time 0.3045 (0.2860) loss 4.0869 (3.5852) grad_norm 1.2834 (nan) [2021-04-16 03:42:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][720/1251] eta 0:02:31 lr 0.000590 time 0.2551 (0.2862) loss 3.2613 (3.5839) grad_norm 1.3632 (nan) [2021-04-16 03:42:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][730/1251] eta 0:02:29 lr 0.000590 time 0.2935 (0.2861) loss 3.0592 (3.5855) grad_norm 1.2812 (nan) [2021-04-16 03:42:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][740/1251] eta 0:02:26 lr 0.000590 time 0.2794 (0.2863) loss 4.1320 (3.5857) grad_norm 1.4010 (nan) [2021-04-16 03:42:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][750/1251] eta 0:02:23 lr 0.000590 time 0.2637 (0.2862) loss 3.8535 (3.5822) grad_norm 1.4470 (nan) [2021-04-16 03:42:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][760/1251] eta 0:02:20 lr 0.000590 time 0.2768 (0.2862) loss 4.3205 (3.5855) grad_norm 1.4556 (nan) [2021-04-16 03:42:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][770/1251] eta 0:02:17 lr 0.000590 time 0.3088 (0.2862) loss 3.3582 (3.5864) grad_norm 1.2853 (nan) [2021-04-16 03:42:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][780/1251] eta 0:02:14 lr 0.000589 time 0.2752 (0.2861) loss 3.6852 (3.5899) grad_norm 1.3168 (nan) [2021-04-16 03:42:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][790/1251] eta 0:02:11 lr 0.000589 time 0.2690 (0.2862) loss 3.4115 (3.5926) grad_norm 1.3243 (nan) [2021-04-16 03:42:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][800/1251] eta 0:02:09 lr 0.000589 time 0.3065 (0.2862) loss 4.3461 (3.5921) grad_norm 1.6719 (nan) [2021-04-16 03:42:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][810/1251] eta 0:02:06 lr 0.000589 time 0.2600 (0.2860) loss 3.5860 (3.5915) grad_norm 1.1931 (nan) [2021-04-16 03:42:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][820/1251] eta 0:02:03 lr 0.000589 time 0.3083 (0.2860) loss 3.5438 (3.5893) grad_norm 1.5234 (nan) [2021-04-16 03:42:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][830/1251] eta 0:02:00 lr 0.000589 time 0.2457 (0.2861) loss 3.6263 (3.5882) grad_norm 1.6048 (nan) [2021-04-16 03:43:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][840/1251] eta 0:01:57 lr 0.000589 time 0.2589 (0.2864) loss 3.1800 (3.5906) grad_norm 1.3960 (nan) [2021-04-16 03:43:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][850/1251] eta 0:01:54 lr 0.000589 time 0.2904 (0.2863) loss 2.9720 (3.5915) grad_norm 1.4480 (nan) [2021-04-16 03:43:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][860/1251] eta 0:01:51 lr 0.000589 time 0.2733 (0.2862) loss 3.1360 (3.5916) grad_norm 1.2738 (nan) [2021-04-16 03:43:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][870/1251] eta 0:01:49 lr 0.000589 time 0.2737 (0.2862) loss 3.9241 (3.5925) grad_norm 1.4788 (nan) [2021-04-16 03:43:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][880/1251] eta 0:01:46 lr 0.000589 time 0.3054 (0.2861) loss 2.5189 (3.5886) grad_norm 1.4460 (nan) [2021-04-16 03:43:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][890/1251] eta 0:01:43 lr 0.000589 time 0.2861 (0.2861) loss 3.8810 (3.5851) grad_norm 1.3044 (nan) [2021-04-16 03:43:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][900/1251] eta 0:01:40 lr 0.000589 time 0.2796 (0.2860) loss 4.1700 (3.5864) grad_norm 1.3494 (nan) [2021-04-16 03:43:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][910/1251] eta 0:01:37 lr 0.000589 time 0.2908 (0.2859) loss 4.4675 (3.5914) grad_norm 1.5922 (nan) [2021-04-16 03:43:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][920/1251] eta 0:01:34 lr 0.000589 time 0.2522 (0.2859) loss 2.8731 (3.5878) grad_norm 1.4566 (nan) [2021-04-16 03:43:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][930/1251] eta 0:01:31 lr 0.000589 time 0.2696 (0.2859) loss 3.7294 (3.5863) grad_norm 1.3022 (nan) [2021-04-16 03:43:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][940/1251] eta 0:01:28 lr 0.000589 time 0.2417 (0.2860) loss 3.7914 (3.5861) grad_norm 1.3602 (nan) [2021-04-16 03:43:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][950/1251] eta 0:01:26 lr 0.000589 time 0.2824 (0.2860) loss 3.6690 (3.5859) grad_norm 1.2749 (nan) [2021-04-16 03:43:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][960/1251] eta 0:01:23 lr 0.000589 time 0.2899 (0.2860) loss 3.9912 (3.5848) grad_norm 1.4269 (nan) [2021-04-16 03:43:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][970/1251] eta 0:01:20 lr 0.000589 time 0.2912 (0.2860) loss 3.8224 (3.5866) grad_norm 1.2410 (nan) [2021-04-16 03:43:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][980/1251] eta 0:01:17 lr 0.000589 time 0.2829 (0.2859) loss 4.2375 (3.5860) grad_norm 1.4763 (nan) [2021-04-16 03:43:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][990/1251] eta 0:01:14 lr 0.000589 time 0.2633 (0.2860) loss 4.0767 (3.5866) grad_norm 1.3290 (nan) [2021-04-16 03:43:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][1000/1251] eta 0:01:11 lr 0.000589 time 0.2756 (0.2859) loss 4.0279 (3.5844) grad_norm 1.4720 (nan) [2021-04-16 03:43:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][1010/1251] eta 0:01:08 lr 0.000589 time 0.2719 (0.2858) loss 3.2262 (3.5839) grad_norm 1.4129 (nan) [2021-04-16 03:43:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][1020/1251] eta 0:01:06 lr 0.000588 time 0.2930 (0.2858) loss 2.2554 (3.5851) grad_norm 1.3547 (nan) [2021-04-16 03:43:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][1030/1251] eta 0:01:03 lr 0.000588 time 0.2822 (0.2858) loss 4.5041 (3.5861) grad_norm 1.3329 (nan) [2021-04-16 03:43:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][1040/1251] eta 0:01:00 lr 0.000588 time 0.3128 (0.2858) loss 3.6353 (3.5859) grad_norm 1.4312 (nan) [2021-04-16 03:44:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][1050/1251] eta 0:00:57 lr 0.000588 time 0.2806 (0.2857) loss 4.4038 (3.5869) grad_norm 1.3805 (nan) [2021-04-16 03:44:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][1060/1251] eta 0:00:54 lr 0.000588 time 0.2839 (0.2857) loss 4.0654 (3.5869) grad_norm 1.3080 (nan) [2021-04-16 03:44:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][1070/1251] eta 0:00:51 lr 0.000588 time 0.2901 (0.2856) loss 3.8409 (3.5873) grad_norm 1.5699 (nan) [2021-04-16 03:44:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][1080/1251] eta 0:00:48 lr 0.000588 time 0.2689 (0.2856) loss 3.5104 (3.5889) grad_norm 1.4412 (nan) [2021-04-16 03:44:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][1090/1251] eta 0:00:45 lr 0.000588 time 0.2803 (0.2856) loss 3.2437 (3.5858) grad_norm 1.3450 (nan) [2021-04-16 03:44:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][1100/1251] eta 0:00:43 lr 0.000588 time 0.2916 (0.2855) loss 3.3039 (3.5812) grad_norm 1.5360 (nan) [2021-04-16 03:44:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][1110/1251] eta 0:00:40 lr 0.000588 time 0.2859 (0.2855) loss 3.4895 (3.5830) grad_norm 1.4553 (nan) [2021-04-16 03:44:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][1120/1251] eta 0:00:37 lr 0.000588 time 0.2977 (0.2854) loss 3.8966 (3.5839) grad_norm 1.4895 (nan) [2021-04-16 03:44:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][1130/1251] eta 0:00:34 lr 0.000588 time 0.2618 (0.2855) loss 3.9596 (3.5862) grad_norm 1.4757 (nan) [2021-04-16 03:44:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][1140/1251] eta 0:00:31 lr 0.000588 time 0.2816 (0.2854) loss 3.9771 (3.5860) grad_norm 1.2323 (nan) [2021-04-16 03:44:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][1150/1251] eta 0:00:28 lr 0.000588 time 0.2766 (0.2855) loss 4.3616 (3.5890) grad_norm 1.4784 (nan) [2021-04-16 03:44:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][1160/1251] eta 0:00:25 lr 0.000588 time 0.2777 (0.2856) loss 3.7152 (3.5894) grad_norm 1.6237 (nan) [2021-04-16 03:44:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][1170/1251] eta 0:00:23 lr 0.000588 time 0.2707 (0.2856) loss 3.7561 (3.5878) grad_norm 1.5326 (nan) [2021-04-16 03:44:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][1180/1251] eta 0:00:20 lr 0.000588 time 0.2751 (0.2856) loss 3.7040 (3.5872) grad_norm 1.5134 (nan) [2021-04-16 03:44:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][1190/1251] eta 0:00:17 lr 0.000588 time 0.2799 (0.2855) loss 2.4186 (3.5858) grad_norm 1.4237 (nan) [2021-04-16 03:44:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][1200/1251] eta 0:00:14 lr 0.000588 time 0.2680 (0.2854) loss 3.7253 (3.5864) grad_norm 1.7433 (nan) [2021-04-16 03:44:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][1210/1251] eta 0:00:11 lr 0.000588 time 0.2710 (0.2854) loss 3.7594 (3.5886) grad_norm 1.3915 (nan) [2021-04-16 03:44:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][1220/1251] eta 0:00:08 lr 0.000588 time 0.2753 (0.2853) loss 3.3949 (3.5876) grad_norm 1.4103 (nan) [2021-04-16 03:44:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][1230/1251] eta 0:00:05 lr 0.000588 time 0.2831 (0.2853) loss 4.4504 (3.5869) grad_norm 1.3757 (nan) [2021-04-16 03:44:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][1240/1251] eta 0:00:03 lr 0.000588 time 0.2675 (0.2852) loss 4.2486 (3.5877) grad_norm 1.3690 (nan) [2021-04-16 03:44:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [133/300][1250/1251] eta 0:00:00 lr 0.000588 time 0.2563 (0.2849) loss 3.5531 (3.5854) grad_norm 1.4911 (nan) [2021-04-16 03:45:00 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 133 training takes 0:05:59 [2021-04-16 03:45:00 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_133.pth saving...... [2021-04-16 03:45:12 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_133.pth saved !!! [2021-04-16 03:45:13 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.062 (1.062) Loss 1.1166 (1.1166) Acc@1 74.414 (74.414) Acc@5 91.699 (91.699) [2021-04-16 03:45:15 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.791 (0.256) Loss 1.0874 (1.0794) Acc@1 75.098 (75.098) Acc@5 93.359 (92.844) [2021-04-16 03:45:17 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.110 (0.231) Loss 1.0245 (1.0670) Acc@1 76.172 (75.414) Acc@5 93.262 (92.936) [2021-04-16 03:45:19 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.112 (0.220) Loss 0.9957 (1.0701) Acc@1 77.734 (75.148) Acc@5 93.066 (92.824) [2021-04-16 03:45:21 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.209) Loss 1.0775 (1.0690) Acc@1 75.488 (75.195) Acc@5 93.066 (92.838) [2021-04-16 03:45:27 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 75.114 Acc@5 92.846 [2021-04-16 03:45:27 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 75.1% [2021-04-16 03:45:27 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 75.21% [2021-04-16 03:45:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][0/1251] eta 2:23:22 lr 0.000588 time 6.8767 (6.8767) loss 3.7445 (3.7445) grad_norm 1.7957 (1.7957) [2021-04-16 03:45:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][10/1251] eta 0:18:21 lr 0.000588 time 0.3903 (0.8875) loss 2.9991 (3.3850) grad_norm 1.4564 (1.4712) [2021-04-16 03:45:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][20/1251] eta 0:12:15 lr 0.000587 time 0.2759 (0.5976) loss 4.1212 (3.5053) grad_norm 1.2085 (1.4256) [2021-04-16 03:45:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][30/1251] eta 0:10:03 lr 0.000587 time 0.2726 (0.4942) loss 4.3589 (3.5282) grad_norm 1.2583 (1.4209) [2021-04-16 03:45:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][40/1251] eta 0:08:53 lr 0.000587 time 0.2751 (0.4407) loss 3.6863 (3.5712) grad_norm 1.3164 (1.3973) [2021-04-16 03:45:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][50/1251] eta 0:08:09 lr 0.000587 time 0.2734 (0.4076) loss 4.1093 (3.5726) grad_norm 1.3740 (1.3872) [2021-04-16 03:45:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][60/1251] eta 0:07:39 lr 0.000587 time 0.2639 (0.3861) loss 3.0836 (3.5175) grad_norm 1.3501 (1.3745) [2021-04-16 03:45:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][70/1251] eta 0:07:18 lr 0.000587 time 0.2629 (0.3709) loss 3.7964 (3.5000) grad_norm 1.2917 (1.3661) [2021-04-16 03:45:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][80/1251] eta 0:07:00 lr 0.000587 time 0.2842 (0.3592) loss 3.9290 (3.5053) grad_norm 1.3843 (1.3648) [2021-04-16 03:45:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][90/1251] eta 0:06:46 lr 0.000587 time 0.2973 (0.3502) loss 4.4102 (3.5285) grad_norm 1.2877 (1.3866) [2021-04-16 03:46:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][100/1251] eta 0:06:34 lr 0.000587 time 0.2995 (0.3430) loss 3.8270 (3.5248) grad_norm 1.4479 (1.3975) [2021-04-16 03:46:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][110/1251] eta 0:06:24 lr 0.000587 time 0.2726 (0.3369) loss 3.0878 (3.5409) grad_norm 1.3864 (1.4016) [2021-04-16 03:46:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][120/1251] eta 0:06:15 lr 0.000587 time 0.2779 (0.3319) loss 3.7060 (3.5441) grad_norm 1.3172 (1.3973) [2021-04-16 03:46:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][130/1251] eta 0:06:07 lr 0.000587 time 0.2588 (0.3278) loss 3.1554 (3.5299) grad_norm 1.4271 (1.3965) [2021-04-16 03:46:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][140/1251] eta 0:06:01 lr 0.000587 time 0.2636 (0.3250) loss 4.4822 (3.5360) grad_norm 1.5570 (1.3963) [2021-04-16 03:46:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][150/1251] eta 0:05:54 lr 0.000587 time 0.2929 (0.3219) loss 3.7884 (3.5389) grad_norm 1.7093 (1.3965) [2021-04-16 03:46:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][160/1251] eta 0:05:48 lr 0.000587 time 0.2639 (0.3191) loss 3.2153 (3.5291) grad_norm 1.1939 (1.3960) [2021-04-16 03:46:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][170/1251] eta 0:05:42 lr 0.000587 time 0.2719 (0.3168) loss 3.1514 (3.5382) grad_norm 1.4692 (1.3969) [2021-04-16 03:46:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][180/1251] eta 0:05:36 lr 0.000587 time 0.2743 (0.3145) loss 3.8549 (3.5343) grad_norm 1.4243 (1.3973) [2021-04-16 03:46:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][190/1251] eta 0:05:31 lr 0.000587 time 0.2862 (0.3123) loss 3.6113 (3.5328) grad_norm 1.3651 (1.3989) [2021-04-16 03:46:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][200/1251] eta 0:05:27 lr 0.000587 time 0.2678 (0.3116) loss 4.2162 (3.5412) grad_norm 1.5025 (1.3997) [2021-04-16 03:46:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][210/1251] eta 0:05:22 lr 0.000587 time 0.2573 (0.3098) loss 3.8941 (3.5447) grad_norm 1.5003 (1.4029) [2021-04-16 03:46:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][220/1251] eta 0:05:17 lr 0.000587 time 0.2779 (0.3082) loss 4.4152 (3.5559) grad_norm 1.8324 (1.4075) [2021-04-16 03:46:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][230/1251] eta 0:05:13 lr 0.000587 time 0.2691 (0.3068) loss 3.7152 (3.5616) grad_norm 1.3035 (1.4087) [2021-04-16 03:46:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][240/1251] eta 0:05:09 lr 0.000587 time 0.2765 (0.3060) loss 4.0016 (3.5614) grad_norm 1.3146 (1.4075) [2021-04-16 03:46:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][250/1251] eta 0:05:05 lr 0.000587 time 0.2721 (0.3048) loss 3.8326 (3.5625) grad_norm 1.5853 (1.4071) [2021-04-16 03:46:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][260/1251] eta 0:05:00 lr 0.000586 time 0.2967 (0.3037) loss 2.8324 (3.5595) grad_norm 1.6019 (1.4094) [2021-04-16 03:46:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][270/1251] eta 0:04:56 lr 0.000586 time 0.2659 (0.3026) loss 3.4147 (3.5613) grad_norm 1.5282 (1.4121) [2021-04-16 03:46:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][280/1251] eta 0:04:53 lr 0.000586 time 0.2710 (0.3018) loss 4.3026 (3.5475) grad_norm 1.7841 (1.4168) [2021-04-16 03:46:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][290/1251] eta 0:04:49 lr 0.000586 time 0.2695 (0.3009) loss 3.7282 (3.5481) grad_norm 1.7912 (1.4227) [2021-04-16 03:46:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][300/1251] eta 0:04:45 lr 0.000586 time 0.2635 (0.3000) loss 3.3591 (3.5433) grad_norm 1.3601 (1.4246) [2021-04-16 03:47:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][310/1251] eta 0:04:42 lr 0.000586 time 0.2922 (0.2997) loss 3.2336 (3.5521) grad_norm 1.4623 (1.4258) [2021-04-16 03:47:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][320/1251] eta 0:04:38 lr 0.000586 time 0.2847 (0.2989) loss 3.2371 (3.5526) grad_norm 1.3749 (1.4281) [2021-04-16 03:47:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][330/1251] eta 0:04:35 lr 0.000586 time 0.2571 (0.2987) loss 3.2324 (3.5498) grad_norm 1.2748 (1.4286) [2021-04-16 03:47:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][340/1251] eta 0:04:31 lr 0.000586 time 0.2843 (0.2980) loss 3.4377 (3.5531) grad_norm 1.2971 (1.4285) [2021-04-16 03:47:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][350/1251] eta 0:04:28 lr 0.000586 time 0.4540 (0.2980) loss 3.4007 (3.5489) grad_norm 1.3149 (1.4294) [2021-04-16 03:47:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][360/1251] eta 0:04:25 lr 0.000586 time 0.2820 (0.2977) loss 2.7595 (3.5489) grad_norm 1.3521 (1.4275) [2021-04-16 03:47:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][370/1251] eta 0:04:22 lr 0.000586 time 0.2897 (0.2975) loss 3.8007 (3.5557) grad_norm 1.2082 (1.4260) [2021-04-16 03:47:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][380/1251] eta 0:04:18 lr 0.000586 time 0.2765 (0.2969) loss 3.9707 (3.5645) grad_norm 1.6145 (1.4246) [2021-04-16 03:47:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][390/1251] eta 0:04:15 lr 0.000586 time 0.3099 (0.2964) loss 3.6673 (3.5670) grad_norm 1.2055 (1.4242) [2021-04-16 03:47:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][400/1251] eta 0:04:11 lr 0.000586 time 0.2846 (0.2958) loss 3.1808 (3.5619) grad_norm 1.2614 (1.4250) [2021-04-16 03:47:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][410/1251] eta 0:04:08 lr 0.000586 time 0.2600 (0.2952) loss 2.9346 (3.5603) grad_norm 1.2199 (1.4236) [2021-04-16 03:47:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][420/1251] eta 0:04:04 lr 0.000586 time 0.2659 (0.2948) loss 3.9352 (3.5604) grad_norm 1.8014 (1.4264) [2021-04-16 03:47:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][430/1251] eta 0:04:01 lr 0.000586 time 0.2829 (0.2943) loss 3.2483 (3.5519) grad_norm 1.8367 (1.4275) [2021-04-16 03:47:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][440/1251] eta 0:03:58 lr 0.000586 time 0.2632 (0.2939) loss 2.3248 (3.5513) grad_norm 1.3352 (1.4280) [2021-04-16 03:47:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][450/1251] eta 0:03:55 lr 0.000586 time 0.2823 (0.2936) loss 4.2448 (3.5524) grad_norm 1.4538 (1.4267) [2021-04-16 03:47:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][460/1251] eta 0:03:51 lr 0.000586 time 0.2856 (0.2932) loss 3.8986 (3.5544) grad_norm 1.2900 (1.4254) [2021-04-16 03:47:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][470/1251] eta 0:03:48 lr 0.000586 time 0.2531 (0.2928) loss 3.3276 (3.5615) grad_norm 1.3941 (1.4233) [2021-04-16 03:47:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][480/1251] eta 0:03:45 lr 0.000586 time 0.2714 (0.2924) loss 3.0829 (3.5644) grad_norm 1.2301 (1.4225) [2021-04-16 03:47:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][490/1251] eta 0:03:42 lr 0.000586 time 0.2638 (0.2920) loss 4.1810 (3.5659) grad_norm 1.3870 (1.4212) [2021-04-16 03:47:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][500/1251] eta 0:03:39 lr 0.000586 time 0.2787 (0.2917) loss 3.5538 (3.5675) grad_norm 1.5100 (1.4237) [2021-04-16 03:47:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][510/1251] eta 0:03:35 lr 0.000585 time 0.2587 (0.2915) loss 4.1172 (3.5638) grad_norm 1.5057 (1.4232) [2021-04-16 03:47:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][520/1251] eta 0:03:32 lr 0.000585 time 0.2753 (0.2911) loss 3.1930 (3.5664) grad_norm 1.4381 (1.4247) [2021-04-16 03:48:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][530/1251] eta 0:03:29 lr 0.000585 time 0.2642 (0.2908) loss 3.7487 (3.5652) grad_norm 1.3585 (1.4247) [2021-04-16 03:48:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][540/1251] eta 0:03:26 lr 0.000585 time 0.3027 (0.2909) loss 3.2266 (3.5651) grad_norm 1.5513 (1.4251) [2021-04-16 03:48:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][550/1251] eta 0:03:23 lr 0.000585 time 0.2497 (0.2907) loss 3.7167 (3.5714) grad_norm 1.4445 (1.4264) [2021-04-16 03:48:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][560/1251] eta 0:03:20 lr 0.000585 time 0.2708 (0.2904) loss 3.7211 (3.5695) grad_norm 1.5423 (1.4262) [2021-04-16 03:48:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][570/1251] eta 0:03:17 lr 0.000585 time 0.3011 (0.2906) loss 3.8554 (3.5741) grad_norm 1.5258 (1.4274) [2021-04-16 03:48:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][580/1251] eta 0:03:14 lr 0.000585 time 0.2513 (0.2904) loss 4.3490 (3.5735) grad_norm 1.2942 (1.4266) [2021-04-16 03:48:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][590/1251] eta 0:03:11 lr 0.000585 time 0.2794 (0.2903) loss 4.3867 (3.5748) grad_norm 1.5025 (1.4273) [2021-04-16 03:48:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][600/1251] eta 0:03:08 lr 0.000585 time 0.2931 (0.2903) loss 3.3657 (3.5773) grad_norm 1.5123 (1.4279) [2021-04-16 03:48:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][610/1251] eta 0:03:05 lr 0.000585 time 0.2764 (0.2900) loss 3.6199 (3.5787) grad_norm 1.3149 (1.4284) [2021-04-16 03:48:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][620/1251] eta 0:03:02 lr 0.000585 time 0.2757 (0.2898) loss 3.4366 (3.5791) grad_norm 2.2239 (1.4318) [2021-04-16 03:48:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][630/1251] eta 0:02:59 lr 0.000585 time 0.2814 (0.2896) loss 3.7041 (3.5825) grad_norm 1.6389 (1.4329) [2021-04-16 03:48:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][640/1251] eta 0:02:56 lr 0.000585 time 0.2650 (0.2894) loss 3.5057 (3.5796) grad_norm 1.3990 (1.4328) [2021-04-16 03:48:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][650/1251] eta 0:02:53 lr 0.000585 time 0.2774 (0.2892) loss 4.1073 (3.5816) grad_norm 1.5667 (1.4314) [2021-04-16 03:48:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][660/1251] eta 0:02:50 lr 0.000585 time 0.2960 (0.2890) loss 3.8679 (3.5845) grad_norm 1.3913 (1.4306) [2021-04-16 03:48:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][670/1251] eta 0:02:47 lr 0.000585 time 0.2910 (0.2888) loss 3.3578 (3.5868) grad_norm 1.5144 (1.4323) [2021-04-16 03:48:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][680/1251] eta 0:02:44 lr 0.000585 time 0.3076 (0.2887) loss 4.3943 (3.5856) grad_norm 1.3517 (1.4321) [2021-04-16 03:48:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][690/1251] eta 0:02:41 lr 0.000585 time 0.2772 (0.2885) loss 4.3588 (3.5861) grad_norm 1.2253 (1.4302) [2021-04-16 03:48:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][700/1251] eta 0:02:38 lr 0.000585 time 0.2715 (0.2883) loss 2.2641 (3.5861) grad_norm 1.3864 (1.4297) [2021-04-16 03:48:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][710/1251] eta 0:02:35 lr 0.000585 time 0.2684 (0.2881) loss 3.4871 (3.5844) grad_norm 1.7007 (1.4301) [2021-04-16 03:48:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][720/1251] eta 0:02:33 lr 0.000585 time 0.2882 (0.2881) loss 3.7575 (3.5809) grad_norm 1.4180 (1.4305) [2021-04-16 03:48:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][730/1251] eta 0:02:30 lr 0.000585 time 0.2846 (0.2880) loss 3.4194 (3.5796) grad_norm 1.6221 (1.4316) [2021-04-16 03:49:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][740/1251] eta 0:02:27 lr 0.000585 time 0.2588 (0.2878) loss 3.1000 (3.5779) grad_norm 1.3446 (1.4338) [2021-04-16 03:49:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][750/1251] eta 0:02:24 lr 0.000584 time 0.3166 (0.2877) loss 2.7941 (3.5747) grad_norm 1.3238 (1.4351) [2021-04-16 03:49:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][760/1251] eta 0:02:21 lr 0.000584 time 0.2873 (0.2875) loss 3.7647 (3.5791) grad_norm 1.4099 (1.4350) [2021-04-16 03:49:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][770/1251] eta 0:02:18 lr 0.000584 time 0.2833 (0.2874) loss 3.1827 (3.5771) grad_norm 1.3990 (1.4342) [2021-04-16 03:49:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][780/1251] eta 0:02:15 lr 0.000584 time 0.2678 (0.2874) loss 4.1159 (3.5769) grad_norm 1.4217 (1.4332) [2021-04-16 03:49:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][790/1251] eta 0:02:12 lr 0.000584 time 0.2995 (0.2872) loss 3.4933 (3.5788) grad_norm 1.5876 (1.4329) [2021-04-16 03:49:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][800/1251] eta 0:02:09 lr 0.000584 time 0.2849 (0.2872) loss 3.8742 (3.5803) grad_norm 1.4539 (1.4325) [2021-04-16 03:49:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][810/1251] eta 0:02:06 lr 0.000584 time 0.2507 (0.2871) loss 3.6986 (3.5820) grad_norm 1.3799 (1.4330) [2021-04-16 03:49:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][820/1251] eta 0:02:03 lr 0.000584 time 0.2715 (0.2871) loss 3.7313 (3.5805) grad_norm 1.2574 (1.4327) [2021-04-16 03:49:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][830/1251] eta 0:02:00 lr 0.000584 time 0.2634 (0.2871) loss 4.0110 (3.5806) grad_norm 1.4414 (1.4329) [2021-04-16 03:49:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][840/1251] eta 0:01:57 lr 0.000584 time 0.2725 (0.2869) loss 3.9221 (3.5784) grad_norm 1.4640 (1.4347) [2021-04-16 03:49:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][850/1251] eta 0:01:54 lr 0.000584 time 0.2696 (0.2868) loss 4.1394 (3.5783) grad_norm 1.4860 (1.4342) [2021-04-16 03:49:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][860/1251] eta 0:01:52 lr 0.000584 time 0.2844 (0.2867) loss 3.9707 (3.5787) grad_norm 1.3128 (1.4344) [2021-04-16 03:49:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][870/1251] eta 0:01:49 lr 0.000584 time 0.2731 (0.2865) loss 3.9205 (3.5824) grad_norm 1.3936 (1.4336) [2021-04-16 03:49:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][880/1251] eta 0:01:46 lr 0.000584 time 0.4489 (0.2866) loss 3.7874 (3.5829) grad_norm 1.5735 (1.4333) [2021-04-16 03:49:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][890/1251] eta 0:01:43 lr 0.000584 time 0.2762 (0.2864) loss 3.2048 (3.5850) grad_norm 1.2933 (1.4328) [2021-04-16 03:49:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][900/1251] eta 0:01:40 lr 0.000584 time 0.2838 (0.2863) loss 2.7937 (3.5848) grad_norm 1.3896 (1.4322) [2021-04-16 03:49:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][910/1251] eta 0:01:37 lr 0.000584 time 0.2662 (0.2864) loss 3.4272 (3.5823) grad_norm 1.4098 (1.4318) [2021-04-16 03:49:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][920/1251] eta 0:01:34 lr 0.000584 time 0.2666 (0.2863) loss 3.5980 (3.5848) grad_norm 1.4150 (1.4321) [2021-04-16 03:49:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][930/1251] eta 0:01:31 lr 0.000584 time 0.2976 (0.2864) loss 3.7005 (3.5869) grad_norm 1.2534 (1.4325) [2021-04-16 03:49:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][940/1251] eta 0:01:29 lr 0.000584 time 0.3610 (0.2865) loss 2.7929 (3.5837) grad_norm 1.4057 (1.4325) [2021-04-16 03:49:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][950/1251] eta 0:01:26 lr 0.000584 time 0.2897 (0.2864) loss 3.2624 (3.5820) grad_norm 1.3463 (1.4328) [2021-04-16 03:50:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][960/1251] eta 0:01:23 lr 0.000584 time 0.2742 (0.2862) loss 3.8593 (3.5803) grad_norm 1.5745 (1.4333) [2021-04-16 03:50:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][970/1251] eta 0:01:20 lr 0.000584 time 0.2962 (0.2861) loss 3.6881 (3.5793) grad_norm 1.3766 (1.4349) [2021-04-16 03:50:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][980/1251] eta 0:01:17 lr 0.000584 time 0.2718 (0.2860) loss 3.7358 (3.5795) grad_norm 1.3338 (1.4352) [2021-04-16 03:50:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][990/1251] eta 0:01:14 lr 0.000584 time 0.2665 (0.2859) loss 3.7829 (3.5809) grad_norm 1.4488 (1.4352) [2021-04-16 03:50:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][1000/1251] eta 0:01:11 lr 0.000583 time 0.2743 (0.2858) loss 4.0369 (3.5810) grad_norm 1.3516 (1.4347) [2021-04-16 03:50:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][1010/1251] eta 0:01:08 lr 0.000583 time 0.2724 (0.2857) loss 4.3743 (3.5830) grad_norm 1.4316 (1.4348) [2021-04-16 03:50:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][1020/1251] eta 0:01:05 lr 0.000583 time 0.2712 (0.2856) loss 2.8376 (3.5839) grad_norm 1.4914 (1.4342) [2021-04-16 03:50:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][1030/1251] eta 0:01:03 lr 0.000583 time 0.2838 (0.2855) loss 4.1646 (3.5837) grad_norm 1.4774 (1.4342) [2021-04-16 03:50:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][1040/1251] eta 0:01:00 lr 0.000583 time 0.2769 (0.2854) loss 3.6734 (3.5841) grad_norm 1.1489 (1.4335) [2021-04-16 03:50:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][1050/1251] eta 0:00:57 lr 0.000583 time 0.2747 (0.2853) loss 3.8131 (3.5871) grad_norm 1.5207 (1.4328) [2021-04-16 03:50:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][1060/1251] eta 0:00:54 lr 0.000583 time 0.2895 (0.2852) loss 4.0438 (3.5869) grad_norm 1.2901 (1.4324) [2021-04-16 03:50:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][1070/1251] eta 0:00:51 lr 0.000583 time 0.2790 (0.2851) loss 3.2955 (3.5844) grad_norm 1.5401 (1.4319) [2021-04-16 03:50:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][1080/1251] eta 0:00:48 lr 0.000583 time 0.2833 (0.2850) loss 3.4390 (3.5846) grad_norm 1.2808 (1.4317) [2021-04-16 03:50:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][1090/1251] eta 0:00:45 lr 0.000583 time 0.2772 (0.2849) loss 3.8012 (3.5873) grad_norm 1.5140 (1.4313) [2021-04-16 03:50:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][1100/1251] eta 0:00:43 lr 0.000583 time 0.2702 (0.2849) loss 3.7996 (3.5886) grad_norm 1.8183 (1.4325) [2021-04-16 03:50:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][1110/1251] eta 0:00:40 lr 0.000583 time 0.2702 (0.2848) loss 4.0126 (3.5901) grad_norm 1.3343 (1.4329) [2021-04-16 03:50:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][1120/1251] eta 0:00:37 lr 0.000583 time 0.2555 (0.2848) loss 4.5882 (3.5915) grad_norm 1.4701 (1.4334) [2021-04-16 03:50:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][1130/1251] eta 0:00:34 lr 0.000583 time 0.2870 (0.2848) loss 4.2196 (3.5916) grad_norm 1.6853 (1.4337) [2021-04-16 03:50:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][1140/1251] eta 0:00:31 lr 0.000583 time 0.2794 (0.2848) loss 3.6182 (3.5908) grad_norm 1.6567 (1.4341) [2021-04-16 03:50:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][1150/1251] eta 0:00:28 lr 0.000583 time 0.2626 (0.2848) loss 2.6262 (3.5900) grad_norm 1.4722 (1.4344) [2021-04-16 03:50:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][1160/1251] eta 0:00:25 lr 0.000583 time 0.2752 (0.2847) loss 3.6272 (3.5910) grad_norm 1.3340 (1.4351) [2021-04-16 03:51:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][1170/1251] eta 0:00:23 lr 0.000583 time 0.2769 (0.2847) loss 2.8821 (3.5919) grad_norm 1.5798 (1.4372) [2021-04-16 03:51:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][1180/1251] eta 0:00:20 lr 0.000583 time 0.2796 (0.2846) loss 3.8147 (3.5933) grad_norm 1.4669 (1.4385) [2021-04-16 03:51:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][1190/1251] eta 0:00:17 lr 0.000583 time 0.2721 (0.2845) loss 4.0905 (3.5933) grad_norm 1.3002 (1.4386) [2021-04-16 03:51:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][1200/1251] eta 0:00:14 lr 0.000583 time 0.2802 (0.2844) loss 4.4936 (3.5927) grad_norm 1.4104 (1.4384) [2021-04-16 03:51:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][1210/1251] eta 0:00:11 lr 0.000583 time 0.2797 (0.2844) loss 3.9645 (3.5952) grad_norm 1.4206 (1.4385) [2021-04-16 03:51:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][1220/1251] eta 0:00:08 lr 0.000583 time 0.2584 (0.2843) loss 3.2383 (3.5936) grad_norm 1.5003 (1.4385) [2021-04-16 03:51:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][1230/1251] eta 0:00:05 lr 0.000583 time 0.2985 (0.2844) loss 4.4988 (3.5925) grad_norm 1.5145 (1.4384) [2021-04-16 03:51:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][1240/1251] eta 0:00:03 lr 0.000582 time 0.2482 (0.2842) loss 3.7066 (3.5927) grad_norm 1.1883 (1.4382) [2021-04-16 03:51:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [134/300][1250/1251] eta 0:00:00 lr 0.000582 time 0.2491 (0.2839) loss 3.7508 (3.5930) grad_norm 1.4265 (1.4382) [2021-04-16 03:51:25 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 134 training takes 0:05:57 [2021-04-16 03:51:25 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_134.pth saving...... [2021-04-16 03:51:35 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_134.pth saved !!! [2021-04-16 03:51:37 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.228 (1.228) Loss 1.0838 (1.0838) Acc@1 74.805 (74.805) Acc@5 92.188 (92.188) [2021-04-16 03:51:38 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.113 (0.266) Loss 0.9489 (1.0569) Acc@1 76.270 (74.867) Acc@5 94.727 (93.058) [2021-04-16 03:51:40 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.121 (0.232) Loss 1.0248 (1.0434) Acc@1 75.195 (75.349) Acc@5 93.457 (93.085) [2021-04-16 03:51:43 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.111 (0.238) Loss 1.0079 (1.0411) Acc@1 75.195 (75.545) Acc@5 94.043 (93.032) [2021-04-16 03:51:44 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.215) Loss 1.0385 (1.0515) Acc@1 75.000 (75.286) Acc@5 93.359 (92.909) [2021-04-16 03:51:48 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 75.252 Acc@5 92.888 [2021-04-16 03:51:48 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 75.3% [2021-04-16 03:51:48 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 75.25% [2021-04-16 03:51:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][0/1251] eta 1:29:27 lr 0.000582 time 4.2903 (4.2903) loss 3.2012 (3.2012) grad_norm 1.2793 (1.2793) [2021-04-16 03:51:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][10/1251] eta 0:13:08 lr 0.000582 time 0.2515 (0.6354) loss 2.9396 (3.4769) grad_norm 1.3208 (1.4492) [2021-04-16 03:51:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][20/1251] eta 0:09:31 lr 0.000582 time 0.2652 (0.4644) loss 4.0436 (3.5344) grad_norm 1.2493 (1.4828) [2021-04-16 03:52:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][30/1251] eta 0:08:16 lr 0.000582 time 0.2813 (0.4064) loss 3.3612 (3.5881) grad_norm 1.3608 (1.4804) [2021-04-16 03:52:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][40/1251] eta 0:07:33 lr 0.000582 time 0.2814 (0.3744) loss 3.3115 (3.6209) grad_norm 1.5773 (1.4734) [2021-04-16 03:52:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][50/1251] eta 0:07:06 lr 0.000582 time 0.2769 (0.3554) loss 2.6462 (3.6323) grad_norm 1.4236 (1.4589) [2021-04-16 03:52:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][60/1251] eta 0:06:47 lr 0.000582 time 0.2923 (0.3425) loss 3.7917 (3.5984) grad_norm 1.3119 (1.4507) [2021-04-16 03:52:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][70/1251] eta 0:06:35 lr 0.000582 time 0.2523 (0.3347) loss 3.5960 (3.5863) grad_norm 1.2344 (1.4379) [2021-04-16 03:52:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][80/1251] eta 0:06:23 lr 0.000582 time 0.2617 (0.3279) loss 4.0496 (3.5603) grad_norm 1.3106 (1.4328) [2021-04-16 03:52:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][90/1251] eta 0:06:14 lr 0.000582 time 0.2816 (0.3222) loss 3.4553 (3.5659) grad_norm 1.5168 (1.4283) [2021-04-16 03:52:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][100/1251] eta 0:06:06 lr 0.000582 time 0.3014 (0.3187) loss 4.0788 (3.5936) grad_norm 1.2522 (1.4254) [2021-04-16 03:52:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][110/1251] eta 0:05:59 lr 0.000582 time 0.2692 (0.3150) loss 3.8660 (3.6041) grad_norm 1.3261 (1.4336) [2021-04-16 03:52:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][120/1251] eta 0:05:52 lr 0.000582 time 0.2988 (0.3120) loss 3.7588 (3.5753) grad_norm 1.4292 (1.4368) [2021-04-16 03:52:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][130/1251] eta 0:05:47 lr 0.000582 time 0.2754 (0.3102) loss 3.4939 (3.5834) grad_norm 1.5843 (1.4398) [2021-04-16 03:52:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][140/1251] eta 0:05:41 lr 0.000582 time 0.3183 (0.3076) loss 3.1800 (3.5692) grad_norm 1.3832 (1.4360) [2021-04-16 03:52:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][150/1251] eta 0:05:37 lr 0.000582 time 0.3836 (0.3062) loss 3.2241 (3.5666) grad_norm 1.4724 (1.4353) [2021-04-16 03:52:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][160/1251] eta 0:05:32 lr 0.000582 time 0.2710 (0.3050) loss 4.0589 (3.5807) grad_norm 1.4606 (1.4338) [2021-04-16 03:52:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][170/1251] eta 0:05:27 lr 0.000582 time 0.2597 (0.3033) loss 3.7776 (3.5776) grad_norm 1.2246 (1.4346) [2021-04-16 03:52:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][180/1251] eta 0:05:23 lr 0.000582 time 0.2514 (0.3017) loss 4.2331 (3.5882) grad_norm 1.3891 (1.4328) [2021-04-16 03:52:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][190/1251] eta 0:05:18 lr 0.000582 time 0.2944 (0.3004) loss 4.1060 (3.5808) grad_norm 1.2489 (1.4305) [2021-04-16 03:52:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][200/1251] eta 0:05:14 lr 0.000582 time 0.2798 (0.2992) loss 3.7140 (3.5843) grad_norm 1.3777 (1.4284) [2021-04-16 03:52:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][210/1251] eta 0:05:10 lr 0.000582 time 0.2788 (0.2983) loss 3.6530 (3.5821) grad_norm 1.3062 (1.4257) [2021-04-16 03:52:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][220/1251] eta 0:05:06 lr 0.000582 time 0.2661 (0.2971) loss 4.5785 (3.5807) grad_norm 1.4344 (nan) [2021-04-16 03:52:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][230/1251] eta 0:05:02 lr 0.000581 time 0.2769 (0.2961) loss 3.8875 (3.5854) grad_norm 1.4757 (nan) [2021-04-16 03:52:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][240/1251] eta 0:04:58 lr 0.000581 time 0.2771 (0.2954) loss 4.6687 (3.5796) grad_norm 1.5260 (nan) [2021-04-16 03:53:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][250/1251] eta 0:04:55 lr 0.000581 time 0.3070 (0.2949) loss 3.0883 (3.5813) grad_norm 1.3991 (nan) [2021-04-16 03:53:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][260/1251] eta 0:04:51 lr 0.000581 time 0.2836 (0.2941) loss 3.4123 (3.5878) grad_norm 1.4003 (nan) [2021-04-16 03:53:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][270/1251] eta 0:04:47 lr 0.000581 time 0.2652 (0.2934) loss 3.0365 (3.5981) grad_norm 1.3310 (nan) [2021-04-16 03:53:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][280/1251] eta 0:04:44 lr 0.000581 time 0.2725 (0.2928) loss 3.0140 (3.5998) grad_norm 1.5873 (nan) [2021-04-16 03:53:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][290/1251] eta 0:04:40 lr 0.000581 time 0.2697 (0.2921) loss 3.6214 (3.6016) grad_norm 1.6090 (nan) [2021-04-16 03:53:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][300/1251] eta 0:04:37 lr 0.000581 time 0.2641 (0.2914) loss 2.7136 (3.6050) grad_norm 1.4914 (nan) [2021-04-16 03:53:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][310/1251] eta 0:04:33 lr 0.000581 time 0.3079 (0.2911) loss 4.0405 (3.6103) grad_norm 1.9468 (nan) [2021-04-16 03:53:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][320/1251] eta 0:04:30 lr 0.000581 time 0.2472 (0.2908) loss 3.9750 (3.6111) grad_norm 1.4585 (nan) [2021-04-16 03:53:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][330/1251] eta 0:04:27 lr 0.000581 time 0.2753 (0.2903) loss 3.8645 (3.6121) grad_norm 1.2773 (nan) [2021-04-16 03:53:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][340/1251] eta 0:04:24 lr 0.000581 time 0.2454 (0.2904) loss 2.5776 (3.6092) grad_norm 1.4892 (nan) [2021-04-16 03:53:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][350/1251] eta 0:04:21 lr 0.000581 time 0.2777 (0.2906) loss 4.0753 (3.6104) grad_norm 1.2596 (nan) [2021-04-16 03:53:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][360/1251] eta 0:04:18 lr 0.000581 time 0.2673 (0.2905) loss 3.9501 (3.6058) grad_norm 1.4999 (nan) [2021-04-16 03:53:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][370/1251] eta 0:04:15 lr 0.000581 time 0.2736 (0.2901) loss 2.8441 (3.6034) grad_norm 1.5467 (nan) [2021-04-16 03:53:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][380/1251] eta 0:04:12 lr 0.000581 time 0.2770 (0.2900) loss 4.1621 (3.6122) grad_norm 1.4851 (nan) [2021-04-16 03:53:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][390/1251] eta 0:04:09 lr 0.000581 time 0.2732 (0.2896) loss 2.4571 (3.6075) grad_norm 1.3498 (nan) [2021-04-16 03:53:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][400/1251] eta 0:04:06 lr 0.000581 time 0.2928 (0.2893) loss 2.2646 (3.6114) grad_norm 1.3454 (nan) [2021-04-16 03:53:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][410/1251] eta 0:04:02 lr 0.000581 time 0.3032 (0.2889) loss 3.4588 (3.6072) grad_norm 1.3416 (nan) [2021-04-16 03:53:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][420/1251] eta 0:03:59 lr 0.000581 time 0.2824 (0.2886) loss 4.4812 (3.6068) grad_norm 1.3079 (nan) [2021-04-16 03:53:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][430/1251] eta 0:03:56 lr 0.000581 time 0.2654 (0.2882) loss 3.4197 (3.6095) grad_norm 1.3557 (nan) [2021-04-16 03:53:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][440/1251] eta 0:03:53 lr 0.000581 time 0.2761 (0.2880) loss 3.8500 (3.6130) grad_norm 1.3513 (nan) [2021-04-16 03:53:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][450/1251] eta 0:03:50 lr 0.000581 time 0.2547 (0.2878) loss 2.9812 (3.6147) grad_norm 1.3731 (nan) [2021-04-16 03:54:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][460/1251] eta 0:03:47 lr 0.000581 time 0.2719 (0.2876) loss 3.1521 (3.6081) grad_norm 1.5099 (nan) [2021-04-16 03:54:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][470/1251] eta 0:03:44 lr 0.000581 time 0.2709 (0.2873) loss 3.2100 (3.6082) grad_norm 1.1988 (nan) [2021-04-16 03:54:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][480/1251] eta 0:03:41 lr 0.000580 time 0.2818 (0.2872) loss 4.4107 (3.6114) grad_norm 1.4710 (nan) [2021-04-16 03:54:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][490/1251] eta 0:03:38 lr 0.000580 time 0.2785 (0.2870) loss 3.7885 (3.6113) grad_norm 1.2523 (nan) [2021-04-16 03:54:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][500/1251] eta 0:03:35 lr 0.000580 time 0.2654 (0.2867) loss 3.5320 (3.6117) grad_norm 1.5406 (nan) [2021-04-16 03:54:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][510/1251] eta 0:03:32 lr 0.000580 time 0.2562 (0.2866) loss 4.1105 (3.6139) grad_norm 1.3165 (nan) [2021-04-16 03:54:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][520/1251] eta 0:03:29 lr 0.000580 time 0.2865 (0.2864) loss 3.5248 (3.6134) grad_norm 1.2678 (nan) [2021-04-16 03:54:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][530/1251] eta 0:03:26 lr 0.000580 time 0.2704 (0.2863) loss 2.3936 (3.6111) grad_norm 1.3782 (nan) [2021-04-16 03:54:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][540/1251] eta 0:03:23 lr 0.000580 time 0.2818 (0.2861) loss 3.6251 (3.6136) grad_norm 1.6991 (nan) [2021-04-16 03:54:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][550/1251] eta 0:03:20 lr 0.000580 time 0.4177 (0.2861) loss 3.8791 (3.6165) grad_norm 1.3961 (nan) [2021-04-16 03:54:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][560/1251] eta 0:03:17 lr 0.000580 time 0.2636 (0.2860) loss 2.2085 (3.6150) grad_norm 1.3315 (nan) [2021-04-16 03:54:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][570/1251] eta 0:03:14 lr 0.000580 time 0.2593 (0.2858) loss 3.6433 (3.6189) grad_norm 1.4753 (nan) [2021-04-16 03:54:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][580/1251] eta 0:03:11 lr 0.000580 time 0.2625 (0.2858) loss 3.6460 (3.6204) grad_norm 1.4515 (nan) [2021-04-16 03:54:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][590/1251] eta 0:03:08 lr 0.000580 time 0.2634 (0.2856) loss 2.8036 (3.6153) grad_norm 1.2316 (nan) [2021-04-16 03:54:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][600/1251] eta 0:03:05 lr 0.000580 time 0.2890 (0.2855) loss 3.5621 (3.6163) grad_norm 1.3040 (nan) [2021-04-16 03:54:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][610/1251] eta 0:03:03 lr 0.000580 time 0.2868 (0.2856) loss 3.4736 (3.6136) grad_norm 1.5015 (nan) [2021-04-16 03:54:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][620/1251] eta 0:03:00 lr 0.000580 time 0.2823 (0.2855) loss 3.6133 (3.6140) grad_norm 1.5014 (nan) [2021-04-16 03:54:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][630/1251] eta 0:02:57 lr 0.000580 time 0.2534 (0.2853) loss 2.9738 (3.6147) grad_norm 1.4981 (nan) [2021-04-16 03:54:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][640/1251] eta 0:02:54 lr 0.000580 time 0.2470 (0.2854) loss 3.6112 (3.6121) grad_norm 1.3998 (nan) [2021-04-16 03:54:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][650/1251] eta 0:02:51 lr 0.000580 time 0.2847 (0.2852) loss 4.1921 (3.6101) grad_norm 1.3611 (nan) [2021-04-16 03:54:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][660/1251] eta 0:02:48 lr 0.000580 time 0.2579 (0.2850) loss 3.8179 (3.6103) grad_norm 1.5915 (nan) [2021-04-16 03:54:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][670/1251] eta 0:02:45 lr 0.000580 time 0.2866 (0.2848) loss 4.4616 (3.6102) grad_norm 1.3406 (nan) [2021-04-16 03:55:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][680/1251] eta 0:02:42 lr 0.000580 time 0.3116 (0.2849) loss 3.8157 (3.6087) grad_norm 1.5404 (nan) [2021-04-16 03:55:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][690/1251] eta 0:02:39 lr 0.000580 time 0.2897 (0.2848) loss 4.6778 (3.6107) grad_norm 1.4423 (nan) [2021-04-16 03:55:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][700/1251] eta 0:02:36 lr 0.000580 time 0.2790 (0.2847) loss 2.5742 (3.6110) grad_norm 1.6695 (nan) [2021-04-16 03:55:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][710/1251] eta 0:02:34 lr 0.000580 time 0.2735 (0.2847) loss 3.6751 (3.6125) grad_norm 1.4510 (nan) [2021-04-16 03:55:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][720/1251] eta 0:02:31 lr 0.000579 time 0.2760 (0.2846) loss 3.3511 (3.6126) grad_norm 1.3931 (nan) [2021-04-16 03:55:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][730/1251] eta 0:02:28 lr 0.000579 time 0.4072 (0.2848) loss 4.1932 (3.6146) grad_norm 1.8397 (nan) [2021-04-16 03:55:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][740/1251] eta 0:02:25 lr 0.000579 time 0.2849 (0.2847) loss 3.5706 (3.6113) grad_norm 1.1975 (nan) [2021-04-16 03:55:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][750/1251] eta 0:02:22 lr 0.000579 time 0.2868 (0.2846) loss 4.3099 (3.6135) grad_norm 1.2989 (nan) [2021-04-16 03:55:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][760/1251] eta 0:02:19 lr 0.000579 time 0.2621 (0.2845) loss 2.5444 (3.6112) grad_norm 1.4592 (nan) [2021-04-16 03:55:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][770/1251] eta 0:02:16 lr 0.000579 time 0.2811 (0.2844) loss 3.7505 (3.6118) grad_norm 1.9520 (nan) [2021-04-16 03:55:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][780/1251] eta 0:02:13 lr 0.000579 time 0.2573 (0.2842) loss 3.6504 (3.6116) grad_norm 1.5717 (nan) [2021-04-16 03:55:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][790/1251] eta 0:02:10 lr 0.000579 time 0.2863 (0.2841) loss 2.5525 (3.6100) grad_norm 1.5004 (nan) [2021-04-16 03:55:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][800/1251] eta 0:02:08 lr 0.000579 time 0.2629 (0.2840) loss 3.3020 (3.6069) grad_norm 1.3223 (nan) [2021-04-16 03:55:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][810/1251] eta 0:02:05 lr 0.000579 time 0.3037 (0.2839) loss 3.5869 (3.6062) grad_norm 1.3066 (nan) [2021-04-16 03:55:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][820/1251] eta 0:02:02 lr 0.000579 time 0.2567 (0.2838) loss 4.3331 (3.6083) grad_norm 1.6513 (nan) [2021-04-16 03:55:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][830/1251] eta 0:01:59 lr 0.000579 time 0.3001 (0.2838) loss 3.9190 (3.6132) grad_norm 1.4436 (nan) [2021-04-16 03:55:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][840/1251] eta 0:01:56 lr 0.000579 time 0.3007 (0.2838) loss 3.1593 (3.6111) grad_norm 1.4375 (nan) [2021-04-16 03:55:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][850/1251] eta 0:01:53 lr 0.000579 time 0.2844 (0.2836) loss 3.2199 (3.6119) grad_norm 1.4804 (nan) [2021-04-16 03:55:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][860/1251] eta 0:01:50 lr 0.000579 time 0.2814 (0.2835) loss 3.6813 (3.6134) grad_norm 1.3307 (nan) [2021-04-16 03:55:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][870/1251] eta 0:01:47 lr 0.000579 time 0.2522 (0.2834) loss 2.9368 (3.6145) grad_norm 1.4016 (nan) [2021-04-16 03:55:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][880/1251] eta 0:01:45 lr 0.000579 time 0.2758 (0.2834) loss 3.4906 (3.6141) grad_norm 1.3421 (nan) [2021-04-16 03:56:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][890/1251] eta 0:01:42 lr 0.000579 time 0.2867 (0.2833) loss 2.4027 (3.6140) grad_norm 1.2640 (nan) [2021-04-16 03:56:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][900/1251] eta 0:01:39 lr 0.000579 time 0.2819 (0.2832) loss 3.4440 (3.6139) grad_norm 1.3985 (nan) [2021-04-16 03:56:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][910/1251] eta 0:01:36 lr 0.000579 time 0.2561 (0.2831) loss 4.0037 (3.6138) grad_norm 1.5504 (nan) [2021-04-16 03:56:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][920/1251] eta 0:01:33 lr 0.000579 time 0.2839 (0.2831) loss 2.1607 (3.6109) grad_norm 1.5616 (nan) [2021-04-16 03:56:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][930/1251] eta 0:01:30 lr 0.000579 time 0.2608 (0.2832) loss 3.4041 (3.6082) grad_norm 1.3538 (nan) [2021-04-16 03:56:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][940/1251] eta 0:01:28 lr 0.000579 time 0.2733 (0.2833) loss 3.7652 (3.6093) grad_norm 1.2814 (nan) [2021-04-16 03:56:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][950/1251] eta 0:01:25 lr 0.000579 time 0.2882 (0.2832) loss 3.0556 (3.6103) grad_norm 1.6280 (nan) [2021-04-16 03:56:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][960/1251] eta 0:01:22 lr 0.000579 time 0.2961 (0.2832) loss 2.2435 (3.6095) grad_norm 1.7727 (nan) [2021-04-16 03:56:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][970/1251] eta 0:01:19 lr 0.000578 time 0.2833 (0.2832) loss 3.8820 (3.6096) grad_norm 1.6311 (nan) [2021-04-16 03:56:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][980/1251] eta 0:01:16 lr 0.000578 time 0.2829 (0.2831) loss 3.5289 (3.6108) grad_norm 1.7018 (nan) [2021-04-16 03:56:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][990/1251] eta 0:01:13 lr 0.000578 time 0.2704 (0.2830) loss 4.3927 (3.6120) grad_norm 1.6887 (nan) [2021-04-16 03:56:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][1000/1251] eta 0:01:11 lr 0.000578 time 0.3072 (0.2830) loss 3.0775 (3.6124) grad_norm 1.3911 (nan) [2021-04-16 03:56:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][1010/1251] eta 0:01:08 lr 0.000578 time 0.2580 (0.2829) loss 3.6047 (3.6128) grad_norm 1.3486 (nan) [2021-04-16 03:56:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][1020/1251] eta 0:01:05 lr 0.000578 time 0.2933 (0.2829) loss 3.9221 (3.6107) grad_norm 1.3234 (nan) [2021-04-16 03:56:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][1030/1251] eta 0:01:02 lr 0.000578 time 0.2744 (0.2828) loss 3.8269 (3.6124) grad_norm 1.3516 (nan) [2021-04-16 03:56:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][1040/1251] eta 0:00:59 lr 0.000578 time 0.2921 (0.2827) loss 4.2406 (3.6109) grad_norm 1.2201 (nan) [2021-04-16 03:56:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][1050/1251] eta 0:00:56 lr 0.000578 time 0.3107 (0.2827) loss 4.0656 (3.6129) grad_norm 1.2795 (nan) [2021-04-16 03:56:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][1060/1251] eta 0:00:53 lr 0.000578 time 0.2682 (0.2826) loss 3.0272 (3.6141) grad_norm 1.3105 (nan) [2021-04-16 03:56:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][1070/1251] eta 0:00:51 lr 0.000578 time 0.2825 (0.2826) loss 4.0025 (3.6141) grad_norm 1.4361 (nan) [2021-04-16 03:56:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][1080/1251] eta 0:00:48 lr 0.000578 time 0.2812 (0.2825) loss 3.4205 (3.6118) grad_norm 1.4034 (nan) [2021-04-16 03:56:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][1090/1251] eta 0:00:45 lr 0.000578 time 0.2661 (0.2825) loss 4.0621 (3.6106) grad_norm 1.4819 (nan) [2021-04-16 03:56:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][1100/1251] eta 0:00:42 lr 0.000578 time 0.3015 (0.2824) loss 3.8878 (3.6108) grad_norm 1.3856 (nan) [2021-04-16 03:57:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][1110/1251] eta 0:00:39 lr 0.000578 time 0.2583 (0.2824) loss 3.9679 (3.6096) grad_norm 1.5546 (nan) [2021-04-16 03:57:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][1120/1251] eta 0:00:36 lr 0.000578 time 0.2582 (0.2823) loss 3.1624 (3.6065) grad_norm 1.2974 (nan) [2021-04-16 03:57:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][1130/1251] eta 0:00:34 lr 0.000578 time 0.2752 (0.2824) loss 3.3021 (3.6061) grad_norm 1.5438 (nan) [2021-04-16 03:57:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][1140/1251] eta 0:00:31 lr 0.000578 time 0.2778 (0.2823) loss 3.6336 (3.6070) grad_norm 1.4252 (nan) [2021-04-16 03:57:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][1150/1251] eta 0:00:28 lr 0.000578 time 0.2623 (0.2824) loss 3.4873 (3.6064) grad_norm 1.3873 (nan) [2021-04-16 03:57:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][1160/1251] eta 0:00:25 lr 0.000578 time 0.2752 (0.2823) loss 3.1921 (3.6069) grad_norm 1.4640 (nan) [2021-04-16 03:57:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][1170/1251] eta 0:00:22 lr 0.000578 time 0.2771 (0.2823) loss 4.0359 (3.6059) grad_norm 1.4625 (nan) [2021-04-16 03:57:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][1180/1251] eta 0:00:20 lr 0.000578 time 0.2761 (0.2823) loss 2.7005 (3.6062) grad_norm 1.3920 (nan) [2021-04-16 03:57:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][1190/1251] eta 0:00:17 lr 0.000578 time 0.2610 (0.2823) loss 4.5094 (3.6054) grad_norm 1.4299 (nan) [2021-04-16 03:57:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][1200/1251] eta 0:00:14 lr 0.000578 time 0.2794 (0.2822) loss 4.4180 (3.6052) grad_norm 1.3213 (nan) [2021-04-16 03:57:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][1210/1251] eta 0:00:11 lr 0.000577 time 0.3146 (0.2822) loss 3.4982 (3.6077) grad_norm 1.4780 (nan) [2021-04-16 03:57:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][1220/1251] eta 0:00:08 lr 0.000577 time 0.2959 (0.2823) loss 4.1949 (3.6085) grad_norm 1.4038 (nan) [2021-04-16 03:57:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][1230/1251] eta 0:00:05 lr 0.000577 time 0.2874 (0.2822) loss 3.9948 (3.6101) grad_norm 1.2212 (nan) [2021-04-16 03:57:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][1240/1251] eta 0:00:03 lr 0.000577 time 0.2566 (0.2820) loss 4.5823 (3.6101) grad_norm 1.2907 (nan) [2021-04-16 03:57:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [135/300][1250/1251] eta 0:00:00 lr 0.000577 time 0.2485 (0.2818) loss 2.5050 (3.6103) grad_norm 1.4374 (nan) [2021-04-16 03:57:44 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 135 training takes 0:05:56 [2021-04-16 03:57:44 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_135.pth saving...... [2021-04-16 03:57:54 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_135.pth saved !!! [2021-04-16 03:57:55 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.205 (1.205) Loss 1.0510 (1.0510) Acc@1 76.660 (76.660) Acc@5 91.992 (91.992) [2021-04-16 03:57:57 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.164 (0.244) Loss 1.0220 (1.0628) Acc@1 75.879 (75.053) Acc@5 93.164 (92.765) [2021-04-16 03:57:59 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.161 (0.239) Loss 1.0048 (1.0501) Acc@1 75.977 (75.395) Acc@5 92.871 (92.764) [2021-04-16 03:58:01 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.128 (0.232) Loss 0.9769 (1.0489) Acc@1 77.148 (75.428) Acc@5 94.531 (92.896) [2021-04-16 03:58:03 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.224) Loss 1.0038 (1.0471) Acc@1 75.293 (75.569) Acc@5 94.336 (92.938) [2021-04-16 03:58:07 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 75.622 Acc@5 92.960 [2021-04-16 03:58:07 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 75.6% [2021-04-16 03:58:07 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 75.62% [2021-04-16 03:58:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][0/1251] eta 1:40:09 lr 0.000577 time 4.8034 (4.8034) loss 2.8595 (2.8595) grad_norm 1.2394 (1.2394) [2021-04-16 03:58:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][10/1251] eta 0:14:14 lr 0.000577 time 0.2739 (0.6884) loss 4.0706 (3.7174) grad_norm 1.2760 (1.4064) [2021-04-16 03:58:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][20/1251] eta 0:10:08 lr 0.000577 time 0.2859 (0.4946) loss 3.8032 (3.6695) grad_norm 1.3982 (1.4344) [2021-04-16 03:58:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][30/1251] eta 0:08:38 lr 0.000577 time 0.2901 (0.4245) loss 4.1047 (3.7138) grad_norm 1.1902 (1.4030) [2021-04-16 03:58:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][40/1251] eta 0:07:52 lr 0.000577 time 0.2899 (0.3905) loss 3.7447 (3.6365) grad_norm 1.3977 (1.4168) [2021-04-16 03:58:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][50/1251] eta 0:07:22 lr 0.000577 time 0.3017 (0.3681) loss 3.0171 (3.6117) grad_norm 1.7320 (1.4445) [2021-04-16 03:58:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][60/1251] eta 0:07:01 lr 0.000577 time 0.3012 (0.3535) loss 3.2574 (3.5779) grad_norm 1.8287 (1.4494) [2021-04-16 03:58:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][70/1251] eta 0:06:46 lr 0.000577 time 0.2969 (0.3441) loss 2.3964 (3.5804) grad_norm 1.2553 (1.4466) [2021-04-16 03:58:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][80/1251] eta 0:06:32 lr 0.000577 time 0.2736 (0.3354) loss 2.5814 (3.5601) grad_norm 1.4231 (1.4463) [2021-04-16 03:58:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][90/1251] eta 0:06:22 lr 0.000577 time 0.2717 (0.3291) loss 4.1689 (3.5565) grad_norm 1.5038 (1.4432) [2021-04-16 03:58:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][100/1251] eta 0:06:14 lr 0.000577 time 0.2514 (0.3255) loss 3.7625 (3.5434) grad_norm 1.4564 (1.4454) [2021-04-16 03:58:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][110/1251] eta 0:06:06 lr 0.000577 time 0.2691 (0.3211) loss 3.3032 (3.5219) grad_norm 1.2025 (1.4393) [2021-04-16 03:58:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][120/1251] eta 0:06:00 lr 0.000577 time 0.2710 (0.3187) loss 4.2357 (3.5287) grad_norm 1.3775 (1.4325) [2021-04-16 03:58:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][130/1251] eta 0:05:54 lr 0.000577 time 0.2721 (0.3163) loss 3.9417 (3.5311) grad_norm 1.2716 (1.4332) [2021-04-16 03:58:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][140/1251] eta 0:05:49 lr 0.000577 time 0.2713 (0.3144) loss 3.4418 (3.5207) grad_norm 1.6410 (1.4360) [2021-04-16 03:58:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][150/1251] eta 0:05:44 lr 0.000577 time 0.2423 (0.3132) loss 3.9546 (3.5298) grad_norm 1.3358 (1.4364) [2021-04-16 03:58:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][160/1251] eta 0:05:39 lr 0.000577 time 0.2716 (0.3109) loss 2.3442 (3.5102) grad_norm 1.3117 (1.4385) [2021-04-16 03:59:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][170/1251] eta 0:05:33 lr 0.000577 time 0.2453 (0.3087) loss 2.9549 (3.5120) grad_norm 1.4354 (1.4416) [2021-04-16 03:59:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][180/1251] eta 0:05:29 lr 0.000577 time 0.2430 (0.3073) loss 3.2363 (3.5170) grad_norm 1.4251 (1.4440) [2021-04-16 03:59:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][190/1251] eta 0:05:24 lr 0.000577 time 0.2979 (0.3060) loss 2.9888 (3.5172) grad_norm 1.1912 (1.4479) [2021-04-16 03:59:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][200/1251] eta 0:05:20 lr 0.000576 time 0.2750 (0.3046) loss 3.1326 (3.5213) grad_norm 1.3863 (1.4446) [2021-04-16 03:59:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][210/1251] eta 0:05:16 lr 0.000576 time 0.2808 (0.3037) loss 3.4095 (3.5163) grad_norm 1.7756 (1.4441) [2021-04-16 03:59:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][220/1251] eta 0:05:12 lr 0.000576 time 0.2962 (0.3029) loss 4.7050 (3.5203) grad_norm 1.2682 (1.4433) [2021-04-16 03:59:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][230/1251] eta 0:05:07 lr 0.000576 time 0.2824 (0.3015) loss 3.6126 (3.5295) grad_norm 1.2396 (1.4385) [2021-04-16 03:59:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][240/1251] eta 0:05:04 lr 0.000576 time 0.2894 (0.3007) loss 3.6306 (3.5383) grad_norm 1.3742 (1.4410) [2021-04-16 03:59:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][250/1251] eta 0:05:00 lr 0.000576 time 0.2901 (0.2998) loss 4.2434 (3.5421) grad_norm 1.7707 (1.4444) [2021-04-16 03:59:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][260/1251] eta 0:04:56 lr 0.000576 time 0.2959 (0.2988) loss 4.3378 (3.5492) grad_norm 1.6257 (1.4480) [2021-04-16 03:59:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][270/1251] eta 0:04:52 lr 0.000576 time 0.2917 (0.2979) loss 3.9688 (3.5561) grad_norm 1.4168 (1.4458) [2021-04-16 03:59:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][280/1251] eta 0:04:48 lr 0.000576 time 0.2519 (0.2970) loss 4.0630 (3.5537) grad_norm 1.2396 (1.4457) [2021-04-16 03:59:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][290/1251] eta 0:04:44 lr 0.000576 time 0.2985 (0.2963) loss 3.8502 (3.5449) grad_norm 1.2126 (1.4451) [2021-04-16 03:59:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][300/1251] eta 0:04:41 lr 0.000576 time 0.2910 (0.2956) loss 3.4557 (3.5449) grad_norm 1.2908 (1.4434) [2021-04-16 03:59:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][310/1251] eta 0:04:37 lr 0.000576 time 0.2848 (0.2953) loss 4.1068 (3.5465) grad_norm 1.5236 (1.4393) [2021-04-16 03:59:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][320/1251] eta 0:04:34 lr 0.000576 time 0.2987 (0.2946) loss 3.4683 (3.5478) grad_norm 1.2774 (1.4401) [2021-04-16 03:59:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][330/1251] eta 0:04:30 lr 0.000576 time 0.2981 (0.2940) loss 3.5905 (3.5458) grad_norm 1.3538 (1.4384) [2021-04-16 03:59:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][340/1251] eta 0:04:27 lr 0.000576 time 0.2797 (0.2940) loss 2.4759 (3.5431) grad_norm 1.4140 (1.4407) [2021-04-16 03:59:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][350/1251] eta 0:04:24 lr 0.000576 time 0.2644 (0.2933) loss 2.9185 (3.5450) grad_norm 1.3817 (1.4405) [2021-04-16 03:59:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][360/1251] eta 0:04:21 lr 0.000576 time 0.2685 (0.2933) loss 3.6424 (3.5489) grad_norm 1.3268 (1.4403) [2021-04-16 03:59:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][370/1251] eta 0:04:17 lr 0.000576 time 0.2841 (0.2927) loss 4.3150 (3.5562) grad_norm 1.4233 (1.4402) [2021-04-16 03:59:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][380/1251] eta 0:04:14 lr 0.000576 time 0.2759 (0.2926) loss 3.4505 (3.5611) grad_norm 1.3730 (1.4393) [2021-04-16 04:00:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][390/1251] eta 0:04:11 lr 0.000576 time 0.2687 (0.2921) loss 2.6675 (3.5614) grad_norm 1.3490 (1.4383) [2021-04-16 04:00:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][400/1251] eta 0:04:08 lr 0.000576 time 0.2774 (0.2921) loss 3.4356 (3.5544) grad_norm 1.4738 (1.4375) [2021-04-16 04:00:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][410/1251] eta 0:04:05 lr 0.000576 time 0.2900 (0.2917) loss 4.1416 (3.5627) grad_norm 1.5795 (1.4376) [2021-04-16 04:00:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][420/1251] eta 0:04:02 lr 0.000576 time 0.2737 (0.2914) loss 3.5250 (3.5690) grad_norm 1.6178 (1.4409) [2021-04-16 04:00:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][430/1251] eta 0:03:58 lr 0.000576 time 0.2679 (0.2911) loss 3.7120 (3.5727) grad_norm 1.4705 (1.4416) [2021-04-16 04:00:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][440/1251] eta 0:03:55 lr 0.000576 time 0.2671 (0.2908) loss 3.0530 (3.5705) grad_norm 1.3579 (1.4420) [2021-04-16 04:00:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][450/1251] eta 0:03:52 lr 0.000575 time 0.2631 (0.2904) loss 3.8040 (3.5713) grad_norm 1.2427 (1.4414) [2021-04-16 04:00:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][460/1251] eta 0:03:49 lr 0.000575 time 0.2467 (0.2900) loss 3.3258 (3.5691) grad_norm 1.2128 (1.4403) [2021-04-16 04:00:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][470/1251] eta 0:03:46 lr 0.000575 time 0.2612 (0.2897) loss 4.6515 (3.5734) grad_norm 1.4329 (1.4391) [2021-04-16 04:00:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][480/1251] eta 0:03:43 lr 0.000575 time 0.2676 (0.2895) loss 3.7119 (3.5780) grad_norm 1.5331 (1.4372) [2021-04-16 04:00:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][490/1251] eta 0:03:40 lr 0.000575 time 0.2724 (0.2895) loss 4.0611 (3.5807) grad_norm 1.2220 (1.4355) [2021-04-16 04:00:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][500/1251] eta 0:03:37 lr 0.000575 time 0.2959 (0.2893) loss 4.0127 (3.5805) grad_norm 1.4367 (1.4362) [2021-04-16 04:00:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][510/1251] eta 0:03:34 lr 0.000575 time 0.2814 (0.2892) loss 4.2168 (3.5802) grad_norm 1.2484 (1.4355) [2021-04-16 04:00:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][520/1251] eta 0:03:31 lr 0.000575 time 0.2428 (0.2893) loss 3.7085 (3.5803) grad_norm 1.3393 (1.4352) [2021-04-16 04:00:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][530/1251] eta 0:03:28 lr 0.000575 time 0.2722 (0.2891) loss 3.9704 (3.5790) grad_norm 1.6573 (1.4373) [2021-04-16 04:00:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][540/1251] eta 0:03:25 lr 0.000575 time 0.2772 (0.2890) loss 3.9334 (3.5806) grad_norm 1.8661 (1.4399) [2021-04-16 04:00:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][550/1251] eta 0:03:22 lr 0.000575 time 0.2995 (0.2888) loss 2.3818 (3.5810) grad_norm 1.6013 (1.4392) [2021-04-16 04:00:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][560/1251] eta 0:03:19 lr 0.000575 time 0.2858 (0.2885) loss 3.4744 (3.5802) grad_norm 1.9196 (1.4393) [2021-04-16 04:00:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][570/1251] eta 0:03:16 lr 0.000575 time 0.2878 (0.2885) loss 4.0017 (3.5794) grad_norm 1.2439 (1.4393) [2021-04-16 04:00:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][580/1251] eta 0:03:13 lr 0.000575 time 0.2686 (0.2886) loss 4.0196 (3.5776) grad_norm 1.4283 (1.4387) [2021-04-16 04:00:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][590/1251] eta 0:03:10 lr 0.000575 time 0.2657 (0.2884) loss 3.7058 (3.5760) grad_norm 1.2604 (1.4385) [2021-04-16 04:01:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][600/1251] eta 0:03:07 lr 0.000575 time 0.2610 (0.2881) loss 2.8231 (3.5765) grad_norm 1.3453 (1.4382) [2021-04-16 04:01:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][610/1251] eta 0:03:04 lr 0.000575 time 0.2846 (0.2880) loss 3.9729 (3.5827) grad_norm 1.4597 (1.4377) [2021-04-16 04:01:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][620/1251] eta 0:03:01 lr 0.000575 time 0.2594 (0.2878) loss 3.2488 (3.5828) grad_norm 1.5169 (1.4374) [2021-04-16 04:01:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][630/1251] eta 0:02:58 lr 0.000575 time 0.2698 (0.2876) loss 4.0249 (3.5817) grad_norm 1.4362 (1.4372) [2021-04-16 04:01:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][640/1251] eta 0:02:55 lr 0.000575 time 0.2760 (0.2874) loss 3.1222 (3.5784) grad_norm 1.4076 (1.4360) [2021-04-16 04:01:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][650/1251] eta 0:02:52 lr 0.000575 time 0.2721 (0.2872) loss 3.6140 (3.5819) grad_norm 1.3969 (1.4348) [2021-04-16 04:01:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][660/1251] eta 0:02:49 lr 0.000575 time 0.2455 (0.2870) loss 4.1731 (3.5830) grad_norm 1.4061 (1.4339) [2021-04-16 04:01:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][670/1251] eta 0:02:46 lr 0.000575 time 0.2756 (0.2868) loss 3.3980 (3.5828) grad_norm 1.6542 (1.4354) [2021-04-16 04:01:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][680/1251] eta 0:02:43 lr 0.000575 time 0.2785 (0.2869) loss 2.5539 (3.5846) grad_norm 1.6367 (1.4380) [2021-04-16 04:01:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][690/1251] eta 0:02:41 lr 0.000574 time 0.2762 (0.2870) loss 3.1347 (3.5839) grad_norm 1.2689 (1.4373) [2021-04-16 04:01:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][700/1251] eta 0:02:38 lr 0.000574 time 0.3855 (0.2870) loss 4.1752 (3.5827) grad_norm 1.3961 (1.4364) [2021-04-16 04:01:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][710/1251] eta 0:02:35 lr 0.000574 time 0.4464 (0.2871) loss 4.3263 (3.5845) grad_norm 1.3667 (1.4377) [2021-04-16 04:01:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][720/1251] eta 0:02:32 lr 0.000574 time 0.3172 (0.2870) loss 2.3734 (3.5885) grad_norm 1.4720 (1.4378) [2021-04-16 04:01:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][730/1251] eta 0:02:29 lr 0.000574 time 0.2959 (0.2870) loss 4.0030 (3.5896) grad_norm 1.4939 (1.4387) [2021-04-16 04:01:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][740/1251] eta 0:02:26 lr 0.000574 time 0.2840 (0.2869) loss 2.6287 (3.5929) grad_norm 1.5348 (1.4384) [2021-04-16 04:01:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][750/1251] eta 0:02:23 lr 0.000574 time 0.2507 (0.2867) loss 3.8795 (3.5946) grad_norm 1.3154 (1.4390) [2021-04-16 04:01:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][760/1251] eta 0:02:20 lr 0.000574 time 0.2801 (0.2866) loss 3.1806 (3.5946) grad_norm 1.7225 (1.4404) [2021-04-16 04:01:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][770/1251] eta 0:02:17 lr 0.000574 time 0.2692 (0.2864) loss 2.9474 (3.5939) grad_norm 1.6073 (1.4404) [2021-04-16 04:01:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][780/1251] eta 0:02:14 lr 0.000574 time 0.2707 (0.2863) loss 3.9532 (3.5975) grad_norm 1.3890 (1.4409) [2021-04-16 04:01:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][790/1251] eta 0:02:11 lr 0.000574 time 0.2855 (0.2862) loss 2.7242 (3.5951) grad_norm 1.3679 (1.4412) [2021-04-16 04:01:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][800/1251] eta 0:02:08 lr 0.000574 time 0.2489 (0.2860) loss 3.1165 (3.5945) grad_norm 1.4443 (1.4418) [2021-04-16 04:01:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][810/1251] eta 0:02:06 lr 0.000574 time 0.2702 (0.2859) loss 4.1248 (3.5954) grad_norm 1.5041 (1.4427) [2021-04-16 04:02:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][820/1251] eta 0:02:03 lr 0.000574 time 0.2550 (0.2858) loss 2.8120 (3.5923) grad_norm 1.3978 (1.4419) [2021-04-16 04:02:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][830/1251] eta 0:02:00 lr 0.000574 time 0.2661 (0.2856) loss 3.1925 (3.5917) grad_norm 1.6599 (1.4413) [2021-04-16 04:02:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][840/1251] eta 0:01:57 lr 0.000574 time 0.2872 (0.2855) loss 4.4156 (3.5913) grad_norm 1.3090 (1.4413) [2021-04-16 04:02:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][850/1251] eta 0:01:54 lr 0.000574 time 0.2696 (0.2854) loss 3.4095 (3.5929) grad_norm 1.4663 (1.4422) [2021-04-16 04:02:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][860/1251] eta 0:01:51 lr 0.000574 time 0.2760 (0.2853) loss 4.3565 (3.5940) grad_norm 1.3255 (1.4424) [2021-04-16 04:02:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][870/1251] eta 0:01:48 lr 0.000574 time 0.2600 (0.2853) loss 3.6700 (3.5929) grad_norm 1.2381 (1.4415) [2021-04-16 04:02:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][880/1251] eta 0:01:45 lr 0.000574 time 0.2687 (0.2851) loss 3.8834 (3.5932) grad_norm 1.2117 (1.4425) [2021-04-16 04:02:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][890/1251] eta 0:01:42 lr 0.000574 time 0.2864 (0.2851) loss 3.0530 (3.5948) grad_norm 1.5738 (1.4422) [2021-04-16 04:02:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][900/1251] eta 0:01:40 lr 0.000574 time 0.3027 (0.2851) loss 4.3463 (3.5960) grad_norm 1.5265 (1.4419) [2021-04-16 04:02:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][910/1251] eta 0:01:37 lr 0.000574 time 0.2747 (0.2850) loss 3.9208 (3.5957) grad_norm 1.4302 (1.4410) [2021-04-16 04:02:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][920/1251] eta 0:01:34 lr 0.000574 time 0.2962 (0.2850) loss 3.4758 (3.5946) grad_norm 1.3407 (1.4402) [2021-04-16 04:02:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][930/1251] eta 0:01:31 lr 0.000573 time 0.2499 (0.2849) loss 3.6768 (3.5947) grad_norm 1.4818 (1.4403) [2021-04-16 04:02:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][940/1251] eta 0:01:28 lr 0.000573 time 0.2889 (0.2850) loss 4.1282 (3.5931) grad_norm 1.4467 (1.4405) [2021-04-16 04:02:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][950/1251] eta 0:01:25 lr 0.000573 time 0.2992 (0.2849) loss 3.2700 (3.5909) grad_norm 1.7978 (1.4413) [2021-04-16 04:02:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][960/1251] eta 0:01:22 lr 0.000573 time 0.3070 (0.2850) loss 2.9394 (3.5917) grad_norm 1.5823 (1.4435) [2021-04-16 04:02:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][970/1251] eta 0:01:20 lr 0.000573 time 0.2908 (0.2848) loss 3.8391 (3.5936) grad_norm 1.6859 (1.4462) [2021-04-16 04:02:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][980/1251] eta 0:01:17 lr 0.000573 time 0.2790 (0.2848) loss 3.1389 (3.5969) grad_norm 1.3604 (1.4481) [2021-04-16 04:02:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][990/1251] eta 0:01:14 lr 0.000573 time 0.3107 (0.2848) loss 3.7959 (3.5968) grad_norm 1.6779 (1.4493) [2021-04-16 04:02:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][1000/1251] eta 0:01:11 lr 0.000573 time 0.2723 (0.2847) loss 3.1897 (3.5947) grad_norm 1.3412 (1.4494) [2021-04-16 04:02:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][1010/1251] eta 0:01:08 lr 0.000573 time 0.2889 (0.2845) loss 3.8324 (3.5939) grad_norm 1.4803 (1.4490) [2021-04-16 04:02:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][1020/1251] eta 0:01:05 lr 0.000573 time 0.2864 (0.2845) loss 3.4679 (3.5922) grad_norm 1.3838 (1.4498) [2021-04-16 04:03:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][1030/1251] eta 0:01:02 lr 0.000573 time 0.2807 (0.2844) loss 3.3792 (3.5951) grad_norm 1.3401 (1.4491) [2021-04-16 04:03:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][1040/1251] eta 0:01:00 lr 0.000573 time 0.2794 (0.2844) loss 3.9828 (3.5963) grad_norm 1.4063 (1.4492) [2021-04-16 04:03:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][1050/1251] eta 0:00:57 lr 0.000573 time 0.2688 (0.2843) loss 3.9857 (3.5970) grad_norm 1.5732 (1.4501) [2021-04-16 04:03:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][1060/1251] eta 0:00:54 lr 0.000573 time 0.2591 (0.2842) loss 3.4091 (3.5949) grad_norm 1.3810 (1.4500) [2021-04-16 04:03:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][1070/1251] eta 0:00:51 lr 0.000573 time 0.2931 (0.2843) loss 3.4065 (3.5961) grad_norm 1.2976 (1.4491) [2021-04-16 04:03:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][1080/1251] eta 0:00:48 lr 0.000573 time 0.2647 (0.2842) loss 3.0750 (3.5934) grad_norm 1.3309 (1.4488) [2021-04-16 04:03:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][1090/1251] eta 0:00:45 lr 0.000573 time 0.2589 (0.2844) loss 3.8420 (3.5930) grad_norm 1.4906 (1.4498) [2021-04-16 04:03:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][1100/1251] eta 0:00:42 lr 0.000573 time 0.2763 (0.2842) loss 3.2918 (3.5929) grad_norm 1.3503 (1.4495) [2021-04-16 04:03:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][1110/1251] eta 0:00:40 lr 0.000573 time 0.2644 (0.2841) loss 2.4358 (3.5894) grad_norm 1.2640 (1.4487) [2021-04-16 04:03:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][1120/1251] eta 0:00:37 lr 0.000573 time 0.2810 (0.2841) loss 2.7295 (3.5868) grad_norm 1.6550 (1.4493) [2021-04-16 04:03:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][1130/1251] eta 0:00:34 lr 0.000573 time 0.2860 (0.2841) loss 4.5199 (3.5879) grad_norm 1.3172 (1.4486) [2021-04-16 04:03:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][1140/1251] eta 0:00:31 lr 0.000573 time 0.3020 (0.2842) loss 3.9352 (3.5892) grad_norm 1.3861 (1.4489) [2021-04-16 04:03:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][1150/1251] eta 0:00:28 lr 0.000573 time 0.2977 (0.2841) loss 4.1913 (3.5886) grad_norm 1.4364 (1.4490) [2021-04-16 04:03:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][1160/1251] eta 0:00:25 lr 0.000573 time 0.2785 (0.2842) loss 3.9116 (3.5870) grad_norm 1.5270 (1.4488) [2021-04-16 04:03:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][1170/1251] eta 0:00:23 lr 0.000573 time 0.2700 (0.2841) loss 4.1232 (3.5840) grad_norm 1.4047 (1.4495) [2021-04-16 04:03:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][1180/1251] eta 0:00:20 lr 0.000572 time 0.2650 (0.2840) loss 3.9064 (3.5867) grad_norm 1.3388 (1.4512) [2021-04-16 04:03:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][1190/1251] eta 0:00:17 lr 0.000572 time 0.2739 (0.2840) loss 3.4392 (3.5856) grad_norm 1.5392 (1.4516) [2021-04-16 04:03:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][1200/1251] eta 0:00:14 lr 0.000572 time 0.2662 (0.2839) loss 3.2912 (3.5865) grad_norm 1.3470 (1.4521) [2021-04-16 04:03:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][1210/1251] eta 0:00:11 lr 0.000572 time 0.2622 (0.2839) loss 3.3070 (3.5865) grad_norm 1.7886 (1.4522) [2021-04-16 04:03:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][1220/1251] eta 0:00:08 lr 0.000572 time 0.2961 (0.2839) loss 4.3088 (3.5891) grad_norm 1.4682 (1.4528) [2021-04-16 04:03:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][1230/1251] eta 0:00:05 lr 0.000572 time 0.2902 (0.2838) loss 3.8800 (3.5921) grad_norm 1.2993 (1.4524) [2021-04-16 04:03:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][1240/1251] eta 0:00:03 lr 0.000572 time 0.2484 (0.2837) loss 3.0914 (3.5920) grad_norm 1.3401 (1.4517) [2021-04-16 04:04:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [136/300][1250/1251] eta 0:00:00 lr 0.000572 time 0.2491 (0.2834) loss 2.8187 (3.5910) grad_norm 1.3126 (1.4511) [2021-04-16 04:04:05 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 136 training takes 0:05:58 [2021-04-16 04:04:05 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_136.pth saving...... [2021-04-16 04:04:13 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_136.pth saved !!! [2021-04-16 04:04:15 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.071 (1.071) Loss 1.0268 (1.0268) Acc@1 76.074 (76.074) Acc@5 94.043 (94.043) [2021-04-16 04:04:16 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.145 (0.232) Loss 0.9868 (1.0732) Acc@1 76.562 (75.018) Acc@5 93.555 (92.907) [2021-04-16 04:04:18 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.270 (0.212) Loss 1.0851 (1.0744) Acc@1 75.195 (74.893) Acc@5 92.773 (92.815) [2021-04-16 04:04:20 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.162 (0.220) Loss 1.0799 (1.0692) Acc@1 75.195 (75.173) Acc@5 93.164 (92.893) [2021-04-16 04:04:22 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.091 (0.212) Loss 1.0415 (1.0650) Acc@1 76.562 (75.255) Acc@5 93.359 (92.888) [2021-04-16 04:04:26 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 75.392 Acc@5 92.966 [2021-04-16 04:04:26 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 75.4% [2021-04-16 04:04:26 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 75.62% [2021-04-16 04:04:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][0/1251] eta 1:38:22 lr 0.000572 time 4.7184 (4.7184) loss 3.7264 (3.7264) grad_norm 1.4377 (1.4377) [2021-04-16 04:04:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][10/1251] eta 0:14:12 lr 0.000572 time 0.2609 (0.6871) loss 4.1086 (3.4456) grad_norm 1.3809 (1.4351) [2021-04-16 04:04:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][20/1251] eta 0:10:06 lr 0.000572 time 0.2627 (0.4929) loss 3.8424 (3.6147) grad_norm 1.5107 (1.4038) [2021-04-16 04:04:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][30/1251] eta 0:08:38 lr 0.000572 time 0.2785 (0.4243) loss 2.7036 (3.5927) grad_norm 1.4485 (1.4293) [2021-04-16 04:04:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][40/1251] eta 0:07:49 lr 0.000572 time 0.2673 (0.3876) loss 2.9655 (3.6067) grad_norm 1.4970 (1.4321) [2021-04-16 04:04:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][50/1251] eta 0:07:19 lr 0.000572 time 0.2953 (0.3663) loss 2.6750 (3.5858) grad_norm 1.6048 (1.4367) [2021-04-16 04:04:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][60/1251] eta 0:06:58 lr 0.000572 time 0.2864 (0.3513) loss 3.5129 (3.5564) grad_norm 1.3130 (1.4239) [2021-04-16 04:04:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][70/1251] eta 0:06:46 lr 0.000572 time 0.3515 (0.3443) loss 3.1330 (3.5047) grad_norm 1.2628 (1.4294) [2021-04-16 04:04:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][80/1251] eta 0:06:32 lr 0.000572 time 0.2636 (0.3355) loss 4.3084 (3.5312) grad_norm 1.5815 (1.4427) [2021-04-16 04:04:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][90/1251] eta 0:06:21 lr 0.000572 time 0.2470 (0.3290) loss 3.3515 (3.5199) grad_norm 1.3068 (1.4480) [2021-04-16 04:04:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][100/1251] eta 0:06:12 lr 0.000572 time 0.2898 (0.3238) loss 2.6725 (3.4927) grad_norm 1.5217 (1.4418) [2021-04-16 04:05:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][110/1251] eta 0:06:05 lr 0.000572 time 0.2631 (0.3202) loss 4.0453 (3.4987) grad_norm 1.2667 (1.4358) [2021-04-16 04:05:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][120/1251] eta 0:05:59 lr 0.000572 time 0.2613 (0.3175) loss 3.0401 (3.5019) grad_norm 1.5075 (1.4348) [2021-04-16 04:05:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][130/1251] eta 0:05:52 lr 0.000572 time 0.2963 (0.3145) loss 3.9090 (3.5150) grad_norm 1.3308 (1.4301) [2021-04-16 04:05:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][140/1251] eta 0:05:46 lr 0.000572 time 0.2431 (0.3120) loss 3.3679 (3.5117) grad_norm 1.3345 (1.4297) [2021-04-16 04:05:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][150/1251] eta 0:05:41 lr 0.000572 time 0.2785 (0.3106) loss 3.2547 (3.5102) grad_norm 1.4639 (1.4329) [2021-04-16 04:05:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][160/1251] eta 0:05:36 lr 0.000572 time 0.2793 (0.3086) loss 3.8709 (3.5029) grad_norm 1.4549 (1.4366) [2021-04-16 04:05:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][170/1251] eta 0:05:32 lr 0.000571 time 0.2952 (0.3076) loss 4.2061 (3.5039) grad_norm 1.4571 (1.4408) [2021-04-16 04:05:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][180/1251] eta 0:05:28 lr 0.000571 time 0.4431 (0.3070) loss 3.8964 (3.5040) grad_norm 1.3066 (1.4359) [2021-04-16 04:05:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][190/1251] eta 0:05:23 lr 0.000571 time 0.2680 (0.3053) loss 3.2346 (3.5042) grad_norm 1.5606 (1.4335) [2021-04-16 04:05:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][200/1251] eta 0:05:19 lr 0.000571 time 0.2690 (0.3039) loss 3.5888 (3.5120) grad_norm 1.3280 (1.4329) [2021-04-16 04:05:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][210/1251] eta 0:05:14 lr 0.000571 time 0.2745 (0.3023) loss 3.8881 (3.5197) grad_norm 1.3267 (1.4316) [2021-04-16 04:05:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][220/1251] eta 0:05:10 lr 0.000571 time 0.2590 (0.3011) loss 3.0987 (3.5252) grad_norm 1.4002 (1.4310) [2021-04-16 04:05:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][230/1251] eta 0:05:06 lr 0.000571 time 0.2900 (0.3003) loss 4.1416 (3.5259) grad_norm 1.4690 (1.4272) [2021-04-16 04:05:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][240/1251] eta 0:05:02 lr 0.000571 time 0.2757 (0.2994) loss 4.5013 (3.5274) grad_norm 1.3895 (1.4277) [2021-04-16 04:05:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][250/1251] eta 0:04:58 lr 0.000571 time 0.2743 (0.2984) loss 4.4013 (3.5277) grad_norm 1.6424 (1.4290) [2021-04-16 04:05:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][260/1251] eta 0:04:55 lr 0.000571 time 0.2696 (0.2977) loss 3.1053 (3.5230) grad_norm 1.4286 (1.4320) [2021-04-16 04:05:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][270/1251] eta 0:04:51 lr 0.000571 time 0.2554 (0.2967) loss 4.2189 (3.5309) grad_norm 1.5120 (1.4338) [2021-04-16 04:05:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][280/1251] eta 0:04:47 lr 0.000571 time 0.2952 (0.2960) loss 3.6286 (3.5283) grad_norm 1.7223 (1.4334) [2021-04-16 04:05:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][290/1251] eta 0:04:44 lr 0.000571 time 0.2748 (0.2956) loss 4.3397 (3.5394) grad_norm 1.7879 (1.4378) [2021-04-16 04:05:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][300/1251] eta 0:04:41 lr 0.000571 time 0.2816 (0.2955) loss 3.7953 (3.5274) grad_norm 1.2169 (1.4362) [2021-04-16 04:05:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][310/1251] eta 0:04:37 lr 0.000571 time 0.2759 (0.2949) loss 2.5112 (3.5204) grad_norm 1.5561 (1.4356) [2021-04-16 04:06:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][320/1251] eta 0:04:34 lr 0.000571 time 0.2941 (0.2946) loss 3.9275 (3.5272) grad_norm 1.3502 (1.4364) [2021-04-16 04:06:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][330/1251] eta 0:04:30 lr 0.000571 time 0.2707 (0.2941) loss 3.2994 (3.5255) grad_norm 1.6951 (1.4388) [2021-04-16 04:06:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][340/1251] eta 0:04:27 lr 0.000571 time 0.2989 (0.2938) loss 3.2830 (3.5250) grad_norm 1.5501 (1.4385) [2021-04-16 04:06:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][350/1251] eta 0:04:24 lr 0.000571 time 0.2841 (0.2937) loss 3.3687 (3.5285) grad_norm 1.4763 (1.4410) [2021-04-16 04:06:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][360/1251] eta 0:04:21 lr 0.000571 time 0.2955 (0.2936) loss 4.0214 (3.5303) grad_norm 1.4228 (1.4438) [2021-04-16 04:06:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][370/1251] eta 0:04:18 lr 0.000571 time 0.2803 (0.2932) loss 3.8446 (3.5266) grad_norm 1.3846 (1.4413) [2021-04-16 04:06:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][380/1251] eta 0:04:15 lr 0.000571 time 0.2512 (0.2928) loss 4.0752 (3.5222) grad_norm 1.5197 (1.4414) [2021-04-16 04:06:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][390/1251] eta 0:04:11 lr 0.000571 time 0.2923 (0.2924) loss 4.6605 (3.5268) grad_norm 1.4971 (1.4418) [2021-04-16 04:06:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][400/1251] eta 0:04:08 lr 0.000571 time 0.2910 (0.2920) loss 3.7445 (3.5227) grad_norm 1.2881 (1.4410) [2021-04-16 04:06:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][410/1251] eta 0:04:05 lr 0.000570 time 0.2662 (0.2915) loss 3.8876 (3.5252) grad_norm 1.6061 (1.4413) [2021-04-16 04:06:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][420/1251] eta 0:04:02 lr 0.000570 time 0.2881 (0.2912) loss 4.0530 (3.5307) grad_norm 1.3495 (1.4429) [2021-04-16 04:06:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][430/1251] eta 0:03:58 lr 0.000570 time 0.2826 (0.2910) loss 2.8719 (3.5366) grad_norm 1.2250 (1.4429) [2021-04-16 04:06:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][440/1251] eta 0:03:55 lr 0.000570 time 0.2863 (0.2907) loss 1.9577 (3.5326) grad_norm 1.3339 (1.4425) [2021-04-16 04:06:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][450/1251] eta 0:03:52 lr 0.000570 time 0.2844 (0.2904) loss 2.7534 (3.5314) grad_norm 1.5292 (1.4405) [2021-04-16 04:06:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][460/1251] eta 0:03:49 lr 0.000570 time 0.3121 (0.2902) loss 3.9720 (3.5283) grad_norm 1.3476 (1.4395) [2021-04-16 04:06:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][470/1251] eta 0:03:46 lr 0.000570 time 0.2746 (0.2903) loss 4.1159 (3.5289) grad_norm 1.6363 (1.4400) [2021-04-16 04:06:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][480/1251] eta 0:03:43 lr 0.000570 time 0.2838 (0.2899) loss 4.0501 (3.5342) grad_norm 1.3512 (1.4402) [2021-04-16 04:06:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][490/1251] eta 0:03:40 lr 0.000570 time 0.2629 (0.2900) loss 3.8969 (3.5357) grad_norm 1.3200 (1.4415) [2021-04-16 04:06:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][500/1251] eta 0:03:37 lr 0.000570 time 0.2800 (0.2898) loss 4.1642 (3.5342) grad_norm 1.3516 (1.4426) [2021-04-16 04:06:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][510/1251] eta 0:03:34 lr 0.000570 time 0.2731 (0.2895) loss 4.3086 (3.5339) grad_norm 1.3609 (1.4424) [2021-04-16 04:06:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][520/1251] eta 0:03:31 lr 0.000570 time 0.2766 (0.2892) loss 4.1973 (3.5301) grad_norm 1.4260 (1.4423) [2021-04-16 04:07:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][530/1251] eta 0:03:28 lr 0.000570 time 0.3012 (0.2890) loss 4.0888 (3.5317) grad_norm 1.2270 (1.4411) [2021-04-16 04:07:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][540/1251] eta 0:03:25 lr 0.000570 time 0.2769 (0.2889) loss 3.5911 (3.5332) grad_norm 1.4157 (1.4396) [2021-04-16 04:07:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][550/1251] eta 0:03:22 lr 0.000570 time 0.3062 (0.2886) loss 3.5382 (3.5345) grad_norm 1.4589 (1.4382) [2021-04-16 04:07:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][560/1251] eta 0:03:19 lr 0.000570 time 0.2860 (0.2885) loss 3.2268 (3.5346) grad_norm 1.4540 (1.4397) [2021-04-16 04:07:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][570/1251] eta 0:03:16 lr 0.000570 time 0.3471 (0.2887) loss 3.5403 (3.5352) grad_norm 1.4799 (1.4404) [2021-04-16 04:07:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][580/1251] eta 0:03:13 lr 0.000570 time 0.2631 (0.2884) loss 2.4997 (3.5317) grad_norm 1.3407 (nan) [2021-04-16 04:07:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][590/1251] eta 0:03:10 lr 0.000570 time 0.2879 (0.2881) loss 3.8376 (3.5292) grad_norm 1.6859 (nan) [2021-04-16 04:07:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][600/1251] eta 0:03:07 lr 0.000570 time 0.2893 (0.2879) loss 2.8075 (3.5305) grad_norm 1.4378 (nan) [2021-04-16 04:07:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][610/1251] eta 0:03:04 lr 0.000570 time 0.3019 (0.2877) loss 4.4174 (3.5273) grad_norm 1.3649 (nan) [2021-04-16 04:07:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][620/1251] eta 0:03:01 lr 0.000570 time 0.2658 (0.2874) loss 4.1344 (3.5313) grad_norm 1.6053 (nan) [2021-04-16 04:07:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][630/1251] eta 0:02:58 lr 0.000570 time 0.2725 (0.2874) loss 4.1360 (3.5314) grad_norm 1.3787 (nan) [2021-04-16 04:07:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][640/1251] eta 0:02:55 lr 0.000570 time 0.2551 (0.2871) loss 4.4507 (3.5368) grad_norm 1.6719 (nan) [2021-04-16 04:07:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][650/1251] eta 0:02:52 lr 0.000570 time 0.2659 (0.2869) loss 4.1368 (3.5434) grad_norm 1.4043 (nan) [2021-04-16 04:07:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][660/1251] eta 0:02:49 lr 0.000569 time 0.2754 (0.2867) loss 3.6784 (3.5419) grad_norm 1.3910 (nan) [2021-04-16 04:07:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][670/1251] eta 0:02:46 lr 0.000569 time 0.2453 (0.2865) loss 2.6898 (3.5436) grad_norm 1.6369 (nan) [2021-04-16 04:07:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][680/1251] eta 0:02:43 lr 0.000569 time 0.2760 (0.2864) loss 3.8926 (3.5466) grad_norm 1.4571 (nan) [2021-04-16 04:07:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][690/1251] eta 0:02:40 lr 0.000569 time 0.2581 (0.2864) loss 3.4087 (3.5498) grad_norm 1.8417 (nan) [2021-04-16 04:07:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][700/1251] eta 0:02:37 lr 0.000569 time 0.2553 (0.2862) loss 3.6894 (3.5524) grad_norm 1.4327 (nan) [2021-04-16 04:07:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][710/1251] eta 0:02:34 lr 0.000569 time 0.2674 (0.2860) loss 3.4808 (3.5533) grad_norm 1.4253 (nan) [2021-04-16 04:07:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][720/1251] eta 0:02:31 lr 0.000569 time 0.2835 (0.2859) loss 3.8668 (3.5547) grad_norm 1.2225 (nan) [2021-04-16 04:07:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][730/1251] eta 0:02:29 lr 0.000569 time 0.2800 (0.2860) loss 4.1722 (3.5585) grad_norm 1.3617 (nan) [2021-04-16 04:07:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][740/1251] eta 0:02:26 lr 0.000569 time 0.3731 (0.2860) loss 4.2051 (3.5600) grad_norm 1.5013 (nan) [2021-04-16 04:08:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][750/1251] eta 0:02:23 lr 0.000569 time 0.2807 (0.2859) loss 3.3231 (3.5571) grad_norm 1.2341 (nan) [2021-04-16 04:08:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][760/1251] eta 0:02:20 lr 0.000569 time 0.2885 (0.2860) loss 3.6223 (3.5608) grad_norm 1.3510 (nan) [2021-04-16 04:08:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][770/1251] eta 0:02:17 lr 0.000569 time 0.2741 (0.2859) loss 4.2754 (3.5640) grad_norm 1.3985 (nan) [2021-04-16 04:08:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][780/1251] eta 0:02:14 lr 0.000569 time 0.2857 (0.2858) loss 3.4379 (3.5614) grad_norm 1.3666 (nan) [2021-04-16 04:08:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][790/1251] eta 0:02:11 lr 0.000569 time 0.2701 (0.2857) loss 3.7286 (3.5620) grad_norm 1.4867 (nan) [2021-04-16 04:08:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][800/1251] eta 0:02:08 lr 0.000569 time 0.3074 (0.2855) loss 3.6559 (3.5644) grad_norm 1.3266 (nan) [2021-04-16 04:08:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][810/1251] eta 0:02:05 lr 0.000569 time 0.2926 (0.2854) loss 4.0403 (3.5682) grad_norm 1.3026 (nan) [2021-04-16 04:08:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][820/1251] eta 0:02:02 lr 0.000569 time 0.2744 (0.2853) loss 4.2605 (3.5723) grad_norm 1.2624 (nan) [2021-04-16 04:08:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][830/1251] eta 0:02:00 lr 0.000569 time 0.2737 (0.2853) loss 3.3041 (3.5738) grad_norm 1.4367 (nan) [2021-04-16 04:08:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][840/1251] eta 0:01:57 lr 0.000569 time 0.2820 (0.2851) loss 3.8578 (3.5736) grad_norm 1.4743 (nan) [2021-04-16 04:08:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][850/1251] eta 0:01:54 lr 0.000569 time 0.2624 (0.2850) loss 3.7990 (3.5721) grad_norm 1.6291 (nan) [2021-04-16 04:08:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][860/1251] eta 0:01:51 lr 0.000569 time 0.2911 (0.2850) loss 3.2056 (3.5709) grad_norm 1.3825 (nan) [2021-04-16 04:08:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][870/1251] eta 0:01:48 lr 0.000569 time 0.2945 (0.2851) loss 4.1149 (3.5677) grad_norm 1.4202 (nan) [2021-04-16 04:08:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][880/1251] eta 0:01:45 lr 0.000569 time 0.2754 (0.2851) loss 4.2670 (3.5712) grad_norm 1.2860 (nan) [2021-04-16 04:08:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][890/1251] eta 0:01:42 lr 0.000569 time 0.2557 (0.2852) loss 3.8455 (3.5741) grad_norm 1.5710 (nan) [2021-04-16 04:08:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][900/1251] eta 0:01:40 lr 0.000568 time 0.2766 (0.2851) loss 3.6191 (3.5726) grad_norm 1.5101 (nan) [2021-04-16 04:08:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][910/1251] eta 0:01:37 lr 0.000568 time 0.2652 (0.2851) loss 4.1692 (3.5766) grad_norm 1.2832 (nan) [2021-04-16 04:08:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][920/1251] eta 0:01:34 lr 0.000568 time 0.2725 (0.2850) loss 3.9033 (3.5759) grad_norm 1.5769 (nan) [2021-04-16 04:08:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][930/1251] eta 0:01:31 lr 0.000568 time 0.2729 (0.2852) loss 3.1925 (3.5758) grad_norm 1.3743 (nan) [2021-04-16 04:08:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][940/1251] eta 0:01:28 lr 0.000568 time 0.2741 (0.2850) loss 4.3473 (3.5745) grad_norm 1.6998 (nan) [2021-04-16 04:08:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][950/1251] eta 0:01:25 lr 0.000568 time 0.2536 (0.2849) loss 3.2715 (3.5724) grad_norm 1.3693 (nan) [2021-04-16 04:09:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][960/1251] eta 0:01:22 lr 0.000568 time 0.3045 (0.2848) loss 4.1047 (3.5729) grad_norm 1.5392 (nan) [2021-04-16 04:09:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][970/1251] eta 0:01:20 lr 0.000568 time 0.2762 (0.2848) loss 3.6556 (3.5757) grad_norm 1.5181 (nan) [2021-04-16 04:09:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][980/1251] eta 0:01:17 lr 0.000568 time 0.3091 (0.2847) loss 3.9130 (3.5774) grad_norm 1.2113 (nan) [2021-04-16 04:09:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][990/1251] eta 0:01:14 lr 0.000568 time 0.3028 (0.2847) loss 3.2640 (3.5740) grad_norm 1.6472 (nan) [2021-04-16 04:09:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][1000/1251] eta 0:01:11 lr 0.000568 time 0.2601 (0.2845) loss 3.8628 (3.5746) grad_norm 1.5511 (nan) [2021-04-16 04:09:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][1010/1251] eta 0:01:08 lr 0.000568 time 0.2770 (0.2845) loss 4.0301 (3.5741) grad_norm 1.9990 (nan) [2021-04-16 04:09:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][1020/1251] eta 0:01:05 lr 0.000568 time 0.2799 (0.2844) loss 3.7384 (3.5716) grad_norm 1.7523 (nan) [2021-04-16 04:09:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][1030/1251] eta 0:01:02 lr 0.000568 time 0.2671 (0.2843) loss 3.9751 (3.5718) grad_norm 1.3371 (nan) [2021-04-16 04:09:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][1040/1251] eta 0:01:00 lr 0.000568 time 0.2875 (0.2844) loss 3.8760 (3.5726) grad_norm 1.4168 (nan) [2021-04-16 04:09:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][1050/1251] eta 0:00:57 lr 0.000568 time 0.2742 (0.2843) loss 3.6881 (3.5723) grad_norm 1.5290 (nan) [2021-04-16 04:09:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][1060/1251] eta 0:00:54 lr 0.000568 time 0.2681 (0.2842) loss 3.8070 (3.5732) grad_norm 1.4488 (nan) [2021-04-16 04:09:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][1070/1251] eta 0:00:51 lr 0.000568 time 0.2637 (0.2843) loss 2.7371 (3.5758) grad_norm 1.9429 (nan) [2021-04-16 04:09:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][1080/1251] eta 0:00:48 lr 0.000568 time 0.2931 (0.2842) loss 2.7180 (3.5705) grad_norm 1.3168 (nan) [2021-04-16 04:09:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][1090/1251] eta 0:00:45 lr 0.000568 time 0.2702 (0.2841) loss 4.0967 (3.5688) grad_norm 1.5564 (nan) [2021-04-16 04:09:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][1100/1251] eta 0:00:42 lr 0.000568 time 0.2811 (0.2840) loss 3.4437 (3.5687) grad_norm 1.8472 (nan) [2021-04-16 04:09:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][1110/1251] eta 0:00:40 lr 0.000568 time 0.2813 (0.2840) loss 2.6312 (3.5678) grad_norm 1.4620 (nan) [2021-04-16 04:09:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][1120/1251] eta 0:00:37 lr 0.000568 time 0.2526 (0.2840) loss 4.2559 (3.5704) grad_norm 1.3290 (nan) [2021-04-16 04:09:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][1130/1251] eta 0:00:34 lr 0.000568 time 0.2648 (0.2839) loss 4.0460 (3.5720) grad_norm 1.4774 (nan) [2021-04-16 04:09:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][1140/1251] eta 0:00:31 lr 0.000567 time 0.2695 (0.2838) loss 3.5811 (3.5713) grad_norm 1.4558 (nan) [2021-04-16 04:09:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][1150/1251] eta 0:00:28 lr 0.000567 time 0.2649 (0.2839) loss 3.7235 (3.5711) grad_norm 1.4419 (nan) [2021-04-16 04:09:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][1160/1251] eta 0:00:25 lr 0.000567 time 0.2685 (0.2839) loss 3.7032 (3.5719) grad_norm 1.4238 (nan) [2021-04-16 04:09:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][1170/1251] eta 0:00:22 lr 0.000567 time 0.2926 (0.2838) loss 2.7762 (3.5717) grad_norm 1.4486 (nan) [2021-04-16 04:10:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][1180/1251] eta 0:00:20 lr 0.000567 time 0.2726 (0.2837) loss 2.9350 (3.5702) grad_norm 1.2974 (nan) [2021-04-16 04:10:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][1190/1251] eta 0:00:17 lr 0.000567 time 0.2774 (0.2837) loss 4.1713 (3.5723) grad_norm 1.4201 (nan) [2021-04-16 04:10:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][1200/1251] eta 0:00:14 lr 0.000567 time 0.2755 (0.2836) loss 4.0404 (3.5730) grad_norm 1.3322 (nan) [2021-04-16 04:10:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][1210/1251] eta 0:00:11 lr 0.000567 time 0.2813 (0.2836) loss 4.2167 (3.5743) grad_norm 1.3581 (nan) [2021-04-16 04:10:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][1220/1251] eta 0:00:08 lr 0.000567 time 0.2673 (0.2835) loss 2.6199 (3.5724) grad_norm 1.5410 (nan) [2021-04-16 04:10:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][1230/1251] eta 0:00:05 lr 0.000567 time 0.2937 (0.2835) loss 3.7670 (3.5727) grad_norm 1.3766 (nan) [2021-04-16 04:10:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][1240/1251] eta 0:00:03 lr 0.000567 time 0.2485 (0.2833) loss 4.0575 (3.5721) grad_norm 1.4872 (nan) [2021-04-16 04:10:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [137/300][1250/1251] eta 0:00:00 lr 0.000567 time 0.2485 (0.2830) loss 3.5175 (3.5726) grad_norm 1.6164 (nan) [2021-04-16 04:10:25 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 137 training takes 0:05:59 [2021-04-16 04:10:25 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_137.pth saving...... [2021-04-16 04:10:38 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_137.pth saved !!! [2021-04-16 04:10:39 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.169 (1.169) Loss 1.0651 (1.0651) Acc@1 74.512 (74.512) Acc@5 92.871 (92.871) [2021-04-16 04:10:41 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.108 (0.252) Loss 1.1351 (1.0377) Acc@1 73.145 (75.710) Acc@5 91.992 (93.075) [2021-04-16 04:10:43 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.166 (0.234) Loss 1.0525 (1.0383) Acc@1 74.707 (75.832) Acc@5 92.578 (93.001) [2021-04-16 04:10:45 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.126 (0.226) Loss 1.0255 (1.0475) Acc@1 75.977 (75.520) Acc@5 93.066 (92.836) [2021-04-16 04:10:47 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.213) Loss 1.0094 (1.0432) Acc@1 75.098 (75.536) Acc@5 94.727 (93.007) [2021-04-16 04:10:53 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 75.562 Acc@5 93.046 [2021-04-16 04:10:53 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 75.6% [2021-04-16 04:10:53 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 75.62% [2021-04-16 04:10:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][0/1251] eta 1:22:50 lr 0.000567 time 3.9732 (3.9732) loss 4.1034 (4.1034) grad_norm 1.3449 (1.3449) [2021-04-16 04:10:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][10/1251] eta 0:12:40 lr 0.000567 time 0.2804 (0.6130) loss 3.1167 (3.4615) grad_norm 1.4452 (1.4370) [2021-04-16 04:11:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][20/1251] eta 0:09:19 lr 0.000567 time 0.2684 (0.4542) loss 3.4507 (3.5375) grad_norm 1.4793 (1.4498) [2021-04-16 04:11:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][30/1251] eta 0:08:08 lr 0.000567 time 0.2907 (0.4003) loss 3.5289 (3.5069) grad_norm 1.3719 (1.4359) [2021-04-16 04:11:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][40/1251] eta 0:07:29 lr 0.000567 time 0.2860 (0.3709) loss 2.8383 (3.4932) grad_norm 1.2943 (1.4472) [2021-04-16 04:11:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][50/1251] eta 0:07:04 lr 0.000567 time 0.2980 (0.3532) loss 4.1312 (3.4961) grad_norm 1.3433 (1.4435) [2021-04-16 04:11:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][60/1251] eta 0:06:46 lr 0.000567 time 0.2840 (0.3409) loss 2.3545 (3.4783) grad_norm 1.8044 (1.4443) [2021-04-16 04:11:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][70/1251] eta 0:06:30 lr 0.000567 time 0.2682 (0.3308) loss 3.4051 (3.4637) grad_norm 1.7936 (1.4551) [2021-04-16 04:11:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][80/1251] eta 0:06:20 lr 0.000567 time 0.2614 (0.3251) loss 3.6537 (3.4931) grad_norm 1.3978 (1.4530) [2021-04-16 04:11:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][90/1251] eta 0:06:13 lr 0.000567 time 0.2929 (0.3213) loss 2.7626 (3.5083) grad_norm 1.4894 (1.4635) [2021-04-16 04:11:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][100/1251] eta 0:06:04 lr 0.000567 time 0.2868 (0.3168) loss 3.6729 (3.4889) grad_norm 1.3709 (1.4587) [2021-04-16 04:11:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][110/1251] eta 0:05:57 lr 0.000567 time 0.2687 (0.3130) loss 4.2539 (3.5090) grad_norm 1.4428 (1.4597) [2021-04-16 04:11:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][120/1251] eta 0:05:51 lr 0.000567 time 0.2661 (0.3104) loss 3.8400 (3.5132) grad_norm 1.3420 (1.4531) [2021-04-16 04:11:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][130/1251] eta 0:05:45 lr 0.000567 time 0.2825 (0.3085) loss 3.8986 (3.5060) grad_norm 1.3646 (1.4501) [2021-04-16 04:11:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][140/1251] eta 0:05:41 lr 0.000566 time 0.2621 (0.3072) loss 4.1391 (3.5262) grad_norm 1.3059 (1.4518) [2021-04-16 04:11:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][150/1251] eta 0:05:36 lr 0.000566 time 0.2758 (0.3055) loss 3.4467 (3.5289) grad_norm 1.3660 (1.4565) [2021-04-16 04:11:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][160/1251] eta 0:05:31 lr 0.000566 time 0.2877 (0.3039) loss 3.4472 (3.5351) grad_norm 1.3370 (1.4567) [2021-04-16 04:11:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][170/1251] eta 0:05:26 lr 0.000566 time 0.2717 (0.3022) loss 2.9882 (3.5453) grad_norm 1.5246 (1.4541) [2021-04-16 04:11:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][180/1251] eta 0:05:22 lr 0.000566 time 0.2784 (0.3007) loss 2.7184 (3.5580) grad_norm 1.5972 (1.4588) [2021-04-16 04:11:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][190/1251] eta 0:05:18 lr 0.000566 time 0.2918 (0.2997) loss 3.3034 (3.5638) grad_norm 1.2649 (1.4591) [2021-04-16 04:11:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][200/1251] eta 0:05:13 lr 0.000566 time 0.2977 (0.2985) loss 3.7145 (3.5621) grad_norm 1.6046 (1.4564) [2021-04-16 04:11:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][210/1251] eta 0:05:10 lr 0.000566 time 0.3032 (0.2979) loss 3.3245 (3.5640) grad_norm 1.4534 (1.4534) [2021-04-16 04:11:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][220/1251] eta 0:05:06 lr 0.000566 time 0.2753 (0.2974) loss 3.7068 (3.5752) grad_norm 1.2815 (1.4516) [2021-04-16 04:12:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][230/1251] eta 0:05:02 lr 0.000566 time 0.2591 (0.2962) loss 2.4378 (3.5689) grad_norm 1.4625 (1.4491) [2021-04-16 04:12:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][240/1251] eta 0:04:58 lr 0.000566 time 0.2771 (0.2953) loss 2.7796 (3.5651) grad_norm 1.4266 (1.4470) [2021-04-16 04:12:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][250/1251] eta 0:04:55 lr 0.000566 time 0.2712 (0.2952) loss 3.4531 (3.5544) grad_norm 1.6050 (1.4470) [2021-04-16 04:12:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][260/1251] eta 0:04:51 lr 0.000566 time 0.2772 (0.2944) loss 3.8282 (3.5535) grad_norm 1.7254 (1.4505) [2021-04-16 04:12:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][270/1251] eta 0:04:48 lr 0.000566 time 0.2740 (0.2936) loss 3.9664 (3.5490) grad_norm 1.5485 (1.4518) [2021-04-16 04:12:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][280/1251] eta 0:04:44 lr 0.000566 time 0.2838 (0.2930) loss 4.3471 (3.5437) grad_norm 1.4901 (1.4522) [2021-04-16 04:12:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][290/1251] eta 0:04:41 lr 0.000566 time 0.2851 (0.2929) loss 3.9902 (3.5437) grad_norm 1.2001 (1.4490) [2021-04-16 04:12:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][300/1251] eta 0:04:38 lr 0.000566 time 0.2617 (0.2923) loss 3.9821 (3.5373) grad_norm 1.6369 (1.4519) [2021-04-16 04:12:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][310/1251] eta 0:04:34 lr 0.000566 time 0.2743 (0.2918) loss 3.7839 (3.5489) grad_norm 1.4101 (1.4543) [2021-04-16 04:12:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][320/1251] eta 0:04:31 lr 0.000566 time 0.2595 (0.2916) loss 2.8134 (3.5476) grad_norm 1.4501 (1.4554) [2021-04-16 04:12:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][330/1251] eta 0:04:28 lr 0.000566 time 0.2770 (0.2912) loss 2.3126 (3.5502) grad_norm 1.5488 (1.4560) [2021-04-16 04:12:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][340/1251] eta 0:04:24 lr 0.000566 time 0.2609 (0.2909) loss 3.9783 (3.5630) grad_norm 1.6015 (1.4576) [2021-04-16 04:12:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][350/1251] eta 0:04:21 lr 0.000566 time 0.2496 (0.2904) loss 3.9957 (3.5657) grad_norm 1.1956 (1.4591) [2021-04-16 04:12:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][360/1251] eta 0:04:18 lr 0.000566 time 0.2658 (0.2904) loss 3.4144 (3.5605) grad_norm 1.6576 (1.4588) [2021-04-16 04:12:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][370/1251] eta 0:04:15 lr 0.000566 time 0.2598 (0.2903) loss 4.1432 (3.5664) grad_norm 1.4003 (1.4578) [2021-04-16 04:12:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][380/1251] eta 0:04:12 lr 0.000565 time 0.2531 (0.2898) loss 3.7572 (3.5634) grad_norm 1.4687 (1.4575) [2021-04-16 04:12:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][390/1251] eta 0:04:09 lr 0.000565 time 0.2990 (0.2895) loss 2.9116 (3.5567) grad_norm 1.4388 (1.4569) [2021-04-16 04:12:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][400/1251] eta 0:04:06 lr 0.000565 time 0.2873 (0.2891) loss 4.2464 (3.5673) grad_norm 1.6633 (1.4563) [2021-04-16 04:12:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][410/1251] eta 0:04:02 lr 0.000565 time 0.2793 (0.2888) loss 3.7905 (3.5719) grad_norm 1.3550 (1.4549) [2021-04-16 04:12:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][420/1251] eta 0:03:59 lr 0.000565 time 0.2723 (0.2885) loss 2.6615 (3.5754) grad_norm 1.2462 (1.4536) [2021-04-16 04:12:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][430/1251] eta 0:03:56 lr 0.000565 time 0.2720 (0.2882) loss 4.0238 (3.5717) grad_norm 1.4683 (1.4552) [2021-04-16 04:13:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][440/1251] eta 0:03:53 lr 0.000565 time 0.2795 (0.2879) loss 4.0939 (3.5715) grad_norm 1.4425 (1.4548) [2021-04-16 04:13:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][450/1251] eta 0:03:50 lr 0.000565 time 0.3087 (0.2878) loss 4.1619 (3.5760) grad_norm 1.5546 (1.4566) [2021-04-16 04:13:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][460/1251] eta 0:03:47 lr 0.000565 time 0.2636 (0.2874) loss 2.7340 (3.5768) grad_norm 1.3060 (1.4557) [2021-04-16 04:13:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][470/1251] eta 0:03:44 lr 0.000565 time 0.2775 (0.2871) loss 4.4616 (3.5761) grad_norm 1.8341 (1.4560) [2021-04-16 04:13:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][480/1251] eta 0:03:41 lr 0.000565 time 0.2957 (0.2868) loss 2.4519 (3.5706) grad_norm 1.2870 (1.4559) [2021-04-16 04:13:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][490/1251] eta 0:03:38 lr 0.000565 time 0.2843 (0.2866) loss 3.5254 (3.5668) grad_norm 1.9943 (1.4553) [2021-04-16 04:13:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][500/1251] eta 0:03:34 lr 0.000565 time 0.2685 (0.2863) loss 3.9134 (3.5705) grad_norm 1.3135 (1.4562) [2021-04-16 04:13:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][510/1251] eta 0:03:32 lr 0.000565 time 0.2645 (0.2861) loss 4.3281 (3.5742) grad_norm 1.3224 (1.4564) [2021-04-16 04:13:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][520/1251] eta 0:03:29 lr 0.000565 time 0.2815 (0.2861) loss 3.2927 (3.5724) grad_norm 1.5089 (1.4555) [2021-04-16 04:13:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][530/1251] eta 0:03:26 lr 0.000565 time 0.2796 (0.2860) loss 3.9858 (3.5718) grad_norm 1.3513 (1.4539) [2021-04-16 04:13:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][540/1251] eta 0:03:23 lr 0.000565 time 0.2430 (0.2859) loss 2.9785 (3.5736) grad_norm 1.3975 (1.4535) [2021-04-16 04:13:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][550/1251] eta 0:03:20 lr 0.000565 time 0.2821 (0.2857) loss 3.2547 (3.5777) grad_norm 1.4140 (1.4528) [2021-04-16 04:13:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][560/1251] eta 0:03:17 lr 0.000565 time 0.2834 (0.2856) loss 3.3882 (3.5788) grad_norm 1.4023 (1.4539) [2021-04-16 04:13:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][570/1251] eta 0:03:14 lr 0.000565 time 0.2631 (0.2857) loss 3.3287 (3.5803) grad_norm 1.3789 (1.4543) [2021-04-16 04:13:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][580/1251] eta 0:03:11 lr 0.000565 time 0.2430 (0.2857) loss 3.7880 (3.5855) grad_norm 1.4951 (1.4551) [2021-04-16 04:13:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][590/1251] eta 0:03:08 lr 0.000565 time 0.2856 (0.2856) loss 3.8973 (3.5851) grad_norm 1.2903 (1.4545) [2021-04-16 04:13:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][600/1251] eta 0:03:05 lr 0.000565 time 0.2767 (0.2854) loss 4.2578 (3.5886) grad_norm 1.5795 (1.4548) [2021-04-16 04:13:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][610/1251] eta 0:03:02 lr 0.000565 time 0.2648 (0.2852) loss 3.2435 (3.5843) grad_norm 1.7100 (1.4555) [2021-04-16 04:13:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][620/1251] eta 0:02:59 lr 0.000564 time 0.2945 (0.2851) loss 4.2703 (3.5873) grad_norm 1.4200 (1.4563) [2021-04-16 04:13:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][630/1251] eta 0:02:56 lr 0.000564 time 0.2623 (0.2848) loss 2.8549 (3.5853) grad_norm 1.6146 (1.4573) [2021-04-16 04:13:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][640/1251] eta 0:02:53 lr 0.000564 time 0.2721 (0.2847) loss 4.6421 (3.5876) grad_norm 1.3916 (1.4579) [2021-04-16 04:13:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][650/1251] eta 0:02:51 lr 0.000564 time 0.2892 (0.2848) loss 2.5442 (3.5914) grad_norm 1.5803 (1.4585) [2021-04-16 04:14:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][660/1251] eta 0:02:48 lr 0.000564 time 0.2943 (0.2847) loss 4.3391 (3.5908) grad_norm 1.2520 (1.4590) [2021-04-16 04:14:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][670/1251] eta 0:02:45 lr 0.000564 time 0.2783 (0.2846) loss 4.0608 (3.5883) grad_norm 1.3222 (1.4582) [2021-04-16 04:14:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][680/1251] eta 0:02:42 lr 0.000564 time 0.2900 (0.2846) loss 3.5913 (3.5897) grad_norm 1.6544 (1.4577) [2021-04-16 04:14:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][690/1251] eta 0:02:39 lr 0.000564 time 0.2870 (0.2846) loss 4.2081 (3.5875) grad_norm 1.3775 (1.4575) [2021-04-16 04:14:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][700/1251] eta 0:02:36 lr 0.000564 time 0.3243 (0.2846) loss 3.9922 (3.5884) grad_norm 1.4748 (1.4563) [2021-04-16 04:14:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][710/1251] eta 0:02:33 lr 0.000564 time 0.2892 (0.2844) loss 4.2948 (3.5871) grad_norm 1.5779 (1.4558) [2021-04-16 04:14:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][720/1251] eta 0:02:31 lr 0.000564 time 0.2846 (0.2846) loss 3.7606 (3.5854) grad_norm 1.7510 (1.4559) [2021-04-16 04:14:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][730/1251] eta 0:02:28 lr 0.000564 time 0.2680 (0.2844) loss 3.7975 (3.5873) grad_norm 1.2444 (1.4567) [2021-04-16 04:14:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][740/1251] eta 0:02:25 lr 0.000564 time 0.2970 (0.2843) loss 4.0129 (3.5842) grad_norm 1.4183 (1.4566) [2021-04-16 04:14:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][750/1251] eta 0:02:22 lr 0.000564 time 0.2457 (0.2842) loss 4.4146 (3.5871) grad_norm 1.3888 (1.4565) [2021-04-16 04:14:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][760/1251] eta 0:02:19 lr 0.000564 time 0.2434 (0.2840) loss 3.0111 (3.5873) grad_norm 1.3060 (1.4564) [2021-04-16 04:14:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][770/1251] eta 0:02:16 lr 0.000564 time 0.2680 (0.2840) loss 3.4604 (3.5869) grad_norm 1.3924 (1.4560) [2021-04-16 04:14:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][780/1251] eta 0:02:13 lr 0.000564 time 0.2795 (0.2839) loss 3.2785 (3.5891) grad_norm 1.3300 (1.4564) [2021-04-16 04:14:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][790/1251] eta 0:02:10 lr 0.000564 time 0.2557 (0.2839) loss 3.9947 (3.5875) grad_norm 1.2565 (1.4565) [2021-04-16 04:14:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][800/1251] eta 0:02:07 lr 0.000564 time 0.2930 (0.2838) loss 3.6932 (3.5891) grad_norm 1.5449 (1.4566) [2021-04-16 04:14:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][810/1251] eta 0:02:05 lr 0.000564 time 0.2957 (0.2837) loss 2.7749 (3.5872) grad_norm 1.4671 (1.4571) [2021-04-16 04:14:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][820/1251] eta 0:02:02 lr 0.000564 time 0.2933 (0.2836) loss 4.0335 (3.5882) grad_norm 1.2706 (1.4573) [2021-04-16 04:14:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][830/1251] eta 0:01:59 lr 0.000564 time 0.2675 (0.2837) loss 3.3623 (3.5875) grad_norm 1.4290 (1.4572) [2021-04-16 04:14:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][840/1251] eta 0:01:56 lr 0.000564 time 0.2809 (0.2836) loss 4.1463 (3.5882) grad_norm 1.5085 (1.4579) [2021-04-16 04:14:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][850/1251] eta 0:01:53 lr 0.000564 time 0.2587 (0.2835) loss 3.7534 (3.5884) grad_norm 1.7240 (1.4577) [2021-04-16 04:14:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][860/1251] eta 0:01:50 lr 0.000564 time 0.2849 (0.2835) loss 2.6735 (3.5874) grad_norm 1.5165 (1.4578) [2021-04-16 04:15:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][870/1251] eta 0:01:48 lr 0.000563 time 0.3070 (0.2835) loss 3.7079 (3.5885) grad_norm 1.4870 (1.4591) [2021-04-16 04:15:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][880/1251] eta 0:01:45 lr 0.000563 time 0.2702 (0.2834) loss 3.8166 (3.5870) grad_norm 1.3958 (1.4597) [2021-04-16 04:15:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][890/1251] eta 0:01:42 lr 0.000563 time 0.2645 (0.2835) loss 4.3160 (3.5867) grad_norm 1.4249 (1.4619) [2021-04-16 04:15:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][900/1251] eta 0:01:39 lr 0.000563 time 0.2705 (0.2834) loss 3.7164 (3.5854) grad_norm 1.3776 (1.4618) [2021-04-16 04:15:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][910/1251] eta 0:01:36 lr 0.000563 time 0.2863 (0.2833) loss 3.6040 (3.5846) grad_norm 1.5977 (1.4631) [2021-04-16 04:15:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][920/1251] eta 0:01:33 lr 0.000563 time 0.2856 (0.2832) loss 4.3045 (3.5824) grad_norm 1.5193 (1.4629) [2021-04-16 04:15:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][930/1251] eta 0:01:30 lr 0.000563 time 0.2880 (0.2831) loss 3.9596 (3.5808) grad_norm 1.5613 (1.4635) [2021-04-16 04:15:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][940/1251] eta 0:01:28 lr 0.000563 time 0.2912 (0.2832) loss 3.8326 (3.5815) grad_norm 1.2987 (1.4645) [2021-04-16 04:15:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][950/1251] eta 0:01:25 lr 0.000563 time 0.2683 (0.2833) loss 3.4816 (3.5831) grad_norm 1.3193 (1.4640) [2021-04-16 04:15:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][960/1251] eta 0:01:22 lr 0.000563 time 0.2977 (0.2832) loss 4.1715 (3.5851) grad_norm 1.5009 (1.4636) [2021-04-16 04:15:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][970/1251] eta 0:01:19 lr 0.000563 time 0.2671 (0.2831) loss 3.6499 (3.5865) grad_norm 1.3742 (1.4627) [2021-04-16 04:15:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][980/1251] eta 0:01:16 lr 0.000563 time 0.2713 (0.2831) loss 4.2790 (3.5866) grad_norm 1.5063 (1.4627) [2021-04-16 04:15:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][990/1251] eta 0:01:13 lr 0.000563 time 0.2628 (0.2830) loss 4.0453 (3.5854) grad_norm 1.4153 (1.4623) [2021-04-16 04:15:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][1000/1251] eta 0:01:11 lr 0.000563 time 0.2801 (0.2830) loss 3.7195 (3.5850) grad_norm 1.6807 (1.4627) [2021-04-16 04:15:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][1010/1251] eta 0:01:08 lr 0.000563 time 0.2824 (0.2831) loss 2.6673 (3.5837) grad_norm 1.6625 (1.4628) [2021-04-16 04:15:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][1020/1251] eta 0:01:05 lr 0.000563 time 0.2557 (0.2830) loss 3.8887 (3.5839) grad_norm 1.2509 (1.4623) [2021-04-16 04:15:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][1030/1251] eta 0:01:02 lr 0.000563 time 0.2770 (0.2829) loss 3.9672 (3.5839) grad_norm 1.2194 (1.4616) [2021-04-16 04:15:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][1040/1251] eta 0:00:59 lr 0.000563 time 0.2662 (0.2829) loss 3.7758 (3.5844) grad_norm 1.4536 (1.4615) [2021-04-16 04:15:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][1050/1251] eta 0:00:56 lr 0.000563 time 0.2673 (0.2828) loss 3.8274 (3.5834) grad_norm 1.2824 (1.4605) [2021-04-16 04:15:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][1060/1251] eta 0:00:54 lr 0.000563 time 0.2627 (0.2827) loss 3.9698 (3.5852) grad_norm 1.5242 (1.4600) [2021-04-16 04:15:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][1070/1251] eta 0:00:51 lr 0.000563 time 0.2895 (0.2827) loss 2.7388 (3.5845) grad_norm 1.3663 (1.4595) [2021-04-16 04:15:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][1080/1251] eta 0:00:48 lr 0.000563 time 0.2779 (0.2827) loss 2.8512 (3.5857) grad_norm 1.4441 (1.4588) [2021-04-16 04:16:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][1090/1251] eta 0:00:45 lr 0.000563 time 0.2443 (0.2827) loss 3.9081 (3.5852) grad_norm 1.4188 (1.4588) [2021-04-16 04:16:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][1100/1251] eta 0:00:42 lr 0.000563 time 0.2856 (0.2826) loss 3.7679 (3.5857) grad_norm 1.5802 (1.4584) [2021-04-16 04:16:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][1110/1251] eta 0:00:39 lr 0.000562 time 0.2703 (0.2826) loss 3.8260 (3.5828) grad_norm 1.2976 (1.4585) [2021-04-16 04:16:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][1120/1251] eta 0:00:37 lr 0.000562 time 0.3000 (0.2826) loss 3.6894 (3.5812) grad_norm 1.2335 (1.4585) [2021-04-16 04:16:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][1130/1251] eta 0:00:34 lr 0.000562 time 0.3074 (0.2825) loss 3.9287 (3.5812) grad_norm 1.2535 (1.4586) [2021-04-16 04:16:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][1140/1251] eta 0:00:31 lr 0.000562 time 0.2876 (0.2824) loss 4.1461 (3.5817) grad_norm 1.5938 (1.4589) [2021-04-16 04:16:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][1150/1251] eta 0:00:28 lr 0.000562 time 0.2894 (0.2826) loss 4.0118 (3.5821) grad_norm 1.3480 (1.4593) [2021-04-16 04:16:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][1160/1251] eta 0:00:25 lr 0.000562 time 0.2657 (0.2827) loss 3.8920 (3.5841) grad_norm 1.3359 (1.4591) [2021-04-16 04:16:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][1170/1251] eta 0:00:22 lr 0.000562 time 0.2817 (0.2826) loss 3.8857 (3.5840) grad_norm 1.5596 (1.4592) [2021-04-16 04:16:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][1180/1251] eta 0:00:20 lr 0.000562 time 0.2702 (0.2826) loss 2.8212 (3.5850) grad_norm 1.5322 (1.4596) [2021-04-16 04:16:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][1190/1251] eta 0:00:17 lr 0.000562 time 0.2681 (0.2825) loss 4.1622 (3.5854) grad_norm 1.2772 (1.4596) [2021-04-16 04:16:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][1200/1251] eta 0:00:14 lr 0.000562 time 0.2747 (0.2825) loss 4.0025 (3.5849) grad_norm 1.3177 (1.4593) [2021-04-16 04:16:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][1210/1251] eta 0:00:11 lr 0.000562 time 0.3047 (0.2825) loss 2.9616 (3.5826) grad_norm 1.2232 (1.4602) [2021-04-16 04:16:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][1220/1251] eta 0:00:08 lr 0.000562 time 0.2719 (0.2824) loss 2.2857 (3.5828) grad_norm 1.4833 (1.4612) [2021-04-16 04:16:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][1230/1251] eta 0:00:05 lr 0.000562 time 0.2978 (0.2825) loss 2.7421 (3.5814) grad_norm 1.4819 (1.4618) [2021-04-16 04:16:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][1240/1251] eta 0:00:03 lr 0.000562 time 0.2478 (0.2824) loss 3.8204 (3.5812) grad_norm 1.2862 (1.4626) [2021-04-16 04:16:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [138/300][1250/1251] eta 0:00:00 lr 0.000562 time 0.2479 (0.2821) loss 4.0846 (3.5826) grad_norm 1.2960 (1.4627) [2021-04-16 04:16:51 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 138 training takes 0:05:57 [2021-04-16 04:16:51 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_138.pth saving...... [2021-04-16 04:17:03 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_138.pth saved !!! [2021-04-16 04:17:04 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.283 (1.283) Loss 1.0102 (1.0102) Acc@1 76.855 (76.855) Acc@5 93.555 (93.555) [2021-04-16 04:17:06 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.138 (0.231) Loss 1.1055 (1.0455) Acc@1 75.781 (75.568) Acc@5 91.895 (92.853) [2021-04-16 04:17:08 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.137 (0.225) Loss 1.0461 (1.0460) Acc@1 73.535 (75.474) Acc@5 93.164 (92.908) [2021-04-16 04:17:10 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.123 (0.235) Loss 0.9997 (1.0441) Acc@1 76.758 (75.529) Acc@5 94.531 (92.975) [2021-04-16 04:17:12 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.220) Loss 1.0291 (1.0380) Acc@1 76.562 (75.693) Acc@5 93.262 (93.064) [2021-04-16 04:17:18 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 75.592 Acc@5 93.008 [2021-04-16 04:17:18 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 75.6% [2021-04-16 04:17:18 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 75.62% [2021-04-16 04:17:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][0/1251] eta 1:02:26 lr 0.000562 time 2.9945 (2.9945) loss 3.8255 (3.8255) grad_norm 1.3958 (1.3958) [2021-04-16 04:17:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][10/1251] eta 0:10:47 lr 0.000562 time 0.2685 (0.5219) loss 2.8253 (3.4545) grad_norm 1.4316 (1.3497) [2021-04-16 04:17:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][20/1251] eta 0:08:22 lr 0.000562 time 0.2642 (0.4080) loss 3.4730 (3.4740) grad_norm 1.2538 (1.3447) [2021-04-16 04:17:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][30/1251] eta 0:07:26 lr 0.000562 time 0.2477 (0.3657) loss 2.6613 (3.4652) grad_norm 1.3007 (1.3574) [2021-04-16 04:17:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][40/1251] eta 0:07:02 lr 0.000562 time 0.2866 (0.3490) loss 3.8672 (3.5002) grad_norm 1.2655 (1.3974) [2021-04-16 04:17:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][50/1251] eta 0:06:42 lr 0.000562 time 0.2833 (0.3347) loss 3.9091 (3.5049) grad_norm 1.7352 (1.4058) [2021-04-16 04:17:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][60/1251] eta 0:06:29 lr 0.000562 time 0.4110 (0.3267) loss 3.8913 (3.5462) grad_norm 1.2470 (1.3993) [2021-04-16 04:17:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][70/1251] eta 0:06:16 lr 0.000562 time 0.2563 (0.3191) loss 2.9669 (3.5540) grad_norm 1.5944 (1.4040) [2021-04-16 04:17:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][80/1251] eta 0:06:10 lr 0.000562 time 0.3109 (0.3167) loss 3.1752 (3.5476) grad_norm 1.5985 (1.4105) [2021-04-16 04:17:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][90/1251] eta 0:06:05 lr 0.000562 time 0.2786 (0.3152) loss 3.9926 (3.5011) grad_norm 1.5129 (1.4187) [2021-04-16 04:17:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][100/1251] eta 0:05:58 lr 0.000561 time 0.2981 (0.3117) loss 4.1747 (3.5222) grad_norm 1.4331 (1.4165) [2021-04-16 04:17:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][110/1251] eta 0:05:52 lr 0.000561 time 0.2941 (0.3088) loss 3.9879 (3.5325) grad_norm 1.3216 (1.4134) [2021-04-16 04:17:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][120/1251] eta 0:05:46 lr 0.000561 time 0.2792 (0.3065) loss 2.9454 (3.5463) grad_norm 1.3043 (1.4161) [2021-04-16 04:17:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][130/1251] eta 0:05:40 lr 0.000561 time 0.2755 (0.3041) loss 4.0551 (3.5629) grad_norm 1.4475 (1.4213) [2021-04-16 04:18:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][140/1251] eta 0:05:35 lr 0.000561 time 0.2440 (0.3022) loss 4.1018 (3.5365) grad_norm 1.3916 (1.4193) [2021-04-16 04:18:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][150/1251] eta 0:05:32 lr 0.000561 time 0.2773 (0.3022) loss 3.7722 (3.5367) grad_norm 1.3808 (1.4217) [2021-04-16 04:18:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][160/1251] eta 0:05:27 lr 0.000561 time 0.2990 (0.3006) loss 3.7978 (3.5344) grad_norm 1.3314 (1.4222) [2021-04-16 04:18:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][170/1251] eta 0:05:23 lr 0.000561 time 0.2677 (0.2992) loss 3.9046 (3.5281) grad_norm 1.5991 (1.4231) [2021-04-16 04:18:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][180/1251] eta 0:05:19 lr 0.000561 time 0.2723 (0.2981) loss 2.7412 (3.5351) grad_norm 1.4192 (1.4237) [2021-04-16 04:18:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][190/1251] eta 0:05:15 lr 0.000561 time 0.2913 (0.2970) loss 2.8085 (3.5372) grad_norm 1.2237 (1.4219) [2021-04-16 04:18:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][200/1251] eta 0:05:11 lr 0.000561 time 0.2594 (0.2962) loss 4.3976 (3.5387) grad_norm 1.3524 (1.4235) [2021-04-16 04:18:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][210/1251] eta 0:05:07 lr 0.000561 time 0.2802 (0.2955) loss 2.5794 (3.5411) grad_norm 1.4470 (1.4224) [2021-04-16 04:18:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][220/1251] eta 0:05:04 lr 0.000561 time 0.2693 (0.2951) loss 3.0192 (3.5508) grad_norm 1.2327 (1.4258) [2021-04-16 04:18:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][230/1251] eta 0:05:00 lr 0.000561 time 0.2565 (0.2943) loss 2.3878 (3.5533) grad_norm 1.4736 (1.4267) [2021-04-16 04:18:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][240/1251] eta 0:04:56 lr 0.000561 time 0.2682 (0.2936) loss 3.4831 (3.5660) grad_norm 1.4382 (1.4270) [2021-04-16 04:18:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][250/1251] eta 0:04:53 lr 0.000561 time 0.2693 (0.2932) loss 3.9413 (3.5662) grad_norm 1.9809 (1.4306) [2021-04-16 04:18:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][260/1251] eta 0:04:49 lr 0.000561 time 0.2653 (0.2925) loss 2.7915 (3.5633) grad_norm 1.5721 (1.4313) [2021-04-16 04:18:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][270/1251] eta 0:04:46 lr 0.000561 time 0.2647 (0.2924) loss 3.0028 (3.5679) grad_norm 1.3938 (1.4312) [2021-04-16 04:18:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][280/1251] eta 0:04:43 lr 0.000561 time 0.2681 (0.2920) loss 3.4245 (3.5642) grad_norm 1.4168 (1.4309) [2021-04-16 04:18:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][290/1251] eta 0:04:40 lr 0.000561 time 0.2787 (0.2916) loss 2.5505 (3.5660) grad_norm 1.4377 (1.4290) [2021-04-16 04:18:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][300/1251] eta 0:04:37 lr 0.000561 time 0.2862 (0.2917) loss 3.0432 (3.5676) grad_norm 1.6646 (1.4312) [2021-04-16 04:18:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][310/1251] eta 0:04:34 lr 0.000561 time 0.2824 (0.2912) loss 3.6246 (3.5684) grad_norm 1.6070 (1.4303) [2021-04-16 04:18:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][320/1251] eta 0:04:30 lr 0.000561 time 0.2757 (0.2908) loss 3.4612 (3.5684) grad_norm 1.3508 (1.4342) [2021-04-16 04:18:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][330/1251] eta 0:04:27 lr 0.000561 time 0.2791 (0.2905) loss 3.7903 (3.5686) grad_norm 1.7785 (1.4359) [2021-04-16 04:18:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][340/1251] eta 0:04:24 lr 0.000560 time 0.2723 (0.2902) loss 4.0528 (3.5723) grad_norm 1.2110 (1.4365) [2021-04-16 04:19:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][350/1251] eta 0:04:21 lr 0.000560 time 0.4589 (0.2903) loss 2.7108 (3.5791) grad_norm 1.3455 (1.4374) [2021-04-16 04:19:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][360/1251] eta 0:04:18 lr 0.000560 time 0.2623 (0.2903) loss 3.5439 (3.5775) grad_norm 1.6485 (1.4354) [2021-04-16 04:19:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][370/1251] eta 0:04:15 lr 0.000560 time 0.2424 (0.2900) loss 3.8271 (3.5800) grad_norm 1.2860 (1.4349) [2021-04-16 04:19:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][380/1251] eta 0:04:12 lr 0.000560 time 0.2814 (0.2897) loss 4.4023 (3.5823) grad_norm 1.3710 (1.4371) [2021-04-16 04:19:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][390/1251] eta 0:04:09 lr 0.000560 time 0.2690 (0.2894) loss 3.8361 (3.5830) grad_norm 1.3502 (1.4380) [2021-04-16 04:19:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][400/1251] eta 0:04:05 lr 0.000560 time 0.2831 (0.2890) loss 3.4048 (3.5770) grad_norm 1.3435 (1.4415) [2021-04-16 04:19:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][410/1251] eta 0:04:02 lr 0.000560 time 0.2856 (0.2888) loss 3.9815 (3.5771) grad_norm 1.3795 (1.4417) [2021-04-16 04:19:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][420/1251] eta 0:03:59 lr 0.000560 time 0.2749 (0.2887) loss 3.2862 (3.5789) grad_norm 1.4587 (1.4404) [2021-04-16 04:19:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][430/1251] eta 0:03:56 lr 0.000560 time 0.2567 (0.2886) loss 3.7723 (3.5847) grad_norm 1.4827 (1.4394) [2021-04-16 04:19:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][440/1251] eta 0:03:54 lr 0.000560 time 0.2799 (0.2888) loss 3.6235 (3.5893) grad_norm 1.3259 (1.4381) [2021-04-16 04:19:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][450/1251] eta 0:03:51 lr 0.000560 time 0.2834 (0.2885) loss 2.4317 (3.5870) grad_norm 1.2969 (1.4378) [2021-04-16 04:19:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][460/1251] eta 0:03:47 lr 0.000560 time 0.2603 (0.2882) loss 3.1362 (3.5880) grad_norm 1.3799 (1.4410) [2021-04-16 04:19:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][470/1251] eta 0:03:45 lr 0.000560 time 0.2898 (0.2881) loss 3.6444 (3.5897) grad_norm 1.6668 (1.4420) [2021-04-16 04:19:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][480/1251] eta 0:03:41 lr 0.000560 time 0.2795 (0.2878) loss 3.1185 (3.5879) grad_norm 1.3054 (1.4411) [2021-04-16 04:19:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][490/1251] eta 0:03:38 lr 0.000560 time 0.3091 (0.2876) loss 3.4912 (3.5894) grad_norm 1.6441 (1.4420) [2021-04-16 04:19:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][500/1251] eta 0:03:35 lr 0.000560 time 0.3011 (0.2874) loss 3.5231 (3.5896) grad_norm 1.4004 (1.4454) [2021-04-16 04:19:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][510/1251] eta 0:03:32 lr 0.000560 time 0.2953 (0.2872) loss 3.1146 (3.5876) grad_norm 1.5898 (1.4453) [2021-04-16 04:19:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][520/1251] eta 0:03:29 lr 0.000560 time 0.2836 (0.2870) loss 4.1885 (3.5865) grad_norm 1.4946 (1.4459) [2021-04-16 04:19:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][530/1251] eta 0:03:26 lr 0.000560 time 0.2620 (0.2868) loss 3.7312 (3.5832) grad_norm 1.5058 (1.4466) [2021-04-16 04:19:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][540/1251] eta 0:03:24 lr 0.000560 time 0.3010 (0.2870) loss 3.3236 (3.5856) grad_norm 1.2498 (1.4472) [2021-04-16 04:19:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][550/1251] eta 0:03:21 lr 0.000560 time 0.2583 (0.2869) loss 3.0463 (3.5797) grad_norm 1.4092 (1.4473) [2021-04-16 04:19:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][560/1251] eta 0:03:18 lr 0.000560 time 0.2537 (0.2867) loss 2.9863 (3.5823) grad_norm 1.4902 (1.4496) [2021-04-16 04:20:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][570/1251] eta 0:03:15 lr 0.000560 time 0.4081 (0.2869) loss 3.9544 (3.5859) grad_norm 1.2706 (1.4492) [2021-04-16 04:20:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][580/1251] eta 0:03:12 lr 0.000560 time 0.2783 (0.2868) loss 3.3886 (3.5829) grad_norm 1.8799 (1.4507) [2021-04-16 04:20:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][590/1251] eta 0:03:09 lr 0.000559 time 0.2721 (0.2866) loss 4.1165 (3.5816) grad_norm 1.3773 (1.4509) [2021-04-16 04:20:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][600/1251] eta 0:03:06 lr 0.000559 time 0.2652 (0.2864) loss 3.6104 (3.5845) grad_norm 1.4476 (1.4521) [2021-04-16 04:20:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][610/1251] eta 0:03:03 lr 0.000559 time 0.2461 (0.2863) loss 3.2193 (3.5844) grad_norm 1.5896 (1.4536) [2021-04-16 04:20:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][620/1251] eta 0:03:00 lr 0.000559 time 0.2926 (0.2863) loss 3.6800 (3.5848) grad_norm 1.3220 (1.4538) [2021-04-16 04:20:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][630/1251] eta 0:02:57 lr 0.000559 time 0.2901 (0.2861) loss 3.9234 (3.5858) grad_norm 1.3436 (1.4542) [2021-04-16 04:20:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][640/1251] eta 0:02:54 lr 0.000559 time 0.2971 (0.2862) loss 3.9522 (3.5849) grad_norm 1.4763 (1.4543) [2021-04-16 04:20:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][650/1251] eta 0:02:52 lr 0.000559 time 0.3042 (0.2863) loss 3.5446 (3.5856) grad_norm 1.4344 (1.4547) [2021-04-16 04:20:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][660/1251] eta 0:02:49 lr 0.000559 time 0.2710 (0.2861) loss 3.0399 (3.5856) grad_norm 1.2744 (1.4548) [2021-04-16 04:20:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][670/1251] eta 0:02:46 lr 0.000559 time 0.2721 (0.2861) loss 2.6309 (3.5841) grad_norm 1.6216 (1.4544) [2021-04-16 04:20:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][680/1251] eta 0:02:43 lr 0.000559 time 0.3013 (0.2860) loss 3.8036 (3.5855) grad_norm 1.4662 (1.4537) [2021-04-16 04:20:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][690/1251] eta 0:02:40 lr 0.000559 time 0.2733 (0.2859) loss 2.9412 (3.5826) grad_norm 1.3949 (1.4537) [2021-04-16 04:20:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][700/1251] eta 0:02:37 lr 0.000559 time 0.2940 (0.2859) loss 4.0845 (3.5805) grad_norm 1.3337 (1.4531) [2021-04-16 04:20:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][710/1251] eta 0:02:34 lr 0.000559 time 0.2787 (0.2858) loss 3.7021 (3.5848) grad_norm 1.3472 (1.4522) [2021-04-16 04:20:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][720/1251] eta 0:02:31 lr 0.000559 time 0.2711 (0.2859) loss 3.9278 (3.5829) grad_norm 1.7083 (1.4528) [2021-04-16 04:20:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][730/1251] eta 0:02:28 lr 0.000559 time 0.2747 (0.2859) loss 2.9770 (3.5809) grad_norm 1.2787 (1.4544) [2021-04-16 04:20:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][740/1251] eta 0:02:26 lr 0.000559 time 0.2554 (0.2858) loss 3.7472 (3.5837) grad_norm 1.6369 (inf) [2021-04-16 04:20:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][750/1251] eta 0:02:23 lr 0.000559 time 0.2819 (0.2857) loss 2.7940 (3.5868) grad_norm 1.2293 (inf) [2021-04-16 04:20:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][760/1251] eta 0:02:20 lr 0.000559 time 0.2740 (0.2857) loss 3.1330 (3.5886) grad_norm 1.5503 (inf) [2021-04-16 04:20:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][770/1251] eta 0:02:17 lr 0.000559 time 0.2572 (0.2856) loss 4.1273 (3.5860) grad_norm 1.5452 (inf) [2021-04-16 04:21:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][780/1251] eta 0:02:14 lr 0.000559 time 0.2680 (0.2855) loss 4.4570 (3.5848) grad_norm 1.3414 (inf) [2021-04-16 04:21:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][790/1251] eta 0:02:11 lr 0.000559 time 0.2741 (0.2855) loss 4.2701 (3.5824) grad_norm 1.7292 (inf) [2021-04-16 04:21:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][800/1251] eta 0:02:08 lr 0.000559 time 0.2772 (0.2853) loss 4.3518 (3.5838) grad_norm 1.4933 (inf) [2021-04-16 04:21:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][810/1251] eta 0:02:05 lr 0.000559 time 0.2722 (0.2853) loss 3.8375 (3.5867) grad_norm 1.4361 (inf) [2021-04-16 04:21:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][820/1251] eta 0:02:02 lr 0.000559 time 0.2513 (0.2852) loss 2.9844 (3.5867) grad_norm 1.3283 (inf) [2021-04-16 04:21:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][830/1251] eta 0:02:00 lr 0.000558 time 0.2981 (0.2851) loss 3.2834 (3.5862) grad_norm 1.3313 (inf) [2021-04-16 04:21:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][840/1251] eta 0:01:57 lr 0.000558 time 0.2754 (0.2850) loss 4.3256 (3.5894) grad_norm 1.5836 (inf) [2021-04-16 04:21:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][850/1251] eta 0:01:54 lr 0.000558 time 0.2972 (0.2850) loss 2.5124 (3.5845) grad_norm 1.8153 (inf) [2021-04-16 04:21:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][860/1251] eta 0:01:51 lr 0.000558 time 0.2707 (0.2848) loss 3.3467 (3.5829) grad_norm 1.4709 (inf) [2021-04-16 04:21:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][870/1251] eta 0:01:48 lr 0.000558 time 0.2889 (0.2848) loss 2.5627 (3.5805) grad_norm 1.2538 (inf) [2021-04-16 04:21:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][880/1251] eta 0:01:45 lr 0.000558 time 0.2503 (0.2849) loss 3.9941 (3.5837) grad_norm 1.2338 (inf) [2021-04-16 04:21:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][890/1251] eta 0:01:42 lr 0.000558 time 0.2807 (0.2848) loss 3.7184 (3.5846) grad_norm 1.5174 (inf) [2021-04-16 04:21:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][900/1251] eta 0:01:39 lr 0.000558 time 0.2695 (0.2847) loss 4.0788 (3.5851) grad_norm 1.4345 (inf) [2021-04-16 04:21:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][910/1251] eta 0:01:37 lr 0.000558 time 0.2641 (0.2846) loss 3.5022 (3.5876) grad_norm 1.1954 (inf) [2021-04-16 04:21:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][920/1251] eta 0:01:34 lr 0.000558 time 0.3026 (0.2846) loss 2.9341 (3.5872) grad_norm 1.4925 (inf) [2021-04-16 04:21:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][930/1251] eta 0:01:31 lr 0.000558 time 0.2551 (0.2847) loss 3.9484 (3.5864) grad_norm 1.3798 (inf) [2021-04-16 04:21:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][940/1251] eta 0:01:28 lr 0.000558 time 0.2767 (0.2849) loss 3.7198 (3.5890) grad_norm 1.4345 (inf) [2021-04-16 04:21:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][950/1251] eta 0:01:25 lr 0.000558 time 0.2735 (0.2848) loss 4.3055 (3.5904) grad_norm 1.4435 (inf) [2021-04-16 04:21:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][960/1251] eta 0:01:22 lr 0.000558 time 0.2450 (0.2847) loss 2.5454 (3.5903) grad_norm 1.4151 (inf) [2021-04-16 04:21:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][970/1251] eta 0:01:20 lr 0.000558 time 0.2961 (0.2848) loss 4.1795 (3.5900) grad_norm 1.4236 (inf) [2021-04-16 04:21:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][980/1251] eta 0:01:17 lr 0.000558 time 0.2877 (0.2847) loss 3.1115 (3.5893) grad_norm 1.5328 (inf) [2021-04-16 04:22:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][990/1251] eta 0:01:14 lr 0.000558 time 0.3034 (0.2847) loss 3.3055 (3.5901) grad_norm 1.2566 (inf) [2021-04-16 04:22:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][1000/1251] eta 0:01:11 lr 0.000558 time 0.2536 (0.2846) loss 3.9606 (3.5907) grad_norm 1.4233 (inf) [2021-04-16 04:22:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][1010/1251] eta 0:01:08 lr 0.000558 time 0.2665 (0.2846) loss 3.9045 (3.5902) grad_norm 1.3683 (inf) [2021-04-16 04:22:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][1020/1251] eta 0:01:05 lr 0.000558 time 0.2751 (0.2848) loss 4.0921 (3.5901) grad_norm 1.4877 (inf) [2021-04-16 04:22:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][1030/1251] eta 0:01:02 lr 0.000558 time 0.2908 (0.2847) loss 2.3531 (3.5886) grad_norm 1.4313 (inf) [2021-04-16 04:22:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][1040/1251] eta 0:01:00 lr 0.000558 time 0.3211 (0.2846) loss 3.3508 (3.5884) grad_norm 1.2941 (inf) [2021-04-16 04:22:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][1050/1251] eta 0:00:57 lr 0.000558 time 0.3012 (0.2846) loss 3.6979 (3.5868) grad_norm 1.3491 (inf) [2021-04-16 04:22:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][1060/1251] eta 0:00:54 lr 0.000558 time 0.2884 (0.2845) loss 4.0123 (3.5901) grad_norm 1.4296 (inf) [2021-04-16 04:22:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][1070/1251] eta 0:00:51 lr 0.000557 time 0.2680 (0.2845) loss 3.9037 (3.5901) grad_norm 1.5363 (inf) [2021-04-16 04:22:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][1080/1251] eta 0:00:48 lr 0.000557 time 0.2764 (0.2844) loss 3.8051 (3.5930) grad_norm 1.5491 (inf) [2021-04-16 04:22:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][1090/1251] eta 0:00:45 lr 0.000557 time 0.2916 (0.2843) loss 3.6665 (3.5927) grad_norm 1.8190 (inf) [2021-04-16 04:22:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][1100/1251] eta 0:00:42 lr 0.000557 time 0.2896 (0.2843) loss 4.0060 (3.5927) grad_norm 1.3028 (inf) [2021-04-16 04:22:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][1110/1251] eta 0:00:40 lr 0.000557 time 0.2954 (0.2843) loss 3.7944 (3.5915) grad_norm 1.5344 (inf) [2021-04-16 04:22:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][1120/1251] eta 0:00:37 lr 0.000557 time 0.3111 (0.2842) loss 3.8196 (3.5918) grad_norm 1.4457 (inf) [2021-04-16 04:22:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][1130/1251] eta 0:00:34 lr 0.000557 time 0.2781 (0.2842) loss 3.2014 (3.5907) grad_norm 1.3653 (inf) [2021-04-16 04:22:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][1140/1251] eta 0:00:31 lr 0.000557 time 0.2689 (0.2842) loss 4.1353 (3.5912) grad_norm 1.5278 (inf) [2021-04-16 04:22:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][1150/1251] eta 0:00:28 lr 0.000557 time 0.2647 (0.2842) loss 3.3515 (3.5924) grad_norm 1.6417 (inf) [2021-04-16 04:22:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][1160/1251] eta 0:00:25 lr 0.000557 time 0.2518 (0.2841) loss 4.0337 (3.5925) grad_norm 1.4778 (inf) [2021-04-16 04:22:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][1170/1251] eta 0:00:23 lr 0.000557 time 0.2921 (0.2841) loss 3.2181 (3.5934) grad_norm 1.3380 (inf) [2021-04-16 04:22:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][1180/1251] eta 0:00:20 lr 0.000557 time 0.2808 (0.2841) loss 3.7432 (3.5967) grad_norm 1.4472 (inf) [2021-04-16 04:22:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][1190/1251] eta 0:00:17 lr 0.000557 time 0.2584 (0.2841) loss 3.7385 (3.5952) grad_norm 1.6558 (inf) [2021-04-16 04:22:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][1200/1251] eta 0:00:14 lr 0.000557 time 0.2655 (0.2840) loss 2.8408 (3.5932) grad_norm 1.6232 (inf) [2021-04-16 04:23:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][1210/1251] eta 0:00:11 lr 0.000557 time 0.2869 (0.2839) loss 3.9958 (3.5956) grad_norm 1.2778 (inf) [2021-04-16 04:23:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][1220/1251] eta 0:00:08 lr 0.000557 time 0.2723 (0.2840) loss 3.0830 (3.5933) grad_norm 1.5532 (inf) [2021-04-16 04:23:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][1230/1251] eta 0:00:05 lr 0.000557 time 0.2535 (0.2841) loss 2.9374 (3.5924) grad_norm 1.2116 (inf) [2021-04-16 04:23:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][1240/1251] eta 0:00:03 lr 0.000557 time 0.3170 (0.2840) loss 4.4106 (3.5923) grad_norm 1.5597 (inf) [2021-04-16 04:23:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [139/300][1250/1251] eta 0:00:00 lr 0.000557 time 0.2484 (0.2837) loss 2.7223 (3.5927) grad_norm 1.5224 (inf) [2021-04-16 04:23:17 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 139 training takes 0:05:59 [2021-04-16 04:23:17 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_139.pth saving...... [2021-04-16 04:23:32 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_139.pth saved !!! [2021-04-16 04:23:33 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.202 (1.202) Loss 1.0131 (1.0131) Acc@1 77.734 (77.734) Acc@5 92.969 (92.969) [2021-04-16 04:23:35 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.425 (0.265) Loss 1.0458 (1.0396) Acc@1 75.098 (75.613) Acc@5 92.383 (92.844) [2021-04-16 04:23:37 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.091 (0.232) Loss 1.0307 (1.0434) Acc@1 76.562 (75.758) Acc@5 92.676 (92.834) [2021-04-16 04:23:39 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.139 (0.224) Loss 1.0573 (1.0412) Acc@1 73.730 (75.643) Acc@5 93.555 (92.956) [2021-04-16 04:23:40 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.070 (0.208) Loss 0.9781 (1.0424) Acc@1 75.977 (75.572) Acc@5 93.750 (93.024) [2021-04-16 04:23:46 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 75.552 Acc@5 93.036 [2021-04-16 04:23:46 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 75.6% [2021-04-16 04:23:46 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 75.62% [2021-04-16 04:23:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][0/1251] eta 3:45:58 lr 0.000557 time 10.8382 (10.8382) loss 3.2473 (3.2473) grad_norm 1.6222 (1.6222) [2021-04-16 04:24:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][10/1251] eta 0:25:32 lr 0.000557 time 0.2798 (1.2351) loss 2.4443 (3.2279) grad_norm 1.4385 (1.4288) [2021-04-16 04:24:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][20/1251] eta 0:15:59 lr 0.000557 time 0.2873 (0.7791) loss 4.5539 (3.5184) grad_norm 1.3658 (1.4370) [2021-04-16 04:24:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][30/1251] eta 0:12:34 lr 0.000557 time 0.2850 (0.6183) loss 2.7531 (3.4445) grad_norm 1.3748 (1.4423) [2021-04-16 04:24:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][40/1251] eta 0:10:50 lr 0.000557 time 0.2866 (0.5368) loss 3.5607 (3.4058) grad_norm 1.5826 (1.4469) [2021-04-16 04:24:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][50/1251] eta 0:09:43 lr 0.000557 time 0.2811 (0.4858) loss 4.1308 (3.4394) grad_norm 1.4433 (1.4560) [2021-04-16 04:24:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][60/1251] eta 0:08:58 lr 0.000556 time 0.2761 (0.4522) loss 3.8080 (3.4966) grad_norm 1.3275 (1.4432) [2021-04-16 04:24:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][70/1251] eta 0:08:24 lr 0.000556 time 0.2638 (0.4274) loss 4.0985 (3.5232) grad_norm 1.2911 (1.4450) [2021-04-16 04:24:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][80/1251] eta 0:07:58 lr 0.000556 time 0.2828 (0.4088) loss 3.5288 (3.5190) grad_norm 1.5118 (1.4435) [2021-04-16 04:24:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][90/1251] eta 0:07:39 lr 0.000556 time 0.2649 (0.3955) loss 4.0464 (3.5371) grad_norm 1.5859 (1.4494) [2021-04-16 04:24:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][100/1251] eta 0:07:21 lr 0.000556 time 0.2654 (0.3839) loss 3.9003 (3.5298) grad_norm 1.4579 (1.4525) [2021-04-16 04:24:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][110/1251] eta 0:07:06 lr 0.000556 time 0.2907 (0.3739) loss 3.4207 (3.5373) grad_norm 1.4366 (1.4499) [2021-04-16 04:24:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][120/1251] eta 0:06:53 lr 0.000556 time 0.2655 (0.3658) loss 3.6907 (3.5702) grad_norm 1.6867 (1.4512) [2021-04-16 04:24:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][130/1251] eta 0:06:43 lr 0.000556 time 0.2539 (0.3598) loss 3.3385 (3.5734) grad_norm 1.2368 (1.4488) [2021-04-16 04:24:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][140/1251] eta 0:06:34 lr 0.000556 time 0.2784 (0.3550) loss 4.0896 (3.5731) grad_norm 1.3800 (1.4474) [2021-04-16 04:24:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][150/1251] eta 0:06:25 lr 0.000556 time 0.2429 (0.3498) loss 3.8159 (3.5850) grad_norm 1.6354 (1.4532) [2021-04-16 04:24:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][160/1251] eta 0:06:17 lr 0.000556 time 0.2856 (0.3464) loss 3.5578 (3.5687) grad_norm 1.5412 (1.4549) [2021-04-16 04:24:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][170/1251] eta 0:06:10 lr 0.000556 time 0.2743 (0.3423) loss 3.2585 (3.5752) grad_norm 1.5463 (1.4542) [2021-04-16 04:24:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][180/1251] eta 0:06:02 lr 0.000556 time 0.2680 (0.3385) loss 3.7705 (3.5604) grad_norm 1.3731 (1.4515) [2021-04-16 04:24:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][190/1251] eta 0:05:55 lr 0.000556 time 0.2630 (0.3355) loss 3.5229 (3.5632) grad_norm 1.5058 (1.4536) [2021-04-16 04:24:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][200/1251] eta 0:05:49 lr 0.000556 time 0.2544 (0.3323) loss 3.9328 (3.5579) grad_norm 1.5660 (1.4547) [2021-04-16 04:24:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][210/1251] eta 0:05:43 lr 0.000556 time 0.2714 (0.3295) loss 3.7377 (3.5736) grad_norm 1.6127 (1.4575) [2021-04-16 04:24:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][220/1251] eta 0:05:37 lr 0.000556 time 0.2931 (0.3275) loss 3.2635 (3.5728) grad_norm 1.3217 (1.4547) [2021-04-16 04:25:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][230/1251] eta 0:05:32 lr 0.000556 time 0.2586 (0.3259) loss 3.9294 (3.5733) grad_norm 1.3414 (1.4535) [2021-04-16 04:25:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][240/1251] eta 0:05:27 lr 0.000556 time 0.2516 (0.3243) loss 3.7334 (3.5738) grad_norm 1.3795 (1.4578) [2021-04-16 04:25:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][250/1251] eta 0:05:22 lr 0.000556 time 0.2789 (0.3224) loss 3.1449 (3.5761) grad_norm 1.2837 (1.4615) [2021-04-16 04:25:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][260/1251] eta 0:05:17 lr 0.000556 time 0.2864 (0.3206) loss 4.0393 (3.5839) grad_norm 1.6684 (1.4610) [2021-04-16 04:25:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][270/1251] eta 0:05:13 lr 0.000556 time 0.3093 (0.3191) loss 4.0634 (3.6000) grad_norm 1.4682 (1.4625) [2021-04-16 04:25:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][280/1251] eta 0:05:08 lr 0.000556 time 0.3005 (0.3176) loss 3.8004 (3.6030) grad_norm 1.5984 (1.4636) [2021-04-16 04:25:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][290/1251] eta 0:05:03 lr 0.000556 time 0.2914 (0.3162) loss 3.7082 (3.6118) grad_norm 1.3892 (1.4628) [2021-04-16 04:25:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][300/1251] eta 0:04:59 lr 0.000556 time 0.2616 (0.3148) loss 2.8876 (3.6090) grad_norm 1.4306 (1.4639) [2021-04-16 04:25:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][310/1251] eta 0:04:55 lr 0.000555 time 0.2757 (0.3136) loss 3.0195 (3.6006) grad_norm 1.6474 (1.4638) [2021-04-16 04:25:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][320/1251] eta 0:04:50 lr 0.000555 time 0.2771 (0.3125) loss 3.1432 (3.5988) grad_norm 1.3645 (1.4654) [2021-04-16 04:25:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][330/1251] eta 0:04:46 lr 0.000555 time 0.2889 (0.3114) loss 3.7854 (3.5978) grad_norm 1.6556 (1.4650) [2021-04-16 04:25:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][340/1251] eta 0:04:43 lr 0.000555 time 0.3134 (0.3108) loss 3.9172 (3.5982) grad_norm 1.5248 (1.4639) [2021-04-16 04:25:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][350/1251] eta 0:04:39 lr 0.000555 time 0.2533 (0.3101) loss 2.4731 (3.5957) grad_norm 1.4739 (1.4628) [2021-04-16 04:25:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][360/1251] eta 0:04:35 lr 0.000555 time 0.2633 (0.3091) loss 4.4607 (3.5881) grad_norm 1.4092 (1.4604) [2021-04-16 04:25:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][370/1251] eta 0:04:31 lr 0.000555 time 0.3073 (0.3085) loss 3.3082 (3.5793) grad_norm 1.6392 (1.4607) [2021-04-16 04:25:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][380/1251] eta 0:04:27 lr 0.000555 time 0.2858 (0.3076) loss 4.0007 (3.5837) grad_norm 1.5346 (1.4595) [2021-04-16 04:25:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][390/1251] eta 0:04:24 lr 0.000555 time 0.2648 (0.3072) loss 3.3168 (3.5851) grad_norm 1.4268 (1.4588) [2021-04-16 04:25:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][400/1251] eta 0:04:20 lr 0.000555 time 0.2818 (0.3064) loss 3.9197 (3.5876) grad_norm 1.4387 (1.4580) [2021-04-16 04:25:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][410/1251] eta 0:04:17 lr 0.000555 time 0.2824 (0.3058) loss 2.8495 (3.5888) grad_norm 1.7004 (1.4621) [2021-04-16 04:25:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][420/1251] eta 0:04:13 lr 0.000555 time 0.2501 (0.3054) loss 2.6801 (3.5858) grad_norm 1.3420 (1.4613) [2021-04-16 04:25:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][430/1251] eta 0:04:10 lr 0.000555 time 0.2655 (0.3050) loss 3.3384 (3.5859) grad_norm 1.9646 (1.4614) [2021-04-16 04:26:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][440/1251] eta 0:04:07 lr 0.000555 time 0.2681 (0.3047) loss 3.0344 (3.5786) grad_norm 1.4688 (1.4621) [2021-04-16 04:26:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][450/1251] eta 0:04:03 lr 0.000555 time 0.2657 (0.3041) loss 3.2740 (3.5801) grad_norm 1.4165 (1.4642) [2021-04-16 04:26:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][460/1251] eta 0:03:59 lr 0.000555 time 0.2761 (0.3034) loss 2.6398 (3.5760) grad_norm 1.3981 (1.4633) [2021-04-16 04:26:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][470/1251] eta 0:03:56 lr 0.000555 time 0.2709 (0.3029) loss 2.3831 (3.5715) grad_norm 1.2878 (1.4634) [2021-04-16 04:26:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][480/1251] eta 0:03:53 lr 0.000555 time 0.2718 (0.3023) loss 3.6898 (3.5683) grad_norm 1.6246 (1.4634) [2021-04-16 04:26:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][490/1251] eta 0:03:49 lr 0.000555 time 0.2855 (0.3018) loss 3.9574 (3.5677) grad_norm 1.4176 (1.4636) [2021-04-16 04:26:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][500/1251] eta 0:03:46 lr 0.000555 time 0.2841 (0.3012) loss 4.5382 (3.5659) grad_norm 1.4226 (1.4623) [2021-04-16 04:26:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][510/1251] eta 0:03:42 lr 0.000555 time 0.3015 (0.3008) loss 3.9690 (3.5658) grad_norm 1.3062 (1.4611) [2021-04-16 04:26:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][520/1251] eta 0:03:39 lr 0.000555 time 0.2850 (0.3004) loss 3.7351 (3.5712) grad_norm 1.3626 (1.4605) [2021-04-16 04:26:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][530/1251] eta 0:03:36 lr 0.000555 time 0.2939 (0.3000) loss 3.3436 (3.5691) grad_norm 1.5228 (1.4617) [2021-04-16 04:26:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][540/1251] eta 0:03:33 lr 0.000555 time 0.2688 (0.2997) loss 2.7285 (3.5655) grad_norm 1.5677 (1.4629) [2021-04-16 04:26:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][550/1251] eta 0:03:29 lr 0.000554 time 0.2846 (0.2994) loss 3.4305 (3.5672) grad_norm 1.2071 (1.4626) [2021-04-16 04:26:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][560/1251] eta 0:03:26 lr 0.000554 time 0.2672 (0.2990) loss 4.2318 (3.5699) grad_norm 1.3986 (1.4629) [2021-04-16 04:26:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][570/1251] eta 0:03:23 lr 0.000554 time 0.2654 (0.2988) loss 3.4896 (3.5728) grad_norm 1.7535 (1.4642) [2021-04-16 04:26:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][580/1251] eta 0:03:20 lr 0.000554 time 0.2827 (0.2985) loss 2.6600 (3.5705) grad_norm 1.4364 (1.4630) [2021-04-16 04:26:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][590/1251] eta 0:03:17 lr 0.000554 time 0.2751 (0.2982) loss 3.9694 (3.5733) grad_norm 1.5114 (1.4623) [2021-04-16 04:26:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][600/1251] eta 0:03:13 lr 0.000554 time 0.2740 (0.2978) loss 3.1961 (3.5733) grad_norm 1.6873 (1.4627) [2021-04-16 04:26:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][610/1251] eta 0:03:10 lr 0.000554 time 0.2538 (0.2974) loss 3.9561 (3.5704) grad_norm 1.5443 (1.4615) [2021-04-16 04:26:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][620/1251] eta 0:03:07 lr 0.000554 time 0.2819 (0.2970) loss 4.3087 (3.5769) grad_norm 1.4427 (1.4622) [2021-04-16 04:26:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][630/1251] eta 0:03:04 lr 0.000554 time 0.2905 (0.2967) loss 3.6874 (3.5732) grad_norm 1.3768 (1.4617) [2021-04-16 04:26:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][640/1251] eta 0:03:01 lr 0.000554 time 0.2643 (0.2963) loss 3.1070 (3.5721) grad_norm 1.4095 (1.4620) [2021-04-16 04:26:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][650/1251] eta 0:02:57 lr 0.000554 time 0.2711 (0.2960) loss 2.3967 (3.5656) grad_norm 1.2689 (1.4611) [2021-04-16 04:27:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][660/1251] eta 0:02:54 lr 0.000554 time 0.2872 (0.2957) loss 3.5673 (3.5647) grad_norm 1.2489 (1.4608) [2021-04-16 04:27:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][670/1251] eta 0:02:51 lr 0.000554 time 0.2605 (0.2956) loss 2.8099 (3.5609) grad_norm 1.2897 (1.4624) [2021-04-16 04:27:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][680/1251] eta 0:02:48 lr 0.000554 time 0.2876 (0.2952) loss 3.7190 (3.5632) grad_norm 1.4065 (1.4617) [2021-04-16 04:27:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][690/1251] eta 0:02:45 lr 0.000554 time 0.2869 (0.2949) loss 2.5192 (3.5620) grad_norm 1.2076 (1.4623) [2021-04-16 04:27:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][700/1251] eta 0:02:42 lr 0.000554 time 0.2645 (0.2947) loss 3.2491 (3.5590) grad_norm 1.4818 (1.4614) [2021-04-16 04:27:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][710/1251] eta 0:02:39 lr 0.000554 time 0.2839 (0.2948) loss 4.0500 (3.5606) grad_norm 1.4460 (1.4613) [2021-04-16 04:27:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][720/1251] eta 0:02:36 lr 0.000554 time 0.3088 (0.2947) loss 2.7321 (3.5590) grad_norm 1.3618 (1.4616) [2021-04-16 04:27:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][730/1251] eta 0:02:33 lr 0.000554 time 0.2584 (0.2946) loss 3.0068 (3.5599) grad_norm 1.4010 (1.4626) [2021-04-16 04:27:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][740/1251] eta 0:02:30 lr 0.000554 time 0.2908 (0.2944) loss 4.1658 (3.5603) grad_norm 1.4543 (1.4625) [2021-04-16 04:27:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][750/1251] eta 0:02:27 lr 0.000554 time 0.2716 (0.2941) loss 2.5697 (3.5565) grad_norm 1.4082 (1.4625) [2021-04-16 04:27:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][760/1251] eta 0:02:24 lr 0.000554 time 0.2560 (0.2938) loss 3.8455 (3.5543) grad_norm 1.8141 (1.4643) [2021-04-16 04:27:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][770/1251] eta 0:02:21 lr 0.000554 time 0.2838 (0.2936) loss 3.6138 (3.5575) grad_norm 1.5331 (1.4658) [2021-04-16 04:27:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][780/1251] eta 0:02:18 lr 0.000554 time 0.2495 (0.2934) loss 3.9603 (3.5587) grad_norm 1.3114 (1.4674) [2021-04-16 04:27:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][790/1251] eta 0:02:15 lr 0.000553 time 0.2734 (0.2933) loss 3.2762 (3.5551) grad_norm 1.4386 (1.4671) [2021-04-16 04:27:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][800/1251] eta 0:02:12 lr 0.000553 time 0.3977 (0.2933) loss 3.7297 (3.5564) grad_norm 1.3011 (1.4663) [2021-04-16 04:27:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][810/1251] eta 0:02:09 lr 0.000553 time 0.2913 (0.2932) loss 3.5252 (3.5533) grad_norm 1.5646 (1.4667) [2021-04-16 04:27:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][820/1251] eta 0:02:06 lr 0.000553 time 0.2738 (0.2931) loss 2.9522 (3.5545) grad_norm 1.2537 (1.4660) [2021-04-16 04:27:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][830/1251] eta 0:02:03 lr 0.000553 time 0.2980 (0.2930) loss 3.0858 (3.5533) grad_norm 1.5411 (1.4653) [2021-04-16 04:27:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][840/1251] eta 0:02:00 lr 0.000553 time 0.2911 (0.2928) loss 4.1684 (3.5513) grad_norm 1.5224 (1.4643) [2021-04-16 04:27:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][850/1251] eta 0:01:57 lr 0.000553 time 0.2751 (0.2926) loss 4.1149 (3.5541) grad_norm 1.5775 (1.4639) [2021-04-16 04:27:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][860/1251] eta 0:01:54 lr 0.000553 time 0.2617 (0.2923) loss 4.0498 (3.5560) grad_norm 1.7518 (1.4638) [2021-04-16 04:28:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][870/1251] eta 0:01:51 lr 0.000553 time 0.2678 (0.2921) loss 2.9741 (3.5536) grad_norm 1.5500 (1.4641) [2021-04-16 04:28:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][880/1251] eta 0:01:48 lr 0.000553 time 0.2758 (0.2919) loss 3.6830 (3.5517) grad_norm 1.3281 (1.4630) [2021-04-16 04:28:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][890/1251] eta 0:01:45 lr 0.000553 time 0.2801 (0.2917) loss 3.8201 (3.5520) grad_norm 1.3886 (1.4633) [2021-04-16 04:28:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][900/1251] eta 0:01:42 lr 0.000553 time 0.2688 (0.2915) loss 3.6708 (3.5557) grad_norm 1.3580 (1.4628) [2021-04-16 04:28:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][910/1251] eta 0:01:39 lr 0.000553 time 0.2870 (0.2913) loss 3.0015 (3.5546) grad_norm 1.4081 (1.4621) [2021-04-16 04:28:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][920/1251] eta 0:01:36 lr 0.000553 time 0.2702 (0.2912) loss 2.6178 (3.5538) grad_norm 1.7835 (1.4635) [2021-04-16 04:28:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][930/1251] eta 0:01:33 lr 0.000553 time 0.2808 (0.2912) loss 3.8061 (3.5526) grad_norm 1.6038 (1.4631) [2021-04-16 04:28:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][940/1251] eta 0:01:30 lr 0.000553 time 0.2697 (0.2910) loss 3.8536 (3.5536) grad_norm 1.3281 (1.4620) [2021-04-16 04:28:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][950/1251] eta 0:01:27 lr 0.000553 time 0.2688 (0.2909) loss 3.1015 (3.5534) grad_norm 1.5660 (1.4622) [2021-04-16 04:28:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][960/1251] eta 0:01:24 lr 0.000553 time 0.3093 (0.2908) loss 4.1802 (3.5527) grad_norm 1.3302 (1.4612) [2021-04-16 04:28:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][970/1251] eta 0:01:21 lr 0.000553 time 0.2736 (0.2907) loss 2.7978 (3.5533) grad_norm 1.4939 (1.4608) [2021-04-16 04:28:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][980/1251] eta 0:01:18 lr 0.000553 time 0.2646 (0.2905) loss 3.1013 (3.5534) grad_norm 1.5186 (1.4613) [2021-04-16 04:28:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][990/1251] eta 0:01:15 lr 0.000553 time 0.2941 (0.2906) loss 3.8746 (3.5539) grad_norm 1.3457 (1.4616) [2021-04-16 04:28:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][1000/1251] eta 0:01:12 lr 0.000553 time 0.2793 (0.2904) loss 3.5502 (3.5565) grad_norm 1.3448 (1.4609) [2021-04-16 04:28:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][1010/1251] eta 0:01:10 lr 0.000553 time 0.2760 (0.2905) loss 4.1069 (3.5592) grad_norm 1.4027 (1.4602) [2021-04-16 04:28:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][1020/1251] eta 0:01:07 lr 0.000553 time 0.2776 (0.2904) loss 4.1574 (3.5596) grad_norm 1.4026 (1.4594) [2021-04-16 04:28:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][1030/1251] eta 0:01:04 lr 0.000552 time 0.2736 (0.2903) loss 3.0188 (3.5576) grad_norm 1.5363 (1.4598) [2021-04-16 04:28:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][1040/1251] eta 0:01:01 lr 0.000552 time 0.2757 (0.2902) loss 3.5502 (3.5571) grad_norm 1.2216 (1.4595) [2021-04-16 04:28:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][1050/1251] eta 0:00:58 lr 0.000552 time 0.2671 (0.2900) loss 4.0012 (3.5548) grad_norm 1.4298 (1.4598) [2021-04-16 04:28:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][1060/1251] eta 0:00:55 lr 0.000552 time 0.2707 (0.2899) loss 2.7677 (3.5528) grad_norm 1.6982 (1.4599) [2021-04-16 04:28:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][1070/1251] eta 0:00:52 lr 0.000552 time 0.2959 (0.2898) loss 3.6528 (3.5523) grad_norm 1.2985 (1.4598) [2021-04-16 04:28:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][1080/1251] eta 0:00:49 lr 0.000552 time 0.2737 (0.2896) loss 2.9698 (3.5553) grad_norm 2.0255 (1.4601) [2021-04-16 04:29:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][1090/1251] eta 0:00:46 lr 0.000552 time 0.2964 (0.2895) loss 3.9944 (3.5536) grad_norm 1.6835 (1.4612) [2021-04-16 04:29:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][1100/1251] eta 0:00:43 lr 0.000552 time 0.2912 (0.2894) loss 3.8408 (3.5527) grad_norm 1.3898 (1.4618) [2021-04-16 04:29:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][1110/1251] eta 0:00:40 lr 0.000552 time 0.2673 (0.2892) loss 4.1874 (3.5538) grad_norm 1.4758 (1.4619) [2021-04-16 04:29:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][1120/1251] eta 0:00:37 lr 0.000552 time 0.2673 (0.2891) loss 3.5099 (3.5555) grad_norm 1.6232 (1.4617) [2021-04-16 04:29:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][1130/1251] eta 0:00:34 lr 0.000552 time 0.2851 (0.2890) loss 3.9286 (3.5565) grad_norm 1.3996 (1.4620) [2021-04-16 04:29:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][1140/1251] eta 0:00:32 lr 0.000552 time 0.2853 (0.2888) loss 2.9018 (3.5572) grad_norm 1.4517 (1.4620) [2021-04-16 04:29:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][1150/1251] eta 0:00:29 lr 0.000552 time 0.2722 (0.2889) loss 3.7960 (3.5561) grad_norm 1.7377 (1.4629) [2021-04-16 04:29:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][1160/1251] eta 0:00:26 lr 0.000552 time 0.2679 (0.2888) loss 3.1832 (3.5567) grad_norm 1.3757 (1.4632) [2021-04-16 04:29:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][1170/1251] eta 0:00:23 lr 0.000552 time 0.2858 (0.2887) loss 3.5925 (3.5597) grad_norm 1.5777 (1.4638) [2021-04-16 04:29:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][1180/1251] eta 0:00:20 lr 0.000552 time 0.2727 (0.2886) loss 4.2230 (3.5608) grad_norm 1.4170 (1.4636) [2021-04-16 04:29:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][1190/1251] eta 0:00:17 lr 0.000552 time 0.3028 (0.2885) loss 3.8137 (3.5619) grad_norm 1.5753 (1.4633) [2021-04-16 04:29:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][1200/1251] eta 0:00:14 lr 0.000552 time 0.2731 (0.2884) loss 3.6239 (3.5618) grad_norm 1.3859 (1.4629) [2021-04-16 04:29:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][1210/1251] eta 0:00:11 lr 0.000552 time 0.2849 (0.2883) loss 4.1159 (3.5626) grad_norm 1.3345 (1.4623) [2021-04-16 04:29:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][1220/1251] eta 0:00:08 lr 0.000552 time 0.2877 (0.2882) loss 4.2276 (3.5645) grad_norm 1.3363 (1.4614) [2021-04-16 04:29:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][1230/1251] eta 0:00:06 lr 0.000552 time 0.2831 (0.2881) loss 4.2164 (3.5658) grad_norm 1.4765 (1.4611) [2021-04-16 04:29:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][1240/1251] eta 0:00:03 lr 0.000552 time 0.2486 (0.2879) loss 2.4114 (3.5668) grad_norm 1.4266 (1.4619) [2021-04-16 04:29:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [140/300][1250/1251] eta 0:00:00 lr 0.000552 time 0.2546 (0.2876) loss 3.6165 (3.5674) grad_norm 1.3857 (1.4613) [2021-04-16 04:29:49 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 140 training takes 0:06:02 [2021-04-16 04:29:49 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_140.pth saving...... [2021-04-16 04:30:03 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_140.pth saved !!! [2021-04-16 04:30:04 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.389 (1.389) Loss 1.0380 (1.0380) Acc@1 75.879 (75.879) Acc@5 92.773 (92.773) [2021-04-16 04:30:05 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.106 (0.217) Loss 1.0301 (1.0199) Acc@1 77.441 (75.950) Acc@5 91.895 (93.146) [2021-04-16 04:30:08 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.136 (0.218) Loss 1.0408 (1.0134) Acc@1 75.293 (75.977) Acc@5 93.750 (93.355) [2021-04-16 04:30:10 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.124 (0.237) Loss 1.0396 (1.0237) Acc@1 74.707 (75.658) Acc@5 92.285 (93.186) [2021-04-16 04:30:12 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.223) Loss 0.9967 (1.0292) Acc@1 76.172 (75.619) Acc@5 93.750 (93.116) [2021-04-16 04:30:16 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 75.624 Acc@5 93.076 [2021-04-16 04:30:16 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 75.6% [2021-04-16 04:30:16 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 75.62% [2021-04-16 04:30:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][0/1251] eta 2:23:12 lr 0.000552 time 6.8684 (6.8684) loss 4.1164 (4.1164) grad_norm 1.2435 (1.2435) [2021-04-16 04:30:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][10/1251] eta 0:18:20 lr 0.000552 time 0.4207 (0.8868) loss 3.5860 (3.5877) grad_norm 1.3629 (1.4453) [2021-04-16 04:30:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][20/1251] eta 0:12:21 lr 0.000552 time 0.2882 (0.6025) loss 3.9045 (3.6145) grad_norm 1.3398 (1.4364) [2021-04-16 04:30:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][30/1251] eta 0:10:05 lr 0.000551 time 0.3081 (0.4962) loss 3.9473 (3.5576) grad_norm 1.6810 (1.4386) [2021-04-16 04:30:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][40/1251] eta 0:08:57 lr 0.000551 time 0.2541 (0.4435) loss 3.7910 (3.5664) grad_norm 1.2847 (1.4576) [2021-04-16 04:30:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][50/1251] eta 0:08:14 lr 0.000551 time 0.2866 (0.4113) loss 3.6610 (3.5867) grad_norm 1.4080 (1.4496) [2021-04-16 04:30:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][60/1251] eta 0:07:43 lr 0.000551 time 0.2642 (0.3889) loss 2.9617 (3.5893) grad_norm 1.4393 (1.4458) [2021-04-16 04:30:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][70/1251] eta 0:07:20 lr 0.000551 time 0.2832 (0.3730) loss 3.7684 (3.6138) grad_norm 1.4128 (1.4404) [2021-04-16 04:30:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][80/1251] eta 0:07:02 lr 0.000551 time 0.2790 (0.3609) loss 3.7325 (3.5911) grad_norm 1.2340 (1.4403) [2021-04-16 04:30:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][90/1251] eta 0:06:47 lr 0.000551 time 0.2745 (0.3511) loss 3.2185 (3.5952) grad_norm 1.5027 (1.4500) [2021-04-16 04:30:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][100/1251] eta 0:06:36 lr 0.000551 time 0.2915 (0.3447) loss 3.8467 (3.6167) grad_norm 1.5524 (1.4634) [2021-04-16 04:30:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][110/1251] eta 0:06:26 lr 0.000551 time 0.2550 (0.3385) loss 2.8536 (3.6259) grad_norm 1.5247 (1.4701) [2021-04-16 04:30:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][120/1251] eta 0:06:17 lr 0.000551 time 0.2485 (0.3335) loss 3.6192 (3.6186) grad_norm 1.5563 (1.4751) [2021-04-16 04:30:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][130/1251] eta 0:06:10 lr 0.000551 time 0.2960 (0.3306) loss 4.1012 (3.6107) grad_norm 1.4355 (1.4703) [2021-04-16 04:31:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][140/1251] eta 0:06:03 lr 0.000551 time 0.3038 (0.3274) loss 3.2203 (3.5997) grad_norm 1.3012 (1.4718) [2021-04-16 04:31:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][150/1251] eta 0:05:57 lr 0.000551 time 0.2979 (0.3246) loss 2.8038 (3.5892) grad_norm 1.5215 (1.4735) [2021-04-16 04:31:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][160/1251] eta 0:05:52 lr 0.000551 time 0.2774 (0.3228) loss 3.6192 (3.5980) grad_norm 1.4477 (1.4739) [2021-04-16 04:31:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][170/1251] eta 0:05:46 lr 0.000551 time 0.2911 (0.3202) loss 4.0343 (3.5996) grad_norm 1.6278 (1.4748) [2021-04-16 04:31:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][180/1251] eta 0:05:41 lr 0.000551 time 0.2485 (0.3185) loss 3.5119 (3.5906) grad_norm 1.4155 (1.4734) [2021-04-16 04:31:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][190/1251] eta 0:05:35 lr 0.000551 time 0.2791 (0.3163) loss 4.2245 (3.5980) grad_norm 1.3858 (1.4689) [2021-04-16 04:31:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][200/1251] eta 0:05:31 lr 0.000551 time 0.2845 (0.3154) loss 4.3592 (3.6004) grad_norm 1.4834 (1.4677) [2021-04-16 04:31:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][210/1251] eta 0:05:26 lr 0.000551 time 0.2846 (0.3134) loss 2.7653 (3.5905) grad_norm 1.7099 (1.4680) [2021-04-16 04:31:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][220/1251] eta 0:05:22 lr 0.000551 time 0.2700 (0.3123) loss 4.0406 (3.5979) grad_norm 1.3840 (1.4707) [2021-04-16 04:31:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][230/1251] eta 0:05:17 lr 0.000551 time 0.2764 (0.3109) loss 2.5685 (3.5836) grad_norm 1.7195 (1.4748) [2021-04-16 04:31:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][240/1251] eta 0:05:13 lr 0.000551 time 0.2769 (0.3101) loss 3.9644 (3.5826) grad_norm 1.2519 (1.4742) [2021-04-16 04:31:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][250/1251] eta 0:05:09 lr 0.000551 time 0.2804 (0.3088) loss 3.9377 (3.5872) grad_norm 1.3564 (1.4730) [2021-04-16 04:31:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][260/1251] eta 0:05:04 lr 0.000551 time 0.2504 (0.3075) loss 4.3092 (3.5867) grad_norm 1.3257 (1.4700) [2021-04-16 04:31:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][270/1251] eta 0:05:00 lr 0.000550 time 0.2653 (0.3064) loss 3.2879 (3.5915) grad_norm 1.2231 (1.4674) [2021-04-16 04:31:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][280/1251] eta 0:04:56 lr 0.000550 time 0.2905 (0.3056) loss 3.9057 (3.6068) grad_norm 1.4691 (1.4659) [2021-04-16 04:31:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][290/1251] eta 0:04:52 lr 0.000550 time 0.2750 (0.3046) loss 4.5461 (3.6124) grad_norm 1.5451 (1.4625) [2021-04-16 04:31:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][300/1251] eta 0:04:48 lr 0.000550 time 0.2737 (0.3038) loss 2.4065 (3.6042) grad_norm 1.2781 (1.4610) [2021-04-16 04:31:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][310/1251] eta 0:04:45 lr 0.000550 time 0.3038 (0.3031) loss 3.8994 (3.6034) grad_norm 1.2991 (1.4591) [2021-04-16 04:31:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][320/1251] eta 0:04:41 lr 0.000550 time 0.2954 (0.3022) loss 4.2946 (3.6062) grad_norm 1.5809 (1.4601) [2021-04-16 04:31:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][330/1251] eta 0:04:37 lr 0.000550 time 0.2789 (0.3014) loss 4.1183 (3.5929) grad_norm 1.3683 (1.4602) [2021-04-16 04:31:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][340/1251] eta 0:04:34 lr 0.000550 time 0.2878 (0.3010) loss 3.4559 (3.5963) grad_norm 1.3486 (1.4616) [2021-04-16 04:32:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][350/1251] eta 0:04:30 lr 0.000550 time 0.2702 (0.3004) loss 4.2604 (3.5937) grad_norm 1.8337 (1.4640) [2021-04-16 04:32:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][360/1251] eta 0:04:27 lr 0.000550 time 0.2768 (0.3005) loss 4.4509 (3.5950) grad_norm 1.7053 (1.4653) [2021-04-16 04:32:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][370/1251] eta 0:04:24 lr 0.000550 time 0.2669 (0.2999) loss 3.3935 (3.5924) grad_norm 1.3457 (1.4669) [2021-04-16 04:32:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][380/1251] eta 0:04:20 lr 0.000550 time 0.2694 (0.2995) loss 3.1452 (3.5905) grad_norm 1.4094 (1.4654) [2021-04-16 04:32:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][390/1251] eta 0:04:17 lr 0.000550 time 0.2873 (0.2991) loss 3.6522 (3.5944) grad_norm 1.7802 (1.4656) [2021-04-16 04:32:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][400/1251] eta 0:04:14 lr 0.000550 time 0.2831 (0.2987) loss 3.5339 (3.5911) grad_norm 1.5450 (1.4642) [2021-04-16 04:32:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][410/1251] eta 0:04:10 lr 0.000550 time 0.2562 (0.2981) loss 3.1612 (3.5916) grad_norm 1.8440 (1.4671) [2021-04-16 04:32:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][420/1251] eta 0:04:07 lr 0.000550 time 0.3136 (0.2979) loss 3.7815 (3.5928) grad_norm 1.3976 (1.4657) [2021-04-16 04:32:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][430/1251] eta 0:04:04 lr 0.000550 time 0.2781 (0.2975) loss 3.3972 (3.5939) grad_norm 1.3391 (1.4653) [2021-04-16 04:32:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][440/1251] eta 0:04:01 lr 0.000550 time 0.2958 (0.2973) loss 3.7744 (3.5926) grad_norm 1.6983 (1.4658) [2021-04-16 04:32:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][450/1251] eta 0:03:57 lr 0.000550 time 0.2709 (0.2969) loss 4.0190 (3.5904) grad_norm 1.2367 (1.4653) [2021-04-16 04:32:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][460/1251] eta 0:03:54 lr 0.000550 time 0.2689 (0.2964) loss 4.5963 (3.5885) grad_norm 1.4760 (1.4627) [2021-04-16 04:32:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][470/1251] eta 0:03:51 lr 0.000550 time 0.2892 (0.2962) loss 2.7834 (3.5934) grad_norm 1.5734 (1.4622) [2021-04-16 04:32:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][480/1251] eta 0:03:48 lr 0.000550 time 0.2810 (0.2960) loss 4.3391 (3.5912) grad_norm 1.5816 (1.4611) [2021-04-16 04:32:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][490/1251] eta 0:03:44 lr 0.000550 time 0.2689 (0.2956) loss 3.9119 (3.5927) grad_norm 1.4334 (1.4612) [2021-04-16 04:32:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][500/1251] eta 0:03:41 lr 0.000550 time 0.2889 (0.2953) loss 3.1773 (3.5910) grad_norm 1.3095 (1.4606) [2021-04-16 04:32:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][510/1251] eta 0:03:38 lr 0.000549 time 0.2723 (0.2950) loss 3.3097 (3.5873) grad_norm 1.4175 (1.4608) [2021-04-16 04:32:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][520/1251] eta 0:03:35 lr 0.000549 time 0.2823 (0.2947) loss 3.8455 (3.5901) grad_norm 1.4793 (1.4610) [2021-04-16 04:32:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][530/1251] eta 0:03:32 lr 0.000549 time 0.2779 (0.2944) loss 4.3019 (3.5872) grad_norm 1.7170 (1.4605) [2021-04-16 04:32:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][540/1251] eta 0:03:29 lr 0.000549 time 0.2574 (0.2942) loss 3.3800 (3.5880) grad_norm 1.3524 (1.4630) [2021-04-16 04:32:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][550/1251] eta 0:03:25 lr 0.000549 time 0.2733 (0.2939) loss 4.2135 (3.5897) grad_norm 1.4312 (1.4643) [2021-04-16 04:33:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][560/1251] eta 0:03:22 lr 0.000549 time 0.2556 (0.2935) loss 2.7137 (3.5897) grad_norm 1.5335 (1.4660) [2021-04-16 04:33:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][570/1251] eta 0:03:19 lr 0.000549 time 0.3028 (0.2937) loss 3.8959 (3.5879) grad_norm 1.7136 (1.4666) [2021-04-16 04:33:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][580/1251] eta 0:03:16 lr 0.000549 time 0.2900 (0.2935) loss 3.8945 (3.5869) grad_norm 1.4458 (1.4663) [2021-04-16 04:33:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][590/1251] eta 0:03:14 lr 0.000549 time 0.2560 (0.2936) loss 3.9568 (3.5916) grad_norm 1.4068 (1.4673) [2021-04-16 04:33:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][600/1251] eta 0:03:11 lr 0.000549 time 0.2796 (0.2934) loss 3.0545 (3.5919) grad_norm 1.7072 (1.4684) [2021-04-16 04:33:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][610/1251] eta 0:03:07 lr 0.000549 time 0.2871 (0.2932) loss 4.1040 (3.5924) grad_norm 1.5611 (1.4687) [2021-04-16 04:33:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][620/1251] eta 0:03:04 lr 0.000549 time 0.2703 (0.2929) loss 2.6438 (3.5909) grad_norm 1.3623 (1.4695) [2021-04-16 04:33:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][630/1251] eta 0:03:01 lr 0.000549 time 0.2683 (0.2927) loss 3.8185 (3.5913) grad_norm 1.2956 (1.4686) [2021-04-16 04:33:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][640/1251] eta 0:02:58 lr 0.000549 time 0.2451 (0.2926) loss 4.0152 (3.5921) grad_norm 1.4592 (1.4702) [2021-04-16 04:33:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][650/1251] eta 0:02:55 lr 0.000549 time 0.2845 (0.2924) loss 3.7953 (3.5943) grad_norm 1.3470 (1.4696) [2021-04-16 04:33:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][660/1251] eta 0:02:52 lr 0.000549 time 0.2978 (0.2922) loss 3.8112 (3.5895) grad_norm 1.3297 (1.4678) [2021-04-16 04:33:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][670/1251] eta 0:02:49 lr 0.000549 time 0.2615 (0.2920) loss 4.1645 (3.5882) grad_norm 1.4952 (1.4684) [2021-04-16 04:33:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][680/1251] eta 0:02:46 lr 0.000549 time 0.2813 (0.2918) loss 3.6544 (3.5889) grad_norm 1.5116 (1.4686) [2021-04-16 04:33:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][690/1251] eta 0:02:43 lr 0.000549 time 0.2719 (0.2916) loss 3.3827 (3.5939) grad_norm 1.6909 (1.4698) [2021-04-16 04:33:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][700/1251] eta 0:02:40 lr 0.000549 time 0.2864 (0.2914) loss 3.2730 (3.5992) grad_norm 1.3711 (1.4704) [2021-04-16 04:33:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][710/1251] eta 0:02:37 lr 0.000549 time 0.2994 (0.2914) loss 2.6553 (3.5965) grad_norm 1.5980 (1.4708) [2021-04-16 04:33:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][720/1251] eta 0:02:34 lr 0.000549 time 0.2512 (0.2913) loss 3.5185 (3.5951) grad_norm 1.3614 (1.4691) [2021-04-16 04:33:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][730/1251] eta 0:02:31 lr 0.000549 time 0.2534 (0.2910) loss 2.9713 (3.5922) grad_norm 1.6508 (1.4686) [2021-04-16 04:33:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][740/1251] eta 0:02:28 lr 0.000549 time 0.2414 (0.2909) loss 3.6313 (3.5895) grad_norm 1.3892 (1.4682) [2021-04-16 04:33:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][750/1251] eta 0:02:25 lr 0.000548 time 0.2654 (0.2907) loss 3.5332 (3.5889) grad_norm 1.4780 (1.4681) [2021-04-16 04:33:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][760/1251] eta 0:02:22 lr 0.000548 time 0.3082 (0.2907) loss 2.5062 (3.5841) grad_norm 1.4145 (1.4676) [2021-04-16 04:34:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][770/1251] eta 0:02:19 lr 0.000548 time 0.2922 (0.2905) loss 3.9651 (3.5847) grad_norm 1.3536 (1.4674) [2021-04-16 04:34:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][780/1251] eta 0:02:16 lr 0.000548 time 0.2713 (0.2905) loss 2.7055 (3.5829) grad_norm 1.4000 (1.4665) [2021-04-16 04:34:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][790/1251] eta 0:02:13 lr 0.000548 time 0.2937 (0.2906) loss 2.5924 (3.5834) grad_norm 1.4505 (1.4652) [2021-04-16 04:34:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][800/1251] eta 0:02:10 lr 0.000548 time 0.2708 (0.2903) loss 4.0868 (3.5836) grad_norm 1.3450 (1.4654) [2021-04-16 04:34:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][810/1251] eta 0:02:07 lr 0.000548 time 0.2851 (0.2902) loss 4.3762 (3.5847) grad_norm 1.6375 (1.4664) [2021-04-16 04:34:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][820/1251] eta 0:02:05 lr 0.000548 time 0.2525 (0.2902) loss 3.6105 (3.5846) grad_norm 1.4400 (1.4666) [2021-04-16 04:34:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][830/1251] eta 0:02:02 lr 0.000548 time 0.2724 (0.2900) loss 4.1933 (3.5862) grad_norm 1.3839 (1.4664) [2021-04-16 04:34:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][840/1251] eta 0:01:59 lr 0.000548 time 0.2656 (0.2900) loss 3.9264 (3.5882) grad_norm 1.4507 (1.4661) [2021-04-16 04:34:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][850/1251] eta 0:01:56 lr 0.000548 time 0.2491 (0.2898) loss 3.3702 (3.5913) grad_norm 1.5538 (1.4665) [2021-04-16 04:34:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][860/1251] eta 0:01:53 lr 0.000548 time 0.2775 (0.2896) loss 4.5102 (3.5899) grad_norm 1.6475 (1.4666) [2021-04-16 04:34:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][870/1251] eta 0:01:50 lr 0.000548 time 0.2823 (0.2895) loss 3.1781 (3.5900) grad_norm 1.2854 (1.4674) [2021-04-16 04:34:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][880/1251] eta 0:01:47 lr 0.000548 time 0.2942 (0.2894) loss 3.5257 (3.5916) grad_norm 1.4106 (1.4694) [2021-04-16 04:34:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][890/1251] eta 0:01:44 lr 0.000548 time 0.2834 (0.2893) loss 2.6837 (3.5904) grad_norm 1.6225 (1.4694) [2021-04-16 04:34:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][900/1251] eta 0:01:41 lr 0.000548 time 0.2898 (0.2893) loss 4.0423 (3.5872) grad_norm 1.3866 (1.4689) [2021-04-16 04:34:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][910/1251] eta 0:01:38 lr 0.000548 time 0.2671 (0.2891) loss 4.0251 (3.5831) grad_norm 1.3437 (1.4697) [2021-04-16 04:34:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][920/1251] eta 0:01:35 lr 0.000548 time 0.2725 (0.2890) loss 3.2343 (3.5811) grad_norm 1.3731 (1.4700) [2021-04-16 04:34:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][930/1251] eta 0:01:32 lr 0.000548 time 0.2431 (0.2890) loss 3.2515 (3.5806) grad_norm 1.8304 (1.4694) [2021-04-16 04:34:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][940/1251] eta 0:01:29 lr 0.000548 time 0.2712 (0.2890) loss 4.1678 (3.5805) grad_norm 1.4830 (1.4689) [2021-04-16 04:34:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][950/1251] eta 0:01:26 lr 0.000548 time 0.3057 (0.2889) loss 2.7440 (3.5795) grad_norm 1.4057 (1.4688) [2021-04-16 04:34:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][960/1251] eta 0:01:24 lr 0.000548 time 0.2926 (0.2888) loss 3.8435 (3.5782) grad_norm 1.4649 (1.4692) [2021-04-16 04:34:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][970/1251] eta 0:01:21 lr 0.000548 time 0.2645 (0.2887) loss 4.1861 (3.5789) grad_norm 1.6005 (1.4691) [2021-04-16 04:34:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][980/1251] eta 0:01:18 lr 0.000548 time 0.2665 (0.2886) loss 2.7981 (3.5769) grad_norm 1.3643 (1.4699) [2021-04-16 04:35:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][990/1251] eta 0:01:15 lr 0.000547 time 0.2630 (0.2885) loss 4.2239 (3.5799) grad_norm 1.4143 (1.4707) [2021-04-16 04:35:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][1000/1251] eta 0:01:12 lr 0.000547 time 0.2909 (0.2885) loss 3.9913 (3.5809) grad_norm 1.8007 (1.4712) [2021-04-16 04:35:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][1010/1251] eta 0:01:09 lr 0.000547 time 0.2887 (0.2883) loss 3.6768 (3.5816) grad_norm 1.4458 (1.4718) [2021-04-16 04:35:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][1020/1251] eta 0:01:06 lr 0.000547 time 0.2871 (0.2883) loss 3.6942 (3.5839) grad_norm 1.3795 (1.4726) [2021-04-16 04:35:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][1030/1251] eta 0:01:03 lr 0.000547 time 0.2607 (0.2882) loss 3.4921 (3.5850) grad_norm 1.2799 (1.4721) [2021-04-16 04:35:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][1040/1251] eta 0:01:00 lr 0.000547 time 0.2573 (0.2881) loss 3.4699 (3.5811) grad_norm 1.3616 (1.4709) [2021-04-16 04:35:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][1050/1251] eta 0:00:57 lr 0.000547 time 0.2449 (0.2880) loss 3.4543 (3.5809) grad_norm 1.6536 (1.4707) [2021-04-16 04:35:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][1060/1251] eta 0:00:55 lr 0.000547 time 0.2767 (0.2880) loss 3.4604 (3.5816) grad_norm 1.6888 (1.4711) [2021-04-16 04:35:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][1070/1251] eta 0:00:52 lr 0.000547 time 0.2876 (0.2879) loss 4.3059 (3.5821) grad_norm 1.4116 (1.4713) [2021-04-16 04:35:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][1080/1251] eta 0:00:49 lr 0.000547 time 0.2793 (0.2877) loss 3.0986 (3.5827) grad_norm 1.2757 (1.4716) [2021-04-16 04:35:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][1090/1251] eta 0:00:46 lr 0.000547 time 0.2908 (0.2877) loss 3.8744 (3.5813) grad_norm 1.7715 (1.4722) [2021-04-16 04:35:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][1100/1251] eta 0:00:43 lr 0.000547 time 0.2548 (0.2876) loss 3.0980 (3.5821) grad_norm 1.4857 (1.4719) [2021-04-16 04:35:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][1110/1251] eta 0:00:40 lr 0.000547 time 0.3059 (0.2875) loss 4.1361 (3.5839) grad_norm 1.3452 (1.4713) [2021-04-16 04:35:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][1120/1251] eta 0:00:37 lr 0.000547 time 0.2592 (0.2875) loss 4.1772 (3.5828) grad_norm 1.5465 (inf) [2021-04-16 04:35:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][1130/1251] eta 0:00:34 lr 0.000547 time 0.3055 (0.2875) loss 4.5239 (3.5840) grad_norm 1.4193 (inf) [2021-04-16 04:35:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][1140/1251] eta 0:00:31 lr 0.000547 time 0.2436 (0.2874) loss 3.2339 (3.5834) grad_norm 1.4824 (inf) [2021-04-16 04:35:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][1150/1251] eta 0:00:29 lr 0.000547 time 0.2485 (0.2874) loss 3.4556 (3.5840) grad_norm 1.3663 (inf) [2021-04-16 04:35:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][1160/1251] eta 0:00:26 lr 0.000547 time 0.2681 (0.2873) loss 4.0716 (3.5821) grad_norm 1.5912 (inf) [2021-04-16 04:35:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][1170/1251] eta 0:00:23 lr 0.000547 time 0.3082 (0.2874) loss 4.5319 (3.5816) grad_norm 1.3748 (inf) [2021-04-16 04:35:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][1180/1251] eta 0:00:20 lr 0.000547 time 0.2782 (0.2874) loss 4.3793 (3.5802) grad_norm 1.2334 (inf) [2021-04-16 04:35:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][1190/1251] eta 0:00:17 lr 0.000547 time 0.2894 (0.2873) loss 3.6244 (3.5805) grad_norm 1.6976 (inf) [2021-04-16 04:36:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][1200/1251] eta 0:00:14 lr 0.000547 time 0.2766 (0.2873) loss 3.6870 (3.5795) grad_norm 1.2972 (inf) [2021-04-16 04:36:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][1210/1251] eta 0:00:11 lr 0.000547 time 0.3060 (0.2873) loss 4.1940 (3.5821) grad_norm 1.3718 (inf) [2021-04-16 04:36:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][1220/1251] eta 0:00:08 lr 0.000547 time 0.2811 (0.2872) loss 3.9416 (3.5829) grad_norm 1.5024 (inf) [2021-04-16 04:36:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][1230/1251] eta 0:00:06 lr 0.000547 time 0.2842 (0.2871) loss 3.8510 (3.5831) grad_norm 1.3248 (inf) [2021-04-16 04:36:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][1240/1251] eta 0:00:03 lr 0.000546 time 0.2486 (0.2870) loss 3.4350 (3.5850) grad_norm 1.1450 (inf) [2021-04-16 04:36:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [141/300][1250/1251] eta 0:00:00 lr 0.000546 time 0.2484 (0.2867) loss 4.0974 (3.5871) grad_norm 1.3400 (inf) [2021-04-16 04:36:17 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 141 training takes 0:06:01 [2021-04-16 04:36:17 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_141.pth saving...... [2021-04-16 04:36:25 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_141.pth saved !!! [2021-04-16 04:36:26 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.113 (1.113) Loss 1.1130 (1.1130) Acc@1 74.121 (74.121) Acc@5 92.090 (92.090) [2021-04-16 04:36:28 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.180 (0.267) Loss 0.9815 (1.0251) Acc@1 77.637 (75.542) Acc@5 94.043 (93.013) [2021-04-16 04:36:30 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.157 (0.221) Loss 0.9948 (1.0346) Acc@1 76.465 (75.298) Acc@5 93.262 (92.932) [2021-04-16 04:36:33 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.127 (0.241) Loss 1.0646 (1.0265) Acc@1 75.488 (75.611) Acc@5 93.457 (93.048) [2021-04-16 04:36:34 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.222) Loss 1.1038 (1.0218) Acc@1 72.852 (75.674) Acc@5 91.602 (93.059) [2021-04-16 04:36:39 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 75.574 Acc@5 93.060 [2021-04-16 04:36:39 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 75.6% [2021-04-16 04:36:39 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 75.62% [2021-04-16 04:36:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][0/1251] eta 2:03:18 lr 0.000546 time 5.9140 (5.9140) loss 4.1246 (4.1246) grad_norm 1.4648 (1.4648) [2021-04-16 04:36:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][10/1251] eta 0:16:30 lr 0.000546 time 0.4208 (0.7978) loss 3.8654 (3.6975) grad_norm 1.5100 (1.4713) [2021-04-16 04:36:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][20/1251] eta 0:11:21 lr 0.000546 time 0.2632 (0.5537) loss 3.9759 (3.5593) grad_norm 1.4338 (1.4738) [2021-04-16 04:36:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][30/1251] eta 0:09:28 lr 0.000546 time 0.2742 (0.4653) loss 4.1632 (3.5284) grad_norm 1.5208 (1.4618) [2021-04-16 04:36:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][40/1251] eta 0:08:27 lr 0.000546 time 0.2610 (0.4190) loss 3.1043 (3.5387) grad_norm 1.4698 (1.4494) [2021-04-16 04:36:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][50/1251] eta 0:07:49 lr 0.000546 time 0.2606 (0.3909) loss 2.2714 (3.5037) grad_norm 1.4193 (1.4523) [2021-04-16 04:37:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][60/1251] eta 0:07:24 lr 0.000546 time 0.2528 (0.3729) loss 3.2941 (3.5310) grad_norm 2.1155 (1.4727) [2021-04-16 04:37:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][70/1251] eta 0:07:04 lr 0.000546 time 0.2798 (0.3597) loss 2.2270 (3.5326) grad_norm 1.4866 (1.4782) [2021-04-16 04:37:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][80/1251] eta 0:06:49 lr 0.000546 time 0.2584 (0.3496) loss 3.7739 (3.4852) grad_norm 1.6009 (1.4864) [2021-04-16 04:37:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][90/1251] eta 0:06:36 lr 0.000546 time 0.2696 (0.3416) loss 3.9515 (3.4801) grad_norm 2.0878 (1.5058) [2021-04-16 04:37:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][100/1251] eta 0:06:26 lr 0.000546 time 0.2772 (0.3356) loss 3.3125 (3.4884) grad_norm 1.3998 (1.5135) [2021-04-16 04:37:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][110/1251] eta 0:06:17 lr 0.000546 time 0.2833 (0.3305) loss 3.1787 (3.4952) grad_norm 1.2717 (1.5103) [2021-04-16 04:37:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][120/1251] eta 0:06:08 lr 0.000546 time 0.2761 (0.3259) loss 4.2257 (3.5089) grad_norm 1.6860 (1.5224) [2021-04-16 04:37:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][130/1251] eta 0:06:01 lr 0.000546 time 0.2907 (0.3224) loss 3.9323 (3.4957) grad_norm 1.5152 (1.5232) [2021-04-16 04:37:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][140/1251] eta 0:05:55 lr 0.000546 time 0.2744 (0.3200) loss 3.8156 (3.4968) grad_norm 1.7988 (1.5205) [2021-04-16 04:37:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][150/1251] eta 0:05:49 lr 0.000546 time 0.2846 (0.3173) loss 3.1383 (3.5116) grad_norm 1.4007 (1.5149) [2021-04-16 04:37:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][160/1251] eta 0:05:44 lr 0.000546 time 0.2718 (0.3155) loss 2.4675 (3.5054) grad_norm 1.5043 (1.5117) [2021-04-16 04:37:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][170/1251] eta 0:05:38 lr 0.000546 time 0.2548 (0.3132) loss 4.1362 (3.5195) grad_norm 1.5410 (1.5077) [2021-04-16 04:37:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][180/1251] eta 0:05:33 lr 0.000546 time 0.2816 (0.3114) loss 3.6936 (3.5296) grad_norm 1.2987 (1.5031) [2021-04-16 04:37:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][190/1251] eta 0:05:28 lr 0.000546 time 0.2682 (0.3096) loss 4.5524 (3.5539) grad_norm 1.3291 (1.5076) [2021-04-16 04:37:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][200/1251] eta 0:05:23 lr 0.000546 time 0.2867 (0.3081) loss 3.9170 (3.5569) grad_norm 1.5914 (1.5092) [2021-04-16 04:37:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][210/1251] eta 0:05:19 lr 0.000546 time 0.2808 (0.3065) loss 3.3646 (3.5359) grad_norm 1.3600 (1.5071) [2021-04-16 04:37:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][220/1251] eta 0:05:14 lr 0.000546 time 0.2742 (0.3054) loss 3.8646 (3.5454) grad_norm 1.4774 (1.5098) [2021-04-16 04:37:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][230/1251] eta 0:05:10 lr 0.000545 time 0.2652 (0.3044) loss 3.4149 (3.5444) grad_norm 1.2895 (1.5053) [2021-04-16 04:37:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][240/1251] eta 0:05:06 lr 0.000545 time 0.3059 (0.3035) loss 4.0024 (3.5482) grad_norm 1.5001 (1.5006) [2021-04-16 04:37:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][250/1251] eta 0:05:02 lr 0.000545 time 0.2909 (0.3026) loss 4.1220 (3.5597) grad_norm 1.5998 (1.5030) [2021-04-16 04:37:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][260/1251] eta 0:04:58 lr 0.000545 time 0.2806 (0.3016) loss 3.4761 (3.5604) grad_norm 1.2982 (1.5062) [2021-04-16 04:38:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][270/1251] eta 0:04:55 lr 0.000545 time 0.3034 (0.3008) loss 4.1217 (3.5596) grad_norm 1.2133 (1.5018) [2021-04-16 04:38:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][280/1251] eta 0:04:51 lr 0.000545 time 0.2901 (0.2998) loss 3.4240 (3.5550) grad_norm 1.4844 (1.5028) [2021-04-16 04:38:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][290/1251] eta 0:04:47 lr 0.000545 time 0.2515 (0.2989) loss 3.3881 (3.5564) grad_norm 1.6549 (1.5061) [2021-04-16 04:38:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][300/1251] eta 0:04:43 lr 0.000545 time 0.2710 (0.2982) loss 3.8263 (3.5494) grad_norm 1.4548 (1.5030) [2021-04-16 04:38:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][310/1251] eta 0:04:40 lr 0.000545 time 0.3092 (0.2978) loss 3.2155 (3.5477) grad_norm 1.5015 (1.5029) [2021-04-16 04:38:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][320/1251] eta 0:04:37 lr 0.000545 time 0.2738 (0.2978) loss 3.4852 (3.5505) grad_norm 1.5317 (1.5050) [2021-04-16 04:38:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][330/1251] eta 0:04:33 lr 0.000545 time 0.2858 (0.2972) loss 3.4495 (3.5578) grad_norm 1.4288 (1.5057) [2021-04-16 04:38:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][340/1251] eta 0:04:30 lr 0.000545 time 0.2948 (0.2966) loss 2.7284 (3.5557) grad_norm 1.8085 (1.5059) [2021-04-16 04:38:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][350/1251] eta 0:04:27 lr 0.000545 time 0.2715 (0.2966) loss 3.6791 (3.5537) grad_norm 1.3194 (1.5045) [2021-04-16 04:38:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][360/1251] eta 0:04:23 lr 0.000545 time 0.2651 (0.2959) loss 3.8506 (3.5567) grad_norm 1.6485 (1.5031) [2021-04-16 04:38:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][370/1251] eta 0:04:20 lr 0.000545 time 0.2593 (0.2961) loss 3.6284 (3.5598) grad_norm 1.3817 (1.5025) [2021-04-16 04:38:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][380/1251] eta 0:04:17 lr 0.000545 time 0.2809 (0.2956) loss 4.1384 (3.5611) grad_norm 1.5918 (1.5027) [2021-04-16 04:38:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][390/1251] eta 0:04:14 lr 0.000545 time 0.2596 (0.2952) loss 3.9593 (3.5673) grad_norm 1.5121 (1.5010) [2021-04-16 04:38:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][400/1251] eta 0:04:10 lr 0.000545 time 0.2860 (0.2949) loss 3.5447 (3.5654) grad_norm 1.4934 (1.5011) [2021-04-16 04:38:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][410/1251] eta 0:04:07 lr 0.000545 time 0.2750 (0.2947) loss 4.2589 (3.5622) grad_norm 1.4528 (1.4997) [2021-04-16 04:38:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][420/1251] eta 0:04:04 lr 0.000545 time 0.2859 (0.2942) loss 4.1175 (3.5631) grad_norm 1.4443 (1.4978) [2021-04-16 04:38:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][430/1251] eta 0:04:01 lr 0.000545 time 0.2910 (0.2939) loss 3.7556 (3.5622) grad_norm 1.6364 (1.4986) [2021-04-16 04:38:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][440/1251] eta 0:03:58 lr 0.000545 time 0.2667 (0.2937) loss 3.8552 (3.5579) grad_norm 1.5990 (1.4996) [2021-04-16 04:38:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][450/1251] eta 0:03:55 lr 0.000545 time 0.3008 (0.2934) loss 3.2651 (3.5623) grad_norm 1.6586 (1.4999) [2021-04-16 04:38:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][460/1251] eta 0:03:51 lr 0.000545 time 0.2715 (0.2932) loss 2.7841 (3.5606) grad_norm 1.4831 (1.4993) [2021-04-16 04:38:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][470/1251] eta 0:03:48 lr 0.000544 time 0.2726 (0.2927) loss 3.0769 (3.5568) grad_norm 1.4499 (1.4991) [2021-04-16 04:39:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][480/1251] eta 0:03:45 lr 0.000544 time 0.3043 (0.2927) loss 3.8879 (3.5589) grad_norm 1.5981 (1.4977) [2021-04-16 04:39:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][490/1251] eta 0:03:42 lr 0.000544 time 0.2925 (0.2925) loss 4.0049 (3.5581) grad_norm 1.6292 (1.4979) [2021-04-16 04:39:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][500/1251] eta 0:03:39 lr 0.000544 time 0.2653 (0.2921) loss 3.5271 (3.5560) grad_norm 1.4533 (1.4997) [2021-04-16 04:39:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][510/1251] eta 0:03:36 lr 0.000544 time 0.2681 (0.2918) loss 3.4033 (3.5518) grad_norm 1.4538 (1.4983) [2021-04-16 04:39:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][520/1251] eta 0:03:33 lr 0.000544 time 0.3044 (0.2918) loss 3.9421 (3.5495) grad_norm 1.4541 (1.4959) [2021-04-16 04:39:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][530/1251] eta 0:03:30 lr 0.000544 time 0.2851 (0.2917) loss 3.7369 (3.5525) grad_norm 1.5346 (1.4962) [2021-04-16 04:39:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][540/1251] eta 0:03:27 lr 0.000544 time 0.2741 (0.2914) loss 2.5280 (3.5463) grad_norm 1.5854 (1.4982) [2021-04-16 04:39:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][550/1251] eta 0:03:24 lr 0.000544 time 0.3180 (0.2913) loss 4.1479 (3.5506) grad_norm 1.3595 (1.4986) [2021-04-16 04:39:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][560/1251] eta 0:03:21 lr 0.000544 time 0.2925 (0.2913) loss 3.8274 (3.5553) grad_norm 1.5029 (1.4992) [2021-04-16 04:39:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][570/1251] eta 0:03:18 lr 0.000544 time 0.2887 (0.2913) loss 2.8936 (3.5550) grad_norm 1.4754 (1.5002) [2021-04-16 04:39:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][580/1251] eta 0:03:15 lr 0.000544 time 0.2745 (0.2912) loss 3.1665 (3.5534) grad_norm 1.4419 (1.4995) [2021-04-16 04:39:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][590/1251] eta 0:03:12 lr 0.000544 time 0.2640 (0.2914) loss 3.9174 (3.5548) grad_norm 1.3908 (1.4989) [2021-04-16 04:39:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][600/1251] eta 0:03:09 lr 0.000544 time 0.3057 (0.2911) loss 3.6728 (3.5549) grad_norm 1.4674 (1.4992) [2021-04-16 04:39:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][610/1251] eta 0:03:06 lr 0.000544 time 0.2991 (0.2910) loss 4.0127 (3.5527) grad_norm 1.3464 (1.4991) [2021-04-16 04:39:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][620/1251] eta 0:03:03 lr 0.000544 time 0.2954 (0.2908) loss 4.1146 (3.5578) grad_norm 1.3670 (1.4985) [2021-04-16 04:39:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][630/1251] eta 0:03:00 lr 0.000544 time 0.2784 (0.2906) loss 2.9527 (3.5561) grad_norm 1.3747 (1.4989) [2021-04-16 04:39:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][640/1251] eta 0:02:57 lr 0.000544 time 0.2901 (0.2906) loss 3.9826 (3.5581) grad_norm 1.5739 (1.4989) [2021-04-16 04:39:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][650/1251] eta 0:02:54 lr 0.000544 time 0.2808 (0.2905) loss 3.3652 (3.5613) grad_norm 1.5583 (1.4990) [2021-04-16 04:39:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][660/1251] eta 0:02:51 lr 0.000544 time 0.3170 (0.2904) loss 3.2399 (3.5608) grad_norm 1.7211 (1.4990) [2021-04-16 04:39:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][670/1251] eta 0:02:48 lr 0.000544 time 0.2670 (0.2901) loss 2.6504 (3.5598) grad_norm 1.4231 (1.4975) [2021-04-16 04:39:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][680/1251] eta 0:02:45 lr 0.000544 time 0.2820 (0.2901) loss 3.8114 (3.5582) grad_norm 1.4083 (1.4961) [2021-04-16 04:40:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][690/1251] eta 0:02:42 lr 0.000544 time 0.2894 (0.2899) loss 3.1959 (3.5592) grad_norm 1.3835 (1.4945) [2021-04-16 04:40:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][700/1251] eta 0:02:39 lr 0.000544 time 0.2879 (0.2897) loss 3.7702 (3.5631) grad_norm 1.5082 (1.4940) [2021-04-16 04:40:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][710/1251] eta 0:02:36 lr 0.000543 time 0.2636 (0.2895) loss 3.7338 (3.5601) grad_norm 1.7969 (1.4938) [2021-04-16 04:40:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][720/1251] eta 0:02:33 lr 0.000543 time 0.2668 (0.2897) loss 4.0071 (3.5589) grad_norm 1.3551 (1.4931) [2021-04-16 04:40:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][730/1251] eta 0:02:30 lr 0.000543 time 0.3903 (0.2897) loss 4.1888 (3.5608) grad_norm 1.2863 (1.4925) [2021-04-16 04:40:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][740/1251] eta 0:02:27 lr 0.000543 time 0.2667 (0.2895) loss 3.5041 (3.5644) grad_norm 1.4168 (1.4925) [2021-04-16 04:40:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][750/1251] eta 0:02:25 lr 0.000543 time 0.3002 (0.2894) loss 3.3996 (3.5659) grad_norm 1.5241 (1.4919) [2021-04-16 04:40:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][760/1251] eta 0:02:22 lr 0.000543 time 0.2901 (0.2893) loss 3.5023 (3.5651) grad_norm 1.4528 (1.4906) [2021-04-16 04:40:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][770/1251] eta 0:02:19 lr 0.000543 time 0.2841 (0.2892) loss 4.3129 (3.5679) grad_norm 1.3329 (1.4894) [2021-04-16 04:40:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][780/1251] eta 0:02:16 lr 0.000543 time 0.2529 (0.2891) loss 3.1922 (3.5604) grad_norm 1.5424 (1.4895) [2021-04-16 04:40:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][790/1251] eta 0:02:13 lr 0.000543 time 0.3181 (0.2891) loss 2.3940 (3.5581) grad_norm 1.4205 (1.4888) [2021-04-16 04:40:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][800/1251] eta 0:02:10 lr 0.000543 time 0.3671 (0.2890) loss 2.7733 (3.5554) grad_norm 1.5432 (1.4891) [2021-04-16 04:40:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][810/1251] eta 0:02:07 lr 0.000543 time 0.2978 (0.2889) loss 3.8158 (3.5568) grad_norm 1.4242 (1.4886) [2021-04-16 04:40:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][820/1251] eta 0:02:04 lr 0.000543 time 0.2797 (0.2888) loss 3.5102 (3.5575) grad_norm 1.4541 (1.4902) [2021-04-16 04:40:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][830/1251] eta 0:02:01 lr 0.000543 time 0.2837 (0.2887) loss 4.2838 (3.5573) grad_norm 1.6299 (1.4910) [2021-04-16 04:40:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][840/1251] eta 0:01:58 lr 0.000543 time 0.2791 (0.2885) loss 4.0345 (3.5583) grad_norm 1.4411 (1.4921) [2021-04-16 04:40:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][850/1251] eta 0:01:55 lr 0.000543 time 0.2976 (0.2883) loss 3.5608 (3.5597) grad_norm 2.1086 (1.4923) [2021-04-16 04:40:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][860/1251] eta 0:01:52 lr 0.000543 time 0.2596 (0.2881) loss 2.3155 (3.5580) grad_norm 1.2847 (1.4916) [2021-04-16 04:40:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][870/1251] eta 0:01:49 lr 0.000543 time 0.2716 (0.2880) loss 2.8766 (3.5540) grad_norm 1.9123 (1.4920) [2021-04-16 04:40:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][880/1251] eta 0:01:46 lr 0.000543 time 0.2825 (0.2880) loss 2.7275 (3.5531) grad_norm 1.2635 (1.4924) [2021-04-16 04:40:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][890/1251] eta 0:01:43 lr 0.000543 time 0.2522 (0.2879) loss 4.5453 (3.5491) grad_norm 1.5559 (1.4929) [2021-04-16 04:40:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][900/1251] eta 0:01:41 lr 0.000543 time 0.2671 (0.2879) loss 3.3063 (3.5507) grad_norm 1.4682 (1.4929) [2021-04-16 04:41:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][910/1251] eta 0:01:38 lr 0.000543 time 0.2620 (0.2878) loss 3.1896 (3.5513) grad_norm 1.4450 (1.4921) [2021-04-16 04:41:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][920/1251] eta 0:01:35 lr 0.000543 time 0.2790 (0.2878) loss 3.7900 (3.5543) grad_norm 1.3467 (1.4911) [2021-04-16 04:41:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][930/1251] eta 0:01:32 lr 0.000543 time 0.2759 (0.2877) loss 3.6606 (3.5523) grad_norm 1.5245 (1.4915) [2021-04-16 04:41:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][940/1251] eta 0:01:29 lr 0.000543 time 0.2708 (0.2877) loss 4.3801 (3.5514) grad_norm 1.8369 (1.4936) [2021-04-16 04:41:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][950/1251] eta 0:01:26 lr 0.000542 time 0.2558 (0.2876) loss 4.0851 (3.5526) grad_norm 1.2889 (1.4938) [2021-04-16 04:41:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][960/1251] eta 0:01:23 lr 0.000542 time 0.2734 (0.2875) loss 4.2133 (3.5526) grad_norm 1.4074 (1.4934) [2021-04-16 04:41:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][970/1251] eta 0:01:20 lr 0.000542 time 0.2757 (0.2874) loss 3.7885 (3.5543) grad_norm 1.3319 (1.4923) [2021-04-16 04:41:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][980/1251] eta 0:01:17 lr 0.000542 time 0.2739 (0.2873) loss 3.8418 (3.5552) grad_norm 1.2200 (1.4921) [2021-04-16 04:41:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][990/1251] eta 0:01:14 lr 0.000542 time 0.2813 (0.2872) loss 2.5832 (3.5544) grad_norm 1.1965 (1.4915) [2021-04-16 04:41:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][1000/1251] eta 0:01:12 lr 0.000542 time 0.2915 (0.2872) loss 3.9665 (3.5541) grad_norm 1.4436 (1.4914) [2021-04-16 04:41:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][1010/1251] eta 0:01:09 lr 0.000542 time 0.2642 (0.2870) loss 4.0217 (3.5540) grad_norm 1.4407 (1.4910) [2021-04-16 04:41:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][1020/1251] eta 0:01:06 lr 0.000542 time 0.2725 (0.2869) loss 2.6434 (3.5549) grad_norm 1.8480 (1.4915) [2021-04-16 04:41:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][1030/1251] eta 0:01:03 lr 0.000542 time 0.2750 (0.2868) loss 4.1294 (3.5520) grad_norm 1.6196 (1.4917) [2021-04-16 04:41:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][1040/1251] eta 0:01:00 lr 0.000542 time 0.2424 (0.2867) loss 2.9034 (3.5510) grad_norm 1.5306 (1.4924) [2021-04-16 04:41:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][1050/1251] eta 0:00:57 lr 0.000542 time 0.2567 (0.2866) loss 2.6620 (3.5512) grad_norm 1.5465 (1.4920) [2021-04-16 04:41:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][1060/1251] eta 0:00:54 lr 0.000542 time 0.2851 (0.2866) loss 3.5217 (3.5515) grad_norm 1.4544 (1.4923) [2021-04-16 04:41:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][1070/1251] eta 0:00:51 lr 0.000542 time 0.2936 (0.2865) loss 3.3642 (3.5506) grad_norm 1.5584 (1.4921) [2021-04-16 04:41:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][1080/1251] eta 0:00:48 lr 0.000542 time 0.2780 (0.2865) loss 3.0941 (3.5489) grad_norm 1.7484 (1.4922) [2021-04-16 04:41:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][1090/1251] eta 0:00:46 lr 0.000542 time 0.2906 (0.2866) loss 4.1229 (3.5493) grad_norm 1.4476 (1.4921) [2021-04-16 04:41:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][1100/1251] eta 0:00:43 lr 0.000542 time 0.2769 (0.2866) loss 3.9166 (3.5501) grad_norm 1.4484 (1.4919) [2021-04-16 04:41:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][1110/1251] eta 0:00:40 lr 0.000542 time 0.2770 (0.2865) loss 3.1709 (3.5492) grad_norm 1.3628 (1.4918) [2021-04-16 04:42:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][1120/1251] eta 0:00:37 lr 0.000542 time 0.2738 (0.2864) loss 3.9047 (3.5486) grad_norm 1.5042 (1.4919) [2021-04-16 04:42:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][1130/1251] eta 0:00:34 lr 0.000542 time 0.2882 (0.2864) loss 3.2874 (3.5483) grad_norm 1.4039 (1.4918) [2021-04-16 04:42:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][1140/1251] eta 0:00:31 lr 0.000542 time 0.2577 (0.2864) loss 3.9180 (3.5486) grad_norm 1.5097 (1.4912) [2021-04-16 04:42:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][1150/1251] eta 0:00:28 lr 0.000542 time 0.2869 (0.2864) loss 3.9002 (3.5473) grad_norm 1.6185 (1.4911) [2021-04-16 04:42:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][1160/1251] eta 0:00:26 lr 0.000542 time 0.2936 (0.2863) loss 3.9652 (3.5484) grad_norm 1.5832 (1.4916) [2021-04-16 04:42:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][1170/1251] eta 0:00:23 lr 0.000542 time 0.2850 (0.2864) loss 3.2291 (3.5492) grad_norm 1.5542 (1.4920) [2021-04-16 04:42:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][1180/1251] eta 0:00:20 lr 0.000542 time 0.2617 (0.2863) loss 4.0355 (3.5497) grad_norm 1.3160 (1.4923) [2021-04-16 04:42:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][1190/1251] eta 0:00:17 lr 0.000542 time 0.3018 (0.2862) loss 2.7645 (3.5475) grad_norm 1.3233 (1.4916) [2021-04-16 04:42:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][1200/1251] eta 0:00:14 lr 0.000541 time 0.2782 (0.2861) loss 2.9305 (3.5482) grad_norm 1.7559 (1.4916) [2021-04-16 04:42:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][1210/1251] eta 0:00:11 lr 0.000541 time 0.2839 (0.2860) loss 4.0926 (3.5484) grad_norm 1.8011 (1.4914) [2021-04-16 04:42:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][1220/1251] eta 0:00:08 lr 0.000541 time 0.2777 (0.2859) loss 3.7795 (3.5499) grad_norm 1.4191 (1.4918) [2021-04-16 04:42:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][1230/1251] eta 0:00:06 lr 0.000541 time 0.2665 (0.2859) loss 2.5140 (3.5497) grad_norm 1.4239 (1.4913) [2021-04-16 04:42:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][1240/1251] eta 0:00:03 lr 0.000541 time 0.2657 (0.2857) loss 3.7316 (3.5508) grad_norm 1.6500 (1.4908) [2021-04-16 04:42:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [142/300][1250/1251] eta 0:00:00 lr 0.000541 time 0.2484 (0.2855) loss 4.3600 (3.5487) grad_norm 1.8414 (1.4914) [2021-04-16 04:42:39 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 142 training takes 0:05:59 [2021-04-16 04:42:39 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_142.pth saving...... [2021-04-16 04:42:52 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_142.pth saved !!! [2021-04-16 04:42:53 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.191 (1.191) Loss 0.9917 (0.9917) Acc@1 76.660 (76.660) Acc@5 93.555 (93.555) [2021-04-16 04:42:55 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.134 (0.245) Loss 1.1167 (1.0266) Acc@1 73.926 (75.843) Acc@5 91.699 (93.208) [2021-04-16 04:42:57 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.880 (0.253) Loss 1.0251 (1.0404) Acc@1 76.270 (75.739) Acc@5 93.945 (93.141) [2021-04-16 04:42:59 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.107 (0.226) Loss 0.9276 (1.0367) Acc@1 78.711 (75.794) Acc@5 94.434 (93.155) [2021-04-16 04:43:01 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.289 (0.223) Loss 1.0696 (1.0424) Acc@1 74.414 (75.598) Acc@5 91.992 (93.114) [2021-04-16 04:43:09 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 75.590 Acc@5 93.170 [2021-04-16 04:43:09 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 75.6% [2021-04-16 04:43:09 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 75.62% [2021-04-16 04:43:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][0/1251] eta 0:34:12 lr 0.000541 time 1.6406 (1.6406) loss 3.6651 (3.6651) grad_norm 1.4152 (1.4152) [2021-04-16 04:43:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][10/1251] eta 0:08:19 lr 0.000541 time 0.3867 (0.4023) loss 3.1794 (3.3945) grad_norm 1.3121 (1.4782) [2021-04-16 04:43:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][20/1251] eta 0:07:01 lr 0.000541 time 0.2846 (0.3420) loss 3.7029 (3.6012) grad_norm 1.6390 (1.4560) [2021-04-16 04:43:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][30/1251] eta 0:06:31 lr 0.000541 time 0.2806 (0.3208) loss 2.9494 (3.4651) grad_norm 1.5042 (1.4649) [2021-04-16 04:43:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][40/1251] eta 0:06:15 lr 0.000541 time 0.2593 (0.3097) loss 3.6929 (3.5034) grad_norm 1.6252 (1.4613) [2021-04-16 04:43:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][50/1251] eta 0:06:06 lr 0.000541 time 0.2787 (0.3051) loss 3.5244 (3.5291) grad_norm 1.5335 (1.4561) [2021-04-16 04:43:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][60/1251] eta 0:06:00 lr 0.000541 time 0.2979 (0.3027) loss 3.8391 (3.5571) grad_norm 1.2973 (1.4575) [2021-04-16 04:43:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][70/1251] eta 0:05:53 lr 0.000541 time 0.2668 (0.2992) loss 2.9787 (3.5827) grad_norm 1.4230 (1.4649) [2021-04-16 04:43:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][80/1251] eta 0:05:47 lr 0.000541 time 0.3190 (0.2968) loss 3.0629 (3.5982) grad_norm 1.5447 (1.4636) [2021-04-16 04:43:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][90/1251] eta 0:05:42 lr 0.000541 time 0.2778 (0.2947) loss 3.0262 (3.6095) grad_norm 1.5233 (1.4647) [2021-04-16 04:43:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][100/1251] eta 0:05:38 lr 0.000541 time 0.2634 (0.2943) loss 4.2291 (3.5908) grad_norm 1.3609 (1.4715) [2021-04-16 04:43:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][110/1251] eta 0:05:34 lr 0.000541 time 0.2920 (0.2929) loss 2.4704 (3.5853) grad_norm 1.3407 (1.4698) [2021-04-16 04:43:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][120/1251] eta 0:05:30 lr 0.000541 time 0.3081 (0.2920) loss 3.8426 (3.6146) grad_norm 1.4427 (1.4637) [2021-04-16 04:43:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][130/1251] eta 0:05:26 lr 0.000541 time 0.2804 (0.2917) loss 3.1781 (3.5902) grad_norm 1.3779 (1.4646) [2021-04-16 04:43:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][140/1251] eta 0:05:24 lr 0.000541 time 0.2951 (0.2918) loss 3.1113 (3.5849) grad_norm 1.5346 (1.4663) [2021-04-16 04:43:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][150/1251] eta 0:05:20 lr 0.000541 time 0.2504 (0.2909) loss 3.8107 (3.5770) grad_norm 1.3252 (1.4658) [2021-04-16 04:43:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][160/1251] eta 0:05:17 lr 0.000541 time 0.2776 (0.2909) loss 3.4877 (3.5750) grad_norm 1.3111 (1.4735) [2021-04-16 04:43:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][170/1251] eta 0:05:13 lr 0.000541 time 0.2952 (0.2902) loss 4.1757 (3.5821) grad_norm 1.4443 (1.4730) [2021-04-16 04:44:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][180/1251] eta 0:05:10 lr 0.000541 time 0.2883 (0.2900) loss 2.7593 (3.5798) grad_norm 1.6532 (1.4750) [2021-04-16 04:44:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][190/1251] eta 0:05:06 lr 0.000540 time 0.2641 (0.2893) loss 3.7593 (3.5903) grad_norm 1.3624 (1.4718) [2021-04-16 04:44:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][200/1251] eta 0:05:03 lr 0.000540 time 0.2823 (0.2888) loss 4.2906 (3.5856) grad_norm 1.3198 (1.4729) [2021-04-16 04:44:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][210/1251] eta 0:04:59 lr 0.000540 time 0.2759 (0.2881) loss 4.4653 (3.5846) grad_norm 1.7843 (1.4736) [2021-04-16 04:44:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][220/1251] eta 0:04:56 lr 0.000540 time 0.2923 (0.2876) loss 3.1176 (3.5795) grad_norm 1.3240 (1.4761) [2021-04-16 04:44:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][230/1251] eta 0:04:53 lr 0.000540 time 0.2967 (0.2871) loss 3.5818 (3.5791) grad_norm 1.5694 (1.4797) [2021-04-16 04:44:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][240/1251] eta 0:04:49 lr 0.000540 time 0.2703 (0.2867) loss 3.6016 (3.5789) grad_norm 1.4427 (1.4791) [2021-04-16 04:44:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][250/1251] eta 0:04:46 lr 0.000540 time 0.2917 (0.2866) loss 4.4535 (3.5754) grad_norm 1.5968 (1.4774) [2021-04-16 04:44:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][260/1251] eta 0:04:43 lr 0.000540 time 0.2718 (0.2865) loss 3.3588 (3.5678) grad_norm 1.5348 (1.4784) [2021-04-16 04:44:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][270/1251] eta 0:04:41 lr 0.000540 time 0.2787 (0.2867) loss 3.4608 (3.5717) grad_norm 1.4007 (1.4777) [2021-04-16 04:44:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][280/1251] eta 0:04:38 lr 0.000540 time 0.3124 (0.2868) loss 4.0813 (3.5652) grad_norm 1.6833 (1.4798) [2021-04-16 04:44:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][290/1251] eta 0:04:35 lr 0.000540 time 0.2721 (0.2871) loss 3.3034 (3.5601) grad_norm 1.4250 (1.4803) [2021-04-16 04:44:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][300/1251] eta 0:04:32 lr 0.000540 time 0.2645 (0.2867) loss 3.8711 (3.5568) grad_norm 1.4455 (1.4844) [2021-04-16 04:44:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][310/1251] eta 0:04:29 lr 0.000540 time 0.3023 (0.2865) loss 3.8044 (3.5535) grad_norm 1.7406 (1.4841) [2021-04-16 04:44:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][320/1251] eta 0:04:26 lr 0.000540 time 0.2934 (0.2862) loss 3.2464 (3.5519) grad_norm 1.3414 (1.4841) [2021-04-16 04:44:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][330/1251] eta 0:04:23 lr 0.000540 time 0.2917 (0.2863) loss 3.9403 (3.5516) grad_norm 1.5393 (1.4822) [2021-04-16 04:44:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][340/1251] eta 0:04:21 lr 0.000540 time 0.2711 (0.2865) loss 3.7070 (3.5555) grad_norm 1.2291 (1.4836) [2021-04-16 04:44:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][350/1251] eta 0:04:18 lr 0.000540 time 0.2728 (0.2864) loss 3.6097 (3.5519) grad_norm 1.5658 (1.4859) [2021-04-16 04:44:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][360/1251] eta 0:04:15 lr 0.000540 time 0.3037 (0.2867) loss 3.5586 (3.5596) grad_norm 1.5917 (1.4861) [2021-04-16 04:44:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][370/1251] eta 0:04:12 lr 0.000540 time 0.2820 (0.2869) loss 3.5677 (3.5646) grad_norm 1.6508 (1.4847) [2021-04-16 04:44:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][380/1251] eta 0:04:09 lr 0.000540 time 0.2915 (0.2870) loss 3.6122 (3.5668) grad_norm 1.3707 (1.4846) [2021-04-16 04:45:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][390/1251] eta 0:04:06 lr 0.000540 time 0.2643 (0.2867) loss 3.7662 (3.5633) grad_norm 1.7004 (1.4841) [2021-04-16 04:45:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][400/1251] eta 0:04:03 lr 0.000540 time 0.2995 (0.2864) loss 3.9226 (3.5628) grad_norm 1.4499 (1.4839) [2021-04-16 04:45:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][410/1251] eta 0:04:00 lr 0.000540 time 0.2792 (0.2861) loss 3.9643 (3.5590) grad_norm 1.4259 (1.4852) [2021-04-16 04:45:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][420/1251] eta 0:03:57 lr 0.000540 time 0.2762 (0.2859) loss 2.2183 (3.5620) grad_norm 1.6200 (1.4860) [2021-04-16 04:45:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][430/1251] eta 0:03:54 lr 0.000539 time 0.2828 (0.2858) loss 2.6647 (3.5623) grad_norm 1.4876 (1.4844) [2021-04-16 04:45:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][440/1251] eta 0:03:51 lr 0.000539 time 0.2695 (0.2858) loss 4.3431 (3.5604) grad_norm 1.5028 (1.4835) [2021-04-16 04:45:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][450/1251] eta 0:03:48 lr 0.000539 time 0.2612 (0.2856) loss 2.7393 (3.5560) grad_norm 1.4295 (1.4830) [2021-04-16 04:45:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][460/1251] eta 0:03:45 lr 0.000539 time 0.2831 (0.2854) loss 3.1356 (3.5602) grad_norm 1.6398 (1.4821) [2021-04-16 04:45:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][470/1251] eta 0:03:42 lr 0.000539 time 0.3106 (0.2854) loss 3.7197 (3.5617) grad_norm 1.6932 (1.4826) [2021-04-16 04:45:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][480/1251] eta 0:03:39 lr 0.000539 time 0.2758 (0.2852) loss 3.5921 (3.5640) grad_norm 1.5946 (1.4833) [2021-04-16 04:45:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][490/1251] eta 0:03:36 lr 0.000539 time 0.2926 (0.2850) loss 3.5106 (3.5586) grad_norm 1.8631 (1.4849) [2021-04-16 04:45:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][500/1251] eta 0:03:34 lr 0.000539 time 0.2565 (0.2850) loss 3.9776 (3.5620) grad_norm 1.3228 (1.4850) [2021-04-16 04:45:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][510/1251] eta 0:03:31 lr 0.000539 time 0.2802 (0.2849) loss 3.7570 (3.5565) grad_norm 1.4127 (1.4874) [2021-04-16 04:45:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][520/1251] eta 0:03:28 lr 0.000539 time 0.3001 (0.2849) loss 2.7136 (3.5527) grad_norm 1.4657 (1.4871) [2021-04-16 04:45:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][530/1251] eta 0:03:25 lr 0.000539 time 0.2748 (0.2847) loss 4.0629 (3.5537) grad_norm 1.3285 (1.4861) [2021-04-16 04:45:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][540/1251] eta 0:03:22 lr 0.000539 time 0.2824 (0.2845) loss 4.0387 (3.5596) grad_norm 1.5463 (1.4867) [2021-04-16 04:45:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][550/1251] eta 0:03:19 lr 0.000539 time 0.2971 (0.2845) loss 3.5352 (3.5641) grad_norm 1.8464 (1.4905) [2021-04-16 04:45:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][560/1251] eta 0:03:16 lr 0.000539 time 0.2780 (0.2843) loss 3.6788 (3.5624) grad_norm 1.4396 (1.4921) [2021-04-16 04:45:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][570/1251] eta 0:03:13 lr 0.000539 time 0.2746 (0.2842) loss 3.7955 (3.5611) grad_norm 1.3338 (1.4923) [2021-04-16 04:45:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][580/1251] eta 0:03:11 lr 0.000539 time 0.2841 (0.2847) loss 4.1769 (3.5583) grad_norm 1.2995 (1.4908) [2021-04-16 04:45:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][590/1251] eta 0:03:08 lr 0.000539 time 0.2666 (0.2849) loss 3.6109 (3.5569) grad_norm 1.4102 (1.4886) [2021-04-16 04:46:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][600/1251] eta 0:03:05 lr 0.000539 time 0.2790 (0.2847) loss 3.1058 (3.5574) grad_norm 1.3932 (1.4866) [2021-04-16 04:46:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][610/1251] eta 0:03:02 lr 0.000539 time 0.2679 (0.2847) loss 4.4052 (3.5591) grad_norm 1.5318 (1.4861) [2021-04-16 04:46:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][620/1251] eta 0:02:59 lr 0.000539 time 0.2825 (0.2846) loss 2.4060 (3.5583) grad_norm 1.7392 (1.4863) [2021-04-16 04:46:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][630/1251] eta 0:02:56 lr 0.000539 time 0.2840 (0.2847) loss 3.1518 (3.5532) grad_norm 1.5776 (1.4860) [2021-04-16 04:46:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][640/1251] eta 0:02:53 lr 0.000539 time 0.2798 (0.2845) loss 3.6556 (3.5520) grad_norm 1.7594 (1.4862) [2021-04-16 04:46:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][650/1251] eta 0:02:50 lr 0.000539 time 0.2951 (0.2845) loss 4.1068 (3.5511) grad_norm 1.2940 (1.4860) [2021-04-16 04:46:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][660/1251] eta 0:02:48 lr 0.000539 time 0.3068 (0.2845) loss 4.0571 (3.5508) grad_norm 1.4818 (1.4861) [2021-04-16 04:46:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][670/1251] eta 0:02:45 lr 0.000538 time 0.2760 (0.2844) loss 3.4484 (3.5482) grad_norm 1.4745 (1.4856) [2021-04-16 04:46:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][680/1251] eta 0:02:42 lr 0.000538 time 0.2772 (0.2845) loss 3.5098 (3.5477) grad_norm 1.5816 (1.4868) [2021-04-16 04:46:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][690/1251] eta 0:02:39 lr 0.000538 time 0.2936 (0.2845) loss 3.8062 (3.5506) grad_norm 1.8897 (1.4901) [2021-04-16 04:46:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][700/1251] eta 0:02:36 lr 0.000538 time 0.2558 (0.2844) loss 3.9010 (3.5533) grad_norm 1.4027 (1.4901) [2021-04-16 04:46:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][710/1251] eta 0:02:33 lr 0.000538 time 0.2620 (0.2845) loss 4.2383 (3.5515) grad_norm 1.3683 (1.4899) [2021-04-16 04:46:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][720/1251] eta 0:02:31 lr 0.000538 time 0.2664 (0.2845) loss 4.3541 (3.5564) grad_norm 1.4202 (1.4895) [2021-04-16 04:46:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][730/1251] eta 0:02:28 lr 0.000538 time 0.2653 (0.2846) loss 4.1717 (3.5548) grad_norm 1.4555 (1.4891) [2021-04-16 04:46:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][740/1251] eta 0:02:25 lr 0.000538 time 0.3091 (0.2845) loss 3.6388 (3.5505) grad_norm 1.5115 (1.4887) [2021-04-16 04:46:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][750/1251] eta 0:02:22 lr 0.000538 time 0.2829 (0.2844) loss 3.9928 (3.5536) grad_norm 1.3770 (1.4885) [2021-04-16 04:46:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][760/1251] eta 0:02:19 lr 0.000538 time 0.2902 (0.2843) loss 3.4700 (3.5524) grad_norm 1.4829 (1.4891) [2021-04-16 04:46:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][770/1251] eta 0:02:16 lr 0.000538 time 0.2645 (0.2842) loss 3.9948 (3.5527) grad_norm 1.7603 (1.4906) [2021-04-16 04:46:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][780/1251] eta 0:02:13 lr 0.000538 time 0.2693 (0.2841) loss 2.5949 (3.5467) grad_norm 1.6368 (1.4901) [2021-04-16 04:46:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][790/1251] eta 0:02:10 lr 0.000538 time 0.2781 (0.2840) loss 3.9005 (3.5504) grad_norm 1.3803 (1.4909) [2021-04-16 04:46:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][800/1251] eta 0:02:08 lr 0.000538 time 0.2862 (0.2840) loss 3.2475 (3.5514) grad_norm 1.3009 (1.4900) [2021-04-16 04:46:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][810/1251] eta 0:02:05 lr 0.000538 time 0.2835 (0.2839) loss 3.5726 (3.5532) grad_norm 1.3231 (1.4886) [2021-04-16 04:47:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][820/1251] eta 0:02:02 lr 0.000538 time 0.2721 (0.2839) loss 4.2744 (3.5522) grad_norm 1.2442 (1.4879) [2021-04-16 04:47:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][830/1251] eta 0:01:59 lr 0.000538 time 0.2549 (0.2838) loss 3.7390 (3.5526) grad_norm 1.2823 (1.4882) [2021-04-16 04:47:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][840/1251] eta 0:01:56 lr 0.000538 time 0.2756 (0.2837) loss 4.6941 (3.5513) grad_norm 1.6816 (1.4886) [2021-04-16 04:47:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][850/1251] eta 0:01:53 lr 0.000538 time 0.2732 (0.2836) loss 3.6406 (3.5523) grad_norm 1.4688 (1.4891) [2021-04-16 04:47:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][860/1251] eta 0:01:50 lr 0.000538 time 0.3062 (0.2836) loss 3.8000 (3.5557) grad_norm 1.5425 (1.4885) [2021-04-16 04:47:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][870/1251] eta 0:01:48 lr 0.000538 time 0.2676 (0.2836) loss 2.9107 (3.5547) grad_norm 1.4996 (1.4879) [2021-04-16 04:47:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][880/1251] eta 0:01:45 lr 0.000538 time 0.2690 (0.2835) loss 3.6379 (3.5551) grad_norm 1.3468 (1.4869) [2021-04-16 04:47:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][890/1251] eta 0:01:42 lr 0.000538 time 0.2853 (0.2835) loss 2.5081 (3.5519) grad_norm 1.3008 (1.4863) [2021-04-16 04:47:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][900/1251] eta 0:01:39 lr 0.000538 time 0.2747 (0.2834) loss 3.7835 (3.5505) grad_norm 1.4499 (1.4869) [2021-04-16 04:47:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][910/1251] eta 0:01:36 lr 0.000537 time 0.2785 (0.2835) loss 2.2812 (3.5526) grad_norm 1.6288 (1.4870) [2021-04-16 04:47:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][920/1251] eta 0:01:33 lr 0.000537 time 0.2655 (0.2836) loss 2.8186 (3.5494) grad_norm 1.5280 (1.4880) [2021-04-16 04:47:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][930/1251] eta 0:01:31 lr 0.000537 time 0.2976 (0.2835) loss 4.2904 (3.5521) grad_norm 1.8098 (1.4893) [2021-04-16 04:47:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][940/1251] eta 0:01:28 lr 0.000537 time 0.2906 (0.2836) loss 3.7163 (3.5524) grad_norm 1.6375 (1.4896) [2021-04-16 04:47:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][950/1251] eta 0:01:25 lr 0.000537 time 0.2679 (0.2837) loss 4.0657 (3.5506) grad_norm 1.4309 (nan) [2021-04-16 04:47:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][960/1251] eta 0:01:22 lr 0.000537 time 0.3123 (0.2838) loss 3.1243 (3.5498) grad_norm 1.4730 (nan) [2021-04-16 04:47:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][970/1251] eta 0:01:19 lr 0.000537 time 0.2752 (0.2837) loss 4.0849 (3.5518) grad_norm 1.4763 (nan) [2021-04-16 04:47:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][980/1251] eta 0:01:16 lr 0.000537 time 0.2522 (0.2836) loss 3.4451 (3.5546) grad_norm 1.7552 (nan) [2021-04-16 04:47:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][990/1251] eta 0:01:14 lr 0.000537 time 0.3068 (0.2836) loss 3.4206 (3.5552) grad_norm 1.2907 (nan) [2021-04-16 04:47:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][1000/1251] eta 0:01:11 lr 0.000537 time 0.2798 (0.2835) loss 4.0144 (3.5553) grad_norm 1.4059 (nan) [2021-04-16 04:47:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][1010/1251] eta 0:01:08 lr 0.000537 time 0.2670 (0.2834) loss 3.6999 (3.5564) grad_norm 1.4878 (nan) [2021-04-16 04:47:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][1020/1251] eta 0:01:05 lr 0.000537 time 0.2541 (0.2834) loss 3.2947 (3.5565) grad_norm 1.5929 (nan) [2021-04-16 04:48:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][1030/1251] eta 0:01:02 lr 0.000537 time 0.2691 (0.2833) loss 3.9276 (3.5556) grad_norm 1.2583 (nan) [2021-04-16 04:48:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][1040/1251] eta 0:00:59 lr 0.000537 time 0.2614 (0.2832) loss 3.4484 (3.5562) grad_norm 1.8774 (nan) [2021-04-16 04:48:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][1050/1251] eta 0:00:56 lr 0.000537 time 0.2625 (0.2832) loss 3.4667 (3.5550) grad_norm 1.8092 (nan) [2021-04-16 04:48:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][1060/1251] eta 0:00:54 lr 0.000537 time 0.2584 (0.2832) loss 3.7430 (3.5531) grad_norm 1.5730 (nan) [2021-04-16 04:48:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][1070/1251] eta 0:00:51 lr 0.000537 time 0.2863 (0.2831) loss 2.8696 (3.5548) grad_norm 1.3885 (nan) [2021-04-16 04:48:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][1080/1251] eta 0:00:48 lr 0.000537 time 0.2575 (0.2831) loss 2.3394 (3.5508) grad_norm 1.4439 (nan) [2021-04-16 04:48:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][1090/1251] eta 0:00:45 lr 0.000537 time 0.2928 (0.2830) loss 3.8383 (3.5494) grad_norm 1.3266 (nan) [2021-04-16 04:48:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][1100/1251] eta 0:00:42 lr 0.000537 time 0.2687 (0.2830) loss 3.6384 (3.5516) grad_norm 1.2808 (nan) [2021-04-16 04:48:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][1110/1251] eta 0:00:39 lr 0.000537 time 0.2743 (0.2830) loss 4.0619 (3.5535) grad_norm 1.5998 (nan) [2021-04-16 04:48:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][1120/1251] eta 0:00:37 lr 0.000537 time 0.2960 (0.2829) loss 2.9693 (3.5520) grad_norm 1.4348 (nan) [2021-04-16 04:48:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][1130/1251] eta 0:00:34 lr 0.000537 time 0.2547 (0.2829) loss 4.5059 (3.5519) grad_norm 1.4951 (nan) [2021-04-16 04:48:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][1140/1251] eta 0:00:31 lr 0.000537 time 0.2803 (0.2829) loss 3.7805 (3.5502) grad_norm 1.3825 (nan) [2021-04-16 04:48:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][1150/1251] eta 0:00:28 lr 0.000536 time 0.3064 (0.2828) loss 3.8991 (3.5529) grad_norm 1.3905 (nan) [2021-04-16 04:48:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][1160/1251] eta 0:00:25 lr 0.000536 time 0.2817 (0.2830) loss 3.7251 (3.5557) grad_norm 1.3355 (nan) [2021-04-16 04:48:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][1170/1251] eta 0:00:22 lr 0.000536 time 0.2767 (0.2831) loss 2.7585 (3.5545) grad_norm 1.7242 (nan) [2021-04-16 04:48:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][1180/1251] eta 0:00:20 lr 0.000536 time 0.2750 (0.2832) loss 3.7811 (3.5521) grad_norm 1.3142 (nan) [2021-04-16 04:48:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][1190/1251] eta 0:00:17 lr 0.000536 time 0.2807 (0.2831) loss 2.7515 (3.5501) grad_norm 1.2422 (nan) [2021-04-16 04:48:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][1200/1251] eta 0:00:14 lr 0.000536 time 0.2426 (0.2831) loss 3.5066 (3.5509) grad_norm 1.6740 (nan) [2021-04-16 04:48:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][1210/1251] eta 0:00:11 lr 0.000536 time 0.2674 (0.2832) loss 3.8047 (3.5513) grad_norm 1.9515 (nan) [2021-04-16 04:48:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][1220/1251] eta 0:00:08 lr 0.000536 time 0.2860 (0.2832) loss 3.8645 (3.5515) grad_norm 1.9053 (nan) [2021-04-16 04:48:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][1230/1251] eta 0:00:05 lr 0.000536 time 0.2860 (0.2832) loss 3.3319 (3.5512) grad_norm 1.5011 (nan) [2021-04-16 04:49:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][1240/1251] eta 0:00:03 lr 0.000536 time 0.2489 (0.2831) loss 3.4940 (3.5486) grad_norm 1.4425 (nan) [2021-04-16 04:49:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [143/300][1250/1251] eta 0:00:00 lr 0.000536 time 0.2484 (0.2828) loss 2.8123 (3.5476) grad_norm 1.5320 (nan) [2021-04-16 04:49:06 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 143 training takes 0:05:57 [2021-04-16 04:49:06 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_143.pth saving...... [2021-04-16 04:49:21 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_143.pth saved !!! [2021-04-16 04:49:22 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.129 (1.129) Loss 1.0149 (1.0149) Acc@1 75.781 (75.781) Acc@5 93.262 (93.262) [2021-04-16 04:49:23 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.137 (0.226) Loss 1.0679 (1.0401) Acc@1 74.023 (75.932) Acc@5 94.043 (93.191) [2021-04-16 04:49:26 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.114 (0.243) Loss 0.9383 (1.0260) Acc@1 77.832 (76.028) Acc@5 94.922 (93.304) [2021-04-16 04:49:27 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.195 (0.214) Loss 0.9964 (1.0226) Acc@1 76.660 (76.156) Acc@5 93.066 (93.300) [2021-04-16 04:49:30 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.219) Loss 1.0422 (1.0272) Acc@1 76.270 (76.119) Acc@5 93.066 (93.228) [2021-04-16 04:49:35 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 75.936 Acc@5 93.176 [2021-04-16 04:49:35 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 75.9% [2021-04-16 04:49:35 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 75.94% [2021-04-16 04:49:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][0/1251] eta 2:15:14 lr 0.000536 time 6.4861 (6.4861) loss 4.0791 (4.0791) grad_norm 1.3265 (1.3265) [2021-04-16 04:49:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][10/1251] eta 0:17:19 lr 0.000536 time 0.2952 (0.8378) loss 3.5706 (3.5122) grad_norm 1.6739 (1.4317) [2021-04-16 04:49:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][20/1251] eta 0:11:43 lr 0.000536 time 0.2807 (0.5713) loss 3.7479 (3.6313) grad_norm 1.3024 (1.4128) [2021-04-16 04:49:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][30/1251] eta 0:09:40 lr 0.000536 time 0.2676 (0.4756) loss 3.4278 (3.5960) grad_norm 1.3986 (1.4331) [2021-04-16 04:49:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][40/1251] eta 0:08:36 lr 0.000536 time 0.2822 (0.4268) loss 3.7561 (3.5556) grad_norm 1.4350 (1.4528) [2021-04-16 04:49:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][50/1251] eta 0:07:57 lr 0.000536 time 0.2927 (0.3978) loss 3.0526 (3.4905) grad_norm 1.5275 (1.4546) [2021-04-16 04:49:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][60/1251] eta 0:07:30 lr 0.000536 time 0.2920 (0.3786) loss 4.0381 (3.4480) grad_norm 1.6281 (1.4507) [2021-04-16 04:50:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][70/1251] eta 0:07:11 lr 0.000536 time 0.2799 (0.3650) loss 2.6448 (3.4480) grad_norm 1.5914 (1.4658) [2021-04-16 04:50:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][80/1251] eta 0:06:56 lr 0.000536 time 0.2886 (0.3560) loss 2.7908 (3.4534) grad_norm 1.4629 (1.4652) [2021-04-16 04:50:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][90/1251] eta 0:06:43 lr 0.000536 time 0.2817 (0.3477) loss 2.8725 (3.4737) grad_norm 1.4761 (1.4737) [2021-04-16 04:50:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][100/1251] eta 0:06:33 lr 0.000536 time 0.2823 (0.3419) loss 4.0229 (3.4838) grad_norm 1.4971 (1.4786) [2021-04-16 04:50:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][110/1251] eta 0:06:23 lr 0.000536 time 0.3047 (0.3359) loss 3.8208 (3.5053) grad_norm 1.3267 (1.4940) [2021-04-16 04:50:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][120/1251] eta 0:06:15 lr 0.000536 time 0.3154 (0.3321) loss 4.2142 (3.5227) grad_norm 1.5946 (1.4978) [2021-04-16 04:50:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][130/1251] eta 0:06:07 lr 0.000536 time 0.2741 (0.3279) loss 3.6633 (3.5215) grad_norm 1.5343 (1.5006) [2021-04-16 04:50:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][140/1251] eta 0:06:01 lr 0.000536 time 0.2883 (0.3252) loss 4.3233 (3.5206) grad_norm 1.4608 (1.4993) [2021-04-16 04:50:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][150/1251] eta 0:05:54 lr 0.000535 time 0.2740 (0.3221) loss 3.8839 (3.5211) grad_norm 1.5879 (1.5034) [2021-04-16 04:50:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][160/1251] eta 0:05:49 lr 0.000535 time 0.2687 (0.3204) loss 2.4624 (3.5161) grad_norm 1.4710 (1.5070) [2021-04-16 04:50:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][170/1251] eta 0:05:44 lr 0.000535 time 0.2674 (0.3187) loss 3.9782 (3.5170) grad_norm 1.4778 (1.5066) [2021-04-16 04:50:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][180/1251] eta 0:05:39 lr 0.000535 time 0.2730 (0.3168) loss 3.8383 (3.5230) grad_norm 1.4085 (1.5101) [2021-04-16 04:50:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][190/1251] eta 0:05:33 lr 0.000535 time 0.2636 (0.3147) loss 4.1331 (3.5351) grad_norm 1.4452 (1.5085) [2021-04-16 04:50:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][200/1251] eta 0:05:28 lr 0.000535 time 0.2851 (0.3129) loss 4.1579 (3.5337) grad_norm 1.4667 (1.5102) [2021-04-16 04:50:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][210/1251] eta 0:05:23 lr 0.000535 time 0.2886 (0.3112) loss 3.5673 (3.5363) grad_norm 1.5409 (1.5143) [2021-04-16 04:50:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][220/1251] eta 0:05:19 lr 0.000535 time 0.2681 (0.3095) loss 3.9796 (3.5456) grad_norm 1.2401 (1.5102) [2021-04-16 04:50:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][230/1251] eta 0:05:14 lr 0.000535 time 0.2773 (0.3080) loss 3.3841 (3.5485) grad_norm 1.5404 (1.5075) [2021-04-16 04:50:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][240/1251] eta 0:05:10 lr 0.000535 time 0.3222 (0.3068) loss 3.8136 (3.5481) grad_norm 1.4124 (1.5046) [2021-04-16 04:50:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][250/1251] eta 0:05:05 lr 0.000535 time 0.2534 (0.3056) loss 3.6398 (3.5446) grad_norm 1.5878 (1.5083) [2021-04-16 04:50:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][260/1251] eta 0:05:01 lr 0.000535 time 0.2415 (0.3045) loss 3.6997 (3.5404) grad_norm 1.7659 (1.5088) [2021-04-16 04:50:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][270/1251] eta 0:04:57 lr 0.000535 time 0.2745 (0.3035) loss 3.1049 (3.5487) grad_norm 1.3878 (1.5111) [2021-04-16 04:51:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][280/1251] eta 0:04:53 lr 0.000535 time 0.2716 (0.3026) loss 3.4135 (3.5487) grad_norm 1.4112 (1.5115) [2021-04-16 04:51:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][290/1251] eta 0:04:49 lr 0.000535 time 0.2723 (0.3017) loss 2.7972 (3.5442) grad_norm 1.5345 (1.5125) [2021-04-16 04:51:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][300/1251] eta 0:04:46 lr 0.000535 time 0.2723 (0.3010) loss 3.6836 (3.5481) grad_norm 1.5963 (1.5085) [2021-04-16 04:51:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][310/1251] eta 0:04:42 lr 0.000535 time 0.2482 (0.3003) loss 3.9205 (3.5453) grad_norm 1.3677 (1.5074) [2021-04-16 04:51:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][320/1251] eta 0:04:38 lr 0.000535 time 0.2841 (0.2995) loss 3.2653 (3.5544) grad_norm 1.6864 (1.5074) [2021-04-16 04:51:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][330/1251] eta 0:04:35 lr 0.000535 time 0.2770 (0.2990) loss 3.5936 (3.5622) grad_norm 1.5416 (1.5034) [2021-04-16 04:51:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][340/1251] eta 0:04:31 lr 0.000535 time 0.2677 (0.2982) loss 3.1123 (3.5633) grad_norm 1.2878 (1.5033) [2021-04-16 04:51:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][350/1251] eta 0:04:28 lr 0.000535 time 0.2867 (0.2977) loss 3.4512 (3.5621) grad_norm 1.2830 (1.4993) [2021-04-16 04:51:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][360/1251] eta 0:04:24 lr 0.000535 time 0.2758 (0.2974) loss 3.8465 (3.5669) grad_norm 1.3237 (1.4964) [2021-04-16 04:51:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][370/1251] eta 0:04:22 lr 0.000535 time 0.2857 (0.2980) loss 4.0737 (3.5695) grad_norm 1.2694 (1.4966) [2021-04-16 04:51:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][380/1251] eta 0:04:19 lr 0.000535 time 0.2740 (0.2974) loss 2.7567 (3.5714) grad_norm 1.3890 (1.4949) [2021-04-16 04:51:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][390/1251] eta 0:04:15 lr 0.000534 time 0.2740 (0.2969) loss 2.9705 (3.5739) grad_norm 1.2827 (1.4937) [2021-04-16 04:51:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][400/1251] eta 0:04:12 lr 0.000534 time 0.3028 (0.2964) loss 3.5643 (3.5780) grad_norm 1.7794 (1.4937) [2021-04-16 04:51:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][410/1251] eta 0:04:08 lr 0.000534 time 0.2716 (0.2959) loss 3.6562 (3.5759) grad_norm 1.5285 (1.4952) [2021-04-16 04:51:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][420/1251] eta 0:04:05 lr 0.000534 time 0.2510 (0.2954) loss 2.5840 (3.5694) grad_norm 1.6115 (1.4955) [2021-04-16 04:51:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][430/1251] eta 0:04:02 lr 0.000534 time 0.2979 (0.2950) loss 2.8087 (3.5700) grad_norm 1.6057 (1.4947) [2021-04-16 04:51:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][440/1251] eta 0:03:59 lr 0.000534 time 0.3068 (0.2947) loss 4.4494 (3.5775) grad_norm 1.2648 (1.4929) [2021-04-16 04:51:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][450/1251] eta 0:03:55 lr 0.000534 time 0.2680 (0.2943) loss 2.4884 (3.5671) grad_norm 1.4682 (1.4925) [2021-04-16 04:51:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][460/1251] eta 0:03:52 lr 0.000534 time 0.2689 (0.2941) loss 4.2293 (3.5711) grad_norm 1.2234 (1.4898) [2021-04-16 04:51:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][470/1251] eta 0:03:49 lr 0.000534 time 0.2831 (0.2942) loss 3.5425 (3.5734) grad_norm 1.6394 (1.4891) [2021-04-16 04:51:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][480/1251] eta 0:03:46 lr 0.000534 time 0.2786 (0.2939) loss 2.9416 (3.5748) grad_norm 1.6157 (1.4882) [2021-04-16 04:51:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][490/1251] eta 0:03:43 lr 0.000534 time 0.2992 (0.2936) loss 3.8486 (3.5713) grad_norm 1.2771 (1.4882) [2021-04-16 04:52:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][500/1251] eta 0:03:40 lr 0.000534 time 0.3005 (0.2934) loss 3.9051 (3.5691) grad_norm 1.3088 (1.4860) [2021-04-16 04:52:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][510/1251] eta 0:03:37 lr 0.000534 time 0.2555 (0.2931) loss 3.7717 (3.5638) grad_norm 1.5800 (1.4871) [2021-04-16 04:52:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][520/1251] eta 0:03:34 lr 0.000534 time 0.2837 (0.2928) loss 2.7692 (3.5651) grad_norm 1.4079 (1.4859) [2021-04-16 04:52:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][530/1251] eta 0:03:30 lr 0.000534 time 0.2698 (0.2924) loss 4.1420 (3.5614) grad_norm 1.4063 (1.4842) [2021-04-16 04:52:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][540/1251] eta 0:03:27 lr 0.000534 time 0.2974 (0.2924) loss 2.9893 (3.5610) grad_norm 1.4983 (1.4834) [2021-04-16 04:52:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][550/1251] eta 0:03:24 lr 0.000534 time 0.2723 (0.2921) loss 4.0251 (3.5599) grad_norm 1.3677 (1.4824) [2021-04-16 04:52:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][560/1251] eta 0:03:21 lr 0.000534 time 0.2896 (0.2919) loss 4.5415 (3.5657) grad_norm 1.5701 (1.4818) [2021-04-16 04:52:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][570/1251] eta 0:03:18 lr 0.000534 time 0.2663 (0.2917) loss 3.9224 (3.5671) grad_norm 1.4029 (1.4830) [2021-04-16 04:52:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][580/1251] eta 0:03:15 lr 0.000534 time 0.2829 (0.2917) loss 3.6633 (3.5689) grad_norm 1.7638 (1.4829) [2021-04-16 04:52:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][590/1251] eta 0:03:12 lr 0.000534 time 0.2791 (0.2915) loss 2.6877 (3.5658) grad_norm 1.5835 (1.4840) [2021-04-16 04:52:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][600/1251] eta 0:03:09 lr 0.000534 time 0.2523 (0.2912) loss 3.2623 (3.5678) grad_norm 1.3771 (1.4826) [2021-04-16 04:52:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][610/1251] eta 0:03:06 lr 0.000534 time 0.2820 (0.2911) loss 2.3123 (3.5656) grad_norm 1.8711 (1.4832) [2021-04-16 04:52:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][620/1251] eta 0:03:03 lr 0.000534 time 0.2746 (0.2908) loss 3.5125 (3.5676) grad_norm 1.2976 (1.4840) [2021-04-16 04:52:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][630/1251] eta 0:03:00 lr 0.000533 time 0.2794 (0.2907) loss 3.9453 (3.5669) grad_norm 1.3989 (1.4851) [2021-04-16 04:52:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][640/1251] eta 0:02:57 lr 0.000533 time 0.3076 (0.2905) loss 3.7226 (3.5664) grad_norm 1.5527 (1.4854) [2021-04-16 04:52:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][650/1251] eta 0:02:54 lr 0.000533 time 0.4279 (0.2906) loss 3.7772 (3.5673) grad_norm 1.4774 (1.4842) [2021-04-16 04:52:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][660/1251] eta 0:02:51 lr 0.000533 time 0.2758 (0.2904) loss 4.6924 (3.5649) grad_norm 1.7973 (1.4871) [2021-04-16 04:52:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][670/1251] eta 0:02:48 lr 0.000533 time 0.2937 (0.2903) loss 3.6081 (3.5673) grad_norm 1.4056 (1.4867) [2021-04-16 04:52:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][680/1251] eta 0:02:45 lr 0.000533 time 0.2864 (0.2904) loss 2.3853 (3.5654) grad_norm 1.2392 (1.4867) [2021-04-16 04:52:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][690/1251] eta 0:02:42 lr 0.000533 time 0.2795 (0.2904) loss 2.8474 (3.5645) grad_norm 1.3960 (1.4872) [2021-04-16 04:52:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][700/1251] eta 0:02:39 lr 0.000533 time 0.2704 (0.2902) loss 3.5064 (3.5640) grad_norm 1.4776 (1.4874) [2021-04-16 04:53:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][710/1251] eta 0:02:36 lr 0.000533 time 0.2759 (0.2900) loss 3.9329 (3.5629) grad_norm 1.4466 (1.4861) [2021-04-16 04:53:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][720/1251] eta 0:02:33 lr 0.000533 time 0.2755 (0.2899) loss 3.3956 (3.5662) grad_norm 1.2514 (1.4859) [2021-04-16 04:53:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][730/1251] eta 0:02:31 lr 0.000533 time 0.3992 (0.2899) loss 2.7309 (3.5652) grad_norm 1.4583 (1.4860) [2021-04-16 04:53:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][740/1251] eta 0:02:28 lr 0.000533 time 0.2589 (0.2900) loss 4.1280 (3.5675) grad_norm 1.4810 (1.4870) [2021-04-16 04:53:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][750/1251] eta 0:02:25 lr 0.000533 time 0.2889 (0.2898) loss 3.1622 (3.5685) grad_norm 1.4191 (1.4872) [2021-04-16 04:53:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][760/1251] eta 0:02:22 lr 0.000533 time 0.2669 (0.2896) loss 3.6390 (3.5712) grad_norm 1.5747 (1.4877) [2021-04-16 04:53:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][770/1251] eta 0:02:19 lr 0.000533 time 0.2907 (0.2894) loss 3.9571 (3.5749) grad_norm 1.5110 (1.4891) [2021-04-16 04:53:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][780/1251] eta 0:02:16 lr 0.000533 time 0.3029 (0.2893) loss 3.7808 (3.5731) grad_norm 1.4444 (1.4887) [2021-04-16 04:53:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][790/1251] eta 0:02:13 lr 0.000533 time 0.2756 (0.2891) loss 4.2718 (3.5763) grad_norm 1.7532 (1.4893) [2021-04-16 04:53:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][800/1251] eta 0:02:10 lr 0.000533 time 0.2905 (0.2890) loss 3.8100 (3.5750) grad_norm 1.6495 (1.4896) [2021-04-16 04:53:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][810/1251] eta 0:02:07 lr 0.000533 time 0.2900 (0.2888) loss 4.1524 (3.5740) grad_norm 1.3481 (1.4894) [2021-04-16 04:53:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][820/1251] eta 0:02:04 lr 0.000533 time 0.2682 (0.2886) loss 4.2546 (3.5751) grad_norm 1.5442 (1.4902) [2021-04-16 04:53:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][830/1251] eta 0:02:01 lr 0.000533 time 0.2657 (0.2884) loss 4.0816 (3.5767) grad_norm 1.3976 (1.4898) [2021-04-16 04:53:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][840/1251] eta 0:01:58 lr 0.000533 time 0.2640 (0.2882) loss 2.9032 (3.5761) grad_norm 1.6612 (1.4895) [2021-04-16 04:53:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][850/1251] eta 0:01:55 lr 0.000533 time 0.2707 (0.2881) loss 4.0815 (3.5772) grad_norm 1.2363 (1.4893) [2021-04-16 04:53:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][860/1251] eta 0:01:52 lr 0.000533 time 0.2781 (0.2879) loss 4.1548 (3.5825) grad_norm 1.5848 (1.4887) [2021-04-16 04:53:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][870/1251] eta 0:01:49 lr 0.000532 time 0.2989 (0.2878) loss 2.5130 (3.5808) grad_norm 1.4226 (1.4888) [2021-04-16 04:53:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][880/1251] eta 0:01:46 lr 0.000532 time 0.2718 (0.2876) loss 2.4924 (3.5802) grad_norm 1.3773 (1.4896) [2021-04-16 04:53:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][890/1251] eta 0:01:43 lr 0.000532 time 0.2726 (0.2876) loss 3.6430 (3.5809) grad_norm 1.5670 (1.4901) [2021-04-16 04:53:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][900/1251] eta 0:01:40 lr 0.000532 time 0.2906 (0.2877) loss 3.5396 (3.5851) grad_norm 1.3573 (1.4898) [2021-04-16 04:53:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][910/1251] eta 0:01:38 lr 0.000532 time 0.3044 (0.2876) loss 3.6692 (3.5870) grad_norm 1.5254 (1.4903) [2021-04-16 04:54:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][920/1251] eta 0:01:35 lr 0.000532 time 0.2723 (0.2876) loss 3.4806 (3.5876) grad_norm 1.5774 (1.4900) [2021-04-16 04:54:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][930/1251] eta 0:01:32 lr 0.000532 time 0.2866 (0.2876) loss 4.1204 (3.5899) grad_norm 1.4045 (1.4893) [2021-04-16 04:54:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][940/1251] eta 0:01:29 lr 0.000532 time 0.3795 (0.2878) loss 3.1012 (3.5901) grad_norm 1.3604 (1.4891) [2021-04-16 04:54:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][950/1251] eta 0:01:26 lr 0.000532 time 0.2786 (0.2877) loss 2.3549 (3.5864) grad_norm 1.4247 (1.4889) [2021-04-16 04:54:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][960/1251] eta 0:01:23 lr 0.000532 time 0.2916 (0.2877) loss 3.5307 (3.5850) grad_norm 1.4139 (1.4887) [2021-04-16 04:54:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][970/1251] eta 0:01:20 lr 0.000532 time 0.2633 (0.2875) loss 3.9296 (3.5851) grad_norm 1.4626 (1.4885) [2021-04-16 04:54:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][980/1251] eta 0:01:17 lr 0.000532 time 0.3177 (0.2875) loss 3.5292 (3.5863) grad_norm 1.3956 (1.4884) [2021-04-16 04:54:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][990/1251] eta 0:01:15 lr 0.000532 time 0.3010 (0.2874) loss 2.5542 (3.5872) grad_norm 1.3469 (1.4882) [2021-04-16 04:54:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][1000/1251] eta 0:01:12 lr 0.000532 time 0.2745 (0.2872) loss 3.4943 (3.5863) grad_norm 2.1210 (1.4896) [2021-04-16 04:54:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][1010/1251] eta 0:01:09 lr 0.000532 time 0.2878 (0.2872) loss 4.3188 (3.5880) grad_norm 1.4119 (1.4900) [2021-04-16 04:54:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][1020/1251] eta 0:01:06 lr 0.000532 time 0.3072 (0.2871) loss 4.0563 (3.5909) grad_norm 1.7048 (1.4909) [2021-04-16 04:54:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][1030/1251] eta 0:01:03 lr 0.000532 time 0.2820 (0.2870) loss 3.0259 (3.5926) grad_norm 1.4561 (1.4915) [2021-04-16 04:54:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][1040/1251] eta 0:01:00 lr 0.000532 time 0.2657 (0.2871) loss 3.0672 (3.5922) grad_norm 1.5195 (1.4930) [2021-04-16 04:54:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][1050/1251] eta 0:00:57 lr 0.000532 time 0.3045 (0.2870) loss 3.4834 (3.5922) grad_norm 1.5169 (1.4934) [2021-04-16 04:54:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][1060/1251] eta 0:00:54 lr 0.000532 time 0.2811 (0.2869) loss 3.7170 (3.5924) grad_norm 1.3540 (1.4929) [2021-04-16 04:54:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][1070/1251] eta 0:00:51 lr 0.000532 time 0.2957 (0.2868) loss 4.5045 (3.5960) grad_norm 1.5429 (1.4930) [2021-04-16 04:54:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][1080/1251] eta 0:00:49 lr 0.000532 time 0.2889 (0.2867) loss 3.5722 (3.5925) grad_norm 1.5798 (1.4937) [2021-04-16 04:54:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][1090/1251] eta 0:00:46 lr 0.000532 time 0.2663 (0.2866) loss 3.1640 (3.5905) grad_norm 1.5568 (1.4936) [2021-04-16 04:54:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][1100/1251] eta 0:00:43 lr 0.000532 time 0.2826 (0.2866) loss 3.9750 (3.5912) grad_norm 1.5393 (1.4938) [2021-04-16 04:54:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][1110/1251] eta 0:00:40 lr 0.000531 time 0.2649 (0.2865) loss 2.4841 (3.5894) grad_norm 1.4609 (1.4945) [2021-04-16 04:54:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][1120/1251] eta 0:00:37 lr 0.000531 time 0.2636 (0.2865) loss 3.8257 (3.5882) grad_norm 1.9193 (1.4958) [2021-04-16 04:54:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][1130/1251] eta 0:00:34 lr 0.000531 time 0.2825 (0.2865) loss 3.6642 (3.5874) grad_norm 1.7407 (1.4960) [2021-04-16 04:55:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][1140/1251] eta 0:00:31 lr 0.000531 time 0.2818 (0.2864) loss 3.0810 (3.5841) grad_norm 1.4475 (1.4960) [2021-04-16 04:55:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][1150/1251] eta 0:00:28 lr 0.000531 time 0.2908 (0.2863) loss 2.4560 (3.5824) grad_norm 1.3162 (1.4970) [2021-04-16 04:55:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][1160/1251] eta 0:00:26 lr 0.000531 time 0.2688 (0.2863) loss 4.2022 (3.5829) grad_norm 1.6932 (1.4977) [2021-04-16 04:55:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][1170/1251] eta 0:00:23 lr 0.000531 time 0.3002 (0.2863) loss 3.3607 (3.5837) grad_norm 1.4875 (1.4975) [2021-04-16 04:55:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][1180/1251] eta 0:00:20 lr 0.000531 time 0.2520 (0.2862) loss 4.0845 (3.5835) grad_norm 1.5186 (1.4971) [2021-04-16 04:55:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][1190/1251] eta 0:00:17 lr 0.000531 time 0.2552 (0.2862) loss 3.6540 (3.5858) grad_norm 1.3567 (1.4966) [2021-04-16 04:55:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][1200/1251] eta 0:00:14 lr 0.000531 time 0.2545 (0.2861) loss 3.6740 (3.5848) grad_norm 1.5436 (1.4965) [2021-04-16 04:55:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][1210/1251] eta 0:00:11 lr 0.000531 time 0.2781 (0.2860) loss 3.3474 (3.5834) grad_norm 1.4800 (1.4959) [2021-04-16 04:55:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][1220/1251] eta 0:00:08 lr 0.000531 time 0.2827 (0.2859) loss 3.7065 (3.5814) grad_norm 1.2610 (1.4953) [2021-04-16 04:55:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][1230/1251] eta 0:00:06 lr 0.000531 time 0.2748 (0.2860) loss 2.5736 (3.5814) grad_norm 1.5645 (1.4949) [2021-04-16 04:55:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][1240/1251] eta 0:00:03 lr 0.000531 time 0.2481 (0.2858) loss 4.1809 (3.5817) grad_norm 1.3913 (1.4954) [2021-04-16 04:55:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [144/300][1250/1251] eta 0:00:00 lr 0.000531 time 0.2482 (0.2855) loss 3.8148 (3.5832) grad_norm 1.4281 (1.4952) [2021-04-16 04:55:35 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 144 training takes 0:06:00 [2021-04-16 04:55:35 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_144.pth saving...... [2021-04-16 04:55:43 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_144.pth saved !!! [2021-04-16 04:55:44 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.149 (1.149) Loss 1.0415 (1.0415) Acc@1 74.707 (74.707) Acc@5 93.066 (93.066) [2021-04-16 04:55:46 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.108 (0.242) Loss 0.9359 (1.0114) Acc@1 76.660 (76.403) Acc@5 95.020 (93.430) [2021-04-16 04:55:48 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.134 (0.236) Loss 1.0542 (1.0201) Acc@1 74.805 (76.125) Acc@5 92.383 (93.257) [2021-04-16 04:55:51 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.076 (0.232) Loss 1.1265 (1.0263) Acc@1 72.852 (75.907) Acc@5 91.895 (93.098) [2021-04-16 04:55:52 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.221) Loss 1.0014 (1.0248) Acc@1 75.586 (75.931) Acc@5 93.652 (93.143) [2021-04-16 04:55:58 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 75.980 Acc@5 93.164 [2021-04-16 04:55:58 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 76.0% [2021-04-16 04:55:58 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 75.98% [2021-04-16 04:56:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][0/1251] eta 1:19:26 lr 0.000531 time 3.8102 (3.8102) loss 3.6886 (3.6886) grad_norm 1.4574 (1.4574) [2021-04-16 04:56:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][10/1251] eta 0:12:14 lr 0.000531 time 0.2777 (0.5921) loss 3.5906 (3.4939) grad_norm 1.3712 (1.4674) [2021-04-16 04:56:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][20/1251] eta 0:09:03 lr 0.000531 time 0.3127 (0.4417) loss 4.3288 (3.6277) grad_norm 1.5650 (1.4897) [2021-04-16 04:56:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][30/1251] eta 0:07:52 lr 0.000531 time 0.2856 (0.3868) loss 4.1659 (3.6863) grad_norm 1.3133 (1.4684) [2021-04-16 04:56:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][40/1251] eta 0:07:14 lr 0.000531 time 0.2564 (0.3591) loss 4.2941 (3.7304) grad_norm 1.3417 (1.4773) [2021-04-16 04:56:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][50/1251] eta 0:06:51 lr 0.000531 time 0.2460 (0.3426) loss 2.2982 (3.7112) grad_norm 1.6750 (1.4844) [2021-04-16 04:56:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][60/1251] eta 0:06:35 lr 0.000531 time 0.2800 (0.3320) loss 3.3261 (3.7017) grad_norm 1.2409 (1.4777) [2021-04-16 04:56:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][70/1251] eta 0:06:22 lr 0.000531 time 0.2799 (0.3241) loss 3.5095 (3.6675) grad_norm 1.4142 (1.4728) [2021-04-16 04:56:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][80/1251] eta 0:06:12 lr 0.000531 time 0.2719 (0.3181) loss 3.9324 (3.6359) grad_norm 1.6924 (1.4665) [2021-04-16 04:56:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][90/1251] eta 0:06:04 lr 0.000531 time 0.2923 (0.3136) loss 3.4999 (3.6381) grad_norm 1.6309 (1.4814) [2021-04-16 04:56:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][100/1251] eta 0:05:56 lr 0.000530 time 0.2906 (0.3099) loss 2.7992 (3.5862) grad_norm 1.4879 (1.4817) [2021-04-16 04:56:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][110/1251] eta 0:05:50 lr 0.000530 time 0.2595 (0.3068) loss 4.1365 (3.5960) grad_norm 1.5424 (1.4829) [2021-04-16 04:56:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][120/1251] eta 0:05:44 lr 0.000530 time 0.2856 (0.3042) loss 3.8324 (3.5919) grad_norm 1.3621 (1.4878) [2021-04-16 04:56:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][130/1251] eta 0:05:39 lr 0.000530 time 0.2770 (0.3028) loss 4.1715 (3.6026) grad_norm 1.5097 (1.4917) [2021-04-16 04:56:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][140/1251] eta 0:05:35 lr 0.000530 time 0.2495 (0.3021) loss 3.0018 (3.5772) grad_norm 1.4677 (1.4983) [2021-04-16 04:56:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][150/1251] eta 0:05:31 lr 0.000530 time 0.2503 (0.3007) loss 3.5030 (3.5958) grad_norm 1.5076 (1.4925) [2021-04-16 04:56:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][160/1251] eta 0:05:26 lr 0.000530 time 0.2673 (0.2991) loss 3.7404 (3.6054) grad_norm 1.5141 (1.4939) [2021-04-16 04:56:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][170/1251] eta 0:05:21 lr 0.000530 time 0.2776 (0.2977) loss 4.3086 (3.6176) grad_norm 1.5910 (1.5035) [2021-04-16 04:56:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][180/1251] eta 0:05:17 lr 0.000530 time 0.2951 (0.2968) loss 3.0887 (3.6202) grad_norm 1.4196 (1.5045) [2021-04-16 04:56:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][190/1251] eta 0:05:14 lr 0.000530 time 0.3215 (0.2960) loss 3.9077 (3.6335) grad_norm 1.4654 (1.5043) [2021-04-16 04:56:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][200/1251] eta 0:05:09 lr 0.000530 time 0.2899 (0.2949) loss 3.7194 (3.6194) grad_norm 1.3642 (1.5082) [2021-04-16 04:57:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][210/1251] eta 0:05:06 lr 0.000530 time 0.2954 (0.2943) loss 2.6531 (3.6120) grad_norm 1.4327 (1.5119) [2021-04-16 04:57:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][220/1251] eta 0:05:02 lr 0.000530 time 0.2752 (0.2935) loss 3.6982 (3.6228) grad_norm 1.3206 (1.5108) [2021-04-16 04:57:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][230/1251] eta 0:04:59 lr 0.000530 time 0.2739 (0.2929) loss 4.0596 (3.6249) grad_norm 1.4377 (1.5081) [2021-04-16 04:57:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][240/1251] eta 0:04:55 lr 0.000530 time 0.2770 (0.2922) loss 4.1270 (3.6028) grad_norm 1.4453 (1.5043) [2021-04-16 04:57:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][250/1251] eta 0:04:52 lr 0.000530 time 0.2677 (0.2919) loss 2.2488 (3.5879) grad_norm 1.2240 (1.5012) [2021-04-16 04:57:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][260/1251] eta 0:04:48 lr 0.000530 time 0.2811 (0.2912) loss 3.0963 (3.5845) grad_norm 1.6044 (1.5008) [2021-04-16 04:57:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][270/1251] eta 0:04:45 lr 0.000530 time 0.2474 (0.2906) loss 3.9940 (3.5902) grad_norm 1.5035 (1.5013) [2021-04-16 04:57:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][280/1251] eta 0:04:41 lr 0.000530 time 0.2646 (0.2902) loss 4.2658 (3.6016) grad_norm 1.7434 (1.5044) [2021-04-16 04:57:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][290/1251] eta 0:04:38 lr 0.000530 time 0.2913 (0.2898) loss 3.9531 (3.6089) grad_norm 1.6260 (1.5034) [2021-04-16 04:57:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][300/1251] eta 0:04:35 lr 0.000530 time 0.2792 (0.2894) loss 4.0694 (3.6155) grad_norm 1.6632 (1.5017) [2021-04-16 04:57:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][310/1251] eta 0:04:31 lr 0.000530 time 0.3081 (0.2890) loss 3.4624 (3.6180) grad_norm 1.6517 (1.5030) [2021-04-16 04:57:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][320/1251] eta 0:04:28 lr 0.000530 time 0.2694 (0.2888) loss 3.6438 (3.6135) grad_norm 1.3030 (1.5006) [2021-04-16 04:57:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][330/1251] eta 0:04:25 lr 0.000530 time 0.2574 (0.2888) loss 3.9284 (3.6111) grad_norm 1.6757 (1.5012) [2021-04-16 04:57:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][340/1251] eta 0:04:22 lr 0.000529 time 0.2606 (0.2884) loss 2.8058 (3.5983) grad_norm 1.4984 (1.5011) [2021-04-16 04:57:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][350/1251] eta 0:04:20 lr 0.000529 time 0.4311 (0.2886) loss 3.3399 (3.5910) grad_norm 1.2785 (1.4987) [2021-04-16 04:57:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][360/1251] eta 0:04:16 lr 0.000529 time 0.2827 (0.2883) loss 3.6317 (3.5897) grad_norm 1.2618 (1.4971) [2021-04-16 04:57:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][370/1251] eta 0:04:14 lr 0.000529 time 0.2741 (0.2884) loss 3.9300 (3.5950) grad_norm 1.5069 (1.4987) [2021-04-16 04:57:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][380/1251] eta 0:04:10 lr 0.000529 time 0.2415 (0.2881) loss 4.2261 (3.5958) grad_norm 1.5113 (1.4976) [2021-04-16 04:57:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][390/1251] eta 0:04:07 lr 0.000529 time 0.2636 (0.2878) loss 3.5035 (3.5921) grad_norm 1.8862 (1.4998) [2021-04-16 04:57:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][400/1251] eta 0:04:04 lr 0.000529 time 0.2680 (0.2875) loss 4.0822 (3.5954) grad_norm 1.3651 (1.5014) [2021-04-16 04:57:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][410/1251] eta 0:04:01 lr 0.000529 time 0.2694 (0.2871) loss 4.1663 (3.5926) grad_norm 1.4165 (1.5015) [2021-04-16 04:57:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][420/1251] eta 0:03:58 lr 0.000529 time 0.2565 (0.2869) loss 3.8800 (3.5919) grad_norm 1.5591 (1.4999) [2021-04-16 04:58:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][430/1251] eta 0:03:55 lr 0.000529 time 0.2650 (0.2870) loss 2.3527 (3.5972) grad_norm 1.3643 (1.4973) [2021-04-16 04:58:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][440/1251] eta 0:03:52 lr 0.000529 time 0.2980 (0.2867) loss 4.4157 (3.5949) grad_norm 1.4518 (1.4974) [2021-04-16 04:58:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][450/1251] eta 0:03:49 lr 0.000529 time 0.2608 (0.2864) loss 3.0945 (3.5957) grad_norm 1.4134 (1.4978) [2021-04-16 04:58:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][460/1251] eta 0:03:46 lr 0.000529 time 0.2735 (0.2861) loss 2.4662 (3.5904) grad_norm 1.2984 (1.4983) [2021-04-16 04:58:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][470/1251] eta 0:03:43 lr 0.000529 time 0.2795 (0.2858) loss 3.4481 (3.5911) grad_norm 1.8085 (1.4984) [2021-04-16 04:58:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][480/1251] eta 0:03:40 lr 0.000529 time 0.2681 (0.2856) loss 2.8911 (3.5904) grad_norm 1.3979 (1.4966) [2021-04-16 04:58:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][490/1251] eta 0:03:37 lr 0.000529 time 0.2561 (0.2856) loss 3.6147 (3.5922) grad_norm 1.5132 (1.4967) [2021-04-16 04:58:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][500/1251] eta 0:03:34 lr 0.000529 time 0.3122 (0.2855) loss 4.5805 (3.5924) grad_norm 1.3054 (1.4962) [2021-04-16 04:58:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][510/1251] eta 0:03:31 lr 0.000529 time 0.2872 (0.2855) loss 3.9936 (3.5932) grad_norm 1.5434 (1.5006) [2021-04-16 04:58:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][520/1251] eta 0:03:28 lr 0.000529 time 0.2850 (0.2858) loss 3.7409 (3.5941) grad_norm 1.7128 (1.5008) [2021-04-16 04:58:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][530/1251] eta 0:03:25 lr 0.000529 time 0.2755 (0.2856) loss 4.0693 (3.5949) grad_norm 1.7480 (1.5009) [2021-04-16 04:58:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][540/1251] eta 0:03:22 lr 0.000529 time 0.2670 (0.2853) loss 4.2446 (3.5897) grad_norm 1.5256 (1.4993) [2021-04-16 04:58:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][550/1251] eta 0:03:19 lr 0.000529 time 0.2937 (0.2852) loss 3.9536 (3.5943) grad_norm 1.4157 (1.4992) [2021-04-16 04:58:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][560/1251] eta 0:03:17 lr 0.000529 time 0.2807 (0.2851) loss 3.7751 (3.5958) grad_norm 1.2529 (1.4991) [2021-04-16 04:58:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][570/1251] eta 0:03:14 lr 0.000529 time 0.2678 (0.2851) loss 3.6381 (3.5970) grad_norm 1.3243 (1.4983) [2021-04-16 04:58:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][580/1251] eta 0:03:11 lr 0.000529 time 0.2757 (0.2854) loss 2.9485 (3.5944) grad_norm 1.7367 (1.4985) [2021-04-16 04:58:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][590/1251] eta 0:03:08 lr 0.000528 time 0.2921 (0.2852) loss 3.0761 (3.6001) grad_norm 1.4307 (1.4977) [2021-04-16 04:58:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][600/1251] eta 0:03:05 lr 0.000528 time 0.2596 (0.2851) loss 3.1893 (3.5950) grad_norm 1.4393 (1.4967) [2021-04-16 04:58:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][610/1251] eta 0:03:02 lr 0.000528 time 0.2961 (0.2850) loss 2.5921 (3.5935) grad_norm 1.4324 (1.4954) [2021-04-16 04:58:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][620/1251] eta 0:02:59 lr 0.000528 time 0.2689 (0.2849) loss 3.0384 (3.5883) grad_norm 1.4726 (1.4954) [2021-04-16 04:58:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][630/1251] eta 0:02:56 lr 0.000528 time 0.2629 (0.2847) loss 3.9966 (3.5901) grad_norm 1.4773 (1.4947) [2021-04-16 04:59:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][640/1251] eta 0:02:53 lr 0.000528 time 0.2714 (0.2845) loss 4.1773 (3.5865) grad_norm 1.6449 (1.4942) [2021-04-16 04:59:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][650/1251] eta 0:02:50 lr 0.000528 time 0.2599 (0.2844) loss 3.8477 (3.5867) grad_norm 1.4426 (1.4937) [2021-04-16 04:59:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][660/1251] eta 0:02:48 lr 0.000528 time 0.2778 (0.2843) loss 2.9481 (3.5851) grad_norm 1.3712 (1.4937) [2021-04-16 04:59:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][670/1251] eta 0:02:45 lr 0.000528 time 0.2540 (0.2842) loss 3.8809 (3.5803) grad_norm 1.7233 (1.4941) [2021-04-16 04:59:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][680/1251] eta 0:02:42 lr 0.000528 time 0.2586 (0.2840) loss 3.8353 (3.5835) grad_norm 1.3633 (1.4942) [2021-04-16 04:59:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][690/1251] eta 0:02:39 lr 0.000528 time 0.2642 (0.2839) loss 3.9883 (3.5829) grad_norm 1.4017 (1.4937) [2021-04-16 04:59:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][700/1251] eta 0:02:36 lr 0.000528 time 0.2538 (0.2840) loss 3.8089 (3.5817) grad_norm 1.5695 (1.4949) [2021-04-16 04:59:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][710/1251] eta 0:02:33 lr 0.000528 time 0.2652 (0.2838) loss 3.6716 (3.5828) grad_norm 1.7990 (1.4947) [2021-04-16 04:59:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][720/1251] eta 0:02:30 lr 0.000528 time 0.2455 (0.2839) loss 3.9906 (3.5824) grad_norm 1.3773 (1.4940) [2021-04-16 04:59:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][730/1251] eta 0:02:27 lr 0.000528 time 0.2974 (0.2838) loss 2.6588 (3.5843) grad_norm 1.3135 (1.4938) [2021-04-16 04:59:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][740/1251] eta 0:02:24 lr 0.000528 time 0.2481 (0.2837) loss 4.0415 (3.5850) grad_norm 1.5651 (1.4934) [2021-04-16 04:59:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][750/1251] eta 0:02:22 lr 0.000528 time 0.2889 (0.2835) loss 4.0250 (3.5854) grad_norm 1.4106 (1.4934) [2021-04-16 04:59:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][760/1251] eta 0:02:19 lr 0.000528 time 0.2641 (0.2835) loss 3.3343 (3.5876) grad_norm 1.5933 (1.4931) [2021-04-16 04:59:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][770/1251] eta 0:02:16 lr 0.000528 time 0.2734 (0.2833) loss 2.7118 (3.5833) grad_norm 2.1194 (1.4934) [2021-04-16 04:59:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][780/1251] eta 0:02:13 lr 0.000528 time 0.2722 (0.2832) loss 2.9918 (3.5881) grad_norm 1.6228 (1.4945) [2021-04-16 04:59:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][790/1251] eta 0:02:10 lr 0.000528 time 0.2960 (0.2832) loss 2.4911 (3.5855) grad_norm 1.4878 (1.4949) [2021-04-16 04:59:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][800/1251] eta 0:02:07 lr 0.000528 time 0.2798 (0.2831) loss 3.9484 (3.5844) grad_norm 1.3468 (1.4942) [2021-04-16 04:59:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][810/1251] eta 0:02:04 lr 0.000528 time 0.2607 (0.2830) loss 3.8228 (3.5829) grad_norm 1.5009 (1.4932) [2021-04-16 04:59:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][820/1251] eta 0:02:01 lr 0.000528 time 0.2784 (0.2829) loss 3.8908 (3.5776) grad_norm 1.3398 (1.4930) [2021-04-16 04:59:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][830/1251] eta 0:01:59 lr 0.000527 time 0.2791 (0.2830) loss 4.0727 (3.5768) grad_norm 1.4783 (1.4925) [2021-04-16 04:59:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][840/1251] eta 0:01:56 lr 0.000527 time 0.2903 (0.2829) loss 3.7955 (3.5798) grad_norm 1.2631 (1.4929) [2021-04-16 04:59:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][850/1251] eta 0:01:53 lr 0.000527 time 0.2668 (0.2829) loss 2.3611 (3.5796) grad_norm 1.5217 (1.4923) [2021-04-16 05:00:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][860/1251] eta 0:01:50 lr 0.000527 time 0.2826 (0.2828) loss 4.0427 (3.5803) grad_norm 1.3711 (1.4925) [2021-04-16 05:00:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][870/1251] eta 0:01:47 lr 0.000527 time 0.2534 (0.2827) loss 3.9350 (3.5834) grad_norm 1.4663 (1.4917) [2021-04-16 05:00:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][880/1251] eta 0:01:44 lr 0.000527 time 0.2684 (0.2825) loss 3.6102 (3.5824) grad_norm 1.3776 (1.4910) [2021-04-16 05:00:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][890/1251] eta 0:01:42 lr 0.000527 time 0.2572 (0.2826) loss 4.5819 (3.5844) grad_norm 1.4829 (1.4900) [2021-04-16 05:00:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][900/1251] eta 0:01:39 lr 0.000527 time 0.3710 (0.2826) loss 3.5310 (3.5878) grad_norm 1.3562 (1.4895) [2021-04-16 05:00:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][910/1251] eta 0:01:36 lr 0.000527 time 0.2514 (0.2825) loss 4.3252 (3.5918) grad_norm 1.3736 (1.4908) [2021-04-16 05:00:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][920/1251] eta 0:01:33 lr 0.000527 time 0.2487 (0.2824) loss 4.2815 (3.5911) grad_norm 1.3253 (1.4902) [2021-04-16 05:00:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][930/1251] eta 0:01:30 lr 0.000527 time 0.2509 (0.2824) loss 3.8615 (3.5915) grad_norm 1.5898 (1.4901) [2021-04-16 05:00:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][940/1251] eta 0:01:27 lr 0.000527 time 0.2499 (0.2825) loss 2.8099 (3.5909) grad_norm 1.5085 (1.4894) [2021-04-16 05:00:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][950/1251] eta 0:01:25 lr 0.000527 time 0.2907 (0.2825) loss 4.3783 (3.5940) grad_norm 1.3715 (1.4899) [2021-04-16 05:00:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][960/1251] eta 0:01:22 lr 0.000527 time 0.2597 (0.2825) loss 3.1993 (3.5955) grad_norm 1.4662 (1.4894) [2021-04-16 05:00:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][970/1251] eta 0:01:19 lr 0.000527 time 0.2704 (0.2824) loss 3.5419 (3.5968) grad_norm 1.4683 (1.4894) [2021-04-16 05:00:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][980/1251] eta 0:01:16 lr 0.000527 time 0.2713 (0.2823) loss 3.7265 (3.5955) grad_norm 1.6097 (1.4899) [2021-04-16 05:00:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][990/1251] eta 0:01:13 lr 0.000527 time 0.2661 (0.2822) loss 3.5858 (3.5949) grad_norm 1.9713 (1.4912) [2021-04-16 05:00:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][1000/1251] eta 0:01:10 lr 0.000527 time 0.2671 (0.2822) loss 3.5988 (3.5948) grad_norm 1.2521 (1.4917) [2021-04-16 05:00:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][1010/1251] eta 0:01:08 lr 0.000527 time 0.2981 (0.2822) loss 3.0727 (3.5961) grad_norm 1.4654 (1.4920) [2021-04-16 05:00:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][1020/1251] eta 0:01:05 lr 0.000527 time 0.2956 (0.2821) loss 2.4359 (3.5961) grad_norm 1.3145 (1.4919) [2021-04-16 05:00:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][1030/1251] eta 0:01:02 lr 0.000527 time 0.2694 (0.2821) loss 3.7737 (3.5935) grad_norm 1.3980 (1.4911) [2021-04-16 05:00:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][1040/1251] eta 0:00:59 lr 0.000527 time 0.2765 (0.2820) loss 2.5293 (3.5917) grad_norm 1.4039 (1.4909) [2021-04-16 05:00:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][1050/1251] eta 0:00:56 lr 0.000527 time 0.2678 (0.2820) loss 3.5021 (3.5919) grad_norm 1.6366 (1.4909) [2021-04-16 05:00:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][1060/1251] eta 0:00:53 lr 0.000527 time 0.2586 (0.2819) loss 3.6011 (3.5912) grad_norm 1.3706 (1.4907) [2021-04-16 05:01:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][1070/1251] eta 0:00:51 lr 0.000526 time 0.2813 (0.2820) loss 2.3505 (3.5883) grad_norm 1.7387 (1.4907) [2021-04-16 05:01:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][1080/1251] eta 0:00:48 lr 0.000526 time 0.3133 (0.2820) loss 3.6540 (3.5884) grad_norm 1.3699 (1.4903) [2021-04-16 05:01:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][1090/1251] eta 0:00:45 lr 0.000526 time 0.2765 (0.2821) loss 3.4907 (3.5871) grad_norm 1.4917 (1.4896) [2021-04-16 05:01:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][1100/1251] eta 0:00:42 lr 0.000526 time 0.3140 (0.2820) loss 4.1033 (3.5869) grad_norm 1.6860 (1.4901) [2021-04-16 05:01:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][1110/1251] eta 0:00:39 lr 0.000526 time 0.2498 (0.2819) loss 4.2060 (3.5881) grad_norm 1.3691 (1.4895) [2021-04-16 05:01:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][1120/1251] eta 0:00:36 lr 0.000526 time 0.3000 (0.2819) loss 4.2385 (3.5872) grad_norm 1.7841 (1.4900) [2021-04-16 05:01:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][1130/1251] eta 0:00:34 lr 0.000526 time 0.2729 (0.2818) loss 4.4376 (3.5890) grad_norm 1.6036 (1.4911) [2021-04-16 05:01:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][1140/1251] eta 0:00:31 lr 0.000526 time 0.2851 (0.2819) loss 3.9365 (3.5891) grad_norm 1.3193 (1.4912) [2021-04-16 05:01:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][1150/1251] eta 0:00:28 lr 0.000526 time 0.3025 (0.2818) loss 4.2196 (3.5879) grad_norm 1.6023 (1.4921) [2021-04-16 05:01:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][1160/1251] eta 0:00:25 lr 0.000526 time 0.2939 (0.2819) loss 3.9793 (3.5876) grad_norm 1.6013 (1.4919) [2021-04-16 05:01:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][1170/1251] eta 0:00:22 lr 0.000526 time 0.2724 (0.2819) loss 4.3754 (3.5881) grad_norm 1.5729 (1.4919) [2021-04-16 05:01:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][1180/1251] eta 0:00:20 lr 0.000526 time 0.2626 (0.2818) loss 3.3055 (3.5868) grad_norm 1.4650 (1.4915) [2021-04-16 05:01:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][1190/1251] eta 0:00:17 lr 0.000526 time 0.2507 (0.2817) loss 2.7618 (3.5852) grad_norm 1.4930 (1.4920) [2021-04-16 05:01:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][1200/1251] eta 0:00:14 lr 0.000526 time 0.2881 (0.2817) loss 4.5022 (3.5892) grad_norm 1.3700 (1.4935) [2021-04-16 05:01:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][1210/1251] eta 0:00:11 lr 0.000526 time 0.2944 (0.2816) loss 2.5224 (3.5895) grad_norm 1.3294 (1.4939) [2021-04-16 05:01:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][1220/1251] eta 0:00:08 lr 0.000526 time 0.2803 (0.2816) loss 3.5251 (3.5901) grad_norm 1.9755 (1.4942) [2021-04-16 05:01:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][1230/1251] eta 0:00:05 lr 0.000526 time 0.2790 (0.2815) loss 4.0560 (3.5907) grad_norm 1.3998 (1.4949) [2021-04-16 05:01:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][1240/1251] eta 0:00:03 lr 0.000526 time 0.2485 (0.2814) loss 3.2424 (3.5909) grad_norm 1.2959 (1.4940) [2021-04-16 05:01:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [145/300][1250/1251] eta 0:00:00 lr 0.000526 time 0.2484 (0.2811) loss 3.8394 (3.5928) grad_norm 1.4232 (1.4938) [2021-04-16 05:01:53 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 145 training takes 0:05:55 [2021-04-16 05:01:53 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_145.pth saving...... [2021-04-16 05:02:01 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_145.pth saved !!! [2021-04-16 05:02:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.172 (1.172) Loss 1.0651 (1.0651) Acc@1 75.195 (75.195) Acc@5 91.699 (91.699) [2021-04-16 05:02:03 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.134 (0.228) Loss 0.9604 (1.0040) Acc@1 76.367 (76.474) Acc@5 94.141 (93.226) [2021-04-16 05:02:05 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.155 (0.203) Loss 0.9538 (1.0125) Acc@1 77.734 (76.153) Acc@5 93.652 (93.248) [2021-04-16 05:02:08 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.106 (0.231) Loss 1.0050 (1.0164) Acc@1 75.391 (75.973) Acc@5 93.359 (93.164) [2021-04-16 05:02:10 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.218) Loss 0.9378 (1.0156) Acc@1 77.539 (75.912) Acc@5 93.848 (93.216) [2021-04-16 05:02:14 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 75.802 Acc@5 93.174 [2021-04-16 05:02:14 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 75.8% [2021-04-16 05:02:14 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 75.98% [2021-04-16 05:02:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][0/1251] eta 1:25:17 lr 0.000526 time 4.0905 (4.0905) loss 3.9761 (3.9761) grad_norm 1.6547 (1.6547) [2021-04-16 05:02:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][10/1251] eta 0:12:52 lr 0.000526 time 0.2966 (0.6224) loss 4.0853 (3.7672) grad_norm 1.8345 (1.5250) [2021-04-16 05:02:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][20/1251] eta 0:09:26 lr 0.000526 time 0.2657 (0.4605) loss 3.0568 (3.5821) grad_norm 1.4481 (1.5986) [2021-04-16 05:02:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][30/1251] eta 0:08:11 lr 0.000526 time 0.2973 (0.4022) loss 2.3657 (3.5515) grad_norm 1.2952 (1.5875) [2021-04-16 05:02:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][40/1251] eta 0:07:32 lr 0.000526 time 0.2703 (0.3739) loss 4.1327 (3.5803) grad_norm 1.5585 (1.5484) [2021-04-16 05:02:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][50/1251] eta 0:07:07 lr 0.000526 time 0.2968 (0.3559) loss 4.0784 (3.5999) grad_norm 1.9986 (1.5620) [2021-04-16 05:02:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][60/1251] eta 0:06:48 lr 0.000525 time 0.2770 (0.3428) loss 4.2981 (3.6068) grad_norm 1.3235 (1.5485) [2021-04-16 05:02:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][70/1251] eta 0:06:33 lr 0.000525 time 0.2557 (0.3334) loss 3.4815 (3.5168) grad_norm 1.3019 (1.5293) [2021-04-16 05:02:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][80/1251] eta 0:06:22 lr 0.000525 time 0.2856 (0.3263) loss 4.1028 (3.5375) grad_norm 1.2801 (1.5181) [2021-04-16 05:02:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][90/1251] eta 0:06:13 lr 0.000525 time 0.2604 (0.3216) loss 3.7058 (3.5672) grad_norm 1.7587 (1.5145) [2021-04-16 05:02:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][100/1251] eta 0:06:06 lr 0.000525 time 0.2647 (0.3187) loss 3.8113 (3.5626) grad_norm 1.4410 (1.5222) [2021-04-16 05:02:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][110/1251] eta 0:05:59 lr 0.000525 time 0.2560 (0.3149) loss 3.9327 (3.5776) grad_norm 1.4725 (1.5174) [2021-04-16 05:02:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][120/1251] eta 0:05:52 lr 0.000525 time 0.2827 (0.3119) loss 3.7768 (3.5933) grad_norm 1.4889 (1.5132) [2021-04-16 05:02:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][130/1251] eta 0:05:47 lr 0.000525 time 0.2747 (0.3099) loss 3.5014 (3.5639) grad_norm 1.5395 (1.5100) [2021-04-16 05:02:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][140/1251] eta 0:05:42 lr 0.000525 time 0.2537 (0.3084) loss 4.1264 (3.5601) grad_norm 1.3539 (1.5035) [2021-04-16 05:03:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][150/1251] eta 0:05:37 lr 0.000525 time 0.2914 (0.3063) loss 3.7304 (3.5433) grad_norm 1.6358 (1.4994) [2021-04-16 05:03:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][160/1251] eta 0:05:31 lr 0.000525 time 0.2763 (0.3042) loss 2.8035 (3.5441) grad_norm 1.4826 (1.5052) [2021-04-16 05:03:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][170/1251] eta 0:05:26 lr 0.000525 time 0.2867 (0.3025) loss 3.8560 (3.5420) grad_norm 1.2737 (1.5064) [2021-04-16 05:03:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][180/1251] eta 0:05:22 lr 0.000525 time 0.2816 (0.3011) loss 4.5849 (3.5325) grad_norm 1.3198 (1.5058) [2021-04-16 05:03:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][190/1251] eta 0:05:18 lr 0.000525 time 0.2770 (0.2998) loss 3.3279 (3.5443) grad_norm 1.4349 (1.5061) [2021-04-16 05:03:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][200/1251] eta 0:05:13 lr 0.000525 time 0.2705 (0.2985) loss 2.6716 (3.5442) grad_norm 1.6555 (1.5066) [2021-04-16 05:03:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][210/1251] eta 0:05:09 lr 0.000525 time 0.3023 (0.2976) loss 4.3253 (3.5426) grad_norm 1.6487 (1.5073) [2021-04-16 05:03:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][220/1251] eta 0:05:06 lr 0.000525 time 0.2767 (0.2975) loss 3.9196 (3.5361) grad_norm 1.5635 (1.5027) [2021-04-16 05:03:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][230/1251] eta 0:05:02 lr 0.000525 time 0.2891 (0.2966) loss 3.9822 (3.5268) grad_norm 1.2666 (1.4998) [2021-04-16 05:03:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][240/1251] eta 0:04:59 lr 0.000525 time 0.2702 (0.2958) loss 4.3408 (3.5309) grad_norm 1.2571 (1.4945) [2021-04-16 05:03:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][250/1251] eta 0:04:55 lr 0.000525 time 0.2528 (0.2950) loss 3.6240 (3.5285) grad_norm 1.7246 (1.4973) [2021-04-16 05:03:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][260/1251] eta 0:04:51 lr 0.000525 time 0.2714 (0.2943) loss 4.1113 (3.5342) grad_norm 1.6276 (1.4981) [2021-04-16 05:03:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][270/1251] eta 0:04:47 lr 0.000525 time 0.2468 (0.2934) loss 3.7387 (3.5339) grad_norm 1.5019 (1.4986) [2021-04-16 05:03:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][280/1251] eta 0:04:44 lr 0.000525 time 0.3007 (0.2929) loss 3.9140 (3.5433) grad_norm 1.4036 (1.4998) [2021-04-16 05:03:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][290/1251] eta 0:04:41 lr 0.000525 time 0.2793 (0.2928) loss 3.5234 (3.5443) grad_norm 1.4357 (1.5020) [2021-04-16 05:03:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][300/1251] eta 0:04:38 lr 0.000524 time 0.2717 (0.2927) loss 3.1755 (3.5467) grad_norm 1.3527 (1.5021) [2021-04-16 05:03:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][310/1251] eta 0:04:34 lr 0.000524 time 0.2798 (0.2919) loss 4.3297 (3.5581) grad_norm 1.3800 (1.5030) [2021-04-16 05:03:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][320/1251] eta 0:04:31 lr 0.000524 time 0.2944 (0.2915) loss 3.0888 (3.5523) grad_norm 1.4495 (1.4998) [2021-04-16 05:03:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][330/1251] eta 0:04:28 lr 0.000524 time 0.2525 (0.2911) loss 3.1933 (3.5534) grad_norm 1.4853 (1.4979) [2021-04-16 05:03:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][340/1251] eta 0:04:25 lr 0.000524 time 0.2727 (0.2909) loss 2.9963 (3.5472) grad_norm 1.4477 (1.4957) [2021-04-16 05:03:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][350/1251] eta 0:04:21 lr 0.000524 time 0.2623 (0.2904) loss 3.9714 (3.5472) grad_norm 1.4656 (1.4969) [2021-04-16 05:03:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][360/1251] eta 0:04:18 lr 0.000524 time 0.2731 (0.2904) loss 3.7174 (3.5567) grad_norm 1.3732 (1.4973) [2021-04-16 05:04:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][370/1251] eta 0:04:16 lr 0.000524 time 0.2892 (0.2906) loss 3.6760 (3.5535) grad_norm 1.5900 (1.4979) [2021-04-16 05:04:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][380/1251] eta 0:04:12 lr 0.000524 time 0.2809 (0.2901) loss 3.4910 (3.5462) grad_norm 1.2222 (1.4947) [2021-04-16 05:04:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][390/1251] eta 0:04:09 lr 0.000524 time 0.3014 (0.2898) loss 3.2006 (3.5474) grad_norm 1.3962 (1.4920) [2021-04-16 05:04:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][400/1251] eta 0:04:06 lr 0.000524 time 0.2505 (0.2894) loss 3.6941 (3.5441) grad_norm 1.2789 (1.4915) [2021-04-16 05:04:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][410/1251] eta 0:04:03 lr 0.000524 time 0.2805 (0.2890) loss 3.1160 (3.5433) grad_norm 1.5070 (1.4918) [2021-04-16 05:04:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][420/1251] eta 0:03:59 lr 0.000524 time 0.2749 (0.2887) loss 4.0070 (3.5402) grad_norm 1.4708 (1.4902) [2021-04-16 05:04:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][430/1251] eta 0:03:56 lr 0.000524 time 0.2760 (0.2884) loss 3.8642 (3.5392) grad_norm 1.8248 (1.4917) [2021-04-16 05:04:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][440/1251] eta 0:03:53 lr 0.000524 time 0.2918 (0.2880) loss 3.9613 (3.5405) grad_norm 1.5095 (1.4926) [2021-04-16 05:04:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][450/1251] eta 0:03:50 lr 0.000524 time 0.2810 (0.2877) loss 3.4373 (3.5324) grad_norm 1.6746 (1.4928) [2021-04-16 05:04:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][460/1251] eta 0:03:47 lr 0.000524 time 0.3166 (0.2876) loss 4.0030 (3.5305) grad_norm 1.9236 (1.4923) [2021-04-16 05:04:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][470/1251] eta 0:03:44 lr 0.000524 time 0.2899 (0.2874) loss 3.8006 (3.5302) grad_norm 1.3941 (1.4940) [2021-04-16 05:04:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][480/1251] eta 0:03:41 lr 0.000524 time 0.2663 (0.2872) loss 3.6492 (3.5237) grad_norm 1.7083 (1.4956) [2021-04-16 05:04:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][490/1251] eta 0:03:38 lr 0.000524 time 0.2821 (0.2870) loss 2.5118 (3.5210) grad_norm 1.5106 (1.4964) [2021-04-16 05:04:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][500/1251] eta 0:03:35 lr 0.000524 time 0.2926 (0.2867) loss 4.4853 (3.5203) grad_norm 1.4156 (1.4961) [2021-04-16 05:04:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][510/1251] eta 0:03:32 lr 0.000524 time 0.2845 (0.2865) loss 4.0540 (3.5212) grad_norm 1.9201 (1.4978) [2021-04-16 05:04:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][520/1251] eta 0:03:29 lr 0.000524 time 0.2825 (0.2863) loss 3.4699 (3.5239) grad_norm 1.5368 (1.4984) [2021-04-16 05:04:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][530/1251] eta 0:03:26 lr 0.000524 time 0.2849 (0.2860) loss 3.7916 (3.5192) grad_norm 1.4117 (1.4986) [2021-04-16 05:04:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][540/1251] eta 0:03:23 lr 0.000523 time 0.2886 (0.2862) loss 3.9791 (3.5210) grad_norm 1.3903 (1.4990) [2021-04-16 05:04:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][550/1251] eta 0:03:20 lr 0.000523 time 0.2792 (0.2860) loss 4.1605 (3.5238) grad_norm 1.7021 (1.4978) [2021-04-16 05:04:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][560/1251] eta 0:03:17 lr 0.000523 time 0.2751 (0.2858) loss 2.7681 (3.5200) grad_norm 1.3904 (1.4970) [2021-04-16 05:04:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][570/1251] eta 0:03:14 lr 0.000523 time 0.2818 (0.2859) loss 2.3807 (3.5206) grad_norm 1.2856 (1.4960) [2021-04-16 05:05:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][580/1251] eta 0:03:11 lr 0.000523 time 0.2606 (0.2858) loss 4.0778 (3.5242) grad_norm 1.3680 (1.4968) [2021-04-16 05:05:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][590/1251] eta 0:03:08 lr 0.000523 time 0.3149 (0.2858) loss 3.8593 (3.5214) grad_norm 1.4057 (1.4957) [2021-04-16 05:05:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][600/1251] eta 0:03:06 lr 0.000523 time 0.2827 (0.2857) loss 4.2750 (3.5262) grad_norm 1.4523 (1.4958) [2021-04-16 05:05:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][610/1251] eta 0:03:03 lr 0.000523 time 0.3113 (0.2857) loss 2.5124 (3.5250) grad_norm 1.6319 (1.4959) [2021-04-16 05:05:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][620/1251] eta 0:03:00 lr 0.000523 time 0.2836 (0.2857) loss 4.1599 (3.5289) grad_norm 1.7593 (1.4973) [2021-04-16 05:05:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][630/1251] eta 0:02:57 lr 0.000523 time 0.2819 (0.2855) loss 3.5210 (3.5249) grad_norm 1.4311 (1.4967) [2021-04-16 05:05:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][640/1251] eta 0:02:54 lr 0.000523 time 0.2837 (0.2853) loss 2.9362 (3.5254) grad_norm 1.8293 (1.4956) [2021-04-16 05:05:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][650/1251] eta 0:02:51 lr 0.000523 time 0.2825 (0.2852) loss 2.9582 (3.5226) grad_norm 1.5632 (1.4954) [2021-04-16 05:05:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][660/1251] eta 0:02:48 lr 0.000523 time 0.2945 (0.2850) loss 3.9697 (3.5227) grad_norm 1.6339 (1.4961) [2021-04-16 05:05:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][670/1251] eta 0:02:45 lr 0.000523 time 0.2728 (0.2849) loss 4.0649 (3.5228) grad_norm 1.4260 (1.4964) [2021-04-16 05:05:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][680/1251] eta 0:02:42 lr 0.000523 time 0.2770 (0.2850) loss 3.4011 (3.5240) grad_norm 1.3534 (1.4962) [2021-04-16 05:05:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][690/1251] eta 0:02:39 lr 0.000523 time 0.2730 (0.2849) loss 4.3413 (3.5256) grad_norm 1.5255 (1.4958) [2021-04-16 05:05:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][700/1251] eta 0:02:36 lr 0.000523 time 0.2816 (0.2849) loss 3.7855 (3.5267) grad_norm 1.7518 (1.4967) [2021-04-16 05:05:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][710/1251] eta 0:02:34 lr 0.000523 time 0.2975 (0.2850) loss 3.8671 (3.5322) grad_norm 1.2828 (1.4967) [2021-04-16 05:05:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][720/1251] eta 0:02:31 lr 0.000523 time 0.2616 (0.2850) loss 4.0837 (3.5321) grad_norm 1.4357 (inf) [2021-04-16 05:05:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][730/1251] eta 0:02:28 lr 0.000523 time 0.2971 (0.2849) loss 3.2833 (3.5282) grad_norm 1.4043 (inf) [2021-04-16 05:05:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][740/1251] eta 0:02:25 lr 0.000523 time 0.2697 (0.2848) loss 3.3798 (3.5288) grad_norm 1.3769 (inf) [2021-04-16 05:05:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][750/1251] eta 0:02:22 lr 0.000523 time 0.2740 (0.2847) loss 2.4231 (3.5311) grad_norm 1.5105 (inf) [2021-04-16 05:05:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][760/1251] eta 0:02:19 lr 0.000523 time 0.2821 (0.2846) loss 3.3692 (3.5314) grad_norm 1.5805 (inf) [2021-04-16 05:05:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][770/1251] eta 0:02:16 lr 0.000523 time 0.2670 (0.2844) loss 2.9550 (3.5268) grad_norm 1.5835 (inf) [2021-04-16 05:05:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][780/1251] eta 0:02:13 lr 0.000522 time 0.2967 (0.2843) loss 3.9823 (3.5276) grad_norm 1.3752 (inf) [2021-04-16 05:05:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][790/1251] eta 0:02:11 lr 0.000522 time 0.2749 (0.2844) loss 4.5341 (3.5275) grad_norm 1.4740 (inf) [2021-04-16 05:06:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][800/1251] eta 0:02:08 lr 0.000522 time 0.2916 (0.2843) loss 3.5516 (3.5281) grad_norm 1.7315 (inf) [2021-04-16 05:06:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][810/1251] eta 0:02:05 lr 0.000522 time 0.2971 (0.2842) loss 2.6347 (3.5271) grad_norm 1.4838 (inf) [2021-04-16 05:06:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][820/1251] eta 0:02:02 lr 0.000522 time 0.2563 (0.2840) loss 3.2470 (3.5299) grad_norm 1.6183 (inf) [2021-04-16 05:06:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][830/1251] eta 0:01:59 lr 0.000522 time 0.2981 (0.2840) loss 3.3490 (3.5332) grad_norm 1.4297 (inf) [2021-04-16 05:06:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][840/1251] eta 0:01:56 lr 0.000522 time 0.2643 (0.2838) loss 3.0135 (3.5329) grad_norm 1.5472 (inf) [2021-04-16 05:06:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][850/1251] eta 0:01:53 lr 0.000522 time 0.2957 (0.2837) loss 4.0346 (3.5356) grad_norm 1.5353 (inf) [2021-04-16 05:06:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][860/1251] eta 0:01:50 lr 0.000522 time 0.2861 (0.2836) loss 3.8141 (3.5354) grad_norm 1.5049 (inf) [2021-04-16 05:06:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][870/1251] eta 0:01:48 lr 0.000522 time 0.2628 (0.2836) loss 3.2412 (3.5355) grad_norm 1.4328 (inf) [2021-04-16 05:06:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][880/1251] eta 0:01:45 lr 0.000522 time 0.2743 (0.2836) loss 3.9380 (3.5357) grad_norm 1.4840 (inf) [2021-04-16 05:06:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][890/1251] eta 0:01:42 lr 0.000522 time 0.2891 (0.2836) loss 2.8163 (3.5342) grad_norm 1.4204 (inf) [2021-04-16 05:06:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][900/1251] eta 0:01:39 lr 0.000522 time 0.3015 (0.2836) loss 4.2967 (3.5365) grad_norm 1.4025 (inf) [2021-04-16 05:06:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][910/1251] eta 0:01:36 lr 0.000522 time 0.2567 (0.2835) loss 3.6150 (3.5360) grad_norm 1.5621 (inf) [2021-04-16 05:06:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][920/1251] eta 0:01:33 lr 0.000522 time 0.2863 (0.2836) loss 4.0689 (3.5372) grad_norm 1.7511 (inf) [2021-04-16 05:06:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][930/1251] eta 0:01:31 lr 0.000522 time 0.2802 (0.2837) loss 3.9160 (3.5350) grad_norm 1.6947 (inf) [2021-04-16 05:06:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][940/1251] eta 0:01:28 lr 0.000522 time 0.2649 (0.2837) loss 3.3603 (3.5340) grad_norm 1.4994 (inf) [2021-04-16 05:06:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][950/1251] eta 0:01:25 lr 0.000522 time 0.2644 (0.2836) loss 3.7959 (3.5322) grad_norm 1.4787 (inf) [2021-04-16 05:06:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][960/1251] eta 0:01:22 lr 0.000522 time 0.2759 (0.2835) loss 4.3273 (3.5377) grad_norm 1.4370 (inf) [2021-04-16 05:06:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][970/1251] eta 0:01:19 lr 0.000522 time 0.3111 (0.2835) loss 3.5822 (3.5348) grad_norm 2.1768 (inf) [2021-04-16 05:06:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][980/1251] eta 0:01:16 lr 0.000522 time 0.2693 (0.2834) loss 2.5766 (3.5351) grad_norm 1.4813 (inf) [2021-04-16 05:06:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][990/1251] eta 0:01:13 lr 0.000522 time 0.2773 (0.2834) loss 3.2640 (3.5332) grad_norm 1.3528 (inf) [2021-04-16 05:06:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][1000/1251] eta 0:01:11 lr 0.000522 time 0.2807 (0.2833) loss 3.5980 (3.5339) grad_norm 1.7108 (inf) [2021-04-16 05:07:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][1010/1251] eta 0:01:08 lr 0.000522 time 0.2425 (0.2832) loss 2.7707 (3.5333) grad_norm 2.0668 (inf) [2021-04-16 05:07:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][1020/1251] eta 0:01:05 lr 0.000522 time 0.2799 (0.2832) loss 3.5223 (3.5338) grad_norm 1.5717 (inf) [2021-04-16 05:07:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][1030/1251] eta 0:01:02 lr 0.000521 time 0.2945 (0.2831) loss 4.2635 (3.5359) grad_norm 1.7369 (inf) [2021-04-16 05:07:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][1040/1251] eta 0:00:59 lr 0.000521 time 0.2678 (0.2830) loss 3.7588 (3.5361) grad_norm 1.5670 (inf) [2021-04-16 05:07:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][1050/1251] eta 0:00:56 lr 0.000521 time 0.2506 (0.2830) loss 3.9030 (3.5383) grad_norm 1.5232 (inf) [2021-04-16 05:07:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][1060/1251] eta 0:00:54 lr 0.000521 time 0.2678 (0.2829) loss 4.2020 (3.5395) grad_norm 1.6695 (inf) [2021-04-16 05:07:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][1070/1251] eta 0:00:51 lr 0.000521 time 0.2880 (0.2828) loss 3.6076 (3.5396) grad_norm 1.4286 (inf) [2021-04-16 05:07:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][1080/1251] eta 0:00:48 lr 0.000521 time 0.2651 (0.2827) loss 3.8150 (3.5410) grad_norm 1.5388 (inf) [2021-04-16 05:07:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][1090/1251] eta 0:00:45 lr 0.000521 time 0.2897 (0.2827) loss 3.1335 (3.5406) grad_norm 1.7214 (inf) [2021-04-16 05:07:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][1100/1251] eta 0:00:42 lr 0.000521 time 0.2757 (0.2827) loss 3.7281 (3.5430) grad_norm 1.4751 (inf) [2021-04-16 05:07:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][1110/1251] eta 0:00:39 lr 0.000521 time 0.2650 (0.2826) loss 3.3833 (3.5445) grad_norm 1.4062 (inf) [2021-04-16 05:07:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][1120/1251] eta 0:00:37 lr 0.000521 time 0.3095 (0.2827) loss 3.3135 (3.5461) grad_norm 1.4543 (inf) [2021-04-16 05:07:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][1130/1251] eta 0:00:34 lr 0.000521 time 0.2722 (0.2826) loss 2.9162 (3.5440) grad_norm 1.4539 (inf) [2021-04-16 05:07:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][1140/1251] eta 0:00:31 lr 0.000521 time 0.2641 (0.2826) loss 3.9073 (3.5448) grad_norm 1.8344 (inf) [2021-04-16 05:07:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][1150/1251] eta 0:00:28 lr 0.000521 time 0.2717 (0.2825) loss 3.6734 (3.5425) grad_norm 1.4783 (inf) [2021-04-16 05:07:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][1160/1251] eta 0:00:25 lr 0.000521 time 0.2812 (0.2826) loss 3.7162 (3.5427) grad_norm 1.2903 (inf) [2021-04-16 05:07:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][1170/1251] eta 0:00:22 lr 0.000521 time 0.2721 (0.2825) loss 3.9535 (3.5445) grad_norm 1.3585 (inf) [2021-04-16 05:07:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][1180/1251] eta 0:00:20 lr 0.000521 time 0.2701 (0.2825) loss 3.9760 (3.5455) grad_norm 1.5958 (inf) [2021-04-16 05:07:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][1190/1251] eta 0:00:17 lr 0.000521 time 0.2890 (0.2825) loss 4.1311 (3.5465) grad_norm 1.3361 (inf) [2021-04-16 05:07:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][1200/1251] eta 0:00:14 lr 0.000521 time 0.2609 (0.2824) loss 2.8270 (3.5464) grad_norm 1.5336 (inf) [2021-04-16 05:07:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][1210/1251] eta 0:00:11 lr 0.000521 time 0.2829 (0.2824) loss 3.6369 (3.5472) grad_norm 1.5753 (inf) [2021-04-16 05:07:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][1220/1251] eta 0:00:08 lr 0.000521 time 0.2688 (0.2823) loss 3.3554 (3.5473) grad_norm 1.3326 (inf) [2021-04-16 05:08:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][1230/1251] eta 0:00:05 lr 0.000521 time 0.2828 (0.2823) loss 3.8091 (3.5477) grad_norm 1.3630 (inf) [2021-04-16 05:08:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][1240/1251] eta 0:00:03 lr 0.000521 time 0.2513 (0.2821) loss 3.7158 (3.5480) grad_norm 1.4236 (inf) [2021-04-16 05:08:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [146/300][1250/1251] eta 0:00:00 lr 0.000521 time 0.2483 (0.2819) loss 2.3252 (3.5472) grad_norm 1.2735 (inf) [2021-04-16 05:08:10 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 146 training takes 0:05:56 [2021-04-16 05:08:10 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_146.pth saving...... [2021-04-16 05:08:26 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_146.pth saved !!! [2021-04-16 05:08:27 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.224 (1.224) Loss 1.0705 (1.0705) Acc@1 74.707 (74.707) Acc@5 93.066 (93.066) [2021-04-16 05:08:28 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.110 (0.207) Loss 1.0125 (1.0363) Acc@1 75.977 (76.154) Acc@5 93.750 (93.333) [2021-04-16 05:08:31 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.125 (0.230) Loss 1.0463 (1.0387) Acc@1 75.879 (76.037) Acc@5 93.457 (93.159) [2021-04-16 05:08:33 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.518 (0.246) Loss 1.0359 (1.0387) Acc@1 77.051 (75.816) Acc@5 92.871 (93.155) [2021-04-16 05:08:35 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.084 (0.218) Loss 1.1080 (1.0362) Acc@1 75.000 (75.853) Acc@5 91.797 (93.207) [2021-04-16 05:08:39 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 75.866 Acc@5 93.238 [2021-04-16 05:08:39 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 75.9% [2021-04-16 05:08:39 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 75.98% [2021-04-16 05:08:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][0/1251] eta 1:55:41 lr 0.000521 time 5.5491 (5.5491) loss 3.8651 (3.8651) grad_norm 1.4188 (1.4188) [2021-04-16 05:08:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][10/1251] eta 0:15:33 lr 0.000521 time 0.2696 (0.7525) loss 3.5187 (3.2018) grad_norm 1.7275 (1.4678) [2021-04-16 05:08:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][20/1251] eta 0:10:58 lr 0.000520 time 0.2841 (0.5349) loss 2.9563 (3.3544) grad_norm 1.7270 (1.5072) [2021-04-16 05:08:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][30/1251] eta 0:09:13 lr 0.000520 time 0.2652 (0.4531) loss 3.8972 (3.3831) grad_norm 1.5896 (1.5090) [2021-04-16 05:08:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][40/1251] eta 0:08:16 lr 0.000520 time 0.2788 (0.4098) loss 3.6129 (3.4125) grad_norm 1.3785 (1.5044) [2021-04-16 05:08:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][50/1251] eta 0:07:40 lr 0.000520 time 0.2867 (0.3833) loss 4.4738 (3.4091) grad_norm 1.4835 (1.5110) [2021-04-16 05:09:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][60/1251] eta 0:07:15 lr 0.000520 time 0.3047 (0.3658) loss 4.2088 (3.4541) grad_norm 1.5023 (1.5055) [2021-04-16 05:09:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][70/1251] eta 0:06:56 lr 0.000520 time 0.2666 (0.3525) loss 2.6355 (3.4327) grad_norm 1.6023 (1.5065) [2021-04-16 05:09:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][80/1251] eta 0:06:44 lr 0.000520 time 0.2582 (0.3453) loss 4.0250 (3.4680) grad_norm 2.1148 (1.5126) [2021-04-16 05:09:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][90/1251] eta 0:06:33 lr 0.000520 time 0.2650 (0.3390) loss 3.4640 (3.4580) grad_norm 1.5006 (1.5283) [2021-04-16 05:09:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][100/1251] eta 0:06:23 lr 0.000520 time 0.2852 (0.3328) loss 2.8968 (3.4528) grad_norm 1.3408 (1.5246) [2021-04-16 05:09:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][110/1251] eta 0:06:14 lr 0.000520 time 0.3019 (0.3280) loss 2.9827 (3.4442) grad_norm 1.5136 (1.5142) [2021-04-16 05:09:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][120/1251] eta 0:06:06 lr 0.000520 time 0.2769 (0.3245) loss 3.1911 (3.4428) grad_norm 1.2781 (1.5108) [2021-04-16 05:09:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][130/1251] eta 0:05:59 lr 0.000520 time 0.2705 (0.3209) loss 4.0640 (3.4564) grad_norm 1.2615 (1.5075) [2021-04-16 05:09:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][140/1251] eta 0:05:54 lr 0.000520 time 0.3144 (0.3195) loss 4.0167 (3.4607) grad_norm 1.3111 (1.5000) [2021-04-16 05:09:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][150/1251] eta 0:05:48 lr 0.000520 time 0.2749 (0.3166) loss 2.2549 (3.4546) grad_norm 1.3168 (1.4947) [2021-04-16 05:09:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][160/1251] eta 0:05:43 lr 0.000520 time 0.2735 (0.3148) loss 3.7197 (3.4565) grad_norm 1.3755 (1.4905) [2021-04-16 05:09:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][170/1251] eta 0:05:38 lr 0.000520 time 0.3052 (0.3128) loss 3.1718 (3.4700) grad_norm 1.3200 (1.4922) [2021-04-16 05:09:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][180/1251] eta 0:05:32 lr 0.000520 time 0.2642 (0.3107) loss 3.6212 (3.4731) grad_norm 1.5269 (1.4967) [2021-04-16 05:09:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][190/1251] eta 0:05:27 lr 0.000520 time 0.3050 (0.3091) loss 3.8839 (3.4895) grad_norm 1.3248 (1.4942) [2021-04-16 05:09:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][200/1251] eta 0:05:22 lr 0.000520 time 0.2683 (0.3072) loss 2.6631 (3.4962) grad_norm 1.5348 (1.5038) [2021-04-16 05:09:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][210/1251] eta 0:05:18 lr 0.000520 time 0.2815 (0.3061) loss 3.6640 (3.4899) grad_norm 1.6588 (1.5091) [2021-04-16 05:09:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][220/1251] eta 0:05:14 lr 0.000520 time 0.2825 (0.3049) loss 4.0614 (3.4934) grad_norm 1.4604 (1.5046) [2021-04-16 05:09:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][230/1251] eta 0:05:09 lr 0.000520 time 0.2970 (0.3036) loss 2.6160 (3.4954) grad_norm 1.3234 (1.5000) [2021-04-16 05:09:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][240/1251] eta 0:05:05 lr 0.000520 time 0.2506 (0.3023) loss 3.9254 (3.5030) grad_norm 1.6646 (1.4969) [2021-04-16 05:09:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][250/1251] eta 0:05:01 lr 0.000520 time 0.2567 (0.3012) loss 3.9806 (3.5046) grad_norm 1.5005 (1.4958) [2021-04-16 05:09:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][260/1251] eta 0:04:57 lr 0.000519 time 0.2557 (0.3004) loss 2.9315 (3.5039) grad_norm 1.4742 (1.4954) [2021-04-16 05:10:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][270/1251] eta 0:04:53 lr 0.000519 time 0.2990 (0.2994) loss 4.1744 (3.5095) grad_norm 1.7391 (1.4991) [2021-04-16 05:10:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][280/1251] eta 0:04:50 lr 0.000519 time 0.2688 (0.2987) loss 4.1327 (3.5173) grad_norm 1.7124 (1.4997) [2021-04-16 05:10:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][290/1251] eta 0:04:46 lr 0.000519 time 0.2435 (0.2979) loss 2.9594 (3.5142) grad_norm 1.4010 (1.4986) [2021-04-16 05:10:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][300/1251] eta 0:04:42 lr 0.000519 time 0.2662 (0.2971) loss 3.2549 (3.5073) grad_norm 1.6369 (1.4981) [2021-04-16 05:10:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][310/1251] eta 0:04:38 lr 0.000519 time 0.2703 (0.2963) loss 2.7253 (3.5022) grad_norm 1.6762 (1.4995) [2021-04-16 05:10:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][320/1251] eta 0:04:35 lr 0.000519 time 0.2770 (0.2960) loss 4.4436 (3.5057) grad_norm 1.5011 (1.4995) [2021-04-16 05:10:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][330/1251] eta 0:04:32 lr 0.000519 time 0.2667 (0.2955) loss 3.9273 (3.5114) grad_norm 1.6533 (1.4997) [2021-04-16 05:10:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][340/1251] eta 0:04:29 lr 0.000519 time 0.2584 (0.2953) loss 3.4360 (3.5172) grad_norm 1.3800 (1.4965) [2021-04-16 05:10:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][350/1251] eta 0:04:25 lr 0.000519 time 0.2800 (0.2951) loss 4.3272 (3.5231) grad_norm 1.3630 (1.4965) [2021-04-16 05:10:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][360/1251] eta 0:04:22 lr 0.000519 time 0.2487 (0.2949) loss 3.2739 (3.5128) grad_norm 1.6694 (1.4973) [2021-04-16 05:10:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][370/1251] eta 0:04:19 lr 0.000519 time 0.2882 (0.2944) loss 3.6564 (3.5171) grad_norm 2.3888 (1.5010) [2021-04-16 05:10:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][380/1251] eta 0:04:16 lr 0.000519 time 0.2862 (0.2939) loss 2.4897 (3.5188) grad_norm 1.3286 (1.5021) [2021-04-16 05:10:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][390/1251] eta 0:04:12 lr 0.000519 time 0.2734 (0.2938) loss 3.8521 (3.5289) grad_norm 1.6603 (1.5024) [2021-04-16 05:10:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][400/1251] eta 0:04:09 lr 0.000519 time 0.2549 (0.2932) loss 4.1125 (3.5292) grad_norm 1.5375 (1.5034) [2021-04-16 05:10:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][410/1251] eta 0:04:06 lr 0.000519 time 0.2976 (0.2929) loss 3.4776 (3.5264) grad_norm 1.4565 (1.5035) [2021-04-16 05:10:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][420/1251] eta 0:04:03 lr 0.000519 time 0.2767 (0.2924) loss 3.5209 (3.5272) grad_norm 1.4720 (1.5032) [2021-04-16 05:10:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][430/1251] eta 0:03:59 lr 0.000519 time 0.2744 (0.2920) loss 4.0244 (3.5234) grad_norm 1.4388 (1.5024) [2021-04-16 05:10:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][440/1251] eta 0:03:56 lr 0.000519 time 0.2675 (0.2915) loss 2.8790 (3.5235) grad_norm 1.4767 (1.5021) [2021-04-16 05:10:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][450/1251] eta 0:03:53 lr 0.000519 time 0.2644 (0.2911) loss 3.8674 (3.5253) grad_norm 1.3029 (1.5003) [2021-04-16 05:10:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][460/1251] eta 0:03:50 lr 0.000519 time 0.2624 (0.2908) loss 3.1362 (3.5267) grad_norm 1.5087 (1.5013) [2021-04-16 05:10:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][470/1251] eta 0:03:47 lr 0.000519 time 0.2712 (0.2907) loss 4.0396 (3.5273) grad_norm 1.5838 (1.5027) [2021-04-16 05:10:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][480/1251] eta 0:03:43 lr 0.000519 time 0.3016 (0.2905) loss 3.5668 (3.5267) grad_norm 1.4916 (1.5037) [2021-04-16 05:11:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][490/1251] eta 0:03:40 lr 0.000519 time 0.2781 (0.2902) loss 4.0556 (3.5254) grad_norm 1.2576 (1.5024) [2021-04-16 05:11:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][500/1251] eta 0:03:37 lr 0.000518 time 0.2756 (0.2900) loss 3.7976 (3.5275) grad_norm 1.6858 (1.5026) [2021-04-16 05:11:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][510/1251] eta 0:03:34 lr 0.000518 time 0.3028 (0.2899) loss 3.5429 (3.5272) grad_norm 1.4521 (1.5063) [2021-04-16 05:11:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][520/1251] eta 0:03:31 lr 0.000518 time 0.2851 (0.2896) loss 3.2133 (3.5303) grad_norm 1.3272 (1.5064) [2021-04-16 05:11:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][530/1251] eta 0:03:28 lr 0.000518 time 0.2678 (0.2894) loss 3.8764 (3.5378) grad_norm 1.6807 (1.5070) [2021-04-16 05:11:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][540/1251] eta 0:03:25 lr 0.000518 time 0.2678 (0.2894) loss 4.0606 (3.5367) grad_norm 1.4685 (1.5068) [2021-04-16 05:11:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][550/1251] eta 0:03:22 lr 0.000518 time 0.2898 (0.2891) loss 3.9417 (3.5398) grad_norm 1.4589 (1.5067) [2021-04-16 05:11:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][560/1251] eta 0:03:19 lr 0.000518 time 0.2727 (0.2889) loss 3.3288 (3.5430) grad_norm 1.6015 (1.5062) [2021-04-16 05:11:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][570/1251] eta 0:03:16 lr 0.000518 time 0.2717 (0.2887) loss 4.0517 (3.5446) grad_norm 1.4248 (1.5075) [2021-04-16 05:11:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][580/1251] eta 0:03:13 lr 0.000518 time 0.2930 (0.2888) loss 2.4719 (3.5415) grad_norm 1.5753 (1.5080) [2021-04-16 05:11:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][590/1251] eta 0:03:11 lr 0.000518 time 0.2836 (0.2890) loss 4.0516 (3.5433) grad_norm 1.5248 (1.5084) [2021-04-16 05:11:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][600/1251] eta 0:03:08 lr 0.000518 time 0.2683 (0.2888) loss 3.6275 (3.5425) grad_norm 1.5462 (1.5092) [2021-04-16 05:11:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][610/1251] eta 0:03:05 lr 0.000518 time 0.2857 (0.2887) loss 3.9715 (3.5421) grad_norm 1.6148 (1.5090) [2021-04-16 05:11:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][620/1251] eta 0:03:02 lr 0.000518 time 0.2511 (0.2885) loss 3.2493 (3.5427) grad_norm 1.5011 (1.5097) [2021-04-16 05:11:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][630/1251] eta 0:02:59 lr 0.000518 time 0.3112 (0.2886) loss 3.1621 (3.5393) grad_norm 1.8095 (1.5089) [2021-04-16 05:11:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][640/1251] eta 0:02:56 lr 0.000518 time 0.2838 (0.2883) loss 3.5983 (3.5435) grad_norm 1.2827 (1.5088) [2021-04-16 05:11:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][650/1251] eta 0:02:53 lr 0.000518 time 0.2862 (0.2883) loss 3.9332 (3.5403) grad_norm 1.5883 (1.5090) [2021-04-16 05:11:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][660/1251] eta 0:02:50 lr 0.000518 time 0.2723 (0.2881) loss 3.7651 (3.5427) grad_norm 1.5207 (1.5096) [2021-04-16 05:11:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][670/1251] eta 0:02:47 lr 0.000518 time 0.2769 (0.2880) loss 3.1256 (3.5347) grad_norm 1.3360 (1.5092) [2021-04-16 05:11:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][680/1251] eta 0:02:44 lr 0.000518 time 0.2777 (0.2880) loss 4.3295 (3.5342) grad_norm 1.3761 (1.5111) [2021-04-16 05:11:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][690/1251] eta 0:02:41 lr 0.000518 time 0.2578 (0.2880) loss 3.4964 (3.5323) grad_norm 1.5719 (1.5104) [2021-04-16 05:12:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][700/1251] eta 0:02:38 lr 0.000518 time 0.2704 (0.2878) loss 3.6559 (3.5308) grad_norm 1.3007 (1.5096) [2021-04-16 05:12:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][710/1251] eta 0:02:35 lr 0.000518 time 0.2902 (0.2877) loss 3.3189 (3.5300) grad_norm 1.4065 (1.5091) [2021-04-16 05:12:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][720/1251] eta 0:02:32 lr 0.000518 time 0.2898 (0.2878) loss 3.9198 (3.5298) grad_norm 1.5675 (1.5087) [2021-04-16 05:12:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][730/1251] eta 0:02:29 lr 0.000518 time 0.2860 (0.2876) loss 3.3671 (3.5304) grad_norm 1.3930 (1.5073) [2021-04-16 05:12:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][740/1251] eta 0:02:26 lr 0.000517 time 0.2628 (0.2874) loss 4.2273 (3.5344) grad_norm 1.4156 (1.5066) [2021-04-16 05:12:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][750/1251] eta 0:02:23 lr 0.000517 time 0.2547 (0.2872) loss 3.5413 (3.5300) grad_norm 1.8787 (1.5082) [2021-04-16 05:12:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][760/1251] eta 0:02:20 lr 0.000517 time 0.2802 (0.2871) loss 3.3991 (3.5298) grad_norm 2.0550 (1.5092) [2021-04-16 05:12:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][770/1251] eta 0:02:18 lr 0.000517 time 0.3143 (0.2870) loss 3.9609 (3.5299) grad_norm 1.5262 (1.5097) [2021-04-16 05:12:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][780/1251] eta 0:02:15 lr 0.000517 time 0.2744 (0.2868) loss 4.3310 (3.5283) grad_norm 1.5845 (1.5099) [2021-04-16 05:12:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][790/1251] eta 0:02:12 lr 0.000517 time 0.2747 (0.2866) loss 3.6573 (3.5288) grad_norm 1.3430 (1.5107) [2021-04-16 05:12:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][800/1251] eta 0:02:09 lr 0.000517 time 0.2610 (0.2865) loss 2.3115 (3.5294) grad_norm 1.6044 (1.5110) [2021-04-16 05:12:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][810/1251] eta 0:02:06 lr 0.000517 time 0.2942 (0.2864) loss 4.0903 (3.5327) grad_norm 1.4070 (1.5104) [2021-04-16 05:12:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][820/1251] eta 0:02:03 lr 0.000517 time 0.2800 (0.2863) loss 3.2341 (3.5355) grad_norm 1.6151 (1.5103) [2021-04-16 05:12:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][830/1251] eta 0:02:00 lr 0.000517 time 0.2740 (0.2861) loss 3.0120 (3.5330) grad_norm 1.5484 (1.5106) [2021-04-16 05:12:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][840/1251] eta 0:01:57 lr 0.000517 time 0.3170 (0.2861) loss 4.0732 (3.5337) grad_norm 1.4469 (1.5105) [2021-04-16 05:12:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][850/1251] eta 0:01:54 lr 0.000517 time 0.2837 (0.2860) loss 4.3120 (3.5324) grad_norm 1.9413 (1.5109) [2021-04-16 05:12:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][860/1251] eta 0:01:51 lr 0.000517 time 0.2853 (0.2859) loss 4.0070 (3.5329) grad_norm 1.5359 (1.5115) [2021-04-16 05:12:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][870/1251] eta 0:01:48 lr 0.000517 time 0.2893 (0.2858) loss 3.0355 (3.5301) grad_norm 1.3198 (1.5112) [2021-04-16 05:12:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][880/1251] eta 0:01:45 lr 0.000517 time 0.2811 (0.2857) loss 4.1046 (3.5259) grad_norm 1.3271 (1.5106) [2021-04-16 05:12:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][890/1251] eta 0:01:43 lr 0.000517 time 0.2927 (0.2858) loss 3.0316 (3.5285) grad_norm 1.7854 (1.5109) [2021-04-16 05:12:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][900/1251] eta 0:01:40 lr 0.000517 time 0.2535 (0.2857) loss 3.4322 (3.5306) grad_norm 1.5984 (1.5109) [2021-04-16 05:12:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][910/1251] eta 0:01:37 lr 0.000517 time 0.2436 (0.2856) loss 3.3499 (3.5311) grad_norm 1.7281 (1.5106) [2021-04-16 05:13:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][920/1251] eta 0:01:34 lr 0.000517 time 0.2757 (0.2856) loss 3.2210 (3.5327) grad_norm 1.2893 (1.5102) [2021-04-16 05:13:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][930/1251] eta 0:01:31 lr 0.000517 time 0.2698 (0.2856) loss 3.6869 (3.5327) grad_norm 1.5039 (1.5105) [2021-04-16 05:13:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][940/1251] eta 0:01:28 lr 0.000517 time 0.2761 (0.2855) loss 4.1084 (3.5358) grad_norm 1.8964 (1.5099) [2021-04-16 05:13:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][950/1251] eta 0:01:25 lr 0.000517 time 0.2557 (0.2854) loss 3.2807 (3.5379) grad_norm 1.6031 (1.5109) [2021-04-16 05:13:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][960/1251] eta 0:01:23 lr 0.000517 time 0.2569 (0.2853) loss 2.6588 (3.5386) grad_norm 1.5479 (1.5105) [2021-04-16 05:13:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][970/1251] eta 0:01:20 lr 0.000517 time 0.2579 (0.2853) loss 4.0991 (3.5394) grad_norm 1.4329 (1.5109) [2021-04-16 05:13:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][980/1251] eta 0:01:17 lr 0.000516 time 0.2789 (0.2851) loss 2.4269 (3.5374) grad_norm 1.4404 (1.5106) [2021-04-16 05:13:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][990/1251] eta 0:01:14 lr 0.000516 time 0.2606 (0.2850) loss 3.5861 (3.5381) grad_norm 1.5540 (1.5107) [2021-04-16 05:13:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][1000/1251] eta 0:01:11 lr 0.000516 time 0.2723 (0.2849) loss 3.3675 (3.5373) grad_norm 1.4605 (1.5113) [2021-04-16 05:13:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][1010/1251] eta 0:01:08 lr 0.000516 time 0.2725 (0.2848) loss 2.7708 (3.5360) grad_norm 1.3975 (1.5107) [2021-04-16 05:13:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][1020/1251] eta 0:01:05 lr 0.000516 time 0.2455 (0.2847) loss 2.9378 (3.5361) grad_norm 2.0773 (1.5120) [2021-04-16 05:13:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][1030/1251] eta 0:01:02 lr 0.000516 time 0.2919 (0.2847) loss 3.4508 (3.5377) grad_norm 1.3377 (1.5123) [2021-04-16 05:13:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][1040/1251] eta 0:01:00 lr 0.000516 time 0.2652 (0.2847) loss 4.4681 (3.5373) grad_norm 1.3990 (1.5118) [2021-04-16 05:13:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][1050/1251] eta 0:00:57 lr 0.000516 time 0.2925 (0.2847) loss 4.1504 (3.5350) grad_norm 1.6087 (1.5117) [2021-04-16 05:13:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][1060/1251] eta 0:00:54 lr 0.000516 time 0.2741 (0.2846) loss 3.7606 (3.5376) grad_norm 1.3971 (1.5120) [2021-04-16 05:13:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][1070/1251] eta 0:00:51 lr 0.000516 time 0.2767 (0.2846) loss 3.9783 (3.5406) grad_norm 1.6755 (1.5126) [2021-04-16 05:13:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][1080/1251] eta 0:00:48 lr 0.000516 time 0.2983 (0.2845) loss 2.8294 (3.5411) grad_norm 1.3732 (1.5125) [2021-04-16 05:13:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][1090/1251] eta 0:00:45 lr 0.000516 time 0.2917 (0.2844) loss 3.3219 (3.5396) grad_norm 1.3941 (1.5121) [2021-04-16 05:13:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][1100/1251] eta 0:00:42 lr 0.000516 time 0.2864 (0.2843) loss 2.6479 (3.5381) grad_norm 1.4526 (1.5110) [2021-04-16 05:13:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][1110/1251] eta 0:00:40 lr 0.000516 time 0.2806 (0.2843) loss 4.1428 (3.5388) grad_norm 1.3704 (1.5108) [2021-04-16 05:13:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][1120/1251] eta 0:00:37 lr 0.000516 time 0.2853 (0.2844) loss 2.9349 (3.5402) grad_norm 1.7577 (1.5116) [2021-04-16 05:14:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][1130/1251] eta 0:00:34 lr 0.000516 time 0.2683 (0.2843) loss 3.2092 (3.5394) grad_norm 1.5434 (1.5120) [2021-04-16 05:14:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][1140/1251] eta 0:00:31 lr 0.000516 time 0.2828 (0.2842) loss 3.1928 (3.5400) grad_norm 1.4461 (1.5126) [2021-04-16 05:14:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][1150/1251] eta 0:00:28 lr 0.000516 time 0.2581 (0.2842) loss 3.4999 (3.5412) grad_norm 1.5581 (1.5125) [2021-04-16 05:14:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][1160/1251] eta 0:00:25 lr 0.000516 time 0.2803 (0.2842) loss 3.0940 (3.5429) grad_norm 1.7173 (1.5124) [2021-04-16 05:14:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][1170/1251] eta 0:00:23 lr 0.000516 time 0.3062 (0.2843) loss 3.5966 (3.5454) grad_norm 1.3439 (1.5129) [2021-04-16 05:14:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][1180/1251] eta 0:00:20 lr 0.000516 time 0.2865 (0.2842) loss 3.3444 (3.5466) grad_norm 1.3423 (1.5138) [2021-04-16 05:14:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][1190/1251] eta 0:00:17 lr 0.000516 time 0.2647 (0.2842) loss 2.7522 (3.5453) grad_norm 1.4971 (1.5138) [2021-04-16 05:14:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][1200/1251] eta 0:00:14 lr 0.000516 time 0.2920 (0.2841) loss 3.0480 (3.5457) grad_norm 1.3509 (1.5135) [2021-04-16 05:14:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][1210/1251] eta 0:00:11 lr 0.000516 time 0.2792 (0.2841) loss 3.7082 (3.5461) grad_norm 1.2118 (1.5133) [2021-04-16 05:14:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][1220/1251] eta 0:00:08 lr 0.000515 time 0.2628 (0.2840) loss 3.9414 (3.5444) grad_norm 1.2706 (1.5131) [2021-04-16 05:14:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][1230/1251] eta 0:00:05 lr 0.000515 time 0.2761 (0.2841) loss 3.7169 (3.5432) grad_norm 1.6754 (1.5130) [2021-04-16 05:14:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][1240/1251] eta 0:00:03 lr 0.000515 time 0.2483 (0.2840) loss 4.1398 (3.5439) grad_norm 1.7432 (1.5133) [2021-04-16 05:14:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [147/300][1250/1251] eta 0:00:00 lr 0.000515 time 0.2518 (0.2837) loss 3.5295 (3.5425) grad_norm 1.2700 (1.5139) [2021-04-16 05:14:37 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 147 training takes 0:05:57 [2021-04-16 05:14:37 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_147.pth saving...... [2021-04-16 05:14:50 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_147.pth saved !!! [2021-04-16 05:14:51 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.137 (1.137) Loss 1.0898 (1.0898) Acc@1 74.023 (74.023) Acc@5 93.164 (93.164) [2021-04-16 05:14:53 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.129 (0.205) Loss 0.9941 (1.0252) Acc@1 77.051 (75.692) Acc@5 93.945 (93.368) [2021-04-16 05:14:55 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.136 (0.228) Loss 0.9829 (1.0321) Acc@1 75.977 (75.512) Acc@5 94.043 (93.262) [2021-04-16 05:14:57 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.127 (0.229) Loss 1.0129 (1.0300) Acc@1 76.855 (75.573) Acc@5 92.383 (93.252) [2021-04-16 05:14:59 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.219) Loss 1.0228 (1.0331) Acc@1 74.902 (75.353) Acc@5 93.457 (93.216) [2021-04-16 05:15:04 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 75.500 Acc@5 93.222 [2021-04-16 05:15:04 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 75.5% [2021-04-16 05:15:04 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 75.98% [2021-04-16 05:15:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][0/1251] eta 3:43:20 lr 0.000515 time 10.7119 (10.7119) loss 3.9673 (3.9673) grad_norm 1.2828 (1.2828) [2021-04-16 05:15:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][10/1251] eta 0:25:48 lr 0.000515 time 0.2634 (1.2477) loss 4.1626 (3.6722) grad_norm 1.4777 (1.4992) [2021-04-16 05:15:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][20/1251] eta 0:16:06 lr 0.000515 time 0.2660 (0.7854) loss 3.1142 (3.4315) grad_norm 1.3205 (1.4621) [2021-04-16 05:15:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][30/1251] eta 0:12:39 lr 0.000515 time 0.2795 (0.6219) loss 2.5891 (3.4057) grad_norm 1.3827 (1.4510) [2021-04-16 05:15:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][40/1251] eta 0:10:52 lr 0.000515 time 0.2688 (0.5388) loss 3.9170 (3.4495) grad_norm 1.6366 (1.4928) [2021-04-16 05:15:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][50/1251] eta 0:09:45 lr 0.000515 time 0.2567 (0.4877) loss 3.9370 (3.4678) grad_norm 1.7460 (1.4974) [2021-04-16 05:15:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][60/1251] eta 0:09:00 lr 0.000515 time 0.2739 (0.4538) loss 3.2768 (3.4635) grad_norm 1.5245 (1.4990) [2021-04-16 05:15:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][70/1251] eta 0:08:26 lr 0.000515 time 0.2902 (0.4289) loss 3.9662 (3.4756) grad_norm 1.7236 (1.4970) [2021-04-16 05:15:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][80/1251] eta 0:07:59 lr 0.000515 time 0.2445 (0.4099) loss 3.7997 (3.4702) grad_norm 1.6095 (1.5052) [2021-04-16 05:15:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][90/1251] eta 0:07:38 lr 0.000515 time 0.2638 (0.3951) loss 3.3340 (3.4852) grad_norm 1.5505 (1.5075) [2021-04-16 05:15:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][100/1251] eta 0:07:22 lr 0.000515 time 0.2551 (0.3844) loss 4.1327 (3.4909) grad_norm 1.2656 (1.4983) [2021-04-16 05:15:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][110/1251] eta 0:07:07 lr 0.000515 time 0.2680 (0.3748) loss 3.2728 (3.4985) grad_norm 1.5850 (1.4942) [2021-04-16 05:15:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][120/1251] eta 0:06:56 lr 0.000515 time 0.3103 (0.3680) loss 3.5979 (3.5097) grad_norm 1.5133 (1.4984) [2021-04-16 05:15:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][130/1251] eta 0:06:45 lr 0.000515 time 0.2809 (0.3614) loss 3.5899 (3.5117) grad_norm 1.4859 (1.4972) [2021-04-16 05:15:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][140/1251] eta 0:06:35 lr 0.000515 time 0.2730 (0.3558) loss 3.8608 (3.5173) grad_norm 1.6665 (1.4968) [2021-04-16 05:15:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][150/1251] eta 0:06:26 lr 0.000515 time 0.2704 (0.3509) loss 3.9079 (3.5269) grad_norm 1.7182 (1.4950) [2021-04-16 05:16:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][160/1251] eta 0:06:18 lr 0.000515 time 0.3135 (0.3466) loss 3.3970 (3.5323) grad_norm 1.6101 (1.5014) [2021-04-16 05:16:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][170/1251] eta 0:06:10 lr 0.000515 time 0.2767 (0.3431) loss 3.9062 (3.5416) grad_norm 1.4001 (1.4982) [2021-04-16 05:16:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][180/1251] eta 0:06:03 lr 0.000515 time 0.2808 (0.3395) loss 3.3565 (3.5362) grad_norm 1.6798 (1.4944) [2021-04-16 05:16:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][190/1251] eta 0:05:56 lr 0.000515 time 0.2764 (0.3362) loss 2.6560 (3.5475) grad_norm 1.7615 (1.4982) [2021-04-16 05:16:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][200/1251] eta 0:05:50 lr 0.000515 time 0.2543 (0.3332) loss 3.7923 (3.5539) grad_norm 1.6483 (1.5040) [2021-04-16 05:16:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][210/1251] eta 0:05:44 lr 0.000514 time 0.2884 (0.3307) loss 2.3683 (3.5531) grad_norm 1.8104 (1.5095) [2021-04-16 05:16:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][220/1251] eta 0:05:38 lr 0.000514 time 0.2596 (0.3282) loss 3.0859 (3.5542) grad_norm 1.7640 (1.5112) [2021-04-16 05:16:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][230/1251] eta 0:05:32 lr 0.000514 time 0.2873 (0.3258) loss 2.9598 (3.5572) grad_norm 1.3152 (1.5200) [2021-04-16 05:16:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][240/1251] eta 0:05:27 lr 0.000514 time 0.3070 (0.3239) loss 3.6242 (3.5577) grad_norm 1.4923 (1.5188) [2021-04-16 05:16:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][250/1251] eta 0:05:22 lr 0.000514 time 0.2627 (0.3226) loss 3.5028 (3.5619) grad_norm 1.3905 (1.5188) [2021-04-16 05:16:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][260/1251] eta 0:05:17 lr 0.000514 time 0.2738 (0.3207) loss 4.3506 (3.5557) grad_norm 1.4892 (1.5201) [2021-04-16 05:16:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][270/1251] eta 0:05:12 lr 0.000514 time 0.2885 (0.3190) loss 3.4591 (3.5510) grad_norm 1.8591 (1.5219) [2021-04-16 05:16:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][280/1251] eta 0:05:08 lr 0.000514 time 0.2796 (0.3174) loss 3.4993 (3.5458) grad_norm 1.6728 (1.5245) [2021-04-16 05:16:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][290/1251] eta 0:05:03 lr 0.000514 time 0.2685 (0.3159) loss 3.4922 (3.5467) grad_norm 1.5106 (1.5234) [2021-04-16 05:16:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][300/1251] eta 0:04:59 lr 0.000514 time 0.2581 (0.3146) loss 2.4510 (3.5395) grad_norm 1.6594 (1.5252) [2021-04-16 05:16:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][310/1251] eta 0:04:54 lr 0.000514 time 0.2743 (0.3133) loss 3.9245 (3.5470) grad_norm 1.5474 (1.5269) [2021-04-16 05:16:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][320/1251] eta 0:04:50 lr 0.000514 time 0.3284 (0.3121) loss 3.7620 (3.5488) grad_norm 1.5587 (1.5310) [2021-04-16 05:16:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][330/1251] eta 0:04:46 lr 0.000514 time 0.2693 (0.3116) loss 4.0685 (3.5459) grad_norm 1.3875 (1.5282) [2021-04-16 05:16:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][340/1251] eta 0:04:42 lr 0.000514 time 0.2675 (0.3105) loss 3.7320 (3.5466) grad_norm 1.6198 (1.5285) [2021-04-16 05:16:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][350/1251] eta 0:04:38 lr 0.000514 time 0.2777 (0.3096) loss 2.8782 (3.5485) grad_norm 1.4701 (1.5284) [2021-04-16 05:16:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][360/1251] eta 0:04:35 lr 0.000514 time 0.2932 (0.3091) loss 3.8253 (3.5497) grad_norm 1.3585 (1.5274) [2021-04-16 05:16:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][370/1251] eta 0:04:31 lr 0.000514 time 0.2677 (0.3087) loss 2.7929 (3.5478) grad_norm 1.3906 (1.5240) [2021-04-16 05:17:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][380/1251] eta 0:04:28 lr 0.000514 time 0.2746 (0.3080) loss 4.3190 (3.5493) grad_norm 1.3295 (1.5218) [2021-04-16 05:17:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][390/1251] eta 0:04:24 lr 0.000514 time 0.2885 (0.3073) loss 3.9763 (3.5527) grad_norm 1.6793 (1.5203) [2021-04-16 05:17:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][400/1251] eta 0:04:20 lr 0.000514 time 0.2781 (0.3066) loss 2.3717 (3.5418) grad_norm 1.2987 (1.5209) [2021-04-16 05:17:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][410/1251] eta 0:04:17 lr 0.000514 time 0.2682 (0.3059) loss 3.3052 (3.5398) grad_norm 1.4154 (1.5213) [2021-04-16 05:17:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][420/1251] eta 0:04:13 lr 0.000514 time 0.2572 (0.3052) loss 2.8453 (3.5328) grad_norm 1.8102 (1.5213) [2021-04-16 05:17:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][430/1251] eta 0:04:10 lr 0.000514 time 0.2726 (0.3049) loss 3.7560 (3.5272) grad_norm 1.7475 (1.5223) [2021-04-16 05:17:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][440/1251] eta 0:04:06 lr 0.000514 time 0.2471 (0.3044) loss 4.3434 (3.5283) grad_norm 1.5366 (1.5224) [2021-04-16 05:17:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][450/1251] eta 0:04:03 lr 0.000514 time 0.2708 (0.3038) loss 3.1429 (3.5335) grad_norm 1.3837 (1.5216) [2021-04-16 05:17:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][460/1251] eta 0:03:59 lr 0.000513 time 0.3117 (0.3033) loss 4.2942 (3.5396) grad_norm 1.3820 (1.5207) [2021-04-16 05:17:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][470/1251] eta 0:03:56 lr 0.000513 time 0.2820 (0.3028) loss 3.5044 (3.5421) grad_norm 1.7513 (1.5208) [2021-04-16 05:17:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][480/1251] eta 0:03:53 lr 0.000513 time 0.2902 (0.3024) loss 2.1597 (3.5341) grad_norm 2.1442 (1.5240) [2021-04-16 05:17:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][490/1251] eta 0:03:49 lr 0.000513 time 0.2866 (0.3022) loss 3.9764 (3.5363) grad_norm 1.3976 (1.5246) [2021-04-16 05:17:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][500/1251] eta 0:03:46 lr 0.000513 time 0.2972 (0.3017) loss 3.2074 (3.5394) grad_norm 1.8091 (1.5275) [2021-04-16 05:17:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][510/1251] eta 0:03:43 lr 0.000513 time 0.2957 (0.3013) loss 4.1215 (3.5396) grad_norm 1.3906 (1.5267) [2021-04-16 05:17:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][520/1251] eta 0:03:40 lr 0.000513 time 0.2818 (0.3010) loss 4.0648 (3.5452) grad_norm 1.5809 (1.5268) [2021-04-16 05:17:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][530/1251] eta 0:03:36 lr 0.000513 time 0.2865 (0.3007) loss 4.1869 (3.5480) grad_norm 1.3816 (1.5259) [2021-04-16 05:17:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][540/1251] eta 0:03:33 lr 0.000513 time 0.2756 (0.3002) loss 4.3280 (3.5481) grad_norm 1.5128 (1.5272) [2021-04-16 05:17:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][550/1251] eta 0:03:30 lr 0.000513 time 0.2732 (0.2997) loss 3.9596 (3.5504) grad_norm 1.3738 (1.5255) [2021-04-16 05:17:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][560/1251] eta 0:03:26 lr 0.000513 time 0.2518 (0.2993) loss 3.4218 (3.5503) grad_norm 1.4930 (1.5241) [2021-04-16 05:17:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][570/1251] eta 0:03:23 lr 0.000513 time 0.2718 (0.2990) loss 4.2537 (3.5557) grad_norm 1.5620 (1.5243) [2021-04-16 05:17:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][580/1251] eta 0:03:20 lr 0.000513 time 0.2718 (0.2988) loss 3.8154 (3.5611) grad_norm 1.4228 (1.5246) [2021-04-16 05:18:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][590/1251] eta 0:03:17 lr 0.000513 time 0.2657 (0.2984) loss 4.2135 (3.5555) grad_norm 1.5163 (1.5254) [2021-04-16 05:18:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][600/1251] eta 0:03:14 lr 0.000513 time 0.2816 (0.2980) loss 3.8840 (3.5555) grad_norm 1.6690 (1.5249) [2021-04-16 05:18:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][610/1251] eta 0:03:10 lr 0.000513 time 0.2779 (0.2977) loss 2.7376 (3.5489) grad_norm 1.4965 (1.5239) [2021-04-16 05:18:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][620/1251] eta 0:03:07 lr 0.000513 time 0.2860 (0.2973) loss 3.8868 (3.5481) grad_norm 1.3192 (1.5229) [2021-04-16 05:18:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][630/1251] eta 0:03:04 lr 0.000513 time 0.2797 (0.2970) loss 2.3819 (3.5446) grad_norm 1.7397 (1.5226) [2021-04-16 05:18:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][640/1251] eta 0:03:01 lr 0.000513 time 0.2721 (0.2968) loss 3.3263 (3.5453) grad_norm 1.6323 (1.5227) [2021-04-16 05:18:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][650/1251] eta 0:02:58 lr 0.000513 time 0.2770 (0.2965) loss 2.3375 (3.5434) grad_norm 1.3130 (1.5222) [2021-04-16 05:18:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][660/1251] eta 0:02:55 lr 0.000513 time 0.2880 (0.2963) loss 2.1504 (3.5421) grad_norm 1.4604 (1.5227) [2021-04-16 05:18:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][670/1251] eta 0:02:51 lr 0.000513 time 0.2805 (0.2960) loss 4.3571 (3.5438) grad_norm 1.2961 (1.5217) [2021-04-16 05:18:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][680/1251] eta 0:02:49 lr 0.000513 time 0.3018 (0.2960) loss 3.5659 (3.5417) grad_norm 1.4411 (1.5219) [2021-04-16 05:18:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][690/1251] eta 0:02:46 lr 0.000513 time 0.2775 (0.2960) loss 3.0805 (3.5403) grad_norm 1.3173 (1.5218) [2021-04-16 05:18:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][700/1251] eta 0:02:42 lr 0.000512 time 0.2957 (0.2958) loss 4.1379 (3.5370) grad_norm 1.5835 (1.5227) [2021-04-16 05:18:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][710/1251] eta 0:02:39 lr 0.000512 time 0.2831 (0.2955) loss 4.0170 (3.5373) grad_norm 1.5766 (1.5233) [2021-04-16 05:18:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][720/1251] eta 0:02:36 lr 0.000512 time 0.2467 (0.2954) loss 3.7025 (3.5415) grad_norm 1.6288 (1.5235) [2021-04-16 05:18:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][730/1251] eta 0:02:33 lr 0.000512 time 0.2829 (0.2952) loss 3.3219 (3.5423) grad_norm 1.5820 (1.5234) [2021-04-16 05:18:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][740/1251] eta 0:02:30 lr 0.000512 time 0.2700 (0.2950) loss 3.2998 (3.5417) grad_norm 2.0953 (1.5226) [2021-04-16 05:18:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][750/1251] eta 0:02:27 lr 0.000512 time 0.2899 (0.2947) loss 3.6959 (3.5449) grad_norm 1.5779 (1.5228) [2021-04-16 05:18:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][760/1251] eta 0:02:24 lr 0.000512 time 0.2810 (0.2944) loss 4.0761 (3.5452) grad_norm 1.2817 (1.5228) [2021-04-16 05:18:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][770/1251] eta 0:02:21 lr 0.000512 time 0.2825 (0.2943) loss 3.5343 (3.5445) grad_norm 1.3010 (1.5218) [2021-04-16 05:18:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][780/1251] eta 0:02:18 lr 0.000512 time 0.2535 (0.2940) loss 3.5612 (3.5411) grad_norm 1.3232 (1.5206) [2021-04-16 05:18:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][790/1251] eta 0:02:15 lr 0.000512 time 0.2669 (0.2937) loss 3.8450 (3.5414) grad_norm 1.2308 (1.5196) [2021-04-16 05:18:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][800/1251] eta 0:02:12 lr 0.000512 time 0.2664 (0.2935) loss 4.2319 (3.5408) grad_norm 1.7706 (1.5205) [2021-04-16 05:19:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][810/1251] eta 0:02:09 lr 0.000512 time 0.2860 (0.2932) loss 3.7550 (3.5417) grad_norm 1.6516 (1.5214) [2021-04-16 05:19:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][820/1251] eta 0:02:06 lr 0.000512 time 0.2552 (0.2930) loss 4.0665 (3.5425) grad_norm 1.5379 (1.5222) [2021-04-16 05:19:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][830/1251] eta 0:02:03 lr 0.000512 time 0.2739 (0.2929) loss 3.9270 (3.5429) grad_norm 1.3625 (1.5233) [2021-04-16 05:19:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][840/1251] eta 0:02:00 lr 0.000512 time 0.2929 (0.2927) loss 3.7812 (3.5430) grad_norm 1.3182 (1.5221) [2021-04-16 05:19:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][850/1251] eta 0:01:57 lr 0.000512 time 0.2501 (0.2925) loss 3.9789 (3.5425) grad_norm 1.3777 (1.5215) [2021-04-16 05:19:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][860/1251] eta 0:01:54 lr 0.000512 time 0.2974 (0.2924) loss 4.1018 (3.5434) grad_norm 1.4057 (1.5219) [2021-04-16 05:19:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][870/1251] eta 0:01:51 lr 0.000512 time 0.3134 (0.2922) loss 3.9694 (3.5439) grad_norm 1.5926 (1.5232) [2021-04-16 05:19:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][880/1251] eta 0:01:48 lr 0.000512 time 0.2823 (0.2920) loss 2.7493 (3.5433) grad_norm 1.3663 (1.5234) [2021-04-16 05:19:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][890/1251] eta 0:01:45 lr 0.000512 time 0.2758 (0.2918) loss 3.7599 (3.5422) grad_norm 1.3951 (1.5230) [2021-04-16 05:19:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][900/1251] eta 0:01:42 lr 0.000512 time 0.3908 (0.2918) loss 4.3252 (3.5399) grad_norm 1.5692 (1.5231) [2021-04-16 05:19:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][910/1251] eta 0:01:39 lr 0.000512 time 0.2565 (0.2916) loss 2.6745 (3.5393) grad_norm 1.2807 (1.5222) [2021-04-16 05:19:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][920/1251] eta 0:01:36 lr 0.000512 time 0.2693 (0.2915) loss 4.0495 (3.5399) grad_norm 1.2472 (1.5208) [2021-04-16 05:19:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][930/1251] eta 0:01:33 lr 0.000512 time 0.2858 (0.2913) loss 4.0899 (3.5419) grad_norm 1.6039 (1.5210) [2021-04-16 05:19:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][940/1251] eta 0:01:30 lr 0.000511 time 0.4257 (0.2915) loss 3.9967 (3.5362) grad_norm 1.5332 (1.5204) [2021-04-16 05:19:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][950/1251] eta 0:01:27 lr 0.000511 time 0.2600 (0.2913) loss 3.2658 (3.5366) grad_norm 1.4161 (1.5204) [2021-04-16 05:19:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][960/1251] eta 0:01:24 lr 0.000511 time 0.2886 (0.2911) loss 2.4009 (3.5345) grad_norm 1.3605 (1.5198) [2021-04-16 05:19:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][970/1251] eta 0:01:21 lr 0.000511 time 0.2788 (0.2910) loss 3.7729 (3.5327) grad_norm 1.7285 (1.5191) [2021-04-16 05:19:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][980/1251] eta 0:01:18 lr 0.000511 time 0.2873 (0.2908) loss 3.9268 (3.5352) grad_norm 1.4286 (1.5193) [2021-04-16 05:19:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][990/1251] eta 0:01:15 lr 0.000511 time 0.2888 (0.2906) loss 2.7505 (3.5339) grad_norm 1.5480 (1.5186) [2021-04-16 05:19:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][1000/1251] eta 0:01:12 lr 0.000511 time 0.2893 (0.2905) loss 3.6073 (3.5336) grad_norm 1.4127 (1.5184) [2021-04-16 05:19:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][1010/1251] eta 0:01:09 lr 0.000511 time 0.2680 (0.2905) loss 3.2912 (3.5318) grad_norm 1.6931 (1.5200) [2021-04-16 05:20:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][1020/1251] eta 0:01:07 lr 0.000511 time 0.2657 (0.2904) loss 3.3218 (3.5302) grad_norm 1.3023 (1.5196) [2021-04-16 05:20:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][1030/1251] eta 0:01:04 lr 0.000511 time 0.2755 (0.2903) loss 3.5718 (3.5325) grad_norm 1.3685 (1.5190) [2021-04-16 05:20:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][1040/1251] eta 0:01:01 lr 0.000511 time 0.2747 (0.2901) loss 3.7846 (3.5352) grad_norm 1.3061 (1.5193) [2021-04-16 05:20:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][1050/1251] eta 0:00:58 lr 0.000511 time 0.2533 (0.2900) loss 2.9720 (3.5340) grad_norm 1.4405 (1.5196) [2021-04-16 05:20:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][1060/1251] eta 0:00:55 lr 0.000511 time 0.2907 (0.2900) loss 2.9050 (3.5359) grad_norm 1.3850 (1.5199) [2021-04-16 05:20:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][1070/1251] eta 0:00:52 lr 0.000511 time 0.2767 (0.2900) loss 4.3155 (3.5388) grad_norm 1.5503 (1.5194) [2021-04-16 05:20:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][1080/1251] eta 0:00:49 lr 0.000511 time 0.2879 (0.2899) loss 3.9655 (3.5377) grad_norm 1.3720 (1.5190) [2021-04-16 05:20:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][1090/1251] eta 0:00:46 lr 0.000511 time 0.2586 (0.2897) loss 3.8264 (3.5358) grad_norm 1.5356 (1.5191) [2021-04-16 05:20:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][1100/1251] eta 0:00:43 lr 0.000511 time 0.3050 (0.2896) loss 3.6174 (3.5371) grad_norm 1.7822 (1.5198) [2021-04-16 05:20:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][1110/1251] eta 0:00:40 lr 0.000511 time 0.2846 (0.2895) loss 3.8164 (3.5382) grad_norm 1.5063 (1.5202) [2021-04-16 05:20:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][1120/1251] eta 0:00:37 lr 0.000511 time 0.2715 (0.2894) loss 4.1128 (3.5387) grad_norm 1.7247 (1.5199) [2021-04-16 05:20:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][1130/1251] eta 0:00:35 lr 0.000511 time 0.2691 (0.2893) loss 3.7186 (3.5402) grad_norm 1.4293 (1.5195) [2021-04-16 05:20:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][1140/1251] eta 0:00:32 lr 0.000511 time 0.2746 (0.2893) loss 3.3466 (3.5394) grad_norm 1.4688 (1.5185) [2021-04-16 05:20:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][1150/1251] eta 0:00:29 lr 0.000511 time 0.2913 (0.2892) loss 3.2041 (3.5407) grad_norm 1.4595 (1.5183) [2021-04-16 05:20:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][1160/1251] eta 0:00:26 lr 0.000511 time 0.2897 (0.2892) loss 4.1202 (3.5428) grad_norm 1.5524 (1.5187) [2021-04-16 05:20:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][1170/1251] eta 0:00:23 lr 0.000511 time 0.2658 (0.2891) loss 3.5906 (3.5448) grad_norm 1.4425 (1.5190) [2021-04-16 05:20:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][1180/1251] eta 0:00:20 lr 0.000510 time 0.2627 (0.2890) loss 2.2414 (3.5441) grad_norm 1.6756 (1.5190) [2021-04-16 05:20:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][1190/1251] eta 0:00:17 lr 0.000510 time 0.2621 (0.2889) loss 2.5200 (3.5421) grad_norm 1.6002 (1.5195) [2021-04-16 05:20:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][1200/1251] eta 0:00:14 lr 0.000510 time 0.2660 (0.2888) loss 3.0904 (3.5422) grad_norm 1.6042 (1.5200) [2021-04-16 05:20:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][1210/1251] eta 0:00:11 lr 0.000510 time 0.3974 (0.2888) loss 2.8685 (3.5436) grad_norm 1.4748 (1.5197) [2021-04-16 05:20:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][1220/1251] eta 0:00:08 lr 0.000510 time 0.4076 (0.2890) loss 2.8737 (3.5433) grad_norm 1.4503 (1.5201) [2021-04-16 05:21:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][1230/1251] eta 0:00:06 lr 0.000510 time 0.2724 (0.2889) loss 3.5157 (3.5421) grad_norm 1.4100 (1.5193) [2021-04-16 05:21:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][1240/1251] eta 0:00:03 lr 0.000510 time 0.2486 (0.2887) loss 3.0576 (3.5418) grad_norm 1.6244 (1.5191) [2021-04-16 05:21:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [148/300][1250/1251] eta 0:00:00 lr 0.000510 time 0.2493 (0.2884) loss 3.4363 (3.5405) grad_norm 1.6585 (1.5181) [2021-04-16 05:21:07 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 148 training takes 0:06:03 [2021-04-16 05:21:07 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_148.pth saving...... [2021-04-16 05:21:14 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_148.pth saved !!! [2021-04-16 05:21:15 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.156 (1.156) Loss 1.0756 (1.0756) Acc@1 73.926 (73.926) Acc@5 92.676 (92.676) [2021-04-16 05:21:16 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.115 (0.230) Loss 1.0811 (1.0241) Acc@1 74.512 (75.639) Acc@5 92.090 (93.164) [2021-04-16 05:21:19 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.164 (0.222) Loss 0.9705 (1.0187) Acc@1 76.562 (75.846) Acc@5 94.531 (93.294) [2021-04-16 05:21:21 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.378 (0.233) Loss 1.0328 (1.0182) Acc@1 74.121 (75.923) Acc@5 92.676 (93.249) [2021-04-16 05:21:23 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.213) Loss 1.0013 (1.0182) Acc@1 76.562 (75.843) Acc@5 94.043 (93.274) [2021-04-16 05:21:27 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 75.860 Acc@5 93.282 [2021-04-16 05:21:27 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 75.9% [2021-04-16 05:21:27 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 75.98% [2021-04-16 05:21:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][0/1251] eta 2:44:35 lr 0.000510 time 7.8939 (7.8939) loss 3.5678 (3.5678) grad_norm 1.3150 (1.3150) [2021-04-16 05:21:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][10/1251] eta 0:20:00 lr 0.000510 time 0.2894 (0.9673) loss 3.4393 (3.6387) grad_norm 1.6000 (1.5762) [2021-04-16 05:21:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][20/1251] eta 0:13:07 lr 0.000510 time 0.2692 (0.6399) loss 2.7262 (3.5200) grad_norm 1.5064 (1.5801) [2021-04-16 05:21:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][30/1251] eta 0:10:43 lr 0.000510 time 0.2578 (0.5269) loss 2.6531 (3.4920) grad_norm 1.3638 (1.5632) [2021-04-16 05:21:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][40/1251] eta 0:09:23 lr 0.000510 time 0.2764 (0.4657) loss 3.1306 (3.5003) grad_norm 1.5179 (1.5808) [2021-04-16 05:21:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][50/1251] eta 0:08:34 lr 0.000510 time 0.2683 (0.4287) loss 3.8656 (3.5137) grad_norm 1.3752 (1.5781) [2021-04-16 05:21:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][60/1251] eta 0:08:00 lr 0.000510 time 0.2720 (0.4034) loss 2.3958 (3.5071) grad_norm 1.5365 (1.5779) [2021-04-16 05:21:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][70/1251] eta 0:07:35 lr 0.000510 time 0.2613 (0.3857) loss 3.7497 (3.5055) grad_norm 1.3853 (1.5934) [2021-04-16 05:21:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][80/1251] eta 0:07:15 lr 0.000510 time 0.2813 (0.3722) loss 3.8518 (3.5220) grad_norm 1.5198 (1.5901) [2021-04-16 05:22:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][90/1251] eta 0:06:59 lr 0.000510 time 0.2782 (0.3615) loss 2.3628 (3.4964) grad_norm 1.5303 (1.5772) [2021-04-16 05:22:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][100/1251] eta 0:06:47 lr 0.000510 time 0.2690 (0.3539) loss 2.6630 (3.4751) grad_norm 1.4403 (1.5601) [2021-04-16 05:22:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][110/1251] eta 0:06:36 lr 0.000510 time 0.2976 (0.3471) loss 3.3908 (3.4876) grad_norm 1.7273 (1.5514) [2021-04-16 05:22:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][120/1251] eta 0:06:26 lr 0.000510 time 0.2969 (0.3416) loss 3.7481 (3.5037) grad_norm 1.3510 (1.5506) [2021-04-16 05:22:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][130/1251] eta 0:06:17 lr 0.000510 time 0.2666 (0.3366) loss 3.9218 (3.4990) grad_norm 1.4887 (1.5431) [2021-04-16 05:22:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][140/1251] eta 0:06:09 lr 0.000510 time 0.2846 (0.3326) loss 2.9620 (3.5022) grad_norm 1.5658 (1.5438) [2021-04-16 05:22:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][150/1251] eta 0:06:04 lr 0.000510 time 0.4088 (0.3310) loss 3.0792 (3.4849) grad_norm 1.4243 (1.5393) [2021-04-16 05:22:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][160/1251] eta 0:05:57 lr 0.000510 time 0.2829 (0.3279) loss 3.7453 (3.4793) grad_norm 1.4685 (1.5349) [2021-04-16 05:22:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][170/1251] eta 0:05:51 lr 0.000509 time 0.2875 (0.3248) loss 2.8503 (3.4596) grad_norm 1.3861 (1.5332) [2021-04-16 05:22:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][180/1251] eta 0:05:45 lr 0.000509 time 0.2860 (0.3222) loss 2.8563 (3.4587) grad_norm 1.4394 (1.5336) [2021-04-16 05:22:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][190/1251] eta 0:05:39 lr 0.000509 time 0.2710 (0.3197) loss 4.1212 (3.4675) grad_norm 1.4272 (1.5350) [2021-04-16 05:22:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][200/1251] eta 0:05:33 lr 0.000509 time 0.2972 (0.3176) loss 3.8732 (3.4652) grad_norm 1.5280 (1.5344) [2021-04-16 05:22:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][210/1251] eta 0:05:28 lr 0.000509 time 0.2786 (0.3158) loss 3.5158 (3.4807) grad_norm 1.7848 (1.5378) [2021-04-16 05:22:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][220/1251] eta 0:05:24 lr 0.000509 time 0.2846 (0.3148) loss 3.7431 (3.4893) grad_norm 1.3386 (1.5357) [2021-04-16 05:22:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][230/1251] eta 0:05:20 lr 0.000509 time 0.2695 (0.3137) loss 2.7727 (3.4893) grad_norm 1.5090 (1.5365) [2021-04-16 05:22:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][240/1251] eta 0:05:15 lr 0.000509 time 0.2894 (0.3121) loss 3.0600 (3.4829) grad_norm 1.4103 (1.5392) [2021-04-16 05:22:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][250/1251] eta 0:05:11 lr 0.000509 time 0.2866 (0.3108) loss 3.0552 (3.4885) grad_norm 1.4407 (1.5378) [2021-04-16 05:22:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][260/1251] eta 0:05:06 lr 0.000509 time 0.2853 (0.3095) loss 4.1783 (3.4952) grad_norm 1.8939 (1.5370) [2021-04-16 05:22:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][270/1251] eta 0:05:02 lr 0.000509 time 0.2808 (0.3083) loss 3.2799 (3.4943) grad_norm 1.6137 (1.5357) [2021-04-16 05:22:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][280/1251] eta 0:04:58 lr 0.000509 time 0.2559 (0.3074) loss 4.1473 (3.5051) grad_norm 1.5102 (1.5345) [2021-04-16 05:22:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][290/1251] eta 0:04:55 lr 0.000509 time 0.2792 (0.3070) loss 3.9472 (3.5162) grad_norm 2.0839 (1.5376) [2021-04-16 05:22:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][300/1251] eta 0:04:51 lr 0.000509 time 0.2862 (0.3061) loss 4.0504 (3.5172) grad_norm 1.3600 (1.5351) [2021-04-16 05:23:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][310/1251] eta 0:04:47 lr 0.000509 time 0.2612 (0.3050) loss 3.1810 (3.5115) grad_norm 1.3819 (1.5373) [2021-04-16 05:23:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][320/1251] eta 0:04:43 lr 0.000509 time 0.2907 (0.3042) loss 3.9863 (3.5134) grad_norm 1.6196 (1.5359) [2021-04-16 05:23:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][330/1251] eta 0:04:39 lr 0.000509 time 0.2951 (0.3034) loss 3.9633 (3.5041) grad_norm 1.4389 (1.5351) [2021-04-16 05:23:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][340/1251] eta 0:04:35 lr 0.000509 time 0.3026 (0.3027) loss 3.8888 (3.5018) grad_norm 1.5167 (1.5326) [2021-04-16 05:23:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][350/1251] eta 0:04:32 lr 0.000509 time 0.2412 (0.3027) loss 2.4428 (3.4996) grad_norm 1.8536 (1.5314) [2021-04-16 05:23:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][360/1251] eta 0:04:29 lr 0.000509 time 0.2561 (0.3021) loss 3.2394 (3.4964) grad_norm 1.5118 (1.5302) [2021-04-16 05:23:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][370/1251] eta 0:04:25 lr 0.000509 time 0.2884 (0.3014) loss 3.5175 (3.5033) grad_norm 1.5375 (1.5283) [2021-04-16 05:23:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][380/1251] eta 0:04:21 lr 0.000509 time 0.3272 (0.3007) loss 4.2023 (3.5053) grad_norm 1.5673 (1.5256) [2021-04-16 05:23:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][390/1251] eta 0:04:18 lr 0.000509 time 0.2861 (0.3001) loss 3.6622 (3.5119) grad_norm 1.7155 (1.5284) [2021-04-16 05:23:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][400/1251] eta 0:04:14 lr 0.000509 time 0.2738 (0.2994) loss 4.3975 (3.5229) grad_norm 1.6374 (1.5266) [2021-04-16 05:23:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][410/1251] eta 0:04:11 lr 0.000508 time 0.2708 (0.2989) loss 3.0386 (3.5237) grad_norm 1.3667 (1.5270) [2021-04-16 05:23:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][420/1251] eta 0:04:08 lr 0.000508 time 0.2893 (0.2986) loss 2.3184 (3.5165) grad_norm 1.9038 (1.5269) [2021-04-16 05:23:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][430/1251] eta 0:04:04 lr 0.000508 time 0.2569 (0.2981) loss 2.8331 (3.5174) grad_norm 1.4555 (1.5271) [2021-04-16 05:23:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][440/1251] eta 0:04:01 lr 0.000508 time 0.2935 (0.2978) loss 3.1603 (3.5084) grad_norm 1.3996 (1.5265) [2021-04-16 05:23:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][450/1251] eta 0:03:58 lr 0.000508 time 0.2652 (0.2972) loss 4.2243 (3.5020) grad_norm 1.3896 (1.5250) [2021-04-16 05:23:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][460/1251] eta 0:03:54 lr 0.000508 time 0.2903 (0.2968) loss 3.1639 (3.5015) grad_norm 1.6474 (1.5233) [2021-04-16 05:23:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][470/1251] eta 0:03:51 lr 0.000508 time 0.2573 (0.2968) loss 3.7970 (3.5085) grad_norm 1.8538 (1.5237) [2021-04-16 05:23:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][480/1251] eta 0:03:48 lr 0.000508 time 0.2835 (0.2967) loss 3.2269 (3.5050) grad_norm 1.7157 (1.5223) [2021-04-16 05:23:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][490/1251] eta 0:03:45 lr 0.000508 time 0.2709 (0.2966) loss 4.0659 (3.5054) grad_norm 1.5086 (1.5230) [2021-04-16 05:23:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][500/1251] eta 0:03:42 lr 0.000508 time 0.2910 (0.2962) loss 3.4185 (3.5092) grad_norm 1.3444 (1.5217) [2021-04-16 05:23:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][510/1251] eta 0:03:39 lr 0.000508 time 0.2667 (0.2959) loss 4.0722 (3.5124) grad_norm 1.5298 (1.5244) [2021-04-16 05:24:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][520/1251] eta 0:03:36 lr 0.000508 time 0.2687 (0.2957) loss 3.5090 (3.5080) grad_norm 1.4178 (1.5248) [2021-04-16 05:24:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][530/1251] eta 0:03:32 lr 0.000508 time 0.2734 (0.2953) loss 3.7883 (3.5087) grad_norm 1.4750 (1.5249) [2021-04-16 05:24:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][540/1251] eta 0:03:29 lr 0.000508 time 0.2888 (0.2949) loss 4.1190 (3.5115) grad_norm 1.4026 (1.5240) [2021-04-16 05:24:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][550/1251] eta 0:03:26 lr 0.000508 time 0.2944 (0.2946) loss 3.6502 (3.5146) grad_norm 1.5221 (1.5229) [2021-04-16 05:24:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][560/1251] eta 0:03:23 lr 0.000508 time 0.3048 (0.2943) loss 4.3438 (3.5118) grad_norm 1.4075 (1.5252) [2021-04-16 05:24:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][570/1251] eta 0:03:20 lr 0.000508 time 0.2933 (0.2942) loss 3.7314 (3.5158) grad_norm 1.4762 (1.5254) [2021-04-16 05:24:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][580/1251] eta 0:03:17 lr 0.000508 time 0.2704 (0.2942) loss 3.3809 (3.5194) grad_norm 1.4780 (1.5244) [2021-04-16 05:24:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][590/1251] eta 0:03:14 lr 0.000508 time 0.2761 (0.2939) loss 2.7309 (3.5210) grad_norm 1.5100 (1.5238) [2021-04-16 05:24:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][600/1251] eta 0:03:11 lr 0.000508 time 0.4095 (0.2937) loss 2.9160 (3.5163) grad_norm 1.8421 (1.5244) [2021-04-16 05:24:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][610/1251] eta 0:03:08 lr 0.000508 time 0.2936 (0.2935) loss 4.0287 (3.5177) grad_norm 1.5610 (1.5239) [2021-04-16 05:24:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][620/1251] eta 0:03:05 lr 0.000508 time 0.3177 (0.2933) loss 4.2128 (3.5234) grad_norm 1.7453 (1.5246) [2021-04-16 05:24:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][630/1251] eta 0:03:01 lr 0.000508 time 0.2909 (0.2931) loss 4.0730 (3.5245) grad_norm 1.5323 (1.5260) [2021-04-16 05:24:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][640/1251] eta 0:02:58 lr 0.000508 time 0.2788 (0.2927) loss 4.2939 (3.5297) grad_norm 1.6186 (1.5255) [2021-04-16 05:24:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][650/1251] eta 0:02:55 lr 0.000507 time 0.2778 (0.2925) loss 3.0720 (3.5324) grad_norm 1.4330 (1.5259) [2021-04-16 05:24:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][660/1251] eta 0:02:52 lr 0.000507 time 0.2749 (0.2922) loss 3.2932 (3.5330) grad_norm 1.5713 (1.5270) [2021-04-16 05:24:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][670/1251] eta 0:02:49 lr 0.000507 time 0.2730 (0.2920) loss 4.1358 (3.5335) grad_norm 1.2506 (1.5254) [2021-04-16 05:24:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][680/1251] eta 0:02:46 lr 0.000507 time 0.2798 (0.2919) loss 3.6297 (3.5324) grad_norm 1.4051 (1.5243) [2021-04-16 05:24:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][690/1251] eta 0:02:43 lr 0.000507 time 0.2743 (0.2917) loss 3.0287 (3.5301) grad_norm 1.3560 (1.5234) [2021-04-16 05:24:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][700/1251] eta 0:02:40 lr 0.000507 time 0.2800 (0.2915) loss 2.8626 (3.5224) grad_norm 1.5025 (1.5232) [2021-04-16 05:24:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][710/1251] eta 0:02:37 lr 0.000507 time 0.2589 (0.2913) loss 2.9599 (3.5210) grad_norm 1.2898 (1.5213) [2021-04-16 05:24:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][720/1251] eta 0:02:34 lr 0.000507 time 0.2864 (0.2911) loss 3.5756 (3.5218) grad_norm 1.6095 (1.5206) [2021-04-16 05:25:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][730/1251] eta 0:02:31 lr 0.000507 time 0.2735 (0.2912) loss 2.8697 (3.5213) grad_norm 1.4483 (1.5203) [2021-04-16 05:25:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][740/1251] eta 0:02:28 lr 0.000507 time 0.2851 (0.2910) loss 3.3018 (3.5199) grad_norm 1.5247 (1.5201) [2021-04-16 05:25:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][750/1251] eta 0:02:25 lr 0.000507 time 0.3118 (0.2908) loss 3.0218 (3.5233) grad_norm 1.5619 (1.5192) [2021-04-16 05:25:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][760/1251] eta 0:02:22 lr 0.000507 time 0.2698 (0.2907) loss 3.9401 (3.5210) grad_norm 1.4727 (1.5189) [2021-04-16 05:25:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][770/1251] eta 0:02:19 lr 0.000507 time 0.2842 (0.2906) loss 3.5367 (3.5230) grad_norm 1.4995 (1.5195) [2021-04-16 05:25:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][780/1251] eta 0:02:16 lr 0.000507 time 0.2819 (0.2904) loss 3.9073 (3.5263) grad_norm 1.7603 (1.5203) [2021-04-16 05:25:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][790/1251] eta 0:02:13 lr 0.000507 time 0.3128 (0.2904) loss 3.1186 (3.5265) grad_norm 1.6486 (1.5197) [2021-04-16 05:25:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][800/1251] eta 0:02:10 lr 0.000507 time 0.3670 (0.2904) loss 2.5183 (3.5272) grad_norm 1.6213 (1.5213) [2021-04-16 05:25:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][810/1251] eta 0:02:08 lr 0.000507 time 0.2958 (0.2903) loss 4.1605 (3.5309) grad_norm 1.6838 (1.5216) [2021-04-16 05:25:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][820/1251] eta 0:02:05 lr 0.000507 time 0.2824 (0.2902) loss 3.3851 (3.5300) grad_norm 1.4632 (1.5208) [2021-04-16 05:25:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][830/1251] eta 0:02:02 lr 0.000507 time 0.2572 (0.2900) loss 4.4409 (3.5325) grad_norm 1.5591 (1.5212) [2021-04-16 05:25:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][840/1251] eta 0:01:59 lr 0.000507 time 0.3101 (0.2898) loss 3.4810 (3.5323) grad_norm 1.4007 (1.5206) [2021-04-16 05:25:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][850/1251] eta 0:01:56 lr 0.000507 time 0.2469 (0.2896) loss 3.4562 (3.5318) grad_norm 1.7977 (1.5206) [2021-04-16 05:25:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][860/1251] eta 0:01:53 lr 0.000507 time 0.2597 (0.2894) loss 3.8186 (3.5299) grad_norm 1.3395 (1.5201) [2021-04-16 05:25:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][870/1251] eta 0:01:50 lr 0.000507 time 0.2783 (0.2895) loss 4.1363 (3.5298) grad_norm 1.5700 (1.5194) [2021-04-16 05:25:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][880/1251] eta 0:01:47 lr 0.000507 time 0.2754 (0.2893) loss 4.3710 (3.5297) grad_norm 1.5927 (1.5217) [2021-04-16 05:25:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][890/1251] eta 0:01:44 lr 0.000506 time 0.2767 (0.2892) loss 3.7656 (3.5305) grad_norm 1.4428 (1.5213) [2021-04-16 05:25:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][900/1251] eta 0:01:41 lr 0.000506 time 0.2721 (0.2890) loss 4.0024 (3.5331) grad_norm 1.3994 (1.5200) [2021-04-16 05:25:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][910/1251] eta 0:01:38 lr 0.000506 time 0.2652 (0.2889) loss 3.8526 (3.5351) grad_norm 1.6590 (1.5194) [2021-04-16 05:25:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][920/1251] eta 0:01:35 lr 0.000506 time 0.2812 (0.2887) loss 3.1023 (3.5359) grad_norm 1.4863 (1.5199) [2021-04-16 05:25:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][930/1251] eta 0:01:32 lr 0.000506 time 0.2750 (0.2887) loss 3.5620 (3.5350) grad_norm 1.3916 (1.5205) [2021-04-16 05:25:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][940/1251] eta 0:01:29 lr 0.000506 time 0.2704 (0.2886) loss 3.4987 (3.5341) grad_norm 1.3692 (1.5199) [2021-04-16 05:26:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][950/1251] eta 0:01:26 lr 0.000506 time 0.2886 (0.2885) loss 3.7207 (3.5332) grad_norm 1.6817 (1.5203) [2021-04-16 05:26:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][960/1251] eta 0:01:23 lr 0.000506 time 0.2791 (0.2884) loss 3.5432 (3.5335) grad_norm 1.4747 (1.5205) [2021-04-16 05:26:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][970/1251] eta 0:01:20 lr 0.000506 time 0.2681 (0.2882) loss 3.7713 (3.5362) grad_norm 1.3918 (1.5210) [2021-04-16 05:26:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][980/1251] eta 0:01:18 lr 0.000506 time 0.2912 (0.2881) loss 3.2789 (3.5358) grad_norm 1.4852 (1.5214) [2021-04-16 05:26:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][990/1251] eta 0:01:15 lr 0.000506 time 0.2754 (0.2880) loss 3.8900 (3.5317) grad_norm 1.3801 (1.5209) [2021-04-16 05:26:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][1000/1251] eta 0:01:12 lr 0.000506 time 0.2937 (0.2880) loss 4.5326 (3.5329) grad_norm 1.2639 (1.5199) [2021-04-16 05:26:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][1010/1251] eta 0:01:09 lr 0.000506 time 0.2796 (0.2879) loss 3.4913 (3.5356) grad_norm 1.4593 (1.5200) [2021-04-16 05:26:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][1020/1251] eta 0:01:06 lr 0.000506 time 0.3028 (0.2878) loss 3.7423 (3.5352) grad_norm 1.7786 (1.5203) [2021-04-16 05:26:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][1030/1251] eta 0:01:03 lr 0.000506 time 0.2520 (0.2876) loss 2.3634 (3.5350) grad_norm 1.2875 (1.5208) [2021-04-16 05:26:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][1040/1251] eta 0:01:00 lr 0.000506 time 0.2670 (0.2878) loss 3.8266 (3.5351) grad_norm 1.5571 (1.5217) [2021-04-16 05:26:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][1050/1251] eta 0:00:57 lr 0.000506 time 0.4287 (0.2878) loss 3.0210 (3.5350) grad_norm 1.7529 (1.5217) [2021-04-16 05:26:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][1060/1251] eta 0:00:54 lr 0.000506 time 0.2704 (0.2877) loss 2.5483 (3.5368) grad_norm 1.4783 (1.5220) [2021-04-16 05:26:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][1070/1251] eta 0:00:52 lr 0.000506 time 0.2856 (0.2878) loss 3.9575 (3.5361) grad_norm 1.7538 (1.5238) [2021-04-16 05:26:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][1080/1251] eta 0:00:49 lr 0.000506 time 0.2878 (0.2876) loss 3.3736 (3.5347) grad_norm 1.4163 (1.5261) [2021-04-16 05:26:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][1090/1251] eta 0:00:46 lr 0.000506 time 0.2532 (0.2875) loss 3.5631 (3.5341) grad_norm 2.1921 (1.5267) [2021-04-16 05:26:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][1100/1251] eta 0:00:43 lr 0.000506 time 0.2866 (0.2874) loss 3.5543 (3.5350) grad_norm 1.6684 (1.5268) [2021-04-16 05:26:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][1110/1251] eta 0:00:40 lr 0.000506 time 0.2688 (0.2874) loss 3.2629 (3.5372) grad_norm 1.4745 (1.5264) [2021-04-16 05:26:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][1120/1251] eta 0:00:37 lr 0.000506 time 0.2716 (0.2873) loss 3.6616 (3.5346) grad_norm 1.3258 (1.5258) [2021-04-16 05:26:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][1130/1251] eta 0:00:34 lr 0.000506 time 0.2816 (0.2872) loss 2.5063 (3.5359) grad_norm 1.6151 (1.5253) [2021-04-16 05:26:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][1140/1251] eta 0:00:31 lr 0.000505 time 0.2753 (0.2872) loss 4.1563 (3.5374) grad_norm 1.4797 (1.5252) [2021-04-16 05:26:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][1150/1251] eta 0:00:29 lr 0.000505 time 0.4059 (0.2872) loss 4.0177 (3.5400) grad_norm 1.4393 (1.5250) [2021-04-16 05:27:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][1160/1251] eta 0:00:26 lr 0.000505 time 0.2827 (0.2871) loss 3.1938 (3.5397) grad_norm 1.3947 (1.5251) [2021-04-16 05:27:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][1170/1251] eta 0:00:23 lr 0.000505 time 0.2717 (0.2869) loss 2.6045 (3.5395) grad_norm 1.3445 (1.5253) [2021-04-16 05:27:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][1180/1251] eta 0:00:20 lr 0.000505 time 0.2826 (0.2869) loss 3.8460 (3.5398) grad_norm 1.4814 (1.5252) [2021-04-16 05:27:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][1190/1251] eta 0:00:17 lr 0.000505 time 0.2892 (0.2869) loss 3.5791 (3.5369) grad_norm 1.3788 (1.5252) [2021-04-16 05:27:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][1200/1251] eta 0:00:14 lr 0.000505 time 0.2727 (0.2868) loss 3.4100 (3.5374) grad_norm 1.3941 (1.5246) [2021-04-16 05:27:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][1210/1251] eta 0:00:11 lr 0.000505 time 0.2890 (0.2867) loss 3.7733 (3.5382) grad_norm 1.3454 (1.5246) [2021-04-16 05:27:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][1220/1251] eta 0:00:08 lr 0.000505 time 0.2691 (0.2865) loss 3.0011 (3.5381) grad_norm 1.3881 (1.5245) [2021-04-16 05:27:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][1230/1251] eta 0:00:06 lr 0.000505 time 0.2720 (0.2865) loss 3.8849 (3.5395) grad_norm 1.5440 (1.5247) [2021-04-16 05:27:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][1240/1251] eta 0:00:03 lr 0.000505 time 0.2518 (0.2863) loss 2.8423 (3.5365) grad_norm 1.6929 (1.5252) [2021-04-16 05:27:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [149/300][1250/1251] eta 0:00:00 lr 0.000505 time 0.2498 (0.2860) loss 4.0609 (3.5361) grad_norm 1.4179 (1.5255) [2021-04-16 05:27:28 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 149 training takes 0:06:00 [2021-04-16 05:27:28 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_149.pth saving...... [2021-04-16 05:27:36 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_149.pth saved !!! [2021-04-16 05:27:37 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.207 (1.207) Loss 1.0921 (1.0921) Acc@1 73.828 (73.828) Acc@5 91.504 (91.504) [2021-04-16 05:27:38 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.138 (0.233) Loss 1.0429 (1.0378) Acc@1 74.707 (75.559) Acc@5 94.043 (93.066) [2021-04-16 05:27:40 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.250 (0.213) Loss 0.9530 (1.0223) Acc@1 78.223 (76.177) Acc@5 94.531 (93.290) [2021-04-16 05:27:43 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.123 (0.228) Loss 1.0483 (1.0299) Acc@1 74.414 (75.923) Acc@5 94.043 (93.218) [2021-04-16 05:27:44 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.218) Loss 1.0415 (1.0265) Acc@1 75.488 (76.065) Acc@5 92.090 (93.269) [2021-04-16 05:27:50 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 76.048 Acc@5 93.304 [2021-04-16 05:27:50 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 76.0% [2021-04-16 05:27:50 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 76.05% [2021-04-16 05:27:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][0/1251] eta 1:28:59 lr 0.000505 time 4.2681 (4.2681) loss 3.9050 (3.9050) grad_norm 1.5536 (1.5536) [2021-04-16 05:27:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][10/1251] eta 0:13:36 lr 0.000505 time 0.4563 (0.6581) loss 3.2451 (3.5231) grad_norm 1.6060 (1.5125) [2021-04-16 05:28:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][20/1251] eta 0:09:49 lr 0.000505 time 0.2934 (0.4788) loss 3.8583 (3.5718) grad_norm 1.7598 (1.5040) [2021-04-16 05:28:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][30/1251] eta 0:08:28 lr 0.000505 time 0.2979 (0.4164) loss 3.2558 (3.5178) grad_norm 1.5812 (1.5092) [2021-04-16 05:28:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][40/1251] eta 0:07:43 lr 0.000505 time 0.2912 (0.3825) loss 3.7683 (3.5304) grad_norm 1.3503 (1.5114) [2021-04-16 05:28:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][50/1251] eta 0:07:13 lr 0.000505 time 0.2693 (0.3611) loss 3.9995 (3.5215) grad_norm 1.4326 (1.4900) [2021-04-16 05:28:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][60/1251] eta 0:06:58 lr 0.000505 time 0.4470 (0.3518) loss 4.5309 (3.5613) grad_norm 1.5486 (1.4798) [2021-04-16 05:28:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][70/1251] eta 0:06:43 lr 0.000505 time 0.2725 (0.3415) loss 3.2609 (3.5328) grad_norm 1.6570 (1.4754) [2021-04-16 05:28:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][80/1251] eta 0:06:30 lr 0.000505 time 0.2719 (0.3336) loss 3.8341 (3.5663) grad_norm 1.5729 (1.4842) [2021-04-16 05:28:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][90/1251] eta 0:06:20 lr 0.000505 time 0.2594 (0.3279) loss 3.5408 (3.5953) grad_norm 1.3827 (1.4869) [2021-04-16 05:28:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][100/1251] eta 0:06:12 lr 0.000505 time 0.2976 (0.3235) loss 3.7519 (3.5771) grad_norm 1.4205 (1.4879) [2021-04-16 05:28:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][110/1251] eta 0:06:04 lr 0.000505 time 0.2860 (0.3193) loss 3.7621 (3.5734) grad_norm 1.4324 (1.4869) [2021-04-16 05:28:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][120/1251] eta 0:05:56 lr 0.000505 time 0.2706 (0.3155) loss 2.8170 (3.5775) grad_norm 1.5560 (inf) [2021-04-16 05:28:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][130/1251] eta 0:05:50 lr 0.000504 time 0.2788 (0.3131) loss 3.3374 (3.5818) grad_norm 1.6942 (inf) [2021-04-16 05:28:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][140/1251] eta 0:05:46 lr 0.000504 time 0.2767 (0.3115) loss 2.5998 (3.5852) grad_norm 1.5117 (inf) [2021-04-16 05:28:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][150/1251] eta 0:05:41 lr 0.000504 time 0.2742 (0.3097) loss 4.2041 (3.5888) grad_norm 1.7223 (inf) [2021-04-16 05:28:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][160/1251] eta 0:05:35 lr 0.000504 time 0.2844 (0.3076) loss 3.4306 (3.6070) grad_norm 1.3180 (inf) [2021-04-16 05:28:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][170/1251] eta 0:05:30 lr 0.000504 time 0.2749 (0.3057) loss 3.8896 (3.6036) grad_norm 1.7042 (inf) [2021-04-16 05:28:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][180/1251] eta 0:05:26 lr 0.000504 time 0.2794 (0.3045) loss 3.5574 (3.6159) grad_norm 1.3687 (inf) [2021-04-16 05:28:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][190/1251] eta 0:05:21 lr 0.000504 time 0.2831 (0.3032) loss 4.1049 (3.6320) grad_norm 1.2966 (inf) [2021-04-16 05:28:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][200/1251] eta 0:05:17 lr 0.000504 time 0.2861 (0.3021) loss 4.1729 (3.6281) grad_norm 1.6052 (inf) [2021-04-16 05:28:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][210/1251] eta 0:05:13 lr 0.000504 time 0.2807 (0.3013) loss 3.8850 (3.6124) grad_norm 1.6653 (inf) [2021-04-16 05:28:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][220/1251] eta 0:05:09 lr 0.000504 time 0.2817 (0.3004) loss 3.1054 (3.6070) grad_norm 1.3221 (inf) [2021-04-16 05:28:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][230/1251] eta 0:05:05 lr 0.000504 time 0.2825 (0.2995) loss 3.0039 (3.6026) grad_norm 1.5652 (inf) [2021-04-16 05:29:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][240/1251] eta 0:05:01 lr 0.000504 time 0.2633 (0.2986) loss 3.4877 (3.6049) grad_norm 1.4631 (inf) [2021-04-16 05:29:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][250/1251] eta 0:04:58 lr 0.000504 time 0.2709 (0.2986) loss 3.1573 (3.6030) grad_norm 1.9769 (inf) [2021-04-16 05:29:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][260/1251] eta 0:04:55 lr 0.000504 time 0.4097 (0.2981) loss 2.7126 (3.5959) grad_norm 1.5224 (inf) [2021-04-16 05:29:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][270/1251] eta 0:04:51 lr 0.000504 time 0.2775 (0.2972) loss 3.6658 (3.5869) grad_norm 1.8323 (inf) [2021-04-16 05:29:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][280/1251] eta 0:04:47 lr 0.000504 time 0.2814 (0.2965) loss 4.0177 (3.5883) grad_norm 1.6096 (inf) [2021-04-16 05:29:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][290/1251] eta 0:04:44 lr 0.000504 time 0.2785 (0.2964) loss 3.8294 (3.5904) grad_norm 1.3164 (inf) [2021-04-16 05:29:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][300/1251] eta 0:04:41 lr 0.000504 time 0.2714 (0.2957) loss 3.4560 (3.5805) grad_norm 1.6020 (inf) [2021-04-16 05:29:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][310/1251] eta 0:04:37 lr 0.000504 time 0.2649 (0.2954) loss 3.6934 (3.5895) grad_norm 1.7757 (inf) [2021-04-16 05:29:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][320/1251] eta 0:04:34 lr 0.000504 time 0.2657 (0.2950) loss 4.2597 (3.5956) grad_norm 1.3590 (inf) [2021-04-16 05:29:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][330/1251] eta 0:04:31 lr 0.000504 time 0.2806 (0.2946) loss 3.6394 (3.5968) grad_norm 1.5084 (inf) [2021-04-16 05:29:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][340/1251] eta 0:04:28 lr 0.000504 time 0.2696 (0.2946) loss 3.1594 (3.5853) grad_norm 1.3468 (inf) [2021-04-16 05:29:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][350/1251] eta 0:04:25 lr 0.000504 time 0.2907 (0.2942) loss 3.6485 (3.5811) grad_norm 1.2488 (inf) [2021-04-16 05:29:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][360/1251] eta 0:04:22 lr 0.000504 time 0.2945 (0.2942) loss 3.8521 (3.5770) grad_norm 1.4318 (inf) [2021-04-16 05:29:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][370/1251] eta 0:04:19 lr 0.000503 time 0.2967 (0.2940) loss 4.1434 (3.5776) grad_norm 1.4237 (inf) [2021-04-16 05:29:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][380/1251] eta 0:04:16 lr 0.000503 time 0.2913 (0.2940) loss 3.8530 (3.5802) grad_norm 1.6118 (inf) [2021-04-16 05:29:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][390/1251] eta 0:04:12 lr 0.000503 time 0.2646 (0.2935) loss 2.3653 (3.5852) grad_norm 1.2571 (inf) [2021-04-16 05:29:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][400/1251] eta 0:04:09 lr 0.000503 time 0.2759 (0.2932) loss 3.3821 (3.5802) grad_norm 1.3335 (inf) [2021-04-16 05:29:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][410/1251] eta 0:04:06 lr 0.000503 time 0.2617 (0.2928) loss 2.7687 (3.5791) grad_norm 1.5355 (inf) [2021-04-16 05:29:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][420/1251] eta 0:04:03 lr 0.000503 time 0.2892 (0.2924) loss 3.0436 (3.5751) grad_norm 1.7507 (inf) [2021-04-16 05:29:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][430/1251] eta 0:03:59 lr 0.000503 time 0.2725 (0.2921) loss 3.9089 (3.5751) grad_norm 1.7217 (inf) [2021-04-16 05:29:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][440/1251] eta 0:03:56 lr 0.000503 time 0.2739 (0.2918) loss 4.1206 (3.5797) grad_norm 1.4711 (inf) [2021-04-16 05:30:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][450/1251] eta 0:03:53 lr 0.000503 time 0.2679 (0.2915) loss 2.3645 (3.5822) grad_norm 1.4208 (inf) [2021-04-16 05:30:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][460/1251] eta 0:03:50 lr 0.000503 time 0.2966 (0.2915) loss 3.4142 (3.5821) grad_norm 1.5290 (inf) [2021-04-16 05:30:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][470/1251] eta 0:03:47 lr 0.000503 time 0.3072 (0.2913) loss 3.7566 (3.5845) grad_norm 1.4708 (inf) [2021-04-16 05:30:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][480/1251] eta 0:03:44 lr 0.000503 time 0.2818 (0.2913) loss 3.0037 (3.5874) grad_norm 2.0244 (inf) [2021-04-16 05:30:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][490/1251] eta 0:03:41 lr 0.000503 time 0.2746 (0.2912) loss 4.1010 (3.5921) grad_norm 1.3558 (nan) [2021-04-16 05:30:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][500/1251] eta 0:03:38 lr 0.000503 time 0.2702 (0.2908) loss 3.2768 (3.5862) grad_norm 1.8740 (nan) [2021-04-16 05:30:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][510/1251] eta 0:03:35 lr 0.000503 time 0.3027 (0.2907) loss 3.2930 (3.5863) grad_norm 1.5690 (nan) [2021-04-16 05:30:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][520/1251] eta 0:03:32 lr 0.000503 time 0.2823 (0.2905) loss 2.9938 (3.5921) grad_norm 1.3153 (nan) [2021-04-16 05:30:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][530/1251] eta 0:03:29 lr 0.000503 time 0.2592 (0.2903) loss 4.0729 (3.5881) grad_norm 1.6713 (nan) [2021-04-16 05:30:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][540/1251] eta 0:03:26 lr 0.000503 time 0.2764 (0.2901) loss 4.0929 (3.5910) grad_norm 1.3260 (nan) [2021-04-16 05:30:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][550/1251] eta 0:03:23 lr 0.000503 time 0.2697 (0.2902) loss 3.6341 (3.5926) grad_norm 1.7344 (nan) [2021-04-16 05:30:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][560/1251] eta 0:03:20 lr 0.000503 time 0.2724 (0.2900) loss 2.6680 (3.5890) grad_norm 1.3493 (nan) [2021-04-16 05:30:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][570/1251] eta 0:03:17 lr 0.000503 time 0.2831 (0.2901) loss 3.8978 (3.5884) grad_norm 1.7689 (nan) [2021-04-16 05:30:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][580/1251] eta 0:03:14 lr 0.000503 time 0.2736 (0.2900) loss 4.1608 (3.5885) grad_norm 1.5081 (nan) [2021-04-16 05:30:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][590/1251] eta 0:03:11 lr 0.000503 time 0.2605 (0.2901) loss 3.5404 (3.5859) grad_norm 1.5400 (nan) [2021-04-16 05:30:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][600/1251] eta 0:03:08 lr 0.000503 time 0.2598 (0.2899) loss 3.6488 (3.5870) grad_norm 1.6349 (nan) [2021-04-16 05:30:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][610/1251] eta 0:03:05 lr 0.000502 time 0.2609 (0.2897) loss 3.4116 (3.5899) grad_norm 1.4670 (nan) [2021-04-16 05:30:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][620/1251] eta 0:03:02 lr 0.000502 time 0.2700 (0.2896) loss 4.0833 (3.5883) grad_norm 1.3861 (nan) [2021-04-16 05:30:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][630/1251] eta 0:02:59 lr 0.000502 time 0.2866 (0.2895) loss 3.8355 (3.5839) grad_norm 1.5514 (nan) [2021-04-16 05:30:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][640/1251] eta 0:02:56 lr 0.000502 time 0.2794 (0.2896) loss 4.0823 (3.5841) grad_norm 1.5019 (nan) [2021-04-16 05:30:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][650/1251] eta 0:02:53 lr 0.000502 time 0.2653 (0.2894) loss 3.1540 (3.5830) grad_norm 1.5151 (nan) [2021-04-16 05:31:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][660/1251] eta 0:02:50 lr 0.000502 time 0.2854 (0.2892) loss 3.5065 (3.5820) grad_norm 1.5241 (nan) [2021-04-16 05:31:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][670/1251] eta 0:02:47 lr 0.000502 time 0.2979 (0.2891) loss 4.0399 (3.5818) grad_norm 1.9645 (nan) [2021-04-16 05:31:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][680/1251] eta 0:02:44 lr 0.000502 time 0.3018 (0.2889) loss 3.3922 (3.5823) grad_norm 1.8962 (nan) [2021-04-16 05:31:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][690/1251] eta 0:02:42 lr 0.000502 time 0.2611 (0.2888) loss 4.4040 (3.5853) grad_norm 1.4470 (nan) [2021-04-16 05:31:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][700/1251] eta 0:02:39 lr 0.000502 time 0.2594 (0.2886) loss 4.0128 (3.5824) grad_norm 1.5079 (nan) [2021-04-16 05:31:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][710/1251] eta 0:02:36 lr 0.000502 time 0.4401 (0.2889) loss 3.6955 (3.5812) grad_norm 1.5828 (nan) [2021-04-16 05:31:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][720/1251] eta 0:02:33 lr 0.000502 time 0.2801 (0.2887) loss 4.6067 (3.5852) grad_norm 1.3008 (nan) [2021-04-16 05:31:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][730/1251] eta 0:02:30 lr 0.000502 time 0.2683 (0.2886) loss 2.8362 (3.5839) grad_norm 1.7611 (nan) [2021-04-16 05:31:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][740/1251] eta 0:02:27 lr 0.000502 time 0.2622 (0.2884) loss 3.4825 (3.5803) grad_norm 1.5375 (nan) [2021-04-16 05:31:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][750/1251] eta 0:02:24 lr 0.000502 time 0.2675 (0.2883) loss 4.1536 (3.5830) grad_norm 1.4600 (nan) [2021-04-16 05:31:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][760/1251] eta 0:02:21 lr 0.000502 time 0.2772 (0.2882) loss 2.8516 (3.5851) grad_norm 1.4135 (nan) [2021-04-16 05:31:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][770/1251] eta 0:02:18 lr 0.000502 time 0.2795 (0.2880) loss 4.3865 (3.5863) grad_norm 1.3344 (nan) [2021-04-16 05:31:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][780/1251] eta 0:02:15 lr 0.000502 time 0.2749 (0.2880) loss 3.2183 (3.5867) grad_norm 1.7592 (nan) [2021-04-16 05:31:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][790/1251] eta 0:02:12 lr 0.000502 time 0.2945 (0.2880) loss 4.0111 (3.5861) grad_norm 1.9294 (nan) [2021-04-16 05:31:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][800/1251] eta 0:02:09 lr 0.000502 time 0.3129 (0.2880) loss 4.3657 (3.5890) grad_norm 1.4413 (nan) [2021-04-16 05:31:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][810/1251] eta 0:02:06 lr 0.000502 time 0.2712 (0.2878) loss 4.0401 (3.5913) grad_norm 1.4760 (nan) [2021-04-16 05:31:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][820/1251] eta 0:02:03 lr 0.000502 time 0.2703 (0.2877) loss 3.0367 (3.5882) grad_norm 1.5039 (nan) [2021-04-16 05:31:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][830/1251] eta 0:02:01 lr 0.000502 time 0.3135 (0.2878) loss 3.6114 (3.5859) grad_norm 1.5577 (nan) [2021-04-16 05:31:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][840/1251] eta 0:01:58 lr 0.000502 time 0.2627 (0.2879) loss 2.6438 (3.5838) grad_norm 1.3827 (nan) [2021-04-16 05:31:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][850/1251] eta 0:01:55 lr 0.000501 time 0.4537 (0.2880) loss 4.0632 (3.5877) grad_norm 1.2917 (nan) [2021-04-16 05:31:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][860/1251] eta 0:01:52 lr 0.000501 time 0.2796 (0.2878) loss 3.2799 (3.5869) grad_norm 1.7797 (nan) [2021-04-16 05:32:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][870/1251] eta 0:01:49 lr 0.000501 time 0.2792 (0.2879) loss 3.9622 (3.5871) grad_norm 1.4993 (nan) [2021-04-16 05:32:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][880/1251] eta 0:01:46 lr 0.000501 time 0.2711 (0.2878) loss 3.7466 (3.5865) grad_norm 1.2788 (nan) [2021-04-16 05:32:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][890/1251] eta 0:01:43 lr 0.000501 time 0.2803 (0.2877) loss 3.9349 (3.5862) grad_norm 1.5021 (nan) [2021-04-16 05:32:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][900/1251] eta 0:01:40 lr 0.000501 time 0.2826 (0.2877) loss 3.3247 (3.5830) grad_norm 1.8988 (nan) [2021-04-16 05:32:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][910/1251] eta 0:01:38 lr 0.000501 time 0.2651 (0.2875) loss 3.4660 (3.5839) grad_norm 1.4112 (nan) [2021-04-16 05:32:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][920/1251] eta 0:01:35 lr 0.000501 time 0.2872 (0.2876) loss 3.7156 (3.5831) grad_norm 1.6015 (nan) [2021-04-16 05:32:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][930/1251] eta 0:01:32 lr 0.000501 time 0.3049 (0.2875) loss 3.7706 (3.5801) grad_norm 1.5541 (nan) [2021-04-16 05:32:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][940/1251] eta 0:01:29 lr 0.000501 time 0.2962 (0.2876) loss 4.1686 (3.5791) grad_norm 1.8353 (nan) [2021-04-16 05:32:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][950/1251] eta 0:01:26 lr 0.000501 time 0.3007 (0.2875) loss 4.0473 (3.5791) grad_norm 1.6640 (nan) [2021-04-16 05:32:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][960/1251] eta 0:01:23 lr 0.000501 time 0.2767 (0.2875) loss 3.8067 (3.5796) grad_norm 1.4658 (nan) [2021-04-16 05:32:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][970/1251] eta 0:01:20 lr 0.000501 time 0.3045 (0.2874) loss 3.6055 (3.5762) grad_norm 1.5735 (nan) [2021-04-16 05:32:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][980/1251] eta 0:01:17 lr 0.000501 time 0.2794 (0.2873) loss 3.9931 (3.5785) grad_norm 1.5752 (nan) [2021-04-16 05:32:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][990/1251] eta 0:01:14 lr 0.000501 time 0.3089 (0.2872) loss 4.3343 (3.5794) grad_norm 1.5495 (nan) [2021-04-16 05:32:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][1000/1251] eta 0:01:12 lr 0.000501 time 0.3033 (0.2871) loss 2.8650 (3.5794) grad_norm 1.5806 (nan) [2021-04-16 05:32:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][1010/1251] eta 0:01:09 lr 0.000501 time 0.2636 (0.2871) loss 3.5412 (3.5779) grad_norm 1.8252 (nan) [2021-04-16 05:32:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][1020/1251] eta 0:01:06 lr 0.000501 time 0.2771 (0.2870) loss 3.3944 (3.5777) grad_norm 1.8223 (nan) [2021-04-16 05:32:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][1030/1251] eta 0:01:03 lr 0.000501 time 0.2942 (0.2869) loss 2.3475 (3.5752) grad_norm 1.3748 (nan) [2021-04-16 05:32:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][1040/1251] eta 0:01:00 lr 0.000501 time 0.2784 (0.2869) loss 4.1570 (3.5746) grad_norm 1.5161 (nan) [2021-04-16 05:32:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][1050/1251] eta 0:00:57 lr 0.000501 time 0.4395 (0.2870) loss 4.7532 (3.5758) grad_norm 1.4762 (nan) [2021-04-16 05:32:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][1060/1251] eta 0:00:54 lr 0.000501 time 0.2467 (0.2869) loss 3.5677 (3.5758) grad_norm 1.5196 (nan) [2021-04-16 05:32:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][1070/1251] eta 0:00:51 lr 0.000501 time 0.3012 (0.2868) loss 3.7317 (3.5762) grad_norm 1.6996 (nan) [2021-04-16 05:33:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][1080/1251] eta 0:00:49 lr 0.000501 time 0.2653 (0.2867) loss 2.8955 (3.5761) grad_norm 1.8116 (nan) [2021-04-16 05:33:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][1090/1251] eta 0:00:46 lr 0.000500 time 0.2968 (0.2867) loss 3.0754 (3.5736) grad_norm 1.4724 (nan) [2021-04-16 05:33:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][1100/1251] eta 0:00:43 lr 0.000500 time 0.2738 (0.2866) loss 3.9362 (3.5759) grad_norm 1.4637 (nan) [2021-04-16 05:33:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][1110/1251] eta 0:00:40 lr 0.000500 time 0.2871 (0.2865) loss 3.2891 (3.5744) grad_norm 1.5791 (nan) [2021-04-16 05:33:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][1120/1251] eta 0:00:37 lr 0.000500 time 0.2976 (0.2864) loss 4.4063 (3.5727) grad_norm 1.4533 (nan) [2021-04-16 05:33:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][1130/1251] eta 0:00:34 lr 0.000500 time 0.2666 (0.2863) loss 4.2418 (3.5718) grad_norm 1.7667 (nan) [2021-04-16 05:33:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][1140/1251] eta 0:00:31 lr 0.000500 time 0.2697 (0.2864) loss 3.5198 (3.5726) grad_norm 1.5236 (nan) [2021-04-16 05:33:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][1150/1251] eta 0:00:28 lr 0.000500 time 0.2958 (0.2864) loss 4.3109 (3.5727) grad_norm 1.5382 (nan) [2021-04-16 05:33:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][1160/1251] eta 0:00:26 lr 0.000500 time 0.2932 (0.2864) loss 4.1928 (3.5733) grad_norm 1.8735 (nan) [2021-04-16 05:33:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][1170/1251] eta 0:00:23 lr 0.000500 time 0.2559 (0.2865) loss 3.0411 (3.5730) grad_norm 1.3871 (nan) [2021-04-16 05:33:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][1180/1251] eta 0:00:20 lr 0.000500 time 0.2947 (0.2864) loss 3.4376 (3.5722) grad_norm 1.5241 (nan) [2021-04-16 05:33:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][1190/1251] eta 0:00:17 lr 0.000500 time 0.2746 (0.2864) loss 2.2297 (3.5706) grad_norm 1.3858 (nan) [2021-04-16 05:33:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][1200/1251] eta 0:00:14 lr 0.000500 time 0.2939 (0.2864) loss 3.2449 (3.5697) grad_norm 1.4510 (nan) [2021-04-16 05:33:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][1210/1251] eta 0:00:11 lr 0.000500 time 0.2873 (0.2863) loss 3.6692 (3.5674) grad_norm 1.4561 (nan) [2021-04-16 05:33:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][1220/1251] eta 0:00:08 lr 0.000500 time 0.3008 (0.2865) loss 3.3689 (3.5666) grad_norm 2.0060 (nan) [2021-04-16 05:33:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][1230/1251] eta 0:00:06 lr 0.000500 time 0.2792 (0.2865) loss 2.8916 (3.5671) grad_norm 1.5354 (nan) [2021-04-16 05:33:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][1240/1251] eta 0:00:03 lr 0.000500 time 0.2484 (0.2864) loss 3.4232 (3.5665) grad_norm 1.5070 (nan) [2021-04-16 05:33:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [150/300][1250/1251] eta 0:00:00 lr 0.000500 time 0.2476 (0.2861) loss 2.3517 (3.5665) grad_norm 1.6671 (nan) [2021-04-16 05:33:51 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 150 training takes 0:06:01 [2021-04-16 05:33:51 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_150.pth saving...... [2021-04-16 05:34:02 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_150.pth saved !!! [2021-04-16 05:34:03 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.206 (1.206) Loss 1.0556 (1.0556) Acc@1 74.902 (74.902) Acc@5 93.066 (93.066) [2021-04-16 05:34:04 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.126 (0.216) Loss 1.0767 (1.0070) Acc@1 73.535 (76.314) Acc@5 92.383 (93.217) [2021-04-16 05:34:07 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.403 (0.241) Loss 1.0756 (1.0136) Acc@1 75.098 (76.065) Acc@5 91.895 (93.127) [2021-04-16 05:34:09 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.105 (0.229) Loss 1.0149 (1.0182) Acc@1 74.121 (76.014) Acc@5 94.238 (93.205) [2021-04-16 05:34:11 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.217) Loss 0.9557 (1.0116) Acc@1 76.758 (76.167) Acc@5 93.750 (93.305) [2021-04-16 05:34:19 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 76.108 Acc@5 93.270 [2021-04-16 05:34:19 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 76.1% [2021-04-16 05:34:19 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 76.11% [2021-04-16 05:34:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][0/1251] eta 0:34:38 lr 0.000500 time 1.6619 (1.6619) loss 3.4167 (3.4167) grad_norm 1.4903 (1.4903) [2021-04-16 05:34:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][10/1251] eta 0:08:24 lr 0.000500 time 0.2975 (0.4068) loss 3.3875 (3.2228) grad_norm 1.3903 (1.5279) [2021-04-16 05:34:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][20/1251] eta 0:07:06 lr 0.000500 time 0.2607 (0.3462) loss 2.4373 (3.3508) grad_norm 1.5970 (1.5643) [2021-04-16 05:34:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][30/1251] eta 0:06:39 lr 0.000500 time 0.3046 (0.3270) loss 3.5525 (3.4027) grad_norm 1.6456 (1.5922) [2021-04-16 05:34:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][40/1251] eta 0:06:21 lr 0.000500 time 0.2982 (0.3146) loss 3.7368 (3.4688) grad_norm 1.8234 (1.6149) [2021-04-16 05:34:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][50/1251] eta 0:06:10 lr 0.000500 time 0.2637 (0.3088) loss 3.1712 (3.4316) grad_norm 1.9060 (1.6007) [2021-04-16 05:34:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][60/1251] eta 0:06:03 lr 0.000500 time 0.4352 (0.3054) loss 3.0995 (3.4233) grad_norm 1.6499 (1.5893) [2021-04-16 05:34:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][70/1251] eta 0:05:56 lr 0.000500 time 0.3002 (0.3016) loss 3.1778 (3.4149) grad_norm 1.3131 (1.5659) [2021-04-16 05:34:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][80/1251] eta 0:05:49 lr 0.000499 time 0.2718 (0.2986) loss 4.3239 (3.4240) grad_norm 1.5835 (1.5683) [2021-04-16 05:34:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][90/1251] eta 0:05:43 lr 0.000499 time 0.2771 (0.2959) loss 4.0855 (3.4393) grad_norm 1.6216 (1.5750) [2021-04-16 05:34:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][100/1251] eta 0:05:38 lr 0.000499 time 0.2650 (0.2943) loss 4.0231 (3.4396) grad_norm 1.6150 (1.5675) [2021-04-16 05:34:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][110/1251] eta 0:05:33 lr 0.000499 time 0.2689 (0.2925) loss 2.8278 (3.4495) grad_norm 1.4595 (1.5623) [2021-04-16 05:34:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][120/1251] eta 0:05:30 lr 0.000499 time 0.2801 (0.2922) loss 3.7060 (3.4867) grad_norm 1.6781 (1.5636) [2021-04-16 05:34:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][130/1251] eta 0:05:26 lr 0.000499 time 0.2530 (0.2910) loss 3.7971 (3.4966) grad_norm 1.5931 (1.5573) [2021-04-16 05:35:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][140/1251] eta 0:05:23 lr 0.000499 time 0.2835 (0.2912) loss 3.9679 (3.5050) grad_norm 1.4393 (1.5538) [2021-04-16 05:35:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][150/1251] eta 0:05:19 lr 0.000499 time 0.2755 (0.2899) loss 3.8170 (3.5287) grad_norm 1.7825 (1.5491) [2021-04-16 05:35:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][160/1251] eta 0:05:15 lr 0.000499 time 0.2784 (0.2891) loss 3.4919 (3.5421) grad_norm 1.4203 (1.5492) [2021-04-16 05:35:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][170/1251] eta 0:05:12 lr 0.000499 time 0.2844 (0.2889) loss 4.2999 (3.5610) grad_norm 1.3230 (1.5414) [2021-04-16 05:35:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][180/1251] eta 0:05:08 lr 0.000499 time 0.2824 (0.2884) loss 3.5977 (3.5641) grad_norm 1.7115 (1.5412) [2021-04-16 05:35:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][190/1251] eta 0:05:05 lr 0.000499 time 0.2749 (0.2875) loss 4.0528 (3.5717) grad_norm 1.7221 (1.5408) [2021-04-16 05:35:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][200/1251] eta 0:05:01 lr 0.000499 time 0.2815 (0.2872) loss 3.7460 (3.5744) grad_norm 1.5310 (1.5402) [2021-04-16 05:35:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][210/1251] eta 0:04:58 lr 0.000499 time 0.2842 (0.2870) loss 3.6891 (3.5733) grad_norm 1.4563 (1.5432) [2021-04-16 05:35:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][220/1251] eta 0:04:55 lr 0.000499 time 0.2775 (0.2866) loss 3.7930 (3.5676) grad_norm 1.4823 (1.5448) [2021-04-16 05:35:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][230/1251] eta 0:04:52 lr 0.000499 time 0.2902 (0.2863) loss 3.9678 (3.5652) grad_norm 1.4299 (1.5458) [2021-04-16 05:35:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][240/1251] eta 0:04:49 lr 0.000499 time 0.2817 (0.2861) loss 3.2125 (3.5601) grad_norm 1.7506 (1.5476) [2021-04-16 05:35:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][250/1251] eta 0:04:46 lr 0.000499 time 0.2659 (0.2858) loss 4.0788 (3.5663) grad_norm 1.7450 (1.5497) [2021-04-16 05:35:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][260/1251] eta 0:04:43 lr 0.000499 time 0.3879 (0.2858) loss 3.7347 (3.5654) grad_norm 1.5001 (1.5469) [2021-04-16 05:35:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][270/1251] eta 0:04:39 lr 0.000499 time 0.2726 (0.2854) loss 2.2574 (3.5620) grad_norm 1.4032 (1.5487) [2021-04-16 05:35:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][280/1251] eta 0:04:37 lr 0.000499 time 0.2789 (0.2853) loss 3.1969 (3.5534) grad_norm 1.5927 (1.5491) [2021-04-16 05:35:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][290/1251] eta 0:04:33 lr 0.000499 time 0.2670 (0.2851) loss 3.5567 (3.5576) grad_norm 1.6213 (1.5503) [2021-04-16 05:35:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][300/1251] eta 0:04:30 lr 0.000499 time 0.2733 (0.2849) loss 3.8905 (3.5594) grad_norm 1.4875 (1.5519) [2021-04-16 05:35:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][310/1251] eta 0:04:27 lr 0.000499 time 0.2977 (0.2848) loss 3.9494 (3.5667) grad_norm 1.6646 (1.5505) [2021-04-16 05:35:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][320/1251] eta 0:04:24 lr 0.000498 time 0.3100 (0.2846) loss 4.0213 (3.5700) grad_norm 1.5937 (1.5465) [2021-04-16 05:35:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][330/1251] eta 0:04:22 lr 0.000498 time 0.3036 (0.2845) loss 3.3495 (3.5645) grad_norm 1.7226 (1.5461) [2021-04-16 05:35:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][340/1251] eta 0:04:19 lr 0.000498 time 0.2784 (0.2849) loss 4.0300 (3.5722) grad_norm 1.3780 (1.5432) [2021-04-16 05:35:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][350/1251] eta 0:04:16 lr 0.000498 time 0.3005 (0.2847) loss 3.7184 (3.5685) grad_norm 1.5512 (1.5411) [2021-04-16 05:36:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][360/1251] eta 0:04:13 lr 0.000498 time 0.2847 (0.2848) loss 3.9530 (3.5724) grad_norm 1.4140 (1.5396) [2021-04-16 05:36:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][370/1251] eta 0:04:11 lr 0.000498 time 0.2832 (0.2852) loss 3.7432 (3.5712) grad_norm 1.5308 (1.5421) [2021-04-16 05:36:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][380/1251] eta 0:04:08 lr 0.000498 time 0.2813 (0.2851) loss 3.3973 (3.5685) grad_norm 1.6003 (1.5442) [2021-04-16 05:36:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][390/1251] eta 0:04:05 lr 0.000498 time 0.2816 (0.2850) loss 3.5654 (3.5672) grad_norm 1.6376 (1.5434) [2021-04-16 05:36:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][400/1251] eta 0:04:02 lr 0.000498 time 0.2975 (0.2850) loss 2.7112 (3.5613) grad_norm 1.9048 (1.5454) [2021-04-16 05:36:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][410/1251] eta 0:03:59 lr 0.000498 time 0.2756 (0.2848) loss 3.2664 (3.5549) grad_norm 1.7892 (1.5462) [2021-04-16 05:36:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][420/1251] eta 0:03:56 lr 0.000498 time 0.2782 (0.2848) loss 3.9793 (3.5610) grad_norm 1.4176 (1.5443) [2021-04-16 05:36:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][430/1251] eta 0:03:53 lr 0.000498 time 0.2594 (0.2846) loss 3.3642 (3.5642) grad_norm 1.4371 (1.5439) [2021-04-16 05:36:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][440/1251] eta 0:03:50 lr 0.000498 time 0.2911 (0.2844) loss 4.2908 (3.5625) grad_norm 1.9075 (1.5445) [2021-04-16 05:36:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][450/1251] eta 0:03:47 lr 0.000498 time 0.2685 (0.2841) loss 2.8605 (3.5605) grad_norm 1.5881 (1.5440) [2021-04-16 05:36:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][460/1251] eta 0:03:44 lr 0.000498 time 0.2745 (0.2839) loss 3.9786 (3.5594) grad_norm 1.4105 (1.5423) [2021-04-16 05:36:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][470/1251] eta 0:03:41 lr 0.000498 time 0.2757 (0.2839) loss 3.6953 (3.5638) grad_norm 1.4785 (1.5413) [2021-04-16 05:36:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][480/1251] eta 0:03:38 lr 0.000498 time 0.2685 (0.2838) loss 2.2978 (3.5567) grad_norm 1.4913 (1.5409) [2021-04-16 05:36:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][490/1251] eta 0:03:35 lr 0.000498 time 0.3055 (0.2836) loss 3.6964 (3.5526) grad_norm 1.4845 (1.5401) [2021-04-16 05:36:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][500/1251] eta 0:03:32 lr 0.000498 time 0.2531 (0.2835) loss 4.6248 (3.5563) grad_norm 1.7313 (1.5400) [2021-04-16 05:36:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][510/1251] eta 0:03:30 lr 0.000498 time 0.2749 (0.2835) loss 3.6367 (3.5556) grad_norm 1.6815 (1.5415) [2021-04-16 05:36:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][520/1251] eta 0:03:27 lr 0.000498 time 0.2796 (0.2836) loss 2.2341 (3.5512) grad_norm 1.6944 (1.5424) [2021-04-16 05:36:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][530/1251] eta 0:03:24 lr 0.000498 time 0.2862 (0.2836) loss 4.4186 (3.5546) grad_norm 1.4900 (1.5408) [2021-04-16 05:36:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][540/1251] eta 0:03:21 lr 0.000498 time 0.2763 (0.2834) loss 4.0315 (3.5566) grad_norm 1.5372 (1.5394) [2021-04-16 05:36:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][550/1251] eta 0:03:18 lr 0.000498 time 0.2858 (0.2835) loss 2.2896 (3.5584) grad_norm 1.4339 (1.5387) [2021-04-16 05:36:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][560/1251] eta 0:03:15 lr 0.000497 time 0.2549 (0.2834) loss 2.4474 (3.5594) grad_norm 1.6056 (1.5388) [2021-04-16 05:37:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][570/1251] eta 0:03:13 lr 0.000497 time 0.2787 (0.2836) loss 3.1367 (3.5582) grad_norm 1.5901 (1.5392) [2021-04-16 05:37:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][580/1251] eta 0:03:10 lr 0.000497 time 0.2661 (0.2834) loss 3.4010 (3.5585) grad_norm 1.6035 (1.5389) [2021-04-16 05:37:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][590/1251] eta 0:03:07 lr 0.000497 time 0.2873 (0.2833) loss 3.9090 (3.5579) grad_norm 1.6458 (1.5404) [2021-04-16 05:37:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][600/1251] eta 0:03:04 lr 0.000497 time 0.3918 (0.2833) loss 3.6248 (3.5560) grad_norm 1.4798 (1.5406) [2021-04-16 05:37:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][610/1251] eta 0:03:01 lr 0.000497 time 0.2757 (0.2832) loss 3.6930 (3.5574) grad_norm 1.7252 (1.5414) [2021-04-16 05:37:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][620/1251] eta 0:02:58 lr 0.000497 time 0.2690 (0.2830) loss 2.3215 (3.5519) grad_norm 1.5061 (1.5405) [2021-04-16 05:37:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][630/1251] eta 0:02:55 lr 0.000497 time 0.2773 (0.2831) loss 3.1750 (3.5553) grad_norm 1.5438 (1.5400) [2021-04-16 05:37:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][640/1251] eta 0:02:53 lr 0.000497 time 0.2875 (0.2832) loss 3.6465 (3.5557) grad_norm 1.4990 (1.5404) [2021-04-16 05:37:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][650/1251] eta 0:02:50 lr 0.000497 time 0.2885 (0.2830) loss 2.4132 (3.5559) grad_norm 1.5029 (1.5398) [2021-04-16 05:37:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][660/1251] eta 0:02:47 lr 0.000497 time 0.2925 (0.2830) loss 3.9181 (3.5559) grad_norm 1.3620 (1.5385) [2021-04-16 05:37:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][670/1251] eta 0:02:44 lr 0.000497 time 0.2988 (0.2830) loss 3.0743 (3.5549) grad_norm 1.7139 (1.5387) [2021-04-16 05:37:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][680/1251] eta 0:02:41 lr 0.000497 time 0.2725 (0.2829) loss 3.6073 (3.5562) grad_norm 1.6622 (1.5398) [2021-04-16 05:37:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][690/1251] eta 0:02:38 lr 0.000497 time 0.2681 (0.2830) loss 3.5623 (3.5570) grad_norm 1.6764 (1.5394) [2021-04-16 05:37:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][700/1251] eta 0:02:35 lr 0.000497 time 0.2902 (0.2829) loss 4.4653 (3.5558) grad_norm 1.5466 (1.5396) [2021-04-16 05:37:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][710/1251] eta 0:02:32 lr 0.000497 time 0.2906 (0.2828) loss 3.3690 (3.5547) grad_norm 1.3874 (1.5400) [2021-04-16 05:37:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][720/1251] eta 0:02:30 lr 0.000497 time 0.2892 (0.2829) loss 4.0333 (3.5575) grad_norm 1.6027 (1.5403) [2021-04-16 05:37:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][730/1251] eta 0:02:27 lr 0.000497 time 0.2567 (0.2828) loss 3.7341 (3.5580) grad_norm 1.4480 (1.5413) [2021-04-16 05:37:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][740/1251] eta 0:02:24 lr 0.000497 time 0.2764 (0.2829) loss 3.4622 (3.5590) grad_norm 1.3786 (1.5414) [2021-04-16 05:37:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][750/1251] eta 0:02:21 lr 0.000497 time 0.2826 (0.2828) loss 4.0925 (3.5575) grad_norm 1.8257 (1.5406) [2021-04-16 05:37:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][760/1251] eta 0:02:18 lr 0.000497 time 0.3029 (0.2829) loss 3.6578 (3.5607) grad_norm 1.4311 (1.5421) [2021-04-16 05:37:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][770/1251] eta 0:02:16 lr 0.000497 time 0.2907 (0.2828) loss 3.8685 (3.5575) grad_norm 1.5513 (1.5404) [2021-04-16 05:38:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][780/1251] eta 0:02:13 lr 0.000497 time 0.2660 (0.2827) loss 3.9834 (3.5583) grad_norm 1.5377 (1.5405) [2021-04-16 05:38:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][790/1251] eta 0:02:10 lr 0.000497 time 0.3112 (0.2826) loss 3.9732 (3.5569) grad_norm 1.4232 (1.5393) [2021-04-16 05:38:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][800/1251] eta 0:02:07 lr 0.000497 time 0.2926 (0.2825) loss 3.6716 (3.5572) grad_norm 1.7405 (1.5390) [2021-04-16 05:38:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][810/1251] eta 0:02:04 lr 0.000496 time 0.2858 (0.2824) loss 3.6581 (3.5543) grad_norm 1.4938 (1.5402) [2021-04-16 05:38:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][820/1251] eta 0:02:01 lr 0.000496 time 0.2917 (0.2824) loss 4.5320 (3.5562) grad_norm 1.8592 (1.5410) [2021-04-16 05:38:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][830/1251] eta 0:01:58 lr 0.000496 time 0.2799 (0.2824) loss 4.4591 (3.5590) grad_norm 1.6938 (1.5422) [2021-04-16 05:38:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][840/1251] eta 0:01:56 lr 0.000496 time 0.2796 (0.2824) loss 3.8075 (3.5615) grad_norm 1.4370 (1.5432) [2021-04-16 05:38:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][850/1251] eta 0:01:53 lr 0.000496 time 0.2533 (0.2823) loss 2.4325 (3.5607) grad_norm 2.0851 (1.5450) [2021-04-16 05:38:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][860/1251] eta 0:01:50 lr 0.000496 time 0.2634 (0.2822) loss 4.3084 (3.5608) grad_norm 1.4530 (1.5469) [2021-04-16 05:38:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][870/1251] eta 0:01:47 lr 0.000496 time 0.2779 (0.2821) loss 2.5658 (3.5584) grad_norm 1.5060 (1.5467) [2021-04-16 05:38:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][880/1251] eta 0:01:44 lr 0.000496 time 0.2812 (0.2820) loss 3.8279 (3.5575) grad_norm 1.2232 (1.5454) [2021-04-16 05:38:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][890/1251] eta 0:01:41 lr 0.000496 time 0.2845 (0.2819) loss 3.6741 (3.5571) grad_norm 1.4496 (1.5446) [2021-04-16 05:38:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][900/1251] eta 0:01:38 lr 0.000496 time 0.2874 (0.2819) loss 4.0586 (3.5552) grad_norm 1.5758 (1.5446) [2021-04-16 05:38:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][910/1251] eta 0:01:36 lr 0.000496 time 0.2737 (0.2818) loss 2.2674 (3.5530) grad_norm 1.5272 (1.5448) [2021-04-16 05:38:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][920/1251] eta 0:01:33 lr 0.000496 time 0.2821 (0.2819) loss 3.6421 (3.5532) grad_norm 1.6711 (1.5442) [2021-04-16 05:38:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][930/1251] eta 0:01:30 lr 0.000496 time 0.2623 (0.2818) loss 3.6720 (3.5526) grad_norm 1.4855 (1.5433) [2021-04-16 05:38:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][940/1251] eta 0:01:27 lr 0.000496 time 0.2874 (0.2820) loss 3.9831 (3.5485) grad_norm 1.5764 (1.5418) [2021-04-16 05:38:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][950/1251] eta 0:01:24 lr 0.000496 time 0.2512 (0.2819) loss 2.2187 (3.5467) grad_norm 1.4808 (1.5418) [2021-04-16 05:38:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][960/1251] eta 0:01:22 lr 0.000496 time 0.2496 (0.2819) loss 2.9526 (3.5474) grad_norm 1.5213 (1.5412) [2021-04-16 05:38:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][970/1251] eta 0:01:19 lr 0.000496 time 0.2690 (0.2818) loss 4.4896 (3.5490) grad_norm 1.5543 (1.5410) [2021-04-16 05:38:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][980/1251] eta 0:01:16 lr 0.000496 time 0.2586 (0.2818) loss 3.4620 (3.5474) grad_norm 1.9175 (1.5419) [2021-04-16 05:38:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][990/1251] eta 0:01:13 lr 0.000496 time 0.2601 (0.2818) loss 3.4758 (3.5447) grad_norm 1.4207 (1.5422) [2021-04-16 05:39:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][1000/1251] eta 0:01:10 lr 0.000496 time 0.3115 (0.2818) loss 3.8537 (3.5445) grad_norm 1.4260 (1.5423) [2021-04-16 05:39:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][1010/1251] eta 0:01:07 lr 0.000496 time 0.2439 (0.2817) loss 4.1006 (3.5434) grad_norm 1.3483 (1.5416) [2021-04-16 05:39:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][1020/1251] eta 0:01:05 lr 0.000496 time 0.2785 (0.2817) loss 2.8947 (3.5431) grad_norm 1.4429 (1.5412) [2021-04-16 05:39:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][1030/1251] eta 0:01:02 lr 0.000496 time 0.2977 (0.2817) loss 4.6314 (3.5440) grad_norm 2.0776 (1.5429) [2021-04-16 05:39:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][1040/1251] eta 0:00:59 lr 0.000496 time 0.2810 (0.2816) loss 3.2395 (3.5458) grad_norm 1.8173 (1.5437) [2021-04-16 05:39:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][1050/1251] eta 0:00:56 lr 0.000495 time 0.2868 (0.2816) loss 2.6327 (3.5427) grad_norm 1.5266 (1.5437) [2021-04-16 05:39:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][1060/1251] eta 0:00:53 lr 0.000495 time 0.2649 (0.2816) loss 4.4201 (3.5453) grad_norm 1.5411 (1.5433) [2021-04-16 05:39:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][1070/1251] eta 0:00:50 lr 0.000495 time 0.3023 (0.2816) loss 2.9680 (3.5429) grad_norm 1.3838 (1.5427) [2021-04-16 05:39:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][1080/1251] eta 0:00:48 lr 0.000495 time 0.2692 (0.2815) loss 3.3009 (3.5451) grad_norm 1.6455 (1.5426) [2021-04-16 05:39:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][1090/1251] eta 0:00:45 lr 0.000495 time 0.2898 (0.2816) loss 2.4482 (3.5454) grad_norm 1.2817 (1.5426) [2021-04-16 05:39:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][1100/1251] eta 0:00:42 lr 0.000495 time 0.2800 (0.2815) loss 3.4166 (3.5426) grad_norm 1.6872 (1.5430) [2021-04-16 05:39:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][1110/1251] eta 0:00:39 lr 0.000495 time 0.2630 (0.2814) loss 3.7786 (3.5423) grad_norm 1.3638 (1.5427) [2021-04-16 05:39:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][1120/1251] eta 0:00:36 lr 0.000495 time 0.2778 (0.2814) loss 3.6698 (3.5437) grad_norm 1.3740 (1.5433) [2021-04-16 05:39:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][1130/1251] eta 0:00:34 lr 0.000495 time 0.2778 (0.2813) loss 3.8756 (3.5446) grad_norm 1.2950 (1.5436) [2021-04-16 05:39:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][1140/1251] eta 0:00:31 lr 0.000495 time 0.2617 (0.2813) loss 4.0438 (3.5440) grad_norm 1.8278 (1.5438) [2021-04-16 05:39:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][1150/1251] eta 0:00:28 lr 0.000495 time 0.2657 (0.2813) loss 4.1186 (3.5463) grad_norm 1.4787 (1.5441) [2021-04-16 05:39:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][1160/1251] eta 0:00:25 lr 0.000495 time 0.2863 (0.2813) loss 3.7106 (3.5458) grad_norm 1.4982 (1.5439) [2021-04-16 05:39:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][1170/1251] eta 0:00:22 lr 0.000495 time 0.2786 (0.2812) loss 3.9082 (3.5472) grad_norm 1.4288 (1.5446) [2021-04-16 05:39:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][1180/1251] eta 0:00:19 lr 0.000495 time 0.2602 (0.2813) loss 3.7260 (3.5472) grad_norm 1.6366 (1.5453) [2021-04-16 05:39:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][1190/1251] eta 0:00:17 lr 0.000495 time 0.3023 (0.2813) loss 4.2058 (3.5462) grad_norm 1.4941 (1.5447) [2021-04-16 05:39:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][1200/1251] eta 0:00:14 lr 0.000495 time 0.2658 (0.2812) loss 3.7537 (3.5459) grad_norm 1.6469 (1.5443) [2021-04-16 05:40:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][1210/1251] eta 0:00:11 lr 0.000495 time 0.2699 (0.2813) loss 3.1256 (3.5446) grad_norm 1.5110 (1.5439) [2021-04-16 05:40:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][1220/1251] eta 0:00:08 lr 0.000495 time 0.2731 (0.2813) loss 3.4208 (3.5437) grad_norm 1.4515 (1.5435) [2021-04-16 05:40:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][1230/1251] eta 0:00:05 lr 0.000495 time 0.2834 (0.2814) loss 3.7262 (3.5450) grad_norm 1.3584 (1.5428) [2021-04-16 05:40:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][1240/1251] eta 0:00:03 lr 0.000495 time 0.2620 (0.2813) loss 4.0837 (3.5465) grad_norm 1.3385 (1.5426) [2021-04-16 05:40:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [151/300][1250/1251] eta 0:00:00 lr 0.000495 time 0.2590 (0.2811) loss 3.7809 (3.5464) grad_norm 1.5220 (1.5428) [2021-04-16 05:40:16 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 151 training takes 0:05:56 [2021-04-16 05:40:16 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_151.pth saving...... [2021-04-16 05:40:31 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_151.pth saved !!! [2021-04-16 05:40:32 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.159 (1.159) Loss 1.0532 (1.0532) Acc@1 76.562 (76.562) Acc@5 92.676 (92.676) [2021-04-16 05:40:33 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.129 (0.210) Loss 1.0226 (1.0394) Acc@1 77.539 (75.994) Acc@5 93.066 (93.173) [2021-04-16 05:40:35 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.297 (0.198) Loss 1.0241 (1.0158) Acc@1 77.734 (76.595) Acc@5 94.238 (93.536) [2021-04-16 05:40:38 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.476 (0.223) Loss 1.0272 (1.0293) Acc@1 75.000 (76.194) Acc@5 93.848 (93.391) [2021-04-16 05:40:40 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.076 (0.209) Loss 1.0380 (1.0347) Acc@1 76.074 (75.988) Acc@5 92.480 (93.333) [2021-04-16 05:40:49 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 75.882 Acc@5 93.290 [2021-04-16 05:40:49 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 75.9% [2021-04-16 05:40:49 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 76.11% [2021-04-16 05:40:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][0/1251] eta 0:57:58 lr 0.000495 time 2.7804 (2.7804) loss 3.3491 (3.3491) grad_norm 1.5133 (1.5133) [2021-04-16 05:40:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][10/1251] eta 0:10:23 lr 0.000495 time 0.2935 (0.5023) loss 3.0550 (3.5907) grad_norm 1.6990 (1.5625) [2021-04-16 05:40:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][20/1251] eta 0:08:11 lr 0.000495 time 0.2791 (0.3996) loss 4.2904 (3.4736) grad_norm 1.5127 (1.5763) [2021-04-16 05:41:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][30/1251] eta 0:07:22 lr 0.000495 time 0.2859 (0.3626) loss 3.5843 (3.4597) grad_norm 1.5018 (1.5669) [2021-04-16 05:41:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][40/1251] eta 0:06:59 lr 0.000494 time 0.2931 (0.3464) loss 3.8726 (3.4788) grad_norm 1.4462 (1.5674) [2021-04-16 05:41:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][50/1251] eta 0:06:42 lr 0.000494 time 0.2775 (0.3352) loss 3.4095 (3.4852) grad_norm 1.6065 (1.5583) [2021-04-16 05:41:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][60/1251] eta 0:06:29 lr 0.000494 time 0.2876 (0.3274) loss 2.3704 (3.4486) grad_norm 1.3569 (1.5625) [2021-04-16 05:41:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][70/1251] eta 0:06:19 lr 0.000494 time 0.2789 (0.3210) loss 2.7384 (3.4513) grad_norm 1.4734 (1.5674) [2021-04-16 05:41:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][80/1251] eta 0:06:09 lr 0.000494 time 0.2721 (0.3159) loss 3.7705 (3.4758) grad_norm 1.5590 (1.5558) [2021-04-16 05:41:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][90/1251] eta 0:06:02 lr 0.000494 time 0.3138 (0.3124) loss 3.7490 (3.4892) grad_norm 1.4714 (1.5516) [2021-04-16 05:41:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][100/1251] eta 0:05:54 lr 0.000494 time 0.2423 (0.3084) loss 2.3615 (3.4961) grad_norm 1.5418 (1.5562) [2021-04-16 05:41:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][110/1251] eta 0:05:48 lr 0.000494 time 0.2642 (0.3057) loss 4.1296 (3.4960) grad_norm 1.8415 (1.5581) [2021-04-16 05:41:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][120/1251] eta 0:05:44 lr 0.000494 time 0.2845 (0.3048) loss 3.6295 (3.5075) grad_norm 1.5943 (1.5633) [2021-04-16 05:41:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][130/1251] eta 0:05:39 lr 0.000494 time 0.2823 (0.3029) loss 3.8293 (3.5003) grad_norm 1.8623 (1.5569) [2021-04-16 05:41:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][140/1251] eta 0:05:36 lr 0.000494 time 0.2985 (0.3030) loss 4.4518 (3.5092) grad_norm 1.8387 (1.5534) [2021-04-16 05:41:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][150/1251] eta 0:05:32 lr 0.000494 time 0.2691 (0.3022) loss 4.1347 (3.5009) grad_norm 1.4127 (1.5598) [2021-04-16 05:41:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][160/1251] eta 0:05:27 lr 0.000494 time 0.2702 (0.3005) loss 3.3318 (3.4980) grad_norm 1.5224 (1.5657) [2021-04-16 05:41:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][170/1251] eta 0:05:23 lr 0.000494 time 0.3142 (0.2994) loss 3.6868 (3.5003) grad_norm 1.7108 (1.5614) [2021-04-16 05:41:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][180/1251] eta 0:05:19 lr 0.000494 time 0.2598 (0.2983) loss 3.6669 (3.5003) grad_norm 1.5515 (1.5619) [2021-04-16 05:41:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][190/1251] eta 0:05:15 lr 0.000494 time 0.2768 (0.2974) loss 3.6331 (3.5130) grad_norm 1.3990 (1.5588) [2021-04-16 05:41:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][200/1251] eta 0:05:12 lr 0.000494 time 0.2726 (0.2972) loss 3.8040 (3.5195) grad_norm 1.3485 (1.5596) [2021-04-16 05:41:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][210/1251] eta 0:05:09 lr 0.000494 time 0.2755 (0.2970) loss 4.4302 (3.5234) grad_norm 1.5330 (1.5577) [2021-04-16 05:41:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][220/1251] eta 0:05:05 lr 0.000494 time 0.2873 (0.2961) loss 4.1383 (3.5136) grad_norm 1.4296 (1.5568) [2021-04-16 05:41:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][230/1251] eta 0:05:01 lr 0.000494 time 0.2640 (0.2956) loss 3.9083 (3.5138) grad_norm 1.3138 (1.5559) [2021-04-16 05:42:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][240/1251] eta 0:04:58 lr 0.000494 time 0.2782 (0.2949) loss 3.8340 (3.5046) grad_norm 1.5697 (1.5590) [2021-04-16 05:42:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][250/1251] eta 0:04:54 lr 0.000494 time 0.2823 (0.2942) loss 3.2060 (3.5045) grad_norm 1.6669 (1.5607) [2021-04-16 05:42:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][260/1251] eta 0:04:51 lr 0.000494 time 0.3160 (0.2938) loss 4.5056 (3.5153) grad_norm 1.5296 (1.5608) [2021-04-16 05:42:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][270/1251] eta 0:04:47 lr 0.000494 time 0.2801 (0.2933) loss 3.9756 (3.5144) grad_norm 1.4576 (1.5576) [2021-04-16 05:42:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][280/1251] eta 0:04:44 lr 0.000493 time 0.2900 (0.2931) loss 4.1271 (3.5192) grad_norm 1.3691 (1.5562) [2021-04-16 05:42:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][290/1251] eta 0:04:41 lr 0.000493 time 0.2446 (0.2927) loss 3.6817 (3.5247) grad_norm 1.4997 (1.5516) [2021-04-16 05:42:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][300/1251] eta 0:04:38 lr 0.000493 time 0.2504 (0.2926) loss 3.2342 (3.5190) grad_norm 1.6751 (1.5517) [2021-04-16 05:42:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][310/1251] eta 0:04:34 lr 0.000493 time 0.2990 (0.2922) loss 2.9959 (3.5173) grad_norm 1.6460 (1.5507) [2021-04-16 05:42:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][320/1251] eta 0:04:32 lr 0.000493 time 0.2906 (0.2922) loss 3.6614 (3.5094) grad_norm 1.5493 (1.5517) [2021-04-16 05:42:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][330/1251] eta 0:04:28 lr 0.000493 time 0.2987 (0.2919) loss 2.8591 (3.5085) grad_norm 1.5302 (1.5547) [2021-04-16 05:42:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][340/1251] eta 0:04:25 lr 0.000493 time 0.3246 (0.2919) loss 3.5373 (3.5047) grad_norm 1.5483 (1.5544) [2021-04-16 05:42:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][350/1251] eta 0:04:22 lr 0.000493 time 0.2863 (0.2916) loss 2.8939 (3.5093) grad_norm 1.5136 (1.5531) [2021-04-16 05:42:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][360/1251] eta 0:04:19 lr 0.000493 time 0.2654 (0.2917) loss 2.9725 (3.5078) grad_norm 1.6111 (1.5540) [2021-04-16 05:42:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][370/1251] eta 0:04:16 lr 0.000493 time 0.2714 (0.2914) loss 3.7489 (3.5043) grad_norm 1.8710 (1.5564) [2021-04-16 05:42:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][380/1251] eta 0:04:14 lr 0.000493 time 0.2898 (0.2917) loss 4.5976 (3.5066) grad_norm 1.7727 (1.5586) [2021-04-16 05:42:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][390/1251] eta 0:04:11 lr 0.000493 time 0.2893 (0.2916) loss 2.1786 (3.5004) grad_norm 1.5846 (1.5563) [2021-04-16 05:42:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][400/1251] eta 0:04:08 lr 0.000493 time 0.2716 (0.2917) loss 3.7532 (3.5129) grad_norm 1.5163 (1.5542) [2021-04-16 05:42:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][410/1251] eta 0:04:05 lr 0.000493 time 0.2760 (0.2913) loss 3.8510 (3.5064) grad_norm 1.6852 (1.5532) [2021-04-16 05:42:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][420/1251] eta 0:04:02 lr 0.000493 time 0.2737 (0.2913) loss 3.4567 (3.5080) grad_norm 1.4975 (1.5527) [2021-04-16 05:42:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][430/1251] eta 0:03:58 lr 0.000493 time 0.2816 (0.2910) loss 2.8919 (3.5105) grad_norm 1.3714 (1.5521) [2021-04-16 05:42:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][440/1251] eta 0:03:55 lr 0.000493 time 0.3027 (0.2907) loss 2.3795 (3.5055) grad_norm 1.4528 (1.5516) [2021-04-16 05:43:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][450/1251] eta 0:03:52 lr 0.000493 time 0.2676 (0.2905) loss 2.4769 (3.5032) grad_norm 1.3588 (1.5507) [2021-04-16 05:43:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][460/1251] eta 0:03:49 lr 0.000493 time 0.2541 (0.2902) loss 3.3050 (3.5004) grad_norm 1.3948 (1.5496) [2021-04-16 05:43:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][470/1251] eta 0:03:46 lr 0.000493 time 0.2932 (0.2899) loss 3.8324 (3.5004) grad_norm 1.4461 (1.5520) [2021-04-16 05:43:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][480/1251] eta 0:03:43 lr 0.000493 time 0.2952 (0.2897) loss 3.5061 (3.4973) grad_norm 1.6229 (1.5521) [2021-04-16 05:43:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][490/1251] eta 0:03:40 lr 0.000493 time 0.2658 (0.2898) loss 4.1624 (3.5053) grad_norm 1.8999 (1.5554) [2021-04-16 05:43:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][500/1251] eta 0:03:37 lr 0.000493 time 0.2658 (0.2894) loss 3.3438 (3.5070) grad_norm 1.5401 (1.5554) [2021-04-16 05:43:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][510/1251] eta 0:03:34 lr 0.000493 time 0.2688 (0.2892) loss 2.3220 (3.5070) grad_norm 1.7351 (1.5573) [2021-04-16 05:43:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][520/1251] eta 0:03:31 lr 0.000492 time 0.2691 (0.2891) loss 3.6606 (3.4978) grad_norm 1.9095 (1.5561) [2021-04-16 05:43:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][530/1251] eta 0:03:28 lr 0.000492 time 0.2537 (0.2889) loss 2.8981 (3.4962) grad_norm 1.8581 (1.5568) [2021-04-16 05:43:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][540/1251] eta 0:03:25 lr 0.000492 time 0.2672 (0.2888) loss 2.2973 (3.4923) grad_norm 1.5848 (1.5567) [2021-04-16 05:43:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][550/1251] eta 0:03:22 lr 0.000492 time 0.2953 (0.2887) loss 4.1596 (3.4914) grad_norm 1.5707 (1.5552) [2021-04-16 05:43:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][560/1251] eta 0:03:19 lr 0.000492 time 0.2845 (0.2886) loss 4.3170 (3.4927) grad_norm 1.6871 (1.5548) [2021-04-16 05:43:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][570/1251] eta 0:03:16 lr 0.000492 time 0.2639 (0.2890) loss 3.5791 (3.4871) grad_norm 1.6879 (1.5549) [2021-04-16 05:43:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][580/1251] eta 0:03:13 lr 0.000492 time 0.3081 (0.2888) loss 3.3364 (3.4905) grad_norm 1.4584 (1.5556) [2021-04-16 05:43:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][590/1251] eta 0:03:10 lr 0.000492 time 0.2791 (0.2888) loss 3.6995 (3.4908) grad_norm 1.6367 (1.5555) [2021-04-16 05:43:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][600/1251] eta 0:03:07 lr 0.000492 time 0.2571 (0.2887) loss 3.7326 (3.4909) grad_norm 1.7239 (1.5559) [2021-04-16 05:43:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][610/1251] eta 0:03:04 lr 0.000492 time 0.2814 (0.2885) loss 4.2268 (3.4990) grad_norm 1.5659 (1.5582) [2021-04-16 05:43:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][620/1251] eta 0:03:02 lr 0.000492 time 0.2706 (0.2885) loss 3.7514 (3.5005) grad_norm 1.7887 (1.5584) [2021-04-16 05:43:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][630/1251] eta 0:02:59 lr 0.000492 time 0.2965 (0.2883) loss 2.7827 (3.5012) grad_norm 1.5510 (1.5588) [2021-04-16 05:43:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][640/1251] eta 0:02:56 lr 0.000492 time 0.2845 (0.2881) loss 2.8218 (3.4987) grad_norm 1.6312 (1.5580) [2021-04-16 05:43:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][650/1251] eta 0:02:53 lr 0.000492 time 0.2790 (0.2880) loss 3.5994 (3.5007) grad_norm 1.4443 (1.5571) [2021-04-16 05:43:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][660/1251] eta 0:02:50 lr 0.000492 time 0.2873 (0.2880) loss 3.1910 (3.4997) grad_norm 1.3652 (1.5568) [2021-04-16 05:44:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][670/1251] eta 0:02:47 lr 0.000492 time 0.2816 (0.2878) loss 2.5871 (3.5012) grad_norm 1.4724 (1.5566) [2021-04-16 05:44:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][680/1251] eta 0:02:44 lr 0.000492 time 0.2616 (0.2877) loss 3.8541 (3.5004) grad_norm 1.6400 (1.5545) [2021-04-16 05:44:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][690/1251] eta 0:02:41 lr 0.000492 time 0.2782 (0.2877) loss 2.8140 (3.4988) grad_norm 1.3696 (1.5534) [2021-04-16 05:44:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][700/1251] eta 0:02:38 lr 0.000492 time 0.2805 (0.2876) loss 2.4851 (3.4985) grad_norm 1.3426 (1.5519) [2021-04-16 05:44:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][710/1251] eta 0:02:35 lr 0.000492 time 0.2832 (0.2876) loss 3.0750 (3.4929) grad_norm 1.8984 (1.5545) [2021-04-16 05:44:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][720/1251] eta 0:02:32 lr 0.000492 time 0.2958 (0.2876) loss 3.9410 (3.4948) grad_norm 1.3150 (1.5547) [2021-04-16 05:44:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][730/1251] eta 0:02:29 lr 0.000492 time 0.2740 (0.2876) loss 2.7632 (3.4944) grad_norm 1.4286 (1.5557) [2021-04-16 05:44:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][740/1251] eta 0:02:26 lr 0.000492 time 0.2981 (0.2876) loss 3.1771 (3.4932) grad_norm 1.2812 (1.5547) [2021-04-16 05:44:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][750/1251] eta 0:02:24 lr 0.000492 time 0.2561 (0.2875) loss 2.5568 (3.4931) grad_norm 1.3462 (1.5535) [2021-04-16 05:44:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][760/1251] eta 0:02:21 lr 0.000491 time 0.2793 (0.2873) loss 2.0931 (3.4935) grad_norm 1.3649 (1.5542) [2021-04-16 05:44:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][770/1251] eta 0:02:18 lr 0.000491 time 0.2779 (0.2872) loss 2.8771 (3.4938) grad_norm 1.5174 (1.5551) [2021-04-16 05:44:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][780/1251] eta 0:02:15 lr 0.000491 time 0.2882 (0.2873) loss 3.7326 (3.4931) grad_norm 1.7939 (1.5566) [2021-04-16 05:44:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][790/1251] eta 0:02:12 lr 0.000491 time 0.2822 (0.2872) loss 2.4843 (3.4932) grad_norm 1.5315 (1.5570) [2021-04-16 05:44:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][800/1251] eta 0:02:09 lr 0.000491 time 0.2912 (0.2871) loss 3.3771 (3.4968) grad_norm 1.6224 (1.5563) [2021-04-16 05:44:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][810/1251] eta 0:02:06 lr 0.000491 time 0.3096 (0.2870) loss 2.6888 (3.4980) grad_norm 1.5297 (1.5571) [2021-04-16 05:44:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][820/1251] eta 0:02:03 lr 0.000491 time 0.2742 (0.2869) loss 3.9351 (3.4976) grad_norm 1.4685 (1.5581) [2021-04-16 05:44:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][830/1251] eta 0:02:00 lr 0.000491 time 0.2875 (0.2868) loss 2.7521 (3.4938) grad_norm 1.4225 (1.5582) [2021-04-16 05:44:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][840/1251] eta 0:01:57 lr 0.000491 time 0.2775 (0.2866) loss 4.4259 (3.4948) grad_norm 1.3374 (1.5578) [2021-04-16 05:44:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][850/1251] eta 0:01:54 lr 0.000491 time 0.2872 (0.2867) loss 3.8143 (3.4960) grad_norm 1.4046 (1.5580) [2021-04-16 05:44:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][860/1251] eta 0:01:52 lr 0.000491 time 0.2706 (0.2866) loss 2.8806 (3.4986) grad_norm 1.7725 (1.5587) [2021-04-16 05:44:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][870/1251] eta 0:01:49 lr 0.000491 time 0.2790 (0.2865) loss 3.5385 (3.4984) grad_norm 1.5149 (1.5605) [2021-04-16 05:45:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][880/1251] eta 0:01:46 lr 0.000491 time 0.2580 (0.2866) loss 3.8053 (3.5020) grad_norm 1.3373 (1.5599) [2021-04-16 05:45:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][890/1251] eta 0:01:43 lr 0.000491 time 0.2755 (0.2867) loss 2.9055 (3.5029) grad_norm 1.5606 (1.5589) [2021-04-16 05:45:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][900/1251] eta 0:01:40 lr 0.000491 time 0.2836 (0.2866) loss 3.9017 (3.5024) grad_norm 1.5003 (1.5582) [2021-04-16 05:45:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][910/1251] eta 0:01:37 lr 0.000491 time 0.2770 (0.2865) loss 3.7648 (3.5065) grad_norm 1.7415 (1.5583) [2021-04-16 05:45:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][920/1251] eta 0:01:34 lr 0.000491 time 0.2708 (0.2863) loss 4.1571 (3.5053) grad_norm 1.4602 (1.5573) [2021-04-16 05:45:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][930/1251] eta 0:01:31 lr 0.000491 time 0.2781 (0.2862) loss 4.3462 (3.5059) grad_norm 1.5571 (1.5571) [2021-04-16 05:45:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][940/1251] eta 0:01:29 lr 0.000491 time 0.2708 (0.2862) loss 3.5656 (3.5051) grad_norm 1.4691 (1.5570) [2021-04-16 05:45:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][950/1251] eta 0:01:26 lr 0.000491 time 0.2688 (0.2862) loss 3.3364 (3.5065) grad_norm 1.6201 (1.5563) [2021-04-16 05:45:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][960/1251] eta 0:01:23 lr 0.000491 time 0.2688 (0.2862) loss 4.1313 (3.5101) grad_norm 1.4683 (1.5563) [2021-04-16 05:45:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][970/1251] eta 0:01:20 lr 0.000491 time 0.2761 (0.2862) loss 2.3825 (3.5093) grad_norm 1.3779 (1.5557) [2021-04-16 05:45:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][980/1251] eta 0:01:17 lr 0.000491 time 0.2478 (0.2861) loss 3.3993 (3.5135) grad_norm 1.5756 (1.5559) [2021-04-16 05:45:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][990/1251] eta 0:01:14 lr 0.000491 time 0.2883 (0.2861) loss 3.8892 (3.5162) grad_norm 1.5161 (1.5554) [2021-04-16 05:45:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][1000/1251] eta 0:01:11 lr 0.000490 time 0.2885 (0.2860) loss 3.0650 (3.5150) grad_norm 1.4597 (1.5562) [2021-04-16 05:45:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][1010/1251] eta 0:01:08 lr 0.000490 time 0.2856 (0.2859) loss 2.7744 (3.5127) grad_norm 1.8037 (1.5583) [2021-04-16 05:45:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][1020/1251] eta 0:01:06 lr 0.000490 time 0.2735 (0.2858) loss 3.9007 (3.5155) grad_norm 1.4757 (1.5594) [2021-04-16 05:45:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][1030/1251] eta 0:01:03 lr 0.000490 time 0.2695 (0.2858) loss 3.5401 (3.5165) grad_norm 1.5915 (1.5592) [2021-04-16 05:45:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][1040/1251] eta 0:01:00 lr 0.000490 time 0.2741 (0.2857) loss 3.5348 (3.5172) grad_norm 1.9788 (1.5593) [2021-04-16 05:45:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][1050/1251] eta 0:00:57 lr 0.000490 time 0.2534 (0.2856) loss 3.7019 (3.5171) grad_norm 1.4084 (1.5592) [2021-04-16 05:45:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][1060/1251] eta 0:00:54 lr 0.000490 time 0.2628 (0.2855) loss 4.3607 (3.5198) grad_norm 1.5247 (1.5595) [2021-04-16 05:45:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][1070/1251] eta 0:00:51 lr 0.000490 time 0.2769 (0.2856) loss 3.9610 (3.5200) grad_norm 1.4381 (1.5594) [2021-04-16 05:45:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][1080/1251] eta 0:00:48 lr 0.000490 time 0.2788 (0.2856) loss 2.7036 (3.5195) grad_norm 1.2851 (1.5595) [2021-04-16 05:46:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][1090/1251] eta 0:00:45 lr 0.000490 time 0.2583 (0.2856) loss 3.7890 (3.5228) grad_norm 1.3913 (1.5599) [2021-04-16 05:46:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][1100/1251] eta 0:00:43 lr 0.000490 time 0.2959 (0.2856) loss 2.9664 (3.5220) grad_norm 1.8278 (1.5596) [2021-04-16 05:46:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][1110/1251] eta 0:00:40 lr 0.000490 time 0.2648 (0.2855) loss 3.8438 (3.5241) grad_norm 1.2498 (1.5587) [2021-04-16 05:46:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][1120/1251] eta 0:00:37 lr 0.000490 time 0.2873 (0.2855) loss 4.2368 (3.5246) grad_norm 1.5030 (1.5587) [2021-04-16 05:46:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][1130/1251] eta 0:00:34 lr 0.000490 time 0.2906 (0.2854) loss 4.0129 (3.5235) grad_norm 1.9568 (1.5591) [2021-04-16 05:46:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][1140/1251] eta 0:00:31 lr 0.000490 time 0.2811 (0.2854) loss 3.9543 (3.5223) grad_norm 1.4739 (1.5589) [2021-04-16 05:46:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][1150/1251] eta 0:00:28 lr 0.000490 time 0.2989 (0.2854) loss 3.9827 (3.5218) grad_norm 1.4799 (1.5588) [2021-04-16 05:46:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][1160/1251] eta 0:00:25 lr 0.000490 time 0.2425 (0.2853) loss 3.6336 (3.5204) grad_norm 1.4511 (1.5586) [2021-04-16 05:46:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][1170/1251] eta 0:00:23 lr 0.000490 time 0.2634 (0.2852) loss 2.7241 (3.5217) grad_norm 1.4666 (1.5579) [2021-04-16 05:46:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][1180/1251] eta 0:00:20 lr 0.000490 time 0.2962 (0.2852) loss 2.4152 (3.5210) grad_norm 1.4687 (1.5573) [2021-04-16 05:46:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][1190/1251] eta 0:00:17 lr 0.000490 time 0.3108 (0.2853) loss 3.3660 (3.5192) grad_norm 1.3831 (1.5560) [2021-04-16 05:46:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][1200/1251] eta 0:00:14 lr 0.000490 time 0.2823 (0.2853) loss 4.2544 (3.5199) grad_norm 1.6318 (1.5556) [2021-04-16 05:46:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][1210/1251] eta 0:00:11 lr 0.000490 time 0.2825 (0.2852) loss 3.3523 (3.5198) grad_norm 1.5600 (1.5556) [2021-04-16 05:46:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][1220/1251] eta 0:00:08 lr 0.000490 time 0.2811 (0.2852) loss 4.2837 (3.5212) grad_norm 1.4157 (1.5552) [2021-04-16 05:46:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][1230/1251] eta 0:00:05 lr 0.000490 time 0.2653 (0.2851) loss 3.8581 (3.5235) grad_norm 1.5921 (1.5550) [2021-04-16 05:46:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][1240/1251] eta 0:00:03 lr 0.000489 time 0.2591 (0.2850) loss 3.7262 (3.5225) grad_norm 1.7501 (1.5549) [2021-04-16 05:46:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [152/300][1250/1251] eta 0:00:00 lr 0.000489 time 0.2480 (0.2847) loss 3.2152 (3.5220) grad_norm 1.2702 (1.5549) [2021-04-16 05:46:50 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 152 training takes 0:06:00 [2021-04-16 05:46:50 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_152.pth saving...... [2021-04-16 05:47:02 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_152.pth saved !!! [2021-04-16 05:47:03 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.075 (1.075) Loss 1.0011 (1.0011) Acc@1 76.465 (76.465) Acc@5 93.848 (93.848) [2021-04-16 05:47:04 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.112 (0.194) Loss 0.9272 (0.9989) Acc@1 77.734 (76.784) Acc@5 95.117 (93.679) [2021-04-16 05:47:07 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.560 (0.233) Loss 1.0194 (1.0214) Acc@1 76.855 (76.195) Acc@5 92.676 (93.304) [2021-04-16 05:47:09 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.092 (0.225) Loss 0.9715 (1.0152) Acc@1 77.832 (76.200) Acc@5 94.727 (93.407) [2021-04-16 05:47:10 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.214) Loss 0.9579 (1.0164) Acc@1 77.832 (76.210) Acc@5 93.164 (93.333) [2021-04-16 05:47:17 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 76.086 Acc@5 93.256 [2021-04-16 05:47:17 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 76.1% [2021-04-16 05:47:17 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 76.11% [2021-04-16 05:47:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][0/1251] eta 1:35:21 lr 0.000489 time 4.5732 (4.5732) loss 3.7383 (3.7383) grad_norm 1.6829 (1.6829) [2021-04-16 05:47:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][10/1251] eta 0:13:47 lr 0.000489 time 0.2845 (0.6665) loss 3.0661 (3.4451) grad_norm 1.9111 (1.5558) [2021-04-16 05:47:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][20/1251] eta 0:09:50 lr 0.000489 time 0.2566 (0.4793) loss 4.1847 (3.5777) grad_norm 1.3726 (1.5170) [2021-04-16 05:47:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][30/1251] eta 0:08:25 lr 0.000489 time 0.2678 (0.4143) loss 2.6733 (3.4934) grad_norm 1.5685 (1.5308) [2021-04-16 05:47:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][40/1251] eta 0:07:41 lr 0.000489 time 0.2714 (0.3810) loss 3.4108 (3.4460) grad_norm 1.3603 (1.5210) [2021-04-16 05:47:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][50/1251] eta 0:07:13 lr 0.000489 time 0.2631 (0.3610) loss 3.3926 (3.4687) grad_norm 1.6261 (1.5243) [2021-04-16 05:47:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][60/1251] eta 0:06:53 lr 0.000489 time 0.2734 (0.3473) loss 3.2926 (3.5132) grad_norm 1.7017 (1.5308) [2021-04-16 05:47:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][70/1251] eta 0:06:39 lr 0.000489 time 0.2948 (0.3381) loss 4.1720 (3.5359) grad_norm 1.5429 (1.5350) [2021-04-16 05:47:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][80/1251] eta 0:06:28 lr 0.000489 time 0.2783 (0.3321) loss 4.0316 (3.5310) grad_norm 1.3628 (1.5343) [2021-04-16 05:47:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][90/1251] eta 0:06:19 lr 0.000489 time 0.2968 (0.3264) loss 3.8327 (3.5102) grad_norm 1.5675 (1.5508) [2021-04-16 05:47:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][100/1251] eta 0:06:10 lr 0.000489 time 0.2675 (0.3217) loss 2.3939 (3.5273) grad_norm 1.7022 (1.5488) [2021-04-16 05:47:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][110/1251] eta 0:06:02 lr 0.000489 time 0.3016 (0.3181) loss 3.8729 (3.5487) grad_norm 1.5547 (1.5548) [2021-04-16 05:47:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][120/1251] eta 0:05:55 lr 0.000489 time 0.2722 (0.3143) loss 2.4361 (3.5259) grad_norm 1.6716 (1.5555) [2021-04-16 05:47:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][130/1251] eta 0:05:48 lr 0.000489 time 0.2669 (0.3113) loss 3.0949 (3.5462) grad_norm 1.6096 (1.5484) [2021-04-16 05:48:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][140/1251] eta 0:05:43 lr 0.000489 time 0.2726 (0.3090) loss 4.2575 (3.5533) grad_norm 1.4548 (1.5433) [2021-04-16 05:48:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][150/1251] eta 0:05:38 lr 0.000489 time 0.2847 (0.3077) loss 3.7592 (3.5521) grad_norm 1.3069 (1.5363) [2021-04-16 05:48:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][160/1251] eta 0:05:34 lr 0.000489 time 0.2696 (0.3062) loss 3.8975 (3.5658) grad_norm 1.4243 (1.5312) [2021-04-16 05:48:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][170/1251] eta 0:05:30 lr 0.000489 time 0.2639 (0.3053) loss 2.9519 (3.5585) grad_norm 1.3394 (1.5281) [2021-04-16 05:48:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][180/1251] eta 0:05:25 lr 0.000489 time 0.2798 (0.3038) loss 4.2058 (3.5585) grad_norm 1.5359 (1.5320) [2021-04-16 05:48:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][190/1251] eta 0:05:21 lr 0.000489 time 0.2907 (0.3032) loss 3.8927 (3.5560) grad_norm 1.5230 (1.5346) [2021-04-16 05:48:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][200/1251] eta 0:05:17 lr 0.000489 time 0.2554 (0.3022) loss 3.8164 (3.5645) grad_norm 1.6860 (1.5384) [2021-04-16 05:48:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][210/1251] eta 0:05:13 lr 0.000489 time 0.2755 (0.3012) loss 3.9200 (3.5702) grad_norm 1.8501 (1.5418) [2021-04-16 05:48:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][220/1251] eta 0:05:09 lr 0.000489 time 0.2881 (0.3001) loss 2.2318 (3.5699) grad_norm 1.3612 (1.5427) [2021-04-16 05:48:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][230/1251] eta 0:05:05 lr 0.000488 time 0.2771 (0.2992) loss 3.5644 (3.5629) grad_norm 1.5129 (1.5403) [2021-04-16 05:48:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][240/1251] eta 0:05:01 lr 0.000488 time 0.2529 (0.2981) loss 3.2999 (3.5629) grad_norm 1.4604 (1.5352) [2021-04-16 05:48:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][250/1251] eta 0:04:57 lr 0.000488 time 0.2586 (0.2972) loss 2.7311 (3.5526) grad_norm 1.5114 (1.5356) [2021-04-16 05:48:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][260/1251] eta 0:04:54 lr 0.000488 time 0.2921 (0.2968) loss 3.8745 (3.5649) grad_norm 1.5851 (1.5412) [2021-04-16 05:48:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][270/1251] eta 0:04:50 lr 0.000488 time 0.2689 (0.2961) loss 3.6027 (3.5678) grad_norm 1.5121 (1.5419) [2021-04-16 05:48:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][280/1251] eta 0:04:46 lr 0.000488 time 0.2895 (0.2955) loss 4.1521 (3.5661) grad_norm 1.3749 (1.5411) [2021-04-16 05:48:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][290/1251] eta 0:04:43 lr 0.000488 time 0.2774 (0.2952) loss 2.5140 (3.5677) grad_norm 1.6950 (1.5418) [2021-04-16 05:48:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][300/1251] eta 0:04:40 lr 0.000488 time 0.3129 (0.2948) loss 2.4521 (3.5567) grad_norm 1.5963 (1.5387) [2021-04-16 05:48:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][310/1251] eta 0:04:37 lr 0.000488 time 0.2972 (0.2944) loss 2.7812 (3.5429) grad_norm 1.6827 (1.5421) [2021-04-16 05:48:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][320/1251] eta 0:04:33 lr 0.000488 time 0.2793 (0.2939) loss 3.0941 (3.5413) grad_norm 1.5387 (1.5433) [2021-04-16 05:48:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][330/1251] eta 0:04:30 lr 0.000488 time 0.2944 (0.2933) loss 3.4097 (3.5287) grad_norm 1.7159 (1.5426) [2021-04-16 05:48:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][340/1251] eta 0:04:26 lr 0.000488 time 0.2654 (0.2927) loss 2.5240 (3.5279) grad_norm 1.5479 (1.5414) [2021-04-16 05:49:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][350/1251] eta 0:04:23 lr 0.000488 time 0.2690 (0.2926) loss 3.6811 (3.5324) grad_norm 1.4436 (nan) [2021-04-16 05:49:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][360/1251] eta 0:04:20 lr 0.000488 time 0.2717 (0.2920) loss 2.7175 (3.5282) grad_norm 1.6145 (nan) [2021-04-16 05:49:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][370/1251] eta 0:04:16 lr 0.000488 time 0.2574 (0.2916) loss 3.9778 (3.5286) grad_norm 1.5905 (nan) [2021-04-16 05:49:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][380/1251] eta 0:04:13 lr 0.000488 time 0.2670 (0.2911) loss 2.8363 (3.5252) grad_norm 1.4161 (nan) [2021-04-16 05:49:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][390/1251] eta 0:04:10 lr 0.000488 time 0.2568 (0.2912) loss 3.1766 (3.5187) grad_norm 1.7403 (nan) [2021-04-16 05:49:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][400/1251] eta 0:04:07 lr 0.000488 time 0.2825 (0.2908) loss 2.9606 (3.5177) grad_norm 1.7286 (nan) [2021-04-16 05:49:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][410/1251] eta 0:04:04 lr 0.000488 time 0.2669 (0.2903) loss 3.0062 (3.5159) grad_norm 1.6675 (nan) [2021-04-16 05:49:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][420/1251] eta 0:04:00 lr 0.000488 time 0.2904 (0.2900) loss 4.2599 (3.5166) grad_norm 1.8831 (nan) [2021-04-16 05:49:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][430/1251] eta 0:03:57 lr 0.000488 time 0.2857 (0.2899) loss 3.6058 (3.5201) grad_norm 1.6697 (nan) [2021-04-16 05:49:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][440/1251] eta 0:03:54 lr 0.000488 time 0.3036 (0.2897) loss 4.3661 (3.5186) grad_norm 1.5231 (nan) [2021-04-16 05:49:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][450/1251] eta 0:03:51 lr 0.000488 time 0.3115 (0.2896) loss 3.8276 (3.5131) grad_norm 1.4169 (nan) [2021-04-16 05:49:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][460/1251] eta 0:03:49 lr 0.000488 time 0.2875 (0.2895) loss 3.4042 (3.5124) grad_norm 1.5511 (nan) [2021-04-16 05:49:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][470/1251] eta 0:03:45 lr 0.000488 time 0.2425 (0.2892) loss 3.6080 (3.5116) grad_norm 1.7995 (nan) [2021-04-16 05:49:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][480/1251] eta 0:03:42 lr 0.000487 time 0.2673 (0.2892) loss 3.6379 (3.5136) grad_norm 1.4912 (nan) [2021-04-16 05:49:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][490/1251] eta 0:03:40 lr 0.000487 time 0.2739 (0.2891) loss 3.8140 (3.5140) grad_norm 1.4968 (nan) [2021-04-16 05:49:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][500/1251] eta 0:03:36 lr 0.000487 time 0.2683 (0.2888) loss 3.8304 (3.5153) grad_norm 2.1996 (nan) [2021-04-16 05:49:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][510/1251] eta 0:03:33 lr 0.000487 time 0.2772 (0.2886) loss 3.6145 (3.5179) grad_norm 1.6156 (nan) [2021-04-16 05:49:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][520/1251] eta 0:03:30 lr 0.000487 time 0.2910 (0.2885) loss 2.8626 (3.5168) grad_norm 1.4923 (nan) [2021-04-16 05:49:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][530/1251] eta 0:03:27 lr 0.000487 time 0.2712 (0.2882) loss 3.3743 (3.5184) grad_norm 1.5234 (nan) [2021-04-16 05:49:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][540/1251] eta 0:03:25 lr 0.000487 time 0.2930 (0.2884) loss 3.0782 (3.5236) grad_norm 1.3568 (nan) [2021-04-16 05:49:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][550/1251] eta 0:03:22 lr 0.000487 time 0.4615 (0.2887) loss 2.5311 (3.5161) grad_norm 1.5876 (nan) [2021-04-16 05:49:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][560/1251] eta 0:03:19 lr 0.000487 time 0.2636 (0.2884) loss 2.7765 (3.5093) grad_norm 1.8295 (nan) [2021-04-16 05:50:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][570/1251] eta 0:03:16 lr 0.000487 time 0.3052 (0.2886) loss 2.6591 (3.5106) grad_norm 1.5352 (nan) [2021-04-16 05:50:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][580/1251] eta 0:03:13 lr 0.000487 time 0.2456 (0.2886) loss 2.1550 (3.5052) grad_norm 1.6839 (nan) [2021-04-16 05:50:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][590/1251] eta 0:03:10 lr 0.000487 time 0.2749 (0.2884) loss 3.7804 (3.4976) grad_norm 1.3871 (nan) [2021-04-16 05:50:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][600/1251] eta 0:03:07 lr 0.000487 time 0.2763 (0.2881) loss 3.4084 (3.4974) grad_norm 1.7490 (nan) [2021-04-16 05:50:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][610/1251] eta 0:03:04 lr 0.000487 time 0.2748 (0.2880) loss 3.6133 (3.4970) grad_norm 1.4944 (nan) [2021-04-16 05:50:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][620/1251] eta 0:03:01 lr 0.000487 time 0.3033 (0.2879) loss 3.9594 (3.5012) grad_norm 1.4877 (nan) [2021-04-16 05:50:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][630/1251] eta 0:02:58 lr 0.000487 time 0.2673 (0.2876) loss 2.2862 (3.4959) grad_norm 1.7425 (nan) [2021-04-16 05:50:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][640/1251] eta 0:02:55 lr 0.000487 time 0.2780 (0.2875) loss 3.9820 (3.5015) grad_norm 1.6891 (nan) [2021-04-16 05:50:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][650/1251] eta 0:02:52 lr 0.000487 time 0.3413 (0.2875) loss 4.1128 (3.4977) grad_norm 1.5518 (nan) [2021-04-16 05:50:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][660/1251] eta 0:02:49 lr 0.000487 time 0.2983 (0.2873) loss 4.0985 (3.5014) grad_norm 1.6024 (nan) [2021-04-16 05:50:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][670/1251] eta 0:02:46 lr 0.000487 time 0.2947 (0.2872) loss 4.0436 (3.5029) grad_norm 1.5661 (nan) [2021-04-16 05:50:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][680/1251] eta 0:02:43 lr 0.000487 time 0.2746 (0.2871) loss 2.9615 (3.5032) grad_norm 1.7219 (nan) [2021-04-16 05:50:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][690/1251] eta 0:02:40 lr 0.000487 time 0.2673 (0.2869) loss 2.7790 (3.5084) grad_norm 1.3224 (nan) [2021-04-16 05:50:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][700/1251] eta 0:02:38 lr 0.000487 time 0.2770 (0.2868) loss 3.0928 (3.5087) grad_norm 1.3390 (nan) [2021-04-16 05:50:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][710/1251] eta 0:02:35 lr 0.000487 time 0.2517 (0.2867) loss 2.5745 (3.5075) grad_norm 1.5201 (nan) [2021-04-16 05:50:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][720/1251] eta 0:02:32 lr 0.000486 time 0.2815 (0.2868) loss 3.9519 (3.5048) grad_norm 1.5807 (nan) [2021-04-16 05:50:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][730/1251] eta 0:02:29 lr 0.000486 time 0.2748 (0.2866) loss 3.2298 (3.5059) grad_norm 1.5210 (nan) [2021-04-16 05:50:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][740/1251] eta 0:02:26 lr 0.000486 time 0.2701 (0.2868) loss 3.6584 (3.5059) grad_norm 1.6045 (nan) [2021-04-16 05:50:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][750/1251] eta 0:02:23 lr 0.000486 time 0.2802 (0.2867) loss 2.6297 (3.5076) grad_norm 1.3500 (nan) [2021-04-16 05:50:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][760/1251] eta 0:02:20 lr 0.000486 time 0.2736 (0.2865) loss 3.2191 (3.5062) grad_norm 1.3975 (nan) [2021-04-16 05:50:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][770/1251] eta 0:02:17 lr 0.000486 time 0.2875 (0.2865) loss 3.5860 (3.5039) grad_norm 1.5707 (nan) [2021-04-16 05:51:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][780/1251] eta 0:02:14 lr 0.000486 time 0.2716 (0.2865) loss 3.3107 (3.5087) grad_norm 1.4626 (nan) [2021-04-16 05:51:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][790/1251] eta 0:02:12 lr 0.000486 time 0.2647 (0.2864) loss 4.2363 (3.5088) grad_norm 1.5278 (nan) [2021-04-16 05:51:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][800/1251] eta 0:02:09 lr 0.000486 time 0.2864 (0.2863) loss 3.2128 (3.5077) grad_norm 1.4241 (nan) [2021-04-16 05:51:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][810/1251] eta 0:02:06 lr 0.000486 time 0.2821 (0.2862) loss 3.1102 (3.5106) grad_norm 1.5053 (nan) [2021-04-16 05:51:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][820/1251] eta 0:02:03 lr 0.000486 time 0.2691 (0.2861) loss 3.1307 (3.5109) grad_norm 1.5571 (nan) [2021-04-16 05:51:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][830/1251] eta 0:02:00 lr 0.000486 time 0.2509 (0.2860) loss 3.9956 (3.5115) grad_norm 1.5811 (nan) [2021-04-16 05:51:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][840/1251] eta 0:01:57 lr 0.000486 time 0.2719 (0.2859) loss 4.0107 (3.5120) grad_norm 1.9577 (nan) [2021-04-16 05:51:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][850/1251] eta 0:01:54 lr 0.000486 time 0.2998 (0.2858) loss 4.2075 (3.5138) grad_norm 1.4235 (nan) [2021-04-16 05:51:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][860/1251] eta 0:01:51 lr 0.000486 time 0.2722 (0.2857) loss 3.8032 (3.5159) grad_norm 1.4551 (nan) [2021-04-16 05:51:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][870/1251] eta 0:01:48 lr 0.000486 time 0.2730 (0.2857) loss 4.1859 (3.5150) grad_norm 1.7133 (nan) [2021-04-16 05:51:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][880/1251] eta 0:01:46 lr 0.000486 time 0.4273 (0.2858) loss 3.6295 (3.5190) grad_norm 1.4105 (nan) [2021-04-16 05:51:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][890/1251] eta 0:01:43 lr 0.000486 time 0.2759 (0.2857) loss 3.8809 (3.5220) grad_norm 1.7838 (nan) [2021-04-16 05:51:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][900/1251] eta 0:01:40 lr 0.000486 time 0.2744 (0.2856) loss 2.9485 (3.5213) grad_norm 1.6843 (nan) [2021-04-16 05:51:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][910/1251] eta 0:01:37 lr 0.000486 time 0.2662 (0.2855) loss 3.6428 (3.5219) grad_norm 1.4980 (nan) [2021-04-16 05:51:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][920/1251] eta 0:01:34 lr 0.000486 time 0.2662 (0.2854) loss 4.0580 (3.5256) grad_norm 1.5996 (nan) [2021-04-16 05:51:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][930/1251] eta 0:01:31 lr 0.000486 time 0.2706 (0.2855) loss 3.6299 (3.5247) grad_norm 1.3409 (nan) [2021-04-16 05:51:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][940/1251] eta 0:01:28 lr 0.000486 time 0.2715 (0.2855) loss 3.9859 (3.5248) grad_norm 1.5875 (nan) [2021-04-16 05:51:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][950/1251] eta 0:01:25 lr 0.000486 time 0.2592 (0.2854) loss 3.6903 (3.5273) grad_norm 1.4805 (nan) [2021-04-16 05:51:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][960/1251] eta 0:01:23 lr 0.000485 time 0.2753 (0.2853) loss 3.4531 (3.5255) grad_norm 1.4770 (nan) [2021-04-16 05:51:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][970/1251] eta 0:01:20 lr 0.000485 time 0.2890 (0.2854) loss 3.3108 (3.5245) grad_norm 1.4763 (nan) [2021-04-16 05:51:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][980/1251] eta 0:01:17 lr 0.000485 time 0.2759 (0.2853) loss 4.5675 (3.5260) grad_norm 1.3447 (nan) [2021-04-16 05:52:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][990/1251] eta 0:01:14 lr 0.000485 time 0.2633 (0.2852) loss 3.6862 (3.5264) grad_norm 1.5435 (nan) [2021-04-16 05:52:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][1000/1251] eta 0:01:11 lr 0.000485 time 0.2823 (0.2851) loss 4.1315 (3.5295) grad_norm 2.0176 (nan) [2021-04-16 05:52:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][1010/1251] eta 0:01:08 lr 0.000485 time 0.2751 (0.2851) loss 3.2923 (3.5312) grad_norm 1.6571 (nan) [2021-04-16 05:52:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][1020/1251] eta 0:01:05 lr 0.000485 time 0.2928 (0.2850) loss 3.7402 (3.5335) grad_norm 1.7052 (nan) [2021-04-16 05:52:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][1030/1251] eta 0:01:02 lr 0.000485 time 0.2790 (0.2849) loss 3.5788 (3.5332) grad_norm 1.4739 (nan) [2021-04-16 05:52:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][1040/1251] eta 0:01:00 lr 0.000485 time 0.2886 (0.2849) loss 3.9220 (3.5331) grad_norm 1.5121 (nan) [2021-04-16 05:52:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][1050/1251] eta 0:00:57 lr 0.000485 time 0.2641 (0.2848) loss 2.6955 (3.5311) grad_norm 1.5824 (nan) [2021-04-16 05:52:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][1060/1251] eta 0:00:54 lr 0.000485 time 0.2757 (0.2848) loss 2.5668 (3.5296) grad_norm 1.3807 (nan) [2021-04-16 05:52:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][1070/1251] eta 0:00:51 lr 0.000485 time 0.2759 (0.2848) loss 4.3740 (3.5310) grad_norm 1.6475 (nan) [2021-04-16 05:52:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][1080/1251] eta 0:00:48 lr 0.000485 time 0.2505 (0.2847) loss 3.3536 (3.5321) grad_norm 1.3329 (nan) [2021-04-16 05:52:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][1090/1251] eta 0:00:45 lr 0.000485 time 0.2812 (0.2846) loss 3.5676 (3.5322) grad_norm 1.5354 (nan) [2021-04-16 05:52:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][1100/1251] eta 0:00:42 lr 0.000485 time 0.2831 (0.2845) loss 3.8342 (3.5347) grad_norm 1.4088 (nan) [2021-04-16 05:52:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][1110/1251] eta 0:00:40 lr 0.000485 time 0.2595 (0.2845) loss 4.2094 (3.5357) grad_norm 1.6741 (nan) [2021-04-16 05:52:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][1120/1251] eta 0:00:37 lr 0.000485 time 0.2671 (0.2845) loss 3.4312 (3.5368) grad_norm 1.6134 (nan) [2021-04-16 05:52:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][1130/1251] eta 0:00:34 lr 0.000485 time 0.2763 (0.2846) loss 3.2665 (3.5368) grad_norm 1.6068 (nan) [2021-04-16 05:52:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][1140/1251] eta 0:00:31 lr 0.000485 time 0.2788 (0.2847) loss 3.9473 (3.5352) grad_norm 1.3887 (nan) [2021-04-16 05:52:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][1150/1251] eta 0:00:28 lr 0.000485 time 0.2765 (0.2846) loss 4.0558 (3.5366) grad_norm 1.4354 (nan) [2021-04-16 05:52:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][1160/1251] eta 0:00:25 lr 0.000485 time 0.2985 (0.2846) loss 3.4473 (3.5329) grad_norm 1.4962 (nan) [2021-04-16 05:52:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][1170/1251] eta 0:00:23 lr 0.000485 time 0.2659 (0.2845) loss 3.2263 (3.5350) grad_norm 1.5900 (nan) [2021-04-16 05:52:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][1180/1251] eta 0:00:20 lr 0.000485 time 0.2960 (0.2845) loss 3.6451 (3.5351) grad_norm 1.5028 (nan) [2021-04-16 05:52:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][1190/1251] eta 0:00:17 lr 0.000485 time 0.2707 (0.2845) loss 2.9300 (3.5349) grad_norm 1.6428 (nan) [2021-04-16 05:52:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][1200/1251] eta 0:00:14 lr 0.000484 time 0.2983 (0.2845) loss 3.9364 (3.5336) grad_norm 1.3624 (nan) [2021-04-16 05:53:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][1210/1251] eta 0:00:11 lr 0.000484 time 0.2683 (0.2844) loss 4.1154 (3.5356) grad_norm 1.5198 (nan) [2021-04-16 05:53:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][1220/1251] eta 0:00:08 lr 0.000484 time 0.2932 (0.2844) loss 4.2190 (3.5362) grad_norm 1.5350 (nan) [2021-04-16 05:53:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][1230/1251] eta 0:00:05 lr 0.000484 time 0.2852 (0.2845) loss 4.0189 (3.5372) grad_norm 1.5934 (nan) [2021-04-16 05:53:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][1240/1251] eta 0:00:03 lr 0.000484 time 0.2520 (0.2844) loss 2.6589 (3.5367) grad_norm 1.7141 (nan) [2021-04-16 05:53:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [153/300][1250/1251] eta 0:00:00 lr 0.000484 time 0.2496 (0.2841) loss 3.5925 (3.5391) grad_norm 1.4224 (nan) [2021-04-16 05:53:17 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 153 training takes 0:05:59 [2021-04-16 05:53:17 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_153.pth saving...... [2021-04-16 05:53:33 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_153.pth saved !!! [2021-04-16 05:53:34 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.172 (1.172) Loss 1.0035 (1.0035) Acc@1 76.758 (76.758) Acc@5 93.750 (93.750) [2021-04-16 05:53:36 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.316 (0.226) Loss 0.9404 (1.0100) Acc@1 78.418 (76.225) Acc@5 93.848 (93.350) [2021-04-16 05:53:38 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.119 (0.218) Loss 1.0144 (1.0095) Acc@1 75.684 (76.135) Acc@5 93.262 (93.471) [2021-04-16 05:53:40 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.112 (0.226) Loss 1.0034 (1.0135) Acc@1 77.930 (76.099) Acc@5 93.164 (93.331) [2021-04-16 05:53:42 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.213) Loss 0.9924 (1.0102) Acc@1 75.684 (76.081) Acc@5 94.434 (93.452) [2021-04-16 05:53:46 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 76.094 Acc@5 93.410 [2021-04-16 05:53:46 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 76.1% [2021-04-16 05:53:46 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 76.11% [2021-04-16 05:53:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][0/1251] eta 3:45:55 lr 0.000484 time 10.8357 (10.8357) loss 2.2025 (2.2025) grad_norm 1.6611 (1.6611) [2021-04-16 05:54:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][10/1251] eta 0:25:23 lr 0.000484 time 0.2655 (1.2279) loss 3.8193 (3.0767) grad_norm 1.6921 (1.5991) [2021-04-16 05:54:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][20/1251] eta 0:15:54 lr 0.000484 time 0.2922 (0.7757) loss 3.9424 (3.3827) grad_norm 1.8030 (1.5728) [2021-04-16 05:54:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][30/1251] eta 0:12:32 lr 0.000484 time 0.2560 (0.6165) loss 3.0619 (3.4312) grad_norm 1.6559 (1.6186) [2021-04-16 05:54:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][40/1251] eta 0:10:45 lr 0.000484 time 0.2713 (0.5330) loss 3.8842 (3.4826) grad_norm 1.7222 (1.5911) [2021-04-16 05:54:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][50/1251] eta 0:09:40 lr 0.000484 time 0.2923 (0.4834) loss 3.6061 (3.4494) grad_norm 1.5597 (1.5925) [2021-04-16 05:54:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][60/1251] eta 0:08:54 lr 0.000484 time 0.2661 (0.4492) loss 4.3286 (3.4846) grad_norm 1.5739 (1.5863) [2021-04-16 05:54:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][70/1251] eta 0:08:23 lr 0.000484 time 0.2636 (0.4263) loss 2.2312 (3.4495) grad_norm 1.5672 (1.5832) [2021-04-16 05:54:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][80/1251] eta 0:07:59 lr 0.000484 time 0.2719 (0.4093) loss 2.2514 (3.4373) grad_norm 1.5345 (1.5745) [2021-04-16 05:54:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][90/1251] eta 0:07:38 lr 0.000484 time 0.2731 (0.3952) loss 3.5408 (3.4391) grad_norm 1.4797 (1.5768) [2021-04-16 05:54:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][100/1251] eta 0:07:21 lr 0.000484 time 0.3101 (0.3836) loss 3.6658 (3.4388) grad_norm 1.6089 (1.5776) [2021-04-16 05:54:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][110/1251] eta 0:07:06 lr 0.000484 time 0.2728 (0.3736) loss 2.6227 (3.4393) grad_norm 1.3989 (1.5795) [2021-04-16 05:54:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][120/1251] eta 0:06:53 lr 0.000484 time 0.2890 (0.3657) loss 3.7556 (3.4657) grad_norm 1.6140 (1.5746) [2021-04-16 05:54:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][130/1251] eta 0:06:43 lr 0.000484 time 0.2687 (0.3597) loss 4.0646 (3.4801) grad_norm 1.5438 (1.5740) [2021-04-16 05:54:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][140/1251] eta 0:06:33 lr 0.000484 time 0.2762 (0.3539) loss 3.5080 (3.4822) grad_norm 1.5659 (1.5765) [2021-04-16 05:54:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][150/1251] eta 0:06:24 lr 0.000484 time 0.2810 (0.3493) loss 3.7634 (3.4970) grad_norm 2.1240 (1.5756) [2021-04-16 05:54:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][160/1251] eta 0:06:15 lr 0.000484 time 0.2794 (0.3446) loss 4.1103 (3.5027) grad_norm 1.5315 (1.5723) [2021-04-16 05:54:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][170/1251] eta 0:06:07 lr 0.000484 time 0.2557 (0.3404) loss 3.4591 (3.4947) grad_norm 1.3636 (1.5751) [2021-04-16 05:54:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][180/1251] eta 0:06:01 lr 0.000484 time 0.2531 (0.3376) loss 4.0118 (3.4866) grad_norm 1.3122 (1.5724) [2021-04-16 05:54:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][190/1251] eta 0:05:54 lr 0.000483 time 0.2736 (0.3343) loss 4.0077 (3.4981) grad_norm 1.3343 (1.5732) [2021-04-16 05:54:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][200/1251] eta 0:05:48 lr 0.000483 time 0.2823 (0.3318) loss 2.9623 (3.4996) grad_norm 1.6230 (1.5750) [2021-04-16 05:54:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][210/1251] eta 0:05:42 lr 0.000483 time 0.2575 (0.3290) loss 2.9455 (3.5039) grad_norm 1.4603 (1.5754) [2021-04-16 05:54:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][220/1251] eta 0:05:36 lr 0.000483 time 0.2636 (0.3267) loss 3.8915 (3.4944) grad_norm 1.5221 (1.5742) [2021-04-16 05:55:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][230/1251] eta 0:05:31 lr 0.000483 time 0.2749 (0.3245) loss 3.8712 (3.4986) grad_norm 1.7609 (1.5736) [2021-04-16 05:55:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][240/1251] eta 0:05:26 lr 0.000483 time 0.3058 (0.3227) loss 4.2257 (3.5040) grad_norm 1.3269 (1.5763) [2021-04-16 05:55:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][250/1251] eta 0:05:21 lr 0.000483 time 0.2695 (0.3210) loss 3.7790 (3.5078) grad_norm 1.3792 (1.5777) [2021-04-16 05:55:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][260/1251] eta 0:05:16 lr 0.000483 time 0.2842 (0.3192) loss 3.6612 (3.5063) grad_norm 1.9430 (1.5790) [2021-04-16 05:55:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][270/1251] eta 0:05:11 lr 0.000483 time 0.2698 (0.3177) loss 3.0409 (3.5081) grad_norm 1.4414 (1.5757) [2021-04-16 05:55:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][280/1251] eta 0:05:07 lr 0.000483 time 0.3015 (0.3164) loss 3.7154 (3.5022) grad_norm 1.3548 (1.5726) [2021-04-16 05:55:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][290/1251] eta 0:05:03 lr 0.000483 time 0.2867 (0.3155) loss 3.3663 (3.4934) grad_norm 1.5687 (1.5721) [2021-04-16 05:55:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][300/1251] eta 0:04:58 lr 0.000483 time 0.2815 (0.3144) loss 4.1574 (3.4987) grad_norm 1.5906 (1.5702) [2021-04-16 05:55:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][310/1251] eta 0:04:54 lr 0.000483 time 0.2719 (0.3133) loss 3.5580 (3.5035) grad_norm 1.5142 (1.5699) [2021-04-16 05:55:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][320/1251] eta 0:04:50 lr 0.000483 time 0.2680 (0.3126) loss 4.1105 (3.5079) grad_norm 1.5984 (1.5673) [2021-04-16 05:55:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][330/1251] eta 0:04:46 lr 0.000483 time 0.2761 (0.3115) loss 2.4189 (3.5031) grad_norm 1.3804 (1.5655) [2021-04-16 05:55:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][340/1251] eta 0:04:43 lr 0.000483 time 0.2759 (0.3109) loss 2.7440 (3.5002) grad_norm 1.5304 (1.5654) [2021-04-16 05:55:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][350/1251] eta 0:04:39 lr 0.000483 time 0.2891 (0.3101) loss 3.3495 (3.4931) grad_norm 1.2761 (1.5613) [2021-04-16 05:55:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][360/1251] eta 0:04:35 lr 0.000483 time 0.2754 (0.3095) loss 4.2008 (3.4927) grad_norm 1.7537 (1.5628) [2021-04-16 05:55:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][370/1251] eta 0:04:31 lr 0.000483 time 0.2684 (0.3086) loss 4.2584 (3.4971) grad_norm 1.5305 (1.5678) [2021-04-16 05:55:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][380/1251] eta 0:04:28 lr 0.000483 time 0.2911 (0.3078) loss 4.1993 (3.5044) grad_norm 1.6178 (1.5658) [2021-04-16 05:55:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][390/1251] eta 0:04:24 lr 0.000483 time 0.2883 (0.3071) loss 3.4427 (3.5040) grad_norm 1.4257 (1.5649) [2021-04-16 05:55:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][400/1251] eta 0:04:20 lr 0.000483 time 0.2941 (0.3062) loss 3.0646 (3.5092) grad_norm 1.4484 (1.5632) [2021-04-16 05:55:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][410/1251] eta 0:04:16 lr 0.000483 time 0.2896 (0.3055) loss 3.1366 (3.5102) grad_norm 1.5745 (1.5621) [2021-04-16 05:55:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][420/1251] eta 0:04:13 lr 0.000483 time 0.2755 (0.3048) loss 3.4951 (3.5081) grad_norm 1.4308 (1.5612) [2021-04-16 05:55:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][430/1251] eta 0:04:10 lr 0.000482 time 0.2729 (0.3046) loss 3.9533 (3.5130) grad_norm 1.6387 (1.5602) [2021-04-16 05:56:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][440/1251] eta 0:04:06 lr 0.000482 time 0.2790 (0.3040) loss 2.6042 (3.5133) grad_norm 1.4662 (1.5579) [2021-04-16 05:56:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][450/1251] eta 0:04:03 lr 0.000482 time 0.2674 (0.3034) loss 2.7420 (3.5108) grad_norm 1.4021 (1.5563) [2021-04-16 05:56:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][460/1251] eta 0:03:59 lr 0.000482 time 0.3013 (0.3030) loss 3.3772 (3.5101) grad_norm 1.4844 (1.5565) [2021-04-16 05:56:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][470/1251] eta 0:03:56 lr 0.000482 time 0.2636 (0.3023) loss 3.1375 (3.5094) grad_norm 1.6788 (1.5553) [2021-04-16 05:56:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][480/1251] eta 0:03:52 lr 0.000482 time 0.2842 (0.3018) loss 4.2247 (3.5131) grad_norm 1.6012 (1.5559) [2021-04-16 05:56:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][490/1251] eta 0:03:49 lr 0.000482 time 0.2848 (0.3014) loss 3.3677 (3.5113) grad_norm 1.6657 (1.5558) [2021-04-16 05:56:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][500/1251] eta 0:03:45 lr 0.000482 time 0.2706 (0.3008) loss 3.2093 (3.5127) grad_norm 1.8619 (1.5557) [2021-04-16 05:56:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][510/1251] eta 0:03:42 lr 0.000482 time 0.2459 (0.3003) loss 2.5967 (3.5030) grad_norm 1.5079 (1.5537) [2021-04-16 05:56:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][520/1251] eta 0:03:39 lr 0.000482 time 0.3225 (0.2999) loss 4.4180 (3.5061) grad_norm 1.5540 (1.5538) [2021-04-16 05:56:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][530/1251] eta 0:03:35 lr 0.000482 time 0.2661 (0.2995) loss 3.8403 (3.5089) grad_norm 1.3612 (1.5529) [2021-04-16 05:56:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][540/1251] eta 0:03:32 lr 0.000482 time 0.2889 (0.2991) loss 3.2985 (3.5124) grad_norm 1.6824 (1.5520) [2021-04-16 05:56:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][550/1251] eta 0:03:29 lr 0.000482 time 0.2878 (0.2990) loss 3.5305 (3.5102) grad_norm 1.3620 (1.5517) [2021-04-16 05:56:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][560/1251] eta 0:03:26 lr 0.000482 time 0.2803 (0.2987) loss 3.6774 (3.5109) grad_norm 1.6977 (1.5512) [2021-04-16 05:56:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][570/1251] eta 0:03:23 lr 0.000482 time 0.2719 (0.2987) loss 3.8245 (3.5096) grad_norm 1.4557 (1.5507) [2021-04-16 05:56:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][580/1251] eta 0:03:20 lr 0.000482 time 0.2851 (0.2986) loss 3.9629 (3.5107) grad_norm 1.5370 (1.5509) [2021-04-16 05:56:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][590/1251] eta 0:03:17 lr 0.000482 time 0.2710 (0.2982) loss 3.8290 (3.5100) grad_norm 1.5988 (1.5498) [2021-04-16 05:56:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][600/1251] eta 0:03:13 lr 0.000482 time 0.2782 (0.2979) loss 3.9561 (3.5135) grad_norm 1.5884 (1.5503) [2021-04-16 05:56:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][610/1251] eta 0:03:10 lr 0.000482 time 0.2463 (0.2975) loss 3.9260 (3.5136) grad_norm 1.7335 (1.5509) [2021-04-16 05:56:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][620/1251] eta 0:03:07 lr 0.000482 time 0.2993 (0.2973) loss 3.7764 (3.5121) grad_norm 1.4233 (1.5513) [2021-04-16 05:56:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][630/1251] eta 0:03:04 lr 0.000482 time 0.2861 (0.2970) loss 3.9908 (3.5073) grad_norm 1.6163 (1.5520) [2021-04-16 05:56:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][640/1251] eta 0:03:01 lr 0.000482 time 0.2611 (0.2967) loss 3.4618 (3.5097) grad_norm 1.5904 (1.5522) [2021-04-16 05:57:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][650/1251] eta 0:02:58 lr 0.000482 time 0.2903 (0.2966) loss 2.4411 (3.5103) grad_norm 1.5183 (1.5530) [2021-04-16 05:57:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][660/1251] eta 0:02:55 lr 0.000482 time 0.3052 (0.2963) loss 2.3548 (3.5118) grad_norm 1.5538 (1.5561) [2021-04-16 05:57:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][670/1251] eta 0:02:51 lr 0.000481 time 0.2704 (0.2959) loss 3.6235 (3.5103) grad_norm 1.4344 (1.5558) [2021-04-16 05:57:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][680/1251] eta 0:02:48 lr 0.000481 time 0.2812 (0.2957) loss 2.7798 (3.5122) grad_norm 1.3440 (1.5578) [2021-04-16 05:57:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][690/1251] eta 0:02:45 lr 0.000481 time 0.2932 (0.2955) loss 2.7129 (3.5117) grad_norm 1.4514 (1.5581) [2021-04-16 05:57:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][700/1251] eta 0:02:42 lr 0.000481 time 0.3048 (0.2952) loss 4.1167 (3.5141) grad_norm 1.7045 (1.5583) [2021-04-16 05:57:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][710/1251] eta 0:02:39 lr 0.000481 time 0.2682 (0.2950) loss 3.2078 (3.5123) grad_norm 1.4066 (1.5590) [2021-04-16 05:57:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][720/1251] eta 0:02:36 lr 0.000481 time 0.2784 (0.2948) loss 3.4677 (3.5078) grad_norm 1.7393 (1.5581) [2021-04-16 05:57:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][730/1251] eta 0:02:33 lr 0.000481 time 0.2748 (0.2946) loss 3.8329 (3.5104) grad_norm 1.6510 (1.5587) [2021-04-16 05:57:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][740/1251] eta 0:02:30 lr 0.000481 time 0.2657 (0.2943) loss 2.4554 (3.5064) grad_norm 1.5817 (1.5601) [2021-04-16 05:57:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][750/1251] eta 0:02:27 lr 0.000481 time 0.3046 (0.2941) loss 3.9507 (3.5063) grad_norm 1.5443 (1.5601) [2021-04-16 05:57:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][760/1251] eta 0:02:24 lr 0.000481 time 0.2743 (0.2940) loss 3.7653 (3.5049) grad_norm 1.5120 (1.5603) [2021-04-16 05:57:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][770/1251] eta 0:02:21 lr 0.000481 time 0.2596 (0.2937) loss 3.5674 (3.5049) grad_norm 1.5527 (1.5601) [2021-04-16 05:57:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][780/1251] eta 0:02:18 lr 0.000481 time 0.2939 (0.2935) loss 3.9549 (3.5058) grad_norm 1.3295 (1.5614) [2021-04-16 05:57:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][790/1251] eta 0:02:15 lr 0.000481 time 0.2814 (0.2933) loss 2.5434 (3.5006) grad_norm 1.5215 (1.5616) [2021-04-16 05:57:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][800/1251] eta 0:02:12 lr 0.000481 time 0.2664 (0.2930) loss 3.5340 (3.5014) grad_norm 1.4753 (1.5612) [2021-04-16 05:57:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][810/1251] eta 0:02:09 lr 0.000481 time 0.2496 (0.2929) loss 3.7521 (3.5033) grad_norm 1.7301 (1.5617) [2021-04-16 05:57:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][820/1251] eta 0:02:06 lr 0.000481 time 0.2960 (0.2927) loss 2.7508 (3.5032) grad_norm 1.5788 (1.5619) [2021-04-16 05:57:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][830/1251] eta 0:02:03 lr 0.000481 time 0.2794 (0.2924) loss 3.6816 (3.5011) grad_norm 1.5681 (1.5619) [2021-04-16 05:57:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][840/1251] eta 0:02:00 lr 0.000481 time 0.2957 (0.2924) loss 3.3877 (3.4987) grad_norm 1.9841 (1.5621) [2021-04-16 05:57:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][850/1251] eta 0:01:57 lr 0.000481 time 0.2666 (0.2921) loss 4.1421 (3.4987) grad_norm 1.6454 (1.5630) [2021-04-16 05:57:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][860/1251] eta 0:01:54 lr 0.000481 time 0.2788 (0.2920) loss 3.1725 (3.4957) grad_norm 1.4148 (1.5621) [2021-04-16 05:58:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][870/1251] eta 0:01:51 lr 0.000481 time 0.2562 (0.2918) loss 2.2865 (3.4980) grad_norm 1.3467 (1.5618) [2021-04-16 05:58:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][880/1251] eta 0:01:48 lr 0.000481 time 0.2973 (0.2917) loss 3.5495 (3.4978) grad_norm 1.6209 (1.5617) [2021-04-16 05:58:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][890/1251] eta 0:01:45 lr 0.000481 time 0.2741 (0.2917) loss 3.7170 (3.4985) grad_norm 1.5249 (1.5617) [2021-04-16 05:58:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][900/1251] eta 0:01:42 lr 0.000481 time 0.2718 (0.2916) loss 4.1276 (3.4999) grad_norm 1.6996 (1.5614) [2021-04-16 05:58:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][910/1251] eta 0:01:39 lr 0.000481 time 0.2752 (0.2914) loss 3.2830 (3.5023) grad_norm 1.3417 (1.5608) [2021-04-16 05:58:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][920/1251] eta 0:01:36 lr 0.000480 time 0.2678 (0.2914) loss 2.4936 (3.5005) grad_norm 1.4297 (1.5602) [2021-04-16 05:58:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][930/1251] eta 0:01:33 lr 0.000480 time 0.3761 (0.2913) loss 3.3875 (3.5003) grad_norm 1.5852 (1.5599) [2021-04-16 05:58:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][940/1251] eta 0:01:30 lr 0.000480 time 0.2587 (0.2912) loss 3.8657 (3.4992) grad_norm 1.3265 (1.5595) [2021-04-16 05:58:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][950/1251] eta 0:01:27 lr 0.000480 time 0.2721 (0.2910) loss 4.3210 (3.5013) grad_norm 1.3245 (1.5587) [2021-04-16 05:58:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][960/1251] eta 0:01:24 lr 0.000480 time 0.2648 (0.2908) loss 2.9816 (3.5033) grad_norm 1.5642 (1.5585) [2021-04-16 05:58:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][970/1251] eta 0:01:21 lr 0.000480 time 0.2899 (0.2908) loss 3.8204 (3.5008) grad_norm 1.7963 (1.5583) [2021-04-16 05:58:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][980/1251] eta 0:01:18 lr 0.000480 time 0.2631 (0.2906) loss 4.1461 (3.5015) grad_norm 1.7019 (1.5588) [2021-04-16 05:58:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][990/1251] eta 0:01:15 lr 0.000480 time 0.2769 (0.2904) loss 2.7862 (3.4992) grad_norm 1.5230 (1.5587) [2021-04-16 05:58:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][1000/1251] eta 0:01:12 lr 0.000480 time 0.2777 (0.2903) loss 3.9065 (3.5017) grad_norm 2.0705 (1.5595) [2021-04-16 05:58:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][1010/1251] eta 0:01:09 lr 0.000480 time 0.2859 (0.2902) loss 2.2367 (3.5001) grad_norm 1.4200 (1.5587) [2021-04-16 05:58:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][1020/1251] eta 0:01:06 lr 0.000480 time 0.2668 (0.2900) loss 2.7054 (3.4999) grad_norm 1.3239 (1.5578) [2021-04-16 05:58:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][1030/1251] eta 0:01:04 lr 0.000480 time 0.2677 (0.2899) loss 3.4485 (3.5013) grad_norm 1.3650 (1.5583) [2021-04-16 05:58:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][1040/1251] eta 0:01:01 lr 0.000480 time 0.2828 (0.2898) loss 3.9613 (3.5035) grad_norm 1.4916 (1.5589) [2021-04-16 05:58:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][1050/1251] eta 0:00:58 lr 0.000480 time 0.2580 (0.2897) loss 3.2729 (3.4996) grad_norm 1.4844 (1.5591) [2021-04-16 05:58:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][1060/1251] eta 0:00:55 lr 0.000480 time 0.2619 (0.2895) loss 3.0164 (3.5008) grad_norm 1.9986 (1.5588) [2021-04-16 05:58:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][1070/1251] eta 0:00:52 lr 0.000480 time 0.2763 (0.2894) loss 3.5265 (3.5002) grad_norm 1.4679 (1.5586) [2021-04-16 05:58:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][1080/1251] eta 0:00:49 lr 0.000480 time 0.2660 (0.2893) loss 4.2186 (3.5020) grad_norm 1.4337 (1.5581) [2021-04-16 05:59:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][1090/1251] eta 0:00:46 lr 0.000480 time 0.3043 (0.2892) loss 2.5609 (3.5006) grad_norm 1.5012 (1.5582) [2021-04-16 05:59:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][1100/1251] eta 0:00:43 lr 0.000480 time 0.2758 (0.2891) loss 4.0334 (3.5021) grad_norm 1.5212 (1.5582) [2021-04-16 05:59:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][1110/1251] eta 0:00:40 lr 0.000480 time 0.2580 (0.2890) loss 3.3584 (3.5019) grad_norm 1.3544 (1.5578) [2021-04-16 05:59:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][1120/1251] eta 0:00:37 lr 0.000480 time 0.2488 (0.2889) loss 2.7018 (3.5016) grad_norm 1.3325 (1.5579) [2021-04-16 05:59:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][1130/1251] eta 0:00:34 lr 0.000480 time 0.2682 (0.2889) loss 3.8358 (3.5016) grad_norm 1.4207 (1.5581) [2021-04-16 05:59:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][1140/1251] eta 0:00:32 lr 0.000480 time 0.2663 (0.2890) loss 4.2498 (3.5002) grad_norm 1.6031 (1.5580) [2021-04-16 05:59:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][1150/1251] eta 0:00:29 lr 0.000480 time 0.2680 (0.2889) loss 3.5051 (3.5029) grad_norm 1.7803 (1.5594) [2021-04-16 05:59:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][1160/1251] eta 0:00:26 lr 0.000479 time 0.2799 (0.2888) loss 3.7465 (3.5040) grad_norm 1.4081 (1.5590) [2021-04-16 05:59:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][1170/1251] eta 0:00:23 lr 0.000479 time 0.2729 (0.2887) loss 4.2104 (3.5041) grad_norm 1.4291 (1.5590) [2021-04-16 05:59:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][1180/1251] eta 0:00:20 lr 0.000479 time 0.2915 (0.2886) loss 3.9778 (3.5053) grad_norm 1.6191 (1.5599) [2021-04-16 05:59:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][1190/1251] eta 0:00:17 lr 0.000479 time 0.2829 (0.2886) loss 2.7079 (3.5030) grad_norm 1.5881 (1.5596) [2021-04-16 05:59:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][1200/1251] eta 0:00:14 lr 0.000479 time 0.2724 (0.2885) loss 4.1373 (3.5052) grad_norm 1.4975 (1.5596) [2021-04-16 05:59:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][1210/1251] eta 0:00:11 lr 0.000479 time 0.2789 (0.2884) loss 3.2951 (3.5067) grad_norm 1.4513 (1.5591) [2021-04-16 05:59:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][1220/1251] eta 0:00:08 lr 0.000479 time 0.2653 (0.2884) loss 3.6949 (3.5074) grad_norm 1.6094 (1.5597) [2021-04-16 05:59:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][1230/1251] eta 0:00:06 lr 0.000479 time 0.2673 (0.2883) loss 3.3515 (3.5090) grad_norm 1.9856 (1.5612) [2021-04-16 05:59:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][1240/1251] eta 0:00:03 lr 0.000479 time 0.2482 (0.2881) loss 3.0904 (3.5097) grad_norm 1.3418 (1.5614) [2021-04-16 05:59:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [154/300][1250/1251] eta 0:00:00 lr 0.000479 time 0.2480 (0.2878) loss 4.2684 (3.5108) grad_norm 1.6981 (1.5620) [2021-04-16 05:59:50 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 154 training takes 0:06:03 [2021-04-16 05:59:50 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_154.pth saving...... [2021-04-16 06:00:03 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_154.pth saved !!! [2021-04-16 06:00:05 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.179 (1.179) Loss 1.0973 (1.0973) Acc@1 72.656 (72.656) Acc@5 92.578 (92.578) [2021-04-16 06:00:06 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.148 (0.280) Loss 1.0003 (0.9992) Acc@1 76.562 (76.225) Acc@5 93.164 (93.750) [2021-04-16 06:00:08 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.097 (0.239) Loss 0.9693 (0.9845) Acc@1 76.660 (76.502) Acc@5 94.238 (93.834) [2021-04-16 06:00:11 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.100 (0.236) Loss 0.9633 (0.9826) Acc@1 76.660 (76.421) Acc@5 92.871 (93.772) [2021-04-16 06:00:12 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.201 (0.216) Loss 0.9519 (0.9896) Acc@1 77.539 (76.236) Acc@5 93.750 (93.574) [2021-04-16 06:00:18 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 76.112 Acc@5 93.500 [2021-04-16 06:00:18 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 76.1% [2021-04-16 06:00:18 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 76.11% [2021-04-16 06:00:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][0/1251] eta 1:29:33 lr 0.000479 time 4.2950 (4.2950) loss 4.2829 (4.2829) grad_norm 1.5432 (1.5432) [2021-04-16 06:00:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][10/1251] eta 0:13:22 lr 0.000479 time 0.2710 (0.6463) loss 3.9302 (3.6374) grad_norm 1.6056 (1.5538) [2021-04-16 06:00:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][20/1251] eta 0:09:45 lr 0.000479 time 0.2780 (0.4752) loss 3.0329 (3.4531) grad_norm 1.4013 (1.5351) [2021-04-16 06:00:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][30/1251] eta 0:08:24 lr 0.000479 time 0.2970 (0.4136) loss 3.7393 (3.5128) grad_norm 1.4022 (1.5025) [2021-04-16 06:00:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][40/1251] eta 0:07:41 lr 0.000479 time 0.2803 (0.3814) loss 4.3165 (3.4876) grad_norm 1.7950 (1.5338) [2021-04-16 06:00:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][50/1251] eta 0:07:14 lr 0.000479 time 0.2951 (0.3619) loss 4.2626 (3.4910) grad_norm 1.7460 (1.5381) [2021-04-16 06:00:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][60/1251] eta 0:06:56 lr 0.000479 time 0.3269 (0.3495) loss 4.2323 (3.5338) grad_norm 1.6885 (1.5301) [2021-04-16 06:00:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][70/1251] eta 0:06:41 lr 0.000479 time 0.3073 (0.3402) loss 4.0524 (3.5129) grad_norm 1.5905 (1.5401) [2021-04-16 06:00:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][80/1251] eta 0:06:31 lr 0.000479 time 0.2969 (0.3343) loss 3.8950 (3.4815) grad_norm 1.4949 (1.5547) [2021-04-16 06:00:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][90/1251] eta 0:06:21 lr 0.000479 time 0.2870 (0.3287) loss 3.6363 (3.4893) grad_norm 1.2793 (1.5421) [2021-04-16 06:00:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][100/1251] eta 0:06:14 lr 0.000479 time 0.2692 (0.3253) loss 3.0569 (3.5098) grad_norm 1.3452 (1.5423) [2021-04-16 06:00:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][110/1251] eta 0:06:06 lr 0.000479 time 0.2930 (0.3214) loss 3.5832 (3.5312) grad_norm 1.7224 (1.5392) [2021-04-16 06:00:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][120/1251] eta 0:06:01 lr 0.000479 time 0.2969 (0.3192) loss 2.5186 (3.5252) grad_norm 1.5560 (1.5347) [2021-04-16 06:01:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][130/1251] eta 0:05:55 lr 0.000479 time 0.2827 (0.3170) loss 3.6906 (3.4993) grad_norm 1.8336 (1.5390) [2021-04-16 06:01:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][140/1251] eta 0:05:50 lr 0.000479 time 0.4253 (0.3154) loss 3.1572 (3.4862) grad_norm 1.8146 (1.5453) [2021-04-16 06:01:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][150/1251] eta 0:05:45 lr 0.000478 time 0.2957 (0.3138) loss 4.1349 (3.4767) grad_norm 1.5534 (1.5455) [2021-04-16 06:01:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][160/1251] eta 0:05:40 lr 0.000478 time 0.2944 (0.3118) loss 4.1615 (3.4861) grad_norm 1.6021 (1.5475) [2021-04-16 06:01:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][170/1251] eta 0:05:35 lr 0.000478 time 0.2464 (0.3100) loss 3.3920 (3.4891) grad_norm 1.5286 (1.5483) [2021-04-16 06:01:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][180/1251] eta 0:05:30 lr 0.000478 time 0.2878 (0.3087) loss 3.7930 (3.4947) grad_norm 1.8620 (1.5565) [2021-04-16 06:01:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][190/1251] eta 0:05:25 lr 0.000478 time 0.2657 (0.3071) loss 3.7655 (3.4856) grad_norm 1.4919 (1.5578) [2021-04-16 06:01:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][200/1251] eta 0:05:20 lr 0.000478 time 0.2905 (0.3054) loss 3.6312 (3.5012) grad_norm 1.6498 (1.5576) [2021-04-16 06:01:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][210/1251] eta 0:05:16 lr 0.000478 time 0.2706 (0.3043) loss 3.0155 (3.4939) grad_norm 1.5149 (1.5556) [2021-04-16 06:01:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][220/1251] eta 0:05:13 lr 0.000478 time 0.2592 (0.3039) loss 3.6720 (3.5059) grad_norm 1.5804 (1.5524) [2021-04-16 06:01:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][230/1251] eta 0:05:09 lr 0.000478 time 0.2834 (0.3029) loss 3.3304 (3.5002) grad_norm 1.5222 (1.5529) [2021-04-16 06:01:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][240/1251] eta 0:05:05 lr 0.000478 time 0.2930 (0.3020) loss 2.9474 (3.4895) grad_norm 1.4901 (1.5507) [2021-04-16 06:01:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][250/1251] eta 0:05:01 lr 0.000478 time 0.2726 (0.3008) loss 3.1463 (3.4870) grad_norm 1.6880 (1.5520) [2021-04-16 06:01:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][260/1251] eta 0:04:57 lr 0.000478 time 0.2941 (0.3000) loss 4.0261 (3.4847) grad_norm 1.2918 (1.5500) [2021-04-16 06:01:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][270/1251] eta 0:04:53 lr 0.000478 time 0.2700 (0.2993) loss 2.8705 (3.4898) grad_norm 1.4965 (1.5475) [2021-04-16 06:01:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][280/1251] eta 0:04:49 lr 0.000478 time 0.2869 (0.2985) loss 4.3741 (3.4967) grad_norm 1.6008 (1.5476) [2021-04-16 06:01:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][290/1251] eta 0:04:45 lr 0.000478 time 0.2689 (0.2975) loss 3.6913 (3.4951) grad_norm 1.5320 (1.5489) [2021-04-16 06:01:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][300/1251] eta 0:04:42 lr 0.000478 time 0.2738 (0.2969) loss 4.1367 (3.4995) grad_norm 1.4601 (1.5468) [2021-04-16 06:01:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][310/1251] eta 0:04:38 lr 0.000478 time 0.2641 (0.2963) loss 2.7961 (3.4963) grad_norm 1.4363 (1.5426) [2021-04-16 06:01:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][320/1251] eta 0:04:35 lr 0.000478 time 0.2655 (0.2955) loss 4.1750 (3.5042) grad_norm 2.0452 (1.5422) [2021-04-16 06:01:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][330/1251] eta 0:04:32 lr 0.000478 time 0.2980 (0.2956) loss 2.1773 (3.5003) grad_norm 1.8740 (1.5460) [2021-04-16 06:01:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][340/1251] eta 0:04:29 lr 0.000478 time 0.3053 (0.2955) loss 3.5132 (3.4944) grad_norm 1.7188 (1.5467) [2021-04-16 06:02:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][350/1251] eta 0:04:25 lr 0.000478 time 0.2610 (0.2952) loss 3.6608 (3.4975) grad_norm 1.6686 (1.5472) [2021-04-16 06:02:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][360/1251] eta 0:04:23 lr 0.000478 time 0.3174 (0.2953) loss 2.9394 (3.5009) grad_norm 1.9591 (1.5492) [2021-04-16 06:02:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][370/1251] eta 0:04:20 lr 0.000478 time 0.3211 (0.2952) loss 3.7403 (3.5025) grad_norm 2.1590 (1.5514) [2021-04-16 06:02:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][380/1251] eta 0:04:16 lr 0.000478 time 0.2905 (0.2950) loss 3.7457 (3.5021) grad_norm 1.4307 (1.5515) [2021-04-16 06:02:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][390/1251] eta 0:04:13 lr 0.000477 time 0.2909 (0.2947) loss 3.7676 (3.5050) grad_norm 1.4177 (1.5507) [2021-04-16 06:02:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][400/1251] eta 0:04:10 lr 0.000477 time 0.2840 (0.2947) loss 3.2325 (3.5022) grad_norm 1.6637 (1.5541) [2021-04-16 06:02:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][410/1251] eta 0:04:07 lr 0.000477 time 0.2812 (0.2942) loss 2.5191 (3.5026) grad_norm 1.4297 (1.5554) [2021-04-16 06:02:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][420/1251] eta 0:04:04 lr 0.000477 time 0.2722 (0.2938) loss 3.3106 (3.5075) grad_norm 1.8150 (1.5578) [2021-04-16 06:02:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][430/1251] eta 0:04:01 lr 0.000477 time 0.2871 (0.2939) loss 3.6680 (3.5074) grad_norm 1.9046 (1.5584) [2021-04-16 06:02:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][440/1251] eta 0:03:57 lr 0.000477 time 0.2653 (0.2935) loss 3.1693 (3.5064) grad_norm 1.6519 (1.5588) [2021-04-16 06:02:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][450/1251] eta 0:03:54 lr 0.000477 time 0.2687 (0.2930) loss 3.1057 (3.5042) grad_norm 1.3724 (1.5583) [2021-04-16 06:02:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][460/1251] eta 0:03:51 lr 0.000477 time 0.2644 (0.2926) loss 4.0198 (3.5016) grad_norm 1.7253 (1.5593) [2021-04-16 06:02:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][470/1251] eta 0:03:48 lr 0.000477 time 0.3013 (0.2924) loss 4.2307 (3.5049) grad_norm 1.4650 (1.5607) [2021-04-16 06:02:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][480/1251] eta 0:03:45 lr 0.000477 time 0.2748 (0.2921) loss 4.3334 (3.5095) grad_norm 1.6617 (1.5607) [2021-04-16 06:02:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][490/1251] eta 0:03:42 lr 0.000477 time 0.2892 (0.2918) loss 3.8273 (3.5133) grad_norm 1.3053 (1.5602) [2021-04-16 06:02:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][500/1251] eta 0:03:38 lr 0.000477 time 0.2762 (0.2915) loss 3.3033 (3.5137) grad_norm 1.6517 (1.5594) [2021-04-16 06:02:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][510/1251] eta 0:03:35 lr 0.000477 time 0.2691 (0.2913) loss 3.3585 (3.5153) grad_norm 1.5087 (1.5584) [2021-04-16 06:02:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][520/1251] eta 0:03:32 lr 0.000477 time 0.2841 (0.2913) loss 3.0347 (3.5154) grad_norm 1.4020 (1.5568) [2021-04-16 06:02:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][530/1251] eta 0:03:29 lr 0.000477 time 0.2739 (0.2911) loss 3.8565 (3.5168) grad_norm 1.5916 (1.5571) [2021-04-16 06:02:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][540/1251] eta 0:03:26 lr 0.000477 time 0.2891 (0.2908) loss 2.5259 (3.5146) grad_norm 1.8522 (1.5580) [2021-04-16 06:02:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][550/1251] eta 0:03:23 lr 0.000477 time 0.2629 (0.2909) loss 4.2151 (3.5172) grad_norm 1.8038 (1.5584) [2021-04-16 06:03:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][560/1251] eta 0:03:20 lr 0.000477 time 0.2730 (0.2907) loss 3.1878 (3.5137) grad_norm 1.6751 (1.5586) [2021-04-16 06:03:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][570/1251] eta 0:03:17 lr 0.000477 time 0.2564 (0.2904) loss 2.9081 (3.5122) grad_norm 1.5727 (1.5583) [2021-04-16 06:03:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][580/1251] eta 0:03:14 lr 0.000477 time 0.2881 (0.2902) loss 3.5574 (3.5149) grad_norm 1.5615 (1.5600) [2021-04-16 06:03:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][590/1251] eta 0:03:11 lr 0.000477 time 0.2726 (0.2900) loss 4.2005 (3.5183) grad_norm 1.9817 (1.5626) [2021-04-16 06:03:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][600/1251] eta 0:03:08 lr 0.000477 time 0.2714 (0.2897) loss 2.7327 (3.5141) grad_norm 1.6643 (1.5623) [2021-04-16 06:03:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][610/1251] eta 0:03:05 lr 0.000477 time 0.2910 (0.2898) loss 3.6928 (3.5144) grad_norm 1.3221 (1.5607) [2021-04-16 06:03:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][620/1251] eta 0:03:02 lr 0.000477 time 0.2998 (0.2896) loss 4.2994 (3.5124) grad_norm 1.5677 (1.5610) [2021-04-16 06:03:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][630/1251] eta 0:02:59 lr 0.000476 time 0.2630 (0.2895) loss 2.6281 (3.5127) grad_norm 1.4298 (1.5617) [2021-04-16 06:03:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][640/1251] eta 0:02:56 lr 0.000476 time 0.2709 (0.2894) loss 3.9810 (3.5157) grad_norm 1.5165 (1.5618) [2021-04-16 06:03:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][650/1251] eta 0:02:54 lr 0.000476 time 0.2864 (0.2895) loss 3.7134 (3.5187) grad_norm 1.4121 (1.5611) [2021-04-16 06:03:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][660/1251] eta 0:02:50 lr 0.000476 time 0.3047 (0.2893) loss 3.9481 (3.5225) grad_norm 1.4355 (1.5615) [2021-04-16 06:03:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][670/1251] eta 0:02:48 lr 0.000476 time 0.2463 (0.2893) loss 3.2803 (3.5239) grad_norm 1.5223 (1.5613) [2021-04-16 06:03:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][680/1251] eta 0:02:45 lr 0.000476 time 0.2871 (0.2894) loss 3.6229 (3.5252) grad_norm 1.3919 (1.5614) [2021-04-16 06:03:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][690/1251] eta 0:02:42 lr 0.000476 time 0.2718 (0.2894) loss 3.5430 (3.5209) grad_norm 1.6872 (1.5607) [2021-04-16 06:03:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][700/1251] eta 0:02:39 lr 0.000476 time 0.2752 (0.2892) loss 3.6871 (3.5192) grad_norm 1.3966 (1.5587) [2021-04-16 06:03:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][710/1251] eta 0:02:36 lr 0.000476 time 0.2523 (0.2892) loss 3.2892 (3.5188) grad_norm 1.4912 (1.5585) [2021-04-16 06:03:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][720/1251] eta 0:02:33 lr 0.000476 time 0.2661 (0.2891) loss 3.6812 (3.5183) grad_norm 1.3650 (1.5584) [2021-04-16 06:03:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][730/1251] eta 0:02:30 lr 0.000476 time 0.2496 (0.2889) loss 3.7677 (3.5203) grad_norm 1.4327 (1.5575) [2021-04-16 06:03:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][740/1251] eta 0:02:27 lr 0.000476 time 0.2738 (0.2889) loss 2.8776 (3.5205) grad_norm 1.4742 (1.5580) [2021-04-16 06:03:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][750/1251] eta 0:02:24 lr 0.000476 time 0.2755 (0.2888) loss 4.1820 (3.5173) grad_norm 1.5158 (1.5590) [2021-04-16 06:03:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][760/1251] eta 0:02:21 lr 0.000476 time 0.2805 (0.2887) loss 3.8771 (3.5170) grad_norm 1.4817 (1.5581) [2021-04-16 06:04:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][770/1251] eta 0:02:18 lr 0.000476 time 0.2974 (0.2885) loss 2.8842 (3.5150) grad_norm 1.5175 (1.5577) [2021-04-16 06:04:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][780/1251] eta 0:02:15 lr 0.000476 time 0.2973 (0.2884) loss 3.1303 (3.5149) grad_norm 1.4020 (1.5571) [2021-04-16 06:04:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][790/1251] eta 0:02:13 lr 0.000476 time 0.2926 (0.2887) loss 2.8544 (3.5107) grad_norm 1.3714 (1.5562) [2021-04-16 06:04:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][800/1251] eta 0:02:10 lr 0.000476 time 0.3043 (0.2885) loss 3.8720 (3.5117) grad_norm 1.5305 (1.5562) [2021-04-16 06:04:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][810/1251] eta 0:02:07 lr 0.000476 time 0.2855 (0.2885) loss 3.6192 (3.5158) grad_norm 1.6260 (1.5565) [2021-04-16 06:04:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][820/1251] eta 0:02:04 lr 0.000476 time 0.2936 (0.2884) loss 3.9562 (3.5158) grad_norm 1.4898 (1.5567) [2021-04-16 06:04:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][830/1251] eta 0:02:01 lr 0.000476 time 0.2842 (0.2883) loss 3.1993 (3.5174) grad_norm 1.8199 (1.5564) [2021-04-16 06:04:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][840/1251] eta 0:01:58 lr 0.000476 time 0.2961 (0.2882) loss 2.9675 (3.5185) grad_norm 1.5488 (1.5571) [2021-04-16 06:04:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][850/1251] eta 0:01:55 lr 0.000476 time 0.3095 (0.2882) loss 3.4490 (3.5162) grad_norm 1.6517 (1.5565) [2021-04-16 06:04:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][860/1251] eta 0:01:52 lr 0.000476 time 0.3048 (0.2880) loss 3.6977 (3.5167) grad_norm 1.7367 (1.5560) [2021-04-16 06:04:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][870/1251] eta 0:01:49 lr 0.000475 time 0.2681 (0.2880) loss 4.1552 (3.5148) grad_norm 1.4526 (1.5558) [2021-04-16 06:04:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][880/1251] eta 0:01:46 lr 0.000475 time 0.2614 (0.2879) loss 3.7169 (3.5172) grad_norm 1.6632 (1.5551) [2021-04-16 06:04:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][890/1251] eta 0:01:43 lr 0.000475 time 0.2799 (0.2877) loss 3.7803 (3.5188) grad_norm 1.9959 (1.5565) [2021-04-16 06:04:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][900/1251] eta 0:01:40 lr 0.000475 time 0.2761 (0.2877) loss 3.0414 (3.5191) grad_norm 1.5913 (1.5565) [2021-04-16 06:04:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][910/1251] eta 0:01:38 lr 0.000475 time 0.2885 (0.2878) loss 3.3473 (3.5192) grad_norm 1.5853 (1.5567) [2021-04-16 06:04:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][920/1251] eta 0:01:35 lr 0.000475 time 0.2980 (0.2879) loss 2.9025 (3.5156) grad_norm 1.3786 (1.5566) [2021-04-16 06:04:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][930/1251] eta 0:01:32 lr 0.000475 time 0.2803 (0.2879) loss 3.4163 (3.5132) grad_norm 1.6644 (1.5558) [2021-04-16 06:04:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][940/1251] eta 0:01:29 lr 0.000475 time 0.2713 (0.2878) loss 2.3418 (3.5107) grad_norm 1.7635 (1.5550) [2021-04-16 06:04:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][950/1251] eta 0:01:26 lr 0.000475 time 0.2970 (0.2878) loss 4.0046 (3.5126) grad_norm 1.7328 (1.5549) [2021-04-16 06:04:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][960/1251] eta 0:01:23 lr 0.000475 time 0.2850 (0.2878) loss 3.7879 (3.5158) grad_norm 1.4255 (1.5553) [2021-04-16 06:04:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][970/1251] eta 0:01:20 lr 0.000475 time 0.2677 (0.2877) loss 3.8840 (3.5132) grad_norm 1.5021 (1.5556) [2021-04-16 06:05:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][980/1251] eta 0:01:17 lr 0.000475 time 0.2579 (0.2877) loss 2.6177 (3.5133) grad_norm 1.4380 (1.5556) [2021-04-16 06:05:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][990/1251] eta 0:01:15 lr 0.000475 time 0.2716 (0.2876) loss 3.5993 (3.5144) grad_norm 1.8466 (1.5563) [2021-04-16 06:05:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][1000/1251] eta 0:01:12 lr 0.000475 time 0.2685 (0.2875) loss 3.5185 (3.5153) grad_norm 1.6838 (1.5565) [2021-04-16 06:05:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][1010/1251] eta 0:01:09 lr 0.000475 time 0.2692 (0.2877) loss 3.6068 (3.5165) grad_norm 1.4788 (1.5567) [2021-04-16 06:05:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][1020/1251] eta 0:01:06 lr 0.000475 time 0.2847 (0.2876) loss 3.9484 (3.5153) grad_norm 1.5102 (1.5562) [2021-04-16 06:05:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][1030/1251] eta 0:01:03 lr 0.000475 time 0.2695 (0.2875) loss 2.7301 (3.5153) grad_norm 1.6583 (1.5561) [2021-04-16 06:05:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][1040/1251] eta 0:01:00 lr 0.000475 time 0.2725 (0.2874) loss 4.1145 (3.5136) grad_norm 2.2178 (1.5571) [2021-04-16 06:05:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][1050/1251] eta 0:00:57 lr 0.000475 time 0.2872 (0.2873) loss 3.5549 (3.5134) grad_norm 1.5783 (1.5579) [2021-04-16 06:05:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][1060/1251] eta 0:00:54 lr 0.000475 time 0.2985 (0.2873) loss 2.7912 (3.5148) grad_norm 1.6068 (1.5583) [2021-04-16 06:05:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][1070/1251] eta 0:00:51 lr 0.000475 time 0.2785 (0.2873) loss 3.9718 (3.5171) grad_norm 1.6447 (1.5590) [2021-04-16 06:05:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][1080/1251] eta 0:00:49 lr 0.000475 time 0.2694 (0.2872) loss 3.6034 (3.5163) grad_norm 1.6132 (1.5592) [2021-04-16 06:05:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][1090/1251] eta 0:00:46 lr 0.000475 time 0.2818 (0.2873) loss 2.5107 (3.5141) grad_norm 1.5285 (1.5592) [2021-04-16 06:05:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][1100/1251] eta 0:00:43 lr 0.000475 time 0.2792 (0.2873) loss 3.8360 (3.5153) grad_norm 1.4196 (inf) [2021-04-16 06:05:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][1110/1251] eta 0:00:40 lr 0.000475 time 0.2761 (0.2871) loss 3.8487 (3.5161) grad_norm 1.4349 (inf) [2021-04-16 06:05:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][1120/1251] eta 0:00:37 lr 0.000474 time 0.3077 (0.2871) loss 2.9194 (3.5139) grad_norm 1.5291 (inf) [2021-04-16 06:05:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][1130/1251] eta 0:00:34 lr 0.000474 time 0.2721 (0.2872) loss 3.7211 (3.5156) grad_norm 1.3889 (inf) [2021-04-16 06:05:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][1140/1251] eta 0:00:31 lr 0.000474 time 0.2941 (0.2871) loss 3.4226 (3.5174) grad_norm 1.5698 (inf) [2021-04-16 06:05:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][1150/1251] eta 0:00:28 lr 0.000474 time 0.2697 (0.2871) loss 3.7587 (3.5179) grad_norm 1.6721 (inf) [2021-04-16 06:05:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][1160/1251] eta 0:00:26 lr 0.000474 time 0.2802 (0.2870) loss 3.6443 (3.5192) grad_norm 1.6304 (inf) [2021-04-16 06:05:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][1170/1251] eta 0:00:23 lr 0.000474 time 0.2862 (0.2870) loss 3.5363 (3.5169) grad_norm 1.6273 (inf) [2021-04-16 06:05:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][1180/1251] eta 0:00:20 lr 0.000474 time 0.3069 (0.2869) loss 3.8796 (3.5169) grad_norm 1.5464 (inf) [2021-04-16 06:06:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][1190/1251] eta 0:00:17 lr 0.000474 time 0.2872 (0.2869) loss 4.4239 (3.5179) grad_norm 1.4971 (inf) [2021-04-16 06:06:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][1200/1251] eta 0:00:14 lr 0.000474 time 0.2862 (0.2868) loss 3.8958 (3.5202) grad_norm 1.5366 (inf) [2021-04-16 06:06:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][1210/1251] eta 0:00:11 lr 0.000474 time 0.2554 (0.2868) loss 3.0321 (3.5217) grad_norm 1.4433 (inf) [2021-04-16 06:06:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][1220/1251] eta 0:00:08 lr 0.000474 time 0.2849 (0.2868) loss 1.9927 (3.5208) grad_norm 1.4554 (inf) [2021-04-16 06:06:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][1230/1251] eta 0:00:06 lr 0.000474 time 0.2675 (0.2868) loss 4.2269 (3.5236) grad_norm 1.4128 (inf) [2021-04-16 06:06:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][1240/1251] eta 0:00:03 lr 0.000474 time 0.2491 (0.2867) loss 4.0107 (3.5242) grad_norm 1.9673 (inf) [2021-04-16 06:06:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [155/300][1250/1251] eta 0:00:00 lr 0.000474 time 0.2486 (0.2864) loss 3.6863 (3.5231) grad_norm 1.5400 (inf) [2021-04-16 06:06:22 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 155 training takes 0:06:03 [2021-04-16 06:06:22 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_155.pth saving...... [2021-04-16 06:06:30 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_155.pth saved !!! [2021-04-16 06:06:31 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.278 (1.278) Loss 1.0514 (1.0514) Acc@1 75.000 (75.000) Acc@5 94.043 (94.043) [2021-04-16 06:06:32 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.069 (0.217) Loss 1.0060 (1.0309) Acc@1 76.758 (75.985) Acc@5 92.871 (93.137) [2021-04-16 06:06:35 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.392 (0.224) Loss 1.0014 (1.0123) Acc@1 76.953 (76.265) Acc@5 94.141 (93.424) [2021-04-16 06:06:37 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.182 (0.230) Loss 0.9774 (1.0113) Acc@1 76.953 (76.304) Acc@5 93.848 (93.523) [2021-04-16 06:06:39 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.221) Loss 1.1274 (1.0187) Acc@1 74.609 (76.131) Acc@5 92.188 (93.445) [2021-04-16 06:06:45 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 76.256 Acc@5 93.500 [2021-04-16 06:06:45 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 76.3% [2021-04-16 06:06:45 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 76.26% [2021-04-16 06:06:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][0/1251] eta 2:41:21 lr 0.000474 time 7.7386 (7.7386) loss 2.9356 (2.9356) grad_norm 1.8935 (1.8935) [2021-04-16 06:06:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][10/1251] eta 0:19:42 lr 0.000474 time 0.2781 (0.9529) loss 3.4341 (3.4864) grad_norm 1.5723 (1.5827) [2021-04-16 06:06:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][20/1251] eta 0:13:02 lr 0.000474 time 0.2801 (0.6359) loss 3.6231 (3.5567) grad_norm 1.5280 (1.5690) [2021-04-16 06:07:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][30/1251] eta 0:10:35 lr 0.000474 time 0.2749 (0.5207) loss 2.4370 (3.5030) grad_norm 1.8104 (1.5842) [2021-04-16 06:07:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][40/1251] eta 0:09:17 lr 0.000474 time 0.2808 (0.4606) loss 3.3014 (3.5014) grad_norm 1.6006 (1.5695) [2021-04-16 06:07:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][50/1251] eta 0:08:28 lr 0.000474 time 0.2552 (0.4237) loss 3.4854 (3.5180) grad_norm 1.4921 (1.5648) [2021-04-16 06:07:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][60/1251] eta 0:07:56 lr 0.000474 time 0.2984 (0.3998) loss 2.4849 (3.4945) grad_norm 1.6329 (1.5724) [2021-04-16 06:07:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][70/1251] eta 0:07:31 lr 0.000474 time 0.2812 (0.3825) loss 4.1776 (3.5003) grad_norm 1.7076 (1.5696) [2021-04-16 06:07:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][80/1251] eta 0:07:12 lr 0.000474 time 0.2913 (0.3690) loss 4.0885 (3.4756) grad_norm 1.5842 (1.5671) [2021-04-16 06:07:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][90/1251] eta 0:06:57 lr 0.000474 time 0.2784 (0.3593) loss 2.2185 (3.4916) grad_norm 1.5520 (1.5676) [2021-04-16 06:07:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][100/1251] eta 0:06:45 lr 0.000474 time 0.2613 (0.3522) loss 3.0673 (3.4921) grad_norm 1.5203 (1.5632) [2021-04-16 06:07:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][110/1251] eta 0:06:35 lr 0.000473 time 0.2725 (0.3463) loss 3.1070 (3.4665) grad_norm 1.5894 (1.5608) [2021-04-16 06:07:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][120/1251] eta 0:06:25 lr 0.000473 time 0.2738 (0.3412) loss 3.5416 (3.4721) grad_norm 1.6293 (1.5708) [2021-04-16 06:07:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][130/1251] eta 0:06:17 lr 0.000473 time 0.2669 (0.3363) loss 3.9065 (3.4784) grad_norm 1.6821 (1.5715) [2021-04-16 06:07:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][140/1251] eta 0:06:10 lr 0.000473 time 0.2644 (0.3333) loss 3.3623 (3.4634) grad_norm 1.4280 (1.5687) [2021-04-16 06:07:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][150/1251] eta 0:06:02 lr 0.000473 time 0.2602 (0.3291) loss 2.3905 (3.4466) grad_norm 1.7612 (1.5646) [2021-04-16 06:07:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][160/1251] eta 0:05:55 lr 0.000473 time 0.2887 (0.3259) loss 3.9682 (3.4370) grad_norm 2.0949 (1.5768) [2021-04-16 06:07:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][170/1251] eta 0:05:49 lr 0.000473 time 0.2791 (0.3231) loss 3.3299 (3.4220) grad_norm 1.7040 (1.5778) [2021-04-16 06:07:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][180/1251] eta 0:05:43 lr 0.000473 time 0.4105 (0.3211) loss 3.6883 (3.4412) grad_norm 1.5596 (1.5719) [2021-04-16 06:07:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][190/1251] eta 0:05:38 lr 0.000473 time 0.2608 (0.3187) loss 3.2397 (3.4284) grad_norm 1.4623 (1.5711) [2021-04-16 06:07:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][200/1251] eta 0:05:32 lr 0.000473 time 0.2591 (0.3164) loss 3.7125 (3.4350) grad_norm 1.3222 (1.5672) [2021-04-16 06:07:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][210/1251] eta 0:05:27 lr 0.000473 time 0.2734 (0.3146) loss 2.4995 (3.4360) grad_norm 1.3060 (1.5645) [2021-04-16 06:07:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][220/1251] eta 0:05:23 lr 0.000473 time 0.2757 (0.3133) loss 4.1458 (3.4368) grad_norm 1.5150 (1.5630) [2021-04-16 06:07:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][230/1251] eta 0:05:18 lr 0.000473 time 0.2993 (0.3119) loss 3.2316 (3.4307) grad_norm 2.3111 (1.5659) [2021-04-16 06:07:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][240/1251] eta 0:05:13 lr 0.000473 time 0.2622 (0.3105) loss 3.6630 (3.4401) grad_norm 1.4003 (1.5650) [2021-04-16 06:08:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][250/1251] eta 0:05:09 lr 0.000473 time 0.2681 (0.3092) loss 2.7425 (3.4418) grad_norm 1.5642 (1.5629) [2021-04-16 06:08:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][260/1251] eta 0:05:05 lr 0.000473 time 0.2878 (0.3081) loss 3.7972 (3.4394) grad_norm 1.6012 (1.5636) [2021-04-16 06:08:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][270/1251] eta 0:05:01 lr 0.000473 time 0.3047 (0.3071) loss 2.4589 (3.4334) grad_norm 1.6388 (1.5668) [2021-04-16 06:08:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][280/1251] eta 0:04:57 lr 0.000473 time 0.2675 (0.3060) loss 2.6932 (3.4397) grad_norm 1.6558 (1.5644) [2021-04-16 06:08:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][290/1251] eta 0:04:53 lr 0.000473 time 0.2905 (0.3050) loss 3.0520 (3.4394) grad_norm 1.4746 (1.5649) [2021-04-16 06:08:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][300/1251] eta 0:04:49 lr 0.000473 time 0.2691 (0.3044) loss 2.6672 (3.4419) grad_norm 1.5580 (1.5603) [2021-04-16 06:08:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][310/1251] eta 0:04:45 lr 0.000473 time 0.2787 (0.3034) loss 3.5357 (3.4317) grad_norm 1.5023 (1.5595) [2021-04-16 06:08:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][320/1251] eta 0:04:41 lr 0.000473 time 0.2590 (0.3025) loss 2.6679 (3.4211) grad_norm 1.3714 (1.5588) [2021-04-16 06:08:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][330/1251] eta 0:04:38 lr 0.000473 time 0.2772 (0.3021) loss 4.0088 (3.4167) grad_norm 1.7550 (1.5577) [2021-04-16 06:08:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][340/1251] eta 0:04:34 lr 0.000473 time 0.2766 (0.3012) loss 3.7209 (3.4154) grad_norm 1.3527 (1.5591) [2021-04-16 06:08:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][350/1251] eta 0:04:30 lr 0.000472 time 0.2716 (0.3004) loss 4.1228 (3.4248) grad_norm 1.4962 (1.5570) [2021-04-16 06:08:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][360/1251] eta 0:04:27 lr 0.000472 time 0.2898 (0.2998) loss 3.4691 (3.4318) grad_norm 1.6156 (1.5595) [2021-04-16 06:08:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][370/1251] eta 0:04:23 lr 0.000472 time 0.2873 (0.2994) loss 2.6109 (3.4274) grad_norm 1.4834 (1.5581) [2021-04-16 06:08:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][380/1251] eta 0:04:20 lr 0.000472 time 0.2621 (0.2989) loss 4.4784 (3.4305) grad_norm 1.5852 (1.5600) [2021-04-16 06:08:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][390/1251] eta 0:04:16 lr 0.000472 time 0.2620 (0.2984) loss 3.6703 (3.4355) grad_norm 1.8188 (1.5631) [2021-04-16 06:08:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][400/1251] eta 0:04:13 lr 0.000472 time 0.2773 (0.2979) loss 3.8902 (3.4351) grad_norm 1.5675 (1.5621) [2021-04-16 06:08:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][410/1251] eta 0:04:10 lr 0.000472 time 0.2979 (0.2977) loss 3.6750 (3.4350) grad_norm 1.4727 (1.5627) [2021-04-16 06:08:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][420/1251] eta 0:04:06 lr 0.000472 time 0.2583 (0.2971) loss 3.4478 (3.4394) grad_norm 1.6401 (1.5656) [2021-04-16 06:08:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][430/1251] eta 0:04:03 lr 0.000472 time 0.2910 (0.2968) loss 3.8310 (3.4371) grad_norm 1.4573 (1.5678) [2021-04-16 06:08:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][440/1251] eta 0:04:00 lr 0.000472 time 0.2693 (0.2965) loss 3.3860 (3.4327) grad_norm 1.4699 (1.5681) [2021-04-16 06:08:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][450/1251] eta 0:03:56 lr 0.000472 time 0.2651 (0.2959) loss 4.0846 (3.4334) grad_norm 1.5914 (1.5663) [2021-04-16 06:09:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][460/1251] eta 0:03:53 lr 0.000472 time 0.2812 (0.2954) loss 3.2055 (3.4332) grad_norm 1.4398 (1.5677) [2021-04-16 06:09:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][470/1251] eta 0:03:50 lr 0.000472 time 0.2579 (0.2955) loss 3.1347 (3.4338) grad_norm 1.5091 (1.5666) [2021-04-16 06:09:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][480/1251] eta 0:03:47 lr 0.000472 time 0.2887 (0.2952) loss 4.1275 (3.4387) grad_norm 1.5948 (1.5678) [2021-04-16 06:09:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][490/1251] eta 0:03:44 lr 0.000472 time 0.2665 (0.2950) loss 3.6597 (3.4424) grad_norm 1.4780 (1.5685) [2021-04-16 06:09:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][500/1251] eta 0:03:41 lr 0.000472 time 0.2752 (0.2947) loss 2.9906 (3.4354) grad_norm 1.6727 (1.5698) [2021-04-16 06:09:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][510/1251] eta 0:03:38 lr 0.000472 time 0.2765 (0.2944) loss 2.2130 (3.4392) grad_norm 1.4013 (1.5708) [2021-04-16 06:09:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][520/1251] eta 0:03:35 lr 0.000472 time 0.2686 (0.2945) loss 2.7962 (3.4433) grad_norm 1.4994 (1.5707) [2021-04-16 06:09:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][530/1251] eta 0:03:32 lr 0.000472 time 0.2863 (0.2942) loss 4.2474 (3.4421) grad_norm 1.3849 (1.5712) [2021-04-16 06:09:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][540/1251] eta 0:03:29 lr 0.000472 time 0.2609 (0.2940) loss 3.5009 (3.4443) grad_norm 1.3819 (1.5718) [2021-04-16 06:09:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][550/1251] eta 0:03:25 lr 0.000472 time 0.2740 (0.2936) loss 3.9188 (3.4429) grad_norm 1.5812 (1.5710) [2021-04-16 06:09:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][560/1251] eta 0:03:22 lr 0.000472 time 0.2748 (0.2933) loss 3.9524 (3.4471) grad_norm 1.8057 (1.5729) [2021-04-16 06:09:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][570/1251] eta 0:03:19 lr 0.000472 time 0.3021 (0.2930) loss 3.6808 (3.4444) grad_norm 1.4271 (1.5733) [2021-04-16 06:09:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][580/1251] eta 0:03:16 lr 0.000472 time 0.2652 (0.2930) loss 3.0605 (3.4482) grad_norm 1.7872 (1.5730) [2021-04-16 06:09:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][590/1251] eta 0:03:13 lr 0.000471 time 0.2853 (0.2929) loss 3.7770 (3.4522) grad_norm 1.6216 (1.5726) [2021-04-16 06:09:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][600/1251] eta 0:03:10 lr 0.000471 time 0.2644 (0.2926) loss 4.2433 (3.4505) grad_norm 1.5097 (1.5716) [2021-04-16 06:09:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][610/1251] eta 0:03:07 lr 0.000471 time 0.2735 (0.2924) loss 2.8644 (3.4516) grad_norm 1.7739 (1.5708) [2021-04-16 06:09:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][620/1251] eta 0:03:04 lr 0.000471 time 0.2633 (0.2921) loss 2.8015 (3.4494) grad_norm 1.7879 (1.5700) [2021-04-16 06:09:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][630/1251] eta 0:03:01 lr 0.000471 time 0.2589 (0.2918) loss 4.0247 (3.4542) grad_norm 1.5768 (1.5712) [2021-04-16 06:09:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][640/1251] eta 0:02:58 lr 0.000471 time 0.2425 (0.2915) loss 3.9800 (3.4593) grad_norm 1.7427 (1.5721) [2021-04-16 06:09:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][650/1251] eta 0:02:55 lr 0.000471 time 0.2981 (0.2915) loss 3.1105 (3.4608) grad_norm 1.3190 (1.5725) [2021-04-16 06:09:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][660/1251] eta 0:02:52 lr 0.000471 time 0.2658 (0.2913) loss 4.0616 (3.4642) grad_norm 1.5494 (1.5731) [2021-04-16 06:10:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][670/1251] eta 0:02:49 lr 0.000471 time 0.2745 (0.2911) loss 3.3416 (3.4668) grad_norm 1.6809 (1.5737) [2021-04-16 06:10:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][680/1251] eta 0:02:46 lr 0.000471 time 0.2546 (0.2910) loss 3.0081 (3.4655) grad_norm 1.8135 (1.5740) [2021-04-16 06:10:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][690/1251] eta 0:02:43 lr 0.000471 time 0.2607 (0.2909) loss 3.5688 (3.4668) grad_norm 1.4259 (1.5747) [2021-04-16 06:10:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][700/1251] eta 0:02:40 lr 0.000471 time 0.2773 (0.2907) loss 3.7183 (3.4662) grad_norm 1.7505 (1.5748) [2021-04-16 06:10:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][710/1251] eta 0:02:37 lr 0.000471 time 0.2879 (0.2906) loss 3.8135 (3.4671) grad_norm 1.8099 (1.5745) [2021-04-16 06:10:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][720/1251] eta 0:02:34 lr 0.000471 time 0.2885 (0.2906) loss 3.9061 (3.4692) grad_norm 1.5597 (1.5742) [2021-04-16 06:10:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][730/1251] eta 0:02:31 lr 0.000471 time 0.2927 (0.2904) loss 3.3809 (3.4687) grad_norm 1.5971 (1.5739) [2021-04-16 06:10:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][740/1251] eta 0:02:28 lr 0.000471 time 0.2701 (0.2903) loss 3.7743 (3.4683) grad_norm 1.3952 (1.5741) [2021-04-16 06:10:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][750/1251] eta 0:02:25 lr 0.000471 time 0.2912 (0.2902) loss 2.9463 (3.4700) grad_norm 1.5951 (1.5740) [2021-04-16 06:10:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][760/1251] eta 0:02:22 lr 0.000471 time 0.2696 (0.2903) loss 4.2860 (3.4731) grad_norm 1.5949 (1.5746) [2021-04-16 06:10:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][770/1251] eta 0:02:19 lr 0.000471 time 0.2826 (0.2902) loss 3.9998 (3.4762) grad_norm 1.6410 (1.5764) [2021-04-16 06:10:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][780/1251] eta 0:02:16 lr 0.000471 time 0.2830 (0.2901) loss 3.9235 (3.4761) grad_norm 1.8050 (1.5769) [2021-04-16 06:10:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][790/1251] eta 0:02:13 lr 0.000471 time 0.2851 (0.2901) loss 2.4593 (3.4732) grad_norm 1.5938 (1.5776) [2021-04-16 06:10:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][800/1251] eta 0:02:10 lr 0.000471 time 0.2801 (0.2899) loss 3.4973 (3.4706) grad_norm 1.3704 (1.5760) [2021-04-16 06:10:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][810/1251] eta 0:02:07 lr 0.000471 time 0.2510 (0.2897) loss 3.8275 (3.4748) grad_norm 1.6327 (1.5754) [2021-04-16 06:10:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][820/1251] eta 0:02:04 lr 0.000471 time 0.2850 (0.2895) loss 3.2291 (3.4754) grad_norm 1.8903 (1.5752) [2021-04-16 06:10:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][830/1251] eta 0:02:01 lr 0.000470 time 0.2673 (0.2894) loss 3.4798 (3.4758) grad_norm 1.4852 (1.5757) [2021-04-16 06:10:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][840/1251] eta 0:01:58 lr 0.000470 time 0.2560 (0.2892) loss 4.2061 (3.4801) grad_norm 1.4791 (1.5757) [2021-04-16 06:10:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][850/1251] eta 0:01:55 lr 0.000470 time 0.2682 (0.2891) loss 2.7485 (3.4834) grad_norm 1.4875 (1.5764) [2021-04-16 06:10:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][860/1251] eta 0:01:52 lr 0.000470 time 0.2731 (0.2890) loss 2.6812 (3.4837) grad_norm 1.9001 (1.5778) [2021-04-16 06:10:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][870/1251] eta 0:01:50 lr 0.000470 time 0.3028 (0.2888) loss 3.8968 (3.4819) grad_norm 1.6343 (1.5777) [2021-04-16 06:10:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][880/1251] eta 0:01:47 lr 0.000470 time 0.2888 (0.2888) loss 2.2062 (3.4795) grad_norm 1.5808 (1.5768) [2021-04-16 06:11:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][890/1251] eta 0:01:44 lr 0.000470 time 0.2658 (0.2887) loss 4.0070 (3.4834) grad_norm 1.4933 (1.5786) [2021-04-16 06:11:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][900/1251] eta 0:01:41 lr 0.000470 time 0.2774 (0.2886) loss 3.1472 (3.4843) grad_norm 1.6706 (1.5804) [2021-04-16 06:11:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][910/1251] eta 0:01:38 lr 0.000470 time 0.2754 (0.2886) loss 3.2170 (3.4824) grad_norm 1.4462 (1.5819) [2021-04-16 06:11:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][920/1251] eta 0:01:35 lr 0.000470 time 0.2663 (0.2884) loss 3.7062 (3.4831) grad_norm 1.4768 (1.5822) [2021-04-16 06:11:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][930/1251] eta 0:01:32 lr 0.000470 time 0.2662 (0.2883) loss 3.7573 (3.4869) grad_norm 1.5667 (1.5816) [2021-04-16 06:11:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][940/1251] eta 0:01:29 lr 0.000470 time 0.2896 (0.2881) loss 3.8384 (3.4869) grad_norm 1.6852 (1.5822) [2021-04-16 06:11:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][950/1251] eta 0:01:26 lr 0.000470 time 0.2590 (0.2880) loss 2.8762 (3.4884) grad_norm 1.6430 (1.5820) [2021-04-16 06:11:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][960/1251] eta 0:01:23 lr 0.000470 time 0.3026 (0.2878) loss 3.1946 (3.4859) grad_norm 1.4250 (1.5817) [2021-04-16 06:11:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][970/1251] eta 0:01:20 lr 0.000470 time 0.2663 (0.2877) loss 3.3162 (3.4903) grad_norm 1.5545 (1.5823) [2021-04-16 06:11:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][980/1251] eta 0:01:17 lr 0.000470 time 0.2651 (0.2875) loss 3.5608 (3.4911) grad_norm 1.7020 (1.5826) [2021-04-16 06:11:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][990/1251] eta 0:01:15 lr 0.000470 time 0.2620 (0.2874) loss 4.1358 (3.4931) grad_norm 1.4085 (1.5824) [2021-04-16 06:11:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][1000/1251] eta 0:01:12 lr 0.000470 time 0.2667 (0.2873) loss 3.5549 (3.4938) grad_norm 1.4354 (1.5823) [2021-04-16 06:11:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][1010/1251] eta 0:01:09 lr 0.000470 time 0.2759 (0.2873) loss 3.6911 (3.4971) grad_norm 1.5441 (1.5822) [2021-04-16 06:11:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][1020/1251] eta 0:01:06 lr 0.000470 time 0.2601 (0.2873) loss 4.0984 (3.5008) grad_norm 1.4629 (1.5818) [2021-04-16 06:11:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][1030/1251] eta 0:01:03 lr 0.000470 time 0.3035 (0.2874) loss 3.9792 (3.5020) grad_norm 1.6148 (1.5824) [2021-04-16 06:11:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][1040/1251] eta 0:01:00 lr 0.000470 time 0.2655 (0.2874) loss 2.6424 (3.5009) grad_norm 1.8556 (1.5837) [2021-04-16 06:11:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][1050/1251] eta 0:00:57 lr 0.000470 time 0.2721 (0.2873) loss 3.7037 (3.5035) grad_norm 1.5270 (1.5831) [2021-04-16 06:11:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][1060/1251] eta 0:00:54 lr 0.000470 time 0.3006 (0.2872) loss 3.9829 (3.5052) grad_norm 1.6212 (1.5832) [2021-04-16 06:11:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][1070/1251] eta 0:00:51 lr 0.000469 time 0.2922 (0.2872) loss 2.6335 (3.5059) grad_norm 1.5448 (1.5826) [2021-04-16 06:11:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][1080/1251] eta 0:00:49 lr 0.000469 time 0.2684 (0.2871) loss 2.7824 (3.5072) grad_norm 1.3808 (1.5820) [2021-04-16 06:11:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][1090/1251] eta 0:00:46 lr 0.000469 time 0.2829 (0.2872) loss 3.6018 (3.5069) grad_norm 1.4819 (1.5822) [2021-04-16 06:12:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][1100/1251] eta 0:00:43 lr 0.000469 time 0.2889 (0.2870) loss 3.2296 (3.5060) grad_norm 1.6074 (1.5823) [2021-04-16 06:12:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][1110/1251] eta 0:00:40 lr 0.000469 time 0.2726 (0.2870) loss 4.6197 (3.5076) grad_norm 1.5086 (1.5817) [2021-04-16 06:12:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][1120/1251] eta 0:00:37 lr 0.000469 time 0.2825 (0.2870) loss 3.4475 (3.5084) grad_norm 2.0184 (1.5828) [2021-04-16 06:12:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][1130/1251] eta 0:00:34 lr 0.000469 time 0.2694 (0.2869) loss 2.7768 (3.5093) grad_norm 1.3489 (1.5824) [2021-04-16 06:12:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][1140/1251] eta 0:00:31 lr 0.000469 time 0.2886 (0.2868) loss 3.7316 (3.5128) grad_norm 1.6400 (1.5823) [2021-04-16 06:12:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][1150/1251] eta 0:00:28 lr 0.000469 time 0.2871 (0.2867) loss 2.9211 (3.5131) grad_norm 1.7077 (1.5829) [2021-04-16 06:12:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][1160/1251] eta 0:00:26 lr 0.000469 time 0.2872 (0.2868) loss 3.2175 (3.5144) grad_norm 1.4612 (1.5829) [2021-04-16 06:12:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][1170/1251] eta 0:00:23 lr 0.000469 time 0.2512 (0.2868) loss 3.8780 (3.5142) grad_norm 1.8932 (1.5833) [2021-04-16 06:12:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][1180/1251] eta 0:00:20 lr 0.000469 time 0.2667 (0.2867) loss 4.2165 (3.5148) grad_norm 1.4293 (1.5835) [2021-04-16 06:12:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][1190/1251] eta 0:00:17 lr 0.000469 time 0.2702 (0.2866) loss 3.0325 (3.5146) grad_norm 1.5230 (1.5831) [2021-04-16 06:12:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][1200/1251] eta 0:00:14 lr 0.000469 time 0.2616 (0.2865) loss 3.4589 (3.5142) grad_norm 1.7088 (1.5831) [2021-04-16 06:12:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][1210/1251] eta 0:00:11 lr 0.000469 time 0.2823 (0.2864) loss 4.2595 (3.5128) grad_norm 1.4418 (1.5828) [2021-04-16 06:12:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][1220/1251] eta 0:00:08 lr 0.000469 time 0.3014 (0.2863) loss 3.4700 (3.5121) grad_norm 1.6279 (1.5827) [2021-04-16 06:12:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][1230/1251] eta 0:00:06 lr 0.000469 time 0.2582 (0.2863) loss 3.9537 (3.5140) grad_norm 1.5365 (1.5820) [2021-04-16 06:12:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][1240/1251] eta 0:00:03 lr 0.000469 time 0.2482 (0.2861) loss 2.8342 (3.5152) grad_norm 1.4364 (1.5815) [2021-04-16 06:12:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [156/300][1250/1251] eta 0:00:00 lr 0.000469 time 0.2492 (0.2858) loss 3.8379 (3.5167) grad_norm 1.6526 (1.5818) [2021-04-16 06:12:46 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 156 training takes 0:06:00 [2021-04-16 06:12:46 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_156.pth saving...... [2021-04-16 06:12:53 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_156.pth saved !!! [2021-04-16 06:12:54 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.164 (1.164) Loss 0.9626 (0.9626) Acc@1 77.246 (77.246) Acc@5 94.531 (94.531) [2021-04-16 06:12:55 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.118 (0.241) Loss 1.0001 (1.0097) Acc@1 77.246 (76.465) Acc@5 93.945 (93.519) [2021-04-16 06:12:57 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.552 (0.226) Loss 1.0949 (1.0190) Acc@1 74.023 (76.311) Acc@5 92.383 (93.364) [2021-04-16 06:13:00 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.112 (0.240) Loss 0.9916 (1.0136) Acc@1 78.125 (76.449) Acc@5 93.945 (93.463) [2021-04-16 06:13:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.217) Loss 0.9793 (1.0145) Acc@1 78.613 (76.341) Acc@5 93.066 (93.457) [2021-04-16 06:13:07 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 76.198 Acc@5 93.460 [2021-04-16 06:13:07 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 76.2% [2021-04-16 06:13:07 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 76.26% [2021-04-16 06:13:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][0/1251] eta 3:24:53 lr 0.000469 time 9.8271 (9.8271) loss 2.8076 (2.8076) grad_norm 1.4822 (1.4822) [2021-04-16 06:13:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][10/1251] eta 0:23:46 lr 0.000469 time 0.4052 (1.1497) loss 3.1076 (3.4150) grad_norm 1.4208 (1.4766) [2021-04-16 06:13:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][20/1251] eta 0:15:00 lr 0.000469 time 0.2857 (0.7316) loss 3.9267 (3.3935) grad_norm 1.9336 (1.5554) [2021-04-16 06:13:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][30/1251] eta 0:11:53 lr 0.000469 time 0.2713 (0.5848) loss 2.2285 (3.4277) grad_norm 1.6348 (1.5780) [2021-04-16 06:13:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][40/1251] eta 0:10:15 lr 0.000469 time 0.2470 (0.5086) loss 4.1079 (3.4017) grad_norm 1.3468 (1.5735) [2021-04-16 06:13:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][50/1251] eta 0:09:17 lr 0.000469 time 0.2975 (0.4640) loss 3.7001 (3.4877) grad_norm 1.4433 (1.5730) [2021-04-16 06:13:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][60/1251] eta 0:08:36 lr 0.000468 time 0.2954 (0.4335) loss 3.5000 (3.4870) grad_norm 1.4597 (1.5747) [2021-04-16 06:13:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][70/1251] eta 0:08:05 lr 0.000468 time 0.2851 (0.4114) loss 3.8510 (3.4833) grad_norm 1.4783 (1.5680) [2021-04-16 06:13:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][80/1251] eta 0:07:44 lr 0.000468 time 0.2894 (0.3968) loss 4.2587 (3.4765) grad_norm 1.6901 (1.5704) [2021-04-16 06:13:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][90/1251] eta 0:07:27 lr 0.000468 time 0.2953 (0.3853) loss 4.1623 (3.4769) grad_norm 1.5524 (1.5690) [2021-04-16 06:13:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][100/1251] eta 0:07:13 lr 0.000468 time 0.2478 (0.3767) loss 3.7255 (3.4889) grad_norm 1.4856 (1.5741) [2021-04-16 06:13:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][110/1251] eta 0:06:59 lr 0.000468 time 0.2750 (0.3675) loss 3.5174 (3.5045) grad_norm 1.4638 (1.5718) [2021-04-16 06:13:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][120/1251] eta 0:06:47 lr 0.000468 time 0.3175 (0.3604) loss 4.0012 (3.5247) grad_norm 2.1054 (1.5786) [2021-04-16 06:13:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][130/1251] eta 0:06:37 lr 0.000468 time 0.2779 (0.3543) loss 3.2775 (3.4994) grad_norm 1.7585 (1.5829) [2021-04-16 06:13:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][140/1251] eta 0:06:27 lr 0.000468 time 0.2433 (0.3484) loss 3.5553 (3.4871) grad_norm 1.4149 (1.5821) [2021-04-16 06:13:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][150/1251] eta 0:06:18 lr 0.000468 time 0.2630 (0.3437) loss 3.3918 (3.4857) grad_norm 1.4593 (1.5794) [2021-04-16 06:14:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][160/1251] eta 0:06:10 lr 0.000468 time 0.2597 (0.3395) loss 4.0394 (3.4890) grad_norm 1.7831 (1.5793) [2021-04-16 06:14:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][170/1251] eta 0:06:03 lr 0.000468 time 0.2907 (0.3359) loss 2.9670 (3.4934) grad_norm 1.4563 (1.5797) [2021-04-16 06:14:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][180/1251] eta 0:05:56 lr 0.000468 time 0.2811 (0.3326) loss 3.3397 (3.4778) grad_norm 1.4100 (1.5773) [2021-04-16 06:14:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][190/1251] eta 0:05:49 lr 0.000468 time 0.2589 (0.3294) loss 3.5963 (3.4694) grad_norm 1.4307 (1.5789) [2021-04-16 06:14:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][200/1251] eta 0:05:43 lr 0.000468 time 0.2828 (0.3268) loss 3.8290 (3.4831) grad_norm 1.4892 (1.5849) [2021-04-16 06:14:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][210/1251] eta 0:05:37 lr 0.000468 time 0.2857 (0.3243) loss 3.3966 (3.4867) grad_norm 1.8347 (1.5878) [2021-04-16 06:14:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][220/1251] eta 0:05:33 lr 0.000468 time 0.2768 (0.3230) loss 3.5663 (3.4870) grad_norm 1.6913 (1.5906) [2021-04-16 06:14:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][230/1251] eta 0:05:27 lr 0.000468 time 0.2884 (0.3208) loss 3.6535 (3.4821) grad_norm 1.6872 (1.5935) [2021-04-16 06:14:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][240/1251] eta 0:05:23 lr 0.000468 time 0.2572 (0.3196) loss 2.9938 (3.4841) grad_norm 1.5689 (1.5887) [2021-04-16 06:14:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][250/1251] eta 0:05:18 lr 0.000468 time 0.2755 (0.3184) loss 3.1710 (3.4916) grad_norm 1.5979 (1.5843) [2021-04-16 06:14:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][260/1251] eta 0:05:13 lr 0.000468 time 0.2604 (0.3167) loss 3.9265 (3.4989) grad_norm 1.4147 (1.5818) [2021-04-16 06:14:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][270/1251] eta 0:05:09 lr 0.000468 time 0.2980 (0.3152) loss 2.5116 (3.4937) grad_norm 1.4946 (1.5812) [2021-04-16 06:14:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][280/1251] eta 0:05:04 lr 0.000468 time 0.2691 (0.3137) loss 2.6198 (3.4951) grad_norm 1.3992 (1.5792) [2021-04-16 06:14:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][290/1251] eta 0:05:00 lr 0.000468 time 0.2745 (0.3128) loss 3.7143 (3.4947) grad_norm 1.4988 (1.5818) [2021-04-16 06:14:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][300/1251] eta 0:04:56 lr 0.000468 time 0.2719 (0.3120) loss 3.5177 (3.4954) grad_norm 1.4807 (1.5796) [2021-04-16 06:14:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][310/1251] eta 0:04:52 lr 0.000467 time 0.2760 (0.3109) loss 3.1745 (3.4845) grad_norm 1.4660 (1.5788) [2021-04-16 06:14:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][320/1251] eta 0:04:48 lr 0.000467 time 0.2624 (0.3097) loss 3.1827 (3.4854) grad_norm 1.6781 (1.5786) [2021-04-16 06:14:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][330/1251] eta 0:04:44 lr 0.000467 time 0.3008 (0.3089) loss 4.1303 (3.4827) grad_norm 1.7847 (1.5773) [2021-04-16 06:14:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][340/1251] eta 0:04:40 lr 0.000467 time 0.2900 (0.3080) loss 3.2584 (3.4784) grad_norm 1.6011 (1.5776) [2021-04-16 06:14:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][350/1251] eta 0:04:37 lr 0.000467 time 0.4396 (0.3076) loss 2.4415 (3.4812) grad_norm 1.7046 (1.5780) [2021-04-16 06:14:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][360/1251] eta 0:04:33 lr 0.000467 time 0.2606 (0.3067) loss 3.7178 (3.4848) grad_norm 1.5830 (1.5806) [2021-04-16 06:15:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][370/1251] eta 0:04:29 lr 0.000467 time 0.2686 (0.3060) loss 2.7444 (3.4839) grad_norm 1.8882 (1.5840) [2021-04-16 06:15:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][380/1251] eta 0:04:25 lr 0.000467 time 0.2680 (0.3054) loss 4.1846 (3.4833) grad_norm 1.5825 (1.5864) [2021-04-16 06:15:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][390/1251] eta 0:04:22 lr 0.000467 time 0.2982 (0.3047) loss 3.8709 (3.4878) grad_norm 1.3181 (1.5849) [2021-04-16 06:15:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][400/1251] eta 0:04:18 lr 0.000467 time 0.2726 (0.3040) loss 3.8749 (3.4838) grad_norm 1.8002 (1.5876) [2021-04-16 06:15:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][410/1251] eta 0:04:15 lr 0.000467 time 0.2445 (0.3033) loss 2.5389 (3.4854) grad_norm 1.3114 (1.5863) [2021-04-16 06:15:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][420/1251] eta 0:04:11 lr 0.000467 time 0.3069 (0.3027) loss 4.1980 (3.4863) grad_norm 1.6386 (1.5846) [2021-04-16 06:15:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][430/1251] eta 0:04:08 lr 0.000467 time 0.2699 (0.3021) loss 3.8950 (3.4833) grad_norm 1.4788 (1.5846) [2021-04-16 06:15:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][440/1251] eta 0:04:04 lr 0.000467 time 0.2717 (0.3015) loss 3.2307 (3.4810) grad_norm 1.5800 (1.5853) [2021-04-16 06:15:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][450/1251] eta 0:04:00 lr 0.000467 time 0.2534 (0.3009) loss 2.9277 (3.4740) grad_norm 1.6032 (1.5854) [2021-04-16 06:15:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][460/1251] eta 0:03:57 lr 0.000467 time 0.2636 (0.3003) loss 3.4834 (3.4753) grad_norm 1.5806 (1.5871) [2021-04-16 06:15:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][470/1251] eta 0:03:54 lr 0.000467 time 0.2541 (0.2998) loss 4.2537 (3.4806) grad_norm 1.4429 (1.5872) [2021-04-16 06:15:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][480/1251] eta 0:03:50 lr 0.000467 time 0.2925 (0.2994) loss 3.1581 (3.4804) grad_norm 1.7165 (1.5857) [2021-04-16 06:15:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][490/1251] eta 0:03:47 lr 0.000467 time 0.2736 (0.2992) loss 3.1209 (3.4791) grad_norm 1.4394 (1.5839) [2021-04-16 06:15:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][500/1251] eta 0:03:44 lr 0.000467 time 0.2620 (0.2988) loss 3.1807 (3.4766) grad_norm 1.6371 (1.5842) [2021-04-16 06:15:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][510/1251] eta 0:03:41 lr 0.000467 time 0.2902 (0.2984) loss 3.6724 (3.4731) grad_norm 1.7012 (1.5842) [2021-04-16 06:15:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][520/1251] eta 0:03:38 lr 0.000467 time 0.2581 (0.2983) loss 3.6982 (3.4718) grad_norm 1.3361 (1.5844) [2021-04-16 06:15:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][530/1251] eta 0:03:34 lr 0.000467 time 0.2946 (0.2979) loss 3.6761 (3.4706) grad_norm 1.7458 (1.5852) [2021-04-16 06:15:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][540/1251] eta 0:03:31 lr 0.000467 time 0.2766 (0.2975) loss 3.3436 (3.4657) grad_norm 1.3249 (1.5831) [2021-04-16 06:15:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][550/1251] eta 0:03:28 lr 0.000466 time 0.2780 (0.2971) loss 3.8895 (3.4716) grad_norm 1.3472 (1.5814) [2021-04-16 06:15:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][560/1251] eta 0:03:25 lr 0.000466 time 0.2587 (0.2968) loss 3.9305 (3.4723) grad_norm 1.3632 (1.5797) [2021-04-16 06:15:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][570/1251] eta 0:03:21 lr 0.000466 time 0.2782 (0.2964) loss 3.3441 (3.4709) grad_norm 1.3804 (1.5789) [2021-04-16 06:16:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][580/1251] eta 0:03:18 lr 0.000466 time 0.2671 (0.2965) loss 3.6615 (3.4760) grad_norm 1.6696 (1.5788) [2021-04-16 06:16:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][590/1251] eta 0:03:15 lr 0.000466 time 0.2662 (0.2964) loss 2.5809 (3.4713) grad_norm 1.9297 (1.5802) [2021-04-16 06:16:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][600/1251] eta 0:03:12 lr 0.000466 time 0.2717 (0.2961) loss 3.6662 (3.4699) grad_norm 1.7556 (1.5798) [2021-04-16 06:16:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][610/1251] eta 0:03:09 lr 0.000466 time 0.2767 (0.2958) loss 3.7056 (3.4729) grad_norm 1.5310 (1.5813) [2021-04-16 06:16:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][620/1251] eta 0:03:06 lr 0.000466 time 0.2672 (0.2955) loss 3.4393 (3.4738) grad_norm 1.4096 (1.5807) [2021-04-16 06:16:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][630/1251] eta 0:03:03 lr 0.000466 time 0.2606 (0.2952) loss 2.8440 (3.4730) grad_norm 1.4773 (1.5799) [2021-04-16 06:16:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][640/1251] eta 0:03:00 lr 0.000466 time 0.2890 (0.2949) loss 2.9867 (3.4752) grad_norm 1.6729 (1.5802) [2021-04-16 06:16:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][650/1251] eta 0:02:57 lr 0.000466 time 0.4292 (0.2948) loss 3.5657 (3.4733) grad_norm 1.7093 (1.5815) [2021-04-16 06:16:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][660/1251] eta 0:02:54 lr 0.000466 time 0.2563 (0.2945) loss 3.8525 (3.4737) grad_norm 1.8824 (1.5824) [2021-04-16 06:16:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][670/1251] eta 0:02:51 lr 0.000466 time 0.3052 (0.2946) loss 4.2422 (3.4775) grad_norm 1.8213 (1.5846) [2021-04-16 06:16:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][680/1251] eta 0:02:48 lr 0.000466 time 0.2815 (0.2944) loss 2.3124 (3.4770) grad_norm 1.5827 (1.5846) [2021-04-16 06:16:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][690/1251] eta 0:02:45 lr 0.000466 time 0.2816 (0.2941) loss 2.3509 (3.4731) grad_norm 1.5454 (1.5847) [2021-04-16 06:16:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][700/1251] eta 0:02:41 lr 0.000466 time 0.2862 (0.2939) loss 3.3083 (3.4741) grad_norm 1.5783 (1.5855) [2021-04-16 06:16:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][710/1251] eta 0:02:38 lr 0.000466 time 0.2418 (0.2936) loss 3.8952 (3.4735) grad_norm 1.4276 (1.5840) [2021-04-16 06:16:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][720/1251] eta 0:02:35 lr 0.000466 time 0.2739 (0.2934) loss 4.0983 (3.4785) grad_norm 1.3621 (1.5841) [2021-04-16 06:16:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][730/1251] eta 0:02:32 lr 0.000466 time 0.2955 (0.2931) loss 3.4941 (3.4759) grad_norm 1.6130 (1.5845) [2021-04-16 06:16:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][740/1251] eta 0:02:29 lr 0.000466 time 0.2566 (0.2929) loss 2.3978 (3.4719) grad_norm 1.4637 (1.5854) [2021-04-16 06:16:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][750/1251] eta 0:02:26 lr 0.000466 time 0.2985 (0.2927) loss 3.0927 (3.4753) grad_norm 1.3565 (1.5859) [2021-04-16 06:16:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][760/1251] eta 0:02:23 lr 0.000466 time 0.2826 (0.2924) loss 3.7537 (3.4753) grad_norm 1.3073 (1.5855) [2021-04-16 06:16:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][770/1251] eta 0:02:20 lr 0.000466 time 0.2900 (0.2922) loss 3.4576 (3.4754) grad_norm 1.4004 (1.5846) [2021-04-16 06:16:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][780/1251] eta 0:02:17 lr 0.000466 time 0.3000 (0.2920) loss 3.7702 (3.4758) grad_norm 1.7934 (1.5848) [2021-04-16 06:16:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][790/1251] eta 0:02:14 lr 0.000465 time 0.2667 (0.2921) loss 3.8984 (3.4781) grad_norm 1.7555 (1.5855) [2021-04-16 06:17:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][800/1251] eta 0:02:11 lr 0.000465 time 0.2670 (0.2918) loss 4.1271 (3.4777) grad_norm 1.5362 (1.5848) [2021-04-16 06:17:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][810/1251] eta 0:02:08 lr 0.000465 time 0.2563 (0.2916) loss 3.3776 (3.4749) grad_norm 1.6683 (1.5844) [2021-04-16 06:17:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][820/1251] eta 0:02:05 lr 0.000465 time 0.2717 (0.2916) loss 3.8841 (3.4735) grad_norm 1.6722 (1.5849) [2021-04-16 06:17:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][830/1251] eta 0:02:02 lr 0.000465 time 0.2924 (0.2915) loss 3.7158 (3.4741) grad_norm 1.5278 (1.5852) [2021-04-16 06:17:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][840/1251] eta 0:01:59 lr 0.000465 time 0.2640 (0.2914) loss 3.8065 (3.4752) grad_norm 1.6558 (1.5849) [2021-04-16 06:17:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][850/1251] eta 0:01:56 lr 0.000465 time 0.2881 (0.2912) loss 3.8212 (3.4791) grad_norm 1.4664 (1.5851) [2021-04-16 06:17:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][860/1251] eta 0:01:53 lr 0.000465 time 0.2937 (0.2911) loss 3.7379 (3.4795) grad_norm 1.4397 (1.5847) [2021-04-16 06:17:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][870/1251] eta 0:01:50 lr 0.000465 time 0.2752 (0.2910) loss 3.6436 (3.4840) grad_norm 1.6227 (1.5851) [2021-04-16 06:17:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][880/1251] eta 0:01:47 lr 0.000465 time 0.2674 (0.2909) loss 4.0209 (3.4863) grad_norm 1.4374 (1.5849) [2021-04-16 06:17:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][890/1251] eta 0:01:44 lr 0.000465 time 0.2881 (0.2908) loss 2.5897 (3.4876) grad_norm 1.6163 (1.5847) [2021-04-16 06:17:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][900/1251] eta 0:01:42 lr 0.000465 time 0.2757 (0.2906) loss 3.5148 (3.4882) grad_norm 1.7435 (1.5857) [2021-04-16 06:17:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][910/1251] eta 0:01:39 lr 0.000465 time 0.2889 (0.2905) loss 3.8311 (3.4892) grad_norm 1.5417 (1.5865) [2021-04-16 06:17:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][920/1251] eta 0:01:36 lr 0.000465 time 0.2882 (0.2903) loss 3.8115 (3.4923) grad_norm 1.4428 (1.5864) [2021-04-16 06:17:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][930/1251] eta 0:01:33 lr 0.000465 time 0.2915 (0.2903) loss 3.1092 (3.4940) grad_norm 1.4093 (1.5872) [2021-04-16 06:17:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][940/1251] eta 0:01:30 lr 0.000465 time 0.3657 (0.2903) loss 3.3330 (3.4924) grad_norm 1.5753 (1.5875) [2021-04-16 06:17:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][950/1251] eta 0:01:27 lr 0.000465 time 0.2696 (0.2902) loss 4.0005 (3.4946) grad_norm 1.5725 (1.5893) [2021-04-16 06:17:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][960/1251] eta 0:01:24 lr 0.000465 time 0.2764 (0.2900) loss 3.1423 (3.4946) grad_norm 1.4028 (1.5897) [2021-04-16 06:17:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][970/1251] eta 0:01:21 lr 0.000465 time 0.2864 (0.2899) loss 4.4344 (3.4967) grad_norm 1.6032 (1.5888) [2021-04-16 06:17:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][980/1251] eta 0:01:18 lr 0.000465 time 0.2524 (0.2898) loss 3.3605 (3.4966) grad_norm 1.5497 (1.5893) [2021-04-16 06:17:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][990/1251] eta 0:01:15 lr 0.000465 time 0.2719 (0.2896) loss 3.9418 (3.4959) grad_norm 1.8513 (1.5894) [2021-04-16 06:17:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][1000/1251] eta 0:01:12 lr 0.000465 time 0.2728 (0.2895) loss 3.8643 (3.4945) grad_norm 1.4653 (1.5888) [2021-04-16 06:18:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][1010/1251] eta 0:01:09 lr 0.000465 time 0.2615 (0.2894) loss 3.7167 (3.4978) grad_norm 1.6506 (1.5884) [2021-04-16 06:18:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][1020/1251] eta 0:01:06 lr 0.000465 time 0.2777 (0.2893) loss 3.9315 (3.4991) grad_norm 2.1324 (1.5893) [2021-04-16 06:18:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][1030/1251] eta 0:01:03 lr 0.000464 time 0.2817 (0.2891) loss 3.2858 (3.4965) grad_norm 1.6778 (1.5909) [2021-04-16 06:18:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][1040/1251] eta 0:01:00 lr 0.000464 time 0.2514 (0.2890) loss 2.7833 (3.4968) grad_norm 1.8153 (1.5909) [2021-04-16 06:18:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][1050/1251] eta 0:00:58 lr 0.000464 time 0.3014 (0.2889) loss 2.8357 (3.4957) grad_norm 1.5000 (1.5922) [2021-04-16 06:18:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][1060/1251] eta 0:00:55 lr 0.000464 time 0.2797 (0.2888) loss 3.7942 (3.4952) grad_norm 1.5349 (1.5927) [2021-04-16 06:18:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][1070/1251] eta 0:00:52 lr 0.000464 time 0.2818 (0.2889) loss 2.4858 (3.4958) grad_norm 1.5241 (1.5926) [2021-04-16 06:18:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][1080/1251] eta 0:00:49 lr 0.000464 time 0.2723 (0.2889) loss 2.1580 (3.4951) grad_norm 1.4977 (1.5923) [2021-04-16 06:18:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][1090/1251] eta 0:00:46 lr 0.000464 time 0.2925 (0.2890) loss 3.2670 (3.4948) grad_norm 1.3640 (1.5922) [2021-04-16 06:18:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][1100/1251] eta 0:00:43 lr 0.000464 time 0.2853 (0.2889) loss 3.8588 (3.4962) grad_norm 1.3793 (1.5912) [2021-04-16 06:18:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][1110/1251] eta 0:00:40 lr 0.000464 time 0.2885 (0.2888) loss 3.9917 (3.4950) grad_norm 1.4940 (1.5920) [2021-04-16 06:18:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][1120/1251] eta 0:00:37 lr 0.000464 time 0.2526 (0.2887) loss 3.3575 (3.4911) grad_norm 1.6859 (1.5921) [2021-04-16 06:18:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][1130/1251] eta 0:00:34 lr 0.000464 time 0.2607 (0.2885) loss 2.8120 (3.4911) grad_norm 1.3713 (1.5915) [2021-04-16 06:18:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][1140/1251] eta 0:00:32 lr 0.000464 time 0.2845 (0.2884) loss 3.3306 (3.4911) grad_norm 1.3843 (1.5904) [2021-04-16 06:18:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][1150/1251] eta 0:00:29 lr 0.000464 time 0.3057 (0.2883) loss 3.4584 (3.4913) grad_norm 1.4430 (1.5900) [2021-04-16 06:18:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][1160/1251] eta 0:00:26 lr 0.000464 time 0.2735 (0.2884) loss 3.3250 (3.4920) grad_norm 1.5681 (1.5898) [2021-04-16 06:18:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][1170/1251] eta 0:00:23 lr 0.000464 time 0.2891 (0.2885) loss 3.4454 (3.4935) grad_norm 1.5892 (1.5894) [2021-04-16 06:18:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][1180/1251] eta 0:00:20 lr 0.000464 time 0.2776 (0.2884) loss 3.2780 (3.4934) grad_norm 1.3957 (1.5890) [2021-04-16 06:18:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][1190/1251] eta 0:00:17 lr 0.000464 time 0.3106 (0.2883) loss 3.8194 (3.4962) grad_norm 1.6027 (1.5884) [2021-04-16 06:18:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][1200/1251] eta 0:00:14 lr 0.000464 time 0.2913 (0.2882) loss 2.4415 (3.4941) grad_norm 1.3786 (1.5873) [2021-04-16 06:18:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][1210/1251] eta 0:00:11 lr 0.000464 time 0.2663 (0.2881) loss 4.0953 (3.4949) grad_norm 1.5508 (1.5876) [2021-04-16 06:18:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][1220/1251] eta 0:00:08 lr 0.000464 time 0.2647 (0.2880) loss 3.0614 (3.4954) grad_norm 1.4475 (1.5866) [2021-04-16 06:19:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][1230/1251] eta 0:00:06 lr 0.000464 time 0.2640 (0.2880) loss 3.7180 (3.4939) grad_norm 1.6099 (1.5866) [2021-04-16 06:19:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][1240/1251] eta 0:00:03 lr 0.000464 time 0.2487 (0.2878) loss 3.4504 (3.4937) grad_norm 1.5632 (1.5864) [2021-04-16 06:19:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [157/300][1250/1251] eta 0:00:00 lr 0.000464 time 0.2496 (0.2875) loss 3.8472 (3.4949) grad_norm 1.6376 (1.5876) [2021-04-16 06:19:11 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 157 training takes 0:06:04 [2021-04-16 06:19:11 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_157.pth saving...... [2021-04-16 06:19:34 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_157.pth saved !!! [2021-04-16 06:19:35 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.150 (1.150) Loss 1.0354 (1.0354) Acc@1 75.781 (75.781) Acc@5 93.945 (93.945) [2021-04-16 06:19:36 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.120 (0.201) Loss 1.0390 (1.0227) Acc@1 76.367 (76.669) Acc@5 93.164 (93.368) [2021-04-16 06:19:38 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.155 (0.213) Loss 1.0353 (1.0158) Acc@1 77.148 (76.786) Acc@5 93.262 (93.587) [2021-04-16 06:19:41 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.128 (0.226) Loss 1.0642 (1.0232) Acc@1 74.902 (76.566) Acc@5 93.555 (93.526) [2021-04-16 06:19:43 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.215) Loss 0.9683 (1.0259) Acc@1 78.027 (76.446) Acc@5 94.238 (93.521) [2021-04-16 06:19:48 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 76.192 Acc@5 93.458 [2021-04-16 06:19:48 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 76.2% [2021-04-16 06:19:48 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 76.26% [2021-04-16 06:19:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][0/1251] eta 3:01:01 lr 0.000464 time 8.6823 (8.6823) loss 2.8403 (2.8403) grad_norm 1.4118 (1.4118) [2021-04-16 06:19:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][10/1251] eta 0:21:42 lr 0.000464 time 0.4314 (1.0494) loss 4.3835 (3.6102) grad_norm 1.6744 (1.6259) [2021-04-16 06:20:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][20/1251] eta 0:13:56 lr 0.000463 time 0.2674 (0.6798) loss 2.6406 (3.6262) grad_norm 1.6226 (1.5663) [2021-04-16 06:20:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][30/1251] eta 0:11:17 lr 0.000463 time 0.2607 (0.5551) loss 2.6112 (3.5568) grad_norm 1.3797 (1.5341) [2021-04-16 06:20:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][40/1251] eta 0:09:50 lr 0.000463 time 0.2877 (0.4876) loss 3.7203 (3.4935) grad_norm 1.8830 (1.5822) [2021-04-16 06:20:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][50/1251] eta 0:08:56 lr 0.000463 time 0.2557 (0.4470) loss 3.7480 (3.5268) grad_norm 1.4552 (1.5806) [2021-04-16 06:20:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][60/1251] eta 0:08:21 lr 0.000463 time 0.3711 (0.4210) loss 4.1917 (3.5357) grad_norm 1.4149 (1.5766) [2021-04-16 06:20:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][70/1251] eta 0:07:52 lr 0.000463 time 0.2758 (0.4001) loss 3.7267 (3.5291) grad_norm 1.8809 (1.5776) [2021-04-16 06:20:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][80/1251] eta 0:07:30 lr 0.000463 time 0.2435 (0.3848) loss 3.2479 (3.5440) grad_norm 1.5121 (1.5727) [2021-04-16 06:20:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][90/1251] eta 0:07:15 lr 0.000463 time 0.2551 (0.3748) loss 3.6697 (3.5476) grad_norm 1.7167 (1.5756) [2021-04-16 06:20:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][100/1251] eta 0:06:59 lr 0.000463 time 0.2942 (0.3649) loss 3.3268 (3.5480) grad_norm 1.4717 (1.5784) [2021-04-16 06:20:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][110/1251] eta 0:06:47 lr 0.000463 time 0.2913 (0.3567) loss 3.6381 (3.5340) grad_norm 1.7613 (1.5855) [2021-04-16 06:20:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][120/1251] eta 0:06:36 lr 0.000463 time 0.2809 (0.3503) loss 3.3588 (3.5282) grad_norm 1.6966 (1.5820) [2021-04-16 06:20:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][130/1251] eta 0:06:27 lr 0.000463 time 0.2679 (0.3453) loss 3.5799 (3.5215) grad_norm 1.5533 (1.5850) [2021-04-16 06:20:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][140/1251] eta 0:06:18 lr 0.000463 time 0.2697 (0.3407) loss 2.7643 (3.5025) grad_norm 1.8096 (1.5874) [2021-04-16 06:20:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][150/1251] eta 0:06:11 lr 0.000463 time 0.4152 (0.3373) loss 3.6369 (3.5062) grad_norm 1.6055 (1.5888) [2021-04-16 06:20:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][160/1251] eta 0:06:03 lr 0.000463 time 0.2978 (0.3332) loss 4.1480 (3.5237) grad_norm 1.5909 (1.5861) [2021-04-16 06:20:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][170/1251] eta 0:05:56 lr 0.000463 time 0.2819 (0.3300) loss 3.1092 (3.5134) grad_norm 1.5010 (1.5820) [2021-04-16 06:20:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][180/1251] eta 0:05:50 lr 0.000463 time 0.2910 (0.3274) loss 2.9063 (3.5197) grad_norm 1.4475 (1.5861) [2021-04-16 06:20:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][190/1251] eta 0:05:44 lr 0.000463 time 0.3074 (0.3251) loss 4.0023 (3.5311) grad_norm 1.4547 (1.5947) [2021-04-16 06:20:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][200/1251] eta 0:05:39 lr 0.000463 time 0.2833 (0.3229) loss 2.8685 (3.5235) grad_norm 1.5387 (1.5948) [2021-04-16 06:20:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][210/1251] eta 0:05:33 lr 0.000463 time 0.2647 (0.3207) loss 3.7127 (3.5216) grad_norm 1.5663 (1.5960) [2021-04-16 06:20:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][220/1251] eta 0:05:28 lr 0.000463 time 0.2693 (0.3187) loss 4.2092 (3.5183) grad_norm 1.5629 (1.5993) [2021-04-16 06:21:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][230/1251] eta 0:05:23 lr 0.000463 time 0.2940 (0.3170) loss 3.6713 (3.5274) grad_norm 1.6594 (1.5987) [2021-04-16 06:21:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][240/1251] eta 0:05:18 lr 0.000463 time 0.2674 (0.3154) loss 2.8514 (3.5201) grad_norm 1.5460 (1.5962) [2021-04-16 06:21:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][250/1251] eta 0:05:13 lr 0.000463 time 0.2588 (0.3137) loss 3.5816 (3.5312) grad_norm 1.3597 (1.5922) [2021-04-16 06:21:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][260/1251] eta 0:05:09 lr 0.000463 time 0.2876 (0.3123) loss 3.0643 (3.5209) grad_norm 1.6801 (1.5914) [2021-04-16 06:21:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][270/1251] eta 0:05:04 lr 0.000462 time 0.2696 (0.3109) loss 4.0715 (3.5128) grad_norm 1.7224 (1.5920) [2021-04-16 06:21:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][280/1251] eta 0:05:00 lr 0.000462 time 0.2844 (0.3097) loss 2.8973 (3.5067) grad_norm 1.3963 (1.5898) [2021-04-16 06:21:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][290/1251] eta 0:04:56 lr 0.000462 time 0.2719 (0.3089) loss 3.5744 (3.5020) grad_norm 1.6274 (1.5874) [2021-04-16 06:21:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][300/1251] eta 0:04:52 lr 0.000462 time 0.2670 (0.3080) loss 3.2601 (3.4969) grad_norm 1.5796 (1.5865) [2021-04-16 06:21:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][310/1251] eta 0:04:48 lr 0.000462 time 0.2854 (0.3070) loss 3.3186 (3.5030) grad_norm 1.8836 (1.5863) [2021-04-16 06:21:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][320/1251] eta 0:04:44 lr 0.000462 time 0.2797 (0.3060) loss 4.2103 (3.5084) grad_norm 1.5320 (1.5843) [2021-04-16 06:21:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][330/1251] eta 0:04:41 lr 0.000462 time 0.2919 (0.3054) loss 3.9846 (3.5057) grad_norm 1.5036 (1.5858) [2021-04-16 06:21:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][340/1251] eta 0:04:37 lr 0.000462 time 0.2998 (0.3047) loss 3.6666 (3.5070) grad_norm 1.6582 (1.5877) [2021-04-16 06:21:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][350/1251] eta 0:04:33 lr 0.000462 time 0.2996 (0.3039) loss 2.3257 (3.4955) grad_norm 1.6258 (1.5850) [2021-04-16 06:21:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][360/1251] eta 0:04:30 lr 0.000462 time 0.2871 (0.3035) loss 4.2360 (3.4897) grad_norm 1.9732 (1.5865) [2021-04-16 06:21:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][370/1251] eta 0:04:26 lr 0.000462 time 0.2635 (0.3030) loss 3.7565 (3.4882) grad_norm 1.7874 (1.5928) [2021-04-16 06:21:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][380/1251] eta 0:04:23 lr 0.000462 time 0.2933 (0.3027) loss 3.8319 (3.4871) grad_norm 1.7303 (1.5924) [2021-04-16 06:21:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][390/1251] eta 0:04:20 lr 0.000462 time 0.2732 (0.3021) loss 4.0654 (3.4890) grad_norm 1.5400 (1.5944) [2021-04-16 06:21:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][400/1251] eta 0:04:16 lr 0.000462 time 0.2690 (0.3015) loss 3.7794 (3.4910) grad_norm 1.5315 (1.5958) [2021-04-16 06:21:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][410/1251] eta 0:04:13 lr 0.000462 time 0.2898 (0.3010) loss 3.3881 (3.4910) grad_norm 1.4301 (1.5947) [2021-04-16 06:21:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][420/1251] eta 0:04:09 lr 0.000462 time 0.2597 (0.3003) loss 4.1173 (3.4926) grad_norm 1.7595 (1.5940) [2021-04-16 06:21:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][430/1251] eta 0:04:06 lr 0.000462 time 0.2841 (0.2999) loss 3.3368 (3.4870) grad_norm 1.6686 (1.5932) [2021-04-16 06:22:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][440/1251] eta 0:04:03 lr 0.000462 time 0.2781 (0.2997) loss 2.1489 (3.4872) grad_norm 1.6368 (1.5921) [2021-04-16 06:22:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][450/1251] eta 0:03:59 lr 0.000462 time 0.2643 (0.2990) loss 3.6866 (3.4847) grad_norm 1.4146 (1.5913) [2021-04-16 06:22:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][460/1251] eta 0:03:56 lr 0.000462 time 0.2706 (0.2985) loss 4.1400 (3.4904) grad_norm 1.6028 (1.5918) [2021-04-16 06:22:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][470/1251] eta 0:03:52 lr 0.000462 time 0.2624 (0.2980) loss 2.0483 (3.4834) grad_norm 1.5059 (1.5917) [2021-04-16 06:22:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][480/1251] eta 0:03:49 lr 0.000462 time 0.3041 (0.2976) loss 3.2144 (3.4873) grad_norm 1.4668 (1.5920) [2021-04-16 06:22:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][490/1251] eta 0:03:46 lr 0.000462 time 0.2910 (0.2975) loss 3.5901 (3.4814) grad_norm 1.5336 (1.5919) [2021-04-16 06:22:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][500/1251] eta 0:03:43 lr 0.000462 time 0.2674 (0.2970) loss 3.3706 (3.4789) grad_norm 1.7423 (1.5915) [2021-04-16 06:22:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][510/1251] eta 0:03:39 lr 0.000461 time 0.2832 (0.2966) loss 2.9203 (3.4797) grad_norm 1.4222 (1.5895) [2021-04-16 06:22:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][520/1251] eta 0:03:36 lr 0.000461 time 0.2751 (0.2963) loss 3.3125 (3.4801) grad_norm 1.7982 (1.5911) [2021-04-16 06:22:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][530/1251] eta 0:03:33 lr 0.000461 time 0.2831 (0.2958) loss 3.4809 (3.4828) grad_norm 1.4250 (1.5924) [2021-04-16 06:22:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][540/1251] eta 0:03:30 lr 0.000461 time 0.2805 (0.2955) loss 3.9738 (3.4821) grad_norm 1.7201 (1.5928) [2021-04-16 06:22:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][550/1251] eta 0:03:26 lr 0.000461 time 0.2758 (0.2952) loss 3.3537 (3.4866) grad_norm 1.4360 (1.5933) [2021-04-16 06:22:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][560/1251] eta 0:03:23 lr 0.000461 time 0.2644 (0.2948) loss 4.0435 (3.4907) grad_norm 1.6823 (1.5937) [2021-04-16 06:22:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][570/1251] eta 0:03:20 lr 0.000461 time 0.2587 (0.2944) loss 4.3271 (3.4981) grad_norm 1.5890 (1.5934) [2021-04-16 06:22:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][580/1251] eta 0:03:17 lr 0.000461 time 0.2503 (0.2941) loss 3.6275 (3.4969) grad_norm 1.6123 (1.5931) [2021-04-16 06:22:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][590/1251] eta 0:03:14 lr 0.000461 time 0.2710 (0.2939) loss 3.3313 (3.4973) grad_norm 1.6060 (1.5934) [2021-04-16 06:22:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][600/1251] eta 0:03:11 lr 0.000461 time 0.3893 (0.2938) loss 4.0695 (3.4943) grad_norm 1.5139 (1.5933) [2021-04-16 06:22:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][610/1251] eta 0:03:08 lr 0.000461 time 0.2568 (0.2935) loss 3.8082 (3.4960) grad_norm 1.8138 (1.5940) [2021-04-16 06:22:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][620/1251] eta 0:03:05 lr 0.000461 time 0.2848 (0.2932) loss 3.0451 (3.4920) grad_norm 2.1002 (1.5944) [2021-04-16 06:22:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][630/1251] eta 0:03:02 lr 0.000461 time 0.2970 (0.2931) loss 4.0296 (3.4891) grad_norm 1.4051 (1.5925) [2021-04-16 06:22:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][640/1251] eta 0:02:59 lr 0.000461 time 0.2500 (0.2930) loss 2.5530 (3.4880) grad_norm 1.4384 (1.5920) [2021-04-16 06:22:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][650/1251] eta 0:02:56 lr 0.000461 time 0.2955 (0.2929) loss 4.4077 (3.4864) grad_norm 1.4158 (1.5913) [2021-04-16 06:23:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][660/1251] eta 0:02:53 lr 0.000461 time 0.2622 (0.2928) loss 4.0219 (3.4892) grad_norm 1.8637 (1.5913) [2021-04-16 06:23:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][670/1251] eta 0:02:50 lr 0.000461 time 0.2582 (0.2929) loss 4.0485 (3.4900) grad_norm 1.8946 (1.5928) [2021-04-16 06:23:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][680/1251] eta 0:02:47 lr 0.000461 time 0.2818 (0.2926) loss 3.2393 (3.4902) grad_norm 1.4803 (1.5923) [2021-04-16 06:23:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][690/1251] eta 0:02:44 lr 0.000461 time 0.2601 (0.2924) loss 3.2407 (3.4888) grad_norm 1.7468 (1.5936) [2021-04-16 06:23:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][700/1251] eta 0:02:40 lr 0.000461 time 0.2651 (0.2921) loss 3.8613 (3.4871) grad_norm 1.4869 (1.5936) [2021-04-16 06:23:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][710/1251] eta 0:02:38 lr 0.000461 time 0.3119 (0.2921) loss 3.6483 (3.4852) grad_norm 1.6410 (1.5953) [2021-04-16 06:23:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][720/1251] eta 0:02:34 lr 0.000461 time 0.2775 (0.2918) loss 2.8424 (3.4813) grad_norm 1.5492 (1.5948) [2021-04-16 06:23:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][730/1251] eta 0:02:32 lr 0.000461 time 0.2855 (0.2918) loss 4.0073 (3.4813) grad_norm 1.6027 (1.5947) [2021-04-16 06:23:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][740/1251] eta 0:02:29 lr 0.000461 time 0.2480 (0.2917) loss 3.0230 (3.4814) grad_norm 1.3353 (1.5939) [2021-04-16 06:23:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][750/1251] eta 0:02:25 lr 0.000460 time 0.2666 (0.2914) loss 2.8240 (3.4782) grad_norm 1.5299 (1.5939) [2021-04-16 06:23:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][760/1251] eta 0:02:22 lr 0.000460 time 0.2614 (0.2912) loss 3.9566 (3.4775) grad_norm 1.6830 (1.5938) [2021-04-16 06:23:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][770/1251] eta 0:02:19 lr 0.000460 time 0.2813 (0.2910) loss 3.4533 (3.4786) grad_norm 1.4577 (1.5926) [2021-04-16 06:23:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][780/1251] eta 0:02:16 lr 0.000460 time 0.2561 (0.2907) loss 2.6084 (3.4742) grad_norm 1.4939 (1.5934) [2021-04-16 06:23:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][790/1251] eta 0:02:13 lr 0.000460 time 0.2810 (0.2905) loss 2.8782 (3.4731) grad_norm 1.5381 (1.5940) [2021-04-16 06:23:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][800/1251] eta 0:02:10 lr 0.000460 time 0.2932 (0.2903) loss 4.0309 (3.4739) grad_norm 1.5915 (1.5941) [2021-04-16 06:23:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][810/1251] eta 0:02:07 lr 0.000460 time 0.2592 (0.2902) loss 3.5057 (3.4747) grad_norm 1.5459 (1.5932) [2021-04-16 06:23:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][820/1251] eta 0:02:05 lr 0.000460 time 0.2775 (0.2900) loss 3.9120 (3.4748) grad_norm 1.5236 (1.5936) [2021-04-16 06:23:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][830/1251] eta 0:02:02 lr 0.000460 time 0.2748 (0.2900) loss 3.6594 (3.4778) grad_norm 1.4220 (1.5930) [2021-04-16 06:23:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][840/1251] eta 0:01:59 lr 0.000460 time 0.2689 (0.2899) loss 3.5594 (3.4754) grad_norm 1.5285 (1.5930) [2021-04-16 06:23:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][850/1251] eta 0:01:56 lr 0.000460 time 0.2622 (0.2898) loss 3.2412 (3.4790) grad_norm 1.6560 (1.5932) [2021-04-16 06:23:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][860/1251] eta 0:01:53 lr 0.000460 time 0.2961 (0.2896) loss 3.8949 (3.4788) grad_norm 1.3107 (1.5926) [2021-04-16 06:24:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][870/1251] eta 0:01:50 lr 0.000460 time 0.2864 (0.2896) loss 3.2159 (3.4781) grad_norm 1.5404 (1.5928) [2021-04-16 06:24:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][880/1251] eta 0:01:47 lr 0.000460 time 0.2706 (0.2897) loss 4.1936 (3.4787) grad_norm 1.3660 (1.5921) [2021-04-16 06:24:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][890/1251] eta 0:01:44 lr 0.000460 time 0.2790 (0.2896) loss 2.7753 (3.4794) grad_norm 1.4997 (1.5910) [2021-04-16 06:24:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][900/1251] eta 0:01:41 lr 0.000460 time 0.2510 (0.2894) loss 4.3512 (3.4784) grad_norm 1.4308 (1.5909) [2021-04-16 06:24:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][910/1251] eta 0:01:38 lr 0.000460 time 0.2918 (0.2893) loss 4.0847 (3.4798) grad_norm 1.4077 (1.5911) [2021-04-16 06:24:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][920/1251] eta 0:01:35 lr 0.000460 time 0.2458 (0.2891) loss 3.1143 (3.4773) grad_norm 1.4591 (1.5908) [2021-04-16 06:24:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][930/1251] eta 0:01:32 lr 0.000460 time 0.2729 (0.2889) loss 2.8186 (3.4797) grad_norm 1.7402 (1.5901) [2021-04-16 06:24:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][940/1251] eta 0:01:29 lr 0.000460 time 0.2793 (0.2891) loss 3.9078 (3.4802) grad_norm 1.4702 (1.5896) [2021-04-16 06:24:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][950/1251] eta 0:01:26 lr 0.000460 time 0.2665 (0.2890) loss 2.6669 (3.4804) grad_norm 1.4657 (1.5885) [2021-04-16 06:24:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][960/1251] eta 0:01:24 lr 0.000460 time 0.2837 (0.2890) loss 4.2648 (3.4825) grad_norm 1.5578 (1.5885) [2021-04-16 06:24:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][970/1251] eta 0:01:21 lr 0.000460 time 0.2556 (0.2889) loss 3.4098 (3.4796) grad_norm 1.6481 (1.5893) [2021-04-16 06:24:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][980/1251] eta 0:01:18 lr 0.000460 time 0.2748 (0.2888) loss 2.6760 (3.4781) grad_norm 1.6269 (1.5891) [2021-04-16 06:24:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][990/1251] eta 0:01:15 lr 0.000459 time 0.2735 (0.2887) loss 3.2213 (3.4780) grad_norm 1.5010 (1.5886) [2021-04-16 06:24:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][1000/1251] eta 0:01:12 lr 0.000459 time 0.2761 (0.2885) loss 2.7449 (3.4761) grad_norm 1.5510 (1.5883) [2021-04-16 06:24:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][1010/1251] eta 0:01:09 lr 0.000459 time 0.2832 (0.2885) loss 4.0968 (3.4758) grad_norm 1.3868 (1.5884) [2021-04-16 06:24:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][1020/1251] eta 0:01:06 lr 0.000459 time 0.2575 (0.2884) loss 3.6231 (3.4737) grad_norm 1.5570 (1.5877) [2021-04-16 06:24:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][1030/1251] eta 0:01:03 lr 0.000459 time 0.2680 (0.2883) loss 3.6502 (3.4749) grad_norm 1.5759 (1.5885) [2021-04-16 06:24:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][1040/1251] eta 0:01:00 lr 0.000459 time 0.2539 (0.2882) loss 3.0603 (3.4714) grad_norm 1.5311 (1.5881) [2021-04-16 06:24:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][1050/1251] eta 0:00:57 lr 0.000459 time 0.2749 (0.2880) loss 2.8720 (3.4683) grad_norm 1.6044 (1.5884) [2021-04-16 06:24:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][1060/1251] eta 0:00:54 lr 0.000459 time 0.2624 (0.2879) loss 3.1519 (3.4683) grad_norm 1.4714 (1.5887) [2021-04-16 06:24:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][1070/1251] eta 0:00:52 lr 0.000459 time 0.3094 (0.2881) loss 4.1544 (3.4695) grad_norm 1.7444 (1.5904) [2021-04-16 06:24:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][1080/1251] eta 0:00:49 lr 0.000459 time 0.2643 (0.2880) loss 3.1788 (3.4676) grad_norm 2.0021 (1.5916) [2021-04-16 06:25:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][1090/1251] eta 0:00:46 lr 0.000459 time 0.3044 (0.2878) loss 3.8691 (3.4687) grad_norm 1.5956 (1.5922) [2021-04-16 06:25:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][1100/1251] eta 0:00:43 lr 0.000459 time 0.2632 (0.2877) loss 3.8122 (3.4706) grad_norm 1.5353 (1.5924) [2021-04-16 06:25:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][1110/1251] eta 0:00:40 lr 0.000459 time 0.2957 (0.2876) loss 3.8462 (3.4733) grad_norm 1.5318 (1.5931) [2021-04-16 06:25:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][1120/1251] eta 0:00:37 lr 0.000459 time 0.2691 (0.2876) loss 2.9231 (3.4719) grad_norm 1.8122 (1.5940) [2021-04-16 06:25:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][1130/1251] eta 0:00:34 lr 0.000459 time 0.2730 (0.2875) loss 3.6889 (3.4724) grad_norm 1.6830 (1.5941) [2021-04-16 06:25:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][1140/1251] eta 0:00:31 lr 0.000459 time 0.2887 (0.2874) loss 3.5375 (3.4732) grad_norm 1.5998 (1.5941) [2021-04-16 06:25:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][1150/1251] eta 0:00:29 lr 0.000459 time 0.2891 (0.2873) loss 2.8858 (3.4725) grad_norm 1.5827 (1.5950) [2021-04-16 06:25:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][1160/1251] eta 0:00:26 lr 0.000459 time 0.2434 (0.2872) loss 2.6808 (3.4713) grad_norm 2.0049 (1.5967) [2021-04-16 06:25:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][1170/1251] eta 0:00:23 lr 0.000459 time 0.2619 (0.2872) loss 3.8583 (3.4726) grad_norm 1.5987 (1.5969) [2021-04-16 06:25:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][1180/1251] eta 0:00:20 lr 0.000459 time 0.2965 (0.2873) loss 3.7962 (3.4744) grad_norm 1.5232 (1.5972) [2021-04-16 06:25:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][1190/1251] eta 0:00:17 lr 0.000459 time 0.2761 (0.2871) loss 3.5548 (3.4754) grad_norm 1.5708 (1.5975) [2021-04-16 06:25:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][1200/1251] eta 0:00:14 lr 0.000459 time 0.2894 (0.2871) loss 4.5732 (3.4770) grad_norm 1.4766 (1.5974) [2021-04-16 06:25:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][1210/1251] eta 0:00:11 lr 0.000459 time 0.2610 (0.2870) loss 4.0222 (3.4759) grad_norm 1.5442 (1.5969) [2021-04-16 06:25:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][1220/1251] eta 0:00:08 lr 0.000459 time 0.2711 (0.2870) loss 3.8331 (3.4774) grad_norm 1.6538 (1.5969) [2021-04-16 06:25:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][1230/1251] eta 0:00:06 lr 0.000459 time 0.2564 (0.2869) loss 3.7835 (3.4770) grad_norm 1.7133 (1.5971) [2021-04-16 06:25:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][1240/1251] eta 0:00:03 lr 0.000458 time 0.2489 (0.2868) loss 3.6562 (3.4788) grad_norm 1.5463 (1.5977) [2021-04-16 06:25:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [158/300][1250/1251] eta 0:00:00 lr 0.000458 time 0.2484 (0.2865) loss 3.4305 (3.4801) grad_norm 1.5984 (1.5984) [2021-04-16 06:25:50 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 158 training takes 0:06:02 [2021-04-16 06:25:50 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_158.pth saving...... [2021-04-16 06:26:04 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_158.pth saved !!! [2021-04-16 06:26:05 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.125 (1.125) Loss 1.0005 (1.0005) Acc@1 76.172 (76.172) Acc@5 93.848 (93.848) [2021-04-16 06:26:06 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.178 (0.228) Loss 1.0507 (0.9947) Acc@1 75.879 (76.545) Acc@5 92.090 (93.404) [2021-04-16 06:26:09 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.138 (0.232) Loss 0.9194 (0.9925) Acc@1 78.320 (76.488) Acc@5 93.848 (93.517) [2021-04-16 06:26:12 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.192 (0.244) Loss 0.9403 (0.9938) Acc@1 76.855 (76.333) Acc@5 93.945 (93.539) [2021-04-16 06:26:13 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.215) Loss 1.0166 (0.9937) Acc@1 76.074 (76.281) Acc@5 93.359 (93.624) [2021-04-16 06:26:19 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 76.260 Acc@5 93.584 [2021-04-16 06:26:19 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 76.3% [2021-04-16 06:26:19 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 76.26% [2021-04-16 06:26:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][0/1251] eta 2:38:00 lr 0.000458 time 7.5787 (7.5787) loss 3.9608 (3.9608) grad_norm 1.3816 (1.3816) [2021-04-16 06:26:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][10/1251] eta 0:19:27 lr 0.000458 time 0.2766 (0.9406) loss 3.5091 (3.5465) grad_norm 1.6527 (1.6155) [2021-04-16 06:26:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][20/1251] eta 0:12:47 lr 0.000458 time 0.2788 (0.6237) loss 3.9544 (3.4321) grad_norm 1.4689 (1.6251) [2021-04-16 06:26:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][30/1251] eta 0:10:28 lr 0.000458 time 0.2828 (0.5145) loss 3.8387 (3.3676) grad_norm 1.4363 (1.6322) [2021-04-16 06:26:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][40/1251] eta 0:09:13 lr 0.000458 time 0.2839 (0.4569) loss 2.9985 (3.3558) grad_norm 1.8169 (1.6269) [2021-04-16 06:26:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][50/1251] eta 0:08:27 lr 0.000458 time 0.2683 (0.4229) loss 3.8373 (3.3569) grad_norm 1.3964 (1.6259) [2021-04-16 06:26:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][60/1251] eta 0:07:56 lr 0.000458 time 0.2782 (0.3997) loss 3.4715 (3.4473) grad_norm 1.7497 (1.6216) [2021-04-16 06:26:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][70/1251] eta 0:07:32 lr 0.000458 time 0.2885 (0.3833) loss 3.5579 (3.4800) grad_norm 1.6352 (1.6243) [2021-04-16 06:26:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][80/1251] eta 0:07:15 lr 0.000458 time 0.2557 (0.3720) loss 2.7035 (3.4663) grad_norm 1.3627 (1.6265) [2021-04-16 06:26:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][90/1251] eta 0:07:00 lr 0.000458 time 0.2730 (0.3622) loss 3.2069 (3.4324) grad_norm 1.4258 (1.6222) [2021-04-16 06:26:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][100/1251] eta 0:06:47 lr 0.000458 time 0.2684 (0.3541) loss 3.5990 (3.4611) grad_norm 1.4661 (1.6228) [2021-04-16 06:26:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][110/1251] eta 0:06:36 lr 0.000458 time 0.2955 (0.3472) loss 3.8521 (3.4604) grad_norm 1.8510 (1.6162) [2021-04-16 06:27:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][120/1251] eta 0:06:26 lr 0.000458 time 0.2986 (0.3414) loss 3.5552 (3.4506) grad_norm 1.4355 (1.6119) [2021-04-16 06:27:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][130/1251] eta 0:06:17 lr 0.000458 time 0.2792 (0.3367) loss 3.3518 (3.4511) grad_norm 1.4232 (1.6103) [2021-04-16 06:27:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][140/1251] eta 0:06:10 lr 0.000458 time 0.2766 (0.3337) loss 3.5590 (3.4364) grad_norm 1.5249 (1.6054) [2021-04-16 06:27:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][150/1251] eta 0:06:03 lr 0.000458 time 0.2768 (0.3305) loss 3.8652 (3.4153) grad_norm 1.4574 (1.6032) [2021-04-16 06:27:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][160/1251] eta 0:05:57 lr 0.000458 time 0.2684 (0.3277) loss 3.2415 (3.4213) grad_norm 1.7901 (1.6084) [2021-04-16 06:27:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][170/1251] eta 0:05:51 lr 0.000458 time 0.2717 (0.3251) loss 4.2479 (3.4217) grad_norm 1.4296 (1.6083) [2021-04-16 06:27:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][180/1251] eta 0:05:45 lr 0.000458 time 0.2781 (0.3223) loss 2.5704 (3.4117) grad_norm 1.6486 (1.6043) [2021-04-16 06:27:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][190/1251] eta 0:05:39 lr 0.000458 time 0.3070 (0.3199) loss 3.9358 (3.4248) grad_norm 1.6499 (1.6097) [2021-04-16 06:27:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][200/1251] eta 0:05:34 lr 0.000458 time 0.2765 (0.3179) loss 3.4279 (3.4195) grad_norm 1.4886 (1.6097) [2021-04-16 06:27:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][210/1251] eta 0:05:29 lr 0.000458 time 0.2871 (0.3161) loss 2.5624 (3.4070) grad_norm 1.3349 (1.6085) [2021-04-16 06:27:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][220/1251] eta 0:05:24 lr 0.000458 time 0.2747 (0.3149) loss 3.6963 (3.4156) grad_norm 1.8623 (1.6110) [2021-04-16 06:27:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][230/1251] eta 0:05:20 lr 0.000457 time 0.2848 (0.3137) loss 3.5865 (3.4156) grad_norm 1.6248 (1.6127) [2021-04-16 06:27:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][240/1251] eta 0:05:15 lr 0.000457 time 0.2674 (0.3124) loss 2.7628 (3.4210) grad_norm 1.8365 (1.6140) [2021-04-16 06:27:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][250/1251] eta 0:05:11 lr 0.000457 time 0.2876 (0.3108) loss 3.6155 (3.4271) grad_norm 1.6182 (1.6185) [2021-04-16 06:27:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][260/1251] eta 0:05:06 lr 0.000457 time 0.2710 (0.3095) loss 3.6158 (3.4268) grad_norm 1.4471 (1.6206) [2021-04-16 06:27:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][270/1251] eta 0:05:02 lr 0.000457 time 0.2812 (0.3085) loss 4.0832 (3.4317) grad_norm 1.4546 (1.6196) [2021-04-16 06:27:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][280/1251] eta 0:04:58 lr 0.000457 time 0.2946 (0.3074) loss 4.2999 (3.4320) grad_norm 1.5821 (1.6211) [2021-04-16 06:27:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][290/1251] eta 0:04:54 lr 0.000457 time 0.2927 (0.3068) loss 3.6761 (3.4269) grad_norm 1.5761 (1.6217) [2021-04-16 06:27:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][300/1251] eta 0:04:50 lr 0.000457 time 0.2639 (0.3058) loss 3.8168 (3.4341) grad_norm 1.5326 (inf) [2021-04-16 06:27:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][310/1251] eta 0:04:46 lr 0.000457 time 0.2722 (0.3048) loss 3.7785 (3.4411) grad_norm 1.5062 (inf) [2021-04-16 06:27:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][320/1251] eta 0:04:42 lr 0.000457 time 0.2624 (0.3038) loss 3.9139 (3.4429) grad_norm 1.5876 (inf) [2021-04-16 06:27:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][330/1251] eta 0:04:38 lr 0.000457 time 0.2580 (0.3029) loss 3.1215 (3.4446) grad_norm 1.5522 (inf) [2021-04-16 06:28:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][340/1251] eta 0:04:35 lr 0.000457 time 0.2781 (0.3021) loss 3.4402 (3.4484) grad_norm 1.7816 (inf) [2021-04-16 06:28:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][350/1251] eta 0:04:31 lr 0.000457 time 0.2914 (0.3013) loss 3.4656 (3.4559) grad_norm 1.6090 (inf) [2021-04-16 06:28:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][360/1251] eta 0:04:27 lr 0.000457 time 0.2870 (0.3005) loss 3.3931 (3.4575) grad_norm 1.5463 (inf) [2021-04-16 06:28:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][370/1251] eta 0:04:24 lr 0.000457 time 0.2778 (0.3002) loss 4.1548 (3.4527) grad_norm 1.9998 (inf) [2021-04-16 06:28:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][380/1251] eta 0:04:21 lr 0.000457 time 0.2754 (0.2999) loss 3.5838 (3.4583) grad_norm 1.6904 (inf) [2021-04-16 06:28:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][390/1251] eta 0:04:17 lr 0.000457 time 0.2742 (0.2993) loss 3.4952 (3.4600) grad_norm 1.6162 (inf) [2021-04-16 06:28:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][400/1251] eta 0:04:14 lr 0.000457 time 0.2732 (0.2987) loss 3.9097 (3.4604) grad_norm 1.4835 (inf) [2021-04-16 06:28:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][410/1251] eta 0:04:10 lr 0.000457 time 0.2715 (0.2981) loss 3.8645 (3.4683) grad_norm 1.7081 (inf) [2021-04-16 06:28:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][420/1251] eta 0:04:07 lr 0.000457 time 0.3096 (0.2980) loss 4.0288 (3.4687) grad_norm 1.6228 (inf) [2021-04-16 06:28:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][430/1251] eta 0:04:04 lr 0.000457 time 0.2697 (0.2974) loss 3.9498 (3.4702) grad_norm 1.4622 (inf) [2021-04-16 06:28:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][440/1251] eta 0:04:00 lr 0.000457 time 0.2702 (0.2970) loss 3.7568 (3.4700) grad_norm 1.7166 (inf) [2021-04-16 06:28:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][450/1251] eta 0:03:57 lr 0.000457 time 0.3070 (0.2968) loss 3.9281 (3.4698) grad_norm 1.6263 (inf) [2021-04-16 06:28:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][460/1251] eta 0:03:54 lr 0.000457 time 0.2622 (0.2963) loss 4.2749 (3.4673) grad_norm 1.5999 (inf) [2021-04-16 06:28:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][470/1251] eta 0:03:51 lr 0.000456 time 0.2766 (0.2960) loss 4.1790 (3.4635) grad_norm 1.6675 (inf) [2021-04-16 06:28:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][480/1251] eta 0:03:47 lr 0.000456 time 0.2978 (0.2956) loss 3.9182 (3.4573) grad_norm 1.4683 (inf) [2021-04-16 06:28:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][490/1251] eta 0:03:44 lr 0.000456 time 0.3118 (0.2953) loss 4.0708 (3.4633) grad_norm 2.0916 (inf) [2021-04-16 06:28:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][500/1251] eta 0:03:41 lr 0.000456 time 0.2977 (0.2950) loss 2.7517 (3.4614) grad_norm 1.5763 (inf) [2021-04-16 06:28:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][510/1251] eta 0:03:38 lr 0.000456 time 0.2851 (0.2945) loss 3.8297 (3.4684) grad_norm 1.7737 (inf) [2021-04-16 06:28:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][520/1251] eta 0:03:35 lr 0.000456 time 0.2622 (0.2941) loss 3.9666 (3.4668) grad_norm 1.8209 (inf) [2021-04-16 06:28:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][530/1251] eta 0:03:31 lr 0.000456 time 0.2593 (0.2940) loss 3.9875 (3.4718) grad_norm 1.7474 (inf) [2021-04-16 06:28:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][540/1251] eta 0:03:28 lr 0.000456 time 0.2754 (0.2938) loss 3.7093 (3.4766) grad_norm 1.5212 (inf) [2021-04-16 06:29:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][550/1251] eta 0:03:25 lr 0.000456 time 0.2732 (0.2936) loss 3.7632 (3.4740) grad_norm 1.6438 (inf) [2021-04-16 06:29:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][560/1251] eta 0:03:22 lr 0.000456 time 0.2580 (0.2932) loss 3.1233 (3.4765) grad_norm 1.6377 (inf) [2021-04-16 06:29:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][570/1251] eta 0:03:19 lr 0.000456 time 0.2652 (0.2930) loss 4.0301 (3.4794) grad_norm 1.7695 (inf) [2021-04-16 06:29:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][580/1251] eta 0:03:16 lr 0.000456 time 0.2884 (0.2928) loss 4.3427 (3.4822) grad_norm 1.6181 (inf) [2021-04-16 06:29:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][590/1251] eta 0:03:13 lr 0.000456 time 0.2988 (0.2925) loss 3.4259 (3.4820) grad_norm 1.6933 (inf) [2021-04-16 06:29:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][600/1251] eta 0:03:10 lr 0.000456 time 0.2616 (0.2921) loss 2.8680 (3.4817) grad_norm 1.8342 (inf) [2021-04-16 06:29:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][610/1251] eta 0:03:07 lr 0.000456 time 0.2854 (0.2919) loss 3.4489 (3.4753) grad_norm 1.7479 (inf) [2021-04-16 06:29:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][620/1251] eta 0:03:04 lr 0.000456 time 0.2915 (0.2918) loss 2.5046 (3.4758) grad_norm 1.6204 (inf) [2021-04-16 06:29:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][630/1251] eta 0:03:01 lr 0.000456 time 0.2918 (0.2915) loss 3.7638 (3.4820) grad_norm 1.4581 (inf) [2021-04-16 06:29:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][640/1251] eta 0:02:58 lr 0.000456 time 0.2812 (0.2914) loss 3.3802 (3.4773) grad_norm 1.4538 (inf) [2021-04-16 06:29:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][650/1251] eta 0:02:55 lr 0.000456 time 0.2550 (0.2913) loss 4.7557 (3.4782) grad_norm 1.7114 (inf) [2021-04-16 06:29:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][660/1251] eta 0:02:52 lr 0.000456 time 0.2877 (0.2910) loss 3.9091 (3.4765) grad_norm 1.4768 (inf) [2021-04-16 06:29:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][670/1251] eta 0:02:49 lr 0.000456 time 0.2684 (0.2909) loss 4.1273 (3.4757) grad_norm 1.5826 (inf) [2021-04-16 06:29:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][680/1251] eta 0:02:46 lr 0.000456 time 0.2992 (0.2908) loss 2.9329 (3.4728) grad_norm 1.5627 (inf) [2021-04-16 06:29:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][690/1251] eta 0:02:43 lr 0.000456 time 0.3032 (0.2907) loss 3.7111 (3.4725) grad_norm 1.4736 (inf) [2021-04-16 06:29:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][700/1251] eta 0:02:40 lr 0.000456 time 0.2689 (0.2905) loss 3.6035 (3.4711) grad_norm 1.4350 (inf) [2021-04-16 06:29:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][710/1251] eta 0:02:37 lr 0.000455 time 0.2728 (0.2903) loss 3.4765 (3.4709) grad_norm 1.4138 (inf) [2021-04-16 06:29:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][720/1251] eta 0:02:34 lr 0.000455 time 0.2602 (0.2902) loss 2.2863 (3.4709) grad_norm 1.7827 (inf) [2021-04-16 06:29:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][730/1251] eta 0:02:31 lr 0.000455 time 0.2442 (0.2904) loss 4.0489 (3.4695) grad_norm 1.7112 (inf) [2021-04-16 06:29:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][740/1251] eta 0:02:28 lr 0.000455 time 0.2646 (0.2901) loss 4.1929 (3.4689) grad_norm 1.7670 (inf) [2021-04-16 06:29:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][750/1251] eta 0:02:25 lr 0.000455 time 0.2832 (0.2900) loss 2.1877 (3.4685) grad_norm 1.5267 (inf) [2021-04-16 06:30:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][760/1251] eta 0:02:22 lr 0.000455 time 0.2704 (0.2898) loss 3.6285 (3.4711) grad_norm 1.6774 (inf) [2021-04-16 06:30:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][770/1251] eta 0:02:19 lr 0.000455 time 0.2744 (0.2896) loss 3.5285 (3.4714) grad_norm 1.5596 (inf) [2021-04-16 06:30:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][780/1251] eta 0:02:16 lr 0.000455 time 0.2858 (0.2894) loss 2.3805 (3.4706) grad_norm 1.5551 (inf) [2021-04-16 06:30:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][790/1251] eta 0:02:13 lr 0.000455 time 0.3076 (0.2894) loss 3.1715 (3.4702) grad_norm 1.5050 (inf) [2021-04-16 06:30:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][800/1251] eta 0:02:10 lr 0.000455 time 0.2934 (0.2893) loss 4.5510 (3.4740) grad_norm 1.5206 (inf) [2021-04-16 06:30:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][810/1251] eta 0:02:07 lr 0.000455 time 0.2830 (0.2892) loss 3.1555 (3.4735) grad_norm 1.4503 (inf) [2021-04-16 06:30:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][820/1251] eta 0:02:04 lr 0.000455 time 0.2628 (0.2890) loss 3.5914 (3.4726) grad_norm 1.5551 (inf) [2021-04-16 06:30:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][830/1251] eta 0:02:01 lr 0.000455 time 0.3100 (0.2888) loss 2.7885 (3.4730) grad_norm 1.7015 (inf) [2021-04-16 06:30:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][840/1251] eta 0:01:58 lr 0.000455 time 0.2799 (0.2887) loss 3.0043 (3.4725) grad_norm 1.6011 (inf) [2021-04-16 06:30:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][850/1251] eta 0:01:55 lr 0.000455 time 0.2632 (0.2885) loss 2.7171 (3.4700) grad_norm 1.4618 (inf) [2021-04-16 06:30:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][860/1251] eta 0:01:52 lr 0.000455 time 0.2798 (0.2884) loss 3.4439 (3.4697) grad_norm 1.4449 (inf) [2021-04-16 06:30:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][870/1251] eta 0:01:49 lr 0.000455 time 0.3130 (0.2885) loss 2.6077 (3.4706) grad_norm 1.7160 (inf) [2021-04-16 06:30:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][880/1251] eta 0:01:46 lr 0.000455 time 0.2683 (0.2883) loss 2.8682 (3.4692) grad_norm 1.4990 (inf) [2021-04-16 06:30:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][890/1251] eta 0:01:44 lr 0.000455 time 0.2758 (0.2882) loss 4.5159 (3.4714) grad_norm 1.7136 (inf) [2021-04-16 06:30:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][900/1251] eta 0:01:41 lr 0.000455 time 0.2810 (0.2881) loss 3.9563 (3.4730) grad_norm 1.6215 (inf) [2021-04-16 06:30:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][910/1251] eta 0:01:38 lr 0.000455 time 0.2684 (0.2880) loss 2.4516 (3.4746) grad_norm 1.4309 (inf) [2021-04-16 06:30:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][920/1251] eta 0:01:35 lr 0.000455 time 0.2988 (0.2879) loss 3.9749 (3.4766) grad_norm 1.4757 (inf) [2021-04-16 06:30:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][930/1251] eta 0:01:32 lr 0.000455 time 0.2665 (0.2878) loss 3.1566 (3.4779) grad_norm 1.7608 (inf) [2021-04-16 06:30:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][940/1251] eta 0:01:29 lr 0.000455 time 0.2770 (0.2878) loss 3.5258 (3.4788) grad_norm 1.3414 (inf) [2021-04-16 06:30:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][950/1251] eta 0:01:26 lr 0.000454 time 0.2439 (0.2876) loss 3.1018 (3.4837) grad_norm 1.5615 (inf) [2021-04-16 06:30:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][960/1251] eta 0:01:23 lr 0.000454 time 0.2934 (0.2877) loss 3.3260 (3.4814) grad_norm 1.7190 (inf) [2021-04-16 06:30:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][970/1251] eta 0:01:20 lr 0.000454 time 0.2661 (0.2875) loss 3.1525 (3.4797) grad_norm 1.5206 (inf) [2021-04-16 06:31:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][980/1251] eta 0:01:17 lr 0.000454 time 0.2764 (0.2875) loss 2.7695 (3.4787) grad_norm 1.8861 (inf) [2021-04-16 06:31:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][990/1251] eta 0:01:15 lr 0.000454 time 0.2664 (0.2875) loss 3.9746 (3.4765) grad_norm 1.8117 (inf) [2021-04-16 06:31:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][1000/1251] eta 0:01:12 lr 0.000454 time 0.2434 (0.2874) loss 3.4889 (3.4752) grad_norm 1.4928 (inf) [2021-04-16 06:31:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][1010/1251] eta 0:01:09 lr 0.000454 time 0.2859 (0.2873) loss 3.2851 (3.4745) grad_norm 1.5565 (inf) [2021-04-16 06:31:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][1020/1251] eta 0:01:06 lr 0.000454 time 0.2678 (0.2873) loss 3.0235 (3.4717) grad_norm 1.5560 (inf) [2021-04-16 06:31:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][1030/1251] eta 0:01:03 lr 0.000454 time 0.2767 (0.2872) loss 3.5164 (3.4734) grad_norm 1.4095 (inf) [2021-04-16 06:31:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][1040/1251] eta 0:01:00 lr 0.000454 time 0.2875 (0.2871) loss 3.6763 (3.4719) grad_norm 1.7047 (inf) [2021-04-16 06:31:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][1050/1251] eta 0:00:57 lr 0.000454 time 0.2625 (0.2870) loss 3.5707 (3.4733) grad_norm 1.6210 (inf) [2021-04-16 06:31:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][1060/1251] eta 0:00:54 lr 0.000454 time 0.3068 (0.2869) loss 4.6077 (3.4757) grad_norm 1.7290 (inf) [2021-04-16 06:31:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][1070/1251] eta 0:00:51 lr 0.000454 time 0.2731 (0.2868) loss 4.0563 (3.4777) grad_norm 1.5067 (inf) [2021-04-16 06:31:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][1080/1251] eta 0:00:49 lr 0.000454 time 0.2824 (0.2867) loss 2.7355 (3.4799) grad_norm 1.5071 (inf) [2021-04-16 06:31:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][1090/1251] eta 0:00:46 lr 0.000454 time 0.2959 (0.2866) loss 3.7097 (3.4794) grad_norm 1.4697 (inf) [2021-04-16 06:31:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][1100/1251] eta 0:00:43 lr 0.000454 time 0.2935 (0.2864) loss 3.5190 (3.4778) grad_norm 1.7518 (inf) [2021-04-16 06:31:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][1110/1251] eta 0:00:40 lr 0.000454 time 0.2623 (0.2864) loss 3.5397 (3.4777) grad_norm 2.2272 (inf) [2021-04-16 06:31:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][1120/1251] eta 0:00:37 lr 0.000454 time 0.2671 (0.2863) loss 3.9862 (3.4770) grad_norm 1.6352 (inf) [2021-04-16 06:31:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][1130/1251] eta 0:00:34 lr 0.000454 time 0.2563 (0.2862) loss 3.2141 (3.4775) grad_norm 1.6319 (inf) [2021-04-16 06:31:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][1140/1251] eta 0:00:31 lr 0.000454 time 0.2793 (0.2862) loss 2.8388 (3.4789) grad_norm 1.9896 (inf) [2021-04-16 06:31:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][1150/1251] eta 0:00:28 lr 0.000454 time 0.2933 (0.2862) loss 3.6909 (3.4786) grad_norm 1.8402 (inf) [2021-04-16 06:31:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][1160/1251] eta 0:00:26 lr 0.000454 time 0.2798 (0.2861) loss 4.1940 (3.4796) grad_norm 1.6448 (inf) [2021-04-16 06:31:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][1170/1251] eta 0:00:23 lr 0.000454 time 0.2699 (0.2860) loss 2.7277 (3.4800) grad_norm 2.0843 (inf) [2021-04-16 06:31:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][1180/1251] eta 0:00:20 lr 0.000454 time 0.2980 (0.2859) loss 3.8865 (3.4802) grad_norm 1.4215 (inf) [2021-04-16 06:31:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][1190/1251] eta 0:00:17 lr 0.000454 time 0.2922 (0.2858) loss 2.6393 (3.4827) grad_norm 2.0461 (inf) [2021-04-16 06:32:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][1200/1251] eta 0:00:14 lr 0.000453 time 0.2995 (0.2857) loss 4.1881 (3.4830) grad_norm 1.6037 (inf) [2021-04-16 06:32:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][1210/1251] eta 0:00:11 lr 0.000453 time 0.3000 (0.2857) loss 4.0477 (3.4834) grad_norm 1.8160 (inf) [2021-04-16 06:32:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][1220/1251] eta 0:00:08 lr 0.000453 time 0.2658 (0.2856) loss 4.6013 (3.4836) grad_norm 1.6201 (inf) [2021-04-16 06:32:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][1230/1251] eta 0:00:05 lr 0.000453 time 0.2735 (0.2856) loss 3.3878 (3.4832) grad_norm 1.4845 (inf) [2021-04-16 06:32:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][1240/1251] eta 0:00:03 lr 0.000453 time 0.2612 (0.2855) loss 3.3691 (3.4835) grad_norm 1.4521 (inf) [2021-04-16 06:32:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [159/300][1250/1251] eta 0:00:00 lr 0.000453 time 0.2481 (0.2852) loss 2.9638 (3.4830) grad_norm 1.8978 (inf) [2021-04-16 06:32:21 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 159 training takes 0:06:01 [2021-04-16 06:32:21 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_159.pth saving...... [2021-04-16 06:32:39 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_159.pth saved !!! [2021-04-16 06:32:41 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.377 (1.377) Loss 1.0164 (1.0164) Acc@1 76.172 (76.172) Acc@5 93.555 (93.555) [2021-04-16 06:32:42 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.147 (0.251) Loss 0.9947 (1.0126) Acc@1 77.539 (76.349) Acc@5 92.578 (93.049) [2021-04-16 06:32:45 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.202 (0.247) Loss 0.9532 (1.0074) Acc@1 79.199 (76.469) Acc@5 93.262 (93.276) [2021-04-16 06:32:47 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.109 (0.246) Loss 1.0328 (1.0045) Acc@1 76.074 (76.449) Acc@5 93.164 (93.391) [2021-04-16 06:32:48 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.211) Loss 1.0966 (1.0062) Acc@1 74.902 (76.403) Acc@5 92.773 (93.390) [2021-04-16 06:33:02 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 76.342 Acc@5 93.386 [2021-04-16 06:33:02 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 76.3% [2021-04-16 06:33:02 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 76.34% [2021-04-16 06:33:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][0/1251] eta 0:43:19 lr 0.000453 time 2.0782 (2.0782) loss 2.5264 (2.5264) grad_norm 1.5760 (1.5760) [2021-04-16 06:33:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][10/1251] eta 0:09:03 lr 0.000453 time 0.2900 (0.4383) loss 3.3674 (3.2912) grad_norm 1.3251 (1.5877) [2021-04-16 06:33:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][20/1251] eta 0:07:35 lr 0.000453 time 0.2923 (0.3702) loss 3.7901 (3.3611) grad_norm 1.4327 (1.6380) [2021-04-16 06:33:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][30/1251] eta 0:06:56 lr 0.000453 time 0.2951 (0.3411) loss 3.1390 (3.2963) grad_norm 1.2681 (1.6089) [2021-04-16 06:33:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][40/1251] eta 0:06:35 lr 0.000453 time 0.2850 (0.3263) loss 3.9100 (3.3222) grad_norm 1.7422 (1.6292) [2021-04-16 06:33:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][50/1251] eta 0:06:20 lr 0.000453 time 0.2840 (0.3170) loss 3.6437 (3.3583) grad_norm 1.9173 (1.6265) [2021-04-16 06:33:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][60/1251] eta 0:06:13 lr 0.000453 time 0.4661 (0.3139) loss 4.2476 (3.3382) grad_norm 1.5128 (1.6120) [2021-04-16 06:33:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][70/1251] eta 0:06:05 lr 0.000453 time 0.3155 (0.3093) loss 2.7307 (3.3924) grad_norm 1.4958 (1.5982) [2021-04-16 06:33:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][80/1251] eta 0:05:58 lr 0.000453 time 0.2855 (0.3063) loss 2.4269 (3.4031) grad_norm 1.5889 (1.5864) [2021-04-16 06:33:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][90/1251] eta 0:05:52 lr 0.000453 time 0.2692 (0.3034) loss 3.5019 (3.4243) grad_norm 1.5420 (1.5697) [2021-04-16 06:33:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][100/1251] eta 0:05:47 lr 0.000453 time 0.2666 (0.3015) loss 2.7919 (3.4319) grad_norm 1.5982 (1.5774) [2021-04-16 06:33:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][110/1251] eta 0:05:41 lr 0.000453 time 0.2611 (0.2995) loss 3.0814 (3.4291) grad_norm 1.4734 (1.5837) [2021-04-16 06:33:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][120/1251] eta 0:05:36 lr 0.000453 time 0.2905 (0.2978) loss 4.0636 (3.4457) grad_norm 1.3515 (1.5761) [2021-04-16 06:33:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][130/1251] eta 0:05:31 lr 0.000453 time 0.2517 (0.2961) loss 3.9189 (3.4645) grad_norm 1.8454 (1.5748) [2021-04-16 06:33:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][140/1251] eta 0:05:28 lr 0.000453 time 0.2644 (0.2960) loss 4.0847 (3.4658) grad_norm 1.4656 (1.5770) [2021-04-16 06:33:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][150/1251] eta 0:05:24 lr 0.000453 time 0.2920 (0.2949) loss 3.4179 (3.4900) grad_norm 1.9657 (1.5859) [2021-04-16 06:33:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][160/1251] eta 0:05:22 lr 0.000453 time 0.2871 (0.2954) loss 4.0791 (3.4874) grad_norm 1.5817 (1.5864) [2021-04-16 06:33:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][170/1251] eta 0:05:19 lr 0.000453 time 0.2847 (0.2954) loss 3.8277 (3.4959) grad_norm 1.7044 (1.5906) [2021-04-16 06:33:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][180/1251] eta 0:05:15 lr 0.000453 time 0.2614 (0.2943) loss 2.9047 (3.4957) grad_norm 1.5827 (1.5902) [2021-04-16 06:33:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][190/1251] eta 0:05:11 lr 0.000452 time 0.3081 (0.2935) loss 2.1380 (3.4819) grad_norm 1.5823 (1.5866) [2021-04-16 06:34:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][200/1251] eta 0:05:08 lr 0.000452 time 0.2473 (0.2933) loss 2.6929 (3.4783) grad_norm 2.0387 (1.5898) [2021-04-16 06:34:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][210/1251] eta 0:05:04 lr 0.000452 time 0.2810 (0.2925) loss 3.5880 (3.4939) grad_norm 2.0355 (1.5943) [2021-04-16 06:34:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][220/1251] eta 0:05:00 lr 0.000452 time 0.2666 (0.2918) loss 3.2476 (3.4828) grad_norm 1.5486 (1.6002) [2021-04-16 06:34:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][230/1251] eta 0:04:57 lr 0.000452 time 0.2865 (0.2917) loss 4.1989 (3.4965) grad_norm 1.7040 (1.6033) [2021-04-16 06:34:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][240/1251] eta 0:04:54 lr 0.000452 time 0.2528 (0.2911) loss 2.2951 (3.4928) grad_norm 1.4908 (1.6043) [2021-04-16 06:34:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][250/1251] eta 0:04:50 lr 0.000452 time 0.2768 (0.2907) loss 2.9778 (3.4894) grad_norm 1.5211 (1.6060) [2021-04-16 06:34:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][260/1251] eta 0:04:47 lr 0.000452 time 0.2616 (0.2903) loss 4.2796 (3.4853) grad_norm 1.6239 (1.6053) [2021-04-16 06:34:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][270/1251] eta 0:04:44 lr 0.000452 time 0.2796 (0.2899) loss 3.8261 (3.4905) grad_norm 1.5831 (1.6041) [2021-04-16 06:34:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][280/1251] eta 0:04:41 lr 0.000452 time 0.2792 (0.2897) loss 3.6002 (3.4975) grad_norm 1.6963 (1.6034) [2021-04-16 06:34:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][290/1251] eta 0:04:38 lr 0.000452 time 0.2910 (0.2894) loss 4.3103 (3.4995) grad_norm 1.8025 (1.6055) [2021-04-16 06:34:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][300/1251] eta 0:04:34 lr 0.000452 time 0.2642 (0.2890) loss 3.9917 (3.5141) grad_norm 1.7582 (1.6071) [2021-04-16 06:34:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][310/1251] eta 0:04:31 lr 0.000452 time 0.3109 (0.2887) loss 4.0219 (3.5048) grad_norm 1.4722 (1.6071) [2021-04-16 06:34:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][320/1251] eta 0:04:28 lr 0.000452 time 0.2896 (0.2889) loss 3.0108 (3.5041) grad_norm 1.5763 (1.6085) [2021-04-16 06:34:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][330/1251] eta 0:04:25 lr 0.000452 time 0.2657 (0.2884) loss 3.3421 (3.5042) grad_norm 1.4270 (1.6067) [2021-04-16 06:34:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][340/1251] eta 0:04:22 lr 0.000452 time 0.2868 (0.2884) loss 2.9654 (3.5040) grad_norm 1.4738 (1.6045) [2021-04-16 06:34:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][350/1251] eta 0:04:19 lr 0.000452 time 0.2867 (0.2881) loss 3.7190 (3.5070) grad_norm 1.4400 (1.6048) [2021-04-16 06:34:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][360/1251] eta 0:04:16 lr 0.000452 time 0.2680 (0.2878) loss 3.1480 (3.5040) grad_norm 1.5954 (1.6033) [2021-04-16 06:34:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][370/1251] eta 0:04:13 lr 0.000452 time 0.2952 (0.2876) loss 3.5774 (3.5041) grad_norm 1.6301 (1.6032) [2021-04-16 06:34:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][380/1251] eta 0:04:10 lr 0.000452 time 0.2727 (0.2873) loss 3.2835 (3.5088) grad_norm 1.5275 (1.6046) [2021-04-16 06:34:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][390/1251] eta 0:04:07 lr 0.000452 time 0.2933 (0.2871) loss 3.9614 (3.5153) grad_norm 1.7285 (1.6063) [2021-04-16 06:34:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][400/1251] eta 0:04:04 lr 0.000452 time 0.2609 (0.2870) loss 3.4475 (3.5131) grad_norm 1.3926 (1.6047) [2021-04-16 06:35:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][410/1251] eta 0:04:01 lr 0.000452 time 0.2819 (0.2867) loss 4.2458 (3.5107) grad_norm 1.5697 (1.6018) [2021-04-16 06:35:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][420/1251] eta 0:03:58 lr 0.000452 time 0.2812 (0.2867) loss 3.7988 (3.5119) grad_norm 1.7475 (1.6016) [2021-04-16 06:35:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][430/1251] eta 0:03:55 lr 0.000451 time 0.2635 (0.2868) loss 3.6434 (3.5113) grad_norm 1.6195 (1.6008) [2021-04-16 06:35:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][440/1251] eta 0:03:52 lr 0.000451 time 0.2955 (0.2868) loss 2.8985 (3.5054) grad_norm 1.4128 (1.6019) [2021-04-16 06:35:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][450/1251] eta 0:03:49 lr 0.000451 time 0.2904 (0.2867) loss 3.2419 (3.5023) grad_norm 2.1586 (1.6034) [2021-04-16 06:35:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][460/1251] eta 0:03:46 lr 0.000451 time 0.2818 (0.2863) loss 2.4659 (3.5019) grad_norm 1.5006 (1.6054) [2021-04-16 06:35:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][470/1251] eta 0:03:43 lr 0.000451 time 0.2807 (0.2862) loss 4.3096 (3.5033) grad_norm 1.4676 (1.6059) [2021-04-16 06:35:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][480/1251] eta 0:03:40 lr 0.000451 time 0.2762 (0.2861) loss 2.8384 (3.5022) grad_norm 1.9118 (1.6059) [2021-04-16 06:35:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][490/1251] eta 0:03:37 lr 0.000451 time 0.2732 (0.2860) loss 4.3471 (3.5001) grad_norm 1.6276 (1.6042) [2021-04-16 06:35:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][500/1251] eta 0:03:34 lr 0.000451 time 0.2626 (0.2858) loss 3.1763 (3.4965) grad_norm 1.4649 (1.6043) [2021-04-16 06:35:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][510/1251] eta 0:03:31 lr 0.000451 time 0.2914 (0.2856) loss 4.1134 (3.4974) grad_norm 1.4480 (1.6045) [2021-04-16 06:35:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][520/1251] eta 0:03:28 lr 0.000451 time 0.3071 (0.2855) loss 3.5388 (3.5029) grad_norm 1.4501 (1.6020) [2021-04-16 06:35:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][530/1251] eta 0:03:25 lr 0.000451 time 0.2908 (0.2854) loss 4.1319 (3.5112) grad_norm 1.6110 (1.6017) [2021-04-16 06:35:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][540/1251] eta 0:03:23 lr 0.000451 time 0.2956 (0.2856) loss 3.6731 (3.5115) grad_norm 1.6732 (1.6032) [2021-04-16 06:35:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][550/1251] eta 0:03:20 lr 0.000451 time 0.2999 (0.2856) loss 3.7391 (3.5110) grad_norm 1.4184 (1.6021) [2021-04-16 06:35:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][560/1251] eta 0:03:17 lr 0.000451 time 0.2755 (0.2856) loss 2.4218 (3.5078) grad_norm 1.5524 (1.6028) [2021-04-16 06:35:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][570/1251] eta 0:03:14 lr 0.000451 time 0.2599 (0.2854) loss 3.7424 (3.5095) grad_norm 1.4244 (1.6029) [2021-04-16 06:35:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][580/1251] eta 0:03:11 lr 0.000451 time 0.2702 (0.2855) loss 4.1285 (3.5071) grad_norm 1.7027 (1.6042) [2021-04-16 06:35:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][590/1251] eta 0:03:08 lr 0.000451 time 0.2766 (0.2858) loss 3.0922 (3.5079) grad_norm 1.4227 (1.6036) [2021-04-16 06:35:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][600/1251] eta 0:03:06 lr 0.000451 time 0.2910 (0.2857) loss 3.8362 (3.5102) grad_norm 1.5775 (1.6036) [2021-04-16 06:35:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][610/1251] eta 0:03:03 lr 0.000451 time 0.2694 (0.2857) loss 3.8043 (3.5114) grad_norm 1.5863 (1.6039) [2021-04-16 06:35:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][620/1251] eta 0:03:00 lr 0.000451 time 0.2892 (0.2856) loss 2.4635 (3.5120) grad_norm 1.5479 (1.6056) [2021-04-16 06:36:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][630/1251] eta 0:02:57 lr 0.000451 time 0.2916 (0.2854) loss 4.0007 (3.5151) grad_norm 1.8682 (1.6050) [2021-04-16 06:36:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][640/1251] eta 0:02:54 lr 0.000451 time 0.2760 (0.2857) loss 3.7325 (3.5154) grad_norm 1.5364 (1.6048) [2021-04-16 06:36:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][650/1251] eta 0:02:51 lr 0.000451 time 0.2650 (0.2855) loss 4.0212 (3.5145) grad_norm 1.9110 (1.6048) [2021-04-16 06:36:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][660/1251] eta 0:02:48 lr 0.000451 time 0.2869 (0.2854) loss 4.1017 (3.5152) grad_norm 1.4942 (1.6045) [2021-04-16 06:36:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][670/1251] eta 0:02:45 lr 0.000450 time 0.2939 (0.2853) loss 2.4821 (3.5112) grad_norm 1.5250 (1.6037) [2021-04-16 06:36:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][680/1251] eta 0:02:42 lr 0.000450 time 0.2918 (0.2852) loss 3.8361 (3.5104) grad_norm 1.4617 (1.6034) [2021-04-16 06:36:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][690/1251] eta 0:02:39 lr 0.000450 time 0.2656 (0.2851) loss 3.8933 (3.5134) grad_norm 1.6663 (1.6055) [2021-04-16 06:36:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][700/1251] eta 0:02:37 lr 0.000450 time 0.2665 (0.2849) loss 3.6311 (3.5111) grad_norm 1.4979 (1.6047) [2021-04-16 06:36:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][710/1251] eta 0:02:34 lr 0.000450 time 0.2847 (0.2849) loss 2.7361 (3.5097) grad_norm 1.4251 (1.6056) [2021-04-16 06:36:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][720/1251] eta 0:02:31 lr 0.000450 time 0.3169 (0.2851) loss 2.1738 (3.5081) grad_norm 1.4478 (1.6049) [2021-04-16 06:36:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][730/1251] eta 0:02:28 lr 0.000450 time 0.2918 (0.2852) loss 2.8796 (3.5070) grad_norm 1.7240 (1.6045) [2021-04-16 06:36:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][740/1251] eta 0:02:25 lr 0.000450 time 0.3046 (0.2854) loss 3.7941 (3.5049) grad_norm 1.4092 (1.6036) [2021-04-16 06:36:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][750/1251] eta 0:02:22 lr 0.000450 time 0.2963 (0.2853) loss 4.0584 (3.5078) grad_norm 1.4376 (1.6034) [2021-04-16 06:36:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][760/1251] eta 0:02:20 lr 0.000450 time 0.2891 (0.2852) loss 3.0430 (3.5048) grad_norm 1.7899 (1.6038) [2021-04-16 06:36:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][770/1251] eta 0:02:17 lr 0.000450 time 0.2903 (0.2852) loss 3.7778 (3.5057) grad_norm 1.5507 (1.6042) [2021-04-16 06:36:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][780/1251] eta 0:02:14 lr 0.000450 time 0.2629 (0.2852) loss 3.8266 (3.5100) grad_norm 2.0685 (1.6057) [2021-04-16 06:36:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][790/1251] eta 0:02:11 lr 0.000450 time 0.2899 (0.2852) loss 3.4791 (3.5077) grad_norm 1.4388 (1.6059) [2021-04-16 06:36:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][800/1251] eta 0:02:08 lr 0.000450 time 0.4087 (0.2853) loss 2.7028 (3.5094) grad_norm 1.7579 (1.6060) [2021-04-16 06:36:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][810/1251] eta 0:02:05 lr 0.000450 time 0.2639 (0.2853) loss 3.4354 (3.5111) grad_norm 1.6166 (1.6067) [2021-04-16 06:36:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][820/1251] eta 0:02:02 lr 0.000450 time 0.2703 (0.2851) loss 2.5822 (3.5081) grad_norm 1.6931 (1.6055) [2021-04-16 06:36:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][830/1251] eta 0:02:00 lr 0.000450 time 0.2832 (0.2850) loss 3.3957 (3.5076) grad_norm 1.4448 (1.6049) [2021-04-16 06:37:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][840/1251] eta 0:01:57 lr 0.000450 time 0.2883 (0.2850) loss 2.6801 (3.5100) grad_norm 1.7896 (1.6057) [2021-04-16 06:37:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][850/1251] eta 0:01:54 lr 0.000450 time 0.2931 (0.2849) loss 3.1785 (3.5091) grad_norm 1.9057 (1.6060) [2021-04-16 06:37:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][860/1251] eta 0:01:51 lr 0.000450 time 0.2583 (0.2848) loss 3.3047 (3.5069) grad_norm 1.5722 (1.6058) [2021-04-16 06:37:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][870/1251] eta 0:01:48 lr 0.000450 time 0.2831 (0.2848) loss 2.6435 (3.5022) grad_norm 1.6678 (1.6076) [2021-04-16 06:37:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][880/1251] eta 0:01:45 lr 0.000450 time 0.2819 (0.2848) loss 3.0765 (3.5016) grad_norm 1.6339 (1.6079) [2021-04-16 06:37:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][890/1251] eta 0:01:42 lr 0.000450 time 0.2805 (0.2849) loss 2.0471 (3.4989) grad_norm 1.5644 (1.6079) [2021-04-16 06:37:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][900/1251] eta 0:01:39 lr 0.000450 time 0.2832 (0.2848) loss 3.7950 (3.5005) grad_norm 1.8290 (1.6070) [2021-04-16 06:37:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][910/1251] eta 0:01:37 lr 0.000450 time 0.2833 (0.2850) loss 3.9397 (3.5001) grad_norm 1.6792 (1.6082) [2021-04-16 06:37:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][920/1251] eta 0:01:34 lr 0.000449 time 0.2733 (0.2851) loss 3.4476 (3.5011) grad_norm 1.5032 (1.6090) [2021-04-16 06:37:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][930/1251] eta 0:01:31 lr 0.000449 time 0.2628 (0.2850) loss 3.0399 (3.5012) grad_norm 1.5066 (1.6084) [2021-04-16 06:37:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][940/1251] eta 0:01:28 lr 0.000449 time 0.2953 (0.2849) loss 4.1095 (3.5011) grad_norm 1.7087 (1.6078) [2021-04-16 06:37:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][950/1251] eta 0:01:25 lr 0.000449 time 0.2741 (0.2848) loss 3.5712 (3.5001) grad_norm 1.7710 (1.6076) [2021-04-16 06:37:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][960/1251] eta 0:01:22 lr 0.000449 time 0.3023 (0.2847) loss 3.3147 (3.4995) grad_norm 1.7586 (1.6079) [2021-04-16 06:37:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][970/1251] eta 0:01:19 lr 0.000449 time 0.2875 (0.2847) loss 3.8938 (3.4984) grad_norm 1.5802 (1.6085) [2021-04-16 06:37:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][980/1251] eta 0:01:17 lr 0.000449 time 0.3160 (0.2846) loss 2.6255 (3.5011) grad_norm 1.5711 (1.6084) [2021-04-16 06:37:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][990/1251] eta 0:01:14 lr 0.000449 time 0.2824 (0.2846) loss 3.4210 (3.4992) grad_norm 1.6270 (1.6086) [2021-04-16 06:37:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][1000/1251] eta 0:01:11 lr 0.000449 time 0.2763 (0.2845) loss 2.6512 (3.4993) grad_norm 1.6791 (1.6081) [2021-04-16 06:37:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][1010/1251] eta 0:01:08 lr 0.000449 time 0.2702 (0.2846) loss 4.0360 (3.4990) grad_norm 1.6319 (1.6078) [2021-04-16 06:37:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][1020/1251] eta 0:01:05 lr 0.000449 time 0.2810 (0.2845) loss 2.5528 (3.4964) grad_norm 1.7398 (1.6097) [2021-04-16 06:37:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][1030/1251] eta 0:01:02 lr 0.000449 time 0.3083 (0.2846) loss 3.5244 (3.4956) grad_norm 1.6001 (1.6104) [2021-04-16 06:37:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][1040/1251] eta 0:01:00 lr 0.000449 time 0.2742 (0.2846) loss 3.5362 (3.4965) grad_norm 1.6303 (1.6103) [2021-04-16 06:38:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][1050/1251] eta 0:00:57 lr 0.000449 time 0.2900 (0.2845) loss 4.0022 (3.4943) grad_norm 2.2099 (1.6120) [2021-04-16 06:38:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][1060/1251] eta 0:00:54 lr 0.000449 time 0.2421 (0.2844) loss 4.0634 (3.4950) grad_norm 1.3563 (1.6141) [2021-04-16 06:38:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][1070/1251] eta 0:00:51 lr 0.000449 time 0.2660 (0.2845) loss 3.2528 (3.4948) grad_norm 1.7983 (1.6142) [2021-04-16 06:38:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][1080/1251] eta 0:00:48 lr 0.000449 time 0.2761 (0.2844) loss 3.4750 (3.4953) grad_norm 1.5640 (1.6137) [2021-04-16 06:38:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][1090/1251] eta 0:00:45 lr 0.000449 time 0.2738 (0.2844) loss 2.6975 (3.4955) grad_norm 1.6390 (1.6135) [2021-04-16 06:38:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][1100/1251] eta 0:00:42 lr 0.000449 time 0.2925 (0.2844) loss 4.5193 (3.4965) grad_norm 1.7359 (1.6136) [2021-04-16 06:38:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][1110/1251] eta 0:00:40 lr 0.000449 time 0.2842 (0.2844) loss 3.7297 (3.4975) grad_norm 1.7505 (1.6134) [2021-04-16 06:38:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][1120/1251] eta 0:00:37 lr 0.000449 time 0.2922 (0.2844) loss 2.7692 (3.4974) grad_norm 1.4540 (1.6132) [2021-04-16 06:38:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][1130/1251] eta 0:00:34 lr 0.000449 time 0.2749 (0.2844) loss 3.6592 (3.4978) grad_norm 2.0765 (1.6133) [2021-04-16 06:38:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][1140/1251] eta 0:00:31 lr 0.000449 time 0.2784 (0.2843) loss 2.7691 (3.4983) grad_norm 1.5924 (1.6135) [2021-04-16 06:38:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][1150/1251] eta 0:00:28 lr 0.000449 time 0.2538 (0.2844) loss 3.9621 (3.4966) grad_norm 1.6385 (1.6128) [2021-04-16 06:38:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][1160/1251] eta 0:00:25 lr 0.000448 time 0.2693 (0.2844) loss 3.4449 (3.4971) grad_norm 1.5537 (1.6129) [2021-04-16 06:38:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][1170/1251] eta 0:00:23 lr 0.000448 time 0.2719 (0.2844) loss 3.4768 (3.4986) grad_norm 1.8209 (1.6123) [2021-04-16 06:38:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][1180/1251] eta 0:00:20 lr 0.000448 time 0.2903 (0.2843) loss 3.8826 (3.5001) grad_norm 1.6659 (1.6124) [2021-04-16 06:38:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][1190/1251] eta 0:00:17 lr 0.000448 time 0.2678 (0.2843) loss 3.3882 (3.4989) grad_norm 1.6325 (1.6125) [2021-04-16 06:38:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][1200/1251] eta 0:00:14 lr 0.000448 time 0.2884 (0.2843) loss 4.0376 (3.5013) grad_norm 1.6938 (1.6128) [2021-04-16 06:38:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][1210/1251] eta 0:00:11 lr 0.000448 time 0.2585 (0.2841) loss 3.8219 (3.5019) grad_norm 1.5539 (1.6132) [2021-04-16 06:38:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][1220/1251] eta 0:00:08 lr 0.000448 time 0.2946 (0.2842) loss 4.3000 (3.5029) grad_norm 1.8739 (1.6127) [2021-04-16 06:38:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][1230/1251] eta 0:00:05 lr 0.000448 time 0.2825 (0.2842) loss 3.5013 (3.5022) grad_norm 1.6553 (1.6134) [2021-04-16 06:38:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][1240/1251] eta 0:00:03 lr 0.000448 time 0.2494 (0.2840) loss 3.6767 (3.5040) grad_norm 1.4774 (1.6130) [2021-04-16 06:38:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [160/300][1250/1251] eta 0:00:00 lr 0.000448 time 0.2464 (0.2838) loss 3.6873 (3.5033) grad_norm 1.5289 (nan) [2021-04-16 06:39:01 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 160 training takes 0:05:58 [2021-04-16 06:39:01 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_160.pth saving...... [2021-04-16 06:39:11 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_160.pth saved !!! [2021-04-16 06:39:12 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.161 (1.161) Loss 1.0368 (1.0368) Acc@1 74.414 (74.414) Acc@5 93.457 (93.457) [2021-04-16 06:39:13 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.132 (0.202) Loss 0.9775 (1.0108) Acc@1 76.953 (76.003) Acc@5 93.164 (93.226) [2021-04-16 06:39:15 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.115 (0.214) Loss 0.9527 (1.0064) Acc@1 78.027 (76.274) Acc@5 93.555 (93.248) [2021-04-16 06:39:18 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.097 (0.231) Loss 1.0170 (0.9964) Acc@1 76.074 (76.462) Acc@5 92.969 (93.444) [2021-04-16 06:39:19 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.215) Loss 1.0180 (0.9958) Acc@1 76.074 (76.520) Acc@5 91.992 (93.426) [2021-04-16 06:39:28 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 76.466 Acc@5 93.402 [2021-04-16 06:39:28 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 76.5% [2021-04-16 06:39:28 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 76.47% [2021-04-16 06:39:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][0/1251] eta 1:50:18 lr 0.000448 time 5.2907 (5.2907) loss 4.0108 (4.0108) grad_norm 1.6828 (1.6828) [2021-04-16 06:39:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][10/1251] eta 0:15:23 lr 0.000448 time 0.4430 (0.7439) loss 4.2702 (3.5509) grad_norm 1.5897 (1.5336) [2021-04-16 06:39:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][20/1251] eta 0:10:41 lr 0.000448 time 0.2740 (0.5212) loss 2.7486 (3.4537) grad_norm 1.8003 (1.5590) [2021-04-16 06:39:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][30/1251] eta 0:09:03 lr 0.000448 time 0.3172 (0.4455) loss 3.5949 (3.4573) grad_norm 1.6390 (1.6066) [2021-04-16 06:39:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][40/1251] eta 0:08:10 lr 0.000448 time 0.2892 (0.4054) loss 3.4579 (3.4071) grad_norm 1.5229 (1.6059) [2021-04-16 06:39:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][50/1251] eta 0:07:37 lr 0.000448 time 0.2931 (0.3808) loss 4.5257 (3.4317) grad_norm 1.5987 (1.5869) [2021-04-16 06:39:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][60/1251] eta 0:07:13 lr 0.000448 time 0.2862 (0.3643) loss 2.2259 (3.3700) grad_norm 1.4056 (1.6145) [2021-04-16 06:39:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][70/1251] eta 0:06:56 lr 0.000448 time 0.2537 (0.3526) loss 4.0925 (3.3696) grad_norm 1.6953 (1.6163) [2021-04-16 06:39:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][80/1251] eta 0:06:42 lr 0.000448 time 0.2692 (0.3437) loss 3.8400 (3.3590) grad_norm 1.7599 (1.6247) [2021-04-16 06:39:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][90/1251] eta 0:06:32 lr 0.000448 time 0.2843 (0.3381) loss 4.1242 (3.3603) grad_norm 1.5355 (1.6291) [2021-04-16 06:40:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][100/1251] eta 0:06:23 lr 0.000448 time 0.2885 (0.3331) loss 2.0572 (3.3849) grad_norm 1.5909 (1.6242) [2021-04-16 06:40:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][110/1251] eta 0:06:13 lr 0.000448 time 0.2637 (0.3277) loss 3.7343 (3.3733) grad_norm 2.1602 (1.6276) [2021-04-16 06:40:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][120/1251] eta 0:06:06 lr 0.000448 time 0.2809 (0.3244) loss 2.8149 (3.3664) grad_norm 1.5713 (1.6313) [2021-04-16 06:40:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][130/1251] eta 0:05:59 lr 0.000448 time 0.2977 (0.3209) loss 3.3404 (3.3572) grad_norm 1.5151 (1.6279) [2021-04-16 06:40:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][140/1251] eta 0:05:52 lr 0.000448 time 0.2553 (0.3176) loss 3.8881 (3.3800) grad_norm 1.6406 (1.6385) [2021-04-16 06:40:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][150/1251] eta 0:05:47 lr 0.000447 time 0.2907 (0.3152) loss 3.0914 (3.3846) grad_norm 1.5416 (1.6308) [2021-04-16 06:40:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][160/1251] eta 0:05:41 lr 0.000447 time 0.2508 (0.3126) loss 3.2744 (3.3939) grad_norm 1.5307 (1.6256) [2021-04-16 06:40:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][170/1251] eta 0:05:36 lr 0.000447 time 0.2790 (0.3109) loss 3.8887 (3.3959) grad_norm 1.5955 (1.6247) [2021-04-16 06:40:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][180/1251] eta 0:05:31 lr 0.000447 time 0.3007 (0.3094) loss 3.5510 (3.4145) grad_norm 1.6726 (1.6303) [2021-04-16 06:40:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][190/1251] eta 0:05:26 lr 0.000447 time 0.2953 (0.3079) loss 3.4302 (3.4198) grad_norm 1.8536 (1.6340) [2021-04-16 06:40:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][200/1251] eta 0:05:22 lr 0.000447 time 0.2674 (0.3070) loss 2.3304 (3.4158) grad_norm 1.4952 (1.6328) [2021-04-16 06:40:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][210/1251] eta 0:05:18 lr 0.000447 time 0.2533 (0.3056) loss 3.9829 (3.4275) grad_norm 1.6171 (1.6298) [2021-04-16 06:40:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][220/1251] eta 0:05:14 lr 0.000447 time 0.2989 (0.3051) loss 3.4418 (3.4232) grad_norm 1.6450 (1.6290) [2021-04-16 06:40:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][230/1251] eta 0:05:10 lr 0.000447 time 0.2749 (0.3042) loss 3.6524 (3.4278) grad_norm 1.8289 (1.6300) [2021-04-16 06:40:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][240/1251] eta 0:05:06 lr 0.000447 time 0.2975 (0.3032) loss 3.8879 (3.4310) grad_norm 1.3717 (1.6282) [2021-04-16 06:40:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][250/1251] eta 0:05:02 lr 0.000447 time 0.3142 (0.3024) loss 3.6233 (3.4367) grad_norm 1.8862 (1.6266) [2021-04-16 06:40:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][260/1251] eta 0:04:58 lr 0.000447 time 0.2848 (0.3016) loss 2.3620 (3.4399) grad_norm 1.4698 (1.6216) [2021-04-16 06:40:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][270/1251] eta 0:04:55 lr 0.000447 time 0.2736 (0.3009) loss 2.6780 (3.4377) grad_norm 1.8151 (1.6217) [2021-04-16 06:40:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][280/1251] eta 0:04:51 lr 0.000447 time 0.2746 (0.3001) loss 2.3134 (3.4324) grad_norm 1.6339 (1.6204) [2021-04-16 06:40:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][290/1251] eta 0:04:47 lr 0.000447 time 0.2855 (0.2995) loss 4.2296 (3.4302) grad_norm 1.4408 (1.6176) [2021-04-16 06:40:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][300/1251] eta 0:04:44 lr 0.000447 time 0.2766 (0.2987) loss 2.9454 (3.4331) grad_norm 1.6403 (1.6148) [2021-04-16 06:41:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][310/1251] eta 0:04:40 lr 0.000447 time 0.2679 (0.2981) loss 3.6594 (3.4402) grad_norm 2.0538 (1.6162) [2021-04-16 06:41:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][320/1251] eta 0:04:37 lr 0.000447 time 0.2498 (0.2983) loss 3.6612 (3.4412) grad_norm 1.5371 (1.6177) [2021-04-16 06:41:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][330/1251] eta 0:04:34 lr 0.000447 time 0.2880 (0.2977) loss 3.4846 (3.4386) grad_norm 1.6769 (1.6152) [2021-04-16 06:41:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][340/1251] eta 0:04:30 lr 0.000447 time 0.2857 (0.2971) loss 2.5395 (3.4379) grad_norm 1.3906 (1.6143) [2021-04-16 06:41:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][350/1251] eta 0:04:27 lr 0.000447 time 0.3020 (0.2968) loss 3.0149 (3.4389) grad_norm 1.5610 (1.6139) [2021-04-16 06:41:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][360/1251] eta 0:04:24 lr 0.000447 time 0.2818 (0.2965) loss 3.0129 (3.4387) grad_norm 1.6771 (1.6144) [2021-04-16 06:41:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][370/1251] eta 0:04:21 lr 0.000447 time 0.2870 (0.2965) loss 4.0076 (3.4430) grad_norm 1.5616 (1.6192) [2021-04-16 06:41:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][380/1251] eta 0:04:17 lr 0.000447 time 0.3079 (0.2962) loss 3.5067 (3.4465) grad_norm 1.4265 (1.6179) [2021-04-16 06:41:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][390/1251] eta 0:04:14 lr 0.000447 time 0.2736 (0.2958) loss 4.0391 (3.4408) grad_norm 1.6876 (1.6196) [2021-04-16 06:41:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][400/1251] eta 0:04:11 lr 0.000446 time 0.2965 (0.2954) loss 4.1850 (3.4443) grad_norm 1.4496 (1.6189) [2021-04-16 06:41:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][410/1251] eta 0:04:08 lr 0.000446 time 0.2785 (0.2951) loss 2.8021 (3.4372) grad_norm 1.6343 (1.6207) [2021-04-16 06:41:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][420/1251] eta 0:04:05 lr 0.000446 time 0.2766 (0.2950) loss 3.8614 (3.4461) grad_norm 1.6527 (1.6189) [2021-04-16 06:41:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][430/1251] eta 0:04:01 lr 0.000446 time 0.2567 (0.2946) loss 3.2733 (3.4412) grad_norm 1.5696 (1.6186) [2021-04-16 06:41:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][440/1251] eta 0:03:58 lr 0.000446 time 0.2887 (0.2943) loss 4.1460 (3.4459) grad_norm 2.2885 (1.6193) [2021-04-16 06:41:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][450/1251] eta 0:03:55 lr 0.000446 time 0.2851 (0.2940) loss 3.0220 (3.4497) grad_norm 1.5565 (1.6206) [2021-04-16 06:41:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][460/1251] eta 0:03:52 lr 0.000446 time 0.3025 (0.2938) loss 3.8175 (3.4485) grad_norm 1.6036 (1.6209) [2021-04-16 06:41:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][470/1251] eta 0:03:49 lr 0.000446 time 0.2589 (0.2935) loss 3.7447 (3.4495) grad_norm 1.6339 (1.6214) [2021-04-16 06:41:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][480/1251] eta 0:03:46 lr 0.000446 time 0.2698 (0.2931) loss 3.7422 (3.4588) grad_norm 1.4892 (1.6215) [2021-04-16 06:41:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][490/1251] eta 0:03:42 lr 0.000446 time 0.2746 (0.2928) loss 3.0446 (3.4587) grad_norm 1.6140 (1.6209) [2021-04-16 06:41:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][500/1251] eta 0:03:39 lr 0.000446 time 0.2990 (0.2924) loss 2.5490 (3.4608) grad_norm 1.7180 (1.6200) [2021-04-16 06:41:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][510/1251] eta 0:03:36 lr 0.000446 time 0.2874 (0.2923) loss 2.2646 (3.4611) grad_norm 1.5770 (1.6190) [2021-04-16 06:42:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][520/1251] eta 0:03:33 lr 0.000446 time 0.2920 (0.2921) loss 3.2453 (3.4588) grad_norm 1.4371 (1.6202) [2021-04-16 06:42:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][530/1251] eta 0:03:30 lr 0.000446 time 0.2892 (0.2919) loss 3.9090 (3.4573) grad_norm 1.6116 (1.6198) [2021-04-16 06:42:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][540/1251] eta 0:03:27 lr 0.000446 time 0.2527 (0.2920) loss 3.7132 (3.4587) grad_norm 1.6087 (1.6196) [2021-04-16 06:42:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][550/1251] eta 0:03:24 lr 0.000446 time 0.4121 (0.2920) loss 3.2481 (3.4598) grad_norm 1.5897 (1.6211) [2021-04-16 06:42:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][560/1251] eta 0:03:21 lr 0.000446 time 0.2916 (0.2918) loss 3.6991 (3.4570) grad_norm 1.4574 (1.6200) [2021-04-16 06:42:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][570/1251] eta 0:03:18 lr 0.000446 time 0.2543 (0.2918) loss 3.2934 (3.4543) grad_norm 1.5960 (1.6201) [2021-04-16 06:42:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][580/1251] eta 0:03:15 lr 0.000446 time 0.2872 (0.2918) loss 2.7441 (3.4540) grad_norm 1.7106 (1.6189) [2021-04-16 06:42:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][590/1251] eta 0:03:12 lr 0.000446 time 0.2760 (0.2920) loss 2.6930 (3.4549) grad_norm 2.0110 (1.6190) [2021-04-16 06:42:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][600/1251] eta 0:03:09 lr 0.000446 time 0.2736 (0.2917) loss 3.9104 (3.4589) grad_norm 1.7094 (1.6205) [2021-04-16 06:42:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][610/1251] eta 0:03:06 lr 0.000446 time 0.3016 (0.2916) loss 3.8426 (3.4557) grad_norm 1.8976 (1.6207) [2021-04-16 06:42:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][620/1251] eta 0:03:03 lr 0.000446 time 0.3064 (0.2916) loss 3.4321 (3.4554) grad_norm 1.9448 (1.6216) [2021-04-16 06:42:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][630/1251] eta 0:03:00 lr 0.000446 time 0.2636 (0.2914) loss 3.7127 (3.4558) grad_norm 1.6335 (1.6193) [2021-04-16 06:42:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][640/1251] eta 0:02:58 lr 0.000445 time 0.2948 (0.2914) loss 3.6421 (3.4577) grad_norm 1.5518 (1.6189) [2021-04-16 06:42:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][650/1251] eta 0:02:54 lr 0.000445 time 0.2693 (0.2911) loss 3.9273 (3.4640) grad_norm 1.5776 (1.6190) [2021-04-16 06:42:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][660/1251] eta 0:02:51 lr 0.000445 time 0.2713 (0.2910) loss 3.5848 (3.4663) grad_norm 1.7307 (1.6180) [2021-04-16 06:42:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][670/1251] eta 0:02:49 lr 0.000445 time 0.2794 (0.2909) loss 3.1044 (3.4685) grad_norm 1.5740 (1.6169) [2021-04-16 06:42:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][680/1251] eta 0:02:45 lr 0.000445 time 0.2888 (0.2907) loss 3.7425 (3.4637) grad_norm 1.9159 (1.6177) [2021-04-16 06:42:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][690/1251] eta 0:02:43 lr 0.000445 time 0.2599 (0.2907) loss 2.7420 (3.4612) grad_norm 1.6575 (1.6182) [2021-04-16 06:42:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][700/1251] eta 0:02:40 lr 0.000445 time 0.2966 (0.2906) loss 2.5785 (3.4625) grad_norm 1.6230 (1.6182) [2021-04-16 06:42:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][710/1251] eta 0:02:37 lr 0.000445 time 0.2550 (0.2904) loss 3.9942 (3.4606) grad_norm 1.4821 (1.6188) [2021-04-16 06:42:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][720/1251] eta 0:02:34 lr 0.000445 time 0.2754 (0.2902) loss 3.3132 (3.4600) grad_norm 1.5599 (1.6179) [2021-04-16 06:43:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][730/1251] eta 0:02:31 lr 0.000445 time 0.2988 (0.2902) loss 3.3389 (3.4589) grad_norm 1.7773 (1.6215) [2021-04-16 06:43:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][740/1251] eta 0:02:28 lr 0.000445 time 0.2827 (0.2900) loss 4.3774 (3.4621) grad_norm 1.4321 (1.6219) [2021-04-16 06:43:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][750/1251] eta 0:02:25 lr 0.000445 time 0.2714 (0.2899) loss 3.9743 (3.4623) grad_norm 1.5272 (1.6226) [2021-04-16 06:43:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][760/1251] eta 0:02:22 lr 0.000445 time 0.2812 (0.2897) loss 3.4625 (3.4608) grad_norm 1.4851 (1.6218) [2021-04-16 06:43:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][770/1251] eta 0:02:19 lr 0.000445 time 0.2879 (0.2897) loss 2.3330 (3.4631) grad_norm 1.4368 (1.6221) [2021-04-16 06:43:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][780/1251] eta 0:02:16 lr 0.000445 time 0.2472 (0.2897) loss 3.1375 (3.4624) grad_norm 1.4808 (1.6232) [2021-04-16 06:43:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][790/1251] eta 0:02:13 lr 0.000445 time 0.4531 (0.2898) loss 3.4971 (3.4611) grad_norm 1.2917 (1.6234) [2021-04-16 06:43:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][800/1251] eta 0:02:10 lr 0.000445 time 0.2714 (0.2895) loss 3.7237 (3.4612) grad_norm 1.6182 (1.6231) [2021-04-16 06:43:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][810/1251] eta 0:02:07 lr 0.000445 time 0.2614 (0.2895) loss 4.3330 (3.4607) grad_norm 1.6032 (1.6221) [2021-04-16 06:43:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][820/1251] eta 0:02:04 lr 0.000445 time 0.2427 (0.2893) loss 3.4246 (3.4613) grad_norm 1.9449 (1.6221) [2021-04-16 06:43:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][830/1251] eta 0:02:01 lr 0.000445 time 0.2783 (0.2893) loss 3.2649 (3.4597) grad_norm 1.7820 (1.6226) [2021-04-16 06:43:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][840/1251] eta 0:01:58 lr 0.000445 time 0.2740 (0.2892) loss 3.5583 (3.4574) grad_norm 1.7065 (1.6223) [2021-04-16 06:43:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][850/1251] eta 0:01:55 lr 0.000445 time 0.2993 (0.2891) loss 3.0304 (3.4578) grad_norm 1.4505 (1.6218) [2021-04-16 06:43:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][860/1251] eta 0:01:52 lr 0.000445 time 0.2662 (0.2890) loss 3.8212 (3.4574) grad_norm 1.4643 (1.6211) [2021-04-16 06:43:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][870/1251] eta 0:01:50 lr 0.000445 time 0.2902 (0.2889) loss 3.7256 (3.4571) grad_norm 1.4594 (1.6208) [2021-04-16 06:43:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][880/1251] eta 0:01:47 lr 0.000444 time 0.2986 (0.2888) loss 3.2956 (3.4600) grad_norm 1.7584 (1.6223) [2021-04-16 06:43:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][890/1251] eta 0:01:44 lr 0.000444 time 0.2654 (0.2889) loss 3.4351 (3.4608) grad_norm 1.5970 (1.6218) [2021-04-16 06:43:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][900/1251] eta 0:01:41 lr 0.000444 time 0.2841 (0.2889) loss 2.7627 (3.4619) grad_norm 1.3803 (1.6209) [2021-04-16 06:43:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][910/1251] eta 0:01:38 lr 0.000444 time 0.2727 (0.2888) loss 2.6797 (3.4657) grad_norm 1.6651 (1.6202) [2021-04-16 06:43:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][920/1251] eta 0:01:35 lr 0.000444 time 0.2686 (0.2887) loss 4.1939 (3.4682) grad_norm 1.4611 (1.6205) [2021-04-16 06:43:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][930/1251] eta 0:01:32 lr 0.000444 time 0.2978 (0.2887) loss 3.7112 (3.4660) grad_norm 1.9054 (1.6209) [2021-04-16 06:43:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][940/1251] eta 0:01:29 lr 0.000444 time 0.4066 (0.2888) loss 3.7120 (3.4680) grad_norm 1.8245 (1.6213) [2021-04-16 06:44:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][950/1251] eta 0:01:26 lr 0.000444 time 0.2660 (0.2886) loss 2.5090 (3.4689) grad_norm 1.5105 (1.6213) [2021-04-16 06:44:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][960/1251] eta 0:01:23 lr 0.000444 time 0.2579 (0.2886) loss 2.7027 (3.4686) grad_norm 1.6863 (1.6220) [2021-04-16 06:44:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][970/1251] eta 0:01:21 lr 0.000444 time 0.2988 (0.2885) loss 3.1632 (3.4678) grad_norm 1.7483 (1.6231) [2021-04-16 06:44:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][980/1251] eta 0:01:18 lr 0.000444 time 0.2612 (0.2885) loss 4.2182 (3.4689) grad_norm 1.5522 (1.6244) [2021-04-16 06:44:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][990/1251] eta 0:01:15 lr 0.000444 time 0.2886 (0.2885) loss 2.7216 (3.4674) grad_norm 1.5755 (1.6247) [2021-04-16 06:44:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][1000/1251] eta 0:01:12 lr 0.000444 time 0.2752 (0.2883) loss 3.8540 (3.4687) grad_norm 1.6217 (1.6246) [2021-04-16 06:44:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][1010/1251] eta 0:01:09 lr 0.000444 time 0.2853 (0.2882) loss 2.1932 (3.4676) grad_norm 1.7612 (1.6244) [2021-04-16 06:44:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][1020/1251] eta 0:01:06 lr 0.000444 time 0.2786 (0.2882) loss 4.0843 (3.4690) grad_norm 1.5586 (1.6241) [2021-04-16 06:44:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][1030/1251] eta 0:01:03 lr 0.000444 time 0.2759 (0.2880) loss 3.7207 (3.4694) grad_norm 1.5669 (1.6234) [2021-04-16 06:44:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][1040/1251] eta 0:01:00 lr 0.000444 time 0.2522 (0.2879) loss 2.6680 (3.4704) grad_norm 1.4366 (1.6227) [2021-04-16 06:44:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][1050/1251] eta 0:00:57 lr 0.000444 time 0.3066 (0.2879) loss 4.0491 (3.4709) grad_norm 1.7331 (1.6220) [2021-04-16 06:44:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][1060/1251] eta 0:00:54 lr 0.000444 time 0.3015 (0.2879) loss 3.8985 (3.4735) grad_norm 1.5190 (1.6218) [2021-04-16 06:44:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][1070/1251] eta 0:00:52 lr 0.000444 time 0.2734 (0.2877) loss 3.3202 (3.4708) grad_norm 1.9148 (1.6214) [2021-04-16 06:44:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][1080/1251] eta 0:00:49 lr 0.000444 time 0.2761 (0.2877) loss 4.2186 (3.4697) grad_norm 1.4346 (1.6219) [2021-04-16 06:44:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][1090/1251] eta 0:00:46 lr 0.000444 time 0.2425 (0.2876) loss 4.3138 (3.4734) grad_norm 1.6545 (1.6217) [2021-04-16 06:44:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][1100/1251] eta 0:00:43 lr 0.000444 time 0.3116 (0.2875) loss 2.1045 (3.4711) grad_norm 1.4783 (1.6216) [2021-04-16 06:44:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][1110/1251] eta 0:00:40 lr 0.000444 time 0.2848 (0.2875) loss 3.6405 (3.4713) grad_norm 1.4043 (1.6207) [2021-04-16 06:44:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][1120/1251] eta 0:00:37 lr 0.000443 time 0.2955 (0.2875) loss 2.9728 (3.4706) grad_norm 1.4519 (1.6203) [2021-04-16 06:44:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][1130/1251] eta 0:00:34 lr 0.000443 time 0.2824 (0.2874) loss 3.6022 (3.4719) grad_norm 1.6335 (1.6210) [2021-04-16 06:44:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][1140/1251] eta 0:00:31 lr 0.000443 time 0.2572 (0.2874) loss 4.0583 (3.4710) grad_norm 1.3706 (1.6203) [2021-04-16 06:44:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][1150/1251] eta 0:00:29 lr 0.000443 time 0.2721 (0.2873) loss 3.2803 (3.4709) grad_norm 1.4225 (1.6204) [2021-04-16 06:45:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][1160/1251] eta 0:00:26 lr 0.000443 time 0.2468 (0.2873) loss 3.0362 (3.4712) grad_norm 1.5484 (1.6196) [2021-04-16 06:45:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][1170/1251] eta 0:00:23 lr 0.000443 time 0.2529 (0.2873) loss 3.8698 (3.4730) grad_norm 1.7297 (1.6191) [2021-04-16 06:45:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][1180/1251] eta 0:00:20 lr 0.000443 time 0.3065 (0.2873) loss 3.6691 (3.4736) grad_norm 1.7375 (1.6194) [2021-04-16 06:45:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][1190/1251] eta 0:00:17 lr 0.000443 time 0.2725 (0.2873) loss 4.0613 (3.4739) grad_norm 1.4433 (1.6196) [2021-04-16 06:45:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][1200/1251] eta 0:00:14 lr 0.000443 time 0.2849 (0.2872) loss 4.3467 (3.4727) grad_norm 1.5811 (1.6197) [2021-04-16 06:45:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][1210/1251] eta 0:00:11 lr 0.000443 time 0.2933 (0.2872) loss 3.8569 (3.4730) grad_norm 1.9661 (1.6202) [2021-04-16 06:45:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][1220/1251] eta 0:00:08 lr 0.000443 time 0.2840 (0.2872) loss 4.3124 (3.4741) grad_norm 1.4984 (1.6201) [2021-04-16 06:45:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][1230/1251] eta 0:00:06 lr 0.000443 time 0.3158 (0.2871) loss 4.0764 (3.4740) grad_norm 1.7873 (1.6200) [2021-04-16 06:45:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][1240/1251] eta 0:00:03 lr 0.000443 time 0.3173 (0.2871) loss 2.5212 (3.4742) grad_norm 1.3970 (1.6191) [2021-04-16 06:45:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [161/300][1250/1251] eta 0:00:00 lr 0.000443 time 0.2483 (0.2868) loss 3.8397 (3.4736) grad_norm 1.7281 (1.6190) [2021-04-16 06:45:30 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 161 training takes 0:06:01 [2021-04-16 06:45:30 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_161.pth saving...... [2021-04-16 06:45:39 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_161.pth saved !!! [2021-04-16 06:45:40 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.150 (1.150) Loss 0.9697 (0.9697) Acc@1 79.199 (79.199) Acc@5 92.773 (92.773) [2021-04-16 06:45:42 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.128 (0.238) Loss 0.9788 (0.9962) Acc@1 78.320 (77.379) Acc@5 94.043 (93.422) [2021-04-16 06:45:44 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.101 (0.242) Loss 1.0953 (1.0127) Acc@1 74.609 (76.851) Acc@5 92.285 (93.341) [2021-04-16 06:45:46 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.146 (0.223) Loss 1.0303 (1.0133) Acc@1 77.051 (76.755) Acc@5 92.969 (93.356) [2021-04-16 06:45:48 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.214) Loss 1.0037 (1.0138) Acc@1 77.539 (76.746) Acc@5 93.262 (93.357) [2021-04-16 06:45:58 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 76.632 Acc@5 93.378 [2021-04-16 06:45:58 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 76.6% [2021-04-16 06:45:58 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 76.63% [2021-04-16 06:46:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][0/1251] eta 1:00:09 lr 0.000443 time 2.8852 (2.8852) loss 3.5225 (3.5225) grad_norm 1.7351 (1.7351) [2021-04-16 06:46:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][10/1251] eta 0:10:30 lr 0.000443 time 0.3019 (0.5080) loss 3.2119 (3.1525) grad_norm 1.5461 (1.5823) [2021-04-16 06:46:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][20/1251] eta 0:08:16 lr 0.000443 time 0.2800 (0.4030) loss 3.4914 (3.3587) grad_norm 1.8453 (1.6471) [2021-04-16 06:46:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][30/1251] eta 0:07:23 lr 0.000443 time 0.2911 (0.3629) loss 3.3693 (3.4304) grad_norm 1.6110 (1.6630) [2021-04-16 06:46:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][40/1251] eta 0:06:53 lr 0.000443 time 0.2592 (0.3412) loss 4.1341 (3.4229) grad_norm 1.5129 (1.6497) [2021-04-16 06:46:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][50/1251] eta 0:06:35 lr 0.000443 time 0.2937 (0.3294) loss 2.5625 (3.4068) grad_norm 1.5591 (1.6398) [2021-04-16 06:46:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][60/1251] eta 0:06:24 lr 0.000443 time 0.2561 (0.3231) loss 2.8093 (3.3680) grad_norm 1.5487 (1.6325) [2021-04-16 06:46:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][70/1251] eta 0:06:13 lr 0.000443 time 0.2727 (0.3166) loss 4.1228 (3.4304) grad_norm 1.4825 (1.6329) [2021-04-16 06:46:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][80/1251] eta 0:06:05 lr 0.000443 time 0.2881 (0.3120) loss 3.8247 (3.4514) grad_norm 1.6503 (1.6360) [2021-04-16 06:46:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][90/1251] eta 0:06:00 lr 0.000443 time 0.2818 (0.3101) loss 4.0667 (3.4291) grad_norm 1.5906 (1.6329) [2021-04-16 06:46:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][100/1251] eta 0:05:53 lr 0.000443 time 0.2570 (0.3072) loss 3.9523 (3.4585) grad_norm 1.7703 (1.6305) [2021-04-16 06:46:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][110/1251] eta 0:05:47 lr 0.000443 time 0.2685 (0.3047) loss 3.5881 (3.4562) grad_norm 1.7920 (1.6310) [2021-04-16 06:46:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][120/1251] eta 0:05:42 lr 0.000442 time 0.2795 (0.3025) loss 2.9605 (3.4363) grad_norm 1.9008 (1.6526) [2021-04-16 06:46:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][130/1251] eta 0:05:36 lr 0.000442 time 0.2858 (0.3003) loss 2.6626 (3.4392) grad_norm 1.5631 (1.6700) [2021-04-16 06:46:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][140/1251] eta 0:05:32 lr 0.000442 time 0.2743 (0.2995) loss 3.6994 (3.4271) grad_norm 1.6962 (1.6773) [2021-04-16 06:46:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][150/1251] eta 0:05:28 lr 0.000442 time 0.3978 (0.2988) loss 3.5309 (3.4242) grad_norm 1.4354 (1.6716) [2021-04-16 06:46:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][160/1251] eta 0:05:24 lr 0.000442 time 0.2773 (0.2974) loss 3.1300 (3.4394) grad_norm 1.5740 (1.6715) [2021-04-16 06:46:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][170/1251] eta 0:05:20 lr 0.000442 time 0.2541 (0.2963) loss 3.9173 (3.4408) grad_norm 1.6444 (1.6644) [2021-04-16 06:46:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][180/1251] eta 0:05:16 lr 0.000442 time 0.2811 (0.2955) loss 3.8148 (3.4554) grad_norm 1.5939 (1.6555) [2021-04-16 06:46:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][190/1251] eta 0:05:12 lr 0.000442 time 0.2672 (0.2945) loss 4.2982 (3.4599) grad_norm 1.5643 (1.6531) [2021-04-16 06:46:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][200/1251] eta 0:05:09 lr 0.000442 time 0.2812 (0.2946) loss 3.0377 (3.4503) grad_norm 1.5677 (1.6520) [2021-04-16 06:47:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][210/1251] eta 0:05:06 lr 0.000442 time 0.2572 (0.2941) loss 3.6502 (3.4456) grad_norm 1.5619 (1.6505) [2021-04-16 06:47:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][220/1251] eta 0:05:02 lr 0.000442 time 0.2941 (0.2936) loss 3.9140 (3.4351) grad_norm 1.6959 (1.6490) [2021-04-16 06:47:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][230/1251] eta 0:04:58 lr 0.000442 time 0.2757 (0.2926) loss 4.2206 (3.4490) grad_norm 1.6706 (1.6459) [2021-04-16 06:47:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][240/1251] eta 0:04:55 lr 0.000442 time 0.2957 (0.2919) loss 3.3575 (3.4564) grad_norm 1.4324 (1.6440) [2021-04-16 06:47:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][250/1251] eta 0:04:51 lr 0.000442 time 0.3080 (0.2916) loss 3.1805 (3.4621) grad_norm 1.5333 (1.6455) [2021-04-16 06:47:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][260/1251] eta 0:04:48 lr 0.000442 time 0.3044 (0.2909) loss 3.4551 (3.4578) grad_norm 1.7495 (1.6469) [2021-04-16 06:47:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][270/1251] eta 0:04:44 lr 0.000442 time 0.2920 (0.2903) loss 3.2084 (3.4607) grad_norm 2.0026 (1.6529) [2021-04-16 06:47:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][280/1251] eta 0:04:41 lr 0.000442 time 0.2830 (0.2897) loss 3.0222 (3.4440) grad_norm 1.5430 (1.6546) [2021-04-16 06:47:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][290/1251] eta 0:04:38 lr 0.000442 time 0.2894 (0.2893) loss 2.7063 (3.4467) grad_norm 1.6486 (1.6519) [2021-04-16 06:47:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][300/1251] eta 0:04:34 lr 0.000442 time 0.2484 (0.2890) loss 3.0501 (3.4547) grad_norm 1.4868 (1.6515) [2021-04-16 06:47:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][310/1251] eta 0:04:31 lr 0.000442 time 0.2761 (0.2888) loss 3.7902 (3.4590) grad_norm 1.6246 (1.6498) [2021-04-16 06:47:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][320/1251] eta 0:04:28 lr 0.000442 time 0.2809 (0.2889) loss 4.0213 (3.4543) grad_norm 1.4784 (1.6484) [2021-04-16 06:47:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][330/1251] eta 0:04:25 lr 0.000442 time 0.2718 (0.2885) loss 3.5165 (3.4529) grad_norm 1.4987 (1.6440) [2021-04-16 06:47:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][340/1251] eta 0:04:23 lr 0.000442 time 0.2660 (0.2888) loss 2.3781 (3.4511) grad_norm 1.5048 (1.6414) [2021-04-16 06:47:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][350/1251] eta 0:04:19 lr 0.000442 time 0.2540 (0.2883) loss 3.0691 (3.4521) grad_norm 1.6680 (1.6399) [2021-04-16 06:47:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][360/1251] eta 0:04:16 lr 0.000441 time 0.2985 (0.2882) loss 4.4465 (3.4563) grad_norm 1.4731 (1.6392) [2021-04-16 06:47:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][370/1251] eta 0:04:13 lr 0.000441 time 0.2475 (0.2880) loss 3.8422 (3.4485) grad_norm 1.4489 (1.6378) [2021-04-16 06:47:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][380/1251] eta 0:04:10 lr 0.000441 time 0.2566 (0.2878) loss 3.3386 (3.4532) grad_norm 1.5816 (1.6369) [2021-04-16 06:47:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][390/1251] eta 0:04:07 lr 0.000441 time 0.2716 (0.2875) loss 3.6921 (3.4487) grad_norm 1.5598 (1.6394) [2021-04-16 06:47:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][400/1251] eta 0:04:04 lr 0.000441 time 0.2749 (0.2873) loss 2.1742 (3.4500) grad_norm 1.9985 (1.6414) [2021-04-16 06:47:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][410/1251] eta 0:04:01 lr 0.000441 time 0.2674 (0.2871) loss 3.6213 (3.4552) grad_norm 1.5802 (1.6404) [2021-04-16 06:47:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][420/1251] eta 0:03:58 lr 0.000441 time 0.2776 (0.2871) loss 3.6745 (3.4624) grad_norm 1.5774 (1.6401) [2021-04-16 06:48:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][430/1251] eta 0:03:55 lr 0.000441 time 0.2674 (0.2868) loss 3.5391 (3.4686) grad_norm 1.4852 (1.6379) [2021-04-16 06:48:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][440/1251] eta 0:03:52 lr 0.000441 time 0.2643 (0.2864) loss 4.0512 (3.4730) grad_norm 1.7098 (1.6363) [2021-04-16 06:48:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][450/1251] eta 0:03:49 lr 0.000441 time 0.4093 (0.2865) loss 4.2987 (3.4729) grad_norm 1.8815 (1.6385) [2021-04-16 06:48:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][460/1251] eta 0:03:46 lr 0.000441 time 0.2756 (0.2866) loss 3.6075 (3.4752) grad_norm 1.7628 (1.6366) [2021-04-16 06:48:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][470/1251] eta 0:03:43 lr 0.000441 time 0.2514 (0.2864) loss 2.1319 (3.4681) grad_norm 1.4351 (1.6356) [2021-04-16 06:48:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][480/1251] eta 0:03:40 lr 0.000441 time 0.2717 (0.2863) loss 2.8761 (3.4652) grad_norm 1.7657 (1.6352) [2021-04-16 06:48:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][490/1251] eta 0:03:38 lr 0.000441 time 0.2966 (0.2865) loss 3.5884 (3.4647) grad_norm 1.7036 (1.6359) [2021-04-16 06:48:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][500/1251] eta 0:03:35 lr 0.000441 time 0.2913 (0.2863) loss 3.3342 (3.4618) grad_norm 1.5070 (1.6339) [2021-04-16 06:48:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][510/1251] eta 0:03:32 lr 0.000441 time 0.2830 (0.2861) loss 3.5629 (3.4648) grad_norm 1.7354 (1.6329) [2021-04-16 06:48:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][520/1251] eta 0:03:28 lr 0.000441 time 0.2625 (0.2859) loss 4.1181 (3.4629) grad_norm 1.7070 (1.6333) [2021-04-16 06:48:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][530/1251] eta 0:03:26 lr 0.000441 time 0.2681 (0.2857) loss 3.8206 (3.4607) grad_norm 1.5540 (1.6326) [2021-04-16 06:48:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][540/1251] eta 0:03:23 lr 0.000441 time 0.2777 (0.2856) loss 3.8731 (3.4606) grad_norm 1.4574 (1.6318) [2021-04-16 06:48:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][550/1251] eta 0:03:20 lr 0.000441 time 0.4167 (0.2856) loss 3.8330 (3.4633) grad_norm 1.5962 (1.6336) [2021-04-16 06:48:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][560/1251] eta 0:03:17 lr 0.000441 time 0.2765 (0.2854) loss 4.0362 (3.4699) grad_norm 1.6128 (1.6345) [2021-04-16 06:48:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][570/1251] eta 0:03:14 lr 0.000441 time 0.2706 (0.2853) loss 4.5343 (3.4720) grad_norm 1.8240 (1.6340) [2021-04-16 06:48:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][580/1251] eta 0:03:11 lr 0.000441 time 0.2816 (0.2851) loss 3.8616 (3.4673) grad_norm 1.5160 (1.6332) [2021-04-16 06:48:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][590/1251] eta 0:03:08 lr 0.000441 time 0.2543 (0.2851) loss 3.4478 (3.4688) grad_norm 1.5395 (1.6320) [2021-04-16 06:48:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][600/1251] eta 0:03:05 lr 0.000440 time 0.2885 (0.2850) loss 2.7702 (3.4678) grad_norm 1.5148 (1.6317) [2021-04-16 06:48:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][610/1251] eta 0:03:02 lr 0.000440 time 0.2766 (0.2849) loss 2.3498 (3.4660) grad_norm 1.8238 (1.6311) [2021-04-16 06:48:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][620/1251] eta 0:02:59 lr 0.000440 time 0.2998 (0.2847) loss 3.3249 (3.4691) grad_norm 1.5840 (1.6314) [2021-04-16 06:48:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][630/1251] eta 0:02:56 lr 0.000440 time 0.2574 (0.2848) loss 4.1798 (3.4704) grad_norm 1.5631 (1.6317) [2021-04-16 06:49:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][640/1251] eta 0:02:53 lr 0.000440 time 0.2909 (0.2847) loss 3.3331 (3.4692) grad_norm 1.9318 (1.6320) [2021-04-16 06:49:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][650/1251] eta 0:02:51 lr 0.000440 time 0.2626 (0.2846) loss 2.9283 (3.4653) grad_norm 1.6168 (1.6309) [2021-04-16 06:49:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][660/1251] eta 0:02:48 lr 0.000440 time 0.2887 (0.2845) loss 3.4866 (3.4653) grad_norm 1.4292 (1.6306) [2021-04-16 06:49:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][670/1251] eta 0:02:45 lr 0.000440 time 0.2675 (0.2845) loss 4.1008 (3.4644) grad_norm 1.5988 (1.6303) [2021-04-16 06:49:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][680/1251] eta 0:02:42 lr 0.000440 time 0.2679 (0.2844) loss 2.8170 (3.4606) grad_norm 1.9085 (1.6321) [2021-04-16 06:49:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][690/1251] eta 0:02:39 lr 0.000440 time 0.2630 (0.2843) loss 3.5486 (3.4551) grad_norm 1.6474 (1.6311) [2021-04-16 06:49:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][700/1251] eta 0:02:36 lr 0.000440 time 0.2585 (0.2842) loss 3.8865 (3.4531) grad_norm 1.7692 (1.6307) [2021-04-16 06:49:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][710/1251] eta 0:02:33 lr 0.000440 time 0.4181 (0.2843) loss 2.5046 (3.4529) grad_norm 1.4893 (1.6300) [2021-04-16 06:49:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][720/1251] eta 0:02:30 lr 0.000440 time 0.2945 (0.2842) loss 3.3526 (3.4559) grad_norm 1.4989 (1.6310) [2021-04-16 06:49:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][730/1251] eta 0:02:28 lr 0.000440 time 0.2663 (0.2843) loss 3.5219 (3.4554) grad_norm 1.6216 (1.6296) [2021-04-16 06:49:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][740/1251] eta 0:02:25 lr 0.000440 time 0.2850 (0.2842) loss 4.2663 (3.4611) grad_norm 1.4002 (1.6297) [2021-04-16 06:49:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][750/1251] eta 0:02:22 lr 0.000440 time 0.2502 (0.2842) loss 3.1465 (3.4664) grad_norm 2.1230 (1.6296) [2021-04-16 06:49:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][760/1251] eta 0:02:19 lr 0.000440 time 0.2927 (0.2842) loss 3.9101 (3.4687) grad_norm 1.9147 (1.6303) [2021-04-16 06:49:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][770/1251] eta 0:02:16 lr 0.000440 time 0.2654 (0.2840) loss 4.0071 (3.4687) grad_norm 1.6953 (1.6294) [2021-04-16 06:49:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][780/1251] eta 0:02:13 lr 0.000440 time 0.2677 (0.2840) loss 2.2595 (3.4683) grad_norm 1.5644 (1.6283) [2021-04-16 06:49:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][790/1251] eta 0:02:10 lr 0.000440 time 0.3065 (0.2840) loss 3.2176 (3.4699) grad_norm 1.3867 (1.6271) [2021-04-16 06:49:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][800/1251] eta 0:02:08 lr 0.000440 time 0.2729 (0.2839) loss 3.8554 (3.4699) grad_norm 1.3829 (1.6255) [2021-04-16 06:49:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][810/1251] eta 0:02:05 lr 0.000440 time 0.3075 (0.2839) loss 3.3397 (3.4721) grad_norm 1.7497 (1.6244) [2021-04-16 06:49:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][820/1251] eta 0:02:02 lr 0.000440 time 0.2654 (0.2838) loss 4.1295 (3.4755) grad_norm 1.5727 (1.6241) [2021-04-16 06:49:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][830/1251] eta 0:01:59 lr 0.000440 time 0.2779 (0.2837) loss 3.4067 (3.4728) grad_norm 1.6317 (1.6237) [2021-04-16 06:49:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][840/1251] eta 0:01:56 lr 0.000440 time 0.2624 (0.2837) loss 3.8178 (3.4745) grad_norm 1.8960 (1.6242) [2021-04-16 06:49:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][850/1251] eta 0:01:53 lr 0.000439 time 0.2915 (0.2837) loss 3.6126 (3.4763) grad_norm 1.3200 (1.6239) [2021-04-16 06:50:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][860/1251] eta 0:01:50 lr 0.000439 time 0.2689 (0.2836) loss 3.8857 (3.4775) grad_norm 1.5516 (1.6222) [2021-04-16 06:50:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][870/1251] eta 0:01:48 lr 0.000439 time 0.2623 (0.2835) loss 3.2906 (3.4756) grad_norm 1.6924 (1.6224) [2021-04-16 06:50:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][880/1251] eta 0:01:45 lr 0.000439 time 0.2536 (0.2835) loss 4.1224 (3.4779) grad_norm 1.3605 (1.6211) [2021-04-16 06:50:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][890/1251] eta 0:01:42 lr 0.000439 time 0.2668 (0.2835) loss 3.3290 (3.4780) grad_norm 1.7349 (1.6206) [2021-04-16 06:50:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][900/1251] eta 0:01:39 lr 0.000439 time 0.3157 (0.2835) loss 3.5019 (3.4766) grad_norm 1.5360 (1.6203) [2021-04-16 06:50:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][910/1251] eta 0:01:36 lr 0.000439 time 0.2880 (0.2834) loss 3.7478 (3.4748) grad_norm 1.6817 (1.6202) [2021-04-16 06:50:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][920/1251] eta 0:01:33 lr 0.000439 time 0.2673 (0.2835) loss 4.1578 (3.4746) grad_norm 1.6750 (1.6197) [2021-04-16 06:50:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][930/1251] eta 0:01:30 lr 0.000439 time 0.2856 (0.2834) loss 2.3189 (3.4722) grad_norm 1.9068 (1.6205) [2021-04-16 06:50:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][940/1251] eta 0:01:28 lr 0.000439 time 0.2635 (0.2836) loss 3.6990 (3.4704) grad_norm 1.5778 (1.6209) [2021-04-16 06:50:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][950/1251] eta 0:01:25 lr 0.000439 time 0.2929 (0.2835) loss 2.3454 (3.4699) grad_norm 1.7397 (1.6210) [2021-04-16 06:50:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][960/1251] eta 0:01:22 lr 0.000439 time 0.2686 (0.2835) loss 3.2353 (3.4646) grad_norm 1.8036 (1.6217) [2021-04-16 06:50:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][970/1251] eta 0:01:19 lr 0.000439 time 0.2975 (0.2834) loss 3.8481 (3.4635) grad_norm 1.5176 (1.6209) [2021-04-16 06:50:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][980/1251] eta 0:01:16 lr 0.000439 time 0.2873 (0.2834) loss 3.7580 (3.4636) grad_norm 1.4054 (1.6202) [2021-04-16 06:50:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][990/1251] eta 0:01:13 lr 0.000439 time 0.2669 (0.2833) loss 2.9040 (3.4660) grad_norm 1.8480 (1.6206) [2021-04-16 06:50:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][1000/1251] eta 0:01:11 lr 0.000439 time 0.3096 (0.2833) loss 3.1237 (3.4665) grad_norm 1.3867 (1.6210) [2021-04-16 06:50:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][1010/1251] eta 0:01:08 lr 0.000439 time 0.2919 (0.2833) loss 2.8832 (3.4684) grad_norm 1.6949 (1.6202) [2021-04-16 06:50:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][1020/1251] eta 0:01:05 lr 0.000439 time 0.2742 (0.2832) loss 3.9831 (3.4692) grad_norm 1.5808 (1.6205) [2021-04-16 06:50:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][1030/1251] eta 0:01:02 lr 0.000439 time 0.2785 (0.2833) loss 3.7497 (3.4692) grad_norm 1.7296 (1.6217) [2021-04-16 06:50:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][1040/1251] eta 0:00:59 lr 0.000439 time 0.2676 (0.2832) loss 3.7852 (3.4685) grad_norm 1.9678 (1.6226) [2021-04-16 06:50:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][1050/1251] eta 0:00:56 lr 0.000439 time 0.2709 (0.2831) loss 3.6254 (3.4685) grad_norm 1.5102 (1.6231) [2021-04-16 06:50:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][1060/1251] eta 0:00:54 lr 0.000439 time 0.2739 (0.2830) loss 3.0900 (3.4649) grad_norm 1.7563 (1.6234) [2021-04-16 06:51:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][1070/1251] eta 0:00:51 lr 0.000439 time 0.2677 (0.2831) loss 4.1032 (3.4655) grad_norm 1.4202 (1.6235) [2021-04-16 06:51:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][1080/1251] eta 0:00:48 lr 0.000439 time 0.2914 (0.2831) loss 3.0705 (3.4653) grad_norm 1.5479 (1.6227) [2021-04-16 06:51:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][1090/1251] eta 0:00:45 lr 0.000438 time 0.2709 (0.2831) loss 2.5382 (3.4652) grad_norm 1.4723 (1.6220) [2021-04-16 06:51:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][1100/1251] eta 0:00:42 lr 0.000438 time 0.2704 (0.2830) loss 3.8923 (3.4670) grad_norm 1.3477 (1.6216) [2021-04-16 06:51:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][1110/1251] eta 0:00:39 lr 0.000438 time 0.2813 (0.2830) loss 4.0198 (3.4672) grad_norm 1.5581 (1.6217) [2021-04-16 06:51:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][1120/1251] eta 0:00:37 lr 0.000438 time 0.2859 (0.2829) loss 2.8242 (3.4666) grad_norm 1.6300 (1.6222) [2021-04-16 06:51:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][1130/1251] eta 0:00:34 lr 0.000438 time 0.2675 (0.2830) loss 3.6469 (3.4655) grad_norm 1.5894 (1.6222) [2021-04-16 06:51:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][1140/1251] eta 0:00:31 lr 0.000438 time 0.2855 (0.2829) loss 3.8894 (3.4645) grad_norm 1.4434 (1.6211) [2021-04-16 06:51:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][1150/1251] eta 0:00:28 lr 0.000438 time 0.2962 (0.2828) loss 3.8046 (3.4660) grad_norm 1.5107 (1.6206) [2021-04-16 06:51:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][1160/1251] eta 0:00:25 lr 0.000438 time 0.3031 (0.2828) loss 3.7493 (3.4660) grad_norm 1.6053 (1.6204) [2021-04-16 06:51:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][1170/1251] eta 0:00:22 lr 0.000438 time 0.2533 (0.2829) loss 3.9042 (3.4684) grad_norm 1.4757 (1.6198) [2021-04-16 06:51:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][1180/1251] eta 0:00:20 lr 0.000438 time 0.2865 (0.2828) loss 3.9083 (3.4688) grad_norm 1.8187 (1.6201) [2021-04-16 06:51:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][1190/1251] eta 0:00:17 lr 0.000438 time 0.2817 (0.2828) loss 3.7158 (3.4682) grad_norm 1.5196 (1.6205) [2021-04-16 06:51:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][1200/1251] eta 0:00:14 lr 0.000438 time 0.2650 (0.2828) loss 3.6426 (3.4673) grad_norm 1.7291 (1.6211) [2021-04-16 06:51:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][1210/1251] eta 0:00:11 lr 0.000438 time 0.2639 (0.2828) loss 2.8360 (3.4674) grad_norm 1.6353 (1.6215) [2021-04-16 06:51:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][1220/1251] eta 0:00:08 lr 0.000438 time 0.2871 (0.2828) loss 3.2252 (3.4689) grad_norm 1.7010 (1.6221) [2021-04-16 06:51:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][1230/1251] eta 0:00:05 lr 0.000438 time 0.2907 (0.2828) loss 2.7384 (3.4671) grad_norm 1.5254 (1.6215) [2021-04-16 06:51:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][1240/1251] eta 0:00:03 lr 0.000438 time 0.3196 (0.2827) loss 3.0811 (3.4676) grad_norm 1.6925 (1.6212) [2021-04-16 06:51:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [162/300][1250/1251] eta 0:00:00 lr 0.000438 time 0.2442 (0.2825) loss 4.0412 (3.4701) grad_norm 1.8506 (1.6214) [2021-04-16 06:51:54 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 162 training takes 0:05:56 [2021-04-16 06:51:54 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_162.pth saving...... [2021-04-16 06:52:10 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_162.pth saved !!! [2021-04-16 06:52:11 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.123 (1.123) Loss 0.9617 (0.9617) Acc@1 76.855 (76.855) Acc@5 94.434 (94.434) [2021-04-16 06:52:12 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.105 (0.251) Loss 0.9646 (1.0032) Acc@1 77.930 (76.536) Acc@5 93.457 (93.368) [2021-04-16 06:52:15 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.420 (0.235) Loss 0.9431 (1.0068) Acc@1 79.004 (76.511) Acc@5 93.848 (93.406) [2021-04-16 06:52:17 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.132 (0.226) Loss 0.9863 (1.0001) Acc@1 75.781 (76.443) Acc@5 93.848 (93.470) [2021-04-16 06:52:19 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.221) Loss 1.0586 (0.9981) Acc@1 75.391 (76.472) Acc@5 92.871 (93.514) [2021-04-16 06:52:23 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 76.606 Acc@5 93.576 [2021-04-16 06:52:23 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 76.6% [2021-04-16 06:52:23 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 76.63% [2021-04-16 06:52:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][0/1251] eta 1:51:22 lr 0.000438 time 5.3420 (5.3420) loss 3.2501 (3.2501) grad_norm 1.7262 (1.7262) [2021-04-16 06:52:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][10/1251] eta 0:15:12 lr 0.000438 time 0.2932 (0.7350) loss 3.5883 (3.6245) grad_norm 1.5378 (1.5630) [2021-04-16 06:52:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][20/1251] eta 0:10:41 lr 0.000438 time 0.2738 (0.5215) loss 3.8079 (3.6358) grad_norm 1.5380 (1.5660) [2021-04-16 06:52:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][30/1251] eta 0:09:02 lr 0.000438 time 0.3050 (0.4440) loss 3.8620 (3.5647) grad_norm 1.6280 (1.5823) [2021-04-16 06:52:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][40/1251] eta 0:08:07 lr 0.000438 time 0.2861 (0.4028) loss 3.7903 (3.5570) grad_norm 1.7170 (1.5845) [2021-04-16 06:52:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][50/1251] eta 0:07:34 lr 0.000438 time 0.2666 (0.3784) loss 3.0876 (3.5220) grad_norm 1.5124 (1.5719) [2021-04-16 06:52:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][60/1251] eta 0:07:11 lr 0.000438 time 0.2964 (0.3620) loss 2.8338 (3.5041) grad_norm 1.5899 (1.5719) [2021-04-16 06:52:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][70/1251] eta 0:06:53 lr 0.000438 time 0.2641 (0.3503) loss 4.0531 (3.4995) grad_norm 1.6139 (1.5796) [2021-04-16 06:52:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][80/1251] eta 0:06:40 lr 0.000437 time 0.2697 (0.3417) loss 4.5961 (3.5018) grad_norm 1.4486 (1.5778) [2021-04-16 06:52:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][90/1251] eta 0:06:27 lr 0.000437 time 0.2512 (0.3334) loss 3.9985 (3.5304) grad_norm 1.8313 (nan) [2021-04-16 06:52:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][100/1251] eta 0:06:19 lr 0.000437 time 0.2768 (0.3300) loss 2.3174 (3.4986) grad_norm 1.5388 (nan) [2021-04-16 06:53:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][110/1251] eta 0:06:11 lr 0.000437 time 0.2810 (0.3257) loss 3.6443 (3.4845) grad_norm 1.8877 (nan) [2021-04-16 06:53:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][120/1251] eta 0:06:06 lr 0.000437 time 0.2770 (0.3240) loss 3.6948 (3.4884) grad_norm 1.8213 (nan) [2021-04-16 06:53:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][130/1251] eta 0:05:59 lr 0.000437 time 0.3020 (0.3205) loss 3.1990 (3.5041) grad_norm 1.8389 (nan) [2021-04-16 06:53:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][140/1251] eta 0:05:53 lr 0.000437 time 0.2902 (0.3182) loss 3.2436 (3.4947) grad_norm 2.0626 (nan) [2021-04-16 06:53:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][150/1251] eta 0:05:47 lr 0.000437 time 0.2805 (0.3158) loss 2.0599 (3.4782) grad_norm 1.4892 (nan) [2021-04-16 06:53:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][160/1251] eta 0:05:42 lr 0.000437 time 0.3050 (0.3142) loss 3.2038 (3.4683) grad_norm 1.7156 (nan) [2021-04-16 06:53:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][170/1251] eta 0:05:37 lr 0.000437 time 0.2924 (0.3124) loss 2.7665 (3.4482) grad_norm 1.5911 (nan) [2021-04-16 06:53:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][180/1251] eta 0:05:32 lr 0.000437 time 0.2773 (0.3106) loss 3.6100 (3.4606) grad_norm 1.5631 (nan) [2021-04-16 06:53:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][190/1251] eta 0:05:27 lr 0.000437 time 0.2756 (0.3090) loss 3.7112 (3.4604) grad_norm 1.3849 (nan) [2021-04-16 06:53:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][200/1251] eta 0:05:23 lr 0.000437 time 0.2559 (0.3074) loss 3.9861 (3.4616) grad_norm 1.5932 (nan) [2021-04-16 06:53:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][210/1251] eta 0:05:18 lr 0.000437 time 0.2903 (0.3059) loss 4.5335 (3.4617) grad_norm 1.5875 (nan) [2021-04-16 06:53:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][220/1251] eta 0:05:14 lr 0.000437 time 0.2867 (0.3049) loss 3.5725 (3.4621) grad_norm 1.5524 (nan) [2021-04-16 06:53:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][230/1251] eta 0:05:09 lr 0.000437 time 0.2666 (0.3034) loss 3.4229 (3.4550) grad_norm 1.4710 (nan) [2021-04-16 06:53:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][240/1251] eta 0:05:05 lr 0.000437 time 0.2773 (0.3026) loss 3.6297 (3.4443) grad_norm 1.5689 (nan) [2021-04-16 06:53:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][250/1251] eta 0:05:02 lr 0.000437 time 0.2814 (0.3023) loss 3.8826 (3.4455) grad_norm 1.6102 (nan) [2021-04-16 06:53:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][260/1251] eta 0:04:58 lr 0.000437 time 0.2637 (0.3014) loss 4.1369 (3.4618) grad_norm 1.4791 (nan) [2021-04-16 06:53:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][270/1251] eta 0:04:55 lr 0.000437 time 0.2802 (0.3007) loss 3.2715 (3.4644) grad_norm 1.5632 (nan) [2021-04-16 06:53:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][280/1251] eta 0:04:51 lr 0.000437 time 0.2663 (0.3000) loss 3.6315 (3.4593) grad_norm 1.6956 (nan) [2021-04-16 06:53:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][290/1251] eta 0:04:48 lr 0.000437 time 0.2914 (0.2998) loss 4.0120 (3.4620) grad_norm 1.5990 (nan) [2021-04-16 06:53:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][300/1251] eta 0:04:44 lr 0.000437 time 0.2914 (0.2990) loss 3.9790 (3.4542) grad_norm 1.4860 (nan) [2021-04-16 06:53:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][310/1251] eta 0:04:40 lr 0.000437 time 0.2622 (0.2984) loss 3.7741 (3.4577) grad_norm 1.5695 (nan) [2021-04-16 06:53:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][320/1251] eta 0:04:37 lr 0.000437 time 0.2876 (0.2979) loss 3.5212 (3.4615) grad_norm 1.4966 (nan) [2021-04-16 06:54:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][330/1251] eta 0:04:33 lr 0.000436 time 0.2579 (0.2973) loss 4.0095 (3.4716) grad_norm 1.7951 (nan) [2021-04-16 06:54:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][340/1251] eta 0:04:30 lr 0.000436 time 0.2787 (0.2972) loss 4.3918 (3.4663) grad_norm 2.3434 (nan) [2021-04-16 06:54:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][350/1251] eta 0:04:27 lr 0.000436 time 0.2836 (0.2966) loss 3.6896 (3.4692) grad_norm 1.6249 (nan) [2021-04-16 06:54:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][360/1251] eta 0:04:23 lr 0.000436 time 0.2677 (0.2960) loss 3.8367 (3.4623) grad_norm 1.4800 (nan) [2021-04-16 06:54:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][370/1251] eta 0:04:20 lr 0.000436 time 0.2548 (0.2958) loss 4.1326 (3.4559) grad_norm 1.7383 (nan) [2021-04-16 06:54:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][380/1251] eta 0:04:17 lr 0.000436 time 0.2621 (0.2955) loss 3.8197 (3.4672) grad_norm 1.6587 (nan) [2021-04-16 06:54:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][390/1251] eta 0:04:13 lr 0.000436 time 0.2731 (0.2949) loss 4.2753 (3.4701) grad_norm 1.6109 (nan) [2021-04-16 06:54:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][400/1251] eta 0:04:10 lr 0.000436 time 0.2788 (0.2944) loss 4.0911 (3.4729) grad_norm 2.0173 (nan) [2021-04-16 06:54:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][410/1251] eta 0:04:07 lr 0.000436 time 0.3113 (0.2940) loss 3.6026 (3.4682) grad_norm 1.4520 (nan) [2021-04-16 06:54:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][420/1251] eta 0:04:04 lr 0.000436 time 0.2783 (0.2937) loss 2.9609 (3.4647) grad_norm 1.5377 (nan) [2021-04-16 06:54:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][430/1251] eta 0:04:00 lr 0.000436 time 0.3200 (0.2934) loss 3.2283 (3.4659) grad_norm 1.6180 (nan) [2021-04-16 06:54:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][440/1251] eta 0:03:57 lr 0.000436 time 0.2538 (0.2932) loss 3.3082 (3.4657) grad_norm 1.6722 (nan) [2021-04-16 06:54:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][450/1251] eta 0:03:54 lr 0.000436 time 0.4309 (0.2931) loss 3.5868 (3.4593) grad_norm 1.6434 (nan) [2021-04-16 06:54:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][460/1251] eta 0:03:51 lr 0.000436 time 0.2752 (0.2927) loss 3.5058 (3.4629) grad_norm 1.5418 (nan) [2021-04-16 06:54:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][470/1251] eta 0:03:48 lr 0.000436 time 0.2826 (0.2923) loss 3.5881 (3.4666) grad_norm 1.6191 (nan) [2021-04-16 06:54:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][480/1251] eta 0:03:45 lr 0.000436 time 0.2908 (0.2921) loss 4.0523 (3.4693) grad_norm 1.6442 (nan) [2021-04-16 06:54:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][490/1251] eta 0:03:42 lr 0.000436 time 0.2529 (0.2918) loss 3.1914 (3.4685) grad_norm 1.7574 (nan) [2021-04-16 06:54:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][500/1251] eta 0:03:39 lr 0.000436 time 0.2925 (0.2917) loss 3.4705 (3.4708) grad_norm 1.5750 (nan) [2021-04-16 06:54:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][510/1251] eta 0:03:35 lr 0.000436 time 0.2729 (0.2914) loss 3.9957 (3.4726) grad_norm 1.6929 (nan) [2021-04-16 06:54:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][520/1251] eta 0:03:32 lr 0.000436 time 0.2947 (0.2911) loss 4.0338 (3.4753) grad_norm 1.4396 (nan) [2021-04-16 06:54:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][530/1251] eta 0:03:29 lr 0.000436 time 0.2813 (0.2909) loss 3.7774 (3.4740) grad_norm 1.8480 (nan) [2021-04-16 06:55:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][540/1251] eta 0:03:26 lr 0.000436 time 0.2787 (0.2906) loss 3.3411 (3.4720) grad_norm 1.9070 (nan) [2021-04-16 06:55:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][550/1251] eta 0:03:23 lr 0.000436 time 0.3689 (0.2906) loss 2.7537 (3.4692) grad_norm 1.5425 (nan) [2021-04-16 06:55:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][560/1251] eta 0:03:20 lr 0.000436 time 0.2656 (0.2904) loss 3.6904 (3.4744) grad_norm 1.4936 (nan) [2021-04-16 06:55:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][570/1251] eta 0:03:17 lr 0.000435 time 0.2595 (0.2906) loss 3.8157 (3.4755) grad_norm 1.4686 (nan) [2021-04-16 06:55:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][580/1251] eta 0:03:14 lr 0.000435 time 0.2781 (0.2904) loss 3.5728 (3.4806) grad_norm 1.6765 (nan) [2021-04-16 06:55:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][590/1251] eta 0:03:11 lr 0.000435 time 0.2584 (0.2901) loss 3.5098 (3.4793) grad_norm 1.4077 (nan) [2021-04-16 06:55:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][600/1251] eta 0:03:08 lr 0.000435 time 0.2836 (0.2899) loss 4.1574 (3.4803) grad_norm 1.5558 (nan) [2021-04-16 06:55:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][610/1251] eta 0:03:05 lr 0.000435 time 0.2644 (0.2897) loss 4.2065 (3.4809) grad_norm 1.5689 (nan) [2021-04-16 06:55:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][620/1251] eta 0:03:02 lr 0.000435 time 0.2948 (0.2896) loss 3.7537 (3.4761) grad_norm 1.6601 (nan) [2021-04-16 06:55:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][630/1251] eta 0:02:59 lr 0.000435 time 0.2748 (0.2894) loss 3.5062 (3.4722) grad_norm 1.7347 (nan) [2021-04-16 06:55:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][640/1251] eta 0:02:56 lr 0.000435 time 0.2985 (0.2893) loss 3.9984 (3.4714) grad_norm 1.4273 (nan) [2021-04-16 06:55:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][650/1251] eta 0:02:53 lr 0.000435 time 0.2757 (0.2890) loss 3.9154 (3.4780) grad_norm 1.7841 (nan) [2021-04-16 06:55:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][660/1251] eta 0:02:50 lr 0.000435 time 0.2962 (0.2889) loss 3.0667 (3.4749) grad_norm 1.6228 (nan) [2021-04-16 06:55:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][670/1251] eta 0:02:47 lr 0.000435 time 0.2813 (0.2888) loss 3.4913 (3.4728) grad_norm 1.7900 (nan) [2021-04-16 06:55:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][680/1251] eta 0:02:44 lr 0.000435 time 0.2759 (0.2888) loss 3.1460 (3.4784) grad_norm 1.5887 (nan) [2021-04-16 06:55:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][690/1251] eta 0:02:42 lr 0.000435 time 0.2982 (0.2889) loss 3.7614 (3.4770) grad_norm 1.4506 (nan) [2021-04-16 06:55:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][700/1251] eta 0:02:39 lr 0.000435 time 0.2736 (0.2887) loss 3.5118 (3.4787) grad_norm 1.6946 (nan) [2021-04-16 06:55:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][710/1251] eta 0:02:36 lr 0.000435 time 0.3067 (0.2886) loss 3.3104 (3.4780) grad_norm 1.5030 (nan) [2021-04-16 06:55:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][720/1251] eta 0:02:33 lr 0.000435 time 0.2953 (0.2887) loss 4.1077 (3.4806) grad_norm 1.6009 (nan) [2021-04-16 06:55:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][730/1251] eta 0:02:30 lr 0.000435 time 0.2742 (0.2886) loss 2.9291 (3.4795) grad_norm 1.5021 (nan) [2021-04-16 06:55:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][740/1251] eta 0:02:27 lr 0.000435 time 0.2937 (0.2886) loss 2.8119 (3.4764) grad_norm 1.5245 (nan) [2021-04-16 06:56:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][750/1251] eta 0:02:24 lr 0.000435 time 0.2647 (0.2883) loss 3.8420 (3.4756) grad_norm 1.5340 (nan) [2021-04-16 06:56:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][760/1251] eta 0:02:21 lr 0.000435 time 0.2813 (0.2882) loss 4.2125 (3.4761) grad_norm 1.5780 (nan) [2021-04-16 06:56:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][770/1251] eta 0:02:18 lr 0.000435 time 0.2712 (0.2880) loss 3.9173 (3.4736) grad_norm 1.5277 (nan) [2021-04-16 06:56:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][780/1251] eta 0:02:15 lr 0.000435 time 0.2563 (0.2879) loss 3.7713 (3.4704) grad_norm 1.6587 (nan) [2021-04-16 06:56:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][790/1251] eta 0:02:12 lr 0.000435 time 0.3027 (0.2878) loss 3.4242 (3.4713) grad_norm 1.7202 (nan) [2021-04-16 06:56:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][800/1251] eta 0:02:09 lr 0.000435 time 0.2790 (0.2878) loss 4.1428 (3.4701) grad_norm 1.3833 (nan) [2021-04-16 06:56:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][810/1251] eta 0:02:06 lr 0.000434 time 0.2663 (0.2877) loss 4.0038 (3.4723) grad_norm 1.5061 (nan) [2021-04-16 06:56:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][820/1251] eta 0:02:03 lr 0.000434 time 0.2885 (0.2877) loss 3.9969 (3.4736) grad_norm 1.6160 (nan) [2021-04-16 06:56:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][830/1251] eta 0:02:01 lr 0.000434 time 0.2572 (0.2876) loss 3.7388 (3.4775) grad_norm 1.5353 (nan) [2021-04-16 06:56:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][840/1251] eta 0:01:58 lr 0.000434 time 0.2720 (0.2875) loss 2.8571 (3.4766) grad_norm 1.6399 (nan) [2021-04-16 06:56:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][850/1251] eta 0:01:55 lr 0.000434 time 0.2467 (0.2873) loss 3.6591 (3.4770) grad_norm 1.4042 (nan) [2021-04-16 06:56:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][860/1251] eta 0:01:52 lr 0.000434 time 0.2694 (0.2873) loss 3.1604 (3.4798) grad_norm 1.6145 (nan) [2021-04-16 06:56:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][870/1251] eta 0:01:49 lr 0.000434 time 0.2873 (0.2873) loss 3.1502 (3.4779) grad_norm 1.8517 (nan) [2021-04-16 06:56:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][880/1251] eta 0:01:46 lr 0.000434 time 0.2662 (0.2872) loss 3.2154 (3.4771) grad_norm 1.6201 (nan) [2021-04-16 06:56:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][890/1251] eta 0:01:43 lr 0.000434 time 0.2535 (0.2870) loss 3.9824 (3.4770) grad_norm 1.5813 (nan) [2021-04-16 06:56:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][900/1251] eta 0:01:40 lr 0.000434 time 0.2576 (0.2869) loss 3.0510 (3.4748) grad_norm 1.5899 (nan) [2021-04-16 06:56:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][910/1251] eta 0:01:37 lr 0.000434 time 0.2703 (0.2869) loss 4.0924 (3.4747) grad_norm 1.6223 (nan) [2021-04-16 06:56:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][920/1251] eta 0:01:35 lr 0.000434 time 0.2782 (0.2870) loss 3.6566 (3.4710) grad_norm 1.9671 (nan) [2021-04-16 06:56:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][930/1251] eta 0:01:32 lr 0.000434 time 0.2903 (0.2869) loss 3.5456 (3.4727) grad_norm 1.6984 (nan) [2021-04-16 06:56:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][940/1251] eta 0:01:29 lr 0.000434 time 0.3871 (0.2869) loss 4.0948 (3.4748) grad_norm 1.6133 (nan) [2021-04-16 06:56:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][950/1251] eta 0:01:26 lr 0.000434 time 0.2814 (0.2868) loss 3.3306 (3.4739) grad_norm 1.3554 (nan) [2021-04-16 06:56:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][960/1251] eta 0:01:23 lr 0.000434 time 0.2556 (0.2867) loss 3.9848 (3.4756) grad_norm 1.5512 (nan) [2021-04-16 06:57:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][970/1251] eta 0:01:20 lr 0.000434 time 0.2772 (0.2867) loss 3.9392 (3.4787) grad_norm 1.5869 (nan) [2021-04-16 06:57:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][980/1251] eta 0:01:17 lr 0.000434 time 0.2807 (0.2866) loss 3.9854 (3.4824) grad_norm 1.5793 (nan) [2021-04-16 06:57:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][990/1251] eta 0:01:14 lr 0.000434 time 0.2726 (0.2866) loss 2.9964 (3.4818) grad_norm 1.8553 (nan) [2021-04-16 06:57:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][1000/1251] eta 0:01:11 lr 0.000434 time 0.2646 (0.2865) loss 2.4126 (3.4845) grad_norm 1.3966 (nan) [2021-04-16 06:57:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][1010/1251] eta 0:01:09 lr 0.000434 time 0.2613 (0.2863) loss 3.8384 (3.4855) grad_norm 1.3198 (nan) [2021-04-16 06:57:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][1020/1251] eta 0:01:06 lr 0.000434 time 0.3044 (0.2863) loss 4.0354 (3.4866) grad_norm 1.7693 (nan) [2021-04-16 06:57:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][1030/1251] eta 0:01:03 lr 0.000434 time 0.2663 (0.2863) loss 3.7570 (3.4870) grad_norm 1.5440 (nan) [2021-04-16 06:57:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][1040/1251] eta 0:01:00 lr 0.000434 time 0.2673 (0.2862) loss 2.8152 (3.4878) grad_norm 1.3888 (nan) [2021-04-16 06:57:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][1050/1251] eta 0:00:57 lr 0.000434 time 0.2758 (0.2861) loss 3.7863 (3.4867) grad_norm 1.6836 (nan) [2021-04-16 06:57:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][1060/1251] eta 0:00:54 lr 0.000433 time 0.2987 (0.2862) loss 2.9058 (3.4867) grad_norm 1.9570 (nan) [2021-04-16 06:57:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][1070/1251] eta 0:00:51 lr 0.000433 time 0.2881 (0.2861) loss 2.1028 (3.4859) grad_norm 1.6825 (nan) [2021-04-16 06:57:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][1080/1251] eta 0:00:48 lr 0.000433 time 0.2653 (0.2860) loss 3.9074 (3.4865) grad_norm 1.6678 (nan) [2021-04-16 06:57:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][1090/1251] eta 0:00:46 lr 0.000433 time 0.2782 (0.2859) loss 3.9772 (3.4869) grad_norm 1.5235 (nan) [2021-04-16 06:57:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][1100/1251] eta 0:00:43 lr 0.000433 time 0.2891 (0.2858) loss 3.4852 (3.4858) grad_norm 1.5998 (nan) [2021-04-16 06:57:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][1110/1251] eta 0:00:40 lr 0.000433 time 0.2617 (0.2857) loss 3.9988 (3.4873) grad_norm 1.4557 (nan) [2021-04-16 06:57:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][1120/1251] eta 0:00:37 lr 0.000433 time 0.2905 (0.2857) loss 3.0846 (3.4876) grad_norm 1.6300 (nan) [2021-04-16 06:57:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][1130/1251] eta 0:00:34 lr 0.000433 time 0.2624 (0.2857) loss 3.5321 (3.4879) grad_norm 1.7323 (nan) [2021-04-16 06:57:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][1140/1251] eta 0:00:31 lr 0.000433 time 0.2904 (0.2858) loss 3.3669 (3.4864) grad_norm 1.5926 (nan) [2021-04-16 06:57:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][1150/1251] eta 0:00:28 lr 0.000433 time 0.2734 (0.2857) loss 2.9383 (3.4825) grad_norm 1.4146 (nan) [2021-04-16 06:57:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][1160/1251] eta 0:00:25 lr 0.000433 time 0.2814 (0.2856) loss 3.3341 (3.4823) grad_norm 1.5333 (nan) [2021-04-16 06:57:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][1170/1251] eta 0:00:23 lr 0.000433 time 0.2635 (0.2855) loss 4.0726 (3.4828) grad_norm 1.6126 (nan) [2021-04-16 06:58:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][1180/1251] eta 0:00:20 lr 0.000433 time 0.2926 (0.2854) loss 3.3775 (3.4843) grad_norm 1.5203 (nan) [2021-04-16 06:58:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][1190/1251] eta 0:00:17 lr 0.000433 time 0.2949 (0.2854) loss 4.2180 (3.4864) grad_norm 1.6169 (nan) [2021-04-16 06:58:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][1200/1251] eta 0:00:14 lr 0.000433 time 0.2721 (0.2854) loss 3.8578 (3.4873) grad_norm 2.8424 (nan) [2021-04-16 06:58:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][1210/1251] eta 0:00:11 lr 0.000433 time 0.2929 (0.2854) loss 4.0978 (3.4880) grad_norm 1.6632 (nan) [2021-04-16 06:58:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][1220/1251] eta 0:00:08 lr 0.000433 time 0.2863 (0.2854) loss 3.9108 (3.4890) grad_norm 1.6522 (nan) [2021-04-16 06:58:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][1230/1251] eta 0:00:05 lr 0.000433 time 0.2895 (0.2853) loss 4.4010 (3.4902) grad_norm 1.5949 (nan) [2021-04-16 06:58:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][1240/1251] eta 0:00:03 lr 0.000433 time 0.2493 (0.2852) loss 2.9233 (3.4876) grad_norm 1.4960 (nan) [2021-04-16 06:58:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [163/300][1250/1251] eta 0:00:00 lr 0.000433 time 0.2482 (0.2849) loss 3.9178 (3.4872) grad_norm 1.5340 (nan) [2021-04-16 06:58:23 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 163 training takes 0:05:59 [2021-04-16 06:58:23 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_163.pth saving...... [2021-04-16 06:58:32 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_163.pth saved !!! [2021-04-16 06:58:33 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.190 (1.190) Loss 1.0042 (1.0042) Acc@1 75.977 (75.977) Acc@5 92.969 (92.969) [2021-04-16 06:58:34 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.076 (0.213) Loss 0.9858 (1.0090) Acc@1 76.953 (76.580) Acc@5 93.359 (93.422) [2021-04-16 06:58:37 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.154 (0.242) Loss 0.9628 (1.0010) Acc@1 76.855 (76.693) Acc@5 93.262 (93.536) [2021-04-16 06:58:39 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.336 (0.222) Loss 0.9905 (1.0047) Acc@1 76.953 (76.711) Acc@5 93.262 (93.460) [2021-04-16 06:58:40 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.209 (0.216) Loss 0.9933 (1.0024) Acc@1 75.195 (76.770) Acc@5 94.824 (93.509) [2021-04-16 06:58:45 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 76.688 Acc@5 93.474 [2021-04-16 06:58:45 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 76.7% [2021-04-16 06:58:45 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 76.69% [2021-04-16 06:58:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][0/1251] eta 1:50:09 lr 0.000433 time 5.2837 (5.2837) loss 3.8220 (3.8220) grad_norm 1.9301 (1.9301) [2021-04-16 06:58:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][10/1251] eta 0:15:05 lr 0.000433 time 0.2949 (0.7297) loss 3.9680 (3.2684) grad_norm 2.0219 (1.6738) [2021-04-16 06:58:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][20/1251] eta 0:10:33 lr 0.000433 time 0.2689 (0.5150) loss 2.8622 (3.3377) grad_norm 1.6104 (1.7121) [2021-04-16 06:58:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][30/1251] eta 0:08:56 lr 0.000433 time 0.3157 (0.4395) loss 3.9734 (3.3813) grad_norm 1.5877 (1.6822) [2021-04-16 06:59:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][40/1251] eta 0:08:08 lr 0.000433 time 0.3003 (0.4038) loss 3.6790 (3.3581) grad_norm 1.7550 (1.6777) [2021-04-16 06:59:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][50/1251] eta 0:07:34 lr 0.000432 time 0.3005 (0.3784) loss 3.6012 (3.3938) grad_norm 1.4129 (1.6768) [2021-04-16 06:59:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][60/1251] eta 0:07:12 lr 0.000432 time 0.2877 (0.3631) loss 3.3782 (3.4334) grad_norm 1.7034 (1.6711) [2021-04-16 06:59:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][70/1251] eta 0:06:57 lr 0.000432 time 0.2933 (0.3536) loss 3.8186 (3.4387) grad_norm 1.6768 (1.6615) [2021-04-16 06:59:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][80/1251] eta 0:06:45 lr 0.000432 time 0.2694 (0.3459) loss 4.2229 (3.4322) grad_norm 1.6610 (1.6510) [2021-04-16 06:59:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][90/1251] eta 0:06:32 lr 0.000432 time 0.2782 (0.3383) loss 2.4549 (3.4482) grad_norm 1.7099 (1.6490) [2021-04-16 06:59:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][100/1251] eta 0:06:22 lr 0.000432 time 0.2733 (0.3320) loss 3.7544 (3.4611) grad_norm 1.5499 (1.6439) [2021-04-16 06:59:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][110/1251] eta 0:06:13 lr 0.000432 time 0.2476 (0.3273) loss 3.1452 (3.4685) grad_norm 2.0634 (1.6398) [2021-04-16 06:59:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][120/1251] eta 0:06:06 lr 0.000432 time 0.2629 (0.3240) loss 4.2158 (3.4856) grad_norm 1.5959 (1.6440) [2021-04-16 06:59:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][130/1251] eta 0:06:00 lr 0.000432 time 0.2918 (0.3215) loss 3.7610 (3.4623) grad_norm 1.9515 (1.6485) [2021-04-16 06:59:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][140/1251] eta 0:05:53 lr 0.000432 time 0.2498 (0.3184) loss 3.7751 (3.4766) grad_norm 1.5841 (1.6509) [2021-04-16 06:59:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][150/1251] eta 0:05:48 lr 0.000432 time 0.3876 (0.3170) loss 2.9380 (3.4670) grad_norm 1.5308 (1.6532) [2021-04-16 06:59:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][160/1251] eta 0:05:43 lr 0.000432 time 0.3027 (0.3149) loss 3.6049 (3.4773) grad_norm 1.9697 (1.6533) [2021-04-16 06:59:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][170/1251] eta 0:05:37 lr 0.000432 time 0.2934 (0.3126) loss 4.4074 (3.4838) grad_norm 1.8068 (1.6639) [2021-04-16 06:59:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][180/1251] eta 0:05:32 lr 0.000432 time 0.2767 (0.3107) loss 2.7334 (3.4731) grad_norm 1.7197 (1.6617) [2021-04-16 06:59:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][190/1251] eta 0:05:27 lr 0.000432 time 0.2856 (0.3090) loss 3.3318 (3.4850) grad_norm 1.6330 (1.6578) [2021-04-16 06:59:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][200/1251] eta 0:05:22 lr 0.000432 time 0.2672 (0.3072) loss 3.7225 (3.4874) grad_norm 1.5223 (1.6549) [2021-04-16 06:59:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][210/1251] eta 0:05:18 lr 0.000432 time 0.2505 (0.3056) loss 3.6743 (3.4785) grad_norm 1.5110 (1.6494) [2021-04-16 06:59:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][220/1251] eta 0:05:13 lr 0.000432 time 0.2708 (0.3044) loss 3.7625 (3.4655) grad_norm 1.4146 (1.6414) [2021-04-16 06:59:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][230/1251] eta 0:05:09 lr 0.000432 time 0.2876 (0.3033) loss 3.0908 (3.4663) grad_norm 1.8805 (1.6423) [2021-04-16 06:59:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][240/1251] eta 0:05:05 lr 0.000432 time 0.2736 (0.3020) loss 3.6838 (3.4796) grad_norm 1.9619 (1.6448) [2021-04-16 07:00:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][250/1251] eta 0:05:01 lr 0.000432 time 0.2880 (0.3016) loss 3.5886 (3.4913) grad_norm 1.8505 (1.6439) [2021-04-16 07:00:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][260/1251] eta 0:04:58 lr 0.000432 time 0.2727 (0.3007) loss 2.6537 (3.4921) grad_norm 1.5342 (1.6411) [2021-04-16 07:00:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][270/1251] eta 0:04:54 lr 0.000432 time 0.2749 (0.2997) loss 3.1358 (3.4924) grad_norm 1.7901 (1.6424) [2021-04-16 07:00:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][280/1251] eta 0:04:50 lr 0.000432 time 0.2602 (0.2989) loss 2.2871 (3.4809) grad_norm 1.8698 (1.6398) [2021-04-16 07:00:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][290/1251] eta 0:04:46 lr 0.000432 time 0.2729 (0.2982) loss 3.9662 (3.4839) grad_norm 1.3642 (1.6389) [2021-04-16 07:00:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][300/1251] eta 0:04:42 lr 0.000431 time 0.2628 (0.2974) loss 3.1517 (3.4887) grad_norm 1.5194 (1.6403) [2021-04-16 07:00:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][310/1251] eta 0:04:39 lr 0.000431 time 0.2750 (0.2971) loss 3.9909 (3.4898) grad_norm 1.6482 (1.6422) [2021-04-16 07:00:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][320/1251] eta 0:04:35 lr 0.000431 time 0.2825 (0.2964) loss 2.6133 (3.4827) grad_norm 1.5103 (1.6428) [2021-04-16 07:00:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][330/1251] eta 0:04:32 lr 0.000431 time 0.2632 (0.2959) loss 3.6902 (3.4816) grad_norm 1.5282 (1.6447) [2021-04-16 07:00:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][340/1251] eta 0:04:29 lr 0.000431 time 0.2651 (0.2958) loss 3.4756 (3.4816) grad_norm 1.7484 (1.6434) [2021-04-16 07:00:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][350/1251] eta 0:04:26 lr 0.000431 time 0.2637 (0.2957) loss 3.7226 (3.4846) grad_norm 1.5590 (1.6409) [2021-04-16 07:00:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][360/1251] eta 0:04:23 lr 0.000431 time 0.2813 (0.2954) loss 3.2341 (3.4855) grad_norm 1.7999 (1.6420) [2021-04-16 07:00:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][370/1251] eta 0:04:19 lr 0.000431 time 0.2842 (0.2950) loss 3.1465 (3.4931) grad_norm 1.6155 (1.6406) [2021-04-16 07:00:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][380/1251] eta 0:04:16 lr 0.000431 time 0.2985 (0.2945) loss 3.0449 (3.4912) grad_norm 1.5142 (1.6385) [2021-04-16 07:00:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][390/1251] eta 0:04:13 lr 0.000431 time 0.2901 (0.2939) loss 3.4188 (3.4898) grad_norm 1.5783 (1.6375) [2021-04-16 07:00:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][400/1251] eta 0:04:09 lr 0.000431 time 0.2642 (0.2935) loss 3.8291 (3.4920) grad_norm 1.5609 (1.6379) [2021-04-16 07:00:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][410/1251] eta 0:04:06 lr 0.000431 time 0.2573 (0.2931) loss 3.5160 (3.4943) grad_norm 2.0492 (1.6410) [2021-04-16 07:00:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][420/1251] eta 0:04:03 lr 0.000431 time 0.2933 (0.2928) loss 2.8743 (3.4873) grad_norm 1.6002 (1.6414) [2021-04-16 07:00:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][430/1251] eta 0:04:00 lr 0.000431 time 0.2571 (0.2924) loss 4.1960 (3.4904) grad_norm 1.6442 (1.6414) [2021-04-16 07:00:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][440/1251] eta 0:03:56 lr 0.000431 time 0.2675 (0.2921) loss 3.3098 (3.4917) grad_norm 1.6469 (1.6428) [2021-04-16 07:00:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][450/1251] eta 0:03:53 lr 0.000431 time 0.2589 (0.2918) loss 2.2450 (3.4891) grad_norm 1.7507 (1.6427) [2021-04-16 07:00:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][460/1251] eta 0:03:50 lr 0.000431 time 0.2842 (0.2915) loss 2.9088 (3.4873) grad_norm 1.8103 (1.6435) [2021-04-16 07:01:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][470/1251] eta 0:03:47 lr 0.000431 time 0.3070 (0.2914) loss 3.3516 (3.4884) grad_norm 2.0616 (1.6426) [2021-04-16 07:01:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][480/1251] eta 0:03:44 lr 0.000431 time 0.3146 (0.2914) loss 3.0304 (3.4901) grad_norm 1.5987 (1.6419) [2021-04-16 07:01:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][490/1251] eta 0:03:41 lr 0.000431 time 0.2687 (0.2914) loss 3.8755 (3.4855) grad_norm 1.6892 (1.6450) [2021-04-16 07:01:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][500/1251] eta 0:03:38 lr 0.000431 time 0.2744 (0.2910) loss 4.4792 (3.4884) grad_norm 1.5259 (1.6453) [2021-04-16 07:01:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][510/1251] eta 0:03:35 lr 0.000431 time 0.2809 (0.2909) loss 4.0449 (3.4883) grad_norm 1.7548 (1.6445) [2021-04-16 07:01:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][520/1251] eta 0:03:32 lr 0.000431 time 0.2806 (0.2910) loss 3.8043 (3.4850) grad_norm 1.8643 (1.6435) [2021-04-16 07:01:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][530/1251] eta 0:03:29 lr 0.000431 time 0.2828 (0.2908) loss 2.6973 (3.4781) grad_norm 1.9546 (1.6429) [2021-04-16 07:01:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][540/1251] eta 0:03:26 lr 0.000430 time 0.2723 (0.2908) loss 3.7024 (3.4811) grad_norm 1.6155 (1.6433) [2021-04-16 07:01:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][550/1251] eta 0:03:23 lr 0.000430 time 0.2674 (0.2905) loss 4.4281 (3.4820) grad_norm 1.7042 (1.6437) [2021-04-16 07:01:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][560/1251] eta 0:03:20 lr 0.000430 time 0.2617 (0.2902) loss 2.6472 (3.4832) grad_norm 1.3753 (1.6417) [2021-04-16 07:01:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][570/1251] eta 0:03:17 lr 0.000430 time 0.2559 (0.2899) loss 3.5405 (3.4819) grad_norm 1.7609 (1.6429) [2021-04-16 07:01:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][580/1251] eta 0:03:14 lr 0.000430 time 0.2485 (0.2900) loss 3.4835 (3.4809) grad_norm 1.5549 (1.6434) [2021-04-16 07:01:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][590/1251] eta 0:03:11 lr 0.000430 time 0.2836 (0.2899) loss 3.5292 (3.4801) grad_norm 1.8150 (1.6437) [2021-04-16 07:01:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][600/1251] eta 0:03:08 lr 0.000430 time 0.2774 (0.2897) loss 2.9483 (3.4798) grad_norm 1.7337 (1.6455) [2021-04-16 07:01:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][610/1251] eta 0:03:05 lr 0.000430 time 0.2453 (0.2894) loss 2.8326 (3.4789) grad_norm 1.8099 (1.6456) [2021-04-16 07:01:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][620/1251] eta 0:03:02 lr 0.000430 time 0.2965 (0.2895) loss 3.9326 (3.4824) grad_norm 1.6583 (1.6470) [2021-04-16 07:01:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][630/1251] eta 0:02:59 lr 0.000430 time 0.2882 (0.2893) loss 2.9233 (3.4826) grad_norm 1.6430 (1.6459) [2021-04-16 07:01:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][640/1251] eta 0:02:56 lr 0.000430 time 0.2975 (0.2891) loss 3.9368 (3.4821) grad_norm 1.5646 (1.6456) [2021-04-16 07:01:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][650/1251] eta 0:02:53 lr 0.000430 time 0.2639 (0.2892) loss 2.1499 (3.4796) grad_norm 1.6267 (1.6452) [2021-04-16 07:01:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][660/1251] eta 0:02:50 lr 0.000430 time 0.2768 (0.2889) loss 3.6121 (3.4776) grad_norm 1.6578 (1.6452) [2021-04-16 07:01:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][670/1251] eta 0:02:47 lr 0.000430 time 0.2910 (0.2887) loss 4.0119 (3.4795) grad_norm 1.7099 (1.6435) [2021-04-16 07:02:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][680/1251] eta 0:02:44 lr 0.000430 time 0.2944 (0.2885) loss 3.6513 (3.4803) grad_norm 2.1335 (1.6438) [2021-04-16 07:02:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][690/1251] eta 0:02:41 lr 0.000430 time 0.2829 (0.2885) loss 2.7764 (3.4787) grad_norm 1.7050 (1.6437) [2021-04-16 07:02:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][700/1251] eta 0:02:39 lr 0.000430 time 0.2462 (0.2886) loss 2.1249 (3.4746) grad_norm 1.6403 (1.6433) [2021-04-16 07:02:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][710/1251] eta 0:02:36 lr 0.000430 time 0.2821 (0.2884) loss 4.0912 (3.4718) grad_norm 1.6941 (1.6442) [2021-04-16 07:02:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][720/1251] eta 0:02:33 lr 0.000430 time 0.2558 (0.2883) loss 3.5564 (3.4728) grad_norm 1.5853 (1.6434) [2021-04-16 07:02:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][730/1251] eta 0:02:30 lr 0.000430 time 0.2643 (0.2883) loss 2.0980 (3.4722) grad_norm 1.7133 (1.6424) [2021-04-16 07:02:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][740/1251] eta 0:02:27 lr 0.000430 time 0.2657 (0.2882) loss 3.0908 (3.4719) grad_norm 1.6876 (1.6423) [2021-04-16 07:02:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][750/1251] eta 0:02:24 lr 0.000430 time 0.2704 (0.2880) loss 4.1843 (3.4696) grad_norm 1.5658 (1.6423) [2021-04-16 07:02:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][760/1251] eta 0:02:21 lr 0.000430 time 0.2906 (0.2879) loss 2.9080 (3.4706) grad_norm 1.7136 (1.6433) [2021-04-16 07:02:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][770/1251] eta 0:02:18 lr 0.000430 time 0.2729 (0.2878) loss 3.8506 (3.4668) grad_norm 1.7126 (1.6428) [2021-04-16 07:02:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][780/1251] eta 0:02:15 lr 0.000429 time 0.2730 (0.2877) loss 2.6397 (3.4660) grad_norm 1.5532 (1.6425) [2021-04-16 07:02:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][790/1251] eta 0:02:12 lr 0.000429 time 0.2857 (0.2876) loss 2.7330 (3.4665) grad_norm 1.6477 (1.6418) [2021-04-16 07:02:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][800/1251] eta 0:02:09 lr 0.000429 time 0.2875 (0.2875) loss 3.6958 (3.4640) grad_norm 1.5407 (1.6420) [2021-04-16 07:02:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][810/1251] eta 0:02:06 lr 0.000429 time 0.2760 (0.2874) loss 3.6408 (3.4623) grad_norm 1.5431 (1.6411) [2021-04-16 07:02:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][820/1251] eta 0:02:03 lr 0.000429 time 0.2576 (0.2872) loss 4.1318 (3.4617) grad_norm 1.3768 (1.6409) [2021-04-16 07:02:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][830/1251] eta 0:02:00 lr 0.000429 time 0.2546 (0.2872) loss 2.9949 (3.4607) grad_norm 1.6351 (1.6403) [2021-04-16 07:02:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][840/1251] eta 0:01:57 lr 0.000429 time 0.2651 (0.2871) loss 4.0352 (3.4625) grad_norm 1.6314 (1.6394) [2021-04-16 07:02:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][850/1251] eta 0:01:55 lr 0.000429 time 0.2733 (0.2870) loss 2.5352 (3.4625) grad_norm 1.6791 (1.6399) [2021-04-16 07:02:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][860/1251] eta 0:01:52 lr 0.000429 time 0.2822 (0.2869) loss 2.7297 (3.4592) grad_norm 1.3627 (1.6399) [2021-04-16 07:02:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][870/1251] eta 0:01:49 lr 0.000429 time 0.2717 (0.2867) loss 3.4440 (3.4577) grad_norm 1.6097 (1.6394) [2021-04-16 07:02:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][880/1251] eta 0:01:46 lr 0.000429 time 0.3004 (0.2866) loss 2.6975 (3.4561) grad_norm 1.9380 (1.6400) [2021-04-16 07:03:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][890/1251] eta 0:01:43 lr 0.000429 time 0.2613 (0.2864) loss 4.3027 (3.4578) grad_norm 1.5579 (1.6401) [2021-04-16 07:03:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][900/1251] eta 0:01:40 lr 0.000429 time 0.2523 (0.2863) loss 3.8975 (3.4598) grad_norm 1.5704 (1.6395) [2021-04-16 07:03:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][910/1251] eta 0:01:37 lr 0.000429 time 0.2699 (0.2862) loss 3.6417 (3.4626) grad_norm 1.5999 (1.6390) [2021-04-16 07:03:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][920/1251] eta 0:01:34 lr 0.000429 time 0.2655 (0.2862) loss 3.5764 (3.4620) grad_norm 1.8650 (1.6397) [2021-04-16 07:03:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][930/1251] eta 0:01:31 lr 0.000429 time 0.2794 (0.2862) loss 3.8874 (3.4608) grad_norm 2.1451 (1.6411) [2021-04-16 07:03:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][940/1251] eta 0:01:29 lr 0.000429 time 0.2664 (0.2862) loss 4.3852 (3.4627) grad_norm 1.8622 (1.6413) [2021-04-16 07:03:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][950/1251] eta 0:01:26 lr 0.000429 time 0.2774 (0.2861) loss 3.2489 (3.4644) grad_norm 1.8270 (1.6415) [2021-04-16 07:03:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][960/1251] eta 0:01:23 lr 0.000429 time 0.2931 (0.2861) loss 3.6788 (3.4655) grad_norm 1.6098 (1.6415) [2021-04-16 07:03:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][970/1251] eta 0:01:20 lr 0.000429 time 0.2906 (0.2860) loss 2.8327 (3.4649) grad_norm 1.6232 (1.6410) [2021-04-16 07:03:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][980/1251] eta 0:01:17 lr 0.000429 time 0.2647 (0.2858) loss 3.2629 (3.4656) grad_norm 1.4110 (1.6409) [2021-04-16 07:03:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][990/1251] eta 0:01:14 lr 0.000429 time 0.2526 (0.2858) loss 2.4422 (3.4660) grad_norm 1.5912 (1.6408) [2021-04-16 07:03:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][1000/1251] eta 0:01:11 lr 0.000429 time 0.2705 (0.2857) loss 3.2626 (3.4660) grad_norm 1.5774 (1.6407) [2021-04-16 07:03:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][1010/1251] eta 0:01:08 lr 0.000429 time 0.2796 (0.2856) loss 3.9124 (3.4654) grad_norm 1.5805 (1.6410) [2021-04-16 07:03:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][1020/1251] eta 0:01:05 lr 0.000429 time 0.2907 (0.2855) loss 3.9281 (3.4676) grad_norm 1.5313 (1.6407) [2021-04-16 07:03:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][1030/1251] eta 0:01:03 lr 0.000428 time 0.2806 (0.2854) loss 3.3319 (3.4691) grad_norm 1.3871 (1.6403) [2021-04-16 07:03:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][1040/1251] eta 0:01:00 lr 0.000428 time 0.2792 (0.2855) loss 4.4875 (3.4683) grad_norm 1.4844 (1.6394) [2021-04-16 07:03:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][1050/1251] eta 0:00:57 lr 0.000428 time 0.2739 (0.2854) loss 3.8471 (3.4666) grad_norm 1.8175 (1.6400) [2021-04-16 07:03:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][1060/1251] eta 0:00:54 lr 0.000428 time 0.2761 (0.2854) loss 3.8133 (3.4658) grad_norm 1.5801 (1.6397) [2021-04-16 07:03:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][1070/1251] eta 0:00:51 lr 0.000428 time 0.2599 (0.2854) loss 2.6359 (3.4648) grad_norm 1.5700 (1.6409) [2021-04-16 07:03:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][1080/1251] eta 0:00:48 lr 0.000428 time 0.2694 (0.2853) loss 3.3075 (3.4651) grad_norm 2.0488 (1.6406) [2021-04-16 07:03:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][1090/1251] eta 0:00:45 lr 0.000428 time 0.2771 (0.2853) loss 3.4829 (3.4654) grad_norm 1.8434 (1.6409) [2021-04-16 07:03:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][1100/1251] eta 0:00:43 lr 0.000428 time 0.2570 (0.2852) loss 3.4906 (3.4669) grad_norm 1.7943 (1.6417) [2021-04-16 07:04:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][1110/1251] eta 0:00:40 lr 0.000428 time 0.2790 (0.2851) loss 3.8607 (3.4659) grad_norm 1.5026 (1.6419) [2021-04-16 07:04:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][1120/1251] eta 0:00:37 lr 0.000428 time 0.2960 (0.2852) loss 4.0452 (3.4639) grad_norm 1.5260 (1.6419) [2021-04-16 07:04:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][1130/1251] eta 0:00:34 lr 0.000428 time 0.2765 (0.2852) loss 3.8675 (3.4655) grad_norm 1.5136 (1.6416) [2021-04-16 07:04:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][1140/1251] eta 0:00:31 lr 0.000428 time 0.3007 (0.2852) loss 3.7049 (3.4648) grad_norm 1.5387 (1.6422) [2021-04-16 07:04:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][1150/1251] eta 0:00:28 lr 0.000428 time 0.2531 (0.2853) loss 4.0077 (3.4670) grad_norm 1.6212 (1.6421) [2021-04-16 07:04:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][1160/1251] eta 0:00:25 lr 0.000428 time 0.2813 (0.2852) loss 3.6510 (3.4674) grad_norm 1.6213 (1.6412) [2021-04-16 07:04:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][1170/1251] eta 0:00:23 lr 0.000428 time 0.3006 (0.2852) loss 3.9909 (3.4675) grad_norm 1.5372 (1.6403) [2021-04-16 07:04:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][1180/1251] eta 0:00:20 lr 0.000428 time 0.3022 (0.2852) loss 3.8888 (3.4645) grad_norm 1.5597 (1.6403) [2021-04-16 07:04:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][1190/1251] eta 0:00:17 lr 0.000428 time 0.3928 (0.2852) loss 3.3634 (3.4644) grad_norm 2.0060 (1.6407) [2021-04-16 07:04:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][1200/1251] eta 0:00:14 lr 0.000428 time 0.2692 (0.2851) loss 3.4145 (3.4635) grad_norm 2.0197 (1.6409) [2021-04-16 07:04:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][1210/1251] eta 0:00:11 lr 0.000428 time 0.2743 (0.2851) loss 3.6347 (3.4644) grad_norm 1.5071 (1.6405) [2021-04-16 07:04:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][1220/1251] eta 0:00:08 lr 0.000428 time 0.3111 (0.2850) loss 3.6557 (3.4662) grad_norm 1.4181 (1.6404) [2021-04-16 07:04:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][1230/1251] eta 0:00:05 lr 0.000428 time 0.2827 (0.2851) loss 3.6971 (3.4652) grad_norm 1.5081 (1.6407) [2021-04-16 07:04:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][1240/1251] eta 0:00:03 lr 0.000428 time 0.2531 (0.2849) loss 2.1121 (3.4627) grad_norm 1.5509 (1.6405) [2021-04-16 07:04:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [164/300][1250/1251] eta 0:00:00 lr 0.000428 time 0.2470 (0.2846) loss 3.7706 (3.4626) grad_norm 1.7028 (1.6408) [2021-04-16 07:04:44 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 164 training takes 0:05:59 [2021-04-16 07:04:44 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_164.pth saving...... [2021-04-16 07:04:57 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_164.pth saved !!! [2021-04-16 07:04:58 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.084 (1.084) Loss 0.9144 (0.9144) Acc@1 78.027 (78.027) Acc@5 95.898 (95.898) [2021-04-16 07:04:59 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.099 (0.201) Loss 0.9989 (0.9804) Acc@1 75.879 (76.491) Acc@5 93.066 (93.812) [2021-04-16 07:05:01 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.154 (0.187) Loss 0.9819 (0.9911) Acc@1 76.074 (76.288) Acc@5 93.945 (93.662) [2021-04-16 07:05:04 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.131 (0.225) Loss 0.9641 (0.9973) Acc@1 77.637 (76.380) Acc@5 93.262 (93.552) [2021-04-16 07:05:06 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.263 (0.214) Loss 1.0075 (0.9928) Acc@1 75.293 (76.522) Acc@5 94.141 (93.609) [2021-04-16 07:05:13 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 76.542 Acc@5 93.600 [2021-04-16 07:05:13 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 76.5% [2021-04-16 07:05:13 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 76.69% [2021-04-16 07:05:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][0/1251] eta 3:12:45 lr 0.000428 time 9.2452 (9.2452) loss 4.1932 (4.1932) grad_norm 1.7333 (1.7333) [2021-04-16 07:05:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][10/1251] eta 0:22:25 lr 0.000428 time 0.2674 (1.0843) loss 2.4289 (3.4915) grad_norm 1.6086 (1.6884) [2021-04-16 07:05:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][20/1251] eta 0:14:23 lr 0.000427 time 0.2871 (0.7011) loss 3.8572 (3.5094) grad_norm 2.1282 (1.6928) [2021-04-16 07:05:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][30/1251] eta 0:11:29 lr 0.000427 time 0.2656 (0.5650) loss 2.4006 (3.4780) grad_norm 1.5548 (1.6616) [2021-04-16 07:05:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][40/1251] eta 0:10:01 lr 0.000427 time 0.2723 (0.4968) loss 3.9313 (3.5261) grad_norm 1.4995 (1.6365) [2021-04-16 07:05:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][50/1251] eta 0:09:04 lr 0.000427 time 0.2942 (0.4534) loss 4.0635 (3.5723) grad_norm 1.8213 (1.6468) [2021-04-16 07:05:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][60/1251] eta 0:08:25 lr 0.000427 time 0.2706 (0.4240) loss 3.5811 (3.5603) grad_norm 1.7254 (1.6445) [2021-04-16 07:05:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][70/1251] eta 0:07:56 lr 0.000427 time 0.2677 (0.4034) loss 2.8672 (3.5346) grad_norm 1.3903 (1.6452) [2021-04-16 07:05:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][80/1251] eta 0:07:34 lr 0.000427 time 0.2809 (0.3879) loss 4.0815 (3.5196) grad_norm 1.5092 (1.6345) [2021-04-16 07:05:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][90/1251] eta 0:07:16 lr 0.000427 time 0.2691 (0.3756) loss 3.2529 (3.5333) grad_norm 1.6016 (1.6267) [2021-04-16 07:05:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][100/1251] eta 0:07:01 lr 0.000427 time 0.2589 (0.3659) loss 2.1481 (3.5048) grad_norm 1.6081 (1.6285) [2021-04-16 07:05:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][110/1251] eta 0:06:48 lr 0.000427 time 0.2627 (0.3578) loss 2.8766 (3.5036) grad_norm 1.3729 (1.6258) [2021-04-16 07:05:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][120/1251] eta 0:06:38 lr 0.000427 time 0.2575 (0.3520) loss 3.1216 (3.4925) grad_norm 1.6437 (1.6291) [2021-04-16 07:05:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][130/1251] eta 0:06:30 lr 0.000427 time 0.2962 (0.3480) loss 2.9450 (3.4638) grad_norm 1.6999 (1.6290) [2021-04-16 07:06:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][140/1251] eta 0:06:21 lr 0.000427 time 0.2821 (0.3431) loss 4.0035 (3.4675) grad_norm 1.7715 (1.6414) [2021-04-16 07:06:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][150/1251] eta 0:06:13 lr 0.000427 time 0.2746 (0.3395) loss 3.8283 (3.4749) grad_norm 1.5250 (1.6427) [2021-04-16 07:06:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][160/1251] eta 0:06:06 lr 0.000427 time 0.2848 (0.3357) loss 2.5772 (3.4685) grad_norm 1.7035 (1.6417) [2021-04-16 07:06:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][170/1251] eta 0:05:59 lr 0.000427 time 0.2909 (0.3322) loss 4.0087 (3.4745) grad_norm 1.4680 (1.6362) [2021-04-16 07:06:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][180/1251] eta 0:05:52 lr 0.000427 time 0.2830 (0.3288) loss 2.4135 (3.4834) grad_norm 1.5062 (1.6341) [2021-04-16 07:06:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][190/1251] eta 0:05:45 lr 0.000427 time 0.2750 (0.3258) loss 2.3062 (3.4826) grad_norm 1.7551 (1.6346) [2021-04-16 07:06:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][200/1251] eta 0:05:39 lr 0.000427 time 0.2665 (0.3235) loss 2.9291 (3.4811) grad_norm 1.5562 (1.6327) [2021-04-16 07:06:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][210/1251] eta 0:05:34 lr 0.000427 time 0.2899 (0.3214) loss 3.3692 (3.4890) grad_norm 1.4654 (1.6276) [2021-04-16 07:06:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][220/1251] eta 0:05:29 lr 0.000427 time 0.2641 (0.3197) loss 3.9734 (3.4942) grad_norm 1.8100 (1.6255) [2021-04-16 07:06:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][230/1251] eta 0:05:24 lr 0.000427 time 0.2892 (0.3180) loss 3.9631 (3.4970) grad_norm 2.0071 (1.6319) [2021-04-16 07:06:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][240/1251] eta 0:05:19 lr 0.000427 time 0.2892 (0.3163) loss 3.7274 (3.4818) grad_norm 1.6543 (1.6351) [2021-04-16 07:06:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][250/1251] eta 0:05:15 lr 0.000427 time 0.2933 (0.3150) loss 3.0401 (3.4797) grad_norm 1.6142 (1.6358) [2021-04-16 07:06:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][260/1251] eta 0:05:10 lr 0.000427 time 0.2955 (0.3135) loss 3.9956 (3.4840) grad_norm 1.7412 (1.6419) [2021-04-16 07:06:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][270/1251] eta 0:05:06 lr 0.000426 time 0.2653 (0.3121) loss 2.6024 (3.4900) grad_norm 1.6625 (1.6415) [2021-04-16 07:06:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][280/1251] eta 0:05:01 lr 0.000426 time 0.2724 (0.3109) loss 3.9468 (3.4891) grad_norm 1.6569 (1.6417) [2021-04-16 07:06:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][290/1251] eta 0:04:57 lr 0.000426 time 0.2711 (0.3100) loss 3.5356 (3.4876) grad_norm 1.6901 (1.6447) [2021-04-16 07:06:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][300/1251] eta 0:04:54 lr 0.000426 time 0.2923 (0.3094) loss 3.8553 (3.4773) grad_norm 1.9887 (1.6494) [2021-04-16 07:06:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][310/1251] eta 0:04:50 lr 0.000426 time 0.2703 (0.3084) loss 3.6220 (3.4738) grad_norm 1.5371 (1.6503) [2021-04-16 07:06:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][320/1251] eta 0:04:46 lr 0.000426 time 0.2874 (0.3079) loss 2.3351 (3.4737) grad_norm 1.5578 (1.6485) [2021-04-16 07:06:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][330/1251] eta 0:04:42 lr 0.000426 time 0.2696 (0.3068) loss 3.7563 (3.4784) grad_norm 1.6562 (1.6476) [2021-04-16 07:06:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][340/1251] eta 0:04:38 lr 0.000426 time 0.2802 (0.3059) loss 3.3211 (3.4736) grad_norm 1.3280 (1.6465) [2021-04-16 07:07:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][350/1251] eta 0:04:34 lr 0.000426 time 0.2581 (0.3049) loss 4.0224 (3.4793) grad_norm 1.4629 (1.6472) [2021-04-16 07:07:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][360/1251] eta 0:04:31 lr 0.000426 time 0.2804 (0.3046) loss 2.4667 (3.4709) grad_norm 1.5470 (1.6459) [2021-04-16 07:07:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][370/1251] eta 0:04:27 lr 0.000426 time 0.2688 (0.3037) loss 2.8109 (3.4642) grad_norm 1.5296 (1.6453) [2021-04-16 07:07:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][380/1251] eta 0:04:23 lr 0.000426 time 0.2692 (0.3029) loss 3.2791 (3.4543) grad_norm 1.5612 (1.6459) [2021-04-16 07:07:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][390/1251] eta 0:04:20 lr 0.000426 time 0.2580 (0.3026) loss 3.7984 (3.4548) grad_norm 1.6413 (1.6446) [2021-04-16 07:07:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][400/1251] eta 0:04:17 lr 0.000426 time 0.2807 (0.3020) loss 2.9114 (3.4543) grad_norm 1.7104 (1.6452) [2021-04-16 07:07:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][410/1251] eta 0:04:13 lr 0.000426 time 0.2566 (0.3013) loss 3.4499 (3.4542) grad_norm 1.6202 (1.6432) [2021-04-16 07:07:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][420/1251] eta 0:04:10 lr 0.000426 time 0.2703 (0.3009) loss 2.9170 (3.4472) grad_norm 1.7848 (1.6445) [2021-04-16 07:07:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][430/1251] eta 0:04:06 lr 0.000426 time 0.2851 (0.3007) loss 3.8539 (3.4506) grad_norm 1.6034 (1.6454) [2021-04-16 07:07:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][440/1251] eta 0:04:03 lr 0.000426 time 0.2809 (0.3002) loss 3.0029 (3.4501) grad_norm 1.5623 (1.6440) [2021-04-16 07:07:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][450/1251] eta 0:04:00 lr 0.000426 time 0.2803 (0.2998) loss 3.3653 (3.4530) grad_norm 1.7365 (1.6437) [2021-04-16 07:07:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][460/1251] eta 0:03:56 lr 0.000426 time 0.2678 (0.2994) loss 2.5907 (3.4447) grad_norm 1.5404 (1.6455) [2021-04-16 07:07:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][470/1251] eta 0:03:53 lr 0.000426 time 0.2904 (0.2989) loss 3.4089 (3.4419) grad_norm 1.7669 (1.6449) [2021-04-16 07:07:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][480/1251] eta 0:03:50 lr 0.000426 time 0.2854 (0.2986) loss 3.6037 (3.4398) grad_norm 1.5095 (1.6450) [2021-04-16 07:07:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][490/1251] eta 0:03:46 lr 0.000426 time 0.3029 (0.2982) loss 3.3595 (3.4394) grad_norm 1.6062 (1.6451) [2021-04-16 07:07:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][500/1251] eta 0:03:43 lr 0.000426 time 0.2780 (0.2978) loss 3.9142 (3.4371) grad_norm 1.4211 (1.6432) [2021-04-16 07:07:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][510/1251] eta 0:03:40 lr 0.000425 time 0.2763 (0.2974) loss 3.1429 (3.4383) grad_norm 1.7218 (1.6427) [2021-04-16 07:07:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][520/1251] eta 0:03:37 lr 0.000425 time 0.3037 (0.2973) loss 3.4956 (3.4377) grad_norm 1.6368 (1.6411) [2021-04-16 07:07:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][530/1251] eta 0:03:34 lr 0.000425 time 0.2907 (0.2970) loss 4.0780 (3.4390) grad_norm 1.6350 (1.6426) [2021-04-16 07:07:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][540/1251] eta 0:03:30 lr 0.000425 time 0.2695 (0.2965) loss 3.8467 (3.4426) grad_norm 1.6099 (1.6428) [2021-04-16 07:07:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][550/1251] eta 0:03:27 lr 0.000425 time 0.3042 (0.2962) loss 3.7058 (3.4432) grad_norm 1.5324 (1.6416) [2021-04-16 07:07:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][560/1251] eta 0:03:24 lr 0.000425 time 0.2672 (0.2958) loss 3.7457 (3.4450) grad_norm 1.6724 (1.6406) [2021-04-16 07:08:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][570/1251] eta 0:03:21 lr 0.000425 time 0.2826 (0.2957) loss 3.9080 (3.4495) grad_norm 1.6773 (1.6418) [2021-04-16 07:08:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][580/1251] eta 0:03:18 lr 0.000425 time 0.2858 (0.2959) loss 2.9488 (3.4506) grad_norm 1.5221 (1.6413) [2021-04-16 07:08:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][590/1251] eta 0:03:15 lr 0.000425 time 0.2688 (0.2956) loss 3.3215 (3.4441) grad_norm 1.5043 (1.6397) [2021-04-16 07:08:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][600/1251] eta 0:03:12 lr 0.000425 time 0.2757 (0.2952) loss 3.0487 (3.4449) grad_norm 1.7278 (1.6398) [2021-04-16 07:08:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][610/1251] eta 0:03:09 lr 0.000425 time 0.2488 (0.2949) loss 3.6458 (3.4451) grad_norm 1.5555 (1.6393) [2021-04-16 07:08:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][620/1251] eta 0:03:06 lr 0.000425 time 0.2974 (0.2949) loss 3.4584 (3.4434) grad_norm 1.4727 (1.6390) [2021-04-16 07:08:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][630/1251] eta 0:03:02 lr 0.000425 time 0.2652 (0.2945) loss 2.7379 (3.4427) grad_norm 1.4646 (1.6395) [2021-04-16 07:08:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][640/1251] eta 0:02:59 lr 0.000425 time 0.2581 (0.2942) loss 3.0155 (3.4425) grad_norm 1.6538 (1.6400) [2021-04-16 07:08:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][650/1251] eta 0:02:56 lr 0.000425 time 0.2612 (0.2939) loss 4.1227 (3.4480) grad_norm 1.6355 (1.6396) [2021-04-16 07:08:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][660/1251] eta 0:02:53 lr 0.000425 time 0.2904 (0.2937) loss 4.1491 (3.4489) grad_norm 1.3728 (1.6392) [2021-04-16 07:08:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][670/1251] eta 0:02:50 lr 0.000425 time 0.2888 (0.2934) loss 2.6846 (3.4496) grad_norm 1.4879 (1.6385) [2021-04-16 07:08:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][680/1251] eta 0:02:47 lr 0.000425 time 0.2933 (0.2931) loss 2.6697 (3.4513) grad_norm 1.6819 (1.6378) [2021-04-16 07:08:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][690/1251] eta 0:02:44 lr 0.000425 time 0.2725 (0.2932) loss 3.5898 (3.4482) grad_norm 1.7395 (1.6372) [2021-04-16 07:08:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][700/1251] eta 0:02:41 lr 0.000425 time 0.4127 (0.2932) loss 2.9515 (3.4499) grad_norm 1.6633 (1.6374) [2021-04-16 07:08:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][710/1251] eta 0:02:38 lr 0.000425 time 0.3035 (0.2929) loss 3.5964 (3.4552) grad_norm 1.9063 (1.6374) [2021-04-16 07:08:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][720/1251] eta 0:02:35 lr 0.000425 time 0.2691 (0.2929) loss 3.3175 (3.4534) grad_norm 1.6987 (1.6373) [2021-04-16 07:08:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][730/1251] eta 0:02:32 lr 0.000425 time 0.2868 (0.2926) loss 3.5840 (3.4515) grad_norm 1.6181 (1.6373) [2021-04-16 07:08:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][740/1251] eta 0:02:29 lr 0.000425 time 0.2474 (0.2924) loss 4.0599 (3.4530) grad_norm 2.0769 (1.6379) [2021-04-16 07:08:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][750/1251] eta 0:02:26 lr 0.000424 time 0.2682 (0.2921) loss 3.5462 (3.4543) grad_norm 1.5270 (1.6385) [2021-04-16 07:08:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][760/1251] eta 0:02:23 lr 0.000424 time 0.2811 (0.2919) loss 2.7937 (3.4546) grad_norm 1.5335 (1.6379) [2021-04-16 07:08:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][770/1251] eta 0:02:20 lr 0.000424 time 0.3139 (0.2918) loss 2.2253 (3.4529) grad_norm 1.7206 (1.6373) [2021-04-16 07:09:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][780/1251] eta 0:02:17 lr 0.000424 time 0.2813 (0.2916) loss 4.2207 (3.4523) grad_norm 1.4732 (1.6365) [2021-04-16 07:09:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][790/1251] eta 0:02:14 lr 0.000424 time 0.2979 (0.2916) loss 4.1096 (3.4519) grad_norm 1.6550 (1.6357) [2021-04-16 07:09:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][800/1251] eta 0:02:11 lr 0.000424 time 0.2794 (0.2913) loss 2.8233 (3.4519) grad_norm 1.8090 (1.6368) [2021-04-16 07:09:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][810/1251] eta 0:02:08 lr 0.000424 time 0.2649 (0.2911) loss 3.9625 (3.4554) grad_norm 1.6509 (1.6358) [2021-04-16 07:09:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][820/1251] eta 0:02:05 lr 0.000424 time 0.2936 (0.2910) loss 3.7656 (3.4580) grad_norm 1.6471 (1.6357) [2021-04-16 07:09:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][830/1251] eta 0:02:02 lr 0.000424 time 0.3065 (0.2909) loss 3.0672 (3.4522) grad_norm 1.5217 (1.6360) [2021-04-16 07:09:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][840/1251] eta 0:01:59 lr 0.000424 time 0.2790 (0.2908) loss 3.9352 (3.4516) grad_norm 1.4955 (1.6379) [2021-04-16 07:09:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][850/1251] eta 0:01:56 lr 0.000424 time 0.2827 (0.2906) loss 3.4519 (3.4526) grad_norm 1.6900 (1.6384) [2021-04-16 07:09:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][860/1251] eta 0:01:53 lr 0.000424 time 0.3118 (0.2905) loss 3.9479 (3.4543) grad_norm 1.8224 (1.6398) [2021-04-16 07:09:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][870/1251] eta 0:01:50 lr 0.000424 time 0.2710 (0.2904) loss 4.2591 (3.4540) grad_norm 1.9899 (1.6414) [2021-04-16 07:09:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][880/1251] eta 0:01:47 lr 0.000424 time 0.2593 (0.2904) loss 3.8008 (3.4530) grad_norm 1.3920 (1.6406) [2021-04-16 07:09:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][890/1251] eta 0:01:44 lr 0.000424 time 0.2852 (0.2903) loss 3.6038 (3.4541) grad_norm 1.4173 (1.6398) [2021-04-16 07:09:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][900/1251] eta 0:01:41 lr 0.000424 time 0.2765 (0.2901) loss 3.5577 (3.4568) grad_norm 1.4730 (1.6383) [2021-04-16 07:09:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][910/1251] eta 0:01:38 lr 0.000424 time 0.2813 (0.2900) loss 2.6821 (3.4540) grad_norm 1.8498 (1.6386) [2021-04-16 07:09:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][920/1251] eta 0:01:35 lr 0.000424 time 0.2751 (0.2899) loss 2.9803 (3.4521) grad_norm 1.5789 (1.6383) [2021-04-16 07:09:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][930/1251] eta 0:01:33 lr 0.000424 time 0.2888 (0.2898) loss 3.7272 (3.4521) grad_norm 1.9015 (1.6388) [2021-04-16 07:09:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][940/1251] eta 0:01:30 lr 0.000424 time 0.3009 (0.2898) loss 3.9373 (3.4545) grad_norm 1.5427 (1.6378) [2021-04-16 07:09:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][950/1251] eta 0:01:27 lr 0.000424 time 0.2843 (0.2897) loss 2.7009 (3.4542) grad_norm 1.6086 (1.6370) [2021-04-16 07:09:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][960/1251] eta 0:01:24 lr 0.000424 time 0.2810 (0.2897) loss 3.4581 (3.4551) grad_norm 1.8135 (1.6378) [2021-04-16 07:09:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][970/1251] eta 0:01:21 lr 0.000424 time 0.2740 (0.2897) loss 3.4673 (3.4537) grad_norm 1.7357 (1.6373) [2021-04-16 07:09:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][980/1251] eta 0:01:18 lr 0.000424 time 0.2644 (0.2896) loss 4.6142 (3.4556) grad_norm 1.6065 (1.6373) [2021-04-16 07:10:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][990/1251] eta 0:01:15 lr 0.000424 time 0.2721 (0.2894) loss 3.0959 (3.4542) grad_norm 1.9899 (1.6373) [2021-04-16 07:10:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][1000/1251] eta 0:01:12 lr 0.000423 time 0.2666 (0.2893) loss 3.2055 (3.4545) grad_norm 1.5944 (1.6366) [2021-04-16 07:10:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][1010/1251] eta 0:01:09 lr 0.000423 time 0.2823 (0.2893) loss 3.9462 (3.4548) grad_norm 1.6193 (1.6363) [2021-04-16 07:10:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][1020/1251] eta 0:01:06 lr 0.000423 time 0.2782 (0.2892) loss 3.4844 (3.4560) grad_norm 1.6022 (1.6359) [2021-04-16 07:10:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][1030/1251] eta 0:01:03 lr 0.000423 time 0.2802 (0.2892) loss 3.3918 (3.4589) grad_norm 1.5966 (1.6355) [2021-04-16 07:10:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][1040/1251] eta 0:01:01 lr 0.000423 time 0.2902 (0.2891) loss 3.1674 (3.4570) grad_norm 1.5385 (1.6347) [2021-04-16 07:10:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][1050/1251] eta 0:00:58 lr 0.000423 time 0.2797 (0.2890) loss 2.0937 (3.4570) grad_norm 1.6928 (1.6348) [2021-04-16 07:10:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][1060/1251] eta 0:00:55 lr 0.000423 time 0.2747 (0.2890) loss 4.3130 (3.4609) grad_norm 1.4793 (1.6350) [2021-04-16 07:10:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][1070/1251] eta 0:00:52 lr 0.000423 time 0.2788 (0.2889) loss 2.5224 (3.4604) grad_norm 1.8536 (1.6349) [2021-04-16 07:10:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][1080/1251] eta 0:00:49 lr 0.000423 time 0.2543 (0.2887) loss 2.6529 (3.4599) grad_norm 1.6202 (1.6348) [2021-04-16 07:10:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][1090/1251] eta 0:00:46 lr 0.000423 time 0.2659 (0.2888) loss 4.0217 (3.4611) grad_norm 1.7114 (1.6341) [2021-04-16 07:10:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][1100/1251] eta 0:00:43 lr 0.000423 time 0.2595 (0.2886) loss 3.6074 (3.4623) grad_norm 1.6759 (1.6336) [2021-04-16 07:10:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][1110/1251] eta 0:00:40 lr 0.000423 time 0.2663 (0.2885) loss 2.5334 (3.4576) grad_norm 1.5872 (1.6332) [2021-04-16 07:10:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][1120/1251] eta 0:00:37 lr 0.000423 time 0.3096 (0.2884) loss 2.5787 (3.4572) grad_norm 1.4670 (1.6336) [2021-04-16 07:10:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][1130/1251] eta 0:00:34 lr 0.000423 time 0.2618 (0.2883) loss 3.1648 (3.4549) grad_norm 1.9770 (1.6349) [2021-04-16 07:10:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][1140/1251] eta 0:00:32 lr 0.000423 time 0.2539 (0.2883) loss 3.2236 (3.4524) grad_norm 1.9329 (1.6345) [2021-04-16 07:10:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][1150/1251] eta 0:00:29 lr 0.000423 time 0.2777 (0.2882) loss 2.8955 (3.4511) grad_norm 1.6834 (1.6341) [2021-04-16 07:10:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][1160/1251] eta 0:00:26 lr 0.000423 time 0.2929 (0.2883) loss 3.9782 (3.4502) grad_norm 1.4285 (1.6334) [2021-04-16 07:10:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][1170/1251] eta 0:00:23 lr 0.000423 time 0.2740 (0.2883) loss 3.2456 (3.4506) grad_norm 1.4578 (1.6336) [2021-04-16 07:10:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][1180/1251] eta 0:00:20 lr 0.000423 time 0.2585 (0.2882) loss 2.7402 (3.4532) grad_norm 1.4743 (1.6327) [2021-04-16 07:10:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][1190/1251] eta 0:00:17 lr 0.000423 time 0.3864 (0.2882) loss 3.5052 (3.4538) grad_norm 1.6450 (1.6322) [2021-04-16 07:10:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][1200/1251] eta 0:00:14 lr 0.000423 time 0.2832 (0.2881) loss 3.2916 (3.4537) grad_norm 1.6104 (1.6329) [2021-04-16 07:11:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][1210/1251] eta 0:00:11 lr 0.000423 time 0.2720 (0.2879) loss 3.1401 (3.4540) grad_norm 1.4609 (1.6325) [2021-04-16 07:11:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][1220/1251] eta 0:00:08 lr 0.000423 time 0.2838 (0.2879) loss 3.4996 (3.4571) grad_norm 2.2821 (inf) [2021-04-16 07:11:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][1230/1251] eta 0:00:06 lr 0.000423 time 0.2697 (0.2878) loss 3.2773 (3.4565) grad_norm 1.6655 (inf) [2021-04-16 07:11:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][1240/1251] eta 0:00:03 lr 0.000422 time 0.2488 (0.2876) loss 3.9721 (3.4587) grad_norm 1.5624 (inf) [2021-04-16 07:11:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [165/300][1250/1251] eta 0:00:00 lr 0.000422 time 0.2489 (0.2873) loss 3.9802 (3.4560) grad_norm 1.5120 (inf) [2021-04-16 07:11:17 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 165 training takes 0:06:03 [2021-04-16 07:11:17 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_165.pth saving...... [2021-04-16 07:11:28 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_165.pth saved !!! [2021-04-16 07:11:29 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.188 (1.188) Loss 1.0193 (1.0193) Acc@1 75.879 (75.879) Acc@5 92.969 (92.969) [2021-04-16 07:11:31 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.272 (0.257) Loss 1.0020 (1.0151) Acc@1 76.660 (75.914) Acc@5 93.457 (93.244) [2021-04-16 07:11:34 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.368 (0.254) Loss 0.9336 (0.9974) Acc@1 77.539 (76.321) Acc@5 94.238 (93.494) [2021-04-16 07:11:36 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.357 (0.246) Loss 1.0436 (0.9908) Acc@1 75.293 (76.499) Acc@5 93.457 (93.555) [2021-04-16 07:11:37 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.224) Loss 1.0267 (0.9920) Acc@1 74.902 (76.562) Acc@5 92.090 (93.464) [2021-04-16 07:11:45 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 76.624 Acc@5 93.536 [2021-04-16 07:11:45 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 76.6% [2021-04-16 07:11:45 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 76.69% [2021-04-16 07:11:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][0/1251] eta 2:16:14 lr 0.000422 time 6.5342 (6.5342) loss 3.8017 (3.8017) grad_norm 1.4847 (1.4847) [2021-04-16 07:11:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][10/1251] eta 0:17:30 lr 0.000422 time 0.3057 (0.8465) loss 3.3893 (3.5776) grad_norm 1.5928 (1.5806) [2021-04-16 07:11:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][20/1251] eta 0:11:45 lr 0.000422 time 0.2592 (0.5731) loss 4.1321 (3.5632) grad_norm 1.6110 (1.6049) [2021-04-16 07:12:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][30/1251] eta 0:09:46 lr 0.000422 time 0.2715 (0.4801) loss 4.2390 (3.5822) grad_norm 1.4986 (1.6057) [2021-04-16 07:12:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][40/1251] eta 0:08:43 lr 0.000422 time 0.2697 (0.4320) loss 3.8047 (3.5778) grad_norm 1.8678 (1.6089) [2021-04-16 07:12:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][50/1251] eta 0:08:02 lr 0.000422 time 0.2798 (0.4015) loss 3.8641 (3.5635) grad_norm 1.7166 (1.6226) [2021-04-16 07:12:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][60/1251] eta 0:07:33 lr 0.000422 time 0.2807 (0.3807) loss 4.4038 (3.5793) grad_norm 1.7358 (1.6359) [2021-04-16 07:12:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][70/1251] eta 0:07:11 lr 0.000422 time 0.2895 (0.3654) loss 3.0878 (3.5399) grad_norm 1.9477 (1.6458) [2021-04-16 07:12:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][80/1251] eta 0:06:55 lr 0.000422 time 0.2576 (0.3547) loss 3.8611 (3.5407) grad_norm 1.8244 (1.6560) [2021-04-16 07:12:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][90/1251] eta 0:06:43 lr 0.000422 time 0.2684 (0.3478) loss 3.0062 (3.5422) grad_norm 1.4821 (1.6612) [2021-04-16 07:12:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][100/1251] eta 0:06:32 lr 0.000422 time 0.2884 (0.3408) loss 3.3299 (3.5383) grad_norm 1.5524 (1.6628) [2021-04-16 07:12:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][110/1251] eta 0:06:22 lr 0.000422 time 0.2606 (0.3348) loss 3.9095 (3.5245) grad_norm 1.4634 (1.6538) [2021-04-16 07:12:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][120/1251] eta 0:06:12 lr 0.000422 time 0.2840 (0.3298) loss 3.9433 (3.5511) grad_norm 1.5690 (1.6472) [2021-04-16 07:12:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][130/1251] eta 0:06:05 lr 0.000422 time 0.2873 (0.3260) loss 2.9195 (3.5653) grad_norm 1.5569 (1.6437) [2021-04-16 07:12:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][140/1251] eta 0:05:58 lr 0.000422 time 0.2554 (0.3230) loss 3.7570 (3.5741) grad_norm 1.7188 (1.6465) [2021-04-16 07:12:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][150/1251] eta 0:05:53 lr 0.000422 time 0.2787 (0.3210) loss 3.5730 (3.5483) grad_norm 1.7241 (1.6437) [2021-04-16 07:12:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][160/1251] eta 0:05:47 lr 0.000422 time 0.2689 (0.3182) loss 3.9811 (3.5511) grad_norm 1.6812 (1.6470) [2021-04-16 07:12:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][170/1251] eta 0:05:41 lr 0.000422 time 0.2689 (0.3158) loss 2.8176 (3.5431) grad_norm 1.9359 (1.6491) [2021-04-16 07:12:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][180/1251] eta 0:05:36 lr 0.000422 time 0.2873 (0.3145) loss 3.7779 (3.5309) grad_norm 1.7832 (1.6499) [2021-04-16 07:12:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][190/1251] eta 0:05:31 lr 0.000422 time 0.2832 (0.3124) loss 3.7626 (3.5184) grad_norm 1.5217 (1.6569) [2021-04-16 07:12:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][200/1251] eta 0:05:26 lr 0.000422 time 0.2623 (0.3105) loss 2.7399 (3.5097) grad_norm 1.7277 (1.6599) [2021-04-16 07:12:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][210/1251] eta 0:05:21 lr 0.000422 time 0.2867 (0.3090) loss 3.2370 (3.5034) grad_norm 1.6124 (1.6572) [2021-04-16 07:12:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][220/1251] eta 0:05:17 lr 0.000422 time 0.2858 (0.3080) loss 4.3499 (3.5059) grad_norm 1.7268 (1.6577) [2021-04-16 07:12:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][230/1251] eta 0:05:12 lr 0.000422 time 0.2805 (0.3064) loss 4.2117 (3.5119) grad_norm 1.5297 (1.6574) [2021-04-16 07:12:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][240/1251] eta 0:05:08 lr 0.000421 time 0.2907 (0.3050) loss 3.6220 (3.5145) grad_norm 1.7679 (1.6586) [2021-04-16 07:13:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][250/1251] eta 0:05:04 lr 0.000421 time 0.2678 (0.3038) loss 4.2326 (3.5225) grad_norm 1.6798 (1.6645) [2021-04-16 07:13:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][260/1251] eta 0:05:00 lr 0.000421 time 0.2784 (0.3029) loss 3.2061 (3.5237) grad_norm 1.5078 (1.6619) [2021-04-16 07:13:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][270/1251] eta 0:04:56 lr 0.000421 time 0.2721 (0.3020) loss 2.3349 (3.5162) grad_norm 1.7868 (1.6595) [2021-04-16 07:13:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][280/1251] eta 0:04:52 lr 0.000421 time 0.2578 (0.3010) loss 2.9028 (3.5051) grad_norm 1.6920 (1.6578) [2021-04-16 07:13:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][290/1251] eta 0:04:48 lr 0.000421 time 0.2819 (0.3002) loss 3.1897 (3.5002) grad_norm 1.5237 (1.6560) [2021-04-16 07:13:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][300/1251] eta 0:04:45 lr 0.000421 time 0.2556 (0.3001) loss 3.8561 (3.4974) grad_norm 1.5882 (1.6535) [2021-04-16 07:13:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][310/1251] eta 0:04:41 lr 0.000421 time 0.2686 (0.2994) loss 2.4635 (3.4908) grad_norm 1.6418 (1.6504) [2021-04-16 07:13:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][320/1251] eta 0:04:38 lr 0.000421 time 0.2642 (0.2988) loss 3.3454 (3.4939) grad_norm 1.4182 (1.6473) [2021-04-16 07:13:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][330/1251] eta 0:04:34 lr 0.000421 time 0.2689 (0.2981) loss 2.9575 (3.4932) grad_norm 2.1938 (1.6482) [2021-04-16 07:13:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][340/1251] eta 0:04:31 lr 0.000421 time 0.2764 (0.2978) loss 3.6662 (3.4965) grad_norm 1.5613 (1.6483) [2021-04-16 07:13:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][350/1251] eta 0:04:27 lr 0.000421 time 0.2707 (0.2971) loss 3.6978 (3.4940) grad_norm 1.6861 (1.6474) [2021-04-16 07:13:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][360/1251] eta 0:04:24 lr 0.000421 time 0.2533 (0.2970) loss 3.8057 (3.4965) grad_norm 1.5231 (1.6500) [2021-04-16 07:13:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][370/1251] eta 0:04:21 lr 0.000421 time 0.2891 (0.2965) loss 3.7927 (3.4948) grad_norm 1.6469 (1.6503) [2021-04-16 07:13:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][380/1251] eta 0:04:17 lr 0.000421 time 0.2817 (0.2959) loss 3.4799 (3.4949) grad_norm 1.8331 (1.6509) [2021-04-16 07:13:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][390/1251] eta 0:04:14 lr 0.000421 time 0.2838 (0.2956) loss 3.6986 (3.4993) grad_norm 1.5470 (1.6495) [2021-04-16 07:13:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][400/1251] eta 0:04:11 lr 0.000421 time 0.3052 (0.2952) loss 2.5495 (3.5047) grad_norm 1.3682 (1.6462) [2021-04-16 07:13:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][410/1251] eta 0:04:07 lr 0.000421 time 0.2960 (0.2948) loss 3.3278 (3.5080) grad_norm 1.5515 (1.6479) [2021-04-16 07:13:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][420/1251] eta 0:04:04 lr 0.000421 time 0.2680 (0.2943) loss 3.4015 (3.5061) grad_norm 1.8051 (1.6473) [2021-04-16 07:13:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][430/1251] eta 0:04:01 lr 0.000421 time 0.3010 (0.2939) loss 3.0581 (3.5025) grad_norm 1.7432 (1.6481) [2021-04-16 07:13:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][440/1251] eta 0:03:58 lr 0.000421 time 0.3102 (0.2936) loss 3.8296 (3.5062) grad_norm 1.7920 (1.6502) [2021-04-16 07:13:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][450/1251] eta 0:03:54 lr 0.000421 time 0.2767 (0.2931) loss 3.4825 (3.5027) grad_norm 1.4200 (1.6550) [2021-04-16 07:14:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][460/1251] eta 0:03:51 lr 0.000421 time 0.2746 (0.2927) loss 3.1015 (3.4991) grad_norm 1.5326 (1.6536) [2021-04-16 07:14:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][470/1251] eta 0:03:48 lr 0.000421 time 0.2732 (0.2925) loss 3.0492 (3.4974) grad_norm 1.5005 (1.6527) [2021-04-16 07:14:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][480/1251] eta 0:03:45 lr 0.000420 time 0.2611 (0.2922) loss 3.6087 (3.4951) grad_norm 1.5996 (1.6536) [2021-04-16 07:14:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][490/1251] eta 0:03:42 lr 0.000420 time 0.3015 (0.2921) loss 4.0342 (3.4900) grad_norm 1.6265 (1.6522) [2021-04-16 07:14:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][500/1251] eta 0:03:39 lr 0.000420 time 0.2440 (0.2918) loss 3.5900 (3.4966) grad_norm 1.6378 (1.6504) [2021-04-16 07:14:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][510/1251] eta 0:03:36 lr 0.000420 time 0.2917 (0.2916) loss 2.5519 (3.4972) grad_norm 1.7423 (1.6516) [2021-04-16 07:14:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][520/1251] eta 0:03:33 lr 0.000420 time 0.2473 (0.2916) loss 4.1196 (3.5017) grad_norm 1.4495 (1.6511) [2021-04-16 07:14:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][530/1251] eta 0:03:29 lr 0.000420 time 0.2629 (0.2912) loss 3.8018 (3.5021) grad_norm 1.6096 (1.6512) [2021-04-16 07:14:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][540/1251] eta 0:03:26 lr 0.000420 time 0.2727 (0.2909) loss 4.1319 (3.5064) grad_norm 1.7102 (1.6523) [2021-04-16 07:14:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][550/1251] eta 0:03:23 lr 0.000420 time 0.2937 (0.2907) loss 4.0233 (3.5084) grad_norm 1.5908 (1.6516) [2021-04-16 07:14:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][560/1251] eta 0:03:20 lr 0.000420 time 0.2593 (0.2904) loss 3.2119 (3.5086) grad_norm 1.6944 (1.6514) [2021-04-16 07:14:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][570/1251] eta 0:03:17 lr 0.000420 time 0.4050 (0.2904) loss 2.8510 (3.5075) grad_norm 1.5505 (1.6528) [2021-04-16 07:14:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][580/1251] eta 0:03:14 lr 0.000420 time 0.2988 (0.2904) loss 4.2307 (3.5086) grad_norm 1.7031 (1.6553) [2021-04-16 07:14:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][590/1251] eta 0:03:11 lr 0.000420 time 0.2926 (0.2903) loss 3.6126 (3.5082) grad_norm 1.3632 (1.6553) [2021-04-16 07:14:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][600/1251] eta 0:03:09 lr 0.000420 time 0.4082 (0.2904) loss 2.6640 (3.5059) grad_norm 1.6535 (1.6560) [2021-04-16 07:14:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][610/1251] eta 0:03:05 lr 0.000420 time 0.2777 (0.2902) loss 3.3410 (3.5060) grad_norm 2.0221 (1.6557) [2021-04-16 07:14:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][620/1251] eta 0:03:02 lr 0.000420 time 0.2661 (0.2899) loss 3.4506 (3.5037) grad_norm 1.6456 (1.6570) [2021-04-16 07:14:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][630/1251] eta 0:02:59 lr 0.000420 time 0.2907 (0.2897) loss 3.8090 (3.5023) grad_norm 1.9009 (1.6573) [2021-04-16 07:14:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][640/1251] eta 0:02:56 lr 0.000420 time 0.2810 (0.2895) loss 3.5094 (3.5001) grad_norm 1.6384 (1.6569) [2021-04-16 07:14:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][650/1251] eta 0:02:53 lr 0.000420 time 0.2771 (0.2895) loss 3.7312 (3.5014) grad_norm 1.4178 (1.6563) [2021-04-16 07:14:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][660/1251] eta 0:02:50 lr 0.000420 time 0.2526 (0.2892) loss 3.5838 (3.5014) grad_norm 1.3991 (1.6547) [2021-04-16 07:14:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][670/1251] eta 0:02:48 lr 0.000420 time 0.2497 (0.2892) loss 3.5799 (3.5017) grad_norm 1.5474 (1.6536) [2021-04-16 07:15:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][680/1251] eta 0:02:45 lr 0.000420 time 0.2805 (0.2890) loss 3.7510 (3.4996) grad_norm 1.5194 (1.6515) [2021-04-16 07:15:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][690/1251] eta 0:02:42 lr 0.000420 time 0.2674 (0.2888) loss 2.5308 (3.4982) grad_norm 1.5149 (1.6519) [2021-04-16 07:15:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][700/1251] eta 0:02:39 lr 0.000420 time 0.2855 (0.2887) loss 3.0569 (3.4960) grad_norm 1.7359 (1.6527) [2021-04-16 07:15:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][710/1251] eta 0:02:36 lr 0.000420 time 0.2788 (0.2885) loss 2.6480 (3.4960) grad_norm 1.6523 (1.6524) [2021-04-16 07:15:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][720/1251] eta 0:02:33 lr 0.000420 time 0.2663 (0.2883) loss 3.7701 (3.4949) grad_norm 1.6452 (1.6540) [2021-04-16 07:15:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][730/1251] eta 0:02:30 lr 0.000419 time 0.2788 (0.2884) loss 3.8083 (3.4971) grad_norm 1.5472 (1.6526) [2021-04-16 07:15:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][740/1251] eta 0:02:27 lr 0.000419 time 0.2863 (0.2882) loss 2.7594 (3.4971) grad_norm 1.9795 (1.6536) [2021-04-16 07:15:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][750/1251] eta 0:02:24 lr 0.000419 time 0.2639 (0.2880) loss 3.8021 (3.4974) grad_norm 1.7981 (1.6540) [2021-04-16 07:15:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][760/1251] eta 0:02:21 lr 0.000419 time 0.2822 (0.2880) loss 3.7563 (3.4971) grad_norm 1.5698 (1.6529) [2021-04-16 07:15:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][770/1251] eta 0:02:18 lr 0.000419 time 0.2715 (0.2878) loss 3.2351 (3.5007) grad_norm 1.7610 (1.6515) [2021-04-16 07:15:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][780/1251] eta 0:02:15 lr 0.000419 time 0.3036 (0.2877) loss 3.6833 (3.5010) grad_norm 1.6333 (1.6504) [2021-04-16 07:15:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][790/1251] eta 0:02:12 lr 0.000419 time 0.2688 (0.2877) loss 2.3712 (3.4970) grad_norm 1.4879 (1.6496) [2021-04-16 07:15:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][800/1251] eta 0:02:09 lr 0.000419 time 0.2573 (0.2875) loss 3.5709 (3.4991) grad_norm 1.4153 (1.6494) [2021-04-16 07:15:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][810/1251] eta 0:02:06 lr 0.000419 time 0.3063 (0.2874) loss 3.9978 (3.5009) grad_norm 1.7903 (1.6491) [2021-04-16 07:15:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][820/1251] eta 0:02:03 lr 0.000419 time 0.2845 (0.2873) loss 4.0357 (3.5012) grad_norm 1.7660 (1.6506) [2021-04-16 07:15:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][830/1251] eta 0:02:00 lr 0.000419 time 0.2745 (0.2872) loss 3.5774 (3.5016) grad_norm 1.6813 (1.6499) [2021-04-16 07:15:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][840/1251] eta 0:01:57 lr 0.000419 time 0.2582 (0.2871) loss 2.9206 (3.5011) grad_norm 1.8099 (1.6505) [2021-04-16 07:15:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][850/1251] eta 0:01:55 lr 0.000419 time 0.2816 (0.2870) loss 3.0742 (3.4962) grad_norm 1.6300 (1.6507) [2021-04-16 07:15:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][860/1251] eta 0:01:52 lr 0.000419 time 0.2722 (0.2868) loss 3.8164 (3.4981) grad_norm 2.1668 (1.6528) [2021-04-16 07:15:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][870/1251] eta 0:01:49 lr 0.000419 time 0.2766 (0.2867) loss 2.9554 (3.4984) grad_norm 1.5611 (1.6519) [2021-04-16 07:15:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][880/1251] eta 0:01:46 lr 0.000419 time 0.2858 (0.2867) loss 2.6212 (3.5026) grad_norm 1.5525 (1.6511) [2021-04-16 07:16:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][890/1251] eta 0:01:43 lr 0.000419 time 0.3223 (0.2866) loss 3.1613 (3.5019) grad_norm 1.6333 (1.6507) [2021-04-16 07:16:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][900/1251] eta 0:01:40 lr 0.000419 time 0.2970 (0.2866) loss 4.1096 (3.5001) grad_norm 1.4531 (1.6507) [2021-04-16 07:16:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][910/1251] eta 0:01:37 lr 0.000419 time 0.2661 (0.2864) loss 3.4309 (3.4991) grad_norm 1.7793 (1.6507) [2021-04-16 07:16:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][920/1251] eta 0:01:34 lr 0.000419 time 0.2821 (0.2866) loss 3.7475 (3.5010) grad_norm 1.6669 (1.6502) [2021-04-16 07:16:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][930/1251] eta 0:01:31 lr 0.000419 time 0.2664 (0.2864) loss 2.2527 (3.4988) grad_norm 1.5878 (1.6502) [2021-04-16 07:16:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][940/1251] eta 0:01:29 lr 0.000419 time 0.2735 (0.2865) loss 3.5506 (3.4967) grad_norm 1.4048 (1.6501) [2021-04-16 07:16:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][950/1251] eta 0:01:26 lr 0.000419 time 0.2744 (0.2863) loss 3.1449 (3.4958) grad_norm 1.8543 (1.6507) [2021-04-16 07:16:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][960/1251] eta 0:01:23 lr 0.000419 time 0.2908 (0.2862) loss 3.2100 (3.4948) grad_norm 1.4419 (1.6508) [2021-04-16 07:16:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][970/1251] eta 0:01:20 lr 0.000418 time 0.2456 (0.2863) loss 2.7964 (3.4948) grad_norm 1.5297 (1.6499) [2021-04-16 07:16:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][980/1251] eta 0:01:17 lr 0.000418 time 0.2627 (0.2862) loss 2.3096 (3.4919) grad_norm 1.7668 (1.6496) [2021-04-16 07:16:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][990/1251] eta 0:01:14 lr 0.000418 time 0.2650 (0.2861) loss 3.6234 (3.4927) grad_norm 1.4923 (1.6498) [2021-04-16 07:16:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][1000/1251] eta 0:01:11 lr 0.000418 time 0.2813 (0.2860) loss 3.4994 (3.4944) grad_norm 1.5187 (1.6494) [2021-04-16 07:16:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][1010/1251] eta 0:01:08 lr 0.000418 time 0.2781 (0.2859) loss 3.6502 (3.4984) grad_norm 2.1619 (1.6501) [2021-04-16 07:16:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][1020/1251] eta 0:01:06 lr 0.000418 time 0.2653 (0.2858) loss 3.4704 (3.4985) grad_norm 1.5220 (1.6495) [2021-04-16 07:16:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][1030/1251] eta 0:01:03 lr 0.000418 time 0.2708 (0.2857) loss 3.6071 (3.5000) grad_norm 2.2632 (1.6505) [2021-04-16 07:16:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][1040/1251] eta 0:01:00 lr 0.000418 time 0.2688 (0.2856) loss 3.5656 (3.5021) grad_norm 1.4224 (1.6500) [2021-04-16 07:16:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][1050/1251] eta 0:00:57 lr 0.000418 time 0.2897 (0.2855) loss 3.6843 (3.5011) grad_norm 1.5536 (1.6494) [2021-04-16 07:16:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][1060/1251] eta 0:00:54 lr 0.000418 time 0.2872 (0.2855) loss 2.9158 (3.4985) grad_norm 1.6643 (1.6486) [2021-04-16 07:16:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][1070/1251] eta 0:00:51 lr 0.000418 time 0.2730 (0.2855) loss 3.4418 (3.4992) grad_norm 1.5752 (1.6491) [2021-04-16 07:16:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][1080/1251] eta 0:00:48 lr 0.000418 time 0.2659 (0.2853) loss 2.3839 (3.4994) grad_norm 1.7850 (1.6487) [2021-04-16 07:16:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][1090/1251] eta 0:00:45 lr 0.000418 time 0.2673 (0.2854) loss 3.8625 (3.4999) grad_norm 1.5396 (1.6483) [2021-04-16 07:16:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][1100/1251] eta 0:00:43 lr 0.000418 time 0.2743 (0.2854) loss 3.0836 (3.5005) grad_norm 1.3924 (1.6479) [2021-04-16 07:17:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][1110/1251] eta 0:00:40 lr 0.000418 time 0.2779 (0.2853) loss 3.8042 (3.5022) grad_norm 1.4097 (1.6473) [2021-04-16 07:17:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][1120/1251] eta 0:00:37 lr 0.000418 time 0.2751 (0.2852) loss 4.0395 (3.5049) grad_norm 1.6759 (1.6478) [2021-04-16 07:17:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][1130/1251] eta 0:00:34 lr 0.000418 time 0.2746 (0.2851) loss 4.0085 (3.5023) grad_norm 1.5165 (1.6482) [2021-04-16 07:17:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][1140/1251] eta 0:00:31 lr 0.000418 time 0.3079 (0.2851) loss 3.0771 (3.4999) grad_norm 1.5704 (1.6475) [2021-04-16 07:17:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][1150/1251] eta 0:00:28 lr 0.000418 time 0.3109 (0.2851) loss 3.0449 (3.4976) grad_norm 1.7101 (1.6476) [2021-04-16 07:17:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][1160/1251] eta 0:00:25 lr 0.000418 time 0.2523 (0.2851) loss 3.5999 (3.4989) grad_norm 1.5505 (1.6479) [2021-04-16 07:17:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][1170/1251] eta 0:00:23 lr 0.000418 time 0.2961 (0.2850) loss 2.9384 (3.5004) grad_norm 2.3294 (1.6539) [2021-04-16 07:17:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][1180/1251] eta 0:00:20 lr 0.000418 time 0.2623 (0.2850) loss 3.2725 (3.5028) grad_norm 2.0093 (1.6553) [2021-04-16 07:17:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][1190/1251] eta 0:00:17 lr 0.000418 time 0.2731 (0.2850) loss 3.3356 (3.5031) grad_norm 1.9855 (1.6560) [2021-04-16 07:17:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][1200/1251] eta 0:00:14 lr 0.000418 time 0.2705 (0.2849) loss 2.4071 (3.5019) grad_norm 1.5144 (1.6566) [2021-04-16 07:17:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][1210/1251] eta 0:00:11 lr 0.000418 time 0.2997 (0.2848) loss 3.3561 (3.5015) grad_norm 1.5323 (1.6563) [2021-04-16 07:17:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][1220/1251] eta 0:00:08 lr 0.000417 time 0.2885 (0.2847) loss 2.5505 (3.5006) grad_norm 1.5692 (1.6556) [2021-04-16 07:17:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][1230/1251] eta 0:00:05 lr 0.000417 time 0.2685 (0.2847) loss 3.5448 (3.5009) grad_norm 1.4655 (1.6562) [2021-04-16 07:17:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][1240/1251] eta 0:00:03 lr 0.000417 time 0.2681 (0.2847) loss 4.0326 (3.5035) grad_norm 1.6445 (1.6557) [2021-04-16 07:17:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [166/300][1250/1251] eta 0:00:00 lr 0.000417 time 0.2490 (0.2844) loss 3.5026 (3.5047) grad_norm 1.5006 (1.6552) [2021-04-16 07:17:46 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 166 training takes 0:06:00 [2021-04-16 07:17:46 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_166.pth saving...... [2021-04-16 07:18:00 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_166.pth saved !!! [2021-04-16 07:18:01 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.219 (1.219) Loss 0.9757 (0.9757) Acc@1 76.367 (76.367) Acc@5 93.848 (93.848) [2021-04-16 07:18:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.134 (0.219) Loss 0.9233 (0.9724) Acc@1 79.297 (77.344) Acc@5 94.824 (93.839) [2021-04-16 07:18:05 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.541 (0.238) Loss 1.0423 (0.9821) Acc@1 75.586 (77.000) Acc@5 92.383 (93.694) [2021-04-16 07:18:08 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.158 (0.245) Loss 0.9838 (0.9952) Acc@1 76.953 (76.752) Acc@5 94.141 (93.523) [2021-04-16 07:18:09 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.224) Loss 0.9168 (0.9919) Acc@1 78.418 (76.734) Acc@5 94.727 (93.626) [2021-04-16 07:18:23 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 76.864 Acc@5 93.682 [2021-04-16 07:18:23 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 76.9% [2021-04-16 07:18:23 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 76.86% [2021-04-16 07:18:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][0/1251] eta 0:29:16 lr 0.000417 time 1.4042 (1.4042) loss 4.2480 (4.2480) grad_norm 1.7712 (1.7712) [2021-04-16 07:18:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][10/1251] eta 0:08:07 lr 0.000417 time 0.4561 (0.3925) loss 3.4253 (3.6530) grad_norm 1.5823 (1.6862) [2021-04-16 07:18:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][20/1251] eta 0:06:56 lr 0.000417 time 0.2679 (0.3382) loss 3.0886 (3.5345) grad_norm 1.5024 (1.6704) [2021-04-16 07:18:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][30/1251] eta 0:06:30 lr 0.000417 time 0.2654 (0.3199) loss 2.8061 (3.4434) grad_norm 1.6306 (1.6916) [2021-04-16 07:18:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][40/1251] eta 0:06:14 lr 0.000417 time 0.2779 (0.3095) loss 2.6961 (3.4230) grad_norm 1.4399 (1.6953) [2021-04-16 07:18:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][50/1251] eta 0:06:04 lr 0.000417 time 0.2696 (0.3038) loss 3.4269 (3.4952) grad_norm 1.7411 (1.6775) [2021-04-16 07:18:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][60/1251] eta 0:05:57 lr 0.000417 time 0.2841 (0.2998) loss 2.3018 (3.4731) grad_norm 1.6335 (1.6791) [2021-04-16 07:18:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][70/1251] eta 0:05:50 lr 0.000417 time 0.2818 (0.2968) loss 3.7005 (3.4757) grad_norm 1.9290 (1.6842) [2021-04-16 07:18:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][80/1251] eta 0:05:44 lr 0.000417 time 0.2831 (0.2944) loss 3.7186 (3.4842) grad_norm 1.6001 (1.6804) [2021-04-16 07:18:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][90/1251] eta 0:05:41 lr 0.000417 time 0.2819 (0.2941) loss 2.7374 (3.4591) grad_norm 1.5094 (1.6652) [2021-04-16 07:18:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][100/1251] eta 0:05:36 lr 0.000417 time 0.2617 (0.2928) loss 4.0399 (3.4649) grad_norm 1.5003 (1.6577) [2021-04-16 07:18:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][110/1251] eta 0:05:32 lr 0.000417 time 0.2752 (0.2914) loss 3.9300 (3.4616) grad_norm 1.5946 (1.6486) [2021-04-16 07:18:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][120/1251] eta 0:05:29 lr 0.000417 time 0.2718 (0.2917) loss 4.1803 (3.4645) grad_norm 1.4435 (1.6490) [2021-04-16 07:19:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][130/1251] eta 0:05:26 lr 0.000417 time 0.2678 (0.2912) loss 4.1486 (3.4805) grad_norm 1.8428 (1.6539) [2021-04-16 07:19:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][140/1251] eta 0:05:23 lr 0.000417 time 0.2621 (0.2911) loss 4.2198 (3.4959) grad_norm 1.6308 (1.6542) [2021-04-16 07:19:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][150/1251] eta 0:05:20 lr 0.000417 time 0.2819 (0.2910) loss 3.6068 (3.4888) grad_norm 1.8365 (1.6605) [2021-04-16 07:19:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][160/1251] eta 0:05:15 lr 0.000417 time 0.2668 (0.2896) loss 2.4142 (3.4796) grad_norm 1.7619 (1.6689) [2021-04-16 07:19:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][170/1251] eta 0:05:12 lr 0.000417 time 0.2906 (0.2892) loss 2.9659 (3.4818) grad_norm 1.5879 (1.6721) [2021-04-16 07:19:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][180/1251] eta 0:05:09 lr 0.000417 time 0.2932 (0.2893) loss 3.1523 (3.4807) grad_norm 1.6570 (1.6698) [2021-04-16 07:19:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][190/1251] eta 0:05:06 lr 0.000417 time 0.2760 (0.2884) loss 3.7121 (3.4925) grad_norm 1.4510 (1.6691) [2021-04-16 07:19:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][200/1251] eta 0:05:02 lr 0.000417 time 0.2605 (0.2881) loss 2.2926 (3.4895) grad_norm 1.6099 (1.6675) [2021-04-16 07:19:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][210/1251] eta 0:04:59 lr 0.000416 time 0.2761 (0.2877) loss 3.6285 (3.4860) grad_norm 1.4672 (1.6662) [2021-04-16 07:19:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][220/1251] eta 0:04:56 lr 0.000416 time 0.3013 (0.2875) loss 2.1645 (3.4867) grad_norm 1.6829 (1.6623) [2021-04-16 07:19:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][230/1251] eta 0:04:52 lr 0.000416 time 0.2852 (0.2869) loss 4.0740 (3.4771) grad_norm 1.8289 (1.6677) [2021-04-16 07:19:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][240/1251] eta 0:04:49 lr 0.000416 time 0.2669 (0.2865) loss 2.4184 (3.4854) grad_norm 1.6338 (1.6705) [2021-04-16 07:19:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][250/1251] eta 0:04:46 lr 0.000416 time 0.2511 (0.2861) loss 3.6651 (3.4867) grad_norm 1.4406 (1.6671) [2021-04-16 07:19:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][260/1251] eta 0:04:43 lr 0.000416 time 0.2746 (0.2857) loss 3.1100 (3.4831) grad_norm 1.7765 (1.6675) [2021-04-16 07:19:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][270/1251] eta 0:04:40 lr 0.000416 time 0.2911 (0.2856) loss 4.1332 (3.4893) grad_norm 1.4261 (1.6649) [2021-04-16 07:19:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][280/1251] eta 0:04:37 lr 0.000416 time 0.3100 (0.2853) loss 3.6535 (3.4964) grad_norm 1.5814 (1.6626) [2021-04-16 07:19:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][290/1251] eta 0:04:34 lr 0.000416 time 0.2700 (0.2858) loss 3.5762 (3.4983) grad_norm 1.4930 (1.6613) [2021-04-16 07:19:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][300/1251] eta 0:04:31 lr 0.000416 time 0.2706 (0.2855) loss 2.5761 (3.5008) grad_norm 1.7380 (1.6623) [2021-04-16 07:19:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][310/1251] eta 0:04:28 lr 0.000416 time 0.2532 (0.2851) loss 3.5217 (3.4979) grad_norm 1.4318 (1.6607) [2021-04-16 07:19:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][320/1251] eta 0:04:25 lr 0.000416 time 0.2823 (0.2849) loss 2.2332 (3.4983) grad_norm 1.6467 (1.6601) [2021-04-16 07:19:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][330/1251] eta 0:04:22 lr 0.000416 time 0.2670 (0.2847) loss 3.2847 (3.4959) grad_norm 1.3348 (1.6604) [2021-04-16 07:20:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][340/1251] eta 0:04:19 lr 0.000416 time 0.2828 (0.2845) loss 3.3303 (3.4958) grad_norm 1.7387 (1.6580) [2021-04-16 07:20:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][350/1251] eta 0:04:16 lr 0.000416 time 0.2901 (0.2844) loss 3.0042 (3.4918) grad_norm 1.6420 (1.6576) [2021-04-16 07:20:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][360/1251] eta 0:04:13 lr 0.000416 time 0.3154 (0.2847) loss 4.3142 (3.4873) grad_norm 1.6515 (1.6579) [2021-04-16 07:20:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][370/1251] eta 0:04:10 lr 0.000416 time 0.2596 (0.2849) loss 3.2781 (3.4856) grad_norm 1.7024 (1.6560) [2021-04-16 07:20:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][380/1251] eta 0:04:08 lr 0.000416 time 0.2886 (0.2849) loss 3.6129 (3.4822) grad_norm 1.6716 (1.6535) [2021-04-16 07:20:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][390/1251] eta 0:04:05 lr 0.000416 time 0.2676 (0.2847) loss 3.3218 (3.4793) grad_norm 1.7179 (1.6530) [2021-04-16 07:20:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][400/1251] eta 0:04:02 lr 0.000416 time 0.2514 (0.2844) loss 3.4023 (3.4786) grad_norm 1.7647 (1.6524) [2021-04-16 07:20:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][410/1251] eta 0:03:59 lr 0.000416 time 0.2794 (0.2844) loss 4.2079 (3.4800) grad_norm 1.7482 (1.6506) [2021-04-16 07:20:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][420/1251] eta 0:03:56 lr 0.000416 time 0.2617 (0.2842) loss 3.6545 (3.4782) grad_norm 1.5773 (1.6497) [2021-04-16 07:20:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][430/1251] eta 0:03:53 lr 0.000416 time 0.2545 (0.2844) loss 4.2954 (3.4792) grad_norm 1.4920 (1.6485) [2021-04-16 07:20:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][440/1251] eta 0:03:50 lr 0.000416 time 0.2765 (0.2845) loss 2.8748 (3.4710) grad_norm 1.5273 (1.6478) [2021-04-16 07:20:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][450/1251] eta 0:03:48 lr 0.000416 time 0.4871 (0.2849) loss 3.5639 (3.4677) grad_norm 2.0086 (1.6510) [2021-04-16 07:20:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][460/1251] eta 0:03:45 lr 0.000415 time 0.2539 (0.2848) loss 2.8955 (3.4662) grad_norm 1.7086 (1.6495) [2021-04-16 07:20:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][470/1251] eta 0:03:42 lr 0.000415 time 0.2720 (0.2846) loss 3.2441 (3.4696) grad_norm 1.8182 (1.6513) [2021-04-16 07:20:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][480/1251] eta 0:03:39 lr 0.000415 time 0.2515 (0.2844) loss 3.4430 (3.4701) grad_norm 1.8491 (1.6519) [2021-04-16 07:20:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][490/1251] eta 0:03:36 lr 0.000415 time 0.2752 (0.2842) loss 3.4838 (3.4699) grad_norm 1.9874 (1.6531) [2021-04-16 07:20:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][500/1251] eta 0:03:33 lr 0.000415 time 0.2780 (0.2841) loss 3.9163 (3.4740) grad_norm 1.7752 (1.6546) [2021-04-16 07:20:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][510/1251] eta 0:03:30 lr 0.000415 time 0.2634 (0.2839) loss 3.5650 (3.4693) grad_norm 1.6770 (1.6565) [2021-04-16 07:20:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][520/1251] eta 0:03:27 lr 0.000415 time 0.2513 (0.2837) loss 3.6555 (3.4644) grad_norm 1.5922 (1.6574) [2021-04-16 07:20:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][530/1251] eta 0:03:24 lr 0.000415 time 0.3026 (0.2836) loss 3.9616 (3.4673) grad_norm 1.5638 (1.6568) [2021-04-16 07:20:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][540/1251] eta 0:03:22 lr 0.000415 time 0.2947 (0.2844) loss 3.9770 (3.4722) grad_norm 2.0442 (1.6572) [2021-04-16 07:21:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][550/1251] eta 0:03:19 lr 0.000415 time 0.3035 (0.2842) loss 3.5853 (3.4690) grad_norm 1.7548 (1.6563) [2021-04-16 07:21:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][560/1251] eta 0:03:16 lr 0.000415 time 0.2843 (0.2842) loss 3.4171 (3.4707) grad_norm 1.5987 (1.6561) [2021-04-16 07:21:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][570/1251] eta 0:03:13 lr 0.000415 time 0.2706 (0.2843) loss 3.8269 (3.4689) grad_norm 1.6590 (1.6558) [2021-04-16 07:21:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][580/1251] eta 0:03:10 lr 0.000415 time 0.2723 (0.2845) loss 3.7871 (3.4656) grad_norm 1.8809 (1.6578) [2021-04-16 07:21:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][590/1251] eta 0:03:08 lr 0.000415 time 0.2868 (0.2847) loss 2.6442 (3.4610) grad_norm 1.4530 (1.6582) [2021-04-16 07:21:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][600/1251] eta 0:03:05 lr 0.000415 time 0.2675 (0.2844) loss 3.5983 (3.4585) grad_norm 1.6073 (1.6575) [2021-04-16 07:21:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][610/1251] eta 0:03:02 lr 0.000415 time 0.3083 (0.2844) loss 2.8725 (3.4598) grad_norm 1.5555 (1.6589) [2021-04-16 07:21:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][620/1251] eta 0:02:59 lr 0.000415 time 0.2625 (0.2842) loss 3.3405 (3.4629) grad_norm 1.5777 (1.6598) [2021-04-16 07:21:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][630/1251] eta 0:02:56 lr 0.000415 time 0.2612 (0.2841) loss 3.3610 (3.4602) grad_norm 1.7808 (1.6601) [2021-04-16 07:21:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][640/1251] eta 0:02:53 lr 0.000415 time 0.2809 (0.2840) loss 3.1744 (3.4616) grad_norm 1.6987 (1.6591) [2021-04-16 07:21:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][650/1251] eta 0:02:50 lr 0.000415 time 0.2620 (0.2839) loss 2.8922 (3.4614) grad_norm 1.7113 (1.6591) [2021-04-16 07:21:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][660/1251] eta 0:02:47 lr 0.000415 time 0.2663 (0.2838) loss 3.4490 (3.4604) grad_norm 1.6401 (1.6613) [2021-04-16 07:21:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][670/1251] eta 0:02:44 lr 0.000415 time 0.3035 (0.2840) loss 4.0105 (3.4634) grad_norm 1.7284 (1.6615) [2021-04-16 07:21:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][680/1251] eta 0:02:42 lr 0.000415 time 0.2754 (0.2839) loss 2.7031 (3.4637) grad_norm 1.6403 (1.6613) [2021-04-16 07:21:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][690/1251] eta 0:02:39 lr 0.000415 time 0.2652 (0.2839) loss 3.5534 (3.4627) grad_norm 1.6441 (1.6610) [2021-04-16 07:21:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][700/1251] eta 0:02:36 lr 0.000414 time 0.2814 (0.2838) loss 2.9799 (3.4641) grad_norm 2.1407 (1.6629) [2021-04-16 07:21:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][710/1251] eta 0:02:33 lr 0.000414 time 0.2985 (0.2838) loss 3.9441 (3.4640) grad_norm 1.6086 (1.6622) [2021-04-16 07:21:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][720/1251] eta 0:02:30 lr 0.000414 time 0.2625 (0.2840) loss 4.0517 (3.4649) grad_norm 1.6921 (1.6624) [2021-04-16 07:21:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][730/1251] eta 0:02:27 lr 0.000414 time 0.2776 (0.2838) loss 3.0714 (3.4644) grad_norm 1.6220 (1.6618) [2021-04-16 07:21:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][740/1251] eta 0:02:24 lr 0.000414 time 0.2896 (0.2837) loss 2.4599 (3.4624) grad_norm 1.6748 (1.6633) [2021-04-16 07:21:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][750/1251] eta 0:02:22 lr 0.000414 time 0.2923 (0.2837) loss 3.1656 (3.4588) grad_norm 1.8492 (1.6646) [2021-04-16 07:21:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][760/1251] eta 0:02:19 lr 0.000414 time 0.2668 (0.2837) loss 3.5134 (3.4603) grad_norm 1.8023 (1.6661) [2021-04-16 07:22:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][770/1251] eta 0:02:16 lr 0.000414 time 0.2830 (0.2837) loss 2.6363 (3.4608) grad_norm 1.6692 (1.6665) [2021-04-16 07:22:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][780/1251] eta 0:02:13 lr 0.000414 time 0.2897 (0.2836) loss 2.3595 (3.4621) grad_norm 1.7087 (1.6666) [2021-04-16 07:22:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][790/1251] eta 0:02:10 lr 0.000414 time 0.2899 (0.2835) loss 3.2813 (3.4589) grad_norm 1.4741 (1.6658) [2021-04-16 07:22:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][800/1251] eta 0:02:07 lr 0.000414 time 0.2469 (0.2834) loss 3.3841 (3.4557) grad_norm 1.7015 (1.6649) [2021-04-16 07:22:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][810/1251] eta 0:02:04 lr 0.000414 time 0.2738 (0.2833) loss 2.8035 (3.4558) grad_norm 2.0300 (1.6656) [2021-04-16 07:22:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][820/1251] eta 0:02:02 lr 0.000414 time 0.2766 (0.2833) loss 3.7218 (3.4592) grad_norm 1.8416 (1.6661) [2021-04-16 07:22:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][830/1251] eta 0:01:59 lr 0.000414 time 0.2703 (0.2831) loss 3.6082 (3.4589) grad_norm 1.7320 (1.6657) [2021-04-16 07:22:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][840/1251] eta 0:01:56 lr 0.000414 time 0.2455 (0.2830) loss 3.9273 (3.4633) grad_norm 1.6902 (1.6651) [2021-04-16 07:22:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][850/1251] eta 0:01:53 lr 0.000414 time 0.2782 (0.2829) loss 3.2971 (3.4646) grad_norm 1.6133 (1.6647) [2021-04-16 07:22:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][860/1251] eta 0:01:50 lr 0.000414 time 0.2772 (0.2829) loss 3.3884 (3.4672) grad_norm 1.6753 (1.6643) [2021-04-16 07:22:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][870/1251] eta 0:01:47 lr 0.000414 time 0.2676 (0.2830) loss 3.7310 (3.4678) grad_norm 1.6048 (1.6650) [2021-04-16 07:22:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][880/1251] eta 0:01:44 lr 0.000414 time 0.2922 (0.2829) loss 3.8938 (3.4691) grad_norm 1.5897 (1.6662) [2021-04-16 07:22:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][890/1251] eta 0:01:42 lr 0.000414 time 0.2654 (0.2829) loss 4.0598 (3.4708) grad_norm 1.5180 (1.6667) [2021-04-16 07:22:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][900/1251] eta 0:01:39 lr 0.000414 time 0.2856 (0.2828) loss 3.6885 (3.4719) grad_norm 1.5438 (1.6677) [2021-04-16 07:22:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][910/1251] eta 0:01:36 lr 0.000414 time 0.2618 (0.2827) loss 4.3449 (3.4717) grad_norm 1.7259 (1.6671) [2021-04-16 07:22:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][920/1251] eta 0:01:33 lr 0.000414 time 0.2893 (0.2826) loss 3.6513 (3.4726) grad_norm 1.5620 (1.6666) [2021-04-16 07:22:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][930/1251] eta 0:01:30 lr 0.000414 time 0.2899 (0.2826) loss 2.6617 (3.4746) grad_norm 1.7235 (1.6662) [2021-04-16 07:22:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][940/1251] eta 0:01:27 lr 0.000414 time 0.2664 (0.2826) loss 4.2217 (3.4742) grad_norm 1.5818 (1.6658) [2021-04-16 07:22:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][950/1251] eta 0:01:25 lr 0.000413 time 0.2477 (0.2826) loss 3.2603 (3.4748) grad_norm 1.3808 (1.6651) [2021-04-16 07:22:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][960/1251] eta 0:01:22 lr 0.000413 time 0.2840 (0.2828) loss 3.9384 (3.4758) grad_norm 1.6667 (1.6642) [2021-04-16 07:22:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][970/1251] eta 0:01:19 lr 0.000413 time 0.2794 (0.2828) loss 2.3481 (3.4751) grad_norm 1.5742 (1.6633) [2021-04-16 07:23:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][980/1251] eta 0:01:16 lr 0.000413 time 0.2713 (0.2828) loss 3.4605 (3.4773) grad_norm 1.6496 (1.6634) [2021-04-16 07:23:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][990/1251] eta 0:01:13 lr 0.000413 time 0.2591 (0.2827) loss 2.7884 (3.4774) grad_norm 1.4721 (1.6642) [2021-04-16 07:23:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][1000/1251] eta 0:01:10 lr 0.000413 time 0.2819 (0.2827) loss 3.9014 (3.4772) grad_norm 1.7814 (1.6641) [2021-04-16 07:23:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][1010/1251] eta 0:01:08 lr 0.000413 time 0.2689 (0.2828) loss 3.8860 (3.4770) grad_norm 1.7503 (1.6640) [2021-04-16 07:23:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][1020/1251] eta 0:01:05 lr 0.000413 time 0.2428 (0.2829) loss 2.8284 (3.4717) grad_norm 1.5936 (1.6639) [2021-04-16 07:23:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][1030/1251] eta 0:01:02 lr 0.000413 time 0.2632 (0.2829) loss 2.7516 (3.4692) grad_norm 1.7367 (1.6642) [2021-04-16 07:23:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][1040/1251] eta 0:00:59 lr 0.000413 time 0.2754 (0.2828) loss 3.8253 (3.4698) grad_norm 1.7278 (1.6647) [2021-04-16 07:23:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][1050/1251] eta 0:00:56 lr 0.000413 time 0.2541 (0.2827) loss 3.6309 (3.4669) grad_norm 2.0073 (1.6656) [2021-04-16 07:23:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][1060/1251] eta 0:00:53 lr 0.000413 time 0.2611 (0.2827) loss 3.2648 (3.4679) grad_norm 1.6382 (1.6663) [2021-04-16 07:23:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][1070/1251] eta 0:00:51 lr 0.000413 time 0.2905 (0.2826) loss 3.3786 (3.4694) grad_norm 1.4515 (1.6663) [2021-04-16 07:23:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][1080/1251] eta 0:00:48 lr 0.000413 time 0.2521 (0.2825) loss 3.8740 (3.4730) grad_norm 1.6098 (1.6656) [2021-04-16 07:23:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][1090/1251] eta 0:00:45 lr 0.000413 time 0.2727 (0.2824) loss 3.0552 (3.4730) grad_norm 1.5665 (1.6656) [2021-04-16 07:23:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][1100/1251] eta 0:00:42 lr 0.000413 time 0.2902 (0.2824) loss 4.1473 (3.4736) grad_norm 1.6588 (1.6659) [2021-04-16 07:23:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][1110/1251] eta 0:00:39 lr 0.000413 time 0.2934 (0.2824) loss 3.7040 (3.4743) grad_norm 1.6789 (1.6656) [2021-04-16 07:23:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][1120/1251] eta 0:00:36 lr 0.000413 time 0.3838 (0.2824) loss 3.2412 (3.4752) grad_norm 1.6626 (inf) [2021-04-16 07:23:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][1130/1251] eta 0:00:34 lr 0.000413 time 0.2825 (0.2824) loss 2.8570 (3.4751) grad_norm 1.3715 (nan) [2021-04-16 07:23:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][1140/1251] eta 0:00:31 lr 0.000413 time 0.3039 (0.2823) loss 3.1585 (3.4728) grad_norm 1.5861 (nan) [2021-04-16 07:23:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][1150/1251] eta 0:00:28 lr 0.000413 time 0.2750 (0.2824) loss 3.4303 (3.4734) grad_norm 1.6210 (nan) [2021-04-16 07:23:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][1160/1251] eta 0:00:25 lr 0.000413 time 0.2580 (0.2824) loss 4.0196 (3.4750) grad_norm 1.7779 (nan) [2021-04-16 07:23:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][1170/1251] eta 0:00:22 lr 0.000413 time 0.2770 (0.2826) loss 3.9544 (3.4767) grad_norm 1.7634 (nan) [2021-04-16 07:23:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][1180/1251] eta 0:00:20 lr 0.000413 time 0.2756 (0.2826) loss 3.5349 (3.4780) grad_norm 1.8525 (nan) [2021-04-16 07:24:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][1190/1251] eta 0:00:17 lr 0.000412 time 0.2866 (0.2826) loss 2.9888 (3.4783) grad_norm 1.7393 (nan) [2021-04-16 07:24:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][1200/1251] eta 0:00:14 lr 0.000412 time 0.2589 (0.2825) loss 4.4695 (3.4801) grad_norm 1.6011 (nan) [2021-04-16 07:24:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][1210/1251] eta 0:00:11 lr 0.000412 time 0.2860 (0.2825) loss 3.5060 (3.4816) grad_norm 1.5148 (nan) [2021-04-16 07:24:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][1220/1251] eta 0:00:08 lr 0.000412 time 0.3017 (0.2824) loss 4.1772 (3.4810) grad_norm 1.5973 (nan) [2021-04-16 07:24:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][1230/1251] eta 0:00:05 lr 0.000412 time 0.2785 (0.2824) loss 2.9246 (3.4805) grad_norm 1.8877 (nan) [2021-04-16 07:24:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][1240/1251] eta 0:00:03 lr 0.000412 time 0.2491 (0.2824) loss 3.1200 (3.4802) grad_norm 1.5415 (nan) [2021-04-16 07:24:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [167/300][1250/1251] eta 0:00:00 lr 0.000412 time 0.2483 (0.2821) loss 3.1138 (3.4771) grad_norm 1.9427 (nan) [2021-04-16 07:24:22 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 167 training takes 0:05:58 [2021-04-16 07:24:22 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_167.pth saving...... [2021-04-16 07:24:32 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_167.pth saved !!! [2021-04-16 07:24:33 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.063 (1.063) Loss 0.9708 (0.9708) Acc@1 76.367 (76.367) Acc@5 93.945 (93.945) [2021-04-16 07:24:34 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.083 (0.198) Loss 0.9494 (0.9791) Acc@1 76.855 (76.687) Acc@5 93.945 (93.581) [2021-04-16 07:24:36 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.146 (0.211) Loss 0.9339 (0.9713) Acc@1 77.539 (76.958) Acc@5 94.238 (93.741) [2021-04-16 07:24:39 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.174 (0.242) Loss 0.8934 (0.9682) Acc@1 77.734 (76.903) Acc@5 94.824 (93.734) [2021-04-16 07:24:41 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.221) Loss 0.9794 (0.9747) Acc@1 76.855 (76.817) Acc@5 92.773 (93.631) [2021-04-16 07:24:49 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 76.852 Acc@5 93.662 [2021-04-16 07:24:49 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 76.9% [2021-04-16 07:24:49 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 76.86% [2021-04-16 07:24:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][0/1251] eta 2:26:22 lr 0.000412 time 7.0203 (7.0203) loss 3.4553 (3.4553) grad_norm 1.5898 (1.5898) [2021-04-16 07:24:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][10/1251] eta 0:18:17 lr 0.000412 time 0.2654 (0.8843) loss 3.2882 (3.0038) grad_norm 1.7097 (1.8291) [2021-04-16 07:25:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][20/1251] eta 0:12:15 lr 0.000412 time 0.2734 (0.5973) loss 3.3720 (3.1291) grad_norm 1.3870 (1.7164) [2021-04-16 07:25:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][30/1251] eta 0:10:03 lr 0.000412 time 0.2597 (0.4945) loss 2.2003 (3.1360) grad_norm 1.5421 (1.6894) [2021-04-16 07:25:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][40/1251] eta 0:08:53 lr 0.000412 time 0.2525 (0.4403) loss 2.9864 (3.2503) grad_norm 1.6508 (1.6544) [2021-04-16 07:25:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][50/1251] eta 0:08:11 lr 0.000412 time 0.2776 (0.4089) loss 3.2254 (3.2271) grad_norm 1.6875 (1.6436) [2021-04-16 07:25:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][60/1251] eta 0:07:42 lr 0.000412 time 0.2907 (0.3881) loss 4.2438 (3.2968) grad_norm 1.6081 (1.6552) [2021-04-16 07:25:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][70/1251] eta 0:07:20 lr 0.000412 time 0.2790 (0.3732) loss 3.7196 (3.3413) grad_norm 1.7423 (1.6615) [2021-04-16 07:25:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][80/1251] eta 0:07:04 lr 0.000412 time 0.2811 (0.3626) loss 2.7710 (3.3512) grad_norm 1.5025 (1.6564) [2021-04-16 07:25:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][90/1251] eta 0:06:50 lr 0.000412 time 0.2709 (0.3535) loss 3.4939 (3.3474) grad_norm 1.6242 (1.6519) [2021-04-16 07:25:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][100/1251] eta 0:06:37 lr 0.000412 time 0.2556 (0.3457) loss 3.2006 (3.3490) grad_norm 1.7998 (1.6568) [2021-04-16 07:25:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][110/1251] eta 0:06:27 lr 0.000412 time 0.2712 (0.3400) loss 3.4500 (3.3686) grad_norm 1.7486 (1.6563) [2021-04-16 07:25:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][120/1251] eta 0:06:18 lr 0.000412 time 0.3147 (0.3350) loss 3.5248 (3.3666) grad_norm 1.3890 (1.6482) [2021-04-16 07:25:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][130/1251] eta 0:06:10 lr 0.000412 time 0.2991 (0.3307) loss 2.9997 (3.3594) grad_norm 1.4897 (1.6415) [2021-04-16 07:25:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][140/1251] eta 0:06:04 lr 0.000412 time 0.2653 (0.3277) loss 3.8224 (3.3634) grad_norm 1.5869 (1.6447) [2021-04-16 07:25:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][150/1251] eta 0:05:57 lr 0.000412 time 0.2444 (0.3243) loss 3.8159 (3.3866) grad_norm 1.8522 (1.6525) [2021-04-16 07:25:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][160/1251] eta 0:05:51 lr 0.000412 time 0.2598 (0.3218) loss 4.3505 (3.3807) grad_norm 1.6953 (1.6488) [2021-04-16 07:25:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][170/1251] eta 0:05:46 lr 0.000412 time 0.2881 (0.3204) loss 3.1066 (3.3801) grad_norm 1.7052 (1.6515) [2021-04-16 07:25:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][180/1251] eta 0:05:40 lr 0.000412 time 0.2565 (0.3178) loss 3.9254 (3.4001) grad_norm 1.8230 (1.6540) [2021-04-16 07:25:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][190/1251] eta 0:05:34 lr 0.000411 time 0.2919 (0.3157) loss 2.7755 (3.4046) grad_norm 1.6934 (1.6499) [2021-04-16 07:25:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][200/1251] eta 0:05:30 lr 0.000411 time 0.2661 (0.3141) loss 2.8684 (3.3868) grad_norm 1.8755 (1.6484) [2021-04-16 07:25:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][210/1251] eta 0:05:25 lr 0.000411 time 0.2692 (0.3122) loss 3.7435 (3.3783) grad_norm 1.8661 (1.6522) [2021-04-16 07:25:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][220/1251] eta 0:05:20 lr 0.000411 time 0.2643 (0.3109) loss 4.2151 (3.3886) grad_norm 1.5468 (1.6501) [2021-04-16 07:26:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][230/1251] eta 0:05:16 lr 0.000411 time 0.2875 (0.3104) loss 2.4795 (3.3861) grad_norm 1.5904 (1.6501) [2021-04-16 07:26:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][240/1251] eta 0:05:12 lr 0.000411 time 0.2703 (0.3089) loss 2.9842 (3.3989) grad_norm 1.7635 (1.6473) [2021-04-16 07:26:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][250/1251] eta 0:05:08 lr 0.000411 time 0.2850 (0.3084) loss 3.4558 (3.4074) grad_norm 1.6349 (1.6445) [2021-04-16 07:26:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][260/1251] eta 0:05:04 lr 0.000411 time 0.2902 (0.3072) loss 3.7311 (3.4173) grad_norm 2.1146 (1.6498) [2021-04-16 07:26:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][270/1251] eta 0:05:00 lr 0.000411 time 0.2559 (0.3060) loss 3.6191 (3.4198) grad_norm 1.6644 (1.6491) [2021-04-16 07:26:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][280/1251] eta 0:04:56 lr 0.000411 time 0.2881 (0.3051) loss 3.0040 (3.4219) grad_norm 1.8302 (1.6496) [2021-04-16 07:26:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][290/1251] eta 0:04:52 lr 0.000411 time 0.2645 (0.3039) loss 2.8697 (3.4108) grad_norm 1.8120 (1.6502) [2021-04-16 07:26:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][300/1251] eta 0:04:48 lr 0.000411 time 0.2574 (0.3030) loss 3.7573 (3.4138) grad_norm 1.5220 (1.6498) [2021-04-16 07:26:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][310/1251] eta 0:04:44 lr 0.000411 time 0.2692 (0.3021) loss 3.1094 (3.4179) grad_norm 2.1796 (1.6523) [2021-04-16 07:26:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][320/1251] eta 0:04:40 lr 0.000411 time 0.2975 (0.3013) loss 3.7013 (3.4185) grad_norm 2.0337 (1.6543) [2021-04-16 07:26:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][330/1251] eta 0:04:36 lr 0.000411 time 0.2807 (0.3005) loss 3.5588 (3.4204) grad_norm 1.6910 (1.6562) [2021-04-16 07:26:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][340/1251] eta 0:04:33 lr 0.000411 time 0.2747 (0.2999) loss 4.3723 (3.4214) grad_norm 1.6500 (1.6578) [2021-04-16 07:26:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][350/1251] eta 0:04:29 lr 0.000411 time 0.2850 (0.2995) loss 3.7474 (3.4190) grad_norm 1.6219 (1.6583) [2021-04-16 07:26:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][360/1251] eta 0:04:26 lr 0.000411 time 0.2493 (0.2993) loss 3.8948 (3.4179) grad_norm 1.5945 (1.6617) [2021-04-16 07:26:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][370/1251] eta 0:04:23 lr 0.000411 time 0.3097 (0.2992) loss 3.6904 (3.4182) grad_norm 1.5435 (1.6622) [2021-04-16 07:26:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][380/1251] eta 0:04:20 lr 0.000411 time 0.2959 (0.2986) loss 3.2943 (3.4150) grad_norm 1.8514 (1.6610) [2021-04-16 07:26:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][390/1251] eta 0:04:16 lr 0.000411 time 0.2996 (0.2981) loss 4.5798 (3.4170) grad_norm 1.6617 (1.6589) [2021-04-16 07:26:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][400/1251] eta 0:04:13 lr 0.000411 time 0.2905 (0.2977) loss 2.9554 (3.4204) grad_norm 1.4106 (1.6566) [2021-04-16 07:26:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][410/1251] eta 0:04:09 lr 0.000411 time 0.2597 (0.2971) loss 2.6099 (3.4234) grad_norm 1.4542 (1.6560) [2021-04-16 07:26:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][420/1251] eta 0:04:06 lr 0.000411 time 0.2828 (0.2966) loss 3.9875 (3.4238) grad_norm 1.4623 (1.6558) [2021-04-16 07:26:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][430/1251] eta 0:04:03 lr 0.000410 time 0.2514 (0.2962) loss 4.2404 (3.4216) grad_norm 1.6997 (1.6571) [2021-04-16 07:27:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][440/1251] eta 0:04:00 lr 0.000410 time 0.2695 (0.2961) loss 3.1694 (3.4219) grad_norm 1.6316 (1.6583) [2021-04-16 07:27:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][450/1251] eta 0:03:56 lr 0.000410 time 0.2859 (0.2957) loss 4.1324 (3.4274) grad_norm 1.4461 (1.6589) [2021-04-16 07:27:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][460/1251] eta 0:03:53 lr 0.000410 time 0.2539 (0.2952) loss 3.9115 (3.4288) grad_norm 1.5451 (1.6599) [2021-04-16 07:27:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][470/1251] eta 0:03:50 lr 0.000410 time 0.2609 (0.2948) loss 3.3814 (3.4321) grad_norm 1.7874 (1.6614) [2021-04-16 07:27:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][480/1251] eta 0:03:46 lr 0.000410 time 0.2668 (0.2943) loss 3.9433 (3.4299) grad_norm 1.7791 (1.6607) [2021-04-16 07:27:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][490/1251] eta 0:03:43 lr 0.000410 time 0.2793 (0.2941) loss 3.9320 (3.4304) grad_norm 1.6211 (1.6613) [2021-04-16 07:27:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][500/1251] eta 0:03:40 lr 0.000410 time 0.2889 (0.2940) loss 3.4792 (3.4266) grad_norm 1.9034 (1.6605) [2021-04-16 07:27:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][510/1251] eta 0:03:37 lr 0.000410 time 0.2646 (0.2937) loss 3.9822 (3.4285) grad_norm 1.6952 (1.6608) [2021-04-16 07:27:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][520/1251] eta 0:03:34 lr 0.000410 time 0.2851 (0.2935) loss 3.8525 (3.4291) grad_norm 2.0766 (1.6648) [2021-04-16 07:27:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][530/1251] eta 0:03:31 lr 0.000410 time 0.2762 (0.2932) loss 2.5271 (3.4270) grad_norm 1.8137 (1.6679) [2021-04-16 07:27:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][540/1251] eta 0:03:28 lr 0.000410 time 0.2830 (0.2928) loss 2.8000 (3.4268) grad_norm 1.4721 (1.6698) [2021-04-16 07:27:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][550/1251] eta 0:03:25 lr 0.000410 time 0.2749 (0.2925) loss 3.4973 (3.4282) grad_norm 1.5832 (1.6702) [2021-04-16 07:27:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][560/1251] eta 0:03:21 lr 0.000410 time 0.2642 (0.2923) loss 4.2291 (3.4296) grad_norm 1.4926 (1.6700) [2021-04-16 07:27:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][570/1251] eta 0:03:18 lr 0.000410 time 0.2712 (0.2921) loss 4.0002 (3.4335) grad_norm 1.6786 (1.6694) [2021-04-16 07:27:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][580/1251] eta 0:03:16 lr 0.000410 time 0.2766 (0.2923) loss 2.2383 (3.4248) grad_norm 1.8173 (1.6699) [2021-04-16 07:27:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][590/1251] eta 0:03:13 lr 0.000410 time 0.2614 (0.2922) loss 3.2087 (3.4270) grad_norm 1.5899 (1.6695) [2021-04-16 07:27:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][600/1251] eta 0:03:10 lr 0.000410 time 0.2674 (0.2921) loss 2.9575 (3.4337) grad_norm 1.5955 (1.6723) [2021-04-16 07:27:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][610/1251] eta 0:03:07 lr 0.000410 time 0.2967 (0.2920) loss 2.3037 (3.4341) grad_norm 1.4973 (1.6716) [2021-04-16 07:27:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][620/1251] eta 0:03:04 lr 0.000410 time 0.2710 (0.2919) loss 3.1650 (3.4374) grad_norm 1.4841 (1.6692) [2021-04-16 07:27:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][630/1251] eta 0:03:01 lr 0.000410 time 0.2848 (0.2916) loss 3.9373 (3.4375) grad_norm 1.9345 (1.6683) [2021-04-16 07:27:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][640/1251] eta 0:02:58 lr 0.000410 time 0.2896 (0.2916) loss 3.0036 (3.4389) grad_norm 1.5297 (1.6670) [2021-04-16 07:27:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][650/1251] eta 0:02:55 lr 0.000410 time 0.2719 (0.2916) loss 4.0957 (3.4366) grad_norm 1.5353 (1.6687) [2021-04-16 07:28:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][660/1251] eta 0:02:52 lr 0.000410 time 0.3008 (0.2913) loss 3.5366 (3.4395) grad_norm 1.7484 (1.6686) [2021-04-16 07:28:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][670/1251] eta 0:02:49 lr 0.000410 time 0.2728 (0.2912) loss 2.8276 (3.4321) grad_norm 1.7027 (1.6679) [2021-04-16 07:28:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][680/1251] eta 0:02:46 lr 0.000409 time 0.3117 (0.2910) loss 3.3395 (3.4338) grad_norm 1.6793 (1.6679) [2021-04-16 07:28:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][690/1251] eta 0:02:43 lr 0.000409 time 0.2775 (0.2908) loss 2.6746 (3.4322) grad_norm 1.6578 (1.6681) [2021-04-16 07:28:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][700/1251] eta 0:02:40 lr 0.000409 time 0.2598 (0.2905) loss 3.0368 (3.4299) grad_norm 1.7132 (1.6680) [2021-04-16 07:28:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][710/1251] eta 0:02:37 lr 0.000409 time 0.2656 (0.2903) loss 3.2805 (3.4271) grad_norm 2.1572 (1.6685) [2021-04-16 07:28:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][720/1251] eta 0:02:34 lr 0.000409 time 0.2627 (0.2902) loss 3.1591 (3.4306) grad_norm 1.5864 (1.6678) [2021-04-16 07:28:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][730/1251] eta 0:02:31 lr 0.000409 time 0.2904 (0.2900) loss 3.0500 (3.4320) grad_norm 1.7091 (1.6678) [2021-04-16 07:28:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][740/1251] eta 0:02:28 lr 0.000409 time 0.3037 (0.2901) loss 3.8007 (3.4344) grad_norm 1.6817 (1.6666) [2021-04-16 07:28:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][750/1251] eta 0:02:25 lr 0.000409 time 0.2844 (0.2900) loss 3.4420 (3.4352) grad_norm 1.6243 (1.6667) [2021-04-16 07:28:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][760/1251] eta 0:02:22 lr 0.000409 time 0.2664 (0.2897) loss 4.0999 (3.4381) grad_norm 1.6238 (1.6660) [2021-04-16 07:28:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][770/1251] eta 0:02:19 lr 0.000409 time 0.2874 (0.2895) loss 3.5195 (3.4362) grad_norm 1.8262 (1.6656) [2021-04-16 07:28:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][780/1251] eta 0:02:16 lr 0.000409 time 0.2980 (0.2894) loss 3.8120 (3.4369) grad_norm 1.7119 (1.6657) [2021-04-16 07:28:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][790/1251] eta 0:02:13 lr 0.000409 time 0.2420 (0.2894) loss 3.6058 (3.4349) grad_norm 1.6670 (1.6649) [2021-04-16 07:28:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][800/1251] eta 0:02:10 lr 0.000409 time 0.4163 (0.2893) loss 3.5059 (3.4364) grad_norm 1.8298 (1.6650) [2021-04-16 07:28:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][810/1251] eta 0:02:07 lr 0.000409 time 0.2837 (0.2892) loss 3.5645 (3.4362) grad_norm 1.3882 (1.6639) [2021-04-16 07:28:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][820/1251] eta 0:02:04 lr 0.000409 time 0.2793 (0.2890) loss 3.7546 (3.4388) grad_norm 1.8096 (1.6647) [2021-04-16 07:28:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][830/1251] eta 0:02:01 lr 0.000409 time 0.2433 (0.2890) loss 3.2093 (3.4399) grad_norm 1.6525 (1.6654) [2021-04-16 07:28:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][840/1251] eta 0:01:58 lr 0.000409 time 0.2747 (0.2888) loss 3.4467 (3.4401) grad_norm 1.5084 (1.6642) [2021-04-16 07:28:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][850/1251] eta 0:01:55 lr 0.000409 time 0.2663 (0.2887) loss 3.4502 (3.4425) grad_norm 1.5289 (1.6640) [2021-04-16 07:28:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][860/1251] eta 0:01:52 lr 0.000409 time 0.2727 (0.2886) loss 3.7589 (3.4384) grad_norm 1.5023 (1.6630) [2021-04-16 07:29:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][870/1251] eta 0:01:49 lr 0.000409 time 0.2452 (0.2884) loss 3.7359 (3.4409) grad_norm 1.7681 (1.6635) [2021-04-16 07:29:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][880/1251] eta 0:01:46 lr 0.000409 time 0.2803 (0.2883) loss 3.2301 (3.4382) grad_norm 1.6921 (1.6643) [2021-04-16 07:29:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][890/1251] eta 0:01:44 lr 0.000409 time 0.2569 (0.2882) loss 2.4454 (3.4348) grad_norm 2.0190 (1.6639) [2021-04-16 07:29:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][900/1251] eta 0:01:41 lr 0.000409 time 0.2871 (0.2880) loss 3.4888 (3.4373) grad_norm 1.7828 (1.6637) [2021-04-16 07:29:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][910/1251] eta 0:01:38 lr 0.000409 time 0.2915 (0.2878) loss 3.2748 (3.4366) grad_norm 1.5128 (1.6633) [2021-04-16 07:29:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][920/1251] eta 0:01:35 lr 0.000409 time 0.2745 (0.2877) loss 3.8921 (3.4376) grad_norm 2.3529 (1.6642) [2021-04-16 07:29:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][930/1251] eta 0:01:32 lr 0.000408 time 0.2696 (0.2877) loss 3.8131 (3.4378) grad_norm 1.4495 (1.6643) [2021-04-16 07:29:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][940/1251] eta 0:01:29 lr 0.000408 time 0.2580 (0.2879) loss 3.5168 (3.4416) grad_norm 1.5601 (1.6648) [2021-04-16 07:29:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][950/1251] eta 0:01:26 lr 0.000408 time 0.2767 (0.2877) loss 3.7676 (3.4431) grad_norm 1.7327 (1.6650) [2021-04-16 07:29:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][960/1251] eta 0:01:23 lr 0.000408 time 0.2799 (0.2876) loss 3.0218 (3.4436) grad_norm 1.4928 (1.6643) [2021-04-16 07:29:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][970/1251] eta 0:01:20 lr 0.000408 time 0.2688 (0.2874) loss 2.6149 (3.4440) grad_norm 1.7978 (1.6651) [2021-04-16 07:29:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][980/1251] eta 0:01:17 lr 0.000408 time 0.2772 (0.2873) loss 3.1773 (3.4464) grad_norm 1.6387 (1.6650) [2021-04-16 07:29:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][990/1251] eta 0:01:14 lr 0.000408 time 0.2609 (0.2872) loss 2.7840 (3.4464) grad_norm 1.8769 (1.6666) [2021-04-16 07:29:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][1000/1251] eta 0:01:12 lr 0.000408 time 0.2571 (0.2870) loss 3.2278 (3.4453) grad_norm 1.7327 (1.6689) [2021-04-16 07:29:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][1010/1251] eta 0:01:09 lr 0.000408 time 0.3063 (0.2870) loss 2.9481 (3.4447) grad_norm 1.4690 (1.6690) [2021-04-16 07:29:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][1020/1251] eta 0:01:06 lr 0.000408 time 0.2801 (0.2871) loss 2.9229 (3.4460) grad_norm 2.0110 (1.6688) [2021-04-16 07:29:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][1030/1251] eta 0:01:03 lr 0.000408 time 0.2719 (0.2870) loss 4.0281 (3.4465) grad_norm 1.5705 (1.6692) [2021-04-16 07:29:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][1040/1251] eta 0:01:00 lr 0.000408 time 0.2953 (0.2869) loss 2.8726 (3.4462) grad_norm 1.7465 (1.6689) [2021-04-16 07:29:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][1050/1251] eta 0:00:57 lr 0.000408 time 0.2901 (0.2868) loss 4.2478 (3.4490) grad_norm 1.5820 (1.6679) [2021-04-16 07:29:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][1060/1251] eta 0:00:54 lr 0.000408 time 0.2804 (0.2867) loss 3.4321 (3.4507) grad_norm 1.4908 (1.6674) [2021-04-16 07:29:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][1070/1251] eta 0:00:51 lr 0.000408 time 0.2785 (0.2867) loss 3.7690 (3.4521) grad_norm 1.5703 (1.6672) [2021-04-16 07:29:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][1080/1251] eta 0:00:49 lr 0.000408 time 0.2830 (0.2866) loss 2.5440 (3.4475) grad_norm 1.5417 (1.6672) [2021-04-16 07:30:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][1090/1251] eta 0:00:46 lr 0.000408 time 0.2822 (0.2865) loss 3.8658 (3.4471) grad_norm 1.7275 (1.6680) [2021-04-16 07:30:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][1100/1251] eta 0:00:43 lr 0.000408 time 0.2746 (0.2865) loss 3.7714 (3.4466) grad_norm 1.6556 (1.6680) [2021-04-16 07:30:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][1110/1251] eta 0:00:40 lr 0.000408 time 0.2647 (0.2864) loss 3.1896 (3.4446) grad_norm 1.5044 (1.6682) [2021-04-16 07:30:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][1120/1251] eta 0:00:37 lr 0.000408 time 0.2730 (0.2863) loss 2.4139 (3.4452) grad_norm 1.8330 (1.6683) [2021-04-16 07:30:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][1130/1251] eta 0:00:34 lr 0.000408 time 0.2688 (0.2862) loss 2.4641 (3.4446) grad_norm 1.6461 (1.6687) [2021-04-16 07:30:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][1140/1251] eta 0:00:31 lr 0.000408 time 0.2665 (0.2861) loss 3.3767 (3.4436) grad_norm 1.6233 (1.6692) [2021-04-16 07:30:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][1150/1251] eta 0:00:28 lr 0.000408 time 0.3892 (0.2863) loss 3.8732 (3.4435) grad_norm 1.4661 (1.6691) [2021-04-16 07:30:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][1160/1251] eta 0:00:26 lr 0.000408 time 0.2837 (0.2863) loss 3.4715 (3.4405) grad_norm 1.5100 (1.6681) [2021-04-16 07:30:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][1170/1251] eta 0:00:23 lr 0.000407 time 0.2749 (0.2863) loss 2.2973 (3.4381) grad_norm 1.3336 (1.6676) [2021-04-16 07:30:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][1180/1251] eta 0:00:20 lr 0.000407 time 0.2862 (0.2862) loss 3.3181 (3.4376) grad_norm 1.6118 (1.6673) [2021-04-16 07:30:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][1190/1251] eta 0:00:17 lr 0.000407 time 0.2846 (0.2862) loss 4.0741 (3.4364) grad_norm 2.2191 (1.6674) [2021-04-16 07:30:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][1200/1251] eta 0:00:14 lr 0.000407 time 0.2720 (0.2861) loss 2.9464 (3.4352) grad_norm 1.6994 (1.6674) [2021-04-16 07:30:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][1210/1251] eta 0:00:11 lr 0.000407 time 0.2799 (0.2861) loss 3.3916 (3.4351) grad_norm 1.6687 (1.6675) [2021-04-16 07:30:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][1220/1251] eta 0:00:08 lr 0.000407 time 0.2597 (0.2861) loss 3.1088 (3.4380) grad_norm 1.4611 (1.6679) [2021-04-16 07:30:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][1230/1251] eta 0:00:06 lr 0.000407 time 0.2683 (0.2860) loss 3.7410 (3.4366) grad_norm 1.9612 (1.6684) [2021-04-16 07:30:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][1240/1251] eta 0:00:03 lr 0.000407 time 0.2484 (0.2859) loss 4.2545 (3.4373) grad_norm 1.8060 (1.6684) [2021-04-16 07:30:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [168/300][1250/1251] eta 0:00:00 lr 0.000407 time 0.2482 (0.2856) loss 3.2785 (3.4381) grad_norm 1.5267 (1.6687) [2021-04-16 07:30:53 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 168 training takes 0:06:04 [2021-04-16 07:30:53 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_168.pth saving...... [2021-04-16 07:31:08 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_168.pth saved !!! [2021-04-16 07:31:09 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.253 (1.253) Loss 1.1013 (1.1013) Acc@1 73.145 (73.145) Acc@5 91.895 (91.895) [2021-04-16 07:31:10 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.106 (0.210) Loss 0.9153 (0.9487) Acc@1 78.613 (77.601) Acc@5 93.555 (93.501) [2021-04-16 07:31:12 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.145 (0.217) Loss 0.8961 (0.9606) Acc@1 78.418 (77.200) Acc@5 94.043 (93.448) [2021-04-16 07:31:15 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.358 (0.229) Loss 0.9391 (0.9579) Acc@1 77.344 (77.199) Acc@5 94.141 (93.633) [2021-04-16 07:31:17 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.217) Loss 0.9240 (0.9568) Acc@1 77.148 (77.070) Acc@5 93.945 (93.721) [2021-04-16 07:31:24 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 77.076 Acc@5 93.788 [2021-04-16 07:31:24 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 77.1% [2021-04-16 07:31:24 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 77.08% [2021-04-16 07:31:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][0/1251] eta 3:41:07 lr 0.000407 time 10.6057 (10.6057) loss 3.0950 (3.0950) grad_norm 1.5974 (1.5974) [2021-04-16 07:31:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][10/1251] eta 0:25:22 lr 0.000407 time 0.4868 (1.2269) loss 2.1729 (3.3936) grad_norm 1.5431 (1.6863) [2021-04-16 07:31:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][20/1251] eta 0:15:51 lr 0.000407 time 0.2676 (0.7727) loss 3.7740 (3.3657) grad_norm 1.3265 (1.6937) [2021-04-16 07:31:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][30/1251] eta 0:12:25 lr 0.000407 time 0.2464 (0.6106) loss 3.5636 (3.3272) grad_norm 2.0360 (1.6780) [2021-04-16 07:31:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][40/1251] eta 0:10:46 lr 0.000407 time 0.2976 (0.5335) loss 3.2461 (3.3277) grad_norm 1.4531 (1.6857) [2021-04-16 07:31:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][50/1251] eta 0:09:43 lr 0.000407 time 0.2773 (0.4857) loss 3.6196 (3.3125) grad_norm 1.5700 (1.6919) [2021-04-16 07:31:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][60/1251] eta 0:08:57 lr 0.000407 time 0.2735 (0.4514) loss 3.5451 (3.3319) grad_norm 1.5218 (1.6784) [2021-04-16 07:31:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][70/1251] eta 0:08:24 lr 0.000407 time 0.2631 (0.4269) loss 2.9056 (3.3394) grad_norm 1.8633 (1.6762) [2021-04-16 07:31:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][80/1251] eta 0:07:58 lr 0.000407 time 0.2819 (0.4087) loss 3.9039 (3.3772) grad_norm 1.6855 (1.6796) [2021-04-16 07:32:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][90/1251] eta 0:07:38 lr 0.000407 time 0.2720 (0.3952) loss 3.5449 (3.4074) grad_norm 1.9153 (1.6897) [2021-04-16 07:32:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][100/1251] eta 0:07:21 lr 0.000407 time 0.2842 (0.3833) loss 4.1398 (3.4090) grad_norm 1.5218 (1.6862) [2021-04-16 07:32:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][110/1251] eta 0:07:06 lr 0.000407 time 0.2959 (0.3736) loss 3.7853 (3.4080) grad_norm 1.5097 (1.6842) [2021-04-16 07:32:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][120/1251] eta 0:06:53 lr 0.000407 time 0.2892 (0.3654) loss 2.6242 (3.4284) grad_norm 1.5688 (1.6827) [2021-04-16 07:32:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][130/1251] eta 0:06:42 lr 0.000407 time 0.2759 (0.3594) loss 4.2471 (3.4293) grad_norm 1.5996 (1.6921) [2021-04-16 07:32:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][140/1251] eta 0:06:33 lr 0.000407 time 0.2563 (0.3546) loss 3.4319 (3.4364) grad_norm 1.5395 (1.6891) [2021-04-16 07:32:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][150/1251] eta 0:06:25 lr 0.000407 time 0.2660 (0.3499) loss 3.8038 (3.4354) grad_norm 1.8698 (1.6935) [2021-04-16 07:32:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][160/1251] eta 0:06:16 lr 0.000407 time 0.2510 (0.3452) loss 4.1686 (3.4342) grad_norm 1.5272 (1.6922) [2021-04-16 07:32:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][170/1251] eta 0:06:08 lr 0.000406 time 0.2687 (0.3412) loss 3.4593 (3.4347) grad_norm 1.4693 (1.6902) [2021-04-16 07:32:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][180/1251] eta 0:06:01 lr 0.000406 time 0.2718 (0.3376) loss 4.3691 (3.4301) grad_norm 1.6535 (1.6938) [2021-04-16 07:32:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][190/1251] eta 0:05:54 lr 0.000406 time 0.2734 (0.3344) loss 3.2371 (3.4407) grad_norm 1.5072 (1.6949) [2021-04-16 07:32:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][200/1251] eta 0:05:48 lr 0.000406 time 0.2926 (0.3319) loss 3.8484 (3.4521) grad_norm 2.2655 (1.6995) [2021-04-16 07:32:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][210/1251] eta 0:05:43 lr 0.000406 time 0.2629 (0.3295) loss 2.9772 (3.4484) grad_norm 1.6178 (1.7001) [2021-04-16 07:32:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][220/1251] eta 0:05:37 lr 0.000406 time 0.2665 (0.3271) loss 3.6550 (3.4527) grad_norm 1.6352 (1.6952) [2021-04-16 07:32:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][230/1251] eta 0:05:32 lr 0.000406 time 0.2745 (0.3256) loss 3.1515 (3.4434) grad_norm 1.6007 (1.6964) [2021-04-16 07:32:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][240/1251] eta 0:05:27 lr 0.000406 time 0.2797 (0.3237) loss 2.8590 (3.4441) grad_norm 1.9243 (1.6965) [2021-04-16 07:32:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][250/1251] eta 0:05:22 lr 0.000406 time 0.2839 (0.3218) loss 3.9904 (3.4502) grad_norm 1.6040 (1.6956) [2021-04-16 07:32:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][260/1251] eta 0:05:17 lr 0.000406 time 0.3068 (0.3201) loss 2.9536 (3.4465) grad_norm 1.6225 (1.6946) [2021-04-16 07:32:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][270/1251] eta 0:05:12 lr 0.000406 time 0.2748 (0.3185) loss 3.8411 (3.4469) grad_norm 1.7764 (1.6952) [2021-04-16 07:32:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][280/1251] eta 0:05:07 lr 0.000406 time 0.2976 (0.3170) loss 2.5386 (3.4325) grad_norm 1.7213 (1.6964) [2021-04-16 07:32:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][290/1251] eta 0:05:03 lr 0.000406 time 0.2865 (0.3161) loss 3.2534 (3.4320) grad_norm 1.3547 (1.6971) [2021-04-16 07:32:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][300/1251] eta 0:04:59 lr 0.000406 time 0.2750 (0.3147) loss 3.3197 (3.4332) grad_norm 1.4821 (1.6939) [2021-04-16 07:33:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][310/1251] eta 0:04:55 lr 0.000406 time 0.2704 (0.3136) loss 3.5034 (3.4286) grad_norm 1.7168 (1.6922) [2021-04-16 07:33:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][320/1251] eta 0:04:50 lr 0.000406 time 0.2688 (0.3125) loss 3.3537 (3.4207) grad_norm 1.4367 (1.6895) [2021-04-16 07:33:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][330/1251] eta 0:04:46 lr 0.000406 time 0.3050 (0.3115) loss 3.6386 (3.4212) grad_norm 1.8714 (1.6902) [2021-04-16 07:33:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][340/1251] eta 0:04:42 lr 0.000406 time 0.2590 (0.3106) loss 3.6791 (3.4278) grad_norm 1.4420 (1.6880) [2021-04-16 07:33:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][350/1251] eta 0:04:39 lr 0.000406 time 0.2636 (0.3098) loss 3.4448 (3.4231) grad_norm 1.8883 (1.6891) [2021-04-16 07:33:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][360/1251] eta 0:04:36 lr 0.000406 time 0.2823 (0.3099) loss 3.7767 (3.4200) grad_norm 1.6475 (1.6926) [2021-04-16 07:33:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][370/1251] eta 0:04:32 lr 0.000406 time 0.2821 (0.3091) loss 2.5260 (3.4223) grad_norm 1.5421 (1.6909) [2021-04-16 07:33:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][380/1251] eta 0:04:28 lr 0.000406 time 0.2759 (0.3082) loss 3.7320 (3.4232) grad_norm 2.0839 (1.6905) [2021-04-16 07:33:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][390/1251] eta 0:04:24 lr 0.000406 time 0.2714 (0.3073) loss 3.5552 (3.4317) grad_norm 1.6561 (1.6896) [2021-04-16 07:33:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][400/1251] eta 0:04:20 lr 0.000406 time 0.2857 (0.3066) loss 3.9302 (3.4307) grad_norm 1.8470 (1.6904) [2021-04-16 07:33:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][410/1251] eta 0:04:17 lr 0.000405 time 0.2866 (0.3059) loss 3.1402 (3.4275) grad_norm 1.6001 (1.6893) [2021-04-16 07:33:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][420/1251] eta 0:04:13 lr 0.000405 time 0.2827 (0.3054) loss 2.6179 (3.4253) grad_norm 1.9635 (1.6901) [2021-04-16 07:33:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][430/1251] eta 0:04:10 lr 0.000405 time 0.2762 (0.3049) loss 3.3557 (3.4297) grad_norm 1.6352 (1.6916) [2021-04-16 07:33:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][440/1251] eta 0:04:06 lr 0.000405 time 0.2877 (0.3042) loss 3.2630 (3.4256) grad_norm 1.6601 (1.6894) [2021-04-16 07:33:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][450/1251] eta 0:04:03 lr 0.000405 time 0.2708 (0.3036) loss 3.9566 (3.4308) grad_norm 1.6530 (1.6893) [2021-04-16 07:33:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][460/1251] eta 0:03:59 lr 0.000405 time 0.2671 (0.3029) loss 3.8076 (3.4236) grad_norm 1.6920 (1.6870) [2021-04-16 07:33:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][470/1251] eta 0:03:56 lr 0.000405 time 0.2902 (0.3024) loss 3.8872 (3.4298) grad_norm 1.6316 (1.6866) [2021-04-16 07:33:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][480/1251] eta 0:03:52 lr 0.000405 time 0.2858 (0.3020) loss 2.8905 (3.4262) grad_norm 1.7684 (1.6874) [2021-04-16 07:33:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][490/1251] eta 0:03:49 lr 0.000405 time 0.2790 (0.3014) loss 2.9192 (3.4254) grad_norm 1.8301 (1.6870) [2021-04-16 07:33:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][500/1251] eta 0:03:45 lr 0.000405 time 0.2731 (0.3008) loss 3.7156 (3.4241) grad_norm 1.5733 (1.6867) [2021-04-16 07:33:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][510/1251] eta 0:03:42 lr 0.000405 time 0.2976 (0.3003) loss 3.4562 (3.4251) grad_norm 1.6944 (1.6881) [2021-04-16 07:34:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][520/1251] eta 0:03:39 lr 0.000405 time 0.2846 (0.2999) loss 4.1400 (3.4283) grad_norm 1.8769 (1.6912) [2021-04-16 07:34:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][530/1251] eta 0:03:35 lr 0.000405 time 0.2786 (0.2994) loss 4.2000 (3.4277) grad_norm 1.9522 (1.6908) [2021-04-16 07:34:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][540/1251] eta 0:03:32 lr 0.000405 time 0.2699 (0.2991) loss 3.5844 (3.4262) grad_norm 1.4809 (1.6904) [2021-04-16 07:34:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][550/1251] eta 0:03:29 lr 0.000405 time 0.2840 (0.2986) loss 3.0733 (3.4262) grad_norm 1.6662 (1.6889) [2021-04-16 07:34:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][560/1251] eta 0:03:26 lr 0.000405 time 0.2826 (0.2984) loss 3.9798 (3.4294) grad_norm 1.7013 (1.6891) [2021-04-16 07:34:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][570/1251] eta 0:03:23 lr 0.000405 time 0.2982 (0.2985) loss 3.1991 (3.4227) grad_norm 1.4132 (1.6877) [2021-04-16 07:34:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][580/1251] eta 0:03:20 lr 0.000405 time 0.2786 (0.2983) loss 2.0648 (3.4269) grad_norm 1.8897 (1.6889) [2021-04-16 07:34:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][590/1251] eta 0:03:17 lr 0.000405 time 0.2658 (0.2982) loss 3.6015 (3.4296) grad_norm 1.6583 (1.6877) [2021-04-16 07:34:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][600/1251] eta 0:03:13 lr 0.000405 time 0.2737 (0.2978) loss 2.5444 (3.4272) grad_norm 1.8039 (1.6878) [2021-04-16 07:34:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][610/1251] eta 0:03:10 lr 0.000405 time 0.2998 (0.2975) loss 3.2334 (3.4248) grad_norm 1.6504 (1.6894) [2021-04-16 07:34:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][620/1251] eta 0:03:07 lr 0.000405 time 0.2961 (0.2975) loss 4.0049 (3.4294) grad_norm 1.7544 (1.6889) [2021-04-16 07:34:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][630/1251] eta 0:03:04 lr 0.000405 time 0.2706 (0.2973) loss 2.7502 (3.4310) grad_norm 1.5533 (1.6880) [2021-04-16 07:34:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][640/1251] eta 0:03:01 lr 0.000405 time 0.2881 (0.2970) loss 3.4283 (3.4346) grad_norm 1.6304 (1.6866) [2021-04-16 07:34:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][650/1251] eta 0:02:58 lr 0.000405 time 0.2751 (0.2966) loss 3.4197 (3.4379) grad_norm 1.6797 (1.6864) [2021-04-16 07:34:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][660/1251] eta 0:02:55 lr 0.000404 time 0.2892 (0.2963) loss 3.9335 (3.4374) grad_norm 1.8125 (1.6854) [2021-04-16 07:34:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][670/1251] eta 0:02:51 lr 0.000404 time 0.2850 (0.2960) loss 2.3196 (3.4355) grad_norm 1.9562 (1.6847) [2021-04-16 07:34:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][680/1251] eta 0:02:48 lr 0.000404 time 0.2740 (0.2957) loss 3.2882 (3.4370) grad_norm 1.6392 (1.6853) [2021-04-16 07:34:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][690/1251] eta 0:02:45 lr 0.000404 time 0.2901 (0.2955) loss 3.9825 (3.4442) grad_norm 1.6916 (1.6846) [2021-04-16 07:34:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][700/1251] eta 0:02:42 lr 0.000404 time 0.2821 (0.2952) loss 3.8193 (3.4457) grad_norm 2.9046 (1.6858) [2021-04-16 07:34:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][710/1251] eta 0:02:39 lr 0.000404 time 0.2643 (0.2951) loss 4.1317 (3.4489) grad_norm 1.8730 (1.6877) [2021-04-16 07:34:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][720/1251] eta 0:02:36 lr 0.000404 time 0.2775 (0.2951) loss 4.0014 (3.4533) grad_norm 1.6384 (1.6876) [2021-04-16 07:35:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][730/1251] eta 0:02:33 lr 0.000404 time 0.2531 (0.2948) loss 3.8428 (3.4534) grad_norm 1.5943 (1.6879) [2021-04-16 07:35:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][740/1251] eta 0:02:30 lr 0.000404 time 0.2851 (0.2945) loss 3.7165 (3.4503) grad_norm 1.5249 (1.6884) [2021-04-16 07:35:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][750/1251] eta 0:02:27 lr 0.000404 time 0.2637 (0.2942) loss 4.1650 (3.4510) grad_norm 1.4685 (1.6875) [2021-04-16 07:35:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][760/1251] eta 0:02:24 lr 0.000404 time 0.3121 (0.2940) loss 3.8393 (3.4476) grad_norm 1.6265 (1.6869) [2021-04-16 07:35:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][770/1251] eta 0:02:21 lr 0.000404 time 0.2614 (0.2938) loss 4.0127 (3.4489) grad_norm 1.5901 (1.6859) [2021-04-16 07:35:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][780/1251] eta 0:02:18 lr 0.000404 time 0.2584 (0.2936) loss 3.6359 (3.4495) grad_norm 1.6866 (1.6857) [2021-04-16 07:35:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][790/1251] eta 0:02:15 lr 0.000404 time 0.2757 (0.2934) loss 2.7120 (3.4467) grad_norm 1.5459 (1.6853) [2021-04-16 07:35:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][800/1251] eta 0:02:12 lr 0.000404 time 0.4201 (0.2934) loss 4.3928 (3.4515) grad_norm 1.8231 (1.6850) [2021-04-16 07:35:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][810/1251] eta 0:02:09 lr 0.000404 time 0.2912 (0.2932) loss 3.9140 (3.4508) grad_norm 1.6230 (1.6850) [2021-04-16 07:35:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][820/1251] eta 0:02:06 lr 0.000404 time 0.2679 (0.2930) loss 3.8085 (3.4484) grad_norm 1.7614 (1.6866) [2021-04-16 07:35:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][830/1251] eta 0:02:03 lr 0.000404 time 0.2868 (0.2928) loss 2.8599 (3.4474) grad_norm 1.5810 (1.6863) [2021-04-16 07:35:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][840/1251] eta 0:02:00 lr 0.000404 time 0.2718 (0.2926) loss 3.2101 (3.4435) grad_norm 1.8117 (1.6859) [2021-04-16 07:35:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][850/1251] eta 0:01:57 lr 0.000404 time 0.2657 (0.2925) loss 3.8036 (3.4465) grad_norm 1.5567 (1.6874) [2021-04-16 07:35:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][860/1251] eta 0:01:54 lr 0.000404 time 0.2887 (0.2923) loss 3.1258 (3.4428) grad_norm 1.6158 (1.6876) [2021-04-16 07:35:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][870/1251] eta 0:01:51 lr 0.000404 time 0.2792 (0.2922) loss 3.1788 (3.4403) grad_norm 1.8239 (1.6885) [2021-04-16 07:35:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][880/1251] eta 0:01:48 lr 0.000404 time 0.2834 (0.2920) loss 2.9518 (3.4409) grad_norm 1.7228 (1.6888) [2021-04-16 07:35:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][890/1251] eta 0:01:45 lr 0.000404 time 0.2761 (0.2918) loss 2.7747 (3.4395) grad_norm 1.6383 (1.6886) [2021-04-16 07:35:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][900/1251] eta 0:01:42 lr 0.000404 time 0.2794 (0.2917) loss 2.7386 (3.4373) grad_norm 1.9949 (1.6890) [2021-04-16 07:35:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][910/1251] eta 0:01:39 lr 0.000403 time 0.2835 (0.2915) loss 3.3633 (3.4344) grad_norm 1.6059 (1.6890) [2021-04-16 07:35:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][920/1251] eta 0:01:36 lr 0.000403 time 0.2871 (0.2913) loss 2.9709 (3.4342) grad_norm 2.0820 (1.6904) [2021-04-16 07:35:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][930/1251] eta 0:01:33 lr 0.000403 time 0.2754 (0.2912) loss 3.6622 (3.4322) grad_norm 1.5389 (1.6900) [2021-04-16 07:35:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][940/1251] eta 0:01:30 lr 0.000403 time 0.2677 (0.2913) loss 3.9541 (3.4363) grad_norm 1.6163 (1.6889) [2021-04-16 07:36:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][950/1251] eta 0:01:27 lr 0.000403 time 0.2602 (0.2911) loss 4.2026 (3.4361) grad_norm 1.4176 (1.6882) [2021-04-16 07:36:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][960/1251] eta 0:01:24 lr 0.000403 time 0.2990 (0.2910) loss 3.4963 (3.4369) grad_norm 1.7745 (1.6873) [2021-04-16 07:36:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][970/1251] eta 0:01:21 lr 0.000403 time 0.2594 (0.2908) loss 4.1204 (3.4355) grad_norm 1.7410 (1.6878) [2021-04-16 07:36:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][980/1251] eta 0:01:18 lr 0.000403 time 0.2747 (0.2907) loss 2.9591 (3.4355) grad_norm 1.6959 (1.6870) [2021-04-16 07:36:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][990/1251] eta 0:01:15 lr 0.000403 time 0.2699 (0.2906) loss 3.6127 (3.4315) grad_norm 1.4841 (1.6864) [2021-04-16 07:36:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][1000/1251] eta 0:01:12 lr 0.000403 time 0.2651 (0.2905) loss 4.2689 (3.4304) grad_norm 1.5708 (1.6858) [2021-04-16 07:36:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][1010/1251] eta 0:01:10 lr 0.000403 time 0.3053 (0.2906) loss 3.7479 (3.4279) grad_norm 1.7245 (1.6862) [2021-04-16 07:36:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][1020/1251] eta 0:01:07 lr 0.000403 time 0.2853 (0.2905) loss 2.8715 (3.4293) grad_norm 1.6640 (1.6853) [2021-04-16 07:36:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][1030/1251] eta 0:01:04 lr 0.000403 time 0.2662 (0.2904) loss 2.5828 (3.4294) grad_norm 1.5712 (1.6854) [2021-04-16 07:36:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][1040/1251] eta 0:01:01 lr 0.000403 time 0.2694 (0.2903) loss 3.2181 (3.4306) grad_norm 1.4895 (1.6844) [2021-04-16 07:36:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][1050/1251] eta 0:00:58 lr 0.000403 time 0.2841 (0.2901) loss 3.6189 (3.4304) grad_norm 1.7125 (1.6842) [2021-04-16 07:36:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][1060/1251] eta 0:00:55 lr 0.000403 time 0.2753 (0.2900) loss 4.3775 (3.4330) grad_norm 1.7138 (1.6840) [2021-04-16 07:36:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][1070/1251] eta 0:00:52 lr 0.000403 time 0.2650 (0.2899) loss 4.1028 (3.4335) grad_norm 1.7618 (1.6845) [2021-04-16 07:36:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][1080/1251] eta 0:00:49 lr 0.000403 time 0.2943 (0.2898) loss 3.4073 (3.4326) grad_norm 1.4100 (1.6852) [2021-04-16 07:36:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][1090/1251] eta 0:00:46 lr 0.000403 time 0.2840 (0.2897) loss 4.1501 (3.4340) grad_norm 1.6718 (1.6866) [2021-04-16 07:36:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][1100/1251] eta 0:00:43 lr 0.000403 time 0.2543 (0.2896) loss 3.9342 (3.4324) grad_norm 2.0566 (1.6878) [2021-04-16 07:36:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][1110/1251] eta 0:00:40 lr 0.000403 time 0.2861 (0.2895) loss 3.8787 (3.4327) grad_norm 1.8276 (1.6883) [2021-04-16 07:36:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][1120/1251] eta 0:00:37 lr 0.000403 time 0.2742 (0.2894) loss 3.7291 (3.4315) grad_norm 1.6574 (1.6880) [2021-04-16 07:36:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][1130/1251] eta 0:00:35 lr 0.000403 time 0.3204 (0.2893) loss 2.7974 (3.4309) grad_norm 1.5399 (1.6876) [2021-04-16 07:36:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][1140/1251] eta 0:00:32 lr 0.000403 time 0.2557 (0.2893) loss 3.5888 (3.4327) grad_norm 1.8937 (1.6870) [2021-04-16 07:36:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][1150/1251] eta 0:00:29 lr 0.000402 time 0.2788 (0.2893) loss 3.2616 (3.4343) grad_norm 1.8441 (1.6876) [2021-04-16 07:37:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][1160/1251] eta 0:00:26 lr 0.000402 time 0.2970 (0.2893) loss 3.3169 (3.4342) grad_norm 1.5701 (1.6871) [2021-04-16 07:37:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][1170/1251] eta 0:00:23 lr 0.000402 time 0.2725 (0.2892) loss 2.9192 (3.4358) grad_norm 1.6669 (1.6863) [2021-04-16 07:37:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][1180/1251] eta 0:00:20 lr 0.000402 time 0.3178 (0.2892) loss 2.7868 (3.4360) grad_norm 1.7726 (1.6865) [2021-04-16 07:37:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][1190/1251] eta 0:00:17 lr 0.000402 time 0.4051 (0.2892) loss 4.1183 (3.4357) grad_norm 1.6051 (1.6868) [2021-04-16 07:37:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][1200/1251] eta 0:00:14 lr 0.000402 time 0.2668 (0.2890) loss 2.8981 (3.4341) grad_norm 1.6769 (1.6870) [2021-04-16 07:37:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][1210/1251] eta 0:00:11 lr 0.000402 time 0.2766 (0.2889) loss 3.4098 (3.4312) grad_norm 1.7640 (1.6865) [2021-04-16 07:37:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][1220/1251] eta 0:00:08 lr 0.000402 time 0.2785 (0.2888) loss 3.4963 (3.4298) grad_norm 1.7108 (1.6872) [2021-04-16 07:37:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][1230/1251] eta 0:00:06 lr 0.000402 time 0.2824 (0.2886) loss 3.8492 (3.4312) grad_norm 1.6014 (1.6865) [2021-04-16 07:37:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][1240/1251] eta 0:00:03 lr 0.000402 time 0.2479 (0.2884) loss 4.1314 (3.4336) grad_norm 1.6016 (1.6859) [2021-04-16 07:37:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [169/300][1250/1251] eta 0:00:00 lr 0.000402 time 0.2504 (0.2881) loss 2.6579 (3.4361) grad_norm 1.6206 (1.6853) [2021-04-16 07:37:29 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 169 training takes 0:06:04 [2021-04-16 07:37:29 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_169.pth saving...... [2021-04-16 07:37:41 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_169.pth saved !!! [2021-04-16 07:37:42 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.192 (1.192) Loss 1.0072 (1.0072) Acc@1 76.855 (76.855) Acc@5 93.555 (93.555) [2021-04-16 07:37:43 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.153 (0.221) Loss 0.9941 (0.9880) Acc@1 75.684 (76.989) Acc@5 94.336 (93.963) [2021-04-16 07:37:45 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.125 (0.206) Loss 1.0044 (0.9968) Acc@1 76.367 (76.809) Acc@5 93.652 (93.745) [2021-04-16 07:37:48 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.131 (0.222) Loss 1.0075 (0.9999) Acc@1 75.195 (76.815) Acc@5 94.531 (93.709) [2021-04-16 07:37:50 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.215) Loss 1.0836 (1.0013) Acc@1 74.316 (76.801) Acc@5 92.285 (93.648) [2021-04-16 07:37:55 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 76.814 Acc@5 93.662 [2021-04-16 07:37:55 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 76.8% [2021-04-16 07:37:55 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 77.08% [2021-04-16 07:38:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][0/1251] eta 2:07:45 lr 0.000402 time 6.1278 (6.1278) loss 3.8539 (3.8539) grad_norm 1.6337 (1.6337) [2021-04-16 07:38:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][10/1251] eta 0:16:57 lr 0.000402 time 0.4643 (0.8197) loss 3.6205 (3.4290) grad_norm 1.6195 (1.6235) [2021-04-16 07:38:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][20/1251] eta 0:11:30 lr 0.000402 time 0.2892 (0.5612) loss 3.7003 (3.4583) grad_norm 1.7576 (1.6527) [2021-04-16 07:38:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][30/1251] eta 0:09:35 lr 0.000402 time 0.2917 (0.4716) loss 3.9715 (3.4449) grad_norm 1.7976 (1.6427) [2021-04-16 07:38:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][40/1251] eta 0:08:34 lr 0.000402 time 0.2879 (0.4248) loss 3.4512 (3.4870) grad_norm 1.7093 (1.6360) [2021-04-16 07:38:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][50/1251] eta 0:07:56 lr 0.000402 time 0.2731 (0.3968) loss 3.6576 (3.4941) grad_norm 1.6334 (1.6346) [2021-04-16 07:38:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][60/1251] eta 0:07:30 lr 0.000402 time 0.3067 (0.3786) loss 3.9385 (3.5055) grad_norm 1.8608 (1.6628) [2021-04-16 07:38:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][70/1251] eta 0:07:11 lr 0.000402 time 0.2801 (0.3655) loss 3.0615 (3.4698) grad_norm 1.5313 (1.6631) [2021-04-16 07:38:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][80/1251] eta 0:06:56 lr 0.000402 time 0.2678 (0.3561) loss 2.6770 (3.4325) grad_norm 2.0567 (1.6765) [2021-04-16 07:38:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][90/1251] eta 0:06:43 lr 0.000402 time 0.2772 (0.3477) loss 3.3410 (3.4254) grad_norm 1.5552 (1.6760) [2021-04-16 07:38:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][100/1251] eta 0:06:32 lr 0.000402 time 0.2601 (0.3411) loss 3.7429 (3.4046) grad_norm 1.7677 (1.6732) [2021-04-16 07:38:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][110/1251] eta 0:06:23 lr 0.000402 time 0.2881 (0.3361) loss 3.2510 (3.3992) grad_norm 1.5756 (1.6770) [2021-04-16 07:38:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][120/1251] eta 0:06:14 lr 0.000402 time 0.2760 (0.3312) loss 3.3928 (3.4064) grad_norm 1.5334 (1.6755) [2021-04-16 07:38:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][130/1251] eta 0:06:06 lr 0.000402 time 0.2776 (0.3270) loss 2.2295 (3.4156) grad_norm 1.5211 (1.6841) [2021-04-16 07:38:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][140/1251] eta 0:06:01 lr 0.000402 time 0.2938 (0.3252) loss 2.7776 (3.4089) grad_norm 1.7248 (1.6877) [2021-04-16 07:38:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][150/1251] eta 0:05:55 lr 0.000401 time 0.2814 (0.3232) loss 3.5620 (3.4267) grad_norm 1.7293 (1.6871) [2021-04-16 07:38:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][160/1251] eta 0:05:50 lr 0.000401 time 0.3005 (0.3211) loss 3.4853 (3.4427) grad_norm 1.5425 (1.6836) [2021-04-16 07:38:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][170/1251] eta 0:05:44 lr 0.000401 time 0.2592 (0.3188) loss 3.0559 (3.4552) grad_norm 1.5454 (1.6829) [2021-04-16 07:38:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][180/1251] eta 0:05:39 lr 0.000401 time 0.2961 (0.3168) loss 2.4974 (3.4404) grad_norm 1.9715 (1.6809) [2021-04-16 07:38:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][190/1251] eta 0:05:34 lr 0.000401 time 0.2756 (0.3148) loss 3.4476 (3.4447) grad_norm 1.5089 (1.6808) [2021-04-16 07:38:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][200/1251] eta 0:05:28 lr 0.000401 time 0.2534 (0.3128) loss 3.4776 (3.4350) grad_norm 2.0522 (1.6862) [2021-04-16 07:39:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][210/1251] eta 0:05:24 lr 0.000401 time 0.2552 (0.3112) loss 2.7227 (3.4304) grad_norm 1.6555 (1.6812) [2021-04-16 07:39:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][220/1251] eta 0:05:20 lr 0.000401 time 0.2691 (0.3106) loss 3.1910 (3.4237) grad_norm 1.5694 (1.6828) [2021-04-16 07:39:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][230/1251] eta 0:05:15 lr 0.000401 time 0.2839 (0.3093) loss 3.5074 (3.4273) grad_norm 1.8627 (1.6835) [2021-04-16 07:39:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][240/1251] eta 0:05:11 lr 0.000401 time 0.2959 (0.3082) loss 3.1424 (3.4215) grad_norm 1.7947 (1.6842) [2021-04-16 07:39:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][250/1251] eta 0:05:07 lr 0.000401 time 0.2740 (0.3072) loss 2.6673 (3.4273) grad_norm 2.0733 (1.6894) [2021-04-16 07:39:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][260/1251] eta 0:05:03 lr 0.000401 time 0.2986 (0.3062) loss 3.9003 (3.4247) grad_norm 1.6175 (1.6911) [2021-04-16 07:39:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][270/1251] eta 0:04:59 lr 0.000401 time 0.3014 (0.3055) loss 2.8073 (3.4266) grad_norm 1.4883 (1.6920) [2021-04-16 07:39:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][280/1251] eta 0:04:55 lr 0.000401 time 0.2598 (0.3045) loss 3.2865 (3.4292) grad_norm 1.6943 (1.6919) [2021-04-16 07:39:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][290/1251] eta 0:04:51 lr 0.000401 time 0.2798 (0.3038) loss 3.2727 (3.4364) grad_norm 1.5049 (1.6907) [2021-04-16 07:39:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][300/1251] eta 0:04:48 lr 0.000401 time 0.2890 (0.3032) loss 2.9277 (3.4367) grad_norm 1.4974 (1.6893) [2021-04-16 07:39:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][310/1251] eta 0:04:44 lr 0.000401 time 0.2883 (0.3025) loss 4.0408 (3.4368) grad_norm 1.5889 (1.6869) [2021-04-16 07:39:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][320/1251] eta 0:04:41 lr 0.000401 time 0.2692 (0.3019) loss 3.3660 (3.4361) grad_norm 2.0735 (1.6886) [2021-04-16 07:39:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][330/1251] eta 0:04:37 lr 0.000401 time 0.2986 (0.3014) loss 3.6726 (3.4340) grad_norm 1.8772 (1.6874) [2021-04-16 07:39:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][340/1251] eta 0:04:34 lr 0.000401 time 0.2677 (0.3010) loss 3.8823 (3.4370) grad_norm 1.6701 (1.6850) [2021-04-16 07:39:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][350/1251] eta 0:04:30 lr 0.000401 time 0.2886 (0.3006) loss 3.1875 (3.4410) grad_norm 1.7214 (1.6876) [2021-04-16 07:39:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][360/1251] eta 0:04:27 lr 0.000401 time 0.2693 (0.3002) loss 2.4189 (3.4349) grad_norm 1.3786 (1.6876) [2021-04-16 07:39:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][370/1251] eta 0:04:24 lr 0.000401 time 0.2816 (0.3003) loss 3.3951 (3.4328) grad_norm 1.6887 (1.6859) [2021-04-16 07:39:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][380/1251] eta 0:04:21 lr 0.000401 time 0.2759 (0.2997) loss 3.4420 (3.4346) grad_norm 1.5156 (1.6838) [2021-04-16 07:39:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][390/1251] eta 0:04:17 lr 0.000401 time 0.2734 (0.2995) loss 4.0932 (3.4369) grad_norm 1.5527 (1.6833) [2021-04-16 07:39:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][400/1251] eta 0:04:14 lr 0.000400 time 0.2708 (0.2989) loss 3.7075 (3.4324) grad_norm 1.4435 (1.6809) [2021-04-16 07:39:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][410/1251] eta 0:04:10 lr 0.000400 time 0.2953 (0.2984) loss 3.3738 (3.4293) grad_norm 1.8323 (1.6822) [2021-04-16 07:40:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][420/1251] eta 0:04:07 lr 0.000400 time 0.2581 (0.2980) loss 3.6245 (3.4267) grad_norm 1.7233 (1.6856) [2021-04-16 07:40:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][430/1251] eta 0:04:04 lr 0.000400 time 0.2936 (0.2976) loss 3.4780 (3.4265) grad_norm 1.5617 (1.6843) [2021-04-16 07:40:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][440/1251] eta 0:04:00 lr 0.000400 time 0.2723 (0.2971) loss 3.8196 (3.4262) grad_norm 1.4350 (1.6827) [2021-04-16 07:40:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][450/1251] eta 0:03:57 lr 0.000400 time 0.2969 (0.2968) loss 4.1416 (3.4253) grad_norm 1.5425 (1.6810) [2021-04-16 07:40:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][460/1251] eta 0:03:54 lr 0.000400 time 0.2859 (0.2964) loss 3.5702 (3.4229) grad_norm 1.7646 (1.6832) [2021-04-16 07:40:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][470/1251] eta 0:03:51 lr 0.000400 time 0.2494 (0.2959) loss 3.2279 (3.4268) grad_norm 1.8201 (1.6842) [2021-04-16 07:40:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][480/1251] eta 0:03:47 lr 0.000400 time 0.3012 (0.2956) loss 3.6386 (3.4210) grad_norm 1.3841 (1.6832) [2021-04-16 07:40:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][490/1251] eta 0:03:44 lr 0.000400 time 0.2795 (0.2952) loss 3.8031 (3.4272) grad_norm 1.6810 (1.6829) [2021-04-16 07:40:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][500/1251] eta 0:03:41 lr 0.000400 time 0.2899 (0.2950) loss 2.9008 (3.4267) grad_norm 1.6210 (1.6817) [2021-04-16 07:40:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][510/1251] eta 0:03:38 lr 0.000400 time 0.2669 (0.2947) loss 3.6712 (3.4290) grad_norm 1.6263 (1.6819) [2021-04-16 07:40:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][520/1251] eta 0:03:35 lr 0.000400 time 0.2667 (0.2945) loss 3.5498 (3.4246) grad_norm 1.6444 (1.6813) [2021-04-16 07:40:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][530/1251] eta 0:03:32 lr 0.000400 time 0.2824 (0.2943) loss 3.0161 (3.4186) grad_norm 1.7867 (1.6796) [2021-04-16 07:40:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][540/1251] eta 0:03:29 lr 0.000400 time 0.3052 (0.2945) loss 3.6490 (3.4217) grad_norm 1.6856 (1.6804) [2021-04-16 07:40:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][550/1251] eta 0:03:26 lr 0.000400 time 0.2656 (0.2942) loss 2.8268 (3.4183) grad_norm 1.5662 (1.6827) [2021-04-16 07:40:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][560/1251] eta 0:03:23 lr 0.000400 time 0.2802 (0.2939) loss 2.3805 (3.4162) grad_norm 1.6626 (1.6837) [2021-04-16 07:40:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][570/1251] eta 0:03:19 lr 0.000400 time 0.2800 (0.2936) loss 3.2087 (3.4134) grad_norm 2.1338 (1.6846) [2021-04-16 07:40:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][580/1251] eta 0:03:17 lr 0.000400 time 0.2818 (0.2936) loss 3.6472 (3.4161) grad_norm 1.5457 (1.6857) [2021-04-16 07:40:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][590/1251] eta 0:03:14 lr 0.000400 time 0.2905 (0.2936) loss 2.7179 (3.4204) grad_norm 1.6474 (1.6856) [2021-04-16 07:40:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][600/1251] eta 0:03:11 lr 0.000400 time 0.2918 (0.2934) loss 3.7217 (3.4236) grad_norm 1.6575 (1.6855) [2021-04-16 07:40:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][610/1251] eta 0:03:07 lr 0.000400 time 0.2743 (0.2931) loss 3.3349 (3.4252) grad_norm 1.7642 (1.6859) [2021-04-16 07:40:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][620/1251] eta 0:03:04 lr 0.000400 time 0.3064 (0.2929) loss 4.1222 (3.4255) grad_norm 1.7517 (1.6860) [2021-04-16 07:41:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][630/1251] eta 0:03:01 lr 0.000400 time 0.2666 (0.2928) loss 3.0472 (3.4238) grad_norm 1.5874 (1.6865) [2021-04-16 07:41:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][640/1251] eta 0:02:58 lr 0.000399 time 0.2874 (0.2926) loss 3.9471 (3.4258) grad_norm 1.5924 (1.6852) [2021-04-16 07:41:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][650/1251] eta 0:02:55 lr 0.000399 time 0.2646 (0.2925) loss 3.2514 (3.4252) grad_norm 1.4435 (1.6840) [2021-04-16 07:41:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][660/1251] eta 0:02:52 lr 0.000399 time 0.2781 (0.2923) loss 3.5600 (3.4253) grad_norm 1.9946 (1.6842) [2021-04-16 07:41:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][670/1251] eta 0:02:49 lr 0.000399 time 0.2549 (0.2920) loss 2.6547 (3.4270) grad_norm 1.4650 (1.6837) [2021-04-16 07:41:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][680/1251] eta 0:02:46 lr 0.000399 time 0.2917 (0.2918) loss 3.8468 (3.4297) grad_norm 1.8232 (1.6845) [2021-04-16 07:41:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][690/1251] eta 0:02:43 lr 0.000399 time 0.2772 (0.2917) loss 4.3563 (3.4306) grad_norm 1.5637 (1.6855) [2021-04-16 07:41:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][700/1251] eta 0:02:40 lr 0.000399 time 0.3168 (0.2915) loss 2.6413 (3.4337) grad_norm 1.7458 (1.6850) [2021-04-16 07:41:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][710/1251] eta 0:02:37 lr 0.000399 time 0.2833 (0.2913) loss 2.7674 (3.4353) grad_norm 1.7098 (1.6849) [2021-04-16 07:41:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][720/1251] eta 0:02:34 lr 0.000399 time 0.2599 (0.2914) loss 3.2893 (3.4381) grad_norm 1.6008 (1.6842) [2021-04-16 07:41:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][730/1251] eta 0:02:31 lr 0.000399 time 0.2728 (0.2914) loss 3.9668 (3.4425) grad_norm 1.7051 (1.6846) [2021-04-16 07:41:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][740/1251] eta 0:02:28 lr 0.000399 time 0.2824 (0.2912) loss 4.1417 (3.4445) grad_norm 1.9104 (1.6838) [2021-04-16 07:41:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][750/1251] eta 0:02:25 lr 0.000399 time 0.2840 (0.2911) loss 1.9512 (3.4454) grad_norm 1.6132 (1.6838) [2021-04-16 07:41:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][760/1251] eta 0:02:22 lr 0.000399 time 0.2685 (0.2909) loss 4.0446 (3.4457) grad_norm 1.7605 (1.6835) [2021-04-16 07:41:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][770/1251] eta 0:02:19 lr 0.000399 time 0.2660 (0.2907) loss 4.0937 (3.4430) grad_norm 1.5262 (1.6831) [2021-04-16 07:41:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][780/1251] eta 0:02:16 lr 0.000399 time 0.2662 (0.2905) loss 3.8068 (3.4424) grad_norm 1.8414 (1.6835) [2021-04-16 07:41:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][790/1251] eta 0:02:14 lr 0.000399 time 0.4068 (0.2907) loss 3.8110 (3.4433) grad_norm 1.7199 (1.6838) [2021-04-16 07:41:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][800/1251] eta 0:02:11 lr 0.000399 time 0.2711 (0.2906) loss 3.3898 (3.4445) grad_norm 1.7864 (1.6836) [2021-04-16 07:41:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][810/1251] eta 0:02:08 lr 0.000399 time 0.3024 (0.2905) loss 4.0082 (3.4447) grad_norm 1.6101 (1.6833) [2021-04-16 07:41:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][820/1251] eta 0:02:05 lr 0.000399 time 0.3008 (0.2903) loss 3.9497 (3.4442) grad_norm 1.9841 (1.6839) [2021-04-16 07:41:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][830/1251] eta 0:02:02 lr 0.000399 time 0.2686 (0.2901) loss 3.5648 (3.4482) grad_norm 2.2828 (1.6838) [2021-04-16 07:41:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][840/1251] eta 0:01:59 lr 0.000399 time 0.2883 (0.2900) loss 3.4766 (3.4482) grad_norm 1.8774 (1.6854) [2021-04-16 07:42:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][850/1251] eta 0:01:56 lr 0.000399 time 0.3026 (0.2899) loss 3.0637 (3.4479) grad_norm 1.7386 (1.6860) [2021-04-16 07:42:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][860/1251] eta 0:01:53 lr 0.000399 time 0.2780 (0.2898) loss 4.1054 (3.4467) grad_norm 1.4952 (1.6850) [2021-04-16 07:42:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][870/1251] eta 0:01:50 lr 0.000399 time 0.2899 (0.2897) loss 4.1160 (3.4476) grad_norm 1.7734 (1.6851) [2021-04-16 07:42:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][880/1251] eta 0:01:47 lr 0.000399 time 0.2906 (0.2895) loss 3.9522 (3.4485) grad_norm 1.6653 (1.6849) [2021-04-16 07:42:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][890/1251] eta 0:01:44 lr 0.000398 time 0.2766 (0.2895) loss 4.0199 (3.4506) grad_norm 2.2257 (1.6863) [2021-04-16 07:42:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][900/1251] eta 0:01:41 lr 0.000398 time 0.2729 (0.2895) loss 2.1789 (3.4509) grad_norm 1.5856 (1.6865) [2021-04-16 07:42:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][910/1251] eta 0:01:38 lr 0.000398 time 0.2714 (0.2894) loss 2.8924 (3.4483) grad_norm 1.8325 (1.6875) [2021-04-16 07:42:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][920/1251] eta 0:01:35 lr 0.000398 time 0.2771 (0.2895) loss 3.8908 (3.4477) grad_norm 1.6368 (1.6888) [2021-04-16 07:42:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][930/1251] eta 0:01:32 lr 0.000398 time 0.2782 (0.2894) loss 4.2346 (3.4493) grad_norm 1.8057 (1.6892) [2021-04-16 07:42:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][940/1251] eta 0:01:30 lr 0.000398 time 0.2888 (0.2897) loss 3.8229 (3.4514) grad_norm 1.6643 (1.6900) [2021-04-16 07:42:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][950/1251] eta 0:01:27 lr 0.000398 time 0.2618 (0.2896) loss 3.7558 (3.4505) grad_norm 1.6368 (1.6892) [2021-04-16 07:42:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][960/1251] eta 0:01:24 lr 0.000398 time 0.2837 (0.2895) loss 3.8086 (3.4515) grad_norm 1.7136 (1.6883) [2021-04-16 07:42:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][970/1251] eta 0:01:21 lr 0.000398 time 0.2982 (0.2896) loss 4.1411 (3.4505) grad_norm 1.5377 (1.6869) [2021-04-16 07:42:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][980/1251] eta 0:01:18 lr 0.000398 time 0.2890 (0.2895) loss 2.6525 (3.4508) grad_norm 1.5849 (1.6868) [2021-04-16 07:42:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][990/1251] eta 0:01:15 lr 0.000398 time 0.2794 (0.2894) loss 3.7914 (3.4506) grad_norm 1.6190 (1.6867) [2021-04-16 07:42:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][1000/1251] eta 0:01:12 lr 0.000398 time 0.2470 (0.2893) loss 3.9331 (3.4523) grad_norm 1.7691 (1.6871) [2021-04-16 07:42:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][1010/1251] eta 0:01:09 lr 0.000398 time 0.3137 (0.2892) loss 2.9150 (3.4508) grad_norm 1.5741 (1.6869) [2021-04-16 07:42:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][1020/1251] eta 0:01:06 lr 0.000398 time 0.2730 (0.2890) loss 2.8263 (3.4502) grad_norm 1.5195 (1.6871) [2021-04-16 07:42:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][1030/1251] eta 0:01:03 lr 0.000398 time 0.2608 (0.2890) loss 3.8334 (3.4532) grad_norm 1.6789 (1.6871) [2021-04-16 07:42:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][1040/1251] eta 0:01:00 lr 0.000398 time 0.2852 (0.2890) loss 3.8324 (3.4555) grad_norm 1.8981 (1.6875) [2021-04-16 07:42:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][1050/1251] eta 0:00:58 lr 0.000398 time 0.2824 (0.2889) loss 4.0224 (3.4561) grad_norm 1.8144 (1.6874) [2021-04-16 07:43:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][1060/1251] eta 0:00:55 lr 0.000398 time 0.2753 (0.2887) loss 3.8103 (3.4546) grad_norm 1.4773 (1.6868) [2021-04-16 07:43:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][1070/1251] eta 0:00:52 lr 0.000398 time 0.2641 (0.2886) loss 3.4531 (3.4556) grad_norm 1.5726 (1.6867) [2021-04-16 07:43:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][1080/1251] eta 0:00:49 lr 0.000398 time 0.2795 (0.2886) loss 2.3621 (3.4535) grad_norm 1.8962 (1.6868) [2021-04-16 07:43:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][1090/1251] eta 0:00:46 lr 0.000398 time 0.2724 (0.2886) loss 2.5599 (3.4533) grad_norm 1.3527 (1.6862) [2021-04-16 07:43:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][1100/1251] eta 0:00:43 lr 0.000398 time 0.2637 (0.2884) loss 2.8001 (3.4522) grad_norm 1.6249 (1.6855) [2021-04-16 07:43:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][1110/1251] eta 0:00:40 lr 0.000398 time 0.2651 (0.2883) loss 2.7819 (3.4516) grad_norm 1.5083 (1.6863) [2021-04-16 07:43:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][1120/1251] eta 0:00:37 lr 0.000398 time 0.2723 (0.2884) loss 3.5802 (3.4510) grad_norm 1.6128 (1.6864) [2021-04-16 07:43:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][1130/1251] eta 0:00:34 lr 0.000398 time 0.2728 (0.2883) loss 3.4458 (3.4500) grad_norm 1.8972 (1.6861) [2021-04-16 07:43:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][1140/1251] eta 0:00:32 lr 0.000397 time 0.3029 (0.2883) loss 3.6338 (3.4505) grad_norm 1.5942 (1.6865) [2021-04-16 07:43:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][1150/1251] eta 0:00:29 lr 0.000397 time 0.2744 (0.2882) loss 3.9297 (3.4492) grad_norm 1.4962 (1.6871) [2021-04-16 07:43:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][1160/1251] eta 0:00:26 lr 0.000397 time 0.2722 (0.2882) loss 3.7711 (3.4502) grad_norm 1.6646 (1.6881) [2021-04-16 07:43:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][1170/1251] eta 0:00:23 lr 0.000397 time 0.2783 (0.2882) loss 3.7427 (3.4495) grad_norm 1.3885 (1.6873) [2021-04-16 07:43:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][1180/1251] eta 0:00:20 lr 0.000397 time 0.2830 (0.2882) loss 2.6128 (3.4494) grad_norm 1.5992 (1.6865) [2021-04-16 07:43:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][1190/1251] eta 0:00:17 lr 0.000397 time 0.2902 (0.2881) loss 3.6256 (3.4519) grad_norm 1.7102 (1.6863) [2021-04-16 07:43:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][1200/1251] eta 0:00:14 lr 0.000397 time 0.3132 (0.2882) loss 2.8408 (3.4515) grad_norm 1.6975 (1.6868) [2021-04-16 07:43:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][1210/1251] eta 0:00:11 lr 0.000397 time 0.3004 (0.2880) loss 3.4053 (3.4494) grad_norm 1.4537 (1.6861) [2021-04-16 07:43:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][1220/1251] eta 0:00:08 lr 0.000397 time 0.2732 (0.2880) loss 2.4552 (3.4462) grad_norm 1.7095 (1.6865) [2021-04-16 07:43:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][1230/1251] eta 0:00:06 lr 0.000397 time 0.2775 (0.2880) loss 3.1934 (3.4444) grad_norm 1.5824 (1.6867) [2021-04-16 07:43:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][1240/1251] eta 0:00:03 lr 0.000397 time 0.2619 (0.2879) loss 3.7787 (3.4446) grad_norm 1.6805 (1.6864) [2021-04-16 07:43:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [170/300][1250/1251] eta 0:00:00 lr 0.000397 time 0.2489 (0.2876) loss 3.7330 (3.4461) grad_norm 1.6665 (1.6858) [2021-04-16 07:44:03 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 170 training takes 0:06:07 [2021-04-16 07:44:03 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_170.pth saving...... [2021-04-16 07:44:13 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_170.pth saved !!! [2021-04-16 07:44:14 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.115 (1.115) Loss 0.9255 (0.9255) Acc@1 78.418 (78.418) Acc@5 94.043 (94.043) [2021-04-16 07:44:16 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.093 (0.225) Loss 0.9627 (0.9654) Acc@1 78.320 (77.459) Acc@5 93.750 (94.078) [2021-04-16 07:44:18 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.313 (0.228) Loss 0.9841 (0.9709) Acc@1 77.930 (77.297) Acc@5 92.676 (93.890) [2021-04-16 07:44:21 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.254 (0.238) Loss 1.0298 (0.9771) Acc@1 75.684 (77.161) Acc@5 92.578 (93.734) [2021-04-16 07:44:22 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.212) Loss 0.9289 (0.9765) Acc@1 77.246 (77.106) Acc@5 93.555 (93.762) [2021-04-16 07:44:34 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 77.072 Acc@5 93.782 [2021-04-16 07:44:34 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 77.1% [2021-04-16 07:44:34 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 77.08% [2021-04-16 07:44:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][0/1251] eta 0:59:43 lr 0.000397 time 2.8648 (2.8648) loss 3.9831 (3.9831) grad_norm 1.5714 (1.5714) [2021-04-16 07:44:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][10/1251] eta 0:10:47 lr 0.000397 time 0.4170 (0.5221) loss 3.0172 (3.5247) grad_norm 1.5168 (1.6641) [2021-04-16 07:44:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][20/1251] eta 0:08:20 lr 0.000397 time 0.2634 (0.4067) loss 3.6408 (3.5294) grad_norm 1.5582 (1.6258) [2021-04-16 07:44:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][30/1251] eta 0:07:27 lr 0.000397 time 0.2734 (0.3665) loss 2.6286 (3.5441) grad_norm 1.5302 (1.6298) [2021-04-16 07:44:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][40/1251] eta 0:06:58 lr 0.000397 time 0.2801 (0.3456) loss 3.5419 (3.5425) grad_norm 1.9910 (1.6263) [2021-04-16 07:44:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][50/1251] eta 0:06:41 lr 0.000397 time 0.2734 (0.3346) loss 4.0280 (3.5372) grad_norm 1.6328 (1.6363) [2021-04-16 07:44:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][60/1251] eta 0:06:27 lr 0.000397 time 0.2851 (0.3251) loss 3.7052 (3.4973) grad_norm 1.7123 (1.6447) [2021-04-16 07:44:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][70/1251] eta 0:06:16 lr 0.000397 time 0.3065 (0.3190) loss 3.2733 (3.4744) grad_norm 1.5652 (1.6305) [2021-04-16 07:45:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][80/1251] eta 0:06:07 lr 0.000397 time 0.2831 (0.3142) loss 2.3462 (3.4286) grad_norm 1.8756 (1.6373) [2021-04-16 07:45:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][90/1251] eta 0:05:59 lr 0.000397 time 0.2989 (0.3099) loss 3.9425 (3.4625) grad_norm 2.0259 (1.6513) [2021-04-16 07:45:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][100/1251] eta 0:05:53 lr 0.000397 time 0.2970 (0.3072) loss 3.8549 (3.4716) grad_norm 1.6131 (1.6627) [2021-04-16 07:45:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][110/1251] eta 0:05:47 lr 0.000397 time 0.2699 (0.3048) loss 2.7319 (3.4978) grad_norm 1.6837 (1.6586) [2021-04-16 07:45:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][120/1251] eta 0:05:44 lr 0.000397 time 0.2701 (0.3045) loss 3.7440 (3.4908) grad_norm 1.5969 (1.6529) [2021-04-16 07:45:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][130/1251] eta 0:05:38 lr 0.000396 time 0.2534 (0.3023) loss 3.2708 (3.4717) grad_norm 1.6181 (1.6520) [2021-04-16 07:45:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][140/1251] eta 0:05:34 lr 0.000396 time 0.2660 (0.3012) loss 2.2933 (3.4737) grad_norm 1.7365 (1.6603) [2021-04-16 07:45:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][150/1251] eta 0:05:31 lr 0.000396 time 0.2491 (0.3006) loss 2.9817 (3.4755) grad_norm 1.4861 (1.6606) [2021-04-16 07:45:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][160/1251] eta 0:05:26 lr 0.000396 time 0.2802 (0.2991) loss 3.3026 (3.4724) grad_norm 2.1357 (1.6692) [2021-04-16 07:45:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][170/1251] eta 0:05:21 lr 0.000396 time 0.3020 (0.2977) loss 4.0865 (3.4604) grad_norm 1.5511 (1.6677) [2021-04-16 07:45:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][180/1251] eta 0:05:18 lr 0.000396 time 0.2908 (0.2978) loss 4.0141 (3.4728) grad_norm 1.5662 (1.6636) [2021-04-16 07:45:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][190/1251] eta 0:05:15 lr 0.000396 time 0.2770 (0.2969) loss 2.8559 (3.4608) grad_norm 2.0619 (1.6683) [2021-04-16 07:45:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][200/1251] eta 0:05:10 lr 0.000396 time 0.2671 (0.2957) loss 3.7044 (3.4561) grad_norm 1.5537 (1.6678) [2021-04-16 07:45:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][210/1251] eta 0:05:06 lr 0.000396 time 0.2660 (0.2949) loss 3.4402 (3.4588) grad_norm 1.5877 (nan) [2021-04-16 07:45:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][220/1251] eta 0:05:03 lr 0.000396 time 0.2820 (0.2941) loss 3.5949 (3.4594) grad_norm 1.5813 (nan) [2021-04-16 07:45:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][230/1251] eta 0:04:59 lr 0.000396 time 0.2880 (0.2938) loss 2.4856 (3.4461) grad_norm 1.4579 (nan) [2021-04-16 07:45:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][240/1251] eta 0:04:56 lr 0.000396 time 0.2653 (0.2929) loss 3.6353 (3.4444) grad_norm 1.7057 (nan) [2021-04-16 07:45:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][250/1251] eta 0:04:52 lr 0.000396 time 0.2513 (0.2925) loss 3.7288 (3.4354) grad_norm 1.5210 (nan) [2021-04-16 07:45:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][260/1251] eta 0:04:49 lr 0.000396 time 0.2613 (0.2918) loss 3.8294 (3.4340) grad_norm 1.8882 (nan) [2021-04-16 07:45:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][270/1251] eta 0:04:45 lr 0.000396 time 0.2648 (0.2912) loss 3.1029 (3.4408) grad_norm 1.9175 (nan) [2021-04-16 07:45:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][280/1251] eta 0:04:42 lr 0.000396 time 0.2763 (0.2907) loss 4.1316 (3.4393) grad_norm 1.8378 (nan) [2021-04-16 07:45:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][290/1251] eta 0:04:38 lr 0.000396 time 0.2773 (0.2903) loss 3.2408 (3.4337) grad_norm 1.6839 (nan) [2021-04-16 07:46:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][300/1251] eta 0:04:36 lr 0.000396 time 0.3029 (0.2903) loss 3.5421 (3.4409) grad_norm 1.8306 (nan) [2021-04-16 07:46:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][310/1251] eta 0:04:32 lr 0.000396 time 0.2741 (0.2900) loss 3.7608 (3.4406) grad_norm 1.4979 (nan) [2021-04-16 07:46:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][320/1251] eta 0:04:30 lr 0.000396 time 0.2845 (0.2900) loss 4.1311 (3.4491) grad_norm 1.5705 (nan) [2021-04-16 07:46:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][330/1251] eta 0:04:26 lr 0.000396 time 0.2585 (0.2897) loss 3.9672 (3.4509) grad_norm 1.7350 (nan) [2021-04-16 07:46:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][340/1251] eta 0:04:23 lr 0.000396 time 0.2745 (0.2894) loss 4.1204 (3.4480) grad_norm 1.9967 (nan) [2021-04-16 07:46:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][350/1251] eta 0:04:20 lr 0.000396 time 0.2837 (0.2891) loss 3.5062 (3.4364) grad_norm 1.7670 (nan) [2021-04-16 07:46:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][360/1251] eta 0:04:17 lr 0.000396 time 0.2663 (0.2889) loss 3.6414 (3.4340) grad_norm 1.7768 (nan) [2021-04-16 07:46:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][370/1251] eta 0:04:14 lr 0.000396 time 0.2714 (0.2892) loss 3.4775 (3.4316) grad_norm 1.5442 (nan) [2021-04-16 07:46:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][380/1251] eta 0:04:11 lr 0.000395 time 0.2981 (0.2891) loss 3.2449 (3.4288) grad_norm 1.7360 (nan) [2021-04-16 07:46:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][390/1251] eta 0:04:08 lr 0.000395 time 0.2767 (0.2888) loss 2.5709 (3.4216) grad_norm 1.4697 (nan) [2021-04-16 07:46:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][400/1251] eta 0:04:05 lr 0.000395 time 0.2762 (0.2887) loss 3.9857 (3.4250) grad_norm 1.6668 (nan) [2021-04-16 07:46:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][410/1251] eta 0:04:02 lr 0.000395 time 0.2754 (0.2884) loss 3.6783 (3.4307) grad_norm 1.5817 (nan) [2021-04-16 07:46:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][420/1251] eta 0:03:59 lr 0.000395 time 0.2753 (0.2880) loss 3.0940 (3.4358) grad_norm 1.7040 (nan) [2021-04-16 07:46:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][430/1251] eta 0:03:56 lr 0.000395 time 0.2638 (0.2882) loss 3.6745 (3.4416) grad_norm 1.7356 (nan) [2021-04-16 07:46:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][440/1251] eta 0:03:53 lr 0.000395 time 0.2967 (0.2881) loss 3.6246 (3.4475) grad_norm 1.6589 (nan) [2021-04-16 07:46:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][450/1251] eta 0:03:50 lr 0.000395 time 0.2913 (0.2879) loss 3.6003 (3.4534) grad_norm 1.5519 (nan) [2021-04-16 07:46:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][460/1251] eta 0:03:47 lr 0.000395 time 0.2946 (0.2878) loss 2.4033 (3.4551) grad_norm 1.5232 (nan) [2021-04-16 07:46:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][470/1251] eta 0:03:44 lr 0.000395 time 0.2666 (0.2875) loss 3.7973 (3.4503) grad_norm 2.1381 (nan) [2021-04-16 07:46:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][480/1251] eta 0:03:41 lr 0.000395 time 0.3046 (0.2873) loss 4.2574 (3.4502) grad_norm 1.8996 (nan) [2021-04-16 07:46:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][490/1251] eta 0:03:38 lr 0.000395 time 0.2885 (0.2871) loss 2.5560 (3.4487) grad_norm 1.5227 (nan) [2021-04-16 07:46:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][500/1251] eta 0:03:35 lr 0.000395 time 0.2775 (0.2870) loss 3.8202 (3.4443) grad_norm 1.6907 (nan) [2021-04-16 07:47:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][510/1251] eta 0:03:32 lr 0.000395 time 0.2906 (0.2870) loss 3.4173 (3.4433) grad_norm 1.7962 (nan) [2021-04-16 07:47:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][520/1251] eta 0:03:29 lr 0.000395 time 0.2853 (0.2869) loss 3.6606 (3.4458) grad_norm 1.6559 (nan) [2021-04-16 07:47:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][530/1251] eta 0:03:26 lr 0.000395 time 0.2598 (0.2867) loss 3.2730 (3.4429) grad_norm 1.6877 (nan) [2021-04-16 07:47:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][540/1251] eta 0:03:23 lr 0.000395 time 0.2660 (0.2866) loss 3.0851 (3.4388) grad_norm 1.4903 (nan) [2021-04-16 07:47:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][550/1251] eta 0:03:20 lr 0.000395 time 0.2631 (0.2865) loss 3.8043 (3.4420) grad_norm 1.7476 (nan) [2021-04-16 07:47:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][560/1251] eta 0:03:17 lr 0.000395 time 0.2855 (0.2864) loss 2.1755 (3.4390) grad_norm 1.9305 (nan) [2021-04-16 07:47:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][570/1251] eta 0:03:15 lr 0.000395 time 0.2556 (0.2866) loss 3.5122 (3.4372) grad_norm 1.7084 (nan) [2021-04-16 07:47:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][580/1251] eta 0:03:12 lr 0.000395 time 0.2807 (0.2864) loss 4.0862 (3.4409) grad_norm 1.8848 (nan) [2021-04-16 07:47:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][590/1251] eta 0:03:09 lr 0.000395 time 0.2643 (0.2865) loss 2.5386 (3.4409) grad_norm 1.5132 (nan) [2021-04-16 07:47:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][600/1251] eta 0:03:06 lr 0.000395 time 0.2687 (0.2863) loss 3.9227 (3.4422) grad_norm 1.5359 (nan) [2021-04-16 07:47:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][610/1251] eta 0:03:03 lr 0.000395 time 0.2856 (0.2861) loss 3.8338 (3.4409) grad_norm 1.6315 (nan) [2021-04-16 07:47:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][620/1251] eta 0:03:00 lr 0.000395 time 0.2920 (0.2860) loss 3.4692 (3.4412) grad_norm 1.9394 (nan) [2021-04-16 07:47:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][630/1251] eta 0:02:57 lr 0.000394 time 0.2752 (0.2860) loss 3.9129 (3.4414) grad_norm 2.1157 (nan) [2021-04-16 07:47:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][640/1251] eta 0:02:54 lr 0.000394 time 0.2698 (0.2858) loss 3.4437 (3.4372) grad_norm 1.6649 (nan) [2021-04-16 07:47:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][650/1251] eta 0:02:51 lr 0.000394 time 0.3016 (0.2858) loss 3.9175 (3.4415) grad_norm 1.5535 (nan) [2021-04-16 07:47:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][660/1251] eta 0:02:48 lr 0.000394 time 0.2762 (0.2856) loss 3.9128 (3.4397) grad_norm 1.5614 (nan) [2021-04-16 07:47:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][670/1251] eta 0:02:45 lr 0.000394 time 0.2765 (0.2855) loss 3.7006 (3.4395) grad_norm 1.5384 (nan) [2021-04-16 07:47:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][680/1251] eta 0:02:42 lr 0.000394 time 0.3008 (0.2854) loss 2.6160 (3.4382) grad_norm 1.5769 (nan) [2021-04-16 07:47:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][690/1251] eta 0:02:40 lr 0.000394 time 0.2828 (0.2854) loss 4.0425 (3.4431) grad_norm 1.7333 (nan) [2021-04-16 07:47:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][700/1251] eta 0:02:37 lr 0.000394 time 0.3010 (0.2852) loss 3.6276 (3.4443) grad_norm 1.7292 (nan) [2021-04-16 07:47:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][710/1251] eta 0:02:34 lr 0.000394 time 0.2622 (0.2850) loss 3.6741 (3.4459) grad_norm 1.9975 (nan) [2021-04-16 07:48:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][720/1251] eta 0:02:31 lr 0.000394 time 0.3143 (0.2854) loss 3.1246 (3.4449) grad_norm 1.5992 (nan) [2021-04-16 07:48:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][730/1251] eta 0:02:28 lr 0.000394 time 0.2740 (0.2853) loss 4.3189 (3.4464) grad_norm 1.4951 (nan) [2021-04-16 07:48:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][740/1251] eta 0:02:25 lr 0.000394 time 0.2805 (0.2851) loss 3.8217 (3.4481) grad_norm 1.7069 (nan) [2021-04-16 07:48:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][750/1251] eta 0:02:22 lr 0.000394 time 0.3255 (0.2850) loss 2.6632 (3.4492) grad_norm 1.6908 (nan) [2021-04-16 07:48:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][760/1251] eta 0:02:19 lr 0.000394 time 0.2531 (0.2851) loss 3.0760 (3.4494) grad_norm 1.5767 (nan) [2021-04-16 07:48:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][770/1251] eta 0:02:17 lr 0.000394 time 0.2755 (0.2850) loss 2.5599 (3.4444) grad_norm 1.6243 (nan) [2021-04-16 07:48:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][780/1251] eta 0:02:14 lr 0.000394 time 0.2750 (0.2849) loss 4.0832 (3.4440) grad_norm 1.5604 (nan) [2021-04-16 07:48:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][790/1251] eta 0:02:11 lr 0.000394 time 0.2771 (0.2848) loss 3.5888 (3.4427) grad_norm 1.3936 (nan) [2021-04-16 07:48:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][800/1251] eta 0:02:08 lr 0.000394 time 0.2842 (0.2847) loss 3.2817 (3.4430) grad_norm 1.7091 (nan) [2021-04-16 07:48:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][810/1251] eta 0:02:05 lr 0.000394 time 0.2973 (0.2847) loss 2.6448 (3.4440) grad_norm 1.6595 (nan) [2021-04-16 07:48:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][820/1251] eta 0:02:02 lr 0.000394 time 0.2593 (0.2846) loss 3.7028 (3.4433) grad_norm 2.0606 (nan) [2021-04-16 07:48:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][830/1251] eta 0:01:59 lr 0.000394 time 0.2796 (0.2845) loss 3.5680 (3.4440) grad_norm 1.7693 (nan) [2021-04-16 07:48:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][840/1251] eta 0:01:56 lr 0.000394 time 0.2899 (0.2845) loss 3.7185 (3.4477) grad_norm 2.2445 (nan) [2021-04-16 07:48:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][850/1251] eta 0:01:54 lr 0.000394 time 0.2896 (0.2844) loss 3.9373 (3.4505) grad_norm 1.6984 (nan) [2021-04-16 07:48:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][860/1251] eta 0:01:51 lr 0.000394 time 0.2575 (0.2843) loss 3.2183 (3.4517) grad_norm 1.6707 (nan) [2021-04-16 07:48:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][870/1251] eta 0:01:48 lr 0.000394 time 0.2743 (0.2842) loss 2.7641 (3.4504) grad_norm 1.6723 (nan) [2021-04-16 07:48:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][880/1251] eta 0:01:45 lr 0.000393 time 0.2586 (0.2842) loss 3.9092 (3.4501) grad_norm 1.6700 (nan) [2021-04-16 07:48:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][890/1251] eta 0:01:42 lr 0.000393 time 0.2675 (0.2843) loss 3.5723 (3.4497) grad_norm 1.5527 (nan) [2021-04-16 07:48:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][900/1251] eta 0:01:39 lr 0.000393 time 0.2817 (0.2842) loss 3.8678 (3.4504) grad_norm 1.8315 (nan) [2021-04-16 07:48:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][910/1251] eta 0:01:36 lr 0.000393 time 0.3003 (0.2841) loss 3.9695 (3.4529) grad_norm 1.6427 (nan) [2021-04-16 07:48:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][920/1251] eta 0:01:34 lr 0.000393 time 0.2762 (0.2841) loss 3.4170 (3.4555) grad_norm 1.5874 (nan) [2021-04-16 07:48:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][930/1251] eta 0:01:31 lr 0.000393 time 0.2791 (0.2841) loss 3.7759 (3.4580) grad_norm 1.3822 (nan) [2021-04-16 07:49:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][940/1251] eta 0:01:28 lr 0.000393 time 0.2738 (0.2842) loss 2.3894 (3.4571) grad_norm 1.9111 (nan) [2021-04-16 07:49:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][950/1251] eta 0:01:25 lr 0.000393 time 0.3034 (0.2841) loss 3.0538 (3.4561) grad_norm 1.8914 (nan) [2021-04-16 07:49:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][960/1251] eta 0:01:22 lr 0.000393 time 0.3004 (0.2840) loss 2.9097 (3.4556) grad_norm 1.5208 (nan) [2021-04-16 07:49:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][970/1251] eta 0:01:19 lr 0.000393 time 0.2716 (0.2839) loss 3.9452 (3.4541) grad_norm 1.6992 (nan) [2021-04-16 07:49:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][980/1251] eta 0:01:16 lr 0.000393 time 0.2550 (0.2839) loss 3.6754 (3.4523) grad_norm 1.7367 (nan) [2021-04-16 07:49:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][990/1251] eta 0:01:14 lr 0.000393 time 0.2910 (0.2839) loss 3.3939 (3.4531) grad_norm 1.8374 (nan) [2021-04-16 07:49:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][1000/1251] eta 0:01:11 lr 0.000393 time 0.2877 (0.2839) loss 4.1114 (3.4543) grad_norm 1.5448 (nan) [2021-04-16 07:49:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][1010/1251] eta 0:01:08 lr 0.000393 time 0.2647 (0.2839) loss 2.5400 (3.4534) grad_norm 1.6552 (nan) [2021-04-16 07:49:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][1020/1251] eta 0:01:05 lr 0.000393 time 0.2831 (0.2839) loss 3.1848 (3.4539) grad_norm 1.6139 (nan) [2021-04-16 07:49:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][1030/1251] eta 0:01:02 lr 0.000393 time 0.3005 (0.2838) loss 4.0616 (3.4535) grad_norm 1.8058 (nan) [2021-04-16 07:49:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][1040/1251] eta 0:00:59 lr 0.000393 time 0.2596 (0.2838) loss 3.3613 (3.4551) grad_norm 1.6319 (nan) [2021-04-16 07:49:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][1050/1251] eta 0:00:57 lr 0.000393 time 0.2947 (0.2838) loss 3.5238 (3.4551) grad_norm 1.5016 (nan) [2021-04-16 07:49:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][1060/1251] eta 0:00:54 lr 0.000393 time 0.2905 (0.2839) loss 2.5722 (3.4561) grad_norm 1.6273 (nan) [2021-04-16 07:49:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][1070/1251] eta 0:00:51 lr 0.000393 time 0.2796 (0.2838) loss 4.0932 (3.4586) grad_norm 1.6053 (nan) [2021-04-16 07:49:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][1080/1251] eta 0:00:48 lr 0.000393 time 0.2755 (0.2837) loss 4.0610 (3.4576) grad_norm 1.6777 (nan) [2021-04-16 07:49:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][1090/1251] eta 0:00:45 lr 0.000393 time 0.3164 (0.2837) loss 2.6758 (3.4574) grad_norm 2.1852 (nan) [2021-04-16 07:49:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][1100/1251] eta 0:00:42 lr 0.000393 time 0.2667 (0.2837) loss 3.6469 (3.4580) grad_norm 1.5873 (nan) [2021-04-16 07:49:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][1110/1251] eta 0:00:39 lr 0.000393 time 0.2729 (0.2836) loss 3.6646 (3.4580) grad_norm 1.8344 (nan) [2021-04-16 07:49:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][1120/1251] eta 0:00:37 lr 0.000392 time 0.2751 (0.2836) loss 2.7604 (3.4538) grad_norm 1.9995 (nan) [2021-04-16 07:49:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][1130/1251] eta 0:00:34 lr 0.000392 time 0.2681 (0.2836) loss 4.1734 (3.4520) grad_norm 1.5581 (nan) [2021-04-16 07:49:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][1140/1251] eta 0:00:31 lr 0.000392 time 0.2652 (0.2835) loss 3.7656 (3.4526) grad_norm 1.5338 (nan) [2021-04-16 07:50:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][1150/1251] eta 0:00:28 lr 0.000392 time 0.2783 (0.2836) loss 2.3787 (3.4546) grad_norm 1.7604 (nan) [2021-04-16 07:50:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][1160/1251] eta 0:00:25 lr 0.000392 time 0.2685 (0.2836) loss 4.1941 (3.4548) grad_norm 1.5560 (nan) [2021-04-16 07:50:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][1170/1251] eta 0:00:22 lr 0.000392 time 0.2760 (0.2836) loss 3.4578 (3.4553) grad_norm 1.5885 (nan) [2021-04-16 07:50:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][1180/1251] eta 0:00:20 lr 0.000392 time 0.2742 (0.2836) loss 3.0733 (3.4553) grad_norm 1.6874 (nan) [2021-04-16 07:50:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][1190/1251] eta 0:00:17 lr 0.000392 time 0.2770 (0.2836) loss 3.9867 (3.4550) grad_norm 1.7450 (nan) [2021-04-16 07:50:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][1200/1251] eta 0:00:14 lr 0.000392 time 0.2826 (0.2835) loss 2.5581 (3.4556) grad_norm 1.8036 (nan) [2021-04-16 07:50:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][1210/1251] eta 0:00:11 lr 0.000392 time 0.2813 (0.2834) loss 3.7900 (3.4565) grad_norm 1.5741 (nan) [2021-04-16 07:50:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][1220/1251] eta 0:00:08 lr 0.000392 time 0.2820 (0.2834) loss 4.0189 (3.4586) grad_norm 1.6430 (nan) [2021-04-16 07:50:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][1230/1251] eta 0:00:05 lr 0.000392 time 0.3068 (0.2834) loss 4.1883 (3.4593) grad_norm 1.7669 (nan) [2021-04-16 07:50:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][1240/1251] eta 0:00:03 lr 0.000392 time 0.2521 (0.2832) loss 3.1977 (3.4596) grad_norm 1.5881 (nan) [2021-04-16 07:50:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [171/300][1250/1251] eta 0:00:00 lr 0.000392 time 0.2488 (0.2830) loss 3.0524 (3.4592) grad_norm 1.5513 (nan) [2021-04-16 07:50:32 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 171 training takes 0:05:58 [2021-04-16 07:50:32 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_171.pth saving...... [2021-04-16 07:50:42 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_171.pth saved !!! [2021-04-16 07:50:43 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.155 (1.155) Loss 1.0214 (1.0214) Acc@1 74.805 (74.805) Acc@5 93.555 (93.555) [2021-04-16 07:50:44 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.338 (0.229) Loss 0.9576 (0.9669) Acc@1 78.711 (77.086) Acc@5 93.262 (93.750) [2021-04-16 07:50:46 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.126 (0.224) Loss 0.9230 (0.9544) Acc@1 78.027 (77.386) Acc@5 94.043 (93.848) [2021-04-16 07:50:49 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.101 (0.237) Loss 0.9155 (0.9554) Acc@1 78.613 (77.416) Acc@5 94.531 (93.832) [2021-04-16 07:50:51 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.216) Loss 0.9416 (0.9518) Acc@1 76.465 (77.480) Acc@5 95.020 (93.914) [2021-04-16 07:50:56 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 77.334 Acc@5 93.822 [2021-04-16 07:50:56 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 77.3% [2021-04-16 07:50:56 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 77.33% [2021-04-16 07:51:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][0/1251] eta 1:21:14 lr 0.000392 time 3.8964 (3.8964) loss 2.8079 (2.8079) grad_norm 1.6336 (1.6336) [2021-04-16 07:51:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][10/1251] eta 0:12:35 lr 0.000392 time 0.2583 (0.6090) loss 3.5778 (3.3170) grad_norm 1.9866 (1.7118) [2021-04-16 07:51:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][20/1251] eta 0:09:17 lr 0.000392 time 0.2717 (0.4526) loss 3.8360 (3.4811) grad_norm 1.6736 (1.7283) [2021-04-16 07:51:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][30/1251] eta 0:08:02 lr 0.000392 time 0.2854 (0.3952) loss 2.5717 (3.4546) grad_norm 2.2729 (1.7310) [2021-04-16 07:51:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][40/1251] eta 0:07:25 lr 0.000392 time 0.2985 (0.3680) loss 3.5419 (3.4616) grad_norm 1.6051 (1.7247) [2021-04-16 07:51:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][50/1251] eta 0:07:01 lr 0.000392 time 0.2732 (0.3507) loss 3.7657 (3.4855) grad_norm 1.9338 (1.7332) [2021-04-16 07:51:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][60/1251] eta 0:06:44 lr 0.000392 time 0.2767 (0.3393) loss 3.5897 (3.4986) grad_norm 1.7898 (1.7233) [2021-04-16 07:51:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][70/1251] eta 0:06:31 lr 0.000392 time 0.2782 (0.3318) loss 2.3470 (3.4730) grad_norm 1.4248 (1.7101) [2021-04-16 07:51:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][80/1251] eta 0:06:22 lr 0.000392 time 0.2938 (0.3265) loss 3.9647 (3.4797) grad_norm 1.4925 (1.6963) [2021-04-16 07:51:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][90/1251] eta 0:06:14 lr 0.000392 time 0.2893 (0.3223) loss 3.7043 (3.4864) grad_norm 1.4996 (1.7030) [2021-04-16 07:51:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][100/1251] eta 0:06:07 lr 0.000392 time 0.2637 (0.3195) loss 3.5672 (3.4513) grad_norm 1.9402 (1.7185) [2021-04-16 07:51:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][110/1251] eta 0:06:01 lr 0.000392 time 0.2864 (0.3164) loss 2.4813 (3.4433) grad_norm 1.6993 (1.7192) [2021-04-16 07:51:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][120/1251] eta 0:05:55 lr 0.000391 time 0.2688 (0.3144) loss 2.4728 (3.4348) grad_norm 1.5502 (1.7395) [2021-04-16 07:51:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][130/1251] eta 0:05:49 lr 0.000391 time 0.2796 (0.3122) loss 3.9733 (3.4300) grad_norm 1.6806 (1.7376) [2021-04-16 07:51:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][140/1251] eta 0:05:46 lr 0.000391 time 0.2665 (0.3121) loss 3.3129 (3.4193) grad_norm 1.9547 (1.7378) [2021-04-16 07:51:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][150/1251] eta 0:05:41 lr 0.000391 time 0.2709 (0.3102) loss 2.7470 (3.4168) grad_norm 2.0086 (1.7316) [2021-04-16 07:51:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][160/1251] eta 0:05:36 lr 0.000391 time 0.2739 (0.3084) loss 2.8078 (3.4192) grad_norm 1.6743 (1.7346) [2021-04-16 07:51:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][170/1251] eta 0:05:31 lr 0.000391 time 0.2831 (0.3068) loss 3.6106 (3.4184) grad_norm 1.5402 (1.7342) [2021-04-16 07:51:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][180/1251] eta 0:05:26 lr 0.000391 time 0.2771 (0.3052) loss 3.5733 (3.4151) grad_norm 1.4068 (1.7295) [2021-04-16 07:51:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][190/1251] eta 0:05:21 lr 0.000391 time 0.2657 (0.3035) loss 2.8607 (3.4104) grad_norm 1.6834 (1.7235) [2021-04-16 07:51:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][200/1251] eta 0:05:17 lr 0.000391 time 0.2726 (0.3022) loss 3.8791 (3.4282) grad_norm 1.8084 (1.7219) [2021-04-16 07:51:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][210/1251] eta 0:05:13 lr 0.000391 time 0.2881 (0.3012) loss 3.6033 (3.4381) grad_norm 1.5980 (1.7226) [2021-04-16 07:52:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][220/1251] eta 0:05:10 lr 0.000391 time 0.2684 (0.3008) loss 3.9086 (3.4516) grad_norm 2.0019 (1.7269) [2021-04-16 07:52:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][230/1251] eta 0:05:06 lr 0.000391 time 0.2895 (0.2999) loss 4.2347 (3.4480) grad_norm 1.6158 (1.7243) [2021-04-16 07:52:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][240/1251] eta 0:05:02 lr 0.000391 time 0.3037 (0.2992) loss 3.6828 (3.4522) grad_norm 1.6176 (1.7197) [2021-04-16 07:52:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][250/1251] eta 0:04:58 lr 0.000391 time 0.3024 (0.2985) loss 3.8220 (3.4597) grad_norm 1.5017 (1.7145) [2021-04-16 07:52:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][260/1251] eta 0:04:54 lr 0.000391 time 0.2963 (0.2976) loss 3.7492 (3.4574) grad_norm 1.8538 (1.7134) [2021-04-16 07:52:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][270/1251] eta 0:04:51 lr 0.000391 time 0.2650 (0.2969) loss 2.5408 (3.4526) grad_norm 1.9530 (1.7130) [2021-04-16 07:52:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][280/1251] eta 0:04:47 lr 0.000391 time 0.3022 (0.2961) loss 3.7446 (3.4552) grad_norm 1.4703 (1.7114) [2021-04-16 07:52:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][290/1251] eta 0:04:43 lr 0.000391 time 0.2549 (0.2955) loss 4.1995 (3.4586) grad_norm 1.6035 (1.7113) [2021-04-16 07:52:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][300/1251] eta 0:04:40 lr 0.000391 time 0.2541 (0.2949) loss 2.3472 (3.4483) grad_norm 1.9011 (1.7112) [2021-04-16 07:52:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][310/1251] eta 0:04:37 lr 0.000391 time 0.3119 (0.2949) loss 2.9503 (3.4472) grad_norm 1.5378 (1.7130) [2021-04-16 07:52:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][320/1251] eta 0:04:34 lr 0.000391 time 0.2964 (0.2946) loss 4.1103 (3.4486) grad_norm 1.7389 (1.7112) [2021-04-16 07:52:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][330/1251] eta 0:04:30 lr 0.000391 time 0.2819 (0.2941) loss 3.3982 (3.4442) grad_norm 1.7686 (1.7124) [2021-04-16 07:52:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][340/1251] eta 0:04:27 lr 0.000391 time 0.2947 (0.2936) loss 2.5693 (3.4446) grad_norm 1.6106 (1.7093) [2021-04-16 07:52:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][350/1251] eta 0:04:24 lr 0.000391 time 0.3009 (0.2934) loss 3.9554 (3.4479) grad_norm 1.5870 (1.7056) [2021-04-16 07:52:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][360/1251] eta 0:04:21 lr 0.000391 time 0.2851 (0.2936) loss 2.8530 (3.4486) grad_norm 1.6179 (1.7061) [2021-04-16 07:52:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][370/1251] eta 0:04:18 lr 0.000390 time 0.2480 (0.2938) loss 2.9424 (3.4373) grad_norm 1.5016 (1.7025) [2021-04-16 07:52:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][380/1251] eta 0:04:15 lr 0.000390 time 0.2730 (0.2933) loss 3.4814 (3.4379) grad_norm 1.5987 (1.7039) [2021-04-16 07:52:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][390/1251] eta 0:04:12 lr 0.000390 time 0.2811 (0.2931) loss 2.7572 (3.4419) grad_norm 1.7229 (1.7052) [2021-04-16 07:52:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][400/1251] eta 0:04:09 lr 0.000390 time 0.2769 (0.2928) loss 3.8764 (3.4453) grad_norm 1.8064 (1.7077) [2021-04-16 07:52:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][410/1251] eta 0:04:05 lr 0.000390 time 0.2651 (0.2925) loss 3.8753 (3.4422) grad_norm 1.5175 (1.7093) [2021-04-16 07:52:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][420/1251] eta 0:04:02 lr 0.000390 time 0.2986 (0.2923) loss 3.8081 (3.4462) grad_norm 1.9244 (1.7106) [2021-04-16 07:53:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][430/1251] eta 0:03:59 lr 0.000390 time 0.2911 (0.2921) loss 3.4604 (3.4432) grad_norm 1.9546 (1.7103) [2021-04-16 07:53:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][440/1251] eta 0:03:56 lr 0.000390 time 0.2827 (0.2918) loss 2.4521 (3.4379) grad_norm 1.6133 (1.7119) [2021-04-16 07:53:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][450/1251] eta 0:03:53 lr 0.000390 time 0.3026 (0.2916) loss 2.4160 (3.4336) grad_norm 1.7071 (1.7120) [2021-04-16 07:53:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][460/1251] eta 0:03:50 lr 0.000390 time 0.2637 (0.2915) loss 3.6095 (3.4389) grad_norm 1.4718 (1.7101) [2021-04-16 07:53:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][470/1251] eta 0:03:47 lr 0.000390 time 0.2678 (0.2913) loss 3.8569 (3.4412) grad_norm 1.5670 (1.7095) [2021-04-16 07:53:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][480/1251] eta 0:03:44 lr 0.000390 time 0.2587 (0.2911) loss 4.0910 (3.4399) grad_norm 1.6386 (1.7079) [2021-04-16 07:53:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][490/1251] eta 0:03:41 lr 0.000390 time 0.2855 (0.2913) loss 4.3077 (3.4386) grad_norm 1.5452 (1.7057) [2021-04-16 07:53:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][500/1251] eta 0:03:38 lr 0.000390 time 0.2972 (0.2911) loss 3.6376 (3.4401) grad_norm 1.7155 (1.7056) [2021-04-16 07:53:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][510/1251] eta 0:03:35 lr 0.000390 time 0.2923 (0.2909) loss 3.5735 (3.4384) grad_norm 1.7035 (1.7045) [2021-04-16 07:53:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][520/1251] eta 0:03:32 lr 0.000390 time 0.2672 (0.2906) loss 3.5344 (3.4410) grad_norm 2.0770 (1.7033) [2021-04-16 07:53:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][530/1251] eta 0:03:29 lr 0.000390 time 0.4247 (0.2907) loss 3.8321 (3.4359) grad_norm 1.7110 (1.7030) [2021-04-16 07:53:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][540/1251] eta 0:03:26 lr 0.000390 time 0.2665 (0.2909) loss 3.1722 (3.4354) grad_norm 1.5857 (1.7026) [2021-04-16 07:53:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][550/1251] eta 0:03:23 lr 0.000390 time 0.2832 (0.2907) loss 3.2603 (3.4362) grad_norm 1.8429 (1.7032) [2021-04-16 07:53:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][560/1251] eta 0:03:20 lr 0.000390 time 0.2803 (0.2905) loss 3.9691 (3.4390) grad_norm 1.5456 (1.7031) [2021-04-16 07:53:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][570/1251] eta 0:03:17 lr 0.000390 time 0.2838 (0.2903) loss 3.9659 (3.4415) grad_norm 1.6007 (1.7033) [2021-04-16 07:53:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][580/1251] eta 0:03:15 lr 0.000390 time 0.2763 (0.2908) loss 3.3964 (3.4399) grad_norm 1.6896 (1.7050) [2021-04-16 07:53:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][590/1251] eta 0:03:12 lr 0.000390 time 0.2877 (0.2906) loss 3.4288 (3.4411) grad_norm 1.5487 (1.7045) [2021-04-16 07:53:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][600/1251] eta 0:03:09 lr 0.000390 time 0.2580 (0.2904) loss 3.5966 (3.4417) grad_norm 1.6334 (1.7044) [2021-04-16 07:53:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][610/1251] eta 0:03:06 lr 0.000390 time 0.2873 (0.2902) loss 3.8937 (3.4432) grad_norm 2.0277 (1.7056) [2021-04-16 07:53:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][620/1251] eta 0:03:02 lr 0.000389 time 0.2559 (0.2900) loss 4.0018 (3.4432) grad_norm 1.5518 (1.7055) [2021-04-16 07:53:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][630/1251] eta 0:02:59 lr 0.000389 time 0.3024 (0.2898) loss 3.1575 (3.4460) grad_norm 1.5198 (1.7041) [2021-04-16 07:54:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][640/1251] eta 0:02:57 lr 0.000389 time 0.2779 (0.2898) loss 3.8176 (3.4430) grad_norm 1.9266 (1.7053) [2021-04-16 07:54:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][650/1251] eta 0:02:54 lr 0.000389 time 0.3842 (0.2898) loss 3.6780 (3.4419) grad_norm 1.6802 (1.7068) [2021-04-16 07:54:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][660/1251] eta 0:02:51 lr 0.000389 time 0.2764 (0.2896) loss 3.4041 (3.4413) grad_norm 1.7644 (1.7084) [2021-04-16 07:54:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][670/1251] eta 0:02:48 lr 0.000389 time 0.2865 (0.2894) loss 2.6303 (3.4397) grad_norm 1.5939 (1.7079) [2021-04-16 07:54:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][680/1251] eta 0:02:45 lr 0.000389 time 0.2767 (0.2894) loss 3.5371 (3.4420) grad_norm 1.6844 (1.7068) [2021-04-16 07:54:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][690/1251] eta 0:02:42 lr 0.000389 time 0.2577 (0.2893) loss 4.2982 (3.4463) grad_norm 1.8506 (1.7082) [2021-04-16 07:54:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][700/1251] eta 0:02:39 lr 0.000389 time 0.2610 (0.2891) loss 3.7249 (3.4466) grad_norm 1.8078 (1.7080) [2021-04-16 07:54:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][710/1251] eta 0:02:36 lr 0.000389 time 0.2758 (0.2890) loss 4.3114 (3.4403) grad_norm 1.8318 (1.7087) [2021-04-16 07:54:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][720/1251] eta 0:02:33 lr 0.000389 time 0.2908 (0.2891) loss 3.8754 (3.4425) grad_norm 1.5034 (1.7085) [2021-04-16 07:54:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][730/1251] eta 0:02:30 lr 0.000389 time 0.2731 (0.2889) loss 2.2283 (3.4406) grad_norm 1.6214 (1.7077) [2021-04-16 07:54:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][740/1251] eta 0:02:27 lr 0.000389 time 0.2618 (0.2888) loss 3.3676 (3.4416) grad_norm 1.9250 (1.7079) [2021-04-16 07:54:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][750/1251] eta 0:02:24 lr 0.000389 time 0.2574 (0.2886) loss 3.8066 (3.4429) grad_norm 1.5944 (1.7076) [2021-04-16 07:54:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][760/1251] eta 0:02:21 lr 0.000389 time 0.2899 (0.2887) loss 2.7824 (3.4429) grad_norm 1.7128 (1.7076) [2021-04-16 07:54:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][770/1251] eta 0:02:18 lr 0.000389 time 0.3010 (0.2886) loss 4.4080 (3.4445) grad_norm 1.7768 (1.7083) [2021-04-16 07:54:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][780/1251] eta 0:02:15 lr 0.000389 time 0.2580 (0.2884) loss 3.4314 (3.4467) grad_norm 1.7322 (1.7089) [2021-04-16 07:54:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][790/1251] eta 0:02:13 lr 0.000389 time 0.2807 (0.2885) loss 2.1145 (3.4398) grad_norm 1.5096 (1.7081) [2021-04-16 07:54:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][800/1251] eta 0:02:10 lr 0.000389 time 0.2794 (0.2884) loss 2.3220 (3.4379) grad_norm 1.6788 (1.7081) [2021-04-16 07:54:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][810/1251] eta 0:02:07 lr 0.000389 time 0.2801 (0.2884) loss 3.7250 (3.4372) grad_norm 1.9158 (1.7088) [2021-04-16 07:54:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][820/1251] eta 0:02:04 lr 0.000389 time 0.3012 (0.2883) loss 3.5206 (3.4388) grad_norm 1.4463 (1.7081) [2021-04-16 07:54:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][830/1251] eta 0:02:01 lr 0.000389 time 0.2672 (0.2882) loss 3.2570 (3.4378) grad_norm 1.7733 (1.7086) [2021-04-16 07:54:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][840/1251] eta 0:01:58 lr 0.000389 time 0.2758 (0.2881) loss 2.2993 (3.4363) grad_norm 1.4791 (1.7072) [2021-04-16 07:55:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][850/1251] eta 0:01:55 lr 0.000389 time 0.2930 (0.2881) loss 3.3696 (3.4402) grad_norm 1.4434 (1.7059) [2021-04-16 07:55:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][860/1251] eta 0:01:52 lr 0.000388 time 0.2955 (0.2881) loss 3.2106 (3.4393) grad_norm 1.8494 (1.7064) [2021-04-16 07:55:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][870/1251] eta 0:01:49 lr 0.000388 time 0.2901 (0.2880) loss 4.0130 (3.4417) grad_norm 1.6656 (1.7065) [2021-04-16 07:55:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][880/1251] eta 0:01:46 lr 0.000388 time 0.4070 (0.2881) loss 3.2938 (3.4439) grad_norm 1.5154 (1.7062) [2021-04-16 07:55:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][890/1251] eta 0:01:43 lr 0.000388 time 0.2874 (0.2879) loss 4.0574 (3.4438) grad_norm 2.1574 (1.7082) [2021-04-16 07:55:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][900/1251] eta 0:01:41 lr 0.000388 time 0.2574 (0.2879) loss 4.1356 (3.4486) grad_norm 1.5100 (1.7090) [2021-04-16 07:55:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][910/1251] eta 0:01:38 lr 0.000388 time 0.3025 (0.2878) loss 3.3951 (3.4503) grad_norm 2.0032 (1.7105) [2021-04-16 07:55:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][920/1251] eta 0:01:35 lr 0.000388 time 0.2845 (0.2877) loss 3.1177 (3.4515) grad_norm 2.0660 (1.7111) [2021-04-16 07:55:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][930/1251] eta 0:01:32 lr 0.000388 time 0.2706 (0.2877) loss 2.3586 (3.4485) grad_norm 1.5900 (1.7108) [2021-04-16 07:55:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][940/1251] eta 0:01:29 lr 0.000388 time 0.2708 (0.2879) loss 3.8275 (3.4483) grad_norm 1.6288 (1.7105) [2021-04-16 07:55:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][950/1251] eta 0:01:26 lr 0.000388 time 0.2964 (0.2880) loss 3.5480 (3.4484) grad_norm 1.6455 (1.7100) [2021-04-16 07:55:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][960/1251] eta 0:01:23 lr 0.000388 time 0.2776 (0.2880) loss 3.0609 (3.4485) grad_norm 1.9322 (1.7098) [2021-04-16 07:55:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][970/1251] eta 0:01:20 lr 0.000388 time 0.2675 (0.2879) loss 2.7140 (3.4450) grad_norm 1.5061 (1.7093) [2021-04-16 07:55:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][980/1251] eta 0:01:17 lr 0.000388 time 0.2887 (0.2878) loss 4.2069 (3.4444) grad_norm 1.7049 (1.7087) [2021-04-16 07:55:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][990/1251] eta 0:01:15 lr 0.000388 time 0.2762 (0.2876) loss 3.1069 (3.4416) grad_norm 1.6758 (1.7081) [2021-04-16 07:55:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][1000/1251] eta 0:01:12 lr 0.000388 time 0.2912 (0.2876) loss 4.1082 (3.4432) grad_norm 1.6828 (1.7074) [2021-04-16 07:55:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][1010/1251] eta 0:01:09 lr 0.000388 time 0.2727 (0.2875) loss 3.9201 (3.4434) grad_norm 1.5417 (1.7090) [2021-04-16 07:55:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][1020/1251] eta 0:01:06 lr 0.000388 time 0.2528 (0.2874) loss 4.4205 (3.4438) grad_norm 1.8092 (1.7092) [2021-04-16 07:55:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][1030/1251] eta 0:01:03 lr 0.000388 time 0.2731 (0.2874) loss 3.8060 (3.4423) grad_norm 1.9862 (1.7093) [2021-04-16 07:55:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][1040/1251] eta 0:01:00 lr 0.000388 time 0.2969 (0.2873) loss 3.9618 (3.4421) grad_norm 2.1804 (1.7095) [2021-04-16 07:55:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][1050/1251] eta 0:00:57 lr 0.000388 time 0.2688 (0.2872) loss 3.7347 (3.4447) grad_norm 1.4938 (1.7087) [2021-04-16 07:56:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][1060/1251] eta 0:00:54 lr 0.000388 time 0.2746 (0.2872) loss 4.0224 (3.4449) grad_norm 1.6053 (1.7083) [2021-04-16 07:56:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][1070/1251] eta 0:00:52 lr 0.000388 time 0.3023 (0.2873) loss 3.8888 (3.4441) grad_norm 1.8411 (1.7083) [2021-04-16 07:56:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][1080/1251] eta 0:00:49 lr 0.000388 time 0.2948 (0.2873) loss 3.1295 (3.4422) grad_norm 1.5459 (1.7085) [2021-04-16 07:56:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][1090/1251] eta 0:00:46 lr 0.000388 time 0.2628 (0.2872) loss 3.4647 (3.4403) grad_norm 1.6601 (1.7081) [2021-04-16 07:56:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][1100/1251] eta 0:00:43 lr 0.000388 time 0.2897 (0.2872) loss 4.2237 (3.4404) grad_norm 1.7509 (1.7089) [2021-04-16 07:56:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][1110/1251] eta 0:00:40 lr 0.000387 time 0.2926 (0.2871) loss 4.0979 (3.4391) grad_norm 2.0311 (1.7100) [2021-04-16 07:56:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][1120/1251] eta 0:00:37 lr 0.000387 time 0.2805 (0.2872) loss 3.5573 (3.4393) grad_norm 1.6423 (1.7104) [2021-04-16 07:56:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][1130/1251] eta 0:00:34 lr 0.000387 time 0.2832 (0.2871) loss 3.4322 (3.4407) grad_norm 1.6179 (1.7104) [2021-04-16 07:56:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][1140/1251] eta 0:00:31 lr 0.000387 time 0.2857 (0.2871) loss 3.3436 (3.4430) grad_norm 1.8183 (1.7112) [2021-04-16 07:56:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][1150/1251] eta 0:00:29 lr 0.000387 time 0.2706 (0.2871) loss 2.7046 (3.4418) grad_norm 1.7026 (1.7109) [2021-04-16 07:56:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][1160/1251] eta 0:00:26 lr 0.000387 time 0.3085 (0.2874) loss 4.3639 (3.4419) grad_norm 1.5518 (1.7109) [2021-04-16 07:56:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][1170/1251] eta 0:00:23 lr 0.000387 time 0.2786 (0.2873) loss 3.6631 (3.4403) grad_norm 1.5991 (1.7102) [2021-04-16 07:56:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][1180/1251] eta 0:00:20 lr 0.000387 time 0.2945 (0.2873) loss 3.3189 (3.4401) grad_norm 1.6472 (1.7093) [2021-04-16 07:56:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][1190/1251] eta 0:00:17 lr 0.000387 time 0.2748 (0.2872) loss 3.6548 (3.4394) grad_norm 1.8154 (1.7093) [2021-04-16 07:56:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][1200/1251] eta 0:00:14 lr 0.000387 time 0.2630 (0.2871) loss 3.9147 (3.4403) grad_norm 1.5301 (1.7085) [2021-04-16 07:56:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][1210/1251] eta 0:00:11 lr 0.000387 time 0.3054 (0.2871) loss 4.0279 (3.4422) grad_norm 1.6710 (1.7081) [2021-04-16 07:56:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][1220/1251] eta 0:00:08 lr 0.000387 time 0.2944 (0.2871) loss 3.6428 (3.4423) grad_norm 1.6496 (1.7073) [2021-04-16 07:56:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][1230/1251] eta 0:00:06 lr 0.000387 time 0.2700 (0.2871) loss 3.2933 (3.4404) grad_norm 1.5174 (1.7065) [2021-04-16 07:56:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][1240/1251] eta 0:00:03 lr 0.000387 time 0.2488 (0.2869) loss 2.7653 (3.4393) grad_norm 1.9262 (1.7068) [2021-04-16 07:56:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [172/300][1250/1251] eta 0:00:00 lr 0.000387 time 0.2486 (0.2866) loss 3.3006 (3.4388) grad_norm 1.6463 (1.7070) [2021-04-16 07:56:58 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 172 training takes 0:06:02 [2021-04-16 07:56:58 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_172.pth saving...... [2021-04-16 07:57:06 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_172.pth saved !!! [2021-04-16 07:57:07 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.151 (1.151) Loss 1.0282 (1.0282) Acc@1 75.977 (75.977) Acc@5 92.578 (92.578) [2021-04-16 07:57:08 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.098 (0.202) Loss 1.0330 (0.9729) Acc@1 75.977 (76.705) Acc@5 93.652 (93.759) [2021-04-16 07:57:11 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.809 (0.240) Loss 0.9984 (0.9667) Acc@1 77.539 (76.921) Acc@5 93.652 (93.764) [2021-04-16 07:57:13 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.092 (0.234) Loss 0.8991 (0.9661) Acc@1 78.613 (77.004) Acc@5 94.141 (93.728) [2021-04-16 07:57:15 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.221) Loss 0.9524 (0.9687) Acc@1 78.613 (77.001) Acc@5 94.531 (93.707) [2021-04-16 07:57:20 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 77.010 Acc@5 93.740 [2021-04-16 07:57:20 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 77.0% [2021-04-16 07:57:20 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 77.33% [2021-04-16 07:57:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][0/1251] eta 0:55:58 lr 0.000387 time 2.6843 (2.6843) loss 2.5810 (2.5810) grad_norm 1.6563 (1.6563) [2021-04-16 07:57:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][10/1251] eta 0:10:30 lr 0.000387 time 0.3885 (0.5084) loss 3.2522 (3.3374) grad_norm 1.7255 (1.7980) [2021-04-16 07:57:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][20/1251] eta 0:08:13 lr 0.000387 time 0.2686 (0.4012) loss 2.2831 (3.2577) grad_norm 1.5640 (1.7392) [2021-04-16 07:57:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][30/1251] eta 0:07:22 lr 0.000387 time 0.2654 (0.3625) loss 3.4017 (3.3172) grad_norm 1.4931 (1.7410) [2021-04-16 07:57:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][40/1251] eta 0:06:54 lr 0.000387 time 0.2594 (0.3425) loss 3.7299 (3.3423) grad_norm 1.8533 (1.7271) [2021-04-16 07:57:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][50/1251] eta 0:06:37 lr 0.000387 time 0.2866 (0.3311) loss 3.9115 (3.3778) grad_norm 1.8912 (1.7190) [2021-04-16 07:57:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][60/1251] eta 0:06:24 lr 0.000387 time 0.2735 (0.3226) loss 3.5621 (3.4115) grad_norm 1.4833 (1.7218) [2021-04-16 07:57:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][70/1251] eta 0:06:17 lr 0.000387 time 0.2996 (0.3196) loss 3.3980 (3.4383) grad_norm 1.9358 (1.7295) [2021-04-16 07:57:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][80/1251] eta 0:06:08 lr 0.000387 time 0.2575 (0.3148) loss 4.2725 (3.4626) grad_norm 1.7322 (1.7256) [2021-04-16 07:57:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][90/1251] eta 0:06:00 lr 0.000387 time 0.2680 (0.3105) loss 2.8736 (3.4409) grad_norm 1.9831 (1.7371) [2021-04-16 07:57:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][100/1251] eta 0:05:54 lr 0.000387 time 0.2980 (0.3076) loss 4.3816 (3.4678) grad_norm 1.4192 (1.7340) [2021-04-16 07:57:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][110/1251] eta 0:05:48 lr 0.000386 time 0.2608 (0.3056) loss 3.9108 (3.4767) grad_norm 1.7429 (1.7366) [2021-04-16 07:57:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][120/1251] eta 0:05:43 lr 0.000386 time 0.2800 (0.3036) loss 2.6146 (3.4593) grad_norm 1.7354 (1.7334) [2021-04-16 07:58:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][130/1251] eta 0:05:38 lr 0.000386 time 0.2642 (0.3018) loss 2.5130 (3.4611) grad_norm 1.4644 (1.7319) [2021-04-16 07:58:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][140/1251] eta 0:05:34 lr 0.000386 time 0.2821 (0.3015) loss 3.8148 (3.4728) grad_norm 1.6595 (1.7279) [2021-04-16 07:58:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][150/1251] eta 0:05:32 lr 0.000386 time 0.2798 (0.3016) loss 3.6642 (3.4740) grad_norm 1.8146 (1.7283) [2021-04-16 07:58:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][160/1251] eta 0:05:28 lr 0.000386 time 0.2936 (0.3012) loss 4.0507 (3.4785) grad_norm 1.6708 (1.7260) [2021-04-16 07:58:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][170/1251] eta 0:05:24 lr 0.000386 time 0.2702 (0.2997) loss 3.6324 (3.4678) grad_norm 1.6787 (1.7255) [2021-04-16 07:58:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][180/1251] eta 0:05:20 lr 0.000386 time 0.2581 (0.2990) loss 4.1191 (3.4480) grad_norm 1.8529 (1.7307) [2021-04-16 07:58:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][190/1251] eta 0:05:16 lr 0.000386 time 0.2809 (0.2980) loss 4.0129 (3.4627) grad_norm 1.7831 (1.7322) [2021-04-16 07:58:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][200/1251] eta 0:05:12 lr 0.000386 time 0.2840 (0.2970) loss 3.2582 (3.4508) grad_norm 1.4991 (1.7325) [2021-04-16 07:58:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][210/1251] eta 0:05:08 lr 0.000386 time 0.2993 (0.2961) loss 4.3868 (3.4628) grad_norm 1.6124 (1.7320) [2021-04-16 07:58:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][220/1251] eta 0:05:04 lr 0.000386 time 0.2922 (0.2955) loss 2.3525 (3.4530) grad_norm 2.0382 (1.7299) [2021-04-16 07:58:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][230/1251] eta 0:05:00 lr 0.000386 time 0.2772 (0.2947) loss 4.3911 (3.4397) grad_norm 1.4363 (1.7275) [2021-04-16 07:58:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][240/1251] eta 0:04:57 lr 0.000386 time 0.2724 (0.2944) loss 3.8889 (3.4346) grad_norm 2.1851 (1.7264) [2021-04-16 07:58:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][250/1251] eta 0:04:54 lr 0.000386 time 0.2842 (0.2940) loss 3.1725 (3.4401) grad_norm 1.5045 (1.7269) [2021-04-16 07:58:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][260/1251] eta 0:04:50 lr 0.000386 time 0.2943 (0.2934) loss 3.2922 (3.4385) grad_norm 1.8163 (1.7278) [2021-04-16 07:58:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][270/1251] eta 0:04:47 lr 0.000386 time 0.2600 (0.2929) loss 4.0796 (3.4402) grad_norm 1.7088 (1.7270) [2021-04-16 07:58:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][280/1251] eta 0:04:43 lr 0.000386 time 0.2878 (0.2925) loss 2.9909 (3.4416) grad_norm 1.7158 (1.7260) [2021-04-16 07:58:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][290/1251] eta 0:04:41 lr 0.000386 time 0.2666 (0.2925) loss 3.4996 (3.4380) grad_norm 1.7794 (1.7217) [2021-04-16 07:58:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][300/1251] eta 0:04:37 lr 0.000386 time 0.2618 (0.2918) loss 3.8136 (3.4308) grad_norm 1.7050 (1.7236) [2021-04-16 07:58:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][310/1251] eta 0:04:34 lr 0.000386 time 0.2881 (0.2915) loss 2.8192 (3.4274) grad_norm 2.0373 (1.7223) [2021-04-16 07:58:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][320/1251] eta 0:04:31 lr 0.000386 time 0.2617 (0.2916) loss 3.5252 (3.4245) grad_norm 1.7254 (1.7249) [2021-04-16 07:58:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][330/1251] eta 0:04:28 lr 0.000386 time 0.2883 (0.2918) loss 3.4176 (3.4254) grad_norm 1.6918 (1.7250) [2021-04-16 07:59:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][340/1251] eta 0:04:25 lr 0.000386 time 0.2649 (0.2912) loss 3.3347 (3.4308) grad_norm 1.9826 (1.7249) [2021-04-16 07:59:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][350/1251] eta 0:04:22 lr 0.000386 time 0.2497 (0.2908) loss 3.8052 (3.4326) grad_norm 1.6566 (1.7244) [2021-04-16 07:59:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][360/1251] eta 0:04:19 lr 0.000385 time 0.2423 (0.2911) loss 2.6972 (3.4326) grad_norm 1.7054 (1.7209) [2021-04-16 07:59:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][370/1251] eta 0:04:16 lr 0.000385 time 0.2910 (0.2908) loss 4.0043 (3.4316) grad_norm 1.7408 (1.7220) [2021-04-16 07:59:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][380/1251] eta 0:04:13 lr 0.000385 time 0.2703 (0.2906) loss 2.7852 (3.4208) grad_norm 1.6244 (1.7211) [2021-04-16 07:59:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][390/1251] eta 0:04:10 lr 0.000385 time 0.2789 (0.2904) loss 3.2357 (3.4255) grad_norm 1.8485 (1.7265) [2021-04-16 07:59:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][400/1251] eta 0:04:06 lr 0.000385 time 0.2876 (0.2902) loss 3.2187 (3.4311) grad_norm 1.6071 (1.7294) [2021-04-16 07:59:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][410/1251] eta 0:04:03 lr 0.000385 time 0.3146 (0.2901) loss 3.3625 (3.4285) grad_norm 1.8040 (1.7289) [2021-04-16 07:59:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][420/1251] eta 0:04:00 lr 0.000385 time 0.2609 (0.2897) loss 3.0621 (3.4224) grad_norm 1.5173 (1.7268) [2021-04-16 07:59:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][430/1251] eta 0:03:57 lr 0.000385 time 0.2784 (0.2895) loss 3.5974 (3.4229) grad_norm 2.2665 (1.7282) [2021-04-16 07:59:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][440/1251] eta 0:03:54 lr 0.000385 time 0.2543 (0.2892) loss 2.9542 (3.4203) grad_norm 1.5634 (1.7278) [2021-04-16 07:59:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][450/1251] eta 0:03:51 lr 0.000385 time 0.2743 (0.2889) loss 4.6181 (3.4211) grad_norm 2.0715 (1.7296) [2021-04-16 07:59:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][460/1251] eta 0:03:48 lr 0.000385 time 0.2982 (0.2888) loss 3.8287 (3.4237) grad_norm 1.5828 (1.7290) [2021-04-16 07:59:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][470/1251] eta 0:03:45 lr 0.000385 time 0.2801 (0.2886) loss 3.5292 (3.4236) grad_norm 1.8953 (1.7294) [2021-04-16 07:59:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][480/1251] eta 0:03:42 lr 0.000385 time 0.2648 (0.2887) loss 3.7114 (3.4301) grad_norm 1.6485 (1.7312) [2021-04-16 07:59:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][490/1251] eta 0:03:39 lr 0.000385 time 0.2818 (0.2886) loss 3.7836 (3.4337) grad_norm 1.7281 (1.7310) [2021-04-16 07:59:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][500/1251] eta 0:03:36 lr 0.000385 time 0.2931 (0.2884) loss 3.1751 (3.4323) grad_norm 1.8637 (1.7296) [2021-04-16 07:59:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][510/1251] eta 0:03:33 lr 0.000385 time 0.2492 (0.2881) loss 3.0637 (3.4290) grad_norm 1.7572 (1.7284) [2021-04-16 07:59:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][520/1251] eta 0:03:30 lr 0.000385 time 0.2514 (0.2878) loss 2.7351 (3.4259) grad_norm 1.7032 (1.7276) [2021-04-16 07:59:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][530/1251] eta 0:03:27 lr 0.000385 time 0.2697 (0.2875) loss 2.6851 (3.4295) grad_norm 1.4549 (1.7259) [2021-04-16 07:59:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][540/1251] eta 0:03:24 lr 0.000385 time 0.2984 (0.2877) loss 3.1858 (3.4283) grad_norm 1.7664 (1.7253) [2021-04-16 07:59:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][550/1251] eta 0:03:21 lr 0.000385 time 0.2906 (0.2875) loss 3.6555 (3.4314) grad_norm 1.5858 (1.7253) [2021-04-16 08:00:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][560/1251] eta 0:03:18 lr 0.000385 time 0.2961 (0.2874) loss 2.4122 (3.4296) grad_norm 1.5725 (1.7256) [2021-04-16 08:00:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][570/1251] eta 0:03:15 lr 0.000385 time 0.2798 (0.2872) loss 3.6516 (3.4316) grad_norm 1.4783 (1.7249) [2021-04-16 08:00:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][580/1251] eta 0:03:12 lr 0.000385 time 0.2980 (0.2873) loss 3.3958 (3.4359) grad_norm 1.7847 (1.7239) [2021-04-16 08:00:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][590/1251] eta 0:03:09 lr 0.000385 time 0.2651 (0.2873) loss 3.6510 (3.4387) grad_norm 1.7874 (1.7231) [2021-04-16 08:00:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][600/1251] eta 0:03:06 lr 0.000385 time 0.2856 (0.2872) loss 3.7164 (3.4447) grad_norm 1.7084 (1.7257) [2021-04-16 08:00:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][610/1251] eta 0:03:04 lr 0.000384 time 0.2751 (0.2871) loss 2.8832 (3.4467) grad_norm 2.0515 (1.7294) [2021-04-16 08:00:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][620/1251] eta 0:03:01 lr 0.000384 time 0.2774 (0.2869) loss 3.2647 (3.4476) grad_norm 1.6697 (1.7296) [2021-04-16 08:00:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][630/1251] eta 0:02:58 lr 0.000384 time 0.2759 (0.2868) loss 3.4840 (3.4501) grad_norm 1.6950 (1.7281) [2021-04-16 08:00:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][640/1251] eta 0:02:55 lr 0.000384 time 0.2710 (0.2867) loss 3.6558 (3.4538) grad_norm 1.6517 (1.7275) [2021-04-16 08:00:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][650/1251] eta 0:02:52 lr 0.000384 time 0.2719 (0.2865) loss 3.8319 (3.4593) grad_norm 1.6425 (1.7262) [2021-04-16 08:00:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][660/1251] eta 0:02:49 lr 0.000384 time 0.2793 (0.2864) loss 1.8210 (3.4572) grad_norm 1.4116 (1.7260) [2021-04-16 08:00:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][670/1251] eta 0:02:46 lr 0.000384 time 0.2693 (0.2862) loss 3.8322 (3.4573) grad_norm 1.7666 (1.7258) [2021-04-16 08:00:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][680/1251] eta 0:02:43 lr 0.000384 time 0.2621 (0.2862) loss 2.9350 (3.4547) grad_norm 1.7422 (1.7276) [2021-04-16 08:00:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][690/1251] eta 0:02:40 lr 0.000384 time 0.2788 (0.2860) loss 3.2161 (3.4564) grad_norm 2.0232 (1.7267) [2021-04-16 08:00:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][700/1251] eta 0:02:37 lr 0.000384 time 0.2756 (0.2859) loss 3.3320 (3.4544) grad_norm 1.4730 (1.7267) [2021-04-16 08:00:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][710/1251] eta 0:02:34 lr 0.000384 time 0.2820 (0.2858) loss 3.2880 (3.4550) grad_norm 1.6362 (1.7274) [2021-04-16 08:00:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][720/1251] eta 0:02:31 lr 0.000384 time 0.2623 (0.2859) loss 3.8426 (3.4555) grad_norm 1.4692 (1.7280) [2021-04-16 08:00:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][730/1251] eta 0:02:29 lr 0.000384 time 0.2868 (0.2862) loss 3.2177 (3.4544) grad_norm 1.8265 (1.7274) [2021-04-16 08:00:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][740/1251] eta 0:02:26 lr 0.000384 time 0.2836 (0.2862) loss 2.6929 (3.4538) grad_norm 1.7553 (1.7267) [2021-04-16 08:00:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][750/1251] eta 0:02:23 lr 0.000384 time 0.2820 (0.2860) loss 2.8414 (3.4489) grad_norm 1.4505 (1.7257) [2021-04-16 08:00:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][760/1251] eta 0:02:20 lr 0.000384 time 0.2619 (0.2859) loss 3.2632 (3.4485) grad_norm 2.2960 (1.7266) [2021-04-16 08:01:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][770/1251] eta 0:02:17 lr 0.000384 time 0.2733 (0.2858) loss 3.3242 (3.4492) grad_norm 1.9196 (nan) [2021-04-16 08:01:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][780/1251] eta 0:02:14 lr 0.000384 time 0.2730 (0.2857) loss 3.6356 (3.4510) grad_norm 1.5817 (nan) [2021-04-16 08:01:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][790/1251] eta 0:02:11 lr 0.000384 time 0.2721 (0.2856) loss 3.5303 (3.4515) grad_norm 1.7515 (nan) [2021-04-16 08:01:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][800/1251] eta 0:02:08 lr 0.000384 time 0.2555 (0.2855) loss 2.3824 (3.4499) grad_norm 1.5729 (nan) [2021-04-16 08:01:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][810/1251] eta 0:02:05 lr 0.000384 time 0.2665 (0.2854) loss 2.8650 (3.4466) grad_norm 1.5244 (nan) [2021-04-16 08:01:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][820/1251] eta 0:02:02 lr 0.000384 time 0.2633 (0.2854) loss 4.1913 (3.4507) grad_norm 1.5613 (nan) [2021-04-16 08:01:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][830/1251] eta 0:02:00 lr 0.000384 time 0.2573 (0.2853) loss 3.7223 (3.4508) grad_norm 1.6363 (nan) [2021-04-16 08:01:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][840/1251] eta 0:01:57 lr 0.000384 time 0.2846 (0.2851) loss 3.6200 (3.4535) grad_norm 1.4729 (nan) [2021-04-16 08:01:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][850/1251] eta 0:01:54 lr 0.000384 time 0.2676 (0.2851) loss 4.3126 (3.4543) grad_norm 1.7390 (nan) [2021-04-16 08:01:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][860/1251] eta 0:01:51 lr 0.000383 time 0.2775 (0.2850) loss 3.7877 (3.4535) grad_norm 1.7892 (nan) [2021-04-16 08:01:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][870/1251] eta 0:01:48 lr 0.000383 time 0.2691 (0.2851) loss 3.9012 (3.4560) grad_norm 1.5694 (nan) [2021-04-16 08:01:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][880/1251] eta 0:01:45 lr 0.000383 time 0.2545 (0.2850) loss 3.3212 (3.4544) grad_norm 1.7995 (nan) [2021-04-16 08:01:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][890/1251] eta 0:01:42 lr 0.000383 time 0.2747 (0.2851) loss 2.0592 (3.4534) grad_norm 3.4033 (nan) [2021-04-16 08:01:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][900/1251] eta 0:01:40 lr 0.000383 time 0.3895 (0.2852) loss 3.8825 (3.4557) grad_norm 1.7107 (nan) [2021-04-16 08:01:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][910/1251] eta 0:01:37 lr 0.000383 time 0.2749 (0.2851) loss 3.2946 (3.4545) grad_norm 1.7888 (nan) [2021-04-16 08:01:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][920/1251] eta 0:01:34 lr 0.000383 time 0.3018 (0.2850) loss 2.3476 (3.4546) grad_norm 1.5215 (nan) [2021-04-16 08:01:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][930/1251] eta 0:01:31 lr 0.000383 time 0.2986 (0.2851) loss 3.6495 (3.4554) grad_norm 1.8360 (nan) [2021-04-16 08:01:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][940/1251] eta 0:01:28 lr 0.000383 time 0.2945 (0.2854) loss 4.0554 (3.4560) grad_norm 1.5121 (nan) [2021-04-16 08:01:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][950/1251] eta 0:01:25 lr 0.000383 time 0.2710 (0.2853) loss 4.0543 (3.4567) grad_norm 1.8343 (nan) [2021-04-16 08:01:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][960/1251] eta 0:01:22 lr 0.000383 time 0.2914 (0.2852) loss 3.3773 (3.4571) grad_norm 1.5754 (nan) [2021-04-16 08:01:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][970/1251] eta 0:01:20 lr 0.000383 time 0.3140 (0.2852) loss 2.9047 (3.4566) grad_norm 1.7089 (nan) [2021-04-16 08:02:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][980/1251] eta 0:01:17 lr 0.000383 time 0.2814 (0.2851) loss 4.1641 (3.4547) grad_norm 1.5562 (nan) [2021-04-16 08:02:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][990/1251] eta 0:01:14 lr 0.000383 time 0.2657 (0.2851) loss 3.3681 (3.4573) grad_norm 1.7518 (nan) [2021-04-16 08:02:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][1000/1251] eta 0:01:11 lr 0.000383 time 0.2776 (0.2849) loss 3.6529 (3.4596) grad_norm 1.5360 (nan) [2021-04-16 08:02:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][1010/1251] eta 0:01:08 lr 0.000383 time 0.3024 (0.2850) loss 3.8228 (3.4613) grad_norm 1.8305 (nan) [2021-04-16 08:02:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][1020/1251] eta 0:01:05 lr 0.000383 time 0.2713 (0.2849) loss 2.7567 (3.4598) grad_norm 1.5087 (nan) [2021-04-16 08:02:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][1030/1251] eta 0:01:02 lr 0.000383 time 0.2588 (0.2847) loss 3.7815 (3.4614) grad_norm 1.7775 (nan) [2021-04-16 08:02:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][1040/1251] eta 0:01:00 lr 0.000383 time 0.2887 (0.2847) loss 2.6275 (3.4618) grad_norm 1.6667 (nan) [2021-04-16 08:02:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][1050/1251] eta 0:00:57 lr 0.000383 time 0.4278 (0.2849) loss 2.9372 (3.4581) grad_norm 1.6320 (nan) [2021-04-16 08:02:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][1060/1251] eta 0:00:54 lr 0.000383 time 0.2778 (0.2848) loss 3.2433 (3.4602) grad_norm 1.7058 (nan) [2021-04-16 08:02:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][1070/1251] eta 0:00:51 lr 0.000383 time 0.2829 (0.2848) loss 2.5715 (3.4590) grad_norm 1.8930 (nan) [2021-04-16 08:02:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][1080/1251] eta 0:00:48 lr 0.000383 time 0.2805 (0.2847) loss 2.6971 (3.4593) grad_norm 1.6939 (nan) [2021-04-16 08:02:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][1090/1251] eta 0:00:45 lr 0.000383 time 0.2661 (0.2846) loss 2.9857 (3.4596) grad_norm 2.1577 (nan) [2021-04-16 08:02:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][1100/1251] eta 0:00:42 lr 0.000383 time 0.2984 (0.2845) loss 4.0637 (3.4620) grad_norm 1.6786 (nan) [2021-04-16 08:02:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][1110/1251] eta 0:00:40 lr 0.000382 time 0.2686 (0.2845) loss 4.0037 (3.4635) grad_norm 1.6129 (nan) [2021-04-16 08:02:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][1120/1251] eta 0:00:37 lr 0.000382 time 0.2742 (0.2846) loss 3.6659 (3.4630) grad_norm 2.0203 (nan) [2021-04-16 08:02:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][1130/1251] eta 0:00:34 lr 0.000382 time 0.2783 (0.2845) loss 3.4585 (3.4629) grad_norm 1.9033 (nan) [2021-04-16 08:02:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][1140/1251] eta 0:00:31 lr 0.000382 time 0.2769 (0.2845) loss 3.3794 (3.4601) grad_norm 1.4582 (nan) [2021-04-16 08:02:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][1150/1251] eta 0:00:28 lr 0.000382 time 0.2684 (0.2844) loss 2.7002 (3.4604) grad_norm 1.8597 (nan) [2021-04-16 08:02:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][1160/1251] eta 0:00:25 lr 0.000382 time 0.2616 (0.2844) loss 3.5275 (3.4607) grad_norm 1.8993 (nan) [2021-04-16 08:02:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][1170/1251] eta 0:00:23 lr 0.000382 time 0.2649 (0.2844) loss 3.2872 (3.4606) grad_norm 1.6810 (nan) [2021-04-16 08:02:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][1180/1251] eta 0:00:20 lr 0.000382 time 0.2600 (0.2843) loss 3.2607 (3.4607) grad_norm 1.8102 (nan) [2021-04-16 08:02:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][1190/1251] eta 0:00:17 lr 0.000382 time 0.2614 (0.2843) loss 3.6398 (3.4598) grad_norm 1.6339 (nan) [2021-04-16 08:03:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][1200/1251] eta 0:00:14 lr 0.000382 time 0.2796 (0.2842) loss 4.0031 (3.4615) grad_norm 1.6299 (nan) [2021-04-16 08:03:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][1210/1251] eta 0:00:11 lr 0.000382 time 0.2622 (0.2842) loss 3.7548 (3.4614) grad_norm 1.6085 (nan) [2021-04-16 08:03:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][1220/1251] eta 0:00:08 lr 0.000382 time 0.2738 (0.2842) loss 3.2878 (3.4607) grad_norm 1.7424 (nan) [2021-04-16 08:03:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][1230/1251] eta 0:00:05 lr 0.000382 time 0.2880 (0.2841) loss 3.3329 (3.4601) grad_norm 1.9773 (nan) [2021-04-16 08:03:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][1240/1251] eta 0:00:03 lr 0.000382 time 0.2531 (0.2840) loss 3.5765 (3.4601) grad_norm 1.6744 (nan) [2021-04-16 08:03:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [173/300][1250/1251] eta 0:00:00 lr 0.000382 time 0.2483 (0.2837) loss 4.0138 (3.4603) grad_norm 1.4595 (nan) [2021-04-16 08:03:18 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 173 training takes 0:05:58 [2021-04-16 08:03:18 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_173.pth saving...... [2021-04-16 08:03:25 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_173.pth saved !!! [2021-04-16 08:03:27 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.183 (1.183) Loss 0.9601 (0.9601) Acc@1 78.027 (78.027) Acc@5 94.434 (94.434) [2021-04-16 08:03:28 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.458 (0.246) Loss 0.9501 (0.9811) Acc@1 77.930 (76.873) Acc@5 93.652 (93.954) [2021-04-16 08:03:31 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.376 (0.261) Loss 1.0129 (0.9769) Acc@1 77.539 (77.093) Acc@5 93.652 (93.996) [2021-04-16 08:03:32 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.116 (0.216) Loss 0.9294 (0.9748) Acc@1 78.809 (77.296) Acc@5 93.848 (93.980) [2021-04-16 08:03:34 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.210) Loss 1.0318 (0.9780) Acc@1 74.805 (77.160) Acc@5 93.652 (93.936) [2021-04-16 08:03:39 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 77.178 Acc@5 93.882 [2021-04-16 08:03:39 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 77.2% [2021-04-16 08:03:39 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 77.33% [2021-04-16 08:03:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][0/1251] eta 1:14:33 lr 0.000382 time 3.5756 (3.5756) loss 3.4601 (3.4601) grad_norm 1.4973 (1.4973) [2021-04-16 08:03:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][10/1251] eta 0:11:59 lr 0.000382 time 0.2902 (0.5800) loss 3.2805 (3.5200) grad_norm 1.6021 (1.6762) [2021-04-16 08:03:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][20/1251] eta 0:09:00 lr 0.000382 time 0.3184 (0.4387) loss 3.2343 (3.5887) grad_norm 1.6141 (1.6619) [2021-04-16 08:03:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][30/1251] eta 0:07:52 lr 0.000382 time 0.2662 (0.3869) loss 2.3267 (3.4984) grad_norm 2.0390 (1.6601) [2021-04-16 08:03:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][40/1251] eta 0:07:15 lr 0.000382 time 0.2811 (0.3598) loss 3.9050 (3.4927) grad_norm 1.6590 (1.6765) [2021-04-16 08:03:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][50/1251] eta 0:06:53 lr 0.000382 time 0.3097 (0.3444) loss 3.5458 (3.5288) grad_norm 2.1672 (1.6999) [2021-04-16 08:04:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][60/1251] eta 0:06:39 lr 0.000382 time 0.4086 (0.3356) loss 3.5612 (3.4932) grad_norm 1.4945 (1.6933) [2021-04-16 08:04:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][70/1251] eta 0:06:26 lr 0.000382 time 0.2764 (0.3269) loss 3.8551 (3.4546) grad_norm 1.4736 (1.6995) [2021-04-16 08:04:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][80/1251] eta 0:06:15 lr 0.000382 time 0.2625 (0.3204) loss 2.5716 (3.4383) grad_norm 1.7292 (1.7000) [2021-04-16 08:04:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][90/1251] eta 0:06:08 lr 0.000382 time 0.2860 (0.3172) loss 3.8054 (3.4271) grad_norm 1.9667 (1.7074) [2021-04-16 08:04:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][100/1251] eta 0:06:00 lr 0.000381 time 0.2821 (0.3136) loss 4.0131 (3.4294) grad_norm 2.3802 (1.7149) [2021-04-16 08:04:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][110/1251] eta 0:05:55 lr 0.000381 time 0.4255 (0.3118) loss 3.7098 (3.4367) grad_norm 1.7149 (1.7233) [2021-04-16 08:04:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][120/1251] eta 0:05:50 lr 0.000381 time 0.2855 (0.3100) loss 3.2601 (3.4165) grad_norm 1.8972 (1.7211) [2021-04-16 08:04:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][130/1251] eta 0:05:44 lr 0.000381 time 0.2497 (0.3074) loss 3.8860 (3.4215) grad_norm 1.6769 (1.7217) [2021-04-16 08:04:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][140/1251] eta 0:05:42 lr 0.000381 time 0.2600 (0.3081) loss 2.9790 (3.4161) grad_norm 1.6286 (1.7271) [2021-04-16 08:04:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][150/1251] eta 0:05:36 lr 0.000381 time 0.2503 (0.3057) loss 4.0619 (3.4062) grad_norm 1.5790 (1.7294) [2021-04-16 08:04:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][160/1251] eta 0:05:33 lr 0.000381 time 0.2497 (0.3053) loss 4.3544 (3.4221) grad_norm 1.8496 (1.7245) [2021-04-16 08:04:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][170/1251] eta 0:05:28 lr 0.000381 time 0.2666 (0.3037) loss 3.3990 (3.4108) grad_norm 1.5691 (1.7209) [2021-04-16 08:04:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][180/1251] eta 0:05:23 lr 0.000381 time 0.2762 (0.3023) loss 3.0532 (3.3975) grad_norm 1.5240 (1.7177) [2021-04-16 08:04:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][190/1251] eta 0:05:19 lr 0.000381 time 0.3047 (0.3014) loss 4.3000 (3.4010) grad_norm 1.6832 (1.7183) [2021-04-16 08:04:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][200/1251] eta 0:05:15 lr 0.000381 time 0.2798 (0.3002) loss 3.2855 (3.4094) grad_norm 2.0329 (1.7202) [2021-04-16 08:04:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][210/1251] eta 0:05:11 lr 0.000381 time 0.2727 (0.2991) loss 2.3696 (3.4070) grad_norm 1.7813 (1.7227) [2021-04-16 08:04:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][220/1251] eta 0:05:08 lr 0.000381 time 0.2908 (0.2988) loss 3.8829 (3.4214) grad_norm 1.5761 (1.7209) [2021-04-16 08:04:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][230/1251] eta 0:05:04 lr 0.000381 time 0.2661 (0.2979) loss 3.4406 (3.4255) grad_norm 1.6270 (1.7222) [2021-04-16 08:04:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][240/1251] eta 0:05:00 lr 0.000381 time 0.2837 (0.2972) loss 2.5385 (3.4276) grad_norm 1.5268 (1.7225) [2021-04-16 08:04:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][250/1251] eta 0:04:56 lr 0.000381 time 0.2868 (0.2964) loss 2.9971 (3.4235) grad_norm 2.0777 (1.7248) [2021-04-16 08:04:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][260/1251] eta 0:04:53 lr 0.000381 time 0.4141 (0.2963) loss 3.5172 (3.4072) grad_norm 1.3767 (1.7215) [2021-04-16 08:04:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][270/1251] eta 0:04:50 lr 0.000381 time 0.2636 (0.2958) loss 3.4166 (3.4143) grad_norm 1.9325 (1.7243) [2021-04-16 08:05:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][280/1251] eta 0:04:46 lr 0.000381 time 0.2718 (0.2950) loss 3.6698 (3.4168) grad_norm 1.5788 (1.7290) [2021-04-16 08:05:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][290/1251] eta 0:04:42 lr 0.000381 time 0.2604 (0.2942) loss 3.8131 (3.4296) grad_norm 2.0482 (1.7294) [2021-04-16 08:05:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][300/1251] eta 0:04:39 lr 0.000381 time 0.3056 (0.2937) loss 3.9246 (3.4398) grad_norm 1.4054 (1.7273) [2021-04-16 08:05:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][310/1251] eta 0:04:35 lr 0.000381 time 0.2785 (0.2932) loss 3.5328 (3.4459) grad_norm 1.4707 (1.7253) [2021-04-16 08:05:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][320/1251] eta 0:04:33 lr 0.000381 time 0.2890 (0.2933) loss 3.3821 (3.4486) grad_norm 1.8187 (1.7239) [2021-04-16 08:05:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][330/1251] eta 0:04:29 lr 0.000381 time 0.2915 (0.2928) loss 3.3899 (3.4560) grad_norm 1.7931 (1.7259) [2021-04-16 08:05:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][340/1251] eta 0:04:26 lr 0.000381 time 0.2838 (0.2923) loss 3.4909 (3.4466) grad_norm 1.6545 (1.7255) [2021-04-16 08:05:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][350/1251] eta 0:04:23 lr 0.000380 time 0.2964 (0.2920) loss 3.4823 (3.4445) grad_norm 1.7290 (1.7254) [2021-04-16 08:05:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][360/1251] eta 0:04:19 lr 0.000380 time 0.2773 (0.2916) loss 2.1512 (3.4367) grad_norm 1.5740 (1.7247) [2021-04-16 08:05:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][370/1251] eta 0:04:16 lr 0.000380 time 0.2797 (0.2914) loss 3.9899 (3.4237) grad_norm 1.6060 (1.7250) [2021-04-16 08:05:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][380/1251] eta 0:04:13 lr 0.000380 time 0.2763 (0.2910) loss 2.7309 (3.4238) grad_norm 1.6991 (1.7223) [2021-04-16 08:05:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][390/1251] eta 0:04:10 lr 0.000380 time 0.2551 (0.2907) loss 4.2532 (3.4248) grad_norm 1.7684 (1.7228) [2021-04-16 08:05:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][400/1251] eta 0:04:07 lr 0.000380 time 0.2859 (0.2906) loss 3.1862 (3.4249) grad_norm 1.6223 (1.7217) [2021-04-16 08:05:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][410/1251] eta 0:04:04 lr 0.000380 time 0.2650 (0.2905) loss 3.0384 (3.4224) grad_norm 1.4518 (1.7179) [2021-04-16 08:05:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][420/1251] eta 0:04:01 lr 0.000380 time 0.2887 (0.2903) loss 3.9083 (3.4252) grad_norm 1.6893 (1.7161) [2021-04-16 08:05:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][430/1251] eta 0:03:58 lr 0.000380 time 0.2645 (0.2901) loss 3.1451 (3.4171) grad_norm 1.5418 (1.7152) [2021-04-16 08:05:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][440/1251] eta 0:03:54 lr 0.000380 time 0.2829 (0.2897) loss 2.5707 (3.4138) grad_norm 1.5866 (1.7138) [2021-04-16 08:05:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][450/1251] eta 0:03:51 lr 0.000380 time 0.2724 (0.2895) loss 3.6826 (3.4142) grad_norm 1.6320 (1.7115) [2021-04-16 08:05:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][460/1251] eta 0:03:48 lr 0.000380 time 0.3046 (0.2893) loss 3.4107 (3.4178) grad_norm 1.7838 (1.7094) [2021-04-16 08:05:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][470/1251] eta 0:03:45 lr 0.000380 time 0.2629 (0.2891) loss 4.0968 (3.4149) grad_norm 1.7045 (1.7135) [2021-04-16 08:05:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][480/1251] eta 0:03:42 lr 0.000380 time 0.2825 (0.2889) loss 4.6181 (3.4111) grad_norm 1.8758 (1.7141) [2021-04-16 08:06:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][490/1251] eta 0:03:39 lr 0.000380 time 0.3078 (0.2887) loss 3.0551 (3.4153) grad_norm 1.6518 (1.7155) [2021-04-16 08:06:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][500/1251] eta 0:03:36 lr 0.000380 time 0.2603 (0.2883) loss 2.4868 (3.4096) grad_norm 1.7841 (1.7155) [2021-04-16 08:06:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][510/1251] eta 0:03:33 lr 0.000380 time 0.2656 (0.2881) loss 2.9874 (3.4080) grad_norm 1.5588 (1.7175) [2021-04-16 08:06:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][520/1251] eta 0:03:30 lr 0.000380 time 0.2760 (0.2881) loss 3.0305 (3.4019) grad_norm 2.0763 (1.7188) [2021-04-16 08:06:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][530/1251] eta 0:03:27 lr 0.000380 time 0.2824 (0.2880) loss 3.9522 (3.4060) grad_norm 1.9814 (1.7215) [2021-04-16 08:06:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][540/1251] eta 0:03:24 lr 0.000380 time 0.2824 (0.2879) loss 3.8551 (3.4117) grad_norm 1.6129 (1.7220) [2021-04-16 08:06:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][550/1251] eta 0:03:21 lr 0.000380 time 0.2740 (0.2878) loss 3.1682 (3.4101) grad_norm 1.5889 (1.7219) [2021-04-16 08:06:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][560/1251] eta 0:03:18 lr 0.000380 time 0.2890 (0.2876) loss 3.2069 (3.4067) grad_norm 2.0533 (1.7228) [2021-04-16 08:06:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][570/1251] eta 0:03:15 lr 0.000380 time 0.3003 (0.2877) loss 2.6985 (3.4068) grad_norm 1.7571 (1.7218) [2021-04-16 08:06:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][580/1251] eta 0:03:13 lr 0.000380 time 0.2627 (0.2879) loss 3.9589 (3.4064) grad_norm 1.6025 (1.7222) [2021-04-16 08:06:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][590/1251] eta 0:03:10 lr 0.000380 time 0.2934 (0.2878) loss 3.8165 (3.4079) grad_norm 1.6470 (1.7224) [2021-04-16 08:06:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][600/1251] eta 0:03:07 lr 0.000379 time 0.2487 (0.2876) loss 2.9084 (3.4090) grad_norm 1.5217 (1.7215) [2021-04-16 08:06:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][610/1251] eta 0:03:04 lr 0.000379 time 0.2780 (0.2875) loss 2.8382 (3.4070) grad_norm 1.5195 (1.7213) [2021-04-16 08:06:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][620/1251] eta 0:03:01 lr 0.000379 time 0.2617 (0.2873) loss 3.1663 (3.4040) grad_norm 1.6832 (1.7228) [2021-04-16 08:06:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][630/1251] eta 0:02:58 lr 0.000379 time 0.2623 (0.2871) loss 3.9113 (3.4057) grad_norm 1.7122 (1.7243) [2021-04-16 08:06:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][640/1251] eta 0:02:55 lr 0.000379 time 0.2641 (0.2871) loss 3.7491 (3.4080) grad_norm 1.7598 (1.7241) [2021-04-16 08:06:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][650/1251] eta 0:02:52 lr 0.000379 time 0.2776 (0.2870) loss 3.3565 (3.4117) grad_norm 1.6427 (1.7230) [2021-04-16 08:06:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][660/1251] eta 0:02:49 lr 0.000379 time 0.2968 (0.2869) loss 2.9822 (3.4064) grad_norm 1.5769 (1.7221) [2021-04-16 08:06:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][670/1251] eta 0:02:46 lr 0.000379 time 0.2600 (0.2868) loss 3.9084 (3.4077) grad_norm 1.8495 (1.7224) [2021-04-16 08:06:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][680/1251] eta 0:02:43 lr 0.000379 time 0.2810 (0.2867) loss 4.0077 (3.4107) grad_norm 1.6022 (1.7227) [2021-04-16 08:06:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][690/1251] eta 0:02:40 lr 0.000379 time 0.2739 (0.2867) loss 3.0292 (3.4110) grad_norm 1.5896 (1.7224) [2021-04-16 08:07:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][700/1251] eta 0:02:37 lr 0.000379 time 0.2726 (0.2866) loss 3.3660 (3.4160) grad_norm 1.6346 (1.7229) [2021-04-16 08:07:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][710/1251] eta 0:02:35 lr 0.000379 time 0.3656 (0.2866) loss 4.1845 (3.4205) grad_norm 1.5668 (1.7217) [2021-04-16 08:07:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][720/1251] eta 0:02:32 lr 0.000379 time 0.2597 (0.2868) loss 3.8457 (3.4197) grad_norm 1.7354 (1.7232) [2021-04-16 08:07:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][730/1251] eta 0:02:29 lr 0.000379 time 0.2881 (0.2867) loss 3.3133 (3.4188) grad_norm 1.5109 (1.7225) [2021-04-16 08:07:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][740/1251] eta 0:02:26 lr 0.000379 time 0.3158 (0.2866) loss 4.1948 (3.4215) grad_norm 1.6731 (1.7217) [2021-04-16 08:07:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][750/1251] eta 0:02:23 lr 0.000379 time 0.2756 (0.2865) loss 3.7669 (3.4245) grad_norm 1.6377 (1.7210) [2021-04-16 08:07:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][760/1251] eta 0:02:20 lr 0.000379 time 0.3021 (0.2864) loss 4.0829 (3.4253) grad_norm 1.7084 (1.7211) [2021-04-16 08:07:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][770/1251] eta 0:02:17 lr 0.000379 time 0.2683 (0.2862) loss 2.4745 (3.4221) grad_norm 2.0037 (1.7228) [2021-04-16 08:07:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][780/1251] eta 0:02:14 lr 0.000379 time 0.2965 (0.2861) loss 3.4769 (3.4240) grad_norm 1.8346 (1.7239) [2021-04-16 08:07:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][790/1251] eta 0:02:11 lr 0.000379 time 0.3131 (0.2863) loss 3.4842 (3.4227) grad_norm 1.5945 (1.7242) [2021-04-16 08:07:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][800/1251] eta 0:02:09 lr 0.000379 time 0.2673 (0.2861) loss 3.6783 (3.4232) grad_norm 1.5480 (1.7240) [2021-04-16 08:07:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][810/1251] eta 0:02:06 lr 0.000379 time 0.2804 (0.2860) loss 3.6159 (3.4249) grad_norm 1.7015 (1.7235) [2021-04-16 08:07:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][820/1251] eta 0:02:03 lr 0.000379 time 0.2783 (0.2858) loss 3.1234 (3.4220) grad_norm 2.0357 (1.7238) [2021-04-16 08:07:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][830/1251] eta 0:02:00 lr 0.000379 time 0.2823 (0.2858) loss 3.7398 (3.4224) grad_norm 1.6797 (1.7231) [2021-04-16 08:07:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][840/1251] eta 0:01:57 lr 0.000379 time 0.2792 (0.2858) loss 3.7090 (3.4226) grad_norm 1.5499 (1.7245) [2021-04-16 08:07:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][850/1251] eta 0:01:54 lr 0.000378 time 0.2735 (0.2858) loss 3.4744 (3.4203) grad_norm 1.8055 (1.7242) [2021-04-16 08:07:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][860/1251] eta 0:01:51 lr 0.000378 time 0.2706 (0.2857) loss 3.7718 (3.4220) grad_norm 1.6680 (1.7240) [2021-04-16 08:07:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][870/1251] eta 0:01:48 lr 0.000378 time 0.2657 (0.2857) loss 3.3687 (3.4227) grad_norm 1.7781 (1.7236) [2021-04-16 08:07:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][880/1251] eta 0:01:45 lr 0.000378 time 0.2884 (0.2856) loss 2.3990 (3.4240) grad_norm 1.6768 (1.7238) [2021-04-16 08:07:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][890/1251] eta 0:01:43 lr 0.000378 time 0.2677 (0.2856) loss 3.9881 (3.4278) grad_norm 1.6814 (1.7256) [2021-04-16 08:07:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][900/1251] eta 0:01:40 lr 0.000378 time 0.2685 (0.2854) loss 4.0880 (3.4302) grad_norm 1.7638 (1.7247) [2021-04-16 08:07:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][910/1251] eta 0:01:37 lr 0.000378 time 0.2819 (0.2854) loss 3.2979 (3.4297) grad_norm 1.8435 (1.7241) [2021-04-16 08:08:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][920/1251] eta 0:01:34 lr 0.000378 time 0.2871 (0.2853) loss 3.8295 (3.4285) grad_norm 2.1656 (1.7247) [2021-04-16 08:08:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][930/1251] eta 0:01:31 lr 0.000378 time 0.2668 (0.2852) loss 3.3107 (3.4286) grad_norm 1.6412 (1.7244) [2021-04-16 08:08:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][940/1251] eta 0:01:28 lr 0.000378 time 0.2755 (0.2853) loss 3.2677 (3.4283) grad_norm 1.8861 (1.7250) [2021-04-16 08:08:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][950/1251] eta 0:01:25 lr 0.000378 time 0.2711 (0.2852) loss 4.2404 (3.4253) grad_norm 1.6180 (1.7249) [2021-04-16 08:08:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][960/1251] eta 0:01:22 lr 0.000378 time 0.2906 (0.2851) loss 3.9531 (3.4265) grad_norm 1.7077 (1.7246) [2021-04-16 08:08:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][970/1251] eta 0:01:20 lr 0.000378 time 0.2953 (0.2851) loss 2.4110 (3.4275) grad_norm 1.8660 (1.7245) [2021-04-16 08:08:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][980/1251] eta 0:01:17 lr 0.000378 time 0.2598 (0.2852) loss 3.8299 (3.4269) grad_norm 1.5741 (1.7241) [2021-04-16 08:08:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][990/1251] eta 0:01:14 lr 0.000378 time 0.2580 (0.2851) loss 3.4345 (3.4278) grad_norm 1.5096 (1.7237) [2021-04-16 08:08:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][1000/1251] eta 0:01:11 lr 0.000378 time 0.2767 (0.2850) loss 4.0937 (3.4279) grad_norm 1.7767 (1.7238) [2021-04-16 08:08:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][1010/1251] eta 0:01:08 lr 0.000378 time 0.2697 (0.2850) loss 3.5405 (3.4264) grad_norm 1.6140 (1.7234) [2021-04-16 08:08:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][1020/1251] eta 0:01:05 lr 0.000378 time 0.2657 (0.2849) loss 3.6416 (3.4263) grad_norm 2.0497 (1.7241) [2021-04-16 08:08:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][1030/1251] eta 0:01:02 lr 0.000378 time 0.2684 (0.2848) loss 3.2598 (3.4252) grad_norm 1.5668 (1.7249) [2021-04-16 08:08:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][1040/1251] eta 0:01:00 lr 0.000378 time 0.2706 (0.2848) loss 3.9928 (3.4244) grad_norm 1.5735 (1.7244) [2021-04-16 08:08:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][1050/1251] eta 0:00:57 lr 0.000378 time 0.2776 (0.2847) loss 2.1777 (3.4216) grad_norm 1.7805 (1.7247) [2021-04-16 08:08:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][1060/1251] eta 0:00:54 lr 0.000378 time 0.2660 (0.2846) loss 2.6175 (3.4223) grad_norm 1.7123 (1.7259) [2021-04-16 08:08:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][1070/1251] eta 0:00:51 lr 0.000378 time 0.2611 (0.2845) loss 3.8764 (3.4229) grad_norm 1.4686 (1.7254) [2021-04-16 08:08:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][1080/1251] eta 0:00:48 lr 0.000378 time 0.2755 (0.2845) loss 3.8785 (3.4253) grad_norm 1.6669 (1.7254) [2021-04-16 08:08:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][1090/1251] eta 0:00:45 lr 0.000378 time 0.2957 (0.2845) loss 4.0698 (3.4271) grad_norm 2.0022 (1.7252) [2021-04-16 08:08:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][1100/1251] eta 0:00:42 lr 0.000377 time 0.3241 (0.2844) loss 3.6943 (3.4265) grad_norm 1.4723 (1.7251) [2021-04-16 08:08:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][1110/1251] eta 0:00:40 lr 0.000377 time 0.2810 (0.2844) loss 4.0894 (3.4264) grad_norm 1.9351 (1.7250) [2021-04-16 08:08:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][1120/1251] eta 0:00:37 lr 0.000377 time 0.2791 (0.2844) loss 3.3560 (3.4259) grad_norm 1.8227 (1.7248) [2021-04-16 08:09:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][1130/1251] eta 0:00:34 lr 0.000377 time 0.2673 (0.2844) loss 4.3947 (3.4273) grad_norm 1.9023 (1.7243) [2021-04-16 08:09:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][1140/1251] eta 0:00:31 lr 0.000377 time 0.2928 (0.2844) loss 3.5309 (3.4262) grad_norm 1.7660 (1.7251) [2021-04-16 08:09:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][1150/1251] eta 0:00:28 lr 0.000377 time 0.2881 (0.2844) loss 2.4148 (3.4257) grad_norm 1.7177 (1.7251) [2021-04-16 08:09:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][1160/1251] eta 0:00:25 lr 0.000377 time 0.2700 (0.2846) loss 2.4670 (3.4263) grad_norm 1.4044 (1.7249) [2021-04-16 08:09:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][1170/1251] eta 0:00:23 lr 0.000377 time 0.2858 (0.2846) loss 4.0613 (3.4261) grad_norm 1.7171 (1.7244) [2021-04-16 08:09:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][1180/1251] eta 0:00:20 lr 0.000377 time 0.2759 (0.2846) loss 2.5010 (3.4253) grad_norm 1.6372 (1.7250) [2021-04-16 08:09:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][1190/1251] eta 0:00:17 lr 0.000377 time 0.2721 (0.2845) loss 2.6180 (3.4248) grad_norm 1.6278 (1.7253) [2021-04-16 08:09:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][1200/1251] eta 0:00:14 lr 0.000377 time 0.2818 (0.2845) loss 3.1267 (3.4227) grad_norm 1.5985 (1.7255) [2021-04-16 08:09:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][1210/1251] eta 0:00:11 lr 0.000377 time 0.2586 (0.2844) loss 2.6458 (3.4223) grad_norm 1.9816 (1.7263) [2021-04-16 08:09:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][1220/1251] eta 0:00:08 lr 0.000377 time 0.2703 (0.2844) loss 2.5505 (3.4235) grad_norm 2.1734 (1.7270) [2021-04-16 08:09:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][1230/1251] eta 0:00:05 lr 0.000377 time 0.2870 (0.2844) loss 2.9197 (3.4233) grad_norm 1.5351 (1.7270) [2021-04-16 08:09:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][1240/1251] eta 0:00:03 lr 0.000377 time 0.3048 (0.2843) loss 3.8639 (3.4235) grad_norm 1.4716 (1.7270) [2021-04-16 08:09:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [174/300][1250/1251] eta 0:00:00 lr 0.000377 time 0.2484 (0.2840) loss 2.7678 (3.4232) grad_norm 1.5362 (1.7266) [2021-04-16 08:09:40 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 174 training takes 0:06:00 [2021-04-16 08:09:40 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_174.pth saving...... [2021-04-16 08:09:53 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_174.pth saved !!! [2021-04-16 08:09:54 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.219 (1.219) Loss 0.9133 (0.9133) Acc@1 78.906 (78.906) Acc@5 93.848 (93.848) [2021-04-16 08:09:56 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.159 (0.265) Loss 0.9467 (0.9617) Acc@1 78.320 (77.317) Acc@5 94.043 (93.732) [2021-04-16 08:09:58 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.394 (0.235) Loss 0.9014 (0.9513) Acc@1 77.148 (77.386) Acc@5 95.898 (94.020) [2021-04-16 08:10:01 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.130 (0.253) Loss 0.8592 (0.9459) Acc@1 80.469 (77.580) Acc@5 94.531 (94.015) [2021-04-16 08:10:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.222) Loss 0.9257 (0.9524) Acc@1 79.004 (77.425) Acc@5 94.043 (93.874) [2021-04-16 08:10:10 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 77.272 Acc@5 93.782 [2021-04-16 08:10:10 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 77.3% [2021-04-16 08:10:10 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 77.33% [2021-04-16 08:10:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][0/1251] eta 1:36:23 lr 0.000377 time 4.6234 (4.6234) loss 4.0948 (4.0948) grad_norm 1.9771 (1.9771) [2021-04-16 08:10:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][10/1251] eta 0:13:59 lr 0.000377 time 0.3104 (0.6762) loss 3.9086 (3.4020) grad_norm 1.6235 (1.7401) [2021-04-16 08:10:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][20/1251] eta 0:09:59 lr 0.000377 time 0.2569 (0.4873) loss 3.6441 (3.5916) grad_norm 1.6571 (1.7068) [2021-04-16 08:10:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][30/1251] eta 0:08:33 lr 0.000377 time 0.2590 (0.4203) loss 3.1429 (3.5698) grad_norm 1.9795 (1.7481) [2021-04-16 08:10:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][40/1251] eta 0:07:50 lr 0.000377 time 0.2839 (0.3885) loss 4.3183 (3.5654) grad_norm 2.0410 (1.7623) [2021-04-16 08:10:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][50/1251] eta 0:07:20 lr 0.000377 time 0.2779 (0.3672) loss 2.9652 (3.5127) grad_norm 1.6976 (1.7468) [2021-04-16 08:10:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][60/1251] eta 0:07:00 lr 0.000377 time 0.2698 (0.3533) loss 4.0937 (3.5694) grad_norm 1.6009 (1.7422) [2021-04-16 08:10:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][70/1251] eta 0:06:44 lr 0.000377 time 0.2663 (0.3429) loss 2.1440 (3.5060) grad_norm 1.6455 (1.7381) [2021-04-16 08:10:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][80/1251] eta 0:06:32 lr 0.000377 time 0.3003 (0.3349) loss 3.8502 (3.4846) grad_norm 1.5927 (1.7423) [2021-04-16 08:10:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][90/1251] eta 0:06:22 lr 0.000377 time 0.2567 (0.3298) loss 2.9635 (3.4526) grad_norm 1.7864 (1.7438) [2021-04-16 08:10:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][100/1251] eta 0:06:13 lr 0.000376 time 0.2775 (0.3249) loss 3.7093 (3.4696) grad_norm 1.7331 (1.7391) [2021-04-16 08:10:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][110/1251] eta 0:06:06 lr 0.000376 time 0.3026 (0.3210) loss 3.3350 (3.4510) grad_norm 1.8505 (1.7424) [2021-04-16 08:10:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][120/1251] eta 0:05:59 lr 0.000376 time 0.3015 (0.3180) loss 3.3783 (3.4459) grad_norm 1.7555 (1.7416) [2021-04-16 08:10:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][130/1251] eta 0:05:52 lr 0.000376 time 0.2559 (0.3146) loss 3.3480 (3.4343) grad_norm 1.5312 (1.7465) [2021-04-16 08:10:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][140/1251] eta 0:05:47 lr 0.000376 time 0.2413 (0.3123) loss 2.1884 (3.4064) grad_norm 1.7755 (1.7414) [2021-04-16 08:10:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][150/1251] eta 0:05:41 lr 0.000376 time 0.2894 (0.3106) loss 3.9751 (3.4092) grad_norm 1.6556 (1.7359) [2021-04-16 08:10:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][160/1251] eta 0:05:37 lr 0.000376 time 0.2753 (0.3089) loss 3.2919 (3.3996) grad_norm 1.7207 (1.7329) [2021-04-16 08:11:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][170/1251] eta 0:05:31 lr 0.000376 time 0.2788 (0.3071) loss 3.8745 (3.4219) grad_norm 1.6478 (1.7315) [2021-04-16 08:11:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][180/1251] eta 0:05:28 lr 0.000376 time 0.4054 (0.3064) loss 3.7354 (3.4245) grad_norm 1.9308 (1.7365) [2021-04-16 08:11:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][190/1251] eta 0:05:23 lr 0.000376 time 0.2697 (0.3050) loss 4.2673 (3.4341) grad_norm 1.9039 (1.7376) [2021-04-16 08:11:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][200/1251] eta 0:05:19 lr 0.000376 time 0.2742 (0.3036) loss 3.5241 (3.4221) grad_norm 1.9169 (1.7371) [2021-04-16 08:11:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][210/1251] eta 0:05:14 lr 0.000376 time 0.2926 (0.3024) loss 3.9123 (3.4312) grad_norm 1.7369 (1.7339) [2021-04-16 08:11:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][220/1251] eta 0:05:10 lr 0.000376 time 0.2847 (0.3015) loss 3.4086 (3.4388) grad_norm 1.5480 (1.7374) [2021-04-16 08:11:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][230/1251] eta 0:05:06 lr 0.000376 time 0.2922 (0.3006) loss 3.8594 (3.4411) grad_norm 2.0895 (1.7385) [2021-04-16 08:11:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][240/1251] eta 0:05:02 lr 0.000376 time 0.2780 (0.2996) loss 2.0279 (3.4355) grad_norm 1.6485 (1.7380) [2021-04-16 08:11:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][250/1251] eta 0:04:58 lr 0.000376 time 0.2565 (0.2987) loss 3.0037 (3.4300) grad_norm 1.6871 (1.7388) [2021-04-16 08:11:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][260/1251] eta 0:04:55 lr 0.000376 time 0.2887 (0.2978) loss 2.4846 (3.4169) grad_norm 1.4771 (1.7330) [2021-04-16 08:11:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][270/1251] eta 0:04:51 lr 0.000376 time 0.2680 (0.2970) loss 3.0038 (3.4030) grad_norm 1.6945 (1.7350) [2021-04-16 08:11:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][280/1251] eta 0:04:47 lr 0.000376 time 0.2984 (0.2964) loss 3.3638 (3.4094) grad_norm 1.6677 (1.7343) [2021-04-16 08:11:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][290/1251] eta 0:04:44 lr 0.000376 time 0.2628 (0.2959) loss 4.2209 (3.4138) grad_norm 1.6583 (1.7373) [2021-04-16 08:11:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][300/1251] eta 0:04:41 lr 0.000376 time 0.2923 (0.2955) loss 2.7846 (3.4126) grad_norm 1.4684 (1.7342) [2021-04-16 08:11:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][310/1251] eta 0:04:37 lr 0.000376 time 0.2810 (0.2951) loss 2.6165 (3.4098) grad_norm 1.5254 (1.7322) [2021-04-16 08:11:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][320/1251] eta 0:04:34 lr 0.000376 time 0.2806 (0.2948) loss 4.0239 (3.4188) grad_norm 1.8428 (1.7331) [2021-04-16 08:11:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][330/1251] eta 0:04:31 lr 0.000376 time 0.2740 (0.2944) loss 3.2881 (3.4150) grad_norm 1.5470 (1.7319) [2021-04-16 08:11:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][340/1251] eta 0:04:27 lr 0.000376 time 0.2650 (0.2939) loss 2.2937 (3.4103) grad_norm 1.7603 (1.7314) [2021-04-16 08:11:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][350/1251] eta 0:04:24 lr 0.000375 time 0.2934 (0.2937) loss 3.8669 (3.4110) grad_norm 1.7683 (1.7317) [2021-04-16 08:11:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][360/1251] eta 0:04:21 lr 0.000375 time 0.2804 (0.2939) loss 3.9120 (3.4104) grad_norm 1.5878 (1.7318) [2021-04-16 08:11:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][370/1251] eta 0:04:18 lr 0.000375 time 0.2747 (0.2937) loss 2.7773 (3.4066) grad_norm 1.6572 (1.7295) [2021-04-16 08:12:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][380/1251] eta 0:04:15 lr 0.000375 time 0.2748 (0.2934) loss 2.3913 (3.4089) grad_norm 2.1455 (1.7348) [2021-04-16 08:12:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][390/1251] eta 0:04:12 lr 0.000375 time 0.2979 (0.2932) loss 2.4037 (3.4110) grad_norm 1.9482 (1.7399) [2021-04-16 08:12:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][400/1251] eta 0:04:09 lr 0.000375 time 0.3234 (0.2929) loss 4.1476 (3.4086) grad_norm 1.7854 (1.7383) [2021-04-16 08:12:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][410/1251] eta 0:04:06 lr 0.000375 time 0.2799 (0.2925) loss 3.3606 (3.4098) grad_norm 1.6397 (1.7402) [2021-04-16 08:12:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][420/1251] eta 0:04:03 lr 0.000375 time 0.3015 (0.2927) loss 4.2493 (3.4119) grad_norm 1.6815 (1.7401) [2021-04-16 08:12:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][430/1251] eta 0:04:00 lr 0.000375 time 0.2857 (0.2924) loss 3.6186 (3.4143) grad_norm 1.7920 (1.7379) [2021-04-16 08:12:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][440/1251] eta 0:03:56 lr 0.000375 time 0.2902 (0.2920) loss 3.3006 (3.4183) grad_norm 1.6195 (1.7375) [2021-04-16 08:12:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][450/1251] eta 0:03:53 lr 0.000375 time 0.4128 (0.2919) loss 2.6280 (3.4072) grad_norm 2.1042 (1.7386) [2021-04-16 08:12:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][460/1251] eta 0:03:50 lr 0.000375 time 0.2824 (0.2916) loss 2.8111 (3.4024) grad_norm 1.8836 (1.7399) [2021-04-16 08:12:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][470/1251] eta 0:03:47 lr 0.000375 time 0.3066 (0.2914) loss 3.6528 (3.4001) grad_norm 1.5278 (1.7381) [2021-04-16 08:12:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][480/1251] eta 0:03:44 lr 0.000375 time 0.2556 (0.2913) loss 2.5031 (3.3987) grad_norm 2.3681 (1.7376) [2021-04-16 08:12:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][490/1251] eta 0:03:41 lr 0.000375 time 0.2819 (0.2912) loss 3.7485 (3.3944) grad_norm 1.6284 (1.7381) [2021-04-16 08:12:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][500/1251] eta 0:03:38 lr 0.000375 time 0.2827 (0.2908) loss 2.5535 (3.3906) grad_norm 1.7332 (1.7388) [2021-04-16 08:12:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][510/1251] eta 0:03:35 lr 0.000375 time 0.2688 (0.2905) loss 2.9433 (3.3933) grad_norm 1.8501 (1.7390) [2021-04-16 08:12:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][520/1251] eta 0:03:32 lr 0.000375 time 0.2795 (0.2903) loss 3.2827 (3.3939) grad_norm 1.7521 (1.7379) [2021-04-16 08:12:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][530/1251] eta 0:03:29 lr 0.000375 time 0.2666 (0.2902) loss 2.6811 (3.3930) grad_norm 1.5415 (1.7372) [2021-04-16 08:12:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][540/1251] eta 0:03:26 lr 0.000375 time 0.2697 (0.2900) loss 2.6363 (3.3936) grad_norm 1.6042 (1.7368) [2021-04-16 08:12:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][550/1251] eta 0:03:23 lr 0.000375 time 0.2745 (0.2898) loss 2.8922 (3.3867) grad_norm 2.1614 (1.7376) [2021-04-16 08:12:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][560/1251] eta 0:03:20 lr 0.000375 time 0.2569 (0.2895) loss 3.9519 (3.3918) grad_norm 1.5687 (1.7393) [2021-04-16 08:12:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][570/1251] eta 0:03:17 lr 0.000375 time 0.2714 (0.2895) loss 2.1267 (3.3844) grad_norm 1.7082 (1.7414) [2021-04-16 08:12:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][580/1251] eta 0:03:14 lr 0.000375 time 0.2680 (0.2897) loss 3.8357 (3.3889) grad_norm 1.7084 (1.7423) [2021-04-16 08:13:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][590/1251] eta 0:03:11 lr 0.000375 time 0.2557 (0.2897) loss 3.5859 (3.3933) grad_norm 1.6678 (1.7422) [2021-04-16 08:13:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][600/1251] eta 0:03:08 lr 0.000374 time 0.2830 (0.2895) loss 4.1938 (3.3956) grad_norm 1.8463 (1.7424) [2021-04-16 08:13:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][610/1251] eta 0:03:05 lr 0.000374 time 0.2851 (0.2893) loss 3.9645 (3.3939) grad_norm 1.7192 (1.7411) [2021-04-16 08:13:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][620/1251] eta 0:03:02 lr 0.000374 time 0.2899 (0.2894) loss 3.2784 (3.3973) grad_norm 1.5726 (1.7393) [2021-04-16 08:13:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][630/1251] eta 0:02:59 lr 0.000374 time 0.2757 (0.2893) loss 4.0570 (3.3954) grad_norm 1.7388 (1.7376) [2021-04-16 08:13:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][640/1251] eta 0:02:56 lr 0.000374 time 0.2920 (0.2891) loss 4.0293 (3.3958) grad_norm 1.9736 (1.7384) [2021-04-16 08:13:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][650/1251] eta 0:02:53 lr 0.000374 time 0.2762 (0.2890) loss 2.7788 (3.3985) grad_norm 1.9408 (1.7383) [2021-04-16 08:13:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][660/1251] eta 0:02:50 lr 0.000374 time 0.2830 (0.2888) loss 3.6107 (3.3926) grad_norm 1.5378 (1.7376) [2021-04-16 08:13:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][670/1251] eta 0:02:47 lr 0.000374 time 0.2714 (0.2889) loss 3.4442 (3.3967) grad_norm 1.8396 (1.7373) [2021-04-16 08:13:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][680/1251] eta 0:02:44 lr 0.000374 time 0.2475 (0.2887) loss 3.3706 (3.3996) grad_norm 1.8188 (1.7375) [2021-04-16 08:13:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][690/1251] eta 0:02:41 lr 0.000374 time 0.2885 (0.2886) loss 3.1790 (3.3959) grad_norm 2.3730 (1.7380) [2021-04-16 08:13:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][700/1251] eta 0:02:38 lr 0.000374 time 0.2597 (0.2885) loss 2.5603 (3.3954) grad_norm 1.6263 (1.7377) [2021-04-16 08:13:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][710/1251] eta 0:02:36 lr 0.000374 time 0.2893 (0.2884) loss 3.0338 (3.3927) grad_norm 1.6709 (1.7382) [2021-04-16 08:13:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][720/1251] eta 0:02:33 lr 0.000374 time 0.2725 (0.2884) loss 2.0101 (3.3939) grad_norm 1.4688 (1.7381) [2021-04-16 08:13:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][730/1251] eta 0:02:30 lr 0.000374 time 0.2565 (0.2882) loss 3.5271 (3.3934) grad_norm 1.5927 (1.7390) [2021-04-16 08:13:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][740/1251] eta 0:02:27 lr 0.000374 time 0.2691 (0.2881) loss 3.8082 (3.3915) grad_norm 1.6371 (1.7372) [2021-04-16 08:13:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][750/1251] eta 0:02:24 lr 0.000374 time 0.2700 (0.2880) loss 2.8824 (3.3906) grad_norm 1.6873 (1.7375) [2021-04-16 08:13:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][760/1251] eta 0:02:21 lr 0.000374 time 0.2907 (0.2880) loss 3.9455 (3.3933) grad_norm 1.7034 (1.7373) [2021-04-16 08:13:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][770/1251] eta 0:02:18 lr 0.000374 time 0.2834 (0.2878) loss 3.4626 (3.3940) grad_norm 1.7887 (1.7374) [2021-04-16 08:13:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][780/1251] eta 0:02:15 lr 0.000374 time 0.2653 (0.2877) loss 3.0833 (3.3947) grad_norm 1.5043 (1.7372) [2021-04-16 08:13:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][790/1251] eta 0:02:12 lr 0.000374 time 0.2643 (0.2875) loss 2.9597 (3.3913) grad_norm 1.4827 (1.7357) [2021-04-16 08:14:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][800/1251] eta 0:02:09 lr 0.000374 time 0.2788 (0.2874) loss 3.6784 (3.3922) grad_norm 1.7629 (1.7351) [2021-04-16 08:14:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][810/1251] eta 0:02:06 lr 0.000374 time 0.2816 (0.2873) loss 3.1230 (3.3902) grad_norm 2.0728 (1.7368) [2021-04-16 08:14:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][820/1251] eta 0:02:03 lr 0.000374 time 0.2696 (0.2872) loss 3.4632 (3.3940) grad_norm 2.2990 (1.7377) [2021-04-16 08:14:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][830/1251] eta 0:02:00 lr 0.000374 time 0.2463 (0.2871) loss 2.8988 (3.3914) grad_norm 1.6570 (1.7363) [2021-04-16 08:14:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][840/1251] eta 0:01:57 lr 0.000374 time 0.3117 (0.2871) loss 3.8348 (3.3921) grad_norm 1.5336 (1.7353) [2021-04-16 08:14:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][850/1251] eta 0:01:55 lr 0.000373 time 0.2651 (0.2869) loss 3.4824 (3.3882) grad_norm 1.7071 (1.7357) [2021-04-16 08:14:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][860/1251] eta 0:01:52 lr 0.000373 time 0.2847 (0.2869) loss 2.8218 (3.3908) grad_norm 1.7234 (1.7369) [2021-04-16 08:14:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][870/1251] eta 0:01:49 lr 0.000373 time 0.3040 (0.2868) loss 3.7274 (3.3906) grad_norm 1.7129 (1.7365) [2021-04-16 08:14:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][880/1251] eta 0:01:46 lr 0.000373 time 0.3091 (0.2867) loss 2.4205 (3.3921) grad_norm 1.9555 (1.7360) [2021-04-16 08:14:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][890/1251] eta 0:01:43 lr 0.000373 time 0.2494 (0.2867) loss 3.5321 (3.3946) grad_norm 1.6003 (1.7365) [2021-04-16 08:14:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][900/1251] eta 0:01:40 lr 0.000373 time 0.2776 (0.2867) loss 3.8421 (3.3969) grad_norm 1.5319 (1.7368) [2021-04-16 08:14:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][910/1251] eta 0:01:37 lr 0.000373 time 0.2751 (0.2865) loss 2.7626 (3.3969) grad_norm 1.6496 (1.7360) [2021-04-16 08:14:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][920/1251] eta 0:01:34 lr 0.000373 time 0.2693 (0.2865) loss 2.8655 (3.3960) grad_norm 1.5228 (nan) [2021-04-16 08:14:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][930/1251] eta 0:01:31 lr 0.000373 time 0.2695 (0.2865) loss 3.9592 (3.3943) grad_norm 2.0267 (nan) [2021-04-16 08:14:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][940/1251] eta 0:01:29 lr 0.000373 time 0.2752 (0.2867) loss 3.9185 (3.3954) grad_norm 1.6737 (nan) [2021-04-16 08:14:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][950/1251] eta 0:01:26 lr 0.000373 time 0.2698 (0.2866) loss 4.1689 (3.3966) grad_norm 2.3995 (nan) [2021-04-16 08:14:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][960/1251] eta 0:01:23 lr 0.000373 time 0.2832 (0.2865) loss 3.6946 (3.3959) grad_norm 2.0183 (nan) [2021-04-16 08:14:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][970/1251] eta 0:01:20 lr 0.000373 time 0.2864 (0.2865) loss 4.0720 (3.3982) grad_norm 2.0546 (nan) [2021-04-16 08:14:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][980/1251] eta 0:01:17 lr 0.000373 time 0.2786 (0.2864) loss 3.6044 (3.3990) grad_norm 2.2081 (nan) [2021-04-16 08:14:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][990/1251] eta 0:01:14 lr 0.000373 time 0.2704 (0.2863) loss 3.5335 (3.3994) grad_norm 2.0863 (nan) [2021-04-16 08:14:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][1000/1251] eta 0:01:11 lr 0.000373 time 0.2615 (0.2863) loss 4.2652 (3.4011) grad_norm 1.7291 (nan) [2021-04-16 08:14:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][1010/1251] eta 0:01:09 lr 0.000373 time 0.2550 (0.2863) loss 2.6509 (3.4003) grad_norm 1.5445 (nan) [2021-04-16 08:15:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][1020/1251] eta 0:01:06 lr 0.000373 time 0.2644 (0.2862) loss 2.0887 (3.3978) grad_norm 1.5781 (nan) [2021-04-16 08:15:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][1030/1251] eta 0:01:03 lr 0.000373 time 0.2947 (0.2863) loss 4.0555 (3.3954) grad_norm 1.6558 (nan) [2021-04-16 08:15:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][1040/1251] eta 0:01:00 lr 0.000373 time 0.2746 (0.2862) loss 3.5469 (3.3951) grad_norm 1.6560 (nan) [2021-04-16 08:15:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][1050/1251] eta 0:00:57 lr 0.000373 time 0.2670 (0.2861) loss 3.7512 (3.3938) grad_norm 1.6419 (nan) [2021-04-16 08:15:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][1060/1251] eta 0:00:54 lr 0.000373 time 0.2552 (0.2860) loss 3.7742 (3.3934) grad_norm 1.6361 (nan) [2021-04-16 08:15:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][1070/1251] eta 0:00:51 lr 0.000373 time 0.2844 (0.2859) loss 2.4449 (3.3896) grad_norm 1.7033 (nan) [2021-04-16 08:15:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][1080/1251] eta 0:00:48 lr 0.000373 time 0.2648 (0.2859) loss 4.1117 (3.3916) grad_norm 1.8718 (nan) [2021-04-16 08:15:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][1090/1251] eta 0:00:46 lr 0.000373 time 0.2822 (0.2858) loss 4.3314 (3.3921) grad_norm 1.7633 (nan) [2021-04-16 08:15:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][1100/1251] eta 0:00:43 lr 0.000372 time 0.2848 (0.2858) loss 3.8798 (3.3926) grad_norm 1.7716 (nan) [2021-04-16 08:15:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][1110/1251] eta 0:00:40 lr 0.000372 time 0.2520 (0.2857) loss 3.4225 (3.3937) grad_norm 1.5109 (nan) [2021-04-16 08:15:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][1120/1251] eta 0:00:37 lr 0.000372 time 0.2706 (0.2857) loss 3.3041 (3.3938) grad_norm 1.6452 (nan) [2021-04-16 08:15:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][1130/1251] eta 0:00:34 lr 0.000372 time 0.2725 (0.2856) loss 2.6400 (3.3951) grad_norm 2.1161 (nan) [2021-04-16 08:15:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][1140/1251] eta 0:00:31 lr 0.000372 time 0.2654 (0.2855) loss 2.6921 (3.3968) grad_norm 1.7277 (nan) [2021-04-16 08:15:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][1150/1251] eta 0:00:28 lr 0.000372 time 0.2564 (0.2857) loss 2.4924 (3.3953) grad_norm 1.5766 (nan) [2021-04-16 08:15:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][1160/1251] eta 0:00:25 lr 0.000372 time 0.3006 (0.2857) loss 3.3314 (3.3934) grad_norm 1.7097 (nan) [2021-04-16 08:15:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][1170/1251] eta 0:00:23 lr 0.000372 time 0.2648 (0.2857) loss 3.7402 (3.3934) grad_norm 1.7043 (nan) [2021-04-16 08:15:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][1180/1251] eta 0:00:20 lr 0.000372 time 0.2757 (0.2856) loss 2.9957 (3.3932) grad_norm 1.8257 (nan) [2021-04-16 08:15:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][1190/1251] eta 0:00:17 lr 0.000372 time 0.3004 (0.2857) loss 4.2839 (3.3969) grad_norm 1.5971 (nan) [2021-04-16 08:15:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][1200/1251] eta 0:00:14 lr 0.000372 time 0.2815 (0.2856) loss 3.5922 (3.3980) grad_norm 1.6062 (nan) [2021-04-16 08:15:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][1210/1251] eta 0:00:11 lr 0.000372 time 0.2802 (0.2856) loss 4.5342 (3.3966) grad_norm 1.8657 (nan) [2021-04-16 08:15:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][1220/1251] eta 0:00:08 lr 0.000372 time 0.2928 (0.2856) loss 4.2907 (3.3963) grad_norm 1.6912 (nan) [2021-04-16 08:16:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][1230/1251] eta 0:00:05 lr 0.000372 time 0.2835 (0.2855) loss 2.7194 (3.3941) grad_norm 1.6242 (nan) [2021-04-16 08:16:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][1240/1251] eta 0:00:03 lr 0.000372 time 0.3343 (0.2855) loss 2.8079 (3.3901) grad_norm 1.8762 (nan) [2021-04-16 08:16:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [175/300][1250/1251] eta 0:00:00 lr 0.000372 time 0.2481 (0.2852) loss 3.5605 (3.3889) grad_norm 1.5582 (nan) [2021-04-16 08:16:10 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 175 training takes 0:06:00 [2021-04-16 08:16:10 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_175.pth saving...... [2021-04-16 08:16:27 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_175.pth saved !!! [2021-04-16 08:16:28 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.155 (1.155) Loss 0.9447 (0.9447) Acc@1 78.027 (78.027) Acc@5 94.336 (94.336) [2021-04-16 08:16:29 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.158 (0.224) Loss 0.9948 (0.9794) Acc@1 76.758 (77.362) Acc@5 93.848 (93.910) [2021-04-16 08:16:31 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.563 (0.217) Loss 0.8986 (0.9754) Acc@1 79.395 (77.293) Acc@5 94.629 (94.127) [2021-04-16 08:16:34 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.260 (0.232) Loss 0.9983 (0.9822) Acc@1 75.488 (77.041) Acc@5 94.336 (94.056) [2021-04-16 08:16:36 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.212) Loss 0.9046 (0.9856) Acc@1 77.930 (77.034) Acc@5 95.020 (93.917) [2021-04-16 08:16:39 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 77.146 Acc@5 93.902 [2021-04-16 08:16:39 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 77.1% [2021-04-16 08:16:39 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 77.33% [2021-04-16 08:16:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][0/1251] eta 3:52:46 lr 0.000372 time 11.1642 (11.1642) loss 3.6257 (3.6257) grad_norm 1.9405 (1.9405) [2021-04-16 08:16:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][10/1251] eta 0:26:06 lr 0.000372 time 0.2742 (1.2620) loss 3.8649 (3.1662) grad_norm 1.6623 (1.8031) [2021-04-16 08:16:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][20/1251] eta 0:16:14 lr 0.000372 time 0.2788 (0.7919) loss 3.8252 (3.3243) grad_norm 1.4412 (1.6966) [2021-04-16 08:16:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][30/1251] eta 0:12:44 lr 0.000372 time 0.2741 (0.6259) loss 2.4237 (3.2811) grad_norm 1.5056 (1.7061) [2021-04-16 08:17:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][40/1251] eta 0:10:55 lr 0.000372 time 0.2940 (0.5410) loss 3.4437 (3.3177) grad_norm 1.4881 (1.6827) [2021-04-16 08:17:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][50/1251] eta 0:09:47 lr 0.000372 time 0.2803 (0.4895) loss 3.4489 (3.3258) grad_norm 1.7158 (1.6967) [2021-04-16 08:17:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][60/1251] eta 0:09:01 lr 0.000372 time 0.2806 (0.4551) loss 3.4668 (3.3291) grad_norm 1.8200 (1.7101) [2021-04-16 08:17:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][70/1251] eta 0:08:27 lr 0.000372 time 0.2813 (0.4298) loss 3.6422 (3.3647) grad_norm 1.5892 (1.7202) [2021-04-16 08:17:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][80/1251] eta 0:08:01 lr 0.000372 time 0.2709 (0.4108) loss 3.1518 (3.3593) grad_norm 1.6990 (1.7144) [2021-04-16 08:17:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][90/1251] eta 0:07:42 lr 0.000372 time 0.2746 (0.3980) loss 4.0040 (3.3426) grad_norm 1.5731 (1.7117) [2021-04-16 08:17:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][100/1251] eta 0:07:24 lr 0.000371 time 0.2564 (0.3858) loss 3.2994 (3.3319) grad_norm 2.1900 (1.7204) [2021-04-16 08:17:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][110/1251] eta 0:07:09 lr 0.000371 time 0.2826 (0.3761) loss 3.2287 (3.3332) grad_norm 1.6945 (1.7249) [2021-04-16 08:17:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][120/1251] eta 0:06:55 lr 0.000371 time 0.2721 (0.3676) loss 3.3928 (3.3266) grad_norm 1.9759 (1.7295) [2021-04-16 08:17:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][130/1251] eta 0:06:44 lr 0.000371 time 0.2749 (0.3606) loss 2.9062 (3.3274) grad_norm 1.6619 (1.7242) [2021-04-16 08:17:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][140/1251] eta 0:06:36 lr 0.000371 time 0.2460 (0.3568) loss 3.3464 (3.3186) grad_norm 1.6152 (1.7217) [2021-04-16 08:17:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][150/1251] eta 0:06:28 lr 0.000371 time 0.2549 (0.3527) loss 3.2582 (3.3307) grad_norm 1.8649 (1.7215) [2021-04-16 08:17:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][160/1251] eta 0:06:20 lr 0.000371 time 0.2978 (0.3487) loss 2.9874 (3.3367) grad_norm 1.6225 (1.7191) [2021-04-16 08:17:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][170/1251] eta 0:06:12 lr 0.000371 time 0.2901 (0.3443) loss 3.4038 (3.3488) grad_norm 1.7692 (1.7205) [2021-04-16 08:17:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][180/1251] eta 0:06:04 lr 0.000371 time 0.2830 (0.3406) loss 2.4682 (3.3433) grad_norm 1.8325 (1.7258) [2021-04-16 08:17:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][190/1251] eta 0:05:57 lr 0.000371 time 0.2829 (0.3372) loss 2.6637 (3.3439) grad_norm 1.6637 (1.7279) [2021-04-16 08:17:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][200/1251] eta 0:05:52 lr 0.000371 time 0.2675 (0.3350) loss 2.9836 (3.3448) grad_norm 1.7795 (1.7248) [2021-04-16 08:17:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][210/1251] eta 0:05:45 lr 0.000371 time 0.2931 (0.3322) loss 3.6114 (3.3559) grad_norm 1.7309 (1.7262) [2021-04-16 08:17:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][220/1251] eta 0:05:39 lr 0.000371 time 0.2782 (0.3296) loss 3.0226 (3.3559) grad_norm 1.5489 (1.7264) [2021-04-16 08:17:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][230/1251] eta 0:05:34 lr 0.000371 time 0.2658 (0.3275) loss 2.9604 (3.3567) grad_norm 1.5857 (1.7238) [2021-04-16 08:17:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][240/1251] eta 0:05:28 lr 0.000371 time 0.2946 (0.3252) loss 3.5904 (3.3696) grad_norm 1.7427 (1.7240) [2021-04-16 08:18:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][250/1251] eta 0:05:24 lr 0.000371 time 0.2644 (0.3239) loss 2.6861 (3.3667) grad_norm 1.6305 (1.7213) [2021-04-16 08:18:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][260/1251] eta 0:05:19 lr 0.000371 time 0.2742 (0.3219) loss 2.9719 (3.3725) grad_norm 1.6548 (1.7216) [2021-04-16 08:18:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][270/1251] eta 0:05:14 lr 0.000371 time 0.2831 (0.3205) loss 3.8279 (3.3806) grad_norm 1.6138 (1.7222) [2021-04-16 08:18:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][280/1251] eta 0:05:09 lr 0.000371 time 0.2787 (0.3189) loss 3.0656 (3.3748) grad_norm 1.5770 (1.7227) [2021-04-16 08:18:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][290/1251] eta 0:05:05 lr 0.000371 time 0.2672 (0.3174) loss 3.4173 (3.3713) grad_norm 1.8066 (1.7238) [2021-04-16 08:18:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][300/1251] eta 0:05:00 lr 0.000371 time 0.2894 (0.3164) loss 4.3298 (3.3694) grad_norm 1.7710 (1.7225) [2021-04-16 08:18:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][310/1251] eta 0:04:56 lr 0.000371 time 0.2707 (0.3151) loss 3.3842 (3.3775) grad_norm 1.7834 (1.7236) [2021-04-16 08:18:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][320/1251] eta 0:04:52 lr 0.000371 time 0.2772 (0.3139) loss 3.9632 (3.3725) grad_norm 1.6251 (1.7233) [2021-04-16 08:18:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][330/1251] eta 0:04:48 lr 0.000371 time 0.2553 (0.3129) loss 3.7409 (3.3713) grad_norm 1.5531 (1.7229) [2021-04-16 08:18:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][340/1251] eta 0:04:44 lr 0.000371 time 0.3009 (0.3120) loss 3.8701 (3.3763) grad_norm 1.7887 (1.7235) [2021-04-16 08:18:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][350/1251] eta 0:04:40 lr 0.000370 time 0.3835 (0.3114) loss 3.5899 (3.3798) grad_norm 2.1224 (1.7240) [2021-04-16 08:18:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][360/1251] eta 0:04:37 lr 0.000370 time 0.2720 (0.3111) loss 2.6128 (3.3794) grad_norm 1.5052 (1.7262) [2021-04-16 08:18:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][370/1251] eta 0:04:33 lr 0.000370 time 0.2713 (0.3105) loss 3.3828 (3.3839) grad_norm 1.6574 (1.7286) [2021-04-16 08:18:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][380/1251] eta 0:04:29 lr 0.000370 time 0.2957 (0.3097) loss 2.8175 (3.3854) grad_norm 1.4977 (1.7292) [2021-04-16 08:18:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][390/1251] eta 0:04:26 lr 0.000370 time 0.2906 (0.3093) loss 4.0562 (3.3911) grad_norm 1.5518 (1.7290) [2021-04-16 08:18:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][400/1251] eta 0:04:22 lr 0.000370 time 0.2813 (0.3086) loss 4.1603 (3.3950) grad_norm 1.5736 (1.7279) [2021-04-16 08:18:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][410/1251] eta 0:04:18 lr 0.000370 time 0.2636 (0.3078) loss 4.2824 (3.4002) grad_norm 1.9943 (1.7274) [2021-04-16 08:18:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][420/1251] eta 0:04:15 lr 0.000370 time 0.2482 (0.3070) loss 2.7582 (3.4061) grad_norm 1.6181 (1.7289) [2021-04-16 08:18:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][430/1251] eta 0:04:11 lr 0.000370 time 0.2776 (0.3063) loss 3.6868 (3.3976) grad_norm 1.9504 (1.7298) [2021-04-16 08:18:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][440/1251] eta 0:04:07 lr 0.000370 time 0.2837 (0.3057) loss 2.6723 (3.3979) grad_norm 1.8603 (1.7304) [2021-04-16 08:18:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][450/1251] eta 0:04:04 lr 0.000370 time 0.4131 (0.3053) loss 3.6200 (3.4056) grad_norm 1.8234 (1.7320) [2021-04-16 08:19:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][460/1251] eta 0:04:01 lr 0.000370 time 0.2683 (0.3047) loss 3.3043 (3.4037) grad_norm 1.7767 (1.7313) [2021-04-16 08:19:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][470/1251] eta 0:03:57 lr 0.000370 time 0.2692 (0.3040) loss 2.8959 (3.4048) grad_norm 1.7010 (1.7306) [2021-04-16 08:19:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][480/1251] eta 0:03:53 lr 0.000370 time 0.2826 (0.3034) loss 2.3950 (3.3994) grad_norm 1.5846 (1.7307) [2021-04-16 08:19:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][490/1251] eta 0:03:50 lr 0.000370 time 0.2796 (0.3029) loss 3.3275 (3.4016) grad_norm 1.7129 (1.7309) [2021-04-16 08:19:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][500/1251] eta 0:03:47 lr 0.000370 time 0.2803 (0.3024) loss 3.0749 (3.4007) grad_norm 1.6338 (1.7307) [2021-04-16 08:19:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][510/1251] eta 0:03:43 lr 0.000370 time 0.2729 (0.3018) loss 2.8116 (3.4012) grad_norm 2.0561 (1.7315) [2021-04-16 08:19:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][520/1251] eta 0:03:40 lr 0.000370 time 0.2878 (0.3014) loss 3.4955 (3.3975) grad_norm 1.6948 (1.7337) [2021-04-16 08:19:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][530/1251] eta 0:03:36 lr 0.000370 time 0.2570 (0.3008) loss 4.5799 (3.4004) grad_norm 1.6163 (1.7354) [2021-04-16 08:19:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][540/1251] eta 0:03:33 lr 0.000370 time 0.2700 (0.3007) loss 3.7812 (3.3996) grad_norm 1.8891 (1.7348) [2021-04-16 08:19:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][550/1251] eta 0:03:30 lr 0.000370 time 0.2832 (0.3002) loss 3.0719 (3.4013) grad_norm 1.4752 (1.7347) [2021-04-16 08:19:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][560/1251] eta 0:03:27 lr 0.000370 time 0.3122 (0.2999) loss 2.1091 (3.3999) grad_norm 1.8419 (1.7342) [2021-04-16 08:19:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][570/1251] eta 0:03:23 lr 0.000370 time 0.2589 (0.2995) loss 2.5623 (3.3957) grad_norm 1.6150 (1.7331) [2021-04-16 08:19:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][580/1251] eta 0:03:20 lr 0.000370 time 0.2853 (0.2995) loss 3.4117 (3.3935) grad_norm 1.8450 (1.7344) [2021-04-16 08:19:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][590/1251] eta 0:03:17 lr 0.000370 time 0.2645 (0.2991) loss 3.0605 (3.3972) grad_norm 1.6850 (1.7338) [2021-04-16 08:19:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][600/1251] eta 0:03:14 lr 0.000369 time 0.2531 (0.2987) loss 4.2031 (3.3953) grad_norm 1.8327 (1.7346) [2021-04-16 08:19:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][610/1251] eta 0:03:11 lr 0.000369 time 0.2985 (0.2985) loss 2.8654 (3.3957) grad_norm 1.6834 (1.7346) [2021-04-16 08:19:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][620/1251] eta 0:03:08 lr 0.000369 time 0.2694 (0.2981) loss 4.0465 (3.4004) grad_norm 1.7122 (1.7369) [2021-04-16 08:19:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][630/1251] eta 0:03:04 lr 0.000369 time 0.2878 (0.2978) loss 3.5715 (3.4013) grad_norm 1.4861 (1.7360) [2021-04-16 08:19:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][640/1251] eta 0:03:01 lr 0.000369 time 0.2909 (0.2975) loss 2.4809 (3.3998) grad_norm 1.5543 (1.7356) [2021-04-16 08:19:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][650/1251] eta 0:02:58 lr 0.000369 time 0.2551 (0.2972) loss 3.8367 (3.4048) grad_norm 1.6310 (1.7355) [2021-04-16 08:19:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][660/1251] eta 0:02:55 lr 0.000369 time 0.2916 (0.2969) loss 3.2724 (3.4052) grad_norm 1.7387 (1.7354) [2021-04-16 08:19:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][670/1251] eta 0:02:52 lr 0.000369 time 0.2734 (0.2968) loss 3.2129 (3.4047) grad_norm 1.9369 (1.7351) [2021-04-16 08:20:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][680/1251] eta 0:02:49 lr 0.000369 time 0.2609 (0.2965) loss 3.5866 (3.4032) grad_norm 1.7118 (1.7361) [2021-04-16 08:20:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][690/1251] eta 0:02:46 lr 0.000369 time 0.2797 (0.2962) loss 4.0535 (3.4061) grad_norm 1.7980 (1.7393) [2021-04-16 08:20:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][700/1251] eta 0:02:43 lr 0.000369 time 0.2602 (0.2959) loss 4.0610 (3.4049) grad_norm 1.9916 (1.7396) [2021-04-16 08:20:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][710/1251] eta 0:02:39 lr 0.000369 time 0.2948 (0.2956) loss 3.0972 (3.4054) grad_norm 1.7829 (1.7389) [2021-04-16 08:20:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][720/1251] eta 0:02:37 lr 0.000369 time 0.2729 (0.2958) loss 4.1939 (3.4086) grad_norm 1.4302 (1.7381) [2021-04-16 08:20:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][730/1251] eta 0:02:33 lr 0.000369 time 0.2562 (0.2955) loss 3.5117 (3.4112) grad_norm 1.8329 (1.7386) [2021-04-16 08:20:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][740/1251] eta 0:02:30 lr 0.000369 time 0.2848 (0.2954) loss 3.2433 (3.4124) grad_norm 1.6130 (1.7376) [2021-04-16 08:20:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][750/1251] eta 0:02:27 lr 0.000369 time 0.2917 (0.2951) loss 2.8215 (3.4117) grad_norm 1.8686 (1.7369) [2021-04-16 08:20:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][760/1251] eta 0:02:24 lr 0.000369 time 0.2558 (0.2949) loss 3.9890 (3.4124) grad_norm 1.7563 (1.7360) [2021-04-16 08:20:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][770/1251] eta 0:02:21 lr 0.000369 time 0.2750 (0.2947) loss 3.9311 (3.4122) grad_norm 1.5700 (1.7365) [2021-04-16 08:20:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][780/1251] eta 0:02:18 lr 0.000369 time 0.2905 (0.2947) loss 3.7569 (3.4143) grad_norm 1.7937 (1.7386) [2021-04-16 08:20:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][790/1251] eta 0:02:15 lr 0.000369 time 0.2891 (0.2945) loss 3.0847 (3.4173) grad_norm 1.7928 (1.7400) [2021-04-16 08:20:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][800/1251] eta 0:02:12 lr 0.000369 time 0.2597 (0.2942) loss 3.5758 (3.4167) grad_norm 1.7195 (1.7395) [2021-04-16 08:20:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][810/1251] eta 0:02:09 lr 0.000369 time 0.2893 (0.2939) loss 3.5383 (3.4182) grad_norm 1.4605 (1.7390) [2021-04-16 08:20:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][820/1251] eta 0:02:06 lr 0.000369 time 0.2838 (0.2937) loss 3.5442 (3.4184) grad_norm 1.4496 (1.7392) [2021-04-16 08:20:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][830/1251] eta 0:02:03 lr 0.000369 time 0.2744 (0.2936) loss 2.3258 (3.4147) grad_norm 1.9001 (1.7394) [2021-04-16 08:20:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][840/1251] eta 0:02:00 lr 0.000369 time 0.2726 (0.2935) loss 3.1023 (3.4124) grad_norm 1.7144 (1.7395) [2021-04-16 08:20:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][850/1251] eta 0:01:57 lr 0.000368 time 0.2839 (0.2932) loss 3.7096 (3.4177) grad_norm 1.6398 (1.7399) [2021-04-16 08:20:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][860/1251] eta 0:01:54 lr 0.000368 time 0.2715 (0.2931) loss 1.9997 (3.4182) grad_norm 1.7257 (1.7398) [2021-04-16 08:20:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][870/1251] eta 0:01:51 lr 0.000368 time 0.2803 (0.2929) loss 2.5640 (3.4208) grad_norm 2.1929 (1.7400) [2021-04-16 08:20:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][880/1251] eta 0:01:48 lr 0.000368 time 0.2798 (0.2927) loss 3.3736 (3.4177) grad_norm 1.6460 (1.7402) [2021-04-16 08:21:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][890/1251] eta 0:01:45 lr 0.000368 time 0.2696 (0.2926) loss 3.9860 (3.4187) grad_norm 2.0121 (1.7396) [2021-04-16 08:21:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][900/1251] eta 0:01:42 lr 0.000368 time 0.2662 (0.2924) loss 4.0465 (3.4230) grad_norm 1.6552 (1.7397) [2021-04-16 08:21:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][910/1251] eta 0:01:39 lr 0.000368 time 0.2641 (0.2922) loss 4.2558 (3.4240) grad_norm 2.0190 (1.7410) [2021-04-16 08:21:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][920/1251] eta 0:01:36 lr 0.000368 time 0.3100 (0.2921) loss 3.9683 (3.4251) grad_norm 1.6014 (1.7419) [2021-04-16 08:21:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][930/1251] eta 0:01:33 lr 0.000368 time 0.2855 (0.2920) loss 2.9122 (3.4240) grad_norm 1.7175 (1.7415) [2021-04-16 08:21:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][940/1251] eta 0:01:30 lr 0.000368 time 0.4127 (0.2923) loss 3.2816 (3.4229) grad_norm 1.7351 (1.7411) [2021-04-16 08:21:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][950/1251] eta 0:01:27 lr 0.000368 time 0.2766 (0.2921) loss 3.4818 (3.4236) grad_norm 1.8248 (1.7419) [2021-04-16 08:21:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][960/1251] eta 0:01:24 lr 0.000368 time 0.2584 (0.2919) loss 3.1201 (3.4220) grad_norm 1.7514 (1.7416) [2021-04-16 08:21:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][970/1251] eta 0:01:21 lr 0.000368 time 0.2505 (0.2918) loss 3.9774 (3.4252) grad_norm 1.5525 (1.7420) [2021-04-16 08:21:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][980/1251] eta 0:01:19 lr 0.000368 time 0.3108 (0.2916) loss 3.5113 (3.4263) grad_norm 1.5851 (1.7421) [2021-04-16 08:21:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][990/1251] eta 0:01:16 lr 0.000368 time 0.2427 (0.2914) loss 3.6251 (3.4246) grad_norm 2.1805 (1.7423) [2021-04-16 08:21:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][1000/1251] eta 0:01:13 lr 0.000368 time 0.2764 (0.2913) loss 3.3047 (3.4227) grad_norm 1.8593 (1.7422) [2021-04-16 08:21:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][1010/1251] eta 0:01:10 lr 0.000368 time 0.2556 (0.2911) loss 3.9581 (3.4232) grad_norm 2.1810 (1.7427) [2021-04-16 08:21:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][1020/1251] eta 0:01:07 lr 0.000368 time 0.2496 (0.2910) loss 3.1043 (3.4237) grad_norm 1.4767 (1.7419) [2021-04-16 08:21:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][1030/1251] eta 0:01:04 lr 0.000368 time 0.2817 (0.2910) loss 3.5537 (3.4256) grad_norm 1.9631 (1.7421) [2021-04-16 08:21:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][1040/1251] eta 0:01:01 lr 0.000368 time 0.2704 (0.2908) loss 2.0310 (3.4247) grad_norm 1.6305 (1.7420) [2021-04-16 08:21:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][1050/1251] eta 0:00:58 lr 0.000368 time 0.2535 (0.2907) loss 3.6755 (3.4255) grad_norm 2.3939 (1.7419) [2021-04-16 08:21:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][1060/1251] eta 0:00:55 lr 0.000368 time 0.2725 (0.2905) loss 3.5263 (3.4235) grad_norm 2.2919 (1.7422) [2021-04-16 08:21:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][1070/1251] eta 0:00:52 lr 0.000368 time 0.2768 (0.2904) loss 1.9973 (3.4205) grad_norm 1.6159 (1.7419) [2021-04-16 08:21:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][1080/1251] eta 0:00:49 lr 0.000368 time 0.2696 (0.2903) loss 3.8418 (3.4199) grad_norm 1.6526 (1.7422) [2021-04-16 08:21:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][1090/1251] eta 0:00:46 lr 0.000368 time 0.3100 (0.2902) loss 3.6626 (3.4207) grad_norm 1.7298 (1.7411) [2021-04-16 08:21:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][1100/1251] eta 0:00:43 lr 0.000368 time 0.2806 (0.2900) loss 4.3046 (3.4224) grad_norm 1.7788 (1.7411) [2021-04-16 08:22:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][1110/1251] eta 0:00:40 lr 0.000367 time 0.2850 (0.2899) loss 2.7013 (3.4210) grad_norm 1.7125 (1.7409) [2021-04-16 08:22:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][1120/1251] eta 0:00:37 lr 0.000367 time 0.2802 (0.2899) loss 3.6917 (3.4226) grad_norm 1.7306 (1.7405) [2021-04-16 08:22:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][1130/1251] eta 0:00:35 lr 0.000367 time 0.2617 (0.2898) loss 2.9505 (3.4233) grad_norm 1.9042 (1.7406) [2021-04-16 08:22:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][1140/1251] eta 0:00:32 lr 0.000367 time 0.2611 (0.2898) loss 4.2045 (3.4249) grad_norm 1.7250 (1.7403) [2021-04-16 08:22:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][1150/1251] eta 0:00:29 lr 0.000367 time 0.2659 (0.2897) loss 3.7930 (3.4250) grad_norm 1.6006 (1.7399) [2021-04-16 08:22:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][1160/1251] eta 0:00:26 lr 0.000367 time 0.2886 (0.2897) loss 2.2312 (3.4237) grad_norm 1.8021 (1.7401) [2021-04-16 08:22:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][1170/1251] eta 0:00:23 lr 0.000367 time 0.2556 (0.2896) loss 2.8146 (3.4217) grad_norm 2.0004 (1.7417) [2021-04-16 08:22:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][1180/1251] eta 0:00:20 lr 0.000367 time 0.2679 (0.2895) loss 3.2371 (3.4210) grad_norm 1.5956 (1.7438) [2021-04-16 08:22:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][1190/1251] eta 0:00:17 lr 0.000367 time 0.2937 (0.2894) loss 2.3581 (3.4198) grad_norm 1.7296 (1.7438) [2021-04-16 08:22:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][1200/1251] eta 0:00:14 lr 0.000367 time 0.2713 (0.2893) loss 3.8017 (3.4202) grad_norm 1.7280 (1.7432) [2021-04-16 08:22:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][1210/1251] eta 0:00:11 lr 0.000367 time 0.2713 (0.2891) loss 4.1680 (3.4190) grad_norm 1.6114 (1.7423) [2021-04-16 08:22:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][1220/1251] eta 0:00:08 lr 0.000367 time 0.2780 (0.2890) loss 3.4887 (3.4181) grad_norm 1.7506 (1.7415) [2021-04-16 08:22:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][1230/1251] eta 0:00:06 lr 0.000367 time 0.2855 (0.2890) loss 4.0047 (3.4212) grad_norm 1.9202 (1.7414) [2021-04-16 08:22:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][1240/1251] eta 0:00:03 lr 0.000367 time 0.2974 (0.2888) loss 4.1528 (3.4240) grad_norm 1.6421 (1.7412) [2021-04-16 08:22:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [176/300][1250/1251] eta 0:00:00 lr 0.000367 time 0.2651 (0.2886) loss 3.1200 (3.4247) grad_norm 1.6221 (1.7410) [2021-04-16 08:22:45 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 176 training takes 0:06:05 [2021-04-16 08:22:45 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_176.pth saving...... [2021-04-16 08:22:57 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_176.pth saved !!! [2021-04-16 08:22:58 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.218 (1.218) Loss 0.9567 (0.9567) Acc@1 78.125 (78.125) Acc@5 94.043 (94.043) [2021-04-16 08:23:00 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.428 (0.256) Loss 0.8867 (0.9663) Acc@1 78.516 (77.255) Acc@5 95.605 (93.812) [2021-04-16 08:23:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.144 (0.241) Loss 0.9672 (0.9682) Acc@1 76.270 (77.167) Acc@5 94.824 (93.964) [2021-04-16 08:23:04 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.136 (0.232) Loss 0.9705 (0.9754) Acc@1 77.539 (77.060) Acc@5 93.555 (93.857) [2021-04-16 08:23:06 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.215) Loss 0.9870 (0.9714) Acc@1 77.246 (77.087) Acc@5 93.848 (93.902) [2021-04-16 08:23:13 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 77.120 Acc@5 93.860 [2021-04-16 08:23:13 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 77.1% [2021-04-16 08:23:13 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 77.33% [2021-04-16 08:23:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][0/1251] eta 3:15:12 lr 0.000367 time 9.3627 (9.3627) loss 2.7961 (2.7961) grad_norm 1.8467 (1.8467) [2021-04-16 08:23:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][10/1251] eta 0:22:38 lr 0.000367 time 0.2645 (1.0945) loss 3.8492 (3.2563) grad_norm 1.7510 (1.7574) [2021-04-16 08:23:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][20/1251] eta 0:14:31 lr 0.000367 time 0.3183 (0.7080) loss 3.3930 (3.2847) grad_norm 1.6653 (1.7605) [2021-04-16 08:23:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][30/1251] eta 0:11:38 lr 0.000367 time 0.2795 (0.5720) loss 3.6583 (3.3134) grad_norm 1.5351 (1.7709) [2021-04-16 08:23:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][40/1251] eta 0:10:04 lr 0.000367 time 0.2639 (0.4990) loss 4.1120 (3.3292) grad_norm 1.7452 (1.7497) [2021-04-16 08:23:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][50/1251] eta 0:09:07 lr 0.000367 time 0.2722 (0.4556) loss 2.8925 (3.3578) grad_norm 1.8019 (1.7377) [2021-04-16 08:23:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][60/1251] eta 0:08:29 lr 0.000367 time 0.2826 (0.4275) loss 3.4441 (3.3319) grad_norm 1.7373 (1.7649) [2021-04-16 08:23:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][70/1251] eta 0:07:59 lr 0.000367 time 0.2465 (0.4059) loss 2.0352 (3.3130) grad_norm 1.5576 (1.7523) [2021-04-16 08:23:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][80/1251] eta 0:07:37 lr 0.000367 time 0.2735 (0.3906) loss 3.8875 (3.2884) grad_norm 1.6221 (1.7463) [2021-04-16 08:23:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][90/1251] eta 0:07:18 lr 0.000367 time 0.2554 (0.3781) loss 3.8758 (3.3202) grad_norm 1.5066 (1.7327) [2021-04-16 08:23:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][100/1251] eta 0:07:03 lr 0.000367 time 0.2732 (0.3683) loss 2.7532 (3.3340) grad_norm 1.9822 (1.7388) [2021-04-16 08:23:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][110/1251] eta 0:06:50 lr 0.000366 time 0.2709 (0.3599) loss 2.6547 (3.3504) grad_norm 1.6531 (1.7367) [2021-04-16 08:23:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][120/1251] eta 0:06:39 lr 0.000366 time 0.2621 (0.3534) loss 2.7021 (3.3306) grad_norm 1.9469 (1.7486) [2021-04-16 08:23:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][130/1251] eta 0:06:31 lr 0.000366 time 0.2735 (0.3489) loss 3.9267 (3.3486) grad_norm 1.9870 (1.7531) [2021-04-16 08:24:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][140/1251] eta 0:06:23 lr 0.000366 time 0.2548 (0.3448) loss 3.8499 (3.3572) grad_norm 1.6213 (1.7486) [2021-04-16 08:24:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][150/1251] eta 0:06:17 lr 0.000366 time 0.4131 (0.3426) loss 3.1863 (3.3544) grad_norm 1.6456 (1.7451) [2021-04-16 08:24:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][160/1251] eta 0:06:09 lr 0.000366 time 0.2729 (0.3385) loss 3.2666 (3.3488) grad_norm 2.0401 (1.7424) [2021-04-16 08:24:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][170/1251] eta 0:06:01 lr 0.000366 time 0.2553 (0.3348) loss 3.1874 (3.3530) grad_norm 1.7948 (1.7368) [2021-04-16 08:24:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][180/1251] eta 0:05:55 lr 0.000366 time 0.2591 (0.3323) loss 3.1095 (3.3534) grad_norm 1.5230 (1.7309) [2021-04-16 08:24:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][190/1251] eta 0:05:49 lr 0.000366 time 0.2591 (0.3291) loss 2.2819 (3.3546) grad_norm 1.8721 (1.7335) [2021-04-16 08:24:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][200/1251] eta 0:05:43 lr 0.000366 time 0.2983 (0.3266) loss 3.3653 (3.3552) grad_norm 1.7349 (1.7330) [2021-04-16 08:24:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][210/1251] eta 0:05:37 lr 0.000366 time 0.2788 (0.3241) loss 3.9826 (3.3448) grad_norm 2.2425 (1.7338) [2021-04-16 08:24:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][220/1251] eta 0:05:32 lr 0.000366 time 0.2646 (0.3223) loss 3.6887 (3.3376) grad_norm 1.7328 (1.7315) [2021-04-16 08:24:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][230/1251] eta 0:05:27 lr 0.000366 time 0.2784 (0.3206) loss 2.7705 (3.3460) grad_norm 1.6226 (1.7348) [2021-04-16 08:24:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][240/1251] eta 0:05:22 lr 0.000366 time 0.2682 (0.3186) loss 4.3896 (3.3469) grad_norm 1.7428 (1.7331) [2021-04-16 08:24:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][250/1251] eta 0:05:18 lr 0.000366 time 0.2810 (0.3178) loss 3.3750 (3.3529) grad_norm 1.7326 (1.7300) [2021-04-16 08:24:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][260/1251] eta 0:05:13 lr 0.000366 time 0.3017 (0.3161) loss 2.6569 (3.3477) grad_norm 1.7538 (1.7310) [2021-04-16 08:24:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][270/1251] eta 0:05:08 lr 0.000366 time 0.2865 (0.3147) loss 2.6155 (3.3468) grad_norm 1.7027 (1.7343) [2021-04-16 08:24:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][280/1251] eta 0:05:04 lr 0.000366 time 0.2726 (0.3132) loss 3.2287 (3.3437) grad_norm 1.6699 (1.7376) [2021-04-16 08:24:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][290/1251] eta 0:04:59 lr 0.000366 time 0.2732 (0.3121) loss 3.6463 (3.3493) grad_norm 1.7646 (1.7427) [2021-04-16 08:24:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][300/1251] eta 0:04:55 lr 0.000366 time 0.2586 (0.3109) loss 3.4053 (3.3613) grad_norm 1.7445 (1.7418) [2021-04-16 08:24:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][310/1251] eta 0:04:51 lr 0.000366 time 0.2720 (0.3099) loss 3.8309 (3.3712) grad_norm 1.6644 (1.7418) [2021-04-16 08:24:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][320/1251] eta 0:04:48 lr 0.000366 time 0.2710 (0.3095) loss 2.6633 (3.3684) grad_norm 1.7382 (1.7405) [2021-04-16 08:24:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][330/1251] eta 0:04:44 lr 0.000366 time 0.2585 (0.3084) loss 4.1159 (3.3695) grad_norm 1.5847 (1.7392) [2021-04-16 08:24:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][340/1251] eta 0:04:40 lr 0.000366 time 0.2623 (0.3075) loss 3.9747 (3.3692) grad_norm 1.7289 (1.7411) [2021-04-16 08:25:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][350/1251] eta 0:04:36 lr 0.000366 time 0.2983 (0.3072) loss 3.0273 (3.3726) grad_norm 1.4637 (1.7397) [2021-04-16 08:25:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][360/1251] eta 0:04:33 lr 0.000365 time 0.2657 (0.3067) loss 2.9479 (3.3711) grad_norm 1.7702 (1.7414) [2021-04-16 08:25:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][370/1251] eta 0:04:29 lr 0.000365 time 0.2841 (0.3059) loss 3.5556 (3.3729) grad_norm 1.8488 (1.7445) [2021-04-16 08:25:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][380/1251] eta 0:04:25 lr 0.000365 time 0.2954 (0.3051) loss 3.3722 (3.3791) grad_norm 1.6494 (1.7511) [2021-04-16 08:25:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][390/1251] eta 0:04:22 lr 0.000365 time 0.2861 (0.3044) loss 3.1399 (3.3761) grad_norm 1.7601 (1.7495) [2021-04-16 08:25:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][400/1251] eta 0:04:18 lr 0.000365 time 0.2702 (0.3037) loss 3.4115 (3.3788) grad_norm 1.5389 (1.7481) [2021-04-16 08:25:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][410/1251] eta 0:04:14 lr 0.000365 time 0.2767 (0.3031) loss 4.2810 (3.3744) grad_norm 1.9082 (1.7475) [2021-04-16 08:25:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][420/1251] eta 0:04:11 lr 0.000365 time 0.2546 (0.3023) loss 2.7583 (3.3788) grad_norm 1.5953 (1.7526) [2021-04-16 08:25:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][430/1251] eta 0:04:07 lr 0.000365 time 0.2698 (0.3017) loss 2.5135 (3.3814) grad_norm 1.8261 (1.7537) [2021-04-16 08:25:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][440/1251] eta 0:04:04 lr 0.000365 time 0.2659 (0.3014) loss 4.1992 (3.3862) grad_norm 1.6778 (1.7541) [2021-04-16 08:25:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][450/1251] eta 0:04:00 lr 0.000365 time 0.2812 (0.3008) loss 3.7849 (3.3885) grad_norm 1.4435 (1.7517) [2021-04-16 08:25:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][460/1251] eta 0:03:57 lr 0.000365 time 0.2581 (0.3003) loss 4.1014 (3.3798) grad_norm 1.6404 (1.7511) [2021-04-16 08:25:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][470/1251] eta 0:03:54 lr 0.000365 time 0.2869 (0.2997) loss 3.7857 (3.3834) grad_norm 1.8571 (1.7526) [2021-04-16 08:25:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][480/1251] eta 0:03:50 lr 0.000365 time 0.2939 (0.2994) loss 3.6140 (3.3815) grad_norm 1.6133 (1.7515) [2021-04-16 08:25:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][490/1251] eta 0:03:47 lr 0.000365 time 0.2852 (0.2989) loss 2.5026 (3.3838) grad_norm 1.7051 (1.7490) [2021-04-16 08:25:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][500/1251] eta 0:03:44 lr 0.000365 time 0.2645 (0.2984) loss 3.7600 (3.3859) grad_norm 1.7181 (1.7477) [2021-04-16 08:25:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][510/1251] eta 0:03:40 lr 0.000365 time 0.2744 (0.2981) loss 3.6397 (3.3926) grad_norm 1.6590 (1.7479) [2021-04-16 08:25:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][520/1251] eta 0:03:37 lr 0.000365 time 0.2503 (0.2977) loss 3.4525 (3.3974) grad_norm 1.9071 (1.7473) [2021-04-16 08:25:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][530/1251] eta 0:03:34 lr 0.000365 time 0.2881 (0.2973) loss 3.1918 (3.4006) grad_norm 1.9383 (1.7496) [2021-04-16 08:25:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][540/1251] eta 0:03:31 lr 0.000365 time 0.2712 (0.2969) loss 3.9420 (3.4085) grad_norm 1.6685 (1.7489) [2021-04-16 08:25:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][550/1251] eta 0:03:27 lr 0.000365 time 0.3975 (0.2967) loss 4.2244 (3.4091) grad_norm 2.7258 (1.7486) [2021-04-16 08:25:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][560/1251] eta 0:03:24 lr 0.000365 time 0.2766 (0.2963) loss 3.9175 (3.4055) grad_norm 1.7567 (1.7482) [2021-04-16 08:26:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][570/1251] eta 0:03:21 lr 0.000365 time 0.2732 (0.2960) loss 2.6660 (3.4046) grad_norm 1.5924 (1.7471) [2021-04-16 08:26:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][580/1251] eta 0:03:18 lr 0.000365 time 0.2575 (0.2963) loss 4.1166 (3.4035) grad_norm 1.7133 (1.7462) [2021-04-16 08:26:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][590/1251] eta 0:03:15 lr 0.000365 time 0.2844 (0.2959) loss 3.7183 (3.4060) grad_norm 1.6253 (nan) [2021-04-16 08:26:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][600/1251] eta 0:03:12 lr 0.000365 time 0.2990 (0.2956) loss 2.5064 (3.4052) grad_norm 2.0587 (nan) [2021-04-16 08:26:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][610/1251] eta 0:03:09 lr 0.000364 time 0.2623 (0.2955) loss 2.4725 (3.4044) grad_norm 1.5599 (nan) [2021-04-16 08:26:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][620/1251] eta 0:03:06 lr 0.000364 time 0.2727 (0.2952) loss 3.9920 (3.4094) grad_norm 1.5185 (nan) [2021-04-16 08:26:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][630/1251] eta 0:03:03 lr 0.000364 time 0.3007 (0.2949) loss 3.0119 (3.4078) grad_norm 1.8757 (nan) [2021-04-16 08:26:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][640/1251] eta 0:02:59 lr 0.000364 time 0.2957 (0.2946) loss 3.8092 (3.4076) grad_norm 1.7868 (nan) [2021-04-16 08:26:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][650/1251] eta 0:02:56 lr 0.000364 time 0.2992 (0.2943) loss 4.5899 (3.4103) grad_norm 1.6411 (nan) [2021-04-16 08:26:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][660/1251] eta 0:02:53 lr 0.000364 time 0.2980 (0.2941) loss 4.1656 (3.4101) grad_norm 2.0113 (nan) [2021-04-16 08:26:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][670/1251] eta 0:02:50 lr 0.000364 time 0.2905 (0.2939) loss 4.1320 (3.4118) grad_norm 1.5682 (nan) [2021-04-16 08:26:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][680/1251] eta 0:02:47 lr 0.000364 time 0.2610 (0.2937) loss 3.7314 (3.4106) grad_norm 1.9742 (nan) [2021-04-16 08:26:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][690/1251] eta 0:02:44 lr 0.000364 time 0.2854 (0.2935) loss 3.1556 (3.4136) grad_norm 1.6082 (nan) [2021-04-16 08:26:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][700/1251] eta 0:02:41 lr 0.000364 time 0.2551 (0.2932) loss 3.9973 (3.4173) grad_norm 1.6149 (nan) [2021-04-16 08:26:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][710/1251] eta 0:02:38 lr 0.000364 time 0.2879 (0.2933) loss 3.6093 (3.4192) grad_norm 1.8543 (nan) [2021-04-16 08:26:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][720/1251] eta 0:02:35 lr 0.000364 time 0.2693 (0.2933) loss 3.8996 (3.4265) grad_norm 1.6177 (nan) [2021-04-16 08:26:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][730/1251] eta 0:02:32 lr 0.000364 time 0.2741 (0.2934) loss 2.9336 (3.4209) grad_norm 1.7058 (nan) [2021-04-16 08:26:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][740/1251] eta 0:02:29 lr 0.000364 time 0.2821 (0.2931) loss 2.7814 (3.4206) grad_norm 1.6138 (nan) [2021-04-16 08:26:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][750/1251] eta 0:02:26 lr 0.000364 time 0.3032 (0.2928) loss 3.5137 (3.4185) grad_norm 1.5536 (nan) [2021-04-16 08:26:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][760/1251] eta 0:02:23 lr 0.000364 time 0.2797 (0.2929) loss 3.9809 (3.4215) grad_norm 1.8196 (nan) [2021-04-16 08:26:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][770/1251] eta 0:02:20 lr 0.000364 time 0.2810 (0.2927) loss 3.7030 (3.4232) grad_norm 1.6431 (nan) [2021-04-16 08:27:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][780/1251] eta 0:02:17 lr 0.000364 time 0.2625 (0.2925) loss 3.1905 (3.4221) grad_norm 1.6338 (nan) [2021-04-16 08:27:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][790/1251] eta 0:02:14 lr 0.000364 time 0.2745 (0.2923) loss 3.6157 (3.4206) grad_norm 1.5576 (nan) [2021-04-16 08:27:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][800/1251] eta 0:02:11 lr 0.000364 time 0.2897 (0.2922) loss 3.9269 (3.4209) grad_norm 1.5276 (nan) [2021-04-16 08:27:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][810/1251] eta 0:02:08 lr 0.000364 time 0.2464 (0.2920) loss 3.6257 (3.4234) grad_norm 1.5154 (nan) [2021-04-16 08:27:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][820/1251] eta 0:02:05 lr 0.000364 time 0.2844 (0.2918) loss 3.7937 (3.4237) grad_norm 1.5163 (nan) [2021-04-16 08:27:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][830/1251] eta 0:02:02 lr 0.000364 time 0.2755 (0.2918) loss 3.9639 (3.4218) grad_norm 1.6671 (nan) [2021-04-16 08:27:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][840/1251] eta 0:01:59 lr 0.000364 time 0.2777 (0.2916) loss 3.7391 (3.4201) grad_norm 1.8132 (nan) [2021-04-16 08:27:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][850/1251] eta 0:01:56 lr 0.000364 time 0.2894 (0.2915) loss 3.9342 (3.4183) grad_norm 1.6713 (nan) [2021-04-16 08:27:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][860/1251] eta 0:01:53 lr 0.000363 time 0.2557 (0.2913) loss 3.0823 (3.4201) grad_norm 2.0279 (nan) [2021-04-16 08:27:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][870/1251] eta 0:01:50 lr 0.000363 time 0.2773 (0.2911) loss 3.6804 (3.4202) grad_norm 1.7627 (nan) [2021-04-16 08:27:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][880/1251] eta 0:01:47 lr 0.000363 time 0.2601 (0.2909) loss 2.8078 (3.4181) grad_norm 1.7489 (nan) [2021-04-16 08:27:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][890/1251] eta 0:01:45 lr 0.000363 time 0.2907 (0.2909) loss 3.2709 (3.4167) grad_norm 2.0462 (nan) [2021-04-16 08:27:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][900/1251] eta 0:01:42 lr 0.000363 time 0.2879 (0.2908) loss 3.4583 (3.4156) grad_norm 1.5269 (nan) [2021-04-16 08:27:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][910/1251] eta 0:01:39 lr 0.000363 time 0.2692 (0.2906) loss 3.7892 (3.4173) grad_norm 1.8048 (nan) [2021-04-16 08:27:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][920/1251] eta 0:01:36 lr 0.000363 time 0.2616 (0.2904) loss 3.8003 (3.4178) grad_norm 1.8489 (nan) [2021-04-16 08:27:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][930/1251] eta 0:01:33 lr 0.000363 time 0.2619 (0.2904) loss 3.6963 (3.4193) grad_norm 1.7266 (nan) [2021-04-16 08:27:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][940/1251] eta 0:01:30 lr 0.000363 time 0.2854 (0.2904) loss 3.3334 (3.4199) grad_norm 1.5837 (nan) [2021-04-16 08:27:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][950/1251] eta 0:01:27 lr 0.000363 time 0.2800 (0.2903) loss 3.1988 (3.4188) grad_norm 1.7301 (nan) [2021-04-16 08:27:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][960/1251] eta 0:01:24 lr 0.000363 time 0.2437 (0.2901) loss 3.7285 (3.4162) grad_norm 1.6800 (nan) [2021-04-16 08:27:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][970/1251] eta 0:01:21 lr 0.000363 time 0.2611 (0.2899) loss 2.7634 (3.4159) grad_norm 1.7061 (nan) [2021-04-16 08:27:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][980/1251] eta 0:01:18 lr 0.000363 time 0.3007 (0.2898) loss 4.0297 (3.4118) grad_norm 1.7165 (nan) [2021-04-16 08:28:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][990/1251] eta 0:01:15 lr 0.000363 time 0.2875 (0.2897) loss 2.8907 (3.4135) grad_norm 1.7708 (nan) [2021-04-16 08:28:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][1000/1251] eta 0:01:12 lr 0.000363 time 0.2820 (0.2896) loss 2.7094 (3.4139) grad_norm 1.7866 (nan) [2021-04-16 08:28:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][1010/1251] eta 0:01:09 lr 0.000363 time 0.2803 (0.2895) loss 3.8217 (3.4149) grad_norm 1.7296 (nan) [2021-04-16 08:28:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][1020/1251] eta 0:01:06 lr 0.000363 time 0.2832 (0.2894) loss 3.7460 (3.4161) grad_norm 2.1192 (nan) [2021-04-16 08:28:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][1030/1251] eta 0:01:03 lr 0.000363 time 0.2922 (0.2893) loss 3.7724 (3.4180) grad_norm 1.5404 (nan) [2021-04-16 08:28:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][1040/1251] eta 0:01:01 lr 0.000363 time 0.2794 (0.2892) loss 3.6773 (3.4193) grad_norm 1.7686 (nan) [2021-04-16 08:28:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][1050/1251] eta 0:00:58 lr 0.000363 time 0.2778 (0.2890) loss 3.2180 (3.4199) grad_norm 1.8150 (nan) [2021-04-16 08:28:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][1060/1251] eta 0:00:55 lr 0.000363 time 0.2916 (0.2889) loss 3.8864 (3.4200) grad_norm 1.7192 (nan) [2021-04-16 08:28:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][1070/1251] eta 0:00:52 lr 0.000363 time 0.2844 (0.2888) loss 3.1594 (3.4208) grad_norm 2.6184 (nan) [2021-04-16 08:28:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][1080/1251] eta 0:00:49 lr 0.000363 time 0.2820 (0.2887) loss 3.3768 (3.4208) grad_norm 1.5694 (nan) [2021-04-16 08:28:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][1090/1251] eta 0:00:46 lr 0.000363 time 0.2628 (0.2887) loss 2.7922 (3.4202) grad_norm 1.9294 (nan) [2021-04-16 08:28:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][1100/1251] eta 0:00:43 lr 0.000363 time 0.2745 (0.2885) loss 4.3289 (3.4206) grad_norm 1.7674 (nan) [2021-04-16 08:28:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][1110/1251] eta 0:00:40 lr 0.000362 time 0.2593 (0.2884) loss 3.5768 (3.4185) grad_norm 1.6906 (nan) [2021-04-16 08:28:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][1120/1251] eta 0:00:37 lr 0.000362 time 0.2774 (0.2883) loss 3.8000 (3.4203) grad_norm 1.5703 (nan) [2021-04-16 08:28:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][1130/1251] eta 0:00:34 lr 0.000362 time 0.2611 (0.2883) loss 2.8388 (3.4192) grad_norm 1.6249 (nan) [2021-04-16 08:28:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][1140/1251] eta 0:00:31 lr 0.000362 time 0.2660 (0.2882) loss 4.2214 (3.4228) grad_norm 1.5911 (nan) [2021-04-16 08:28:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][1150/1251] eta 0:00:29 lr 0.000362 time 0.2671 (0.2882) loss 3.8527 (3.4234) grad_norm 1.5766 (nan) [2021-04-16 08:28:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][1160/1251] eta 0:00:26 lr 0.000362 time 0.2595 (0.2883) loss 3.6912 (3.4244) grad_norm 1.6245 (nan) [2021-04-16 08:28:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][1170/1251] eta 0:00:23 lr 0.000362 time 0.2581 (0.2882) loss 4.1921 (3.4272) grad_norm 1.8612 (nan) [2021-04-16 08:28:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][1180/1251] eta 0:00:20 lr 0.000362 time 0.2821 (0.2882) loss 3.3715 (3.4274) grad_norm 1.4820 (nan) [2021-04-16 08:28:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][1190/1251] eta 0:00:17 lr 0.000362 time 0.3022 (0.2881) loss 2.9836 (3.4289) grad_norm 1.7222 (nan) [2021-04-16 08:28:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][1200/1251] eta 0:00:14 lr 0.000362 time 0.2848 (0.2879) loss 3.8944 (3.4300) grad_norm 1.9090 (nan) [2021-04-16 08:29:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][1210/1251] eta 0:00:11 lr 0.000362 time 0.2689 (0.2878) loss 3.3648 (3.4298) grad_norm 2.1892 (nan) [2021-04-16 08:29:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][1220/1251] eta 0:00:08 lr 0.000362 time 0.2880 (0.2877) loss 3.7490 (3.4289) grad_norm 1.7462 (nan) [2021-04-16 08:29:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][1230/1251] eta 0:00:06 lr 0.000362 time 0.2710 (0.2876) loss 3.0819 (3.4269) grad_norm 1.7798 (nan) [2021-04-16 08:29:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][1240/1251] eta 0:00:03 lr 0.000362 time 0.2485 (0.2875) loss 3.7648 (3.4266) grad_norm 1.7761 (nan) [2021-04-16 08:29:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [177/300][1250/1251] eta 0:00:00 lr 0.000362 time 0.2486 (0.2872) loss 3.8236 (3.4308) grad_norm 1.7048 (nan) [2021-04-16 08:29:15 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 177 training takes 0:06:02 [2021-04-16 08:29:15 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_177.pth saving...... [2021-04-16 08:29:36 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_177.pth saved !!! [2021-04-16 08:29:37 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.112 (1.112) Loss 0.9293 (0.9293) Acc@1 78.320 (78.320) Acc@5 94.043 (94.043) [2021-04-16 08:29:38 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.390 (0.245) Loss 0.9232 (0.9309) Acc@1 78.320 (77.708) Acc@5 94.629 (94.016) [2021-04-16 08:29:40 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.350 (0.219) Loss 1.0080 (0.9558) Acc@1 75.977 (77.307) Acc@5 93.555 (93.797) [2021-04-16 08:29:43 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.293 (0.236) Loss 0.9677 (0.9577) Acc@1 77.441 (77.252) Acc@5 94.141 (93.835) [2021-04-16 08:29:45 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.216) Loss 0.9762 (0.9570) Acc@1 76.465 (77.282) Acc@5 93.555 (93.826) [2021-04-16 08:29:49 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 77.166 Acc@5 93.824 [2021-04-16 08:29:49 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 77.2% [2021-04-16 08:29:49 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 77.33% [2021-04-16 08:29:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][0/1251] eta 2:24:48 lr 0.000362 time 6.9453 (6.9453) loss 3.7465 (3.7465) grad_norm 1.7249 (1.7249) [2021-04-16 08:29:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][10/1251] eta 0:18:17 lr 0.000362 time 0.2864 (0.8847) loss 3.9132 (3.4401) grad_norm 2.1522 (1.7147) [2021-04-16 08:30:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][20/1251] eta 0:12:16 lr 0.000362 time 0.2898 (0.5979) loss 3.8320 (3.5373) grad_norm 1.6775 (1.7001) [2021-04-16 08:30:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][30/1251] eta 0:10:05 lr 0.000362 time 0.2721 (0.4957) loss 4.6561 (3.5402) grad_norm 1.8788 (1.7400) [2021-04-16 08:30:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][40/1251] eta 0:08:58 lr 0.000362 time 0.2695 (0.4451) loss 3.2782 (3.5043) grad_norm 1.4728 (1.7539) [2021-04-16 08:30:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][50/1251] eta 0:08:16 lr 0.000362 time 0.2720 (0.4130) loss 3.9402 (3.4983) grad_norm 1.6065 (1.7437) [2021-04-16 08:30:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][60/1251] eta 0:07:46 lr 0.000362 time 0.2667 (0.3913) loss 3.2111 (3.4758) grad_norm 2.0264 (1.7533) [2021-04-16 08:30:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][70/1251] eta 0:07:22 lr 0.000362 time 0.2881 (0.3750) loss 3.4660 (3.4874) grad_norm 1.6045 (1.7704) [2021-04-16 08:30:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][80/1251] eta 0:07:04 lr 0.000362 time 0.2758 (0.3628) loss 4.1600 (3.4642) grad_norm 1.6705 (1.7663) [2021-04-16 08:30:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][90/1251] eta 0:06:52 lr 0.000362 time 0.2964 (0.3552) loss 3.2738 (3.4571) grad_norm 2.4373 (1.7831) [2021-04-16 08:30:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][100/1251] eta 0:06:39 lr 0.000362 time 0.2828 (0.3472) loss 2.6531 (3.4579) grad_norm 1.5870 (1.7735) [2021-04-16 08:30:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][110/1251] eta 0:06:29 lr 0.000361 time 0.2799 (0.3411) loss 3.7629 (3.4667) grad_norm 1.7473 (1.7739) [2021-04-16 08:30:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][120/1251] eta 0:06:21 lr 0.000361 time 0.2977 (0.3370) loss 3.9261 (3.4616) grad_norm 1.8472 (1.7857) [2021-04-16 08:30:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][130/1251] eta 0:06:12 lr 0.000361 time 0.2833 (0.3323) loss 3.7865 (3.4666) grad_norm 1.8606 (1.7851) [2021-04-16 08:30:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][140/1251] eta 0:06:05 lr 0.000361 time 0.2851 (0.3294) loss 3.9872 (3.4717) grad_norm 1.5993 (1.7740) [2021-04-16 08:30:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][150/1251] eta 0:05:59 lr 0.000361 time 0.2754 (0.3262) loss 3.9060 (3.4631) grad_norm 1.8620 (1.7702) [2021-04-16 08:30:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][160/1251] eta 0:05:53 lr 0.000361 time 0.2830 (0.3240) loss 2.6178 (3.4419) grad_norm 1.6890 (1.7672) [2021-04-16 08:30:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][170/1251] eta 0:05:47 lr 0.000361 time 0.2927 (0.3214) loss 2.8482 (3.4247) grad_norm 1.5090 (1.7658) [2021-04-16 08:30:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][180/1251] eta 0:05:41 lr 0.000361 time 0.2582 (0.3187) loss 4.0073 (3.4275) grad_norm 1.8715 (1.7613) [2021-04-16 08:30:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][190/1251] eta 0:05:36 lr 0.000361 time 0.2918 (0.3169) loss 4.2823 (3.4191) grad_norm 1.8064 (1.7610) [2021-04-16 08:30:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][200/1251] eta 0:05:30 lr 0.000361 time 0.2593 (0.3148) loss 3.3970 (3.4319) grad_norm 1.8317 (1.7603) [2021-04-16 08:30:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][210/1251] eta 0:05:26 lr 0.000361 time 0.2861 (0.3132) loss 3.5125 (3.4318) grad_norm 1.7905 (1.7598) [2021-04-16 08:30:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][220/1251] eta 0:05:21 lr 0.000361 time 0.2971 (0.3115) loss 4.3441 (3.4326) grad_norm 1.5622 (1.7579) [2021-04-16 08:31:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][230/1251] eta 0:05:16 lr 0.000361 time 0.2973 (0.3100) loss 2.9811 (3.4256) grad_norm 1.8283 (1.7591) [2021-04-16 08:31:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][240/1251] eta 0:05:12 lr 0.000361 time 0.2675 (0.3091) loss 3.0808 (3.4334) grad_norm 1.4585 (1.7564) [2021-04-16 08:31:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][250/1251] eta 0:05:08 lr 0.000361 time 0.2612 (0.3078) loss 4.1599 (3.4415) grad_norm 1.6206 (1.7522) [2021-04-16 08:31:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][260/1251] eta 0:05:03 lr 0.000361 time 0.2681 (0.3066) loss 3.5562 (3.4379) grad_norm 1.6605 (1.7520) [2021-04-16 08:31:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][270/1251] eta 0:04:59 lr 0.000361 time 0.2638 (0.3053) loss 3.0324 (3.4385) grad_norm 1.8504 (1.7561) [2021-04-16 08:31:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][280/1251] eta 0:04:55 lr 0.000361 time 0.2767 (0.3044) loss 3.2555 (3.4299) grad_norm 1.9283 (1.7577) [2021-04-16 08:31:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][290/1251] eta 0:04:51 lr 0.000361 time 0.2594 (0.3034) loss 3.4144 (3.4151) grad_norm 1.8888 (1.7580) [2021-04-16 08:31:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][300/1251] eta 0:04:47 lr 0.000361 time 0.2962 (0.3025) loss 2.4471 (3.4108) grad_norm 1.7506 (1.7561) [2021-04-16 08:31:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][310/1251] eta 0:04:43 lr 0.000361 time 0.2680 (0.3017) loss 3.9279 (3.4207) grad_norm 1.6153 (1.7533) [2021-04-16 08:31:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][320/1251] eta 0:04:40 lr 0.000361 time 0.2674 (0.3010) loss 3.3812 (3.4294) grad_norm 1.5281 (1.7556) [2021-04-16 08:31:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][330/1251] eta 0:04:36 lr 0.000361 time 0.2687 (0.3005) loss 3.8549 (3.4346) grad_norm 1.7164 (1.7531) [2021-04-16 08:31:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][340/1251] eta 0:04:33 lr 0.000361 time 0.2786 (0.3004) loss 3.5814 (3.4288) grad_norm 1.9858 (1.7527) [2021-04-16 08:31:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][350/1251] eta 0:04:29 lr 0.000361 time 0.2627 (0.2997) loss 3.6859 (3.4269) grad_norm 1.5422 (1.7514) [2021-04-16 08:31:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][360/1251] eta 0:04:26 lr 0.000361 time 0.2738 (0.2997) loss 2.9571 (3.4292) grad_norm 1.6839 (1.7542) [2021-04-16 08:31:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][370/1251] eta 0:04:23 lr 0.000360 time 0.3106 (0.2995) loss 3.4626 (3.4318) grad_norm 1.9728 (1.7537) [2021-04-16 08:31:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][380/1251] eta 0:04:20 lr 0.000360 time 0.2642 (0.2992) loss 2.6635 (3.4348) grad_norm 1.5100 (1.7518) [2021-04-16 08:31:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][390/1251] eta 0:04:17 lr 0.000360 time 0.2560 (0.2988) loss 2.3579 (3.4313) grad_norm 1.6405 (1.7509) [2021-04-16 08:31:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][400/1251] eta 0:04:13 lr 0.000360 time 0.2772 (0.2984) loss 3.3254 (3.4340) grad_norm 1.5370 (1.7499) [2021-04-16 08:31:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][410/1251] eta 0:04:10 lr 0.000360 time 0.2814 (0.2979) loss 2.7165 (3.4290) grad_norm 1.7545 (1.7523) [2021-04-16 08:31:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][420/1251] eta 0:04:07 lr 0.000360 time 0.2808 (0.2973) loss 4.3748 (3.4306) grad_norm 1.7218 (1.7558) [2021-04-16 08:31:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][430/1251] eta 0:04:03 lr 0.000360 time 0.2859 (0.2969) loss 3.8909 (3.4321) grad_norm 1.9603 (1.7547) [2021-04-16 08:31:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][440/1251] eta 0:04:00 lr 0.000360 time 0.2771 (0.2964) loss 3.8660 (3.4312) grad_norm 1.6405 (1.7546) [2021-04-16 08:32:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][450/1251] eta 0:03:57 lr 0.000360 time 0.2863 (0.2959) loss 3.3256 (3.4347) grad_norm 1.7992 (1.7556) [2021-04-16 08:32:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][460/1251] eta 0:03:53 lr 0.000360 time 0.2762 (0.2955) loss 3.5935 (3.4392) grad_norm 1.7832 (1.7555) [2021-04-16 08:32:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][470/1251] eta 0:03:50 lr 0.000360 time 0.2883 (0.2951) loss 2.3006 (3.4290) grad_norm 1.5067 (1.7538) [2021-04-16 08:32:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][480/1251] eta 0:03:47 lr 0.000360 time 0.2857 (0.2948) loss 3.2967 (3.4321) grad_norm 1.7110 (1.7560) [2021-04-16 08:32:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][490/1251] eta 0:03:44 lr 0.000360 time 0.2750 (0.2945) loss 3.3846 (3.4292) grad_norm 1.4277 (1.7590) [2021-04-16 08:32:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][500/1251] eta 0:03:40 lr 0.000360 time 0.2639 (0.2941) loss 2.5072 (3.4304) grad_norm 1.8009 (1.7584) [2021-04-16 08:32:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][510/1251] eta 0:03:37 lr 0.000360 time 0.2620 (0.2939) loss 3.6027 (3.4310) grad_norm 1.7487 (1.7570) [2021-04-16 08:32:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][520/1251] eta 0:03:34 lr 0.000360 time 0.3024 (0.2936) loss 3.3890 (3.4313) grad_norm 1.5692 (1.7550) [2021-04-16 08:32:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][530/1251] eta 0:03:31 lr 0.000360 time 0.2975 (0.2933) loss 3.3637 (3.4375) grad_norm 1.6732 (1.7544) [2021-04-16 08:32:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][540/1251] eta 0:03:28 lr 0.000360 time 0.2596 (0.2932) loss 3.5162 (3.4408) grad_norm 1.9401 (1.7541) [2021-04-16 08:32:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][550/1251] eta 0:03:25 lr 0.000360 time 0.2505 (0.2932) loss 3.5808 (3.4448) grad_norm 1.5654 (1.7530) [2021-04-16 08:32:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][560/1251] eta 0:03:22 lr 0.000360 time 0.2629 (0.2928) loss 2.9563 (3.4446) grad_norm 2.1425 (1.7525) [2021-04-16 08:32:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][570/1251] eta 0:03:19 lr 0.000360 time 0.2732 (0.2929) loss 2.4737 (3.4462) grad_norm 1.7958 (1.7519) [2021-04-16 08:32:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][580/1251] eta 0:03:16 lr 0.000360 time 0.2890 (0.2929) loss 2.9380 (3.4450) grad_norm 1.8444 (1.7529) [2021-04-16 08:32:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][590/1251] eta 0:03:13 lr 0.000360 time 0.2764 (0.2926) loss 2.7305 (3.4442) grad_norm 1.7773 (1.7530) [2021-04-16 08:32:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][600/1251] eta 0:03:10 lr 0.000360 time 0.2825 (0.2925) loss 3.1999 (3.4479) grad_norm 2.0500 (1.7535) [2021-04-16 08:32:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][610/1251] eta 0:03:07 lr 0.000360 time 0.2903 (0.2922) loss 4.1303 (3.4484) grad_norm 1.6283 (1.7531) [2021-04-16 08:32:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][620/1251] eta 0:03:04 lr 0.000359 time 0.2746 (0.2921) loss 3.6162 (3.4439) grad_norm 1.8199 (1.7537) [2021-04-16 08:32:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][630/1251] eta 0:03:01 lr 0.000359 time 0.2982 (0.2919) loss 4.4936 (3.4463) grad_norm 1.6831 (1.7545) [2021-04-16 08:32:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][640/1251] eta 0:02:58 lr 0.000359 time 0.2602 (0.2917) loss 3.2802 (3.4468) grad_norm 1.5363 (1.7536) [2021-04-16 08:32:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][650/1251] eta 0:02:55 lr 0.000359 time 0.3074 (0.2916) loss 3.1034 (3.4453) grad_norm 2.0049 (1.7537) [2021-04-16 08:33:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][660/1251] eta 0:02:52 lr 0.000359 time 0.2791 (0.2914) loss 3.8356 (3.4458) grad_norm 1.7913 (1.7551) [2021-04-16 08:33:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][670/1251] eta 0:02:49 lr 0.000359 time 0.3269 (0.2914) loss 2.8803 (3.4440) grad_norm 1.6947 (1.7562) [2021-04-16 08:33:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][680/1251] eta 0:02:46 lr 0.000359 time 0.2994 (0.2912) loss 3.6871 (3.4444) grad_norm 1.6108 (1.7544) [2021-04-16 08:33:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][690/1251] eta 0:02:43 lr 0.000359 time 0.2785 (0.2911) loss 3.4042 (3.4409) grad_norm 1.8600 (1.7545) [2021-04-16 08:33:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][700/1251] eta 0:02:40 lr 0.000359 time 0.2985 (0.2909) loss 3.9780 (3.4418) grad_norm 1.6513 (1.7531) [2021-04-16 08:33:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][710/1251] eta 0:02:37 lr 0.000359 time 0.3761 (0.2909) loss 3.6346 (3.4370) grad_norm 1.5871 (1.7519) [2021-04-16 08:33:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][720/1251] eta 0:02:34 lr 0.000359 time 0.2500 (0.2906) loss 2.8745 (3.4368) grad_norm 1.4927 (1.7514) [2021-04-16 08:33:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][730/1251] eta 0:02:31 lr 0.000359 time 0.2558 (0.2906) loss 4.1016 (3.4378) grad_norm 1.5148 (1.7509) [2021-04-16 08:33:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][740/1251] eta 0:02:28 lr 0.000359 time 0.2799 (0.2904) loss 3.2171 (3.4372) grad_norm 1.7831 (1.7502) [2021-04-16 08:33:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][750/1251] eta 0:02:25 lr 0.000359 time 0.2681 (0.2902) loss 3.4579 (3.4375) grad_norm 1.5210 (1.7501) [2021-04-16 08:33:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][760/1251] eta 0:02:22 lr 0.000359 time 0.2756 (0.2901) loss 3.7956 (3.4403) grad_norm 1.5510 (1.7490) [2021-04-16 08:33:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][770/1251] eta 0:02:19 lr 0.000359 time 0.2896 (0.2899) loss 3.5801 (3.4401) grad_norm 1.7362 (1.7481) [2021-04-16 08:33:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][780/1251] eta 0:02:16 lr 0.000359 time 0.2874 (0.2897) loss 3.4833 (3.4398) grad_norm 1.7059 (1.7486) [2021-04-16 08:33:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][790/1251] eta 0:02:13 lr 0.000359 time 0.2548 (0.2895) loss 3.3466 (3.4359) grad_norm 1.8784 (1.7476) [2021-04-16 08:33:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][800/1251] eta 0:02:10 lr 0.000359 time 0.3084 (0.2894) loss 3.7667 (3.4343) grad_norm 2.0260 (1.7495) [2021-04-16 08:33:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][810/1251] eta 0:02:07 lr 0.000359 time 0.2796 (0.2892) loss 3.3248 (3.4327) grad_norm 1.9375 (1.7492) [2021-04-16 08:33:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][820/1251] eta 0:02:04 lr 0.000359 time 0.2914 (0.2892) loss 3.6728 (3.4337) grad_norm 2.0687 (1.7498) [2021-04-16 08:33:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][830/1251] eta 0:02:01 lr 0.000359 time 0.2659 (0.2891) loss 3.7783 (3.4352) grad_norm 1.5690 (1.7498) [2021-04-16 08:33:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][840/1251] eta 0:01:58 lr 0.000359 time 0.2586 (0.2890) loss 4.3789 (3.4360) grad_norm 1.7204 (1.7509) [2021-04-16 08:33:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][850/1251] eta 0:01:55 lr 0.000359 time 0.2890 (0.2888) loss 3.6908 (3.4381) grad_norm 1.7599 (1.7505) [2021-04-16 08:33:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][860/1251] eta 0:01:52 lr 0.000359 time 0.2678 (0.2886) loss 3.6721 (3.4407) grad_norm 1.4848 (1.7496) [2021-04-16 08:34:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][870/1251] eta 0:01:49 lr 0.000358 time 0.2708 (0.2885) loss 4.2111 (3.4437) grad_norm 1.6622 (1.7500) [2021-04-16 08:34:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][880/1251] eta 0:01:47 lr 0.000358 time 0.2988 (0.2884) loss 2.3037 (3.4429) grad_norm 1.5912 (1.7504) [2021-04-16 08:34:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][890/1251] eta 0:01:44 lr 0.000358 time 0.2824 (0.2884) loss 3.9503 (3.4404) grad_norm 1.5348 (1.7509) [2021-04-16 08:34:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][900/1251] eta 0:01:41 lr 0.000358 time 0.2552 (0.2882) loss 3.8490 (3.4392) grad_norm 1.8517 (1.7508) [2021-04-16 08:34:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][910/1251] eta 0:01:38 lr 0.000358 time 0.2688 (0.2881) loss 3.0851 (3.4380) grad_norm 2.1512 (1.7496) [2021-04-16 08:34:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][920/1251] eta 0:01:35 lr 0.000358 time 0.2872 (0.2882) loss 4.0049 (3.4403) grad_norm 1.7834 (1.7494) [2021-04-16 08:34:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][930/1251] eta 0:01:32 lr 0.000358 time 0.3099 (0.2882) loss 3.7865 (3.4391) grad_norm 1.8012 (1.7486) [2021-04-16 08:34:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][940/1251] eta 0:01:29 lr 0.000358 time 0.2584 (0.2883) loss 3.9351 (3.4378) grad_norm 1.8985 (1.7483) [2021-04-16 08:34:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][950/1251] eta 0:01:26 lr 0.000358 time 0.2824 (0.2882) loss 3.7735 (3.4375) grad_norm 1.8787 (1.7488) [2021-04-16 08:34:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][960/1251] eta 0:01:23 lr 0.000358 time 0.2741 (0.2882) loss 3.3175 (3.4365) grad_norm 1.9672 (1.7492) [2021-04-16 08:34:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][970/1251] eta 0:01:20 lr 0.000358 time 0.2822 (0.2881) loss 3.5490 (3.4376) grad_norm 1.7903 (1.7481) [2021-04-16 08:34:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][980/1251] eta 0:01:18 lr 0.000358 time 0.2564 (0.2880) loss 3.3372 (3.4391) grad_norm 1.5797 (1.7475) [2021-04-16 08:34:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][990/1251] eta 0:01:15 lr 0.000358 time 0.2708 (0.2879) loss 3.0715 (3.4383) grad_norm 1.5938 (1.7472) [2021-04-16 08:34:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][1000/1251] eta 0:01:12 lr 0.000358 time 0.2706 (0.2878) loss 2.4324 (3.4374) grad_norm 1.8955 (1.7481) [2021-04-16 08:34:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][1010/1251] eta 0:01:09 lr 0.000358 time 0.2612 (0.2877) loss 3.7683 (3.4358) grad_norm 1.9518 (1.7484) [2021-04-16 08:34:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][1020/1251] eta 0:01:06 lr 0.000358 time 0.2752 (0.2876) loss 3.6597 (3.4341) grad_norm 1.9022 (1.7486) [2021-04-16 08:34:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][1030/1251] eta 0:01:03 lr 0.000358 time 0.2872 (0.2875) loss 3.9215 (3.4352) grad_norm 1.5620 (1.7490) [2021-04-16 08:34:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][1040/1251] eta 0:01:00 lr 0.000358 time 0.2934 (0.2874) loss 3.9211 (3.4372) grad_norm 1.7880 (1.7498) [2021-04-16 08:34:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][1050/1251] eta 0:00:57 lr 0.000358 time 0.2647 (0.2872) loss 3.5636 (3.4374) grad_norm 2.1029 (1.7511) [2021-04-16 08:34:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][1060/1251] eta 0:00:54 lr 0.000358 time 0.2789 (0.2871) loss 3.0580 (3.4378) grad_norm 1.7439 (1.7511) [2021-04-16 08:34:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][1070/1251] eta 0:00:51 lr 0.000358 time 0.2891 (0.2870) loss 3.7663 (3.4371) grad_norm 1.8012 (1.7512) [2021-04-16 08:34:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][1080/1251] eta 0:00:49 lr 0.000358 time 0.2663 (0.2870) loss 3.5224 (3.4346) grad_norm 1.8357 (1.7510) [2021-04-16 08:35:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][1090/1251] eta 0:00:46 lr 0.000358 time 0.2850 (0.2869) loss 3.1251 (3.4322) grad_norm 1.9376 (1.7512) [2021-04-16 08:35:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][1100/1251] eta 0:00:43 lr 0.000358 time 0.2910 (0.2868) loss 3.8731 (3.4339) grad_norm 1.7347 (1.7510) [2021-04-16 08:35:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][1110/1251] eta 0:00:40 lr 0.000358 time 0.2816 (0.2867) loss 3.2033 (3.4341) grad_norm 1.9688 (1.7528) [2021-04-16 08:35:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][1120/1251] eta 0:00:37 lr 0.000357 time 0.2669 (0.2867) loss 2.9886 (3.4329) grad_norm 1.6613 (1.7529) [2021-04-16 08:35:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][1130/1251] eta 0:00:34 lr 0.000357 time 0.2735 (0.2867) loss 3.6482 (3.4332) grad_norm 1.6314 (1.7535) [2021-04-16 08:35:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][1140/1251] eta 0:00:31 lr 0.000357 time 0.2602 (0.2867) loss 2.4640 (3.4309) grad_norm 1.7796 (1.7531) [2021-04-16 08:35:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][1150/1251] eta 0:00:28 lr 0.000357 time 0.3066 (0.2867) loss 3.5820 (3.4299) grad_norm 1.6702 (1.7529) [2021-04-16 08:35:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][1160/1251] eta 0:00:26 lr 0.000357 time 0.2689 (0.2869) loss 3.2659 (3.4277) grad_norm 1.7263 (1.7524) [2021-04-16 08:35:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][1170/1251] eta 0:00:23 lr 0.000357 time 0.2832 (0.2868) loss 4.0153 (3.4287) grad_norm 1.8228 (1.7524) [2021-04-16 08:35:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][1180/1251] eta 0:00:20 lr 0.000357 time 0.2961 (0.2867) loss 3.7209 (3.4270) grad_norm 1.7319 (1.7526) [2021-04-16 08:35:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][1190/1251] eta 0:00:17 lr 0.000357 time 0.4129 (0.2867) loss 3.6639 (3.4275) grad_norm 1.7528 (1.7530) [2021-04-16 08:35:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][1200/1251] eta 0:00:14 lr 0.000357 time 0.2899 (0.2866) loss 3.9789 (3.4276) grad_norm 1.6368 (1.7527) [2021-04-16 08:35:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][1210/1251] eta 0:00:11 lr 0.000357 time 0.2739 (0.2865) loss 3.9280 (3.4283) grad_norm 1.7297 (1.7534) [2021-04-16 08:35:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][1220/1251] eta 0:00:08 lr 0.000357 time 0.2638 (0.2865) loss 3.9643 (3.4293) grad_norm 1.8136 (1.7541) [2021-04-16 08:35:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][1230/1251] eta 0:00:06 lr 0.000357 time 0.2964 (0.2864) loss 3.6990 (3.4280) grad_norm 1.6431 (1.7543) [2021-04-16 08:35:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][1240/1251] eta 0:00:03 lr 0.000357 time 0.3306 (0.2863) loss 3.5616 (3.4270) grad_norm 1.4675 (1.7550) [2021-04-16 08:35:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [178/300][1250/1251] eta 0:00:00 lr 0.000357 time 0.2484 (0.2861) loss 3.1275 (3.4271) grad_norm 1.5327 (1.7545) [2021-04-16 08:35:51 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 178 training takes 0:06:02 [2021-04-16 08:35:51 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_178.pth saving...... [2021-04-16 08:36:09 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_178.pth saved !!! [2021-04-16 08:36:10 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.225 (1.225) Loss 1.0360 (1.0360) Acc@1 75.879 (75.879) Acc@5 92.676 (92.676) [2021-04-16 08:36:11 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.160 (0.239) Loss 0.9975 (0.9895) Acc@1 75.488 (76.855) Acc@5 93.848 (93.768) [2021-04-16 08:36:14 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.269 (0.252) Loss 0.9442 (0.9776) Acc@1 77.832 (76.981) Acc@5 93.555 (93.848) [2021-04-16 08:36:15 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.100 (0.210) Loss 0.9567 (0.9728) Acc@1 77.344 (77.199) Acc@5 94.238 (93.889) [2021-04-16 08:36:17 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.209) Loss 1.0089 (0.9772) Acc@1 76.172 (77.113) Acc@5 93.359 (93.829) [2021-04-16 08:36:26 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 77.054 Acc@5 93.810 [2021-04-16 08:36:26 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 77.1% [2021-04-16 08:36:26 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 77.33% [2021-04-16 08:36:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][0/1251] eta 1:50:01 lr 0.000357 time 5.2769 (5.2769) loss 3.7054 (3.7054) grad_norm 1.8787 (1.8787) [2021-04-16 08:36:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][10/1251] eta 0:15:01 lr 0.000357 time 0.2660 (0.7268) loss 2.9998 (3.5714) grad_norm 1.7183 (1.7128) [2021-04-16 08:36:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][20/1251] eta 0:10:32 lr 0.000357 time 0.2887 (0.5135) loss 3.1316 (3.5755) grad_norm 1.8741 (1.7694) [2021-04-16 08:36:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][30/1251] eta 0:08:58 lr 0.000357 time 0.2679 (0.4407) loss 3.6374 (3.5365) grad_norm 1.7593 (1.7968) [2021-04-16 08:36:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][40/1251] eta 0:08:05 lr 0.000357 time 0.2701 (0.4009) loss 3.3600 (3.4940) grad_norm 1.6384 (1.7962) [2021-04-16 08:36:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][50/1251] eta 0:07:33 lr 0.000357 time 0.2785 (0.3775) loss 4.1849 (3.5336) grad_norm 1.5906 (1.7916) [2021-04-16 08:36:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][60/1251] eta 0:07:09 lr 0.000357 time 0.2539 (0.3609) loss 3.3194 (3.4721) grad_norm 1.6758 (1.7946) [2021-04-16 08:36:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][70/1251] eta 0:06:52 lr 0.000357 time 0.2957 (0.3494) loss 3.6604 (3.4975) grad_norm 1.8296 (1.8053) [2021-04-16 08:36:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][80/1251] eta 0:06:39 lr 0.000357 time 0.3115 (0.3413) loss 2.6117 (3.5244) grad_norm 1.8261 (1.7921) [2021-04-16 08:36:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][90/1251] eta 0:06:27 lr 0.000357 time 0.2709 (0.3341) loss 3.0759 (3.5385) grad_norm 1.8226 (1.7845) [2021-04-16 08:36:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][100/1251] eta 0:06:18 lr 0.000357 time 0.2866 (0.3288) loss 2.8619 (3.5455) grad_norm 1.7824 (1.7800) [2021-04-16 08:37:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][110/1251] eta 0:06:14 lr 0.000357 time 0.2966 (0.3282) loss 3.7027 (3.5280) grad_norm 1.9209 (1.7842) [2021-04-16 08:37:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][120/1251] eta 0:06:08 lr 0.000357 time 0.2906 (0.3254) loss 2.7094 (3.5125) grad_norm 1.6697 (1.7796) [2021-04-16 08:37:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][130/1251] eta 0:06:00 lr 0.000356 time 0.2674 (0.3219) loss 2.7679 (3.5020) grad_norm 1.6539 (1.7811) [2021-04-16 08:37:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][140/1251] eta 0:05:55 lr 0.000356 time 0.2617 (0.3203) loss 3.2715 (3.5003) grad_norm 1.6885 (1.7966) [2021-04-16 08:37:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][150/1251] eta 0:05:50 lr 0.000356 time 0.2972 (0.3187) loss 4.2976 (3.5012) grad_norm 1.7656 (1.7947) [2021-04-16 08:37:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][160/1251] eta 0:05:45 lr 0.000356 time 0.2926 (0.3163) loss 2.5815 (3.5050) grad_norm 1.6295 (1.7877) [2021-04-16 08:37:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][170/1251] eta 0:05:40 lr 0.000356 time 0.2782 (0.3151) loss 3.8524 (3.5033) grad_norm 1.5394 (1.7815) [2021-04-16 08:37:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][180/1251] eta 0:05:35 lr 0.000356 time 0.2895 (0.3137) loss 3.6438 (3.5065) grad_norm 1.6011 (1.7810) [2021-04-16 08:37:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][190/1251] eta 0:05:30 lr 0.000356 time 0.2762 (0.3120) loss 4.1294 (3.5105) grad_norm 1.6759 (1.7788) [2021-04-16 08:37:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][200/1251] eta 0:05:26 lr 0.000356 time 0.2658 (0.3105) loss 3.1285 (3.5067) grad_norm 1.7756 (1.7744) [2021-04-16 08:37:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][210/1251] eta 0:05:21 lr 0.000356 time 0.2692 (0.3089) loss 4.1613 (3.5138) grad_norm 1.7360 (1.7754) [2021-04-16 08:37:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][220/1251] eta 0:05:17 lr 0.000356 time 0.2978 (0.3076) loss 2.7120 (3.5010) grad_norm 1.8782 (1.7735) [2021-04-16 08:37:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][230/1251] eta 0:05:12 lr 0.000356 time 0.2852 (0.3062) loss 4.1739 (3.4981) grad_norm 1.6438 (1.7723) [2021-04-16 08:37:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][240/1251] eta 0:05:08 lr 0.000356 time 0.2708 (0.3052) loss 3.5997 (3.4830) grad_norm 1.8630 (1.7748) [2021-04-16 08:37:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][250/1251] eta 0:05:04 lr 0.000356 time 0.2672 (0.3042) loss 4.2537 (3.4787) grad_norm 1.6171 (1.7783) [2021-04-16 08:37:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][260/1251] eta 0:05:00 lr 0.000356 time 0.2681 (0.3032) loss 3.4330 (3.4792) grad_norm 1.7822 (1.7768) [2021-04-16 08:37:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][270/1251] eta 0:04:56 lr 0.000356 time 0.2709 (0.3020) loss 2.6607 (3.4642) grad_norm 1.6527 (1.7743) [2021-04-16 08:37:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][280/1251] eta 0:04:52 lr 0.000356 time 0.2669 (0.3013) loss 2.7604 (3.4658) grad_norm 1.7203 (1.7720) [2021-04-16 08:37:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][290/1251] eta 0:04:48 lr 0.000356 time 0.2822 (0.3005) loss 3.4474 (3.4619) grad_norm 2.0378 (1.7746) [2021-04-16 08:37:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][300/1251] eta 0:04:45 lr 0.000356 time 0.2728 (0.2998) loss 3.0731 (3.4603) grad_norm 1.8125 (1.7776) [2021-04-16 08:37:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][310/1251] eta 0:04:41 lr 0.000356 time 0.2635 (0.2990) loss 3.5962 (3.4584) grad_norm 2.0229 (1.7762) [2021-04-16 08:38:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][320/1251] eta 0:04:38 lr 0.000356 time 0.2614 (0.2987) loss 3.5842 (3.4519) grad_norm 1.9026 (1.7781) [2021-04-16 08:38:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][330/1251] eta 0:04:34 lr 0.000356 time 0.2524 (0.2979) loss 3.6686 (3.4557) grad_norm 1.6042 (1.7796) [2021-04-16 08:38:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][340/1251] eta 0:04:31 lr 0.000356 time 0.2803 (0.2981) loss 2.7406 (3.4475) grad_norm 1.8915 (1.7781) [2021-04-16 08:38:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][350/1251] eta 0:04:28 lr 0.000356 time 0.2574 (0.2975) loss 3.5977 (3.4534) grad_norm 1.5106 (1.7760) [2021-04-16 08:38:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][360/1251] eta 0:04:25 lr 0.000356 time 0.2815 (0.2979) loss 3.2010 (3.4457) grad_norm 1.7968 (1.7749) [2021-04-16 08:38:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][370/1251] eta 0:04:22 lr 0.000356 time 0.2507 (0.2974) loss 4.4247 (3.4470) grad_norm 1.7455 (1.7724) [2021-04-16 08:38:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][380/1251] eta 0:04:18 lr 0.000355 time 0.2914 (0.2970) loss 1.9825 (3.4465) grad_norm 1.6075 (1.7728) [2021-04-16 08:38:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][390/1251] eta 0:04:15 lr 0.000355 time 0.2930 (0.2969) loss 2.8476 (3.4498) grad_norm 1.7219 (1.7721) [2021-04-16 08:38:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][400/1251] eta 0:04:12 lr 0.000355 time 0.2789 (0.2965) loss 4.0497 (3.4568) grad_norm 1.9652 (1.7736) [2021-04-16 08:38:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][410/1251] eta 0:04:08 lr 0.000355 time 0.2974 (0.2960) loss 3.5342 (3.4594) grad_norm 1.5684 (1.7729) [2021-04-16 08:38:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][420/1251] eta 0:04:05 lr 0.000355 time 0.2803 (0.2957) loss 2.7620 (3.4604) grad_norm 1.8107 (1.7728) [2021-04-16 08:38:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][430/1251] eta 0:04:02 lr 0.000355 time 0.2735 (0.2952) loss 3.8227 (3.4586) grad_norm 1.8874 (1.7729) [2021-04-16 08:38:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][440/1251] eta 0:03:59 lr 0.000355 time 0.2896 (0.2950) loss 3.3136 (3.4575) grad_norm 1.7382 (1.7745) [2021-04-16 08:38:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][450/1251] eta 0:03:56 lr 0.000355 time 0.4068 (0.2949) loss 2.4780 (3.4575) grad_norm 1.8884 (1.7735) [2021-04-16 08:38:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][460/1251] eta 0:03:53 lr 0.000355 time 0.2945 (0.2946) loss 3.7717 (3.4608) grad_norm 1.6450 (1.7718) [2021-04-16 08:38:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][470/1251] eta 0:03:49 lr 0.000355 time 0.2572 (0.2942) loss 2.5327 (3.4567) grad_norm 1.7879 (1.7721) [2021-04-16 08:38:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][480/1251] eta 0:03:46 lr 0.000355 time 0.2599 (0.2939) loss 3.6264 (3.4511) grad_norm 1.9152 (1.7705) [2021-04-16 08:38:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][490/1251] eta 0:03:43 lr 0.000355 time 0.2837 (0.2937) loss 2.0545 (3.4446) grad_norm 1.5594 (1.7711) [2021-04-16 08:38:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][500/1251] eta 0:03:40 lr 0.000355 time 0.2640 (0.2934) loss 2.9953 (3.4448) grad_norm 1.7119 (1.7715) [2021-04-16 08:38:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][510/1251] eta 0:03:37 lr 0.000355 time 0.2866 (0.2931) loss 3.6316 (3.4396) grad_norm 1.6879 (1.7727) [2021-04-16 08:38:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][520/1251] eta 0:03:34 lr 0.000355 time 0.2818 (0.2928) loss 3.2977 (3.4401) grad_norm 1.8859 (1.7733) [2021-04-16 08:39:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][530/1251] eta 0:03:31 lr 0.000355 time 0.2811 (0.2927) loss 2.4586 (3.4369) grad_norm 2.1116 (1.7741) [2021-04-16 08:39:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][540/1251] eta 0:03:28 lr 0.000355 time 0.2634 (0.2926) loss 3.5208 (3.4377) grad_norm 2.0098 (1.7752) [2021-04-16 08:39:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][550/1251] eta 0:03:24 lr 0.000355 time 0.2775 (0.2923) loss 3.9146 (3.4422) grad_norm 1.7313 (1.7746) [2021-04-16 08:39:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][560/1251] eta 0:03:21 lr 0.000355 time 0.2902 (0.2922) loss 3.5210 (3.4408) grad_norm 1.5761 (1.7734) [2021-04-16 08:39:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][570/1251] eta 0:03:18 lr 0.000355 time 0.2611 (0.2921) loss 2.4782 (3.4367) grad_norm 1.6234 (1.7722) [2021-04-16 08:39:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][580/1251] eta 0:03:15 lr 0.000355 time 0.2725 (0.2919) loss 4.1831 (3.4378) grad_norm 1.8725 (1.7730) [2021-04-16 08:39:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][590/1251] eta 0:03:12 lr 0.000355 time 0.2572 (0.2917) loss 3.5435 (3.4387) grad_norm 1.5994 (1.7718) [2021-04-16 08:39:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][600/1251] eta 0:03:09 lr 0.000355 time 0.2944 (0.2914) loss 3.8168 (3.4416) grad_norm 1.5378 (1.7724) [2021-04-16 08:39:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][610/1251] eta 0:03:06 lr 0.000355 time 0.2880 (0.2914) loss 2.6546 (3.4373) grad_norm 1.9309 (1.7733) [2021-04-16 08:39:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][620/1251] eta 0:03:03 lr 0.000355 time 0.2612 (0.2911) loss 3.9659 (3.4368) grad_norm 1.9656 (1.7734) [2021-04-16 08:39:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][630/1251] eta 0:03:00 lr 0.000354 time 0.2910 (0.2910) loss 2.8236 (3.4358) grad_norm 2.0067 (1.7735) [2021-04-16 08:39:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][640/1251] eta 0:02:57 lr 0.000354 time 0.2727 (0.2906) loss 4.0030 (3.4377) grad_norm 1.6931 (inf) [2021-04-16 08:39:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][650/1251] eta 0:02:54 lr 0.000354 time 0.2826 (0.2906) loss 3.6550 (3.4391) grad_norm 2.0243 (inf) [2021-04-16 08:39:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][660/1251] eta 0:02:51 lr 0.000354 time 0.2668 (0.2904) loss 4.7565 (3.4384) grad_norm 2.5446 (inf) [2021-04-16 08:39:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][670/1251] eta 0:02:48 lr 0.000354 time 0.2761 (0.2902) loss 3.7287 (3.4397) grad_norm 1.5986 (inf) [2021-04-16 08:39:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][680/1251] eta 0:02:45 lr 0.000354 time 0.2720 (0.2900) loss 3.1165 (3.4363) grad_norm 1.8323 (inf) [2021-04-16 08:39:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][690/1251] eta 0:02:42 lr 0.000354 time 0.2763 (0.2900) loss 2.7604 (3.4380) grad_norm 1.6675 (inf) [2021-04-16 08:39:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][700/1251] eta 0:02:39 lr 0.000354 time 0.2948 (0.2899) loss 3.8865 (3.4389) grad_norm 1.7885 (inf) [2021-04-16 08:39:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][710/1251] eta 0:02:36 lr 0.000354 time 0.2995 (0.2899) loss 3.4977 (3.4390) grad_norm 1.5357 (inf) [2021-04-16 08:39:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][720/1251] eta 0:02:33 lr 0.000354 time 0.2863 (0.2899) loss 3.4829 (3.4374) grad_norm 1.7437 (inf) [2021-04-16 08:39:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][730/1251] eta 0:02:31 lr 0.000354 time 0.2777 (0.2900) loss 3.8688 (3.4411) grad_norm 1.5765 (inf) [2021-04-16 08:40:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][740/1251] eta 0:02:28 lr 0.000354 time 0.4262 (0.2900) loss 3.9435 (3.4431) grad_norm 1.7314 (inf) [2021-04-16 08:40:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][750/1251] eta 0:02:25 lr 0.000354 time 0.2706 (0.2898) loss 3.3129 (3.4483) grad_norm 1.5862 (inf) [2021-04-16 08:40:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][760/1251] eta 0:02:22 lr 0.000354 time 0.3080 (0.2897) loss 3.9888 (3.4480) grad_norm 1.8940 (inf) [2021-04-16 08:40:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][770/1251] eta 0:02:19 lr 0.000354 time 0.2871 (0.2896) loss 2.9352 (3.4483) grad_norm 1.7466 (inf) [2021-04-16 08:40:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][780/1251] eta 0:02:16 lr 0.000354 time 0.2759 (0.2894) loss 2.5976 (3.4474) grad_norm 1.7646 (inf) [2021-04-16 08:40:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][790/1251] eta 0:02:13 lr 0.000354 time 0.2785 (0.2893) loss 3.9819 (3.4462) grad_norm 1.7457 (inf) [2021-04-16 08:40:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][800/1251] eta 0:02:10 lr 0.000354 time 0.3025 (0.2892) loss 3.2539 (3.4459) grad_norm 1.5548 (inf) [2021-04-16 08:40:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][810/1251] eta 0:02:07 lr 0.000354 time 0.2696 (0.2891) loss 3.2895 (3.4467) grad_norm 1.7544 (inf) [2021-04-16 08:40:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][820/1251] eta 0:02:04 lr 0.000354 time 0.2664 (0.2890) loss 4.2325 (3.4423) grad_norm 1.6100 (inf) [2021-04-16 08:40:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][830/1251] eta 0:02:01 lr 0.000354 time 0.2932 (0.2889) loss 2.3485 (3.4410) grad_norm 1.7679 (inf) [2021-04-16 08:40:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][840/1251] eta 0:01:58 lr 0.000354 time 0.2907 (0.2888) loss 3.7138 (3.4414) grad_norm 2.3326 (inf) [2021-04-16 08:40:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][850/1251] eta 0:01:55 lr 0.000354 time 0.2704 (0.2886) loss 3.6286 (3.4419) grad_norm 2.0988 (inf) [2021-04-16 08:40:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][860/1251] eta 0:01:52 lr 0.000354 time 0.3124 (0.2886) loss 3.0457 (3.4425) grad_norm 1.5955 (inf) [2021-04-16 08:40:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][870/1251] eta 0:01:49 lr 0.000354 time 0.2915 (0.2885) loss 3.7657 (3.4422) grad_norm 1.6063 (inf) [2021-04-16 08:40:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][880/1251] eta 0:01:46 lr 0.000353 time 0.2575 (0.2884) loss 3.1661 (3.4445) grad_norm 2.0626 (inf) [2021-04-16 08:40:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][890/1251] eta 0:01:44 lr 0.000353 time 0.2561 (0.2884) loss 2.9300 (3.4457) grad_norm 1.8056 (inf) [2021-04-16 08:40:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][900/1251] eta 0:01:41 lr 0.000353 time 0.2844 (0.2884) loss 3.4612 (3.4464) grad_norm 1.6493 (inf) [2021-04-16 08:40:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][910/1251] eta 0:01:38 lr 0.000353 time 0.2900 (0.2883) loss 3.6257 (3.4456) grad_norm 1.5441 (inf) [2021-04-16 08:40:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][920/1251] eta 0:01:35 lr 0.000353 time 0.2783 (0.2884) loss 3.7971 (3.4453) grad_norm 1.7116 (inf) [2021-04-16 08:40:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][930/1251] eta 0:01:32 lr 0.000353 time 0.2553 (0.2885) loss 2.5245 (3.4448) grad_norm 1.6144 (inf) [2021-04-16 08:40:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][940/1251] eta 0:01:29 lr 0.000353 time 0.2906 (0.2885) loss 2.2510 (3.4462) grad_norm 1.5551 (inf) [2021-04-16 08:41:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][950/1251] eta 0:01:26 lr 0.000353 time 0.2657 (0.2884) loss 2.2587 (3.4451) grad_norm 1.5970 (inf) [2021-04-16 08:41:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][960/1251] eta 0:01:23 lr 0.000353 time 0.2860 (0.2884) loss 3.7525 (3.4432) grad_norm 2.0873 (inf) [2021-04-16 08:41:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][970/1251] eta 0:01:21 lr 0.000353 time 0.2779 (0.2884) loss 3.1891 (3.4450) grad_norm 1.6130 (inf) [2021-04-16 08:41:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][980/1251] eta 0:01:18 lr 0.000353 time 0.2866 (0.2883) loss 3.3259 (3.4430) grad_norm 1.6718 (inf) [2021-04-16 08:41:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][990/1251] eta 0:01:15 lr 0.000353 time 0.2909 (0.2882) loss 3.0122 (3.4421) grad_norm 1.9630 (inf) [2021-04-16 08:41:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][1000/1251] eta 0:01:12 lr 0.000353 time 0.2779 (0.2881) loss 2.1630 (3.4394) grad_norm 1.6438 (inf) [2021-04-16 08:41:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][1010/1251] eta 0:01:09 lr 0.000353 time 0.2944 (0.2880) loss 2.1551 (3.4374) grad_norm 2.0727 (inf) [2021-04-16 08:41:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][1020/1251] eta 0:01:06 lr 0.000353 time 0.2474 (0.2879) loss 3.8304 (3.4401) grad_norm 1.6528 (inf) [2021-04-16 08:41:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][1030/1251] eta 0:01:03 lr 0.000353 time 0.2482 (0.2880) loss 3.7369 (3.4406) grad_norm 1.5911 (inf) [2021-04-16 08:41:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][1040/1251] eta 0:01:00 lr 0.000353 time 0.2900 (0.2879) loss 3.9983 (3.4415) grad_norm 1.9518 (inf) [2021-04-16 08:41:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][1050/1251] eta 0:00:57 lr 0.000353 time 0.2608 (0.2878) loss 2.5246 (3.4402) grad_norm 1.7661 (inf) [2021-04-16 08:41:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][1060/1251] eta 0:00:54 lr 0.000353 time 0.2932 (0.2877) loss 3.6466 (3.4427) grad_norm 1.6691 (inf) [2021-04-16 08:41:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][1070/1251] eta 0:00:52 lr 0.000353 time 0.2882 (0.2876) loss 3.0204 (3.4427) grad_norm 1.6650 (inf) [2021-04-16 08:41:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][1080/1251] eta 0:00:49 lr 0.000353 time 0.2808 (0.2875) loss 4.1032 (3.4404) grad_norm 1.6138 (inf) [2021-04-16 08:41:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][1090/1251] eta 0:00:46 lr 0.000353 time 0.2806 (0.2875) loss 1.9637 (3.4378) grad_norm 1.7565 (inf) [2021-04-16 08:41:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][1100/1251] eta 0:00:43 lr 0.000353 time 0.2560 (0.2873) loss 3.6929 (3.4381) grad_norm 1.9117 (inf) [2021-04-16 08:41:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][1110/1251] eta 0:00:40 lr 0.000353 time 0.2906 (0.2873) loss 3.1128 (3.4391) grad_norm 2.1327 (inf) [2021-04-16 08:41:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][1120/1251] eta 0:00:37 lr 0.000353 time 0.2926 (0.2873) loss 3.6586 (3.4415) grad_norm 1.9293 (inf) [2021-04-16 08:41:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][1130/1251] eta 0:00:34 lr 0.000353 time 0.2671 (0.2873) loss 3.8216 (3.4437) grad_norm 1.5279 (inf) [2021-04-16 08:41:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][1140/1251] eta 0:00:31 lr 0.000352 time 0.2907 (0.2872) loss 3.3776 (3.4444) grad_norm 1.5935 (inf) [2021-04-16 08:41:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][1150/1251] eta 0:00:29 lr 0.000352 time 0.2999 (0.2873) loss 3.0046 (3.4464) grad_norm 1.6119 (inf) [2021-04-16 08:41:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][1160/1251] eta 0:00:26 lr 0.000352 time 0.2846 (0.2872) loss 2.4802 (3.4458) grad_norm 1.8172 (inf) [2021-04-16 08:42:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][1170/1251] eta 0:00:23 lr 0.000352 time 0.2823 (0.2871) loss 3.1038 (3.4455) grad_norm 2.2119 (inf) [2021-04-16 08:42:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][1180/1251] eta 0:00:20 lr 0.000352 time 0.2542 (0.2871) loss 3.3718 (3.4456) grad_norm 1.6583 (inf) [2021-04-16 08:42:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][1190/1251] eta 0:00:17 lr 0.000352 time 0.2723 (0.2870) loss 3.6378 (3.4455) grad_norm 1.6306 (inf) [2021-04-16 08:42:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][1200/1251] eta 0:00:14 lr 0.000352 time 0.2774 (0.2870) loss 3.3939 (3.4451) grad_norm 1.8463 (inf) [2021-04-16 08:42:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][1210/1251] eta 0:00:11 lr 0.000352 time 0.2991 (0.2869) loss 3.5754 (3.4477) grad_norm 1.4800 (inf) [2021-04-16 08:42:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][1220/1251] eta 0:00:08 lr 0.000352 time 0.2704 (0.2868) loss 3.2305 (3.4479) grad_norm 1.7222 (inf) [2021-04-16 08:42:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][1230/1251] eta 0:00:06 lr 0.000352 time 0.2849 (0.2867) loss 4.2359 (3.4509) grad_norm 1.6524 (inf) [2021-04-16 08:42:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][1240/1251] eta 0:00:03 lr 0.000352 time 0.2482 (0.2866) loss 4.3553 (3.4511) grad_norm 1.6117 (inf) [2021-04-16 08:42:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [179/300][1250/1251] eta 0:00:00 lr 0.000352 time 0.2491 (0.2863) loss 2.5235 (3.4502) grad_norm 1.7785 (inf) [2021-04-16 08:42:27 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 179 training takes 0:06:01 [2021-04-16 08:42:27 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_179.pth saving...... [2021-04-16 08:42:39 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_179.pth saved !!! [2021-04-16 08:42:41 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.290 (1.290) Loss 0.9240 (0.9240) Acc@1 78.613 (78.613) Acc@5 94.434 (94.434) [2021-04-16 08:42:42 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.114 (0.222) Loss 0.9593 (0.9461) Acc@1 78.711 (77.548) Acc@5 92.871 (93.999) [2021-04-16 08:42:45 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.140 (0.241) Loss 1.0392 (0.9628) Acc@1 75.977 (77.037) Acc@5 93.262 (93.829) [2021-04-16 08:42:47 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.085 (0.251) Loss 0.9118 (0.9521) Acc@1 78.125 (77.287) Acc@5 93.457 (93.904) [2021-04-16 08:42:48 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.214) Loss 0.9167 (0.9480) Acc@1 78.809 (77.434) Acc@5 94.043 (93.964) [2021-04-16 08:42:53 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 77.506 Acc@5 93.968 [2021-04-16 08:42:53 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 77.5% [2021-04-16 08:42:53 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 77.51% [2021-04-16 08:43:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][0/1251] eta 2:20:57 lr 0.000352 time 6.7607 (6.7607) loss 3.2595 (3.2595) grad_norm 1.5747 (1.5747) [2021-04-16 08:43:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][10/1251] eta 0:17:54 lr 0.000352 time 0.2662 (0.8659) loss 3.8014 (3.4237) grad_norm 1.7767 (1.6749) [2021-04-16 08:43:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][20/1251] eta 0:12:03 lr 0.000352 time 0.2987 (0.5877) loss 3.0850 (3.3476) grad_norm 1.8416 (1.7050) [2021-04-16 08:43:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][30/1251] eta 0:09:58 lr 0.000352 time 0.3063 (0.4899) loss 3.4059 (3.4311) grad_norm 1.7243 (1.7014) [2021-04-16 08:43:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][40/1251] eta 0:08:50 lr 0.000352 time 0.2952 (0.4384) loss 3.7498 (3.4336) grad_norm 1.6113 (1.7158) [2021-04-16 08:43:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][50/1251] eta 0:08:08 lr 0.000352 time 0.2672 (0.4066) loss 3.1868 (3.4243) grad_norm 1.8053 (1.7352) [2021-04-16 08:43:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][60/1251] eta 0:07:38 lr 0.000352 time 0.2983 (0.3854) loss 3.1938 (3.4220) grad_norm 1.8331 (1.7464) [2021-04-16 08:43:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][70/1251] eta 0:07:17 lr 0.000352 time 0.2813 (0.3703) loss 3.2412 (3.4705) grad_norm 1.7505 (1.7406) [2021-04-16 08:43:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][80/1251] eta 0:07:00 lr 0.000352 time 0.2722 (0.3589) loss 3.0494 (3.4813) grad_norm 1.5598 (1.7312) [2021-04-16 08:43:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][90/1251] eta 0:06:48 lr 0.000352 time 0.3014 (0.3515) loss 3.5651 (3.4684) grad_norm 1.7416 (1.7316) [2021-04-16 08:43:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][100/1251] eta 0:06:36 lr 0.000352 time 0.2920 (0.3442) loss 3.7947 (3.4753) grad_norm 1.7641 (1.7293) [2021-04-16 08:43:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][110/1251] eta 0:06:25 lr 0.000352 time 0.2809 (0.3380) loss 3.4729 (3.4377) grad_norm 1.7038 (1.7261) [2021-04-16 08:43:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][120/1251] eta 0:06:18 lr 0.000352 time 0.2631 (0.3345) loss 3.3843 (3.4334) grad_norm 1.8210 (1.7393) [2021-04-16 08:43:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][130/1251] eta 0:06:09 lr 0.000352 time 0.2778 (0.3296) loss 3.8347 (3.4243) grad_norm 1.5622 (1.7441) [2021-04-16 08:43:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][140/1251] eta 0:06:03 lr 0.000351 time 0.2699 (0.3270) loss 3.5679 (3.4186) grad_norm 1.6645 (1.7399) [2021-04-16 08:43:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][150/1251] eta 0:05:57 lr 0.000351 time 0.2607 (0.3247) loss 3.1484 (3.4206) grad_norm 2.2890 (1.7492) [2021-04-16 08:43:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][160/1251] eta 0:05:51 lr 0.000351 time 0.2793 (0.3219) loss 3.6035 (3.4193) grad_norm 1.7655 (1.7478) [2021-04-16 08:43:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][170/1251] eta 0:05:45 lr 0.000351 time 0.2993 (0.3193) loss 3.5004 (3.4289) grad_norm 1.5625 (1.7527) [2021-04-16 08:43:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][180/1251] eta 0:05:40 lr 0.000351 time 0.2940 (0.3179) loss 3.4693 (3.4221) grad_norm 1.8345 (1.7591) [2021-04-16 08:43:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][190/1251] eta 0:05:34 lr 0.000351 time 0.2837 (0.3155) loss 3.8721 (3.4213) grad_norm 1.7304 (1.7600) [2021-04-16 08:43:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][200/1251] eta 0:05:29 lr 0.000351 time 0.2900 (0.3138) loss 2.9485 (3.4275) grad_norm 1.5250 (1.7619) [2021-04-16 08:43:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][210/1251] eta 0:05:24 lr 0.000351 time 0.2777 (0.3121) loss 3.3281 (3.4322) grad_norm 1.4598 (1.7601) [2021-04-16 08:44:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][220/1251] eta 0:05:20 lr 0.000351 time 0.2765 (0.3105) loss 3.9205 (3.4296) grad_norm 1.6603 (1.7590) [2021-04-16 08:44:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][230/1251] eta 0:05:15 lr 0.000351 time 0.2881 (0.3092) loss 3.5903 (3.4418) grad_norm 1.5960 (1.7563) [2021-04-16 08:44:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][240/1251] eta 0:05:11 lr 0.000351 time 0.2713 (0.3078) loss 3.5745 (3.4346) grad_norm 1.7761 (1.7582) [2021-04-16 08:44:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][250/1251] eta 0:05:07 lr 0.000351 time 0.2879 (0.3072) loss 3.5074 (3.4404) grad_norm 1.7684 (1.7573) [2021-04-16 08:44:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][260/1251] eta 0:05:03 lr 0.000351 time 0.2821 (0.3059) loss 2.7451 (3.4491) grad_norm 1.5258 (1.7581) [2021-04-16 08:44:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][270/1251] eta 0:04:58 lr 0.000351 time 0.2778 (0.3048) loss 2.5184 (3.4453) grad_norm 1.8156 (1.7568) [2021-04-16 08:44:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][280/1251] eta 0:04:55 lr 0.000351 time 0.2854 (0.3038) loss 2.8969 (3.4379) grad_norm 1.7499 (1.7575) [2021-04-16 08:44:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][290/1251] eta 0:04:51 lr 0.000351 time 0.2884 (0.3029) loss 3.8775 (3.4446) grad_norm 1.8427 (1.7590) [2021-04-16 08:44:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][300/1251] eta 0:04:47 lr 0.000351 time 0.2690 (0.3018) loss 1.8494 (3.4340) grad_norm 1.5950 (1.7606) [2021-04-16 08:44:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][310/1251] eta 0:04:43 lr 0.000351 time 0.2802 (0.3010) loss 3.9449 (3.4381) grad_norm 1.7487 (1.7619) [2021-04-16 08:44:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][320/1251] eta 0:04:40 lr 0.000351 time 0.2768 (0.3008) loss 3.7351 (3.4352) grad_norm 1.6024 (1.7620) [2021-04-16 08:44:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][330/1251] eta 0:04:36 lr 0.000351 time 0.2588 (0.3000) loss 3.8750 (3.4284) grad_norm 1.7391 (1.7635) [2021-04-16 08:44:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][340/1251] eta 0:04:32 lr 0.000351 time 0.2755 (0.2991) loss 3.2394 (3.4323) grad_norm 1.5567 (1.7611) [2021-04-16 08:44:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][350/1251] eta 0:04:28 lr 0.000351 time 0.2700 (0.2983) loss 3.2259 (3.4309) grad_norm 1.7097 (1.7593) [2021-04-16 08:44:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][360/1251] eta 0:04:25 lr 0.000351 time 0.2890 (0.2981) loss 3.9519 (3.4321) grad_norm 1.9266 (1.7593) [2021-04-16 08:44:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][370/1251] eta 0:04:22 lr 0.000351 time 0.2471 (0.2975) loss 2.5136 (3.4250) grad_norm 1.6888 (1.7607) [2021-04-16 08:44:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][380/1251] eta 0:04:18 lr 0.000351 time 0.2891 (0.2972) loss 3.4959 (3.4199) grad_norm 1.7754 (1.7602) [2021-04-16 08:44:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][390/1251] eta 0:04:15 lr 0.000351 time 0.2830 (0.2967) loss 3.7477 (3.4279) grad_norm 2.1770 (1.7650) [2021-04-16 08:44:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][400/1251] eta 0:04:11 lr 0.000350 time 0.2990 (0.2961) loss 3.5164 (3.4301) grad_norm 1.7817 (1.7677) [2021-04-16 08:44:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][410/1251] eta 0:04:08 lr 0.000350 time 0.2559 (0.2955) loss 3.7716 (3.4285) grad_norm 1.8664 (1.7728) [2021-04-16 08:44:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][420/1251] eta 0:04:05 lr 0.000350 time 0.2616 (0.2951) loss 3.9282 (3.4273) grad_norm 1.6419 (1.7722) [2021-04-16 08:45:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][430/1251] eta 0:04:02 lr 0.000350 time 0.2606 (0.2948) loss 4.2293 (3.4305) grad_norm 1.7285 (1.7711) [2021-04-16 08:45:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][440/1251] eta 0:03:58 lr 0.000350 time 0.2604 (0.2943) loss 3.4098 (3.4249) grad_norm 2.0121 (1.7721) [2021-04-16 08:45:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][450/1251] eta 0:03:55 lr 0.000350 time 0.2574 (0.2939) loss 2.4748 (3.4258) grad_norm 1.6995 (1.7724) [2021-04-16 08:45:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][460/1251] eta 0:03:52 lr 0.000350 time 0.2795 (0.2936) loss 3.6663 (3.4289) grad_norm 1.9559 (1.7721) [2021-04-16 08:45:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][470/1251] eta 0:03:48 lr 0.000350 time 0.2821 (0.2932) loss 2.8459 (3.4294) grad_norm 1.6747 (1.7734) [2021-04-16 08:45:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][480/1251] eta 0:03:45 lr 0.000350 time 0.2621 (0.2929) loss 2.8061 (3.4278) grad_norm 1.7907 (1.7715) [2021-04-16 08:45:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][490/1251] eta 0:03:42 lr 0.000350 time 0.2574 (0.2925) loss 3.3458 (3.4318) grad_norm 1.6257 (1.7712) [2021-04-16 08:45:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][500/1251] eta 0:03:39 lr 0.000350 time 0.2685 (0.2921) loss 4.4407 (3.4312) grad_norm 1.9157 (1.7720) [2021-04-16 08:45:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][510/1251] eta 0:03:36 lr 0.000350 time 0.2804 (0.2918) loss 3.3976 (3.4296) grad_norm 1.8935 (1.7735) [2021-04-16 08:45:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][520/1251] eta 0:03:33 lr 0.000350 time 0.2888 (0.2917) loss 3.8134 (3.4308) grad_norm 1.6100 (1.7722) [2021-04-16 08:45:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][530/1251] eta 0:03:30 lr 0.000350 time 0.2853 (0.2914) loss 2.1971 (3.4277) grad_norm 1.9366 (1.7724) [2021-04-16 08:45:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][540/1251] eta 0:03:26 lr 0.000350 time 0.2865 (0.2911) loss 3.7879 (3.4306) grad_norm 1.7643 (1.7742) [2021-04-16 08:45:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][550/1251] eta 0:03:23 lr 0.000350 time 0.2678 (0.2907) loss 4.4440 (3.4330) grad_norm 1.6792 (1.7719) [2021-04-16 08:45:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][560/1251] eta 0:03:20 lr 0.000350 time 0.2953 (0.2905) loss 4.3395 (3.4328) grad_norm 1.6129 (1.7723) [2021-04-16 08:45:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][570/1251] eta 0:03:17 lr 0.000350 time 0.2571 (0.2905) loss 3.2188 (3.4355) grad_norm 2.2727 (1.7738) [2021-04-16 08:45:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][580/1251] eta 0:03:15 lr 0.000350 time 0.2808 (0.2907) loss 2.8943 (3.4329) grad_norm 1.5948 (1.7717) [2021-04-16 08:45:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][590/1251] eta 0:03:11 lr 0.000350 time 0.2729 (0.2904) loss 3.4272 (3.4353) grad_norm 1.9303 (1.7715) [2021-04-16 08:45:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][600/1251] eta 0:03:08 lr 0.000350 time 0.2875 (0.2903) loss 3.3314 (3.4392) grad_norm 1.7415 (1.7706) [2021-04-16 08:45:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][610/1251] eta 0:03:05 lr 0.000350 time 0.2536 (0.2900) loss 3.3422 (3.4384) grad_norm 1.5704 (1.7700) [2021-04-16 08:45:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][620/1251] eta 0:03:02 lr 0.000350 time 0.2633 (0.2898) loss 2.2569 (3.4371) grad_norm 2.3994 (1.7734) [2021-04-16 08:45:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][630/1251] eta 0:02:59 lr 0.000350 time 0.2770 (0.2897) loss 4.1370 (3.4390) grad_norm 1.8238 (1.7739) [2021-04-16 08:45:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][640/1251] eta 0:02:56 lr 0.000350 time 0.3002 (0.2895) loss 4.0333 (3.4384) grad_norm 1.8469 (1.7749) [2021-04-16 08:46:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][650/1251] eta 0:02:53 lr 0.000349 time 0.2659 (0.2892) loss 2.3039 (3.4380) grad_norm 2.1794 (1.7769) [2021-04-16 08:46:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][660/1251] eta 0:02:50 lr 0.000349 time 0.2849 (0.2892) loss 3.8258 (3.4438) grad_norm 1.7066 (1.7782) [2021-04-16 08:46:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][670/1251] eta 0:02:48 lr 0.000349 time 0.2943 (0.2892) loss 3.4976 (3.4482) grad_norm 1.5458 (1.7769) [2021-04-16 08:46:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][680/1251] eta 0:02:45 lr 0.000349 time 0.2926 (0.2890) loss 3.2638 (3.4482) grad_norm 1.5796 (1.7765) [2021-04-16 08:46:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][690/1251] eta 0:02:42 lr 0.000349 time 0.2634 (0.2890) loss 3.8002 (3.4479) grad_norm 1.8543 (1.7761) [2021-04-16 08:46:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][700/1251] eta 0:02:39 lr 0.000349 time 0.2625 (0.2889) loss 3.6098 (3.4505) grad_norm 1.6922 (1.7745) [2021-04-16 08:46:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][710/1251] eta 0:02:36 lr 0.000349 time 0.2903 (0.2888) loss 3.6431 (3.4519) grad_norm 1.6761 (1.7742) [2021-04-16 08:46:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][720/1251] eta 0:02:33 lr 0.000349 time 0.2994 (0.2889) loss 3.9086 (3.4483) grad_norm 1.7114 (1.7727) [2021-04-16 08:46:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][730/1251] eta 0:02:30 lr 0.000349 time 0.2553 (0.2887) loss 3.3809 (3.4491) grad_norm 2.2888 (1.7725) [2021-04-16 08:46:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][740/1251] eta 0:02:27 lr 0.000349 time 0.2937 (0.2885) loss 3.9589 (3.4487) grad_norm 1.5069 (1.7726) [2021-04-16 08:46:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][750/1251] eta 0:02:24 lr 0.000349 time 0.2936 (0.2884) loss 2.5548 (3.4475) grad_norm 1.7458 (1.7713) [2021-04-16 08:46:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][760/1251] eta 0:02:21 lr 0.000349 time 0.2920 (0.2884) loss 3.7020 (3.4464) grad_norm 1.5838 (1.7712) [2021-04-16 08:46:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][770/1251] eta 0:02:18 lr 0.000349 time 0.2558 (0.2882) loss 3.9575 (3.4471) grad_norm 1.8325 (1.7721) [2021-04-16 08:46:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][780/1251] eta 0:02:15 lr 0.000349 time 0.2694 (0.2881) loss 4.1115 (3.4479) grad_norm 1.8920 (1.7731) [2021-04-16 08:46:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][790/1251] eta 0:02:12 lr 0.000349 time 0.2758 (0.2880) loss 3.2745 (3.4490) grad_norm 1.9891 (1.7740) [2021-04-16 08:46:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][800/1251] eta 0:02:09 lr 0.000349 time 0.2900 (0.2879) loss 3.2478 (3.4481) grad_norm 1.6415 (1.7729) [2021-04-16 08:46:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][810/1251] eta 0:02:06 lr 0.000349 time 0.2862 (0.2878) loss 3.5560 (3.4486) grad_norm 1.7979 (1.7730) [2021-04-16 08:46:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][820/1251] eta 0:02:03 lr 0.000349 time 0.2712 (0.2877) loss 2.3762 (3.4492) grad_norm 1.7145 (1.7725) [2021-04-16 08:46:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][830/1251] eta 0:02:01 lr 0.000349 time 0.2719 (0.2877) loss 2.9562 (3.4489) grad_norm 1.5518 (1.7721) [2021-04-16 08:46:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][840/1251] eta 0:01:58 lr 0.000349 time 0.2630 (0.2876) loss 2.6388 (3.4456) grad_norm 2.4562 (1.7728) [2021-04-16 08:46:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][850/1251] eta 0:01:55 lr 0.000349 time 0.2730 (0.2875) loss 3.4317 (3.4454) grad_norm 1.6938 (1.7724) [2021-04-16 08:47:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][860/1251] eta 0:01:52 lr 0.000349 time 0.2768 (0.2874) loss 2.3270 (3.4429) grad_norm 1.7964 (1.7718) [2021-04-16 08:47:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][870/1251] eta 0:01:49 lr 0.000349 time 0.3047 (0.2873) loss 4.3426 (3.4422) grad_norm 1.5920 (1.7714) [2021-04-16 08:47:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][880/1251] eta 0:01:46 lr 0.000349 time 0.3103 (0.2873) loss 3.6535 (3.4446) grad_norm 1.6655 (1.7700) [2021-04-16 08:47:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][890/1251] eta 0:01:43 lr 0.000349 time 0.2569 (0.2871) loss 2.3338 (3.4403) grad_norm 2.4471 (1.7697) [2021-04-16 08:47:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][900/1251] eta 0:01:40 lr 0.000348 time 0.2968 (0.2872) loss 3.0096 (3.4402) grad_norm 1.6341 (1.7713) [2021-04-16 08:47:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][910/1251] eta 0:01:37 lr 0.000348 time 0.2754 (0.2870) loss 2.3076 (3.4378) grad_norm 1.7555 (1.7713) [2021-04-16 08:47:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][920/1251] eta 0:01:34 lr 0.000348 time 0.2841 (0.2869) loss 3.3338 (3.4383) grad_norm 1.5799 (1.7708) [2021-04-16 08:47:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][930/1251] eta 0:01:32 lr 0.000348 time 0.3097 (0.2869) loss 3.3850 (3.4381) grad_norm 1.5593 (1.7711) [2021-04-16 08:47:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][940/1251] eta 0:01:29 lr 0.000348 time 0.2716 (0.2868) loss 3.0571 (3.4334) grad_norm 1.7418 (1.7722) [2021-04-16 08:47:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][950/1251] eta 0:01:26 lr 0.000348 time 0.3034 (0.2867) loss 3.9129 (3.4336) grad_norm 2.0682 (1.7723) [2021-04-16 08:47:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][960/1251] eta 0:01:23 lr 0.000348 time 0.2751 (0.2868) loss 3.0347 (3.4331) grad_norm 2.8049 (1.7735) [2021-04-16 08:47:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][970/1251] eta 0:01:20 lr 0.000348 time 0.2955 (0.2866) loss 3.8652 (3.4334) grad_norm 1.6206 (1.7741) [2021-04-16 08:47:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][980/1251] eta 0:01:17 lr 0.000348 time 0.2792 (0.2865) loss 3.7269 (3.4293) grad_norm 1.9619 (1.7744) [2021-04-16 08:47:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][990/1251] eta 0:01:14 lr 0.000348 time 0.2756 (0.2864) loss 4.0747 (3.4294) grad_norm 1.6186 (1.7738) [2021-04-16 08:47:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][1000/1251] eta 0:01:11 lr 0.000348 time 0.2599 (0.2863) loss 4.0999 (3.4314) grad_norm 1.6906 (1.7747) [2021-04-16 08:47:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][1010/1251] eta 0:01:08 lr 0.000348 time 0.2655 (0.2863) loss 3.5778 (3.4323) grad_norm 1.6899 (1.7753) [2021-04-16 08:47:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][1020/1251] eta 0:01:06 lr 0.000348 time 0.2640 (0.2861) loss 3.9027 (3.4350) grad_norm 1.4465 (1.7755) [2021-04-16 08:47:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][1030/1251] eta 0:01:03 lr 0.000348 time 0.2881 (0.2860) loss 3.6326 (3.4339) grad_norm 1.7764 (1.7762) [2021-04-16 08:47:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][1040/1251] eta 0:01:00 lr 0.000348 time 0.3078 (0.2860) loss 3.0109 (3.4316) grad_norm 2.2045 (1.7780) [2021-04-16 08:47:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][1050/1251] eta 0:00:57 lr 0.000348 time 0.2675 (0.2860) loss 4.0958 (3.4343) grad_norm 1.8107 (1.7797) [2021-04-16 08:47:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][1060/1251] eta 0:00:54 lr 0.000348 time 0.2871 (0.2860) loss 3.7586 (3.4308) grad_norm 1.9608 (1.7796) [2021-04-16 08:47:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][1070/1251] eta 0:00:51 lr 0.000348 time 0.2866 (0.2859) loss 2.3068 (3.4296) grad_norm 1.7337 (1.7786) [2021-04-16 08:48:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][1080/1251] eta 0:00:48 lr 0.000348 time 0.2676 (0.2858) loss 3.5145 (3.4278) grad_norm 1.6406 (1.7777) [2021-04-16 08:48:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][1090/1251] eta 0:00:46 lr 0.000348 time 0.2684 (0.2859) loss 3.6703 (3.4268) grad_norm 1.8912 (1.7779) [2021-04-16 08:48:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][1100/1251] eta 0:00:43 lr 0.000348 time 0.2802 (0.2857) loss 3.9644 (3.4283) grad_norm 1.6946 (1.7792) [2021-04-16 08:48:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][1110/1251] eta 0:00:40 lr 0.000348 time 0.2805 (0.2857) loss 2.5479 (3.4282) grad_norm 1.7232 (1.7790) [2021-04-16 08:48:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][1120/1251] eta 0:00:37 lr 0.000348 time 0.2662 (0.2856) loss 3.3330 (3.4286) grad_norm 1.6870 (1.7790) [2021-04-16 08:48:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][1130/1251] eta 0:00:34 lr 0.000348 time 0.2986 (0.2856) loss 3.5934 (3.4281) grad_norm 1.5801 (1.7780) [2021-04-16 08:48:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][1140/1251] eta 0:00:31 lr 0.000348 time 0.3003 (0.2855) loss 2.7719 (3.4286) grad_norm 1.7782 (1.7786) [2021-04-16 08:48:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][1150/1251] eta 0:00:28 lr 0.000348 time 0.2637 (0.2856) loss 3.3685 (3.4289) grad_norm 1.7689 (1.7782) [2021-04-16 08:48:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][1160/1251] eta 0:00:25 lr 0.000347 time 0.2758 (0.2857) loss 3.0056 (3.4281) grad_norm 1.8931 (1.7770) [2021-04-16 08:48:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][1170/1251] eta 0:00:23 lr 0.000347 time 0.2766 (0.2856) loss 3.1385 (3.4283) grad_norm 1.4613 (1.7762) [2021-04-16 08:48:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][1180/1251] eta 0:00:20 lr 0.000347 time 0.2670 (0.2855) loss 3.6119 (3.4288) grad_norm 1.6441 (1.7763) [2021-04-16 08:48:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][1190/1251] eta 0:00:17 lr 0.000347 time 0.2780 (0.2854) loss 3.9247 (3.4305) grad_norm 1.8188 (1.7763) [2021-04-16 08:48:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][1200/1251] eta 0:00:14 lr 0.000347 time 0.2720 (0.2853) loss 3.9167 (3.4337) grad_norm 1.6375 (1.7761) [2021-04-16 08:48:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][1210/1251] eta 0:00:11 lr 0.000347 time 0.2613 (0.2852) loss 4.1344 (3.4343) grad_norm 1.7065 (1.7759) [2021-04-16 08:48:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][1220/1251] eta 0:00:08 lr 0.000347 time 0.2620 (0.2851) loss 3.4539 (3.4332) grad_norm 1.7997 (1.7757) [2021-04-16 08:48:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][1230/1251] eta 0:00:05 lr 0.000347 time 0.2545 (0.2850) loss 3.9736 (3.4310) grad_norm 1.9534 (1.7759) [2021-04-16 08:48:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][1240/1251] eta 0:00:03 lr 0.000347 time 0.3358 (0.2849) loss 3.8387 (3.4271) grad_norm 1.6513 (1.7758) [2021-04-16 08:48:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [180/300][1250/1251] eta 0:00:00 lr 0.000347 time 0.2491 (0.2847) loss 4.0454 (3.4274) grad_norm 1.7633 (1.7756) [2021-04-16 08:48:52 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 180 training takes 0:05:59 [2021-04-16 08:48:52 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_180.pth saving...... [2021-04-16 08:49:00 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_180.pth saved !!! [2021-04-16 08:49:01 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.165 (1.165) Loss 0.9755 (0.9755) Acc@1 76.855 (76.855) Acc@5 93.652 (93.652) [2021-04-16 08:49:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.408 (0.246) Loss 0.9937 (0.9808) Acc@1 77.637 (77.006) Acc@5 93.555 (93.874) [2021-04-16 08:49:04 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.118 (0.204) Loss 0.9922 (0.9699) Acc@1 77.441 (77.316) Acc@5 93.652 (93.838) [2021-04-16 08:49:07 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 1.188 (0.248) Loss 0.9620 (0.9732) Acc@1 78.223 (77.271) Acc@5 93.555 (93.892) [2021-04-16 08:49:09 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.068 (0.221) Loss 0.9531 (0.9679) Acc@1 76.953 (77.391) Acc@5 93.945 (93.933) [2021-04-16 08:49:18 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 77.216 Acc@5 93.896 [2021-04-16 08:49:18 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 77.2% [2021-04-16 08:49:18 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 77.51% [2021-04-16 08:49:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][0/1251] eta 1:08:07 lr 0.000347 time 3.2672 (3.2672) loss 4.1252 (4.1252) grad_norm 1.8121 (1.8121) [2021-04-16 08:49:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][10/1251] eta 0:11:20 lr 0.000347 time 0.2790 (0.5481) loss 2.9106 (3.3263) grad_norm 2.0216 (1.7517) [2021-04-16 08:49:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][20/1251] eta 0:08:39 lr 0.000347 time 0.2879 (0.4217) loss 2.9935 (3.3429) grad_norm 1.6844 (1.7488) [2021-04-16 08:49:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][30/1251] eta 0:07:36 lr 0.000347 time 0.2835 (0.3737) loss 3.2950 (3.2722) grad_norm 1.8650 (1.7677) [2021-04-16 08:49:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][40/1251] eta 0:07:07 lr 0.000347 time 0.3021 (0.3533) loss 3.0825 (3.3218) grad_norm 1.7653 (1.7749) [2021-04-16 08:49:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][50/1251] eta 0:06:46 lr 0.000347 time 0.2831 (0.3384) loss 3.8797 (3.3413) grad_norm 1.5057 (1.7614) [2021-04-16 08:49:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][60/1251] eta 0:06:33 lr 0.000347 time 0.2678 (0.3305) loss 2.5925 (3.3528) grad_norm 1.7986 (1.7671) [2021-04-16 08:49:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][70/1251] eta 0:06:21 lr 0.000347 time 0.2861 (0.3227) loss 3.4727 (3.3879) grad_norm 1.6576 (1.7638) [2021-04-16 08:49:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][80/1251] eta 0:06:11 lr 0.000347 time 0.2581 (0.3171) loss 3.6596 (3.4379) grad_norm 1.7909 (1.7651) [2021-04-16 08:49:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][90/1251] eta 0:06:02 lr 0.000347 time 0.2524 (0.3124) loss 3.7919 (3.4556) grad_norm 1.6033 (1.7631) [2021-04-16 08:49:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][100/1251] eta 0:05:56 lr 0.000347 time 0.2811 (0.3094) loss 3.5946 (3.4389) grad_norm 1.7716 (1.7691) [2021-04-16 08:49:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][110/1251] eta 0:05:49 lr 0.000347 time 0.2599 (0.3063) loss 3.7231 (3.4208) grad_norm 1.6984 (1.7754) [2021-04-16 08:49:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][120/1251] eta 0:05:43 lr 0.000347 time 0.2667 (0.3039) loss 4.1293 (3.4301) grad_norm 1.5609 (1.7700) [2021-04-16 08:49:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][130/1251] eta 0:05:38 lr 0.000347 time 0.2617 (0.3019) loss 3.5615 (3.4105) grad_norm 1.8985 (1.7687) [2021-04-16 08:50:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][140/1251] eta 0:05:33 lr 0.000347 time 0.2886 (0.3003) loss 3.3341 (3.4209) grad_norm 2.0443 (1.7677) [2021-04-16 08:50:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][150/1251] eta 0:05:29 lr 0.000347 time 0.2594 (0.2995) loss 3.3197 (3.4079) grad_norm 1.7559 (1.7692) [2021-04-16 08:50:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][160/1251] eta 0:05:25 lr 0.000346 time 0.2902 (0.2987) loss 3.6524 (3.4012) grad_norm 1.6487 (1.7681) [2021-04-16 08:50:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][170/1251] eta 0:05:22 lr 0.000346 time 0.2884 (0.2987) loss 2.1308 (3.4004) grad_norm 1.7826 (1.7665) [2021-04-16 08:50:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][180/1251] eta 0:05:18 lr 0.000346 time 0.2947 (0.2977) loss 3.6977 (3.3842) grad_norm 1.6515 (1.7693) [2021-04-16 08:50:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][190/1251] eta 0:05:14 lr 0.000346 time 0.3167 (0.2969) loss 3.9323 (3.3804) grad_norm 1.5053 (1.7669) [2021-04-16 08:50:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][200/1251] eta 0:05:10 lr 0.000346 time 0.2670 (0.2959) loss 3.7775 (3.3899) grad_norm 1.8227 (1.7693) [2021-04-16 08:50:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][210/1251] eta 0:05:07 lr 0.000346 time 0.2687 (0.2952) loss 3.3247 (3.3821) grad_norm 1.6296 (1.7688) [2021-04-16 08:50:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][220/1251] eta 0:05:03 lr 0.000346 time 0.2510 (0.2943) loss 3.7383 (3.3801) grad_norm inf (inf) [2021-04-16 08:50:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][230/1251] eta 0:04:59 lr 0.000346 time 0.2683 (0.2938) loss 3.6182 (3.3809) grad_norm 1.8797 (inf) [2021-04-16 08:50:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][240/1251] eta 0:04:56 lr 0.000346 time 0.2774 (0.2931) loss 3.8276 (3.3767) grad_norm 2.1651 (inf) [2021-04-16 08:50:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][250/1251] eta 0:04:52 lr 0.000346 time 0.2826 (0.2927) loss 4.2674 (3.3723) grad_norm 1.6209 (inf) [2021-04-16 08:50:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][260/1251] eta 0:04:49 lr 0.000346 time 0.2602 (0.2920) loss 3.9924 (3.3840) grad_norm 2.5006 (inf) [2021-04-16 08:50:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][270/1251] eta 0:04:46 lr 0.000346 time 0.2798 (0.2917) loss 2.3917 (3.3860) grad_norm 2.0391 (inf) [2021-04-16 08:50:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][280/1251] eta 0:04:42 lr 0.000346 time 0.2826 (0.2912) loss 3.6228 (3.3924) grad_norm 1.6887 (inf) [2021-04-16 08:50:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][290/1251] eta 0:04:39 lr 0.000346 time 0.2866 (0.2907) loss 3.4220 (3.3909) grad_norm 1.9049 (inf) [2021-04-16 08:50:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][300/1251] eta 0:04:36 lr 0.000346 time 0.2670 (0.2907) loss 3.6725 (3.3950) grad_norm 2.0764 (inf) [2021-04-16 08:50:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][310/1251] eta 0:04:33 lr 0.000346 time 0.2609 (0.2903) loss 3.7262 (3.3979) grad_norm 1.6099 (inf) [2021-04-16 08:50:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][320/1251] eta 0:04:29 lr 0.000346 time 0.2822 (0.2898) loss 2.0069 (3.3904) grad_norm 1.6750 (inf) [2021-04-16 08:50:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][330/1251] eta 0:04:26 lr 0.000346 time 0.2845 (0.2894) loss 3.1774 (3.3978) grad_norm 1.8746 (inf) [2021-04-16 08:50:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][340/1251] eta 0:04:23 lr 0.000346 time 0.2611 (0.2889) loss 4.1463 (3.3978) grad_norm 1.9019 (inf) [2021-04-16 08:50:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][350/1251] eta 0:04:20 lr 0.000346 time 0.2830 (0.2892) loss 3.3996 (3.4022) grad_norm 1.7796 (inf) [2021-04-16 08:51:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][360/1251] eta 0:04:17 lr 0.000346 time 0.2793 (0.2894) loss 3.6044 (3.3952) grad_norm 1.6063 (inf) [2021-04-16 08:51:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][370/1251] eta 0:04:15 lr 0.000346 time 0.2930 (0.2895) loss 3.7650 (3.3993) grad_norm 1.7272 (inf) [2021-04-16 08:51:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][380/1251] eta 0:04:11 lr 0.000346 time 0.2730 (0.2891) loss 3.5409 (3.3993) grad_norm 1.5647 (inf) [2021-04-16 08:51:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][390/1251] eta 0:04:08 lr 0.000346 time 0.2621 (0.2887) loss 3.7169 (3.3977) grad_norm 2.0253 (inf) [2021-04-16 08:51:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][400/1251] eta 0:04:05 lr 0.000346 time 0.2736 (0.2884) loss 2.7452 (3.3914) grad_norm 1.6172 (inf) [2021-04-16 08:51:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][410/1251] eta 0:04:02 lr 0.000346 time 0.3014 (0.2882) loss 3.4516 (3.3887) grad_norm 1.5714 (inf) [2021-04-16 08:51:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][420/1251] eta 0:03:59 lr 0.000345 time 0.2874 (0.2881) loss 2.5669 (3.3893) grad_norm 1.8265 (inf) [2021-04-16 08:51:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][430/1251] eta 0:03:56 lr 0.000345 time 0.2574 (0.2878) loss 3.6377 (3.3900) grad_norm 1.9582 (inf) [2021-04-16 08:51:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][440/1251] eta 0:03:53 lr 0.000345 time 0.2875 (0.2877) loss 3.8124 (3.3888) grad_norm 1.5837 (inf) [2021-04-16 08:51:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][450/1251] eta 0:03:50 lr 0.000345 time 0.4476 (0.2879) loss 3.7980 (3.3812) grad_norm 1.7997 (inf) [2021-04-16 08:51:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][460/1251] eta 0:03:47 lr 0.000345 time 0.2576 (0.2876) loss 3.7428 (3.3841) grad_norm 2.0804 (inf) [2021-04-16 08:51:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][470/1251] eta 0:03:44 lr 0.000345 time 0.2932 (0.2874) loss 2.8587 (3.3822) grad_norm 1.5428 (inf) [2021-04-16 08:51:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][480/1251] eta 0:03:41 lr 0.000345 time 0.2753 (0.2873) loss 3.8512 (3.3797) grad_norm 1.8039 (inf) [2021-04-16 08:51:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][490/1251] eta 0:03:38 lr 0.000345 time 0.2819 (0.2873) loss 4.0837 (3.3824) grad_norm 2.3974 (inf) [2021-04-16 08:51:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][500/1251] eta 0:03:35 lr 0.000345 time 0.2795 (0.2871) loss 3.7281 (3.3828) grad_norm 1.8324 (inf) [2021-04-16 08:51:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][510/1251] eta 0:03:32 lr 0.000345 time 0.2814 (0.2869) loss 2.1438 (3.3827) grad_norm 1.7997 (inf) [2021-04-16 08:51:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][520/1251] eta 0:03:29 lr 0.000345 time 0.2866 (0.2868) loss 4.0101 (3.3825) grad_norm 1.7498 (inf) [2021-04-16 08:51:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][530/1251] eta 0:03:26 lr 0.000345 time 0.2572 (0.2867) loss 3.1274 (3.3835) grad_norm 1.7344 (inf) [2021-04-16 08:51:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][540/1251] eta 0:03:23 lr 0.000345 time 0.2684 (0.2865) loss 3.0134 (3.3826) grad_norm 1.8723 (inf) [2021-04-16 08:51:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][550/1251] eta 0:03:20 lr 0.000345 time 0.2750 (0.2863) loss 2.5935 (3.3796) grad_norm 1.6454 (inf) [2021-04-16 08:51:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][560/1251] eta 0:03:17 lr 0.000345 time 0.3104 (0.2863) loss 3.7753 (3.3840) grad_norm 1.6552 (inf) [2021-04-16 08:52:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][570/1251] eta 0:03:14 lr 0.000345 time 0.2614 (0.2861) loss 3.1580 (3.3869) grad_norm 1.7329 (inf) [2021-04-16 08:52:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][580/1251] eta 0:03:12 lr 0.000345 time 0.2722 (0.2866) loss 3.5692 (3.3869) grad_norm 1.7354 (inf) [2021-04-16 08:52:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][590/1251] eta 0:03:09 lr 0.000345 time 0.2641 (0.2863) loss 3.6047 (3.3878) grad_norm 1.5494 (inf) [2021-04-16 08:52:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][600/1251] eta 0:03:06 lr 0.000345 time 0.2571 (0.2861) loss 3.5238 (3.3905) grad_norm 1.8042 (inf) [2021-04-16 08:52:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][610/1251] eta 0:03:03 lr 0.000345 time 0.2656 (0.2861) loss 3.7326 (3.3908) grad_norm 1.8456 (inf) [2021-04-16 08:52:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][620/1251] eta 0:03:00 lr 0.000345 time 0.2469 (0.2864) loss 3.0981 (3.3907) grad_norm 1.5719 (inf) [2021-04-16 08:52:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][630/1251] eta 0:02:57 lr 0.000345 time 0.2526 (0.2865) loss 4.1205 (3.3935) grad_norm 1.6429 (inf) [2021-04-16 08:52:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][640/1251] eta 0:02:54 lr 0.000345 time 0.2981 (0.2863) loss 3.5170 (3.3938) grad_norm 1.5572 (inf) [2021-04-16 08:52:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][650/1251] eta 0:02:51 lr 0.000345 time 0.2790 (0.2861) loss 3.8505 (3.3946) grad_norm 1.8367 (inf) [2021-04-16 08:52:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][660/1251] eta 0:02:49 lr 0.000345 time 0.2855 (0.2860) loss 3.7822 (3.3939) grad_norm 1.9838 (inf) [2021-04-16 08:52:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][670/1251] eta 0:02:46 lr 0.000344 time 0.2706 (0.2858) loss 3.3316 (3.3942) grad_norm 1.6896 (inf) [2021-04-16 08:52:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][680/1251] eta 0:02:43 lr 0.000344 time 0.3025 (0.2857) loss 3.9632 (3.3938) grad_norm 1.6033 (inf) [2021-04-16 08:52:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][690/1251] eta 0:02:40 lr 0.000344 time 0.2802 (0.2857) loss 3.7018 (3.3955) grad_norm 2.0733 (inf) [2021-04-16 08:52:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][700/1251] eta 0:02:37 lr 0.000344 time 0.2740 (0.2856) loss 3.5232 (3.3974) grad_norm 1.5683 (inf) [2021-04-16 08:52:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][710/1251] eta 0:02:34 lr 0.000344 time 0.2691 (0.2855) loss 3.7023 (3.3958) grad_norm 1.8609 (inf) [2021-04-16 08:52:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][720/1251] eta 0:02:31 lr 0.000344 time 0.2510 (0.2855) loss 3.3617 (3.3997) grad_norm 1.9121 (inf) [2021-04-16 08:52:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][730/1251] eta 0:02:28 lr 0.000344 time 0.2848 (0.2854) loss 2.3822 (3.3993) grad_norm 1.8982 (inf) [2021-04-16 08:52:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][740/1251] eta 0:02:25 lr 0.000344 time 0.4428 (0.2855) loss 3.3553 (3.4015) grad_norm 1.5713 (inf) [2021-04-16 08:52:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][750/1251] eta 0:02:23 lr 0.000344 time 0.2683 (0.2854) loss 3.5421 (3.4009) grad_norm 2.0883 (inf) [2021-04-16 08:52:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][760/1251] eta 0:02:20 lr 0.000344 time 0.2923 (0.2854) loss 3.7129 (3.4025) grad_norm 1.5623 (inf) [2021-04-16 08:52:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][770/1251] eta 0:02:17 lr 0.000344 time 0.2885 (0.2853) loss 3.4943 (3.4010) grad_norm 1.6739 (inf) [2021-04-16 08:53:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][780/1251] eta 0:02:14 lr 0.000344 time 0.2736 (0.2852) loss 3.5727 (3.4020) grad_norm 1.6267 (inf) [2021-04-16 08:53:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][790/1251] eta 0:02:11 lr 0.000344 time 0.2724 (0.2852) loss 2.9405 (3.3998) grad_norm 1.6255 (inf) [2021-04-16 08:53:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][800/1251] eta 0:02:08 lr 0.000344 time 0.2735 (0.2851) loss 2.3308 (3.3985) grad_norm 1.6734 (inf) [2021-04-16 08:53:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][810/1251] eta 0:02:05 lr 0.000344 time 0.2913 (0.2850) loss 3.7709 (3.3976) grad_norm 1.6224 (inf) [2021-04-16 08:53:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][820/1251] eta 0:02:02 lr 0.000344 time 0.2778 (0.2848) loss 2.6369 (3.3945) grad_norm 1.8259 (inf) [2021-04-16 08:53:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][830/1251] eta 0:01:59 lr 0.000344 time 0.2552 (0.2847) loss 3.3031 (3.3958) grad_norm 1.7460 (inf) [2021-04-16 08:53:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][840/1251] eta 0:01:56 lr 0.000344 time 0.2922 (0.2846) loss 3.5628 (3.3955) grad_norm 1.5380 (inf) [2021-04-16 08:53:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][850/1251] eta 0:01:54 lr 0.000344 time 0.3050 (0.2846) loss 3.6554 (3.3951) grad_norm 1.7566 (inf) [2021-04-16 08:53:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][860/1251] eta 0:01:51 lr 0.000344 time 0.2541 (0.2845) loss 3.4303 (3.3957) grad_norm 1.6404 (inf) [2021-04-16 08:53:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][870/1251] eta 0:01:48 lr 0.000344 time 0.2782 (0.2844) loss 3.6046 (3.3978) grad_norm 2.3967 (inf) [2021-04-16 08:53:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][880/1251] eta 0:01:45 lr 0.000344 time 0.2864 (0.2846) loss 3.5297 (3.3969) grad_norm 1.7810 (inf) [2021-04-16 08:53:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][890/1251] eta 0:01:42 lr 0.000344 time 0.2610 (0.2846) loss 3.2485 (3.3957) grad_norm 1.7407 (inf) [2021-04-16 08:53:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][900/1251] eta 0:01:39 lr 0.000344 time 0.2786 (0.2845) loss 3.6402 (3.3957) grad_norm 1.6327 (inf) [2021-04-16 08:53:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][910/1251] eta 0:01:36 lr 0.000344 time 0.3062 (0.2844) loss 3.6018 (3.3972) grad_norm 1.8172 (inf) [2021-04-16 08:53:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][920/1251] eta 0:01:34 lr 0.000344 time 0.2742 (0.2844) loss 2.2067 (3.3963) grad_norm 2.0640 (inf) [2021-04-16 08:53:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][930/1251] eta 0:01:31 lr 0.000343 time 0.2492 (0.2844) loss 3.3901 (3.3997) grad_norm 1.7031 (inf) [2021-04-16 08:53:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][940/1251] eta 0:01:28 lr 0.000343 time 0.2533 (0.2846) loss 3.6037 (3.4011) grad_norm 1.7531 (inf) [2021-04-16 08:53:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][950/1251] eta 0:01:25 lr 0.000343 time 0.2621 (0.2845) loss 3.8997 (3.4042) grad_norm 1.8253 (inf) [2021-04-16 08:53:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][960/1251] eta 0:01:22 lr 0.000343 time 0.2639 (0.2844) loss 3.2497 (3.4056) grad_norm 1.5882 (inf) [2021-04-16 08:53:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][970/1251] eta 0:01:19 lr 0.000343 time 0.2648 (0.2843) loss 3.7128 (3.4038) grad_norm 1.6557 (inf) [2021-04-16 08:53:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][980/1251] eta 0:01:17 lr 0.000343 time 0.2957 (0.2843) loss 2.1530 (3.4023) grad_norm 1.6833 (inf) [2021-04-16 08:54:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][990/1251] eta 0:01:14 lr 0.000343 time 0.2558 (0.2841) loss 2.9127 (3.4006) grad_norm 1.9474 (inf) [2021-04-16 08:54:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][1000/1251] eta 0:01:11 lr 0.000343 time 0.2453 (0.2841) loss 3.8361 (3.4047) grad_norm 1.7414 (inf) [2021-04-16 08:54:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][1010/1251] eta 0:01:08 lr 0.000343 time 0.2934 (0.2841) loss 3.0951 (3.4039) grad_norm 1.5209 (inf) [2021-04-16 08:54:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][1020/1251] eta 0:01:05 lr 0.000343 time 0.2945 (0.2841) loss 4.0487 (3.4028) grad_norm 1.9711 (inf) [2021-04-16 08:54:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][1030/1251] eta 0:01:02 lr 0.000343 time 0.2761 (0.2841) loss 3.2386 (3.4032) grad_norm 1.7434 (inf) [2021-04-16 08:54:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][1040/1251] eta 0:00:59 lr 0.000343 time 0.2684 (0.2840) loss 3.3631 (3.4020) grad_norm 1.9432 (inf) [2021-04-16 08:54:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][1050/1251] eta 0:00:57 lr 0.000343 time 0.2852 (0.2840) loss 3.6991 (3.3991) grad_norm 1.6349 (inf) [2021-04-16 08:54:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][1060/1251] eta 0:00:54 lr 0.000343 time 0.2533 (0.2839) loss 2.9062 (3.4013) grad_norm 1.7033 (inf) [2021-04-16 08:54:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][1070/1251] eta 0:00:51 lr 0.000343 time 0.2666 (0.2839) loss 3.9682 (3.4025) grad_norm 1.6043 (inf) [2021-04-16 08:54:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][1080/1251] eta 0:00:48 lr 0.000343 time 0.2861 (0.2839) loss 3.6003 (3.4024) grad_norm 1.6997 (inf) [2021-04-16 08:54:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][1090/1251] eta 0:00:45 lr 0.000343 time 0.2660 (0.2837) loss 4.0716 (3.4009) grad_norm 1.9848 (inf) [2021-04-16 08:54:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][1100/1251] eta 0:00:42 lr 0.000343 time 0.2620 (0.2837) loss 3.6754 (3.4016) grad_norm 1.7193 (inf) [2021-04-16 08:54:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][1110/1251] eta 0:00:39 lr 0.000343 time 0.3050 (0.2836) loss 3.0167 (3.4025) grad_norm 1.7155 (inf) [2021-04-16 08:54:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][1120/1251] eta 0:00:37 lr 0.000343 time 0.2726 (0.2836) loss 3.5539 (3.4023) grad_norm 1.9463 (inf) [2021-04-16 08:54:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][1130/1251] eta 0:00:34 lr 0.000343 time 0.2832 (0.2835) loss 3.4203 (3.4025) grad_norm 1.6588 (inf) [2021-04-16 08:54:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][1140/1251] eta 0:00:31 lr 0.000343 time 0.2666 (0.2835) loss 2.6341 (3.4033) grad_norm 2.3873 (inf) [2021-04-16 08:54:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][1150/1251] eta 0:00:28 lr 0.000343 time 0.2823 (0.2835) loss 3.7873 (3.4042) grad_norm 1.7754 (inf) [2021-04-16 08:54:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][1160/1251] eta 0:00:25 lr 0.000343 time 0.2618 (0.2836) loss 3.7183 (3.4063) grad_norm 1.7544 (inf) [2021-04-16 08:54:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][1170/1251] eta 0:00:22 lr 0.000343 time 0.2585 (0.2835) loss 4.0263 (3.4080) grad_norm 2.0650 (inf) [2021-04-16 08:54:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][1180/1251] eta 0:00:20 lr 0.000342 time 0.2540 (0.2835) loss 4.0203 (3.4102) grad_norm 1.6725 (inf) [2021-04-16 08:54:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][1190/1251] eta 0:00:17 lr 0.000342 time 0.4039 (0.2835) loss 3.9306 (3.4101) grad_norm 1.6974 (inf) [2021-04-16 08:54:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][1200/1251] eta 0:00:14 lr 0.000342 time 0.2694 (0.2835) loss 3.9932 (3.4113) grad_norm 1.7150 (inf) [2021-04-16 08:55:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][1210/1251] eta 0:00:11 lr 0.000342 time 0.2961 (0.2836) loss 3.9289 (3.4127) grad_norm 1.8583 (inf) [2021-04-16 08:55:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][1220/1251] eta 0:00:08 lr 0.000342 time 0.2678 (0.2835) loss 2.2889 (3.4128) grad_norm 2.2631 (inf) [2021-04-16 08:55:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][1230/1251] eta 0:00:05 lr 0.000342 time 0.2808 (0.2835) loss 3.4000 (3.4155) grad_norm 1.7393 (inf) [2021-04-16 08:55:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][1240/1251] eta 0:00:03 lr 0.000342 time 0.2410 (0.2834) loss 2.7880 (3.4150) grad_norm 1.7211 (inf) [2021-04-16 08:55:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [181/300][1250/1251] eta 0:00:00 lr 0.000342 time 0.2483 (0.2832) loss 3.7619 (3.4132) grad_norm 1.8187 (inf) [2021-04-16 08:55:15 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 181 training takes 0:05:57 [2021-04-16 08:55:15 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_181.pth saving...... [2021-04-16 08:55:23 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_181.pth saved !!! [2021-04-16 08:55:24 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.128 (1.128) Loss 0.9348 (0.9348) Acc@1 77.148 (77.148) Acc@5 93.652 (93.652) [2021-04-16 08:55:26 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.449 (0.285) Loss 0.9719 (0.9284) Acc@1 74.805 (77.477) Acc@5 94.824 (94.380) [2021-04-16 08:55:28 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.094 (0.210) Loss 0.8913 (0.9324) Acc@1 78.320 (77.567) Acc@5 94.531 (94.169) [2021-04-16 08:55:30 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.114 (0.228) Loss 0.8397 (0.9326) Acc@1 79.980 (77.627) Acc@5 94.727 (94.100) [2021-04-16 08:55:32 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.068 (0.213) Loss 0.9928 (0.9400) Acc@1 75.879 (77.544) Acc@5 93.750 (94.017) [2021-04-16 08:55:36 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 77.534 Acc@5 93.930 [2021-04-16 08:55:36 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 77.5% [2021-04-16 08:55:36 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 77.53% [2021-04-16 08:55:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][0/1251] eta 3:10:29 lr 0.000342 time 9.1365 (9.1365) loss 3.8490 (3.8490) grad_norm 1.9706 (1.9706) [2021-04-16 08:55:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][10/1251] eta 0:22:24 lr 0.000342 time 0.2903 (1.0835) loss 3.0660 (3.4346) grad_norm 1.7126 (1.7340) [2021-04-16 08:55:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][20/1251] eta 0:14:26 lr 0.000342 time 0.2606 (0.7038) loss 2.9867 (3.3283) grad_norm 1.5663 (1.7127) [2021-04-16 08:55:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][30/1251] eta 0:11:31 lr 0.000342 time 0.2831 (0.5662) loss 2.6590 (3.3577) grad_norm 2.0253 (1.7430) [2021-04-16 08:55:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][40/1251] eta 0:10:02 lr 0.000342 time 0.2992 (0.4979) loss 3.9187 (3.3996) grad_norm 1.7775 (1.7789) [2021-04-16 08:55:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][50/1251] eta 0:09:06 lr 0.000342 time 0.2799 (0.4546) loss 3.9080 (3.4524) grad_norm 1.6815 (1.8056) [2021-04-16 08:56:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][60/1251] eta 0:08:26 lr 0.000342 time 0.2823 (0.4255) loss 3.2030 (3.4443) grad_norm 2.0177 (1.8126) [2021-04-16 08:56:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][70/1251] eta 0:07:57 lr 0.000342 time 0.2807 (0.4043) loss 3.8504 (3.4230) grad_norm 1.8675 (1.8222) [2021-04-16 08:56:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][80/1251] eta 0:07:34 lr 0.000342 time 0.2925 (0.3885) loss 3.3482 (3.4188) grad_norm 1.7000 (1.8109) [2021-04-16 08:56:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][90/1251] eta 0:07:17 lr 0.000342 time 0.2634 (0.3772) loss 3.4573 (3.4117) grad_norm 1.5319 (1.8130) [2021-04-16 08:56:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][100/1251] eta 0:07:03 lr 0.000342 time 0.2996 (0.3677) loss 4.0398 (3.4296) grad_norm 1.9107 (1.8271) [2021-04-16 08:56:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][110/1251] eta 0:06:50 lr 0.000342 time 0.2827 (0.3596) loss 3.9488 (3.4279) grad_norm 1.9132 (1.8374) [2021-04-16 08:56:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][120/1251] eta 0:06:39 lr 0.000342 time 0.2975 (0.3531) loss 3.2016 (3.4335) grad_norm 1.8402 (1.8363) [2021-04-16 08:56:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][130/1251] eta 0:06:30 lr 0.000342 time 0.2764 (0.3481) loss 3.9801 (3.4321) grad_norm 1.8598 (1.8320) [2021-04-16 08:56:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][140/1251] eta 0:06:22 lr 0.000342 time 0.2735 (0.3443) loss 3.4782 (3.4441) grad_norm 1.7455 (1.8310) [2021-04-16 08:56:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][150/1251] eta 0:06:15 lr 0.000342 time 0.3120 (0.3415) loss 2.5305 (3.4308) grad_norm 2.1375 (1.8373) [2021-04-16 08:56:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][160/1251] eta 0:06:08 lr 0.000342 time 0.2818 (0.3377) loss 3.4697 (3.4420) grad_norm 1.9887 (1.8378) [2021-04-16 08:56:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][170/1251] eta 0:06:01 lr 0.000342 time 0.2809 (0.3343) loss 3.5482 (3.4623) grad_norm 1.6263 (1.8346) [2021-04-16 08:56:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][180/1251] eta 0:05:54 lr 0.000342 time 0.2719 (0.3311) loss 2.9962 (3.4631) grad_norm 1.7887 (1.8320) [2021-04-16 08:56:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][190/1251] eta 0:05:48 lr 0.000341 time 0.3123 (0.3285) loss 2.2097 (3.4735) grad_norm 1.5583 (1.8288) [2021-04-16 08:56:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][200/1251] eta 0:05:42 lr 0.000341 time 0.2826 (0.3258) loss 3.7137 (3.4836) grad_norm 1.8096 (1.8222) [2021-04-16 08:56:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][210/1251] eta 0:05:36 lr 0.000341 time 0.2557 (0.3233) loss 2.9080 (3.4820) grad_norm 1.9961 (1.8201) [2021-04-16 08:56:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][220/1251] eta 0:05:31 lr 0.000341 time 0.2675 (0.3212) loss 3.6882 (3.4842) grad_norm 1.8481 (1.8266) [2021-04-16 08:56:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][230/1251] eta 0:05:25 lr 0.000341 time 0.2684 (0.3192) loss 3.9292 (3.4913) grad_norm 1.9971 (1.8268) [2021-04-16 08:56:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][240/1251] eta 0:05:21 lr 0.000341 time 0.2979 (0.3176) loss 4.0287 (3.4925) grad_norm 2.0307 (1.8256) [2021-04-16 08:56:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][250/1251] eta 0:05:16 lr 0.000341 time 0.3048 (0.3167) loss 2.8757 (3.4815) grad_norm 1.6214 (1.8231) [2021-04-16 08:56:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][260/1251] eta 0:05:12 lr 0.000341 time 0.2658 (0.3155) loss 3.6010 (3.4773) grad_norm 1.7569 (1.8211) [2021-04-16 08:57:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][270/1251] eta 0:05:08 lr 0.000341 time 0.2677 (0.3140) loss 2.5985 (3.4879) grad_norm 1.8717 (1.8201) [2021-04-16 08:57:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][280/1251] eta 0:05:03 lr 0.000341 time 0.2754 (0.3125) loss 3.8687 (3.4864) grad_norm 1.7538 (1.8201) [2021-04-16 08:57:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][290/1251] eta 0:04:59 lr 0.000341 time 0.2921 (0.3120) loss 4.1509 (3.4809) grad_norm 1.7299 (1.8204) [2021-04-16 08:57:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][300/1251] eta 0:04:55 lr 0.000341 time 0.2495 (0.3106) loss 3.8932 (3.4786) grad_norm 1.7937 (1.8231) [2021-04-16 08:57:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][310/1251] eta 0:04:51 lr 0.000341 time 0.3071 (0.3096) loss 2.4247 (3.4697) grad_norm 1.9418 (1.8263) [2021-04-16 08:57:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][320/1251] eta 0:04:47 lr 0.000341 time 0.2685 (0.3087) loss 3.0038 (3.4521) grad_norm 1.7465 (1.8243) [2021-04-16 08:57:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][330/1251] eta 0:04:43 lr 0.000341 time 0.3031 (0.3080) loss 3.5326 (3.4460) grad_norm 1.8806 (1.8233) [2021-04-16 08:57:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][340/1251] eta 0:04:39 lr 0.000341 time 0.2876 (0.3071) loss 3.6293 (3.4468) grad_norm 1.7486 (1.8220) [2021-04-16 08:57:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][350/1251] eta 0:04:35 lr 0.000341 time 0.2817 (0.3063) loss 3.5890 (3.4486) grad_norm 1.8626 (1.8207) [2021-04-16 08:57:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][360/1251] eta 0:04:32 lr 0.000341 time 0.2558 (0.3059) loss 2.9110 (3.4498) grad_norm 1.6975 (1.8194) [2021-04-16 08:57:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][370/1251] eta 0:04:29 lr 0.000341 time 0.2568 (0.3058) loss 2.4276 (3.4458) grad_norm 2.0272 (1.8212) [2021-04-16 08:57:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][380/1251] eta 0:04:25 lr 0.000341 time 0.3004 (0.3052) loss 3.5060 (3.4427) grad_norm 1.7974 (1.8233) [2021-04-16 08:57:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][390/1251] eta 0:04:22 lr 0.000341 time 0.2609 (0.3049) loss 2.5631 (3.4378) grad_norm 1.7553 (1.8281) [2021-04-16 08:57:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][400/1251] eta 0:04:18 lr 0.000341 time 0.2793 (0.3041) loss 4.4690 (3.4400) grad_norm 1.7332 (1.8265) [2021-04-16 08:57:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][410/1251] eta 0:04:15 lr 0.000341 time 0.2786 (0.3034) loss 3.8754 (3.4426) grad_norm 1.8375 (1.8259) [2021-04-16 08:57:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][420/1251] eta 0:04:11 lr 0.000341 time 0.2625 (0.3029) loss 3.8564 (3.4411) grad_norm 1.5718 (1.8235) [2021-04-16 08:57:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][430/1251] eta 0:04:08 lr 0.000341 time 0.2717 (0.3025) loss 2.4522 (3.4326) grad_norm 1.5665 (1.8217) [2021-04-16 08:57:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][440/1251] eta 0:04:04 lr 0.000340 time 0.2956 (0.3019) loss 3.5298 (3.4328) grad_norm 1.6821 (1.8201) [2021-04-16 08:57:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][450/1251] eta 0:04:01 lr 0.000340 time 0.3067 (0.3014) loss 3.4968 (3.4384) grad_norm 1.4947 (1.8183) [2021-04-16 08:57:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][460/1251] eta 0:03:58 lr 0.000340 time 0.3079 (0.3009) loss 3.6351 (3.4403) grad_norm 1.7353 (1.8170) [2021-04-16 08:57:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][470/1251] eta 0:03:54 lr 0.000340 time 0.2770 (0.3004) loss 3.7219 (3.4357) grad_norm 1.5539 (1.8165) [2021-04-16 08:58:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][480/1251] eta 0:03:51 lr 0.000340 time 0.2716 (0.2999) loss 3.8885 (3.4392) grad_norm 1.6069 (1.8145) [2021-04-16 08:58:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][490/1251] eta 0:03:47 lr 0.000340 time 0.2877 (0.2995) loss 3.7557 (3.4371) grad_norm 1.6440 (1.8124) [2021-04-16 08:58:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][500/1251] eta 0:03:44 lr 0.000340 time 0.2652 (0.2992) loss 3.6872 (3.4373) grad_norm 1.9240 (1.8131) [2021-04-16 08:58:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][510/1251] eta 0:03:41 lr 0.000340 time 0.2917 (0.2988) loss 3.4210 (3.4397) grad_norm 2.5433 (1.8151) [2021-04-16 08:58:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][520/1251] eta 0:03:38 lr 0.000340 time 0.2792 (0.2984) loss 4.1719 (3.4461) grad_norm 1.8702 (1.8156) [2021-04-16 08:58:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][530/1251] eta 0:03:34 lr 0.000340 time 0.2917 (0.2980) loss 3.0773 (3.4442) grad_norm 2.1026 (1.8162) [2021-04-16 08:58:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][540/1251] eta 0:03:31 lr 0.000340 time 0.2583 (0.2975) loss 3.6279 (3.4433) grad_norm 1.7944 (1.8179) [2021-04-16 08:58:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][550/1251] eta 0:03:28 lr 0.000340 time 0.2829 (0.2973) loss 3.7914 (3.4442) grad_norm 1.6715 (1.8165) [2021-04-16 08:58:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][560/1251] eta 0:03:25 lr 0.000340 time 0.2568 (0.2970) loss 3.5028 (3.4437) grad_norm 2.0310 (1.8155) [2021-04-16 08:58:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][570/1251] eta 0:03:22 lr 0.000340 time 0.2555 (0.2966) loss 3.8152 (3.4419) grad_norm 1.8253 (1.8145) [2021-04-16 08:58:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][580/1251] eta 0:03:19 lr 0.000340 time 0.2837 (0.2967) loss 2.8042 (3.4465) grad_norm 2.0027 (1.8160) [2021-04-16 08:58:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][590/1251] eta 0:03:15 lr 0.000340 time 0.2974 (0.2964) loss 4.0917 (3.4474) grad_norm 1.6999 (1.8167) [2021-04-16 08:58:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][600/1251] eta 0:03:12 lr 0.000340 time 0.3072 (0.2961) loss 4.3444 (3.4494) grad_norm 2.0659 (1.8189) [2021-04-16 08:58:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][610/1251] eta 0:03:09 lr 0.000340 time 0.2761 (0.2958) loss 3.6329 (3.4511) grad_norm 1.6192 (1.8204) [2021-04-16 08:58:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][620/1251] eta 0:03:06 lr 0.000340 time 0.2893 (0.2955) loss 3.7124 (3.4493) grad_norm 1.5291 (1.8211) [2021-04-16 08:58:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][630/1251] eta 0:03:03 lr 0.000340 time 0.2854 (0.2952) loss 3.4119 (3.4512) grad_norm 1.7903 (1.8211) [2021-04-16 08:58:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][640/1251] eta 0:03:00 lr 0.000340 time 0.2692 (0.2949) loss 3.9427 (3.4496) grad_norm 1.5657 (1.8203) [2021-04-16 08:58:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][650/1251] eta 0:02:57 lr 0.000340 time 0.2715 (0.2949) loss 3.5459 (3.4508) grad_norm 1.6768 (1.8210) [2021-04-16 08:58:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][660/1251] eta 0:02:54 lr 0.000340 time 0.2417 (0.2945) loss 3.9366 (3.4479) grad_norm 1.8413 (1.8204) [2021-04-16 08:58:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][670/1251] eta 0:02:51 lr 0.000340 time 0.2767 (0.2945) loss 3.2726 (3.4507) grad_norm 1.8382 (1.8209) [2021-04-16 08:58:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][680/1251] eta 0:02:48 lr 0.000340 time 0.3108 (0.2943) loss 3.1068 (3.4498) grad_norm 1.6059 (1.8219) [2021-04-16 08:58:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][690/1251] eta 0:02:44 lr 0.000340 time 0.2713 (0.2941) loss 3.8648 (3.4547) grad_norm 1.5449 (1.8198) [2021-04-16 08:59:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][700/1251] eta 0:02:41 lr 0.000339 time 0.2918 (0.2938) loss 4.0042 (3.4526) grad_norm 1.9470 (1.8206) [2021-04-16 08:59:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][710/1251] eta 0:02:38 lr 0.000339 time 0.4003 (0.2939) loss 3.8214 (3.4455) grad_norm 1.9371 (1.8193) [2021-04-16 08:59:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][720/1251] eta 0:02:36 lr 0.000339 time 0.2798 (0.2939) loss 4.5445 (3.4466) grad_norm 1.7039 (1.8192) [2021-04-16 08:59:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][730/1251] eta 0:02:32 lr 0.000339 time 0.2731 (0.2936) loss 3.8556 (3.4429) grad_norm 1.8621 (1.8199) [2021-04-16 08:59:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][740/1251] eta 0:02:29 lr 0.000339 time 0.2653 (0.2934) loss 3.6797 (3.4398) grad_norm 1.8229 (1.8204) [2021-04-16 08:59:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][750/1251] eta 0:02:26 lr 0.000339 time 0.2913 (0.2933) loss 3.6429 (3.4413) grad_norm 1.6837 (1.8198) [2021-04-16 08:59:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][760/1251] eta 0:02:23 lr 0.000339 time 0.2773 (0.2931) loss 3.5154 (3.4433) grad_norm 2.0892 (1.8199) [2021-04-16 08:59:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][770/1251] eta 0:02:20 lr 0.000339 time 0.2948 (0.2929) loss 3.5640 (3.4468) grad_norm 1.7289 (1.8194) [2021-04-16 08:59:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][780/1251] eta 0:02:17 lr 0.000339 time 0.2969 (0.2927) loss 3.7232 (3.4472) grad_norm 2.0532 (1.8206) [2021-04-16 08:59:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][790/1251] eta 0:02:14 lr 0.000339 time 0.2618 (0.2924) loss 3.1419 (3.4455) grad_norm 1.7134 (1.8208) [2021-04-16 08:59:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][800/1251] eta 0:02:11 lr 0.000339 time 0.2621 (0.2922) loss 3.2319 (3.4462) grad_norm 1.4681 (1.8206) [2021-04-16 08:59:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][810/1251] eta 0:02:08 lr 0.000339 time 0.2929 (0.2921) loss 3.0798 (3.4456) grad_norm 1.9782 (1.8213) [2021-04-16 08:59:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][820/1251] eta 0:02:05 lr 0.000339 time 0.2780 (0.2919) loss 3.4701 (3.4444) grad_norm 1.6976 (1.8231) [2021-04-16 08:59:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][830/1251] eta 0:02:02 lr 0.000339 time 0.2570 (0.2919) loss 3.2041 (3.4402) grad_norm 2.1754 (1.8248) [2021-04-16 08:59:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][840/1251] eta 0:01:59 lr 0.000339 time 0.2577 (0.2917) loss 3.6519 (3.4410) grad_norm 1.7287 (1.8256) [2021-04-16 08:59:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][850/1251] eta 0:01:56 lr 0.000339 time 0.2809 (0.2915) loss 3.5617 (3.4399) grad_norm 1.6927 (1.8251) [2021-04-16 08:59:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][860/1251] eta 0:01:53 lr 0.000339 time 0.2770 (0.2913) loss 3.5839 (3.4395) grad_norm 1.7875 (1.8245) [2021-04-16 08:59:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][870/1251] eta 0:01:50 lr 0.000339 time 0.2567 (0.2912) loss 3.7643 (3.4392) grad_norm 1.7854 (1.8230) [2021-04-16 08:59:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][880/1251] eta 0:01:47 lr 0.000339 time 0.2802 (0.2910) loss 3.2962 (3.4387) grad_norm 1.7559 (1.8227) [2021-04-16 08:59:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][890/1251] eta 0:01:44 lr 0.000339 time 0.2901 (0.2908) loss 2.1691 (3.4342) grad_norm 1.7207 (1.8227) [2021-04-16 08:59:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][900/1251] eta 0:01:42 lr 0.000339 time 0.2818 (0.2906) loss 2.7724 (3.4326) grad_norm 2.0358 (1.8220) [2021-04-16 09:00:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][910/1251] eta 0:01:39 lr 0.000339 time 0.2821 (0.2904) loss 2.3840 (3.4341) grad_norm 2.1407 (1.8219) [2021-04-16 09:00:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][920/1251] eta 0:01:36 lr 0.000339 time 0.2524 (0.2902) loss 4.1352 (3.4349) grad_norm 1.6770 (1.8220) [2021-04-16 09:00:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][930/1251] eta 0:01:33 lr 0.000339 time 0.2705 (0.2901) loss 3.4938 (3.4335) grad_norm 1.4830 (1.8209) [2021-04-16 09:00:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][940/1251] eta 0:01:30 lr 0.000339 time 0.3961 (0.2902) loss 3.4656 (3.4332) grad_norm 1.9057 (1.8214) [2021-04-16 09:00:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][950/1251] eta 0:01:27 lr 0.000338 time 0.2660 (0.2900) loss 2.4977 (3.4314) grad_norm 1.6676 (1.8219) [2021-04-16 09:00:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][960/1251] eta 0:01:24 lr 0.000338 time 0.2876 (0.2899) loss 3.5439 (3.4321) grad_norm 1.4960 (1.8213) [2021-04-16 09:00:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][970/1251] eta 0:01:21 lr 0.000338 time 0.2854 (0.2899) loss 3.4243 (3.4354) grad_norm 1.7389 (1.8209) [2021-04-16 09:00:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][980/1251] eta 0:01:18 lr 0.000338 time 0.2793 (0.2897) loss 2.6374 (3.4337) grad_norm 1.7085 (1.8204) [2021-04-16 09:00:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][990/1251] eta 0:01:15 lr 0.000338 time 0.2543 (0.2896) loss 3.6374 (3.4331) grad_norm 1.6364 (1.8207) [2021-04-16 09:00:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][1000/1251] eta 0:01:12 lr 0.000338 time 0.2683 (0.2894) loss 2.4810 (3.4309) grad_norm 1.8128 (1.8199) [2021-04-16 09:00:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][1010/1251] eta 0:01:09 lr 0.000338 time 0.2662 (0.2892) loss 3.3456 (3.4296) grad_norm 1.4412 (1.8193) [2021-04-16 09:00:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][1020/1251] eta 0:01:06 lr 0.000338 time 0.2966 (0.2891) loss 3.3270 (3.4257) grad_norm 1.6351 (1.8189) [2021-04-16 09:00:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][1030/1251] eta 0:01:03 lr 0.000338 time 0.2820 (0.2891) loss 2.3291 (3.4250) grad_norm 2.1283 (1.8188) [2021-04-16 09:00:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][1040/1251] eta 0:01:00 lr 0.000338 time 0.2887 (0.2889) loss 4.0439 (3.4263) grad_norm 1.6909 (1.8193) [2021-04-16 09:00:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][1050/1251] eta 0:00:58 lr 0.000338 time 0.2782 (0.2888) loss 2.6348 (3.4246) grad_norm 1.6493 (1.8188) [2021-04-16 09:00:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][1060/1251] eta 0:00:55 lr 0.000338 time 0.2704 (0.2887) loss 2.2742 (3.4225) grad_norm 1.7850 (1.8186) [2021-04-16 09:00:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][1070/1251] eta 0:00:52 lr 0.000338 time 0.2770 (0.2886) loss 2.5354 (3.4228) grad_norm 1.7286 (1.8178) [2021-04-16 09:00:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][1080/1251] eta 0:00:49 lr 0.000338 time 0.3013 (0.2886) loss 3.8193 (3.4254) grad_norm 1.6425 (1.8172) [2021-04-16 09:00:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][1090/1251] eta 0:00:46 lr 0.000338 time 0.2883 (0.2885) loss 3.3063 (3.4244) grad_norm 1.5113 (1.8165) [2021-04-16 09:00:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][1100/1251] eta 0:00:43 lr 0.000338 time 0.2656 (0.2883) loss 3.6397 (3.4255) grad_norm 1.4274 (1.8163) [2021-04-16 09:00:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][1110/1251] eta 0:00:40 lr 0.000338 time 0.2602 (0.2882) loss 2.9935 (3.4246) grad_norm 1.7302 (1.8151) [2021-04-16 09:00:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][1120/1251] eta 0:00:37 lr 0.000338 time 0.2647 (0.2881) loss 3.7829 (3.4239) grad_norm 2.1149 (1.8149) [2021-04-16 09:01:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][1130/1251] eta 0:00:34 lr 0.000338 time 0.2892 (0.2880) loss 3.8267 (3.4230) grad_norm 1.7610 (1.8148) [2021-04-16 09:01:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][1140/1251] eta 0:00:31 lr 0.000338 time 0.2815 (0.2879) loss 3.6181 (3.4252) grad_norm 2.0348 (1.8156) [2021-04-16 09:01:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][1150/1251] eta 0:00:29 lr 0.000338 time 0.2900 (0.2878) loss 2.6036 (3.4248) grad_norm 1.7940 (1.8148) [2021-04-16 09:01:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][1160/1251] eta 0:00:26 lr 0.000338 time 0.2667 (0.2878) loss 2.2467 (3.4231) grad_norm 1.7657 (1.8145) [2021-04-16 09:01:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][1170/1251] eta 0:00:23 lr 0.000338 time 0.2712 (0.2877) loss 3.2815 (3.4215) grad_norm 1.8592 (1.8143) [2021-04-16 09:01:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][1180/1251] eta 0:00:20 lr 0.000338 time 0.2874 (0.2876) loss 3.7973 (3.4216) grad_norm 1.5751 (1.8134) [2021-04-16 09:01:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][1190/1251] eta 0:00:17 lr 0.000338 time 0.2474 (0.2875) loss 4.0540 (3.4198) grad_norm 1.7855 (1.8127) [2021-04-16 09:01:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][1200/1251] eta 0:00:14 lr 0.000338 time 0.2739 (0.2874) loss 4.0628 (3.4206) grad_norm 2.4091 (1.8135) [2021-04-16 09:01:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][1210/1251] eta 0:00:11 lr 0.000337 time 0.2578 (0.2873) loss 3.0282 (3.4174) grad_norm 1.9679 (1.8135) [2021-04-16 09:01:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][1220/1251] eta 0:00:08 lr 0.000337 time 0.2714 (0.2872) loss 3.4671 (3.4178) grad_norm 1.7267 (1.8130) [2021-04-16 09:01:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][1230/1251] eta 0:00:06 lr 0.000337 time 0.2568 (0.2871) loss 4.4264 (3.4169) grad_norm 1.6971 (1.8124) [2021-04-16 09:01:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][1240/1251] eta 0:00:03 lr 0.000337 time 0.2498 (0.2871) loss 3.7784 (3.4184) grad_norm 2.0293 (1.8120) [2021-04-16 09:01:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [182/300][1250/1251] eta 0:00:00 lr 0.000337 time 0.2485 (0.2868) loss 2.6245 (3.4170) grad_norm 2.0068 (1.8120) [2021-04-16 09:01:39 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 182 training takes 0:06:02 [2021-04-16 09:01:39 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_182.pth saving...... [2021-04-16 09:01:54 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_182.pth saved !!! [2021-04-16 09:01:55 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.059 (1.059) Loss 0.9917 (0.9917) Acc@1 75.977 (75.977) Acc@5 94.336 (94.336) [2021-04-16 09:01:57 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.119 (0.260) Loss 0.9469 (0.9689) Acc@1 77.246 (77.175) Acc@5 93.848 (94.043) [2021-04-16 09:01:59 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.118 (0.219) Loss 1.0203 (0.9686) Acc@1 76.270 (77.367) Acc@5 92.969 (94.062) [2021-04-16 09:02:01 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.120 (0.232) Loss 0.9345 (0.9702) Acc@1 79.492 (77.385) Acc@5 94.824 (94.065) [2021-04-16 09:02:03 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.410 (0.211) Loss 0.9545 (0.9642) Acc@1 77.930 (77.525) Acc@5 93.750 (94.107) [2021-04-16 09:02:10 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 77.486 Acc@5 94.046 [2021-04-16 09:02:10 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 77.5% [2021-04-16 09:02:10 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 77.53% [2021-04-16 09:02:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][0/1251] eta 4:15:47 lr 0.000337 time 12.2684 (12.2684) loss 3.9988 (3.9988) grad_norm 1.6473 (1.6473) [2021-04-16 09:02:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][10/1251] eta 0:28:09 lr 0.000337 time 0.2930 (1.3612) loss 3.5390 (3.4553) grad_norm 1.9256 (1.7008) [2021-04-16 09:02:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][20/1251] eta 0:17:22 lr 0.000337 time 0.2735 (0.8469) loss 3.3524 (3.4145) grad_norm 1.8278 (1.7228) [2021-04-16 09:02:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][30/1251] eta 0:13:29 lr 0.000337 time 0.2904 (0.6630) loss 3.3348 (3.4409) grad_norm 2.0614 (1.7654) [2021-04-16 09:02:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][40/1251] eta 0:11:32 lr 0.000337 time 0.2512 (0.5716) loss 3.7550 (3.4349) grad_norm 2.2095 (1.7979) [2021-04-16 09:02:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][50/1251] eta 0:10:18 lr 0.000337 time 0.2602 (0.5148) loss 3.7387 (3.4551) grad_norm 1.7278 (1.8355) [2021-04-16 09:02:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][60/1251] eta 0:09:28 lr 0.000337 time 0.3587 (0.4770) loss 3.6606 (3.4396) grad_norm 1.7406 (1.8417) [2021-04-16 09:02:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][70/1251] eta 0:08:50 lr 0.000337 time 0.3041 (0.4493) loss 4.1763 (3.4566) grad_norm 1.7951 (1.8323) [2021-04-16 09:02:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][80/1251] eta 0:08:22 lr 0.000337 time 0.2593 (0.4293) loss 3.2595 (3.4346) grad_norm 1.8107 (1.8147) [2021-04-16 09:02:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][90/1251] eta 0:07:59 lr 0.000337 time 0.2771 (0.4127) loss 3.6900 (3.4425) grad_norm 2.5305 (1.8188) [2021-04-16 09:02:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][100/1251] eta 0:07:41 lr 0.000337 time 0.2987 (0.4005) loss 3.7485 (3.4665) grad_norm 2.1094 (1.8226) [2021-04-16 09:02:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][110/1251] eta 0:07:24 lr 0.000337 time 0.2939 (0.3898) loss 3.9998 (3.4467) grad_norm 1.7457 (1.8132) [2021-04-16 09:02:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][120/1251] eta 0:07:10 lr 0.000337 time 0.2632 (0.3807) loss 2.8874 (3.4338) grad_norm 1.5640 (1.8154) [2021-04-16 09:02:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][130/1251] eta 0:06:58 lr 0.000337 time 0.2734 (0.3737) loss 3.5979 (3.4259) grad_norm 1.7355 (1.8088) [2021-04-16 09:03:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][140/1251] eta 0:06:50 lr 0.000337 time 0.2732 (0.3693) loss 2.8562 (3.4279) grad_norm 2.1646 (1.8104) [2021-04-16 09:03:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][150/1251] eta 0:06:39 lr 0.000337 time 0.2557 (0.3631) loss 3.1898 (3.4113) grad_norm 1.7870 (1.8093) [2021-04-16 09:03:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][160/1251] eta 0:06:30 lr 0.000337 time 0.2507 (0.3577) loss 2.7469 (3.4023) grad_norm 2.0362 (1.8058) [2021-04-16 09:03:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][170/1251] eta 0:06:21 lr 0.000337 time 0.2813 (0.3529) loss 3.2454 (3.4049) grad_norm 1.7244 (1.8042) [2021-04-16 09:03:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][180/1251] eta 0:06:13 lr 0.000337 time 0.2702 (0.3492) loss 3.8344 (3.4162) grad_norm 1.6878 (1.8025) [2021-04-16 09:03:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][190/1251] eta 0:06:06 lr 0.000337 time 0.3136 (0.3456) loss 3.2437 (3.3917) grad_norm 1.9962 (1.7999) [2021-04-16 09:03:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][200/1251] eta 0:05:59 lr 0.000337 time 0.2661 (0.3420) loss 2.8708 (3.3995) grad_norm 1.9064 (1.8029) [2021-04-16 09:03:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][210/1251] eta 0:05:52 lr 0.000337 time 0.2865 (0.3388) loss 3.0052 (3.3923) grad_norm 1.9292 (1.8025) [2021-04-16 09:03:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][220/1251] eta 0:05:46 lr 0.000336 time 0.2917 (0.3361) loss 2.8551 (3.3902) grad_norm 1.7488 (1.8008) [2021-04-16 09:03:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][230/1251] eta 0:05:40 lr 0.000336 time 0.2709 (0.3337) loss 3.5909 (3.3891) grad_norm 2.0754 (1.8005) [2021-04-16 09:03:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][240/1251] eta 0:05:35 lr 0.000336 time 0.2927 (0.3316) loss 2.7022 (3.3952) grad_norm 1.7613 (1.8023) [2021-04-16 09:03:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][250/1251] eta 0:05:29 lr 0.000336 time 0.2808 (0.3295) loss 3.7868 (3.3965) grad_norm 2.0819 (1.8031) [2021-04-16 09:03:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][260/1251] eta 0:05:24 lr 0.000336 time 0.2777 (0.3275) loss 2.6244 (3.3973) grad_norm 1.9410 (1.8010) [2021-04-16 09:03:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][270/1251] eta 0:05:19 lr 0.000336 time 0.2899 (0.3258) loss 3.9985 (3.3932) grad_norm 2.2955 (1.8055) [2021-04-16 09:03:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][280/1251] eta 0:05:14 lr 0.000336 time 0.2738 (0.3243) loss 3.4318 (3.3931) grad_norm 1.5605 (1.8014) [2021-04-16 09:03:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][290/1251] eta 0:05:10 lr 0.000336 time 0.2941 (0.3227) loss 3.7603 (3.3936) grad_norm 1.7514 (1.7995) [2021-04-16 09:03:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][300/1251] eta 0:05:05 lr 0.000336 time 0.2706 (0.3212) loss 3.1217 (3.3874) grad_norm 1.8343 (1.8046) [2021-04-16 09:03:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][310/1251] eta 0:05:00 lr 0.000336 time 0.2863 (0.3198) loss 3.1570 (3.3976) grad_norm 1.8308 (1.8056) [2021-04-16 09:03:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][320/1251] eta 0:04:56 lr 0.000336 time 0.2809 (0.3184) loss 4.0031 (3.3956) grad_norm 1.8125 (1.8049) [2021-04-16 09:03:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][330/1251] eta 0:04:52 lr 0.000336 time 0.2813 (0.3173) loss 3.6489 (3.3979) grad_norm 2.0477 (1.8047) [2021-04-16 09:03:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][340/1251] eta 0:04:47 lr 0.000336 time 0.2787 (0.3160) loss 4.0282 (3.3986) grad_norm 1.8300 (1.8101) [2021-04-16 09:04:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][350/1251] eta 0:04:43 lr 0.000336 time 0.3042 (0.3149) loss 2.9510 (3.3941) grad_norm 1.9417 (1.8135) [2021-04-16 09:04:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][360/1251] eta 0:04:39 lr 0.000336 time 0.2488 (0.3140) loss 3.0601 (3.3854) grad_norm 1.9306 (1.8147) [2021-04-16 09:04:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][370/1251] eta 0:04:35 lr 0.000336 time 0.2646 (0.3132) loss 3.6579 (3.3858) grad_norm 1.7346 (1.8166) [2021-04-16 09:04:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][380/1251] eta 0:04:31 lr 0.000336 time 0.2775 (0.3123) loss 3.7450 (3.3879) grad_norm 1.9658 (1.8177) [2021-04-16 09:04:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][390/1251] eta 0:04:28 lr 0.000336 time 0.2982 (0.3115) loss 3.1158 (3.3891) grad_norm 1.4696 (1.8175) [2021-04-16 09:04:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][400/1251] eta 0:04:24 lr 0.000336 time 0.2763 (0.3107) loss 3.8384 (3.3941) grad_norm 1.5218 (1.8190) [2021-04-16 09:04:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][410/1251] eta 0:04:20 lr 0.000336 time 0.2985 (0.3100) loss 4.1101 (3.4008) grad_norm 1.5650 (inf) [2021-04-16 09:04:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][420/1251] eta 0:04:16 lr 0.000336 time 0.2611 (0.3091) loss 3.1027 (3.3948) grad_norm 2.1873 (inf) [2021-04-16 09:04:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][430/1251] eta 0:04:13 lr 0.000336 time 0.2463 (0.3084) loss 2.0316 (3.3884) grad_norm 1.5164 (inf) [2021-04-16 09:04:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][440/1251] eta 0:04:09 lr 0.000336 time 0.2701 (0.3078) loss 2.4602 (3.3851) grad_norm 1.7320 (inf) [2021-04-16 09:04:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][450/1251] eta 0:04:06 lr 0.000336 time 0.3011 (0.3072) loss 3.8967 (3.3849) grad_norm 1.7741 (inf) [2021-04-16 09:04:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][460/1251] eta 0:04:02 lr 0.000336 time 0.2675 (0.3065) loss 3.6450 (3.3865) grad_norm 1.8549 (inf) [2021-04-16 09:04:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][470/1251] eta 0:03:58 lr 0.000335 time 0.2567 (0.3059) loss 4.2396 (3.3900) grad_norm 1.9037 (inf) [2021-04-16 09:04:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][480/1251] eta 0:03:55 lr 0.000335 time 0.2943 (0.3053) loss 2.9670 (3.3894) grad_norm 1.8356 (inf) [2021-04-16 09:04:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][490/1251] eta 0:03:51 lr 0.000335 time 0.2658 (0.3048) loss 2.6330 (3.3871) grad_norm 1.8387 (inf) [2021-04-16 09:04:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][500/1251] eta 0:03:48 lr 0.000335 time 0.2769 (0.3042) loss 2.8011 (3.3833) grad_norm 1.6323 (inf) [2021-04-16 09:04:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][510/1251] eta 0:03:45 lr 0.000335 time 0.3021 (0.3038) loss 3.7224 (3.3836) grad_norm 1.8282 (inf) [2021-04-16 09:04:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][520/1251] eta 0:03:41 lr 0.000335 time 0.2781 (0.3035) loss 4.1528 (3.3793) grad_norm 2.0346 (inf) [2021-04-16 09:04:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][530/1251] eta 0:03:38 lr 0.000335 time 0.2741 (0.3029) loss 3.9421 (3.3738) grad_norm 2.0119 (inf) [2021-04-16 09:04:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][540/1251] eta 0:03:35 lr 0.000335 time 0.3007 (0.3026) loss 3.1922 (3.3688) grad_norm 1.7865 (inf) [2021-04-16 09:04:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][550/1251] eta 0:03:31 lr 0.000335 time 0.2671 (0.3022) loss 4.7055 (3.3743) grad_norm 1.9862 (inf) [2021-04-16 09:04:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][560/1251] eta 0:03:28 lr 0.000335 time 0.2699 (0.3016) loss 3.5015 (3.3776) grad_norm 1.8107 (inf) [2021-04-16 09:05:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][570/1251] eta 0:03:25 lr 0.000335 time 0.2691 (0.3014) loss 3.3933 (3.3754) grad_norm 1.6223 (inf) [2021-04-16 09:05:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][580/1251] eta 0:03:22 lr 0.000335 time 0.2663 (0.3015) loss 3.5301 (3.3783) grad_norm 1.6841 (inf) [2021-04-16 09:05:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][590/1251] eta 0:03:18 lr 0.000335 time 0.2718 (0.3011) loss 3.7936 (3.3786) grad_norm 2.3702 (inf) [2021-04-16 09:05:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][600/1251] eta 0:03:15 lr 0.000335 time 0.2831 (0.3007) loss 3.4828 (3.3780) grad_norm 1.6186 (inf) [2021-04-16 09:05:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][610/1251] eta 0:03:12 lr 0.000335 time 0.2855 (0.3003) loss 3.0885 (3.3767) grad_norm 1.8848 (inf) [2021-04-16 09:05:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][620/1251] eta 0:03:09 lr 0.000335 time 0.2770 (0.3001) loss 4.0930 (3.3760) grad_norm 2.0434 (inf) [2021-04-16 09:05:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][630/1251] eta 0:03:06 lr 0.000335 time 0.2841 (0.2998) loss 3.6666 (3.3719) grad_norm 1.8195 (inf) [2021-04-16 09:05:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][640/1251] eta 0:03:02 lr 0.000335 time 0.2675 (0.2995) loss 2.9968 (3.3717) grad_norm 1.6979 (inf) [2021-04-16 09:05:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][650/1251] eta 0:02:59 lr 0.000335 time 0.2724 (0.2994) loss 4.2603 (3.3724) grad_norm 1.7784 (inf) [2021-04-16 09:05:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][660/1251] eta 0:02:56 lr 0.000335 time 0.3001 (0.2990) loss 3.7272 (3.3747) grad_norm 1.8227 (inf) [2021-04-16 09:05:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][670/1251] eta 0:02:53 lr 0.000335 time 0.2828 (0.2987) loss 3.7426 (3.3778) grad_norm 1.9603 (inf) [2021-04-16 09:05:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][680/1251] eta 0:02:50 lr 0.000335 time 0.3027 (0.2985) loss 3.9143 (3.3778) grad_norm 1.7554 (inf) [2021-04-16 09:05:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][690/1251] eta 0:02:47 lr 0.000335 time 0.2515 (0.2982) loss 4.0188 (3.3781) grad_norm 2.0919 (inf) [2021-04-16 09:05:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][700/1251] eta 0:02:44 lr 0.000335 time 0.2449 (0.2981) loss 2.8402 (3.3732) grad_norm 1.8417 (inf) [2021-04-16 09:05:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][710/1251] eta 0:02:41 lr 0.000335 time 0.2755 (0.2977) loss 3.3638 (3.3721) grad_norm 2.0426 (inf) [2021-04-16 09:05:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][720/1251] eta 0:02:38 lr 0.000335 time 0.2572 (0.2978) loss 3.0534 (3.3748) grad_norm 1.7282 (inf) [2021-04-16 09:05:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][730/1251] eta 0:02:35 lr 0.000334 time 0.2929 (0.2976) loss 3.8652 (3.3781) grad_norm 1.6201 (inf) [2021-04-16 09:05:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][740/1251] eta 0:02:31 lr 0.000334 time 0.2909 (0.2974) loss 3.1061 (3.3792) grad_norm 1.7049 (inf) [2021-04-16 09:05:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][750/1251] eta 0:02:28 lr 0.000334 time 0.2901 (0.2972) loss 2.8555 (3.3783) grad_norm 1.7781 (inf) [2021-04-16 09:05:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][760/1251] eta 0:02:25 lr 0.000334 time 0.2932 (0.2970) loss 2.9051 (3.3742) grad_norm 1.8309 (inf) [2021-04-16 09:05:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][770/1251] eta 0:02:22 lr 0.000334 time 0.2719 (0.2968) loss 3.2216 (3.3758) grad_norm 1.9242 (inf) [2021-04-16 09:06:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][780/1251] eta 0:02:19 lr 0.000334 time 0.2642 (0.2965) loss 3.9793 (3.3790) grad_norm 1.8278 (inf) [2021-04-16 09:06:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][790/1251] eta 0:02:16 lr 0.000334 time 0.2897 (0.2963) loss 3.8318 (3.3797) grad_norm 1.6606 (inf) [2021-04-16 09:06:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][800/1251] eta 0:02:13 lr 0.000334 time 0.2563 (0.2960) loss 3.1179 (3.3821) grad_norm 1.8230 (inf) [2021-04-16 09:06:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][810/1251] eta 0:02:10 lr 0.000334 time 0.3080 (0.2958) loss 3.1937 (3.3845) grad_norm 1.7726 (inf) [2021-04-16 09:06:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][820/1251] eta 0:02:07 lr 0.000334 time 0.2906 (0.2956) loss 3.5584 (3.3880) grad_norm 1.6343 (inf) [2021-04-16 09:06:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][830/1251] eta 0:02:04 lr 0.000334 time 0.2734 (0.2954) loss 3.5963 (3.3930) grad_norm 2.0459 (inf) [2021-04-16 09:06:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][840/1251] eta 0:02:01 lr 0.000334 time 0.2898 (0.2952) loss 3.2155 (3.3962) grad_norm 1.7050 (inf) [2021-04-16 09:06:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][850/1251] eta 0:01:58 lr 0.000334 time 0.2609 (0.2950) loss 3.9718 (3.3982) grad_norm 1.7769 (inf) [2021-04-16 09:06:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][860/1251] eta 0:01:55 lr 0.000334 time 0.2919 (0.2948) loss 3.9159 (3.3986) grad_norm 2.0268 (inf) [2021-04-16 09:06:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][870/1251] eta 0:01:52 lr 0.000334 time 0.2854 (0.2947) loss 3.3638 (3.3976) grad_norm 1.9139 (inf) [2021-04-16 09:06:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][880/1251] eta 0:01:49 lr 0.000334 time 0.2761 (0.2945) loss 2.9779 (3.3999) grad_norm 1.6901 (inf) [2021-04-16 09:06:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][890/1251] eta 0:01:46 lr 0.000334 time 0.2890 (0.2942) loss 3.6433 (3.4006) grad_norm 1.5399 (inf) [2021-04-16 09:06:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][900/1251] eta 0:01:43 lr 0.000334 time 0.2591 (0.2940) loss 3.5220 (3.3990) grad_norm 1.7413 (inf) [2021-04-16 09:06:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][910/1251] eta 0:01:40 lr 0.000334 time 0.3021 (0.2939) loss 3.4907 (3.3987) grad_norm 1.8711 (inf) [2021-04-16 09:06:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][920/1251] eta 0:01:37 lr 0.000334 time 0.2843 (0.2937) loss 3.9562 (3.3968) grad_norm 1.8862 (inf) [2021-04-16 09:06:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][930/1251] eta 0:01:34 lr 0.000334 time 0.2900 (0.2938) loss 3.4845 (3.3946) grad_norm 1.7272 (inf) [2021-04-16 09:06:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][940/1251] eta 0:01:31 lr 0.000334 time 0.3026 (0.2937) loss 3.7257 (3.3912) grad_norm 1.6140 (inf) [2021-04-16 09:06:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][950/1251] eta 0:01:28 lr 0.000334 time 0.2890 (0.2935) loss 3.6051 (3.3935) grad_norm 1.6218 (inf) [2021-04-16 09:06:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][960/1251] eta 0:01:25 lr 0.000334 time 0.2821 (0.2933) loss 4.0276 (3.3969) grad_norm 1.8249 (inf) [2021-04-16 09:06:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][970/1251] eta 0:01:22 lr 0.000334 time 0.2937 (0.2931) loss 3.2845 (3.3960) grad_norm 1.8639 (inf) [2021-04-16 09:06:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][980/1251] eta 0:01:19 lr 0.000334 time 0.2950 (0.2929) loss 3.9356 (3.3981) grad_norm 1.5911 (inf) [2021-04-16 09:07:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][990/1251] eta 0:01:16 lr 0.000333 time 0.2630 (0.2927) loss 2.6042 (3.3979) grad_norm 2.0094 (inf) [2021-04-16 09:07:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][1000/1251] eta 0:01:13 lr 0.000333 time 0.2679 (0.2925) loss 2.1257 (3.3976) grad_norm 1.9059 (inf) [2021-04-16 09:07:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][1010/1251] eta 0:01:10 lr 0.000333 time 0.2579 (0.2923) loss 2.1617 (3.3949) grad_norm 1.7721 (inf) [2021-04-16 09:07:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][1020/1251] eta 0:01:07 lr 0.000333 time 0.2966 (0.2923) loss 3.0903 (3.3960) grad_norm 1.5971 (inf) [2021-04-16 09:07:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][1030/1251] eta 0:01:04 lr 0.000333 time 0.2886 (0.2922) loss 3.6605 (3.3961) grad_norm 1.8410 (inf) [2021-04-16 09:07:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][1040/1251] eta 0:01:01 lr 0.000333 time 0.3059 (0.2921) loss 3.9563 (3.3963) grad_norm 1.6150 (inf) [2021-04-16 09:07:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][1050/1251] eta 0:00:58 lr 0.000333 time 0.2688 (0.2919) loss 3.9279 (3.4009) grad_norm 1.9388 (inf) [2021-04-16 09:07:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][1060/1251] eta 0:00:55 lr 0.000333 time 0.2793 (0.2918) loss 4.1351 (3.4041) grad_norm 1.7401 (inf) [2021-04-16 09:07:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][1070/1251] eta 0:00:52 lr 0.000333 time 0.2815 (0.2917) loss 3.5162 (3.4032) grad_norm 1.4919 (inf) [2021-04-16 09:07:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][1080/1251] eta 0:00:49 lr 0.000333 time 0.2476 (0.2916) loss 2.3000 (3.4044) grad_norm 1.6626 (inf) [2021-04-16 09:07:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][1090/1251] eta 0:00:46 lr 0.000333 time 0.2864 (0.2916) loss 2.8359 (3.4033) grad_norm 1.9417 (inf) [2021-04-16 09:07:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][1100/1251] eta 0:00:44 lr 0.000333 time 0.2723 (0.2914) loss 3.0807 (3.3980) grad_norm 1.6843 (inf) [2021-04-16 09:07:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][1110/1251] eta 0:00:41 lr 0.000333 time 0.2795 (0.2913) loss 2.5460 (3.3941) grad_norm 1.6508 (inf) [2021-04-16 09:07:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][1120/1251] eta 0:00:38 lr 0.000333 time 0.2938 (0.2911) loss 3.7320 (3.3950) grad_norm 1.9465 (inf) [2021-04-16 09:07:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][1130/1251] eta 0:00:35 lr 0.000333 time 0.2484 (0.2910) loss 3.3354 (3.3945) grad_norm 1.7588 (inf) [2021-04-16 09:07:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][1140/1251] eta 0:00:32 lr 0.000333 time 0.2801 (0.2909) loss 3.3555 (3.3936) grad_norm 1.6503 (inf) [2021-04-16 09:07:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][1150/1251] eta 0:00:29 lr 0.000333 time 0.2641 (0.2909) loss 3.3228 (3.3928) grad_norm 1.9193 (inf) [2021-04-16 09:07:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][1160/1251] eta 0:00:26 lr 0.000333 time 0.3060 (0.2909) loss 3.3838 (3.3904) grad_norm 1.7951 (inf) [2021-04-16 09:07:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][1170/1251] eta 0:00:23 lr 0.000333 time 0.2637 (0.2907) loss 2.3893 (3.3895) grad_norm 1.5357 (inf) [2021-04-16 09:07:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][1180/1251] eta 0:00:20 lr 0.000333 time 0.2687 (0.2906) loss 2.5830 (3.3872) grad_norm 1.9940 (inf) [2021-04-16 09:07:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][1190/1251] eta 0:00:17 lr 0.000333 time 0.3931 (0.2905) loss 3.7219 (3.3869) grad_norm 1.8548 (inf) [2021-04-16 09:07:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][1200/1251] eta 0:00:14 lr 0.000333 time 0.2923 (0.2905) loss 3.4237 (3.3854) grad_norm 1.5039 (inf) [2021-04-16 09:08:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][1210/1251] eta 0:00:11 lr 0.000333 time 0.2565 (0.2903) loss 3.6455 (3.3851) grad_norm 1.6555 (inf) [2021-04-16 09:08:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][1220/1251] eta 0:00:08 lr 0.000333 time 0.2918 (0.2903) loss 3.3010 (3.3857) grad_norm 1.5073 (inf) [2021-04-16 09:08:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][1230/1251] eta 0:00:06 lr 0.000333 time 0.2713 (0.2903) loss 3.5756 (3.3882) grad_norm 2.4615 (inf) [2021-04-16 09:08:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][1240/1251] eta 0:00:03 lr 0.000332 time 0.2479 (0.2901) loss 3.5416 (3.3894) grad_norm 1.9113 (inf) [2021-04-16 09:08:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [183/300][1250/1251] eta 0:00:00 lr 0.000332 time 0.2498 (0.2898) loss 4.0202 (3.3894) grad_norm 1.6308 (inf) [2021-04-16 09:08:16 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 183 training takes 0:06:06 [2021-04-16 09:08:16 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_183.pth saving...... [2021-04-16 09:08:25 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_183.pth saved !!! [2021-04-16 09:08:26 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.092 (1.092) Loss 0.9915 (0.9915) Acc@1 76.562 (76.562) Acc@5 94.238 (94.238) [2021-04-16 09:08:27 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.089 (0.213) Loss 0.9587 (0.9689) Acc@1 76.660 (77.157) Acc@5 93.945 (93.839) [2021-04-16 09:08:29 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.131 (0.221) Loss 0.9664 (0.9552) Acc@1 77.734 (77.567) Acc@5 93.652 (93.885) [2021-04-16 09:08:32 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.116 (0.234) Loss 0.9848 (0.9481) Acc@1 76.758 (77.706) Acc@5 94.238 (94.011) [2021-04-16 09:08:33 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.213) Loss 0.9378 (0.9489) Acc@1 77.637 (77.727) Acc@5 94.531 (93.979) [2021-04-16 09:08:39 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 77.730 Acc@5 94.008 [2021-04-16 09:08:39 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 77.7% [2021-04-16 09:08:39 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 77.73% [2021-04-16 09:08:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][0/1251] eta 3:05:03 lr 0.000332 time 8.8757 (8.8757) loss 3.5509 (3.5509) grad_norm 1.9838 (1.9838) [2021-04-16 09:08:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][10/1251] eta 0:21:52 lr 0.000332 time 0.3138 (1.0575) loss 3.4040 (3.0795) grad_norm 1.8020 (1.8461) [2021-04-16 09:08:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][20/1251] eta 0:14:04 lr 0.000332 time 0.3037 (0.6864) loss 3.2584 (3.1468) grad_norm 1.6109 (1.7863) [2021-04-16 09:08:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][30/1251] eta 0:11:16 lr 0.000332 time 0.2762 (0.5540) loss 2.1273 (3.1636) grad_norm 1.6446 (1.7554) [2021-04-16 09:08:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][40/1251] eta 0:09:48 lr 0.000332 time 0.2853 (0.4863) loss 3.2234 (3.1461) grad_norm 1.8905 (1.7655) [2021-04-16 09:09:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][50/1251] eta 0:08:54 lr 0.000332 time 0.2738 (0.4448) loss 3.4541 (3.2318) grad_norm 1.6974 (1.7693) [2021-04-16 09:09:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][60/1251] eta 0:08:18 lr 0.000332 time 0.2762 (0.4185) loss 4.2970 (3.2248) grad_norm 1.8513 (1.7827) [2021-04-16 09:09:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][70/1251] eta 0:07:51 lr 0.000332 time 0.2820 (0.3989) loss 4.1975 (3.2624) grad_norm 1.8045 (1.7872) [2021-04-16 09:09:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][80/1251] eta 0:07:29 lr 0.000332 time 0.2867 (0.3840) loss 3.5843 (3.2845) grad_norm 1.5246 (1.7958) [2021-04-16 09:09:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][90/1251] eta 0:07:12 lr 0.000332 time 0.2779 (0.3727) loss 4.1320 (3.3156) grad_norm 2.1285 (1.8050) [2021-04-16 09:09:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][100/1251] eta 0:06:59 lr 0.000332 time 0.3077 (0.3646) loss 3.5050 (3.3335) grad_norm 2.2337 (1.8155) [2021-04-16 09:09:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][110/1251] eta 0:06:47 lr 0.000332 time 0.2644 (0.3568) loss 3.7975 (3.3395) grad_norm 1.7309 (1.8247) [2021-04-16 09:09:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][120/1251] eta 0:06:35 lr 0.000332 time 0.2638 (0.3500) loss 2.9647 (3.3447) grad_norm 2.3096 (1.8386) [2021-04-16 09:09:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][130/1251] eta 0:06:26 lr 0.000332 time 0.2815 (0.3445) loss 2.8686 (3.3384) grad_norm 1.8028 (1.8330) [2021-04-16 09:09:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][140/1251] eta 0:06:19 lr 0.000332 time 0.2910 (0.3415) loss 3.7929 (3.3526) grad_norm 1.6418 (1.8327) [2021-04-16 09:09:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][150/1251] eta 0:06:11 lr 0.000332 time 0.2973 (0.3372) loss 2.9695 (3.3516) grad_norm 1.9421 (1.8263) [2021-04-16 09:09:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][160/1251] eta 0:06:04 lr 0.000332 time 0.2560 (0.3338) loss 3.0990 (3.3399) grad_norm 1.6029 (1.8195) [2021-04-16 09:09:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][170/1251] eta 0:05:57 lr 0.000332 time 0.2700 (0.3307) loss 3.5739 (3.3356) grad_norm 1.7370 (1.8193) [2021-04-16 09:09:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][180/1251] eta 0:05:52 lr 0.000332 time 0.2822 (0.3293) loss 3.3501 (3.3393) grad_norm 1.5403 (1.8191) [2021-04-16 09:09:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][190/1251] eta 0:05:47 lr 0.000332 time 0.2688 (0.3274) loss 2.4988 (3.3410) grad_norm 1.7161 (1.8162) [2021-04-16 09:09:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][200/1251] eta 0:05:41 lr 0.000332 time 0.2687 (0.3251) loss 4.2396 (3.3378) grad_norm 1.7735 (1.8147) [2021-04-16 09:09:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][210/1251] eta 0:05:36 lr 0.000332 time 0.3050 (0.3233) loss 3.8631 (3.3445) grad_norm 1.7674 (1.8136) [2021-04-16 09:09:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][220/1251] eta 0:05:31 lr 0.000332 time 0.2789 (0.3211) loss 3.5337 (3.3583) grad_norm 1.6811 (1.8093) [2021-04-16 09:09:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][230/1251] eta 0:05:26 lr 0.000332 time 0.2765 (0.3200) loss 3.1439 (3.3656) grad_norm 1.5025 (1.8044) [2021-04-16 09:09:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][240/1251] eta 0:05:21 lr 0.000332 time 0.2636 (0.3182) loss 2.9676 (3.3629) grad_norm 2.1829 (1.8067) [2021-04-16 09:09:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][250/1251] eta 0:05:16 lr 0.000331 time 0.2778 (0.3164) loss 3.0331 (3.3665) grad_norm 1.9602 (1.8076) [2021-04-16 09:10:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][260/1251] eta 0:05:12 lr 0.000331 time 0.2971 (0.3151) loss 3.5560 (3.3543) grad_norm 2.0327 (1.8087) [2021-04-16 09:10:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][270/1251] eta 0:05:07 lr 0.000331 time 0.2818 (0.3136) loss 3.3603 (3.3618) grad_norm 1.9341 (1.8085) [2021-04-16 09:10:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][280/1251] eta 0:05:03 lr 0.000331 time 0.2700 (0.3124) loss 3.2694 (3.3643) grad_norm 1.6321 (1.8120) [2021-04-16 09:10:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][290/1251] eta 0:04:59 lr 0.000331 time 0.2733 (0.3113) loss 3.5601 (3.3652) grad_norm 1.6298 (1.8112) [2021-04-16 09:10:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][300/1251] eta 0:04:55 lr 0.000331 time 0.2725 (0.3106) loss 3.6681 (3.3696) grad_norm 1.6993 (1.8063) [2021-04-16 09:10:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][310/1251] eta 0:04:51 lr 0.000331 time 0.2575 (0.3097) loss 3.3628 (3.3672) grad_norm 1.7179 (1.8070) [2021-04-16 09:10:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][320/1251] eta 0:04:47 lr 0.000331 time 0.2724 (0.3088) loss 3.1660 (3.3682) grad_norm 1.8870 (1.8059) [2021-04-16 09:10:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][330/1251] eta 0:04:43 lr 0.000331 time 0.2833 (0.3079) loss 3.3949 (3.3736) grad_norm 1.6534 (1.8021) [2021-04-16 09:10:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][340/1251] eta 0:04:39 lr 0.000331 time 0.3136 (0.3071) loss 3.5463 (3.3718) grad_norm 1.8685 (1.8022) [2021-04-16 09:10:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][350/1251] eta 0:04:35 lr 0.000331 time 0.2845 (0.3063) loss 2.9873 (3.3688) grad_norm 1.7884 (1.8030) [2021-04-16 09:10:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][360/1251] eta 0:04:32 lr 0.000331 time 0.2461 (0.3063) loss 3.7780 (3.3755) grad_norm 1.8276 (1.8003) [2021-04-16 09:10:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][370/1251] eta 0:04:29 lr 0.000331 time 0.2711 (0.3055) loss 2.8798 (3.3801) grad_norm 1.7055 (1.7988) [2021-04-16 09:10:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][380/1251] eta 0:04:25 lr 0.000331 time 0.2734 (0.3050) loss 3.6187 (3.3812) grad_norm 2.2320 (1.8007) [2021-04-16 09:10:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][390/1251] eta 0:04:22 lr 0.000331 time 0.2756 (0.3045) loss 4.3023 (3.3886) grad_norm 1.5629 (1.7976) [2021-04-16 09:10:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][400/1251] eta 0:04:18 lr 0.000331 time 0.2773 (0.3039) loss 3.3400 (3.3753) grad_norm 1.7988 (1.7972) [2021-04-16 09:10:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][410/1251] eta 0:04:15 lr 0.000331 time 0.3021 (0.3034) loss 3.8109 (3.3787) grad_norm 1.6569 (1.7984) [2021-04-16 09:10:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][420/1251] eta 0:04:11 lr 0.000331 time 0.3078 (0.3031) loss 3.9043 (3.3762) grad_norm 1.8777 (1.8003) [2021-04-16 09:10:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][430/1251] eta 0:04:08 lr 0.000331 time 0.2624 (0.3027) loss 3.4107 (3.3768) grad_norm 1.9142 (1.8008) [2021-04-16 09:10:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][440/1251] eta 0:04:05 lr 0.000331 time 0.2416 (0.3022) loss 3.5526 (3.3847) grad_norm 1.8400 (1.7990) [2021-04-16 09:10:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][450/1251] eta 0:04:01 lr 0.000331 time 0.2682 (0.3016) loss 2.6701 (3.3853) grad_norm 1.6823 (1.7989) [2021-04-16 09:10:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][460/1251] eta 0:03:58 lr 0.000331 time 0.2759 (0.3011) loss 3.8628 (3.3832) grad_norm 1.9270 (1.8009) [2021-04-16 09:11:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][470/1251] eta 0:03:54 lr 0.000331 time 0.2728 (0.3007) loss 3.8351 (3.3885) grad_norm 1.8603 (1.8022) [2021-04-16 09:11:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][480/1251] eta 0:03:51 lr 0.000331 time 0.2783 (0.3003) loss 3.6347 (3.3922) grad_norm 1.6712 (1.8014) [2021-04-16 09:11:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][490/1251] eta 0:03:48 lr 0.000331 time 0.2695 (0.2998) loss 3.5112 (3.3909) grad_norm 1.6234 (1.7994) [2021-04-16 09:11:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][500/1251] eta 0:03:44 lr 0.000331 time 0.2689 (0.2994) loss 3.5275 (3.3911) grad_norm 1.7077 (1.7987) [2021-04-16 09:11:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][510/1251] eta 0:03:41 lr 0.000330 time 0.2841 (0.2991) loss 2.9809 (3.3941) grad_norm 1.6722 (1.7984) [2021-04-16 09:11:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][520/1251] eta 0:03:38 lr 0.000330 time 0.3011 (0.2987) loss 4.3466 (3.3954) grad_norm 1.6969 (1.7971) [2021-04-16 09:11:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][530/1251] eta 0:03:34 lr 0.000330 time 0.2663 (0.2982) loss 3.7266 (3.3960) grad_norm 1.9005 (1.7980) [2021-04-16 09:11:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][540/1251] eta 0:03:31 lr 0.000330 time 0.2679 (0.2978) loss 3.9531 (3.4008) grad_norm 1.9230 (1.7972) [2021-04-16 09:11:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][550/1251] eta 0:03:28 lr 0.000330 time 0.2792 (0.2975) loss 3.5859 (3.4003) grad_norm 1.8282 (1.7976) [2021-04-16 09:11:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][560/1251] eta 0:03:25 lr 0.000330 time 0.2638 (0.2972) loss 3.4509 (3.3981) grad_norm 1.7181 (1.7976) [2021-04-16 09:11:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][570/1251] eta 0:03:22 lr 0.000330 time 0.2972 (0.2970) loss 3.1929 (3.4042) grad_norm 1.7944 (1.7991) [2021-04-16 09:11:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][580/1251] eta 0:03:19 lr 0.000330 time 0.2707 (0.2974) loss 3.7183 (3.4083) grad_norm 1.5419 (1.7987) [2021-04-16 09:11:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][590/1251] eta 0:03:16 lr 0.000330 time 0.2769 (0.2971) loss 2.7294 (3.4102) grad_norm 1.6583 (1.7974) [2021-04-16 09:11:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][600/1251] eta 0:03:13 lr 0.000330 time 0.2691 (0.2967) loss 2.8837 (3.4098) grad_norm 1.8085 (1.7968) [2021-04-16 09:11:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][610/1251] eta 0:03:09 lr 0.000330 time 0.2515 (0.2964) loss 3.7205 (3.4133) grad_norm 2.1861 (1.7974) [2021-04-16 09:11:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][620/1251] eta 0:03:06 lr 0.000330 time 0.3004 (0.2961) loss 2.7149 (3.4116) grad_norm 1.6483 (1.7978) [2021-04-16 09:11:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][630/1251] eta 0:03:03 lr 0.000330 time 0.2916 (0.2959) loss 2.9063 (3.4052) grad_norm 1.8990 (1.7975) [2021-04-16 09:11:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][640/1251] eta 0:03:00 lr 0.000330 time 0.2902 (0.2955) loss 3.8190 (3.4066) grad_norm 1.7148 (1.7964) [2021-04-16 09:11:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][650/1251] eta 0:02:57 lr 0.000330 time 0.2833 (0.2952) loss 3.5760 (3.4091) grad_norm 1.8759 (1.7977) [2021-04-16 09:11:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][660/1251] eta 0:02:54 lr 0.000330 time 0.2707 (0.2950) loss 3.6395 (3.4060) grad_norm 1.8268 (1.8012) [2021-04-16 09:11:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][670/1251] eta 0:02:51 lr 0.000330 time 0.3157 (0.2949) loss 3.6743 (3.4064) grad_norm 1.8020 (1.8024) [2021-04-16 09:11:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][680/1251] eta 0:02:48 lr 0.000330 time 0.2628 (0.2948) loss 3.1636 (3.4057) grad_norm 1.9196 (1.8039) [2021-04-16 09:12:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][690/1251] eta 0:02:45 lr 0.000330 time 0.3034 (0.2946) loss 2.7953 (3.4065) grad_norm 1.8881 (1.8059) [2021-04-16 09:12:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][700/1251] eta 0:02:42 lr 0.000330 time 0.2745 (0.2943) loss 3.6798 (3.4084) grad_norm 1.7299 (1.8048) [2021-04-16 09:12:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][710/1251] eta 0:02:39 lr 0.000330 time 0.2758 (0.2941) loss 3.8129 (3.4083) grad_norm 1.9530 (1.8043) [2021-04-16 09:12:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][720/1251] eta 0:02:36 lr 0.000330 time 0.2872 (0.2942) loss 3.1340 (3.4081) grad_norm 1.6858 (1.8043) [2021-04-16 09:12:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][730/1251] eta 0:02:33 lr 0.000330 time 0.2960 (0.2940) loss 3.3064 (3.4078) grad_norm 1.6817 (1.8029) [2021-04-16 09:12:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][740/1251] eta 0:02:30 lr 0.000330 time 0.2740 (0.2938) loss 3.5155 (3.4077) grad_norm 1.8674 (1.8033) [2021-04-16 09:12:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][750/1251] eta 0:02:27 lr 0.000330 time 0.2985 (0.2937) loss 3.6008 (3.4058) grad_norm 1.8432 (1.8028) [2021-04-16 09:12:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][760/1251] eta 0:02:24 lr 0.000330 time 0.2713 (0.2935) loss 3.4928 (3.4053) grad_norm 1.7865 (1.8027) [2021-04-16 09:12:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][770/1251] eta 0:02:21 lr 0.000329 time 0.2715 (0.2934) loss 4.0634 (3.4045) grad_norm 1.7020 (1.8013) [2021-04-16 09:12:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][780/1251] eta 0:02:18 lr 0.000329 time 0.2823 (0.2933) loss 2.4283 (3.4057) grad_norm 2.0174 (1.8027) [2021-04-16 09:12:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][790/1251] eta 0:02:15 lr 0.000329 time 0.3086 (0.2931) loss 3.4292 (3.4015) grad_norm 1.8280 (1.8037) [2021-04-16 09:12:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][800/1251] eta 0:02:12 lr 0.000329 time 0.2633 (0.2929) loss 4.4592 (3.4029) grad_norm 1.7163 (1.8040) [2021-04-16 09:12:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][810/1251] eta 0:02:09 lr 0.000329 time 0.2768 (0.2929) loss 3.6586 (3.4011) grad_norm 1.8050 (1.8039) [2021-04-16 09:12:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][820/1251] eta 0:02:06 lr 0.000329 time 0.2915 (0.2927) loss 4.0449 (3.4007) grad_norm 2.2045 (1.8038) [2021-04-16 09:12:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][830/1251] eta 0:02:03 lr 0.000329 time 0.2940 (0.2926) loss 3.7233 (3.3997) grad_norm 1.7295 (1.8046) [2021-04-16 09:12:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][840/1251] eta 0:02:00 lr 0.000329 time 0.2659 (0.2924) loss 3.5490 (3.3989) grad_norm 1.8609 (1.8041) [2021-04-16 09:12:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][850/1251] eta 0:01:57 lr 0.000329 time 0.2624 (0.2923) loss 3.5907 (3.3984) grad_norm 1.8663 (1.8040) [2021-04-16 09:12:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][860/1251] eta 0:01:54 lr 0.000329 time 0.2793 (0.2921) loss 3.6794 (3.3996) grad_norm 1.7363 (1.8036) [2021-04-16 09:12:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][870/1251] eta 0:01:51 lr 0.000329 time 0.2637 (0.2920) loss 2.7371 (3.3988) grad_norm 1.7674 (1.8026) [2021-04-16 09:12:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][880/1251] eta 0:01:48 lr 0.000329 time 0.2849 (0.2919) loss 2.2110 (3.3981) grad_norm 2.1389 (1.8030) [2021-04-16 09:12:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][890/1251] eta 0:01:45 lr 0.000329 time 0.2812 (0.2918) loss 2.7082 (3.3982) grad_norm 1.9102 (1.8032) [2021-04-16 09:13:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][900/1251] eta 0:01:42 lr 0.000329 time 0.2542 (0.2916) loss 3.9859 (3.3974) grad_norm 1.8870 (1.8047) [2021-04-16 09:13:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][910/1251] eta 0:01:39 lr 0.000329 time 0.2636 (0.2915) loss 2.2001 (3.3959) grad_norm 1.9923 (1.8044) [2021-04-16 09:13:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][920/1251] eta 0:01:36 lr 0.000329 time 0.2686 (0.2914) loss 3.9939 (3.3971) grad_norm 1.7299 (1.8043) [2021-04-16 09:13:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][930/1251] eta 0:01:33 lr 0.000329 time 0.2602 (0.2914) loss 3.6709 (3.3993) grad_norm 1.6691 (1.8037) [2021-04-16 09:13:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][940/1251] eta 0:01:30 lr 0.000329 time 0.2789 (0.2915) loss 3.9947 (3.3981) grad_norm 1.7247 (1.8035) [2021-04-16 09:13:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][950/1251] eta 0:01:27 lr 0.000329 time 0.2900 (0.2914) loss 3.6075 (3.3983) grad_norm 1.9546 (1.8029) [2021-04-16 09:13:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][960/1251] eta 0:01:24 lr 0.000329 time 0.2513 (0.2912) loss 4.0068 (3.4002) grad_norm 1.7571 (1.8035) [2021-04-16 09:13:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][970/1251] eta 0:01:21 lr 0.000329 time 0.2747 (0.2912) loss 3.7835 (3.4003) grad_norm 1.7772 (1.8034) [2021-04-16 09:13:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][980/1251] eta 0:01:18 lr 0.000329 time 0.2829 (0.2911) loss 3.1104 (3.4020) grad_norm 1.8257 (1.8040) [2021-04-16 09:13:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][990/1251] eta 0:01:15 lr 0.000329 time 0.2767 (0.2910) loss 4.5157 (3.4040) grad_norm 2.1851 (1.8046) [2021-04-16 09:13:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][1000/1251] eta 0:01:12 lr 0.000329 time 0.2512 (0.2908) loss 3.8290 (3.4049) grad_norm 1.6069 (1.8033) [2021-04-16 09:13:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][1010/1251] eta 0:01:10 lr 0.000329 time 0.2695 (0.2908) loss 3.1537 (3.4041) grad_norm 1.4268 (1.8028) [2021-04-16 09:13:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][1020/1251] eta 0:01:07 lr 0.000329 time 0.2902 (0.2907) loss 4.0445 (3.4072) grad_norm 1.6925 (1.8033) [2021-04-16 09:13:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][1030/1251] eta 0:01:04 lr 0.000328 time 0.2905 (0.2906) loss 2.7700 (3.4068) grad_norm 1.9974 (1.8051) [2021-04-16 09:13:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][1040/1251] eta 0:01:01 lr 0.000328 time 0.2826 (0.2905) loss 3.4475 (3.4074) grad_norm 1.8628 (1.8055) [2021-04-16 09:13:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][1050/1251] eta 0:00:58 lr 0.000328 time 0.2738 (0.2904) loss 3.6953 (3.4079) grad_norm 1.8534 (1.8060) [2021-04-16 09:13:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][1060/1251] eta 0:00:55 lr 0.000328 time 0.2793 (0.2902) loss 2.9223 (3.4056) grad_norm 1.9400 (1.8074) [2021-04-16 09:13:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][1070/1251] eta 0:00:52 lr 0.000328 time 0.2636 (0.2901) loss 3.4085 (3.4064) grad_norm 1.6433 (1.8087) [2021-04-16 09:13:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][1080/1251] eta 0:00:49 lr 0.000328 time 0.2774 (0.2900) loss 3.7973 (3.4080) grad_norm 1.9342 (1.8091) [2021-04-16 09:13:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][1090/1251] eta 0:00:46 lr 0.000328 time 0.2723 (0.2899) loss 2.9449 (3.4066) grad_norm 1.7982 (1.8119) [2021-04-16 09:13:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][1100/1251] eta 0:00:43 lr 0.000328 time 0.3043 (0.2898) loss 3.6425 (3.4049) grad_norm 1.8546 (1.8123) [2021-04-16 09:14:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][1110/1251] eta 0:00:40 lr 0.000328 time 0.2799 (0.2897) loss 3.5242 (3.4045) grad_norm 2.1035 (1.8129) [2021-04-16 09:14:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][1120/1251] eta 0:00:37 lr 0.000328 time 0.2571 (0.2896) loss 3.8450 (3.4042) grad_norm 1.7448 (1.8131) [2021-04-16 09:14:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][1130/1251] eta 0:00:35 lr 0.000328 time 0.2817 (0.2895) loss 3.2646 (3.4039) grad_norm 1.8259 (1.8135) [2021-04-16 09:14:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][1140/1251] eta 0:00:32 lr 0.000328 time 0.2870 (0.2895) loss 3.3093 (3.4054) grad_norm 1.7182 (1.8128) [2021-04-16 09:14:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][1150/1251] eta 0:00:29 lr 0.000328 time 0.2642 (0.2896) loss 3.9935 (3.4073) grad_norm 2.0141 (1.8133) [2021-04-16 09:14:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][1160/1251] eta 0:00:26 lr 0.000328 time 0.2952 (0.2896) loss 3.7660 (3.4085) grad_norm 1.8380 (1.8130) [2021-04-16 09:14:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][1170/1251] eta 0:00:23 lr 0.000328 time 0.2661 (0.2895) loss 2.6460 (3.4060) grad_norm 1.7954 (1.8131) [2021-04-16 09:14:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][1180/1251] eta 0:00:20 lr 0.000328 time 0.2884 (0.2894) loss 2.6312 (3.4065) grad_norm 1.9377 (1.8134) [2021-04-16 09:14:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][1190/1251] eta 0:00:17 lr 0.000328 time 0.2795 (0.2892) loss 3.1398 (3.4058) grad_norm 1.4792 (1.8136) [2021-04-16 09:14:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][1200/1251] eta 0:00:14 lr 0.000328 time 0.2747 (0.2892) loss 2.2716 (3.4026) grad_norm 2.1658 (1.8143) [2021-04-16 09:14:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][1210/1251] eta 0:00:11 lr 0.000328 time 0.2795 (0.2890) loss 2.2584 (3.3992) grad_norm 1.7373 (1.8146) [2021-04-16 09:14:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][1220/1251] eta 0:00:08 lr 0.000328 time 0.2928 (0.2889) loss 4.2066 (3.4008) grad_norm 1.6172 (1.8143) [2021-04-16 09:14:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][1230/1251] eta 0:00:06 lr 0.000328 time 0.2811 (0.2888) loss 3.8511 (3.4034) grad_norm 2.0963 (1.8151) [2021-04-16 09:14:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][1240/1251] eta 0:00:03 lr 0.000328 time 0.2625 (0.2887) loss 2.9780 (3.4020) grad_norm 1.7910 (1.8160) [2021-04-16 09:14:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [184/300][1250/1251] eta 0:00:00 lr 0.000328 time 0.2493 (0.2884) loss 2.6247 (3.4033) grad_norm 1.8425 (1.8171) [2021-04-16 09:14:43 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 184 training takes 0:06:04 [2021-04-16 09:14:43 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_184.pth saving...... [2021-04-16 09:14:53 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_184.pth saved !!! [2021-04-16 09:14:54 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.188 (1.188) Loss 0.8481 (0.8481) Acc@1 80.469 (80.469) Acc@5 95.215 (95.215) [2021-04-16 09:14:55 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.080 (0.205) Loss 0.9217 (0.9191) Acc@1 78.027 (78.161) Acc@5 94.629 (94.256) [2021-04-16 09:14:58 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.140 (0.215) Loss 0.9718 (0.9365) Acc@1 76.855 (77.688) Acc@5 94.141 (94.322) [2021-04-16 09:15:00 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.149 (0.236) Loss 0.8333 (0.9381) Acc@1 79.102 (77.593) Acc@5 95.996 (94.194) [2021-04-16 09:15:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.217) Loss 1.0392 (0.9487) Acc@1 76.465 (77.468) Acc@5 93.652 (94.022) [2021-04-16 09:15:09 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 77.560 Acc@5 94.028 [2021-04-16 09:15:09 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 77.6% [2021-04-16 09:15:09 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 77.73% [2021-04-16 09:15:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][0/1251] eta 3:08:17 lr 0.000328 time 9.0312 (9.0312) loss 2.8013 (2.8013) grad_norm 1.9135 (1.9135) [2021-04-16 09:15:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][10/1251] eta 0:22:04 lr 0.000328 time 0.2675 (1.0673) loss 3.6363 (3.1475) grad_norm 1.7243 (1.8605) [2021-04-16 09:15:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][20/1251] eta 0:14:18 lr 0.000328 time 0.2828 (0.6973) loss 3.3366 (3.2104) grad_norm 2.0614 (1.8631) [2021-04-16 09:15:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][30/1251] eta 0:11:28 lr 0.000327 time 0.2594 (0.5641) loss 3.9463 (3.2862) grad_norm 1.9474 (1.8777) [2021-04-16 09:15:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][40/1251] eta 0:09:59 lr 0.000327 time 0.2823 (0.4949) loss 2.6882 (3.3668) grad_norm 2.1773 (1.8859) [2021-04-16 09:15:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][50/1251] eta 0:09:04 lr 0.000327 time 0.2978 (0.4532) loss 3.6470 (3.3766) grad_norm 1.6420 (1.8703) [2021-04-16 09:15:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][60/1251] eta 0:08:28 lr 0.000327 time 0.3230 (0.4266) loss 4.0244 (3.3653) grad_norm 1.7925 (1.8589) [2021-04-16 09:15:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][70/1251] eta 0:07:58 lr 0.000327 time 0.2940 (0.4054) loss 3.9960 (3.4063) grad_norm 1.9205 (1.8483) [2021-04-16 09:15:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][80/1251] eta 0:07:36 lr 0.000327 time 0.2797 (0.3896) loss 3.6486 (3.4074) grad_norm 2.1095 (1.8528) [2021-04-16 09:15:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][90/1251] eta 0:07:20 lr 0.000327 time 0.2663 (0.3796) loss 2.4217 (3.4030) grad_norm 1.7217 (1.8472) [2021-04-16 09:15:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][100/1251] eta 0:07:06 lr 0.000327 time 0.2822 (0.3706) loss 3.8327 (3.4047) grad_norm 1.9525 (1.8480) [2021-04-16 09:15:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][110/1251] eta 0:06:53 lr 0.000327 time 0.2894 (0.3621) loss 3.6941 (3.4209) grad_norm 1.8630 (1.8653) [2021-04-16 09:15:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][120/1251] eta 0:06:41 lr 0.000327 time 0.2749 (0.3554) loss 3.7691 (3.4265) grad_norm 1.6994 (1.8615) [2021-04-16 09:15:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][130/1251] eta 0:06:32 lr 0.000327 time 0.2872 (0.3497) loss 3.1877 (3.4214) grad_norm 1.7877 (1.8548) [2021-04-16 09:15:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][140/1251] eta 0:06:23 lr 0.000327 time 0.2662 (0.3454) loss 3.4396 (3.4068) grad_norm 1.6149 (1.8480) [2021-04-16 09:16:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][150/1251] eta 0:06:17 lr 0.000327 time 0.2730 (0.3427) loss 3.5489 (3.4129) grad_norm 1.9496 (1.8423) [2021-04-16 09:16:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][160/1251] eta 0:06:09 lr 0.000327 time 0.2851 (0.3385) loss 3.0002 (3.4129) grad_norm 1.8506 (inf) [2021-04-16 09:16:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][170/1251] eta 0:06:02 lr 0.000327 time 0.2892 (0.3353) loss 3.2466 (3.3995) grad_norm 2.1136 (inf) [2021-04-16 09:16:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][180/1251] eta 0:05:56 lr 0.000327 time 0.2951 (0.3328) loss 4.1298 (3.4141) grad_norm 2.0606 (inf) [2021-04-16 09:16:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][190/1251] eta 0:05:50 lr 0.000327 time 0.2732 (0.3300) loss 3.1245 (3.4194) grad_norm 1.8039 (inf) [2021-04-16 09:16:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][200/1251] eta 0:05:44 lr 0.000327 time 0.2814 (0.3278) loss 2.9311 (3.4042) grad_norm 1.7253 (inf) [2021-04-16 09:16:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][210/1251] eta 0:05:38 lr 0.000327 time 0.2863 (0.3256) loss 2.7568 (3.4000) grad_norm 1.6952 (inf) [2021-04-16 09:16:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][220/1251] eta 0:05:33 lr 0.000327 time 0.2501 (0.3238) loss 3.9164 (3.3925) grad_norm 2.1276 (inf) [2021-04-16 09:16:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][230/1251] eta 0:05:28 lr 0.000327 time 0.2826 (0.3218) loss 3.4284 (3.3991) grad_norm 1.9295 (inf) [2021-04-16 09:16:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][240/1251] eta 0:05:23 lr 0.000327 time 0.3004 (0.3200) loss 3.5649 (3.3924) grad_norm 1.8175 (inf) [2021-04-16 09:16:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][250/1251] eta 0:05:18 lr 0.000327 time 0.2821 (0.3184) loss 4.0806 (3.3950) grad_norm 1.9667 (inf) [2021-04-16 09:16:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][260/1251] eta 0:05:13 lr 0.000327 time 0.2718 (0.3166) loss 2.8247 (3.3943) grad_norm 1.9161 (inf) [2021-04-16 09:16:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][270/1251] eta 0:05:09 lr 0.000327 time 0.2673 (0.3153) loss 2.6533 (3.3891) grad_norm 1.9063 (inf) [2021-04-16 09:16:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][280/1251] eta 0:05:04 lr 0.000327 time 0.2622 (0.3139) loss 3.8312 (3.3974) grad_norm 1.9954 (inf) [2021-04-16 09:16:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][290/1251] eta 0:05:00 lr 0.000326 time 0.2592 (0.3129) loss 2.9409 (3.3980) grad_norm 1.8711 (inf) [2021-04-16 09:16:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][300/1251] eta 0:04:56 lr 0.000326 time 0.2859 (0.3115) loss 3.8930 (3.4074) grad_norm 1.8181 (inf) [2021-04-16 09:16:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][310/1251] eta 0:04:52 lr 0.000326 time 0.2855 (0.3105) loss 3.1162 (3.4006) grad_norm 1.9994 (inf) [2021-04-16 09:16:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][320/1251] eta 0:04:47 lr 0.000326 time 0.2453 (0.3093) loss 3.3873 (3.4033) grad_norm 1.8232 (inf) [2021-04-16 09:16:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][330/1251] eta 0:04:44 lr 0.000326 time 0.2501 (0.3084) loss 2.1811 (3.4016) grad_norm 1.9319 (inf) [2021-04-16 09:16:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][340/1251] eta 0:04:40 lr 0.000326 time 0.2978 (0.3081) loss 3.6756 (3.4018) grad_norm 2.0448 (inf) [2021-04-16 09:16:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][350/1251] eta 0:04:37 lr 0.000326 time 0.2658 (0.3076) loss 3.1972 (3.4092) grad_norm 1.7889 (inf) [2021-04-16 09:17:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][360/1251] eta 0:04:34 lr 0.000326 time 0.2702 (0.3077) loss 3.8800 (3.4040) grad_norm 1.7435 (inf) [2021-04-16 09:17:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][370/1251] eta 0:04:30 lr 0.000326 time 0.3105 (0.3068) loss 3.7551 (3.4016) grad_norm 2.4208 (inf) [2021-04-16 09:17:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][380/1251] eta 0:04:26 lr 0.000326 time 0.2889 (0.3060) loss 4.0270 (3.4018) grad_norm 1.6070 (inf) [2021-04-16 09:17:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][390/1251] eta 0:04:22 lr 0.000326 time 0.2877 (0.3054) loss 3.0243 (3.3979) grad_norm 1.8361 (inf) [2021-04-16 09:17:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][400/1251] eta 0:04:19 lr 0.000326 time 0.2864 (0.3049) loss 4.0770 (3.3975) grad_norm 1.9622 (inf) [2021-04-16 09:17:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][410/1251] eta 0:04:15 lr 0.000326 time 0.2905 (0.3043) loss 3.6292 (3.4024) grad_norm 2.0687 (inf) [2021-04-16 09:17:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][420/1251] eta 0:04:12 lr 0.000326 time 0.2766 (0.3039) loss 3.0122 (3.4058) grad_norm 1.9606 (inf) [2021-04-16 09:17:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][430/1251] eta 0:04:09 lr 0.000326 time 0.2643 (0.3035) loss 3.2387 (3.4067) grad_norm 1.9207 (inf) [2021-04-16 09:17:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][440/1251] eta 0:04:05 lr 0.000326 time 0.2597 (0.3030) loss 3.4594 (3.4021) grad_norm 1.7717 (inf) [2021-04-16 09:17:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][450/1251] eta 0:04:02 lr 0.000326 time 0.3045 (0.3025) loss 2.8789 (3.3912) grad_norm 1.9707 (inf) [2021-04-16 09:17:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][460/1251] eta 0:03:58 lr 0.000326 time 0.2932 (0.3021) loss 3.8706 (3.3943) grad_norm 2.0638 (inf) [2021-04-16 09:17:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][470/1251] eta 0:03:55 lr 0.000326 time 0.2647 (0.3017) loss 3.8995 (3.3962) grad_norm 1.5052 (inf) [2021-04-16 09:17:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][480/1251] eta 0:03:52 lr 0.000326 time 0.3016 (0.3013) loss 3.0201 (3.3981) grad_norm 1.7101 (inf) [2021-04-16 09:17:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][490/1251] eta 0:03:48 lr 0.000326 time 0.2711 (0.3009) loss 3.0135 (3.3898) grad_norm 1.8230 (inf) [2021-04-16 09:17:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][500/1251] eta 0:03:45 lr 0.000326 time 0.2684 (0.3005) loss 3.9379 (3.3945) grad_norm 1.6998 (inf) [2021-04-16 09:17:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][510/1251] eta 0:03:42 lr 0.000326 time 0.2733 (0.3001) loss 2.9239 (3.3898) grad_norm 1.8945 (inf) [2021-04-16 09:17:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][520/1251] eta 0:03:39 lr 0.000326 time 0.2765 (0.2997) loss 3.2922 (3.3884) grad_norm 1.7195 (inf) [2021-04-16 09:17:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][530/1251] eta 0:03:35 lr 0.000326 time 0.3135 (0.2994) loss 2.8377 (3.3893) grad_norm 1.7968 (inf) [2021-04-16 09:17:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][540/1251] eta 0:03:32 lr 0.000326 time 0.2972 (0.2991) loss 3.1777 (3.3868) grad_norm 1.7321 (inf) [2021-04-16 09:17:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][550/1251] eta 0:03:29 lr 0.000325 time 0.2623 (0.2986) loss 2.6606 (3.3885) grad_norm 1.5779 (inf) [2021-04-16 09:17:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][560/1251] eta 0:03:26 lr 0.000325 time 0.3038 (0.2984) loss 3.2408 (3.3862) grad_norm 2.0264 (inf) [2021-04-16 09:17:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][570/1251] eta 0:03:23 lr 0.000325 time 0.2648 (0.2982) loss 4.2118 (3.3849) grad_norm 1.7685 (inf) [2021-04-16 09:18:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][580/1251] eta 0:03:20 lr 0.000325 time 0.2639 (0.2981) loss 3.5145 (3.3879) grad_norm 1.7980 (inf) [2021-04-16 09:18:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][590/1251] eta 0:03:17 lr 0.000325 time 0.2648 (0.2981) loss 3.4098 (3.3859) grad_norm 2.0557 (inf) [2021-04-16 09:18:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][600/1251] eta 0:03:13 lr 0.000325 time 0.2870 (0.2977) loss 3.7415 (3.3840) grad_norm 1.9295 (inf) [2021-04-16 09:18:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][610/1251] eta 0:03:10 lr 0.000325 time 0.2588 (0.2973) loss 3.1829 (3.3812) grad_norm 1.7414 (inf) [2021-04-16 09:18:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][620/1251] eta 0:03:07 lr 0.000325 time 0.2666 (0.2970) loss 4.2697 (3.3806) grad_norm 1.9333 (inf) [2021-04-16 09:18:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][630/1251] eta 0:03:04 lr 0.000325 time 0.3019 (0.2966) loss 3.4428 (3.3785) grad_norm 1.8463 (inf) [2021-04-16 09:18:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][640/1251] eta 0:03:01 lr 0.000325 time 0.2597 (0.2965) loss 3.3477 (3.3789) grad_norm 2.3621 (inf) [2021-04-16 09:18:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][650/1251] eta 0:02:58 lr 0.000325 time 0.2823 (0.2962) loss 4.0262 (3.3779) grad_norm 1.8984 (inf) [2021-04-16 09:18:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][660/1251] eta 0:02:54 lr 0.000325 time 0.2676 (0.2960) loss 3.4304 (3.3800) grad_norm 1.4531 (inf) [2021-04-16 09:18:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][670/1251] eta 0:02:51 lr 0.000325 time 0.2668 (0.2958) loss 3.0347 (3.3778) grad_norm 2.1538 (inf) [2021-04-16 09:18:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][680/1251] eta 0:02:48 lr 0.000325 time 0.2808 (0.2955) loss 3.1248 (3.3810) grad_norm 2.2855 (inf) [2021-04-16 09:18:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][690/1251] eta 0:02:45 lr 0.000325 time 0.2444 (0.2952) loss 3.6098 (3.3852) grad_norm 1.6558 (inf) [2021-04-16 09:18:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][700/1251] eta 0:02:42 lr 0.000325 time 0.3158 (0.2950) loss 3.7473 (3.3847) grad_norm 2.1398 (inf) [2021-04-16 09:18:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][710/1251] eta 0:02:39 lr 0.000325 time 0.3014 (0.2948) loss 3.8067 (3.3848) grad_norm 1.9491 (inf) [2021-04-16 09:18:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][720/1251] eta 0:02:36 lr 0.000325 time 0.2763 (0.2948) loss 3.1889 (3.3842) grad_norm 1.7750 (inf) [2021-04-16 09:18:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][730/1251] eta 0:02:33 lr 0.000325 time 0.2738 (0.2948) loss 2.3573 (3.3856) grad_norm 1.8377 (inf) [2021-04-16 09:18:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][740/1251] eta 0:02:30 lr 0.000325 time 0.2865 (0.2946) loss 3.7447 (3.3858) grad_norm 1.8612 (inf) [2021-04-16 09:18:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][750/1251] eta 0:02:27 lr 0.000325 time 0.2584 (0.2943) loss 2.3723 (3.3869) grad_norm 1.6122 (inf) [2021-04-16 09:18:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][760/1251] eta 0:02:24 lr 0.000325 time 0.2928 (0.2943) loss 4.6003 (3.3906) grad_norm 1.7199 (inf) [2021-04-16 09:18:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][770/1251] eta 0:02:21 lr 0.000325 time 0.2408 (0.2940) loss 2.5214 (3.3872) grad_norm 2.0398 (inf) [2021-04-16 09:18:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][780/1251] eta 0:02:18 lr 0.000325 time 0.2691 (0.2939) loss 2.8679 (3.3852) grad_norm 2.0518 (inf) [2021-04-16 09:19:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][790/1251] eta 0:02:15 lr 0.000325 time 0.2821 (0.2939) loss 3.8554 (3.3826) grad_norm 1.9761 (inf) [2021-04-16 09:19:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][800/1251] eta 0:02:12 lr 0.000325 time 0.2549 (0.2937) loss 3.9043 (3.3854) grad_norm 1.6500 (inf) [2021-04-16 09:19:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][810/1251] eta 0:02:09 lr 0.000324 time 0.2642 (0.2935) loss 3.9749 (3.3875) grad_norm 1.6602 (inf) [2021-04-16 09:19:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][820/1251] eta 0:02:06 lr 0.000324 time 0.3091 (0.2934) loss 3.9900 (3.3918) grad_norm 1.9278 (inf) [2021-04-16 09:19:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][830/1251] eta 0:02:03 lr 0.000324 time 0.2710 (0.2933) loss 3.9271 (3.3938) grad_norm 1.6796 (inf) [2021-04-16 09:19:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][840/1251] eta 0:02:00 lr 0.000324 time 0.2886 (0.2931) loss 3.3396 (3.3943) grad_norm 1.9246 (inf) [2021-04-16 09:19:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][850/1251] eta 0:01:57 lr 0.000324 time 0.2643 (0.2929) loss 3.6382 (3.3945) grad_norm 1.6283 (inf) [2021-04-16 09:19:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][860/1251] eta 0:01:54 lr 0.000324 time 0.2903 (0.2928) loss 3.1288 (3.3934) grad_norm 1.8177 (inf) [2021-04-16 09:19:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][870/1251] eta 0:01:51 lr 0.000324 time 0.2691 (0.2926) loss 3.9541 (3.3937) grad_norm 1.9434 (inf) [2021-04-16 09:19:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][880/1251] eta 0:01:48 lr 0.000324 time 0.2926 (0.2925) loss 3.9484 (3.3918) grad_norm 1.8168 (inf) [2021-04-16 09:19:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][890/1251] eta 0:01:45 lr 0.000324 time 0.2686 (0.2923) loss 3.9555 (3.3908) grad_norm 1.8112 (inf) [2021-04-16 09:19:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][900/1251] eta 0:01:42 lr 0.000324 time 0.2696 (0.2923) loss 2.5988 (3.3907) grad_norm 1.6296 (inf) [2021-04-16 09:19:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][910/1251] eta 0:01:39 lr 0.000324 time 0.2648 (0.2922) loss 3.1399 (3.3932) grad_norm 2.0216 (inf) [2021-04-16 09:19:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][920/1251] eta 0:01:36 lr 0.000324 time 0.2893 (0.2921) loss 3.9150 (3.3917) grad_norm 1.5860 (inf) [2021-04-16 09:19:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][930/1251] eta 0:01:33 lr 0.000324 time 0.2868 (0.2921) loss 3.4076 (3.3897) grad_norm 2.1198 (inf) [2021-04-16 09:19:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][940/1251] eta 0:01:30 lr 0.000324 time 0.2688 (0.2922) loss 4.1060 (3.3895) grad_norm 1.5950 (inf) [2021-04-16 09:19:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][950/1251] eta 0:01:27 lr 0.000324 time 0.3034 (0.2921) loss 3.8969 (3.3907) grad_norm 1.8674 (inf) [2021-04-16 09:19:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][960/1251] eta 0:01:24 lr 0.000324 time 0.2835 (0.2919) loss 3.9220 (3.3916) grad_norm 1.6958 (inf) [2021-04-16 09:19:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][970/1251] eta 0:01:21 lr 0.000324 time 0.2871 (0.2918) loss 2.5495 (3.3924) grad_norm 1.7651 (inf) [2021-04-16 09:19:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][980/1251] eta 0:01:19 lr 0.000324 time 0.2843 (0.2917) loss 3.4381 (3.3924) grad_norm 1.9836 (inf) [2021-04-16 09:19:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][990/1251] eta 0:01:16 lr 0.000324 time 0.2683 (0.2917) loss 2.4854 (3.3901) grad_norm 1.6797 (inf) [2021-04-16 09:20:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][1000/1251] eta 0:01:13 lr 0.000324 time 0.2747 (0.2915) loss 4.0872 (3.3887) grad_norm 1.7573 (inf) [2021-04-16 09:20:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][1010/1251] eta 0:01:10 lr 0.000324 time 0.2774 (0.2914) loss 3.3196 (3.3886) grad_norm 1.5834 (inf) [2021-04-16 09:20:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][1020/1251] eta 0:01:07 lr 0.000324 time 0.2950 (0.2913) loss 3.7310 (3.3895) grad_norm 1.9120 (inf) [2021-04-16 09:20:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][1030/1251] eta 0:01:04 lr 0.000324 time 0.2658 (0.2912) loss 3.3772 (3.3873) grad_norm 1.6363 (inf) [2021-04-16 09:20:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][1040/1251] eta 0:01:01 lr 0.000324 time 0.2696 (0.2911) loss 3.6620 (3.3881) grad_norm 1.6905 (inf) [2021-04-16 09:20:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][1050/1251] eta 0:00:58 lr 0.000324 time 0.2603 (0.2909) loss 2.3895 (3.3857) grad_norm 1.5673 (inf) [2021-04-16 09:20:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][1060/1251] eta 0:00:55 lr 0.000324 time 0.2888 (0.2908) loss 3.8885 (3.3843) grad_norm 1.6209 (inf) [2021-04-16 09:20:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][1070/1251] eta 0:00:52 lr 0.000323 time 0.2734 (0.2907) loss 2.6045 (3.3834) grad_norm 1.5377 (inf) [2021-04-16 09:20:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][1080/1251] eta 0:00:49 lr 0.000323 time 0.2848 (0.2906) loss 2.4904 (3.3822) grad_norm 1.6514 (inf) [2021-04-16 09:20:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][1090/1251] eta 0:00:46 lr 0.000323 time 0.2587 (0.2905) loss 3.8173 (3.3821) grad_norm 1.8367 (inf) [2021-04-16 09:20:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][1100/1251] eta 0:00:43 lr 0.000323 time 0.2762 (0.2904) loss 4.1854 (3.3830) grad_norm 1.8172 (inf) [2021-04-16 09:20:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][1110/1251] eta 0:00:40 lr 0.000323 time 0.2658 (0.2904) loss 3.1772 (3.3839) grad_norm 1.6782 (inf) [2021-04-16 09:20:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][1120/1251] eta 0:00:38 lr 0.000323 time 0.2725 (0.2903) loss 3.9466 (3.3850) grad_norm 1.6070 (inf) [2021-04-16 09:20:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][1130/1251] eta 0:00:35 lr 0.000323 time 0.2441 (0.2902) loss 3.2920 (3.3858) grad_norm 1.6298 (inf) [2021-04-16 09:20:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][1140/1251] eta 0:00:32 lr 0.000323 time 0.2666 (0.2900) loss 3.2468 (3.3845) grad_norm 1.8662 (inf) [2021-04-16 09:20:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][1150/1251] eta 0:00:29 lr 0.000323 time 0.2865 (0.2901) loss 3.7996 (3.3844) grad_norm 1.8277 (inf) [2021-04-16 09:20:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][1160/1251] eta 0:00:26 lr 0.000323 time 0.2694 (0.2901) loss 3.6611 (3.3859) grad_norm 1.8553 (inf) [2021-04-16 09:20:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][1170/1251] eta 0:00:23 lr 0.000323 time 0.2632 (0.2902) loss 4.1974 (3.3866) grad_norm 1.7514 (inf) [2021-04-16 09:20:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][1180/1251] eta 0:00:20 lr 0.000323 time 0.2730 (0.2902) loss 3.2946 (3.3865) grad_norm 1.8276 (inf) [2021-04-16 09:20:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][1190/1251] eta 0:00:17 lr 0.000323 time 0.2663 (0.2901) loss 3.5773 (3.3866) grad_norm 1.6365 (inf) [2021-04-16 09:20:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][1200/1251] eta 0:00:14 lr 0.000323 time 0.2702 (0.2900) loss 3.9287 (3.3877) grad_norm 1.9457 (inf) [2021-04-16 09:21:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][1210/1251] eta 0:00:11 lr 0.000323 time 0.2831 (0.2899) loss 2.8808 (3.3868) grad_norm 1.8924 (inf) [2021-04-16 09:21:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][1220/1251] eta 0:00:08 lr 0.000323 time 0.2883 (0.2898) loss 3.8019 (3.3859) grad_norm 1.7830 (inf) [2021-04-16 09:21:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][1230/1251] eta 0:00:06 lr 0.000323 time 0.2760 (0.2897) loss 2.8403 (3.3851) grad_norm 1.9901 (inf) [2021-04-16 09:21:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][1240/1251] eta 0:00:03 lr 0.000323 time 0.2485 (0.2896) loss 3.6922 (3.3859) grad_norm 1.7519 (inf) [2021-04-16 09:21:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [185/300][1250/1251] eta 0:00:00 lr 0.000323 time 0.2484 (0.2893) loss 3.1726 (3.3853) grad_norm 1.6988 (inf) [2021-04-16 09:21:16 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 185 training takes 0:06:06 [2021-04-16 09:21:16 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_185.pth saving...... [2021-04-16 09:21:32 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_185.pth saved !!! [2021-04-16 09:21:33 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.095 (1.095) Loss 0.9282 (0.9282) Acc@1 78.906 (78.906) Acc@5 93.750 (93.750) [2021-04-16 09:21:34 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.314 (0.222) Loss 0.9526 (0.9514) Acc@1 79.492 (77.983) Acc@5 93.652 (94.114) [2021-04-16 09:21:37 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.355 (0.240) Loss 0.9911 (0.9618) Acc@1 77.344 (77.860) Acc@5 93.945 (93.973) [2021-04-16 09:21:39 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.119 (0.241) Loss 0.9326 (0.9605) Acc@1 78.418 (77.775) Acc@5 93.848 (93.993) [2021-04-16 09:21:41 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.218) Loss 0.9840 (0.9564) Acc@1 76.855 (77.832) Acc@5 93.457 (94.060) [2021-04-16 09:21:50 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 77.806 Acc@5 94.068 [2021-04-16 09:21:50 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 77.8% [2021-04-16 09:21:50 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 77.81% [2021-04-16 09:21:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][0/1251] eta 1:57:59 lr 0.000323 time 5.6589 (5.6589) loss 4.3053 (4.3053) grad_norm 1.6203 (1.6203) [2021-04-16 09:21:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][10/1251] eta 0:16:02 lr 0.000323 time 0.4413 (0.7754) loss 2.0349 (3.1561) grad_norm 1.7816 (1.8386) [2021-04-16 09:22:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][20/1251] eta 0:11:01 lr 0.000323 time 0.2652 (0.5371) loss 3.7496 (3.2535) grad_norm 1.8389 (1.7811) [2021-04-16 09:22:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][30/1251] eta 0:09:14 lr 0.000323 time 0.2901 (0.4542) loss 3.2692 (3.3135) grad_norm 2.1832 (1.8141) [2021-04-16 09:22:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][40/1251] eta 0:08:19 lr 0.000323 time 0.2849 (0.4123) loss 3.4795 (3.2528) grad_norm 1.6757 (1.8286) [2021-04-16 09:22:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][50/1251] eta 0:07:42 lr 0.000323 time 0.2822 (0.3854) loss 2.3818 (3.2949) grad_norm 1.8037 (1.8454) [2021-04-16 09:22:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][60/1251] eta 0:07:17 lr 0.000323 time 0.2769 (0.3675) loss 3.0087 (3.2717) grad_norm 1.6154 (1.8470) [2021-04-16 09:22:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][70/1251] eta 0:06:58 lr 0.000323 time 0.2735 (0.3547) loss 2.8703 (3.2558) grad_norm 1.7658 (1.8494) [2021-04-16 09:22:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][80/1251] eta 0:06:44 lr 0.000322 time 0.2742 (0.3450) loss 3.1215 (3.2637) grad_norm 2.0451 (1.8625) [2021-04-16 09:22:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][90/1251] eta 0:06:32 lr 0.000322 time 0.2855 (0.3377) loss 3.3395 (3.2658) grad_norm 1.5973 (1.8496) [2021-04-16 09:22:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][100/1251] eta 0:06:21 lr 0.000322 time 0.2576 (0.3315) loss 2.3574 (3.2840) grad_norm 1.6204 (1.8488) [2021-04-16 09:22:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][110/1251] eta 0:06:13 lr 0.000322 time 0.2809 (0.3270) loss 3.6229 (3.2894) grad_norm 1.7117 (1.8457) [2021-04-16 09:22:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][120/1251] eta 0:06:06 lr 0.000322 time 0.2614 (0.3237) loss 3.6693 (3.3053) grad_norm 1.8635 (1.8441) [2021-04-16 09:22:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][130/1251] eta 0:06:00 lr 0.000322 time 0.2903 (0.3214) loss 3.5723 (3.3048) grad_norm 1.7393 (1.8380) [2021-04-16 09:22:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][140/1251] eta 0:05:55 lr 0.000322 time 0.2651 (0.3197) loss 3.6364 (3.3335) grad_norm 1.8304 (1.8411) [2021-04-16 09:22:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][150/1251] eta 0:05:49 lr 0.000322 time 0.2746 (0.3179) loss 3.1258 (3.3404) grad_norm 1.8170 (1.8372) [2021-04-16 09:22:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][160/1251] eta 0:05:44 lr 0.000322 time 0.2770 (0.3154) loss 2.0025 (3.3383) grad_norm 1.4935 (1.8325) [2021-04-16 09:22:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][170/1251] eta 0:05:38 lr 0.000322 time 0.3022 (0.3133) loss 3.6487 (3.3260) grad_norm 1.7201 (1.8261) [2021-04-16 09:22:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][180/1251] eta 0:05:33 lr 0.000322 time 0.3044 (0.3114) loss 2.8696 (3.3212) grad_norm 1.6747 (1.8299) [2021-04-16 09:22:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][190/1251] eta 0:05:28 lr 0.000322 time 0.2857 (0.3099) loss 3.6059 (3.3229) grad_norm 1.6107 (1.8245) [2021-04-16 09:22:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][200/1251] eta 0:05:23 lr 0.000322 time 0.2601 (0.3081) loss 2.7506 (3.3220) grad_norm 1.6315 (1.8215) [2021-04-16 09:22:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][210/1251] eta 0:05:19 lr 0.000322 time 0.2645 (0.3068) loss 3.6207 (3.3174) grad_norm 1.8660 (1.8191) [2021-04-16 09:22:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][220/1251] eta 0:05:15 lr 0.000322 time 0.3098 (0.3057) loss 3.3595 (3.3290) grad_norm 1.7263 (1.8171) [2021-04-16 09:23:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][230/1251] eta 0:05:10 lr 0.000322 time 0.3014 (0.3045) loss 3.0326 (3.3254) grad_norm 1.6980 (1.8173) [2021-04-16 09:23:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][240/1251] eta 0:05:06 lr 0.000322 time 0.2738 (0.3034) loss 4.2041 (3.3306) grad_norm 2.0470 (1.8188) [2021-04-16 09:23:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][250/1251] eta 0:05:02 lr 0.000322 time 0.2670 (0.3024) loss 3.5658 (3.3375) grad_norm 1.9435 (1.8200) [2021-04-16 09:23:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][260/1251] eta 0:04:58 lr 0.000322 time 0.3030 (0.3017) loss 3.7423 (3.3452) grad_norm 1.7777 (1.8206) [2021-04-16 09:23:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][270/1251] eta 0:04:55 lr 0.000322 time 0.2689 (0.3008) loss 3.4866 (3.3534) grad_norm 1.8809 (1.8258) [2021-04-16 09:23:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][280/1251] eta 0:04:51 lr 0.000322 time 0.2663 (0.3000) loss 3.4278 (3.3452) grad_norm 1.9477 (1.8358) [2021-04-16 09:23:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][290/1251] eta 0:04:47 lr 0.000322 time 0.2874 (0.2992) loss 3.5158 (3.3380) grad_norm 2.0977 (1.8367) [2021-04-16 09:23:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][300/1251] eta 0:04:43 lr 0.000322 time 0.2864 (0.2984) loss 3.5322 (3.3286) grad_norm 1.5084 (1.8338) [2021-04-16 09:23:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][310/1251] eta 0:04:40 lr 0.000322 time 0.2867 (0.2979) loss 3.9193 (3.3322) grad_norm 1.7659 (1.8336) [2021-04-16 09:23:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][320/1251] eta 0:04:36 lr 0.000322 time 0.2799 (0.2975) loss 3.6916 (3.3345) grad_norm 1.7872 (1.8329) [2021-04-16 09:23:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][330/1251] eta 0:04:33 lr 0.000322 time 0.2837 (0.2971) loss 3.9301 (3.3464) grad_norm 1.8198 (1.8337) [2021-04-16 09:23:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][340/1251] eta 0:04:30 lr 0.000321 time 0.3013 (0.2965) loss 3.6611 (3.3416) grad_norm 1.6049 (1.8312) [2021-04-16 09:23:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][350/1251] eta 0:04:26 lr 0.000321 time 0.2676 (0.2960) loss 3.6374 (3.3430) grad_norm 1.8188 (1.8300) [2021-04-16 09:23:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][360/1251] eta 0:04:23 lr 0.000321 time 0.2779 (0.2962) loss 4.0150 (3.3429) grad_norm 1.7039 (1.8299) [2021-04-16 09:23:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][370/1251] eta 0:04:21 lr 0.000321 time 0.3002 (0.2964) loss 3.6841 (3.3434) grad_norm 1.7490 (1.8310) [2021-04-16 09:23:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][380/1251] eta 0:04:17 lr 0.000321 time 0.3043 (0.2962) loss 3.6809 (3.3471) grad_norm 1.8167 (1.8350) [2021-04-16 09:23:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][390/1251] eta 0:04:14 lr 0.000321 time 0.2715 (0.2957) loss 3.3093 (3.3522) grad_norm 2.1225 (1.8401) [2021-04-16 09:23:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][400/1251] eta 0:04:11 lr 0.000321 time 0.2839 (0.2954) loss 3.9567 (3.3554) grad_norm 1.9246 (1.8410) [2021-04-16 09:23:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][410/1251] eta 0:04:08 lr 0.000321 time 0.2974 (0.2950) loss 3.7790 (3.3566) grad_norm 1.9269 (1.8422) [2021-04-16 09:23:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][420/1251] eta 0:04:04 lr 0.000321 time 0.2939 (0.2947) loss 3.1899 (3.3625) grad_norm 1.7018 (1.8419) [2021-04-16 09:23:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][430/1251] eta 0:04:01 lr 0.000321 time 0.2937 (0.2946) loss 3.2457 (3.3602) grad_norm 1.6537 (1.8408) [2021-04-16 09:24:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][440/1251] eta 0:03:58 lr 0.000321 time 0.2922 (0.2945) loss 3.5968 (3.3578) grad_norm 2.3621 (1.8395) [2021-04-16 09:24:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][450/1251] eta 0:03:55 lr 0.000321 time 0.3097 (0.2942) loss 3.3194 (3.3573) grad_norm 1.4537 (1.8408) [2021-04-16 09:24:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][460/1251] eta 0:03:52 lr 0.000321 time 0.2763 (0.2940) loss 3.8860 (3.3610) grad_norm 1.8258 (1.8373) [2021-04-16 09:24:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][470/1251] eta 0:03:49 lr 0.000321 time 0.2813 (0.2937) loss 3.4317 (3.3604) grad_norm 1.9036 (1.8391) [2021-04-16 09:24:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][480/1251] eta 0:03:46 lr 0.000321 time 0.2783 (0.2934) loss 3.8987 (3.3662) grad_norm 2.0367 (1.8391) [2021-04-16 09:24:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][490/1251] eta 0:03:43 lr 0.000321 time 0.2746 (0.2931) loss 3.8805 (3.3701) grad_norm 1.6904 (1.8386) [2021-04-16 09:24:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][500/1251] eta 0:03:39 lr 0.000321 time 0.3152 (0.2928) loss 3.4457 (3.3674) grad_norm 1.9879 (1.8395) [2021-04-16 09:24:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][510/1251] eta 0:03:36 lr 0.000321 time 0.2843 (0.2925) loss 3.4220 (3.3675) grad_norm 1.6657 (1.8414) [2021-04-16 09:24:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][520/1251] eta 0:03:33 lr 0.000321 time 0.2764 (0.2921) loss 2.9891 (3.3713) grad_norm 1.6429 (1.8411) [2021-04-16 09:24:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][530/1251] eta 0:03:30 lr 0.000321 time 0.2766 (0.2919) loss 2.2951 (3.3739) grad_norm 2.0218 (1.8413) [2021-04-16 09:24:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][540/1251] eta 0:03:27 lr 0.000321 time 0.2695 (0.2917) loss 2.8131 (3.3706) grad_norm 1.7678 (1.8432) [2021-04-16 09:24:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][550/1251] eta 0:03:24 lr 0.000321 time 0.3056 (0.2915) loss 3.1947 (3.3747) grad_norm 1.7974 (1.8437) [2021-04-16 09:24:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][560/1251] eta 0:03:21 lr 0.000321 time 0.2646 (0.2914) loss 3.4816 (3.3755) grad_norm 2.0771 (1.8460) [2021-04-16 09:24:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][570/1251] eta 0:03:18 lr 0.000321 time 0.2677 (0.2916) loss 3.9764 (3.3749) grad_norm 1.8734 (1.8463) [2021-04-16 09:24:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][580/1251] eta 0:03:15 lr 0.000321 time 0.2796 (0.2917) loss 2.4616 (3.3763) grad_norm 1.7239 (1.8446) [2021-04-16 09:24:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][590/1251] eta 0:03:12 lr 0.000321 time 0.2793 (0.2918) loss 3.0564 (3.3801) grad_norm 1.6362 (1.8438) [2021-04-16 09:24:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][600/1251] eta 0:03:09 lr 0.000320 time 0.2864 (0.2917) loss 2.2327 (3.3823) grad_norm 2.2493 (1.8431) [2021-04-16 09:24:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][610/1251] eta 0:03:06 lr 0.000320 time 0.2858 (0.2915) loss 3.5039 (3.3802) grad_norm 1.9749 (1.8429) [2021-04-16 09:24:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][620/1251] eta 0:03:03 lr 0.000320 time 0.2872 (0.2914) loss 3.8833 (3.3808) grad_norm 1.8621 (1.8431) [2021-04-16 09:24:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][630/1251] eta 0:03:00 lr 0.000320 time 0.2671 (0.2913) loss 3.4821 (3.3825) grad_norm 1.6694 (1.8443) [2021-04-16 09:24:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][640/1251] eta 0:02:57 lr 0.000320 time 0.2633 (0.2911) loss 3.0017 (3.3786) grad_norm 1.6780 (1.8438) [2021-04-16 09:24:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][650/1251] eta 0:02:54 lr 0.000320 time 0.2806 (0.2909) loss 4.0463 (3.3811) grad_norm 1.5743 (1.8429) [2021-04-16 09:25:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][660/1251] eta 0:02:51 lr 0.000320 time 0.2797 (0.2907) loss 3.5089 (3.3811) grad_norm 1.6761 (1.8438) [2021-04-16 09:25:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][670/1251] eta 0:02:48 lr 0.000320 time 0.2925 (0.2905) loss 3.8849 (3.3852) grad_norm 1.5571 (1.8424) [2021-04-16 09:25:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][680/1251] eta 0:02:45 lr 0.000320 time 0.2872 (0.2903) loss 3.6556 (3.3909) grad_norm 1.7615 (1.8415) [2021-04-16 09:25:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][690/1251] eta 0:02:42 lr 0.000320 time 0.2772 (0.2904) loss 3.8172 (3.3921) grad_norm 1.6749 (1.8396) [2021-04-16 09:25:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][700/1251] eta 0:02:39 lr 0.000320 time 0.2884 (0.2901) loss 3.2707 (3.3904) grad_norm 1.6622 (1.8393) [2021-04-16 09:25:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][710/1251] eta 0:02:36 lr 0.000320 time 0.2781 (0.2900) loss 3.0985 (3.3913) grad_norm 1.8038 (1.8388) [2021-04-16 09:25:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][720/1251] eta 0:02:34 lr 0.000320 time 0.4292 (0.2903) loss 4.1734 (3.3933) grad_norm 2.3175 (1.8394) [2021-04-16 09:25:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][730/1251] eta 0:02:31 lr 0.000320 time 0.2558 (0.2901) loss 2.9859 (3.3906) grad_norm 1.8867 (1.8390) [2021-04-16 09:25:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][740/1251] eta 0:02:28 lr 0.000320 time 0.2608 (0.2900) loss 3.5726 (3.3925) grad_norm 1.9378 (1.8388) [2021-04-16 09:25:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][750/1251] eta 0:02:25 lr 0.000320 time 0.2659 (0.2899) loss 3.2126 (3.3937) grad_norm 1.6102 (1.8375) [2021-04-16 09:25:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][760/1251] eta 0:02:22 lr 0.000320 time 0.2886 (0.2897) loss 3.1476 (3.3911) grad_norm 1.5520 (1.8365) [2021-04-16 09:25:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][770/1251] eta 0:02:19 lr 0.000320 time 0.2698 (0.2896) loss 3.8169 (3.3925) grad_norm 1.6041 (1.8359) [2021-04-16 09:25:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][780/1251] eta 0:02:16 lr 0.000320 time 0.2939 (0.2893) loss 3.3738 (3.3919) grad_norm 1.6922 (1.8378) [2021-04-16 09:25:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][790/1251] eta 0:02:13 lr 0.000320 time 0.2969 (0.2892) loss 4.2495 (3.3922) grad_norm 1.8145 (1.8373) [2021-04-16 09:25:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][800/1251] eta 0:02:10 lr 0.000320 time 0.2736 (0.2890) loss 3.5856 (3.3923) grad_norm 1.9494 (1.8365) [2021-04-16 09:25:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][810/1251] eta 0:02:07 lr 0.000320 time 0.2565 (0.2889) loss 2.9233 (3.3900) grad_norm 1.7834 (1.8366) [2021-04-16 09:25:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][820/1251] eta 0:02:04 lr 0.000320 time 0.2645 (0.2887) loss 3.7389 (3.3934) grad_norm 1.9462 (1.8368) [2021-04-16 09:25:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][830/1251] eta 0:02:01 lr 0.000320 time 0.2858 (0.2886) loss 4.0630 (3.3954) grad_norm 1.6679 (1.8366) [2021-04-16 09:25:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][840/1251] eta 0:01:58 lr 0.000320 time 0.2799 (0.2885) loss 3.8674 (3.3977) grad_norm 1.5331 (1.8360) [2021-04-16 09:25:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][850/1251] eta 0:01:55 lr 0.000320 time 0.2698 (0.2883) loss 3.0543 (3.3980) grad_norm 1.8610 (1.8351) [2021-04-16 09:25:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][860/1251] eta 0:01:52 lr 0.000319 time 0.2785 (0.2882) loss 2.5657 (3.3973) grad_norm 1.6543 (1.8358) [2021-04-16 09:26:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][870/1251] eta 0:01:49 lr 0.000319 time 0.2547 (0.2881) loss 3.6934 (3.3936) grad_norm 2.0012 (1.8355) [2021-04-16 09:26:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][880/1251] eta 0:01:46 lr 0.000319 time 0.2674 (0.2880) loss 3.0987 (3.3967) grad_norm 1.7550 (1.8363) [2021-04-16 09:26:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][890/1251] eta 0:01:43 lr 0.000319 time 0.2999 (0.2880) loss 3.6321 (3.3996) grad_norm 1.9129 (1.8380) [2021-04-16 09:26:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][900/1251] eta 0:01:41 lr 0.000319 time 0.2968 (0.2879) loss 3.7585 (3.4016) grad_norm 1.7476 (1.8381) [2021-04-16 09:26:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][910/1251] eta 0:01:38 lr 0.000319 time 0.2732 (0.2878) loss 2.3425 (3.3979) grad_norm 1.8744 (1.8371) [2021-04-16 09:26:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][920/1251] eta 0:01:35 lr 0.000319 time 0.2572 (0.2877) loss 3.6742 (3.3984) grad_norm 1.7487 (1.8366) [2021-04-16 09:26:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][930/1251] eta 0:01:32 lr 0.000319 time 0.2903 (0.2876) loss 2.3054 (3.3950) grad_norm 1.9240 (1.8359) [2021-04-16 09:26:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][940/1251] eta 0:01:29 lr 0.000319 time 0.4279 (0.2879) loss 3.8615 (3.3983) grad_norm 1.9022 (1.8361) [2021-04-16 09:26:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][950/1251] eta 0:01:26 lr 0.000319 time 0.2475 (0.2878) loss 3.9244 (3.4006) grad_norm 1.9554 (1.8356) [2021-04-16 09:26:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][960/1251] eta 0:01:23 lr 0.000319 time 0.2971 (0.2877) loss 3.8782 (3.4004) grad_norm 1.6405 (1.8349) [2021-04-16 09:26:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][970/1251] eta 0:01:20 lr 0.000319 time 0.2436 (0.2876) loss 3.4384 (3.3991) grad_norm 1.7901 (1.8337) [2021-04-16 09:26:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][980/1251] eta 0:01:17 lr 0.000319 time 0.2799 (0.2875) loss 3.7311 (3.3992) grad_norm 1.7138 (1.8335) [2021-04-16 09:26:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][990/1251] eta 0:01:15 lr 0.000319 time 0.2533 (0.2874) loss 2.7890 (3.3993) grad_norm 1.7978 (1.8331) [2021-04-16 09:26:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][1000/1251] eta 0:01:12 lr 0.000319 time 0.2774 (0.2873) loss 2.6432 (3.3967) grad_norm 1.9696 (1.8335) [2021-04-16 09:26:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][1010/1251] eta 0:01:09 lr 0.000319 time 0.3088 (0.2872) loss 3.6086 (3.3972) grad_norm 1.6575 (1.8338) [2021-04-16 09:26:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][1020/1251] eta 0:01:06 lr 0.000319 time 0.2745 (0.2872) loss 4.1857 (3.3982) grad_norm 1.8039 (1.8342) [2021-04-16 09:26:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][1030/1251] eta 0:01:03 lr 0.000319 time 0.2595 (0.2871) loss 2.3916 (3.3963) grad_norm 1.7684 (1.8338) [2021-04-16 09:26:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][1040/1251] eta 0:01:00 lr 0.000319 time 0.2694 (0.2871) loss 4.2467 (3.3975) grad_norm 1.6862 (1.8326) [2021-04-16 09:26:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][1050/1251] eta 0:00:57 lr 0.000319 time 0.2990 (0.2870) loss 3.7160 (3.3976) grad_norm 1.8996 (1.8331) [2021-04-16 09:26:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][1060/1251] eta 0:00:54 lr 0.000319 time 0.2737 (0.2870) loss 3.6803 (3.3968) grad_norm 2.0510 (1.8323) [2021-04-16 09:26:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][1070/1251] eta 0:00:51 lr 0.000319 time 0.2987 (0.2869) loss 2.7421 (3.3943) grad_norm 1.8598 (1.8327) [2021-04-16 09:27:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][1080/1251] eta 0:00:49 lr 0.000319 time 0.2817 (0.2868) loss 3.9432 (3.3939) grad_norm 1.5879 (1.8325) [2021-04-16 09:27:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][1090/1251] eta 0:00:46 lr 0.000319 time 0.2577 (0.2867) loss 3.4279 (3.3969) grad_norm 1.8244 (1.8328) [2021-04-16 09:27:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][1100/1251] eta 0:00:43 lr 0.000319 time 0.2807 (0.2867) loss 3.1365 (3.3959) grad_norm 1.5686 (1.8337) [2021-04-16 09:27:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][1110/1251] eta 0:00:40 lr 0.000319 time 0.2832 (0.2866) loss 3.5739 (3.3959) grad_norm 2.1670 (1.8335) [2021-04-16 09:27:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][1120/1251] eta 0:00:37 lr 0.000318 time 0.2983 (0.2865) loss 3.8738 (3.3960) grad_norm 1.8322 (1.8337) [2021-04-16 09:27:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][1130/1251] eta 0:00:34 lr 0.000318 time 0.2709 (0.2865) loss 3.3326 (3.3948) grad_norm 1.7029 (1.8361) [2021-04-16 09:27:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][1140/1251] eta 0:00:31 lr 0.000318 time 0.2988 (0.2866) loss 3.7272 (3.3968) grad_norm 1.7501 (1.8355) [2021-04-16 09:27:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][1150/1251] eta 0:00:28 lr 0.000318 time 0.3070 (0.2867) loss 4.1407 (3.3994) grad_norm 1.8447 (1.8348) [2021-04-16 09:27:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][1160/1251] eta 0:00:26 lr 0.000318 time 0.2897 (0.2868) loss 3.7475 (3.3996) grad_norm 1.7211 (1.8348) [2021-04-16 09:27:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][1170/1251] eta 0:00:23 lr 0.000318 time 0.2618 (0.2869) loss 3.9070 (3.3974) grad_norm 1.7363 (1.8343) [2021-04-16 09:27:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][1180/1251] eta 0:00:20 lr 0.000318 time 0.2738 (0.2868) loss 2.7380 (3.3958) grad_norm 1.9012 (1.8339) [2021-04-16 09:27:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][1190/1251] eta 0:00:17 lr 0.000318 time 0.2781 (0.2867) loss 3.9140 (3.3965) grad_norm 1.6009 (1.8349) [2021-04-16 09:27:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][1200/1251] eta 0:00:14 lr 0.000318 time 0.2961 (0.2868) loss 4.7378 (3.3980) grad_norm 1.6033 (1.8348) [2021-04-16 09:27:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][1210/1251] eta 0:00:11 lr 0.000318 time 0.2804 (0.2866) loss 3.3878 (3.3990) grad_norm 1.8355 (1.8344) [2021-04-16 09:27:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][1220/1251] eta 0:00:08 lr 0.000318 time 0.2799 (0.2866) loss 2.3804 (3.3995) grad_norm 1.5592 (1.8333) [2021-04-16 09:27:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][1230/1251] eta 0:00:06 lr 0.000318 time 0.2813 (0.2865) loss 2.9947 (3.3996) grad_norm 1.5587 (1.8336) [2021-04-16 09:27:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][1240/1251] eta 0:00:03 lr 0.000318 time 0.2482 (0.2863) loss 3.7370 (3.4007) grad_norm 1.8468 (1.8337) [2021-04-16 09:27:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [186/300][1250/1251] eta 0:00:00 lr 0.000318 time 0.2485 (0.2860) loss 3.7137 (3.4001) grad_norm 1.7263 (1.8348) [2021-04-16 09:27:51 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 186 training takes 0:06:01 [2021-04-16 09:27:51 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_186.pth saving...... [2021-04-16 09:28:01 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_186.pth saved !!! [2021-04-16 09:28:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.116 (1.116) Loss 0.9853 (0.9853) Acc@1 76.172 (76.172) Acc@5 93.945 (93.945) [2021-04-16 09:28:04 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.105 (0.219) Loss 0.9360 (0.9256) Acc@1 76.465 (77.663) Acc@5 94.629 (94.283) [2021-04-16 09:28:07 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.365 (0.260) Loss 0.9927 (0.9370) Acc@1 77.539 (77.832) Acc@5 92.871 (94.089) [2021-04-16 09:28:08 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.129 (0.224) Loss 0.9938 (0.9377) Acc@1 76.270 (77.797) Acc@5 94.238 (94.147) [2021-04-16 09:28:10 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.144 (0.210) Loss 0.9431 (0.9426) Acc@1 76.758 (77.620) Acc@5 94.922 (94.145) [2021-04-16 09:28:15 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 77.546 Acc@5 94.140 [2021-04-16 09:28:15 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 77.5% [2021-04-16 09:28:15 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 77.81% [2021-04-16 09:28:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][0/1251] eta 1:23:56 lr 0.000318 time 4.0262 (4.0262) loss 2.3209 (2.3209) grad_norm 1.7625 (1.7625) [2021-04-16 09:28:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][10/1251] eta 0:12:45 lr 0.000318 time 0.2777 (0.6169) loss 3.2900 (3.4803) grad_norm 1.7115 (1.7570) [2021-04-16 09:28:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][20/1251] eta 0:09:20 lr 0.000318 time 0.2937 (0.4556) loss 4.0413 (3.4296) grad_norm 1.5582 (1.7582) [2021-04-16 09:28:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][30/1251] eta 0:08:03 lr 0.000318 time 0.2748 (0.3962) loss 2.1972 (3.2991) grad_norm nan (nan) [2021-04-16 09:28:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][40/1251] eta 0:07:23 lr 0.000318 time 0.2661 (0.3663) loss 3.1230 (3.2970) grad_norm 1.5688 (nan) [2021-04-16 09:28:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][50/1251] eta 0:06:59 lr 0.000318 time 0.2838 (0.3492) loss 2.1747 (3.3289) grad_norm 1.9975 (nan) [2021-04-16 09:28:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][60/1251] eta 0:06:42 lr 0.000318 time 0.2733 (0.3376) loss 3.9361 (3.3304) grad_norm 1.7244 (nan) [2021-04-16 09:28:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][70/1251] eta 0:06:29 lr 0.000318 time 0.2953 (0.3297) loss 3.6151 (3.3610) grad_norm 1.6807 (nan) [2021-04-16 09:28:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][80/1251] eta 0:06:18 lr 0.000318 time 0.2864 (0.3235) loss 2.9790 (3.3340) grad_norm 1.6879 (nan) [2021-04-16 09:28:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][90/1251] eta 0:06:09 lr 0.000318 time 0.2932 (0.3183) loss 2.9795 (3.3226) grad_norm 1.8774 (nan) [2021-04-16 09:28:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][100/1251] eta 0:06:01 lr 0.000318 time 0.2641 (0.3140) loss 2.9780 (3.2968) grad_norm 1.9856 (nan) [2021-04-16 09:28:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][110/1251] eta 0:05:54 lr 0.000318 time 0.2994 (0.3111) loss 2.3143 (3.2884) grad_norm 1.8306 (nan) [2021-04-16 09:28:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][120/1251] eta 0:05:49 lr 0.000318 time 0.2732 (0.3088) loss 3.3621 (3.3022) grad_norm 1.7095 (nan) [2021-04-16 09:28:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][130/1251] eta 0:05:43 lr 0.000317 time 0.2834 (0.3060) loss 2.5706 (3.3117) grad_norm 1.7107 (nan) [2021-04-16 09:28:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][140/1251] eta 0:05:38 lr 0.000317 time 0.2794 (0.3051) loss 3.8646 (3.3119) grad_norm 1.7252 (nan) [2021-04-16 09:29:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][150/1251] eta 0:05:35 lr 0.000317 time 0.3980 (0.3051) loss 3.2388 (3.3017) grad_norm 1.7530 (nan) [2021-04-16 09:29:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][160/1251] eta 0:05:31 lr 0.000317 time 0.2736 (0.3035) loss 2.4875 (3.2946) grad_norm 2.0349 (nan) [2021-04-16 09:29:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][170/1251] eta 0:05:26 lr 0.000317 time 0.2719 (0.3017) loss 3.7942 (3.3001) grad_norm 1.6702 (nan) [2021-04-16 09:29:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][180/1251] eta 0:05:22 lr 0.000317 time 0.2913 (0.3007) loss 2.9573 (3.2895) grad_norm 1.5476 (nan) [2021-04-16 09:29:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][190/1251] eta 0:05:18 lr 0.000317 time 0.2738 (0.2997) loss 3.0556 (3.2864) grad_norm 1.6062 (nan) [2021-04-16 09:29:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][200/1251] eta 0:05:13 lr 0.000317 time 0.2728 (0.2985) loss 2.8266 (3.2823) grad_norm 1.7441 (nan) [2021-04-16 09:29:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][210/1251] eta 0:05:09 lr 0.000317 time 0.2624 (0.2974) loss 2.2785 (3.2885) grad_norm 1.8223 (nan) [2021-04-16 09:29:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][220/1251] eta 0:05:06 lr 0.000317 time 0.2955 (0.2974) loss 3.7824 (3.2779) grad_norm 1.7769 (nan) [2021-04-16 09:29:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][230/1251] eta 0:05:02 lr 0.000317 time 0.2910 (0.2966) loss 2.7601 (3.2873) grad_norm 1.8900 (nan) [2021-04-16 09:29:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][240/1251] eta 0:04:59 lr 0.000317 time 0.3015 (0.2960) loss 3.2462 (3.2881) grad_norm 1.7273 (nan) [2021-04-16 09:29:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][250/1251] eta 0:04:55 lr 0.000317 time 0.2584 (0.2953) loss 2.2180 (3.2799) grad_norm 1.9768 (nan) [2021-04-16 09:29:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][260/1251] eta 0:04:51 lr 0.000317 time 0.2665 (0.2946) loss 2.8974 (3.2818) grad_norm 1.8501 (nan) [2021-04-16 09:29:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][270/1251] eta 0:04:48 lr 0.000317 time 0.2467 (0.2940) loss 3.4446 (3.2928) grad_norm 2.0257 (nan) [2021-04-16 09:29:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][280/1251] eta 0:04:44 lr 0.000317 time 0.2808 (0.2935) loss 3.3256 (3.2851) grad_norm 1.8295 (nan) [2021-04-16 09:29:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][290/1251] eta 0:04:41 lr 0.000317 time 0.2595 (0.2928) loss 3.3046 (3.2926) grad_norm 1.7051 (nan) [2021-04-16 09:29:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][300/1251] eta 0:04:37 lr 0.000317 time 0.2851 (0.2922) loss 4.1316 (3.3012) grad_norm 1.6013 (nan) [2021-04-16 09:29:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][310/1251] eta 0:04:34 lr 0.000317 time 0.2854 (0.2917) loss 3.3847 (3.3081) grad_norm 2.1643 (nan) [2021-04-16 09:29:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][320/1251] eta 0:04:31 lr 0.000317 time 0.2695 (0.2913) loss 3.6662 (3.3080) grad_norm 2.1514 (nan) [2021-04-16 09:29:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][330/1251] eta 0:04:28 lr 0.000317 time 0.2582 (0.2912) loss 3.9793 (3.3070) grad_norm 1.8176 (nan) [2021-04-16 09:29:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][340/1251] eta 0:04:24 lr 0.000317 time 0.2571 (0.2908) loss 3.4134 (3.3043) grad_norm 1.7030 (nan) [2021-04-16 09:29:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][350/1251] eta 0:04:22 lr 0.000317 time 0.2832 (0.2909) loss 3.9524 (3.3033) grad_norm 1.9451 (nan) [2021-04-16 09:30:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][360/1251] eta 0:04:19 lr 0.000317 time 0.2814 (0.2914) loss 3.5401 (3.3031) grad_norm 1.5585 (nan) [2021-04-16 09:30:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][370/1251] eta 0:04:16 lr 0.000317 time 0.2930 (0.2912) loss 3.4164 (3.3051) grad_norm 1.9608 (nan) [2021-04-16 09:30:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][380/1251] eta 0:04:13 lr 0.000317 time 0.2741 (0.2909) loss 3.0345 (3.2987) grad_norm 1.7100 (nan) [2021-04-16 09:30:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][390/1251] eta 0:04:10 lr 0.000316 time 0.2632 (0.2906) loss 3.1880 (3.2993) grad_norm 1.4500 (nan) [2021-04-16 09:30:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][400/1251] eta 0:04:07 lr 0.000316 time 0.2837 (0.2903) loss 3.6085 (3.3035) grad_norm 2.1723 (nan) [2021-04-16 09:30:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][410/1251] eta 0:04:03 lr 0.000316 time 0.2823 (0.2900) loss 4.0235 (3.3139) grad_norm 1.9608 (nan) [2021-04-16 09:30:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][420/1251] eta 0:04:00 lr 0.000316 time 0.3024 (0.2896) loss 3.2906 (3.3153) grad_norm 1.9111 (nan) [2021-04-16 09:30:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][430/1251] eta 0:03:57 lr 0.000316 time 0.2856 (0.2894) loss 2.6558 (3.3106) grad_norm 2.2149 (nan) [2021-04-16 09:30:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][440/1251] eta 0:03:54 lr 0.000316 time 0.2775 (0.2892) loss 4.3485 (3.3097) grad_norm 1.8910 (nan) [2021-04-16 09:30:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][450/1251] eta 0:03:51 lr 0.000316 time 0.2805 (0.2890) loss 2.2413 (3.3147) grad_norm 1.8962 (nan) [2021-04-16 09:30:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][460/1251] eta 0:03:48 lr 0.000316 time 0.2775 (0.2887) loss 3.0133 (3.3100) grad_norm 1.9448 (nan) [2021-04-16 09:30:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][470/1251] eta 0:03:45 lr 0.000316 time 0.2811 (0.2885) loss 3.9402 (3.3082) grad_norm 1.8458 (nan) [2021-04-16 09:30:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][480/1251] eta 0:03:42 lr 0.000316 time 0.3004 (0.2884) loss 3.7885 (3.3107) grad_norm 1.7340 (nan) [2021-04-16 09:30:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][490/1251] eta 0:03:39 lr 0.000316 time 0.2578 (0.2884) loss 2.4019 (3.3128) grad_norm 1.8323 (nan) [2021-04-16 09:30:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][500/1251] eta 0:03:36 lr 0.000316 time 0.2575 (0.2882) loss 3.7644 (3.3179) grad_norm 2.7572 (nan) [2021-04-16 09:30:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][510/1251] eta 0:03:33 lr 0.000316 time 0.3076 (0.2880) loss 2.8021 (3.3216) grad_norm 1.6787 (nan) [2021-04-16 09:30:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][520/1251] eta 0:03:30 lr 0.000316 time 0.2652 (0.2881) loss 3.3460 (3.3174) grad_norm 2.1791 (nan) [2021-04-16 09:30:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][530/1251] eta 0:03:27 lr 0.000316 time 0.3019 (0.2880) loss 3.5207 (3.3171) grad_norm 1.7966 (nan) [2021-04-16 09:30:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][540/1251] eta 0:03:24 lr 0.000316 time 0.2782 (0.2877) loss 2.3920 (3.3203) grad_norm 1.7513 (nan) [2021-04-16 09:30:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][550/1251] eta 0:03:21 lr 0.000316 time 0.2618 (0.2876) loss 3.8060 (3.3217) grad_norm 1.7759 (nan) [2021-04-16 09:30:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][560/1251] eta 0:03:18 lr 0.000316 time 0.3055 (0.2875) loss 3.8685 (3.3210) grad_norm 2.0333 (nan) [2021-04-16 09:31:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][570/1251] eta 0:03:15 lr 0.000316 time 0.3153 (0.2877) loss 3.8735 (3.3260) grad_norm 1.7260 (nan) [2021-04-16 09:31:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][580/1251] eta 0:03:13 lr 0.000316 time 0.2746 (0.2880) loss 3.9765 (3.3277) grad_norm 1.6924 (nan) [2021-04-16 09:31:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][590/1251] eta 0:03:10 lr 0.000316 time 0.3295 (0.2879) loss 2.8693 (3.3275) grad_norm 1.9964 (nan) [2021-04-16 09:31:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][600/1251] eta 0:03:07 lr 0.000316 time 0.2566 (0.2877) loss 3.5871 (3.3282) grad_norm 1.8248 (nan) [2021-04-16 09:31:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][610/1251] eta 0:03:04 lr 0.000316 time 0.2616 (0.2875) loss 3.4825 (3.3280) grad_norm 1.8948 (nan) [2021-04-16 09:31:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][620/1251] eta 0:03:01 lr 0.000316 time 0.2712 (0.2873) loss 3.3961 (3.3303) grad_norm 1.9329 (nan) [2021-04-16 09:31:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][630/1251] eta 0:02:58 lr 0.000316 time 0.2901 (0.2870) loss 3.6789 (3.3301) grad_norm 2.1941 (nan) [2021-04-16 09:31:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][640/1251] eta 0:02:55 lr 0.000316 time 0.2592 (0.2869) loss 2.5800 (3.3297) grad_norm 1.7110 (nan) [2021-04-16 09:31:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][650/1251] eta 0:02:52 lr 0.000315 time 0.2852 (0.2875) loss 3.5143 (3.3354) grad_norm 1.8895 (nan) [2021-04-16 09:31:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][660/1251] eta 0:02:49 lr 0.000315 time 0.2903 (0.2876) loss 2.6234 (3.3332) grad_norm 2.3664 (nan) [2021-04-16 09:31:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][670/1251] eta 0:02:46 lr 0.000315 time 0.2762 (0.2874) loss 3.9009 (3.3356) grad_norm 1.7891 (nan) [2021-04-16 09:31:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][680/1251] eta 0:02:44 lr 0.000315 time 0.2740 (0.2873) loss 4.1961 (3.3324) grad_norm 1.9980 (nan) [2021-04-16 09:31:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][690/1251] eta 0:02:41 lr 0.000315 time 0.2729 (0.2872) loss 3.0238 (3.3323) grad_norm 1.7769 (nan) [2021-04-16 09:31:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][700/1251] eta 0:02:38 lr 0.000315 time 0.2811 (0.2872) loss 4.1464 (3.3334) grad_norm 1.7590 (nan) [2021-04-16 09:31:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][710/1251] eta 0:02:35 lr 0.000315 time 0.2664 (0.2870) loss 2.1875 (3.3319) grad_norm 1.7622 (nan) [2021-04-16 09:31:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][720/1251] eta 0:02:32 lr 0.000315 time 0.2784 (0.2871) loss 3.8932 (3.3343) grad_norm 2.0806 (nan) [2021-04-16 09:31:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][730/1251] eta 0:02:29 lr 0.000315 time 0.2638 (0.2872) loss 3.5563 (3.3364) grad_norm 1.6155 (nan) [2021-04-16 09:31:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][740/1251] eta 0:02:26 lr 0.000315 time 0.2710 (0.2870) loss 3.3905 (3.3360) grad_norm 1.8574 (nan) [2021-04-16 09:31:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][750/1251] eta 0:02:23 lr 0.000315 time 0.2575 (0.2869) loss 3.9965 (3.3420) grad_norm 2.0062 (nan) [2021-04-16 09:31:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][760/1251] eta 0:02:20 lr 0.000315 time 0.2581 (0.2868) loss 3.9884 (3.3420) grad_norm 1.8656 (nan) [2021-04-16 09:31:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][770/1251] eta 0:02:17 lr 0.000315 time 0.2732 (0.2867) loss 3.5745 (3.3476) grad_norm 2.5130 (nan) [2021-04-16 09:31:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][780/1251] eta 0:02:15 lr 0.000315 time 0.2915 (0.2867) loss 3.8810 (3.3488) grad_norm 1.9637 (nan) [2021-04-16 09:32:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][790/1251] eta 0:02:12 lr 0.000315 time 0.2759 (0.2867) loss 3.7137 (3.3462) grad_norm 1.5750 (nan) [2021-04-16 09:32:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][800/1251] eta 0:02:09 lr 0.000315 time 0.2845 (0.2866) loss 3.7553 (3.3460) grad_norm 1.8106 (nan) [2021-04-16 09:32:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][810/1251] eta 0:02:06 lr 0.000315 time 0.2625 (0.2864) loss 3.5681 (3.3453) grad_norm 1.9227 (nan) [2021-04-16 09:32:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][820/1251] eta 0:02:03 lr 0.000315 time 0.2651 (0.2863) loss 2.9250 (3.3457) grad_norm 1.7796 (nan) [2021-04-16 09:32:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][830/1251] eta 0:02:00 lr 0.000315 time 0.2930 (0.2862) loss 3.1134 (3.3449) grad_norm 2.0052 (nan) [2021-04-16 09:32:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][840/1251] eta 0:01:57 lr 0.000315 time 0.2740 (0.2860) loss 3.7202 (3.3451) grad_norm 1.8187 (nan) [2021-04-16 09:32:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][850/1251] eta 0:01:54 lr 0.000315 time 0.3090 (0.2859) loss 3.8860 (3.3441) grad_norm 1.9224 (nan) [2021-04-16 09:32:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][860/1251] eta 0:01:51 lr 0.000315 time 0.2869 (0.2858) loss 3.6477 (3.3453) grad_norm 1.7826 (nan) [2021-04-16 09:32:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][870/1251] eta 0:01:48 lr 0.000315 time 0.2597 (0.2857) loss 3.4082 (3.3470) grad_norm 1.8018 (nan) [2021-04-16 09:32:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][880/1251] eta 0:01:45 lr 0.000315 time 0.2757 (0.2856) loss 3.6862 (3.3510) grad_norm 1.8927 (nan) [2021-04-16 09:32:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][890/1251] eta 0:01:43 lr 0.000315 time 0.2869 (0.2855) loss 3.0970 (3.3465) grad_norm 1.8691 (nan) [2021-04-16 09:32:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][900/1251] eta 0:01:40 lr 0.000315 time 0.2824 (0.2854) loss 3.6182 (3.3493) grad_norm 1.6282 (nan) [2021-04-16 09:32:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][910/1251] eta 0:01:37 lr 0.000314 time 0.2916 (0.2854) loss 3.4184 (3.3511) grad_norm 1.9791 (nan) [2021-04-16 09:32:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][920/1251] eta 0:01:34 lr 0.000314 time 0.3069 (0.2854) loss 2.2212 (3.3482) grad_norm 1.7118 (nan) [2021-04-16 09:32:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][930/1251] eta 0:01:31 lr 0.000314 time 0.2719 (0.2855) loss 2.8204 (3.3493) grad_norm 1.8100 (nan) [2021-04-16 09:32:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][940/1251] eta 0:01:28 lr 0.000314 time 0.2738 (0.2857) loss 3.3699 (3.3496) grad_norm 1.7428 (nan) [2021-04-16 09:32:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][950/1251] eta 0:01:25 lr 0.000314 time 0.2835 (0.2856) loss 2.4580 (3.3496) grad_norm 1.9164 (nan) [2021-04-16 09:32:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][960/1251] eta 0:01:23 lr 0.000314 time 0.2661 (0.2855) loss 3.6759 (3.3509) grad_norm 2.0910 (nan) [2021-04-16 09:32:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][970/1251] eta 0:01:20 lr 0.000314 time 0.2974 (0.2855) loss 3.7469 (3.3536) grad_norm 2.1012 (nan) [2021-04-16 09:32:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][980/1251] eta 0:01:17 lr 0.000314 time 0.2812 (0.2854) loss 3.0812 (3.3552) grad_norm 1.7310 (nan) [2021-04-16 09:32:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][990/1251] eta 0:01:14 lr 0.000314 time 0.2974 (0.2854) loss 3.7560 (3.3585) grad_norm 1.8423 (nan) [2021-04-16 09:33:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][1000/1251] eta 0:01:11 lr 0.000314 time 0.2954 (0.2853) loss 3.4827 (3.3580) grad_norm 1.6208 (nan) [2021-04-16 09:33:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][1010/1251] eta 0:01:08 lr 0.000314 time 0.3005 (0.2853) loss 3.7040 (3.3590) grad_norm 2.2852 (nan) [2021-04-16 09:33:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][1020/1251] eta 0:01:05 lr 0.000314 time 0.2740 (0.2852) loss 3.5935 (3.3569) grad_norm 1.7291 (nan) [2021-04-16 09:33:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][1030/1251] eta 0:01:03 lr 0.000314 time 0.2763 (0.2851) loss 3.5707 (3.3583) grad_norm 2.0678 (nan) [2021-04-16 09:33:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][1040/1251] eta 0:01:00 lr 0.000314 time 0.3014 (0.2850) loss 3.5171 (3.3585) grad_norm 1.6491 (nan) [2021-04-16 09:33:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][1050/1251] eta 0:00:57 lr 0.000314 time 0.2634 (0.2849) loss 3.6122 (3.3600) grad_norm 2.0976 (nan) [2021-04-16 09:33:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][1060/1251] eta 0:00:54 lr 0.000314 time 0.2599 (0.2848) loss 2.9531 (3.3575) grad_norm 1.9504 (nan) [2021-04-16 09:33:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][1070/1251] eta 0:00:51 lr 0.000314 time 0.2789 (0.2848) loss 3.3776 (3.3568) grad_norm 1.7998 (nan) [2021-04-16 09:33:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][1080/1251] eta 0:00:48 lr 0.000314 time 0.2742 (0.2848) loss 3.6727 (3.3566) grad_norm 1.8035 (nan) [2021-04-16 09:33:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][1090/1251] eta 0:00:45 lr 0.000314 time 0.2679 (0.2848) loss 3.4141 (3.3572) grad_norm 1.7300 (nan) [2021-04-16 09:33:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][1100/1251] eta 0:00:43 lr 0.000314 time 0.3059 (0.2848) loss 3.9225 (3.3574) grad_norm 1.6807 (nan) [2021-04-16 09:33:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][1110/1251] eta 0:00:40 lr 0.000314 time 0.2793 (0.2847) loss 4.2396 (3.3573) grad_norm 1.8486 (nan) [2021-04-16 09:33:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][1120/1251] eta 0:00:37 lr 0.000314 time 0.2563 (0.2846) loss 3.9829 (3.3561) grad_norm 2.0350 (nan) [2021-04-16 09:33:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][1130/1251] eta 0:00:34 lr 0.000314 time 0.2804 (0.2846) loss 4.3112 (3.3550) grad_norm 1.8518 (nan) [2021-04-16 09:33:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][1140/1251] eta 0:00:31 lr 0.000314 time 0.2706 (0.2845) loss 3.5384 (3.3554) grad_norm 2.0654 (nan) [2021-04-16 09:33:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][1150/1251] eta 0:00:28 lr 0.000314 time 0.2718 (0.2846) loss 3.1393 (3.3549) grad_norm 1.8456 (nan) [2021-04-16 09:33:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][1160/1251] eta 0:00:25 lr 0.000314 time 0.2763 (0.2846) loss 3.5416 (3.3560) grad_norm 1.7296 (nan) [2021-04-16 09:33:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][1170/1251] eta 0:00:23 lr 0.000313 time 0.2887 (0.2846) loss 2.6958 (3.3565) grad_norm 1.6440 (nan) [2021-04-16 09:33:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][1180/1251] eta 0:00:20 lr 0.000313 time 0.2644 (0.2845) loss 3.4650 (3.3535) grad_norm 1.6602 (nan) [2021-04-16 09:33:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][1190/1251] eta 0:00:17 lr 0.000313 time 0.2902 (0.2845) loss 2.9769 (3.3532) grad_norm 1.9731 (nan) [2021-04-16 09:33:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][1200/1251] eta 0:00:14 lr 0.000313 time 0.2996 (0.2845) loss 4.1381 (3.3538) grad_norm 1.7445 (nan) [2021-04-16 09:34:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][1210/1251] eta 0:00:11 lr 0.000313 time 0.2672 (0.2845) loss 2.1617 (3.3527) grad_norm 1.9098 (nan) [2021-04-16 09:34:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][1220/1251] eta 0:00:08 lr 0.000313 time 0.2837 (0.2845) loss 3.3691 (3.3540) grad_norm 1.7374 (nan) [2021-04-16 09:34:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][1230/1251] eta 0:00:05 lr 0.000313 time 0.2797 (0.2844) loss 3.9680 (3.3557) grad_norm 2.0960 (nan) [2021-04-16 09:34:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][1240/1251] eta 0:00:03 lr 0.000313 time 0.2488 (0.2843) loss 3.3760 (3.3552) grad_norm 1.8028 (nan) [2021-04-16 09:34:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [187/300][1250/1251] eta 0:00:00 lr 0.000313 time 0.2495 (0.2840) loss 2.7710 (3.3525) grad_norm 1.9276 (nan) [2021-04-16 09:34:14 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 187 training takes 0:05:58 [2021-04-16 09:34:14 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_187.pth saving...... [2021-04-16 09:34:26 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_187.pth saved !!! [2021-04-16 09:34:27 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.153 (1.153) Loss 0.8894 (0.8894) Acc@1 79.004 (79.004) Acc@5 94.434 (94.434) [2021-04-16 09:34:28 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.395 (0.266) Loss 0.8112 (0.9073) Acc@1 79.199 (77.992) Acc@5 95.410 (94.611) [2021-04-16 09:34:31 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.112 (0.242) Loss 0.9342 (0.9274) Acc@1 77.344 (77.660) Acc@5 94.141 (94.252) [2021-04-16 09:34:33 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.347 (0.243) Loss 0.9719 (0.9333) Acc@1 77.734 (77.567) Acc@5 93.848 (94.163) [2021-04-16 09:34:35 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.221) Loss 0.9347 (0.9376) Acc@1 75.781 (77.470) Acc@5 94.238 (94.052) [2021-04-16 09:34:42 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 77.544 Acc@5 94.074 [2021-04-16 09:34:42 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 77.5% [2021-04-16 09:34:42 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 77.81% [2021-04-16 09:34:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][0/1251] eta 2:05:04 lr 0.000313 time 5.9992 (5.9992) loss 3.9966 (3.9966) grad_norm 2.2450 (2.2450) [2021-04-16 09:34:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][10/1251] eta 0:16:38 lr 0.000313 time 0.4102 (0.8045) loss 3.3767 (3.4309) grad_norm 1.9588 (1.9667) [2021-04-16 09:34:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][20/1251] eta 0:11:19 lr 0.000313 time 0.2758 (0.5520) loss 2.8687 (3.3772) grad_norm 1.6789 (1.9214) [2021-04-16 09:34:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][30/1251] eta 0:09:27 lr 0.000313 time 0.2816 (0.4645) loss 4.1073 (3.4705) grad_norm 1.7488 (1.8778) [2021-04-16 09:34:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][40/1251] eta 0:08:29 lr 0.000313 time 0.2746 (0.4205) loss 3.4278 (3.3963) grad_norm 1.5999 (1.8641) [2021-04-16 09:35:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][50/1251] eta 0:07:52 lr 0.000313 time 0.2938 (0.3936) loss 2.9569 (3.3987) grad_norm 1.8189 (inf) [2021-04-16 09:35:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][60/1251] eta 0:07:26 lr 0.000313 time 0.2952 (0.3750) loss 3.4638 (3.3593) grad_norm 1.8388 (inf) [2021-04-16 09:35:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][70/1251] eta 0:07:07 lr 0.000313 time 0.2743 (0.3620) loss 3.7758 (3.3853) grad_norm 1.5428 (inf) [2021-04-16 09:35:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][80/1251] eta 0:06:51 lr 0.000313 time 0.2914 (0.3517) loss 3.3820 (3.4134) grad_norm 1.9306 (inf) [2021-04-16 09:35:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][90/1251] eta 0:06:39 lr 0.000313 time 0.2714 (0.3439) loss 3.4635 (3.4237) grad_norm 1.8339 (inf) [2021-04-16 09:35:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][100/1251] eta 0:06:29 lr 0.000313 time 0.2600 (0.3380) loss 2.9624 (3.3676) grad_norm 1.6005 (inf) [2021-04-16 09:35:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][110/1251] eta 0:06:20 lr 0.000313 time 0.2872 (0.3334) loss 3.5664 (3.3723) grad_norm 1.7506 (inf) [2021-04-16 09:35:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][120/1251] eta 0:06:12 lr 0.000313 time 0.2894 (0.3292) loss 4.2832 (3.3585) grad_norm 1.8111 (inf) [2021-04-16 09:35:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][130/1251] eta 0:06:07 lr 0.000313 time 0.2824 (0.3275) loss 2.8838 (3.3729) grad_norm 1.9330 (inf) [2021-04-16 09:35:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][140/1251] eta 0:06:00 lr 0.000313 time 0.2922 (0.3247) loss 3.6578 (3.3832) grad_norm 1.9531 (inf) [2021-04-16 09:35:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][150/1251] eta 0:05:55 lr 0.000313 time 0.2677 (0.3227) loss 3.5791 (3.3655) grad_norm 2.0281 (inf) [2021-04-16 09:35:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][160/1251] eta 0:05:48 lr 0.000313 time 0.2584 (0.3198) loss 2.1674 (3.3670) grad_norm 1.7487 (inf) [2021-04-16 09:35:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][170/1251] eta 0:05:43 lr 0.000313 time 0.2837 (0.3174) loss 3.5570 (3.3576) grad_norm 1.6644 (inf) [2021-04-16 09:35:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][180/1251] eta 0:05:37 lr 0.000312 time 0.3051 (0.3153) loss 3.4638 (3.3672) grad_norm 1.7671 (inf) [2021-04-16 09:35:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][190/1251] eta 0:05:32 lr 0.000312 time 0.2673 (0.3132) loss 3.7100 (3.3752) grad_norm 1.6526 (inf) [2021-04-16 09:35:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][200/1251] eta 0:05:27 lr 0.000312 time 0.2817 (0.3119) loss 2.8911 (3.3821) grad_norm 1.9568 (inf) [2021-04-16 09:35:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][210/1251] eta 0:05:22 lr 0.000312 time 0.2806 (0.3102) loss 3.3669 (3.3958) grad_norm 1.6835 (inf) [2021-04-16 09:35:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][220/1251] eta 0:05:18 lr 0.000312 time 0.2745 (0.3090) loss 3.6578 (3.3895) grad_norm 1.8414 (inf) [2021-04-16 09:35:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][230/1251] eta 0:05:14 lr 0.000312 time 0.2748 (0.3077) loss 3.9545 (3.3938) grad_norm 1.6257 (inf) [2021-04-16 09:35:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][240/1251] eta 0:05:09 lr 0.000312 time 0.2782 (0.3064) loss 3.9831 (3.3892) grad_norm 1.8454 (inf) [2021-04-16 09:35:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][250/1251] eta 0:05:05 lr 0.000312 time 0.2941 (0.3054) loss 3.4624 (3.4037) grad_norm 1.7928 (inf) [2021-04-16 09:36:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][260/1251] eta 0:05:01 lr 0.000312 time 0.2699 (0.3043) loss 2.3791 (3.3975) grad_norm 1.6885 (inf) [2021-04-16 09:36:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][270/1251] eta 0:04:57 lr 0.000312 time 0.2896 (0.3036) loss 4.3170 (3.4038) grad_norm 1.7355 (inf) [2021-04-16 09:36:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][280/1251] eta 0:04:53 lr 0.000312 time 0.2927 (0.3027) loss 3.3765 (3.3930) grad_norm 2.7679 (inf) [2021-04-16 09:36:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][290/1251] eta 0:04:50 lr 0.000312 time 0.2922 (0.3021) loss 2.8269 (3.3795) grad_norm 1.7088 (inf) [2021-04-16 09:36:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][300/1251] eta 0:04:46 lr 0.000312 time 0.2599 (0.3016) loss 3.5134 (3.3850) grad_norm 2.0743 (inf) [2021-04-16 09:36:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][310/1251] eta 0:04:43 lr 0.000312 time 0.3043 (0.3009) loss 2.6139 (3.3874) grad_norm 2.1981 (inf) [2021-04-16 09:36:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][320/1251] eta 0:04:39 lr 0.000312 time 0.2625 (0.3002) loss 2.6731 (3.3959) grad_norm 1.9829 (inf) [2021-04-16 09:36:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][330/1251] eta 0:04:35 lr 0.000312 time 0.2728 (0.2995) loss 2.7994 (3.3960) grad_norm 2.2677 (inf) [2021-04-16 09:36:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][340/1251] eta 0:04:32 lr 0.000312 time 0.2853 (0.2990) loss 4.2857 (3.3966) grad_norm 1.8604 (inf) [2021-04-16 09:36:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][350/1251] eta 0:04:28 lr 0.000312 time 0.2789 (0.2985) loss 3.2532 (3.3994) grad_norm 2.0413 (inf) [2021-04-16 09:36:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][360/1251] eta 0:04:26 lr 0.000312 time 0.2749 (0.2988) loss 2.9226 (3.3891) grad_norm 1.9217 (inf) [2021-04-16 09:36:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][370/1251] eta 0:04:23 lr 0.000312 time 0.2764 (0.2986) loss 2.9743 (3.3858) grad_norm 1.6812 (inf) [2021-04-16 09:36:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][380/1251] eta 0:04:19 lr 0.000312 time 0.2673 (0.2980) loss 3.4473 (3.3762) grad_norm 1.8084 (inf) [2021-04-16 09:36:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][390/1251] eta 0:04:16 lr 0.000312 time 0.2794 (0.2976) loss 2.8529 (3.3761) grad_norm 1.6550 (inf) [2021-04-16 09:36:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][400/1251] eta 0:04:12 lr 0.000312 time 0.2918 (0.2971) loss 3.4887 (3.3813) grad_norm 2.0585 (inf) [2021-04-16 09:36:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][410/1251] eta 0:04:09 lr 0.000312 time 0.2892 (0.2969) loss 2.9125 (3.3859) grad_norm 1.8048 (inf) [2021-04-16 09:36:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][420/1251] eta 0:04:06 lr 0.000312 time 0.3152 (0.2965) loss 2.6247 (3.3886) grad_norm 1.9731 (inf) [2021-04-16 09:36:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][430/1251] eta 0:04:03 lr 0.000312 time 0.2997 (0.2962) loss 3.1151 (3.3886) grad_norm 2.0773 (inf) [2021-04-16 09:36:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][440/1251] eta 0:03:59 lr 0.000312 time 0.2725 (0.2957) loss 3.8602 (3.3862) grad_norm 1.9203 (inf) [2021-04-16 09:36:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][450/1251] eta 0:03:56 lr 0.000311 time 0.2847 (0.2953) loss 3.2882 (3.3872) grad_norm 1.7331 (inf) [2021-04-16 09:36:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][460/1251] eta 0:03:53 lr 0.000311 time 0.2635 (0.2949) loss 3.2887 (3.3860) grad_norm 1.8085 (inf) [2021-04-16 09:37:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][470/1251] eta 0:03:50 lr 0.000311 time 0.2722 (0.2945) loss 3.7944 (3.3835) grad_norm 2.0574 (inf) [2021-04-16 09:37:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][480/1251] eta 0:03:46 lr 0.000311 time 0.2827 (0.2941) loss 3.5305 (3.3796) grad_norm 1.6805 (inf) [2021-04-16 09:37:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][490/1251] eta 0:03:43 lr 0.000311 time 0.2822 (0.2939) loss 3.0399 (3.3792) grad_norm 1.6625 (inf) [2021-04-16 09:37:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][500/1251] eta 0:03:40 lr 0.000311 time 0.2766 (0.2935) loss 3.5725 (3.3838) grad_norm 1.8379 (inf) [2021-04-16 09:37:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][510/1251] eta 0:03:37 lr 0.000311 time 0.2750 (0.2931) loss 3.9493 (3.3876) grad_norm 1.7279 (inf) [2021-04-16 09:37:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][520/1251] eta 0:03:34 lr 0.000311 time 0.2923 (0.2928) loss 2.6970 (3.3887) grad_norm 1.7853 (inf) [2021-04-16 09:37:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][530/1251] eta 0:03:30 lr 0.000311 time 0.3159 (0.2924) loss 4.0743 (3.3856) grad_norm 1.9854 (inf) [2021-04-16 09:37:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][540/1251] eta 0:03:27 lr 0.000311 time 0.2721 (0.2922) loss 3.3815 (3.3859) grad_norm 1.6994 (inf) [2021-04-16 09:37:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][550/1251] eta 0:03:24 lr 0.000311 time 0.2788 (0.2918) loss 3.6319 (3.3886) grad_norm 1.7153 (inf) [2021-04-16 09:37:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][560/1251] eta 0:03:21 lr 0.000311 time 0.2683 (0.2916) loss 3.1972 (3.3854) grad_norm 1.4912 (inf) [2021-04-16 09:37:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][570/1251] eta 0:03:18 lr 0.000311 time 0.2929 (0.2918) loss 3.6041 (3.3856) grad_norm 1.6047 (inf) [2021-04-16 09:37:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][580/1251] eta 0:03:15 lr 0.000311 time 0.3093 (0.2918) loss 4.1368 (3.3899) grad_norm 2.0377 (inf) [2021-04-16 09:37:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][590/1251] eta 0:03:12 lr 0.000311 time 0.2911 (0.2918) loss 3.5402 (3.3900) grad_norm 1.8469 (inf) [2021-04-16 09:37:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][600/1251] eta 0:03:09 lr 0.000311 time 0.2523 (0.2915) loss 3.3946 (3.3870) grad_norm 1.9214 (inf) [2021-04-16 09:37:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][610/1251] eta 0:03:06 lr 0.000311 time 0.2795 (0.2913) loss 3.6819 (3.3888) grad_norm 1.7007 (inf) [2021-04-16 09:37:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][620/1251] eta 0:03:03 lr 0.000311 time 0.2660 (0.2911) loss 3.8688 (3.3922) grad_norm 1.5089 (inf) [2021-04-16 09:37:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][630/1251] eta 0:03:00 lr 0.000311 time 0.2652 (0.2907) loss 2.3027 (3.3928) grad_norm 1.8653 (inf) [2021-04-16 09:37:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][640/1251] eta 0:02:57 lr 0.000311 time 0.2693 (0.2906) loss 3.2388 (3.3933) grad_norm 2.1998 (inf) [2021-04-16 09:37:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][650/1251] eta 0:02:54 lr 0.000311 time 0.2882 (0.2904) loss 3.8300 (3.3942) grad_norm 1.8141 (inf) [2021-04-16 09:37:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][660/1251] eta 0:02:51 lr 0.000311 time 0.2826 (0.2903) loss 3.5516 (3.3924) grad_norm 1.6353 (inf) [2021-04-16 09:37:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][670/1251] eta 0:02:48 lr 0.000311 time 0.2822 (0.2901) loss 4.0396 (3.3953) grad_norm 1.9229 (inf) [2021-04-16 09:38:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][680/1251] eta 0:02:45 lr 0.000311 time 0.2881 (0.2901) loss 3.4268 (3.3954) grad_norm 2.0622 (inf) [2021-04-16 09:38:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][690/1251] eta 0:02:42 lr 0.000311 time 0.2875 (0.2899) loss 3.1961 (3.3956) grad_norm 1.9078 (inf) [2021-04-16 09:38:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][700/1251] eta 0:02:39 lr 0.000311 time 0.2849 (0.2897) loss 2.8158 (3.3957) grad_norm 2.0149 (inf) [2021-04-16 09:38:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][710/1251] eta 0:02:36 lr 0.000310 time 0.2577 (0.2898) loss 2.7757 (3.3922) grad_norm 1.9481 (inf) [2021-04-16 09:38:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][720/1251] eta 0:02:33 lr 0.000310 time 0.2876 (0.2899) loss 3.3797 (3.3927) grad_norm 2.0754 (inf) [2021-04-16 09:38:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][730/1251] eta 0:02:30 lr 0.000310 time 0.2886 (0.2898) loss 2.6698 (3.3886) grad_norm 1.9986 (inf) [2021-04-16 09:38:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][740/1251] eta 0:02:28 lr 0.000310 time 0.3026 (0.2896) loss 3.4338 (3.3881) grad_norm 1.7452 (inf) [2021-04-16 09:38:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][750/1251] eta 0:02:25 lr 0.000310 time 0.2700 (0.2895) loss 2.4611 (3.3886) grad_norm 1.8156 (inf) [2021-04-16 09:38:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][760/1251] eta 0:02:22 lr 0.000310 time 0.3113 (0.2894) loss 3.2375 (3.3859) grad_norm 2.0982 (inf) [2021-04-16 09:38:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][770/1251] eta 0:02:19 lr 0.000310 time 0.2577 (0.2893) loss 2.5057 (3.3882) grad_norm 1.8277 (inf) [2021-04-16 09:38:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][780/1251] eta 0:02:16 lr 0.000310 time 0.2731 (0.2892) loss 4.2163 (3.3904) grad_norm 1.7471 (inf) [2021-04-16 09:38:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][790/1251] eta 0:02:13 lr 0.000310 time 0.2689 (0.2891) loss 3.6307 (3.3895) grad_norm 1.9654 (inf) [2021-04-16 09:38:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][800/1251] eta 0:02:10 lr 0.000310 time 0.2746 (0.2889) loss 3.6605 (3.3899) grad_norm 1.9810 (inf) [2021-04-16 09:38:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][810/1251] eta 0:02:07 lr 0.000310 time 0.2712 (0.2888) loss 3.3392 (3.3886) grad_norm 1.7706 (inf) [2021-04-16 09:38:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][820/1251] eta 0:02:04 lr 0.000310 time 0.2843 (0.2887) loss 4.0000 (3.3901) grad_norm 2.4977 (inf) [2021-04-16 09:38:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][830/1251] eta 0:02:01 lr 0.000310 time 0.2727 (0.2885) loss 3.5850 (3.3916) grad_norm 1.7466 (inf) [2021-04-16 09:38:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][840/1251] eta 0:01:58 lr 0.000310 time 0.2584 (0.2884) loss 3.8443 (3.3927) grad_norm 1.7393 (inf) [2021-04-16 09:38:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][850/1251] eta 0:01:55 lr 0.000310 time 0.2667 (0.2883) loss 4.3758 (3.3974) grad_norm 1.7529 (inf) [2021-04-16 09:38:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][860/1251] eta 0:01:52 lr 0.000310 time 0.2594 (0.2881) loss 3.8531 (3.3966) grad_norm 1.6983 (inf) [2021-04-16 09:38:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][870/1251] eta 0:01:49 lr 0.000310 time 0.2941 (0.2882) loss 3.2949 (3.3947) grad_norm 1.6696 (inf) [2021-04-16 09:38:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][880/1251] eta 0:01:46 lr 0.000310 time 0.2955 (0.2882) loss 3.1751 (3.3930) grad_norm 2.0590 (inf) [2021-04-16 09:38:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][890/1251] eta 0:01:43 lr 0.000310 time 0.2629 (0.2880) loss 3.5987 (3.3939) grad_norm 1.6414 (inf) [2021-04-16 09:39:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][900/1251] eta 0:01:41 lr 0.000310 time 0.2847 (0.2879) loss 2.6033 (3.3928) grad_norm 1.8130 (inf) [2021-04-16 09:39:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][910/1251] eta 0:01:38 lr 0.000310 time 0.2898 (0.2878) loss 3.5453 (3.3924) grad_norm 1.9917 (inf) [2021-04-16 09:39:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][920/1251] eta 0:01:35 lr 0.000310 time 0.2674 (0.2877) loss 3.1005 (3.3933) grad_norm 1.9153 (inf) [2021-04-16 09:39:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][930/1251] eta 0:01:32 lr 0.000310 time 0.2810 (0.2876) loss 4.3954 (3.3951) grad_norm 1.8054 (inf) [2021-04-16 09:39:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][940/1251] eta 0:01:29 lr 0.000310 time 0.2835 (0.2878) loss 3.4669 (3.3992) grad_norm 1.6855 (inf) [2021-04-16 09:39:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][950/1251] eta 0:01:26 lr 0.000310 time 0.2647 (0.2879) loss 3.6178 (3.3989) grad_norm 1.8671 (inf) [2021-04-16 09:39:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][960/1251] eta 0:01:23 lr 0.000310 time 0.2834 (0.2877) loss 3.8207 (3.3974) grad_norm 1.8819 (inf) [2021-04-16 09:39:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][970/1251] eta 0:01:20 lr 0.000309 time 0.3000 (0.2876) loss 3.8969 (3.3978) grad_norm 2.3681 (inf) [2021-04-16 09:39:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][980/1251] eta 0:01:17 lr 0.000309 time 0.2794 (0.2876) loss 3.5169 (3.3958) grad_norm 1.7408 (inf) [2021-04-16 09:39:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][990/1251] eta 0:01:15 lr 0.000309 time 0.2879 (0.2874) loss 2.8694 (3.3936) grad_norm 1.6159 (inf) [2021-04-16 09:39:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][1000/1251] eta 0:01:12 lr 0.000309 time 0.2670 (0.2873) loss 3.7831 (3.3971) grad_norm 1.7506 (inf) [2021-04-16 09:39:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][1010/1251] eta 0:01:09 lr 0.000309 time 0.2939 (0.2872) loss 2.6384 (3.3970) grad_norm 1.5889 (inf) [2021-04-16 09:39:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][1020/1251] eta 0:01:06 lr 0.000309 time 0.2808 (0.2871) loss 2.6740 (3.3980) grad_norm 1.6223 (inf) [2021-04-16 09:39:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][1030/1251] eta 0:01:03 lr 0.000309 time 0.2622 (0.2870) loss 3.5571 (3.3978) grad_norm 1.9214 (inf) [2021-04-16 09:39:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][1040/1251] eta 0:01:00 lr 0.000309 time 0.2773 (0.2869) loss 3.7122 (3.3982) grad_norm 1.5267 (inf) [2021-04-16 09:39:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][1050/1251] eta 0:00:57 lr 0.000309 time 0.2732 (0.2868) loss 4.0243 (3.3972) grad_norm 1.8456 (inf) [2021-04-16 09:39:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][1060/1251] eta 0:00:54 lr 0.000309 time 0.2822 (0.2867) loss 2.1536 (3.3949) grad_norm 1.8759 (inf) [2021-04-16 09:39:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][1070/1251] eta 0:00:51 lr 0.000309 time 0.2893 (0.2866) loss 2.8641 (3.3950) grad_norm 1.8227 (inf) [2021-04-16 09:39:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][1080/1251] eta 0:00:48 lr 0.000309 time 0.2796 (0.2865) loss 3.8955 (3.3956) grad_norm 1.6452 (inf) [2021-04-16 09:39:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][1090/1251] eta 0:00:46 lr 0.000309 time 0.2807 (0.2865) loss 3.4360 (3.3945) grad_norm 1.6441 (inf) [2021-04-16 09:39:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][1100/1251] eta 0:00:43 lr 0.000309 time 0.2892 (0.2863) loss 2.3982 (3.3913) grad_norm 1.8415 (inf) [2021-04-16 09:40:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][1110/1251] eta 0:00:40 lr 0.000309 time 0.2929 (0.2863) loss 4.0465 (3.3913) grad_norm 1.7772 (inf) [2021-04-16 09:40:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][1120/1251] eta 0:00:37 lr 0.000309 time 0.2476 (0.2861) loss 3.1337 (3.3888) grad_norm 1.8224 (inf) [2021-04-16 09:40:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][1130/1251] eta 0:00:34 lr 0.000309 time 0.3088 (0.2861) loss 3.4497 (3.3867) grad_norm 1.6991 (inf) [2021-04-16 09:40:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][1140/1251] eta 0:00:31 lr 0.000309 time 0.2712 (0.2861) loss 3.0963 (3.3870) grad_norm 1.7865 (inf) [2021-04-16 09:40:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][1150/1251] eta 0:00:28 lr 0.000309 time 0.2452 (0.2862) loss 3.7085 (3.3879) grad_norm 1.8612 (inf) [2021-04-16 09:40:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][1160/1251] eta 0:00:26 lr 0.000309 time 0.2926 (0.2861) loss 3.0899 (3.3858) grad_norm 1.7036 (inf) [2021-04-16 09:40:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][1170/1251] eta 0:00:23 lr 0.000309 time 0.2892 (0.2863) loss 3.8250 (3.3864) grad_norm 1.7888 (inf) [2021-04-16 09:40:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][1180/1251] eta 0:00:20 lr 0.000309 time 0.3040 (0.2862) loss 3.4616 (3.3848) grad_norm 1.7078 (inf) [2021-04-16 09:40:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][1190/1251] eta 0:00:17 lr 0.000309 time 0.2685 (0.2862) loss 3.6709 (3.3879) grad_norm 1.8773 (inf) [2021-04-16 09:40:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][1200/1251] eta 0:00:14 lr 0.000309 time 0.2909 (0.2861) loss 3.5315 (3.3874) grad_norm 1.7928 (inf) [2021-04-16 09:40:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][1210/1251] eta 0:00:11 lr 0.000309 time 0.2845 (0.2860) loss 2.6582 (3.3855) grad_norm 1.8286 (inf) [2021-04-16 09:40:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][1220/1251] eta 0:00:08 lr 0.000309 time 0.2884 (0.2860) loss 2.7812 (3.3867) grad_norm 1.7780 (inf) [2021-04-16 09:40:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][1230/1251] eta 0:00:06 lr 0.000308 time 0.2710 (0.2859) loss 2.9464 (3.3887) grad_norm 1.9808 (inf) [2021-04-16 09:40:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][1240/1251] eta 0:00:03 lr 0.000308 time 0.2486 (0.2857) loss 3.6377 (3.3893) grad_norm 1.8770 (inf) [2021-04-16 09:40:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [188/300][1250/1251] eta 0:00:00 lr 0.000308 time 0.2487 (0.2854) loss 3.7118 (3.3907) grad_norm 1.9246 (inf) [2021-04-16 09:40:43 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 188 training takes 0:06:01 [2021-04-16 09:40:43 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_188.pth saving...... [2021-04-16 09:41:00 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_188.pth saved !!! [2021-04-16 09:41:01 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.090 (1.090) Loss 0.9807 (0.9807) Acc@1 77.441 (77.441) Acc@5 94.238 (94.238) [2021-04-16 09:41:03 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.419 (0.231) Loss 0.9241 (0.9379) Acc@1 78.125 (77.850) Acc@5 93.555 (94.185) [2021-04-16 09:41:05 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.117 (0.240) Loss 0.9515 (0.9480) Acc@1 75.879 (77.372) Acc@5 94.629 (94.052) [2021-04-16 09:41:08 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.100 (0.240) Loss 0.8987 (0.9470) Acc@1 78.516 (77.360) Acc@5 95.215 (94.137) [2021-04-16 09:41:09 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.213) Loss 0.8640 (0.9420) Acc@1 79.590 (77.577) Acc@5 94.727 (94.238) [2021-04-16 09:41:19 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 77.620 Acc@5 94.238 [2021-04-16 09:41:19 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 77.6% [2021-04-16 09:41:19 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 77.81% [2021-04-16 09:41:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][0/1251] eta 2:22:23 lr 0.000308 time 6.8293 (6.8293) loss 2.8112 (2.8112) grad_norm 1.7033 (1.7033) [2021-04-16 09:41:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][10/1251] eta 0:17:57 lr 0.000308 time 0.2625 (0.8685) loss 3.3115 (3.3534) grad_norm 2.2533 (1.8670) [2021-04-16 09:41:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][20/1251] eta 0:12:01 lr 0.000308 time 0.2778 (0.5864) loss 2.8042 (3.2736) grad_norm 1.8345 (1.8275) [2021-04-16 09:41:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][30/1251] eta 0:09:56 lr 0.000308 time 0.3245 (0.4882) loss 3.6640 (3.3319) grad_norm 1.7512 (1.8192) [2021-04-16 09:41:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][40/1251] eta 0:08:48 lr 0.000308 time 0.2630 (0.4367) loss 3.4811 (3.3776) grad_norm 2.0323 (1.8525) [2021-04-16 09:41:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][50/1251] eta 0:08:06 lr 0.000308 time 0.2952 (0.4048) loss 2.7451 (3.4160) grad_norm 2.1934 (1.8645) [2021-04-16 09:41:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][60/1251] eta 0:07:39 lr 0.000308 time 0.4169 (0.3857) loss 3.1106 (3.3455) grad_norm 1.8945 (1.8749) [2021-04-16 09:41:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][70/1251] eta 0:07:16 lr 0.000308 time 0.2927 (0.3695) loss 3.4135 (3.3076) grad_norm 1.6177 (1.8625) [2021-04-16 09:41:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][80/1251] eta 0:07:01 lr 0.000308 time 0.2600 (0.3596) loss 3.6024 (3.3337) grad_norm 1.7717 (1.8506) [2021-04-16 09:41:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][90/1251] eta 0:06:47 lr 0.000308 time 0.2932 (0.3507) loss 3.6860 (3.3398) grad_norm 1.5782 (1.8427) [2021-04-16 09:41:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][100/1251] eta 0:06:35 lr 0.000308 time 0.2932 (0.3435) loss 3.2544 (3.3488) grad_norm 1.7412 (1.8407) [2021-04-16 09:41:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][110/1251] eta 0:06:25 lr 0.000308 time 0.2651 (0.3375) loss 3.8849 (3.3537) grad_norm 1.8048 (1.8477) [2021-04-16 09:41:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][120/1251] eta 0:06:16 lr 0.000308 time 0.2680 (0.3330) loss 2.1389 (3.3437) grad_norm 1.9975 (1.8533) [2021-04-16 09:42:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][130/1251] eta 0:06:08 lr 0.000308 time 0.2799 (0.3288) loss 3.6670 (3.3371) grad_norm 1.8454 (1.8565) [2021-04-16 09:42:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][140/1251] eta 0:06:01 lr 0.000308 time 0.2852 (0.3251) loss 3.6523 (3.3382) grad_norm 1.8857 (1.8619) [2021-04-16 09:42:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][150/1251] eta 0:05:56 lr 0.000308 time 0.2613 (0.3240) loss 4.4240 (3.3411) grad_norm 1.7829 (1.8653) [2021-04-16 09:42:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][160/1251] eta 0:05:51 lr 0.000308 time 0.3053 (0.3220) loss 4.1415 (3.3621) grad_norm 1.7667 (1.8607) [2021-04-16 09:42:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][170/1251] eta 0:05:45 lr 0.000308 time 0.2970 (0.3201) loss 3.7811 (3.3728) grad_norm 1.9180 (1.8542) [2021-04-16 09:42:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][180/1251] eta 0:05:40 lr 0.000308 time 0.2882 (0.3179) loss 3.3929 (3.3734) grad_norm 1.6849 (1.8533) [2021-04-16 09:42:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][190/1251] eta 0:05:35 lr 0.000308 time 0.2844 (0.3160) loss 2.3674 (3.3747) grad_norm 2.0486 (1.8605) [2021-04-16 09:42:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][200/1251] eta 0:05:30 lr 0.000308 time 0.2917 (0.3140) loss 3.8114 (3.3855) grad_norm 1.8162 (1.8658) [2021-04-16 09:42:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][210/1251] eta 0:05:25 lr 0.000308 time 0.2846 (0.3126) loss 3.3768 (3.3941) grad_norm 1.8191 (1.8649) [2021-04-16 09:42:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][220/1251] eta 0:05:20 lr 0.000308 time 0.2844 (0.3111) loss 3.6873 (3.3963) grad_norm 1.8371 (1.8739) [2021-04-16 09:42:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][230/1251] eta 0:05:16 lr 0.000308 time 0.2806 (0.3095) loss 2.4286 (3.3743) grad_norm 1.8957 (1.8756) [2021-04-16 09:42:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][240/1251] eta 0:05:11 lr 0.000307 time 0.2823 (0.3082) loss 3.4975 (3.3798) grad_norm 2.2797 (1.8793) [2021-04-16 09:42:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][250/1251] eta 0:05:07 lr 0.000307 time 0.2761 (0.3071) loss 3.6612 (3.3872) grad_norm 1.9987 (1.8814) [2021-04-16 09:42:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][260/1251] eta 0:05:03 lr 0.000307 time 0.2780 (0.3059) loss 3.5834 (3.3860) grad_norm 1.6576 (1.8807) [2021-04-16 09:42:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][270/1251] eta 0:04:59 lr 0.000307 time 0.2733 (0.3049) loss 4.1521 (3.3897) grad_norm 1.8521 (1.8839) [2021-04-16 09:42:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][280/1251] eta 0:04:54 lr 0.000307 time 0.2905 (0.3038) loss 3.6846 (3.3851) grad_norm 2.2271 (1.8844) [2021-04-16 09:42:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][290/1251] eta 0:04:50 lr 0.000307 time 0.2778 (0.3028) loss 3.0467 (3.3862) grad_norm 1.7403 (1.8810) [2021-04-16 09:42:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][300/1251] eta 0:04:47 lr 0.000307 time 0.2920 (0.3021) loss 2.6401 (3.3836) grad_norm 1.8419 (1.8800) [2021-04-16 09:42:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][310/1251] eta 0:04:43 lr 0.000307 time 0.2558 (0.3013) loss 2.4581 (3.3835) grad_norm 1.7076 (1.8788) [2021-04-16 09:42:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][320/1251] eta 0:04:39 lr 0.000307 time 0.2475 (0.3005) loss 3.5222 (3.3818) grad_norm 1.6483 (1.8772) [2021-04-16 09:42:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][330/1251] eta 0:04:36 lr 0.000307 time 0.2681 (0.2997) loss 3.4223 (3.3886) grad_norm 2.2370 (1.8803) [2021-04-16 09:43:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][340/1251] eta 0:04:32 lr 0.000307 time 0.2827 (0.2995) loss 3.2286 (3.3891) grad_norm 1.8819 (1.8819) [2021-04-16 09:43:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][350/1251] eta 0:04:29 lr 0.000307 time 0.2829 (0.2993) loss 3.5395 (3.3868) grad_norm 1.6451 (1.8831) [2021-04-16 09:43:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][360/1251] eta 0:04:26 lr 0.000307 time 0.2917 (0.2990) loss 4.2747 (3.3917) grad_norm 1.9820 (1.8840) [2021-04-16 09:43:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][370/1251] eta 0:04:23 lr 0.000307 time 0.2631 (0.2988) loss 4.0712 (3.3953) grad_norm 1.9655 (1.8833) [2021-04-16 09:43:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][380/1251] eta 0:04:19 lr 0.000307 time 0.2612 (0.2983) loss 3.6147 (3.3995) grad_norm 1.9589 (1.8875) [2021-04-16 09:43:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][390/1251] eta 0:04:16 lr 0.000307 time 0.3052 (0.2977) loss 3.7694 (3.4022) grad_norm 2.0097 (1.8891) [2021-04-16 09:43:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][400/1251] eta 0:04:12 lr 0.000307 time 0.2782 (0.2971) loss 3.8405 (3.4033) grad_norm 1.9923 (1.8884) [2021-04-16 09:43:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][410/1251] eta 0:04:09 lr 0.000307 time 0.2972 (0.2965) loss 2.1016 (3.4047) grad_norm 1.8094 (1.8892) [2021-04-16 09:43:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][420/1251] eta 0:04:06 lr 0.000307 time 0.2730 (0.2962) loss 3.1058 (3.4006) grad_norm 2.0137 (1.8900) [2021-04-16 09:43:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][430/1251] eta 0:04:02 lr 0.000307 time 0.2605 (0.2957) loss 3.6974 (3.4031) grad_norm 1.7081 (1.8889) [2021-04-16 09:43:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][440/1251] eta 0:03:59 lr 0.000307 time 0.2783 (0.2953) loss 2.4509 (3.4036) grad_norm 2.1270 (1.8913) [2021-04-16 09:43:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][450/1251] eta 0:03:56 lr 0.000307 time 0.2812 (0.2949) loss 3.6394 (3.4059) grad_norm 1.8062 (1.8906) [2021-04-16 09:43:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][460/1251] eta 0:03:52 lr 0.000307 time 0.2708 (0.2945) loss 3.6111 (3.4099) grad_norm 1.8561 (1.8898) [2021-04-16 09:43:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][470/1251] eta 0:03:49 lr 0.000307 time 0.2792 (0.2940) loss 3.4627 (3.4113) grad_norm 1.6574 (1.8894) [2021-04-16 09:43:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][480/1251] eta 0:03:46 lr 0.000307 time 0.2815 (0.2937) loss 2.8719 (3.4137) grad_norm 1.6292 (1.8911) [2021-04-16 09:43:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][490/1251] eta 0:03:43 lr 0.000307 time 0.2884 (0.2933) loss 3.1869 (3.4135) grad_norm 2.0649 (1.8902) [2021-04-16 09:43:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][500/1251] eta 0:03:39 lr 0.000307 time 0.2806 (0.2929) loss 3.3575 (3.4146) grad_norm 1.8618 (1.8912) [2021-04-16 09:43:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][510/1251] eta 0:03:36 lr 0.000306 time 0.2547 (0.2926) loss 2.8840 (3.4127) grad_norm 1.8832 (1.8900) [2021-04-16 09:43:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][520/1251] eta 0:03:33 lr 0.000306 time 0.2845 (0.2925) loss 2.3143 (3.4103) grad_norm 1.7304 (1.8888) [2021-04-16 09:43:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][530/1251] eta 0:03:30 lr 0.000306 time 0.2775 (0.2922) loss 2.4357 (3.4033) grad_norm 1.7687 (1.8873) [2021-04-16 09:43:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][540/1251] eta 0:03:27 lr 0.000306 time 0.2649 (0.2921) loss 3.8760 (3.4056) grad_norm 1.6692 (1.8880) [2021-04-16 09:43:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][550/1251] eta 0:03:24 lr 0.000306 time 0.2597 (0.2918) loss 3.3082 (3.4030) grad_norm 1.6437 (1.8875) [2021-04-16 09:44:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][560/1251] eta 0:03:21 lr 0.000306 time 0.2709 (0.2915) loss 2.4137 (3.4025) grad_norm 2.0763 (1.8875) [2021-04-16 09:44:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][570/1251] eta 0:03:18 lr 0.000306 time 0.2451 (0.2912) loss 3.5397 (3.4038) grad_norm 1.8848 (1.8867) [2021-04-16 09:44:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][580/1251] eta 0:03:15 lr 0.000306 time 0.2674 (0.2913) loss 2.5417 (3.4022) grad_norm 2.2290 (1.8870) [2021-04-16 09:44:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][590/1251] eta 0:03:12 lr 0.000306 time 0.2666 (0.2911) loss 3.7931 (3.4042) grad_norm 1.7804 (1.8872) [2021-04-16 09:44:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][600/1251] eta 0:03:09 lr 0.000306 time 0.2853 (0.2909) loss 3.6005 (3.4039) grad_norm 1.8910 (1.8865) [2021-04-16 09:44:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][610/1251] eta 0:03:06 lr 0.000306 time 0.2598 (0.2908) loss 3.9069 (3.3971) grad_norm 1.6589 (1.8879) [2021-04-16 09:44:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][620/1251] eta 0:03:03 lr 0.000306 time 0.2525 (0.2905) loss 3.3284 (3.3976) grad_norm 2.1702 (1.8881) [2021-04-16 09:44:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][630/1251] eta 0:03:00 lr 0.000306 time 0.2673 (0.2903) loss 3.1049 (3.3990) grad_norm 1.7065 (1.8881) [2021-04-16 09:44:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][640/1251] eta 0:02:57 lr 0.000306 time 0.2778 (0.2903) loss 3.0164 (3.3956) grad_norm 2.0133 (1.8890) [2021-04-16 09:44:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][650/1251] eta 0:02:54 lr 0.000306 time 0.3629 (0.2903) loss 3.5162 (3.3931) grad_norm 1.7490 (1.8876) [2021-04-16 09:44:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][660/1251] eta 0:02:51 lr 0.000306 time 0.2759 (0.2900) loss 3.6389 (3.3923) grad_norm 1.9902 (1.8866) [2021-04-16 09:44:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][670/1251] eta 0:02:48 lr 0.000306 time 0.2542 (0.2898) loss 3.3742 (3.3916) grad_norm 1.8763 (1.8857) [2021-04-16 09:44:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][680/1251] eta 0:02:45 lr 0.000306 time 0.2444 (0.2895) loss 2.9233 (3.3915) grad_norm 1.7803 (1.8848) [2021-04-16 09:44:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][690/1251] eta 0:02:42 lr 0.000306 time 0.2622 (0.2893) loss 3.4565 (3.3911) grad_norm 1.7342 (1.8863) [2021-04-16 09:44:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][700/1251] eta 0:02:39 lr 0.000306 time 0.2782 (0.2891) loss 2.2664 (3.3878) grad_norm 1.6320 (1.8856) [2021-04-16 09:44:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][710/1251] eta 0:02:36 lr 0.000306 time 0.2657 (0.2889) loss 3.1460 (3.3845) grad_norm 1.5516 (1.8845) [2021-04-16 09:44:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][720/1251] eta 0:02:33 lr 0.000306 time 0.2794 (0.2889) loss 4.1545 (3.3806) grad_norm 1.9992 (1.8837) [2021-04-16 09:44:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][730/1251] eta 0:02:30 lr 0.000306 time 0.3006 (0.2890) loss 3.6929 (3.3795) grad_norm 1.6655 (1.8843) [2021-04-16 09:44:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][740/1251] eta 0:02:27 lr 0.000306 time 0.2637 (0.2888) loss 4.0295 (3.3792) grad_norm 2.0776 (1.8844) [2021-04-16 09:44:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][750/1251] eta 0:02:24 lr 0.000306 time 0.2951 (0.2886) loss 3.7449 (3.3825) grad_norm 1.7878 (1.8827) [2021-04-16 09:44:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][760/1251] eta 0:02:21 lr 0.000306 time 0.2839 (0.2885) loss 3.7327 (3.3813) grad_norm 1.7224 (1.8816) [2021-04-16 09:45:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][770/1251] eta 0:02:18 lr 0.000305 time 0.2783 (0.2883) loss 2.3303 (3.3822) grad_norm 1.7950 (1.8801) [2021-04-16 09:45:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][780/1251] eta 0:02:15 lr 0.000305 time 0.2901 (0.2881) loss 4.1253 (3.3807) grad_norm 1.6328 (1.8793) [2021-04-16 09:45:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][790/1251] eta 0:02:12 lr 0.000305 time 0.2905 (0.2880) loss 4.3834 (3.3823) grad_norm 2.2645 (1.8788) [2021-04-16 09:45:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][800/1251] eta 0:02:09 lr 0.000305 time 0.2806 (0.2878) loss 2.8764 (3.3807) grad_norm 1.7629 (1.8785) [2021-04-16 09:45:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][810/1251] eta 0:02:06 lr 0.000305 time 0.2778 (0.2878) loss 3.6345 (3.3827) grad_norm 1.7306 (1.8784) [2021-04-16 09:45:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][820/1251] eta 0:02:03 lr 0.000305 time 0.2766 (0.2876) loss 3.3330 (3.3819) grad_norm 2.1122 (1.8787) [2021-04-16 09:45:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][830/1251] eta 0:02:00 lr 0.000305 time 0.2747 (0.2874) loss 3.9725 (3.3828) grad_norm 2.0412 (1.8780) [2021-04-16 09:45:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][840/1251] eta 0:01:58 lr 0.000305 time 0.2537 (0.2872) loss 2.6890 (3.3818) grad_norm 1.9454 (1.8776) [2021-04-16 09:45:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][850/1251] eta 0:01:55 lr 0.000305 time 0.2734 (0.2871) loss 2.3883 (3.3802) grad_norm 1.7913 (1.8770) [2021-04-16 09:45:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][860/1251] eta 0:01:52 lr 0.000305 time 0.2795 (0.2869) loss 3.7896 (3.3804) grad_norm 1.9815 (1.8780) [2021-04-16 09:45:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][870/1251] eta 0:01:49 lr 0.000305 time 0.2684 (0.2869) loss 2.6248 (3.3761) grad_norm 1.9018 (1.8791) [2021-04-16 09:45:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][880/1251] eta 0:01:46 lr 0.000305 time 0.2518 (0.2867) loss 4.1948 (3.3803) grad_norm 1.6662 (1.8797) [2021-04-16 09:45:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][890/1251] eta 0:01:43 lr 0.000305 time 0.2819 (0.2866) loss 4.0709 (3.3828) grad_norm 1.5576 (1.8789) [2021-04-16 09:45:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][900/1251] eta 0:01:40 lr 0.000305 time 0.2546 (0.2865) loss 3.5353 (3.3830) grad_norm 2.0096 (1.8790) [2021-04-16 09:45:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][910/1251] eta 0:01:37 lr 0.000305 time 0.2943 (0.2865) loss 3.5944 (3.3827) grad_norm 2.1039 (1.8798) [2021-04-16 09:45:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][920/1251] eta 0:01:34 lr 0.000305 time 0.3001 (0.2864) loss 4.0115 (3.3854) grad_norm 1.6808 (1.8803) [2021-04-16 09:45:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][930/1251] eta 0:01:31 lr 0.000305 time 0.2497 (0.2865) loss 3.5767 (3.3837) grad_norm 2.0722 (1.8822) [2021-04-16 09:45:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][940/1251] eta 0:01:29 lr 0.000305 time 0.2933 (0.2866) loss 2.9590 (3.3830) grad_norm 1.7055 (1.8814) [2021-04-16 09:45:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][950/1251] eta 0:01:26 lr 0.000305 time 0.2780 (0.2865) loss 3.8287 (3.3850) grad_norm 2.0318 (1.8820) [2021-04-16 09:45:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][960/1251] eta 0:01:23 lr 0.000305 time 0.2797 (0.2864) loss 3.7311 (3.3875) grad_norm 1.9158 (1.8821) [2021-04-16 09:45:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][970/1251] eta 0:01:20 lr 0.000305 time 0.2753 (0.2863) loss 2.9642 (3.3874) grad_norm 1.7993 (1.8825) [2021-04-16 09:46:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][980/1251] eta 0:01:17 lr 0.000305 time 0.3003 (0.2863) loss 3.4824 (3.3856) grad_norm 1.8529 (1.8820) [2021-04-16 09:46:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][990/1251] eta 0:01:14 lr 0.000305 time 0.2849 (0.2861) loss 3.9431 (3.3866) grad_norm 1.9304 (1.8829) [2021-04-16 09:46:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][1000/1251] eta 0:01:11 lr 0.000305 time 0.2683 (0.2861) loss 3.1679 (3.3809) grad_norm 1.9399 (1.8820) [2021-04-16 09:46:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][1010/1251] eta 0:01:08 lr 0.000305 time 0.2995 (0.2861) loss 3.9408 (3.3830) grad_norm 1.8231 (1.8817) [2021-04-16 09:46:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][1020/1251] eta 0:01:06 lr 0.000305 time 0.2669 (0.2860) loss 3.6224 (3.3816) grad_norm 1.8042 (1.8818) [2021-04-16 09:46:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][1030/1251] eta 0:01:03 lr 0.000305 time 0.2833 (0.2859) loss 3.9814 (3.3834) grad_norm 1.8365 (1.8811) [2021-04-16 09:46:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][1040/1251] eta 0:01:00 lr 0.000304 time 0.2787 (0.2858) loss 3.5354 (3.3828) grad_norm 2.3010 (1.8815) [2021-04-16 09:46:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][1050/1251] eta 0:00:57 lr 0.000304 time 0.2571 (0.2857) loss 2.9189 (3.3818) grad_norm 1.8963 (1.8811) [2021-04-16 09:46:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][1060/1251] eta 0:00:54 lr 0.000304 time 0.2743 (0.2856) loss 3.4901 (3.3831) grad_norm 1.6062 (1.8804) [2021-04-16 09:46:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][1070/1251] eta 0:00:51 lr 0.000304 time 0.2749 (0.2856) loss 4.2341 (3.3828) grad_norm 2.0971 (1.8805) [2021-04-16 09:46:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][1080/1251] eta 0:00:48 lr 0.000304 time 0.2768 (0.2856) loss 2.4774 (3.3822) grad_norm 1.7355 (1.8811) [2021-04-16 09:46:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][1090/1251] eta 0:00:45 lr 0.000304 time 0.2669 (0.2856) loss 3.0093 (3.3829) grad_norm 1.9398 (1.8813) [2021-04-16 09:46:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][1100/1251] eta 0:00:43 lr 0.000304 time 0.2578 (0.2855) loss 2.5663 (3.3825) grad_norm 1.9880 (1.8817) [2021-04-16 09:46:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][1110/1251] eta 0:00:40 lr 0.000304 time 0.2818 (0.2854) loss 3.5272 (3.3829) grad_norm 2.1405 (1.8821) [2021-04-16 09:46:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][1120/1251] eta 0:00:37 lr 0.000304 time 0.2942 (0.2855) loss 3.9098 (3.3857) grad_norm 2.0656 (1.8820) [2021-04-16 09:46:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][1130/1251] eta 0:00:34 lr 0.000304 time 0.2575 (0.2854) loss 3.4163 (3.3877) grad_norm 1.8027 (1.8822) [2021-04-16 09:46:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][1140/1251] eta 0:00:31 lr 0.000304 time 0.2527 (0.2854) loss 3.3829 (3.3875) grad_norm 1.7454 (1.8819) [2021-04-16 09:46:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][1150/1251] eta 0:00:28 lr 0.000304 time 0.2927 (0.2853) loss 3.2627 (3.3881) grad_norm 1.7880 (1.8826) [2021-04-16 09:46:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][1160/1251] eta 0:00:25 lr 0.000304 time 0.2801 (0.2854) loss 3.9695 (3.3892) grad_norm 2.0856 (1.8825) [2021-04-16 09:46:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][1170/1251] eta 0:00:23 lr 0.000304 time 0.2779 (0.2853) loss 3.7936 (3.3877) grad_norm 1.8121 (1.8814) [2021-04-16 09:46:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][1180/1251] eta 0:00:20 lr 0.000304 time 0.2795 (0.2853) loss 2.1854 (3.3853) grad_norm 1.6520 (1.8808) [2021-04-16 09:46:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][1190/1251] eta 0:00:17 lr 0.000304 time 0.2804 (0.2852) loss 2.6969 (3.3842) grad_norm 1.9512 (1.8810) [2021-04-16 09:47:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][1200/1251] eta 0:00:14 lr 0.000304 time 0.2982 (0.2852) loss 3.8227 (3.3847) grad_norm 1.7810 (1.8809) [2021-04-16 09:47:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][1210/1251] eta 0:00:11 lr 0.000304 time 0.2588 (0.2851) loss 2.2054 (3.3838) grad_norm 1.9584 (1.8811) [2021-04-16 09:47:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][1220/1251] eta 0:00:08 lr 0.000304 time 0.2823 (0.2851) loss 3.5201 (3.3832) grad_norm 2.0673 (1.8807) [2021-04-16 09:47:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][1230/1251] eta 0:00:05 lr 0.000304 time 0.2714 (0.2851) loss 3.5316 (3.3834) grad_norm 1.7001 (1.8804) [2021-04-16 09:47:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][1240/1251] eta 0:00:03 lr 0.000304 time 0.2597 (0.2850) loss 4.1834 (3.3818) grad_norm 1.8917 (1.8816) [2021-04-16 09:47:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [189/300][1250/1251] eta 0:00:00 lr 0.000304 time 0.2485 (0.2847) loss 3.2111 (3.3818) grad_norm 1.8258 (1.8808) [2021-04-16 09:47:25 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 189 training takes 0:06:06 [2021-04-16 09:47:25 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_189.pth saving...... [2021-04-16 09:47:43 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_189.pth saved !!! [2021-04-16 09:47:44 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.299 (1.299) Loss 1.0196 (1.0196) Acc@1 75.195 (75.195) Acc@5 94.043 (94.043) [2021-04-16 09:47:46 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.581 (0.267) Loss 0.8791 (0.9513) Acc@1 80.859 (77.539) Acc@5 94.434 (94.052) [2021-04-16 09:47:48 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.553 (0.233) Loss 1.0089 (0.9465) Acc@1 77.246 (77.985) Acc@5 93.066 (94.117) [2021-04-16 09:47:50 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.344 (0.218) Loss 0.9890 (0.9488) Acc@1 76.270 (77.848) Acc@5 91.992 (94.071) [2021-04-16 09:47:52 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.218) Loss 0.9648 (0.9455) Acc@1 76.562 (77.820) Acc@5 93.848 (94.129) [2021-04-16 09:48:01 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 77.874 Acc@5 94.220 [2021-04-16 09:48:01 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 77.9% [2021-04-16 09:48:01 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 77.87% [2021-04-16 09:48:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][0/1251] eta 1:39:07 lr 0.000304 time 4.7543 (4.7543) loss 4.1037 (4.1037) grad_norm 1.7780 (1.7780) [2021-04-16 09:48:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][10/1251] eta 0:14:04 lr 0.000304 time 0.2742 (0.6803) loss 3.2412 (3.4463) grad_norm 1.6522 (1.8825) [2021-04-16 09:48:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][20/1251] eta 0:09:58 lr 0.000304 time 0.2448 (0.4858) loss 3.3448 (3.3567) grad_norm 1.9401 (1.8853) [2021-04-16 09:48:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][30/1251] eta 0:08:31 lr 0.000304 time 0.2708 (0.4187) loss 2.5941 (3.3719) grad_norm 1.6699 (1.9127) [2021-04-16 09:48:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][40/1251] eta 0:07:45 lr 0.000304 time 0.2769 (0.3846) loss 2.4309 (3.3772) grad_norm 1.7555 (1.9337) [2021-04-16 09:48:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][50/1251] eta 0:07:17 lr 0.000303 time 0.2713 (0.3647) loss 3.6016 (3.3883) grad_norm 1.9600 (1.9211) [2021-04-16 09:48:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][60/1251] eta 0:06:58 lr 0.000303 time 0.2852 (0.3510) loss 2.8476 (3.3955) grad_norm 1.8573 (1.9065) [2021-04-16 09:48:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][70/1251] eta 0:06:42 lr 0.000303 time 0.2929 (0.3412) loss 3.6994 (3.4183) grad_norm 2.0098 (nan) [2021-04-16 09:48:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][80/1251] eta 0:06:31 lr 0.000303 time 0.2673 (0.3340) loss 3.4156 (3.3892) grad_norm 1.9530 (nan) [2021-04-16 09:48:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][90/1251] eta 0:06:21 lr 0.000303 time 0.2818 (0.3283) loss 4.1160 (3.4178) grad_norm 1.7251 (nan) [2021-04-16 09:48:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][100/1251] eta 0:06:12 lr 0.000303 time 0.2887 (0.3236) loss 3.4963 (3.4308) grad_norm 1.8733 (nan) [2021-04-16 09:48:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][110/1251] eta 0:06:04 lr 0.000303 time 0.2791 (0.3194) loss 3.7363 (3.4197) grad_norm 1.8208 (nan) [2021-04-16 09:48:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][120/1251] eta 0:05:58 lr 0.000303 time 0.2733 (0.3173) loss 3.4419 (3.4059) grad_norm 1.7401 (nan) [2021-04-16 09:48:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][130/1251] eta 0:05:52 lr 0.000303 time 0.2603 (0.3142) loss 2.6226 (3.4083) grad_norm 1.8819 (nan) [2021-04-16 09:48:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][140/1251] eta 0:05:48 lr 0.000303 time 0.2609 (0.3135) loss 3.0448 (3.3915) grad_norm 1.8618 (nan) [2021-04-16 09:48:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][150/1251] eta 0:05:43 lr 0.000303 time 0.3142 (0.3121) loss 3.5659 (3.3930) grad_norm 2.1311 (nan) [2021-04-16 09:48:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][160/1251] eta 0:05:37 lr 0.000303 time 0.2748 (0.3098) loss 4.0996 (3.3965) grad_norm 2.4477 (nan) [2021-04-16 09:48:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][170/1251] eta 0:05:32 lr 0.000303 time 0.3020 (0.3080) loss 4.0286 (3.3939) grad_norm 1.5623 (nan) [2021-04-16 09:48:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][180/1251] eta 0:05:27 lr 0.000303 time 0.2569 (0.3062) loss 3.8594 (3.3986) grad_norm 1.8550 (nan) [2021-04-16 09:48:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][190/1251] eta 0:05:23 lr 0.000303 time 0.2691 (0.3048) loss 2.3417 (3.3774) grad_norm 1.8126 (nan) [2021-04-16 09:49:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][200/1251] eta 0:05:19 lr 0.000303 time 0.2851 (0.3038) loss 3.8038 (3.3837) grad_norm 3.5556 (nan) [2021-04-16 09:49:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][210/1251] eta 0:05:15 lr 0.000303 time 0.2693 (0.3028) loss 3.6688 (3.3850) grad_norm 1.7103 (nan) [2021-04-16 09:49:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][220/1251] eta 0:05:10 lr 0.000303 time 0.2826 (0.3016) loss 3.1176 (3.3728) grad_norm 1.7194 (nan) [2021-04-16 09:49:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][230/1251] eta 0:05:07 lr 0.000303 time 0.2718 (0.3007) loss 3.0393 (3.3562) grad_norm 1.9604 (nan) [2021-04-16 09:49:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][240/1251] eta 0:05:03 lr 0.000303 time 0.3017 (0.2999) loss 3.3327 (3.3521) grad_norm 1.7507 (nan) [2021-04-16 09:49:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][250/1251] eta 0:04:59 lr 0.000303 time 0.2845 (0.2991) loss 2.2767 (3.3494) grad_norm 2.0278 (nan) [2021-04-16 09:49:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][260/1251] eta 0:04:55 lr 0.000303 time 0.2822 (0.2983) loss 3.9035 (3.3427) grad_norm 1.9185 (nan) [2021-04-16 09:49:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][270/1251] eta 0:04:51 lr 0.000303 time 0.2842 (0.2976) loss 3.6668 (3.3455) grad_norm 1.8913 (nan) [2021-04-16 09:49:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][280/1251] eta 0:04:48 lr 0.000303 time 0.2945 (0.2970) loss 4.0079 (3.3609) grad_norm 2.1101 (nan) [2021-04-16 09:49:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][290/1251] eta 0:04:45 lr 0.000303 time 0.2685 (0.2967) loss 2.8228 (3.3559) grad_norm 2.0601 (nan) [2021-04-16 09:49:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][300/1251] eta 0:04:42 lr 0.000303 time 0.2482 (0.2966) loss 3.3846 (3.3629) grad_norm 1.9522 (nan) [2021-04-16 09:49:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][310/1251] eta 0:04:38 lr 0.000302 time 0.2836 (0.2959) loss 3.5209 (3.3662) grad_norm 2.1786 (nan) [2021-04-16 09:49:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][320/1251] eta 0:04:35 lr 0.000302 time 0.3013 (0.2958) loss 3.1861 (3.3679) grad_norm 1.9216 (nan) [2021-04-16 09:49:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][330/1251] eta 0:04:32 lr 0.000302 time 0.2811 (0.2954) loss 2.1988 (3.3705) grad_norm 1.8495 (nan) [2021-04-16 09:49:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][340/1251] eta 0:04:28 lr 0.000302 time 0.3022 (0.2950) loss 4.1245 (3.3665) grad_norm 2.1945 (nan) [2021-04-16 09:49:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][350/1251] eta 0:04:25 lr 0.000302 time 0.2779 (0.2944) loss 3.6304 (3.3698) grad_norm 2.3495 (nan) [2021-04-16 09:49:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][360/1251] eta 0:04:22 lr 0.000302 time 0.3082 (0.2943) loss 3.1522 (3.3624) grad_norm 1.8525 (nan) [2021-04-16 09:49:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][370/1251] eta 0:04:19 lr 0.000302 time 0.2549 (0.2942) loss 3.1788 (3.3617) grad_norm 1.6941 (nan) [2021-04-16 09:49:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][380/1251] eta 0:04:15 lr 0.000302 time 0.2754 (0.2938) loss 2.8823 (3.3580) grad_norm 2.0161 (nan) [2021-04-16 09:49:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][390/1251] eta 0:04:12 lr 0.000302 time 0.2752 (0.2934) loss 3.0156 (3.3634) grad_norm 2.0533 (nan) [2021-04-16 09:49:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][400/1251] eta 0:04:09 lr 0.000302 time 0.2909 (0.2932) loss 2.5071 (3.3613) grad_norm 1.7055 (nan) [2021-04-16 09:50:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][410/1251] eta 0:04:06 lr 0.000302 time 0.2673 (0.2928) loss 3.9172 (3.3596) grad_norm 1.6455 (nan) [2021-04-16 09:50:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][420/1251] eta 0:04:03 lr 0.000302 time 0.2962 (0.2929) loss 4.1534 (3.3676) grad_norm 1.6607 (nan) [2021-04-16 09:50:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][430/1251] eta 0:04:00 lr 0.000302 time 0.2541 (0.2928) loss 3.6585 (3.3689) grad_norm 2.2837 (nan) [2021-04-16 09:50:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][440/1251] eta 0:03:57 lr 0.000302 time 0.2723 (0.2925) loss 3.2612 (3.3649) grad_norm 1.9618 (nan) [2021-04-16 09:50:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][450/1251] eta 0:03:54 lr 0.000302 time 0.2979 (0.2922) loss 2.7622 (3.3685) grad_norm 2.2939 (nan) [2021-04-16 09:50:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][460/1251] eta 0:03:51 lr 0.000302 time 0.2792 (0.2921) loss 3.6035 (3.3725) grad_norm 1.9596 (nan) [2021-04-16 09:50:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][470/1251] eta 0:03:47 lr 0.000302 time 0.2811 (0.2917) loss 4.0300 (3.3762) grad_norm 2.1301 (nan) [2021-04-16 09:50:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][480/1251] eta 0:03:44 lr 0.000302 time 0.2682 (0.2915) loss 2.7592 (3.3805) grad_norm 2.2517 (nan) [2021-04-16 09:50:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][490/1251] eta 0:03:41 lr 0.000302 time 0.2867 (0.2911) loss 3.8519 (3.3710) grad_norm 2.0424 (nan) [2021-04-16 09:50:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][500/1251] eta 0:03:38 lr 0.000302 time 0.3010 (0.2909) loss 3.6654 (3.3742) grad_norm 1.6936 (nan) [2021-04-16 09:50:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][510/1251] eta 0:03:35 lr 0.000302 time 0.2827 (0.2907) loss 2.4267 (3.3676) grad_norm 1.8708 (nan) [2021-04-16 09:50:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][520/1251] eta 0:03:32 lr 0.000302 time 0.3082 (0.2906) loss 2.4952 (3.3669) grad_norm 1.8621 (nan) [2021-04-16 09:50:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][530/1251] eta 0:03:29 lr 0.000302 time 0.2647 (0.2904) loss 2.6863 (3.3674) grad_norm 1.8156 (nan) [2021-04-16 09:50:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][540/1251] eta 0:03:26 lr 0.000302 time 0.2698 (0.2902) loss 2.7267 (3.3721) grad_norm 1.8510 (nan) [2021-04-16 09:50:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][550/1251] eta 0:03:23 lr 0.000302 time 0.2763 (0.2900) loss 3.5972 (3.3756) grad_norm 1.5942 (nan) [2021-04-16 09:50:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][560/1251] eta 0:03:20 lr 0.000302 time 0.3028 (0.2898) loss 3.5668 (3.3713) grad_norm 1.6826 (nan) [2021-04-16 09:50:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][570/1251] eta 0:03:17 lr 0.000302 time 0.3663 (0.2900) loss 2.3379 (3.3711) grad_norm 1.7864 (nan) [2021-04-16 09:50:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][580/1251] eta 0:03:14 lr 0.000301 time 0.2415 (0.2900) loss 3.7985 (3.3750) grad_norm 1.8040 (nan) [2021-04-16 09:50:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][590/1251] eta 0:03:11 lr 0.000301 time 0.2689 (0.2899) loss 3.7637 (3.3803) grad_norm 1.7627 (nan) [2021-04-16 09:50:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][600/1251] eta 0:03:08 lr 0.000301 time 0.2711 (0.2896) loss 2.9333 (3.3807) grad_norm 1.5936 (nan) [2021-04-16 09:50:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][610/1251] eta 0:03:05 lr 0.000301 time 0.2682 (0.2895) loss 3.2953 (3.3809) grad_norm 1.7936 (nan) [2021-04-16 09:51:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][620/1251] eta 0:03:02 lr 0.000301 time 0.3001 (0.2895) loss 3.4390 (3.3815) grad_norm 1.5147 (nan) [2021-04-16 09:51:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][630/1251] eta 0:02:59 lr 0.000301 time 0.2876 (0.2894) loss 2.5504 (3.3757) grad_norm 1.8136 (nan) [2021-04-16 09:51:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][640/1251] eta 0:02:56 lr 0.000301 time 0.2875 (0.2892) loss 3.3819 (3.3777) grad_norm 1.8833 (nan) [2021-04-16 09:51:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][650/1251] eta 0:02:53 lr 0.000301 time 0.2857 (0.2891) loss 3.4784 (3.3744) grad_norm 1.8401 (nan) [2021-04-16 09:51:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][660/1251] eta 0:02:50 lr 0.000301 time 0.2639 (0.2889) loss 3.3187 (3.3725) grad_norm 1.8564 (nan) [2021-04-16 09:51:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][670/1251] eta 0:02:47 lr 0.000301 time 0.3081 (0.2888) loss 3.5364 (3.3726) grad_norm 2.1291 (nan) [2021-04-16 09:51:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][680/1251] eta 0:02:44 lr 0.000301 time 0.2638 (0.2887) loss 3.9348 (3.3754) grad_norm 1.8070 (nan) [2021-04-16 09:51:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][690/1251] eta 0:02:41 lr 0.000301 time 0.2990 (0.2888) loss 3.1787 (3.3767) grad_norm 2.0620 (nan) [2021-04-16 09:51:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][700/1251] eta 0:02:39 lr 0.000301 time 0.2716 (0.2887) loss 3.4146 (3.3737) grad_norm 2.1464 (nan) [2021-04-16 09:51:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][710/1251] eta 0:02:36 lr 0.000301 time 0.2838 (0.2886) loss 3.3392 (3.3705) grad_norm 1.8180 (nan) [2021-04-16 09:51:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][720/1251] eta 0:02:33 lr 0.000301 time 0.2769 (0.2888) loss 2.9357 (3.3730) grad_norm 1.8189 (nan) [2021-04-16 09:51:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][730/1251] eta 0:02:30 lr 0.000301 time 0.2523 (0.2885) loss 2.4109 (3.3701) grad_norm 1.9680 (nan) [2021-04-16 09:51:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][740/1251] eta 0:02:27 lr 0.000301 time 0.2835 (0.2884) loss 2.7891 (3.3694) grad_norm 1.7557 (nan) [2021-04-16 09:51:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][750/1251] eta 0:02:24 lr 0.000301 time 0.2866 (0.2882) loss 3.3888 (3.3709) grad_norm 1.7248 (nan) [2021-04-16 09:51:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][760/1251] eta 0:02:21 lr 0.000301 time 0.2872 (0.2881) loss 3.4224 (3.3702) grad_norm 1.9326 (nan) [2021-04-16 09:51:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][770/1251] eta 0:02:18 lr 0.000301 time 0.2735 (0.2880) loss 3.3082 (3.3671) grad_norm 1.6812 (nan) [2021-04-16 09:51:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][780/1251] eta 0:02:15 lr 0.000301 time 0.2781 (0.2878) loss 3.3398 (3.3680) grad_norm 2.0034 (nan) [2021-04-16 09:51:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][790/1251] eta 0:02:12 lr 0.000301 time 0.2939 (0.2878) loss 2.9978 (3.3671) grad_norm 2.1586 (nan) [2021-04-16 09:51:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][800/1251] eta 0:02:09 lr 0.000301 time 0.2931 (0.2877) loss 3.8680 (3.3682) grad_norm 1.8747 (nan) [2021-04-16 09:51:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][810/1251] eta 0:02:06 lr 0.000301 time 0.2747 (0.2877) loss 3.6519 (3.3656) grad_norm 2.0407 (nan) [2021-04-16 09:51:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][820/1251] eta 0:02:03 lr 0.000301 time 0.2608 (0.2875) loss 3.5778 (3.3657) grad_norm 2.1259 (nan) [2021-04-16 09:51:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][830/1251] eta 0:02:00 lr 0.000301 time 0.3045 (0.2874) loss 4.2113 (3.3684) grad_norm 2.1634 (nan) [2021-04-16 09:52:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][840/1251] eta 0:01:58 lr 0.000300 time 0.2700 (0.2873) loss 2.0887 (3.3632) grad_norm 1.9834 (nan) [2021-04-16 09:52:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][850/1251] eta 0:01:55 lr 0.000300 time 0.2596 (0.2872) loss 3.9780 (3.3640) grad_norm 1.6043 (nan) [2021-04-16 09:52:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][860/1251] eta 0:01:52 lr 0.000300 time 0.2642 (0.2871) loss 3.7018 (3.3678) grad_norm 2.0273 (nan) [2021-04-16 09:52:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][870/1251] eta 0:01:49 lr 0.000300 time 0.2659 (0.2871) loss 3.5461 (3.3682) grad_norm 1.6351 (nan) [2021-04-16 09:52:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][880/1251] eta 0:01:46 lr 0.000300 time 0.2757 (0.2872) loss 3.3075 (3.3677) grad_norm 1.7582 (nan) [2021-04-16 09:52:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][890/1251] eta 0:01:43 lr 0.000300 time 0.2610 (0.2873) loss 3.2666 (3.3664) grad_norm 2.0716 (nan) [2021-04-16 09:52:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][900/1251] eta 0:01:40 lr 0.000300 time 0.2978 (0.2873) loss 4.0643 (3.3664) grad_norm 2.1281 (nan) [2021-04-16 09:52:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][910/1251] eta 0:01:37 lr 0.000300 time 0.2752 (0.2872) loss 2.9524 (3.3637) grad_norm 1.7343 (nan) [2021-04-16 09:52:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][920/1251] eta 0:01:35 lr 0.000300 time 0.2966 (0.2872) loss 4.1128 (3.3639) grad_norm 1.8197 (nan) [2021-04-16 09:52:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][930/1251] eta 0:01:32 lr 0.000300 time 0.2817 (0.2873) loss 4.1596 (3.3630) grad_norm 2.3491 (nan) [2021-04-16 09:52:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][940/1251] eta 0:01:29 lr 0.000300 time 0.2562 (0.2873) loss 3.4432 (3.3630) grad_norm 1.8308 (nan) [2021-04-16 09:52:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][950/1251] eta 0:01:26 lr 0.000300 time 0.2795 (0.2872) loss 3.3120 (3.3613) grad_norm 2.0493 (nan) [2021-04-16 09:52:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][960/1251] eta 0:01:23 lr 0.000300 time 0.2831 (0.2870) loss 2.9024 (3.3594) grad_norm 1.7692 (nan) [2021-04-16 09:52:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][970/1251] eta 0:01:20 lr 0.000300 time 0.2946 (0.2871) loss 3.3791 (3.3602) grad_norm 1.9839 (nan) [2021-04-16 09:52:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][980/1251] eta 0:01:17 lr 0.000300 time 0.2760 (0.2869) loss 2.7854 (3.3608) grad_norm 1.9848 (nan) [2021-04-16 09:52:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][990/1251] eta 0:01:14 lr 0.000300 time 0.3031 (0.2871) loss 2.4388 (3.3583) grad_norm 2.1122 (nan) [2021-04-16 09:52:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][1000/1251] eta 0:01:12 lr 0.000300 time 0.3219 (0.2870) loss 3.7389 (3.3583) grad_norm 1.8280 (nan) [2021-04-16 09:52:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][1010/1251] eta 0:01:09 lr 0.000300 time 0.2823 (0.2870) loss 3.6746 (3.3590) grad_norm 1.8081 (nan) [2021-04-16 09:52:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][1020/1251] eta 0:01:06 lr 0.000300 time 0.2710 (0.2868) loss 4.1186 (3.3620) grad_norm 2.0870 (nan) [2021-04-16 09:52:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][1030/1251] eta 0:01:03 lr 0.000300 time 0.2621 (0.2867) loss 2.6015 (3.3622) grad_norm 1.8803 (nan) [2021-04-16 09:52:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][1040/1251] eta 0:01:00 lr 0.000300 time 0.2707 (0.2867) loss 2.5087 (3.3599) grad_norm 1.8931 (nan) [2021-04-16 09:53:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][1050/1251] eta 0:00:57 lr 0.000300 time 0.2628 (0.2866) loss 2.1555 (3.3588) grad_norm 1.8779 (nan) [2021-04-16 09:53:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][1060/1251] eta 0:00:54 lr 0.000300 time 0.2540 (0.2865) loss 4.2249 (3.3585) grad_norm 1.8065 (nan) [2021-04-16 09:53:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][1070/1251] eta 0:00:51 lr 0.000300 time 0.2811 (0.2865) loss 3.9061 (3.3598) grad_norm 1.8382 (nan) [2021-04-16 09:53:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][1080/1251] eta 0:00:49 lr 0.000300 time 0.2842 (0.2866) loss 2.6261 (3.3589) grad_norm 2.2281 (nan) [2021-04-16 09:53:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][1090/1251] eta 0:00:46 lr 0.000300 time 0.3006 (0.2866) loss 3.5301 (3.3609) grad_norm 2.1001 (nan) [2021-04-16 09:53:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][1100/1251] eta 0:00:43 lr 0.000300 time 0.2855 (0.2865) loss 3.8383 (3.3625) grad_norm 1.6759 (nan) [2021-04-16 09:53:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][1110/1251] eta 0:00:40 lr 0.000299 time 0.2875 (0.2864) loss 3.5783 (3.3635) grad_norm 1.7909 (nan) [2021-04-16 09:53:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][1120/1251] eta 0:00:37 lr 0.000299 time 0.2947 (0.2864) loss 3.8721 (3.3632) grad_norm 2.0626 (nan) [2021-04-16 09:53:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][1130/1251] eta 0:00:34 lr 0.000299 time 0.2809 (0.2863) loss 3.5175 (3.3648) grad_norm 2.0710 (nan) [2021-04-16 09:53:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][1140/1251] eta 0:00:31 lr 0.000299 time 0.2821 (0.2862) loss 4.1021 (3.3644) grad_norm 1.7975 (nan) [2021-04-16 09:53:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][1150/1251] eta 0:00:28 lr 0.000299 time 0.2781 (0.2863) loss 3.7839 (3.3631) grad_norm 1.8671 (nan) [2021-04-16 09:53:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][1160/1251] eta 0:00:26 lr 0.000299 time 0.2888 (0.2863) loss 3.0073 (3.3624) grad_norm 1.7986 (nan) [2021-04-16 09:53:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][1170/1251] eta 0:00:23 lr 0.000299 time 0.2835 (0.2863) loss 4.0207 (3.3610) grad_norm 2.0144 (nan) [2021-04-16 09:53:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][1180/1251] eta 0:00:20 lr 0.000299 time 0.2777 (0.2862) loss 3.6013 (3.3601) grad_norm 2.0823 (nan) [2021-04-16 09:53:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][1190/1251] eta 0:00:17 lr 0.000299 time 0.2779 (0.2861) loss 2.8809 (3.3595) grad_norm 1.7064 (nan) [2021-04-16 09:53:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][1200/1251] eta 0:00:14 lr 0.000299 time 0.2877 (0.2861) loss 3.8502 (3.3601) grad_norm 1.7740 (nan) [2021-04-16 09:53:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][1210/1251] eta 0:00:11 lr 0.000299 time 0.2771 (0.2860) loss 3.7150 (3.3585) grad_norm 2.2919 (nan) [2021-04-16 09:53:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][1220/1251] eta 0:00:08 lr 0.000299 time 0.2772 (0.2859) loss 2.3525 (3.3575) grad_norm 1.8275 (nan) [2021-04-16 09:53:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][1230/1251] eta 0:00:06 lr 0.000299 time 0.3058 (0.2859) loss 3.7672 (3.3584) grad_norm 1.8691 (nan) [2021-04-16 09:53:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][1240/1251] eta 0:00:03 lr 0.000299 time 0.2487 (0.2857) loss 3.5176 (3.3579) grad_norm 2.1390 (nan) [2021-04-16 09:53:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [190/300][1250/1251] eta 0:00:00 lr 0.000299 time 0.2483 (0.2854) loss 3.1917 (3.3586) grad_norm 1.9115 (nan) [2021-04-16 09:54:01 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 190 training takes 0:06:00 [2021-04-16 09:54:01 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_190.pth saving...... [2021-04-16 09:54:14 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_190.pth saved !!! [2021-04-16 09:54:15 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.223 (1.223) Loss 0.9580 (0.9580) Acc@1 76.074 (76.074) Acc@5 93.750 (93.750) [2021-04-16 09:54:17 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.341 (0.249) Loss 0.9278 (0.9510) Acc@1 79.004 (77.468) Acc@5 93.359 (93.892) [2021-04-16 09:54:19 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.108 (0.243) Loss 0.9888 (0.9593) Acc@1 77.051 (77.320) Acc@5 93.457 (93.936) [2021-04-16 09:54:22 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.091 (0.249) Loss 0.9597 (0.9452) Acc@1 76.953 (77.675) Acc@5 94.238 (94.106) [2021-04-16 09:54:23 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.215) Loss 1.0424 (0.9464) Acc@1 75.000 (77.684) Acc@5 93.262 (94.155) [2021-04-16 09:54:28 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 77.772 Acc@5 94.244 [2021-04-16 09:54:28 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 77.8% [2021-04-16 09:54:28 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 77.87% [2021-04-16 09:54:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][0/1251] eta 4:08:56 lr 0.000299 time 11.9394 (11.9394) loss 3.4359 (3.4359) grad_norm 1.9062 (1.9062) [2021-04-16 09:54:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][10/1251] eta 0:27:37 lr 0.000299 time 0.2693 (1.3360) loss 2.7907 (3.2536) grad_norm 1.5863 (1.8639) [2021-04-16 09:54:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][20/1251] eta 0:17:05 lr 0.000299 time 0.2854 (0.8327) loss 2.6961 (3.3480) grad_norm 2.3313 (1.9455) [2021-04-16 09:54:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][30/1251] eta 0:13:20 lr 0.000299 time 0.2953 (0.6554) loss 3.4058 (3.2600) grad_norm 1.9046 (1.9430) [2021-04-16 09:54:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][40/1251] eta 0:11:21 lr 0.000299 time 0.2704 (0.5629) loss 3.4653 (3.3041) grad_norm 1.6836 (1.9037) [2021-04-16 09:54:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][50/1251] eta 0:10:08 lr 0.000299 time 0.2869 (0.5066) loss 3.2342 (3.3125) grad_norm 2.0075 (1.8866) [2021-04-16 09:54:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][60/1251] eta 0:09:18 lr 0.000299 time 0.2886 (0.4689) loss 3.4697 (3.3085) grad_norm 2.1461 (1.8752) [2021-04-16 09:54:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][70/1251] eta 0:08:42 lr 0.000299 time 0.2783 (0.4420) loss 3.1241 (3.3407) grad_norm 1.9397 (1.8726) [2021-04-16 09:55:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][80/1251] eta 0:08:15 lr 0.000299 time 0.2755 (0.4234) loss 3.3593 (3.3418) grad_norm 1.8462 (1.9025) [2021-04-16 09:55:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][90/1251] eta 0:07:55 lr 0.000299 time 0.2855 (0.4099) loss 3.1776 (3.3361) grad_norm 2.1183 (1.9068) [2021-04-16 09:55:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][100/1251] eta 0:07:38 lr 0.000299 time 0.2611 (0.3984) loss 3.7529 (3.3588) grad_norm 1.8987 (1.9104) [2021-04-16 09:55:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][110/1251] eta 0:07:21 lr 0.000299 time 0.2902 (0.3873) loss 2.2556 (3.3545) grad_norm 1.7228 (1.8941) [2021-04-16 09:55:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][120/1251] eta 0:07:07 lr 0.000298 time 0.2751 (0.3780) loss 2.5091 (3.3509) grad_norm 2.0334 (1.8949) [2021-04-16 09:55:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][130/1251] eta 0:06:55 lr 0.000298 time 0.2971 (0.3703) loss 4.0608 (3.3648) grad_norm 1.7668 (1.8944) [2021-04-16 09:55:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][140/1251] eta 0:06:46 lr 0.000298 time 0.2741 (0.3655) loss 4.2062 (3.3643) grad_norm 1.7466 (1.8925) [2021-04-16 09:55:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][150/1251] eta 0:06:35 lr 0.000298 time 0.2762 (0.3595) loss 3.3769 (3.3556) grad_norm 1.7888 (1.8951) [2021-04-16 09:55:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][160/1251] eta 0:06:26 lr 0.000298 time 0.2677 (0.3543) loss 3.7252 (3.3370) grad_norm 1.7607 (1.8960) [2021-04-16 09:55:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][170/1251] eta 0:06:17 lr 0.000298 time 0.2701 (0.3495) loss 3.6333 (3.3338) grad_norm 1.8225 (1.8968) [2021-04-16 09:55:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][180/1251] eta 0:06:10 lr 0.000298 time 0.2741 (0.3457) loss 3.5453 (3.3453) grad_norm 1.8132 (1.8953) [2021-04-16 09:55:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][190/1251] eta 0:06:02 lr 0.000298 time 0.2750 (0.3418) loss 2.9752 (3.3331) grad_norm 2.0124 (1.8974) [2021-04-16 09:55:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][200/1251] eta 0:05:56 lr 0.000298 time 0.2676 (0.3392) loss 3.8703 (3.3339) grad_norm 1.8506 (1.8979) [2021-04-16 09:55:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][210/1251] eta 0:05:50 lr 0.000298 time 0.2901 (0.3367) loss 3.6378 (3.3302) grad_norm 2.0296 (1.8967) [2021-04-16 09:55:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][220/1251] eta 0:05:44 lr 0.000298 time 0.2938 (0.3340) loss 2.8997 (3.3210) grad_norm 1.8487 (1.8991) [2021-04-16 09:55:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][230/1251] eta 0:05:38 lr 0.000298 time 0.2745 (0.3315) loss 2.7018 (3.3308) grad_norm 1.7702 (1.8960) [2021-04-16 09:55:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][240/1251] eta 0:05:32 lr 0.000298 time 0.2765 (0.3292) loss 3.3172 (3.3236) grad_norm 1.8607 (1.8921) [2021-04-16 09:55:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][250/1251] eta 0:05:27 lr 0.000298 time 0.2924 (0.3273) loss 3.0939 (3.3242) grad_norm 2.1052 (1.8937) [2021-04-16 09:55:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][260/1251] eta 0:05:22 lr 0.000298 time 0.2798 (0.3254) loss 4.1377 (3.3376) grad_norm 1.9963 (1.8940) [2021-04-16 09:55:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][270/1251] eta 0:05:17 lr 0.000298 time 0.2757 (0.3237) loss 3.5634 (3.3506) grad_norm 1.7042 (1.8943) [2021-04-16 09:55:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][280/1251] eta 0:05:12 lr 0.000298 time 0.2868 (0.3222) loss 3.7882 (3.3570) grad_norm 1.9664 (1.8946) [2021-04-16 09:56:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][290/1251] eta 0:05:08 lr 0.000298 time 0.2745 (0.3207) loss 3.3108 (3.3535) grad_norm 1.9451 (1.8947) [2021-04-16 09:56:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][300/1251] eta 0:05:04 lr 0.000298 time 0.2574 (0.3197) loss 3.8136 (3.3463) grad_norm 2.3302 (1.8999) [2021-04-16 09:56:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][310/1251] eta 0:04:59 lr 0.000298 time 0.3001 (0.3187) loss 3.0150 (3.3465) grad_norm 2.3660 (1.9037) [2021-04-16 09:56:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][320/1251] eta 0:04:55 lr 0.000298 time 0.2956 (0.3174) loss 3.7560 (3.3580) grad_norm 1.9179 (1.9022) [2021-04-16 09:56:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][330/1251] eta 0:04:51 lr 0.000298 time 0.2759 (0.3162) loss 4.1367 (3.3608) grad_norm 2.0969 (1.9022) [2021-04-16 09:56:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][340/1251] eta 0:04:47 lr 0.000298 time 0.2796 (0.3152) loss 3.8465 (3.3680) grad_norm 1.6432 (1.9022) [2021-04-16 09:56:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][350/1251] eta 0:04:42 lr 0.000298 time 0.2687 (0.3141) loss 3.3651 (3.3705) grad_norm 1.6804 (1.9015) [2021-04-16 09:56:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][360/1251] eta 0:04:39 lr 0.000298 time 0.3240 (0.3137) loss 3.9600 (3.3754) grad_norm 1.8303 (1.9021) [2021-04-16 09:56:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][370/1251] eta 0:04:35 lr 0.000298 time 0.2771 (0.3127) loss 3.5097 (3.3812) grad_norm 1.9197 (1.9047) [2021-04-16 09:56:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][380/1251] eta 0:04:31 lr 0.000298 time 0.2867 (0.3117) loss 3.9686 (3.3740) grad_norm 2.3179 (1.9041) [2021-04-16 09:56:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][390/1251] eta 0:04:27 lr 0.000297 time 0.2723 (0.3110) loss 2.6051 (3.3772) grad_norm 1.9149 (1.9019) [2021-04-16 09:56:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][400/1251] eta 0:04:23 lr 0.000297 time 0.2911 (0.3101) loss 2.8122 (3.3774) grad_norm 2.2366 (1.9042) [2021-04-16 09:56:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][410/1251] eta 0:04:19 lr 0.000297 time 0.2655 (0.3091) loss 3.2170 (3.3759) grad_norm 1.7418 (1.9061) [2021-04-16 09:56:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][420/1251] eta 0:04:16 lr 0.000297 time 0.2900 (0.3084) loss 3.7622 (3.3768) grad_norm 2.0463 (1.9060) [2021-04-16 09:56:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][430/1251] eta 0:04:12 lr 0.000297 time 0.2697 (0.3077) loss 3.0259 (3.3778) grad_norm 1.6991 (1.9062) [2021-04-16 09:56:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][440/1251] eta 0:04:09 lr 0.000297 time 0.2852 (0.3071) loss 3.6181 (3.3801) grad_norm 2.6149 (1.9108) [2021-04-16 09:56:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][450/1251] eta 0:04:05 lr 0.000297 time 0.2655 (0.3064) loss 3.6572 (3.3862) grad_norm 1.6221 (1.9076) [2021-04-16 09:56:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][460/1251] eta 0:04:01 lr 0.000297 time 0.2596 (0.3057) loss 3.1454 (3.3821) grad_norm 1.8784 (1.9059) [2021-04-16 09:56:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][470/1251] eta 0:03:58 lr 0.000297 time 0.2855 (0.3052) loss 3.5165 (3.3871) grad_norm 2.1032 (1.9033) [2021-04-16 09:56:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][480/1251] eta 0:03:54 lr 0.000297 time 0.2900 (0.3045) loss 2.6389 (3.3857) grad_norm 1.8268 (1.9038) [2021-04-16 09:56:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][490/1251] eta 0:03:51 lr 0.000297 time 0.2750 (0.3041) loss 2.5100 (3.3924) grad_norm 1.7460 (1.9022) [2021-04-16 09:57:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][500/1251] eta 0:03:47 lr 0.000297 time 0.2888 (0.3035) loss 3.2189 (3.3955) grad_norm 1.7147 (1.8986) [2021-04-16 09:57:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][510/1251] eta 0:03:44 lr 0.000297 time 0.2882 (0.3031) loss 2.2196 (3.3848) grad_norm 1.5401 (1.8959) [2021-04-16 09:57:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][520/1251] eta 0:03:41 lr 0.000297 time 0.2824 (0.3026) loss 2.3125 (3.3795) grad_norm 1.9175 (1.8944) [2021-04-16 09:57:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][530/1251] eta 0:03:37 lr 0.000297 time 0.2778 (0.3021) loss 3.6627 (3.3803) grad_norm 1.6520 (1.8940) [2021-04-16 09:57:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][540/1251] eta 0:03:34 lr 0.000297 time 0.2637 (0.3019) loss 3.5414 (3.3868) grad_norm 2.1589 (1.8937) [2021-04-16 09:57:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][550/1251] eta 0:03:31 lr 0.000297 time 0.2691 (0.3015) loss 3.7108 (3.3892) grad_norm 1.7750 (1.8929) [2021-04-16 09:57:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][560/1251] eta 0:03:28 lr 0.000297 time 0.2978 (0.3011) loss 2.4329 (3.3840) grad_norm 2.4053 (1.8937) [2021-04-16 09:57:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][570/1251] eta 0:03:24 lr 0.000297 time 0.2554 (0.3008) loss 3.5227 (3.3844) grad_norm 1.7562 (1.8923) [2021-04-16 09:57:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][580/1251] eta 0:03:21 lr 0.000297 time 0.2486 (0.3006) loss 2.9821 (3.3853) grad_norm 2.1895 (1.8923) [2021-04-16 09:57:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][590/1251] eta 0:03:18 lr 0.000297 time 0.2645 (0.3001) loss 3.0345 (3.3845) grad_norm 1.8432 (1.8931) [2021-04-16 09:57:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][600/1251] eta 0:03:15 lr 0.000297 time 0.2586 (0.2996) loss 3.0645 (3.3874) grad_norm 1.9356 (1.8937) [2021-04-16 09:57:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][610/1251] eta 0:03:11 lr 0.000297 time 0.2877 (0.2992) loss 3.6671 (3.3840) grad_norm 1.5433 (1.8923) [2021-04-16 09:57:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][620/1251] eta 0:03:08 lr 0.000297 time 0.2764 (0.2988) loss 3.9017 (3.3799) grad_norm 1.6058 (1.8917) [2021-04-16 09:57:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][630/1251] eta 0:03:05 lr 0.000297 time 0.2604 (0.2984) loss 3.5879 (3.3797) grad_norm 1.7363 (1.8909) [2021-04-16 09:57:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][640/1251] eta 0:03:02 lr 0.000297 time 0.2805 (0.2982) loss 2.1658 (3.3763) grad_norm 1.8031 (1.8909) [2021-04-16 09:57:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][650/1251] eta 0:02:59 lr 0.000296 time 0.3749 (0.2981) loss 2.9591 (3.3834) grad_norm 1.7304 (1.8890) [2021-04-16 09:57:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][660/1251] eta 0:02:56 lr 0.000296 time 0.2827 (0.2979) loss 3.7946 (3.3837) grad_norm 1.8677 (1.8880) [2021-04-16 09:57:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][670/1251] eta 0:02:52 lr 0.000296 time 0.2583 (0.2977) loss 2.2806 (3.3808) grad_norm 1.8797 (1.8870) [2021-04-16 09:57:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][680/1251] eta 0:02:49 lr 0.000296 time 0.2501 (0.2976) loss 2.6503 (3.3814) grad_norm 1.5634 (1.8865) [2021-04-16 09:57:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][690/1251] eta 0:02:46 lr 0.000296 time 0.2608 (0.2974) loss 3.7299 (3.3818) grad_norm 2.3072 (1.8859) [2021-04-16 09:57:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][700/1251] eta 0:02:43 lr 0.000296 time 0.2611 (0.2971) loss 2.5750 (3.3782) grad_norm 1.9031 (1.8867) [2021-04-16 09:57:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][710/1251] eta 0:02:40 lr 0.000296 time 0.2573 (0.2968) loss 2.6675 (3.3773) grad_norm 1.7152 (1.8862) [2021-04-16 09:58:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][720/1251] eta 0:02:37 lr 0.000296 time 0.2569 (0.2968) loss 3.8528 (3.3799) grad_norm 2.5880 (1.8875) [2021-04-16 09:58:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][730/1251] eta 0:02:34 lr 0.000296 time 0.2923 (0.2965) loss 2.8503 (3.3787) grad_norm 2.2261 (1.8883) [2021-04-16 09:58:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][740/1251] eta 0:02:31 lr 0.000296 time 0.2810 (0.2963) loss 3.7491 (3.3787) grad_norm 2.2830 (1.8902) [2021-04-16 09:58:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][750/1251] eta 0:02:28 lr 0.000296 time 0.2495 (0.2961) loss 3.8448 (3.3794) grad_norm 1.7045 (1.8893) [2021-04-16 09:58:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][760/1251] eta 0:02:25 lr 0.000296 time 0.2735 (0.2958) loss 3.4357 (3.3789) grad_norm 1.9553 (1.8894) [2021-04-16 09:58:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][770/1251] eta 0:02:22 lr 0.000296 time 0.2658 (0.2956) loss 2.7855 (3.3779) grad_norm 1.8004 (1.8899) [2021-04-16 09:58:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][780/1251] eta 0:02:19 lr 0.000296 time 0.2625 (0.2955) loss 3.0214 (3.3774) grad_norm 2.0758 (1.8906) [2021-04-16 09:58:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][790/1251] eta 0:02:16 lr 0.000296 time 0.3009 (0.2953) loss 2.1416 (3.3750) grad_norm 1.7581 (1.8913) [2021-04-16 09:58:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][800/1251] eta 0:02:13 lr 0.000296 time 0.2706 (0.2951) loss 2.2679 (3.3774) grad_norm 1.8489 (1.8917) [2021-04-16 09:58:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][810/1251] eta 0:02:10 lr 0.000296 time 0.2860 (0.2948) loss 3.1082 (3.3735) grad_norm 1.7466 (1.8914) [2021-04-16 09:58:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][820/1251] eta 0:02:06 lr 0.000296 time 0.2922 (0.2946) loss 2.5306 (3.3751) grad_norm 1.7764 (1.8924) [2021-04-16 09:58:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][830/1251] eta 0:02:03 lr 0.000296 time 0.2510 (0.2943) loss 3.2691 (3.3760) grad_norm 1.6365 (1.8927) [2021-04-16 09:58:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][840/1251] eta 0:02:00 lr 0.000296 time 0.2742 (0.2941) loss 2.0787 (3.3724) grad_norm 1.8054 (1.8923) [2021-04-16 09:58:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][850/1251] eta 0:01:57 lr 0.000296 time 0.2766 (0.2939) loss 3.4922 (3.3683) grad_norm 1.8399 (1.8923) [2021-04-16 09:58:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][860/1251] eta 0:01:54 lr 0.000296 time 0.2925 (0.2937) loss 3.7162 (3.3711) grad_norm 2.2116 (1.8929) [2021-04-16 09:58:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][870/1251] eta 0:01:51 lr 0.000296 time 0.2619 (0.2937) loss 3.8688 (3.3713) grad_norm 2.1311 (1.8939) [2021-04-16 09:58:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][880/1251] eta 0:01:48 lr 0.000296 time 0.2754 (0.2934) loss 2.8651 (3.3732) grad_norm 1.7575 (1.8931) [2021-04-16 09:58:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][890/1251] eta 0:01:45 lr 0.000296 time 0.2893 (0.2932) loss 2.3852 (3.3723) grad_norm 2.2691 (1.8954) [2021-04-16 09:58:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][900/1251] eta 0:01:42 lr 0.000296 time 0.2857 (0.2931) loss 3.2259 (3.3699) grad_norm 1.9307 (1.8959) [2021-04-16 09:58:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][910/1251] eta 0:01:39 lr 0.000296 time 0.3021 (0.2930) loss 2.5908 (3.3690) grad_norm 1.8862 (1.8961) [2021-04-16 09:58:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][920/1251] eta 0:01:36 lr 0.000295 time 0.2676 (0.2929) loss 3.1396 (3.3731) grad_norm 1.7453 (1.8952) [2021-04-16 09:59:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][930/1251] eta 0:01:34 lr 0.000295 time 0.2930 (0.2929) loss 3.5487 (3.3721) grad_norm 1.7865 (1.8947) [2021-04-16 09:59:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][940/1251] eta 0:01:31 lr 0.000295 time 0.2919 (0.2927) loss 3.8058 (3.3743) grad_norm 1.7525 (1.8945) [2021-04-16 09:59:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][950/1251] eta 0:01:28 lr 0.000295 time 0.2666 (0.2925) loss 3.3349 (3.3751) grad_norm 1.9126 (1.8942) [2021-04-16 09:59:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][960/1251] eta 0:01:25 lr 0.000295 time 0.2866 (0.2924) loss 3.7017 (3.3737) grad_norm 2.1934 (1.8969) [2021-04-16 09:59:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][970/1251] eta 0:01:22 lr 0.000295 time 0.2700 (0.2922) loss 3.5243 (3.3752) grad_norm 2.0463 (1.8973) [2021-04-16 09:59:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][980/1251] eta 0:01:19 lr 0.000295 time 0.2706 (0.2921) loss 2.4597 (3.3741) grad_norm 1.6612 (1.8969) [2021-04-16 09:59:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][990/1251] eta 0:01:16 lr 0.000295 time 0.2776 (0.2920) loss 2.5041 (3.3742) grad_norm 1.8523 (1.8966) [2021-04-16 09:59:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][1000/1251] eta 0:01:13 lr 0.000295 time 0.3014 (0.2918) loss 3.3976 (3.3741) grad_norm 2.0080 (1.8967) [2021-04-16 09:59:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][1010/1251] eta 0:01:10 lr 0.000295 time 0.2704 (0.2916) loss 3.3997 (3.3746) grad_norm 2.1162 (1.8981) [2021-04-16 09:59:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][1020/1251] eta 0:01:07 lr 0.000295 time 0.2769 (0.2915) loss 3.2669 (3.3736) grad_norm 1.9705 (1.8971) [2021-04-16 09:59:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][1030/1251] eta 0:01:04 lr 0.000295 time 0.2669 (0.2914) loss 3.6988 (3.3738) grad_norm 1.6980 (1.8984) [2021-04-16 09:59:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][1040/1251] eta 0:01:01 lr 0.000295 time 0.3016 (0.2913) loss 3.6700 (3.3734) grad_norm 1.8671 (1.8983) [2021-04-16 09:59:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][1050/1251] eta 0:00:58 lr 0.000295 time 0.2797 (0.2911) loss 3.7167 (3.3746) grad_norm 1.8352 (1.8990) [2021-04-16 09:59:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][1060/1251] eta 0:00:55 lr 0.000295 time 0.2610 (0.2909) loss 3.3106 (3.3748) grad_norm 1.7414 (1.8990) [2021-04-16 09:59:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][1070/1251] eta 0:00:52 lr 0.000295 time 0.2427 (0.2908) loss 3.7818 (3.3747) grad_norm 1.6017 (1.8982) [2021-04-16 09:59:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][1080/1251] eta 0:00:49 lr 0.000295 time 0.2774 (0.2907) loss 4.1240 (3.3755) grad_norm 1.8139 (1.8985) [2021-04-16 09:59:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][1090/1251] eta 0:00:46 lr 0.000295 time 0.2733 (0.2906) loss 3.5218 (3.3751) grad_norm 1.7614 (1.8981) [2021-04-16 09:59:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][1100/1251] eta 0:00:43 lr 0.000295 time 0.2716 (0.2905) loss 3.3387 (3.3725) grad_norm 1.8180 (1.8988) [2021-04-16 09:59:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][1110/1251] eta 0:00:40 lr 0.000295 time 0.2593 (0.2904) loss 3.9802 (3.3725) grad_norm 1.8474 (1.8982) [2021-04-16 09:59:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][1120/1251] eta 0:00:38 lr 0.000295 time 0.2827 (0.2904) loss 3.8210 (3.3722) grad_norm 1.7846 (1.8984) [2021-04-16 09:59:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][1130/1251] eta 0:00:35 lr 0.000295 time 0.2843 (0.2903) loss 3.3751 (3.3724) grad_norm 1.8508 (1.8987) [2021-04-16 09:59:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][1140/1251] eta 0:00:32 lr 0.000295 time 0.2900 (0.2902) loss 3.4617 (3.3732) grad_norm 1.8184 (1.8993) [2021-04-16 10:00:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][1150/1251] eta 0:00:29 lr 0.000295 time 0.2663 (0.2903) loss 2.8220 (3.3714) grad_norm 1.6795 (1.8990) [2021-04-16 10:00:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][1160/1251] eta 0:00:26 lr 0.000295 time 0.2866 (0.2903) loss 2.4027 (3.3694) grad_norm 1.8362 (1.8987) [2021-04-16 10:00:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][1170/1251] eta 0:00:23 lr 0.000295 time 0.2875 (0.2901) loss 3.3547 (3.3712) grad_norm 1.8682 (1.8978) [2021-04-16 10:00:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][1180/1251] eta 0:00:20 lr 0.000295 time 0.2582 (0.2901) loss 3.0399 (3.3705) grad_norm 1.9298 (1.8970) [2021-04-16 10:00:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][1190/1251] eta 0:00:17 lr 0.000294 time 0.2777 (0.2899) loss 3.6160 (3.3673) grad_norm 1.6580 (1.8967) [2021-04-16 10:00:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][1200/1251] eta 0:00:14 lr 0.000294 time 0.2733 (0.2898) loss 2.6891 (3.3669) grad_norm 1.7824 (1.8956) [2021-04-16 10:00:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][1210/1251] eta 0:00:11 lr 0.000294 time 0.2816 (0.2897) loss 2.9014 (3.3660) grad_norm 1.8375 (1.8958) [2021-04-16 10:00:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][1220/1251] eta 0:00:08 lr 0.000294 time 0.2635 (0.2896) loss 3.6705 (3.3670) grad_norm 2.4217 (1.8968) [2021-04-16 10:00:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][1230/1251] eta 0:00:06 lr 0.000294 time 0.2706 (0.2896) loss 2.6239 (3.3683) grad_norm 1.8397 (1.8969) [2021-04-16 10:00:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][1240/1251] eta 0:00:03 lr 0.000294 time 0.2778 (0.2895) loss 3.2162 (3.3673) grad_norm 1.6545 (1.8967) [2021-04-16 10:00:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [191/300][1250/1251] eta 0:00:00 lr 0.000294 time 0.2594 (0.2893) loss 2.6472 (3.3670) grad_norm 1.8240 (1.8968) [2021-04-16 10:00:34 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 191 training takes 0:06:05 [2021-04-16 10:00:34 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_191.pth saving...... [2021-04-16 10:00:45 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_191.pth saved !!! [2021-04-16 10:00:46 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.131 (1.131) Loss 1.0263 (1.0263) Acc@1 74.805 (74.805) Acc@5 93.945 (93.945) [2021-04-16 10:00:47 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.105 (0.216) Loss 0.8776 (0.9396) Acc@1 78.613 (77.947) Acc@5 94.824 (94.176) [2021-04-16 10:00:49 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.157 (0.208) Loss 0.8883 (0.9378) Acc@1 78.418 (78.102) Acc@5 95.020 (94.127) [2021-04-16 10:00:52 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.441 (0.243) Loss 0.8996 (0.9393) Acc@1 78.027 (77.949) Acc@5 94.531 (94.109) [2021-04-16 10:00:53 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.075 (0.209) Loss 0.9466 (0.9364) Acc@1 77.051 (77.994) Acc@5 93.750 (94.164) [2021-04-16 10:01:03 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 77.910 Acc@5 94.152 [2021-04-16 10:01:03 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 77.9% [2021-04-16 10:01:03 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 77.91% [2021-04-16 10:01:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][0/1251] eta 2:14:06 lr 0.000294 time 6.4322 (6.4322) loss 4.0547 (4.0547) grad_norm 1.8911 (1.8911) [2021-04-16 10:01:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][10/1251] eta 0:17:27 lr 0.000294 time 0.4234 (0.8444) loss 2.8621 (3.4499) grad_norm 2.1152 (2.0356) [2021-04-16 10:01:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][20/1251] eta 0:11:45 lr 0.000294 time 0.2661 (0.5731) loss 3.6212 (3.5017) grad_norm 1.8646 (1.9597) [2021-04-16 10:01:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][30/1251] eta 0:09:42 lr 0.000294 time 0.2764 (0.4768) loss 2.1693 (3.4177) grad_norm 1.7839 (1.9250) [2021-04-16 10:01:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][40/1251] eta 0:08:42 lr 0.000294 time 0.2510 (0.4312) loss 3.2415 (3.4188) grad_norm 1.8477 (1.9219) [2021-04-16 10:01:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][50/1251] eta 0:08:00 lr 0.000294 time 0.2635 (0.4001) loss 2.2724 (3.3794) grad_norm 1.9311 (1.9144) [2021-04-16 10:01:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][60/1251] eta 0:07:32 lr 0.000294 time 0.2773 (0.3799) loss 2.9407 (3.4327) grad_norm 1.8159 (1.9077) [2021-04-16 10:01:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][70/1251] eta 0:07:11 lr 0.000294 time 0.2821 (0.3653) loss 3.8889 (3.4005) grad_norm 1.7232 (1.8910) [2021-04-16 10:01:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][80/1251] eta 0:06:55 lr 0.000294 time 0.2619 (0.3549) loss 2.9208 (3.4113) grad_norm 2.1500 (1.9032) [2021-04-16 10:01:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][90/1251] eta 0:06:42 lr 0.000294 time 0.2996 (0.3466) loss 4.0715 (3.4235) grad_norm 2.2546 (1.9038) [2021-04-16 10:01:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][100/1251] eta 0:06:30 lr 0.000294 time 0.2735 (0.3395) loss 2.8426 (3.3823) grad_norm 1.7929 (1.8947) [2021-04-16 10:01:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][110/1251] eta 0:06:20 lr 0.000294 time 0.2871 (0.3338) loss 3.7802 (3.3636) grad_norm 1.6803 (1.8937) [2021-04-16 10:01:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][120/1251] eta 0:06:12 lr 0.000294 time 0.2701 (0.3292) loss 2.5813 (3.3566) grad_norm 1.8307 (1.8939) [2021-04-16 10:01:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][130/1251] eta 0:06:04 lr 0.000294 time 0.2788 (0.3250) loss 3.2450 (3.3611) grad_norm 1.7274 (1.8909) [2021-04-16 10:01:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][140/1251] eta 0:05:59 lr 0.000294 time 0.2595 (0.3233) loss 3.8147 (3.3416) grad_norm 1.9144 (1.8903) [2021-04-16 10:01:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][150/1251] eta 0:05:53 lr 0.000294 time 0.2758 (0.3210) loss 3.7266 (3.3477) grad_norm 2.2076 (1.8862) [2021-04-16 10:01:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][160/1251] eta 0:05:47 lr 0.000294 time 0.2756 (0.3185) loss 2.4180 (3.3448) grad_norm 1.7306 (1.8849) [2021-04-16 10:01:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][170/1251] eta 0:05:41 lr 0.000294 time 0.2854 (0.3162) loss 3.5615 (3.3498) grad_norm 2.1855 (1.8952) [2021-04-16 10:02:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][180/1251] eta 0:05:35 lr 0.000294 time 0.2681 (0.3137) loss 3.7947 (3.3653) grad_norm 2.2588 (1.8972) [2021-04-16 10:02:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][190/1251] eta 0:05:30 lr 0.000294 time 0.2967 (0.3119) loss 2.8555 (3.3581) grad_norm 2.3552 (1.8971) [2021-04-16 10:02:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][200/1251] eta 0:05:25 lr 0.000293 time 0.2675 (0.3102) loss 3.7674 (3.3626) grad_norm 1.7932 (1.9081) [2021-04-16 10:02:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][210/1251] eta 0:05:21 lr 0.000293 time 0.2804 (0.3085) loss 3.0873 (3.3751) grad_norm 2.5317 (1.9122) [2021-04-16 10:02:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][220/1251] eta 0:05:16 lr 0.000293 time 0.2911 (0.3069) loss 3.8098 (3.3848) grad_norm 1.5616 (1.9101) [2021-04-16 10:02:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][230/1251] eta 0:05:12 lr 0.000293 time 0.2958 (0.3060) loss 3.7473 (3.3867) grad_norm 1.7641 (1.9115) [2021-04-16 10:02:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][240/1251] eta 0:05:08 lr 0.000293 time 0.2722 (0.3048) loss 3.9262 (3.3893) grad_norm 2.2111 (1.9106) [2021-04-16 10:02:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][250/1251] eta 0:05:03 lr 0.000293 time 0.2918 (0.3037) loss 2.4695 (3.3845) grad_norm 1.9723 (1.9114) [2021-04-16 10:02:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][260/1251] eta 0:05:00 lr 0.000293 time 0.2749 (0.3028) loss 3.4980 (3.3902) grad_norm 1.9727 (1.9096) [2021-04-16 10:02:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][270/1251] eta 0:04:55 lr 0.000293 time 0.2568 (0.3017) loss 3.1973 (3.3908) grad_norm 1.7666 (1.9053) [2021-04-16 10:02:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][280/1251] eta 0:04:52 lr 0.000293 time 0.2992 (0.3008) loss 3.9314 (3.3842) grad_norm 1.7938 (1.9017) [2021-04-16 10:02:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][290/1251] eta 0:04:48 lr 0.000293 time 0.2649 (0.3005) loss 3.8099 (3.3840) grad_norm 2.1569 (1.9036) [2021-04-16 10:02:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][300/1251] eta 0:04:45 lr 0.000293 time 0.2805 (0.2998) loss 4.1376 (3.3913) grad_norm 2.0299 (1.9042) [2021-04-16 10:02:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][310/1251] eta 0:04:41 lr 0.000293 time 0.2717 (0.2989) loss 1.9626 (3.3718) grad_norm 1.8798 (1.9030) [2021-04-16 10:02:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][320/1251] eta 0:04:37 lr 0.000293 time 0.2796 (0.2986) loss 2.1350 (3.3685) grad_norm 1.6192 (1.9041) [2021-04-16 10:02:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][330/1251] eta 0:04:34 lr 0.000293 time 0.2908 (0.2979) loss 4.2019 (3.3614) grad_norm 1.7279 (1.9037) [2021-04-16 10:02:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][340/1251] eta 0:04:30 lr 0.000293 time 0.3020 (0.2974) loss 3.0391 (3.3624) grad_norm 1.8845 (1.9045) [2021-04-16 10:02:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][350/1251] eta 0:04:27 lr 0.000293 time 0.2729 (0.2969) loss 3.2795 (3.3521) grad_norm 1.7277 (1.9009) [2021-04-16 10:02:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][360/1251] eta 0:04:24 lr 0.000293 time 0.2896 (0.2973) loss 3.9196 (3.3481) grad_norm 2.1869 (1.9015) [2021-04-16 10:02:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][370/1251] eta 0:04:21 lr 0.000293 time 0.3046 (0.2969) loss 3.6905 (3.3472) grad_norm 1.9142 (1.9015) [2021-04-16 10:02:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][380/1251] eta 0:04:18 lr 0.000293 time 0.2664 (0.2965) loss 3.6578 (3.3448) grad_norm 1.7696 (1.9020) [2021-04-16 10:02:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][390/1251] eta 0:04:14 lr 0.000293 time 0.2984 (0.2958) loss 3.6433 (3.3479) grad_norm 2.1218 (nan) [2021-04-16 10:03:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][400/1251] eta 0:04:11 lr 0.000293 time 0.2824 (0.2954) loss 3.8178 (3.3449) grad_norm 1.9338 (nan) [2021-04-16 10:03:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][410/1251] eta 0:04:07 lr 0.000293 time 0.2698 (0.2948) loss 3.7723 (3.3454) grad_norm 1.8201 (nan) [2021-04-16 10:03:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][420/1251] eta 0:04:04 lr 0.000293 time 0.2693 (0.2944) loss 3.0392 (3.3459) grad_norm 1.8948 (nan) [2021-04-16 10:03:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][430/1251] eta 0:04:01 lr 0.000293 time 0.3143 (0.2943) loss 3.6950 (3.3497) grad_norm 1.9873 (nan) [2021-04-16 10:03:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][440/1251] eta 0:03:58 lr 0.000293 time 0.2669 (0.2941) loss 2.9755 (3.3521) grad_norm 2.0334 (nan) [2021-04-16 10:03:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][450/1251] eta 0:03:55 lr 0.000293 time 0.2607 (0.2938) loss 3.3499 (3.3535) grad_norm 2.4303 (nan) [2021-04-16 10:03:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][460/1251] eta 0:03:52 lr 0.000293 time 0.2957 (0.2934) loss 3.6579 (3.3486) grad_norm 1.8459 (nan) [2021-04-16 10:03:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][470/1251] eta 0:03:49 lr 0.000292 time 0.2873 (0.2934) loss 3.2148 (3.3474) grad_norm 1.8675 (nan) [2021-04-16 10:03:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][480/1251] eta 0:03:46 lr 0.000292 time 0.2854 (0.2932) loss 2.3949 (3.3464) grad_norm 1.7289 (nan) [2021-04-16 10:03:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][490/1251] eta 0:03:42 lr 0.000292 time 0.2851 (0.2929) loss 3.4608 (3.3468) grad_norm 1.7080 (nan) [2021-04-16 10:03:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][500/1251] eta 0:03:39 lr 0.000292 time 0.2668 (0.2925) loss 3.8215 (3.3471) grad_norm 2.0427 (nan) [2021-04-16 10:03:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][510/1251] eta 0:03:36 lr 0.000292 time 0.2669 (0.2922) loss 3.5234 (3.3488) grad_norm 1.7293 (nan) [2021-04-16 10:03:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][520/1251] eta 0:03:33 lr 0.000292 time 0.2597 (0.2919) loss 4.3981 (3.3573) grad_norm 2.0082 (nan) [2021-04-16 10:03:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][530/1251] eta 0:03:30 lr 0.000292 time 0.2641 (0.2915) loss 4.0700 (3.3587) grad_norm 2.3150 (nan) [2021-04-16 10:03:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][540/1251] eta 0:03:27 lr 0.000292 time 0.2776 (0.2912) loss 2.5202 (3.3583) grad_norm 1.9015 (nan) [2021-04-16 10:03:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][550/1251] eta 0:03:23 lr 0.000292 time 0.2731 (0.2909) loss 2.5360 (3.3627) grad_norm 1.8000 (nan) [2021-04-16 10:03:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][560/1251] eta 0:03:20 lr 0.000292 time 0.2746 (0.2905) loss 2.8990 (3.3652) grad_norm 1.7093 (nan) [2021-04-16 10:03:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][570/1251] eta 0:03:17 lr 0.000292 time 0.2870 (0.2904) loss 3.7094 (3.3686) grad_norm 1.7254 (nan) [2021-04-16 10:03:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][580/1251] eta 0:03:15 lr 0.000292 time 0.2939 (0.2906) loss 3.6663 (3.3744) grad_norm 1.8701 (nan) [2021-04-16 10:03:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][590/1251] eta 0:03:12 lr 0.000292 time 0.2992 (0.2907) loss 3.1644 (3.3760) grad_norm 2.1827 (nan) [2021-04-16 10:03:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][600/1251] eta 0:03:09 lr 0.000292 time 0.2722 (0.2907) loss 4.1732 (3.3790) grad_norm 2.0092 (nan) [2021-04-16 10:04:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][610/1251] eta 0:03:06 lr 0.000292 time 0.3055 (0.2905) loss 3.6979 (3.3762) grad_norm 1.7009 (nan) [2021-04-16 10:04:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][620/1251] eta 0:03:03 lr 0.000292 time 0.2653 (0.2905) loss 3.5881 (3.3770) grad_norm 1.8800 (nan) [2021-04-16 10:04:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][630/1251] eta 0:03:00 lr 0.000292 time 0.2653 (0.2903) loss 3.8148 (3.3808) grad_norm 1.7202 (nan) [2021-04-16 10:04:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][640/1251] eta 0:02:57 lr 0.000292 time 0.2599 (0.2901) loss 3.7874 (3.3852) grad_norm 1.8543 (nan) [2021-04-16 10:04:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][650/1251] eta 0:02:54 lr 0.000292 time 0.2891 (0.2899) loss 4.2912 (3.3867) grad_norm 1.8183 (nan) [2021-04-16 10:04:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][660/1251] eta 0:02:51 lr 0.000292 time 0.2826 (0.2897) loss 2.7019 (3.3845) grad_norm 1.6386 (nan) [2021-04-16 10:04:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][670/1251] eta 0:02:48 lr 0.000292 time 0.2604 (0.2895) loss 3.5955 (3.3849) grad_norm 1.8851 (nan) [2021-04-16 10:04:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][680/1251] eta 0:02:45 lr 0.000292 time 0.2778 (0.2893) loss 2.4872 (3.3832) grad_norm 1.8490 (nan) [2021-04-16 10:04:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][690/1251] eta 0:02:42 lr 0.000292 time 0.2925 (0.2892) loss 2.9706 (3.3821) grad_norm 1.8179 (nan) [2021-04-16 10:04:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][700/1251] eta 0:02:39 lr 0.000292 time 0.2537 (0.2890) loss 2.1306 (3.3828) grad_norm 1.6248 (nan) [2021-04-16 10:04:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][710/1251] eta 0:02:36 lr 0.000292 time 0.3518 (0.2890) loss 3.8589 (3.3856) grad_norm 1.6635 (nan) [2021-04-16 10:04:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][720/1251] eta 0:02:33 lr 0.000292 time 0.2633 (0.2892) loss 3.4041 (3.3885) grad_norm 1.7605 (nan) [2021-04-16 10:04:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][730/1251] eta 0:02:30 lr 0.000292 time 0.2795 (0.2889) loss 2.5254 (3.3855) grad_norm 1.9500 (nan) [2021-04-16 10:04:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][740/1251] eta 0:02:27 lr 0.000291 time 0.2952 (0.2887) loss 3.5323 (3.3861) grad_norm 1.8875 (nan) [2021-04-16 10:04:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][750/1251] eta 0:02:24 lr 0.000291 time 0.2745 (0.2886) loss 3.8560 (3.3847) grad_norm 1.7331 (nan) [2021-04-16 10:04:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][760/1251] eta 0:02:21 lr 0.000291 time 0.2691 (0.2884) loss 3.5690 (3.3838) grad_norm 1.8295 (nan) [2021-04-16 10:04:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][770/1251] eta 0:02:18 lr 0.000291 time 0.2815 (0.2883) loss 3.4495 (3.3829) grad_norm 2.0632 (nan) [2021-04-16 10:04:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][780/1251] eta 0:02:15 lr 0.000291 time 0.2692 (0.2882) loss 3.0200 (3.3817) grad_norm 1.6974 (nan) [2021-04-16 10:04:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][790/1251] eta 0:02:12 lr 0.000291 time 0.3161 (0.2881) loss 2.1233 (3.3811) grad_norm 1.9255 (nan) [2021-04-16 10:04:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][800/1251] eta 0:02:09 lr 0.000291 time 0.2873 (0.2880) loss 3.6104 (3.3812) grad_norm 2.0098 (nan) [2021-04-16 10:04:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][810/1251] eta 0:02:06 lr 0.000291 time 0.2778 (0.2879) loss 2.2343 (3.3814) grad_norm 1.5803 (nan) [2021-04-16 10:05:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][820/1251] eta 0:02:04 lr 0.000291 time 0.2830 (0.2879) loss 2.6343 (3.3803) grad_norm 2.1098 (nan) [2021-04-16 10:05:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][830/1251] eta 0:02:01 lr 0.000291 time 0.2686 (0.2878) loss 3.2005 (3.3777) grad_norm 1.8362 (nan) [2021-04-16 10:05:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][840/1251] eta 0:01:58 lr 0.000291 time 0.2745 (0.2877) loss 3.3896 (3.3748) grad_norm 2.1303 (nan) [2021-04-16 10:05:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][850/1251] eta 0:01:55 lr 0.000291 time 0.2851 (0.2876) loss 3.5336 (3.3766) grad_norm 2.0511 (nan) [2021-04-16 10:05:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][860/1251] eta 0:01:52 lr 0.000291 time 0.3019 (0.2875) loss 2.8872 (3.3760) grad_norm 2.2719 (nan) [2021-04-16 10:05:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][870/1251] eta 0:01:49 lr 0.000291 time 0.2735 (0.2875) loss 3.7753 (3.3768) grad_norm 1.7132 (nan) [2021-04-16 10:05:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][880/1251] eta 0:01:46 lr 0.000291 time 0.2976 (0.2874) loss 4.0665 (3.3779) grad_norm 2.2158 (nan) [2021-04-16 10:05:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][890/1251] eta 0:01:43 lr 0.000291 time 0.2459 (0.2874) loss 3.3751 (3.3784) grad_norm 1.6235 (nan) [2021-04-16 10:05:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][900/1251] eta 0:01:40 lr 0.000291 time 0.2671 (0.2873) loss 3.6294 (3.3744) grad_norm 1.9419 (nan) [2021-04-16 10:05:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][910/1251] eta 0:01:37 lr 0.000291 time 0.2790 (0.2872) loss 3.8470 (3.3782) grad_norm 1.7385 (nan) [2021-04-16 10:05:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][920/1251] eta 0:01:35 lr 0.000291 time 0.2459 (0.2871) loss 3.4834 (3.3755) grad_norm 1.7336 (nan) [2021-04-16 10:05:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][930/1251] eta 0:01:32 lr 0.000291 time 0.2762 (0.2871) loss 3.4958 (3.3771) grad_norm 1.9872 (nan) [2021-04-16 10:05:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][940/1251] eta 0:01:29 lr 0.000291 time 0.2809 (0.2871) loss 3.0265 (3.3775) grad_norm 1.6783 (nan) [2021-04-16 10:05:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][950/1251] eta 0:01:26 lr 0.000291 time 0.2904 (0.2870) loss 3.0734 (3.3747) grad_norm 1.8453 (nan) [2021-04-16 10:05:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][960/1251] eta 0:01:23 lr 0.000291 time 0.3037 (0.2869) loss 2.5813 (3.3741) grad_norm 2.0568 (nan) [2021-04-16 10:05:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][970/1251] eta 0:01:20 lr 0.000291 time 0.2651 (0.2868) loss 3.2092 (3.3738) grad_norm 1.8879 (nan) [2021-04-16 10:05:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][980/1251] eta 0:01:17 lr 0.000291 time 0.2950 (0.2868) loss 3.1277 (3.3726) grad_norm 2.2326 (nan) [2021-04-16 10:05:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][990/1251] eta 0:01:14 lr 0.000291 time 0.2836 (0.2867) loss 3.8136 (3.3739) grad_norm 1.7663 (nan) [2021-04-16 10:05:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][1000/1251] eta 0:01:11 lr 0.000290 time 0.2858 (0.2866) loss 4.0063 (3.3739) grad_norm 1.6562 (nan) [2021-04-16 10:05:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][1010/1251] eta 0:01:09 lr 0.000290 time 0.2623 (0.2865) loss 3.9464 (3.3730) grad_norm 2.0545 (nan) [2021-04-16 10:05:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][1020/1251] eta 0:01:06 lr 0.000290 time 0.2725 (0.2865) loss 2.3455 (3.3700) grad_norm 1.8117 (nan) [2021-04-16 10:05:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][1030/1251] eta 0:01:03 lr 0.000290 time 0.2797 (0.2864) loss 4.0042 (3.3691) grad_norm 2.0278 (nan) [2021-04-16 10:06:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][1040/1251] eta 0:01:00 lr 0.000290 time 0.2798 (0.2865) loss 3.4929 (3.3704) grad_norm 1.7379 (nan) [2021-04-16 10:06:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][1050/1251] eta 0:00:57 lr 0.000290 time 0.3975 (0.2865) loss 3.1054 (3.3706) grad_norm 1.8124 (nan) [2021-04-16 10:06:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][1060/1251] eta 0:00:54 lr 0.000290 time 0.2740 (0.2864) loss 3.3504 (3.3704) grad_norm 1.7338 (nan) [2021-04-16 10:06:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][1070/1251] eta 0:00:51 lr 0.000290 time 0.2850 (0.2863) loss 3.5893 (3.3712) grad_norm 1.7464 (nan) [2021-04-16 10:06:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][1080/1251] eta 0:00:48 lr 0.000290 time 0.2772 (0.2862) loss 3.6438 (3.3708) grad_norm 1.9599 (nan) [2021-04-16 10:06:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][1090/1251] eta 0:00:46 lr 0.000290 time 0.2905 (0.2862) loss 3.4499 (3.3703) grad_norm 1.6966 (nan) [2021-04-16 10:06:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][1100/1251] eta 0:00:43 lr 0.000290 time 0.2901 (0.2862) loss 3.0192 (3.3695) grad_norm 1.7033 (nan) [2021-04-16 10:06:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][1110/1251] eta 0:00:40 lr 0.000290 time 0.2833 (0.2861) loss 3.9348 (3.3704) grad_norm 1.8141 (nan) [2021-04-16 10:06:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][1120/1251] eta 0:00:37 lr 0.000290 time 0.2696 (0.2860) loss 2.2525 (3.3696) grad_norm 1.6566 (nan) [2021-04-16 10:06:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][1130/1251] eta 0:00:34 lr 0.000290 time 0.2592 (0.2859) loss 4.2151 (3.3689) grad_norm 1.9888 (nan) [2021-04-16 10:06:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][1140/1251] eta 0:00:31 lr 0.000290 time 0.2775 (0.2857) loss 3.6817 (3.3703) grad_norm 1.7941 (nan) [2021-04-16 10:06:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][1150/1251] eta 0:00:28 lr 0.000290 time 0.2715 (0.2858) loss 3.7902 (3.3685) grad_norm 1.9102 (nan) [2021-04-16 10:06:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][1160/1251] eta 0:00:26 lr 0.000290 time 0.2857 (0.2858) loss 4.1093 (3.3701) grad_norm 2.0267 (nan) [2021-04-16 10:06:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][1170/1251] eta 0:00:23 lr 0.000290 time 0.2650 (0.2859) loss 3.0736 (3.3699) grad_norm 2.1430 (nan) [2021-04-16 10:06:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][1180/1251] eta 0:00:20 lr 0.000290 time 0.2963 (0.2859) loss 3.2858 (3.3688) grad_norm 1.8650 (nan) [2021-04-16 10:06:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][1190/1251] eta 0:00:17 lr 0.000290 time 0.2813 (0.2858) loss 3.2517 (3.3682) grad_norm 1.8810 (nan) [2021-04-16 10:06:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][1200/1251] eta 0:00:14 lr 0.000290 time 0.2758 (0.2859) loss 3.4362 (3.3678) grad_norm 1.9312 (nan) [2021-04-16 10:06:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][1210/1251] eta 0:00:11 lr 0.000290 time 0.2675 (0.2858) loss 3.2465 (3.3666) grad_norm 1.8167 (nan) [2021-04-16 10:06:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][1220/1251] eta 0:00:08 lr 0.000290 time 0.2802 (0.2858) loss 3.3557 (3.3682) grad_norm 1.9030 (nan) [2021-04-16 10:06:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][1230/1251] eta 0:00:05 lr 0.000290 time 0.2725 (0.2857) loss 3.7734 (3.3677) grad_norm 1.6667 (nan) [2021-04-16 10:06:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][1240/1251] eta 0:00:03 lr 0.000290 time 0.2486 (0.2855) loss 3.2698 (3.3682) grad_norm 1.8718 (nan) [2021-04-16 10:07:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [192/300][1250/1251] eta 0:00:00 lr 0.000290 time 0.2487 (0.2852) loss 3.0875 (3.3675) grad_norm 1.9172 (nan) [2021-04-16 10:07:04 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 192 training takes 0:06:00 [2021-04-16 10:07:04 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_192.pth saving...... [2021-04-16 10:07:16 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_192.pth saved !!! [2021-04-16 10:07:17 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.237 (1.237) Loss 0.9440 (0.9440) Acc@1 78.320 (78.320) Acc@5 94.238 (94.238) [2021-04-16 10:07:18 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.242 (0.227) Loss 0.9170 (0.9413) Acc@1 78.516 (77.823) Acc@5 94.336 (94.176) [2021-04-16 10:07:21 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.164 (0.229) Loss 0.9106 (0.9434) Acc@1 79.297 (77.790) Acc@5 94.629 (94.150) [2021-04-16 10:07:23 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.128 (0.234) Loss 0.9222 (0.9461) Acc@1 77.832 (77.854) Acc@5 93.945 (94.131) [2021-04-16 10:07:25 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.217) Loss 0.8992 (0.9440) Acc@1 79.102 (77.946) Acc@5 94.336 (94.181) [2021-04-16 10:07:31 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 77.938 Acc@5 94.160 [2021-04-16 10:07:31 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 77.9% [2021-04-16 10:07:31 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 77.94% [2021-04-16 10:07:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][0/1251] eta 5:09:41 lr 0.000290 time 14.8532 (14.8532) loss 2.6493 (2.6493) grad_norm 2.2825 (2.2825) [2021-04-16 10:07:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][10/1251] eta 0:33:03 lr 0.000290 time 0.2701 (1.5986) loss 3.4359 (3.2425) grad_norm 1.9374 (2.0478) [2021-04-16 10:07:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][20/1251] eta 0:19:52 lr 0.000289 time 0.2740 (0.9688) loss 3.8935 (3.3246) grad_norm 1.8535 (1.9909) [2021-04-16 10:07:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][30/1251] eta 0:15:12 lr 0.000289 time 0.2946 (0.7472) loss 3.3958 (3.2648) grad_norm 1.8410 (1.9552) [2021-04-16 10:07:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][40/1251] eta 0:12:46 lr 0.000289 time 0.2835 (0.6333) loss 3.9410 (3.2610) grad_norm 1.9707 (1.9388) [2021-04-16 10:07:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][50/1251] eta 0:11:16 lr 0.000289 time 0.2750 (0.5636) loss 3.1928 (3.2556) grad_norm 1.8988 (1.9218) [2021-04-16 10:08:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][60/1251] eta 0:10:19 lr 0.000289 time 0.4524 (0.5202) loss 4.1209 (3.2600) grad_norm 1.9578 (1.9161) [2021-04-16 10:08:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][70/1251] eta 0:09:33 lr 0.000289 time 0.2777 (0.4859) loss 3.7045 (3.2865) grad_norm 2.2946 (1.9092) [2021-04-16 10:08:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][80/1251] eta 0:08:58 lr 0.000289 time 0.2895 (0.4598) loss 3.3775 (3.2919) grad_norm 2.2107 (1.9056) [2021-04-16 10:08:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][90/1251] eta 0:08:32 lr 0.000289 time 0.2743 (0.4416) loss 3.3521 (3.3160) grad_norm 2.0320 (1.9148) [2021-04-16 10:08:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][100/1251] eta 0:08:09 lr 0.000289 time 0.2720 (0.4256) loss 4.3344 (3.3144) grad_norm 2.0912 (1.9241) [2021-04-16 10:08:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][110/1251] eta 0:07:50 lr 0.000289 time 0.2665 (0.4121) loss 2.9492 (3.3172) grad_norm 2.1345 (1.9282) [2021-04-16 10:08:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][120/1251] eta 0:07:34 lr 0.000289 time 0.2790 (0.4017) loss 3.7940 (3.3380) grad_norm 2.1662 (1.9352) [2021-04-16 10:08:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][130/1251] eta 0:07:20 lr 0.000289 time 0.2822 (0.3925) loss 3.6054 (3.3517) grad_norm 2.3224 (nan) [2021-04-16 10:08:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][140/1251] eta 0:07:09 lr 0.000289 time 0.2574 (0.3867) loss 3.8630 (3.3525) grad_norm 1.9782 (nan) [2021-04-16 10:08:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][150/1251] eta 0:06:58 lr 0.000289 time 0.2936 (0.3797) loss 3.6801 (3.3579) grad_norm 2.1782 (nan) [2021-04-16 10:08:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][160/1251] eta 0:06:47 lr 0.000289 time 0.2886 (0.3735) loss 2.9033 (3.3548) grad_norm 1.9617 (nan) [2021-04-16 10:08:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][170/1251] eta 0:06:38 lr 0.000289 time 0.2941 (0.3682) loss 2.6695 (3.3552) grad_norm 1.9969 (nan) [2021-04-16 10:08:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][180/1251] eta 0:06:28 lr 0.000289 time 0.2588 (0.3631) loss 3.8022 (3.3699) grad_norm 1.9808 (nan) [2021-04-16 10:08:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][190/1251] eta 0:06:20 lr 0.000289 time 0.2624 (0.3586) loss 4.0064 (3.3668) grad_norm 2.0479 (nan) [2021-04-16 10:08:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][200/1251] eta 0:06:12 lr 0.000289 time 0.2883 (0.3546) loss 3.2323 (3.3692) grad_norm 1.7631 (nan) [2021-04-16 10:08:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][210/1251] eta 0:06:05 lr 0.000289 time 0.2498 (0.3508) loss 3.4542 (3.3692) grad_norm 1.6310 (nan) [2021-04-16 10:08:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][220/1251] eta 0:05:58 lr 0.000289 time 0.2700 (0.3481) loss 3.5935 (3.3587) grad_norm 2.1947 (nan) [2021-04-16 10:08:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][230/1251] eta 0:05:52 lr 0.000289 time 0.2755 (0.3457) loss 4.0711 (3.3497) grad_norm 1.9322 (nan) [2021-04-16 10:08:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][240/1251] eta 0:05:46 lr 0.000289 time 0.2764 (0.3428) loss 3.7705 (3.3389) grad_norm 1.9023 (nan) [2021-04-16 10:08:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][250/1251] eta 0:05:40 lr 0.000289 time 0.2646 (0.3406) loss 2.7665 (3.3331) grad_norm 2.1415 (nan) [2021-04-16 10:08:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][260/1251] eta 0:05:35 lr 0.000289 time 0.2819 (0.3381) loss 3.3117 (3.3355) grad_norm 2.0852 (nan) [2021-04-16 10:09:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][270/1251] eta 0:05:29 lr 0.000289 time 0.2990 (0.3359) loss 3.7696 (3.3442) grad_norm 1.7082 (nan) [2021-04-16 10:09:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][280/1251] eta 0:05:23 lr 0.000289 time 0.2580 (0.3336) loss 2.5682 (3.3449) grad_norm 1.7239 (nan) [2021-04-16 10:09:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][290/1251] eta 0:05:18 lr 0.000288 time 0.2925 (0.3319) loss 3.9954 (3.3461) grad_norm 1.9354 (nan) [2021-04-16 10:09:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][300/1251] eta 0:05:14 lr 0.000288 time 0.2743 (0.3302) loss 3.5777 (3.3483) grad_norm 1.8376 (nan) [2021-04-16 10:09:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][310/1251] eta 0:05:09 lr 0.000288 time 0.2816 (0.3287) loss 3.8486 (3.3561) grad_norm 2.2069 (nan) [2021-04-16 10:09:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][320/1251] eta 0:05:04 lr 0.000288 time 0.2674 (0.3270) loss 4.0159 (3.3605) grad_norm 1.8108 (nan) [2021-04-16 10:09:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][330/1251] eta 0:04:59 lr 0.000288 time 0.2943 (0.3257) loss 3.9966 (3.3675) grad_norm 1.7642 (nan) [2021-04-16 10:09:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][340/1251] eta 0:04:55 lr 0.000288 time 0.2749 (0.3243) loss 3.6828 (3.3736) grad_norm 1.7770 (nan) [2021-04-16 10:09:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][350/1251] eta 0:04:51 lr 0.000288 time 0.4233 (0.3234) loss 3.3614 (3.3693) grad_norm 1.9509 (nan) [2021-04-16 10:09:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][360/1251] eta 0:04:47 lr 0.000288 time 0.2641 (0.3227) loss 2.9829 (3.3665) grad_norm 2.7396 (nan) [2021-04-16 10:09:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][370/1251] eta 0:04:43 lr 0.000288 time 0.2644 (0.3222) loss 3.2166 (3.3594) grad_norm 1.8194 (nan) [2021-04-16 10:09:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][380/1251] eta 0:04:39 lr 0.000288 time 0.2917 (0.3211) loss 3.7143 (3.3601) grad_norm 1.8615 (nan) [2021-04-16 10:09:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][390/1251] eta 0:04:35 lr 0.000288 time 0.3081 (0.3202) loss 2.3192 (3.3510) grad_norm 1.9993 (nan) [2021-04-16 10:09:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][400/1251] eta 0:04:31 lr 0.000288 time 0.2852 (0.3191) loss 2.9088 (3.3496) grad_norm 3.2848 (nan) [2021-04-16 10:09:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][410/1251] eta 0:04:27 lr 0.000288 time 0.2786 (0.3181) loss 3.1443 (3.3581) grad_norm 1.7042 (nan) [2021-04-16 10:09:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][420/1251] eta 0:04:23 lr 0.000288 time 0.2596 (0.3170) loss 3.8248 (3.3630) grad_norm 1.8666 (nan) [2021-04-16 10:09:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][430/1251] eta 0:04:19 lr 0.000288 time 0.2918 (0.3163) loss 1.9591 (3.3552) grad_norm 1.7627 (nan) [2021-04-16 10:09:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][440/1251] eta 0:04:15 lr 0.000288 time 0.2724 (0.3154) loss 2.8199 (3.3559) grad_norm 1.8372 (nan) [2021-04-16 10:09:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][450/1251] eta 0:04:11 lr 0.000288 time 0.2648 (0.3145) loss 3.2793 (3.3583) grad_norm 1.8488 (nan) [2021-04-16 10:09:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][460/1251] eta 0:04:08 lr 0.000288 time 0.2817 (0.3137) loss 3.7030 (3.3638) grad_norm 1.9527 (nan) [2021-04-16 10:09:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][470/1251] eta 0:04:04 lr 0.000288 time 0.2755 (0.3129) loss 3.3367 (3.3637) grad_norm 1.6274 (nan) [2021-04-16 10:10:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][480/1251] eta 0:04:00 lr 0.000288 time 0.2997 (0.3121) loss 2.0992 (3.3601) grad_norm 1.7537 (nan) [2021-04-16 10:10:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][490/1251] eta 0:03:56 lr 0.000288 time 0.2812 (0.3114) loss 2.6679 (3.3556) grad_norm 1.7945 (nan) [2021-04-16 10:10:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][500/1251] eta 0:03:53 lr 0.000288 time 0.2649 (0.3106) loss 1.9672 (3.3569) grad_norm 1.7021 (nan) [2021-04-16 10:10:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][510/1251] eta 0:03:49 lr 0.000288 time 0.2691 (0.3099) loss 3.7676 (3.3573) grad_norm 1.7820 (nan) [2021-04-16 10:10:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][520/1251] eta 0:03:46 lr 0.000288 time 0.2844 (0.3093) loss 3.2003 (3.3521) grad_norm 2.1083 (nan) [2021-04-16 10:10:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][530/1251] eta 0:03:42 lr 0.000288 time 0.3003 (0.3087) loss 3.6737 (3.3558) grad_norm 1.7543 (nan) [2021-04-16 10:10:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][540/1251] eta 0:03:39 lr 0.000288 time 0.2533 (0.3082) loss 3.2346 (3.3543) grad_norm 1.8777 (nan) [2021-04-16 10:10:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][550/1251] eta 0:03:35 lr 0.000288 time 0.2803 (0.3077) loss 2.3826 (3.3524) grad_norm 2.1212 (nan) [2021-04-16 10:10:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][560/1251] eta 0:03:32 lr 0.000287 time 0.2544 (0.3071) loss 3.6642 (3.3476) grad_norm 1.7943 (nan) [2021-04-16 10:10:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][570/1251] eta 0:03:28 lr 0.000287 time 0.2678 (0.3067) loss 3.5265 (3.3466) grad_norm 2.0493 (nan) [2021-04-16 10:10:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][580/1251] eta 0:03:25 lr 0.000287 time 0.2690 (0.3066) loss 2.7494 (3.3483) grad_norm 1.7986 (nan) [2021-04-16 10:10:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][590/1251] eta 0:03:22 lr 0.000287 time 0.2827 (0.3061) loss 2.2760 (3.3484) grad_norm 2.0012 (nan) [2021-04-16 10:10:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][600/1251] eta 0:03:19 lr 0.000287 time 0.2879 (0.3058) loss 3.6817 (3.3482) grad_norm 1.8643 (nan) [2021-04-16 10:10:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][610/1251] eta 0:03:15 lr 0.000287 time 0.2541 (0.3053) loss 3.4416 (3.3530) grad_norm 1.8900 (nan) [2021-04-16 10:10:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][620/1251] eta 0:03:12 lr 0.000287 time 0.2906 (0.3049) loss 3.2206 (3.3524) grad_norm 1.8791 (nan) [2021-04-16 10:10:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][630/1251] eta 0:03:09 lr 0.000287 time 0.2755 (0.3045) loss 2.8738 (3.3507) grad_norm 1.8716 (nan) [2021-04-16 10:10:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][640/1251] eta 0:03:05 lr 0.000287 time 0.2702 (0.3044) loss 2.8042 (3.3520) grad_norm 2.0812 (nan) [2021-04-16 10:10:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][650/1251] eta 0:03:02 lr 0.000287 time 0.2683 (0.3040) loss 3.2739 (3.3509) grad_norm 1.9059 (nan) [2021-04-16 10:10:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][660/1251] eta 0:02:59 lr 0.000287 time 0.2769 (0.3036) loss 2.9775 (3.3482) grad_norm 2.1567 (nan) [2021-04-16 10:10:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][670/1251] eta 0:02:56 lr 0.000287 time 0.2922 (0.3034) loss 3.5653 (3.3505) grad_norm 2.2691 (nan) [2021-04-16 10:10:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][680/1251] eta 0:02:53 lr 0.000287 time 0.2720 (0.3031) loss 3.9533 (3.3530) grad_norm 1.8116 (nan) [2021-04-16 10:11:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][690/1251] eta 0:02:49 lr 0.000287 time 0.2591 (0.3028) loss 3.1815 (3.3509) grad_norm 1.8745 (nan) [2021-04-16 10:11:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][700/1251] eta 0:02:46 lr 0.000287 time 0.2735 (0.3024) loss 3.0989 (3.3524) grad_norm 1.7947 (nan) [2021-04-16 10:11:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][710/1251] eta 0:02:43 lr 0.000287 time 0.2953 (0.3022) loss 4.0617 (3.3553) grad_norm 2.1684 (nan) [2021-04-16 10:11:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][720/1251] eta 0:02:40 lr 0.000287 time 0.2766 (0.3021) loss 3.1172 (3.3572) grad_norm 1.6619 (nan) [2021-04-16 10:11:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][730/1251] eta 0:02:37 lr 0.000287 time 0.2602 (0.3017) loss 3.0543 (3.3568) grad_norm 2.1780 (nan) [2021-04-16 10:11:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][740/1251] eta 0:02:34 lr 0.000287 time 0.2874 (0.3014) loss 3.6881 (3.3558) grad_norm 1.7444 (nan) [2021-04-16 10:11:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][750/1251] eta 0:02:30 lr 0.000287 time 0.2953 (0.3011) loss 3.7803 (3.3555) grad_norm 2.0316 (nan) [2021-04-16 10:11:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][760/1251] eta 0:02:27 lr 0.000287 time 0.2667 (0.3007) loss 3.8482 (3.3556) grad_norm 1.9845 (nan) [2021-04-16 10:11:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][770/1251] eta 0:02:24 lr 0.000287 time 0.2670 (0.3004) loss 3.9161 (3.3564) grad_norm 2.0270 (nan) [2021-04-16 10:11:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][780/1251] eta 0:02:21 lr 0.000287 time 0.2854 (0.3002) loss 3.2959 (3.3604) grad_norm 2.0446 (nan) [2021-04-16 10:11:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][790/1251] eta 0:02:18 lr 0.000287 time 0.2730 (0.3000) loss 3.8401 (3.3605) grad_norm 2.0854 (nan) [2021-04-16 10:11:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][800/1251] eta 0:02:15 lr 0.000287 time 0.2750 (0.2997) loss 3.3841 (3.3627) grad_norm 2.0407 (nan) [2021-04-16 10:11:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][810/1251] eta 0:02:12 lr 0.000287 time 0.2837 (0.2996) loss 2.3808 (3.3582) grad_norm 1.6735 (nan) [2021-04-16 10:11:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][820/1251] eta 0:02:08 lr 0.000287 time 0.2754 (0.2993) loss 4.3110 (3.3611) grad_norm 1.9882 (nan) [2021-04-16 10:11:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][830/1251] eta 0:02:05 lr 0.000286 time 0.2948 (0.2992) loss 3.4784 (3.3641) grad_norm 1.7451 (nan) [2021-04-16 10:11:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][840/1251] eta 0:02:02 lr 0.000286 time 0.2821 (0.2989) loss 3.6756 (3.3652) grad_norm 2.1134 (nan) [2021-04-16 10:11:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][850/1251] eta 0:01:59 lr 0.000286 time 0.2814 (0.2987) loss 3.7640 (3.3657) grad_norm 1.6236 (nan) [2021-04-16 10:11:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][860/1251] eta 0:01:56 lr 0.000286 time 0.2689 (0.2984) loss 3.8640 (3.3673) grad_norm 1.7070 (nan) [2021-04-16 10:11:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][870/1251] eta 0:01:53 lr 0.000286 time 0.2664 (0.2981) loss 2.9688 (3.3661) grad_norm 2.0854 (nan) [2021-04-16 10:11:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][880/1251] eta 0:01:50 lr 0.000286 time 0.3035 (0.2980) loss 3.3398 (3.3679) grad_norm 2.0432 (nan) [2021-04-16 10:11:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][890/1251] eta 0:01:47 lr 0.000286 time 0.2677 (0.2978) loss 3.4703 (3.3687) grad_norm 2.0807 (nan) [2021-04-16 10:11:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][900/1251] eta 0:01:44 lr 0.000286 time 0.2705 (0.2975) loss 3.7964 (3.3700) grad_norm 1.6857 (nan) [2021-04-16 10:12:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][910/1251] eta 0:01:41 lr 0.000286 time 0.2682 (0.2973) loss 3.0422 (3.3676) grad_norm 2.1885 (nan) [2021-04-16 10:12:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][920/1251] eta 0:01:38 lr 0.000286 time 0.2786 (0.2971) loss 3.2736 (3.3645) grad_norm 1.8815 (nan) [2021-04-16 10:12:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][930/1251] eta 0:01:35 lr 0.000286 time 0.2668 (0.2970) loss 2.5191 (3.3600) grad_norm 2.0413 (nan) [2021-04-16 10:12:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][940/1251] eta 0:01:32 lr 0.000286 time 0.4266 (0.2973) loss 3.3381 (3.3616) grad_norm 1.8060 (nan) [2021-04-16 10:12:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][950/1251] eta 0:01:29 lr 0.000286 time 0.2896 (0.2970) loss 3.8072 (3.3634) grad_norm 1.9669 (nan) [2021-04-16 10:12:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][960/1251] eta 0:01:26 lr 0.000286 time 0.2721 (0.2968) loss 3.4020 (3.3661) grad_norm 1.7285 (nan) [2021-04-16 10:12:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][970/1251] eta 0:01:23 lr 0.000286 time 0.2670 (0.2966) loss 4.1828 (3.3645) grad_norm 1.8746 (nan) [2021-04-16 10:12:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][980/1251] eta 0:01:20 lr 0.000286 time 0.2865 (0.2964) loss 2.1908 (3.3619) grad_norm 2.0056 (nan) [2021-04-16 10:12:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][990/1251] eta 0:01:17 lr 0.000286 time 0.2916 (0.2962) loss 3.9826 (3.3603) grad_norm 2.1030 (nan) [2021-04-16 10:12:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][1000/1251] eta 0:01:14 lr 0.000286 time 0.2725 (0.2960) loss 2.6486 (3.3618) grad_norm 1.7297 (nan) [2021-04-16 10:12:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][1010/1251] eta 0:01:11 lr 0.000286 time 0.3068 (0.2959) loss 3.4357 (3.3627) grad_norm 1.9192 (nan) [2021-04-16 10:12:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][1020/1251] eta 0:01:08 lr 0.000286 time 0.2740 (0.2957) loss 2.6025 (3.3605) grad_norm 1.9506 (nan) [2021-04-16 10:12:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][1030/1251] eta 0:01:05 lr 0.000286 time 0.2915 (0.2955) loss 3.6260 (3.3633) grad_norm 1.7617 (nan) [2021-04-16 10:12:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][1040/1251] eta 0:01:02 lr 0.000286 time 0.2855 (0.2954) loss 3.7848 (3.3643) grad_norm 1.8506 (nan) [2021-04-16 10:12:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][1050/1251] eta 0:00:59 lr 0.000286 time 0.2541 (0.2952) loss 3.3580 (3.3646) grad_norm 1.8538 (nan) [2021-04-16 10:12:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][1060/1251] eta 0:00:56 lr 0.000286 time 0.2772 (0.2950) loss 3.6408 (3.3642) grad_norm 2.7981 (nan) [2021-04-16 10:12:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][1070/1251] eta 0:00:53 lr 0.000286 time 0.2810 (0.2949) loss 3.3257 (3.3635) grad_norm 1.9474 (nan) [2021-04-16 10:12:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][1080/1251] eta 0:00:50 lr 0.000286 time 0.2860 (0.2947) loss 3.5419 (3.3647) grad_norm 1.8812 (nan) [2021-04-16 10:12:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][1090/1251] eta 0:00:47 lr 0.000286 time 0.2800 (0.2946) loss 3.9107 (3.3637) grad_norm 2.1069 (nan) [2021-04-16 10:12:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][1100/1251] eta 0:00:44 lr 0.000285 time 0.2818 (0.2944) loss 3.8362 (3.3652) grad_norm 1.9064 (nan) [2021-04-16 10:12:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][1110/1251] eta 0:00:41 lr 0.000285 time 0.3056 (0.2942) loss 3.0336 (3.3661) grad_norm 1.9114 (nan) [2021-04-16 10:13:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][1120/1251] eta 0:00:38 lr 0.000285 time 0.2705 (0.2941) loss 3.5070 (3.3640) grad_norm 1.9357 (nan) [2021-04-16 10:13:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][1130/1251] eta 0:00:35 lr 0.000285 time 0.2699 (0.2939) loss 2.6680 (3.3646) grad_norm 2.0762 (nan) [2021-04-16 10:13:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][1140/1251] eta 0:00:32 lr 0.000285 time 0.2550 (0.2937) loss 2.4431 (3.3648) grad_norm 1.7514 (nan) [2021-04-16 10:13:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][1150/1251] eta 0:00:29 lr 0.000285 time 0.2804 (0.2936) loss 3.2538 (3.3659) grad_norm 1.8633 (nan) [2021-04-16 10:13:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][1160/1251] eta 0:00:26 lr 0.000285 time 0.2774 (0.2937) loss 2.2796 (3.3660) grad_norm 1.7792 (nan) [2021-04-16 10:13:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][1170/1251] eta 0:00:23 lr 0.000285 time 0.2682 (0.2935) loss 2.3512 (3.3645) grad_norm 1.7259 (nan) [2021-04-16 10:13:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][1180/1251] eta 0:00:20 lr 0.000285 time 0.2693 (0.2934) loss 3.8495 (3.3655) grad_norm 1.8197 (nan) [2021-04-16 10:13:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][1190/1251] eta 0:00:17 lr 0.000285 time 0.2540 (0.2933) loss 4.1884 (3.3663) grad_norm 1.9207 (nan) [2021-04-16 10:13:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][1200/1251] eta 0:00:14 lr 0.000285 time 0.2822 (0.2932) loss 3.6148 (3.3658) grad_norm 1.8318 (nan) [2021-04-16 10:13:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][1210/1251] eta 0:00:12 lr 0.000285 time 0.2747 (0.2930) loss 2.4861 (3.3635) grad_norm 1.9880 (nan) [2021-04-16 10:13:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][1220/1251] eta 0:00:09 lr 0.000285 time 0.2720 (0.2930) loss 3.5524 (3.3655) grad_norm 1.9505 (nan) [2021-04-16 10:13:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][1230/1251] eta 0:00:06 lr 0.000285 time 0.2600 (0.2929) loss 2.2387 (3.3642) grad_norm 1.9580 (nan) [2021-04-16 10:13:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][1240/1251] eta 0:00:03 lr 0.000285 time 0.3829 (0.2927) loss 3.4927 (3.3649) grad_norm 1.8918 (nan) [2021-04-16 10:13:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [193/300][1250/1251] eta 0:00:00 lr 0.000285 time 0.2471 (0.2924) loss 3.1383 (3.3642) grad_norm 1.7016 (nan) [2021-04-16 10:13:39 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 193 training takes 0:06:08 [2021-04-16 10:13:39 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_193.pth saving...... [2021-04-16 10:13:50 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_193.pth saved !!! [2021-04-16 10:13:51 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.092 (1.092) Loss 0.8889 (0.8889) Acc@1 78.809 (78.809) Acc@5 95.605 (95.605) [2021-04-16 10:13:52 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.101 (0.212) Loss 0.9281 (0.9431) Acc@1 78.711 (77.850) Acc@5 94.434 (94.611) [2021-04-16 10:13:55 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.112 (0.240) Loss 0.9888 (0.9426) Acc@1 75.000 (77.897) Acc@5 93.359 (94.336) [2021-04-16 10:13:56 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.115 (0.210) Loss 0.8612 (0.9305) Acc@1 79.785 (78.251) Acc@5 95.020 (94.383) [2021-04-16 10:13:59 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.217) Loss 1.0326 (0.9353) Acc@1 76.562 (78.070) Acc@5 93.066 (94.307) [2021-04-16 10:14:12 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 78.080 Acc@5 94.272 [2021-04-16 10:14:12 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 78.1% [2021-04-16 10:14:12 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 78.08% [2021-04-16 10:14:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][0/1251] eta 3:17:46 lr 0.000285 time 9.4857 (9.4857) loss 3.4915 (3.4915) grad_norm 2.0993 (2.0993) [2021-04-16 10:14:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][10/1251] eta 0:22:55 lr 0.000285 time 0.2669 (1.1080) loss 3.5829 (3.4419) grad_norm 1.6356 (1.8428) [2021-04-16 10:14:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][20/1251] eta 0:14:44 lr 0.000285 time 0.3153 (0.7182) loss 3.3302 (3.3554) grad_norm 2.5460 (1.9015) [2021-04-16 10:14:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][30/1251] eta 0:11:50 lr 0.000285 time 0.3108 (0.5822) loss 3.9290 (3.4306) grad_norm 1.9755 (1.9015) [2021-04-16 10:14:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][40/1251] eta 0:10:18 lr 0.000285 time 0.2591 (0.5106) loss 3.7181 (3.4265) grad_norm 1.8875 (1.9222) [2021-04-16 10:14:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][50/1251] eta 0:09:18 lr 0.000285 time 0.2534 (0.4652) loss 2.6526 (3.3722) grad_norm 1.7746 (1.9136) [2021-04-16 10:14:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][60/1251] eta 0:08:37 lr 0.000285 time 0.2991 (0.4346) loss 3.3783 (3.3807) grad_norm 1.6655 (1.9017) [2021-04-16 10:14:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][70/1251] eta 0:08:07 lr 0.000285 time 0.2737 (0.4127) loss 3.9870 (3.4034) grad_norm 2.0836 (1.8918) [2021-04-16 10:14:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][80/1251] eta 0:07:43 lr 0.000285 time 0.2769 (0.3958) loss 2.2466 (3.3694) grad_norm 1.7759 (1.9035) [2021-04-16 10:14:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][90/1251] eta 0:07:24 lr 0.000285 time 0.2685 (0.3827) loss 3.4507 (3.3518) grad_norm 1.7123 (1.9053) [2021-04-16 10:14:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][100/1251] eta 0:07:08 lr 0.000285 time 0.2709 (0.3724) loss 2.7436 (3.3108) grad_norm 1.7483 (1.9048) [2021-04-16 10:14:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][110/1251] eta 0:06:55 lr 0.000284 time 0.2711 (0.3639) loss 3.3741 (3.3112) grad_norm 2.0113 (1.8983) [2021-04-16 10:14:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][120/1251] eta 0:06:44 lr 0.000284 time 0.3423 (0.3577) loss 3.7827 (3.3227) grad_norm 2.1349 (1.8946) [2021-04-16 10:14:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][130/1251] eta 0:06:35 lr 0.000284 time 0.3041 (0.3531) loss 3.2672 (3.3201) grad_norm 1.6650 (1.8931) [2021-04-16 10:15:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][140/1251] eta 0:06:27 lr 0.000284 time 0.2626 (0.3485) loss 3.0722 (3.3330) grad_norm 1.6311 (1.8927) [2021-04-16 10:15:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][150/1251] eta 0:06:19 lr 0.000284 time 0.2722 (0.3447) loss 3.8543 (3.3509) grad_norm 2.0028 (1.8941) [2021-04-16 10:15:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][160/1251] eta 0:06:11 lr 0.000284 time 0.2766 (0.3405) loss 3.4696 (3.3584) grad_norm 1.6723 (1.8940) [2021-04-16 10:15:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][170/1251] eta 0:06:03 lr 0.000284 time 0.2618 (0.3367) loss 3.5817 (3.3527) grad_norm 1.8718 (1.8912) [2021-04-16 10:15:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][180/1251] eta 0:05:57 lr 0.000284 time 0.2989 (0.3334) loss 3.1654 (3.3485) grad_norm 2.0833 (1.8910) [2021-04-16 10:15:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][190/1251] eta 0:05:50 lr 0.000284 time 0.2713 (0.3305) loss 3.7538 (3.3382) grad_norm 1.8329 (1.8943) [2021-04-16 10:15:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][200/1251] eta 0:05:44 lr 0.000284 time 0.2517 (0.3278) loss 2.3517 (3.3401) grad_norm 1.7651 (1.8906) [2021-04-16 10:15:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][210/1251] eta 0:05:38 lr 0.000284 time 0.2754 (0.3253) loss 3.8996 (3.3456) grad_norm 1.6791 (1.8896) [2021-04-16 10:15:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][220/1251] eta 0:05:33 lr 0.000284 time 0.2860 (0.3231) loss 2.7995 (3.3498) grad_norm 1.6803 (1.8868) [2021-04-16 10:15:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][230/1251] eta 0:05:27 lr 0.000284 time 0.2711 (0.3211) loss 3.5894 (3.3533) grad_norm 1.8032 (1.8824) [2021-04-16 10:15:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][240/1251] eta 0:05:22 lr 0.000284 time 0.2734 (0.3193) loss 2.3020 (3.3480) grad_norm 1.8723 (1.8883) [2021-04-16 10:15:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][250/1251] eta 0:05:18 lr 0.000284 time 0.2951 (0.3178) loss 3.6459 (3.3486) grad_norm 2.2088 (1.8897) [2021-04-16 10:15:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][260/1251] eta 0:05:13 lr 0.000284 time 0.2740 (0.3163) loss 3.0551 (3.3467) grad_norm 1.9385 (1.8922) [2021-04-16 10:15:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][270/1251] eta 0:05:08 lr 0.000284 time 0.2940 (0.3149) loss 2.8709 (3.3437) grad_norm 1.8250 (1.8925) [2021-04-16 10:15:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][280/1251] eta 0:05:04 lr 0.000284 time 0.2896 (0.3134) loss 3.8942 (3.3341) grad_norm 1.8477 (1.8886) [2021-04-16 10:15:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][290/1251] eta 0:04:59 lr 0.000284 time 0.2861 (0.3120) loss 4.0065 (3.3403) grad_norm 2.0510 (1.8898) [2021-04-16 10:15:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][300/1251] eta 0:04:55 lr 0.000284 time 0.2909 (0.3108) loss 2.6536 (3.3370) grad_norm 1.6223 (1.8871) [2021-04-16 10:15:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][310/1251] eta 0:04:51 lr 0.000284 time 0.2873 (0.3097) loss 2.5941 (3.3313) grad_norm 1.7245 (1.8874) [2021-04-16 10:15:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][320/1251] eta 0:04:47 lr 0.000284 time 0.2956 (0.3087) loss 3.6236 (3.3325) grad_norm 2.0595 (1.8911) [2021-04-16 10:15:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][330/1251] eta 0:04:43 lr 0.000284 time 0.3142 (0.3078) loss 3.5647 (3.3309) grad_norm 1.9312 (1.8940) [2021-04-16 10:15:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][340/1251] eta 0:04:39 lr 0.000284 time 0.2624 (0.3068) loss 3.6847 (3.3389) grad_norm 1.7477 (1.8931) [2021-04-16 10:15:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][350/1251] eta 0:04:36 lr 0.000284 time 0.2829 (0.3065) loss 3.5560 (3.3355) grad_norm 1.7366 (1.8945) [2021-04-16 10:16:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][360/1251] eta 0:04:32 lr 0.000284 time 0.2882 (0.3059) loss 2.4777 (3.3365) grad_norm 1.9156 (1.8932) [2021-04-16 10:16:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][370/1251] eta 0:04:29 lr 0.000284 time 0.2801 (0.3056) loss 2.6847 (3.3364) grad_norm 1.8436 (1.8915) [2021-04-16 10:16:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][380/1251] eta 0:04:25 lr 0.000283 time 0.2523 (0.3047) loss 2.8514 (3.3307) grad_norm 1.8314 (1.8891) [2021-04-16 10:16:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][390/1251] eta 0:04:21 lr 0.000283 time 0.2949 (0.3040) loss 2.9335 (3.3309) grad_norm 2.0020 (1.8882) [2021-04-16 10:16:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][400/1251] eta 0:04:18 lr 0.000283 time 0.2761 (0.3033) loss 2.6516 (3.3278) grad_norm 1.9827 (1.8866) [2021-04-16 10:16:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][410/1251] eta 0:04:14 lr 0.000283 time 0.2827 (0.3026) loss 3.3421 (3.3306) grad_norm 2.2571 (1.8860) [2021-04-16 10:16:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][420/1251] eta 0:04:11 lr 0.000283 time 0.2704 (0.3022) loss 3.7237 (3.3322) grad_norm 2.0964 (1.8865) [2021-04-16 10:16:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][430/1251] eta 0:04:07 lr 0.000283 time 0.2685 (0.3016) loss 3.5020 (3.3326) grad_norm 1.8654 (1.8902) [2021-04-16 10:16:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][440/1251] eta 0:04:04 lr 0.000283 time 0.2878 (0.3011) loss 3.6227 (3.3332) grad_norm 2.0115 (1.8918) [2021-04-16 10:16:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][450/1251] eta 0:04:01 lr 0.000283 time 0.4714 (0.3009) loss 3.6323 (3.3350) grad_norm 2.3160 (1.8925) [2021-04-16 10:16:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][460/1251] eta 0:03:57 lr 0.000283 time 0.2763 (0.3003) loss 3.5496 (3.3377) grad_norm 1.8095 (1.8934) [2021-04-16 10:16:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][470/1251] eta 0:03:54 lr 0.000283 time 0.2677 (0.2999) loss 3.4890 (3.3343) grad_norm 1.9555 (1.8916) [2021-04-16 10:16:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][480/1251] eta 0:03:50 lr 0.000283 time 0.2764 (0.2995) loss 2.2837 (3.3318) grad_norm 1.8939 (1.8936) [2021-04-16 10:16:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][490/1251] eta 0:03:47 lr 0.000283 time 0.2815 (0.2990) loss 3.9192 (3.3316) grad_norm 1.8178 (1.8914) [2021-04-16 10:16:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][500/1251] eta 0:03:44 lr 0.000283 time 0.2817 (0.2986) loss 3.7641 (3.3318) grad_norm 2.1628 (1.8913) [2021-04-16 10:16:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][510/1251] eta 0:03:40 lr 0.000283 time 0.2622 (0.2982) loss 3.8920 (3.3353) grad_norm 1.8490 (1.8923) [2021-04-16 10:16:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][520/1251] eta 0:03:37 lr 0.000283 time 0.2608 (0.2977) loss 3.8430 (3.3377) grad_norm 1.7713 (1.8913) [2021-04-16 10:16:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][530/1251] eta 0:03:34 lr 0.000283 time 0.2995 (0.2974) loss 2.3126 (3.3352) grad_norm 1.7447 (1.8921) [2021-04-16 10:16:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][540/1251] eta 0:03:31 lr 0.000283 time 0.2684 (0.2970) loss 4.0932 (3.3359) grad_norm 2.2635 (1.8913) [2021-04-16 10:16:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][550/1251] eta 0:03:27 lr 0.000283 time 0.2690 (0.2965) loss 3.2146 (3.3377) grad_norm 2.0798 (1.8911) [2021-04-16 10:16:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][560/1251] eta 0:03:24 lr 0.000283 time 0.2901 (0.2962) loss 3.6026 (3.3369) grad_norm 1.6985 (1.8918) [2021-04-16 10:17:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][570/1251] eta 0:03:21 lr 0.000283 time 0.2707 (0.2959) loss 3.4010 (3.3379) grad_norm 1.8095 (1.8921) [2021-04-16 10:17:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][580/1251] eta 0:03:18 lr 0.000283 time 0.2738 (0.2961) loss 3.6257 (3.3400) grad_norm 1.7515 (1.8929) [2021-04-16 10:17:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][590/1251] eta 0:03:15 lr 0.000283 time 0.2490 (0.2959) loss 2.9038 (3.3410) grad_norm 1.9539 (1.8928) [2021-04-16 10:17:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][600/1251] eta 0:03:12 lr 0.000283 time 0.3970 (0.2959) loss 3.5960 (3.3405) grad_norm 2.2981 (1.8923) [2021-04-16 10:17:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][610/1251] eta 0:03:09 lr 0.000283 time 0.2878 (0.2956) loss 2.9172 (3.3449) grad_norm 1.6252 (1.8938) [2021-04-16 10:17:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][620/1251] eta 0:03:06 lr 0.000283 time 0.3075 (0.2956) loss 3.5110 (3.3451) grad_norm 1.9551 (1.8938) [2021-04-16 10:17:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][630/1251] eta 0:03:03 lr 0.000283 time 0.2765 (0.2953) loss 3.4901 (3.3465) grad_norm 2.0760 (1.8961) [2021-04-16 10:17:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][640/1251] eta 0:03:00 lr 0.000283 time 0.2795 (0.2949) loss 2.8142 (3.3453) grad_norm 1.8713 (1.8951) [2021-04-16 10:17:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][650/1251] eta 0:02:57 lr 0.000282 time 0.2727 (0.2946) loss 4.0507 (3.3494) grad_norm 2.0524 (1.8962) [2021-04-16 10:17:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][660/1251] eta 0:02:53 lr 0.000282 time 0.2762 (0.2944) loss 2.6662 (3.3469) grad_norm 1.7874 (1.8967) [2021-04-16 10:17:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][670/1251] eta 0:02:50 lr 0.000282 time 0.2735 (0.2941) loss 4.3675 (3.3487) grad_norm 2.3258 (1.8973) [2021-04-16 10:17:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][680/1251] eta 0:02:47 lr 0.000282 time 0.2604 (0.2938) loss 2.9673 (3.3485) grad_norm 2.0612 (1.8972) [2021-04-16 10:17:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][690/1251] eta 0:02:44 lr 0.000282 time 0.2998 (0.2936) loss 2.6446 (3.3484) grad_norm 2.0137 (1.8997) [2021-04-16 10:17:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][700/1251] eta 0:02:41 lr 0.000282 time 0.2449 (0.2935) loss 3.4235 (3.3467) grad_norm 1.8269 (1.8991) [2021-04-16 10:17:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][710/1251] eta 0:02:38 lr 0.000282 time 0.2718 (0.2934) loss 2.1231 (3.3434) grad_norm 1.9783 (1.8995) [2021-04-16 10:17:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][720/1251] eta 0:02:35 lr 0.000282 time 0.2869 (0.2934) loss 3.8738 (3.3427) grad_norm 2.1620 (1.9006) [2021-04-16 10:17:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][730/1251] eta 0:02:32 lr 0.000282 time 0.2657 (0.2933) loss 3.9448 (3.3451) grad_norm 2.1141 (1.9018) [2021-04-16 10:17:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][740/1251] eta 0:02:29 lr 0.000282 time 0.2580 (0.2932) loss 2.8864 (3.3435) grad_norm 2.0769 (1.9027) [2021-04-16 10:17:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][750/1251] eta 0:02:26 lr 0.000282 time 0.2574 (0.2930) loss 4.3921 (3.3453) grad_norm 2.2744 (1.9049) [2021-04-16 10:17:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][760/1251] eta 0:02:23 lr 0.000282 time 0.2777 (0.2928) loss 3.3314 (3.3474) grad_norm 1.8086 (1.9051) [2021-04-16 10:17:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][770/1251] eta 0:02:20 lr 0.000282 time 0.2510 (0.2927) loss 2.3868 (3.3461) grad_norm 1.8133 (1.9054) [2021-04-16 10:18:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][780/1251] eta 0:02:17 lr 0.000282 time 0.2692 (0.2925) loss 2.6010 (3.3465) grad_norm 1.7410 (1.9064) [2021-04-16 10:18:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][790/1251] eta 0:02:14 lr 0.000282 time 0.3111 (0.2924) loss 2.3747 (3.3414) grad_norm 1.8000 (1.9063) [2021-04-16 10:18:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][800/1251] eta 0:02:11 lr 0.000282 time 0.2827 (0.2922) loss 3.5349 (3.3433) grad_norm 1.9144 (1.9063) [2021-04-16 10:18:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][810/1251] eta 0:02:08 lr 0.000282 time 0.2653 (0.2920) loss 3.2701 (3.3421) grad_norm 1.7494 (1.9062) [2021-04-16 10:18:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][820/1251] eta 0:02:05 lr 0.000282 time 0.2672 (0.2918) loss 3.3159 (3.3418) grad_norm 1.9289 (1.9073) [2021-04-16 10:18:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][830/1251] eta 0:02:02 lr 0.000282 time 0.2869 (0.2916) loss 2.7965 (3.3421) grad_norm 1.9970 (1.9071) [2021-04-16 10:18:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][840/1251] eta 0:01:59 lr 0.000282 time 0.3063 (0.2915) loss 3.4335 (3.3399) grad_norm 1.7123 (1.9085) [2021-04-16 10:18:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][850/1251] eta 0:01:56 lr 0.000282 time 0.2814 (0.2913) loss 3.2996 (3.3372) grad_norm 1.6699 (1.9083) [2021-04-16 10:18:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][860/1251] eta 0:01:53 lr 0.000282 time 0.2611 (0.2912) loss 2.6638 (3.3371) grad_norm 1.9112 (1.9084) [2021-04-16 10:18:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][870/1251] eta 0:01:50 lr 0.000282 time 0.2770 (0.2910) loss 2.5659 (3.3341) grad_norm 1.9126 (1.9073) [2021-04-16 10:18:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][880/1251] eta 0:01:47 lr 0.000282 time 0.2960 (0.2909) loss 4.0975 (3.3369) grad_norm 1.8925 (1.9074) [2021-04-16 10:18:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][890/1251] eta 0:01:44 lr 0.000282 time 0.3054 (0.2907) loss 3.1303 (3.3378) grad_norm 2.5289 (1.9101) [2021-04-16 10:18:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][900/1251] eta 0:01:42 lr 0.000282 time 0.2730 (0.2906) loss 3.9314 (3.3364) grad_norm 1.9028 (1.9111) [2021-04-16 10:18:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][910/1251] eta 0:01:39 lr 0.000282 time 0.2982 (0.2905) loss 2.9691 (3.3371) grad_norm 1.7435 (1.9117) [2021-04-16 10:18:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][920/1251] eta 0:01:36 lr 0.000281 time 0.2846 (0.2903) loss 2.7055 (3.3368) grad_norm 1.6407 (1.9121) [2021-04-16 10:18:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][930/1251] eta 0:01:33 lr 0.000281 time 0.2623 (0.2905) loss 3.6972 (3.3382) grad_norm 1.9290 (1.9127) [2021-04-16 10:18:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][940/1251] eta 0:01:30 lr 0.000281 time 0.2744 (0.2905) loss 2.7794 (3.3356) grad_norm 2.0182 (1.9150) [2021-04-16 10:18:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][950/1251] eta 0:01:27 lr 0.000281 time 0.2689 (0.2903) loss 2.8551 (3.3369) grad_norm 1.8413 (1.9163) [2021-04-16 10:18:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][960/1251] eta 0:01:24 lr 0.000281 time 0.2833 (0.2902) loss 2.4214 (3.3374) grad_norm 2.5431 (1.9185) [2021-04-16 10:18:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][970/1251] eta 0:01:21 lr 0.000281 time 0.2479 (0.2901) loss 2.7741 (3.3378) grad_norm 1.8529 (1.9192) [2021-04-16 10:18:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][980/1251] eta 0:01:18 lr 0.000281 time 0.2842 (0.2899) loss 3.7537 (3.3417) grad_norm 2.4134 (1.9204) [2021-04-16 10:18:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][990/1251] eta 0:01:15 lr 0.000281 time 0.2598 (0.2899) loss 3.5207 (3.3441) grad_norm 1.6715 (1.9199) [2021-04-16 10:19:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][1000/1251] eta 0:01:12 lr 0.000281 time 0.2930 (0.2898) loss 3.5964 (3.3435) grad_norm 2.0693 (1.9208) [2021-04-16 10:19:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][1010/1251] eta 0:01:09 lr 0.000281 time 0.2740 (0.2897) loss 2.3059 (3.3399) grad_norm 1.7959 (1.9215) [2021-04-16 10:19:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][1020/1251] eta 0:01:06 lr 0.000281 time 0.2810 (0.2896) loss 2.5873 (3.3389) grad_norm 1.7240 (1.9204) [2021-04-16 10:19:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][1030/1251] eta 0:01:03 lr 0.000281 time 0.2988 (0.2896) loss 2.5137 (3.3406) grad_norm 1.7933 (1.9206) [2021-04-16 10:19:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][1040/1251] eta 0:01:01 lr 0.000281 time 0.2609 (0.2894) loss 2.6837 (3.3360) grad_norm 2.1049 (1.9214) [2021-04-16 10:19:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][1050/1251] eta 0:00:58 lr 0.000281 time 0.2936 (0.2893) loss 3.3825 (3.3360) grad_norm 1.8018 (1.9220) [2021-04-16 10:19:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][1060/1251] eta 0:00:55 lr 0.000281 time 0.2768 (0.2893) loss 3.4076 (3.3367) grad_norm 1.6925 (1.9215) [2021-04-16 10:19:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][1070/1251] eta 0:00:52 lr 0.000281 time 0.2712 (0.2892) loss 3.3439 (3.3378) grad_norm 1.7831 (1.9203) [2021-04-16 10:19:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][1080/1251] eta 0:00:49 lr 0.000281 time 0.2788 (0.2890) loss 2.7042 (3.3393) grad_norm 1.5903 (1.9197) [2021-04-16 10:19:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][1090/1251] eta 0:00:46 lr 0.000281 time 0.2638 (0.2889) loss 3.5653 (3.3389) grad_norm 1.8152 (1.9195) [2021-04-16 10:19:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][1100/1251] eta 0:00:43 lr 0.000281 time 0.2751 (0.2888) loss 4.0143 (3.3403) grad_norm 1.7751 (1.9193) [2021-04-16 10:19:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][1110/1251] eta 0:00:40 lr 0.000281 time 0.2833 (0.2887) loss 3.6964 (3.3410) grad_norm 1.9722 (1.9202) [2021-04-16 10:19:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][1120/1251] eta 0:00:37 lr 0.000281 time 0.2764 (0.2886) loss 4.0688 (3.3412) grad_norm 2.1019 (1.9200) [2021-04-16 10:19:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][1130/1251] eta 0:00:34 lr 0.000281 time 0.2728 (0.2885) loss 3.3588 (3.3401) grad_norm 2.0530 (1.9200) [2021-04-16 10:19:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][1140/1251] eta 0:00:32 lr 0.000281 time 0.2644 (0.2884) loss 3.5963 (3.3397) grad_norm 1.8357 (1.9195) [2021-04-16 10:19:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][1150/1251] eta 0:00:29 lr 0.000281 time 0.2876 (0.2883) loss 3.6133 (3.3405) grad_norm 2.1319 (1.9195) [2021-04-16 10:19:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][1160/1251] eta 0:00:26 lr 0.000281 time 0.2543 (0.2884) loss 3.4255 (3.3400) grad_norm 2.2165 (1.9188) [2021-04-16 10:19:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][1170/1251] eta 0:00:23 lr 0.000281 time 0.2762 (0.2885) loss 3.6612 (3.3368) grad_norm 1.9020 (1.9191) [2021-04-16 10:19:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][1180/1251] eta 0:00:20 lr 0.000281 time 0.2865 (0.2884) loss 2.6994 (3.3356) grad_norm 2.0356 (1.9198) [2021-04-16 10:19:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][1190/1251] eta 0:00:17 lr 0.000280 time 0.2605 (0.2884) loss 3.4982 (3.3351) grad_norm 1.9407 (1.9200) [2021-04-16 10:19:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][1200/1251] eta 0:00:14 lr 0.000280 time 0.2511 (0.2883) loss 3.8374 (3.3371) grad_norm 2.0163 (1.9193) [2021-04-16 10:20:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][1210/1251] eta 0:00:11 lr 0.000280 time 0.2844 (0.2882) loss 2.5025 (3.3360) grad_norm 1.8069 (1.9187) [2021-04-16 10:20:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][1220/1251] eta 0:00:08 lr 0.000280 time 0.2977 (0.2882) loss 3.7067 (3.3373) grad_norm 2.0921 (1.9180) [2021-04-16 10:20:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][1230/1251] eta 0:00:06 lr 0.000280 time 0.2677 (0.2881) loss 2.4062 (3.3347) grad_norm 1.9217 (1.9179) [2021-04-16 10:20:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][1240/1251] eta 0:00:03 lr 0.000280 time 0.2506 (0.2879) loss 2.1524 (3.3336) grad_norm 1.8281 (1.9175) [2021-04-16 10:20:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [194/300][1250/1251] eta 0:00:00 lr 0.000280 time 0.2486 (0.2876) loss 3.5111 (3.3351) grad_norm 1.8728 (1.9171) [2021-04-16 10:20:15 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 194 training takes 0:06:03 [2021-04-16 10:20:15 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_194.pth saving...... [2021-04-16 10:20:29 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_194.pth saved !!! [2021-04-16 10:20:30 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.220 (1.220) Loss 0.9065 (0.9065) Acc@1 79.297 (79.297) Acc@5 94.238 (94.238) [2021-04-16 10:20:31 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.115 (0.233) Loss 0.9370 (0.9174) Acc@1 78.027 (78.400) Acc@5 94.434 (94.478) [2021-04-16 10:20:34 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.865 (0.244) Loss 0.9142 (0.9215) Acc@1 78.125 (78.241) Acc@5 94.824 (94.368) [2021-04-16 10:20:36 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.238 (0.225) Loss 0.9641 (0.9291) Acc@1 76.562 (78.008) Acc@5 93.848 (94.260) [2021-04-16 10:20:37 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.208) Loss 0.8309 (0.9273) Acc@1 80.859 (78.092) Acc@5 95.215 (94.305) [2021-04-16 10:20:42 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 78.040 Acc@5 94.312 [2021-04-16 10:20:42 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 78.0% [2021-04-16 10:20:42 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 78.08% [2021-04-16 10:20:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][0/1251] eta 2:12:33 lr 0.000280 time 6.3580 (6.3580) loss 3.3966 (3.3966) grad_norm 1.9276 (1.9276) [2021-04-16 10:20:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][10/1251] eta 0:17:15 lr 0.000280 time 0.2816 (0.8344) loss 3.3937 (3.2273) grad_norm 1.7695 (2.0232) [2021-04-16 10:20:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][20/1251] eta 0:11:43 lr 0.000280 time 0.2910 (0.5714) loss 2.5082 (3.2975) grad_norm 1.8293 (1.9975) [2021-04-16 10:20:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][30/1251] eta 0:09:42 lr 0.000280 time 0.2692 (0.4770) loss 2.9817 (3.3331) grad_norm 2.1435 (1.9849) [2021-04-16 10:21:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][40/1251] eta 0:08:39 lr 0.000280 time 0.2831 (0.4288) loss 3.6526 (3.3358) grad_norm 2.3533 (1.9772) [2021-04-16 10:21:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][50/1251] eta 0:08:02 lr 0.000280 time 0.2973 (0.4014) loss 2.8020 (3.3125) grad_norm 2.1442 (1.9857) [2021-04-16 10:21:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][60/1251] eta 0:07:34 lr 0.000280 time 0.2609 (0.3813) loss 3.8366 (3.2840) grad_norm 2.1955 (1.9866) [2021-04-16 10:21:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][70/1251] eta 0:07:13 lr 0.000280 time 0.3059 (0.3674) loss 3.3944 (3.2935) grad_norm 1.9274 (1.9797) [2021-04-16 10:21:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][80/1251] eta 0:06:57 lr 0.000280 time 0.2650 (0.3565) loss 3.8078 (3.3104) grad_norm 2.0638 (1.9686) [2021-04-16 10:21:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][90/1251] eta 0:06:44 lr 0.000280 time 0.3020 (0.3487) loss 3.4283 (3.2798) grad_norm 1.6479 (1.9542) [2021-04-16 10:21:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][100/1251] eta 0:06:32 lr 0.000280 time 0.2665 (0.3413) loss 2.8200 (3.2940) grad_norm 1.8766 (1.9403) [2021-04-16 10:21:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][110/1251] eta 0:06:23 lr 0.000280 time 0.2857 (0.3358) loss 3.9019 (3.2952) grad_norm 2.0825 (1.9351) [2021-04-16 10:21:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][120/1251] eta 0:06:14 lr 0.000280 time 0.2762 (0.3314) loss 2.8721 (3.2775) grad_norm 1.6699 (1.9244) [2021-04-16 10:21:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][130/1251] eta 0:06:08 lr 0.000280 time 0.2714 (0.3290) loss 2.9512 (3.2701) grad_norm 1.8118 (1.9148) [2021-04-16 10:21:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][140/1251] eta 0:06:03 lr 0.000280 time 0.2973 (0.3275) loss 3.8663 (3.2750) grad_norm 1.9595 (1.9135) [2021-04-16 10:21:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][150/1251] eta 0:05:57 lr 0.000280 time 0.2568 (0.3246) loss 2.6899 (3.2803) grad_norm 1.7840 (1.9126) [2021-04-16 10:21:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][160/1251] eta 0:05:51 lr 0.000280 time 0.2933 (0.3222) loss 3.4453 (3.2711) grad_norm 2.2435 (1.9250) [2021-04-16 10:21:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][170/1251] eta 0:05:45 lr 0.000280 time 0.2788 (0.3196) loss 3.7625 (3.2724) grad_norm 1.9457 (1.9255) [2021-04-16 10:21:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][180/1251] eta 0:05:39 lr 0.000280 time 0.2571 (0.3175) loss 4.0758 (3.2872) grad_norm 2.3997 (1.9248) [2021-04-16 10:21:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][190/1251] eta 0:05:34 lr 0.000280 time 0.2735 (0.3153) loss 3.5928 (3.2857) grad_norm 1.8194 (1.9269) [2021-04-16 10:21:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][200/1251] eta 0:05:30 lr 0.000280 time 0.2854 (0.3140) loss 2.9298 (3.2977) grad_norm 1.9366 (1.9315) [2021-04-16 10:21:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][210/1251] eta 0:05:25 lr 0.000279 time 0.2701 (0.3124) loss 3.4499 (3.2946) grad_norm 1.7607 (1.9323) [2021-04-16 10:21:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][220/1251] eta 0:05:20 lr 0.000279 time 0.2931 (0.3112) loss 3.8133 (3.3022) grad_norm 1.7169 (1.9335) [2021-04-16 10:21:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][230/1251] eta 0:05:16 lr 0.000279 time 0.2774 (0.3098) loss 3.6300 (3.3026) grad_norm 1.9640 (1.9339) [2021-04-16 10:21:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][240/1251] eta 0:05:11 lr 0.000279 time 0.2607 (0.3083) loss 2.2427 (3.2897) grad_norm 2.2632 (1.9420) [2021-04-16 10:22:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][250/1251] eta 0:05:08 lr 0.000279 time 0.2677 (0.3078) loss 4.0401 (3.2941) grad_norm 1.8796 (1.9399) [2021-04-16 10:22:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][260/1251] eta 0:05:03 lr 0.000279 time 0.2793 (0.3067) loss 2.7796 (3.2943) grad_norm 1.9970 (1.9371) [2021-04-16 10:22:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][270/1251] eta 0:04:59 lr 0.000279 time 0.2711 (0.3056) loss 3.5165 (3.2934) grad_norm 2.1199 (1.9413) [2021-04-16 10:22:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][280/1251] eta 0:04:55 lr 0.000279 time 0.2746 (0.3045) loss 3.4523 (3.2939) grad_norm 1.6108 (1.9388) [2021-04-16 10:22:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][290/1251] eta 0:04:51 lr 0.000279 time 0.2723 (0.3035) loss 3.8825 (3.2907) grad_norm 2.3940 (1.9364) [2021-04-16 10:22:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][300/1251] eta 0:04:47 lr 0.000279 time 0.3006 (0.3028) loss 3.2886 (3.2956) grad_norm 1.7496 (1.9324) [2021-04-16 10:22:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][310/1251] eta 0:04:44 lr 0.000279 time 0.2673 (0.3019) loss 3.7080 (3.2930) grad_norm 1.8469 (1.9292) [2021-04-16 10:22:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][320/1251] eta 0:04:40 lr 0.000279 time 0.2760 (0.3012) loss 2.2265 (3.2809) grad_norm 1.9652 (1.9290) [2021-04-16 10:22:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][330/1251] eta 0:04:36 lr 0.000279 time 0.2654 (0.3006) loss 3.0683 (3.2948) grad_norm 1.9381 (1.9265) [2021-04-16 10:22:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][340/1251] eta 0:04:33 lr 0.000279 time 0.2977 (0.3002) loss 3.7005 (3.2961) grad_norm 2.1660 (1.9284) [2021-04-16 10:22:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][350/1251] eta 0:04:29 lr 0.000279 time 0.2624 (0.2995) loss 3.2714 (3.3017) grad_norm 2.3543 (1.9270) [2021-04-16 10:22:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][360/1251] eta 0:04:26 lr 0.000279 time 0.2946 (0.2993) loss 3.2003 (3.3064) grad_norm 1.6980 (1.9257) [2021-04-16 10:22:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][370/1251] eta 0:04:23 lr 0.000279 time 0.2666 (0.2993) loss 3.7212 (3.3083) grad_norm 2.0292 (1.9281) [2021-04-16 10:22:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][380/1251] eta 0:04:20 lr 0.000279 time 0.2825 (0.2990) loss 3.2905 (3.3069) grad_norm 2.3366 (1.9268) [2021-04-16 10:22:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][390/1251] eta 0:04:17 lr 0.000279 time 0.2546 (0.2986) loss 3.8890 (3.3109) grad_norm 1.9049 (1.9249) [2021-04-16 10:22:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][400/1251] eta 0:04:13 lr 0.000279 time 0.2729 (0.2980) loss 3.7932 (3.3183) grad_norm 1.8003 (1.9243) [2021-04-16 10:22:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][410/1251] eta 0:04:10 lr 0.000279 time 0.2772 (0.2974) loss 3.3263 (3.3173) grad_norm 1.8621 (1.9242) [2021-04-16 10:22:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][420/1251] eta 0:04:06 lr 0.000279 time 0.2892 (0.2971) loss 3.9701 (3.3199) grad_norm 1.9630 (1.9247) [2021-04-16 10:22:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][430/1251] eta 0:04:03 lr 0.000279 time 0.3092 (0.2967) loss 3.8868 (3.3235) grad_norm 1.6227 (1.9224) [2021-04-16 10:22:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][440/1251] eta 0:04:00 lr 0.000279 time 0.2776 (0.2964) loss 4.1870 (3.3213) grad_norm 2.0121 (1.9225) [2021-04-16 10:22:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][450/1251] eta 0:03:56 lr 0.000279 time 0.2881 (0.2959) loss 2.7331 (3.3302) grad_norm 1.7821 (1.9232) [2021-04-16 10:22:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][460/1251] eta 0:03:53 lr 0.000279 time 0.2796 (0.2956) loss 3.9041 (3.3298) grad_norm 1.6806 (1.9253) [2021-04-16 10:23:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][470/1251] eta 0:03:50 lr 0.000279 time 0.2803 (0.2951) loss 3.5828 (3.3244) grad_norm 2.9735 (1.9311) [2021-04-16 10:23:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][480/1251] eta 0:03:47 lr 0.000279 time 0.2638 (0.2947) loss 2.6660 (3.3261) grad_norm 2.3596 (1.9343) [2021-04-16 10:23:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][490/1251] eta 0:03:44 lr 0.000278 time 0.2931 (0.2944) loss 2.5938 (3.3268) grad_norm 2.3626 (1.9340) [2021-04-16 10:23:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][500/1251] eta 0:03:40 lr 0.000278 time 0.2613 (0.2940) loss 2.4592 (3.3242) grad_norm 1.8511 (1.9348) [2021-04-16 10:23:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][510/1251] eta 0:03:37 lr 0.000278 time 0.2763 (0.2938) loss 3.4916 (3.3282) grad_norm 1.7604 (1.9336) [2021-04-16 10:23:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][520/1251] eta 0:03:34 lr 0.000278 time 0.2702 (0.2935) loss 2.6347 (3.3267) grad_norm 1.7729 (1.9321) [2021-04-16 10:23:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][530/1251] eta 0:03:31 lr 0.000278 time 0.2654 (0.2931) loss 3.4525 (3.3266) grad_norm 1.9151 (1.9317) [2021-04-16 10:23:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][540/1251] eta 0:03:28 lr 0.000278 time 0.2844 (0.2931) loss 3.2128 (3.3263) grad_norm 1.8303 (1.9326) [2021-04-16 10:23:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][550/1251] eta 0:03:25 lr 0.000278 time 0.2668 (0.2927) loss 3.3062 (3.3283) grad_norm 1.6732 (1.9334) [2021-04-16 10:23:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][560/1251] eta 0:03:22 lr 0.000278 time 0.2538 (0.2926) loss 3.8896 (3.3284) grad_norm 1.8107 (1.9337) [2021-04-16 10:23:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][570/1251] eta 0:03:19 lr 0.000278 time 0.2517 (0.2928) loss 3.9155 (3.3307) grad_norm 2.1207 (1.9357) [2021-04-16 10:23:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][580/1251] eta 0:03:16 lr 0.000278 time 0.3050 (0.2929) loss 2.5758 (3.3309) grad_norm 2.3279 (1.9354) [2021-04-16 10:23:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][590/1251] eta 0:03:13 lr 0.000278 time 0.3137 (0.2927) loss 2.6038 (3.3317) grad_norm 1.8574 (1.9371) [2021-04-16 10:23:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][600/1251] eta 0:03:10 lr 0.000278 time 0.2601 (0.2924) loss 3.0956 (3.3326) grad_norm 2.0035 (1.9398) [2021-04-16 10:23:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][610/1251] eta 0:03:07 lr 0.000278 time 0.2720 (0.2921) loss 4.3042 (3.3319) grad_norm 1.9559 (1.9382) [2021-04-16 10:23:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][620/1251] eta 0:03:04 lr 0.000278 time 0.2889 (0.2920) loss 2.6110 (3.3332) grad_norm 1.6336 (1.9373) [2021-04-16 10:23:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][630/1251] eta 0:03:01 lr 0.000278 time 0.2651 (0.2918) loss 2.3233 (3.3326) grad_norm 2.0213 (1.9356) [2021-04-16 10:23:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][640/1251] eta 0:02:58 lr 0.000278 time 0.2906 (0.2916) loss 2.2807 (3.3316) grad_norm 1.9242 (1.9350) [2021-04-16 10:23:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][650/1251] eta 0:02:55 lr 0.000278 time 0.2719 (0.2915) loss 3.2390 (3.3338) grad_norm 1.6582 (1.9344) [2021-04-16 10:23:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][660/1251] eta 0:02:52 lr 0.000278 time 0.2816 (0.2914) loss 3.1206 (3.3356) grad_norm 1.9163 (1.9337) [2021-04-16 10:23:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][670/1251] eta 0:02:49 lr 0.000278 time 0.2877 (0.2913) loss 2.9705 (3.3374) grad_norm 2.2542 (1.9348) [2021-04-16 10:24:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][680/1251] eta 0:02:46 lr 0.000278 time 0.2983 (0.2912) loss 3.6818 (3.3411) grad_norm 2.1898 (1.9344) [2021-04-16 10:24:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][690/1251] eta 0:02:43 lr 0.000278 time 0.2983 (0.2910) loss 3.2742 (3.3420) grad_norm 1.6817 (1.9350) [2021-04-16 10:24:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][700/1251] eta 0:02:40 lr 0.000278 time 0.2883 (0.2909) loss 3.7850 (3.3421) grad_norm 1.7447 (1.9336) [2021-04-16 10:24:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][710/1251] eta 0:02:37 lr 0.000278 time 0.2847 (0.2909) loss 3.3890 (3.3405) grad_norm 1.9652 (1.9335) [2021-04-16 10:24:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][720/1251] eta 0:02:34 lr 0.000278 time 0.2909 (0.2909) loss 4.0453 (3.3441) grad_norm 1.7735 (1.9329) [2021-04-16 10:24:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][730/1251] eta 0:02:31 lr 0.000278 time 0.2759 (0.2908) loss 3.5824 (3.3422) grad_norm 1.7009 (1.9321) [2021-04-16 10:24:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][740/1251] eta 0:02:28 lr 0.000278 time 0.2772 (0.2906) loss 4.0330 (3.3427) grad_norm 2.1675 (1.9337) [2021-04-16 10:24:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][750/1251] eta 0:02:25 lr 0.000278 time 0.2707 (0.2905) loss 2.8942 (3.3421) grad_norm 2.0526 (1.9349) [2021-04-16 10:24:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][760/1251] eta 0:02:22 lr 0.000277 time 0.2767 (0.2903) loss 3.9626 (3.3406) grad_norm 1.9729 (1.9350) [2021-04-16 10:24:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][770/1251] eta 0:02:19 lr 0.000277 time 0.2642 (0.2901) loss 3.6966 (3.3453) grad_norm 2.0131 (1.9374) [2021-04-16 10:24:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][780/1251] eta 0:02:16 lr 0.000277 time 0.2870 (0.2901) loss 4.2082 (3.3483) grad_norm 2.1190 (1.9372) [2021-04-16 10:24:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][790/1251] eta 0:02:13 lr 0.000277 time 0.2788 (0.2900) loss 3.0711 (3.3480) grad_norm 1.8450 (1.9370) [2021-04-16 10:24:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][800/1251] eta 0:02:10 lr 0.000277 time 0.2716 (0.2898) loss 2.7890 (3.3476) grad_norm 1.9356 (1.9357) [2021-04-16 10:24:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][810/1251] eta 0:02:07 lr 0.000277 time 0.2799 (0.2897) loss 3.2614 (3.3485) grad_norm 1.7083 (1.9349) [2021-04-16 10:24:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][820/1251] eta 0:02:04 lr 0.000277 time 0.2904 (0.2896) loss 3.6834 (3.3499) grad_norm 1.8534 (1.9334) [2021-04-16 10:24:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][830/1251] eta 0:02:01 lr 0.000277 time 0.2795 (0.2896) loss 2.3877 (3.3492) grad_norm 1.6554 (1.9329) [2021-04-16 10:24:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][840/1251] eta 0:01:58 lr 0.000277 time 0.2791 (0.2894) loss 3.9439 (3.3476) grad_norm 1.8134 (1.9318) [2021-04-16 10:24:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][850/1251] eta 0:01:56 lr 0.000277 time 0.2762 (0.2893) loss 3.1293 (3.3498) grad_norm 1.8650 (1.9304) [2021-04-16 10:24:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][860/1251] eta 0:01:53 lr 0.000277 time 0.2562 (0.2892) loss 2.7889 (3.3460) grad_norm 1.6383 (1.9292) [2021-04-16 10:24:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][870/1251] eta 0:01:50 lr 0.000277 time 0.2892 (0.2890) loss 3.6952 (3.3448) grad_norm 2.3370 (1.9287) [2021-04-16 10:24:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][880/1251] eta 0:01:47 lr 0.000277 time 0.2806 (0.2889) loss 3.3126 (3.3493) grad_norm 1.7707 (1.9310) [2021-04-16 10:25:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][890/1251] eta 0:01:44 lr 0.000277 time 0.2754 (0.2887) loss 2.8861 (3.3520) grad_norm 1.8362 (1.9304) [2021-04-16 10:25:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][900/1251] eta 0:01:41 lr 0.000277 time 0.3051 (0.2886) loss 3.8731 (3.3526) grad_norm 1.8198 (1.9303) [2021-04-16 10:25:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][910/1251] eta 0:01:38 lr 0.000277 time 0.2762 (0.2885) loss 3.6372 (3.3534) grad_norm 1.8994 (1.9314) [2021-04-16 10:25:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][920/1251] eta 0:01:35 lr 0.000277 time 0.2818 (0.2884) loss 4.0946 (3.3555) grad_norm 1.9279 (1.9325) [2021-04-16 10:25:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][930/1251] eta 0:01:32 lr 0.000277 time 0.3981 (0.2884) loss 2.4763 (3.3548) grad_norm 1.7274 (1.9323) [2021-04-16 10:25:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][940/1251] eta 0:01:29 lr 0.000277 time 0.3750 (0.2885) loss 2.7719 (3.3557) grad_norm 1.6931 (1.9328) [2021-04-16 10:25:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][950/1251] eta 0:01:26 lr 0.000277 time 0.2666 (0.2883) loss 3.4645 (3.3505) grad_norm 1.8186 (1.9348) [2021-04-16 10:25:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][960/1251] eta 0:01:23 lr 0.000277 time 0.2563 (0.2883) loss 2.3963 (3.3515) grad_norm 2.0952 (1.9356) [2021-04-16 10:25:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][970/1251] eta 0:01:21 lr 0.000277 time 0.2688 (0.2884) loss 2.6464 (3.3489) grad_norm 1.9846 (1.9345) [2021-04-16 10:25:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][980/1251] eta 0:01:18 lr 0.000277 time 0.2813 (0.2882) loss 3.9258 (3.3507) grad_norm 2.2046 (1.9347) [2021-04-16 10:25:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][990/1251] eta 0:01:15 lr 0.000277 time 0.2663 (0.2881) loss 3.8803 (3.3498) grad_norm 2.2265 (1.9350) [2021-04-16 10:25:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][1000/1251] eta 0:01:12 lr 0.000277 time 0.2822 (0.2880) loss 3.3166 (3.3472) grad_norm 1.9349 (1.9347) [2021-04-16 10:25:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][1010/1251] eta 0:01:09 lr 0.000277 time 0.2490 (0.2878) loss 3.5605 (3.3485) grad_norm 1.6348 (1.9346) [2021-04-16 10:25:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][1020/1251] eta 0:01:06 lr 0.000277 time 0.2693 (0.2877) loss 3.4715 (3.3470) grad_norm 1.8831 (1.9345) [2021-04-16 10:25:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][1030/1251] eta 0:01:03 lr 0.000276 time 0.2585 (0.2877) loss 3.3705 (3.3457) grad_norm 1.8271 (1.9348) [2021-04-16 10:25:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][1040/1251] eta 0:01:00 lr 0.000276 time 0.2742 (0.2876) loss 3.3914 (3.3455) grad_norm 1.7809 (1.9345) [2021-04-16 10:25:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][1050/1251] eta 0:00:57 lr 0.000276 time 0.2486 (0.2875) loss 3.7082 (3.3467) grad_norm 2.1896 (1.9340) [2021-04-16 10:25:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][1060/1251] eta 0:00:54 lr 0.000276 time 0.2614 (0.2874) loss 2.9642 (3.3484) grad_norm 2.0327 (1.9369) [2021-04-16 10:25:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][1070/1251] eta 0:00:52 lr 0.000276 time 0.2853 (0.2873) loss 3.5874 (3.3482) grad_norm 1.8141 (1.9372) [2021-04-16 10:25:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][1080/1251] eta 0:00:49 lr 0.000276 time 0.2682 (0.2872) loss 2.5912 (3.3476) grad_norm 1.9592 (1.9362) [2021-04-16 10:25:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][1090/1251] eta 0:00:46 lr 0.000276 time 0.2872 (0.2871) loss 3.7815 (3.3483) grad_norm 2.1893 (1.9360) [2021-04-16 10:25:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][1100/1251] eta 0:00:43 lr 0.000276 time 0.2758 (0.2870) loss 3.5297 (3.3497) grad_norm 1.7322 (1.9356) [2021-04-16 10:26:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][1110/1251] eta 0:00:40 lr 0.000276 time 0.2716 (0.2869) loss 3.4020 (3.3497) grad_norm 1.8322 (1.9355) [2021-04-16 10:26:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][1120/1251] eta 0:00:37 lr 0.000276 time 0.2562 (0.2869) loss 2.6599 (3.3483) grad_norm 1.8828 (1.9360) [2021-04-16 10:26:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][1130/1251] eta 0:00:34 lr 0.000276 time 0.2744 (0.2868) loss 3.6195 (3.3488) grad_norm 2.0202 (1.9362) [2021-04-16 10:26:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][1140/1251] eta 0:00:31 lr 0.000276 time 0.2730 (0.2867) loss 3.2803 (3.3496) grad_norm 1.8094 (1.9358) [2021-04-16 10:26:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][1150/1251] eta 0:00:28 lr 0.000276 time 0.2940 (0.2870) loss 3.2950 (3.3486) grad_norm 1.8845 (1.9349) [2021-04-16 10:26:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][1160/1251] eta 0:00:26 lr 0.000276 time 0.2654 (0.2869) loss 3.2930 (3.3473) grad_norm 1.9178 (1.9360) [2021-04-16 10:26:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][1170/1251] eta 0:00:23 lr 0.000276 time 0.2852 (0.2868) loss 3.3696 (3.3463) grad_norm 1.8489 (1.9354) [2021-04-16 10:26:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][1180/1251] eta 0:00:20 lr 0.000276 time 0.2805 (0.2867) loss 3.7495 (3.3460) grad_norm 1.9413 (1.9359) [2021-04-16 10:26:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][1190/1251] eta 0:00:17 lr 0.000276 time 0.2737 (0.2866) loss 3.2852 (3.3448) grad_norm 1.8143 (1.9361) [2021-04-16 10:26:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][1200/1251] eta 0:00:14 lr 0.000276 time 0.2874 (0.2865) loss 3.8295 (3.3464) grad_norm 1.7873 (1.9359) [2021-04-16 10:26:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][1210/1251] eta 0:00:11 lr 0.000276 time 0.2819 (0.2865) loss 4.1286 (3.3460) grad_norm 2.2983 (1.9354) [2021-04-16 10:26:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][1220/1251] eta 0:00:08 lr 0.000276 time 0.2857 (0.2864) loss 3.5742 (3.3488) grad_norm 2.3255 (1.9373) [2021-04-16 10:26:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][1230/1251] eta 0:00:06 lr 0.000276 time 0.2875 (0.2863) loss 3.9190 (3.3508) grad_norm 2.0225 (1.9387) [2021-04-16 10:26:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][1240/1251] eta 0:00:03 lr 0.000276 time 0.2587 (0.2862) loss 3.6520 (3.3519) grad_norm 2.1100 (1.9391) [2021-04-16 10:26:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [195/300][1250/1251] eta 0:00:00 lr 0.000276 time 0.2487 (0.2859) loss 3.4194 (3.3525) grad_norm 1.8725 (1.9410) [2021-04-16 10:26:44 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 195 training takes 0:06:01 [2021-04-16 10:26:44 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_195.pth saving...... [2021-04-16 10:27:03 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_195.pth saved !!! [2021-04-16 10:27:04 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.117 (1.117) Loss 0.8962 (0.8962) Acc@1 79.688 (79.688) Acc@5 93.652 (93.652) [2021-04-16 10:27:05 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.104 (0.224) Loss 0.8644 (0.9155) Acc@1 79.492 (78.436) Acc@5 95.801 (94.105) [2021-04-16 10:27:07 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.154 (0.193) Loss 0.8561 (0.9266) Acc@1 79.199 (78.120) Acc@5 95.215 (94.029) [2021-04-16 10:27:10 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.089 (0.226) Loss 0.9419 (0.9236) Acc@1 78.516 (78.270) Acc@5 93.848 (94.169) [2021-04-16 10:27:12 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.280 (0.222) Loss 0.9231 (0.9266) Acc@1 78.809 (78.085) Acc@5 94.727 (94.155) [2021-04-16 10:27:18 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 78.008 Acc@5 94.178 [2021-04-16 10:27:18 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 78.0% [2021-04-16 10:27:18 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 78.08% [2021-04-16 10:27:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][0/1251] eta 6:05:51 lr 0.000276 time 17.5468 (17.5468) loss 3.6779 (3.6779) grad_norm 1.9033 (1.9033) [2021-04-16 10:27:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][10/1251] eta 0:38:08 lr 0.000276 time 0.3039 (1.8445) loss 3.4808 (3.3971) grad_norm 1.6538 (1.9657) [2021-04-16 10:27:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][20/1251] eta 0:22:32 lr 0.000276 time 0.2959 (1.0991) loss 3.7738 (3.4755) grad_norm 1.8204 (1.9656) [2021-04-16 10:27:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][30/1251] eta 0:17:02 lr 0.000276 time 0.3030 (0.8371) loss 3.9807 (3.5113) grad_norm 1.9244 (1.9802) [2021-04-16 10:27:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][40/1251] eta 0:14:13 lr 0.000276 time 0.3048 (0.7045) loss 2.2942 (3.4479) grad_norm 1.6893 (1.9360) [2021-04-16 10:27:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][50/1251] eta 0:12:26 lr 0.000275 time 0.2615 (0.6216) loss 3.5230 (3.4673) grad_norm 2.3373 (1.9376) [2021-04-16 10:27:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][60/1251] eta 0:11:13 lr 0.000275 time 0.2787 (0.5657) loss 4.2488 (3.4688) grad_norm 2.2788 (1.9491) [2021-04-16 10:27:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][70/1251] eta 0:10:21 lr 0.000275 time 0.2900 (0.5259) loss 3.2772 (3.4753) grad_norm 1.8015 (1.9450) [2021-04-16 10:27:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][80/1251] eta 0:09:40 lr 0.000275 time 0.2761 (0.4959) loss 4.3186 (3.4311) grad_norm 1.8913 (1.9480) [2021-04-16 10:28:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][90/1251] eta 0:09:09 lr 0.000275 time 0.3078 (0.4733) loss 3.4646 (3.4393) grad_norm 2.2705 (1.9534) [2021-04-16 10:28:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][100/1251] eta 0:08:43 lr 0.000275 time 0.2555 (0.4547) loss 2.5540 (3.4446) grad_norm 1.8121 (1.9566) [2021-04-16 10:28:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][110/1251] eta 0:08:21 lr 0.000275 time 0.2705 (0.4396) loss 4.0893 (3.4414) grad_norm 1.9514 (1.9541) [2021-04-16 10:28:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][120/1251] eta 0:08:02 lr 0.000275 time 0.2863 (0.4267) loss 3.3755 (3.4288) grad_norm 2.0915 (1.9575) [2021-04-16 10:28:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][130/1251] eta 0:07:45 lr 0.000275 time 0.3043 (0.4157) loss 3.6868 (3.4234) grad_norm 2.0589 (1.9659) [2021-04-16 10:28:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][140/1251] eta 0:07:33 lr 0.000275 time 0.4360 (0.4080) loss 3.3601 (3.4254) grad_norm 1.8983 (1.9597) [2021-04-16 10:28:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][150/1251] eta 0:07:21 lr 0.000275 time 0.4803 (0.4008) loss 2.9768 (3.4305) grad_norm 1.8628 (1.9511) [2021-04-16 10:28:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][160/1251] eta 0:07:08 lr 0.000275 time 0.2676 (0.3931) loss 3.6605 (3.4386) grad_norm 2.8433 (1.9517) [2021-04-16 10:28:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][170/1251] eta 0:06:59 lr 0.000275 time 0.2773 (0.3876) loss 3.5400 (3.4369) grad_norm 2.0071 (1.9473) [2021-04-16 10:28:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][180/1251] eta 0:06:50 lr 0.000275 time 0.2816 (0.3828) loss 3.0180 (3.4361) grad_norm 1.8277 (1.9476) [2021-04-16 10:28:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][190/1251] eta 0:06:40 lr 0.000275 time 0.2781 (0.3779) loss 3.8530 (3.4239) grad_norm 1.5216 (1.9453) [2021-04-16 10:28:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][200/1251] eta 0:06:32 lr 0.000275 time 0.3126 (0.3732) loss 3.9377 (3.4303) grad_norm 1.8657 (1.9455) [2021-04-16 10:28:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][210/1251] eta 0:06:23 lr 0.000275 time 0.2709 (0.3687) loss 2.9762 (3.4283) grad_norm 2.0073 (1.9446) [2021-04-16 10:28:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][220/1251] eta 0:06:16 lr 0.000275 time 0.2944 (0.3650) loss 2.7876 (3.4204) grad_norm 1.6676 (1.9433) [2021-04-16 10:28:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][230/1251] eta 0:06:08 lr 0.000275 time 0.2646 (0.3611) loss 4.1294 (3.4047) grad_norm 1.8156 (1.9398) [2021-04-16 10:28:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][240/1251] eta 0:06:01 lr 0.000275 time 0.2912 (0.3577) loss 2.2382 (3.4011) grad_norm 1.5711 (1.9368) [2021-04-16 10:28:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][250/1251] eta 0:05:55 lr 0.000275 time 0.2854 (0.3548) loss 2.0068 (3.3975) grad_norm 1.6290 (1.9370) [2021-04-16 10:28:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][260/1251] eta 0:05:48 lr 0.000275 time 0.2698 (0.3520) loss 3.4125 (3.3891) grad_norm 2.2883 (1.9424) [2021-04-16 10:28:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][270/1251] eta 0:05:42 lr 0.000275 time 0.2694 (0.3493) loss 3.4766 (3.3874) grad_norm 1.7519 (1.9436) [2021-04-16 10:28:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][280/1251] eta 0:05:36 lr 0.000275 time 0.2665 (0.3467) loss 3.6302 (3.3876) grad_norm 1.9722 (1.9424) [2021-04-16 10:28:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][290/1251] eta 0:05:30 lr 0.000275 time 0.2807 (0.3444) loss 3.8949 (3.3906) grad_norm 1.7079 (1.9384) [2021-04-16 10:29:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][300/1251] eta 0:05:25 lr 0.000275 time 0.2818 (0.3425) loss 1.9475 (3.3902) grad_norm 1.8665 (1.9422) [2021-04-16 10:29:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][310/1251] eta 0:05:20 lr 0.000275 time 0.2890 (0.3405) loss 3.5437 (3.3900) grad_norm 1.7647 (1.9429) [2021-04-16 10:29:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][320/1251] eta 0:05:15 lr 0.000274 time 0.2795 (0.3391) loss 3.3428 (3.3824) grad_norm 2.0814 (1.9436) [2021-04-16 10:29:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][330/1251] eta 0:05:10 lr 0.000274 time 0.2773 (0.3375) loss 3.2038 (3.3791) grad_norm 1.9948 (1.9421) [2021-04-16 10:29:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][340/1251] eta 0:05:05 lr 0.000274 time 0.2831 (0.3357) loss 3.0290 (3.3741) grad_norm 1.8088 (1.9438) [2021-04-16 10:29:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][350/1251] eta 0:05:01 lr 0.000274 time 0.2648 (0.3347) loss 3.2967 (3.3747) grad_norm 2.0697 (1.9512) [2021-04-16 10:29:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][360/1251] eta 0:04:57 lr 0.000274 time 0.2999 (0.3334) loss 3.8634 (3.3672) grad_norm 1.8847 (1.9532) [2021-04-16 10:29:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][370/1251] eta 0:04:52 lr 0.000274 time 0.3042 (0.3320) loss 3.8672 (3.3654) grad_norm 1.6522 (1.9515) [2021-04-16 10:29:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][380/1251] eta 0:04:47 lr 0.000274 time 0.2677 (0.3306) loss 2.9094 (3.3637) grad_norm 1.7824 (1.9527) [2021-04-16 10:29:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][390/1251] eta 0:04:43 lr 0.000274 time 0.2496 (0.3292) loss 3.6631 (3.3624) grad_norm 1.8938 (1.9487) [2021-04-16 10:29:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][400/1251] eta 0:04:39 lr 0.000274 time 0.2718 (0.3281) loss 2.9224 (3.3611) grad_norm 1.6458 (1.9458) [2021-04-16 10:29:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][410/1251] eta 0:04:34 lr 0.000274 time 0.2915 (0.3269) loss 2.5518 (3.3648) grad_norm 1.7397 (1.9442) [2021-04-16 10:29:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][420/1251] eta 0:04:30 lr 0.000274 time 0.3009 (0.3258) loss 3.4295 (3.3615) grad_norm 2.1512 (1.9451) [2021-04-16 10:29:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][430/1251] eta 0:04:26 lr 0.000274 time 0.2991 (0.3248) loss 3.2639 (3.3573) grad_norm 2.2656 (1.9469) [2021-04-16 10:29:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][440/1251] eta 0:04:22 lr 0.000274 time 0.2865 (0.3238) loss 4.0921 (3.3609) grad_norm 2.3082 (1.9493) [2021-04-16 10:29:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][450/1251] eta 0:04:18 lr 0.000274 time 0.2944 (0.3228) loss 2.5730 (3.3618) grad_norm 1.9392 (1.9509) [2021-04-16 10:29:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][460/1251] eta 0:04:14 lr 0.000274 time 0.2803 (0.3219) loss 3.8527 (3.3648) grad_norm 1.7671 (1.9483) [2021-04-16 10:29:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][470/1251] eta 0:04:10 lr 0.000274 time 0.2819 (0.3211) loss 3.5018 (3.3659) grad_norm 1.6211 (1.9488) [2021-04-16 10:29:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][480/1251] eta 0:04:06 lr 0.000274 time 0.2573 (0.3201) loss 3.8267 (3.3668) grad_norm 1.6251 (1.9476) [2021-04-16 10:29:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][490/1251] eta 0:04:03 lr 0.000274 time 0.3019 (0.3193) loss 3.9037 (3.3683) grad_norm 2.2345 (1.9464) [2021-04-16 10:29:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][500/1251] eta 0:03:59 lr 0.000274 time 0.2640 (0.3185) loss 3.5737 (3.3675) grad_norm 1.8940 (1.9461) [2021-04-16 10:30:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][510/1251] eta 0:03:55 lr 0.000274 time 0.2586 (0.3178) loss 3.4711 (3.3678) grad_norm 1.9300 (1.9475) [2021-04-16 10:30:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][520/1251] eta 0:03:51 lr 0.000274 time 0.2982 (0.3173) loss 3.7931 (3.3672) grad_norm 2.1595 (1.9480) [2021-04-16 10:30:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][530/1251] eta 0:03:48 lr 0.000274 time 0.2771 (0.3164) loss 3.8502 (3.3710) grad_norm 2.0128 (1.9486) [2021-04-16 10:30:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][540/1251] eta 0:03:44 lr 0.000274 time 0.2547 (0.3157) loss 4.0486 (3.3785) grad_norm 2.4288 (1.9529) [2021-04-16 10:30:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][550/1251] eta 0:03:41 lr 0.000274 time 0.4354 (0.3154) loss 2.6608 (3.3743) grad_norm 1.8238 (1.9557) [2021-04-16 10:30:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][560/1251] eta 0:03:37 lr 0.000274 time 0.2878 (0.3149) loss 3.4901 (3.3709) grad_norm 2.1834 (1.9572) [2021-04-16 10:30:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][570/1251] eta 0:03:33 lr 0.000274 time 0.2625 (0.3142) loss 2.2843 (3.3700) grad_norm 2.1145 (1.9567) [2021-04-16 10:30:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][580/1251] eta 0:03:30 lr 0.000274 time 0.2861 (0.3140) loss 2.9963 (3.3691) grad_norm 1.8344 (1.9562) [2021-04-16 10:30:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][590/1251] eta 0:03:27 lr 0.000274 time 0.2793 (0.3135) loss 2.9682 (3.3686) grad_norm 1.8448 (1.9581) [2021-04-16 10:30:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][600/1251] eta 0:03:23 lr 0.000273 time 0.2745 (0.3129) loss 3.5732 (3.3693) grad_norm 1.7364 (1.9578) [2021-04-16 10:30:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][610/1251] eta 0:03:20 lr 0.000273 time 0.3013 (0.3125) loss 3.1297 (3.3695) grad_norm 2.1584 (1.9578) [2021-04-16 10:30:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][620/1251] eta 0:03:16 lr 0.000273 time 0.2765 (0.3122) loss 3.6215 (3.3708) grad_norm 1.7919 (1.9582) [2021-04-16 10:30:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][630/1251] eta 0:03:13 lr 0.000273 time 0.2710 (0.3116) loss 3.7124 (3.3721) grad_norm 2.0294 (1.9566) [2021-04-16 10:30:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][640/1251] eta 0:03:10 lr 0.000273 time 0.2974 (0.3111) loss 3.5544 (3.3681) grad_norm 1.7862 (1.9553) [2021-04-16 10:30:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][650/1251] eta 0:03:06 lr 0.000273 time 0.2860 (0.3107) loss 2.9959 (3.3637) grad_norm 1.8345 (1.9546) [2021-04-16 10:30:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][660/1251] eta 0:03:03 lr 0.000273 time 0.2681 (0.3103) loss 2.5537 (3.3648) grad_norm 2.1775 (1.9547) [2021-04-16 10:30:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][670/1251] eta 0:03:00 lr 0.000273 time 0.2790 (0.3099) loss 2.8334 (3.3653) grad_norm 1.9591 (1.9548) [2021-04-16 10:30:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][680/1251] eta 0:02:56 lr 0.000273 time 0.2930 (0.3094) loss 2.2592 (3.3690) grad_norm 2.0401 (1.9549) [2021-04-16 10:30:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][690/1251] eta 0:02:53 lr 0.000273 time 0.2512 (0.3090) loss 3.8590 (3.3681) grad_norm 1.7706 (1.9556) [2021-04-16 10:30:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][700/1251] eta 0:02:49 lr 0.000273 time 0.2675 (0.3085) loss 3.4887 (3.3663) grad_norm 2.0457 (1.9545) [2021-04-16 10:30:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][710/1251] eta 0:02:46 lr 0.000273 time 0.3142 (0.3083) loss 3.6861 (3.3668) grad_norm 2.2952 (1.9556) [2021-04-16 10:31:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][720/1251] eta 0:02:43 lr 0.000273 time 0.2857 (0.3081) loss 3.2377 (3.3628) grad_norm 2.1096 (1.9555) [2021-04-16 10:31:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][730/1251] eta 0:02:40 lr 0.000273 time 0.2622 (0.3079) loss 2.7874 (3.3613) grad_norm 2.2664 (1.9556) [2021-04-16 10:31:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][740/1251] eta 0:02:37 lr 0.000273 time 0.2841 (0.3079) loss 3.3038 (3.3576) grad_norm 2.1179 (1.9565) [2021-04-16 10:31:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][750/1251] eta 0:02:34 lr 0.000273 time 0.2759 (0.3075) loss 3.7420 (3.3555) grad_norm 1.9724 (1.9568) [2021-04-16 10:31:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][760/1251] eta 0:02:30 lr 0.000273 time 0.2705 (0.3072) loss 2.3242 (3.3521) grad_norm 2.0990 (1.9574) [2021-04-16 10:31:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][770/1251] eta 0:02:27 lr 0.000273 time 0.3005 (0.3069) loss 3.4798 (3.3514) grad_norm 2.0723 (1.9599) [2021-04-16 10:31:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][780/1251] eta 0:02:24 lr 0.000273 time 0.2706 (0.3065) loss 3.8470 (3.3508) grad_norm 2.0294 (1.9596) [2021-04-16 10:31:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][790/1251] eta 0:02:21 lr 0.000273 time 0.2770 (0.3062) loss 3.3388 (3.3514) grad_norm 1.7826 (1.9608) [2021-04-16 10:31:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][800/1251] eta 0:02:18 lr 0.000273 time 0.2938 (0.3060) loss 3.8144 (3.3508) grad_norm 1.9821 (1.9601) [2021-04-16 10:31:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][810/1251] eta 0:02:14 lr 0.000273 time 0.2811 (0.3057) loss 3.7756 (3.3502) grad_norm 1.6734 (1.9591) [2021-04-16 10:31:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][820/1251] eta 0:02:11 lr 0.000273 time 0.2773 (0.3054) loss 3.7398 (3.3501) grad_norm 2.0079 (1.9589) [2021-04-16 10:31:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][830/1251] eta 0:02:08 lr 0.000273 time 0.2800 (0.3051) loss 3.3367 (3.3494) grad_norm 1.8628 (1.9580) [2021-04-16 10:31:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][840/1251] eta 0:02:05 lr 0.000273 time 0.2717 (0.3049) loss 3.8943 (3.3493) grad_norm 2.2394 (1.9577) [2021-04-16 10:31:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][850/1251] eta 0:02:02 lr 0.000273 time 0.2615 (0.3046) loss 3.7592 (3.3498) grad_norm 2.0955 (1.9574) [2021-04-16 10:31:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][860/1251] eta 0:01:58 lr 0.000273 time 0.2712 (0.3043) loss 2.9151 (3.3471) grad_norm 1.9942 (1.9557) [2021-04-16 10:31:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][870/1251] eta 0:01:55 lr 0.000272 time 0.2993 (0.3040) loss 3.7425 (3.3500) grad_norm 1.9361 (1.9555) [2021-04-16 10:31:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][880/1251] eta 0:01:52 lr 0.000272 time 0.2908 (0.3038) loss 3.2710 (3.3487) grad_norm 1.6957 (1.9556) [2021-04-16 10:31:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][890/1251] eta 0:01:49 lr 0.000272 time 0.2742 (0.3037) loss 2.7622 (3.3459) grad_norm 1.7349 (1.9562) [2021-04-16 10:31:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][900/1251] eta 0:01:46 lr 0.000272 time 0.3338 (0.3035) loss 3.7270 (3.3428) grad_norm 1.8984 (1.9558) [2021-04-16 10:31:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][910/1251] eta 0:01:43 lr 0.000272 time 0.2842 (0.3033) loss 3.1957 (3.3395) grad_norm 2.1002 (1.9558) [2021-04-16 10:31:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][920/1251] eta 0:01:40 lr 0.000272 time 0.2681 (0.3030) loss 3.3217 (3.3405) grad_norm 1.9162 (1.9570) [2021-04-16 10:32:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][930/1251] eta 0:01:37 lr 0.000272 time 0.2937 (0.3029) loss 2.5092 (3.3374) grad_norm 1.6029 (1.9565) [2021-04-16 10:32:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][940/1251] eta 0:01:34 lr 0.000272 time 0.2651 (0.3028) loss 3.7303 (3.3382) grad_norm 2.1410 (1.9561) [2021-04-16 10:32:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][950/1251] eta 0:01:31 lr 0.000272 time 0.3027 (0.3026) loss 4.0507 (3.3400) grad_norm 1.7777 (1.9560) [2021-04-16 10:32:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][960/1251] eta 0:01:27 lr 0.000272 time 0.2749 (0.3023) loss 3.5536 (3.3415) grad_norm 1.9578 (1.9557) [2021-04-16 10:32:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][970/1251] eta 0:01:24 lr 0.000272 time 0.2812 (0.3021) loss 3.2182 (3.3419) grad_norm 2.1065 (1.9570) [2021-04-16 10:32:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][980/1251] eta 0:01:21 lr 0.000272 time 0.2710 (0.3018) loss 3.7136 (3.3406) grad_norm 2.0293 (1.9577) [2021-04-16 10:32:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][990/1251] eta 0:01:18 lr 0.000272 time 0.2720 (0.3016) loss 3.7860 (3.3411) grad_norm 2.0734 (1.9570) [2021-04-16 10:32:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][1000/1251] eta 0:01:15 lr 0.000272 time 0.2786 (0.3014) loss 2.9702 (3.3393) grad_norm 1.6706 (1.9577) [2021-04-16 10:32:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][1010/1251] eta 0:01:12 lr 0.000272 time 0.2866 (0.3012) loss 2.4097 (3.3369) grad_norm 1.7680 (1.9572) [2021-04-16 10:32:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][1020/1251] eta 0:01:09 lr 0.000272 time 0.2878 (0.3010) loss 3.2780 (3.3363) grad_norm 2.0659 (1.9570) [2021-04-16 10:32:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][1030/1251] eta 0:01:06 lr 0.000272 time 0.3024 (0.3008) loss 3.2872 (3.3360) grad_norm 1.8298 (1.9561) [2021-04-16 10:32:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][1040/1251] eta 0:01:03 lr 0.000272 time 0.2658 (0.3006) loss 3.6406 (3.3365) grad_norm 1.8817 (1.9568) [2021-04-16 10:32:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][1050/1251] eta 0:01:00 lr 0.000272 time 0.2634 (0.3003) loss 3.9586 (3.3382) grad_norm 1.9827 (1.9557) [2021-04-16 10:32:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][1060/1251] eta 0:00:57 lr 0.000272 time 0.2700 (0.3002) loss 3.7796 (3.3409) grad_norm 1.7937 (1.9552) [2021-04-16 10:32:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][1070/1251] eta 0:00:54 lr 0.000272 time 0.2692 (0.3000) loss 3.5302 (3.3408) grad_norm 2.3235 (1.9560) [2021-04-16 10:32:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][1080/1251] eta 0:00:51 lr 0.000272 time 0.2707 (0.2999) loss 2.2341 (3.3415) grad_norm 1.9982 (1.9552) [2021-04-16 10:32:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][1090/1251] eta 0:00:48 lr 0.000272 time 0.2614 (0.2998) loss 3.1730 (3.3415) grad_norm 1.7430 (1.9560) [2021-04-16 10:32:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][1100/1251] eta 0:00:45 lr 0.000272 time 0.2964 (0.2997) loss 2.5707 (3.3431) grad_norm 2.4065 (1.9567) [2021-04-16 10:32:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][1110/1251] eta 0:00:42 lr 0.000272 time 0.2762 (0.2995) loss 3.5606 (3.3434) grad_norm 1.9330 (1.9570) [2021-04-16 10:32:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][1120/1251] eta 0:00:39 lr 0.000272 time 0.2889 (0.2994) loss 3.3673 (3.3458) grad_norm 2.7214 (1.9568) [2021-04-16 10:32:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][1130/1251] eta 0:00:36 lr 0.000272 time 0.2707 (0.2994) loss 2.2668 (3.3424) grad_norm 1.8950 (1.9565) [2021-04-16 10:32:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][1140/1251] eta 0:00:33 lr 0.000271 time 0.2876 (0.2992) loss 3.8407 (3.3436) grad_norm 2.1976 (1.9573) [2021-04-16 10:33:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][1150/1251] eta 0:00:30 lr 0.000271 time 0.2847 (0.2992) loss 3.4305 (3.3436) grad_norm 1.7053 (1.9568) [2021-04-16 10:33:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][1160/1251] eta 0:00:27 lr 0.000271 time 0.2884 (0.2992) loss 3.2467 (3.3446) grad_norm 2.0008 (1.9577) [2021-04-16 10:33:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][1170/1251] eta 0:00:24 lr 0.000271 time 0.2600 (0.2991) loss 2.1179 (3.3456) grad_norm 1.9518 (1.9575) [2021-04-16 10:33:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][1180/1251] eta 0:00:21 lr 0.000271 time 0.2851 (0.2989) loss 3.4300 (3.3456) grad_norm 2.1049 (1.9572) [2021-04-16 10:33:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][1190/1251] eta 0:00:18 lr 0.000271 time 0.4106 (0.2989) loss 3.7135 (3.3446) grad_norm 1.7007 (1.9566) [2021-04-16 10:33:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][1200/1251] eta 0:00:15 lr 0.000271 time 0.2834 (0.2988) loss 3.2566 (3.3448) grad_norm 2.3764 (1.9569) [2021-04-16 10:33:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][1210/1251] eta 0:00:12 lr 0.000271 time 0.2845 (0.2987) loss 2.2058 (3.3412) grad_norm 1.8880 (1.9574) [2021-04-16 10:33:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][1220/1251] eta 0:00:09 lr 0.000271 time 0.3155 (0.2986) loss 2.8611 (3.3414) grad_norm 2.0960 (1.9578) [2021-04-16 10:33:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][1230/1251] eta 0:00:06 lr 0.000271 time 0.3025 (0.2984) loss 3.3934 (3.3420) grad_norm 2.5388 (1.9590) [2021-04-16 10:33:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][1240/1251] eta 0:00:03 lr 0.000271 time 0.2485 (0.2982) loss 3.8976 (3.3420) grad_norm 1.9549 (1.9591) [2021-04-16 10:33:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [196/300][1250/1251] eta 0:00:00 lr 0.000271 time 0.2516 (0.2978) loss 3.2010 (3.3414) grad_norm 1.8485 (1.9594) [2021-04-16 10:33:34 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 196 training takes 0:06:16 [2021-04-16 10:33:34 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_196.pth saving...... [2021-04-16 10:33:51 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_196.pth saved !!! [2021-04-16 10:33:53 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.251 (1.251) Loss 0.8175 (0.8175) Acc@1 79.980 (79.980) Acc@5 96.094 (96.094) [2021-04-16 10:33:54 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.123 (0.222) Loss 0.9657 (0.9150) Acc@1 77.734 (78.196) Acc@5 93.457 (94.274) [2021-04-16 10:33:56 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.155 (0.210) Loss 0.9634 (0.9173) Acc@1 78.418 (78.172) Acc@5 93.848 (94.345) [2021-04-16 10:33:59 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.328 (0.233) Loss 0.8509 (0.9127) Acc@1 80.176 (78.364) Acc@5 94.434 (94.298) [2021-04-16 10:34:00 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.216) Loss 0.9327 (0.9168) Acc@1 76.953 (78.196) Acc@5 94.043 (94.250) [2021-04-16 10:34:05 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 78.164 Acc@5 94.310 [2021-04-16 10:34:05 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 78.2% [2021-04-16 10:34:05 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 78.16% [2021-04-16 10:34:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][0/1251] eta 3:38:52 lr 0.000271 time 10.4979 (10.4979) loss 2.2291 (2.2291) grad_norm 1.9696 (1.9696) [2021-04-16 10:34:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][10/1251] eta 0:24:45 lr 0.000271 time 0.2630 (1.1971) loss 2.3331 (2.9740) grad_norm 1.7353 (1.8729) [2021-04-16 10:34:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][20/1251] eta 0:15:35 lr 0.000271 time 0.2796 (0.7601) loss 3.7839 (3.2629) grad_norm 2.0661 (1.9506) [2021-04-16 10:34:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][30/1251] eta 0:12:17 lr 0.000271 time 0.2625 (0.6041) loss 3.1550 (3.2844) grad_norm 1.7702 (1.9226) [2021-04-16 10:34:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][40/1251] eta 0:10:35 lr 0.000271 time 0.2999 (0.5251) loss 3.5243 (3.3096) grad_norm 1.8531 (1.9607) [2021-04-16 10:34:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][50/1251] eta 0:09:32 lr 0.000271 time 0.2421 (0.4768) loss 3.5121 (3.3352) grad_norm 1.9322 (1.9849) [2021-04-16 10:34:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][60/1251] eta 0:08:48 lr 0.000271 time 0.2746 (0.4438) loss 3.2204 (3.3314) grad_norm 1.9380 (1.9837) [2021-04-16 10:34:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][70/1251] eta 0:08:18 lr 0.000271 time 0.2820 (0.4220) loss 3.1652 (3.2986) grad_norm 1.9210 (1.9725) [2021-04-16 10:34:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][80/1251] eta 0:07:54 lr 0.000271 time 0.2917 (0.4053) loss 3.7443 (3.3252) grad_norm 1.9564 (1.9532) [2021-04-16 10:34:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][90/1251] eta 0:07:34 lr 0.000271 time 0.2834 (0.3912) loss 3.3868 (3.3209) grad_norm 2.3017 (1.9507) [2021-04-16 10:34:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][100/1251] eta 0:07:19 lr 0.000271 time 0.2742 (0.3821) loss 3.4406 (3.3314) grad_norm 1.8460 (1.9492) [2021-04-16 10:34:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][110/1251] eta 0:07:06 lr 0.000271 time 0.4027 (0.3742) loss 3.9489 (3.3351) grad_norm 1.8133 (1.9502) [2021-04-16 10:34:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][120/1251] eta 0:06:53 lr 0.000271 time 0.2854 (0.3659) loss 2.4108 (3.3248) grad_norm 1.9363 (1.9549) [2021-04-16 10:34:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][130/1251] eta 0:06:44 lr 0.000271 time 0.4385 (0.3605) loss 3.8679 (3.3300) grad_norm 1.9207 (1.9520) [2021-04-16 10:34:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][140/1251] eta 0:06:34 lr 0.000271 time 0.2666 (0.3552) loss 3.6016 (3.3345) grad_norm 2.0321 (1.9505) [2021-04-16 10:34:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][150/1251] eta 0:06:26 lr 0.000271 time 0.2982 (0.3509) loss 2.1317 (3.3333) grad_norm 1.6602 (1.9489) [2021-04-16 10:35:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][160/1251] eta 0:06:17 lr 0.000271 time 0.2694 (0.3462) loss 3.2062 (3.3238) grad_norm 2.1344 (1.9486) [2021-04-16 10:35:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][170/1251] eta 0:06:09 lr 0.000270 time 0.2657 (0.3422) loss 3.6610 (3.3188) grad_norm 2.1071 (1.9519) [2021-04-16 10:35:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][180/1251] eta 0:06:02 lr 0.000270 time 0.2988 (0.3389) loss 2.9151 (3.3155) grad_norm 2.0119 (1.9480) [2021-04-16 10:35:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][190/1251] eta 0:05:56 lr 0.000270 time 0.2992 (0.3357) loss 3.6365 (3.3259) grad_norm 2.0189 (1.9495) [2021-04-16 10:35:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][200/1251] eta 0:05:49 lr 0.000270 time 0.2816 (0.3328) loss 3.4300 (3.3213) grad_norm 1.9918 (1.9527) [2021-04-16 10:35:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][210/1251] eta 0:05:43 lr 0.000270 time 0.2652 (0.3300) loss 2.6894 (3.3188) grad_norm 2.2619 (1.9508) [2021-04-16 10:35:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][220/1251] eta 0:05:38 lr 0.000270 time 0.2786 (0.3281) loss 3.6465 (3.3388) grad_norm 1.7679 (1.9490) [2021-04-16 10:35:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][230/1251] eta 0:05:32 lr 0.000270 time 0.2548 (0.3256) loss 3.7104 (3.3416) grad_norm 1.8505 (1.9498) [2021-04-16 10:35:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][240/1251] eta 0:05:27 lr 0.000270 time 0.2716 (0.3236) loss 3.5566 (3.3277) grad_norm 1.8878 (1.9459) [2021-04-16 10:35:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][250/1251] eta 0:05:22 lr 0.000270 time 0.2803 (0.3221) loss 2.9111 (3.3347) grad_norm 1.7768 (1.9414) [2021-04-16 10:35:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][260/1251] eta 0:05:17 lr 0.000270 time 0.2851 (0.3206) loss 3.4172 (3.3466) grad_norm 2.3221 (1.9411) [2021-04-16 10:35:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][270/1251] eta 0:05:12 lr 0.000270 time 0.2753 (0.3188) loss 2.8428 (3.3404) grad_norm 1.6005 (1.9437) [2021-04-16 10:35:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][280/1251] eta 0:05:08 lr 0.000270 time 0.2773 (0.3174) loss 3.2728 (3.3245) grad_norm 1.8388 (1.9413) [2021-04-16 10:35:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][290/1251] eta 0:05:04 lr 0.000270 time 0.2766 (0.3165) loss 2.6426 (3.3290) grad_norm 1.6936 (1.9386) [2021-04-16 10:35:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][300/1251] eta 0:05:00 lr 0.000270 time 0.2790 (0.3156) loss 4.3232 (3.3343) grad_norm 1.7883 (1.9408) [2021-04-16 10:35:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][310/1251] eta 0:04:55 lr 0.000270 time 0.2744 (0.3142) loss 2.6618 (3.3397) grad_norm 1.8925 (1.9425) [2021-04-16 10:35:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][320/1251] eta 0:04:51 lr 0.000270 time 0.2824 (0.3130) loss 3.4746 (3.3443) grad_norm 2.0206 (1.9465) [2021-04-16 10:35:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][330/1251] eta 0:04:47 lr 0.000270 time 0.2733 (0.3118) loss 3.5709 (3.3503) grad_norm 1.9080 (1.9466) [2021-04-16 10:35:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][340/1251] eta 0:04:43 lr 0.000270 time 0.2830 (0.3115) loss 3.6746 (3.3581) grad_norm 2.6611 (1.9502) [2021-04-16 10:35:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][350/1251] eta 0:04:39 lr 0.000270 time 0.2662 (0.3105) loss 4.3463 (3.3684) grad_norm 2.1805 (1.9535) [2021-04-16 10:35:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][360/1251] eta 0:04:36 lr 0.000270 time 0.2928 (0.3099) loss 3.3559 (3.3762) grad_norm 2.2830 (1.9593) [2021-04-16 10:36:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][370/1251] eta 0:04:32 lr 0.000270 time 0.2641 (0.3095) loss 3.2691 (3.3759) grad_norm 1.9310 (1.9593) [2021-04-16 10:36:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][380/1251] eta 0:04:28 lr 0.000270 time 0.2897 (0.3087) loss 3.6617 (3.3778) grad_norm 1.8697 (1.9602) [2021-04-16 10:36:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][390/1251] eta 0:04:25 lr 0.000270 time 0.2536 (0.3081) loss 3.1814 (3.3758) grad_norm 1.7314 (1.9602) [2021-04-16 10:36:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][400/1251] eta 0:04:21 lr 0.000270 time 0.2604 (0.3075) loss 3.7153 (3.3736) grad_norm 1.7869 (1.9588) [2021-04-16 10:36:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][410/1251] eta 0:04:18 lr 0.000270 time 0.2793 (0.3068) loss 3.5934 (3.3749) grad_norm 1.8119 (1.9588) [2021-04-16 10:36:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][420/1251] eta 0:04:14 lr 0.000270 time 0.2775 (0.3060) loss 3.1499 (3.3789) grad_norm 1.8974 (1.9575) [2021-04-16 10:36:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][430/1251] eta 0:04:10 lr 0.000270 time 0.2666 (0.3054) loss 2.8048 (3.3681) grad_norm 1.9885 (1.9601) [2021-04-16 10:36:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][440/1251] eta 0:04:07 lr 0.000269 time 0.2950 (0.3048) loss 3.4048 (3.3715) grad_norm 1.8997 (1.9583) [2021-04-16 10:36:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][450/1251] eta 0:04:03 lr 0.000269 time 0.3028 (0.3042) loss 3.6714 (3.3693) grad_norm 1.9174 (1.9580) [2021-04-16 10:36:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][460/1251] eta 0:04:00 lr 0.000269 time 0.2846 (0.3037) loss 3.2117 (3.3666) grad_norm 1.8046 (1.9573) [2021-04-16 10:36:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][470/1251] eta 0:03:56 lr 0.000269 time 0.2546 (0.3030) loss 3.3291 (3.3712) grad_norm 1.8391 (1.9552) [2021-04-16 10:36:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][480/1251] eta 0:03:53 lr 0.000269 time 0.2852 (0.3026) loss 2.4136 (3.3690) grad_norm 1.8314 (1.9575) [2021-04-16 10:36:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][490/1251] eta 0:03:49 lr 0.000269 time 0.2707 (0.3022) loss 4.5673 (3.3664) grad_norm 1.8381 (1.9569) [2021-04-16 10:36:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][500/1251] eta 0:03:46 lr 0.000269 time 0.2881 (0.3017) loss 3.4072 (3.3705) grad_norm 2.0037 (1.9601) [2021-04-16 10:36:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][510/1251] eta 0:03:43 lr 0.000269 time 0.2689 (0.3018) loss 3.9499 (3.3728) grad_norm 2.3427 (1.9634) [2021-04-16 10:36:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][520/1251] eta 0:03:40 lr 0.000269 time 0.2797 (0.3014) loss 2.3837 (3.3700) grad_norm 1.6708 (1.9607) [2021-04-16 10:36:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][530/1251] eta 0:03:36 lr 0.000269 time 0.2655 (0.3010) loss 3.5130 (3.3724) grad_norm 1.9319 (1.9595) [2021-04-16 10:36:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][540/1251] eta 0:03:33 lr 0.000269 time 0.2729 (0.3005) loss 2.9020 (3.3710) grad_norm 1.7430 (1.9591) [2021-04-16 10:36:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][550/1251] eta 0:03:30 lr 0.000269 time 0.2777 (0.3003) loss 2.9741 (3.3712) grad_norm 2.0068 (1.9593) [2021-04-16 10:36:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][560/1251] eta 0:03:27 lr 0.000269 time 0.2703 (0.2999) loss 3.8937 (3.3699) grad_norm 1.9173 (1.9599) [2021-04-16 10:36:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][570/1251] eta 0:03:24 lr 0.000269 time 0.2899 (0.2998) loss 3.8381 (3.3704) grad_norm 1.8099 (1.9602) [2021-04-16 10:36:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][580/1251] eta 0:03:21 lr 0.000269 time 0.3042 (0.3000) loss 3.2140 (3.3725) grad_norm 2.0846 (1.9585) [2021-04-16 10:37:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][590/1251] eta 0:03:18 lr 0.000269 time 0.2829 (0.2996) loss 3.5297 (3.3695) grad_norm 2.6933 (1.9589) [2021-04-16 10:37:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][600/1251] eta 0:03:14 lr 0.000269 time 0.2508 (0.2992) loss 3.7634 (3.3665) grad_norm 1.8020 (1.9574) [2021-04-16 10:37:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][610/1251] eta 0:03:11 lr 0.000269 time 0.2580 (0.2988) loss 3.8528 (3.3716) grad_norm 1.8600 (1.9568) [2021-04-16 10:37:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][620/1251] eta 0:03:08 lr 0.000269 time 0.2823 (0.2984) loss 2.4463 (3.3679) grad_norm 1.7753 (1.9547) [2021-04-16 10:37:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][630/1251] eta 0:03:05 lr 0.000269 time 0.2830 (0.2980) loss 2.4661 (3.3662) grad_norm 1.8731 (1.9551) [2021-04-16 10:37:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][640/1251] eta 0:03:01 lr 0.000269 time 0.2627 (0.2978) loss 3.2648 (3.3649) grad_norm 1.5859 (1.9546) [2021-04-16 10:37:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][650/1251] eta 0:02:58 lr 0.000269 time 0.2864 (0.2975) loss 3.1908 (3.3655) grad_norm 2.7797 (1.9560) [2021-04-16 10:37:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][660/1251] eta 0:02:55 lr 0.000269 time 0.2693 (0.2972) loss 3.6677 (3.3670) grad_norm 1.9257 (1.9546) [2021-04-16 10:37:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][670/1251] eta 0:02:52 lr 0.000269 time 0.2975 (0.2970) loss 3.8901 (3.3684) grad_norm 2.0110 (1.9546) [2021-04-16 10:37:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][680/1251] eta 0:02:49 lr 0.000269 time 0.2574 (0.2968) loss 2.9280 (3.3674) grad_norm 2.0311 (1.9549) [2021-04-16 10:37:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][690/1251] eta 0:02:46 lr 0.000269 time 0.2811 (0.2967) loss 3.6250 (3.3651) grad_norm 2.0029 (1.9560) [2021-04-16 10:37:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][700/1251] eta 0:02:43 lr 0.000269 time 0.3132 (0.2965) loss 3.3808 (3.3649) grad_norm 2.0123 (1.9565) [2021-04-16 10:37:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][710/1251] eta 0:02:40 lr 0.000268 time 0.2728 (0.2964) loss 2.5659 (3.3644) grad_norm 1.8911 (1.9557) [2021-04-16 10:37:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][720/1251] eta 0:02:37 lr 0.000268 time 0.2684 (0.2964) loss 4.0934 (3.3639) grad_norm 1.9771 (1.9553) [2021-04-16 10:37:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][730/1251] eta 0:02:34 lr 0.000268 time 0.2828 (0.2961) loss 2.2173 (3.3622) grad_norm 1.6176 (1.9548) [2021-04-16 10:37:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][740/1251] eta 0:02:31 lr 0.000268 time 0.2926 (0.2959) loss 4.2932 (3.3598) grad_norm 2.1813 (1.9559) [2021-04-16 10:37:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][750/1251] eta 0:02:28 lr 0.000268 time 0.2588 (0.2956) loss 2.5978 (3.3593) grad_norm 2.2157 (1.9561) [2021-04-16 10:37:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][760/1251] eta 0:02:25 lr 0.000268 time 0.2744 (0.2954) loss 3.0038 (3.3558) grad_norm 2.4669 (1.9578) [2021-04-16 10:37:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][770/1251] eta 0:02:21 lr 0.000268 time 0.2933 (0.2952) loss 2.2580 (3.3522) grad_norm 2.3996 (1.9603) [2021-04-16 10:37:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][780/1251] eta 0:02:18 lr 0.000268 time 0.2783 (0.2949) loss 4.0045 (3.3518) grad_norm 1.8547 (1.9601) [2021-04-16 10:37:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][790/1251] eta 0:02:15 lr 0.000268 time 0.2886 (0.2948) loss 3.8229 (3.3483) grad_norm 1.8836 (1.9588) [2021-04-16 10:38:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][800/1251] eta 0:02:12 lr 0.000268 time 0.2787 (0.2946) loss 2.6124 (3.3485) grad_norm 1.9261 (1.9581) [2021-04-16 10:38:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][810/1251] eta 0:02:09 lr 0.000268 time 0.2762 (0.2944) loss 3.4501 (3.3480) grad_norm 1.9188 (1.9572) [2021-04-16 10:38:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][820/1251] eta 0:02:06 lr 0.000268 time 0.2802 (0.2941) loss 3.6744 (3.3467) grad_norm 1.8701 (1.9567) [2021-04-16 10:38:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][830/1251] eta 0:02:03 lr 0.000268 time 0.3014 (0.2939) loss 3.0516 (3.3456) grad_norm 1.9503 (1.9562) [2021-04-16 10:38:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][840/1251] eta 0:02:00 lr 0.000268 time 0.2742 (0.2938) loss 3.7168 (3.3481) grad_norm 1.9228 (1.9574) [2021-04-16 10:38:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][850/1251] eta 0:01:57 lr 0.000268 time 0.2762 (0.2935) loss 3.6930 (3.3449) grad_norm 2.2827 (1.9582) [2021-04-16 10:38:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][860/1251] eta 0:01:54 lr 0.000268 time 0.2655 (0.2934) loss 2.8468 (3.3449) grad_norm 2.1860 (1.9596) [2021-04-16 10:38:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][870/1251] eta 0:01:51 lr 0.000268 time 0.3031 (0.2932) loss 4.0846 (3.3456) grad_norm 2.3405 (1.9602) [2021-04-16 10:38:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][880/1251] eta 0:01:48 lr 0.000268 time 0.2867 (0.2932) loss 3.5435 (3.3454) grad_norm 1.7999 (1.9603) [2021-04-16 10:38:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][890/1251] eta 0:01:45 lr 0.000268 time 0.2964 (0.2931) loss 3.6580 (3.3476) grad_norm 1.8775 (1.9601) [2021-04-16 10:38:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][900/1251] eta 0:01:42 lr 0.000268 time 0.2726 (0.2929) loss 3.2482 (3.3482) grad_norm 1.7066 (1.9601) [2021-04-16 10:38:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][910/1251] eta 0:01:39 lr 0.000268 time 0.2927 (0.2927) loss 4.1584 (3.3523) grad_norm 1.8693 (1.9606) [2021-04-16 10:38:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][920/1251] eta 0:01:36 lr 0.000268 time 0.2869 (0.2926) loss 3.2061 (3.3487) grad_norm 1.7658 (1.9591) [2021-04-16 10:38:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][930/1251] eta 0:01:33 lr 0.000268 time 0.2878 (0.2924) loss 2.9998 (3.3524) grad_norm 1.9663 (1.9593) [2021-04-16 10:38:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][940/1251] eta 0:01:30 lr 0.000268 time 0.2941 (0.2926) loss 3.5241 (3.3547) grad_norm 2.0908 (1.9610) [2021-04-16 10:38:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][950/1251] eta 0:01:28 lr 0.000268 time 0.2870 (0.2924) loss 3.0505 (3.3531) grad_norm 1.8438 (1.9618) [2021-04-16 10:38:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][960/1251] eta 0:01:25 lr 0.000268 time 0.2657 (0.2923) loss 3.3870 (3.3532) grad_norm 1.9456 (1.9616) [2021-04-16 10:38:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][970/1251] eta 0:01:22 lr 0.000268 time 0.2707 (0.2923) loss 3.3389 (3.3541) grad_norm 1.6924 (1.9614) [2021-04-16 10:38:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][980/1251] eta 0:01:19 lr 0.000268 time 0.2808 (0.2921) loss 3.5084 (3.3541) grad_norm 1.9125 (1.9611) [2021-04-16 10:38:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][990/1251] eta 0:01:16 lr 0.000267 time 0.2808 (0.2919) loss 3.9229 (3.3529) grad_norm 1.9792 (1.9616) [2021-04-16 10:38:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][1000/1251] eta 0:01:13 lr 0.000267 time 0.2999 (0.2918) loss 3.6142 (3.3534) grad_norm 1.7451 (1.9609) [2021-04-16 10:39:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][1010/1251] eta 0:01:10 lr 0.000267 time 0.2742 (0.2916) loss 3.6836 (3.3556) grad_norm 1.9509 (1.9610) [2021-04-16 10:39:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][1020/1251] eta 0:01:07 lr 0.000267 time 0.2588 (0.2915) loss 2.8868 (3.3526) grad_norm 2.1911 (1.9611) [2021-04-16 10:39:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][1030/1251] eta 0:01:04 lr 0.000267 time 0.2790 (0.2915) loss 3.6729 (3.3521) grad_norm 2.1781 (1.9616) [2021-04-16 10:39:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][1040/1251] eta 0:01:01 lr 0.000267 time 0.2800 (0.2913) loss 3.5337 (3.3517) grad_norm 1.6920 (1.9607) [2021-04-16 10:39:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][1050/1251] eta 0:00:58 lr 0.000267 time 0.2989 (0.2912) loss 3.6764 (3.3529) grad_norm 1.8824 (1.9598) [2021-04-16 10:39:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][1060/1251] eta 0:00:55 lr 0.000267 time 0.2962 (0.2910) loss 3.0231 (3.3538) grad_norm 2.2234 (1.9605) [2021-04-16 10:39:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][1070/1251] eta 0:00:52 lr 0.000267 time 0.3107 (0.2910) loss 3.6732 (3.3554) grad_norm 2.1323 (1.9611) [2021-04-16 10:39:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][1080/1251] eta 0:00:49 lr 0.000267 time 0.2698 (0.2909) loss 3.3617 (3.3556) grad_norm 2.2215 (1.9605) [2021-04-16 10:39:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][1090/1251] eta 0:00:46 lr 0.000267 time 0.2546 (0.2908) loss 4.1088 (3.3555) grad_norm 2.0147 (1.9597) [2021-04-16 10:39:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][1100/1251] eta 0:00:43 lr 0.000267 time 0.2684 (0.2907) loss 3.4703 (3.3561) grad_norm 2.7786 (1.9608) [2021-04-16 10:39:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][1110/1251] eta 0:00:40 lr 0.000267 time 0.2610 (0.2906) loss 2.8897 (3.3537) grad_norm 2.0708 (1.9609) [2021-04-16 10:39:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][1120/1251] eta 0:00:38 lr 0.000267 time 0.2765 (0.2905) loss 3.5535 (3.3552) grad_norm 1.9434 (1.9614) [2021-04-16 10:39:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][1130/1251] eta 0:00:35 lr 0.000267 time 0.2692 (0.2905) loss 3.4705 (3.3553) grad_norm 1.9507 (1.9616) [2021-04-16 10:39:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][1140/1251] eta 0:00:32 lr 0.000267 time 0.2823 (0.2905) loss 2.8094 (3.3551) grad_norm 1.7280 (1.9607) [2021-04-16 10:39:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][1150/1251] eta 0:00:29 lr 0.000267 time 0.2928 (0.2904) loss 3.9129 (3.3559) grad_norm 2.1851 (1.9608) [2021-04-16 10:39:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][1160/1251] eta 0:00:26 lr 0.000267 time 0.2986 (0.2904) loss 3.0865 (3.3560) grad_norm 1.8629 (1.9613) [2021-04-16 10:39:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][1170/1251] eta 0:00:23 lr 0.000267 time 0.2648 (0.2903) loss 3.7776 (3.3567) grad_norm 1.7641 (1.9617) [2021-04-16 10:39:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][1180/1251] eta 0:00:20 lr 0.000267 time 0.2563 (0.2902) loss 2.7948 (3.3568) grad_norm 2.0868 (1.9612) [2021-04-16 10:39:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][1190/1251] eta 0:00:17 lr 0.000267 time 0.2693 (0.2901) loss 4.0093 (3.3562) grad_norm 2.1551 (1.9607) [2021-04-16 10:39:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][1200/1251] eta 0:00:14 lr 0.000267 time 0.2789 (0.2900) loss 3.4210 (3.3544) grad_norm 2.2280 (1.9610) [2021-04-16 10:39:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][1210/1251] eta 0:00:11 lr 0.000267 time 0.2753 (0.2899) loss 3.9483 (3.3535) grad_norm 2.2143 (1.9607) [2021-04-16 10:39:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][1220/1251] eta 0:00:08 lr 0.000267 time 0.2843 (0.2898) loss 2.7099 (3.3533) grad_norm 1.9124 (1.9601) [2021-04-16 10:40:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][1230/1251] eta 0:00:06 lr 0.000267 time 0.2637 (0.2896) loss 3.7410 (3.3527) grad_norm 1.6884 (1.9594) [2021-04-16 10:40:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][1240/1251] eta 0:00:03 lr 0.000267 time 0.2489 (0.2895) loss 2.5176 (3.3529) grad_norm 1.8012 (1.9588) [2021-04-16 10:40:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [197/300][1250/1251] eta 0:00:00 lr 0.000267 time 0.2486 (0.2891) loss 1.9972 (3.3515) grad_norm 1.8386 (1.9577) [2021-04-16 10:40:11 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 197 training takes 0:06:06 [2021-04-16 10:40:11 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_197.pth saving...... [2021-04-16 10:40:28 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_197.pth saved !!! [2021-04-16 10:40:29 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.125 (1.125) Loss 0.9985 (0.9985) Acc@1 76.074 (76.074) Acc@5 92.969 (92.969) [2021-04-16 10:40:31 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.462 (0.235) Loss 0.9228 (0.8940) Acc@1 76.660 (78.746) Acc@5 95.020 (94.531) [2021-04-16 10:40:33 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.187 (0.223) Loss 0.8691 (0.9025) Acc@1 78.906 (78.525) Acc@5 95.312 (94.508) [2021-04-16 10:40:35 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.232 (0.232) Loss 0.9517 (0.9062) Acc@1 77.441 (78.368) Acc@5 93.945 (94.459) [2021-04-16 10:40:37 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.216) Loss 0.9338 (0.9119) Acc@1 76.953 (78.216) Acc@5 94.531 (94.350) [2021-04-16 10:40:46 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 78.266 Acc@5 94.376 [2021-04-16 10:40:46 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 78.3% [2021-04-16 10:40:46 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 78.27% [2021-04-16 10:40:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][0/1251] eta 0:37:53 lr 0.000267 time 1.8175 (1.8175) loss 3.2163 (3.2163) grad_norm 2.1324 (2.1324) [2021-04-16 10:40:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][10/1251] eta 0:08:37 lr 0.000266 time 0.3015 (0.4173) loss 3.5022 (3.1908) grad_norm 2.1443 (2.0159) [2021-04-16 10:40:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][20/1251] eta 0:07:14 lr 0.000266 time 0.2681 (0.3528) loss 4.1059 (3.3099) grad_norm 1.8082 (1.9966) [2021-04-16 10:40:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][30/1251] eta 0:06:41 lr 0.000266 time 0.2776 (0.3286) loss 3.5001 (3.3428) grad_norm 1.8114 (1.9545) [2021-04-16 10:40:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][40/1251] eta 0:06:23 lr 0.000266 time 0.2879 (0.3165) loss 3.6034 (3.3327) grad_norm 1.8348 (1.9457) [2021-04-16 10:41:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][50/1251] eta 0:06:10 lr 0.000266 time 0.2644 (0.3089) loss 3.4547 (3.3196) grad_norm 2.0989 (1.9545) [2021-04-16 10:41:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][60/1251] eta 0:06:01 lr 0.000266 time 0.2618 (0.3033) loss 3.3667 (3.2822) grad_norm 1.5939 (nan) [2021-04-16 10:41:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][70/1251] eta 0:05:53 lr 0.000266 time 0.2828 (0.2991) loss 3.8602 (3.2841) grad_norm 2.3672 (nan) [2021-04-16 10:41:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][80/1251] eta 0:05:46 lr 0.000266 time 0.2929 (0.2960) loss 2.5077 (3.2769) grad_norm 2.0523 (nan) [2021-04-16 10:41:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][90/1251] eta 0:05:42 lr 0.000266 time 0.2851 (0.2950) loss 3.3281 (3.2786) grad_norm 2.2235 (nan) [2021-04-16 10:41:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][100/1251] eta 0:05:37 lr 0.000266 time 0.2766 (0.2932) loss 2.8835 (3.2677) grad_norm 1.9138 (nan) [2021-04-16 10:41:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][110/1251] eta 0:05:33 lr 0.000266 time 0.3000 (0.2919) loss 3.7401 (3.2812) grad_norm 2.1770 (nan) [2021-04-16 10:41:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][120/1251] eta 0:05:29 lr 0.000266 time 0.2615 (0.2912) loss 3.8313 (3.3172) grad_norm 1.7321 (nan) [2021-04-16 10:41:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][130/1251] eta 0:05:25 lr 0.000266 time 0.2735 (0.2902) loss 3.1571 (3.3280) grad_norm 1.8527 (nan) [2021-04-16 10:41:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][140/1251] eta 0:05:23 lr 0.000266 time 0.2995 (0.2909) loss 3.4027 (3.3452) grad_norm 1.8024 (nan) [2021-04-16 10:41:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][150/1251] eta 0:05:20 lr 0.000266 time 0.2902 (0.2910) loss 4.0432 (3.3647) grad_norm 1.9470 (nan) [2021-04-16 10:41:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][160/1251] eta 0:05:16 lr 0.000266 time 0.2859 (0.2900) loss 3.5569 (3.3719) grad_norm 1.9323 (nan) [2021-04-16 10:41:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][170/1251] eta 0:05:12 lr 0.000266 time 0.2656 (0.2894) loss 3.7423 (3.3755) grad_norm 1.6929 (nan) [2021-04-16 10:41:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][180/1251] eta 0:05:09 lr 0.000266 time 0.2749 (0.2893) loss 3.7430 (3.3663) grad_norm 2.3213 (nan) [2021-04-16 10:41:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][190/1251] eta 0:05:06 lr 0.000266 time 0.3090 (0.2886) loss 4.0013 (3.3692) grad_norm 2.1785 (nan) [2021-04-16 10:41:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][200/1251] eta 0:05:02 lr 0.000266 time 0.2847 (0.2880) loss 3.5239 (3.3766) grad_norm 1.9939 (nan) [2021-04-16 10:41:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][210/1251] eta 0:04:59 lr 0.000266 time 0.2818 (0.2874) loss 2.7952 (3.3748) grad_norm 2.5141 (nan) [2021-04-16 10:41:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][220/1251] eta 0:04:55 lr 0.000266 time 0.2731 (0.2868) loss 2.8540 (3.3685) grad_norm 2.0779 (nan) [2021-04-16 10:41:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][230/1251] eta 0:04:52 lr 0.000266 time 0.2521 (0.2862) loss 3.5374 (3.3771) grad_norm 2.5650 (nan) [2021-04-16 10:41:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][240/1251] eta 0:04:49 lr 0.000266 time 0.2715 (0.2859) loss 3.0402 (3.3798) grad_norm 1.9637 (nan) [2021-04-16 10:41:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][250/1251] eta 0:04:45 lr 0.000266 time 0.2872 (0.2855) loss 3.4107 (3.3761) grad_norm 1.7049 (nan) [2021-04-16 10:42:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][260/1251] eta 0:04:42 lr 0.000266 time 0.2705 (0.2851) loss 3.7189 (3.3714) grad_norm 2.2511 (nan) [2021-04-16 10:42:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][270/1251] eta 0:04:39 lr 0.000266 time 0.2794 (0.2848) loss 3.6111 (3.3699) grad_norm 1.9690 (nan) [2021-04-16 10:42:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][280/1251] eta 0:04:36 lr 0.000266 time 0.2875 (0.2845) loss 2.6474 (3.3669) grad_norm 2.3430 (nan) [2021-04-16 10:42:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][290/1251] eta 0:04:33 lr 0.000265 time 0.2811 (0.2842) loss 2.7576 (3.3663) grad_norm 1.9626 (nan) [2021-04-16 10:42:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][300/1251] eta 0:04:30 lr 0.000265 time 0.2534 (0.2840) loss 3.2725 (3.3655) grad_norm 2.0316 (nan) [2021-04-16 10:42:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][310/1251] eta 0:04:27 lr 0.000265 time 0.2960 (0.2839) loss 3.0879 (3.3608) grad_norm 2.2652 (nan) [2021-04-16 10:42:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][320/1251] eta 0:04:24 lr 0.000265 time 0.2562 (0.2836) loss 2.5746 (3.3579) grad_norm 2.2998 (nan) [2021-04-16 10:42:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][330/1251] eta 0:04:21 lr 0.000265 time 0.2793 (0.2834) loss 3.2628 (3.3590) grad_norm 2.1145 (nan) [2021-04-16 10:42:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][340/1251] eta 0:04:18 lr 0.000265 time 0.2547 (0.2837) loss 3.8147 (3.3617) grad_norm 2.0049 (nan) [2021-04-16 10:42:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][350/1251] eta 0:04:16 lr 0.000265 time 0.4385 (0.2846) loss 2.5420 (3.3487) grad_norm 2.0561 (nan) [2021-04-16 10:42:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][360/1251] eta 0:04:13 lr 0.000265 time 0.2895 (0.2850) loss 3.6598 (3.3514) grad_norm 2.0917 (nan) [2021-04-16 10:42:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][370/1251] eta 0:04:10 lr 0.000265 time 0.2814 (0.2847) loss 3.9134 (3.3504) grad_norm 2.1138 (nan) [2021-04-16 10:42:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][380/1251] eta 0:04:07 lr 0.000265 time 0.2793 (0.2846) loss 3.0245 (3.3575) grad_norm 1.9194 (nan) [2021-04-16 10:42:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][390/1251] eta 0:04:04 lr 0.000265 time 0.2667 (0.2843) loss 3.8489 (3.3557) grad_norm 1.9532 (nan) [2021-04-16 10:42:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][400/1251] eta 0:04:01 lr 0.000265 time 0.2758 (0.2843) loss 3.8801 (3.3531) grad_norm 2.3184 (nan) [2021-04-16 10:42:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][410/1251] eta 0:03:58 lr 0.000265 time 0.2832 (0.2842) loss 3.5892 (3.3482) grad_norm 1.9348 (nan) [2021-04-16 10:42:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][420/1251] eta 0:03:56 lr 0.000265 time 0.2673 (0.2840) loss 3.5247 (3.3447) grad_norm 2.0116 (nan) [2021-04-16 10:42:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][430/1251] eta 0:03:53 lr 0.000265 time 0.2608 (0.2839) loss 3.1723 (3.3406) grad_norm 1.8317 (nan) [2021-04-16 10:42:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][440/1251] eta 0:03:50 lr 0.000265 time 0.2696 (0.2836) loss 3.5431 (3.3339) grad_norm 1.8649 (nan) [2021-04-16 10:42:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][450/1251] eta 0:03:47 lr 0.000265 time 0.2691 (0.2836) loss 3.1199 (3.3305) grad_norm 1.8173 (nan) [2021-04-16 10:42:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][460/1251] eta 0:03:44 lr 0.000265 time 0.2787 (0.2833) loss 3.3310 (3.3260) grad_norm 1.8739 (nan) [2021-04-16 10:43:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][470/1251] eta 0:03:41 lr 0.000265 time 0.3124 (0.2833) loss 3.7510 (3.3245) grad_norm 1.9843 (nan) [2021-04-16 10:43:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][480/1251] eta 0:03:38 lr 0.000265 time 0.2759 (0.2836) loss 2.4674 (3.3207) grad_norm 1.6532 (nan) [2021-04-16 10:43:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][490/1251] eta 0:03:35 lr 0.000265 time 0.2950 (0.2835) loss 3.3465 (3.3211) grad_norm 1.9722 (nan) [2021-04-16 10:43:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][500/1251] eta 0:03:32 lr 0.000265 time 0.2852 (0.2833) loss 3.5785 (3.3236) grad_norm 2.5547 (nan) [2021-04-16 10:43:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][510/1251] eta 0:03:29 lr 0.000265 time 0.2726 (0.2832) loss 3.9828 (3.3277) grad_norm 2.1908 (nan) [2021-04-16 10:43:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][520/1251] eta 0:03:26 lr 0.000265 time 0.2878 (0.2831) loss 3.5793 (3.3279) grad_norm 2.4765 (nan) [2021-04-16 10:43:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][530/1251] eta 0:03:24 lr 0.000265 time 0.2667 (0.2830) loss 2.6203 (3.3229) grad_norm 1.7890 (nan) [2021-04-16 10:43:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][540/1251] eta 0:03:21 lr 0.000265 time 0.2634 (0.2828) loss 3.6218 (3.3254) grad_norm 1.9076 (nan) [2021-04-16 10:43:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][550/1251] eta 0:03:18 lr 0.000265 time 0.2610 (0.2826) loss 3.1816 (3.3287) grad_norm 2.0427 (nan) [2021-04-16 10:43:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][560/1251] eta 0:03:15 lr 0.000265 time 0.3119 (0.2826) loss 4.0739 (3.3251) grad_norm 1.7434 (nan) [2021-04-16 10:43:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][570/1251] eta 0:03:12 lr 0.000264 time 0.3009 (0.2828) loss 3.7025 (3.3304) grad_norm 1.8479 (nan) [2021-04-16 10:43:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][580/1251] eta 0:03:09 lr 0.000264 time 0.2839 (0.2830) loss 3.0626 (3.3273) grad_norm 1.7846 (nan) [2021-04-16 10:43:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][590/1251] eta 0:03:06 lr 0.000264 time 0.2698 (0.2828) loss 2.8195 (3.3275) grad_norm 1.8984 (nan) [2021-04-16 10:43:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][600/1251] eta 0:03:04 lr 0.000264 time 0.2785 (0.2827) loss 3.7295 (3.3269) grad_norm 2.2500 (nan) [2021-04-16 10:43:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][610/1251] eta 0:03:01 lr 0.000264 time 0.2987 (0.2826) loss 3.7084 (3.3288) grad_norm 2.1592 (nan) [2021-04-16 10:43:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][620/1251] eta 0:02:58 lr 0.000264 time 0.2659 (0.2825) loss 2.6110 (3.3300) grad_norm 1.9143 (nan) [2021-04-16 10:43:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][630/1251] eta 0:02:55 lr 0.000264 time 0.3016 (0.2824) loss 3.4753 (3.3301) grad_norm 2.0019 (nan) [2021-04-16 10:43:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][640/1251] eta 0:02:52 lr 0.000264 time 0.2717 (0.2824) loss 2.7905 (3.3259) grad_norm 1.7675 (nan) [2021-04-16 10:43:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][650/1251] eta 0:02:49 lr 0.000264 time 0.2545 (0.2823) loss 2.6705 (3.3244) grad_norm 1.7387 (nan) [2021-04-16 10:43:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][660/1251] eta 0:02:46 lr 0.000264 time 0.2667 (0.2823) loss 3.2778 (3.3265) grad_norm 2.0739 (nan) [2021-04-16 10:43:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][670/1251] eta 0:02:44 lr 0.000264 time 0.2721 (0.2823) loss 4.1357 (3.3290) grad_norm 1.9692 (nan) [2021-04-16 10:43:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][680/1251] eta 0:02:41 lr 0.000264 time 0.2857 (0.2822) loss 3.4153 (3.3310) grad_norm 1.9517 (nan) [2021-04-16 10:44:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][690/1251] eta 0:02:38 lr 0.000264 time 0.2588 (0.2823) loss 4.1582 (3.3332) grad_norm 1.9758 (nan) [2021-04-16 10:44:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][700/1251] eta 0:02:35 lr 0.000264 time 0.2624 (0.2822) loss 3.2412 (3.3374) grad_norm 1.8496 (nan) [2021-04-16 10:44:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][710/1251] eta 0:02:32 lr 0.000264 time 0.2804 (0.2822) loss 3.8937 (3.3367) grad_norm 2.1759 (nan) [2021-04-16 10:44:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][720/1251] eta 0:02:29 lr 0.000264 time 0.2707 (0.2824) loss 2.8868 (3.3361) grad_norm 2.2883 (nan) [2021-04-16 10:44:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][730/1251] eta 0:02:27 lr 0.000264 time 0.2745 (0.2825) loss 3.8466 (3.3356) grad_norm 2.0581 (nan) [2021-04-16 10:44:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][740/1251] eta 0:02:24 lr 0.000264 time 0.2705 (0.2824) loss 3.7398 (3.3414) grad_norm 1.7641 (nan) [2021-04-16 10:44:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][750/1251] eta 0:02:21 lr 0.000264 time 0.2792 (0.2823) loss 3.4915 (3.3383) grad_norm 1.7183 (nan) [2021-04-16 10:44:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][760/1251] eta 0:02:18 lr 0.000264 time 0.2675 (0.2824) loss 3.7671 (3.3369) grad_norm 1.7559 (nan) [2021-04-16 10:44:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][770/1251] eta 0:02:15 lr 0.000264 time 0.2631 (0.2823) loss 4.1493 (3.3348) grad_norm 1.9837 (nan) [2021-04-16 10:44:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][780/1251] eta 0:02:12 lr 0.000264 time 0.3201 (0.2823) loss 3.0907 (3.3327) grad_norm 2.0780 (nan) [2021-04-16 10:44:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][790/1251] eta 0:02:10 lr 0.000264 time 0.2809 (0.2824) loss 2.5333 (3.3336) grad_norm 1.8926 (nan) [2021-04-16 10:44:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][800/1251] eta 0:02:07 lr 0.000264 time 0.2801 (0.2823) loss 3.4043 (3.3333) grad_norm 1.8751 (nan) [2021-04-16 10:44:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][810/1251] eta 0:02:04 lr 0.000264 time 0.2873 (0.2823) loss 3.0775 (3.3356) grad_norm 1.9352 (nan) [2021-04-16 10:44:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][820/1251] eta 0:02:01 lr 0.000264 time 0.2710 (0.2822) loss 3.9866 (3.3357) grad_norm 2.1133 (nan) [2021-04-16 10:44:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][830/1251] eta 0:01:58 lr 0.000264 time 0.2662 (0.2822) loss 3.0605 (3.3361) grad_norm 1.7235 (nan) [2021-04-16 10:44:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][840/1251] eta 0:01:55 lr 0.000263 time 0.2679 (0.2821) loss 3.8821 (3.3363) grad_norm 1.7358 (nan) [2021-04-16 10:44:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][850/1251] eta 0:01:53 lr 0.000263 time 0.2553 (0.2820) loss 3.3392 (3.3375) grad_norm 2.3316 (nan) [2021-04-16 10:44:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][860/1251] eta 0:01:50 lr 0.000263 time 0.2815 (0.2820) loss 3.6575 (3.3360) grad_norm 2.4991 (nan) [2021-04-16 10:44:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][870/1251] eta 0:01:47 lr 0.000263 time 0.2741 (0.2820) loss 3.8786 (3.3355) grad_norm 2.1321 (nan) [2021-04-16 10:44:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][880/1251] eta 0:01:44 lr 0.000263 time 0.2661 (0.2819) loss 3.8085 (3.3348) grad_norm 2.1928 (nan) [2021-04-16 10:44:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][890/1251] eta 0:01:41 lr 0.000263 time 0.2432 (0.2818) loss 3.6110 (3.3339) grad_norm 1.9334 (nan) [2021-04-16 10:45:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][900/1251] eta 0:01:38 lr 0.000263 time 0.2784 (0.2818) loss 3.1771 (3.3346) grad_norm 1.7052 (nan) [2021-04-16 10:45:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][910/1251] eta 0:01:36 lr 0.000263 time 0.2913 (0.2817) loss 3.0387 (3.3313) grad_norm 1.8822 (nan) [2021-04-16 10:45:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][920/1251] eta 0:01:33 lr 0.000263 time 0.2514 (0.2817) loss 4.2375 (3.3337) grad_norm 1.9064 (nan) [2021-04-16 10:45:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][930/1251] eta 0:01:30 lr 0.000263 time 0.2690 (0.2820) loss 3.5916 (3.3355) grad_norm 1.7427 (nan) [2021-04-16 10:45:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][940/1251] eta 0:01:27 lr 0.000263 time 0.2674 (0.2821) loss 3.1348 (3.3359) grad_norm 1.8359 (nan) [2021-04-16 10:45:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][950/1251] eta 0:01:24 lr 0.000263 time 0.2672 (0.2820) loss 3.5450 (3.3353) grad_norm 2.0337 (nan) [2021-04-16 10:45:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][960/1251] eta 0:01:22 lr 0.000263 time 0.2530 (0.2820) loss 3.7692 (3.3366) grad_norm 2.3208 (nan) [2021-04-16 10:45:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][970/1251] eta 0:01:19 lr 0.000263 time 0.2875 (0.2819) loss 3.2442 (3.3384) grad_norm 2.0342 (nan) [2021-04-16 10:45:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][980/1251] eta 0:01:16 lr 0.000263 time 0.2541 (0.2819) loss 3.2886 (3.3391) grad_norm 1.9159 (nan) [2021-04-16 10:45:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][990/1251] eta 0:01:13 lr 0.000263 time 0.2857 (0.2819) loss 3.5776 (3.3406) grad_norm 1.9526 (nan) [2021-04-16 10:45:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][1000/1251] eta 0:01:10 lr 0.000263 time 0.2992 (0.2819) loss 2.6935 (3.3400) grad_norm 2.1844 (nan) [2021-04-16 10:45:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][1010/1251] eta 0:01:07 lr 0.000263 time 0.2690 (0.2818) loss 2.0094 (3.3389) grad_norm 2.1878 (nan) [2021-04-16 10:45:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][1020/1251] eta 0:01:05 lr 0.000263 time 0.2826 (0.2818) loss 3.3738 (3.3368) grad_norm 2.2729 (nan) [2021-04-16 10:45:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][1030/1251] eta 0:01:02 lr 0.000263 time 0.2864 (0.2818) loss 3.5352 (3.3357) grad_norm 2.0890 (nan) [2021-04-16 10:45:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][1040/1251] eta 0:00:59 lr 0.000263 time 0.2829 (0.2817) loss 3.4125 (3.3364) grad_norm 1.7257 (nan) [2021-04-16 10:45:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][1050/1251] eta 0:00:56 lr 0.000263 time 0.4557 (0.2819) loss 3.4842 (3.3397) grad_norm 1.9103 (nan) [2021-04-16 10:45:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][1060/1251] eta 0:00:53 lr 0.000263 time 0.2673 (0.2818) loss 3.6584 (3.3392) grad_norm 1.9320 (nan) [2021-04-16 10:45:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][1070/1251] eta 0:00:50 lr 0.000263 time 0.2831 (0.2818) loss 2.6264 (3.3370) grad_norm 1.7141 (nan) [2021-04-16 10:45:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][1080/1251] eta 0:00:48 lr 0.000263 time 0.2901 (0.2818) loss 3.6052 (3.3382) grad_norm 2.0582 (nan) [2021-04-16 10:45:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][1090/1251] eta 0:00:45 lr 0.000263 time 0.2754 (0.2817) loss 3.6469 (3.3373) grad_norm 2.2116 (nan) [2021-04-16 10:45:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][1100/1251] eta 0:00:42 lr 0.000263 time 0.2628 (0.2817) loss 3.0715 (3.3384) grad_norm 1.8139 (nan) [2021-04-16 10:45:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][1110/1251] eta 0:00:39 lr 0.000263 time 0.2775 (0.2817) loss 3.1688 (3.3358) grad_norm 2.0048 (nan) [2021-04-16 10:46:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][1120/1251] eta 0:00:36 lr 0.000262 time 0.2618 (0.2817) loss 3.5744 (3.3329) grad_norm 2.0759 (nan) [2021-04-16 10:46:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][1130/1251] eta 0:00:34 lr 0.000262 time 0.2902 (0.2816) loss 3.9745 (3.3337) grad_norm 1.6183 (nan) [2021-04-16 10:46:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][1140/1251] eta 0:00:31 lr 0.000262 time 0.2996 (0.2816) loss 2.0800 (3.3312) grad_norm 2.0238 (nan) [2021-04-16 10:46:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][1150/1251] eta 0:00:28 lr 0.000262 time 0.3970 (0.2818) loss 3.2624 (3.3286) grad_norm 2.0544 (nan) [2021-04-16 10:46:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][1160/1251] eta 0:00:25 lr 0.000262 time 0.2669 (0.2819) loss 2.8335 (3.3286) grad_norm 1.9495 (nan) [2021-04-16 10:46:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][1170/1251] eta 0:00:22 lr 0.000262 time 0.2640 (0.2818) loss 3.8144 (3.3297) grad_norm 1.8743 (nan) [2021-04-16 10:46:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][1180/1251] eta 0:00:20 lr 0.000262 time 0.2619 (0.2818) loss 3.3384 (3.3273) grad_norm 1.7797 (nan) [2021-04-16 10:46:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][1190/1251] eta 0:00:17 lr 0.000262 time 0.2810 (0.2817) loss 3.7191 (3.3285) grad_norm 2.0211 (nan) [2021-04-16 10:46:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][1200/1251] eta 0:00:14 lr 0.000262 time 0.2815 (0.2816) loss 3.4454 (3.3279) grad_norm 1.7535 (nan) [2021-04-16 10:46:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][1210/1251] eta 0:00:11 lr 0.000262 time 0.2596 (0.2815) loss 3.3330 (3.3272) grad_norm 2.3481 (nan) [2021-04-16 10:46:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][1220/1251] eta 0:00:08 lr 0.000262 time 0.2950 (0.2815) loss 3.8907 (3.3266) grad_norm 1.6955 (nan) [2021-04-16 10:46:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][1230/1251] eta 0:00:05 lr 0.000262 time 0.2810 (0.2815) loss 2.5498 (3.3279) grad_norm 1.6156 (nan) [2021-04-16 10:46:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][1240/1251] eta 0:00:03 lr 0.000262 time 0.2642 (0.2815) loss 3.2015 (3.3289) grad_norm 2.0146 (nan) [2021-04-16 10:46:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [198/300][1250/1251] eta 0:00:00 lr 0.000262 time 0.2484 (0.2812) loss 2.2822 (3.3296) grad_norm 1.8534 (nan) [2021-04-16 10:46:55 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 198 training takes 0:06:08 [2021-04-16 10:46:55 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_198.pth saving...... [2021-04-16 10:47:10 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_198.pth saved !!! [2021-04-16 10:47:12 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.147 (1.147) Loss 1.0106 (1.0106) Acc@1 75.488 (75.488) Acc@5 93.164 (93.164) [2021-04-16 10:47:13 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.110 (0.220) Loss 0.9126 (0.9159) Acc@1 77.539 (78.240) Acc@5 94.434 (94.398) [2021-04-16 10:47:15 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.118 (0.230) Loss 0.8894 (0.9175) Acc@1 78.809 (78.037) Acc@5 94.727 (94.387) [2021-04-16 10:47:18 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.120 (0.245) Loss 0.9209 (0.9147) Acc@1 78.125 (78.122) Acc@5 94.531 (94.452) [2021-04-16 10:47:19 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.136 (0.210) Loss 0.8977 (0.9116) Acc@1 77.734 (78.161) Acc@5 93.652 (94.410) [2021-04-16 10:47:31 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 78.156 Acc@5 94.384 [2021-04-16 10:47:31 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 78.2% [2021-04-16 10:47:31 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 78.27% [2021-04-16 10:47:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][0/1251] eta 2:31:39 lr 0.000262 time 7.2741 (7.2741) loss 2.4625 (2.4625) grad_norm 1.7805 (1.7805) [2021-04-16 10:47:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][10/1251] eta 0:18:47 lr 0.000262 time 0.2726 (0.9087) loss 3.4063 (3.4010) grad_norm 1.8208 (1.9188) [2021-04-16 10:47:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][20/1251] eta 0:12:32 lr 0.000262 time 0.2830 (0.6112) loss 3.8896 (3.3749) grad_norm 2.7927 (1.9833) [2021-04-16 10:47:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][30/1251] eta 0:10:14 lr 0.000262 time 0.2848 (0.5032) loss 3.5605 (3.3359) grad_norm 2.4603 (1.9679) [2021-04-16 10:47:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][40/1251] eta 0:09:06 lr 0.000262 time 0.3167 (0.4515) loss 3.8546 (3.3199) grad_norm 2.1232 (1.9507) [2021-04-16 10:47:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][50/1251] eta 0:08:20 lr 0.000262 time 0.2666 (0.4166) loss 3.0445 (3.3253) grad_norm 1.7124 (1.9456) [2021-04-16 10:47:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][60/1251] eta 0:07:50 lr 0.000262 time 0.3036 (0.3953) loss 3.7300 (3.3282) grad_norm 1.8433 (1.9437) [2021-04-16 10:47:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][70/1251] eta 0:07:27 lr 0.000262 time 0.2571 (0.3790) loss 3.6102 (3.3063) grad_norm 2.1097 (1.9410) [2021-04-16 10:48:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][80/1251] eta 0:07:10 lr 0.000262 time 0.2506 (0.3673) loss 3.7512 (3.3316) grad_norm 1.7570 (1.9410) [2021-04-16 10:48:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][90/1251] eta 0:06:55 lr 0.000262 time 0.2466 (0.3580) loss 3.8131 (3.3260) grad_norm 2.1697 (1.9394) [2021-04-16 10:48:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][100/1251] eta 0:06:43 lr 0.000262 time 0.2517 (0.3506) loss 2.2871 (3.3052) grad_norm 2.1916 (1.9514) [2021-04-16 10:48:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][110/1251] eta 0:06:32 lr 0.000262 time 0.2916 (0.3443) loss 3.6847 (3.3112) grad_norm 1.9357 (1.9494) [2021-04-16 10:48:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][120/1251] eta 0:06:25 lr 0.000262 time 0.3040 (0.3407) loss 3.4100 (3.3081) grad_norm 2.5506 (1.9546) [2021-04-16 10:48:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][130/1251] eta 0:06:18 lr 0.000262 time 0.2698 (0.3381) loss 3.0660 (3.3056) grad_norm 2.1886 (1.9665) [2021-04-16 10:48:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][140/1251] eta 0:06:11 lr 0.000261 time 0.2806 (0.3344) loss 3.6219 (3.2888) grad_norm 1.7902 (1.9663) [2021-04-16 10:48:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][150/1251] eta 0:06:05 lr 0.000261 time 0.2801 (0.3322) loss 3.5411 (3.2955) grad_norm 2.0413 (1.9699) [2021-04-16 10:48:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][160/1251] eta 0:05:58 lr 0.000261 time 0.3068 (0.3290) loss 2.4943 (3.2972) grad_norm 2.0120 (1.9735) [2021-04-16 10:48:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][170/1251] eta 0:05:52 lr 0.000261 time 0.2929 (0.3265) loss 3.4351 (3.3016) grad_norm 2.2357 (1.9806) [2021-04-16 10:48:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][180/1251] eta 0:05:47 lr 0.000261 time 0.2759 (0.3240) loss 3.1580 (3.3123) grad_norm 2.0218 (1.9781) [2021-04-16 10:48:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][190/1251] eta 0:05:41 lr 0.000261 time 0.2854 (0.3220) loss 3.5809 (3.3153) grad_norm 1.8726 (1.9724) [2021-04-16 10:48:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][200/1251] eta 0:05:36 lr 0.000261 time 0.2901 (0.3203) loss 2.6842 (3.3100) grad_norm 2.1540 (1.9714) [2021-04-16 10:48:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][210/1251] eta 0:05:31 lr 0.000261 time 0.3030 (0.3189) loss 3.2210 (3.3259) grad_norm 1.7985 (1.9739) [2021-04-16 10:48:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][220/1251] eta 0:05:27 lr 0.000261 time 0.2845 (0.3174) loss 3.5806 (3.3397) grad_norm 1.7427 (1.9711) [2021-04-16 10:48:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][230/1251] eta 0:05:23 lr 0.000261 time 0.2877 (0.3164) loss 2.6432 (3.3453) grad_norm 2.1379 (1.9760) [2021-04-16 10:48:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][240/1251] eta 0:05:18 lr 0.000261 time 0.2713 (0.3150) loss 2.9387 (3.3501) grad_norm 2.0293 (1.9857) [2021-04-16 10:48:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][250/1251] eta 0:05:13 lr 0.000261 time 0.2806 (0.3136) loss 3.5327 (3.3563) grad_norm 2.1840 (1.9953) [2021-04-16 10:48:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][260/1251] eta 0:05:10 lr 0.000261 time 0.4354 (0.3133) loss 4.1090 (3.3666) grad_norm 2.2180 (1.9972) [2021-04-16 10:48:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][270/1251] eta 0:05:06 lr 0.000261 time 0.2700 (0.3124) loss 3.4370 (3.3614) grad_norm 1.8908 (1.9981) [2021-04-16 10:48:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][280/1251] eta 0:05:02 lr 0.000261 time 0.3016 (0.3115) loss 3.8272 (3.3540) grad_norm 1.7923 (1.9941) [2021-04-16 10:49:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][290/1251] eta 0:04:58 lr 0.000261 time 0.2907 (0.3104) loss 2.8423 (3.3514) grad_norm 1.8013 (1.9964) [2021-04-16 10:49:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][300/1251] eta 0:04:54 lr 0.000261 time 0.2910 (0.3094) loss 3.0996 (3.3523) grad_norm 2.0508 (1.9985) [2021-04-16 10:49:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][310/1251] eta 0:04:50 lr 0.000261 time 0.2786 (0.3084) loss 3.4573 (3.3453) grad_norm 1.7071 (2.0029) [2021-04-16 10:49:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][320/1251] eta 0:04:46 lr 0.000261 time 0.2816 (0.3073) loss 3.7840 (3.3463) grad_norm 2.1342 (2.0023) [2021-04-16 10:49:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][330/1251] eta 0:04:42 lr 0.000261 time 0.2972 (0.3065) loss 2.6707 (3.3516) grad_norm 1.7663 (1.9979) [2021-04-16 10:49:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][340/1251] eta 0:04:38 lr 0.000261 time 0.2761 (0.3057) loss 3.9225 (3.3523) grad_norm 1.9139 (1.9995) [2021-04-16 10:49:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][350/1251] eta 0:04:34 lr 0.000261 time 0.2870 (0.3051) loss 3.1527 (3.3490) grad_norm 1.6551 (1.9986) [2021-04-16 10:49:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][360/1251] eta 0:04:31 lr 0.000261 time 0.2830 (0.3051) loss 3.8156 (3.3502) grad_norm 2.2550 (1.9990) [2021-04-16 10:49:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][370/1251] eta 0:04:28 lr 0.000261 time 0.2921 (0.3045) loss 3.3357 (3.3523) grad_norm 1.9806 (1.9992) [2021-04-16 10:49:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][380/1251] eta 0:04:24 lr 0.000261 time 0.2842 (0.3038) loss 3.0235 (3.3534) grad_norm 1.9066 (1.9967) [2021-04-16 10:49:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][390/1251] eta 0:04:21 lr 0.000261 time 0.2739 (0.3034) loss 3.8872 (3.3605) grad_norm 1.8690 (1.9946) [2021-04-16 10:49:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][400/1251] eta 0:04:17 lr 0.000261 time 0.3165 (0.3029) loss 3.3109 (3.3627) grad_norm 1.6691 (1.9957) [2021-04-16 10:49:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][410/1251] eta 0:04:14 lr 0.000261 time 0.2849 (0.3024) loss 4.2172 (3.3650) grad_norm 1.8652 (1.9957) [2021-04-16 10:49:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][420/1251] eta 0:04:10 lr 0.000260 time 0.2682 (0.3019) loss 3.9551 (3.3621) grad_norm 2.0189 (1.9973) [2021-04-16 10:49:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][430/1251] eta 0:04:07 lr 0.000260 time 0.2826 (0.3014) loss 2.7221 (3.3563) grad_norm 2.6349 (1.9992) [2021-04-16 10:49:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][440/1251] eta 0:04:04 lr 0.000260 time 0.2985 (0.3011) loss 3.9222 (3.3586) grad_norm 1.7245 (1.9994) [2021-04-16 10:49:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][450/1251] eta 0:04:00 lr 0.000260 time 0.2759 (0.3006) loss 3.8675 (3.3634) grad_norm 2.2022 (1.9999) [2021-04-16 10:49:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][460/1251] eta 0:03:57 lr 0.000260 time 0.2801 (0.3002) loss 3.0938 (3.3638) grad_norm 2.0166 (2.0033) [2021-04-16 10:49:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][470/1251] eta 0:03:54 lr 0.000260 time 0.2744 (0.2997) loss 2.8873 (3.3647) grad_norm 1.8053 (2.0027) [2021-04-16 10:49:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][480/1251] eta 0:03:50 lr 0.000260 time 0.3054 (0.2995) loss 3.8112 (3.3733) grad_norm 1.8461 (2.0016) [2021-04-16 10:49:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][490/1251] eta 0:03:47 lr 0.000260 time 0.2908 (0.2991) loss 3.1159 (3.3714) grad_norm 1.9014 (2.0018) [2021-04-16 10:50:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][500/1251] eta 0:03:44 lr 0.000260 time 0.2940 (0.2987) loss 3.0264 (3.3699) grad_norm 1.9199 (2.0012) [2021-04-16 10:50:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][510/1251] eta 0:03:41 lr 0.000260 time 0.2826 (0.2985) loss 2.4696 (3.3750) grad_norm 1.9006 (2.0008) [2021-04-16 10:50:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][520/1251] eta 0:03:38 lr 0.000260 time 0.3206 (0.2987) loss 3.6530 (3.3742) grad_norm 2.1317 (2.0021) [2021-04-16 10:50:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][530/1251] eta 0:03:35 lr 0.000260 time 0.2857 (0.2984) loss 3.0620 (3.3722) grad_norm 1.9205 (2.0034) [2021-04-16 10:50:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][540/1251] eta 0:03:31 lr 0.000260 time 0.2710 (0.2980) loss 2.7461 (3.3725) grad_norm 1.9197 (2.0019) [2021-04-16 10:50:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][550/1251] eta 0:03:28 lr 0.000260 time 0.2853 (0.2977) loss 2.8847 (3.3741) grad_norm 2.3614 (2.0031) [2021-04-16 10:50:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][560/1251] eta 0:03:25 lr 0.000260 time 0.2551 (0.2974) loss 3.7488 (3.3743) grad_norm 1.8862 (2.0036) [2021-04-16 10:50:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][570/1251] eta 0:03:22 lr 0.000260 time 0.2734 (0.2975) loss 3.7776 (3.3760) grad_norm 1.8481 (2.0021) [2021-04-16 10:50:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][580/1251] eta 0:03:19 lr 0.000260 time 0.2838 (0.2976) loss 4.0189 (3.3745) grad_norm 1.8967 (2.0018) [2021-04-16 10:50:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][590/1251] eta 0:03:16 lr 0.000260 time 0.2805 (0.2973) loss 3.7502 (3.3734) grad_norm 1.8390 (2.0022) [2021-04-16 10:50:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][600/1251] eta 0:03:13 lr 0.000260 time 0.3061 (0.2971) loss 3.5355 (3.3770) grad_norm 1.6240 (1.9995) [2021-04-16 10:50:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][610/1251] eta 0:03:10 lr 0.000260 time 0.2952 (0.2969) loss 3.7008 (3.3775) grad_norm 1.7622 (1.9979) [2021-04-16 10:50:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][620/1251] eta 0:03:07 lr 0.000260 time 0.2858 (0.2966) loss 2.2051 (3.3747) grad_norm 1.7384 (1.9963) [2021-04-16 10:50:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][630/1251] eta 0:03:04 lr 0.000260 time 0.2998 (0.2964) loss 4.1398 (3.3799) grad_norm 2.0315 (1.9947) [2021-04-16 10:50:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][640/1251] eta 0:03:00 lr 0.000260 time 0.2699 (0.2961) loss 4.1106 (3.3809) grad_norm 1.7116 (1.9929) [2021-04-16 10:50:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][650/1251] eta 0:02:57 lr 0.000260 time 0.2845 (0.2959) loss 3.5294 (3.3761) grad_norm 1.8164 (1.9909) [2021-04-16 10:50:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][660/1251] eta 0:02:54 lr 0.000260 time 0.2986 (0.2957) loss 3.4387 (3.3735) grad_norm 1.9401 (1.9914) [2021-04-16 10:50:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][670/1251] eta 0:02:51 lr 0.000260 time 0.2514 (0.2955) loss 3.7987 (3.3769) grad_norm 2.1207 (1.9917) [2021-04-16 10:50:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][680/1251] eta 0:02:48 lr 0.000260 time 0.2815 (0.2953) loss 3.6513 (3.3739) grad_norm 2.0403 (1.9926) [2021-04-16 10:50:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][690/1251] eta 0:02:45 lr 0.000260 time 0.2818 (0.2954) loss 2.5245 (3.3766) grad_norm 1.8828 (1.9936) [2021-04-16 10:50:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][700/1251] eta 0:02:42 lr 0.000259 time 0.4419 (0.2955) loss 4.2123 (3.3772) grad_norm 2.0889 (1.9944) [2021-04-16 10:51:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][710/1251] eta 0:02:39 lr 0.000259 time 0.2646 (0.2953) loss 3.6694 (3.3759) grad_norm 1.8627 (1.9935) [2021-04-16 10:51:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][720/1251] eta 0:02:36 lr 0.000259 time 0.2650 (0.2952) loss 2.4588 (3.3692) grad_norm 2.2829 (1.9946) [2021-04-16 10:51:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][730/1251] eta 0:02:33 lr 0.000259 time 0.2795 (0.2952) loss 2.6925 (3.3675) grad_norm 2.3991 (1.9965) [2021-04-16 10:51:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][740/1251] eta 0:02:30 lr 0.000259 time 0.2846 (0.2950) loss 3.9097 (3.3679) grad_norm 2.1365 (1.9953) [2021-04-16 10:51:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][750/1251] eta 0:02:27 lr 0.000259 time 0.2719 (0.2948) loss 3.7944 (3.3677) grad_norm 1.9241 (1.9951) [2021-04-16 10:51:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][760/1251] eta 0:02:24 lr 0.000259 time 0.2785 (0.2947) loss 3.7501 (3.3709) grad_norm 1.8606 (1.9940) [2021-04-16 10:51:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][770/1251] eta 0:02:21 lr 0.000259 time 0.2851 (0.2945) loss 2.9394 (3.3693) grad_norm 2.4045 (1.9938) [2021-04-16 10:51:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][780/1251] eta 0:02:18 lr 0.000259 time 0.2748 (0.2942) loss 3.6484 (3.3672) grad_norm 1.8466 (1.9926) [2021-04-16 10:51:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][790/1251] eta 0:02:15 lr 0.000259 time 0.2744 (0.2940) loss 3.8054 (3.3644) grad_norm 1.8560 (1.9918) [2021-04-16 10:51:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][800/1251] eta 0:02:12 lr 0.000259 time 0.2811 (0.2939) loss 3.9124 (3.3668) grad_norm 1.8798 (1.9911) [2021-04-16 10:51:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][810/1251] eta 0:02:09 lr 0.000259 time 0.2828 (0.2938) loss 3.9035 (3.3645) grad_norm 2.1311 (1.9900) [2021-04-16 10:51:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][820/1251] eta 0:02:06 lr 0.000259 time 0.2747 (0.2936) loss 2.5680 (3.3594) grad_norm 1.9621 (1.9921) [2021-04-16 10:51:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][830/1251] eta 0:02:03 lr 0.000259 time 0.2999 (0.2935) loss 3.2337 (3.3586) grad_norm 1.9307 (1.9917) [2021-04-16 10:51:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][840/1251] eta 0:02:00 lr 0.000259 time 0.2804 (0.2935) loss 3.5288 (3.3605) grad_norm 1.8249 (1.9909) [2021-04-16 10:51:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][850/1251] eta 0:01:57 lr 0.000259 time 0.2949 (0.2933) loss 2.6605 (3.3608) grad_norm 2.1518 (1.9919) [2021-04-16 10:51:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][860/1251] eta 0:01:54 lr 0.000259 time 0.2801 (0.2932) loss 4.0848 (3.3579) grad_norm 1.7952 (1.9929) [2021-04-16 10:51:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][870/1251] eta 0:01:51 lr 0.000259 time 0.2442 (0.2931) loss 3.5989 (3.3584) grad_norm 2.5508 (1.9922) [2021-04-16 10:51:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][880/1251] eta 0:01:48 lr 0.000259 time 0.2805 (0.2929) loss 3.5379 (3.3602) grad_norm 2.0820 (1.9921) [2021-04-16 10:51:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][890/1251] eta 0:01:45 lr 0.000259 time 0.3188 (0.2928) loss 3.5284 (3.3620) grad_norm 2.0225 (1.9918) [2021-04-16 10:51:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][900/1251] eta 0:01:42 lr 0.000259 time 0.2526 (0.2927) loss 3.7727 (3.3618) grad_norm 1.8317 (1.9917) [2021-04-16 10:51:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][910/1251] eta 0:01:39 lr 0.000259 time 0.3061 (0.2926) loss 2.8810 (3.3627) grad_norm 1.7598 (1.9915) [2021-04-16 10:52:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][920/1251] eta 0:01:36 lr 0.000259 time 0.2790 (0.2924) loss 2.8921 (3.3590) grad_norm 1.8943 (1.9906) [2021-04-16 10:52:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][930/1251] eta 0:01:33 lr 0.000259 time 0.2958 (0.2924) loss 2.4139 (3.3555) grad_norm 1.9469 (1.9895) [2021-04-16 10:52:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][940/1251] eta 0:01:30 lr 0.000259 time 0.3051 (0.2924) loss 4.2357 (3.3555) grad_norm 2.1478 (1.9891) [2021-04-16 10:52:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][950/1251] eta 0:01:27 lr 0.000259 time 0.2685 (0.2923) loss 3.0542 (3.3570) grad_norm 2.0189 (1.9897) [2021-04-16 10:52:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][960/1251] eta 0:01:25 lr 0.000259 time 0.2895 (0.2922) loss 4.0420 (3.3554) grad_norm 2.0249 (1.9901) [2021-04-16 10:52:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][970/1251] eta 0:01:22 lr 0.000259 time 0.2744 (0.2921) loss 3.5888 (3.3541) grad_norm 1.8813 (1.9903) [2021-04-16 10:52:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][980/1251] eta 0:01:19 lr 0.000258 time 0.3077 (0.2920) loss 3.0441 (3.3520) grad_norm 1.7360 (1.9894) [2021-04-16 10:52:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][990/1251] eta 0:01:16 lr 0.000258 time 0.3084 (0.2919) loss 3.5054 (3.3543) grad_norm 2.1150 (1.9889) [2021-04-16 10:52:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][1000/1251] eta 0:01:13 lr 0.000258 time 0.2687 (0.2917) loss 2.0761 (3.3507) grad_norm 1.7859 (1.9893) [2021-04-16 10:52:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][1010/1251] eta 0:01:10 lr 0.000258 time 0.2704 (0.2916) loss 3.7234 (3.3497) grad_norm 1.8333 (1.9885) [2021-04-16 10:52:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][1020/1251] eta 0:01:07 lr 0.000258 time 0.2809 (0.2916) loss 3.6485 (3.3492) grad_norm 1.8584 (1.9872) [2021-04-16 10:52:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][1030/1251] eta 0:01:04 lr 0.000258 time 0.2877 (0.2915) loss 3.4440 (3.3498) grad_norm 1.7859 (1.9862) [2021-04-16 10:52:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][1040/1251] eta 0:01:01 lr 0.000258 time 0.2679 (0.2913) loss 2.6625 (3.3504) grad_norm 1.8128 (1.9855) [2021-04-16 10:52:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][1050/1251] eta 0:00:58 lr 0.000258 time 0.2599 (0.2912) loss 3.6231 (3.3491) grad_norm 1.9076 (1.9846) [2021-04-16 10:52:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][1060/1251] eta 0:00:55 lr 0.000258 time 0.2753 (0.2911) loss 3.1704 (3.3497) grad_norm 1.7147 (1.9832) [2021-04-16 10:52:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][1070/1251] eta 0:00:52 lr 0.000258 time 0.2928 (0.2911) loss 4.0073 (3.3488) grad_norm 1.8342 (1.9826) [2021-04-16 10:52:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][1080/1251] eta 0:00:49 lr 0.000258 time 0.2908 (0.2910) loss 3.7724 (3.3495) grad_norm 1.8041 (1.9827) [2021-04-16 10:52:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][1090/1251] eta 0:00:46 lr 0.000258 time 0.2755 (0.2911) loss 2.2763 (3.3492) grad_norm 1.9898 (1.9828) [2021-04-16 10:52:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][1100/1251] eta 0:00:43 lr 0.000258 time 0.2749 (0.2910) loss 2.8250 (3.3480) grad_norm 2.3632 (1.9833) [2021-04-16 10:52:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][1110/1251] eta 0:00:41 lr 0.000258 time 0.2800 (0.2910) loss 2.7345 (3.3460) grad_norm 1.9774 (1.9830) [2021-04-16 10:52:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][1120/1251] eta 0:00:38 lr 0.000258 time 0.2795 (0.2909) loss 3.3872 (3.3451) grad_norm 2.1883 (1.9819) [2021-04-16 10:53:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][1130/1251] eta 0:00:35 lr 0.000258 time 0.2955 (0.2908) loss 4.0751 (3.3447) grad_norm 2.5197 (1.9818) [2021-04-16 10:53:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][1140/1251] eta 0:00:32 lr 0.000258 time 0.2860 (0.2907) loss 3.3264 (3.3455) grad_norm 2.3531 (1.9828) [2021-04-16 10:53:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][1150/1251] eta 0:00:29 lr 0.000258 time 0.2605 (0.2909) loss 3.7479 (3.3441) grad_norm 1.9821 (1.9828) [2021-04-16 10:53:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][1160/1251] eta 0:00:26 lr 0.000258 time 0.2602 (0.2909) loss 3.6885 (3.3469) grad_norm 1.9215 (1.9831) [2021-04-16 10:53:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][1170/1251] eta 0:00:23 lr 0.000258 time 0.2462 (0.2908) loss 4.1137 (3.3477) grad_norm 1.8847 (1.9840) [2021-04-16 10:53:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][1180/1251] eta 0:00:20 lr 0.000258 time 0.2787 (0.2907) loss 2.5522 (3.3493) grad_norm 2.0508 (1.9843) [2021-04-16 10:53:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][1190/1251] eta 0:00:17 lr 0.000258 time 0.2797 (0.2906) loss 3.5861 (3.3513) grad_norm 1.8831 (1.9837) [2021-04-16 10:53:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][1200/1251] eta 0:00:14 lr 0.000258 time 0.2838 (0.2905) loss 3.1719 (3.3518) grad_norm 1.7482 (1.9826) [2021-04-16 10:53:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][1210/1251] eta 0:00:11 lr 0.000258 time 0.2987 (0.2904) loss 4.2294 (3.3522) grad_norm 1.7924 (1.9826) [2021-04-16 10:53:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][1220/1251] eta 0:00:09 lr 0.000258 time 0.2812 (0.2903) loss 3.4970 (3.3524) grad_norm 1.7118 (1.9821) [2021-04-16 10:53:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][1230/1251] eta 0:00:06 lr 0.000258 time 0.3032 (0.2903) loss 2.2489 (3.3515) grad_norm 1.7107 (1.9816) [2021-04-16 10:53:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][1240/1251] eta 0:00:03 lr 0.000258 time 0.2540 (0.2901) loss 2.9693 (3.3517) grad_norm 1.6439 (1.9803) [2021-04-16 10:53:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [199/300][1250/1251] eta 0:00:00 lr 0.000258 time 0.2481 (0.2898) loss 3.2469 (3.3542) grad_norm 1.7412 (1.9798) [2021-04-16 10:53:36 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 199 training takes 0:06:05 [2021-04-16 10:53:36 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_199.pth saving...... [2021-04-16 10:53:43 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_199.pth saved !!! [2021-04-16 10:53:44 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.131 (1.131) Loss 0.9151 (0.9151) Acc@1 78.125 (78.125) Acc@5 94.629 (94.629) [2021-04-16 10:53:45 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.275 (0.218) Loss 0.9097 (0.9222) Acc@1 79.102 (78.516) Acc@5 94.629 (94.558) [2021-04-16 10:53:48 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.329 (0.234) Loss 0.9725 (0.9260) Acc@1 78.027 (78.297) Acc@5 93.164 (94.424) [2021-04-16 10:53:50 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.114 (0.230) Loss 0.9134 (0.9230) Acc@1 78.906 (78.418) Acc@5 94.043 (94.349) [2021-04-16 10:53:52 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.079 (0.229) Loss 0.8508 (0.9276) Acc@1 80.176 (78.258) Acc@5 95.020 (94.331) [2021-04-16 10:54:01 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 78.278 Acc@5 94.346 [2021-04-16 10:54:01 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 78.3% [2021-04-16 10:54:01 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 78.28% [2021-04-16 10:54:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][0/1251] eta 0:31:51 lr 0.000258 time 1.5283 (1.5283) loss 2.9317 (2.9317) grad_norm 1.8123 (1.8123) [2021-04-16 10:54:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][10/1251] eta 0:07:57 lr 0.000257 time 0.2658 (0.3850) loss 3.2392 (3.1860) grad_norm 2.1502 (1.8875) [2021-04-16 10:54:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][20/1251] eta 0:06:52 lr 0.000257 time 0.2816 (0.3350) loss 4.0819 (3.1841) grad_norm 2.3785 (1.9903) [2021-04-16 10:54:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][30/1251] eta 0:06:24 lr 0.000257 time 0.2813 (0.3151) loss 3.4776 (3.2016) grad_norm 1.9958 (1.9937) [2021-04-16 10:54:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][40/1251] eta 0:06:14 lr 0.000257 time 0.2755 (0.3094) loss 2.1217 (3.2210) grad_norm 1.7636 (2.0068) [2021-04-16 10:54:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][50/1251] eta 0:06:04 lr 0.000257 time 0.2888 (0.3032) loss 2.3057 (3.1981) grad_norm 1.9368 (1.9731) [2021-04-16 10:54:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][60/1251] eta 0:05:57 lr 0.000257 time 0.2769 (0.2998) loss 3.7833 (3.2020) grad_norm 2.1373 (1.9696) [2021-04-16 10:54:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][70/1251] eta 0:05:50 lr 0.000257 time 0.2694 (0.2964) loss 2.4345 (3.2222) grad_norm 1.9848 (1.9918) [2021-04-16 10:54:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][80/1251] eta 0:05:45 lr 0.000257 time 0.2959 (0.2948) loss 3.7394 (3.2621) grad_norm 2.2556 (2.0020) [2021-04-16 10:54:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][90/1251] eta 0:05:40 lr 0.000257 time 0.2758 (0.2930) loss 4.1921 (3.2519) grad_norm 2.1438 (1.9979) [2021-04-16 10:54:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][100/1251] eta 0:05:35 lr 0.000257 time 0.2775 (0.2918) loss 2.6709 (3.2488) grad_norm 1.9327 (1.9986) [2021-04-16 10:54:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][110/1251] eta 0:05:31 lr 0.000257 time 0.2867 (0.2906) loss 3.7281 (3.2652) grad_norm 1.8547 (1.9966) [2021-04-16 10:54:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][120/1251] eta 0:05:27 lr 0.000257 time 0.2909 (0.2893) loss 2.7457 (3.2653) grad_norm 1.8438 (1.9941) [2021-04-16 10:54:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][130/1251] eta 0:05:22 lr 0.000257 time 0.2877 (0.2880) loss 3.1897 (3.2859) grad_norm 1.8258 (1.9932) [2021-04-16 10:54:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][140/1251] eta 0:05:20 lr 0.000257 time 0.2565 (0.2889) loss 2.7684 (3.2759) grad_norm 1.7465 (1.9852) [2021-04-16 10:54:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][150/1251] eta 0:05:17 lr 0.000257 time 0.2754 (0.2888) loss 3.7108 (3.3003) grad_norm 2.0571 (1.9822) [2021-04-16 10:54:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][160/1251] eta 0:05:14 lr 0.000257 time 0.2858 (0.2887) loss 3.4764 (3.2918) grad_norm 1.8677 (1.9780) [2021-04-16 10:54:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][170/1251] eta 0:05:11 lr 0.000257 time 0.2916 (0.2878) loss 3.3527 (3.2968) grad_norm 2.0867 (1.9853) [2021-04-16 10:54:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][180/1251] eta 0:05:07 lr 0.000257 time 0.2899 (0.2873) loss 2.4060 (3.3058) grad_norm 1.9125 (1.9855) [2021-04-16 10:54:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][190/1251] eta 0:05:04 lr 0.000257 time 0.3004 (0.2868) loss 2.5626 (3.3014) grad_norm 1.8239 (1.9866) [2021-04-16 10:54:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][200/1251] eta 0:05:00 lr 0.000257 time 0.2867 (0.2861) loss 3.3675 (3.2824) grad_norm 1.9776 (1.9886) [2021-04-16 10:55:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][210/1251] eta 0:04:57 lr 0.000257 time 0.2901 (0.2857) loss 3.6218 (3.2913) grad_norm 1.8430 (1.9907) [2021-04-16 10:55:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][220/1251] eta 0:04:54 lr 0.000257 time 0.2861 (0.2853) loss 2.8621 (3.2878) grad_norm 1.8749 (1.9962) [2021-04-16 10:55:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][230/1251] eta 0:04:51 lr 0.000257 time 0.2912 (0.2854) loss 3.3204 (3.2899) grad_norm 1.7694 (1.9971) [2021-04-16 10:55:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][240/1251] eta 0:04:48 lr 0.000257 time 0.2757 (0.2850) loss 2.9112 (3.2913) grad_norm 1.8296 (1.9965) [2021-04-16 10:55:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][250/1251] eta 0:04:44 lr 0.000257 time 0.2861 (0.2846) loss 4.2988 (3.3030) grad_norm 2.0803 (1.9989) [2021-04-16 10:55:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][260/1251] eta 0:04:41 lr 0.000257 time 0.2674 (0.2842) loss 2.6706 (3.3001) grad_norm 1.8879 (1.9999) [2021-04-16 10:55:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][270/1251] eta 0:04:38 lr 0.000257 time 0.2541 (0.2838) loss 3.2372 (3.2953) grad_norm 1.8247 (1.9990) [2021-04-16 10:55:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][280/1251] eta 0:04:35 lr 0.000256 time 0.2723 (0.2836) loss 2.8927 (3.2906) grad_norm 2.0338 (1.9968) [2021-04-16 10:55:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][290/1251] eta 0:04:32 lr 0.000256 time 0.3032 (0.2835) loss 3.7716 (3.2942) grad_norm 2.2487 (1.9997) [2021-04-16 10:55:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][300/1251] eta 0:04:29 lr 0.000256 time 0.2616 (0.2837) loss 3.0181 (3.2951) grad_norm 1.9070 (2.0008) [2021-04-16 10:55:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][310/1251] eta 0:04:26 lr 0.000256 time 0.2522 (0.2836) loss 2.7884 (3.2952) grad_norm 2.6433 (2.0053) [2021-04-16 10:55:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][320/1251] eta 0:04:23 lr 0.000256 time 0.2558 (0.2834) loss 4.0081 (3.2955) grad_norm 1.9303 (2.0067) [2021-04-16 10:55:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][330/1251] eta 0:04:20 lr 0.000256 time 0.2907 (0.2833) loss 4.0317 (3.2942) grad_norm 2.1166 (2.0048) [2021-04-16 10:55:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][340/1251] eta 0:04:17 lr 0.000256 time 0.2946 (0.2830) loss 3.6296 (3.2993) grad_norm 1.8388 (2.0033) [2021-04-16 10:55:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][350/1251] eta 0:04:14 lr 0.000256 time 0.2717 (0.2828) loss 3.2148 (3.2942) grad_norm 2.2608 (2.0053) [2021-04-16 10:55:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][360/1251] eta 0:04:12 lr 0.000256 time 0.2878 (0.2829) loss 3.1824 (3.2917) grad_norm 1.8312 (2.0055) [2021-04-16 10:55:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][370/1251] eta 0:04:09 lr 0.000256 time 0.2754 (0.2829) loss 3.2043 (3.2979) grad_norm 2.1427 (2.0048) [2021-04-16 10:55:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][380/1251] eta 0:04:06 lr 0.000256 time 0.2950 (0.2828) loss 3.6576 (3.3024) grad_norm 1.7200 (2.0037) [2021-04-16 10:55:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][390/1251] eta 0:04:03 lr 0.000256 time 0.2719 (0.2827) loss 3.7777 (3.2997) grad_norm 2.1764 (1.9991) [2021-04-16 10:55:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][400/1251] eta 0:04:00 lr 0.000256 time 0.2634 (0.2825) loss 3.8392 (3.3025) grad_norm 1.8979 (1.9964) [2021-04-16 10:55:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][410/1251] eta 0:03:57 lr 0.000256 time 0.2837 (0.2823) loss 2.6995 (3.3009) grad_norm 2.2267 (1.9950) [2021-04-16 10:55:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][420/1251] eta 0:03:54 lr 0.000256 time 0.2712 (0.2823) loss 2.7049 (3.2969) grad_norm 2.2593 (1.9970) [2021-04-16 10:56:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][430/1251] eta 0:03:51 lr 0.000256 time 0.2713 (0.2821) loss 3.9258 (3.2960) grad_norm 1.9818 (1.9973) [2021-04-16 10:56:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][440/1251] eta 0:03:48 lr 0.000256 time 0.2857 (0.2820) loss 4.0620 (3.2972) grad_norm 2.4383 (2.0011) [2021-04-16 10:56:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][450/1251] eta 0:03:45 lr 0.000256 time 0.2674 (0.2818) loss 3.2366 (3.3011) grad_norm 1.8114 (1.9976) [2021-04-16 10:56:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][460/1251] eta 0:03:42 lr 0.000256 time 0.3120 (0.2818) loss 3.5686 (3.3000) grad_norm 1.8801 (2.0149) [2021-04-16 10:56:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][470/1251] eta 0:03:39 lr 0.000256 time 0.2823 (0.2817) loss 3.3762 (3.3008) grad_norm 1.7243 (2.0147) [2021-04-16 10:56:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][480/1251] eta 0:03:37 lr 0.000256 time 0.2906 (0.2816) loss 3.6519 (3.3021) grad_norm 1.8398 (2.0135) [2021-04-16 10:56:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][490/1251] eta 0:03:34 lr 0.000256 time 0.2699 (0.2818) loss 3.5375 (3.3020) grad_norm 1.8568 (2.0124) [2021-04-16 10:56:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][500/1251] eta 0:03:31 lr 0.000256 time 0.2664 (0.2816) loss 3.8069 (3.3037) grad_norm 1.6482 (2.0088) [2021-04-16 10:56:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][510/1251] eta 0:03:28 lr 0.000256 time 0.2860 (0.2815) loss 3.5376 (3.3001) grad_norm 2.1772 (2.0074) [2021-04-16 10:56:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][520/1251] eta 0:03:25 lr 0.000256 time 0.3020 (0.2816) loss 3.5314 (3.2969) grad_norm 1.7922 (2.0054) [2021-04-16 10:56:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][530/1251] eta 0:03:23 lr 0.000256 time 0.2834 (0.2816) loss 3.7269 (3.2989) grad_norm 2.0110 (2.0035) [2021-04-16 10:56:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][540/1251] eta 0:03:20 lr 0.000256 time 0.2633 (0.2817) loss 3.6753 (3.3011) grad_norm 2.0538 (2.0041) [2021-04-16 10:56:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][550/1251] eta 0:03:17 lr 0.000256 time 0.2882 (0.2817) loss 2.6747 (3.3000) grad_norm 2.0450 (2.0034) [2021-04-16 10:56:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][560/1251] eta 0:03:14 lr 0.000255 time 0.2751 (0.2815) loss 3.5438 (3.2997) grad_norm 1.8288 (2.0019) [2021-04-16 10:56:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][570/1251] eta 0:03:11 lr 0.000255 time 0.4104 (0.2819) loss 2.7651 (3.2994) grad_norm 2.1654 (2.0012) [2021-04-16 10:56:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][580/1251] eta 0:03:09 lr 0.000255 time 0.2906 (0.2818) loss 3.0343 (3.2992) grad_norm 1.8527 (2.0002) [2021-04-16 10:56:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][590/1251] eta 0:03:06 lr 0.000255 time 0.2936 (0.2817) loss 2.1017 (3.3021) grad_norm 2.2468 (1.9990) [2021-04-16 10:56:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][600/1251] eta 0:03:03 lr 0.000255 time 0.2745 (0.2817) loss 3.3437 (3.3061) grad_norm 1.8539 (1.9980) [2021-04-16 10:56:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][610/1251] eta 0:03:00 lr 0.000255 time 0.2738 (0.2816) loss 3.3838 (3.3034) grad_norm 2.0183 (1.9983) [2021-04-16 10:56:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][620/1251] eta 0:02:57 lr 0.000255 time 0.2783 (0.2818) loss 3.1319 (3.3032) grad_norm 2.2387 (2.0012) [2021-04-16 10:56:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][630/1251] eta 0:02:54 lr 0.000255 time 0.2852 (0.2817) loss 3.3238 (3.3023) grad_norm 1.8692 (2.0002) [2021-04-16 10:57:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][640/1251] eta 0:02:52 lr 0.000255 time 0.2625 (0.2816) loss 3.8054 (3.3043) grad_norm 2.0086 (2.0008) [2021-04-16 10:57:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][650/1251] eta 0:02:49 lr 0.000255 time 0.2769 (0.2815) loss 3.8360 (3.3074) grad_norm 1.7853 (2.0000) [2021-04-16 10:57:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][660/1251] eta 0:02:46 lr 0.000255 time 0.2919 (0.2814) loss 3.2760 (3.3091) grad_norm 1.8274 (1.9989) [2021-04-16 10:57:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][670/1251] eta 0:02:43 lr 0.000255 time 0.2601 (0.2813) loss 3.5069 (3.3103) grad_norm 1.8947 (1.9980) [2021-04-16 10:57:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][680/1251] eta 0:02:40 lr 0.000255 time 0.2752 (0.2813) loss 3.2122 (3.3113) grad_norm 1.9453 (1.9964) [2021-04-16 10:57:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][690/1251] eta 0:02:37 lr 0.000255 time 0.2807 (0.2812) loss 2.7083 (3.3165) grad_norm 1.8807 (1.9973) [2021-04-16 10:57:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][700/1251] eta 0:02:34 lr 0.000255 time 0.2768 (0.2811) loss 3.4570 (3.3183) grad_norm 1.8470 (1.9962) [2021-04-16 10:57:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][710/1251] eta 0:02:32 lr 0.000255 time 0.2978 (0.2812) loss 3.8293 (3.3219) grad_norm 1.9574 (1.9967) [2021-04-16 10:57:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][720/1251] eta 0:02:29 lr 0.000255 time 0.2696 (0.2812) loss 2.7175 (3.3211) grad_norm 1.7392 (1.9942) [2021-04-16 10:57:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][730/1251] eta 0:02:26 lr 0.000255 time 0.2911 (0.2813) loss 4.1649 (3.3233) grad_norm 1.9606 (1.9931) [2021-04-16 10:57:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][740/1251] eta 0:02:23 lr 0.000255 time 0.2960 (0.2813) loss 3.1223 (3.3231) grad_norm 1.7952 (1.9929) [2021-04-16 10:57:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][750/1251] eta 0:02:20 lr 0.000255 time 0.2888 (0.2813) loss 4.0124 (3.3232) grad_norm 2.0389 (1.9945) [2021-04-16 10:57:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][760/1251] eta 0:02:18 lr 0.000255 time 0.2883 (0.2814) loss 3.4947 (3.3255) grad_norm 2.1400 (1.9965) [2021-04-16 10:57:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][770/1251] eta 0:02:15 lr 0.000255 time 0.3340 (0.2815) loss 3.2363 (3.3286) grad_norm 2.1647 (1.9991) [2021-04-16 10:57:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][780/1251] eta 0:02:12 lr 0.000255 time 0.2636 (0.2814) loss 3.0089 (3.3289) grad_norm 3.4819 (2.0018) [2021-04-16 10:57:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][790/1251] eta 0:02:09 lr 0.000255 time 0.2991 (0.2814) loss 3.7031 (3.3295) grad_norm 2.1918 (2.0025) [2021-04-16 10:57:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][800/1251] eta 0:02:06 lr 0.000255 time 0.4231 (0.2815) loss 3.2392 (3.3325) grad_norm 1.9723 (2.0023) [2021-04-16 10:57:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][810/1251] eta 0:02:04 lr 0.000255 time 0.2766 (0.2814) loss 3.7792 (3.3312) grad_norm 1.8159 (2.0020) [2021-04-16 10:57:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][820/1251] eta 0:02:01 lr 0.000255 time 0.2725 (0.2813) loss 3.9780 (3.3338) grad_norm 1.9646 (2.0014) [2021-04-16 10:57:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][830/1251] eta 0:01:58 lr 0.000255 time 0.2826 (0.2813) loss 2.4627 (3.3329) grad_norm 2.3223 (2.0039) [2021-04-16 10:57:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][840/1251] eta 0:01:55 lr 0.000254 time 0.2880 (0.2813) loss 4.1042 (3.3348) grad_norm 2.4400 (2.0050) [2021-04-16 10:58:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][850/1251] eta 0:01:52 lr 0.000254 time 0.2553 (0.2812) loss 3.9392 (3.3364) grad_norm 1.9307 (2.0052) [2021-04-16 10:58:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][860/1251] eta 0:01:49 lr 0.000254 time 0.2759 (0.2811) loss 3.5329 (3.3344) grad_norm 2.3522 (2.0060) [2021-04-16 10:58:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][870/1251] eta 0:01:47 lr 0.000254 time 0.2701 (0.2810) loss 2.8854 (3.3363) grad_norm 1.6950 (2.0045) [2021-04-16 10:58:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][880/1251] eta 0:01:44 lr 0.000254 time 0.2753 (0.2812) loss 3.5129 (3.3321) grad_norm 1.7696 (2.0046) [2021-04-16 10:58:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][890/1251] eta 0:01:41 lr 0.000254 time 0.2698 (0.2811) loss 3.5713 (3.3293) grad_norm 1.9402 (2.0049) [2021-04-16 10:58:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][900/1251] eta 0:01:38 lr 0.000254 time 0.2606 (0.2810) loss 3.2281 (3.3297) grad_norm 1.8898 (2.0047) [2021-04-16 10:58:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][910/1251] eta 0:01:35 lr 0.000254 time 0.2625 (0.2810) loss 3.5663 (3.3318) grad_norm 2.1702 (2.0039) [2021-04-16 10:58:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][920/1251] eta 0:01:33 lr 0.000254 time 0.2901 (0.2810) loss 3.0808 (3.3274) grad_norm 2.2687 (2.0040) [2021-04-16 10:58:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][930/1251] eta 0:01:30 lr 0.000254 time 0.2708 (0.2810) loss 3.4236 (3.3307) grad_norm 1.7429 (2.0029) [2021-04-16 10:58:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][940/1251] eta 0:01:27 lr 0.000254 time 0.3071 (0.2813) loss 3.5874 (3.3297) grad_norm 1.8470 (2.0027) [2021-04-16 10:58:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][950/1251] eta 0:01:24 lr 0.000254 time 0.2579 (0.2812) loss 3.0615 (3.3286) grad_norm 2.2948 (2.0035) [2021-04-16 10:58:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][960/1251] eta 0:01:21 lr 0.000254 time 0.2855 (0.2812) loss 3.7998 (3.3276) grad_norm 2.0658 (2.0031) [2021-04-16 10:58:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][970/1251] eta 0:01:18 lr 0.000254 time 0.2570 (0.2811) loss 4.0948 (3.3269) grad_norm 2.2713 (2.0034) [2021-04-16 10:58:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][980/1251] eta 0:01:16 lr 0.000254 time 0.2701 (0.2811) loss 3.6966 (3.3300) grad_norm 2.1170 (2.0046) [2021-04-16 10:58:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][990/1251] eta 0:01:13 lr 0.000254 time 0.3022 (0.2810) loss 3.6427 (3.3332) grad_norm 1.9725 (2.0049) [2021-04-16 10:58:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][1000/1251] eta 0:01:10 lr 0.000254 time 0.2729 (0.2810) loss 3.3122 (3.3325) grad_norm 2.1100 (nan) [2021-04-16 10:58:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][1010/1251] eta 0:01:07 lr 0.000254 time 0.2494 (0.2810) loss 3.8256 (3.3346) grad_norm 1.9091 (nan) [2021-04-16 10:58:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][1020/1251] eta 0:01:04 lr 0.000254 time 0.2781 (0.2809) loss 3.2449 (3.3338) grad_norm 1.7558 (nan) [2021-04-16 10:58:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][1030/1251] eta 0:01:02 lr 0.000254 time 0.2638 (0.2808) loss 3.7581 (3.3299) grad_norm 2.1602 (nan) [2021-04-16 10:58:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][1040/1251] eta 0:00:59 lr 0.000254 time 0.2696 (0.2808) loss 3.1771 (3.3304) grad_norm 2.0909 (nan) [2021-04-16 10:58:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][1050/1251] eta 0:00:56 lr 0.000254 time 0.2647 (0.2807) loss 3.3411 (3.3322) grad_norm 1.9237 (nan) [2021-04-16 10:58:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][1060/1251] eta 0:00:53 lr 0.000254 time 0.2640 (0.2807) loss 3.7168 (3.3324) grad_norm 1.6833 (nan) [2021-04-16 10:59:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][1070/1251] eta 0:00:50 lr 0.000254 time 0.3074 (0.2808) loss 3.6033 (3.3322) grad_norm 2.1371 (nan) [2021-04-16 10:59:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][1080/1251] eta 0:00:48 lr 0.000254 time 0.2639 (0.2807) loss 2.7332 (3.3331) grad_norm 1.9836 (nan) [2021-04-16 10:59:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][1090/1251] eta 0:00:45 lr 0.000254 time 0.2599 (0.2809) loss 3.6439 (3.3309) grad_norm 2.1351 (nan) [2021-04-16 10:59:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][1100/1251] eta 0:00:42 lr 0.000254 time 0.2885 (0.2809) loss 2.1251 (3.3304) grad_norm 2.3984 (nan) [2021-04-16 10:59:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][1110/1251] eta 0:00:39 lr 0.000254 time 0.2652 (0.2809) loss 4.0388 (3.3300) grad_norm 2.3709 (nan) [2021-04-16 10:59:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][1120/1251] eta 0:00:36 lr 0.000253 time 0.2639 (0.2810) loss 4.0784 (3.3328) grad_norm 2.0374 (nan) [2021-04-16 10:59:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][1130/1251] eta 0:00:33 lr 0.000253 time 0.2840 (0.2810) loss 4.1228 (3.3330) grad_norm 2.1834 (nan) [2021-04-16 10:59:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][1140/1251] eta 0:00:31 lr 0.000253 time 0.2558 (0.2809) loss 3.7743 (3.3311) grad_norm 1.9752 (nan) [2021-04-16 10:59:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][1150/1251] eta 0:00:28 lr 0.000253 time 0.2780 (0.2810) loss 2.9043 (3.3292) grad_norm 1.8549 (nan) [2021-04-16 10:59:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][1160/1251] eta 0:00:25 lr 0.000253 time 0.2996 (0.2810) loss 3.9722 (3.3308) grad_norm 2.0168 (nan) [2021-04-16 10:59:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][1170/1251] eta 0:00:22 lr 0.000253 time 0.2741 (0.2809) loss 3.9464 (3.3322) grad_norm 1.8281 (nan) [2021-04-16 10:59:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][1180/1251] eta 0:00:19 lr 0.000253 time 0.2841 (0.2809) loss 3.6888 (3.3323) grad_norm 1.7581 (nan) [2021-04-16 10:59:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][1190/1251] eta 0:00:17 lr 0.000253 time 0.2864 (0.2809) loss 2.8026 (3.3319) grad_norm 1.9492 (nan) [2021-04-16 10:59:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][1200/1251] eta 0:00:14 lr 0.000253 time 0.2605 (0.2809) loss 3.8918 (3.3318) grad_norm 1.9387 (nan) [2021-04-16 10:59:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][1210/1251] eta 0:00:11 lr 0.000253 time 0.3049 (0.2808) loss 4.0441 (3.3355) grad_norm 2.3406 (nan) [2021-04-16 10:59:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][1220/1251] eta 0:00:08 lr 0.000253 time 0.2577 (0.2808) loss 2.5719 (3.3363) grad_norm 1.9476 (nan) [2021-04-16 10:59:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][1230/1251] eta 0:00:05 lr 0.000253 time 0.2768 (0.2809) loss 3.4659 (3.3376) grad_norm 1.7880 (nan) [2021-04-16 10:59:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][1240/1251] eta 0:00:03 lr 0.000253 time 0.2486 (0.2807) loss 3.1875 (3.3371) grad_norm 1.7453 (nan) [2021-04-16 10:59:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [200/300][1250/1251] eta 0:00:00 lr 0.000253 time 0.2506 (0.2805) loss 3.5572 (3.3358) grad_norm 2.0472 (nan) [2021-04-16 10:59:56 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 200 training takes 0:05:55 [2021-04-16 10:59:56 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_200.pth saving...... [2021-04-16 11:00:13 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_200.pth saved !!! [2021-04-16 11:00:15 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.153 (1.153) Loss 0.8957 (0.8957) Acc@1 78.809 (78.809) Acc@5 94.336 (94.336) [2021-04-16 11:00:16 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.236 (0.217) Loss 0.9463 (0.9042) Acc@1 78.516 (78.924) Acc@5 93.457 (94.221) [2021-04-16 11:00:18 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.112 (0.214) Loss 0.8258 (0.9098) Acc@1 80.762 (78.632) Acc@5 94.629 (94.266) [2021-04-16 11:00:21 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.107 (0.242) Loss 0.9925 (0.9177) Acc@1 77.441 (78.396) Acc@5 92.480 (94.188) [2021-04-16 11:00:23 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.225) Loss 0.9743 (0.9208) Acc@1 77.344 (78.313) Acc@5 93.945 (94.157) [2021-04-16 11:00:27 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 78.326 Acc@5 94.226 [2021-04-16 11:00:27 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 78.3% [2021-04-16 11:00:27 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 78.33% [2021-04-16 11:00:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][0/1251] eta 5:14:05 lr 0.000253 time 15.0641 (15.0641) loss 3.0758 (3.0758) grad_norm 2.5983 (2.5983) [2021-04-16 11:00:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][10/1251] eta 0:33:49 lr 0.000253 time 0.4495 (1.6352) loss 3.8264 (3.4098) grad_norm 2.0124 (2.0642) [2021-04-16 11:00:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][20/1251] eta 0:20:17 lr 0.000253 time 0.2644 (0.9894) loss 4.0317 (3.3592) grad_norm 1.9540 (2.0259) [2021-04-16 11:00:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][30/1251] eta 0:15:27 lr 0.000253 time 0.2889 (0.7597) loss 3.5926 (3.2659) grad_norm 1.8266 (2.0155) [2021-04-16 11:00:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][40/1251] eta 0:12:58 lr 0.000253 time 0.2929 (0.6425) loss 3.1187 (3.2469) grad_norm 1.7949 (2.0031) [2021-04-16 11:00:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][50/1251] eta 0:11:25 lr 0.000253 time 0.2673 (0.5711) loss 2.2633 (3.2104) grad_norm 1.8857 (1.9945) [2021-04-16 11:00:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][60/1251] eta 0:10:23 lr 0.000253 time 0.3052 (0.5232) loss 3.7891 (3.2830) grad_norm 1.7575 (1.9696) [2021-04-16 11:01:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][70/1251] eta 0:09:36 lr 0.000253 time 0.2921 (0.4882) loss 3.9523 (3.3377) grad_norm 1.7250 (1.9729) [2021-04-16 11:01:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][80/1251] eta 0:09:01 lr 0.000253 time 0.3221 (0.4625) loss 2.4917 (3.3284) grad_norm 1.8265 (1.9587) [2021-04-16 11:01:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][90/1251] eta 0:08:35 lr 0.000253 time 0.2650 (0.4438) loss 3.9785 (3.3265) grad_norm 1.6976 (1.9504) [2021-04-16 11:01:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][100/1251] eta 0:08:11 lr 0.000253 time 0.2795 (0.4272) loss 2.2261 (3.3224) grad_norm 1.8827 (1.9556) [2021-04-16 11:01:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][110/1251] eta 0:07:52 lr 0.000253 time 0.2797 (0.4141) loss 3.0341 (3.3353) grad_norm 1.9241 (1.9612) [2021-04-16 11:01:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][120/1251] eta 0:07:36 lr 0.000253 time 0.2728 (0.4033) loss 3.6346 (3.3324) grad_norm 2.2480 (1.9674) [2021-04-16 11:01:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][130/1251] eta 0:07:21 lr 0.000253 time 0.2696 (0.3936) loss 3.6615 (3.3326) grad_norm 2.2192 (1.9681) [2021-04-16 11:01:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][140/1251] eta 0:07:10 lr 0.000253 time 0.2643 (0.3874) loss 2.3413 (3.3266) grad_norm 1.9835 (1.9719) [2021-04-16 11:01:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][150/1251] eta 0:06:59 lr 0.000252 time 0.2723 (0.3812) loss 3.9260 (3.3352) grad_norm 2.2656 (1.9848) [2021-04-16 11:01:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][160/1251] eta 0:06:49 lr 0.000252 time 0.2858 (0.3750) loss 3.0866 (3.3363) grad_norm 1.8172 (1.9824) [2021-04-16 11:01:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][170/1251] eta 0:06:39 lr 0.000252 time 0.3160 (0.3699) loss 3.5188 (3.3404) grad_norm 2.1012 (1.9826) [2021-04-16 11:01:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][180/1251] eta 0:06:30 lr 0.000252 time 0.2803 (0.3646) loss 3.1024 (3.3408) grad_norm 1.7176 (1.9810) [2021-04-16 11:01:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][190/1251] eta 0:06:21 lr 0.000252 time 0.2902 (0.3599) loss 3.7888 (3.3258) grad_norm 2.0829 (1.9786) [2021-04-16 11:01:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][200/1251] eta 0:06:13 lr 0.000252 time 0.2833 (0.3557) loss 3.1231 (3.3187) grad_norm 2.1004 (1.9822) [2021-04-16 11:01:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][210/1251] eta 0:06:06 lr 0.000252 time 0.2774 (0.3521) loss 3.5016 (3.3235) grad_norm 1.9745 (1.9822) [2021-04-16 11:01:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][220/1251] eta 0:05:59 lr 0.000252 time 0.2596 (0.3486) loss 2.3291 (3.3155) grad_norm 1.9414 (1.9820) [2021-04-16 11:01:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][230/1251] eta 0:05:52 lr 0.000252 time 0.2977 (0.3454) loss 3.4785 (3.3285) grad_norm 1.8518 (1.9838) [2021-04-16 11:01:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][240/1251] eta 0:05:46 lr 0.000252 time 0.2719 (0.3423) loss 3.6855 (3.3197) grad_norm 1.8869 (1.9814) [2021-04-16 11:01:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][250/1251] eta 0:05:40 lr 0.000252 time 0.2823 (0.3399) loss 3.5088 (3.3356) grad_norm 2.0448 (1.9874) [2021-04-16 11:01:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][260/1251] eta 0:05:34 lr 0.000252 time 0.2627 (0.3379) loss 3.2888 (3.3317) grad_norm 1.9682 (1.9879) [2021-04-16 11:01:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][270/1251] eta 0:05:29 lr 0.000252 time 0.2760 (0.3357) loss 2.8661 (3.3162) grad_norm 1.8499 (1.9861) [2021-04-16 11:02:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][280/1251] eta 0:05:23 lr 0.000252 time 0.3051 (0.3337) loss 3.9360 (3.3250) grad_norm 2.2054 (1.9852) [2021-04-16 11:02:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][290/1251] eta 0:05:19 lr 0.000252 time 0.3035 (0.3322) loss 3.3154 (3.3206) grad_norm 1.9791 (1.9879) [2021-04-16 11:02:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][300/1251] eta 0:05:14 lr 0.000252 time 0.2914 (0.3308) loss 3.9500 (3.3266) grad_norm 2.1200 (1.9859) [2021-04-16 11:02:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][310/1251] eta 0:05:09 lr 0.000252 time 0.2519 (0.3290) loss 3.8766 (3.3297) grad_norm 2.1704 (1.9886) [2021-04-16 11:02:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][320/1251] eta 0:05:05 lr 0.000252 time 0.2733 (0.3277) loss 4.1476 (3.3345) grad_norm 1.9442 (1.9903) [2021-04-16 11:02:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][330/1251] eta 0:05:00 lr 0.000252 time 0.2810 (0.3264) loss 3.7530 (3.3373) grad_norm 1.7027 (1.9887) [2021-04-16 11:02:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][340/1251] eta 0:04:56 lr 0.000252 time 0.2757 (0.3251) loss 3.4757 (3.3326) grad_norm 2.0738 (1.9868) [2021-04-16 11:02:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][350/1251] eta 0:04:51 lr 0.000252 time 0.2990 (0.3238) loss 3.2920 (3.3347) grad_norm 1.9617 (1.9857) [2021-04-16 11:02:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][360/1251] eta 0:04:47 lr 0.000252 time 0.2583 (0.3227) loss 3.2920 (3.3354) grad_norm 2.0012 (1.9826) [2021-04-16 11:02:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][370/1251] eta 0:04:43 lr 0.000252 time 0.2776 (0.3219) loss 3.7192 (3.3363) grad_norm 2.2557 (1.9842) [2021-04-16 11:02:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][380/1251] eta 0:04:39 lr 0.000252 time 0.2645 (0.3206) loss 3.9383 (3.3405) grad_norm 1.9315 (nan) [2021-04-16 11:02:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][390/1251] eta 0:04:35 lr 0.000252 time 0.2735 (0.3199) loss 3.7014 (3.3410) grad_norm 2.2518 (nan) [2021-04-16 11:02:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][400/1251] eta 0:04:31 lr 0.000252 time 0.2864 (0.3187) loss 3.7332 (3.3345) grad_norm 2.3949 (nan) [2021-04-16 11:02:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][410/1251] eta 0:04:27 lr 0.000252 time 0.2958 (0.3177) loss 2.8993 (3.3315) grad_norm 1.7806 (nan) [2021-04-16 11:02:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][420/1251] eta 0:04:23 lr 0.000252 time 0.2689 (0.3167) loss 3.6427 (3.3305) grad_norm 1.6429 (nan) [2021-04-16 11:02:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][430/1251] eta 0:04:19 lr 0.000251 time 0.2879 (0.3158) loss 3.6450 (3.3279) grad_norm 2.3832 (nan) [2021-04-16 11:02:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][440/1251] eta 0:04:15 lr 0.000251 time 0.2795 (0.3148) loss 2.8485 (3.3271) grad_norm 1.9233 (nan) [2021-04-16 11:02:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][450/1251] eta 0:04:11 lr 0.000251 time 0.2738 (0.3139) loss 3.6008 (3.3273) grad_norm 2.2192 (nan) [2021-04-16 11:02:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][460/1251] eta 0:04:07 lr 0.000251 time 0.2580 (0.3130) loss 3.2760 (3.3271) grad_norm 1.8088 (nan) [2021-04-16 11:02:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][470/1251] eta 0:04:03 lr 0.000251 time 0.2705 (0.3121) loss 3.7285 (3.3230) grad_norm 1.9073 (nan) [2021-04-16 11:02:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][480/1251] eta 0:04:00 lr 0.000251 time 0.2669 (0.3114) loss 4.1841 (3.3250) grad_norm 2.0836 (nan) [2021-04-16 11:02:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][490/1251] eta 0:03:56 lr 0.000251 time 0.2704 (0.3107) loss 3.6891 (3.3210) grad_norm 1.9255 (nan) [2021-04-16 11:03:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][500/1251] eta 0:03:52 lr 0.000251 time 0.2680 (0.3099) loss 2.6308 (3.3149) grad_norm 2.1282 (nan) [2021-04-16 11:03:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][510/1251] eta 0:03:49 lr 0.000251 time 0.2691 (0.3092) loss 3.0549 (3.3183) grad_norm 2.2789 (nan) [2021-04-16 11:03:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][520/1251] eta 0:03:45 lr 0.000251 time 0.2642 (0.3086) loss 2.7878 (3.3126) grad_norm 2.3837 (nan) [2021-04-16 11:03:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][530/1251] eta 0:03:42 lr 0.000251 time 0.2724 (0.3080) loss 3.3747 (3.3177) grad_norm 1.9911 (nan) [2021-04-16 11:03:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][540/1251] eta 0:03:38 lr 0.000251 time 0.2845 (0.3074) loss 2.7346 (3.3162) grad_norm 2.0715 (nan) [2021-04-16 11:03:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][550/1251] eta 0:03:35 lr 0.000251 time 0.2660 (0.3069) loss 2.4913 (3.3160) grad_norm 1.9843 (nan) [2021-04-16 11:03:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][560/1251] eta 0:03:31 lr 0.000251 time 0.2876 (0.3065) loss 4.0383 (3.3099) grad_norm 1.8428 (nan) [2021-04-16 11:03:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][570/1251] eta 0:03:28 lr 0.000251 time 0.2626 (0.3064) loss 3.7559 (3.3133) grad_norm 2.1270 (nan) [2021-04-16 11:03:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][580/1251] eta 0:03:25 lr 0.000251 time 0.3043 (0.3061) loss 3.6950 (3.3163) grad_norm 2.0317 (nan) [2021-04-16 11:03:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][590/1251] eta 0:03:22 lr 0.000251 time 0.2859 (0.3060) loss 3.4590 (3.3126) grad_norm 1.9257 (nan) [2021-04-16 11:03:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][600/1251] eta 0:03:18 lr 0.000251 time 0.2696 (0.3053) loss 3.7661 (3.3149) grad_norm 2.1131 (nan) [2021-04-16 11:03:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][610/1251] eta 0:03:15 lr 0.000251 time 0.2984 (0.3049) loss 3.5228 (3.3097) grad_norm 1.9262 (nan) [2021-04-16 11:03:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][620/1251] eta 0:03:12 lr 0.000251 time 0.2945 (0.3044) loss 3.5544 (3.3089) grad_norm 1.8579 (nan) [2021-04-16 11:03:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][630/1251] eta 0:03:08 lr 0.000251 time 0.2770 (0.3040) loss 3.1122 (3.3067) grad_norm 2.0239 (nan) [2021-04-16 11:03:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][640/1251] eta 0:03:05 lr 0.000251 time 0.2972 (0.3037) loss 3.6705 (3.3070) grad_norm 1.9874 (nan) [2021-04-16 11:03:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][650/1251] eta 0:03:02 lr 0.000251 time 0.2624 (0.3032) loss 2.0836 (3.3047) grad_norm 2.1590 (nan) [2021-04-16 11:03:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][660/1251] eta 0:02:59 lr 0.000251 time 0.2784 (0.3029) loss 3.6031 (3.3047) grad_norm 1.9336 (nan) [2021-04-16 11:03:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][670/1251] eta 0:02:55 lr 0.000251 time 0.2990 (0.3027) loss 3.3368 (3.3055) grad_norm 2.2303 (nan) [2021-04-16 11:03:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][680/1251] eta 0:02:52 lr 0.000251 time 0.2773 (0.3022) loss 3.8396 (3.3097) grad_norm 1.7889 (nan) [2021-04-16 11:03:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][690/1251] eta 0:02:49 lr 0.000251 time 0.2726 (0.3018) loss 2.9568 (3.3126) grad_norm 1.9848 (nan) [2021-04-16 11:03:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][700/1251] eta 0:02:46 lr 0.000251 time 0.2476 (0.3014) loss 3.7258 (3.3130) grad_norm 1.7603 (nan) [2021-04-16 11:04:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][710/1251] eta 0:02:42 lr 0.000250 time 0.2849 (0.3013) loss 3.6196 (3.3148) grad_norm 2.2497 (nan) [2021-04-16 11:04:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][720/1251] eta 0:02:39 lr 0.000250 time 0.2761 (0.3010) loss 2.9595 (3.3143) grad_norm 2.1358 (nan) [2021-04-16 11:04:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][730/1251] eta 0:02:36 lr 0.000250 time 0.2801 (0.3009) loss 4.1878 (3.3195) grad_norm 1.9097 (nan) [2021-04-16 11:04:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][740/1251] eta 0:02:33 lr 0.000250 time 0.2970 (0.3007) loss 3.5641 (3.3213) grad_norm 1.7231 (nan) [2021-04-16 11:04:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][750/1251] eta 0:02:30 lr 0.000250 time 0.2722 (0.3003) loss 3.8456 (3.3201) grad_norm 2.1926 (nan) [2021-04-16 11:04:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][760/1251] eta 0:02:27 lr 0.000250 time 0.2682 (0.3000) loss 2.9958 (3.3213) grad_norm 2.2055 (nan) [2021-04-16 11:04:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][770/1251] eta 0:02:24 lr 0.000250 time 0.2660 (0.2998) loss 2.6819 (3.3203) grad_norm 1.8502 (nan) [2021-04-16 11:04:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][780/1251] eta 0:02:21 lr 0.000250 time 0.2602 (0.2995) loss 2.7573 (3.3190) grad_norm 1.9254 (nan) [2021-04-16 11:04:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][790/1251] eta 0:02:17 lr 0.000250 time 0.2635 (0.2992) loss 3.8322 (3.3208) grad_norm 1.8707 (nan) [2021-04-16 11:04:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][800/1251] eta 0:02:14 lr 0.000250 time 0.2975 (0.2990) loss 4.0617 (3.3218) grad_norm 1.8904 (nan) [2021-04-16 11:04:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][810/1251] eta 0:02:11 lr 0.000250 time 0.2796 (0.2987) loss 2.5969 (3.3185) grad_norm 1.7417 (nan) [2021-04-16 11:04:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][820/1251] eta 0:02:08 lr 0.000250 time 0.2681 (0.2983) loss 3.6485 (3.3201) grad_norm 2.0211 (nan) [2021-04-16 11:04:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][830/1251] eta 0:02:05 lr 0.000250 time 0.2716 (0.2981) loss 3.2403 (3.3207) grad_norm 2.2504 (nan) [2021-04-16 11:04:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][840/1251] eta 0:02:02 lr 0.000250 time 0.2717 (0.2979) loss 4.4032 (3.3206) grad_norm 2.1352 (nan) [2021-04-16 11:04:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][850/1251] eta 0:01:59 lr 0.000250 time 0.2944 (0.2976) loss 2.8813 (3.3214) grad_norm 1.9599 (nan) [2021-04-16 11:04:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][860/1251] eta 0:01:56 lr 0.000250 time 0.2655 (0.2975) loss 3.0258 (3.3197) grad_norm 2.0066 (nan) [2021-04-16 11:04:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][870/1251] eta 0:01:53 lr 0.000250 time 0.2817 (0.2974) loss 2.7443 (3.3201) grad_norm 2.0782 (nan) [2021-04-16 11:04:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][880/1251] eta 0:01:50 lr 0.000250 time 0.2617 (0.2973) loss 3.3886 (3.3202) grad_norm 2.0037 (nan) [2021-04-16 11:04:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][890/1251] eta 0:01:47 lr 0.000250 time 0.2863 (0.2972) loss 2.6387 (3.3226) grad_norm 1.9681 (nan) [2021-04-16 11:04:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][900/1251] eta 0:01:44 lr 0.000250 time 0.2659 (0.2969) loss 3.8660 (3.3248) grad_norm 2.0000 (nan) [2021-04-16 11:04:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][910/1251] eta 0:01:41 lr 0.000250 time 0.2492 (0.2967) loss 3.3106 (3.3228) grad_norm 1.9113 (nan) [2021-04-16 11:05:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][920/1251] eta 0:01:38 lr 0.000250 time 0.2705 (0.2965) loss 3.5253 (3.3217) grad_norm 2.2828 (nan) [2021-04-16 11:05:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][930/1251] eta 0:01:35 lr 0.000250 time 0.2732 (0.2962) loss 2.2839 (3.3190) grad_norm 1.7020 (nan) [2021-04-16 11:05:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][940/1251] eta 0:01:32 lr 0.000250 time 0.2624 (0.2961) loss 3.3719 (3.3187) grad_norm 1.7857 (nan) [2021-04-16 11:05:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][950/1251] eta 0:01:29 lr 0.000250 time 0.3042 (0.2960) loss 2.7331 (3.3171) grad_norm 1.9774 (nan) [2021-04-16 11:05:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][960/1251] eta 0:01:26 lr 0.000250 time 0.2892 (0.2958) loss 2.6771 (3.3150) grad_norm 1.9203 (nan) [2021-04-16 11:05:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][970/1251] eta 0:01:23 lr 0.000250 time 0.2905 (0.2958) loss 3.0405 (3.3131) grad_norm 1.6904 (nan) [2021-04-16 11:05:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][980/1251] eta 0:01:20 lr 0.000250 time 0.2734 (0.2956) loss 3.5182 (3.3156) grad_norm 1.7600 (nan) [2021-04-16 11:05:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][990/1251] eta 0:01:17 lr 0.000250 time 0.2827 (0.2955) loss 4.1723 (3.3170) grad_norm 2.1599 (nan) [2021-04-16 11:05:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][1000/1251] eta 0:01:14 lr 0.000249 time 0.2774 (0.2953) loss 2.7399 (3.3153) grad_norm 2.0660 (nan) [2021-04-16 11:05:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][1010/1251] eta 0:01:11 lr 0.000249 time 0.2887 (0.2952) loss 3.0197 (3.3130) grad_norm 2.2673 (nan) [2021-04-16 11:05:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][1020/1251] eta 0:01:08 lr 0.000249 time 0.2839 (0.2950) loss 3.0946 (3.3120) grad_norm 1.8026 (nan) [2021-04-16 11:05:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][1030/1251] eta 0:01:05 lr 0.000249 time 0.2650 (0.2948) loss 3.1547 (3.3126) grad_norm 2.4009 (nan) [2021-04-16 11:05:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][1040/1251] eta 0:01:02 lr 0.000249 time 0.2511 (0.2946) loss 3.6326 (3.3122) grad_norm 2.2561 (nan) [2021-04-16 11:05:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][1050/1251] eta 0:00:59 lr 0.000249 time 0.2569 (0.2946) loss 3.4897 (3.3130) grad_norm 2.5126 (nan) [2021-04-16 11:05:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][1060/1251] eta 0:00:56 lr 0.000249 time 0.2828 (0.2944) loss 2.9803 (3.3120) grad_norm 1.9331 (nan) [2021-04-16 11:05:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][1070/1251] eta 0:00:53 lr 0.000249 time 0.2679 (0.2942) loss 3.2689 (3.3126) grad_norm 1.9670 (nan) [2021-04-16 11:05:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][1080/1251] eta 0:00:50 lr 0.000249 time 0.2902 (0.2941) loss 2.5730 (3.3096) grad_norm 2.2211 (nan) [2021-04-16 11:05:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][1090/1251] eta 0:00:47 lr 0.000249 time 0.2895 (0.2940) loss 3.6100 (3.3085) grad_norm 2.0132 (nan) [2021-04-16 11:05:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][1100/1251] eta 0:00:44 lr 0.000249 time 0.2942 (0.2938) loss 3.8935 (3.3101) grad_norm 1.8836 (nan) [2021-04-16 11:05:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][1110/1251] eta 0:00:41 lr 0.000249 time 0.2735 (0.2937) loss 3.0024 (3.3073) grad_norm 1.7090 (nan) [2021-04-16 11:05:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][1120/1251] eta 0:00:38 lr 0.000249 time 0.2598 (0.2936) loss 3.0842 (3.3068) grad_norm 1.7810 (nan) [2021-04-16 11:05:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][1130/1251] eta 0:00:35 lr 0.000249 time 0.2844 (0.2934) loss 2.7113 (3.3062) grad_norm 2.1099 (nan) [2021-04-16 11:06:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][1140/1251] eta 0:00:32 lr 0.000249 time 0.2705 (0.2932) loss 4.0530 (3.3049) grad_norm 1.7796 (nan) [2021-04-16 11:06:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][1150/1251] eta 0:00:29 lr 0.000249 time 0.3042 (0.2933) loss 3.1586 (3.3051) grad_norm 2.3243 (nan) [2021-04-16 11:06:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][1160/1251] eta 0:00:26 lr 0.000249 time 0.2659 (0.2933) loss 4.2421 (3.3050) grad_norm 1.9264 (nan) [2021-04-16 11:06:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][1170/1251] eta 0:00:23 lr 0.000249 time 0.2954 (0.2934) loss 3.3635 (3.3040) grad_norm 1.6612 (nan) [2021-04-16 11:06:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][1180/1251] eta 0:00:20 lr 0.000249 time 0.2598 (0.2932) loss 3.8412 (3.3051) grad_norm 1.8822 (nan) [2021-04-16 11:06:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][1190/1251] eta 0:00:17 lr 0.000249 time 0.2812 (0.2931) loss 2.5795 (3.3020) grad_norm 2.0754 (nan) [2021-04-16 11:06:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][1200/1251] eta 0:00:14 lr 0.000249 time 0.2773 (0.2930) loss 3.4632 (3.3020) grad_norm 1.8704 (nan) [2021-04-16 11:06:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][1210/1251] eta 0:00:12 lr 0.000249 time 0.2904 (0.2928) loss 3.1006 (3.3034) grad_norm 2.0260 (nan) [2021-04-16 11:06:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][1220/1251] eta 0:00:09 lr 0.000249 time 0.2732 (0.2927) loss 3.3791 (3.3036) grad_norm 2.0101 (nan) [2021-04-16 11:06:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][1230/1251] eta 0:00:06 lr 0.000249 time 0.2879 (0.2926) loss 3.5358 (3.3049) grad_norm 2.1918 (nan) [2021-04-16 11:06:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][1240/1251] eta 0:00:03 lr 0.000249 time 0.2486 (0.2925) loss 3.5049 (3.3064) grad_norm 2.3502 (nan) [2021-04-16 11:06:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [201/300][1250/1251] eta 0:00:00 lr 0.000249 time 0.2480 (0.2921) loss 3.4994 (3.3067) grad_norm 2.0298 (nan) [2021-04-16 11:06:36 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 201 training takes 0:06:09 [2021-04-16 11:06:36 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_201.pth saving...... [2021-04-16 11:06:47 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_201.pth saved !!! [2021-04-16 11:06:48 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.126 (1.126) Loss 0.9243 (0.9243) Acc@1 78.320 (78.320) Acc@5 94.824 (94.824) [2021-04-16 11:06:49 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.125 (0.214) Loss 0.8787 (0.9013) Acc@1 78.711 (78.533) Acc@5 94.922 (94.593) [2021-04-16 11:06:52 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.151 (0.240) Loss 0.8593 (0.8996) Acc@1 79.688 (78.562) Acc@5 94.531 (94.545) [2021-04-16 11:06:54 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.622 (0.231) Loss 0.8950 (0.9087) Acc@1 78.418 (78.377) Acc@5 95.020 (94.430) [2021-04-16 11:06:56 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.218) Loss 0.9250 (0.9051) Acc@1 77.051 (78.344) Acc@5 94.629 (94.467) [2021-04-16 11:07:02 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 78.372 Acc@5 94.450 [2021-04-16 11:07:02 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 78.4% [2021-04-16 11:07:02 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 78.37% [2021-04-16 11:07:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][0/1251] eta 3:38:33 lr 0.000249 time 10.4828 (10.4828) loss 3.0622 (3.0622) grad_norm 1.6031 (1.6031) [2021-04-16 11:07:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][10/1251] eta 0:24:48 lr 0.000249 time 0.2958 (1.1997) loss 2.7318 (3.2515) grad_norm 1.7631 (1.8817) [2021-04-16 11:07:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][20/1251] eta 0:15:36 lr 0.000249 time 0.2911 (0.7610) loss 3.7970 (3.3690) grad_norm 1.8897 (1.9645) [2021-04-16 11:07:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][30/1251] eta 0:12:17 lr 0.000248 time 0.2783 (0.6043) loss 3.3188 (3.3412) grad_norm 1.9886 (1.9729) [2021-04-16 11:07:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][40/1251] eta 0:10:35 lr 0.000248 time 0.2781 (0.5244) loss 3.3624 (3.4034) grad_norm 2.0510 (2.0034) [2021-04-16 11:07:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][50/1251] eta 0:09:31 lr 0.000248 time 0.2933 (0.4759) loss 3.5297 (3.4009) grad_norm 1.9086 (2.0071) [2021-04-16 11:07:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][60/1251] eta 0:08:47 lr 0.000248 time 0.2668 (0.4426) loss 3.5804 (3.3913) grad_norm 2.2049 (2.0055) [2021-04-16 11:07:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][70/1251] eta 0:08:15 lr 0.000248 time 0.3030 (0.4197) loss 3.7547 (3.4047) grad_norm 2.5808 (2.0397) [2021-04-16 11:07:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][80/1251] eta 0:07:52 lr 0.000248 time 0.2766 (0.4039) loss 3.3437 (3.3858) grad_norm 1.9077 (2.0408) [2021-04-16 11:07:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][90/1251] eta 0:07:34 lr 0.000248 time 0.2925 (0.3916) loss 2.0380 (3.3760) grad_norm 1.9575 (2.0304) [2021-04-16 11:07:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][100/1251] eta 0:07:17 lr 0.000248 time 0.2687 (0.3802) loss 4.1919 (3.3587) grad_norm 2.2903 (2.0491) [2021-04-16 11:07:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][110/1251] eta 0:07:03 lr 0.000248 time 0.2854 (0.3711) loss 2.3658 (3.3447) grad_norm 1.8628 (2.0444) [2021-04-16 11:07:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][120/1251] eta 0:06:50 lr 0.000248 time 0.2608 (0.3633) loss 2.4214 (3.3449) grad_norm 1.9081 (2.0511) [2021-04-16 11:07:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][130/1251] eta 0:06:39 lr 0.000248 time 0.2591 (0.3563) loss 2.7022 (3.3593) grad_norm 1.9128 (2.0556) [2021-04-16 11:07:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][140/1251] eta 0:06:29 lr 0.000248 time 0.2791 (0.3508) loss 3.2877 (3.3680) grad_norm 1.8729 (2.0511) [2021-04-16 11:07:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][150/1251] eta 0:06:22 lr 0.000248 time 0.2788 (0.3470) loss 3.6919 (3.3526) grad_norm 2.0773 (2.0497) [2021-04-16 11:07:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][160/1251] eta 0:06:15 lr 0.000248 time 0.2677 (0.3439) loss 3.3567 (3.3590) grad_norm 1.8904 (2.0548) [2021-04-16 11:08:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][170/1251] eta 0:06:07 lr 0.000248 time 0.2640 (0.3402) loss 3.9133 (3.3724) grad_norm 2.0343 (2.0587) [2021-04-16 11:08:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][180/1251] eta 0:06:01 lr 0.000248 time 0.3070 (0.3375) loss 3.0494 (3.3652) grad_norm 1.9628 (2.0607) [2021-04-16 11:08:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][190/1251] eta 0:05:54 lr 0.000248 time 0.2882 (0.3345) loss 3.3366 (3.3441) grad_norm 1.8457 (2.0580) [2021-04-16 11:08:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][200/1251] eta 0:05:48 lr 0.000248 time 0.2910 (0.3318) loss 3.7500 (3.3519) grad_norm 1.8211 (2.0633) [2021-04-16 11:08:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][210/1251] eta 0:05:42 lr 0.000248 time 0.2711 (0.3293) loss 3.0749 (3.3609) grad_norm 2.3073 (2.0631) [2021-04-16 11:08:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][220/1251] eta 0:05:37 lr 0.000248 time 0.2736 (0.3269) loss 3.5669 (3.3622) grad_norm 1.7775 (2.0558) [2021-04-16 11:08:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][230/1251] eta 0:05:31 lr 0.000248 time 0.2558 (0.3246) loss 3.6010 (3.3646) grad_norm 2.3217 (2.0524) [2021-04-16 11:08:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][240/1251] eta 0:05:26 lr 0.000248 time 0.2724 (0.3226) loss 2.9911 (3.3565) grad_norm 1.7806 (2.0458) [2021-04-16 11:08:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][250/1251] eta 0:05:21 lr 0.000248 time 0.2649 (0.3208) loss 3.5005 (3.3703) grad_norm 1.8670 (2.0461) [2021-04-16 11:08:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][260/1251] eta 0:05:16 lr 0.000248 time 0.2966 (0.3191) loss 3.3977 (3.3732) grad_norm 2.0536 (2.0471) [2021-04-16 11:08:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][270/1251] eta 0:05:11 lr 0.000248 time 0.2723 (0.3176) loss 2.6981 (3.3737) grad_norm 2.0228 (2.0437) [2021-04-16 11:08:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][280/1251] eta 0:05:07 lr 0.000248 time 0.2573 (0.3164) loss 3.7938 (3.3679) grad_norm 1.8579 (2.0409) [2021-04-16 11:08:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][290/1251] eta 0:05:02 lr 0.000248 time 0.2561 (0.3150) loss 3.4072 (3.3723) grad_norm 1.6494 (2.0367) [2021-04-16 11:08:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][300/1251] eta 0:04:58 lr 0.000248 time 0.2918 (0.3138) loss 3.7222 (3.3724) grad_norm 1.9334 (2.0322) [2021-04-16 11:08:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][310/1251] eta 0:04:54 lr 0.000247 time 0.3162 (0.3129) loss 3.8355 (3.3727) grad_norm 2.1417 (2.0334) [2021-04-16 11:08:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][320/1251] eta 0:04:50 lr 0.000247 time 0.2840 (0.3120) loss 2.7892 (3.3701) grad_norm 1.7638 (2.0326) [2021-04-16 11:08:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][330/1251] eta 0:04:46 lr 0.000247 time 0.2704 (0.3111) loss 3.6224 (3.3772) grad_norm 3.0973 (2.0348) [2021-04-16 11:08:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][340/1251] eta 0:04:42 lr 0.000247 time 0.2668 (0.3101) loss 2.4635 (3.3667) grad_norm 2.0068 (2.0357) [2021-04-16 11:08:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][350/1251] eta 0:04:39 lr 0.000247 time 0.2789 (0.3098) loss 2.5321 (3.3585) grad_norm 2.2872 (2.0362) [2021-04-16 11:08:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][360/1251] eta 0:04:35 lr 0.000247 time 0.2782 (0.3092) loss 3.2904 (3.3527) grad_norm 2.0092 (2.0334) [2021-04-16 11:08:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][370/1251] eta 0:04:32 lr 0.000247 time 0.2875 (0.3088) loss 3.6843 (3.3462) grad_norm 2.3978 (2.0387) [2021-04-16 11:08:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][380/1251] eta 0:04:28 lr 0.000247 time 0.3121 (0.3084) loss 4.0106 (3.3435) grad_norm 2.3584 (2.0396) [2021-04-16 11:09:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][390/1251] eta 0:04:24 lr 0.000247 time 0.2907 (0.3076) loss 3.7022 (3.3445) grad_norm 2.0662 (2.0384) [2021-04-16 11:09:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][400/1251] eta 0:04:21 lr 0.000247 time 0.2814 (0.3069) loss 3.8968 (3.3466) grad_norm 2.7695 (2.0403) [2021-04-16 11:09:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][410/1251] eta 0:04:17 lr 0.000247 time 0.2801 (0.3061) loss 3.5240 (3.3411) grad_norm 2.0056 (2.0444) [2021-04-16 11:09:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][420/1251] eta 0:04:13 lr 0.000247 time 0.2977 (0.3054) loss 3.4319 (3.3469) grad_norm 2.0498 (2.0587) [2021-04-16 11:09:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][430/1251] eta 0:04:10 lr 0.000247 time 0.2621 (0.3048) loss 4.0564 (3.3481) grad_norm 2.1103 (2.0593) [2021-04-16 11:09:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][440/1251] eta 0:04:06 lr 0.000247 time 0.2666 (0.3045) loss 2.6430 (3.3536) grad_norm 1.8466 (2.0578) [2021-04-16 11:09:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][450/1251] eta 0:04:03 lr 0.000247 time 0.2800 (0.3043) loss 4.1469 (3.3599) grad_norm 1.9699 (2.0594) [2021-04-16 11:09:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][460/1251] eta 0:04:00 lr 0.000247 time 0.2564 (0.3041) loss 3.6451 (3.3574) grad_norm 1.7738 (2.0575) [2021-04-16 11:09:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][470/1251] eta 0:03:57 lr 0.000247 time 0.2758 (0.3036) loss 3.0491 (3.3598) grad_norm 2.3413 (2.0581) [2021-04-16 11:09:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][480/1251] eta 0:03:53 lr 0.000247 time 0.2768 (0.3031) loss 3.5393 (3.3627) grad_norm 1.8360 (2.0603) [2021-04-16 11:09:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][490/1251] eta 0:03:50 lr 0.000247 time 0.2820 (0.3025) loss 2.8877 (3.3520) grad_norm 2.3047 (2.0599) [2021-04-16 11:09:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][500/1251] eta 0:03:46 lr 0.000247 time 0.2737 (0.3020) loss 2.8247 (3.3490) grad_norm 2.2672 (2.0637) [2021-04-16 11:09:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][510/1251] eta 0:03:43 lr 0.000247 time 0.2956 (0.3015) loss 3.4575 (3.3477) grad_norm 1.8545 (2.0634) [2021-04-16 11:09:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][520/1251] eta 0:03:40 lr 0.000247 time 0.2727 (0.3010) loss 3.4468 (3.3440) grad_norm 1.7726 (2.0621) [2021-04-16 11:09:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][530/1251] eta 0:03:36 lr 0.000247 time 0.2676 (0.3006) loss 3.8857 (3.3474) grad_norm 1.7981 (2.0624) [2021-04-16 11:09:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][540/1251] eta 0:03:33 lr 0.000247 time 0.2685 (0.3002) loss 3.5374 (3.3462) grad_norm 1.6832 (2.0607) [2021-04-16 11:09:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][550/1251] eta 0:03:30 lr 0.000247 time 0.2732 (0.2998) loss 2.2922 (3.3472) grad_norm 2.3529 (2.0612) [2021-04-16 11:09:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][560/1251] eta 0:03:26 lr 0.000247 time 0.2649 (0.2993) loss 3.5219 (3.3495) grad_norm 1.9407 (2.0584) [2021-04-16 11:09:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][570/1251] eta 0:03:23 lr 0.000247 time 0.2656 (0.2992) loss 3.3366 (3.3515) grad_norm 1.9664 (2.0589) [2021-04-16 11:09:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][580/1251] eta 0:03:20 lr 0.000247 time 0.2687 (0.2989) loss 3.4708 (3.3503) grad_norm 2.0498 (2.0601) [2021-04-16 11:09:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][590/1251] eta 0:03:17 lr 0.000246 time 0.2846 (0.2987) loss 3.8190 (3.3511) grad_norm 2.2554 (2.0603) [2021-04-16 11:10:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][600/1251] eta 0:03:14 lr 0.000246 time 0.2786 (0.2983) loss 3.8388 (3.3546) grad_norm 2.5363 (2.0612) [2021-04-16 11:10:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][610/1251] eta 0:03:10 lr 0.000246 time 0.3021 (0.2980) loss 3.5076 (3.3553) grad_norm 2.2225 (2.0611) [2021-04-16 11:10:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][620/1251] eta 0:03:07 lr 0.000246 time 0.2722 (0.2976) loss 3.4023 (3.3518) grad_norm 2.6133 (2.0601) [2021-04-16 11:10:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][630/1251] eta 0:03:04 lr 0.000246 time 0.2595 (0.2973) loss 3.0627 (3.3458) grad_norm 2.2255 (2.0611) [2021-04-16 11:10:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][640/1251] eta 0:03:01 lr 0.000246 time 0.2794 (0.2970) loss 2.8789 (3.3446) grad_norm 1.7820 (2.0597) [2021-04-16 11:10:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][650/1251] eta 0:02:58 lr 0.000246 time 0.2751 (0.2969) loss 3.7206 (3.3441) grad_norm 2.3296 (2.0588) [2021-04-16 11:10:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][660/1251] eta 0:02:55 lr 0.000246 time 0.2805 (0.2965) loss 2.1768 (3.3447) grad_norm 1.8221 (2.0596) [2021-04-16 11:10:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][670/1251] eta 0:02:52 lr 0.000246 time 0.2690 (0.2963) loss 2.6676 (3.3424) grad_norm 1.9968 (2.0575) [2021-04-16 11:10:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][680/1251] eta 0:02:48 lr 0.000246 time 0.2588 (0.2960) loss 3.4254 (3.3420) grad_norm 1.9521 (2.0575) [2021-04-16 11:10:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][690/1251] eta 0:02:45 lr 0.000246 time 0.2922 (0.2957) loss 3.2847 (3.3418) grad_norm 2.1244 (2.0567) [2021-04-16 11:10:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][700/1251] eta 0:02:42 lr 0.000246 time 0.3025 (0.2956) loss 2.9320 (3.3382) grad_norm 2.0009 (2.0552) [2021-04-16 11:10:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][710/1251] eta 0:02:39 lr 0.000246 time 0.2846 (0.2953) loss 3.5500 (3.3329) grad_norm 2.0778 (2.0545) [2021-04-16 11:10:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][720/1251] eta 0:02:36 lr 0.000246 time 0.2923 (0.2952) loss 3.3724 (3.3324) grad_norm 2.1488 (2.0548) [2021-04-16 11:10:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][730/1251] eta 0:02:33 lr 0.000246 time 0.2782 (0.2949) loss 3.1519 (3.3319) grad_norm 2.0173 (2.0533) [2021-04-16 11:10:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][740/1251] eta 0:02:30 lr 0.000246 time 0.3294 (0.2949) loss 3.7755 (3.3326) grad_norm 1.6269 (2.0507) [2021-04-16 11:10:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][750/1251] eta 0:02:27 lr 0.000246 time 0.2712 (0.2946) loss 3.1691 (3.3321) grad_norm 2.0663 (2.0490) [2021-04-16 11:10:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][760/1251] eta 0:02:24 lr 0.000246 time 0.2435 (0.2945) loss 3.6011 (3.3292) grad_norm 1.7886 (2.0484) [2021-04-16 11:10:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][770/1251] eta 0:02:21 lr 0.000246 time 0.2834 (0.2943) loss 2.3713 (3.3290) grad_norm 2.1949 (2.0476) [2021-04-16 11:10:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][780/1251] eta 0:02:18 lr 0.000246 time 0.2851 (0.2940) loss 3.5870 (3.3268) grad_norm 2.2166 (2.0460) [2021-04-16 11:10:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][790/1251] eta 0:02:15 lr 0.000246 time 0.2735 (0.2938) loss 2.8932 (3.3245) grad_norm 1.8579 (2.0451) [2021-04-16 11:10:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][800/1251] eta 0:02:12 lr 0.000246 time 0.2900 (0.2937) loss 2.5778 (3.3240) grad_norm 2.0347 (2.0449) [2021-04-16 11:11:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][810/1251] eta 0:02:09 lr 0.000246 time 0.2623 (0.2935) loss 2.3712 (3.3218) grad_norm 1.9831 (2.0442) [2021-04-16 11:11:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][820/1251] eta 0:02:06 lr 0.000246 time 0.2784 (0.2933) loss 3.8358 (3.3210) grad_norm 1.8996 (2.0427) [2021-04-16 11:11:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][830/1251] eta 0:02:03 lr 0.000246 time 0.2493 (0.2931) loss 3.1520 (3.3205) grad_norm 1.8530 (2.0414) [2021-04-16 11:11:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][840/1251] eta 0:02:00 lr 0.000246 time 0.2661 (0.2929) loss 2.8850 (3.3210) grad_norm 1.9391 (2.0394) [2021-04-16 11:11:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][850/1251] eta 0:01:57 lr 0.000246 time 0.2472 (0.2928) loss 2.5242 (3.3202) grad_norm 2.2794 (2.0397) [2021-04-16 11:11:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][860/1251] eta 0:01:54 lr 0.000246 time 0.2580 (0.2926) loss 3.6155 (3.3215) grad_norm 1.7348 (2.0377) [2021-04-16 11:11:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][870/1251] eta 0:01:51 lr 0.000245 time 0.2615 (0.2924) loss 3.4702 (3.3187) grad_norm 2.4614 (2.0371) [2021-04-16 11:11:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][880/1251] eta 0:01:48 lr 0.000245 time 0.4210 (0.2925) loss 2.8647 (3.3193) grad_norm 2.2284 (2.0389) [2021-04-16 11:11:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][890/1251] eta 0:01:45 lr 0.000245 time 0.2786 (0.2923) loss 2.5606 (3.3137) grad_norm 1.8622 (2.0405) [2021-04-16 11:11:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][900/1251] eta 0:01:42 lr 0.000245 time 0.2625 (0.2921) loss 2.1890 (3.3123) grad_norm 2.0070 (2.0403) [2021-04-16 11:11:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][910/1251] eta 0:01:39 lr 0.000245 time 0.2517 (0.2919) loss 3.7929 (3.3109) grad_norm 1.7390 (2.0403) [2021-04-16 11:11:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][920/1251] eta 0:01:36 lr 0.000245 time 0.2501 (0.2919) loss 3.1115 (3.3125) grad_norm 1.7525 (2.0387) [2021-04-16 11:11:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][930/1251] eta 0:01:33 lr 0.000245 time 0.2992 (0.2917) loss 3.6770 (3.3141) grad_norm 1.9056 (2.0377) [2021-04-16 11:11:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][940/1251] eta 0:01:30 lr 0.000245 time 0.2589 (0.2918) loss 2.5314 (3.3126) grad_norm 1.7940 (2.0367) [2021-04-16 11:11:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][950/1251] eta 0:01:27 lr 0.000245 time 0.2909 (0.2917) loss 3.4539 (3.3091) grad_norm 2.0125 (2.0361) [2021-04-16 11:11:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][960/1251] eta 0:01:24 lr 0.000245 time 0.2759 (0.2916) loss 3.9310 (3.3107) grad_norm 1.7562 (2.0369) [2021-04-16 11:11:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][970/1251] eta 0:01:21 lr 0.000245 time 0.2747 (0.2915) loss 2.3429 (3.3121) grad_norm 1.7950 (2.0363) [2021-04-16 11:11:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][980/1251] eta 0:01:18 lr 0.000245 time 0.2760 (0.2913) loss 3.8459 (3.3152) grad_norm 1.9170 (2.0361) [2021-04-16 11:11:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][990/1251] eta 0:01:15 lr 0.000245 time 0.3020 (0.2912) loss 3.2005 (3.3164) grad_norm 1.8800 (2.0360) [2021-04-16 11:11:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][1000/1251] eta 0:01:13 lr 0.000245 time 0.2702 (0.2910) loss 3.6945 (3.3172) grad_norm 1.9927 (2.0356) [2021-04-16 11:11:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][1010/1251] eta 0:01:10 lr 0.000245 time 0.2604 (0.2909) loss 3.4937 (3.3173) grad_norm 1.9921 (2.0356) [2021-04-16 11:11:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][1020/1251] eta 0:01:07 lr 0.000245 time 0.2550 (0.2909) loss 3.6298 (3.3187) grad_norm 1.8310 (2.0369) [2021-04-16 11:12:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][1030/1251] eta 0:01:04 lr 0.000245 time 0.2767 (0.2908) loss 3.2790 (3.3203) grad_norm 1.8698 (2.0363) [2021-04-16 11:12:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][1040/1251] eta 0:01:01 lr 0.000245 time 0.2857 (0.2907) loss 2.2311 (3.3188) grad_norm 1.7445 (2.0361) [2021-04-16 11:12:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][1050/1251] eta 0:00:58 lr 0.000245 time 0.2843 (0.2906) loss 3.9103 (3.3182) grad_norm 1.8513 (2.0352) [2021-04-16 11:12:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][1060/1251] eta 0:00:55 lr 0.000245 time 0.2899 (0.2904) loss 3.1486 (3.3189) grad_norm 1.8544 (2.0340) [2021-04-16 11:12:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][1070/1251] eta 0:00:52 lr 0.000245 time 0.2780 (0.2903) loss 3.4388 (3.3177) grad_norm 2.3279 (2.0344) [2021-04-16 11:12:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][1080/1251] eta 0:00:49 lr 0.000245 time 0.2903 (0.2902) loss 3.4376 (3.3172) grad_norm 1.7767 (2.0339) [2021-04-16 11:12:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][1090/1251] eta 0:00:46 lr 0.000245 time 0.2721 (0.2901) loss 2.3749 (3.3168) grad_norm 2.0465 (2.0334) [2021-04-16 11:12:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][1100/1251] eta 0:00:43 lr 0.000245 time 0.2765 (0.2899) loss 3.7863 (3.3169) grad_norm 2.0604 (2.0328) [2021-04-16 11:12:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][1110/1251] eta 0:00:40 lr 0.000245 time 0.2873 (0.2899) loss 3.1139 (3.3164) grad_norm 2.1061 (2.0330) [2021-04-16 11:12:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][1120/1251] eta 0:00:37 lr 0.000245 time 0.2890 (0.2898) loss 2.5195 (3.3168) grad_norm 1.6789 (2.0338) [2021-04-16 11:12:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][1130/1251] eta 0:00:35 lr 0.000245 time 0.2702 (0.2897) loss 3.2861 (3.3161) grad_norm 2.3768 (2.0343) [2021-04-16 11:12:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][1140/1251] eta 0:00:32 lr 0.000245 time 0.2656 (0.2897) loss 3.4402 (3.3174) grad_norm 2.7417 (2.0356) [2021-04-16 11:12:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][1150/1251] eta 0:00:29 lr 0.000245 time 0.2713 (0.2898) loss 3.4379 (3.3180) grad_norm 1.9008 (2.0354) [2021-04-16 11:12:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][1160/1251] eta 0:00:26 lr 0.000244 time 0.2684 (0.2897) loss 3.6981 (3.3193) grad_norm 1.7664 (2.0352) [2021-04-16 11:12:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][1170/1251] eta 0:00:23 lr 0.000244 time 0.2631 (0.2895) loss 3.3895 (3.3185) grad_norm 1.8837 (2.0350) [2021-04-16 11:12:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][1180/1251] eta 0:00:20 lr 0.000244 time 0.2921 (0.2895) loss 3.8136 (3.3191) grad_norm 1.8736 (2.0347) [2021-04-16 11:12:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][1190/1251] eta 0:00:17 lr 0.000244 time 0.3105 (0.2894) loss 3.4300 (3.3197) grad_norm 2.1108 (2.0342) [2021-04-16 11:12:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][1200/1251] eta 0:00:14 lr 0.000244 time 0.2830 (0.2893) loss 3.3717 (3.3220) grad_norm 2.0300 (2.0337) [2021-04-16 11:12:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][1210/1251] eta 0:00:11 lr 0.000244 time 0.2954 (0.2893) loss 4.0862 (3.3232) grad_norm 2.3058 (2.0327) [2021-04-16 11:12:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][1220/1251] eta 0:00:08 lr 0.000244 time 0.2793 (0.2891) loss 2.0567 (3.3228) grad_norm 1.8322 (2.0329) [2021-04-16 11:12:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][1230/1251] eta 0:00:06 lr 0.000244 time 0.2694 (0.2893) loss 3.4948 (3.3221) grad_norm 2.0724 (2.0342) [2021-04-16 11:13:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][1240/1251] eta 0:00:03 lr 0.000244 time 0.2609 (0.2892) loss 3.3010 (3.3198) grad_norm 2.0480 (2.0350) [2021-04-16 11:13:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [202/300][1250/1251] eta 0:00:00 lr 0.000244 time 0.2485 (0.2889) loss 3.6842 (3.3185) grad_norm 2.0829 (2.0365) [2021-04-16 11:13:09 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 202 training takes 0:06:07 [2021-04-16 11:13:09 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_202.pth saving...... [2021-04-16 11:13:22 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_202.pth saved !!! [2021-04-16 11:13:23 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.185 (1.185) Loss 0.7970 (0.7970) Acc@1 80.078 (80.078) Acc@5 95.703 (95.703) [2021-04-16 11:13:25 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.253 (0.230) Loss 0.8395 (0.9018) Acc@1 79.395 (78.613) Acc@5 95.508 (94.496) [2021-04-16 11:13:27 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.148 (0.219) Loss 0.9033 (0.9081) Acc@1 79.395 (78.557) Acc@5 94.434 (94.438) [2021-04-16 11:13:29 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.115 (0.223) Loss 0.9118 (0.9100) Acc@1 78.125 (78.484) Acc@5 94.141 (94.415) [2021-04-16 11:13:31 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.225) Loss 0.9379 (0.9124) Acc@1 77.539 (78.418) Acc@5 93.555 (94.415) [2021-04-16 11:13:42 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 78.418 Acc@5 94.426 [2021-04-16 11:13:42 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 78.4% [2021-04-16 11:13:42 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 78.42% [2021-04-16 11:13:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][0/1251] eta 2:05:55 lr 0.000244 time 6.0395 (6.0395) loss 2.3186 (2.3186) grad_norm 1.8175 (1.8175) [2021-04-16 11:13:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][10/1251] eta 0:16:23 lr 0.000244 time 0.2540 (0.7927) loss 3.8268 (3.4023) grad_norm 1.8929 (2.0585) [2021-04-16 11:13:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][20/1251] eta 0:11:15 lr 0.000244 time 0.2727 (0.5485) loss 3.2664 (3.3274) grad_norm 1.7206 (2.0720) [2021-04-16 11:13:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][30/1251] eta 0:09:23 lr 0.000244 time 0.2787 (0.4612) loss 3.0186 (3.3100) grad_norm 2.3155 (2.0854) [2021-04-16 11:13:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][40/1251] eta 0:08:24 lr 0.000244 time 0.2885 (0.4169) loss 2.4411 (3.3231) grad_norm 1.8108 (2.0623) [2021-04-16 11:14:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][50/1251] eta 0:07:47 lr 0.000244 time 0.2525 (0.3894) loss 3.8352 (3.3385) grad_norm 1.9933 (2.0374) [2021-04-16 11:14:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][60/1251] eta 0:07:21 lr 0.000244 time 0.2704 (0.3710) loss 2.7599 (3.3267) grad_norm 2.1409 (2.0305) [2021-04-16 11:14:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][70/1251] eta 0:07:05 lr 0.000244 time 0.2680 (0.3606) loss 3.0562 (3.3228) grad_norm 1.6993 (2.0255) [2021-04-16 11:14:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][80/1251] eta 0:06:49 lr 0.000244 time 0.2842 (0.3499) loss 3.6045 (3.2805) grad_norm 1.8736 (2.0237) [2021-04-16 11:14:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][90/1251] eta 0:06:36 lr 0.000244 time 0.2646 (0.3417) loss 2.9872 (3.2727) grad_norm 1.9644 (2.0088) [2021-04-16 11:14:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][100/1251] eta 0:06:25 lr 0.000244 time 0.2759 (0.3353) loss 3.2993 (3.2731) grad_norm 2.1359 (2.0021) [2021-04-16 11:14:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][110/1251] eta 0:06:17 lr 0.000244 time 0.2780 (0.3305) loss 3.4358 (3.2903) grad_norm 1.8167 (2.0023) [2021-04-16 11:14:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][120/1251] eta 0:06:08 lr 0.000244 time 0.2819 (0.3260) loss 2.7864 (3.2958) grad_norm 1.9284 (2.0021) [2021-04-16 11:14:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][130/1251] eta 0:06:02 lr 0.000244 time 0.2548 (0.3230) loss 3.9481 (3.3008) grad_norm 2.2410 (1.9966) [2021-04-16 11:14:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][140/1251] eta 0:05:56 lr 0.000244 time 0.2835 (0.3208) loss 3.4278 (3.3133) grad_norm 2.3814 (1.9957) [2021-04-16 11:14:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][150/1251] eta 0:05:51 lr 0.000244 time 0.2803 (0.3195) loss 3.2269 (3.3288) grad_norm 1.8018 (1.9993) [2021-04-16 11:14:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][160/1251] eta 0:05:46 lr 0.000244 time 0.2563 (0.3178) loss 2.7945 (3.3201) grad_norm 2.1926 (2.0067) [2021-04-16 11:14:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][170/1251] eta 0:05:40 lr 0.000244 time 0.2549 (0.3152) loss 3.6180 (3.3398) grad_norm 1.7971 (1.9991) [2021-04-16 11:14:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][180/1251] eta 0:05:35 lr 0.000244 time 0.2796 (0.3131) loss 4.0313 (3.3427) grad_norm 2.3240 (1.9999) [2021-04-16 11:14:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][190/1251] eta 0:05:30 lr 0.000243 time 0.2998 (0.3113) loss 2.9446 (3.3406) grad_norm 1.8701 (2.0055) [2021-04-16 11:14:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][200/1251] eta 0:05:25 lr 0.000243 time 0.2751 (0.3097) loss 3.4365 (3.3391) grad_norm 2.0520 (2.0112) [2021-04-16 11:14:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][210/1251] eta 0:05:20 lr 0.000243 time 0.2600 (0.3083) loss 3.6684 (3.3389) grad_norm 2.0516 (2.0115) [2021-04-16 11:14:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][220/1251] eta 0:05:16 lr 0.000243 time 0.2764 (0.3071) loss 2.5689 (3.3396) grad_norm 1.8119 (2.0112) [2021-04-16 11:14:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][230/1251] eta 0:05:13 lr 0.000243 time 0.3160 (0.3070) loss 4.0864 (3.3339) grad_norm 1.9052 (2.0101) [2021-04-16 11:14:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][240/1251] eta 0:05:09 lr 0.000243 time 0.2956 (0.3058) loss 3.8624 (3.3387) grad_norm 1.9949 (2.0154) [2021-04-16 11:14:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][250/1251] eta 0:05:04 lr 0.000243 time 0.2863 (0.3046) loss 3.1400 (3.3401) grad_norm 2.0053 (2.0148) [2021-04-16 11:15:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][260/1251] eta 0:05:00 lr 0.000243 time 0.2759 (0.3036) loss 2.4645 (3.3353) grad_norm 2.0429 (2.0177) [2021-04-16 11:15:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][270/1251] eta 0:04:56 lr 0.000243 time 0.2766 (0.3027) loss 2.9658 (3.3203) grad_norm 1.9828 (2.0187) [2021-04-16 11:15:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][280/1251] eta 0:04:53 lr 0.000243 time 0.2530 (0.3018) loss 3.9716 (3.3198) grad_norm 2.1864 (2.0190) [2021-04-16 11:15:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][290/1251] eta 0:04:49 lr 0.000243 time 0.2853 (0.3012) loss 2.6057 (3.3160) grad_norm 2.0600 (2.0200) [2021-04-16 11:15:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][300/1251] eta 0:04:45 lr 0.000243 time 0.2643 (0.3004) loss 3.2233 (3.3262) grad_norm 2.5944 (2.0224) [2021-04-16 11:15:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][310/1251] eta 0:04:41 lr 0.000243 time 0.2846 (0.2997) loss 3.2552 (3.3205) grad_norm 1.9311 (2.0211) [2021-04-16 11:15:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][320/1251] eta 0:04:38 lr 0.000243 time 0.2796 (0.2988) loss 3.0175 (3.3136) grad_norm 1.8024 (2.0196) [2021-04-16 11:15:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][330/1251] eta 0:04:34 lr 0.000243 time 0.2735 (0.2981) loss 2.9278 (3.3196) grad_norm 2.1548 (2.0177) [2021-04-16 11:15:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][340/1251] eta 0:04:31 lr 0.000243 time 0.2834 (0.2975) loss 3.2501 (3.3179) grad_norm 1.7485 (2.0174) [2021-04-16 11:15:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][350/1251] eta 0:04:27 lr 0.000243 time 0.2709 (0.2969) loss 2.1305 (3.3168) grad_norm 2.0797 (2.0178) [2021-04-16 11:15:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][360/1251] eta 0:04:24 lr 0.000243 time 0.2638 (0.2971) loss 3.0917 (3.3202) grad_norm 1.7547 (2.0129) [2021-04-16 11:15:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][370/1251] eta 0:04:21 lr 0.000243 time 0.2882 (0.2966) loss 3.3606 (3.3255) grad_norm 2.4055 (2.0125) [2021-04-16 11:15:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][380/1251] eta 0:04:17 lr 0.000243 time 0.2787 (0.2960) loss 3.6482 (3.3295) grad_norm 1.8044 (2.0114) [2021-04-16 11:15:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][390/1251] eta 0:04:14 lr 0.000243 time 0.2661 (0.2956) loss 4.1631 (3.3310) grad_norm 1.8996 (2.0112) [2021-04-16 11:15:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][400/1251] eta 0:04:11 lr 0.000243 time 0.2865 (0.2951) loss 3.9766 (3.3388) grad_norm 2.5315 (2.0130) [2021-04-16 11:15:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][410/1251] eta 0:04:07 lr 0.000243 time 0.2607 (0.2946) loss 4.3103 (3.3454) grad_norm 1.9526 (2.0157) [2021-04-16 11:15:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][420/1251] eta 0:04:04 lr 0.000243 time 0.2834 (0.2941) loss 3.9937 (3.3477) grad_norm 2.6479 (2.0190) [2021-04-16 11:15:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][430/1251] eta 0:04:01 lr 0.000243 time 0.2743 (0.2942) loss 3.7287 (3.3471) grad_norm 2.1305 (2.0215) [2021-04-16 11:15:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][440/1251] eta 0:03:58 lr 0.000243 time 0.2971 (0.2941) loss 2.1659 (3.3436) grad_norm 2.1686 (2.0227) [2021-04-16 11:15:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][450/1251] eta 0:03:55 lr 0.000243 time 0.4199 (0.2940) loss 3.5044 (3.3450) grad_norm 1.6597 (2.0230) [2021-04-16 11:15:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][460/1251] eta 0:03:52 lr 0.000243 time 0.2910 (0.2936) loss 2.4911 (3.3419) grad_norm 1.9994 (2.0225) [2021-04-16 11:16:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][470/1251] eta 0:03:48 lr 0.000243 time 0.2788 (0.2931) loss 2.6500 (3.3390) grad_norm 1.8879 (2.0217) [2021-04-16 11:16:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][480/1251] eta 0:03:45 lr 0.000242 time 0.2858 (0.2928) loss 2.3096 (3.3329) grad_norm 2.0830 (2.0207) [2021-04-16 11:16:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][490/1251] eta 0:03:42 lr 0.000242 time 0.2544 (0.2925) loss 3.3946 (3.3328) grad_norm 1.9618 (2.0229) [2021-04-16 11:16:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][500/1251] eta 0:03:39 lr 0.000242 time 0.2856 (0.2922) loss 2.6345 (3.3302) grad_norm 2.2511 (2.0226) [2021-04-16 11:16:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][510/1251] eta 0:03:36 lr 0.000242 time 0.2980 (0.2919) loss 2.5854 (3.3263) grad_norm 2.0426 (2.0231) [2021-04-16 11:16:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][520/1251] eta 0:03:33 lr 0.000242 time 0.2660 (0.2916) loss 3.2605 (3.3238) grad_norm 1.9813 (2.0224) [2021-04-16 11:16:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][530/1251] eta 0:03:29 lr 0.000242 time 0.2889 (0.2912) loss 3.0041 (3.3240) grad_norm 1.8130 (2.0234) [2021-04-16 11:16:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][540/1251] eta 0:03:26 lr 0.000242 time 0.2743 (0.2910) loss 3.2603 (3.3219) grad_norm 1.8407 (2.0248) [2021-04-16 11:16:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][550/1251] eta 0:03:23 lr 0.000242 time 0.2771 (0.2908) loss 3.0773 (3.3193) grad_norm 1.9841 (2.0250) [2021-04-16 11:16:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][560/1251] eta 0:03:20 lr 0.000242 time 0.2602 (0.2904) loss 3.7106 (3.3208) grad_norm 1.9618 (2.0250) [2021-04-16 11:16:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][570/1251] eta 0:03:17 lr 0.000242 time 0.2861 (0.2907) loss 2.2152 (3.3162) grad_norm 2.3979 (2.0287) [2021-04-16 11:16:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][580/1251] eta 0:03:15 lr 0.000242 time 0.2832 (0.2906) loss 3.4663 (3.3162) grad_norm 2.3184 (2.0320) [2021-04-16 11:16:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][590/1251] eta 0:03:11 lr 0.000242 time 0.2743 (0.2905) loss 3.0586 (3.3154) grad_norm 1.8302 (2.0324) [2021-04-16 11:16:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][600/1251] eta 0:03:08 lr 0.000242 time 0.2722 (0.2902) loss 3.6717 (3.3123) grad_norm 2.2180 (2.0330) [2021-04-16 11:16:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][610/1251] eta 0:03:05 lr 0.000242 time 0.2888 (0.2900) loss 3.7475 (3.3086) grad_norm 2.1353 (2.0323) [2021-04-16 11:16:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][620/1251] eta 0:03:02 lr 0.000242 time 0.2761 (0.2898) loss 3.7052 (3.3079) grad_norm 2.1068 (2.0330) [2021-04-16 11:16:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][630/1251] eta 0:02:59 lr 0.000242 time 0.2881 (0.2897) loss 2.5772 (3.3081) grad_norm 1.8455 (2.0311) [2021-04-16 11:16:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][640/1251] eta 0:02:56 lr 0.000242 time 0.2521 (0.2896) loss 3.9495 (3.3111) grad_norm 2.0642 (2.0329) [2021-04-16 11:16:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][650/1251] eta 0:02:53 lr 0.000242 time 0.2766 (0.2894) loss 3.3296 (3.3123) grad_norm 1.9318 (2.0334) [2021-04-16 11:16:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][660/1251] eta 0:02:51 lr 0.000242 time 0.3005 (0.2894) loss 3.2510 (3.3161) grad_norm 1.8046 (2.0330) [2021-04-16 11:16:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][670/1251] eta 0:02:48 lr 0.000242 time 0.2838 (0.2892) loss 3.3386 (3.3179) grad_norm 2.1003 (2.0333) [2021-04-16 11:16:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][680/1251] eta 0:02:45 lr 0.000242 time 0.2740 (0.2891) loss 2.5633 (3.3173) grad_norm 1.7525 (2.0345) [2021-04-16 11:17:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][690/1251] eta 0:02:42 lr 0.000242 time 0.2778 (0.2888) loss 3.1853 (3.3204) grad_norm 2.0392 (2.0370) [2021-04-16 11:17:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][700/1251] eta 0:02:39 lr 0.000242 time 0.2683 (0.2887) loss 2.6293 (3.3214) grad_norm 1.8249 (2.0364) [2021-04-16 11:17:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][710/1251] eta 0:02:36 lr 0.000242 time 0.2903 (0.2887) loss 3.1973 (3.3240) grad_norm 3.8607 (2.0399) [2021-04-16 11:17:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][720/1251] eta 0:02:33 lr 0.000242 time 0.3049 (0.2888) loss 3.2087 (3.3234) grad_norm 1.8182 (2.0404) [2021-04-16 11:17:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][730/1251] eta 0:02:30 lr 0.000242 time 0.4320 (0.2891) loss 3.7909 (3.3221) grad_norm 2.3022 (2.0409) [2021-04-16 11:17:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][740/1251] eta 0:02:27 lr 0.000242 time 0.3105 (0.2889) loss 3.7390 (3.3226) grad_norm 2.0300 (2.0426) [2021-04-16 11:17:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][750/1251] eta 0:02:24 lr 0.000242 time 0.2905 (0.2887) loss 3.5034 (3.3247) grad_norm 2.4059 (2.0438) [2021-04-16 11:17:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][760/1251] eta 0:02:21 lr 0.000241 time 0.2795 (0.2887) loss 3.6253 (3.3232) grad_norm 1.9841 (2.0441) [2021-04-16 11:17:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][770/1251] eta 0:02:18 lr 0.000241 time 0.2713 (0.2885) loss 3.2268 (3.3238) grad_norm 1.9692 (2.0450) [2021-04-16 11:17:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][780/1251] eta 0:02:15 lr 0.000241 time 0.2832 (0.2884) loss 3.0228 (3.3249) grad_norm 1.8629 (2.0437) [2021-04-16 11:17:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][790/1251] eta 0:02:12 lr 0.000241 time 0.2716 (0.2882) loss 3.0164 (3.3266) grad_norm 2.1855 (2.0420) [2021-04-16 11:17:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][800/1251] eta 0:02:10 lr 0.000241 time 0.4293 (0.2883) loss 4.0309 (3.3270) grad_norm 1.7891 (2.0421) [2021-04-16 11:17:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][810/1251] eta 0:02:07 lr 0.000241 time 0.3144 (0.2881) loss 3.3725 (3.3274) grad_norm 1.9460 (2.0412) [2021-04-16 11:17:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][820/1251] eta 0:02:04 lr 0.000241 time 0.2899 (0.2880) loss 2.2333 (3.3258) grad_norm 2.0178 (2.0403) [2021-04-16 11:17:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][830/1251] eta 0:02:01 lr 0.000241 time 0.2532 (0.2878) loss 3.8263 (3.3251) grad_norm 1.8395 (2.0401) [2021-04-16 11:17:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][840/1251] eta 0:01:58 lr 0.000241 time 0.3100 (0.2877) loss 3.0326 (3.3266) grad_norm 1.7925 (2.0404) [2021-04-16 11:17:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][850/1251] eta 0:01:55 lr 0.000241 time 0.2818 (0.2876) loss 4.3154 (3.3252) grad_norm 2.0318 (2.0406) [2021-04-16 11:17:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][860/1251] eta 0:01:52 lr 0.000241 time 0.2868 (0.2875) loss 3.6029 (3.3246) grad_norm 1.8853 (2.0403) [2021-04-16 11:17:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][870/1251] eta 0:01:49 lr 0.000241 time 0.2842 (0.2873) loss 2.7652 (3.3263) grad_norm 2.0082 (2.0405) [2021-04-16 11:17:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][880/1251] eta 0:01:46 lr 0.000241 time 0.2702 (0.2871) loss 4.0301 (3.3245) grad_norm 2.0958 (2.0410) [2021-04-16 11:17:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][890/1251] eta 0:01:43 lr 0.000241 time 0.2919 (0.2870) loss 3.1105 (3.3242) grad_norm 2.0463 (2.0409) [2021-04-16 11:18:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][900/1251] eta 0:01:40 lr 0.000241 time 0.2612 (0.2869) loss 3.1966 (3.3261) grad_norm 1.8450 (2.0410) [2021-04-16 11:18:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][910/1251] eta 0:01:37 lr 0.000241 time 0.2655 (0.2868) loss 3.6213 (3.3276) grad_norm 2.1919 (2.0405) [2021-04-16 11:18:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][920/1251] eta 0:01:34 lr 0.000241 time 0.2735 (0.2867) loss 3.6191 (3.3265) grad_norm 1.8676 (2.0417) [2021-04-16 11:18:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][930/1251] eta 0:01:32 lr 0.000241 time 0.2793 (0.2866) loss 3.9802 (3.3259) grad_norm 1.8886 (2.0400) [2021-04-16 11:18:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][940/1251] eta 0:01:29 lr 0.000241 time 0.2975 (0.2868) loss 3.9434 (3.3256) grad_norm 2.3535 (2.0415) [2021-04-16 11:18:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][950/1251] eta 0:01:26 lr 0.000241 time 0.2771 (0.2867) loss 3.7103 (3.3230) grad_norm 1.8991 (2.0424) [2021-04-16 11:18:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][960/1251] eta 0:01:23 lr 0.000241 time 0.2917 (0.2866) loss 3.6698 (3.3228) grad_norm 2.3401 (2.0443) [2021-04-16 11:18:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][970/1251] eta 0:01:20 lr 0.000241 time 0.3086 (0.2865) loss 2.8574 (3.3206) grad_norm 2.1800 (2.0444) [2021-04-16 11:18:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][980/1251] eta 0:01:17 lr 0.000241 time 0.2916 (0.2864) loss 3.5187 (3.3173) grad_norm 2.3783 (2.0451) [2021-04-16 11:18:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][990/1251] eta 0:01:14 lr 0.000241 time 0.3014 (0.2863) loss 3.0512 (3.3164) grad_norm 2.2186 (2.0450) [2021-04-16 11:18:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][1000/1251] eta 0:01:11 lr 0.000241 time 0.2688 (0.2862) loss 2.7778 (3.3182) grad_norm 1.7246 (2.0437) [2021-04-16 11:18:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][1010/1251] eta 0:01:08 lr 0.000241 time 0.2861 (0.2862) loss 3.6367 (3.3184) grad_norm 2.0320 (2.0433) [2021-04-16 11:18:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][1020/1251] eta 0:01:06 lr 0.000241 time 0.3184 (0.2862) loss 3.7053 (3.3156) grad_norm 2.0549 (2.0428) [2021-04-16 11:18:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][1030/1251] eta 0:01:03 lr 0.000241 time 0.2728 (0.2863) loss 2.3155 (3.3154) grad_norm 1.7646 (2.0417) [2021-04-16 11:18:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][1040/1251] eta 0:01:00 lr 0.000241 time 0.2728 (0.2862) loss 3.4967 (3.3124) grad_norm 2.0542 (2.0411) [2021-04-16 11:18:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][1050/1251] eta 0:00:57 lr 0.000240 time 0.2869 (0.2862) loss 3.4993 (3.3117) grad_norm 2.0411 (2.0400) [2021-04-16 11:18:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][1060/1251] eta 0:00:54 lr 0.000240 time 0.2553 (0.2861) loss 2.7575 (3.3090) grad_norm 1.7594 (2.0396) [2021-04-16 11:18:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][1070/1251] eta 0:00:51 lr 0.000240 time 0.2719 (0.2861) loss 4.0645 (3.3099) grad_norm 1.8192 (2.0388) [2021-04-16 11:18:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][1080/1251] eta 0:00:48 lr 0.000240 time 0.2641 (0.2859) loss 3.7122 (3.3107) grad_norm 2.3801 (2.0384) [2021-04-16 11:18:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][1090/1251] eta 0:00:46 lr 0.000240 time 0.2693 (0.2859) loss 2.9556 (3.3090) grad_norm 2.0664 (2.0390) [2021-04-16 11:18:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][1100/1251] eta 0:00:43 lr 0.000240 time 0.2739 (0.2859) loss 2.3247 (3.3100) grad_norm 1.7277 (2.0384) [2021-04-16 11:19:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][1110/1251] eta 0:00:40 lr 0.000240 time 0.2871 (0.2859) loss 2.4975 (3.3081) grad_norm 2.0911 (2.0370) [2021-04-16 11:19:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][1120/1251] eta 0:00:37 lr 0.000240 time 0.2745 (0.2858) loss 4.4568 (3.3121) grad_norm 2.0030 (2.0377) [2021-04-16 11:19:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][1130/1251] eta 0:00:34 lr 0.000240 time 0.2751 (0.2857) loss 3.5959 (3.3129) grad_norm 1.7940 (2.0388) [2021-04-16 11:19:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][1140/1251] eta 0:00:31 lr 0.000240 time 0.2817 (0.2857) loss 2.9471 (3.3129) grad_norm 1.9611 (2.0402) [2021-04-16 11:19:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][1150/1251] eta 0:00:28 lr 0.000240 time 0.2897 (0.2857) loss 3.7039 (3.3143) grad_norm 2.5514 (2.0404) [2021-04-16 11:19:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][1160/1251] eta 0:00:26 lr 0.000240 time 0.2896 (0.2858) loss 4.0542 (3.3151) grad_norm 1.9207 (2.0407) [2021-04-16 11:19:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][1170/1251] eta 0:00:23 lr 0.000240 time 0.2830 (0.2857) loss 2.6269 (3.3145) grad_norm 2.1540 (2.0411) [2021-04-16 11:19:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][1180/1251] eta 0:00:20 lr 0.000240 time 0.2651 (0.2857) loss 3.0795 (3.3158) grad_norm 2.0900 (2.0416) [2021-04-16 11:19:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][1190/1251] eta 0:00:17 lr 0.000240 time 0.2685 (0.2856) loss 2.9980 (3.3157) grad_norm 2.0353 (2.0414) [2021-04-16 11:19:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][1200/1251] eta 0:00:14 lr 0.000240 time 0.2721 (0.2855) loss 3.6750 (3.3160) grad_norm 2.5280 (2.0416) [2021-04-16 11:19:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][1210/1251] eta 0:00:11 lr 0.000240 time 0.2762 (0.2854) loss 3.8473 (3.3169) grad_norm 2.4736 (2.0413) [2021-04-16 11:19:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][1220/1251] eta 0:00:08 lr 0.000240 time 0.2869 (0.2854) loss 3.5360 (3.3184) grad_norm 2.1950 (2.0422) [2021-04-16 11:19:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][1230/1251] eta 0:00:05 lr 0.000240 time 0.2531 (0.2854) loss 2.7394 (3.3178) grad_norm 2.2563 (2.0419) [2021-04-16 11:19:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][1240/1251] eta 0:00:03 lr 0.000240 time 0.2490 (0.2852) loss 4.0139 (3.3198) grad_norm 1.8865 (2.0413) [2021-04-16 11:19:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [203/300][1250/1251] eta 0:00:00 lr 0.000240 time 0.2476 (0.2849) loss 3.4437 (3.3200) grad_norm 2.3963 (2.0411) [2021-04-16 11:19:41 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 203 training takes 0:05:59 [2021-04-16 11:19:41 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_203.pth saving...... [2021-04-16 11:20:06 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_203.pth saved !!! [2021-04-16 11:20:07 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.268 (1.268) Loss 0.9823 (0.9823) Acc@1 76.367 (76.367) Acc@5 93.848 (93.848) [2021-04-16 11:20:09 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.120 (0.273) Loss 0.9034 (0.9198) Acc@1 77.539 (78.382) Acc@5 94.824 (94.221) [2021-04-16 11:20:10 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.120 (0.205) Loss 0.9415 (0.9164) Acc@1 78.223 (78.283) Acc@5 93.750 (94.169) [2021-04-16 11:20:13 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.282 (0.232) Loss 0.8612 (0.9090) Acc@1 80.762 (78.364) Acc@5 94.629 (94.405) [2021-04-16 11:20:15 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.096 (0.218) Loss 0.9165 (0.9051) Acc@1 77.930 (78.494) Acc@5 95.020 (94.526) [2021-04-16 11:20:21 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 78.502 Acc@5 94.538 [2021-04-16 11:20:21 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 78.5% [2021-04-16 11:20:21 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 78.50% [2021-04-16 11:20:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][0/1251] eta 3:47:44 lr 0.000240 time 10.9228 (10.9228) loss 3.0286 (3.0286) grad_norm 1.8497 (1.8497) [2021-04-16 11:20:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][10/1251] eta 0:25:58 lr 0.000240 time 0.4341 (1.2561) loss 3.8115 (3.4875) grad_norm 1.9516 (2.0824) [2021-04-16 11:20:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][20/1251] eta 0:16:13 lr 0.000240 time 0.2800 (0.7906) loss 2.7753 (3.4688) grad_norm 2.0723 (2.1379) [2021-04-16 11:20:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][30/1251] eta 0:12:43 lr 0.000240 time 0.2710 (0.6252) loss 3.5692 (3.2581) grad_norm 2.4171 (2.1701) [2021-04-16 11:20:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][40/1251] eta 0:10:56 lr 0.000240 time 0.2765 (0.5421) loss 3.4032 (3.1871) grad_norm 2.0552 (2.1672) [2021-04-16 11:20:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][50/1251] eta 0:09:49 lr 0.000240 time 0.2695 (0.4909) loss 3.4336 (3.1968) grad_norm 1.9908 (2.1634) [2021-04-16 11:20:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][60/1251] eta 0:09:02 lr 0.000240 time 0.2870 (0.4558) loss 2.2494 (3.1864) grad_norm 1.8091 (2.1331) [2021-04-16 11:20:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][70/1251] eta 0:08:29 lr 0.000240 time 0.3018 (0.4314) loss 2.5890 (3.1784) grad_norm 1.8395 (2.1351) [2021-04-16 11:20:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][80/1251] eta 0:08:03 lr 0.000239 time 0.2603 (0.4129) loss 3.0346 (3.1755) grad_norm 2.0602 (2.1156) [2021-04-16 11:20:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][90/1251] eta 0:07:42 lr 0.000239 time 0.3111 (0.3986) loss 2.4870 (3.1997) grad_norm 2.0189 (2.1076) [2021-04-16 11:21:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][100/1251] eta 0:07:25 lr 0.000239 time 0.2950 (0.3872) loss 3.5727 (3.2178) grad_norm 2.1832 (2.1099) [2021-04-16 11:21:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][110/1251] eta 0:07:11 lr 0.000239 time 0.2689 (0.3781) loss 2.1908 (3.2065) grad_norm 2.0564 (2.1085) [2021-04-16 11:21:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][120/1251] eta 0:06:58 lr 0.000239 time 0.2689 (0.3704) loss 3.6699 (3.2168) grad_norm 1.9631 (2.1016) [2021-04-16 11:21:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][130/1251] eta 0:06:47 lr 0.000239 time 0.2548 (0.3636) loss 3.3047 (3.2241) grad_norm 2.0154 (2.0943) [2021-04-16 11:21:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][140/1251] eta 0:06:37 lr 0.000239 time 0.3189 (0.3577) loss 3.6410 (3.2508) grad_norm 2.2013 (2.0962) [2021-04-16 11:21:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][150/1251] eta 0:06:30 lr 0.000239 time 0.2900 (0.3550) loss 3.4706 (3.2702) grad_norm 1.9441 (2.0988) [2021-04-16 11:21:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][160/1251] eta 0:06:22 lr 0.000239 time 0.2631 (0.3510) loss 3.6364 (3.2802) grad_norm 2.2424 (2.0950) [2021-04-16 11:21:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][170/1251] eta 0:06:15 lr 0.000239 time 0.2715 (0.3469) loss 2.2234 (3.2805) grad_norm 2.0323 (2.0882) [2021-04-16 11:21:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][180/1251] eta 0:06:08 lr 0.000239 time 0.2771 (0.3437) loss 3.1183 (3.2808) grad_norm 2.0971 (2.0881) [2021-04-16 11:21:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][190/1251] eta 0:06:01 lr 0.000239 time 0.2817 (0.3406) loss 3.5424 (3.2879) grad_norm 1.8982 (2.0814) [2021-04-16 11:21:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][200/1251] eta 0:05:54 lr 0.000239 time 0.2828 (0.3373) loss 2.6204 (3.3007) grad_norm 1.8147 (2.0762) [2021-04-16 11:21:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][210/1251] eta 0:05:48 lr 0.000239 time 0.3053 (0.3347) loss 3.3693 (3.3033) grad_norm 1.7907 (2.0727) [2021-04-16 11:21:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][220/1251] eta 0:05:43 lr 0.000239 time 0.2819 (0.3329) loss 3.7260 (3.3095) grad_norm 2.0152 (2.0716) [2021-04-16 11:21:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][230/1251] eta 0:05:38 lr 0.000239 time 0.2882 (0.3312) loss 2.9728 (3.3102) grad_norm 2.2863 (2.0751) [2021-04-16 11:21:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][240/1251] eta 0:05:32 lr 0.000239 time 0.2851 (0.3290) loss 3.8308 (3.3095) grad_norm 2.1247 (2.0788) [2021-04-16 11:21:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][250/1251] eta 0:05:27 lr 0.000239 time 0.2937 (0.3277) loss 3.6177 (3.3128) grad_norm 2.0141 (2.0772) [2021-04-16 11:21:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][260/1251] eta 0:05:23 lr 0.000239 time 0.2669 (0.3259) loss 4.2252 (3.3226) grad_norm 2.3231 (2.0785) [2021-04-16 11:21:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][270/1251] eta 0:05:17 lr 0.000239 time 0.2608 (0.3241) loss 2.7544 (3.3198) grad_norm 1.8015 (2.0776) [2021-04-16 11:21:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][280/1251] eta 0:05:13 lr 0.000239 time 0.2863 (0.3224) loss 3.3458 (3.3140) grad_norm 2.3771 (2.0777) [2021-04-16 11:21:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][290/1251] eta 0:05:08 lr 0.000239 time 0.2791 (0.3209) loss 2.4769 (3.3092) grad_norm 1.8761 (2.0761) [2021-04-16 11:21:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][300/1251] eta 0:05:03 lr 0.000239 time 0.3003 (0.3196) loss 3.5920 (3.3022) grad_norm 2.1889 (2.0752) [2021-04-16 11:22:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][310/1251] eta 0:04:59 lr 0.000239 time 0.2882 (0.3184) loss 3.8881 (3.2990) grad_norm 1.7230 (2.0715) [2021-04-16 11:22:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][320/1251] eta 0:04:55 lr 0.000239 time 0.2678 (0.3174) loss 3.1902 (3.2968) grad_norm 1.9357 (2.0689) [2021-04-16 11:22:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][330/1251] eta 0:04:51 lr 0.000239 time 0.2911 (0.3164) loss 3.9604 (3.2963) grad_norm 2.2014 (2.0716) [2021-04-16 11:22:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][340/1251] eta 0:04:47 lr 0.000239 time 0.2836 (0.3152) loss 3.6964 (3.2953) grad_norm 2.2505 (2.0730) [2021-04-16 11:22:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][350/1251] eta 0:04:43 lr 0.000239 time 0.2877 (0.3147) loss 2.4278 (3.2911) grad_norm 2.1485 (2.0701) [2021-04-16 11:22:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][360/1251] eta 0:04:39 lr 0.000239 time 0.2687 (0.3142) loss 3.4691 (3.2898) grad_norm 1.8273 (2.0680) [2021-04-16 11:22:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][370/1251] eta 0:04:36 lr 0.000238 time 0.3068 (0.3134) loss 3.2180 (3.2856) grad_norm 1.9507 (2.0733) [2021-04-16 11:22:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][380/1251] eta 0:04:32 lr 0.000238 time 0.2866 (0.3124) loss 3.4175 (3.2795) grad_norm 1.9145 (2.0702) [2021-04-16 11:22:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][390/1251] eta 0:04:28 lr 0.000238 time 0.2819 (0.3115) loss 3.3802 (3.2849) grad_norm 2.4332 (2.0702) [2021-04-16 11:22:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][400/1251] eta 0:04:24 lr 0.000238 time 0.2725 (0.3110) loss 3.5870 (3.2828) grad_norm 1.8302 (2.0684) [2021-04-16 11:22:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][410/1251] eta 0:04:20 lr 0.000238 time 0.2677 (0.3103) loss 3.4044 (3.2755) grad_norm 3.8394 (2.0737) [2021-04-16 11:22:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][420/1251] eta 0:04:17 lr 0.000238 time 0.2633 (0.3096) loss 3.0932 (3.2797) grad_norm 1.9965 (2.0717) [2021-04-16 11:22:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][430/1251] eta 0:04:13 lr 0.000238 time 0.2857 (0.3090) loss 3.0252 (3.2757) grad_norm 2.2878 (2.0719) [2021-04-16 11:22:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][440/1251] eta 0:04:10 lr 0.000238 time 0.2872 (0.3083) loss 2.6871 (3.2730) grad_norm 2.1171 (2.0694) [2021-04-16 11:22:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][450/1251] eta 0:04:06 lr 0.000238 time 0.2761 (0.3076) loss 2.8592 (3.2699) grad_norm 2.1409 (2.0690) [2021-04-16 11:22:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][460/1251] eta 0:04:02 lr 0.000238 time 0.2812 (0.3071) loss 3.5341 (3.2675) grad_norm 2.1723 (2.0690) [2021-04-16 11:22:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][470/1251] eta 0:03:59 lr 0.000238 time 0.2687 (0.3066) loss 4.0718 (3.2707) grad_norm 2.4442 (2.0715) [2021-04-16 11:22:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][480/1251] eta 0:03:56 lr 0.000238 time 0.3006 (0.3062) loss 2.4876 (3.2716) grad_norm 2.1580 (2.0763) [2021-04-16 11:22:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][490/1251] eta 0:03:52 lr 0.000238 time 0.2955 (0.3058) loss 3.9095 (3.2743) grad_norm 2.1501 (2.0736) [2021-04-16 11:22:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][500/1251] eta 0:03:49 lr 0.000238 time 0.2859 (0.3054) loss 3.7972 (3.2730) grad_norm 2.0414 (2.0743) [2021-04-16 11:22:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][510/1251] eta 0:03:45 lr 0.000238 time 0.2737 (0.3049) loss 2.7808 (3.2739) grad_norm 2.0793 (2.0738) [2021-04-16 11:23:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][520/1251] eta 0:03:42 lr 0.000238 time 0.2753 (0.3044) loss 2.4086 (3.2735) grad_norm 1.9855 (2.0745) [2021-04-16 11:23:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][530/1251] eta 0:03:39 lr 0.000238 time 0.2842 (0.3040) loss 3.2907 (3.2738) grad_norm 1.8966 (2.0751) [2021-04-16 11:23:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][540/1251] eta 0:03:36 lr 0.000238 time 0.2578 (0.3039) loss 2.6136 (3.2739) grad_norm 1.9937 (2.0754) [2021-04-16 11:23:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][550/1251] eta 0:03:32 lr 0.000238 time 0.2627 (0.3034) loss 3.4424 (3.2716) grad_norm 1.9723 (2.0753) [2021-04-16 11:23:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][560/1251] eta 0:03:29 lr 0.000238 time 0.2893 (0.3031) loss 3.1190 (3.2704) grad_norm 2.7011 (2.0751) [2021-04-16 11:23:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][570/1251] eta 0:03:26 lr 0.000238 time 0.2710 (0.3027) loss 3.0460 (3.2661) grad_norm 2.0461 (2.0754) [2021-04-16 11:23:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][580/1251] eta 0:03:23 lr 0.000238 time 0.2689 (0.3030) loss 3.3413 (3.2639) grad_norm 2.0212 (2.0746) [2021-04-16 11:23:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][590/1251] eta 0:03:20 lr 0.000238 time 0.2930 (0.3030) loss 3.0889 (3.2626) grad_norm 1.9983 (2.0745) [2021-04-16 11:23:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][600/1251] eta 0:03:16 lr 0.000238 time 0.2605 (0.3026) loss 3.6952 (3.2677) grad_norm 2.0655 (2.0741) [2021-04-16 11:23:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][610/1251] eta 0:03:13 lr 0.000238 time 0.2718 (0.3022) loss 3.9799 (3.2646) grad_norm 1.8672 (2.0724) [2021-04-16 11:23:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][620/1251] eta 0:03:10 lr 0.000238 time 0.2926 (0.3018) loss 2.3322 (3.2662) grad_norm 2.1630 (2.0738) [2021-04-16 11:23:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][630/1251] eta 0:03:07 lr 0.000238 time 0.2552 (0.3014) loss 3.5483 (3.2640) grad_norm 2.9510 (2.0743) [2021-04-16 11:23:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][640/1251] eta 0:03:03 lr 0.000238 time 0.2792 (0.3011) loss 3.4761 (3.2598) grad_norm 1.9845 (2.0748) [2021-04-16 11:23:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][650/1251] eta 0:03:00 lr 0.000237 time 0.2704 (0.3008) loss 3.5178 (3.2635) grad_norm 1.7976 (2.0746) [2021-04-16 11:23:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][660/1251] eta 0:02:57 lr 0.000237 time 0.2524 (0.3005) loss 2.8513 (3.2629) grad_norm 2.1852 (2.0756) [2021-04-16 11:23:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][670/1251] eta 0:02:54 lr 0.000237 time 0.2931 (0.3002) loss 2.6324 (3.2624) grad_norm 1.9869 (2.0781) [2021-04-16 11:23:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][680/1251] eta 0:02:51 lr 0.000237 time 0.2636 (0.2999) loss 4.0402 (3.2624) grad_norm 1.9125 (2.0781) [2021-04-16 11:23:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][690/1251] eta 0:02:48 lr 0.000237 time 0.2739 (0.2995) loss 3.9915 (3.2642) grad_norm 2.0698 (2.0790) [2021-04-16 11:23:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][700/1251] eta 0:02:44 lr 0.000237 time 0.2667 (0.2992) loss 3.3629 (3.2670) grad_norm 2.1007 (2.0786) [2021-04-16 11:23:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][710/1251] eta 0:02:41 lr 0.000237 time 0.2578 (0.2991) loss 2.6352 (3.2684) grad_norm 2.1029 (2.0814) [2021-04-16 11:23:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][720/1251] eta 0:02:38 lr 0.000237 time 0.2896 (0.2991) loss 2.8612 (3.2703) grad_norm 1.7647 (2.0801) [2021-04-16 11:24:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][730/1251] eta 0:02:35 lr 0.000237 time 0.3912 (0.2992) loss 3.5555 (3.2677) grad_norm 1.8881 (2.0800) [2021-04-16 11:24:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][740/1251] eta 0:02:32 lr 0.000237 time 0.2869 (0.2989) loss 2.2890 (3.2644) grad_norm 1.9271 (2.0789) [2021-04-16 11:24:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][750/1251] eta 0:02:29 lr 0.000237 time 0.2858 (0.2988) loss 3.7213 (3.2653) grad_norm 2.4982 (2.0805) [2021-04-16 11:24:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][760/1251] eta 0:02:26 lr 0.000237 time 0.2703 (0.2985) loss 3.3412 (3.2657) grad_norm 1.8727 (2.0820) [2021-04-16 11:24:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][770/1251] eta 0:02:23 lr 0.000237 time 0.2680 (0.2983) loss 3.4964 (3.2676) grad_norm 2.0234 (2.0824) [2021-04-16 11:24:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][780/1251] eta 0:02:20 lr 0.000237 time 0.2726 (0.2981) loss 2.8254 (3.2685) grad_norm 1.8923 (2.0805) [2021-04-16 11:24:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][790/1251] eta 0:02:17 lr 0.000237 time 0.2576 (0.2980) loss 3.2247 (3.2694) grad_norm 2.2686 (2.0806) [2021-04-16 11:24:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][800/1251] eta 0:02:14 lr 0.000237 time 0.2915 (0.2977) loss 3.5805 (3.2718) grad_norm 1.6969 (2.0801) [2021-04-16 11:24:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][810/1251] eta 0:02:11 lr 0.000237 time 0.2886 (0.2977) loss 3.6849 (3.2681) grad_norm 1.8908 (2.0785) [2021-04-16 11:24:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][820/1251] eta 0:02:08 lr 0.000237 time 0.2867 (0.2975) loss 2.4687 (3.2682) grad_norm 1.8661 (2.0786) [2021-04-16 11:24:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][830/1251] eta 0:02:05 lr 0.000237 time 0.2645 (0.2974) loss 3.6751 (3.2703) grad_norm 2.3232 (2.0788) [2021-04-16 11:24:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][840/1251] eta 0:02:02 lr 0.000237 time 0.2859 (0.2973) loss 3.4015 (3.2698) grad_norm 2.0037 (2.0801) [2021-04-16 11:24:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][850/1251] eta 0:01:59 lr 0.000237 time 0.2810 (0.2972) loss 2.5330 (3.2694) grad_norm 1.8699 (2.0800) [2021-04-16 11:24:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][860/1251] eta 0:01:56 lr 0.000237 time 0.2594 (0.2970) loss 3.2189 (3.2679) grad_norm 1.7960 (2.0783) [2021-04-16 11:24:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][870/1251] eta 0:01:53 lr 0.000237 time 0.2593 (0.2967) loss 2.9168 (3.2704) grad_norm 1.8687 (2.0786) [2021-04-16 11:24:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][880/1251] eta 0:01:50 lr 0.000237 time 0.2630 (0.2965) loss 2.8570 (3.2700) grad_norm 1.7688 (2.0782) [2021-04-16 11:24:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][890/1251] eta 0:01:46 lr 0.000237 time 0.2931 (0.2964) loss 3.1821 (3.2703) grad_norm 2.1118 (2.0787) [2021-04-16 11:24:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][900/1251] eta 0:01:43 lr 0.000237 time 0.3138 (0.2962) loss 3.2008 (3.2703) grad_norm 2.0787 (2.0798) [2021-04-16 11:24:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][910/1251] eta 0:01:40 lr 0.000237 time 0.2661 (0.2960) loss 2.1514 (3.2689) grad_norm 1.6246 (2.0796) [2021-04-16 11:24:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][920/1251] eta 0:01:37 lr 0.000237 time 0.2749 (0.2960) loss 3.7639 (3.2696) grad_norm 1.9091 (2.0801) [2021-04-16 11:24:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][930/1251] eta 0:01:35 lr 0.000237 time 0.2597 (0.2960) loss 3.9292 (3.2728) grad_norm 1.8467 (2.0792) [2021-04-16 11:25:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][940/1251] eta 0:01:32 lr 0.000236 time 0.2515 (0.2959) loss 3.3452 (3.2755) grad_norm 2.0334 (2.0810) [2021-04-16 11:25:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][950/1251] eta 0:01:29 lr 0.000236 time 0.3054 (0.2958) loss 2.3358 (3.2734) grad_norm 2.0986 (2.0810) [2021-04-16 11:25:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][960/1251] eta 0:01:26 lr 0.000236 time 0.2851 (0.2956) loss 3.7226 (3.2736) grad_norm 2.3771 (2.0816) [2021-04-16 11:25:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][970/1251] eta 0:01:23 lr 0.000236 time 0.2824 (0.2955) loss 3.5485 (3.2741) grad_norm 2.0251 (2.0814) [2021-04-16 11:25:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][980/1251] eta 0:01:20 lr 0.000236 time 0.2804 (0.2954) loss 4.3719 (3.2752) grad_norm 2.8099 (2.0815) [2021-04-16 11:25:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][990/1251] eta 0:01:17 lr 0.000236 time 0.2716 (0.2952) loss 3.7548 (3.2762) grad_norm 2.4928 (2.0824) [2021-04-16 11:25:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][1000/1251] eta 0:01:14 lr 0.000236 time 0.2783 (0.2951) loss 2.9089 (3.2744) grad_norm 1.9257 (2.0819) [2021-04-16 11:25:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][1010/1251] eta 0:01:11 lr 0.000236 time 0.2708 (0.2950) loss 2.3794 (3.2738) grad_norm 2.3751 (2.0823) [2021-04-16 11:25:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][1020/1251] eta 0:01:08 lr 0.000236 time 0.2772 (0.2948) loss 3.3267 (3.2742) grad_norm 2.0869 (2.0825) [2021-04-16 11:25:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][1030/1251] eta 0:01:05 lr 0.000236 time 0.2718 (0.2946) loss 3.7827 (3.2743) grad_norm 1.9677 (2.0824) [2021-04-16 11:25:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][1040/1251] eta 0:01:02 lr 0.000236 time 0.2941 (0.2945) loss 3.7179 (3.2702) grad_norm 1.8779 (2.0815) [2021-04-16 11:25:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][1050/1251] eta 0:00:59 lr 0.000236 time 0.2964 (0.2944) loss 2.8098 (3.2704) grad_norm 2.1654 (2.0811) [2021-04-16 11:25:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][1060/1251] eta 0:00:56 lr 0.000236 time 0.2815 (0.2942) loss 3.7455 (3.2696) grad_norm 1.7859 (2.0813) [2021-04-16 11:25:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][1070/1251] eta 0:00:53 lr 0.000236 time 0.2847 (0.2941) loss 2.2622 (3.2711) grad_norm 2.1488 (2.0809) [2021-04-16 11:25:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][1080/1251] eta 0:00:50 lr 0.000236 time 0.2875 (0.2940) loss 3.3625 (3.2708) grad_norm 2.2309 (2.0810) [2021-04-16 11:25:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][1090/1251] eta 0:00:47 lr 0.000236 time 0.2869 (0.2939) loss 2.4826 (3.2683) grad_norm 1.9133 (2.0801) [2021-04-16 11:25:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][1100/1251] eta 0:00:44 lr 0.000236 time 0.2814 (0.2937) loss 3.4825 (3.2701) grad_norm 2.2548 (2.0797) [2021-04-16 11:25:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][1110/1251] eta 0:00:41 lr 0.000236 time 0.2872 (0.2936) loss 3.8928 (3.2692) grad_norm 1.8751 (2.0792) [2021-04-16 11:25:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][1120/1251] eta 0:00:38 lr 0.000236 time 0.2843 (0.2937) loss 2.5025 (3.2703) grad_norm 2.2589 (2.0790) [2021-04-16 11:25:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][1130/1251] eta 0:00:35 lr 0.000236 time 0.2672 (0.2935) loss 2.6580 (3.2706) grad_norm 2.0962 (2.0783) [2021-04-16 11:25:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][1140/1251] eta 0:00:32 lr 0.000236 time 0.2814 (0.2935) loss 2.5750 (3.2735) grad_norm 2.4563 (2.0778) [2021-04-16 11:25:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][1150/1251] eta 0:00:29 lr 0.000236 time 0.2960 (0.2935) loss 2.8616 (3.2743) grad_norm 2.0548 (2.0767) [2021-04-16 11:26:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][1160/1251] eta 0:00:26 lr 0.000236 time 0.2674 (0.2936) loss 3.7287 (3.2756) grad_norm 1.9004 (2.0764) [2021-04-16 11:26:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][1170/1251] eta 0:00:23 lr 0.000236 time 0.2881 (0.2936) loss 2.9360 (3.2743) grad_norm 2.7741 (2.0774) [2021-04-16 11:26:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][1180/1251] eta 0:00:20 lr 0.000236 time 0.2676 (0.2935) loss 3.1153 (3.2774) grad_norm 2.0279 (2.0770) [2021-04-16 11:26:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][1190/1251] eta 0:00:17 lr 0.000236 time 0.2864 (0.2935) loss 4.0472 (3.2792) grad_norm 1.9255 (2.0766) [2021-04-16 11:26:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][1200/1251] eta 0:00:14 lr 0.000236 time 0.2667 (0.2933) loss 3.3490 (3.2784) grad_norm 1.8123 (2.0765) [2021-04-16 11:26:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][1210/1251] eta 0:00:12 lr 0.000236 time 0.2750 (0.2932) loss 2.9754 (3.2786) grad_norm 1.9282 (2.0767) [2021-04-16 11:26:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][1220/1251] eta 0:00:09 lr 0.000236 time 0.3194 (0.2931) loss 3.7056 (3.2809) grad_norm 1.9361 (2.0756) [2021-04-16 11:26:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][1230/1251] eta 0:00:06 lr 0.000235 time 0.2942 (0.2930) loss 3.0167 (3.2787) grad_norm 1.7681 (2.0759) [2021-04-16 11:26:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][1240/1251] eta 0:00:03 lr 0.000235 time 0.2492 (0.2928) loss 3.9423 (3.2809) grad_norm 2.4381 (2.0759) [2021-04-16 11:26:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [204/300][1250/1251] eta 0:00:00 lr 0.000235 time 0.2480 (0.2925) loss 2.8742 (3.2828) grad_norm 2.0853 (2.0749) [2021-04-16 11:26:31 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 204 training takes 0:06:09 [2021-04-16 11:26:31 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_204.pth saving...... [2021-04-16 11:26:45 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_204.pth saved !!! [2021-04-16 11:26:46 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.307 (1.307) Loss 0.9049 (0.9049) Acc@1 79.199 (79.199) Acc@5 94.922 (94.922) [2021-04-16 11:26:48 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.235 (0.320) Loss 1.0093 (0.9069) Acc@1 75.977 (78.480) Acc@5 93.262 (94.558) [2021-04-16 11:26:50 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.340 (0.238) Loss 0.8464 (0.9016) Acc@1 79.980 (78.706) Acc@5 95.508 (94.610) [2021-04-16 11:26:51 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.136 (0.221) Loss 0.8774 (0.9094) Acc@1 79.590 (78.585) Acc@5 94.922 (94.525) [2021-04-16 11:26:54 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.228) Loss 0.8894 (0.9097) Acc@1 78.613 (78.523) Acc@5 95.117 (94.543) [2021-04-16 11:27:00 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 78.546 Acc@5 94.530 [2021-04-16 11:27:00 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 78.5% [2021-04-16 11:27:00 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 78.55% [2021-04-16 11:27:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][0/1251] eta 3:33:51 lr 0.000235 time 10.2570 (10.2570) loss 3.4761 (3.4761) grad_norm 1.9739 (1.9739) [2021-04-16 11:27:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][10/1251] eta 0:24:45 lr 0.000235 time 0.3919 (1.1969) loss 2.9743 (3.2908) grad_norm 1.7284 (1.9569) [2021-04-16 11:27:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][20/1251] eta 0:15:40 lr 0.000235 time 0.2882 (0.7638) loss 3.2331 (3.2800) grad_norm 2.8983 (2.0413) [2021-04-16 11:27:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][30/1251] eta 0:12:22 lr 0.000235 time 0.2798 (0.6079) loss 2.1717 (3.2565) grad_norm 2.1727 (2.0913) [2021-04-16 11:27:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][40/1251] eta 0:10:41 lr 0.000235 time 0.3152 (0.5295) loss 2.9123 (3.2288) grad_norm 2.1877 (2.0854) [2021-04-16 11:27:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][50/1251] eta 0:09:36 lr 0.000235 time 0.2947 (0.4800) loss 3.1543 (3.2529) grad_norm 1.6694 (2.0799) [2021-04-16 11:27:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][60/1251] eta 0:08:52 lr 0.000235 time 0.3054 (0.4470) loss 3.0520 (3.2531) grad_norm 2.2367 (2.0866) [2021-04-16 11:27:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][70/1251] eta 0:08:20 lr 0.000235 time 0.2586 (0.4234) loss 2.7276 (3.2317) grad_norm 2.2207 (2.0831) [2021-04-16 11:27:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][80/1251] eta 0:07:54 lr 0.000235 time 0.2559 (0.4051) loss 3.7369 (3.2514) grad_norm 2.0478 (2.0756) [2021-04-16 11:27:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][90/1251] eta 0:07:33 lr 0.000235 time 0.2955 (0.3909) loss 3.9580 (3.2713) grad_norm 2.1333 (2.0783) [2021-04-16 11:27:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][100/1251] eta 0:07:17 lr 0.000235 time 0.2889 (0.3802) loss 3.9019 (3.2619) grad_norm 2.1176 (2.0788) [2021-04-16 11:27:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][110/1251] eta 0:07:03 lr 0.000235 time 0.3064 (0.3714) loss 3.6547 (3.2665) grad_norm 2.3956 (2.0897) [2021-04-16 11:27:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][120/1251] eta 0:06:50 lr 0.000235 time 0.2683 (0.3633) loss 2.1335 (3.2761) grad_norm 2.0446 (2.0863) [2021-04-16 11:27:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][130/1251] eta 0:06:40 lr 0.000235 time 0.2866 (0.3577) loss 3.2895 (3.2912) grad_norm 2.0872 (2.0839) [2021-04-16 11:27:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][140/1251] eta 0:06:32 lr 0.000235 time 0.2625 (0.3532) loss 3.7465 (3.2870) grad_norm 1.8891 (2.0825) [2021-04-16 11:27:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][150/1251] eta 0:06:23 lr 0.000235 time 0.2570 (0.3481) loss 2.3038 (3.2817) grad_norm 2.0873 (2.0756) [2021-04-16 11:27:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][160/1251] eta 0:06:16 lr 0.000235 time 0.3101 (0.3447) loss 3.1336 (3.2780) grad_norm 2.1177 (2.0805) [2021-04-16 11:27:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][170/1251] eta 0:06:08 lr 0.000235 time 0.2891 (0.3408) loss 3.5287 (3.2850) grad_norm 1.9035 (2.0774) [2021-04-16 11:28:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][180/1251] eta 0:06:02 lr 0.000235 time 0.4927 (0.3387) loss 3.5126 (3.2781) grad_norm 1.7835 (2.0782) [2021-04-16 11:28:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][190/1251] eta 0:05:56 lr 0.000235 time 0.2797 (0.3356) loss 3.3495 (3.2921) grad_norm 1.8489 (2.0688) [2021-04-16 11:28:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][200/1251] eta 0:05:49 lr 0.000235 time 0.2817 (0.3330) loss 2.5321 (3.2985) grad_norm 2.3209 (2.0728) [2021-04-16 11:28:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][210/1251] eta 0:05:43 lr 0.000235 time 0.2521 (0.3303) loss 3.1404 (3.3034) grad_norm 2.1060 (2.0715) [2021-04-16 11:28:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][220/1251] eta 0:05:38 lr 0.000235 time 0.2621 (0.3281) loss 3.7948 (3.3035) grad_norm 2.0696 (2.0792) [2021-04-16 11:28:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][230/1251] eta 0:05:32 lr 0.000235 time 0.2881 (0.3260) loss 3.5472 (3.2976) grad_norm 2.2470 (2.0802) [2021-04-16 11:28:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][240/1251] eta 0:05:27 lr 0.000235 time 0.2620 (0.3239) loss 3.2282 (3.2910) grad_norm 2.0601 (2.0802) [2021-04-16 11:28:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][250/1251] eta 0:05:22 lr 0.000235 time 0.2785 (0.3220) loss 3.6472 (3.2962) grad_norm 2.2547 (2.0825) [2021-04-16 11:28:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][260/1251] eta 0:05:17 lr 0.000235 time 0.2764 (0.3202) loss 3.1747 (3.2908) grad_norm 2.1518 (2.0870) [2021-04-16 11:28:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][270/1251] eta 0:05:12 lr 0.000234 time 0.2653 (0.3186) loss 3.4486 (3.2849) grad_norm 2.1699 (2.0847) [2021-04-16 11:28:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][280/1251] eta 0:05:08 lr 0.000234 time 0.2776 (0.3173) loss 3.5096 (3.2804) grad_norm 1.9083 (2.0855) [2021-04-16 11:28:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][290/1251] eta 0:05:03 lr 0.000234 time 0.2651 (0.3160) loss 3.9991 (3.2742) grad_norm 2.3391 (2.0957) [2021-04-16 11:28:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][300/1251] eta 0:04:59 lr 0.000234 time 0.2648 (0.3147) loss 3.5590 (3.2831) grad_norm 1.9198 (2.0952) [2021-04-16 11:28:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][310/1251] eta 0:04:54 lr 0.000234 time 0.2901 (0.3134) loss 3.8026 (3.2825) grad_norm 2.0386 (2.0950) [2021-04-16 11:28:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][320/1251] eta 0:04:51 lr 0.000234 time 0.2735 (0.3127) loss 3.3368 (3.2855) grad_norm 1.9433 (2.0945) [2021-04-16 11:28:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][330/1251] eta 0:04:47 lr 0.000234 time 0.2556 (0.3118) loss 3.3651 (3.2839) grad_norm 1.9956 (2.0935) [2021-04-16 11:28:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][340/1251] eta 0:04:43 lr 0.000234 time 0.2979 (0.3111) loss 3.5885 (3.2861) grad_norm 2.2318 (2.0933) [2021-04-16 11:28:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][350/1251] eta 0:04:39 lr 0.000234 time 0.4055 (0.3106) loss 2.7850 (3.2841) grad_norm 1.9084 (2.0933) [2021-04-16 11:28:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][360/1251] eta 0:04:36 lr 0.000234 time 0.2720 (0.3105) loss 3.8047 (3.2899) grad_norm 1.9590 (2.0910) [2021-04-16 11:28:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][370/1251] eta 0:04:33 lr 0.000234 time 0.2554 (0.3100) loss 3.7061 (3.2812) grad_norm 1.7140 (2.0903) [2021-04-16 11:28:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][380/1251] eta 0:04:29 lr 0.000234 time 0.2456 (0.3091) loss 3.7710 (3.2894) grad_norm 2.2408 (2.0883) [2021-04-16 11:29:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][390/1251] eta 0:04:25 lr 0.000234 time 0.2742 (0.3084) loss 3.2675 (3.2929) grad_norm 1.9596 (2.0870) [2021-04-16 11:29:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][400/1251] eta 0:04:21 lr 0.000234 time 0.2807 (0.3075) loss 3.2395 (3.2890) grad_norm 1.8505 (2.0842) [2021-04-16 11:29:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][410/1251] eta 0:04:18 lr 0.000234 time 0.2841 (0.3068) loss 2.8304 (3.2848) grad_norm 2.1661 (2.0850) [2021-04-16 11:29:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][420/1251] eta 0:04:14 lr 0.000234 time 0.2586 (0.3060) loss 2.3873 (3.2833) grad_norm 2.0561 (2.0872) [2021-04-16 11:29:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][430/1251] eta 0:04:10 lr 0.000234 time 0.2850 (0.3053) loss 3.3610 (3.2843) grad_norm 1.9445 (2.0865) [2021-04-16 11:29:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][440/1251] eta 0:04:07 lr 0.000234 time 0.3053 (0.3049) loss 2.5519 (3.2892) grad_norm 2.0694 (2.0868) [2021-04-16 11:29:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][450/1251] eta 0:04:03 lr 0.000234 time 0.2424 (0.3043) loss 2.6259 (3.2882) grad_norm 2.0008 (2.0857) [2021-04-16 11:29:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][460/1251] eta 0:04:00 lr 0.000234 time 0.2716 (0.3037) loss 3.3579 (3.2850) grad_norm 2.2911 (2.0856) [2021-04-16 11:29:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][470/1251] eta 0:03:56 lr 0.000234 time 0.2800 (0.3031) loss 3.7855 (3.2873) grad_norm 2.0559 (2.0823) [2021-04-16 11:29:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][480/1251] eta 0:03:53 lr 0.000234 time 0.2766 (0.3026) loss 2.8568 (3.2895) grad_norm 1.9586 (2.0781) [2021-04-16 11:29:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][490/1251] eta 0:03:50 lr 0.000234 time 0.2989 (0.3023) loss 3.4088 (3.2877) grad_norm 1.9740 (2.0748) [2021-04-16 11:29:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][500/1251] eta 0:03:46 lr 0.000234 time 0.2663 (0.3017) loss 4.1235 (3.2948) grad_norm 2.0306 (2.0766) [2021-04-16 11:29:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][510/1251] eta 0:03:43 lr 0.000234 time 0.2688 (0.3012) loss 2.6122 (3.2930) grad_norm 1.7850 (2.0740) [2021-04-16 11:29:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][520/1251] eta 0:03:39 lr 0.000234 time 0.2746 (0.3008) loss 3.2096 (3.2973) grad_norm 2.1063 (2.0740) [2021-04-16 11:29:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][530/1251] eta 0:03:36 lr 0.000234 time 0.2783 (0.3003) loss 2.7705 (3.2969) grad_norm 2.0260 (2.0727) [2021-04-16 11:29:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][540/1251] eta 0:03:33 lr 0.000234 time 0.2707 (0.2998) loss 3.0474 (3.2936) grad_norm 2.0540 (2.0718) [2021-04-16 11:29:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][550/1251] eta 0:03:29 lr 0.000233 time 0.2554 (0.2994) loss 3.0769 (3.2937) grad_norm 2.2056 (2.0721) [2021-04-16 11:29:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][560/1251] eta 0:03:26 lr 0.000233 time 0.2848 (0.2989) loss 2.7827 (3.2893) grad_norm 1.8197 (2.0714) [2021-04-16 11:29:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][570/1251] eta 0:03:23 lr 0.000233 time 0.2661 (0.2989) loss 3.2661 (3.2902) grad_norm 1.9875 (2.0710) [2021-04-16 11:29:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][580/1251] eta 0:03:20 lr 0.000233 time 0.2882 (0.2988) loss 2.5897 (3.2868) grad_norm 2.0333 (2.0698) [2021-04-16 11:29:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][590/1251] eta 0:03:17 lr 0.000233 time 0.2777 (0.2987) loss 2.5992 (3.2836) grad_norm 1.9744 (2.0689) [2021-04-16 11:29:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][600/1251] eta 0:03:14 lr 0.000233 time 0.2766 (0.2984) loss 3.2570 (3.2844) grad_norm 2.4036 (2.0697) [2021-04-16 11:30:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][610/1251] eta 0:03:11 lr 0.000233 time 0.2842 (0.2982) loss 3.5320 (3.2882) grad_norm 1.7984 (2.0691) [2021-04-16 11:30:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][620/1251] eta 0:03:08 lr 0.000233 time 0.2703 (0.2980) loss 2.8648 (3.2872) grad_norm 1.9407 (2.0729) [2021-04-16 11:30:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][630/1251] eta 0:03:04 lr 0.000233 time 0.2828 (0.2977) loss 3.3642 (3.2882) grad_norm 1.9077 (2.0747) [2021-04-16 11:30:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][640/1251] eta 0:03:01 lr 0.000233 time 0.2766 (0.2974) loss 2.8461 (3.2877) grad_norm 1.9695 (2.0747) [2021-04-16 11:30:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][650/1251] eta 0:02:58 lr 0.000233 time 0.2827 (0.2971) loss 4.0902 (3.2902) grad_norm 2.3148 (2.0772) [2021-04-16 11:30:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][660/1251] eta 0:02:55 lr 0.000233 time 0.2963 (0.2969) loss 3.8184 (3.2931) grad_norm 1.8251 (2.0779) [2021-04-16 11:30:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][670/1251] eta 0:02:52 lr 0.000233 time 0.2985 (0.2967) loss 2.9992 (3.2918) grad_norm 2.0262 (2.0777) [2021-04-16 11:30:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][680/1251] eta 0:02:49 lr 0.000233 time 0.2635 (0.2963) loss 4.1496 (3.2974) grad_norm 1.8864 (2.0762) [2021-04-16 11:30:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][690/1251] eta 0:02:46 lr 0.000233 time 0.2692 (0.2961) loss 3.5266 (3.2955) grad_norm 1.8710 (2.0769) [2021-04-16 11:30:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][700/1251] eta 0:02:43 lr 0.000233 time 0.3857 (0.2961) loss 3.4295 (3.2978) grad_norm 2.3373 (2.0771) [2021-04-16 11:30:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][710/1251] eta 0:02:40 lr 0.000233 time 0.2938 (0.2958) loss 3.2300 (3.2976) grad_norm 2.0880 (2.0783) [2021-04-16 11:30:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][720/1251] eta 0:02:36 lr 0.000233 time 0.2812 (0.2956) loss 3.1367 (3.2936) grad_norm 1.9954 (2.0775) [2021-04-16 11:30:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][730/1251] eta 0:02:33 lr 0.000233 time 0.3841 (0.2955) loss 3.8226 (3.2929) grad_norm 1.8613 (nan) [2021-04-16 11:30:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][740/1251] eta 0:02:30 lr 0.000233 time 0.2465 (0.2952) loss 2.4448 (3.2904) grad_norm 1.8536 (nan) [2021-04-16 11:30:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][750/1251] eta 0:02:27 lr 0.000233 time 0.2722 (0.2950) loss 3.9359 (3.2903) grad_norm 2.3345 (nan) [2021-04-16 11:30:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][760/1251] eta 0:02:24 lr 0.000233 time 0.2594 (0.2949) loss 2.7036 (3.2906) grad_norm 1.7761 (nan) [2021-04-16 11:30:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][770/1251] eta 0:02:21 lr 0.000233 time 0.2632 (0.2946) loss 4.0862 (3.2917) grad_norm 2.3039 (nan) [2021-04-16 11:30:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][780/1251] eta 0:02:18 lr 0.000233 time 0.2991 (0.2943) loss 3.3409 (3.2914) grad_norm 2.0901 (nan) [2021-04-16 11:30:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][790/1251] eta 0:02:15 lr 0.000233 time 0.2817 (0.2941) loss 3.5197 (3.2905) grad_norm 1.8965 (nan) [2021-04-16 11:30:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][800/1251] eta 0:02:12 lr 0.000233 time 0.2812 (0.2940) loss 3.0352 (3.2904) grad_norm 1.6660 (nan) [2021-04-16 11:30:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][810/1251] eta 0:02:09 lr 0.000233 time 0.2884 (0.2937) loss 4.5032 (3.2895) grad_norm 2.0773 (nan) [2021-04-16 11:31:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][820/1251] eta 0:02:06 lr 0.000233 time 0.2708 (0.2936) loss 3.7254 (3.2909) grad_norm 1.9482 (nan) [2021-04-16 11:31:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][830/1251] eta 0:02:03 lr 0.000233 time 0.2717 (0.2936) loss 3.3064 (3.2907) grad_norm 1.8298 (nan) [2021-04-16 11:31:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][840/1251] eta 0:02:00 lr 0.000232 time 0.2686 (0.2934) loss 3.4060 (3.2908) grad_norm 2.1988 (nan) [2021-04-16 11:31:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][850/1251] eta 0:01:57 lr 0.000232 time 0.2984 (0.2932) loss 3.5057 (3.2902) grad_norm 2.0807 (nan) [2021-04-16 11:31:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][860/1251] eta 0:01:54 lr 0.000232 time 0.2716 (0.2930) loss 3.4292 (3.2894) grad_norm 1.7779 (nan) [2021-04-16 11:31:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][870/1251] eta 0:01:51 lr 0.000232 time 0.2778 (0.2929) loss 3.3470 (3.2909) grad_norm 1.8001 (nan) [2021-04-16 11:31:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][880/1251] eta 0:01:48 lr 0.000232 time 0.2885 (0.2927) loss 3.9713 (3.2879) grad_norm 2.2295 (nan) [2021-04-16 11:31:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][890/1251] eta 0:01:45 lr 0.000232 time 0.2754 (0.2927) loss 3.3540 (3.2882) grad_norm 3.0375 (nan) [2021-04-16 11:31:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][900/1251] eta 0:01:42 lr 0.000232 time 0.2742 (0.2926) loss 2.5666 (3.2882) grad_norm 2.3622 (nan) [2021-04-16 11:31:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][910/1251] eta 0:01:39 lr 0.000232 time 0.2631 (0.2924) loss 3.1268 (3.2879) grad_norm 2.0224 (nan) [2021-04-16 11:31:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][920/1251] eta 0:01:36 lr 0.000232 time 0.2714 (0.2922) loss 2.8945 (3.2867) grad_norm 1.9016 (nan) [2021-04-16 11:31:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][930/1251] eta 0:01:33 lr 0.000232 time 0.2729 (0.2923) loss 3.5308 (3.2858) grad_norm 2.3166 (nan) [2021-04-16 11:31:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][940/1251] eta 0:01:30 lr 0.000232 time 0.2761 (0.2923) loss 3.6703 (3.2865) grad_norm 1.9328 (nan) [2021-04-16 11:31:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][950/1251] eta 0:01:27 lr 0.000232 time 0.2902 (0.2921) loss 2.9735 (3.2846) grad_norm 1.6758 (nan) [2021-04-16 11:31:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][960/1251] eta 0:01:24 lr 0.000232 time 0.2539 (0.2919) loss 3.3875 (3.2852) grad_norm 2.0868 (nan) [2021-04-16 11:31:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][970/1251] eta 0:01:21 lr 0.000232 time 0.2724 (0.2917) loss 2.5564 (3.2864) grad_norm 1.9276 (nan) [2021-04-16 11:31:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][980/1251] eta 0:01:19 lr 0.000232 time 0.2733 (0.2916) loss 3.2887 (3.2864) grad_norm 2.2742 (nan) [2021-04-16 11:31:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][990/1251] eta 0:01:16 lr 0.000232 time 0.2676 (0.2915) loss 3.3478 (3.2889) grad_norm 2.2961 (nan) [2021-04-16 11:31:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][1000/1251] eta 0:01:13 lr 0.000232 time 0.2859 (0.2913) loss 3.5708 (3.2876) grad_norm 2.3010 (nan) [2021-04-16 11:31:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][1010/1251] eta 0:01:10 lr 0.000232 time 0.2943 (0.2912) loss 2.6268 (3.2877) grad_norm 2.1073 (nan) [2021-04-16 11:31:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][1020/1251] eta 0:01:07 lr 0.000232 time 0.2694 (0.2911) loss 2.5541 (3.2872) grad_norm 1.7496 (nan) [2021-04-16 11:32:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][1030/1251] eta 0:01:04 lr 0.000232 time 0.2546 (0.2910) loss 3.4678 (3.2862) grad_norm 2.0528 (nan) [2021-04-16 11:32:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][1040/1251] eta 0:01:01 lr 0.000232 time 0.2773 (0.2908) loss 3.0119 (3.2862) grad_norm 2.3958 (nan) [2021-04-16 11:32:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][1050/1251] eta 0:00:58 lr 0.000232 time 0.2670 (0.2907) loss 4.0991 (3.2878) grad_norm 1.9126 (nan) [2021-04-16 11:32:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][1060/1251] eta 0:00:55 lr 0.000232 time 0.2629 (0.2906) loss 3.3394 (3.2914) grad_norm 2.4429 (nan) [2021-04-16 11:32:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][1070/1251] eta 0:00:52 lr 0.000232 time 0.2689 (0.2904) loss 3.0415 (3.2922) grad_norm 2.0121 (nan) [2021-04-16 11:32:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][1080/1251] eta 0:00:49 lr 0.000232 time 0.2896 (0.2903) loss 3.8935 (3.2915) grad_norm 2.2191 (nan) [2021-04-16 11:32:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][1090/1251] eta 0:00:46 lr 0.000232 time 0.2730 (0.2902) loss 2.8532 (3.2926) grad_norm 2.0868 (nan) [2021-04-16 11:32:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][1100/1251] eta 0:00:43 lr 0.000232 time 0.2820 (0.2901) loss 3.9398 (3.2936) grad_norm 2.2600 (nan) [2021-04-16 11:32:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][1110/1251] eta 0:00:40 lr 0.000232 time 0.2735 (0.2900) loss 3.0346 (3.2947) grad_norm 2.7548 (nan) [2021-04-16 11:32:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][1120/1251] eta 0:00:37 lr 0.000232 time 0.2883 (0.2900) loss 3.9432 (3.2936) grad_norm 2.2316 (nan) [2021-04-16 11:32:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][1130/1251] eta 0:00:35 lr 0.000231 time 0.2699 (0.2898) loss 3.8082 (3.2910) grad_norm 2.0271 (nan) [2021-04-16 11:32:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][1140/1251] eta 0:00:32 lr 0.000231 time 0.2915 (0.2898) loss 3.5569 (3.2908) grad_norm 2.4388 (nan) [2021-04-16 11:32:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][1150/1251] eta 0:00:29 lr 0.000231 time 0.2789 (0.2897) loss 3.6123 (3.2899) grad_norm 2.1997 (nan) [2021-04-16 11:32:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][1160/1251] eta 0:00:26 lr 0.000231 time 0.2468 (0.2897) loss 2.8792 (3.2933) grad_norm 2.1837 (nan) [2021-04-16 11:32:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][1170/1251] eta 0:00:23 lr 0.000231 time 0.2827 (0.2898) loss 3.1868 (3.2943) grad_norm 1.9855 (nan) [2021-04-16 11:32:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][1180/1251] eta 0:00:20 lr 0.000231 time 0.2748 (0.2898) loss 2.2603 (3.2937) grad_norm 1.9973 (nan) [2021-04-16 11:32:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][1190/1251] eta 0:00:17 lr 0.000231 time 0.4301 (0.2898) loss 3.4030 (3.2956) grad_norm 1.8396 (nan) [2021-04-16 11:32:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][1200/1251] eta 0:00:14 lr 0.000231 time 0.2472 (0.2897) loss 1.7950 (3.2918) grad_norm 1.8821 (nan) [2021-04-16 11:32:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][1210/1251] eta 0:00:11 lr 0.000231 time 0.2816 (0.2896) loss 3.1265 (3.2880) grad_norm 2.7431 (nan) [2021-04-16 11:32:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][1220/1251] eta 0:00:08 lr 0.000231 time 0.2532 (0.2895) loss 3.5071 (3.2896) grad_norm 2.3499 (nan) [2021-04-16 11:32:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][1230/1251] eta 0:00:06 lr 0.000231 time 0.2513 (0.2894) loss 2.3408 (3.2890) grad_norm 2.2350 (nan) [2021-04-16 11:32:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][1240/1251] eta 0:00:03 lr 0.000231 time 0.2492 (0.2892) loss 2.6892 (3.2901) grad_norm 2.0029 (nan) [2021-04-16 11:33:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [205/300][1250/1251] eta 0:00:00 lr 0.000231 time 0.2472 (0.2889) loss 2.2335 (3.2893) grad_norm 2.1681 (nan) [2021-04-16 11:33:06 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 205 training takes 0:06:06 [2021-04-16 11:33:06 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_205.pth saving...... [2021-04-16 11:33:16 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_205.pth saved !!! [2021-04-16 11:33:18 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.084 (1.084) Loss 0.9454 (0.9454) Acc@1 78.809 (78.809) Acc@5 93.750 (93.750) [2021-04-16 11:33:19 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.125 (0.263) Loss 1.0158 (0.9107) Acc@1 76.172 (78.587) Acc@5 93.555 (94.513) [2021-04-16 11:33:21 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.103 (0.219) Loss 0.9697 (0.9114) Acc@1 77.051 (78.664) Acc@5 94.434 (94.531) [2021-04-16 11:33:23 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.133 (0.226) Loss 0.8819 (0.9001) Acc@1 78.711 (79.117) Acc@5 95.312 (94.648) [2021-04-16 11:33:25 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.134 (0.210) Loss 0.9694 (0.9035) Acc@1 76.758 (78.840) Acc@5 93.555 (94.686) [2021-04-16 11:33:35 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 78.730 Acc@5 94.572 [2021-04-16 11:33:35 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 78.7% [2021-04-16 11:33:35 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 78.73% [2021-04-16 11:33:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][0/1251] eta 4:51:40 lr 0.000231 time 13.9892 (13.9892) loss 3.4699 (3.4699) grad_norm 2.0565 (2.0565) [2021-04-16 11:33:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][10/1251] eta 0:31:28 lr 0.000231 time 0.2874 (1.5217) loss 3.7057 (3.2115) grad_norm 1.8854 (2.1237) [2021-04-16 11:33:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][20/1251] eta 0:19:05 lr 0.000231 time 0.2622 (0.9305) loss 2.0841 (3.1175) grad_norm 2.7950 (2.1691) [2021-04-16 11:33:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][30/1251] eta 0:14:39 lr 0.000231 time 0.2878 (0.7201) loss 3.0257 (3.0955) grad_norm 1.9240 (2.1380) [2021-04-16 11:34:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][40/1251] eta 0:12:21 lr 0.000231 time 0.2929 (0.6125) loss 2.9056 (3.1034) grad_norm 2.7098 (2.1633) [2021-04-16 11:34:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][50/1251] eta 0:10:56 lr 0.000231 time 0.2648 (0.5467) loss 3.2673 (3.1846) grad_norm 2.0408 (2.1257) [2021-04-16 11:34:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][60/1251] eta 0:09:57 lr 0.000231 time 0.2658 (0.5019) loss 3.0380 (3.2394) grad_norm 1.7620 (2.0946) [2021-04-16 11:34:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][70/1251] eta 0:09:16 lr 0.000231 time 0.2714 (0.4710) loss 2.7601 (3.2501) grad_norm 1.9208 (2.0874) [2021-04-16 11:34:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][80/1251] eta 0:08:42 lr 0.000231 time 0.2784 (0.4466) loss 3.6160 (3.2512) grad_norm 2.5334 (2.1017) [2021-04-16 11:34:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][90/1251] eta 0:08:19 lr 0.000231 time 0.2869 (0.4299) loss 3.1013 (3.2454) grad_norm 1.9585 (2.0880) [2021-04-16 11:34:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][100/1251] eta 0:07:57 lr 0.000231 time 0.2947 (0.4149) loss 3.3059 (3.2542) grad_norm 1.8861 (2.0955) [2021-04-16 11:34:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][110/1251] eta 0:07:38 lr 0.000231 time 0.2877 (0.4020) loss 3.6035 (3.2718) grad_norm 2.3328 (2.1016) [2021-04-16 11:34:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][120/1251] eta 0:07:23 lr 0.000231 time 0.2869 (0.3918) loss 3.6274 (3.2675) grad_norm 2.6525 (2.1001) [2021-04-16 11:34:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][130/1251] eta 0:07:10 lr 0.000231 time 0.2592 (0.3836) loss 3.3110 (3.2659) grad_norm 2.2828 (2.1006) [2021-04-16 11:34:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][140/1251] eta 0:06:59 lr 0.000231 time 0.2523 (0.3774) loss 3.6832 (3.2660) grad_norm 2.3134 (2.1123) [2021-04-16 11:34:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][150/1251] eta 0:06:48 lr 0.000231 time 0.2491 (0.3714) loss 4.0131 (3.2759) grad_norm 2.1548 (2.1175) [2021-04-16 11:34:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][160/1251] eta 0:06:38 lr 0.000231 time 0.2809 (0.3656) loss 3.7738 (3.2929) grad_norm 1.8899 (2.1176) [2021-04-16 11:34:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][170/1251] eta 0:06:29 lr 0.000230 time 0.2699 (0.3606) loss 3.5868 (3.2905) grad_norm 2.5419 (2.1151) [2021-04-16 11:34:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][180/1251] eta 0:06:21 lr 0.000230 time 0.2594 (0.3561) loss 2.8414 (3.2821) grad_norm 2.0684 (2.1123) [2021-04-16 11:34:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][190/1251] eta 0:06:13 lr 0.000230 time 0.2985 (0.3521) loss 3.6883 (3.2804) grad_norm 2.2160 (2.1075) [2021-04-16 11:34:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][200/1251] eta 0:06:06 lr 0.000230 time 0.2617 (0.3483) loss 3.5367 (3.2796) grad_norm 2.2228 (2.1140) [2021-04-16 11:34:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][210/1251] eta 0:05:59 lr 0.000230 time 0.2812 (0.3449) loss 3.4778 (3.2927) grad_norm 2.3039 (2.1150) [2021-04-16 11:34:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][220/1251] eta 0:05:52 lr 0.000230 time 0.2562 (0.3416) loss 3.4379 (3.2872) grad_norm 2.0521 (2.1161) [2021-04-16 11:34:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][230/1251] eta 0:05:45 lr 0.000230 time 0.2732 (0.3388) loss 3.1499 (3.2952) grad_norm 2.5006 (2.1182) [2021-04-16 11:34:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][240/1251] eta 0:05:39 lr 0.000230 time 0.2903 (0.3362) loss 3.0924 (3.2925) grad_norm 2.4266 (2.1241) [2021-04-16 11:34:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][250/1251] eta 0:05:34 lr 0.000230 time 0.2949 (0.3339) loss 4.1449 (3.2912) grad_norm 1.9625 (2.1224) [2021-04-16 11:35:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][260/1251] eta 0:05:28 lr 0.000230 time 0.2729 (0.3316) loss 3.4370 (3.2941) grad_norm 3.0067 (2.1282) [2021-04-16 11:35:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][270/1251] eta 0:05:23 lr 0.000230 time 0.2844 (0.3299) loss 3.0669 (3.2958) grad_norm 2.0669 (2.1262) [2021-04-16 11:35:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][280/1251] eta 0:05:18 lr 0.000230 time 0.2924 (0.3279) loss 3.9539 (3.3075) grad_norm 1.9974 (2.1271) [2021-04-16 11:35:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][290/1251] eta 0:05:13 lr 0.000230 time 0.2896 (0.3263) loss 3.7717 (3.3180) grad_norm 1.9242 (2.1279) [2021-04-16 11:35:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][300/1251] eta 0:05:08 lr 0.000230 time 0.2673 (0.3246) loss 2.8578 (3.3148) grad_norm 2.1362 (2.1309) [2021-04-16 11:35:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][310/1251] eta 0:05:03 lr 0.000230 time 0.2680 (0.3230) loss 3.0974 (3.3159) grad_norm 2.4771 (2.1289) [2021-04-16 11:35:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][320/1251] eta 0:04:59 lr 0.000230 time 0.2680 (0.3216) loss 3.7229 (3.3247) grad_norm 1.9542 (2.1296) [2021-04-16 11:35:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][330/1251] eta 0:04:54 lr 0.000230 time 0.2880 (0.3201) loss 3.8688 (3.3341) grad_norm 2.1103 (2.1274) [2021-04-16 11:35:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][340/1251] eta 0:04:50 lr 0.000230 time 0.2485 (0.3190) loss 3.6542 (3.3316) grad_norm 1.9517 (2.1274) [2021-04-16 11:35:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][350/1251] eta 0:04:46 lr 0.000230 time 0.2850 (0.3178) loss 3.6431 (3.3339) grad_norm 2.2323 (2.1279) [2021-04-16 11:35:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][360/1251] eta 0:04:43 lr 0.000230 time 0.2986 (0.3177) loss 2.7839 (3.3394) grad_norm 2.0648 (2.1268) [2021-04-16 11:35:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][370/1251] eta 0:04:39 lr 0.000230 time 0.2719 (0.3170) loss 3.1731 (3.3372) grad_norm 1.9055 (2.1261) [2021-04-16 11:35:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][380/1251] eta 0:04:35 lr 0.000230 time 0.2685 (0.3163) loss 3.8639 (3.3442) grad_norm 2.1169 (2.1290) [2021-04-16 11:35:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][390/1251] eta 0:04:31 lr 0.000230 time 0.2782 (0.3158) loss 3.9528 (3.3376) grad_norm 2.0484 (2.1320) [2021-04-16 11:35:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][400/1251] eta 0:04:27 lr 0.000230 time 0.2692 (0.3147) loss 3.9971 (3.3371) grad_norm 2.1249 (2.1303) [2021-04-16 11:35:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][410/1251] eta 0:04:23 lr 0.000230 time 0.2496 (0.3138) loss 3.1106 (3.3375) grad_norm 1.9112 (2.1294) [2021-04-16 11:35:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][420/1251] eta 0:04:20 lr 0.000230 time 0.3100 (0.3131) loss 2.4230 (3.3337) grad_norm 1.8441 (2.1271) [2021-04-16 11:35:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][430/1251] eta 0:04:16 lr 0.000230 time 0.2879 (0.3122) loss 3.5383 (3.3336) grad_norm 2.0437 (2.1241) [2021-04-16 11:35:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][440/1251] eta 0:04:12 lr 0.000230 time 0.2934 (0.3116) loss 3.2389 (3.3250) grad_norm 2.0792 (2.1235) [2021-04-16 11:35:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][450/1251] eta 0:04:09 lr 0.000230 time 0.2773 (0.3109) loss 3.8291 (3.3209) grad_norm 2.1811 (2.1208) [2021-04-16 11:35:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][460/1251] eta 0:04:05 lr 0.000229 time 0.2736 (0.3101) loss 3.6768 (3.3162) grad_norm 1.9312 (2.1239) [2021-04-16 11:36:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][470/1251] eta 0:04:01 lr 0.000229 time 0.2651 (0.3093) loss 3.3593 (3.3189) grad_norm 1.7345 (2.1240) [2021-04-16 11:36:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][480/1251] eta 0:03:57 lr 0.000229 time 0.2860 (0.3086) loss 3.3546 (3.3134) grad_norm 2.0417 (2.1251) [2021-04-16 11:36:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][490/1251] eta 0:03:54 lr 0.000229 time 0.2576 (0.3080) loss 3.4514 (3.3135) grad_norm 2.0149 (2.1242) [2021-04-16 11:36:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][500/1251] eta 0:03:50 lr 0.000229 time 0.2944 (0.3075) loss 3.9440 (3.3144) grad_norm 2.3486 (2.1266) [2021-04-16 11:36:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][510/1251] eta 0:03:47 lr 0.000229 time 0.2728 (0.3069) loss 4.0262 (3.3193) grad_norm 1.9603 (2.1270) [2021-04-16 11:36:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][520/1251] eta 0:03:43 lr 0.000229 time 0.2715 (0.3064) loss 3.4662 (3.3163) grad_norm 2.0475 (2.1269) [2021-04-16 11:36:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][530/1251] eta 0:03:40 lr 0.000229 time 0.2624 (0.3057) loss 3.7987 (3.3184) grad_norm 2.0847 (2.1269) [2021-04-16 11:36:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][540/1251] eta 0:03:36 lr 0.000229 time 0.2925 (0.3052) loss 3.0522 (3.3178) grad_norm 2.0414 (2.1268) [2021-04-16 11:36:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][550/1251] eta 0:03:33 lr 0.000229 time 0.2623 (0.3045) loss 3.7567 (3.3189) grad_norm 2.5179 (2.1292) [2021-04-16 11:36:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][560/1251] eta 0:03:30 lr 0.000229 time 0.2767 (0.3041) loss 2.3521 (3.3170) grad_norm 3.2379 (2.1344) [2021-04-16 11:36:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][570/1251] eta 0:03:26 lr 0.000229 time 0.3827 (0.3038) loss 3.9989 (3.3187) grad_norm 1.8586 (2.1329) [2021-04-16 11:36:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][580/1251] eta 0:03:23 lr 0.000229 time 0.2807 (0.3038) loss 3.0591 (3.3173) grad_norm 1.9023 (2.1312) [2021-04-16 11:36:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][590/1251] eta 0:03:20 lr 0.000229 time 0.2721 (0.3033) loss 3.9682 (3.3203) grad_norm 2.3430 (2.1314) [2021-04-16 11:36:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][600/1251] eta 0:03:17 lr 0.000229 time 0.3018 (0.3029) loss 3.8945 (3.3236) grad_norm 2.0937 (2.1309) [2021-04-16 11:36:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][610/1251] eta 0:03:13 lr 0.000229 time 0.2834 (0.3024) loss 3.5565 (3.3241) grad_norm 3.2518 (2.1326) [2021-04-16 11:36:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][620/1251] eta 0:03:10 lr 0.000229 time 0.3003 (0.3020) loss 3.6599 (3.3218) grad_norm 2.1087 (2.1345) [2021-04-16 11:36:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][630/1251] eta 0:03:07 lr 0.000229 time 0.2714 (0.3015) loss 2.8074 (3.3247) grad_norm 2.1800 (2.1353) [2021-04-16 11:36:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][640/1251] eta 0:03:03 lr 0.000229 time 0.3009 (0.3011) loss 3.0372 (3.3190) grad_norm 2.0474 (2.1355) [2021-04-16 11:36:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][650/1251] eta 0:03:00 lr 0.000229 time 0.3057 (0.3009) loss 3.8516 (3.3177) grad_norm 2.0164 (2.1362) [2021-04-16 11:36:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][660/1251] eta 0:02:57 lr 0.000229 time 0.2944 (0.3005) loss 2.4533 (3.3160) grad_norm 1.8451 (2.1340) [2021-04-16 11:36:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][670/1251] eta 0:02:54 lr 0.000229 time 0.2735 (0.3004) loss 3.5496 (3.3157) grad_norm 1.6386 (2.1331) [2021-04-16 11:36:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][680/1251] eta 0:02:51 lr 0.000229 time 0.2781 (0.3000) loss 2.9545 (3.3144) grad_norm 1.8841 (2.1298) [2021-04-16 11:37:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][690/1251] eta 0:02:48 lr 0.000229 time 0.2676 (0.2997) loss 3.5749 (3.3152) grad_norm 1.8171 (2.1269) [2021-04-16 11:37:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][700/1251] eta 0:02:44 lr 0.000229 time 0.2763 (0.2994) loss 2.7227 (3.3160) grad_norm 1.9466 (2.1258) [2021-04-16 11:37:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][710/1251] eta 0:02:41 lr 0.000229 time 0.2664 (0.2992) loss 2.6103 (3.3147) grad_norm 2.2032 (2.1251) [2021-04-16 11:37:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][720/1251] eta 0:02:38 lr 0.000229 time 0.2526 (0.2992) loss 3.4206 (3.3155) grad_norm 1.8090 (2.1253) [2021-04-16 11:37:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][730/1251] eta 0:02:35 lr 0.000229 time 0.2822 (0.2990) loss 3.1065 (3.3160) grad_norm 2.0054 (2.1242) [2021-04-16 11:37:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][740/1251] eta 0:02:32 lr 0.000229 time 0.2830 (0.2987) loss 3.9771 (3.3158) grad_norm 1.9793 (2.1236) [2021-04-16 11:37:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][750/1251] eta 0:02:29 lr 0.000228 time 0.2763 (0.2984) loss 3.2005 (3.3195) grad_norm 2.0436 (2.1245) [2021-04-16 11:37:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][760/1251] eta 0:02:26 lr 0.000228 time 0.2605 (0.2980) loss 3.3652 (3.3193) grad_norm 3.1207 (2.1276) [2021-04-16 11:37:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][770/1251] eta 0:02:23 lr 0.000228 time 0.2883 (0.2978) loss 3.4975 (3.3207) grad_norm 2.1386 (2.1269) [2021-04-16 11:37:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][780/1251] eta 0:02:20 lr 0.000228 time 0.2724 (0.2975) loss 3.8001 (3.3214) grad_norm 1.8664 (2.1282) [2021-04-16 11:37:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][790/1251] eta 0:02:17 lr 0.000228 time 0.2623 (0.2973) loss 3.2313 (3.3207) grad_norm 2.0286 (2.1265) [2021-04-16 11:37:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][800/1251] eta 0:02:13 lr 0.000228 time 0.2513 (0.2970) loss 3.0743 (3.3221) grad_norm 2.4750 (2.1262) [2021-04-16 11:37:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][810/1251] eta 0:02:10 lr 0.000228 time 0.2675 (0.2967) loss 3.1532 (3.3176) grad_norm 2.0141 (2.1272) [2021-04-16 11:37:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][820/1251] eta 0:02:07 lr 0.000228 time 0.3090 (0.2966) loss 3.8685 (3.3185) grad_norm 2.4032 (2.1281) [2021-04-16 11:37:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][830/1251] eta 0:02:04 lr 0.000228 time 0.2899 (0.2963) loss 3.7851 (3.3192) grad_norm 1.9675 (2.1274) [2021-04-16 11:37:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][840/1251] eta 0:02:01 lr 0.000228 time 0.2564 (0.2961) loss 3.4761 (3.3196) grad_norm 1.9090 (2.1267) [2021-04-16 11:37:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][850/1251] eta 0:01:58 lr 0.000228 time 0.2718 (0.2959) loss 3.7206 (3.3231) grad_norm 1.9751 (2.1268) [2021-04-16 11:37:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][860/1251] eta 0:01:55 lr 0.000228 time 0.2858 (0.2957) loss 2.8169 (3.3200) grad_norm 2.0814 (2.1262) [2021-04-16 11:37:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][870/1251] eta 0:01:52 lr 0.000228 time 0.2683 (0.2955) loss 3.3692 (3.3184) grad_norm 2.3693 (2.1260) [2021-04-16 11:37:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][880/1251] eta 0:01:49 lr 0.000228 time 0.2728 (0.2953) loss 3.7673 (3.3200) grad_norm 1.8917 (2.1258) [2021-04-16 11:37:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][890/1251] eta 0:01:46 lr 0.000228 time 0.2731 (0.2950) loss 2.4569 (3.3190) grad_norm 1.9080 (2.1255) [2021-04-16 11:38:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][900/1251] eta 0:01:43 lr 0.000228 time 0.2663 (0.2948) loss 3.4154 (3.3151) grad_norm 2.1759 (2.1248) [2021-04-16 11:38:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][910/1251] eta 0:01:40 lr 0.000228 time 0.2846 (0.2947) loss 2.5563 (3.3108) grad_norm 1.7740 (2.1235) [2021-04-16 11:38:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][920/1251] eta 0:01:37 lr 0.000228 time 0.2884 (0.2945) loss 3.3648 (3.3076) grad_norm 1.7432 (2.1228) [2021-04-16 11:38:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][930/1251] eta 0:01:34 lr 0.000228 time 0.2562 (0.2944) loss 2.8112 (3.3070) grad_norm 1.9688 (2.1230) [2021-04-16 11:38:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][940/1251] eta 0:01:31 lr 0.000228 time 0.3163 (0.2945) loss 3.2582 (3.3058) grad_norm 1.9989 (2.1210) [2021-04-16 11:38:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][950/1251] eta 0:01:28 lr 0.000228 time 0.2803 (0.2942) loss 3.1951 (3.3058) grad_norm 2.3684 (2.1205) [2021-04-16 11:38:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][960/1251] eta 0:01:25 lr 0.000228 time 0.2880 (0.2941) loss 3.3866 (3.3048) grad_norm 1.7721 (2.1193) [2021-04-16 11:38:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][970/1251] eta 0:01:22 lr 0.000228 time 0.2927 (0.2940) loss 4.0647 (3.3037) grad_norm 2.2077 (2.1198) [2021-04-16 11:38:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][980/1251] eta 0:01:19 lr 0.000228 time 0.2616 (0.2939) loss 3.3032 (3.3028) grad_norm 2.0006 (2.1191) [2021-04-16 11:38:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][990/1251] eta 0:01:16 lr 0.000228 time 0.2525 (0.2937) loss 3.4845 (3.3002) grad_norm 1.9940 (2.1210) [2021-04-16 11:38:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][1000/1251] eta 0:01:13 lr 0.000228 time 0.2734 (0.2935) loss 3.7195 (3.2977) grad_norm 1.8937 (2.1202) [2021-04-16 11:38:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][1010/1251] eta 0:01:10 lr 0.000228 time 0.2554 (0.2933) loss 4.1711 (3.2993) grad_norm 1.8055 (2.1191) [2021-04-16 11:38:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][1020/1251] eta 0:01:07 lr 0.000228 time 0.2526 (0.2932) loss 2.1945 (3.2983) grad_norm 1.8374 (2.1186) [2021-04-16 11:38:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][1030/1251] eta 0:01:04 lr 0.000228 time 0.2704 (0.2930) loss 2.6794 (3.2978) grad_norm 3.1028 (2.1183) [2021-04-16 11:38:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][1040/1251] eta 0:01:01 lr 0.000227 time 0.2912 (0.2928) loss 2.8964 (3.2987) grad_norm 2.4214 (2.1178) [2021-04-16 11:38:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][1050/1251] eta 0:00:58 lr 0.000227 time 0.2793 (0.2926) loss 3.7277 (3.2972) grad_norm 1.9734 (2.1167) [2021-04-16 11:38:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][1060/1251] eta 0:00:55 lr 0.000227 time 0.3483 (0.2924) loss 3.1549 (3.2969) grad_norm 1.9744 (2.1151) [2021-04-16 11:38:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][1070/1251] eta 0:00:52 lr 0.000227 time 0.2448 (0.2922) loss 3.1242 (3.2968) grad_norm 2.1270 (2.1135) [2021-04-16 11:38:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][1080/1251] eta 0:00:49 lr 0.000227 time 0.2774 (0.2921) loss 4.0651 (3.2981) grad_norm 2.1039 (2.1124) [2021-04-16 11:38:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][1090/1251] eta 0:00:47 lr 0.000227 time 0.2814 (0.2919) loss 3.5219 (3.2994) grad_norm 2.1314 (2.1118) [2021-04-16 11:38:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][1100/1251] eta 0:00:44 lr 0.000227 time 0.2629 (0.2918) loss 2.7608 (3.3006) grad_norm 1.8081 (2.1108) [2021-04-16 11:38:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][1110/1251] eta 0:00:41 lr 0.000227 time 0.2571 (0.2917) loss 3.3846 (3.2982) grad_norm 2.0607 (2.1093) [2021-04-16 11:39:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][1120/1251] eta 0:00:38 lr 0.000227 time 0.2695 (0.2915) loss 3.7566 (3.3000) grad_norm 2.1092 (2.1090) [2021-04-16 11:39:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][1130/1251] eta 0:00:35 lr 0.000227 time 0.3005 (0.2914) loss 3.1551 (3.3011) grad_norm 2.1821 (2.1087) [2021-04-16 11:39:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][1140/1251] eta 0:00:32 lr 0.000227 time 0.2573 (0.2913) loss 3.9477 (3.3041) grad_norm 2.1377 (2.1082) [2021-04-16 11:39:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][1150/1251] eta 0:00:29 lr 0.000227 time 0.2806 (0.2913) loss 3.6349 (3.3043) grad_norm 1.9219 (2.1067) [2021-04-16 11:39:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][1160/1251] eta 0:00:26 lr 0.000227 time 0.2828 (0.2914) loss 3.1443 (3.3041) grad_norm 1.8896 (2.1057) [2021-04-16 11:39:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][1170/1251] eta 0:00:23 lr 0.000227 time 0.2652 (0.2912) loss 3.4390 (3.3043) grad_norm 1.9821 (2.1057) [2021-04-16 11:39:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][1180/1251] eta 0:00:20 lr 0.000227 time 0.2959 (0.2911) loss 3.8028 (3.3060) grad_norm 1.9012 (2.1070) [2021-04-16 11:39:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][1190/1251] eta 0:00:17 lr 0.000227 time 0.2548 (0.2910) loss 2.1916 (3.3058) grad_norm 1.8714 (2.1079) [2021-04-16 11:39:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][1200/1251] eta 0:00:14 lr 0.000227 time 0.2797 (0.2909) loss 2.2043 (3.3044) grad_norm 1.7996 (2.1082) [2021-04-16 11:39:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][1210/1251] eta 0:00:11 lr 0.000227 time 0.2556 (0.2907) loss 2.7173 (3.3039) grad_norm 1.8622 (2.1080) [2021-04-16 11:39:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][1220/1251] eta 0:00:09 lr 0.000227 time 0.2687 (0.2906) loss 2.9463 (3.3030) grad_norm 1.6631 (2.1064) [2021-04-16 11:39:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][1230/1251] eta 0:00:06 lr 0.000227 time 0.2819 (0.2905) loss 3.1073 (3.3032) grad_norm 2.0947 (2.1061) [2021-04-16 11:39:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][1240/1251] eta 0:00:03 lr 0.000227 time 0.3514 (0.2904) loss 3.5226 (3.3022) grad_norm 2.4230 (2.1060) [2021-04-16 11:39:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [206/300][1250/1251] eta 0:00:00 lr 0.000227 time 0.2486 (0.2901) loss 3.9657 (3.3041) grad_norm 2.0924 (2.1064) [2021-04-16 11:39:42 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 206 training takes 0:06:07 [2021-04-16 11:39:42 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_206.pth saving...... [2021-04-16 11:39:56 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_206.pth saved !!! [2021-04-16 11:39:57 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.151 (1.151) Loss 0.8793 (0.8793) Acc@1 79.785 (79.785) Acc@5 95.215 (95.215) [2021-04-16 11:39:58 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.128 (0.219) Loss 0.9315 (0.8949) Acc@1 78.223 (78.933) Acc@5 94.141 (94.620) [2021-04-16 11:40:00 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.139 (0.211) Loss 0.9072 (0.8931) Acc@1 79.199 (79.153) Acc@5 94.727 (94.634) [2021-04-16 11:40:03 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.143 (0.235) Loss 0.8559 (0.8997) Acc@1 80.176 (78.843) Acc@5 94.824 (94.616) [2021-04-16 11:40:05 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.221) Loss 0.8965 (0.8984) Acc@1 78.711 (78.830) Acc@5 94.922 (94.610) [2021-04-16 11:40:13 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 78.736 Acc@5 94.558 [2021-04-16 11:40:13 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 78.7% [2021-04-16 11:40:13 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 78.74% [2021-04-16 11:40:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][0/1251] eta 3:54:34 lr 0.000227 time 11.2504 (11.2504) loss 2.3155 (2.3155) grad_norm 1.9651 (1.9651) [2021-04-16 11:40:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][10/1251] eta 0:26:18 lr 0.000227 time 0.2792 (1.2717) loss 3.0798 (3.2403) grad_norm 1.9017 (2.0784) [2021-04-16 11:40:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][20/1251] eta 0:16:22 lr 0.000227 time 0.2847 (0.7980) loss 3.6780 (3.4360) grad_norm 2.0570 (2.0399) [2021-04-16 11:40:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][30/1251] eta 0:12:51 lr 0.000227 time 0.3026 (0.6322) loss 3.8683 (3.2963) grad_norm 2.0656 (2.0549) [2021-04-16 11:40:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][40/1251] eta 0:10:58 lr 0.000227 time 0.2515 (0.5438) loss 2.4516 (3.2507) grad_norm 1.9674 (2.0400) [2021-04-16 11:40:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][50/1251] eta 0:09:49 lr 0.000227 time 0.2620 (0.4913) loss 3.2442 (3.1874) grad_norm 2.3305 (2.0609) [2021-04-16 11:40:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][60/1251] eta 0:09:05 lr 0.000227 time 0.2843 (0.4576) loss 3.3724 (3.2355) grad_norm 2.0837 (2.0558) [2021-04-16 11:40:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][70/1251] eta 0:08:29 lr 0.000227 time 0.2866 (0.4317) loss 3.8373 (3.2467) grad_norm 2.0961 (2.0651) [2021-04-16 11:40:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][80/1251] eta 0:08:03 lr 0.000226 time 0.2940 (0.4126) loss 3.5623 (3.2573) grad_norm 2.0079 (2.0652) [2021-04-16 11:40:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][90/1251] eta 0:07:41 lr 0.000226 time 0.2908 (0.3979) loss 3.2139 (3.2630) grad_norm 2.1339 (2.0700) [2021-04-16 11:40:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][100/1251] eta 0:07:24 lr 0.000226 time 0.2671 (0.3858) loss 4.0164 (3.2859) grad_norm 2.1622 (2.0686) [2021-04-16 11:40:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][110/1251] eta 0:07:10 lr 0.000226 time 0.3000 (0.3770) loss 2.5945 (3.2880) grad_norm 1.8066 (2.0613) [2021-04-16 11:40:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][120/1251] eta 0:06:57 lr 0.000226 time 0.2707 (0.3689) loss 2.7117 (3.2646) grad_norm 1.8351 (2.0512) [2021-04-16 11:41:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][130/1251] eta 0:06:45 lr 0.000226 time 0.2677 (0.3615) loss 3.6706 (3.2807) grad_norm 1.9115 (2.0522) [2021-04-16 11:41:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][140/1251] eta 0:06:36 lr 0.000226 time 0.2511 (0.3571) loss 2.2015 (3.2631) grad_norm 1.9311 (2.0514) [2021-04-16 11:41:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][150/1251] eta 0:06:29 lr 0.000226 time 0.3352 (0.3536) loss 2.3065 (3.2503) grad_norm 2.3906 (2.0521) [2021-04-16 11:41:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][160/1251] eta 0:06:20 lr 0.000226 time 0.2720 (0.3488) loss 3.5675 (3.2608) grad_norm 2.1040 (2.0619) [2021-04-16 11:41:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][170/1251] eta 0:06:12 lr 0.000226 time 0.3134 (0.3448) loss 3.6814 (3.2708) grad_norm 1.9719 (2.0637) [2021-04-16 11:41:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][180/1251] eta 0:06:05 lr 0.000226 time 0.2852 (0.3417) loss 3.1546 (3.2669) grad_norm 1.9915 (2.0627) [2021-04-16 11:41:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][190/1251] eta 0:05:58 lr 0.000226 time 0.2605 (0.3383) loss 2.4200 (3.2587) grad_norm 1.9054 (2.0604) [2021-04-16 11:41:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][200/1251] eta 0:05:52 lr 0.000226 time 0.2733 (0.3354) loss 3.5313 (3.2622) grad_norm 2.1373 (2.0682) [2021-04-16 11:41:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][210/1251] eta 0:05:46 lr 0.000226 time 0.2720 (0.3325) loss 3.4589 (3.2600) grad_norm 1.8180 (2.0727) [2021-04-16 11:41:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][220/1251] eta 0:05:39 lr 0.000226 time 0.2776 (0.3298) loss 3.0793 (3.2628) grad_norm 1.9373 (2.0707) [2021-04-16 11:41:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][230/1251] eta 0:05:34 lr 0.000226 time 0.2816 (0.3274) loss 3.7530 (3.2753) grad_norm 1.6613 (2.0713) [2021-04-16 11:41:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][240/1251] eta 0:05:29 lr 0.000226 time 0.2673 (0.3256) loss 3.2105 (3.2739) grad_norm 1.9283 (2.0685) [2021-04-16 11:41:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][250/1251] eta 0:05:24 lr 0.000226 time 0.2607 (0.3238) loss 3.3371 (3.2838) grad_norm 2.2289 (2.0720) [2021-04-16 11:41:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][260/1251] eta 0:05:19 lr 0.000226 time 0.2639 (0.3222) loss 3.6316 (3.2705) grad_norm 2.0004 (2.0691) [2021-04-16 11:41:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][270/1251] eta 0:05:14 lr 0.000226 time 0.2785 (0.3208) loss 2.5430 (3.2684) grad_norm 2.0764 (2.0674) [2021-04-16 11:41:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][280/1251] eta 0:05:10 lr 0.000226 time 0.2471 (0.3193) loss 2.9994 (3.2651) grad_norm 1.9240 (2.0667) [2021-04-16 11:41:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][290/1251] eta 0:05:05 lr 0.000226 time 0.2869 (0.3178) loss 3.4541 (3.2648) grad_norm 2.0235 (2.0646) [2021-04-16 11:41:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][300/1251] eta 0:05:00 lr 0.000226 time 0.2553 (0.3163) loss 2.1122 (3.2671) grad_norm 2.0260 (2.0609) [2021-04-16 11:41:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][310/1251] eta 0:04:56 lr 0.000226 time 0.3093 (0.3153) loss 3.3967 (3.2663) grad_norm 2.3781 (2.0668) [2021-04-16 11:41:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][320/1251] eta 0:04:52 lr 0.000226 time 0.2604 (0.3139) loss 4.1578 (3.2693) grad_norm 2.1526 (2.0663) [2021-04-16 11:41:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][330/1251] eta 0:04:48 lr 0.000226 time 0.2739 (0.3128) loss 3.6212 (3.2767) grad_norm 2.1405 (2.0664) [2021-04-16 11:41:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][340/1251] eta 0:04:44 lr 0.000226 time 0.3617 (0.3121) loss 3.8755 (3.2801) grad_norm 2.1154 (2.0659) [2021-04-16 11:42:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][350/1251] eta 0:04:40 lr 0.000226 time 0.2901 (0.3112) loss 3.0891 (3.2750) grad_norm 1.8755 (2.0644) [2021-04-16 11:42:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][360/1251] eta 0:04:37 lr 0.000226 time 0.2711 (0.3109) loss 2.2782 (3.2734) grad_norm 1.9442 (2.0641) [2021-04-16 11:42:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][370/1251] eta 0:04:33 lr 0.000226 time 0.2497 (0.3107) loss 3.8147 (3.2729) grad_norm 1.9147 (2.0630) [2021-04-16 11:42:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][380/1251] eta 0:04:29 lr 0.000225 time 0.2988 (0.3097) loss 3.4065 (3.2787) grad_norm 2.1006 (2.0624) [2021-04-16 11:42:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][390/1251] eta 0:04:26 lr 0.000225 time 0.2691 (0.3090) loss 3.5689 (3.2819) grad_norm 1.8523 (2.0630) [2021-04-16 11:42:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][400/1251] eta 0:04:22 lr 0.000225 time 0.2805 (0.3083) loss 3.3204 (3.2868) grad_norm 2.1353 (2.0646) [2021-04-16 11:42:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][410/1251] eta 0:04:18 lr 0.000225 time 0.2817 (0.3078) loss 3.5725 (3.2821) grad_norm 2.2320 (2.0641) [2021-04-16 11:42:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][420/1251] eta 0:04:15 lr 0.000225 time 0.2911 (0.3073) loss 2.5088 (3.2852) grad_norm 1.9799 (2.0629) [2021-04-16 11:42:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][430/1251] eta 0:04:11 lr 0.000225 time 0.2869 (0.3066) loss 3.1475 (3.2837) grad_norm 1.9406 (2.0614) [2021-04-16 11:42:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][440/1251] eta 0:04:08 lr 0.000225 time 0.2772 (0.3058) loss 3.1048 (3.2906) grad_norm 1.9139 (2.0617) [2021-04-16 11:42:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][450/1251] eta 0:04:04 lr 0.000225 time 0.5284 (0.3058) loss 3.8010 (3.2807) grad_norm 1.9810 (2.0629) [2021-04-16 11:42:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][460/1251] eta 0:04:01 lr 0.000225 time 0.2579 (0.3050) loss 3.7782 (3.2826) grad_norm 1.9523 (2.0627) [2021-04-16 11:42:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][470/1251] eta 0:03:57 lr 0.000225 time 0.2633 (0.3044) loss 2.6274 (3.2848) grad_norm 1.7190 (2.0628) [2021-04-16 11:42:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][480/1251] eta 0:03:54 lr 0.000225 time 0.2950 (0.3038) loss 3.5387 (3.2870) grad_norm 2.0911 (2.0657) [2021-04-16 11:42:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][490/1251] eta 0:03:50 lr 0.000225 time 0.2761 (0.3033) loss 4.2924 (3.2863) grad_norm 3.3422 (2.0681) [2021-04-16 11:42:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][500/1251] eta 0:03:47 lr 0.000225 time 0.2670 (0.3028) loss 3.7177 (3.2886) grad_norm 1.9202 (2.0675) [2021-04-16 11:42:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][510/1251] eta 0:03:44 lr 0.000225 time 0.2577 (0.3023) loss 3.4691 (3.2882) grad_norm 2.0077 (2.0693) [2021-04-16 11:42:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][520/1251] eta 0:03:40 lr 0.000225 time 0.2761 (0.3018) loss 3.6531 (3.2914) grad_norm 1.9136 (2.0696) [2021-04-16 11:42:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][530/1251] eta 0:03:37 lr 0.000225 time 0.2718 (0.3013) loss 3.1901 (3.2894) grad_norm 2.2774 (2.0713) [2021-04-16 11:42:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][540/1251] eta 0:03:34 lr 0.000225 time 0.2949 (0.3013) loss 3.9296 (3.2958) grad_norm 2.1858 (2.0718) [2021-04-16 11:42:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][550/1251] eta 0:03:30 lr 0.000225 time 0.2799 (0.3008) loss 3.4455 (3.2915) grad_norm 2.5663 (2.0762) [2021-04-16 11:43:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][560/1251] eta 0:03:27 lr 0.000225 time 0.2557 (0.3003) loss 2.5721 (3.2933) grad_norm 2.2378 (2.0805) [2021-04-16 11:43:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][570/1251] eta 0:03:24 lr 0.000225 time 0.2722 (0.3001) loss 2.5456 (3.2897) grad_norm 2.0631 (2.0807) [2021-04-16 11:43:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][580/1251] eta 0:03:21 lr 0.000225 time 0.2885 (0.3001) loss 2.8729 (3.2902) grad_norm 1.9972 (2.0824) [2021-04-16 11:43:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][590/1251] eta 0:03:18 lr 0.000225 time 0.2674 (0.2997) loss 3.4296 (3.2886) grad_norm 1.9686 (2.0815) [2021-04-16 11:43:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][600/1251] eta 0:03:14 lr 0.000225 time 0.2793 (0.2992) loss 3.9455 (3.2900) grad_norm 2.1242 (2.0811) [2021-04-16 11:43:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][610/1251] eta 0:03:11 lr 0.000225 time 0.2833 (0.2988) loss 3.4332 (3.2913) grad_norm 2.0796 (2.0852) [2021-04-16 11:43:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][620/1251] eta 0:03:08 lr 0.000225 time 0.2522 (0.2984) loss 3.3565 (3.2870) grad_norm 2.0274 (2.0867) [2021-04-16 11:43:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][630/1251] eta 0:03:05 lr 0.000225 time 0.2339 (0.2980) loss 2.9455 (3.2870) grad_norm nan (nan) [2021-04-16 11:43:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][640/1251] eta 0:03:01 lr 0.000225 time 0.3060 (0.2976) loss 3.7664 (3.2874) grad_norm 1.8142 (nan) [2021-04-16 11:43:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][650/1251] eta 0:02:58 lr 0.000225 time 0.2682 (0.2973) loss 2.3075 (3.2859) grad_norm 1.9430 (nan) [2021-04-16 11:43:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][660/1251] eta 0:02:55 lr 0.000225 time 0.2835 (0.2970) loss 2.9657 (3.2866) grad_norm 1.8515 (nan) [2021-04-16 11:43:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][670/1251] eta 0:02:52 lr 0.000224 time 0.2919 (0.2966) loss 3.2612 (3.2891) grad_norm 1.8545 (nan) [2021-04-16 11:43:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][680/1251] eta 0:02:49 lr 0.000224 time 0.2697 (0.2964) loss 2.4846 (3.2905) grad_norm 2.1982 (nan) [2021-04-16 11:43:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][690/1251] eta 0:02:46 lr 0.000224 time 0.2754 (0.2960) loss 3.0777 (3.2877) grad_norm 1.9886 (nan) [2021-04-16 11:43:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][700/1251] eta 0:02:42 lr 0.000224 time 0.2756 (0.2957) loss 2.7001 (3.2874) grad_norm 1.9176 (nan) [2021-04-16 11:43:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][710/1251] eta 0:02:39 lr 0.000224 time 0.2677 (0.2955) loss 3.3372 (3.2884) grad_norm 2.2089 (nan) [2021-04-16 11:43:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][720/1251] eta 0:02:37 lr 0.000224 time 0.2456 (0.2957) loss 2.6911 (3.2864) grad_norm 2.2998 (nan) [2021-04-16 11:43:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][730/1251] eta 0:02:34 lr 0.000224 time 0.2650 (0.2956) loss 3.6204 (3.2884) grad_norm 2.1694 (nan) [2021-04-16 11:43:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][740/1251] eta 0:02:30 lr 0.000224 time 0.2784 (0.2954) loss 3.8461 (3.2880) grad_norm 1.8176 (nan) [2021-04-16 11:43:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][750/1251] eta 0:02:27 lr 0.000224 time 0.2621 (0.2951) loss 3.2928 (3.2878) grad_norm 1.7539 (nan) [2021-04-16 11:43:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][760/1251] eta 0:02:24 lr 0.000224 time 0.2532 (0.2950) loss 3.9014 (3.2885) grad_norm 2.0113 (nan) [2021-04-16 11:44:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][770/1251] eta 0:02:21 lr 0.000224 time 0.3009 (0.2949) loss 3.2228 (3.2879) grad_norm 1.8845 (nan) [2021-04-16 11:44:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][780/1251] eta 0:02:18 lr 0.000224 time 0.2734 (0.2946) loss 3.8655 (3.2889) grad_norm 2.1159 (nan) [2021-04-16 11:44:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][790/1251] eta 0:02:15 lr 0.000224 time 0.2725 (0.2943) loss 3.2126 (3.2912) grad_norm 1.7478 (nan) [2021-04-16 11:44:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][800/1251] eta 0:02:12 lr 0.000224 time 0.2767 (0.2941) loss 2.9168 (3.2905) grad_norm 1.9532 (nan) [2021-04-16 11:44:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][810/1251] eta 0:02:09 lr 0.000224 time 0.2803 (0.2939) loss 3.5249 (3.2952) grad_norm 2.3789 (nan) [2021-04-16 11:44:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][820/1251] eta 0:02:06 lr 0.000224 time 0.2880 (0.2937) loss 3.7397 (3.2949) grad_norm 1.8304 (nan) [2021-04-16 11:44:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][830/1251] eta 0:02:03 lr 0.000224 time 0.2947 (0.2935) loss 3.7587 (3.2996) grad_norm 2.0179 (nan) [2021-04-16 11:44:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][840/1251] eta 0:02:00 lr 0.000224 time 0.2603 (0.2933) loss 3.1874 (3.3016) grad_norm 1.8181 (nan) [2021-04-16 11:44:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][850/1251] eta 0:01:57 lr 0.000224 time 0.2791 (0.2931) loss 3.7632 (3.3039) grad_norm 2.2083 (nan) [2021-04-16 11:44:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][860/1251] eta 0:01:54 lr 0.000224 time 0.2727 (0.2929) loss 2.6782 (3.3033) grad_norm 1.9252 (nan) [2021-04-16 11:44:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][870/1251] eta 0:01:51 lr 0.000224 time 0.2680 (0.2927) loss 2.7852 (3.3049) grad_norm 2.4053 (nan) [2021-04-16 11:44:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][880/1251] eta 0:01:48 lr 0.000224 time 0.2964 (0.2925) loss 3.4167 (3.3036) grad_norm 2.0747 (nan) [2021-04-16 11:44:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][890/1251] eta 0:01:45 lr 0.000224 time 0.2791 (0.2923) loss 2.2049 (3.3044) grad_norm 1.8042 (nan) [2021-04-16 11:44:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][900/1251] eta 0:01:42 lr 0.000224 time 0.2925 (0.2921) loss 4.2120 (3.3066) grad_norm 2.1633 (nan) [2021-04-16 11:44:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][910/1251] eta 0:01:39 lr 0.000224 time 0.2649 (0.2920) loss 2.7672 (3.3079) grad_norm 1.8309 (nan) [2021-04-16 11:44:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][920/1251] eta 0:01:36 lr 0.000224 time 0.2743 (0.2919) loss 3.8685 (3.3092) grad_norm 2.3699 (nan) [2021-04-16 11:44:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][930/1251] eta 0:01:33 lr 0.000224 time 0.2839 (0.2919) loss 3.2006 (3.3109) grad_norm 1.9566 (nan) [2021-04-16 11:44:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][940/1251] eta 0:01:30 lr 0.000224 time 0.2636 (0.2920) loss 3.2661 (3.3100) grad_norm 2.0010 (nan) [2021-04-16 11:44:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][950/1251] eta 0:01:27 lr 0.000224 time 0.2930 (0.2921) loss 2.7149 (3.3084) grad_norm 1.9947 (nan) [2021-04-16 11:44:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][960/1251] eta 0:01:24 lr 0.000223 time 0.2892 (0.2921) loss 2.9999 (3.3064) grad_norm 1.9657 (nan) [2021-04-16 11:44:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][970/1251] eta 0:01:22 lr 0.000223 time 0.2700 (0.2919) loss 3.7308 (3.3045) grad_norm 2.0118 (nan) [2021-04-16 11:44:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][980/1251] eta 0:01:19 lr 0.000223 time 0.2802 (0.2917) loss 3.7231 (3.3074) grad_norm 2.3517 (nan) [2021-04-16 11:45:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][990/1251] eta 0:01:16 lr 0.000223 time 0.2559 (0.2916) loss 3.5533 (3.3067) grad_norm 2.0551 (nan) [2021-04-16 11:45:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][1000/1251] eta 0:01:13 lr 0.000223 time 0.2725 (0.2914) loss 2.9814 (3.3049) grad_norm 2.4335 (nan) [2021-04-16 11:45:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][1010/1251] eta 0:01:10 lr 0.000223 time 0.2758 (0.2913) loss 3.6653 (3.3053) grad_norm 1.9926 (nan) [2021-04-16 11:45:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][1020/1251] eta 0:01:07 lr 0.000223 time 0.2570 (0.2911) loss 2.3594 (3.3056) grad_norm 1.9895 (nan) [2021-04-16 11:45:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][1030/1251] eta 0:01:04 lr 0.000223 time 0.3023 (0.2912) loss 2.3641 (3.3059) grad_norm 2.1698 (nan) [2021-04-16 11:45:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][1040/1251] eta 0:01:01 lr 0.000223 time 0.2900 (0.2910) loss 3.8623 (3.3073) grad_norm 2.0271 (nan) [2021-04-16 11:45:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][1050/1251] eta 0:00:58 lr 0.000223 time 0.2810 (0.2909) loss 3.5273 (3.3056) grad_norm 2.0653 (nan) [2021-04-16 11:45:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][1060/1251] eta 0:00:55 lr 0.000223 time 0.2748 (0.2908) loss 3.2293 (3.3010) grad_norm 2.3199 (nan) [2021-04-16 11:45:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][1070/1251] eta 0:00:52 lr 0.000223 time 0.2621 (0.2906) loss 3.5787 (3.3040) grad_norm 2.1086 (nan) [2021-04-16 11:45:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][1080/1251] eta 0:00:49 lr 0.000223 time 0.2895 (0.2905) loss 3.3302 (3.3023) grad_norm 1.9924 (nan) [2021-04-16 11:45:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][1090/1251] eta 0:00:46 lr 0.000223 time 0.3033 (0.2904) loss 3.7565 (3.3022) grad_norm 2.2768 (nan) [2021-04-16 11:45:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][1100/1251] eta 0:00:43 lr 0.000223 time 0.2880 (0.2903) loss 3.3145 (3.3026) grad_norm 2.0707 (nan) [2021-04-16 11:45:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][1110/1251] eta 0:00:40 lr 0.000223 time 0.2593 (0.2902) loss 3.5484 (3.3047) grad_norm 1.8820 (nan) [2021-04-16 11:45:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][1120/1251] eta 0:00:38 lr 0.000223 time 0.2924 (0.2901) loss 2.8690 (3.3041) grad_norm 1.9594 (nan) [2021-04-16 11:45:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][1130/1251] eta 0:00:35 lr 0.000223 time 0.2616 (0.2900) loss 3.7426 (3.3023) grad_norm 1.9342 (nan) [2021-04-16 11:45:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][1140/1251] eta 0:00:32 lr 0.000223 time 0.2586 (0.2899) loss 3.7394 (3.3049) grad_norm 2.1022 (nan) [2021-04-16 11:45:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][1150/1251] eta 0:00:29 lr 0.000223 time 0.2462 (0.2899) loss 3.9804 (3.3047) grad_norm 2.3377 (nan) [2021-04-16 11:45:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][1160/1251] eta 0:00:26 lr 0.000223 time 0.2951 (0.2899) loss 2.7424 (3.3058) grad_norm 1.8705 (nan) [2021-04-16 11:45:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][1170/1251] eta 0:00:23 lr 0.000223 time 0.2888 (0.2899) loss 2.7056 (3.3071) grad_norm 2.3672 (nan) [2021-04-16 11:45:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][1180/1251] eta 0:00:20 lr 0.000223 time 0.3080 (0.2898) loss 3.4016 (3.3089) grad_norm 2.3883 (nan) [2021-04-16 11:45:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][1190/1251] eta 0:00:17 lr 0.000223 time 0.2569 (0.2897) loss 3.4322 (3.3083) grad_norm 1.9779 (nan) [2021-04-16 11:46:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][1200/1251] eta 0:00:14 lr 0.000223 time 0.2851 (0.2896) loss 2.7118 (3.3074) grad_norm 2.7955 (nan) [2021-04-16 11:46:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][1210/1251] eta 0:00:11 lr 0.000223 time 0.2978 (0.2895) loss 3.6871 (3.3081) grad_norm 2.1566 (nan) [2021-04-16 11:46:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][1220/1251] eta 0:00:08 lr 0.000223 time 0.2786 (0.2894) loss 3.7605 (3.3086) grad_norm 1.8830 (nan) [2021-04-16 11:46:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][1230/1251] eta 0:00:06 lr 0.000223 time 0.2752 (0.2892) loss 3.0443 (3.3084) grad_norm 2.0306 (nan) [2021-04-16 11:46:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][1240/1251] eta 0:00:03 lr 0.000223 time 0.2625 (0.2891) loss 3.5162 (3.3074) grad_norm 2.7191 (nan) [2021-04-16 11:46:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [207/300][1250/1251] eta 0:00:00 lr 0.000223 time 0.2737 (0.2888) loss 3.3896 (3.3084) grad_norm 2.6818 (nan) [2021-04-16 11:46:17 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 207 training takes 0:06:04 [2021-04-16 11:46:17 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_207.pth saving...... [2021-04-16 11:46:31 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_207.pth saved !!! [2021-04-16 11:46:32 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.106 (1.106) Loss 0.9174 (0.9174) Acc@1 77.832 (77.832) Acc@5 94.141 (94.141) [2021-04-16 11:46:34 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.433 (0.259) Loss 0.9093 (0.9294) Acc@1 79.688 (77.965) Acc@5 94.629 (94.469) [2021-04-16 11:46:37 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.140 (0.254) Loss 0.9496 (0.9167) Acc@1 77.148 (78.246) Acc@5 93.848 (94.434) [2021-04-16 11:46:39 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.091 (0.237) Loss 0.8915 (0.9114) Acc@1 80.078 (78.462) Acc@5 95.020 (94.481) [2021-04-16 11:46:40 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.222) Loss 0.9381 (0.9102) Acc@1 78.711 (78.513) Acc@5 93.848 (94.450) [2021-04-16 11:46:54 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 78.640 Acc@5 94.496 [2021-04-16 11:46:54 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 78.6% [2021-04-16 11:46:54 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 78.74% [2021-04-16 11:47:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][0/1251] eta 3:04:09 lr 0.000222 time 8.8323 (8.8323) loss 3.9439 (3.9439) grad_norm 2.0115 (2.0115) [2021-04-16 11:47:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][10/1251] eta 0:21:44 lr 0.000222 time 0.2824 (1.0512) loss 3.4790 (3.2964) grad_norm 1.7722 (2.0840) [2021-04-16 11:47:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][20/1251] eta 0:14:02 lr 0.000222 time 0.2761 (0.6842) loss 3.4630 (3.3252) grad_norm 1.9993 (2.1970) [2021-04-16 11:47:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][30/1251] eta 0:11:15 lr 0.000222 time 0.2579 (0.5528) loss 3.8447 (3.2671) grad_norm 2.7284 (2.1902) [2021-04-16 11:47:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][40/1251] eta 0:09:47 lr 0.000222 time 0.2675 (0.4852) loss 4.0522 (3.3200) grad_norm 2.3702 (2.1794) [2021-04-16 11:47:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][50/1251] eta 0:08:54 lr 0.000222 time 0.2674 (0.4449) loss 3.6724 (3.3447) grad_norm 2.5462 (2.2017) [2021-04-16 11:47:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][60/1251] eta 0:08:17 lr 0.000222 time 0.2776 (0.4174) loss 2.9516 (3.3505) grad_norm 2.7020 (2.2105) [2021-04-16 11:47:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][70/1251] eta 0:07:50 lr 0.000222 time 0.2782 (0.3984) loss 3.4500 (3.3209) grad_norm 1.9743 (2.1903) [2021-04-16 11:47:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][80/1251] eta 0:07:29 lr 0.000222 time 0.2667 (0.3836) loss 3.8744 (3.3562) grad_norm 1.9737 (2.1922) [2021-04-16 11:47:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][90/1251] eta 0:07:12 lr 0.000222 time 0.2749 (0.3727) loss 3.1903 (3.3590) grad_norm 2.0036 (2.1797) [2021-04-16 11:47:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][100/1251] eta 0:06:58 lr 0.000222 time 0.2851 (0.3635) loss 3.2782 (3.3448) grad_norm 2.1635 (2.1763) [2021-04-16 11:47:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][110/1251] eta 0:06:46 lr 0.000222 time 0.2885 (0.3561) loss 3.2653 (3.3426) grad_norm 2.2128 (2.1759) [2021-04-16 11:47:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][120/1251] eta 0:06:35 lr 0.000222 time 0.2759 (0.3500) loss 4.0885 (3.3508) grad_norm 2.1621 (2.1686) [2021-04-16 11:47:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][130/1251] eta 0:06:25 lr 0.000222 time 0.2788 (0.3442) loss 3.0004 (3.3116) grad_norm 2.1662 (2.1666) [2021-04-16 11:47:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][140/1251] eta 0:06:19 lr 0.000222 time 0.2647 (0.3420) loss 3.4504 (3.3191) grad_norm 1.9960 (2.1564) [2021-04-16 11:47:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][150/1251] eta 0:06:12 lr 0.000222 time 0.2452 (0.3385) loss 3.5357 (3.3301) grad_norm 2.0762 (2.1575) [2021-04-16 11:47:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][160/1251] eta 0:06:05 lr 0.000222 time 0.2645 (0.3351) loss 3.5754 (3.3273) grad_norm 1.7548 (2.1545) [2021-04-16 11:47:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][170/1251] eta 0:05:58 lr 0.000222 time 0.2574 (0.3318) loss 3.9871 (3.3310) grad_norm 2.0149 (2.1426) [2021-04-16 11:47:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][180/1251] eta 0:05:51 lr 0.000222 time 0.2529 (0.3287) loss 3.5601 (3.3473) grad_norm 2.1012 (2.1428) [2021-04-16 11:47:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][190/1251] eta 0:05:45 lr 0.000222 time 0.2669 (0.3259) loss 3.7189 (3.3459) grad_norm 2.0060 (2.1423) [2021-04-16 11:47:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][200/1251] eta 0:05:39 lr 0.000222 time 0.2654 (0.3233) loss 2.5668 (3.3369) grad_norm 2.0841 (2.1386) [2021-04-16 11:48:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][210/1251] eta 0:05:34 lr 0.000222 time 0.2716 (0.3212) loss 2.7150 (3.3364) grad_norm 2.0989 (2.1332) [2021-04-16 11:48:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][220/1251] eta 0:05:29 lr 0.000222 time 0.2845 (0.3192) loss 3.3813 (3.3335) grad_norm 1.8597 (2.1292) [2021-04-16 11:48:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][230/1251] eta 0:05:23 lr 0.000222 time 0.2666 (0.3172) loss 3.6017 (3.3325) grad_norm 2.0729 (2.1223) [2021-04-16 11:48:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][240/1251] eta 0:05:19 lr 0.000222 time 0.2932 (0.3158) loss 2.4390 (3.3370) grad_norm 2.1742 (2.1200) [2021-04-16 11:48:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][250/1251] eta 0:05:15 lr 0.000222 time 0.2624 (0.3148) loss 2.4582 (3.3278) grad_norm 2.0377 (2.1190) [2021-04-16 11:48:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][260/1251] eta 0:05:10 lr 0.000222 time 0.2863 (0.3132) loss 2.8849 (3.3209) grad_norm 1.8736 (2.1208) [2021-04-16 11:48:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][270/1251] eta 0:05:05 lr 0.000222 time 0.2498 (0.3117) loss 1.9845 (3.3195) grad_norm 2.2940 (2.1293) [2021-04-16 11:48:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][280/1251] eta 0:05:01 lr 0.000222 time 0.2687 (0.3105) loss 3.1075 (3.3216) grad_norm 2.2458 (2.1363) [2021-04-16 11:48:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][290/1251] eta 0:04:57 lr 0.000222 time 0.2752 (0.3095) loss 3.5347 (3.3195) grad_norm 1.8898 (2.1344) [2021-04-16 11:48:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][300/1251] eta 0:04:53 lr 0.000221 time 0.2823 (0.3086) loss 2.3467 (3.3174) grad_norm 2.1205 (2.1327) [2021-04-16 11:48:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][310/1251] eta 0:04:49 lr 0.000221 time 0.2617 (0.3076) loss 3.2642 (3.3193) grad_norm 2.3688 (2.1290) [2021-04-16 11:48:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][320/1251] eta 0:04:45 lr 0.000221 time 0.2973 (0.3071) loss 2.6818 (3.3045) grad_norm 2.2484 (2.1273) [2021-04-16 11:48:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][330/1251] eta 0:04:42 lr 0.000221 time 0.3075 (0.3066) loss 3.8467 (3.3026) grad_norm 1.8629 (2.1262) [2021-04-16 11:48:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][340/1251] eta 0:04:38 lr 0.000221 time 0.2792 (0.3057) loss 2.5861 (3.2978) grad_norm 2.1049 (2.1233) [2021-04-16 11:48:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][350/1251] eta 0:04:34 lr 0.000221 time 0.2753 (0.3047) loss 3.3923 (3.2906) grad_norm 2.1555 (2.1241) [2021-04-16 11:48:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][360/1251] eta 0:04:31 lr 0.000221 time 0.2853 (0.3042) loss 2.5180 (3.2922) grad_norm 2.3441 (2.1239) [2021-04-16 11:48:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][370/1251] eta 0:04:28 lr 0.000221 time 0.2828 (0.3042) loss 3.2610 (3.2850) grad_norm 1.8986 (2.1234) [2021-04-16 11:48:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][380/1251] eta 0:04:24 lr 0.000221 time 0.2625 (0.3035) loss 3.3118 (3.2867) grad_norm 2.0386 (2.1248) [2021-04-16 11:48:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][390/1251] eta 0:04:20 lr 0.000221 time 0.2687 (0.3029) loss 1.9137 (3.2806) grad_norm 2.3365 (2.1245) [2021-04-16 11:48:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][400/1251] eta 0:04:17 lr 0.000221 time 0.2939 (0.3023) loss 3.5445 (3.2822) grad_norm 2.2210 (2.1238) [2021-04-16 11:48:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][410/1251] eta 0:04:13 lr 0.000221 time 0.2985 (0.3020) loss 2.9036 (3.2794) grad_norm 2.5357 (2.1211) [2021-04-16 11:49:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][420/1251] eta 0:04:10 lr 0.000221 time 0.2984 (0.3014) loss 2.8591 (3.2782) grad_norm 2.1259 (2.1194) [2021-04-16 11:49:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][430/1251] eta 0:04:07 lr 0.000221 time 0.2934 (0.3009) loss 4.0303 (3.2750) grad_norm 2.2396 (2.1198) [2021-04-16 11:49:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][440/1251] eta 0:04:03 lr 0.000221 time 0.2791 (0.3005) loss 2.1412 (3.2797) grad_norm 2.2653 (2.1201) [2021-04-16 11:49:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][450/1251] eta 0:04:00 lr 0.000221 time 0.2655 (0.3001) loss 3.3323 (3.2750) grad_norm 2.0662 (2.1221) [2021-04-16 11:49:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][460/1251] eta 0:03:56 lr 0.000221 time 0.2800 (0.2996) loss 3.1077 (3.2756) grad_norm 2.0038 (2.1232) [2021-04-16 11:49:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][470/1251] eta 0:03:53 lr 0.000221 time 0.2691 (0.2991) loss 3.8929 (3.2760) grad_norm 2.0982 (2.1239) [2021-04-16 11:49:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][480/1251] eta 0:03:50 lr 0.000221 time 0.2621 (0.2986) loss 2.5532 (3.2698) grad_norm 2.0704 (2.1213) [2021-04-16 11:49:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][490/1251] eta 0:03:46 lr 0.000221 time 0.2801 (0.2982) loss 2.8000 (3.2657) grad_norm 2.1048 (2.1222) [2021-04-16 11:49:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][500/1251] eta 0:03:43 lr 0.000221 time 0.2800 (0.2977) loss 3.1153 (3.2696) grad_norm 2.3787 (2.1236) [2021-04-16 11:49:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][510/1251] eta 0:03:40 lr 0.000221 time 0.2876 (0.2973) loss 3.7129 (3.2690) grad_norm 1.9441 (2.1242) [2021-04-16 11:49:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][520/1251] eta 0:03:37 lr 0.000221 time 0.2740 (0.2969) loss 3.9418 (3.2670) grad_norm 1.9417 (2.1232) [2021-04-16 11:49:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][530/1251] eta 0:03:33 lr 0.000221 time 0.2738 (0.2965) loss 3.8173 (3.2713) grad_norm 1.9827 (2.1210) [2021-04-16 11:49:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][540/1251] eta 0:03:30 lr 0.000221 time 0.2738 (0.2965) loss 3.7197 (3.2733) grad_norm 2.5476 (2.1207) [2021-04-16 11:49:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][550/1251] eta 0:03:27 lr 0.000221 time 0.2645 (0.2961) loss 3.2854 (3.2720) grad_norm 1.9492 (2.1202) [2021-04-16 11:49:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][560/1251] eta 0:03:24 lr 0.000221 time 0.2891 (0.2958) loss 2.5133 (3.2736) grad_norm 1.7255 (2.1199) [2021-04-16 11:49:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][570/1251] eta 0:03:21 lr 0.000221 time 0.2474 (0.2958) loss 3.4229 (3.2764) grad_norm 1.7367 (2.1183) [2021-04-16 11:49:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][580/1251] eta 0:03:18 lr 0.000221 time 0.2672 (0.2959) loss 3.5397 (3.2747) grad_norm 1.8156 (2.1151) [2021-04-16 11:49:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][590/1251] eta 0:03:15 lr 0.000220 time 0.2709 (0.2955) loss 2.8084 (3.2741) grad_norm 2.1376 (2.1147) [2021-04-16 11:49:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][600/1251] eta 0:03:12 lr 0.000220 time 0.2870 (0.2953) loss 4.1042 (3.2745) grad_norm 3.2485 (2.1178) [2021-04-16 11:49:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][610/1251] eta 0:03:09 lr 0.000220 time 0.2873 (0.2949) loss 2.0940 (3.2752) grad_norm 2.0886 (2.1210) [2021-04-16 11:49:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][620/1251] eta 0:03:05 lr 0.000220 time 0.2679 (0.2946) loss 2.7147 (3.2731) grad_norm 2.0539 (2.1204) [2021-04-16 11:50:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][630/1251] eta 0:03:02 lr 0.000220 time 0.2890 (0.2943) loss 3.8423 (3.2722) grad_norm 2.4229 (2.1213) [2021-04-16 11:50:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][640/1251] eta 0:02:59 lr 0.000220 time 0.2923 (0.2941) loss 3.4368 (3.2698) grad_norm 1.8104 (2.1214) [2021-04-16 11:50:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][650/1251] eta 0:02:56 lr 0.000220 time 0.2911 (0.2939) loss 2.2467 (3.2701) grad_norm 2.0451 (2.1203) [2021-04-16 11:50:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][660/1251] eta 0:02:53 lr 0.000220 time 0.2898 (0.2936) loss 3.3536 (3.2719) grad_norm 1.9564 (2.1186) [2021-04-16 11:50:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][670/1251] eta 0:02:50 lr 0.000220 time 0.2654 (0.2933) loss 3.6816 (3.2769) grad_norm 2.5818 (2.1205) [2021-04-16 11:50:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][680/1251] eta 0:02:47 lr 0.000220 time 0.2836 (0.2932) loss 3.4964 (3.2759) grad_norm 1.7270 (2.1215) [2021-04-16 11:50:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][690/1251] eta 0:02:44 lr 0.000220 time 0.3042 (0.2930) loss 3.5057 (3.2750) grad_norm 1.8361 (2.1215) [2021-04-16 11:50:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][700/1251] eta 0:02:41 lr 0.000220 time 0.2761 (0.2927) loss 3.8431 (3.2797) grad_norm 2.3498 (2.1220) [2021-04-16 11:50:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][710/1251] eta 0:02:38 lr 0.000220 time 0.3275 (0.2926) loss 2.6660 (3.2793) grad_norm 2.0978 (2.1229) [2021-04-16 11:50:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][720/1251] eta 0:02:35 lr 0.000220 time 0.2689 (0.2926) loss 2.9991 (3.2770) grad_norm 1.9780 (2.1235) [2021-04-16 11:50:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][730/1251] eta 0:02:32 lr 0.000220 time 0.2545 (0.2927) loss 2.3262 (3.2760) grad_norm 2.2596 (2.1247) [2021-04-16 11:50:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][740/1251] eta 0:02:29 lr 0.000220 time 0.2544 (0.2925) loss 3.3920 (3.2794) grad_norm 2.7458 (2.1263) [2021-04-16 11:50:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][750/1251] eta 0:02:26 lr 0.000220 time 0.2813 (0.2922) loss 3.7379 (3.2826) grad_norm 2.6655 (2.1303) [2021-04-16 11:50:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][760/1251] eta 0:02:23 lr 0.000220 time 0.2928 (0.2921) loss 4.0108 (3.2841) grad_norm 2.0789 (2.1305) [2021-04-16 11:50:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][770/1251] eta 0:02:20 lr 0.000220 time 0.2888 (0.2919) loss 2.7414 (3.2813) grad_norm 1.8879 (2.1306) [2021-04-16 11:50:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][780/1251] eta 0:02:17 lr 0.000220 time 0.2871 (0.2917) loss 2.9824 (3.2826) grad_norm 2.2697 (2.1305) [2021-04-16 11:50:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][790/1251] eta 0:02:14 lr 0.000220 time 0.3003 (0.2915) loss 3.0107 (3.2855) grad_norm 1.7708 (2.1300) [2021-04-16 11:50:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][800/1251] eta 0:02:11 lr 0.000220 time 0.3054 (0.2915) loss 3.7997 (3.2835) grad_norm 1.9353 (2.1280) [2021-04-16 11:50:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][810/1251] eta 0:02:08 lr 0.000220 time 0.2780 (0.2913) loss 3.7716 (3.2850) grad_norm 2.4715 (2.1279) [2021-04-16 11:50:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][820/1251] eta 0:02:05 lr 0.000220 time 0.3130 (0.2911) loss 4.1269 (3.2830) grad_norm 2.0529 (2.1272) [2021-04-16 11:50:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][830/1251] eta 0:02:02 lr 0.000220 time 0.2781 (0.2910) loss 2.1454 (3.2810) grad_norm 2.0973 (2.1266) [2021-04-16 11:50:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][840/1251] eta 0:01:59 lr 0.000220 time 0.2475 (0.2908) loss 1.7988 (3.2798) grad_norm 2.2304 (2.1270) [2021-04-16 11:51:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][850/1251] eta 0:01:56 lr 0.000220 time 0.2602 (0.2906) loss 3.7636 (3.2806) grad_norm 1.9369 (2.1266) [2021-04-16 11:51:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][860/1251] eta 0:01:53 lr 0.000220 time 0.2759 (0.2904) loss 3.8525 (3.2825) grad_norm 2.1767 (2.1252) [2021-04-16 11:51:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][870/1251] eta 0:01:50 lr 0.000220 time 0.2501 (0.2903) loss 2.8436 (3.2791) grad_norm 1.9720 (2.1245) [2021-04-16 11:51:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][880/1251] eta 0:01:47 lr 0.000220 time 0.2821 (0.2901) loss 3.5033 (3.2793) grad_norm 2.1490 (2.1249) [2021-04-16 11:51:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][890/1251] eta 0:01:44 lr 0.000219 time 0.2799 (0.2902) loss 4.1695 (3.2782) grad_norm 1.9056 (2.1238) [2021-04-16 11:51:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][900/1251] eta 0:01:41 lr 0.000219 time 0.3758 (0.2901) loss 3.5962 (3.2784) grad_norm 2.0846 (nan) [2021-04-16 11:51:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][910/1251] eta 0:01:38 lr 0.000219 time 0.2826 (0.2901) loss 2.1005 (3.2782) grad_norm 1.9806 (nan) [2021-04-16 11:51:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][920/1251] eta 0:01:35 lr 0.000219 time 0.2928 (0.2900) loss 2.6747 (3.2771) grad_norm 1.9401 (nan) [2021-04-16 11:51:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][930/1251] eta 0:01:33 lr 0.000219 time 0.2789 (0.2899) loss 3.5702 (3.2779) grad_norm 1.8467 (nan) [2021-04-16 11:51:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][940/1251] eta 0:01:30 lr 0.000219 time 0.2894 (0.2900) loss 3.1438 (3.2796) grad_norm 2.1303 (nan) [2021-04-16 11:51:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][950/1251] eta 0:01:27 lr 0.000219 time 0.2614 (0.2898) loss 3.6733 (3.2806) grad_norm 2.0662 (nan) [2021-04-16 11:51:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][960/1251] eta 0:01:24 lr 0.000219 time 0.2510 (0.2896) loss 3.2193 (3.2815) grad_norm 2.2245 (nan) [2021-04-16 11:51:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][970/1251] eta 0:01:21 lr 0.000219 time 0.3003 (0.2896) loss 4.2405 (3.2851) grad_norm 2.2477 (nan) [2021-04-16 11:51:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][980/1251] eta 0:01:18 lr 0.000219 time 0.2424 (0.2894) loss 3.4615 (3.2810) grad_norm 1.8995 (nan) [2021-04-16 11:51:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][990/1251] eta 0:01:15 lr 0.000219 time 0.2856 (0.2893) loss 2.7731 (3.2801) grad_norm 2.0464 (nan) [2021-04-16 11:51:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][1000/1251] eta 0:01:12 lr 0.000219 time 0.2536 (0.2891) loss 3.4454 (3.2811) grad_norm 2.0487 (nan) [2021-04-16 11:51:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][1010/1251] eta 0:01:09 lr 0.000219 time 0.2694 (0.2890) loss 2.3745 (3.2806) grad_norm 3.3802 (nan) [2021-04-16 11:51:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][1020/1251] eta 0:01:06 lr 0.000219 time 0.2840 (0.2889) loss 3.7003 (3.2813) grad_norm 2.1983 (nan) [2021-04-16 11:51:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][1030/1251] eta 0:01:03 lr 0.000219 time 0.2611 (0.2888) loss 3.4572 (3.2827) grad_norm 1.9986 (nan) [2021-04-16 11:51:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][1040/1251] eta 0:01:00 lr 0.000219 time 0.2497 (0.2887) loss 3.8682 (3.2811) grad_norm 2.0232 (nan) [2021-04-16 11:51:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][1050/1251] eta 0:00:58 lr 0.000219 time 0.2789 (0.2886) loss 2.6141 (3.2819) grad_norm 1.9028 (nan) [2021-04-16 11:52:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][1060/1251] eta 0:00:55 lr 0.000219 time 0.2891 (0.2885) loss 4.1146 (3.2823) grad_norm 1.8593 (nan) [2021-04-16 11:52:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][1070/1251] eta 0:00:52 lr 0.000219 time 0.2426 (0.2884) loss 2.8268 (3.2833) grad_norm 1.9550 (nan) [2021-04-16 11:52:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][1080/1251] eta 0:00:49 lr 0.000219 time 0.2738 (0.2883) loss 3.4465 (3.2810) grad_norm 2.3277 (nan) [2021-04-16 11:52:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][1090/1251] eta 0:00:46 lr 0.000219 time 0.2821 (0.2882) loss 4.0132 (3.2797) grad_norm 2.0392 (nan) [2021-04-16 11:52:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][1100/1251] eta 0:00:43 lr 0.000219 time 0.2646 (0.2880) loss 2.6848 (3.2759) grad_norm 2.2633 (nan) [2021-04-16 11:52:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][1110/1251] eta 0:00:40 lr 0.000219 time 0.2676 (0.2879) loss 3.2207 (3.2778) grad_norm 2.8832 (nan) [2021-04-16 11:52:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][1120/1251] eta 0:00:37 lr 0.000219 time 0.3016 (0.2879) loss 3.0756 (3.2762) grad_norm 2.7163 (nan) [2021-04-16 11:52:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][1130/1251] eta 0:00:34 lr 0.000219 time 0.2629 (0.2878) loss 3.7242 (3.2755) grad_norm 1.9719 (nan) [2021-04-16 11:52:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][1140/1251] eta 0:00:31 lr 0.000219 time 0.2670 (0.2879) loss 3.0490 (3.2771) grad_norm 2.2725 (nan) [2021-04-16 11:52:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][1150/1251] eta 0:00:29 lr 0.000219 time 0.2579 (0.2878) loss 3.5554 (3.2753) grad_norm 2.3662 (nan) [2021-04-16 11:52:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][1160/1251] eta 0:00:26 lr 0.000219 time 0.2627 (0.2880) loss 3.3793 (3.2755) grad_norm 2.1237 (nan) [2021-04-16 11:52:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][1170/1251] eta 0:00:23 lr 0.000219 time 0.2637 (0.2879) loss 2.7125 (3.2733) grad_norm 2.3239 (nan) [2021-04-16 11:52:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][1180/1251] eta 0:00:20 lr 0.000218 time 0.2715 (0.2878) loss 2.5020 (3.2728) grad_norm 2.2337 (nan) [2021-04-16 11:52:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][1190/1251] eta 0:00:17 lr 0.000218 time 0.2593 (0.2877) loss 3.9839 (3.2731) grad_norm 2.1630 (nan) [2021-04-16 11:52:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][1200/1251] eta 0:00:14 lr 0.000218 time 0.2685 (0.2876) loss 2.6326 (3.2694) grad_norm 2.0559 (nan) [2021-04-16 11:52:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][1210/1251] eta 0:00:11 lr 0.000218 time 0.2572 (0.2875) loss 4.2782 (3.2705) grad_norm 2.0140 (nan) [2021-04-16 11:52:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][1220/1251] eta 0:00:08 lr 0.000218 time 0.2979 (0.2875) loss 3.4029 (3.2723) grad_norm 1.9531 (nan) [2021-04-16 11:52:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][1230/1251] eta 0:00:06 lr 0.000218 time 0.2836 (0.2874) loss 3.2818 (3.2731) grad_norm 1.7484 (nan) [2021-04-16 11:52:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][1240/1251] eta 0:00:03 lr 0.000218 time 0.2499 (0.2872) loss 3.4443 (3.2755) grad_norm 2.1161 (nan) [2021-04-16 11:52:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [208/300][1250/1251] eta 0:00:00 lr 0.000218 time 0.2486 (0.2869) loss 3.1345 (3.2767) grad_norm 1.7303 (nan) [2021-04-16 11:52:57 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 208 training takes 0:06:02 [2021-04-16 11:52:57 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_208.pth saving...... [2021-04-16 11:53:10 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_208.pth saved !!! [2021-04-16 11:53:11 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.238 (1.238) Loss 0.9500 (0.9500) Acc@1 77.930 (77.930) Acc@5 94.922 (94.922) [2021-04-16 11:53:13 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.125 (0.253) Loss 0.9904 (0.9086) Acc@1 77.539 (78.418) Acc@5 93.652 (94.727) [2021-04-16 11:53:15 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.118 (0.249) Loss 0.8954 (0.9075) Acc@1 79.590 (78.632) Acc@5 94.629 (94.713) [2021-04-16 11:53:17 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.150 (0.232) Loss 0.9569 (0.9069) Acc@1 78.613 (78.667) Acc@5 93.555 (94.664) [2021-04-16 11:53:19 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.135 (0.213) Loss 0.9363 (0.9034) Acc@1 78.418 (78.818) Acc@5 93.945 (94.646) [2021-04-16 11:53:26 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 78.708 Acc@5 94.602 [2021-04-16 11:53:26 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 78.7% [2021-04-16 11:53:26 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 78.74% [2021-04-16 11:53:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][0/1251] eta 3:22:22 lr 0.000218 time 9.7059 (9.7059) loss 3.1461 (3.1461) grad_norm 2.0762 (2.0762) [2021-04-16 11:53:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][10/1251] eta 0:23:16 lr 0.000218 time 0.2750 (1.1255) loss 3.9051 (3.3451) grad_norm 2.1372 (2.1947) [2021-04-16 11:53:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][20/1251] eta 0:14:56 lr 0.000218 time 0.2485 (0.7286) loss 3.7617 (3.3664) grad_norm 2.0051 (2.1677) [2021-04-16 11:53:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][30/1251] eta 0:11:56 lr 0.000218 time 0.2981 (0.5871) loss 3.8962 (3.3567) grad_norm 2.1016 (2.1308) [2021-04-16 11:53:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][40/1251] eta 0:10:25 lr 0.000218 time 0.3001 (0.5165) loss 3.4741 (3.3488) grad_norm 1.9899 (2.0908) [2021-04-16 11:53:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][50/1251] eta 0:09:25 lr 0.000218 time 0.2792 (0.4711) loss 3.1162 (3.3374) grad_norm 1.9391 (2.0952) [2021-04-16 11:53:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][60/1251] eta 0:08:44 lr 0.000218 time 0.2643 (0.4405) loss 3.2746 (3.3226) grad_norm 2.3569 (2.1312) [2021-04-16 11:53:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][70/1251] eta 0:08:13 lr 0.000218 time 0.2818 (0.4182) loss 3.6397 (3.3331) grad_norm 2.3068 (2.1449) [2021-04-16 11:53:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][80/1251] eta 0:07:49 lr 0.000218 time 0.2858 (0.4009) loss 3.8322 (3.3568) grad_norm 2.0348 (2.1412) [2021-04-16 11:54:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][90/1251] eta 0:07:32 lr 0.000218 time 0.2643 (0.3895) loss 3.5602 (3.3422) grad_norm 2.1815 (2.1346) [2021-04-16 11:54:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][100/1251] eta 0:07:16 lr 0.000218 time 0.2674 (0.3793) loss 3.3476 (3.3520) grad_norm 2.5128 (2.1371) [2021-04-16 11:54:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][110/1251] eta 0:07:02 lr 0.000218 time 0.2668 (0.3703) loss 2.7983 (3.3562) grad_norm 1.9873 (2.1335) [2021-04-16 11:54:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][120/1251] eta 0:06:50 lr 0.000218 time 0.2795 (0.3630) loss 2.6072 (3.3581) grad_norm 2.1389 (2.1424) [2021-04-16 11:54:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][130/1251] eta 0:06:39 lr 0.000218 time 0.2615 (0.3567) loss 2.4222 (3.3354) grad_norm 2.1356 (2.1422) [2021-04-16 11:54:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][140/1251] eta 0:06:31 lr 0.000218 time 0.2684 (0.3527) loss 2.2181 (3.3140) grad_norm 2.1437 (2.1355) [2021-04-16 11:54:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][150/1251] eta 0:06:24 lr 0.000218 time 0.2912 (0.3491) loss 3.6988 (3.3078) grad_norm 2.1341 (2.1249) [2021-04-16 11:54:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][160/1251] eta 0:06:16 lr 0.000218 time 0.2707 (0.3448) loss 3.1645 (3.2944) grad_norm 2.2831 (2.1251) [2021-04-16 11:54:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][170/1251] eta 0:06:08 lr 0.000218 time 0.3225 (0.3413) loss 3.2863 (3.2969) grad_norm 2.0106 (2.1247) [2021-04-16 11:54:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][180/1251] eta 0:06:01 lr 0.000218 time 0.2780 (0.3377) loss 3.1885 (3.3031) grad_norm 2.7572 (2.1279) [2021-04-16 11:54:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][190/1251] eta 0:05:55 lr 0.000218 time 0.2912 (0.3350) loss 3.4431 (3.3010) grad_norm 2.2959 (2.1255) [2021-04-16 11:54:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][200/1251] eta 0:05:49 lr 0.000218 time 0.2701 (0.3324) loss 3.4906 (3.2854) grad_norm 2.1782 (2.1277) [2021-04-16 11:54:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][210/1251] eta 0:05:43 lr 0.000218 time 0.3002 (0.3299) loss 3.9976 (3.2821) grad_norm 1.9928 (2.1255) [2021-04-16 11:54:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][220/1251] eta 0:05:38 lr 0.000218 time 0.2942 (0.3279) loss 3.5732 (3.2798) grad_norm 2.1067 (2.1227) [2021-04-16 11:54:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][230/1251] eta 0:05:32 lr 0.000217 time 0.2966 (0.3258) loss 2.8362 (3.2808) grad_norm 2.0345 (2.1180) [2021-04-16 11:54:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][240/1251] eta 0:05:27 lr 0.000217 time 0.2697 (0.3238) loss 2.0606 (3.2745) grad_norm 2.1452 (2.1177) [2021-04-16 11:54:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][250/1251] eta 0:05:22 lr 0.000217 time 0.2790 (0.3221) loss 3.9527 (3.2708) grad_norm 2.1664 (2.1117) [2021-04-16 11:54:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][260/1251] eta 0:05:17 lr 0.000217 time 0.2790 (0.3204) loss 3.5873 (3.2649) grad_norm 1.9196 (2.1089) [2021-04-16 11:54:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][270/1251] eta 0:05:12 lr 0.000217 time 0.2695 (0.3191) loss 3.6204 (3.2594) grad_norm 2.1904 (2.1158) [2021-04-16 11:54:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][280/1251] eta 0:05:08 lr 0.000217 time 0.2957 (0.3177) loss 3.9179 (3.2686) grad_norm 2.0244 (2.1192) [2021-04-16 11:54:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][290/1251] eta 0:05:03 lr 0.000217 time 0.2609 (0.3162) loss 3.3512 (3.2745) grad_norm 1.9490 (2.1128) [2021-04-16 11:55:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][300/1251] eta 0:04:59 lr 0.000217 time 0.2688 (0.3152) loss 3.1123 (3.2827) grad_norm 2.0635 (2.1122) [2021-04-16 11:55:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][310/1251] eta 0:04:55 lr 0.000217 time 0.2751 (0.3140) loss 3.3779 (3.2806) grad_norm 1.9571 (2.1111) [2021-04-16 11:55:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][320/1251] eta 0:04:51 lr 0.000217 time 0.2649 (0.3131) loss 4.1306 (3.2820) grad_norm 1.8722 (2.1081) [2021-04-16 11:55:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][330/1251] eta 0:04:47 lr 0.000217 time 0.2872 (0.3126) loss 3.1895 (3.2863) grad_norm 1.9316 (2.1044) [2021-04-16 11:55:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][340/1251] eta 0:04:44 lr 0.000217 time 0.2785 (0.3121) loss 2.9508 (3.2821) grad_norm 1.9596 (2.1036) [2021-04-16 11:55:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][350/1251] eta 0:04:40 lr 0.000217 time 0.2877 (0.3111) loss 3.1251 (3.2807) grad_norm 1.7981 (2.0998) [2021-04-16 11:55:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][360/1251] eta 0:04:36 lr 0.000217 time 0.2831 (0.3107) loss 3.0411 (3.2728) grad_norm 2.1863 (2.1026) [2021-04-16 11:55:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][370/1251] eta 0:04:33 lr 0.000217 time 0.2569 (0.3104) loss 4.0247 (3.2786) grad_norm 1.8182 (2.1025) [2021-04-16 11:55:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][380/1251] eta 0:04:29 lr 0.000217 time 0.2902 (0.3095) loss 2.6367 (3.2769) grad_norm 2.6146 (2.1015) [2021-04-16 11:55:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][390/1251] eta 0:04:25 lr 0.000217 time 0.2682 (0.3088) loss 3.4364 (3.2730) grad_norm 2.1261 (2.0989) [2021-04-16 11:55:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][400/1251] eta 0:04:22 lr 0.000217 time 0.2625 (0.3079) loss 3.5677 (3.2718) grad_norm 2.3944 (2.1007) [2021-04-16 11:55:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][410/1251] eta 0:04:18 lr 0.000217 time 0.2602 (0.3072) loss 2.6050 (3.2669) grad_norm 2.1145 (2.1003) [2021-04-16 11:55:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][420/1251] eta 0:04:14 lr 0.000217 time 0.2681 (0.3064) loss 2.5787 (3.2655) grad_norm 2.1025 (2.1021) [2021-04-16 11:55:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][430/1251] eta 0:04:10 lr 0.000217 time 0.2803 (0.3056) loss 3.6708 (3.2646) grad_norm 2.3381 (2.1053) [2021-04-16 11:55:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][440/1251] eta 0:04:07 lr 0.000217 time 0.3168 (0.3052) loss 3.8306 (3.2696) grad_norm 2.0849 (2.1065) [2021-04-16 11:55:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][450/1251] eta 0:04:04 lr 0.000217 time 0.2762 (0.3046) loss 2.8425 (3.2739) grad_norm 1.8847 (2.1034) [2021-04-16 11:55:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][460/1251] eta 0:04:00 lr 0.000217 time 0.2859 (0.3041) loss 3.8730 (3.2705) grad_norm 1.9699 (2.1005) [2021-04-16 11:55:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][470/1251] eta 0:03:57 lr 0.000217 time 0.2873 (0.3036) loss 3.6037 (3.2645) grad_norm 2.3701 (2.1017) [2021-04-16 11:55:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][480/1251] eta 0:03:53 lr 0.000217 time 0.2927 (0.3032) loss 3.8504 (3.2689) grad_norm 2.1503 (2.1028) [2021-04-16 11:55:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][490/1251] eta 0:03:50 lr 0.000217 time 0.2866 (0.3027) loss 4.0401 (3.2778) grad_norm 2.0196 (2.1043) [2021-04-16 11:55:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][500/1251] eta 0:03:47 lr 0.000217 time 0.2812 (0.3023) loss 3.2743 (3.2707) grad_norm 2.0397 (2.1028) [2021-04-16 11:56:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][510/1251] eta 0:03:43 lr 0.000217 time 0.2626 (0.3017) loss 2.2750 (3.2703) grad_norm 2.2485 (2.0994) [2021-04-16 11:56:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][520/1251] eta 0:03:40 lr 0.000217 time 0.2852 (0.3014) loss 2.2245 (3.2679) grad_norm 2.0880 (2.0996) [2021-04-16 11:56:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][530/1251] eta 0:03:36 lr 0.000216 time 0.2751 (0.3010) loss 2.8672 (3.2703) grad_norm 2.3166 (2.1033) [2021-04-16 11:56:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][540/1251] eta 0:03:33 lr 0.000216 time 0.3075 (0.3008) loss 3.7062 (3.2742) grad_norm 2.0829 (2.1044) [2021-04-16 11:56:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][550/1251] eta 0:03:30 lr 0.000216 time 0.2978 (0.3004) loss 3.2430 (3.2753) grad_norm 2.0860 (2.1061) [2021-04-16 11:56:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][560/1251] eta 0:03:27 lr 0.000216 time 0.2807 (0.3001) loss 3.1198 (3.2814) grad_norm 2.0231 (2.1078) [2021-04-16 11:56:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][570/1251] eta 0:03:24 lr 0.000216 time 0.2663 (0.3000) loss 3.0961 (3.2833) grad_norm 1.9947 (2.1080) [2021-04-16 11:56:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][580/1251] eta 0:03:21 lr 0.000216 time 0.2804 (0.3000) loss 3.7253 (3.2853) grad_norm 2.0268 (2.1070) [2021-04-16 11:56:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][590/1251] eta 0:03:18 lr 0.000216 time 0.2840 (0.2997) loss 3.9862 (3.2885) grad_norm 2.3137 (2.1063) [2021-04-16 11:56:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][600/1251] eta 0:03:14 lr 0.000216 time 0.2903 (0.2994) loss 3.4935 (3.2876) grad_norm 1.9361 (2.1044) [2021-04-16 11:56:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][610/1251] eta 0:03:11 lr 0.000216 time 0.2637 (0.2991) loss 2.2675 (3.2821) grad_norm 2.0350 (2.1034) [2021-04-16 11:56:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][620/1251] eta 0:03:08 lr 0.000216 time 0.3032 (0.2989) loss 3.5300 (3.2808) grad_norm 1.9327 (2.1039) [2021-04-16 11:56:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][630/1251] eta 0:03:05 lr 0.000216 time 0.2790 (0.2986) loss 2.9614 (3.2789) grad_norm 1.9821 (2.1039) [2021-04-16 11:56:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][640/1251] eta 0:03:02 lr 0.000216 time 0.2613 (0.2983) loss 2.1425 (3.2758) grad_norm 1.8097 (2.1027) [2021-04-16 11:56:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][650/1251] eta 0:02:59 lr 0.000216 time 0.2784 (0.2979) loss 3.9013 (3.2812) grad_norm 2.8467 (2.1020) [2021-04-16 11:56:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][660/1251] eta 0:02:55 lr 0.000216 time 0.2800 (0.2976) loss 4.1840 (3.2804) grad_norm 1.7280 (2.1010) [2021-04-16 11:56:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][670/1251] eta 0:02:52 lr 0.000216 time 0.2434 (0.2975) loss 2.4671 (3.2766) grad_norm 1.8607 (2.1009) [2021-04-16 11:56:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][680/1251] eta 0:02:49 lr 0.000216 time 0.2841 (0.2974) loss 2.5862 (3.2735) grad_norm 2.0234 (2.0993) [2021-04-16 11:56:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][690/1251] eta 0:02:46 lr 0.000216 time 0.2742 (0.2972) loss 3.5310 (3.2729) grad_norm 1.8967 (2.0984) [2021-04-16 11:56:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][700/1251] eta 0:02:43 lr 0.000216 time 0.2850 (0.2970) loss 3.3496 (3.2727) grad_norm 2.1669 (2.0969) [2021-04-16 11:56:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][710/1251] eta 0:02:40 lr 0.000216 time 0.2868 (0.2968) loss 3.3260 (3.2748) grad_norm 2.3733 (2.0958) [2021-04-16 11:57:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][720/1251] eta 0:02:37 lr 0.000216 time 0.2741 (0.2969) loss 3.2416 (3.2745) grad_norm 2.3655 (2.0976) [2021-04-16 11:57:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][730/1251] eta 0:02:34 lr 0.000216 time 0.2944 (0.2967) loss 3.2631 (3.2740) grad_norm 2.1676 (2.0997) [2021-04-16 11:57:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][740/1251] eta 0:02:31 lr 0.000216 time 0.2808 (0.2965) loss 3.3442 (3.2776) grad_norm 1.7098 (2.1005) [2021-04-16 11:57:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][750/1251] eta 0:02:28 lr 0.000216 time 0.2755 (0.2963) loss 3.5193 (3.2804) grad_norm 2.3607 (2.1017) [2021-04-16 11:57:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][760/1251] eta 0:02:25 lr 0.000216 time 0.2552 (0.2962) loss 3.3292 (3.2804) grad_norm 1.9446 (2.1008) [2021-04-16 11:57:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][770/1251] eta 0:02:22 lr 0.000216 time 0.2908 (0.2960) loss 3.3875 (3.2771) grad_norm 2.1189 (2.1006) [2021-04-16 11:57:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][780/1251] eta 0:02:19 lr 0.000216 time 0.2681 (0.2958) loss 4.0093 (3.2788) grad_norm 2.2587 (2.1015) [2021-04-16 11:57:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][790/1251] eta 0:02:16 lr 0.000216 time 0.2794 (0.2956) loss 3.6039 (3.2817) grad_norm 1.9909 (2.1031) [2021-04-16 11:57:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][800/1251] eta 0:02:13 lr 0.000216 time 0.2838 (0.2954) loss 3.7781 (3.2822) grad_norm 2.5881 (2.1034) [2021-04-16 11:57:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][810/1251] eta 0:02:10 lr 0.000216 time 0.2737 (0.2952) loss 3.5150 (3.2805) grad_norm 1.8405 (2.1036) [2021-04-16 11:57:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][820/1251] eta 0:02:07 lr 0.000215 time 0.2752 (0.2950) loss 3.6842 (3.2821) grad_norm 2.1103 (2.1034) [2021-04-16 11:57:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][830/1251] eta 0:02:04 lr 0.000215 time 0.2704 (0.2948) loss 3.5648 (3.2798) grad_norm 2.0142 (2.1056) [2021-04-16 11:57:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][840/1251] eta 0:02:01 lr 0.000215 time 0.2454 (0.2947) loss 3.7798 (3.2799) grad_norm 2.0262 (2.1080) [2021-04-16 11:57:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][850/1251] eta 0:01:58 lr 0.000215 time 0.2727 (0.2944) loss 2.6696 (3.2779) grad_norm 2.1507 (2.1103) [2021-04-16 11:57:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][860/1251] eta 0:01:55 lr 0.000215 time 0.3049 (0.2943) loss 2.4205 (3.2774) grad_norm 1.9358 (2.1092) [2021-04-16 11:57:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][870/1251] eta 0:01:52 lr 0.000215 time 0.2925 (0.2941) loss 2.6157 (3.2710) grad_norm 1.9233 (2.1094) [2021-04-16 11:57:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][880/1251] eta 0:01:49 lr 0.000215 time 0.2790 (0.2939) loss 3.6686 (3.2702) grad_norm 1.9797 (2.1098) [2021-04-16 11:57:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][890/1251] eta 0:01:46 lr 0.000215 time 0.2751 (0.2939) loss 2.6794 (3.2659) grad_norm 1.9047 (2.1109) [2021-04-16 11:57:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][900/1251] eta 0:01:43 lr 0.000215 time 0.3855 (0.2938) loss 4.0565 (3.2692) grad_norm 2.1054 (2.1110) [2021-04-16 11:57:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][910/1251] eta 0:01:40 lr 0.000215 time 0.2928 (0.2937) loss 3.7273 (3.2684) grad_norm 2.1145 (2.1103) [2021-04-16 11:57:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][920/1251] eta 0:01:37 lr 0.000215 time 0.2832 (0.2937) loss 3.2101 (3.2680) grad_norm 2.1740 (2.1091) [2021-04-16 11:57:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][930/1251] eta 0:01:34 lr 0.000215 time 0.2910 (0.2938) loss 3.9483 (3.2725) grad_norm 2.4811 (2.1105) [2021-04-16 11:58:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][940/1251] eta 0:01:31 lr 0.000215 time 0.2941 (0.2939) loss 3.7355 (3.2733) grad_norm 2.0428 (2.1111) [2021-04-16 11:58:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][950/1251] eta 0:01:28 lr 0.000215 time 0.2997 (0.2938) loss 2.1990 (3.2733) grad_norm 2.2545 (2.1108) [2021-04-16 11:58:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][960/1251] eta 0:01:25 lr 0.000215 time 0.2723 (0.2937) loss 3.1795 (3.2736) grad_norm 2.3087 (2.1107) [2021-04-16 11:58:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][970/1251] eta 0:01:22 lr 0.000215 time 0.2951 (0.2936) loss 3.3878 (3.2772) grad_norm 2.5113 (2.1122) [2021-04-16 11:58:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][980/1251] eta 0:01:19 lr 0.000215 time 0.2690 (0.2934) loss 3.1042 (3.2785) grad_norm 2.1690 (2.1114) [2021-04-16 11:58:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][990/1251] eta 0:01:16 lr 0.000215 time 0.3149 (0.2933) loss 3.3475 (3.2768) grad_norm 2.2277 (2.1109) [2021-04-16 11:58:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][1000/1251] eta 0:01:13 lr 0.000215 time 0.2804 (0.2932) loss 2.2266 (3.2759) grad_norm 2.0703 (2.1121) [2021-04-16 11:58:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][1010/1251] eta 0:01:10 lr 0.000215 time 0.2877 (0.2932) loss 3.1109 (3.2767) grad_norm 2.5782 (2.1124) [2021-04-16 11:58:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][1020/1251] eta 0:01:07 lr 0.000215 time 0.3023 (0.2931) loss 4.0156 (3.2757) grad_norm 2.0184 (2.1124) [2021-04-16 11:58:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][1030/1251] eta 0:01:04 lr 0.000215 time 0.2635 (0.2929) loss 3.6845 (3.2767) grad_norm 2.0998 (2.1125) [2021-04-16 11:58:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][1040/1251] eta 0:01:01 lr 0.000215 time 0.2736 (0.2928) loss 3.6566 (3.2773) grad_norm 1.7716 (2.1121) [2021-04-16 11:58:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][1050/1251] eta 0:00:58 lr 0.000215 time 0.2687 (0.2927) loss 3.5953 (3.2786) grad_norm 2.4681 (2.1113) [2021-04-16 11:58:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][1060/1251] eta 0:00:55 lr 0.000215 time 0.2807 (0.2925) loss 2.6692 (3.2786) grad_norm 1.9825 (2.1106) [2021-04-16 11:58:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][1070/1251] eta 0:00:52 lr 0.000215 time 0.2884 (0.2924) loss 2.6587 (3.2795) grad_norm 1.9103 (2.1098) [2021-04-16 11:58:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][1080/1251] eta 0:00:49 lr 0.000215 time 0.2814 (0.2922) loss 2.0630 (3.2763) grad_norm 1.9083 (2.1117) [2021-04-16 11:58:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][1090/1251] eta 0:00:47 lr 0.000215 time 0.2831 (0.2921) loss 2.7531 (3.2741) grad_norm 2.0411 (2.1121) [2021-04-16 11:58:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][1100/1251] eta 0:00:44 lr 0.000215 time 0.2701 (0.2920) loss 3.1051 (3.2720) grad_norm 1.9079 (2.1132) [2021-04-16 11:58:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][1110/1251] eta 0:00:41 lr 0.000215 time 0.2730 (0.2919) loss 3.4124 (3.2709) grad_norm 2.5618 (2.1147) [2021-04-16 11:58:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][1120/1251] eta 0:00:38 lr 0.000214 time 0.2762 (0.2919) loss 3.6744 (3.2739) grad_norm 1.9938 (2.1149) [2021-04-16 11:58:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][1130/1251] eta 0:00:35 lr 0.000214 time 0.2780 (0.2918) loss 2.2393 (3.2745) grad_norm 1.8775 (2.1149) [2021-04-16 11:58:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][1140/1251] eta 0:00:32 lr 0.000214 time 0.2790 (0.2917) loss 4.0325 (3.2745) grad_norm 1.9327 (2.1150) [2021-04-16 11:59:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][1150/1251] eta 0:00:29 lr 0.000214 time 0.3013 (0.2918) loss 2.2967 (3.2714) grad_norm 2.3670 (2.1172) [2021-04-16 11:59:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][1160/1251] eta 0:00:26 lr 0.000214 time 0.2615 (0.2917) loss 3.3663 (3.2733) grad_norm 2.1095 (2.1170) [2021-04-16 11:59:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][1170/1251] eta 0:00:23 lr 0.000214 time 0.2767 (0.2915) loss 3.2173 (3.2716) grad_norm 2.0046 (2.1165) [2021-04-16 11:59:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][1180/1251] eta 0:00:20 lr 0.000214 time 0.2766 (0.2914) loss 3.1308 (3.2724) grad_norm 2.3529 (2.1162) [2021-04-16 11:59:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][1190/1251] eta 0:00:17 lr 0.000214 time 0.4319 (0.2914) loss 3.2782 (3.2722) grad_norm 2.5652 (2.1180) [2021-04-16 11:59:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][1200/1251] eta 0:00:14 lr 0.000214 time 0.2694 (0.2913) loss 2.1558 (3.2728) grad_norm 2.0271 (2.1189) [2021-04-16 11:59:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][1210/1251] eta 0:00:11 lr 0.000214 time 0.2762 (0.2912) loss 3.1180 (3.2676) grad_norm 2.5892 (2.1201) [2021-04-16 11:59:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][1220/1251] eta 0:00:09 lr 0.000214 time 0.2766 (0.2911) loss 3.8036 (3.2679) grad_norm 2.7401 (2.1219) [2021-04-16 11:59:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][1230/1251] eta 0:00:06 lr 0.000214 time 0.2653 (0.2909) loss 3.0621 (3.2666) grad_norm 2.1532 (2.1214) [2021-04-16 11:59:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][1240/1251] eta 0:00:03 lr 0.000214 time 0.3604 (0.2909) loss 3.9203 (3.2670) grad_norm 1.9390 (2.1215) [2021-04-16 11:59:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [209/300][1250/1251] eta 0:00:00 lr 0.000214 time 0.2479 (0.2905) loss 3.7612 (3.2694) grad_norm 2.1147 (2.1217) [2021-04-16 11:59:33 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 209 training takes 0:06:07 [2021-04-16 11:59:33 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_209.pth saving...... [2021-04-16 11:59:51 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_209.pth saved !!! [2021-04-16 11:59:52 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.105 (1.105) Loss 0.8667 (0.8667) Acc@1 79.883 (79.883) Acc@5 94.824 (94.824) [2021-04-16 11:59:53 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.157 (0.193) Loss 0.9535 (0.8923) Acc@1 76.953 (79.013) Acc@5 93.164 (94.389) [2021-04-16 11:59:55 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.113 (0.216) Loss 0.9180 (0.8948) Acc@1 78.906 (78.809) Acc@5 94.434 (94.471) [2021-04-16 11:59:58 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.488 (0.240) Loss 0.8863 (0.8948) Acc@1 79.688 (78.805) Acc@5 94.922 (94.493) [2021-04-16 11:59:59 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.205) Loss 0.8927 (0.8998) Acc@1 77.734 (78.673) Acc@5 95.703 (94.553) [2021-04-16 12:00:08 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 78.732 Acc@5 94.634 [2021-04-16 12:00:08 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 78.7% [2021-04-16 12:00:08 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 78.74% [2021-04-16 12:00:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][0/1251] eta 1:58:50 lr 0.000214 time 5.6995 (5.6995) loss 3.8218 (3.8218) grad_norm 2.4804 (2.4804) [2021-04-16 12:00:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][10/1251] eta 0:15:53 lr 0.000214 time 0.2620 (0.7682) loss 3.3146 (3.3119) grad_norm 1.8082 (2.0877) [2021-04-16 12:00:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][20/1251] eta 0:10:57 lr 0.000214 time 0.2559 (0.5345) loss 2.7161 (3.1717) grad_norm 1.9397 (2.0963) [2021-04-16 12:00:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][30/1251] eta 0:09:10 lr 0.000214 time 0.2758 (0.4509) loss 3.0215 (3.1684) grad_norm 1.9843 (2.0858) [2021-04-16 12:00:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][40/1251] eta 0:08:14 lr 0.000214 time 0.2711 (0.4084) loss 3.9748 (3.2540) grad_norm 2.0581 (2.0857) [2021-04-16 12:00:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][50/1251] eta 0:07:41 lr 0.000214 time 0.2963 (0.3843) loss 2.0662 (3.2545) grad_norm 2.5080 (2.0909) [2021-04-16 12:00:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][60/1251] eta 0:07:16 lr 0.000214 time 0.2735 (0.3665) loss 3.7805 (3.3143) grad_norm 2.2277 (2.1604) [2021-04-16 12:00:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][70/1251] eta 0:06:57 lr 0.000214 time 0.2660 (0.3535) loss 2.4331 (3.3084) grad_norm 2.5281 (2.1799) [2021-04-16 12:00:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][80/1251] eta 0:06:44 lr 0.000214 time 0.2872 (0.3453) loss 3.0966 (3.2555) grad_norm 1.9671 (2.1906) [2021-04-16 12:00:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][90/1251] eta 0:06:33 lr 0.000214 time 0.2826 (0.3389) loss 2.9021 (3.2411) grad_norm 2.0371 (2.1748) [2021-04-16 12:00:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][100/1251] eta 0:06:25 lr 0.000214 time 0.2649 (0.3348) loss 3.3239 (3.2512) grad_norm 2.0153 (2.1642) [2021-04-16 12:00:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][110/1251] eta 0:06:16 lr 0.000214 time 0.2759 (0.3296) loss 3.7108 (3.2735) grad_norm 2.6210 (2.1586) [2021-04-16 12:00:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][120/1251] eta 0:06:09 lr 0.000214 time 0.2677 (0.3266) loss 3.2422 (3.2693) grad_norm 2.5077 (2.1538) [2021-04-16 12:00:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][130/1251] eta 0:06:02 lr 0.000214 time 0.2979 (0.3231) loss 3.9026 (3.2792) grad_norm 2.2089 (2.1413) [2021-04-16 12:00:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][140/1251] eta 0:05:57 lr 0.000214 time 0.2561 (0.3221) loss 3.3392 (3.2888) grad_norm 2.0979 (2.1470) [2021-04-16 12:00:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][150/1251] eta 0:05:51 lr 0.000214 time 0.2867 (0.3188) loss 3.6862 (3.2961) grad_norm 2.2329 (2.1507) [2021-04-16 12:00:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][160/1251] eta 0:05:44 lr 0.000214 time 0.2549 (0.3159) loss 3.4943 (3.2797) grad_norm 1.9855 (2.1456) [2021-04-16 12:01:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][170/1251] eta 0:05:39 lr 0.000213 time 0.2952 (0.3140) loss 3.0209 (3.2639) grad_norm 1.8680 (2.1376) [2021-04-16 12:01:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][180/1251] eta 0:05:34 lr 0.000213 time 0.2793 (0.3119) loss 3.6576 (3.2715) grad_norm 2.1221 (2.1289) [2021-04-16 12:01:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][190/1251] eta 0:05:29 lr 0.000213 time 0.2734 (0.3102) loss 3.8711 (3.2686) grad_norm 2.1074 (2.1313) [2021-04-16 12:01:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][200/1251] eta 0:05:24 lr 0.000213 time 0.2632 (0.3086) loss 3.8544 (3.2740) grad_norm 2.6651 (2.1371) [2021-04-16 12:01:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][210/1251] eta 0:05:19 lr 0.000213 time 0.2817 (0.3072) loss 2.3038 (3.2838) grad_norm 2.2986 (2.1422) [2021-04-16 12:01:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][220/1251] eta 0:05:15 lr 0.000213 time 0.2571 (0.3056) loss 3.2664 (3.2871) grad_norm 2.2152 (2.1456) [2021-04-16 12:01:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][230/1251] eta 0:05:10 lr 0.000213 time 0.2707 (0.3042) loss 3.4181 (3.2852) grad_norm 1.8793 (2.1430) [2021-04-16 12:01:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][240/1251] eta 0:05:06 lr 0.000213 time 0.2893 (0.3031) loss 2.8965 (3.2770) grad_norm 2.0101 (2.1373) [2021-04-16 12:01:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][250/1251] eta 0:05:02 lr 0.000213 time 0.2914 (0.3019) loss 3.6444 (3.2755) grad_norm 2.1028 (2.1402) [2021-04-16 12:01:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][260/1251] eta 0:04:58 lr 0.000213 time 0.2689 (0.3009) loss 3.0799 (3.2705) grad_norm 2.0179 (2.1373) [2021-04-16 12:01:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][270/1251] eta 0:04:54 lr 0.000213 time 0.2783 (0.2999) loss 3.6248 (3.2722) grad_norm 2.6570 (2.1410) [2021-04-16 12:01:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][280/1251] eta 0:04:50 lr 0.000213 time 0.2751 (0.2993) loss 3.8001 (3.2868) grad_norm 2.4724 (2.1412) [2021-04-16 12:01:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][290/1251] eta 0:04:46 lr 0.000213 time 0.2893 (0.2986) loss 3.4475 (3.2842) grad_norm 2.2398 (2.1396) [2021-04-16 12:01:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][300/1251] eta 0:04:43 lr 0.000213 time 0.2728 (0.2981) loss 2.4273 (3.2837) grad_norm 2.6263 (2.1445) [2021-04-16 12:01:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][310/1251] eta 0:04:39 lr 0.000213 time 0.2466 (0.2974) loss 3.7490 (3.2827) grad_norm 1.9764 (2.1428) [2021-04-16 12:01:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][320/1251] eta 0:04:36 lr 0.000213 time 0.2835 (0.2973) loss 3.6077 (3.2841) grad_norm 2.1620 (2.1454) [2021-04-16 12:01:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][330/1251] eta 0:04:33 lr 0.000213 time 0.2615 (0.2967) loss 3.5507 (3.2847) grad_norm 1.8596 (2.1418) [2021-04-16 12:01:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][340/1251] eta 0:04:30 lr 0.000213 time 0.2806 (0.2964) loss 3.4273 (3.2857) grad_norm 2.7845 (2.1426) [2021-04-16 12:01:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][350/1251] eta 0:04:26 lr 0.000213 time 0.2727 (0.2958) loss 2.0216 (3.2809) grad_norm 3.1558 (2.1441) [2021-04-16 12:01:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][360/1251] eta 0:04:23 lr 0.000213 time 0.2888 (0.2957) loss 2.5178 (3.2692) grad_norm 1.9315 (2.1427) [2021-04-16 12:01:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][370/1251] eta 0:04:20 lr 0.000213 time 0.2742 (0.2954) loss 3.5996 (3.2677) grad_norm 2.1333 (2.1405) [2021-04-16 12:02:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][380/1251] eta 0:04:16 lr 0.000213 time 0.2562 (0.2950) loss 2.7767 (3.2779) grad_norm 2.0899 (2.1396) [2021-04-16 12:02:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][390/1251] eta 0:04:13 lr 0.000213 time 0.2513 (0.2950) loss 2.5529 (3.2764) grad_norm 2.1698 (2.1370) [2021-04-16 12:02:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][400/1251] eta 0:04:10 lr 0.000213 time 0.2607 (0.2946) loss 3.5128 (3.2792) grad_norm 1.9445 (2.1355) [2021-04-16 12:02:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][410/1251] eta 0:04:07 lr 0.000213 time 0.2855 (0.2941) loss 3.9031 (3.2841) grad_norm 2.0189 (2.1382) [2021-04-16 12:02:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][420/1251] eta 0:04:03 lr 0.000213 time 0.2691 (0.2936) loss 2.7230 (3.2780) grad_norm 2.0628 (2.1369) [2021-04-16 12:02:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][430/1251] eta 0:04:00 lr 0.000213 time 0.2708 (0.2933) loss 3.2779 (3.2809) grad_norm 2.1691 (2.1369) [2021-04-16 12:02:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][440/1251] eta 0:03:57 lr 0.000213 time 0.2444 (0.2929) loss 3.6411 (3.2767) grad_norm 2.0627 (2.1361) [2021-04-16 12:02:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][450/1251] eta 0:03:54 lr 0.000213 time 0.2593 (0.2929) loss 3.1343 (3.2701) grad_norm 3.1002 (2.1408) [2021-04-16 12:02:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][460/1251] eta 0:03:51 lr 0.000213 time 0.3079 (0.2926) loss 2.9190 (3.2685) grad_norm 1.9261 (2.1395) [2021-04-16 12:02:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][470/1251] eta 0:03:48 lr 0.000212 time 0.2656 (0.2922) loss 3.4514 (3.2680) grad_norm 2.8669 (2.1387) [2021-04-16 12:02:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][480/1251] eta 0:03:45 lr 0.000212 time 0.2637 (0.2922) loss 2.1654 (3.2738) grad_norm 2.2124 (2.1376) [2021-04-16 12:02:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][490/1251] eta 0:03:42 lr 0.000212 time 0.2662 (0.2919) loss 2.3161 (3.2690) grad_norm 2.1456 (2.1394) [2021-04-16 12:02:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][500/1251] eta 0:03:38 lr 0.000212 time 0.2523 (0.2916) loss 3.2527 (3.2705) grad_norm 1.9298 (2.1399) [2021-04-16 12:02:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][510/1251] eta 0:03:35 lr 0.000212 time 0.2796 (0.2913) loss 4.0449 (3.2724) grad_norm 2.0495 (2.1386) [2021-04-16 12:02:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][520/1251] eta 0:03:32 lr 0.000212 time 0.2772 (0.2910) loss 2.8033 (3.2743) grad_norm 2.0870 (2.1392) [2021-04-16 12:02:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][530/1251] eta 0:03:29 lr 0.000212 time 0.2535 (0.2906) loss 3.7535 (3.2787) grad_norm 2.0464 (2.1382) [2021-04-16 12:02:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][540/1251] eta 0:03:26 lr 0.000212 time 0.2711 (0.2903) loss 3.2660 (3.2760) grad_norm 2.2641 (2.1404) [2021-04-16 12:02:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][550/1251] eta 0:03:23 lr 0.000212 time 0.2734 (0.2901) loss 2.4859 (3.2746) grad_norm 1.9437 (2.1415) [2021-04-16 12:02:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][560/1251] eta 0:03:20 lr 0.000212 time 0.2887 (0.2899) loss 3.6836 (3.2759) grad_norm 2.1744 (2.1407) [2021-04-16 12:02:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][570/1251] eta 0:03:17 lr 0.000212 time 0.2697 (0.2896) loss 3.0363 (3.2730) grad_norm 2.1216 (2.1405) [2021-04-16 12:02:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][580/1251] eta 0:03:14 lr 0.000212 time 0.2630 (0.2901) loss 3.3333 (3.2689) grad_norm 2.3243 (2.1404) [2021-04-16 12:03:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][590/1251] eta 0:03:11 lr 0.000212 time 0.2790 (0.2898) loss 3.9157 (3.2757) grad_norm 2.2387 (2.1402) [2021-04-16 12:03:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][600/1251] eta 0:03:08 lr 0.000212 time 0.2835 (0.2896) loss 3.4490 (3.2698) grad_norm 1.9573 (2.1405) [2021-04-16 12:03:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][610/1251] eta 0:03:05 lr 0.000212 time 0.2466 (0.2892) loss 3.3670 (3.2685) grad_norm 1.9691 (2.1407) [2021-04-16 12:03:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][620/1251] eta 0:03:02 lr 0.000212 time 0.2438 (0.2890) loss 3.6130 (3.2696) grad_norm 2.2124 (2.1455) [2021-04-16 12:03:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][630/1251] eta 0:02:59 lr 0.000212 time 0.2840 (0.2888) loss 3.0191 (3.2716) grad_norm 2.0362 (2.1469) [2021-04-16 12:03:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][640/1251] eta 0:02:56 lr 0.000212 time 0.2745 (0.2886) loss 3.7054 (3.2740) grad_norm 1.9896 (2.1481) [2021-04-16 12:03:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][650/1251] eta 0:02:53 lr 0.000212 time 0.2959 (0.2884) loss 3.9486 (3.2768) grad_norm 2.0206 (2.1506) [2021-04-16 12:03:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][660/1251] eta 0:02:50 lr 0.000212 time 0.2797 (0.2884) loss 3.4328 (3.2746) grad_norm 1.8222 (2.1483) [2021-04-16 12:03:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][670/1251] eta 0:02:47 lr 0.000212 time 0.2712 (0.2883) loss 3.6216 (3.2768) grad_norm 2.9195 (2.1479) [2021-04-16 12:03:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][680/1251] eta 0:02:44 lr 0.000212 time 0.3063 (0.2884) loss 3.0679 (3.2779) grad_norm 2.0084 (2.1466) [2021-04-16 12:03:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][690/1251] eta 0:02:41 lr 0.000212 time 0.2964 (0.2885) loss 3.1468 (3.2802) grad_norm 2.0106 (2.1450) [2021-04-16 12:03:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][700/1251] eta 0:02:38 lr 0.000212 time 0.2778 (0.2884) loss 2.7749 (3.2775) grad_norm 2.0796 (2.1459) [2021-04-16 12:03:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][710/1251] eta 0:02:35 lr 0.000212 time 0.2830 (0.2882) loss 3.6596 (3.2784) grad_norm 1.8314 (2.1448) [2021-04-16 12:03:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][720/1251] eta 0:02:33 lr 0.000212 time 0.2572 (0.2882) loss 3.6516 (3.2768) grad_norm 2.4002 (2.1466) [2021-04-16 12:03:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][730/1251] eta 0:02:30 lr 0.000212 time 0.2705 (0.2880) loss 3.9044 (3.2786) grad_norm 2.1584 (2.1465) [2021-04-16 12:03:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][740/1251] eta 0:02:27 lr 0.000212 time 0.2983 (0.2879) loss 3.8839 (3.2781) grad_norm 2.2396 (2.1476) [2021-04-16 12:03:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][750/1251] eta 0:02:24 lr 0.000212 time 0.2933 (0.2878) loss 3.6282 (3.2826) grad_norm 2.3255 (2.1497) [2021-04-16 12:03:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][760/1251] eta 0:02:21 lr 0.000212 time 0.2975 (0.2879) loss 3.3016 (3.2817) grad_norm 2.2006 (2.1495) [2021-04-16 12:03:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][770/1251] eta 0:02:18 lr 0.000211 time 0.2969 (0.2880) loss 3.3164 (3.2813) grad_norm 1.8745 (2.1495) [2021-04-16 12:03:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][780/1251] eta 0:02:15 lr 0.000211 time 0.2711 (0.2879) loss 3.2334 (3.2825) grad_norm 2.0138 (2.1490) [2021-04-16 12:03:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][790/1251] eta 0:02:12 lr 0.000211 time 0.2793 (0.2878) loss 2.9293 (3.2840) grad_norm 2.0588 (2.1475) [2021-04-16 12:03:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][800/1251] eta 0:02:09 lr 0.000211 time 0.2762 (0.2877) loss 3.0091 (3.2799) grad_norm 2.1377 (2.1470) [2021-04-16 12:04:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][810/1251] eta 0:02:06 lr 0.000211 time 0.2490 (0.2875) loss 3.3834 (3.2803) grad_norm 2.2958 (2.1476) [2021-04-16 12:04:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][820/1251] eta 0:02:03 lr 0.000211 time 0.2934 (0.2875) loss 3.2356 (3.2814) grad_norm 2.2774 (2.1483) [2021-04-16 12:04:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][830/1251] eta 0:02:00 lr 0.000211 time 0.2933 (0.2874) loss 2.7977 (3.2802) grad_norm 2.1351 (2.1465) [2021-04-16 12:04:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][840/1251] eta 0:01:58 lr 0.000211 time 0.2703 (0.2873) loss 3.6408 (3.2809) grad_norm 1.8597 (2.1444) [2021-04-16 12:04:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][850/1251] eta 0:01:55 lr 0.000211 time 0.2855 (0.2872) loss 2.5031 (3.2779) grad_norm 2.2836 (2.1434) [2021-04-16 12:04:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][860/1251] eta 0:01:52 lr 0.000211 time 0.2668 (0.2871) loss 3.4881 (3.2753) grad_norm 2.1515 (2.1438) [2021-04-16 12:04:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][870/1251] eta 0:01:49 lr 0.000211 time 0.2526 (0.2870) loss 3.4900 (3.2736) grad_norm 2.0867 (2.1427) [2021-04-16 12:04:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][880/1251] eta 0:01:46 lr 0.000211 time 0.2893 (0.2869) loss 3.3403 (3.2724) grad_norm 2.1076 (2.1421) [2021-04-16 12:04:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][890/1251] eta 0:01:43 lr 0.000211 time 0.2789 (0.2868) loss 2.9350 (3.2716) grad_norm 2.2746 (2.1457) [2021-04-16 12:04:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][900/1251] eta 0:01:40 lr 0.000211 time 0.2820 (0.2868) loss 2.8037 (3.2711) grad_norm 2.1913 (2.1463) [2021-04-16 12:04:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][910/1251] eta 0:01:37 lr 0.000211 time 0.2656 (0.2866) loss 3.0129 (3.2715) grad_norm 2.0434 (2.1454) [2021-04-16 12:04:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][920/1251] eta 0:01:34 lr 0.000211 time 0.2875 (0.2867) loss 3.4228 (3.2709) grad_norm 1.9844 (2.1445) [2021-04-16 12:04:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][930/1251] eta 0:01:31 lr 0.000211 time 0.2753 (0.2866) loss 3.3260 (3.2734) grad_norm 2.4779 (2.1452) [2021-04-16 12:04:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][940/1251] eta 0:01:29 lr 0.000211 time 0.2847 (0.2867) loss 3.4724 (3.2733) grad_norm 1.9494 (2.1451) [2021-04-16 12:04:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][950/1251] eta 0:01:26 lr 0.000211 time 0.3061 (0.2866) loss 3.3493 (3.2752) grad_norm 1.8464 (2.1442) [2021-04-16 12:04:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][960/1251] eta 0:01:23 lr 0.000211 time 0.2819 (0.2865) loss 3.9782 (3.2754) grad_norm 1.9100 (2.1436) [2021-04-16 12:04:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][970/1251] eta 0:01:20 lr 0.000211 time 0.2658 (0.2866) loss 3.8336 (3.2779) grad_norm 2.1814 (2.1436) [2021-04-16 12:04:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][980/1251] eta 0:01:17 lr 0.000211 time 0.2647 (0.2864) loss 3.2757 (3.2793) grad_norm 2.1794 (2.1432) [2021-04-16 12:04:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][990/1251] eta 0:01:14 lr 0.000211 time 0.2978 (0.2863) loss 3.6687 (3.2809) grad_norm 1.9209 (2.1424) [2021-04-16 12:04:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][1000/1251] eta 0:01:11 lr 0.000211 time 0.2604 (0.2862) loss 3.8389 (3.2815) grad_norm 2.2911 (2.1412) [2021-04-16 12:04:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][1010/1251] eta 0:01:08 lr 0.000211 time 0.2654 (0.2861) loss 3.3281 (3.2844) grad_norm 1.8854 (2.1404) [2021-04-16 12:05:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][1020/1251] eta 0:01:06 lr 0.000211 time 0.2738 (0.2860) loss 3.4946 (3.2854) grad_norm 2.0950 (2.1396) [2021-04-16 12:05:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][1030/1251] eta 0:01:03 lr 0.000211 time 0.2936 (0.2860) loss 3.6330 (3.2851) grad_norm 2.0374 (2.1389) [2021-04-16 12:05:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][1040/1251] eta 0:01:00 lr 0.000211 time 0.2447 (0.2860) loss 2.8889 (3.2857) grad_norm 2.3498 (2.1389) [2021-04-16 12:05:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][1050/1251] eta 0:00:57 lr 0.000211 time 0.4888 (0.2861) loss 3.0317 (3.2856) grad_norm 2.2790 (2.1379) [2021-04-16 12:05:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][1060/1251] eta 0:00:54 lr 0.000211 time 0.3006 (0.2860) loss 3.3541 (3.2851) grad_norm 1.9373 (2.1378) [2021-04-16 12:05:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][1070/1251] eta 0:00:51 lr 0.000210 time 0.2767 (0.2859) loss 3.9326 (3.2847) grad_norm 2.3995 (2.1375) [2021-04-16 12:05:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][1080/1251] eta 0:00:48 lr 0.000210 time 0.2740 (0.2859) loss 2.5381 (3.2810) grad_norm 2.0554 (2.1386) [2021-04-16 12:05:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][1090/1251] eta 0:00:46 lr 0.000210 time 0.2769 (0.2857) loss 3.9881 (3.2813) grad_norm 2.0514 (2.1392) [2021-04-16 12:05:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][1100/1251] eta 0:00:43 lr 0.000210 time 0.3053 (0.2856) loss 2.7258 (3.2824) grad_norm 2.0347 (2.1394) [2021-04-16 12:05:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][1110/1251] eta 0:00:40 lr 0.000210 time 0.2855 (0.2856) loss 3.1545 (3.2825) grad_norm 2.4066 (2.1390) [2021-04-16 12:05:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][1120/1251] eta 0:00:37 lr 0.000210 time 0.2546 (0.2855) loss 2.4719 (3.2809) grad_norm 2.4053 (2.1396) [2021-04-16 12:05:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][1130/1251] eta 0:00:34 lr 0.000210 time 0.2589 (0.2854) loss 2.4592 (3.2789) grad_norm 2.0088 (2.1396) [2021-04-16 12:05:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][1140/1251] eta 0:00:31 lr 0.000210 time 0.2823 (0.2854) loss 3.0864 (3.2791) grad_norm 2.5476 (2.1415) [2021-04-16 12:05:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][1150/1251] eta 0:00:28 lr 0.000210 time 0.2659 (0.2854) loss 3.9938 (3.2811) grad_norm 2.4518 (2.1414) [2021-04-16 12:05:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][1160/1251] eta 0:00:25 lr 0.000210 time 0.2576 (0.2854) loss 2.8731 (3.2799) grad_norm 2.3128 (2.1414) [2021-04-16 12:05:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][1170/1251] eta 0:00:23 lr 0.000210 time 0.2717 (0.2853) loss 3.4001 (3.2798) grad_norm 2.2344 (2.1411) [2021-04-16 12:05:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][1180/1251] eta 0:00:20 lr 0.000210 time 0.2667 (0.2852) loss 2.6961 (3.2788) grad_norm 2.0926 (2.1412) [2021-04-16 12:05:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][1190/1251] eta 0:00:17 lr 0.000210 time 0.2862 (0.2852) loss 3.4655 (3.2797) grad_norm 1.9375 (2.1412) [2021-04-16 12:05:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][1200/1251] eta 0:00:14 lr 0.000210 time 0.2655 (0.2852) loss 3.3001 (3.2800) grad_norm 2.3388 (2.1410) [2021-04-16 12:05:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][1210/1251] eta 0:00:11 lr 0.000210 time 0.2940 (0.2851) loss 4.1857 (3.2806) grad_norm 2.0295 (2.1408) [2021-04-16 12:05:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][1220/1251] eta 0:00:08 lr 0.000210 time 0.3067 (0.2851) loss 3.8210 (3.2812) grad_norm 2.2062 (2.1408) [2021-04-16 12:05:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][1230/1251] eta 0:00:05 lr 0.000210 time 0.2510 (0.2850) loss 1.9615 (3.2793) grad_norm 1.9035 (2.1393) [2021-04-16 12:06:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][1240/1251] eta 0:00:03 lr 0.000210 time 0.2596 (0.2849) loss 2.5101 (3.2786) grad_norm 2.3920 (2.1394) [2021-04-16 12:06:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [210/300][1250/1251] eta 0:00:00 lr 0.000210 time 0.2486 (0.2846) loss 3.8169 (3.2792) grad_norm 2.4201 (2.1397) [2021-04-16 12:06:11 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 210 training takes 0:06:02 [2021-04-16 12:06:11 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_210.pth saving...... [2021-04-16 12:06:27 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_210.pth saved !!! [2021-04-16 12:06:28 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.125 (1.125) Loss 0.9661 (0.9661) Acc@1 77.832 (77.832) Acc@5 93.457 (93.457) [2021-04-16 12:06:29 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.196 (0.224) Loss 0.9395 (0.8964) Acc@1 78.809 (78.880) Acc@5 93.848 (94.531) [2021-04-16 12:06:31 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.161 (0.233) Loss 0.8873 (0.8930) Acc@1 78.809 (78.767) Acc@5 94.336 (94.489) [2021-04-16 12:06:34 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.392 (0.246) Loss 0.9675 (0.8967) Acc@1 78.320 (78.730) Acc@5 93.262 (94.355) [2021-04-16 12:06:35 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.213) Loss 0.9624 (0.8939) Acc@1 77.051 (78.668) Acc@5 94.238 (94.522) [2021-04-16 12:06:43 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 78.750 Acc@5 94.512 [2021-04-16 12:06:43 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 78.8% [2021-04-16 12:06:43 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 78.75% [2021-04-16 12:06:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][0/1251] eta 5:22:10 lr 0.000210 time 15.4517 (15.4517) loss 3.5699 (3.5699) grad_norm 1.8801 (1.8801) [2021-04-16 12:07:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][10/1251] eta 0:34:09 lr 0.000210 time 0.2875 (1.6519) loss 2.4697 (3.3223) grad_norm 2.0227 (2.0804) [2021-04-16 12:07:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][20/1251] eta 0:20:27 lr 0.000210 time 0.2897 (0.9972) loss 3.2416 (3.0676) grad_norm 2.2394 (2.0800) [2021-04-16 12:07:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][30/1251] eta 0:15:33 lr 0.000210 time 0.2710 (0.7641) loss 3.0018 (3.2120) grad_norm 1.7975 (2.0546) [2021-04-16 12:07:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][40/1251] eta 0:13:06 lr 0.000210 time 0.4314 (0.6492) loss 3.5872 (3.2547) grad_norm 2.0401 (2.0511) [2021-04-16 12:07:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][50/1251] eta 0:11:29 lr 0.000210 time 0.2834 (0.5743) loss 3.0253 (3.2148) grad_norm 2.3462 (2.0727) [2021-04-16 12:07:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][60/1251] eta 0:10:25 lr 0.000210 time 0.2774 (0.5249) loss 2.5374 (3.2115) grad_norm 1.9985 (2.0738) [2021-04-16 12:07:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][70/1251] eta 0:09:39 lr 0.000210 time 0.2811 (0.4904) loss 3.7884 (3.2335) grad_norm 2.1488 (2.0797) [2021-04-16 12:07:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][80/1251] eta 0:09:03 lr 0.000210 time 0.2829 (0.4639) loss 2.9989 (3.2508) grad_norm 2.4435 (2.0956) [2021-04-16 12:07:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][90/1251] eta 0:08:34 lr 0.000210 time 0.3168 (0.4433) loss 3.1026 (3.2603) grad_norm 2.1674 (2.1103) [2021-04-16 12:07:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][100/1251] eta 0:08:12 lr 0.000210 time 0.2855 (0.4277) loss 2.6561 (3.2534) grad_norm 2.5323 (2.1117) [2021-04-16 12:07:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][110/1251] eta 0:07:52 lr 0.000210 time 0.2764 (0.4144) loss 2.5620 (3.2494) grad_norm 2.0161 (2.1125) [2021-04-16 12:07:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][120/1251] eta 0:07:37 lr 0.000209 time 0.2729 (0.4041) loss 3.6527 (3.2321) grad_norm 1.9335 (2.1115) [2021-04-16 12:07:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][130/1251] eta 0:07:22 lr 0.000209 time 0.2674 (0.3944) loss 4.0481 (3.2426) grad_norm 2.1487 (2.1093) [2021-04-16 12:07:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][140/1251] eta 0:07:08 lr 0.000209 time 0.2440 (0.3859) loss 3.5070 (3.2509) grad_norm 1.9824 (2.1100) [2021-04-16 12:07:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][150/1251] eta 0:06:57 lr 0.000209 time 0.2607 (0.3795) loss 3.6583 (3.2565) grad_norm 2.0204 (2.1069) [2021-04-16 12:07:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][160/1251] eta 0:06:47 lr 0.000209 time 0.2878 (0.3738) loss 3.2336 (3.2318) grad_norm 1.9294 (2.1068) [2021-04-16 12:07:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][170/1251] eta 0:06:38 lr 0.000209 time 0.2739 (0.3684) loss 3.0090 (3.2240) grad_norm 2.0170 (2.1063) [2021-04-16 12:07:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][180/1251] eta 0:06:29 lr 0.000209 time 0.2609 (0.3640) loss 3.8466 (3.2343) grad_norm 2.0904 (2.1130) [2021-04-16 12:07:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][190/1251] eta 0:06:21 lr 0.000209 time 0.2856 (0.3595) loss 3.6807 (3.2445) grad_norm 2.3014 (2.1239) [2021-04-16 12:07:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][200/1251] eta 0:06:13 lr 0.000209 time 0.2738 (0.3553) loss 2.1595 (3.2364) grad_norm 2.8253 (2.1321) [2021-04-16 12:07:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][210/1251] eta 0:06:05 lr 0.000209 time 0.2762 (0.3515) loss 2.8718 (3.2300) grad_norm 2.2433 (2.1357) [2021-04-16 12:08:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][220/1251] eta 0:05:59 lr 0.000209 time 0.2932 (0.3482) loss 2.5077 (3.2340) grad_norm 2.3657 (2.1356) [2021-04-16 12:08:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][230/1251] eta 0:05:52 lr 0.000209 time 0.3169 (0.3452) loss 2.7013 (3.2354) grad_norm 2.5307 (2.1449) [2021-04-16 12:08:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][240/1251] eta 0:05:47 lr 0.000209 time 0.2608 (0.3433) loss 3.4413 (3.2317) grad_norm 2.0719 (2.1484) [2021-04-16 12:08:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][250/1251] eta 0:05:40 lr 0.000209 time 0.2685 (0.3406) loss 3.1051 (3.2184) grad_norm 2.1266 (2.1505) [2021-04-16 12:08:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][260/1251] eta 0:05:35 lr 0.000209 time 0.4642 (0.3388) loss 2.1958 (3.2081) grad_norm 2.1495 (2.1488) [2021-04-16 12:08:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][270/1251] eta 0:05:29 lr 0.000209 time 0.2639 (0.3363) loss 2.8844 (3.2078) grad_norm 2.2605 (2.1462) [2021-04-16 12:08:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][280/1251] eta 0:05:24 lr 0.000209 time 0.2861 (0.3342) loss 3.4631 (3.2222) grad_norm 2.1985 (2.1475) [2021-04-16 12:08:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][290/1251] eta 0:05:19 lr 0.000209 time 0.2916 (0.3322) loss 2.3282 (3.2097) grad_norm 1.9966 (2.1449) [2021-04-16 12:08:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][300/1251] eta 0:05:14 lr 0.000209 time 0.2579 (0.3303) loss 3.9118 (3.2136) grad_norm 2.2051 (2.1415) [2021-04-16 12:08:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][310/1251] eta 0:05:09 lr 0.000209 time 0.2542 (0.3285) loss 3.2278 (3.2244) grad_norm 1.7862 (2.1390) [2021-04-16 12:08:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][320/1251] eta 0:05:04 lr 0.000209 time 0.2830 (0.3270) loss 3.0664 (3.2250) grad_norm 2.2770 (2.1439) [2021-04-16 12:08:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][330/1251] eta 0:04:59 lr 0.000209 time 0.2677 (0.3254) loss 3.5345 (3.2264) grad_norm 2.1531 (2.1402) [2021-04-16 12:08:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][340/1251] eta 0:04:55 lr 0.000209 time 0.2827 (0.3238) loss 2.5103 (3.2174) grad_norm 1.8471 (2.1391) [2021-04-16 12:08:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][350/1251] eta 0:04:50 lr 0.000209 time 0.2780 (0.3224) loss 3.2727 (3.2232) grad_norm 2.3090 (2.1376) [2021-04-16 12:08:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][360/1251] eta 0:04:46 lr 0.000209 time 0.2541 (0.3219) loss 2.4517 (3.2258) grad_norm 1.9155 (2.1364) [2021-04-16 12:08:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][370/1251] eta 0:04:42 lr 0.000209 time 0.2845 (0.3206) loss 2.8913 (3.2202) grad_norm 2.0837 (2.1354) [2021-04-16 12:08:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][380/1251] eta 0:04:38 lr 0.000209 time 0.2604 (0.3195) loss 3.9265 (3.2263) grad_norm 2.4957 (2.1382) [2021-04-16 12:08:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][390/1251] eta 0:04:34 lr 0.000209 time 0.2685 (0.3184) loss 3.3004 (3.2210) grad_norm 2.2416 (2.1356) [2021-04-16 12:08:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][400/1251] eta 0:04:30 lr 0.000209 time 0.2599 (0.3173) loss 3.4559 (3.2243) grad_norm 2.2489 (2.1345) [2021-04-16 12:08:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][410/1251] eta 0:04:25 lr 0.000209 time 0.2688 (0.3163) loss 3.5958 (3.2241) grad_norm 1.9398 (2.1320) [2021-04-16 12:08:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][420/1251] eta 0:04:22 lr 0.000208 time 0.2630 (0.3153) loss 4.1373 (3.2265) grad_norm 1.9646 (2.1315) [2021-04-16 12:08:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][430/1251] eta 0:04:18 lr 0.000208 time 0.2982 (0.3144) loss 3.4360 (3.2323) grad_norm 1.8662 (2.1309) [2021-04-16 12:09:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][440/1251] eta 0:04:14 lr 0.000208 time 0.2502 (0.3138) loss 3.2853 (3.2340) grad_norm 1.9919 (2.1276) [2021-04-16 12:09:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][450/1251] eta 0:04:10 lr 0.000208 time 0.2470 (0.3130) loss 3.8777 (3.2369) grad_norm 1.9095 (2.1267) [2021-04-16 12:09:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][460/1251] eta 0:04:06 lr 0.000208 time 0.2669 (0.3122) loss 3.4110 (3.2334) grad_norm 1.8359 (2.1254) [2021-04-16 12:09:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][470/1251] eta 0:04:03 lr 0.000208 time 0.2521 (0.3114) loss 2.2946 (3.2381) grad_norm 2.3842 (2.1251) [2021-04-16 12:09:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][480/1251] eta 0:03:59 lr 0.000208 time 0.2624 (0.3110) loss 3.0169 (3.2354) grad_norm 1.9353 (2.1247) [2021-04-16 12:09:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][490/1251] eta 0:03:56 lr 0.000208 time 0.2695 (0.3107) loss 3.6773 (3.2417) grad_norm 2.0274 (2.1279) [2021-04-16 12:09:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][500/1251] eta 0:03:52 lr 0.000208 time 0.2434 (0.3099) loss 3.3770 (3.2461) grad_norm 2.0104 (2.1310) [2021-04-16 12:09:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][510/1251] eta 0:03:49 lr 0.000208 time 0.2659 (0.3093) loss 2.5683 (3.2462) grad_norm 2.1205 (2.1295) [2021-04-16 12:09:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][520/1251] eta 0:03:45 lr 0.000208 time 0.3117 (0.3086) loss 3.2678 (3.2469) grad_norm 2.0833 (2.1288) [2021-04-16 12:09:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][530/1251] eta 0:03:42 lr 0.000208 time 0.2584 (0.3081) loss 3.8545 (3.2516) grad_norm 2.1771 (2.1282) [2021-04-16 12:09:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][540/1251] eta 0:03:38 lr 0.000208 time 0.2736 (0.3074) loss 2.6429 (3.2508) grad_norm 2.4001 (2.1282) [2021-04-16 12:09:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][550/1251] eta 0:03:35 lr 0.000208 time 0.2483 (0.3069) loss 2.4477 (3.2526) grad_norm 1.9686 (2.1292) [2021-04-16 12:09:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][560/1251] eta 0:03:31 lr 0.000208 time 0.2572 (0.3064) loss 2.9030 (3.2527) grad_norm 1.9422 (2.1283) [2021-04-16 12:09:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][570/1251] eta 0:03:28 lr 0.000208 time 0.2585 (0.3059) loss 3.6952 (3.2551) grad_norm 2.2115 (2.1282) [2021-04-16 12:09:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][580/1251] eta 0:03:25 lr 0.000208 time 0.2955 (0.3059) loss 2.7942 (3.2551) grad_norm 2.0540 (2.1276) [2021-04-16 12:09:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][590/1251] eta 0:03:21 lr 0.000208 time 0.2828 (0.3054) loss 3.4450 (3.2564) grad_norm 2.2674 (2.1307) [2021-04-16 12:09:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][600/1251] eta 0:03:18 lr 0.000208 time 0.2970 (0.3049) loss 3.3728 (3.2572) grad_norm 2.1839 (2.1310) [2021-04-16 12:09:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][610/1251] eta 0:03:15 lr 0.000208 time 0.2818 (0.3045) loss 3.4289 (3.2589) grad_norm 2.3025 (2.1324) [2021-04-16 12:09:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][620/1251] eta 0:03:11 lr 0.000208 time 0.2531 (0.3042) loss 2.2744 (3.2598) grad_norm 2.0452 (2.1329) [2021-04-16 12:09:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][630/1251] eta 0:03:08 lr 0.000208 time 0.2761 (0.3038) loss 3.3952 (3.2597) grad_norm 2.0146 (2.1340) [2021-04-16 12:09:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][640/1251] eta 0:03:05 lr 0.000208 time 0.2835 (0.3034) loss 3.2170 (3.2619) grad_norm 2.1220 (2.1363) [2021-04-16 12:10:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][650/1251] eta 0:03:02 lr 0.000208 time 0.2984 (0.3030) loss 2.6228 (3.2626) grad_norm 2.2708 (2.1371) [2021-04-16 12:10:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][660/1251] eta 0:02:58 lr 0.000208 time 0.2935 (0.3026) loss 3.4362 (3.2671) grad_norm 2.0092 (2.1387) [2021-04-16 12:10:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][670/1251] eta 0:02:55 lr 0.000208 time 0.2699 (0.3023) loss 3.6936 (3.2683) grad_norm 2.1741 (2.1407) [2021-04-16 12:10:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][680/1251] eta 0:02:52 lr 0.000208 time 0.2699 (0.3019) loss 3.3373 (3.2683) grad_norm 2.3730 (2.1418) [2021-04-16 12:10:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][690/1251] eta 0:02:49 lr 0.000208 time 0.2615 (0.3017) loss 2.2535 (3.2652) grad_norm 1.9244 (2.1423) [2021-04-16 12:10:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][700/1251] eta 0:02:46 lr 0.000208 time 0.2605 (0.3014) loss 3.7399 (3.2670) grad_norm 2.5385 (2.1418) [2021-04-16 12:10:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][710/1251] eta 0:02:42 lr 0.000208 time 0.3041 (0.3011) loss 2.4144 (3.2635) grad_norm 1.9440 (2.1401) [2021-04-16 12:10:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][720/1251] eta 0:02:39 lr 0.000207 time 0.2497 (0.3008) loss 3.1588 (3.2641) grad_norm 2.3247 (2.1405) [2021-04-16 12:10:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][730/1251] eta 0:02:36 lr 0.000207 time 0.2916 (0.3006) loss 2.9535 (3.2626) grad_norm 2.0937 (2.1408) [2021-04-16 12:10:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][740/1251] eta 0:02:33 lr 0.000207 time 0.2920 (0.3004) loss 3.4193 (3.2657) grad_norm 2.1143 (2.1403) [2021-04-16 12:10:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][750/1251] eta 0:02:30 lr 0.000207 time 0.2662 (0.3002) loss 3.6268 (3.2660) grad_norm 2.5187 (2.1402) [2021-04-16 12:10:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][760/1251] eta 0:02:27 lr 0.000207 time 0.2741 (0.3000) loss 4.1127 (3.2661) grad_norm 1.8741 (2.1387) [2021-04-16 12:10:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][770/1251] eta 0:02:24 lr 0.000207 time 0.2599 (0.2997) loss 3.7349 (3.2700) grad_norm 2.0680 (2.1425) [2021-04-16 12:10:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][780/1251] eta 0:02:21 lr 0.000207 time 0.3079 (0.2995) loss 2.9647 (3.2679) grad_norm 2.0831 (2.1413) [2021-04-16 12:10:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][790/1251] eta 0:02:17 lr 0.000207 time 0.2745 (0.2991) loss 3.2500 (3.2678) grad_norm 2.2337 (2.1420) [2021-04-16 12:10:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][800/1251] eta 0:02:14 lr 0.000207 time 0.2825 (0.2988) loss 3.6006 (3.2676) grad_norm 2.0390 (2.1416) [2021-04-16 12:10:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][810/1251] eta 0:02:11 lr 0.000207 time 0.2638 (0.2985) loss 3.4552 (3.2690) grad_norm 2.0476 (2.1412) [2021-04-16 12:10:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][820/1251] eta 0:02:08 lr 0.000207 time 0.2687 (0.2984) loss 2.8660 (3.2678) grad_norm 1.8270 (2.1407) [2021-04-16 12:10:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][830/1251] eta 0:02:05 lr 0.000207 time 0.2700 (0.2982) loss 2.8733 (3.2665) grad_norm 2.1676 (2.1409) [2021-04-16 12:10:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][840/1251] eta 0:02:02 lr 0.000207 time 0.2503 (0.2980) loss 3.3194 (3.2663) grad_norm 2.0030 (2.1411) [2021-04-16 12:10:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][850/1251] eta 0:01:59 lr 0.000207 time 0.2801 (0.2978) loss 3.1864 (3.2638) grad_norm 2.2188 (2.1405) [2021-04-16 12:10:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][860/1251] eta 0:01:56 lr 0.000207 time 0.2558 (0.2975) loss 3.2492 (3.2636) grad_norm 1.8469 (2.1392) [2021-04-16 12:11:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][870/1251] eta 0:01:53 lr 0.000207 time 0.2877 (0.2972) loss 2.9130 (3.2608) grad_norm 2.2618 (2.1392) [2021-04-16 12:11:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][880/1251] eta 0:01:50 lr 0.000207 time 0.2885 (0.2970) loss 3.0377 (3.2600) grad_norm 2.7159 (2.1400) [2021-04-16 12:11:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][890/1251] eta 0:01:47 lr 0.000207 time 0.2778 (0.2968) loss 2.6498 (3.2600) grad_norm 2.2432 (2.1391) [2021-04-16 12:11:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][900/1251] eta 0:01:44 lr 0.000207 time 0.2811 (0.2966) loss 2.3653 (3.2591) grad_norm 1.8779 (2.1396) [2021-04-16 12:11:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][910/1251] eta 0:01:41 lr 0.000207 time 0.2868 (0.2964) loss 3.4408 (3.2586) grad_norm 1.9418 (2.1394) [2021-04-16 12:11:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][920/1251] eta 0:01:38 lr 0.000207 time 0.3041 (0.2961) loss 3.5530 (3.2612) grad_norm 2.2615 (2.1382) [2021-04-16 12:11:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][930/1251] eta 0:01:35 lr 0.000207 time 0.4380 (0.2960) loss 3.0148 (3.2595) grad_norm 2.4825 (2.1382) [2021-04-16 12:11:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][940/1251] eta 0:01:32 lr 0.000207 time 0.2764 (0.2960) loss 3.1549 (3.2598) grad_norm 2.0450 (2.1381) [2021-04-16 12:11:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][950/1251] eta 0:01:29 lr 0.000207 time 0.2663 (0.2958) loss 4.0914 (3.2595) grad_norm 1.8850 (2.1376) [2021-04-16 12:11:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][960/1251] eta 0:01:26 lr 0.000207 time 0.2652 (0.2956) loss 3.1211 (3.2613) grad_norm 2.0296 (2.1372) [2021-04-16 12:11:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][970/1251] eta 0:01:23 lr 0.000207 time 0.2810 (0.2954) loss 2.9096 (3.2613) grad_norm 2.3393 (2.1375) [2021-04-16 12:11:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][980/1251] eta 0:01:20 lr 0.000207 time 0.2895 (0.2953) loss 3.9197 (3.2620) grad_norm 2.5013 (2.1371) [2021-04-16 12:11:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][990/1251] eta 0:01:17 lr 0.000207 time 0.2853 (0.2952) loss 3.4508 (3.2625) grad_norm 2.4400 (2.1374) [2021-04-16 12:11:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][1000/1251] eta 0:01:14 lr 0.000207 time 0.2833 (0.2950) loss 2.7459 (3.2657) grad_norm 2.2234 (2.1374) [2021-04-16 12:11:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][1010/1251] eta 0:01:11 lr 0.000207 time 0.2640 (0.2948) loss 3.2861 (3.2635) grad_norm 1.8939 (2.1377) [2021-04-16 12:11:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][1020/1251] eta 0:01:08 lr 0.000206 time 0.2977 (0.2948) loss 3.8586 (3.2660) grad_norm 2.0104 (2.1376) [2021-04-16 12:11:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][1030/1251] eta 0:01:05 lr 0.000206 time 0.2808 (0.2946) loss 4.0620 (3.2677) grad_norm 2.0222 (2.1372) [2021-04-16 12:11:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][1040/1251] eta 0:01:02 lr 0.000206 time 0.2590 (0.2945) loss 3.6614 (3.2698) grad_norm 2.0167 (2.1369) [2021-04-16 12:11:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][1050/1251] eta 0:00:59 lr 0.000206 time 0.4022 (0.2945) loss 3.4268 (3.2713) grad_norm 2.4253 (2.1357) [2021-04-16 12:11:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][1060/1251] eta 0:00:56 lr 0.000206 time 0.2873 (0.2944) loss 3.3907 (3.2714) grad_norm 2.4315 (2.1360) [2021-04-16 12:11:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][1070/1251] eta 0:00:53 lr 0.000206 time 0.2561 (0.2943) loss 2.3955 (3.2694) grad_norm 2.1028 (2.1361) [2021-04-16 12:12:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][1080/1251] eta 0:00:50 lr 0.000206 time 0.2823 (0.2941) loss 2.3764 (3.2691) grad_norm 2.0365 (2.1361) [2021-04-16 12:12:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][1090/1251] eta 0:00:47 lr 0.000206 time 0.2733 (0.2939) loss 3.2323 (3.2696) grad_norm 2.0626 (2.1350) [2021-04-16 12:12:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][1100/1251] eta 0:00:44 lr 0.000206 time 0.2836 (0.2938) loss 2.6197 (3.2696) grad_norm 2.0153 (2.1360) [2021-04-16 12:12:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][1110/1251] eta 0:00:41 lr 0.000206 time 0.2967 (0.2937) loss 2.0241 (3.2674) grad_norm 2.2113 (2.1346) [2021-04-16 12:12:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][1120/1251] eta 0:00:38 lr 0.000206 time 0.2646 (0.2935) loss 3.1710 (3.2696) grad_norm 2.2513 (2.1337) [2021-04-16 12:12:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][1130/1251] eta 0:00:35 lr 0.000206 time 0.2505 (0.2934) loss 3.9480 (3.2709) grad_norm 2.2092 (2.1352) [2021-04-16 12:12:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][1140/1251] eta 0:00:32 lr 0.000206 time 0.2773 (0.2932) loss 3.2461 (3.2731) grad_norm 2.0958 (2.1350) [2021-04-16 12:12:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][1150/1251] eta 0:00:29 lr 0.000206 time 0.2611 (0.2932) loss 2.1380 (3.2725) grad_norm 2.2039 (2.1349) [2021-04-16 12:12:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][1160/1251] eta 0:00:26 lr 0.000206 time 0.2746 (0.2932) loss 3.1597 (3.2734) grad_norm 3.8199 (2.1383) [2021-04-16 12:12:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][1170/1251] eta 0:00:23 lr 0.000206 time 0.2809 (0.2930) loss 2.5708 (3.2722) grad_norm 2.0037 (2.1394) [2021-04-16 12:12:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][1180/1251] eta 0:00:20 lr 0.000206 time 0.2964 (0.2929) loss 3.5336 (3.2719) grad_norm 2.0251 (2.1384) [2021-04-16 12:12:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][1190/1251] eta 0:00:17 lr 0.000206 time 0.2804 (0.2927) loss 3.5559 (3.2729) grad_norm 1.9075 (2.1384) [2021-04-16 12:12:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][1200/1251] eta 0:00:14 lr 0.000206 time 0.2842 (0.2927) loss 3.5273 (3.2746) grad_norm 2.2630 (2.1382) [2021-04-16 12:12:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][1210/1251] eta 0:00:11 lr 0.000206 time 0.2629 (0.2925) loss 3.4612 (3.2756) grad_norm 2.0733 (2.1383) [2021-04-16 12:12:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][1220/1251] eta 0:00:09 lr 0.000206 time 0.2912 (0.2924) loss 4.1522 (3.2742) grad_norm 2.2364 (2.1391) [2021-04-16 12:12:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][1230/1251] eta 0:00:06 lr 0.000206 time 0.2725 (0.2923) loss 3.9504 (3.2744) grad_norm 2.0803 (2.1386) [2021-04-16 12:12:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][1240/1251] eta 0:00:03 lr 0.000206 time 0.2487 (0.2921) loss 3.2304 (3.2746) grad_norm 2.0967 (2.1377) [2021-04-16 12:12:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [211/300][1250/1251] eta 0:00:00 lr 0.000206 time 0.2560 (0.2917) loss 2.3605 (3.2742) grad_norm 2.1792 (2.1374) [2021-04-16 12:12:51 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 211 training takes 0:06:07 [2021-04-16 12:12:51 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_211.pth saving...... [2021-04-16 12:12:57 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_211.pth saved !!! [2021-04-16 12:12:58 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.068 (1.068) Loss 0.8940 (0.8940) Acc@1 79.297 (79.297) Acc@5 94.629 (94.629) [2021-04-16 12:13:00 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.382 (0.252) Loss 0.9478 (0.8955) Acc@1 78.223 (78.888) Acc@5 93.945 (94.798) [2021-04-16 12:13:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.424 (0.249) Loss 0.9461 (0.9004) Acc@1 79.004 (79.050) Acc@5 93.750 (94.708) [2021-04-16 12:13:04 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.107 (0.232) Loss 0.9257 (0.9018) Acc@1 78.223 (78.966) Acc@5 93.750 (94.651) [2021-04-16 12:13:06 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.075 (0.208) Loss 0.8944 (0.9024) Acc@1 78.516 (78.944) Acc@5 95.312 (94.646) [2021-04-16 12:13:14 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 78.920 Acc@5 94.652 [2021-04-16 12:13:14 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 78.9% [2021-04-16 12:13:14 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 78.92% [2021-04-16 12:13:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][0/1251] eta 0:52:10 lr 0.000206 time 2.5026 (2.5026) loss 3.5365 (3.5365) grad_norm 1.9836 (1.9836) [2021-04-16 12:13:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][10/1251] eta 0:09:51 lr 0.000206 time 0.2735 (0.4768) loss 2.3195 (3.2842) grad_norm 2.0779 (2.0670) [2021-04-16 12:13:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][20/1251] eta 0:08:00 lr 0.000206 time 0.3158 (0.3903) loss 3.7452 (3.1176) grad_norm 2.4238 (2.1137) [2021-04-16 12:13:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][30/1251] eta 0:07:15 lr 0.000206 time 0.2890 (0.3565) loss 3.6034 (3.1601) grad_norm 2.2835 (2.1281) [2021-04-16 12:13:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][40/1251] eta 0:06:51 lr 0.000206 time 0.2606 (0.3401) loss 3.6751 (3.1412) grad_norm 1.9220 (2.1084) [2021-04-16 12:13:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][50/1251] eta 0:06:38 lr 0.000206 time 0.3812 (0.3315) loss 3.3958 (3.1025) grad_norm 2.0991 (2.0801) [2021-04-16 12:13:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][60/1251] eta 0:06:23 lr 0.000206 time 0.2617 (0.3223) loss 2.1898 (3.1284) grad_norm 1.7593 (2.0848) [2021-04-16 12:13:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][70/1251] eta 0:06:12 lr 0.000205 time 0.2808 (0.3158) loss 3.5815 (3.1598) grad_norm 2.2096 (2.1031) [2021-04-16 12:13:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][80/1251] eta 0:06:04 lr 0.000205 time 0.2588 (0.3113) loss 3.0417 (3.1678) grad_norm 2.0027 (2.1127) [2021-04-16 12:13:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][90/1251] eta 0:05:57 lr 0.000205 time 0.2650 (0.3075) loss 3.7905 (3.1497) grad_norm 2.1243 (2.1182) [2021-04-16 12:13:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][100/1251] eta 0:05:51 lr 0.000205 time 0.2642 (0.3050) loss 3.5589 (3.1722) grad_norm 1.9712 (2.1211) [2021-04-16 12:13:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][110/1251] eta 0:05:45 lr 0.000205 time 0.2715 (0.3025) loss 3.1656 (3.1654) grad_norm 2.4246 (2.1339) [2021-04-16 12:13:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][120/1251] eta 0:05:40 lr 0.000205 time 0.2687 (0.3007) loss 3.8103 (3.1784) grad_norm 2.0244 (2.1273) [2021-04-16 12:13:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][130/1251] eta 0:05:35 lr 0.000205 time 0.2885 (0.2995) loss 3.5841 (3.1778) grad_norm 2.0845 (2.1370) [2021-04-16 12:13:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][140/1251] eta 0:05:33 lr 0.000205 time 0.3563 (0.2999) loss 2.9571 (3.1820) grad_norm 2.2222 (2.1491) [2021-04-16 12:13:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][150/1251] eta 0:05:28 lr 0.000205 time 0.2809 (0.2982) loss 2.9298 (3.1962) grad_norm 2.0474 (2.1538) [2021-04-16 12:14:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][160/1251] eta 0:05:24 lr 0.000205 time 0.2734 (0.2971) loss 3.2428 (3.2022) grad_norm 2.2844 (2.1562) [2021-04-16 12:14:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][170/1251] eta 0:05:19 lr 0.000205 time 0.2668 (0.2959) loss 3.6373 (3.2046) grad_norm 2.1236 (2.1547) [2021-04-16 12:14:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][180/1251] eta 0:05:16 lr 0.000205 time 0.2902 (0.2955) loss 2.2313 (3.2054) grad_norm 1.9495 (2.1534) [2021-04-16 12:14:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][190/1251] eta 0:05:12 lr 0.000205 time 0.2795 (0.2948) loss 3.4889 (3.2190) grad_norm 2.4585 (2.1565) [2021-04-16 12:14:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][200/1251] eta 0:05:08 lr 0.000205 time 0.2529 (0.2938) loss 3.4084 (3.2239) grad_norm 2.9256 (2.1599) [2021-04-16 12:14:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][210/1251] eta 0:05:05 lr 0.000205 time 0.2865 (0.2933) loss 3.0189 (3.2376) grad_norm 1.8567 (2.1711) [2021-04-16 12:14:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][220/1251] eta 0:05:01 lr 0.000205 time 0.2836 (0.2926) loss 4.1599 (3.2485) grad_norm 2.1461 (2.1741) [2021-04-16 12:14:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][230/1251] eta 0:04:58 lr 0.000205 time 0.2698 (0.2926) loss 2.7622 (3.2474) grad_norm 2.5778 (2.1755) [2021-04-16 12:14:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][240/1251] eta 0:04:55 lr 0.000205 time 0.2812 (0.2921) loss 3.8971 (3.2551) grad_norm 1.9675 (2.1756) [2021-04-16 12:14:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][250/1251] eta 0:04:51 lr 0.000205 time 0.2875 (0.2916) loss 3.9471 (3.2485) grad_norm 2.1712 (2.1827) [2021-04-16 12:14:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][260/1251] eta 0:04:49 lr 0.000205 time 0.4426 (0.2920) loss 2.2252 (3.2478) grad_norm 2.4562 (2.1848) [2021-04-16 12:14:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][270/1251] eta 0:04:46 lr 0.000205 time 0.2837 (0.2916) loss 3.5806 (3.2473) grad_norm 1.9758 (2.1828) [2021-04-16 12:14:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][280/1251] eta 0:04:42 lr 0.000205 time 0.2588 (0.2909) loss 3.2483 (3.2395) grad_norm 1.8902 (2.1814) [2021-04-16 12:14:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][290/1251] eta 0:04:39 lr 0.000205 time 0.2765 (0.2909) loss 3.8399 (3.2465) grad_norm 2.1144 (2.1783) [2021-04-16 12:14:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][300/1251] eta 0:04:36 lr 0.000205 time 0.2829 (0.2906) loss 3.5769 (3.2522) grad_norm 2.0291 (2.1793) [2021-04-16 12:14:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][310/1251] eta 0:04:33 lr 0.000205 time 0.2765 (0.2901) loss 3.5704 (3.2564) grad_norm 1.9215 (2.1796) [2021-04-16 12:14:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][320/1251] eta 0:04:29 lr 0.000205 time 0.2901 (0.2899) loss 2.0590 (3.2507) grad_norm 2.0483 (2.1773) [2021-04-16 12:14:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][330/1251] eta 0:04:26 lr 0.000205 time 0.2749 (0.2895) loss 3.1837 (3.2508) grad_norm 2.3945 (2.1763) [2021-04-16 12:14:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][340/1251] eta 0:04:23 lr 0.000205 time 0.2751 (0.2892) loss 3.4983 (3.2528) grad_norm 2.2667 (2.1795) [2021-04-16 12:14:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][350/1251] eta 0:04:20 lr 0.000205 time 0.2886 (0.2890) loss 3.1650 (3.2537) grad_norm 2.0694 (2.1816) [2021-04-16 12:14:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][360/1251] eta 0:04:18 lr 0.000205 time 0.2730 (0.2897) loss 3.7983 (3.2537) grad_norm 2.2430 (2.1798) [2021-04-16 12:15:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][370/1251] eta 0:04:15 lr 0.000205 time 0.2795 (0.2898) loss 3.1467 (3.2450) grad_norm 1.9866 (2.1786) [2021-04-16 12:15:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][380/1251] eta 0:04:12 lr 0.000204 time 0.2913 (0.2898) loss 3.4123 (3.2373) grad_norm 1.8519 (2.1776) [2021-04-16 12:15:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][390/1251] eta 0:04:09 lr 0.000204 time 0.3061 (0.2895) loss 2.9099 (3.2349) grad_norm 2.6883 (2.1794) [2021-04-16 12:15:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][400/1251] eta 0:04:05 lr 0.000204 time 0.2697 (0.2890) loss 2.9909 (3.2414) grad_norm 2.3189 (2.1812) [2021-04-16 12:15:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][410/1251] eta 0:04:02 lr 0.000204 time 0.2520 (0.2887) loss 3.6811 (3.2414) grad_norm 2.0141 (2.1807) [2021-04-16 12:15:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][420/1251] eta 0:03:59 lr 0.000204 time 0.2772 (0.2885) loss 3.7761 (3.2438) grad_norm 1.8112 (2.1797) [2021-04-16 12:15:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][430/1251] eta 0:03:56 lr 0.000204 time 0.2934 (0.2883) loss 2.7159 (3.2432) grad_norm 2.3723 (2.1779) [2021-04-16 12:15:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][440/1251] eta 0:03:53 lr 0.000204 time 0.2805 (0.2881) loss 2.8838 (3.2465) grad_norm 2.0184 (2.1805) [2021-04-16 12:15:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][450/1251] eta 0:03:50 lr 0.000204 time 0.2640 (0.2878) loss 2.7322 (3.2435) grad_norm 2.5038 (2.1807) [2021-04-16 12:15:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][460/1251] eta 0:03:47 lr 0.000204 time 0.2666 (0.2876) loss 3.1998 (3.2425) grad_norm 2.0255 (2.1810) [2021-04-16 12:15:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][470/1251] eta 0:03:44 lr 0.000204 time 0.2758 (0.2874) loss 3.3781 (3.2416) grad_norm 2.0947 (2.1790) [2021-04-16 12:15:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][480/1251] eta 0:03:41 lr 0.000204 time 0.2888 (0.2873) loss 3.6147 (3.2462) grad_norm 2.0115 (2.1789) [2021-04-16 12:15:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][490/1251] eta 0:03:38 lr 0.000204 time 0.3032 (0.2875) loss 3.6307 (3.2441) grad_norm 1.7817 (2.1770) [2021-04-16 12:15:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][500/1251] eta 0:03:35 lr 0.000204 time 0.2884 (0.2873) loss 3.7377 (3.2473) grad_norm 2.2168 (2.1747) [2021-04-16 12:15:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][510/1251] eta 0:03:32 lr 0.000204 time 0.2839 (0.2871) loss 3.3765 (3.2510) grad_norm 1.8809 (2.1747) [2021-04-16 12:15:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][520/1251] eta 0:03:29 lr 0.000204 time 0.2654 (0.2869) loss 3.2080 (3.2454) grad_norm 2.1858 (2.1749) [2021-04-16 12:15:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][530/1251] eta 0:03:26 lr 0.000204 time 0.2899 (0.2868) loss 3.7947 (3.2473) grad_norm 2.0568 (2.1726) [2021-04-16 12:15:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][540/1251] eta 0:03:23 lr 0.000204 time 0.3142 (0.2867) loss 3.0418 (3.2422) grad_norm 2.3947 (2.1712) [2021-04-16 12:15:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][550/1251] eta 0:03:20 lr 0.000204 time 0.2711 (0.2865) loss 2.5100 (3.2385) grad_norm 1.9095 (2.1718) [2021-04-16 12:15:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][560/1251] eta 0:03:17 lr 0.000204 time 0.2762 (0.2863) loss 2.1297 (3.2362) grad_norm 2.2991 (2.1710) [2021-04-16 12:15:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][570/1251] eta 0:03:15 lr 0.000204 time 0.2632 (0.2865) loss 3.6021 (3.2342) grad_norm 2.3814 (2.1714) [2021-04-16 12:16:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][580/1251] eta 0:03:12 lr 0.000204 time 0.2619 (0.2869) loss 2.6488 (3.2355) grad_norm 2.1728 (2.1695) [2021-04-16 12:16:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][590/1251] eta 0:03:09 lr 0.000204 time 0.2778 (0.2870) loss 3.5068 (3.2361) grad_norm 1.9606 (2.1707) [2021-04-16 12:16:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][600/1251] eta 0:03:06 lr 0.000204 time 0.2825 (0.2868) loss 3.7980 (3.2375) grad_norm 1.9070 (2.1706) [2021-04-16 12:16:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][610/1251] eta 0:03:03 lr 0.000204 time 0.2680 (0.2866) loss 2.3743 (3.2381) grad_norm 2.0595 (2.1710) [2021-04-16 12:16:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][620/1251] eta 0:03:00 lr 0.000204 time 0.3003 (0.2866) loss 2.2810 (3.2380) grad_norm 1.9808 (2.1700) [2021-04-16 12:16:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][630/1251] eta 0:02:57 lr 0.000204 time 0.2706 (0.2864) loss 3.2562 (3.2401) grad_norm 2.1729 (2.1712) [2021-04-16 12:16:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][640/1251] eta 0:02:54 lr 0.000204 time 0.2676 (0.2862) loss 3.8735 (3.2436) grad_norm 2.0074 (2.1721) [2021-04-16 12:16:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][650/1251] eta 0:02:51 lr 0.000204 time 0.2863 (0.2860) loss 3.3379 (3.2423) grad_norm 2.1448 (2.1725) [2021-04-16 12:16:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][660/1251] eta 0:02:48 lr 0.000204 time 0.2726 (0.2859) loss 3.3118 (3.2428) grad_norm 1.9813 (2.1715) [2021-04-16 12:16:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][670/1251] eta 0:02:46 lr 0.000204 time 0.2889 (0.2859) loss 3.2456 (3.2421) grad_norm 2.2173 (2.1742) [2021-04-16 12:16:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][680/1251] eta 0:02:43 lr 0.000203 time 0.2709 (0.2857) loss 3.2946 (3.2443) grad_norm 2.1797 (2.1742) [2021-04-16 12:16:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][690/1251] eta 0:02:40 lr 0.000203 time 0.2591 (0.2856) loss 3.5683 (3.2423) grad_norm 2.3043 (2.1745) [2021-04-16 12:16:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][700/1251] eta 0:02:37 lr 0.000203 time 0.2569 (0.2855) loss 3.2774 (3.2437) grad_norm 2.1191 (2.1740) [2021-04-16 12:16:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][710/1251] eta 0:02:34 lr 0.000203 time 0.2788 (0.2854) loss 3.5620 (3.2466) grad_norm 2.2615 (2.1754) [2021-04-16 12:16:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][720/1251] eta 0:02:31 lr 0.000203 time 0.2675 (0.2856) loss 3.6640 (3.2484) grad_norm 2.2306 (2.1758) [2021-04-16 12:16:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][730/1251] eta 0:02:28 lr 0.000203 time 0.2889 (0.2855) loss 3.5512 (3.2493) grad_norm 2.7730 (2.1773) [2021-04-16 12:16:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][740/1251] eta 0:02:25 lr 0.000203 time 0.2576 (0.2854) loss 3.1520 (3.2524) grad_norm 1.9730 (2.1767) [2021-04-16 12:16:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][750/1251] eta 0:02:22 lr 0.000203 time 0.2615 (0.2853) loss 3.8530 (3.2519) grad_norm 2.4054 (2.1768) [2021-04-16 12:16:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][760/1251] eta 0:02:20 lr 0.000203 time 0.2907 (0.2853) loss 3.7136 (3.2533) grad_norm 2.0067 (2.1758) [2021-04-16 12:16:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][770/1251] eta 0:02:17 lr 0.000203 time 0.2702 (0.2852) loss 3.3383 (3.2502) grad_norm 2.3795 (2.1754) [2021-04-16 12:16:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][780/1251] eta 0:02:14 lr 0.000203 time 0.2913 (0.2851) loss 3.7772 (3.2513) grad_norm 1.8614 (2.1751) [2021-04-16 12:17:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][790/1251] eta 0:02:11 lr 0.000203 time 0.2751 (0.2850) loss 3.0907 (3.2524) grad_norm 2.3017 (2.1747) [2021-04-16 12:17:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][800/1251] eta 0:02:08 lr 0.000203 time 0.2761 (0.2849) loss 2.8738 (3.2527) grad_norm 2.3543 (2.1739) [2021-04-16 12:17:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][810/1251] eta 0:02:05 lr 0.000203 time 0.2922 (0.2850) loss 3.2133 (3.2548) grad_norm 1.9425 (2.1723) [2021-04-16 12:17:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][820/1251] eta 0:02:02 lr 0.000203 time 0.2878 (0.2850) loss 4.2115 (3.2551) grad_norm 2.1948 (2.1716) [2021-04-16 12:17:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][830/1251] eta 0:01:59 lr 0.000203 time 0.2823 (0.2849) loss 3.6397 (3.2580) grad_norm 1.6928 (2.1708) [2021-04-16 12:17:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][840/1251] eta 0:01:57 lr 0.000203 time 0.2722 (0.2849) loss 3.7636 (3.2581) grad_norm 2.0269 (2.1697) [2021-04-16 12:17:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][850/1251] eta 0:01:54 lr 0.000203 time 0.2806 (0.2849) loss 2.3674 (3.2517) grad_norm 1.8071 (2.1676) [2021-04-16 12:17:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][860/1251] eta 0:01:51 lr 0.000203 time 0.2951 (0.2849) loss 3.7110 (3.2543) grad_norm 2.0966 (2.1665) [2021-04-16 12:17:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][870/1251] eta 0:01:48 lr 0.000203 time 0.2903 (0.2850) loss 1.8958 (3.2496) grad_norm 2.0264 (2.1647) [2021-04-16 12:17:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][880/1251] eta 0:01:45 lr 0.000203 time 0.2826 (0.2848) loss 3.7211 (3.2484) grad_norm 2.1491 (2.1646) [2021-04-16 12:17:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][890/1251] eta 0:01:42 lr 0.000203 time 0.2890 (0.2847) loss 3.7977 (3.2452) grad_norm 2.6164 (2.1649) [2021-04-16 12:17:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][900/1251] eta 0:01:39 lr 0.000203 time 0.2432 (0.2846) loss 3.6197 (3.2473) grad_norm 2.3113 (2.1642) [2021-04-16 12:17:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][910/1251] eta 0:01:37 lr 0.000203 time 0.2885 (0.2845) loss 3.5525 (3.2493) grad_norm 2.3593 (2.1650) [2021-04-16 12:17:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][920/1251] eta 0:01:34 lr 0.000203 time 0.2650 (0.2844) loss 3.7012 (3.2506) grad_norm 2.5483 (2.1648) [2021-04-16 12:17:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][930/1251] eta 0:01:31 lr 0.000203 time 0.2828 (0.2845) loss 3.5676 (3.2485) grad_norm 1.9433 (2.1644) [2021-04-16 12:17:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][940/1251] eta 0:01:28 lr 0.000203 time 0.2680 (0.2847) loss 3.5079 (3.2485) grad_norm 2.0427 (2.1636) [2021-04-16 12:17:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][950/1251] eta 0:01:25 lr 0.000203 time 0.3012 (0.2846) loss 2.4864 (3.2474) grad_norm 2.1201 (2.1641) [2021-04-16 12:17:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][960/1251] eta 0:01:22 lr 0.000203 time 0.2704 (0.2845) loss 2.8837 (3.2459) grad_norm 2.0534 (2.1655) [2021-04-16 12:17:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][970/1251] eta 0:01:19 lr 0.000203 time 0.2557 (0.2844) loss 2.3546 (3.2468) grad_norm 2.7014 (2.1673) [2021-04-16 12:17:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][980/1251] eta 0:01:17 lr 0.000202 time 0.2662 (0.2843) loss 3.5265 (3.2497) grad_norm 2.0846 (2.1681) [2021-04-16 12:17:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][990/1251] eta 0:01:14 lr 0.000202 time 0.2865 (0.2842) loss 2.6560 (3.2524) grad_norm 2.2503 (2.1679) [2021-04-16 12:17:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][1000/1251] eta 0:01:11 lr 0.000202 time 0.2644 (0.2841) loss 4.0031 (3.2554) grad_norm 1.8787 (2.1672) [2021-04-16 12:18:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][1010/1251] eta 0:01:08 lr 0.000202 time 0.2738 (0.2840) loss 3.2527 (3.2549) grad_norm 1.9283 (2.1656) [2021-04-16 12:18:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][1020/1251] eta 0:01:05 lr 0.000202 time 0.2960 (0.2840) loss 2.4080 (3.2554) grad_norm 2.1672 (2.1660) [2021-04-16 12:18:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][1030/1251] eta 0:01:02 lr 0.000202 time 0.2495 (0.2839) loss 2.4779 (3.2546) grad_norm 1.8536 (2.1664) [2021-04-16 12:18:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][1040/1251] eta 0:00:59 lr 0.000202 time 0.2764 (0.2839) loss 3.6553 (3.2540) grad_norm 2.3953 (2.1670) [2021-04-16 12:18:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][1050/1251] eta 0:00:57 lr 0.000202 time 0.2979 (0.2838) loss 3.7371 (3.2565) grad_norm 2.1934 (2.1669) [2021-04-16 12:18:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][1060/1251] eta 0:00:54 lr 0.000202 time 0.2825 (0.2837) loss 3.5343 (3.2558) grad_norm 1.9229 (2.1669) [2021-04-16 12:18:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][1070/1251] eta 0:00:51 lr 0.000202 time 0.2890 (0.2838) loss 3.3987 (3.2578) grad_norm 2.0322 (2.1660) [2021-04-16 12:18:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][1080/1251] eta 0:00:48 lr 0.000202 time 0.2696 (0.2837) loss 2.5670 (3.2572) grad_norm 2.1158 (2.1659) [2021-04-16 12:18:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][1090/1251] eta 0:00:45 lr 0.000202 time 0.2687 (0.2836) loss 3.1601 (3.2572) grad_norm 2.1600 (2.1660) [2021-04-16 12:18:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][1100/1251] eta 0:00:42 lr 0.000202 time 0.2805 (0.2836) loss 3.6604 (3.2570) grad_norm 2.1685 (2.1657) [2021-04-16 12:18:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][1110/1251] eta 0:00:39 lr 0.000202 time 0.2726 (0.2835) loss 2.9308 (3.2559) grad_norm 2.6209 (2.1664) [2021-04-16 12:18:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][1120/1251] eta 0:00:37 lr 0.000202 time 0.2733 (0.2835) loss 3.6754 (3.2572) grad_norm 2.8711 (2.1668) [2021-04-16 12:18:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][1130/1251] eta 0:00:34 lr 0.000202 time 0.2537 (0.2834) loss 3.4303 (3.2566) grad_norm 2.3121 (2.1669) [2021-04-16 12:18:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][1140/1251] eta 0:00:31 lr 0.000202 time 0.3063 (0.2834) loss 3.0015 (3.2562) grad_norm 2.2311 (2.1675) [2021-04-16 12:18:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][1150/1251] eta 0:00:28 lr 0.000202 time 0.2919 (0.2835) loss 3.4549 (3.2572) grad_norm 2.8765 (2.1673) [2021-04-16 12:18:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][1160/1251] eta 0:00:25 lr 0.000202 time 0.2474 (0.2835) loss 2.3109 (3.2591) grad_norm 1.9983 (2.1669) [2021-04-16 12:18:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][1170/1251] eta 0:00:22 lr 0.000202 time 0.2695 (0.2836) loss 3.6754 (3.2602) grad_norm 2.0868 (2.1660) [2021-04-16 12:18:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][1180/1251] eta 0:00:20 lr 0.000202 time 0.2600 (0.2835) loss 3.4141 (3.2602) grad_norm 2.3020 (2.1656) [2021-04-16 12:18:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][1190/1251] eta 0:00:17 lr 0.000202 time 0.3101 (0.2836) loss 3.7465 (3.2625) grad_norm 1.9713 (2.1648) [2021-04-16 12:18:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][1200/1251] eta 0:00:14 lr 0.000202 time 0.2789 (0.2835) loss 3.6523 (3.2630) grad_norm 2.5031 (2.1647) [2021-04-16 12:18:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][1210/1251] eta 0:00:11 lr 0.000202 time 0.2678 (0.2835) loss 3.8406 (3.2633) grad_norm 2.4419 (2.1660) [2021-04-16 12:19:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][1220/1251] eta 0:00:08 lr 0.000202 time 0.3106 (0.2834) loss 3.6494 (3.2633) grad_norm 1.9008 (2.1662) [2021-04-16 12:19:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][1230/1251] eta 0:00:05 lr 0.000202 time 0.3039 (0.2835) loss 3.2541 (3.2640) grad_norm 2.2745 (2.1659) [2021-04-16 12:19:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][1240/1251] eta 0:00:03 lr 0.000202 time 0.2495 (0.2834) loss 2.5940 (3.2631) grad_norm 2.3413 (2.1649) [2021-04-16 12:19:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [212/300][1250/1251] eta 0:00:00 lr 0.000202 time 0.2589 (0.2831) loss 3.1734 (3.2641) grad_norm 2.2484 (2.1652) [2021-04-16 12:19:12 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 212 training takes 0:05:57 [2021-04-16 12:19:12 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_212.pth saving...... [2021-04-16 12:19:19 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_212.pth saved !!! [2021-04-16 12:19:23 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 3.904 (3.904) Loss 0.7963 (0.7963) Acc@1 81.152 (81.152) Acc@5 95.312 (95.312) [2021-04-16 12:19:24 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.155 (0.471) Loss 0.8733 (0.8939) Acc@1 79.883 (78.817) Acc@5 94.434 (94.815) [2021-04-16 12:19:27 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.158 (0.378) Loss 0.9233 (0.9056) Acc@1 79.199 (78.562) Acc@5 93.359 (94.657) [2021-04-16 12:19:29 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.139 (0.310) Loss 0.8538 (0.9011) Acc@1 81.055 (78.739) Acc@5 95.020 (94.635) [2021-04-16 12:19:30 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.274) Loss 0.8811 (0.8944) Acc@1 80.469 (78.937) Acc@5 94.824 (94.662) [2021-04-16 12:19:34 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 78.904 Acc@5 94.628 [2021-04-16 12:19:34 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 78.9% [2021-04-16 12:19:34 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 78.92% [2021-04-16 12:19:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][0/1251] eta 2:41:22 lr 0.000202 time 7.7402 (7.7402) loss 3.3443 (3.3443) grad_norm 2.1302 (2.1302) [2021-04-16 12:19:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][10/1251] eta 0:19:40 lr 0.000202 time 0.2878 (0.9513) loss 4.1048 (3.4317) grad_norm 1.9960 (2.1517) [2021-04-16 12:19:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][20/1251] eta 0:13:01 lr 0.000202 time 0.2781 (0.6346) loss 3.6162 (3.5378) grad_norm 2.0880 (2.0840) [2021-04-16 12:19:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][30/1251] eta 0:10:35 lr 0.000202 time 0.2765 (0.5204) loss 3.6399 (3.4710) grad_norm 2.0059 (2.0747) [2021-04-16 12:19:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][40/1251] eta 0:09:22 lr 0.000201 time 0.2766 (0.4645) loss 3.0565 (3.4249) grad_norm 1.8931 (2.0979) [2021-04-16 12:19:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][50/1251] eta 0:08:35 lr 0.000201 time 0.2789 (0.4288) loss 3.1768 (3.3699) grad_norm 1.9168 (2.1427) [2021-04-16 12:19:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][60/1251] eta 0:08:04 lr 0.000201 time 0.4066 (0.4069) loss 2.3811 (3.3363) grad_norm 2.1463 (2.1342) [2021-04-16 12:20:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][70/1251] eta 0:07:37 lr 0.000201 time 0.2619 (0.3878) loss 2.7717 (3.3305) grad_norm 1.7948 (2.1077) [2021-04-16 12:20:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][80/1251] eta 0:07:20 lr 0.000201 time 0.2670 (0.3761) loss 3.4146 (3.3215) grad_norm 2.0046 (2.1053) [2021-04-16 12:20:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][90/1251] eta 0:07:04 lr 0.000201 time 0.2737 (0.3655) loss 2.0743 (3.3114) grad_norm 2.0895 (2.0983) [2021-04-16 12:20:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][100/1251] eta 0:06:51 lr 0.000201 time 0.2971 (0.3571) loss 3.2944 (3.2957) grad_norm 2.1029 (2.1072) [2021-04-16 12:20:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][110/1251] eta 0:06:41 lr 0.000201 time 0.4968 (0.3516) loss 3.1039 (3.3025) grad_norm 2.4819 (2.1107) [2021-04-16 12:20:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][120/1251] eta 0:06:30 lr 0.000201 time 0.3109 (0.3454) loss 3.1854 (3.2784) grad_norm 1.8869 (2.1101) [2021-04-16 12:20:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][130/1251] eta 0:06:21 lr 0.000201 time 0.2832 (0.3402) loss 3.6433 (3.2822) grad_norm 2.2263 (2.1168) [2021-04-16 12:20:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][140/1251] eta 0:06:13 lr 0.000201 time 0.2675 (0.3364) loss 3.8239 (3.2780) grad_norm 2.1038 (2.1205) [2021-04-16 12:20:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][150/1251] eta 0:06:07 lr 0.000201 time 0.2795 (0.3337) loss 3.9203 (3.2875) grad_norm 2.0374 (2.1180) [2021-04-16 12:20:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][160/1251] eta 0:06:00 lr 0.000201 time 0.2744 (0.3302) loss 2.5360 (3.2828) grad_norm 2.5448 (2.1211) [2021-04-16 12:20:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][170/1251] eta 0:05:53 lr 0.000201 time 0.2840 (0.3269) loss 3.1000 (3.2870) grad_norm 2.7358 (2.1313) [2021-04-16 12:20:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][180/1251] eta 0:05:47 lr 0.000201 time 0.2524 (0.3241) loss 3.6109 (3.2759) grad_norm 2.0403 (2.1352) [2021-04-16 12:20:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][190/1251] eta 0:05:41 lr 0.000201 time 0.2620 (0.3220) loss 2.3605 (3.2626) grad_norm 1.9064 (2.1358) [2021-04-16 12:20:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][200/1251] eta 0:05:35 lr 0.000201 time 0.2706 (0.3197) loss 3.8869 (3.2694) grad_norm 2.3003 (2.1374) [2021-04-16 12:20:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][210/1251] eta 0:05:30 lr 0.000201 time 0.2863 (0.3175) loss 3.3495 (3.2695) grad_norm 2.7847 (2.1491) [2021-04-16 12:20:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][220/1251] eta 0:05:25 lr 0.000201 time 0.2673 (0.3161) loss 2.3796 (3.2678) grad_norm 1.8573 (2.1467) [2021-04-16 12:20:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][230/1251] eta 0:05:21 lr 0.000201 time 0.2594 (0.3145) loss 3.4472 (3.2687) grad_norm 2.1646 (2.1447) [2021-04-16 12:20:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][240/1251] eta 0:05:16 lr 0.000201 time 0.2935 (0.3129) loss 3.5187 (3.2638) grad_norm 2.2295 (2.1466) [2021-04-16 12:20:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][250/1251] eta 0:05:11 lr 0.000201 time 0.2698 (0.3114) loss 3.2059 (3.2656) grad_norm 2.0854 (2.1510) [2021-04-16 12:20:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][260/1251] eta 0:05:07 lr 0.000201 time 0.2756 (0.3101) loss 3.8488 (3.2750) grad_norm 1.7696 (2.1521) [2021-04-16 12:20:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][270/1251] eta 0:05:03 lr 0.000201 time 0.2706 (0.3090) loss 3.4150 (3.2693) grad_norm 2.2118 (2.1508) [2021-04-16 12:21:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][280/1251] eta 0:04:58 lr 0.000201 time 0.2926 (0.3079) loss 3.5810 (3.2634) grad_norm 1.8935 (2.1508) [2021-04-16 12:21:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][290/1251] eta 0:04:55 lr 0.000201 time 0.2671 (0.3073) loss 4.0729 (3.2614) grad_norm 2.1647 (2.1492) [2021-04-16 12:21:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][300/1251] eta 0:04:51 lr 0.000201 time 0.2915 (0.3063) loss 2.8637 (3.2510) grad_norm 2.1069 (2.1503) [2021-04-16 12:21:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][310/1251] eta 0:04:47 lr 0.000201 time 0.2665 (0.3055) loss 3.0404 (3.2430) grad_norm 2.0536 (2.1509) [2021-04-16 12:21:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][320/1251] eta 0:04:43 lr 0.000201 time 0.2886 (0.3046) loss 3.6194 (3.2388) grad_norm 2.2825 (2.1494) [2021-04-16 12:21:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][330/1251] eta 0:04:39 lr 0.000201 time 0.2815 (0.3040) loss 3.8066 (3.2365) grad_norm 2.3768 (2.1497) [2021-04-16 12:21:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][340/1251] eta 0:04:36 lr 0.000200 time 0.2890 (0.3031) loss 2.7134 (3.2369) grad_norm 2.9669 (2.1550) [2021-04-16 12:21:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][350/1251] eta 0:04:32 lr 0.000200 time 0.2943 (0.3024) loss 3.3541 (3.2376) grad_norm 2.4951 (inf) [2021-04-16 12:21:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][360/1251] eta 0:04:29 lr 0.000200 time 0.2614 (0.3023) loss 2.3850 (3.2321) grad_norm 1.9542 (inf) [2021-04-16 12:21:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][370/1251] eta 0:04:26 lr 0.000200 time 0.2934 (0.3022) loss 3.5655 (3.2362) grad_norm 1.8518 (inf) [2021-04-16 12:21:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][380/1251] eta 0:04:22 lr 0.000200 time 0.2497 (0.3014) loss 2.7665 (3.2395) grad_norm 1.9771 (inf) [2021-04-16 12:21:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][390/1251] eta 0:04:19 lr 0.000200 time 0.2651 (0.3011) loss 3.5620 (3.2419) grad_norm 2.2420 (inf) [2021-04-16 12:21:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][400/1251] eta 0:04:15 lr 0.000200 time 0.2920 (0.3006) loss 3.2334 (3.2481) grad_norm 2.2123 (inf) [2021-04-16 12:21:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][410/1251] eta 0:04:12 lr 0.000200 time 0.2771 (0.2999) loss 3.8375 (3.2555) grad_norm 2.1103 (inf) [2021-04-16 12:21:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][420/1251] eta 0:04:08 lr 0.000200 time 0.2815 (0.2994) loss 3.7788 (3.2557) grad_norm 2.0706 (inf) [2021-04-16 12:21:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][430/1251] eta 0:04:05 lr 0.000200 time 0.2612 (0.2990) loss 3.3757 (3.2534) grad_norm 2.1887 (inf) [2021-04-16 12:21:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][440/1251] eta 0:04:02 lr 0.000200 time 0.2791 (0.2986) loss 3.7287 (3.2463) grad_norm 2.8197 (inf) [2021-04-16 12:21:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][450/1251] eta 0:03:58 lr 0.000200 time 0.2762 (0.2983) loss 3.4514 (3.2412) grad_norm 2.2659 (inf) [2021-04-16 12:21:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][460/1251] eta 0:03:55 lr 0.000200 time 0.2985 (0.2979) loss 3.6433 (3.2412) grad_norm 2.0058 (inf) [2021-04-16 12:21:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][470/1251] eta 0:03:52 lr 0.000200 time 0.2717 (0.2974) loss 3.6684 (3.2469) grad_norm 2.0703 (inf) [2021-04-16 12:21:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][480/1251] eta 0:03:49 lr 0.000200 time 0.2838 (0.2971) loss 4.0521 (3.2466) grad_norm 2.3567 (inf) [2021-04-16 12:22:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][490/1251] eta 0:03:45 lr 0.000200 time 0.2768 (0.2969) loss 3.1639 (3.2386) grad_norm 2.1403 (inf) [2021-04-16 12:22:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][500/1251] eta 0:03:42 lr 0.000200 time 0.2474 (0.2965) loss 3.5659 (3.2424) grad_norm 2.2683 (inf) [2021-04-16 12:22:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][510/1251] eta 0:03:39 lr 0.000200 time 0.2881 (0.2961) loss 3.6743 (3.2477) grad_norm 1.9457 (inf) [2021-04-16 12:22:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][520/1251] eta 0:03:36 lr 0.000200 time 0.2911 (0.2958) loss 2.7180 (3.2491) grad_norm 2.1075 (inf) [2021-04-16 12:22:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][530/1251] eta 0:03:33 lr 0.000200 time 0.2915 (0.2955) loss 3.3336 (3.2529) grad_norm 2.0701 (inf) [2021-04-16 12:22:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][540/1251] eta 0:03:29 lr 0.000200 time 0.2750 (0.2951) loss 3.1233 (3.2540) grad_norm 2.4099 (inf) [2021-04-16 12:22:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][550/1251] eta 0:03:26 lr 0.000200 time 0.2648 (0.2948) loss 2.3500 (3.2509) grad_norm 1.9710 (inf) [2021-04-16 12:22:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][560/1251] eta 0:03:23 lr 0.000200 time 0.2795 (0.2945) loss 4.0760 (3.2537) grad_norm 2.1700 (inf) [2021-04-16 12:22:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][570/1251] eta 0:03:20 lr 0.000200 time 0.2786 (0.2945) loss 3.6652 (3.2538) grad_norm 1.8413 (inf) [2021-04-16 12:22:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][580/1251] eta 0:03:17 lr 0.000200 time 0.2742 (0.2943) loss 3.5341 (3.2559) grad_norm 2.2562 (inf) [2021-04-16 12:22:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][590/1251] eta 0:03:14 lr 0.000200 time 0.4600 (0.2944) loss 3.3424 (3.2584) grad_norm 2.1541 (inf) [2021-04-16 12:22:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][600/1251] eta 0:03:11 lr 0.000200 time 0.2726 (0.2941) loss 2.8084 (3.2565) grad_norm 1.8996 (inf) [2021-04-16 12:22:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][610/1251] eta 0:03:08 lr 0.000200 time 0.2982 (0.2939) loss 2.9962 (3.2561) grad_norm 2.1034 (inf) [2021-04-16 12:22:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][620/1251] eta 0:03:05 lr 0.000200 time 0.2906 (0.2939) loss 2.3268 (3.2567) grad_norm 2.1123 (inf) [2021-04-16 12:22:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][630/1251] eta 0:03:02 lr 0.000200 time 0.2728 (0.2938) loss 3.2468 (3.2538) grad_norm 1.9006 (inf) [2021-04-16 12:22:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][640/1251] eta 0:02:59 lr 0.000200 time 0.2880 (0.2937) loss 3.8327 (3.2542) grad_norm 2.3287 (inf) [2021-04-16 12:22:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][650/1251] eta 0:02:56 lr 0.000199 time 0.2798 (0.2937) loss 3.3241 (3.2548) grad_norm 1.9548 (inf) [2021-04-16 12:22:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][660/1251] eta 0:02:53 lr 0.000199 time 0.2740 (0.2935) loss 3.7105 (3.2553) grad_norm 2.3646 (inf) [2021-04-16 12:22:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][670/1251] eta 0:02:50 lr 0.000199 time 0.2652 (0.2933) loss 3.6420 (3.2564) grad_norm 2.3728 (inf) [2021-04-16 12:22:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][680/1251] eta 0:02:47 lr 0.000199 time 0.2785 (0.2931) loss 3.0936 (3.2569) grad_norm 2.0191 (inf) [2021-04-16 12:22:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][690/1251] eta 0:02:44 lr 0.000199 time 0.2719 (0.2931) loss 3.1491 (3.2551) grad_norm 2.0186 (inf) [2021-04-16 12:22:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][700/1251] eta 0:02:41 lr 0.000199 time 0.2767 (0.2929) loss 2.9807 (3.2537) grad_norm 1.9862 (inf) [2021-04-16 12:23:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][710/1251] eta 0:02:38 lr 0.000199 time 0.3097 (0.2927) loss 3.7071 (3.2547) grad_norm 2.2906 (inf) [2021-04-16 12:23:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][720/1251] eta 0:02:35 lr 0.000199 time 0.2515 (0.2927) loss 4.2919 (3.2570) grad_norm 2.1516 (inf) [2021-04-16 12:23:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][730/1251] eta 0:02:32 lr 0.000199 time 0.2495 (0.2926) loss 2.4363 (3.2589) grad_norm 2.4317 (inf) [2021-04-16 12:23:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][740/1251] eta 0:02:29 lr 0.000199 time 0.2797 (0.2924) loss 2.7566 (3.2597) grad_norm 1.9407 (inf) [2021-04-16 12:23:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][750/1251] eta 0:02:26 lr 0.000199 time 0.2749 (0.2922) loss 2.8102 (3.2532) grad_norm 2.3158 (inf) [2021-04-16 12:23:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][760/1251] eta 0:02:23 lr 0.000199 time 0.2691 (0.2920) loss 3.2000 (3.2509) grad_norm 2.2480 (inf) [2021-04-16 12:23:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][770/1251] eta 0:02:20 lr 0.000199 time 0.2682 (0.2917) loss 3.5347 (3.2497) grad_norm 2.1114 (inf) [2021-04-16 12:23:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][780/1251] eta 0:02:17 lr 0.000199 time 0.2880 (0.2915) loss 3.4582 (3.2499) grad_norm 2.1057 (inf) [2021-04-16 12:23:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][790/1251] eta 0:02:14 lr 0.000199 time 0.2770 (0.2914) loss 2.8812 (3.2485) grad_norm 2.1976 (inf) [2021-04-16 12:23:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][800/1251] eta 0:02:11 lr 0.000199 time 0.2832 (0.2913) loss 2.8198 (3.2476) grad_norm 1.9445 (inf) [2021-04-16 12:23:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][810/1251] eta 0:02:08 lr 0.000199 time 0.2553 (0.2911) loss 3.4025 (3.2496) grad_norm 2.1218 (inf) [2021-04-16 12:23:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][820/1251] eta 0:02:05 lr 0.000199 time 0.2967 (0.2910) loss 2.7878 (3.2474) grad_norm 2.1502 (inf) [2021-04-16 12:23:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][830/1251] eta 0:02:02 lr 0.000199 time 0.2764 (0.2908) loss 3.4489 (3.2481) grad_norm 2.0114 (inf) [2021-04-16 12:23:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][840/1251] eta 0:01:59 lr 0.000199 time 0.2914 (0.2907) loss 3.4566 (3.2449) grad_norm 2.1787 (inf) [2021-04-16 12:23:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][850/1251] eta 0:01:56 lr 0.000199 time 0.2816 (0.2905) loss 2.4523 (3.2449) grad_norm 2.1766 (inf) [2021-04-16 12:23:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][860/1251] eta 0:01:53 lr 0.000199 time 0.2679 (0.2903) loss 2.8228 (3.2464) grad_norm 2.1140 (inf) [2021-04-16 12:23:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][870/1251] eta 0:01:50 lr 0.000199 time 0.2752 (0.2902) loss 2.1196 (3.2461) grad_norm 1.9546 (inf) [2021-04-16 12:23:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][880/1251] eta 0:01:47 lr 0.000199 time 0.2754 (0.2901) loss 3.1159 (3.2450) grad_norm 2.0845 (inf) [2021-04-16 12:23:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][890/1251] eta 0:01:44 lr 0.000199 time 0.2706 (0.2900) loss 2.6897 (3.2436) grad_norm 2.0051 (inf) [2021-04-16 12:23:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][900/1251] eta 0:01:41 lr 0.000199 time 0.2542 (0.2900) loss 3.0468 (3.2408) grad_norm 2.0773 (inf) [2021-04-16 12:23:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][910/1251] eta 0:01:38 lr 0.000199 time 0.2822 (0.2899) loss 2.4215 (3.2420) grad_norm 1.9902 (inf) [2021-04-16 12:24:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][920/1251] eta 0:01:35 lr 0.000199 time 0.2693 (0.2897) loss 3.0508 (3.2413) grad_norm 2.2411 (inf) [2021-04-16 12:24:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][930/1251] eta 0:01:32 lr 0.000199 time 0.3806 (0.2897) loss 3.4749 (3.2418) grad_norm 2.2213 (inf) [2021-04-16 12:24:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][940/1251] eta 0:01:30 lr 0.000199 time 0.3453 (0.2899) loss 3.5873 (3.2408) grad_norm 2.3402 (inf) [2021-04-16 12:24:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][950/1251] eta 0:01:27 lr 0.000199 time 0.2770 (0.2897) loss 2.5648 (3.2394) grad_norm 2.2066 (inf) [2021-04-16 12:24:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][960/1251] eta 0:01:24 lr 0.000198 time 0.2601 (0.2896) loss 3.4926 (3.2390) grad_norm 2.6071 (inf) [2021-04-16 12:24:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][970/1251] eta 0:01:21 lr 0.000198 time 0.2696 (0.2896) loss 3.4305 (3.2411) grad_norm 2.3315 (inf) [2021-04-16 12:24:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][980/1251] eta 0:01:18 lr 0.000198 time 0.2919 (0.2894) loss 2.8029 (3.2393) grad_norm 1.9254 (inf) [2021-04-16 12:24:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][990/1251] eta 0:01:15 lr 0.000198 time 0.2646 (0.2893) loss 3.5510 (3.2396) grad_norm 2.6846 (inf) [2021-04-16 12:24:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][1000/1251] eta 0:01:12 lr 0.000198 time 0.2788 (0.2893) loss 2.8186 (3.2400) grad_norm 2.2930 (inf) [2021-04-16 12:24:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][1010/1251] eta 0:01:09 lr 0.000198 time 0.2714 (0.2891) loss 3.2071 (3.2388) grad_norm 2.2326 (inf) [2021-04-16 12:24:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][1020/1251] eta 0:01:06 lr 0.000198 time 0.2670 (0.2890) loss 2.4838 (3.2364) grad_norm 2.6886 (inf) [2021-04-16 12:24:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][1030/1251] eta 0:01:03 lr 0.000198 time 0.2808 (0.2889) loss 3.5446 (3.2388) grad_norm 2.1339 (inf) [2021-04-16 12:24:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][1040/1251] eta 0:01:00 lr 0.000198 time 0.2737 (0.2888) loss 3.3746 (3.2382) grad_norm 1.9276 (inf) [2021-04-16 12:24:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][1050/1251] eta 0:00:58 lr 0.000198 time 0.2972 (0.2887) loss 3.0036 (3.2371) grad_norm 2.0411 (inf) [2021-04-16 12:24:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][1060/1251] eta 0:00:55 lr 0.000198 time 0.2894 (0.2885) loss 3.2160 (3.2359) grad_norm 2.3790 (inf) [2021-04-16 12:24:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][1070/1251] eta 0:00:52 lr 0.000198 time 0.2868 (0.2884) loss 3.4291 (3.2379) grad_norm 2.7154 (inf) [2021-04-16 12:24:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][1080/1251] eta 0:00:49 lr 0.000198 time 0.2980 (0.2883) loss 3.9943 (3.2407) grad_norm 2.7641 (inf) [2021-04-16 12:24:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][1090/1251] eta 0:00:46 lr 0.000198 time 0.2673 (0.2881) loss 3.6631 (3.2405) grad_norm 2.3648 (inf) [2021-04-16 12:24:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][1100/1251] eta 0:00:43 lr 0.000198 time 0.3199 (0.2881) loss 3.4945 (3.2400) grad_norm 2.3420 (inf) [2021-04-16 12:24:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][1110/1251] eta 0:00:40 lr 0.000198 time 0.2932 (0.2880) loss 3.3783 (3.2418) grad_norm 2.2406 (inf) [2021-04-16 12:24:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][1120/1251] eta 0:00:37 lr 0.000198 time 0.2574 (0.2879) loss 2.8990 (3.2425) grad_norm 2.0837 (inf) [2021-04-16 12:25:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][1130/1251] eta 0:00:34 lr 0.000198 time 0.2701 (0.2878) loss 3.7362 (3.2420) grad_norm 2.0047 (inf) [2021-04-16 12:25:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][1140/1251] eta 0:00:31 lr 0.000198 time 0.2683 (0.2877) loss 3.4294 (3.2429) grad_norm 2.2691 (inf) [2021-04-16 12:25:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][1150/1251] eta 0:00:29 lr 0.000198 time 0.2806 (0.2878) loss 2.8672 (3.2432) grad_norm 2.0801 (inf) [2021-04-16 12:25:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][1160/1251] eta 0:00:26 lr 0.000198 time 0.2740 (0.2878) loss 2.4116 (3.2419) grad_norm 1.9014 (inf) [2021-04-16 12:25:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][1170/1251] eta 0:00:23 lr 0.000198 time 0.2823 (0.2878) loss 3.5467 (3.2431) grad_norm 2.0257 (inf) [2021-04-16 12:25:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][1180/1251] eta 0:00:20 lr 0.000198 time 0.2632 (0.2877) loss 3.0863 (3.2441) grad_norm 2.2700 (inf) [2021-04-16 12:25:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][1190/1251] eta 0:00:17 lr 0.000198 time 0.4038 (0.2877) loss 3.2191 (3.2437) grad_norm 2.2741 (inf) [2021-04-16 12:25:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][1200/1251] eta 0:00:14 lr 0.000198 time 0.2510 (0.2876) loss 3.4871 (3.2464) grad_norm 2.3016 (inf) [2021-04-16 12:25:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][1210/1251] eta 0:00:11 lr 0.000198 time 0.2646 (0.2875) loss 3.6486 (3.2459) grad_norm 1.8770 (inf) [2021-04-16 12:25:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][1220/1251] eta 0:00:08 lr 0.000198 time 0.2857 (0.2875) loss 3.4591 (3.2457) grad_norm 1.9596 (inf) [2021-04-16 12:25:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][1230/1251] eta 0:00:06 lr 0.000198 time 0.2752 (0.2875) loss 3.2126 (3.2471) grad_norm 1.8126 (inf) [2021-04-16 12:25:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][1240/1251] eta 0:00:03 lr 0.000198 time 0.2490 (0.2873) loss 3.7833 (3.2476) grad_norm 2.3119 (inf) [2021-04-16 12:25:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [213/300][1250/1251] eta 0:00:00 lr 0.000198 time 0.2376 (0.2870) loss 3.4141 (3.2469) grad_norm 2.2001 (inf) [2021-04-16 12:25:37 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 213 training takes 0:06:02 [2021-04-16 12:25:37 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_213.pth saving...... [2021-04-16 12:25:48 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_213.pth saved !!! [2021-04-16 12:25:49 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.196 (1.196) Loss 0.9062 (0.9062) Acc@1 78.418 (78.418) Acc@5 94.336 (94.336) [2021-04-16 12:25:51 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.115 (0.249) Loss 0.7649 (0.8778) Acc@1 82.812 (79.590) Acc@5 95.801 (94.585) [2021-04-16 12:25:53 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.321 (0.220) Loss 0.8826 (0.8736) Acc@1 79.199 (79.432) Acc@5 95.410 (94.703) [2021-04-16 12:25:55 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.123 (0.218) Loss 0.8107 (0.8802) Acc@1 80.078 (79.231) Acc@5 95.312 (94.632) [2021-04-16 12:25:57 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.133 (0.212) Loss 0.8258 (0.8806) Acc@1 81.348 (79.216) Acc@5 95.117 (94.641) [2021-04-16 12:26:08 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 79.160 Acc@5 94.614 [2021-04-16 12:26:08 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 79.2% [2021-04-16 12:26:08 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 79.16% [2021-04-16 12:26:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][0/1251] eta 0:37:19 lr 0.000198 time 1.7899 (1.7899) loss 2.3905 (2.3905) grad_norm 2.2164 (2.2164) [2021-04-16 12:26:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][10/1251] eta 0:08:37 lr 0.000197 time 0.2813 (0.4171) loss 2.3670 (3.1698) grad_norm 2.2268 (2.1265) [2021-04-16 12:26:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][20/1251] eta 0:07:08 lr 0.000197 time 0.2923 (0.3483) loss 3.4105 (3.2673) grad_norm 2.1188 (2.2609) [2021-04-16 12:26:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][30/1251] eta 0:06:37 lr 0.000197 time 0.2680 (0.3254) loss 3.2718 (3.2421) grad_norm 1.9886 (2.2355) [2021-04-16 12:26:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][40/1251] eta 0:06:20 lr 0.000197 time 0.2828 (0.3142) loss 3.6539 (3.2383) grad_norm 1.8373 (2.1830) [2021-04-16 12:26:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][50/1251] eta 0:06:08 lr 0.000197 time 0.2761 (0.3064) loss 2.4753 (3.2475) grad_norm 1.8423 (2.1623) [2021-04-16 12:26:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][60/1251] eta 0:05:58 lr 0.000197 time 0.2648 (0.3014) loss 4.0205 (3.2198) grad_norm 1.8996 (2.1555) [2021-04-16 12:26:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][70/1251] eta 0:05:53 lr 0.000197 time 0.2862 (0.2990) loss 2.9072 (3.2010) grad_norm 1.8127 (2.1488) [2021-04-16 12:26:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][80/1251] eta 0:05:48 lr 0.000197 time 0.2733 (0.2972) loss 3.6428 (3.2244) grad_norm 2.4812 (2.1395) [2021-04-16 12:26:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][90/1251] eta 0:05:42 lr 0.000197 time 0.2539 (0.2954) loss 3.5462 (3.2130) grad_norm 2.0130 (2.1503) [2021-04-16 12:26:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][100/1251] eta 0:05:37 lr 0.000197 time 0.2715 (0.2935) loss 3.8505 (3.1990) grad_norm 2.1396 (2.1447) [2021-04-16 12:26:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][110/1251] eta 0:05:33 lr 0.000197 time 0.2802 (0.2920) loss 3.0636 (3.2181) grad_norm 2.1086 (2.1436) [2021-04-16 12:26:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][120/1251] eta 0:05:28 lr 0.000197 time 0.2725 (0.2904) loss 3.9317 (3.2381) grad_norm 2.4224 (2.1510) [2021-04-16 12:26:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][130/1251] eta 0:05:24 lr 0.000197 time 0.2702 (0.2892) loss 2.8104 (3.2425) grad_norm 2.3048 (2.1529) [2021-04-16 12:26:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][140/1251] eta 0:05:22 lr 0.000197 time 0.3805 (0.2902) loss 3.2336 (3.2419) grad_norm 2.3410 (2.1599) [2021-04-16 12:26:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][150/1251] eta 0:05:20 lr 0.000197 time 0.3949 (0.2910) loss 3.1336 (3.2350) grad_norm 2.2448 (2.1676) [2021-04-16 12:26:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][160/1251] eta 0:05:16 lr 0.000197 time 0.2780 (0.2897) loss 3.2458 (3.2272) grad_norm 2.4102 (2.1686) [2021-04-16 12:26:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][170/1251] eta 0:05:12 lr 0.000197 time 0.2616 (0.2894) loss 2.5537 (3.2320) grad_norm 2.3675 (2.1710) [2021-04-16 12:27:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][180/1251] eta 0:05:10 lr 0.000197 time 0.2836 (0.2895) loss 3.2048 (3.2333) grad_norm 2.3415 (2.1702) [2021-04-16 12:27:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][190/1251] eta 0:05:06 lr 0.000197 time 0.2948 (0.2892) loss 3.1056 (3.2263) grad_norm 2.3992 (2.1760) [2021-04-16 12:27:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][200/1251] eta 0:05:03 lr 0.000197 time 0.2787 (0.2889) loss 2.1334 (3.2179) grad_norm 1.9515 (2.1762) [2021-04-16 12:27:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][210/1251] eta 0:05:00 lr 0.000197 time 0.2679 (0.2883) loss 2.8835 (3.2218) grad_norm 2.5640 (2.1790) [2021-04-16 12:27:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][220/1251] eta 0:04:56 lr 0.000197 time 0.2951 (0.2880) loss 3.6263 (3.2268) grad_norm 2.0306 (2.1747) [2021-04-16 12:27:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][230/1251] eta 0:04:53 lr 0.000197 time 0.2826 (0.2875) loss 3.4694 (3.2297) grad_norm 2.5489 (2.1759) [2021-04-16 12:27:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][240/1251] eta 0:04:50 lr 0.000197 time 0.2876 (0.2869) loss 2.6701 (3.2343) grad_norm 2.3196 (2.1805) [2021-04-16 12:27:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][250/1251] eta 0:04:46 lr 0.000197 time 0.2738 (0.2866) loss 3.5674 (3.2380) grad_norm 2.3566 (2.1764) [2021-04-16 12:27:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][260/1251] eta 0:04:43 lr 0.000197 time 0.2441 (0.2861) loss 3.7747 (3.2383) grad_norm 1.9514 (2.1735) [2021-04-16 12:27:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][270/1251] eta 0:04:40 lr 0.000197 time 0.2757 (0.2862) loss 3.5132 (3.2399) grad_norm 2.3128 (2.1700) [2021-04-16 12:27:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][280/1251] eta 0:04:37 lr 0.000197 time 0.2549 (0.2857) loss 2.4637 (3.2328) grad_norm 1.9626 (2.1669) [2021-04-16 12:27:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][290/1251] eta 0:04:34 lr 0.000197 time 0.2671 (0.2856) loss 2.7521 (3.2399) grad_norm 2.0863 (2.1711) [2021-04-16 12:27:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][300/1251] eta 0:04:31 lr 0.000197 time 0.2826 (0.2853) loss 2.7225 (3.2341) grad_norm 2.2138 (2.1756) [2021-04-16 12:27:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][310/1251] eta 0:04:28 lr 0.000197 time 0.2862 (0.2850) loss 3.8045 (3.2264) grad_norm 2.0729 (2.1759) [2021-04-16 12:27:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][320/1251] eta 0:04:25 lr 0.000196 time 0.2879 (0.2850) loss 3.6823 (3.2301) grad_norm 2.1983 (2.1739) [2021-04-16 12:27:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][330/1251] eta 0:04:22 lr 0.000196 time 0.2474 (0.2848) loss 3.4092 (3.2330) grad_norm 2.3314 (2.1743) [2021-04-16 12:27:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][340/1251] eta 0:04:19 lr 0.000196 time 0.2504 (0.2846) loss 2.2911 (3.2291) grad_norm 2.1381 (2.1741) [2021-04-16 12:27:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][350/1251] eta 0:04:16 lr 0.000196 time 0.4313 (0.2850) loss 2.4453 (3.2272) grad_norm 2.0064 (2.1714) [2021-04-16 12:27:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][360/1251] eta 0:04:14 lr 0.000196 time 0.2675 (0.2851) loss 3.4156 (3.2313) grad_norm 2.1209 (2.1715) [2021-04-16 12:27:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][370/1251] eta 0:04:11 lr 0.000196 time 0.2621 (0.2852) loss 3.6865 (3.2358) grad_norm 2.1877 (2.1722) [2021-04-16 12:27:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][380/1251] eta 0:04:08 lr 0.000196 time 0.2923 (0.2852) loss 2.3167 (3.2317) grad_norm 2.5369 (2.1716) [2021-04-16 12:27:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][390/1251] eta 0:04:05 lr 0.000196 time 0.2586 (0.2853) loss 3.3489 (3.2282) grad_norm 2.2300 (2.1744) [2021-04-16 12:28:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][400/1251] eta 0:04:02 lr 0.000196 time 0.3227 (0.2853) loss 2.8131 (3.2293) grad_norm 2.3527 (2.1750) [2021-04-16 12:28:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][410/1251] eta 0:03:59 lr 0.000196 time 0.3015 (0.2852) loss 3.0996 (3.2270) grad_norm 2.3324 (2.1762) [2021-04-16 12:28:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][420/1251] eta 0:03:57 lr 0.000196 time 0.2565 (0.2852) loss 2.6574 (3.2333) grad_norm 2.0280 (2.1747) [2021-04-16 12:28:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][430/1251] eta 0:03:54 lr 0.000196 time 0.2556 (0.2853) loss 3.2105 (3.2349) grad_norm 2.2859 (2.1739) [2021-04-16 12:28:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][440/1251] eta 0:03:51 lr 0.000196 time 0.2783 (0.2852) loss 2.8995 (3.2375) grad_norm 1.9115 (2.1718) [2021-04-16 12:28:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][450/1251] eta 0:03:48 lr 0.000196 time 0.2646 (0.2850) loss 3.4296 (3.2359) grad_norm 2.0393 (2.1711) [2021-04-16 12:28:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][460/1251] eta 0:03:45 lr 0.000196 time 0.2775 (0.2848) loss 3.6139 (3.2412) grad_norm 2.0425 (2.1695) [2021-04-16 12:28:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][470/1251] eta 0:03:42 lr 0.000196 time 0.2648 (0.2847) loss 2.8294 (3.2363) grad_norm 2.1603 (2.1699) [2021-04-16 12:28:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][480/1251] eta 0:03:39 lr 0.000196 time 0.2879 (0.2846) loss 2.2999 (3.2299) grad_norm 1.9799 (2.1679) [2021-04-16 12:28:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][490/1251] eta 0:03:36 lr 0.000196 time 0.2725 (0.2843) loss 2.5231 (3.2301) grad_norm 1.9317 (2.1663) [2021-04-16 12:28:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][500/1251] eta 0:03:33 lr 0.000196 time 0.2582 (0.2842) loss 2.9709 (3.2231) grad_norm 2.1773 (2.1671) [2021-04-16 12:28:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][510/1251] eta 0:03:30 lr 0.000196 time 0.2823 (0.2841) loss 2.7693 (3.2252) grad_norm 2.0838 (2.1674) [2021-04-16 12:28:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][520/1251] eta 0:03:27 lr 0.000196 time 0.2687 (0.2841) loss 2.9939 (3.2273) grad_norm 2.1108 (2.1683) [2021-04-16 12:28:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][530/1251] eta 0:03:24 lr 0.000196 time 0.2889 (0.2839) loss 2.9506 (3.2207) grad_norm 2.4645 (2.1706) [2021-04-16 12:28:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][540/1251] eta 0:03:21 lr 0.000196 time 0.2581 (0.2837) loss 2.4039 (3.2241) grad_norm 2.1265 (2.1715) [2021-04-16 12:28:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][550/1251] eta 0:03:18 lr 0.000196 time 0.3016 (0.2836) loss 2.2371 (3.2248) grad_norm 1.9606 (2.1714) [2021-04-16 12:28:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][560/1251] eta 0:03:15 lr 0.000196 time 0.2778 (0.2834) loss 3.5287 (3.2271) grad_norm 2.1619 (2.1709) [2021-04-16 12:28:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][570/1251] eta 0:03:13 lr 0.000196 time 0.2683 (0.2836) loss 2.5292 (3.2277) grad_norm 2.5569 (2.1724) [2021-04-16 12:28:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][580/1251] eta 0:03:10 lr 0.000196 time 0.2797 (0.2840) loss 3.6908 (3.2296) grad_norm 2.1213 (2.1732) [2021-04-16 12:28:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][590/1251] eta 0:03:07 lr 0.000196 time 0.2660 (0.2839) loss 2.7785 (3.2302) grad_norm 1.9787 (2.1717) [2021-04-16 12:28:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][600/1251] eta 0:03:04 lr 0.000196 time 0.2682 (0.2838) loss 3.5902 (3.2307) grad_norm 2.2061 (2.1702) [2021-04-16 12:29:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][610/1251] eta 0:03:01 lr 0.000196 time 0.2673 (0.2837) loss 3.3430 (3.2276) grad_norm 2.3360 (2.1691) [2021-04-16 12:29:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][620/1251] eta 0:02:58 lr 0.000196 time 0.3159 (0.2837) loss 3.8118 (3.2328) grad_norm 2.1796 (2.1705) [2021-04-16 12:29:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][630/1251] eta 0:02:56 lr 0.000195 time 0.2793 (0.2835) loss 3.0108 (3.2285) grad_norm 2.1892 (2.1711) [2021-04-16 12:29:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][640/1251] eta 0:02:53 lr 0.000195 time 0.2614 (0.2834) loss 3.8940 (3.2317) grad_norm 2.2781 (2.1722) [2021-04-16 12:29:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][650/1251] eta 0:02:50 lr 0.000195 time 0.2786 (0.2834) loss 3.2381 (3.2358) grad_norm 1.7920 (2.1736) [2021-04-16 12:29:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][660/1251] eta 0:02:47 lr 0.000195 time 0.2909 (0.2833) loss 3.2143 (3.2389) grad_norm 2.2761 (2.1718) [2021-04-16 12:29:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][670/1251] eta 0:02:44 lr 0.000195 time 0.2888 (0.2831) loss 2.4824 (3.2364) grad_norm 2.1968 (2.1739) [2021-04-16 12:29:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][680/1251] eta 0:02:41 lr 0.000195 time 0.2595 (0.2830) loss 3.6502 (3.2361) grad_norm 2.1928 (2.1725) [2021-04-16 12:29:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][690/1251] eta 0:02:38 lr 0.000195 time 0.2593 (0.2829) loss 3.2420 (3.2354) grad_norm 2.8677 (2.1743) [2021-04-16 12:29:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][700/1251] eta 0:02:35 lr 0.000195 time 0.2631 (0.2828) loss 3.4290 (3.2343) grad_norm 3.0852 (2.1749) [2021-04-16 12:29:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][710/1251] eta 0:02:32 lr 0.000195 time 0.2765 (0.2828) loss 2.6102 (3.2371) grad_norm 1.9714 (2.1745) [2021-04-16 12:29:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][720/1251] eta 0:02:30 lr 0.000195 time 0.2893 (0.2832) loss 2.3545 (3.2398) grad_norm 2.0459 (2.1746) [2021-04-16 12:29:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][730/1251] eta 0:02:27 lr 0.000195 time 0.2677 (0.2831) loss 3.1809 (3.2397) grad_norm 2.0306 (2.1736) [2021-04-16 12:29:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][740/1251] eta 0:02:24 lr 0.000195 time 0.2572 (0.2829) loss 3.6298 (3.2411) grad_norm 1.9462 (2.1714) [2021-04-16 12:29:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][750/1251] eta 0:02:21 lr 0.000195 time 0.2756 (0.2829) loss 3.5099 (3.2418) grad_norm 2.1588 (2.1717) [2021-04-16 12:29:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][760/1251] eta 0:02:18 lr 0.000195 time 0.2760 (0.2830) loss 2.6910 (3.2453) grad_norm 1.9448 (2.1717) [2021-04-16 12:29:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][770/1251] eta 0:02:16 lr 0.000195 time 0.2551 (0.2829) loss 3.4937 (3.2435) grad_norm 2.0713 (2.1734) [2021-04-16 12:29:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][780/1251] eta 0:02:13 lr 0.000195 time 0.2461 (0.2827) loss 3.1752 (3.2431) grad_norm 1.9678 (2.1733) [2021-04-16 12:29:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][790/1251] eta 0:02:10 lr 0.000195 time 0.2504 (0.2827) loss 3.6298 (3.2438) grad_norm 2.1379 (2.1782) [2021-04-16 12:29:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][800/1251] eta 0:02:07 lr 0.000195 time 0.2839 (0.2827) loss 3.7754 (3.2425) grad_norm 2.1291 (2.1779) [2021-04-16 12:29:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][810/1251] eta 0:02:04 lr 0.000195 time 0.2675 (0.2827) loss 2.3565 (3.2448) grad_norm 2.2453 (2.1781) [2021-04-16 12:30:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][820/1251] eta 0:02:01 lr 0.000195 time 0.2878 (0.2826) loss 3.5322 (3.2459) grad_norm 2.4522 (2.1783) [2021-04-16 12:30:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][830/1251] eta 0:01:58 lr 0.000195 time 0.2788 (0.2825) loss 2.5487 (3.2467) grad_norm 2.1077 (2.1794) [2021-04-16 12:30:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][840/1251] eta 0:01:56 lr 0.000195 time 0.2662 (0.2824) loss 2.7767 (3.2474) grad_norm 2.1002 (2.1806) [2021-04-16 12:30:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][850/1251] eta 0:01:53 lr 0.000195 time 0.2603 (0.2824) loss 3.6108 (3.2499) grad_norm 1.7431 (2.1798) [2021-04-16 12:30:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][860/1251] eta 0:01:50 lr 0.000195 time 0.2963 (0.2823) loss 3.0464 (3.2498) grad_norm 2.5060 (2.1805) [2021-04-16 12:30:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][870/1251] eta 0:01:47 lr 0.000195 time 0.3048 (0.2822) loss 3.4455 (3.2519) grad_norm 1.8088 (2.1822) [2021-04-16 12:30:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][880/1251] eta 0:01:44 lr 0.000195 time 0.2741 (0.2822) loss 3.6915 (3.2525) grad_norm 1.8718 (2.1823) [2021-04-16 12:30:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][890/1251] eta 0:01:41 lr 0.000195 time 0.2711 (0.2821) loss 2.4864 (3.2527) grad_norm 2.1503 (2.1819) [2021-04-16 12:30:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][900/1251] eta 0:01:38 lr 0.000195 time 0.2705 (0.2820) loss 2.9000 (3.2521) grad_norm 1.8800 (2.1815) [2021-04-16 12:30:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][910/1251] eta 0:01:36 lr 0.000195 time 0.2555 (0.2820) loss 3.5453 (3.2496) grad_norm 2.1965 (2.1811) [2021-04-16 12:30:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][920/1251] eta 0:01:33 lr 0.000195 time 0.2818 (0.2819) loss 3.3415 (3.2486) grad_norm 2.6460 (2.1801) [2021-04-16 12:30:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][930/1251] eta 0:01:30 lr 0.000195 time 0.2845 (0.2819) loss 2.9053 (3.2471) grad_norm 1.8927 (2.1792) [2021-04-16 12:30:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][940/1251] eta 0:01:27 lr 0.000194 time 0.2889 (0.2820) loss 3.5376 (3.2479) grad_norm 2.2170 (2.1786) [2021-04-16 12:30:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][950/1251] eta 0:01:24 lr 0.000194 time 0.2751 (0.2821) loss 3.0082 (3.2506) grad_norm 2.2162 (2.1814) [2021-04-16 12:30:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][960/1251] eta 0:01:22 lr 0.000194 time 0.2777 (0.2821) loss 3.4718 (3.2516) grad_norm 1.9481 (2.1842) [2021-04-16 12:30:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][970/1251] eta 0:01:19 lr 0.000194 time 0.2978 (0.2821) loss 3.6934 (3.2537) grad_norm 2.2658 (2.1839) [2021-04-16 12:30:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][980/1251] eta 0:01:16 lr 0.000194 time 0.3069 (0.2820) loss 3.2224 (3.2516) grad_norm 2.2161 (2.1827) [2021-04-16 12:30:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][990/1251] eta 0:01:13 lr 0.000194 time 0.2948 (0.2820) loss 2.4610 (3.2511) grad_norm 2.1859 (2.1827) [2021-04-16 12:30:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][1000/1251] eta 0:01:10 lr 0.000194 time 0.2629 (0.2820) loss 3.4272 (3.2510) grad_norm 1.8780 (2.1830) [2021-04-16 12:30:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][1010/1251] eta 0:01:07 lr 0.000194 time 0.2524 (0.2820) loss 2.5415 (3.2490) grad_norm 1.9947 (2.1816) [2021-04-16 12:30:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][1020/1251] eta 0:01:05 lr 0.000194 time 0.2672 (0.2819) loss 4.1403 (3.2484) grad_norm 2.0032 (2.1801) [2021-04-16 12:30:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][1030/1251] eta 0:01:02 lr 0.000194 time 0.2844 (0.2819) loss 2.2147 (3.2473) grad_norm 2.3163 (2.1790) [2021-04-16 12:31:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][1040/1251] eta 0:00:59 lr 0.000194 time 0.2631 (0.2819) loss 3.4545 (3.2454) grad_norm 2.5716 (2.1791) [2021-04-16 12:31:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][1050/1251] eta 0:00:56 lr 0.000194 time 0.2814 (0.2818) loss 4.0355 (3.2468) grad_norm 2.2061 (2.1792) [2021-04-16 12:31:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][1060/1251] eta 0:00:53 lr 0.000194 time 0.2809 (0.2818) loss 3.2039 (3.2472) grad_norm 2.0935 (2.1781) [2021-04-16 12:31:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][1070/1251] eta 0:00:50 lr 0.000194 time 0.3052 (0.2818) loss 3.7047 (3.2455) grad_norm 2.1696 (2.1771) [2021-04-16 12:31:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][1080/1251] eta 0:00:48 lr 0.000194 time 0.2720 (0.2817) loss 3.8221 (3.2465) grad_norm 2.1028 (2.1765) [2021-04-16 12:31:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][1090/1251] eta 0:00:45 lr 0.000194 time 0.2888 (0.2817) loss 3.6077 (3.2474) grad_norm 2.1507 (2.1761) [2021-04-16 12:31:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][1100/1251] eta 0:00:42 lr 0.000194 time 0.2932 (0.2817) loss 3.6888 (3.2454) grad_norm 2.5958 (2.1764) [2021-04-16 12:31:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][1110/1251] eta 0:00:39 lr 0.000194 time 0.2834 (0.2816) loss 1.9915 (3.2437) grad_norm 1.9770 (2.1759) [2021-04-16 12:31:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][1120/1251] eta 0:00:36 lr 0.000194 time 0.2785 (0.2816) loss 3.1015 (3.2425) grad_norm 2.1200 (2.1760) [2021-04-16 12:31:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][1130/1251] eta 0:00:34 lr 0.000194 time 0.2828 (0.2816) loss 3.5293 (3.2426) grad_norm 2.3016 (2.1764) [2021-04-16 12:31:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][1140/1251] eta 0:00:31 lr 0.000194 time 0.2699 (0.2816) loss 2.3785 (3.2410) grad_norm 2.1476 (2.1766) [2021-04-16 12:31:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][1150/1251] eta 0:00:28 lr 0.000194 time 0.2794 (0.2818) loss 2.8578 (3.2395) grad_norm 1.9691 (2.1768) [2021-04-16 12:31:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][1160/1251] eta 0:00:25 lr 0.000194 time 0.2430 (0.2818) loss 2.7638 (3.2380) grad_norm nan (nan) [2021-04-16 12:31:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][1170/1251] eta 0:00:22 lr 0.000194 time 0.2696 (0.2817) loss 3.1113 (3.2384) grad_norm 2.1266 (nan) [2021-04-16 12:31:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][1180/1251] eta 0:00:19 lr 0.000194 time 0.2750 (0.2817) loss 2.3347 (3.2385) grad_norm 2.1727 (nan) [2021-04-16 12:31:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][1190/1251] eta 0:00:17 lr 0.000194 time 0.2659 (0.2816) loss 2.4224 (3.2363) grad_norm 1.8345 (nan) [2021-04-16 12:31:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][1200/1251] eta 0:00:14 lr 0.000194 time 0.3126 (0.2816) loss 3.4653 (3.2374) grad_norm 1.9292 (nan) [2021-04-16 12:31:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][1210/1251] eta 0:00:11 lr 0.000194 time 0.2758 (0.2816) loss 3.4143 (3.2395) grad_norm 1.9290 (nan) [2021-04-16 12:31:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][1220/1251] eta 0:00:08 lr 0.000194 time 0.2752 (0.2816) loss 3.8179 (3.2398) grad_norm 2.2552 (nan) [2021-04-16 12:31:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][1230/1251] eta 0:00:05 lr 0.000194 time 0.2730 (0.2816) loss 3.3471 (3.2387) grad_norm 2.0598 (nan) [2021-04-16 12:31:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][1240/1251] eta 0:00:03 lr 0.000194 time 0.2483 (0.2814) loss 3.8797 (3.2394) grad_norm 2.0881 (nan) [2021-04-16 12:31:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [214/300][1250/1251] eta 0:00:00 lr 0.000193 time 0.2670 (0.2812) loss 3.4507 (3.2385) grad_norm 1.9975 (nan) [2021-04-16 12:32:09 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 214 training takes 0:06:00 [2021-04-16 12:32:09 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_214.pth saving...... [2021-04-16 12:32:29 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_214.pth saved !!! [2021-04-16 12:32:30 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.183 (1.183) Loss 0.8716 (0.8716) Acc@1 80.664 (80.664) Acc@5 93.848 (93.848) [2021-04-16 12:32:32 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.462 (0.258) Loss 0.8517 (0.8691) Acc@1 79.883 (79.492) Acc@5 94.531 (94.593) [2021-04-16 12:32:33 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.079 (0.212) Loss 0.8545 (0.8788) Acc@1 79.395 (79.064) Acc@5 95.020 (94.568) [2021-04-16 12:32:36 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.180 (0.226) Loss 0.8853 (0.8753) Acc@1 79.297 (79.237) Acc@5 94.629 (94.657) [2021-04-16 12:32:38 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.216) Loss 0.8993 (0.8772) Acc@1 78.809 (79.140) Acc@5 93.750 (94.700) [2021-04-16 12:32:45 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 79.048 Acc@5 94.686 [2021-04-16 12:32:45 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 79.0% [2021-04-16 12:32:45 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 79.16% [2021-04-16 12:32:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][0/1251] eta 4:31:23 lr 0.000193 time 13.0163 (13.0163) loss 3.0552 (3.0552) grad_norm 2.1512 (2.1512) [2021-04-16 12:33:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][10/1251] eta 0:29:30 lr 0.000193 time 0.2778 (1.4269) loss 2.2697 (3.1601) grad_norm 2.0460 (2.1533) [2021-04-16 12:33:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][20/1251] eta 0:18:07 lr 0.000193 time 0.2813 (0.8830) loss 2.3316 (3.1606) grad_norm 2.0634 (2.1519) [2021-04-16 12:33:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][30/1251] eta 0:13:59 lr 0.000193 time 0.2641 (0.6876) loss 3.5655 (3.1771) grad_norm 1.8609 (2.1181) [2021-04-16 12:33:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][40/1251] eta 0:11:50 lr 0.000193 time 0.2811 (0.5869) loss 3.6576 (3.2414) grad_norm 1.8591 (2.1070) [2021-04-16 12:33:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][50/1251] eta 0:10:31 lr 0.000193 time 0.2540 (0.5258) loss 2.2300 (3.2353) grad_norm 2.3332 (2.1202) [2021-04-16 12:33:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][60/1251] eta 0:09:39 lr 0.000193 time 0.2600 (0.4864) loss 3.0066 (3.2393) grad_norm 2.4081 (2.1298) [2021-04-16 12:33:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][70/1251] eta 0:09:01 lr 0.000193 time 0.2634 (0.4585) loss 3.3862 (3.2457) grad_norm 2.0283 (2.1305) [2021-04-16 12:33:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][80/1251] eta 0:08:30 lr 0.000193 time 0.2973 (0.4356) loss 4.0389 (3.2646) grad_norm 2.4427 (2.1269) [2021-04-16 12:33:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][90/1251] eta 0:08:05 lr 0.000193 time 0.2600 (0.4181) loss 3.5341 (3.2479) grad_norm 2.0651 (2.1350) [2021-04-16 12:33:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][100/1251] eta 0:07:44 lr 0.000193 time 0.2758 (0.4038) loss 3.5108 (3.2540) grad_norm 2.2697 (2.1569) [2021-04-16 12:33:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][110/1251] eta 0:07:28 lr 0.000193 time 0.3067 (0.3927) loss 3.0655 (3.2353) grad_norm 2.1437 (2.1519) [2021-04-16 12:33:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][120/1251] eta 0:07:13 lr 0.000193 time 0.2733 (0.3832) loss 2.4939 (3.2051) grad_norm 2.1935 (2.1474) [2021-04-16 12:33:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][130/1251] eta 0:07:01 lr 0.000193 time 0.2645 (0.3760) loss 3.8857 (3.2116) grad_norm 2.1251 (2.1459) [2021-04-16 12:33:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][140/1251] eta 0:06:51 lr 0.000193 time 0.2516 (0.3701) loss 3.1158 (3.2307) grad_norm 2.0361 (2.1477) [2021-04-16 12:33:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][150/1251] eta 0:06:40 lr 0.000193 time 0.2922 (0.3640) loss 2.3336 (3.2322) grad_norm 2.4441 (2.1525) [2021-04-16 12:33:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][160/1251] eta 0:06:32 lr 0.000193 time 0.2737 (0.3594) loss 4.0667 (3.2261) grad_norm 3.1398 (2.1601) [2021-04-16 12:33:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][170/1251] eta 0:06:23 lr 0.000193 time 0.2834 (0.3549) loss 2.5953 (3.2306) grad_norm 1.9253 (2.1627) [2021-04-16 12:33:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][180/1251] eta 0:06:16 lr 0.000193 time 0.2746 (0.3514) loss 3.4688 (3.2365) grad_norm 1.9396 (2.1689) [2021-04-16 12:33:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][190/1251] eta 0:06:08 lr 0.000193 time 0.2532 (0.3475) loss 3.2137 (3.2406) grad_norm 2.0401 (2.1663) [2021-04-16 12:33:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][200/1251] eta 0:06:02 lr 0.000193 time 0.2528 (0.3445) loss 2.9975 (3.2309) grad_norm 2.2966 (2.1757) [2021-04-16 12:33:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][210/1251] eta 0:05:55 lr 0.000193 time 0.2788 (0.3413) loss 2.8582 (3.2329) grad_norm 2.4311 (2.1847) [2021-04-16 12:34:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][220/1251] eta 0:05:49 lr 0.000193 time 0.2445 (0.3389) loss 3.2036 (3.2367) grad_norm 2.1277 (2.1846) [2021-04-16 12:34:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][230/1251] eta 0:05:43 lr 0.000193 time 0.2640 (0.3361) loss 2.9131 (3.2343) grad_norm 1.9741 (2.1824) [2021-04-16 12:34:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][240/1251] eta 0:05:37 lr 0.000193 time 0.2795 (0.3337) loss 2.5479 (3.2182) grad_norm 2.1935 (2.1855) [2021-04-16 12:34:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][250/1251] eta 0:05:31 lr 0.000193 time 0.2851 (0.3315) loss 3.7363 (3.2205) grad_norm 2.1163 (2.1942) [2021-04-16 12:34:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][260/1251] eta 0:05:26 lr 0.000193 time 0.2888 (0.3293) loss 3.7219 (3.2268) grad_norm 2.2709 (2.1912) [2021-04-16 12:34:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][270/1251] eta 0:05:20 lr 0.000193 time 0.2724 (0.3271) loss 3.6259 (3.2359) grad_norm 2.0340 (2.1989) [2021-04-16 12:34:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][280/1251] eta 0:05:15 lr 0.000193 time 0.2629 (0.3254) loss 3.1427 (3.2345) grad_norm 1.8552 (2.2035) [2021-04-16 12:34:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][290/1251] eta 0:05:11 lr 0.000193 time 0.2866 (0.3238) loss 2.4569 (3.2444) grad_norm 2.2648 (2.2108) [2021-04-16 12:34:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][300/1251] eta 0:05:06 lr 0.000193 time 0.2517 (0.3221) loss 3.2047 (3.2466) grad_norm 1.9940 (2.2110) [2021-04-16 12:34:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][310/1251] eta 0:05:01 lr 0.000192 time 0.2719 (0.3208) loss 2.8709 (3.2502) grad_norm 2.1714 (2.2116) [2021-04-16 12:34:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][320/1251] eta 0:04:57 lr 0.000192 time 0.2657 (0.3194) loss 3.5052 (3.2402) grad_norm 1.8932 (2.2141) [2021-04-16 12:34:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][330/1251] eta 0:04:52 lr 0.000192 time 0.2511 (0.3181) loss 2.4541 (3.2353) grad_norm 1.9518 (2.2097) [2021-04-16 12:34:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][340/1251] eta 0:04:48 lr 0.000192 time 0.3140 (0.3171) loss 3.3743 (3.2346) grad_norm 1.8188 (2.2072) [2021-04-16 12:34:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][350/1251] eta 0:04:44 lr 0.000192 time 0.2680 (0.3160) loss 3.1328 (3.2291) grad_norm 1.8952 (2.2066) [2021-04-16 12:34:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][360/1251] eta 0:04:41 lr 0.000192 time 0.2585 (0.3159) loss 4.0698 (3.2291) grad_norm 1.9376 (2.2059) [2021-04-16 12:34:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][370/1251] eta 0:04:37 lr 0.000192 time 0.2835 (0.3150) loss 3.4379 (3.2377) grad_norm 2.0678 (2.2038) [2021-04-16 12:34:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][380/1251] eta 0:04:33 lr 0.000192 time 0.2704 (0.3140) loss 3.9471 (3.2443) grad_norm 2.1168 (2.2023) [2021-04-16 12:34:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][390/1251] eta 0:04:29 lr 0.000192 time 0.2787 (0.3131) loss 3.4747 (3.2375) grad_norm 2.0767 (2.2011) [2021-04-16 12:34:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][400/1251] eta 0:04:25 lr 0.000192 time 0.2918 (0.3123) loss 3.3073 (3.2409) grad_norm 2.1889 (2.2024) [2021-04-16 12:34:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][410/1251] eta 0:04:22 lr 0.000192 time 0.2762 (0.3116) loss 3.1830 (3.2458) grad_norm 1.9398 (2.2059) [2021-04-16 12:34:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][420/1251] eta 0:04:18 lr 0.000192 time 0.3083 (0.3108) loss 2.1267 (3.2425) grad_norm 2.3695 (2.2048) [2021-04-16 12:34:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][430/1251] eta 0:04:14 lr 0.000192 time 0.2472 (0.3101) loss 3.5625 (3.2436) grad_norm 2.3531 (2.2038) [2021-04-16 12:35:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][440/1251] eta 0:04:11 lr 0.000192 time 0.2966 (0.3095) loss 3.9627 (3.2407) grad_norm 1.9218 (2.2009) [2021-04-16 12:35:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][450/1251] eta 0:04:07 lr 0.000192 time 0.2661 (0.3088) loss 3.8241 (3.2420) grad_norm 2.1167 (2.1988) [2021-04-16 12:35:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][460/1251] eta 0:04:03 lr 0.000192 time 0.2825 (0.3081) loss 3.8667 (3.2459) grad_norm 2.5395 (2.2013) [2021-04-16 12:35:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][470/1251] eta 0:04:00 lr 0.000192 time 0.2896 (0.3075) loss 3.0892 (3.2423) grad_norm 1.9145 (2.2005) [2021-04-16 12:35:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][480/1251] eta 0:03:56 lr 0.000192 time 0.2732 (0.3069) loss 2.9621 (3.2483) grad_norm 2.3327 (2.1995) [2021-04-16 12:35:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][490/1251] eta 0:03:53 lr 0.000192 time 0.2942 (0.3062) loss 2.3224 (3.2444) grad_norm 2.3117 (2.2017) [2021-04-16 12:35:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][500/1251] eta 0:03:49 lr 0.000192 time 0.2876 (0.3056) loss 3.7089 (3.2398) grad_norm 2.4813 (2.2035) [2021-04-16 12:35:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][510/1251] eta 0:03:45 lr 0.000192 time 0.2522 (0.3050) loss 3.0087 (3.2367) grad_norm 1.9611 (2.2019) [2021-04-16 12:35:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][520/1251] eta 0:03:42 lr 0.000192 time 0.2748 (0.3046) loss 2.4828 (3.2310) grad_norm 2.1723 (2.2025) [2021-04-16 12:35:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][530/1251] eta 0:03:39 lr 0.000192 time 0.2723 (0.3042) loss 3.5727 (3.2357) grad_norm 2.0264 (2.2015) [2021-04-16 12:35:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][540/1251] eta 0:03:35 lr 0.000192 time 0.2816 (0.3037) loss 3.5410 (3.2386) grad_norm 2.0081 (2.2042) [2021-04-16 12:35:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][550/1251] eta 0:03:32 lr 0.000192 time 0.2754 (0.3034) loss 2.1503 (3.2380) grad_norm 2.1611 (2.2035) [2021-04-16 12:35:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][560/1251] eta 0:03:29 lr 0.000192 time 0.3017 (0.3030) loss 3.2879 (3.2383) grad_norm 1.9335 (2.2028) [2021-04-16 12:35:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][570/1251] eta 0:03:26 lr 0.000192 time 0.2607 (0.3027) loss 3.2684 (3.2376) grad_norm 2.4684 (2.2025) [2021-04-16 12:35:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][580/1251] eta 0:03:23 lr 0.000192 time 0.2775 (0.3031) loss 3.7469 (3.2399) grad_norm 2.0505 (2.2017) [2021-04-16 12:35:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][590/1251] eta 0:03:20 lr 0.000192 time 0.3006 (0.3026) loss 2.3379 (3.2354) grad_norm 2.1223 (2.2011) [2021-04-16 12:35:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][600/1251] eta 0:03:16 lr 0.000192 time 0.2825 (0.3021) loss 3.2945 (3.2390) grad_norm 1.8853 (2.1996) [2021-04-16 12:35:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][610/1251] eta 0:03:13 lr 0.000192 time 0.2708 (0.3017) loss 3.4419 (3.2391) grad_norm 1.8980 (2.1979) [2021-04-16 12:35:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][620/1251] eta 0:03:10 lr 0.000191 time 0.2976 (0.3013) loss 3.7620 (3.2429) grad_norm 2.4037 (2.1981) [2021-04-16 12:35:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][630/1251] eta 0:03:06 lr 0.000191 time 0.2936 (0.3009) loss 2.3532 (3.2411) grad_norm 1.9685 (2.1974) [2021-04-16 12:35:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][640/1251] eta 0:03:03 lr 0.000191 time 0.2812 (0.3004) loss 3.3738 (3.2426) grad_norm 2.4789 (2.1980) [2021-04-16 12:36:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][650/1251] eta 0:03:00 lr 0.000191 time 0.2836 (0.3001) loss 2.4092 (3.2400) grad_norm 2.8648 (2.2007) [2021-04-16 12:36:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][660/1251] eta 0:02:57 lr 0.000191 time 0.2550 (0.2997) loss 3.5134 (3.2394) grad_norm 2.2270 (2.1993) [2021-04-16 12:36:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][670/1251] eta 0:02:53 lr 0.000191 time 0.2750 (0.2993) loss 3.8932 (3.2417) grad_norm 2.2746 (2.1992) [2021-04-16 12:36:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][680/1251] eta 0:02:50 lr 0.000191 time 0.2871 (0.2990) loss 3.0696 (3.2437) grad_norm 2.2784 (2.1988) [2021-04-16 12:36:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][690/1251] eta 0:02:47 lr 0.000191 time 0.2723 (0.2986) loss 3.2488 (3.2452) grad_norm 1.9268 (2.1972) [2021-04-16 12:36:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][700/1251] eta 0:02:44 lr 0.000191 time 0.2805 (0.2983) loss 2.8503 (3.2467) grad_norm 2.2611 (2.1970) [2021-04-16 12:36:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][710/1251] eta 0:02:41 lr 0.000191 time 0.2723 (0.2983) loss 3.1469 (3.2438) grad_norm 2.0270 (2.1955) [2021-04-16 12:36:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][720/1251] eta 0:02:38 lr 0.000191 time 0.2830 (0.2980) loss 2.9365 (3.2407) grad_norm 2.2952 (2.1945) [2021-04-16 12:36:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][730/1251] eta 0:02:35 lr 0.000191 time 0.2567 (0.2979) loss 2.6696 (3.2433) grad_norm 2.2900 (2.1932) [2021-04-16 12:36:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][740/1251] eta 0:02:32 lr 0.000191 time 0.2546 (0.2975) loss 4.0303 (3.2485) grad_norm 2.0826 (2.1921) [2021-04-16 12:36:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][750/1251] eta 0:02:28 lr 0.000191 time 0.2708 (0.2972) loss 4.1830 (3.2518) grad_norm 2.0720 (2.1918) [2021-04-16 12:36:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][760/1251] eta 0:02:25 lr 0.000191 time 0.2855 (0.2970) loss 3.3563 (3.2532) grad_norm 1.9758 (2.1899) [2021-04-16 12:36:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][770/1251] eta 0:02:22 lr 0.000191 time 0.2812 (0.2968) loss 2.8303 (3.2496) grad_norm 2.0240 (2.1897) [2021-04-16 12:36:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][780/1251] eta 0:02:19 lr 0.000191 time 0.2752 (0.2968) loss 2.2219 (3.2472) grad_norm 2.4029 (2.1890) [2021-04-16 12:36:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][790/1251] eta 0:02:16 lr 0.000191 time 0.2633 (0.2967) loss 3.2441 (3.2496) grad_norm 2.0721 (2.1896) [2021-04-16 12:36:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][800/1251] eta 0:02:13 lr 0.000191 time 0.2641 (0.2965) loss 3.7778 (3.2506) grad_norm 2.0703 (2.1889) [2021-04-16 12:36:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][810/1251] eta 0:02:10 lr 0.000191 time 0.2663 (0.2962) loss 3.9902 (3.2544) grad_norm 2.5015 (2.1882) [2021-04-16 12:36:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][820/1251] eta 0:02:07 lr 0.000191 time 0.2418 (0.2960) loss 2.4615 (3.2520) grad_norm 2.1516 (2.1884) [2021-04-16 12:36:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][830/1251] eta 0:02:04 lr 0.000191 time 0.2657 (0.2958) loss 3.6392 (3.2547) grad_norm 2.8403 (2.1895) [2021-04-16 12:36:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][840/1251] eta 0:02:01 lr 0.000191 time 0.3049 (0.2957) loss 3.6611 (3.2577) grad_norm 2.2127 (2.1906) [2021-04-16 12:36:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][850/1251] eta 0:01:58 lr 0.000191 time 0.2623 (0.2954) loss 2.3657 (3.2547) grad_norm 2.7655 (2.1915) [2021-04-16 12:36:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][860/1251] eta 0:01:55 lr 0.000191 time 0.2995 (0.2952) loss 3.4892 (3.2523) grad_norm 1.9156 (2.1919) [2021-04-16 12:37:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][870/1251] eta 0:01:52 lr 0.000191 time 0.2764 (0.2950) loss 3.0633 (3.2526) grad_norm 2.6674 (2.1932) [2021-04-16 12:37:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][880/1251] eta 0:01:49 lr 0.000191 time 0.3165 (0.2948) loss 3.5127 (3.2513) grad_norm 1.9678 (2.1928) [2021-04-16 12:37:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][890/1251] eta 0:01:46 lr 0.000191 time 0.2712 (0.2946) loss 3.1012 (3.2525) grad_norm 2.0921 (2.1939) [2021-04-16 12:37:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][900/1251] eta 0:01:43 lr 0.000191 time 0.2545 (0.2944) loss 3.6543 (3.2534) grad_norm 1.9940 (2.1940) [2021-04-16 12:37:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][910/1251] eta 0:01:40 lr 0.000191 time 0.2802 (0.2942) loss 3.5872 (3.2552) grad_norm 2.1100 (2.1941) [2021-04-16 12:37:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][920/1251] eta 0:01:37 lr 0.000191 time 0.2785 (0.2940) loss 2.1119 (3.2539) grad_norm 2.1862 (2.1936) [2021-04-16 12:37:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][930/1251] eta 0:01:34 lr 0.000191 time 0.2797 (0.2938) loss 3.5546 (3.2554) grad_norm 2.1328 (2.1933) [2021-04-16 12:37:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][940/1251] eta 0:01:31 lr 0.000190 time 0.2816 (0.2940) loss 3.6361 (3.2576) grad_norm 1.9077 (2.1935) [2021-04-16 12:37:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][950/1251] eta 0:01:28 lr 0.000190 time 0.2819 (0.2938) loss 3.1604 (3.2547) grad_norm 2.4000 (2.1927) [2021-04-16 12:37:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][960/1251] eta 0:01:25 lr 0.000190 time 0.3088 (0.2937) loss 2.5236 (3.2540) grad_norm 2.1249 (2.1924) [2021-04-16 12:37:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][970/1251] eta 0:01:22 lr 0.000190 time 0.2746 (0.2935) loss 3.4760 (3.2524) grad_norm 2.2730 (2.1919) [2021-04-16 12:37:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][980/1251] eta 0:01:19 lr 0.000190 time 0.2729 (0.2934) loss 3.1478 (3.2544) grad_norm 2.1038 (2.1914) [2021-04-16 12:37:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][990/1251] eta 0:01:16 lr 0.000190 time 0.2744 (0.2933) loss 3.7731 (3.2570) grad_norm 2.1445 (2.1920) [2021-04-16 12:37:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][1000/1251] eta 0:01:13 lr 0.000190 time 0.2641 (0.2931) loss 2.4100 (3.2560) grad_norm 2.1897 (2.1910) [2021-04-16 12:37:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][1010/1251] eta 0:01:10 lr 0.000190 time 0.3017 (0.2931) loss 3.5142 (3.2560) grad_norm 2.5185 (2.1919) [2021-04-16 12:37:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][1020/1251] eta 0:01:07 lr 0.000190 time 0.2690 (0.2929) loss 3.7750 (3.2559) grad_norm 2.5019 (2.1950) [2021-04-16 12:37:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][1030/1251] eta 0:01:04 lr 0.000190 time 0.2933 (0.2928) loss 3.6362 (3.2550) grad_norm 2.1606 (2.1972) [2021-04-16 12:37:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][1040/1251] eta 0:01:01 lr 0.000190 time 0.3231 (0.2927) loss 3.0613 (3.2541) grad_norm 1.9922 (2.1969) [2021-04-16 12:37:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][1050/1251] eta 0:00:58 lr 0.000190 time 0.2761 (0.2926) loss 3.8403 (3.2532) grad_norm 2.8759 (2.1961) [2021-04-16 12:37:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][1060/1251] eta 0:00:55 lr 0.000190 time 0.2841 (0.2925) loss 3.3237 (3.2500) grad_norm 2.0553 (2.1956) [2021-04-16 12:37:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][1070/1251] eta 0:00:52 lr 0.000190 time 0.2937 (0.2924) loss 3.2349 (3.2512) grad_norm 2.0373 (2.1948) [2021-04-16 12:38:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][1080/1251] eta 0:00:49 lr 0.000190 time 0.2768 (0.2923) loss 3.5538 (3.2516) grad_norm 2.1174 (2.1958) [2021-04-16 12:38:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][1090/1251] eta 0:00:47 lr 0.000190 time 0.2816 (0.2922) loss 3.4884 (3.2526) grad_norm 2.2368 (2.1959) [2021-04-16 12:38:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][1100/1251] eta 0:00:44 lr 0.000190 time 0.2621 (0.2920) loss 3.4309 (3.2507) grad_norm 2.2716 (2.1956) [2021-04-16 12:38:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][1110/1251] eta 0:00:41 lr 0.000190 time 0.2810 (0.2919) loss 3.6625 (3.2519) grad_norm 2.0672 (2.1959) [2021-04-16 12:38:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][1120/1251] eta 0:00:38 lr 0.000190 time 0.2641 (0.2918) loss 3.5129 (3.2544) grad_norm 2.6055 (2.1966) [2021-04-16 12:38:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][1130/1251] eta 0:00:35 lr 0.000190 time 0.2437 (0.2917) loss 3.3860 (3.2535) grad_norm 2.6000 (2.1963) [2021-04-16 12:38:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][1140/1251] eta 0:00:32 lr 0.000190 time 0.2862 (0.2917) loss 3.5666 (3.2545) grad_norm 2.5070 (2.1973) [2021-04-16 12:38:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][1150/1251] eta 0:00:29 lr 0.000190 time 0.2765 (0.2918) loss 2.5656 (3.2525) grad_norm 2.1569 (2.1972) [2021-04-16 12:38:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][1160/1251] eta 0:00:26 lr 0.000190 time 0.3012 (0.2919) loss 4.2908 (3.2552) grad_norm 1.9391 (2.1969) [2021-04-16 12:38:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][1170/1251] eta 0:00:23 lr 0.000190 time 0.2465 (0.2917) loss 3.3940 (3.2556) grad_norm 2.0287 (2.1969) [2021-04-16 12:38:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][1180/1251] eta 0:00:20 lr 0.000190 time 0.2799 (0.2917) loss 2.4270 (3.2572) grad_norm 1.8873 (2.1959) [2021-04-16 12:38:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][1190/1251] eta 0:00:17 lr 0.000190 time 0.2532 (0.2915) loss 2.9621 (3.2586) grad_norm 2.0901 (2.1959) [2021-04-16 12:38:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][1200/1251] eta 0:00:14 lr 0.000190 time 0.2795 (0.2914) loss 4.0826 (3.2588) grad_norm 2.0961 (2.1963) [2021-04-16 12:38:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][1210/1251] eta 0:00:11 lr 0.000190 time 0.2698 (0.2913) loss 3.1270 (3.2601) grad_norm 2.1718 (2.1964) [2021-04-16 12:38:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][1220/1251] eta 0:00:09 lr 0.000190 time 0.2780 (0.2912) loss 2.6169 (3.2597) grad_norm 2.4906 (2.1957) [2021-04-16 12:38:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][1230/1251] eta 0:00:06 lr 0.000190 time 0.2600 (0.2910) loss 3.2792 (3.2601) grad_norm 2.2151 (2.1964) [2021-04-16 12:38:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][1240/1251] eta 0:00:03 lr 0.000190 time 0.2482 (0.2908) loss 3.8770 (3.2601) grad_norm 2.0725 (2.1972) [2021-04-16 12:38:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [215/300][1250/1251] eta 0:00:00 lr 0.000189 time 0.2485 (0.2905) loss 3.8464 (3.2628) grad_norm 2.4784 (2.1971) [2021-04-16 12:38:51 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 215 training takes 0:06:06 [2021-04-16 12:38:51 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_215.pth saving...... [2021-04-16 12:38:58 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_215.pth saved !!! [2021-04-16 12:39:01 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 3.455 (3.455) Loss 0.9109 (0.9109) Acc@1 78.809 (78.809) Acc@5 94.434 (94.434) [2021-04-16 12:39:03 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.104 (0.429) Loss 0.8739 (0.8763) Acc@1 80.273 (79.199) Acc@5 94.336 (94.762) [2021-04-16 12:39:05 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.111 (0.341) Loss 0.9970 (0.8816) Acc@1 76.270 (78.985) Acc@5 93.066 (94.727) [2021-04-16 12:39:08 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.091 (0.325) Loss 0.9650 (0.8838) Acc@1 76.270 (79.045) Acc@5 93.164 (94.679) [2021-04-16 12:39:09 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.279) Loss 0.8240 (0.8834) Acc@1 80.664 (79.056) Acc@5 95.312 (94.684) [2021-04-16 12:39:16 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 78.978 Acc@5 94.642 [2021-04-16 12:39:16 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 79.0% [2021-04-16 12:39:16 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 79.16% [2021-04-16 12:39:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][0/1251] eta 4:40:26 lr 0.000189 time 13.4506 (13.4506) loss 2.3615 (2.3615) grad_norm 2.2552 (2.2552) [2021-04-16 12:39:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][10/1251] eta 0:30:36 lr 0.000189 time 0.4206 (1.4801) loss 3.6138 (3.3496) grad_norm 2.2736 (2.2129) [2021-04-16 12:39:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][20/1251] eta 0:18:34 lr 0.000189 time 0.2776 (0.9056) loss 3.9384 (3.4267) grad_norm 2.2704 (2.1864) [2021-04-16 12:39:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][30/1251] eta 0:14:20 lr 0.000189 time 0.2790 (0.7048) loss 3.8193 (3.4252) grad_norm 2.2671 (2.1886) [2021-04-16 12:39:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][40/1251] eta 0:12:07 lr 0.000189 time 0.2925 (0.6004) loss 3.2638 (3.3261) grad_norm 2.4906 (2.1849) [2021-04-16 12:39:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][50/1251] eta 0:10:46 lr 0.000189 time 0.2813 (0.5382) loss 3.4453 (3.2719) grad_norm 1.8945 (2.1987) [2021-04-16 12:39:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][60/1251] eta 0:09:49 lr 0.000189 time 0.2786 (0.4952) loss 3.4639 (3.2522) grad_norm 2.0646 (2.1968) [2021-04-16 12:39:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][70/1251] eta 0:09:09 lr 0.000189 time 0.3151 (0.4656) loss 3.1059 (3.2413) grad_norm 1.8490 (2.1923) [2021-04-16 12:39:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][80/1251] eta 0:08:38 lr 0.000189 time 0.2774 (0.4427) loss 4.1603 (3.2564) grad_norm 2.0870 (2.1816) [2021-04-16 12:39:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][90/1251] eta 0:08:13 lr 0.000189 time 0.2904 (0.4252) loss 3.5528 (3.2401) grad_norm 2.0573 (2.1726) [2021-04-16 12:39:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][100/1251] eta 0:07:53 lr 0.000189 time 0.2581 (0.4111) loss 3.2758 (3.2726) grad_norm 1.8782 (2.1620) [2021-04-16 12:40:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][110/1251] eta 0:07:35 lr 0.000189 time 0.2661 (0.3994) loss 3.5653 (3.2839) grad_norm 2.6291 (2.1715) [2021-04-16 12:40:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][120/1251] eta 0:07:20 lr 0.000189 time 0.2936 (0.3898) loss 3.2244 (3.2912) grad_norm 2.0583 (2.1803) [2021-04-16 12:40:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][130/1251] eta 0:07:07 lr 0.000189 time 0.2818 (0.3813) loss 3.8123 (3.2895) grad_norm 2.2048 (2.1750) [2021-04-16 12:40:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][140/1251] eta 0:06:56 lr 0.000189 time 0.2708 (0.3745) loss 3.7605 (3.2951) grad_norm 1.9916 (2.1817) [2021-04-16 12:40:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][150/1251] eta 0:06:47 lr 0.000189 time 0.2892 (0.3705) loss 3.5420 (3.3104) grad_norm 1.8941 (2.1827) [2021-04-16 12:40:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][160/1251] eta 0:06:38 lr 0.000189 time 0.2755 (0.3648) loss 3.6710 (3.3342) grad_norm 2.2125 (2.1774) [2021-04-16 12:40:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][170/1251] eta 0:06:28 lr 0.000189 time 0.2855 (0.3597) loss 2.7411 (3.3292) grad_norm 2.5685 (2.1837) [2021-04-16 12:40:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][180/1251] eta 0:06:20 lr 0.000189 time 0.2742 (0.3552) loss 3.2129 (3.3254) grad_norm 2.1836 (2.1822) [2021-04-16 12:40:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][190/1251] eta 0:06:12 lr 0.000189 time 0.2766 (0.3512) loss 3.1182 (3.3214) grad_norm 2.3224 (2.1829) [2021-04-16 12:40:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][200/1251] eta 0:06:05 lr 0.000189 time 0.2890 (0.3479) loss 3.7959 (3.3264) grad_norm 2.0947 (2.1795) [2021-04-16 12:40:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][210/1251] eta 0:05:59 lr 0.000189 time 0.3046 (0.3449) loss 3.5004 (3.3304) grad_norm 1.8856 (2.1786) [2021-04-16 12:40:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][220/1251] eta 0:05:53 lr 0.000189 time 0.2501 (0.3425) loss 3.9789 (3.3349) grad_norm 2.1185 (2.1745) [2021-04-16 12:40:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][230/1251] eta 0:05:47 lr 0.000189 time 0.3051 (0.3401) loss 3.3061 (3.3353) grad_norm 1.9528 (2.1750) [2021-04-16 12:40:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][240/1251] eta 0:05:41 lr 0.000189 time 0.2959 (0.3378) loss 2.7743 (3.3295) grad_norm 2.3221 (2.1714) [2021-04-16 12:40:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][250/1251] eta 0:05:35 lr 0.000189 time 0.2774 (0.3354) loss 3.5454 (3.3232) grad_norm 2.3510 (2.1697) [2021-04-16 12:40:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][260/1251] eta 0:05:30 lr 0.000189 time 0.2680 (0.3335) loss 1.9871 (3.3001) grad_norm 2.0647 (2.1750) [2021-04-16 12:40:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][270/1251] eta 0:05:25 lr 0.000189 time 0.2740 (0.3313) loss 2.8431 (3.2944) grad_norm 2.4401 (2.1794) [2021-04-16 12:40:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][280/1251] eta 0:05:19 lr 0.000189 time 0.2673 (0.3295) loss 3.3985 (3.2874) grad_norm 1.9711 (2.1803) [2021-04-16 12:40:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][290/1251] eta 0:05:14 lr 0.000189 time 0.2846 (0.3277) loss 3.8452 (3.2807) grad_norm 2.3394 (2.1765) [2021-04-16 12:40:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][300/1251] eta 0:05:10 lr 0.000189 time 0.2790 (0.3260) loss 3.3179 (3.2802) grad_norm 2.0811 (2.1774) [2021-04-16 12:40:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][310/1251] eta 0:05:05 lr 0.000188 time 0.3134 (0.3246) loss 3.3645 (3.2862) grad_norm 1.9171 (2.1775) [2021-04-16 12:41:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][320/1251] eta 0:05:00 lr 0.000188 time 0.2851 (0.3230) loss 2.2274 (3.2797) grad_norm 2.6665 (2.1765) [2021-04-16 12:41:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][330/1251] eta 0:04:56 lr 0.000188 time 0.2784 (0.3216) loss 3.4876 (3.2843) grad_norm 2.6224 (2.1787) [2021-04-16 12:41:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][340/1251] eta 0:04:52 lr 0.000188 time 0.2747 (0.3206) loss 3.4389 (3.2858) grad_norm 1.8983 (2.1778) [2021-04-16 12:41:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][350/1251] eta 0:04:48 lr 0.000188 time 0.4672 (0.3199) loss 3.2579 (3.2839) grad_norm 2.1858 (2.1784) [2021-04-16 12:41:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][360/1251] eta 0:04:44 lr 0.000188 time 0.3014 (0.3195) loss 2.6254 (3.2866) grad_norm 1.7495 (2.1782) [2021-04-16 12:41:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][370/1251] eta 0:04:40 lr 0.000188 time 0.2748 (0.3184) loss 3.3660 (3.2880) grad_norm 2.1347 (2.1781) [2021-04-16 12:41:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][380/1251] eta 0:04:36 lr 0.000188 time 0.2813 (0.3176) loss 3.7098 (3.2865) grad_norm 2.3293 (2.1792) [2021-04-16 12:41:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][390/1251] eta 0:04:32 lr 0.000188 time 0.3074 (0.3166) loss 2.8709 (3.2780) grad_norm 2.0519 (2.1782) [2021-04-16 12:41:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][400/1251] eta 0:04:28 lr 0.000188 time 0.2901 (0.3156) loss 3.5672 (3.2791) grad_norm 2.2455 (2.1798) [2021-04-16 12:41:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][410/1251] eta 0:04:24 lr 0.000188 time 0.2611 (0.3148) loss 3.6875 (3.2778) grad_norm 2.2236 (2.1835) [2021-04-16 12:41:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][420/1251] eta 0:04:20 lr 0.000188 time 0.2734 (0.3139) loss 2.2984 (3.2792) grad_norm 1.7968 (2.1854) [2021-04-16 12:41:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][430/1251] eta 0:04:17 lr 0.000188 time 0.2698 (0.3132) loss 2.7577 (3.2827) grad_norm 2.2933 (2.1854) [2021-04-16 12:41:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][440/1251] eta 0:04:13 lr 0.000188 time 0.2932 (0.3124) loss 2.9974 (3.2852) grad_norm 2.0314 (2.1869) [2021-04-16 12:41:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][450/1251] eta 0:04:09 lr 0.000188 time 0.2808 (0.3116) loss 4.1939 (3.2864) grad_norm 2.1501 (2.1852) [2021-04-16 12:41:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][460/1251] eta 0:04:05 lr 0.000188 time 0.2827 (0.3110) loss 2.9651 (3.2862) grad_norm 2.3980 (2.1856) [2021-04-16 12:41:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][470/1251] eta 0:04:02 lr 0.000188 time 0.2945 (0.3105) loss 3.7281 (3.2885) grad_norm 2.4647 (2.1852) [2021-04-16 12:41:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][480/1251] eta 0:03:58 lr 0.000188 time 0.2717 (0.3099) loss 3.6807 (3.2871) grad_norm 2.1008 (2.1834) [2021-04-16 12:41:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][490/1251] eta 0:03:55 lr 0.000188 time 0.2825 (0.3093) loss 3.9836 (3.2878) grad_norm 2.4407 (2.1856) [2021-04-16 12:41:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][500/1251] eta 0:03:51 lr 0.000188 time 0.2752 (0.3087) loss 3.8507 (3.2896) grad_norm 2.1123 (2.1846) [2021-04-16 12:41:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][510/1251] eta 0:03:48 lr 0.000188 time 0.2508 (0.3080) loss 3.7791 (3.2899) grad_norm 2.3954 (2.1872) [2021-04-16 12:41:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][520/1251] eta 0:03:44 lr 0.000188 time 0.2842 (0.3075) loss 2.4048 (3.2892) grad_norm 2.4503 (2.1901) [2021-04-16 12:41:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][530/1251] eta 0:03:41 lr 0.000188 time 0.2571 (0.3068) loss 2.8244 (3.2882) grad_norm 1.7446 (2.1903) [2021-04-16 12:42:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][540/1251] eta 0:03:37 lr 0.000188 time 0.2814 (0.3065) loss 3.3972 (3.2867) grad_norm 2.1344 (2.1932) [2021-04-16 12:42:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][550/1251] eta 0:03:34 lr 0.000188 time 0.4081 (0.3062) loss 2.9926 (3.2879) grad_norm 2.6766 (2.1963) [2021-04-16 12:42:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][560/1251] eta 0:03:31 lr 0.000188 time 0.2644 (0.3057) loss 3.1865 (3.2866) grad_norm 2.0315 (2.1970) [2021-04-16 12:42:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][570/1251] eta 0:03:27 lr 0.000188 time 0.2755 (0.3052) loss 3.3129 (3.2852) grad_norm 1.9459 (2.1966) [2021-04-16 12:42:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][580/1251] eta 0:03:24 lr 0.000188 time 0.2704 (0.3052) loss 3.0788 (3.2804) grad_norm 2.1057 (2.1986) [2021-04-16 12:42:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][590/1251] eta 0:03:21 lr 0.000188 time 0.2713 (0.3049) loss 3.1319 (3.2748) grad_norm 2.2950 (2.1982) [2021-04-16 12:42:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][600/1251] eta 0:03:18 lr 0.000188 time 0.2983 (0.3045) loss 3.5976 (3.2776) grad_norm 2.7808 (2.1998) [2021-04-16 12:42:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][610/1251] eta 0:03:14 lr 0.000188 time 0.2856 (0.3041) loss 3.0927 (3.2784) grad_norm 2.0875 (2.2004) [2021-04-16 12:42:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][620/1251] eta 0:03:11 lr 0.000187 time 0.2807 (0.3036) loss 3.6086 (3.2750) grad_norm 2.4597 (2.2008) [2021-04-16 12:42:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][630/1251] eta 0:03:08 lr 0.000187 time 0.2471 (0.3032) loss 2.5161 (3.2745) grad_norm 2.2596 (2.2011) [2021-04-16 12:42:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][640/1251] eta 0:03:04 lr 0.000187 time 0.2855 (0.3028) loss 3.5630 (3.2742) grad_norm 2.0848 (2.2009) [2021-04-16 12:42:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][650/1251] eta 0:03:01 lr 0.000187 time 0.2780 (0.3024) loss 3.5968 (3.2751) grad_norm 2.4971 (2.2007) [2021-04-16 12:42:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][660/1251] eta 0:02:58 lr 0.000187 time 0.2878 (0.3021) loss 2.9453 (3.2781) grad_norm 2.0201 (2.2005) [2021-04-16 12:42:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][670/1251] eta 0:02:55 lr 0.000187 time 0.2549 (0.3017) loss 3.1814 (3.2753) grad_norm 2.0882 (2.2023) [2021-04-16 12:42:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][680/1251] eta 0:02:52 lr 0.000187 time 0.2924 (0.3014) loss 3.3947 (3.2777) grad_norm 2.1210 (2.2025) [2021-04-16 12:42:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][690/1251] eta 0:02:48 lr 0.000187 time 0.2771 (0.3011) loss 2.2878 (3.2723) grad_norm 2.1708 (2.2012) [2021-04-16 12:42:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][700/1251] eta 0:02:45 lr 0.000187 time 0.2768 (0.3007) loss 3.6003 (3.2719) grad_norm 1.9035 (2.1993) [2021-04-16 12:42:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][710/1251] eta 0:02:42 lr 0.000187 time 0.2682 (0.3003) loss 3.0708 (3.2747) grad_norm 3.9037 (2.2003) [2021-04-16 12:42:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][720/1251] eta 0:02:39 lr 0.000187 time 0.2632 (0.3002) loss 2.2245 (3.2743) grad_norm 1.9238 (2.2017) [2021-04-16 12:42:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][730/1251] eta 0:02:36 lr 0.000187 time 0.2884 (0.3001) loss 3.7221 (3.2723) grad_norm 1.9245 (2.2007) [2021-04-16 12:42:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][740/1251] eta 0:02:33 lr 0.000187 time 0.2554 (0.2998) loss 2.2679 (3.2690) grad_norm 2.5318 (2.2030) [2021-04-16 12:43:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][750/1251] eta 0:02:30 lr 0.000187 time 0.2413 (0.2996) loss 3.5343 (3.2717) grad_norm 2.4818 (2.2019) [2021-04-16 12:43:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][760/1251] eta 0:02:26 lr 0.000187 time 0.2842 (0.2993) loss 3.4736 (3.2744) grad_norm 2.0904 (2.2012) [2021-04-16 12:43:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][770/1251] eta 0:02:23 lr 0.000187 time 0.2820 (0.2990) loss 2.6515 (3.2769) grad_norm 2.2432 (2.2010) [2021-04-16 12:43:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][780/1251] eta 0:02:20 lr 0.000187 time 0.2894 (0.2988) loss 3.7129 (3.2780) grad_norm 2.2094 (2.2009) [2021-04-16 12:43:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][790/1251] eta 0:02:17 lr 0.000187 time 0.2798 (0.2987) loss 3.0787 (3.2772) grad_norm 2.2879 (2.2024) [2021-04-16 12:43:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][800/1251] eta 0:02:14 lr 0.000187 time 0.2927 (0.2985) loss 3.8697 (3.2756) grad_norm 2.1327 (2.2036) [2021-04-16 12:43:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][810/1251] eta 0:02:11 lr 0.000187 time 0.2764 (0.2983) loss 3.8684 (3.2739) grad_norm 2.3049 (2.2035) [2021-04-16 12:43:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][820/1251] eta 0:02:08 lr 0.000187 time 0.2775 (0.2982) loss 3.2200 (3.2729) grad_norm 1.8616 (2.2021) [2021-04-16 12:43:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][830/1251] eta 0:02:05 lr 0.000187 time 0.2765 (0.2979) loss 3.4080 (3.2743) grad_norm 2.4482 (2.2023) [2021-04-16 12:43:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][840/1251] eta 0:02:02 lr 0.000187 time 0.2775 (0.2976) loss 3.3345 (3.2747) grad_norm 2.0127 (2.2021) [2021-04-16 12:43:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][850/1251] eta 0:01:59 lr 0.000187 time 0.2652 (0.2974) loss 2.8362 (3.2754) grad_norm 1.7853 (2.2011) [2021-04-16 12:43:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][860/1251] eta 0:01:56 lr 0.000187 time 0.2884 (0.2972) loss 3.6537 (3.2785) grad_norm 2.1509 (2.2010) [2021-04-16 12:43:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][870/1251] eta 0:01:53 lr 0.000187 time 0.2794 (0.2970) loss 2.4740 (3.2785) grad_norm 1.9935 (2.1999) [2021-04-16 12:43:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][880/1251] eta 0:01:50 lr 0.000187 time 0.2670 (0.2968) loss 3.4446 (3.2753) grad_norm 2.0648 (2.1986) [2021-04-16 12:43:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][890/1251] eta 0:01:47 lr 0.000187 time 0.2683 (0.2965) loss 3.6718 (3.2772) grad_norm 2.1186 (2.1978) [2021-04-16 12:43:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][900/1251] eta 0:01:44 lr 0.000187 time 0.2692 (0.2964) loss 3.2713 (3.2746) grad_norm 1.9526 (2.1968) [2021-04-16 12:43:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][910/1251] eta 0:01:40 lr 0.000187 time 0.2666 (0.2961) loss 3.5161 (3.2751) grad_norm 2.4344 (2.1981) [2021-04-16 12:43:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][920/1251] eta 0:01:37 lr 0.000187 time 0.2772 (0.2961) loss 3.9608 (3.2779) grad_norm 2.2950 (2.1994) [2021-04-16 12:43:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][930/1251] eta 0:01:35 lr 0.000187 time 0.2948 (0.2962) loss 3.6605 (3.2798) grad_norm 1.8043 (2.1984) [2021-04-16 12:43:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][940/1251] eta 0:01:32 lr 0.000186 time 0.2962 (0.2961) loss 3.9532 (3.2815) grad_norm 2.4116 (2.1986) [2021-04-16 12:43:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][950/1251] eta 0:01:29 lr 0.000186 time 0.2768 (0.2960) loss 3.6673 (3.2832) grad_norm 2.0449 (2.1976) [2021-04-16 12:44:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][960/1251] eta 0:01:26 lr 0.000186 time 0.2811 (0.2958) loss 3.3854 (3.2822) grad_norm 2.2429 (2.1962) [2021-04-16 12:44:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][970/1251] eta 0:01:23 lr 0.000186 time 0.2913 (0.2956) loss 3.7296 (3.2803) grad_norm 2.3426 (2.1979) [2021-04-16 12:44:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][980/1251] eta 0:01:20 lr 0.000186 time 0.2558 (0.2954) loss 3.6428 (3.2816) grad_norm 2.1607 (2.2000) [2021-04-16 12:44:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][990/1251] eta 0:01:17 lr 0.000186 time 0.3038 (0.2953) loss 3.2274 (3.2813) grad_norm 2.1865 (2.2001) [2021-04-16 12:44:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][1000/1251] eta 0:01:14 lr 0.000186 time 0.2791 (0.2953) loss 3.1594 (3.2804) grad_norm 1.9981 (2.1999) [2021-04-16 12:44:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][1010/1251] eta 0:01:11 lr 0.000186 time 0.2783 (0.2950) loss 2.1922 (3.2796) grad_norm 2.1623 (2.1990) [2021-04-16 12:44:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][1020/1251] eta 0:01:08 lr 0.000186 time 0.3108 (0.2949) loss 3.4685 (3.2772) grad_norm 2.2799 (2.1997) [2021-04-16 12:44:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][1030/1251] eta 0:01:05 lr 0.000186 time 0.3009 (0.2947) loss 2.8452 (3.2783) grad_norm 2.1346 (2.1987) [2021-04-16 12:44:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][1040/1251] eta 0:01:02 lr 0.000186 time 0.2720 (0.2946) loss 3.2635 (3.2772) grad_norm 2.3516 (2.1985) [2021-04-16 12:44:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][1050/1251] eta 0:00:59 lr 0.000186 time 0.2826 (0.2944) loss 3.2907 (3.2766) grad_norm 1.9612 (2.1978) [2021-04-16 12:44:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][1060/1251] eta 0:00:56 lr 0.000186 time 0.2652 (0.2943) loss 2.1118 (3.2754) grad_norm 2.2688 (2.1973) [2021-04-16 12:44:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][1070/1251] eta 0:00:53 lr 0.000186 time 0.2841 (0.2941) loss 3.6496 (3.2780) grad_norm 2.7671 (2.1980) [2021-04-16 12:44:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][1080/1251] eta 0:00:50 lr 0.000186 time 0.2755 (0.2939) loss 3.5417 (3.2767) grad_norm 2.3516 (2.1985) [2021-04-16 12:44:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][1090/1251] eta 0:00:47 lr 0.000186 time 0.2688 (0.2938) loss 2.0978 (3.2752) grad_norm 2.4862 (2.1985) [2021-04-16 12:44:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][1100/1251] eta 0:00:44 lr 0.000186 time 0.2470 (0.2936) loss 4.0797 (3.2753) grad_norm nan (nan) [2021-04-16 12:44:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][1110/1251] eta 0:00:41 lr 0.000186 time 0.2717 (0.2935) loss 2.3019 (3.2748) grad_norm 2.3725 (nan) [2021-04-16 12:44:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][1120/1251] eta 0:00:38 lr 0.000186 time 0.2926 (0.2935) loss 3.8307 (3.2728) grad_norm 2.4089 (nan) [2021-04-16 12:44:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][1130/1251] eta 0:00:35 lr 0.000186 time 0.2670 (0.2935) loss 3.4005 (3.2727) grad_norm 2.3879 (nan) [2021-04-16 12:44:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][1140/1251] eta 0:00:32 lr 0.000186 time 0.2660 (0.2934) loss 3.5433 (3.2755) grad_norm 2.1991 (nan) [2021-04-16 12:44:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][1150/1251] eta 0:00:29 lr 0.000186 time 0.2862 (0.2932) loss 2.6705 (3.2741) grad_norm 2.2759 (nan) [2021-04-16 12:44:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][1160/1251] eta 0:00:26 lr 0.000186 time 0.2982 (0.2933) loss 1.9773 (3.2710) grad_norm 2.1038 (nan) [2021-04-16 12:45:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][1170/1251] eta 0:00:23 lr 0.000186 time 0.2746 (0.2933) loss 3.0248 (3.2716) grad_norm 2.6703 (nan) [2021-04-16 12:45:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][1180/1251] eta 0:00:20 lr 0.000186 time 0.2773 (0.2932) loss 3.8719 (3.2731) grad_norm 1.9226 (nan) [2021-04-16 12:45:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][1190/1251] eta 0:00:17 lr 0.000186 time 0.2810 (0.2930) loss 2.5796 (3.2712) grad_norm 1.9075 (nan) [2021-04-16 12:45:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][1200/1251] eta 0:00:14 lr 0.000186 time 0.2784 (0.2930) loss 3.3582 (3.2727) grad_norm 2.1392 (nan) [2021-04-16 12:45:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][1210/1251] eta 0:00:12 lr 0.000186 time 0.2976 (0.2929) loss 2.9980 (3.2737) grad_norm 2.3329 (nan) [2021-04-16 12:45:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][1220/1251] eta 0:00:09 lr 0.000186 time 0.2889 (0.2927) loss 3.0750 (3.2739) grad_norm 2.4051 (nan) [2021-04-16 12:45:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][1230/1251] eta 0:00:06 lr 0.000186 time 0.2716 (0.2926) loss 3.5732 (3.2741) grad_norm 2.3588 (nan) [2021-04-16 12:45:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][1240/1251] eta 0:00:03 lr 0.000186 time 0.2490 (0.2924) loss 3.5957 (3.2746) grad_norm 1.9427 (nan) [2021-04-16 12:45:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [216/300][1250/1251] eta 0:00:00 lr 0.000186 time 0.2491 (0.2921) loss 3.1924 (3.2748) grad_norm 2.2712 (nan) [2021-04-16 12:45:27 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 216 training takes 0:06:10 [2021-04-16 12:45:27 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_216.pth saving...... [2021-04-16 12:45:37 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_216.pth saved !!! [2021-04-16 12:45:38 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.130 (1.130) Loss 0.8645 (0.8645) Acc@1 79.004 (79.004) Acc@5 95.117 (95.117) [2021-04-16 12:45:40 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.614 (0.262) Loss 0.9996 (0.8901) Acc@1 75.781 (79.155) Acc@5 93.652 (94.416) [2021-04-16 12:45:42 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.568 (0.226) Loss 0.8853 (0.8783) Acc@1 79.004 (79.381) Acc@5 94.629 (94.606) [2021-04-16 12:45:44 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.441 (0.240) Loss 0.8333 (0.8828) Acc@1 80.371 (79.272) Acc@5 95.605 (94.660) [2021-04-16 12:45:46 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.214) Loss 0.9424 (0.8864) Acc@1 78.906 (79.216) Acc@5 93.750 (94.610) [2021-04-16 12:45:54 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 79.168 Acc@5 94.648 [2021-04-16 12:45:54 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 79.2% [2021-04-16 12:45:54 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 79.17% [2021-04-16 12:46:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][0/1251] eta 2:30:55 lr 0.000185 time 7.2385 (7.2385) loss 3.7478 (3.7478) grad_norm 2.1292 (2.1292) [2021-04-16 12:46:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][10/1251] eta 0:19:04 lr 0.000185 time 0.4091 (0.9223) loss 2.9173 (2.9466) grad_norm 1.9354 (2.2080) [2021-04-16 12:46:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][20/1251] eta 0:12:39 lr 0.000185 time 0.2800 (0.6170) loss 3.4523 (3.1678) grad_norm 2.2369 (2.1910) [2021-04-16 12:46:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][30/1251] eta 0:10:24 lr 0.000185 time 0.2679 (0.5111) loss 3.7097 (3.2146) grad_norm 2.2891 (2.1980) [2021-04-16 12:46:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][40/1251] eta 0:09:09 lr 0.000185 time 0.2769 (0.4540) loss 3.7245 (3.2587) grad_norm 2.3936 (2.1893) [2021-04-16 12:46:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][50/1251] eta 0:08:24 lr 0.000185 time 0.2786 (0.4201) loss 3.4439 (3.2869) grad_norm 2.0380 (2.1799) [2021-04-16 12:46:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][60/1251] eta 0:07:53 lr 0.000185 time 0.2743 (0.3979) loss 3.4007 (3.3094) grad_norm 1.7637 (2.2109) [2021-04-16 12:46:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][70/1251] eta 0:07:30 lr 0.000185 time 0.2810 (0.3810) loss 3.2216 (3.3430) grad_norm 1.9990 (2.2032) [2021-04-16 12:46:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][80/1251] eta 0:07:11 lr 0.000185 time 0.2822 (0.3688) loss 3.3889 (3.3464) grad_norm 2.0164 (2.2083) [2021-04-16 12:46:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][90/1251] eta 0:06:57 lr 0.000185 time 0.2821 (0.3595) loss 2.3054 (3.3361) grad_norm 2.1298 (2.2214) [2021-04-16 12:46:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][100/1251] eta 0:06:45 lr 0.000185 time 0.3189 (0.3526) loss 3.8507 (3.3383) grad_norm 2.1339 (2.2211) [2021-04-16 12:46:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][110/1251] eta 0:06:34 lr 0.000185 time 0.2621 (0.3458) loss 3.3336 (3.3337) grad_norm 2.7029 (2.2252) [2021-04-16 12:46:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][120/1251] eta 0:06:26 lr 0.000185 time 0.3080 (0.3415) loss 3.5909 (3.3132) grad_norm 1.8297 (2.2166) [2021-04-16 12:46:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][130/1251] eta 0:06:19 lr 0.000185 time 0.2821 (0.3381) loss 2.2261 (3.3010) grad_norm 2.0171 (2.2070) [2021-04-16 12:46:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][140/1251] eta 0:06:12 lr 0.000185 time 0.2751 (0.3351) loss 3.0435 (3.2808) grad_norm 2.4929 (2.2045) [2021-04-16 12:46:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][150/1251] eta 0:06:06 lr 0.000185 time 0.2570 (0.3326) loss 3.6441 (3.2511) grad_norm 2.4289 (2.2159) [2021-04-16 12:46:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][160/1251] eta 0:05:59 lr 0.000185 time 0.2684 (0.3296) loss 3.4978 (3.2599) grad_norm 1.8882 (2.2098) [2021-04-16 12:46:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][170/1251] eta 0:05:53 lr 0.000185 time 0.2815 (0.3268) loss 3.2698 (3.2497) grad_norm 2.5906 (2.2124) [2021-04-16 12:46:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][180/1251] eta 0:05:46 lr 0.000185 time 0.2800 (0.3240) loss 3.8009 (3.2566) grad_norm 2.2184 (2.2063) [2021-04-16 12:46:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][190/1251] eta 0:05:41 lr 0.000185 time 0.2811 (0.3217) loss 3.4033 (3.2703) grad_norm 2.5764 (2.2072) [2021-04-16 12:46:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][200/1251] eta 0:05:36 lr 0.000185 time 0.3022 (0.3198) loss 3.4296 (3.2704) grad_norm 2.1577 (2.2078) [2021-04-16 12:47:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][210/1251] eta 0:05:31 lr 0.000185 time 0.2740 (0.3182) loss 3.4900 (3.2756) grad_norm 2.3217 (2.2086) [2021-04-16 12:47:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][220/1251] eta 0:05:26 lr 0.000185 time 0.2919 (0.3165) loss 2.6828 (3.2664) grad_norm 2.1673 (2.2170) [2021-04-16 12:47:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][230/1251] eta 0:05:21 lr 0.000185 time 0.2836 (0.3147) loss 2.8789 (3.2595) grad_norm 2.0324 (2.2164) [2021-04-16 12:47:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][240/1251] eta 0:05:16 lr 0.000185 time 0.3070 (0.3134) loss 3.8244 (3.2749) grad_norm 2.4152 (2.2170) [2021-04-16 12:47:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][250/1251] eta 0:05:12 lr 0.000185 time 0.2598 (0.3121) loss 3.3730 (3.2669) grad_norm 2.6423 (2.2236) [2021-04-16 12:47:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][260/1251] eta 0:05:08 lr 0.000185 time 0.3073 (0.3108) loss 3.4462 (3.2688) grad_norm 2.1349 (2.2244) [2021-04-16 12:47:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][270/1251] eta 0:05:03 lr 0.000185 time 0.2886 (0.3097) loss 2.9566 (3.2655) grad_norm 2.1616 (2.2252) [2021-04-16 12:47:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][280/1251] eta 0:05:00 lr 0.000185 time 0.2784 (0.3090) loss 3.0305 (3.2629) grad_norm 1.9272 (2.2196) [2021-04-16 12:47:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][290/1251] eta 0:04:56 lr 0.000185 time 0.3024 (0.3081) loss 3.4524 (3.2639) grad_norm 2.1951 (2.2218) [2021-04-16 12:47:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][300/1251] eta 0:04:52 lr 0.000185 time 0.2743 (0.3072) loss 3.4509 (3.2633) grad_norm 2.0092 (2.2163) [2021-04-16 12:47:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][310/1251] eta 0:04:48 lr 0.000185 time 0.2745 (0.3062) loss 2.8640 (3.2579) grad_norm 2.3179 (2.2151) [2021-04-16 12:47:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][320/1251] eta 0:04:44 lr 0.000184 time 0.2994 (0.3060) loss 3.6241 (3.2580) grad_norm 1.9776 (2.2157) [2021-04-16 12:47:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][330/1251] eta 0:04:41 lr 0.000184 time 0.2836 (0.3053) loss 3.0703 (3.2577) grad_norm 2.1009 (2.2208) [2021-04-16 12:47:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][340/1251] eta 0:04:37 lr 0.000184 time 0.2733 (0.3051) loss 1.9417 (3.2507) grad_norm 2.1156 (2.2223) [2021-04-16 12:47:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][350/1251] eta 0:04:34 lr 0.000184 time 0.3954 (0.3046) loss 2.9570 (3.2471) grad_norm 2.4049 (2.2201) [2021-04-16 12:47:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][360/1251] eta 0:04:31 lr 0.000184 time 0.2751 (0.3046) loss 3.4001 (3.2493) grad_norm 1.9334 (2.2181) [2021-04-16 12:47:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][370/1251] eta 0:04:28 lr 0.000184 time 0.2898 (0.3043) loss 2.9355 (3.2454) grad_norm 2.1929 (2.2172) [2021-04-16 12:47:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][380/1251] eta 0:04:24 lr 0.000184 time 0.2564 (0.3040) loss 3.3597 (3.2454) grad_norm 2.2177 (2.2170) [2021-04-16 12:47:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][390/1251] eta 0:04:21 lr 0.000184 time 0.2511 (0.3033) loss 3.5357 (3.2406) grad_norm 2.3923 (2.2148) [2021-04-16 12:47:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][400/1251] eta 0:04:17 lr 0.000184 time 0.2897 (0.3027) loss 4.2235 (3.2455) grad_norm 2.6167 (2.2190) [2021-04-16 12:47:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][410/1251] eta 0:04:14 lr 0.000184 time 0.2722 (0.3022) loss 3.6228 (3.2463) grad_norm 2.0567 (2.2200) [2021-04-16 12:48:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][420/1251] eta 0:04:10 lr 0.000184 time 0.2847 (0.3018) loss 2.7134 (3.2466) grad_norm 2.1534 (2.2202) [2021-04-16 12:48:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][430/1251] eta 0:04:07 lr 0.000184 time 0.2736 (0.3012) loss 3.2881 (3.2486) grad_norm 2.1836 (2.2225) [2021-04-16 12:48:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][440/1251] eta 0:04:03 lr 0.000184 time 0.2930 (0.3008) loss 3.6728 (3.2508) grad_norm 2.2141 (2.2202) [2021-04-16 12:48:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][450/1251] eta 0:04:00 lr 0.000184 time 0.2631 (0.3003) loss 3.1768 (3.2540) grad_norm 2.2225 (2.2196) [2021-04-16 12:48:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][460/1251] eta 0:03:57 lr 0.000184 time 0.3065 (0.2999) loss 3.2641 (3.2515) grad_norm 2.2492 (2.2190) [2021-04-16 12:48:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][470/1251] eta 0:03:53 lr 0.000184 time 0.2700 (0.2994) loss 3.9017 (3.2499) grad_norm 2.3797 (2.2195) [2021-04-16 12:48:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][480/1251] eta 0:03:50 lr 0.000184 time 0.2744 (0.2989) loss 3.4912 (3.2505) grad_norm 2.1160 (2.2187) [2021-04-16 12:48:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][490/1251] eta 0:03:47 lr 0.000184 time 0.2636 (0.2985) loss 3.3365 (3.2516) grad_norm 2.0534 (2.2164) [2021-04-16 12:48:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][500/1251] eta 0:03:43 lr 0.000184 time 0.2644 (0.2981) loss 3.8108 (3.2522) grad_norm 2.4005 (2.2146) [2021-04-16 12:48:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][510/1251] eta 0:03:40 lr 0.000184 time 0.2806 (0.2977) loss 3.2763 (3.2530) grad_norm 2.6646 (2.2145) [2021-04-16 12:48:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][520/1251] eta 0:03:37 lr 0.000184 time 0.2749 (0.2973) loss 2.9831 (3.2516) grad_norm 2.1639 (2.2129) [2021-04-16 12:48:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][530/1251] eta 0:03:34 lr 0.000184 time 0.2773 (0.2970) loss 3.5725 (3.2488) grad_norm 2.5643 (2.2153) [2021-04-16 12:48:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][540/1251] eta 0:03:31 lr 0.000184 time 0.2763 (0.2968) loss 3.4997 (3.2488) grad_norm 2.2985 (2.2150) [2021-04-16 12:48:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][550/1251] eta 0:03:27 lr 0.000184 time 0.2540 (0.2965) loss 3.8594 (3.2531) grad_norm 2.7662 (2.2190) [2021-04-16 12:48:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][560/1251] eta 0:03:24 lr 0.000184 time 0.3113 (0.2965) loss 3.5203 (3.2504) grad_norm 2.1751 (2.2215) [2021-04-16 12:48:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][570/1251] eta 0:03:21 lr 0.000184 time 0.2769 (0.2964) loss 3.2747 (3.2511) grad_norm 2.3037 (2.2245) [2021-04-16 12:48:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][580/1251] eta 0:03:18 lr 0.000184 time 0.2771 (0.2964) loss 3.1350 (3.2521) grad_norm 2.7124 (2.2282) [2021-04-16 12:48:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][590/1251] eta 0:03:15 lr 0.000184 time 0.2743 (0.2963) loss 3.6036 (3.2544) grad_norm 2.5958 (2.2281) [2021-04-16 12:48:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][600/1251] eta 0:03:12 lr 0.000184 time 0.2758 (0.2959) loss 3.9285 (3.2574) grad_norm 2.0384 (2.2270) [2021-04-16 12:48:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][610/1251] eta 0:03:09 lr 0.000184 time 0.2736 (0.2957) loss 3.1637 (3.2590) grad_norm 2.1988 (2.2262) [2021-04-16 12:48:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][620/1251] eta 0:03:06 lr 0.000184 time 0.2767 (0.2955) loss 4.0900 (3.2621) grad_norm 2.5205 (2.2257) [2021-04-16 12:49:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][630/1251] eta 0:03:03 lr 0.000184 time 0.3358 (0.2953) loss 2.9408 (3.2583) grad_norm 1.9885 (2.2255) [2021-04-16 12:49:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][640/1251] eta 0:03:00 lr 0.000183 time 0.3065 (0.2951) loss 3.5540 (3.2611) grad_norm 2.2065 (2.2270) [2021-04-16 12:49:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][650/1251] eta 0:02:57 lr 0.000183 time 0.2775 (0.2950) loss 1.9187 (3.2525) grad_norm 1.7852 (2.2252) [2021-04-16 12:49:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][660/1251] eta 0:02:54 lr 0.000183 time 0.2583 (0.2947) loss 3.5200 (3.2500) grad_norm 2.1219 (2.2260) [2021-04-16 12:49:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][670/1251] eta 0:02:51 lr 0.000183 time 0.2727 (0.2945) loss 3.3137 (3.2558) grad_norm 1.8106 (2.2259) [2021-04-16 12:49:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][680/1251] eta 0:02:48 lr 0.000183 time 0.2760 (0.2943) loss 2.1578 (3.2545) grad_norm 2.2756 (2.2243) [2021-04-16 12:49:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][690/1251] eta 0:02:45 lr 0.000183 time 0.2742 (0.2943) loss 3.0977 (3.2541) grad_norm 2.0135 (2.2229) [2021-04-16 12:49:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][700/1251] eta 0:02:41 lr 0.000183 time 0.2588 (0.2940) loss 3.7476 (3.2535) grad_norm 2.1975 (2.2220) [2021-04-16 12:49:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][710/1251] eta 0:02:38 lr 0.000183 time 0.2611 (0.2939) loss 2.6117 (3.2504) grad_norm 2.1587 (2.2211) [2021-04-16 12:49:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][720/1251] eta 0:02:36 lr 0.000183 time 0.2898 (0.2939) loss 3.1984 (3.2492) grad_norm 2.3165 (2.2214) [2021-04-16 12:49:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][730/1251] eta 0:02:33 lr 0.000183 time 0.2815 (0.2939) loss 3.4346 (3.2500) grad_norm 2.2137 (2.2220) [2021-04-16 12:49:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][740/1251] eta 0:02:30 lr 0.000183 time 0.2750 (0.2937) loss 2.8082 (3.2470) grad_norm 2.0503 (2.2217) [2021-04-16 12:49:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][750/1251] eta 0:02:27 lr 0.000183 time 0.2864 (0.2935) loss 3.8886 (3.2506) grad_norm 3.1070 (2.2285) [2021-04-16 12:49:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][760/1251] eta 0:02:24 lr 0.000183 time 0.2949 (0.2933) loss 3.1890 (3.2494) grad_norm 2.2659 (2.2286) [2021-04-16 12:49:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][770/1251] eta 0:02:21 lr 0.000183 time 0.2548 (0.2931) loss 2.6064 (3.2468) grad_norm 2.1423 (2.2319) [2021-04-16 12:49:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][780/1251] eta 0:02:17 lr 0.000183 time 0.2769 (0.2930) loss 3.3715 (3.2472) grad_norm 2.3503 (2.2312) [2021-04-16 12:49:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][790/1251] eta 0:02:15 lr 0.000183 time 0.2933 (0.2928) loss 2.4444 (3.2476) grad_norm 2.2319 (2.2320) [2021-04-16 12:49:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][800/1251] eta 0:02:11 lr 0.000183 time 0.2855 (0.2927) loss 3.7834 (3.2486) grad_norm 2.1736 (2.2336) [2021-04-16 12:49:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][810/1251] eta 0:02:09 lr 0.000183 time 0.2528 (0.2926) loss 3.7570 (3.2505) grad_norm 2.1035 (2.2348) [2021-04-16 12:49:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][820/1251] eta 0:02:06 lr 0.000183 time 0.2821 (0.2924) loss 4.0021 (3.2509) grad_norm 2.0989 (2.2341) [2021-04-16 12:49:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][830/1251] eta 0:02:03 lr 0.000183 time 0.2742 (0.2923) loss 2.8411 (3.2484) grad_norm 2.2563 (2.2328) [2021-04-16 12:50:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][840/1251] eta 0:02:00 lr 0.000183 time 0.2938 (0.2921) loss 3.0362 (3.2454) grad_norm 1.9718 (2.2321) [2021-04-16 12:50:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][850/1251] eta 0:01:57 lr 0.000183 time 0.2841 (0.2921) loss 3.4599 (3.2441) grad_norm 2.3138 (2.2309) [2021-04-16 12:50:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][860/1251] eta 0:01:54 lr 0.000183 time 0.2780 (0.2919) loss 3.6512 (3.2418) grad_norm 2.4937 (2.2325) [2021-04-16 12:50:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][870/1251] eta 0:01:51 lr 0.000183 time 0.2588 (0.2918) loss 3.5965 (3.2400) grad_norm 2.3916 (2.2327) [2021-04-16 12:50:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][880/1251] eta 0:01:48 lr 0.000183 time 0.2894 (0.2917) loss 3.5631 (3.2379) grad_norm 2.3378 (2.2333) [2021-04-16 12:50:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][890/1251] eta 0:01:45 lr 0.000183 time 0.2661 (0.2917) loss 3.4342 (3.2374) grad_norm 2.5707 (2.2340) [2021-04-16 12:50:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][900/1251] eta 0:01:42 lr 0.000183 time 0.2743 (0.2915) loss 3.2354 (3.2360) grad_norm 3.7118 (2.2365) [2021-04-16 12:50:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][910/1251] eta 0:01:39 lr 0.000183 time 0.2778 (0.2914) loss 3.0965 (3.2326) grad_norm 2.0118 (2.2357) [2021-04-16 12:50:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][920/1251] eta 0:01:36 lr 0.000183 time 0.2683 (0.2914) loss 3.6737 (3.2333) grad_norm 2.0161 (2.2357) [2021-04-16 12:50:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][930/1251] eta 0:01:33 lr 0.000183 time 0.3747 (0.2916) loss 2.6633 (3.2302) grad_norm 2.2303 (2.2358) [2021-04-16 12:50:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][940/1251] eta 0:01:30 lr 0.000183 time 0.3016 (0.2915) loss 3.1112 (3.2308) grad_norm 2.0986 (2.2367) [2021-04-16 12:50:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][950/1251] eta 0:01:27 lr 0.000183 time 0.2847 (0.2914) loss 3.2799 (3.2324) grad_norm 2.1684 (2.2357) [2021-04-16 12:50:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][960/1251] eta 0:01:24 lr 0.000182 time 0.2622 (0.2914) loss 4.0621 (3.2342) grad_norm 2.2532 (2.2358) [2021-04-16 12:50:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][970/1251] eta 0:01:21 lr 0.000182 time 0.2749 (0.2913) loss 2.6793 (3.2326) grad_norm 2.2019 (2.2350) [2021-04-16 12:50:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][980/1251] eta 0:01:18 lr 0.000182 time 0.2929 (0.2911) loss 3.3376 (3.2313) grad_norm 2.2549 (2.2343) [2021-04-16 12:50:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][990/1251] eta 0:01:15 lr 0.000182 time 0.3062 (0.2910) loss 3.4611 (3.2302) grad_norm 2.0657 (2.2345) [2021-04-16 12:50:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][1000/1251] eta 0:01:12 lr 0.000182 time 0.2682 (0.2908) loss 3.5974 (3.2301) grad_norm 2.1772 (2.2341) [2021-04-16 12:50:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][1010/1251] eta 0:01:10 lr 0.000182 time 0.2997 (0.2907) loss 3.2689 (3.2341) grad_norm 2.3516 (2.2351) [2021-04-16 12:50:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][1020/1251] eta 0:01:07 lr 0.000182 time 0.2777 (0.2905) loss 2.0504 (3.2310) grad_norm 2.6834 (2.2342) [2021-04-16 12:50:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][1030/1251] eta 0:01:04 lr 0.000182 time 0.2676 (0.2904) loss 3.5405 (3.2314) grad_norm 2.1712 (2.2342) [2021-04-16 12:50:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][1040/1251] eta 0:01:01 lr 0.000182 time 0.2845 (0.2903) loss 2.1112 (3.2277) grad_norm 1.9250 (2.2342) [2021-04-16 12:50:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][1050/1251] eta 0:00:58 lr 0.000182 time 0.2976 (0.2903) loss 3.5163 (3.2284) grad_norm 2.3020 (2.2342) [2021-04-16 12:51:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][1060/1251] eta 0:00:55 lr 0.000182 time 0.2814 (0.2902) loss 3.9635 (3.2286) grad_norm 2.0554 (2.2341) [2021-04-16 12:51:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][1070/1251] eta 0:00:52 lr 0.000182 time 0.2612 (0.2902) loss 2.8212 (3.2291) grad_norm 2.1600 (2.2340) [2021-04-16 12:51:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][1080/1251] eta 0:00:49 lr 0.000182 time 0.2860 (0.2900) loss 3.2613 (3.2311) grad_norm 2.3463 (2.2331) [2021-04-16 12:51:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][1090/1251] eta 0:00:46 lr 0.000182 time 0.2720 (0.2900) loss 3.5932 (3.2332) grad_norm 2.3047 (2.2328) [2021-04-16 12:51:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][1100/1251] eta 0:00:43 lr 0.000182 time 0.2787 (0.2899) loss 3.2301 (3.2323) grad_norm 1.9395 (2.2318) [2021-04-16 12:51:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][1110/1251] eta 0:00:40 lr 0.000182 time 0.2667 (0.2899) loss 3.8746 (3.2304) grad_norm 2.3903 (2.2319) [2021-04-16 12:51:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][1120/1251] eta 0:00:37 lr 0.000182 time 0.2660 (0.2897) loss 3.7467 (3.2305) grad_norm 2.0014 (2.2322) [2021-04-16 12:51:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][1130/1251] eta 0:00:35 lr 0.000182 time 0.2617 (0.2896) loss 3.1608 (3.2308) grad_norm 1.9258 (2.2313) [2021-04-16 12:51:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][1140/1251] eta 0:00:32 lr 0.000182 time 0.2869 (0.2895) loss 4.1331 (3.2299) grad_norm 2.2022 (2.2316) [2021-04-16 12:51:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][1150/1251] eta 0:00:29 lr 0.000182 time 0.3735 (0.2896) loss 2.3859 (3.2312) grad_norm 2.2997 (2.2318) [2021-04-16 12:51:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][1160/1251] eta 0:00:26 lr 0.000182 time 0.2655 (0.2895) loss 2.3624 (3.2333) grad_norm 2.9942 (2.2314) [2021-04-16 12:51:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][1170/1251] eta 0:00:23 lr 0.000182 time 0.2970 (0.2896) loss 4.2132 (3.2354) grad_norm 2.1801 (2.2308) [2021-04-16 12:51:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][1180/1251] eta 0:00:20 lr 0.000182 time 0.2786 (0.2895) loss 3.2391 (3.2374) grad_norm 1.9249 (2.2302) [2021-04-16 12:51:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][1190/1251] eta 0:00:17 lr 0.000182 time 0.2545 (0.2895) loss 3.5443 (3.2364) grad_norm 2.0083 (2.2296) [2021-04-16 12:51:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][1200/1251] eta 0:00:14 lr 0.000182 time 0.2699 (0.2894) loss 3.2644 (3.2358) grad_norm 2.3488 (2.2294) [2021-04-16 12:51:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][1210/1251] eta 0:00:11 lr 0.000182 time 0.2701 (0.2893) loss 3.4890 (3.2345) grad_norm 2.2698 (2.2301) [2021-04-16 12:51:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][1220/1251] eta 0:00:08 lr 0.000182 time 0.2757 (0.2892) loss 3.4417 (3.2345) grad_norm 2.2350 (2.2304) [2021-04-16 12:51:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][1230/1251] eta 0:00:06 lr 0.000182 time 0.2699 (0.2891) loss 3.3205 (3.2359) grad_norm 2.3615 (2.2300) [2021-04-16 12:51:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][1240/1251] eta 0:00:03 lr 0.000182 time 0.2485 (0.2889) loss 2.1974 (3.2334) grad_norm 2.6035 (2.2292) [2021-04-16 12:51:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [217/300][1250/1251] eta 0:00:00 lr 0.000182 time 0.2488 (0.2886) loss 3.7724 (3.2322) grad_norm 1.9747 (2.2297) [2021-04-16 12:51:59 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 217 training takes 0:06:05 [2021-04-16 12:51:59 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_217.pth saving...... [2021-04-16 12:52:09 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_217.pth saved !!! [2021-04-16 12:52:10 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.069 (1.069) Loss 0.8418 (0.8418) Acc@1 81.152 (81.152) Acc@5 94.629 (94.629) [2021-04-16 12:52:12 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.104 (0.195) Loss 0.9260 (0.8918) Acc@1 77.637 (79.208) Acc@5 94.629 (94.611) [2021-04-16 12:52:14 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.223 (0.198) Loss 0.9092 (0.8879) Acc@1 79.785 (79.404) Acc@5 95.020 (94.694) [2021-04-16 12:52:17 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.441 (0.239) Loss 0.9177 (0.8933) Acc@1 78.125 (79.193) Acc@5 94.727 (94.692) [2021-04-16 12:52:18 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.214) Loss 0.8864 (0.8986) Acc@1 80.273 (78.982) Acc@5 93.457 (94.612) [2021-04-16 12:52:29 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 79.044 Acc@5 94.672 [2021-04-16 12:52:29 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 79.0% [2021-04-16 12:52:29 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 79.17% [2021-04-16 12:52:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][0/1251] eta 2:00:34 lr 0.000182 time 5.7830 (5.7830) loss 2.7703 (2.7703) grad_norm 1.9513 (1.9513) [2021-04-16 12:52:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][10/1251] eta 0:15:59 lr 0.000182 time 0.2696 (0.7730) loss 3.3832 (3.3779) grad_norm 2.0890 (2.1334) [2021-04-16 12:52:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][20/1251] eta 0:11:00 lr 0.000181 time 0.2608 (0.5362) loss 3.5855 (3.4127) grad_norm 2.2032 (2.1909) [2021-04-16 12:52:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][30/1251] eta 0:09:11 lr 0.000181 time 0.2756 (0.4519) loss 2.4839 (3.4250) grad_norm 2.6446 (2.2324) [2021-04-16 12:52:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][40/1251] eta 0:08:16 lr 0.000181 time 0.2870 (0.4101) loss 3.6938 (3.3993) grad_norm 2.5405 (2.2250) [2021-04-16 12:52:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][50/1251] eta 0:07:40 lr 0.000181 time 0.2675 (0.3832) loss 3.7994 (3.3456) grad_norm 2.4722 (2.2283) [2021-04-16 12:52:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][60/1251] eta 0:07:15 lr 0.000181 time 0.2847 (0.3659) loss 3.4299 (3.3531) grad_norm 2.0132 (2.2423) [2021-04-16 12:52:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][70/1251] eta 0:06:58 lr 0.000181 time 0.2687 (0.3543) loss 3.2210 (3.3214) grad_norm 2.0987 (2.2450) [2021-04-16 12:52:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][80/1251] eta 0:06:43 lr 0.000181 time 0.2582 (0.3444) loss 3.7612 (3.3357) grad_norm 2.2901 (2.2549) [2021-04-16 12:52:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][90/1251] eta 0:06:31 lr 0.000181 time 0.2785 (0.3375) loss 3.4705 (3.3238) grad_norm 2.6367 (2.2645) [2021-04-16 12:53:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][100/1251] eta 0:06:23 lr 0.000181 time 0.2725 (0.3328) loss 3.6590 (3.3287) grad_norm 2.1217 (2.2546) [2021-04-16 12:53:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][110/1251] eta 0:06:13 lr 0.000181 time 0.2731 (0.3274) loss 2.5364 (3.3298) grad_norm 2.1685 (2.2614) [2021-04-16 12:53:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][120/1251] eta 0:06:06 lr 0.000181 time 0.2865 (0.3241) loss 3.6997 (3.3114) grad_norm 2.1417 (2.2512) [2021-04-16 12:53:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][130/1251] eta 0:05:59 lr 0.000181 time 0.2777 (0.3210) loss 3.9828 (3.3149) grad_norm 2.1162 (2.2527) [2021-04-16 12:53:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][140/1251] eta 0:05:55 lr 0.000181 time 0.4014 (0.3196) loss 3.8167 (3.3170) grad_norm 1.9529 (2.2564) [2021-04-16 12:53:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][150/1251] eta 0:05:49 lr 0.000181 time 0.2545 (0.3174) loss 3.5943 (3.3183) grad_norm 2.2103 (2.2514) [2021-04-16 12:53:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][160/1251] eta 0:05:43 lr 0.000181 time 0.2571 (0.3147) loss 2.9113 (3.3170) grad_norm 1.7676 (2.2386) [2021-04-16 12:53:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][170/1251] eta 0:05:37 lr 0.000181 time 0.3065 (0.3125) loss 3.3793 (3.3190) grad_norm 2.1494 (2.2298) [2021-04-16 12:53:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][180/1251] eta 0:05:32 lr 0.000181 time 0.2644 (0.3106) loss 3.3858 (3.3137) grad_norm 2.2728 (2.2378) [2021-04-16 12:53:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][190/1251] eta 0:05:27 lr 0.000181 time 0.2737 (0.3086) loss 3.5508 (3.3248) grad_norm 2.4962 (2.2407) [2021-04-16 12:53:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][200/1251] eta 0:05:22 lr 0.000181 time 0.2819 (0.3071) loss 3.9234 (3.3241) grad_norm 2.2414 (2.2448) [2021-04-16 12:53:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][210/1251] eta 0:05:17 lr 0.000181 time 0.2764 (0.3054) loss 2.9382 (3.3255) grad_norm 2.0509 (2.2488) [2021-04-16 12:53:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][220/1251] eta 0:05:13 lr 0.000181 time 0.2475 (0.3040) loss 2.5729 (3.3314) grad_norm 2.3311 (2.2460) [2021-04-16 12:53:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][230/1251] eta 0:05:09 lr 0.000181 time 0.2673 (0.3027) loss 2.4979 (3.3192) grad_norm 2.0879 (2.2424) [2021-04-16 12:53:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][240/1251] eta 0:05:05 lr 0.000181 time 0.2686 (0.3020) loss 3.0800 (3.3186) grad_norm 2.0889 (2.2462) [2021-04-16 12:53:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][250/1251] eta 0:05:01 lr 0.000181 time 0.2708 (0.3009) loss 2.4425 (3.3044) grad_norm 2.3998 (2.2436) [2021-04-16 12:53:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][260/1251] eta 0:04:57 lr 0.000181 time 0.2653 (0.3001) loss 3.3728 (3.3040) grad_norm 1.9377 (2.2444) [2021-04-16 12:53:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][270/1251] eta 0:04:53 lr 0.000181 time 0.2473 (0.2991) loss 3.2711 (3.2892) grad_norm 2.3089 (2.2520) [2021-04-16 12:53:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][280/1251] eta 0:04:49 lr 0.000181 time 0.2660 (0.2983) loss 3.0278 (3.2807) grad_norm 2.0840 (2.2495) [2021-04-16 12:53:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][290/1251] eta 0:04:45 lr 0.000181 time 0.2807 (0.2975) loss 3.4667 (3.2876) grad_norm 2.0413 (2.2468) [2021-04-16 12:53:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][300/1251] eta 0:04:42 lr 0.000181 time 0.2846 (0.2968) loss 3.1289 (3.2747) grad_norm 1.8561 (2.2453) [2021-04-16 12:54:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][310/1251] eta 0:04:38 lr 0.000181 time 0.2631 (0.2961) loss 3.1027 (3.2638) grad_norm 1.9551 (2.2446) [2021-04-16 12:54:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][320/1251] eta 0:04:35 lr 0.000181 time 0.2624 (0.2955) loss 3.0145 (3.2611) grad_norm 2.2681 (2.2511) [2021-04-16 12:54:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][330/1251] eta 0:04:31 lr 0.000181 time 0.2875 (0.2951) loss 3.3098 (3.2626) grad_norm 1.9933 (2.2534) [2021-04-16 12:54:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][340/1251] eta 0:04:28 lr 0.000180 time 0.2902 (0.2945) loss 2.6877 (3.2641) grad_norm 2.1108 (2.2578) [2021-04-16 12:54:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][350/1251] eta 0:04:24 lr 0.000180 time 0.2965 (0.2940) loss 3.0609 (3.2552) grad_norm 2.0729 (2.2625) [2021-04-16 12:54:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][360/1251] eta 0:04:22 lr 0.000180 time 0.3006 (0.2942) loss 2.4213 (3.2519) grad_norm 1.9264 (2.2629) [2021-04-16 12:54:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][370/1251] eta 0:04:18 lr 0.000180 time 0.2750 (0.2939) loss 3.2684 (3.2512) grad_norm 2.3907 (2.2625) [2021-04-16 12:54:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][380/1251] eta 0:04:15 lr 0.000180 time 0.2573 (0.2936) loss 3.8842 (3.2517) grad_norm 2.8399 (2.2626) [2021-04-16 12:54:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][390/1251] eta 0:04:12 lr 0.000180 time 0.2863 (0.2932) loss 3.0367 (3.2514) grad_norm 1.9358 (2.2612) [2021-04-16 12:54:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][400/1251] eta 0:04:09 lr 0.000180 time 0.2889 (0.2928) loss 3.1436 (3.2519) grad_norm 2.3766 (2.2626) [2021-04-16 12:54:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][410/1251] eta 0:04:06 lr 0.000180 time 0.3019 (0.2926) loss 3.4276 (3.2500) grad_norm 2.2878 (2.2629) [2021-04-16 12:54:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][420/1251] eta 0:04:02 lr 0.000180 time 0.2731 (0.2922) loss 3.5092 (3.2553) grad_norm 2.1301 (2.2643) [2021-04-16 12:54:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][430/1251] eta 0:03:59 lr 0.000180 time 0.2932 (0.2919) loss 3.7515 (3.2541) grad_norm 2.2071 (2.2657) [2021-04-16 12:54:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][440/1251] eta 0:03:56 lr 0.000180 time 0.2715 (0.2914) loss 3.6705 (3.2533) grad_norm 2.4968 (2.2659) [2021-04-16 12:54:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][450/1251] eta 0:03:53 lr 0.000180 time 0.2429 (0.2910) loss 3.4574 (3.2498) grad_norm 2.0722 (2.2654) [2021-04-16 12:54:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][460/1251] eta 0:03:49 lr 0.000180 time 0.2589 (0.2906) loss 2.2438 (3.2527) grad_norm 2.2883 (2.2680) [2021-04-16 12:54:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][470/1251] eta 0:03:46 lr 0.000180 time 0.2688 (0.2903) loss 3.8119 (3.2526) grad_norm 2.3893 (2.2704) [2021-04-16 12:54:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][480/1251] eta 0:03:43 lr 0.000180 time 0.2845 (0.2901) loss 3.6091 (3.2545) grad_norm 2.1996 (2.2695) [2021-04-16 12:54:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][490/1251] eta 0:03:40 lr 0.000180 time 0.2755 (0.2900) loss 2.5330 (3.2473) grad_norm 2.0757 (2.2686) [2021-04-16 12:54:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][500/1251] eta 0:03:37 lr 0.000180 time 0.2978 (0.2898) loss 2.0724 (3.2383) grad_norm 2.3876 (2.2687) [2021-04-16 12:54:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][510/1251] eta 0:03:34 lr 0.000180 time 0.2692 (0.2894) loss 3.1633 (3.2384) grad_norm 1.8952 (2.2701) [2021-04-16 12:54:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][520/1251] eta 0:03:31 lr 0.000180 time 0.2489 (0.2895) loss 3.9890 (3.2357) grad_norm 2.2156 (2.2678) [2021-04-16 12:55:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][530/1251] eta 0:03:28 lr 0.000180 time 0.2700 (0.2892) loss 3.2600 (3.2425) grad_norm 2.2805 (2.2647) [2021-04-16 12:55:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][540/1251] eta 0:03:25 lr 0.000180 time 0.2793 (0.2892) loss 3.3702 (3.2403) grad_norm 2.0591 (2.2648) [2021-04-16 12:55:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][550/1251] eta 0:03:22 lr 0.000180 time 0.2599 (0.2888) loss 3.4505 (3.2409) grad_norm 3.6023 (2.2659) [2021-04-16 12:55:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][560/1251] eta 0:03:19 lr 0.000180 time 0.3059 (0.2886) loss 3.7952 (3.2402) grad_norm 2.1829 (2.2640) [2021-04-16 12:55:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][570/1251] eta 0:03:16 lr 0.000180 time 0.2657 (0.2887) loss 3.6470 (3.2457) grad_norm 2.8286 (2.2621) [2021-04-16 12:55:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][580/1251] eta 0:03:13 lr 0.000180 time 0.2802 (0.2887) loss 3.2640 (3.2475) grad_norm 2.0043 (2.2628) [2021-04-16 12:55:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][590/1251] eta 0:03:10 lr 0.000180 time 0.2611 (0.2886) loss 3.6988 (3.2520) grad_norm 2.4862 (2.2620) [2021-04-16 12:55:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][600/1251] eta 0:03:07 lr 0.000180 time 0.2820 (0.2885) loss 3.8716 (3.2546) grad_norm 2.4942 (2.2612) [2021-04-16 12:55:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][610/1251] eta 0:03:04 lr 0.000180 time 0.3015 (0.2883) loss 3.1352 (3.2564) grad_norm 2.2424 (2.2591) [2021-04-16 12:55:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][620/1251] eta 0:03:01 lr 0.000180 time 0.2586 (0.2881) loss 2.0641 (3.2532) grad_norm 1.9629 (2.2562) [2021-04-16 12:55:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][630/1251] eta 0:02:58 lr 0.000180 time 0.2902 (0.2880) loss 2.2497 (3.2555) grad_norm 1.9788 (2.2546) [2021-04-16 12:55:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][640/1251] eta 0:02:55 lr 0.000180 time 0.2556 (0.2878) loss 3.5358 (3.2526) grad_norm 2.1119 (2.2571) [2021-04-16 12:55:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][650/1251] eta 0:02:52 lr 0.000180 time 0.2639 (0.2877) loss 3.7310 (3.2549) grad_norm 2.0436 (2.2554) [2021-04-16 12:55:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][660/1251] eta 0:02:49 lr 0.000179 time 0.2639 (0.2875) loss 2.9598 (3.2550) grad_norm 2.1757 (nan) [2021-04-16 12:55:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][670/1251] eta 0:02:46 lr 0.000179 time 0.2637 (0.2872) loss 3.2681 (3.2523) grad_norm 2.1769 (nan) [2021-04-16 12:55:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][680/1251] eta 0:02:44 lr 0.000179 time 0.2973 (0.2872) loss 2.5765 (3.2529) grad_norm 2.0981 (nan) [2021-04-16 12:55:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][690/1251] eta 0:02:41 lr 0.000179 time 0.2781 (0.2873) loss 3.3433 (3.2531) grad_norm 2.4228 (nan) [2021-04-16 12:55:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][700/1251] eta 0:02:38 lr 0.000179 time 0.2803 (0.2871) loss 2.8216 (3.2551) grad_norm 2.1307 (nan) [2021-04-16 12:55:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][710/1251] eta 0:02:35 lr 0.000179 time 0.2686 (0.2872) loss 3.0138 (3.2575) grad_norm 2.0170 (nan) [2021-04-16 12:55:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][720/1251] eta 0:02:32 lr 0.000179 time 0.2746 (0.2875) loss 3.6723 (3.2572) grad_norm 1.8851 (nan) [2021-04-16 12:55:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][730/1251] eta 0:02:29 lr 0.000179 time 0.3103 (0.2876) loss 3.5566 (3.2583) grad_norm 2.1117 (nan) [2021-04-16 12:56:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][740/1251] eta 0:02:26 lr 0.000179 time 0.2920 (0.2875) loss 3.4638 (3.2593) grad_norm 1.9754 (nan) [2021-04-16 12:56:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][750/1251] eta 0:02:24 lr 0.000179 time 0.2838 (0.2875) loss 2.8343 (3.2595) grad_norm 2.3348 (nan) [2021-04-16 12:56:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][760/1251] eta 0:02:21 lr 0.000179 time 0.2593 (0.2874) loss 3.4915 (3.2579) grad_norm 2.0784 (nan) [2021-04-16 12:56:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][770/1251] eta 0:02:18 lr 0.000179 time 0.2887 (0.2874) loss 3.1723 (3.2605) grad_norm 2.1091 (nan) [2021-04-16 12:56:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][780/1251] eta 0:02:15 lr 0.000179 time 0.2755 (0.2873) loss 3.4683 (3.2615) grad_norm 2.2250 (nan) [2021-04-16 12:56:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][790/1251] eta 0:02:12 lr 0.000179 time 0.2852 (0.2871) loss 3.8169 (3.2625) grad_norm 2.6185 (nan) [2021-04-16 12:56:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][800/1251] eta 0:02:09 lr 0.000179 time 0.2593 (0.2870) loss 3.0907 (3.2651) grad_norm 2.1991 (nan) [2021-04-16 12:56:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][810/1251] eta 0:02:06 lr 0.000179 time 0.2550 (0.2869) loss 3.1405 (3.2649) grad_norm 1.8318 (nan) [2021-04-16 12:56:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][820/1251] eta 0:02:03 lr 0.000179 time 0.2691 (0.2868) loss 3.4566 (3.2605) grad_norm 2.4257 (nan) [2021-04-16 12:56:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][830/1251] eta 0:02:00 lr 0.000179 time 0.2771 (0.2866) loss 1.9767 (3.2583) grad_norm 2.1949 (nan) [2021-04-16 12:56:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][840/1251] eta 0:01:57 lr 0.000179 time 0.2720 (0.2865) loss 3.3492 (3.2567) grad_norm 1.9482 (nan) [2021-04-16 12:56:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][850/1251] eta 0:01:54 lr 0.000179 time 0.2597 (0.2864) loss 3.0218 (3.2548) grad_norm 2.2381 (nan) [2021-04-16 12:56:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][860/1251] eta 0:01:51 lr 0.000179 time 0.2619 (0.2864) loss 3.5260 (3.2562) grad_norm 2.3515 (nan) [2021-04-16 12:56:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][870/1251] eta 0:01:49 lr 0.000179 time 0.2520 (0.2862) loss 2.9022 (3.2568) grad_norm 1.8641 (nan) [2021-04-16 12:56:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][880/1251] eta 0:01:46 lr 0.000179 time 0.2767 (0.2861) loss 3.8696 (3.2578) grad_norm 2.0728 (nan) [2021-04-16 12:56:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][890/1251] eta 0:01:43 lr 0.000179 time 0.2907 (0.2860) loss 3.6519 (3.2594) grad_norm 2.3359 (nan) [2021-04-16 12:56:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][900/1251] eta 0:01:40 lr 0.000179 time 0.2789 (0.2858) loss 3.1383 (3.2599) grad_norm 2.3665 (nan) [2021-04-16 12:56:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][910/1251] eta 0:01:37 lr 0.000179 time 0.2746 (0.2857) loss 3.9153 (3.2614) grad_norm 2.3068 (nan) [2021-04-16 12:56:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][920/1251] eta 0:01:34 lr 0.000179 time 0.2695 (0.2856) loss 3.3829 (3.2630) grad_norm 2.6981 (nan) [2021-04-16 12:56:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][930/1251] eta 0:01:31 lr 0.000179 time 0.2471 (0.2856) loss 3.4913 (3.2625) grad_norm 2.1174 (nan) [2021-04-16 12:56:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][940/1251] eta 0:01:28 lr 0.000179 time 0.2768 (0.2858) loss 3.5988 (3.2639) grad_norm 2.6096 (nan) [2021-04-16 12:57:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][950/1251] eta 0:01:25 lr 0.000179 time 0.2748 (0.2856) loss 2.7909 (3.2625) grad_norm 2.2834 (nan) [2021-04-16 12:57:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][960/1251] eta 0:01:23 lr 0.000179 time 0.2963 (0.2855) loss 2.8870 (3.2624) grad_norm 2.2623 (nan) [2021-04-16 12:57:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][970/1251] eta 0:01:20 lr 0.000179 time 0.2619 (0.2854) loss 3.5742 (3.2632) grad_norm 2.8277 (nan) [2021-04-16 12:57:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][980/1251] eta 0:01:17 lr 0.000178 time 0.2478 (0.2853) loss 2.5946 (3.2613) grad_norm 2.9568 (nan) [2021-04-16 12:57:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][990/1251] eta 0:01:14 lr 0.000178 time 0.2638 (0.2852) loss 3.6751 (3.2638) grad_norm 2.3774 (nan) [2021-04-16 12:57:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][1000/1251] eta 0:01:11 lr 0.000178 time 0.2587 (0.2851) loss 2.2226 (3.2619) grad_norm 2.2841 (nan) [2021-04-16 12:57:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][1010/1251] eta 0:01:08 lr 0.000178 time 0.2704 (0.2850) loss 3.7718 (3.2629) grad_norm 2.3283 (nan) [2021-04-16 12:57:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][1020/1251] eta 0:01:05 lr 0.000178 time 0.2852 (0.2849) loss 3.4660 (3.2623) grad_norm 2.1499 (nan) [2021-04-16 12:57:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][1030/1251] eta 0:01:02 lr 0.000178 time 0.3013 (0.2849) loss 3.6487 (3.2616) grad_norm 2.3209 (nan) [2021-04-16 12:57:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][1040/1251] eta 0:01:00 lr 0.000178 time 0.2421 (0.2848) loss 2.4094 (3.2585) grad_norm 3.0541 (nan) [2021-04-16 12:57:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][1050/1251] eta 0:00:57 lr 0.000178 time 0.3000 (0.2847) loss 3.4977 (3.2580) grad_norm 2.4136 (nan) [2021-04-16 12:57:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][1060/1251] eta 0:00:54 lr 0.000178 time 0.2791 (0.2846) loss 3.8453 (3.2596) grad_norm 2.0240 (nan) [2021-04-16 12:57:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][1070/1251] eta 0:00:51 lr 0.000178 time 0.2757 (0.2846) loss 3.2981 (3.2597) grad_norm 1.8945 (nan) [2021-04-16 12:57:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][1080/1251] eta 0:00:48 lr 0.000178 time 0.2792 (0.2845) loss 3.1350 (3.2578) grad_norm 2.0290 (nan) [2021-04-16 12:57:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][1090/1251] eta 0:00:45 lr 0.000178 time 0.2875 (0.2846) loss 2.8043 (3.2584) grad_norm 2.2783 (nan) [2021-04-16 12:57:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][1100/1251] eta 0:00:42 lr 0.000178 time 0.2583 (0.2845) loss 2.4979 (3.2594) grad_norm 2.3141 (nan) [2021-04-16 12:57:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][1110/1251] eta 0:00:40 lr 0.000178 time 0.2960 (0.2845) loss 3.1414 (3.2575) grad_norm 2.2372 (nan) [2021-04-16 12:57:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][1120/1251] eta 0:00:37 lr 0.000178 time 0.2863 (0.2844) loss 3.5026 (3.2563) grad_norm 2.2058 (nan) [2021-04-16 12:57:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][1130/1251] eta 0:00:34 lr 0.000178 time 0.2964 (0.2843) loss 3.3395 (3.2583) grad_norm 1.9431 (nan) [2021-04-16 12:57:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][1140/1251] eta 0:00:31 lr 0.000178 time 0.2796 (0.2843) loss 3.2352 (3.2585) grad_norm 2.6585 (nan) [2021-04-16 12:57:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][1150/1251] eta 0:00:28 lr 0.000178 time 0.3066 (0.2843) loss 2.6083 (3.2573) grad_norm 2.2389 (nan) [2021-04-16 12:57:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][1160/1251] eta 0:00:25 lr 0.000178 time 0.2555 (0.2843) loss 3.5106 (3.2568) grad_norm 2.8557 (nan) [2021-04-16 12:58:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][1170/1251] eta 0:00:23 lr 0.000178 time 0.2645 (0.2843) loss 3.4052 (3.2578) grad_norm 2.2428 (nan) [2021-04-16 12:58:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][1180/1251] eta 0:00:20 lr 0.000178 time 0.2761 (0.2842) loss 2.0446 (3.2594) grad_norm 2.4619 (nan) [2021-04-16 12:58:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][1190/1251] eta 0:00:17 lr 0.000178 time 0.2749 (0.2842) loss 3.3048 (3.2611) grad_norm 2.0683 (nan) [2021-04-16 12:58:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][1200/1251] eta 0:00:14 lr 0.000178 time 0.2559 (0.2841) loss 3.8367 (3.2634) grad_norm 2.6918 (nan) [2021-04-16 12:58:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][1210/1251] eta 0:00:11 lr 0.000178 time 0.2544 (0.2841) loss 3.3042 (3.2647) grad_norm 1.9898 (nan) [2021-04-16 12:58:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][1220/1251] eta 0:00:08 lr 0.000178 time 0.2927 (0.2840) loss 3.8806 (3.2652) grad_norm 2.2631 (nan) [2021-04-16 12:58:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][1230/1251] eta 0:00:05 lr 0.000178 time 0.2595 (0.2839) loss 3.0161 (3.2645) grad_norm 2.0257 (nan) [2021-04-16 12:58:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][1240/1251] eta 0:00:03 lr 0.000178 time 0.2488 (0.2838) loss 3.4186 (3.2637) grad_norm 2.1695 (nan) [2021-04-16 12:58:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [218/300][1250/1251] eta 0:00:00 lr 0.000178 time 0.2486 (0.2835) loss 2.8333 (3.2625) grad_norm 2.1201 (nan) [2021-04-16 12:58:27 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 218 training takes 0:05:58 [2021-04-16 12:58:27 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_218.pth saving...... [2021-04-16 12:58:37 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_218.pth saved !!! [2021-04-16 12:58:38 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.175 (1.175) Loss 0.9356 (0.9356) Acc@1 78.906 (78.906) Acc@5 94.531 (94.531) [2021-04-16 12:58:40 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.099 (0.227) Loss 0.8410 (0.8906) Acc@1 80.078 (79.288) Acc@5 95.996 (95.002) [2021-04-16 12:58:42 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.144 (0.211) Loss 0.9632 (0.8963) Acc@1 76.562 (79.032) Acc@5 94.043 (94.685) [2021-04-16 12:58:45 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.145 (0.240) Loss 0.8285 (0.8952) Acc@1 80.957 (79.086) Acc@5 95.605 (94.604) [2021-04-16 12:58:46 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.224) Loss 0.8973 (0.8846) Acc@1 78.027 (79.259) Acc@5 94.531 (94.769) [2021-04-16 12:59:02 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 79.142 Acc@5 94.724 [2021-04-16 12:59:02 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 79.1% [2021-04-16 12:59:02 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 79.17% [2021-04-16 12:59:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][0/1251] eta 1:55:02 lr 0.000178 time 5.5177 (5.5177) loss 2.6038 (2.6038) grad_norm 2.0313 (2.0313) [2021-04-16 12:59:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][10/1251] eta 0:15:36 lr 0.000178 time 0.3027 (0.7545) loss 3.4239 (3.3169) grad_norm 2.0329 (2.1117) [2021-04-16 12:59:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][20/1251] eta 0:10:49 lr 0.000178 time 0.2730 (0.5274) loss 3.3998 (3.1828) grad_norm 2.0727 (2.1986) [2021-04-16 12:59:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][30/1251] eta 0:09:05 lr 0.000178 time 0.2961 (0.4465) loss 3.4676 (3.2211) grad_norm 2.2646 (2.1871) [2021-04-16 12:59:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][40/1251] eta 0:08:12 lr 0.000178 time 0.2823 (0.4067) loss 3.3700 (3.2349) grad_norm 2.4643 (2.2073) [2021-04-16 12:59:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][50/1251] eta 0:07:38 lr 0.000177 time 0.2871 (0.3814) loss 4.1317 (3.2846) grad_norm 2.0169 (2.2144) [2021-04-16 12:59:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][60/1251] eta 0:07:13 lr 0.000177 time 0.2486 (0.3638) loss 3.2744 (3.2671) grad_norm 2.2661 (2.2206) [2021-04-16 12:59:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][70/1251] eta 0:06:54 lr 0.000177 time 0.2526 (0.3513) loss 3.2981 (3.2761) grad_norm 2.1527 (2.2102) [2021-04-16 12:59:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][80/1251] eta 0:06:41 lr 0.000177 time 0.2469 (0.3427) loss 2.3863 (3.2672) grad_norm 2.1486 (2.2083) [2021-04-16 12:59:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][90/1251] eta 0:06:29 lr 0.000177 time 0.2804 (0.3356) loss 3.5899 (3.2946) grad_norm 2.5600 (2.2170) [2021-04-16 12:59:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][100/1251] eta 0:06:19 lr 0.000177 time 0.2648 (0.3297) loss 3.2907 (3.3135) grad_norm 2.4834 (2.2214) [2021-04-16 12:59:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][110/1251] eta 0:06:10 lr 0.000177 time 0.2852 (0.3251) loss 3.1998 (3.3127) grad_norm 2.9157 (2.2430) [2021-04-16 12:59:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][120/1251] eta 0:06:02 lr 0.000177 time 0.2720 (0.3209) loss 2.9331 (3.2958) grad_norm 2.3795 (2.2475) [2021-04-16 12:59:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][130/1251] eta 0:05:57 lr 0.000177 time 0.2650 (0.3187) loss 3.3653 (3.2801) grad_norm 1.8956 (2.2505) [2021-04-16 12:59:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][140/1251] eta 0:05:53 lr 0.000177 time 0.2634 (0.3181) loss 3.4502 (3.2767) grad_norm 1.9068 (2.2387) [2021-04-16 12:59:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][150/1251] eta 0:05:48 lr 0.000177 time 0.2927 (0.3163) loss 3.1667 (3.2757) grad_norm 2.1081 (2.2359) [2021-04-16 12:59:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][160/1251] eta 0:05:42 lr 0.000177 time 0.2739 (0.3138) loss 3.2816 (3.2670) grad_norm 2.2529 (2.2383) [2021-04-16 12:59:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][170/1251] eta 0:05:37 lr 0.000177 time 0.2821 (0.3121) loss 3.8182 (3.2735) grad_norm 2.0268 (2.2384) [2021-04-16 12:59:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][180/1251] eta 0:05:31 lr 0.000177 time 0.2849 (0.3100) loss 3.7745 (3.2530) grad_norm 1.9594 (2.2442) [2021-04-16 13:00:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][190/1251] eta 0:05:27 lr 0.000177 time 0.2807 (0.3083) loss 3.7790 (3.2464) grad_norm 2.2087 (2.2608) [2021-04-16 13:00:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][200/1251] eta 0:05:22 lr 0.000177 time 0.2830 (0.3066) loss 3.6126 (3.2660) grad_norm 2.4783 (2.2623) [2021-04-16 13:00:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][210/1251] eta 0:05:17 lr 0.000177 time 0.2774 (0.3052) loss 3.3058 (3.2729) grad_norm 2.8312 (2.2679) [2021-04-16 13:00:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][220/1251] eta 0:05:13 lr 0.000177 time 0.2997 (0.3041) loss 2.4393 (3.2705) grad_norm 3.3124 (2.2743) [2021-04-16 13:00:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][230/1251] eta 0:05:09 lr 0.000177 time 0.2865 (0.3031) loss 2.9339 (3.2723) grad_norm 2.1124 (2.2700) [2021-04-16 13:00:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][240/1251] eta 0:05:05 lr 0.000177 time 0.3034 (0.3020) loss 3.5014 (3.2713) grad_norm 2.4014 (2.2748) [2021-04-16 13:00:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][250/1251] eta 0:05:01 lr 0.000177 time 0.2748 (0.3009) loss 3.7407 (3.2799) grad_norm 2.2289 (2.2701) [2021-04-16 13:00:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][260/1251] eta 0:04:57 lr 0.000177 time 0.2700 (0.3001) loss 3.4875 (3.2825) grad_norm 2.0251 (2.2657) [2021-04-16 13:00:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][270/1251] eta 0:04:53 lr 0.000177 time 0.2521 (0.2990) loss 3.6401 (3.2939) grad_norm 2.0328 (2.2689) [2021-04-16 13:00:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][280/1251] eta 0:04:49 lr 0.000177 time 0.2820 (0.2983) loss 3.5697 (3.2883) grad_norm 2.5987 (2.2672) [2021-04-16 13:00:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][290/1251] eta 0:04:46 lr 0.000177 time 0.2801 (0.2980) loss 2.6007 (3.2747) grad_norm 2.3233 (2.2706) [2021-04-16 13:00:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][300/1251] eta 0:04:43 lr 0.000177 time 0.2500 (0.2978) loss 2.4141 (3.2696) grad_norm 2.0047 (2.2702) [2021-04-16 13:00:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][310/1251] eta 0:04:39 lr 0.000177 time 0.2730 (0.2969) loss 3.6503 (3.2681) grad_norm 2.3171 (2.2717) [2021-04-16 13:00:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][320/1251] eta 0:04:36 lr 0.000177 time 0.2699 (0.2967) loss 3.0129 (3.2735) grad_norm 2.0329 (2.2709) [2021-04-16 13:00:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][330/1251] eta 0:04:32 lr 0.000177 time 0.2923 (0.2961) loss 2.4421 (3.2714) grad_norm 1.9817 (2.2659) [2021-04-16 13:00:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][340/1251] eta 0:04:29 lr 0.000177 time 0.2858 (0.2955) loss 3.0224 (3.2767) grad_norm 1.9226 (2.2630) [2021-04-16 13:00:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][350/1251] eta 0:04:25 lr 0.000177 time 0.2872 (0.2950) loss 2.6008 (3.2765) grad_norm 2.2960 (2.2616) [2021-04-16 13:00:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][360/1251] eta 0:04:22 lr 0.000177 time 0.2651 (0.2948) loss 4.1790 (3.2803) grad_norm 4.7919 (2.2672) [2021-04-16 13:00:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][370/1251] eta 0:04:19 lr 0.000177 time 0.2753 (0.2947) loss 3.6542 (3.2862) grad_norm 2.2198 (2.2686) [2021-04-16 13:00:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][380/1251] eta 0:04:16 lr 0.000176 time 0.2796 (0.2941) loss 3.5083 (3.2868) grad_norm 2.3684 (2.2668) [2021-04-16 13:00:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][390/1251] eta 0:04:12 lr 0.000176 time 0.2688 (0.2936) loss 3.6401 (3.2861) grad_norm 2.0244 (2.2631) [2021-04-16 13:00:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][400/1251] eta 0:04:09 lr 0.000176 time 0.2819 (0.2933) loss 2.8127 (3.2828) grad_norm 2.0715 (2.2649) [2021-04-16 13:01:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][410/1251] eta 0:04:06 lr 0.000176 time 0.2848 (0.2931) loss 3.7061 (3.2761) grad_norm 2.2076 (2.2646) [2021-04-16 13:01:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][420/1251] eta 0:04:03 lr 0.000176 time 0.2842 (0.2927) loss 3.4882 (3.2743) grad_norm 2.3257 (2.2642) [2021-04-16 13:01:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][430/1251] eta 0:03:59 lr 0.000176 time 0.2585 (0.2923) loss 3.8746 (3.2738) grad_norm 2.1451 (2.2642) [2021-04-16 13:01:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][440/1251] eta 0:03:56 lr 0.000176 time 0.2920 (0.2921) loss 3.3542 (3.2740) grad_norm 2.2891 (2.2639) [2021-04-16 13:01:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][450/1251] eta 0:03:53 lr 0.000176 time 0.2547 (0.2916) loss 3.7085 (3.2680) grad_norm 2.0667 (2.2634) [2021-04-16 13:01:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][460/1251] eta 0:03:50 lr 0.000176 time 0.2758 (0.2914) loss 3.2843 (3.2702) grad_norm 2.1289 (2.2640) [2021-04-16 13:01:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][470/1251] eta 0:03:47 lr 0.000176 time 0.2720 (0.2913) loss 3.7375 (3.2700) grad_norm 2.2969 (2.2635) [2021-04-16 13:01:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][480/1251] eta 0:03:44 lr 0.000176 time 0.2692 (0.2909) loss 3.7505 (3.2722) grad_norm 2.1227 (2.2632) [2021-04-16 13:01:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][490/1251] eta 0:03:41 lr 0.000176 time 0.2628 (0.2905) loss 2.6028 (3.2706) grad_norm 2.2001 (2.2630) [2021-04-16 13:01:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][500/1251] eta 0:03:38 lr 0.000176 time 0.2815 (0.2903) loss 3.1269 (3.2675) grad_norm 2.0469 (2.2629) [2021-04-16 13:01:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][510/1251] eta 0:03:34 lr 0.000176 time 0.2700 (0.2900) loss 3.3962 (3.2661) grad_norm 2.8846 (2.2714) [2021-04-16 13:01:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][520/1251] eta 0:03:32 lr 0.000176 time 0.2879 (0.2901) loss 3.4622 (3.2670) grad_norm 2.5835 (2.2719) [2021-04-16 13:01:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][530/1251] eta 0:03:28 lr 0.000176 time 0.2643 (0.2898) loss 3.1064 (3.2672) grad_norm 2.2408 (2.2705) [2021-04-16 13:01:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][540/1251] eta 0:03:26 lr 0.000176 time 0.2781 (0.2898) loss 2.5818 (3.2661) grad_norm 2.0113 (2.2703) [2021-04-16 13:01:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][550/1251] eta 0:03:22 lr 0.000176 time 0.2572 (0.2895) loss 4.1852 (3.2686) grad_norm 2.1105 (2.2689) [2021-04-16 13:01:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][560/1251] eta 0:03:19 lr 0.000176 time 0.3125 (0.2894) loss 3.6461 (3.2782) grad_norm 2.0054 (2.2696) [2021-04-16 13:01:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][570/1251] eta 0:03:16 lr 0.000176 time 0.2751 (0.2891) loss 2.9745 (3.2796) grad_norm 2.2525 (2.2708) [2021-04-16 13:01:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][580/1251] eta 0:03:14 lr 0.000176 time 0.2715 (0.2894) loss 3.3912 (3.2772) grad_norm 2.2534 (2.2700) [2021-04-16 13:01:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][590/1251] eta 0:03:11 lr 0.000176 time 0.3106 (0.2892) loss 2.9499 (3.2741) grad_norm 2.0123 (2.2673) [2021-04-16 13:01:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][600/1251] eta 0:03:08 lr 0.000176 time 0.2642 (0.2889) loss 2.6318 (3.2780) grad_norm 2.6848 (2.2662) [2021-04-16 13:01:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][610/1251] eta 0:03:05 lr 0.000176 time 0.2863 (0.2887) loss 2.9729 (3.2755) grad_norm 2.4248 (2.2665) [2021-04-16 13:02:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][620/1251] eta 0:03:02 lr 0.000176 time 0.2785 (0.2885) loss 3.4298 (3.2720) grad_norm 2.1175 (2.2653) [2021-04-16 13:02:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][630/1251] eta 0:02:59 lr 0.000176 time 0.2959 (0.2883) loss 4.0039 (3.2732) grad_norm 2.1946 (2.2655) [2021-04-16 13:02:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][640/1251] eta 0:02:56 lr 0.000176 time 0.2847 (0.2881) loss 3.1551 (3.2748) grad_norm 2.1252 (2.2643) [2021-04-16 13:02:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][650/1251] eta 0:02:52 lr 0.000176 time 0.2455 (0.2878) loss 3.3649 (3.2663) grad_norm 2.2974 (2.2638) [2021-04-16 13:02:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][660/1251] eta 0:02:50 lr 0.000176 time 0.2815 (0.2877) loss 3.2764 (3.2667) grad_norm 2.0529 (2.2643) [2021-04-16 13:02:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][670/1251] eta 0:02:47 lr 0.000176 time 0.2655 (0.2874) loss 2.6505 (3.2656) grad_norm 2.6346 (2.2666) [2021-04-16 13:02:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][680/1251] eta 0:02:44 lr 0.000176 time 0.2713 (0.2873) loss 2.8191 (3.2688) grad_norm 2.1839 (2.2665) [2021-04-16 13:02:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][690/1251] eta 0:02:41 lr 0.000176 time 0.2717 (0.2871) loss 3.2855 (3.2689) grad_norm 2.1841 (2.2669) [2021-04-16 13:02:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][700/1251] eta 0:02:38 lr 0.000175 time 0.2652 (0.2869) loss 3.5905 (3.2710) grad_norm 1.9688 (2.2665) [2021-04-16 13:02:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][710/1251] eta 0:02:35 lr 0.000175 time 0.2734 (0.2869) loss 3.7662 (3.2738) grad_norm 1.9517 (2.2650) [2021-04-16 13:02:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][720/1251] eta 0:02:32 lr 0.000175 time 0.2795 (0.2873) loss 3.4985 (3.2720) grad_norm 1.9804 (2.2629) [2021-04-16 13:02:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][730/1251] eta 0:02:29 lr 0.000175 time 0.2661 (0.2871) loss 3.0196 (3.2691) grad_norm 1.8621 (2.2618) [2021-04-16 13:02:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][740/1251] eta 0:02:26 lr 0.000175 time 0.2863 (0.2870) loss 3.2356 (3.2703) grad_norm 2.2515 (2.2613) [2021-04-16 13:02:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][750/1251] eta 0:02:23 lr 0.000175 time 0.2616 (0.2869) loss 2.7007 (3.2709) grad_norm 2.1719 (2.2648) [2021-04-16 13:02:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][760/1251] eta 0:02:20 lr 0.000175 time 0.2856 (0.2869) loss 3.5913 (3.2728) grad_norm 2.0504 (2.2648) [2021-04-16 13:02:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][770/1251] eta 0:02:17 lr 0.000175 time 0.2701 (0.2867) loss 3.4047 (3.2713) grad_norm 2.0601 (2.2638) [2021-04-16 13:02:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][780/1251] eta 0:02:15 lr 0.000175 time 0.2927 (0.2867) loss 3.7367 (3.2724) grad_norm 2.0158 (2.2637) [2021-04-16 13:02:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][790/1251] eta 0:02:12 lr 0.000175 time 0.2631 (0.2865) loss 3.4659 (3.2733) grad_norm 2.3380 (2.2657) [2021-04-16 13:02:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][800/1251] eta 0:02:09 lr 0.000175 time 0.2786 (0.2864) loss 2.9172 (3.2721) grad_norm 2.3126 (2.2667) [2021-04-16 13:02:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][810/1251] eta 0:02:06 lr 0.000175 time 0.2636 (0.2863) loss 3.7770 (3.2702) grad_norm 2.6438 (2.2668) [2021-04-16 13:02:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][820/1251] eta 0:02:03 lr 0.000175 time 0.2862 (0.2862) loss 2.7061 (3.2704) grad_norm 2.0944 (2.2651) [2021-04-16 13:03:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][830/1251] eta 0:02:00 lr 0.000175 time 0.2512 (0.2861) loss 3.3738 (3.2743) grad_norm 2.9964 (2.2664) [2021-04-16 13:03:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][840/1251] eta 0:01:57 lr 0.000175 time 0.2742 (0.2860) loss 4.0796 (3.2747) grad_norm 2.3732 (2.2667) [2021-04-16 13:03:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][850/1251] eta 0:01:54 lr 0.000175 time 0.2978 (0.2859) loss 2.2502 (3.2724) grad_norm 2.1765 (2.2665) [2021-04-16 13:03:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][860/1251] eta 0:01:51 lr 0.000175 time 0.2938 (0.2858) loss 3.2814 (3.2723) grad_norm 2.3116 (2.2670) [2021-04-16 13:03:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][870/1251] eta 0:01:48 lr 0.000175 time 0.2726 (0.2858) loss 3.0754 (3.2700) grad_norm 2.2367 (2.2659) [2021-04-16 13:03:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][880/1251] eta 0:01:46 lr 0.000175 time 0.2739 (0.2859) loss 2.5357 (3.2679) grad_norm 2.4412 (2.2654) [2021-04-16 13:03:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][890/1251] eta 0:01:43 lr 0.000175 time 0.2595 (0.2859) loss 2.9634 (3.2680) grad_norm 2.5942 (2.2663) [2021-04-16 13:03:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][900/1251] eta 0:01:40 lr 0.000175 time 0.2581 (0.2858) loss 2.9102 (3.2648) grad_norm 2.6444 (2.2655) [2021-04-16 13:03:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][910/1251] eta 0:01:37 lr 0.000175 time 0.2844 (0.2857) loss 2.6573 (3.2631) grad_norm 2.1247 (2.2685) [2021-04-16 13:03:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][920/1251] eta 0:01:34 lr 0.000175 time 0.2691 (0.2856) loss 3.7028 (3.2639) grad_norm 2.0253 (2.2692) [2021-04-16 13:03:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][930/1251] eta 0:01:31 lr 0.000175 time 0.2850 (0.2855) loss 3.8065 (3.2655) grad_norm 2.3383 (2.2704) [2021-04-16 13:03:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][940/1251] eta 0:01:28 lr 0.000175 time 0.2829 (0.2856) loss 3.6048 (3.2658) grad_norm 2.1482 (2.2696) [2021-04-16 13:03:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][950/1251] eta 0:01:25 lr 0.000175 time 0.2430 (0.2855) loss 3.2340 (3.2649) grad_norm 2.4383 (2.2709) [2021-04-16 13:03:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][960/1251] eta 0:01:23 lr 0.000175 time 0.3153 (0.2855) loss 3.5223 (3.2642) grad_norm 2.7777 (2.2722) [2021-04-16 13:03:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][970/1251] eta 0:01:20 lr 0.000175 time 0.2879 (0.2854) loss 2.0241 (3.2629) grad_norm 2.2370 (2.2712) [2021-04-16 13:03:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][980/1251] eta 0:01:17 lr 0.000175 time 0.2771 (0.2852) loss 2.4379 (3.2628) grad_norm 2.3015 (2.2706) [2021-04-16 13:03:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][990/1251] eta 0:01:14 lr 0.000175 time 0.2821 (0.2851) loss 3.0205 (3.2622) grad_norm 2.4746 (2.2715) [2021-04-16 13:03:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][1000/1251] eta 0:01:11 lr 0.000175 time 0.2792 (0.2850) loss 2.8678 (3.2616) grad_norm 2.0880 (2.2714) [2021-04-16 13:03:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][1010/1251] eta 0:01:08 lr 0.000175 time 0.2768 (0.2849) loss 3.3110 (3.2640) grad_norm 2.7481 (2.2710) [2021-04-16 13:03:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][1020/1251] eta 0:01:05 lr 0.000174 time 0.2749 (0.2848) loss 2.2750 (3.2636) grad_norm 1.9249 (2.2722) [2021-04-16 13:03:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][1030/1251] eta 0:01:02 lr 0.000174 time 0.2995 (0.2847) loss 3.3705 (3.2645) grad_norm 2.3457 (2.2737) [2021-04-16 13:03:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][1040/1251] eta 0:01:00 lr 0.000174 time 0.2905 (0.2848) loss 3.5485 (3.2641) grad_norm 2.5163 (2.2739) [2021-04-16 13:04:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][1050/1251] eta 0:00:57 lr 0.000174 time 0.2655 (0.2847) loss 3.4215 (3.2645) grad_norm 2.2838 (2.2747) [2021-04-16 13:04:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][1060/1251] eta 0:00:54 lr 0.000174 time 0.2621 (0.2846) loss 3.3843 (3.2634) grad_norm 2.4452 (2.2745) [2021-04-16 13:04:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][1070/1251] eta 0:00:51 lr 0.000174 time 0.2529 (0.2845) loss 3.7767 (3.2640) grad_norm 2.7837 (2.2752) [2021-04-16 13:04:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][1080/1251] eta 0:00:48 lr 0.000174 time 0.2647 (0.2844) loss 3.0250 (3.2652) grad_norm 2.1451 (2.2759) [2021-04-16 13:04:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][1090/1251] eta 0:00:45 lr 0.000174 time 0.2762 (0.2845) loss 2.9067 (3.2647) grad_norm 1.9886 (2.2765) [2021-04-16 13:04:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][1100/1251] eta 0:00:42 lr 0.000174 time 0.2645 (0.2844) loss 2.8595 (3.2638) grad_norm 2.1199 (2.2766) [2021-04-16 13:04:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][1110/1251] eta 0:00:40 lr 0.000174 time 0.2593 (0.2844) loss 3.4254 (3.2645) grad_norm 2.0695 (2.2767) [2021-04-16 13:04:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][1120/1251] eta 0:00:37 lr 0.000174 time 0.2897 (0.2844) loss 2.9035 (3.2629) grad_norm 2.1465 (2.2769) [2021-04-16 13:04:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][1130/1251] eta 0:00:34 lr 0.000174 time 0.2893 (0.2844) loss 2.1637 (3.2636) grad_norm 2.6381 (2.2778) [2021-04-16 13:04:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][1140/1251] eta 0:00:31 lr 0.000174 time 0.2700 (0.2844) loss 3.2614 (3.2626) grad_norm 2.2324 (2.2763) [2021-04-16 13:04:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][1150/1251] eta 0:00:28 lr 0.000174 time 0.2741 (0.2845) loss 3.6544 (3.2642) grad_norm 2.5802 (2.2775) [2021-04-16 13:04:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][1160/1251] eta 0:00:25 lr 0.000174 time 0.2698 (0.2846) loss 3.2092 (3.2629) grad_norm 2.2514 (2.2777) [2021-04-16 13:04:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][1170/1251] eta 0:00:23 lr 0.000174 time 0.2800 (0.2845) loss 3.3517 (3.2626) grad_norm 2.1061 (2.2768) [2021-04-16 13:04:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][1180/1251] eta 0:00:20 lr 0.000174 time 0.2762 (0.2845) loss 3.2710 (3.2620) grad_norm 2.3922 (2.2760) [2021-04-16 13:04:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][1190/1251] eta 0:00:17 lr 0.000174 time 0.2718 (0.2844) loss 3.5141 (3.2603) grad_norm 2.1770 (2.2761) [2021-04-16 13:04:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][1200/1251] eta 0:00:14 lr 0.000174 time 0.2760 (0.2844) loss 3.1911 (3.2622) grad_norm 1.9313 (2.2760) [2021-04-16 13:04:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][1210/1251] eta 0:00:11 lr 0.000174 time 0.2735 (0.2843) loss 3.0332 (3.2626) grad_norm 2.5951 (2.2757) [2021-04-16 13:04:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][1220/1251] eta 0:00:08 lr 0.000174 time 0.2503 (0.2842) loss 3.5853 (3.2640) grad_norm 2.7728 (2.2767) [2021-04-16 13:04:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][1230/1251] eta 0:00:05 lr 0.000174 time 0.2708 (0.2842) loss 3.4076 (3.2637) grad_norm 2.1175 (2.2755) [2021-04-16 13:04:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][1240/1251] eta 0:00:03 lr 0.000174 time 0.2614 (0.2840) loss 2.9493 (3.2623) grad_norm 1.9135 (2.2766) [2021-04-16 13:04:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [219/300][1250/1251] eta 0:00:00 lr 0.000174 time 0.2486 (0.2837) loss 2.8107 (3.2624) grad_norm 2.0344 (2.2771) [2021-04-16 13:05:00 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 219 training takes 0:05:58 [2021-04-16 13:05:00 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_219.pth saving...... [2021-04-16 13:05:12 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_219.pth saved !!! [2021-04-16 13:05:13 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.170 (1.170) Loss 0.8529 (0.8529) Acc@1 80.469 (80.469) Acc@5 95.215 (95.215) [2021-04-16 13:05:14 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.342 (0.234) Loss 0.9375 (0.8882) Acc@1 77.344 (78.844) Acc@5 94.629 (94.700) [2021-04-16 13:05:16 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.155 (0.198) Loss 0.8415 (0.8955) Acc@1 80.273 (78.906) Acc@5 95.215 (94.527) [2021-04-16 13:05:19 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.661 (0.245) Loss 0.8482 (0.8895) Acc@1 79.785 (79.010) Acc@5 95.312 (94.623) [2021-04-16 13:05:21 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.129 (0.216) Loss 0.8785 (0.8894) Acc@1 79.102 (79.030) Acc@5 94.336 (94.648) [2021-04-16 13:05:27 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 79.006 Acc@5 94.688 [2021-04-16 13:05:27 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 79.0% [2021-04-16 13:05:27 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 79.17% [2021-04-16 13:05:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][0/1251] eta 2:48:00 lr 0.000174 time 8.0576 (8.0576) loss 3.4707 (3.4707) grad_norm 2.3757 (2.3757) [2021-04-16 13:05:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][10/1251] eta 0:20:19 lr 0.000174 time 0.2954 (0.9828) loss 3.2541 (3.3391) grad_norm 2.0109 (2.1561) [2021-04-16 13:05:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][20/1251] eta 0:13:24 lr 0.000174 time 0.3116 (0.6533) loss 3.2987 (3.2209) grad_norm 2.3595 (2.1928) [2021-04-16 13:05:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][30/1251] eta 0:10:52 lr 0.000174 time 0.2867 (0.5343) loss 4.1380 (3.3050) grad_norm 2.4297 (2.2229) [2021-04-16 13:05:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][40/1251] eta 0:09:30 lr 0.000174 time 0.2800 (0.4715) loss 2.6825 (3.2970) grad_norm 2.0700 (2.2289) [2021-04-16 13:05:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][50/1251] eta 0:08:42 lr 0.000174 time 0.2972 (0.4351) loss 3.2823 (3.2332) grad_norm 1.9173 (2.2286) [2021-04-16 13:05:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][60/1251] eta 0:08:10 lr 0.000174 time 0.2996 (0.4114) loss 4.1131 (3.2597) grad_norm 2.2055 (2.2339) [2021-04-16 13:05:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][70/1251] eta 0:07:44 lr 0.000174 time 0.3117 (0.3935) loss 3.6644 (3.2804) grad_norm 2.3125 (2.2334) [2021-04-16 13:05:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][80/1251] eta 0:07:26 lr 0.000174 time 0.2862 (0.3810) loss 3.5589 (3.3112) grad_norm 2.0129 (2.2398) [2021-04-16 13:06:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][90/1251] eta 0:07:09 lr 0.000174 time 0.2719 (0.3700) loss 3.3031 (3.3313) grad_norm 2.0213 (2.2457) [2021-04-16 13:06:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][100/1251] eta 0:06:56 lr 0.000173 time 0.2666 (0.3623) loss 2.8502 (3.3343) grad_norm 2.1833 (2.2394) [2021-04-16 13:06:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][110/1251] eta 0:06:44 lr 0.000173 time 0.2691 (0.3547) loss 3.3815 (3.3210) grad_norm 2.1211 (2.2355) [2021-04-16 13:06:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][120/1251] eta 0:06:33 lr 0.000173 time 0.2649 (0.3483) loss 2.3185 (3.3173) grad_norm 1.9909 (2.2381) [2021-04-16 13:06:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][130/1251] eta 0:06:24 lr 0.000173 time 0.2603 (0.3432) loss 2.5981 (3.2964) grad_norm 1.9969 (2.2416) [2021-04-16 13:06:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][140/1251] eta 0:06:16 lr 0.000173 time 0.2754 (0.3393) loss 3.7118 (3.2934) grad_norm 2.0424 (2.2334) [2021-04-16 13:06:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][150/1251] eta 0:06:10 lr 0.000173 time 0.2957 (0.3361) loss 3.3996 (3.2674) grad_norm 2.0994 (2.2273) [2021-04-16 13:06:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][160/1251] eta 0:06:02 lr 0.000173 time 0.2858 (0.3326) loss 3.0304 (3.2757) grad_norm 2.4654 (2.2252) [2021-04-16 13:06:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][170/1251] eta 0:05:56 lr 0.000173 time 0.2663 (0.3296) loss 3.3609 (3.2777) grad_norm 2.3020 (2.2387) [2021-04-16 13:06:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][180/1251] eta 0:05:50 lr 0.000173 time 0.2703 (0.3269) loss 3.3704 (3.2686) grad_norm 1.9945 (2.2424) [2021-04-16 13:06:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][190/1251] eta 0:05:44 lr 0.000173 time 0.2859 (0.3249) loss 3.6142 (3.2743) grad_norm 2.2052 (2.2677) [2021-04-16 13:06:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][200/1251] eta 0:05:38 lr 0.000173 time 0.2734 (0.3224) loss 3.8737 (3.2803) grad_norm 2.1690 (2.2672) [2021-04-16 13:06:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][210/1251] eta 0:05:33 lr 0.000173 time 0.2795 (0.3206) loss 3.4163 (3.2692) grad_norm 2.6484 (2.2702) [2021-04-16 13:06:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][220/1251] eta 0:05:29 lr 0.000173 time 0.3008 (0.3194) loss 3.7668 (3.2582) grad_norm 2.4592 (2.2812) [2021-04-16 13:06:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][230/1251] eta 0:05:24 lr 0.000173 time 0.2717 (0.3177) loss 2.4925 (3.2458) grad_norm 2.0270 (2.2811) [2021-04-16 13:06:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][240/1251] eta 0:05:19 lr 0.000173 time 0.2918 (0.3160) loss 3.0254 (3.2433) grad_norm 2.0766 (2.2796) [2021-04-16 13:06:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][250/1251] eta 0:05:14 lr 0.000173 time 0.2627 (0.3147) loss 3.3961 (3.2305) grad_norm 2.1528 (2.2861) [2021-04-16 13:06:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][260/1251] eta 0:05:10 lr 0.000173 time 0.3125 (0.3133) loss 2.5437 (3.2296) grad_norm 2.2654 (2.2878) [2021-04-16 13:06:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][270/1251] eta 0:05:06 lr 0.000173 time 0.2908 (0.3122) loss 3.6556 (3.2345) grad_norm 2.1395 (2.2831) [2021-04-16 13:06:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][280/1251] eta 0:05:02 lr 0.000173 time 0.2785 (0.3111) loss 3.5213 (3.2359) grad_norm 2.0676 (2.2796) [2021-04-16 13:06:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][290/1251] eta 0:04:57 lr 0.000173 time 0.2851 (0.3099) loss 3.2670 (3.2407) grad_norm 2.0535 (2.2776) [2021-04-16 13:07:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][300/1251] eta 0:04:54 lr 0.000173 time 0.2502 (0.3098) loss 3.6858 (3.2416) grad_norm 2.6848 (2.2772) [2021-04-16 13:07:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][310/1251] eta 0:04:50 lr 0.000173 time 0.3024 (0.3087) loss 3.4484 (3.2441) grad_norm 2.4467 (2.2770) [2021-04-16 13:07:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][320/1251] eta 0:04:46 lr 0.000173 time 0.2736 (0.3080) loss 3.1834 (3.2480) grad_norm 2.1328 (2.2786) [2021-04-16 13:07:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][330/1251] eta 0:04:42 lr 0.000173 time 0.3082 (0.3072) loss 3.2641 (3.2482) grad_norm 2.4256 (2.2767) [2021-04-16 13:07:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][340/1251] eta 0:04:38 lr 0.000173 time 0.2833 (0.3062) loss 2.4698 (3.2418) grad_norm 2.4342 (2.2747) [2021-04-16 13:07:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][350/1251] eta 0:04:35 lr 0.000173 time 0.4544 (0.3060) loss 3.2069 (3.2383) grad_norm 2.3892 (2.2752) [2021-04-16 13:07:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][360/1251] eta 0:04:32 lr 0.000173 time 0.2419 (0.3056) loss 2.7705 (3.2383) grad_norm 2.7595 (2.2762) [2021-04-16 13:07:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][370/1251] eta 0:04:28 lr 0.000173 time 0.2687 (0.3048) loss 2.9871 (3.2411) grad_norm 2.1557 (inf) [2021-04-16 13:07:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][380/1251] eta 0:04:24 lr 0.000173 time 0.2818 (0.3041) loss 2.5079 (3.2276) grad_norm 2.6550 (inf) [2021-04-16 13:07:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][390/1251] eta 0:04:21 lr 0.000173 time 0.2844 (0.3037) loss 3.1177 (3.2300) grad_norm 2.1559 (inf) [2021-04-16 13:07:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][400/1251] eta 0:04:17 lr 0.000173 time 0.2600 (0.3029) loss 3.4145 (3.2302) grad_norm 2.4066 (inf) [2021-04-16 13:07:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][410/1251] eta 0:04:14 lr 0.000173 time 0.2798 (0.3024) loss 3.5977 (3.2344) grad_norm 1.9848 (inf) [2021-04-16 13:07:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][420/1251] eta 0:04:10 lr 0.000172 time 0.2955 (0.3020) loss 3.6897 (3.2390) grad_norm 2.1400 (inf) [2021-04-16 13:07:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][430/1251] eta 0:04:07 lr 0.000172 time 0.2804 (0.3014) loss 3.8140 (3.2453) grad_norm 2.5377 (inf) [2021-04-16 13:07:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][440/1251] eta 0:04:04 lr 0.000172 time 0.2800 (0.3009) loss 3.5764 (3.2407) grad_norm 2.0928 (inf) [2021-04-16 13:07:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][450/1251] eta 0:04:00 lr 0.000172 time 0.2784 (0.3004) loss 3.5768 (3.2467) grad_norm 1.9223 (inf) [2021-04-16 13:07:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][460/1251] eta 0:03:57 lr 0.000172 time 0.2509 (0.2999) loss 3.0799 (3.2464) grad_norm 2.4195 (inf) [2021-04-16 13:07:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][470/1251] eta 0:03:53 lr 0.000172 time 0.2563 (0.2995) loss 3.8432 (3.2444) grad_norm 2.3559 (inf) [2021-04-16 13:07:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][480/1251] eta 0:03:50 lr 0.000172 time 0.2773 (0.2992) loss 3.1811 (3.2500) grad_norm 2.4735 (inf) [2021-04-16 13:07:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][490/1251] eta 0:03:47 lr 0.000172 time 0.2901 (0.2989) loss 3.5558 (3.2495) grad_norm 2.1649 (inf) [2021-04-16 13:07:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][500/1251] eta 0:03:44 lr 0.000172 time 0.2750 (0.2987) loss 3.7138 (3.2497) grad_norm 2.4706 (inf) [2021-04-16 13:08:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][510/1251] eta 0:03:41 lr 0.000172 time 0.2878 (0.2983) loss 3.8784 (3.2501) grad_norm 2.3910 (inf) [2021-04-16 13:08:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][520/1251] eta 0:03:37 lr 0.000172 time 0.2851 (0.2980) loss 2.5645 (3.2484) grad_norm 1.8815 (inf) [2021-04-16 13:08:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][530/1251] eta 0:03:34 lr 0.000172 time 0.2844 (0.2977) loss 3.7556 (3.2491) grad_norm 2.4201 (inf) [2021-04-16 13:08:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][540/1251] eta 0:03:31 lr 0.000172 time 0.2690 (0.2975) loss 3.8875 (3.2479) grad_norm 2.2820 (inf) [2021-04-16 13:08:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][550/1251] eta 0:03:28 lr 0.000172 time 0.2767 (0.2971) loss 3.4456 (3.2493) grad_norm 1.9411 (inf) [2021-04-16 13:08:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][560/1251] eta 0:03:24 lr 0.000172 time 0.2812 (0.2967) loss 3.2128 (3.2522) grad_norm 2.2649 (inf) [2021-04-16 13:08:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][570/1251] eta 0:03:21 lr 0.000172 time 0.2561 (0.2965) loss 3.2715 (3.2504) grad_norm 2.6546 (inf) [2021-04-16 13:08:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][580/1251] eta 0:03:18 lr 0.000172 time 0.2907 (0.2966) loss 2.3387 (3.2499) grad_norm 2.1579 (inf) [2021-04-16 13:08:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][590/1251] eta 0:03:15 lr 0.000172 time 0.2962 (0.2963) loss 3.8438 (3.2498) grad_norm 3.7603 (inf) [2021-04-16 13:08:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][600/1251] eta 0:03:12 lr 0.000172 time 0.2887 (0.2960) loss 3.2506 (3.2518) grad_norm 2.2446 (inf) [2021-04-16 13:08:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][610/1251] eta 0:03:09 lr 0.000172 time 0.3024 (0.2957) loss 3.0703 (3.2527) grad_norm 2.5812 (inf) [2021-04-16 13:08:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][620/1251] eta 0:03:06 lr 0.000172 time 0.2854 (0.2952) loss 4.1288 (3.2558) grad_norm 2.3833 (inf) [2021-04-16 13:08:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][630/1251] eta 0:03:03 lr 0.000172 time 0.2774 (0.2950) loss 3.5801 (3.2501) grad_norm 1.9954 (inf) [2021-04-16 13:08:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][640/1251] eta 0:03:00 lr 0.000172 time 0.2685 (0.2948) loss 2.0974 (3.2496) grad_norm 2.2084 (inf) [2021-04-16 13:08:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][650/1251] eta 0:02:57 lr 0.000172 time 0.2936 (0.2949) loss 3.3820 (3.2437) grad_norm 2.1487 (inf) [2021-04-16 13:08:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][660/1251] eta 0:02:54 lr 0.000172 time 0.2798 (0.2946) loss 3.1245 (3.2414) grad_norm 2.7711 (inf) [2021-04-16 13:08:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][670/1251] eta 0:02:51 lr 0.000172 time 0.2969 (0.2944) loss 3.0549 (3.2405) grad_norm 2.1655 (inf) [2021-04-16 13:08:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][680/1251] eta 0:02:48 lr 0.000172 time 0.2718 (0.2943) loss 3.7550 (3.2424) grad_norm 1.8203 (inf) [2021-04-16 13:08:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][690/1251] eta 0:02:45 lr 0.000172 time 0.2927 (0.2942) loss 2.5252 (3.2381) grad_norm 2.7453 (inf) [2021-04-16 13:08:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][700/1251] eta 0:02:41 lr 0.000172 time 0.2844 (0.2940) loss 3.2830 (3.2400) grad_norm 2.3449 (inf) [2021-04-16 13:08:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][710/1251] eta 0:02:38 lr 0.000172 time 0.2864 (0.2938) loss 3.7049 (3.2371) grad_norm 4.6550 (inf) [2021-04-16 13:08:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][720/1251] eta 0:02:36 lr 0.000172 time 0.2577 (0.2938) loss 3.6464 (3.2375) grad_norm 2.2211 (inf) [2021-04-16 13:09:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][730/1251] eta 0:02:32 lr 0.000172 time 0.2751 (0.2936) loss 3.5083 (3.2390) grad_norm 2.1563 (inf) [2021-04-16 13:09:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][740/1251] eta 0:02:29 lr 0.000172 time 0.2814 (0.2935) loss 3.0652 (3.2388) grad_norm 2.2547 (inf) [2021-04-16 13:09:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][750/1251] eta 0:02:26 lr 0.000171 time 0.3010 (0.2933) loss 3.3464 (3.2374) grad_norm 2.4441 (inf) [2021-04-16 13:09:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][760/1251] eta 0:02:23 lr 0.000171 time 0.2795 (0.2930) loss 3.4317 (3.2388) grad_norm 2.2196 (inf) [2021-04-16 13:09:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][770/1251] eta 0:02:20 lr 0.000171 time 0.2508 (0.2928) loss 3.1861 (3.2381) grad_norm 2.5639 (inf) [2021-04-16 13:09:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][780/1251] eta 0:02:17 lr 0.000171 time 0.2630 (0.2927) loss 3.0038 (3.2396) grad_norm 2.2855 (inf) [2021-04-16 13:09:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][790/1251] eta 0:02:14 lr 0.000171 time 0.2775 (0.2925) loss 3.4778 (3.2399) grad_norm 2.1992 (inf) [2021-04-16 13:09:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][800/1251] eta 0:02:11 lr 0.000171 time 0.2945 (0.2924) loss 2.8949 (3.2373) grad_norm 2.1731 (inf) [2021-04-16 13:09:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][810/1251] eta 0:02:08 lr 0.000171 time 0.2749 (0.2925) loss 3.9156 (3.2394) grad_norm 2.4535 (inf) [2021-04-16 13:09:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][820/1251] eta 0:02:05 lr 0.000171 time 0.2761 (0.2923) loss 3.6490 (3.2394) grad_norm 2.1576 (inf) [2021-04-16 13:09:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][830/1251] eta 0:02:03 lr 0.000171 time 0.2631 (0.2922) loss 3.1524 (3.2382) grad_norm 2.2484 (inf) [2021-04-16 13:09:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][840/1251] eta 0:02:00 lr 0.000171 time 0.2749 (0.2921) loss 2.4844 (3.2387) grad_norm 2.1193 (inf) [2021-04-16 13:09:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][850/1251] eta 0:01:57 lr 0.000171 time 0.3086 (0.2920) loss 2.7028 (3.2389) grad_norm 2.4178 (inf) [2021-04-16 13:09:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][860/1251] eta 0:01:54 lr 0.000171 time 0.2644 (0.2919) loss 3.1210 (3.2394) grad_norm 2.1083 (inf) [2021-04-16 13:09:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][870/1251] eta 0:01:51 lr 0.000171 time 0.2987 (0.2918) loss 3.5896 (3.2379) grad_norm 2.3392 (inf) [2021-04-16 13:09:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][880/1251] eta 0:01:48 lr 0.000171 time 0.2762 (0.2917) loss 3.4027 (3.2361) grad_norm 2.2006 (inf) [2021-04-16 13:09:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][890/1251] eta 0:01:45 lr 0.000171 time 0.2843 (0.2918) loss 3.1113 (3.2341) grad_norm 2.1244 (inf) [2021-04-16 13:09:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][900/1251] eta 0:01:42 lr 0.000171 time 0.2821 (0.2917) loss 3.3750 (3.2314) grad_norm 2.1858 (inf) [2021-04-16 13:09:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][910/1251] eta 0:01:39 lr 0.000171 time 0.2810 (0.2915) loss 3.6950 (3.2322) grad_norm 3.7749 (inf) [2021-04-16 13:09:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][920/1251] eta 0:01:36 lr 0.000171 time 0.2690 (0.2914) loss 3.4106 (3.2328) grad_norm 2.1155 (inf) [2021-04-16 13:09:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][930/1251] eta 0:01:33 lr 0.000171 time 0.2704 (0.2914) loss 3.9553 (3.2339) grad_norm 2.1031 (inf) [2021-04-16 13:10:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][940/1251] eta 0:01:30 lr 0.000171 time 0.2835 (0.2913) loss 3.6149 (3.2351) grad_norm 2.3751 (inf) [2021-04-16 13:10:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][950/1251] eta 0:01:27 lr 0.000171 time 0.3116 (0.2912) loss 3.5858 (3.2333) grad_norm 2.1508 (inf) [2021-04-16 13:10:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][960/1251] eta 0:01:24 lr 0.000171 time 0.2780 (0.2911) loss 3.0538 (3.2345) grad_norm 2.3927 (inf) [2021-04-16 13:10:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][970/1251] eta 0:01:21 lr 0.000171 time 0.3133 (0.2911) loss 4.0968 (3.2376) grad_norm 2.9291 (inf) [2021-04-16 13:10:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][980/1251] eta 0:01:18 lr 0.000171 time 0.2695 (0.2910) loss 3.7433 (3.2383) grad_norm 2.1216 (inf) [2021-04-16 13:10:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][990/1251] eta 0:01:15 lr 0.000171 time 0.2906 (0.2908) loss 4.0874 (3.2362) grad_norm 2.5234 (inf) [2021-04-16 13:10:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][1000/1251] eta 0:01:12 lr 0.000171 time 0.2899 (0.2907) loss 2.2858 (3.2348) grad_norm 2.2046 (inf) [2021-04-16 13:10:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][1010/1251] eta 0:01:10 lr 0.000171 time 0.2978 (0.2906) loss 3.4753 (3.2337) grad_norm 1.9952 (inf) [2021-04-16 13:10:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][1020/1251] eta 0:01:07 lr 0.000171 time 0.2555 (0.2905) loss 2.7142 (3.2335) grad_norm 2.0763 (inf) [2021-04-16 13:10:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][1030/1251] eta 0:01:04 lr 0.000171 time 0.2921 (0.2905) loss 3.5961 (3.2321) grad_norm 2.2517 (inf) [2021-04-16 13:10:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][1040/1251] eta 0:01:01 lr 0.000171 time 0.2772 (0.2904) loss 3.5230 (3.2316) grad_norm 1.8890 (inf) [2021-04-16 13:10:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][1050/1251] eta 0:00:58 lr 0.000171 time 0.2912 (0.2903) loss 3.7769 (3.2331) grad_norm 2.2861 (inf) [2021-04-16 13:10:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][1060/1251] eta 0:00:55 lr 0.000171 time 0.3053 (0.2902) loss 3.7342 (3.2354) grad_norm 2.8418 (inf) [2021-04-16 13:10:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][1070/1251] eta 0:00:52 lr 0.000170 time 0.2761 (0.2901) loss 3.7937 (3.2376) grad_norm 2.6722 (inf) [2021-04-16 13:10:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][1080/1251] eta 0:00:49 lr 0.000170 time 0.2914 (0.2900) loss 2.0081 (3.2362) grad_norm 2.2308 (inf) [2021-04-16 13:10:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][1090/1251] eta 0:00:46 lr 0.000170 time 0.2503 (0.2900) loss 2.3865 (3.2371) grad_norm 2.5173 (inf) [2021-04-16 13:10:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][1100/1251] eta 0:00:43 lr 0.000170 time 0.3123 (0.2900) loss 2.0861 (3.2334) grad_norm 2.7026 (inf) [2021-04-16 13:10:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][1110/1251] eta 0:00:40 lr 0.000170 time 0.2590 (0.2899) loss 2.6096 (3.2333) grad_norm 2.3306 (inf) [2021-04-16 13:10:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][1120/1251] eta 0:00:37 lr 0.000170 time 0.2950 (0.2899) loss 3.1383 (3.2355) grad_norm 2.0411 (inf) [2021-04-16 13:10:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][1130/1251] eta 0:00:35 lr 0.000170 time 0.2751 (0.2898) loss 4.2221 (3.2356) grad_norm 2.5462 (inf) [2021-04-16 13:10:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][1140/1251] eta 0:00:32 lr 0.000170 time 0.2804 (0.2898) loss 3.5721 (3.2352) grad_norm 2.4384 (inf) [2021-04-16 13:11:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][1150/1251] eta 0:00:29 lr 0.000170 time 0.2663 (0.2897) loss 3.6036 (3.2339) grad_norm 2.3834 (inf) [2021-04-16 13:11:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][1160/1251] eta 0:00:26 lr 0.000170 time 0.2864 (0.2898) loss 3.3524 (3.2346) grad_norm 2.2551 (inf) [2021-04-16 13:11:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][1170/1251] eta 0:00:23 lr 0.000170 time 0.2985 (0.2897) loss 3.2898 (3.2337) grad_norm 2.5110 (inf) [2021-04-16 13:11:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][1180/1251] eta 0:00:20 lr 0.000170 time 0.3073 (0.2896) loss 2.2513 (3.2352) grad_norm 2.1145 (inf) [2021-04-16 13:11:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][1190/1251] eta 0:00:17 lr 0.000170 time 0.2776 (0.2896) loss 3.2520 (3.2344) grad_norm 2.2271 (inf) [2021-04-16 13:11:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][1200/1251] eta 0:00:14 lr 0.000170 time 0.2730 (0.2895) loss 3.4633 (3.2342) grad_norm 2.0837 (inf) [2021-04-16 13:11:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][1210/1251] eta 0:00:11 lr 0.000170 time 0.2965 (0.2894) loss 3.5921 (3.2350) grad_norm 2.6973 (inf) [2021-04-16 13:11:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][1220/1251] eta 0:00:08 lr 0.000170 time 0.2822 (0.2893) loss 3.5081 (3.2363) grad_norm 2.2782 (inf) [2021-04-16 13:11:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][1230/1251] eta 0:00:06 lr 0.000170 time 0.3017 (0.2894) loss 2.6042 (3.2362) grad_norm 2.3819 (inf) [2021-04-16 13:11:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][1240/1251] eta 0:00:03 lr 0.000170 time 0.2673 (0.2892) loss 3.2072 (3.2362) grad_norm 1.9641 (inf) [2021-04-16 13:11:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [220/300][1250/1251] eta 0:00:00 lr 0.000170 time 0.2556 (0.2890) loss 1.9590 (3.2332) grad_norm 2.1454 (inf) [2021-04-16 13:11:32 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 220 training takes 0:06:05 [2021-04-16 13:11:32 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_220.pth saving...... [2021-04-16 13:11:45 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_220.pth saved !!! [2021-04-16 13:11:46 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.077 (1.077) Loss 0.9026 (0.9026) Acc@1 77.930 (77.930) Acc@5 95.215 (95.215) [2021-04-16 13:11:48 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.221 (0.232) Loss 0.8862 (0.8836) Acc@1 79.980 (79.306) Acc@5 94.629 (94.709) [2021-04-16 13:11:51 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.994 (0.256) Loss 0.8272 (0.8769) Acc@1 79.688 (79.348) Acc@5 95.215 (94.703) [2021-04-16 13:11:53 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.080 (0.248) Loss 0.8028 (0.8732) Acc@1 80.371 (79.316) Acc@5 96.289 (94.727) [2021-04-16 13:11:54 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.223) Loss 0.8506 (0.8712) Acc@1 79.199 (79.247) Acc@5 95.215 (94.765) [2021-04-16 13:12:00 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 79.188 Acc@5 94.752 [2021-04-16 13:12:00 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 79.2% [2021-04-16 13:12:00 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 79.19% [2021-04-16 13:12:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][0/1251] eta 2:43:42 lr 0.000170 time 7.8519 (7.8519) loss 3.6138 (3.6138) grad_norm 2.0998 (2.0998) [2021-04-16 13:12:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][10/1251] eta 0:19:59 lr 0.000170 time 0.2764 (0.9669) loss 2.9969 (3.3217) grad_norm 1.9573 (2.1597) [2021-04-16 13:12:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][20/1251] eta 0:13:05 lr 0.000170 time 0.2857 (0.6378) loss 2.3830 (3.2110) grad_norm 2.1816 (2.1895) [2021-04-16 13:12:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][30/1251] eta 0:10:41 lr 0.000170 time 0.2739 (0.5252) loss 3.2703 (3.1902) grad_norm 1.8861 (2.1831) [2021-04-16 13:12:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][40/1251] eta 0:09:23 lr 0.000170 time 0.2899 (0.4651) loss 3.6647 (3.1928) grad_norm 2.2675 (2.2157) [2021-04-16 13:12:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][50/1251] eta 0:08:35 lr 0.000170 time 0.2918 (0.4288) loss 3.8772 (3.1844) grad_norm 2.1945 (2.2463) [2021-04-16 13:12:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][60/1251] eta 0:08:01 lr 0.000170 time 0.2937 (0.4047) loss 3.4381 (3.1519) grad_norm 2.9122 (2.2714) [2021-04-16 13:12:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][70/1251] eta 0:07:37 lr 0.000170 time 0.2696 (0.3877) loss 2.3140 (3.1659) grad_norm 1.9349 (2.2524) [2021-04-16 13:12:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][80/1251] eta 0:07:18 lr 0.000170 time 0.2973 (0.3749) loss 3.6522 (3.1960) grad_norm 2.7429 (2.2654) [2021-04-16 13:12:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][90/1251] eta 0:07:05 lr 0.000170 time 0.2427 (0.3661) loss 1.8877 (3.1949) grad_norm 2.0696 (2.2497) [2021-04-16 13:12:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][100/1251] eta 0:06:53 lr 0.000170 time 0.2789 (0.3592) loss 2.1529 (3.2021) grad_norm 2.2938 (2.2474) [2021-04-16 13:12:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][110/1251] eta 0:06:41 lr 0.000170 time 0.2845 (0.3521) loss 2.4605 (3.1812) grad_norm 2.5394 (2.2577) [2021-04-16 13:12:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][120/1251] eta 0:06:31 lr 0.000170 time 0.2844 (0.3459) loss 3.4329 (3.1735) grad_norm 2.3756 (2.2605) [2021-04-16 13:12:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][130/1251] eta 0:06:23 lr 0.000170 time 0.2938 (0.3419) loss 2.6044 (3.1542) grad_norm 2.3137 (2.2626) [2021-04-16 13:12:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][140/1251] eta 0:06:14 lr 0.000170 time 0.2809 (0.3374) loss 3.7469 (3.1686) grad_norm 2.6920 (2.2678) [2021-04-16 13:12:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][150/1251] eta 0:06:07 lr 0.000169 time 0.2832 (0.3341) loss 3.9669 (3.1920) grad_norm 1.9273 (2.2735) [2021-04-16 13:12:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][160/1251] eta 0:06:00 lr 0.000169 time 0.2960 (0.3309) loss 3.5419 (3.2159) grad_norm 2.0457 (2.2715) [2021-04-16 13:12:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][170/1251] eta 0:05:54 lr 0.000169 time 0.2722 (0.3277) loss 3.6911 (3.2173) grad_norm 2.2236 (2.2693) [2021-04-16 13:12:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][180/1251] eta 0:05:47 lr 0.000169 time 0.2943 (0.3248) loss 3.6541 (3.2249) grad_norm 2.0077 (2.2637) [2021-04-16 13:13:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][190/1251] eta 0:05:42 lr 0.000169 time 0.2713 (0.3225) loss 3.2417 (3.2215) grad_norm 2.4253 (2.2643) [2021-04-16 13:13:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][200/1251] eta 0:05:36 lr 0.000169 time 0.2593 (0.3205) loss 3.7718 (3.2289) grad_norm 2.0027 (2.2676) [2021-04-16 13:13:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][210/1251] eta 0:05:31 lr 0.000169 time 0.2648 (0.3184) loss 3.5119 (3.2391) grad_norm 2.3093 (2.2681) [2021-04-16 13:13:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][220/1251] eta 0:05:26 lr 0.000169 time 0.2840 (0.3165) loss 4.0514 (3.2420) grad_norm 2.3169 (2.2680) [2021-04-16 13:13:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][230/1251] eta 0:05:21 lr 0.000169 time 0.2556 (0.3147) loss 3.3511 (3.2502) grad_norm 2.1121 (2.2683) [2021-04-16 13:13:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][240/1251] eta 0:05:16 lr 0.000169 time 0.2753 (0.3132) loss 3.3225 (3.2452) grad_norm 3.0245 (2.2755) [2021-04-16 13:13:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][250/1251] eta 0:05:11 lr 0.000169 time 0.2911 (0.3116) loss 3.4239 (3.2381) grad_norm 2.2969 (2.2740) [2021-04-16 13:13:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][260/1251] eta 0:05:07 lr 0.000169 time 0.2483 (0.3101) loss 3.1941 (3.2390) grad_norm 2.1010 (2.2726) [2021-04-16 13:13:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][270/1251] eta 0:05:03 lr 0.000169 time 0.2862 (0.3090) loss 3.0057 (3.2473) grad_norm 2.7491 (2.2689) [2021-04-16 13:13:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][280/1251] eta 0:04:58 lr 0.000169 time 0.3023 (0.3078) loss 2.6220 (3.2527) grad_norm 2.2032 (2.2693) [2021-04-16 13:13:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][290/1251] eta 0:04:55 lr 0.000169 time 0.2510 (0.3070) loss 3.7402 (3.2418) grad_norm 2.0198 (2.2649) [2021-04-16 13:13:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][300/1251] eta 0:04:51 lr 0.000169 time 0.2764 (0.3060) loss 2.1786 (3.2376) grad_norm 2.2657 (2.2657) [2021-04-16 13:13:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][310/1251] eta 0:04:47 lr 0.000169 time 0.3028 (0.3055) loss 3.3706 (3.2345) grad_norm 2.1803 (2.2672) [2021-04-16 13:13:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][320/1251] eta 0:04:43 lr 0.000169 time 0.2706 (0.3045) loss 3.0993 (3.2325) grad_norm 2.5051 (2.2726) [2021-04-16 13:13:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][330/1251] eta 0:04:40 lr 0.000169 time 0.2687 (0.3041) loss 3.5703 (3.2270) grad_norm 2.1383 (2.2752) [2021-04-16 13:13:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][340/1251] eta 0:04:36 lr 0.000169 time 0.3046 (0.3034) loss 3.8601 (3.2319) grad_norm 2.2332 (2.2721) [2021-04-16 13:13:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][350/1251] eta 0:04:33 lr 0.000169 time 0.2779 (0.3030) loss 3.4185 (3.2197) grad_norm 2.2954 (2.2753) [2021-04-16 13:13:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][360/1251] eta 0:04:29 lr 0.000169 time 0.2793 (0.3027) loss 3.2696 (3.2145) grad_norm 1.9064 (2.2722) [2021-04-16 13:13:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][370/1251] eta 0:04:26 lr 0.000169 time 0.2872 (0.3025) loss 3.3951 (3.2173) grad_norm 2.2535 (2.2703) [2021-04-16 13:13:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][380/1251] eta 0:04:22 lr 0.000169 time 0.2907 (0.3018) loss 3.1366 (3.2104) grad_norm 2.1674 (2.2668) [2021-04-16 13:13:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][390/1251] eta 0:04:19 lr 0.000169 time 0.2714 (0.3012) loss 3.4062 (3.2144) grad_norm 2.5721 (2.2674) [2021-04-16 13:14:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][400/1251] eta 0:04:15 lr 0.000169 time 0.2747 (0.3005) loss 3.2310 (3.2164) grad_norm 2.3686 (2.2652) [2021-04-16 13:14:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][410/1251] eta 0:04:12 lr 0.000169 time 0.2589 (0.3000) loss 3.8257 (3.2128) grad_norm 2.5041 (2.2621) [2021-04-16 13:14:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][420/1251] eta 0:04:08 lr 0.000169 time 0.2657 (0.2995) loss 2.8533 (3.2110) grad_norm 2.0871 (2.2657) [2021-04-16 13:14:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][430/1251] eta 0:04:05 lr 0.000169 time 0.2886 (0.2994) loss 2.4182 (3.2050) grad_norm 2.0789 (2.2640) [2021-04-16 13:14:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][440/1251] eta 0:04:02 lr 0.000169 time 0.2839 (0.2988) loss 3.8138 (3.2008) grad_norm 2.5331 (2.2638) [2021-04-16 13:14:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][450/1251] eta 0:03:59 lr 0.000169 time 0.2774 (0.2986) loss 3.0927 (3.2015) grad_norm 2.3560 (2.2657) [2021-04-16 13:14:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][460/1251] eta 0:03:55 lr 0.000169 time 0.2533 (0.2980) loss 3.2127 (3.2096) grad_norm 2.4768 (2.2667) [2021-04-16 13:14:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][470/1251] eta 0:03:52 lr 0.000169 time 0.2780 (0.2980) loss 2.5440 (3.2053) grad_norm 2.2121 (2.2705) [2021-04-16 13:14:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][480/1251] eta 0:03:49 lr 0.000168 time 0.3337 (0.2978) loss 2.0656 (3.2007) grad_norm 2.1771 (2.2710) [2021-04-16 13:14:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][490/1251] eta 0:03:46 lr 0.000168 time 0.2724 (0.2974) loss 2.7135 (3.1953) grad_norm 1.8989 (2.2664) [2021-04-16 13:14:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][500/1251] eta 0:03:43 lr 0.000168 time 0.2823 (0.2970) loss 2.0895 (3.1941) grad_norm 2.5237 (2.2668) [2021-04-16 13:14:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][510/1251] eta 0:03:39 lr 0.000168 time 0.2964 (0.2966) loss 3.3442 (3.1916) grad_norm 2.2742 (2.2651) [2021-04-16 13:14:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][520/1251] eta 0:03:36 lr 0.000168 time 0.2898 (0.2964) loss 3.7745 (3.1957) grad_norm 2.7943 (2.2691) [2021-04-16 13:14:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][530/1251] eta 0:03:33 lr 0.000168 time 0.2734 (0.2960) loss 2.5213 (3.1925) grad_norm 2.3659 (2.2682) [2021-04-16 13:14:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][540/1251] eta 0:03:30 lr 0.000168 time 0.2549 (0.2956) loss 2.6826 (3.1910) grad_norm 1.8831 (2.2657) [2021-04-16 13:14:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][550/1251] eta 0:03:26 lr 0.000168 time 0.3006 (0.2953) loss 3.4346 (3.1946) grad_norm 2.1150 (2.2625) [2021-04-16 13:14:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][560/1251] eta 0:03:23 lr 0.000168 time 0.2502 (0.2949) loss 3.5425 (3.1973) grad_norm 2.2394 (2.2622) [2021-04-16 13:14:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][570/1251] eta 0:03:20 lr 0.000168 time 0.2829 (0.2947) loss 2.3598 (3.1924) grad_norm 2.0291 (2.2634) [2021-04-16 13:14:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][580/1251] eta 0:03:17 lr 0.000168 time 0.2659 (0.2949) loss 2.3522 (3.1878) grad_norm 2.2294 (2.2635) [2021-04-16 13:14:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][590/1251] eta 0:03:14 lr 0.000168 time 0.2881 (0.2947) loss 3.7567 (3.1907) grad_norm 1.9789 (2.2644) [2021-04-16 13:14:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][600/1251] eta 0:03:11 lr 0.000168 time 0.3989 (0.2946) loss 3.4951 (3.1940) grad_norm 2.3235 (2.2662) [2021-04-16 13:15:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][610/1251] eta 0:03:08 lr 0.000168 time 0.2836 (0.2943) loss 3.5101 (3.1960) grad_norm 2.0982 (2.2684) [2021-04-16 13:15:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][620/1251] eta 0:03:05 lr 0.000168 time 0.2819 (0.2942) loss 3.7225 (3.2021) grad_norm 1.9996 (2.2682) [2021-04-16 13:15:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][630/1251] eta 0:03:02 lr 0.000168 time 0.2870 (0.2940) loss 3.7100 (3.1974) grad_norm 2.0852 (2.2661) [2021-04-16 13:15:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][640/1251] eta 0:02:59 lr 0.000168 time 0.2785 (0.2937) loss 3.0718 (3.1984) grad_norm 1.8799 (2.2639) [2021-04-16 13:15:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][650/1251] eta 0:02:56 lr 0.000168 time 0.2790 (0.2935) loss 1.9871 (3.1996) grad_norm 2.7538 (2.2650) [2021-04-16 13:15:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][660/1251] eta 0:02:53 lr 0.000168 time 0.2648 (0.2933) loss 2.6775 (3.1965) grad_norm 2.3315 (2.2622) [2021-04-16 13:15:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][670/1251] eta 0:02:50 lr 0.000168 time 0.2886 (0.2933) loss 3.2160 (3.2002) grad_norm 2.1045 (2.2619) [2021-04-16 13:15:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][680/1251] eta 0:02:47 lr 0.000168 time 0.3168 (0.2933) loss 2.0050 (3.2022) grad_norm 2.0426 (2.2586) [2021-04-16 13:15:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][690/1251] eta 0:02:44 lr 0.000168 time 0.3011 (0.2931) loss 3.6131 (3.2037) grad_norm 2.2928 (2.2597) [2021-04-16 13:15:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][700/1251] eta 0:02:41 lr 0.000168 time 0.2609 (0.2928) loss 3.2480 (3.2059) grad_norm 2.3462 (2.2583) [2021-04-16 13:15:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][710/1251] eta 0:02:38 lr 0.000168 time 0.2630 (0.2928) loss 2.8349 (3.2077) grad_norm 2.1728 (2.2586) [2021-04-16 13:15:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][720/1251] eta 0:02:35 lr 0.000168 time 0.2883 (0.2926) loss 2.6888 (3.2032) grad_norm 1.9543 (2.2578) [2021-04-16 13:15:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][730/1251] eta 0:02:32 lr 0.000168 time 0.2825 (0.2926) loss 2.9464 (3.2033) grad_norm 2.8666 (2.2598) [2021-04-16 13:15:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][740/1251] eta 0:02:29 lr 0.000168 time 0.2652 (0.2924) loss 2.0644 (3.2001) grad_norm 3.1682 (2.2635) [2021-04-16 13:15:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][750/1251] eta 0:02:26 lr 0.000168 time 0.2806 (0.2922) loss 2.6011 (3.1969) grad_norm 1.9896 (2.2632) [2021-04-16 13:15:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][760/1251] eta 0:02:23 lr 0.000168 time 0.2715 (0.2920) loss 3.2242 (3.1979) grad_norm 2.4171 (2.2643) [2021-04-16 13:15:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][770/1251] eta 0:02:20 lr 0.000168 time 0.2756 (0.2918) loss 2.2476 (3.1958) grad_norm 1.8366 (2.2628) [2021-04-16 13:15:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][780/1251] eta 0:02:17 lr 0.000168 time 0.2775 (0.2916) loss 3.5818 (3.1958) grad_norm 2.0576 (2.2619) [2021-04-16 13:15:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][790/1251] eta 0:02:14 lr 0.000168 time 0.2743 (0.2915) loss 3.5456 (3.1947) grad_norm 2.1881 (2.2610) [2021-04-16 13:15:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][800/1251] eta 0:02:11 lr 0.000168 time 0.2667 (0.2913) loss 3.3864 (3.1971) grad_norm 2.1903 (2.2605) [2021-04-16 13:15:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][810/1251] eta 0:02:08 lr 0.000167 time 0.2902 (0.2912) loss 3.5169 (3.1955) grad_norm 2.4785 (2.2621) [2021-04-16 13:15:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][820/1251] eta 0:02:05 lr 0.000167 time 0.2788 (0.2910) loss 3.5192 (3.1977) grad_norm 2.0966 (2.2610) [2021-04-16 13:16:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][830/1251] eta 0:02:02 lr 0.000167 time 0.2943 (0.2909) loss 3.0150 (3.1995) grad_norm 2.2242 (2.2621) [2021-04-16 13:16:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][840/1251] eta 0:01:59 lr 0.000167 time 0.3006 (0.2908) loss 3.8935 (3.2017) grad_norm 3.0109 (2.2648) [2021-04-16 13:16:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][850/1251] eta 0:01:56 lr 0.000167 time 0.2742 (0.2908) loss 3.8253 (3.2039) grad_norm 2.8070 (2.2679) [2021-04-16 13:16:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][860/1251] eta 0:01:53 lr 0.000167 time 0.2772 (0.2907) loss 2.6127 (3.2050) grad_norm 2.1764 (2.2688) [2021-04-16 13:16:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][870/1251] eta 0:01:50 lr 0.000167 time 0.2975 (0.2906) loss 3.6810 (3.2047) grad_norm 2.2747 (2.2699) [2021-04-16 13:16:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][880/1251] eta 0:01:47 lr 0.000167 time 0.4251 (0.2908) loss 2.9111 (3.2069) grad_norm 1.9299 (2.2708) [2021-04-16 13:16:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][890/1251] eta 0:01:44 lr 0.000167 time 0.2703 (0.2906) loss 2.6855 (3.2077) grad_norm 2.5364 (2.2733) [2021-04-16 13:16:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][900/1251] eta 0:01:41 lr 0.000167 time 0.4135 (0.2906) loss 3.3405 (3.2087) grad_norm 2.1714 (2.2738) [2021-04-16 13:16:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][910/1251] eta 0:01:39 lr 0.000167 time 0.2776 (0.2905) loss 2.3864 (3.2089) grad_norm 2.2940 (2.2731) [2021-04-16 13:16:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][920/1251] eta 0:01:36 lr 0.000167 time 0.2915 (0.2904) loss 2.7871 (3.2089) grad_norm 2.2342 (2.2722) [2021-04-16 13:16:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][930/1251] eta 0:01:33 lr 0.000167 time 0.2474 (0.2904) loss 2.5864 (3.2092) grad_norm 2.8089 (2.2753) [2021-04-16 13:16:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][940/1251] eta 0:01:30 lr 0.000167 time 0.2795 (0.2904) loss 2.5950 (3.2098) grad_norm 2.1620 (2.2766) [2021-04-16 13:16:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][950/1251] eta 0:01:27 lr 0.000167 time 0.2785 (0.2903) loss 2.8652 (3.2103) grad_norm 2.1509 (2.2768) [2021-04-16 13:16:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][960/1251] eta 0:01:24 lr 0.000167 time 0.2960 (0.2901) loss 2.5416 (3.2122) grad_norm 2.0499 (2.2776) [2021-04-16 13:16:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][970/1251] eta 0:01:21 lr 0.000167 time 0.2732 (0.2900) loss 3.6386 (3.2128) grad_norm 1.9513 (2.2772) [2021-04-16 13:16:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][980/1251] eta 0:01:18 lr 0.000167 time 0.2638 (0.2898) loss 3.4125 (3.2115) grad_norm 2.5522 (2.2787) [2021-04-16 13:16:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][990/1251] eta 0:01:15 lr 0.000167 time 0.2799 (0.2898) loss 2.6865 (3.2098) grad_norm 2.0770 (2.2801) [2021-04-16 13:16:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][1000/1251] eta 0:01:12 lr 0.000167 time 0.2789 (0.2897) loss 3.2071 (3.2102) grad_norm 2.9926 (2.2819) [2021-04-16 13:16:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][1010/1251] eta 0:01:09 lr 0.000167 time 0.2651 (0.2896) loss 2.8130 (3.2128) grad_norm 2.3709 (2.2838) [2021-04-16 13:16:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][1020/1251] eta 0:01:06 lr 0.000167 time 0.2854 (0.2896) loss 3.5594 (3.2127) grad_norm 2.6387 (2.2850) [2021-04-16 13:16:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][1030/1251] eta 0:01:03 lr 0.000167 time 0.2671 (0.2895) loss 3.3188 (3.2142) grad_norm 2.5138 (2.2859) [2021-04-16 13:17:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][1040/1251] eta 0:01:01 lr 0.000167 time 0.2813 (0.2896) loss 3.1985 (3.2143) grad_norm 1.9576 (2.2854) [2021-04-16 13:17:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][1050/1251] eta 0:00:58 lr 0.000167 time 0.3022 (0.2894) loss 3.6390 (3.2143) grad_norm 2.4136 (2.2854) [2021-04-16 13:17:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][1060/1251] eta 0:00:55 lr 0.000167 time 0.2682 (0.2893) loss 3.7050 (3.2162) grad_norm 2.2101 (2.2851) [2021-04-16 13:17:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][1070/1251] eta 0:00:52 lr 0.000167 time 0.2799 (0.2892) loss 2.3097 (3.2167) grad_norm 2.4131 (2.2842) [2021-04-16 13:17:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][1080/1251] eta 0:00:49 lr 0.000167 time 0.2820 (0.2891) loss 2.0721 (3.2164) grad_norm 2.1415 (2.2849) [2021-04-16 13:17:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][1090/1251] eta 0:00:46 lr 0.000167 time 0.2704 (0.2890) loss 3.2449 (3.2178) grad_norm 1.9690 (2.2860) [2021-04-16 13:17:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][1100/1251] eta 0:00:43 lr 0.000167 time 0.2673 (0.2889) loss 3.4886 (3.2193) grad_norm 2.5531 (2.2855) [2021-04-16 13:17:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][1110/1251] eta 0:00:40 lr 0.000167 time 0.2785 (0.2888) loss 2.0188 (3.2170) grad_norm 1.8735 (2.2849) [2021-04-16 13:17:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][1120/1251] eta 0:00:37 lr 0.000167 time 0.2810 (0.2887) loss 2.8733 (3.2183) grad_norm 2.2208 (2.2850) [2021-04-16 13:17:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][1130/1251] eta 0:00:34 lr 0.000167 time 0.2805 (0.2886) loss 3.7701 (3.2182) grad_norm 2.2876 (2.2860) [2021-04-16 13:17:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][1140/1251] eta 0:00:32 lr 0.000166 time 0.2952 (0.2885) loss 2.2544 (3.2156) grad_norm 2.5843 (2.2868) [2021-04-16 13:17:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][1150/1251] eta 0:00:29 lr 0.000166 time 0.2824 (0.2884) loss 3.6188 (3.2152) grad_norm 2.7450 (2.2879) [2021-04-16 13:17:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][1160/1251] eta 0:00:26 lr 0.000166 time 0.2665 (0.2884) loss 3.3426 (3.2174) grad_norm 2.0849 (2.2878) [2021-04-16 13:17:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][1170/1251] eta 0:00:23 lr 0.000166 time 0.2859 (0.2883) loss 3.8128 (3.2191) grad_norm 3.8312 (2.2883) [2021-04-16 13:17:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][1180/1251] eta 0:00:20 lr 0.000166 time 0.2972 (0.2883) loss 3.0798 (3.2199) grad_norm 2.1577 (2.2881) [2021-04-16 13:17:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][1190/1251] eta 0:00:17 lr 0.000166 time 0.2816 (0.2882) loss 3.2386 (3.2205) grad_norm 2.3736 (2.2886) [2021-04-16 13:17:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][1200/1251] eta 0:00:14 lr 0.000166 time 0.2704 (0.2881) loss 3.5421 (3.2222) grad_norm 1.9579 (2.2890) [2021-04-16 13:17:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][1210/1251] eta 0:00:11 lr 0.000166 time 0.2683 (0.2880) loss 3.7058 (3.2245) grad_norm 2.4970 (2.2889) [2021-04-16 13:17:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][1220/1251] eta 0:00:08 lr 0.000166 time 0.2569 (0.2880) loss 2.6702 (3.2241) grad_norm 2.2348 (2.2882) [2021-04-16 13:17:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][1230/1251] eta 0:00:06 lr 0.000166 time 0.2613 (0.2879) loss 3.7461 (3.2264) grad_norm 2.4152 (2.2884) [2021-04-16 13:17:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][1240/1251] eta 0:00:03 lr 0.000166 time 0.2382 (0.2878) loss 3.1011 (3.2270) grad_norm 2.1845 (2.2882) [2021-04-16 13:18:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [221/300][1250/1251] eta 0:00:00 lr 0.000166 time 0.2490 (0.2875) loss 2.7468 (3.2256) grad_norm 2.0406 (2.2880) [2021-04-16 13:18:04 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 221 training takes 0:06:04 [2021-04-16 13:18:04 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_221.pth saving...... [2021-04-16 13:18:26 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_221.pth saved !!! [2021-04-16 13:18:27 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.111 (1.111) Loss 0.9052 (0.9052) Acc@1 78.906 (78.906) Acc@5 95.117 (95.117) [2021-04-16 13:18:29 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.252 (0.228) Loss 0.9146 (0.8835) Acc@1 78.906 (79.279) Acc@5 94.824 (94.789) [2021-04-16 13:18:31 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.615 (0.252) Loss 0.8187 (0.8824) Acc@1 81.934 (79.148) Acc@5 95.215 (94.708) [2021-04-16 13:18:33 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.146 (0.221) Loss 0.8586 (0.8796) Acc@1 79.980 (79.143) Acc@5 94.824 (94.676) [2021-04-16 13:18:35 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.163 (0.216) Loss 0.8276 (0.8778) Acc@1 80.469 (79.142) Acc@5 95.117 (94.705) [2021-04-16 13:18:51 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 79.342 Acc@5 94.798 [2021-04-16 13:18:51 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 79.3% [2021-04-16 13:18:51 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 79.34% [2021-04-16 13:18:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][0/1251] eta 0:30:35 lr 0.000166 time 1.4676 (1.4676) loss 3.4732 (3.4732) grad_norm 2.2199 (2.2199) [2021-04-16 13:18:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][10/1251] eta 0:07:55 lr 0.000166 time 0.2838 (0.3830) loss 2.8474 (3.4729) grad_norm 2.2935 (2.2326) [2021-04-16 13:18:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][20/1251] eta 0:06:46 lr 0.000166 time 0.2603 (0.3304) loss 3.8098 (3.4175) grad_norm 2.0333 (2.2498) [2021-04-16 13:19:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][30/1251] eta 0:06:22 lr 0.000166 time 0.2799 (0.3134) loss 3.2603 (3.3051) grad_norm 2.5197 (2.2967) [2021-04-16 13:19:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][40/1251] eta 0:06:10 lr 0.000166 time 0.2990 (0.3057) loss 3.7901 (3.3309) grad_norm 2.6116 (2.3015) [2021-04-16 13:19:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][50/1251] eta 0:05:59 lr 0.000166 time 0.2753 (0.2993) loss 2.4634 (3.2496) grad_norm 1.9736 (2.2729) [2021-04-16 13:19:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][60/1251] eta 0:05:52 lr 0.000166 time 0.2933 (0.2956) loss 3.5470 (3.2536) grad_norm 2.0261 (2.2503) [2021-04-16 13:19:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][70/1251] eta 0:05:48 lr 0.000166 time 0.2668 (0.2947) loss 2.4659 (3.1970) grad_norm 2.1388 (2.2333) [2021-04-16 13:19:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][80/1251] eta 0:05:42 lr 0.000166 time 0.2724 (0.2921) loss 3.7329 (3.1869) grad_norm 2.4335 (2.2335) [2021-04-16 13:19:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][90/1251] eta 0:05:37 lr 0.000166 time 0.2621 (0.2903) loss 3.3670 (3.2062) grad_norm 2.3961 (2.2453) [2021-04-16 13:19:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][100/1251] eta 0:05:34 lr 0.000166 time 0.2775 (0.2903) loss 3.4259 (3.1608) grad_norm 2.2607 (2.2535) [2021-04-16 13:19:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][110/1251] eta 0:05:30 lr 0.000166 time 0.2694 (0.2892) loss 3.6272 (3.1685) grad_norm 2.8035 (2.2612) [2021-04-16 13:19:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][120/1251] eta 0:05:26 lr 0.000166 time 0.2965 (0.2886) loss 3.5225 (3.1824) grad_norm 2.3300 (2.2722) [2021-04-16 13:19:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][130/1251] eta 0:05:23 lr 0.000166 time 0.2705 (0.2887) loss 3.0244 (3.1831) grad_norm 2.3522 (2.2768) [2021-04-16 13:19:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][140/1251] eta 0:05:21 lr 0.000166 time 0.2726 (0.2894) loss 3.5257 (3.1777) grad_norm 2.3955 (2.2989) [2021-04-16 13:19:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][150/1251] eta 0:05:17 lr 0.000166 time 0.2589 (0.2885) loss 2.6584 (3.1669) grad_norm 1.9992 (2.3048) [2021-04-16 13:19:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][160/1251] eta 0:05:13 lr 0.000166 time 0.2589 (0.2876) loss 3.2231 (3.1814) grad_norm 2.6239 (2.3107) [2021-04-16 13:19:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][170/1251] eta 0:05:10 lr 0.000166 time 0.2770 (0.2870) loss 3.3978 (3.1914) grad_norm 2.1813 (2.3088) [2021-04-16 13:19:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][180/1251] eta 0:05:06 lr 0.000166 time 0.2848 (0.2865) loss 3.0402 (3.1851) grad_norm 3.7512 (2.3095) [2021-04-16 13:19:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][190/1251] eta 0:05:03 lr 0.000166 time 0.2915 (0.2863) loss 3.5063 (3.1813) grad_norm 2.1268 (2.3071) [2021-04-16 13:19:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][200/1251] eta 0:05:00 lr 0.000166 time 0.2765 (0.2858) loss 3.4973 (3.2004) grad_norm 2.5890 (2.3079) [2021-04-16 13:19:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][210/1251] eta 0:04:57 lr 0.000166 time 0.2836 (0.2855) loss 3.9412 (3.1992) grad_norm 2.0604 (2.3087) [2021-04-16 13:19:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][220/1251] eta 0:04:54 lr 0.000165 time 0.2737 (0.2858) loss 2.8472 (3.2040) grad_norm 2.4177 (2.3072) [2021-04-16 13:19:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][230/1251] eta 0:04:51 lr 0.000165 time 0.2593 (0.2856) loss 2.8312 (3.2018) grad_norm 2.2281 (2.3092) [2021-04-16 13:19:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][240/1251] eta 0:04:48 lr 0.000165 time 0.2758 (0.2853) loss 3.7411 (3.2017) grad_norm 2.2542 (2.3101) [2021-04-16 13:20:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][250/1251] eta 0:04:45 lr 0.000165 time 0.2751 (0.2856) loss 2.8007 (3.1969) grad_norm 2.3311 (2.3106) [2021-04-16 13:20:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][260/1251] eta 0:04:42 lr 0.000165 time 0.3072 (0.2853) loss 4.0371 (3.1967) grad_norm 2.3304 (2.3108) [2021-04-16 13:20:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][270/1251] eta 0:04:39 lr 0.000165 time 0.2816 (0.2850) loss 3.7980 (3.1981) grad_norm 1.9864 (2.3158) [2021-04-16 13:20:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][280/1251] eta 0:04:36 lr 0.000165 time 0.2639 (0.2848) loss 3.6552 (3.1922) grad_norm 2.3621 (2.3128) [2021-04-16 13:20:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][290/1251] eta 0:04:33 lr 0.000165 time 0.2795 (0.2845) loss 3.2544 (3.1991) grad_norm 2.3725 (2.3119) [2021-04-16 13:20:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][300/1251] eta 0:04:30 lr 0.000165 time 0.2582 (0.2843) loss 3.2324 (3.2103) grad_norm 2.5678 (2.3129) [2021-04-16 13:20:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][310/1251] eta 0:04:27 lr 0.000165 time 0.2914 (0.2841) loss 3.4992 (3.2111) grad_norm 2.4523 (2.3120) [2021-04-16 13:20:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][320/1251] eta 0:04:26 lr 0.000165 time 0.2958 (0.2862) loss 2.8172 (3.2093) grad_norm 3.8017 (2.3160) [2021-04-16 13:20:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][330/1251] eta 0:04:23 lr 0.000165 time 0.2906 (0.2858) loss 2.5911 (3.2056) grad_norm 2.3323 (2.3181) [2021-04-16 13:20:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][340/1251] eta 0:04:20 lr 0.000165 time 0.2564 (0.2855) loss 3.5948 (3.2002) grad_norm 2.2425 (2.3163) [2021-04-16 13:20:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][350/1251] eta 0:04:17 lr 0.000165 time 0.2633 (0.2853) loss 2.9137 (3.1930) grad_norm 2.2508 (2.3140) [2021-04-16 13:20:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][360/1251] eta 0:04:14 lr 0.000165 time 0.2820 (0.2857) loss 3.7242 (3.1944) grad_norm 2.3899 (2.3143) [2021-04-16 13:20:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][370/1251] eta 0:04:11 lr 0.000165 time 0.2426 (0.2855) loss 3.5400 (3.1984) grad_norm 2.2726 (2.3162) [2021-04-16 13:20:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][380/1251] eta 0:04:08 lr 0.000165 time 0.2714 (0.2856) loss 3.0975 (3.1992) grad_norm 2.0885 (2.3128) [2021-04-16 13:20:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][390/1251] eta 0:04:05 lr 0.000165 time 0.2553 (0.2854) loss 3.8768 (3.2018) grad_norm 2.2792 (2.3097) [2021-04-16 13:20:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][400/1251] eta 0:04:02 lr 0.000165 time 0.2838 (0.2854) loss 2.7076 (3.2031) grad_norm 2.2027 (2.3084) [2021-04-16 13:20:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][410/1251] eta 0:03:59 lr 0.000165 time 0.2554 (0.2850) loss 2.7301 (3.2025) grad_norm 3.0159 (2.3118) [2021-04-16 13:20:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][420/1251] eta 0:03:56 lr 0.000165 time 0.2524 (0.2849) loss 3.2687 (3.2059) grad_norm 2.6021 (2.3105) [2021-04-16 13:20:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][430/1251] eta 0:03:53 lr 0.000165 time 0.2919 (0.2848) loss 3.5358 (3.2133) grad_norm 2.4434 (2.3131) [2021-04-16 13:20:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][440/1251] eta 0:03:51 lr 0.000165 time 0.3006 (0.2850) loss 3.9129 (3.2124) grad_norm 2.5721 (2.3143) [2021-04-16 13:20:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][450/1251] eta 0:03:48 lr 0.000165 time 0.2911 (0.2849) loss 3.0098 (3.2082) grad_norm 1.9553 (2.3113) [2021-04-16 13:21:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][460/1251] eta 0:03:45 lr 0.000165 time 0.2815 (0.2849) loss 2.4857 (3.2061) grad_norm 2.0498 (2.3075) [2021-04-16 13:21:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][470/1251] eta 0:03:42 lr 0.000165 time 0.2760 (0.2851) loss 3.3097 (3.2068) grad_norm 2.2843 (2.3059) [2021-04-16 13:21:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][480/1251] eta 0:03:39 lr 0.000165 time 0.2820 (0.2849) loss 3.5164 (3.2082) grad_norm 2.1941 (2.3044) [2021-04-16 13:21:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][490/1251] eta 0:03:36 lr 0.000165 time 0.2423 (0.2847) loss 3.5316 (3.2081) grad_norm 2.0697 (2.3036) [2021-04-16 13:21:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][500/1251] eta 0:03:33 lr 0.000165 time 0.2553 (0.2845) loss 3.3102 (3.2111) grad_norm 2.3412 (2.3047) [2021-04-16 13:21:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][510/1251] eta 0:03:30 lr 0.000165 time 0.2931 (0.2844) loss 3.4915 (3.2123) grad_norm 2.1344 (2.3076) [2021-04-16 13:21:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][520/1251] eta 0:03:27 lr 0.000165 time 0.2937 (0.2842) loss 2.6469 (3.2116) grad_norm 2.6310 (2.3089) [2021-04-16 13:21:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][530/1251] eta 0:03:24 lr 0.000165 time 0.2519 (0.2839) loss 3.5192 (3.2094) grad_norm 2.5346 (2.3101) [2021-04-16 13:21:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][540/1251] eta 0:03:21 lr 0.000165 time 0.2849 (0.2840) loss 2.6037 (3.2053) grad_norm 2.0878 (2.3110) [2021-04-16 13:21:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][550/1251] eta 0:03:19 lr 0.000164 time 0.2812 (0.2839) loss 3.7379 (3.2041) grad_norm 2.1751 (2.3122) [2021-04-16 13:21:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][560/1251] eta 0:03:16 lr 0.000164 time 0.2749 (0.2837) loss 2.6086 (3.2041) grad_norm 2.2398 (nan) [2021-04-16 13:21:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][570/1251] eta 0:03:13 lr 0.000164 time 0.2568 (0.2837) loss 2.8996 (3.2031) grad_norm 2.2524 (nan) [2021-04-16 13:21:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][580/1251] eta 0:03:10 lr 0.000164 time 0.2588 (0.2836) loss 3.3456 (3.2001) grad_norm 2.6446 (nan) [2021-04-16 13:21:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][590/1251] eta 0:03:07 lr 0.000164 time 0.2631 (0.2835) loss 2.3525 (3.1990) grad_norm 2.2233 (nan) [2021-04-16 13:21:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][600/1251] eta 0:03:04 lr 0.000164 time 0.2757 (0.2834) loss 3.7713 (3.1978) grad_norm 2.7027 (nan) [2021-04-16 13:21:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][610/1251] eta 0:03:01 lr 0.000164 time 0.2736 (0.2834) loss 2.7336 (3.2024) grad_norm 2.2321 (nan) [2021-04-16 13:21:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][620/1251] eta 0:02:58 lr 0.000164 time 0.2757 (0.2833) loss 2.6942 (3.2020) grad_norm 2.3601 (nan) [2021-04-16 13:21:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][630/1251] eta 0:02:55 lr 0.000164 time 0.2670 (0.2833) loss 3.3286 (3.2046) grad_norm 2.4607 (nan) [2021-04-16 13:21:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][640/1251] eta 0:02:53 lr 0.000164 time 0.2695 (0.2833) loss 3.2485 (3.2021) grad_norm 2.0424 (nan) [2021-04-16 13:21:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][650/1251] eta 0:02:50 lr 0.000164 time 0.2752 (0.2832) loss 3.1742 (3.2011) grad_norm 2.0053 (nan) [2021-04-16 13:21:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][660/1251] eta 0:02:47 lr 0.000164 time 0.2705 (0.2831) loss 2.4189 (3.1963) grad_norm 2.6509 (nan) [2021-04-16 13:22:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][670/1251] eta 0:02:44 lr 0.000164 time 0.2740 (0.2829) loss 2.8196 (3.1944) grad_norm 2.4421 (nan) [2021-04-16 13:22:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][680/1251] eta 0:02:41 lr 0.000164 time 0.2761 (0.2831) loss 2.5430 (3.1914) grad_norm 2.6531 (nan) [2021-04-16 13:22:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][690/1251] eta 0:02:38 lr 0.000164 time 0.2705 (0.2830) loss 3.7216 (3.1905) grad_norm 2.1716 (nan) [2021-04-16 13:22:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][700/1251] eta 0:02:35 lr 0.000164 time 0.2580 (0.2828) loss 3.4439 (3.1929) grad_norm 2.2595 (nan) [2021-04-16 13:22:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][710/1251] eta 0:02:33 lr 0.000164 time 0.3275 (0.2831) loss 3.5613 (3.1951) grad_norm 2.4549 (nan) [2021-04-16 13:22:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][720/1251] eta 0:02:30 lr 0.000164 time 0.2739 (0.2832) loss 4.1767 (3.1943) grad_norm 2.0949 (nan) [2021-04-16 13:22:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][730/1251] eta 0:02:27 lr 0.000164 time 0.2862 (0.2831) loss 3.8452 (3.1920) grad_norm 3.5722 (nan) [2021-04-16 13:22:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][740/1251] eta 0:02:24 lr 0.000164 time 0.2827 (0.2831) loss 3.3654 (3.1920) grad_norm 2.2304 (nan) [2021-04-16 13:22:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][750/1251] eta 0:02:21 lr 0.000164 time 0.2606 (0.2831) loss 3.7336 (3.1874) grad_norm 2.4207 (nan) [2021-04-16 13:22:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][760/1251] eta 0:02:18 lr 0.000164 time 0.2681 (0.2830) loss 2.8490 (3.1912) grad_norm 2.1479 (nan) [2021-04-16 13:22:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][770/1251] eta 0:02:16 lr 0.000164 time 0.2639 (0.2829) loss 2.4892 (3.1909) grad_norm 2.2757 (nan) [2021-04-16 13:22:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][780/1251] eta 0:02:13 lr 0.000164 time 0.2444 (0.2829) loss 3.5425 (3.1938) grad_norm 2.9098 (nan) [2021-04-16 13:22:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][790/1251] eta 0:02:10 lr 0.000164 time 0.2721 (0.2830) loss 3.1685 (3.1917) grad_norm 1.9787 (nan) [2021-04-16 13:22:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][800/1251] eta 0:02:07 lr 0.000164 time 0.2548 (0.2829) loss 3.5974 (3.1903) grad_norm 2.1781 (nan) [2021-04-16 13:22:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][810/1251] eta 0:02:04 lr 0.000164 time 0.2698 (0.2828) loss 4.1113 (3.1915) grad_norm 2.3874 (nan) [2021-04-16 13:22:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][820/1251] eta 0:02:01 lr 0.000164 time 0.3005 (0.2827) loss 3.9159 (3.1946) grad_norm 2.5984 (nan) [2021-04-16 13:22:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][830/1251] eta 0:01:59 lr 0.000164 time 0.2653 (0.2827) loss 3.3133 (3.1925) grad_norm 2.2673 (nan) [2021-04-16 13:22:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][840/1251] eta 0:01:56 lr 0.000164 time 0.2796 (0.2826) loss 2.9272 (3.1901) grad_norm 2.6738 (nan) [2021-04-16 13:22:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][850/1251] eta 0:01:53 lr 0.000164 time 0.3166 (0.2825) loss 3.8526 (3.1905) grad_norm 2.1229 (nan) [2021-04-16 13:22:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][860/1251] eta 0:01:50 lr 0.000164 time 0.2727 (0.2824) loss 3.4215 (3.1919) grad_norm 2.1692 (nan) [2021-04-16 13:22:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][870/1251] eta 0:01:47 lr 0.000164 time 0.2872 (0.2825) loss 3.3405 (3.1890) grad_norm 2.3504 (nan) [2021-04-16 13:22:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][880/1251] eta 0:01:44 lr 0.000164 time 0.2650 (0.2824) loss 3.8178 (3.1910) grad_norm 2.6851 (nan) [2021-04-16 13:23:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][890/1251] eta 0:01:41 lr 0.000163 time 0.2656 (0.2824) loss 3.4506 (3.1896) grad_norm 2.1457 (nan) [2021-04-16 13:23:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][900/1251] eta 0:01:39 lr 0.000163 time 0.2738 (0.2824) loss 3.2617 (3.1871) grad_norm 2.3631 (nan) [2021-04-16 13:23:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][910/1251] eta 0:01:36 lr 0.000163 time 0.2700 (0.2824) loss 3.8845 (3.1911) grad_norm 2.5541 (nan) [2021-04-16 13:23:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][920/1251] eta 0:01:33 lr 0.000163 time 0.2929 (0.2823) loss 3.7725 (3.1948) grad_norm 2.1475 (nan) [2021-04-16 13:23:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][930/1251] eta 0:01:30 lr 0.000163 time 0.2700 (0.2824) loss 2.4324 (3.1938) grad_norm 2.2091 (nan) [2021-04-16 13:23:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][940/1251] eta 0:01:27 lr 0.000163 time 0.2725 (0.2824) loss 3.0539 (3.1928) grad_norm 2.7606 (nan) [2021-04-16 13:23:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][950/1251] eta 0:01:24 lr 0.000163 time 0.2855 (0.2824) loss 3.7251 (3.1923) grad_norm 3.0400 (nan) [2021-04-16 13:23:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][960/1251] eta 0:01:22 lr 0.000163 time 0.2853 (0.2824) loss 3.2438 (3.1950) grad_norm 2.3763 (nan) [2021-04-16 13:23:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][970/1251] eta 0:01:19 lr 0.000163 time 0.2724 (0.2824) loss 3.0186 (3.1941) grad_norm 2.0547 (nan) [2021-04-16 13:23:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][980/1251] eta 0:01:16 lr 0.000163 time 0.2792 (0.2823) loss 3.0138 (3.1924) grad_norm 1.8764 (nan) [2021-04-16 13:23:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][990/1251] eta 0:01:13 lr 0.000163 time 0.2802 (0.2823) loss 3.5315 (3.1949) grad_norm 2.4588 (nan) [2021-04-16 13:23:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][1000/1251] eta 0:01:10 lr 0.000163 time 0.2783 (0.2822) loss 3.6659 (3.1951) grad_norm 1.9910 (nan) [2021-04-16 13:23:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][1010/1251] eta 0:01:08 lr 0.000163 time 0.2940 (0.2823) loss 3.3465 (3.1967) grad_norm 2.2532 (nan) [2021-04-16 13:23:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][1020/1251] eta 0:01:05 lr 0.000163 time 0.2749 (0.2824) loss 2.4340 (3.1971) grad_norm 2.4617 (nan) [2021-04-16 13:23:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][1030/1251] eta 0:01:02 lr 0.000163 time 0.2566 (0.2823) loss 3.1304 (3.1965) grad_norm 2.1141 (nan) [2021-04-16 13:23:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][1040/1251] eta 0:00:59 lr 0.000163 time 0.2803 (0.2824) loss 3.1750 (3.1956) grad_norm 2.5718 (nan) [2021-04-16 13:23:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][1050/1251] eta 0:00:56 lr 0.000163 time 0.2758 (0.2824) loss 3.5577 (3.1943) grad_norm 2.1635 (nan) [2021-04-16 13:23:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][1060/1251] eta 0:00:53 lr 0.000163 time 0.3083 (0.2824) loss 3.2734 (3.1938) grad_norm 2.1652 (nan) [2021-04-16 13:23:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][1070/1251] eta 0:00:51 lr 0.000163 time 0.2909 (0.2823) loss 3.1386 (3.1957) grad_norm 2.9683 (nan) [2021-04-16 13:23:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][1080/1251] eta 0:00:48 lr 0.000163 time 0.2922 (0.2823) loss 3.8819 (3.1995) grad_norm 2.2684 (nan) [2021-04-16 13:23:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][1090/1251] eta 0:00:45 lr 0.000163 time 0.2662 (0.2822) loss 3.5134 (3.2006) grad_norm 3.6587 (nan) [2021-04-16 13:24:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][1100/1251] eta 0:00:42 lr 0.000163 time 0.2729 (0.2821) loss 3.8874 (3.2015) grad_norm 2.2114 (nan) [2021-04-16 13:24:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][1110/1251] eta 0:00:39 lr 0.000163 time 0.2825 (0.2821) loss 2.7695 (3.2006) grad_norm 2.4682 (nan) [2021-04-16 13:24:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][1120/1251] eta 0:00:36 lr 0.000163 time 0.2828 (0.2821) loss 2.8912 (3.2000) grad_norm 2.1341 (nan) [2021-04-16 13:24:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][1130/1251] eta 0:00:34 lr 0.000163 time 0.2773 (0.2821) loss 2.3956 (3.2004) grad_norm 2.2437 (nan) [2021-04-16 13:24:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][1140/1251] eta 0:00:31 lr 0.000163 time 0.2655 (0.2820) loss 2.3464 (3.1983) grad_norm 2.0370 (nan) [2021-04-16 13:24:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][1150/1251] eta 0:00:28 lr 0.000163 time 0.2782 (0.2821) loss 3.0676 (3.1970) grad_norm 2.1249 (nan) [2021-04-16 13:24:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][1160/1251] eta 0:00:25 lr 0.000163 time 0.2707 (0.2821) loss 2.7242 (3.1958) grad_norm 2.4557 (nan) [2021-04-16 13:24:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][1170/1251] eta 0:00:22 lr 0.000163 time 0.2908 (0.2821) loss 3.0041 (3.1958) grad_norm 2.3723 (nan) [2021-04-16 13:24:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][1180/1251] eta 0:00:20 lr 0.000163 time 0.2540 (0.2820) loss 3.2812 (3.1947) grad_norm 1.9942 (nan) [2021-04-16 13:24:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][1190/1251] eta 0:00:17 lr 0.000163 time 0.2728 (0.2820) loss 3.1273 (3.1952) grad_norm 2.0799 (nan) [2021-04-16 13:24:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][1200/1251] eta 0:00:14 lr 0.000163 time 0.2970 (0.2820) loss 2.6860 (3.1945) grad_norm 3.1409 (nan) [2021-04-16 13:24:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][1210/1251] eta 0:00:11 lr 0.000163 time 0.2543 (0.2820) loss 2.6824 (3.1962) grad_norm 2.3717 (nan) [2021-04-16 13:24:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][1220/1251] eta 0:00:08 lr 0.000162 time 0.2753 (0.2820) loss 3.9777 (3.1966) grad_norm 2.4027 (nan) [2021-04-16 13:24:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][1230/1251] eta 0:00:05 lr 0.000162 time 0.2583 (0.2820) loss 3.5196 (3.1958) grad_norm 2.4598 (nan) [2021-04-16 13:24:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][1240/1251] eta 0:00:03 lr 0.000162 time 0.2627 (0.2819) loss 3.2203 (3.1966) grad_norm 2.7075 (nan) [2021-04-16 13:24:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [222/300][1250/1251] eta 0:00:00 lr 0.000162 time 0.2488 (0.2816) loss 3.1175 (3.1970) grad_norm 2.2300 (nan) [2021-04-16 13:24:46 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 222 training takes 0:05:55 [2021-04-16 13:24:46 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_222.pth saving...... [2021-04-16 13:25:01 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_222.pth saved !!! [2021-04-16 13:25:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.206 (1.206) Loss 0.8639 (0.8639) Acc@1 80.078 (80.078) Acc@5 94.629 (94.629) [2021-04-16 13:25:03 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.148 (0.209) Loss 0.8960 (0.8652) Acc@1 78.125 (79.545) Acc@5 94.629 (94.806) [2021-04-16 13:25:05 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.156 (0.201) Loss 0.9655 (0.8727) Acc@1 77.246 (79.162) Acc@5 93.848 (94.792) [2021-04-16 13:25:08 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.149 (0.240) Loss 0.9784 (0.8807) Acc@1 77.344 (78.963) Acc@5 93.164 (94.717) [2021-04-16 13:25:09 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.217) Loss 0.8787 (0.8716) Acc@1 79.199 (79.130) Acc@5 95.215 (94.822) [2021-04-16 13:25:16 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 79.190 Acc@5 94.818 [2021-04-16 13:25:16 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 79.2% [2021-04-16 13:25:16 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 79.34% [2021-04-16 13:25:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][0/1251] eta 3:43:27 lr 0.000162 time 10.7178 (10.7178) loss 3.8230 (3.8230) grad_norm 2.1141 (2.1141) [2021-04-16 13:25:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][10/1251] eta 0:25:27 lr 0.000162 time 0.4227 (1.2310) loss 2.8201 (3.2609) grad_norm 2.1568 (2.2525) [2021-04-16 13:25:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][20/1251] eta 0:15:52 lr 0.000162 time 0.2645 (0.7738) loss 3.8864 (3.0894) grad_norm 2.1797 (2.2881) [2021-04-16 13:25:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][30/1251] eta 0:12:34 lr 0.000162 time 0.2662 (0.6183) loss 3.9558 (3.0644) grad_norm 2.2153 (2.2772) [2021-04-16 13:25:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][40/1251] eta 0:10:50 lr 0.000162 time 0.2853 (0.5371) loss 2.3532 (3.1214) grad_norm 2.2743 (2.2901) [2021-04-16 13:25:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][50/1251] eta 0:09:44 lr 0.000162 time 0.2701 (0.4866) loss 3.3100 (3.1361) grad_norm 2.0813 (2.2831) [2021-04-16 13:25:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][60/1251] eta 0:09:00 lr 0.000162 time 0.2765 (0.4535) loss 3.5118 (3.1423) grad_norm 2.6235 (2.2890) [2021-04-16 13:25:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][70/1251] eta 0:08:25 lr 0.000162 time 0.2724 (0.4284) loss 3.3701 (3.1891) grad_norm 2.3819 (inf) [2021-04-16 13:25:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][80/1251] eta 0:07:59 lr 0.000162 time 0.2613 (0.4091) loss 3.6102 (3.2108) grad_norm 2.4251 (inf) [2021-04-16 13:25:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][90/1251] eta 0:07:37 lr 0.000162 time 0.2993 (0.3943) loss 3.6020 (3.1780) grad_norm 2.0048 (inf) [2021-04-16 13:25:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][100/1251] eta 0:07:20 lr 0.000162 time 0.2506 (0.3823) loss 2.9542 (3.1898) grad_norm 2.8182 (inf) [2021-04-16 13:25:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][110/1251] eta 0:07:05 lr 0.000162 time 0.2690 (0.3726) loss 2.6794 (3.1899) grad_norm 2.3485 (inf) [2021-04-16 13:26:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][120/1251] eta 0:06:51 lr 0.000162 time 0.2759 (0.3641) loss 3.3431 (3.1884) grad_norm 1.9076 (inf) [2021-04-16 13:26:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][130/1251] eta 0:06:40 lr 0.000162 time 0.2943 (0.3573) loss 3.1451 (3.1917) grad_norm 2.3007 (inf) [2021-04-16 13:26:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][140/1251] eta 0:06:32 lr 0.000162 time 0.2815 (0.3535) loss 2.9126 (3.2006) grad_norm 1.9899 (inf) [2021-04-16 13:26:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][150/1251] eta 0:06:24 lr 0.000162 time 0.3168 (0.3491) loss 3.2414 (3.1861) grad_norm 2.5467 (inf) [2021-04-16 13:26:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][160/1251] eta 0:06:16 lr 0.000162 time 0.2873 (0.3447) loss 3.2591 (3.1912) grad_norm 2.6841 (inf) [2021-04-16 13:26:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][170/1251] eta 0:06:09 lr 0.000162 time 0.2880 (0.3414) loss 2.7588 (3.1965) grad_norm 2.2451 (inf) [2021-04-16 13:26:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][180/1251] eta 0:06:01 lr 0.000162 time 0.2458 (0.3376) loss 3.8121 (3.1988) grad_norm 2.1885 (inf) [2021-04-16 13:26:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][190/1251] eta 0:05:54 lr 0.000162 time 0.2778 (0.3346) loss 3.1402 (3.1830) grad_norm 2.3808 (inf) [2021-04-16 13:26:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][200/1251] eta 0:05:48 lr 0.000162 time 0.2558 (0.3319) loss 2.0204 (3.1774) grad_norm 2.0225 (inf) [2021-04-16 13:26:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][210/1251] eta 0:05:42 lr 0.000162 time 0.2906 (0.3294) loss 3.3712 (3.1740) grad_norm 1.9668 (inf) [2021-04-16 13:26:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][220/1251] eta 0:05:37 lr 0.000162 time 0.2718 (0.3270) loss 3.3407 (3.1733) grad_norm 2.3860 (inf) [2021-04-16 13:26:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][230/1251] eta 0:05:32 lr 0.000162 time 0.2602 (0.3255) loss 3.6783 (3.1727) grad_norm 2.2291 (inf) [2021-04-16 13:26:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][240/1251] eta 0:05:27 lr 0.000162 time 0.2798 (0.3241) loss 3.1447 (3.1761) grad_norm 2.2372 (inf) [2021-04-16 13:26:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][250/1251] eta 0:05:22 lr 0.000162 time 0.2523 (0.3227) loss 2.9081 (3.1553) grad_norm 2.1132 (inf) [2021-04-16 13:26:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][260/1251] eta 0:05:18 lr 0.000162 time 0.2725 (0.3209) loss 2.8274 (3.1531) grad_norm 2.4156 (inf) [2021-04-16 13:26:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][270/1251] eta 0:05:13 lr 0.000162 time 0.3141 (0.3196) loss 3.5872 (3.1623) grad_norm 2.3921 (inf) [2021-04-16 13:26:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][280/1251] eta 0:05:08 lr 0.000162 time 0.2849 (0.3180) loss 3.4729 (3.1647) grad_norm 2.2274 (inf) [2021-04-16 13:26:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][290/1251] eta 0:05:04 lr 0.000162 time 0.2773 (0.3165) loss 1.9775 (3.1657) grad_norm 2.1507 (inf) [2021-04-16 13:26:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][300/1251] eta 0:04:59 lr 0.000161 time 0.3135 (0.3153) loss 2.2381 (3.1680) grad_norm 2.2016 (inf) [2021-04-16 13:26:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][310/1251] eta 0:04:55 lr 0.000161 time 0.2828 (0.3140) loss 3.3651 (3.1710) grad_norm 2.1160 (inf) [2021-04-16 13:26:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][320/1251] eta 0:04:51 lr 0.000161 time 0.2769 (0.3131) loss 3.7505 (3.1744) grad_norm 2.1958 (inf) [2021-04-16 13:26:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][330/1251] eta 0:04:47 lr 0.000161 time 0.2832 (0.3119) loss 2.6600 (3.1778) grad_norm 2.1847 (inf) [2021-04-16 13:27:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][340/1251] eta 0:04:43 lr 0.000161 time 0.2906 (0.3110) loss 2.1868 (3.1734) grad_norm 1.8939 (inf) [2021-04-16 13:27:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][350/1251] eta 0:04:39 lr 0.000161 time 0.2505 (0.3099) loss 3.5606 (3.1824) grad_norm 2.3481 (inf) [2021-04-16 13:27:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][360/1251] eta 0:04:35 lr 0.000161 time 0.2821 (0.3096) loss 3.5391 (3.1783) grad_norm 2.1877 (inf) [2021-04-16 13:27:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][370/1251] eta 0:04:31 lr 0.000161 time 0.2475 (0.3087) loss 2.2993 (3.1772) grad_norm 3.4900 (inf) [2021-04-16 13:27:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][380/1251] eta 0:04:28 lr 0.000161 time 0.2564 (0.3078) loss 3.0887 (3.1840) grad_norm 2.2543 (inf) [2021-04-16 13:27:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][390/1251] eta 0:04:24 lr 0.000161 time 0.2461 (0.3069) loss 2.3036 (3.1795) grad_norm 2.0216 (inf) [2021-04-16 13:27:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][400/1251] eta 0:04:20 lr 0.000161 time 0.2726 (0.3064) loss 2.7573 (3.1773) grad_norm 2.1182 (inf) [2021-04-16 13:27:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][410/1251] eta 0:04:17 lr 0.000161 time 0.2861 (0.3056) loss 3.3412 (3.1778) grad_norm 2.1923 (inf) [2021-04-16 13:27:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][420/1251] eta 0:04:13 lr 0.000161 time 0.2615 (0.3050) loss 3.5744 (3.1786) grad_norm 2.2183 (inf) [2021-04-16 13:27:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][430/1251] eta 0:04:09 lr 0.000161 time 0.2679 (0.3043) loss 3.4920 (3.1767) grad_norm 2.3270 (inf) [2021-04-16 13:27:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][440/1251] eta 0:04:06 lr 0.000161 time 0.2714 (0.3037) loss 3.4039 (3.1745) grad_norm 2.1740 (inf) [2021-04-16 13:27:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][450/1251] eta 0:04:02 lr 0.000161 time 0.2637 (0.3032) loss 3.6493 (3.1759) grad_norm 2.1655 (inf) [2021-04-16 13:27:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][460/1251] eta 0:03:59 lr 0.000161 time 0.2642 (0.3025) loss 3.1185 (3.1799) grad_norm 2.7411 (inf) [2021-04-16 13:27:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][470/1251] eta 0:03:56 lr 0.000161 time 0.2780 (0.3023) loss 3.4949 (3.1788) grad_norm 2.0766 (inf) [2021-04-16 13:27:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][480/1251] eta 0:03:52 lr 0.000161 time 0.3006 (0.3020) loss 3.8571 (3.1829) grad_norm 2.2709 (inf) [2021-04-16 13:27:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][490/1251] eta 0:03:49 lr 0.000161 time 0.2839 (0.3015) loss 3.7026 (3.1859) grad_norm 2.3955 (inf) [2021-04-16 13:27:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][500/1251] eta 0:03:46 lr 0.000161 time 0.2764 (0.3011) loss 3.4139 (3.1863) grad_norm 2.0097 (inf) [2021-04-16 13:27:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][510/1251] eta 0:03:42 lr 0.000161 time 0.2730 (0.3006) loss 3.1126 (3.1903) grad_norm 2.7594 (inf) [2021-04-16 13:27:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][520/1251] eta 0:03:39 lr 0.000161 time 0.2648 (0.3002) loss 3.1317 (3.1902) grad_norm 2.3570 (inf) [2021-04-16 13:27:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][530/1251] eta 0:03:36 lr 0.000161 time 0.2979 (0.2998) loss 2.0530 (3.1859) grad_norm 2.6903 (inf) [2021-04-16 13:27:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][540/1251] eta 0:03:32 lr 0.000161 time 0.2644 (0.2994) loss 3.0093 (3.1893) grad_norm 2.1736 (inf) [2021-04-16 13:28:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][550/1251] eta 0:03:29 lr 0.000161 time 0.2608 (0.2989) loss 3.3965 (3.1839) grad_norm 2.3404 (inf) [2021-04-16 13:28:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][560/1251] eta 0:03:26 lr 0.000161 time 0.2715 (0.2985) loss 3.1754 (3.1872) grad_norm 2.3106 (inf) [2021-04-16 13:28:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][570/1251] eta 0:03:23 lr 0.000161 time 0.2735 (0.2981) loss 3.5795 (3.1866) grad_norm 2.0441 (inf) [2021-04-16 13:28:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][580/1251] eta 0:03:20 lr 0.000161 time 0.2896 (0.2982) loss 3.6249 (3.1859) grad_norm 2.1497 (inf) [2021-04-16 13:28:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][590/1251] eta 0:03:17 lr 0.000161 time 0.2641 (0.2981) loss 3.4624 (3.1858) grad_norm 2.4472 (inf) [2021-04-16 13:28:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][600/1251] eta 0:03:13 lr 0.000161 time 0.3741 (0.2979) loss 3.6251 (3.1909) grad_norm 2.9230 (inf) [2021-04-16 13:28:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][610/1251] eta 0:03:10 lr 0.000161 time 0.2832 (0.2976) loss 3.7623 (3.1890) grad_norm 1.9665 (inf) [2021-04-16 13:28:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][620/1251] eta 0:03:07 lr 0.000161 time 0.2669 (0.2972) loss 3.6690 (3.1884) grad_norm 2.5481 (inf) [2021-04-16 13:28:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][630/1251] eta 0:03:04 lr 0.000161 time 0.2742 (0.2969) loss 2.8216 (3.1864) grad_norm 2.1386 (inf) [2021-04-16 13:28:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][640/1251] eta 0:03:01 lr 0.000160 time 0.2837 (0.2966) loss 2.9793 (3.1881) grad_norm 2.0774 (inf) [2021-04-16 13:28:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][650/1251] eta 0:02:58 lr 0.000160 time 0.2599 (0.2963) loss 3.4238 (3.1871) grad_norm 2.2516 (inf) [2021-04-16 13:28:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][660/1251] eta 0:02:54 lr 0.000160 time 0.2708 (0.2960) loss 3.1223 (3.1836) grad_norm 2.2883 (inf) [2021-04-16 13:28:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][670/1251] eta 0:02:51 lr 0.000160 time 0.2513 (0.2957) loss 3.4805 (3.1894) grad_norm 2.1370 (inf) [2021-04-16 13:28:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][680/1251] eta 0:02:48 lr 0.000160 time 0.2819 (0.2954) loss 3.3255 (3.1888) grad_norm 2.2787 (inf) [2021-04-16 13:28:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][690/1251] eta 0:02:45 lr 0.000160 time 0.2886 (0.2952) loss 3.5148 (3.1870) grad_norm 2.3121 (inf) [2021-04-16 13:28:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][700/1251] eta 0:02:42 lr 0.000160 time 0.2819 (0.2949) loss 4.0850 (3.1911) grad_norm 2.3672 (inf) [2021-04-16 13:28:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][710/1251] eta 0:02:39 lr 0.000160 time 0.4552 (0.2949) loss 3.1023 (3.1921) grad_norm 2.4280 (inf) [2021-04-16 13:28:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][720/1251] eta 0:02:36 lr 0.000160 time 0.2866 (0.2947) loss 2.7829 (3.1899) grad_norm 2.2031 (inf) [2021-04-16 13:28:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][730/1251] eta 0:02:33 lr 0.000160 time 0.3005 (0.2945) loss 3.6130 (3.1859) grad_norm 2.1466 (inf) [2021-04-16 13:28:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][740/1251] eta 0:02:30 lr 0.000160 time 0.2567 (0.2944) loss 3.5367 (3.1845) grad_norm 2.4183 (inf) [2021-04-16 13:28:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][750/1251] eta 0:02:27 lr 0.000160 time 0.2767 (0.2941) loss 3.7013 (3.1840) grad_norm 2.1687 (inf) [2021-04-16 13:29:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][760/1251] eta 0:02:24 lr 0.000160 time 0.2762 (0.2938) loss 4.1376 (3.1877) grad_norm 3.1152 (inf) [2021-04-16 13:29:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][770/1251] eta 0:02:21 lr 0.000160 time 0.3116 (0.2936) loss 2.9648 (3.1882) grad_norm 2.0476 (inf) [2021-04-16 13:29:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][780/1251] eta 0:02:18 lr 0.000160 time 0.2669 (0.2933) loss 3.6157 (3.1870) grad_norm 2.4085 (inf) [2021-04-16 13:29:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][790/1251] eta 0:02:15 lr 0.000160 time 0.2446 (0.2931) loss 3.3827 (3.1852) grad_norm 2.2034 (inf) [2021-04-16 13:29:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][800/1251] eta 0:02:12 lr 0.000160 time 0.4256 (0.2931) loss 2.7321 (3.1865) grad_norm 2.5885 (inf) [2021-04-16 13:29:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][810/1251] eta 0:02:09 lr 0.000160 time 0.2594 (0.2929) loss 2.6084 (3.1845) grad_norm 2.6986 (inf) [2021-04-16 13:29:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][820/1251] eta 0:02:06 lr 0.000160 time 0.2555 (0.2929) loss 3.5579 (3.1830) grad_norm 2.0851 (inf) [2021-04-16 13:29:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][830/1251] eta 0:02:03 lr 0.000160 time 0.2774 (0.2928) loss 1.9789 (3.1823) grad_norm 2.1064 (inf) [2021-04-16 13:29:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][840/1251] eta 0:02:00 lr 0.000160 time 0.2414 (0.2925) loss 3.5682 (3.1815) grad_norm 2.2887 (inf) [2021-04-16 13:29:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][850/1251] eta 0:01:57 lr 0.000160 time 0.2811 (0.2923) loss 3.7116 (3.1842) grad_norm 2.2846 (inf) [2021-04-16 13:29:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][860/1251] eta 0:01:54 lr 0.000160 time 0.2714 (0.2922) loss 3.5605 (3.1824) grad_norm 2.2171 (inf) [2021-04-16 13:29:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][870/1251] eta 0:01:51 lr 0.000160 time 0.2622 (0.2920) loss 2.3255 (3.1780) grad_norm 2.0874 (inf) [2021-04-16 13:29:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][880/1251] eta 0:01:48 lr 0.000160 time 0.2891 (0.2918) loss 2.8142 (3.1794) grad_norm 2.3406 (inf) [2021-04-16 13:29:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][890/1251] eta 0:01:45 lr 0.000160 time 0.2928 (0.2918) loss 3.4949 (3.1806) grad_norm 3.3602 (inf) [2021-04-16 13:29:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][900/1251] eta 0:01:42 lr 0.000160 time 0.2979 (0.2916) loss 3.4226 (3.1818) grad_norm 1.9452 (inf) [2021-04-16 13:29:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][910/1251] eta 0:01:39 lr 0.000160 time 0.2578 (0.2915) loss 3.9055 (3.1808) grad_norm 2.9789 (inf) [2021-04-16 13:29:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][920/1251] eta 0:01:36 lr 0.000160 time 0.2753 (0.2913) loss 3.5568 (3.1849) grad_norm 2.3104 (inf) [2021-04-16 13:29:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][930/1251] eta 0:01:33 lr 0.000160 time 0.3274 (0.2912) loss 3.4101 (3.1868) grad_norm 2.7039 (inf) [2021-04-16 13:29:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][940/1251] eta 0:01:30 lr 0.000160 time 0.2883 (0.2911) loss 3.7910 (3.1859) grad_norm 2.1568 (inf) [2021-04-16 13:29:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][950/1251] eta 0:01:27 lr 0.000160 time 0.2938 (0.2909) loss 3.2470 (3.1862) grad_norm 2.6038 (inf) [2021-04-16 13:29:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][960/1251] eta 0:01:24 lr 0.000160 time 0.2588 (0.2908) loss 3.0445 (3.1860) grad_norm 2.2750 (inf) [2021-04-16 13:29:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][970/1251] eta 0:01:21 lr 0.000159 time 0.2917 (0.2907) loss 3.8444 (3.1881) grad_norm 2.3727 (inf) [2021-04-16 13:30:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][980/1251] eta 0:01:18 lr 0.000159 time 0.3233 (0.2907) loss 2.6618 (3.1852) grad_norm 2.3185 (inf) [2021-04-16 13:30:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][990/1251] eta 0:01:15 lr 0.000159 time 0.2587 (0.2906) loss 3.0411 (3.1855) grad_norm 2.6132 (inf) [2021-04-16 13:30:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][1000/1251] eta 0:01:12 lr 0.000159 time 0.2685 (0.2905) loss 3.7911 (3.1877) grad_norm 2.1318 (inf) [2021-04-16 13:30:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][1010/1251] eta 0:01:09 lr 0.000159 time 0.2930 (0.2904) loss 3.6970 (3.1875) grad_norm 2.2297 (inf) [2021-04-16 13:30:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][1020/1251] eta 0:01:07 lr 0.000159 time 0.2847 (0.2903) loss 3.3110 (3.1881) grad_norm 2.5196 (inf) [2021-04-16 13:30:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][1030/1251] eta 0:01:04 lr 0.000159 time 0.2574 (0.2901) loss 3.2302 (3.1886) grad_norm 2.2472 (inf) [2021-04-16 13:30:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][1040/1251] eta 0:01:01 lr 0.000159 time 0.2726 (0.2901) loss 3.1930 (3.1901) grad_norm 2.4571 (inf) [2021-04-16 13:30:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][1050/1251] eta 0:00:58 lr 0.000159 time 0.2903 (0.2901) loss 3.7129 (3.1895) grad_norm 2.1819 (inf) [2021-04-16 13:30:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][1060/1251] eta 0:00:55 lr 0.000159 time 0.2538 (0.2900) loss 3.7425 (3.1899) grad_norm 2.1640 (inf) [2021-04-16 13:30:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][1070/1251] eta 0:00:52 lr 0.000159 time 0.2873 (0.2899) loss 3.2087 (3.1898) grad_norm 2.2578 (inf) [2021-04-16 13:30:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][1080/1251] eta 0:00:49 lr 0.000159 time 0.2659 (0.2897) loss 3.4152 (3.1893) grad_norm 2.6154 (inf) [2021-04-16 13:30:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][1090/1251] eta 0:00:46 lr 0.000159 time 0.2684 (0.2897) loss 2.2184 (3.1896) grad_norm 2.2839 (inf) [2021-04-16 13:30:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][1100/1251] eta 0:00:43 lr 0.000159 time 0.2822 (0.2896) loss 2.3777 (3.1899) grad_norm 2.2193 (inf) [2021-04-16 13:30:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][1110/1251] eta 0:00:40 lr 0.000159 time 0.2885 (0.2895) loss 2.2943 (3.1876) grad_norm 2.5985 (inf) [2021-04-16 13:30:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][1120/1251] eta 0:00:37 lr 0.000159 time 0.2591 (0.2893) loss 3.7087 (3.1904) grad_norm 2.1044 (inf) [2021-04-16 13:30:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][1130/1251] eta 0:00:34 lr 0.000159 time 0.2676 (0.2892) loss 3.1712 (3.1914) grad_norm 2.1782 (inf) [2021-04-16 13:30:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][1140/1251] eta 0:00:32 lr 0.000159 time 0.2680 (0.2891) loss 3.6348 (3.1915) grad_norm 2.3482 (inf) [2021-04-16 13:30:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][1150/1251] eta 0:00:29 lr 0.000159 time 0.2656 (0.2891) loss 3.4469 (3.1898) grad_norm 2.5755 (inf) [2021-04-16 13:30:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][1160/1251] eta 0:00:26 lr 0.000159 time 0.2774 (0.2890) loss 3.5107 (3.1910) grad_norm 2.1075 (inf) [2021-04-16 13:30:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][1170/1251] eta 0:00:23 lr 0.000159 time 0.2429 (0.2891) loss 3.4525 (3.1903) grad_norm 2.3280 (inf) [2021-04-16 13:30:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][1180/1251] eta 0:00:20 lr 0.000159 time 0.2538 (0.2891) loss 2.9340 (3.1903) grad_norm 2.4022 (inf) [2021-04-16 13:31:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][1190/1251] eta 0:00:17 lr 0.000159 time 0.2718 (0.2890) loss 3.2044 (3.1921) grad_norm 2.8496 (inf) [2021-04-16 13:31:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][1200/1251] eta 0:00:14 lr 0.000159 time 0.2823 (0.2889) loss 3.4375 (3.1930) grad_norm 2.4461 (inf) [2021-04-16 13:31:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][1210/1251] eta 0:00:11 lr 0.000159 time 0.2944 (0.2888) loss 3.6285 (3.1944) grad_norm 2.7230 (inf) [2021-04-16 13:31:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][1220/1251] eta 0:00:08 lr 0.000159 time 0.2596 (0.2887) loss 3.6119 (3.1949) grad_norm 2.0927 (inf) [2021-04-16 13:31:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][1230/1251] eta 0:00:06 lr 0.000159 time 0.2754 (0.2886) loss 3.0073 (3.1966) grad_norm 2.4317 (inf) [2021-04-16 13:31:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][1240/1251] eta 0:00:03 lr 0.000159 time 0.2573 (0.2885) loss 2.6105 (3.1957) grad_norm 2.3221 (inf) [2021-04-16 13:31:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [223/300][1250/1251] eta 0:00:00 lr 0.000159 time 0.2484 (0.2882) loss 3.0674 (3.1953) grad_norm 2.1368 (inf) [2021-04-16 13:31:20 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 223 training takes 0:06:04 [2021-04-16 13:31:20 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_223.pth saving...... [2021-04-16 13:31:31 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_223.pth saved !!! [2021-04-16 13:31:32 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.092 (1.092) Loss 0.8705 (0.8705) Acc@1 79.102 (79.102) Acc@5 94.727 (94.727) [2021-04-16 13:31:34 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.124 (0.270) Loss 0.8592 (0.8692) Acc@1 78.125 (79.705) Acc@5 95.703 (95.099) [2021-04-16 13:31:36 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.129 (0.245) Loss 0.9918 (0.8879) Acc@1 76.758 (79.264) Acc@5 93.262 (94.792) [2021-04-16 13:31:38 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.100 (0.213) Loss 0.8271 (0.8855) Acc@1 80.957 (79.313) Acc@5 96.387 (94.875) [2021-04-16 13:31:40 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.207) Loss 0.8791 (0.8882) Acc@1 77.637 (79.240) Acc@5 95.117 (94.853) [2021-04-16 13:31:45 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 79.282 Acc@5 94.816 [2021-04-16 13:31:45 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 79.3% [2021-04-16 13:31:45 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 79.34% [2021-04-16 13:31:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][0/1251] eta 2:58:22 lr 0.000159 time 8.5551 (8.5551) loss 2.9451 (2.9451) grad_norm 1.8493 (1.8493) [2021-04-16 13:31:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][10/1251] eta 0:21:31 lr 0.000159 time 0.4589 (1.0410) loss 3.4889 (3.1197) grad_norm 2.4904 (2.2616) [2021-04-16 13:31:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][20/1251] eta 0:13:55 lr 0.000159 time 0.3179 (0.6786) loss 3.3903 (3.1994) grad_norm 2.2005 (2.2904) [2021-04-16 13:32:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][30/1251] eta 0:11:17 lr 0.000159 time 0.2611 (0.5549) loss 3.5521 (3.2083) grad_norm 2.3833 (2.3281) [2021-04-16 13:32:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][40/1251] eta 0:09:50 lr 0.000159 time 0.2785 (0.4874) loss 2.8773 (3.2040) grad_norm 2.3760 (2.3897) [2021-04-16 13:32:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][50/1251] eta 0:08:56 lr 0.000159 time 0.2682 (0.4466) loss 3.4444 (3.2418) grad_norm 2.2044 (2.3774) [2021-04-16 13:32:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][60/1251] eta 0:08:19 lr 0.000158 time 0.2814 (0.4194) loss 3.6327 (3.2514) grad_norm 3.5264 (2.4091) [2021-04-16 13:32:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][70/1251] eta 0:07:51 lr 0.000158 time 0.2806 (0.3996) loss 2.6243 (3.2348) grad_norm 2.4038 (2.4157) [2021-04-16 13:32:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][80/1251] eta 0:07:31 lr 0.000158 time 0.2827 (0.3852) loss 3.4606 (3.2556) grad_norm 2.0558 (2.4289) [2021-04-16 13:32:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][90/1251] eta 0:07:13 lr 0.000158 time 0.2960 (0.3732) loss 3.5453 (3.2387) grad_norm 2.3809 (2.4142) [2021-04-16 13:32:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][100/1251] eta 0:07:00 lr 0.000158 time 0.2740 (0.3652) loss 3.0449 (3.2676) grad_norm 2.2877 (2.3974) [2021-04-16 13:32:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][110/1251] eta 0:06:48 lr 0.000158 time 0.2873 (0.3577) loss 3.6338 (3.2555) grad_norm 2.0551 (2.3786) [2021-04-16 13:32:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][120/1251] eta 0:06:38 lr 0.000158 time 0.2836 (0.3520) loss 2.0957 (3.2345) grad_norm 2.4744 (2.3767) [2021-04-16 13:32:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][130/1251] eta 0:06:28 lr 0.000158 time 0.2892 (0.3464) loss 2.3354 (3.2342) grad_norm 2.6208 (2.3708) [2021-04-16 13:32:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][140/1251] eta 0:06:19 lr 0.000158 time 0.2847 (0.3414) loss 3.5830 (3.2492) grad_norm 2.1961 (2.3696) [2021-04-16 13:32:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][150/1251] eta 0:06:12 lr 0.000158 time 0.2722 (0.3380) loss 3.3981 (3.2481) grad_norm 2.1743 (2.3730) [2021-04-16 13:32:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][160/1251] eta 0:06:04 lr 0.000158 time 0.2590 (0.3342) loss 3.3838 (3.2354) grad_norm 2.0914 (2.3634) [2021-04-16 13:32:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][170/1251] eta 0:05:58 lr 0.000158 time 0.2820 (0.3314) loss 3.2729 (3.2502) grad_norm 2.2084 (2.3549) [2021-04-16 13:32:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][180/1251] eta 0:05:52 lr 0.000158 time 0.4259 (0.3292) loss 2.7895 (3.2450) grad_norm 2.3386 (2.3571) [2021-04-16 13:32:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][190/1251] eta 0:05:46 lr 0.000158 time 0.2942 (0.3264) loss 3.2862 (3.2328) grad_norm 2.1962 (2.3570) [2021-04-16 13:32:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][200/1251] eta 0:05:40 lr 0.000158 time 0.2779 (0.3240) loss 2.3278 (3.2214) grad_norm 2.4921 (2.3579) [2021-04-16 13:32:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][210/1251] eta 0:05:34 lr 0.000158 time 0.2616 (0.3217) loss 2.9438 (3.2181) grad_norm 2.6065 (2.3667) [2021-04-16 13:32:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][220/1251] eta 0:05:29 lr 0.000158 time 0.3131 (0.3200) loss 3.5480 (3.2101) grad_norm 2.4140 (2.3723) [2021-04-16 13:32:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][230/1251] eta 0:05:24 lr 0.000158 time 0.2555 (0.3180) loss 2.5940 (3.2074) grad_norm 2.2621 (2.3735) [2021-04-16 13:33:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][240/1251] eta 0:05:19 lr 0.000158 time 0.2874 (0.3163) loss 3.5452 (3.1978) grad_norm 2.5341 (2.3725) [2021-04-16 13:33:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][250/1251] eta 0:05:15 lr 0.000158 time 0.2586 (0.3152) loss 2.6991 (3.2025) grad_norm 2.6073 (2.3735) [2021-04-16 13:33:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][260/1251] eta 0:05:10 lr 0.000158 time 0.2782 (0.3137) loss 3.1391 (3.2118) grad_norm 2.2641 (2.3763) [2021-04-16 13:33:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][270/1251] eta 0:05:06 lr 0.000158 time 0.2787 (0.3125) loss 2.2716 (3.2146) grad_norm 2.6139 (2.3744) [2021-04-16 13:33:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][280/1251] eta 0:05:02 lr 0.000158 time 0.2684 (0.3113) loss 3.7270 (3.2090) grad_norm 2.1432 (2.3737) [2021-04-16 13:33:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][290/1251] eta 0:04:58 lr 0.000158 time 0.2853 (0.3102) loss 3.0203 (3.2008) grad_norm 2.5387 (2.3750) [2021-04-16 13:33:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][300/1251] eta 0:04:54 lr 0.000158 time 0.2693 (0.3097) loss 3.3597 (3.1999) grad_norm 2.2290 (2.3737) [2021-04-16 13:33:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][310/1251] eta 0:04:50 lr 0.000158 time 0.2727 (0.3087) loss 3.0785 (3.2062) grad_norm 2.0983 (2.3692) [2021-04-16 13:33:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][320/1251] eta 0:04:46 lr 0.000158 time 0.2482 (0.3076) loss 2.8006 (3.2060) grad_norm 2.3095 (2.3642) [2021-04-16 13:33:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][330/1251] eta 0:04:42 lr 0.000158 time 0.2639 (0.3068) loss 3.1926 (3.2025) grad_norm 2.2771 (2.3629) [2021-04-16 13:33:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][340/1251] eta 0:04:38 lr 0.000158 time 0.2555 (0.3060) loss 3.3003 (3.1973) grad_norm 2.4481 (2.3589) [2021-04-16 13:33:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][350/1251] eta 0:04:35 lr 0.000158 time 0.2830 (0.3053) loss 3.3247 (3.1993) grad_norm 2.1547 (2.3563) [2021-04-16 13:33:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][360/1251] eta 0:04:31 lr 0.000158 time 0.2807 (0.3048) loss 3.4025 (3.1959) grad_norm 2.2373 (2.3527) [2021-04-16 13:33:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][370/1251] eta 0:04:28 lr 0.000158 time 0.2879 (0.3048) loss 3.9196 (3.1947) grad_norm 2.2932 (2.3523) [2021-04-16 13:33:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][380/1251] eta 0:04:25 lr 0.000158 time 0.2636 (0.3043) loss 2.5934 (3.1965) grad_norm 2.1267 (2.3477) [2021-04-16 13:33:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][390/1251] eta 0:04:21 lr 0.000158 time 0.2880 (0.3038) loss 3.4538 (3.1932) grad_norm 2.5676 (2.3461) [2021-04-16 13:33:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][400/1251] eta 0:04:18 lr 0.000157 time 0.2894 (0.3034) loss 3.4906 (3.1932) grad_norm 2.0804 (2.3439) [2021-04-16 13:33:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][410/1251] eta 0:04:14 lr 0.000157 time 0.2503 (0.3027) loss 3.2435 (3.2005) grad_norm 2.2002 (2.3396) [2021-04-16 13:33:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][420/1251] eta 0:04:11 lr 0.000157 time 0.2925 (0.3023) loss 3.5209 (3.1972) grad_norm 2.0511 (2.3398) [2021-04-16 13:33:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][430/1251] eta 0:04:07 lr 0.000157 time 0.2811 (0.3017) loss 3.6120 (3.1964) grad_norm 2.3773 (2.3414) [2021-04-16 13:33:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][440/1251] eta 0:04:04 lr 0.000157 time 0.2802 (0.3010) loss 3.7610 (3.1948) grad_norm 2.1990 (2.3431) [2021-04-16 13:34:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][450/1251] eta 0:04:00 lr 0.000157 time 0.2884 (0.3004) loss 3.8155 (3.1944) grad_norm 1.9752 (2.3398) [2021-04-16 13:34:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][460/1251] eta 0:03:57 lr 0.000157 time 0.2602 (0.3001) loss 2.8662 (3.1958) grad_norm 2.1312 (2.3374) [2021-04-16 13:34:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][470/1251] eta 0:03:54 lr 0.000157 time 0.2886 (0.2998) loss 3.1983 (3.1926) grad_norm 2.2792 (2.3364) [2021-04-16 13:34:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][480/1251] eta 0:03:50 lr 0.000157 time 0.2624 (0.2993) loss 3.3499 (3.1861) grad_norm 2.3977 (2.3364) [2021-04-16 13:34:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][490/1251] eta 0:03:47 lr 0.000157 time 0.2897 (0.2989) loss 4.1487 (3.1878) grad_norm 2.5106 (2.3371) [2021-04-16 13:34:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][500/1251] eta 0:03:44 lr 0.000157 time 0.2789 (0.2985) loss 3.5518 (3.1933) grad_norm 2.6794 (2.3358) [2021-04-16 13:34:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][510/1251] eta 0:03:40 lr 0.000157 time 0.2745 (0.2981) loss 3.5477 (3.2008) grad_norm 2.4531 (2.3368) [2021-04-16 13:34:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][520/1251] eta 0:03:37 lr 0.000157 time 0.2883 (0.2977) loss 3.8531 (3.1936) grad_norm 2.3385 (2.3346) [2021-04-16 13:34:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][530/1251] eta 0:03:34 lr 0.000157 time 0.2896 (0.2974) loss 2.8474 (3.1936) grad_norm 2.3468 (2.3324) [2021-04-16 13:34:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][540/1251] eta 0:03:31 lr 0.000157 time 0.2991 (0.2972) loss 3.4800 (3.1940) grad_norm 2.3610 (2.3303) [2021-04-16 13:34:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][550/1251] eta 0:03:28 lr 0.000157 time 0.2596 (0.2969) loss 3.7097 (3.1907) grad_norm 2.2442 (2.3303) [2021-04-16 13:34:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][560/1251] eta 0:03:24 lr 0.000157 time 0.2950 (0.2967) loss 2.0144 (3.1853) grad_norm 2.1468 (2.3332) [2021-04-16 13:34:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][570/1251] eta 0:03:22 lr 0.000157 time 0.2957 (0.2967) loss 3.9763 (3.1875) grad_norm 2.1091 (2.3313) [2021-04-16 13:34:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][580/1251] eta 0:03:19 lr 0.000157 time 0.2817 (0.2966) loss 3.4977 (3.1896) grad_norm 2.6944 (2.3324) [2021-04-16 13:34:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][590/1251] eta 0:03:16 lr 0.000157 time 0.2800 (0.2966) loss 3.1613 (3.1879) grad_norm 2.2438 (2.3319) [2021-04-16 13:34:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][600/1251] eta 0:03:13 lr 0.000157 time 0.4375 (0.2965) loss 2.1041 (3.1788) grad_norm 2.0097 (2.3299) [2021-04-16 13:34:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][610/1251] eta 0:03:09 lr 0.000157 time 0.2788 (0.2962) loss 3.7744 (3.1823) grad_norm 2.2851 (2.3316) [2021-04-16 13:34:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][620/1251] eta 0:03:06 lr 0.000157 time 0.2666 (0.2959) loss 3.4847 (3.1830) grad_norm 1.9664 (2.3289) [2021-04-16 13:34:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][630/1251] eta 0:03:03 lr 0.000157 time 0.2902 (0.2958) loss 3.5057 (3.1821) grad_norm 2.1637 (2.3285) [2021-04-16 13:34:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][640/1251] eta 0:03:00 lr 0.000157 time 0.2770 (0.2956) loss 3.6612 (3.1889) grad_norm 2.1444 (2.3271) [2021-04-16 13:34:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][650/1251] eta 0:02:57 lr 0.000157 time 0.2626 (0.2953) loss 3.1497 (3.1891) grad_norm 2.1900 (2.3264) [2021-04-16 13:35:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][660/1251] eta 0:02:54 lr 0.000157 time 0.2719 (0.2949) loss 3.4380 (3.1897) grad_norm 2.6161 (2.3271) [2021-04-16 13:35:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][670/1251] eta 0:02:51 lr 0.000157 time 0.2728 (0.2947) loss 3.6525 (3.1900) grad_norm 2.4505 (2.3249) [2021-04-16 13:35:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][680/1251] eta 0:02:48 lr 0.000157 time 0.2712 (0.2946) loss 3.8437 (3.1835) grad_norm 2.3158 (2.3231) [2021-04-16 13:35:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][690/1251] eta 0:02:45 lr 0.000157 time 0.2787 (0.2943) loss 3.1966 (3.1850) grad_norm 2.1132 (2.3234) [2021-04-16 13:35:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][700/1251] eta 0:02:42 lr 0.000157 time 0.3027 (0.2941) loss 2.8114 (3.1904) grad_norm 2.1295 (2.3224) [2021-04-16 13:35:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][710/1251] eta 0:02:38 lr 0.000157 time 0.2783 (0.2939) loss 2.9778 (3.1918) grad_norm 2.0618 (2.3201) [2021-04-16 13:35:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][720/1251] eta 0:02:35 lr 0.000157 time 0.2710 (0.2937) loss 2.1919 (3.1947) grad_norm 2.2612 (2.3218) [2021-04-16 13:35:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][730/1251] eta 0:02:32 lr 0.000157 time 0.2687 (0.2934) loss 2.8261 (3.1936) grad_norm 2.5270 (2.3237) [2021-04-16 13:35:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][740/1251] eta 0:02:29 lr 0.000156 time 0.2641 (0.2933) loss 3.7864 (3.1951) grad_norm 2.7253 (2.3234) [2021-04-16 13:35:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][750/1251] eta 0:02:26 lr 0.000156 time 0.2809 (0.2931) loss 3.0001 (3.1998) grad_norm 2.2509 (2.3230) [2021-04-16 13:35:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][760/1251] eta 0:02:23 lr 0.000156 time 0.2547 (0.2930) loss 3.6761 (3.1995) grad_norm 2.3493 (2.3221) [2021-04-16 13:35:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][770/1251] eta 0:02:20 lr 0.000156 time 0.2670 (0.2928) loss 2.8533 (3.1975) grad_norm 2.1193 (2.3208) [2021-04-16 13:35:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][780/1251] eta 0:02:17 lr 0.000156 time 0.2865 (0.2927) loss 3.4997 (3.1960) grad_norm 2.3625 (2.3189) [2021-04-16 13:35:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][790/1251] eta 0:02:14 lr 0.000156 time 0.2736 (0.2924) loss 2.8756 (3.1954) grad_norm 2.6534 (2.3200) [2021-04-16 13:35:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][800/1251] eta 0:02:11 lr 0.000156 time 0.2738 (0.2922) loss 2.7995 (3.1967) grad_norm 2.3774 (2.3215) [2021-04-16 13:35:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][810/1251] eta 0:02:08 lr 0.000156 time 0.2738 (0.2920) loss 3.0039 (3.1973) grad_norm 2.2727 (2.3210) [2021-04-16 13:35:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][820/1251] eta 0:02:05 lr 0.000156 time 0.2766 (0.2918) loss 3.2329 (3.1984) grad_norm 2.2221 (2.3211) [2021-04-16 13:35:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][830/1251] eta 0:02:02 lr 0.000156 time 0.2811 (0.2917) loss 3.3397 (3.1976) grad_norm 2.4342 (2.3212) [2021-04-16 13:35:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][840/1251] eta 0:01:59 lr 0.000156 time 0.2558 (0.2916) loss 3.1780 (3.1956) grad_norm 1.9621 (2.3216) [2021-04-16 13:35:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][850/1251] eta 0:01:56 lr 0.000156 time 0.2852 (0.2915) loss 2.5037 (3.1978) grad_norm 2.2131 (2.3215) [2021-04-16 13:35:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][860/1251] eta 0:01:53 lr 0.000156 time 0.2851 (0.2913) loss 3.2785 (3.1979) grad_norm 2.6288 (2.3229) [2021-04-16 13:35:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][870/1251] eta 0:01:50 lr 0.000156 time 0.3009 (0.2912) loss 3.3172 (3.1987) grad_norm 2.1441 (2.3233) [2021-04-16 13:36:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][880/1251] eta 0:01:48 lr 0.000156 time 0.2540 (0.2912) loss 3.3547 (3.1982) grad_norm 2.4592 (2.3250) [2021-04-16 13:36:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][890/1251] eta 0:01:45 lr 0.000156 time 0.2875 (0.2911) loss 2.1500 (3.1962) grad_norm 1.9737 (2.3250) [2021-04-16 13:36:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][900/1251] eta 0:01:42 lr 0.000156 time 0.3320 (0.2910) loss 4.1709 (3.1986) grad_norm 2.5979 (2.3259) [2021-04-16 13:36:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][910/1251] eta 0:01:39 lr 0.000156 time 0.2953 (0.2909) loss 3.4248 (3.2006) grad_norm 2.3150 (2.3257) [2021-04-16 13:36:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][920/1251] eta 0:01:36 lr 0.000156 time 0.2799 (0.2908) loss 3.1663 (3.2013) grad_norm 2.3641 (2.3255) [2021-04-16 13:36:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][930/1251] eta 0:01:33 lr 0.000156 time 0.2765 (0.2907) loss 3.2595 (3.2020) grad_norm 2.5477 (2.3254) [2021-04-16 13:36:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][940/1251] eta 0:01:30 lr 0.000156 time 0.3608 (0.2909) loss 3.1736 (3.2023) grad_norm 2.1589 (2.3265) [2021-04-16 13:36:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][950/1251] eta 0:01:27 lr 0.000156 time 0.2972 (0.2909) loss 3.3336 (3.2022) grad_norm 2.6718 (2.3305) [2021-04-16 13:36:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][960/1251] eta 0:01:24 lr 0.000156 time 0.2671 (0.2909) loss 3.5012 (3.2034) grad_norm 2.3154 (2.3337) [2021-04-16 13:36:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][970/1251] eta 0:01:21 lr 0.000156 time 0.2910 (0.2908) loss 3.4435 (3.2054) grad_norm 2.0881 (2.3347) [2021-04-16 13:36:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][980/1251] eta 0:01:18 lr 0.000156 time 0.2863 (0.2906) loss 3.3991 (3.2059) grad_norm 2.4333 (2.3339) [2021-04-16 13:36:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][990/1251] eta 0:01:15 lr 0.000156 time 0.2932 (0.2906) loss 3.7733 (3.2061) grad_norm 2.1274 (2.3329) [2021-04-16 13:36:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][1000/1251] eta 0:01:12 lr 0.000156 time 0.2745 (0.2904) loss 3.7929 (3.2085) grad_norm 1.9921 (2.3324) [2021-04-16 13:36:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][1010/1251] eta 0:01:09 lr 0.000156 time 0.2585 (0.2903) loss 2.6302 (3.2095) grad_norm 2.2969 (2.3316) [2021-04-16 13:36:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][1020/1251] eta 0:01:07 lr 0.000156 time 0.2782 (0.2902) loss 3.2242 (3.2126) grad_norm 2.2716 (2.3311) [2021-04-16 13:36:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][1030/1251] eta 0:01:04 lr 0.000156 time 0.2769 (0.2900) loss 3.1412 (3.2121) grad_norm 2.0928 (2.3312) [2021-04-16 13:36:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][1040/1251] eta 0:01:01 lr 0.000156 time 0.2770 (0.2899) loss 3.8224 (3.2139) grad_norm 2.2313 (2.3307) [2021-04-16 13:36:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][1050/1251] eta 0:00:58 lr 0.000156 time 0.2753 (0.2897) loss 3.4104 (3.2128) grad_norm 2.4667 (2.3319) [2021-04-16 13:36:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][1060/1251] eta 0:00:55 lr 0.000156 time 0.2879 (0.2896) loss 2.5095 (3.2165) grad_norm 2.1374 (2.3321) [2021-04-16 13:36:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][1070/1251] eta 0:00:52 lr 0.000156 time 0.2759 (0.2896) loss 3.7124 (3.2158) grad_norm 2.3949 (2.3313) [2021-04-16 13:36:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][1080/1251] eta 0:00:49 lr 0.000155 time 0.2858 (0.2895) loss 3.7640 (3.2159) grad_norm 2.5489 (2.3310) [2021-04-16 13:37:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][1090/1251] eta 0:00:46 lr 0.000155 time 0.2936 (0.2894) loss 2.9260 (3.2162) grad_norm 3.3801 (2.3315) [2021-04-16 13:37:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][1100/1251] eta 0:00:43 lr 0.000155 time 0.2816 (0.2894) loss 2.4982 (3.2163) grad_norm 2.0129 (2.3319) [2021-04-16 13:37:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][1110/1251] eta 0:00:40 lr 0.000155 time 0.2759 (0.2892) loss 2.8981 (3.2158) grad_norm 2.1681 (2.3323) [2021-04-16 13:37:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][1120/1251] eta 0:00:37 lr 0.000155 time 0.2767 (0.2891) loss 3.9308 (3.2162) grad_norm 2.1960 (2.3337) [2021-04-16 13:37:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][1130/1251] eta 0:00:34 lr 0.000155 time 0.2694 (0.2890) loss 3.3009 (3.2167) grad_norm 2.3128 (2.3346) [2021-04-16 13:37:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][1140/1251] eta 0:00:32 lr 0.000155 time 0.3002 (0.2890) loss 3.3760 (3.2149) grad_norm 2.0134 (nan) [2021-04-16 13:37:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][1150/1251] eta 0:00:29 lr 0.000155 time 0.2621 (0.2890) loss 3.0684 (3.2135) grad_norm 2.4393 (nan) [2021-04-16 13:37:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][1160/1251] eta 0:00:26 lr 0.000155 time 0.2951 (0.2890) loss 3.5888 (3.2143) grad_norm 2.3542 (nan) [2021-04-16 13:37:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][1170/1251] eta 0:00:23 lr 0.000155 time 0.2618 (0.2890) loss 3.1224 (3.2139) grad_norm 2.2890 (nan) [2021-04-16 13:37:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][1180/1251] eta 0:00:20 lr 0.000155 time 0.2889 (0.2891) loss 3.2297 (3.2144) grad_norm 2.1173 (nan) [2021-04-16 13:37:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][1190/1251] eta 0:00:17 lr 0.000155 time 0.3011 (0.2891) loss 2.2523 (3.2143) grad_norm 2.3616 (nan) [2021-04-16 13:37:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][1200/1251] eta 0:00:14 lr 0.000155 time 0.2605 (0.2889) loss 3.1768 (3.2147) grad_norm 2.2738 (nan) [2021-04-16 13:37:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][1210/1251] eta 0:00:11 lr 0.000155 time 0.2818 (0.2888) loss 2.2873 (3.2124) grad_norm 2.1314 (nan) [2021-04-16 13:37:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][1220/1251] eta 0:00:08 lr 0.000155 time 0.2692 (0.2888) loss 3.1471 (3.2132) grad_norm 2.4097 (nan) [2021-04-16 13:37:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][1230/1251] eta 0:00:06 lr 0.000155 time 0.2576 (0.2887) loss 3.1131 (3.2122) grad_norm 2.4438 (nan) [2021-04-16 13:37:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][1240/1251] eta 0:00:03 lr 0.000155 time 0.2485 (0.2885) loss 3.2991 (3.2118) grad_norm 2.0237 (nan) [2021-04-16 13:37:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [224/300][1250/1251] eta 0:00:00 lr 0.000155 time 0.2514 (0.2882) loss 2.5964 (3.2102) grad_norm 2.6117 (nan) [2021-04-16 13:37:49 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 224 training takes 0:06:03 [2021-04-16 13:37:49 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_224.pth saving...... [2021-04-16 13:38:00 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_224.pth saved !!! [2021-04-16 13:38:01 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.185 (1.185) Loss 0.8297 (0.8297) Acc@1 80.371 (80.371) Acc@5 95.605 (95.605) [2021-04-16 13:38:03 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.270 (0.301) Loss 0.9704 (0.8600) Acc@1 76.660 (79.332) Acc@5 93.359 (95.064) [2021-04-16 13:38:04 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.190 (0.213) Loss 0.8342 (0.8672) Acc@1 79.395 (79.404) Acc@5 95.312 (94.922) [2021-04-16 13:38:06 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.094 (0.213) Loss 0.8553 (0.8682) Acc@1 80.273 (79.442) Acc@5 94.434 (94.808) [2021-04-16 13:38:09 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.216) Loss 0.8638 (0.8667) Acc@1 79.004 (79.506) Acc@5 95.117 (94.848) [2021-04-16 13:38:16 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 79.504 Acc@5 94.884 [2021-04-16 13:38:16 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 79.5% [2021-04-16 13:38:16 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 79.50% [2021-04-16 13:38:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][0/1251] eta 3:10:18 lr 0.000155 time 9.1276 (9.1276) loss 3.6026 (3.6026) grad_norm 2.3056 (2.3056) [2021-04-16 13:38:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][10/1251] eta 0:22:17 lr 0.000155 time 0.2701 (1.0775) loss 2.2464 (3.0632) grad_norm 2.5180 (2.3926) [2021-04-16 13:38:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][20/1251] eta 0:14:19 lr 0.000155 time 0.2660 (0.6979) loss 2.3809 (3.1688) grad_norm 2.0828 (2.3543) [2021-04-16 13:38:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][30/1251] eta 0:11:30 lr 0.000155 time 0.2780 (0.5655) loss 2.7662 (3.2281) grad_norm 2.3120 (2.3742) [2021-04-16 13:38:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][40/1251] eta 0:09:58 lr 0.000155 time 0.2891 (0.4945) loss 3.6717 (3.2018) grad_norm 2.2573 (2.3710) [2021-04-16 13:38:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][50/1251] eta 0:09:02 lr 0.000155 time 0.2835 (0.4519) loss 2.8355 (3.1537) grad_norm 2.0995 (2.3576) [2021-04-16 13:38:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][60/1251] eta 0:08:24 lr 0.000155 time 0.2837 (0.4235) loss 2.9827 (3.1497) grad_norm 2.5056 (2.3570) [2021-04-16 13:38:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][70/1251] eta 0:07:56 lr 0.000155 time 0.3107 (0.4031) loss 3.8811 (3.1904) grad_norm 2.8557 (2.3708) [2021-04-16 13:38:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][80/1251] eta 0:07:35 lr 0.000155 time 0.2546 (0.3889) loss 4.0630 (3.1917) grad_norm 2.6311 (2.3663) [2021-04-16 13:38:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][90/1251] eta 0:07:16 lr 0.000155 time 0.2749 (0.3763) loss 3.5846 (3.1483) grad_norm 2.6715 (2.3640) [2021-04-16 13:38:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][100/1251] eta 0:07:01 lr 0.000155 time 0.2773 (0.3664) loss 2.3126 (3.1429) grad_norm 1.9402 (2.3551) [2021-04-16 13:38:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][110/1251] eta 0:06:49 lr 0.000155 time 0.2705 (0.3585) loss 3.7660 (3.1416) grad_norm 2.4324 (2.3521) [2021-04-16 13:38:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][120/1251] eta 0:06:37 lr 0.000155 time 0.2612 (0.3519) loss 3.7477 (3.1415) grad_norm 2.3757 (2.3645) [2021-04-16 13:39:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][130/1251] eta 0:06:28 lr 0.000155 time 0.2938 (0.3463) loss 3.4560 (3.1487) grad_norm 2.2093 (2.3584) [2021-04-16 13:39:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][140/1251] eta 0:06:20 lr 0.000155 time 0.2594 (0.3423) loss 2.4561 (3.1539) grad_norm 2.1465 (2.3494) [2021-04-16 13:39:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][150/1251] eta 0:06:14 lr 0.000155 time 0.4682 (0.3406) loss 3.4684 (3.1632) grad_norm 2.3496 (2.3406) [2021-04-16 13:39:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][160/1251] eta 0:06:06 lr 0.000155 time 0.2606 (0.3362) loss 2.0752 (3.1610) grad_norm 2.2240 (2.3418) [2021-04-16 13:39:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][170/1251] eta 0:06:00 lr 0.000154 time 0.2792 (0.3338) loss 3.1489 (3.1657) grad_norm 2.3439 (2.3475) [2021-04-16 13:39:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][180/1251] eta 0:05:53 lr 0.000154 time 0.2558 (0.3303) loss 3.2369 (3.1763) grad_norm 2.1226 (2.3408) [2021-04-16 13:39:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][190/1251] eta 0:05:47 lr 0.000154 time 0.2840 (0.3276) loss 3.3911 (3.1852) grad_norm 2.6931 (2.3546) [2021-04-16 13:39:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][200/1251] eta 0:05:41 lr 0.000154 time 0.2731 (0.3252) loss 3.1453 (3.1877) grad_norm 2.2033 (2.3493) [2021-04-16 13:39:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][210/1251] eta 0:05:36 lr 0.000154 time 0.2780 (0.3229) loss 3.8911 (3.2079) grad_norm 2.2019 (2.3425) [2021-04-16 13:39:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][220/1251] eta 0:05:30 lr 0.000154 time 0.2800 (0.3206) loss 2.7378 (3.2028) grad_norm 3.1625 (2.3445) [2021-04-16 13:39:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][230/1251] eta 0:05:25 lr 0.000154 time 0.2818 (0.3191) loss 3.9141 (3.2196) grad_norm 2.5700 (2.3521) [2021-04-16 13:39:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][240/1251] eta 0:05:20 lr 0.000154 time 0.2623 (0.3173) loss 3.1807 (3.2177) grad_norm 2.4982 (2.3523) [2021-04-16 13:39:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][250/1251] eta 0:05:15 lr 0.000154 time 0.2489 (0.3157) loss 3.5704 (3.2184) grad_norm 2.2163 (2.3487) [2021-04-16 13:39:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][260/1251] eta 0:05:11 lr 0.000154 time 0.2988 (0.3144) loss 3.6804 (3.2192) grad_norm 3.1129 (2.3516) [2021-04-16 13:39:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][270/1251] eta 0:05:06 lr 0.000154 time 0.2893 (0.3128) loss 2.5652 (3.2088) grad_norm 2.2949 (2.3472) [2021-04-16 13:39:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][280/1251] eta 0:05:02 lr 0.000154 time 0.2672 (0.3115) loss 3.6575 (3.2068) grad_norm 2.2774 (2.3452) [2021-04-16 13:39:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][290/1251] eta 0:04:58 lr 0.000154 time 0.2880 (0.3107) loss 2.7057 (3.2167) grad_norm 2.8428 (2.3485) [2021-04-16 13:39:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][300/1251] eta 0:04:54 lr 0.000154 time 0.2848 (0.3098) loss 3.6571 (3.2141) grad_norm 2.2768 (2.3455) [2021-04-16 13:39:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][310/1251] eta 0:04:50 lr 0.000154 time 0.2788 (0.3089) loss 3.9955 (3.2221) grad_norm 2.1087 (2.3500) [2021-04-16 13:39:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][320/1251] eta 0:04:46 lr 0.000154 time 0.2572 (0.3079) loss 3.1713 (3.2151) grad_norm 2.1913 (2.3542) [2021-04-16 13:39:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][330/1251] eta 0:04:42 lr 0.000154 time 0.2512 (0.3070) loss 3.5882 (3.2249) grad_norm 2.2903 (2.3499) [2021-04-16 13:40:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][340/1251] eta 0:04:38 lr 0.000154 time 0.2444 (0.3061) loss 3.1677 (3.2230) grad_norm 1.9984 (2.3495) [2021-04-16 13:40:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][350/1251] eta 0:04:35 lr 0.000154 time 0.2782 (0.3057) loss 3.6411 (3.2237) grad_norm 2.3405 (2.3481) [2021-04-16 13:40:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][360/1251] eta 0:04:31 lr 0.000154 time 0.2605 (0.3052) loss 3.4343 (3.2272) grad_norm 3.0771 (2.3510) [2021-04-16 13:40:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][370/1251] eta 0:04:28 lr 0.000154 time 0.2671 (0.3049) loss 3.4414 (3.2270) grad_norm 2.2563 (2.3550) [2021-04-16 13:40:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][380/1251] eta 0:04:25 lr 0.000154 time 0.2647 (0.3047) loss 3.7173 (3.2238) grad_norm 2.3167 (2.3590) [2021-04-16 13:40:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][390/1251] eta 0:04:21 lr 0.000154 time 0.2775 (0.3041) loss 4.2329 (3.2237) grad_norm 2.1034 (2.3602) [2021-04-16 13:40:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][400/1251] eta 0:04:18 lr 0.000154 time 0.3060 (0.3036) loss 3.2055 (3.2255) grad_norm 2.1685 (2.3593) [2021-04-16 13:40:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][410/1251] eta 0:04:14 lr 0.000154 time 0.2449 (0.3030) loss 2.1626 (3.2263) grad_norm 2.5070 (2.3585) [2021-04-16 13:40:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][420/1251] eta 0:04:11 lr 0.000154 time 0.2851 (0.3025) loss 3.3712 (3.2273) grad_norm 2.7831 (2.3604) [2021-04-16 13:40:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][430/1251] eta 0:04:07 lr 0.000154 time 0.2780 (0.3020) loss 3.6476 (3.2318) grad_norm 2.4182 (2.3604) [2021-04-16 13:40:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][440/1251] eta 0:04:04 lr 0.000154 time 0.2645 (0.3014) loss 3.0419 (3.2223) grad_norm 2.4861 (2.3604) [2021-04-16 13:40:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][450/1251] eta 0:04:00 lr 0.000154 time 0.2646 (0.3008) loss 3.0966 (3.2124) grad_norm 3.5381 (2.3608) [2021-04-16 13:40:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][460/1251] eta 0:03:57 lr 0.000154 time 0.2720 (0.3003) loss 2.1815 (3.2114) grad_norm 2.5042 (2.3611) [2021-04-16 13:40:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][470/1251] eta 0:03:54 lr 0.000154 time 0.2953 (0.3001) loss 3.1809 (3.2074) grad_norm 2.3590 (2.3604) [2021-04-16 13:40:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][480/1251] eta 0:03:50 lr 0.000154 time 0.2669 (0.2995) loss 2.4614 (3.2034) grad_norm 3.7681 (2.3614) [2021-04-16 13:40:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][490/1251] eta 0:03:47 lr 0.000154 time 0.2922 (0.2990) loss 3.5353 (3.2046) grad_norm 2.2383 (2.3601) [2021-04-16 13:40:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][500/1251] eta 0:03:44 lr 0.000154 time 0.2702 (0.2985) loss 3.5427 (3.2046) grad_norm 2.3550 (2.3595) [2021-04-16 13:40:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][510/1251] eta 0:03:40 lr 0.000153 time 0.2714 (0.2980) loss 3.1279 (3.2035) grad_norm 2.6949 (2.3600) [2021-04-16 13:40:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][520/1251] eta 0:03:37 lr 0.000153 time 0.2723 (0.2976) loss 3.6009 (3.2084) grad_norm 2.3412 (2.3590) [2021-04-16 13:40:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][530/1251] eta 0:03:34 lr 0.000153 time 0.2924 (0.2974) loss 3.6987 (3.2033) grad_norm 2.4845 (2.3575) [2021-04-16 13:40:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][540/1251] eta 0:03:31 lr 0.000153 time 0.2856 (0.2971) loss 3.0921 (3.2046) grad_norm 2.8959 (2.3571) [2021-04-16 13:41:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][550/1251] eta 0:03:27 lr 0.000153 time 0.2591 (0.2967) loss 2.9592 (3.2059) grad_norm 2.1613 (2.3570) [2021-04-16 13:41:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][560/1251] eta 0:03:24 lr 0.000153 time 0.2768 (0.2964) loss 3.2124 (3.2075) grad_norm 2.6417 (2.3587) [2021-04-16 13:41:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][570/1251] eta 0:03:21 lr 0.000153 time 0.2939 (0.2961) loss 3.5571 (3.2059) grad_norm 2.5249 (2.3600) [2021-04-16 13:41:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][580/1251] eta 0:03:18 lr 0.000153 time 0.2940 (0.2960) loss 3.5276 (3.2057) grad_norm 2.5509 (2.3604) [2021-04-16 13:41:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][590/1251] eta 0:03:15 lr 0.000153 time 0.2576 (0.2958) loss 2.7969 (3.2075) grad_norm 2.3706 (2.3587) [2021-04-16 13:41:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][600/1251] eta 0:03:12 lr 0.000153 time 0.3860 (0.2956) loss 3.5351 (3.2006) grad_norm 2.1719 (2.3595) [2021-04-16 13:41:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][610/1251] eta 0:03:09 lr 0.000153 time 0.2782 (0.2952) loss 3.3503 (3.2003) grad_norm 2.4251 (2.3606) [2021-04-16 13:41:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][620/1251] eta 0:03:06 lr 0.000153 time 0.2651 (0.2949) loss 3.1440 (3.2049) grad_norm 2.0888 (2.3594) [2021-04-16 13:41:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][630/1251] eta 0:03:02 lr 0.000153 time 0.2919 (0.2946) loss 2.7326 (3.2067) grad_norm 2.0494 (2.3635) [2021-04-16 13:41:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][640/1251] eta 0:02:59 lr 0.000153 time 0.2596 (0.2944) loss 3.4998 (3.2083) grad_norm 2.2141 (2.3620) [2021-04-16 13:41:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][650/1251] eta 0:02:56 lr 0.000153 time 0.4822 (0.2944) loss 2.7423 (3.2126) grad_norm 2.2169 (2.3633) [2021-04-16 13:41:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][660/1251] eta 0:02:53 lr 0.000153 time 0.2717 (0.2940) loss 2.2297 (3.2135) grad_norm 2.5874 (2.3637) [2021-04-16 13:41:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][670/1251] eta 0:02:50 lr 0.000153 time 0.2659 (0.2937) loss 2.8910 (3.2128) grad_norm 2.2950 (2.3664) [2021-04-16 13:41:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][680/1251] eta 0:02:47 lr 0.000153 time 0.2699 (0.2935) loss 2.3855 (3.2144) grad_norm 2.1547 (2.3678) [2021-04-16 13:41:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][690/1251] eta 0:02:44 lr 0.000153 time 0.2504 (0.2932) loss 3.0879 (3.2184) grad_norm 2.3483 (2.3684) [2021-04-16 13:41:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][700/1251] eta 0:02:41 lr 0.000153 time 0.2834 (0.2929) loss 3.4692 (3.2174) grad_norm 2.3228 (2.3685) [2021-04-16 13:41:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][710/1251] eta 0:02:38 lr 0.000153 time 0.2741 (0.2926) loss 2.7224 (3.2162) grad_norm 2.2678 (2.3687) [2021-04-16 13:41:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][720/1251] eta 0:02:35 lr 0.000153 time 0.2901 (0.2926) loss 3.7717 (3.2190) grad_norm 2.2717 (2.3667) [2021-04-16 13:41:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][730/1251] eta 0:02:32 lr 0.000153 time 0.2804 (0.2927) loss 3.7679 (3.2218) grad_norm 2.0716 (2.3689) [2021-04-16 13:41:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][740/1251] eta 0:02:29 lr 0.000153 time 0.2620 (0.2927) loss 3.9338 (3.2200) grad_norm 2.5733 (2.3693) [2021-04-16 13:41:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][750/1251] eta 0:02:26 lr 0.000153 time 0.2778 (0.2924) loss 3.5639 (3.2183) grad_norm 3.1173 (2.3704) [2021-04-16 13:41:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][760/1251] eta 0:02:23 lr 0.000153 time 0.2779 (0.2922) loss 3.8434 (3.2150) grad_norm 2.2149 (2.3724) [2021-04-16 13:42:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][770/1251] eta 0:02:20 lr 0.000153 time 0.2919 (0.2920) loss 2.4368 (3.2140) grad_norm 2.0481 (2.3712) [2021-04-16 13:42:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][780/1251] eta 0:02:17 lr 0.000153 time 0.2704 (0.2922) loss 3.7719 (3.2153) grad_norm 2.6446 (2.3724) [2021-04-16 13:42:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][790/1251] eta 0:02:14 lr 0.000153 time 0.2582 (0.2922) loss 3.2788 (3.2139) grad_norm 2.3854 (2.3712) [2021-04-16 13:42:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][800/1251] eta 0:02:11 lr 0.000153 time 0.2836 (0.2920) loss 2.7855 (3.2153) grad_norm 2.2012 (2.3698) [2021-04-16 13:42:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][810/1251] eta 0:02:08 lr 0.000153 time 0.2812 (0.2918) loss 2.9355 (3.2145) grad_norm 2.3692 (2.3691) [2021-04-16 13:42:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][820/1251] eta 0:02:05 lr 0.000153 time 0.2632 (0.2916) loss 3.0075 (3.2128) grad_norm 2.6286 (2.3680) [2021-04-16 13:42:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][830/1251] eta 0:02:02 lr 0.000153 time 0.2665 (0.2915) loss 3.5267 (3.2127) grad_norm 2.0405 (2.3679) [2021-04-16 13:42:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][840/1251] eta 0:01:59 lr 0.000153 time 0.2650 (0.2913) loss 3.2299 (3.2149) grad_norm 2.3832 (2.3676) [2021-04-16 13:42:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][850/1251] eta 0:01:56 lr 0.000153 time 0.2894 (0.2911) loss 3.1129 (3.2176) grad_norm 2.7666 (2.3693) [2021-04-16 13:42:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][860/1251] eta 0:01:53 lr 0.000152 time 0.2748 (0.2909) loss 3.6545 (3.2189) grad_norm 2.6920 (2.3705) [2021-04-16 13:42:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][870/1251] eta 0:01:50 lr 0.000152 time 0.2710 (0.2908) loss 3.7043 (3.2205) grad_norm 2.0986 (2.3701) [2021-04-16 13:42:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][880/1251] eta 0:01:47 lr 0.000152 time 0.2887 (0.2906) loss 3.2929 (3.2213) grad_norm 2.8176 (2.3723) [2021-04-16 13:42:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][890/1251] eta 0:01:44 lr 0.000152 time 0.2857 (0.2905) loss 3.9814 (3.2219) grad_norm 2.1596 (2.3750) [2021-04-16 13:42:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][900/1251] eta 0:01:41 lr 0.000152 time 0.2695 (0.2903) loss 3.4278 (3.2237) grad_norm 2.1868 (2.3759) [2021-04-16 13:42:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][910/1251] eta 0:01:38 lr 0.000152 time 0.3051 (0.2902) loss 3.4994 (3.2221) grad_norm 2.3628 (2.3766) [2021-04-16 13:42:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][920/1251] eta 0:01:35 lr 0.000152 time 0.2671 (0.2900) loss 4.2465 (3.2239) grad_norm 2.3879 (2.3752) [2021-04-16 13:42:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][930/1251] eta 0:01:33 lr 0.000152 time 0.2519 (0.2900) loss 2.6412 (3.2208) grad_norm 2.3277 (2.3756) [2021-04-16 13:42:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][940/1251] eta 0:01:30 lr 0.000152 time 0.2548 (0.2902) loss 3.6974 (3.2189) grad_norm 2.4124 (2.3751) [2021-04-16 13:42:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][950/1251] eta 0:01:27 lr 0.000152 time 0.2651 (0.2900) loss 3.3681 (3.2219) grad_norm 2.3294 (2.3756) [2021-04-16 13:42:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][960/1251] eta 0:01:24 lr 0.000152 time 0.2530 (0.2900) loss 2.9970 (3.2212) grad_norm 2.4330 (2.3760) [2021-04-16 13:42:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][970/1251] eta 0:01:21 lr 0.000152 time 0.2810 (0.2899) loss 2.4840 (3.2199) grad_norm 1.9277 (2.3758) [2021-04-16 13:43:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][980/1251] eta 0:01:18 lr 0.000152 time 0.2456 (0.2898) loss 2.3055 (3.2189) grad_norm 2.9507 (2.3761) [2021-04-16 13:43:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][990/1251] eta 0:01:15 lr 0.000152 time 0.2922 (0.2897) loss 2.6519 (3.2174) grad_norm 2.4455 (2.3752) [2021-04-16 13:43:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][1000/1251] eta 0:01:12 lr 0.000152 time 0.2690 (0.2895) loss 3.9362 (3.2195) grad_norm 2.4950 (2.3752) [2021-04-16 13:43:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][1010/1251] eta 0:01:09 lr 0.000152 time 0.2840 (0.2894) loss 3.8796 (3.2184) grad_norm 2.3215 (2.3755) [2021-04-16 13:43:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][1020/1251] eta 0:01:06 lr 0.000152 time 0.3121 (0.2893) loss 3.5578 (3.2195) grad_norm 2.1624 (2.3758) [2021-04-16 13:43:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][1030/1251] eta 0:01:03 lr 0.000152 time 0.2884 (0.2892) loss 2.4840 (3.2188) grad_norm 2.2131 (2.3747) [2021-04-16 13:43:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][1040/1251] eta 0:01:01 lr 0.000152 time 0.2737 (0.2893) loss 2.9606 (3.2190) grad_norm 2.2503 (2.3749) [2021-04-16 13:43:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][1050/1251] eta 0:00:58 lr 0.000152 time 0.2673 (0.2891) loss 3.2797 (3.2199) grad_norm 2.6290 (2.3753) [2021-04-16 13:43:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][1060/1251] eta 0:00:55 lr 0.000152 time 0.2854 (0.2890) loss 2.5790 (3.2188) grad_norm 3.1319 (2.3770) [2021-04-16 13:43:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][1070/1251] eta 0:00:52 lr 0.000152 time 0.2758 (0.2889) loss 3.1806 (3.2181) grad_norm 2.2891 (2.3766) [2021-04-16 13:43:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][1080/1251] eta 0:00:49 lr 0.000152 time 0.2646 (0.2887) loss 2.0810 (3.2166) grad_norm 3.2878 (2.3777) [2021-04-16 13:43:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][1090/1251] eta 0:00:46 lr 0.000152 time 0.2770 (0.2887) loss 3.7806 (3.2153) grad_norm 2.3613 (2.3782) [2021-04-16 13:43:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][1100/1251] eta 0:00:43 lr 0.000152 time 0.2728 (0.2885) loss 3.1380 (3.2139) grad_norm 2.3425 (2.3770) [2021-04-16 13:43:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][1110/1251] eta 0:00:40 lr 0.000152 time 0.2694 (0.2885) loss 3.1474 (3.2160) grad_norm 2.2327 (2.3778) [2021-04-16 13:43:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][1120/1251] eta 0:00:37 lr 0.000152 time 0.3113 (0.2885) loss 2.9790 (3.2139) grad_norm 2.1353 (2.3769) [2021-04-16 13:43:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][1130/1251] eta 0:00:34 lr 0.000152 time 0.2609 (0.2884) loss 2.2138 (3.2132) grad_norm 2.1239 (2.3764) [2021-04-16 13:43:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][1140/1251] eta 0:00:31 lr 0.000152 time 0.2952 (0.2883) loss 2.5776 (3.2127) grad_norm 2.6829 (2.3768) [2021-04-16 13:43:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][1150/1251] eta 0:00:29 lr 0.000152 time 0.2708 (0.2882) loss 3.1504 (3.2132) grad_norm 2.4499 (2.3776) [2021-04-16 13:43:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][1160/1251] eta 0:00:26 lr 0.000152 time 0.2917 (0.2882) loss 3.5889 (3.2122) grad_norm 2.1907 (2.3767) [2021-04-16 13:43:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][1170/1251] eta 0:00:23 lr 0.000152 time 0.2640 (0.2881) loss 3.3288 (3.2105) grad_norm 2.1392 (2.3776) [2021-04-16 13:43:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][1180/1251] eta 0:00:20 lr 0.000152 time 0.2733 (0.2882) loss 3.1966 (3.2101) grad_norm 2.0170 (2.3776) [2021-04-16 13:43:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][1190/1251] eta 0:00:17 lr 0.000152 time 0.2449 (0.2881) loss 2.6700 (3.2099) grad_norm 2.0175 (2.3767) [2021-04-16 13:44:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][1200/1251] eta 0:00:14 lr 0.000151 time 0.2594 (0.2880) loss 2.9247 (3.2087) grad_norm 2.7760 (2.3767) [2021-04-16 13:44:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][1210/1251] eta 0:00:11 lr 0.000151 time 0.2789 (0.2879) loss 3.0842 (3.2066) grad_norm 2.0070 (2.3753) [2021-04-16 13:44:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][1220/1251] eta 0:00:08 lr 0.000151 time 0.2731 (0.2878) loss 2.6362 (3.2059) grad_norm 2.1683 (2.3740) [2021-04-16 13:44:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][1230/1251] eta 0:00:06 lr 0.000151 time 0.2635 (0.2878) loss 4.1823 (3.2077) grad_norm 2.0289 (2.3723) [2021-04-16 13:44:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][1240/1251] eta 0:00:03 lr 0.000151 time 0.2501 (0.2877) loss 3.0406 (3.2059) grad_norm 1.9309 (2.3709) [2021-04-16 13:44:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [225/300][1250/1251] eta 0:00:00 lr 0.000151 time 0.2487 (0.2874) loss 3.6592 (3.2065) grad_norm 2.3814 (2.3703) [2021-04-16 13:44:19 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 225 training takes 0:06:02 [2021-04-16 13:44:19 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_225.pth saving...... [2021-04-16 13:44:28 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_225.pth saved !!! [2021-04-16 13:44:29 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.267 (1.267) Loss 0.8637 (0.8637) Acc@1 80.273 (80.273) Acc@5 95.117 (95.117) [2021-04-16 13:44:31 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.111 (0.281) Loss 0.7939 (0.8504) Acc@1 80.176 (79.998) Acc@5 95.020 (94.957) [2021-04-16 13:44:32 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.316 (0.222) Loss 0.9045 (0.8662) Acc@1 78.320 (79.590) Acc@5 94.629 (94.801) [2021-04-16 13:44:35 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.271 (0.234) Loss 0.7978 (0.8631) Acc@1 82.227 (79.593) Acc@5 95.117 (94.957) [2021-04-16 13:44:36 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.077 (0.211) Loss 0.8357 (0.8642) Acc@1 79.980 (79.483) Acc@5 95.117 (94.936) [2021-04-16 13:44:41 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 79.436 Acc@5 94.942 [2021-04-16 13:44:41 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 79.4% [2021-04-16 13:44:41 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 79.50% [2021-04-16 13:44:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][0/1251] eta 1:57:01 lr 0.000151 time 5.6128 (5.6128) loss 3.0886 (3.0886) grad_norm 2.4290 (2.4290) [2021-04-16 13:44:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][10/1251] eta 0:15:46 lr 0.000151 time 0.2971 (0.7626) loss 3.7872 (3.0868) grad_norm 2.3013 (2.3816) [2021-04-16 13:44:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][20/1251] eta 0:10:55 lr 0.000151 time 0.3103 (0.5327) loss 2.7250 (2.9618) grad_norm 2.0796 (2.3420) [2021-04-16 13:44:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][30/1251] eta 0:09:10 lr 0.000151 time 0.2766 (0.4509) loss 2.6034 (3.0017) grad_norm 2.4914 (2.3482) [2021-04-16 13:44:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][40/1251] eta 0:08:14 lr 0.000151 time 0.2891 (0.4087) loss 2.3896 (3.0139) grad_norm 2.1346 (2.4097) [2021-04-16 13:45:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][50/1251] eta 0:07:40 lr 0.000151 time 0.3027 (0.3834) loss 2.7903 (3.0484) grad_norm 1.9730 (2.4074) [2021-04-16 13:45:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][60/1251] eta 0:07:15 lr 0.000151 time 0.2541 (0.3656) loss 3.6728 (3.0880) grad_norm 2.1341 (2.4116) [2021-04-16 13:45:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][70/1251] eta 0:06:57 lr 0.000151 time 0.2582 (0.3536) loss 3.5167 (3.0612) grad_norm 2.4341 (2.4040) [2021-04-16 13:45:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][80/1251] eta 0:06:42 lr 0.000151 time 0.2617 (0.3441) loss 2.5068 (3.0698) grad_norm 2.1867 (2.3801) [2021-04-16 13:45:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][90/1251] eta 0:06:31 lr 0.000151 time 0.2836 (0.3370) loss 2.6791 (3.0623) grad_norm 2.4475 (2.3767) [2021-04-16 13:45:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][100/1251] eta 0:06:22 lr 0.000151 time 0.2991 (0.3323) loss 2.4522 (3.0769) grad_norm 1.9783 (2.3657) [2021-04-16 13:45:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][110/1251] eta 0:06:13 lr 0.000151 time 0.2888 (0.3273) loss 3.0950 (3.0924) grad_norm 2.2673 (2.3563) [2021-04-16 13:45:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][120/1251] eta 0:06:05 lr 0.000151 time 0.2753 (0.3234) loss 2.9925 (3.0939) grad_norm 2.2597 (2.3696) [2021-04-16 13:45:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][130/1251] eta 0:06:00 lr 0.000151 time 0.2700 (0.3213) loss 3.9517 (3.0946) grad_norm 2.5924 (2.3753) [2021-04-16 13:45:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][140/1251] eta 0:05:55 lr 0.000151 time 0.2670 (0.3196) loss 3.1954 (3.1019) grad_norm 2.7238 (2.3747) [2021-04-16 13:45:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][150/1251] eta 0:05:49 lr 0.000151 time 0.2725 (0.3177) loss 3.3454 (3.1043) grad_norm 2.3228 (2.3765) [2021-04-16 13:45:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][160/1251] eta 0:05:43 lr 0.000151 time 0.2526 (0.3152) loss 3.1803 (3.0939) grad_norm 2.0673 (2.3998) [2021-04-16 13:45:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][170/1251] eta 0:05:39 lr 0.000151 time 0.2788 (0.3143) loss 2.8698 (3.0836) grad_norm 2.0638 (2.3977) [2021-04-16 13:45:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][180/1251] eta 0:05:34 lr 0.000151 time 0.2755 (0.3122) loss 3.7798 (3.0870) grad_norm 2.1812 (2.3928) [2021-04-16 13:45:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][190/1251] eta 0:05:29 lr 0.000151 time 0.2660 (0.3103) loss 2.4902 (3.0884) grad_norm 2.2206 (2.3887) [2021-04-16 13:45:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][200/1251] eta 0:05:24 lr 0.000151 time 0.2991 (0.3088) loss 2.6357 (3.0827) grad_norm 1.8317 (2.3864) [2021-04-16 13:45:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][210/1251] eta 0:05:19 lr 0.000151 time 0.2563 (0.3073) loss 3.3620 (3.0937) grad_norm 2.3169 (2.3831) [2021-04-16 13:45:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][220/1251] eta 0:05:15 lr 0.000151 time 0.2723 (0.3060) loss 3.2951 (3.1101) grad_norm 2.5093 (2.3946) [2021-04-16 13:45:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][230/1251] eta 0:05:11 lr 0.000151 time 0.2755 (0.3051) loss 3.2790 (3.1190) grad_norm 2.7002 (2.3999) [2021-04-16 13:45:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][240/1251] eta 0:05:07 lr 0.000151 time 0.3046 (0.3040) loss 3.5099 (3.1276) grad_norm 2.1431 (2.3971) [2021-04-16 13:45:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][250/1251] eta 0:05:03 lr 0.000151 time 0.2607 (0.3035) loss 3.2202 (3.1259) grad_norm 2.2905 (2.3903) [2021-04-16 13:46:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][260/1251] eta 0:04:59 lr 0.000151 time 0.2947 (0.3025) loss 3.3502 (3.1105) grad_norm 2.3099 (2.3932) [2021-04-16 13:46:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][270/1251] eta 0:04:56 lr 0.000151 time 0.2771 (0.3018) loss 3.4227 (3.1081) grad_norm 2.2341 (2.3914) [2021-04-16 13:46:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][280/1251] eta 0:04:52 lr 0.000151 time 0.2983 (0.3009) loss 3.4938 (3.1224) grad_norm 2.1731 (2.3866) [2021-04-16 13:46:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][290/1251] eta 0:04:48 lr 0.000150 time 0.2795 (0.3001) loss 3.6551 (3.1194) grad_norm 2.2012 (2.3885) [2021-04-16 13:46:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][300/1251] eta 0:04:44 lr 0.000150 time 0.2871 (0.2993) loss 3.3694 (3.1230) grad_norm 2.4038 (2.3870) [2021-04-16 13:46:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][310/1251] eta 0:04:40 lr 0.000150 time 0.2689 (0.2986) loss 3.2721 (3.1287) grad_norm 2.8764 (2.3890) [2021-04-16 13:46:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][320/1251] eta 0:04:37 lr 0.000150 time 0.2808 (0.2978) loss 3.0406 (3.1321) grad_norm 2.3018 (2.3850) [2021-04-16 13:46:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][330/1251] eta 0:04:33 lr 0.000150 time 0.2844 (0.2973) loss 3.2763 (3.1300) grad_norm 2.2136 (2.3850) [2021-04-16 13:46:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][340/1251] eta 0:04:30 lr 0.000150 time 0.2597 (0.2966) loss 3.8461 (3.1341) grad_norm 2.1838 (2.3798) [2021-04-16 13:46:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][350/1251] eta 0:04:26 lr 0.000150 time 0.2836 (0.2959) loss 3.4068 (3.1349) grad_norm 2.4980 (2.3802) [2021-04-16 13:46:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][360/1251] eta 0:04:23 lr 0.000150 time 0.2631 (0.2961) loss 3.4373 (3.1319) grad_norm 2.4472 (2.3786) [2021-04-16 13:46:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][370/1251] eta 0:04:20 lr 0.000150 time 0.2421 (0.2955) loss 3.1871 (3.1287) grad_norm 2.2867 (2.3808) [2021-04-16 13:46:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][380/1251] eta 0:04:17 lr 0.000150 time 0.2647 (0.2956) loss 2.7656 (3.1288) grad_norm 2.1906 (2.3827) [2021-04-16 13:46:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][390/1251] eta 0:04:14 lr 0.000150 time 0.2887 (0.2951) loss 2.9141 (3.1214) grad_norm 2.2773 (2.3825) [2021-04-16 13:46:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][400/1251] eta 0:04:10 lr 0.000150 time 0.2902 (0.2948) loss 2.9991 (3.1258) grad_norm 2.4769 (2.3874) [2021-04-16 13:46:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][410/1251] eta 0:04:07 lr 0.000150 time 0.2634 (0.2945) loss 2.8909 (3.1267) grad_norm 2.2069 (2.3875) [2021-04-16 13:46:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][420/1251] eta 0:04:04 lr 0.000150 time 0.2900 (0.2942) loss 3.6612 (3.1347) grad_norm 2.1962 (2.3866) [2021-04-16 13:46:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][430/1251] eta 0:04:01 lr 0.000150 time 0.2852 (0.2941) loss 2.8859 (3.1337) grad_norm 2.4624 (2.3887) [2021-04-16 13:46:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][440/1251] eta 0:03:58 lr 0.000150 time 0.2638 (0.2937) loss 3.8141 (3.1435) grad_norm 2.3863 (2.3915) [2021-04-16 13:46:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][450/1251] eta 0:03:54 lr 0.000150 time 0.2542 (0.2933) loss 2.1928 (3.1502) grad_norm 1.9915 (2.3874) [2021-04-16 13:46:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][460/1251] eta 0:03:51 lr 0.000150 time 0.2547 (0.2929) loss 3.7955 (3.1546) grad_norm 2.2834 (2.3861) [2021-04-16 13:46:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][470/1251] eta 0:03:48 lr 0.000150 time 0.2914 (0.2926) loss 3.5013 (3.1581) grad_norm 2.1349 (2.3882) [2021-04-16 13:47:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][480/1251] eta 0:03:45 lr 0.000150 time 0.2684 (0.2923) loss 2.6683 (3.1540) grad_norm 2.2974 (2.3892) [2021-04-16 13:47:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][490/1251] eta 0:03:42 lr 0.000150 time 0.2935 (0.2921) loss 3.0086 (3.1484) grad_norm 2.9049 (2.3905) [2021-04-16 13:47:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][500/1251] eta 0:03:39 lr 0.000150 time 0.2707 (0.2916) loss 3.4826 (3.1451) grad_norm 2.7849 (2.3892) [2021-04-16 13:47:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][510/1251] eta 0:03:35 lr 0.000150 time 0.2837 (0.2914) loss 3.5293 (3.1503) grad_norm 2.6635 (2.3894) [2021-04-16 13:47:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][520/1251] eta 0:03:32 lr 0.000150 time 0.2662 (0.2911) loss 3.5924 (3.1537) grad_norm 2.3941 (2.3877) [2021-04-16 13:47:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][530/1251] eta 0:03:29 lr 0.000150 time 0.2913 (0.2909) loss 3.8884 (3.1606) grad_norm 2.0856 (2.3880) [2021-04-16 13:47:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][540/1251] eta 0:03:26 lr 0.000150 time 0.2691 (0.2908) loss 2.2157 (3.1634) grad_norm 2.7838 (2.3881) [2021-04-16 13:47:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][550/1251] eta 0:03:23 lr 0.000150 time 0.2750 (0.2907) loss 2.5251 (3.1629) grad_norm 2.4704 (2.3895) [2021-04-16 13:47:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][560/1251] eta 0:03:20 lr 0.000150 time 0.2894 (0.2904) loss 3.0958 (3.1624) grad_norm 2.2741 (2.3870) [2021-04-16 13:47:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][570/1251] eta 0:03:17 lr 0.000150 time 0.2549 (0.2906) loss 2.7545 (3.1580) grad_norm 2.1116 (2.3852) [2021-04-16 13:47:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][580/1251] eta 0:03:15 lr 0.000150 time 0.2876 (0.2906) loss 1.9933 (3.1533) grad_norm 2.4558 (2.3827) [2021-04-16 13:47:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][590/1251] eta 0:03:11 lr 0.000150 time 0.2750 (0.2904) loss 3.2720 (3.1598) grad_norm 2.0436 (2.3830) [2021-04-16 13:47:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][600/1251] eta 0:03:08 lr 0.000150 time 0.2926 (0.2902) loss 3.1301 (3.1610) grad_norm 2.5929 (2.3832) [2021-04-16 13:47:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][610/1251] eta 0:03:05 lr 0.000150 time 0.2870 (0.2901) loss 3.6157 (3.1646) grad_norm 2.1886 (2.3832) [2021-04-16 13:47:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][620/1251] eta 0:03:02 lr 0.000150 time 0.2449 (0.2900) loss 3.5139 (3.1699) grad_norm 2.2853 (2.3833) [2021-04-16 13:47:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][630/1251] eta 0:02:59 lr 0.000150 time 0.2795 (0.2897) loss 3.2471 (3.1737) grad_norm 2.3109 (2.3850) [2021-04-16 13:47:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][640/1251] eta 0:02:56 lr 0.000149 time 0.2737 (0.2896) loss 2.6483 (3.1697) grad_norm 2.2418 (2.3845) [2021-04-16 13:47:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][650/1251] eta 0:02:53 lr 0.000149 time 0.2712 (0.2894) loss 3.5176 (3.1699) grad_norm 3.0721 (2.3863) [2021-04-16 13:47:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][660/1251] eta 0:02:50 lr 0.000149 time 0.2451 (0.2891) loss 3.5284 (3.1690) grad_norm 2.8837 (2.3852) [2021-04-16 13:47:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][670/1251] eta 0:02:47 lr 0.000149 time 0.2782 (0.2890) loss 2.9677 (3.1718) grad_norm 2.3559 (2.3858) [2021-04-16 13:47:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][680/1251] eta 0:02:44 lr 0.000149 time 0.2897 (0.2889) loss 2.4057 (3.1719) grad_norm 2.1771 (2.3848) [2021-04-16 13:48:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][690/1251] eta 0:02:42 lr 0.000149 time 0.2737 (0.2889) loss 2.6463 (3.1714) grad_norm 2.2649 (2.3835) [2021-04-16 13:48:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][700/1251] eta 0:02:39 lr 0.000149 time 0.3553 (0.2889) loss 2.6371 (3.1725) grad_norm 2.0812 (2.3818) [2021-04-16 13:48:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][710/1251] eta 0:02:36 lr 0.000149 time 0.2786 (0.2886) loss 3.1758 (3.1750) grad_norm 2.0118 (2.3795) [2021-04-16 13:48:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][720/1251] eta 0:02:33 lr 0.000149 time 0.2705 (0.2888) loss 3.1225 (3.1752) grad_norm 2.1952 (2.3786) [2021-04-16 13:48:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][730/1251] eta 0:02:30 lr 0.000149 time 0.2733 (0.2886) loss 3.5655 (3.1774) grad_norm 2.1547 (2.3778) [2021-04-16 13:48:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][740/1251] eta 0:02:27 lr 0.000149 time 0.2649 (0.2886) loss 3.3170 (3.1773) grad_norm 2.0348 (2.3783) [2021-04-16 13:48:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][750/1251] eta 0:02:24 lr 0.000149 time 0.2627 (0.2887) loss 3.2642 (3.1788) grad_norm 2.1634 (2.3775) [2021-04-16 13:48:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][760/1251] eta 0:02:21 lr 0.000149 time 0.2665 (0.2886) loss 2.4769 (3.1802) grad_norm 2.4145 (2.3765) [2021-04-16 13:48:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][770/1251] eta 0:02:18 lr 0.000149 time 0.2934 (0.2884) loss 3.6473 (3.1820) grad_norm 2.2965 (2.3747) [2021-04-16 13:48:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][780/1251] eta 0:02:15 lr 0.000149 time 0.2677 (0.2882) loss 2.8865 (3.1807) grad_norm 3.0899 (2.3747) [2021-04-16 13:48:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][790/1251] eta 0:02:12 lr 0.000149 time 0.2614 (0.2881) loss 2.9785 (3.1802) grad_norm 2.4003 (2.3754) [2021-04-16 13:48:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][800/1251] eta 0:02:09 lr 0.000149 time 0.2736 (0.2880) loss 2.7739 (3.1801) grad_norm 2.2471 (2.3749) [2021-04-16 13:48:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][810/1251] eta 0:02:06 lr 0.000149 time 0.2822 (0.2879) loss 4.0435 (3.1862) grad_norm 2.4377 (2.3750) [2021-04-16 13:48:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][820/1251] eta 0:02:04 lr 0.000149 time 0.2941 (0.2878) loss 3.1475 (3.1848) grad_norm 2.5176 (2.3761) [2021-04-16 13:48:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][830/1251] eta 0:02:01 lr 0.000149 time 0.2891 (0.2879) loss 3.0085 (3.1864) grad_norm 2.4074 (2.3775) [2021-04-16 13:48:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][840/1251] eta 0:01:58 lr 0.000149 time 0.2738 (0.2878) loss 3.7234 (3.1872) grad_norm 1.9411 (2.3772) [2021-04-16 13:48:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][850/1251] eta 0:01:55 lr 0.000149 time 0.2580 (0.2876) loss 3.6647 (3.1882) grad_norm 2.5908 (2.3749) [2021-04-16 13:48:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][860/1251] eta 0:01:52 lr 0.000149 time 0.2805 (0.2875) loss 3.2578 (3.1876) grad_norm 2.1140 (2.3740) [2021-04-16 13:48:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][870/1251] eta 0:01:49 lr 0.000149 time 0.2857 (0.2874) loss 3.5468 (3.1887) grad_norm 2.7666 (2.3736) [2021-04-16 13:48:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][880/1251] eta 0:01:46 lr 0.000149 time 0.2900 (0.2872) loss 3.6767 (3.1867) grad_norm 2.0879 (2.3721) [2021-04-16 13:48:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][890/1251] eta 0:01:43 lr 0.000149 time 0.3051 (0.2872) loss 3.5500 (3.1862) grad_norm 2.2957 (2.3732) [2021-04-16 13:48:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][900/1251] eta 0:01:40 lr 0.000149 time 0.2902 (0.2871) loss 2.3892 (3.1862) grad_norm 2.4083 (2.3745) [2021-04-16 13:49:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][910/1251] eta 0:01:37 lr 0.000149 time 0.3160 (0.2871) loss 2.8739 (3.1845) grad_norm 2.5177 (2.3750) [2021-04-16 13:49:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][920/1251] eta 0:01:34 lr 0.000149 time 0.2742 (0.2870) loss 3.6412 (3.1861) grad_norm 2.5947 (2.3763) [2021-04-16 13:49:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][930/1251] eta 0:01:32 lr 0.000149 time 0.2817 (0.2870) loss 2.8792 (3.1857) grad_norm 2.4377 (2.3766) [2021-04-16 13:49:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][940/1251] eta 0:01:29 lr 0.000149 time 0.2909 (0.2871) loss 2.4858 (3.1850) grad_norm 2.5205 (2.3756) [2021-04-16 13:49:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][950/1251] eta 0:01:26 lr 0.000149 time 0.2984 (0.2870) loss 3.2302 (3.1839) grad_norm 3.1534 (2.3751) [2021-04-16 13:49:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][960/1251] eta 0:01:23 lr 0.000149 time 0.2849 (0.2871) loss 2.5607 (3.1848) grad_norm 2.2541 (2.3757) [2021-04-16 13:49:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][970/1251] eta 0:01:20 lr 0.000149 time 0.2710 (0.2870) loss 2.9491 (3.1855) grad_norm 2.0452 (2.3754) [2021-04-16 13:49:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][980/1251] eta 0:01:17 lr 0.000149 time 0.2832 (0.2869) loss 2.8464 (3.1867) grad_norm 2.1570 (2.3754) [2021-04-16 13:49:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][990/1251] eta 0:01:14 lr 0.000148 time 0.2621 (0.2869) loss 3.4845 (3.1879) grad_norm 2.2291 (2.3752) [2021-04-16 13:49:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][1000/1251] eta 0:01:11 lr 0.000148 time 0.2956 (0.2868) loss 3.2522 (3.1883) grad_norm 2.5033 (2.3752) [2021-04-16 13:49:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][1010/1251] eta 0:01:09 lr 0.000148 time 0.2770 (0.2868) loss 2.0682 (3.1881) grad_norm 2.5255 (2.3743) [2021-04-16 13:49:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][1020/1251] eta 0:01:06 lr 0.000148 time 0.2810 (0.2867) loss 3.4981 (3.1882) grad_norm 2.2710 (2.3735) [2021-04-16 13:49:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][1030/1251] eta 0:01:03 lr 0.000148 time 0.2801 (0.2866) loss 3.2456 (3.1881) grad_norm 2.8679 (2.3746) [2021-04-16 13:49:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][1040/1251] eta 0:01:00 lr 0.000148 time 0.2782 (0.2865) loss 2.1833 (3.1860) grad_norm 2.1674 (2.3740) [2021-04-16 13:49:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][1050/1251] eta 0:00:57 lr 0.000148 time 0.2793 (0.2864) loss 3.2763 (3.1863) grad_norm 2.1247 (2.3732) [2021-04-16 13:49:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][1060/1251] eta 0:00:54 lr 0.000148 time 0.2558 (0.2862) loss 3.7932 (3.1880) grad_norm 2.7696 (2.3751) [2021-04-16 13:49:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][1070/1251] eta 0:00:51 lr 0.000148 time 0.2877 (0.2862) loss 3.3647 (3.1895) grad_norm 2.1524 (2.3747) [2021-04-16 13:49:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][1080/1251] eta 0:00:48 lr 0.000148 time 0.2947 (0.2861) loss 2.7528 (3.1875) grad_norm 2.7735 (2.3776) [2021-04-16 13:49:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][1090/1251] eta 0:00:46 lr 0.000148 time 0.2772 (0.2860) loss 1.9982 (3.1854) grad_norm 2.5457 (2.3782) [2021-04-16 13:49:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][1100/1251] eta 0:00:43 lr 0.000148 time 0.2684 (0.2858) loss 3.5756 (3.1829) grad_norm 2.3560 (2.3788) [2021-04-16 13:49:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][1110/1251] eta 0:00:40 lr 0.000148 time 0.2968 (0.2858) loss 2.4443 (3.1823) grad_norm 2.7551 (2.3792) [2021-04-16 13:50:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][1120/1251] eta 0:00:37 lr 0.000148 time 0.2573 (0.2858) loss 3.9555 (3.1820) grad_norm 3.3704 (2.3785) [2021-04-16 13:50:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][1130/1251] eta 0:00:34 lr 0.000148 time 0.3040 (0.2857) loss 3.7621 (3.1833) grad_norm 1.9210 (2.3779) [2021-04-16 13:50:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][1140/1251] eta 0:00:31 lr 0.000148 time 0.2568 (0.2858) loss 3.4596 (3.1836) grad_norm 2.3640 (2.3781) [2021-04-16 13:50:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][1150/1251] eta 0:00:28 lr 0.000148 time 0.2693 (0.2859) loss 2.6990 (3.1825) grad_norm 2.2735 (2.3791) [2021-04-16 13:50:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][1160/1251] eta 0:00:26 lr 0.000148 time 0.2819 (0.2858) loss 2.4811 (3.1822) grad_norm 2.2987 (2.3797) [2021-04-16 13:50:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][1170/1251] eta 0:00:23 lr 0.000148 time 0.2969 (0.2857) loss 3.1310 (3.1826) grad_norm 2.0529 (2.3800) [2021-04-16 13:50:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][1180/1251] eta 0:00:20 lr 0.000148 time 0.3022 (0.2856) loss 3.7286 (3.1821) grad_norm 1.9806 (2.3801) [2021-04-16 13:50:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][1190/1251] eta 0:00:17 lr 0.000148 time 0.3111 (0.2856) loss 2.6167 (3.1805) grad_norm 2.1373 (2.3800) [2021-04-16 13:50:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][1200/1251] eta 0:00:14 lr 0.000148 time 0.2610 (0.2854) loss 2.4668 (3.1787) grad_norm 2.5752 (2.3807) [2021-04-16 13:50:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][1210/1251] eta 0:00:11 lr 0.000148 time 0.2753 (0.2854) loss 2.3755 (3.1778) grad_norm 2.0568 (2.3802) [2021-04-16 13:50:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][1220/1251] eta 0:00:08 lr 0.000148 time 0.2591 (0.2853) loss 3.9612 (3.1786) grad_norm 2.2603 (2.3799) [2021-04-16 13:50:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][1230/1251] eta 0:00:05 lr 0.000148 time 0.2858 (0.2853) loss 2.6690 (3.1803) grad_norm 2.1527 (2.3806) [2021-04-16 13:50:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][1240/1251] eta 0:00:03 lr 0.000148 time 0.2488 (0.2851) loss 3.1387 (3.1802) grad_norm 2.7263 (2.3810) [2021-04-16 13:50:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [226/300][1250/1251] eta 0:00:00 lr 0.000148 time 0.2488 (0.2849) loss 3.1073 (3.1807) grad_norm 2.2730 (2.3820) [2021-04-16 13:50:41 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 226 training takes 0:06:00 [2021-04-16 13:50:41 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_226.pth saving...... [2021-04-16 13:50:52 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_226.pth saved !!! [2021-04-16 13:50:53 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.074 (1.074) Loss 0.8422 (0.8422) Acc@1 81.055 (81.055) Acc@5 95.801 (95.801) [2021-04-16 13:50:55 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.302 (0.240) Loss 0.8012 (0.8681) Acc@1 81.934 (79.812) Acc@5 95.703 (94.904) [2021-04-16 13:50:57 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.130 (0.222) Loss 0.8426 (0.8704) Acc@1 79.199 (79.818) Acc@5 95.312 (94.852) [2021-04-16 13:50:59 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.130 (0.232) Loss 0.8794 (0.8728) Acc@1 78.711 (79.577) Acc@5 94.434 (94.878) [2021-04-16 13:51:01 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.216) Loss 0.8869 (0.8754) Acc@1 80.371 (79.523) Acc@5 94.629 (94.879) [2021-04-16 13:51:07 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 79.498 Acc@5 94.842 [2021-04-16 13:51:07 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 79.5% [2021-04-16 13:51:07 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 79.50% [2021-04-16 13:51:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][0/1251] eta 5:05:53 lr 0.000148 time 14.6711 (14.6711) loss 2.3746 (2.3746) grad_norm 2.2939 (2.2939) [2021-04-16 13:51:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][10/1251] eta 0:32:41 lr 0.000148 time 0.2783 (1.5805) loss 2.7460 (3.1262) grad_norm 2.1184 (2.4371) [2021-04-16 13:51:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][20/1251] eta 0:19:44 lr 0.000148 time 0.2682 (0.9625) loss 3.5784 (3.1816) grad_norm 2.5958 (2.4762) [2021-04-16 13:51:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][30/1251] eta 0:15:09 lr 0.000148 time 0.2701 (0.7447) loss 3.0064 (3.1324) grad_norm 2.2668 (2.3830) [2021-04-16 13:51:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][40/1251] eta 0:12:44 lr 0.000148 time 0.2820 (0.6311) loss 4.2184 (3.1177) grad_norm 2.2071 (2.3740) [2021-04-16 13:51:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][50/1251] eta 0:11:15 lr 0.000148 time 0.2971 (0.5628) loss 3.5363 (3.1412) grad_norm 2.7386 (2.3855) [2021-04-16 13:51:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][60/1251] eta 0:10:17 lr 0.000148 time 0.2821 (0.5186) loss 2.1173 (3.1224) grad_norm 2.3982 (2.3998) [2021-04-16 13:51:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][70/1251] eta 0:09:32 lr 0.000148 time 0.2924 (0.4845) loss 3.0729 (3.1504) grad_norm 2.7011 (2.3824) [2021-04-16 13:51:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][80/1251] eta 0:08:57 lr 0.000147 time 0.2808 (0.4589) loss 3.9176 (3.1663) grad_norm 2.7119 (2.3866) [2021-04-16 13:51:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][90/1251] eta 0:08:30 lr 0.000147 time 0.2762 (0.4397) loss 3.3597 (3.1430) grad_norm 5.7963 (2.4159) [2021-04-16 13:51:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][100/1251] eta 0:08:07 lr 0.000147 time 0.2693 (0.4238) loss 2.0531 (3.1568) grad_norm 3.0329 (2.4629) [2021-04-16 13:51:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][110/1251] eta 0:07:48 lr 0.000147 time 0.2772 (0.4108) loss 2.3613 (3.1507) grad_norm 2.2606 (2.4879) [2021-04-16 13:51:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][120/1251] eta 0:07:31 lr 0.000147 time 0.2728 (0.3996) loss 3.5444 (3.1473) grad_norm 1.9613 (2.4812) [2021-04-16 13:51:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][130/1251] eta 0:07:17 lr 0.000147 time 0.2871 (0.3906) loss 3.6161 (3.1787) grad_norm 2.5866 (2.4759) [2021-04-16 13:52:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][140/1251] eta 0:07:06 lr 0.000147 time 0.3025 (0.3836) loss 3.2249 (3.1803) grad_norm 1.8397 (2.4573) [2021-04-16 13:52:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][150/1251] eta 0:06:55 lr 0.000147 time 0.2798 (0.3771) loss 3.8074 (3.1999) grad_norm 2.1909 (2.4468) [2021-04-16 13:52:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][160/1251] eta 0:06:45 lr 0.000147 time 0.2818 (0.3713) loss 3.6566 (3.2044) grad_norm 2.6353 (2.4473) [2021-04-16 13:52:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][170/1251] eta 0:06:36 lr 0.000147 time 0.3003 (0.3669) loss 3.6804 (3.2111) grad_norm 2.3908 (2.4398) [2021-04-16 13:52:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][180/1251] eta 0:06:27 lr 0.000147 time 0.2813 (0.3620) loss 3.2504 (3.2080) grad_norm 2.1572 (2.4358) [2021-04-16 13:52:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][190/1251] eta 0:06:19 lr 0.000147 time 0.2712 (0.3575) loss 2.8885 (3.2136) grad_norm 2.1754 (2.4273) [2021-04-16 13:52:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][200/1251] eta 0:06:11 lr 0.000147 time 0.2946 (0.3537) loss 3.4962 (3.2173) grad_norm 2.4109 (2.4217) [2021-04-16 13:52:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][210/1251] eta 0:06:04 lr 0.000147 time 0.2812 (0.3501) loss 2.8202 (3.2176) grad_norm 2.3511 (2.4158) [2021-04-16 13:52:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][220/1251] eta 0:05:58 lr 0.000147 time 0.2811 (0.3475) loss 2.4622 (3.2160) grad_norm 2.5611 (2.4203) [2021-04-16 13:52:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][230/1251] eta 0:05:51 lr 0.000147 time 0.2916 (0.3445) loss 2.5186 (3.2168) grad_norm 2.6765 (2.4236) [2021-04-16 13:52:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][240/1251] eta 0:05:45 lr 0.000147 time 0.2731 (0.3415) loss 2.6964 (3.1972) grad_norm 2.5465 (2.4274) [2021-04-16 13:52:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][250/1251] eta 0:05:39 lr 0.000147 time 0.2710 (0.3393) loss 3.1682 (3.1970) grad_norm 2.2209 (2.4340) [2021-04-16 13:52:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][260/1251] eta 0:05:33 lr 0.000147 time 0.2672 (0.3370) loss 3.2849 (3.1867) grad_norm 2.1491 (2.4336) [2021-04-16 13:52:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][270/1251] eta 0:05:28 lr 0.000147 time 0.2855 (0.3348) loss 3.4158 (3.1807) grad_norm 2.4640 (2.4421) [2021-04-16 13:52:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][280/1251] eta 0:05:23 lr 0.000147 time 0.3132 (0.3328) loss 3.4934 (3.1782) grad_norm 2.2422 (2.4392) [2021-04-16 13:52:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][290/1251] eta 0:05:18 lr 0.000147 time 0.3028 (0.3311) loss 3.6921 (3.1753) grad_norm 2.7571 (2.4463) [2021-04-16 13:52:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][300/1251] eta 0:05:13 lr 0.000147 time 0.3009 (0.3294) loss 2.6069 (3.1725) grad_norm 2.5697 (2.4431) [2021-04-16 13:52:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][310/1251] eta 0:05:08 lr 0.000147 time 0.2925 (0.3278) loss 3.9345 (3.1662) grad_norm 5.1834 (2.4470) [2021-04-16 13:52:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][320/1251] eta 0:05:04 lr 0.000147 time 0.2770 (0.3267) loss 2.9755 (3.1575) grad_norm 2.3380 (2.4491) [2021-04-16 13:52:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][330/1251] eta 0:04:59 lr 0.000147 time 0.2729 (0.3254) loss 3.7481 (3.1537) grad_norm 2.2676 (2.4458) [2021-04-16 13:52:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][340/1251] eta 0:04:55 lr 0.000147 time 0.3052 (0.3246) loss 2.7623 (3.1512) grad_norm 2.0346 (2.4406) [2021-04-16 13:53:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][350/1251] eta 0:04:51 lr 0.000147 time 0.3736 (0.3240) loss 3.5126 (3.1554) grad_norm 2.2800 (2.4378) [2021-04-16 13:53:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][360/1251] eta 0:04:48 lr 0.000147 time 0.2773 (0.3234) loss 3.4335 (3.1601) grad_norm 2.3256 (2.4380) [2021-04-16 13:53:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][370/1251] eta 0:04:43 lr 0.000147 time 0.2619 (0.3224) loss 2.9990 (3.1612) grad_norm 2.9639 (2.4410) [2021-04-16 13:53:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][380/1251] eta 0:04:39 lr 0.000147 time 0.3014 (0.3213) loss 2.7598 (3.1544) grad_norm 2.5569 (2.4431) [2021-04-16 13:53:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][390/1251] eta 0:04:35 lr 0.000147 time 0.2886 (0.3205) loss 2.1094 (3.1471) grad_norm 2.4782 (2.4404) [2021-04-16 13:53:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][400/1251] eta 0:04:31 lr 0.000147 time 0.2745 (0.3195) loss 3.2976 (3.1473) grad_norm 2.2728 (2.4366) [2021-04-16 13:53:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][410/1251] eta 0:04:27 lr 0.000147 time 0.2851 (0.3185) loss 2.4040 (3.1500) grad_norm 2.3998 (2.4360) [2021-04-16 13:53:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][420/1251] eta 0:04:23 lr 0.000147 time 0.2917 (0.3176) loss 3.6078 (3.1466) grad_norm 2.3093 (2.4302) [2021-04-16 13:53:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][430/1251] eta 0:04:19 lr 0.000146 time 0.2857 (0.3167) loss 2.4690 (3.1432) grad_norm 2.1614 (2.4307) [2021-04-16 13:53:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][440/1251] eta 0:04:16 lr 0.000146 time 0.2706 (0.3159) loss 2.6689 (3.1435) grad_norm 2.1070 (2.4283) [2021-04-16 13:53:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][450/1251] eta 0:04:12 lr 0.000146 time 0.2516 (0.3150) loss 3.7337 (3.1485) grad_norm 2.2575 (2.4250) [2021-04-16 13:53:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][460/1251] eta 0:04:08 lr 0.000146 time 0.3037 (0.3141) loss 2.6910 (3.1473) grad_norm 2.1103 (2.4232) [2021-04-16 13:53:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][470/1251] eta 0:04:04 lr 0.000146 time 0.2744 (0.3133) loss 3.4636 (3.1470) grad_norm 2.1010 (2.4189) [2021-04-16 13:53:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][480/1251] eta 0:04:01 lr 0.000146 time 0.2965 (0.3126) loss 3.3050 (3.1477) grad_norm 1.9701 (2.4181) [2021-04-16 13:53:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][490/1251] eta 0:03:57 lr 0.000146 time 0.2848 (0.3118) loss 3.0238 (3.1447) grad_norm 2.1511 (2.4162) [2021-04-16 13:53:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][500/1251] eta 0:03:53 lr 0.000146 time 0.2745 (0.3113) loss 3.6912 (3.1443) grad_norm 2.1129 (2.4164) [2021-04-16 13:53:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][510/1251] eta 0:03:50 lr 0.000146 time 0.2777 (0.3107) loss 3.5470 (3.1422) grad_norm 2.3962 (2.4139) [2021-04-16 13:53:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][520/1251] eta 0:03:46 lr 0.000146 time 0.2692 (0.3104) loss 3.2756 (3.1460) grad_norm 2.7999 (2.4152) [2021-04-16 13:53:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][530/1251] eta 0:03:43 lr 0.000146 time 0.3061 (0.3098) loss 3.6435 (3.1464) grad_norm 2.3503 (2.4143) [2021-04-16 13:53:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][540/1251] eta 0:03:40 lr 0.000146 time 0.2731 (0.3095) loss 3.0730 (3.1471) grad_norm 2.2140 (2.4142) [2021-04-16 13:53:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][550/1251] eta 0:03:36 lr 0.000146 time 0.2840 (0.3090) loss 3.1559 (3.1497) grad_norm 2.4610 (2.4144) [2021-04-16 13:54:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][560/1251] eta 0:03:33 lr 0.000146 time 0.2747 (0.3085) loss 3.5058 (3.1521) grad_norm 2.4453 (2.4140) [2021-04-16 13:54:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][570/1251] eta 0:03:29 lr 0.000146 time 0.3541 (0.3083) loss 3.5873 (3.1550) grad_norm 2.3663 (2.4135) [2021-04-16 13:54:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][580/1251] eta 0:03:26 lr 0.000146 time 0.2578 (0.3083) loss 3.5122 (3.1543) grad_norm 2.7735 (2.4114) [2021-04-16 13:54:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][590/1251] eta 0:03:23 lr 0.000146 time 0.2698 (0.3076) loss 2.4870 (3.1544) grad_norm 2.6231 (2.4110) [2021-04-16 13:54:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][600/1251] eta 0:03:20 lr 0.000146 time 0.4159 (0.3075) loss 2.8613 (3.1559) grad_norm 2.8220 (2.4120) [2021-04-16 13:54:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][610/1251] eta 0:03:16 lr 0.000146 time 0.2814 (0.3070) loss 2.6817 (3.1573) grad_norm 2.3694 (2.4124) [2021-04-16 13:54:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][620/1251] eta 0:03:13 lr 0.000146 time 0.2691 (0.3066) loss 3.2111 (3.1593) grad_norm 2.8260 (2.4122) [2021-04-16 13:54:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][630/1251] eta 0:03:10 lr 0.000146 time 0.2969 (0.3062) loss 3.2367 (3.1573) grad_norm 2.2593 (2.4105) [2021-04-16 13:54:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][640/1251] eta 0:03:06 lr 0.000146 time 0.2611 (0.3058) loss 2.4983 (3.1554) grad_norm 2.4520 (2.4117) [2021-04-16 13:54:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][650/1251] eta 0:03:03 lr 0.000146 time 0.2738 (0.3055) loss 3.1693 (3.1579) grad_norm 3.1344 (2.4115) [2021-04-16 13:54:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][660/1251] eta 0:03:00 lr 0.000146 time 0.2626 (0.3051) loss 3.3753 (3.1603) grad_norm 2.5952 (2.4126) [2021-04-16 13:54:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][670/1251] eta 0:02:57 lr 0.000146 time 0.2783 (0.3047) loss 3.4580 (3.1633) grad_norm 2.6404 (nan) [2021-04-16 13:54:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][680/1251] eta 0:02:53 lr 0.000146 time 0.2799 (0.3043) loss 3.0501 (3.1620) grad_norm 2.4293 (nan) [2021-04-16 13:54:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][690/1251] eta 0:02:50 lr 0.000146 time 0.2695 (0.3039) loss 2.4349 (3.1591) grad_norm 2.5782 (nan) [2021-04-16 13:54:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][700/1251] eta 0:02:47 lr 0.000146 time 0.2762 (0.3036) loss 3.0953 (3.1596) grad_norm 2.1990 (nan) [2021-04-16 13:54:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][710/1251] eta 0:02:44 lr 0.000146 time 0.2932 (0.3032) loss 3.3494 (3.1605) grad_norm 2.4980 (nan) [2021-04-16 13:54:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][720/1251] eta 0:02:40 lr 0.000146 time 0.2638 (0.3031) loss 3.1240 (3.1607) grad_norm 2.3299 (nan) [2021-04-16 13:54:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][730/1251] eta 0:02:37 lr 0.000146 time 0.2543 (0.3029) loss 3.4970 (3.1615) grad_norm 1.8989 (nan) [2021-04-16 13:54:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][740/1251] eta 0:02:34 lr 0.000146 time 0.2848 (0.3028) loss 3.2131 (3.1568) grad_norm 2.3892 (nan) [2021-04-16 13:54:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][750/1251] eta 0:02:31 lr 0.000146 time 0.2888 (0.3025) loss 3.6130 (3.1577) grad_norm 2.1823 (nan) [2021-04-16 13:54:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][760/1251] eta 0:02:28 lr 0.000146 time 0.2779 (0.3021) loss 3.1436 (3.1549) grad_norm 2.2530 (nan) [2021-04-16 13:54:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][770/1251] eta 0:02:25 lr 0.000146 time 0.2892 (0.3019) loss 3.2898 (3.1546) grad_norm 2.2170 (nan) [2021-04-16 13:55:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][780/1251] eta 0:02:22 lr 0.000145 time 0.2893 (0.3016) loss 4.2293 (3.1558) grad_norm 2.2504 (nan) [2021-04-16 13:55:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][790/1251] eta 0:02:18 lr 0.000145 time 0.2521 (0.3014) loss 2.6817 (3.1591) grad_norm 2.7980 (nan) [2021-04-16 13:55:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][800/1251] eta 0:02:15 lr 0.000145 time 0.2713 (0.3011) loss 3.2979 (3.1614) grad_norm 2.6230 (nan) [2021-04-16 13:55:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][810/1251] eta 0:02:12 lr 0.000145 time 0.2793 (0.3009) loss 3.0852 (3.1621) grad_norm 2.3527 (nan) [2021-04-16 13:55:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][820/1251] eta 0:02:09 lr 0.000145 time 0.2676 (0.3006) loss 3.2772 (3.1592) grad_norm 2.3165 (nan) [2021-04-16 13:55:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][830/1251] eta 0:02:06 lr 0.000145 time 0.2728 (0.3003) loss 2.5174 (3.1583) grad_norm 2.3115 (nan) [2021-04-16 13:55:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][840/1251] eta 0:02:03 lr 0.000145 time 0.2877 (0.3001) loss 2.8629 (3.1565) grad_norm 2.3485 (nan) [2021-04-16 13:55:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][850/1251] eta 0:02:00 lr 0.000145 time 0.2642 (0.2999) loss 2.9683 (3.1571) grad_norm 2.4423 (nan) [2021-04-16 13:55:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][860/1251] eta 0:01:57 lr 0.000145 time 0.2767 (0.2996) loss 3.0051 (3.1569) grad_norm 2.2641 (nan) [2021-04-16 13:55:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][870/1251] eta 0:01:54 lr 0.000145 time 0.2917 (0.2994) loss 3.4935 (3.1563) grad_norm 2.1113 (nan) [2021-04-16 13:55:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][880/1251] eta 0:01:50 lr 0.000145 time 0.2654 (0.2992) loss 3.5682 (3.1601) grad_norm 2.0732 (nan) [2021-04-16 13:55:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][890/1251] eta 0:01:47 lr 0.000145 time 0.2819 (0.2991) loss 2.8606 (3.1588) grad_norm 2.9527 (nan) [2021-04-16 13:55:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][900/1251] eta 0:01:44 lr 0.000145 time 0.2683 (0.2988) loss 3.6145 (3.1617) grad_norm 2.5840 (nan) [2021-04-16 13:55:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][910/1251] eta 0:01:41 lr 0.000145 time 0.2704 (0.2985) loss 3.0502 (3.1626) grad_norm 2.9768 (nan) [2021-04-16 13:55:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][920/1251] eta 0:01:38 lr 0.000145 time 0.2705 (0.2984) loss 3.5883 (3.1630) grad_norm 2.4013 (nan) [2021-04-16 13:55:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][930/1251] eta 0:01:35 lr 0.000145 time 0.2770 (0.2983) loss 3.6136 (3.1647) grad_norm 2.3201 (nan) [2021-04-16 13:55:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][940/1251] eta 0:01:32 lr 0.000145 time 0.2780 (0.2984) loss 2.7735 (3.1622) grad_norm 2.1339 (nan) [2021-04-16 13:55:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][950/1251] eta 0:01:29 lr 0.000145 time 0.2906 (0.2982) loss 2.4347 (3.1607) grad_norm 2.5924 (nan) [2021-04-16 13:55:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][960/1251] eta 0:01:26 lr 0.000145 time 0.2878 (0.2980) loss 3.0093 (3.1620) grad_norm 2.1427 (nan) [2021-04-16 13:55:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][970/1251] eta 0:01:23 lr 0.000145 time 0.2583 (0.2977) loss 2.8023 (3.1615) grad_norm 2.5667 (nan) [2021-04-16 13:55:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][980/1251] eta 0:01:20 lr 0.000145 time 0.2731 (0.2976) loss 2.3479 (3.1600) grad_norm 2.2623 (nan) [2021-04-16 13:56:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][990/1251] eta 0:01:17 lr 0.000145 time 0.2637 (0.2973) loss 2.1597 (3.1610) grad_norm 2.3442 (nan) [2021-04-16 13:56:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][1000/1251] eta 0:01:14 lr 0.000145 time 0.2773 (0.2972) loss 3.4520 (3.1616) grad_norm 2.1664 (nan) [2021-04-16 13:56:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][1010/1251] eta 0:01:11 lr 0.000145 time 0.2779 (0.2970) loss 3.5160 (3.1627) grad_norm 2.4827 (nan) [2021-04-16 13:56:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][1020/1251] eta 0:01:08 lr 0.000145 time 0.2582 (0.2968) loss 2.2887 (3.1613) grad_norm 2.1944 (nan) [2021-04-16 13:56:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][1030/1251] eta 0:01:05 lr 0.000145 time 0.2671 (0.2965) loss 2.7980 (3.1600) grad_norm 2.4179 (nan) [2021-04-16 13:56:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][1040/1251] eta 0:01:02 lr 0.000145 time 0.3100 (0.2965) loss 2.9011 (3.1598) grad_norm 2.6325 (nan) [2021-04-16 13:56:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][1050/1251] eta 0:00:59 lr 0.000145 time 0.2679 (0.2963) loss 3.1139 (3.1578) grad_norm 2.5764 (nan) [2021-04-16 13:56:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][1060/1251] eta 0:00:56 lr 0.000145 time 0.2708 (0.2962) loss 3.6343 (3.1588) grad_norm 2.3432 (nan) [2021-04-16 13:56:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][1070/1251] eta 0:00:53 lr 0.000145 time 0.2743 (0.2960) loss 2.9453 (3.1580) grad_norm 2.3064 (nan) [2021-04-16 13:56:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][1080/1251] eta 0:00:50 lr 0.000145 time 0.2529 (0.2958) loss 2.3877 (3.1587) grad_norm 2.2395 (nan) [2021-04-16 13:56:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][1090/1251] eta 0:00:47 lr 0.000145 time 0.2816 (0.2958) loss 3.6163 (3.1587) grad_norm 2.2399 (nan) [2021-04-16 13:56:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][1100/1251] eta 0:00:44 lr 0.000145 time 0.2811 (0.2956) loss 2.5117 (3.1573) grad_norm 2.1374 (nan) [2021-04-16 13:56:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][1110/1251] eta 0:00:41 lr 0.000145 time 0.2803 (0.2954) loss 3.0260 (3.1576) grad_norm 2.2782 (nan) [2021-04-16 13:56:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][1120/1251] eta 0:00:38 lr 0.000145 time 0.3004 (0.2954) loss 3.2287 (3.1580) grad_norm 2.1525 (nan) [2021-04-16 13:56:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][1130/1251] eta 0:00:35 lr 0.000145 time 0.3086 (0.2954) loss 3.3987 (3.1561) grad_norm 2.2498 (nan) [2021-04-16 13:56:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][1140/1251] eta 0:00:32 lr 0.000144 time 0.2839 (0.2953) loss 2.8617 (3.1577) grad_norm 2.5396 (nan) [2021-04-16 13:56:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][1150/1251] eta 0:00:29 lr 0.000144 time 0.2820 (0.2953) loss 3.1283 (3.1570) grad_norm 2.3030 (nan) [2021-04-16 13:56:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][1160/1251] eta 0:00:26 lr 0.000144 time 0.3057 (0.2953) loss 3.1297 (3.1577) grad_norm 2.0450 (nan) [2021-04-16 13:56:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][1170/1251] eta 0:00:23 lr 0.000144 time 0.2712 (0.2952) loss 3.5398 (3.1557) grad_norm 2.2853 (nan) [2021-04-16 13:56:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][1180/1251] eta 0:00:20 lr 0.000144 time 0.2830 (0.2951) loss 2.9615 (3.1562) grad_norm 2.2020 (nan) [2021-04-16 13:56:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][1190/1251] eta 0:00:17 lr 0.000144 time 0.2850 (0.2949) loss 3.0016 (3.1581) grad_norm 2.1963 (nan) [2021-04-16 13:57:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][1200/1251] eta 0:00:15 lr 0.000144 time 0.2788 (0.2948) loss 3.3460 (3.1580) grad_norm 2.2756 (nan) [2021-04-16 13:57:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][1210/1251] eta 0:00:12 lr 0.000144 time 0.2828 (0.2946) loss 3.5635 (3.1578) grad_norm 2.4715 (nan) [2021-04-16 13:57:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][1220/1251] eta 0:00:09 lr 0.000144 time 0.2593 (0.2946) loss 3.5437 (3.1596) grad_norm 2.7896 (nan) [2021-04-16 13:57:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][1230/1251] eta 0:00:06 lr 0.000144 time 0.2844 (0.2944) loss 2.1021 (3.1582) grad_norm 2.5657 (nan) [2021-04-16 13:57:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][1240/1251] eta 0:00:03 lr 0.000144 time 0.2484 (0.2942) loss 3.0138 (3.1568) grad_norm 2.3479 (nan) [2021-04-16 13:57:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [227/300][1250/1251] eta 0:00:00 lr 0.000144 time 0.2486 (0.2938) loss 2.5809 (3.1560) grad_norm 2.1547 (nan) [2021-04-16 13:57:18 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 227 training takes 0:06:11 [2021-04-16 13:57:18 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_227.pth saving...... [2021-04-16 13:57:27 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_227.pth saved !!! [2021-04-16 13:57:29 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.246 (1.246) Loss 0.8380 (0.8380) Acc@1 80.176 (80.176) Acc@5 95.312 (95.312) [2021-04-16 13:57:30 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.118 (0.217) Loss 0.7869 (0.8382) Acc@1 81.641 (79.528) Acc@5 95.703 (95.233) [2021-04-16 13:57:33 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.352 (0.253) Loss 0.7846 (0.8487) Acc@1 80.859 (79.525) Acc@5 94.922 (94.959) [2021-04-16 13:57:35 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.115 (0.240) Loss 0.8881 (0.8523) Acc@1 77.637 (79.347) Acc@5 94.922 (94.966) [2021-04-16 13:57:36 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.220) Loss 0.7896 (0.8562) Acc@1 79.492 (79.347) Acc@5 95.996 (94.886) [2021-04-16 13:57:43 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 79.458 Acc@5 94.960 [2021-04-16 13:57:43 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 79.5% [2021-04-16 13:57:43 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 79.50% [2021-04-16 13:57:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][0/1251] eta 2:20:14 lr 0.000144 time 6.7265 (6.7265) loss 3.6824 (3.6824) grad_norm 2.1954 (2.1954) [2021-04-16 13:57:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][10/1251] eta 0:18:01 lr 0.000144 time 0.4549 (0.8714) loss 3.6086 (3.3863) grad_norm 2.3125 (2.2880) [2021-04-16 13:57:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][20/1251] eta 0:12:04 lr 0.000144 time 0.2777 (0.5884) loss 3.2173 (3.2414) grad_norm 2.2530 (2.2949) [2021-04-16 13:57:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][30/1251] eta 0:09:58 lr 0.000144 time 0.2864 (0.4903) loss 2.1419 (3.1324) grad_norm 2.3647 (2.2810) [2021-04-16 13:58:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][40/1251] eta 0:08:50 lr 0.000144 time 0.2848 (0.4382) loss 2.6960 (3.1360) grad_norm 2.3769 (2.2650) [2021-04-16 13:58:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][50/1251] eta 0:08:09 lr 0.000144 time 0.2915 (0.4079) loss 3.1598 (3.1544) grad_norm 2.0799 (2.2760) [2021-04-16 13:58:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][60/1251] eta 0:07:40 lr 0.000144 time 0.2730 (0.3871) loss 2.2284 (3.1148) grad_norm 2.5240 (2.2929) [2021-04-16 13:58:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][70/1251] eta 0:07:19 lr 0.000144 time 0.2457 (0.3724) loss 3.4674 (3.1297) grad_norm 2.4433 (2.2947) [2021-04-16 13:58:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][80/1251] eta 0:07:02 lr 0.000144 time 0.3106 (0.3609) loss 3.7312 (3.1213) grad_norm 2.1906 (2.3022) [2021-04-16 13:58:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][90/1251] eta 0:06:50 lr 0.000144 time 0.2884 (0.3539) loss 3.3305 (3.1116) grad_norm 3.1803 (2.3088) [2021-04-16 13:58:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][100/1251] eta 0:06:39 lr 0.000144 time 0.2901 (0.3472) loss 4.1869 (3.1373) grad_norm 2.0346 (2.3028) [2021-04-16 13:58:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][110/1251] eta 0:06:28 lr 0.000144 time 0.2634 (0.3407) loss 3.2897 (3.1467) grad_norm 2.1199 (2.3020) [2021-04-16 13:58:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][120/1251] eta 0:06:20 lr 0.000144 time 0.2807 (0.3366) loss 3.5406 (3.1596) grad_norm 2.5280 (2.3020) [2021-04-16 13:58:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][130/1251] eta 0:06:13 lr 0.000144 time 0.2774 (0.3329) loss 3.1994 (3.1727) grad_norm 2.6228 (2.3120) [2021-04-16 13:58:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][140/1251] eta 0:06:05 lr 0.000144 time 0.2780 (0.3293) loss 2.5300 (3.1700) grad_norm 2.2810 (2.3109) [2021-04-16 13:58:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][150/1251] eta 0:06:00 lr 0.000144 time 0.3043 (0.3277) loss 2.6326 (3.1782) grad_norm 2.7894 (2.3198) [2021-04-16 13:58:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][160/1251] eta 0:05:54 lr 0.000144 time 0.2811 (0.3248) loss 2.9600 (3.1767) grad_norm 2.2716 (2.3277) [2021-04-16 13:58:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][170/1251] eta 0:05:48 lr 0.000144 time 0.2532 (0.3220) loss 3.7034 (3.1659) grad_norm 2.2551 (2.3318) [2021-04-16 13:58:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][180/1251] eta 0:05:42 lr 0.000144 time 0.2752 (0.3197) loss 3.4295 (3.1694) grad_norm 2.4923 (2.3409) [2021-04-16 13:58:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][190/1251] eta 0:05:37 lr 0.000144 time 0.2679 (0.3177) loss 2.5092 (3.1716) grad_norm 2.0441 (2.3385) [2021-04-16 13:58:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][200/1251] eta 0:05:31 lr 0.000144 time 0.2844 (0.3159) loss 3.3847 (3.1794) grad_norm 2.0734 (2.3406) [2021-04-16 13:58:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][210/1251] eta 0:05:27 lr 0.000144 time 0.3006 (0.3145) loss 2.3274 (3.1694) grad_norm 1.9989 (2.3390) [2021-04-16 13:58:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][220/1251] eta 0:05:22 lr 0.000144 time 0.2551 (0.3127) loss 2.8125 (3.1651) grad_norm 2.3443 (2.3369) [2021-04-16 13:58:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][230/1251] eta 0:05:17 lr 0.000144 time 0.2865 (0.3113) loss 3.4651 (3.1616) grad_norm 2.2773 (2.3418) [2021-04-16 13:58:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][240/1251] eta 0:05:13 lr 0.000143 time 0.2851 (0.3102) loss 3.3240 (3.1742) grad_norm 2.3222 (2.3403) [2021-04-16 13:59:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][250/1251] eta 0:05:09 lr 0.000143 time 0.2921 (0.3094) loss 3.0658 (3.1700) grad_norm 2.5912 (2.3431) [2021-04-16 13:59:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][260/1251] eta 0:05:05 lr 0.000143 time 0.2851 (0.3086) loss 3.4223 (3.1837) grad_norm 2.6031 (2.3487) [2021-04-16 13:59:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][270/1251] eta 0:05:01 lr 0.000143 time 0.2849 (0.3075) loss 3.6370 (3.1881) grad_norm 2.1950 (2.3485) [2021-04-16 13:59:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][280/1251] eta 0:04:57 lr 0.000143 time 0.2646 (0.3062) loss 3.5073 (3.1837) grad_norm 2.7257 (2.3493) [2021-04-16 13:59:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][290/1251] eta 0:04:53 lr 0.000143 time 0.2820 (0.3053) loss 3.1469 (3.1866) grad_norm 2.2977 (2.3552) [2021-04-16 13:59:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][300/1251] eta 0:04:49 lr 0.000143 time 0.2676 (0.3048) loss 3.8060 (3.1897) grad_norm 2.0776 (2.3537) [2021-04-16 13:59:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][310/1251] eta 0:04:45 lr 0.000143 time 0.2724 (0.3038) loss 3.9273 (3.1923) grad_norm 2.6932 (2.3584) [2021-04-16 13:59:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][320/1251] eta 0:04:42 lr 0.000143 time 0.3052 (0.3036) loss 3.2318 (3.1870) grad_norm 2.2604 (2.3640) [2021-04-16 13:59:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][330/1251] eta 0:04:38 lr 0.000143 time 0.2715 (0.3029) loss 2.6968 (3.1869) grad_norm 2.4818 (2.3618) [2021-04-16 13:59:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][340/1251] eta 0:04:35 lr 0.000143 time 0.3036 (0.3027) loss 3.8454 (3.1779) grad_norm 2.3201 (2.3598) [2021-04-16 13:59:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][350/1251] eta 0:04:31 lr 0.000143 time 0.2625 (0.3018) loss 2.9654 (3.1787) grad_norm 2.7067 (2.3631) [2021-04-16 13:59:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][360/1251] eta 0:04:29 lr 0.000143 time 0.3201 (0.3020) loss 3.5534 (3.1781) grad_norm 2.5262 (2.3642) [2021-04-16 13:59:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][370/1251] eta 0:04:26 lr 0.000143 time 0.2631 (0.3021) loss 2.8795 (3.1805) grad_norm 2.3941 (2.3627) [2021-04-16 13:59:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][380/1251] eta 0:04:22 lr 0.000143 time 0.2934 (0.3018) loss 3.4594 (3.1800) grad_norm 2.9454 (2.3639) [2021-04-16 13:59:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][390/1251] eta 0:04:19 lr 0.000143 time 0.2797 (0.3012) loss 3.5015 (3.1824) grad_norm 2.1227 (2.3637) [2021-04-16 13:59:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][400/1251] eta 0:04:15 lr 0.000143 time 0.2600 (0.3006) loss 2.5160 (3.1837) grad_norm 2.3858 (2.3638) [2021-04-16 13:59:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][410/1251] eta 0:04:12 lr 0.000143 time 0.2919 (0.3000) loss 3.1236 (3.1848) grad_norm 2.7483 (2.3636) [2021-04-16 13:59:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][420/1251] eta 0:04:09 lr 0.000143 time 0.2831 (0.2997) loss 2.7177 (3.1834) grad_norm 2.7651 (2.3634) [2021-04-16 13:59:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][430/1251] eta 0:04:05 lr 0.000143 time 0.2979 (0.2993) loss 2.9299 (3.1867) grad_norm 2.3455 (2.3692) [2021-04-16 13:59:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][440/1251] eta 0:04:02 lr 0.000143 time 0.2794 (0.2989) loss 3.4655 (3.1872) grad_norm 2.3409 (2.3696) [2021-04-16 13:59:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][450/1251] eta 0:03:59 lr 0.000143 time 0.3019 (0.2985) loss 2.4179 (3.1837) grad_norm 2.7252 (2.3746) [2021-04-16 14:00:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][460/1251] eta 0:03:55 lr 0.000143 time 0.2806 (0.2980) loss 2.7570 (3.1779) grad_norm 3.4127 (2.3803) [2021-04-16 14:00:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][470/1251] eta 0:03:52 lr 0.000143 time 0.2948 (0.2976) loss 3.6370 (3.1751) grad_norm 2.4345 (2.3848) [2021-04-16 14:00:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][480/1251] eta 0:03:49 lr 0.000143 time 0.2787 (0.2973) loss 3.6658 (3.1707) grad_norm 2.3379 (2.3869) [2021-04-16 14:00:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][490/1251] eta 0:03:45 lr 0.000143 time 0.2765 (0.2969) loss 2.9211 (3.1664) grad_norm 2.0519 (2.3888) [2021-04-16 14:00:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][500/1251] eta 0:03:42 lr 0.000143 time 0.2973 (0.2968) loss 2.3563 (3.1641) grad_norm 2.6187 (2.3912) [2021-04-16 14:00:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][510/1251] eta 0:03:39 lr 0.000143 time 0.2993 (0.2965) loss 3.6606 (3.1654) grad_norm 2.1986 (2.3908) [2021-04-16 14:00:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][520/1251] eta 0:03:36 lr 0.000143 time 0.2969 (0.2962) loss 3.4956 (3.1640) grad_norm 2.4610 (2.3912) [2021-04-16 14:00:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][530/1251] eta 0:03:33 lr 0.000143 time 0.2661 (0.2959) loss 3.5300 (3.1653) grad_norm 2.1269 (2.3877) [2021-04-16 14:00:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][540/1251] eta 0:03:30 lr 0.000143 time 0.2706 (0.2960) loss 3.4487 (3.1634) grad_norm 2.1054 (2.3857) [2021-04-16 14:00:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][550/1251] eta 0:03:27 lr 0.000143 time 0.2675 (0.2956) loss 3.4389 (3.1615) grad_norm 2.5577 (2.3872) [2021-04-16 14:00:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][560/1251] eta 0:03:24 lr 0.000143 time 0.2589 (0.2954) loss 2.6999 (3.1604) grad_norm 2.0568 (2.3849) [2021-04-16 14:00:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][570/1251] eta 0:03:21 lr 0.000143 time 0.2982 (0.2954) loss 3.7043 (3.1661) grad_norm 2.5307 (2.3816) [2021-04-16 14:00:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][580/1251] eta 0:03:18 lr 0.000143 time 0.2830 (0.2955) loss 3.9436 (3.1723) grad_norm 2.7772 (2.3814) [2021-04-16 14:00:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][590/1251] eta 0:03:15 lr 0.000142 time 0.2756 (0.2955) loss 3.4661 (3.1738) grad_norm 2.5268 (2.3842) [2021-04-16 14:00:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][600/1251] eta 0:03:12 lr 0.000142 time 0.2774 (0.2952) loss 3.5462 (3.1733) grad_norm 2.3350 (2.3831) [2021-04-16 14:00:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][610/1251] eta 0:03:09 lr 0.000142 time 0.2539 (0.2950) loss 3.9865 (3.1797) grad_norm 2.4924 (2.3836) [2021-04-16 14:00:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][620/1251] eta 0:03:05 lr 0.000142 time 0.2816 (0.2947) loss 3.8457 (3.1784) grad_norm 2.2584 (2.3823) [2021-04-16 14:00:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][630/1251] eta 0:03:02 lr 0.000142 time 0.2758 (0.2944) loss 2.4870 (3.1796) grad_norm 1.9369 (2.3812) [2021-04-16 14:00:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][640/1251] eta 0:02:59 lr 0.000142 time 0.2923 (0.2941) loss 3.3691 (3.1803) grad_norm 2.3402 (2.3789) [2021-04-16 14:00:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][650/1251] eta 0:02:56 lr 0.000142 time 0.3025 (0.2940) loss 2.8978 (3.1795) grad_norm 2.0164 (2.3806) [2021-04-16 14:00:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][660/1251] eta 0:02:53 lr 0.000142 time 0.2845 (0.2937) loss 3.3603 (3.1800) grad_norm 2.1972 (2.3833) [2021-04-16 14:01:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][670/1251] eta 0:02:50 lr 0.000142 time 0.2739 (0.2935) loss 3.6456 (3.1831) grad_norm 2.3019 (2.3837) [2021-04-16 14:01:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][680/1251] eta 0:02:47 lr 0.000142 time 0.2856 (0.2933) loss 3.6158 (3.1847) grad_norm 2.4642 (2.3828) [2021-04-16 14:01:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][690/1251] eta 0:02:44 lr 0.000142 time 0.2790 (0.2932) loss 3.1412 (3.1866) grad_norm 2.3100 (2.3833) [2021-04-16 14:01:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][700/1251] eta 0:02:41 lr 0.000142 time 0.4162 (0.2932) loss 2.6428 (3.1839) grad_norm 2.2766 (2.3819) [2021-04-16 14:01:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][710/1251] eta 0:02:38 lr 0.000142 time 0.2541 (0.2931) loss 2.3564 (3.1868) grad_norm 2.0699 (2.3818) [2021-04-16 14:01:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][720/1251] eta 0:02:35 lr 0.000142 time 0.2644 (0.2930) loss 3.4026 (3.1848) grad_norm 2.4831 (2.3826) [2021-04-16 14:01:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][730/1251] eta 0:02:32 lr 0.000142 time 0.2902 (0.2930) loss 3.0492 (3.1823) grad_norm 2.5720 (2.3822) [2021-04-16 14:01:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][740/1251] eta 0:02:29 lr 0.000142 time 0.2719 (0.2927) loss 2.3222 (3.1829) grad_norm 2.0988 (2.3837) [2021-04-16 14:01:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][750/1251] eta 0:02:26 lr 0.000142 time 0.2809 (0.2926) loss 3.7588 (3.1853) grad_norm 2.2484 (2.3832) [2021-04-16 14:01:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][760/1251] eta 0:02:23 lr 0.000142 time 0.2781 (0.2923) loss 3.1877 (3.1851) grad_norm 2.4015 (2.3845) [2021-04-16 14:01:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][770/1251] eta 0:02:20 lr 0.000142 time 0.2611 (0.2922) loss 3.0465 (3.1849) grad_norm 2.1971 (2.3835) [2021-04-16 14:01:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][780/1251] eta 0:02:17 lr 0.000142 time 0.2890 (0.2920) loss 3.4177 (3.1805) grad_norm 2.1181 (2.3815) [2021-04-16 14:01:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][790/1251] eta 0:02:14 lr 0.000142 time 0.2720 (0.2918) loss 2.5796 (3.1813) grad_norm 1.9835 (2.3825) [2021-04-16 14:01:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][800/1251] eta 0:02:11 lr 0.000142 time 0.2777 (0.2916) loss 3.6638 (3.1829) grad_norm 2.5618 (2.3829) [2021-04-16 14:01:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][810/1251] eta 0:02:08 lr 0.000142 time 0.2796 (0.2915) loss 2.4978 (3.1816) grad_norm 2.2388 (2.3815) [2021-04-16 14:01:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][820/1251] eta 0:02:05 lr 0.000142 time 0.2681 (0.2913) loss 3.2095 (3.1832) grad_norm 2.4176 (2.3810) [2021-04-16 14:01:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][830/1251] eta 0:02:02 lr 0.000142 time 0.2767 (0.2912) loss 2.9663 (3.1837) grad_norm 2.4588 (2.3810) [2021-04-16 14:01:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][840/1251] eta 0:01:59 lr 0.000142 time 0.2771 (0.2910) loss 3.2939 (3.1834) grad_norm 2.2268 (2.3821) [2021-04-16 14:01:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][850/1251] eta 0:01:56 lr 0.000142 time 0.2481 (0.2908) loss 2.8628 (3.1826) grad_norm 2.7416 (2.3825) [2021-04-16 14:01:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][860/1251] eta 0:01:53 lr 0.000142 time 0.2772 (0.2906) loss 3.3030 (3.1816) grad_norm 2.7003 (2.3855) [2021-04-16 14:01:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][870/1251] eta 0:01:50 lr 0.000142 time 0.3023 (0.2905) loss 3.5109 (3.1784) grad_norm 2.3866 (2.3861) [2021-04-16 14:01:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][880/1251] eta 0:01:47 lr 0.000142 time 0.2857 (0.2905) loss 3.4776 (3.1788) grad_norm 2.2814 (2.3847) [2021-04-16 14:02:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][890/1251] eta 0:01:44 lr 0.000142 time 0.2702 (0.2905) loss 3.2987 (3.1784) grad_norm 2.3650 (2.3833) [2021-04-16 14:02:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][900/1251] eta 0:01:41 lr 0.000142 time 0.2469 (0.2903) loss 3.5977 (3.1767) grad_norm 2.2124 (2.3844) [2021-04-16 14:02:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][910/1251] eta 0:01:38 lr 0.000142 time 0.2611 (0.2902) loss 2.8370 (3.1759) grad_norm 2.4533 (2.3847) [2021-04-16 14:02:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][920/1251] eta 0:01:36 lr 0.000142 time 0.2530 (0.2902) loss 2.3605 (3.1770) grad_norm 2.7357 (2.3845) [2021-04-16 14:02:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][930/1251] eta 0:01:33 lr 0.000142 time 0.2874 (0.2902) loss 2.8820 (3.1796) grad_norm 2.7799 (2.3845) [2021-04-16 14:02:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][940/1251] eta 0:01:30 lr 0.000142 time 0.2785 (0.2902) loss 3.5058 (3.1787) grad_norm 2.2364 (2.3871) [2021-04-16 14:02:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][950/1251] eta 0:01:27 lr 0.000141 time 0.2714 (0.2901) loss 3.5819 (3.1794) grad_norm 2.6023 (2.3895) [2021-04-16 14:02:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][960/1251] eta 0:01:24 lr 0.000141 time 0.2687 (0.2901) loss 3.2089 (3.1792) grad_norm 2.3590 (2.3902) [2021-04-16 14:02:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][970/1251] eta 0:01:21 lr 0.000141 time 0.2823 (0.2899) loss 3.6504 (3.1791) grad_norm 2.3535 (2.3885) [2021-04-16 14:02:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][980/1251] eta 0:01:18 lr 0.000141 time 0.2629 (0.2898) loss 3.8560 (3.1771) grad_norm 2.8869 (2.3894) [2021-04-16 14:02:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][990/1251] eta 0:01:15 lr 0.000141 time 0.2894 (0.2897) loss 3.2999 (3.1783) grad_norm 2.2919 (2.3919) [2021-04-16 14:02:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][1000/1251] eta 0:01:12 lr 0.000141 time 0.2598 (0.2896) loss 3.1247 (3.1785) grad_norm 2.2684 (2.3907) [2021-04-16 14:02:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][1010/1251] eta 0:01:09 lr 0.000141 time 0.3144 (0.2895) loss 3.1666 (3.1799) grad_norm 2.1660 (2.3914) [2021-04-16 14:02:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][1020/1251] eta 0:01:06 lr 0.000141 time 0.2739 (0.2895) loss 2.5708 (3.1807) grad_norm 2.4137 (2.3917) [2021-04-16 14:02:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][1030/1251] eta 0:01:03 lr 0.000141 time 0.2690 (0.2893) loss 2.5448 (3.1780) grad_norm 2.6573 (2.3927) [2021-04-16 14:02:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][1040/1251] eta 0:01:01 lr 0.000141 time 0.2577 (0.2893) loss 2.7572 (3.1802) grad_norm 2.4926 (2.3928) [2021-04-16 14:02:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][1050/1251] eta 0:00:58 lr 0.000141 time 0.2712 (0.2891) loss 3.5714 (3.1812) grad_norm 2.4589 (2.3960) [2021-04-16 14:02:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][1060/1251] eta 0:00:55 lr 0.000141 time 0.2930 (0.2891) loss 3.5142 (3.1807) grad_norm 2.5780 (2.3959) [2021-04-16 14:02:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][1070/1251] eta 0:00:52 lr 0.000141 time 0.2799 (0.2890) loss 4.0198 (3.1824) grad_norm 2.4732 (2.3967) [2021-04-16 14:02:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][1080/1251] eta 0:00:49 lr 0.000141 time 0.3014 (0.2889) loss 3.3849 (3.1838) grad_norm 2.2457 (2.3959) [2021-04-16 14:02:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][1090/1251] eta 0:00:46 lr 0.000141 time 0.2651 (0.2888) loss 3.0605 (3.1848) grad_norm 2.5447 (2.3957) [2021-04-16 14:03:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][1100/1251] eta 0:00:43 lr 0.000141 time 0.2744 (0.2887) loss 3.3748 (3.1866) grad_norm 2.5976 (2.3968) [2021-04-16 14:03:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][1110/1251] eta 0:00:40 lr 0.000141 time 0.2700 (0.2886) loss 3.5527 (3.1862) grad_norm 2.6622 (2.3962) [2021-04-16 14:03:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][1120/1251] eta 0:00:37 lr 0.000141 time 0.2944 (0.2886) loss 3.2707 (3.1846) grad_norm 2.4409 (2.3960) [2021-04-16 14:03:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][1130/1251] eta 0:00:34 lr 0.000141 time 0.2894 (0.2886) loss 2.5620 (3.1857) grad_norm 2.3060 (2.3956) [2021-04-16 14:03:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][1140/1251] eta 0:00:32 lr 0.000141 time 0.2765 (0.2886) loss 2.8569 (3.1847) grad_norm 2.1120 (2.3966) [2021-04-16 14:03:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][1150/1251] eta 0:00:29 lr 0.000141 time 0.2749 (0.2885) loss 2.9341 (3.1850) grad_norm 2.3896 (2.3958) [2021-04-16 14:03:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][1160/1251] eta 0:00:26 lr 0.000141 time 0.2596 (0.2885) loss 2.9918 (3.1848) grad_norm 2.4366 (2.3955) [2021-04-16 14:03:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][1170/1251] eta 0:00:23 lr 0.000141 time 0.2875 (0.2886) loss 3.7999 (3.1833) grad_norm 2.5076 (2.3945) [2021-04-16 14:03:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][1180/1251] eta 0:00:20 lr 0.000141 time 0.2709 (0.2885) loss 3.7030 (3.1857) grad_norm 2.1889 (2.3944) [2021-04-16 14:03:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][1190/1251] eta 0:00:17 lr 0.000141 time 0.2572 (0.2884) loss 3.7056 (3.1851) grad_norm 2.2420 (2.3957) [2021-04-16 14:03:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][1200/1251] eta 0:00:14 lr 0.000141 time 0.2631 (0.2884) loss 3.8310 (3.1852) grad_norm 2.6327 (2.3949) [2021-04-16 14:03:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][1210/1251] eta 0:00:11 lr 0.000141 time 0.2948 (0.2882) loss 3.7268 (3.1864) grad_norm 2.5465 (2.3961) [2021-04-16 14:03:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][1220/1251] eta 0:00:08 lr 0.000141 time 0.2593 (0.2881) loss 3.2741 (3.1870) grad_norm 2.9247 (2.3959) [2021-04-16 14:03:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][1230/1251] eta 0:00:06 lr 0.000141 time 0.2653 (0.2880) loss 3.4115 (3.1886) grad_norm 2.5109 (2.3963) [2021-04-16 14:03:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][1240/1251] eta 0:00:03 lr 0.000141 time 0.3208 (0.2879) loss 3.9921 (3.1899) grad_norm 2.4006 (2.3974) [2021-04-16 14:03:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [228/300][1250/1251] eta 0:00:00 lr 0.000141 time 0.2484 (0.2876) loss 2.0712 (3.1888) grad_norm 2.3757 (2.3982) [2021-04-16 14:03:46 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 228 training takes 0:06:03 [2021-04-16 14:03:46 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_228.pth saving...... [2021-04-16 14:03:54 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_228.pth saved !!! [2021-04-16 14:03:55 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.163 (1.163) Loss 0.9382 (0.9382) Acc@1 79.492 (79.492) Acc@5 93.945 (93.945) [2021-04-16 14:03:57 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.549 (0.289) Loss 0.8818 (0.8622) Acc@1 80.566 (80.060) Acc@5 94.531 (94.877) [2021-04-16 14:03:59 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.516 (0.231) Loss 0.8295 (0.8610) Acc@1 80.762 (80.036) Acc@5 95.020 (94.871) [2021-04-16 14:04:01 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.119 (0.229) Loss 0.8619 (0.8638) Acc@1 79.980 (79.921) Acc@5 94.727 (94.871) [2021-04-16 14:04:03 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.078 (0.203) Loss 0.8690 (0.8646) Acc@1 79.590 (79.809) Acc@5 94.629 (94.874) [2021-04-16 14:04:08 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 79.694 Acc@5 94.864 [2021-04-16 14:04:08 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 79.7% [2021-04-16 14:04:08 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 79.69% [2021-04-16 14:04:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][0/1251] eta 2:51:07 lr 0.000141 time 8.2076 (8.2076) loss 3.5397 (3.5397) grad_norm 2.3680 (2.3680) [2021-04-16 14:04:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][10/1251] eta 0:20:34 lr 0.000141 time 0.2762 (0.9946) loss 2.8379 (3.2451) grad_norm 2.2270 (2.5100) [2021-04-16 14:04:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][20/1251] eta 0:13:20 lr 0.000141 time 0.2602 (0.6506) loss 3.5078 (3.2323) grad_norm 2.2325 (2.4556) [2021-04-16 14:04:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][30/1251] eta 0:10:47 lr 0.000141 time 0.2513 (0.5304) loss 3.8972 (3.2553) grad_norm 2.6211 (2.4542) [2021-04-16 14:04:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][40/1251] eta 0:09:28 lr 0.000141 time 0.2878 (0.4696) loss 3.2570 (3.1749) grad_norm 2.3179 (2.4444) [2021-04-16 14:04:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][50/1251] eta 0:08:38 lr 0.000140 time 0.3061 (0.4319) loss 3.4898 (3.1820) grad_norm 2.6206 (2.4303) [2021-04-16 14:04:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][60/1251] eta 0:08:04 lr 0.000140 time 0.3024 (0.4070) loss 2.7020 (3.1794) grad_norm 2.4232 (2.4128) [2021-04-16 14:04:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][70/1251] eta 0:07:38 lr 0.000140 time 0.2868 (0.3880) loss 3.0253 (3.1485) grad_norm 2.3009 (2.4211) [2021-04-16 14:04:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][80/1251] eta 0:07:20 lr 0.000140 time 0.2878 (0.3761) loss 2.4771 (3.1560) grad_norm 2.7779 (2.4245) [2021-04-16 14:04:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][90/1251] eta 0:07:04 lr 0.000140 time 0.3005 (0.3654) loss 3.2663 (3.1762) grad_norm 2.1396 (2.4104) [2021-04-16 14:04:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][100/1251] eta 0:06:52 lr 0.000140 time 0.2888 (0.3584) loss 4.0074 (3.1868) grad_norm 3.1704 (2.4223) [2021-04-16 14:04:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][110/1251] eta 0:06:40 lr 0.000140 time 0.2902 (0.3508) loss 3.5218 (3.1938) grad_norm 2.2827 (2.4222) [2021-04-16 14:04:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][120/1251] eta 0:06:30 lr 0.000140 time 0.2463 (0.3456) loss 3.5246 (3.2190) grad_norm 2.2362 (2.4183) [2021-04-16 14:04:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][130/1251] eta 0:06:21 lr 0.000140 time 0.2627 (0.3401) loss 2.1223 (3.2073) grad_norm 2.3543 (2.4129) [2021-04-16 14:04:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][140/1251] eta 0:06:14 lr 0.000140 time 0.2636 (0.3371) loss 3.4688 (3.2134) grad_norm 2.2156 (2.4013) [2021-04-16 14:04:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][150/1251] eta 0:06:06 lr 0.000140 time 0.2777 (0.3330) loss 3.6414 (3.1976) grad_norm 2.0554 (2.3949) [2021-04-16 14:05:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][160/1251] eta 0:06:00 lr 0.000140 time 0.2745 (0.3306) loss 3.4739 (3.2101) grad_norm 2.3351 (2.4076) [2021-04-16 14:05:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][170/1251] eta 0:05:54 lr 0.000140 time 0.2573 (0.3280) loss 3.2450 (3.2068) grad_norm 2.3802 (2.4060) [2021-04-16 14:05:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][180/1251] eta 0:05:48 lr 0.000140 time 0.2816 (0.3253) loss 3.7221 (3.2035) grad_norm 2.5195 (2.4011) [2021-04-16 14:05:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][190/1251] eta 0:05:42 lr 0.000140 time 0.2429 (0.3226) loss 2.8424 (3.2007) grad_norm 2.1891 (2.3991) [2021-04-16 14:05:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][200/1251] eta 0:05:36 lr 0.000140 time 0.2749 (0.3204) loss 3.2667 (3.2045) grad_norm 2.2130 (2.4006) [2021-04-16 14:05:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][210/1251] eta 0:05:31 lr 0.000140 time 0.2683 (0.3181) loss 3.2663 (3.2143) grad_norm 2.7222 (2.4004) [2021-04-16 14:05:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][220/1251] eta 0:05:26 lr 0.000140 time 0.2934 (0.3163) loss 3.7216 (3.2156) grad_norm 2.3069 (2.3968) [2021-04-16 14:05:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][230/1251] eta 0:05:20 lr 0.000140 time 0.2605 (0.3144) loss 3.3326 (3.2124) grad_norm 2.0920 (2.3950) [2021-04-16 14:05:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][240/1251] eta 0:05:16 lr 0.000140 time 0.2820 (0.3127) loss 3.3313 (3.2086) grad_norm 2.3085 (2.3978) [2021-04-16 14:05:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][250/1251] eta 0:05:11 lr 0.000140 time 0.3018 (0.3113) loss 2.3889 (3.2061) grad_norm 2.3446 (2.3966) [2021-04-16 14:05:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][260/1251] eta 0:05:07 lr 0.000140 time 0.2706 (0.3100) loss 2.9131 (3.2114) grad_norm 2.3816 (2.4024) [2021-04-16 14:05:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][270/1251] eta 0:05:02 lr 0.000140 time 0.2748 (0.3087) loss 2.1656 (3.2147) grad_norm 2.2418 (2.4012) [2021-04-16 14:05:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][280/1251] eta 0:04:58 lr 0.000140 time 0.2984 (0.3076) loss 3.6263 (3.2122) grad_norm 2.4589 (2.4041) [2021-04-16 14:05:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][290/1251] eta 0:04:54 lr 0.000140 time 0.2705 (0.3066) loss 2.7593 (3.2221) grad_norm 2.5671 (2.4068) [2021-04-16 14:05:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][300/1251] eta 0:04:50 lr 0.000140 time 0.2760 (0.3055) loss 3.1992 (3.2192) grad_norm 2.3608 (2.4079) [2021-04-16 14:05:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][310/1251] eta 0:04:46 lr 0.000140 time 0.2751 (0.3048) loss 3.3500 (3.2086) grad_norm 2.2962 (2.4033) [2021-04-16 14:05:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][320/1251] eta 0:04:43 lr 0.000140 time 0.2493 (0.3044) loss 2.0326 (3.2077) grad_norm 2.2864 (2.3994) [2021-04-16 14:05:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][330/1251] eta 0:04:39 lr 0.000140 time 0.2908 (0.3034) loss 3.4760 (3.2137) grad_norm 2.6128 (2.4010) [2021-04-16 14:05:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][340/1251] eta 0:04:35 lr 0.000140 time 0.2808 (0.3027) loss 2.6801 (3.2023) grad_norm 2.2811 (2.4147) [2021-04-16 14:05:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][350/1251] eta 0:04:32 lr 0.000140 time 0.2882 (0.3023) loss 3.4840 (3.1917) grad_norm 2.3056 (2.4189) [2021-04-16 14:05:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][360/1251] eta 0:04:28 lr 0.000140 time 0.2748 (0.3015) loss 3.3046 (3.1967) grad_norm 2.3583 (2.4165) [2021-04-16 14:06:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][370/1251] eta 0:04:25 lr 0.000140 time 0.2640 (0.3013) loss 2.7401 (3.1916) grad_norm 2.2143 (2.4117) [2021-04-16 14:06:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][380/1251] eta 0:04:22 lr 0.000140 time 0.2725 (0.3008) loss 2.2033 (3.1848) grad_norm 2.5694 (2.4112) [2021-04-16 14:06:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][390/1251] eta 0:04:18 lr 0.000140 time 0.2692 (0.3002) loss 3.7924 (3.1951) grad_norm 2.5195 (2.4086) [2021-04-16 14:06:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][400/1251] eta 0:04:15 lr 0.000140 time 0.2859 (0.2997) loss 3.5773 (3.1977) grad_norm 2.7827 (2.4087) [2021-04-16 14:06:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][410/1251] eta 0:04:11 lr 0.000139 time 0.2948 (0.2991) loss 2.1307 (3.1960) grad_norm 2.8144 (2.4105) [2021-04-16 14:06:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][420/1251] eta 0:04:08 lr 0.000139 time 0.2622 (0.2986) loss 3.9432 (3.2043) grad_norm 2.6131 (2.4114) [2021-04-16 14:06:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][430/1251] eta 0:04:04 lr 0.000139 time 0.2568 (0.2981) loss 3.6717 (3.2025) grad_norm 2.0975 (2.4085) [2021-04-16 14:06:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][440/1251] eta 0:04:01 lr 0.000139 time 0.2882 (0.2978) loss 3.8528 (3.2028) grad_norm 2.5654 (2.4080) [2021-04-16 14:06:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][450/1251] eta 0:03:58 lr 0.000139 time 0.2794 (0.2974) loss 3.0920 (3.1988) grad_norm 2.1274 (2.4105) [2021-04-16 14:06:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][460/1251] eta 0:03:54 lr 0.000139 time 0.2962 (0.2970) loss 3.7349 (3.1952) grad_norm 2.8175 (2.4083) [2021-04-16 14:06:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][470/1251] eta 0:03:51 lr 0.000139 time 0.2791 (0.2966) loss 3.4661 (3.1949) grad_norm 2.4046 (2.4071) [2021-04-16 14:06:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][480/1251] eta 0:03:48 lr 0.000139 time 0.2836 (0.2962) loss 3.3980 (3.1933) grad_norm 2.3134 (2.4061) [2021-04-16 14:06:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][490/1251] eta 0:03:45 lr 0.000139 time 0.2742 (0.2960) loss 3.9641 (3.1979) grad_norm 3.6236 (2.4081) [2021-04-16 14:06:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][500/1251] eta 0:03:41 lr 0.000139 time 0.2660 (0.2956) loss 2.8672 (3.1980) grad_norm 9.8422 (2.4221) [2021-04-16 14:06:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][510/1251] eta 0:03:38 lr 0.000139 time 0.2935 (0.2952) loss 3.7200 (3.1993) grad_norm 2.8077 (2.4225) [2021-04-16 14:06:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][520/1251] eta 0:03:35 lr 0.000139 time 0.2793 (0.2949) loss 3.5055 (3.2047) grad_norm 2.3816 (2.4243) [2021-04-16 14:06:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][530/1251] eta 0:03:32 lr 0.000139 time 0.2757 (0.2946) loss 2.8816 (3.2001) grad_norm 2.4662 (2.4229) [2021-04-16 14:06:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][540/1251] eta 0:03:29 lr 0.000139 time 0.2512 (0.2942) loss 3.0486 (3.1999) grad_norm 2.2763 (2.4207) [2021-04-16 14:06:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][550/1251] eta 0:03:26 lr 0.000139 time 0.2690 (0.2940) loss 3.4838 (3.2012) grad_norm 2.2440 (2.4222) [2021-04-16 14:06:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][560/1251] eta 0:03:22 lr 0.000139 time 0.2663 (0.2936) loss 3.8181 (3.2069) grad_norm 2.3731 (inf) [2021-04-16 14:06:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][570/1251] eta 0:03:19 lr 0.000139 time 0.3573 (0.2936) loss 3.3533 (3.2066) grad_norm 2.5995 (inf) [2021-04-16 14:06:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][580/1251] eta 0:03:16 lr 0.000139 time 0.2702 (0.2934) loss 3.7628 (3.2049) grad_norm 2.3387 (inf) [2021-04-16 14:07:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][590/1251] eta 0:03:13 lr 0.000139 time 0.2616 (0.2931) loss 2.4951 (3.2062) grad_norm 2.4118 (inf) [2021-04-16 14:07:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][600/1251] eta 0:03:10 lr 0.000139 time 0.2594 (0.2929) loss 3.1396 (3.2092) grad_norm 2.0856 (inf) [2021-04-16 14:07:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][610/1251] eta 0:03:07 lr 0.000139 time 0.2930 (0.2926) loss 3.0760 (3.2108) grad_norm 2.1849 (inf) [2021-04-16 14:07:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][620/1251] eta 0:03:04 lr 0.000139 time 0.2615 (0.2923) loss 2.5446 (3.2097) grad_norm 2.6528 (inf) [2021-04-16 14:07:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][630/1251] eta 0:03:01 lr 0.000139 time 0.2495 (0.2920) loss 3.0029 (3.2058) grad_norm 2.4695 (inf) [2021-04-16 14:07:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][640/1251] eta 0:02:58 lr 0.000139 time 0.2655 (0.2918) loss 3.8148 (3.2063) grad_norm 2.2263 (inf) [2021-04-16 14:07:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][650/1251] eta 0:02:55 lr 0.000139 time 0.3846 (0.2917) loss 3.3058 (3.2064) grad_norm 2.4790 (inf) [2021-04-16 14:07:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][660/1251] eta 0:02:52 lr 0.000139 time 0.2746 (0.2915) loss 2.6940 (3.2008) grad_norm 2.4770 (inf) [2021-04-16 14:07:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][670/1251] eta 0:02:49 lr 0.000139 time 0.2853 (0.2913) loss 2.7998 (3.2032) grad_norm 2.4394 (inf) [2021-04-16 14:07:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][680/1251] eta 0:02:46 lr 0.000139 time 0.2941 (0.2912) loss 3.6971 (3.1994) grad_norm 2.1536 (inf) [2021-04-16 14:07:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][690/1251] eta 0:02:43 lr 0.000139 time 0.2578 (0.2911) loss 2.5378 (3.1967) grad_norm 2.5415 (inf) [2021-04-16 14:07:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][700/1251] eta 0:02:40 lr 0.000139 time 0.2791 (0.2908) loss 3.7184 (3.1958) grad_norm 3.4411 (inf) [2021-04-16 14:07:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][710/1251] eta 0:02:37 lr 0.000139 time 0.2932 (0.2906) loss 2.8818 (3.1970) grad_norm 2.1888 (inf) [2021-04-16 14:07:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][720/1251] eta 0:02:34 lr 0.000139 time 0.2502 (0.2907) loss 3.1220 (3.1997) grad_norm 2.5010 (inf) [2021-04-16 14:07:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][730/1251] eta 0:02:31 lr 0.000139 time 0.2994 (0.2906) loss 2.1665 (3.1984) grad_norm 1.9933 (inf) [2021-04-16 14:07:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][740/1251] eta 0:02:28 lr 0.000139 time 0.2749 (0.2907) loss 3.7468 (3.1986) grad_norm 2.2874 (inf) [2021-04-16 14:07:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][750/1251] eta 0:02:25 lr 0.000139 time 0.2879 (0.2905) loss 2.5645 (3.1955) grad_norm 2.1862 (inf) [2021-04-16 14:07:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][760/1251] eta 0:02:22 lr 0.000139 time 0.2733 (0.2903) loss 3.4270 (3.1943) grad_norm 2.5125 (inf) [2021-04-16 14:07:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][770/1251] eta 0:02:19 lr 0.000138 time 0.2575 (0.2900) loss 3.3162 (3.1930) grad_norm 2.0885 (inf) [2021-04-16 14:07:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][780/1251] eta 0:02:16 lr 0.000138 time 0.2547 (0.2898) loss 3.6671 (3.1968) grad_norm 2.2819 (inf) [2021-04-16 14:07:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][790/1251] eta 0:02:13 lr 0.000138 time 0.2765 (0.2897) loss 2.4948 (3.1960) grad_norm 2.1586 (inf) [2021-04-16 14:08:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][800/1251] eta 0:02:10 lr 0.000138 time 0.2896 (0.2895) loss 2.7586 (3.1970) grad_norm 2.2742 (inf) [2021-04-16 14:08:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][810/1251] eta 0:02:07 lr 0.000138 time 0.2786 (0.2893) loss 2.3043 (3.1950) grad_norm 2.6614 (inf) [2021-04-16 14:08:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][820/1251] eta 0:02:04 lr 0.000138 time 0.2659 (0.2892) loss 3.8933 (3.1976) grad_norm 2.5740 (inf) [2021-04-16 14:08:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][830/1251] eta 0:02:01 lr 0.000138 time 0.2784 (0.2890) loss 3.5142 (3.1988) grad_norm 2.1724 (inf) [2021-04-16 14:08:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][840/1251] eta 0:01:58 lr 0.000138 time 0.2933 (0.2889) loss 3.7460 (3.1977) grad_norm 2.5623 (inf) [2021-04-16 14:08:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][850/1251] eta 0:01:55 lr 0.000138 time 0.2855 (0.2887) loss 2.9836 (3.1965) grad_norm 2.2047 (inf) [2021-04-16 14:08:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][860/1251] eta 0:01:52 lr 0.000138 time 0.2931 (0.2886) loss 3.4279 (3.1977) grad_norm 2.1973 (inf) [2021-04-16 14:08:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][870/1251] eta 0:01:49 lr 0.000138 time 0.2499 (0.2884) loss 3.2771 (3.1973) grad_norm 2.1877 (inf) [2021-04-16 14:08:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][880/1251] eta 0:01:46 lr 0.000138 time 0.2808 (0.2883) loss 3.4263 (3.1985) grad_norm 2.1789 (inf) [2021-04-16 14:08:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][890/1251] eta 0:01:44 lr 0.000138 time 0.2764 (0.2881) loss 3.4917 (3.2007) grad_norm 2.3721 (inf) [2021-04-16 14:08:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][900/1251] eta 0:01:41 lr 0.000138 time 0.3103 (0.2881) loss 3.5800 (3.2003) grad_norm 2.2862 (inf) [2021-04-16 14:08:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][910/1251] eta 0:01:38 lr 0.000138 time 0.2737 (0.2880) loss 3.1733 (3.2005) grad_norm 2.3478 (inf) [2021-04-16 14:08:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][920/1251] eta 0:01:35 lr 0.000138 time 0.2770 (0.2879) loss 3.8521 (3.2021) grad_norm 2.5248 (inf) [2021-04-16 14:08:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][930/1251] eta 0:01:32 lr 0.000138 time 0.2870 (0.2879) loss 3.3479 (3.1997) grad_norm 2.1356 (inf) [2021-04-16 14:08:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][940/1251] eta 0:01:29 lr 0.000138 time 0.2593 (0.2879) loss 3.0706 (3.2002) grad_norm 2.7814 (inf) [2021-04-16 14:08:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][950/1251] eta 0:01:26 lr 0.000138 time 0.2705 (0.2878) loss 3.9000 (3.2029) grad_norm 2.5779 (inf) [2021-04-16 14:08:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][960/1251] eta 0:01:23 lr 0.000138 time 0.2999 (0.2876) loss 3.7089 (3.2032) grad_norm 2.4002 (inf) [2021-04-16 14:08:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][970/1251] eta 0:01:20 lr 0.000138 time 0.2706 (0.2876) loss 2.7758 (3.2001) grad_norm 2.6242 (inf) [2021-04-16 14:08:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][980/1251] eta 0:01:17 lr 0.000138 time 0.2690 (0.2875) loss 3.1982 (3.2019) grad_norm 2.3261 (inf) [2021-04-16 14:08:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][990/1251] eta 0:01:14 lr 0.000138 time 0.2776 (0.2873) loss 2.1690 (3.2016) grad_norm 2.5066 (inf) [2021-04-16 14:08:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][1000/1251] eta 0:01:12 lr 0.000138 time 0.2529 (0.2872) loss 3.0840 (3.2011) grad_norm 2.3586 (inf) [2021-04-16 14:08:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][1010/1251] eta 0:01:09 lr 0.000138 time 0.2881 (0.2871) loss 3.7179 (3.2017) grad_norm 2.2199 (inf) [2021-04-16 14:09:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][1020/1251] eta 0:01:06 lr 0.000138 time 0.2671 (0.2871) loss 3.2782 (3.2021) grad_norm 2.1908 (inf) [2021-04-16 14:09:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][1030/1251] eta 0:01:03 lr 0.000138 time 0.3027 (0.2870) loss 4.2256 (3.2011) grad_norm 2.4362 (inf) [2021-04-16 14:09:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][1040/1251] eta 0:01:00 lr 0.000138 time 0.2726 (0.2868) loss 3.9375 (3.2036) grad_norm 2.1523 (inf) [2021-04-16 14:09:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][1050/1251] eta 0:00:57 lr 0.000138 time 0.3331 (0.2868) loss 3.0793 (3.2038) grad_norm 2.5446 (inf) [2021-04-16 14:09:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][1060/1251] eta 0:00:54 lr 0.000138 time 0.2901 (0.2867) loss 2.3576 (3.2015) grad_norm 2.2985 (inf) [2021-04-16 14:09:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][1070/1251] eta 0:00:51 lr 0.000138 time 0.2703 (0.2867) loss 2.9614 (3.2007) grad_norm 2.4145 (inf) [2021-04-16 14:09:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][1080/1251] eta 0:00:49 lr 0.000138 time 0.2754 (0.2866) loss 2.5410 (3.1996) grad_norm 2.5840 (inf) [2021-04-16 14:09:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][1090/1251] eta 0:00:46 lr 0.000138 time 0.2895 (0.2865) loss 3.8634 (3.1994) grad_norm 2.7963 (inf) [2021-04-16 14:09:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][1100/1251] eta 0:00:43 lr 0.000138 time 0.2706 (0.2865) loss 2.2435 (3.1988) grad_norm 2.2250 (inf) [2021-04-16 14:09:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][1110/1251] eta 0:00:40 lr 0.000138 time 0.2849 (0.2864) loss 3.4197 (3.1993) grad_norm 2.3419 (inf) [2021-04-16 14:09:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][1120/1251] eta 0:00:37 lr 0.000138 time 0.2694 (0.2863) loss 3.8097 (3.2002) grad_norm 2.5245 (inf) [2021-04-16 14:09:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][1130/1251] eta 0:00:34 lr 0.000137 time 0.2554 (0.2862) loss 3.9189 (3.2016) grad_norm 2.7121 (inf) [2021-04-16 14:09:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][1140/1251] eta 0:00:31 lr 0.000137 time 0.2840 (0.2861) loss 3.3519 (3.2003) grad_norm 2.0411 (inf) [2021-04-16 14:09:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][1150/1251] eta 0:00:28 lr 0.000137 time 0.2889 (0.2862) loss 4.0149 (3.1999) grad_norm 2.4521 (inf) [2021-04-16 14:09:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][1160/1251] eta 0:00:26 lr 0.000137 time 0.2455 (0.2861) loss 2.5906 (3.1985) grad_norm 2.2723 (inf) [2021-04-16 14:09:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][1170/1251] eta 0:00:23 lr 0.000137 time 0.2849 (0.2860) loss 3.5439 (3.1997) grad_norm 2.4754 (inf) [2021-04-16 14:09:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][1180/1251] eta 0:00:20 lr 0.000137 time 0.2719 (0.2860) loss 2.8212 (3.1973) grad_norm 2.3503 (inf) [2021-04-16 14:09:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][1190/1251] eta 0:00:17 lr 0.000137 time 0.2592 (0.2859) loss 2.8420 (3.1968) grad_norm 2.1799 (inf) [2021-04-16 14:09:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][1200/1251] eta 0:00:14 lr 0.000137 time 0.2894 (0.2858) loss 3.7184 (3.1976) grad_norm 2.0916 (inf) [2021-04-16 14:09:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][1210/1251] eta 0:00:11 lr 0.000137 time 0.2866 (0.2857) loss 3.9616 (3.1989) grad_norm 1.9265 (inf) [2021-04-16 14:09:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][1220/1251] eta 0:00:08 lr 0.000137 time 0.2953 (0.2857) loss 2.1429 (3.1968) grad_norm 3.0008 (inf) [2021-04-16 14:09:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][1230/1251] eta 0:00:05 lr 0.000137 time 0.2574 (0.2857) loss 3.5056 (3.1974) grad_norm 2.6864 (inf) [2021-04-16 14:10:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][1240/1251] eta 0:00:03 lr 0.000137 time 0.2455 (0.2855) loss 2.2752 (3.1978) grad_norm 2.1344 (inf) [2021-04-16 14:10:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [229/300][1250/1251] eta 0:00:00 lr 0.000137 time 0.2491 (0.2853) loss 2.7701 (3.1962) grad_norm 2.5290 (inf) [2021-04-16 14:10:09 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 229 training takes 0:06:01 [2021-04-16 14:10:09 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_229.pth saving...... [2021-04-16 14:10:18 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_229.pth saved !!! [2021-04-16 14:10:19 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.104 (1.104) Loss 0.9013 (0.9013) Acc@1 79.297 (79.297) Acc@5 94.629 (94.629) [2021-04-16 14:10:20 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.126 (0.195) Loss 0.8520 (0.8732) Acc@1 78.125 (79.315) Acc@5 95.703 (94.656) [2021-04-16 14:10:23 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.104 (0.243) Loss 0.8548 (0.8814) Acc@1 81.348 (79.190) Acc@5 94.629 (94.689) [2021-04-16 14:10:25 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.173 (0.240) Loss 0.8760 (0.8778) Acc@1 79.688 (79.221) Acc@5 94.922 (94.749) [2021-04-16 14:10:27 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.214) Loss 0.9081 (0.8717) Acc@1 79.297 (79.495) Acc@5 94.629 (94.786) [2021-04-16 14:10:36 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 79.562 Acc@5 94.832 [2021-04-16 14:10:36 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 79.6% [2021-04-16 14:10:36 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 79.69% [2021-04-16 14:10:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][0/1251] eta 1:46:35 lr 0.000137 time 5.1120 (5.1120) loss 2.8364 (2.8364) grad_norm 2.3524 (2.3524) [2021-04-16 14:10:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][10/1251] eta 0:14:45 lr 0.000137 time 0.2849 (0.7139) loss 3.2854 (3.2855) grad_norm 2.4329 (2.4750) [2021-04-16 14:10:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][20/1251] eta 0:10:23 lr 0.000137 time 0.2912 (0.5067) loss 3.7038 (3.2740) grad_norm 2.2172 (2.4693) [2021-04-16 14:10:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][30/1251] eta 0:08:47 lr 0.000137 time 0.2829 (0.4321) loss 3.0668 (3.3034) grad_norm 2.3334 (2.4079) [2021-04-16 14:10:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][40/1251] eta 0:07:57 lr 0.000137 time 0.2726 (0.3941) loss 3.6127 (3.2989) grad_norm 2.4745 (2.4187) [2021-04-16 14:10:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][50/1251] eta 0:07:25 lr 0.000137 time 0.2616 (0.3713) loss 3.3205 (3.3414) grad_norm 2.1729 (2.3971) [2021-04-16 14:10:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][60/1251] eta 0:07:06 lr 0.000137 time 0.2558 (0.3580) loss 3.3845 (3.2775) grad_norm 2.1337 (2.4026) [2021-04-16 14:11:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][70/1251] eta 0:06:49 lr 0.000137 time 0.2780 (0.3470) loss 2.9567 (3.2921) grad_norm 2.2725 (2.3871) [2021-04-16 14:11:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][80/1251] eta 0:06:36 lr 0.000137 time 0.2752 (0.3383) loss 2.3614 (3.2636) grad_norm 2.0196 (2.3886) [2021-04-16 14:11:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][90/1251] eta 0:06:25 lr 0.000137 time 0.2960 (0.3323) loss 3.4027 (3.2536) grad_norm 2.2057 (2.3876) [2021-04-16 14:11:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][100/1251] eta 0:06:16 lr 0.000137 time 0.2879 (0.3272) loss 3.4490 (3.2173) grad_norm 2.8399 (2.3916) [2021-04-16 14:11:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][110/1251] eta 0:06:07 lr 0.000137 time 0.2619 (0.3225) loss 2.1461 (3.2152) grad_norm 2.1607 (2.3960) [2021-04-16 14:11:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][120/1251] eta 0:06:00 lr 0.000137 time 0.2743 (0.3190) loss 2.8729 (3.2041) grad_norm 2.2692 (2.4091) [2021-04-16 14:11:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][130/1251] eta 0:05:54 lr 0.000137 time 0.2453 (0.3165) loss 2.9901 (3.2056) grad_norm 2.5682 (2.4206) [2021-04-16 14:11:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][140/1251] eta 0:05:49 lr 0.000137 time 0.2448 (0.3147) loss 3.5447 (3.1895) grad_norm 2.1299 (2.4237) [2021-04-16 14:11:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][150/1251] eta 0:05:43 lr 0.000137 time 0.2632 (0.3120) loss 2.9369 (3.1950) grad_norm 2.1840 (2.4310) [2021-04-16 14:11:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][160/1251] eta 0:05:38 lr 0.000137 time 0.2783 (0.3099) loss 2.8573 (3.1847) grad_norm 2.2979 (2.4318) [2021-04-16 14:11:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][170/1251] eta 0:05:33 lr 0.000137 time 0.3049 (0.3081) loss 3.4211 (3.1788) grad_norm 2.4291 (2.4556) [2021-04-16 14:11:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][180/1251] eta 0:05:27 lr 0.000137 time 0.2675 (0.3062) loss 2.6950 (3.1819) grad_norm 2.0543 (2.4544) [2021-04-16 14:11:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][190/1251] eta 0:05:23 lr 0.000137 time 0.2904 (0.3045) loss 3.5336 (3.1831) grad_norm 2.3437 (2.4534) [2021-04-16 14:11:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][200/1251] eta 0:05:18 lr 0.000137 time 0.2685 (0.3032) loss 4.0492 (3.1916) grad_norm 4.9381 (2.4594) [2021-04-16 14:11:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][210/1251] eta 0:05:14 lr 0.000137 time 0.2871 (0.3020) loss 2.7820 (3.1916) grad_norm 2.3459 (2.4578) [2021-04-16 14:11:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][220/1251] eta 0:05:10 lr 0.000137 time 0.2679 (0.3008) loss 3.4101 (3.1889) grad_norm 2.1899 (2.4525) [2021-04-16 14:11:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][230/1251] eta 0:05:06 lr 0.000137 time 0.2615 (0.2997) loss 3.7568 (3.1840) grad_norm 2.5299 (2.4511) [2021-04-16 14:11:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][240/1251] eta 0:05:02 lr 0.000136 time 0.3006 (0.2996) loss 3.9920 (3.1761) grad_norm 2.4520 (2.4501) [2021-04-16 14:11:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][250/1251] eta 0:04:58 lr 0.000136 time 0.2899 (0.2987) loss 3.3327 (3.1826) grad_norm 2.3714 (2.4436) [2021-04-16 14:11:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][260/1251] eta 0:04:55 lr 0.000136 time 0.2517 (0.2978) loss 3.4607 (3.1810) grad_norm 2.2456 (2.4355) [2021-04-16 14:11:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][270/1251] eta 0:04:51 lr 0.000136 time 0.2989 (0.2970) loss 1.9519 (3.1715) grad_norm 2.4575 (2.4296) [2021-04-16 14:12:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][280/1251] eta 0:04:47 lr 0.000136 time 0.2605 (0.2962) loss 3.7975 (3.1789) grad_norm 2.1706 (2.4234) [2021-04-16 14:12:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][290/1251] eta 0:04:44 lr 0.000136 time 0.2921 (0.2963) loss 3.4407 (3.1843) grad_norm 2.2918 (2.4196) [2021-04-16 14:12:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][300/1251] eta 0:04:41 lr 0.000136 time 0.2871 (0.2957) loss 3.3681 (3.1871) grad_norm 3.3173 (2.4197) [2021-04-16 14:12:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][310/1251] eta 0:04:37 lr 0.000136 time 0.3014 (0.2951) loss 2.4741 (3.1827) grad_norm 2.4430 (2.4209) [2021-04-16 14:12:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][320/1251] eta 0:04:34 lr 0.000136 time 0.2796 (0.2948) loss 2.1680 (3.1823) grad_norm 2.3918 (2.4225) [2021-04-16 14:12:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][330/1251] eta 0:04:31 lr 0.000136 time 0.2612 (0.2943) loss 3.0625 (3.1757) grad_norm 2.1664 (2.4212) [2021-04-16 14:12:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][340/1251] eta 0:04:27 lr 0.000136 time 0.2751 (0.2940) loss 3.2678 (3.1807) grad_norm 2.0911 (2.4195) [2021-04-16 14:12:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][350/1251] eta 0:04:24 lr 0.000136 time 0.2837 (0.2935) loss 2.0873 (3.1800) grad_norm 2.2405 (2.4175) [2021-04-16 14:12:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][360/1251] eta 0:04:21 lr 0.000136 time 0.2882 (0.2938) loss 3.4553 (3.1782) grad_norm 2.4118 (2.4131) [2021-04-16 14:12:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][370/1251] eta 0:04:18 lr 0.000136 time 0.2679 (0.2937) loss 3.8160 (3.1805) grad_norm 2.4709 (2.4106) [2021-04-16 14:12:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][380/1251] eta 0:04:15 lr 0.000136 time 0.2567 (0.2931) loss 3.2071 (3.1835) grad_norm 2.6673 (2.4074) [2021-04-16 14:12:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][390/1251] eta 0:04:12 lr 0.000136 time 0.2763 (0.2927) loss 3.5165 (3.1872) grad_norm 2.6020 (2.4068) [2021-04-16 14:12:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][400/1251] eta 0:04:08 lr 0.000136 time 0.2742 (0.2923) loss 2.8734 (3.1864) grad_norm 2.2273 (2.4192) [2021-04-16 14:12:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][410/1251] eta 0:04:05 lr 0.000136 time 0.2913 (0.2919) loss 3.7094 (3.1795) grad_norm 2.6456 (2.4198) [2021-04-16 14:12:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][420/1251] eta 0:04:02 lr 0.000136 time 0.2882 (0.2915) loss 3.9366 (3.1850) grad_norm 2.3103 (2.4215) [2021-04-16 14:12:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][430/1251] eta 0:03:59 lr 0.000136 time 0.2942 (0.2914) loss 3.5101 (3.1945) grad_norm 2.2187 (2.4217) [2021-04-16 14:12:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][440/1251] eta 0:03:56 lr 0.000136 time 0.2891 (0.2911) loss 2.1736 (3.1918) grad_norm 2.3702 (2.4237) [2021-04-16 14:12:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][450/1251] eta 0:03:52 lr 0.000136 time 0.3024 (0.2908) loss 3.0610 (3.1945) grad_norm 1.9039 (2.4229) [2021-04-16 14:12:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][460/1251] eta 0:03:49 lr 0.000136 time 0.2602 (0.2906) loss 3.0764 (3.1929) grad_norm 2.4585 (2.4230) [2021-04-16 14:12:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][470/1251] eta 0:03:46 lr 0.000136 time 0.2791 (0.2902) loss 2.9260 (3.1925) grad_norm 2.4258 (2.4220) [2021-04-16 14:12:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][480/1251] eta 0:03:43 lr 0.000136 time 0.2751 (0.2902) loss 3.4373 (3.1937) grad_norm 2.2113 (2.4181) [2021-04-16 14:12:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][490/1251] eta 0:03:40 lr 0.000136 time 0.2946 (0.2903) loss 3.8251 (3.1936) grad_norm 2.2511 (2.4158) [2021-04-16 14:13:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][500/1251] eta 0:03:37 lr 0.000136 time 0.2674 (0.2902) loss 3.5734 (3.1932) grad_norm 2.6520 (2.4134) [2021-04-16 14:13:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][510/1251] eta 0:03:34 lr 0.000136 time 0.2658 (0.2900) loss 3.0038 (3.1918) grad_norm 2.5459 (2.4138) [2021-04-16 14:13:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][520/1251] eta 0:03:31 lr 0.000136 time 0.2823 (0.2896) loss 3.4637 (3.1953) grad_norm 2.0996 (2.4152) [2021-04-16 14:13:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][530/1251] eta 0:03:28 lr 0.000136 time 0.2807 (0.2895) loss 3.8057 (3.1935) grad_norm 2.6279 (2.4156) [2021-04-16 14:13:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][540/1251] eta 0:03:25 lr 0.000136 time 0.2954 (0.2895) loss 3.0090 (3.1901) grad_norm 2.3827 (2.4146) [2021-04-16 14:13:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][550/1251] eta 0:03:22 lr 0.000136 time 0.2926 (0.2893) loss 2.4367 (3.1887) grad_norm 2.2360 (2.4123) [2021-04-16 14:13:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][560/1251] eta 0:03:19 lr 0.000136 time 0.2744 (0.2890) loss 2.6563 (3.1909) grad_norm 2.2831 (2.4112) [2021-04-16 14:13:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][570/1251] eta 0:03:17 lr 0.000136 time 0.2780 (0.2894) loss 3.4215 (3.1877) grad_norm 2.1778 (2.4133) [2021-04-16 14:13:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][580/1251] eta 0:03:14 lr 0.000136 time 0.2567 (0.2892) loss 2.3574 (3.1868) grad_norm 2.3509 (2.4150) [2021-04-16 14:13:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][590/1251] eta 0:03:11 lr 0.000136 time 0.2876 (0.2890) loss 3.2281 (3.1912) grad_norm 2.4572 (2.4161) [2021-04-16 14:13:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][600/1251] eta 0:03:07 lr 0.000135 time 0.2731 (0.2888) loss 2.0928 (3.1863) grad_norm 2.5970 (2.4191) [2021-04-16 14:13:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][610/1251] eta 0:03:04 lr 0.000135 time 0.2716 (0.2886) loss 3.8719 (3.1846) grad_norm 2.1300 (2.4192) [2021-04-16 14:13:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][620/1251] eta 0:03:01 lr 0.000135 time 0.2643 (0.2884) loss 3.3726 (3.1866) grad_norm 2.4767 (2.4202) [2021-04-16 14:13:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][630/1251] eta 0:02:59 lr 0.000135 time 0.2683 (0.2883) loss 2.4190 (3.1893) grad_norm 2.4774 (2.4205) [2021-04-16 14:13:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][640/1251] eta 0:02:55 lr 0.000135 time 0.2767 (0.2880) loss 3.5626 (3.1895) grad_norm 2.3414 (2.4228) [2021-04-16 14:13:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][650/1251] eta 0:02:52 lr 0.000135 time 0.2740 (0.2878) loss 3.7442 (3.1888) grad_norm 3.1553 (2.4265) [2021-04-16 14:13:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][660/1251] eta 0:02:49 lr 0.000135 time 0.2590 (0.2875) loss 3.2837 (3.1879) grad_norm 3.0216 (2.4309) [2021-04-16 14:13:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][670/1251] eta 0:02:46 lr 0.000135 time 0.2673 (0.2874) loss 3.3589 (3.1874) grad_norm 2.5029 (2.4336) [2021-04-16 14:13:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][680/1251] eta 0:02:44 lr 0.000135 time 0.3120 (0.2873) loss 2.4396 (3.1859) grad_norm 2.3248 (2.4327) [2021-04-16 14:13:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][690/1251] eta 0:02:41 lr 0.000135 time 0.2774 (0.2872) loss 3.4208 (3.1899) grad_norm 2.5964 (2.4316) [2021-04-16 14:13:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][700/1251] eta 0:02:38 lr 0.000135 time 0.4311 (0.2872) loss 2.4471 (3.1843) grad_norm 3.7595 (2.4356) [2021-04-16 14:14:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][710/1251] eta 0:02:35 lr 0.000135 time 0.2720 (0.2870) loss 2.7689 (3.1858) grad_norm 2.2975 (2.4354) [2021-04-16 14:14:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][720/1251] eta 0:02:32 lr 0.000135 time 0.2936 (0.2871) loss 3.4455 (3.1868) grad_norm 2.1871 (2.4343) [2021-04-16 14:14:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][730/1251] eta 0:02:29 lr 0.000135 time 0.2663 (0.2871) loss 3.3104 (3.1873) grad_norm 2.1582 (2.4337) [2021-04-16 14:14:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][740/1251] eta 0:02:26 lr 0.000135 time 0.2593 (0.2869) loss 2.1992 (3.1865) grad_norm 2.5520 (2.4322) [2021-04-16 14:14:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][750/1251] eta 0:02:23 lr 0.000135 time 0.2719 (0.2868) loss 3.5053 (3.1905) grad_norm 2.4248 (2.4327) [2021-04-16 14:14:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][760/1251] eta 0:02:20 lr 0.000135 time 0.2812 (0.2867) loss 2.1704 (3.1870) grad_norm 2.1806 (2.4326) [2021-04-16 14:14:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][770/1251] eta 0:02:17 lr 0.000135 time 0.2522 (0.2865) loss 3.1927 (3.1882) grad_norm 2.2145 (2.4331) [2021-04-16 14:14:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][780/1251] eta 0:02:14 lr 0.000135 time 0.2708 (0.2864) loss 3.4251 (3.1863) grad_norm 2.6163 (2.4344) [2021-04-16 14:14:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][790/1251] eta 0:02:11 lr 0.000135 time 0.2636 (0.2863) loss 2.5162 (3.1828) grad_norm 2.6239 (2.4361) [2021-04-16 14:14:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][800/1251] eta 0:02:09 lr 0.000135 time 0.2547 (0.2861) loss 3.3033 (3.1833) grad_norm 2.3203 (2.4398) [2021-04-16 14:14:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][810/1251] eta 0:02:06 lr 0.000135 time 0.2867 (0.2861) loss 3.3412 (3.1831) grad_norm 2.6745 (2.4431) [2021-04-16 14:14:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][820/1251] eta 0:02:03 lr 0.000135 time 0.2793 (0.2862) loss 3.8636 (3.1824) grad_norm 2.3287 (2.4448) [2021-04-16 14:14:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][830/1251] eta 0:02:00 lr 0.000135 time 0.2947 (0.2861) loss 3.8500 (3.1806) grad_norm 2.5356 (2.4443) [2021-04-16 14:14:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][840/1251] eta 0:01:57 lr 0.000135 time 0.2597 (0.2860) loss 3.4660 (3.1836) grad_norm 2.6917 (2.4442) [2021-04-16 14:14:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][850/1251] eta 0:01:54 lr 0.000135 time 0.2593 (0.2859) loss 2.7764 (3.1824) grad_norm 2.7181 (2.4454) [2021-04-16 14:14:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][860/1251] eta 0:01:51 lr 0.000135 time 0.2736 (0.2858) loss 3.5449 (3.1813) grad_norm 2.4694 (2.4464) [2021-04-16 14:14:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][870/1251] eta 0:01:48 lr 0.000135 time 0.2932 (0.2859) loss 3.8129 (3.1808) grad_norm 3.0316 (2.4498) [2021-04-16 14:14:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][880/1251] eta 0:01:46 lr 0.000135 time 0.2616 (0.2858) loss 2.9294 (3.1816) grad_norm 2.1569 (2.4502) [2021-04-16 14:14:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][890/1251] eta 0:01:43 lr 0.000135 time 0.2801 (0.2857) loss 3.4211 (3.1822) grad_norm 2.3822 (2.4486) [2021-04-16 14:14:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][900/1251] eta 0:01:40 lr 0.000135 time 0.2667 (0.2858) loss 3.3253 (3.1845) grad_norm 2.3253 (2.4480) [2021-04-16 14:14:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][910/1251] eta 0:01:37 lr 0.000135 time 0.2807 (0.2858) loss 2.2755 (3.1836) grad_norm 2.5055 (2.4465) [2021-04-16 14:14:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][920/1251] eta 0:01:34 lr 0.000135 time 0.2768 (0.2856) loss 3.4005 (3.1847) grad_norm 2.2049 (2.4443) [2021-04-16 14:15:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][930/1251] eta 0:01:31 lr 0.000135 time 0.4244 (0.2857) loss 2.8741 (3.1844) grad_norm 2.0801 (2.4421) [2021-04-16 14:15:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][940/1251] eta 0:01:28 lr 0.000135 time 0.2649 (0.2857) loss 3.1876 (3.1854) grad_norm 2.2784 (2.4414) [2021-04-16 14:15:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][950/1251] eta 0:01:26 lr 0.000135 time 0.2774 (0.2857) loss 3.9012 (3.1832) grad_norm 2.0961 (2.4407) [2021-04-16 14:15:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][960/1251] eta 0:01:23 lr 0.000134 time 0.2784 (0.2857) loss 3.2827 (3.1847) grad_norm 2.3207 (2.4401) [2021-04-16 14:15:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][970/1251] eta 0:01:20 lr 0.000134 time 0.2891 (0.2857) loss 3.5546 (3.1832) grad_norm 2.5824 (2.4404) [2021-04-16 14:15:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][980/1251] eta 0:01:17 lr 0.000134 time 0.2690 (0.2856) loss 3.1660 (3.1820) grad_norm 2.1630 (2.4401) [2021-04-16 14:15:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][990/1251] eta 0:01:14 lr 0.000134 time 0.3060 (0.2855) loss 2.1912 (3.1807) grad_norm 2.5204 (2.4401) [2021-04-16 14:15:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][1000/1251] eta 0:01:11 lr 0.000134 time 0.2478 (0.2855) loss 3.4563 (3.1820) grad_norm 2.0032 (2.4386) [2021-04-16 14:15:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][1010/1251] eta 0:01:08 lr 0.000134 time 0.2900 (0.2855) loss 3.5725 (3.1802) grad_norm 2.6984 (2.4391) [2021-04-16 14:15:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][1020/1251] eta 0:01:05 lr 0.000134 time 0.2823 (0.2855) loss 2.8839 (3.1812) grad_norm 2.2495 (2.4388) [2021-04-16 14:15:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][1030/1251] eta 0:01:03 lr 0.000134 time 0.2968 (0.2855) loss 3.6236 (3.1830) grad_norm 1.9567 (2.4376) [2021-04-16 14:15:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][1040/1251] eta 0:01:00 lr 0.000134 time 0.2830 (0.2854) loss 3.4827 (3.1848) grad_norm 2.1875 (2.4372) [2021-04-16 14:15:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][1050/1251] eta 0:00:57 lr 0.000134 time 0.4268 (0.2855) loss 3.3607 (3.1870) grad_norm 2.4760 (2.4378) [2021-04-16 14:15:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][1060/1251] eta 0:00:54 lr 0.000134 time 0.2945 (0.2854) loss 3.2188 (3.1856) grad_norm 2.5979 (2.4391) [2021-04-16 14:15:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][1070/1251] eta 0:00:51 lr 0.000134 time 0.2880 (0.2855) loss 3.4207 (3.1849) grad_norm 2.4569 (2.4409) [2021-04-16 14:15:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][1080/1251] eta 0:00:48 lr 0.000134 time 0.2693 (0.2854) loss 3.7103 (3.1842) grad_norm 2.4461 (2.4410) [2021-04-16 14:15:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][1090/1251] eta 0:00:45 lr 0.000134 time 0.2710 (0.2853) loss 2.7064 (3.1818) grad_norm 2.7260 (2.4420) [2021-04-16 14:15:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][1100/1251] eta 0:00:43 lr 0.000134 time 0.2557 (0.2852) loss 2.3471 (3.1833) grad_norm 2.4557 (2.4414) [2021-04-16 14:15:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][1110/1251] eta 0:00:40 lr 0.000134 time 0.2668 (0.2852) loss 2.3643 (3.1840) grad_norm 2.7637 (2.4423) [2021-04-16 14:15:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][1120/1251] eta 0:00:37 lr 0.000134 time 0.2916 (0.2852) loss 2.3093 (3.1836) grad_norm 2.7442 (2.4422) [2021-04-16 14:15:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][1130/1251] eta 0:00:34 lr 0.000134 time 0.2746 (0.2852) loss 3.3700 (3.1839) grad_norm 2.6537 (2.4418) [2021-04-16 14:16:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][1140/1251] eta 0:00:31 lr 0.000134 time 0.2810 (0.2853) loss 2.0862 (3.1818) grad_norm 2.3113 (2.4424) [2021-04-16 14:16:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][1150/1251] eta 0:00:28 lr 0.000134 time 0.3233 (0.2853) loss 3.4389 (3.1824) grad_norm 2.2000 (2.4431) [2021-04-16 14:16:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][1160/1251] eta 0:00:25 lr 0.000134 time 0.2434 (0.2852) loss 3.3264 (3.1815) grad_norm 2.4146 (2.4428) [2021-04-16 14:16:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][1170/1251] eta 0:00:23 lr 0.000134 time 0.2753 (0.2851) loss 2.5559 (3.1774) grad_norm 2.1497 (2.4420) [2021-04-16 14:16:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][1180/1251] eta 0:00:20 lr 0.000134 time 0.2936 (0.2850) loss 3.2220 (3.1796) grad_norm 2.0680 (2.4413) [2021-04-16 14:16:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][1190/1251] eta 0:00:17 lr 0.000134 time 0.2765 (0.2849) loss 3.1765 (3.1776) grad_norm 2.8745 (2.4411) [2021-04-16 14:16:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][1200/1251] eta 0:00:14 lr 0.000134 time 0.2663 (0.2849) loss 3.2503 (3.1777) grad_norm 2.1354 (2.4415) [2021-04-16 14:16:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][1210/1251] eta 0:00:11 lr 0.000134 time 0.2485 (0.2849) loss 3.5891 (3.1774) grad_norm 2.5037 (2.4408) [2021-04-16 14:16:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][1220/1251] eta 0:00:08 lr 0.000134 time 0.3006 (0.2848) loss 3.5330 (3.1796) grad_norm 2.6624 (2.4411) [2021-04-16 14:16:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][1230/1251] eta 0:00:05 lr 0.000134 time 0.2852 (0.2848) loss 3.6052 (3.1787) grad_norm 2.3266 (2.4404) [2021-04-16 14:16:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][1240/1251] eta 0:00:03 lr 0.000134 time 0.2483 (0.2846) loss 3.0765 (3.1781) grad_norm 2.8618 (2.4430) [2021-04-16 14:16:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [230/300][1250/1251] eta 0:00:00 lr 0.000134 time 0.2482 (0.2843) loss 2.7820 (3.1777) grad_norm 1.9630 (2.4429) [2021-04-16 14:16:48 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 230 training takes 0:06:11 [2021-04-16 14:16:48 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_230.pth saving...... [2021-04-16 14:17:14 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_230.pth saved !!! [2021-04-16 14:17:15 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.136 (1.136) Loss 0.8442 (0.8442) Acc@1 81.250 (81.250) Acc@5 94.922 (94.922) [2021-04-16 14:17:16 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.129 (0.211) Loss 0.8347 (0.8645) Acc@1 79.492 (79.963) Acc@5 96.191 (95.224) [2021-04-16 14:17:18 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.116 (0.228) Loss 0.8579 (0.8655) Acc@1 79.590 (79.822) Acc@5 95.117 (95.080) [2021-04-16 14:17:21 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.092 (0.228) Loss 0.8801 (0.8682) Acc@1 80.762 (79.877) Acc@5 95.215 (94.963) [2021-04-16 14:17:23 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.217) Loss 0.8623 (0.8710) Acc@1 79.004 (79.738) Acc@5 94.336 (94.941) [2021-04-16 14:17:34 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 79.656 Acc@5 94.966 [2021-04-16 14:17:34 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 79.7% [2021-04-16 14:17:34 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 79.69% [2021-04-16 14:17:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][0/1251] eta 2:25:33 lr 0.000134 time 6.9809 (6.9809) loss 3.0420 (3.0420) grad_norm 2.2096 (2.2096) [2021-04-16 14:17:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][10/1251] eta 0:18:08 lr 0.000134 time 0.2588 (0.8774) loss 3.3078 (3.1147) grad_norm 2.5205 (2.5605) [2021-04-16 14:17:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][20/1251] eta 0:12:07 lr 0.000134 time 0.2774 (0.5907) loss 3.4098 (3.2028) grad_norm 2.3679 (2.5270) [2021-04-16 14:17:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][30/1251] eta 0:10:03 lr 0.000134 time 0.2717 (0.4939) loss 3.7606 (3.1647) grad_norm 2.4678 (2.5104) [2021-04-16 14:17:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][40/1251] eta 0:08:54 lr 0.000134 time 0.2839 (0.4416) loss 3.5400 (3.1595) grad_norm 2.0595 (2.4469) [2021-04-16 14:17:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][50/1251] eta 0:08:11 lr 0.000134 time 0.2883 (0.4092) loss 2.9975 (3.1432) grad_norm 2.4008 (2.5107) [2021-04-16 14:17:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][60/1251] eta 0:07:43 lr 0.000134 time 0.2857 (0.3893) loss 3.2485 (3.1486) grad_norm 2.5506 (2.4851) [2021-04-16 14:18:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][70/1251] eta 0:07:21 lr 0.000134 time 0.2676 (0.3737) loss 2.8411 (3.1525) grad_norm 2.4074 (2.4704) [2021-04-16 14:18:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][80/1251] eta 0:07:05 lr 0.000133 time 0.2956 (0.3632) loss 3.4214 (3.1382) grad_norm 2.3099 (2.4776) [2021-04-16 14:18:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][90/1251] eta 0:06:50 lr 0.000133 time 0.2790 (0.3534) loss 3.2859 (3.1529) grad_norm 2.4644 (2.4867) [2021-04-16 14:18:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][100/1251] eta 0:06:38 lr 0.000133 time 0.3152 (0.3464) loss 3.9078 (3.1713) grad_norm 2.5084 (2.4722) [2021-04-16 14:18:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][110/1251] eta 0:06:28 lr 0.000133 time 0.2832 (0.3408) loss 3.2067 (3.1724) grad_norm 2.3734 (2.4727) [2021-04-16 14:18:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][120/1251] eta 0:06:19 lr 0.000133 time 0.2722 (0.3358) loss 2.3856 (3.1547) grad_norm 2.4463 (2.4689) [2021-04-16 14:18:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][130/1251] eta 0:06:12 lr 0.000133 time 0.2708 (0.3321) loss 3.4012 (3.1489) grad_norm 2.4390 (2.4638) [2021-04-16 14:18:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][140/1251] eta 0:06:05 lr 0.000133 time 0.4455 (0.3293) loss 1.9159 (3.1195) grad_norm 2.3274 (2.4606) [2021-04-16 14:18:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][150/1251] eta 0:05:59 lr 0.000133 time 0.2934 (0.3269) loss 2.7547 (3.1123) grad_norm 2.0322 (2.4586) [2021-04-16 14:18:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][160/1251] eta 0:05:53 lr 0.000133 time 0.2665 (0.3237) loss 3.8061 (3.1321) grad_norm 2.4381 (2.4641) [2021-04-16 14:18:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][170/1251] eta 0:05:46 lr 0.000133 time 0.2522 (0.3209) loss 3.8138 (3.1360) grad_norm 2.2588 (2.4626) [2021-04-16 14:18:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][180/1251] eta 0:05:41 lr 0.000133 time 0.2692 (0.3186) loss 2.9748 (3.1374) grad_norm 2.4543 (2.4609) [2021-04-16 14:18:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][190/1251] eta 0:05:35 lr 0.000133 time 0.2524 (0.3165) loss 3.4450 (3.1435) grad_norm 2.6797 (2.4667) [2021-04-16 14:18:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][200/1251] eta 0:05:30 lr 0.000133 time 0.2564 (0.3148) loss 3.2911 (3.1443) grad_norm 3.2849 (2.4724) [2021-04-16 14:18:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][210/1251] eta 0:05:26 lr 0.000133 time 0.2852 (0.3132) loss 2.7889 (3.1378) grad_norm 2.2155 (2.4724) [2021-04-16 14:18:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][220/1251] eta 0:05:22 lr 0.000133 time 0.2860 (0.3128) loss 2.6206 (3.1331) grad_norm 2.3233 (2.4754) [2021-04-16 14:18:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][230/1251] eta 0:05:18 lr 0.000133 time 0.2814 (0.3115) loss 3.8746 (3.1352) grad_norm 2.8302 (2.4761) [2021-04-16 14:18:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][240/1251] eta 0:05:13 lr 0.000133 time 0.2754 (0.3102) loss 3.3454 (3.1306) grad_norm 2.6999 (2.4747) [2021-04-16 14:18:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][250/1251] eta 0:05:08 lr 0.000133 time 0.2450 (0.3087) loss 3.4904 (3.1252) grad_norm 2.6364 (2.4720) [2021-04-16 14:18:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][260/1251] eta 0:05:05 lr 0.000133 time 0.4030 (0.3082) loss 3.9146 (3.1187) grad_norm 2.6380 (2.4791) [2021-04-16 14:18:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][270/1251] eta 0:05:01 lr 0.000133 time 0.2931 (0.3070) loss 3.6254 (3.1271) grad_norm 2.6422 (2.4764) [2021-04-16 14:19:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][280/1251] eta 0:04:56 lr 0.000133 time 0.2774 (0.3057) loss 2.2263 (3.1345) grad_norm 3.0312 (2.4797) [2021-04-16 14:19:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][290/1251] eta 0:04:53 lr 0.000133 time 0.2424 (0.3053) loss 3.4134 (3.1377) grad_norm 2.7794 (2.4768) [2021-04-16 14:19:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][300/1251] eta 0:04:49 lr 0.000133 time 0.2699 (0.3046) loss 3.4346 (3.1411) grad_norm 2.5021 (2.4754) [2021-04-16 14:19:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][310/1251] eta 0:04:45 lr 0.000133 time 0.2553 (0.3037) loss 2.1525 (3.1302) grad_norm 2.5774 (2.4716) [2021-04-16 14:19:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][320/1251] eta 0:04:42 lr 0.000133 time 0.2964 (0.3036) loss 2.6789 (3.1300) grad_norm 2.8451 (2.4724) [2021-04-16 14:19:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][330/1251] eta 0:04:38 lr 0.000133 time 0.2585 (0.3028) loss 3.7587 (3.1351) grad_norm 2.3049 (2.4694) [2021-04-16 14:19:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][340/1251] eta 0:04:35 lr 0.000133 time 0.2752 (0.3024) loss 2.4903 (3.1342) grad_norm 2.3724 (2.4687) [2021-04-16 14:19:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][350/1251] eta 0:04:32 lr 0.000133 time 0.2842 (0.3019) loss 2.6872 (3.1373) grad_norm 2.9677 (2.4751) [2021-04-16 14:19:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][360/1251] eta 0:04:28 lr 0.000133 time 0.2628 (0.3015) loss 3.7355 (3.1414) grad_norm 2.3046 (2.4727) [2021-04-16 14:19:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][370/1251] eta 0:04:25 lr 0.000133 time 0.2476 (0.3012) loss 2.4654 (3.1417) grad_norm 3.3851 (2.4809) [2021-04-16 14:19:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][380/1251] eta 0:04:22 lr 0.000133 time 0.2729 (0.3008) loss 3.4705 (3.1504) grad_norm 2.4991 (2.4814) [2021-04-16 14:19:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][390/1251] eta 0:04:18 lr 0.000133 time 0.3033 (0.3004) loss 3.3454 (3.1531) grad_norm 2.4262 (2.4834) [2021-04-16 14:19:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][400/1251] eta 0:04:15 lr 0.000133 time 0.2766 (0.2999) loss 3.3741 (3.1559) grad_norm 2.3838 (2.4864) [2021-04-16 14:19:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][410/1251] eta 0:04:11 lr 0.000133 time 0.2577 (0.2992) loss 2.4075 (3.1527) grad_norm 1.9413 (2.4836) [2021-04-16 14:19:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][420/1251] eta 0:04:08 lr 0.000133 time 0.2790 (0.2986) loss 3.5651 (3.1530) grad_norm 2.1250 (2.4801) [2021-04-16 14:19:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][430/1251] eta 0:04:04 lr 0.000133 time 0.2588 (0.2981) loss 3.3170 (3.1504) grad_norm 2.2301 (2.4783) [2021-04-16 14:19:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][440/1251] eta 0:04:01 lr 0.000132 time 0.2581 (0.2976) loss 2.3857 (3.1494) grad_norm 2.2454 (2.4782) [2021-04-16 14:19:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][450/1251] eta 0:03:58 lr 0.000132 time 0.2905 (0.2971) loss 3.3069 (3.1479) grad_norm 2.4608 (2.4763) [2021-04-16 14:19:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][460/1251] eta 0:03:54 lr 0.000132 time 0.3079 (0.2968) loss 3.3433 (3.1524) grad_norm 2.3486 (2.4760) [2021-04-16 14:19:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][470/1251] eta 0:03:51 lr 0.000132 time 0.2615 (0.2962) loss 2.8237 (3.1518) grad_norm 2.3636 (2.4740) [2021-04-16 14:19:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][480/1251] eta 0:03:48 lr 0.000132 time 0.2862 (0.2958) loss 3.3145 (3.1532) grad_norm 2.2608 (2.4713) [2021-04-16 14:19:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][490/1251] eta 0:03:44 lr 0.000132 time 0.2791 (0.2953) loss 3.0849 (3.1569) grad_norm 2.1874 (2.4708) [2021-04-16 14:20:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][500/1251] eta 0:03:41 lr 0.000132 time 0.2723 (0.2949) loss 3.2745 (3.1611) grad_norm 2.3667 (2.4676) [2021-04-16 14:20:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][510/1251] eta 0:03:38 lr 0.000132 time 0.2714 (0.2945) loss 3.2130 (3.1617) grad_norm 2.4246 (2.4662) [2021-04-16 14:20:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][520/1251] eta 0:03:35 lr 0.000132 time 0.2735 (0.2942) loss 3.2754 (3.1608) grad_norm 2.6421 (2.4640) [2021-04-16 14:20:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][530/1251] eta 0:03:31 lr 0.000132 time 0.2995 (0.2938) loss 2.0519 (3.1600) grad_norm 2.1823 (2.4616) [2021-04-16 14:20:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][540/1251] eta 0:03:28 lr 0.000132 time 0.2800 (0.2937) loss 2.0387 (3.1543) grad_norm 2.2414 (2.4602) [2021-04-16 14:20:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][550/1251] eta 0:03:25 lr 0.000132 time 0.2787 (0.2933) loss 3.2559 (3.1553) grad_norm 2.3575 (2.4569) [2021-04-16 14:20:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][560/1251] eta 0:03:22 lr 0.000132 time 0.2791 (0.2933) loss 3.5421 (3.1570) grad_norm 2.1870 (2.4575) [2021-04-16 14:20:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][570/1251] eta 0:03:19 lr 0.000132 time 0.4232 (0.2932) loss 3.0442 (3.1556) grad_norm 2.7534 (2.4608) [2021-04-16 14:20:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][580/1251] eta 0:03:16 lr 0.000132 time 0.2928 (0.2929) loss 2.8917 (3.1488) grad_norm 2.7339 (2.4595) [2021-04-16 14:20:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][590/1251] eta 0:03:13 lr 0.000132 time 0.2752 (0.2926) loss 3.6357 (3.1451) grad_norm 2.1528 (2.4594) [2021-04-16 14:20:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][600/1251] eta 0:03:10 lr 0.000132 time 0.2689 (0.2923) loss 3.0620 (3.1411) grad_norm 2.1446 (2.4563) [2021-04-16 14:20:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][610/1251] eta 0:03:07 lr 0.000132 time 0.3021 (0.2922) loss 4.1488 (3.1404) grad_norm 2.0177 (2.4585) [2021-04-16 14:20:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][620/1251] eta 0:03:04 lr 0.000132 time 0.2977 (0.2919) loss 2.7380 (3.1412) grad_norm 2.3475 (2.4612) [2021-04-16 14:20:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][630/1251] eta 0:03:01 lr 0.000132 time 0.2885 (0.2916) loss 3.7106 (3.1398) grad_norm 2.5971 (2.4632) [2021-04-16 14:20:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][640/1251] eta 0:02:58 lr 0.000132 time 0.2836 (0.2914) loss 2.9525 (3.1419) grad_norm 2.9884 (2.4647) [2021-04-16 14:20:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][650/1251] eta 0:02:55 lr 0.000132 time 0.2597 (0.2915) loss 3.0342 (3.1406) grad_norm 2.8129 (2.4640) [2021-04-16 14:20:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][660/1251] eta 0:02:52 lr 0.000132 time 0.2591 (0.2912) loss 3.3880 (3.1443) grad_norm 2.4495 (2.4638) [2021-04-16 14:20:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][670/1251] eta 0:02:49 lr 0.000132 time 0.2697 (0.2910) loss 2.2254 (3.1440) grad_norm 2.3586 (2.4646) [2021-04-16 14:20:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][680/1251] eta 0:02:46 lr 0.000132 time 0.3048 (0.2908) loss 2.2728 (3.1415) grad_norm 2.3200 (2.4617) [2021-04-16 14:20:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][690/1251] eta 0:02:43 lr 0.000132 time 0.2849 (0.2907) loss 2.9495 (3.1457) grad_norm 2.2599 (2.4598) [2021-04-16 14:20:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][700/1251] eta 0:02:40 lr 0.000132 time 0.2900 (0.2906) loss 3.5273 (3.1429) grad_norm 2.2374 (2.4582) [2021-04-16 14:21:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][710/1251] eta 0:02:37 lr 0.000132 time 0.3081 (0.2905) loss 2.8555 (3.1419) grad_norm 2.8593 (2.4587) [2021-04-16 14:21:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][720/1251] eta 0:02:34 lr 0.000132 time 0.2838 (0.2906) loss 4.2192 (3.1411) grad_norm 8.0643 (2.4663) [2021-04-16 14:21:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][730/1251] eta 0:02:31 lr 0.000132 time 0.2529 (0.2905) loss 3.5532 (3.1397) grad_norm 2.5416 (2.4684) [2021-04-16 14:21:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][740/1251] eta 0:02:28 lr 0.000132 time 0.2882 (0.2904) loss 2.3019 (3.1375) grad_norm 2.5825 (2.4690) [2021-04-16 14:21:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][750/1251] eta 0:02:25 lr 0.000132 time 0.2715 (0.2902) loss 3.3336 (3.1402) grad_norm 2.4486 (2.4718) [2021-04-16 14:21:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][760/1251] eta 0:02:22 lr 0.000132 time 0.2777 (0.2900) loss 3.3963 (3.1403) grad_norm 2.3341 (2.4711) [2021-04-16 14:21:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][770/1251] eta 0:02:19 lr 0.000132 time 0.2611 (0.2899) loss 3.5423 (3.1423) grad_norm 2.4832 (2.4706) [2021-04-16 14:21:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][780/1251] eta 0:02:16 lr 0.000132 time 0.2753 (0.2897) loss 3.2760 (3.1416) grad_norm 2.2852 (2.4714) [2021-04-16 14:21:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][790/1251] eta 0:02:13 lr 0.000132 time 0.2998 (0.2898) loss 2.0957 (3.1375) grad_norm 2.5390 (2.4712) [2021-04-16 14:21:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][800/1251] eta 0:02:10 lr 0.000132 time 0.2726 (0.2897) loss 3.0081 (3.1377) grad_norm 2.3196 (2.4690) [2021-04-16 14:21:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][810/1251] eta 0:02:07 lr 0.000131 time 0.2635 (0.2895) loss 3.7245 (3.1383) grad_norm 2.2738 (2.4678) [2021-04-16 14:21:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][820/1251] eta 0:02:04 lr 0.000131 time 0.2689 (0.2893) loss 2.2352 (3.1368) grad_norm 2.4727 (2.4671) [2021-04-16 14:21:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][830/1251] eta 0:02:01 lr 0.000131 time 0.2882 (0.2892) loss 3.1284 (3.1366) grad_norm 2.4238 (2.4667) [2021-04-16 14:21:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][840/1251] eta 0:01:58 lr 0.000131 time 0.2807 (0.2892) loss 2.1315 (3.1347) grad_norm 2.5622 (2.4665) [2021-04-16 14:21:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][850/1251] eta 0:01:55 lr 0.000131 time 0.2807 (0.2890) loss 2.2671 (3.1335) grad_norm 2.5472 (2.4665) [2021-04-16 14:21:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][860/1251] eta 0:01:52 lr 0.000131 time 0.2693 (0.2888) loss 3.1500 (3.1352) grad_norm 2.2955 (2.4654) [2021-04-16 14:21:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][870/1251] eta 0:01:50 lr 0.000131 time 0.2582 (0.2888) loss 2.5205 (3.1372) grad_norm 2.2574 (2.4631) [2021-04-16 14:21:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][880/1251] eta 0:01:47 lr 0.000131 time 0.2618 (0.2887) loss 3.8346 (3.1414) grad_norm 2.4079 (2.4631) [2021-04-16 14:21:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][890/1251] eta 0:01:44 lr 0.000131 time 0.2511 (0.2887) loss 3.6963 (3.1431) grad_norm 2.3604 (2.4624) [2021-04-16 14:21:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][900/1251] eta 0:01:41 lr 0.000131 time 0.3029 (0.2886) loss 3.4597 (3.1430) grad_norm 2.6027 (2.4606) [2021-04-16 14:21:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][910/1251] eta 0:01:38 lr 0.000131 time 0.2658 (0.2885) loss 3.1179 (3.1431) grad_norm 2.2745 (2.4589) [2021-04-16 14:22:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][920/1251] eta 0:01:35 lr 0.000131 time 0.2769 (0.2886) loss 2.9035 (3.1415) grad_norm 2.0239 (2.4586) [2021-04-16 14:22:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][930/1251] eta 0:01:32 lr 0.000131 time 0.2766 (0.2887) loss 3.6032 (3.1420) grad_norm 3.1160 (2.4584) [2021-04-16 14:22:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][940/1251] eta 0:01:29 lr 0.000131 time 0.3922 (0.2889) loss 3.6460 (3.1438) grad_norm 2.2996 (2.4588) [2021-04-16 14:22:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][950/1251] eta 0:01:26 lr 0.000131 time 0.2744 (0.2887) loss 3.4767 (3.1483) grad_norm 2.6094 (2.4601) [2021-04-16 14:22:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][960/1251] eta 0:01:23 lr 0.000131 time 0.2642 (0.2885) loss 3.1090 (3.1473) grad_norm 2.2099 (2.4609) [2021-04-16 14:22:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][970/1251] eta 0:01:21 lr 0.000131 time 0.2926 (0.2884) loss 3.7738 (3.1484) grad_norm 2.5585 (2.4625) [2021-04-16 14:22:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][980/1251] eta 0:01:18 lr 0.000131 time 0.2832 (0.2883) loss 3.5444 (3.1472) grad_norm 2.7955 (2.4621) [2021-04-16 14:22:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][990/1251] eta 0:01:15 lr 0.000131 time 0.2720 (0.2882) loss 2.0093 (3.1483) grad_norm 2.0926 (2.4607) [2021-04-16 14:22:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][1000/1251] eta 0:01:12 lr 0.000131 time 0.2664 (0.2881) loss 2.5696 (3.1469) grad_norm 2.2773 (2.4606) [2021-04-16 14:22:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][1010/1251] eta 0:01:09 lr 0.000131 time 0.2797 (0.2881) loss 3.1824 (3.1465) grad_norm 2.4654 (2.4606) [2021-04-16 14:22:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][1020/1251] eta 0:01:06 lr 0.000131 time 0.2587 (0.2879) loss 2.8025 (3.1461) grad_norm 2.2022 (2.4598) [2021-04-16 14:22:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][1030/1251] eta 0:01:03 lr 0.000131 time 0.2720 (0.2878) loss 3.2345 (3.1462) grad_norm 2.4101 (2.4594) [2021-04-16 14:22:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][1040/1251] eta 0:01:00 lr 0.000131 time 0.2913 (0.2877) loss 3.5378 (3.1477) grad_norm 2.4850 (2.4600) [2021-04-16 14:22:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][1050/1251] eta 0:00:57 lr 0.000131 time 0.2844 (0.2876) loss 3.7461 (3.1488) grad_norm 2.3030 (2.4589) [2021-04-16 14:22:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][1060/1251] eta 0:00:54 lr 0.000131 time 0.2753 (0.2875) loss 3.2559 (3.1516) grad_norm 2.1863 (2.4590) [2021-04-16 14:22:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][1070/1251] eta 0:00:52 lr 0.000131 time 0.2737 (0.2874) loss 3.5643 (3.1543) grad_norm 2.3769 (2.4576) [2021-04-16 14:22:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][1080/1251] eta 0:00:49 lr 0.000131 time 0.2788 (0.2873) loss 4.0020 (3.1556) grad_norm 2.2508 (2.4559) [2021-04-16 14:22:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][1090/1251] eta 0:00:46 lr 0.000131 time 0.2624 (0.2872) loss 3.8574 (3.1604) grad_norm 2.4088 (2.4558) [2021-04-16 14:22:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][1100/1251] eta 0:00:43 lr 0.000131 time 0.2630 (0.2871) loss 3.9427 (3.1635) grad_norm 2.2352 (2.4551) [2021-04-16 14:22:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][1110/1251] eta 0:00:40 lr 0.000131 time 0.2571 (0.2870) loss 3.9391 (3.1642) grad_norm 2.3657 (2.4549) [2021-04-16 14:22:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][1120/1251] eta 0:00:37 lr 0.000131 time 0.3028 (0.2871) loss 1.9502 (3.1655) grad_norm 2.1003 (2.4546) [2021-04-16 14:22:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][1130/1251] eta 0:00:34 lr 0.000131 time 0.2569 (0.2869) loss 3.3647 (3.1666) grad_norm 2.5894 (2.4546) [2021-04-16 14:23:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][1140/1251] eta 0:00:31 lr 0.000131 time 0.2911 (0.2868) loss 2.5212 (3.1664) grad_norm 2.4595 (2.4549) [2021-04-16 14:23:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][1150/1251] eta 0:00:28 lr 0.000131 time 0.2694 (0.2869) loss 2.5641 (3.1642) grad_norm 2.6485 (2.4559) [2021-04-16 14:23:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][1160/1251] eta 0:00:26 lr 0.000131 time 0.2977 (0.2869) loss 3.6417 (3.1638) grad_norm 2.0546 (2.4556) [2021-04-16 14:23:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][1170/1251] eta 0:00:23 lr 0.000131 time 0.2430 (0.2867) loss 2.4528 (3.1614) grad_norm 2.4477 (2.4558) [2021-04-16 14:23:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][1180/1251] eta 0:00:20 lr 0.000130 time 0.3113 (0.2868) loss 3.9421 (3.1639) grad_norm 2.3619 (2.4555) [2021-04-16 14:23:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][1190/1251] eta 0:00:17 lr 0.000130 time 0.2619 (0.2866) loss 3.4679 (3.1642) grad_norm 2.4125 (2.4544) [2021-04-16 14:23:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][1200/1251] eta 0:00:14 lr 0.000130 time 0.2731 (0.2865) loss 3.7362 (3.1655) grad_norm 2.3956 (2.4542) [2021-04-16 14:23:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][1210/1251] eta 0:00:11 lr 0.000130 time 0.2750 (0.2864) loss 2.3848 (3.1653) grad_norm 2.8876 (2.4536) [2021-04-16 14:23:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][1220/1251] eta 0:00:08 lr 0.000130 time 0.2822 (0.2864) loss 3.5283 (3.1670) grad_norm 2.3957 (2.4551) [2021-04-16 14:23:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][1230/1251] eta 0:00:06 lr 0.000130 time 0.2612 (0.2863) loss 2.1284 (3.1657) grad_norm 3.1932 (2.4571) [2021-04-16 14:23:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][1240/1251] eta 0:00:03 lr 0.000130 time 0.2477 (0.2862) loss 3.7088 (3.1667) grad_norm 2.5022 (2.4559) [2021-04-16 14:23:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [231/300][1250/1251] eta 0:00:00 lr 0.000130 time 0.2506 (0.2859) loss 3.5575 (3.1676) grad_norm 2.2285 (2.4556) [2021-04-16 14:23:45 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 231 training takes 0:06:11 [2021-04-16 14:23:45 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_231.pth saving...... [2021-04-16 14:24:06 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_231.pth saved !!! [2021-04-16 14:24:07 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.104 (1.104) Loss 0.8243 (0.8243) Acc@1 81.934 (81.934) Acc@5 94.824 (94.824) [2021-04-16 14:24:08 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.358 (0.233) Loss 0.8757 (0.8696) Acc@1 80.469 (79.812) Acc@5 94.141 (94.753) [2021-04-16 14:24:10 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.096 (0.206) Loss 0.9551 (0.8841) Acc@1 77.441 (79.455) Acc@5 93.652 (94.596) [2021-04-16 14:24:13 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.158 (0.231) Loss 0.8796 (0.8733) Acc@1 78.027 (79.483) Acc@5 94.922 (94.783) [2021-04-16 14:24:14 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.212) Loss 0.8125 (0.8708) Acc@1 80.273 (79.452) Acc@5 95.898 (94.846) [2021-04-16 14:24:28 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 79.486 Acc@5 94.928 [2021-04-16 14:24:28 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 79.5% [2021-04-16 14:24:28 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 79.69% [2021-04-16 14:24:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][0/1251] eta 3:54:43 lr 0.000130 time 11.2580 (11.2580) loss 3.1716 (3.1716) grad_norm 2.5154 (2.5154) [2021-04-16 14:24:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][10/1251] eta 0:26:18 lr 0.000130 time 0.2803 (1.2721) loss 2.7424 (3.0653) grad_norm 2.4465 (2.3987) [2021-04-16 14:24:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][20/1251] eta 0:16:25 lr 0.000130 time 0.2835 (0.8005) loss 3.8646 (3.3230) grad_norm 2.4774 (2.4465) [2021-04-16 14:24:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][30/1251] eta 0:12:51 lr 0.000130 time 0.2641 (0.6320) loss 2.1052 (3.2229) grad_norm 2.2244 (2.5310) [2021-04-16 14:24:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][40/1251] eta 0:11:05 lr 0.000130 time 0.2586 (0.5499) loss 2.2730 (3.1976) grad_norm 2.3529 (2.5511) [2021-04-16 14:24:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][50/1251] eta 0:09:55 lr 0.000130 time 0.2643 (0.4960) loss 3.2940 (3.1365) grad_norm 2.8665 (2.5583) [2021-04-16 14:24:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][60/1251] eta 0:09:07 lr 0.000130 time 0.2556 (0.4595) loss 2.8886 (3.1494) grad_norm 2.6570 (2.5671) [2021-04-16 14:24:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][70/1251] eta 0:08:32 lr 0.000130 time 0.2714 (0.4338) loss 3.5546 (3.1682) grad_norm 2.5397 (2.5638) [2021-04-16 14:25:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][80/1251] eta 0:08:06 lr 0.000130 time 0.2627 (0.4157) loss 3.3706 (3.1885) grad_norm 2.4241 (2.5707) [2021-04-16 14:25:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][90/1251] eta 0:07:45 lr 0.000130 time 0.2630 (0.4010) loss 2.1562 (3.1588) grad_norm 2.0058 (2.5401) [2021-04-16 14:25:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][100/1251] eta 0:07:28 lr 0.000130 time 0.2621 (0.3900) loss 3.0416 (3.1770) grad_norm 2.4921 (2.5397) [2021-04-16 14:25:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][110/1251] eta 0:07:13 lr 0.000130 time 0.2588 (0.3799) loss 3.6039 (3.2002) grad_norm 3.1254 (2.5475) [2021-04-16 14:25:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][120/1251] eta 0:07:01 lr 0.000130 time 0.2435 (0.3727) loss 2.0013 (3.1687) grad_norm 2.2874 (2.5446) [2021-04-16 14:25:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][130/1251] eta 0:06:51 lr 0.000130 time 0.2700 (0.3668) loss 3.1023 (3.1879) grad_norm 2.3092 (2.5442) [2021-04-16 14:25:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][140/1251] eta 0:06:40 lr 0.000130 time 0.2687 (0.3603) loss 2.5001 (3.1811) grad_norm 2.2338 (2.5356) [2021-04-16 14:25:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][150/1251] eta 0:06:30 lr 0.000130 time 0.2764 (0.3546) loss 2.3250 (3.1763) grad_norm 2.3109 (2.5308) [2021-04-16 14:25:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][160/1251] eta 0:06:22 lr 0.000130 time 0.2794 (0.3509) loss 3.9528 (3.1786) grad_norm 2.5027 (2.5249) [2021-04-16 14:25:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][170/1251] eta 0:06:14 lr 0.000130 time 0.2768 (0.3463) loss 3.5806 (3.1888) grad_norm 2.0943 (2.5227) [2021-04-16 14:25:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][180/1251] eta 0:06:06 lr 0.000130 time 0.2869 (0.3426) loss 3.3285 (3.1873) grad_norm 2.1907 (2.5192) [2021-04-16 14:25:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][190/1251] eta 0:05:59 lr 0.000130 time 0.2841 (0.3392) loss 2.0380 (3.1870) grad_norm 2.1674 (2.5155) [2021-04-16 14:25:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][200/1251] eta 0:05:53 lr 0.000130 time 0.2867 (0.3360) loss 3.4064 (3.1869) grad_norm 2.4293 (2.5051) [2021-04-16 14:25:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][210/1251] eta 0:05:46 lr 0.000130 time 0.2708 (0.3333) loss 3.8493 (3.2016) grad_norm 2.5241 (2.5074) [2021-04-16 14:25:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][220/1251] eta 0:05:40 lr 0.000130 time 0.2685 (0.3306) loss 3.9103 (3.1968) grad_norm 2.3090 (2.5085) [2021-04-16 14:25:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][230/1251] eta 0:05:37 lr 0.000130 time 0.3697 (0.3302) loss 2.5347 (3.1790) grad_norm 3.0875 (2.5115) [2021-04-16 14:25:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][240/1251] eta 0:05:31 lr 0.000130 time 0.2667 (0.3280) loss 3.2346 (3.1809) grad_norm 2.1780 (2.5087) [2021-04-16 14:25:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][250/1251] eta 0:05:26 lr 0.000130 time 0.2939 (0.3261) loss 3.0433 (3.1698) grad_norm 2.5285 (2.5108) [2021-04-16 14:25:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][260/1251] eta 0:05:21 lr 0.000130 time 0.2968 (0.3244) loss 2.9701 (3.1598) grad_norm 2.4782 (2.5089) [2021-04-16 14:25:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][270/1251] eta 0:05:16 lr 0.000130 time 0.2860 (0.3228) loss 3.5822 (3.1617) grad_norm 2.3233 (2.5074) [2021-04-16 14:25:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][280/1251] eta 0:05:11 lr 0.000130 time 0.2796 (0.3210) loss 2.4866 (3.1711) grad_norm 2.4072 (2.5073) [2021-04-16 14:26:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][290/1251] eta 0:05:06 lr 0.000130 time 0.2697 (0.3194) loss 2.8974 (3.1655) grad_norm 2.5803 (2.5091) [2021-04-16 14:26:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][300/1251] eta 0:05:02 lr 0.000129 time 0.2565 (0.3181) loss 2.8263 (3.1640) grad_norm 2.5736 (2.5081) [2021-04-16 14:26:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][310/1251] eta 0:04:58 lr 0.000129 time 0.2689 (0.3168) loss 3.6042 (3.1640) grad_norm 3.3537 (2.5099) [2021-04-16 14:26:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][320/1251] eta 0:04:54 lr 0.000129 time 0.2872 (0.3160) loss 3.2581 (3.1653) grad_norm 2.7989 (2.5212) [2021-04-16 14:26:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][330/1251] eta 0:04:50 lr 0.000129 time 0.3132 (0.3151) loss 2.3523 (3.1684) grad_norm 2.7109 (2.5305) [2021-04-16 14:26:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][340/1251] eta 0:04:45 lr 0.000129 time 0.2868 (0.3138) loss 2.4745 (3.1646) grad_norm 2.3137 (2.5310) [2021-04-16 14:26:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][350/1251] eta 0:04:41 lr 0.000129 time 0.2470 (0.3126) loss 3.2896 (3.1723) grad_norm 2.2982 (2.5246) [2021-04-16 14:26:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][360/1251] eta 0:04:37 lr 0.000129 time 0.2744 (0.3119) loss 2.4086 (3.1727) grad_norm 2.5057 (2.5217) [2021-04-16 14:26:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][370/1251] eta 0:04:33 lr 0.000129 time 0.2419 (0.3109) loss 2.4887 (3.1670) grad_norm 2.6440 (2.5173) [2021-04-16 14:26:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][380/1251] eta 0:04:30 lr 0.000129 time 0.2626 (0.3103) loss 2.2714 (3.1669) grad_norm 2.2604 (2.5158) [2021-04-16 14:26:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][390/1251] eta 0:04:26 lr 0.000129 time 0.2666 (0.3094) loss 2.8911 (3.1655) grad_norm 2.6545 (2.5143) [2021-04-16 14:26:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][400/1251] eta 0:04:22 lr 0.000129 time 0.2839 (0.3086) loss 3.7116 (3.1663) grad_norm 2.2057 (2.5125) [2021-04-16 14:26:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][410/1251] eta 0:04:18 lr 0.000129 time 0.2965 (0.3078) loss 2.0679 (3.1645) grad_norm 2.1519 (2.5089) [2021-04-16 14:26:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][420/1251] eta 0:04:15 lr 0.000129 time 0.2751 (0.3070) loss 3.4998 (3.1626) grad_norm 2.5144 (2.5062) [2021-04-16 14:26:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][430/1251] eta 0:04:11 lr 0.000129 time 0.2863 (0.3066) loss 3.9570 (3.1657) grad_norm 2.4071 (2.5049) [2021-04-16 14:26:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][440/1251] eta 0:04:08 lr 0.000129 time 0.2840 (0.3060) loss 3.6467 (3.1638) grad_norm 2.2539 (2.5026) [2021-04-16 14:26:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][450/1251] eta 0:04:04 lr 0.000129 time 0.2857 (0.3052) loss 2.5997 (3.1592) grad_norm 2.5133 (2.5036) [2021-04-16 14:26:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][460/1251] eta 0:04:00 lr 0.000129 time 0.2775 (0.3046) loss 3.3336 (3.1608) grad_norm 2.3585 (2.5020) [2021-04-16 14:26:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][470/1251] eta 0:03:57 lr 0.000129 time 0.2613 (0.3039) loss 3.7912 (3.1652) grad_norm 2.2905 (2.5017) [2021-04-16 14:26:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][480/1251] eta 0:03:53 lr 0.000129 time 0.2757 (0.3034) loss 3.1230 (3.1631) grad_norm 2.9544 (2.5013) [2021-04-16 14:26:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][490/1251] eta 0:03:50 lr 0.000129 time 0.2844 (0.3029) loss 3.4353 (3.1650) grad_norm 2.7770 (2.5037) [2021-04-16 14:27:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][500/1251] eta 0:03:47 lr 0.000129 time 0.2681 (0.3024) loss 2.5278 (3.1677) grad_norm 2.2756 (2.5037) [2021-04-16 14:27:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][510/1251] eta 0:03:43 lr 0.000129 time 0.2793 (0.3019) loss 1.9337 (3.1684) grad_norm 2.1219 (2.5021) [2021-04-16 14:27:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][520/1251] eta 0:03:40 lr 0.000129 time 0.2806 (0.3014) loss 3.2437 (3.1655) grad_norm 2.4748 (2.4987) [2021-04-16 14:27:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][530/1251] eta 0:03:37 lr 0.000129 time 0.3145 (0.3010) loss 2.8324 (3.1670) grad_norm 2.2637 (2.4987) [2021-04-16 14:27:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][540/1251] eta 0:03:33 lr 0.000129 time 0.2666 (0.3004) loss 2.4489 (3.1663) grad_norm 2.4366 (2.5021) [2021-04-16 14:27:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][550/1251] eta 0:03:30 lr 0.000129 time 0.2758 (0.3001) loss 3.2271 (3.1708) grad_norm 2.3739 (2.5014) [2021-04-16 14:27:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][560/1251] eta 0:03:27 lr 0.000129 time 0.2715 (0.2997) loss 3.6315 (3.1732) grad_norm 2.0935 (2.4999) [2021-04-16 14:27:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][570/1251] eta 0:03:23 lr 0.000129 time 0.2688 (0.2995) loss 2.3126 (3.1733) grad_norm 2.4444 (2.4992) [2021-04-16 14:27:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][580/1251] eta 0:03:20 lr 0.000129 time 0.2564 (0.2993) loss 3.5658 (3.1796) grad_norm 2.5960 (2.4996) [2021-04-16 14:27:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][590/1251] eta 0:03:17 lr 0.000129 time 0.2762 (0.2992) loss 2.6903 (3.1809) grad_norm 2.2715 (2.4981) [2021-04-16 14:27:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][600/1251] eta 0:03:14 lr 0.000129 time 0.2758 (0.2989) loss 2.5459 (3.1793) grad_norm 2.4417 (2.4960) [2021-04-16 14:27:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][610/1251] eta 0:03:11 lr 0.000129 time 0.2786 (0.2985) loss 3.1651 (3.1790) grad_norm 2.1855 (2.4957) [2021-04-16 14:27:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][620/1251] eta 0:03:08 lr 0.000129 time 0.2913 (0.2983) loss 3.7024 (3.1788) grad_norm 1.9008 (2.4924) [2021-04-16 14:27:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][630/1251] eta 0:03:05 lr 0.000129 time 0.2841 (0.2979) loss 2.7615 (3.1784) grad_norm 2.2451 (2.4917) [2021-04-16 14:27:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][640/1251] eta 0:03:01 lr 0.000129 time 0.2926 (0.2976) loss 3.2682 (3.1766) grad_norm 2.6624 (2.4924) [2021-04-16 14:27:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][650/1251] eta 0:02:58 lr 0.000129 time 0.2863 (0.2975) loss 3.6720 (3.1796) grad_norm 2.6840 (2.4929) [2021-04-16 14:27:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][660/1251] eta 0:02:55 lr 0.000129 time 0.2720 (0.2972) loss 3.3488 (3.1781) grad_norm 2.5143 (2.4923) [2021-04-16 14:27:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][670/1251] eta 0:02:52 lr 0.000128 time 0.2839 (0.2968) loss 3.6013 (3.1772) grad_norm 2.3603 (2.4913) [2021-04-16 14:27:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][680/1251] eta 0:02:49 lr 0.000128 time 0.2864 (0.2966) loss 2.7752 (3.1787) grad_norm 2.5632 (2.4906) [2021-04-16 14:27:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][690/1251] eta 0:02:46 lr 0.000128 time 0.2593 (0.2966) loss 2.1592 (3.1729) grad_norm 2.6051 (2.4938) [2021-04-16 14:27:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][700/1251] eta 0:02:43 lr 0.000128 time 0.2578 (0.2964) loss 3.4293 (3.1720) grad_norm 2.7771 (2.4937) [2021-04-16 14:27:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][710/1251] eta 0:02:40 lr 0.000128 time 0.2684 (0.2960) loss 3.5023 (3.1713) grad_norm 2.9144 (2.4941) [2021-04-16 14:28:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][720/1251] eta 0:02:37 lr 0.000128 time 0.2618 (0.2958) loss 3.4874 (3.1701) grad_norm 2.5681 (2.4969) [2021-04-16 14:28:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][730/1251] eta 0:02:34 lr 0.000128 time 0.4081 (0.2957) loss 3.4851 (3.1695) grad_norm 2.2853 (2.4969) [2021-04-16 14:28:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][740/1251] eta 0:02:30 lr 0.000128 time 0.2752 (0.2954) loss 2.6391 (3.1660) grad_norm 2.0141 (2.4953) [2021-04-16 14:28:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][750/1251] eta 0:02:27 lr 0.000128 time 0.2963 (0.2952) loss 2.7884 (3.1641) grad_norm 2.2344 (2.4940) [2021-04-16 14:28:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][760/1251] eta 0:02:24 lr 0.000128 time 0.2688 (0.2950) loss 3.3415 (3.1656) grad_norm 2.8141 (2.4939) [2021-04-16 14:28:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][770/1251] eta 0:02:21 lr 0.000128 time 0.2696 (0.2948) loss 3.6918 (3.1685) grad_norm 2.6279 (2.4947) [2021-04-16 14:28:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][780/1251] eta 0:02:18 lr 0.000128 time 0.2955 (0.2946) loss 2.0712 (3.1671) grad_norm 2.1900 (2.4921) [2021-04-16 14:28:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][790/1251] eta 0:02:15 lr 0.000128 time 0.2588 (0.2943) loss 2.6751 (3.1655) grad_norm 2.9575 (2.4931) [2021-04-16 14:28:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][800/1251] eta 0:02:12 lr 0.000128 time 0.2848 (0.2941) loss 2.2922 (3.1658) grad_norm 3.0278 (2.4973) [2021-04-16 14:28:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][810/1251] eta 0:02:09 lr 0.000128 time 0.2738 (0.2938) loss 3.2194 (3.1649) grad_norm 2.3641 (2.4963) [2021-04-16 14:28:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][820/1251] eta 0:02:06 lr 0.000128 time 0.2565 (0.2936) loss 3.8841 (3.1614) grad_norm 2.6117 (2.4978) [2021-04-16 14:28:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][830/1251] eta 0:02:03 lr 0.000128 time 0.2909 (0.2934) loss 2.9060 (3.1611) grad_norm 2.2816 (2.4985) [2021-04-16 14:28:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][840/1251] eta 0:02:00 lr 0.000128 time 0.3153 (0.2932) loss 3.4346 (3.1593) grad_norm 2.2407 (2.5008) [2021-04-16 14:28:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][850/1251] eta 0:01:57 lr 0.000128 time 0.2883 (0.2930) loss 4.1579 (3.1620) grad_norm 2.2020 (2.5020) [2021-04-16 14:28:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][860/1251] eta 0:01:54 lr 0.000128 time 0.2764 (0.2928) loss 3.3591 (3.1615) grad_norm 2.7037 (inf) [2021-04-16 14:28:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][870/1251] eta 0:01:51 lr 0.000128 time 0.2554 (0.2926) loss 3.2120 (3.1605) grad_norm 2.4697 (inf) [2021-04-16 14:28:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][880/1251] eta 0:01:48 lr 0.000128 time 0.2561 (0.2924) loss 3.5046 (3.1616) grad_norm 2.8463 (inf) [2021-04-16 14:28:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][890/1251] eta 0:01:45 lr 0.000128 time 0.2656 (0.2923) loss 3.4988 (3.1623) grad_norm 2.6657 (inf) [2021-04-16 14:28:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][900/1251] eta 0:01:42 lr 0.000128 time 0.2726 (0.2922) loss 3.3625 (3.1625) grad_norm 2.4937 (inf) [2021-04-16 14:28:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][910/1251] eta 0:01:39 lr 0.000128 time 0.3009 (0.2920) loss 3.2799 (3.1624) grad_norm 2.4908 (inf) [2021-04-16 14:28:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][920/1251] eta 0:01:36 lr 0.000128 time 0.2538 (0.2918) loss 3.6465 (3.1600) grad_norm 2.3401 (inf) [2021-04-16 14:28:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][930/1251] eta 0:01:33 lr 0.000128 time 0.2588 (0.2916) loss 2.5768 (3.1594) grad_norm 2.7622 (inf) [2021-04-16 14:29:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][940/1251] eta 0:01:30 lr 0.000128 time 0.2559 (0.2915) loss 3.3323 (3.1598) grad_norm 2.2128 (inf) [2021-04-16 14:29:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][950/1251] eta 0:01:27 lr 0.000128 time 0.2663 (0.2915) loss 2.9008 (3.1627) grad_norm 2.3075 (inf) [2021-04-16 14:29:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][960/1251] eta 0:01:24 lr 0.000128 time 0.2804 (0.2914) loss 3.2344 (3.1630) grad_norm 2.5650 (inf) [2021-04-16 14:29:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][970/1251] eta 0:01:21 lr 0.000128 time 0.2787 (0.2913) loss 3.6190 (3.1643) grad_norm 2.5599 (inf) [2021-04-16 14:29:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][980/1251] eta 0:01:18 lr 0.000128 time 0.2958 (0.2912) loss 2.8947 (3.1618) grad_norm 2.7678 (inf) [2021-04-16 14:29:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][990/1251] eta 0:01:15 lr 0.000128 time 0.2871 (0.2910) loss 2.8091 (3.1596) grad_norm 2.7353 (inf) [2021-04-16 14:29:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][1000/1251] eta 0:01:13 lr 0.000128 time 0.2891 (0.2909) loss 3.3665 (3.1593) grad_norm 2.3858 (inf) [2021-04-16 14:29:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][1010/1251] eta 0:01:10 lr 0.000128 time 0.2736 (0.2908) loss 2.7325 (3.1591) grad_norm 2.4269 (inf) [2021-04-16 14:29:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][1020/1251] eta 0:01:07 lr 0.000128 time 0.2676 (0.2907) loss 3.0307 (3.1598) grad_norm 2.4811 (inf) [2021-04-16 14:29:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][1030/1251] eta 0:01:04 lr 0.000128 time 0.2870 (0.2906) loss 3.5528 (3.1591) grad_norm 2.2199 (inf) [2021-04-16 14:29:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][1040/1251] eta 0:01:01 lr 0.000127 time 0.2672 (0.2904) loss 1.8706 (3.1575) grad_norm 2.3969 (inf) [2021-04-16 14:29:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][1050/1251] eta 0:00:58 lr 0.000127 time 0.2826 (0.2904) loss 3.1391 (3.1564) grad_norm 2.5581 (inf) [2021-04-16 14:29:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][1060/1251] eta 0:00:55 lr 0.000127 time 0.2991 (0.2902) loss 3.7485 (3.1536) grad_norm 2.5022 (inf) [2021-04-16 14:29:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][1070/1251] eta 0:00:52 lr 0.000127 time 0.3021 (0.2901) loss 3.2081 (3.1505) grad_norm 2.3248 (inf) [2021-04-16 14:29:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][1080/1251] eta 0:00:49 lr 0.000127 time 0.2701 (0.2900) loss 2.1280 (3.1493) grad_norm 2.3644 (inf) [2021-04-16 14:29:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][1090/1251] eta 0:00:46 lr 0.000127 time 0.3015 (0.2900) loss 2.7650 (3.1485) grad_norm 2.3505 (inf) [2021-04-16 14:29:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][1100/1251] eta 0:00:43 lr 0.000127 time 0.2760 (0.2898) loss 3.2881 (3.1473) grad_norm 2.8003 (inf) [2021-04-16 14:29:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][1110/1251] eta 0:00:40 lr 0.000127 time 0.2827 (0.2897) loss 2.7182 (3.1455) grad_norm 2.7223 (inf) [2021-04-16 14:29:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][1120/1251] eta 0:00:37 lr 0.000127 time 0.2704 (0.2896) loss 3.8811 (3.1460) grad_norm 2.2010 (inf) [2021-04-16 14:29:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][1130/1251] eta 0:00:35 lr 0.000127 time 0.2713 (0.2894) loss 2.1474 (3.1473) grad_norm 2.1210 (inf) [2021-04-16 14:29:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][1140/1251] eta 0:00:32 lr 0.000127 time 0.2702 (0.2893) loss 2.2358 (3.1463) grad_norm 3.4097 (inf) [2021-04-16 14:30:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][1150/1251] eta 0:00:29 lr 0.000127 time 0.2667 (0.2893) loss 3.1736 (3.1475) grad_norm 2.1597 (inf) [2021-04-16 14:30:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][1160/1251] eta 0:00:26 lr 0.000127 time 0.2796 (0.2893) loss 3.5142 (3.1470) grad_norm 2.0641 (inf) [2021-04-16 14:30:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][1170/1251] eta 0:00:23 lr 0.000127 time 0.2629 (0.2893) loss 2.8054 (3.1480) grad_norm 2.7095 (inf) [2021-04-16 14:30:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][1180/1251] eta 0:00:20 lr 0.000127 time 0.2668 (0.2891) loss 3.1419 (3.1487) grad_norm 2.8621 (inf) [2021-04-16 14:30:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][1190/1251] eta 0:00:17 lr 0.000127 time 0.2505 (0.2891) loss 3.5486 (3.1491) grad_norm 2.5375 (inf) [2021-04-16 14:30:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][1200/1251] eta 0:00:14 lr 0.000127 time 0.2916 (0.2890) loss 3.3004 (3.1480) grad_norm 2.5326 (inf) [2021-04-16 14:30:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][1210/1251] eta 0:00:11 lr 0.000127 time 0.2708 (0.2889) loss 3.3698 (3.1485) grad_norm 2.3642 (inf) [2021-04-16 14:30:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][1220/1251] eta 0:00:08 lr 0.000127 time 0.2871 (0.2887) loss 3.4928 (3.1479) grad_norm 2.6528 (inf) [2021-04-16 14:30:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][1230/1251] eta 0:00:06 lr 0.000127 time 0.2745 (0.2888) loss 3.6478 (3.1496) grad_norm 2.1610 (inf) [2021-04-16 14:30:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][1240/1251] eta 0:00:03 lr 0.000127 time 0.2484 (0.2886) loss 1.8821 (3.1497) grad_norm 2.2903 (inf) [2021-04-16 14:30:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [232/300][1250/1251] eta 0:00:00 lr 0.000127 time 0.2504 (0.2882) loss 3.7208 (3.1511) grad_norm 2.4891 (inf) [2021-04-16 14:30:43 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 232 training takes 0:06:14 [2021-04-16 14:30:43 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_232.pth saving...... [2021-04-16 14:31:06 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_232.pth saved !!! [2021-04-16 14:31:07 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.137 (1.137) Loss 0.8907 (0.8907) Acc@1 78.418 (78.418) Acc@5 94.141 (94.141) [2021-04-16 14:31:09 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.133 (0.256) Loss 0.8232 (0.8875) Acc@1 80.566 (79.208) Acc@5 95.020 (94.771) [2021-04-16 14:31:11 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.142 (0.228) Loss 0.9404 (0.8712) Acc@1 77.246 (79.608) Acc@5 94.434 (94.950) [2021-04-16 14:31:13 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.097 (0.233) Loss 0.8281 (0.8632) Acc@1 79.688 (79.684) Acc@5 95.312 (94.994) [2021-04-16 14:31:15 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.350 (0.216) Loss 0.8512 (0.8643) Acc@1 81.152 (79.616) Acc@5 95.020 (95.015) [2021-04-16 14:31:34 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 79.570 Acc@5 95.000 [2021-04-16 14:31:34 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 79.6% [2021-04-16 14:31:34 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 79.69% [2021-04-16 14:31:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][0/1251] eta 4:25:41 lr 0.000127 time 12.7427 (12.7427) loss 2.9236 (2.9236) grad_norm 2.6091 (2.6091) [2021-04-16 14:31:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][10/1251] eta 0:28:59 lr 0.000127 time 0.2604 (1.4014) loss 2.0065 (3.1757) grad_norm 2.5006 (2.6518) [2021-04-16 14:31:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][20/1251] eta 0:17:45 lr 0.000127 time 0.2855 (0.8658) loss 2.8295 (3.2628) grad_norm 2.3616 (2.6141) [2021-04-16 14:31:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][30/1251] eta 0:13:46 lr 0.000127 time 0.2518 (0.6765) loss 2.3838 (3.1585) grad_norm 2.2477 (2.5664) [2021-04-16 14:31:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][40/1251] eta 0:11:42 lr 0.000127 time 0.2734 (0.5800) loss 3.6587 (3.2360) grad_norm 2.6179 (2.5438) [2021-04-16 14:32:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][50/1251] eta 0:10:25 lr 0.000127 time 0.2617 (0.5209) loss 2.8875 (3.2214) grad_norm 2.5857 (2.5143) [2021-04-16 14:32:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][60/1251] eta 0:09:34 lr 0.000127 time 0.2919 (0.4826) loss 2.9429 (3.1628) grad_norm 2.6559 (2.5104) [2021-04-16 14:32:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][70/1251] eta 0:08:55 lr 0.000127 time 0.2594 (0.4531) loss 4.0280 (3.1898) grad_norm 2.2802 (2.4959) [2021-04-16 14:32:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][80/1251] eta 0:08:25 lr 0.000127 time 0.2820 (0.4318) loss 3.9952 (3.1989) grad_norm 2.4282 (2.4787) [2021-04-16 14:32:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][90/1251] eta 0:08:01 lr 0.000127 time 0.2695 (0.4146) loss 2.0675 (3.2014) grad_norm 2.1833 (2.5139) [2021-04-16 14:32:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][100/1251] eta 0:07:43 lr 0.000127 time 0.2806 (0.4028) loss 2.3599 (3.1828) grad_norm 2.2255 (2.4984) [2021-04-16 14:32:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][110/1251] eta 0:07:26 lr 0.000127 time 0.2803 (0.3913) loss 3.7785 (3.1945) grad_norm 2.4283 (2.4943) [2021-04-16 14:32:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][120/1251] eta 0:07:11 lr 0.000127 time 0.2712 (0.3816) loss 3.3558 (3.1980) grad_norm 2.2290 (2.4896) [2021-04-16 14:32:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][130/1251] eta 0:06:58 lr 0.000127 time 0.2984 (0.3736) loss 2.8532 (3.1957) grad_norm 2.4070 (2.4817) [2021-04-16 14:32:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][140/1251] eta 0:06:47 lr 0.000127 time 0.2823 (0.3667) loss 3.5905 (3.1929) grad_norm 2.6248 (2.4848) [2021-04-16 14:32:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][150/1251] eta 0:06:37 lr 0.000127 time 0.2568 (0.3614) loss 3.3499 (3.1820) grad_norm 2.3498 (2.4805) [2021-04-16 14:32:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][160/1251] eta 0:06:29 lr 0.000126 time 0.2714 (0.3567) loss 2.2639 (3.1887) grad_norm 2.5888 (2.4804) [2021-04-16 14:32:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][170/1251] eta 0:06:21 lr 0.000126 time 0.2846 (0.3529) loss 3.1768 (3.1825) grad_norm 2.5814 (2.4762) [2021-04-16 14:32:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][180/1251] eta 0:06:13 lr 0.000126 time 0.2768 (0.3485) loss 3.6957 (3.1758) grad_norm 2.7026 (2.4808) [2021-04-16 14:32:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][190/1251] eta 0:06:05 lr 0.000126 time 0.2864 (0.3448) loss 3.8031 (3.1803) grad_norm 3.4611 (2.5082) [2021-04-16 14:32:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][200/1251] eta 0:05:58 lr 0.000126 time 0.2808 (0.3412) loss 3.5045 (3.1932) grad_norm 2.5614 (2.5086) [2021-04-16 14:32:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][210/1251] eta 0:05:52 lr 0.000126 time 0.2801 (0.3382) loss 3.7457 (3.2035) grad_norm 2.3885 (2.5211) [2021-04-16 14:32:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][220/1251] eta 0:05:46 lr 0.000126 time 0.2643 (0.3360) loss 3.5398 (3.2063) grad_norm 2.6237 (2.5218) [2021-04-16 14:32:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][230/1251] eta 0:05:40 lr 0.000126 time 0.2796 (0.3336) loss 3.2204 (3.1994) grad_norm 2.4748 (2.5341) [2021-04-16 14:32:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][240/1251] eta 0:05:35 lr 0.000126 time 0.2746 (0.3315) loss 3.9307 (3.2007) grad_norm 2.0355 (2.5372) [2021-04-16 14:32:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][250/1251] eta 0:05:29 lr 0.000126 time 0.2768 (0.3293) loss 3.4418 (3.2092) grad_norm 2.5159 (2.5350) [2021-04-16 14:32:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][260/1251] eta 0:05:24 lr 0.000126 time 0.2683 (0.3275) loss 2.7793 (3.2041) grad_norm 2.4936 (2.5395) [2021-04-16 14:33:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][270/1251] eta 0:05:19 lr 0.000126 time 0.2801 (0.3258) loss 2.2987 (3.1983) grad_norm 2.1489 (2.5361) [2021-04-16 14:33:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][280/1251] eta 0:05:14 lr 0.000126 time 0.2664 (0.3242) loss 2.4671 (3.1916) grad_norm 2.0831 (2.5311) [2021-04-16 14:33:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][290/1251] eta 0:05:09 lr 0.000126 time 0.2668 (0.3226) loss 2.8964 (3.1745) grad_norm 2.2427 (2.5235) [2021-04-16 14:33:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][300/1251] eta 0:05:05 lr 0.000126 time 0.2610 (0.3210) loss 3.2329 (3.1771) grad_norm 2.5516 (2.5190) [2021-04-16 14:33:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][310/1251] eta 0:05:00 lr 0.000126 time 0.2778 (0.3197) loss 1.9609 (3.1736) grad_norm 2.2350 (2.5168) [2021-04-16 14:33:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][320/1251] eta 0:04:56 lr 0.000126 time 0.2761 (0.3182) loss 4.0195 (3.1692) grad_norm 2.3216 (2.5134) [2021-04-16 14:33:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][330/1251] eta 0:04:52 lr 0.000126 time 0.2786 (0.3172) loss 3.1420 (3.1722) grad_norm 2.4098 (2.5121) [2021-04-16 14:33:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][340/1251] eta 0:04:47 lr 0.000126 time 0.2496 (0.3161) loss 3.8248 (3.1732) grad_norm 2.4583 (2.5050) [2021-04-16 14:33:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][350/1251] eta 0:04:43 lr 0.000126 time 0.2969 (0.3150) loss 2.8157 (3.1772) grad_norm 2.3067 (2.5069) [2021-04-16 14:33:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][360/1251] eta 0:04:40 lr 0.000126 time 0.2693 (0.3147) loss 2.9519 (3.1782) grad_norm 2.2080 (2.5031) [2021-04-16 14:33:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][370/1251] eta 0:04:36 lr 0.000126 time 0.2529 (0.3138) loss 3.6026 (3.1787) grad_norm 2.3474 (2.4990) [2021-04-16 14:33:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][380/1251] eta 0:04:32 lr 0.000126 time 0.2874 (0.3129) loss 2.9335 (3.1715) grad_norm 1.9968 (2.4992) [2021-04-16 14:33:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][390/1251] eta 0:04:29 lr 0.000126 time 0.2740 (0.3127) loss 2.5602 (3.1719) grad_norm 2.2680 (2.4990) [2021-04-16 14:33:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][400/1251] eta 0:04:25 lr 0.000126 time 0.2555 (0.3118) loss 3.4356 (3.1717) grad_norm 2.0753 (2.5028) [2021-04-16 14:33:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][410/1251] eta 0:04:21 lr 0.000126 time 0.2969 (0.3110) loss 3.1963 (3.1684) grad_norm 2.0762 (2.5018) [2021-04-16 14:33:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][420/1251] eta 0:04:17 lr 0.000126 time 0.2724 (0.3103) loss 2.4534 (3.1707) grad_norm 2.3907 (2.5029) [2021-04-16 14:33:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][430/1251] eta 0:04:14 lr 0.000126 time 0.2782 (0.3098) loss 3.6514 (3.1704) grad_norm 2.3670 (2.5010) [2021-04-16 14:33:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][440/1251] eta 0:04:10 lr 0.000126 time 0.2919 (0.3091) loss 3.8928 (3.1700) grad_norm 2.2490 (2.4954) [2021-04-16 14:33:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][450/1251] eta 0:04:06 lr 0.000126 time 0.2719 (0.3083) loss 3.8732 (3.1699) grad_norm 2.1879 (2.4907) [2021-04-16 14:33:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][460/1251] eta 0:04:03 lr 0.000126 time 0.2595 (0.3076) loss 3.4727 (3.1726) grad_norm 2.5985 (2.4913) [2021-04-16 14:33:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][470/1251] eta 0:03:59 lr 0.000126 time 0.2594 (0.3069) loss 3.0903 (3.1689) grad_norm 2.1794 (2.4886) [2021-04-16 14:34:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][480/1251] eta 0:03:56 lr 0.000126 time 0.2932 (0.3067) loss 2.9293 (3.1687) grad_norm 2.4989 (2.4842) [2021-04-16 14:34:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][490/1251] eta 0:03:53 lr 0.000126 time 0.2792 (0.3064) loss 2.6662 (3.1683) grad_norm 2.5159 (2.4855) [2021-04-16 14:34:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][500/1251] eta 0:03:49 lr 0.000126 time 0.2871 (0.3058) loss 3.0883 (3.1676) grad_norm 2.1329 (2.4852) [2021-04-16 14:34:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][510/1251] eta 0:03:46 lr 0.000126 time 0.2954 (0.3053) loss 3.4141 (3.1687) grad_norm 2.2971 (2.4830) [2021-04-16 14:34:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][520/1251] eta 0:03:42 lr 0.000126 time 0.2964 (0.3050) loss 3.0785 (3.1678) grad_norm 2.3041 (2.4828) [2021-04-16 14:34:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][530/1251] eta 0:03:39 lr 0.000126 time 0.2814 (0.3043) loss 3.6541 (3.1683) grad_norm 2.4147 (2.4808) [2021-04-16 14:34:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][540/1251] eta 0:03:36 lr 0.000125 time 0.3149 (0.3039) loss 3.6067 (3.1712) grad_norm 2.3875 (2.4821) [2021-04-16 14:34:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][550/1251] eta 0:03:32 lr 0.000125 time 0.2551 (0.3034) loss 3.0433 (3.1701) grad_norm 2.2150 (2.4810) [2021-04-16 14:34:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][560/1251] eta 0:03:29 lr 0.000125 time 0.2598 (0.3030) loss 3.2814 (3.1704) grad_norm 2.8036 (2.4816) [2021-04-16 14:34:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][570/1251] eta 0:03:25 lr 0.000125 time 0.2775 (0.3025) loss 2.2530 (3.1635) grad_norm 2.8091 (2.4808) [2021-04-16 14:34:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][580/1251] eta 0:03:22 lr 0.000125 time 0.2773 (0.3024) loss 3.7874 (3.1677) grad_norm 2.6541 (2.4855) [2021-04-16 14:34:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][590/1251] eta 0:03:19 lr 0.000125 time 0.2744 (0.3019) loss 3.7366 (3.1690) grad_norm 2.4360 (2.4864) [2021-04-16 14:34:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][600/1251] eta 0:03:16 lr 0.000125 time 0.2654 (0.3015) loss 3.8140 (3.1714) grad_norm 2.5237 (2.4857) [2021-04-16 14:34:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][610/1251] eta 0:03:12 lr 0.000125 time 0.2993 (0.3011) loss 3.1375 (3.1661) grad_norm 2.1333 (2.4866) [2021-04-16 14:34:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][620/1251] eta 0:03:09 lr 0.000125 time 0.2857 (0.3007) loss 3.0244 (3.1668) grad_norm 2.4584 (2.4857) [2021-04-16 14:34:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][630/1251] eta 0:03:06 lr 0.000125 time 0.2769 (0.3004) loss 2.0979 (3.1685) grad_norm 2.4347 (2.4868) [2021-04-16 14:34:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][640/1251] eta 0:03:03 lr 0.000125 time 0.2743 (0.3001) loss 3.2231 (3.1657) grad_norm 2.5936 (2.4851) [2021-04-16 14:34:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][650/1251] eta 0:03:00 lr 0.000125 time 0.2761 (0.2998) loss 4.0182 (3.1683) grad_norm 2.7861 (2.4832) [2021-04-16 14:34:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][660/1251] eta 0:02:56 lr 0.000125 time 0.2821 (0.2994) loss 3.1888 (3.1687) grad_norm 2.3929 (2.4836) [2021-04-16 14:34:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][670/1251] eta 0:02:53 lr 0.000125 time 0.2820 (0.2992) loss 2.0539 (3.1687) grad_norm 2.8282 (2.4846) [2021-04-16 14:34:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][680/1251] eta 0:02:50 lr 0.000125 time 0.2473 (0.2990) loss 2.9986 (3.1684) grad_norm 2.1860 (2.4834) [2021-04-16 14:35:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][690/1251] eta 0:02:47 lr 0.000125 time 0.2805 (0.2987) loss 3.5419 (3.1679) grad_norm 2.3218 (2.4832) [2021-04-16 14:35:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][700/1251] eta 0:02:44 lr 0.000125 time 0.2419 (0.2983) loss 3.5247 (3.1711) grad_norm 2.3917 (2.4817) [2021-04-16 14:35:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][710/1251] eta 0:02:41 lr 0.000125 time 0.2615 (0.2981) loss 2.3588 (3.1683) grad_norm 2.6381 (2.4800) [2021-04-16 14:35:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][720/1251] eta 0:02:38 lr 0.000125 time 0.2789 (0.2978) loss 2.7665 (3.1652) grad_norm 2.1406 (2.4784) [2021-04-16 14:35:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][730/1251] eta 0:02:35 lr 0.000125 time 0.2733 (0.2977) loss 3.2915 (3.1636) grad_norm 2.1598 (2.4760) [2021-04-16 14:35:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][740/1251] eta 0:02:32 lr 0.000125 time 0.3620 (0.2978) loss 3.1380 (3.1689) grad_norm 3.1608 (2.4806) [2021-04-16 14:35:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][750/1251] eta 0:02:29 lr 0.000125 time 0.2824 (0.2975) loss 2.7656 (3.1702) grad_norm 2.3158 (nan) [2021-04-16 14:35:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][760/1251] eta 0:02:25 lr 0.000125 time 0.2754 (0.2972) loss 3.3977 (3.1734) grad_norm 2.6174 (nan) [2021-04-16 14:35:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][770/1251] eta 0:02:22 lr 0.000125 time 0.2782 (0.2969) loss 3.6499 (3.1724) grad_norm 2.2630 (nan) [2021-04-16 14:35:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][780/1251] eta 0:02:19 lr 0.000125 time 0.2755 (0.2966) loss 3.6645 (3.1723) grad_norm 2.8118 (nan) [2021-04-16 14:35:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][790/1251] eta 0:02:16 lr 0.000125 time 0.2860 (0.2965) loss 2.4076 (3.1702) grad_norm 2.3333 (nan) [2021-04-16 14:35:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][800/1251] eta 0:02:13 lr 0.000125 time 0.2952 (0.2962) loss 3.0275 (3.1724) grad_norm 2.2135 (nan) [2021-04-16 14:35:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][810/1251] eta 0:02:10 lr 0.000125 time 0.2618 (0.2959) loss 2.9202 (3.1747) grad_norm 2.5583 (nan) [2021-04-16 14:35:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][820/1251] eta 0:02:07 lr 0.000125 time 0.2973 (0.2958) loss 3.2665 (3.1740) grad_norm 2.2867 (nan) [2021-04-16 14:35:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][830/1251] eta 0:02:04 lr 0.000125 time 0.2801 (0.2955) loss 3.0539 (3.1743) grad_norm 2.6374 (nan) [2021-04-16 14:35:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][840/1251] eta 0:02:01 lr 0.000125 time 0.2853 (0.2952) loss 3.9082 (3.1759) grad_norm 1.9193 (nan) [2021-04-16 14:35:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][850/1251] eta 0:01:58 lr 0.000125 time 0.2678 (0.2951) loss 3.1701 (3.1776) grad_norm 2.5145 (nan) [2021-04-16 14:35:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][860/1251] eta 0:01:55 lr 0.000125 time 0.2983 (0.2949) loss 3.4960 (3.1775) grad_norm 2.3067 (nan) [2021-04-16 14:35:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][870/1251] eta 0:01:52 lr 0.000125 time 0.2711 (0.2947) loss 3.7803 (3.1746) grad_norm 2.2971 (nan) [2021-04-16 14:35:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][880/1251] eta 0:01:49 lr 0.000125 time 0.2910 (0.2945) loss 3.6293 (3.1731) grad_norm 2.5180 (nan) [2021-04-16 14:35:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][890/1251] eta 0:01:46 lr 0.000125 time 0.3000 (0.2942) loss 2.7374 (3.1750) grad_norm 2.6363 (nan) [2021-04-16 14:35:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][900/1251] eta 0:01:43 lr 0.000125 time 0.2461 (0.2940) loss 3.6276 (3.1754) grad_norm 2.4393 (nan) [2021-04-16 14:36:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][910/1251] eta 0:01:40 lr 0.000124 time 0.2668 (0.2939) loss 3.2913 (3.1755) grad_norm 2.2397 (nan) [2021-04-16 14:36:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][920/1251] eta 0:01:37 lr 0.000124 time 0.2542 (0.2937) loss 3.6553 (3.1767) grad_norm 2.6453 (nan) [2021-04-16 14:36:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][930/1251] eta 0:01:34 lr 0.000124 time 0.2683 (0.2936) loss 3.5030 (3.1773) grad_norm 2.2585 (nan) [2021-04-16 14:36:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][940/1251] eta 0:01:31 lr 0.000124 time 0.3900 (0.2937) loss 3.2388 (3.1767) grad_norm 2.4007 (nan) [2021-04-16 14:36:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][950/1251] eta 0:01:28 lr 0.000124 time 0.2733 (0.2935) loss 2.3748 (3.1743) grad_norm 2.1778 (nan) [2021-04-16 14:36:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][960/1251] eta 0:01:25 lr 0.000124 time 0.2641 (0.2933) loss 3.1155 (3.1725) grad_norm 2.2224 (nan) [2021-04-16 14:36:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][970/1251] eta 0:01:22 lr 0.000124 time 0.2564 (0.2932) loss 3.4557 (3.1726) grad_norm 2.3753 (nan) [2021-04-16 14:36:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][980/1251] eta 0:01:19 lr 0.000124 time 0.2767 (0.2931) loss 3.1216 (3.1737) grad_norm 2.4503 (nan) [2021-04-16 14:36:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][990/1251] eta 0:01:16 lr 0.000124 time 0.2819 (0.2928) loss 2.4540 (3.1733) grad_norm 3.5883 (nan) [2021-04-16 14:36:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][1000/1251] eta 0:01:13 lr 0.000124 time 0.2768 (0.2927) loss 3.3354 (3.1712) grad_norm 2.7908 (nan) [2021-04-16 14:36:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][1010/1251] eta 0:01:10 lr 0.000124 time 0.2693 (0.2926) loss 2.5085 (3.1693) grad_norm 2.6710 (nan) [2021-04-16 14:36:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][1020/1251] eta 0:01:07 lr 0.000124 time 0.2869 (0.2924) loss 2.8414 (3.1679) grad_norm 2.4197 (nan) [2021-04-16 14:36:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][1030/1251] eta 0:01:04 lr 0.000124 time 0.2637 (0.2922) loss 3.2651 (3.1688) grad_norm 2.1198 (nan) [2021-04-16 14:36:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][1040/1251] eta 0:01:01 lr 0.000124 time 0.2584 (0.2920) loss 3.4435 (3.1640) grad_norm 2.1557 (nan) [2021-04-16 14:36:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][1050/1251] eta 0:00:58 lr 0.000124 time 0.4486 (0.2921) loss 3.1921 (3.1637) grad_norm 2.3683 (nan) [2021-04-16 14:36:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][1060/1251] eta 0:00:55 lr 0.000124 time 0.2598 (0.2920) loss 3.4512 (3.1656) grad_norm 2.5170 (nan) [2021-04-16 14:36:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][1070/1251] eta 0:00:52 lr 0.000124 time 0.2912 (0.2920) loss 3.6070 (3.1673) grad_norm 2.7278 (nan) [2021-04-16 14:36:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][1080/1251] eta 0:00:49 lr 0.000124 time 0.2710 (0.2919) loss 2.8112 (3.1697) grad_norm 2.5522 (nan) [2021-04-16 14:36:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][1090/1251] eta 0:00:46 lr 0.000124 time 0.2668 (0.2919) loss 3.6311 (3.1708) grad_norm 2.6022 (nan) [2021-04-16 14:36:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][1100/1251] eta 0:00:44 lr 0.000124 time 0.3045 (0.2917) loss 2.2854 (3.1694) grad_norm 2.3999 (nan) [2021-04-16 14:36:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][1110/1251] eta 0:00:41 lr 0.000124 time 0.2849 (0.2916) loss 2.7159 (3.1698) grad_norm 2.4099 (nan) [2021-04-16 14:37:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][1120/1251] eta 0:00:38 lr 0.000124 time 0.2730 (0.2915) loss 3.4524 (3.1685) grad_norm 2.3167 (nan) [2021-04-16 14:37:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][1130/1251] eta 0:00:35 lr 0.000124 time 0.2981 (0.2913) loss 3.5686 (3.1698) grad_norm 2.0317 (nan) [2021-04-16 14:37:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][1140/1251] eta 0:00:32 lr 0.000124 time 0.2679 (0.2913) loss 3.6285 (3.1703) grad_norm 2.6555 (nan) [2021-04-16 14:37:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][1150/1251] eta 0:00:29 lr 0.000124 time 0.2740 (0.2912) loss 3.1525 (3.1706) grad_norm 2.4473 (nan) [2021-04-16 14:37:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][1160/1251] eta 0:00:26 lr 0.000124 time 0.2596 (0.2912) loss 2.1854 (3.1683) grad_norm 2.5135 (nan) [2021-04-16 14:37:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][1170/1251] eta 0:00:23 lr 0.000124 time 0.2941 (0.2911) loss 3.9461 (3.1722) grad_norm 2.6496 (nan) [2021-04-16 14:37:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][1180/1251] eta 0:00:20 lr 0.000124 time 0.4133 (0.2910) loss 2.2278 (3.1707) grad_norm 2.2250 (nan) [2021-04-16 14:37:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][1190/1251] eta 0:00:17 lr 0.000124 time 0.2897 (0.2909) loss 3.4736 (3.1703) grad_norm 2.3404 (nan) [2021-04-16 14:37:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][1200/1251] eta 0:00:14 lr 0.000124 time 0.2882 (0.2907) loss 3.3475 (3.1719) grad_norm 2.4358 (nan) [2021-04-16 14:37:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][1210/1251] eta 0:00:11 lr 0.000124 time 0.2748 (0.2907) loss 2.8710 (3.1703) grad_norm 2.4832 (nan) [2021-04-16 14:37:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][1220/1251] eta 0:00:09 lr 0.000124 time 0.2629 (0.2906) loss 3.2245 (3.1676) grad_norm 2.1966 (nan) [2021-04-16 14:37:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][1230/1251] eta 0:00:06 lr 0.000124 time 0.2941 (0.2905) loss 2.8541 (3.1666) grad_norm 2.6434 (nan) [2021-04-16 14:37:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][1240/1251] eta 0:00:03 lr 0.000124 time 0.2529 (0.2903) loss 3.5945 (3.1681) grad_norm 2.2636 (nan) [2021-04-16 14:37:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [233/300][1250/1251] eta 0:00:00 lr 0.000124 time 0.2513 (0.2900) loss 3.1967 (3.1699) grad_norm 2.6740 (nan) [2021-04-16 14:37:54 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 233 training takes 0:06:20 [2021-04-16 14:37:54 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_233.pth saving...... [2021-04-16 14:38:20 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_233.pth saved !!! [2021-04-16 14:38:21 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.092 (1.092) Loss 0.8583 (0.8583) Acc@1 79.102 (79.102) Acc@5 95.703 (95.703) [2021-04-16 14:38:22 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.686 (0.258) Loss 0.9135 (0.8864) Acc@1 79.688 (79.395) Acc@5 95.020 (94.735) [2021-04-16 14:38:25 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.502 (0.249) Loss 0.8868 (0.8679) Acc@1 79.199 (79.734) Acc@5 95.117 (94.927) [2021-04-16 14:38:27 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.103 (0.246) Loss 0.7991 (0.8637) Acc@1 80.469 (79.829) Acc@5 95.703 (94.931) [2021-04-16 14:38:29 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.085 (0.218) Loss 0.7831 (0.8567) Acc@1 81.836 (79.871) Acc@5 96.582 (94.972) [2021-04-16 14:38:43 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 79.778 Acc@5 94.978 [2021-04-16 14:38:43 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 79.8% [2021-04-16 14:38:43 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 79.78% [2021-04-16 14:38:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][0/1251] eta 2:50:38 lr 0.000124 time 8.1844 (8.1844) loss 3.0967 (3.0967) grad_norm 2.6621 (2.6621) [2021-04-16 14:38:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][10/1251] eta 0:20:41 lr 0.000124 time 0.3734 (1.0004) loss 2.6547 (3.1675) grad_norm 3.1833 (2.6539) [2021-04-16 14:38:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][20/1251] eta 0:13:29 lr 0.000124 time 0.2755 (0.6577) loss 1.9818 (3.0997) grad_norm 2.8036 (2.6168) [2021-04-16 14:39:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][30/1251] eta 0:10:51 lr 0.000124 time 0.2709 (0.5334) loss 3.6518 (3.1382) grad_norm 2.4166 (2.5817) [2021-04-16 14:39:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][40/1251] eta 0:09:30 lr 0.000123 time 0.3045 (0.4714) loss 3.3910 (3.1632) grad_norm 2.4298 (2.5606) [2021-04-16 14:39:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][50/1251] eta 0:08:39 lr 0.000123 time 0.2738 (0.4329) loss 3.2174 (3.1822) grad_norm 2.4169 (2.6125) [2021-04-16 14:39:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][60/1251] eta 0:08:06 lr 0.000123 time 0.3058 (0.4083) loss 1.9490 (3.1489) grad_norm 1.9506 (2.5936) [2021-04-16 14:39:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][70/1251] eta 0:07:41 lr 0.000123 time 0.2574 (0.3905) loss 3.6735 (3.1600) grad_norm 2.3017 (2.6033) [2021-04-16 14:39:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][80/1251] eta 0:07:20 lr 0.000123 time 0.2924 (0.3766) loss 2.1986 (3.1403) grad_norm 1.9640 (2.5557) [2021-04-16 14:39:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][90/1251] eta 0:07:05 lr 0.000123 time 0.2645 (0.3661) loss 3.3866 (3.1282) grad_norm 2.1344 (2.5350) [2021-04-16 14:39:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][100/1251] eta 0:06:51 lr 0.000123 time 0.2727 (0.3575) loss 2.9217 (3.1260) grad_norm 2.2963 (2.5262) [2021-04-16 14:39:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][110/1251] eta 0:06:39 lr 0.000123 time 0.2631 (0.3501) loss 3.4015 (3.1522) grad_norm 2.2885 (2.5141) [2021-04-16 14:39:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][120/1251] eta 0:06:28 lr 0.000123 time 0.2742 (0.3439) loss 3.6767 (3.1547) grad_norm 2.4422 (2.4970) [2021-04-16 14:39:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][130/1251] eta 0:06:19 lr 0.000123 time 0.2796 (0.3386) loss 2.9013 (3.1442) grad_norm 2.1332 (2.4851) [2021-04-16 14:39:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][140/1251] eta 0:06:12 lr 0.000123 time 0.2607 (0.3352) loss 3.7514 (3.1318) grad_norm 2.5070 (2.4796) [2021-04-16 14:39:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][150/1251] eta 0:06:05 lr 0.000123 time 0.3913 (0.3323) loss 2.3798 (3.1424) grad_norm 2.1448 (2.4848) [2021-04-16 14:39:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][160/1251] eta 0:05:59 lr 0.000123 time 0.2728 (0.3296) loss 3.8696 (3.1529) grad_norm 2.4475 (2.4810) [2021-04-16 14:39:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][170/1251] eta 0:05:52 lr 0.000123 time 0.2750 (0.3265) loss 3.8455 (3.1545) grad_norm 2.1072 (2.4803) [2021-04-16 14:39:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][180/1251] eta 0:05:47 lr 0.000123 time 0.2823 (0.3247) loss 2.9685 (3.1700) grad_norm 2.2346 (2.4857) [2021-04-16 14:39:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][190/1251] eta 0:05:41 lr 0.000123 time 0.2919 (0.3222) loss 3.0363 (3.1601) grad_norm 2.3383 (2.4898) [2021-04-16 14:39:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][200/1251] eta 0:05:36 lr 0.000123 time 0.2788 (0.3203) loss 3.5411 (3.1493) grad_norm 2.6382 (2.4849) [2021-04-16 14:39:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][210/1251] eta 0:05:31 lr 0.000123 time 0.2503 (0.3184) loss 3.8771 (3.1600) grad_norm 2.2413 (2.4864) [2021-04-16 14:39:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][220/1251] eta 0:05:27 lr 0.000123 time 0.2658 (0.3172) loss 2.9233 (3.1633) grad_norm 2.3565 (2.4867) [2021-04-16 14:39:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][230/1251] eta 0:05:21 lr 0.000123 time 0.2689 (0.3153) loss 3.4122 (3.1550) grad_norm 2.4486 (2.4819) [2021-04-16 14:39:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][240/1251] eta 0:05:17 lr 0.000123 time 0.2788 (0.3136) loss 2.2590 (3.1586) grad_norm 2.3460 (2.4812) [2021-04-16 14:40:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][250/1251] eta 0:05:12 lr 0.000123 time 0.2675 (0.3123) loss 3.4166 (3.1629) grad_norm 2.6075 (2.4769) [2021-04-16 14:40:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][260/1251] eta 0:05:08 lr 0.000123 time 0.4764 (0.3116) loss 2.5359 (3.1598) grad_norm 2.2656 (2.4812) [2021-04-16 14:40:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][270/1251] eta 0:05:04 lr 0.000123 time 0.2698 (0.3101) loss 3.3514 (3.1588) grad_norm 2.6803 (2.4845) [2021-04-16 14:40:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][280/1251] eta 0:04:59 lr 0.000123 time 0.2650 (0.3089) loss 3.3398 (3.1619) grad_norm 2.3410 (2.4809) [2021-04-16 14:40:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][290/1251] eta 0:04:56 lr 0.000123 time 0.2985 (0.3082) loss 3.9616 (3.1619) grad_norm 2.2186 (2.4762) [2021-04-16 14:40:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][300/1251] eta 0:04:52 lr 0.000123 time 0.2561 (0.3079) loss 2.7029 (3.1607) grad_norm 4.3312 (2.4826) [2021-04-16 14:40:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][310/1251] eta 0:04:48 lr 0.000123 time 0.2756 (0.3068) loss 3.3473 (3.1577) grad_norm 2.5452 (2.4835) [2021-04-16 14:40:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][320/1251] eta 0:04:44 lr 0.000123 time 0.2715 (0.3060) loss 3.5566 (3.1583) grad_norm 2.5915 (2.4853) [2021-04-16 14:40:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][330/1251] eta 0:04:41 lr 0.000123 time 0.2830 (0.3051) loss 2.8239 (3.1568) grad_norm 2.4391 (2.4847) [2021-04-16 14:40:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][340/1251] eta 0:04:37 lr 0.000123 time 0.2698 (0.3042) loss 2.8876 (3.1495) grad_norm 2.1510 (2.4793) [2021-04-16 14:40:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][350/1251] eta 0:04:33 lr 0.000123 time 0.2753 (0.3036) loss 3.2189 (3.1516) grad_norm 2.6017 (2.4754) [2021-04-16 14:40:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][360/1251] eta 0:04:30 lr 0.000123 time 0.2459 (0.3035) loss 3.3084 (3.1484) grad_norm 2.2363 (2.4749) [2021-04-16 14:40:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][370/1251] eta 0:04:26 lr 0.000123 time 0.2813 (0.3027) loss 3.6019 (3.1512) grad_norm 2.6258 (2.4702) [2021-04-16 14:40:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][380/1251] eta 0:04:22 lr 0.000123 time 0.2656 (0.3019) loss 3.3079 (3.1526) grad_norm 2.1422 (2.4670) [2021-04-16 14:40:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][390/1251] eta 0:04:19 lr 0.000123 time 0.2758 (0.3012) loss 3.3356 (3.1533) grad_norm 2.4308 (2.4709) [2021-04-16 14:40:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][400/1251] eta 0:04:15 lr 0.000123 time 0.2814 (0.3006) loss 3.5629 (3.1537) grad_norm 2.5475 (2.4741) [2021-04-16 14:40:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][410/1251] eta 0:04:12 lr 0.000123 time 0.2832 (0.2999) loss 3.0259 (3.1577) grad_norm 2.3410 (2.4773) [2021-04-16 14:40:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][420/1251] eta 0:04:08 lr 0.000122 time 0.2782 (0.2994) loss 3.4541 (3.1594) grad_norm 2.8066 (2.4764) [2021-04-16 14:40:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][430/1251] eta 0:04:05 lr 0.000122 time 0.2691 (0.2989) loss 3.3921 (3.1568) grad_norm 2.4402 (2.4733) [2021-04-16 14:40:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][440/1251] eta 0:04:02 lr 0.000122 time 0.2982 (0.2985) loss 3.2180 (3.1543) grad_norm 2.2730 (2.4695) [2021-04-16 14:40:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][450/1251] eta 0:03:58 lr 0.000122 time 0.2938 (0.2983) loss 2.3840 (3.1477) grad_norm 2.7503 (2.4698) [2021-04-16 14:41:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][460/1251] eta 0:03:55 lr 0.000122 time 0.2637 (0.2978) loss 2.9428 (3.1449) grad_norm 2.5467 (2.4684) [2021-04-16 14:41:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][470/1251] eta 0:03:52 lr 0.000122 time 0.2797 (0.2973) loss 3.0306 (3.1439) grad_norm 2.1996 (2.4688) [2021-04-16 14:41:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][480/1251] eta 0:03:49 lr 0.000122 time 0.2648 (0.2971) loss 2.4287 (3.1344) grad_norm 2.6897 (2.4671) [2021-04-16 14:41:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][490/1251] eta 0:03:45 lr 0.000122 time 0.2784 (0.2968) loss 3.6616 (3.1387) grad_norm 2.5572 (2.4700) [2021-04-16 14:41:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][500/1251] eta 0:03:42 lr 0.000122 time 0.2879 (0.2964) loss 2.4914 (3.1425) grad_norm 2.6147 (2.4724) [2021-04-16 14:41:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][510/1251] eta 0:03:39 lr 0.000122 time 0.2930 (0.2961) loss 2.2294 (3.1420) grad_norm 2.4019 (2.4795) [2021-04-16 14:41:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][520/1251] eta 0:03:36 lr 0.000122 time 0.2669 (0.2958) loss 2.1742 (3.1422) grad_norm 2.2159 (2.4792) [2021-04-16 14:41:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][530/1251] eta 0:03:33 lr 0.000122 time 0.2489 (0.2955) loss 3.9615 (3.1403) grad_norm 2.2995 (2.4766) [2021-04-16 14:41:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][540/1251] eta 0:03:29 lr 0.000122 time 0.2548 (0.2951) loss 3.0119 (3.1378) grad_norm 2.2745 (2.4743) [2021-04-16 14:41:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][550/1251] eta 0:03:26 lr 0.000122 time 0.2636 (0.2948) loss 2.7401 (3.1416) grad_norm 2.5842 (2.4736) [2021-04-16 14:41:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][560/1251] eta 0:03:23 lr 0.000122 time 0.2663 (0.2944) loss 3.8853 (3.1450) grad_norm 2.3568 (2.4731) [2021-04-16 14:41:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][570/1251] eta 0:03:20 lr 0.000122 time 0.2626 (0.2943) loss 3.3191 (3.1509) grad_norm 2.0994 (2.4711) [2021-04-16 14:41:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][580/1251] eta 0:03:17 lr 0.000122 time 0.2802 (0.2942) loss 2.7551 (3.1525) grad_norm 2.6568 (2.4731) [2021-04-16 14:41:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][590/1251] eta 0:03:14 lr 0.000122 time 0.2943 (0.2942) loss 2.0031 (3.1475) grad_norm 2.2350 (2.4723) [2021-04-16 14:41:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][600/1251] eta 0:03:11 lr 0.000122 time 0.2606 (0.2939) loss 3.2777 (3.1461) grad_norm 2.1592 (2.4725) [2021-04-16 14:41:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][610/1251] eta 0:03:08 lr 0.000122 time 0.2518 (0.2936) loss 3.2854 (3.1439) grad_norm 2.1995 (2.4726) [2021-04-16 14:41:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][620/1251] eta 0:03:05 lr 0.000122 time 0.2611 (0.2933) loss 2.5837 (3.1450) grad_norm 2.5259 (2.4737) [2021-04-16 14:41:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][630/1251] eta 0:03:01 lr 0.000122 time 0.2770 (0.2931) loss 1.9249 (3.1487) grad_norm 2.0388 (2.4728) [2021-04-16 14:41:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][640/1251] eta 0:02:58 lr 0.000122 time 0.2883 (0.2928) loss 1.9365 (3.1432) grad_norm 2.4273 (2.4721) [2021-04-16 14:41:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][650/1251] eta 0:02:55 lr 0.000122 time 0.3030 (0.2926) loss 3.3878 (3.1423) grad_norm 2.6688 (2.4743) [2021-04-16 14:41:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][660/1251] eta 0:02:52 lr 0.000122 time 0.3228 (0.2925) loss 3.5539 (3.1448) grad_norm 2.4722 (2.4753) [2021-04-16 14:41:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][670/1251] eta 0:02:49 lr 0.000122 time 0.2764 (0.2922) loss 3.3684 (3.1478) grad_norm 2.7150 (2.4761) [2021-04-16 14:42:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][680/1251] eta 0:02:46 lr 0.000122 time 0.2674 (0.2919) loss 3.4926 (3.1483) grad_norm 2.9087 (2.4751) [2021-04-16 14:42:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][690/1251] eta 0:02:43 lr 0.000122 time 0.2783 (0.2917) loss 3.1544 (3.1479) grad_norm 2.7553 (2.4775) [2021-04-16 14:42:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][700/1251] eta 0:02:40 lr 0.000122 time 0.2768 (0.2915) loss 2.9378 (3.1475) grad_norm 2.2227 (2.4762) [2021-04-16 14:42:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][710/1251] eta 0:02:37 lr 0.000122 time 0.2602 (0.2913) loss 3.5203 (3.1448) grad_norm 2.0867 (2.4756) [2021-04-16 14:42:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][720/1251] eta 0:02:34 lr 0.000122 time 0.2792 (0.2912) loss 3.8346 (3.1448) grad_norm 2.4818 (2.4770) [2021-04-16 14:42:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][730/1251] eta 0:02:31 lr 0.000122 time 0.4334 (0.2914) loss 3.2906 (3.1442) grad_norm 2.4346 (2.4761) [2021-04-16 14:42:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][740/1251] eta 0:02:28 lr 0.000122 time 0.2619 (0.2911) loss 3.7403 (3.1477) grad_norm 2.3340 (2.4754) [2021-04-16 14:42:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][750/1251] eta 0:02:25 lr 0.000122 time 0.2606 (0.2909) loss 3.5365 (3.1467) grad_norm 2.6989 (2.4744) [2021-04-16 14:42:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][760/1251] eta 0:02:22 lr 0.000122 time 0.2582 (0.2908) loss 3.4453 (3.1475) grad_norm 2.1206 (2.4729) [2021-04-16 14:42:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][770/1251] eta 0:02:19 lr 0.000122 time 0.2642 (0.2907) loss 3.5943 (3.1470) grad_norm 2.5894 (2.4733) [2021-04-16 14:42:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][780/1251] eta 0:02:16 lr 0.000122 time 0.2560 (0.2906) loss 3.4701 (3.1492) grad_norm 2.0956 (2.4723) [2021-04-16 14:42:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][790/1251] eta 0:02:13 lr 0.000122 time 0.2686 (0.2905) loss 3.5567 (3.1506) grad_norm 2.8022 (2.4741) [2021-04-16 14:42:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][800/1251] eta 0:02:10 lr 0.000121 time 0.2785 (0.2904) loss 3.0395 (3.1486) grad_norm 2.2110 (2.4794) [2021-04-16 14:42:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][810/1251] eta 0:02:08 lr 0.000121 time 0.2795 (0.2903) loss 3.4831 (3.1492) grad_norm 3.0174 (2.4822) [2021-04-16 14:42:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][820/1251] eta 0:02:05 lr 0.000121 time 0.2760 (0.2901) loss 3.6825 (3.1498) grad_norm 2.4241 (2.4811) [2021-04-16 14:42:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][830/1251] eta 0:02:02 lr 0.000121 time 0.2743 (0.2899) loss 3.3935 (3.1517) grad_norm 2.1525 (2.4808) [2021-04-16 14:42:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][840/1251] eta 0:01:59 lr 0.000121 time 0.2982 (0.2899) loss 1.9322 (3.1521) grad_norm 2.2095 (2.4807) [2021-04-16 14:42:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][850/1251] eta 0:01:56 lr 0.000121 time 0.2862 (0.2897) loss 3.7712 (3.1518) grad_norm 2.5446 (2.4802) [2021-04-16 14:42:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][860/1251] eta 0:01:53 lr 0.000121 time 0.2943 (0.2897) loss 3.6383 (3.1547) grad_norm 2.3017 (2.4796) [2021-04-16 14:42:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][870/1251] eta 0:01:50 lr 0.000121 time 0.2839 (0.2896) loss 2.8383 (3.1527) grad_norm 2.8928 (2.4813) [2021-04-16 14:42:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][880/1251] eta 0:01:47 lr 0.000121 time 0.2750 (0.2896) loss 3.3229 (3.1546) grad_norm 2.6412 (2.4815) [2021-04-16 14:43:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][890/1251] eta 0:01:44 lr 0.000121 time 0.2794 (0.2895) loss 2.5712 (3.1568) grad_norm 2.1383 (2.4822) [2021-04-16 14:43:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][900/1251] eta 0:01:41 lr 0.000121 time 0.2787 (0.2894) loss 2.9720 (3.1541) grad_norm 2.5473 (2.4819) [2021-04-16 14:43:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][910/1251] eta 0:01:38 lr 0.000121 time 0.2887 (0.2893) loss 3.3853 (3.1545) grad_norm 2.5619 (2.4833) [2021-04-16 14:43:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][920/1251] eta 0:01:35 lr 0.000121 time 0.2565 (0.2892) loss 3.1290 (3.1557) grad_norm 2.4347 (2.4830) [2021-04-16 14:43:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][930/1251] eta 0:01:32 lr 0.000121 time 0.2756 (0.2892) loss 4.0426 (3.1556) grad_norm 2.7376 (2.4824) [2021-04-16 14:43:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][940/1251] eta 0:01:29 lr 0.000121 time 0.2748 (0.2893) loss 3.9148 (3.1546) grad_norm 2.2300 (2.4816) [2021-04-16 14:43:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][950/1251] eta 0:01:27 lr 0.000121 time 0.2747 (0.2892) loss 3.5950 (3.1562) grad_norm 2.5236 (2.4822) [2021-04-16 14:43:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][960/1251] eta 0:01:24 lr 0.000121 time 0.2678 (0.2891) loss 2.5949 (3.1506) grad_norm 2.7047 (2.4808) [2021-04-16 14:43:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][970/1251] eta 0:01:21 lr 0.000121 time 0.2789 (0.2890) loss 3.2703 (3.1503) grad_norm 2.3869 (2.4806) [2021-04-16 14:43:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][980/1251] eta 0:01:18 lr 0.000121 time 0.2718 (0.2888) loss 3.0252 (3.1514) grad_norm 2.9038 (2.4806) [2021-04-16 14:43:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][990/1251] eta 0:01:15 lr 0.000121 time 0.2888 (0.2887) loss 2.8093 (3.1516) grad_norm 2.1418 (2.4816) [2021-04-16 14:43:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][1000/1251] eta 0:01:12 lr 0.000121 time 0.2671 (0.2887) loss 2.8091 (3.1515) grad_norm 2.6946 (2.4847) [2021-04-16 14:43:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][1010/1251] eta 0:01:09 lr 0.000121 time 0.2672 (0.2885) loss 3.2188 (3.1506) grad_norm 2.4955 (2.4857) [2021-04-16 14:43:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][1020/1251] eta 0:01:06 lr 0.000121 time 0.2716 (0.2884) loss 3.6519 (3.1491) grad_norm 2.5297 (2.4847) [2021-04-16 14:43:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][1030/1251] eta 0:01:03 lr 0.000121 time 0.2989 (0.2883) loss 2.7407 (3.1518) grad_norm 3.2222 (2.4852) [2021-04-16 14:43:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][1040/1251] eta 0:01:00 lr 0.000121 time 0.2794 (0.2882) loss 3.5007 (3.1514) grad_norm 2.2758 (2.4839) [2021-04-16 14:43:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][1050/1251] eta 0:00:57 lr 0.000121 time 0.3951 (0.2882) loss 3.2016 (3.1497) grad_norm 2.5423 (2.4867) [2021-04-16 14:43:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][1060/1251] eta 0:00:55 lr 0.000121 time 0.2545 (0.2881) loss 3.3309 (3.1515) grad_norm 2.6740 (2.4868) [2021-04-16 14:43:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][1070/1251] eta 0:00:52 lr 0.000121 time 0.2569 (0.2880) loss 3.6657 (3.1501) grad_norm 2.3053 (2.4863) [2021-04-16 14:43:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][1080/1251] eta 0:00:49 lr 0.000121 time 0.2578 (0.2880) loss 3.1826 (3.1477) grad_norm 2.6797 (2.4865) [2021-04-16 14:43:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][1090/1251] eta 0:00:46 lr 0.000121 time 0.2859 (0.2879) loss 3.3837 (3.1493) grad_norm 2.5092 (2.4880) [2021-04-16 14:44:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][1100/1251] eta 0:00:43 lr 0.000121 time 0.2558 (0.2877) loss 2.2213 (3.1508) grad_norm 2.3036 (2.4912) [2021-04-16 14:44:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][1110/1251] eta 0:00:40 lr 0.000121 time 0.2819 (0.2877) loss 3.0213 (3.1522) grad_norm 2.4267 (2.4911) [2021-04-16 14:44:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][1120/1251] eta 0:00:37 lr 0.000121 time 0.2754 (0.2875) loss 3.3170 (3.1525) grad_norm 2.2894 (2.4918) [2021-04-16 14:44:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][1130/1251] eta 0:00:34 lr 0.000121 time 0.2611 (0.2875) loss 3.5489 (3.1513) grad_norm 2.3895 (2.4916) [2021-04-16 14:44:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][1140/1251] eta 0:00:31 lr 0.000121 time 0.2791 (0.2874) loss 2.7817 (3.1508) grad_norm 2.5443 (2.4916) [2021-04-16 14:44:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][1150/1251] eta 0:00:29 lr 0.000121 time 0.3040 (0.2874) loss 3.3214 (3.1504) grad_norm 2.3449 (2.4914) [2021-04-16 14:44:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][1160/1251] eta 0:00:26 lr 0.000121 time 0.2747 (0.2874) loss 3.4352 (3.1504) grad_norm 2.4692 (2.4926) [2021-04-16 14:44:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][1170/1251] eta 0:00:23 lr 0.000121 time 0.3051 (0.2874) loss 2.2841 (3.1494) grad_norm 2.2935 (2.4929) [2021-04-16 14:44:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][1180/1251] eta 0:00:20 lr 0.000120 time 0.2789 (0.2874) loss 3.1107 (3.1504) grad_norm 2.6818 (2.4938) [2021-04-16 14:44:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][1190/1251] eta 0:00:17 lr 0.000120 time 0.3106 (0.2873) loss 3.5804 (3.1502) grad_norm 2.3529 (2.4930) [2021-04-16 14:44:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][1200/1251] eta 0:00:14 lr 0.000120 time 0.2737 (0.2872) loss 2.8474 (3.1483) grad_norm 2.1711 (2.4940) [2021-04-16 14:44:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][1210/1251] eta 0:00:11 lr 0.000120 time 0.2693 (0.2871) loss 3.7160 (3.1475) grad_norm 3.4126 (2.4956) [2021-04-16 14:44:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][1220/1251] eta 0:00:08 lr 0.000120 time 0.2781 (0.2870) loss 3.0361 (3.1480) grad_norm 2.4967 (2.4953) [2021-04-16 14:44:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][1230/1251] eta 0:00:06 lr 0.000120 time 0.2802 (0.2869) loss 3.1878 (3.1471) grad_norm 2.5693 (2.4949) [2021-04-16 14:44:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][1240/1251] eta 0:00:03 lr 0.000120 time 0.2486 (0.2868) loss 3.3579 (3.1473) grad_norm 2.5942 (2.4953) [2021-04-16 14:44:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [234/300][1250/1251] eta 0:00:00 lr 0.000120 time 0.2480 (0.2865) loss 3.3670 (3.1495) grad_norm 2.3815 (2.4955) [2021-04-16 14:45:08 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 234 training takes 0:06:24 [2021-04-16 14:45:08 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_234.pth saving...... [2021-04-16 14:45:29 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_234.pth saved !!! [2021-04-16 14:45:30 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.131 (1.131) Loss 0.8492 (0.8492) Acc@1 79.395 (79.395) Acc@5 95.508 (95.508) [2021-04-16 14:45:31 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.125 (0.245) Loss 0.8994 (0.8560) Acc@1 80.859 (79.794) Acc@5 94.336 (95.135) [2021-04-16 14:45:33 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.128 (0.219) Loss 0.9171 (0.8618) Acc@1 77.246 (79.734) Acc@5 94.922 (94.982) [2021-04-16 14:45:35 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.189 (0.210) Loss 0.7977 (0.8619) Acc@1 80.273 (79.647) Acc@5 96.484 (94.994) [2021-04-16 14:45:38 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.220) Loss 0.8616 (0.8608) Acc@1 79.688 (79.630) Acc@5 94.434 (94.965) [2021-04-16 14:45:51 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 79.728 Acc@5 94.978 [2021-04-16 14:45:51 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 79.7% [2021-04-16 14:45:51 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 79.78% [2021-04-16 14:45:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][0/1251] eta 2:37:56 lr 0.000120 time 7.5748 (7.5748) loss 3.3701 (3.3701) grad_norm 2.5393 (2.5393) [2021-04-16 14:46:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][10/1251] eta 0:19:41 lr 0.000120 time 0.4470 (0.9519) loss 2.6139 (3.0732) grad_norm 2.5879 (2.3648) [2021-04-16 14:46:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][20/1251] eta 0:12:56 lr 0.000120 time 0.3133 (0.6309) loss 3.2347 (3.0884) grad_norm 2.7266 (2.3809) [2021-04-16 14:46:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][30/1251] eta 0:10:30 lr 0.000120 time 0.2682 (0.5163) loss 3.0695 (3.1260) grad_norm 2.3957 (2.3675) [2021-04-16 14:46:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][40/1251] eta 0:09:14 lr 0.000120 time 0.2711 (0.4579) loss 3.0481 (3.1315) grad_norm 2.3047 (2.4114) [2021-04-16 14:46:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][50/1251] eta 0:08:31 lr 0.000120 time 0.2863 (0.4262) loss 3.8480 (3.1792) grad_norm 2.7816 (2.4213) [2021-04-16 14:46:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][60/1251] eta 0:07:58 lr 0.000120 time 0.2534 (0.4018) loss 3.5282 (3.1762) grad_norm 2.5226 (2.4377) [2021-04-16 14:46:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][70/1251] eta 0:07:34 lr 0.000120 time 0.3006 (0.3848) loss 3.5813 (3.1594) grad_norm 2.4938 (2.4422) [2021-04-16 14:46:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][80/1251] eta 0:07:16 lr 0.000120 time 0.2502 (0.3729) loss 3.3577 (3.1807) grad_norm 2.5647 (2.4472) [2021-04-16 14:46:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][90/1251] eta 0:07:03 lr 0.000120 time 0.3128 (0.3649) loss 3.6528 (3.1876) grad_norm 2.7564 (2.4760) [2021-04-16 14:46:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][100/1251] eta 0:06:49 lr 0.000120 time 0.2695 (0.3557) loss 3.6666 (3.1704) grad_norm 2.9770 (2.4739) [2021-04-16 14:46:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][110/1251] eta 0:06:37 lr 0.000120 time 0.2467 (0.3486) loss 2.7713 (3.1526) grad_norm 2.9629 (2.4763) [2021-04-16 14:46:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][120/1251] eta 0:06:27 lr 0.000120 time 0.2495 (0.3425) loss 3.4287 (3.1729) grad_norm 2.4395 (2.4830) [2021-04-16 14:46:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][130/1251] eta 0:06:18 lr 0.000120 time 0.2710 (0.3377) loss 3.5211 (3.1698) grad_norm 2.6458 (2.4812) [2021-04-16 14:46:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][140/1251] eta 0:06:11 lr 0.000120 time 0.2610 (0.3346) loss 3.1129 (3.1797) grad_norm 2.4166 (2.4784) [2021-04-16 14:46:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][150/1251] eta 0:06:05 lr 0.000120 time 0.2823 (0.3317) loss 3.7682 (3.1974) grad_norm 2.3246 (2.4772) [2021-04-16 14:46:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][160/1251] eta 0:05:58 lr 0.000120 time 0.2827 (0.3282) loss 3.7807 (3.2179) grad_norm 2.2108 (2.4815) [2021-04-16 14:46:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][170/1251] eta 0:05:51 lr 0.000120 time 0.2888 (0.3252) loss 2.7693 (3.2200) grad_norm 2.8619 (2.4815) [2021-04-16 14:46:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][180/1251] eta 0:05:45 lr 0.000120 time 0.2793 (0.3224) loss 2.5939 (3.1888) grad_norm 2.6088 (2.4783) [2021-04-16 14:46:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][190/1251] eta 0:05:39 lr 0.000120 time 0.2494 (0.3198) loss 3.5925 (3.1941) grad_norm 2.4363 (2.4729) [2021-04-16 14:46:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][200/1251] eta 0:05:33 lr 0.000120 time 0.2868 (0.3175) loss 3.4530 (3.2045) grad_norm 2.4750 (2.4719) [2021-04-16 14:46:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][210/1251] eta 0:05:28 lr 0.000120 time 0.2675 (0.3155) loss 4.1654 (3.2017) grad_norm 2.2560 (2.4721) [2021-04-16 14:47:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][220/1251] eta 0:05:23 lr 0.000120 time 0.2875 (0.3137) loss 2.2415 (3.2047) grad_norm 2.4891 (2.4704) [2021-04-16 14:47:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][230/1251] eta 0:05:18 lr 0.000120 time 0.2871 (0.3122) loss 3.3115 (3.1969) grad_norm 2.2645 (2.4717) [2021-04-16 14:47:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][240/1251] eta 0:05:14 lr 0.000120 time 0.2913 (0.3108) loss 3.4853 (3.2061) grad_norm 2.5585 (2.4790) [2021-04-16 14:47:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][250/1251] eta 0:05:09 lr 0.000120 time 0.2751 (0.3093) loss 3.4410 (3.2102) grad_norm 2.0779 (2.4769) [2021-04-16 14:47:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][260/1251] eta 0:05:06 lr 0.000120 time 0.4272 (0.3092) loss 2.8242 (3.2100) grad_norm 2.3790 (2.4774) [2021-04-16 14:47:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][270/1251] eta 0:05:02 lr 0.000120 time 0.2698 (0.3079) loss 2.9058 (3.2077) grad_norm 2.4819 (2.4825) [2021-04-16 14:47:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][280/1251] eta 0:04:57 lr 0.000120 time 0.2835 (0.3068) loss 3.0443 (3.2154) grad_norm 2.4600 (2.4891) [2021-04-16 14:47:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][290/1251] eta 0:04:53 lr 0.000120 time 0.2991 (0.3059) loss 3.1099 (3.2037) grad_norm 2.5366 (2.4903) [2021-04-16 14:47:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][300/1251] eta 0:04:50 lr 0.000120 time 0.2820 (0.3050) loss 3.0044 (3.1970) grad_norm 2.5213 (2.4889) [2021-04-16 14:47:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][310/1251] eta 0:04:46 lr 0.000120 time 0.2677 (0.3041) loss 2.3068 (3.1877) grad_norm 2.6976 (2.4869) [2021-04-16 14:47:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][320/1251] eta 0:04:42 lr 0.000119 time 0.2639 (0.3038) loss 3.3177 (3.1864) grad_norm 2.2799 (2.4873) [2021-04-16 14:47:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][330/1251] eta 0:04:39 lr 0.000119 time 0.2694 (0.3030) loss 3.3792 (3.1915) grad_norm 2.6400 (2.4873) [2021-04-16 14:47:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][340/1251] eta 0:04:35 lr 0.000119 time 0.2881 (0.3022) loss 2.4960 (3.1764) grad_norm 2.6403 (2.4875) [2021-04-16 14:47:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][350/1251] eta 0:04:32 lr 0.000119 time 0.3010 (0.3020) loss 3.1998 (3.1766) grad_norm 2.2403 (2.4903) [2021-04-16 14:47:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][360/1251] eta 0:04:28 lr 0.000119 time 0.2737 (0.3017) loss 3.3473 (3.1816) grad_norm 2.0154 (2.4859) [2021-04-16 14:47:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][370/1251] eta 0:04:25 lr 0.000119 time 0.2487 (0.3011) loss 3.6254 (3.1816) grad_norm 2.1742 (2.4831) [2021-04-16 14:47:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][380/1251] eta 0:04:21 lr 0.000119 time 0.2870 (0.3005) loss 3.7123 (3.1853) grad_norm 2.3230 (2.4827) [2021-04-16 14:47:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][390/1251] eta 0:04:18 lr 0.000119 time 0.2747 (0.2998) loss 3.8504 (3.1814) grad_norm 2.1331 (2.4786) [2021-04-16 14:47:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][400/1251] eta 0:04:14 lr 0.000119 time 0.2462 (0.2992) loss 2.0883 (3.1838) grad_norm 4.5706 (2.4854) [2021-04-16 14:47:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][410/1251] eta 0:04:11 lr 0.000119 time 0.2586 (0.2987) loss 2.2369 (3.1802) grad_norm 2.3787 (2.4877) [2021-04-16 14:47:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][420/1251] eta 0:04:07 lr 0.000119 time 0.2715 (0.2982) loss 3.4567 (3.1861) grad_norm 2.1990 (2.4875) [2021-04-16 14:47:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][430/1251] eta 0:04:04 lr 0.000119 time 0.2913 (0.2977) loss 2.7168 (3.1823) grad_norm 2.6455 (2.4901) [2021-04-16 14:48:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][440/1251] eta 0:04:01 lr 0.000119 time 0.3025 (0.2973) loss 1.8388 (3.1779) grad_norm 2.4289 (2.4889) [2021-04-16 14:48:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][450/1251] eta 0:03:57 lr 0.000119 time 0.2938 (0.2968) loss 2.5460 (3.1715) grad_norm 2.6125 (2.4889) [2021-04-16 14:48:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][460/1251] eta 0:03:54 lr 0.000119 time 0.2759 (0.2963) loss 3.1332 (3.1717) grad_norm 2.4416 (2.4913) [2021-04-16 14:48:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][470/1251] eta 0:03:51 lr 0.000119 time 0.2952 (0.2958) loss 3.2463 (3.1721) grad_norm 2.2857 (2.4931) [2021-04-16 14:48:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][480/1251] eta 0:03:47 lr 0.000119 time 0.2673 (0.2953) loss 3.7478 (3.1760) grad_norm 2.5215 (2.4936) [2021-04-16 14:48:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][490/1251] eta 0:03:44 lr 0.000119 time 0.2789 (0.2949) loss 2.4227 (3.1783) grad_norm 2.4298 (2.4942) [2021-04-16 14:48:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][500/1251] eta 0:03:41 lr 0.000119 time 0.2889 (0.2945) loss 3.3733 (3.1728) grad_norm 2.5273 (2.4916) [2021-04-16 14:48:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][510/1251] eta 0:03:37 lr 0.000119 time 0.2863 (0.2942) loss 3.5586 (3.1744) grad_norm 2.2781 (2.4987) [2021-04-16 14:48:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][520/1251] eta 0:03:34 lr 0.000119 time 0.2730 (0.2939) loss 3.1721 (3.1715) grad_norm 2.4337 (2.5057) [2021-04-16 14:48:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][530/1251] eta 0:03:31 lr 0.000119 time 0.2883 (0.2935) loss 2.1268 (3.1683) grad_norm 2.7063 (2.5052) [2021-04-16 14:48:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][540/1251] eta 0:03:28 lr 0.000119 time 0.2761 (0.2932) loss 3.7625 (3.1701) grad_norm 2.5726 (2.5035) [2021-04-16 14:48:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][550/1251] eta 0:03:25 lr 0.000119 time 0.2424 (0.2933) loss 2.9975 (3.1715) grad_norm 2.3078 (2.5009) [2021-04-16 14:48:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][560/1251] eta 0:03:22 lr 0.000119 time 0.2800 (0.2930) loss 3.7309 (3.1698) grad_norm 2.3247 (2.5032) [2021-04-16 14:48:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][570/1251] eta 0:03:19 lr 0.000119 time 0.2797 (0.2930) loss 3.7649 (3.1690) grad_norm 2.3814 (2.5026) [2021-04-16 14:48:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][580/1251] eta 0:03:16 lr 0.000119 time 0.3068 (0.2933) loss 3.3682 (3.1704) grad_norm 2.5605 (2.5020) [2021-04-16 14:48:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][590/1251] eta 0:03:13 lr 0.000119 time 0.2657 (0.2932) loss 3.4877 (3.1660) grad_norm 2.4460 (2.5004) [2021-04-16 14:48:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][600/1251] eta 0:03:10 lr 0.000119 time 0.3018 (0.2929) loss 3.1635 (3.1612) grad_norm 2.3705 (2.4996) [2021-04-16 14:48:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][610/1251] eta 0:03:07 lr 0.000119 time 0.2717 (0.2927) loss 2.8769 (3.1639) grad_norm 2.6882 (2.4989) [2021-04-16 14:48:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][620/1251] eta 0:03:04 lr 0.000119 time 0.2801 (0.2925) loss 3.3630 (3.1647) grad_norm 2.3529 (2.4996) [2021-04-16 14:48:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][630/1251] eta 0:03:01 lr 0.000119 time 0.2857 (0.2924) loss 3.6767 (3.1665) grad_norm 2.9420 (2.5006) [2021-04-16 14:48:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][640/1251] eta 0:02:58 lr 0.000119 time 0.2738 (0.2921) loss 3.9309 (3.1684) grad_norm 2.5343 (2.4995) [2021-04-16 14:49:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][650/1251] eta 0:02:55 lr 0.000119 time 0.2511 (0.2920) loss 3.2948 (3.1698) grad_norm 2.7647 (2.5004) [2021-04-16 14:49:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][660/1251] eta 0:02:52 lr 0.000119 time 0.2790 (0.2918) loss 3.4140 (3.1692) grad_norm 3.2595 (2.5024) [2021-04-16 14:49:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][670/1251] eta 0:02:49 lr 0.000119 time 0.2804 (0.2917) loss 3.4650 (3.1730) grad_norm 2.1858 (2.5024) [2021-04-16 14:49:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][680/1251] eta 0:02:46 lr 0.000119 time 0.2604 (0.2914) loss 2.5289 (3.1722) grad_norm 2.1180 (2.5007) [2021-04-16 14:49:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][690/1251] eta 0:02:43 lr 0.000119 time 0.2709 (0.2913) loss 2.7159 (3.1742) grad_norm 2.4841 (2.4998) [2021-04-16 14:49:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][700/1251] eta 0:02:40 lr 0.000118 time 0.2779 (0.2910) loss 3.6107 (3.1760) grad_norm 2.5483 (2.4992) [2021-04-16 14:49:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][710/1251] eta 0:02:37 lr 0.000118 time 0.2506 (0.2908) loss 2.2845 (3.1762) grad_norm 2.2668 (2.4996) [2021-04-16 14:49:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][720/1251] eta 0:02:34 lr 0.000118 time 0.2456 (0.2910) loss 3.5571 (3.1773) grad_norm 3.0166 (2.5011) [2021-04-16 14:49:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][730/1251] eta 0:02:31 lr 0.000118 time 0.2713 (0.2907) loss 3.5512 (3.1785) grad_norm 2.6761 (2.5033) [2021-04-16 14:49:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][740/1251] eta 0:02:28 lr 0.000118 time 0.2685 (0.2905) loss 2.5731 (3.1761) grad_norm 2.5139 (2.5027) [2021-04-16 14:49:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][750/1251] eta 0:02:25 lr 0.000118 time 0.2817 (0.2904) loss 3.4650 (3.1775) grad_norm 3.1415 (2.5027) [2021-04-16 14:49:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][760/1251] eta 0:02:22 lr 0.000118 time 0.3034 (0.2903) loss 2.8204 (3.1754) grad_norm 2.3056 (2.5031) [2021-04-16 14:49:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][770/1251] eta 0:02:19 lr 0.000118 time 0.2744 (0.2901) loss 3.2551 (3.1748) grad_norm 2.5924 (2.5017) [2021-04-16 14:49:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][780/1251] eta 0:02:16 lr 0.000118 time 0.2776 (0.2900) loss 3.2711 (3.1765) grad_norm 2.2674 (2.5013) [2021-04-16 14:49:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][790/1251] eta 0:02:13 lr 0.000118 time 0.2839 (0.2898) loss 3.9356 (3.1753) grad_norm 2.4410 (2.5068) [2021-04-16 14:49:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][800/1251] eta 0:02:10 lr 0.000118 time 0.2660 (0.2896) loss 2.9565 (3.1738) grad_norm 2.6652 (2.5066) [2021-04-16 14:49:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][810/1251] eta 0:02:07 lr 0.000118 time 0.2628 (0.2894) loss 3.5652 (3.1762) grad_norm 2.8597 (2.5062) [2021-04-16 14:49:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][820/1251] eta 0:02:04 lr 0.000118 time 0.2642 (0.2891) loss 2.3639 (3.1732) grad_norm 2.8487 (2.5059) [2021-04-16 14:49:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][830/1251] eta 0:02:01 lr 0.000118 time 0.2863 (0.2890) loss 2.8958 (3.1744) grad_norm 2.3502 (2.5070) [2021-04-16 14:49:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][840/1251] eta 0:01:58 lr 0.000118 time 0.2752 (0.2891) loss 3.1595 (3.1760) grad_norm 2.5380 (2.5066) [2021-04-16 14:49:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][850/1251] eta 0:01:55 lr 0.000118 time 0.2685 (0.2890) loss 3.0266 (3.1763) grad_norm 2.2449 (2.5067) [2021-04-16 14:50:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][860/1251] eta 0:01:52 lr 0.000118 time 0.2818 (0.2888) loss 2.4006 (3.1768) grad_norm 2.3024 (2.5105) [2021-04-16 14:50:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][870/1251] eta 0:01:49 lr 0.000118 time 0.2804 (0.2887) loss 3.4340 (3.1770) grad_norm 2.3643 (2.5121) [2021-04-16 14:50:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][880/1251] eta 0:01:47 lr 0.000118 time 0.2575 (0.2886) loss 3.8940 (3.1779) grad_norm 2.4112 (2.5130) [2021-04-16 14:50:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][890/1251] eta 0:01:44 lr 0.000118 time 0.2683 (0.2885) loss 2.5156 (3.1770) grad_norm 2.1344 (2.5140) [2021-04-16 14:50:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][900/1251] eta 0:01:41 lr 0.000118 time 0.2707 (0.2884) loss 2.9269 (3.1758) grad_norm 2.6418 (2.5165) [2021-04-16 14:50:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][910/1251] eta 0:01:38 lr 0.000118 time 0.2599 (0.2882) loss 2.6767 (3.1748) grad_norm 2.6207 (2.5147) [2021-04-16 14:50:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][920/1251] eta 0:01:35 lr 0.000118 time 0.2662 (0.2881) loss 4.0153 (3.1734) grad_norm 2.3981 (2.5154) [2021-04-16 14:50:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][930/1251] eta 0:01:32 lr 0.000118 time 0.2564 (0.2882) loss 3.6139 (3.1743) grad_norm 3.9283 (2.5164) [2021-04-16 14:50:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][940/1251] eta 0:01:29 lr 0.000118 time 0.2760 (0.2882) loss 3.2488 (3.1775) grad_norm 2.9419 (2.5165) [2021-04-16 14:50:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][950/1251] eta 0:01:26 lr 0.000118 time 0.2681 (0.2880) loss 2.8100 (3.1767) grad_norm 2.9580 (2.5160) [2021-04-16 14:50:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][960/1251] eta 0:01:23 lr 0.000118 time 0.2963 (0.2879) loss 3.5267 (3.1792) grad_norm 2.4649 (2.5142) [2021-04-16 14:50:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][970/1251] eta 0:01:20 lr 0.000118 time 0.2554 (0.2877) loss 3.6664 (3.1798) grad_norm 2.2863 (2.5160) [2021-04-16 14:50:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][980/1251] eta 0:01:17 lr 0.000118 time 0.2455 (0.2876) loss 3.3792 (3.1811) grad_norm 2.3627 (2.5149) [2021-04-16 14:50:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][990/1251] eta 0:01:15 lr 0.000118 time 0.2592 (0.2875) loss 3.6037 (3.1838) grad_norm 2.7596 (2.5157) [2021-04-16 14:50:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][1000/1251] eta 0:01:12 lr 0.000118 time 0.2567 (0.2873) loss 2.5426 (3.1826) grad_norm 2.3228 (2.5158) [2021-04-16 14:50:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][1010/1251] eta 0:01:09 lr 0.000118 time 0.2665 (0.2872) loss 3.2973 (3.1820) grad_norm 2.5428 (2.5166) [2021-04-16 14:50:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][1020/1251] eta 0:01:06 lr 0.000118 time 0.2680 (0.2871) loss 3.3849 (3.1817) grad_norm 2.1391 (2.5170) [2021-04-16 14:50:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][1030/1251] eta 0:01:03 lr 0.000118 time 0.3144 (0.2870) loss 3.3804 (3.1837) grad_norm 2.6478 (2.5170) [2021-04-16 14:50:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][1040/1251] eta 0:01:00 lr 0.000118 time 0.2654 (0.2869) loss 3.4389 (3.1820) grad_norm 2.3965 (2.5179) [2021-04-16 14:50:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][1050/1251] eta 0:00:57 lr 0.000118 time 0.2747 (0.2868) loss 2.4374 (3.1802) grad_norm 2.4817 (2.5176) [2021-04-16 14:50:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][1060/1251] eta 0:00:54 lr 0.000118 time 0.2640 (0.2867) loss 3.3148 (3.1792) grad_norm 2.5991 (2.5159) [2021-04-16 14:50:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][1070/1251] eta 0:00:51 lr 0.000118 time 0.2726 (0.2866) loss 2.5717 (3.1773) grad_norm 2.4053 (2.5169) [2021-04-16 14:51:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][1080/1251] eta 0:00:48 lr 0.000118 time 0.2986 (0.2865) loss 2.8872 (3.1786) grad_norm 2.6404 (2.5165) [2021-04-16 14:51:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][1090/1251] eta 0:00:46 lr 0.000117 time 0.2869 (0.2863) loss 2.8722 (3.1762) grad_norm 2.4012 (2.5167) [2021-04-16 14:51:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][1100/1251] eta 0:00:43 lr 0.000117 time 0.2939 (0.2862) loss 2.7460 (3.1702) grad_norm 2.3477 (2.5170) [2021-04-16 14:51:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][1110/1251] eta 0:00:40 lr 0.000117 time 0.2943 (0.2862) loss 2.8369 (3.1709) grad_norm 3.2769 (2.5174) [2021-04-16 14:51:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][1120/1251] eta 0:00:37 lr 0.000117 time 0.2577 (0.2862) loss 3.6355 (3.1703) grad_norm 2.5372 (2.5177) [2021-04-16 14:51:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][1130/1251] eta 0:00:34 lr 0.000117 time 0.2627 (0.2862) loss 3.4461 (3.1709) grad_norm 2.8264 (2.5184) [2021-04-16 14:51:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][1140/1251] eta 0:00:31 lr 0.000117 time 0.2630 (0.2862) loss 3.4387 (3.1722) grad_norm 2.4215 (2.5189) [2021-04-16 14:51:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][1150/1251] eta 0:00:28 lr 0.000117 time 0.2913 (0.2861) loss 2.2241 (3.1718) grad_norm 2.5261 (2.5179) [2021-04-16 14:51:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][1160/1251] eta 0:00:26 lr 0.000117 time 0.2601 (0.2862) loss 3.9830 (3.1727) grad_norm 2.5411 (2.5182) [2021-04-16 14:51:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][1170/1251] eta 0:00:23 lr 0.000117 time 0.2489 (0.2862) loss 3.1248 (3.1722) grad_norm 2.3196 (2.5181) [2021-04-16 14:51:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][1180/1251] eta 0:00:20 lr 0.000117 time 0.2781 (0.2861) loss 3.2301 (3.1722) grad_norm 2.0682 (2.5172) [2021-04-16 14:51:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][1190/1251] eta 0:00:17 lr 0.000117 time 0.2718 (0.2860) loss 3.3999 (3.1699) grad_norm 2.5842 (2.5180) [2021-04-16 14:51:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][1200/1251] eta 0:00:14 lr 0.000117 time 0.2754 (0.2859) loss 3.6925 (3.1722) grad_norm 2.2647 (2.5183) [2021-04-16 14:51:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][1210/1251] eta 0:00:11 lr 0.000117 time 0.2689 (0.2860) loss 3.1822 (3.1717) grad_norm 2.3699 (2.5178) [2021-04-16 14:51:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][1220/1251] eta 0:00:08 lr 0.000117 time 0.2786 (0.2859) loss 4.0727 (3.1720) grad_norm 2.5450 (2.5172) [2021-04-16 14:51:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][1230/1251] eta 0:00:06 lr 0.000117 time 0.2753 (0.2859) loss 3.7375 (3.1716) grad_norm 2.9542 (2.5173) [2021-04-16 14:51:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][1240/1251] eta 0:00:03 lr 0.000117 time 0.3166 (0.2858) loss 3.8121 (3.1741) grad_norm 2.3737 (2.5170) [2021-04-16 14:51:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [235/300][1250/1251] eta 0:00:00 lr 0.000117 time 0.2481 (0.2855) loss 3.4633 (3.1738) grad_norm 2.6984 (2.5189) [2021-04-16 14:52:10 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 235 training takes 0:06:18 [2021-04-16 14:52:10 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_235.pth saving...... [2021-04-16 14:52:31 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_235.pth saved !!! [2021-04-16 14:52:32 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.119 (1.119) Loss 0.7788 (0.7788) Acc@1 81.836 (81.836) Acc@5 95.898 (95.898) [2021-04-16 14:52:34 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.186 (0.204) Loss 0.9114 (0.8517) Acc@1 78.613 (79.945) Acc@5 94.629 (95.117) [2021-04-16 14:52:36 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.082 (0.236) Loss 0.7506 (0.8463) Acc@1 82.812 (80.222) Acc@5 96.387 (95.159) [2021-04-16 14:52:39 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.093 (0.236) Loss 0.9132 (0.8427) Acc@1 79.980 (80.365) Acc@5 94.824 (95.209) [2021-04-16 14:52:40 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.218) Loss 0.8919 (0.8528) Acc@1 79.004 (80.111) Acc@5 94.141 (95.081) [2021-04-16 14:52:57 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 79.942 Acc@5 94.978 [2021-04-16 14:52:57 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 79.9% [2021-04-16 14:52:57 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 79.94% [2021-04-16 14:53:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][0/1251] eta 2:37:43 lr 0.000117 time 7.5648 (7.5648) loss 2.8986 (2.8986) grad_norm 2.4396 (2.4396) [2021-04-16 14:53:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][10/1251] eta 0:19:20 lr 0.000117 time 0.2830 (0.9354) loss 3.8306 (3.2165) grad_norm 3.1722 (2.5824) [2021-04-16 14:53:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][20/1251] eta 0:12:46 lr 0.000117 time 0.2576 (0.6223) loss 2.5123 (3.1999) grad_norm 2.2904 (2.5715) [2021-04-16 14:53:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][30/1251] eta 0:10:24 lr 0.000117 time 0.2851 (0.5114) loss 2.6033 (3.1732) grad_norm 2.5716 (2.5580) [2021-04-16 14:53:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][40/1251] eta 0:09:14 lr 0.000117 time 0.2933 (0.4576) loss 3.4018 (3.1984) grad_norm 2.4324 (2.5796) [2021-04-16 14:53:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][50/1251] eta 0:08:27 lr 0.000117 time 0.2800 (0.4226) loss 2.6529 (3.2251) grad_norm 2.5346 (2.6146) [2021-04-16 14:53:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][60/1251] eta 0:07:55 lr 0.000117 time 0.2789 (0.3990) loss 2.8440 (3.2043) grad_norm 2.6364 (2.5948) [2021-04-16 14:53:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][70/1251] eta 0:07:30 lr 0.000117 time 0.2649 (0.3813) loss 3.3921 (3.1712) grad_norm 2.6196 (2.5817) [2021-04-16 14:53:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][80/1251] eta 0:07:10 lr 0.000117 time 0.2817 (0.3680) loss 3.8443 (3.1855) grad_norm 2.9044 (2.5576) [2021-04-16 14:53:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][90/1251] eta 0:06:56 lr 0.000117 time 0.2995 (0.3592) loss 3.3671 (3.1733) grad_norm 2.7322 (2.5638) [2021-04-16 14:53:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][100/1251] eta 0:06:44 lr 0.000117 time 0.3192 (0.3517) loss 2.2162 (3.1374) grad_norm 2.1501 (2.5587) [2021-04-16 14:53:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][110/1251] eta 0:06:33 lr 0.000117 time 0.2566 (0.3451) loss 3.2308 (3.1438) grad_norm 2.1034 (2.5418) [2021-04-16 14:53:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][120/1251] eta 0:06:23 lr 0.000117 time 0.2753 (0.3394) loss 2.3537 (3.1414) grad_norm 2.2490 (2.5422) [2021-04-16 14:53:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][130/1251] eta 0:06:16 lr 0.000117 time 0.2876 (0.3361) loss 3.2682 (3.1316) grad_norm 2.4797 (2.5402) [2021-04-16 14:53:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][140/1251] eta 0:06:09 lr 0.000117 time 0.2601 (0.3326) loss 3.0714 (3.1362) grad_norm 2.2010 (2.5331) [2021-04-16 14:53:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][150/1251] eta 0:06:03 lr 0.000117 time 0.2824 (0.3297) loss 3.2533 (3.1326) grad_norm 2.1749 (2.5293) [2021-04-16 14:53:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][160/1251] eta 0:05:56 lr 0.000117 time 0.2643 (0.3270) loss 3.3382 (3.1537) grad_norm 2.6047 (2.5230) [2021-04-16 14:53:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][170/1251] eta 0:05:50 lr 0.000117 time 0.2983 (0.3241) loss 3.1306 (3.1428) grad_norm 2.1521 (2.5138) [2021-04-16 14:53:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][180/1251] eta 0:05:44 lr 0.000117 time 0.2829 (0.3220) loss 2.4588 (3.1376) grad_norm 2.2539 (2.5135) [2021-04-16 14:53:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][190/1251] eta 0:05:39 lr 0.000117 time 0.2999 (0.3196) loss 3.2251 (3.1372) grad_norm 2.4582 (2.5098) [2021-04-16 14:54:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][200/1251] eta 0:05:33 lr 0.000117 time 0.2903 (0.3174) loss 2.8092 (3.1499) grad_norm 2.4017 (2.5118) [2021-04-16 14:54:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][210/1251] eta 0:05:28 lr 0.000117 time 0.2866 (0.3153) loss 4.2143 (3.1566) grad_norm 2.5233 (2.5115) [2021-04-16 14:54:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][220/1251] eta 0:05:23 lr 0.000117 time 0.2744 (0.3135) loss 2.2760 (3.1483) grad_norm 2.4072 (2.5102) [2021-04-16 14:54:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][230/1251] eta 0:05:18 lr 0.000116 time 0.2898 (0.3121) loss 3.2475 (3.1582) grad_norm 2.5225 (2.5115) [2021-04-16 14:54:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][240/1251] eta 0:05:14 lr 0.000116 time 0.2497 (0.3106) loss 2.6783 (3.1620) grad_norm 2.3010 (2.5071) [2021-04-16 14:54:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][250/1251] eta 0:05:09 lr 0.000116 time 0.2706 (0.3092) loss 3.9111 (3.1695) grad_norm 2.4384 (2.5078) [2021-04-16 14:54:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][260/1251] eta 0:05:05 lr 0.000116 time 0.2630 (0.3079) loss 3.3540 (3.1761) grad_norm 2.4218 (2.5089) [2021-04-16 14:54:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][270/1251] eta 0:05:00 lr 0.000116 time 0.2749 (0.3067) loss 3.6526 (3.1873) grad_norm 2.6788 (2.5040) [2021-04-16 14:54:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][280/1251] eta 0:04:56 lr 0.000116 time 0.2683 (0.3056) loss 2.3870 (3.1826) grad_norm 2.3334 (2.5051) [2021-04-16 14:54:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][290/1251] eta 0:04:52 lr 0.000116 time 0.2815 (0.3046) loss 3.4572 (3.1714) grad_norm 2.2220 (2.5045) [2021-04-16 14:54:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][300/1251] eta 0:04:48 lr 0.000116 time 0.2714 (0.3036) loss 3.8121 (3.1719) grad_norm 2.4111 (2.5089) [2021-04-16 14:54:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][310/1251] eta 0:04:44 lr 0.000116 time 0.2812 (0.3027) loss 3.1289 (3.1614) grad_norm 3.0344 (2.5157) [2021-04-16 14:54:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][320/1251] eta 0:04:41 lr 0.000116 time 0.2781 (0.3019) loss 3.3565 (3.1625) grad_norm 2.6284 (2.5132) [2021-04-16 14:54:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][330/1251] eta 0:04:37 lr 0.000116 time 0.2749 (0.3012) loss 3.2197 (3.1558) grad_norm 2.4762 (2.5197) [2021-04-16 14:54:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][340/1251] eta 0:04:34 lr 0.000116 time 0.2716 (0.3008) loss 2.8210 (3.1577) grad_norm 2.3360 (2.5188) [2021-04-16 14:54:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][350/1251] eta 0:04:30 lr 0.000116 time 0.2728 (0.3008) loss 3.2615 (3.1569) grad_norm 2.3759 (2.5183) [2021-04-16 14:54:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][360/1251] eta 0:04:27 lr 0.000116 time 0.2764 (0.3006) loss 1.9186 (3.1524) grad_norm 2.3623 (2.5145) [2021-04-16 14:54:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][370/1251] eta 0:04:24 lr 0.000116 time 0.2625 (0.3002) loss 2.7369 (3.1515) grad_norm 2.4731 (2.5177) [2021-04-16 14:54:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][380/1251] eta 0:04:20 lr 0.000116 time 0.2694 (0.2996) loss 3.9700 (3.1535) grad_norm 2.3611 (2.5224) [2021-04-16 14:54:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][390/1251] eta 0:04:17 lr 0.000116 time 0.2980 (0.2990) loss 3.3507 (3.1494) grad_norm 2.7688 (2.5256) [2021-04-16 14:54:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][400/1251] eta 0:04:14 lr 0.000116 time 0.2639 (0.2988) loss 3.1692 (3.1533) grad_norm 3.3815 (2.5260) [2021-04-16 14:54:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][410/1251] eta 0:04:10 lr 0.000116 time 0.2662 (0.2983) loss 3.5273 (3.1510) grad_norm 3.0148 (2.5242) [2021-04-16 14:55:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][420/1251] eta 0:04:07 lr 0.000116 time 0.2772 (0.2977) loss 4.0723 (3.1511) grad_norm 2.3411 (2.5250) [2021-04-16 14:55:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][430/1251] eta 0:04:04 lr 0.000116 time 0.2617 (0.2973) loss 3.3356 (3.1467) grad_norm 2.2125 (2.5255) [2021-04-16 14:55:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][440/1251] eta 0:04:00 lr 0.000116 time 0.2818 (0.2968) loss 2.8181 (3.1407) grad_norm 2.5252 (2.5254) [2021-04-16 14:55:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][450/1251] eta 0:03:57 lr 0.000116 time 0.3968 (0.2968) loss 2.9126 (3.1408) grad_norm 2.0235 (2.5248) [2021-04-16 14:55:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][460/1251] eta 0:03:54 lr 0.000116 time 0.2453 (0.2963) loss 3.5289 (3.1451) grad_norm 2.3525 (2.5237) [2021-04-16 14:55:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][470/1251] eta 0:03:51 lr 0.000116 time 0.2759 (0.2959) loss 3.1198 (3.1421) grad_norm 2.5737 (2.5248) [2021-04-16 14:55:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][480/1251] eta 0:03:47 lr 0.000116 time 0.2924 (0.2954) loss 3.6299 (3.1399) grad_norm 2.9116 (2.5294) [2021-04-16 14:55:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][490/1251] eta 0:03:44 lr 0.000116 time 0.2858 (0.2950) loss 3.8555 (3.1446) grad_norm 2.1976 (2.5303) [2021-04-16 14:55:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][500/1251] eta 0:03:41 lr 0.000116 time 0.2802 (0.2946) loss 3.5796 (3.1393) grad_norm 2.1867 (2.5310) [2021-04-16 14:55:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][510/1251] eta 0:03:38 lr 0.000116 time 0.2757 (0.2942) loss 2.2849 (3.1416) grad_norm 2.4617 (2.5368) [2021-04-16 14:55:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][520/1251] eta 0:03:34 lr 0.000116 time 0.2654 (0.2940) loss 3.0484 (3.1420) grad_norm 2.5559 (2.5358) [2021-04-16 14:55:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][530/1251] eta 0:03:31 lr 0.000116 time 0.2955 (0.2938) loss 3.7114 (3.1496) grad_norm 2.2699 (2.5390) [2021-04-16 14:55:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][540/1251] eta 0:03:28 lr 0.000116 time 0.2783 (0.2934) loss 3.6743 (3.1497) grad_norm 2.2746 (2.5366) [2021-04-16 14:55:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][550/1251] eta 0:03:25 lr 0.000116 time 0.2795 (0.2932) loss 3.2405 (3.1492) grad_norm 2.2760 (2.5337) [2021-04-16 14:55:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][560/1251] eta 0:03:22 lr 0.000116 time 0.2797 (0.2929) loss 1.8319 (3.1510) grad_norm 2.5265 (2.5340) [2021-04-16 14:55:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][570/1251] eta 0:03:19 lr 0.000116 time 0.2485 (0.2927) loss 3.8089 (3.1581) grad_norm 2.5132 (2.5325) [2021-04-16 14:55:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][580/1251] eta 0:03:16 lr 0.000116 time 0.2674 (0.2927) loss 3.8726 (3.1554) grad_norm 2.8838 (2.5307) [2021-04-16 14:55:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][590/1251] eta 0:03:13 lr 0.000116 time 0.3102 (0.2925) loss 3.8064 (3.1538) grad_norm 3.6778 (2.5324) [2021-04-16 14:55:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][600/1251] eta 0:03:10 lr 0.000116 time 0.2872 (0.2922) loss 3.5054 (3.1537) grad_norm 2.2830 (2.5316) [2021-04-16 14:55:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][610/1251] eta 0:03:07 lr 0.000116 time 0.2793 (0.2918) loss 3.0781 (3.1526) grad_norm 2.3917 (2.5291) [2021-04-16 14:55:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][620/1251] eta 0:03:04 lr 0.000115 time 0.2436 (0.2917) loss 2.4721 (3.1531) grad_norm 2.2635 (2.5297) [2021-04-16 14:56:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][630/1251] eta 0:03:01 lr 0.000115 time 0.2647 (0.2915) loss 2.7908 (3.1535) grad_norm 2.0924 (2.5279) [2021-04-16 14:56:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][640/1251] eta 0:02:57 lr 0.000115 time 0.2695 (0.2912) loss 2.7593 (3.1571) grad_norm 2.5383 (2.5287) [2021-04-16 14:56:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][650/1251] eta 0:02:54 lr 0.000115 time 0.2886 (0.2910) loss 3.4955 (3.1577) grad_norm 2.1996 (2.5271) [2021-04-16 14:56:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][660/1251] eta 0:02:51 lr 0.000115 time 0.2936 (0.2908) loss 3.6939 (3.1630) grad_norm 2.5576 (2.5273) [2021-04-16 14:56:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][670/1251] eta 0:02:48 lr 0.000115 time 0.2709 (0.2905) loss 3.2530 (3.1608) grad_norm 2.4968 (2.5274) [2021-04-16 14:56:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][680/1251] eta 0:02:45 lr 0.000115 time 0.2600 (0.2903) loss 3.4997 (3.1581) grad_norm 2.7815 (2.5287) [2021-04-16 14:56:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][690/1251] eta 0:02:42 lr 0.000115 time 0.3012 (0.2901) loss 3.7038 (3.1593) grad_norm 2.8028 (2.5321) [2021-04-16 14:56:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][700/1251] eta 0:02:39 lr 0.000115 time 0.2773 (0.2898) loss 3.7471 (3.1615) grad_norm 2.9718 (2.5320) [2021-04-16 14:56:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][710/1251] eta 0:02:36 lr 0.000115 time 0.3030 (0.2898) loss 2.9795 (3.1602) grad_norm 2.5601 (2.5339) [2021-04-16 14:56:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][720/1251] eta 0:02:33 lr 0.000115 time 0.2805 (0.2899) loss 3.1296 (3.1579) grad_norm 2.4968 (2.5352) [2021-04-16 14:56:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][730/1251] eta 0:02:31 lr 0.000115 time 0.3626 (0.2900) loss 3.5416 (3.1614) grad_norm 2.7337 (2.5368) [2021-04-16 14:56:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][740/1251] eta 0:02:28 lr 0.000115 time 0.2804 (0.2899) loss 2.2345 (3.1609) grad_norm 2.6765 (2.5399) [2021-04-16 14:56:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][750/1251] eta 0:02:25 lr 0.000115 time 0.2484 (0.2897) loss 3.4689 (3.1577) grad_norm 2.6994 (2.5377) [2021-04-16 14:56:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][760/1251] eta 0:02:22 lr 0.000115 time 0.2962 (0.2896) loss 2.1920 (3.1561) grad_norm 2.0425 (2.5376) [2021-04-16 14:56:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][770/1251] eta 0:02:19 lr 0.000115 time 0.2915 (0.2894) loss 3.2067 (3.1550) grad_norm 2.7234 (2.5366) [2021-04-16 14:56:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][780/1251] eta 0:02:16 lr 0.000115 time 0.3066 (0.2892) loss 3.4654 (3.1565) grad_norm 2.3848 (2.5399) [2021-04-16 14:56:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][790/1251] eta 0:02:13 lr 0.000115 time 0.2941 (0.2890) loss 3.1788 (3.1569) grad_norm 2.4972 (2.5395) [2021-04-16 14:56:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][800/1251] eta 0:02:10 lr 0.000115 time 0.2996 (0.2889) loss 3.3979 (3.1560) grad_norm 2.4313 (2.5384) [2021-04-16 14:56:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][810/1251] eta 0:02:07 lr 0.000115 time 0.2667 (0.2887) loss 3.2176 (3.1539) grad_norm 2.3202 (2.5378) [2021-04-16 14:56:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][820/1251] eta 0:02:04 lr 0.000115 time 0.2620 (0.2885) loss 2.1302 (3.1523) grad_norm 3.5812 (2.5400) [2021-04-16 14:56:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][830/1251] eta 0:02:01 lr 0.000115 time 0.2676 (0.2884) loss 2.6748 (3.1500) grad_norm 3.3222 (2.5435) [2021-04-16 14:56:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][840/1251] eta 0:01:58 lr 0.000115 time 0.2807 (0.2882) loss 3.0140 (3.1502) grad_norm 2.4790 (2.5457) [2021-04-16 14:57:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][850/1251] eta 0:01:55 lr 0.000115 time 0.2791 (0.2881) loss 2.9758 (3.1470) grad_norm 2.6222 (2.5469) [2021-04-16 14:57:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][860/1251] eta 0:01:52 lr 0.000115 time 0.2671 (0.2880) loss 3.4914 (3.1475) grad_norm 3.8649 (2.5496) [2021-04-16 14:57:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][870/1251] eta 0:01:49 lr 0.000115 time 0.2705 (0.2878) loss 2.6040 (3.1457) grad_norm 2.3770 (2.5496) [2021-04-16 14:57:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][880/1251] eta 0:01:46 lr 0.000115 time 0.2691 (0.2877) loss 3.4679 (3.1480) grad_norm 2.3771 (2.5483) [2021-04-16 14:57:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][890/1251] eta 0:01:43 lr 0.000115 time 0.2686 (0.2876) loss 2.4319 (3.1451) grad_norm 2.5259 (2.5488) [2021-04-16 14:57:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][900/1251] eta 0:01:40 lr 0.000115 time 0.2652 (0.2874) loss 3.2293 (3.1439) grad_norm 2.5706 (2.5482) [2021-04-16 14:57:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][910/1251] eta 0:01:37 lr 0.000115 time 0.2987 (0.2873) loss 3.1952 (3.1431) grad_norm 2.3306 (2.5466) [2021-04-16 14:57:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][920/1251] eta 0:01:35 lr 0.000115 time 0.2596 (0.2873) loss 3.0716 (3.1442) grad_norm 2.1943 (2.5466) [2021-04-16 14:57:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][930/1251] eta 0:01:32 lr 0.000115 time 0.2513 (0.2874) loss 3.4797 (3.1443) grad_norm 2.3838 (2.5457) [2021-04-16 14:57:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][940/1251] eta 0:01:29 lr 0.000115 time 0.2858 (0.2874) loss 3.1479 (3.1435) grad_norm 2.3925 (2.5441) [2021-04-16 14:57:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][950/1251] eta 0:01:26 lr 0.000115 time 0.2758 (0.2873) loss 3.0604 (3.1433) grad_norm 2.3703 (2.5420) [2021-04-16 14:57:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][960/1251] eta 0:01:23 lr 0.000115 time 0.2854 (0.2872) loss 2.3087 (3.1430) grad_norm 2.6155 (2.5432) [2021-04-16 14:57:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][970/1251] eta 0:01:20 lr 0.000115 time 0.2945 (0.2871) loss 2.0434 (3.1456) grad_norm 2.5120 (2.5446) [2021-04-16 14:57:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][980/1251] eta 0:01:17 lr 0.000115 time 0.2798 (0.2871) loss 3.0481 (3.1430) grad_norm 2.2652 (2.5440) [2021-04-16 14:57:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][990/1251] eta 0:01:14 lr 0.000115 time 0.2513 (0.2870) loss 3.9467 (3.1437) grad_norm 2.8163 (2.5429) [2021-04-16 14:57:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][1000/1251] eta 0:01:12 lr 0.000115 time 0.2914 (0.2869) loss 3.8074 (3.1466) grad_norm 2.5987 (2.5425) [2021-04-16 14:57:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][1010/1251] eta 0:01:09 lr 0.000114 time 0.2743 (0.2869) loss 3.4554 (3.1467) grad_norm 2.3232 (2.5405) [2021-04-16 14:57:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][1020/1251] eta 0:01:06 lr 0.000114 time 0.2645 (0.2867) loss 3.6827 (3.1447) grad_norm 2.3021 (2.5430) [2021-04-16 14:57:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][1030/1251] eta 0:01:03 lr 0.000114 time 0.2535 (0.2867) loss 3.5558 (3.1441) grad_norm 2.2440 (2.5441) [2021-04-16 14:57:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][1040/1251] eta 0:01:00 lr 0.000114 time 0.2603 (0.2866) loss 2.0518 (3.1421) grad_norm 2.3981 (2.5444) [2021-04-16 14:57:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][1050/1251] eta 0:00:57 lr 0.000114 time 0.2886 (0.2865) loss 3.1142 (3.1431) grad_norm 2.7736 (2.5443) [2021-04-16 14:58:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][1060/1251] eta 0:00:54 lr 0.000114 time 0.2759 (0.2864) loss 3.5069 (3.1421) grad_norm 2.2651 (2.5448) [2021-04-16 14:58:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][1070/1251] eta 0:00:51 lr 0.000114 time 0.2869 (0.2863) loss 2.6974 (3.1439) grad_norm 2.0991 (2.5444) [2021-04-16 14:58:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][1080/1251] eta 0:00:48 lr 0.000114 time 0.2752 (0.2862) loss 3.3617 (3.1442) grad_norm 2.6624 (2.5427) [2021-04-16 14:58:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][1090/1251] eta 0:00:46 lr 0.000114 time 0.2981 (0.2861) loss 2.0373 (3.1414) grad_norm 2.3105 (2.5419) [2021-04-16 14:58:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][1100/1251] eta 0:00:43 lr 0.000114 time 0.2658 (0.2861) loss 2.1305 (3.1415) grad_norm 2.2786 (2.5433) [2021-04-16 14:58:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][1110/1251] eta 0:00:40 lr 0.000114 time 0.2809 (0.2860) loss 3.4031 (3.1410) grad_norm 2.3041 (2.5447) [2021-04-16 14:58:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][1120/1251] eta 0:00:37 lr 0.000114 time 0.2770 (0.2859) loss 3.5225 (3.1416) grad_norm 2.2858 (2.5473) [2021-04-16 14:58:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][1130/1251] eta 0:00:34 lr 0.000114 time 0.2878 (0.2858) loss 3.6122 (3.1420) grad_norm 2.2995 (2.5474) [2021-04-16 14:58:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][1140/1251] eta 0:00:31 lr 0.000114 time 0.2643 (0.2859) loss 2.9960 (3.1397) grad_norm 2.5143 (2.5472) [2021-04-16 14:58:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][1150/1251] eta 0:00:28 lr 0.000114 time 0.4195 (0.2859) loss 2.7897 (3.1390) grad_norm 2.2660 (2.5471) [2021-04-16 14:58:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][1160/1251] eta 0:00:26 lr 0.000114 time 0.2717 (0.2858) loss 2.5332 (3.1380) grad_norm 2.3428 (2.5478) [2021-04-16 14:58:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][1170/1251] eta 0:00:23 lr 0.000114 time 0.2786 (0.2857) loss 2.8549 (3.1364) grad_norm 2.9059 (2.5478) [2021-04-16 14:58:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][1180/1251] eta 0:00:20 lr 0.000114 time 0.2799 (0.2856) loss 2.0154 (3.1330) grad_norm 2.3772 (2.5478) [2021-04-16 14:58:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][1190/1251] eta 0:00:17 lr 0.000114 time 0.2775 (0.2856) loss 2.3874 (3.1328) grad_norm 2.1039 (inf) [2021-04-16 14:58:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][1200/1251] eta 0:00:14 lr 0.000114 time 0.2797 (0.2855) loss 3.0943 (3.1350) grad_norm 2.6797 (inf) [2021-04-16 14:58:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][1210/1251] eta 0:00:11 lr 0.000114 time 0.2729 (0.2854) loss 3.6368 (3.1344) grad_norm 2.3159 (inf) [2021-04-16 14:58:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][1220/1251] eta 0:00:08 lr 0.000114 time 0.2696 (0.2853) loss 3.5337 (3.1348) grad_norm 2.3894 (inf) [2021-04-16 14:58:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][1230/1251] eta 0:00:05 lr 0.000114 time 0.2815 (0.2852) loss 3.8329 (3.1353) grad_norm 2.4121 (inf) [2021-04-16 14:58:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][1240/1251] eta 0:00:03 lr 0.000114 time 0.2478 (0.2850) loss 2.0309 (3.1353) grad_norm 3.4798 (inf) [2021-04-16 14:58:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [236/300][1250/1251] eta 0:00:00 lr 0.000114 time 0.2480 (0.2847) loss 3.1267 (3.1352) grad_norm 2.6513 (inf) [2021-04-16 14:59:10 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 236 training takes 0:06:13 [2021-04-16 14:59:10 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_236.pth saving...... [2021-04-16 14:59:25 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_236.pth saved !!! [2021-04-16 14:59:26 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.214 (1.214) Loss 0.8544 (0.8544) Acc@1 78.516 (78.516) Acc@5 95.410 (95.410) [2021-04-16 14:59:27 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.306 (0.261) Loss 0.8293 (0.8589) Acc@1 80.469 (79.892) Acc@5 95.020 (95.117) [2021-04-16 14:59:29 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.081 (0.235) Loss 0.9186 (0.8758) Acc@1 78.516 (79.348) Acc@5 94.922 (94.936) [2021-04-16 14:59:32 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.100 (0.231) Loss 0.8684 (0.8627) Acc@1 78.906 (79.751) Acc@5 95.020 (94.972) [2021-04-16 14:59:34 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.233 (0.228) Loss 0.8913 (0.8595) Acc@1 77.832 (79.838) Acc@5 95.215 (95.000) [2021-04-16 14:59:45 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 79.958 Acc@5 95.002 [2021-04-16 14:59:45 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 80.0% [2021-04-16 14:59:45 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 79.96% [2021-04-16 14:59:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][0/1251] eta 4:20:11 lr 0.000114 time 12.4795 (12.4795) loss 3.1219 (3.1219) grad_norm 2.5469 (2.5469) [2021-04-16 15:00:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][10/1251] eta 0:28:39 lr 0.000114 time 0.2808 (1.3854) loss 3.3428 (3.2430) grad_norm 2.4603 (2.4487) [2021-04-16 15:00:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][20/1251] eta 0:17:36 lr 0.000114 time 0.2950 (0.8583) loss 2.2049 (3.2053) grad_norm 2.2549 (2.5053) [2021-04-16 15:00:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][30/1251] eta 0:13:40 lr 0.000114 time 0.2983 (0.6720) loss 2.8894 (3.2079) grad_norm 2.8195 (2.6119) [2021-04-16 15:00:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][40/1251] eta 0:11:36 lr 0.000114 time 0.2991 (0.5749) loss 2.1203 (3.1592) grad_norm 2.5908 (2.5726) [2021-04-16 15:00:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][50/1251] eta 0:10:19 lr 0.000114 time 0.2774 (0.5160) loss 2.9363 (3.1165) grad_norm 3.7580 (2.5857) [2021-04-16 15:00:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][60/1251] eta 0:09:28 lr 0.000114 time 0.2799 (0.4773) loss 3.4627 (3.0703) grad_norm 2.9570 (2.5982) [2021-04-16 15:00:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][70/1251] eta 0:08:51 lr 0.000114 time 0.2890 (0.4496) loss 2.9895 (3.0739) grad_norm 2.4631 (2.5888) [2021-04-16 15:00:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][80/1251] eta 0:08:21 lr 0.000114 time 0.2898 (0.4282) loss 3.3545 (3.0914) grad_norm 2.3338 (2.6063) [2021-04-16 15:00:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][90/1251] eta 0:07:58 lr 0.000114 time 0.2938 (0.4119) loss 2.2149 (3.0405) grad_norm 2.5390 (2.6096) [2021-04-16 15:00:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][100/1251] eta 0:07:39 lr 0.000114 time 0.2672 (0.3990) loss 3.4237 (3.0137) grad_norm 2.8495 (2.6138) [2021-04-16 15:00:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][110/1251] eta 0:07:22 lr 0.000114 time 0.2888 (0.3880) loss 3.2768 (3.0132) grad_norm 2.2673 (2.6096) [2021-04-16 15:00:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][120/1251] eta 0:07:09 lr 0.000114 time 0.2782 (0.3798) loss 3.0795 (3.0276) grad_norm 2.5992 (2.6142) [2021-04-16 15:00:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][130/1251] eta 0:06:59 lr 0.000114 time 0.2905 (0.3739) loss 3.1751 (3.0338) grad_norm 2.2710 (2.5981) [2021-04-16 15:00:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][140/1251] eta 0:06:50 lr 0.000114 time 0.2743 (0.3691) loss 3.4707 (3.0266) grad_norm 2.4495 (2.5894) [2021-04-16 15:00:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][150/1251] eta 0:06:39 lr 0.000113 time 0.2602 (0.3625) loss 3.2743 (3.0280) grad_norm 2.3920 (2.5810) [2021-04-16 15:00:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][160/1251] eta 0:06:29 lr 0.000113 time 0.2952 (0.3574) loss 3.0257 (3.0331) grad_norm 2.6212 (2.5819) [2021-04-16 15:00:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][170/1251] eta 0:06:21 lr 0.000113 time 0.2907 (0.3527) loss 3.7722 (3.0433) grad_norm 2.6236 (2.5790) [2021-04-16 15:00:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][180/1251] eta 0:06:13 lr 0.000113 time 0.2719 (0.3484) loss 3.5478 (3.0495) grad_norm 3.1269 (2.5803) [2021-04-16 15:00:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][190/1251] eta 0:06:06 lr 0.000113 time 0.2720 (0.3451) loss 3.7102 (3.0446) grad_norm 2.4877 (2.5799) [2021-04-16 15:00:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][200/1251] eta 0:05:59 lr 0.000113 time 0.2732 (0.3416) loss 3.3904 (3.0477) grad_norm 2.2437 (2.5707) [2021-04-16 15:00:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][210/1251] eta 0:05:52 lr 0.000113 time 0.2580 (0.3383) loss 3.8292 (3.0687) grad_norm 2.3053 (2.5684) [2021-04-16 15:00:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][220/1251] eta 0:05:46 lr 0.000113 time 0.2910 (0.3361) loss 3.1238 (3.0721) grad_norm 2.7938 (2.5702) [2021-04-16 15:01:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][230/1251] eta 0:05:40 lr 0.000113 time 0.2565 (0.3336) loss 3.3941 (3.0670) grad_norm 3.0759 (2.5708) [2021-04-16 15:01:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][240/1251] eta 0:05:34 lr 0.000113 time 0.2669 (0.3312) loss 3.4426 (3.0669) grad_norm 2.7435 (2.5730) [2021-04-16 15:01:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][250/1251] eta 0:05:29 lr 0.000113 time 0.2664 (0.3296) loss 3.3713 (3.0724) grad_norm 2.8349 (2.5691) [2021-04-16 15:01:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][260/1251] eta 0:05:24 lr 0.000113 time 0.2820 (0.3275) loss 2.9816 (3.0665) grad_norm 2.4374 (2.5633) [2021-04-16 15:01:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][270/1251] eta 0:05:19 lr 0.000113 time 0.2780 (0.3258) loss 3.0933 (3.0747) grad_norm 2.1339 (2.5570) [2021-04-16 15:01:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][280/1251] eta 0:05:14 lr 0.000113 time 0.2734 (0.3239) loss 2.5899 (3.0789) grad_norm 2.9491 (2.5597) [2021-04-16 15:01:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][290/1251] eta 0:05:09 lr 0.000113 time 0.2774 (0.3224) loss 2.8802 (3.0867) grad_norm 3.7240 (2.5671) [2021-04-16 15:01:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][300/1251] eta 0:05:05 lr 0.000113 time 0.2729 (0.3209) loss 2.4986 (3.0802) grad_norm 3.1009 (2.5771) [2021-04-16 15:01:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][310/1251] eta 0:05:00 lr 0.000113 time 0.2613 (0.3194) loss 3.2000 (3.0774) grad_norm 3.6792 (2.5870) [2021-04-16 15:01:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][320/1251] eta 0:04:56 lr 0.000113 time 0.2460 (0.3181) loss 2.0960 (3.0716) grad_norm 3.2639 (2.5854) [2021-04-16 15:01:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][330/1251] eta 0:04:51 lr 0.000113 time 0.2587 (0.3169) loss 3.4781 (3.0735) grad_norm 2.4857 (2.5864) [2021-04-16 15:01:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][340/1251] eta 0:04:47 lr 0.000113 time 0.2894 (0.3159) loss 3.4513 (3.0759) grad_norm 2.5888 (2.5865) [2021-04-16 15:01:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][350/1251] eta 0:04:43 lr 0.000113 time 0.3130 (0.3148) loss 2.4671 (3.0697) grad_norm 2.4867 (2.5838) [2021-04-16 15:01:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][360/1251] eta 0:04:39 lr 0.000113 time 0.2699 (0.3142) loss 2.0201 (3.0722) grad_norm 2.6957 (2.5827) [2021-04-16 15:01:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][370/1251] eta 0:04:35 lr 0.000113 time 0.2559 (0.3131) loss 3.3232 (3.0771) grad_norm 2.8079 (2.5824) [2021-04-16 15:01:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][380/1251] eta 0:04:31 lr 0.000113 time 0.2600 (0.3122) loss 3.0156 (3.0707) grad_norm 3.1967 (2.5907) [2021-04-16 15:01:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][390/1251] eta 0:04:28 lr 0.000113 time 0.2914 (0.3116) loss 2.6740 (3.0721) grad_norm 2.7864 (2.5901) [2021-04-16 15:01:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][400/1251] eta 0:04:24 lr 0.000113 time 0.2803 (0.3108) loss 3.3101 (3.0741) grad_norm 2.5570 (2.5909) [2021-04-16 15:01:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][410/1251] eta 0:04:20 lr 0.000113 time 0.2807 (0.3099) loss 2.1488 (3.0696) grad_norm 2.6798 (2.5932) [2021-04-16 15:01:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][420/1251] eta 0:04:16 lr 0.000113 time 0.2887 (0.3092) loss 3.5522 (3.0765) grad_norm 2.2375 (2.5936) [2021-04-16 15:01:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][430/1251] eta 0:04:13 lr 0.000113 time 0.2853 (0.3086) loss 3.9400 (3.0835) grad_norm 2.9021 (2.5978) [2021-04-16 15:02:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][440/1251] eta 0:04:09 lr 0.000113 time 0.2888 (0.3079) loss 3.5449 (3.0827) grad_norm 2.7982 (2.6023) [2021-04-16 15:02:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][450/1251] eta 0:04:05 lr 0.000113 time 0.2973 (0.3071) loss 2.7301 (3.0794) grad_norm 4.0128 (2.6030) [2021-04-16 15:02:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][460/1251] eta 0:04:02 lr 0.000113 time 0.2729 (0.3064) loss 2.9057 (3.0821) grad_norm 3.1686 (2.6077) [2021-04-16 15:02:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][470/1251] eta 0:03:58 lr 0.000113 time 0.2679 (0.3057) loss 2.1114 (3.0744) grad_norm 2.6405 (2.6089) [2021-04-16 15:02:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][480/1251] eta 0:03:55 lr 0.000113 time 0.2905 (0.3051) loss 4.0285 (3.0817) grad_norm 2.4995 (2.6089) [2021-04-16 15:02:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][490/1251] eta 0:03:51 lr 0.000113 time 0.2613 (0.3045) loss 3.4035 (3.0795) grad_norm 2.2740 (2.6092) [2021-04-16 15:02:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][500/1251] eta 0:03:48 lr 0.000113 time 0.2544 (0.3038) loss 2.2567 (3.0749) grad_norm 2.5191 (2.6076) [2021-04-16 15:02:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][510/1251] eta 0:03:44 lr 0.000113 time 0.2888 (0.3033) loss 3.8094 (3.0780) grad_norm 2.4216 (2.6102) [2021-04-16 15:02:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][520/1251] eta 0:03:41 lr 0.000113 time 0.2662 (0.3030) loss 3.0304 (3.0746) grad_norm 2.4028 (2.6127) [2021-04-16 15:02:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][530/1251] eta 0:03:38 lr 0.000113 time 0.2759 (0.3024) loss 3.7783 (3.0739) grad_norm 2.4059 (2.6093) [2021-04-16 15:02:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][540/1251] eta 0:03:34 lr 0.000113 time 0.2699 (0.3023) loss 2.6290 (3.0727) grad_norm 2.6051 (2.6094) [2021-04-16 15:02:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][550/1251] eta 0:03:31 lr 0.000112 time 0.2488 (0.3018) loss 2.6635 (3.0699) grad_norm 2.3674 (2.6081) [2021-04-16 15:02:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][560/1251] eta 0:03:28 lr 0.000112 time 0.3060 (0.3014) loss 3.1570 (3.0753) grad_norm 3.0504 (2.6055) [2021-04-16 15:02:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][570/1251] eta 0:03:24 lr 0.000112 time 0.2822 (0.3009) loss 3.4719 (3.0773) grad_norm 2.6958 (2.6030) [2021-04-16 15:02:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][580/1251] eta 0:03:21 lr 0.000112 time 0.2665 (0.3007) loss 2.9261 (3.0790) grad_norm 2.5148 (2.6056) [2021-04-16 15:02:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][590/1251] eta 0:03:18 lr 0.000112 time 0.2649 (0.3002) loss 2.0747 (3.0760) grad_norm 2.1633 (2.6036) [2021-04-16 15:02:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][600/1251] eta 0:03:15 lr 0.000112 time 0.2684 (0.2998) loss 3.7534 (3.0769) grad_norm 2.7364 (2.6019) [2021-04-16 15:02:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][610/1251] eta 0:03:12 lr 0.000112 time 0.2821 (0.2996) loss 3.4816 (3.0794) grad_norm 2.2926 (2.6022) [2021-04-16 15:02:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][620/1251] eta 0:03:08 lr 0.000112 time 0.2926 (0.2993) loss 2.5059 (3.0774) grad_norm 2.6446 (2.6013) [2021-04-16 15:02:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][630/1251] eta 0:03:05 lr 0.000112 time 0.2772 (0.2990) loss 2.9310 (3.0786) grad_norm 2.5813 (2.6023) [2021-04-16 15:02:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][640/1251] eta 0:03:02 lr 0.000112 time 0.2641 (0.2986) loss 2.2346 (3.0772) grad_norm 2.5110 (2.6028) [2021-04-16 15:02:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][650/1251] eta 0:02:59 lr 0.000112 time 0.2989 (0.2983) loss 3.1630 (3.0767) grad_norm 2.2972 (2.6035) [2021-04-16 15:03:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][660/1251] eta 0:02:56 lr 0.000112 time 0.2494 (0.2979) loss 2.3595 (3.0750) grad_norm 2.6854 (2.6043) [2021-04-16 15:03:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][670/1251] eta 0:02:52 lr 0.000112 time 0.2892 (0.2975) loss 3.2649 (3.0750) grad_norm 3.0074 (2.6014) [2021-04-16 15:03:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][680/1251] eta 0:02:49 lr 0.000112 time 0.2789 (0.2971) loss 2.7188 (3.0746) grad_norm 2.7924 (2.5998) [2021-04-16 15:03:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][690/1251] eta 0:02:46 lr 0.000112 time 0.2637 (0.2970) loss 3.7961 (3.0748) grad_norm 2.4429 (2.6066) [2021-04-16 15:03:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][700/1251] eta 0:02:43 lr 0.000112 time 0.4228 (0.2969) loss 3.5839 (3.0753) grad_norm 2.5179 (2.6082) [2021-04-16 15:03:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][710/1251] eta 0:02:40 lr 0.000112 time 0.2620 (0.2966) loss 1.9594 (3.0713) grad_norm 3.3892 (2.6110) [2021-04-16 15:03:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][720/1251] eta 0:02:37 lr 0.000112 time 0.2831 (0.2966) loss 3.3157 (3.0747) grad_norm 2.4356 (2.6106) [2021-04-16 15:03:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][730/1251] eta 0:02:34 lr 0.000112 time 0.2836 (0.2964) loss 3.6379 (3.0736) grad_norm 2.4876 (2.6088) [2021-04-16 15:03:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][740/1251] eta 0:02:31 lr 0.000112 time 0.2689 (0.2962) loss 3.3335 (3.0735) grad_norm 2.4897 (2.6063) [2021-04-16 15:03:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][750/1251] eta 0:02:28 lr 0.000112 time 0.2539 (0.2959) loss 3.1622 (3.0729) grad_norm 2.7086 (2.6036) [2021-04-16 15:03:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][760/1251] eta 0:02:25 lr 0.000112 time 0.2781 (0.2956) loss 2.9420 (3.0750) grad_norm 2.9683 (2.6048) [2021-04-16 15:03:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][770/1251] eta 0:02:22 lr 0.000112 time 0.2783 (0.2955) loss 3.5668 (3.0754) grad_norm 2.3081 (2.6037) [2021-04-16 15:03:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][780/1251] eta 0:02:19 lr 0.000112 time 0.2827 (0.2953) loss 3.5224 (3.0760) grad_norm 2.3905 (2.6002) [2021-04-16 15:03:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][790/1251] eta 0:02:16 lr 0.000112 time 0.2603 (0.2951) loss 2.8343 (3.0720) grad_norm 2.7279 (2.5994) [2021-04-16 15:03:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][800/1251] eta 0:02:12 lr 0.000112 time 0.2772 (0.2949) loss 3.7378 (3.0727) grad_norm 2.5002 (2.5976) [2021-04-16 15:03:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][810/1251] eta 0:02:09 lr 0.000112 time 0.2635 (0.2947) loss 2.7983 (3.0745) grad_norm 2.5254 (2.5988) [2021-04-16 15:03:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][820/1251] eta 0:02:06 lr 0.000112 time 0.2693 (0.2945) loss 3.5757 (3.0771) grad_norm 2.2257 (2.5993) [2021-04-16 15:03:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][830/1251] eta 0:02:03 lr 0.000112 time 0.2503 (0.2944) loss 3.2544 (3.0777) grad_norm 2.6132 (2.5995) [2021-04-16 15:03:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][840/1251] eta 0:02:00 lr 0.000112 time 0.2665 (0.2942) loss 3.3425 (3.0775) grad_norm 3.5133 (2.5983) [2021-04-16 15:03:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][850/1251] eta 0:01:57 lr 0.000112 time 0.2775 (0.2940) loss 3.0037 (3.0793) grad_norm 2.8454 (2.5987) [2021-04-16 15:03:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][860/1251] eta 0:01:54 lr 0.000112 time 0.2959 (0.2937) loss 3.4823 (3.0814) grad_norm 2.6357 (2.5968) [2021-04-16 15:04:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][870/1251] eta 0:01:51 lr 0.000112 time 0.2858 (0.2935) loss 3.4419 (3.0815) grad_norm 2.3299 (2.5974) [2021-04-16 15:04:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][880/1251] eta 0:01:48 lr 0.000112 time 0.2818 (0.2933) loss 2.8982 (3.0808) grad_norm 2.3908 (2.5951) [2021-04-16 15:04:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][890/1251] eta 0:01:45 lr 0.000112 time 0.2674 (0.2931) loss 3.5247 (3.0820) grad_norm 2.2448 (2.5942) [2021-04-16 15:04:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][900/1251] eta 0:01:42 lr 0.000112 time 0.2842 (0.2930) loss 2.5423 (3.0827) grad_norm 2.3664 (2.5946) [2021-04-16 15:04:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][910/1251] eta 0:01:39 lr 0.000112 time 0.2945 (0.2928) loss 3.4101 (3.0798) grad_norm 2.5163 (2.5951) [2021-04-16 15:04:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][920/1251] eta 0:01:36 lr 0.000112 time 0.2830 (0.2927) loss 2.6219 (3.0819) grad_norm 2.3891 (2.5943) [2021-04-16 15:04:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][930/1251] eta 0:01:33 lr 0.000112 time 0.2560 (0.2925) loss 3.6313 (3.0840) grad_norm 2.1947 (2.5939) [2021-04-16 15:04:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][940/1251] eta 0:01:30 lr 0.000111 time 0.2709 (0.2924) loss 3.2802 (3.0822) grad_norm 2.6744 (2.5957) [2021-04-16 15:04:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][950/1251] eta 0:01:27 lr 0.000111 time 0.2823 (0.2923) loss 3.2934 (3.0839) grad_norm 2.3204 (2.5964) [2021-04-16 15:04:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][960/1251] eta 0:01:25 lr 0.000111 time 0.3187 (0.2921) loss 3.4583 (3.0868) grad_norm 2.4211 (2.5953) [2021-04-16 15:04:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][970/1251] eta 0:01:22 lr 0.000111 time 0.2790 (0.2921) loss 3.0462 (3.0893) grad_norm 2.9287 (2.5963) [2021-04-16 15:04:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][980/1251] eta 0:01:19 lr 0.000111 time 0.2685 (0.2919) loss 2.7246 (3.0907) grad_norm 3.4918 (2.5968) [2021-04-16 15:04:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][990/1251] eta 0:01:16 lr 0.000111 time 0.2828 (0.2917) loss 3.2404 (3.0895) grad_norm 2.5740 (2.5958) [2021-04-16 15:04:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][1000/1251] eta 0:01:13 lr 0.000111 time 0.2737 (0.2915) loss 3.5035 (3.0897) grad_norm 2.3494 (2.5964) [2021-04-16 15:04:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][1010/1251] eta 0:01:10 lr 0.000111 time 0.2592 (0.2913) loss 2.1310 (3.0872) grad_norm 2.3925 (inf) [2021-04-16 15:04:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][1020/1251] eta 0:01:07 lr 0.000111 time 0.2527 (0.2911) loss 3.1079 (3.0887) grad_norm 2.8955 (inf) [2021-04-16 15:04:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][1030/1251] eta 0:01:04 lr 0.000111 time 0.2573 (0.2909) loss 2.6799 (3.0877) grad_norm 2.9106 (inf) [2021-04-16 15:04:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][1040/1251] eta 0:01:01 lr 0.000111 time 0.2703 (0.2908) loss 3.3507 (3.0865) grad_norm 2.3849 (inf) [2021-04-16 15:04:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][1050/1251] eta 0:00:58 lr 0.000111 time 0.2657 (0.2906) loss 3.3997 (3.0881) grad_norm 3.2046 (inf) [2021-04-16 15:04:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][1060/1251] eta 0:00:55 lr 0.000111 time 0.2797 (0.2905) loss 2.5975 (3.0853) grad_norm 2.3918 (inf) [2021-04-16 15:04:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][1070/1251] eta 0:00:52 lr 0.000111 time 0.2815 (0.2904) loss 3.4861 (3.0862) grad_norm 2.4845 (inf) [2021-04-16 15:04:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][1080/1251] eta 0:00:49 lr 0.000111 time 0.2855 (0.2902) loss 3.1608 (3.0867) grad_norm 2.2353 (inf) [2021-04-16 15:05:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][1090/1251] eta 0:00:46 lr 0.000111 time 0.2738 (0.2903) loss 3.0662 (3.0870) grad_norm 2.4370 (inf) [2021-04-16 15:05:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][1100/1251] eta 0:00:43 lr 0.000111 time 0.2939 (0.2901) loss 2.8854 (3.0894) grad_norm 2.4985 (inf) [2021-04-16 15:05:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][1110/1251] eta 0:00:40 lr 0.000111 time 0.2905 (0.2900) loss 3.5273 (3.0902) grad_norm 2.7499 (inf) [2021-04-16 15:05:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][1120/1251] eta 0:00:37 lr 0.000111 time 0.2733 (0.2901) loss 3.4483 (3.0897) grad_norm 3.1360 (inf) [2021-04-16 15:05:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][1130/1251] eta 0:00:35 lr 0.000111 time 0.2581 (0.2899) loss 3.5588 (3.0924) grad_norm 2.4499 (inf) [2021-04-16 15:05:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][1140/1251] eta 0:00:32 lr 0.000111 time 0.2655 (0.2898) loss 2.8687 (3.0927) grad_norm 2.2133 (inf) [2021-04-16 15:05:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][1150/1251] eta 0:00:29 lr 0.000111 time 0.2838 (0.2898) loss 3.5127 (3.0946) grad_norm 2.4322 (inf) [2021-04-16 15:05:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][1160/1251] eta 0:00:26 lr 0.000111 time 0.2806 (0.2898) loss 3.2688 (3.0938) grad_norm 2.2415 (inf) [2021-04-16 15:05:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][1170/1251] eta 0:00:23 lr 0.000111 time 0.3007 (0.2897) loss 3.1036 (3.0928) grad_norm 2.5502 (inf) [2021-04-16 15:05:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][1180/1251] eta 0:00:20 lr 0.000111 time 0.2664 (0.2896) loss 3.1910 (3.0925) grad_norm 3.2520 (inf) [2021-04-16 15:05:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][1190/1251] eta 0:00:17 lr 0.000111 time 0.2425 (0.2895) loss 2.9843 (3.0930) grad_norm 2.2683 (inf) [2021-04-16 15:05:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][1200/1251] eta 0:00:14 lr 0.000111 time 0.2803 (0.2894) loss 3.4993 (3.0931) grad_norm 2.5425 (inf) [2021-04-16 15:05:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][1210/1251] eta 0:00:11 lr 0.000111 time 0.2798 (0.2893) loss 3.8650 (3.0942) grad_norm 2.5398 (inf) [2021-04-16 15:05:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][1220/1251] eta 0:00:08 lr 0.000111 time 0.2964 (0.2891) loss 3.1153 (3.0964) grad_norm 3.1053 (inf) [2021-04-16 15:05:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][1230/1251] eta 0:00:06 lr 0.000111 time 0.3041 (0.2890) loss 2.2701 (3.0968) grad_norm 2.9437 (inf) [2021-04-16 15:05:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][1240/1251] eta 0:00:03 lr 0.000111 time 0.2476 (0.2889) loss 3.2676 (3.0961) grad_norm 3.3121 (inf) [2021-04-16 15:05:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [237/300][1250/1251] eta 0:00:00 lr 0.000111 time 0.2480 (0.2886) loss 3.1500 (3.0946) grad_norm 3.0502 (inf) [2021-04-16 15:05:56 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 237 training takes 0:06:11 [2021-04-16 15:05:56 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_237.pth saving...... [2021-04-16 15:06:22 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_237.pth saved !!! [2021-04-16 15:06:23 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.163 (1.163) Loss 0.8915 (0.8915) Acc@1 79.102 (79.102) Acc@5 94.922 (94.922) [2021-04-16 15:06:24 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.124 (0.228) Loss 0.8096 (0.8685) Acc@1 81.250 (79.883) Acc@5 95.801 (95.179) [2021-04-16 15:06:27 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.144 (0.227) Loss 0.8755 (0.8651) Acc@1 79.590 (79.915) Acc@5 95.312 (95.052) [2021-04-16 15:06:29 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.108 (0.229) Loss 0.8468 (0.8691) Acc@1 80.078 (79.854) Acc@5 94.824 (94.963) [2021-04-16 15:06:31 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.185 (0.222) Loss 0.8276 (0.8688) Acc@1 81.836 (79.849) Acc@5 95.801 (94.950) [2021-04-16 15:06:47 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 79.918 Acc@5 94.984 [2021-04-16 15:06:47 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 79.9% [2021-04-16 15:06:47 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 79.96% [2021-04-16 15:06:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][0/1251] eta 3:27:38 lr 0.000111 time 9.9589 (9.9589) loss 3.1785 (3.1785) grad_norm 2.3340 (2.3340) [2021-04-16 15:06:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][10/1251] eta 0:24:05 lr 0.000111 time 0.3501 (1.1652) loss 2.5301 (2.8174) grad_norm 2.9433 (2.7285) [2021-04-16 15:07:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][20/1251] eta 0:15:15 lr 0.000111 time 0.3142 (0.7439) loss 3.2538 (2.9405) grad_norm 2.7163 (2.6696) [2021-04-16 15:07:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][30/1251] eta 0:12:10 lr 0.000111 time 0.3012 (0.5979) loss 4.1258 (3.0800) grad_norm 2.6208 (2.6315) [2021-04-16 15:07:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][40/1251] eta 0:10:30 lr 0.000111 time 0.2862 (0.5202) loss 3.4353 (3.1161) grad_norm 2.3120 (2.5667) [2021-04-16 15:07:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][50/1251] eta 0:09:27 lr 0.000111 time 0.2637 (0.4723) loss 3.5603 (3.1117) grad_norm 2.9010 (2.5550) [2021-04-16 15:07:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][60/1251] eta 0:08:44 lr 0.000111 time 0.2746 (0.4401) loss 3.1892 (3.1001) grad_norm 2.4009 (2.5477) [2021-04-16 15:07:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][70/1251] eta 0:08:12 lr 0.000111 time 0.2591 (0.4167) loss 3.7078 (3.1164) grad_norm 2.5917 (2.5446) [2021-04-16 15:07:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][80/1251] eta 0:07:48 lr 0.000111 time 0.2859 (0.3997) loss 3.7471 (3.1335) grad_norm 2.4769 (2.5333) [2021-04-16 15:07:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][90/1251] eta 0:07:28 lr 0.000110 time 0.2989 (0.3860) loss 4.0542 (3.1450) grad_norm 2.3615 (2.5267) [2021-04-16 15:07:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][100/1251] eta 0:07:11 lr 0.000110 time 0.2744 (0.3753) loss 2.5745 (3.1331) grad_norm 2.2316 (2.6120) [2021-04-16 15:07:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][110/1251] eta 0:06:57 lr 0.000110 time 0.2500 (0.3662) loss 2.2765 (3.1348) grad_norm 2.2129 (2.6015) [2021-04-16 15:07:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][120/1251] eta 0:06:47 lr 0.000110 time 0.2786 (0.3601) loss 1.4431 (3.1125) grad_norm 2.7801 (2.6045) [2021-04-16 15:07:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][130/1251] eta 0:06:36 lr 0.000110 time 0.2740 (0.3536) loss 3.2843 (3.1268) grad_norm 2.4034 (2.6116) [2021-04-16 15:07:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][140/1251] eta 0:06:28 lr 0.000110 time 0.2711 (0.3495) loss 3.1243 (3.1224) grad_norm 2.3076 (2.6073) [2021-04-16 15:07:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][150/1251] eta 0:06:18 lr 0.000110 time 0.2543 (0.3441) loss 3.3059 (3.1175) grad_norm 2.6839 (2.6095) [2021-04-16 15:07:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][160/1251] eta 0:06:10 lr 0.000110 time 0.2593 (0.3398) loss 3.2747 (3.1212) grad_norm 3.0007 (2.6146) [2021-04-16 15:07:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][170/1251] eta 0:06:03 lr 0.000110 time 0.2601 (0.3360) loss 2.1114 (3.1123) grad_norm 2.2741 (2.6030) [2021-04-16 15:07:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][180/1251] eta 0:05:56 lr 0.000110 time 0.2757 (0.3333) loss 3.4574 (3.1289) grad_norm 2.3654 (2.5989) [2021-04-16 15:07:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][190/1251] eta 0:05:50 lr 0.000110 time 0.2783 (0.3302) loss 3.7180 (3.1309) grad_norm 2.3438 (2.6025) [2021-04-16 15:07:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][200/1251] eta 0:05:44 lr 0.000110 time 0.2886 (0.3277) loss 3.4630 (3.1453) grad_norm 2.4567 (2.5980) [2021-04-16 15:07:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][210/1251] eta 0:05:38 lr 0.000110 time 0.2568 (0.3251) loss 3.0849 (3.1369) grad_norm 2.4281 (2.5903) [2021-04-16 15:07:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][220/1251] eta 0:05:32 lr 0.000110 time 0.2583 (0.3229) loss 2.7049 (3.1352) grad_norm 2.8270 (2.5881) [2021-04-16 15:08:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][230/1251] eta 0:05:27 lr 0.000110 time 0.2919 (0.3208) loss 3.8262 (3.1470) grad_norm 2.9977 (2.5826) [2021-04-16 15:08:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][240/1251] eta 0:05:22 lr 0.000110 time 0.2652 (0.3191) loss 2.7464 (3.1496) grad_norm 3.2170 (2.5850) [2021-04-16 15:08:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][250/1251] eta 0:05:17 lr 0.000110 time 0.2869 (0.3175) loss 3.0293 (3.1359) grad_norm 2.5898 (2.5889) [2021-04-16 15:08:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][260/1251] eta 0:05:13 lr 0.000110 time 0.2589 (0.3158) loss 3.3002 (3.1473) grad_norm 2.7549 (2.5889) [2021-04-16 15:08:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][270/1251] eta 0:05:08 lr 0.000110 time 0.2964 (0.3146) loss 2.9143 (3.1475) grad_norm 3.2415 (2.5906) [2021-04-16 15:08:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][280/1251] eta 0:05:04 lr 0.000110 time 0.2752 (0.3133) loss 3.2068 (3.1355) grad_norm 2.5332 (2.5913) [2021-04-16 15:08:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][290/1251] eta 0:04:59 lr 0.000110 time 0.2607 (0.3119) loss 2.9794 (3.1236) grad_norm 2.3655 (2.5942) [2021-04-16 15:08:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][300/1251] eta 0:04:56 lr 0.000110 time 0.2833 (0.3113) loss 3.2556 (3.1182) grad_norm 2.6329 (2.5918) [2021-04-16 15:08:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][310/1251] eta 0:04:52 lr 0.000110 time 0.2847 (0.3103) loss 3.1309 (3.1204) grad_norm 2.5110 (2.5873) [2021-04-16 15:08:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][320/1251] eta 0:04:48 lr 0.000110 time 0.2829 (0.3098) loss 3.3309 (3.1206) grad_norm 2.7600 (2.5888) [2021-04-16 15:08:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][330/1251] eta 0:04:44 lr 0.000110 time 0.2798 (0.3088) loss 2.8590 (3.1208) grad_norm 2.9918 (2.5957) [2021-04-16 15:08:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][340/1251] eta 0:04:40 lr 0.000110 time 0.2913 (0.3078) loss 4.1720 (3.1222) grad_norm 2.8581 (2.6024) [2021-04-16 15:08:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][350/1251] eta 0:04:36 lr 0.000110 time 0.2922 (0.3072) loss 3.0633 (3.1293) grad_norm 2.8333 (2.6056) [2021-04-16 15:08:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][360/1251] eta 0:04:33 lr 0.000110 time 0.2816 (0.3068) loss 2.8432 (3.1309) grad_norm 2.7151 (2.6160) [2021-04-16 15:08:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][370/1251] eta 0:04:29 lr 0.000110 time 0.2592 (0.3058) loss 3.5134 (3.1323) grad_norm 2.4083 (2.6123) [2021-04-16 15:08:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][380/1251] eta 0:04:25 lr 0.000110 time 0.2585 (0.3050) loss 3.7283 (3.1367) grad_norm 3.0709 (2.6113) [2021-04-16 15:08:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][390/1251] eta 0:04:22 lr 0.000110 time 0.2656 (0.3043) loss 2.5786 (3.1362) grad_norm 2.3121 (2.6090) [2021-04-16 15:08:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][400/1251] eta 0:04:18 lr 0.000110 time 0.2673 (0.3038) loss 3.6596 (3.1370) grad_norm 2.9893 (2.6064) [2021-04-16 15:08:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][410/1251] eta 0:04:14 lr 0.000110 time 0.2619 (0.3031) loss 3.2640 (3.1428) grad_norm 2.6299 (2.6027) [2021-04-16 15:08:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][420/1251] eta 0:04:11 lr 0.000110 time 0.3004 (0.3025) loss 3.8949 (3.1502) grad_norm 2.3987 (2.5986) [2021-04-16 15:08:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][430/1251] eta 0:04:07 lr 0.000110 time 0.2496 (0.3019) loss 3.1782 (3.1494) grad_norm 2.5860 (2.5969) [2021-04-16 15:08:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][440/1251] eta 0:04:04 lr 0.000110 time 0.2984 (0.3013) loss 3.3309 (3.1488) grad_norm 2.5783 (2.6043) [2021-04-16 15:09:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][450/1251] eta 0:04:00 lr 0.000110 time 0.2867 (0.3008) loss 2.0874 (3.1504) grad_norm 2.1954 (2.6065) [2021-04-16 15:09:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][460/1251] eta 0:03:57 lr 0.000110 time 0.2797 (0.3002) loss 3.1675 (3.1506) grad_norm 2.5799 (2.6033) [2021-04-16 15:09:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][470/1251] eta 0:03:54 lr 0.000110 time 0.2655 (0.2997) loss 3.5682 (3.1531) grad_norm 2.1012 (2.5989) [2021-04-16 15:09:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][480/1251] eta 0:03:50 lr 0.000110 time 0.2820 (0.2993) loss 3.7034 (3.1580) grad_norm 2.5689 (2.5993) [2021-04-16 15:09:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][490/1251] eta 0:03:47 lr 0.000109 time 0.2659 (0.2991) loss 3.4960 (3.1504) grad_norm 2.7017 (2.5959) [2021-04-16 15:09:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][500/1251] eta 0:03:44 lr 0.000109 time 0.3371 (0.2989) loss 3.4250 (3.1502) grad_norm 2.5056 (2.5935) [2021-04-16 15:09:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][510/1251] eta 0:03:41 lr 0.000109 time 0.2676 (0.2985) loss 3.5761 (3.1543) grad_norm 2.5308 (2.5938) [2021-04-16 15:09:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][520/1251] eta 0:03:37 lr 0.000109 time 0.2787 (0.2980) loss 3.4699 (3.1536) grad_norm 2.3313 (2.5930) [2021-04-16 15:09:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][530/1251] eta 0:03:34 lr 0.000109 time 0.2937 (0.2976) loss 3.3132 (3.1529) grad_norm 2.7123 (2.5941) [2021-04-16 15:09:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][540/1251] eta 0:03:31 lr 0.000109 time 0.2811 (0.2974) loss 2.9842 (3.1506) grad_norm 2.4334 (2.5935) [2021-04-16 15:09:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][550/1251] eta 0:03:28 lr 0.000109 time 0.2597 (0.2971) loss 3.6389 (3.1485) grad_norm 2.2673 (2.5921) [2021-04-16 15:09:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][560/1251] eta 0:03:25 lr 0.000109 time 0.2629 (0.2967) loss 3.3666 (3.1516) grad_norm 2.5534 (2.5919) [2021-04-16 15:09:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][570/1251] eta 0:03:21 lr 0.000109 time 0.2797 (0.2965) loss 2.5710 (3.1509) grad_norm 2.7622 (2.5916) [2021-04-16 15:09:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][580/1251] eta 0:03:18 lr 0.000109 time 0.3017 (0.2965) loss 2.7766 (3.1509) grad_norm 2.9259 (2.5928) [2021-04-16 15:09:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][590/1251] eta 0:03:15 lr 0.000109 time 0.2556 (0.2963) loss 2.8958 (3.1504) grad_norm 3.2889 (2.5955) [2021-04-16 15:09:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][600/1251] eta 0:03:12 lr 0.000109 time 0.2738 (0.2960) loss 3.0632 (3.1535) grad_norm 2.7815 (2.5958) [2021-04-16 15:09:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][610/1251] eta 0:03:09 lr 0.000109 time 0.2894 (0.2959) loss 3.3531 (3.1557) grad_norm 2.4954 (2.5976) [2021-04-16 15:09:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][620/1251] eta 0:03:06 lr 0.000109 time 0.2786 (0.2956) loss 3.1771 (3.1564) grad_norm 3.1238 (2.6009) [2021-04-16 15:09:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][630/1251] eta 0:03:03 lr 0.000109 time 0.2913 (0.2954) loss 3.7796 (3.1598) grad_norm 2.7251 (2.6025) [2021-04-16 15:09:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][640/1251] eta 0:03:00 lr 0.000109 time 0.2893 (0.2952) loss 2.7546 (3.1635) grad_norm 2.3724 (2.6032) [2021-04-16 15:09:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][650/1251] eta 0:02:57 lr 0.000109 time 0.2913 (0.2948) loss 3.9875 (3.1621) grad_norm 3.2548 (2.6042) [2021-04-16 15:10:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][660/1251] eta 0:02:54 lr 0.000109 time 0.2780 (0.2946) loss 3.3577 (3.1623) grad_norm 2.4692 (2.6034) [2021-04-16 15:10:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][670/1251] eta 0:02:51 lr 0.000109 time 0.2866 (0.2943) loss 2.8425 (3.1630) grad_norm 2.5874 (2.6045) [2021-04-16 15:10:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][680/1251] eta 0:02:47 lr 0.000109 time 0.2724 (0.2941) loss 2.9113 (3.1642) grad_norm 2.3885 (2.6029) [2021-04-16 15:10:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][690/1251] eta 0:02:44 lr 0.000109 time 0.2686 (0.2940) loss 2.7655 (3.1643) grad_norm 2.3304 (2.6016) [2021-04-16 15:10:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][700/1251] eta 0:02:41 lr 0.000109 time 0.2792 (0.2938) loss 3.5286 (3.1663) grad_norm 2.5636 (2.6004) [2021-04-16 15:10:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][710/1251] eta 0:02:38 lr 0.000109 time 0.2802 (0.2936) loss 3.2697 (3.1637) grad_norm 2.9417 (2.6036) [2021-04-16 15:10:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][720/1251] eta 0:02:35 lr 0.000109 time 0.2908 (0.2936) loss 3.4170 (3.1614) grad_norm 3.1479 (2.6083) [2021-04-16 15:10:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][730/1251] eta 0:02:32 lr 0.000109 time 0.2747 (0.2933) loss 2.9410 (3.1620) grad_norm 2.9284 (2.6081) [2021-04-16 15:10:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][740/1251] eta 0:02:29 lr 0.000109 time 0.2816 (0.2931) loss 3.0128 (3.1575) grad_norm 2.7135 (2.6066) [2021-04-16 15:10:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][750/1251] eta 0:02:26 lr 0.000109 time 0.2806 (0.2928) loss 3.4011 (3.1581) grad_norm 2.6318 (2.6040) [2021-04-16 15:10:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][760/1251] eta 0:02:23 lr 0.000109 time 0.2822 (0.2928) loss 3.1880 (3.1593) grad_norm 2.2565 (2.6026) [2021-04-16 15:10:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][770/1251] eta 0:02:20 lr 0.000109 time 0.2749 (0.2926) loss 3.0674 (3.1593) grad_norm 2.5818 (2.6037) [2021-04-16 15:10:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][780/1251] eta 0:02:17 lr 0.000109 time 0.2544 (0.2924) loss 3.4735 (3.1595) grad_norm 2.7858 (2.6031) [2021-04-16 15:10:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][790/1251] eta 0:02:14 lr 0.000109 time 0.2509 (0.2922) loss 2.7836 (3.1537) grad_norm 2.5315 (2.6031) [2021-04-16 15:10:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][800/1251] eta 0:02:11 lr 0.000109 time 0.2904 (0.2920) loss 3.2883 (3.1536) grad_norm 3.1593 (2.6056) [2021-04-16 15:10:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][810/1251] eta 0:02:08 lr 0.000109 time 0.3049 (0.2918) loss 3.1416 (3.1517) grad_norm 2.7554 (2.6051) [2021-04-16 15:10:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][820/1251] eta 0:02:05 lr 0.000109 time 0.2674 (0.2916) loss 2.6776 (3.1503) grad_norm 2.4491 (2.6051) [2021-04-16 15:10:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][830/1251] eta 0:02:02 lr 0.000109 time 0.2411 (0.2914) loss 2.6017 (3.1506) grad_norm 4.0260 (2.6059) [2021-04-16 15:10:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][840/1251] eta 0:01:59 lr 0.000109 time 0.2917 (0.2913) loss 3.3333 (3.1511) grad_norm 2.3639 (2.6059) [2021-04-16 15:10:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][850/1251] eta 0:01:56 lr 0.000109 time 0.2999 (0.2912) loss 3.0953 (3.1525) grad_norm 2.2079 (2.6062) [2021-04-16 15:10:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][860/1251] eta 0:01:53 lr 0.000109 time 0.2540 (0.2911) loss 2.4077 (3.1519) grad_norm 2.0786 (2.6037) [2021-04-16 15:11:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][870/1251] eta 0:01:50 lr 0.000109 time 0.2831 (0.2910) loss 3.0195 (3.1517) grad_norm 2.4739 (2.6033) [2021-04-16 15:11:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][880/1251] eta 0:01:47 lr 0.000109 time 0.3043 (0.2910) loss 2.7937 (3.1490) grad_norm 2.2800 (2.6021) [2021-04-16 15:11:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][890/1251] eta 0:01:45 lr 0.000108 time 0.2716 (0.2910) loss 3.2271 (3.1491) grad_norm 2.3316 (2.6023) [2021-04-16 15:11:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][900/1251] eta 0:01:42 lr 0.000108 time 0.2424 (0.2908) loss 3.1784 (3.1481) grad_norm 3.3939 (2.6032) [2021-04-16 15:11:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][910/1251] eta 0:01:39 lr 0.000108 time 0.2742 (0.2906) loss 3.4582 (3.1503) grad_norm 2.3503 (2.6017) [2021-04-16 15:11:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][920/1251] eta 0:01:36 lr 0.000108 time 0.3027 (0.2905) loss 3.7638 (3.1499) grad_norm 2.3313 (2.6023) [2021-04-16 15:11:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][930/1251] eta 0:01:33 lr 0.000108 time 0.2648 (0.2905) loss 3.4304 (3.1502) grad_norm 2.4957 (2.6020) [2021-04-16 15:11:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][940/1251] eta 0:01:30 lr 0.000108 time 0.2624 (0.2903) loss 2.6731 (3.1512) grad_norm 2.7180 (2.6017) [2021-04-16 15:11:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][950/1251] eta 0:01:27 lr 0.000108 time 0.2754 (0.2902) loss 2.3283 (3.1483) grad_norm 2.3718 (2.6007) [2021-04-16 15:11:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][960/1251] eta 0:01:24 lr 0.000108 time 0.3012 (0.2901) loss 3.6549 (3.1490) grad_norm 2.1314 (2.5989) [2021-04-16 15:11:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][970/1251] eta 0:01:21 lr 0.000108 time 0.3094 (0.2899) loss 3.7367 (3.1499) grad_norm 2.7228 (2.5973) [2021-04-16 15:11:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][980/1251] eta 0:01:18 lr 0.000108 time 0.2895 (0.2898) loss 3.1930 (3.1489) grad_norm 2.6953 (2.5982) [2021-04-16 15:11:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][990/1251] eta 0:01:15 lr 0.000108 time 0.2749 (0.2896) loss 3.7280 (3.1497) grad_norm 2.5701 (2.5997) [2021-04-16 15:11:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][1000/1251] eta 0:01:12 lr 0.000108 time 0.2550 (0.2894) loss 3.7059 (3.1506) grad_norm 2.2550 (2.5999) [2021-04-16 15:11:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][1010/1251] eta 0:01:09 lr 0.000108 time 0.2653 (0.2893) loss 3.1343 (3.1509) grad_norm 2.6657 (2.5984) [2021-04-16 15:11:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][1020/1251] eta 0:01:06 lr 0.000108 time 0.2459 (0.2891) loss 3.4109 (3.1531) grad_norm 3.1778 (2.5979) [2021-04-16 15:11:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][1030/1251] eta 0:01:03 lr 0.000108 time 0.2461 (0.2891) loss 3.6171 (3.1546) grad_norm 2.3375 (2.5984) [2021-04-16 15:11:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][1040/1251] eta 0:01:00 lr 0.000108 time 0.2636 (0.2889) loss 2.2118 (3.1526) grad_norm 2.3809 (2.5985) [2021-04-16 15:11:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][1050/1251] eta 0:00:58 lr 0.000108 time 0.2856 (0.2888) loss 2.1080 (3.1487) grad_norm 2.5664 (2.5976) [2021-04-16 15:11:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][1060/1251] eta 0:00:55 lr 0.000108 time 0.2760 (0.2887) loss 2.8632 (3.1489) grad_norm 2.5678 (2.5971) [2021-04-16 15:11:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][1070/1251] eta 0:00:52 lr 0.000108 time 0.2604 (0.2887) loss 3.2209 (3.1518) grad_norm 2.5751 (2.5959) [2021-04-16 15:11:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][1080/1251] eta 0:00:49 lr 0.000108 time 0.2841 (0.2886) loss 4.0740 (3.1519) grad_norm 2.8178 (2.5948) [2021-04-16 15:12:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][1090/1251] eta 0:00:46 lr 0.000108 time 0.2921 (0.2884) loss 2.8396 (3.1513) grad_norm 2.1612 (2.5934) [2021-04-16 15:12:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][1100/1251] eta 0:00:43 lr 0.000108 time 0.2853 (0.2883) loss 2.4418 (3.1499) grad_norm 2.7249 (2.5927) [2021-04-16 15:12:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][1110/1251] eta 0:00:40 lr 0.000108 time 0.2761 (0.2883) loss 3.2723 (3.1478) grad_norm 2.9386 (2.5914) [2021-04-16 15:12:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][1120/1251] eta 0:00:37 lr 0.000108 time 0.2736 (0.2882) loss 2.9924 (3.1467) grad_norm 2.4713 (2.5939) [2021-04-16 15:12:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][1130/1251] eta 0:00:34 lr 0.000108 time 0.2735 (0.2882) loss 3.6176 (3.1480) grad_norm 2.5043 (2.5931) [2021-04-16 15:12:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][1140/1251] eta 0:00:31 lr 0.000108 time 0.2963 (0.2882) loss 3.8642 (3.1482) grad_norm 2.5218 (2.5930) [2021-04-16 15:12:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][1150/1251] eta 0:00:29 lr 0.000108 time 0.2736 (0.2883) loss 3.0137 (3.1491) grad_norm 2.3027 (2.5935) [2021-04-16 15:12:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][1160/1251] eta 0:00:26 lr 0.000108 time 0.2828 (0.2881) loss 2.6333 (3.1464) grad_norm 2.7050 (2.5941) [2021-04-16 15:12:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][1170/1251] eta 0:00:23 lr 0.000108 time 0.2599 (0.2882) loss 3.3954 (3.1443) grad_norm 2.5820 (2.5931) [2021-04-16 15:12:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][1180/1251] eta 0:00:20 lr 0.000108 time 0.2632 (0.2882) loss 3.5780 (3.1454) grad_norm 2.4284 (2.5915) [2021-04-16 15:12:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][1190/1251] eta 0:00:17 lr 0.000108 time 0.2830 (0.2881) loss 3.5217 (3.1445) grad_norm 2.2796 (2.5906) [2021-04-16 15:12:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][1200/1251] eta 0:00:14 lr 0.000108 time 0.2740 (0.2881) loss 1.8429 (3.1427) grad_norm 2.3636 (2.5895) [2021-04-16 15:12:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][1210/1251] eta 0:00:11 lr 0.000108 time 0.2935 (0.2880) loss 3.9273 (3.1431) grad_norm 2.2994 (2.5899) [2021-04-16 15:12:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][1220/1251] eta 0:00:08 lr 0.000108 time 0.2916 (0.2880) loss 3.2354 (3.1441) grad_norm 3.0072 (2.5893) [2021-04-16 15:12:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][1230/1251] eta 0:00:06 lr 0.000108 time 0.2715 (0.2879) loss 3.3024 (3.1435) grad_norm 2.4392 (2.5886) [2021-04-16 15:12:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][1240/1251] eta 0:00:03 lr 0.000108 time 0.2479 (0.2877) loss 3.8310 (3.1419) grad_norm 2.8618 (2.5902) [2021-04-16 15:12:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [238/300][1250/1251] eta 0:00:00 lr 0.000108 time 0.2528 (0.2874) loss 3.1551 (3.1436) grad_norm 2.6361 (2.5898) [2021-04-16 15:13:00 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 238 training takes 0:06:13 [2021-04-16 15:13:00 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_238.pth saving...... [2021-04-16 15:13:28 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_238.pth saved !!! [2021-04-16 15:13:29 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.218 (1.218) Loss 0.8594 (0.8594) Acc@1 79.199 (79.199) Acc@5 94.531 (94.531) [2021-04-16 15:13:30 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.129 (0.225) Loss 0.8886 (0.8810) Acc@1 77.930 (79.173) Acc@5 94.922 (94.682) [2021-04-16 15:13:32 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.104 (0.233) Loss 0.8383 (0.8606) Acc@1 81.348 (79.753) Acc@5 95.410 (94.987) [2021-04-16 15:13:34 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.151 (0.224) Loss 0.8888 (0.8596) Acc@1 79.004 (79.839) Acc@5 95.020 (95.004) [2021-04-16 15:13:36 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.094 (0.211) Loss 0.8764 (0.8561) Acc@1 78.711 (79.869) Acc@5 94.727 (95.039) [2021-04-16 15:14:03 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 79.872 Acc@5 95.022 [2021-04-16 15:14:03 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 79.9% [2021-04-16 15:14:03 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 79.96% [2021-04-16 15:14:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][0/1251] eta 0:29:14 lr 0.000108 time 1.4022 (1.4022) loss 3.3255 (3.3255) grad_norm 2.5513 (2.5513) [2021-04-16 15:14:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][10/1251] eta 0:07:43 lr 0.000108 time 0.2702 (0.3732) loss 3.1319 (3.2710) grad_norm 2.5268 (2.6795) [2021-04-16 15:14:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][20/1251] eta 0:06:43 lr 0.000108 time 0.2902 (0.3275) loss 3.1001 (3.2405) grad_norm 2.5177 (2.6486) [2021-04-16 15:14:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][30/1251] eta 0:06:20 lr 0.000108 time 0.2615 (0.3118) loss 1.8258 (3.1959) grad_norm 2.6187 (2.6424) [2021-04-16 15:14:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][40/1251] eta 0:06:07 lr 0.000108 time 0.2660 (0.3036) loss 3.6849 (3.1597) grad_norm 2.6974 (2.6245) [2021-04-16 15:14:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][50/1251] eta 0:05:59 lr 0.000107 time 0.2716 (0.2992) loss 2.9546 (3.2066) grad_norm 2.2163 (2.5777) [2021-04-16 15:14:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][60/1251] eta 0:05:52 lr 0.000107 time 0.2925 (0.2957) loss 3.6282 (3.2199) grad_norm 2.7500 (2.5878) [2021-04-16 15:14:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][70/1251] eta 0:05:45 lr 0.000107 time 0.2848 (0.2930) loss 2.9745 (3.1862) grad_norm 3.7924 (2.6020) [2021-04-16 15:14:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][80/1251] eta 0:05:41 lr 0.000107 time 0.2974 (0.2915) loss 3.6965 (3.2074) grad_norm 2.4000 (2.5985) [2021-04-16 15:14:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][90/1251] eta 0:05:39 lr 0.000107 time 0.2970 (0.2923) loss 3.8401 (3.1871) grad_norm 2.4719 (2.6103) [2021-04-16 15:14:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][100/1251] eta 0:05:36 lr 0.000107 time 0.2609 (0.2924) loss 3.2857 (3.1936) grad_norm 2.2632 (2.6083) [2021-04-16 15:14:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][110/1251] eta 0:05:31 lr 0.000107 time 0.2884 (0.2909) loss 2.1144 (3.1931) grad_norm 2.2320 (2.6012) [2021-04-16 15:14:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][120/1251] eta 0:05:27 lr 0.000107 time 0.2676 (0.2899) loss 1.9811 (3.1981) grad_norm 2.5455 (2.5896) [2021-04-16 15:14:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][130/1251] eta 0:05:23 lr 0.000107 time 0.2723 (0.2890) loss 3.1730 (3.1820) grad_norm 2.7750 (2.5847) [2021-04-16 15:14:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][140/1251] eta 0:05:21 lr 0.000107 time 0.2642 (0.2898) loss 2.7937 (3.1705) grad_norm 2.5649 (2.5940) [2021-04-16 15:14:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][150/1251] eta 0:05:18 lr 0.000107 time 0.3088 (0.2895) loss 1.9660 (3.1595) grad_norm 2.1607 (2.5929) [2021-04-16 15:14:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][160/1251] eta 0:05:14 lr 0.000107 time 0.2939 (0.2884) loss 2.8653 (3.1274) grad_norm 2.2788 (2.5840) [2021-04-16 15:14:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][170/1251] eta 0:05:11 lr 0.000107 time 0.2951 (0.2881) loss 3.2554 (3.1163) grad_norm 2.1928 (2.5812) [2021-04-16 15:14:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][180/1251] eta 0:05:08 lr 0.000107 time 0.2491 (0.2880) loss 2.4419 (3.0993) grad_norm 2.4524 (2.6240) [2021-04-16 15:14:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][190/1251] eta 0:05:05 lr 0.000107 time 0.2709 (0.2875) loss 3.2580 (3.1063) grad_norm 2.2566 (2.6250) [2021-04-16 15:15:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][200/1251] eta 0:05:01 lr 0.000107 time 0.2748 (0.2873) loss 3.0909 (3.1043) grad_norm 2.5238 (2.6197) [2021-04-16 15:15:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][210/1251] eta 0:04:58 lr 0.000107 time 0.2621 (0.2866) loss 3.1315 (3.1094) grad_norm 2.1685 (2.6154) [2021-04-16 15:15:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][220/1251] eta 0:04:55 lr 0.000107 time 0.2675 (0.2865) loss 2.0956 (3.0997) grad_norm 2.7607 (2.6142) [2021-04-16 15:15:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][230/1251] eta 0:04:52 lr 0.000107 time 0.2845 (0.2862) loss 2.9024 (3.0996) grad_norm 2.7087 (2.6177) [2021-04-16 15:15:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][240/1251] eta 0:04:49 lr 0.000107 time 0.3227 (0.2862) loss 2.2347 (3.1062) grad_norm 2.2972 (2.6157) [2021-04-16 15:15:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][250/1251] eta 0:04:46 lr 0.000107 time 0.2792 (0.2860) loss 3.2249 (3.1069) grad_norm 2.7517 (2.6212) [2021-04-16 15:15:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][260/1251] eta 0:04:42 lr 0.000107 time 0.3087 (0.2855) loss 1.8605 (3.1010) grad_norm 2.3058 (2.6183) [2021-04-16 15:15:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][270/1251] eta 0:04:39 lr 0.000107 time 0.2555 (0.2850) loss 2.5591 (3.1075) grad_norm 2.3471 (2.6124) [2021-04-16 15:15:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][280/1251] eta 0:04:36 lr 0.000107 time 0.2978 (0.2848) loss 3.4270 (3.1155) grad_norm 2.3942 (2.6146) [2021-04-16 15:15:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][290/1251] eta 0:04:34 lr 0.000107 time 0.2707 (0.2852) loss 3.6717 (3.1248) grad_norm 2.9690 (2.6196) [2021-04-16 15:15:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][300/1251] eta 0:04:30 lr 0.000107 time 0.2496 (0.2849) loss 3.7355 (3.1236) grad_norm 2.2798 (2.6187) [2021-04-16 15:15:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][310/1251] eta 0:04:27 lr 0.000107 time 0.2651 (0.2846) loss 3.1621 (3.1264) grad_norm 2.2556 (2.6149) [2021-04-16 15:15:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][320/1251] eta 0:04:25 lr 0.000107 time 0.2740 (0.2848) loss 3.6850 (3.1417) grad_norm 2.6110 (2.6141) [2021-04-16 15:15:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][330/1251] eta 0:04:22 lr 0.000107 time 0.2446 (0.2846) loss 3.9540 (3.1428) grad_norm 2.8486 (2.6179) [2021-04-16 15:15:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][340/1251] eta 0:04:19 lr 0.000107 time 0.2871 (0.2850) loss 3.5266 (3.1413) grad_norm 2.4601 (2.6163) [2021-04-16 15:15:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][350/1251] eta 0:04:16 lr 0.000107 time 0.2581 (0.2846) loss 2.5356 (3.1367) grad_norm 2.6482 (2.6127) [2021-04-16 15:15:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][360/1251] eta 0:04:13 lr 0.000107 time 0.2873 (0.2848) loss 3.2838 (3.1387) grad_norm 2.6832 (2.6135) [2021-04-16 15:15:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][370/1251] eta 0:04:10 lr 0.000107 time 0.2663 (0.2849) loss 2.7443 (3.1397) grad_norm 2.7639 (2.6201) [2021-04-16 15:15:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][380/1251] eta 0:04:08 lr 0.000107 time 0.2751 (0.2854) loss 2.4077 (3.1351) grad_norm 2.4422 (2.6219) [2021-04-16 15:15:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][390/1251] eta 0:04:05 lr 0.000107 time 0.2700 (0.2853) loss 3.1330 (3.1380) grad_norm 2.4809 (2.6280) [2021-04-16 15:15:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][400/1251] eta 0:04:02 lr 0.000107 time 0.2817 (0.2850) loss 3.2346 (3.1376) grad_norm 2.9909 (2.6316) [2021-04-16 15:16:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][410/1251] eta 0:03:59 lr 0.000107 time 0.2761 (0.2848) loss 2.9406 (3.1372) grad_norm 2.8920 (2.6309) [2021-04-16 15:16:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][420/1251] eta 0:03:56 lr 0.000107 time 0.2934 (0.2846) loss 2.7605 (3.1369) grad_norm 2.1631 (2.6283) [2021-04-16 15:16:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][430/1251] eta 0:03:53 lr 0.000107 time 0.3078 (0.2846) loss 2.8534 (3.1320) grad_norm 2.3459 (2.6277) [2021-04-16 15:16:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][440/1251] eta 0:03:50 lr 0.000107 time 0.2735 (0.2844) loss 3.2675 (3.1300) grad_norm 2.4220 (2.6247) [2021-04-16 15:16:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][450/1251] eta 0:03:47 lr 0.000106 time 0.2568 (0.2844) loss 4.0101 (3.1397) grad_norm 2.5744 (2.6277) [2021-04-16 15:16:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][460/1251] eta 0:03:44 lr 0.000106 time 0.2709 (0.2841) loss 3.8210 (3.1440) grad_norm 2.5600 (2.6245) [2021-04-16 15:16:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][470/1251] eta 0:03:41 lr 0.000106 time 0.3081 (0.2841) loss 3.4534 (3.1477) grad_norm 2.2484 (2.6244) [2021-04-16 15:16:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][480/1251] eta 0:03:38 lr 0.000106 time 0.2806 (0.2840) loss 3.3668 (3.1451) grad_norm 2.2218 (2.6253) [2021-04-16 15:16:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][490/1251] eta 0:03:36 lr 0.000106 time 0.2683 (0.2841) loss 2.8896 (3.1439) grad_norm 2.3079 (2.6209) [2021-04-16 15:16:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][500/1251] eta 0:03:33 lr 0.000106 time 0.2621 (0.2839) loss 2.8873 (3.1453) grad_norm 2.3336 (2.6181) [2021-04-16 15:16:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][510/1251] eta 0:03:30 lr 0.000106 time 0.2775 (0.2840) loss 2.1747 (3.1444) grad_norm 2.4526 (2.6149) [2021-04-16 15:16:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][520/1251] eta 0:03:27 lr 0.000106 time 0.2922 (0.2839) loss 3.3375 (3.1418) grad_norm 2.2141 (2.6132) [2021-04-16 15:16:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][530/1251] eta 0:03:24 lr 0.000106 time 0.2834 (0.2838) loss 3.1717 (3.1425) grad_norm 2.5684 (2.6164) [2021-04-16 15:16:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][540/1251] eta 0:03:21 lr 0.000106 time 0.2660 (0.2837) loss 3.2671 (3.1456) grad_norm 2.3692 (2.6140) [2021-04-16 15:16:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][550/1251] eta 0:03:18 lr 0.000106 time 0.2583 (0.2838) loss 2.2715 (3.1461) grad_norm 2.4353 (2.6143) [2021-04-16 15:16:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][560/1251] eta 0:03:15 lr 0.000106 time 0.2667 (0.2836) loss 2.6732 (3.1438) grad_norm 2.2516 (2.6104) [2021-04-16 15:16:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][570/1251] eta 0:03:13 lr 0.000106 time 0.2701 (0.2836) loss 3.5185 (3.1471) grad_norm 2.5948 (2.6112) [2021-04-16 15:16:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][580/1251] eta 0:03:10 lr 0.000106 time 0.2827 (0.2836) loss 2.9183 (3.1433) grad_norm 2.8800 (2.6125) [2021-04-16 15:16:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][590/1251] eta 0:03:07 lr 0.000106 time 0.2889 (0.2835) loss 3.3177 (3.1456) grad_norm 2.5987 (2.6142) [2021-04-16 15:16:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][600/1251] eta 0:03:04 lr 0.000106 time 0.2562 (0.2833) loss 3.7194 (3.1474) grad_norm 2.5211 (2.6152) [2021-04-16 15:16:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][610/1251] eta 0:03:01 lr 0.000106 time 0.2764 (0.2831) loss 3.5095 (3.1472) grad_norm 2.9711 (2.6156) [2021-04-16 15:16:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][620/1251] eta 0:02:58 lr 0.000106 time 0.2612 (0.2831) loss 3.4201 (3.1382) grad_norm 2.2719 (2.6152) [2021-04-16 15:17:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][630/1251] eta 0:02:55 lr 0.000106 time 0.2881 (0.2830) loss 3.0762 (3.1409) grad_norm 2.2272 (2.6139) [2021-04-16 15:17:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][640/1251] eta 0:02:52 lr 0.000106 time 0.2950 (0.2829) loss 2.2931 (3.1378) grad_norm 2.3129 (2.6161) [2021-04-16 15:17:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][650/1251] eta 0:02:49 lr 0.000106 time 0.2823 (0.2827) loss 3.0654 (3.1388) grad_norm 2.1154 (2.6146) [2021-04-16 15:17:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][660/1251] eta 0:02:47 lr 0.000106 time 0.2733 (0.2826) loss 3.6275 (3.1397) grad_norm 2.9630 (2.6124) [2021-04-16 15:17:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][670/1251] eta 0:02:44 lr 0.000106 time 0.2805 (0.2828) loss 3.4832 (3.1438) grad_norm 2.4302 (2.6103) [2021-04-16 15:17:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][680/1251] eta 0:02:41 lr 0.000106 time 0.2837 (0.2828) loss 3.8871 (3.1465) grad_norm 2.2581 (2.6106) [2021-04-16 15:17:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][690/1251] eta 0:02:38 lr 0.000106 time 0.2894 (0.2827) loss 3.6775 (3.1471) grad_norm 2.2059 (2.6113) [2021-04-16 15:17:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][700/1251] eta 0:02:35 lr 0.000106 time 0.2703 (0.2826) loss 3.3719 (3.1464) grad_norm 2.5682 (2.6096) [2021-04-16 15:17:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][710/1251] eta 0:02:32 lr 0.000106 time 0.2525 (0.2826) loss 2.6629 (3.1462) grad_norm 2.5464 (2.6081) [2021-04-16 15:17:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][720/1251] eta 0:02:30 lr 0.000106 time 0.2785 (0.2827) loss 3.4453 (3.1462) grad_norm 2.3554 (2.6083) [2021-04-16 15:17:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][730/1251] eta 0:02:27 lr 0.000106 time 0.2977 (0.2827) loss 3.3943 (3.1449) grad_norm 2.6932 (2.6086) [2021-04-16 15:17:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][740/1251] eta 0:02:24 lr 0.000106 time 0.2556 (0.2827) loss 2.4613 (3.1457) grad_norm 2.4119 (2.6105) [2021-04-16 15:17:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][750/1251] eta 0:02:21 lr 0.000106 time 0.2787 (0.2826) loss 2.9964 (3.1421) grad_norm 3.0343 (2.6091) [2021-04-16 15:17:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][760/1251] eta 0:02:18 lr 0.000106 time 0.3059 (0.2827) loss 3.3015 (3.1388) grad_norm 2.5980 (2.6101) [2021-04-16 15:17:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][770/1251] eta 0:02:15 lr 0.000106 time 0.2725 (0.2827) loss 3.3538 (3.1374) grad_norm 2.7692 (2.6084) [2021-04-16 15:17:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][780/1251] eta 0:02:13 lr 0.000106 time 0.2846 (0.2826) loss 3.3624 (3.1366) grad_norm 2.4994 (2.6066) [2021-04-16 15:17:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][790/1251] eta 0:02:10 lr 0.000106 time 0.2451 (0.2825) loss 2.6666 (3.1337) grad_norm 2.3449 (2.6071) [2021-04-16 15:17:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][800/1251] eta 0:02:07 lr 0.000106 time 0.2666 (0.2824) loss 2.5605 (3.1316) grad_norm 2.2067 (2.6045) [2021-04-16 15:17:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][810/1251] eta 0:02:04 lr 0.000106 time 0.2927 (0.2825) loss 3.2663 (3.1320) grad_norm 2.3955 (2.6030) [2021-04-16 15:17:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][820/1251] eta 0:02:01 lr 0.000106 time 0.2745 (0.2825) loss 2.8854 (3.1311) grad_norm 2.3130 (2.6029) [2021-04-16 15:17:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][830/1251] eta 0:01:58 lr 0.000106 time 0.2777 (0.2824) loss 3.7175 (3.1341) grad_norm 2.6240 (2.6011) [2021-04-16 15:18:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][840/1251] eta 0:01:56 lr 0.000106 time 0.2816 (0.2824) loss 2.1458 (3.1339) grad_norm 2.7913 (2.6008) [2021-04-16 15:18:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][850/1251] eta 0:01:53 lr 0.000106 time 0.2829 (0.2823) loss 3.7815 (3.1353) grad_norm 2.7147 (2.6013) [2021-04-16 15:18:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][860/1251] eta 0:01:50 lr 0.000105 time 0.2631 (0.2823) loss 2.9235 (3.1340) grad_norm 2.5135 (2.5999) [2021-04-16 15:18:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][870/1251] eta 0:01:47 lr 0.000105 time 0.2774 (0.2824) loss 2.3980 (3.1350) grad_norm 2.1003 (2.5986) [2021-04-16 15:18:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][880/1251] eta 0:01:44 lr 0.000105 time 0.2811 (0.2825) loss 3.3031 (3.1366) grad_norm 2.2343 (2.5995) [2021-04-16 15:18:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][890/1251] eta 0:01:41 lr 0.000105 time 0.2670 (0.2825) loss 3.4349 (3.1375) grad_norm 2.2106 (2.5992) [2021-04-16 15:18:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][900/1251] eta 0:01:39 lr 0.000105 time 0.2906 (0.2825) loss 3.3562 (3.1397) grad_norm 2.3656 (2.5985) [2021-04-16 15:18:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][910/1251] eta 0:01:36 lr 0.000105 time 0.2588 (0.2824) loss 3.6302 (3.1364) grad_norm 2.5562 (2.5987) [2021-04-16 15:18:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][920/1251] eta 0:01:33 lr 0.000105 time 0.2670 (0.2826) loss 3.1611 (3.1375) grad_norm 2.3371 (2.5995) [2021-04-16 15:18:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][930/1251] eta 0:01:30 lr 0.000105 time 0.2890 (0.2825) loss 2.3749 (3.1352) grad_norm 2.5060 (2.5994) [2021-04-16 15:18:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][940/1251] eta 0:01:27 lr 0.000105 time 0.2746 (0.2828) loss 2.1851 (3.1346) grad_norm 2.5082 (2.5983) [2021-04-16 15:18:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][950/1251] eta 0:01:25 lr 0.000105 time 0.3910 (0.2828) loss 3.8028 (3.1368) grad_norm 2.4352 (2.5983) [2021-04-16 15:18:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][960/1251] eta 0:01:22 lr 0.000105 time 0.3002 (0.2828) loss 3.6122 (3.1393) grad_norm 2.2149 (2.5998) [2021-04-16 15:18:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][970/1251] eta 0:01:19 lr 0.000105 time 0.2828 (0.2828) loss 3.4735 (3.1397) grad_norm 2.3780 (2.6013) [2021-04-16 15:18:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][980/1251] eta 0:01:16 lr 0.000105 time 0.2452 (0.2827) loss 3.1217 (3.1389) grad_norm 2.9598 (2.6021) [2021-04-16 15:18:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][990/1251] eta 0:01:13 lr 0.000105 time 0.2539 (0.2826) loss 3.6219 (3.1409) grad_norm 2.3710 (2.6019) [2021-04-16 15:18:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][1000/1251] eta 0:01:10 lr 0.000105 time 0.2635 (0.2825) loss 3.4723 (3.1395) grad_norm 2.2921 (2.6024) [2021-04-16 15:18:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][1010/1251] eta 0:01:08 lr 0.000105 time 0.2715 (0.2825) loss 3.5748 (3.1379) grad_norm 2.2609 (2.6003) [2021-04-16 15:18:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][1020/1251] eta 0:01:05 lr 0.000105 time 0.2933 (0.2825) loss 3.5219 (3.1380) grad_norm 2.5415 (2.5998) [2021-04-16 15:18:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][1030/1251] eta 0:01:02 lr 0.000105 time 0.2563 (0.2824) loss 3.3018 (3.1371) grad_norm 2.2634 (2.5989) [2021-04-16 15:18:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][1040/1251] eta 0:00:59 lr 0.000105 time 0.2689 (0.2824) loss 3.5949 (3.1371) grad_norm 2.2889 (2.5989) [2021-04-16 15:19:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][1050/1251] eta 0:00:56 lr 0.000105 time 0.2875 (0.2823) loss 3.6842 (3.1382) grad_norm 2.4149 (2.5994) [2021-04-16 15:19:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][1060/1251] eta 0:00:53 lr 0.000105 time 0.2967 (0.2823) loss 3.3134 (3.1387) grad_norm 3.0769 (2.6013) [2021-04-16 15:19:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][1070/1251] eta 0:00:51 lr 0.000105 time 0.2542 (0.2824) loss 2.8577 (3.1358) grad_norm 2.8543 (2.6024) [2021-04-16 15:19:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][1080/1251] eta 0:00:48 lr 0.000105 time 0.3093 (0.2823) loss 3.1854 (3.1367) grad_norm 2.9783 (2.6040) [2021-04-16 15:19:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][1090/1251] eta 0:00:45 lr 0.000105 time 0.2948 (0.2822) loss 2.3323 (3.1351) grad_norm 3.3437 (2.6028) [2021-04-16 15:19:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][1100/1251] eta 0:00:42 lr 0.000105 time 0.2784 (0.2822) loss 3.1656 (3.1361) grad_norm 2.9186 (2.6029) [2021-04-16 15:19:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][1110/1251] eta 0:00:39 lr 0.000105 time 0.2899 (0.2822) loss 3.4143 (3.1350) grad_norm 2.4635 (2.6028) [2021-04-16 15:19:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][1120/1251] eta 0:00:36 lr 0.000105 time 0.2806 (0.2822) loss 3.7812 (3.1324) grad_norm 2.2598 (2.6019) [2021-04-16 15:19:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][1130/1251] eta 0:00:34 lr 0.000105 time 0.2606 (0.2823) loss 3.1293 (3.1325) grad_norm 2.4303 (2.6013) [2021-04-16 15:19:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][1140/1251] eta 0:00:31 lr 0.000105 time 0.2784 (0.2822) loss 2.5264 (3.1333) grad_norm 2.6264 (2.6025) [2021-04-16 15:19:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][1150/1251] eta 0:00:28 lr 0.000105 time 0.2812 (0.2823) loss 3.1530 (3.1343) grad_norm 3.0690 (2.6048) [2021-04-16 15:19:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][1160/1251] eta 0:00:25 lr 0.000105 time 0.2833 (0.2824) loss 4.0336 (3.1350) grad_norm 2.8551 (2.6057) [2021-04-16 15:19:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][1170/1251] eta 0:00:22 lr 0.000105 time 0.2721 (0.2823) loss 3.3225 (3.1361) grad_norm 2.9310 (2.6047) [2021-04-16 15:19:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][1180/1251] eta 0:00:20 lr 0.000105 time 0.2754 (0.2823) loss 2.3642 (3.1366) grad_norm 2.7659 (2.6044) [2021-04-16 15:19:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][1190/1251] eta 0:00:17 lr 0.000105 time 0.2733 (0.2822) loss 3.4688 (3.1356) grad_norm 2.5357 (2.6052) [2021-04-16 15:19:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][1200/1251] eta 0:00:14 lr 0.000105 time 0.2977 (0.2822) loss 3.6478 (3.1367) grad_norm 2.2652 (2.6039) [2021-04-16 15:19:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][1210/1251] eta 0:00:11 lr 0.000105 time 0.2700 (0.2821) loss 3.1781 (3.1377) grad_norm 2.4813 (2.6033) [2021-04-16 15:19:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][1220/1251] eta 0:00:08 lr 0.000105 time 0.2743 (0.2821) loss 2.6038 (3.1364) grad_norm 2.5572 (2.6040) [2021-04-16 15:19:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][1230/1251] eta 0:00:05 lr 0.000105 time 0.2941 (0.2821) loss 3.2671 (3.1350) grad_norm 3.1814 (2.6067) [2021-04-16 15:19:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][1240/1251] eta 0:00:03 lr 0.000105 time 0.2481 (0.2820) loss 3.0409 (3.1359) grad_norm 2.4539 (2.6066) [2021-04-16 15:19:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [239/300][1250/1251] eta 0:00:00 lr 0.000105 time 0.2480 (0.2817) loss 2.4476 (3.1342) grad_norm 2.5363 (2.6088) [2021-04-16 15:20:08 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 239 training takes 0:06:05 [2021-04-16 15:20:08 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_239.pth saving...... [2021-04-16 15:20:29 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_239.pth saved !!! [2021-04-16 15:20:31 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.166 (1.166) Loss 0.8556 (0.8556) Acc@1 80.664 (80.664) Acc@5 95.508 (95.508) [2021-04-16 15:20:32 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.180 (0.222) Loss 0.7931 (0.8404) Acc@1 80.762 (80.433) Acc@5 96.191 (95.206) [2021-04-16 15:20:35 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.123 (0.250) Loss 0.9067 (0.8451) Acc@1 78.809 (80.069) Acc@5 94.043 (95.089) [2021-04-16 15:20:36 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.128 (0.220) Loss 0.8362 (0.8464) Acc@1 81.152 (79.974) Acc@5 95.117 (95.089) [2021-04-16 15:20:38 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.544 (0.217) Loss 0.8462 (0.8480) Acc@1 79.590 (79.959) Acc@5 95.312 (95.062) [2021-04-16 15:20:52 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.006 Acc@5 95.066 [2021-04-16 15:20:52 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 80.0% [2021-04-16 15:20:52 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 80.01% [2021-04-16 15:20:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][0/1251] eta 2:35:28 lr 0.000105 time 7.4566 (7.4566) loss 2.0335 (2.0335) grad_norm 2.4008 (2.4008) [2021-04-16 15:21:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][10/1251] eta 0:19:06 lr 0.000105 time 0.2862 (0.9236) loss 2.9215 (2.8029) grad_norm 2.7657 (2.5591) [2021-04-16 15:21:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][20/1251] eta 0:12:36 lr 0.000104 time 0.2593 (0.6147) loss 3.2273 (2.8222) grad_norm 2.9458 (2.8063) [2021-04-16 15:21:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][30/1251] eta 0:10:16 lr 0.000104 time 0.2711 (0.5049) loss 3.6833 (2.8291) grad_norm 2.6447 (2.8140) [2021-04-16 15:21:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][40/1251] eta 0:09:03 lr 0.000104 time 0.2491 (0.4489) loss 3.2306 (2.8143) grad_norm 2.3356 (2.7392) [2021-04-16 15:21:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][50/1251] eta 0:08:19 lr 0.000104 time 0.3000 (0.4159) loss 2.8411 (2.8594) grad_norm 2.4460 (2.6931) [2021-04-16 15:21:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][60/1251] eta 0:07:50 lr 0.000104 time 0.4681 (0.3954) loss 2.3535 (2.9460) grad_norm 2.4209 (2.6767) [2021-04-16 15:21:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][70/1251] eta 0:07:26 lr 0.000104 time 0.2768 (0.3782) loss 2.6694 (2.9241) grad_norm 3.0124 (2.6497) [2021-04-16 15:21:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][80/1251] eta 0:07:09 lr 0.000104 time 0.2718 (0.3666) loss 2.7788 (2.9259) grad_norm 2.7160 (2.6505) [2021-04-16 15:21:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][90/1251] eta 0:06:54 lr 0.000104 time 0.2736 (0.3570) loss 3.4158 (2.9598) grad_norm 2.8896 (2.6665) [2021-04-16 15:21:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][100/1251] eta 0:06:42 lr 0.000104 time 0.3015 (0.3500) loss 2.9053 (2.9867) grad_norm 2.8133 (2.6558) [2021-04-16 15:21:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][110/1251] eta 0:06:31 lr 0.000104 time 0.2975 (0.3433) loss 3.0454 (3.0041) grad_norm 2.6583 (2.6568) [2021-04-16 15:21:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][120/1251] eta 0:06:22 lr 0.000104 time 0.2621 (0.3383) loss 2.9173 (3.0163) grad_norm 2.2016 (2.6378) [2021-04-16 15:21:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][130/1251] eta 0:06:13 lr 0.000104 time 0.2807 (0.3335) loss 3.1271 (3.0048) grad_norm 2.6123 (2.6599) [2021-04-16 15:21:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][140/1251] eta 0:06:07 lr 0.000104 time 0.2916 (0.3312) loss 3.1532 (3.0037) grad_norm 2.2281 (2.6535) [2021-04-16 15:21:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][150/1251] eta 0:06:02 lr 0.000104 time 0.2753 (0.3293) loss 3.3349 (3.0109) grad_norm 2.9922 (2.6570) [2021-04-16 15:21:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][160/1251] eta 0:05:55 lr 0.000104 time 0.2567 (0.3261) loss 2.7614 (3.0101) grad_norm 2.3811 (2.6677) [2021-04-16 15:21:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][170/1251] eta 0:05:50 lr 0.000104 time 0.2882 (0.3239) loss 2.7070 (3.0207) grad_norm 2.4469 (2.6854) [2021-04-16 15:21:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][180/1251] eta 0:05:44 lr 0.000104 time 0.2620 (0.3213) loss 3.4110 (3.0123) grad_norm 2.8381 (2.6768) [2021-04-16 15:21:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][190/1251] eta 0:05:38 lr 0.000104 time 0.2946 (0.3189) loss 4.0085 (3.0175) grad_norm 2.8023 (2.6737) [2021-04-16 15:21:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][200/1251] eta 0:05:32 lr 0.000104 time 0.3023 (0.3168) loss 3.0313 (3.0225) grad_norm 2.2108 (2.6711) [2021-04-16 15:21:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][210/1251] eta 0:05:27 lr 0.000104 time 0.2774 (0.3148) loss 3.1709 (3.0271) grad_norm 3.2492 (2.6797) [2021-04-16 15:22:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][220/1251] eta 0:05:22 lr 0.000104 time 0.2674 (0.3131) loss 2.1775 (3.0259) grad_norm 2.7022 (2.6867) [2021-04-16 15:22:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][230/1251] eta 0:05:17 lr 0.000104 time 0.2934 (0.3114) loss 3.7442 (3.0346) grad_norm 2.3169 (2.6792) [2021-04-16 15:22:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][240/1251] eta 0:05:13 lr 0.000104 time 0.2573 (0.3097) loss 2.9090 (3.0425) grad_norm 2.5891 (2.6877) [2021-04-16 15:22:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][250/1251] eta 0:05:08 lr 0.000104 time 0.2818 (0.3083) loss 3.4926 (3.0478) grad_norm 3.1606 (2.6942) [2021-04-16 15:22:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][260/1251] eta 0:05:04 lr 0.000104 time 0.2691 (0.3070) loss 2.6154 (3.0499) grad_norm 2.1388 (2.6927) [2021-04-16 15:22:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][270/1251] eta 0:05:00 lr 0.000104 time 0.2674 (0.3058) loss 3.0032 (3.0488) grad_norm 2.6352 (2.6890) [2021-04-16 15:22:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][280/1251] eta 0:04:55 lr 0.000104 time 0.2682 (0.3048) loss 3.6199 (3.0448) grad_norm 2.4958 (2.6826) [2021-04-16 15:22:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][290/1251] eta 0:04:52 lr 0.000104 time 0.2764 (0.3045) loss 2.9372 (3.0449) grad_norm 2.7141 (2.6765) [2021-04-16 15:22:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][300/1251] eta 0:04:48 lr 0.000104 time 0.2731 (0.3035) loss 2.3919 (3.0422) grad_norm 2.3718 (2.6804) [2021-04-16 15:22:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][310/1251] eta 0:04:45 lr 0.000104 time 0.2703 (0.3031) loss 3.7395 (3.0321) grad_norm 2.7537 (2.6798) [2021-04-16 15:22:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][320/1251] eta 0:04:41 lr 0.000104 time 0.2806 (0.3022) loss 2.8946 (3.0301) grad_norm 2.4432 (2.6736) [2021-04-16 15:22:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][330/1251] eta 0:04:37 lr 0.000104 time 0.2610 (0.3017) loss 3.5237 (3.0261) grad_norm 2.8535 (2.6779) [2021-04-16 15:22:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][340/1251] eta 0:04:34 lr 0.000104 time 0.2691 (0.3010) loss 2.7107 (3.0166) grad_norm 2.7019 (2.6775) [2021-04-16 15:22:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][350/1251] eta 0:04:30 lr 0.000104 time 0.2737 (0.3003) loss 3.2990 (3.0177) grad_norm 2.6226 (2.6729) [2021-04-16 15:22:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][360/1251] eta 0:04:26 lr 0.000104 time 0.2721 (0.2996) loss 3.3846 (3.0145) grad_norm 2.2354 (2.6730) [2021-04-16 15:22:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][370/1251] eta 0:04:23 lr 0.000104 time 0.2532 (0.2990) loss 3.2263 (3.0131) grad_norm 4.1403 (2.6779) [2021-04-16 15:22:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][380/1251] eta 0:04:19 lr 0.000104 time 0.3279 (0.2985) loss 2.6088 (3.0140) grad_norm 2.7970 (2.6758) [2021-04-16 15:22:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][390/1251] eta 0:04:16 lr 0.000104 time 0.2605 (0.2980) loss 2.9696 (3.0190) grad_norm 2.2977 (2.6732) [2021-04-16 15:22:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][400/1251] eta 0:04:13 lr 0.000104 time 0.2711 (0.2975) loss 3.2291 (3.0176) grad_norm 2.2768 (2.6681) [2021-04-16 15:22:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][410/1251] eta 0:04:09 lr 0.000104 time 0.2723 (0.2972) loss 3.0317 (3.0214) grad_norm 2.5989 (2.6660) [2021-04-16 15:22:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][420/1251] eta 0:04:06 lr 0.000104 time 0.2625 (0.2968) loss 3.4078 (3.0288) grad_norm 2.3024 (2.6673) [2021-04-16 15:22:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][430/1251] eta 0:04:03 lr 0.000103 time 0.2946 (0.2962) loss 2.3311 (3.0329) grad_norm 2.4553 (2.6687) [2021-04-16 15:23:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][440/1251] eta 0:04:00 lr 0.000103 time 0.2679 (0.2961) loss 2.3967 (3.0396) grad_norm 2.6352 (2.6654) [2021-04-16 15:23:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][450/1251] eta 0:03:56 lr 0.000103 time 0.2880 (0.2956) loss 3.6135 (3.0400) grad_norm 2.3652 (2.6638) [2021-04-16 15:23:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][460/1251] eta 0:03:53 lr 0.000103 time 0.2837 (0.2952) loss 3.0698 (3.0428) grad_norm 2.4775 (2.6688) [2021-04-16 15:23:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][470/1251] eta 0:03:50 lr 0.000103 time 0.2687 (0.2951) loss 3.1290 (3.0405) grad_norm 2.6043 (2.6662) [2021-04-16 15:23:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][480/1251] eta 0:03:47 lr 0.000103 time 0.2496 (0.2947) loss 3.7878 (3.0441) grad_norm 2.2018 (2.6664) [2021-04-16 15:23:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][490/1251] eta 0:03:44 lr 0.000103 time 0.2830 (0.2945) loss 3.1634 (3.0444) grad_norm 2.6147 (2.6645) [2021-04-16 15:23:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][500/1251] eta 0:03:40 lr 0.000103 time 0.2671 (0.2941) loss 3.1686 (3.0427) grad_norm 2.3435 (2.6643) [2021-04-16 15:23:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][510/1251] eta 0:03:37 lr 0.000103 time 0.3001 (0.2937) loss 3.3785 (3.0408) grad_norm 2.7248 (2.6654) [2021-04-16 15:23:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][520/1251] eta 0:03:34 lr 0.000103 time 0.2890 (0.2935) loss 3.6309 (3.0444) grad_norm 2.4061 (2.6632) [2021-04-16 15:23:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][530/1251] eta 0:03:31 lr 0.000103 time 0.2905 (0.2932) loss 3.5614 (3.0456) grad_norm 2.3723 (2.6605) [2021-04-16 15:23:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][540/1251] eta 0:03:28 lr 0.000103 time 0.2647 (0.2929) loss 3.3811 (3.0457) grad_norm 2.4501 (2.6615) [2021-04-16 15:23:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][550/1251] eta 0:03:25 lr 0.000103 time 0.2732 (0.2926) loss 3.5616 (3.0519) grad_norm 2.2054 (2.6597) [2021-04-16 15:23:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][560/1251] eta 0:03:21 lr 0.000103 time 0.2890 (0.2923) loss 2.2087 (3.0506) grad_norm 2.5046 (2.6600) [2021-04-16 15:23:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][570/1251] eta 0:03:18 lr 0.000103 time 0.2876 (0.2921) loss 3.5603 (3.0547) grad_norm 2.7421 (2.6583) [2021-04-16 15:23:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][580/1251] eta 0:03:15 lr 0.000103 time 0.2821 (0.2920) loss 3.2075 (3.0537) grad_norm 2.6939 (inf) [2021-04-16 15:23:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][590/1251] eta 0:03:12 lr 0.000103 time 0.3146 (0.2918) loss 2.3037 (3.0520) grad_norm 2.7673 (inf) [2021-04-16 15:23:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][600/1251] eta 0:03:09 lr 0.000103 time 0.2605 (0.2915) loss 2.3372 (3.0515) grad_norm 2.4019 (inf) [2021-04-16 15:23:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][610/1251] eta 0:03:06 lr 0.000103 time 0.2731 (0.2913) loss 3.1572 (3.0543) grad_norm 2.3016 (inf) [2021-04-16 15:23:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][620/1251] eta 0:03:03 lr 0.000103 time 0.2509 (0.2910) loss 3.4418 (3.0530) grad_norm 2.7245 (inf) [2021-04-16 15:23:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][630/1251] eta 0:03:00 lr 0.000103 time 0.2862 (0.2908) loss 2.8393 (3.0530) grad_norm 2.4053 (inf) [2021-04-16 15:23:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][640/1251] eta 0:02:57 lr 0.000103 time 0.2583 (0.2908) loss 3.7697 (3.0554) grad_norm 2.3489 (inf) [2021-04-16 15:24:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][650/1251] eta 0:02:54 lr 0.000103 time 0.2717 (0.2908) loss 2.7924 (3.0564) grad_norm 2.6161 (inf) [2021-04-16 15:24:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][660/1251] eta 0:02:51 lr 0.000103 time 0.2863 (0.2906) loss 2.5814 (3.0563) grad_norm 2.5447 (inf) [2021-04-16 15:24:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][670/1251] eta 0:02:48 lr 0.000103 time 0.2737 (0.2904) loss 3.2478 (3.0604) grad_norm 2.6220 (inf) [2021-04-16 15:24:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][680/1251] eta 0:02:45 lr 0.000103 time 0.3024 (0.2904) loss 2.6076 (3.0606) grad_norm 2.5262 (inf) [2021-04-16 15:24:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][690/1251] eta 0:02:42 lr 0.000103 time 0.2767 (0.2904) loss 2.0404 (3.0641) grad_norm 2.7283 (inf) [2021-04-16 15:24:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][700/1251] eta 0:02:39 lr 0.000103 time 0.3015 (0.2902) loss 2.3922 (3.0620) grad_norm 3.0941 (inf) [2021-04-16 15:24:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][710/1251] eta 0:02:36 lr 0.000103 time 0.2770 (0.2901) loss 3.4927 (3.0634) grad_norm 2.7475 (inf) [2021-04-16 15:24:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][720/1251] eta 0:02:33 lr 0.000103 time 0.2739 (0.2900) loss 3.7405 (3.0673) grad_norm 3.1404 (inf) [2021-04-16 15:24:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][730/1251] eta 0:02:31 lr 0.000103 time 0.2869 (0.2902) loss 1.9102 (3.0657) grad_norm 2.2358 (inf) [2021-04-16 15:24:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][740/1251] eta 0:02:28 lr 0.000103 time 0.3528 (0.2901) loss 2.8028 (3.0680) grad_norm 2.6954 (inf) [2021-04-16 15:24:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][750/1251] eta 0:02:25 lr 0.000103 time 0.2743 (0.2898) loss 2.2342 (3.0686) grad_norm 2.6208 (inf) [2021-04-16 15:24:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][760/1251] eta 0:02:22 lr 0.000103 time 0.2663 (0.2897) loss 3.4359 (3.0682) grad_norm 2.7895 (inf) [2021-04-16 15:24:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][770/1251] eta 0:02:19 lr 0.000103 time 0.2770 (0.2896) loss 2.0759 (3.0663) grad_norm 2.5886 (inf) [2021-04-16 15:24:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][780/1251] eta 0:02:16 lr 0.000103 time 0.2637 (0.2894) loss 3.3482 (3.0652) grad_norm 2.7844 (inf) [2021-04-16 15:24:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][790/1251] eta 0:02:13 lr 0.000103 time 0.2876 (0.2892) loss 3.8183 (3.0661) grad_norm 2.2236 (inf) [2021-04-16 15:24:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][800/1251] eta 0:02:10 lr 0.000103 time 0.2619 (0.2890) loss 2.9642 (3.0680) grad_norm 3.1472 (inf) [2021-04-16 15:24:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][810/1251] eta 0:02:07 lr 0.000103 time 0.2813 (0.2888) loss 2.8468 (3.0728) grad_norm 2.4271 (inf) [2021-04-16 15:24:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][820/1251] eta 0:02:04 lr 0.000103 time 0.2840 (0.2887) loss 2.4103 (3.0713) grad_norm 3.5070 (inf) [2021-04-16 15:24:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][830/1251] eta 0:02:01 lr 0.000103 time 0.2642 (0.2885) loss 2.7598 (3.0718) grad_norm 2.5834 (inf) [2021-04-16 15:24:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][840/1251] eta 0:01:58 lr 0.000103 time 0.2698 (0.2884) loss 3.7055 (3.0749) grad_norm 2.4598 (inf) [2021-04-16 15:24:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][850/1251] eta 0:01:55 lr 0.000102 time 0.3028 (0.2883) loss 3.4102 (3.0774) grad_norm 2.5715 (inf) [2021-04-16 15:25:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][860/1251] eta 0:01:52 lr 0.000102 time 0.2728 (0.2881) loss 2.8237 (3.0759) grad_norm 2.3395 (inf) [2021-04-16 15:25:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][870/1251] eta 0:01:49 lr 0.000102 time 0.2978 (0.2882) loss 3.0027 (3.0753) grad_norm 2.4178 (inf) [2021-04-16 15:25:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][880/1251] eta 0:01:46 lr 0.000102 time 0.4521 (0.2882) loss 3.7095 (3.0773) grad_norm 2.3814 (inf) [2021-04-16 15:25:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][890/1251] eta 0:01:44 lr 0.000102 time 0.2687 (0.2881) loss 3.5503 (3.0748) grad_norm 2.7818 (inf) [2021-04-16 15:25:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][900/1251] eta 0:01:41 lr 0.000102 time 0.2665 (0.2880) loss 2.5290 (3.0755) grad_norm 2.6753 (inf) [2021-04-16 15:25:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][910/1251] eta 0:01:38 lr 0.000102 time 0.2791 (0.2880) loss 3.3538 (3.0784) grad_norm 2.5406 (inf) [2021-04-16 15:25:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][920/1251] eta 0:01:35 lr 0.000102 time 0.2524 (0.2878) loss 2.8925 (3.0792) grad_norm 2.1423 (inf) [2021-04-16 15:25:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][930/1251] eta 0:01:32 lr 0.000102 time 0.2607 (0.2877) loss 3.2502 (3.0807) grad_norm 2.4423 (inf) [2021-04-16 15:25:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][940/1251] eta 0:01:29 lr 0.000102 time 0.2697 (0.2876) loss 3.5290 (3.0823) grad_norm 2.4105 (inf) [2021-04-16 15:25:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][950/1251] eta 0:01:26 lr 0.000102 time 0.2608 (0.2875) loss 2.5237 (3.0816) grad_norm 2.7327 (inf) [2021-04-16 15:25:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][960/1251] eta 0:01:23 lr 0.000102 time 0.2887 (0.2873) loss 2.8396 (3.0793) grad_norm 2.6100 (inf) [2021-04-16 15:25:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][970/1251] eta 0:01:20 lr 0.000102 time 0.2746 (0.2871) loss 2.6706 (3.0800) grad_norm 2.3970 (inf) [2021-04-16 15:25:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][980/1251] eta 0:01:17 lr 0.000102 time 0.2586 (0.2871) loss 3.7725 (3.0816) grad_norm 2.2578 (inf) [2021-04-16 15:25:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][990/1251] eta 0:01:14 lr 0.000102 time 0.2873 (0.2869) loss 3.5842 (3.0816) grad_norm 2.4018 (inf) [2021-04-16 15:25:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][1000/1251] eta 0:01:11 lr 0.000102 time 0.2955 (0.2868) loss 2.8574 (3.0824) grad_norm 2.2500 (inf) [2021-04-16 15:25:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][1010/1251] eta 0:01:09 lr 0.000102 time 0.2573 (0.2868) loss 3.8290 (3.0844) grad_norm 2.4530 (inf) [2021-04-16 15:25:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][1020/1251] eta 0:01:06 lr 0.000102 time 0.2726 (0.2868) loss 3.5899 (3.0864) grad_norm 2.4032 (inf) [2021-04-16 15:25:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][1030/1251] eta 0:01:03 lr 0.000102 time 0.2949 (0.2867) loss 3.4582 (3.0862) grad_norm 3.1576 (inf) [2021-04-16 15:25:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][1040/1251] eta 0:01:00 lr 0.000102 time 0.2719 (0.2868) loss 2.7060 (3.0868) grad_norm 2.5486 (inf) [2021-04-16 15:25:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][1050/1251] eta 0:00:57 lr 0.000102 time 0.2839 (0.2867) loss 3.0935 (3.0876) grad_norm 2.2240 (inf) [2021-04-16 15:25:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][1060/1251] eta 0:00:54 lr 0.000102 time 0.2935 (0.2866) loss 3.1430 (3.0871) grad_norm 2.8554 (inf) [2021-04-16 15:25:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][1070/1251] eta 0:00:51 lr 0.000102 time 0.2476 (0.2865) loss 3.3148 (3.0883) grad_norm 2.4059 (inf) [2021-04-16 15:26:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][1080/1251] eta 0:00:48 lr 0.000102 time 0.2630 (0.2864) loss 3.6824 (3.0910) grad_norm 3.3348 (inf) [2021-04-16 15:26:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][1090/1251] eta 0:00:46 lr 0.000102 time 0.2759 (0.2863) loss 3.4978 (3.0929) grad_norm 3.3666 (inf) [2021-04-16 15:26:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][1100/1251] eta 0:00:43 lr 0.000102 time 0.2618 (0.2862) loss 3.6597 (3.0950) grad_norm 3.3342 (inf) [2021-04-16 15:26:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][1110/1251] eta 0:00:40 lr 0.000102 time 0.2831 (0.2861) loss 2.0658 (3.0946) grad_norm 2.1973 (inf) [2021-04-16 15:26:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][1120/1251] eta 0:00:37 lr 0.000102 time 0.2675 (0.2860) loss 2.9314 (3.0930) grad_norm 2.3645 (inf) [2021-04-16 15:26:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][1130/1251] eta 0:00:34 lr 0.000102 time 0.2415 (0.2859) loss 3.7273 (3.0930) grad_norm 2.9024 (inf) [2021-04-16 15:26:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][1140/1251] eta 0:00:31 lr 0.000102 time 0.2630 (0.2859) loss 2.7870 (3.0906) grad_norm 2.3477 (inf) [2021-04-16 15:26:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][1150/1251] eta 0:00:28 lr 0.000102 time 0.2979 (0.2858) loss 3.6544 (3.0927) grad_norm 2.5374 (inf) [2021-04-16 15:26:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][1160/1251] eta 0:00:26 lr 0.000102 time 0.2626 (0.2858) loss 3.6428 (3.0950) grad_norm 2.3445 (inf) [2021-04-16 15:26:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][1170/1251] eta 0:00:23 lr 0.000102 time 0.2648 (0.2857) loss 3.4455 (3.0935) grad_norm 2.5711 (inf) [2021-04-16 15:26:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][1180/1251] eta 0:00:20 lr 0.000102 time 0.2853 (0.2856) loss 3.1713 (3.0951) grad_norm 2.9703 (inf) [2021-04-16 15:26:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][1190/1251] eta 0:00:17 lr 0.000102 time 0.2529 (0.2855) loss 3.7836 (3.0952) grad_norm 2.6696 (inf) [2021-04-16 15:26:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][1200/1251] eta 0:00:14 lr 0.000102 time 0.2771 (0.2854) loss 3.9217 (3.0950) grad_norm 2.7277 (inf) [2021-04-16 15:26:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][1210/1251] eta 0:00:11 lr 0.000102 time 0.2913 (0.2853) loss 1.9851 (3.0948) grad_norm 2.7070 (inf) [2021-04-16 15:26:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][1220/1251] eta 0:00:08 lr 0.000102 time 0.2808 (0.2854) loss 3.5357 (3.0953) grad_norm 2.4272 (inf) [2021-04-16 15:26:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][1230/1251] eta 0:00:05 lr 0.000102 time 0.2612 (0.2854) loss 2.9410 (3.0947) grad_norm 2.4189 (inf) [2021-04-16 15:26:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][1240/1251] eta 0:00:03 lr 0.000102 time 0.2482 (0.2852) loss 3.4845 (3.0973) grad_norm 2.2508 (inf) [2021-04-16 15:26:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [240/300][1250/1251] eta 0:00:00 lr 0.000102 time 0.2479 (0.2849) loss 3.4570 (3.0979) grad_norm 2.6001 (inf) [2021-04-16 15:27:03 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 240 training takes 0:06:11 [2021-04-16 15:27:03 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_240.pth saving...... [2021-04-16 15:27:21 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_240.pth saved !!! [2021-04-16 15:27:22 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.163 (1.163) Loss 0.8311 (0.8311) Acc@1 79.395 (79.395) Acc@5 96.484 (96.484) [2021-04-16 15:27:23 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.156 (0.225) Loss 0.7926 (0.8506) Acc@1 82.715 (80.407) Acc@5 94.824 (95.144) [2021-04-16 15:27:26 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.708 (0.248) Loss 0.8729 (0.8567) Acc@1 79.004 (80.153) Acc@5 95.508 (95.052) [2021-04-16 15:27:28 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.122 (0.226) Loss 0.6880 (0.8477) Acc@1 83.691 (80.330) Acc@5 97.168 (95.105) [2021-04-16 15:27:30 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.261 (0.222) Loss 0.7578 (0.8506) Acc@1 81.055 (80.166) Acc@5 96.777 (95.115) [2021-04-16 15:27:44 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.136 Acc@5 95.110 [2021-04-16 15:27:44 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 80.1% [2021-04-16 15:27:44 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 80.14% [2021-04-16 15:27:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][0/1251] eta 3:27:48 lr 0.000102 time 9.9669 (9.9669) loss 3.4304 (3.4304) grad_norm 2.3299 (2.3299) [2021-04-16 15:27:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][10/1251] eta 0:23:53 lr 0.000101 time 0.2844 (1.1550) loss 3.1927 (3.0909) grad_norm 2.6308 (2.6070) [2021-04-16 15:28:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][20/1251] eta 0:15:07 lr 0.000101 time 0.2940 (0.7373) loss 2.9027 (3.0280) grad_norm 2.2557 (2.6236) [2021-04-16 15:28:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][30/1251] eta 0:11:58 lr 0.000101 time 0.2792 (0.5883) loss 2.3352 (3.0710) grad_norm 2.8800 (2.6218) [2021-04-16 15:28:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][40/1251] eta 0:10:20 lr 0.000101 time 0.2562 (0.5120) loss 3.4911 (3.1190) grad_norm 2.6107 (2.6357) [2021-04-16 15:28:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][50/1251] eta 0:09:19 lr 0.000101 time 0.2672 (0.4659) loss 3.2037 (3.1530) grad_norm 2.6246 (2.6371) [2021-04-16 15:28:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][60/1251] eta 0:08:39 lr 0.000101 time 0.2958 (0.4360) loss 2.5542 (3.0768) grad_norm 3.2538 (2.6288) [2021-04-16 15:28:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][70/1251] eta 0:08:10 lr 0.000101 time 0.2694 (0.4154) loss 2.9438 (3.1036) grad_norm 2.6224 (2.6346) [2021-04-16 15:28:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][80/1251] eta 0:07:48 lr 0.000101 time 0.2753 (0.4002) loss 3.3762 (3.1000) grad_norm 3.1979 (2.6235) [2021-04-16 15:28:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][90/1251] eta 0:07:28 lr 0.000101 time 0.2673 (0.3862) loss 3.9669 (3.1305) grad_norm 2.9342 (2.6848) [2021-04-16 15:28:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][100/1251] eta 0:07:13 lr 0.000101 time 0.2557 (0.3763) loss 2.5443 (3.1139) grad_norm 3.0892 (2.7076) [2021-04-16 15:28:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][110/1251] eta 0:06:59 lr 0.000101 time 0.2952 (0.3675) loss 3.1821 (3.1238) grad_norm 2.3472 (2.6904) [2021-04-16 15:28:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][120/1251] eta 0:06:47 lr 0.000101 time 0.2502 (0.3600) loss 3.8132 (3.1488) grad_norm 2.6748 (2.7117) [2021-04-16 15:28:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][130/1251] eta 0:06:36 lr 0.000101 time 0.2697 (0.3536) loss 3.3325 (3.1447) grad_norm 2.6595 (2.7122) [2021-04-16 15:28:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][140/1251] eta 0:06:27 lr 0.000101 time 0.2702 (0.3490) loss 3.1557 (3.1513) grad_norm 3.0757 (2.7090) [2021-04-16 15:28:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][150/1251] eta 0:06:19 lr 0.000101 time 0.2970 (0.3443) loss 2.2020 (3.1309) grad_norm 2.4594 (2.7035) [2021-04-16 15:28:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][160/1251] eta 0:06:10 lr 0.000101 time 0.2697 (0.3399) loss 3.5443 (3.1325) grad_norm 2.8105 (2.6910) [2021-04-16 15:28:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][170/1251] eta 0:06:03 lr 0.000101 time 0.2462 (0.3359) loss 3.8299 (3.1393) grad_norm 2.8278 (2.6806) [2021-04-16 15:28:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][180/1251] eta 0:05:56 lr 0.000101 time 0.2746 (0.3326) loss 3.3427 (3.1400) grad_norm 2.9289 (2.6850) [2021-04-16 15:28:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][190/1251] eta 0:05:49 lr 0.000101 time 0.2918 (0.3296) loss 3.5606 (3.1347) grad_norm 3.4713 (2.6940) [2021-04-16 15:28:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][200/1251] eta 0:05:43 lr 0.000101 time 0.2785 (0.3271) loss 3.6677 (3.1363) grad_norm 3.4444 (2.7001) [2021-04-16 15:28:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][210/1251] eta 0:05:37 lr 0.000101 time 0.2912 (0.3247) loss 2.2612 (3.1206) grad_norm 3.3637 (2.6991) [2021-04-16 15:28:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][220/1251] eta 0:05:32 lr 0.000101 time 0.2605 (0.3225) loss 2.8457 (3.1062) grad_norm 2.6332 (2.6968) [2021-04-16 15:28:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][230/1251] eta 0:05:27 lr 0.000101 time 0.2677 (0.3210) loss 3.6843 (3.1087) grad_norm 2.3212 (2.7134) [2021-04-16 15:29:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][240/1251] eta 0:05:22 lr 0.000101 time 0.3019 (0.3194) loss 3.3478 (3.1078) grad_norm 2.9655 (2.7963) [2021-04-16 15:29:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][250/1251] eta 0:05:18 lr 0.000101 time 0.2626 (0.3182) loss 3.5067 (3.1068) grad_norm 2.6833 (2.7905) [2021-04-16 15:29:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][260/1251] eta 0:05:14 lr 0.000101 time 0.2786 (0.3169) loss 3.0433 (3.1086) grad_norm 2.3822 (2.7834) [2021-04-16 15:29:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][270/1251] eta 0:05:09 lr 0.000101 time 0.2857 (0.3154) loss 3.2220 (3.1123) grad_norm 2.3718 (2.7891) [2021-04-16 15:29:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][280/1251] eta 0:05:05 lr 0.000101 time 0.3120 (0.3143) loss 3.6826 (3.1191) grad_norm 2.7095 (2.7796) [2021-04-16 15:29:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][290/1251] eta 0:05:00 lr 0.000101 time 0.2545 (0.3132) loss 2.1497 (3.1075) grad_norm 2.9771 (2.7753) [2021-04-16 15:29:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][300/1251] eta 0:04:56 lr 0.000101 time 0.2840 (0.3120) loss 3.3183 (3.1099) grad_norm 3.2335 (2.7732) [2021-04-16 15:29:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][310/1251] eta 0:04:52 lr 0.000101 time 0.2743 (0.3107) loss 3.6405 (3.1157) grad_norm 2.6968 (2.7674) [2021-04-16 15:29:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][320/1251] eta 0:04:48 lr 0.000101 time 0.2709 (0.3097) loss 3.4776 (3.1182) grad_norm 2.4687 (2.7668) [2021-04-16 15:29:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][330/1251] eta 0:04:44 lr 0.000101 time 0.2675 (0.3087) loss 2.7078 (3.1180) grad_norm 2.9116 (2.7662) [2021-04-16 15:29:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][340/1251] eta 0:04:40 lr 0.000101 time 0.2524 (0.3082) loss 1.9857 (3.1102) grad_norm 2.5702 (2.7610) [2021-04-16 15:29:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][350/1251] eta 0:04:36 lr 0.000101 time 0.2567 (0.3072) loss 3.4644 (3.1156) grad_norm 2.5131 (2.7618) [2021-04-16 15:29:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][360/1251] eta 0:04:32 lr 0.000101 time 0.2692 (0.3063) loss 3.6493 (3.1139) grad_norm 3.0593 (2.7638) [2021-04-16 15:29:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][370/1251] eta 0:04:29 lr 0.000101 time 0.2796 (0.3059) loss 3.3936 (3.1111) grad_norm 2.8630 (2.7614) [2021-04-16 15:29:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][380/1251] eta 0:04:25 lr 0.000101 time 0.2739 (0.3053) loss 3.6095 (3.1057) grad_norm 2.3184 (2.7624) [2021-04-16 15:29:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][390/1251] eta 0:04:22 lr 0.000101 time 0.2865 (0.3046) loss 3.5301 (3.1080) grad_norm 2.3868 (2.7594) [2021-04-16 15:29:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][400/1251] eta 0:04:18 lr 0.000101 time 0.2666 (0.3038) loss 3.1477 (3.1130) grad_norm 4.7989 (2.7604) [2021-04-16 15:29:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][410/1251] eta 0:04:14 lr 0.000101 time 0.2719 (0.3031) loss 3.1489 (3.1142) grad_norm 2.8583 (2.7622) [2021-04-16 15:29:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][420/1251] eta 0:04:11 lr 0.000101 time 0.2824 (0.3028) loss 3.1798 (3.1138) grad_norm 2.3648 (2.7614) [2021-04-16 15:29:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][430/1251] eta 0:04:08 lr 0.000100 time 0.2860 (0.3025) loss 3.4163 (3.1169) grad_norm 2.7897 (2.7670) [2021-04-16 15:29:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][440/1251] eta 0:04:04 lr 0.000100 time 0.2696 (0.3019) loss 2.9962 (3.1167) grad_norm 3.1130 (2.7664) [2021-04-16 15:30:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][450/1251] eta 0:04:01 lr 0.000100 time 0.2740 (0.3013) loss 2.8499 (3.1189) grad_norm 2.5760 (2.7610) [2021-04-16 15:30:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][460/1251] eta 0:03:57 lr 0.000100 time 0.2667 (0.3009) loss 3.4159 (3.1184) grad_norm 2.4601 (2.7564) [2021-04-16 15:30:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][470/1251] eta 0:03:54 lr 0.000100 time 0.3001 (0.3006) loss 3.2980 (3.1227) grad_norm 2.7230 (2.7549) [2021-04-16 15:30:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][480/1251] eta 0:03:51 lr 0.000100 time 0.3117 (0.3002) loss 3.5005 (3.1196) grad_norm 2.7845 (2.7531) [2021-04-16 15:30:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][490/1251] eta 0:03:48 lr 0.000100 time 0.2701 (0.3000) loss 2.5433 (3.1113) grad_norm 2.6623 (2.7550) [2021-04-16 15:30:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][500/1251] eta 0:03:45 lr 0.000100 time 0.2740 (0.2997) loss 3.0643 (3.1137) grad_norm 2.4200 (2.7522) [2021-04-16 15:30:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][510/1251] eta 0:03:41 lr 0.000100 time 0.2710 (0.2992) loss 1.9602 (3.1037) grad_norm 2.5399 (2.7465) [2021-04-16 15:30:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][520/1251] eta 0:03:38 lr 0.000100 time 0.2516 (0.2987) loss 2.9438 (3.1051) grad_norm 2.8435 (2.7427) [2021-04-16 15:30:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][530/1251] eta 0:03:35 lr 0.000100 time 0.2886 (0.2984) loss 3.9856 (3.1035) grad_norm 2.5775 (2.7416) [2021-04-16 15:30:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][540/1251] eta 0:03:32 lr 0.000100 time 0.2812 (0.2982) loss 3.4615 (3.1018) grad_norm 2.6394 (2.7397) [2021-04-16 15:30:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][550/1251] eta 0:03:28 lr 0.000100 time 0.2981 (0.2979) loss 3.5174 (3.1015) grad_norm 2.5205 (2.7399) [2021-04-16 15:30:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][560/1251] eta 0:03:25 lr 0.000100 time 0.2955 (0.2975) loss 3.8530 (3.1051) grad_norm 2.6610 (2.7408) [2021-04-16 15:30:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][570/1251] eta 0:03:22 lr 0.000100 time 0.2922 (0.2973) loss 2.9809 (3.1110) grad_norm 2.5582 (2.7393) [2021-04-16 15:30:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][580/1251] eta 0:03:19 lr 0.000100 time 0.2840 (0.2974) loss 2.7006 (3.1099) grad_norm 2.8235 (2.7367) [2021-04-16 15:30:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][590/1251] eta 0:03:16 lr 0.000100 time 0.2902 (0.2969) loss 3.7761 (3.1144) grad_norm 2.3216 (2.7342) [2021-04-16 15:30:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][600/1251] eta 0:03:13 lr 0.000100 time 0.2564 (0.2966) loss 3.0609 (3.1095) grad_norm 2.6797 (2.7320) [2021-04-16 15:30:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][610/1251] eta 0:03:09 lr 0.000100 time 0.2580 (0.2963) loss 3.5404 (3.1119) grad_norm 2.5931 (2.7328) [2021-04-16 15:30:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][620/1251] eta 0:03:06 lr 0.000100 time 0.2794 (0.2960) loss 3.7963 (3.1109) grad_norm 2.4534 (2.7311) [2021-04-16 15:30:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][630/1251] eta 0:03:03 lr 0.000100 time 0.2577 (0.2957) loss 3.4657 (3.1124) grad_norm 2.4506 (2.7291) [2021-04-16 15:30:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][640/1251] eta 0:03:00 lr 0.000100 time 0.2672 (0.2956) loss 2.5108 (3.1173) grad_norm 2.3030 (2.7266) [2021-04-16 15:30:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][650/1251] eta 0:02:57 lr 0.000100 time 0.2849 (0.2954) loss 3.5496 (3.1169) grad_norm 2.6373 (2.7231) [2021-04-16 15:30:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][660/1251] eta 0:02:54 lr 0.000100 time 0.2670 (0.2952) loss 3.4130 (3.1160) grad_norm 2.6039 (2.7226) [2021-04-16 15:31:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][670/1251] eta 0:02:51 lr 0.000100 time 0.2741 (0.2949) loss 2.9796 (3.1150) grad_norm 2.3545 (2.7199) [2021-04-16 15:31:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][680/1251] eta 0:02:48 lr 0.000100 time 0.2828 (0.2948) loss 3.2253 (3.1164) grad_norm 2.6650 (2.7189) [2021-04-16 15:31:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][690/1251] eta 0:02:45 lr 0.000100 time 0.2826 (0.2946) loss 2.8856 (3.1150) grad_norm 2.6467 (2.7165) [2021-04-16 15:31:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][700/1251] eta 0:02:42 lr 0.000100 time 0.2446 (0.2943) loss 3.3997 (3.1179) grad_norm 2.3531 (2.7159) [2021-04-16 15:31:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][710/1251] eta 0:02:39 lr 0.000100 time 0.2660 (0.2941) loss 3.0060 (3.1164) grad_norm 2.1621 (2.7132) [2021-04-16 15:31:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][720/1251] eta 0:02:36 lr 0.000100 time 0.3184 (0.2939) loss 3.3467 (3.1172) grad_norm 2.8226 (2.7107) [2021-04-16 15:31:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][730/1251] eta 0:02:33 lr 0.000100 time 0.2580 (0.2937) loss 2.7299 (3.1163) grad_norm 2.5772 (2.7090) [2021-04-16 15:31:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][740/1251] eta 0:02:29 lr 0.000100 time 0.3081 (0.2935) loss 3.6383 (3.1146) grad_norm 3.0586 (2.7077) [2021-04-16 15:31:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][750/1251] eta 0:02:26 lr 0.000100 time 0.2808 (0.2932) loss 3.5860 (3.1176) grad_norm 2.8742 (2.7066) [2021-04-16 15:31:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][760/1251] eta 0:02:23 lr 0.000100 time 0.3042 (0.2930) loss 3.5604 (3.1226) grad_norm 2.2043 (2.7041) [2021-04-16 15:31:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][770/1251] eta 0:02:20 lr 0.000100 time 0.2816 (0.2928) loss 3.5300 (3.1257) grad_norm 2.5166 (2.7025) [2021-04-16 15:31:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][780/1251] eta 0:02:17 lr 0.000100 time 0.2678 (0.2927) loss 3.7548 (3.1297) grad_norm 2.2919 (2.6984) [2021-04-16 15:31:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][790/1251] eta 0:02:14 lr 0.000100 time 0.2653 (0.2924) loss 3.5057 (3.1289) grad_norm 2.4380 (2.6970) [2021-04-16 15:31:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][800/1251] eta 0:02:11 lr 0.000100 time 0.2649 (0.2922) loss 3.7574 (3.1289) grad_norm 2.7664 (2.6950) [2021-04-16 15:31:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][810/1251] eta 0:02:08 lr 0.000100 time 0.2567 (0.2921) loss 3.5819 (3.1302) grad_norm 2.2437 (2.6934) [2021-04-16 15:31:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][820/1251] eta 0:02:05 lr 0.000100 time 0.2782 (0.2920) loss 3.6848 (3.1297) grad_norm 2.3401 (2.6907) [2021-04-16 15:31:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][830/1251] eta 0:02:02 lr 0.000100 time 0.2661 (0.2919) loss 3.9096 (3.1323) grad_norm 2.8185 (2.6880) [2021-04-16 15:31:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][840/1251] eta 0:01:59 lr 0.000100 time 0.2643 (0.2917) loss 3.3956 (3.1340) grad_norm 3.1656 (2.6871) [2021-04-16 15:31:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][850/1251] eta 0:01:56 lr 0.000099 time 0.2689 (0.2915) loss 3.6073 (3.1355) grad_norm 3.2317 (2.6868) [2021-04-16 15:31:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][860/1251] eta 0:01:53 lr 0.000099 time 0.2732 (0.2913) loss 3.6059 (3.1369) grad_norm 2.3330 (2.6842) [2021-04-16 15:31:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][870/1251] eta 0:01:50 lr 0.000099 time 0.2697 (0.2912) loss 3.6499 (3.1390) grad_norm 2.4892 (2.6834) [2021-04-16 15:32:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][880/1251] eta 0:01:47 lr 0.000099 time 0.2838 (0.2911) loss 2.1122 (3.1386) grad_norm 2.5061 (2.6829) [2021-04-16 15:32:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][890/1251] eta 0:01:45 lr 0.000099 time 0.2740 (0.2909) loss 3.3629 (3.1388) grad_norm 2.9384 (2.6835) [2021-04-16 15:32:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][900/1251] eta 0:01:42 lr 0.000099 time 0.2725 (0.2909) loss 2.8020 (3.1403) grad_norm 2.6110 (2.6814) [2021-04-16 15:32:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][910/1251] eta 0:01:39 lr 0.000099 time 0.2821 (0.2907) loss 3.4244 (3.1424) grad_norm 2.5114 (2.6814) [2021-04-16 15:32:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][920/1251] eta 0:01:36 lr 0.000099 time 0.2596 (0.2907) loss 3.3738 (3.1419) grad_norm 2.5138 (2.6797) [2021-04-16 15:32:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][930/1251] eta 0:01:33 lr 0.000099 time 0.2969 (0.2905) loss 2.8882 (3.1411) grad_norm 2.5306 (2.6773) [2021-04-16 15:32:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][940/1251] eta 0:01:30 lr 0.000099 time 0.2497 (0.2905) loss 3.1981 (3.1404) grad_norm 2.5795 (2.6767) [2021-04-16 15:32:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][950/1251] eta 0:01:27 lr 0.000099 time 0.2827 (0.2903) loss 3.2535 (3.1396) grad_norm 2.4428 (2.6763) [2021-04-16 15:32:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][960/1251] eta 0:01:24 lr 0.000099 time 0.2873 (0.2901) loss 3.4352 (3.1390) grad_norm 3.3275 (2.6771) [2021-04-16 15:32:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][970/1251] eta 0:01:21 lr 0.000099 time 0.2841 (0.2900) loss 2.1523 (3.1389) grad_norm 2.9518 (2.6768) [2021-04-16 15:32:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][980/1251] eta 0:01:18 lr 0.000099 time 0.2921 (0.2899) loss 3.3345 (3.1399) grad_norm 2.2708 (2.6749) [2021-04-16 15:32:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][990/1251] eta 0:01:15 lr 0.000099 time 0.2881 (0.2899) loss 2.3731 (3.1392) grad_norm 2.7371 (2.6725) [2021-04-16 15:32:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][1000/1251] eta 0:01:12 lr 0.000099 time 0.2684 (0.2897) loss 2.9134 (3.1392) grad_norm 2.3560 (2.6719) [2021-04-16 15:32:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][1010/1251] eta 0:01:09 lr 0.000099 time 0.2750 (0.2897) loss 3.7536 (3.1408) grad_norm 2.8690 (2.6693) [2021-04-16 15:32:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][1020/1251] eta 0:01:06 lr 0.000099 time 0.2642 (0.2896) loss 1.6731 (3.1407) grad_norm 2.6443 (2.6685) [2021-04-16 15:32:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][1030/1251] eta 0:01:03 lr 0.000099 time 0.2908 (0.2894) loss 3.6678 (3.1423) grad_norm 2.4534 (2.6667) [2021-04-16 15:32:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][1040/1251] eta 0:01:01 lr 0.000099 time 0.2561 (0.2894) loss 3.0747 (3.1435) grad_norm 2.4446 (2.6655) [2021-04-16 15:32:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][1050/1251] eta 0:00:58 lr 0.000099 time 0.2788 (0.2893) loss 3.9696 (3.1423) grad_norm 2.2636 (2.6642) [2021-04-16 15:32:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][1060/1251] eta 0:00:55 lr 0.000099 time 0.2929 (0.2892) loss 3.8749 (3.1427) grad_norm 2.6235 (2.6647) [2021-04-16 15:32:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][1070/1251] eta 0:00:52 lr 0.000099 time 0.2925 (0.2892) loss 3.5701 (3.1416) grad_norm 2.6357 (2.6652) [2021-04-16 15:32:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][1080/1251] eta 0:00:49 lr 0.000099 time 0.2637 (0.2890) loss 2.4064 (3.1401) grad_norm 2.7666 (2.6667) [2021-04-16 15:32:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][1090/1251] eta 0:00:46 lr 0.000099 time 0.2655 (0.2889) loss 2.7333 (3.1404) grad_norm 2.3324 (2.6644) [2021-04-16 15:33:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][1100/1251] eta 0:00:43 lr 0.000099 time 0.2880 (0.2888) loss 3.3952 (3.1407) grad_norm 2.4625 (2.6626) [2021-04-16 15:33:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][1110/1251] eta 0:00:40 lr 0.000099 time 0.2792 (0.2887) loss 3.6083 (3.1405) grad_norm 2.9314 (2.6643) [2021-04-16 15:33:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][1120/1251] eta 0:00:37 lr 0.000099 time 0.2991 (0.2887) loss 3.2176 (3.1401) grad_norm 2.7969 (2.6638) [2021-04-16 15:33:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][1130/1251] eta 0:00:34 lr 0.000099 time 0.2964 (0.2886) loss 3.3556 (3.1398) grad_norm 2.7098 (2.6634) [2021-04-16 15:33:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][1140/1251] eta 0:00:32 lr 0.000099 time 0.2706 (0.2885) loss 3.3697 (3.1390) grad_norm 2.7794 (2.6647) [2021-04-16 15:33:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][1150/1251] eta 0:00:29 lr 0.000099 time 0.2925 (0.2885) loss 3.5681 (3.1395) grad_norm 2.4106 (2.6650) [2021-04-16 15:33:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][1160/1251] eta 0:00:26 lr 0.000099 time 0.2752 (0.2884) loss 4.0874 (3.1399) grad_norm 2.7691 (2.6647) [2021-04-16 15:33:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][1170/1251] eta 0:00:23 lr 0.000099 time 0.2676 (0.2883) loss 2.5943 (3.1393) grad_norm 3.4916 (2.6645) [2021-04-16 15:33:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][1180/1251] eta 0:00:20 lr 0.000099 time 0.2861 (0.2882) loss 3.4912 (3.1411) grad_norm 2.9250 (2.6651) [2021-04-16 15:33:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][1190/1251] eta 0:00:17 lr 0.000099 time 0.2820 (0.2881) loss 3.4125 (3.1406) grad_norm 3.9756 (2.6698) [2021-04-16 15:33:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][1200/1251] eta 0:00:14 lr 0.000099 time 0.2661 (0.2880) loss 3.2480 (3.1402) grad_norm 2.3077 (2.6722) [2021-04-16 15:33:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][1210/1251] eta 0:00:11 lr 0.000099 time 0.2862 (0.2878) loss 3.6220 (3.1399) grad_norm 3.7931 (2.6728) [2021-04-16 15:33:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][1220/1251] eta 0:00:08 lr 0.000099 time 0.2769 (0.2878) loss 3.1606 (3.1410) grad_norm 2.7089 (2.6725) [2021-04-16 15:33:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][1230/1251] eta 0:00:06 lr 0.000099 time 0.2665 (0.2878) loss 3.7847 (3.1406) grad_norm 2.5271 (2.6741) [2021-04-16 15:33:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][1240/1251] eta 0:00:03 lr 0.000099 time 0.2490 (0.2876) loss 3.5874 (3.1416) grad_norm 2.3312 (2.6724) [2021-04-16 15:33:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [241/300][1250/1251] eta 0:00:00 lr 0.000099 time 0.2616 (0.2873) loss 3.2366 (3.1428) grad_norm 2.5443 (2.6711) [2021-04-16 15:33:59 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 241 training takes 0:06:14 [2021-04-16 15:33:59 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_241.pth saving...... [2021-04-16 15:34:23 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_241.pth saved !!! [2021-04-16 15:34:24 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.182 (1.182) Loss 0.8566 (0.8566) Acc@1 79.395 (79.395) Acc@5 95.312 (95.312) [2021-04-16 15:34:26 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.159 (0.242) Loss 0.8471 (0.8448) Acc@1 80.762 (80.344) Acc@5 95.312 (95.197) [2021-04-16 15:34:28 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.201 (0.221) Loss 0.8862 (0.8499) Acc@1 79.980 (80.027) Acc@5 94.922 (95.094) [2021-04-16 15:34:30 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.149 (0.218) Loss 0.8954 (0.8453) Acc@1 78.906 (80.144) Acc@5 94.824 (95.114) [2021-04-16 15:34:32 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.210) Loss 0.8152 (0.8463) Acc@1 81.641 (80.095) Acc@5 96.094 (95.098) [2021-04-16 15:34:45 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.152 Acc@5 95.066 [2021-04-16 15:34:45 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 80.2% [2021-04-16 15:34:45 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 80.15% [2021-04-16 15:35:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][0/1251] eta 6:49:52 lr 0.000099 time 19.6585 (19.6585) loss 2.3638 (2.3638) grad_norm 2.5323 (2.5323) [2021-04-16 15:35:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][10/1251] eta 0:42:06 lr 0.000099 time 0.2776 (2.0359) loss 3.4233 (2.9306) grad_norm 2.2849 (2.6184) [2021-04-16 15:35:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][20/1251] eta 0:24:44 lr 0.000098 time 0.2773 (1.2063) loss 3.4716 (3.1544) grad_norm 2.3284 (2.6075) [2021-04-16 15:35:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][30/1251] eta 0:18:30 lr 0.000098 time 0.2757 (0.9093) loss 3.7508 (3.1370) grad_norm 2.4455 (2.5954) [2021-04-16 15:35:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][40/1251] eta 0:15:14 lr 0.000098 time 0.2575 (0.7550) loss 3.6065 (3.1246) grad_norm 2.4836 (2.5655) [2021-04-16 15:35:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][50/1251] eta 0:13:14 lr 0.000098 time 0.2705 (0.6615) loss 3.7011 (3.1167) grad_norm 2.3035 (2.5666) [2021-04-16 15:35:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][60/1251] eta 0:11:52 lr 0.000098 time 0.2744 (0.5981) loss 3.7349 (3.1344) grad_norm 2.3243 (2.5909) [2021-04-16 15:35:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][70/1251] eta 0:10:53 lr 0.000098 time 0.2612 (0.5530) loss 3.6161 (3.1594) grad_norm 2.8872 (2.5974) [2021-04-16 15:35:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][80/1251] eta 0:10:08 lr 0.000098 time 0.2969 (0.5193) loss 3.2790 (3.1781) grad_norm 2.3256 (2.6090) [2021-04-16 15:35:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][90/1251] eta 0:09:32 lr 0.000098 time 0.2764 (0.4927) loss 2.7201 (3.1751) grad_norm 2.4572 (2.5959) [2021-04-16 15:35:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][100/1251] eta 0:09:02 lr 0.000098 time 0.2863 (0.4716) loss 2.6666 (3.1462) grad_norm 2.8582 (2.6105) [2021-04-16 15:35:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][110/1251] eta 0:08:38 lr 0.000098 time 0.2698 (0.4541) loss 3.2912 (3.1562) grad_norm 2.2945 (2.6129) [2021-04-16 15:35:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][120/1251] eta 0:08:18 lr 0.000098 time 0.2780 (0.4404) loss 3.2052 (3.1592) grad_norm 2.3367 (2.6114) [2021-04-16 15:35:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][130/1251] eta 0:07:59 lr 0.000098 time 0.2703 (0.4279) loss 2.7574 (3.1550) grad_norm 2.2948 (2.5992) [2021-04-16 15:35:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][140/1251] eta 0:07:44 lr 0.000098 time 0.2510 (0.4180) loss 2.4045 (3.1521) grad_norm 3.0872 (2.6026) [2021-04-16 15:35:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][150/1251] eta 0:07:29 lr 0.000098 time 0.2833 (0.4086) loss 3.6634 (3.1539) grad_norm 2.3495 (nan) [2021-04-16 15:35:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][160/1251] eta 0:07:16 lr 0.000098 time 0.2671 (0.4003) loss 3.7217 (3.1543) grad_norm 2.9195 (nan) [2021-04-16 15:35:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][170/1251] eta 0:07:04 lr 0.000098 time 0.2754 (0.3931) loss 3.7525 (3.1557) grad_norm 2.4531 (nan) [2021-04-16 15:35:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][180/1251] eta 0:06:53 lr 0.000098 time 0.2719 (0.3864) loss 3.6818 (3.1625) grad_norm 2.6091 (nan) [2021-04-16 15:35:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][190/1251] eta 0:06:44 lr 0.000098 time 0.2643 (0.3808) loss 3.5498 (3.1705) grad_norm 2.8478 (nan) [2021-04-16 15:36:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][200/1251] eta 0:06:35 lr 0.000098 time 0.2906 (0.3767) loss 2.9793 (3.1663) grad_norm 2.6392 (nan) [2021-04-16 15:36:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][210/1251] eta 0:06:26 lr 0.000098 time 0.2545 (0.3717) loss 3.3835 (3.1551) grad_norm 3.0281 (nan) [2021-04-16 15:36:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][220/1251] eta 0:06:19 lr 0.000098 time 0.2608 (0.3680) loss 2.7748 (3.1581) grad_norm 2.2038 (nan) [2021-04-16 15:36:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][230/1251] eta 0:06:11 lr 0.000098 time 0.3014 (0.3641) loss 2.4293 (3.1443) grad_norm 2.2223 (nan) [2021-04-16 15:36:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][240/1251] eta 0:06:04 lr 0.000098 time 0.2905 (0.3603) loss 3.6553 (3.1473) grad_norm 2.2277 (nan) [2021-04-16 15:36:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][250/1251] eta 0:05:57 lr 0.000098 time 0.2719 (0.3574) loss 2.7113 (3.1509) grad_norm 2.4839 (nan) [2021-04-16 15:36:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][260/1251] eta 0:05:51 lr 0.000098 time 0.2697 (0.3543) loss 2.0304 (3.1429) grad_norm 2.4902 (nan) [2021-04-16 15:36:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][270/1251] eta 0:05:44 lr 0.000098 time 0.2708 (0.3514) loss 2.6223 (3.1476) grad_norm 2.3298 (nan) [2021-04-16 15:36:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][280/1251] eta 0:05:38 lr 0.000098 time 0.2665 (0.3487) loss 2.9671 (3.1409) grad_norm 2.2344 (nan) [2021-04-16 15:36:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][290/1251] eta 0:05:32 lr 0.000098 time 0.2843 (0.3465) loss 2.0354 (3.1365) grad_norm 3.8070 (nan) [2021-04-16 15:36:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][300/1251] eta 0:05:27 lr 0.000098 time 0.2799 (0.3442) loss 3.0742 (3.1392) grad_norm 2.3799 (nan) [2021-04-16 15:36:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][310/1251] eta 0:05:21 lr 0.000098 time 0.2692 (0.3420) loss 3.6997 (3.1423) grad_norm 2.6548 (nan) [2021-04-16 15:36:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][320/1251] eta 0:05:16 lr 0.000098 time 0.2655 (0.3403) loss 2.7219 (3.1391) grad_norm 2.2665 (nan) [2021-04-16 15:36:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][330/1251] eta 0:05:11 lr 0.000098 time 0.2974 (0.3384) loss 3.4298 (3.1348) grad_norm 2.7620 (nan) [2021-04-16 15:36:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][340/1251] eta 0:05:06 lr 0.000098 time 0.2838 (0.3367) loss 2.1191 (3.1372) grad_norm 3.3091 (nan) [2021-04-16 15:36:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][350/1251] eta 0:05:01 lr 0.000098 time 0.2671 (0.3350) loss 3.5336 (3.1401) grad_norm 2.4335 (nan) [2021-04-16 15:36:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][360/1251] eta 0:04:57 lr 0.000098 time 0.2803 (0.3343) loss 2.8377 (3.1385) grad_norm 2.5172 (nan) [2021-04-16 15:36:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][370/1251] eta 0:04:53 lr 0.000098 time 0.3006 (0.3331) loss 3.0014 (3.1339) grad_norm 2.6786 (nan) [2021-04-16 15:36:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][380/1251] eta 0:04:49 lr 0.000098 time 0.3164 (0.3318) loss 2.8985 (3.1279) grad_norm 2.7908 (nan) [2021-04-16 15:36:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][390/1251] eta 0:04:44 lr 0.000098 time 0.2650 (0.3306) loss 3.5926 (3.1292) grad_norm 3.0394 (nan) [2021-04-16 15:36:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][400/1251] eta 0:04:40 lr 0.000098 time 0.2694 (0.3296) loss 2.4159 (3.1223) grad_norm 2.8177 (nan) [2021-04-16 15:37:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][410/1251] eta 0:04:36 lr 0.000098 time 0.2993 (0.3283) loss 2.6974 (3.1194) grad_norm 2.2587 (nan) [2021-04-16 15:37:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][420/1251] eta 0:04:31 lr 0.000098 time 0.2578 (0.3271) loss 3.2559 (3.1194) grad_norm 2.5783 (nan) [2021-04-16 15:37:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][430/1251] eta 0:04:27 lr 0.000098 time 0.2664 (0.3263) loss 3.5692 (3.1229) grad_norm 3.1830 (nan) [2021-04-16 15:37:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][440/1251] eta 0:04:23 lr 0.000097 time 0.3128 (0.3252) loss 2.8842 (3.1240) grad_norm 2.4120 (nan) [2021-04-16 15:37:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][450/1251] eta 0:04:19 lr 0.000097 time 0.2873 (0.3241) loss 3.0250 (3.1249) grad_norm 2.5609 (nan) [2021-04-16 15:37:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][460/1251] eta 0:04:15 lr 0.000097 time 0.2728 (0.3231) loss 3.6949 (3.1310) grad_norm 2.5031 (nan) [2021-04-16 15:37:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][470/1251] eta 0:04:11 lr 0.000097 time 0.2821 (0.3223) loss 3.2010 (3.1303) grad_norm 2.5582 (nan) [2021-04-16 15:37:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][480/1251] eta 0:04:07 lr 0.000097 time 0.2679 (0.3212) loss 2.5957 (3.1314) grad_norm 2.5020 (nan) [2021-04-16 15:37:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][490/1251] eta 0:04:03 lr 0.000097 time 0.2606 (0.3206) loss 3.2544 (3.1282) grad_norm 2.4728 (nan) [2021-04-16 15:37:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][500/1251] eta 0:04:00 lr 0.000097 time 0.2754 (0.3197) loss 3.4221 (3.1245) grad_norm 2.7290 (nan) [2021-04-16 15:37:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][510/1251] eta 0:03:56 lr 0.000097 time 0.2599 (0.3188) loss 3.3162 (3.1234) grad_norm 2.4965 (nan) [2021-04-16 15:37:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][520/1251] eta 0:03:52 lr 0.000097 time 0.2926 (0.3181) loss 3.3929 (3.1249) grad_norm 2.4649 (nan) [2021-04-16 15:37:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][530/1251] eta 0:03:48 lr 0.000097 time 0.2761 (0.3172) loss 3.3144 (3.1277) grad_norm 2.8695 (nan) [2021-04-16 15:37:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][540/1251] eta 0:03:45 lr 0.000097 time 0.2863 (0.3165) loss 2.0323 (3.1273) grad_norm 3.0658 (nan) [2021-04-16 15:37:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][550/1251] eta 0:03:41 lr 0.000097 time 0.2976 (0.3158) loss 1.9197 (3.1237) grad_norm 2.6498 (nan) [2021-04-16 15:37:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][560/1251] eta 0:03:37 lr 0.000097 time 0.2806 (0.3150) loss 2.4974 (3.1250) grad_norm 2.6068 (nan) [2021-04-16 15:37:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][570/1251] eta 0:03:34 lr 0.000097 time 0.2635 (0.3143) loss 3.8769 (3.1271) grad_norm 2.7023 (nan) [2021-04-16 15:37:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][580/1251] eta 0:03:30 lr 0.000097 time 0.2603 (0.3135) loss 3.7759 (3.1278) grad_norm 2.3683 (nan) [2021-04-16 15:37:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][590/1251] eta 0:03:27 lr 0.000097 time 0.2711 (0.3132) loss 2.6802 (3.1239) grad_norm 2.6086 (nan) [2021-04-16 15:37:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][600/1251] eta 0:03:23 lr 0.000097 time 0.3725 (0.3128) loss 4.2499 (3.1258) grad_norm 2.8404 (nan) [2021-04-16 15:37:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][610/1251] eta 0:03:20 lr 0.000097 time 0.2948 (0.3123) loss 2.2984 (3.1256) grad_norm 2.4622 (nan) [2021-04-16 15:37:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][620/1251] eta 0:03:16 lr 0.000097 time 0.2691 (0.3116) loss 3.0024 (3.1255) grad_norm 2.2112 (nan) [2021-04-16 15:38:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][630/1251] eta 0:03:13 lr 0.000097 time 0.2844 (0.3111) loss 2.3054 (3.1264) grad_norm 2.7304 (nan) [2021-04-16 15:38:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][640/1251] eta 0:03:09 lr 0.000097 time 0.2543 (0.3105) loss 2.6817 (3.1264) grad_norm 2.0348 (nan) [2021-04-16 15:38:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][650/1251] eta 0:03:06 lr 0.000097 time 0.2942 (0.3101) loss 3.6961 (3.1273) grad_norm 2.5554 (nan) [2021-04-16 15:38:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][660/1251] eta 0:03:02 lr 0.000097 time 0.2693 (0.3096) loss 2.7335 (3.1245) grad_norm 2.6877 (nan) [2021-04-16 15:38:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][670/1251] eta 0:02:59 lr 0.000097 time 0.2947 (0.3090) loss 3.0613 (3.1216) grad_norm 2.7865 (nan) [2021-04-16 15:38:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][680/1251] eta 0:02:56 lr 0.000097 time 0.2966 (0.3085) loss 2.4996 (3.1233) grad_norm 2.4635 (nan) [2021-04-16 15:38:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][690/1251] eta 0:02:52 lr 0.000097 time 0.2760 (0.3083) loss 3.4115 (3.1234) grad_norm 2.6519 (nan) [2021-04-16 15:38:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][700/1251] eta 0:02:49 lr 0.000097 time 0.2803 (0.3078) loss 3.9764 (3.1269) grad_norm 2.3388 (nan) [2021-04-16 15:38:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][710/1251] eta 0:02:46 lr 0.000097 time 0.2858 (0.3074) loss 3.1641 (3.1266) grad_norm 2.4746 (nan) [2021-04-16 15:38:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][720/1251] eta 0:02:43 lr 0.000097 time 0.2686 (0.3072) loss 3.2731 (3.1244) grad_norm 3.1573 (nan) [2021-04-16 15:38:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][730/1251] eta 0:02:39 lr 0.000097 time 0.2660 (0.3066) loss 3.3278 (3.1249) grad_norm 2.7634 (nan) [2021-04-16 15:38:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][740/1251] eta 0:02:36 lr 0.000097 time 0.2631 (0.3062) loss 3.4257 (3.1257) grad_norm 3.2133 (nan) [2021-04-16 15:38:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][750/1251] eta 0:02:33 lr 0.000097 time 0.2933 (0.3058) loss 3.0946 (3.1264) grad_norm 2.8897 (nan) [2021-04-16 15:38:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][760/1251] eta 0:02:29 lr 0.000097 time 0.3063 (0.3055) loss 3.4035 (3.1261) grad_norm 2.3406 (nan) [2021-04-16 15:38:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][770/1251] eta 0:02:26 lr 0.000097 time 0.2924 (0.3051) loss 3.3617 (3.1280) grad_norm 2.6670 (nan) [2021-04-16 15:38:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][780/1251] eta 0:02:23 lr 0.000097 time 0.2742 (0.3048) loss 3.8583 (3.1318) grad_norm 2.5273 (nan) [2021-04-16 15:38:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][790/1251] eta 0:02:20 lr 0.000097 time 0.2773 (0.3045) loss 3.1878 (3.1291) grad_norm 2.5273 (nan) [2021-04-16 15:38:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][800/1251] eta 0:02:17 lr 0.000097 time 0.2667 (0.3041) loss 3.7611 (3.1300) grad_norm 2.4176 (nan) [2021-04-16 15:38:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][810/1251] eta 0:02:13 lr 0.000097 time 0.2582 (0.3037) loss 3.3347 (3.1302) grad_norm 2.4522 (nan) [2021-04-16 15:38:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][820/1251] eta 0:02:10 lr 0.000097 time 0.2556 (0.3034) loss 2.4801 (3.1300) grad_norm 2.6589 (nan) [2021-04-16 15:38:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][830/1251] eta 0:02:07 lr 0.000097 time 0.2569 (0.3032) loss 2.7526 (3.1330) grad_norm 2.3824 (nan) [2021-04-16 15:39:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][840/1251] eta 0:02:04 lr 0.000097 time 0.2855 (0.3030) loss 1.9337 (3.1315) grad_norm 2.9714 (nan) [2021-04-16 15:39:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][850/1251] eta 0:02:01 lr 0.000097 time 0.2825 (0.3027) loss 2.5206 (3.1288) grad_norm 2.0115 (nan) [2021-04-16 15:39:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][860/1251] eta 0:01:58 lr 0.000097 time 0.3009 (0.3024) loss 2.9356 (3.1310) grad_norm 2.5703 (nan) [2021-04-16 15:39:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][870/1251] eta 0:01:55 lr 0.000096 time 0.2728 (0.3022) loss 2.3844 (3.1266) grad_norm 3.2258 (nan) [2021-04-16 15:39:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][880/1251] eta 0:01:51 lr 0.000096 time 0.2780 (0.3019) loss 2.3345 (3.1255) grad_norm 2.4508 (nan) [2021-04-16 15:39:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][890/1251] eta 0:01:48 lr 0.000096 time 0.2756 (0.3017) loss 2.9825 (3.1263) grad_norm 2.3178 (nan) [2021-04-16 15:39:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][900/1251] eta 0:01:45 lr 0.000096 time 0.2762 (0.3014) loss 3.6562 (3.1274) grad_norm 2.6373 (nan) [2021-04-16 15:39:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][910/1251] eta 0:01:42 lr 0.000096 time 0.2623 (0.3011) loss 2.0674 (3.1270) grad_norm 2.4426 (nan) [2021-04-16 15:39:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][920/1251] eta 0:01:39 lr 0.000096 time 0.2855 (0.3009) loss 3.4828 (3.1240) grad_norm 2.4673 (nan) [2021-04-16 15:39:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][930/1251] eta 0:01:36 lr 0.000096 time 0.2772 (0.3006) loss 3.3647 (3.1264) grad_norm 2.5949 (nan) [2021-04-16 15:39:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][940/1251] eta 0:01:33 lr 0.000096 time 0.2594 (0.3008) loss 3.2710 (3.1243) grad_norm 2.3079 (nan) [2021-04-16 15:39:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][950/1251] eta 0:01:30 lr 0.000096 time 0.2752 (0.3005) loss 2.1981 (3.1204) grad_norm 2.2815 (nan) [2021-04-16 15:39:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][960/1251] eta 0:01:27 lr 0.000096 time 0.2845 (0.3003) loss 2.8244 (3.1183) grad_norm 2.5619 (nan) [2021-04-16 15:39:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][970/1251] eta 0:01:24 lr 0.000096 time 0.2501 (0.3000) loss 3.2465 (3.1198) grad_norm 2.5156 (nan) [2021-04-16 15:39:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][980/1251] eta 0:01:21 lr 0.000096 time 0.2600 (0.2999) loss 3.4005 (3.1184) grad_norm 2.8925 (nan) [2021-04-16 15:39:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][990/1251] eta 0:01:18 lr 0.000096 time 0.2625 (0.2997) loss 3.5077 (3.1163) grad_norm 2.9316 (nan) [2021-04-16 15:39:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][1000/1251] eta 0:01:15 lr 0.000096 time 0.2542 (0.2995) loss 3.6036 (3.1177) grad_norm 2.6389 (nan) [2021-04-16 15:39:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][1010/1251] eta 0:01:12 lr 0.000096 time 0.2744 (0.2993) loss 2.6530 (3.1182) grad_norm 2.7857 (nan) [2021-04-16 15:39:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][1020/1251] eta 0:01:09 lr 0.000096 time 0.2668 (0.2991) loss 3.2894 (3.1167) grad_norm 2.4938 (nan) [2021-04-16 15:39:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][1030/1251] eta 0:01:06 lr 0.000096 time 0.2955 (0.2990) loss 3.4919 (3.1183) grad_norm 2.6509 (nan) [2021-04-16 15:39:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][1040/1251] eta 0:01:03 lr 0.000096 time 0.2801 (0.2988) loss 3.4366 (3.1212) grad_norm 2.5143 (nan) [2021-04-16 15:39:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][1050/1251] eta 0:01:00 lr 0.000096 time 0.2798 (0.2986) loss 3.3438 (3.1228) grad_norm 2.5603 (nan) [2021-04-16 15:40:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][1060/1251] eta 0:00:56 lr 0.000096 time 0.2896 (0.2984) loss 2.3318 (3.1192) grad_norm 2.3187 (nan) [2021-04-16 15:40:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][1070/1251] eta 0:00:54 lr 0.000096 time 0.2964 (0.2984) loss 2.8957 (3.1202) grad_norm 2.4315 (nan) [2021-04-16 15:40:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][1080/1251] eta 0:00:50 lr 0.000096 time 0.2869 (0.2981) loss 3.2767 (3.1190) grad_norm 2.9662 (nan) [2021-04-16 15:40:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][1090/1251] eta 0:00:47 lr 0.000096 time 0.2619 (0.2979) loss 2.3836 (3.1201) grad_norm 2.4809 (nan) [2021-04-16 15:40:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][1100/1251] eta 0:00:44 lr 0.000096 time 0.2636 (0.2977) loss 2.6960 (3.1195) grad_norm 2.8339 (nan) [2021-04-16 15:40:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][1110/1251] eta 0:00:41 lr 0.000096 time 0.2561 (0.2974) loss 3.3182 (3.1173) grad_norm 3.0293 (nan) [2021-04-16 15:40:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][1120/1251] eta 0:00:38 lr 0.000096 time 0.2814 (0.2973) loss 2.4305 (3.1154) grad_norm 2.1284 (nan) [2021-04-16 15:40:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][1130/1251] eta 0:00:35 lr 0.000096 time 0.2588 (0.2970) loss 2.0523 (3.1164) grad_norm 2.4556 (nan) [2021-04-16 15:40:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][1140/1251] eta 0:00:32 lr 0.000096 time 0.2947 (0.2968) loss 2.1762 (3.1166) grad_norm 2.4279 (nan) [2021-04-16 15:40:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][1150/1251] eta 0:00:29 lr 0.000096 time 0.2754 (0.2967) loss 3.5640 (3.1166) grad_norm 2.5778 (nan) [2021-04-16 15:40:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][1160/1251] eta 0:00:26 lr 0.000096 time 0.2818 (0.2965) loss 3.1122 (3.1171) grad_norm 2.5063 (nan) [2021-04-16 15:40:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][1170/1251] eta 0:00:24 lr 0.000096 time 0.2786 (0.2966) loss 3.5161 (3.1166) grad_norm 2.5215 (nan) [2021-04-16 15:40:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][1180/1251] eta 0:00:21 lr 0.000096 time 0.2684 (0.2965) loss 2.3671 (3.1172) grad_norm 2.6537 (nan) [2021-04-16 15:40:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][1190/1251] eta 0:00:18 lr 0.000096 time 0.2713 (0.2963) loss 3.7272 (3.1189) grad_norm 2.7007 (nan) [2021-04-16 15:40:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][1200/1251] eta 0:00:15 lr 0.000096 time 0.2913 (0.2961) loss 2.4805 (3.1178) grad_norm 2.3925 (nan) [2021-04-16 15:40:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][1210/1251] eta 0:00:12 lr 0.000096 time 0.2638 (0.2959) loss 3.4705 (3.1170) grad_norm 3.2375 (nan) [2021-04-16 15:40:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][1220/1251] eta 0:00:09 lr 0.000096 time 0.2954 (0.2958) loss 2.4976 (3.1175) grad_norm 2.7798 (nan) [2021-04-16 15:40:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][1230/1251] eta 0:00:06 lr 0.000096 time 0.2883 (0.2956) loss 3.0540 (3.1158) grad_norm 3.0327 (nan) [2021-04-16 15:40:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][1240/1251] eta 0:00:03 lr 0.000096 time 0.2482 (0.2954) loss 2.9644 (3.1135) grad_norm 3.2828 (nan) [2021-04-16 15:40:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [242/300][1250/1251] eta 0:00:00 lr 0.000096 time 0.2624 (0.2951) loss 2.3354 (3.1124) grad_norm 3.3122 (nan) [2021-04-16 15:41:09 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 242 training takes 0:06:23 [2021-04-16 15:41:09 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_242.pth saving...... [2021-04-16 15:41:32 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_242.pth saved !!! [2021-04-16 15:41:34 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.064 (1.064) Loss 0.8975 (0.8975) Acc@1 79.883 (79.883) Acc@5 94.824 (94.824) [2021-04-16 15:41:35 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.304 (0.223) Loss 0.8686 (0.8706) Acc@1 80.566 (79.608) Acc@5 95.703 (95.046) [2021-04-16 15:41:37 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.759 (0.240) Loss 0.8303 (0.8563) Acc@1 80.664 (80.083) Acc@5 95.898 (95.099) [2021-04-16 15:41:40 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.151 (0.230) Loss 0.8423 (0.8621) Acc@1 80.469 (79.962) Acc@5 95.215 (95.045) [2021-04-16 15:41:41 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.128 (0.219) Loss 0.8179 (0.8595) Acc@1 81.250 (80.083) Acc@5 95.020 (95.008) [2021-04-16 15:41:59 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.086 Acc@5 95.092 [2021-04-16 15:41:59 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 80.1% [2021-04-16 15:41:59 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 80.15% [2021-04-16 15:42:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][0/1251] eta 1:36:54 lr 0.000096 time 4.6476 (4.6476) loss 3.5667 (3.5667) grad_norm 2.7050 (2.7050) [2021-04-16 15:42:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][10/1251] eta 0:14:07 lr 0.000096 time 0.4058 (0.6831) loss 3.1785 (3.0215) grad_norm 2.8568 (2.6399) [2021-04-16 15:42:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][20/1251] eta 0:10:04 lr 0.000096 time 0.2698 (0.4909) loss 2.6347 (3.0558) grad_norm 2.6075 (2.6674) [2021-04-16 15:42:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][30/1251] eta 0:08:33 lr 0.000096 time 0.2758 (0.4208) loss 2.1169 (3.1256) grad_norm 2.6199 (2.6696) [2021-04-16 15:42:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][40/1251] eta 0:07:50 lr 0.000096 time 0.2921 (0.3884) loss 3.0718 (3.1140) grad_norm 2.9759 (2.7043) [2021-04-16 15:42:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][50/1251] eta 0:07:20 lr 0.000095 time 0.2879 (0.3664) loss 2.0865 (3.1062) grad_norm 2.6457 (2.6777) [2021-04-16 15:42:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][60/1251] eta 0:06:58 lr 0.000095 time 0.2896 (0.3515) loss 3.2129 (3.1408) grad_norm 2.8251 (2.6666) [2021-04-16 15:42:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][70/1251] eta 0:06:44 lr 0.000095 time 0.2924 (0.3422) loss 2.6750 (3.1158) grad_norm 2.2536 (2.6466) [2021-04-16 15:42:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][80/1251] eta 0:06:32 lr 0.000095 time 0.2973 (0.3356) loss 3.5944 (3.1328) grad_norm 2.5177 (2.6540) [2021-04-16 15:42:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][90/1251] eta 0:06:23 lr 0.000095 time 0.2770 (0.3302) loss 2.5540 (3.1138) grad_norm 2.3607 (2.6403) [2021-04-16 15:42:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][100/1251] eta 0:06:14 lr 0.000095 time 0.2531 (0.3254) loss 3.4082 (3.1295) grad_norm 2.8538 (2.6361) [2021-04-16 15:42:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][110/1251] eta 0:06:06 lr 0.000095 time 0.2600 (0.3208) loss 3.4118 (3.1133) grad_norm 2.7672 (2.6821) [2021-04-16 15:42:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][120/1251] eta 0:05:58 lr 0.000095 time 0.2626 (0.3171) loss 3.2573 (3.1202) grad_norm 2.2207 (2.6626) [2021-04-16 15:42:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][130/1251] eta 0:05:51 lr 0.000095 time 0.2676 (0.3140) loss 3.4689 (3.1256) grad_norm 3.1673 (2.6615) [2021-04-16 15:42:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][140/1251] eta 0:05:47 lr 0.000095 time 0.2961 (0.3130) loss 3.3562 (3.1088) grad_norm 5.5556 (2.6995) [2021-04-16 15:42:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][150/1251] eta 0:05:43 lr 0.000095 time 0.2846 (0.3120) loss 1.7908 (3.0948) grad_norm 2.7059 (2.6914) [2021-04-16 15:42:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][160/1251] eta 0:05:37 lr 0.000095 time 0.2752 (0.3098) loss 2.1953 (3.0893) grad_norm 2.4798 (2.6933) [2021-04-16 15:42:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][170/1251] eta 0:05:33 lr 0.000095 time 0.2910 (0.3082) loss 2.9547 (3.0912) grad_norm 2.5296 (2.6917) [2021-04-16 15:42:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][180/1251] eta 0:05:28 lr 0.000095 time 0.4390 (0.3072) loss 2.6709 (3.0828) grad_norm 2.7659 (2.6883) [2021-04-16 15:42:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][190/1251] eta 0:05:23 lr 0.000095 time 0.2720 (0.3053) loss 2.4683 (3.0787) grad_norm 2.4805 (2.6840) [2021-04-16 15:43:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][200/1251] eta 0:05:19 lr 0.000095 time 0.2519 (0.3039) loss 2.6096 (3.0943) grad_norm 2.5763 (2.6794) [2021-04-16 15:43:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][210/1251] eta 0:05:15 lr 0.000095 time 0.2921 (0.3026) loss 2.4470 (3.0981) grad_norm 3.5982 (2.6793) [2021-04-16 15:43:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][220/1251] eta 0:05:11 lr 0.000095 time 0.2754 (0.3017) loss 3.0751 (3.1018) grad_norm 3.2322 (2.6806) [2021-04-16 15:43:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][230/1251] eta 0:05:06 lr 0.000095 time 0.2749 (0.3006) loss 3.3431 (3.1031) grad_norm 2.6130 (2.6794) [2021-04-16 15:43:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][240/1251] eta 0:05:02 lr 0.000095 time 0.2764 (0.2996) loss 3.5740 (3.1137) grad_norm 2.8538 (2.6751) [2021-04-16 15:43:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][250/1251] eta 0:04:59 lr 0.000095 time 0.2609 (0.2988) loss 2.7967 (3.1061) grad_norm 3.1254 (2.6782) [2021-04-16 15:43:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][260/1251] eta 0:04:55 lr 0.000095 time 0.2665 (0.2979) loss 2.4227 (3.1033) grad_norm 2.6916 (2.6848) [2021-04-16 15:43:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][270/1251] eta 0:04:51 lr 0.000095 time 0.2574 (0.2970) loss 3.2091 (3.1073) grad_norm 2.7161 (2.6925) [2021-04-16 15:43:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][280/1251] eta 0:04:47 lr 0.000095 time 0.2782 (0.2964) loss 2.8431 (3.1216) grad_norm 2.6855 (2.6954) [2021-04-16 15:43:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][290/1251] eta 0:04:44 lr 0.000095 time 0.2813 (0.2964) loss 3.2786 (3.1211) grad_norm 3.1149 (2.6955) [2021-04-16 15:43:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][300/1251] eta 0:04:41 lr 0.000095 time 0.2645 (0.2958) loss 3.9572 (3.1282) grad_norm 2.5357 (2.6943) [2021-04-16 15:43:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][310/1251] eta 0:04:37 lr 0.000095 time 0.2949 (0.2951) loss 2.7893 (3.1323) grad_norm 2.6096 (2.6936) [2021-04-16 15:43:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][320/1251] eta 0:04:34 lr 0.000095 time 0.2635 (0.2944) loss 2.3516 (3.1305) grad_norm 2.8918 (2.6938) [2021-04-16 15:43:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][330/1251] eta 0:04:30 lr 0.000095 time 0.2430 (0.2938) loss 3.5774 (3.1287) grad_norm 2.5199 (2.6972) [2021-04-16 15:43:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][340/1251] eta 0:04:27 lr 0.000095 time 0.2667 (0.2933) loss 3.0673 (3.1244) grad_norm 2.2539 (2.6967) [2021-04-16 15:43:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][350/1251] eta 0:04:23 lr 0.000095 time 0.2844 (0.2927) loss 3.7322 (3.1288) grad_norm 2.1938 (2.6956) [2021-04-16 15:43:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][360/1251] eta 0:04:20 lr 0.000095 time 0.2630 (0.2922) loss 3.7540 (3.1292) grad_norm 2.8336 (2.6968) [2021-04-16 15:43:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][370/1251] eta 0:04:17 lr 0.000095 time 0.2775 (0.2921) loss 3.1207 (3.1313) grad_norm 2.9353 (2.6989) [2021-04-16 15:43:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][380/1251] eta 0:04:14 lr 0.000095 time 0.2667 (0.2919) loss 3.1042 (3.1306) grad_norm 2.5560 (2.6968) [2021-04-16 15:43:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][390/1251] eta 0:04:11 lr 0.000095 time 0.2890 (0.2915) loss 3.4060 (3.1251) grad_norm 4.0177 (2.7009) [2021-04-16 15:43:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][400/1251] eta 0:04:07 lr 0.000095 time 0.2589 (0.2910) loss 3.5445 (3.1226) grad_norm 2.6537 (2.6983) [2021-04-16 15:43:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][410/1251] eta 0:04:04 lr 0.000095 time 0.2918 (0.2907) loss 3.5701 (3.1198) grad_norm 2.7396 (2.6966) [2021-04-16 15:44:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][420/1251] eta 0:04:01 lr 0.000095 time 0.2678 (0.2905) loss 3.7474 (3.1205) grad_norm 2.6322 (2.6980) [2021-04-16 15:44:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][430/1251] eta 0:03:58 lr 0.000095 time 0.2932 (0.2902) loss 3.3344 (3.1186) grad_norm 2.8463 (2.6993) [2021-04-16 15:44:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][440/1251] eta 0:03:55 lr 0.000095 time 0.2711 (0.2898) loss 3.2222 (3.1271) grad_norm 2.4784 (2.6986) [2021-04-16 15:44:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][450/1251] eta 0:03:51 lr 0.000095 time 0.2546 (0.2896) loss 3.5267 (3.1222) grad_norm 2.4970 (2.6964) [2021-04-16 15:44:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][460/1251] eta 0:03:48 lr 0.000095 time 0.2733 (0.2894) loss 3.9976 (3.1255) grad_norm 2.3271 (2.6926) [2021-04-16 15:44:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][470/1251] eta 0:03:45 lr 0.000095 time 0.2760 (0.2892) loss 3.3204 (3.1289) grad_norm 2.2934 (2.6890) [2021-04-16 15:44:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][480/1251] eta 0:03:42 lr 0.000094 time 0.2707 (0.2891) loss 2.9733 (3.1296) grad_norm 2.8023 (2.6861) [2021-04-16 15:44:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][490/1251] eta 0:03:39 lr 0.000094 time 0.2760 (0.2891) loss 2.9194 (3.1270) grad_norm 2.4803 (2.6858) [2021-04-16 15:44:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][500/1251] eta 0:03:36 lr 0.000094 time 0.2757 (0.2888) loss 2.6426 (3.1290) grad_norm 2.8498 (2.6843) [2021-04-16 15:44:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][510/1251] eta 0:03:33 lr 0.000094 time 0.2915 (0.2886) loss 3.9314 (3.1326) grad_norm 2.5140 (2.6837) [2021-04-16 15:44:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][520/1251] eta 0:03:31 lr 0.000094 time 0.2772 (0.2887) loss 3.3570 (3.1290) grad_norm 2.3778 (2.6803) [2021-04-16 15:44:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][530/1251] eta 0:03:28 lr 0.000094 time 0.2898 (0.2885) loss 2.3767 (3.1290) grad_norm 2.3663 (2.6855) [2021-04-16 15:44:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][540/1251] eta 0:03:24 lr 0.000094 time 0.2989 (0.2883) loss 3.1411 (3.1295) grad_norm 2.6808 (2.6830) [2021-04-16 15:44:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][550/1251] eta 0:03:22 lr 0.000094 time 0.4363 (0.2883) loss 2.7274 (3.1293) grad_norm 2.3730 (2.6842) [2021-04-16 15:44:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][560/1251] eta 0:03:19 lr 0.000094 time 0.2764 (0.2880) loss 3.1167 (3.1229) grad_norm 2.9435 (2.6885) [2021-04-16 15:44:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][570/1251] eta 0:03:16 lr 0.000094 time 0.2735 (0.2879) loss 2.3256 (3.1201) grad_norm 3.0543 (2.6884) [2021-04-16 15:44:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][580/1251] eta 0:03:13 lr 0.000094 time 0.2586 (0.2877) loss 3.4238 (3.1191) grad_norm 2.6513 (2.6907) [2021-04-16 15:44:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][590/1251] eta 0:03:10 lr 0.000094 time 0.3008 (0.2878) loss 3.8233 (3.1242) grad_norm 2.7181 (2.6881) [2021-04-16 15:44:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][600/1251] eta 0:03:07 lr 0.000094 time 0.3059 (0.2876) loss 3.4182 (3.1275) grad_norm 2.6236 (2.6856) [2021-04-16 15:44:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][610/1251] eta 0:03:04 lr 0.000094 time 0.2699 (0.2873) loss 3.5797 (3.1270) grad_norm 2.7520 (2.6877) [2021-04-16 15:44:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][620/1251] eta 0:03:01 lr 0.000094 time 0.2975 (0.2873) loss 3.4542 (3.1303) grad_norm 2.5175 (2.6877) [2021-04-16 15:45:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][630/1251] eta 0:02:58 lr 0.000094 time 0.2795 (0.2870) loss 1.9061 (3.1301) grad_norm 2.3605 (2.6890) [2021-04-16 15:45:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][640/1251] eta 0:02:55 lr 0.000094 time 0.2712 (0.2869) loss 3.8730 (3.1344) grad_norm 3.2816 (2.6912) [2021-04-16 15:45:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][650/1251] eta 0:02:52 lr 0.000094 time 0.3924 (0.2870) loss 3.4521 (3.1370) grad_norm 2.1447 (2.6891) [2021-04-16 15:45:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][660/1251] eta 0:02:49 lr 0.000094 time 0.2678 (0.2867) loss 3.3355 (3.1389) grad_norm 2.2389 (2.6861) [2021-04-16 15:45:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][670/1251] eta 0:02:46 lr 0.000094 time 0.2772 (0.2868) loss 3.3705 (3.1353) grad_norm 2.5353 (2.6878) [2021-04-16 15:45:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][680/1251] eta 0:02:43 lr 0.000094 time 0.2941 (0.2867) loss 3.8584 (3.1337) grad_norm 2.5269 (2.6915) [2021-04-16 15:45:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][690/1251] eta 0:02:40 lr 0.000094 time 0.2680 (0.2865) loss 3.5143 (3.1349) grad_norm 2.8661 (2.6901) [2021-04-16 15:45:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][700/1251] eta 0:02:37 lr 0.000094 time 0.2814 (0.2863) loss 3.3715 (3.1357) grad_norm 2.4332 (2.6883) [2021-04-16 15:45:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][710/1251] eta 0:02:34 lr 0.000094 time 0.2872 (0.2863) loss 2.3111 (3.1341) grad_norm 2.8530 (2.6881) [2021-04-16 15:45:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][720/1251] eta 0:02:32 lr 0.000094 time 0.2863 (0.2865) loss 3.1741 (3.1342) grad_norm 2.7773 (2.6869) [2021-04-16 15:45:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][730/1251] eta 0:02:29 lr 0.000094 time 0.2776 (0.2865) loss 3.8214 (3.1347) grad_norm 2.5276 (2.6840) [2021-04-16 15:45:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][740/1251] eta 0:02:26 lr 0.000094 time 0.2697 (0.2865) loss 3.2987 (3.1373) grad_norm 2.2229 (2.6830) [2021-04-16 15:45:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][750/1251] eta 0:02:23 lr 0.000094 time 0.2915 (0.2863) loss 2.3325 (3.1332) grad_norm 2.3588 (2.6835) [2021-04-16 15:45:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][760/1251] eta 0:02:20 lr 0.000094 time 0.2711 (0.2864) loss 3.2447 (3.1303) grad_norm 2.4558 (2.6816) [2021-04-16 15:45:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][770/1251] eta 0:02:17 lr 0.000094 time 0.2774 (0.2862) loss 3.1244 (3.1314) grad_norm 2.8060 (2.6813) [2021-04-16 15:45:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][780/1251] eta 0:02:14 lr 0.000094 time 0.2695 (0.2861) loss 2.2563 (3.1279) grad_norm 3.1257 (2.6842) [2021-04-16 15:45:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][790/1251] eta 0:02:11 lr 0.000094 time 0.2873 (0.2861) loss 2.3111 (3.1230) grad_norm 2.2362 (2.6836) [2021-04-16 15:45:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][800/1251] eta 0:02:08 lr 0.000094 time 0.2729 (0.2860) loss 3.4149 (3.1214) grad_norm 2.5586 (2.6820) [2021-04-16 15:45:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][810/1251] eta 0:02:06 lr 0.000094 time 0.2950 (0.2858) loss 3.4386 (3.1228) grad_norm 2.6729 (2.6806) [2021-04-16 15:45:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][820/1251] eta 0:02:03 lr 0.000094 time 0.2567 (0.2857) loss 3.0234 (3.1233) grad_norm 2.4475 (2.6788) [2021-04-16 15:45:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][830/1251] eta 0:02:00 lr 0.000094 time 0.2608 (0.2855) loss 3.1091 (3.1224) grad_norm 2.6008 (2.6773) [2021-04-16 15:45:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][840/1251] eta 0:01:57 lr 0.000094 time 0.2617 (0.2854) loss 3.2001 (3.1244) grad_norm 3.1286 (2.6776) [2021-04-16 15:46:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][850/1251] eta 0:01:54 lr 0.000094 time 0.2670 (0.2853) loss 3.5784 (3.1246) grad_norm 2.8084 (2.6773) [2021-04-16 15:46:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][860/1251] eta 0:01:51 lr 0.000094 time 0.2627 (0.2851) loss 3.0799 (3.1246) grad_norm 3.7259 (2.6765) [2021-04-16 15:46:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][870/1251] eta 0:01:48 lr 0.000094 time 0.2928 (0.2852) loss 3.5598 (3.1263) grad_norm 2.9764 (2.6760) [2021-04-16 15:46:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][880/1251] eta 0:01:45 lr 0.000094 time 0.2891 (0.2851) loss 1.9636 (3.1246) grad_norm 2.4070 (2.6786) [2021-04-16 15:46:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][890/1251] eta 0:01:42 lr 0.000094 time 0.2564 (0.2849) loss 2.4298 (3.1205) grad_norm 2.3394 (2.6784) [2021-04-16 15:46:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][900/1251] eta 0:01:39 lr 0.000094 time 0.2775 (0.2848) loss 1.8344 (3.1213) grad_norm 2.4985 (2.6788) [2021-04-16 15:46:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][910/1251] eta 0:01:37 lr 0.000093 time 0.2726 (0.2847) loss 3.9533 (3.1229) grad_norm 2.7280 (2.6801) [2021-04-16 15:46:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][920/1251] eta 0:01:34 lr 0.000093 time 0.2549 (0.2847) loss 2.4812 (3.1248) grad_norm 2.3410 (2.6795) [2021-04-16 15:46:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][930/1251] eta 0:01:31 lr 0.000093 time 0.2969 (0.2847) loss 2.5320 (3.1255) grad_norm 2.9154 (2.6802) [2021-04-16 15:46:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][940/1251] eta 0:01:28 lr 0.000093 time 0.3695 (0.2847) loss 3.6342 (3.1254) grad_norm 2.3148 (2.6787) [2021-04-16 15:46:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][950/1251] eta 0:01:25 lr 0.000093 time 0.2855 (0.2846) loss 2.9287 (3.1202) grad_norm 2.6103 (2.6773) [2021-04-16 15:46:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][960/1251] eta 0:01:22 lr 0.000093 time 0.2860 (0.2845) loss 3.8159 (3.1234) grad_norm 2.4929 (nan) [2021-04-16 15:46:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][970/1251] eta 0:01:19 lr 0.000093 time 0.2968 (0.2845) loss 3.1227 (3.1253) grad_norm 3.0566 (nan) [2021-04-16 15:46:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][980/1251] eta 0:01:17 lr 0.000093 time 0.2672 (0.2843) loss 3.2943 (3.1282) grad_norm 2.5864 (nan) [2021-04-16 15:46:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][990/1251] eta 0:01:14 lr 0.000093 time 0.2836 (0.2843) loss 3.3320 (3.1286) grad_norm 2.8962 (nan) [2021-04-16 15:46:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][1000/1251] eta 0:01:11 lr 0.000093 time 0.2884 (0.2842) loss 2.3914 (3.1273) grad_norm 3.0248 (nan) [2021-04-16 15:46:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][1010/1251] eta 0:01:08 lr 0.000093 time 0.2913 (0.2842) loss 3.7013 (3.1305) grad_norm 2.3099 (nan) [2021-04-16 15:46:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][1020/1251] eta 0:01:05 lr 0.000093 time 0.2997 (0.2841) loss 1.8705 (3.1306) grad_norm 2.8641 (nan) [2021-04-16 15:46:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][1030/1251] eta 0:01:02 lr 0.000093 time 0.3236 (0.2840) loss 3.6469 (3.1311) grad_norm 2.3700 (nan) [2021-04-16 15:46:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][1040/1251] eta 0:00:59 lr 0.000093 time 0.2615 (0.2839) loss 3.3681 (3.1315) grad_norm 2.5377 (nan) [2021-04-16 15:46:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][1050/1251] eta 0:00:57 lr 0.000093 time 0.2894 (0.2838) loss 3.7895 (3.1306) grad_norm 2.3435 (nan) [2021-04-16 15:47:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][1060/1251] eta 0:00:54 lr 0.000093 time 0.2845 (0.2838) loss 3.6082 (3.1292) grad_norm 2.5776 (nan) [2021-04-16 15:47:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][1070/1251] eta 0:00:51 lr 0.000093 time 0.2719 (0.2839) loss 3.6386 (3.1311) grad_norm 3.2550 (nan) [2021-04-16 15:47:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][1080/1251] eta 0:00:48 lr 0.000093 time 0.2450 (0.2838) loss 2.1849 (3.1296) grad_norm 2.8435 (nan) [2021-04-16 15:47:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][1090/1251] eta 0:00:45 lr 0.000093 time 0.2692 (0.2838) loss 3.7755 (3.1303) grad_norm 3.2307 (nan) [2021-04-16 15:47:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][1100/1251] eta 0:00:42 lr 0.000093 time 0.2605 (0.2837) loss 3.4052 (3.1327) grad_norm 2.9586 (nan) [2021-04-16 15:47:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][1110/1251] eta 0:00:39 lr 0.000093 time 0.2718 (0.2837) loss 3.3293 (3.1338) grad_norm 2.5220 (nan) [2021-04-16 15:47:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][1120/1251] eta 0:00:37 lr 0.000093 time 0.2669 (0.2837) loss 3.7103 (3.1357) grad_norm 2.7712 (nan) [2021-04-16 15:47:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][1130/1251] eta 0:00:34 lr 0.000093 time 0.2870 (0.2837) loss 3.3194 (3.1339) grad_norm 2.3083 (nan) [2021-04-16 15:47:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][1140/1251] eta 0:00:31 lr 0.000093 time 0.2981 (0.2836) loss 2.3220 (3.1318) grad_norm 2.6202 (nan) [2021-04-16 15:47:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][1150/1251] eta 0:00:28 lr 0.000093 time 0.2848 (0.2836) loss 3.1803 (3.1305) grad_norm 2.4868 (nan) [2021-04-16 15:47:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][1160/1251] eta 0:00:25 lr 0.000093 time 0.2846 (0.2835) loss 3.1023 (3.1310) grad_norm 2.5468 (nan) [2021-04-16 15:47:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][1170/1251] eta 0:00:22 lr 0.000093 time 0.2759 (0.2836) loss 3.2351 (3.1317) grad_norm 2.3951 (nan) [2021-04-16 15:47:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][1180/1251] eta 0:00:20 lr 0.000093 time 0.2656 (0.2836) loss 2.6079 (3.1331) grad_norm 3.0033 (nan) [2021-04-16 15:47:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][1190/1251] eta 0:00:17 lr 0.000093 time 0.2726 (0.2837) loss 2.7559 (3.1333) grad_norm 2.4975 (nan) [2021-04-16 15:47:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][1200/1251] eta 0:00:14 lr 0.000093 time 0.2663 (0.2836) loss 3.3899 (3.1348) grad_norm 2.7345 (nan) [2021-04-16 15:47:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][1210/1251] eta 0:00:11 lr 0.000093 time 0.2696 (0.2835) loss 3.5694 (3.1353) grad_norm 2.3357 (nan) [2021-04-16 15:47:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][1220/1251] eta 0:00:08 lr 0.000093 time 0.2881 (0.2834) loss 3.4875 (3.1321) grad_norm 2.3708 (nan) [2021-04-16 15:47:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][1230/1251] eta 0:00:05 lr 0.000093 time 0.2635 (0.2835) loss 2.2121 (3.1310) grad_norm 2.3337 (nan) [2021-04-16 15:47:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][1240/1251] eta 0:00:03 lr 0.000093 time 0.3341 (0.2834) loss 2.9241 (3.1314) grad_norm 2.4400 (nan) [2021-04-16 15:47:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [243/300][1250/1251] eta 0:00:00 lr 0.000093 time 0.2473 (0.2831) loss 3.3926 (3.1282) grad_norm 2.5751 (nan) [2021-04-16 15:48:06 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 243 training takes 0:06:06 [2021-04-16 15:48:06 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_243.pth saving...... [2021-04-16 15:48:22 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_243.pth saved !!! [2021-04-16 15:48:23 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.164 (1.164) Loss 0.8567 (0.8567) Acc@1 80.176 (80.176) Acc@5 95.508 (95.508) [2021-04-16 15:48:25 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.117 (0.229) Loss 0.8235 (0.8518) Acc@1 79.688 (80.105) Acc@5 95.996 (95.179) [2021-04-16 15:48:27 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.328 (0.211) Loss 0.7959 (0.8461) Acc@1 81.641 (80.325) Acc@5 95.410 (95.168) [2021-04-16 15:48:29 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.448 (0.233) Loss 0.7897 (0.8467) Acc@1 79.785 (80.258) Acc@5 95.312 (95.117) [2021-04-16 15:48:31 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.216) Loss 0.7852 (0.8443) Acc@1 81.934 (80.245) Acc@5 95.801 (95.153) [2021-04-16 15:48:45 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.106 Acc@5 95.136 [2021-04-16 15:48:45 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 80.1% [2021-04-16 15:48:45 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 80.15% [2021-04-16 15:48:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][0/1251] eta 3:58:28 lr 0.000093 time 11.4376 (11.4376) loss 2.3429 (2.3429) grad_norm 2.6606 (2.6606) [2021-04-16 15:48:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][10/1251] eta 0:27:02 lr 0.000093 time 0.4358 (1.3076) loss 3.5631 (3.1749) grad_norm 2.4958 (2.5598) [2021-04-16 15:49:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][20/1251] eta 0:16:41 lr 0.000093 time 0.2746 (0.8137) loss 3.0996 (3.1461) grad_norm 3.2279 (2.6967) [2021-04-16 15:49:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][30/1251] eta 0:13:02 lr 0.000093 time 0.2717 (0.6406) loss 3.3596 (3.1099) grad_norm 2.5439 (2.6935) [2021-04-16 15:49:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][40/1251] eta 0:11:07 lr 0.000093 time 0.2736 (0.5514) loss 3.3703 (3.1034) grad_norm 2.7418 (2.6878) [2021-04-16 15:49:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][50/1251] eta 0:09:57 lr 0.000093 time 0.2849 (0.4972) loss 2.7780 (3.0735) grad_norm 2.6761 (2.6908) [2021-04-16 15:49:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][60/1251] eta 0:09:09 lr 0.000093 time 0.2956 (0.4613) loss 2.2136 (3.0627) grad_norm 2.3287 (2.7010) [2021-04-16 15:49:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][70/1251] eta 0:08:34 lr 0.000093 time 0.2648 (0.4353) loss 2.6533 (3.0447) grad_norm 2.2187 (2.6888) [2021-04-16 15:49:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][80/1251] eta 0:08:09 lr 0.000093 time 0.2624 (0.4177) loss 3.5526 (3.0768) grad_norm 2.7739 (2.6941) [2021-04-16 15:49:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][90/1251] eta 0:07:48 lr 0.000092 time 0.2818 (0.4033) loss 3.2395 (3.0724) grad_norm 2.7861 (2.7035) [2021-04-16 15:49:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][100/1251] eta 0:07:29 lr 0.000092 time 0.2785 (0.3909) loss 2.7657 (3.0819) grad_norm 2.2872 (2.7048) [2021-04-16 15:49:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][110/1251] eta 0:07:14 lr 0.000092 time 0.2991 (0.3804) loss 3.3677 (3.0760) grad_norm 2.8786 (2.7012) [2021-04-16 15:49:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][120/1251] eta 0:07:00 lr 0.000092 time 0.2745 (0.3716) loss 3.3447 (3.0510) grad_norm 2.3445 (2.6880) [2021-04-16 15:49:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][130/1251] eta 0:06:48 lr 0.000092 time 0.3058 (0.3645) loss 3.1023 (3.0439) grad_norm 2.4452 (2.6838) [2021-04-16 15:49:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][140/1251] eta 0:06:37 lr 0.000092 time 0.2620 (0.3578) loss 3.2318 (3.0470) grad_norm 3.3296 (2.6880) [2021-04-16 15:49:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][150/1251] eta 0:06:29 lr 0.000092 time 0.4245 (0.3534) loss 3.1089 (3.0372) grad_norm 2.6769 (2.6861) [2021-04-16 15:49:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][160/1251] eta 0:06:19 lr 0.000092 time 0.2776 (0.3483) loss 3.5705 (3.0612) grad_norm 2.3956 (2.6769) [2021-04-16 15:49:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][170/1251] eta 0:06:12 lr 0.000092 time 0.2798 (0.3450) loss 3.8476 (3.0731) grad_norm 2.4242 (2.6761) [2021-04-16 15:49:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][180/1251] eta 0:06:05 lr 0.000092 time 0.2638 (0.3411) loss 3.4841 (3.0899) grad_norm 2.7022 (2.6853) [2021-04-16 15:49:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][190/1251] eta 0:05:58 lr 0.000092 time 0.2652 (0.3378) loss 2.3033 (3.0674) grad_norm 2.2882 (2.6819) [2021-04-16 15:49:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][200/1251] eta 0:05:51 lr 0.000092 time 0.2702 (0.3347) loss 3.0179 (3.0731) grad_norm 3.1455 (2.6808) [2021-04-16 15:49:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][210/1251] eta 0:05:45 lr 0.000092 time 0.2634 (0.3319) loss 3.3046 (3.0708) grad_norm 2.7256 (2.6796) [2021-04-16 15:49:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][220/1251] eta 0:05:39 lr 0.000092 time 0.2784 (0.3292) loss 3.0038 (3.0791) grad_norm 2.7285 (2.6759) [2021-04-16 15:50:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][230/1251] eta 0:05:33 lr 0.000092 time 0.2825 (0.3269) loss 3.2061 (3.0760) grad_norm 2.5084 (2.6775) [2021-04-16 15:50:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][240/1251] eta 0:05:29 lr 0.000092 time 0.2949 (0.3254) loss 3.4294 (3.0827) grad_norm 2.5904 (2.6799) [2021-04-16 15:50:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][250/1251] eta 0:05:23 lr 0.000092 time 0.2645 (0.3234) loss 3.1288 (3.0745) grad_norm 2.8448 (2.6825) [2021-04-16 15:50:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][260/1251] eta 0:05:18 lr 0.000092 time 0.2841 (0.3217) loss 2.7155 (3.0738) grad_norm 2.3103 (2.6865) [2021-04-16 15:50:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][270/1251] eta 0:05:14 lr 0.000092 time 0.2766 (0.3203) loss 3.5732 (3.0740) grad_norm 2.7089 (2.6843) [2021-04-16 15:50:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][280/1251] eta 0:05:09 lr 0.000092 time 0.2874 (0.3192) loss 3.2780 (3.0742) grad_norm 3.0234 (2.6884) [2021-04-16 15:50:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][290/1251] eta 0:05:05 lr 0.000092 time 0.2873 (0.3177) loss 3.3003 (3.0829) grad_norm 2.5926 (2.6855) [2021-04-16 15:50:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][300/1251] eta 0:05:01 lr 0.000092 time 0.2735 (0.3169) loss 3.6746 (3.0887) grad_norm 2.6967 (2.6838) [2021-04-16 15:50:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][310/1251] eta 0:04:56 lr 0.000092 time 0.2552 (0.3156) loss 3.1191 (3.0871) grad_norm 2.6330 (2.6851) [2021-04-16 15:50:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][320/1251] eta 0:04:52 lr 0.000092 time 0.2806 (0.3144) loss 3.4056 (3.0814) grad_norm 2.5286 (2.6855) [2021-04-16 15:50:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][330/1251] eta 0:04:48 lr 0.000092 time 0.2883 (0.3134) loss 2.9470 (3.0797) grad_norm 2.4398 (2.6955) [2021-04-16 15:50:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][340/1251] eta 0:04:44 lr 0.000092 time 0.2596 (0.3128) loss 3.4822 (3.0853) grad_norm 2.8649 (2.6979) [2021-04-16 15:50:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][350/1251] eta 0:04:40 lr 0.000092 time 0.2821 (0.3117) loss 3.6427 (3.0835) grad_norm 2.8137 (2.6986) [2021-04-16 15:50:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][360/1251] eta 0:04:36 lr 0.000092 time 0.2770 (0.3106) loss 3.3833 (3.0878) grad_norm 2.5213 (2.6971) [2021-04-16 15:50:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][370/1251] eta 0:04:33 lr 0.000092 time 0.2897 (0.3100) loss 3.3143 (3.0922) grad_norm 2.1658 (2.6968) [2021-04-16 15:50:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][380/1251] eta 0:04:29 lr 0.000092 time 0.2767 (0.3093) loss 2.4547 (3.0873) grad_norm 2.4390 (2.6902) [2021-04-16 15:50:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][390/1251] eta 0:04:25 lr 0.000092 time 0.2713 (0.3086) loss 3.3529 (3.0802) grad_norm 2.1729 (2.6888) [2021-04-16 15:50:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][400/1251] eta 0:04:21 lr 0.000092 time 0.2624 (0.3078) loss 3.0097 (3.0790) grad_norm 2.7619 (2.6858) [2021-04-16 15:50:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][410/1251] eta 0:04:18 lr 0.000092 time 0.2596 (0.3069) loss 3.4097 (3.0780) grad_norm 2.3265 (2.6852) [2021-04-16 15:50:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][420/1251] eta 0:04:14 lr 0.000092 time 0.2706 (0.3062) loss 2.6589 (3.0741) grad_norm 2.3293 (2.6860) [2021-04-16 15:50:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][430/1251] eta 0:04:11 lr 0.000092 time 0.2824 (0.3058) loss 2.5731 (3.0651) grad_norm 2.5410 (2.6876) [2021-04-16 15:50:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][440/1251] eta 0:04:07 lr 0.000092 time 0.2803 (0.3051) loss 3.2632 (3.0622) grad_norm 2.7661 (2.6888) [2021-04-16 15:51:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][450/1251] eta 0:04:04 lr 0.000092 time 0.4294 (0.3048) loss 2.1110 (3.0608) grad_norm 2.5279 (2.6899) [2021-04-16 15:51:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][460/1251] eta 0:04:00 lr 0.000092 time 0.2838 (0.3041) loss 2.9525 (3.0667) grad_norm 2.7019 (2.6907) [2021-04-16 15:51:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][470/1251] eta 0:03:57 lr 0.000092 time 0.2852 (0.3036) loss 2.9818 (3.0661) grad_norm 2.9884 (2.6925) [2021-04-16 15:51:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][480/1251] eta 0:03:53 lr 0.000092 time 0.2875 (0.3030) loss 2.8468 (3.0687) grad_norm 2.4160 (2.6894) [2021-04-16 15:51:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][490/1251] eta 0:03:50 lr 0.000092 time 0.2509 (0.3023) loss 3.6388 (3.0747) grad_norm 2.6490 (2.6850) [2021-04-16 15:51:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][500/1251] eta 0:03:46 lr 0.000092 time 0.2917 (0.3019) loss 2.6785 (3.0700) grad_norm 2.9049 (2.6851) [2021-04-16 15:51:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][510/1251] eta 0:03:43 lr 0.000092 time 0.2833 (0.3014) loss 2.0739 (3.0702) grad_norm 2.3550 (2.6824) [2021-04-16 15:51:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][520/1251] eta 0:03:40 lr 0.000092 time 0.2725 (0.3010) loss 3.6159 (3.0722) grad_norm 2.7996 (2.6808) [2021-04-16 15:51:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][530/1251] eta 0:03:36 lr 0.000091 time 0.3042 (0.3006) loss 2.4455 (3.0739) grad_norm 2.6825 (2.6798) [2021-04-16 15:51:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][540/1251] eta 0:03:33 lr 0.000091 time 0.2668 (0.3002) loss 3.6850 (3.0785) grad_norm 3.0819 (2.6837) [2021-04-16 15:51:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][550/1251] eta 0:03:30 lr 0.000091 time 0.2820 (0.2998) loss 3.0372 (3.0836) grad_norm 2.4252 (2.6839) [2021-04-16 15:51:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][560/1251] eta 0:03:26 lr 0.000091 time 0.2568 (0.2993) loss 3.4730 (3.0857) grad_norm 2.9912 (2.6843) [2021-04-16 15:51:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][570/1251] eta 0:03:23 lr 0.000091 time 0.2772 (0.2993) loss 3.3882 (3.0902) grad_norm 2.4018 (2.6865) [2021-04-16 15:51:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][580/1251] eta 0:03:20 lr 0.000091 time 0.2767 (0.2991) loss 2.4492 (3.0909) grad_norm 2.6175 (2.6863) [2021-04-16 15:51:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][590/1251] eta 0:03:17 lr 0.000091 time 0.2798 (0.2990) loss 2.9883 (3.0892) grad_norm 2.4976 (2.6863) [2021-04-16 15:51:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][600/1251] eta 0:03:14 lr 0.000091 time 0.2572 (0.2986) loss 3.4611 (3.0841) grad_norm 2.5915 (2.6849) [2021-04-16 15:51:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][610/1251] eta 0:03:11 lr 0.000091 time 0.2776 (0.2983) loss 3.4601 (3.0870) grad_norm 2.5355 (2.6883) [2021-04-16 15:51:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][620/1251] eta 0:03:08 lr 0.000091 time 0.2755 (0.2979) loss 2.8885 (3.0859) grad_norm 2.3093 (2.6867) [2021-04-16 15:51:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][630/1251] eta 0:03:04 lr 0.000091 time 0.2710 (0.2976) loss 2.7139 (3.0875) grad_norm 2.7445 (2.6875) [2021-04-16 15:51:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][640/1251] eta 0:03:01 lr 0.000091 time 0.3050 (0.2972) loss 3.0701 (3.0888) grad_norm 3.3124 (2.6899) [2021-04-16 15:51:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][650/1251] eta 0:02:58 lr 0.000091 time 0.3009 (0.2971) loss 3.6620 (3.0885) grad_norm 3.1119 (2.6900) [2021-04-16 15:52:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][660/1251] eta 0:02:55 lr 0.000091 time 0.2745 (0.2967) loss 3.4294 (3.0900) grad_norm 2.6444 (2.6896) [2021-04-16 15:52:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][670/1251] eta 0:02:52 lr 0.000091 time 0.3054 (0.2965) loss 3.6009 (3.0889) grad_norm 3.4221 (2.6923) [2021-04-16 15:52:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][680/1251] eta 0:02:49 lr 0.000091 time 0.2538 (0.2961) loss 3.0084 (3.0876) grad_norm 2.3314 (2.6936) [2021-04-16 15:52:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][690/1251] eta 0:02:46 lr 0.000091 time 0.2647 (0.2959) loss 3.3306 (3.0863) grad_norm 2.9448 (2.6957) [2021-04-16 15:52:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][700/1251] eta 0:02:42 lr 0.000091 time 0.2655 (0.2957) loss 3.6720 (3.0868) grad_norm 2.5410 (2.6951) [2021-04-16 15:52:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][710/1251] eta 0:02:39 lr 0.000091 time 0.2570 (0.2954) loss 2.8940 (3.0846) grad_norm 3.2381 (2.6962) [2021-04-16 15:52:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][720/1251] eta 0:02:36 lr 0.000091 time 0.2620 (0.2952) loss 3.1576 (3.0873) grad_norm 2.9709 (2.6978) [2021-04-16 15:52:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][730/1251] eta 0:02:33 lr 0.000091 time 0.2726 (0.2951) loss 2.4714 (3.0833) grad_norm 2.2629 (2.6966) [2021-04-16 15:52:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][740/1251] eta 0:02:30 lr 0.000091 time 0.2801 (0.2950) loss 2.1559 (3.0841) grad_norm 2.4420 (2.6962) [2021-04-16 15:52:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][750/1251] eta 0:02:27 lr 0.000091 time 0.2576 (0.2947) loss 1.9906 (3.0828) grad_norm 2.6283 (2.6953) [2021-04-16 15:52:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][760/1251] eta 0:02:24 lr 0.000091 time 0.2932 (0.2945) loss 3.0448 (3.0802) grad_norm 2.2643 (2.6928) [2021-04-16 15:52:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][770/1251] eta 0:02:21 lr 0.000091 time 0.2681 (0.2942) loss 3.5774 (3.0806) grad_norm 2.5856 (2.6933) [2021-04-16 15:52:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][780/1251] eta 0:02:18 lr 0.000091 time 0.2814 (0.2940) loss 3.1280 (3.0795) grad_norm 3.2596 (2.6955) [2021-04-16 15:52:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][790/1251] eta 0:02:15 lr 0.000091 time 0.3089 (0.2937) loss 2.4496 (3.0732) grad_norm 2.9616 (2.6949) [2021-04-16 15:52:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][800/1251] eta 0:02:12 lr 0.000091 time 0.2676 (0.2937) loss 2.8229 (3.0726) grad_norm 3.0811 (2.6938) [2021-04-16 15:52:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][810/1251] eta 0:02:09 lr 0.000091 time 0.2577 (0.2935) loss 2.2408 (3.0716) grad_norm 2.7092 (2.6943) [2021-04-16 15:52:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][820/1251] eta 0:02:06 lr 0.000091 time 0.3012 (0.2933) loss 3.1701 (3.0709) grad_norm 3.4008 (2.6940) [2021-04-16 15:52:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][830/1251] eta 0:02:03 lr 0.000091 time 0.2593 (0.2930) loss 3.3166 (3.0732) grad_norm 2.2411 (2.6929) [2021-04-16 15:52:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][840/1251] eta 0:02:00 lr 0.000091 time 0.2994 (0.2929) loss 3.2054 (3.0761) grad_norm 2.4539 (2.6924) [2021-04-16 15:52:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][850/1251] eta 0:01:57 lr 0.000091 time 0.2538 (0.2926) loss 3.4787 (3.0734) grad_norm 3.9084 (2.6917) [2021-04-16 15:52:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][860/1251] eta 0:01:54 lr 0.000091 time 0.2820 (0.2926) loss 3.2003 (3.0730) grad_norm 2.6207 (2.6940) [2021-04-16 15:52:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][870/1251] eta 0:01:51 lr 0.000091 time 0.2632 (0.2924) loss 2.8018 (3.0737) grad_norm 2.7393 (2.6943) [2021-04-16 15:53:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][880/1251] eta 0:01:48 lr 0.000091 time 0.2767 (0.2923) loss 3.7287 (3.0727) grad_norm 2.8146 (2.6946) [2021-04-16 15:53:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][890/1251] eta 0:01:45 lr 0.000091 time 0.2676 (0.2921) loss 2.3091 (3.0740) grad_norm 2.5187 (2.6946) [2021-04-16 15:53:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][900/1251] eta 0:01:42 lr 0.000091 time 0.2724 (0.2919) loss 3.3726 (3.0763) grad_norm 2.3905 (2.6936) [2021-04-16 15:53:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][910/1251] eta 0:01:39 lr 0.000091 time 0.2729 (0.2917) loss 3.4691 (3.0765) grad_norm 2.5590 (2.6941) [2021-04-16 15:53:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][920/1251] eta 0:01:36 lr 0.000091 time 0.2599 (0.2917) loss 3.3023 (3.0749) grad_norm 4.0374 (2.6953) [2021-04-16 15:53:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][930/1251] eta 0:01:33 lr 0.000091 time 0.2769 (0.2915) loss 2.9493 (3.0768) grad_norm 3.0767 (2.6948) [2021-04-16 15:53:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][940/1251] eta 0:01:30 lr 0.000091 time 0.2658 (0.2915) loss 3.6475 (3.0757) grad_norm 2.8244 (2.6950) [2021-04-16 15:53:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][950/1251] eta 0:01:27 lr 0.000091 time 0.2614 (0.2913) loss 3.0991 (3.0757) grad_norm 2.5577 (2.6952) [2021-04-16 15:53:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][960/1251] eta 0:01:24 lr 0.000091 time 0.2690 (0.2911) loss 3.4277 (3.0774) grad_norm 2.7958 (2.6962) [2021-04-16 15:53:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][970/1251] eta 0:01:21 lr 0.000090 time 0.2807 (0.2911) loss 3.1084 (3.0773) grad_norm 2.6415 (2.6960) [2021-04-16 15:53:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][980/1251] eta 0:01:18 lr 0.000090 time 0.2588 (0.2909) loss 3.5225 (3.0749) grad_norm 3.0845 (2.6989) [2021-04-16 15:53:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][990/1251] eta 0:01:15 lr 0.000090 time 0.2703 (0.2907) loss 3.2097 (3.0743) grad_norm 2.6204 (2.6973) [2021-04-16 15:53:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][1000/1251] eta 0:01:12 lr 0.000090 time 0.2800 (0.2906) loss 2.7004 (3.0754) grad_norm 2.6050 (2.6973) [2021-04-16 15:53:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][1010/1251] eta 0:01:10 lr 0.000090 time 0.2499 (0.2905) loss 3.4459 (3.0731) grad_norm 2.9877 (2.6963) [2021-04-16 15:53:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][1020/1251] eta 0:01:07 lr 0.000090 time 0.2670 (0.2904) loss 3.5832 (3.0744) grad_norm 2.4703 (2.6958) [2021-04-16 15:53:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][1030/1251] eta 0:01:04 lr 0.000090 time 0.2827 (0.2905) loss 3.2376 (3.0757) grad_norm 5.8460 (2.6968) [2021-04-16 15:53:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][1040/1251] eta 0:01:01 lr 0.000090 time 0.2561 (0.2903) loss 2.5679 (3.0757) grad_norm 2.7472 (2.6977) [2021-04-16 15:53:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][1050/1251] eta 0:00:58 lr 0.000090 time 0.2757 (0.2901) loss 3.3847 (3.0788) grad_norm 2.5447 (2.6973) [2021-04-16 15:53:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][1060/1251] eta 0:00:55 lr 0.000090 time 0.2711 (0.2900) loss 3.1408 (3.0790) grad_norm 2.6764 (2.6960) [2021-04-16 15:53:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][1070/1251] eta 0:00:52 lr 0.000090 time 0.2765 (0.2899) loss 3.2818 (3.0787) grad_norm 2.6656 (2.6979) [2021-04-16 15:53:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][1080/1251] eta 0:00:49 lr 0.000090 time 0.3104 (0.2897) loss 3.5741 (3.0815) grad_norm 3.0260 (2.6999) [2021-04-16 15:54:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][1090/1251] eta 0:00:46 lr 0.000090 time 0.2660 (0.2898) loss 3.7522 (3.0819) grad_norm 2.8102 (2.7013) [2021-04-16 15:54:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][1100/1251] eta 0:00:43 lr 0.000090 time 0.2675 (0.2897) loss 2.9902 (3.0837) grad_norm 2.8218 (2.7003) [2021-04-16 15:54:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][1110/1251] eta 0:00:40 lr 0.000090 time 0.2814 (0.2896) loss 2.9820 (3.0836) grad_norm 2.7396 (2.6987) [2021-04-16 15:54:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][1120/1251] eta 0:00:37 lr 0.000090 time 0.2653 (0.2894) loss 3.6975 (3.0841) grad_norm 2.4514 (2.6979) [2021-04-16 15:54:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][1130/1251] eta 0:00:35 lr 0.000090 time 0.2890 (0.2894) loss 3.5168 (3.0863) grad_norm 2.9421 (2.6996) [2021-04-16 15:54:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][1140/1251] eta 0:00:32 lr 0.000090 time 0.2672 (0.2892) loss 3.4784 (3.0850) grad_norm 2.6023 (2.6993) [2021-04-16 15:54:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][1150/1251] eta 0:00:29 lr 0.000090 time 0.2856 (0.2894) loss 2.0649 (3.0858) grad_norm 2.4445 (2.6990) [2021-04-16 15:54:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][1160/1251] eta 0:00:26 lr 0.000090 time 0.3098 (0.2894) loss 3.4780 (3.0886) grad_norm 3.1130 (2.7008) [2021-04-16 15:54:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][1170/1251] eta 0:00:23 lr 0.000090 time 0.2826 (0.2894) loss 2.8568 (3.0900) grad_norm 2.6769 (2.7026) [2021-04-16 15:54:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][1180/1251] eta 0:00:20 lr 0.000090 time 0.3105 (0.2893) loss 3.2972 (3.0916) grad_norm 2.5822 (2.7028) [2021-04-16 15:54:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][1190/1251] eta 0:00:17 lr 0.000090 time 0.2801 (0.2892) loss 2.3801 (3.0926) grad_norm 2.5050 (2.7037) [2021-04-16 15:54:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][1200/1251] eta 0:00:14 lr 0.000090 time 0.2557 (0.2890) loss 3.2571 (3.0944) grad_norm 2.6288 (2.7049) [2021-04-16 15:54:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][1210/1251] eta 0:00:11 lr 0.000090 time 0.2426 (0.2889) loss 3.7870 (3.0948) grad_norm 2.3195 (2.7063) [2021-04-16 15:54:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][1220/1251] eta 0:00:08 lr 0.000090 time 0.2742 (0.2887) loss 1.9054 (3.0925) grad_norm 2.5835 (2.7079) [2021-04-16 15:54:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][1230/1251] eta 0:00:06 lr 0.000090 time 0.2642 (0.2887) loss 3.5207 (3.0930) grad_norm 2.8051 (2.7062) [2021-04-16 15:54:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][1240/1251] eta 0:00:03 lr 0.000090 time 0.2478 (0.2885) loss 3.8466 (3.0949) grad_norm 2.3855 (2.7049) [2021-04-16 15:54:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [244/300][1250/1251] eta 0:00:00 lr 0.000090 time 0.2485 (0.2882) loss 2.6057 (3.0942) grad_norm 2.5086 (2.7055) [2021-04-16 15:54:56 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 244 training takes 0:06:11 [2021-04-16 15:54:56 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_244.pth saving...... [2021-04-16 15:55:15 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_244.pth saved !!! [2021-04-16 15:55:17 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.278 (1.278) Loss 0.9579 (0.9579) Acc@1 77.246 (77.246) Acc@5 93.945 (93.945) [2021-04-16 15:55:18 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.105 (0.215) Loss 0.7849 (0.8482) Acc@1 80.762 (79.963) Acc@5 96.387 (95.091) [2021-04-16 15:55:20 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.134 (0.220) Loss 0.8458 (0.8474) Acc@1 79.297 (80.004) Acc@5 94.531 (95.103) [2021-04-16 15:55:23 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.166 (0.248) Loss 0.8099 (0.8420) Acc@1 81.445 (80.182) Acc@5 95.215 (95.171) [2021-04-16 15:55:24 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.216) Loss 0.8419 (0.8451) Acc@1 80.566 (80.111) Acc@5 94.336 (95.048) [2021-04-16 15:55:45 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.104 Acc@5 95.100 [2021-04-16 15:55:45 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 80.1% [2021-04-16 15:55:45 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 80.15% [2021-04-16 15:55:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][0/1251] eta 2:00:20 lr 0.000090 time 5.7721 (5.7721) loss 3.1327 (3.1327) grad_norm 2.4971 (2.4971) [2021-04-16 15:55:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][10/1251] eta 0:15:59 lr 0.000090 time 0.2713 (0.7735) loss 3.2279 (3.1636) grad_norm 2.6887 (2.7066) [2021-04-16 15:55:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][20/1251] eta 0:11:01 lr 0.000090 time 0.2907 (0.5373) loss 3.7187 (3.0198) grad_norm 2.3725 (2.7767) [2021-04-16 15:55:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][30/1251] eta 0:09:16 lr 0.000090 time 0.2834 (0.4556) loss 3.1611 (3.0417) grad_norm 2.5818 (2.7924) [2021-04-16 15:56:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][40/1251] eta 0:08:18 lr 0.000090 time 0.2968 (0.4117) loss 3.1591 (3.0554) grad_norm 2.2150 (2.7703) [2021-04-16 15:56:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][50/1251] eta 0:07:41 lr 0.000090 time 0.2780 (0.3842) loss 3.3832 (3.0402) grad_norm 2.7781 (2.7210) [2021-04-16 15:56:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][60/1251] eta 0:07:16 lr 0.000090 time 0.2689 (0.3664) loss 2.3721 (3.0384) grad_norm 2.5534 (2.6848) [2021-04-16 15:56:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][70/1251] eta 0:07:00 lr 0.000090 time 0.2818 (0.3559) loss 3.0742 (2.9979) grad_norm 2.9733 (2.6647) [2021-04-16 15:56:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][80/1251] eta 0:06:45 lr 0.000090 time 0.2538 (0.3462) loss 3.2978 (3.0069) grad_norm 2.3697 (2.6613) [2021-04-16 15:56:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][90/1251] eta 0:06:35 lr 0.000090 time 0.2834 (0.3404) loss 3.0278 (3.0221) grad_norm 2.6417 (2.6463) [2021-04-16 15:56:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][100/1251] eta 0:06:24 lr 0.000090 time 0.2850 (0.3343) loss 2.5202 (3.0339) grad_norm 2.5764 (2.6486) [2021-04-16 15:56:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][110/1251] eta 0:06:16 lr 0.000090 time 0.3026 (0.3298) loss 3.3431 (3.0463) grad_norm 2.4666 (2.6460) [2021-04-16 15:56:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][120/1251] eta 0:06:09 lr 0.000090 time 0.2724 (0.3268) loss 2.4345 (3.0431) grad_norm 2.6754 (2.6575) [2021-04-16 15:56:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][130/1251] eta 0:06:02 lr 0.000090 time 0.2598 (0.3231) loss 2.1784 (3.0322) grad_norm 2.6924 (2.6634) [2021-04-16 15:56:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][140/1251] eta 0:05:56 lr 0.000090 time 0.2959 (0.3206) loss 3.0077 (3.0374) grad_norm 2.4617 (2.6932) [2021-04-16 15:56:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][150/1251] eta 0:05:49 lr 0.000090 time 0.2683 (0.3176) loss 3.1944 (3.0395) grad_norm 2.3462 (2.7139) [2021-04-16 15:56:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][160/1251] eta 0:05:43 lr 0.000089 time 0.2837 (0.3151) loss 2.9685 (3.0485) grad_norm 2.7738 (2.7285) [2021-04-16 15:56:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][170/1251] eta 0:05:37 lr 0.000089 time 0.2766 (0.3126) loss 3.2589 (3.0544) grad_norm 2.5107 (2.7215) [2021-04-16 15:56:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][180/1251] eta 0:05:33 lr 0.000089 time 0.2916 (0.3115) loss 2.1893 (3.0371) grad_norm 2.3362 (2.7137) [2021-04-16 15:56:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][190/1251] eta 0:05:28 lr 0.000089 time 0.2906 (0.3099) loss 2.7172 (3.0369) grad_norm 3.0601 (2.7062) [2021-04-16 15:56:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][200/1251] eta 0:05:24 lr 0.000089 time 0.2719 (0.3083) loss 3.5993 (3.0444) grad_norm 3.0772 (2.6975) [2021-04-16 15:56:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][210/1251] eta 0:05:19 lr 0.000089 time 0.2724 (0.3067) loss 1.9709 (3.0514) grad_norm 2.7595 (2.6974) [2021-04-16 15:56:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][220/1251] eta 0:05:15 lr 0.000089 time 0.2764 (0.3059) loss 2.1215 (3.0524) grad_norm 2.2642 (2.6881) [2021-04-16 15:56:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][230/1251] eta 0:05:11 lr 0.000089 time 0.3066 (0.3047) loss 2.7590 (3.0389) grad_norm 2.7779 (2.6882) [2021-04-16 15:56:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][240/1251] eta 0:05:07 lr 0.000089 time 0.2677 (0.3037) loss 3.4867 (3.0388) grad_norm 2.3874 (2.6870) [2021-04-16 15:57:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][250/1251] eta 0:05:03 lr 0.000089 time 0.2943 (0.3037) loss 3.0953 (3.0289) grad_norm 2.3856 (2.6861) [2021-04-16 15:57:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][260/1251] eta 0:04:59 lr 0.000089 time 0.2674 (0.3025) loss 2.3470 (3.0236) grad_norm 2.5055 (2.6886) [2021-04-16 15:57:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][270/1251] eta 0:04:55 lr 0.000089 time 0.2964 (0.3016) loss 4.1864 (3.0299) grad_norm 2.9179 (2.6849) [2021-04-16 15:57:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][280/1251] eta 0:04:52 lr 0.000089 time 0.2738 (0.3011) loss 2.3808 (3.0338) grad_norm 3.6338 (2.6847) [2021-04-16 15:57:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][290/1251] eta 0:04:48 lr 0.000089 time 0.3004 (0.3003) loss 3.2683 (3.0283) grad_norm 2.5362 (2.6882) [2021-04-16 15:57:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][300/1251] eta 0:04:45 lr 0.000089 time 0.2715 (0.3002) loss 3.4197 (3.0293) grad_norm 2.5605 (2.6907) [2021-04-16 15:57:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][310/1251] eta 0:04:41 lr 0.000089 time 0.2919 (0.2996) loss 3.7243 (3.0333) grad_norm 2.5204 (2.6889) [2021-04-16 15:57:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][320/1251] eta 0:04:38 lr 0.000089 time 0.2748 (0.2988) loss 2.9434 (3.0361) grad_norm 2.7058 (2.6944) [2021-04-16 15:57:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][330/1251] eta 0:04:34 lr 0.000089 time 0.2992 (0.2983) loss 3.2630 (3.0316) grad_norm 2.6826 (2.6987) [2021-04-16 15:57:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][340/1251] eta 0:04:31 lr 0.000089 time 0.2565 (0.2975) loss 3.5513 (3.0366) grad_norm 2.4701 (2.6969) [2021-04-16 15:57:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][350/1251] eta 0:04:27 lr 0.000089 time 0.2926 (0.2972) loss 1.6746 (3.0323) grad_norm 2.3610 (2.6930) [2021-04-16 15:57:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][360/1251] eta 0:04:25 lr 0.000089 time 0.2655 (0.2977) loss 2.3054 (3.0307) grad_norm 2.4592 (2.6914) [2021-04-16 15:57:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][370/1251] eta 0:04:21 lr 0.000089 time 0.2619 (0.2974) loss 2.4415 (3.0321) grad_norm 2.6402 (2.6924) [2021-04-16 15:57:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][380/1251] eta 0:04:18 lr 0.000089 time 0.3083 (0.2969) loss 3.9455 (3.0354) grad_norm 2.5890 (2.6920) [2021-04-16 15:57:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][390/1251] eta 0:04:15 lr 0.000089 time 0.2654 (0.2962) loss 2.2664 (3.0349) grad_norm 2.5370 (2.6925) [2021-04-16 15:57:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][400/1251] eta 0:04:11 lr 0.000089 time 0.2724 (0.2957) loss 3.5193 (3.0400) grad_norm 2.7540 (2.6929) [2021-04-16 15:57:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][410/1251] eta 0:04:08 lr 0.000089 time 0.2919 (0.2954) loss 2.4789 (3.0365) grad_norm 2.8064 (2.6939) [2021-04-16 15:57:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][420/1251] eta 0:04:05 lr 0.000089 time 0.2592 (0.2951) loss 2.1075 (3.0384) grad_norm 2.5870 (2.6889) [2021-04-16 15:57:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][430/1251] eta 0:04:02 lr 0.000089 time 0.2883 (0.2951) loss 3.5014 (3.0382) grad_norm 2.3485 (2.6969) [2021-04-16 15:57:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][440/1251] eta 0:03:58 lr 0.000089 time 0.2546 (0.2947) loss 3.2886 (3.0368) grad_norm 2.5744 (2.6957) [2021-04-16 15:57:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][450/1251] eta 0:03:55 lr 0.000089 time 0.2929 (0.2943) loss 2.7883 (3.0395) grad_norm 2.5284 (2.7007) [2021-04-16 15:58:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][460/1251] eta 0:03:52 lr 0.000089 time 0.2982 (0.2940) loss 2.3906 (3.0386) grad_norm 2.6298 (2.6982) [2021-04-16 15:58:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][470/1251] eta 0:03:49 lr 0.000089 time 0.2962 (0.2937) loss 3.5198 (3.0379) grad_norm 2.5100 (2.6971) [2021-04-16 15:58:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][480/1251] eta 0:03:46 lr 0.000089 time 0.2744 (0.2934) loss 3.0263 (3.0389) grad_norm 2.5385 (2.6915) [2021-04-16 15:58:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][490/1251] eta 0:03:42 lr 0.000089 time 0.2626 (0.2930) loss 3.7802 (3.0396) grad_norm 2.9090 (2.6896) [2021-04-16 15:58:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][500/1251] eta 0:03:39 lr 0.000089 time 0.2712 (0.2926) loss 3.6507 (3.0471) grad_norm 2.4097 (2.6862) [2021-04-16 15:58:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][510/1251] eta 0:03:36 lr 0.000089 time 0.2559 (0.2922) loss 2.4039 (3.0502) grad_norm 2.6595 (2.6897) [2021-04-16 15:58:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][520/1251] eta 0:03:33 lr 0.000089 time 0.3059 (0.2920) loss 3.1183 (3.0508) grad_norm 2.5906 (2.6865) [2021-04-16 15:58:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][530/1251] eta 0:03:30 lr 0.000089 time 0.2671 (0.2917) loss 3.3598 (3.0547) grad_norm 2.5665 (2.6854) [2021-04-16 15:58:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][540/1251] eta 0:03:27 lr 0.000089 time 0.2663 (0.2913) loss 3.8256 (3.0544) grad_norm 2.4704 (2.6823) [2021-04-16 15:58:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][550/1251] eta 0:03:24 lr 0.000089 time 0.2897 (0.2911) loss 3.4364 (3.0568) grad_norm 2.5083 (2.6836) [2021-04-16 15:58:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][560/1251] eta 0:03:20 lr 0.000089 time 0.2873 (0.2909) loss 3.5525 (3.0607) grad_norm 2.7149 (2.6804) [2021-04-16 15:58:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][570/1251] eta 0:03:18 lr 0.000089 time 0.2990 (0.2909) loss 3.4235 (3.0651) grad_norm 3.2407 (2.6781) [2021-04-16 15:58:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][580/1251] eta 0:03:15 lr 0.000089 time 0.2759 (0.2907) loss 3.6866 (3.0684) grad_norm 2.3163 (2.6800) [2021-04-16 15:58:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][590/1251] eta 0:03:12 lr 0.000089 time 0.2664 (0.2905) loss 3.1009 (3.0661) grad_norm 2.9486 (2.6800) [2021-04-16 15:58:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][600/1251] eta 0:03:09 lr 0.000089 time 0.2808 (0.2904) loss 3.4879 (3.0666) grad_norm 2.6767 (2.6815) [2021-04-16 15:58:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][610/1251] eta 0:03:05 lr 0.000088 time 0.2408 (0.2901) loss 3.2592 (3.0654) grad_norm 2.5154 (2.6831) [2021-04-16 15:58:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][620/1251] eta 0:03:02 lr 0.000088 time 0.3032 (0.2899) loss 3.2175 (3.0643) grad_norm 3.0311 (2.6863) [2021-04-16 15:58:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][630/1251] eta 0:02:59 lr 0.000088 time 0.2954 (0.2898) loss 2.8663 (3.0640) grad_norm 3.4029 (2.6915) [2021-04-16 15:58:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][640/1251] eta 0:02:57 lr 0.000088 time 0.2776 (0.2897) loss 3.1182 (3.0637) grad_norm 2.3547 (2.6930) [2021-04-16 15:58:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][650/1251] eta 0:02:53 lr 0.000088 time 0.2602 (0.2894) loss 2.5774 (3.0650) grad_norm 2.6126 (2.6968) [2021-04-16 15:58:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][660/1251] eta 0:02:51 lr 0.000088 time 0.2550 (0.2894) loss 3.1570 (3.0627) grad_norm 2.9593 (2.6985) [2021-04-16 15:58:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][670/1251] eta 0:02:48 lr 0.000088 time 0.2858 (0.2893) loss 2.7894 (3.0641) grad_norm 2.6132 (2.6972) [2021-04-16 15:59:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][680/1251] eta 0:02:45 lr 0.000088 time 0.2553 (0.2891) loss 3.2741 (3.0601) grad_norm 2.3305 (2.6984) [2021-04-16 15:59:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][690/1251] eta 0:02:42 lr 0.000088 time 0.2930 (0.2891) loss 2.2141 (3.0557) grad_norm 2.7173 (2.7003) [2021-04-16 15:59:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][700/1251] eta 0:02:39 lr 0.000088 time 0.2527 (0.2888) loss 3.4476 (3.0566) grad_norm 2.4813 (2.7012) [2021-04-16 15:59:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][710/1251] eta 0:02:36 lr 0.000088 time 0.2768 (0.2886) loss 3.5337 (3.0609) grad_norm 3.2932 (2.7019) [2021-04-16 15:59:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][720/1251] eta 0:02:33 lr 0.000088 time 0.2835 (0.2886) loss 3.4061 (3.0627) grad_norm 3.0191 (2.7008) [2021-04-16 15:59:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][730/1251] eta 0:02:30 lr 0.000088 time 0.2905 (0.2884) loss 3.2508 (3.0637) grad_norm 2.5707 (2.7031) [2021-04-16 15:59:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][740/1251] eta 0:02:27 lr 0.000088 time 0.2690 (0.2882) loss 3.5871 (3.0653) grad_norm 3.1981 (2.7035) [2021-04-16 15:59:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][750/1251] eta 0:02:24 lr 0.000088 time 0.2818 (0.2880) loss 3.1978 (3.0681) grad_norm 2.3684 (2.7024) [2021-04-16 15:59:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][760/1251] eta 0:02:21 lr 0.000088 time 0.2759 (0.2879) loss 2.8791 (3.0657) grad_norm 2.5225 (2.7011) [2021-04-16 15:59:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][770/1251] eta 0:02:18 lr 0.000088 time 0.2724 (0.2878) loss 3.7247 (3.0681) grad_norm 2.4286 (2.7010) [2021-04-16 15:59:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][780/1251] eta 0:02:15 lr 0.000088 time 0.2859 (0.2876) loss 2.5036 (3.0657) grad_norm 2.7608 (2.7011) [2021-04-16 15:59:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][790/1251] eta 0:02:12 lr 0.000088 time 0.2914 (0.2876) loss 2.8126 (3.0657) grad_norm 3.1100 (2.7025) [2021-04-16 15:59:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][800/1251] eta 0:02:09 lr 0.000088 time 0.2810 (0.2875) loss 2.8387 (3.0657) grad_norm 2.7440 (2.7025) [2021-04-16 15:59:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][810/1251] eta 0:02:06 lr 0.000088 time 0.2803 (0.2875) loss 3.2451 (3.0630) grad_norm 3.2370 (2.7030) [2021-04-16 15:59:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][820/1251] eta 0:02:03 lr 0.000088 time 0.2695 (0.2874) loss 3.5489 (3.0641) grad_norm 2.6817 (2.7040) [2021-04-16 15:59:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][830/1251] eta 0:02:01 lr 0.000088 time 0.2824 (0.2874) loss 3.4551 (3.0612) grad_norm 2.7088 (2.7053) [2021-04-16 15:59:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][840/1251] eta 0:01:58 lr 0.000088 time 0.2979 (0.2872) loss 3.2073 (3.0605) grad_norm 2.4466 (2.7045) [2021-04-16 15:59:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][850/1251] eta 0:01:55 lr 0.000088 time 0.2977 (0.2871) loss 2.5107 (3.0638) grad_norm 2.5252 (2.7024) [2021-04-16 15:59:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][860/1251] eta 0:01:52 lr 0.000088 time 0.2794 (0.2870) loss 2.2973 (3.0648) grad_norm 2.8664 (2.7011) [2021-04-16 15:59:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][870/1251] eta 0:01:49 lr 0.000088 time 0.2922 (0.2869) loss 2.6286 (3.0617) grad_norm 2.4160 (2.7009) [2021-04-16 15:59:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][880/1251] eta 0:01:46 lr 0.000088 time 0.2738 (0.2869) loss 3.0078 (3.0621) grad_norm 3.4833 (2.7022) [2021-04-16 16:00:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][890/1251] eta 0:01:43 lr 0.000088 time 0.2818 (0.2867) loss 3.6410 (3.0644) grad_norm 2.6297 (2.7010) [2021-04-16 16:00:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][900/1251] eta 0:01:40 lr 0.000088 time 0.3168 (0.2866) loss 3.3961 (3.0613) grad_norm 2.2226 (2.7014) [2021-04-16 16:00:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][910/1251] eta 0:01:37 lr 0.000088 time 0.2758 (0.2866) loss 3.4432 (3.0610) grad_norm 2.6691 (2.7010) [2021-04-16 16:00:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][920/1251] eta 0:01:34 lr 0.000088 time 0.2744 (0.2865) loss 2.6592 (3.0621) grad_norm 2.4074 (2.7001) [2021-04-16 16:00:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][930/1251] eta 0:01:31 lr 0.000088 time 0.2463 (0.2866) loss 3.3335 (3.0620) grad_norm 2.5871 (2.7009) [2021-04-16 16:00:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][940/1251] eta 0:01:29 lr 0.000088 time 0.3855 (0.2868) loss 3.3653 (3.0655) grad_norm 2.4981 (2.7001) [2021-04-16 16:00:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][950/1251] eta 0:01:26 lr 0.000088 time 0.2707 (0.2866) loss 3.4402 (3.0658) grad_norm 2.9580 (2.7022) [2021-04-16 16:00:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][960/1251] eta 0:01:23 lr 0.000088 time 0.2616 (0.2865) loss 3.3023 (3.0665) grad_norm 2.3132 (2.6996) [2021-04-16 16:00:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][970/1251] eta 0:01:20 lr 0.000088 time 0.2492 (0.2864) loss 3.6286 (3.0662) grad_norm 2.6779 (2.7004) [2021-04-16 16:00:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][980/1251] eta 0:01:17 lr 0.000088 time 0.2795 (0.2863) loss 3.2171 (3.0680) grad_norm 2.7219 (2.7002) [2021-04-16 16:00:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][990/1251] eta 0:01:14 lr 0.000088 time 0.2787 (0.2861) loss 3.8373 (3.0692) grad_norm 2.7063 (2.6981) [2021-04-16 16:00:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][1000/1251] eta 0:01:11 lr 0.000088 time 0.2911 (0.2860) loss 3.3604 (3.0720) grad_norm 2.7829 (2.6973) [2021-04-16 16:00:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][1010/1251] eta 0:01:08 lr 0.000088 time 0.2845 (0.2861) loss 3.6987 (3.0723) grad_norm 2.6108 (2.6966) [2021-04-16 16:00:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][1020/1251] eta 0:01:06 lr 0.000088 time 0.3109 (0.2860) loss 3.2300 (3.0736) grad_norm 2.4282 (2.6949) [2021-04-16 16:00:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][1030/1251] eta 0:01:03 lr 0.000088 time 0.2739 (0.2859) loss 2.4509 (3.0753) grad_norm 2.9382 (2.6953) [2021-04-16 16:00:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][1040/1251] eta 0:01:00 lr 0.000088 time 0.2773 (0.2858) loss 2.9700 (3.0724) grad_norm 2.6189 (2.6944) [2021-04-16 16:00:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][1050/1251] eta 0:00:57 lr 0.000088 time 0.2892 (0.2857) loss 2.3800 (3.0721) grad_norm 2.5668 (2.6934) [2021-04-16 16:00:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][1060/1251] eta 0:00:54 lr 0.000087 time 0.2584 (0.2856) loss 3.0503 (3.0743) grad_norm 2.7537 (2.6925) [2021-04-16 16:00:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][1070/1251] eta 0:00:51 lr 0.000087 time 0.2712 (0.2856) loss 3.7994 (3.0744) grad_norm 2.4790 (2.6913) [2021-04-16 16:00:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][1080/1251] eta 0:00:48 lr 0.000087 time 0.2811 (0.2856) loss 3.3047 (3.0735) grad_norm 2.2937 (2.6890) [2021-04-16 16:00:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][1090/1251] eta 0:00:45 lr 0.000087 time 0.2922 (0.2855) loss 2.8394 (3.0734) grad_norm 2.4270 (2.6882) [2021-04-16 16:00:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][1100/1251] eta 0:00:43 lr 0.000087 time 0.2713 (0.2854) loss 3.0891 (3.0756) grad_norm 2.4914 (2.6876) [2021-04-16 16:01:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][1110/1251] eta 0:00:40 lr 0.000087 time 0.2490 (0.2853) loss 2.9111 (3.0742) grad_norm 2.6264 (2.6871) [2021-04-16 16:01:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][1120/1251] eta 0:00:37 lr 0.000087 time 0.2822 (0.2852) loss 2.7015 (3.0727) grad_norm 2.6053 (2.6882) [2021-04-16 16:01:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][1130/1251] eta 0:00:34 lr 0.000087 time 0.2764 (0.2852) loss 2.9562 (3.0701) grad_norm 2.6018 (2.6911) [2021-04-16 16:01:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][1140/1251] eta 0:00:31 lr 0.000087 time 0.2717 (0.2853) loss 3.2685 (3.0715) grad_norm 2.9548 (2.6907) [2021-04-16 16:01:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][1150/1251] eta 0:00:28 lr 0.000087 time 0.2593 (0.2853) loss 3.3331 (3.0723) grad_norm 2.4999 (inf) [2021-04-16 16:01:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][1160/1251] eta 0:00:25 lr 0.000087 time 0.2574 (0.2852) loss 2.1722 (3.0723) grad_norm 3.3660 (inf) [2021-04-16 16:01:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][1170/1251] eta 0:00:23 lr 0.000087 time 0.2556 (0.2851) loss 3.4189 (3.0732) grad_norm 2.3872 (inf) [2021-04-16 16:01:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][1180/1251] eta 0:00:20 lr 0.000087 time 0.2713 (0.2850) loss 3.4204 (3.0737) grad_norm 2.3462 (inf) [2021-04-16 16:01:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][1190/1251] eta 0:00:17 lr 0.000087 time 0.2813 (0.2849) loss 3.4730 (3.0758) grad_norm 2.4896 (inf) [2021-04-16 16:01:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][1200/1251] eta 0:00:14 lr 0.000087 time 0.2839 (0.2849) loss 3.2769 (3.0762) grad_norm 2.5334 (inf) [2021-04-16 16:01:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][1210/1251] eta 0:00:11 lr 0.000087 time 0.2743 (0.2848) loss 2.5234 (3.0779) grad_norm 2.7002 (inf) [2021-04-16 16:01:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][1220/1251] eta 0:00:08 lr 0.000087 time 0.2731 (0.2849) loss 3.2140 (3.0791) grad_norm 2.4136 (inf) [2021-04-16 16:01:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][1230/1251] eta 0:00:05 lr 0.000087 time 0.2809 (0.2848) loss 3.2896 (3.0800) grad_norm 2.5334 (inf) [2021-04-16 16:01:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][1240/1251] eta 0:00:03 lr 0.000087 time 0.2487 (0.2847) loss 3.5981 (3.0799) grad_norm 2.9508 (inf) [2021-04-16 16:01:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [245/300][1250/1251] eta 0:00:00 lr 0.000087 time 0.2479 (0.2844) loss 3.2285 (3.0817) grad_norm 2.6345 (inf) [2021-04-16 16:01:56 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 245 training takes 0:06:10 [2021-04-16 16:01:56 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_245.pth saving...... [2021-04-16 16:02:12 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_245.pth saved !!! [2021-04-16 16:02:13 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.125 (1.125) Loss 0.8778 (0.8778) Acc@1 78.809 (78.809) Acc@5 95.410 (95.410) [2021-04-16 16:02:14 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.120 (0.212) Loss 0.8062 (0.8570) Acc@1 80.664 (80.202) Acc@5 95.410 (94.984) [2021-04-16 16:02:17 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.204 (0.237) Loss 0.8456 (0.8590) Acc@1 80.664 (80.199) Acc@5 95.703 (95.038) [2021-04-16 16:02:19 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.114 (0.224) Loss 0.8790 (0.8633) Acc@1 79.492 (80.088) Acc@5 95.020 (94.972) [2021-04-16 16:02:21 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.220) Loss 0.8498 (0.8565) Acc@1 80.273 (80.164) Acc@5 95.508 (95.093) [2021-04-16 16:02:37 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.202 Acc@5 95.148 [2021-04-16 16:02:37 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 80.2% [2021-04-16 16:02:37 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 80.20% [2021-04-16 16:02:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][0/1251] eta 2:06:07 lr 0.000087 time 6.0495 (6.0495) loss 2.7896 (2.7896) grad_norm 2.5722 (2.5722) [2021-04-16 16:02:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][10/1251] eta 0:16:32 lr 0.000087 time 0.2987 (0.8001) loss 3.7721 (3.1638) grad_norm 3.0327 (2.6960) [2021-04-16 16:02:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][20/1251] eta 0:11:24 lr 0.000087 time 0.2488 (0.5560) loss 3.2014 (3.1197) grad_norm 2.4536 (2.8054) [2021-04-16 16:02:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][30/1251] eta 0:09:27 lr 0.000087 time 0.2434 (0.4652) loss 3.2558 (3.0865) grad_norm 3.0435 (2.7916) [2021-04-16 16:02:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][40/1251] eta 0:08:30 lr 0.000087 time 0.2966 (0.4215) loss 3.3665 (3.0961) grad_norm 2.7150 (2.8970) [2021-04-16 16:02:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][50/1251] eta 0:07:51 lr 0.000087 time 0.2698 (0.3925) loss 3.4756 (3.1007) grad_norm 2.4579 (2.8827) [2021-04-16 16:03:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][60/1251] eta 0:07:25 lr 0.000087 time 0.2762 (0.3741) loss 3.6468 (3.1711) grad_norm 2.7220 (2.8305) [2021-04-16 16:03:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][70/1251] eta 0:07:05 lr 0.000087 time 0.2513 (0.3602) loss 3.2249 (3.1665) grad_norm 3.0143 (2.8305) [2021-04-16 16:03:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][80/1251] eta 0:06:49 lr 0.000087 time 0.2469 (0.3496) loss 2.1040 (3.1618) grad_norm 2.5830 (2.8134) [2021-04-16 16:03:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][90/1251] eta 0:06:38 lr 0.000087 time 0.2839 (0.3436) loss 2.6983 (3.1360) grad_norm 3.5815 (2.8108) [2021-04-16 16:03:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][100/1251] eta 0:06:27 lr 0.000087 time 0.2792 (0.3367) loss 3.3031 (3.1490) grad_norm 2.5829 (2.7869) [2021-04-16 16:03:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][110/1251] eta 0:06:18 lr 0.000087 time 0.2764 (0.3318) loss 2.1704 (3.1356) grad_norm 2.4153 (2.7658) [2021-04-16 16:03:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][120/1251] eta 0:06:09 lr 0.000087 time 0.2707 (0.3271) loss 2.4491 (3.1129) grad_norm 3.4338 (2.7594) [2021-04-16 16:03:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][130/1251] eta 0:06:03 lr 0.000087 time 0.2761 (0.3245) loss 3.3130 (3.1163) grad_norm 2.9699 (2.7454) [2021-04-16 16:03:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][140/1251] eta 0:05:57 lr 0.000087 time 0.2910 (0.3220) loss 3.4738 (3.1248) grad_norm 2.3481 (2.7459) [2021-04-16 16:03:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][150/1251] eta 0:05:52 lr 0.000087 time 0.3782 (0.3204) loss 3.3692 (3.1290) grad_norm 3.0158 (2.7461) [2021-04-16 16:03:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][160/1251] eta 0:05:47 lr 0.000087 time 0.2622 (0.3181) loss 3.2082 (3.1281) grad_norm 2.8498 (2.7463) [2021-04-16 16:03:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][170/1251] eta 0:05:41 lr 0.000087 time 0.2664 (0.3159) loss 3.7108 (3.1292) grad_norm 2.5869 (2.7405) [2021-04-16 16:03:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][180/1251] eta 0:05:36 lr 0.000087 time 0.3152 (0.3143) loss 2.1185 (3.1313) grad_norm 2.9509 (2.7397) [2021-04-16 16:03:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][190/1251] eta 0:05:34 lr 0.000087 time 0.2908 (0.3153) loss 2.9852 (3.1399) grad_norm 3.8644 (2.7569) [2021-04-16 16:03:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][200/1251] eta 0:05:29 lr 0.000087 time 0.2931 (0.3135) loss 2.6168 (3.1467) grad_norm 2.5826 (2.7498) [2021-04-16 16:03:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][210/1251] eta 0:05:24 lr 0.000087 time 0.2757 (0.3121) loss 3.6977 (3.1396) grad_norm 3.0287 (2.7457) [2021-04-16 16:03:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][220/1251] eta 0:05:20 lr 0.000087 time 0.2781 (0.3111) loss 3.4037 (3.1415) grad_norm 2.5667 (2.7450) [2021-04-16 16:03:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][230/1251] eta 0:05:16 lr 0.000087 time 0.2559 (0.3096) loss 2.8793 (3.1415) grad_norm 2.4548 (2.7432) [2021-04-16 16:03:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][240/1251] eta 0:05:11 lr 0.000087 time 0.2915 (0.3082) loss 2.8815 (3.1354) grad_norm 2.6677 (2.7356) [2021-04-16 16:03:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][250/1251] eta 0:05:07 lr 0.000087 time 0.2632 (0.3069) loss 2.8329 (3.1304) grad_norm 2.3681 (2.7329) [2021-04-16 16:03:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][260/1251] eta 0:05:03 lr 0.000086 time 0.2842 (0.3059) loss 2.7387 (3.1281) grad_norm 2.7202 (2.7291) [2021-04-16 16:04:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][270/1251] eta 0:04:59 lr 0.000086 time 0.2806 (0.3052) loss 2.7428 (3.1232) grad_norm 2.7145 (2.7285) [2021-04-16 16:04:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][280/1251] eta 0:04:55 lr 0.000086 time 0.2738 (0.3043) loss 4.0691 (3.1255) grad_norm 3.0924 (2.7274) [2021-04-16 16:04:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][290/1251] eta 0:04:51 lr 0.000086 time 0.2847 (0.3036) loss 2.3069 (3.1145) grad_norm 2.6400 (2.7277) [2021-04-16 16:04:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][300/1251] eta 0:04:47 lr 0.000086 time 0.2776 (0.3028) loss 3.2225 (3.1128) grad_norm 2.8735 (2.7248) [2021-04-16 16:04:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][310/1251] eta 0:04:44 lr 0.000086 time 0.2572 (0.3019) loss 3.1789 (3.1088) grad_norm 2.5232 (2.7232) [2021-04-16 16:04:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][320/1251] eta 0:04:40 lr 0.000086 time 0.2937 (0.3010) loss 3.3956 (3.1120) grad_norm 2.9449 (2.7223) [2021-04-16 16:04:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][330/1251] eta 0:04:36 lr 0.000086 time 0.2876 (0.3003) loss 3.2118 (3.1058) grad_norm 2.5394 (2.7192) [2021-04-16 16:04:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][340/1251] eta 0:04:32 lr 0.000086 time 0.2698 (0.2995) loss 3.8009 (3.1002) grad_norm 2.7714 (2.7195) [2021-04-16 16:04:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][350/1251] eta 0:04:29 lr 0.000086 time 0.2600 (0.2993) loss 3.3224 (3.0963) grad_norm 2.9604 (2.7252) [2021-04-16 16:04:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][360/1251] eta 0:04:25 lr 0.000086 time 0.2500 (0.2985) loss 3.6935 (3.0937) grad_norm 2.7720 (2.7267) [2021-04-16 16:04:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][370/1251] eta 0:04:22 lr 0.000086 time 0.2708 (0.2980) loss 2.8035 (3.0958) grad_norm 2.5840 (2.7309) [2021-04-16 16:04:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][380/1251] eta 0:04:19 lr 0.000086 time 0.2747 (0.2975) loss 2.9669 (3.0963) grad_norm 2.6465 (2.7381) [2021-04-16 16:04:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][390/1251] eta 0:04:15 lr 0.000086 time 0.2941 (0.2971) loss 3.4821 (3.1035) grad_norm 3.0221 (2.7401) [2021-04-16 16:04:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][400/1251] eta 0:04:12 lr 0.000086 time 0.2512 (0.2966) loss 3.1168 (3.1081) grad_norm 2.2320 (2.7366) [2021-04-16 16:04:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][410/1251] eta 0:04:09 lr 0.000086 time 0.2622 (0.2963) loss 3.3757 (3.1097) grad_norm 4.2698 (2.7380) [2021-04-16 16:04:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][420/1251] eta 0:04:06 lr 0.000086 time 0.2988 (0.2962) loss 3.0812 (3.1115) grad_norm 2.5485 (2.7349) [2021-04-16 16:04:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][430/1251] eta 0:04:02 lr 0.000086 time 0.2514 (0.2958) loss 3.1488 (3.1024) grad_norm 2.6937 (2.7357) [2021-04-16 16:04:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][440/1251] eta 0:03:59 lr 0.000086 time 0.2770 (0.2959) loss 3.1244 (3.1046) grad_norm 2.4871 (2.7375) [2021-04-16 16:04:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][450/1251] eta 0:03:56 lr 0.000086 time 0.2796 (0.2954) loss 3.3267 (3.1077) grad_norm 2.4366 (2.7396) [2021-04-16 16:04:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][460/1251] eta 0:03:53 lr 0.000086 time 0.2471 (0.2953) loss 3.3301 (3.1095) grad_norm 2.8261 (2.7378) [2021-04-16 16:04:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][470/1251] eta 0:03:50 lr 0.000086 time 0.2807 (0.2950) loss 2.6243 (3.1088) grad_norm 2.7405 (2.7374) [2021-04-16 16:04:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][480/1251] eta 0:03:47 lr 0.000086 time 0.2723 (0.2946) loss 3.0348 (3.1069) grad_norm 2.5172 (2.7345) [2021-04-16 16:05:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][490/1251] eta 0:03:43 lr 0.000086 time 0.2775 (0.2941) loss 3.1196 (3.1054) grad_norm 2.8949 (2.7364) [2021-04-16 16:05:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][500/1251] eta 0:03:40 lr 0.000086 time 0.2844 (0.2937) loss 4.1017 (3.1080) grad_norm 3.3736 (2.7360) [2021-04-16 16:05:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][510/1251] eta 0:03:37 lr 0.000086 time 0.2811 (0.2934) loss 3.0158 (3.1058) grad_norm 2.3236 (2.7405) [2021-04-16 16:05:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][520/1251] eta 0:03:34 lr 0.000086 time 0.2800 (0.2933) loss 3.6243 (3.1072) grad_norm 2.3976 (2.7422) [2021-04-16 16:05:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][530/1251] eta 0:03:31 lr 0.000086 time 0.2822 (0.2930) loss 3.0524 (3.1100) grad_norm 2.4306 (2.7406) [2021-04-16 16:05:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][540/1251] eta 0:03:28 lr 0.000086 time 0.2707 (0.2929) loss 4.1931 (3.1142) grad_norm 2.5726 (2.7438) [2021-04-16 16:05:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][550/1251] eta 0:03:25 lr 0.000086 time 0.2804 (0.2926) loss 3.2531 (3.1103) grad_norm 2.4210 (2.7428) [2021-04-16 16:05:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][560/1251] eta 0:03:22 lr 0.000086 time 0.2701 (0.2923) loss 3.2003 (3.1043) grad_norm 2.9543 (2.7447) [2021-04-16 16:05:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][570/1251] eta 0:03:19 lr 0.000086 time 0.2843 (0.2923) loss 3.4893 (3.1002) grad_norm 4.1733 (2.7461) [2021-04-16 16:05:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][580/1251] eta 0:03:15 lr 0.000086 time 0.2751 (0.2920) loss 3.5346 (3.0985) grad_norm 2.5330 (2.7458) [2021-04-16 16:05:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][590/1251] eta 0:03:13 lr 0.000086 time 0.2503 (0.2920) loss 2.9098 (3.0985) grad_norm 2.6191 (2.7438) [2021-04-16 16:05:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][600/1251] eta 0:03:09 lr 0.000086 time 0.2817 (0.2917) loss 3.3936 (3.1013) grad_norm 3.4263 (2.7435) [2021-04-16 16:05:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][610/1251] eta 0:03:06 lr 0.000086 time 0.2554 (0.2915) loss 3.1182 (3.1050) grad_norm 3.1304 (2.7464) [2021-04-16 16:05:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][620/1251] eta 0:03:03 lr 0.000086 time 0.2884 (0.2915) loss 3.1247 (3.1066) grad_norm 2.9409 (2.7462) [2021-04-16 16:05:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][630/1251] eta 0:03:00 lr 0.000086 time 0.2909 (0.2913) loss 2.9764 (3.1019) grad_norm 2.5741 (2.7443) [2021-04-16 16:05:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][640/1251] eta 0:02:57 lr 0.000086 time 0.2662 (0.2910) loss 3.0587 (3.1025) grad_norm 2.5679 (2.7433) [2021-04-16 16:05:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][650/1251] eta 0:02:54 lr 0.000086 time 0.2574 (0.2908) loss 3.5025 (3.1068) grad_norm 2.7641 (2.7412) [2021-04-16 16:05:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][660/1251] eta 0:02:51 lr 0.000086 time 0.2733 (0.2906) loss 3.4576 (3.1070) grad_norm 3.0681 (2.7411) [2021-04-16 16:05:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][670/1251] eta 0:02:48 lr 0.000086 time 0.2695 (0.2906) loss 2.9907 (3.1045) grad_norm 2.5604 (2.7398) [2021-04-16 16:05:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][680/1251] eta 0:02:45 lr 0.000086 time 0.2850 (0.2905) loss 3.1033 (3.1014) grad_norm 2.6118 (2.7411) [2021-04-16 16:05:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][690/1251] eta 0:02:42 lr 0.000086 time 0.2575 (0.2904) loss 3.1279 (3.1015) grad_norm 2.1624 (2.7405) [2021-04-16 16:06:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][700/1251] eta 0:02:39 lr 0.000086 time 0.2744 (0.2902) loss 2.7482 (3.1013) grad_norm 2.4901 (2.7390) [2021-04-16 16:06:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][710/1251] eta 0:02:36 lr 0.000085 time 0.2748 (0.2901) loss 3.5809 (3.1028) grad_norm 2.8258 (2.7399) [2021-04-16 16:06:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][720/1251] eta 0:02:34 lr 0.000085 time 0.3035 (0.2903) loss 3.4231 (3.1045) grad_norm 2.5901 (2.7409) [2021-04-16 16:06:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][730/1251] eta 0:02:31 lr 0.000085 time 0.2655 (0.2904) loss 3.3841 (3.1078) grad_norm 4.2425 (2.7427) [2021-04-16 16:06:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][740/1251] eta 0:02:28 lr 0.000085 time 0.2595 (0.2902) loss 3.2606 (3.1051) grad_norm 2.5491 (2.7449) [2021-04-16 16:06:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][750/1251] eta 0:02:25 lr 0.000085 time 0.2821 (0.2900) loss 3.8405 (3.1039) grad_norm 2.6699 (2.7448) [2021-04-16 16:06:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][760/1251] eta 0:02:22 lr 0.000085 time 0.2682 (0.2897) loss 2.9334 (3.1035) grad_norm 2.9501 (2.7457) [2021-04-16 16:06:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][770/1251] eta 0:02:19 lr 0.000085 time 0.2647 (0.2895) loss 2.1313 (3.1042) grad_norm 2.9332 (2.7469) [2021-04-16 16:06:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][780/1251] eta 0:02:16 lr 0.000085 time 0.2827 (0.2893) loss 2.8848 (3.1034) grad_norm 2.6965 (2.7479) [2021-04-16 16:06:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][790/1251] eta 0:02:13 lr 0.000085 time 0.2764 (0.2894) loss 3.3298 (3.1058) grad_norm 3.2096 (2.7466) [2021-04-16 16:06:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][800/1251] eta 0:02:10 lr 0.000085 time 0.2674 (0.2892) loss 3.0146 (3.1075) grad_norm 2.7591 (2.7495) [2021-04-16 16:06:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][810/1251] eta 0:02:07 lr 0.000085 time 0.2736 (0.2890) loss 3.7657 (3.1078) grad_norm 2.4877 (2.7537) [2021-04-16 16:06:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][820/1251] eta 0:02:04 lr 0.000085 time 0.2887 (0.2889) loss 3.4960 (3.1041) grad_norm 2.4844 (2.7542) [2021-04-16 16:06:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][830/1251] eta 0:02:01 lr 0.000085 time 0.2829 (0.2887) loss 2.7185 (3.1046) grad_norm 2.9198 (2.7533) [2021-04-16 16:06:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][840/1251] eta 0:01:58 lr 0.000085 time 0.2644 (0.2885) loss 3.4626 (3.1049) grad_norm 2.4052 (2.7558) [2021-04-16 16:06:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][850/1251] eta 0:01:55 lr 0.000085 time 0.2773 (0.2884) loss 3.1703 (3.1046) grad_norm 3.2418 (2.7572) [2021-04-16 16:06:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][860/1251] eta 0:01:52 lr 0.000085 time 0.2878 (0.2883) loss 3.4240 (3.1035) grad_norm 2.4699 (2.7570) [2021-04-16 16:06:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][870/1251] eta 0:01:49 lr 0.000085 time 0.2553 (0.2881) loss 3.6078 (3.1029) grad_norm 2.7172 (2.7575) [2021-04-16 16:06:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][880/1251] eta 0:01:46 lr 0.000085 time 0.2674 (0.2880) loss 3.7936 (3.1042) grad_norm 2.7204 (2.7569) [2021-04-16 16:06:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][890/1251] eta 0:01:43 lr 0.000085 time 0.2779 (0.2879) loss 3.1888 (3.1011) grad_norm 2.5929 (2.7564) [2021-04-16 16:06:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][900/1251] eta 0:01:41 lr 0.000085 time 0.2515 (0.2878) loss 3.3200 (3.1006) grad_norm 2.8591 (2.7580) [2021-04-16 16:07:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][910/1251] eta 0:01:38 lr 0.000085 time 0.2798 (0.2877) loss 3.7546 (3.0999) grad_norm 2.9123 (2.7603) [2021-04-16 16:07:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][920/1251] eta 0:01:35 lr 0.000085 time 0.2585 (0.2876) loss 2.8764 (3.0999) grad_norm 2.4933 (2.7604) [2021-04-16 16:07:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][930/1251] eta 0:01:32 lr 0.000085 time 0.2597 (0.2876) loss 3.2341 (3.0969) grad_norm 2.3871 (2.7603) [2021-04-16 16:07:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][940/1251] eta 0:01:29 lr 0.000085 time 0.2503 (0.2875) loss 3.5649 (3.0987) grad_norm 2.8116 (2.7618) [2021-04-16 16:07:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][950/1251] eta 0:01:26 lr 0.000085 time 0.2888 (0.2874) loss 3.4647 (3.1001) grad_norm 2.6972 (2.7622) [2021-04-16 16:07:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][960/1251] eta 0:01:23 lr 0.000085 time 0.2782 (0.2872) loss 3.1444 (3.0987) grad_norm 2.7703 (2.7615) [2021-04-16 16:07:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][970/1251] eta 0:01:20 lr 0.000085 time 0.2788 (0.2871) loss 2.2938 (3.0957) grad_norm 3.0972 (2.7607) [2021-04-16 16:07:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][980/1251] eta 0:01:17 lr 0.000085 time 0.2500 (0.2870) loss 2.5779 (3.0962) grad_norm 2.9871 (2.7593) [2021-04-16 16:07:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][990/1251] eta 0:01:14 lr 0.000085 time 0.3705 (0.2870) loss 3.4710 (3.0972) grad_norm 3.0267 (2.7566) [2021-04-16 16:07:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][1000/1251] eta 0:01:12 lr 0.000085 time 0.2712 (0.2870) loss 3.0063 (3.0983) grad_norm 2.2681 (2.7561) [2021-04-16 16:07:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][1010/1251] eta 0:01:09 lr 0.000085 time 0.3390 (0.2868) loss 3.2750 (3.0970) grad_norm 2.9041 (2.7568) [2021-04-16 16:07:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][1020/1251] eta 0:01:06 lr 0.000085 time 0.2926 (0.2869) loss 2.9077 (3.0950) grad_norm 3.0043 (2.7573) [2021-04-16 16:07:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][1030/1251] eta 0:01:03 lr 0.000085 time 0.2760 (0.2868) loss 2.5312 (3.0969) grad_norm 2.7085 (2.7571) [2021-04-16 16:07:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][1040/1251] eta 0:01:00 lr 0.000085 time 0.2717 (0.2867) loss 2.3764 (3.0946) grad_norm 2.5106 (2.7559) [2021-04-16 16:07:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][1050/1251] eta 0:00:57 lr 0.000085 time 0.2548 (0.2866) loss 3.7015 (3.0934) grad_norm 3.2668 (2.7566) [2021-04-16 16:07:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][1060/1251] eta 0:00:54 lr 0.000085 time 0.2893 (0.2865) loss 2.9569 (3.0934) grad_norm 2.5220 (2.7569) [2021-04-16 16:07:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][1070/1251] eta 0:00:51 lr 0.000085 time 0.2715 (0.2865) loss 2.9451 (3.0957) grad_norm 2.5060 (2.7567) [2021-04-16 16:07:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][1080/1251] eta 0:00:48 lr 0.000085 time 0.2656 (0.2864) loss 2.1450 (3.0939) grad_norm 3.6671 (2.7578) [2021-04-16 16:07:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][1090/1251] eta 0:00:46 lr 0.000085 time 0.2682 (0.2865) loss 2.8415 (3.0934) grad_norm 2.6869 (2.7585) [2021-04-16 16:07:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][1100/1251] eta 0:00:43 lr 0.000085 time 0.2782 (0.2864) loss 2.5477 (3.0910) grad_norm 2.9934 (2.7598) [2021-04-16 16:07:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][1110/1251] eta 0:00:40 lr 0.000085 time 0.2697 (0.2863) loss 2.8424 (3.0885) grad_norm 2.3816 (2.7619) [2021-04-16 16:07:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][1120/1251] eta 0:00:37 lr 0.000085 time 0.2668 (0.2862) loss 3.6946 (3.0912) grad_norm 2.7952 (2.7620) [2021-04-16 16:08:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][1130/1251] eta 0:00:34 lr 0.000085 time 0.2596 (0.2861) loss 3.3189 (3.0929) grad_norm 2.7263 (2.7615) [2021-04-16 16:08:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][1140/1251] eta 0:00:31 lr 0.000085 time 0.2832 (0.2861) loss 3.6149 (3.0936) grad_norm 2.3368 (2.7599) [2021-04-16 16:08:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][1150/1251] eta 0:00:28 lr 0.000085 time 0.2654 (0.2860) loss 3.7396 (3.0933) grad_norm 2.5811 (2.7589) [2021-04-16 16:08:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][1160/1251] eta 0:00:26 lr 0.000085 time 0.2671 (0.2860) loss 2.6617 (3.0938) grad_norm 2.5945 (2.7583) [2021-04-16 16:08:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][1170/1251] eta 0:00:23 lr 0.000084 time 0.3029 (0.2861) loss 2.7782 (3.0943) grad_norm 2.5643 (2.7568) [2021-04-16 16:08:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][1180/1251] eta 0:00:20 lr 0.000084 time 0.2609 (0.2860) loss 3.2603 (3.0943) grad_norm 3.1952 (2.7566) [2021-04-16 16:08:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][1190/1251] eta 0:00:17 lr 0.000084 time 0.3823 (0.2860) loss 3.6083 (3.0941) grad_norm 2.5830 (2.7575) [2021-04-16 16:08:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][1200/1251] eta 0:00:14 lr 0.000084 time 0.2534 (0.2859) loss 3.2805 (3.0954) grad_norm 2.9266 (2.7596) [2021-04-16 16:08:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][1210/1251] eta 0:00:11 lr 0.000084 time 0.2772 (0.2858) loss 3.1634 (3.0951) grad_norm 3.1965 (2.7625) [2021-04-16 16:08:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][1220/1251] eta 0:00:08 lr 0.000084 time 0.2943 (0.2858) loss 3.4071 (3.0934) grad_norm 3.0248 (2.7646) [2021-04-16 16:08:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][1230/1251] eta 0:00:05 lr 0.000084 time 0.2798 (0.2857) loss 2.7359 (3.0951) grad_norm 2.9658 (2.7624) [2021-04-16 16:08:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][1240/1251] eta 0:00:03 lr 0.000084 time 0.2460 (0.2856) loss 3.4230 (3.0946) grad_norm 2.9811 (2.7631) [2021-04-16 16:08:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [246/300][1250/1251] eta 0:00:00 lr 0.000084 time 0.2476 (0.2853) loss 3.5210 (3.0967) grad_norm 3.2264 (2.7617) [2021-04-16 16:08:39 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 246 training takes 0:06:01 [2021-04-16 16:08:39 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_246.pth saving...... [2021-04-16 16:08:51 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_246.pth saved !!! [2021-04-16 16:08:52 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.266 (1.266) Loss 0.9067 (0.9067) Acc@1 78.320 (78.320) Acc@5 94.727 (94.727) [2021-04-16 16:08:54 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.753 (0.298) Loss 0.8201 (0.8245) Acc@1 83.105 (80.726) Acc@5 94.434 (95.481) [2021-04-16 16:08:56 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.094 (0.243) Loss 0.8719 (0.8381) Acc@1 78.809 (80.394) Acc@5 94.922 (95.266) [2021-04-16 16:08:58 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.118 (0.236) Loss 0.8630 (0.8458) Acc@1 79.297 (80.267) Acc@5 95.117 (95.095) [2021-04-16 16:09:00 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.250 (0.226) Loss 0.8304 (0.8463) Acc@1 81.641 (80.250) Acc@5 95.215 (95.131) [2021-04-16 16:09:09 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.252 Acc@5 95.224 [2021-04-16 16:09:09 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 80.3% [2021-04-16 16:09:09 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 80.25% [2021-04-16 16:09:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][0/1251] eta 2:28:36 lr 0.000084 time 7.1278 (7.1278) loss 3.4161 (3.4161) grad_norm 2.8091 (2.8091) [2021-04-16 16:09:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][10/1251] eta 0:18:43 lr 0.000084 time 0.3918 (0.9054) loss 3.4188 (3.2424) grad_norm 3.6144 (2.9281) [2021-04-16 16:09:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][20/1251] eta 0:12:27 lr 0.000084 time 0.2580 (0.6071) loss 3.6824 (3.2572) grad_norm 2.5544 (2.8285) [2021-04-16 16:09:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][30/1251] eta 0:10:11 lr 0.000084 time 0.2882 (0.5006) loss 3.7931 (3.2745) grad_norm 2.6214 (2.8113) [2021-04-16 16:09:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][40/1251] eta 0:09:00 lr 0.000084 time 0.2987 (0.4466) loss 3.5674 (3.2628) grad_norm 2.4328 (2.7411) [2021-04-16 16:09:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][50/1251] eta 0:08:15 lr 0.000084 time 0.2585 (0.4125) loss 3.0427 (3.2306) grad_norm 3.1913 (2.7422) [2021-04-16 16:09:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][60/1251] eta 0:07:44 lr 0.000084 time 0.3025 (0.3903) loss 3.4394 (3.2018) grad_norm 2.3807 (2.7276) [2021-04-16 16:09:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][70/1251] eta 0:07:21 lr 0.000084 time 0.2667 (0.3738) loss 3.1535 (3.1732) grad_norm 2.4048 (2.7476) [2021-04-16 16:09:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][80/1251] eta 0:07:03 lr 0.000084 time 0.2848 (0.3615) loss 3.1829 (3.1702) grad_norm 2.6385 (2.7424) [2021-04-16 16:09:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][90/1251] eta 0:06:48 lr 0.000084 time 0.2651 (0.3518) loss 3.2531 (3.1698) grad_norm 2.8450 (2.7299) [2021-04-16 16:09:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][100/1251] eta 0:06:36 lr 0.000084 time 0.2679 (0.3443) loss 3.2541 (3.1766) grad_norm 2.8936 (2.7522) [2021-04-16 16:09:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][110/1251] eta 0:06:25 lr 0.000084 time 0.2947 (0.3381) loss 2.5806 (3.1670) grad_norm 2.5127 (2.7442) [2021-04-16 16:09:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][120/1251] eta 0:06:16 lr 0.000084 time 0.3006 (0.3331) loss 3.3582 (3.1541) grad_norm 3.1970 (2.7556) [2021-04-16 16:09:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][130/1251] eta 0:06:11 lr 0.000084 time 0.2732 (0.3311) loss 2.3146 (3.1357) grad_norm 2.7596 (2.7645) [2021-04-16 16:09:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][140/1251] eta 0:06:03 lr 0.000084 time 0.2568 (0.3272) loss 3.3908 (3.1283) grad_norm 3.5236 (2.7726) [2021-04-16 16:09:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][150/1251] eta 0:05:56 lr 0.000084 time 0.2799 (0.3239) loss 2.8158 (3.1130) grad_norm 2.7374 (2.7783) [2021-04-16 16:10:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][160/1251] eta 0:05:50 lr 0.000084 time 0.2818 (0.3210) loss 2.7678 (3.0973) grad_norm 3.0922 (2.7835) [2021-04-16 16:10:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][170/1251] eta 0:05:44 lr 0.000084 time 0.2871 (0.3186) loss 3.2562 (3.0945) grad_norm 4.0164 (2.7902) [2021-04-16 16:10:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][180/1251] eta 0:05:38 lr 0.000084 time 0.2698 (0.3164) loss 3.1481 (3.0888) grad_norm 3.8968 (2.7906) [2021-04-16 16:10:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][190/1251] eta 0:05:33 lr 0.000084 time 0.3112 (0.3144) loss 3.6380 (3.0926) grad_norm 2.8987 (2.7912) [2021-04-16 16:10:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][200/1251] eta 0:05:28 lr 0.000084 time 0.3773 (0.3130) loss 3.6142 (3.0981) grad_norm 2.7197 (2.7862) [2021-04-16 16:10:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][210/1251] eta 0:05:23 lr 0.000084 time 0.2640 (0.3112) loss 2.2739 (3.0986) grad_norm 3.2809 (2.7922) [2021-04-16 16:10:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][220/1251] eta 0:05:19 lr 0.000084 time 0.2753 (0.3095) loss 3.5138 (3.1007) grad_norm 6.6555 (2.8068) [2021-04-16 16:10:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][230/1251] eta 0:05:15 lr 0.000084 time 0.2917 (0.3086) loss 3.2263 (3.0944) grad_norm 4.2479 (2.8229) [2021-04-16 16:10:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][240/1251] eta 0:05:10 lr 0.000084 time 0.2624 (0.3071) loss 3.3270 (3.0928) grad_norm 2.7730 (2.8212) [2021-04-16 16:10:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][250/1251] eta 0:05:06 lr 0.000084 time 0.2815 (0.3059) loss 3.7076 (3.0937) grad_norm 2.8507 (2.8200) [2021-04-16 16:10:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][260/1251] eta 0:05:01 lr 0.000084 time 0.2621 (0.3047) loss 2.0213 (3.0964) grad_norm 2.3660 (2.8197) [2021-04-16 16:10:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][270/1251] eta 0:04:57 lr 0.000084 time 0.2912 (0.3037) loss 2.9885 (3.0962) grad_norm 2.6780 (2.8211) [2021-04-16 16:10:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][280/1251] eta 0:04:54 lr 0.000084 time 0.3284 (0.3031) loss 2.8206 (3.0993) grad_norm 3.0725 (2.8195) [2021-04-16 16:10:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][290/1251] eta 0:04:50 lr 0.000084 time 0.2769 (0.3021) loss 3.5588 (3.0868) grad_norm 2.9478 (2.8146) [2021-04-16 16:10:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][300/1251] eta 0:04:46 lr 0.000084 time 0.2680 (0.3015) loss 2.1193 (3.0902) grad_norm 2.7920 (2.8158) [2021-04-16 16:10:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][310/1251] eta 0:04:43 lr 0.000084 time 0.2926 (0.3009) loss 3.4656 (3.0857) grad_norm 2.8137 (2.8112) [2021-04-16 16:10:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][320/1251] eta 0:04:39 lr 0.000084 time 0.2682 (0.3006) loss 2.7427 (3.0826) grad_norm 2.7770 (2.8177) [2021-04-16 16:10:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][330/1251] eta 0:04:36 lr 0.000084 time 0.2744 (0.2999) loss 3.4403 (3.0843) grad_norm 2.7232 (2.8177) [2021-04-16 16:10:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][340/1251] eta 0:04:32 lr 0.000084 time 0.3108 (0.2995) loss 3.2391 (3.0936) grad_norm 3.6469 (2.8185) [2021-04-16 16:10:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][350/1251] eta 0:04:29 lr 0.000084 time 0.2547 (0.2991) loss 1.9511 (3.0859) grad_norm 2.4499 (2.8117) [2021-04-16 16:10:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][360/1251] eta 0:04:25 lr 0.000084 time 0.3062 (0.2984) loss 2.4431 (3.0854) grad_norm 3.2303 (2.8059) [2021-04-16 16:11:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][370/1251] eta 0:04:22 lr 0.000083 time 0.2905 (0.2982) loss 3.2361 (3.0880) grad_norm 2.6783 (2.8032) [2021-04-16 16:11:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][380/1251] eta 0:04:19 lr 0.000083 time 0.2476 (0.2975) loss 3.0829 (3.0888) grad_norm 2.8653 (2.8043) [2021-04-16 16:11:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][390/1251] eta 0:04:16 lr 0.000083 time 0.2525 (0.2976) loss 1.9301 (3.0816) grad_norm 3.2228 (2.8029) [2021-04-16 16:11:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][400/1251] eta 0:04:12 lr 0.000083 time 0.2820 (0.2971) loss 2.9839 (3.0845) grad_norm 3.1768 (2.8030) [2021-04-16 16:11:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][410/1251] eta 0:04:09 lr 0.000083 time 0.2757 (0.2968) loss 3.2690 (3.0802) grad_norm 2.6908 (2.7980) [2021-04-16 16:11:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][420/1251] eta 0:04:06 lr 0.000083 time 0.3000 (0.2965) loss 3.6929 (3.0869) grad_norm 2.3250 (2.7980) [2021-04-16 16:11:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][430/1251] eta 0:04:03 lr 0.000083 time 0.2867 (0.2962) loss 3.2406 (3.0886) grad_norm 2.3561 (2.7960) [2021-04-16 16:11:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][440/1251] eta 0:04:00 lr 0.000083 time 0.2495 (0.2960) loss 2.5268 (3.0841) grad_norm 2.2914 (2.7924) [2021-04-16 16:11:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][450/1251] eta 0:03:56 lr 0.000083 time 0.2741 (0.2955) loss 3.2135 (3.0842) grad_norm 2.9794 (2.7883) [2021-04-16 16:11:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][460/1251] eta 0:03:53 lr 0.000083 time 0.2581 (0.2951) loss 3.4500 (3.0876) grad_norm 2.6666 (2.7853) [2021-04-16 16:11:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][470/1251] eta 0:03:50 lr 0.000083 time 0.2653 (0.2947) loss 3.2597 (3.0874) grad_norm 3.1111 (2.7856) [2021-04-16 16:11:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][480/1251] eta 0:03:46 lr 0.000083 time 0.2757 (0.2944) loss 2.4073 (3.0826) grad_norm 2.5837 (2.7897) [2021-04-16 16:11:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][490/1251] eta 0:03:43 lr 0.000083 time 0.2903 (0.2940) loss 3.4818 (3.0827) grad_norm 2.4947 (2.7867) [2021-04-16 16:11:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][500/1251] eta 0:03:40 lr 0.000083 time 0.2778 (0.2935) loss 2.9614 (3.0849) grad_norm 2.6957 (2.7888) [2021-04-16 16:11:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][510/1251] eta 0:03:37 lr 0.000083 time 0.2817 (0.2932) loss 2.5986 (3.0842) grad_norm 3.6319 (2.7916) [2021-04-16 16:11:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][520/1251] eta 0:03:34 lr 0.000083 time 0.2987 (0.2929) loss 3.1720 (3.0879) grad_norm 2.7494 (2.7899) [2021-04-16 16:11:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][530/1251] eta 0:03:30 lr 0.000083 time 0.2823 (0.2926) loss 3.2077 (3.0895) grad_norm 3.0876 (2.7951) [2021-04-16 16:11:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][540/1251] eta 0:03:27 lr 0.000083 time 0.2860 (0.2925) loss 2.9534 (3.0883) grad_norm 2.3095 (2.7920) [2021-04-16 16:11:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][550/1251] eta 0:03:24 lr 0.000083 time 0.2893 (0.2923) loss 3.6825 (3.0889) grad_norm 3.2197 (2.7944) [2021-04-16 16:11:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][560/1251] eta 0:03:21 lr 0.000083 time 0.2647 (0.2920) loss 3.0080 (3.0865) grad_norm 3.4106 (2.7944) [2021-04-16 16:11:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][570/1251] eta 0:03:18 lr 0.000083 time 0.2703 (0.2920) loss 2.2985 (3.0821) grad_norm 2.3803 (2.7952) [2021-04-16 16:11:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][580/1251] eta 0:03:16 lr 0.000083 time 0.2621 (0.2922) loss 3.6760 (3.0858) grad_norm 3.8580 (2.7970) [2021-04-16 16:12:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][590/1251] eta 0:03:13 lr 0.000083 time 0.2518 (0.2922) loss 3.3982 (3.0817) grad_norm 2.5789 (2.7912) [2021-04-16 16:12:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][600/1251] eta 0:03:10 lr 0.000083 time 0.2771 (0.2920) loss 3.5436 (3.0831) grad_norm 2.8330 (2.7933) [2021-04-16 16:12:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][610/1251] eta 0:03:07 lr 0.000083 time 0.3030 (0.2919) loss 3.4411 (3.0838) grad_norm 3.0275 (2.7962) [2021-04-16 16:12:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][620/1251] eta 0:03:04 lr 0.000083 time 0.2873 (0.2917) loss 2.1960 (3.0816) grad_norm 2.7691 (2.7948) [2021-04-16 16:12:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][630/1251] eta 0:03:00 lr 0.000083 time 0.2904 (0.2914) loss 3.0384 (3.0796) grad_norm 2.5405 (2.7931) [2021-04-16 16:12:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][640/1251] eta 0:02:57 lr 0.000083 time 0.2670 (0.2910) loss 3.4171 (3.0802) grad_norm 3.0873 (2.7946) [2021-04-16 16:12:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][650/1251] eta 0:02:54 lr 0.000083 time 0.2664 (0.2908) loss 3.3321 (3.0843) grad_norm 2.9014 (2.7959) [2021-04-16 16:12:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][660/1251] eta 0:02:51 lr 0.000083 time 0.2919 (0.2905) loss 3.0318 (3.0858) grad_norm 2.9788 (2.7957) [2021-04-16 16:12:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][670/1251] eta 0:02:48 lr 0.000083 time 0.2467 (0.2902) loss 3.3863 (3.0877) grad_norm 2.8396 (2.7966) [2021-04-16 16:12:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][680/1251] eta 0:02:45 lr 0.000083 time 0.2706 (0.2900) loss 3.7613 (3.0860) grad_norm 2.5495 (2.7952) [2021-04-16 16:12:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][690/1251] eta 0:02:42 lr 0.000083 time 0.2780 (0.2898) loss 3.3266 (3.0873) grad_norm 3.7871 (2.7964) [2021-04-16 16:12:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][700/1251] eta 0:02:39 lr 0.000083 time 0.4251 (0.2898) loss 3.8264 (3.0907) grad_norm 2.3725 (2.7943) [2021-04-16 16:12:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][710/1251] eta 0:02:36 lr 0.000083 time 0.3058 (0.2896) loss 3.3736 (3.0861) grad_norm 2.9889 (2.7924) [2021-04-16 16:12:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][720/1251] eta 0:02:33 lr 0.000083 time 0.2567 (0.2895) loss 3.5191 (3.0890) grad_norm 2.7880 (2.7934) [2021-04-16 16:12:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][730/1251] eta 0:02:30 lr 0.000083 time 0.2831 (0.2893) loss 2.5944 (3.0868) grad_norm 2.9158 (2.7926) [2021-04-16 16:12:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][740/1251] eta 0:02:27 lr 0.000083 time 0.3042 (0.2892) loss 3.4440 (3.0839) grad_norm 2.9488 (2.7940) [2021-04-16 16:12:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][750/1251] eta 0:02:24 lr 0.000083 time 0.2864 (0.2890) loss 3.5507 (3.0853) grad_norm 2.4574 (2.7935) [2021-04-16 16:12:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][760/1251] eta 0:02:21 lr 0.000083 time 0.2781 (0.2888) loss 3.4885 (3.0865) grad_norm 2.7131 (2.7930) [2021-04-16 16:12:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][770/1251] eta 0:02:18 lr 0.000083 time 0.2685 (0.2886) loss 2.9242 (3.0859) grad_norm 2.5949 (2.7941) [2021-04-16 16:12:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][780/1251] eta 0:02:15 lr 0.000083 time 0.2649 (0.2886) loss 2.1054 (3.0836) grad_norm 3.3415 (2.7937) [2021-04-16 16:12:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][790/1251] eta 0:02:12 lr 0.000083 time 0.2497 (0.2883) loss 2.6711 (3.0830) grad_norm 2.7275 (2.7938) [2021-04-16 16:13:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][800/1251] eta 0:02:09 lr 0.000083 time 0.2610 (0.2882) loss 3.5095 (3.0834) grad_norm 2.5390 (2.7981) [2021-04-16 16:13:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][810/1251] eta 0:02:07 lr 0.000083 time 0.2691 (0.2882) loss 2.6322 (3.0819) grad_norm 2.7564 (2.8005) [2021-04-16 16:13:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][820/1251] eta 0:02:04 lr 0.000083 time 0.2935 (0.2881) loss 2.8865 (3.0821) grad_norm 3.2607 (2.8011) [2021-04-16 16:13:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][830/1251] eta 0:02:01 lr 0.000083 time 0.2942 (0.2880) loss 3.2067 (3.0849) grad_norm 3.4328 (2.8020) [2021-04-16 16:13:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][840/1251] eta 0:01:58 lr 0.000082 time 0.2872 (0.2879) loss 2.0532 (3.0821) grad_norm 2.4876 (2.8014) [2021-04-16 16:13:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][850/1251] eta 0:01:55 lr 0.000082 time 0.2810 (0.2878) loss 3.2710 (3.0808) grad_norm 2.7623 (2.8008) [2021-04-16 16:13:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][860/1251] eta 0:01:52 lr 0.000082 time 0.2771 (0.2877) loss 3.5071 (3.0816) grad_norm 2.6104 (2.7995) [2021-04-16 16:13:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][870/1251] eta 0:01:49 lr 0.000082 time 0.2850 (0.2876) loss 2.9744 (3.0800) grad_norm 2.4093 (2.7980) [2021-04-16 16:13:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][880/1251] eta 0:01:46 lr 0.000082 time 0.2733 (0.2876) loss 3.6838 (3.0776) grad_norm 2.5195 (2.7976) [2021-04-16 16:13:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][890/1251] eta 0:01:43 lr 0.000082 time 0.2660 (0.2876) loss 2.6419 (3.0754) grad_norm 2.1874 (2.7983) [2021-04-16 16:13:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][900/1251] eta 0:01:40 lr 0.000082 time 0.2646 (0.2875) loss 2.5391 (3.0747) grad_norm 2.6016 (2.8014) [2021-04-16 16:13:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][910/1251] eta 0:01:37 lr 0.000082 time 0.2958 (0.2874) loss 3.0503 (3.0746) grad_norm 2.8117 (2.8011) [2021-04-16 16:13:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][920/1251] eta 0:01:35 lr 0.000082 time 0.2901 (0.2873) loss 3.0869 (3.0740) grad_norm 2.9468 (2.8011) [2021-04-16 16:13:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][930/1251] eta 0:01:32 lr 0.000082 time 0.2840 (0.2872) loss 3.8283 (3.0769) grad_norm 2.5722 (2.8108) [2021-04-16 16:13:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][940/1251] eta 0:01:29 lr 0.000082 time 0.4439 (0.2873) loss 3.3596 (3.0800) grad_norm 3.2139 (2.8136) [2021-04-16 16:13:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][950/1251] eta 0:01:26 lr 0.000082 time 0.2698 (0.2871) loss 3.4733 (3.0831) grad_norm 2.5855 (2.8151) [2021-04-16 16:13:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][960/1251] eta 0:01:23 lr 0.000082 time 0.2705 (0.2870) loss 3.0103 (3.0820) grad_norm 2.5670 (2.8128) [2021-04-16 16:13:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][970/1251] eta 0:01:20 lr 0.000082 time 0.2628 (0.2871) loss 2.6731 (3.0809) grad_norm 2.5440 (2.8111) [2021-04-16 16:13:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][980/1251] eta 0:01:17 lr 0.000082 time 0.2581 (0.2869) loss 3.1783 (3.0822) grad_norm 2.4645 (2.8092) [2021-04-16 16:13:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][990/1251] eta 0:01:14 lr 0.000082 time 0.2743 (0.2868) loss 2.8498 (3.0823) grad_norm 2.6426 (2.8076) [2021-04-16 16:13:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][1000/1251] eta 0:01:11 lr 0.000082 time 0.2824 (0.2867) loss 3.4244 (3.0818) grad_norm 2.5707 (2.8053) [2021-04-16 16:13:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][1010/1251] eta 0:01:09 lr 0.000082 time 0.2815 (0.2866) loss 3.4284 (3.0832) grad_norm 2.3546 (2.8056) [2021-04-16 16:14:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][1020/1251] eta 0:01:06 lr 0.000082 time 0.2563 (0.2867) loss 2.7358 (3.0834) grad_norm 2.4703 (2.8058) [2021-04-16 16:14:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][1030/1251] eta 0:01:03 lr 0.000082 time 0.2979 (0.2866) loss 2.9970 (3.0835) grad_norm 2.4802 (2.8063) [2021-04-16 16:14:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][1040/1251] eta 0:01:00 lr 0.000082 time 0.2866 (0.2865) loss 3.4905 (3.0841) grad_norm 3.0345 (2.8048) [2021-04-16 16:14:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][1050/1251] eta 0:00:57 lr 0.000082 time 0.2667 (0.2864) loss 3.2773 (3.0825) grad_norm 3.2026 (2.8039) [2021-04-16 16:14:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][1060/1251] eta 0:00:54 lr 0.000082 time 0.2741 (0.2862) loss 2.0823 (3.0836) grad_norm 2.4308 (2.8027) [2021-04-16 16:14:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][1070/1251] eta 0:00:51 lr 0.000082 time 0.2702 (0.2861) loss 3.4595 (3.0841) grad_norm 3.3045 (2.8043) [2021-04-16 16:14:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][1080/1251] eta 0:00:48 lr 0.000082 time 0.2924 (0.2860) loss 2.1560 (3.0849) grad_norm 2.4969 (2.8055) [2021-04-16 16:14:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][1090/1251] eta 0:00:46 lr 0.000082 time 0.2727 (0.2859) loss 2.7749 (3.0856) grad_norm 2.9497 (2.8059) [2021-04-16 16:14:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][1100/1251] eta 0:00:43 lr 0.000082 time 0.2604 (0.2858) loss 2.6935 (3.0826) grad_norm 2.4272 (2.8052) [2021-04-16 16:14:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][1110/1251] eta 0:00:40 lr 0.000082 time 0.2452 (0.2857) loss 3.3523 (3.0834) grad_norm 2.6668 (2.8043) [2021-04-16 16:14:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][1120/1251] eta 0:00:37 lr 0.000082 time 0.2965 (0.2858) loss 3.5518 (3.0850) grad_norm 2.9659 (2.8027) [2021-04-16 16:14:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][1130/1251] eta 0:00:34 lr 0.000082 time 0.2763 (0.2856) loss 2.4110 (3.0847) grad_norm 2.5275 (2.8024) [2021-04-16 16:14:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][1140/1251] eta 0:00:31 lr 0.000082 time 0.2704 (0.2857) loss 2.9725 (3.0845) grad_norm 2.7657 (2.8021) [2021-04-16 16:14:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][1150/1251] eta 0:00:28 lr 0.000082 time 0.2754 (0.2856) loss 3.2163 (3.0839) grad_norm 2.9320 (2.8039) [2021-04-16 16:14:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][1160/1251] eta 0:00:26 lr 0.000082 time 0.2483 (0.2857) loss 2.5444 (3.0828) grad_norm 2.9785 (2.8029) [2021-04-16 16:14:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][1170/1251] eta 0:00:23 lr 0.000082 time 0.2823 (0.2857) loss 3.6397 (3.0816) grad_norm 2.7731 (2.8027) [2021-04-16 16:14:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][1180/1251] eta 0:00:20 lr 0.000082 time 0.2884 (0.2856) loss 2.0475 (3.0796) grad_norm 2.5267 (2.8036) [2021-04-16 16:14:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][1190/1251] eta 0:00:17 lr 0.000082 time 0.2981 (0.2856) loss 2.1251 (3.0794) grad_norm 2.5328 (2.8027) [2021-04-16 16:14:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][1200/1251] eta 0:00:14 lr 0.000082 time 0.2940 (0.2855) loss 3.2479 (3.0793) grad_norm 2.5624 (2.8021) [2021-04-16 16:14:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][1210/1251] eta 0:00:11 lr 0.000082 time 0.2738 (0.2854) loss 3.6016 (3.0811) grad_norm 2.7037 (2.8018) [2021-04-16 16:14:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][1220/1251] eta 0:00:08 lr 0.000082 time 0.2624 (0.2854) loss 3.7469 (3.0816) grad_norm 2.6679 (2.8030) [2021-04-16 16:15:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][1230/1251] eta 0:00:05 lr 0.000082 time 0.2951 (0.2853) loss 3.1700 (3.0804) grad_norm 2.3271 (2.8027) [2021-04-16 16:15:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][1240/1251] eta 0:00:03 lr 0.000082 time 0.2494 (0.2852) loss 3.8217 (3.0822) grad_norm 2.4180 (2.8019) [2021-04-16 16:15:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [247/300][1250/1251] eta 0:00:00 lr 0.000082 time 0.2486 (0.2850) loss 3.1274 (3.0844) grad_norm 3.1006 (2.8035) [2021-04-16 16:15:09 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 247 training takes 0:06:00 [2021-04-16 16:15:09 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_247.pth saving...... [2021-04-16 16:15:29 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_247.pth saved !!! [2021-04-16 16:15:30 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.078 (1.078) Loss 0.8009 (0.8009) Acc@1 82.715 (82.715) Acc@5 95.508 (95.508) [2021-04-16 16:15:32 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.462 (0.248) Loss 0.8200 (0.8276) Acc@1 81.543 (80.771) Acc@5 96.191 (95.250) [2021-04-16 16:15:34 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.101 (0.253) Loss 0.8701 (0.8425) Acc@1 80.078 (80.469) Acc@5 94.434 (95.057) [2021-04-16 16:15:36 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.122 (0.233) Loss 0.8380 (0.8485) Acc@1 80.078 (80.242) Acc@5 94.727 (95.120) [2021-04-16 16:15:38 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.309 (0.216) Loss 0.8053 (0.8451) Acc@1 80.371 (80.266) Acc@5 95.703 (95.186) [2021-04-16 16:15:55 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.278 Acc@5 95.138 [2021-04-16 16:15:55 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 80.3% [2021-04-16 16:15:55 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 80.28% [2021-04-16 16:16:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][0/1251] eta 2:30:27 lr 0.000082 time 7.2159 (7.2159) loss 3.0400 (3.0400) grad_norm 4.1900 (4.1900) [2021-04-16 16:16:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][10/1251] eta 0:18:39 lr 0.000082 time 0.2420 (0.9024) loss 3.5663 (3.2455) grad_norm 4.1783 (2.9810) [2021-04-16 16:16:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][20/1251] eta 0:12:33 lr 0.000082 time 0.2702 (0.6121) loss 3.8521 (3.2417) grad_norm 3.0596 (2.8963) [2021-04-16 16:16:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][30/1251] eta 0:10:16 lr 0.000082 time 0.2762 (0.5047) loss 3.0693 (3.2319) grad_norm 2.5505 (2.8721) [2021-04-16 16:16:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][40/1251] eta 0:09:04 lr 0.000082 time 0.2583 (0.4499) loss 3.3206 (3.2215) grad_norm 2.2611 (2.8505) [2021-04-16 16:16:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][50/1251] eta 0:08:20 lr 0.000081 time 0.3019 (0.4163) loss 3.8314 (3.2339) grad_norm 2.5131 (2.8373) [2021-04-16 16:16:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][60/1251] eta 0:07:47 lr 0.000081 time 0.2797 (0.3929) loss 3.7178 (3.2460) grad_norm 3.0685 (2.8567) [2021-04-16 16:16:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][70/1251] eta 0:07:24 lr 0.000081 time 0.2655 (0.3766) loss 3.2636 (3.2577) grad_norm 3.3447 (2.8846) [2021-04-16 16:16:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][80/1251] eta 0:07:06 lr 0.000081 time 0.3113 (0.3641) loss 3.3578 (3.2522) grad_norm 2.9740 (2.8863) [2021-04-16 16:16:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][90/1251] eta 0:06:52 lr 0.000081 time 0.2714 (0.3557) loss 3.3082 (3.2382) grad_norm 2.6274 (2.8940) [2021-04-16 16:16:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][100/1251] eta 0:06:41 lr 0.000081 time 0.2806 (0.3490) loss 1.9433 (3.1827) grad_norm 2.9286 (2.8734) [2021-04-16 16:16:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][110/1251] eta 0:06:31 lr 0.000081 time 0.2862 (0.3434) loss 3.9029 (3.1863) grad_norm 2.3361 (2.8833) [2021-04-16 16:16:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][120/1251] eta 0:06:22 lr 0.000081 time 0.2689 (0.3382) loss 2.8213 (3.1797) grad_norm 3.7248 (2.8964) [2021-04-16 16:16:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][130/1251] eta 0:06:14 lr 0.000081 time 0.2824 (0.3345) loss 3.1746 (3.1771) grad_norm 2.3114 (2.8826) [2021-04-16 16:16:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][140/1251] eta 0:06:06 lr 0.000081 time 0.2656 (0.3300) loss 3.1183 (3.1915) grad_norm 2.5382 (2.8780) [2021-04-16 16:16:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][150/1251] eta 0:06:00 lr 0.000081 time 0.4711 (0.3278) loss 2.7480 (3.1979) grad_norm 2.6179 (2.8812) [2021-04-16 16:16:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][160/1251] eta 0:05:53 lr 0.000081 time 0.2921 (0.3243) loss 2.3660 (3.1880) grad_norm 2.6721 (2.8634) [2021-04-16 16:16:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][170/1251] eta 0:05:47 lr 0.000081 time 0.2696 (0.3216) loss 2.4464 (3.1901) grad_norm 2.4655 (2.8527) [2021-04-16 16:16:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][180/1251] eta 0:05:42 lr 0.000081 time 0.2718 (0.3197) loss 2.4246 (3.1846) grad_norm 2.4742 (2.8449) [2021-04-16 16:16:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][190/1251] eta 0:05:36 lr 0.000081 time 0.3128 (0.3175) loss 2.0882 (3.1757) grad_norm 3.0079 (2.8394) [2021-04-16 16:16:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][200/1251] eta 0:05:31 lr 0.000081 time 0.3035 (0.3157) loss 3.5779 (3.1737) grad_norm 2.7830 (2.8330) [2021-04-16 16:17:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][210/1251] eta 0:05:26 lr 0.000081 time 0.2752 (0.3139) loss 3.0772 (3.1853) grad_norm 2.6704 (2.8261) [2021-04-16 16:17:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][220/1251] eta 0:05:21 lr 0.000081 time 0.2616 (0.3122) loss 3.5509 (3.1852) grad_norm 2.4414 (2.8244) [2021-04-16 16:17:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][230/1251] eta 0:05:18 lr 0.000081 time 0.2809 (0.3115) loss 3.4963 (3.1924) grad_norm 3.9928 (2.8259) [2021-04-16 16:17:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][240/1251] eta 0:05:13 lr 0.000081 time 0.2902 (0.3100) loss 3.0959 (3.1786) grad_norm 2.9695 (2.8266) [2021-04-16 16:17:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][250/1251] eta 0:05:08 lr 0.000081 time 0.2586 (0.3087) loss 2.7451 (3.1700) grad_norm 3.2183 (2.8273) [2021-04-16 16:17:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][260/1251] eta 0:05:04 lr 0.000081 time 0.2674 (0.3075) loss 3.1984 (3.1710) grad_norm 2.6490 (2.8250) [2021-04-16 16:17:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][270/1251] eta 0:05:00 lr 0.000081 time 0.2790 (0.3064) loss 3.2915 (3.1666) grad_norm 2.3945 (2.8218) [2021-04-16 16:17:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][280/1251] eta 0:04:56 lr 0.000081 time 0.2843 (0.3053) loss 2.7208 (3.1642) grad_norm 2.4475 (2.8175) [2021-04-16 16:17:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][290/1251] eta 0:04:52 lr 0.000081 time 0.2857 (0.3044) loss 2.6906 (3.1492) grad_norm 2.8791 (2.8151) [2021-04-16 16:17:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][300/1251] eta 0:04:48 lr 0.000081 time 0.3023 (0.3036) loss 3.4780 (3.1452) grad_norm 2.3009 (2.8091) [2021-04-16 16:17:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][310/1251] eta 0:04:44 lr 0.000081 time 0.2697 (0.3027) loss 3.4935 (3.1456) grad_norm 2.8323 (2.8140) [2021-04-16 16:17:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][320/1251] eta 0:04:41 lr 0.000081 time 0.2716 (0.3024) loss 3.6985 (3.1412) grad_norm 3.9221 (2.8188) [2021-04-16 16:17:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][330/1251] eta 0:04:37 lr 0.000081 time 0.2901 (0.3015) loss 3.3492 (3.1374) grad_norm 3.5189 (2.8206) [2021-04-16 16:17:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][340/1251] eta 0:04:34 lr 0.000081 time 0.2675 (0.3012) loss 2.8111 (3.1392) grad_norm 2.5552 (2.8185) [2021-04-16 16:17:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][350/1251] eta 0:04:31 lr 0.000081 time 0.4288 (0.3013) loss 3.6555 (3.1452) grad_norm 2.8001 (2.8153) [2021-04-16 16:17:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][360/1251] eta 0:04:28 lr 0.000081 time 0.2706 (0.3009) loss 3.1877 (3.1497) grad_norm 2.6046 (2.8126) [2021-04-16 16:17:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][370/1251] eta 0:04:24 lr 0.000081 time 0.2667 (0.3006) loss 2.3411 (3.1454) grad_norm 2.5188 (2.8091) [2021-04-16 16:17:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][380/1251] eta 0:04:21 lr 0.000081 time 0.2464 (0.2998) loss 1.9808 (3.1473) grad_norm 2.3203 (2.8117) [2021-04-16 16:17:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][390/1251] eta 0:04:17 lr 0.000081 time 0.2531 (0.2992) loss 3.1736 (3.1430) grad_norm 2.3508 (2.8090) [2021-04-16 16:17:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][400/1251] eta 0:04:14 lr 0.000081 time 0.2580 (0.2985) loss 3.8945 (3.1544) grad_norm 2.9012 (2.8097) [2021-04-16 16:17:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][410/1251] eta 0:04:10 lr 0.000081 time 0.2751 (0.2979) loss 3.9104 (3.1594) grad_norm 2.8281 (2.8094) [2021-04-16 16:18:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][420/1251] eta 0:04:07 lr 0.000081 time 0.2673 (0.2974) loss 3.1501 (3.1597) grad_norm 2.4863 (2.8051) [2021-04-16 16:18:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][430/1251] eta 0:04:03 lr 0.000081 time 0.2570 (0.2968) loss 3.6987 (3.1660) grad_norm 2.3531 (2.8042) [2021-04-16 16:18:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][440/1251] eta 0:04:00 lr 0.000081 time 0.2628 (0.2964) loss 2.3520 (3.1572) grad_norm 2.4767 (2.8018) [2021-04-16 16:18:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][450/1251] eta 0:03:57 lr 0.000081 time 0.2820 (0.2960) loss 3.5813 (3.1579) grad_norm 2.9752 (2.7993) [2021-04-16 16:18:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][460/1251] eta 0:03:53 lr 0.000081 time 0.3093 (0.2956) loss 3.0971 (3.1579) grad_norm 3.4571 (2.7995) [2021-04-16 16:18:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][470/1251] eta 0:03:50 lr 0.000081 time 0.2760 (0.2952) loss 2.8223 (3.1602) grad_norm 2.3400 (2.7990) [2021-04-16 16:18:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][480/1251] eta 0:03:47 lr 0.000081 time 0.2591 (0.2948) loss 3.4320 (3.1620) grad_norm 2.4113 (2.7968) [2021-04-16 16:18:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][490/1251] eta 0:03:44 lr 0.000081 time 0.3088 (0.2944) loss 3.6282 (3.1606) grad_norm 2.3727 (2.7923) [2021-04-16 16:18:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][500/1251] eta 0:03:40 lr 0.000081 time 0.2578 (0.2940) loss 3.2853 (3.1594) grad_norm 2.5676 (2.7907) [2021-04-16 16:18:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][510/1251] eta 0:03:37 lr 0.000081 time 0.2739 (0.2936) loss 3.1004 (3.1591) grad_norm 2.6608 (2.7883) [2021-04-16 16:18:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][520/1251] eta 0:03:34 lr 0.000080 time 0.2736 (0.2932) loss 2.9922 (3.1584) grad_norm 2.2649 (2.7854) [2021-04-16 16:18:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][530/1251] eta 0:03:31 lr 0.000080 time 0.2530 (0.2929) loss 3.7234 (3.1597) grad_norm 3.0621 (2.7851) [2021-04-16 16:18:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][540/1251] eta 0:03:28 lr 0.000080 time 0.2763 (0.2926) loss 3.9796 (3.1629) grad_norm 2.2779 (2.7853) [2021-04-16 16:18:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][550/1251] eta 0:03:24 lr 0.000080 time 0.2553 (0.2923) loss 3.9369 (3.1647) grad_norm 3.1048 (2.7832) [2021-04-16 16:18:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][560/1251] eta 0:03:21 lr 0.000080 time 0.2513 (0.2920) loss 2.3441 (3.1621) grad_norm 2.3874 (2.7851) [2021-04-16 16:18:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][570/1251] eta 0:03:18 lr 0.000080 time 0.4199 (0.2920) loss 3.7325 (3.1664) grad_norm 2.6460 (2.7874) [2021-04-16 16:18:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][580/1251] eta 0:03:15 lr 0.000080 time 0.2924 (0.2919) loss 2.0580 (3.1638) grad_norm 2.5307 (2.7876) [2021-04-16 16:18:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][590/1251] eta 0:03:12 lr 0.000080 time 0.2822 (0.2917) loss 3.4669 (3.1573) grad_norm 4.0627 (2.7884) [2021-04-16 16:18:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][600/1251] eta 0:03:09 lr 0.000080 time 0.2632 (0.2914) loss 3.0449 (3.1559) grad_norm 2.8095 (2.7880) [2021-04-16 16:18:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][610/1251] eta 0:03:06 lr 0.000080 time 0.2812 (0.2911) loss 1.8510 (3.1493) grad_norm 2.2018 (2.7882) [2021-04-16 16:18:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][620/1251] eta 0:03:03 lr 0.000080 time 0.2575 (0.2908) loss 2.6366 (3.1473) grad_norm 2.6854 (2.7878) [2021-04-16 16:18:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][630/1251] eta 0:03:00 lr 0.000080 time 0.2890 (0.2906) loss 3.3413 (3.1497) grad_norm 2.3879 (2.7855) [2021-04-16 16:19:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][640/1251] eta 0:02:57 lr 0.000080 time 0.2552 (0.2903) loss 2.8870 (3.1491) grad_norm 2.4313 (2.7810) [2021-04-16 16:19:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][650/1251] eta 0:02:54 lr 0.000080 time 0.2882 (0.2901) loss 3.3074 (3.1462) grad_norm 3.5575 (2.7786) [2021-04-16 16:19:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][660/1251] eta 0:02:51 lr 0.000080 time 0.2900 (0.2900) loss 3.1477 (3.1457) grad_norm 2.2536 (2.7750) [2021-04-16 16:19:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][670/1251] eta 0:02:48 lr 0.000080 time 0.2901 (0.2901) loss 3.3952 (3.1430) grad_norm 2.4625 (2.7730) [2021-04-16 16:19:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][680/1251] eta 0:02:45 lr 0.000080 time 0.2949 (0.2901) loss 3.4029 (3.1455) grad_norm 2.6022 (2.7732) [2021-04-16 16:19:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][690/1251] eta 0:02:42 lr 0.000080 time 0.2912 (0.2899) loss 3.4532 (3.1451) grad_norm 2.5805 (2.7742) [2021-04-16 16:19:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][700/1251] eta 0:02:39 lr 0.000080 time 0.3737 (0.2898) loss 3.3504 (3.1421) grad_norm 2.3861 (2.7776) [2021-04-16 16:19:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][710/1251] eta 0:02:36 lr 0.000080 time 0.2751 (0.2896) loss 3.4122 (3.1404) grad_norm 2.7407 (2.7744) [2021-04-16 16:19:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][720/1251] eta 0:02:33 lr 0.000080 time 0.2777 (0.2894) loss 2.6814 (3.1363) grad_norm 2.7060 (2.7707) [2021-04-16 16:19:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][730/1251] eta 0:02:30 lr 0.000080 time 0.2699 (0.2894) loss 3.3044 (3.1362) grad_norm 2.5464 (2.7691) [2021-04-16 16:19:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][740/1251] eta 0:02:27 lr 0.000080 time 0.2538 (0.2892) loss 3.3523 (3.1358) grad_norm 2.8649 (2.7695) [2021-04-16 16:19:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][750/1251] eta 0:02:24 lr 0.000080 time 0.2747 (0.2890) loss 2.4645 (3.1346) grad_norm 3.5136 (2.7696) [2021-04-16 16:19:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][760/1251] eta 0:02:21 lr 0.000080 time 0.2658 (0.2890) loss 3.4708 (3.1316) grad_norm 3.6501 (2.7684) [2021-04-16 16:19:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][770/1251] eta 0:02:18 lr 0.000080 time 0.2536 (0.2888) loss 3.4654 (3.1276) grad_norm 3.2172 (2.7670) [2021-04-16 16:19:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][780/1251] eta 0:02:15 lr 0.000080 time 0.2722 (0.2887) loss 2.8119 (3.1243) grad_norm 3.3769 (2.7693) [2021-04-16 16:19:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][790/1251] eta 0:02:12 lr 0.000080 time 0.2677 (0.2885) loss 3.8327 (3.1255) grad_norm 3.0303 (2.7689) [2021-04-16 16:19:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][800/1251] eta 0:02:10 lr 0.000080 time 0.4801 (0.2886) loss 3.1714 (3.1226) grad_norm 2.4741 (2.7668) [2021-04-16 16:19:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][810/1251] eta 0:02:07 lr 0.000080 time 0.2817 (0.2884) loss 3.8333 (3.1252) grad_norm 2.8576 (2.7686) [2021-04-16 16:19:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][820/1251] eta 0:02:04 lr 0.000080 time 0.2673 (0.2882) loss 3.5628 (3.1256) grad_norm 2.3799 (2.7687) [2021-04-16 16:19:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][830/1251] eta 0:02:01 lr 0.000080 time 0.2543 (0.2881) loss 2.9974 (3.1237) grad_norm 3.0308 (2.7710) [2021-04-16 16:19:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][840/1251] eta 0:01:58 lr 0.000080 time 0.2754 (0.2879) loss 3.7750 (3.1259) grad_norm 2.7803 (2.7712) [2021-04-16 16:20:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][850/1251] eta 0:01:55 lr 0.000080 time 0.2635 (0.2878) loss 3.0808 (3.1272) grad_norm 2.6900 (2.7720) [2021-04-16 16:20:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][860/1251] eta 0:01:52 lr 0.000080 time 0.2748 (0.2876) loss 2.0034 (3.1277) grad_norm 2.7253 (2.7706) [2021-04-16 16:20:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][870/1251] eta 0:01:49 lr 0.000080 time 0.2798 (0.2875) loss 3.0144 (3.1222) grad_norm 2.8197 (2.7708) [2021-04-16 16:20:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][880/1251] eta 0:01:46 lr 0.000080 time 0.2939 (0.2874) loss 3.2344 (3.1224) grad_norm 2.7140 (2.7706) [2021-04-16 16:20:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][890/1251] eta 0:01:43 lr 0.000080 time 0.2908 (0.2873) loss 2.7525 (3.1199) grad_norm 2.9642 (2.7704) [2021-04-16 16:20:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][900/1251] eta 0:01:40 lr 0.000080 time 0.2716 (0.2873) loss 3.0709 (3.1191) grad_norm 2.4505 (2.7687) [2021-04-16 16:20:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][910/1251] eta 0:01:37 lr 0.000080 time 0.2528 (0.2872) loss 3.4110 (3.1206) grad_norm 2.7264 (2.7689) [2021-04-16 16:20:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][920/1251] eta 0:01:35 lr 0.000080 time 0.2657 (0.2872) loss 2.0100 (3.1212) grad_norm 2.8671 (2.7677) [2021-04-16 16:20:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][930/1251] eta 0:01:32 lr 0.000080 time 0.2975 (0.2875) loss 3.4746 (3.1225) grad_norm 2.6768 (2.7685) [2021-04-16 16:20:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][940/1251] eta 0:01:29 lr 0.000080 time 0.2611 (0.2875) loss 1.9596 (3.1231) grad_norm 2.7271 (2.7677) [2021-04-16 16:20:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][950/1251] eta 0:01:26 lr 0.000080 time 0.2883 (0.2873) loss 2.4190 (3.1215) grad_norm 2.7035 (2.7678) [2021-04-16 16:20:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][960/1251] eta 0:01:23 lr 0.000080 time 0.2642 (0.2872) loss 1.7788 (3.1208) grad_norm 2.8006 (2.7686) [2021-04-16 16:20:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][970/1251] eta 0:01:20 lr 0.000080 time 0.2786 (0.2871) loss 2.6186 (3.1209) grad_norm 2.5258 (2.7672) [2021-04-16 16:20:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][980/1251] eta 0:01:17 lr 0.000080 time 0.2967 (0.2870) loss 3.9235 (3.1230) grad_norm 2.9239 (2.7686) [2021-04-16 16:20:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][990/1251] eta 0:01:14 lr 0.000079 time 0.2638 (0.2869) loss 3.4589 (3.1254) grad_norm 2.4237 (2.7673) [2021-04-16 16:20:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][1000/1251] eta 0:01:11 lr 0.000079 time 0.2856 (0.2867) loss 2.5301 (3.1245) grad_norm 3.0710 (2.7667) [2021-04-16 16:20:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][1010/1251] eta 0:01:09 lr 0.000079 time 0.2649 (0.2866) loss 2.7866 (3.1255) grad_norm 3.0686 (2.7692) [2021-04-16 16:20:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][1020/1251] eta 0:01:06 lr 0.000079 time 0.2819 (0.2865) loss 3.3846 (3.1272) grad_norm 2.2010 (2.7687) [2021-04-16 16:20:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][1030/1251] eta 0:01:03 lr 0.000079 time 0.2967 (0.2864) loss 2.7981 (3.1281) grad_norm 2.6905 (2.7676) [2021-04-16 16:20:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][1040/1251] eta 0:01:00 lr 0.000079 time 0.2842 (0.2863) loss 3.4639 (3.1293) grad_norm 3.0729 (2.7663) [2021-04-16 16:20:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][1050/1251] eta 0:00:57 lr 0.000079 time 0.2657 (0.2862) loss 2.1970 (3.1298) grad_norm 2.6944 (2.7660) [2021-04-16 16:20:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][1060/1251] eta 0:00:54 lr 0.000079 time 0.2692 (0.2862) loss 3.3985 (3.1300) grad_norm 3.2784 (2.7657) [2021-04-16 16:21:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][1070/1251] eta 0:00:51 lr 0.000079 time 0.2814 (0.2861) loss 2.5072 (3.1308) grad_norm 2.6840 (2.7653) [2021-04-16 16:21:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][1080/1251] eta 0:00:48 lr 0.000079 time 0.2902 (0.2860) loss 3.1493 (3.1287) grad_norm 2.9571 (2.7656) [2021-04-16 16:21:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][1090/1251] eta 0:00:46 lr 0.000079 time 0.2785 (0.2859) loss 3.3715 (3.1288) grad_norm 3.0401 (2.7641) [2021-04-16 16:21:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][1100/1251] eta 0:00:43 lr 0.000079 time 0.3031 (0.2859) loss 3.5862 (3.1318) grad_norm 2.6033 (2.7635) [2021-04-16 16:21:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][1110/1251] eta 0:00:40 lr 0.000079 time 0.2799 (0.2858) loss 3.3745 (3.1329) grad_norm 3.0487 (2.7627) [2021-04-16 16:21:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][1120/1251] eta 0:00:37 lr 0.000079 time 0.2843 (0.2857) loss 3.3970 (3.1321) grad_norm 2.3893 (2.7618) [2021-04-16 16:21:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][1130/1251] eta 0:00:34 lr 0.000079 time 0.2890 (0.2856) loss 3.3092 (3.1322) grad_norm 4.8890 (2.7638) [2021-04-16 16:21:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][1140/1251] eta 0:00:31 lr 0.000079 time 0.2871 (0.2855) loss 3.4878 (3.1298) grad_norm 2.6580 (2.7638) [2021-04-16 16:21:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][1150/1251] eta 0:00:28 lr 0.000079 time 0.2641 (0.2856) loss 2.4825 (3.1300) grad_norm 2.3577 (2.7632) [2021-04-16 16:21:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][1160/1251] eta 0:00:25 lr 0.000079 time 0.2803 (0.2856) loss 3.3957 (3.1288) grad_norm 2.5734 (2.7629) [2021-04-16 16:21:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][1170/1251] eta 0:00:23 lr 0.000079 time 0.2849 (0.2856) loss 3.1748 (3.1296) grad_norm 2.5238 (2.7631) [2021-04-16 16:21:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][1180/1251] eta 0:00:20 lr 0.000079 time 0.3057 (0.2855) loss 2.9874 (3.1309) grad_norm 2.5149 (2.7624) [2021-04-16 16:21:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][1190/1251] eta 0:00:17 lr 0.000079 time 0.2581 (0.2853) loss 3.0034 (3.1305) grad_norm 2.9434 (2.7633) [2021-04-16 16:21:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][1200/1251] eta 0:00:14 lr 0.000079 time 0.2703 (0.2852) loss 4.0940 (3.1308) grad_norm 2.9881 (2.7635) [2021-04-16 16:21:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][1210/1251] eta 0:00:11 lr 0.000079 time 0.2728 (0.2852) loss 3.2278 (3.1312) grad_norm 2.6179 (2.7622) [2021-04-16 16:21:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][1220/1251] eta 0:00:08 lr 0.000079 time 0.2675 (0.2851) loss 2.7511 (3.1304) grad_norm 2.7487 (2.7618) [2021-04-16 16:21:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][1230/1251] eta 0:00:05 lr 0.000079 time 0.2804 (0.2851) loss 1.9826 (3.1279) grad_norm 2.5437 (2.7615) [2021-04-16 16:21:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][1240/1251] eta 0:00:03 lr 0.000079 time 0.2479 (0.2850) loss 3.4536 (3.1304) grad_norm 2.8527 (2.7621) [2021-04-16 16:21:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [248/300][1250/1251] eta 0:00:00 lr 0.000079 time 0.2475 (0.2847) loss 3.5226 (3.1331) grad_norm 2.4017 (2.7607) [2021-04-16 16:21:59 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 248 training takes 0:06:04 [2021-04-16 16:21:59 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_248.pth saving...... [2021-04-16 16:22:20 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_248.pth saved !!! [2021-04-16 16:22:21 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.139 (1.139) Loss 0.8277 (0.8277) Acc@1 79.004 (79.004) Acc@5 95.898 (95.898) [2021-04-16 16:22:22 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.120 (0.209) Loss 0.8153 (0.8465) Acc@1 80.957 (79.643) Acc@5 94.922 (95.197) [2021-04-16 16:22:25 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 1.050 (0.232) Loss 0.8361 (0.8385) Acc@1 80.762 (79.827) Acc@5 95.410 (95.401) [2021-04-16 16:22:28 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.090 (0.252) Loss 0.8197 (0.8381) Acc@1 81.152 (80.106) Acc@5 94.336 (95.234) [2021-04-16 16:22:29 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.219) Loss 0.8635 (0.8354) Acc@1 80.469 (80.281) Acc@5 95.215 (95.243) [2021-04-16 16:22:41 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.282 Acc@5 95.246 [2021-04-16 16:22:41 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 80.3% [2021-04-16 16:22:41 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 80.28% [2021-04-16 16:22:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][0/1251] eta 2:52:06 lr 0.000079 time 8.2544 (8.2544) loss 3.3487 (3.3487) grad_norm 2.9678 (2.9678) [2021-04-16 16:22:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][10/1251] eta 0:20:57 lr 0.000079 time 0.4475 (1.0133) loss 3.0831 (3.3409) grad_norm 2.8513 (2.6682) [2021-04-16 16:22:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][20/1251] eta 0:13:35 lr 0.000079 time 0.2967 (0.6628) loss 3.4416 (3.3073) grad_norm 2.9102 (2.7254) [2021-04-16 16:22:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][30/1251] eta 0:10:56 lr 0.000079 time 0.2771 (0.5377) loss 3.5193 (3.2411) grad_norm 2.7518 (2.7202) [2021-04-16 16:23:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][40/1251] eta 0:09:41 lr 0.000079 time 0.4575 (0.4799) loss 3.1999 (3.2356) grad_norm 2.4441 (2.7305) [2021-04-16 16:23:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][50/1251] eta 0:08:48 lr 0.000079 time 0.2788 (0.4400) loss 3.8129 (3.2611) grad_norm 2.5324 (2.7472) [2021-04-16 16:23:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][60/1251] eta 0:08:12 lr 0.000079 time 0.2805 (0.4135) loss 3.2630 (3.1971) grad_norm 2.5498 (2.7284) [2021-04-16 16:23:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][70/1251] eta 0:07:46 lr 0.000079 time 0.2667 (0.3948) loss 3.1783 (3.2069) grad_norm 2.7533 (2.7148) [2021-04-16 16:23:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][80/1251] eta 0:07:25 lr 0.000079 time 0.2828 (0.3802) loss 2.4938 (3.2067) grad_norm 2.6520 (2.7412) [2021-04-16 16:23:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][90/1251] eta 0:07:09 lr 0.000079 time 0.2677 (0.3700) loss 3.1380 (3.2075) grad_norm 2.5826 (2.7571) [2021-04-16 16:23:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][100/1251] eta 0:06:55 lr 0.000079 time 0.2658 (0.3610) loss 3.4427 (3.1975) grad_norm 3.6435 (2.7534) [2021-04-16 16:23:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][110/1251] eta 0:06:43 lr 0.000079 time 0.3270 (0.3540) loss 2.0020 (3.1694) grad_norm 3.0420 (2.7733) [2021-04-16 16:23:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][120/1251] eta 0:06:35 lr 0.000079 time 0.2854 (0.3493) loss 3.5392 (3.1674) grad_norm 2.6419 (2.7822) [2021-04-16 16:23:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][130/1251] eta 0:06:27 lr 0.000079 time 0.2782 (0.3454) loss 3.4104 (3.1593) grad_norm 2.3894 (2.7682) [2021-04-16 16:23:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][140/1251] eta 0:06:20 lr 0.000079 time 0.2729 (0.3423) loss 2.7933 (3.1554) grad_norm 2.7843 (2.7757) [2021-04-16 16:23:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][150/1251] eta 0:06:12 lr 0.000079 time 0.2417 (0.3383) loss 2.0958 (3.1446) grad_norm 2.8187 (2.7794) [2021-04-16 16:23:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][160/1251] eta 0:06:04 lr 0.000079 time 0.2621 (0.3344) loss 2.7481 (3.1418) grad_norm 2.5437 (2.7757) [2021-04-16 16:23:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][170/1251] eta 0:05:58 lr 0.000079 time 0.3023 (0.3312) loss 2.4715 (3.1287) grad_norm 2.7755 (2.7621) [2021-04-16 16:23:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][180/1251] eta 0:05:51 lr 0.000079 time 0.2792 (0.3281) loss 3.4018 (3.1368) grad_norm 2.9289 (2.7577) [2021-04-16 16:23:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][190/1251] eta 0:05:45 lr 0.000079 time 0.2682 (0.3255) loss 3.6874 (3.1371) grad_norm 3.0445 (2.7697) [2021-04-16 16:23:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][200/1251] eta 0:05:39 lr 0.000079 time 0.2741 (0.3232) loss 2.5900 (3.1254) grad_norm 2.0809 (2.7671) [2021-04-16 16:23:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][210/1251] eta 0:05:34 lr 0.000078 time 0.2620 (0.3211) loss 3.7940 (3.1117) grad_norm 2.5155 (2.7612) [2021-04-16 16:23:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][220/1251] eta 0:05:29 lr 0.000078 time 0.2918 (0.3192) loss 3.5341 (3.1064) grad_norm 2.4726 (2.7570) [2021-04-16 16:23:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][230/1251] eta 0:05:24 lr 0.000078 time 0.2605 (0.3175) loss 2.6351 (3.1063) grad_norm 3.1581 (2.7572) [2021-04-16 16:23:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][240/1251] eta 0:05:19 lr 0.000078 time 0.3033 (0.3158) loss 3.4887 (3.1029) grad_norm 2.7898 (2.7539) [2021-04-16 16:24:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][250/1251] eta 0:05:14 lr 0.000078 time 0.2836 (0.3141) loss 3.4672 (3.1026) grad_norm 2.9677 (2.7466) [2021-04-16 16:24:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][260/1251] eta 0:05:09 lr 0.000078 time 0.2740 (0.3125) loss 2.7359 (3.0945) grad_norm 2.9893 (2.7434) [2021-04-16 16:24:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][270/1251] eta 0:05:05 lr 0.000078 time 0.2575 (0.3111) loss 3.0912 (3.0996) grad_norm 3.0153 (2.7565) [2021-04-16 16:24:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][280/1251] eta 0:05:00 lr 0.000078 time 0.2941 (0.3099) loss 3.0726 (3.0968) grad_norm 2.4087 (2.7569) [2021-04-16 16:24:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][290/1251] eta 0:04:56 lr 0.000078 time 0.2606 (0.3086) loss 3.6204 (3.0958) grad_norm 2.7374 (2.7549) [2021-04-16 16:24:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][300/1251] eta 0:04:52 lr 0.000078 time 0.2792 (0.3077) loss 3.1151 (3.0963) grad_norm 2.8823 (2.7555) [2021-04-16 16:24:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][310/1251] eta 0:04:48 lr 0.000078 time 0.2729 (0.3066) loss 2.7560 (3.1040) grad_norm 2.3920 (2.7583) [2021-04-16 16:24:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][320/1251] eta 0:04:44 lr 0.000078 time 0.2875 (0.3058) loss 3.5066 (3.1048) grad_norm 2.3200 (2.7538) [2021-04-16 16:24:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][330/1251] eta 0:04:40 lr 0.000078 time 0.2534 (0.3050) loss 3.5633 (3.1107) grad_norm 2.8244 (2.7553) [2021-04-16 16:24:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][340/1251] eta 0:04:37 lr 0.000078 time 0.2928 (0.3042) loss 2.7046 (3.1105) grad_norm 2.7863 (2.7556) [2021-04-16 16:24:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][350/1251] eta 0:04:33 lr 0.000078 time 0.4171 (0.3038) loss 2.8243 (3.1023) grad_norm 2.5017 (2.7552) [2021-04-16 16:24:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][360/1251] eta 0:04:29 lr 0.000078 time 0.2846 (0.3030) loss 3.6228 (3.0970) grad_norm 2.8518 (2.7549) [2021-04-16 16:24:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][370/1251] eta 0:04:27 lr 0.000078 time 0.2825 (0.3031) loss 3.6817 (3.1027) grad_norm 2.3629 (2.7566) [2021-04-16 16:24:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][380/1251] eta 0:04:23 lr 0.000078 time 0.2624 (0.3022) loss 2.2086 (3.1032) grad_norm 2.6670 (2.7565) [2021-04-16 16:24:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][390/1251] eta 0:04:19 lr 0.000078 time 0.2652 (0.3016) loss 2.1626 (3.1049) grad_norm 2.6720 (2.7571) [2021-04-16 16:24:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][400/1251] eta 0:04:16 lr 0.000078 time 0.2668 (0.3009) loss 2.5388 (3.1043) grad_norm 2.7409 (2.7535) [2021-04-16 16:24:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][410/1251] eta 0:04:12 lr 0.000078 time 0.2743 (0.3004) loss 2.0492 (3.0986) grad_norm 2.6597 (2.7535) [2021-04-16 16:24:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][420/1251] eta 0:04:09 lr 0.000078 time 0.2664 (0.2998) loss 3.7302 (3.0997) grad_norm 2.8038 (2.7526) [2021-04-16 16:24:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][430/1251] eta 0:04:05 lr 0.000078 time 0.2667 (0.2992) loss 2.4035 (3.0965) grad_norm 3.1676 (2.7512) [2021-04-16 16:24:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][440/1251] eta 0:04:02 lr 0.000078 time 0.2796 (0.2991) loss 2.7554 (3.0910) grad_norm 2.6018 (2.7502) [2021-04-16 16:24:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][450/1251] eta 0:03:59 lr 0.000078 time 0.2889 (0.2986) loss 3.2926 (3.0902) grad_norm 2.4884 (2.7458) [2021-04-16 16:24:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][460/1251] eta 0:03:55 lr 0.000078 time 0.2644 (0.2981) loss 3.5810 (3.0875) grad_norm 2.7671 (2.7484) [2021-04-16 16:25:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][470/1251] eta 0:03:52 lr 0.000078 time 0.2753 (0.2978) loss 1.8131 (3.0864) grad_norm 2.5841 (2.7497) [2021-04-16 16:25:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][480/1251] eta 0:03:49 lr 0.000078 time 0.2637 (0.2976) loss 2.7437 (3.0874) grad_norm 2.7356 (2.7482) [2021-04-16 16:25:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][490/1251] eta 0:03:46 lr 0.000078 time 0.2748 (0.2973) loss 3.5378 (3.0884) grad_norm 2.2789 (2.7535) [2021-04-16 16:25:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][500/1251] eta 0:03:42 lr 0.000078 time 0.2826 (0.2968) loss 3.0780 (3.0909) grad_norm 2.5237 (2.7570) [2021-04-16 16:25:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][510/1251] eta 0:03:39 lr 0.000078 time 0.2640 (0.2964) loss 3.2361 (3.0941) grad_norm 2.5337 (2.7562) [2021-04-16 16:25:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][520/1251] eta 0:03:36 lr 0.000078 time 0.3129 (0.2961) loss 3.3103 (3.0935) grad_norm 3.1692 (2.7590) [2021-04-16 16:25:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][530/1251] eta 0:03:33 lr 0.000078 time 0.2775 (0.2957) loss 3.0639 (3.0940) grad_norm 2.5938 (2.7593) [2021-04-16 16:25:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][540/1251] eta 0:03:30 lr 0.000078 time 0.2770 (0.2954) loss 3.2833 (3.1004) grad_norm 2.5847 (2.7568) [2021-04-16 16:25:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][550/1251] eta 0:03:26 lr 0.000078 time 0.2946 (0.2950) loss 3.3134 (3.1010) grad_norm 2.9514 (2.7535) [2021-04-16 16:25:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][560/1251] eta 0:03:23 lr 0.000078 time 0.2689 (0.2948) loss 2.7054 (3.1024) grad_norm 2.3898 (2.7534) [2021-04-16 16:25:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][570/1251] eta 0:03:20 lr 0.000078 time 0.3807 (0.2950) loss 3.4051 (3.1039) grad_norm 2.4736 (2.7556) [2021-04-16 16:25:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][580/1251] eta 0:03:17 lr 0.000078 time 0.2695 (0.2949) loss 3.2988 (3.1027) grad_norm 2.5203 (2.7562) [2021-04-16 16:25:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][590/1251] eta 0:03:14 lr 0.000078 time 0.2666 (0.2948) loss 2.8512 (3.1018) grad_norm 2.7657 (2.7609) [2021-04-16 16:25:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][600/1251] eta 0:03:11 lr 0.000078 time 0.2872 (0.2946) loss 4.0237 (3.1032) grad_norm 2.5864 (2.7625) [2021-04-16 16:25:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][610/1251] eta 0:03:08 lr 0.000078 time 0.2672 (0.2943) loss 2.0964 (3.1019) grad_norm 2.4996 (2.7604) [2021-04-16 16:25:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][620/1251] eta 0:03:05 lr 0.000078 time 0.2729 (0.2942) loss 3.4188 (3.1025) grad_norm 3.1221 (2.7601) [2021-04-16 16:25:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][630/1251] eta 0:03:02 lr 0.000078 time 0.2722 (0.2939) loss 2.7509 (3.1017) grad_norm 2.8369 (2.7624) [2021-04-16 16:25:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][640/1251] eta 0:02:59 lr 0.000078 time 0.2758 (0.2938) loss 3.3741 (3.0999) grad_norm 2.5971 (2.7631) [2021-04-16 16:25:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][650/1251] eta 0:02:56 lr 0.000078 time 0.2849 (0.2934) loss 3.0581 (3.1011) grad_norm 3.1042 (2.7613) [2021-04-16 16:25:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][660/1251] eta 0:02:53 lr 0.000078 time 0.2598 (0.2932) loss 2.7905 (3.1027) grad_norm 2.4599 (2.7594) [2021-04-16 16:25:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][670/1251] eta 0:02:50 lr 0.000078 time 0.2887 (0.2930) loss 3.5322 (3.1032) grad_norm 2.4045 (2.7568) [2021-04-16 16:26:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][680/1251] eta 0:02:47 lr 0.000078 time 0.3047 (0.2927) loss 3.4875 (3.1027) grad_norm 3.1349 (2.7610) [2021-04-16 16:26:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][690/1251] eta 0:02:44 lr 0.000077 time 0.2851 (0.2927) loss 1.8868 (3.1018) grad_norm 2.5244 (2.7616) [2021-04-16 16:26:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][700/1251] eta 0:02:41 lr 0.000077 time 0.3652 (0.2925) loss 3.3562 (3.1021) grad_norm 2.5234 (2.7613) [2021-04-16 16:26:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][710/1251] eta 0:02:38 lr 0.000077 time 0.2599 (0.2924) loss 3.2999 (3.1032) grad_norm 2.4141 (2.7625) [2021-04-16 16:26:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][720/1251] eta 0:02:35 lr 0.000077 time 0.2689 (0.2924) loss 3.5642 (3.1027) grad_norm 2.4698 (2.7608) [2021-04-16 16:26:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][730/1251] eta 0:02:32 lr 0.000077 time 0.2714 (0.2921) loss 3.0993 (3.1034) grad_norm 2.6978 (2.7584) [2021-04-16 16:26:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][740/1251] eta 0:02:29 lr 0.000077 time 0.2915 (0.2919) loss 2.3856 (3.1027) grad_norm 3.3515 (2.7588) [2021-04-16 16:26:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][750/1251] eta 0:02:26 lr 0.000077 time 0.2854 (0.2918) loss 3.1367 (3.1017) grad_norm 2.8987 (2.7592) [2021-04-16 16:26:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][760/1251] eta 0:02:23 lr 0.000077 time 0.2736 (0.2915) loss 3.2829 (3.1026) grad_norm 3.0526 (2.7595) [2021-04-16 16:26:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][770/1251] eta 0:02:20 lr 0.000077 time 0.2764 (0.2913) loss 3.2813 (3.1043) grad_norm 2.4273 (2.7588) [2021-04-16 16:26:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][780/1251] eta 0:02:17 lr 0.000077 time 0.2695 (0.2911) loss 3.3473 (3.1058) grad_norm 2.5553 (2.7572) [2021-04-16 16:26:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][790/1251] eta 0:02:14 lr 0.000077 time 0.2784 (0.2909) loss 2.2465 (3.1061) grad_norm 2.5296 (2.7568) [2021-04-16 16:26:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][800/1251] eta 0:02:11 lr 0.000077 time 0.2879 (0.2908) loss 2.3185 (3.1050) grad_norm 3.0439 (2.7580) [2021-04-16 16:26:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][810/1251] eta 0:02:08 lr 0.000077 time 0.2672 (0.2906) loss 3.1159 (3.1062) grad_norm 2.4648 (2.7567) [2021-04-16 16:26:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][820/1251] eta 0:02:05 lr 0.000077 time 0.2727 (0.2905) loss 3.3248 (3.1040) grad_norm 2.2645 (2.7570) [2021-04-16 16:26:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][830/1251] eta 0:02:02 lr 0.000077 time 0.2714 (0.2903) loss 3.6843 (3.1049) grad_norm 2.7347 (2.7613) [2021-04-16 16:26:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][840/1251] eta 0:01:59 lr 0.000077 time 0.2733 (0.2903) loss 2.3772 (3.1043) grad_norm 3.4523 (2.7669) [2021-04-16 16:26:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][850/1251] eta 0:01:56 lr 0.000077 time 0.2600 (0.2901) loss 3.0467 (3.1023) grad_norm 2.8006 (2.7665) [2021-04-16 16:26:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][860/1251] eta 0:01:53 lr 0.000077 time 0.2917 (0.2899) loss 2.5045 (3.1022) grad_norm 2.5510 (2.7694) [2021-04-16 16:26:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][870/1251] eta 0:01:50 lr 0.000077 time 0.2611 (0.2899) loss 3.2447 (3.1014) grad_norm 2.9521 (2.7711) [2021-04-16 16:26:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][880/1251] eta 0:01:47 lr 0.000077 time 0.2746 (0.2898) loss 3.0344 (3.1006) grad_norm 2.5383 (2.7699) [2021-04-16 16:26:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][890/1251] eta 0:01:44 lr 0.000077 time 0.2869 (0.2897) loss 3.0631 (3.1026) grad_norm 2.4583 (2.7696) [2021-04-16 16:27:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][900/1251] eta 0:01:41 lr 0.000077 time 0.3024 (0.2896) loss 3.1823 (3.1043) grad_norm 2.4335 (2.7678) [2021-04-16 16:27:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][910/1251] eta 0:01:38 lr 0.000077 time 0.2648 (0.2894) loss 3.3452 (3.1037) grad_norm 2.2786 (2.7648) [2021-04-16 16:27:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][920/1251] eta 0:01:35 lr 0.000077 time 0.2795 (0.2893) loss 3.6828 (3.1033) grad_norm 2.8147 (2.7637) [2021-04-16 16:27:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][930/1251] eta 0:01:32 lr 0.000077 time 0.2766 (0.2894) loss 2.4027 (3.1031) grad_norm 2.3870 (2.7629) [2021-04-16 16:27:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][940/1251] eta 0:01:30 lr 0.000077 time 0.3802 (0.2895) loss 3.1701 (3.1039) grad_norm 2.7086 (2.7639) [2021-04-16 16:27:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][950/1251] eta 0:01:27 lr 0.000077 time 0.2719 (0.2895) loss 2.3737 (3.1020) grad_norm 2.9929 (2.7637) [2021-04-16 16:27:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][960/1251] eta 0:01:24 lr 0.000077 time 0.2911 (0.2893) loss 3.2181 (3.1012) grad_norm 2.7993 (2.7639) [2021-04-16 16:27:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][970/1251] eta 0:01:21 lr 0.000077 time 0.3033 (0.2893) loss 3.6481 (3.1017) grad_norm 2.4910 (2.7672) [2021-04-16 16:27:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][980/1251] eta 0:01:18 lr 0.000077 time 0.2856 (0.2892) loss 3.3862 (3.1027) grad_norm 2.8748 (2.7672) [2021-04-16 16:27:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][990/1251] eta 0:01:15 lr 0.000077 time 0.2811 (0.2890) loss 3.1416 (3.1018) grad_norm 2.6869 (2.7668) [2021-04-16 16:27:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][1000/1251] eta 0:01:12 lr 0.000077 time 0.2771 (0.2889) loss 3.0982 (3.1012) grad_norm 2.5605 (2.7672) [2021-04-16 16:27:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][1010/1251] eta 0:01:09 lr 0.000077 time 0.2845 (0.2888) loss 2.9012 (3.1000) grad_norm 2.2578 (2.7666) [2021-04-16 16:27:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][1020/1251] eta 0:01:06 lr 0.000077 time 0.2846 (0.2888) loss 3.2849 (3.0990) grad_norm 2.8366 (2.7664) [2021-04-16 16:27:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][1030/1251] eta 0:01:03 lr 0.000077 time 0.2794 (0.2887) loss 3.1225 (3.1001) grad_norm 2.9383 (2.7659) [2021-04-16 16:27:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][1040/1251] eta 0:01:00 lr 0.000077 time 0.2888 (0.2886) loss 2.2579 (3.0979) grad_norm 2.5562 (2.7649) [2021-04-16 16:27:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][1050/1251] eta 0:00:58 lr 0.000077 time 0.4368 (0.2886) loss 3.0875 (3.1004) grad_norm 2.7755 (2.7659) [2021-04-16 16:27:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][1060/1251] eta 0:00:55 lr 0.000077 time 0.2760 (0.2885) loss 3.1919 (3.1009) grad_norm 2.4807 (2.7650) [2021-04-16 16:27:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][1070/1251] eta 0:00:52 lr 0.000077 time 0.2751 (0.2885) loss 3.5055 (3.1023) grad_norm 2.6843 (2.7644) [2021-04-16 16:27:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][1080/1251] eta 0:00:49 lr 0.000077 time 0.2697 (0.2884) loss 2.3465 (3.1006) grad_norm 3.0480 (2.7667) [2021-04-16 16:27:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][1090/1251] eta 0:00:46 lr 0.000077 time 0.2925 (0.2883) loss 3.1845 (3.1019) grad_norm 2.6272 (2.7682) [2021-04-16 16:27:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][1100/1251] eta 0:00:43 lr 0.000077 time 0.2874 (0.2883) loss 3.7023 (3.1014) grad_norm 3.3714 (2.7696) [2021-04-16 16:28:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][1110/1251] eta 0:00:40 lr 0.000077 time 0.2855 (0.2883) loss 3.5806 (3.1019) grad_norm 2.6732 (2.7702) [2021-04-16 16:28:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][1120/1251] eta 0:00:37 lr 0.000077 time 0.2877 (0.2882) loss 3.1845 (3.1019) grad_norm 2.4078 (2.7730) [2021-04-16 16:28:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][1130/1251] eta 0:00:34 lr 0.000077 time 0.2647 (0.2880) loss 2.9109 (3.0985) grad_norm 3.8289 (2.7741) [2021-04-16 16:28:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][1140/1251] eta 0:00:31 lr 0.000077 time 0.2541 (0.2880) loss 3.4933 (3.0986) grad_norm 2.9259 (2.7747) [2021-04-16 16:28:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][1150/1251] eta 0:00:29 lr 0.000077 time 0.2618 (0.2880) loss 3.5810 (3.0980) grad_norm 2.4997 (2.7767) [2021-04-16 16:28:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][1160/1251] eta 0:00:26 lr 0.000077 time 0.2964 (0.2881) loss 4.0200 (3.0985) grad_norm 2.3741 (2.7767) [2021-04-16 16:28:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][1170/1251] eta 0:00:23 lr 0.000076 time 0.2518 (0.2881) loss 2.4606 (3.0979) grad_norm 2.5794 (2.7772) [2021-04-16 16:28:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][1180/1251] eta 0:00:20 lr 0.000076 time 0.2634 (0.2879) loss 2.6377 (3.0964) grad_norm 2.9486 (2.7783) [2021-04-16 16:28:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][1190/1251] eta 0:00:17 lr 0.000076 time 0.2646 (0.2879) loss 2.5512 (3.0958) grad_norm 2.5435 (2.7773) [2021-04-16 16:28:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][1200/1251] eta 0:00:14 lr 0.000076 time 0.2806 (0.2879) loss 3.1519 (3.0970) grad_norm 2.6943 (2.7775) [2021-04-16 16:28:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][1210/1251] eta 0:00:11 lr 0.000076 time 0.2888 (0.2877) loss 2.6926 (3.0956) grad_norm 3.0621 (2.7786) [2021-04-16 16:28:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][1220/1251] eta 0:00:08 lr 0.000076 time 0.2584 (0.2877) loss 3.1252 (3.0931) grad_norm 2.8920 (2.7795) [2021-04-16 16:28:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][1230/1251] eta 0:00:06 lr 0.000076 time 0.3010 (0.2876) loss 2.8521 (3.0932) grad_norm 2.7270 (2.7790) [2021-04-16 16:28:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][1240/1251] eta 0:00:03 lr 0.000076 time 0.2481 (0.2874) loss 3.4699 (3.0941) grad_norm 2.4922 (2.7769) [2021-04-16 16:28:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [249/300][1250/1251] eta 0:00:00 lr 0.000076 time 0.2515 (0.2871) loss 3.4506 (3.0954) grad_norm 2.6910 (2.7755) [2021-04-16 16:28:44 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 249 training takes 0:06:03 [2021-04-16 16:28:44 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_249.pth saving...... [2021-04-16 16:28:58 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_249.pth saved !!! [2021-04-16 16:28:59 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.229 (1.229) Loss 0.8979 (0.8979) Acc@1 79.492 (79.492) Acc@5 94.531 (94.531) [2021-04-16 16:29:00 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.099 (0.206) Loss 0.8214 (0.8415) Acc@1 80.566 (80.202) Acc@5 95.703 (95.037) [2021-04-16 16:29:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.178 (0.194) Loss 0.7642 (0.8385) Acc@1 81.055 (80.227) Acc@5 95.996 (95.173) [2021-04-16 16:29:06 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.530 (0.250) Loss 0.8770 (0.8399) Acc@1 79.492 (80.261) Acc@5 94.922 (95.155) [2021-04-16 16:29:07 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.075 (0.225) Loss 0.7904 (0.8373) Acc@1 81.543 (80.238) Acc@5 95.508 (95.203) [2021-04-16 16:29:12 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.388 Acc@5 95.202 [2021-04-16 16:29:12 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 80.4% [2021-04-16 16:29:12 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 80.39% [2021-04-16 16:29:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][0/1251] eta 4:50:54 lr 0.000076 time 13.9522 (13.9522) loss 2.6237 (2.6237) grad_norm 3.2756 (3.2756) [2021-04-16 16:29:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][10/1251] eta 0:31:26 lr 0.000076 time 0.2800 (1.5198) loss 3.4088 (3.0424) grad_norm 2.6936 (2.8647) [2021-04-16 16:29:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][20/1251] eta 0:18:59 lr 0.000076 time 0.2721 (0.9257) loss 3.3148 (3.0547) grad_norm 2.3342 (2.8049) [2021-04-16 16:29:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][30/1251] eta 0:14:34 lr 0.000076 time 0.2882 (0.7161) loss 2.9123 (3.0555) grad_norm 3.2631 (2.8192) [2021-04-16 16:29:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][40/1251] eta 0:12:21 lr 0.000076 time 0.2955 (0.6120) loss 3.2942 (3.0831) grad_norm 2.8543 (2.8256) [2021-04-16 16:29:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][50/1251] eta 0:10:54 lr 0.000076 time 0.2612 (0.5453) loss 3.5412 (3.0985) grad_norm 3.0120 (2.7994) [2021-04-16 16:29:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][60/1251] eta 0:09:57 lr 0.000076 time 0.2897 (0.5019) loss 2.6376 (3.0403) grad_norm 3.0031 (2.8086) [2021-04-16 16:29:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][70/1251] eta 0:09:15 lr 0.000076 time 0.2845 (0.4707) loss 3.4620 (3.0631) grad_norm 2.6591 (2.8306) [2021-04-16 16:29:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][80/1251] eta 0:08:43 lr 0.000076 time 0.2906 (0.4469) loss 3.4959 (3.0526) grad_norm 2.9022 (2.8325) [2021-04-16 16:29:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][90/1251] eta 0:08:17 lr 0.000076 time 0.2677 (0.4288) loss 1.6342 (3.0444) grad_norm 2.6466 (2.9030) [2021-04-16 16:29:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][100/1251] eta 0:07:56 lr 0.000076 time 0.2578 (0.4138) loss 3.2870 (3.0244) grad_norm 2.3419 (2.9151) [2021-04-16 16:29:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][110/1251] eta 0:07:38 lr 0.000076 time 0.2948 (0.4015) loss 3.3174 (3.0112) grad_norm 3.4177 (2.9039) [2021-04-16 16:30:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][120/1251] eta 0:07:22 lr 0.000076 time 0.2821 (0.3911) loss 3.2853 (3.0204) grad_norm 2.8318 (2.9009) [2021-04-16 16:30:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][130/1251] eta 0:07:09 lr 0.000076 time 0.2694 (0.3834) loss 3.2505 (3.0282) grad_norm 2.8491 (2.9102) [2021-04-16 16:30:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][140/1251] eta 0:06:58 lr 0.000076 time 0.2942 (0.3764) loss 2.0399 (3.0395) grad_norm 2.8908 (2.9028) [2021-04-16 16:30:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][150/1251] eta 0:06:47 lr 0.000076 time 0.3987 (0.3705) loss 3.5963 (3.0398) grad_norm 2.8865 (2.8834) [2021-04-16 16:30:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][160/1251] eta 0:06:39 lr 0.000076 time 0.3121 (0.3663) loss 3.4465 (3.0338) grad_norm 3.4470 (2.8763) [2021-04-16 16:30:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][170/1251] eta 0:06:30 lr 0.000076 time 0.2963 (0.3615) loss 3.8081 (3.0470) grad_norm 2.5778 (2.8651) [2021-04-16 16:30:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][180/1251] eta 0:06:22 lr 0.000076 time 0.2923 (0.3572) loss 3.3273 (3.0625) grad_norm 2.8430 (2.8489) [2021-04-16 16:30:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][190/1251] eta 0:06:14 lr 0.000076 time 0.2700 (0.3532) loss 3.3747 (3.0661) grad_norm 2.6292 (2.8400) [2021-04-16 16:30:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][200/1251] eta 0:06:07 lr 0.000076 time 0.2912 (0.3497) loss 2.7246 (3.0726) grad_norm 3.3705 (2.8488) [2021-04-16 16:30:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][210/1251] eta 0:06:00 lr 0.000076 time 0.2954 (0.3462) loss 3.4440 (3.0745) grad_norm 2.8054 (2.8485) [2021-04-16 16:30:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][220/1251] eta 0:05:53 lr 0.000076 time 0.2821 (0.3433) loss 3.3436 (3.0706) grad_norm 2.2764 (2.8517) [2021-04-16 16:30:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][230/1251] eta 0:05:48 lr 0.000076 time 0.2908 (0.3411) loss 3.5455 (3.0588) grad_norm 2.2995 (2.8433) [2021-04-16 16:30:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][240/1251] eta 0:05:42 lr 0.000076 time 0.2613 (0.3383) loss 3.4576 (3.0663) grad_norm 2.8582 (2.8307) [2021-04-16 16:30:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][250/1251] eta 0:05:36 lr 0.000076 time 0.2727 (0.3359) loss 3.2501 (3.0648) grad_norm 2.5504 (2.8252) [2021-04-16 16:30:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][260/1251] eta 0:05:31 lr 0.000076 time 0.3974 (0.3343) loss 3.8922 (3.0722) grad_norm 3.3765 (2.8274) [2021-04-16 16:30:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][270/1251] eta 0:05:25 lr 0.000076 time 0.2820 (0.3321) loss 2.8848 (3.0799) grad_norm 2.3837 (2.8189) [2021-04-16 16:30:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][280/1251] eta 0:05:20 lr 0.000076 time 0.2830 (0.3302) loss 2.1589 (3.0799) grad_norm 2.4395 (2.8123) [2021-04-16 16:30:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][290/1251] eta 0:05:15 lr 0.000076 time 0.2775 (0.3281) loss 3.4728 (3.0781) grad_norm 2.9140 (2.8084) [2021-04-16 16:30:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][300/1251] eta 0:05:10 lr 0.000076 time 0.3072 (0.3265) loss 3.4244 (3.0850) grad_norm 3.0839 (2.8059) [2021-04-16 16:30:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][310/1251] eta 0:05:05 lr 0.000076 time 0.2648 (0.3248) loss 2.3053 (3.0864) grad_norm 3.0651 (2.8064) [2021-04-16 16:30:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][320/1251] eta 0:05:01 lr 0.000076 time 0.2861 (0.3233) loss 3.6646 (3.0882) grad_norm 3.1373 (2.8096) [2021-04-16 16:30:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][330/1251] eta 0:04:56 lr 0.000076 time 0.2609 (0.3219) loss 3.4808 (3.0888) grad_norm 2.6766 (2.8113) [2021-04-16 16:31:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][340/1251] eta 0:04:52 lr 0.000076 time 0.2832 (0.3208) loss 3.0312 (3.0893) grad_norm 2.5369 (2.8101) [2021-04-16 16:31:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][350/1251] eta 0:04:48 lr 0.000076 time 0.2754 (0.3201) loss 2.0807 (3.0789) grad_norm 2.8740 (2.8060) [2021-04-16 16:31:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][360/1251] eta 0:04:44 lr 0.000076 time 0.2825 (0.3191) loss 3.0296 (3.0777) grad_norm 2.5576 (2.8031) [2021-04-16 16:31:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][370/1251] eta 0:04:40 lr 0.000076 time 0.2637 (0.3188) loss 3.3205 (3.0804) grad_norm 2.5896 (2.8064) [2021-04-16 16:31:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][380/1251] eta 0:04:36 lr 0.000076 time 0.2940 (0.3176) loss 3.7029 (3.0814) grad_norm 2.6794 (2.8045) [2021-04-16 16:31:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][390/1251] eta 0:04:32 lr 0.000076 time 0.2760 (0.3165) loss 1.8091 (3.0804) grad_norm 2.5963 (2.8025) [2021-04-16 16:31:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][400/1251] eta 0:04:28 lr 0.000075 time 0.2931 (0.3155) loss 2.7967 (3.0750) grad_norm 2.8036 (2.8041) [2021-04-16 16:31:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][410/1251] eta 0:04:24 lr 0.000075 time 0.3118 (0.3148) loss 2.6838 (3.0734) grad_norm 2.5245 (2.8013) [2021-04-16 16:31:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][420/1251] eta 0:04:20 lr 0.000075 time 0.2757 (0.3138) loss 3.3778 (3.0716) grad_norm 2.8291 (2.7977) [2021-04-16 16:31:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][430/1251] eta 0:04:17 lr 0.000075 time 0.3311 (0.3132) loss 3.3987 (3.0742) grad_norm 2.3807 (2.7996) [2021-04-16 16:31:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][440/1251] eta 0:04:13 lr 0.000075 time 0.2483 (0.3123) loss 3.6572 (3.0747) grad_norm 2.7260 (2.8006) [2021-04-16 16:31:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][450/1251] eta 0:04:09 lr 0.000075 time 0.2763 (0.3116) loss 3.6947 (3.0738) grad_norm 2.9705 (2.8031) [2021-04-16 16:31:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][460/1251] eta 0:04:05 lr 0.000075 time 0.3005 (0.3109) loss 3.0413 (3.0718) grad_norm 3.5244 (2.8023) [2021-04-16 16:31:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][470/1251] eta 0:04:02 lr 0.000075 time 0.2963 (0.3101) loss 3.3914 (3.0711) grad_norm 2.5574 (2.8050) [2021-04-16 16:31:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][480/1251] eta 0:03:58 lr 0.000075 time 0.2623 (0.3095) loss 2.6616 (3.0676) grad_norm 2.9419 (2.8079) [2021-04-16 16:31:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][490/1251] eta 0:03:55 lr 0.000075 time 0.2765 (0.3089) loss 2.1877 (3.0704) grad_norm 2.5788 (2.8066) [2021-04-16 16:31:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][500/1251] eta 0:03:51 lr 0.000075 time 0.2478 (0.3082) loss 2.4066 (3.0683) grad_norm 2.6872 (2.8069) [2021-04-16 16:31:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][510/1251] eta 0:03:47 lr 0.000075 time 0.2691 (0.3076) loss 2.1365 (3.0686) grad_norm 3.0363 (2.8053) [2021-04-16 16:31:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][520/1251] eta 0:03:44 lr 0.000075 time 0.2814 (0.3073) loss 1.9807 (3.0669) grad_norm 2.5404 (2.8068) [2021-04-16 16:31:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][530/1251] eta 0:03:41 lr 0.000075 time 0.2834 (0.3067) loss 2.8912 (3.0714) grad_norm 2.4911 (2.8044) [2021-04-16 16:31:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][540/1251] eta 0:03:37 lr 0.000075 time 0.2914 (0.3064) loss 3.2853 (3.0724) grad_norm 2.9836 (2.8011) [2021-04-16 16:32:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][550/1251] eta 0:03:34 lr 0.000075 time 0.2697 (0.3059) loss 3.0440 (3.0767) grad_norm 2.4552 (2.7975) [2021-04-16 16:32:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][560/1251] eta 0:03:30 lr 0.000075 time 0.2841 (0.3053) loss 3.5943 (3.0754) grad_norm 2.5735 (2.7934) [2021-04-16 16:32:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][570/1251] eta 0:03:27 lr 0.000075 time 0.3643 (0.3050) loss 3.5055 (3.0754) grad_norm 2.6598 (2.7912) [2021-04-16 16:32:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][580/1251] eta 0:03:24 lr 0.000075 time 0.2648 (0.3048) loss 3.0405 (3.0735) grad_norm 2.7928 (2.7937) [2021-04-16 16:32:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][590/1251] eta 0:03:21 lr 0.000075 time 0.2917 (0.3042) loss 3.1563 (3.0739) grad_norm 2.5483 (2.7944) [2021-04-16 16:32:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][600/1251] eta 0:03:17 lr 0.000075 time 0.2817 (0.3037) loss 3.1569 (3.0773) grad_norm 2.7527 (2.7944) [2021-04-16 16:32:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][610/1251] eta 0:03:14 lr 0.000075 time 0.2700 (0.3034) loss 2.5140 (3.0762) grad_norm 2.4081 (2.7941) [2021-04-16 16:32:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][620/1251] eta 0:03:11 lr 0.000075 time 0.2817 (0.3029) loss 2.9348 (3.0814) grad_norm 2.8008 (2.7913) [2021-04-16 16:32:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][630/1251] eta 0:03:07 lr 0.000075 time 0.2840 (0.3026) loss 2.7452 (3.0819) grad_norm 2.9118 (2.7941) [2021-04-16 16:32:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][640/1251] eta 0:03:04 lr 0.000075 time 0.3080 (0.3022) loss 3.7804 (3.0837) grad_norm 2.5144 (2.7944) [2021-04-16 16:32:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][650/1251] eta 0:03:01 lr 0.000075 time 0.2828 (0.3018) loss 3.5855 (3.0833) grad_norm 2.5573 (2.7921) [2021-04-16 16:32:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][660/1251] eta 0:02:58 lr 0.000075 time 0.2945 (0.3014) loss 2.6164 (3.0835) grad_norm 2.8376 (2.7910) [2021-04-16 16:32:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][670/1251] eta 0:02:54 lr 0.000075 time 0.2818 (0.3012) loss 1.6415 (3.0818) grad_norm 3.0914 (2.7914) [2021-04-16 16:32:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][680/1251] eta 0:02:51 lr 0.000075 time 0.2836 (0.3008) loss 2.7558 (3.0804) grad_norm 2.7778 (2.7950) [2021-04-16 16:32:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][690/1251] eta 0:02:48 lr 0.000075 time 0.2532 (0.3004) loss 3.1157 (3.0782) grad_norm 2.3263 (2.7958) [2021-04-16 16:32:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][700/1251] eta 0:02:45 lr 0.000075 time 0.2665 (0.3001) loss 3.5803 (3.0794) grad_norm 2.5738 (2.7946) [2021-04-16 16:32:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][710/1251] eta 0:02:42 lr 0.000075 time 0.2779 (0.3000) loss 3.3438 (3.0815) grad_norm 2.5316 (2.7953) [2021-04-16 16:32:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][720/1251] eta 0:02:39 lr 0.000075 time 0.2999 (0.2998) loss 3.7210 (3.0821) grad_norm 3.1184 (2.7954) [2021-04-16 16:32:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][730/1251] eta 0:02:36 lr 0.000075 time 0.2650 (0.2997) loss 1.9675 (3.0786) grad_norm 2.4727 (2.7956) [2021-04-16 16:32:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][740/1251] eta 0:02:33 lr 0.000075 time 0.2896 (0.2995) loss 3.3124 (3.0798) grad_norm 2.5617 (2.7945) [2021-04-16 16:32:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][750/1251] eta 0:02:29 lr 0.000075 time 0.2880 (0.2992) loss 2.2806 (3.0810) grad_norm 2.8174 (2.7931) [2021-04-16 16:33:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][760/1251] eta 0:02:26 lr 0.000075 time 0.2608 (0.2988) loss 2.0559 (3.0826) grad_norm 2.7300 (2.7936) [2021-04-16 16:33:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][770/1251] eta 0:02:23 lr 0.000075 time 0.2784 (0.2985) loss 3.2665 (3.0830) grad_norm 3.4709 (2.7939) [2021-04-16 16:33:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][780/1251] eta 0:02:20 lr 0.000075 time 0.2773 (0.2983) loss 2.8407 (3.0815) grad_norm 2.5075 (2.7942) [2021-04-16 16:33:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][790/1251] eta 0:02:17 lr 0.000075 time 0.2818 (0.2981) loss 3.3490 (3.0841) grad_norm 3.7018 (2.7974) [2021-04-16 16:33:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][800/1251] eta 0:02:14 lr 0.000075 time 0.4901 (0.2981) loss 3.3451 (3.0841) grad_norm 2.9790 (2.7975) [2021-04-16 16:33:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][810/1251] eta 0:02:11 lr 0.000075 time 0.3193 (0.2978) loss 2.2426 (3.0832) grad_norm 2.5873 (2.7981) [2021-04-16 16:33:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][820/1251] eta 0:02:08 lr 0.000075 time 0.2644 (0.2976) loss 3.7689 (3.0840) grad_norm 2.4853 (2.8002) [2021-04-16 16:33:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][830/1251] eta 0:02:05 lr 0.000075 time 0.2899 (0.2973) loss 3.4593 (3.0855) grad_norm 2.4194 (2.7981) [2021-04-16 16:33:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][840/1251] eta 0:02:02 lr 0.000075 time 0.2526 (0.2971) loss 3.6302 (3.0891) grad_norm 2.5582 (2.7990) [2021-04-16 16:33:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][850/1251] eta 0:01:59 lr 0.000075 time 0.2910 (0.2970) loss 3.3598 (3.0912) grad_norm 2.4192 (2.7981) [2021-04-16 16:33:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][860/1251] eta 0:01:56 lr 0.000075 time 0.3145 (0.2968) loss 3.4621 (3.0926) grad_norm 3.0812 (2.7984) [2021-04-16 16:33:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][870/1251] eta 0:01:52 lr 0.000075 time 0.2783 (0.2966) loss 3.2859 (3.0953) grad_norm 2.2824 (2.7973) [2021-04-16 16:33:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][880/1251] eta 0:01:49 lr 0.000075 time 0.2942 (0.2964) loss 3.2953 (3.0936) grad_norm 2.7220 (2.7981) [2021-04-16 16:33:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][890/1251] eta 0:01:46 lr 0.000074 time 0.2841 (0.2963) loss 2.0654 (3.0906) grad_norm 2.9367 (2.7992) [2021-04-16 16:33:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][900/1251] eta 0:01:43 lr 0.000074 time 0.2684 (0.2961) loss 3.8586 (3.0937) grad_norm 2.9773 (2.7986) [2021-04-16 16:33:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][910/1251] eta 0:01:40 lr 0.000074 time 0.2935 (0.2959) loss 2.9166 (3.0920) grad_norm 2.6557 (2.7973) [2021-04-16 16:33:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][920/1251] eta 0:01:37 lr 0.000074 time 0.2575 (0.2957) loss 3.3373 (3.0912) grad_norm 2.8105 (2.7958) [2021-04-16 16:33:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][930/1251] eta 0:01:34 lr 0.000074 time 0.2739 (0.2958) loss 3.5263 (3.0893) grad_norm 2.5684 (2.7960) [2021-04-16 16:33:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][940/1251] eta 0:01:31 lr 0.000074 time 0.2770 (0.2957) loss 3.3123 (3.0891) grad_norm 2.8713 (2.7963) [2021-04-16 16:33:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][950/1251] eta 0:01:28 lr 0.000074 time 0.2763 (0.2956) loss 2.7249 (3.0882) grad_norm 2.7476 (2.8023) [2021-04-16 16:33:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][960/1251] eta 0:01:25 lr 0.000074 time 0.2648 (0.2954) loss 3.1148 (3.0863) grad_norm 4.1531 (2.8028) [2021-04-16 16:33:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][970/1251] eta 0:01:22 lr 0.000074 time 0.2889 (0.2952) loss 3.4353 (3.0889) grad_norm 2.7175 (2.8037) [2021-04-16 16:34:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][980/1251] eta 0:01:19 lr 0.000074 time 0.2866 (0.2951) loss 4.1160 (3.0898) grad_norm 2.6261 (2.8022) [2021-04-16 16:34:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][990/1251] eta 0:01:16 lr 0.000074 time 0.2733 (0.2949) loss 2.9526 (3.0907) grad_norm 3.1862 (2.8038) [2021-04-16 16:34:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][1000/1251] eta 0:01:13 lr 0.000074 time 0.3080 (0.2947) loss 3.0397 (3.0897) grad_norm 2.9436 (2.8033) [2021-04-16 16:34:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][1010/1251] eta 0:01:10 lr 0.000074 time 0.2758 (0.2945) loss 3.6221 (3.0926) grad_norm 2.4941 (2.8023) [2021-04-16 16:34:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][1020/1251] eta 0:01:07 lr 0.000074 time 0.2777 (0.2943) loss 3.8332 (3.0923) grad_norm 2.9306 (2.8021) [2021-04-16 16:34:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][1030/1251] eta 0:01:05 lr 0.000074 time 0.3016 (0.2941) loss 2.7915 (3.0922) grad_norm 2.2533 (2.8014) [2021-04-16 16:34:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][1040/1251] eta 0:01:02 lr 0.000074 time 0.3121 (0.2941) loss 3.4810 (3.0951) grad_norm 3.0038 (2.8003) [2021-04-16 16:34:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][1050/1251] eta 0:00:59 lr 0.000074 time 0.2719 (0.2939) loss 3.6203 (3.0966) grad_norm 3.2134 (2.8026) [2021-04-16 16:34:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][1060/1251] eta 0:00:56 lr 0.000074 time 0.2525 (0.2938) loss 3.7483 (3.0976) grad_norm 2.8025 (2.8030) [2021-04-16 16:34:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][1070/1251] eta 0:00:53 lr 0.000074 time 0.2654 (0.2936) loss 3.6128 (3.0969) grad_norm 3.1784 (2.8030) [2021-04-16 16:34:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][1080/1251] eta 0:00:50 lr 0.000074 time 0.2906 (0.2935) loss 2.9806 (3.0970) grad_norm 2.5129 (2.8027) [2021-04-16 16:34:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][1090/1251] eta 0:00:47 lr 0.000074 time 0.2691 (0.2934) loss 3.7662 (3.0955) grad_norm 2.8956 (2.8017) [2021-04-16 16:34:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][1100/1251] eta 0:00:44 lr 0.000074 time 0.2612 (0.2933) loss 2.8413 (3.0953) grad_norm 3.2896 (2.8013) [2021-04-16 16:34:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][1110/1251] eta 0:00:41 lr 0.000074 time 0.2830 (0.2931) loss 3.6029 (3.0960) grad_norm 2.9423 (2.8014) [2021-04-16 16:34:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][1120/1251] eta 0:00:38 lr 0.000074 time 0.2643 (0.2931) loss 3.1645 (3.0962) grad_norm 2.6194 (2.8008) [2021-04-16 16:34:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][1130/1251] eta 0:00:35 lr 0.000074 time 0.2808 (0.2931) loss 2.4975 (3.0951) grad_norm 3.1909 (2.7994) [2021-04-16 16:34:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][1140/1251] eta 0:00:32 lr 0.000074 time 0.2835 (0.2930) loss 3.0305 (3.0944) grad_norm 2.8743 (2.7996) [2021-04-16 16:34:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][1150/1251] eta 0:00:29 lr 0.000074 time 0.2655 (0.2929) loss 3.1365 (3.0933) grad_norm 2.3034 (2.7995) [2021-04-16 16:34:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][1160/1251] eta 0:00:26 lr 0.000074 time 0.2456 (0.2928) loss 2.9866 (3.0932) grad_norm 2.5775 (2.7986) [2021-04-16 16:34:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][1170/1251] eta 0:00:23 lr 0.000074 time 0.2845 (0.2927) loss 3.3783 (3.0939) grad_norm 2.8913 (inf) [2021-04-16 16:34:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][1180/1251] eta 0:00:20 lr 0.000074 time 0.2722 (0.2926) loss 3.3895 (3.0946) grad_norm 2.4274 (inf) [2021-04-16 16:35:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][1190/1251] eta 0:00:17 lr 0.000074 time 0.3076 (0.2925) loss 2.4464 (3.0948) grad_norm 2.6998 (inf) [2021-04-16 16:35:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][1200/1251] eta 0:00:14 lr 0.000074 time 0.2668 (0.2924) loss 2.3118 (3.0934) grad_norm 3.0108 (inf) [2021-04-16 16:35:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][1210/1251] eta 0:00:11 lr 0.000074 time 0.2571 (0.2923) loss 2.2323 (3.0895) grad_norm 2.5704 (inf) [2021-04-16 16:35:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][1220/1251] eta 0:00:09 lr 0.000074 time 0.2790 (0.2922) loss 3.5182 (3.0891) grad_norm 2.6118 (inf) [2021-04-16 16:35:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][1230/1251] eta 0:00:06 lr 0.000074 time 0.2727 (0.2921) loss 2.4109 (3.0867) grad_norm 2.9434 (inf) [2021-04-16 16:35:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][1240/1251] eta 0:00:03 lr 0.000074 time 0.2481 (0.2919) loss 3.1568 (3.0851) grad_norm 2.5901 (inf) [2021-04-16 16:35:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [250/300][1250/1251] eta 0:00:00 lr 0.000074 time 0.2486 (0.2915) loss 3.0322 (3.0857) grad_norm 2.9218 (inf) [2021-04-16 16:35:22 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 250 training takes 0:06:09 [2021-04-16 16:35:22 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_250.pth saving...... [2021-04-16 16:35:35 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_250.pth saved !!! [2021-04-16 16:35:37 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.173 (1.173) Loss 0.8442 (0.8442) Acc@1 81.641 (81.641) Acc@5 95.020 (95.020) [2021-04-16 16:35:38 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.115 (0.227) Loss 0.8990 (0.8307) Acc@1 78.320 (80.167) Acc@5 94.336 (95.401) [2021-04-16 16:35:41 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.090 (0.244) Loss 0.7878 (0.8209) Acc@1 80.664 (80.487) Acc@5 95.801 (95.485) [2021-04-16 16:35:42 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.291 (0.221) Loss 0.8804 (0.8281) Acc@1 79.883 (80.425) Acc@5 94.238 (95.278) [2021-04-16 16:35:44 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.076 (0.209) Loss 0.8364 (0.8332) Acc@1 82.129 (80.383) Acc@5 95.215 (95.282) [2021-04-16 16:35:53 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.464 Acc@5 95.264 [2021-04-16 16:35:53 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 80.5% [2021-04-16 16:35:53 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 80.46% [2021-04-16 16:35:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][0/1251] eta 1:43:48 lr 0.000074 time 4.9791 (4.9791) loss 3.6659 (3.6659) grad_norm 2.8287 (2.8287) [2021-04-16 16:36:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][10/1251] eta 0:14:47 lr 0.000074 time 0.4570 (0.7151) loss 3.7942 (3.2903) grad_norm 3.1304 (2.7501) [2021-04-16 16:36:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][20/1251] eta 0:10:20 lr 0.000074 time 0.2898 (0.5044) loss 2.4032 (3.1103) grad_norm 2.4363 (2.7802) [2021-04-16 16:36:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][30/1251] eta 0:08:47 lr 0.000074 time 0.2705 (0.4316) loss 2.9014 (3.1312) grad_norm 2.2661 (2.7351) [2021-04-16 16:36:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][40/1251] eta 0:08:00 lr 0.000074 time 0.2760 (0.3972) loss 3.3512 (3.1361) grad_norm 2.5661 (2.7604) [2021-04-16 16:36:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][50/1251] eta 0:07:31 lr 0.000074 time 0.2739 (0.3761) loss 3.6644 (3.1187) grad_norm 2.3212 (2.7373) [2021-04-16 16:36:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][60/1251] eta 0:07:08 lr 0.000074 time 0.2525 (0.3602) loss 3.2710 (3.1236) grad_norm 3.3061 (2.7777) [2021-04-16 16:36:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][70/1251] eta 0:06:51 lr 0.000074 time 0.2627 (0.3485) loss 3.9058 (3.1190) grad_norm 2.7553 (2.8695) [2021-04-16 16:36:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][80/1251] eta 0:06:38 lr 0.000074 time 0.2809 (0.3402) loss 3.4097 (3.1059) grad_norm 2.4928 (2.8525) [2021-04-16 16:36:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][90/1251] eta 0:06:27 lr 0.000074 time 0.2844 (0.3333) loss 2.2555 (3.0969) grad_norm 2.2394 (2.8375) [2021-04-16 16:36:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][100/1251] eta 0:06:16 lr 0.000074 time 0.2820 (0.3275) loss 2.5810 (3.0939) grad_norm 2.8966 (2.8191) [2021-04-16 16:36:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][110/1251] eta 0:06:10 lr 0.000074 time 0.4319 (0.3243) loss 3.5644 (3.0770) grad_norm 2.6848 (2.8326) [2021-04-16 16:36:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][120/1251] eta 0:06:01 lr 0.000074 time 0.2818 (0.3201) loss 2.5608 (3.0912) grad_norm 2.9158 (2.8207) [2021-04-16 16:36:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][130/1251] eta 0:05:56 lr 0.000073 time 0.2668 (0.3183) loss 3.8838 (3.1058) grad_norm 2.9665 (2.8148) [2021-04-16 16:36:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][140/1251] eta 0:05:51 lr 0.000073 time 0.2770 (0.3164) loss 2.5280 (3.1013) grad_norm 2.9092 (2.8265) [2021-04-16 16:36:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][150/1251] eta 0:05:45 lr 0.000073 time 0.2780 (0.3139) loss 3.1754 (3.0866) grad_norm 3.5202 (2.8418) [2021-04-16 16:36:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][160/1251] eta 0:05:40 lr 0.000073 time 0.2440 (0.3121) loss 3.6030 (3.0784) grad_norm 2.7062 (2.8533) [2021-04-16 16:36:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][170/1251] eta 0:05:35 lr 0.000073 time 0.2620 (0.3100) loss 3.2813 (3.0727) grad_norm 2.9264 (2.8530) [2021-04-16 16:36:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][180/1251] eta 0:05:30 lr 0.000073 time 0.3050 (0.3082) loss 2.9240 (3.0692) grad_norm 2.7056 (2.8612) [2021-04-16 16:36:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][190/1251] eta 0:05:25 lr 0.000073 time 0.2689 (0.3065) loss 3.5434 (3.0720) grad_norm 2.7117 (2.8687) [2021-04-16 16:36:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][200/1251] eta 0:05:20 lr 0.000073 time 0.2869 (0.3049) loss 3.1915 (3.0751) grad_norm 3.0986 (2.8670) [2021-04-16 16:36:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][210/1251] eta 0:05:15 lr 0.000073 time 0.2666 (0.3034) loss 3.1220 (3.0831) grad_norm 2.7564 (2.8693) [2021-04-16 16:37:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][220/1251] eta 0:05:11 lr 0.000073 time 0.3085 (0.3023) loss 3.4565 (3.0745) grad_norm 2.9353 (2.8627) [2021-04-16 16:37:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][230/1251] eta 0:05:07 lr 0.000073 time 0.2964 (0.3014) loss 2.3148 (3.0729) grad_norm 2.9941 (2.8603) [2021-04-16 16:37:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][240/1251] eta 0:05:03 lr 0.000073 time 0.2765 (0.3002) loss 2.7790 (3.0689) grad_norm 2.4765 (2.8581) [2021-04-16 16:37:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][250/1251] eta 0:04:59 lr 0.000073 time 0.2856 (0.2991) loss 2.9924 (3.0547) grad_norm 2.7952 (2.8548) [2021-04-16 16:37:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][260/1251] eta 0:04:55 lr 0.000073 time 0.2468 (0.2982) loss 3.2011 (3.0467) grad_norm 2.7891 (2.8523) [2021-04-16 16:37:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][270/1251] eta 0:04:51 lr 0.000073 time 0.2659 (0.2973) loss 2.0589 (3.0445) grad_norm 3.0762 (2.8530) [2021-04-16 16:37:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][280/1251] eta 0:04:47 lr 0.000073 time 0.2588 (0.2964) loss 2.1880 (3.0443) grad_norm 2.5332 (2.8560) [2021-04-16 16:37:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][290/1251] eta 0:04:44 lr 0.000073 time 0.2943 (0.2958) loss 3.4751 (3.0509) grad_norm 2.8641 (2.8521) [2021-04-16 16:37:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][300/1251] eta 0:04:41 lr 0.000073 time 0.2743 (0.2956) loss 3.6806 (3.0619) grad_norm 3.4229 (2.8530) [2021-04-16 16:37:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][310/1251] eta 0:04:37 lr 0.000073 time 0.2776 (0.2949) loss 3.4295 (3.0679) grad_norm 8.0451 (2.8671) [2021-04-16 16:37:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][320/1251] eta 0:04:34 lr 0.000073 time 0.2782 (0.2946) loss 3.1857 (3.0745) grad_norm 2.6629 (2.8657) [2021-04-16 16:37:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][330/1251] eta 0:04:31 lr 0.000073 time 0.2647 (0.2944) loss 2.1902 (3.0752) grad_norm 2.7651 (2.8625) [2021-04-16 16:37:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][340/1251] eta 0:04:27 lr 0.000073 time 0.3251 (0.2939) loss 3.1132 (3.0726) grad_norm 2.3814 (2.8636) [2021-04-16 16:37:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][350/1251] eta 0:04:24 lr 0.000073 time 0.2843 (0.2934) loss 2.9928 (3.0707) grad_norm 2.7671 (2.8617) [2021-04-16 16:37:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][360/1251] eta 0:04:21 lr 0.000073 time 0.2997 (0.2932) loss 3.2721 (3.0703) grad_norm 3.3410 (2.8611) [2021-04-16 16:37:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][370/1251] eta 0:04:18 lr 0.000073 time 0.2594 (0.2935) loss 3.4960 (3.0705) grad_norm 3.4072 (2.8591) [2021-04-16 16:37:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][380/1251] eta 0:04:15 lr 0.000073 time 0.2728 (0.2928) loss 3.2935 (3.0681) grad_norm 3.0759 (nan) [2021-04-16 16:37:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][390/1251] eta 0:04:11 lr 0.000073 time 0.2663 (0.2923) loss 2.4449 (3.0724) grad_norm 2.9091 (nan) [2021-04-16 16:37:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][400/1251] eta 0:04:08 lr 0.000073 time 0.2796 (0.2921) loss 3.2144 (3.0678) grad_norm 2.6495 (nan) [2021-04-16 16:37:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][410/1251] eta 0:04:05 lr 0.000073 time 0.2592 (0.2917) loss 2.5468 (3.0686) grad_norm 2.8584 (nan) [2021-04-16 16:37:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][420/1251] eta 0:04:02 lr 0.000073 time 0.2595 (0.2913) loss 3.5339 (3.0702) grad_norm 3.9519 (nan) [2021-04-16 16:37:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][430/1251] eta 0:03:58 lr 0.000073 time 0.2558 (0.2909) loss 2.1647 (3.0636) grad_norm 2.4416 (nan) [2021-04-16 16:38:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][440/1251] eta 0:03:55 lr 0.000073 time 0.2657 (0.2905) loss 3.4994 (3.0683) grad_norm 2.2954 (nan) [2021-04-16 16:38:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][450/1251] eta 0:03:52 lr 0.000073 time 0.2878 (0.2903) loss 3.4431 (3.0733) grad_norm 3.1474 (nan) [2021-04-16 16:38:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][460/1251] eta 0:03:49 lr 0.000073 time 0.2779 (0.2901) loss 3.8544 (3.0745) grad_norm 2.5408 (nan) [2021-04-16 16:38:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][470/1251] eta 0:03:46 lr 0.000073 time 0.2764 (0.2898) loss 2.0742 (3.0728) grad_norm 2.7903 (nan) [2021-04-16 16:38:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][480/1251] eta 0:03:43 lr 0.000073 time 0.2805 (0.2895) loss 3.5481 (3.0730) grad_norm 3.1426 (nan) [2021-04-16 16:38:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][490/1251] eta 0:03:40 lr 0.000073 time 0.2888 (0.2893) loss 3.6130 (3.0764) grad_norm 2.8376 (nan) [2021-04-16 16:38:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][500/1251] eta 0:03:37 lr 0.000073 time 0.2827 (0.2890) loss 2.5307 (3.0742) grad_norm 2.9132 (nan) [2021-04-16 16:38:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][510/1251] eta 0:03:33 lr 0.000073 time 0.2599 (0.2887) loss 3.0541 (3.0724) grad_norm 2.6745 (nan) [2021-04-16 16:38:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][520/1251] eta 0:03:30 lr 0.000073 time 0.2543 (0.2884) loss 3.6120 (3.0703) grad_norm 2.5264 (nan) [2021-04-16 16:38:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][530/1251] eta 0:03:27 lr 0.000073 time 0.2588 (0.2882) loss 3.5421 (3.0732) grad_norm 2.6886 (nan) [2021-04-16 16:38:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][540/1251] eta 0:03:24 lr 0.000073 time 0.2843 (0.2881) loss 3.4282 (3.0746) grad_norm 3.1756 (nan) [2021-04-16 16:38:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][550/1251] eta 0:03:21 lr 0.000073 time 0.2897 (0.2880) loss 3.2485 (3.0746) grad_norm 3.0583 (nan) [2021-04-16 16:38:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][560/1251] eta 0:03:18 lr 0.000073 time 0.2721 (0.2877) loss 3.1944 (3.0716) grad_norm 2.7069 (nan) [2021-04-16 16:38:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][570/1251] eta 0:03:16 lr 0.000073 time 0.2976 (0.2880) loss 2.8774 (3.0711) grad_norm 2.4756 (nan) [2021-04-16 16:38:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][580/1251] eta 0:03:13 lr 0.000073 time 0.2664 (0.2878) loss 3.2296 (3.0702) grad_norm 2.8426 (nan) [2021-04-16 16:38:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][590/1251] eta 0:03:10 lr 0.000073 time 0.2738 (0.2877) loss 3.4531 (3.0727) grad_norm 3.0109 (nan) [2021-04-16 16:38:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][600/1251] eta 0:03:07 lr 0.000073 time 0.2892 (0.2875) loss 3.8702 (3.0725) grad_norm 3.3677 (nan) [2021-04-16 16:38:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][610/1251] eta 0:03:04 lr 0.000073 time 0.2610 (0.2874) loss 3.6707 (3.0726) grad_norm 2.7702 (nan) [2021-04-16 16:38:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][620/1251] eta 0:03:01 lr 0.000072 time 0.2991 (0.2874) loss 3.6903 (3.0755) grad_norm 3.0493 (nan) [2021-04-16 16:38:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][630/1251] eta 0:02:58 lr 0.000072 time 0.2677 (0.2872) loss 2.7226 (3.0769) grad_norm 2.7940 (nan) [2021-04-16 16:38:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][640/1251] eta 0:02:55 lr 0.000072 time 0.2640 (0.2871) loss 2.2907 (3.0745) grad_norm 2.5842 (nan) [2021-04-16 16:39:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][650/1251] eta 0:02:52 lr 0.000072 time 0.2838 (0.2870) loss 3.6968 (3.0746) grad_norm 2.5090 (nan) [2021-04-16 16:39:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][660/1251] eta 0:02:49 lr 0.000072 time 0.2740 (0.2868) loss 3.8513 (3.0777) grad_norm 3.3869 (nan) [2021-04-16 16:39:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][670/1251] eta 0:02:46 lr 0.000072 time 0.2899 (0.2867) loss 3.6595 (3.0729) grad_norm 2.8307 (nan) [2021-04-16 16:39:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][680/1251] eta 0:02:43 lr 0.000072 time 0.3062 (0.2866) loss 3.7193 (3.0693) grad_norm 3.2494 (nan) [2021-04-16 16:39:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][690/1251] eta 0:02:40 lr 0.000072 time 0.2929 (0.2865) loss 2.8285 (3.0704) grad_norm 2.6823 (nan) [2021-04-16 16:39:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][700/1251] eta 0:02:37 lr 0.000072 time 0.3875 (0.2865) loss 3.8329 (3.0701) grad_norm 2.6517 (nan) [2021-04-16 16:39:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][710/1251] eta 0:02:34 lr 0.000072 time 0.2660 (0.2864) loss 2.8063 (3.0715) grad_norm 2.8992 (nan) [2021-04-16 16:39:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][720/1251] eta 0:02:32 lr 0.000072 time 0.2593 (0.2864) loss 3.1472 (3.0725) grad_norm 2.8251 (nan) [2021-04-16 16:39:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][730/1251] eta 0:02:29 lr 0.000072 time 0.2982 (0.2863) loss 3.6444 (3.0766) grad_norm 2.4485 (nan) [2021-04-16 16:39:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][740/1251] eta 0:02:26 lr 0.000072 time 0.2463 (0.2863) loss 2.6108 (3.0754) grad_norm 2.5909 (nan) [2021-04-16 16:39:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][750/1251] eta 0:02:23 lr 0.000072 time 0.2919 (0.2862) loss 3.1282 (3.0753) grad_norm 2.4697 (nan) [2021-04-16 16:39:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][760/1251] eta 0:02:20 lr 0.000072 time 0.2618 (0.2860) loss 3.7379 (3.0756) grad_norm 2.2039 (nan) [2021-04-16 16:39:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][770/1251] eta 0:02:17 lr 0.000072 time 0.2732 (0.2859) loss 2.0923 (3.0743) grad_norm 2.4974 (nan) [2021-04-16 16:39:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][780/1251] eta 0:02:14 lr 0.000072 time 0.2478 (0.2858) loss 2.3415 (3.0742) grad_norm 2.3891 (nan) [2021-04-16 16:39:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][790/1251] eta 0:02:11 lr 0.000072 time 0.2816 (0.2857) loss 3.2705 (3.0757) grad_norm 2.8149 (nan) [2021-04-16 16:39:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][800/1251] eta 0:02:08 lr 0.000072 time 0.2697 (0.2856) loss 3.1651 (3.0736) grad_norm 3.0166 (nan) [2021-04-16 16:39:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][810/1251] eta 0:02:05 lr 0.000072 time 0.2705 (0.2854) loss 2.5430 (3.0724) grad_norm 3.0970 (nan) [2021-04-16 16:39:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][820/1251] eta 0:02:02 lr 0.000072 time 0.2707 (0.2853) loss 2.7784 (3.0708) grad_norm 3.0527 (nan) [2021-04-16 16:39:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][830/1251] eta 0:02:00 lr 0.000072 time 0.2893 (0.2852) loss 2.7018 (3.0708) grad_norm 3.0621 (nan) [2021-04-16 16:39:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][840/1251] eta 0:01:57 lr 0.000072 time 0.2996 (0.2851) loss 2.6011 (3.0679) grad_norm 2.7149 (nan) [2021-04-16 16:39:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][850/1251] eta 0:01:54 lr 0.000072 time 0.2626 (0.2850) loss 3.5643 (3.0710) grad_norm 3.3680 (nan) [2021-04-16 16:39:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][860/1251] eta 0:01:51 lr 0.000072 time 0.2751 (0.2848) loss 3.2937 (3.0712) grad_norm 2.5623 (nan) [2021-04-16 16:40:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][870/1251] eta 0:01:48 lr 0.000072 time 0.2722 (0.2848) loss 2.9941 (3.0710) grad_norm 2.8012 (nan) [2021-04-16 16:40:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][880/1251] eta 0:01:45 lr 0.000072 time 0.2758 (0.2848) loss 3.4106 (3.0693) grad_norm 3.0064 (nan) [2021-04-16 16:40:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][890/1251] eta 0:01:42 lr 0.000072 time 0.2719 (0.2848) loss 3.1174 (3.0690) grad_norm 2.9147 (nan) [2021-04-16 16:40:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][900/1251] eta 0:01:39 lr 0.000072 time 0.4301 (0.2849) loss 3.1662 (3.0698) grad_norm 2.7578 (nan) [2021-04-16 16:40:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][910/1251] eta 0:01:37 lr 0.000072 time 0.2932 (0.2848) loss 2.5977 (3.0687) grad_norm 2.4996 (nan) [2021-04-16 16:40:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][920/1251] eta 0:01:34 lr 0.000072 time 0.2995 (0.2847) loss 2.5361 (3.0700) grad_norm 2.9894 (nan) [2021-04-16 16:40:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][930/1251] eta 0:01:31 lr 0.000072 time 0.2629 (0.2846) loss 3.4332 (3.0701) grad_norm 2.3094 (nan) [2021-04-16 16:40:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][940/1251] eta 0:01:28 lr 0.000072 time 0.2466 (0.2846) loss 3.2938 (3.0693) grad_norm 2.8305 (nan) [2021-04-16 16:40:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][950/1251] eta 0:01:25 lr 0.000072 time 0.2868 (0.2847) loss 3.7746 (3.0715) grad_norm 2.7172 (nan) [2021-04-16 16:40:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][960/1251] eta 0:01:22 lr 0.000072 time 0.2890 (0.2846) loss 3.1601 (3.0719) grad_norm 2.9842 (nan) [2021-04-16 16:40:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][970/1251] eta 0:01:19 lr 0.000072 time 0.2583 (0.2845) loss 3.2602 (3.0713) grad_norm 3.2684 (nan) [2021-04-16 16:40:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][980/1251] eta 0:01:17 lr 0.000072 time 0.2662 (0.2844) loss 3.1269 (3.0701) grad_norm 2.9233 (nan) [2021-04-16 16:40:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][990/1251] eta 0:01:14 lr 0.000072 time 0.2775 (0.2844) loss 3.1750 (3.0697) grad_norm 3.0114 (nan) [2021-04-16 16:40:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][1000/1251] eta 0:01:11 lr 0.000072 time 0.2872 (0.2843) loss 3.5273 (3.0681) grad_norm 2.6432 (nan) [2021-04-16 16:40:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][1010/1251] eta 0:01:08 lr 0.000072 time 0.2791 (0.2842) loss 2.4913 (3.0684) grad_norm 2.7304 (nan) [2021-04-16 16:40:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][1020/1251] eta 0:01:05 lr 0.000072 time 0.2533 (0.2842) loss 3.6368 (3.0700) grad_norm 2.6753 (nan) [2021-04-16 16:40:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][1030/1251] eta 0:01:02 lr 0.000072 time 0.2790 (0.2842) loss 3.4606 (3.0704) grad_norm 4.7660 (nan) [2021-04-16 16:40:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][1040/1251] eta 0:00:59 lr 0.000072 time 0.2646 (0.2841) loss 2.9661 (3.0710) grad_norm 3.1987 (nan) [2021-04-16 16:40:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][1050/1251] eta 0:00:57 lr 0.000072 time 0.2617 (0.2840) loss 3.1676 (3.0709) grad_norm 3.5148 (nan) [2021-04-16 16:40:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][1060/1251] eta 0:00:54 lr 0.000072 time 0.2720 (0.2840) loss 2.7259 (3.0685) grad_norm 2.9432 (nan) [2021-04-16 16:40:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][1070/1251] eta 0:00:51 lr 0.000072 time 0.2786 (0.2839) loss 3.0767 (3.0683) grad_norm 2.4885 (nan) [2021-04-16 16:41:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][1080/1251] eta 0:00:48 lr 0.000072 time 0.2770 (0.2839) loss 3.1031 (3.0700) grad_norm 2.8484 (nan) [2021-04-16 16:41:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][1090/1251] eta 0:00:45 lr 0.000072 time 0.2679 (0.2838) loss 3.4872 (3.0719) grad_norm 3.0842 (nan) [2021-04-16 16:41:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][1100/1251] eta 0:00:42 lr 0.000072 time 0.2748 (0.2838) loss 2.6488 (3.0727) grad_norm 2.7325 (nan) [2021-04-16 16:41:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][1110/1251] eta 0:00:40 lr 0.000072 time 0.2704 (0.2837) loss 2.7246 (3.0726) grad_norm 3.3877 (nan) [2021-04-16 16:41:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][1120/1251] eta 0:00:37 lr 0.000071 time 0.2794 (0.2837) loss 3.6230 (3.0737) grad_norm 2.7610 (nan) [2021-04-16 16:41:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][1130/1251] eta 0:00:34 lr 0.000071 time 0.2844 (0.2837) loss 3.6676 (3.0714) grad_norm 2.4732 (nan) [2021-04-16 16:41:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][1140/1251] eta 0:00:31 lr 0.000071 time 0.2848 (0.2837) loss 2.9905 (3.0722) grad_norm 3.0579 (nan) [2021-04-16 16:41:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][1150/1251] eta 0:00:28 lr 0.000071 time 0.2782 (0.2837) loss 3.4771 (3.0738) grad_norm 2.7002 (nan) [2021-04-16 16:41:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][1160/1251] eta 0:00:25 lr 0.000071 time 0.2664 (0.2837) loss 3.5529 (3.0747) grad_norm 2.5034 (nan) [2021-04-16 16:41:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][1170/1251] eta 0:00:22 lr 0.000071 time 0.2966 (0.2837) loss 3.5749 (3.0753) grad_norm 2.7691 (nan) [2021-04-16 16:41:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][1180/1251] eta 0:00:20 lr 0.000071 time 0.2613 (0.2837) loss 2.6525 (3.0757) grad_norm 2.7173 (nan) [2021-04-16 16:41:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][1190/1251] eta 0:00:17 lr 0.000071 time 0.2744 (0.2838) loss 3.1871 (3.0755) grad_norm 2.5479 (nan) [2021-04-16 16:41:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][1200/1251] eta 0:00:14 lr 0.000071 time 0.2447 (0.2837) loss 3.5048 (3.0764) grad_norm 3.2357 (nan) [2021-04-16 16:41:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][1210/1251] eta 0:00:11 lr 0.000071 time 0.2771 (0.2836) loss 3.0411 (3.0782) grad_norm 3.0743 (nan) [2021-04-16 16:41:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][1220/1251] eta 0:00:08 lr 0.000071 time 0.2695 (0.2836) loss 3.6150 (3.0785) grad_norm 2.5028 (nan) [2021-04-16 16:41:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][1230/1251] eta 0:00:05 lr 0.000071 time 0.2989 (0.2836) loss 3.0815 (3.0800) grad_norm 2.6624 (nan) [2021-04-16 16:41:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][1240/1251] eta 0:00:03 lr 0.000071 time 0.2483 (0.2835) loss 3.2290 (3.0812) grad_norm 2.3754 (nan) [2021-04-16 16:41:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [251/300][1250/1251] eta 0:00:00 lr 0.000071 time 0.2488 (0.2832) loss 3.3227 (3.0807) grad_norm 2.9537 (nan) [2021-04-16 16:41:50 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 251 training takes 0:05:57 [2021-04-16 16:41:50 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_251.pth saving...... [2021-04-16 16:42:00 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_251.pth saved !!! [2021-04-16 16:42:01 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.075 (1.075) Loss 0.8416 (0.8416) Acc@1 80.371 (80.371) Acc@5 95.312 (95.312) [2021-04-16 16:42:03 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.120 (0.241) Loss 0.7944 (0.8469) Acc@1 82.227 (80.211) Acc@5 95.215 (94.993) [2021-04-16 16:42:05 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.151 (0.216) Loss 0.8839 (0.8268) Acc@1 78.906 (80.422) Acc@5 94.629 (95.382) [2021-04-16 16:42:07 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.154 (0.216) Loss 0.7874 (0.8257) Acc@1 82.324 (80.538) Acc@5 96.094 (95.398) [2021-04-16 16:42:09 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.210) Loss 0.8735 (0.8308) Acc@1 79.785 (80.440) Acc@5 94.434 (95.282) [2021-04-16 16:42:19 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.386 Acc@5 95.236 [2021-04-16 16:42:19 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 80.4% [2021-04-16 16:42:19 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 80.46% [2021-04-16 16:42:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][0/1251] eta 1:57:38 lr 0.000071 time 5.6421 (5.6421) loss 3.0252 (3.0252) grad_norm 2.7458 (2.7458) [2021-04-16 16:42:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][10/1251] eta 0:15:46 lr 0.000071 time 0.2609 (0.7627) loss 3.3359 (3.2852) grad_norm 2.7745 (2.8339) [2021-04-16 16:42:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][20/1251] eta 0:10:55 lr 0.000071 time 0.2480 (0.5327) loss 2.4939 (3.0426) grad_norm 2.3557 (2.7745) [2021-04-16 16:42:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][30/1251] eta 0:09:13 lr 0.000071 time 0.2919 (0.4537) loss 3.0776 (3.0710) grad_norm 2.5470 (2.7620) [2021-04-16 16:42:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][40/1251] eta 0:08:18 lr 0.000071 time 0.2764 (0.4116) loss 3.2527 (3.0882) grad_norm 3.4072 (2.8009) [2021-04-16 16:42:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][50/1251] eta 0:07:43 lr 0.000071 time 0.2706 (0.3859) loss 3.4422 (3.0874) grad_norm 2.4711 (2.7811) [2021-04-16 16:42:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][60/1251] eta 0:07:17 lr 0.000071 time 0.2974 (0.3677) loss 3.0960 (3.0352) grad_norm 2.5942 (2.7992) [2021-04-16 16:42:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][70/1251] eta 0:06:59 lr 0.000071 time 0.2891 (0.3555) loss 2.9714 (3.0225) grad_norm 2.6170 (2.8195) [2021-04-16 16:42:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][80/1251] eta 0:06:46 lr 0.000071 time 0.2812 (0.3468) loss 2.5304 (3.0044) grad_norm 3.2839 (2.8593) [2021-04-16 16:42:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][90/1251] eta 0:06:34 lr 0.000071 time 0.2796 (0.3394) loss 3.1378 (3.0103) grad_norm 2.6270 (2.8472) [2021-04-16 16:42:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][100/1251] eta 0:06:25 lr 0.000071 time 0.2621 (0.3350) loss 2.0180 (3.0060) grad_norm 3.4436 (2.8651) [2021-04-16 16:42:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][110/1251] eta 0:06:15 lr 0.000071 time 0.2613 (0.3292) loss 2.3292 (3.0044) grad_norm 2.8365 (2.8784) [2021-04-16 16:42:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][120/1251] eta 0:06:07 lr 0.000071 time 0.2938 (0.3251) loss 3.3377 (3.0138) grad_norm 2.9590 (2.8713) [2021-04-16 16:43:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][130/1251] eta 0:06:00 lr 0.000071 time 0.2600 (0.3212) loss 2.0982 (2.9802) grad_norm 2.8597 (2.8754) [2021-04-16 16:43:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][140/1251] eta 0:05:54 lr 0.000071 time 0.2845 (0.3191) loss 2.5445 (2.9614) grad_norm 2.4950 (2.8675) [2021-04-16 16:43:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][150/1251] eta 0:05:49 lr 0.000071 time 0.2758 (0.3171) loss 3.1397 (2.9714) grad_norm 2.7532 (2.8620) [2021-04-16 16:43:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][160/1251] eta 0:05:43 lr 0.000071 time 0.2839 (0.3144) loss 3.3468 (2.9778) grad_norm 2.2171 (2.8648) [2021-04-16 16:43:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][170/1251] eta 0:05:37 lr 0.000071 time 0.2946 (0.3123) loss 2.6935 (2.9851) grad_norm 2.9570 (2.8569) [2021-04-16 16:43:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][180/1251] eta 0:05:32 lr 0.000071 time 0.2757 (0.3101) loss 3.5869 (3.0064) grad_norm 2.4777 (2.8547) [2021-04-16 16:43:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][190/1251] eta 0:05:27 lr 0.000071 time 0.2949 (0.3090) loss 2.3542 (3.0161) grad_norm 2.6243 (2.8543) [2021-04-16 16:43:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][200/1251] eta 0:05:23 lr 0.000071 time 0.3025 (0.3077) loss 3.6172 (3.0228) grad_norm 2.3002 (2.8445) [2021-04-16 16:43:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][210/1251] eta 0:05:18 lr 0.000071 time 0.2736 (0.3061) loss 2.5659 (3.0231) grad_norm 2.4453 (2.8354) [2021-04-16 16:43:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][220/1251] eta 0:05:14 lr 0.000071 time 0.2786 (0.3048) loss 3.1812 (3.0193) grad_norm 3.0561 (2.8505) [2021-04-16 16:43:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][230/1251] eta 0:05:09 lr 0.000071 time 0.2547 (0.3035) loss 3.3518 (3.0272) grad_norm 2.8714 (2.8453) [2021-04-16 16:43:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][240/1251] eta 0:05:05 lr 0.000071 time 0.2753 (0.3024) loss 3.5993 (3.0359) grad_norm 2.6767 (2.8417) [2021-04-16 16:43:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][250/1251] eta 0:05:01 lr 0.000071 time 0.2577 (0.3014) loss 3.3142 (3.0409) grad_norm 2.7487 (2.8368) [2021-04-16 16:43:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][260/1251] eta 0:04:57 lr 0.000071 time 0.2668 (0.3005) loss 3.2847 (3.0481) grad_norm 2.8124 (2.8366) [2021-04-16 16:43:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][270/1251] eta 0:04:54 lr 0.000071 time 0.2894 (0.2999) loss 3.8280 (3.0456) grad_norm 2.6095 (2.8339) [2021-04-16 16:43:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][280/1251] eta 0:04:50 lr 0.000071 time 0.2633 (0.2991) loss 3.3796 (3.0526) grad_norm 3.0291 (2.8311) [2021-04-16 16:43:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][290/1251] eta 0:04:46 lr 0.000071 time 0.2753 (0.2985) loss 2.4664 (3.0486) grad_norm 2.3361 (2.8279) [2021-04-16 16:43:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][300/1251] eta 0:04:43 lr 0.000071 time 0.2535 (0.2978) loss 3.7880 (3.0521) grad_norm 2.8579 (2.8257) [2021-04-16 16:43:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][310/1251] eta 0:04:39 lr 0.000071 time 0.2937 (0.2971) loss 3.1804 (3.0559) grad_norm 2.6945 (2.8231) [2021-04-16 16:43:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][320/1251] eta 0:04:36 lr 0.000071 time 0.2731 (0.2968) loss 2.4554 (3.0553) grad_norm 2.9309 (2.8220) [2021-04-16 16:43:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][330/1251] eta 0:04:32 lr 0.000071 time 0.2693 (0.2962) loss 2.3762 (3.0549) grad_norm 2.9560 (2.8216) [2021-04-16 16:44:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][340/1251] eta 0:04:29 lr 0.000071 time 0.2701 (0.2960) loss 2.5278 (3.0517) grad_norm 3.2243 (2.8201) [2021-04-16 16:44:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][350/1251] eta 0:04:26 lr 0.000071 time 0.2647 (0.2958) loss 3.1400 (3.0518) grad_norm 3.4231 (2.8210) [2021-04-16 16:44:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][360/1251] eta 0:04:23 lr 0.000071 time 0.2607 (0.2954) loss 3.4315 (3.0589) grad_norm 2.6796 (2.8221) [2021-04-16 16:44:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][370/1251] eta 0:04:20 lr 0.000070 time 0.2795 (0.2952) loss 3.0137 (3.0605) grad_norm 3.6502 (2.8246) [2021-04-16 16:44:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][380/1251] eta 0:04:16 lr 0.000070 time 0.2605 (0.2947) loss 2.7945 (3.0598) grad_norm 2.9386 (2.8282) [2021-04-16 16:44:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][390/1251] eta 0:04:13 lr 0.000070 time 0.2727 (0.2950) loss 2.4688 (3.0568) grad_norm 2.3920 (2.8291) [2021-04-16 16:44:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][400/1251] eta 0:04:10 lr 0.000070 time 0.2830 (0.2946) loss 2.9264 (3.0598) grad_norm 3.2132 (2.8312) [2021-04-16 16:44:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][410/1251] eta 0:04:07 lr 0.000070 time 0.2572 (0.2943) loss 3.4639 (3.0570) grad_norm 2.6859 (2.8302) [2021-04-16 16:44:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][420/1251] eta 0:04:04 lr 0.000070 time 0.2756 (0.2941) loss 2.7988 (3.0617) grad_norm 2.6147 (2.8289) [2021-04-16 16:44:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][430/1251] eta 0:04:01 lr 0.000070 time 0.2592 (0.2936) loss 3.3163 (3.0618) grad_norm 2.3344 (2.8277) [2021-04-16 16:44:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][440/1251] eta 0:03:58 lr 0.000070 time 0.2663 (0.2935) loss 3.0727 (3.0603) grad_norm 3.0042 (2.8261) [2021-04-16 16:44:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][450/1251] eta 0:03:54 lr 0.000070 time 0.2957 (0.2932) loss 3.1952 (3.0669) grad_norm 2.4790 (2.8313) [2021-04-16 16:44:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][460/1251] eta 0:03:51 lr 0.000070 time 0.2891 (0.2928) loss 2.1041 (3.0650) grad_norm 3.1514 (2.8332) [2021-04-16 16:44:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][470/1251] eta 0:03:48 lr 0.000070 time 0.2544 (0.2924) loss 2.5485 (3.0621) grad_norm 2.7201 (2.8347) [2021-04-16 16:44:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][480/1251] eta 0:03:45 lr 0.000070 time 0.2753 (0.2921) loss 2.9063 (3.0631) grad_norm 2.5782 (2.8356) [2021-04-16 16:44:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][490/1251] eta 0:03:42 lr 0.000070 time 0.2910 (0.2919) loss 2.3335 (3.0624) grad_norm 2.7745 (2.8388) [2021-04-16 16:44:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][500/1251] eta 0:03:38 lr 0.000070 time 0.2728 (0.2916) loss 3.1863 (3.0613) grad_norm 2.8519 (2.8377) [2021-04-16 16:44:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][510/1251] eta 0:03:35 lr 0.000070 time 0.2886 (0.2913) loss 2.1857 (3.0589) grad_norm 3.3519 (2.8432) [2021-04-16 16:44:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][520/1251] eta 0:03:32 lr 0.000070 time 0.2718 (0.2911) loss 3.1914 (3.0612) grad_norm 2.6739 (2.8436) [2021-04-16 16:44:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][530/1251] eta 0:03:29 lr 0.000070 time 0.2851 (0.2909) loss 1.9840 (3.0587) grad_norm 3.7468 (2.8482) [2021-04-16 16:44:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][540/1251] eta 0:03:26 lr 0.000070 time 0.2715 (0.2908) loss 3.4200 (3.0589) grad_norm 2.5100 (2.8624) [2021-04-16 16:44:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][550/1251] eta 0:03:23 lr 0.000070 time 0.2650 (0.2909) loss 2.6988 (3.0581) grad_norm 2.6351 (2.8664) [2021-04-16 16:45:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][560/1251] eta 0:03:20 lr 0.000070 time 0.2686 (0.2905) loss 2.5439 (3.0562) grad_norm 3.1964 (2.8682) [2021-04-16 16:45:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][570/1251] eta 0:03:18 lr 0.000070 time 0.4254 (0.2908) loss 3.3576 (3.0593) grad_norm 2.4975 (2.8667) [2021-04-16 16:45:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][580/1251] eta 0:03:15 lr 0.000070 time 0.2599 (0.2907) loss 3.5221 (3.0643) grad_norm 2.6420 (2.8673) [2021-04-16 16:45:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][590/1251] eta 0:03:11 lr 0.000070 time 0.2951 (0.2904) loss 3.2735 (3.0702) grad_norm 3.0527 (2.8644) [2021-04-16 16:45:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][600/1251] eta 0:03:08 lr 0.000070 time 0.2640 (0.2901) loss 3.6803 (3.0711) grad_norm 2.4529 (2.8637) [2021-04-16 16:45:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][610/1251] eta 0:03:05 lr 0.000070 time 0.2608 (0.2899) loss 2.4434 (3.0696) grad_norm 3.8166 (2.8632) [2021-04-16 16:45:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][620/1251] eta 0:03:02 lr 0.000070 time 0.2778 (0.2897) loss 3.3077 (3.0756) grad_norm 2.3442 (2.8624) [2021-04-16 16:45:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][630/1251] eta 0:02:59 lr 0.000070 time 0.2695 (0.2895) loss 3.4302 (3.0745) grad_norm 3.4013 (2.8614) [2021-04-16 16:45:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][640/1251] eta 0:02:56 lr 0.000070 time 0.2765 (0.2892) loss 2.3809 (3.0750) grad_norm 2.5076 (2.8575) [2021-04-16 16:45:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][650/1251] eta 0:02:53 lr 0.000070 time 0.3730 (0.2891) loss 3.4898 (3.0745) grad_norm 2.9976 (2.8562) [2021-04-16 16:45:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][660/1251] eta 0:02:50 lr 0.000070 time 0.2611 (0.2889) loss 2.4224 (3.0738) grad_norm 3.3388 (2.8576) [2021-04-16 16:45:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][670/1251] eta 0:02:47 lr 0.000070 time 0.2632 (0.2888) loss 3.0265 (3.0729) grad_norm 2.7388 (2.8569) [2021-04-16 16:45:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][680/1251] eta 0:02:44 lr 0.000070 time 0.2646 (0.2887) loss 2.3757 (3.0705) grad_norm 2.6713 (2.8589) [2021-04-16 16:45:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][690/1251] eta 0:02:41 lr 0.000070 time 0.2671 (0.2886) loss 2.9914 (3.0674) grad_norm 2.4071 (2.8577) [2021-04-16 16:45:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][700/1251] eta 0:02:38 lr 0.000070 time 0.2783 (0.2885) loss 3.8326 (3.0697) grad_norm 2.5437 (2.8553) [2021-04-16 16:45:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][710/1251] eta 0:02:35 lr 0.000070 time 0.2766 (0.2883) loss 1.6425 (3.0675) grad_norm 2.7464 (2.8593) [2021-04-16 16:45:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][720/1251] eta 0:02:33 lr 0.000070 time 0.2518 (0.2883) loss 2.2593 (3.0649) grad_norm 3.0816 (2.8591) [2021-04-16 16:45:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][730/1251] eta 0:02:30 lr 0.000070 time 0.2697 (0.2883) loss 2.2183 (3.0664) grad_norm 3.2510 (2.8590) [2021-04-16 16:45:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][740/1251] eta 0:02:27 lr 0.000070 time 0.3019 (0.2881) loss 3.6754 (3.0636) grad_norm 2.6614 (2.8625) [2021-04-16 16:45:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][750/1251] eta 0:02:24 lr 0.000070 time 0.2732 (0.2879) loss 3.5910 (3.0634) grad_norm 3.0796 (2.8614) [2021-04-16 16:45:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][760/1251] eta 0:02:21 lr 0.000070 time 0.2803 (0.2878) loss 3.4876 (3.0604) grad_norm 2.7138 (2.8598) [2021-04-16 16:46:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][770/1251] eta 0:02:18 lr 0.000070 time 0.2864 (0.2876) loss 3.3471 (3.0625) grad_norm 2.7886 (2.8584) [2021-04-16 16:46:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][780/1251] eta 0:02:15 lr 0.000070 time 0.2755 (0.2875) loss 2.5622 (3.0625) grad_norm 2.8577 (2.8579) [2021-04-16 16:46:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][790/1251] eta 0:02:12 lr 0.000070 time 0.2440 (0.2873) loss 2.3546 (3.0630) grad_norm 2.6895 (2.8573) [2021-04-16 16:46:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][800/1251] eta 0:02:09 lr 0.000070 time 0.2490 (0.2872) loss 3.3473 (3.0634) grad_norm 2.7821 (2.8552) [2021-04-16 16:46:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][810/1251] eta 0:02:06 lr 0.000070 time 0.2885 (0.2870) loss 3.4654 (3.0651) grad_norm 3.0138 (2.8566) [2021-04-16 16:46:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][820/1251] eta 0:02:03 lr 0.000070 time 0.2707 (0.2870) loss 3.7105 (3.0631) grad_norm 2.5985 (2.8569) [2021-04-16 16:46:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][830/1251] eta 0:02:00 lr 0.000070 time 0.2763 (0.2868) loss 3.2887 (3.0639) grad_norm 2.8596 (2.8569) [2021-04-16 16:46:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][840/1251] eta 0:01:57 lr 0.000070 time 0.2747 (0.2867) loss 2.3367 (3.0605) grad_norm 3.0500 (2.8553) [2021-04-16 16:46:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][850/1251] eta 0:01:54 lr 0.000070 time 0.2822 (0.2865) loss 3.0892 (3.0610) grad_norm 2.9243 (2.8557) [2021-04-16 16:46:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][860/1251] eta 0:01:51 lr 0.000070 time 0.2687 (0.2864) loss 3.8464 (3.0632) grad_norm 2.8409 (2.8560) [2021-04-16 16:46:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][870/1251] eta 0:01:49 lr 0.000070 time 0.2711 (0.2862) loss 2.5867 (3.0641) grad_norm 2.3477 (2.8549) [2021-04-16 16:46:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][880/1251] eta 0:01:46 lr 0.000069 time 0.2757 (0.2861) loss 3.7723 (3.0623) grad_norm 2.7010 (2.8561) [2021-04-16 16:46:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][890/1251] eta 0:01:43 lr 0.000069 time 0.2774 (0.2861) loss 2.5339 (3.0640) grad_norm 2.8702 (2.8596) [2021-04-16 16:46:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][900/1251] eta 0:01:40 lr 0.000069 time 0.2503 (0.2860) loss 2.0724 (3.0638) grad_norm 2.8428 (2.8614) [2021-04-16 16:46:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][910/1251] eta 0:01:37 lr 0.000069 time 0.2976 (0.2860) loss 3.6157 (3.0655) grad_norm 3.1006 (2.8595) [2021-04-16 16:46:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][920/1251] eta 0:01:34 lr 0.000069 time 0.3905 (0.2861) loss 3.5667 (3.0642) grad_norm 2.6259 (2.8589) [2021-04-16 16:46:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][930/1251] eta 0:01:31 lr 0.000069 time 0.2550 (0.2860) loss 3.4536 (3.0637) grad_norm 2.7041 (2.8583) [2021-04-16 16:46:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][940/1251] eta 0:01:28 lr 0.000069 time 0.2824 (0.2861) loss 3.4387 (3.0607) grad_norm 3.0210 (2.8582) [2021-04-16 16:46:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][950/1251] eta 0:01:26 lr 0.000069 time 0.2654 (0.2859) loss 3.5691 (3.0650) grad_norm 2.7829 (2.8580) [2021-04-16 16:46:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][960/1251] eta 0:01:23 lr 0.000069 time 0.2509 (0.2858) loss 2.7993 (3.0653) grad_norm 2.6487 (2.8575) [2021-04-16 16:46:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][970/1251] eta 0:01:20 lr 0.000069 time 0.2744 (0.2858) loss 3.6477 (3.0633) grad_norm 3.1744 (2.8555) [2021-04-16 16:46:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][980/1251] eta 0:01:17 lr 0.000069 time 0.2756 (0.2858) loss 3.2859 (3.0640) grad_norm 3.0597 (2.8543) [2021-04-16 16:47:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][990/1251] eta 0:01:14 lr 0.000069 time 0.2979 (0.2858) loss 3.5233 (3.0662) grad_norm 2.8081 (2.8542) [2021-04-16 16:47:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][1000/1251] eta 0:01:11 lr 0.000069 time 0.2653 (0.2857) loss 2.7589 (3.0668) grad_norm 2.5349 (2.8543) [2021-04-16 16:47:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][1010/1251] eta 0:01:08 lr 0.000069 time 0.2894 (0.2856) loss 2.9521 (3.0654) grad_norm 2.6393 (2.8536) [2021-04-16 16:47:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][1020/1251] eta 0:01:05 lr 0.000069 time 0.2864 (0.2855) loss 3.6481 (3.0674) grad_norm 2.5488 (2.8518) [2021-04-16 16:47:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][1030/1251] eta 0:01:03 lr 0.000069 time 0.3027 (0.2855) loss 2.3918 (3.0637) grad_norm 3.3979 (2.8523) [2021-04-16 16:47:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][1040/1251] eta 0:01:00 lr 0.000069 time 0.2974 (0.2855) loss 3.4907 (3.0654) grad_norm 3.0739 (2.8554) [2021-04-16 16:47:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][1050/1251] eta 0:00:57 lr 0.000069 time 0.2711 (0.2854) loss 2.7614 (3.0666) grad_norm 3.0141 (2.8563) [2021-04-16 16:47:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][1060/1251] eta 0:00:54 lr 0.000069 time 0.2930 (0.2853) loss 2.4842 (3.0650) grad_norm 2.5394 (2.8549) [2021-04-16 16:47:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][1070/1251] eta 0:00:51 lr 0.000069 time 0.2618 (0.2852) loss 2.2336 (3.0653) grad_norm 2.4149 (2.8550) [2021-04-16 16:47:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][1080/1251] eta 0:00:48 lr 0.000069 time 0.2774 (0.2851) loss 3.3837 (3.0625) grad_norm 3.0126 (2.8553) [2021-04-16 16:47:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][1090/1251] eta 0:00:45 lr 0.000069 time 0.2704 (0.2852) loss 3.7645 (3.0638) grad_norm 2.3155 (2.8539) [2021-04-16 16:47:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][1100/1251] eta 0:00:43 lr 0.000069 time 0.2728 (0.2851) loss 2.7972 (3.0634) grad_norm 3.3074 (2.8535) [2021-04-16 16:47:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][1110/1251] eta 0:00:40 lr 0.000069 time 0.2735 (0.2851) loss 3.4755 (3.0643) grad_norm 2.7161 (2.8531) [2021-04-16 16:47:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][1120/1251] eta 0:00:37 lr 0.000069 time 0.2915 (0.2850) loss 3.2756 (3.0652) grad_norm 2.8396 (2.8537) [2021-04-16 16:47:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][1130/1251] eta 0:00:34 lr 0.000069 time 0.2669 (0.2852) loss 2.2801 (3.0661) grad_norm 3.0200 (2.8532) [2021-04-16 16:47:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][1140/1251] eta 0:00:31 lr 0.000069 time 0.2782 (0.2851) loss 3.0099 (3.0654) grad_norm 2.8585 (2.8529) [2021-04-16 16:47:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][1150/1251] eta 0:00:28 lr 0.000069 time 0.2684 (0.2852) loss 2.2967 (3.0642) grad_norm 3.0014 (2.8540) [2021-04-16 16:47:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][1160/1251] eta 0:00:25 lr 0.000069 time 0.2602 (0.2852) loss 3.2599 (3.0638) grad_norm 2.4594 (2.8536) [2021-04-16 16:47:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][1170/1251] eta 0:00:23 lr 0.000069 time 0.2844 (0.2851) loss 3.4444 (3.0638) grad_norm 2.7680 (2.8532) [2021-04-16 16:47:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][1180/1251] eta 0:00:20 lr 0.000069 time 0.3051 (0.2851) loss 3.4426 (3.0641) grad_norm 2.7346 (2.8538) [2021-04-16 16:47:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][1190/1251] eta 0:00:17 lr 0.000069 time 0.2717 (0.2851) loss 3.1838 (3.0649) grad_norm 2.4398 (2.8554) [2021-04-16 16:48:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][1200/1251] eta 0:00:14 lr 0.000069 time 0.2754 (0.2850) loss 3.5182 (3.0650) grad_norm 2.6233 (2.8529) [2021-04-16 16:48:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][1210/1251] eta 0:00:11 lr 0.000069 time 0.2739 (0.2849) loss 3.2375 (3.0665) grad_norm 2.7242 (2.8524) [2021-04-16 16:48:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][1220/1251] eta 0:00:08 lr 0.000069 time 0.2681 (0.2848) loss 3.1807 (3.0655) grad_norm 2.5506 (2.8527) [2021-04-16 16:48:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][1230/1251] eta 0:00:05 lr 0.000069 time 0.2569 (0.2849) loss 3.3051 (3.0656) grad_norm 2.4469 (2.8518) [2021-04-16 16:48:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][1240/1251] eta 0:00:03 lr 0.000069 time 0.2493 (0.2848) loss 3.7024 (3.0685) grad_norm 2.6419 (2.8522) [2021-04-16 16:48:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [252/300][1250/1251] eta 0:00:00 lr 0.000069 time 0.2482 (0.2845) loss 3.2516 (3.0692) grad_norm 2.6151 (2.8514) [2021-04-16 16:48:20 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 252 training takes 0:06:01 [2021-04-16 16:48:20 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_252.pth saving...... [2021-04-16 16:48:37 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_252.pth saved !!! [2021-04-16 16:48:38 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.221 (1.221) Loss 0.8181 (0.8181) Acc@1 80.176 (80.176) Acc@5 95.215 (95.215) [2021-04-16 16:48:40 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.113 (0.225) Loss 0.8725 (0.8425) Acc@1 79.590 (80.220) Acc@5 95.703 (95.099) [2021-04-16 16:48:42 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.149 (0.225) Loss 0.8334 (0.8335) Acc@1 79.883 (80.506) Acc@5 95.898 (95.178) [2021-04-16 16:48:44 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.134 (0.238) Loss 0.8349 (0.8351) Acc@1 79.590 (80.406) Acc@5 95.605 (95.224) [2021-04-16 16:48:46 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.215) Loss 0.8638 (0.8356) Acc@1 79.980 (80.376) Acc@5 94.434 (95.243) [2021-04-16 16:48:56 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.456 Acc@5 95.262 [2021-04-16 16:48:56 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 80.5% [2021-04-16 16:48:56 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 80.46% [2021-04-16 16:49:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][0/1251] eta 3:02:14 lr 0.000069 time 8.7403 (8.7403) loss 3.2675 (3.2675) grad_norm 2.7212 (2.7212) [2021-04-16 16:49:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][10/1251] eta 0:21:28 lr 0.000069 time 0.2829 (1.0385) loss 2.2358 (3.1079) grad_norm 2.5413 (2.9459) [2021-04-16 16:49:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][20/1251] eta 0:13:57 lr 0.000069 time 0.2625 (0.6802) loss 2.7607 (3.2391) grad_norm 2.6366 (2.8945) [2021-04-16 16:49:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][30/1251] eta 0:11:13 lr 0.000069 time 0.2806 (0.5514) loss 3.3160 (3.2261) grad_norm 3.6067 (2.9167) [2021-04-16 16:49:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][40/1251] eta 0:09:44 lr 0.000069 time 0.2634 (0.4829) loss 3.4084 (3.2521) grad_norm 3.0565 (2.9541) [2021-04-16 16:49:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][50/1251] eta 0:08:51 lr 0.000069 time 0.2630 (0.4423) loss 3.6240 (3.2221) grad_norm 3.2226 (2.9394) [2021-04-16 16:49:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][60/1251] eta 0:08:15 lr 0.000069 time 0.2754 (0.4162) loss 3.7185 (3.2211) grad_norm 3.5023 (2.9534) [2021-04-16 16:49:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][70/1251] eta 0:07:49 lr 0.000069 time 0.2847 (0.3971) loss 3.4512 (3.2006) grad_norm 3.4335 (2.9241) [2021-04-16 16:49:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][80/1251] eta 0:07:28 lr 0.000069 time 0.3060 (0.3831) loss 2.9616 (3.1926) grad_norm 2.6503 (2.9430) [2021-04-16 16:49:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][90/1251] eta 0:07:11 lr 0.000069 time 0.2683 (0.3719) loss 3.1894 (3.1846) grad_norm 3.3798 (2.9357) [2021-04-16 16:49:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][100/1251] eta 0:06:57 lr 0.000069 time 0.2604 (0.3624) loss 3.6422 (3.1852) grad_norm 2.6036 (2.9213) [2021-04-16 16:49:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][110/1251] eta 0:06:45 lr 0.000069 time 0.2893 (0.3551) loss 2.1208 (3.1614) grad_norm 3.0377 (2.9023) [2021-04-16 16:49:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][120/1251] eta 0:06:35 lr 0.000069 time 0.2748 (0.3495) loss 3.4845 (3.1489) grad_norm 3.0840 (2.8838) [2021-04-16 16:49:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][130/1251] eta 0:06:26 lr 0.000069 time 0.4472 (0.3449) loss 3.2115 (3.1468) grad_norm 2.7556 (2.8606) [2021-04-16 16:49:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][140/1251] eta 0:06:18 lr 0.000068 time 0.2988 (0.3408) loss 3.1299 (3.1449) grad_norm 2.3111 (2.8571) [2021-04-16 16:49:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][150/1251] eta 0:06:10 lr 0.000068 time 0.2452 (0.3364) loss 2.5007 (3.1404) grad_norm 2.6267 (2.8443) [2021-04-16 16:49:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][160/1251] eta 0:06:03 lr 0.000068 time 0.2622 (0.3334) loss 3.0162 (3.1523) grad_norm 3.1351 (2.8572) [2021-04-16 16:49:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][170/1251] eta 0:05:57 lr 0.000068 time 0.2920 (0.3305) loss 3.2007 (3.1479) grad_norm 2.6696 (2.8510) [2021-04-16 16:49:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][180/1251] eta 0:05:51 lr 0.000068 time 0.2696 (0.3279) loss 3.4384 (3.1461) grad_norm 3.0954 (2.8589) [2021-04-16 16:49:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][190/1251] eta 0:05:45 lr 0.000068 time 0.2630 (0.3253) loss 3.2311 (3.1495) grad_norm 2.7641 (2.8598) [2021-04-16 16:50:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][200/1251] eta 0:05:39 lr 0.000068 time 0.2892 (0.3229) loss 3.1427 (3.1400) grad_norm 2.8334 (2.8523) [2021-04-16 16:50:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][210/1251] eta 0:05:33 lr 0.000068 time 0.2893 (0.3208) loss 3.5071 (3.1439) grad_norm 3.0918 (2.8596) [2021-04-16 16:50:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][220/1251] eta 0:05:28 lr 0.000068 time 0.2529 (0.3188) loss 3.1148 (3.1445) grad_norm 2.5076 (2.8588) [2021-04-16 16:50:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][230/1251] eta 0:05:23 lr 0.000068 time 0.2671 (0.3168) loss 2.4623 (3.1329) grad_norm 2.7826 (2.8648) [2021-04-16 16:50:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][240/1251] eta 0:05:18 lr 0.000068 time 0.2834 (0.3150) loss 3.3133 (3.1344) grad_norm 2.8041 (2.8645) [2021-04-16 16:50:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][250/1251] eta 0:05:13 lr 0.000068 time 0.2446 (0.3133) loss 2.1741 (3.1273) grad_norm 2.6112 (2.8677) [2021-04-16 16:50:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][260/1251] eta 0:05:09 lr 0.000068 time 0.2757 (0.3121) loss 2.5372 (3.1250) grad_norm 2.6766 (2.8730) [2021-04-16 16:50:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][270/1251] eta 0:05:04 lr 0.000068 time 0.2945 (0.3107) loss 2.0135 (3.1117) grad_norm 2.9557 (2.8778) [2021-04-16 16:50:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][280/1251] eta 0:05:00 lr 0.000068 time 0.2547 (0.3096) loss 3.3510 (3.1137) grad_norm 3.0575 (2.8752) [2021-04-16 16:50:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][290/1251] eta 0:04:56 lr 0.000068 time 0.3236 (0.3089) loss 3.2448 (3.1165) grad_norm 2.8922 (2.8754) [2021-04-16 16:50:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][300/1251] eta 0:04:53 lr 0.000068 time 0.2937 (0.3082) loss 3.3058 (3.1142) grad_norm 3.1392 (2.8700) [2021-04-16 16:50:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][310/1251] eta 0:04:48 lr 0.000068 time 0.2563 (0.3070) loss 2.7867 (3.1165) grad_norm 2.4865 (2.8678) [2021-04-16 16:50:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][320/1251] eta 0:04:44 lr 0.000068 time 0.2709 (0.3060) loss 2.9308 (3.1203) grad_norm 3.8578 (2.8715) [2021-04-16 16:50:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][330/1251] eta 0:04:41 lr 0.000068 time 0.2587 (0.3054) loss 3.6021 (3.1241) grad_norm 2.9849 (2.8793) [2021-04-16 16:50:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][340/1251] eta 0:04:37 lr 0.000068 time 0.2862 (0.3046) loss 3.4601 (3.1130) grad_norm 2.6238 (2.8758) [2021-04-16 16:50:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][350/1251] eta 0:04:34 lr 0.000068 time 0.2983 (0.3045) loss 3.6796 (3.1187) grad_norm 2.8210 (2.8758) [2021-04-16 16:50:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][360/1251] eta 0:04:31 lr 0.000068 time 0.2832 (0.3042) loss 2.9355 (3.1196) grad_norm 2.4628 (2.8744) [2021-04-16 16:50:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][370/1251] eta 0:04:27 lr 0.000068 time 0.2611 (0.3037) loss 2.3223 (3.1188) grad_norm 2.5262 (2.8753) [2021-04-16 16:50:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][380/1251] eta 0:04:23 lr 0.000068 time 0.2666 (0.3029) loss 3.6148 (3.1233) grad_norm 2.8964 (2.8746) [2021-04-16 16:50:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][390/1251] eta 0:04:20 lr 0.000068 time 0.3024 (0.3022) loss 3.2645 (3.1235) grad_norm 2.6834 (2.8758) [2021-04-16 16:50:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][400/1251] eta 0:04:16 lr 0.000068 time 0.2532 (0.3015) loss 3.2270 (3.1295) grad_norm 2.1974 (2.8690) [2021-04-16 16:50:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][410/1251] eta 0:04:12 lr 0.000068 time 0.2814 (0.3008) loss 2.4144 (3.1190) grad_norm 3.2383 (2.8675) [2021-04-16 16:51:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][420/1251] eta 0:04:09 lr 0.000068 time 0.2680 (0.3002) loss 2.3432 (3.1161) grad_norm 2.5291 (2.8634) [2021-04-16 16:51:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][430/1251] eta 0:04:06 lr 0.000068 time 0.2763 (0.2998) loss 2.5415 (3.1146) grad_norm 2.6791 (2.8600) [2021-04-16 16:51:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][440/1251] eta 0:04:02 lr 0.000068 time 0.2763 (0.2993) loss 2.5589 (3.1120) grad_norm 2.7235 (2.8548) [2021-04-16 16:51:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][450/1251] eta 0:03:59 lr 0.000068 time 0.2976 (0.2988) loss 3.5606 (3.1163) grad_norm 2.9777 (2.8576) [2021-04-16 16:51:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][460/1251] eta 0:03:55 lr 0.000068 time 0.2522 (0.2981) loss 3.2231 (3.1120) grad_norm 2.5832 (2.8626) [2021-04-16 16:51:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][470/1251] eta 0:03:52 lr 0.000068 time 0.2726 (0.2980) loss 3.2210 (3.1101) grad_norm 2.7353 (2.8609) [2021-04-16 16:51:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][480/1251] eta 0:03:49 lr 0.000068 time 0.2567 (0.2979) loss 3.4558 (3.1092) grad_norm 2.6072 (2.8587) [2021-04-16 16:51:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][490/1251] eta 0:03:46 lr 0.000068 time 0.2553 (0.2973) loss 2.8976 (3.0970) grad_norm 3.2333 (2.8554) [2021-04-16 16:51:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][500/1251] eta 0:03:43 lr 0.000068 time 0.2885 (0.2970) loss 2.9886 (3.0971) grad_norm 2.3593 (2.8555) [2021-04-16 16:51:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][510/1251] eta 0:03:39 lr 0.000068 time 0.2761 (0.2966) loss 2.5410 (3.0987) grad_norm 2.7325 (2.8518) [2021-04-16 16:51:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][520/1251] eta 0:03:36 lr 0.000068 time 0.2678 (0.2964) loss 3.5661 (3.1022) grad_norm 3.0956 (2.8479) [2021-04-16 16:51:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][530/1251] eta 0:03:33 lr 0.000068 time 0.3100 (0.2959) loss 3.5881 (3.1059) grad_norm 2.5882 (inf) [2021-04-16 16:51:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][540/1251] eta 0:03:30 lr 0.000068 time 0.2673 (0.2956) loss 3.3957 (3.1033) grad_norm 2.4405 (inf) [2021-04-16 16:51:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][550/1251] eta 0:03:26 lr 0.000068 time 0.2545 (0.2951) loss 2.8729 (3.1013) grad_norm 2.8310 (inf) [2021-04-16 16:51:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][560/1251] eta 0:03:23 lr 0.000068 time 0.2844 (0.2948) loss 3.5586 (3.1006) grad_norm 2.9295 (inf) [2021-04-16 16:51:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][570/1251] eta 0:03:20 lr 0.000068 time 0.2717 (0.2946) loss 2.7978 (3.0994) grad_norm 2.2398 (inf) [2021-04-16 16:51:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][580/1251] eta 0:03:17 lr 0.000068 time 0.2903 (0.2946) loss 3.3235 (3.0952) grad_norm 2.9624 (inf) [2021-04-16 16:51:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][590/1251] eta 0:03:14 lr 0.000068 time 0.2890 (0.2943) loss 2.7029 (3.0966) grad_norm 2.5994 (inf) [2021-04-16 16:51:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][600/1251] eta 0:03:11 lr 0.000068 time 0.2719 (0.2939) loss 2.3806 (3.0922) grad_norm 3.0521 (inf) [2021-04-16 16:51:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][610/1251] eta 0:03:08 lr 0.000068 time 0.2732 (0.2937) loss 3.6449 (3.0932) grad_norm 2.6496 (inf) [2021-04-16 16:51:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][620/1251] eta 0:03:05 lr 0.000068 time 0.2916 (0.2935) loss 3.4556 (3.0948) grad_norm 3.2668 (inf) [2021-04-16 16:52:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][630/1251] eta 0:03:02 lr 0.000068 time 0.2719 (0.2932) loss 2.2472 (3.0944) grad_norm 2.6885 (inf) [2021-04-16 16:52:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][640/1251] eta 0:02:58 lr 0.000068 time 0.2578 (0.2929) loss 3.0002 (3.0930) grad_norm 3.0998 (inf) [2021-04-16 16:52:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][650/1251] eta 0:02:55 lr 0.000067 time 0.2722 (0.2927) loss 2.7407 (3.0923) grad_norm 3.1717 (inf) [2021-04-16 16:52:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][660/1251] eta 0:02:52 lr 0.000067 time 0.2720 (0.2924) loss 2.5249 (3.0889) grad_norm 2.6068 (inf) [2021-04-16 16:52:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][670/1251] eta 0:02:49 lr 0.000067 time 0.3012 (0.2923) loss 2.4570 (3.0891) grad_norm 2.8849 (inf) [2021-04-16 16:52:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][680/1251] eta 0:02:46 lr 0.000067 time 0.2791 (0.2921) loss 1.8501 (3.0841) grad_norm 2.4425 (inf) [2021-04-16 16:52:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][690/1251] eta 0:02:43 lr 0.000067 time 0.2720 (0.2920) loss 3.2229 (3.0822) grad_norm 2.6180 (inf) [2021-04-16 16:52:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][700/1251] eta 0:02:40 lr 0.000067 time 0.2694 (0.2917) loss 3.3405 (3.0845) grad_norm 2.6150 (inf) [2021-04-16 16:52:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][710/1251] eta 0:02:37 lr 0.000067 time 0.2749 (0.2916) loss 3.1857 (3.0884) grad_norm 2.7874 (inf) [2021-04-16 16:52:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][720/1251] eta 0:02:34 lr 0.000067 time 0.2619 (0.2915) loss 3.4673 (3.0889) grad_norm 3.0899 (inf) [2021-04-16 16:52:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][730/1251] eta 0:02:31 lr 0.000067 time 0.4536 (0.2915) loss 3.6821 (3.0921) grad_norm 2.7808 (inf) [2021-04-16 16:52:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][740/1251] eta 0:02:28 lr 0.000067 time 0.2736 (0.2914) loss 3.5226 (3.0875) grad_norm 2.4187 (inf) [2021-04-16 16:52:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][750/1251] eta 0:02:25 lr 0.000067 time 0.2544 (0.2912) loss 2.5798 (3.0871) grad_norm 2.7556 (inf) [2021-04-16 16:52:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][760/1251] eta 0:02:22 lr 0.000067 time 0.2743 (0.2912) loss 3.2022 (3.0879) grad_norm 2.5243 (inf) [2021-04-16 16:52:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][770/1251] eta 0:02:19 lr 0.000067 time 0.2626 (0.2909) loss 3.1432 (3.0875) grad_norm 2.4585 (inf) [2021-04-16 16:52:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][780/1251] eta 0:02:16 lr 0.000067 time 0.2908 (0.2907) loss 3.5567 (3.0910) grad_norm 3.6743 (inf) [2021-04-16 16:52:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][790/1251] eta 0:02:13 lr 0.000067 time 0.2902 (0.2905) loss 3.3456 (3.0939) grad_norm 2.9290 (inf) [2021-04-16 16:52:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][800/1251] eta 0:02:10 lr 0.000067 time 0.2576 (0.2904) loss 3.3582 (3.0900) grad_norm 2.5879 (inf) [2021-04-16 16:52:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][810/1251] eta 0:02:08 lr 0.000067 time 0.2827 (0.2903) loss 2.3320 (3.0889) grad_norm 2.3186 (inf) [2021-04-16 16:52:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][820/1251] eta 0:02:05 lr 0.000067 time 0.2789 (0.2901) loss 2.0621 (3.0873) grad_norm 2.8295 (inf) [2021-04-16 16:52:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][830/1251] eta 0:02:02 lr 0.000067 time 0.2780 (0.2900) loss 2.3082 (3.0847) grad_norm 2.8439 (inf) [2021-04-16 16:52:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][840/1251] eta 0:01:59 lr 0.000067 time 0.2749 (0.2898) loss 3.0482 (3.0872) grad_norm 2.5705 (inf) [2021-04-16 16:53:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][850/1251] eta 0:01:56 lr 0.000067 time 0.2836 (0.2896) loss 3.0669 (3.0879) grad_norm 3.1228 (inf) [2021-04-16 16:53:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][860/1251] eta 0:01:53 lr 0.000067 time 0.2671 (0.2895) loss 2.3588 (3.0874) grad_norm 2.6935 (inf) [2021-04-16 16:53:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][870/1251] eta 0:01:50 lr 0.000067 time 0.2567 (0.2893) loss 2.7597 (3.0866) grad_norm 3.0093 (inf) [2021-04-16 16:53:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][880/1251] eta 0:01:47 lr 0.000067 time 0.2720 (0.2893) loss 3.5189 (3.0838) grad_norm 2.9779 (inf) [2021-04-16 16:53:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][890/1251] eta 0:01:44 lr 0.000067 time 0.2701 (0.2892) loss 3.6509 (3.0835) grad_norm 2.6591 (inf) [2021-04-16 16:53:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][900/1251] eta 0:01:41 lr 0.000067 time 0.2787 (0.2890) loss 3.0406 (3.0879) grad_norm 2.4966 (inf) [2021-04-16 16:53:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][910/1251] eta 0:01:38 lr 0.000067 time 0.2772 (0.2890) loss 3.0585 (3.0893) grad_norm 2.8685 (inf) [2021-04-16 16:53:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][920/1251] eta 0:01:35 lr 0.000067 time 0.3110 (0.2890) loss 3.1521 (3.0890) grad_norm 3.1544 (inf) [2021-04-16 16:53:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][930/1251] eta 0:01:32 lr 0.000067 time 0.2855 (0.2889) loss 3.0142 (3.0892) grad_norm 3.0571 (inf) [2021-04-16 16:53:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][940/1251] eta 0:01:29 lr 0.000067 time 0.2747 (0.2888) loss 2.6922 (3.0897) grad_norm 2.6238 (inf) [2021-04-16 16:53:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][950/1251] eta 0:01:26 lr 0.000067 time 0.2531 (0.2889) loss 3.1185 (3.0908) grad_norm 3.2963 (inf) [2021-04-16 16:53:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][960/1251] eta 0:01:24 lr 0.000067 time 0.2716 (0.2887) loss 3.3377 (3.0931) grad_norm 2.2748 (inf) [2021-04-16 16:53:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][970/1251] eta 0:01:21 lr 0.000067 time 0.2871 (0.2887) loss 3.3005 (3.0918) grad_norm 2.6267 (inf) [2021-04-16 16:53:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][980/1251] eta 0:01:18 lr 0.000067 time 0.3066 (0.2886) loss 2.4950 (3.0931) grad_norm 2.7699 (inf) [2021-04-16 16:53:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][990/1251] eta 0:01:15 lr 0.000067 time 0.2952 (0.2885) loss 2.9814 (3.0944) grad_norm 3.0014 (inf) [2021-04-16 16:53:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][1000/1251] eta 0:01:12 lr 0.000067 time 0.2658 (0.2883) loss 3.1168 (3.0936) grad_norm 2.8521 (inf) [2021-04-16 16:53:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][1010/1251] eta 0:01:09 lr 0.000067 time 0.2610 (0.2882) loss 3.2094 (3.0946) grad_norm 2.9034 (inf) [2021-04-16 16:53:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][1020/1251] eta 0:01:06 lr 0.000067 time 0.2675 (0.2881) loss 3.3555 (3.0968) grad_norm 2.2249 (inf) [2021-04-16 16:53:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][1030/1251] eta 0:01:03 lr 0.000067 time 0.2788 (0.2880) loss 2.5064 (3.0972) grad_norm 2.7792 (inf) [2021-04-16 16:53:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][1040/1251] eta 0:01:00 lr 0.000067 time 0.2648 (0.2879) loss 2.0333 (3.0951) grad_norm 3.8778 (inf) [2021-04-16 16:53:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][1050/1251] eta 0:00:57 lr 0.000067 time 0.4067 (0.2880) loss 3.7093 (3.0951) grad_norm 3.1451 (inf) [2021-04-16 16:54:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][1060/1251] eta 0:00:54 lr 0.000067 time 0.2845 (0.2878) loss 2.8265 (3.0936) grad_norm 2.6992 (inf) [2021-04-16 16:54:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][1070/1251] eta 0:00:52 lr 0.000067 time 0.2672 (0.2877) loss 3.1038 (3.0924) grad_norm 2.9593 (inf) [2021-04-16 16:54:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][1080/1251] eta 0:00:49 lr 0.000067 time 0.2715 (0.2876) loss 3.0505 (3.0950) grad_norm 3.1987 (inf) [2021-04-16 16:54:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][1090/1251] eta 0:00:46 lr 0.000067 time 0.2638 (0.2875) loss 2.9564 (3.0916) grad_norm 3.0282 (inf) [2021-04-16 16:54:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][1100/1251] eta 0:00:43 lr 0.000067 time 0.2570 (0.2874) loss 3.5122 (3.0932) grad_norm 3.1103 (inf) [2021-04-16 16:54:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][1110/1251] eta 0:00:40 lr 0.000067 time 0.2697 (0.2873) loss 3.5242 (3.0910) grad_norm 3.3064 (inf) [2021-04-16 16:54:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][1120/1251] eta 0:00:37 lr 0.000067 time 0.2792 (0.2872) loss 2.7968 (3.0905) grad_norm 3.0175 (inf) [2021-04-16 16:54:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][1130/1251] eta 0:00:34 lr 0.000067 time 0.2492 (0.2871) loss 2.6999 (3.0899) grad_norm 2.8630 (inf) [2021-04-16 16:54:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][1140/1251] eta 0:00:31 lr 0.000067 time 0.2632 (0.2870) loss 3.4867 (3.0890) grad_norm 2.9279 (inf) [2021-04-16 16:54:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][1150/1251] eta 0:00:28 lr 0.000067 time 0.2870 (0.2870) loss 3.5437 (3.0892) grad_norm 3.2656 (inf) [2021-04-16 16:54:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][1160/1251] eta 0:00:26 lr 0.000067 time 0.2824 (0.2870) loss 3.0655 (3.0902) grad_norm 3.3079 (inf) [2021-04-16 16:54:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][1170/1251] eta 0:00:23 lr 0.000066 time 0.2787 (0.2869) loss 2.3836 (3.0894) grad_norm 2.7566 (inf) [2021-04-16 16:54:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][1180/1251] eta 0:00:20 lr 0.000066 time 0.2546 (0.2868) loss 3.3300 (3.0885) grad_norm 2.8975 (inf) [2021-04-16 16:54:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][1190/1251] eta 0:00:17 lr 0.000066 time 0.2768 (0.2867) loss 3.3721 (3.0903) grad_norm 2.9435 (inf) [2021-04-16 16:54:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][1200/1251] eta 0:00:14 lr 0.000066 time 0.2655 (0.2867) loss 2.9675 (3.0904) grad_norm 2.1144 (inf) [2021-04-16 16:54:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][1210/1251] eta 0:00:11 lr 0.000066 time 0.2872 (0.2866) loss 3.5165 (3.0900) grad_norm 2.9606 (inf) [2021-04-16 16:54:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][1220/1251] eta 0:00:08 lr 0.000066 time 0.2779 (0.2865) loss 3.3436 (3.0883) grad_norm 3.8918 (inf) [2021-04-16 16:54:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][1230/1251] eta 0:00:06 lr 0.000066 time 0.2607 (0.2864) loss 2.8963 (3.0869) grad_norm 2.7328 (inf) [2021-04-16 16:54:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][1240/1251] eta 0:00:03 lr 0.000066 time 0.3213 (0.2863) loss 3.5204 (3.0868) grad_norm 2.7034 (inf) [2021-04-16 16:54:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [253/300][1250/1251] eta 0:00:00 lr 0.000066 time 0.2500 (0.2860) loss 3.1810 (3.0877) grad_norm 3.0645 (inf) [2021-04-16 16:55:07 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 253 training takes 0:06:11 [2021-04-16 16:55:07 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_253.pth saving...... [2021-04-16 16:55:18 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_253.pth saved !!! [2021-04-16 16:55:20 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.128 (1.128) Loss 0.7988 (0.7988) Acc@1 81.445 (81.445) Acc@5 95.605 (95.605) [2021-04-16 16:55:21 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.423 (0.272) Loss 0.7817 (0.8299) Acc@1 80.762 (80.682) Acc@5 95.410 (95.295) [2021-04-16 16:55:23 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.106 (0.200) Loss 0.8327 (0.8377) Acc@1 81.348 (80.483) Acc@5 95.117 (95.117) [2021-04-16 16:55:25 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.128 (0.214) Loss 0.8268 (0.8356) Acc@1 81.738 (80.336) Acc@5 95.312 (95.275) [2021-04-16 16:55:27 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.215) Loss 0.8194 (0.8317) Acc@1 80.371 (80.414) Acc@5 95.410 (95.310) [2021-04-16 16:55:46 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.494 Acc@5 95.354 [2021-04-16 16:55:46 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 80.5% [2021-04-16 16:55:46 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 80.49% [2021-04-16 16:55:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][0/1251] eta 1:46:06 lr 0.000066 time 5.0893 (5.0893) loss 3.4786 (3.4786) grad_norm 3.2626 (3.2626) [2021-04-16 16:55:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][10/1251] eta 0:14:43 lr 0.000066 time 0.2755 (0.7120) loss 3.5221 (3.0063) grad_norm 4.2033 (3.3626) [2021-04-16 16:55:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][20/1251] eta 0:10:18 lr 0.000066 time 0.2651 (0.5022) loss 3.0535 (3.0202) grad_norm 3.0221 (3.1275) [2021-04-16 16:56:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][30/1251] eta 0:08:42 lr 0.000066 time 0.2533 (0.4280) loss 2.5456 (2.9536) grad_norm 4.0388 (3.0774) [2021-04-16 16:56:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][40/1251] eta 0:07:53 lr 0.000066 time 0.2775 (0.3913) loss 3.7529 (3.0099) grad_norm 2.9683 (3.0271) [2021-04-16 16:56:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][50/1251] eta 0:07:22 lr 0.000066 time 0.2693 (0.3689) loss 3.2903 (3.0416) grad_norm 2.5249 (2.9649) [2021-04-16 16:56:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][60/1251] eta 0:07:00 lr 0.000066 time 0.2801 (0.3529) loss 3.1185 (3.0241) grad_norm 3.1180 (2.9505) [2021-04-16 16:56:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][70/1251] eta 0:06:45 lr 0.000066 time 0.2927 (0.3437) loss 2.3877 (3.0182) grad_norm 2.7147 (2.9306) [2021-04-16 16:56:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][80/1251] eta 0:06:32 lr 0.000066 time 0.2725 (0.3355) loss 3.3717 (3.0064) grad_norm 2.5560 (2.9049) [2021-04-16 16:56:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][90/1251] eta 0:06:23 lr 0.000066 time 0.2619 (0.3305) loss 2.8964 (3.0073) grad_norm 2.9690 (2.9073) [2021-04-16 16:56:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][100/1251] eta 0:06:15 lr 0.000066 time 0.2709 (0.3266) loss 2.4144 (3.0126) grad_norm 2.6148 (2.9062) [2021-04-16 16:56:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][110/1251] eta 0:06:07 lr 0.000066 time 0.2620 (0.3219) loss 2.1210 (3.0175) grad_norm 2.9505 (2.9140) [2021-04-16 16:56:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][120/1251] eta 0:06:00 lr 0.000066 time 0.2629 (0.3184) loss 3.4704 (3.0371) grad_norm 2.7187 (2.9141) [2021-04-16 16:56:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][130/1251] eta 0:05:53 lr 0.000066 time 0.2827 (0.3153) loss 3.3405 (3.0554) grad_norm 2.7287 (2.9002) [2021-04-16 16:56:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][140/1251] eta 0:05:47 lr 0.000066 time 0.2790 (0.3124) loss 2.0381 (3.0561) grad_norm 2.9176 (2.8869) [2021-04-16 16:56:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][150/1251] eta 0:05:43 lr 0.000066 time 0.2838 (0.3117) loss 3.1366 (3.0749) grad_norm 2.3469 (2.8815) [2021-04-16 16:56:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][160/1251] eta 0:05:39 lr 0.000066 time 0.2595 (0.3109) loss 3.3113 (3.0756) grad_norm 2.7369 (2.8812) [2021-04-16 16:56:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][170/1251] eta 0:05:33 lr 0.000066 time 0.2627 (0.3090) loss 2.9001 (3.0804) grad_norm 2.9350 (2.8884) [2021-04-16 16:56:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][180/1251] eta 0:05:29 lr 0.000066 time 0.2800 (0.3074) loss 3.6830 (3.0790) grad_norm 2.6347 (2.8898) [2021-04-16 16:56:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][190/1251] eta 0:05:24 lr 0.000066 time 0.2900 (0.3056) loss 2.2457 (3.0767) grad_norm 2.8224 (2.8867) [2021-04-16 16:56:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][200/1251] eta 0:05:19 lr 0.000066 time 0.2722 (0.3040) loss 2.1618 (3.0851) grad_norm 3.1235 (2.8852) [2021-04-16 16:56:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][210/1251] eta 0:05:15 lr 0.000066 time 0.2828 (0.3026) loss 2.6587 (3.0840) grad_norm 2.6343 (2.8859) [2021-04-16 16:56:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][220/1251] eta 0:05:11 lr 0.000066 time 0.2873 (0.3022) loss 3.1033 (3.0889) grad_norm 3.6152 (2.8938) [2021-04-16 16:56:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][230/1251] eta 0:05:07 lr 0.000066 time 0.2546 (0.3009) loss 3.5886 (3.0861) grad_norm 2.5158 (2.8887) [2021-04-16 16:56:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][240/1251] eta 0:05:03 lr 0.000066 time 0.2765 (0.3004) loss 3.1057 (3.0883) grad_norm 2.6409 (2.8845) [2021-04-16 16:57:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][250/1251] eta 0:05:00 lr 0.000066 time 0.2563 (0.2998) loss 3.4199 (3.0865) grad_norm 2.6112 (2.8818) [2021-04-16 16:57:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][260/1251] eta 0:04:56 lr 0.000066 time 0.4725 (0.2997) loss 3.2489 (3.0815) grad_norm 2.8458 (2.8809) [2021-04-16 16:57:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][270/1251] eta 0:04:52 lr 0.000066 time 0.2833 (0.2985) loss 3.0111 (3.0802) grad_norm 2.8753 (2.8769) [2021-04-16 16:57:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][280/1251] eta 0:04:49 lr 0.000066 time 0.2825 (0.2979) loss 3.2667 (3.0867) grad_norm 2.9621 (2.8730) [2021-04-16 16:57:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][290/1251] eta 0:04:45 lr 0.000066 time 0.2885 (0.2972) loss 2.3129 (3.0876) grad_norm 2.7983 (2.8703) [2021-04-16 16:57:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][300/1251] eta 0:04:42 lr 0.000066 time 0.2687 (0.2966) loss 3.1279 (3.0836) grad_norm 2.8895 (2.8682) [2021-04-16 16:57:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][310/1251] eta 0:04:38 lr 0.000066 time 0.2899 (0.2961) loss 3.6072 (3.0845) grad_norm 2.8052 (2.8646) [2021-04-16 16:57:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][320/1251] eta 0:04:35 lr 0.000066 time 0.2720 (0.2955) loss 2.5082 (3.0793) grad_norm 2.9514 (2.8857) [2021-04-16 16:57:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][330/1251] eta 0:04:31 lr 0.000066 time 0.2456 (0.2948) loss 3.0320 (3.0838) grad_norm 2.8295 (2.8862) [2021-04-16 16:57:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][340/1251] eta 0:04:28 lr 0.000066 time 0.2611 (0.2942) loss 3.0094 (3.0842) grad_norm 2.7511 (2.8928) [2021-04-16 16:57:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][350/1251] eta 0:04:24 lr 0.000066 time 0.2771 (0.2937) loss 2.6762 (3.0834) grad_norm 3.7474 (2.8929) [2021-04-16 16:57:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][360/1251] eta 0:04:21 lr 0.000066 time 0.2606 (0.2940) loss 2.8391 (3.0825) grad_norm 2.6425 (2.8882) [2021-04-16 16:57:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][370/1251] eta 0:04:18 lr 0.000066 time 0.2832 (0.2934) loss 3.6660 (3.0802) grad_norm 2.5960 (2.8908) [2021-04-16 16:57:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][380/1251] eta 0:04:15 lr 0.000066 time 0.2677 (0.2929) loss 3.3572 (3.0832) grad_norm 2.3857 (2.8856) [2021-04-16 16:57:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][390/1251] eta 0:04:12 lr 0.000066 time 0.2761 (0.2927) loss 3.2362 (3.0847) grad_norm 2.4827 (2.8812) [2021-04-16 16:57:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][400/1251] eta 0:04:08 lr 0.000066 time 0.2954 (0.2924) loss 3.4258 (3.0896) grad_norm 3.4276 (2.8870) [2021-04-16 16:57:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][410/1251] eta 0:04:05 lr 0.000066 time 0.2957 (0.2921) loss 3.2363 (3.0890) grad_norm 3.3586 (2.8867) [2021-04-16 16:57:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][420/1251] eta 0:04:02 lr 0.000066 time 0.2797 (0.2917) loss 3.3507 (3.0830) grad_norm 2.7631 (2.8835) [2021-04-16 16:57:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][430/1251] eta 0:03:59 lr 0.000066 time 0.2894 (0.2918) loss 3.1965 (3.0834) grad_norm 3.0107 (2.8828) [2021-04-16 16:57:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][440/1251] eta 0:03:56 lr 0.000065 time 0.3137 (0.2915) loss 2.5127 (3.0831) grad_norm 2.4463 (2.8809) [2021-04-16 16:57:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][450/1251] eta 0:03:53 lr 0.000065 time 0.3305 (0.2912) loss 2.0066 (3.0831) grad_norm 2.8539 (2.8771) [2021-04-16 16:58:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][460/1251] eta 0:03:50 lr 0.000065 time 0.2656 (0.2908) loss 3.5035 (3.0885) grad_norm 2.9001 (2.8734) [2021-04-16 16:58:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][470/1251] eta 0:03:46 lr 0.000065 time 0.2714 (0.2905) loss 3.7705 (3.0907) grad_norm 5.6496 (2.8756) [2021-04-16 16:58:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][480/1251] eta 0:03:43 lr 0.000065 time 0.2970 (0.2903) loss 1.8881 (3.0788) grad_norm 2.5603 (2.8751) [2021-04-16 16:58:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][490/1251] eta 0:03:40 lr 0.000065 time 0.2586 (0.2902) loss 3.9460 (3.0810) grad_norm 2.8668 (2.8734) [2021-04-16 16:58:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][500/1251] eta 0:03:37 lr 0.000065 time 0.2644 (0.2899) loss 2.5692 (3.0823) grad_norm 2.6509 (2.8724) [2021-04-16 16:58:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][510/1251] eta 0:03:34 lr 0.000065 time 0.2623 (0.2896) loss 3.7729 (3.0821) grad_norm 2.7938 (2.8727) [2021-04-16 16:58:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][520/1251] eta 0:03:31 lr 0.000065 time 0.2756 (0.2894) loss 3.4364 (3.0828) grad_norm 2.8176 (2.8737) [2021-04-16 16:58:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][530/1251] eta 0:03:28 lr 0.000065 time 0.2779 (0.2891) loss 3.4329 (3.0801) grad_norm 2.6890 (2.8728) [2021-04-16 16:58:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][540/1251] eta 0:03:25 lr 0.000065 time 0.2538 (0.2890) loss 2.9043 (3.0793) grad_norm 2.7474 (2.8752) [2021-04-16 16:58:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][550/1251] eta 0:03:22 lr 0.000065 time 0.2519 (0.2887) loss 3.5314 (3.0809) grad_norm 3.2459 (2.8751) [2021-04-16 16:58:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][560/1251] eta 0:03:19 lr 0.000065 time 0.2571 (0.2884) loss 3.0142 (3.0827) grad_norm 2.8725 (2.8762) [2021-04-16 16:58:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][570/1251] eta 0:03:16 lr 0.000065 time 0.2685 (0.2882) loss 3.5603 (3.0842) grad_norm 2.8139 (2.8769) [2021-04-16 16:58:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][580/1251] eta 0:03:13 lr 0.000065 time 0.2629 (0.2884) loss 2.9222 (3.0806) grad_norm 2.6895 (2.8743) [2021-04-16 16:58:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][590/1251] eta 0:03:10 lr 0.000065 time 0.2768 (0.2881) loss 3.4230 (3.0818) grad_norm 2.7306 (2.8722) [2021-04-16 16:58:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][600/1251] eta 0:03:07 lr 0.000065 time 0.2884 (0.2879) loss 3.6200 (3.0877) grad_norm 2.5660 (2.8705) [2021-04-16 16:58:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][610/1251] eta 0:03:04 lr 0.000065 time 0.2804 (0.2877) loss 2.9799 (3.0891) grad_norm 2.4285 (2.8686) [2021-04-16 16:58:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][620/1251] eta 0:03:01 lr 0.000065 time 0.2690 (0.2876) loss 3.3398 (3.0899) grad_norm 2.9067 (2.8691) [2021-04-16 16:58:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][630/1251] eta 0:02:58 lr 0.000065 time 0.2642 (0.2873) loss 2.6526 (3.0838) grad_norm 3.1755 (2.8718) [2021-04-16 16:58:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][640/1251] eta 0:02:55 lr 0.000065 time 0.2428 (0.2872) loss 3.6564 (3.0836) grad_norm 2.7037 (2.8706) [2021-04-16 16:58:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][650/1251] eta 0:02:52 lr 0.000065 time 0.2772 (0.2871) loss 2.2470 (3.0837) grad_norm 2.4506 (2.8748) [2021-04-16 16:58:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][660/1251] eta 0:02:49 lr 0.000065 time 0.2873 (0.2868) loss 3.2684 (3.0847) grad_norm 3.2225 (2.8774) [2021-04-16 16:58:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][670/1251] eta 0:02:46 lr 0.000065 time 0.2477 (0.2869) loss 3.3424 (3.0809) grad_norm 2.7902 (2.8757) [2021-04-16 16:59:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][680/1251] eta 0:02:43 lr 0.000065 time 0.2857 (0.2869) loss 3.3304 (3.0800) grad_norm 2.6312 (2.8752) [2021-04-16 16:59:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][690/1251] eta 0:02:40 lr 0.000065 time 0.2502 (0.2867) loss 4.0682 (3.0811) grad_norm 3.0149 (2.8774) [2021-04-16 16:59:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][700/1251] eta 0:02:37 lr 0.000065 time 0.2629 (0.2866) loss 3.5540 (3.0829) grad_norm 2.9673 (2.8793) [2021-04-16 16:59:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][710/1251] eta 0:02:34 lr 0.000065 time 0.2713 (0.2865) loss 3.1366 (3.0792) grad_norm 2.8920 (2.8826) [2021-04-16 16:59:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][720/1251] eta 0:02:32 lr 0.000065 time 0.2446 (0.2866) loss 3.1988 (3.0799) grad_norm 2.6431 (2.8804) [2021-04-16 16:59:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][730/1251] eta 0:02:29 lr 0.000065 time 0.2679 (0.2864) loss 3.6682 (3.0803) grad_norm 2.8223 (2.8822) [2021-04-16 16:59:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][740/1251] eta 0:02:26 lr 0.000065 time 0.2927 (0.2864) loss 3.1375 (3.0793) grad_norm 3.0721 (2.8828) [2021-04-16 16:59:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][750/1251] eta 0:02:23 lr 0.000065 time 0.3030 (0.2863) loss 3.4460 (3.0815) grad_norm 2.7272 (2.8822) [2021-04-16 16:59:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][760/1251] eta 0:02:20 lr 0.000065 time 0.2886 (0.2862) loss 2.7124 (3.0840) grad_norm 2.5326 (2.8823) [2021-04-16 16:59:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][770/1251] eta 0:02:17 lr 0.000065 time 0.2587 (0.2860) loss 3.4189 (3.0847) grad_norm 3.0119 (2.8835) [2021-04-16 16:59:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][780/1251] eta 0:02:14 lr 0.000065 time 0.2741 (0.2859) loss 3.3157 (3.0846) grad_norm 2.9706 (2.8846) [2021-04-16 16:59:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][790/1251] eta 0:02:11 lr 0.000065 time 0.2854 (0.2860) loss 2.9481 (3.0864) grad_norm 2.7436 (2.8836) [2021-04-16 16:59:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][800/1251] eta 0:02:08 lr 0.000065 time 0.2734 (0.2859) loss 3.7789 (3.0817) grad_norm 3.0674 (2.8828) [2021-04-16 16:59:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][810/1251] eta 0:02:06 lr 0.000065 time 0.2573 (0.2858) loss 3.2559 (3.0810) grad_norm 2.9912 (2.8826) [2021-04-16 16:59:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][820/1251] eta 0:02:03 lr 0.000065 time 0.2712 (0.2858) loss 3.2334 (3.0776) grad_norm 3.5049 (2.8805) [2021-04-16 16:59:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][830/1251] eta 0:02:00 lr 0.000065 time 0.2671 (0.2858) loss 3.4835 (3.0799) grad_norm 2.5941 (2.8806) [2021-04-16 16:59:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][840/1251] eta 0:01:57 lr 0.000065 time 0.2872 (0.2858) loss 3.3038 (3.0803) grad_norm 2.9497 (2.8803) [2021-04-16 16:59:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][850/1251] eta 0:01:54 lr 0.000065 time 0.2877 (0.2856) loss 2.3449 (3.0793) grad_norm 3.2029 (2.8804) [2021-04-16 16:59:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][860/1251] eta 0:01:51 lr 0.000065 time 0.2815 (0.2856) loss 3.0633 (3.0761) grad_norm 2.6828 (2.8783) [2021-04-16 16:59:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][870/1251] eta 0:01:48 lr 0.000065 time 0.2518 (0.2855) loss 3.0665 (3.0760) grad_norm 2.9287 (2.8761) [2021-04-16 16:59:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][880/1251] eta 0:01:45 lr 0.000065 time 0.2584 (0.2853) loss 3.0735 (3.0771) grad_norm 3.7675 (2.8781) [2021-04-16 17:00:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][890/1251] eta 0:01:42 lr 0.000065 time 0.2866 (0.2852) loss 2.9203 (3.0765) grad_norm 3.3769 (2.8791) [2021-04-16 17:00:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][900/1251] eta 0:01:40 lr 0.000065 time 0.2629 (0.2851) loss 2.1202 (3.0747) grad_norm 2.9595 (2.8794) [2021-04-16 17:00:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][910/1251] eta 0:01:37 lr 0.000065 time 0.2837 (0.2851) loss 2.8473 (3.0735) grad_norm 2.8239 (2.8822) [2021-04-16 17:00:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][920/1251] eta 0:01:34 lr 0.000065 time 0.2875 (0.2849) loss 3.4628 (3.0740) grad_norm 3.1337 (2.8826) [2021-04-16 17:00:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][930/1251] eta 0:01:31 lr 0.000065 time 0.3063 (0.2850) loss 3.0555 (3.0717) grad_norm 2.5608 (2.8819) [2021-04-16 17:00:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][940/1251] eta 0:01:28 lr 0.000065 time 0.2672 (0.2851) loss 3.3548 (3.0722) grad_norm 2.9599 (2.8826) [2021-04-16 17:00:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][950/1251] eta 0:01:25 lr 0.000065 time 0.2539 (0.2850) loss 3.0163 (3.0710) grad_norm 3.0726 (2.8821) [2021-04-16 17:00:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][960/1251] eta 0:01:22 lr 0.000065 time 0.2998 (0.2849) loss 3.3753 (3.0705) grad_norm 2.6754 (2.8812) [2021-04-16 17:00:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][970/1251] eta 0:01:20 lr 0.000064 time 0.3021 (0.2848) loss 3.2983 (3.0683) grad_norm 2.6375 (2.8825) [2021-04-16 17:00:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][980/1251] eta 0:01:17 lr 0.000064 time 0.2488 (0.2847) loss 3.4362 (3.0661) grad_norm 2.9943 (2.8834) [2021-04-16 17:00:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][990/1251] eta 0:01:14 lr 0.000064 time 0.2781 (0.2846) loss 2.6644 (3.0617) grad_norm 2.6167 (2.8827) [2021-04-16 17:00:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][1000/1251] eta 0:01:11 lr 0.000064 time 0.2927 (0.2845) loss 3.2104 (3.0613) grad_norm 2.9518 (2.8835) [2021-04-16 17:00:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][1010/1251] eta 0:01:08 lr 0.000064 time 0.2928 (0.2844) loss 3.5216 (3.0646) grad_norm 2.9721 (2.8855) [2021-04-16 17:00:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][1020/1251] eta 0:01:05 lr 0.000064 time 0.2760 (0.2843) loss 2.3261 (3.0641) grad_norm 2.5491 (2.8860) [2021-04-16 17:00:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][1030/1251] eta 0:01:02 lr 0.000064 time 0.2843 (0.2843) loss 3.1178 (3.0622) grad_norm 3.3126 (2.8865) [2021-04-16 17:00:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][1040/1251] eta 0:00:59 lr 0.000064 time 0.2951 (0.2842) loss 2.5932 (3.0642) grad_norm 3.0480 (2.8877) [2021-04-16 17:00:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][1050/1251] eta 0:00:57 lr 0.000064 time 0.2576 (0.2841) loss 3.4934 (3.0638) grad_norm 2.8982 (2.8883) [2021-04-16 17:00:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][1060/1251] eta 0:00:54 lr 0.000064 time 0.2839 (0.2840) loss 2.8848 (3.0643) grad_norm 2.9754 (2.8902) [2021-04-16 17:00:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][1070/1251] eta 0:00:51 lr 0.000064 time 0.2607 (0.2840) loss 3.0221 (3.0642) grad_norm 3.2801 (2.8921) [2021-04-16 17:00:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][1080/1251] eta 0:00:48 lr 0.000064 time 0.2832 (0.2840) loss 3.5057 (3.0644) grad_norm 2.7137 (2.8921) [2021-04-16 17:00:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][1090/1251] eta 0:00:45 lr 0.000064 time 0.2494 (0.2839) loss 3.6827 (3.0637) grad_norm 3.4132 (2.9003) [2021-04-16 17:00:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][1100/1251] eta 0:00:42 lr 0.000064 time 0.2673 (0.2838) loss 3.4509 (3.0638) grad_norm 2.4074 (2.8989) [2021-04-16 17:01:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][1110/1251] eta 0:00:40 lr 0.000064 time 0.2707 (0.2837) loss 2.5164 (3.0665) grad_norm 2.9666 (2.8989) [2021-04-16 17:01:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][1120/1251] eta 0:00:37 lr 0.000064 time 0.2803 (0.2838) loss 3.3500 (3.0687) grad_norm 3.3954 (2.8988) [2021-04-16 17:01:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][1130/1251] eta 0:00:34 lr 0.000064 time 0.2796 (0.2837) loss 3.3466 (3.0690) grad_norm 3.0155 (2.8978) [2021-04-16 17:01:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][1140/1251] eta 0:00:31 lr 0.000064 time 0.2583 (0.2836) loss 3.3709 (3.0703) grad_norm 2.9720 (2.8982) [2021-04-16 17:01:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][1150/1251] eta 0:00:28 lr 0.000064 time 0.2563 (0.2835) loss 3.5857 (3.0707) grad_norm 4.6013 (2.8997) [2021-04-16 17:01:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][1160/1251] eta 0:00:25 lr 0.000064 time 0.2720 (0.2837) loss 2.8990 (3.0712) grad_norm 3.1198 (2.8998) [2021-04-16 17:01:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][1170/1251] eta 0:00:22 lr 0.000064 time 0.2947 (0.2837) loss 2.4389 (3.0709) grad_norm 2.7648 (2.8997) [2021-04-16 17:01:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][1180/1251] eta 0:00:20 lr 0.000064 time 0.2531 (0.2836) loss 3.0443 (3.0710) grad_norm 2.7119 (2.8983) [2021-04-16 17:01:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][1190/1251] eta 0:00:17 lr 0.000064 time 0.2675 (0.2835) loss 2.1182 (3.0707) grad_norm 2.5557 (2.8961) [2021-04-16 17:01:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][1200/1251] eta 0:00:14 lr 0.000064 time 0.2693 (0.2834) loss 3.3646 (3.0708) grad_norm 2.8105 (2.8956) [2021-04-16 17:01:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][1210/1251] eta 0:00:11 lr 0.000064 time 0.2878 (0.2834) loss 3.5029 (3.0713) grad_norm 3.0003 (2.8953) [2021-04-16 17:01:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][1220/1251] eta 0:00:08 lr 0.000064 time 0.2581 (0.2833) loss 3.7519 (3.0720) grad_norm 2.7748 (2.8946) [2021-04-16 17:01:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][1230/1251] eta 0:00:05 lr 0.000064 time 0.2716 (0.2832) loss 2.9628 (3.0718) grad_norm 3.2974 (2.8936) [2021-04-16 17:01:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][1240/1251] eta 0:00:03 lr 0.000064 time 0.2486 (0.2831) loss 3.7866 (3.0722) grad_norm 2.9045 (2.8935) [2021-04-16 17:01:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [254/300][1250/1251] eta 0:00:00 lr 0.000064 time 0.2479 (0.2828) loss 3.3438 (3.0720) grad_norm 2.6090 (2.8943) [2021-04-16 17:01:44 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 254 training takes 0:05:58 [2021-04-16 17:01:44 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_254.pth saving...... [2021-04-16 17:01:59 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_254.pth saved !!! [2021-04-16 17:02:00 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.178 (1.178) Loss 0.8291 (0.8291) Acc@1 80.371 (80.371) Acc@5 95.312 (95.312) [2021-04-16 17:02:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.134 (0.238) Loss 0.7585 (0.8231) Acc@1 82.422 (80.966) Acc@5 96.191 (95.392) [2021-04-16 17:02:04 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.125 (0.239) Loss 0.7732 (0.8213) Acc@1 81.836 (80.985) Acc@5 95.801 (95.438) [2021-04-16 17:02:07 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.100 (0.244) Loss 0.8205 (0.8269) Acc@1 81.836 (80.926) Acc@5 95.508 (95.335) [2021-04-16 17:02:08 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.221) Loss 0.8208 (0.8351) Acc@1 80.859 (80.774) Acc@5 95.801 (95.312) [2021-04-16 17:02:19 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.606 Acc@5 95.332 [2021-04-16 17:02:19 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 80.6% [2021-04-16 17:02:19 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 80.61% [2021-04-16 17:02:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][0/1251] eta 3:42:49 lr 0.000064 time 10.6872 (10.6872) loss 2.2686 (2.2686) grad_norm 2.6040 (2.6040) [2021-04-16 17:02:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][10/1251] eta 0:25:25 lr 0.000064 time 0.4219 (1.2295) loss 2.7285 (2.8915) grad_norm 3.6686 (2.9783) [2021-04-16 17:02:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][20/1251] eta 0:15:52 lr 0.000064 time 0.2703 (0.7736) loss 3.2510 (3.0571) grad_norm 2.7898 (3.0628) [2021-04-16 17:02:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][30/1251] eta 0:12:35 lr 0.000064 time 0.2763 (0.6186) loss 2.0217 (3.0533) grad_norm 2.8674 (2.9970) [2021-04-16 17:02:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][40/1251] eta 0:10:47 lr 0.000064 time 0.2908 (0.5345) loss 3.3208 (3.0490) grad_norm 2.1856 (2.9143) [2021-04-16 17:02:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][50/1251] eta 0:09:43 lr 0.000064 time 0.2589 (0.4856) loss 2.9876 (3.0356) grad_norm 2.8728 (2.9088) [2021-04-16 17:02:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][60/1251] eta 0:08:57 lr 0.000064 time 0.2848 (0.4514) loss 2.9873 (3.0725) grad_norm 2.4521 (2.8951) [2021-04-16 17:02:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][70/1251] eta 0:08:24 lr 0.000064 time 0.2965 (0.4276) loss 3.2592 (3.0848) grad_norm 2.6100 (2.8636) [2021-04-16 17:02:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][80/1251] eta 0:08:00 lr 0.000064 time 0.2642 (0.4106) loss 3.3997 (3.0935) grad_norm 2.7614 (2.8688) [2021-04-16 17:02:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][90/1251] eta 0:07:39 lr 0.000064 time 0.2847 (0.3962) loss 3.3532 (3.0967) grad_norm 2.5960 (2.8660) [2021-04-16 17:02:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][100/1251] eta 0:07:22 lr 0.000064 time 0.2924 (0.3846) loss 3.4982 (3.0988) grad_norm 2.6643 (2.8554) [2021-04-16 17:03:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][110/1251] eta 0:07:07 lr 0.000064 time 0.2798 (0.3750) loss 3.1470 (3.0991) grad_norm 2.7728 (2.8542) [2021-04-16 17:03:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][120/1251] eta 0:06:54 lr 0.000064 time 0.2716 (0.3663) loss 2.8412 (3.1102) grad_norm 3.0610 (2.8555) [2021-04-16 17:03:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][130/1251] eta 0:06:43 lr 0.000064 time 0.2756 (0.3598) loss 2.9887 (3.1188) grad_norm 3.3924 (2.8695) [2021-04-16 17:03:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][140/1251] eta 0:06:33 lr 0.000064 time 0.2762 (0.3539) loss 3.8599 (3.1207) grad_norm 2.6846 (2.8738) [2021-04-16 17:03:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][150/1251] eta 0:06:25 lr 0.000064 time 0.2735 (0.3500) loss 2.8538 (3.1167) grad_norm 2.8255 (2.8798) [2021-04-16 17:03:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][160/1251] eta 0:06:17 lr 0.000064 time 0.2781 (0.3462) loss 3.5753 (3.1079) grad_norm 2.7183 (2.8757) [2021-04-16 17:03:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][170/1251] eta 0:06:10 lr 0.000064 time 0.2565 (0.3425) loss 2.5775 (3.0919) grad_norm 3.5274 (2.8729) [2021-04-16 17:03:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][180/1251] eta 0:06:02 lr 0.000064 time 0.2679 (0.3387) loss 2.1950 (3.0849) grad_norm 2.7875 (2.8708) [2021-04-16 17:03:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][190/1251] eta 0:05:56 lr 0.000064 time 0.3056 (0.3357) loss 1.9253 (3.0697) grad_norm 2.4847 (2.8667) [2021-04-16 17:03:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][200/1251] eta 0:05:49 lr 0.000064 time 0.2865 (0.3328) loss 3.5403 (3.0727) grad_norm 2.5813 (2.8702) [2021-04-16 17:03:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][210/1251] eta 0:05:44 lr 0.000064 time 0.2825 (0.3305) loss 2.0559 (3.0716) grad_norm 3.5073 (2.8771) [2021-04-16 17:03:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][220/1251] eta 0:05:38 lr 0.000064 time 0.2866 (0.3284) loss 3.6345 (3.0801) grad_norm 2.4690 (2.8913) [2021-04-16 17:03:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][230/1251] eta 0:05:33 lr 0.000064 time 0.2989 (0.3263) loss 3.3158 (3.0864) grad_norm 2.8038 (2.8899) [2021-04-16 17:03:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][240/1251] eta 0:05:27 lr 0.000064 time 0.2663 (0.3243) loss 3.5882 (3.0894) grad_norm 2.4182 (2.8854) [2021-04-16 17:03:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][250/1251] eta 0:05:23 lr 0.000063 time 0.2696 (0.3229) loss 3.4918 (3.0867) grad_norm 3.3804 (2.8881) [2021-04-16 17:03:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][260/1251] eta 0:05:18 lr 0.000063 time 0.2825 (0.3211) loss 3.0634 (3.0850) grad_norm 3.8424 (2.8943) [2021-04-16 17:03:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][270/1251] eta 0:05:13 lr 0.000063 time 0.2998 (0.3196) loss 3.3633 (3.0890) grad_norm 3.5762 (2.9054) [2021-04-16 17:03:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][280/1251] eta 0:05:09 lr 0.000063 time 0.2676 (0.3183) loss 3.0535 (3.0808) grad_norm 3.8712 (2.9110) [2021-04-16 17:03:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][290/1251] eta 0:05:04 lr 0.000063 time 0.2648 (0.3172) loss 3.1740 (3.0837) grad_norm 3.8668 (2.9176) [2021-04-16 17:03:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][300/1251] eta 0:05:00 lr 0.000063 time 0.2774 (0.3160) loss 2.8276 (3.0821) grad_norm 2.8992 (2.9205) [2021-04-16 17:03:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][310/1251] eta 0:04:56 lr 0.000063 time 0.2799 (0.3148) loss 3.5910 (3.0802) grad_norm 2.6694 (2.9195) [2021-04-16 17:04:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][320/1251] eta 0:04:52 lr 0.000063 time 0.2679 (0.3141) loss 2.6309 (3.0803) grad_norm 3.2990 (2.9193) [2021-04-16 17:04:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][330/1251] eta 0:04:48 lr 0.000063 time 0.2653 (0.3128) loss 3.0538 (3.0795) grad_norm nan (nan) [2021-04-16 17:04:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][340/1251] eta 0:04:44 lr 0.000063 time 0.3024 (0.3118) loss 2.7535 (3.0752) grad_norm 3.0126 (nan) [2021-04-16 17:04:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][350/1251] eta 0:04:40 lr 0.000063 time 0.2718 (0.3108) loss 3.2741 (3.0740) grad_norm 2.8232 (nan) [2021-04-16 17:04:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][360/1251] eta 0:04:36 lr 0.000063 time 0.3024 (0.3107) loss 3.1669 (3.0735) grad_norm 3.0920 (nan) [2021-04-16 17:04:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][370/1251] eta 0:04:33 lr 0.000063 time 0.2899 (0.3100) loss 3.5288 (3.0804) grad_norm 2.9443 (nan) [2021-04-16 17:04:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][380/1251] eta 0:04:29 lr 0.000063 time 0.2682 (0.3092) loss 3.4856 (3.0775) grad_norm 3.0657 (nan) [2021-04-16 17:04:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][390/1251] eta 0:04:25 lr 0.000063 time 0.2776 (0.3084) loss 3.3200 (3.0771) grad_norm 2.4051 (nan) [2021-04-16 17:04:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][400/1251] eta 0:04:21 lr 0.000063 time 0.2557 (0.3075) loss 3.2036 (3.0770) grad_norm 2.5452 (nan) [2021-04-16 17:04:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][410/1251] eta 0:04:18 lr 0.000063 time 0.2859 (0.3068) loss 3.4812 (3.0792) grad_norm 3.6713 (nan) [2021-04-16 17:04:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][420/1251] eta 0:04:14 lr 0.000063 time 0.2760 (0.3062) loss 3.4824 (3.0774) grad_norm 2.7110 (nan) [2021-04-16 17:04:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][430/1251] eta 0:04:10 lr 0.000063 time 0.2496 (0.3054) loss 3.9444 (3.0795) grad_norm 3.3634 (nan) [2021-04-16 17:04:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][440/1251] eta 0:04:07 lr 0.000063 time 0.2770 (0.3050) loss 2.1387 (3.0740) grad_norm 2.7485 (nan) [2021-04-16 17:04:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][450/1251] eta 0:04:03 lr 0.000063 time 0.2416 (0.3043) loss 2.4737 (3.0773) grad_norm 2.6337 (nan) [2021-04-16 17:04:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][460/1251] eta 0:04:00 lr 0.000063 time 0.2836 (0.3037) loss 1.9755 (3.0736) grad_norm 2.5639 (nan) [2021-04-16 17:04:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][470/1251] eta 0:03:56 lr 0.000063 time 0.2720 (0.3033) loss 3.6532 (3.0703) grad_norm 2.8815 (nan) [2021-04-16 17:04:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][480/1251] eta 0:03:53 lr 0.000063 time 0.2771 (0.3027) loss 2.5083 (3.0745) grad_norm 3.6051 (nan) [2021-04-16 17:04:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][490/1251] eta 0:03:49 lr 0.000063 time 0.2748 (0.3022) loss 3.5077 (3.0797) grad_norm 2.5866 (nan) [2021-04-16 17:04:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][500/1251] eta 0:03:46 lr 0.000063 time 0.2609 (0.3022) loss 2.8725 (3.0837) grad_norm 3.5761 (nan) [2021-04-16 17:04:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][510/1251] eta 0:03:43 lr 0.000063 time 0.2686 (0.3017) loss 2.3085 (3.0765) grad_norm 3.0380 (nan) [2021-04-16 17:04:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][520/1251] eta 0:03:40 lr 0.000063 time 0.2834 (0.3012) loss 2.5630 (3.0737) grad_norm 2.7139 (nan) [2021-04-16 17:04:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][530/1251] eta 0:03:36 lr 0.000063 time 0.2683 (0.3008) loss 2.3163 (3.0718) grad_norm 2.4844 (nan) [2021-04-16 17:05:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][540/1251] eta 0:03:33 lr 0.000063 time 0.2655 (0.3003) loss 2.7554 (3.0724) grad_norm 3.2496 (nan) [2021-04-16 17:05:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][550/1251] eta 0:03:30 lr 0.000063 time 0.2640 (0.2998) loss 3.1277 (3.0728) grad_norm 2.6803 (nan) [2021-04-16 17:05:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][560/1251] eta 0:03:26 lr 0.000063 time 0.2746 (0.2994) loss 3.4975 (3.0731) grad_norm 3.8250 (nan) [2021-04-16 17:05:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][570/1251] eta 0:03:24 lr 0.000063 time 0.2948 (0.2996) loss 1.9450 (3.0729) grad_norm 2.4448 (nan) [2021-04-16 17:05:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][580/1251] eta 0:03:20 lr 0.000063 time 0.2668 (0.2993) loss 3.3116 (3.0748) grad_norm 2.6126 (nan) [2021-04-16 17:05:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][590/1251] eta 0:03:17 lr 0.000063 time 0.2504 (0.2992) loss 2.6546 (3.0746) grad_norm 2.9204 (nan) [2021-04-16 17:05:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][600/1251] eta 0:03:14 lr 0.000063 time 0.4053 (0.2989) loss 2.8074 (3.0689) grad_norm 4.1746 (nan) [2021-04-16 17:05:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][610/1251] eta 0:03:11 lr 0.000063 time 0.2715 (0.2986) loss 2.6840 (3.0682) grad_norm 2.8896 (nan) [2021-04-16 17:05:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][620/1251] eta 0:03:08 lr 0.000063 time 0.3213 (0.2983) loss 3.0806 (3.0683) grad_norm 2.8686 (nan) [2021-04-16 17:05:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][630/1251] eta 0:03:05 lr 0.000063 time 0.2801 (0.2980) loss 2.7313 (3.0716) grad_norm 2.6902 (nan) [2021-04-16 17:05:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][640/1251] eta 0:03:01 lr 0.000063 time 0.2627 (0.2977) loss 3.3955 (3.0760) grad_norm 3.2244 (nan) [2021-04-16 17:05:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][650/1251] eta 0:02:58 lr 0.000063 time 0.2873 (0.2975) loss 3.1888 (3.0746) grad_norm 2.7574 (nan) [2021-04-16 17:05:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][660/1251] eta 0:02:55 lr 0.000063 time 0.2447 (0.2971) loss 2.0964 (3.0738) grad_norm 2.6822 (nan) [2021-04-16 17:05:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][670/1251] eta 0:02:52 lr 0.000063 time 0.2775 (0.2969) loss 3.3361 (3.0764) grad_norm 3.1060 (nan) [2021-04-16 17:05:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][680/1251] eta 0:02:49 lr 0.000063 time 0.3033 (0.2967) loss 3.0482 (3.0787) grad_norm 2.7535 (nan) [2021-04-16 17:05:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][690/1251] eta 0:02:46 lr 0.000063 time 0.2881 (0.2966) loss 3.4949 (3.0782) grad_norm 2.9786 (nan) [2021-04-16 17:05:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][700/1251] eta 0:02:43 lr 0.000063 time 0.3054 (0.2965) loss 2.3417 (3.0785) grad_norm 5.2667 (nan) [2021-04-16 17:05:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][710/1251] eta 0:02:40 lr 0.000063 time 0.2916 (0.2962) loss 2.5896 (3.0751) grad_norm 2.6718 (nan) [2021-04-16 17:05:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][720/1251] eta 0:02:37 lr 0.000063 time 0.2907 (0.2960) loss 3.8954 (3.0764) grad_norm 2.7326 (nan) [2021-04-16 17:05:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][730/1251] eta 0:02:34 lr 0.000063 time 0.4202 (0.2960) loss 2.9317 (3.0759) grad_norm 3.0014 (nan) [2021-04-16 17:05:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][740/1251] eta 0:02:31 lr 0.000063 time 0.2536 (0.2958) loss 2.7052 (3.0775) grad_norm 2.8880 (nan) [2021-04-16 17:06:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][750/1251] eta 0:02:28 lr 0.000063 time 0.3209 (0.2956) loss 3.1188 (3.0800) grad_norm 2.8170 (nan) [2021-04-16 17:06:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][760/1251] eta 0:02:25 lr 0.000063 time 0.2507 (0.2954) loss 3.1897 (3.0786) grad_norm 2.6131 (nan) [2021-04-16 17:06:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][770/1251] eta 0:02:21 lr 0.000063 time 0.2897 (0.2951) loss 3.7938 (3.0766) grad_norm 2.6739 (nan) [2021-04-16 17:06:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][780/1251] eta 0:02:18 lr 0.000062 time 0.2648 (0.2949) loss 2.8030 (3.0783) grad_norm 2.8398 (nan) [2021-04-16 17:06:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][790/1251] eta 0:02:15 lr 0.000062 time 0.2758 (0.2947) loss 2.8450 (3.0775) grad_norm 2.8418 (nan) [2021-04-16 17:06:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][800/1251] eta 0:02:12 lr 0.000062 time 0.2927 (0.2946) loss 2.2636 (3.0753) grad_norm 2.7383 (nan) [2021-04-16 17:06:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][810/1251] eta 0:02:09 lr 0.000062 time 0.2550 (0.2944) loss 3.2085 (3.0727) grad_norm 2.5005 (nan) [2021-04-16 17:06:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][820/1251] eta 0:02:06 lr 0.000062 time 0.2651 (0.2943) loss 3.4973 (3.0755) grad_norm 3.0658 (nan) [2021-04-16 17:06:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][830/1251] eta 0:02:03 lr 0.000062 time 0.2859 (0.2942) loss 3.3093 (3.0788) grad_norm 3.1745 (nan) [2021-04-16 17:06:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][840/1251] eta 0:02:00 lr 0.000062 time 0.2752 (0.2939) loss 2.4384 (3.0787) grad_norm 3.2245 (nan) [2021-04-16 17:06:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][850/1251] eta 0:01:57 lr 0.000062 time 0.2559 (0.2937) loss 3.5231 (3.0796) grad_norm 3.2287 (nan) [2021-04-16 17:06:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][860/1251] eta 0:01:54 lr 0.000062 time 0.2746 (0.2935) loss 2.8507 (3.0807) grad_norm 2.8387 (nan) [2021-04-16 17:06:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][870/1251] eta 0:01:51 lr 0.000062 time 0.2740 (0.2935) loss 3.3276 (3.0827) grad_norm 2.9518 (nan) [2021-04-16 17:06:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][880/1251] eta 0:01:48 lr 0.000062 time 0.2776 (0.2932) loss 3.3585 (3.0836) grad_norm 2.9843 (nan) [2021-04-16 17:06:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][890/1251] eta 0:01:45 lr 0.000062 time 0.2824 (0.2932) loss 2.9255 (3.0851) grad_norm 2.8061 (nan) [2021-04-16 17:06:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][900/1251] eta 0:01:42 lr 0.000062 time 0.2596 (0.2931) loss 3.8170 (3.0851) grad_norm 2.8018 (nan) [2021-04-16 17:06:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][910/1251] eta 0:01:39 lr 0.000062 time 0.2923 (0.2929) loss 3.2993 (3.0846) grad_norm 2.5477 (nan) [2021-04-16 17:06:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][920/1251] eta 0:01:36 lr 0.000062 time 0.2672 (0.2927) loss 2.6311 (3.0808) grad_norm 2.6281 (nan) [2021-04-16 17:06:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][930/1251] eta 0:01:33 lr 0.000062 time 0.2551 (0.2927) loss 3.3162 (3.0779) grad_norm 2.9262 (nan) [2021-04-16 17:06:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][940/1251] eta 0:01:31 lr 0.000062 time 0.2904 (0.2927) loss 2.7022 (3.0771) grad_norm 2.6907 (nan) [2021-04-16 17:06:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][950/1251] eta 0:01:28 lr 0.000062 time 0.2833 (0.2925) loss 3.6564 (3.0769) grad_norm 3.3067 (nan) [2021-04-16 17:07:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][960/1251] eta 0:01:25 lr 0.000062 time 0.2762 (0.2923) loss 3.0419 (3.0758) grad_norm 3.0740 (nan) [2021-04-16 17:07:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][970/1251] eta 0:01:22 lr 0.000062 time 0.2716 (0.2921) loss 3.4435 (3.0773) grad_norm 3.1810 (nan) [2021-04-16 17:07:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][980/1251] eta 0:01:19 lr 0.000062 time 0.2642 (0.2920) loss 3.0366 (3.0764) grad_norm 2.7139 (nan) [2021-04-16 17:07:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][990/1251] eta 0:01:16 lr 0.000062 time 0.2501 (0.2919) loss 3.2490 (3.0780) grad_norm 3.3846 (nan) [2021-04-16 17:07:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][1000/1251] eta 0:01:13 lr 0.000062 time 0.3179 (0.2917) loss 3.4284 (3.0801) grad_norm 3.3754 (nan) [2021-04-16 17:07:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][1010/1251] eta 0:01:10 lr 0.000062 time 0.2529 (0.2916) loss 3.2203 (3.0796) grad_norm 3.0053 (nan) [2021-04-16 17:07:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][1020/1251] eta 0:01:07 lr 0.000062 time 0.2974 (0.2916) loss 3.6352 (3.0791) grad_norm 2.8141 (nan) [2021-04-16 17:07:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][1030/1251] eta 0:01:04 lr 0.000062 time 0.2666 (0.2914) loss 3.1532 (3.0779) grad_norm 2.5391 (nan) [2021-04-16 17:07:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][1040/1251] eta 0:01:01 lr 0.000062 time 0.2574 (0.2913) loss 2.8673 (3.0784) grad_norm 2.7528 (nan) [2021-04-16 17:07:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][1050/1251] eta 0:00:58 lr 0.000062 time 0.2980 (0.2912) loss 2.2078 (3.0784) grad_norm 2.9563 (nan) [2021-04-16 17:07:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][1060/1251] eta 0:00:55 lr 0.000062 time 0.2769 (0.2910) loss 2.7662 (3.0791) grad_norm 2.9614 (nan) [2021-04-16 17:07:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][1070/1251] eta 0:00:52 lr 0.000062 time 0.2814 (0.2909) loss 3.3982 (3.0784) grad_norm 2.5117 (nan) [2021-04-16 17:07:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][1080/1251] eta 0:00:49 lr 0.000062 time 0.2853 (0.2908) loss 2.4156 (3.0749) grad_norm 3.6220 (nan) [2021-04-16 17:07:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][1090/1251] eta 0:00:46 lr 0.000062 time 0.3065 (0.2908) loss 3.2113 (3.0736) grad_norm 2.4677 (nan) [2021-04-16 17:07:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][1100/1251] eta 0:00:43 lr 0.000062 time 0.2827 (0.2906) loss 3.4976 (3.0729) grad_norm 2.6264 (nan) [2021-04-16 17:07:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][1110/1251] eta 0:00:40 lr 0.000062 time 0.2825 (0.2906) loss 3.0471 (3.0717) grad_norm 2.5176 (nan) [2021-04-16 17:07:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][1120/1251] eta 0:00:38 lr 0.000062 time 0.2478 (0.2904) loss 2.5902 (3.0696) grad_norm 2.6104 (nan) [2021-04-16 17:07:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][1130/1251] eta 0:00:35 lr 0.000062 time 0.2777 (0.2904) loss 3.2414 (3.0692) grad_norm 3.1503 (nan) [2021-04-16 17:07:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][1140/1251] eta 0:00:32 lr 0.000062 time 0.2757 (0.2903) loss 2.8517 (3.0686) grad_norm 2.5601 (nan) [2021-04-16 17:07:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][1150/1251] eta 0:00:29 lr 0.000062 time 0.2839 (0.2903) loss 1.8819 (3.0689) grad_norm 2.8915 (nan) [2021-04-16 17:07:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][1160/1251] eta 0:00:26 lr 0.000062 time 0.2771 (0.2903) loss 3.1495 (3.0694) grad_norm 2.8948 (nan) [2021-04-16 17:07:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][1170/1251] eta 0:00:23 lr 0.000062 time 0.2760 (0.2904) loss 2.9598 (3.0693) grad_norm 2.9815 (nan) [2021-04-16 17:08:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][1180/1251] eta 0:00:20 lr 0.000062 time 0.2899 (0.2904) loss 3.5699 (3.0693) grad_norm 3.4865 (nan) [2021-04-16 17:08:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][1190/1251] eta 0:00:17 lr 0.000062 time 0.2767 (0.2902) loss 3.4795 (3.0693) grad_norm 2.6711 (nan) [2021-04-16 17:08:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][1200/1251] eta 0:00:14 lr 0.000062 time 0.2560 (0.2902) loss 2.8131 (3.0675) grad_norm 2.3678 (nan) [2021-04-16 17:08:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][1210/1251] eta 0:00:11 lr 0.000062 time 0.2678 (0.2902) loss 2.6748 (3.0669) grad_norm 2.5287 (nan) [2021-04-16 17:08:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][1220/1251] eta 0:00:08 lr 0.000062 time 0.2657 (0.2901) loss 2.6735 (3.0650) grad_norm 3.4483 (nan) [2021-04-16 17:08:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][1230/1251] eta 0:00:06 lr 0.000062 time 0.3037 (0.2900) loss 3.2345 (3.0658) grad_norm 2.7145 (nan) [2021-04-16 17:08:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][1240/1251] eta 0:00:03 lr 0.000062 time 0.2480 (0.2898) loss 2.5737 (3.0657) grad_norm 2.4864 (nan) [2021-04-16 17:08:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [255/300][1250/1251] eta 0:00:00 lr 0.000062 time 0.2480 (0.2895) loss 2.7870 (3.0652) grad_norm 3.0417 (nan) [2021-04-16 17:08:25 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 255 training takes 0:06:06 [2021-04-16 17:08:25 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_255.pth saving...... [2021-04-16 17:08:37 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_255.pth saved !!! [2021-04-16 17:08:40 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 3.011 (3.011) Loss 0.8727 (0.8727) Acc@1 80.957 (80.957) Acc@5 94.727 (94.727) [2021-04-16 17:08:41 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.132 (0.397) Loss 0.8981 (0.8613) Acc@1 78.320 (80.282) Acc@5 94.824 (94.771) [2021-04-16 17:08:44 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.298 (0.334) Loss 0.8004 (0.8513) Acc@1 81.348 (80.352) Acc@5 95.508 (95.033) [2021-04-16 17:08:46 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.164 (0.289) Loss 0.9152 (0.8452) Acc@1 79.395 (80.327) Acc@5 94.727 (95.199) [2021-04-16 17:08:47 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.215 (0.261) Loss 0.8366 (0.8382) Acc@1 79.785 (80.452) Acc@5 95.898 (95.312) [2021-04-16 17:08:53 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.466 Acc@5 95.260 [2021-04-16 17:08:53 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 80.5% [2021-04-16 17:08:53 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 80.61% [2021-04-16 17:09:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][0/1251] eta 5:49:30 lr 0.000062 time 16.7633 (16.7633) loss 3.0712 (3.0712) grad_norm 2.9181 (2.9181) [2021-04-16 17:09:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][10/1251] eta 0:36:40 lr 0.000062 time 0.2894 (1.7729) loss 2.4666 (2.8599) grad_norm 3.0650 (3.2748) [2021-04-16 17:09:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][20/1251] eta 0:21:50 lr 0.000062 time 0.3027 (1.0642) loss 3.6403 (2.9032) grad_norm 2.6660 (3.2440) [2021-04-16 17:09:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][30/1251] eta 0:16:27 lr 0.000062 time 0.2772 (0.8090) loss 3.1977 (2.9661) grad_norm 2.7748 (3.1081) [2021-04-16 17:09:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][40/1251] eta 0:13:46 lr 0.000062 time 0.2902 (0.6827) loss 3.3049 (2.9929) grad_norm 2.9534 (3.0475) [2021-04-16 17:09:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][50/1251] eta 0:12:03 lr 0.000062 time 0.2788 (0.6024) loss 3.0956 (2.9987) grad_norm 2.5280 (3.0094) [2021-04-16 17:09:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][60/1251] eta 0:10:55 lr 0.000062 time 0.2625 (0.5506) loss 2.6930 (2.9920) grad_norm 2.8264 (3.0364) [2021-04-16 17:09:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][70/1251] eta 0:10:05 lr 0.000061 time 0.2752 (0.5123) loss 2.1972 (2.9855) grad_norm 2.9096 (3.0182) [2021-04-16 17:09:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][80/1251] eta 0:09:27 lr 0.000061 time 0.2668 (0.4845) loss 3.2433 (2.9954) grad_norm 2.8971 (2.9955) [2021-04-16 17:09:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][90/1251] eta 0:08:56 lr 0.000061 time 0.2724 (0.4622) loss 2.9618 (3.0251) grad_norm 2.8331 (2.9743) [2021-04-16 17:09:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][100/1251] eta 0:08:31 lr 0.000061 time 0.2678 (0.4446) loss 4.0688 (3.0524) grad_norm 2.7380 (2.9569) [2021-04-16 17:09:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][110/1251] eta 0:08:10 lr 0.000061 time 0.2812 (0.4299) loss 2.5895 (3.0619) grad_norm 3.7814 (2.9522) [2021-04-16 17:09:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][120/1251] eta 0:07:51 lr 0.000061 time 0.2754 (0.4171) loss 1.7163 (3.0515) grad_norm 2.7494 (2.9807) [2021-04-16 17:09:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][130/1251] eta 0:07:35 lr 0.000061 time 0.2871 (0.4067) loss 3.5985 (3.0629) grad_norm 2.6514 (2.9781) [2021-04-16 17:09:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][140/1251] eta 0:07:25 lr 0.000061 time 0.2698 (0.4008) loss 3.4269 (3.0555) grad_norm 2.3286 (2.9614) [2021-04-16 17:09:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][150/1251] eta 0:07:11 lr 0.000061 time 0.2599 (0.3923) loss 3.4311 (3.0640) grad_norm 2.7743 (2.9529) [2021-04-16 17:09:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][160/1251] eta 0:07:00 lr 0.000061 time 0.2650 (0.3850) loss 3.3156 (3.0669) grad_norm 3.0349 (2.9497) [2021-04-16 17:09:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][170/1251] eta 0:06:49 lr 0.000061 time 0.2740 (0.3788) loss 3.3558 (3.0605) grad_norm 2.6481 (2.9430) [2021-04-16 17:10:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][180/1251] eta 0:06:39 lr 0.000061 time 0.2899 (0.3735) loss 3.0307 (3.0649) grad_norm 2.4400 (2.9285) [2021-04-16 17:10:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][190/1251] eta 0:06:31 lr 0.000061 time 0.2889 (0.3686) loss 3.0858 (3.0634) grad_norm 2.5104 (2.9205) [2021-04-16 17:10:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][200/1251] eta 0:06:22 lr 0.000061 time 0.2683 (0.3638) loss 3.7228 (3.0608) grad_norm 2.7147 (2.9124) [2021-04-16 17:10:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][210/1251] eta 0:06:14 lr 0.000061 time 0.2826 (0.3598) loss 3.3817 (3.0735) grad_norm 2.7904 (2.9104) [2021-04-16 17:10:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][220/1251] eta 0:06:06 lr 0.000061 time 0.2530 (0.3558) loss 3.6121 (3.0768) grad_norm 2.6876 (2.9142) [2021-04-16 17:10:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][230/1251] eta 0:06:00 lr 0.000061 time 0.2436 (0.3528) loss 3.3466 (3.0819) grad_norm 2.9822 (2.9179) [2021-04-16 17:10:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][240/1251] eta 0:05:53 lr 0.000061 time 0.2800 (0.3498) loss 3.5753 (3.0747) grad_norm 3.0633 (2.9149) [2021-04-16 17:10:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][250/1251] eta 0:05:47 lr 0.000061 time 0.2834 (0.3469) loss 3.2210 (3.0739) grad_norm 3.1103 (2.9230) [2021-04-16 17:10:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][260/1251] eta 0:05:41 lr 0.000061 time 0.2665 (0.3442) loss 2.8607 (3.0750) grad_norm 2.5767 (2.9231) [2021-04-16 17:10:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][270/1251] eta 0:05:35 lr 0.000061 time 0.2672 (0.3422) loss 2.1439 (3.0778) grad_norm 2.6800 (2.9262) [2021-04-16 17:10:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][280/1251] eta 0:05:30 lr 0.000061 time 0.2706 (0.3399) loss 3.0156 (3.0715) grad_norm 3.3945 (2.9245) [2021-04-16 17:10:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][290/1251] eta 0:05:24 lr 0.000061 time 0.2431 (0.3379) loss 3.3740 (3.0741) grad_norm 2.7133 (2.9271) [2021-04-16 17:10:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][300/1251] eta 0:05:19 lr 0.000061 time 0.2736 (0.3363) loss 3.0681 (3.0718) grad_norm 3.0722 (2.9265) [2021-04-16 17:10:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][310/1251] eta 0:05:14 lr 0.000061 time 0.2900 (0.3343) loss 2.2641 (3.0676) grad_norm 2.4379 (2.9223) [2021-04-16 17:10:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][320/1251] eta 0:05:09 lr 0.000061 time 0.2762 (0.3326) loss 3.3302 (3.0626) grad_norm 2.8745 (2.9301) [2021-04-16 17:10:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][330/1251] eta 0:05:04 lr 0.000061 time 0.2756 (0.3311) loss 2.0880 (3.0562) grad_norm 2.6610 (2.9481) [2021-04-16 17:10:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][340/1251] eta 0:05:00 lr 0.000061 time 0.2749 (0.3295) loss 2.6243 (3.0508) grad_norm 2.7497 (2.9482) [2021-04-16 17:10:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][350/1251] eta 0:04:55 lr 0.000061 time 0.2658 (0.3283) loss 3.3203 (3.0557) grad_norm 2.7776 (2.9458) [2021-04-16 17:10:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][360/1251] eta 0:04:51 lr 0.000061 time 0.2913 (0.3275) loss 2.8204 (3.0548) grad_norm 3.2718 (2.9428) [2021-04-16 17:10:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][370/1251] eta 0:04:47 lr 0.000061 time 0.2950 (0.3266) loss 1.9079 (3.0509) grad_norm 2.7352 (2.9390) [2021-04-16 17:10:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][380/1251] eta 0:04:43 lr 0.000061 time 0.2708 (0.3258) loss 2.3092 (3.0524) grad_norm 2.9250 (2.9337) [2021-04-16 17:11:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][390/1251] eta 0:04:39 lr 0.000061 time 0.2658 (0.3245) loss 3.2227 (3.0574) grad_norm 3.2036 (2.9270) [2021-04-16 17:11:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][400/1251] eta 0:04:35 lr 0.000061 time 0.2772 (0.3237) loss 3.3947 (3.0599) grad_norm 2.8850 (2.9340) [2021-04-16 17:11:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][410/1251] eta 0:04:31 lr 0.000061 time 0.2862 (0.3226) loss 3.0627 (3.0621) grad_norm 2.7535 (2.9318) [2021-04-16 17:11:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][420/1251] eta 0:04:27 lr 0.000061 time 0.2632 (0.3215) loss 3.3803 (3.0611) grad_norm 2.7882 (2.9366) [2021-04-16 17:11:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][430/1251] eta 0:04:23 lr 0.000061 time 0.2831 (0.3205) loss 2.5554 (3.0628) grad_norm 2.9113 (2.9388) [2021-04-16 17:11:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][440/1251] eta 0:04:19 lr 0.000061 time 0.2981 (0.3198) loss 3.7181 (3.0694) grad_norm 2.5275 (2.9376) [2021-04-16 17:11:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][450/1251] eta 0:04:15 lr 0.000061 time 0.2874 (0.3188) loss 3.1489 (3.0708) grad_norm 2.7931 (2.9348) [2021-04-16 17:11:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][460/1251] eta 0:04:11 lr 0.000061 time 0.2727 (0.3180) loss 3.4759 (3.0738) grad_norm 3.3022 (2.9354) [2021-04-16 17:11:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][470/1251] eta 0:04:07 lr 0.000061 time 0.2753 (0.3172) loss 2.4170 (3.0760) grad_norm 3.0029 (2.9315) [2021-04-16 17:11:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][480/1251] eta 0:04:04 lr 0.000061 time 0.2696 (0.3165) loss 2.3879 (3.0728) grad_norm 2.7273 (2.9288) [2021-04-16 17:11:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][490/1251] eta 0:04:00 lr 0.000061 time 0.2773 (0.3157) loss 3.5249 (3.0783) grad_norm 2.5907 (2.9285) [2021-04-16 17:11:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][500/1251] eta 0:03:56 lr 0.000061 time 0.2585 (0.3149) loss 2.6584 (3.0732) grad_norm 2.9305 (2.9263) [2021-04-16 17:11:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][510/1251] eta 0:03:52 lr 0.000061 time 0.2746 (0.3142) loss 3.0438 (3.0714) grad_norm 2.7260 (2.9210) [2021-04-16 17:11:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][520/1251] eta 0:03:49 lr 0.000061 time 0.3025 (0.3136) loss 3.1813 (3.0700) grad_norm 2.7426 (2.9174) [2021-04-16 17:11:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][530/1251] eta 0:03:45 lr 0.000061 time 0.2790 (0.3130) loss 3.4565 (3.0662) grad_norm 2.3445 (2.9132) [2021-04-16 17:11:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][540/1251] eta 0:03:42 lr 0.000061 time 0.2763 (0.3123) loss 2.4266 (3.0617) grad_norm 3.0088 (2.9147) [2021-04-16 17:11:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][550/1251] eta 0:03:38 lr 0.000061 time 0.2935 (0.3117) loss 3.2989 (3.0600) grad_norm 2.9825 (2.9144) [2021-04-16 17:11:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][560/1251] eta 0:03:35 lr 0.000061 time 0.2914 (0.3112) loss 3.3253 (3.0668) grad_norm 2.7450 (2.9134) [2021-04-16 17:11:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][570/1251] eta 0:03:31 lr 0.000061 time 0.2672 (0.3109) loss 2.4055 (3.0664) grad_norm 2.8144 (2.9128) [2021-04-16 17:11:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][580/1251] eta 0:03:28 lr 0.000061 time 0.2889 (0.3106) loss 2.2259 (3.0641) grad_norm 3.2155 (2.9130) [2021-04-16 17:11:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][590/1251] eta 0:03:24 lr 0.000061 time 0.2871 (0.3100) loss 3.1241 (3.0626) grad_norm 2.9282 (2.9129) [2021-04-16 17:11:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][600/1251] eta 0:03:21 lr 0.000061 time 0.2555 (0.3095) loss 3.5524 (3.0676) grad_norm 2.4259 (2.9124) [2021-04-16 17:12:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][610/1251] eta 0:03:18 lr 0.000061 time 0.2420 (0.3089) loss 3.8830 (3.0691) grad_norm 2.5586 (2.9135) [2021-04-16 17:12:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][620/1251] eta 0:03:14 lr 0.000060 time 0.2708 (0.3086) loss 2.9243 (3.0670) grad_norm 3.7836 (2.9165) [2021-04-16 17:12:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][630/1251] eta 0:03:11 lr 0.000060 time 0.2756 (0.3080) loss 3.4215 (3.0671) grad_norm 2.8629 (2.9149) [2021-04-16 17:12:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][640/1251] eta 0:03:07 lr 0.000060 time 0.2818 (0.3076) loss 3.2341 (3.0653) grad_norm 2.4202 (2.9122) [2021-04-16 17:12:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][650/1251] eta 0:03:04 lr 0.000060 time 0.2772 (0.3073) loss 2.5373 (3.0627) grad_norm 3.0540 (2.9118) [2021-04-16 17:12:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][660/1251] eta 0:03:01 lr 0.000060 time 0.2614 (0.3068) loss 3.1807 (3.0635) grad_norm 2.3836 (2.9096) [2021-04-16 17:12:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][670/1251] eta 0:02:58 lr 0.000060 time 0.2662 (0.3064) loss 2.8757 (3.0663) grad_norm 2.5899 (2.9084) [2021-04-16 17:12:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][680/1251] eta 0:02:54 lr 0.000060 time 0.2680 (0.3061) loss 2.9765 (3.0632) grad_norm 4.0454 (2.9077) [2021-04-16 17:12:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][690/1251] eta 0:02:51 lr 0.000060 time 0.2997 (0.3058) loss 3.6665 (3.0615) grad_norm 2.5497 (2.9065) [2021-04-16 17:12:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][700/1251] eta 0:02:48 lr 0.000060 time 0.2416 (0.3054) loss 2.4577 (3.0601) grad_norm 3.6301 (2.9112) [2021-04-16 17:12:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][710/1251] eta 0:02:44 lr 0.000060 time 0.2833 (0.3050) loss 3.3183 (3.0611) grad_norm 2.9207 (2.9103) [2021-04-16 17:12:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][720/1251] eta 0:02:41 lr 0.000060 time 0.2875 (0.3050) loss 3.7460 (3.0630) grad_norm 2.6607 (2.9093) [2021-04-16 17:12:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][730/1251] eta 0:02:38 lr 0.000060 time 0.2739 (0.3046) loss 2.4779 (3.0621) grad_norm 2.8677 (2.9091) [2021-04-16 17:12:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][740/1251] eta 0:02:35 lr 0.000060 time 0.2412 (0.3043) loss 2.8306 (3.0637) grad_norm 3.2657 (2.9080) [2021-04-16 17:12:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][750/1251] eta 0:02:32 lr 0.000060 time 0.3012 (0.3040) loss 2.7423 (3.0641) grad_norm 2.7676 (2.9067) [2021-04-16 17:12:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][760/1251] eta 0:02:29 lr 0.000060 time 0.2906 (0.3036) loss 2.8761 (3.0644) grad_norm 2.4569 (2.9027) [2021-04-16 17:12:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][770/1251] eta 0:02:25 lr 0.000060 time 0.2864 (0.3032) loss 2.7383 (3.0647) grad_norm 2.5946 (2.9002) [2021-04-16 17:12:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][780/1251] eta 0:02:22 lr 0.000060 time 0.2784 (0.3029) loss 3.2897 (3.0639) grad_norm 2.6808 (2.8995) [2021-04-16 17:12:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][790/1251] eta 0:02:19 lr 0.000060 time 0.2623 (0.3026) loss 3.4433 (3.0639) grad_norm 2.6275 (2.8991) [2021-04-16 17:12:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][800/1251] eta 0:02:16 lr 0.000060 time 0.2704 (0.3022) loss 3.6319 (3.0646) grad_norm 2.5149 (2.8983) [2021-04-16 17:12:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][810/1251] eta 0:02:13 lr 0.000060 time 0.2771 (0.3020) loss 1.9767 (3.0637) grad_norm 4.8226 (2.9003) [2021-04-16 17:13:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][820/1251] eta 0:02:10 lr 0.000060 time 0.2944 (0.3017) loss 3.8217 (3.0620) grad_norm 2.7041 (2.8990) [2021-04-16 17:13:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][830/1251] eta 0:02:06 lr 0.000060 time 0.2831 (0.3014) loss 2.2894 (3.0618) grad_norm 2.9607 (2.8981) [2021-04-16 17:13:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][840/1251] eta 0:02:03 lr 0.000060 time 0.2825 (0.3012) loss 3.3479 (3.0579) grad_norm 2.3231 (2.8984) [2021-04-16 17:13:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][850/1251] eta 0:02:00 lr 0.000060 time 0.2738 (0.3009) loss 2.9548 (3.0596) grad_norm 2.7017 (2.8999) [2021-04-16 17:13:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][860/1251] eta 0:01:57 lr 0.000060 time 0.2416 (0.3005) loss 3.5918 (3.0633) grad_norm 3.2426 (2.9032) [2021-04-16 17:13:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][870/1251] eta 0:01:54 lr 0.000060 time 0.3089 (0.3003) loss 3.3611 (3.0661) grad_norm 2.7239 (2.9040) [2021-04-16 17:13:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][880/1251] eta 0:01:51 lr 0.000060 time 0.2825 (0.3001) loss 3.0795 (3.0697) grad_norm 2.7958 (2.9030) [2021-04-16 17:13:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][890/1251] eta 0:01:48 lr 0.000060 time 0.2655 (0.2998) loss 3.3483 (3.0678) grad_norm 2.8223 (2.9014) [2021-04-16 17:13:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][900/1251] eta 0:01:45 lr 0.000060 time 0.2841 (0.2996) loss 2.9660 (3.0702) grad_norm 2.6485 (2.9026) [2021-04-16 17:13:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][910/1251] eta 0:01:42 lr 0.000060 time 0.2896 (0.2993) loss 2.9157 (3.0705) grad_norm 2.9051 (2.9009) [2021-04-16 17:13:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][920/1251] eta 0:01:39 lr 0.000060 time 0.4179 (0.2993) loss 3.1508 (3.0705) grad_norm 2.6484 (2.9001) [2021-04-16 17:13:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][930/1251] eta 0:01:36 lr 0.000060 time 0.2696 (0.2992) loss 3.3621 (3.0714) grad_norm 3.1547 (2.8992) [2021-04-16 17:13:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][940/1251] eta 0:01:33 lr 0.000060 time 0.2603 (0.2992) loss 3.5980 (3.0725) grad_norm 2.7855 (2.8997) [2021-04-16 17:13:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][950/1251] eta 0:01:30 lr 0.000060 time 0.3006 (0.2990) loss 2.1276 (3.0702) grad_norm 3.6660 (2.9015) [2021-04-16 17:13:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][960/1251] eta 0:01:26 lr 0.000060 time 0.2790 (0.2989) loss 2.9467 (3.0698) grad_norm 2.8376 (2.9056) [2021-04-16 17:13:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][970/1251] eta 0:01:23 lr 0.000060 time 0.2784 (0.2987) loss 3.1765 (3.0677) grad_norm 3.6416 (2.9058) [2021-04-16 17:13:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][980/1251] eta 0:01:20 lr 0.000060 time 0.2930 (0.2986) loss 3.3901 (3.0703) grad_norm 3.3330 (2.9056) [2021-04-16 17:13:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][990/1251] eta 0:01:17 lr 0.000060 time 0.3052 (0.2985) loss 2.2447 (3.0685) grad_norm 2.8770 (2.9063) [2021-04-16 17:13:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][1000/1251] eta 0:01:14 lr 0.000060 time 0.2881 (0.2983) loss 1.9768 (3.0676) grad_norm 2.8506 (2.9043) [2021-04-16 17:13:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][1010/1251] eta 0:01:11 lr 0.000060 time 0.2830 (0.2980) loss 2.2778 (3.0662) grad_norm 2.9556 (2.9036) [2021-04-16 17:13:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][1020/1251] eta 0:01:08 lr 0.000060 time 0.2798 (0.2978) loss 3.0913 (3.0663) grad_norm 2.7728 (2.9044) [2021-04-16 17:14:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][1030/1251] eta 0:01:05 lr 0.000060 time 0.2918 (0.2976) loss 3.4395 (3.0663) grad_norm 2.8190 (2.9037) [2021-04-16 17:14:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][1040/1251] eta 0:01:02 lr 0.000060 time 0.3140 (0.2974) loss 3.6611 (3.0672) grad_norm 2.9463 (2.9041) [2021-04-16 17:14:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][1050/1251] eta 0:00:59 lr 0.000060 time 0.2657 (0.2973) loss 3.3827 (3.0690) grad_norm 3.1490 (2.9051) [2021-04-16 17:14:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][1060/1251] eta 0:00:56 lr 0.000060 time 0.2700 (0.2971) loss 3.8393 (3.0699) grad_norm 2.5421 (2.9046) [2021-04-16 17:14:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][1070/1251] eta 0:00:53 lr 0.000060 time 0.2722 (0.2969) loss 2.8913 (3.0694) grad_norm 3.1873 (2.9036) [2021-04-16 17:14:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][1080/1251] eta 0:00:50 lr 0.000060 time 0.2788 (0.2967) loss 2.9388 (3.0704) grad_norm 2.7154 (2.9040) [2021-04-16 17:14:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][1090/1251] eta 0:00:47 lr 0.000060 time 0.2831 (0.2965) loss 3.5295 (3.0704) grad_norm 2.8524 (2.9043) [2021-04-16 17:14:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][1100/1251] eta 0:00:44 lr 0.000060 time 0.2582 (0.2964) loss 3.5894 (3.0693) grad_norm 2.7805 (2.9036) [2021-04-16 17:14:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][1110/1251] eta 0:00:41 lr 0.000060 time 0.2693 (0.2962) loss 3.3071 (3.0697) grad_norm 3.4162 (2.9035) [2021-04-16 17:14:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][1120/1251] eta 0:00:38 lr 0.000060 time 0.2829 (0.2960) loss 3.3766 (3.0701) grad_norm 2.9008 (2.9046) [2021-04-16 17:14:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][1130/1251] eta 0:00:35 lr 0.000060 time 0.2587 (0.2958) loss 3.8173 (3.0685) grad_norm 3.5070 (2.9043) [2021-04-16 17:14:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][1140/1251] eta 0:00:32 lr 0.000060 time 0.2656 (0.2956) loss 3.7862 (3.0683) grad_norm 2.6015 (2.9024) [2021-04-16 17:14:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][1150/1251] eta 0:00:29 lr 0.000060 time 0.2693 (0.2956) loss 3.3122 (3.0686) grad_norm 3.5820 (2.9040) [2021-04-16 17:14:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][1160/1251] eta 0:00:26 lr 0.000060 time 0.2709 (0.2955) loss 2.8989 (3.0689) grad_norm 3.7164 (2.9043) [2021-04-16 17:14:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][1170/1251] eta 0:00:23 lr 0.000059 time 0.3095 (0.2953) loss 3.6137 (3.0703) grad_norm 2.4560 (2.9048) [2021-04-16 17:14:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][1180/1251] eta 0:00:20 lr 0.000059 time 0.2987 (0.2952) loss 3.2693 (3.0697) grad_norm 2.8387 (2.9052) [2021-04-16 17:14:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][1190/1251] eta 0:00:18 lr 0.000059 time 0.4148 (0.2952) loss 3.1130 (3.0723) grad_norm 3.2478 (2.9046) [2021-04-16 17:14:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][1200/1251] eta 0:00:15 lr 0.000059 time 0.2959 (0.2950) loss 3.3464 (3.0719) grad_norm 2.9653 (2.9048) [2021-04-16 17:14:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][1210/1251] eta 0:00:12 lr 0.000059 time 0.2742 (0.2949) loss 3.0274 (3.0731) grad_norm 2.8352 (2.9056) [2021-04-16 17:14:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][1220/1251] eta 0:00:09 lr 0.000059 time 0.2826 (0.2948) loss 3.1704 (3.0732) grad_norm 3.0855 (2.9065) [2021-04-16 17:14:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][1230/1251] eta 0:00:06 lr 0.000059 time 0.2689 (0.2947) loss 2.7656 (3.0716) grad_norm 3.2673 (2.9070) [2021-04-16 17:14:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][1240/1251] eta 0:00:03 lr 0.000059 time 0.2484 (0.2945) loss 3.3578 (3.0734) grad_norm 2.7759 (2.9074) [2021-04-16 17:15:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [256/300][1250/1251] eta 0:00:00 lr 0.000059 time 0.2487 (0.2942) loss 3.2353 (3.0744) grad_norm 4.0462 (2.9094) [2021-04-16 17:15:06 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 256 training takes 0:06:12 [2021-04-16 17:15:06 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_256.pth saving...... [2021-04-16 17:15:14 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_256.pth saved !!! [2021-04-16 17:15:15 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.097 (1.097) Loss 0.8368 (0.8368) Acc@1 81.934 (81.934) Acc@5 94.824 (94.824) [2021-04-16 17:15:16 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.377 (0.225) Loss 0.9154 (0.8473) Acc@1 79.297 (80.504) Acc@5 94.531 (95.268) [2021-04-16 17:15:18 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.164 (0.209) Loss 0.8035 (0.8439) Acc@1 81.836 (80.459) Acc@5 96.094 (95.275) [2021-04-16 17:15:21 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.142 (0.224) Loss 0.8637 (0.8340) Acc@1 79.492 (80.563) Acc@5 95.703 (95.350) [2021-04-16 17:15:23 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.147 (0.219) Loss 0.8072 (0.8353) Acc@1 80.371 (80.569) Acc@5 95.996 (95.334) [2021-04-16 17:15:28 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.528 Acc@5 95.330 [2021-04-16 17:15:28 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 80.5% [2021-04-16 17:15:28 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 80.61% [2021-04-16 17:15:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][0/1251] eta 4:30:27 lr 0.000059 time 12.9720 (12.9720) loss 3.2320 (3.2320) grad_norm 2.9159 (2.9159) [2021-04-16 17:15:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][10/1251] eta 0:29:31 lr 0.000059 time 0.2569 (1.4273) loss 2.8215 (2.8768) grad_norm 3.0072 (2.9505) [2021-04-16 17:15:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][20/1251] eta 0:18:10 lr 0.000059 time 0.2699 (0.8860) loss 2.6186 (3.0170) grad_norm 2.7081 (2.8855) [2021-04-16 17:15:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][30/1251] eta 0:14:03 lr 0.000059 time 0.2988 (0.6908) loss 2.4322 (3.0821) grad_norm 2.6920 (2.8523) [2021-04-16 17:15:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][40/1251] eta 0:11:56 lr 0.000059 time 0.3001 (0.5914) loss 3.1814 (3.0632) grad_norm 3.2983 (2.8860) [2021-04-16 17:15:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][50/1251] eta 0:10:36 lr 0.000059 time 0.2899 (0.5300) loss 2.2452 (3.0135) grad_norm 3.5130 (2.8896) [2021-04-16 17:15:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][60/1251] eta 0:09:40 lr 0.000059 time 0.2815 (0.4878) loss 3.3513 (2.9924) grad_norm 2.9546 (2.9217) [2021-04-16 17:16:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][70/1251] eta 0:09:01 lr 0.000059 time 0.2594 (0.4581) loss 3.5156 (3.0095) grad_norm 3.2845 (2.9333) [2021-04-16 17:16:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][80/1251] eta 0:08:30 lr 0.000059 time 0.2703 (0.4361) loss 3.3718 (3.0162) grad_norm 2.6149 (2.9203) [2021-04-16 17:16:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][90/1251] eta 0:08:07 lr 0.000059 time 0.2913 (0.4199) loss 2.7881 (3.0367) grad_norm 3.1390 (2.9085) [2021-04-16 17:16:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][100/1251] eta 0:07:46 lr 0.000059 time 0.2880 (0.4056) loss 3.4499 (3.0382) grad_norm 3.3946 (2.9237) [2021-04-16 17:16:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][110/1251] eta 0:07:29 lr 0.000059 time 0.2630 (0.3938) loss 3.3405 (3.0438) grad_norm 3.0617 (2.9181) [2021-04-16 17:16:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][120/1251] eta 0:07:14 lr 0.000059 time 0.2420 (0.3842) loss 3.5721 (3.0447) grad_norm 3.3539 (2.9249) [2021-04-16 17:16:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][130/1251] eta 0:07:02 lr 0.000059 time 0.2558 (0.3768) loss 2.2630 (3.0420) grad_norm 3.3893 (2.9286) [2021-04-16 17:16:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][140/1251] eta 0:06:52 lr 0.000059 time 0.2727 (0.3715) loss 3.3695 (3.0272) grad_norm 2.7693 (2.9153) [2021-04-16 17:16:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][150/1251] eta 0:06:41 lr 0.000059 time 0.2432 (0.3649) loss 3.2927 (3.0415) grad_norm 2.7082 (2.9223) [2021-04-16 17:16:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][160/1251] eta 0:06:33 lr 0.000059 time 0.2826 (0.3604) loss 2.3937 (3.0414) grad_norm 3.7259 (2.9275) [2021-04-16 17:16:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][170/1251] eta 0:06:25 lr 0.000059 time 0.2788 (0.3566) loss 3.5460 (3.0441) grad_norm 2.7565 (2.9204) [2021-04-16 17:16:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][180/1251] eta 0:06:17 lr 0.000059 time 0.2602 (0.3522) loss 3.4829 (3.0348) grad_norm 3.0497 (2.9220) [2021-04-16 17:16:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][190/1251] eta 0:06:10 lr 0.000059 time 0.2709 (0.3491) loss 3.4793 (3.0290) grad_norm 2.6223 (2.9223) [2021-04-16 17:16:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][200/1251] eta 0:06:02 lr 0.000059 time 0.2646 (0.3453) loss 3.7642 (3.0311) grad_norm 3.0923 (2.9140) [2021-04-16 17:16:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][210/1251] eta 0:05:56 lr 0.000059 time 0.2796 (0.3421) loss 2.8455 (3.0344) grad_norm 2.9381 (2.9109) [2021-04-16 17:16:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][220/1251] eta 0:05:49 lr 0.000059 time 0.2812 (0.3390) loss 3.5143 (3.0477) grad_norm 2.8595 (2.9048) [2021-04-16 17:16:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][230/1251] eta 0:05:43 lr 0.000059 time 0.2623 (0.3363) loss 3.5946 (3.0504) grad_norm 2.7678 (2.8991) [2021-04-16 17:16:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][240/1251] eta 0:05:37 lr 0.000059 time 0.3001 (0.3341) loss 3.2302 (3.0344) grad_norm 3.1697 (2.8956) [2021-04-16 17:16:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][250/1251] eta 0:05:32 lr 0.000059 time 0.2798 (0.3318) loss 3.4220 (3.0378) grad_norm 2.9586 (2.8940) [2021-04-16 17:16:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][260/1251] eta 0:05:26 lr 0.000059 time 0.2521 (0.3295) loss 3.2647 (3.0379) grad_norm 3.3793 (2.8896) [2021-04-16 17:16:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][270/1251] eta 0:05:21 lr 0.000059 time 0.2560 (0.3274) loss 3.2832 (3.0361) grad_norm 2.7725 (2.8923) [2021-04-16 17:17:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][280/1251] eta 0:05:16 lr 0.000059 time 0.2732 (0.3255) loss 2.5787 (3.0376) grad_norm 2.7016 (2.8917) [2021-04-16 17:17:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][290/1251] eta 0:05:11 lr 0.000059 time 0.2720 (0.3238) loss 3.7352 (3.0342) grad_norm 2.8272 (2.8913) [2021-04-16 17:17:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][300/1251] eta 0:05:06 lr 0.000059 time 0.2873 (0.3223) loss 3.5412 (3.0382) grad_norm 2.8825 (2.8934) [2021-04-16 17:17:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][310/1251] eta 0:05:01 lr 0.000059 time 0.3118 (0.3209) loss 2.7511 (3.0316) grad_norm 2.6274 (2.8913) [2021-04-16 17:17:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][320/1251] eta 0:04:57 lr 0.000059 time 0.2824 (0.3195) loss 3.3486 (3.0320) grad_norm 2.7027 (2.8940) [2021-04-16 17:17:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][330/1251] eta 0:04:53 lr 0.000059 time 0.2903 (0.3181) loss 3.1682 (3.0341) grad_norm 2.8510 (2.8940) [2021-04-16 17:17:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][340/1251] eta 0:04:48 lr 0.000059 time 0.2431 (0.3169) loss 3.3932 (3.0428) grad_norm 3.0514 (2.9002) [2021-04-16 17:17:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][350/1251] eta 0:04:44 lr 0.000059 time 0.2776 (0.3159) loss 3.2605 (3.0396) grad_norm 3.0016 (2.8974) [2021-04-16 17:17:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][360/1251] eta 0:04:41 lr 0.000059 time 0.2461 (0.3154) loss 3.6514 (3.0397) grad_norm 3.0940 (2.9069) [2021-04-16 17:17:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][370/1251] eta 0:04:36 lr 0.000059 time 0.2784 (0.3143) loss 3.4209 (3.0385) grad_norm 3.4014 (2.9143) [2021-04-16 17:17:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][380/1251] eta 0:04:32 lr 0.000059 time 0.2697 (0.3133) loss 3.6854 (3.0409) grad_norm 2.8287 (2.9141) [2021-04-16 17:17:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][390/1251] eta 0:04:29 lr 0.000059 time 0.2963 (0.3126) loss 2.9753 (3.0406) grad_norm 2.7456 (2.9168) [2021-04-16 17:17:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][400/1251] eta 0:04:25 lr 0.000059 time 0.2990 (0.3117) loss 3.6325 (3.0360) grad_norm 3.1379 (2.9153) [2021-04-16 17:17:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][410/1251] eta 0:04:21 lr 0.000059 time 0.2699 (0.3107) loss 4.0780 (3.0403) grad_norm 3.2576 (2.9153) [2021-04-16 17:17:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][420/1251] eta 0:04:17 lr 0.000059 time 0.2723 (0.3099) loss 3.7315 (3.0470) grad_norm 2.6742 (2.9128) [2021-04-16 17:17:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][430/1251] eta 0:04:14 lr 0.000059 time 0.3161 (0.3095) loss 3.6593 (3.0480) grad_norm 3.0390 (2.9086) [2021-04-16 17:17:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][440/1251] eta 0:04:10 lr 0.000059 time 0.3045 (0.3087) loss 3.2649 (3.0491) grad_norm 3.4105 (2.9106) [2021-04-16 17:17:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][450/1251] eta 0:04:06 lr 0.000059 time 0.2839 (0.3079) loss 3.1807 (3.0538) grad_norm 3.1050 (2.9141) [2021-04-16 17:17:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][460/1251] eta 0:04:03 lr 0.000059 time 0.2689 (0.3073) loss 3.5247 (3.0482) grad_norm 3.0927 (2.9150) [2021-04-16 17:17:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][470/1251] eta 0:03:59 lr 0.000058 time 0.2801 (0.3066) loss 3.2835 (3.0504) grad_norm 3.3671 (2.9154) [2021-04-16 17:17:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][480/1251] eta 0:03:56 lr 0.000058 time 0.2592 (0.3062) loss 3.5950 (3.0499) grad_norm 2.8783 (2.9170) [2021-04-16 17:17:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][490/1251] eta 0:03:52 lr 0.000058 time 0.2806 (0.3057) loss 2.5731 (3.0497) grad_norm 2.6450 (2.9144) [2021-04-16 17:18:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][500/1251] eta 0:03:49 lr 0.000058 time 0.2758 (0.3052) loss 3.2047 (3.0473) grad_norm 2.7881 (2.9123) [2021-04-16 17:18:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][510/1251] eta 0:03:45 lr 0.000058 time 0.3024 (0.3046) loss 3.7553 (3.0530) grad_norm 2.5455 (2.9105) [2021-04-16 17:18:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][520/1251] eta 0:03:42 lr 0.000058 time 0.3119 (0.3040) loss 3.2787 (3.0566) grad_norm 3.0518 (2.9098) [2021-04-16 17:18:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][530/1251] eta 0:03:38 lr 0.000058 time 0.2574 (0.3035) loss 3.6892 (3.0601) grad_norm 3.1606 (2.9121) [2021-04-16 17:18:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][540/1251] eta 0:03:35 lr 0.000058 time 0.2713 (0.3030) loss 3.0066 (3.0558) grad_norm 3.0526 (2.9160) [2021-04-16 17:18:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][550/1251] eta 0:03:32 lr 0.000058 time 0.2605 (0.3025) loss 2.2627 (3.0529) grad_norm 3.2070 (2.9157) [2021-04-16 17:18:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][560/1251] eta 0:03:28 lr 0.000058 time 0.2904 (0.3020) loss 3.1063 (3.0515) grad_norm 2.8462 (2.9150) [2021-04-16 17:18:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][570/1251] eta 0:03:25 lr 0.000058 time 0.2634 (0.3016) loss 3.5213 (3.0539) grad_norm 2.8687 (inf) [2021-04-16 17:18:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][580/1251] eta 0:03:22 lr 0.000058 time 0.2781 (0.3017) loss 3.2453 (3.0541) grad_norm 4.4458 (inf) [2021-04-16 17:18:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][590/1251] eta 0:03:19 lr 0.000058 time 0.2610 (0.3015) loss 3.0421 (3.0574) grad_norm 3.0043 (inf) [2021-04-16 17:18:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][600/1251] eta 0:03:15 lr 0.000058 time 0.2561 (0.3010) loss 3.4934 (3.0548) grad_norm 2.2649 (inf) [2021-04-16 17:18:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][610/1251] eta 0:03:12 lr 0.000058 time 0.2847 (0.3006) loss 3.5809 (3.0550) grad_norm 3.0481 (inf) [2021-04-16 17:18:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][620/1251] eta 0:03:09 lr 0.000058 time 0.2664 (0.3002) loss 1.9447 (3.0554) grad_norm 6.3725 (inf) [2021-04-16 17:18:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][630/1251] eta 0:03:06 lr 0.000058 time 0.2815 (0.2998) loss 2.6368 (3.0558) grad_norm 2.8884 (inf) [2021-04-16 17:18:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][640/1251] eta 0:03:02 lr 0.000058 time 0.2574 (0.2994) loss 2.5600 (3.0550) grad_norm 2.4174 (inf) [2021-04-16 17:18:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][650/1251] eta 0:02:59 lr 0.000058 time 0.3015 (0.2992) loss 3.1428 (3.0558) grad_norm 2.5004 (inf) [2021-04-16 17:18:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][660/1251] eta 0:02:56 lr 0.000058 time 0.2918 (0.2988) loss 3.2679 (3.0578) grad_norm 3.2567 (inf) [2021-04-16 17:18:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][670/1251] eta 0:02:53 lr 0.000058 time 0.2982 (0.2986) loss 3.1261 (3.0601) grad_norm 3.3068 (inf) [2021-04-16 17:18:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][680/1251] eta 0:02:50 lr 0.000058 time 0.2492 (0.2984) loss 3.1079 (3.0583) grad_norm 2.9800 (inf) [2021-04-16 17:18:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][690/1251] eta 0:02:47 lr 0.000058 time 0.2808 (0.2980) loss 2.3485 (3.0573) grad_norm 2.6145 (inf) [2021-04-16 17:18:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][700/1251] eta 0:02:44 lr 0.000058 time 0.2831 (0.2977) loss 3.5282 (3.0561) grad_norm 3.5638 (inf) [2021-04-16 17:19:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][710/1251] eta 0:02:40 lr 0.000058 time 0.2982 (0.2975) loss 2.3567 (3.0510) grad_norm 2.8119 (inf) [2021-04-16 17:19:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][720/1251] eta 0:02:37 lr 0.000058 time 0.2886 (0.2975) loss 3.5183 (3.0485) grad_norm 2.6953 (inf) [2021-04-16 17:19:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][730/1251] eta 0:02:34 lr 0.000058 time 0.3713 (0.2974) loss 2.3479 (3.0525) grad_norm 3.3038 (inf) [2021-04-16 17:19:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][740/1251] eta 0:02:31 lr 0.000058 time 0.2889 (0.2974) loss 2.9270 (3.0528) grad_norm 2.5432 (inf) [2021-04-16 17:19:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][750/1251] eta 0:02:28 lr 0.000058 time 0.2877 (0.2970) loss 3.5061 (3.0569) grad_norm 2.8328 (inf) [2021-04-16 17:19:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][760/1251] eta 0:02:25 lr 0.000058 time 0.2607 (0.2967) loss 3.6526 (3.0601) grad_norm 3.8076 (inf) [2021-04-16 17:19:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][770/1251] eta 0:02:22 lr 0.000058 time 0.2706 (0.2966) loss 2.7615 (3.0621) grad_norm 2.8321 (inf) [2021-04-16 17:19:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][780/1251] eta 0:02:19 lr 0.000058 time 0.2816 (0.2963) loss 3.4749 (3.0616) grad_norm 3.2286 (inf) [2021-04-16 17:19:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][790/1251] eta 0:02:16 lr 0.000058 time 0.2683 (0.2961) loss 3.1291 (3.0631) grad_norm 2.4375 (inf) [2021-04-16 17:19:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][800/1251] eta 0:02:13 lr 0.000058 time 0.2974 (0.2958) loss 3.4033 (3.0637) grad_norm 2.8323 (inf) [2021-04-16 17:19:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][810/1251] eta 0:02:10 lr 0.000058 time 0.2622 (0.2956) loss 2.3930 (3.0669) grad_norm 2.7576 (inf) [2021-04-16 17:19:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][820/1251] eta 0:02:07 lr 0.000058 time 0.2902 (0.2954) loss 2.4956 (3.0659) grad_norm 2.9055 (inf) [2021-04-16 17:19:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][830/1251] eta 0:02:04 lr 0.000058 time 0.2552 (0.2951) loss 3.4101 (3.0647) grad_norm 3.0692 (inf) [2021-04-16 17:19:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][840/1251] eta 0:02:01 lr 0.000058 time 0.2580 (0.2949) loss 3.4332 (3.0656) grad_norm 2.6040 (inf) [2021-04-16 17:19:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][850/1251] eta 0:01:58 lr 0.000058 time 0.2751 (0.2947) loss 3.4128 (3.0635) grad_norm 2.6699 (inf) [2021-04-16 17:19:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][860/1251] eta 0:01:55 lr 0.000058 time 0.2724 (0.2945) loss 3.8333 (3.0656) grad_norm 2.4545 (inf) [2021-04-16 17:19:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][870/1251] eta 0:01:52 lr 0.000058 time 0.3192 (0.2943) loss 2.9741 (3.0617) grad_norm 3.3111 (inf) [2021-04-16 17:19:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][880/1251] eta 0:01:49 lr 0.000058 time 0.2708 (0.2940) loss 3.2009 (3.0609) grad_norm 2.5800 (inf) [2021-04-16 17:19:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][890/1251] eta 0:01:46 lr 0.000058 time 0.2861 (0.2938) loss 3.2228 (3.0597) grad_norm 2.8928 (inf) [2021-04-16 17:19:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][900/1251] eta 0:01:43 lr 0.000058 time 0.2779 (0.2936) loss 2.5351 (3.0583) grad_norm 2.5330 (inf) [2021-04-16 17:19:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][910/1251] eta 0:01:40 lr 0.000058 time 0.2866 (0.2934) loss 3.6654 (3.0580) grad_norm 2.8528 (inf) [2021-04-16 17:19:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][920/1251] eta 0:01:37 lr 0.000058 time 0.2486 (0.2932) loss 3.4969 (3.0570) grad_norm 3.0619 (inf) [2021-04-16 17:20:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][930/1251] eta 0:01:34 lr 0.000058 time 0.2861 (0.2932) loss 3.4430 (3.0577) grad_norm 3.3420 (inf) [2021-04-16 17:20:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][940/1251] eta 0:01:31 lr 0.000058 time 0.2653 (0.2931) loss 3.4691 (3.0557) grad_norm 2.6373 (inf) [2021-04-16 17:20:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][950/1251] eta 0:01:28 lr 0.000058 time 0.2853 (0.2929) loss 3.2439 (3.0553) grad_norm 2.8307 (inf) [2021-04-16 17:20:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][960/1251] eta 0:01:25 lr 0.000058 time 0.2474 (0.2928) loss 3.4032 (3.0559) grad_norm 3.2540 (inf) [2021-04-16 17:20:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][970/1251] eta 0:01:22 lr 0.000058 time 0.3020 (0.2928) loss 2.3651 (3.0563) grad_norm 3.2427 (inf) [2021-04-16 17:20:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][980/1251] eta 0:01:19 lr 0.000058 time 0.2498 (0.2926) loss 2.8476 (3.0551) grad_norm 3.0257 (inf) [2021-04-16 17:20:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][990/1251] eta 0:01:16 lr 0.000058 time 0.2662 (0.2925) loss 3.3020 (3.0566) grad_norm 2.5795 (inf) [2021-04-16 17:20:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][1000/1251] eta 0:01:13 lr 0.000058 time 0.2696 (0.2923) loss 3.5595 (3.0596) grad_norm 2.5663 (inf) [2021-04-16 17:20:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][1010/1251] eta 0:01:10 lr 0.000058 time 0.2824 (0.2923) loss 2.4746 (3.0587) grad_norm 2.8994 (inf) [2021-04-16 17:20:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][1020/1251] eta 0:01:07 lr 0.000058 time 0.2555 (0.2921) loss 3.4249 (3.0586) grad_norm 2.7830 (inf) [2021-04-16 17:20:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][1030/1251] eta 0:01:04 lr 0.000058 time 0.2964 (0.2919) loss 3.4080 (3.0585) grad_norm 2.5217 (inf) [2021-04-16 17:20:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][1040/1251] eta 0:01:01 lr 0.000057 time 0.2746 (0.2917) loss 2.2934 (3.0578) grad_norm 2.7599 (inf) [2021-04-16 17:20:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][1050/1251] eta 0:00:58 lr 0.000057 time 0.4937 (0.2917) loss 2.3555 (3.0556) grad_norm 2.6551 (inf) [2021-04-16 17:20:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][1060/1251] eta 0:00:55 lr 0.000057 time 0.2828 (0.2916) loss 3.6108 (3.0557) grad_norm 3.3415 (inf) [2021-04-16 17:20:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][1070/1251] eta 0:00:52 lr 0.000057 time 0.2584 (0.2914) loss 2.3210 (3.0559) grad_norm 3.2871 (inf) [2021-04-16 17:20:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][1080/1251] eta 0:00:49 lr 0.000057 time 0.3048 (0.2912) loss 3.8657 (3.0561) grad_norm 3.4068 (inf) [2021-04-16 17:20:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][1090/1251] eta 0:00:46 lr 0.000057 time 0.2651 (0.2911) loss 3.2688 (3.0555) grad_norm 2.9708 (inf) [2021-04-16 17:20:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][1100/1251] eta 0:00:43 lr 0.000057 time 0.2536 (0.2909) loss 1.8625 (3.0515) grad_norm 2.3161 (inf) [2021-04-16 17:20:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][1110/1251] eta 0:00:41 lr 0.000057 time 0.2526 (0.2908) loss 1.9720 (3.0494) grad_norm 2.6617 (inf) [2021-04-16 17:20:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][1120/1251] eta 0:00:38 lr 0.000057 time 0.2814 (0.2907) loss 2.5919 (3.0473) grad_norm 2.7518 (inf) [2021-04-16 17:20:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][1130/1251] eta 0:00:35 lr 0.000057 time 0.2661 (0.2906) loss 3.1103 (3.0453) grad_norm 3.0002 (inf) [2021-04-16 17:21:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][1140/1251] eta 0:00:32 lr 0.000057 time 0.2689 (0.2905) loss 2.3283 (3.0439) grad_norm 2.7384 (inf) [2021-04-16 17:21:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][1150/1251] eta 0:00:29 lr 0.000057 time 0.2692 (0.2905) loss 2.3513 (3.0436) grad_norm 2.8788 (inf) [2021-04-16 17:21:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][1160/1251] eta 0:00:26 lr 0.000057 time 0.2734 (0.2904) loss 3.2082 (3.0436) grad_norm 2.5277 (inf) [2021-04-16 17:21:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][1170/1251] eta 0:00:23 lr 0.000057 time 0.2663 (0.2903) loss 3.5839 (3.0424) grad_norm 2.7378 (inf) [2021-04-16 17:21:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][1180/1251] eta 0:00:20 lr 0.000057 time 0.2913 (0.2902) loss 3.4026 (3.0417) grad_norm 3.0821 (inf) [2021-04-16 17:21:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][1190/1251] eta 0:00:17 lr 0.000057 time 0.2541 (0.2901) loss 3.4426 (3.0409) grad_norm 2.6025 (inf) [2021-04-16 17:21:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][1200/1251] eta 0:00:14 lr 0.000057 time 0.2952 (0.2900) loss 3.2875 (3.0390) grad_norm 2.7271 (inf) [2021-04-16 17:21:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][1210/1251] eta 0:00:11 lr 0.000057 time 0.2863 (0.2899) loss 3.4822 (3.0404) grad_norm 3.1147 (inf) [2021-04-16 17:21:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][1220/1251] eta 0:00:08 lr 0.000057 time 0.2949 (0.2898) loss 3.6936 (3.0403) grad_norm 3.2658 (inf) [2021-04-16 17:21:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][1230/1251] eta 0:00:06 lr 0.000057 time 0.2653 (0.2897) loss 3.3395 (3.0414) grad_norm 2.9237 (inf) [2021-04-16 17:21:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][1240/1251] eta 0:00:03 lr 0.000057 time 0.2528 (0.2895) loss 2.1387 (3.0393) grad_norm 2.7252 (inf) [2021-04-16 17:21:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [257/300][1250/1251] eta 0:00:00 lr 0.000057 time 0.2500 (0.2892) loss 3.2481 (3.0378) grad_norm 3.1184 (inf) [2021-04-16 17:21:47 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 257 training takes 0:06:18 [2021-04-16 17:21:47 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_257.pth saving...... [2021-04-16 17:22:10 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_257.pth saved !!! [2021-04-16 17:22:12 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.100 (1.100) Loss 0.7940 (0.7940) Acc@1 80.371 (80.371) Acc@5 95.898 (95.898) [2021-04-16 17:22:13 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.110 (0.264) Loss 0.8209 (0.8324) Acc@1 81.641 (80.824) Acc@5 96.094 (95.339) [2021-04-16 17:22:15 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.107 (0.233) Loss 0.8158 (0.8207) Acc@1 80.371 (80.878) Acc@5 95.312 (95.466) [2021-04-16 17:22:18 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.099 (0.236) Loss 0.7991 (0.8309) Acc@1 82.031 (80.522) Acc@5 95.898 (95.369) [2021-04-16 17:22:20 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.221) Loss 0.8436 (0.8305) Acc@1 80.176 (80.616) Acc@5 94.336 (95.365) [2021-04-16 17:22:31 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.624 Acc@5 95.376 [2021-04-16 17:22:31 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 80.6% [2021-04-16 17:22:31 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 80.62% [2021-04-16 17:22:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][0/1251] eta 3:53:57 lr 0.000057 time 11.2208 (11.2208) loss 3.0830 (3.0830) grad_norm 2.7804 (2.7804) [2021-04-16 17:22:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][10/1251] eta 0:26:13 lr 0.000057 time 0.2470 (1.2683) loss 2.7732 (3.2343) grad_norm 4.9527 (3.0273) [2021-04-16 17:22:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][20/1251] eta 0:16:18 lr 0.000057 time 0.2451 (0.7949) loss 3.4004 (3.3326) grad_norm 2.6197 (3.0295) [2021-04-16 17:22:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][30/1251] eta 0:12:46 lr 0.000057 time 0.2841 (0.6279) loss 3.0389 (3.2627) grad_norm 2.5053 (2.9438) [2021-04-16 17:22:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][40/1251] eta 0:10:57 lr 0.000057 time 0.2917 (0.5433) loss 2.8000 (3.2151) grad_norm 3.0085 (2.9372) [2021-04-16 17:22:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][50/1251] eta 0:09:50 lr 0.000057 time 0.3086 (0.4915) loss 3.2401 (3.2266) grad_norm 3.3656 (2.9273) [2021-04-16 17:22:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][60/1251] eta 0:09:03 lr 0.000057 time 0.2842 (0.4566) loss 3.0206 (3.1940) grad_norm 2.9461 (2.9383) [2021-04-16 17:23:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][70/1251] eta 0:08:29 lr 0.000057 time 0.2714 (0.4318) loss 2.6104 (3.1304) grad_norm 4.5719 (2.9506) [2021-04-16 17:23:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][80/1251] eta 0:08:02 lr 0.000057 time 0.2800 (0.4121) loss 3.7242 (3.1314) grad_norm 2.8823 (2.9651) [2021-04-16 17:23:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][90/1251] eta 0:07:41 lr 0.000057 time 0.2959 (0.3972) loss 3.6440 (3.1189) grad_norm 3.3152 (2.9617) [2021-04-16 17:23:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][100/1251] eta 0:07:22 lr 0.000057 time 0.2703 (0.3849) loss 3.0975 (3.1292) grad_norm 2.5779 (2.9385) [2021-04-16 17:23:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][110/1251] eta 0:07:07 lr 0.000057 time 0.2668 (0.3750) loss 2.6339 (3.1114) grad_norm 2.4464 (2.9096) [2021-04-16 17:23:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][120/1251] eta 0:06:54 lr 0.000057 time 0.2777 (0.3668) loss 3.2159 (3.0907) grad_norm 3.0740 (2.9137) [2021-04-16 17:23:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][130/1251] eta 0:06:43 lr 0.000057 time 0.2890 (0.3597) loss 2.6810 (3.0891) grad_norm 2.8156 (2.9298) [2021-04-16 17:23:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][140/1251] eta 0:06:34 lr 0.000057 time 0.2747 (0.3553) loss 3.5663 (3.1018) grad_norm 2.8382 (2.9284) [2021-04-16 17:23:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][150/1251] eta 0:06:25 lr 0.000057 time 0.2531 (0.3505) loss 2.7182 (3.0967) grad_norm 2.8520 (2.9323) [2021-04-16 17:23:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][160/1251] eta 0:06:17 lr 0.000057 time 0.3008 (0.3464) loss 3.6910 (3.0979) grad_norm 2.3520 (2.9254) [2021-04-16 17:23:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][170/1251] eta 0:06:09 lr 0.000057 time 0.2635 (0.3423) loss 3.1106 (3.0942) grad_norm 2.5581 (2.9260) [2021-04-16 17:23:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][180/1251] eta 0:06:03 lr 0.000057 time 0.2964 (0.3396) loss 3.1640 (3.0921) grad_norm 2.8440 (2.9285) [2021-04-16 17:23:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][190/1251] eta 0:05:57 lr 0.000057 time 0.2814 (0.3368) loss 3.4360 (3.0904) grad_norm 2.9174 (2.9213) [2021-04-16 17:23:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][200/1251] eta 0:05:50 lr 0.000057 time 0.2842 (0.3338) loss 2.3550 (3.0843) grad_norm 2.5673 (2.9096) [2021-04-16 17:23:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][210/1251] eta 0:05:44 lr 0.000057 time 0.2880 (0.3311) loss 2.6090 (3.0928) grad_norm 2.8007 (2.9031) [2021-04-16 17:23:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][220/1251] eta 0:05:39 lr 0.000057 time 0.2884 (0.3292) loss 3.5120 (3.0913) grad_norm 2.7237 (2.9043) [2021-04-16 17:23:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][230/1251] eta 0:05:33 lr 0.000057 time 0.2753 (0.3268) loss 2.3265 (3.0780) grad_norm 3.4540 (2.8994) [2021-04-16 17:23:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][240/1251] eta 0:05:28 lr 0.000057 time 0.2820 (0.3247) loss 2.8177 (3.0741) grad_norm 4.0797 (2.9065) [2021-04-16 17:23:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][250/1251] eta 0:05:23 lr 0.000057 time 0.2826 (0.3228) loss 3.4131 (3.0765) grad_norm 3.1860 (2.9143) [2021-04-16 17:23:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][260/1251] eta 0:05:18 lr 0.000057 time 0.4248 (0.3216) loss 2.5516 (3.0715) grad_norm 3.3932 (2.9212) [2021-04-16 17:23:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][270/1251] eta 0:05:13 lr 0.000057 time 0.2763 (0.3198) loss 3.4840 (3.0710) grad_norm 3.8767 (2.9320) [2021-04-16 17:24:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][280/1251] eta 0:05:09 lr 0.000057 time 0.2818 (0.3183) loss 3.4257 (3.0655) grad_norm 3.0510 (2.9307) [2021-04-16 17:24:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][290/1251] eta 0:05:04 lr 0.000057 time 0.2762 (0.3169) loss 3.0330 (3.0673) grad_norm 2.9208 (2.9375) [2021-04-16 17:24:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][300/1251] eta 0:05:00 lr 0.000057 time 0.2795 (0.3156) loss 3.2867 (3.0695) grad_norm 2.6797 (2.9364) [2021-04-16 17:24:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][310/1251] eta 0:04:55 lr 0.000057 time 0.2753 (0.3143) loss 3.1934 (3.0711) grad_norm 3.5498 (2.9343) [2021-04-16 17:24:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][320/1251] eta 0:04:51 lr 0.000057 time 0.2808 (0.3132) loss 3.2312 (3.0716) grad_norm 3.1941 (2.9343) [2021-04-16 17:24:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][330/1251] eta 0:04:47 lr 0.000057 time 0.2511 (0.3120) loss 3.6747 (3.0764) grad_norm 2.7507 (2.9323) [2021-04-16 17:24:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][340/1251] eta 0:04:43 lr 0.000057 time 0.2848 (0.3109) loss 2.4651 (3.0704) grad_norm 2.5360 (2.9261) [2021-04-16 17:24:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][350/1251] eta 0:04:39 lr 0.000056 time 0.2836 (0.3101) loss 3.1951 (3.0689) grad_norm 2.7608 (2.9244) [2021-04-16 17:24:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][360/1251] eta 0:04:35 lr 0.000056 time 0.2534 (0.3096) loss 3.3295 (3.0701) grad_norm 2.6557 (2.9368) [2021-04-16 17:24:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][370/1251] eta 0:04:32 lr 0.000056 time 0.2859 (0.3093) loss 3.6651 (3.0704) grad_norm 2.6678 (2.9321) [2021-04-16 17:24:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][380/1251] eta 0:04:28 lr 0.000056 time 0.2835 (0.3085) loss 3.5443 (3.0750) grad_norm 3.3643 (2.9356) [2021-04-16 17:24:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][390/1251] eta 0:04:24 lr 0.000056 time 0.2593 (0.3075) loss 2.8070 (3.0773) grad_norm 3.0064 (2.9360) [2021-04-16 17:24:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][400/1251] eta 0:04:21 lr 0.000056 time 0.2705 (0.3068) loss 3.5715 (3.0836) grad_norm 2.9483 (2.9429) [2021-04-16 17:24:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][410/1251] eta 0:04:17 lr 0.000056 time 0.2838 (0.3062) loss 2.4648 (3.0855) grad_norm 2.8379 (2.9404) [2021-04-16 17:24:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][420/1251] eta 0:04:13 lr 0.000056 time 0.2815 (0.3055) loss 2.2502 (3.0807) grad_norm 2.9409 (2.9397) [2021-04-16 17:24:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][430/1251] eta 0:04:10 lr 0.000056 time 0.2853 (0.3049) loss 2.4607 (3.0795) grad_norm 2.6175 (2.9407) [2021-04-16 17:24:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][440/1251] eta 0:04:06 lr 0.000056 time 0.2740 (0.3045) loss 2.9368 (3.0722) grad_norm 5.8019 (2.9475) [2021-04-16 17:24:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][450/1251] eta 0:04:03 lr 0.000056 time 0.2955 (0.3039) loss 3.2506 (3.0717) grad_norm 2.6438 (2.9482) [2021-04-16 17:24:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][460/1251] eta 0:03:59 lr 0.000056 time 0.2727 (0.3032) loss 2.5365 (3.0660) grad_norm 2.5977 (2.9477) [2021-04-16 17:24:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][470/1251] eta 0:03:56 lr 0.000056 time 0.2540 (0.3026) loss 3.3151 (3.0690) grad_norm 2.9556 (2.9496) [2021-04-16 17:24:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][480/1251] eta 0:03:53 lr 0.000056 time 0.2674 (0.3022) loss 3.4406 (3.0686) grad_norm 3.5131 (2.9589) [2021-04-16 17:24:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][490/1251] eta 0:03:49 lr 0.000056 time 0.2698 (0.3017) loss 2.5685 (3.0684) grad_norm 3.4395 (2.9616) [2021-04-16 17:25:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][500/1251] eta 0:03:46 lr 0.000056 time 0.2584 (0.3011) loss 3.6449 (3.0723) grad_norm 3.3008 (2.9625) [2021-04-16 17:25:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][510/1251] eta 0:03:42 lr 0.000056 time 0.2679 (0.3005) loss 3.4114 (3.0752) grad_norm 3.2114 (2.9655) [2021-04-16 17:25:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][520/1251] eta 0:03:39 lr 0.000056 time 0.2708 (0.3000) loss 3.4354 (3.0767) grad_norm 2.5170 (2.9668) [2021-04-16 17:25:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][530/1251] eta 0:03:35 lr 0.000056 time 0.2728 (0.2995) loss 3.4387 (3.0769) grad_norm 2.9659 (2.9647) [2021-04-16 17:25:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][540/1251] eta 0:03:32 lr 0.000056 time 0.2797 (0.2994) loss 2.9123 (3.0801) grad_norm 2.6465 (2.9635) [2021-04-16 17:25:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][550/1251] eta 0:03:29 lr 0.000056 time 0.3215 (0.2994) loss 2.6858 (3.0811) grad_norm 3.6886 (2.9631) [2021-04-16 17:25:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][560/1251] eta 0:03:26 lr 0.000056 time 0.2805 (0.2990) loss 2.9503 (3.0793) grad_norm 3.3090 (2.9621) [2021-04-16 17:25:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][570/1251] eta 0:03:23 lr 0.000056 time 0.2558 (0.2991) loss 2.1910 (3.0771) grad_norm 2.8527 (2.9611) [2021-04-16 17:25:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][580/1251] eta 0:03:20 lr 0.000056 time 0.2875 (0.2990) loss 3.5040 (3.0794) grad_norm 2.7460 (2.9626) [2021-04-16 17:25:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][590/1251] eta 0:03:17 lr 0.000056 time 0.2898 (0.2986) loss 3.7418 (3.0800) grad_norm 2.9195 (2.9629) [2021-04-16 17:25:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][600/1251] eta 0:03:14 lr 0.000056 time 0.2865 (0.2983) loss 2.7810 (3.0796) grad_norm 3.0279 (2.9630) [2021-04-16 17:25:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][610/1251] eta 0:03:11 lr 0.000056 time 0.2862 (0.2980) loss 3.1807 (3.0779) grad_norm 2.7802 (2.9605) [2021-04-16 17:25:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][620/1251] eta 0:03:07 lr 0.000056 time 0.2857 (0.2976) loss 3.5525 (3.0822) grad_norm 2.7908 (2.9594) [2021-04-16 17:25:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][630/1251] eta 0:03:04 lr 0.000056 time 0.2615 (0.2972) loss 2.7701 (3.0812) grad_norm 3.1759 (2.9594) [2021-04-16 17:25:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][640/1251] eta 0:03:01 lr 0.000056 time 0.2758 (0.2969) loss 3.5158 (3.0819) grad_norm 2.7238 (2.9581) [2021-04-16 17:25:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][650/1251] eta 0:02:58 lr 0.000056 time 0.2877 (0.2966) loss 3.0742 (3.0784) grad_norm 2.6778 (2.9577) [2021-04-16 17:25:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][660/1251] eta 0:02:55 lr 0.000056 time 0.2744 (0.2963) loss 1.6339 (3.0752) grad_norm 3.0040 (2.9574) [2021-04-16 17:25:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][670/1251] eta 0:02:51 lr 0.000056 time 0.2648 (0.2960) loss 3.2918 (3.0783) grad_norm 2.8145 (2.9572) [2021-04-16 17:25:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][680/1251] eta 0:02:48 lr 0.000056 time 0.2584 (0.2957) loss 3.1930 (3.0787) grad_norm 3.2051 (2.9580) [2021-04-16 17:25:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][690/1251] eta 0:02:45 lr 0.000056 time 0.2664 (0.2954) loss 3.2317 (3.0808) grad_norm 2.9207 (2.9559) [2021-04-16 17:25:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][700/1251] eta 0:02:42 lr 0.000056 time 0.2663 (0.2951) loss 2.3353 (3.0788) grad_norm 3.0198 (2.9586) [2021-04-16 17:26:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][710/1251] eta 0:02:39 lr 0.000056 time 0.2854 (0.2948) loss 3.3522 (3.0768) grad_norm 2.3091 (2.9587) [2021-04-16 17:26:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][720/1251] eta 0:02:36 lr 0.000056 time 0.2827 (0.2949) loss 3.1801 (3.0762) grad_norm 2.5978 (2.9565) [2021-04-16 17:26:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][730/1251] eta 0:02:33 lr 0.000056 time 0.2444 (0.2946) loss 2.1411 (3.0777) grad_norm 2.8011 (2.9568) [2021-04-16 17:26:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][740/1251] eta 0:02:30 lr 0.000056 time 0.2877 (0.2944) loss 2.1925 (3.0717) grad_norm 3.5312 (2.9557) [2021-04-16 17:26:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][750/1251] eta 0:02:27 lr 0.000056 time 0.2720 (0.2941) loss 2.3745 (3.0676) grad_norm 2.6302 (2.9533) [2021-04-16 17:26:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][760/1251] eta 0:02:24 lr 0.000056 time 0.2792 (0.2941) loss 3.6971 (3.0677) grad_norm 3.1836 (2.9537) [2021-04-16 17:26:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][770/1251] eta 0:02:21 lr 0.000056 time 0.2704 (0.2938) loss 2.2378 (3.0692) grad_norm 3.0275 (2.9528) [2021-04-16 17:26:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][780/1251] eta 0:02:18 lr 0.000056 time 0.2685 (0.2935) loss 3.4541 (3.0667) grad_norm 2.4834 (2.9517) [2021-04-16 17:26:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][790/1251] eta 0:02:15 lr 0.000056 time 0.2630 (0.2935) loss 3.4153 (3.0680) grad_norm 3.3473 (2.9580) [2021-04-16 17:26:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][800/1251] eta 0:02:12 lr 0.000056 time 0.2785 (0.2933) loss 3.2072 (3.0696) grad_norm 2.6595 (2.9581) [2021-04-16 17:26:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][810/1251] eta 0:02:09 lr 0.000056 time 0.2802 (0.2931) loss 3.3294 (3.0697) grad_norm 2.5031 (2.9570) [2021-04-16 17:26:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][820/1251] eta 0:02:06 lr 0.000056 time 0.2631 (0.2929) loss 2.0744 (3.0676) grad_norm 2.5816 (2.9559) [2021-04-16 17:26:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][830/1251] eta 0:02:03 lr 0.000056 time 0.2890 (0.2927) loss 2.0985 (3.0666) grad_norm 2.7029 (2.9553) [2021-04-16 17:26:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][840/1251] eta 0:02:00 lr 0.000056 time 0.2836 (0.2925) loss 2.9417 (3.0665) grad_norm 3.6396 (2.9576) [2021-04-16 17:26:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][850/1251] eta 0:01:57 lr 0.000056 time 0.2423 (0.2923) loss 3.8405 (3.0675) grad_norm 3.5860 (2.9584) [2021-04-16 17:26:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][860/1251] eta 0:01:54 lr 0.000056 time 0.3157 (0.2922) loss 2.6028 (3.0654) grad_norm 2.6724 (2.9565) [2021-04-16 17:26:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][870/1251] eta 0:01:51 lr 0.000056 time 0.2833 (0.2920) loss 2.6327 (3.0651) grad_norm 2.8233 (2.9582) [2021-04-16 17:26:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][880/1251] eta 0:01:48 lr 0.000056 time 0.2790 (0.2918) loss 2.9985 (3.0643) grad_norm 3.6127 (2.9590) [2021-04-16 17:26:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][890/1251] eta 0:01:45 lr 0.000056 time 0.2796 (0.2916) loss 3.0593 (3.0680) grad_norm 3.0813 (2.9574) [2021-04-16 17:26:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][900/1251] eta 0:01:42 lr 0.000056 time 0.2683 (0.2914) loss 3.1777 (3.0670) grad_norm 3.0680 (2.9573) [2021-04-16 17:26:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][910/1251] eta 0:01:39 lr 0.000056 time 0.2748 (0.2912) loss 2.3806 (3.0688) grad_norm 2.8617 (2.9595) [2021-04-16 17:26:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][920/1251] eta 0:01:36 lr 0.000056 time 0.2935 (0.2910) loss 3.1456 (3.0701) grad_norm 3.3616 (2.9592) [2021-04-16 17:27:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][930/1251] eta 0:01:33 lr 0.000055 time 0.2489 (0.2909) loss 3.4103 (3.0714) grad_norm 2.8157 (2.9583) [2021-04-16 17:27:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][940/1251] eta 0:01:30 lr 0.000055 time 0.3268 (0.2911) loss 3.2913 (3.0742) grad_norm 2.8170 (2.9562) [2021-04-16 17:27:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][950/1251] eta 0:01:27 lr 0.000055 time 0.2743 (0.2910) loss 2.2263 (3.0736) grad_norm 3.0210 (2.9599) [2021-04-16 17:27:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][960/1251] eta 0:01:24 lr 0.000055 time 0.2907 (0.2910) loss 3.2646 (3.0739) grad_norm 2.8394 (2.9595) [2021-04-16 17:27:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][970/1251] eta 0:01:21 lr 0.000055 time 0.2759 (0.2908) loss 2.1400 (3.0727) grad_norm 2.5040 (2.9616) [2021-04-16 17:27:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][980/1251] eta 0:01:18 lr 0.000055 time 0.2576 (0.2907) loss 3.6176 (3.0756) grad_norm 3.9327 (2.9617) [2021-04-16 17:27:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][990/1251] eta 0:01:15 lr 0.000055 time 0.2743 (0.2905) loss 3.1909 (3.0769) grad_norm 3.4564 (2.9636) [2021-04-16 17:27:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][1000/1251] eta 0:01:12 lr 0.000055 time 0.2933 (0.2904) loss 3.1917 (3.0762) grad_norm 3.1896 (2.9622) [2021-04-16 17:27:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][1010/1251] eta 0:01:09 lr 0.000055 time 0.2873 (0.2902) loss 3.8324 (3.0771) grad_norm 2.8947 (2.9633) [2021-04-16 17:27:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][1020/1251] eta 0:01:07 lr 0.000055 time 0.2563 (0.2902) loss 3.6119 (3.0782) grad_norm 2.8849 (2.9619) [2021-04-16 17:27:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][1030/1251] eta 0:01:04 lr 0.000055 time 0.3008 (0.2900) loss 3.0964 (3.0810) grad_norm 2.5628 (2.9618) [2021-04-16 17:27:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][1040/1251] eta 0:01:01 lr 0.000055 time 0.2614 (0.2899) loss 3.1284 (3.0837) grad_norm 3.2387 (2.9614) [2021-04-16 17:27:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][1050/1251] eta 0:00:58 lr 0.000055 time 0.2807 (0.2898) loss 2.5471 (3.0820) grad_norm 3.3696 (2.9615) [2021-04-16 17:27:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][1060/1251] eta 0:00:55 lr 0.000055 time 0.2656 (0.2897) loss 3.2891 (3.0809) grad_norm 2.9244 (2.9622) [2021-04-16 17:27:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][1070/1251] eta 0:00:52 lr 0.000055 time 0.2702 (0.2895) loss 3.4303 (3.0801) grad_norm 2.7387 (2.9605) [2021-04-16 17:27:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][1080/1251] eta 0:00:49 lr 0.000055 time 0.2662 (0.2894) loss 3.0946 (3.0820) grad_norm 2.7081 (2.9610) [2021-04-16 17:27:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][1090/1251] eta 0:00:46 lr 0.000055 time 0.2703 (0.2892) loss 3.4603 (3.0812) grad_norm 3.2661 (2.9613) [2021-04-16 17:27:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][1100/1251] eta 0:00:43 lr 0.000055 time 0.2914 (0.2891) loss 3.2586 (3.0799) grad_norm 2.6849 (2.9606) [2021-04-16 17:27:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][1110/1251] eta 0:00:40 lr 0.000055 time 0.2787 (0.2890) loss 2.8811 (3.0792) grad_norm 2.8782 (2.9608) [2021-04-16 17:27:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][1120/1251] eta 0:00:37 lr 0.000055 time 0.2699 (0.2889) loss 2.8094 (3.0753) grad_norm 3.3288 (2.9601) [2021-04-16 17:27:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][1130/1251] eta 0:00:34 lr 0.000055 time 0.2782 (0.2890) loss 3.2805 (3.0766) grad_norm 2.4176 (2.9583) [2021-04-16 17:28:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][1140/1251] eta 0:00:32 lr 0.000055 time 0.2656 (0.2890) loss 2.3180 (3.0777) grad_norm 3.3193 (2.9577) [2021-04-16 17:28:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][1150/1251] eta 0:00:29 lr 0.000055 time 0.2645 (0.2890) loss 3.4696 (3.0778) grad_norm 3.2197 (2.9562) [2021-04-16 17:28:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][1160/1251] eta 0:00:26 lr 0.000055 time 0.2759 (0.2889) loss 2.3966 (3.0748) grad_norm 2.4974 (2.9579) [2021-04-16 17:28:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][1170/1251] eta 0:00:23 lr 0.000055 time 0.2871 (0.2888) loss 2.9392 (3.0718) grad_norm 3.9346 (2.9591) [2021-04-16 17:28:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][1180/1251] eta 0:00:20 lr 0.000055 time 0.2738 (0.2887) loss 3.6970 (3.0708) grad_norm 2.9090 (2.9583) [2021-04-16 17:28:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][1190/1251] eta 0:00:17 lr 0.000055 time 0.2828 (0.2887) loss 3.4084 (3.0680) grad_norm 3.2011 (2.9574) [2021-04-16 17:28:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][1200/1251] eta 0:00:14 lr 0.000055 time 0.2825 (0.2886) loss 2.9024 (3.0649) grad_norm 3.0324 (2.9583) [2021-04-16 17:28:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][1210/1251] eta 0:00:11 lr 0.000055 time 0.2944 (0.2886) loss 2.5414 (3.0630) grad_norm 2.8316 (2.9597) [2021-04-16 17:28:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][1220/1251] eta 0:00:08 lr 0.000055 time 0.2721 (0.2884) loss 3.2347 (3.0624) grad_norm 4.1431 (2.9607) [2021-04-16 17:28:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][1230/1251] eta 0:00:06 lr 0.000055 time 0.2787 (0.2883) loss 3.1303 (3.0630) grad_norm 2.7137 (2.9612) [2021-04-16 17:28:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][1240/1251] eta 0:00:03 lr 0.000055 time 0.2478 (0.2881) loss 2.2938 (3.0631) grad_norm 3.6461 (2.9614) [2021-04-16 17:28:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [258/300][1250/1251] eta 0:00:00 lr 0.000055 time 0.2482 (0.2878) loss 3.4303 (3.0631) grad_norm 2.9346 (2.9625) [2021-04-16 17:28:36 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 258 training takes 0:06:04 [2021-04-16 17:28:36 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_258.pth saving...... [2021-04-16 17:28:49 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_258.pth saved !!! [2021-04-16 17:28:54 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 4.423 (4.423) Loss 0.9018 (0.9018) Acc@1 79.297 (79.297) Acc@5 93.945 (93.945) [2021-04-16 17:28:55 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.105 (0.502) Loss 0.8434 (0.8495) Acc@1 81.738 (80.531) Acc@5 94.922 (95.153) [2021-04-16 17:28:58 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.144 (0.399) Loss 0.8976 (0.8405) Acc@1 79.004 (80.585) Acc@5 94.922 (95.196) [2021-04-16 17:29:00 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.107 (0.339) Loss 0.8911 (0.8401) Acc@1 79.883 (80.604) Acc@5 94.336 (95.240) [2021-04-16 17:29:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.298) Loss 0.8525 (0.8363) Acc@1 80.176 (80.590) Acc@5 94.922 (95.310) [2021-04-16 17:29:20 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.584 Acc@5 95.342 [2021-04-16 17:29:20 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 80.6% [2021-04-16 17:29:20 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 80.62% [2021-04-16 17:29:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][0/1251] eta 1:56:34 lr 0.000055 time 5.5913 (5.5913) loss 2.4566 (2.4566) grad_norm 2.8334 (2.8334) [2021-04-16 17:29:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][10/1251] eta 0:15:44 lr 0.000055 time 0.3650 (0.7609) loss 2.8494 (3.0840) grad_norm 2.4394 (2.7209) [2021-04-16 17:29:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][20/1251] eta 0:10:52 lr 0.000055 time 0.2811 (0.5299) loss 3.5260 (3.1113) grad_norm 3.4417 (2.8209) [2021-04-16 17:29:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][30/1251] eta 0:09:08 lr 0.000055 time 0.2975 (0.4492) loss 3.5905 (3.1479) grad_norm 2.5308 (2.7678) [2021-04-16 17:29:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][40/1251] eta 0:08:17 lr 0.000055 time 0.2927 (0.4110) loss 2.0270 (3.1393) grad_norm 2.5835 (2.7650) [2021-04-16 17:29:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][50/1251] eta 0:07:40 lr 0.000055 time 0.2714 (0.3836) loss 3.3433 (3.1231) grad_norm 2.6205 (2.7944) [2021-04-16 17:29:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][60/1251] eta 0:07:15 lr 0.000055 time 0.2738 (0.3657) loss 3.4926 (3.1408) grad_norm 2.4954 (2.7913) [2021-04-16 17:29:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][70/1251] eta 0:06:56 lr 0.000055 time 0.2589 (0.3530) loss 3.2463 (3.1529) grad_norm 2.7128 (2.7995) [2021-04-16 17:29:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][80/1251] eta 0:06:43 lr 0.000055 time 0.3031 (0.3448) loss 3.4061 (3.1435) grad_norm 2.4950 (2.7975) [2021-04-16 17:29:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][90/1251] eta 0:06:31 lr 0.000055 time 0.2593 (0.3371) loss 3.3073 (3.1264) grad_norm 2.5686 (2.7925) [2021-04-16 17:29:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][100/1251] eta 0:06:21 lr 0.000055 time 0.2899 (0.3312) loss 2.8524 (3.1210) grad_norm 3.0072 (2.8199) [2021-04-16 17:29:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][110/1251] eta 0:06:11 lr 0.000055 time 0.2743 (0.3259) loss 2.8891 (3.1374) grad_norm 2.8759 (2.8211) [2021-04-16 17:29:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][120/1251] eta 0:06:03 lr 0.000055 time 0.2800 (0.3216) loss 2.0971 (3.1357) grad_norm 2.7743 (2.8277) [2021-04-16 17:30:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][130/1251] eta 0:05:57 lr 0.000055 time 0.2744 (0.3189) loss 2.4920 (3.1348) grad_norm 3.0282 (2.8432) [2021-04-16 17:30:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][140/1251] eta 0:05:52 lr 0.000055 time 0.3145 (0.3171) loss 3.2355 (3.1111) grad_norm 3.7103 (2.8870) [2021-04-16 17:30:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][150/1251] eta 0:05:47 lr 0.000055 time 0.3893 (0.3160) loss 2.1825 (3.1048) grad_norm 2.4134 (2.8835) [2021-04-16 17:30:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][160/1251] eta 0:05:41 lr 0.000055 time 0.2774 (0.3133) loss 2.8025 (3.1015) grad_norm 2.3291 (2.8899) [2021-04-16 17:30:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][170/1251] eta 0:05:36 lr 0.000055 time 0.2806 (0.3111) loss 2.9968 (3.1042) grad_norm 2.5794 (2.9002) [2021-04-16 17:30:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][180/1251] eta 0:05:31 lr 0.000055 time 0.2840 (0.3095) loss 2.5184 (3.1008) grad_norm 2.8969 (2.9023) [2021-04-16 17:30:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][190/1251] eta 0:05:26 lr 0.000055 time 0.2884 (0.3081) loss 2.5437 (3.0899) grad_norm 4.4129 (2.8966) [2021-04-16 17:30:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][200/1251] eta 0:05:22 lr 0.000055 time 0.2727 (0.3065) loss 2.0656 (3.0803) grad_norm 2.8348 (2.9067) [2021-04-16 17:30:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][210/1251] eta 0:05:17 lr 0.000055 time 0.2831 (0.3048) loss 3.5269 (3.0858) grad_norm 3.5403 (2.9077) [2021-04-16 17:30:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][220/1251] eta 0:05:13 lr 0.000055 time 0.3029 (0.3038) loss 2.7340 (3.0750) grad_norm 2.8198 (2.9085) [2021-04-16 17:30:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][230/1251] eta 0:05:09 lr 0.000055 time 0.2612 (0.3033) loss 2.2343 (3.0680) grad_norm 3.3811 (2.9115) [2021-04-16 17:30:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][240/1251] eta 0:05:05 lr 0.000055 time 0.2444 (0.3021) loss 3.7823 (3.0741) grad_norm 3.7893 (2.9134) [2021-04-16 17:30:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][250/1251] eta 0:05:01 lr 0.000054 time 0.2746 (0.3010) loss 2.4583 (3.0717) grad_norm 3.2149 (2.9332) [2021-04-16 17:30:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][260/1251] eta 0:04:57 lr 0.000054 time 0.2671 (0.3000) loss 2.4368 (3.0702) grad_norm 2.7168 (2.9517) [2021-04-16 17:30:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][270/1251] eta 0:04:53 lr 0.000054 time 0.2601 (0.2993) loss 3.6895 (3.0755) grad_norm 3.4305 (2.9535) [2021-04-16 17:30:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][280/1251] eta 0:04:49 lr 0.000054 time 0.2789 (0.2987) loss 2.4925 (3.0693) grad_norm 2.9652 (2.9528) [2021-04-16 17:30:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][290/1251] eta 0:04:46 lr 0.000054 time 0.2724 (0.2978) loss 3.1870 (3.0768) grad_norm 3.0205 (2.9524) [2021-04-16 17:30:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][300/1251] eta 0:04:42 lr 0.000054 time 0.2816 (0.2971) loss 2.8631 (3.0736) grad_norm 2.9043 (2.9490) [2021-04-16 17:30:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][310/1251] eta 0:04:39 lr 0.000054 time 0.2783 (0.2967) loss 3.6053 (3.0720) grad_norm 2.6615 (2.9503) [2021-04-16 17:30:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][320/1251] eta 0:04:35 lr 0.000054 time 0.2664 (0.2960) loss 2.8279 (3.0748) grad_norm 2.6504 (2.9519) [2021-04-16 17:30:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][330/1251] eta 0:04:32 lr 0.000054 time 0.2957 (0.2959) loss 2.9637 (3.0687) grad_norm 3.6971 (2.9552) [2021-04-16 17:31:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][340/1251] eta 0:04:29 lr 0.000054 time 0.2878 (0.2956) loss 3.2571 (3.0655) grad_norm 3.3379 (2.9574) [2021-04-16 17:31:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][350/1251] eta 0:04:26 lr 0.000054 time 0.2740 (0.2952) loss 3.1756 (3.0660) grad_norm 2.8180 (2.9658) [2021-04-16 17:31:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][360/1251] eta 0:04:23 lr 0.000054 time 0.2942 (0.2953) loss 3.5254 (3.0742) grad_norm 3.4484 (2.9650) [2021-04-16 17:31:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][370/1251] eta 0:04:19 lr 0.000054 time 0.2618 (0.2948) loss 2.9623 (3.0627) grad_norm 2.7526 (2.9633) [2021-04-16 17:31:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][380/1251] eta 0:04:16 lr 0.000054 time 0.2683 (0.2943) loss 3.3095 (3.0637) grad_norm 2.5938 (2.9633) [2021-04-16 17:31:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][390/1251] eta 0:04:13 lr 0.000054 time 0.2774 (0.2939) loss 3.0762 (3.0600) grad_norm 3.7161 (2.9727) [2021-04-16 17:31:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][400/1251] eta 0:04:09 lr 0.000054 time 0.2796 (0.2934) loss 2.9792 (3.0538) grad_norm 4.0198 (2.9807) [2021-04-16 17:31:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][410/1251] eta 0:04:06 lr 0.000054 time 0.2612 (0.2928) loss 3.1951 (3.0602) grad_norm 2.6270 (2.9775) [2021-04-16 17:31:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][420/1251] eta 0:04:03 lr 0.000054 time 0.2561 (0.2924) loss 3.1242 (3.0587) grad_norm 3.0748 (2.9725) [2021-04-16 17:31:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][430/1251] eta 0:03:59 lr 0.000054 time 0.2653 (0.2921) loss 2.6475 (3.0654) grad_norm 2.5004 (2.9714) [2021-04-16 17:31:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][440/1251] eta 0:03:56 lr 0.000054 time 0.2774 (0.2919) loss 3.7903 (3.0654) grad_norm 2.7208 (2.9706) [2021-04-16 17:31:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][450/1251] eta 0:03:53 lr 0.000054 time 0.2862 (0.2916) loss 2.0132 (3.0576) grad_norm 2.8446 (2.9703) [2021-04-16 17:31:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][460/1251] eta 0:03:50 lr 0.000054 time 0.3007 (0.2913) loss 3.2974 (3.0563) grad_norm 3.1937 (2.9695) [2021-04-16 17:31:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][470/1251] eta 0:03:47 lr 0.000054 time 0.2529 (0.2910) loss 1.8784 (3.0555) grad_norm 2.8432 (2.9675) [2021-04-16 17:31:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][480/1251] eta 0:03:44 lr 0.000054 time 0.2751 (0.2907) loss 3.7201 (3.0604) grad_norm 2.7982 (2.9678) [2021-04-16 17:31:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][490/1251] eta 0:03:40 lr 0.000054 time 0.2545 (0.2903) loss 3.6570 (3.0585) grad_norm 3.2353 (2.9676) [2021-04-16 17:31:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][500/1251] eta 0:03:37 lr 0.000054 time 0.2836 (0.2901) loss 3.1398 (3.0561) grad_norm 2.5633 (2.9658) [2021-04-16 17:31:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][510/1251] eta 0:03:34 lr 0.000054 time 0.2496 (0.2898) loss 3.7382 (3.0633) grad_norm 2.5782 (2.9634) [2021-04-16 17:31:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][520/1251] eta 0:03:31 lr 0.000054 time 0.2921 (0.2896) loss 3.0984 (3.0616) grad_norm 3.0065 (2.9689) [2021-04-16 17:31:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][530/1251] eta 0:03:28 lr 0.000054 time 0.2779 (0.2893) loss 3.0875 (3.0589) grad_norm 2.5683 (2.9708) [2021-04-16 17:31:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][540/1251] eta 0:03:25 lr 0.000054 time 0.2888 (0.2891) loss 2.4801 (3.0571) grad_norm 2.6329 (2.9692) [2021-04-16 17:31:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][550/1251] eta 0:03:22 lr 0.000054 time 0.2992 (0.2889) loss 3.3945 (3.0584) grad_norm 2.6703 (2.9698) [2021-04-16 17:32:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][560/1251] eta 0:03:19 lr 0.000054 time 0.2752 (0.2886) loss 3.3534 (3.0549) grad_norm 2.8931 (2.9665) [2021-04-16 17:32:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][570/1251] eta 0:03:16 lr 0.000054 time 0.2935 (0.2889) loss 2.5201 (3.0585) grad_norm 2.6202 (2.9710) [2021-04-16 17:32:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][580/1251] eta 0:03:13 lr 0.000054 time 0.2432 (0.2890) loss 1.9831 (3.0561) grad_norm 3.0987 (2.9705) [2021-04-16 17:32:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][590/1251] eta 0:03:11 lr 0.000054 time 0.2828 (0.2890) loss 2.5131 (3.0557) grad_norm 3.1589 (2.9712) [2021-04-16 17:32:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][600/1251] eta 0:03:08 lr 0.000054 time 0.2722 (0.2888) loss 3.1709 (3.0569) grad_norm 2.7290 (2.9700) [2021-04-16 17:32:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][610/1251] eta 0:03:04 lr 0.000054 time 0.2855 (0.2885) loss 3.3376 (3.0541) grad_norm 2.9830 (2.9713) [2021-04-16 17:32:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][620/1251] eta 0:03:02 lr 0.000054 time 0.2882 (0.2885) loss 3.1696 (3.0594) grad_norm 2.7269 (2.9686) [2021-04-16 17:32:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][630/1251] eta 0:02:59 lr 0.000054 time 0.2757 (0.2883) loss 2.2504 (3.0578) grad_norm 2.9776 (nan) [2021-04-16 17:32:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][640/1251] eta 0:02:56 lr 0.000054 time 0.2690 (0.2881) loss 3.6308 (3.0542) grad_norm 3.6566 (nan) [2021-04-16 17:32:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][650/1251] eta 0:02:53 lr 0.000054 time 0.3180 (0.2880) loss 3.4284 (3.0519) grad_norm 2.6227 (nan) [2021-04-16 17:32:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][660/1251] eta 0:02:50 lr 0.000054 time 0.2660 (0.2878) loss 3.2582 (3.0515) grad_norm 3.1716 (nan) [2021-04-16 17:32:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][670/1251] eta 0:02:47 lr 0.000054 time 0.2872 (0.2875) loss 3.3530 (3.0499) grad_norm 3.1108 (nan) [2021-04-16 17:32:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][680/1251] eta 0:02:44 lr 0.000054 time 0.2995 (0.2875) loss 2.8190 (3.0488) grad_norm 3.0080 (nan) [2021-04-16 17:32:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][690/1251] eta 0:02:41 lr 0.000054 time 0.2817 (0.2874) loss 3.5670 (3.0502) grad_norm 3.1809 (nan) [2021-04-16 17:32:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][700/1251] eta 0:02:38 lr 0.000054 time 0.3906 (0.2873) loss 3.0575 (3.0509) grad_norm 3.9338 (nan) [2021-04-16 17:32:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][710/1251] eta 0:02:35 lr 0.000054 time 0.2847 (0.2871) loss 3.3258 (3.0515) grad_norm 2.7876 (nan) [2021-04-16 17:32:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][720/1251] eta 0:02:32 lr 0.000054 time 0.2834 (0.2871) loss 3.2660 (3.0519) grad_norm 2.9038 (nan) [2021-04-16 17:32:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][730/1251] eta 0:02:29 lr 0.000054 time 0.2537 (0.2872) loss 2.0071 (3.0482) grad_norm 2.7568 (nan) [2021-04-16 17:32:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][740/1251] eta 0:02:26 lr 0.000054 time 0.2491 (0.2870) loss 2.2223 (3.0429) grad_norm 2.5530 (nan) [2021-04-16 17:32:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][750/1251] eta 0:02:23 lr 0.000054 time 0.2611 (0.2867) loss 2.8644 (3.0413) grad_norm 2.4547 (nan) [2021-04-16 17:32:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][760/1251] eta 0:02:20 lr 0.000054 time 0.3162 (0.2867) loss 3.7325 (3.0406) grad_norm 2.5738 (nan) [2021-04-16 17:33:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][770/1251] eta 0:02:17 lr 0.000054 time 0.2771 (0.2865) loss 2.4624 (3.0402) grad_norm 3.0993 (nan) [2021-04-16 17:33:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][780/1251] eta 0:02:14 lr 0.000054 time 0.2909 (0.2864) loss 2.7368 (3.0417) grad_norm 3.0659 (nan) [2021-04-16 17:33:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][790/1251] eta 0:02:11 lr 0.000054 time 0.2743 (0.2863) loss 2.9054 (3.0439) grad_norm 2.9887 (nan) [2021-04-16 17:33:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][800/1251] eta 0:02:09 lr 0.000054 time 0.4358 (0.2863) loss 2.9723 (3.0452) grad_norm 2.4372 (nan) [2021-04-16 17:33:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][810/1251] eta 0:02:06 lr 0.000054 time 0.2570 (0.2861) loss 3.5446 (3.0448) grad_norm 2.8405 (nan) [2021-04-16 17:33:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][820/1251] eta 0:02:03 lr 0.000054 time 0.3238 (0.2861) loss 3.1128 (3.0446) grad_norm 3.2492 (nan) [2021-04-16 17:33:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][830/1251] eta 0:02:00 lr 0.000054 time 0.2639 (0.2860) loss 3.4346 (3.0471) grad_norm 3.1791 (nan) [2021-04-16 17:33:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][840/1251] eta 0:01:57 lr 0.000053 time 0.2454 (0.2858) loss 3.6609 (3.0483) grad_norm 2.8041 (nan) [2021-04-16 17:33:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][850/1251] eta 0:01:54 lr 0.000053 time 0.2940 (0.2857) loss 3.7024 (3.0488) grad_norm 2.7325 (nan) [2021-04-16 17:33:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][860/1251] eta 0:01:51 lr 0.000053 time 0.3015 (0.2856) loss 3.0289 (3.0495) grad_norm 3.3432 (nan) [2021-04-16 17:33:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][870/1251] eta 0:01:48 lr 0.000053 time 0.2920 (0.2854) loss 2.7704 (3.0499) grad_norm 2.4534 (nan) [2021-04-16 17:33:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][880/1251] eta 0:01:45 lr 0.000053 time 0.4270 (0.2855) loss 1.9341 (3.0503) grad_norm 2.6275 (nan) [2021-04-16 17:33:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][890/1251] eta 0:01:43 lr 0.000053 time 0.2840 (0.2853) loss 2.0698 (3.0478) grad_norm 2.8816 (nan) [2021-04-16 17:33:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][900/1251] eta 0:01:40 lr 0.000053 time 0.4151 (0.2855) loss 3.3508 (3.0502) grad_norm 2.9541 (nan) [2021-04-16 17:33:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][910/1251] eta 0:01:37 lr 0.000053 time 0.2753 (0.2854) loss 3.5387 (3.0518) grad_norm 2.9237 (nan) [2021-04-16 17:33:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][920/1251] eta 0:01:34 lr 0.000053 time 0.2770 (0.2854) loss 2.6403 (3.0524) grad_norm 2.6011 (nan) [2021-04-16 17:33:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][930/1251] eta 0:01:31 lr 0.000053 time 0.2946 (0.2854) loss 3.2333 (3.0540) grad_norm 3.3657 (nan) [2021-04-16 17:33:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][940/1251] eta 0:01:28 lr 0.000053 time 0.2763 (0.2853) loss 2.8238 (3.0547) grad_norm 3.5454 (nan) [2021-04-16 17:33:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][950/1251] eta 0:01:25 lr 0.000053 time 0.3008 (0.2853) loss 3.2998 (3.0535) grad_norm 3.1226 (nan) [2021-04-16 17:33:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][960/1251] eta 0:01:22 lr 0.000053 time 0.2649 (0.2852) loss 3.7902 (3.0536) grad_norm 2.5273 (nan) [2021-04-16 17:33:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][970/1251] eta 0:01:20 lr 0.000053 time 0.2669 (0.2852) loss 2.5589 (3.0539) grad_norm 2.7498 (nan) [2021-04-16 17:34:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][980/1251] eta 0:01:17 lr 0.000053 time 0.2795 (0.2851) loss 3.4169 (3.0531) grad_norm 3.1630 (nan) [2021-04-16 17:34:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][990/1251] eta 0:01:14 lr 0.000053 time 0.2681 (0.2849) loss 3.4419 (3.0539) grad_norm 2.6831 (nan) [2021-04-16 17:34:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][1000/1251] eta 0:01:11 lr 0.000053 time 0.2765 (0.2848) loss 3.5967 (3.0542) grad_norm 2.6175 (nan) [2021-04-16 17:34:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][1010/1251] eta 0:01:08 lr 0.000053 time 0.2686 (0.2847) loss 3.5723 (3.0532) grad_norm 2.8519 (nan) [2021-04-16 17:34:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][1020/1251] eta 0:01:05 lr 0.000053 time 0.2791 (0.2846) loss 3.5804 (3.0557) grad_norm 3.0339 (nan) [2021-04-16 17:34:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][1030/1251] eta 0:01:02 lr 0.000053 time 0.2736 (0.2846) loss 3.3243 (3.0550) grad_norm 2.9419 (nan) [2021-04-16 17:34:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][1040/1251] eta 0:01:00 lr 0.000053 time 0.2634 (0.2845) loss 3.1869 (3.0551) grad_norm 2.9148 (nan) [2021-04-16 17:34:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][1050/1251] eta 0:00:57 lr 0.000053 time 0.2731 (0.2844) loss 2.1220 (3.0538) grad_norm 2.9610 (nan) [2021-04-16 17:34:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][1060/1251] eta 0:00:54 lr 0.000053 time 0.2861 (0.2843) loss 2.9876 (3.0554) grad_norm 2.8623 (nan) [2021-04-16 17:34:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][1070/1251] eta 0:00:51 lr 0.000053 time 0.2639 (0.2842) loss 2.9980 (3.0572) grad_norm 2.5195 (nan) [2021-04-16 17:34:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][1080/1251] eta 0:00:48 lr 0.000053 time 0.2797 (0.2842) loss 3.9911 (3.0570) grad_norm 2.6342 (nan) [2021-04-16 17:34:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][1090/1251] eta 0:00:45 lr 0.000053 time 0.2845 (0.2842) loss 2.4121 (3.0538) grad_norm 3.1722 (nan) [2021-04-16 17:34:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][1100/1251] eta 0:00:42 lr 0.000053 time 0.2584 (0.2841) loss 3.3401 (3.0551) grad_norm 2.8548 (nan) [2021-04-16 17:34:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][1110/1251] eta 0:00:40 lr 0.000053 time 0.2597 (0.2840) loss 1.9235 (3.0550) grad_norm 2.9155 (nan) [2021-04-16 17:34:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][1120/1251] eta 0:00:37 lr 0.000053 time 0.2744 (0.2839) loss 3.1682 (3.0556) grad_norm 2.8677 (nan) [2021-04-16 17:34:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][1130/1251] eta 0:00:34 lr 0.000053 time 0.2615 (0.2839) loss 3.5809 (3.0584) grad_norm 2.7816 (nan) [2021-04-16 17:34:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][1140/1251] eta 0:00:31 lr 0.000053 time 0.2682 (0.2839) loss 3.2607 (3.0582) grad_norm 2.7480 (nan) [2021-04-16 17:34:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][1150/1251] eta 0:00:28 lr 0.000053 time 0.2930 (0.2838) loss 3.3892 (3.0562) grad_norm 2.4016 (nan) [2021-04-16 17:34:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][1160/1251] eta 0:00:25 lr 0.000053 time 0.2660 (0.2839) loss 3.6296 (3.0565) grad_norm 3.4590 (nan) [2021-04-16 17:34:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][1170/1251] eta 0:00:22 lr 0.000053 time 0.2736 (0.2839) loss 3.1230 (3.0581) grad_norm 2.6541 (nan) [2021-04-16 17:34:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][1180/1251] eta 0:00:20 lr 0.000053 time 0.2623 (0.2838) loss 3.6658 (3.0581) grad_norm 3.3839 (nan) [2021-04-16 17:34:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][1190/1251] eta 0:00:17 lr 0.000053 time 0.2683 (0.2838) loss 3.2039 (3.0587) grad_norm 2.8360 (nan) [2021-04-16 17:35:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][1200/1251] eta 0:00:14 lr 0.000053 time 0.2997 (0.2838) loss 3.2631 (3.0587) grad_norm 2.6226 (nan) [2021-04-16 17:35:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][1210/1251] eta 0:00:11 lr 0.000053 time 0.2819 (0.2837) loss 2.7762 (3.0566) grad_norm 7.6901 (nan) [2021-04-16 17:35:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][1220/1251] eta 0:00:08 lr 0.000053 time 0.2448 (0.2836) loss 3.5325 (3.0589) grad_norm 3.0379 (nan) [2021-04-16 17:35:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][1230/1251] eta 0:00:05 lr 0.000053 time 0.2855 (0.2836) loss 3.4816 (3.0602) grad_norm 3.2060 (nan) [2021-04-16 17:35:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][1240/1251] eta 0:00:03 lr 0.000053 time 0.2481 (0.2834) loss 3.6922 (3.0601) grad_norm 2.5234 (nan) [2021-04-16 17:35:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [259/300][1250/1251] eta 0:00:00 lr 0.000053 time 0.2473 (0.2831) loss 2.9820 (3.0594) grad_norm 2.8870 (nan) [2021-04-16 17:35:18 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 259 training takes 0:05:57 [2021-04-16 17:35:18 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_259.pth saving...... [2021-04-16 17:35:29 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_259.pth saved !!! [2021-04-16 17:35:31 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 2.028 (2.028) Loss 0.8021 (0.8021) Acc@1 79.590 (79.590) Acc@5 95.801 (95.801) [2021-04-16 17:35:32 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.163 (0.304) Loss 0.8574 (0.8492) Acc@1 81.055 (79.838) Acc@5 94.824 (94.966) [2021-04-16 17:35:35 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.183 (0.269) Loss 0.8041 (0.8390) Acc@1 80.859 (80.194) Acc@5 95.410 (95.131) [2021-04-16 17:35:37 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.108 (0.268) Loss 0.7544 (0.8298) Acc@1 82.129 (80.459) Acc@5 95.410 (95.221) [2021-04-16 17:35:39 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.243) Loss 0.8544 (0.8291) Acc@1 78.906 (80.454) Acc@5 95.508 (95.251) [2021-04-16 17:35:49 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.528 Acc@5 95.260 [2021-04-16 17:35:49 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 80.5% [2021-04-16 17:35:49 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 80.62% [2021-04-16 17:36:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][0/1251] eta 5:43:08 lr 0.000053 time 16.4578 (16.4578) loss 2.9602 (2.9602) grad_norm 2.6659 (2.6659) [2021-04-16 17:36:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][10/1251] eta 0:36:22 lr 0.000053 time 0.4415 (1.7588) loss 2.4146 (2.9839) grad_norm 3.3625 (2.8838) [2021-04-16 17:36:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][20/1251] eta 0:21:36 lr 0.000053 time 0.2809 (1.0528) loss 2.9049 (3.0574) grad_norm 2.6884 (2.8962) [2021-04-16 17:36:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][30/1251] eta 0:16:19 lr 0.000053 time 0.2648 (0.8025) loss 3.1076 (3.0849) grad_norm 2.5749 (2.8828) [2021-04-16 17:36:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][40/1251] eta 0:13:35 lr 0.000053 time 0.2428 (0.6738) loss 3.4654 (3.0658) grad_norm 2.6610 (2.8880) [2021-04-16 17:36:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][50/1251] eta 0:11:55 lr 0.000053 time 0.2869 (0.5959) loss 3.2189 (3.1136) grad_norm 3.2904 (2.8880) [2021-04-16 17:36:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][60/1251] eta 0:10:47 lr 0.000053 time 0.2686 (0.5437) loss 2.2811 (3.0729) grad_norm 3.2045 (2.8920) [2021-04-16 17:36:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][70/1251] eta 0:09:59 lr 0.000053 time 0.2793 (0.5079) loss 2.8537 (3.0744) grad_norm 3.5653 (2.9041) [2021-04-16 17:36:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][80/1251] eta 0:09:21 lr 0.000053 time 0.2915 (0.4796) loss 3.4418 (3.0650) grad_norm 3.2658 (2.9067) [2021-04-16 17:36:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][90/1251] eta 0:08:50 lr 0.000053 time 0.2879 (0.4570) loss 3.4168 (3.0798) grad_norm 2.4944 (2.8945) [2021-04-16 17:36:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][100/1251] eta 0:08:26 lr 0.000053 time 0.2789 (0.4403) loss 2.6642 (3.0843) grad_norm 2.7376 (2.9052) [2021-04-16 17:36:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][110/1251] eta 0:08:05 lr 0.000053 time 0.3060 (0.4253) loss 2.1968 (3.0483) grad_norm 2.3437 (2.9071) [2021-04-16 17:36:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][120/1251] eta 0:07:48 lr 0.000053 time 0.2602 (0.4145) loss 3.2268 (3.0475) grad_norm 3.2020 (2.9181) [2021-04-16 17:36:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][130/1251] eta 0:07:34 lr 0.000053 time 0.2627 (0.4053) loss 3.5563 (3.0528) grad_norm 3.5581 (2.9245) [2021-04-16 17:36:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][140/1251] eta 0:07:19 lr 0.000053 time 0.2876 (0.3960) loss 3.7255 (3.0529) grad_norm 2.5998 (2.9277) [2021-04-16 17:36:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][150/1251] eta 0:07:07 lr 0.000053 time 0.2907 (0.3884) loss 3.2386 (3.0457) grad_norm 2.6332 (2.9264) [2021-04-16 17:36:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][160/1251] eta 0:06:56 lr 0.000053 time 0.2798 (0.3819) loss 2.2334 (3.0345) grad_norm 2.5709 (2.9240) [2021-04-16 17:36:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][170/1251] eta 0:06:46 lr 0.000053 time 0.2844 (0.3758) loss 4.0278 (3.0448) grad_norm 2.7311 (2.9241) [2021-04-16 17:36:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][180/1251] eta 0:06:36 lr 0.000052 time 0.2576 (0.3704) loss 3.4319 (3.0400) grad_norm 3.0540 (2.9261) [2021-04-16 17:36:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][190/1251] eta 0:06:27 lr 0.000052 time 0.2822 (0.3654) loss 2.8026 (3.0348) grad_norm 3.0333 (2.9243) [2021-04-16 17:37:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][200/1251] eta 0:06:19 lr 0.000052 time 0.2689 (0.3609) loss 3.9074 (3.0347) grad_norm 2.8237 (2.9344) [2021-04-16 17:37:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][210/1251] eta 0:06:11 lr 0.000052 time 0.2793 (0.3568) loss 2.8255 (3.0391) grad_norm 3.4268 (2.9356) [2021-04-16 17:37:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][220/1251] eta 0:06:03 lr 0.000052 time 0.2667 (0.3530) loss 2.0208 (3.0380) grad_norm 2.8909 (2.9439) [2021-04-16 17:37:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][230/1251] eta 0:05:57 lr 0.000052 time 0.2848 (0.3497) loss 3.4981 (3.0447) grad_norm 2.9014 (2.9455) [2021-04-16 17:37:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][240/1251] eta 0:05:50 lr 0.000052 time 0.3077 (0.3468) loss 3.6725 (3.0418) grad_norm 2.8301 (2.9507) [2021-04-16 17:37:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][250/1251] eta 0:05:44 lr 0.000052 time 0.2797 (0.3440) loss 2.3726 (3.0224) grad_norm 2.8095 (2.9434) [2021-04-16 17:37:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][260/1251] eta 0:05:38 lr 0.000052 time 0.2772 (0.3414) loss 2.3817 (3.0199) grad_norm 2.9159 (2.9443) [2021-04-16 17:37:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][270/1251] eta 0:05:32 lr 0.000052 time 0.2888 (0.3393) loss 3.0214 (3.0216) grad_norm 2.4817 (2.9423) [2021-04-16 17:37:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][280/1251] eta 0:05:27 lr 0.000052 time 0.2604 (0.3371) loss 2.9632 (3.0194) grad_norm 2.9409 (2.9423) [2021-04-16 17:37:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][290/1251] eta 0:05:21 lr 0.000052 time 0.2836 (0.3349) loss 3.5883 (3.0225) grad_norm 2.4221 (2.9391) [2021-04-16 17:37:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][300/1251] eta 0:05:16 lr 0.000052 time 0.2723 (0.3330) loss 3.5284 (3.0300) grad_norm 3.3312 (2.9396) [2021-04-16 17:37:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][310/1251] eta 0:05:12 lr 0.000052 time 0.2795 (0.3317) loss 3.1213 (3.0225) grad_norm 2.6028 (2.9424) [2021-04-16 17:37:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][320/1251] eta 0:05:07 lr 0.000052 time 0.2697 (0.3299) loss 3.7755 (3.0328) grad_norm 3.2910 (2.9566) [2021-04-16 17:37:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][330/1251] eta 0:05:02 lr 0.000052 time 0.2577 (0.3282) loss 2.6840 (3.0301) grad_norm 2.6848 (2.9571) [2021-04-16 17:37:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][340/1251] eta 0:04:57 lr 0.000052 time 0.2949 (0.3265) loss 3.3640 (3.0261) grad_norm 3.5947 (2.9573) [2021-04-16 17:37:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][350/1251] eta 0:04:53 lr 0.000052 time 0.2779 (0.3253) loss 2.9288 (3.0271) grad_norm 2.8065 (2.9576) [2021-04-16 17:37:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][360/1251] eta 0:04:48 lr 0.000052 time 0.2810 (0.3242) loss 3.1319 (3.0342) grad_norm 3.0945 (2.9631) [2021-04-16 17:37:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][370/1251] eta 0:04:44 lr 0.000052 time 0.2686 (0.3233) loss 3.4151 (3.0295) grad_norm 2.9384 (2.9653) [2021-04-16 17:37:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][380/1251] eta 0:04:40 lr 0.000052 time 0.2619 (0.3225) loss 3.1997 (3.0233) grad_norm 2.8722 (2.9714) [2021-04-16 17:37:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][390/1251] eta 0:04:36 lr 0.000052 time 0.2678 (0.3215) loss 2.5765 (3.0246) grad_norm 2.5578 (2.9673) [2021-04-16 17:37:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][400/1251] eta 0:04:32 lr 0.000052 time 0.2438 (0.3202) loss 3.4738 (3.0280) grad_norm 2.5656 (2.9629) [2021-04-16 17:38:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][410/1251] eta 0:04:28 lr 0.000052 time 0.2660 (0.3192) loss 3.5036 (3.0296) grad_norm 2.3482 (2.9582) [2021-04-16 17:38:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][420/1251] eta 0:04:24 lr 0.000052 time 0.2743 (0.3180) loss 2.0985 (3.0295) grad_norm 3.1900 (2.9620) [2021-04-16 17:38:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][430/1251] eta 0:04:20 lr 0.000052 time 0.2899 (0.3171) loss 2.5179 (3.0190) grad_norm 2.7651 (2.9599) [2021-04-16 17:38:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][440/1251] eta 0:04:16 lr 0.000052 time 0.2901 (0.3161) loss 3.2653 (3.0228) grad_norm 2.7540 (2.9613) [2021-04-16 17:38:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][450/1251] eta 0:04:12 lr 0.000052 time 0.2633 (0.3151) loss 2.2326 (3.0226) grad_norm 2.7056 (2.9662) [2021-04-16 17:38:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][460/1251] eta 0:04:08 lr 0.000052 time 0.2763 (0.3142) loss 3.4377 (3.0262) grad_norm 2.6994 (2.9632) [2021-04-16 17:38:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][470/1251] eta 0:04:04 lr 0.000052 time 0.2433 (0.3134) loss 2.1234 (3.0241) grad_norm 3.0650 (2.9687) [2021-04-16 17:38:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][480/1251] eta 0:04:01 lr 0.000052 time 0.2969 (0.3127) loss 3.0681 (3.0268) grad_norm 2.7806 (2.9695) [2021-04-16 17:38:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][490/1251] eta 0:03:57 lr 0.000052 time 0.2759 (0.3121) loss 3.0530 (3.0260) grad_norm 2.7339 (2.9688) [2021-04-16 17:38:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][500/1251] eta 0:03:53 lr 0.000052 time 0.2679 (0.3114) loss 3.3440 (3.0269) grad_norm 2.7690 (2.9689) [2021-04-16 17:38:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][510/1251] eta 0:03:50 lr 0.000052 time 0.2919 (0.3107) loss 3.7787 (3.0238) grad_norm 2.9484 (2.9753) [2021-04-16 17:38:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][520/1251] eta 0:03:46 lr 0.000052 time 0.2743 (0.3101) loss 3.1301 (3.0281) grad_norm 2.7284 (2.9758) [2021-04-16 17:38:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][530/1251] eta 0:03:43 lr 0.000052 time 0.2695 (0.3094) loss 3.1478 (3.0214) grad_norm 2.3971 (2.9779) [2021-04-16 17:38:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][540/1251] eta 0:03:39 lr 0.000052 time 0.2648 (0.3088) loss 2.9080 (3.0227) grad_norm 2.9141 (2.9797) [2021-04-16 17:38:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][550/1251] eta 0:03:36 lr 0.000052 time 0.2942 (0.3082) loss 3.1478 (3.0247) grad_norm 4.7288 (2.9835) [2021-04-16 17:38:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][560/1251] eta 0:03:32 lr 0.000052 time 0.2794 (0.3076) loss 3.0117 (3.0248) grad_norm 2.6733 (2.9832) [2021-04-16 17:38:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][570/1251] eta 0:03:29 lr 0.000052 time 0.4197 (0.3078) loss 2.6631 (3.0232) grad_norm 3.2804 (2.9871) [2021-04-16 17:38:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][580/1251] eta 0:03:26 lr 0.000052 time 0.2693 (0.3072) loss 3.8040 (3.0227) grad_norm 2.8739 (2.9848) [2021-04-16 17:38:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][590/1251] eta 0:03:22 lr 0.000052 time 0.2432 (0.3068) loss 2.4062 (3.0235) grad_norm 3.1207 (2.9902) [2021-04-16 17:38:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][600/1251] eta 0:03:19 lr 0.000052 time 0.3059 (0.3064) loss 2.7022 (3.0243) grad_norm 3.3753 (2.9947) [2021-04-16 17:38:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][610/1251] eta 0:03:16 lr 0.000052 time 0.2751 (0.3059) loss 2.4421 (3.0235) grad_norm 2.7322 (2.9924) [2021-04-16 17:38:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][620/1251] eta 0:03:12 lr 0.000052 time 0.2620 (0.3054) loss 3.3119 (3.0244) grad_norm 3.6191 (2.9906) [2021-04-16 17:39:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][630/1251] eta 0:03:09 lr 0.000052 time 0.2541 (0.3049) loss 3.2864 (3.0214) grad_norm 2.5406 (2.9920) [2021-04-16 17:39:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][640/1251] eta 0:03:06 lr 0.000052 time 0.2585 (0.3047) loss 3.7191 (3.0247) grad_norm 2.9927 (2.9926) [2021-04-16 17:39:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][650/1251] eta 0:03:02 lr 0.000052 time 0.2936 (0.3043) loss 3.3785 (3.0277) grad_norm 3.4676 (2.9908) [2021-04-16 17:39:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][660/1251] eta 0:02:59 lr 0.000052 time 0.2956 (0.3039) loss 3.2758 (3.0310) grad_norm 3.1190 (2.9895) [2021-04-16 17:39:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][670/1251] eta 0:02:56 lr 0.000052 time 0.2845 (0.3034) loss 2.9446 (3.0317) grad_norm 3.0333 (2.9896) [2021-04-16 17:39:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][680/1251] eta 0:02:53 lr 0.000052 time 0.2653 (0.3032) loss 3.4010 (3.0310) grad_norm 2.9941 (2.9868) [2021-04-16 17:39:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][690/1251] eta 0:02:49 lr 0.000052 time 0.2829 (0.3029) loss 3.6869 (3.0352) grad_norm 3.2046 (2.9845) [2021-04-16 17:39:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][700/1251] eta 0:02:46 lr 0.000052 time 0.2749 (0.3025) loss 3.0631 (3.0417) grad_norm 3.0321 (2.9851) [2021-04-16 17:39:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][710/1251] eta 0:02:43 lr 0.000052 time 0.2501 (0.3023) loss 2.7992 (3.0398) grad_norm 2.5444 (2.9848) [2021-04-16 17:39:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][720/1251] eta 0:02:40 lr 0.000052 time 0.2844 (0.3019) loss 3.1731 (3.0405) grad_norm 2.4567 (2.9834) [2021-04-16 17:39:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][730/1251] eta 0:02:37 lr 0.000052 time 0.2762 (0.3015) loss 2.6810 (3.0424) grad_norm 3.3799 (2.9850) [2021-04-16 17:39:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][740/1251] eta 0:02:33 lr 0.000052 time 0.2908 (0.3012) loss 3.4125 (3.0428) grad_norm 3.7315 (3.0067) [2021-04-16 17:39:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][750/1251] eta 0:02:30 lr 0.000052 time 0.2723 (0.3009) loss 3.3911 (3.0445) grad_norm 2.7321 (3.0079) [2021-04-16 17:39:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][760/1251] eta 0:02:27 lr 0.000052 time 0.2859 (0.3005) loss 2.3282 (3.0465) grad_norm 3.1619 (3.0098) [2021-04-16 17:39:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][770/1251] eta 0:02:24 lr 0.000052 time 0.2761 (0.3002) loss 3.4172 (3.0431) grad_norm 3.2758 (3.0106) [2021-04-16 17:39:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][780/1251] eta 0:02:21 lr 0.000051 time 0.2624 (0.2999) loss 3.1556 (3.0402) grad_norm 3.1276 (3.0111) [2021-04-16 17:39:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][790/1251] eta 0:02:18 lr 0.000051 time 0.2574 (0.2995) loss 2.7850 (3.0377) grad_norm 2.5846 (3.0097) [2021-04-16 17:39:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][800/1251] eta 0:02:15 lr 0.000051 time 0.3071 (0.2994) loss 2.3144 (3.0356) grad_norm 2.8988 (3.0068) [2021-04-16 17:39:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][810/1251] eta 0:02:11 lr 0.000051 time 0.2572 (0.2990) loss 2.8088 (3.0340) grad_norm 3.2576 (3.0046) [2021-04-16 17:39:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][820/1251] eta 0:02:08 lr 0.000051 time 0.2603 (0.2987) loss 3.5626 (3.0370) grad_norm 2.7970 (3.0029) [2021-04-16 17:39:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][830/1251] eta 0:02:05 lr 0.000051 time 0.2597 (0.2985) loss 2.0814 (3.0360) grad_norm 3.1450 (3.0026) [2021-04-16 17:40:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][840/1251] eta 0:02:02 lr 0.000051 time 0.2922 (0.2982) loss 3.9203 (3.0375) grad_norm 2.7980 (3.0030) [2021-04-16 17:40:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][850/1251] eta 0:01:59 lr 0.000051 time 0.2605 (0.2980) loss 2.7952 (3.0385) grad_norm 3.0434 (3.0023) [2021-04-16 17:40:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][860/1251] eta 0:01:56 lr 0.000051 time 0.2723 (0.2977) loss 2.1942 (3.0348) grad_norm 2.5082 (3.0033) [2021-04-16 17:40:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][870/1251] eta 0:01:53 lr 0.000051 time 0.2598 (0.2975) loss 3.0345 (3.0342) grad_norm 2.6202 (3.0014) [2021-04-16 17:40:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][880/1251] eta 0:01:50 lr 0.000051 time 0.2591 (0.2972) loss 3.8194 (3.0354) grad_norm 2.9424 (2.9999) [2021-04-16 17:40:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][890/1251] eta 0:01:47 lr 0.000051 time 0.2918 (0.2970) loss 3.4487 (3.0379) grad_norm 3.2114 (2.9978) [2021-04-16 17:40:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][900/1251] eta 0:01:44 lr 0.000051 time 0.2781 (0.2967) loss 2.7272 (3.0361) grad_norm 2.7681 (2.9967) [2021-04-16 17:40:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][910/1251] eta 0:01:41 lr 0.000051 time 0.2817 (0.2966) loss 2.3507 (3.0384) grad_norm 3.0859 (2.9962) [2021-04-16 17:40:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][920/1251] eta 0:01:38 lr 0.000051 time 0.3087 (0.2964) loss 2.4532 (3.0390) grad_norm 2.5647 (2.9960) [2021-04-16 17:40:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][930/1251] eta 0:01:35 lr 0.000051 time 0.2713 (0.2964) loss 2.9888 (3.0415) grad_norm 3.1988 (2.9971) [2021-04-16 17:40:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][940/1251] eta 0:01:32 lr 0.000051 time 0.2685 (0.2963) loss 3.4428 (3.0392) grad_norm 2.7010 (2.9959) [2021-04-16 17:40:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][950/1251] eta 0:01:29 lr 0.000051 time 0.2825 (0.2961) loss 3.7105 (3.0404) grad_norm 3.1757 (2.9955) [2021-04-16 17:40:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][960/1251] eta 0:01:26 lr 0.000051 time 0.2801 (0.2960) loss 3.4216 (3.0387) grad_norm 3.0190 (2.9959) [2021-04-16 17:40:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][970/1251] eta 0:01:23 lr 0.000051 time 0.2928 (0.2959) loss 3.3161 (3.0393) grad_norm 2.6137 (2.9955) [2021-04-16 17:40:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][980/1251] eta 0:01:20 lr 0.000051 time 0.2964 (0.2957) loss 3.1628 (3.0380) grad_norm 2.8533 (2.9949) [2021-04-16 17:40:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][990/1251] eta 0:01:17 lr 0.000051 time 0.2773 (0.2956) loss 3.2986 (3.0404) grad_norm 2.6697 (2.9947) [2021-04-16 17:40:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][1000/1251] eta 0:01:14 lr 0.000051 time 0.2683 (0.2954) loss 3.0989 (3.0403) grad_norm 2.7017 (2.9947) [2021-04-16 17:40:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][1010/1251] eta 0:01:11 lr 0.000051 time 0.2735 (0.2953) loss 3.7324 (3.0414) grad_norm 2.4638 (2.9923) [2021-04-16 17:40:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][1020/1251] eta 0:01:08 lr 0.000051 time 0.2424 (0.2951) loss 2.2576 (3.0424) grad_norm 2.7515 (2.9910) [2021-04-16 17:40:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][1030/1251] eta 0:01:05 lr 0.000051 time 0.2524 (0.2949) loss 3.5687 (3.0443) grad_norm 3.5232 (2.9971) [2021-04-16 17:40:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][1040/1251] eta 0:01:02 lr 0.000051 time 0.2499 (0.2946) loss 2.9107 (3.0461) grad_norm 3.0228 (2.9967) [2021-04-16 17:40:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][1050/1251] eta 0:00:59 lr 0.000051 time 0.2719 (0.2945) loss 3.8753 (3.0473) grad_norm 3.1532 (2.9965) [2021-04-16 17:41:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][1060/1251] eta 0:00:56 lr 0.000051 time 0.2806 (0.2943) loss 2.6857 (3.0462) grad_norm 2.9975 (2.9955) [2021-04-16 17:41:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][1070/1251] eta 0:00:53 lr 0.000051 time 0.2623 (0.2941) loss 3.4674 (3.0462) grad_norm 2.9718 (2.9969) [2021-04-16 17:41:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][1080/1251] eta 0:00:50 lr 0.000051 time 0.2728 (0.2940) loss 2.1594 (3.0435) grad_norm 3.1372 (3.0032) [2021-04-16 17:41:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][1090/1251] eta 0:00:47 lr 0.000051 time 0.2720 (0.2939) loss 3.4470 (3.0444) grad_norm 3.2748 (3.0038) [2021-04-16 17:41:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][1100/1251] eta 0:00:44 lr 0.000051 time 0.2685 (0.2937) loss 3.5258 (3.0428) grad_norm 2.8191 (3.0057) [2021-04-16 17:41:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][1110/1251] eta 0:00:41 lr 0.000051 time 0.2626 (0.2936) loss 3.1810 (3.0427) grad_norm 3.3379 (3.0064) [2021-04-16 17:41:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][1120/1251] eta 0:00:38 lr 0.000051 time 0.2753 (0.2934) loss 3.2033 (3.0427) grad_norm 3.0588 (3.0060) [2021-04-16 17:41:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][1130/1251] eta 0:00:35 lr 0.000051 time 0.2608 (0.2933) loss 3.2492 (3.0417) grad_norm 2.7176 (3.0060) [2021-04-16 17:41:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][1140/1251] eta 0:00:32 lr 0.000051 time 0.2634 (0.2932) loss 3.1306 (3.0414) grad_norm 2.8870 (3.0059) [2021-04-16 17:41:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][1150/1251] eta 0:00:29 lr 0.000051 time 0.2607 (0.2933) loss 3.2718 (3.0397) grad_norm 2.8195 (3.0046) [2021-04-16 17:41:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][1160/1251] eta 0:00:26 lr 0.000051 time 0.2816 (0.2932) loss 3.5664 (3.0404) grad_norm 2.9013 (3.0052) [2021-04-16 17:41:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][1170/1251] eta 0:00:23 lr 0.000051 time 0.2619 (0.2932) loss 3.4467 (3.0409) grad_norm 4.1487 (3.0080) [2021-04-16 17:41:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][1180/1251] eta 0:00:20 lr 0.000051 time 0.2810 (0.2930) loss 3.2418 (3.0413) grad_norm 3.0205 (3.0111) [2021-04-16 17:41:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][1190/1251] eta 0:00:17 lr 0.000051 time 0.2791 (0.2928) loss 3.3103 (3.0406) grad_norm 3.0945 (3.0105) [2021-04-16 17:41:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][1200/1251] eta 0:00:14 lr 0.000051 time 0.2919 (0.2927) loss 3.4285 (3.0421) grad_norm 3.0908 (3.0124) [2021-04-16 17:41:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][1210/1251] eta 0:00:11 lr 0.000051 time 0.2763 (0.2925) loss 3.1669 (3.0405) grad_norm 3.3649 (3.0128) [2021-04-16 17:41:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][1220/1251] eta 0:00:09 lr 0.000051 time 0.2503 (0.2925) loss 2.8567 (3.0435) grad_norm 3.7993 (3.0120) [2021-04-16 17:41:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][1230/1251] eta 0:00:06 lr 0.000051 time 0.2751 (0.2924) loss 2.6473 (3.0427) grad_norm 3.0446 (3.0111) [2021-04-16 17:41:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][1240/1251] eta 0:00:03 lr 0.000051 time 0.2483 (0.2921) loss 1.8010 (3.0420) grad_norm 2.3459 (3.0093) [2021-04-16 17:41:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [260/300][1250/1251] eta 0:00:00 lr 0.000051 time 0.2477 (0.2918) loss 2.3782 (3.0418) grad_norm 2.8792 (3.0079) [2021-04-16 17:42:02 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 260 training takes 0:06:12 [2021-04-16 17:42:02 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_260.pth saving...... [2021-04-16 17:42:24 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_260.pth saved !!! [2021-04-16 17:42:25 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.187 (1.187) Loss 0.7973 (0.7973) Acc@1 82.715 (82.715) Acc@5 95.605 (95.605) [2021-04-16 17:42:26 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.145 (0.226) Loss 0.8131 (0.8291) Acc@1 82.324 (80.850) Acc@5 94.434 (95.126) [2021-04-16 17:42:28 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.121 (0.222) Loss 0.7983 (0.8206) Acc@1 81.055 (80.827) Acc@5 95.312 (95.257) [2021-04-16 17:42:31 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.112 (0.224) Loss 0.7879 (0.8200) Acc@1 82.031 (80.935) Acc@5 94.824 (95.297) [2021-04-16 17:42:32 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.080 (0.213) Loss 0.8010 (0.8183) Acc@1 79.980 (80.909) Acc@5 95.703 (95.348) [2021-04-16 17:42:47 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.780 Acc@5 95.372 [2021-04-16 17:42:47 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 80.8% [2021-04-16 17:42:47 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 80.78% [2021-04-16 17:43:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][0/1251] eta 5:23:09 lr 0.000051 time 15.4991 (15.4991) loss 3.8854 (3.8854) grad_norm 3.6258 (3.6258) [2021-04-16 17:43:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][10/1251] eta 0:34:14 lr 0.000051 time 0.2777 (1.6553) loss 3.4149 (3.2987) grad_norm 3.0380 (2.9312) [2021-04-16 17:43:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][20/1251] eta 0:20:27 lr 0.000051 time 0.2560 (0.9969) loss 2.4285 (3.1378) grad_norm 3.0168 (2.9196) [2021-04-16 17:43:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][30/1251] eta 0:15:36 lr 0.000051 time 0.2848 (0.7671) loss 3.3470 (3.1934) grad_norm 2.7013 (2.9543) [2021-04-16 17:43:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][40/1251] eta 0:13:04 lr 0.000051 time 0.2809 (0.6476) loss 3.2065 (3.1184) grad_norm 2.8781 (2.9811) [2021-04-16 17:43:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][50/1251] eta 0:11:31 lr 0.000051 time 0.2770 (0.5759) loss 2.1445 (3.0807) grad_norm 2.6918 (3.0197) [2021-04-16 17:43:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][60/1251] eta 0:10:27 lr 0.000051 time 0.2669 (0.5268) loss 3.4945 (3.0864) grad_norm 3.2914 (3.0021) [2021-04-16 17:43:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][70/1251] eta 0:09:41 lr 0.000051 time 0.2833 (0.4921) loss 3.3974 (3.0668) grad_norm 2.7866 (2.9910) [2021-04-16 17:43:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][80/1251] eta 0:09:05 lr 0.000051 time 0.2608 (0.4655) loss 3.2445 (3.0623) grad_norm 2.7840 (2.9748) [2021-04-16 17:43:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][90/1251] eta 0:08:36 lr 0.000051 time 0.2766 (0.4448) loss 2.6004 (3.0279) grad_norm 3.6600 (2.9685) [2021-04-16 17:43:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][100/1251] eta 0:08:12 lr 0.000051 time 0.2630 (0.4282) loss 3.4849 (3.0344) grad_norm 2.7174 (2.9749) [2021-04-16 17:43:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][110/1251] eta 0:07:53 lr 0.000051 time 0.2856 (0.4150) loss 3.1823 (3.0374) grad_norm 2.8614 (2.9739) [2021-04-16 17:43:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][120/1251] eta 0:07:36 lr 0.000051 time 0.2551 (0.4032) loss 2.7509 (3.0237) grad_norm 2.9195 (2.9773) [2021-04-16 17:43:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][130/1251] eta 0:07:21 lr 0.000050 time 0.2970 (0.3942) loss 3.2298 (3.0248) grad_norm 2.9014 (3.0086) [2021-04-16 17:43:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][140/1251] eta 0:07:11 lr 0.000050 time 0.2655 (0.3882) loss 2.5060 (3.0091) grad_norm 3.8366 (3.0172) [2021-04-16 17:43:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][150/1251] eta 0:07:00 lr 0.000050 time 0.2708 (0.3819) loss 3.5037 (3.0003) grad_norm 3.1213 (3.0155) [2021-04-16 17:43:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][160/1251] eta 0:06:49 lr 0.000050 time 0.2620 (0.3752) loss 2.0116 (3.0167) grad_norm 2.6428 (3.0163) [2021-04-16 17:43:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][170/1251] eta 0:06:39 lr 0.000050 time 0.2660 (0.3693) loss 2.2495 (3.0025) grad_norm 2.7512 (3.0002) [2021-04-16 17:43:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][180/1251] eta 0:06:31 lr 0.000050 time 0.2813 (0.3651) loss 2.6492 (3.0028) grad_norm 3.3628 (2.9951) [2021-04-16 17:43:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][190/1251] eta 0:06:22 lr 0.000050 time 0.2576 (0.3604) loss 2.8398 (3.0012) grad_norm 2.4707 (3.0047) [2021-04-16 17:43:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][200/1251] eta 0:06:14 lr 0.000050 time 0.2725 (0.3561) loss 3.4462 (3.0006) grad_norm 2.7358 (3.0000) [2021-04-16 17:44:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][210/1251] eta 0:06:07 lr 0.000050 time 0.2595 (0.3526) loss 2.5984 (3.0031) grad_norm 3.3697 (2.9990) [2021-04-16 17:44:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][220/1251] eta 0:05:59 lr 0.000050 time 0.2651 (0.3489) loss 3.0316 (3.0135) grad_norm 2.9886 (3.0009) [2021-04-16 17:44:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][230/1251] eta 0:05:52 lr 0.000050 time 0.2761 (0.3456) loss 2.4010 (3.0186) grad_norm 2.6612 (2.9969) [2021-04-16 17:44:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][240/1251] eta 0:05:46 lr 0.000050 time 0.2916 (0.3430) loss 3.1674 (3.0123) grad_norm 2.7135 (2.9989) [2021-04-16 17:44:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][250/1251] eta 0:05:40 lr 0.000050 time 0.2596 (0.3406) loss 3.4532 (3.0178) grad_norm 3.3347 (3.0056) [2021-04-16 17:44:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][260/1251] eta 0:05:35 lr 0.000050 time 0.2984 (0.3382) loss 3.3533 (3.0213) grad_norm 3.2453 (3.0109) [2021-04-16 17:44:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][270/1251] eta 0:05:29 lr 0.000050 time 0.2793 (0.3358) loss 2.0598 (3.0032) grad_norm 3.4118 (3.0020) [2021-04-16 17:44:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][280/1251] eta 0:05:23 lr 0.000050 time 0.2607 (0.3336) loss 3.5136 (3.0159) grad_norm 2.6303 (3.0021) [2021-04-16 17:44:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][290/1251] eta 0:05:18 lr 0.000050 time 0.3112 (0.3317) loss 3.4315 (3.0195) grad_norm 5.5395 (3.0187) [2021-04-16 17:44:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][300/1251] eta 0:05:13 lr 0.000050 time 0.2909 (0.3299) loss 2.7830 (3.0143) grad_norm 2.8691 (3.0215) [2021-04-16 17:44:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][310/1251] eta 0:05:08 lr 0.000050 time 0.2775 (0.3283) loss 3.2249 (3.0263) grad_norm 3.1979 (3.0267) [2021-04-16 17:44:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][320/1251] eta 0:05:04 lr 0.000050 time 0.2905 (0.3269) loss 2.1701 (3.0259) grad_norm 3.0726 (3.0247) [2021-04-16 17:44:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][330/1251] eta 0:04:59 lr 0.000050 time 0.2890 (0.3254) loss 3.2945 (3.0303) grad_norm 3.3160 (nan) [2021-04-16 17:44:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][340/1251] eta 0:04:55 lr 0.000050 time 0.2918 (0.3244) loss 2.6595 (3.0336) grad_norm 2.8615 (nan) [2021-04-16 17:44:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][350/1251] eta 0:04:51 lr 0.000050 time 0.2793 (0.3230) loss 3.1446 (3.0326) grad_norm 3.0296 (nan) [2021-04-16 17:44:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][360/1251] eta 0:04:47 lr 0.000050 time 0.3040 (0.3225) loss 3.0053 (3.0373) grad_norm 2.6492 (nan) [2021-04-16 17:44:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][370/1251] eta 0:04:43 lr 0.000050 time 0.2651 (0.3220) loss 2.1930 (3.0391) grad_norm 2.6066 (nan) [2021-04-16 17:44:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][380/1251] eta 0:04:39 lr 0.000050 time 0.2860 (0.3209) loss 3.5319 (3.0414) grad_norm 2.5368 (nan) [2021-04-16 17:44:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][390/1251] eta 0:04:35 lr 0.000050 time 0.2931 (0.3197) loss 3.1513 (3.0435) grad_norm 2.9315 (nan) [2021-04-16 17:44:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][400/1251] eta 0:04:31 lr 0.000050 time 0.2714 (0.3191) loss 2.4656 (3.0374) grad_norm 2.4587 (nan) [2021-04-16 17:44:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][410/1251] eta 0:04:27 lr 0.000050 time 0.2815 (0.3182) loss 3.6392 (3.0398) grad_norm 2.4746 (nan) [2021-04-16 17:45:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][420/1251] eta 0:04:23 lr 0.000050 time 0.2415 (0.3170) loss 1.7893 (3.0375) grad_norm 2.9385 (nan) [2021-04-16 17:45:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][430/1251] eta 0:04:19 lr 0.000050 time 0.2984 (0.3161) loss 3.6924 (3.0404) grad_norm 3.9803 (nan) [2021-04-16 17:45:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][440/1251] eta 0:04:15 lr 0.000050 time 0.2749 (0.3151) loss 2.3300 (3.0393) grad_norm 2.7194 (nan) [2021-04-16 17:45:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][450/1251] eta 0:04:11 lr 0.000050 time 0.2951 (0.3144) loss 3.6807 (3.0440) grad_norm 3.7993 (nan) [2021-04-16 17:45:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][460/1251] eta 0:04:08 lr 0.000050 time 0.2989 (0.3136) loss 3.1844 (3.0464) grad_norm 2.8195 (nan) [2021-04-16 17:45:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][470/1251] eta 0:04:04 lr 0.000050 time 0.2988 (0.3128) loss 3.2311 (3.0487) grad_norm 2.4861 (nan) [2021-04-16 17:45:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][480/1251] eta 0:04:00 lr 0.000050 time 0.2709 (0.3122) loss 3.4627 (3.0534) grad_norm 2.7786 (nan) [2021-04-16 17:45:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][490/1251] eta 0:03:57 lr 0.000050 time 0.2597 (0.3117) loss 3.1270 (3.0601) grad_norm 3.3928 (nan) [2021-04-16 17:45:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][500/1251] eta 0:03:53 lr 0.000050 time 0.2794 (0.3109) loss 2.4988 (3.0571) grad_norm 3.2429 (nan) [2021-04-16 17:45:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][510/1251] eta 0:03:49 lr 0.000050 time 0.3069 (0.3103) loss 2.6433 (3.0578) grad_norm 2.9158 (nan) [2021-04-16 17:45:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][520/1251] eta 0:03:46 lr 0.000050 time 0.2562 (0.3096) loss 3.8237 (3.0571) grad_norm 3.1537 (nan) [2021-04-16 17:45:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][530/1251] eta 0:03:42 lr 0.000050 time 0.2720 (0.3089) loss 3.1220 (3.0515) grad_norm 2.8894 (nan) [2021-04-16 17:45:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][540/1251] eta 0:03:39 lr 0.000050 time 0.2802 (0.3084) loss 1.9061 (3.0463) grad_norm 3.0975 (nan) [2021-04-16 17:45:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][550/1251] eta 0:03:35 lr 0.000050 time 0.2447 (0.3078) loss 3.6607 (3.0463) grad_norm 3.0639 (nan) [2021-04-16 17:45:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][560/1251] eta 0:03:32 lr 0.000050 time 0.2688 (0.3072) loss 2.0057 (3.0413) grad_norm 2.8580 (nan) [2021-04-16 17:45:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][570/1251] eta 0:03:28 lr 0.000050 time 0.2667 (0.3069) loss 3.0489 (3.0423) grad_norm 2.7368 (nan) [2021-04-16 17:45:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][580/1251] eta 0:03:25 lr 0.000050 time 0.2757 (0.3063) loss 2.9520 (3.0404) grad_norm 2.7463 (nan) [2021-04-16 17:45:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][590/1251] eta 0:03:22 lr 0.000050 time 0.2702 (0.3058) loss 2.2809 (3.0345) grad_norm 3.5126 (nan) [2021-04-16 17:45:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][600/1251] eta 0:03:18 lr 0.000050 time 0.2856 (0.3054) loss 3.3742 (3.0325) grad_norm 5.0892 (nan) [2021-04-16 17:45:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][610/1251] eta 0:03:15 lr 0.000050 time 0.2454 (0.3048) loss 3.5436 (3.0366) grad_norm 2.5398 (nan) [2021-04-16 17:45:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][620/1251] eta 0:03:12 lr 0.000050 time 0.2832 (0.3045) loss 3.7057 (3.0417) grad_norm 2.6974 (nan) [2021-04-16 17:45:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][630/1251] eta 0:03:08 lr 0.000050 time 0.2882 (0.3040) loss 3.5234 (3.0400) grad_norm 2.3402 (nan) [2021-04-16 17:46:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][640/1251] eta 0:03:05 lr 0.000050 time 0.2867 (0.3036) loss 3.6429 (3.0388) grad_norm 3.0763 (nan) [2021-04-16 17:46:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][650/1251] eta 0:03:02 lr 0.000050 time 0.2787 (0.3032) loss 3.2642 (3.0407) grad_norm 2.7873 (nan) [2021-04-16 17:46:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][660/1251] eta 0:02:59 lr 0.000050 time 0.2828 (0.3029) loss 2.1449 (3.0367) grad_norm 2.6513 (nan) [2021-04-16 17:46:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][670/1251] eta 0:02:55 lr 0.000050 time 0.2745 (0.3025) loss 2.2658 (3.0333) grad_norm 2.6775 (nan) [2021-04-16 17:46:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][680/1251] eta 0:02:52 lr 0.000050 time 0.2910 (0.3022) loss 3.2843 (3.0312) grad_norm 3.5033 (nan) [2021-04-16 17:46:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][690/1251] eta 0:02:49 lr 0.000050 time 0.2526 (0.3018) loss 2.2117 (3.0278) grad_norm 2.4308 (nan) [2021-04-16 17:46:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][700/1251] eta 0:02:46 lr 0.000050 time 0.3928 (0.3016) loss 2.3532 (3.0238) grad_norm 2.4712 (nan) [2021-04-16 17:46:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][710/1251] eta 0:02:42 lr 0.000050 time 0.2821 (0.3012) loss 3.0700 (3.0181) grad_norm 3.1194 (nan) [2021-04-16 17:46:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][720/1251] eta 0:02:39 lr 0.000050 time 0.2943 (0.3010) loss 3.6582 (3.0191) grad_norm 2.5831 (nan) [2021-04-16 17:46:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][730/1251] eta 0:02:36 lr 0.000050 time 0.2857 (0.3008) loss 3.0855 (3.0191) grad_norm 3.3326 (nan) [2021-04-16 17:46:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][740/1251] eta 0:02:33 lr 0.000050 time 0.2841 (0.3005) loss 2.4301 (3.0205) grad_norm 3.2671 (nan) [2021-04-16 17:46:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][750/1251] eta 0:02:30 lr 0.000049 time 0.2926 (0.3002) loss 2.5544 (3.0193) grad_norm 3.1999 (nan) [2021-04-16 17:46:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][760/1251] eta 0:02:27 lr 0.000049 time 0.3134 (0.3000) loss 2.9582 (3.0185) grad_norm 2.9219 (nan) [2021-04-16 17:46:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][770/1251] eta 0:02:24 lr 0.000049 time 0.2982 (0.2997) loss 2.3144 (3.0203) grad_norm 2.9973 (nan) [2021-04-16 17:46:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][780/1251] eta 0:02:20 lr 0.000049 time 0.2832 (0.2993) loss 2.5172 (3.0202) grad_norm 2.9583 (nan) [2021-04-16 17:46:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][790/1251] eta 0:02:17 lr 0.000049 time 0.2813 (0.2991) loss 3.1581 (3.0216) grad_norm 2.6139 (nan) [2021-04-16 17:46:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][800/1251] eta 0:02:14 lr 0.000049 time 0.2695 (0.2988) loss 3.3790 (3.0229) grad_norm 3.2498 (nan) [2021-04-16 17:46:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][810/1251] eta 0:02:11 lr 0.000049 time 0.3017 (0.2985) loss 2.9131 (3.0239) grad_norm 2.9798 (nan) [2021-04-16 17:46:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][820/1251] eta 0:02:08 lr 0.000049 time 0.2897 (0.2982) loss 2.9971 (3.0252) grad_norm 2.6863 (nan) [2021-04-16 17:46:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][830/1251] eta 0:02:05 lr 0.000049 time 0.2704 (0.2979) loss 2.5457 (3.0198) grad_norm 3.1281 (nan) [2021-04-16 17:46:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][840/1251] eta 0:02:02 lr 0.000049 time 0.2886 (0.2977) loss 3.5501 (3.0220) grad_norm 3.7171 (nan) [2021-04-16 17:47:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][850/1251] eta 0:01:59 lr 0.000049 time 0.2427 (0.2974) loss 2.3817 (3.0231) grad_norm 2.7746 (nan) [2021-04-16 17:47:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][860/1251] eta 0:01:56 lr 0.000049 time 0.2799 (0.2971) loss 3.5656 (3.0255) grad_norm 2.4660 (nan) [2021-04-16 17:47:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][870/1251] eta 0:01:53 lr 0.000049 time 0.3300 (0.2969) loss 3.1960 (3.0262) grad_norm 2.6536 (nan) [2021-04-16 17:47:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][880/1251] eta 0:01:50 lr 0.000049 time 0.2659 (0.2967) loss 3.7926 (3.0273) grad_norm 2.6054 (nan) [2021-04-16 17:47:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][890/1251] eta 0:01:47 lr 0.000049 time 0.2931 (0.2965) loss 3.1544 (3.0270) grad_norm 2.8393 (nan) [2021-04-16 17:47:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][900/1251] eta 0:01:43 lr 0.000049 time 0.2969 (0.2963) loss 3.2116 (3.0271) grad_norm 2.5388 (nan) [2021-04-16 17:47:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][910/1251] eta 0:01:40 lr 0.000049 time 0.3017 (0.2961) loss 3.0324 (3.0286) grad_norm 3.1943 (nan) [2021-04-16 17:47:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][920/1251] eta 0:01:37 lr 0.000049 time 0.2754 (0.2959) loss 3.2656 (3.0306) grad_norm 2.7696 (nan) [2021-04-16 17:47:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][930/1251] eta 0:01:34 lr 0.000049 time 0.2707 (0.2957) loss 3.3965 (3.0328) grad_norm 2.7933 (nan) [2021-04-16 17:47:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][940/1251] eta 0:01:32 lr 0.000049 time 0.2772 (0.2958) loss 3.3391 (3.0337) grad_norm 2.9296 (nan) [2021-04-16 17:47:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][950/1251] eta 0:01:29 lr 0.000049 time 0.2739 (0.2958) loss 3.0666 (3.0350) grad_norm 3.3601 (nan) [2021-04-16 17:47:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][960/1251] eta 0:01:26 lr 0.000049 time 0.2985 (0.2956) loss 3.1793 (3.0348) grad_norm 2.9113 (nan) [2021-04-16 17:47:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][970/1251] eta 0:01:23 lr 0.000049 time 0.2786 (0.2954) loss 3.0473 (3.0355) grad_norm 2.8899 (nan) [2021-04-16 17:47:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][980/1251] eta 0:01:20 lr 0.000049 time 0.3007 (0.2954) loss 2.5327 (3.0353) grad_norm 3.3267 (nan) [2021-04-16 17:47:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][990/1251] eta 0:01:17 lr 0.000049 time 0.2677 (0.2952) loss 2.5312 (3.0366) grad_norm 3.2036 (nan) [2021-04-16 17:47:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][1000/1251] eta 0:01:14 lr 0.000049 time 0.2664 (0.2950) loss 3.5797 (3.0357) grad_norm 3.4661 (nan) [2021-04-16 17:47:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][1010/1251] eta 0:01:11 lr 0.000049 time 0.3067 (0.2948) loss 3.3320 (3.0356) grad_norm 2.8899 (nan) [2021-04-16 17:47:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][1020/1251] eta 0:01:08 lr 0.000049 time 0.2920 (0.2946) loss 3.0776 (3.0342) grad_norm 2.7825 (nan) [2021-04-16 17:47:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][1030/1251] eta 0:01:05 lr 0.000049 time 0.2622 (0.2944) loss 3.8109 (3.0368) grad_norm 2.7911 (nan) [2021-04-16 17:47:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][1040/1251] eta 0:01:02 lr 0.000049 time 0.2462 (0.2942) loss 3.2147 (3.0345) grad_norm 3.1042 (nan) [2021-04-16 17:47:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][1050/1251] eta 0:00:59 lr 0.000049 time 0.3698 (0.2942) loss 3.5663 (3.0346) grad_norm 2.6953 (nan) [2021-04-16 17:47:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][1060/1251] eta 0:00:56 lr 0.000049 time 0.3070 (0.2941) loss 3.2577 (3.0355) grad_norm 2.7028 (nan) [2021-04-16 17:48:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][1070/1251] eta 0:00:53 lr 0.000049 time 0.2852 (0.2941) loss 2.5591 (3.0342) grad_norm 3.0597 (nan) [2021-04-16 17:48:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][1080/1251] eta 0:00:50 lr 0.000049 time 0.2864 (0.2940) loss 2.3359 (3.0329) grad_norm 2.8034 (nan) [2021-04-16 17:48:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][1090/1251] eta 0:00:47 lr 0.000049 time 0.2675 (0.2938) loss 2.7165 (3.0318) grad_norm 2.9789 (nan) [2021-04-16 17:48:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][1100/1251] eta 0:00:44 lr 0.000049 time 0.2704 (0.2936) loss 2.2203 (3.0300) grad_norm 3.5752 (nan) [2021-04-16 17:48:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][1110/1251] eta 0:00:41 lr 0.000049 time 0.2650 (0.2934) loss 2.9235 (3.0287) grad_norm 4.2056 (nan) [2021-04-16 17:48:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][1120/1251] eta 0:00:38 lr 0.000049 time 0.2603 (0.2933) loss 3.4023 (3.0300) grad_norm 2.6601 (nan) [2021-04-16 17:48:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][1130/1251] eta 0:00:35 lr 0.000049 time 0.3082 (0.2932) loss 2.9162 (3.0311) grad_norm 2.8028 (nan) [2021-04-16 17:48:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][1140/1251] eta 0:00:32 lr 0.000049 time 0.2800 (0.2932) loss 3.1045 (3.0300) grad_norm 3.5744 (nan) [2021-04-16 17:48:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][1150/1251] eta 0:00:29 lr 0.000049 time 0.2918 (0.2931) loss 2.6570 (3.0293) grad_norm 3.9373 (nan) [2021-04-16 17:48:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][1160/1251] eta 0:00:26 lr 0.000049 time 0.2705 (0.2930) loss 3.2966 (3.0285) grad_norm 2.9802 (nan) [2021-04-16 17:48:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][1170/1251] eta 0:00:23 lr 0.000049 time 0.3034 (0.2929) loss 3.5381 (3.0290) grad_norm 2.9088 (nan) [2021-04-16 17:48:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][1180/1251] eta 0:00:20 lr 0.000049 time 0.2678 (0.2928) loss 3.5614 (3.0315) grad_norm 2.4849 (nan) [2021-04-16 17:48:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][1190/1251] eta 0:00:17 lr 0.000049 time 0.2422 (0.2926) loss 2.0179 (3.0299) grad_norm 2.9061 (nan) [2021-04-16 17:48:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][1200/1251] eta 0:00:14 lr 0.000049 time 0.3028 (0.2925) loss 3.1257 (3.0310) grad_norm 3.2013 (nan) [2021-04-16 17:48:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][1210/1251] eta 0:00:11 lr 0.000049 time 0.2641 (0.2924) loss 2.2222 (3.0294) grad_norm 3.0667 (nan) [2021-04-16 17:48:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][1220/1251] eta 0:00:09 lr 0.000049 time 0.2926 (0.2923) loss 3.9428 (3.0311) grad_norm 3.3882 (nan) [2021-04-16 17:48:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][1230/1251] eta 0:00:06 lr 0.000049 time 0.2905 (0.2922) loss 2.3291 (3.0301) grad_norm 2.6264 (nan) [2021-04-16 17:48:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][1240/1251] eta 0:00:03 lr 0.000049 time 0.2482 (0.2920) loss 3.2032 (3.0319) grad_norm 3.4230 (nan) [2021-04-16 17:48:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [261/300][1250/1251] eta 0:00:00 lr 0.000049 time 0.2476 (0.2916) loss 2.8568 (3.0328) grad_norm 3.2759 (nan) [2021-04-16 17:48:59 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 261 training takes 0:06:12 [2021-04-16 17:48:59 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_261.pth saving...... [2021-04-16 17:49:23 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_261.pth saved !!! [2021-04-16 17:49:24 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.326 (1.326) Loss 0.7694 (0.7694) Acc@1 82.422 (82.422) Acc@5 96.289 (96.289) [2021-04-16 17:49:26 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.115 (0.283) Loss 0.9169 (0.8308) Acc@1 78.613 (80.176) Acc@5 93.945 (95.543) [2021-04-16 17:49:28 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.121 (0.239) Loss 0.8448 (0.8319) Acc@1 81.641 (80.339) Acc@5 94.531 (95.433) [2021-04-16 17:49:30 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.334 (0.240) Loss 0.9052 (0.8338) Acc@1 79.395 (80.406) Acc@5 94.629 (95.385) [2021-04-16 17:49:32 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.173 (0.221) Loss 0.7749 (0.8354) Acc@1 81.543 (80.576) Acc@5 96.387 (95.324) [2021-04-16 17:49:47 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.738 Acc@5 95.346 [2021-04-16 17:49:47 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 80.7% [2021-04-16 17:49:47 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 80.78% [2021-04-16 17:49:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][0/1251] eta 3:13:54 lr 0.000049 time 9.3004 (9.3004) loss 2.2612 (2.2612) grad_norm 2.7244 (2.7244) [2021-04-16 17:49:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][10/1251] eta 0:22:43 lr 0.000049 time 0.2680 (1.0988) loss 3.5534 (3.1009) grad_norm 2.4188 (2.7399) [2021-04-16 17:50:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][20/1251] eta 0:14:36 lr 0.000049 time 0.2971 (0.7124) loss 3.3175 (3.0978) grad_norm 2.6658 (2.8421) [2021-04-16 17:50:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][30/1251] eta 0:11:39 lr 0.000049 time 0.2706 (0.5725) loss 3.1914 (3.0303) grad_norm 3.2099 (2.8787) [2021-04-16 17:50:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][40/1251] eta 0:10:07 lr 0.000049 time 0.2757 (0.5015) loss 3.6910 (3.0268) grad_norm 3.4466 (2.8675) [2021-04-16 17:50:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][50/1251] eta 0:09:10 lr 0.000049 time 0.2980 (0.4581) loss 3.1966 (3.0024) grad_norm 3.6985 (2.8992) [2021-04-16 17:50:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][60/1251] eta 0:08:30 lr 0.000049 time 0.2698 (0.4290) loss 3.6500 (3.0193) grad_norm 2.4678 (2.9138) [2021-04-16 17:50:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][70/1251] eta 0:08:01 lr 0.000049 time 0.2693 (0.4074) loss 3.4004 (3.0225) grad_norm 2.7620 (2.9098) [2021-04-16 17:50:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][80/1251] eta 0:07:38 lr 0.000049 time 0.2702 (0.3915) loss 2.1288 (3.0010) grad_norm 2.6693 (2.9113) [2021-04-16 17:50:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][90/1251] eta 0:07:20 lr 0.000049 time 0.2566 (0.3793) loss 2.7848 (2.9941) grad_norm 2.5627 (2.9166) [2021-04-16 17:50:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][100/1251] eta 0:07:05 lr 0.000049 time 0.2934 (0.3698) loss 3.0449 (2.9807) grad_norm 3.3074 (2.9198) [2021-04-16 17:50:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][110/1251] eta 0:06:52 lr 0.000049 time 0.2587 (0.3614) loss 1.9496 (2.9860) grad_norm 4.8035 (2.9400) [2021-04-16 17:50:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][120/1251] eta 0:06:42 lr 0.000048 time 0.2665 (0.3554) loss 3.1439 (2.9721) grad_norm 2.9208 (2.9473) [2021-04-16 17:50:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][130/1251] eta 0:06:31 lr 0.000048 time 0.3220 (0.3495) loss 3.4310 (2.9795) grad_norm 3.1551 (2.9448) [2021-04-16 17:50:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][140/1251] eta 0:06:23 lr 0.000048 time 0.2665 (0.3453) loss 3.0015 (2.9828) grad_norm 2.8683 (2.9499) [2021-04-16 17:50:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][150/1251] eta 0:06:17 lr 0.000048 time 0.2921 (0.3432) loss 3.2247 (2.9978) grad_norm 2.7210 (2.9458) [2021-04-16 17:50:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][160/1251] eta 0:06:10 lr 0.000048 time 0.2773 (0.3392) loss 3.3166 (2.9893) grad_norm 3.1171 (2.9459) [2021-04-16 17:50:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][170/1251] eta 0:06:02 lr 0.000048 time 0.2864 (0.3355) loss 2.6576 (2.9930) grad_norm 2.9416 (2.9453) [2021-04-16 17:50:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][180/1251] eta 0:05:56 lr 0.000048 time 0.4303 (0.3332) loss 3.0721 (3.0131) grad_norm 2.8092 (2.9420) [2021-04-16 17:50:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][190/1251] eta 0:05:50 lr 0.000048 time 0.2720 (0.3303) loss 2.4553 (3.0120) grad_norm 2.8908 (2.9346) [2021-04-16 17:50:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][200/1251] eta 0:05:44 lr 0.000048 time 0.3077 (0.3279) loss 2.6407 (2.9991) grad_norm 2.9616 (2.9369) [2021-04-16 17:50:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][210/1251] eta 0:05:38 lr 0.000048 time 0.2708 (0.3255) loss 3.0965 (2.9910) grad_norm 2.6348 (2.9356) [2021-04-16 17:50:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][220/1251] eta 0:05:33 lr 0.000048 time 0.2789 (0.3233) loss 3.0394 (3.0025) grad_norm 3.4318 (2.9409) [2021-04-16 17:51:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][230/1251] eta 0:05:28 lr 0.000048 time 0.2697 (0.3217) loss 3.0548 (3.0076) grad_norm 2.9788 (2.9781) [2021-04-16 17:51:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][240/1251] eta 0:05:23 lr 0.000048 time 0.2767 (0.3198) loss 3.4338 (3.0098) grad_norm 2.5673 (2.9676) [2021-04-16 17:51:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][250/1251] eta 0:05:18 lr 0.000048 time 0.2759 (0.3181) loss 3.4131 (3.0086) grad_norm 3.4609 (2.9772) [2021-04-16 17:51:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][260/1251] eta 0:05:14 lr 0.000048 time 0.4874 (0.3173) loss 2.4743 (3.0151) grad_norm 3.3204 (2.9719) [2021-04-16 17:51:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][270/1251] eta 0:05:09 lr 0.000048 time 0.2594 (0.3157) loss 3.4419 (3.0209) grad_norm 3.0994 (2.9740) [2021-04-16 17:51:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][280/1251] eta 0:05:05 lr 0.000048 time 0.2902 (0.3144) loss 2.3166 (3.0185) grad_norm 2.6239 (2.9756) [2021-04-16 17:51:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][290/1251] eta 0:05:01 lr 0.000048 time 0.2970 (0.3137) loss 3.1540 (3.0220) grad_norm 2.5975 (2.9762) [2021-04-16 17:51:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][300/1251] eta 0:04:57 lr 0.000048 time 0.2744 (0.3127) loss 2.6793 (3.0239) grad_norm 2.8589 (2.9750) [2021-04-16 17:51:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][310/1251] eta 0:04:53 lr 0.000048 time 0.2604 (0.3115) loss 3.2221 (3.0282) grad_norm 2.8405 (2.9764) [2021-04-16 17:51:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][320/1251] eta 0:04:48 lr 0.000048 time 0.2704 (0.3103) loss 2.8402 (3.0277) grad_norm 3.2333 (2.9784) [2021-04-16 17:51:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][330/1251] eta 0:04:44 lr 0.000048 time 0.2957 (0.3094) loss 3.3120 (3.0348) grad_norm 2.7850 (2.9813) [2021-04-16 17:51:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][340/1251] eta 0:04:40 lr 0.000048 time 0.2522 (0.3084) loss 3.0630 (3.0330) grad_norm 2.5606 (2.9792) [2021-04-16 17:51:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][350/1251] eta 0:04:37 lr 0.000048 time 0.2716 (0.3081) loss 4.1029 (3.0380) grad_norm 2.9921 (2.9779) [2021-04-16 17:51:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][360/1251] eta 0:04:33 lr 0.000048 time 0.3075 (0.3074) loss 2.9796 (3.0394) grad_norm 3.5765 (2.9760) [2021-04-16 17:51:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][370/1251] eta 0:04:30 lr 0.000048 time 0.2944 (0.3066) loss 2.7499 (3.0332) grad_norm 2.8660 (2.9772) [2021-04-16 17:51:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][380/1251] eta 0:04:26 lr 0.000048 time 0.2633 (0.3058) loss 3.7900 (3.0366) grad_norm 3.0168 (2.9793) [2021-04-16 17:51:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][390/1251] eta 0:04:22 lr 0.000048 time 0.2600 (0.3050) loss 2.5622 (3.0341) grad_norm 3.5055 (2.9856) [2021-04-16 17:51:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][400/1251] eta 0:04:19 lr 0.000048 time 0.2934 (0.3045) loss 3.3339 (3.0356) grad_norm 3.5619 (2.9863) [2021-04-16 17:51:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][410/1251] eta 0:04:15 lr 0.000048 time 0.2635 (0.3039) loss 3.3771 (3.0416) grad_norm 3.4076 (2.9864) [2021-04-16 17:51:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][420/1251] eta 0:04:11 lr 0.000048 time 0.2818 (0.3031) loss 2.6045 (3.0349) grad_norm 2.7549 (2.9837) [2021-04-16 17:51:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][430/1251] eta 0:04:08 lr 0.000048 time 0.2630 (0.3024) loss 3.3908 (3.0326) grad_norm 2.4018 (2.9834) [2021-04-16 17:52:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][440/1251] eta 0:04:04 lr 0.000048 time 0.2835 (0.3019) loss 2.9229 (3.0301) grad_norm 3.7493 (2.9860) [2021-04-16 17:52:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][450/1251] eta 0:04:01 lr 0.000048 time 0.2632 (0.3014) loss 2.2109 (3.0206) grad_norm 2.7992 (2.9838) [2021-04-16 17:52:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][460/1251] eta 0:03:58 lr 0.000048 time 0.2983 (0.3012) loss 3.6910 (3.0156) grad_norm 2.6090 (2.9804) [2021-04-16 17:52:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][470/1251] eta 0:03:54 lr 0.000048 time 0.2685 (0.3007) loss 3.3088 (3.0127) grad_norm 3.2534 (2.9838) [2021-04-16 17:52:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][480/1251] eta 0:03:51 lr 0.000048 time 0.2586 (0.3003) loss 3.1798 (3.0145) grad_norm 2.7541 (2.9879) [2021-04-16 17:52:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][490/1251] eta 0:03:48 lr 0.000048 time 0.2981 (0.2999) loss 3.6657 (3.0195) grad_norm 3.4169 (2.9887) [2021-04-16 17:52:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][500/1251] eta 0:03:44 lr 0.000048 time 0.2704 (0.2994) loss 3.1779 (3.0188) grad_norm 4.1859 (2.9933) [2021-04-16 17:52:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][510/1251] eta 0:03:41 lr 0.000048 time 0.2608 (0.2990) loss 3.5404 (3.0182) grad_norm 2.9689 (2.9920) [2021-04-16 17:52:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][520/1251] eta 0:03:38 lr 0.000048 time 0.2906 (0.2987) loss 2.9041 (3.0194) grad_norm 3.1477 (2.9911) [2021-04-16 17:52:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][530/1251] eta 0:03:35 lr 0.000048 time 0.3184 (0.2985) loss 2.6967 (3.0212) grad_norm 2.9553 (2.9917) [2021-04-16 17:52:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][540/1251] eta 0:03:32 lr 0.000048 time 0.2698 (0.2983) loss 3.7947 (3.0230) grad_norm 3.3332 (2.9921) [2021-04-16 17:52:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][550/1251] eta 0:03:29 lr 0.000048 time 0.2809 (0.2983) loss 3.0291 (3.0273) grad_norm 3.0582 (2.9940) [2021-04-16 17:52:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][560/1251] eta 0:03:25 lr 0.000048 time 0.2739 (0.2980) loss 2.7677 (3.0252) grad_norm 2.7064 (2.9929) [2021-04-16 17:52:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][570/1251] eta 0:03:22 lr 0.000048 time 0.4194 (0.2979) loss 3.4696 (3.0269) grad_norm 2.7673 (2.9906) [2021-04-16 17:52:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][580/1251] eta 0:03:19 lr 0.000048 time 0.2473 (0.2980) loss 3.0678 (3.0315) grad_norm 3.0555 (2.9901) [2021-04-16 17:52:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][590/1251] eta 0:03:16 lr 0.000048 time 0.2730 (0.2977) loss 2.5699 (3.0294) grad_norm 3.9470 (2.9920) [2021-04-16 17:52:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][600/1251] eta 0:03:13 lr 0.000048 time 0.2818 (0.2974) loss 2.1257 (3.0303) grad_norm 2.6936 (2.9906) [2021-04-16 17:52:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][610/1251] eta 0:03:10 lr 0.000048 time 0.2685 (0.2970) loss 3.5321 (3.0325) grad_norm 3.5294 (2.9894) [2021-04-16 17:52:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][620/1251] eta 0:03:07 lr 0.000048 time 0.2672 (0.2966) loss 3.0264 (3.0340) grad_norm 3.1142 (2.9883) [2021-04-16 17:52:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][630/1251] eta 0:03:04 lr 0.000048 time 0.2830 (0.2964) loss 3.0990 (3.0324) grad_norm 3.2626 (2.9888) [2021-04-16 17:52:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][640/1251] eta 0:03:00 lr 0.000048 time 0.2776 (0.2961) loss 3.2950 (3.0320) grad_norm 2.9667 (2.9885) [2021-04-16 17:52:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][650/1251] eta 0:02:57 lr 0.000048 time 0.3021 (0.2960) loss 3.5208 (3.0309) grad_norm 2.7058 (2.9877) [2021-04-16 17:53:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][660/1251] eta 0:02:54 lr 0.000048 time 0.2607 (0.2957) loss 2.8168 (3.0317) grad_norm 3.5705 (2.9889) [2021-04-16 17:53:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][670/1251] eta 0:02:51 lr 0.000048 time 0.2688 (0.2955) loss 3.1828 (3.0328) grad_norm 2.6924 (2.9900) [2021-04-16 17:53:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][680/1251] eta 0:02:48 lr 0.000048 time 0.2579 (0.2952) loss 3.1069 (3.0312) grad_norm 2.7220 (2.9897) [2021-04-16 17:53:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][690/1251] eta 0:02:45 lr 0.000048 time 0.2725 (0.2952) loss 3.0268 (3.0316) grad_norm 3.9101 (2.9902) [2021-04-16 17:53:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][700/1251] eta 0:02:42 lr 0.000048 time 0.2741 (0.2950) loss 2.1075 (3.0294) grad_norm 2.8968 (2.9918) [2021-04-16 17:53:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][710/1251] eta 0:02:39 lr 0.000048 time 0.2690 (0.2947) loss 2.5854 (3.0314) grad_norm 3.2461 (2.9898) [2021-04-16 17:53:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][720/1251] eta 0:02:36 lr 0.000048 time 0.2521 (0.2947) loss 2.9848 (3.0309) grad_norm 3.2429 (2.9916) [2021-04-16 17:53:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][730/1251] eta 0:02:33 lr 0.000048 time 0.2909 (0.2947) loss 3.4335 (3.0347) grad_norm 2.8349 (2.9919) [2021-04-16 17:53:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][740/1251] eta 0:02:30 lr 0.000047 time 0.2719 (0.2945) loss 3.4092 (3.0367) grad_norm 4.0117 (2.9956) [2021-04-16 17:53:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][750/1251] eta 0:02:27 lr 0.000047 time 0.2684 (0.2942) loss 3.2662 (3.0371) grad_norm 2.5453 (2.9927) [2021-04-16 17:53:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][760/1251] eta 0:02:24 lr 0.000047 time 0.2553 (0.2941) loss 3.7317 (3.0388) grad_norm 3.1878 (2.9923) [2021-04-16 17:53:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][770/1251] eta 0:02:21 lr 0.000047 time 0.2550 (0.2939) loss 3.6203 (3.0420) grad_norm 3.7735 (2.9936) [2021-04-16 17:53:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][780/1251] eta 0:02:18 lr 0.000047 time 0.2969 (0.2937) loss 3.7962 (3.0459) grad_norm 3.1291 (2.9918) [2021-04-16 17:53:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][790/1251] eta 0:02:15 lr 0.000047 time 0.3027 (0.2936) loss 3.4245 (3.0484) grad_norm 2.8979 (2.9925) [2021-04-16 17:53:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][800/1251] eta 0:02:12 lr 0.000047 time 0.3102 (0.2934) loss 2.4972 (3.0446) grad_norm 3.2950 (2.9909) [2021-04-16 17:53:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][810/1251] eta 0:02:09 lr 0.000047 time 0.3120 (0.2932) loss 3.6287 (3.0458) grad_norm 2.8647 (2.9914) [2021-04-16 17:53:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][820/1251] eta 0:02:06 lr 0.000047 time 0.2955 (0.2932) loss 4.0465 (3.0472) grad_norm 2.9633 (2.9925) [2021-04-16 17:53:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][830/1251] eta 0:02:03 lr 0.000047 time 0.2918 (0.2929) loss 2.7187 (3.0472) grad_norm 2.8596 (nan) [2021-04-16 17:53:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][840/1251] eta 0:02:00 lr 0.000047 time 0.2811 (0.2930) loss 2.7904 (3.0498) grad_norm 3.2687 (nan) [2021-04-16 17:53:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][850/1251] eta 0:01:57 lr 0.000047 time 0.3067 (0.2930) loss 3.1549 (3.0497) grad_norm 2.8758 (nan) [2021-04-16 17:53:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][860/1251] eta 0:01:54 lr 0.000047 time 0.2756 (0.2928) loss 3.0748 (3.0521) grad_norm 3.2842 (nan) [2021-04-16 17:54:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][870/1251] eta 0:01:51 lr 0.000047 time 0.2763 (0.2929) loss 3.3520 (3.0533) grad_norm 3.8035 (nan) [2021-04-16 17:54:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][880/1251] eta 0:01:48 lr 0.000047 time 0.2657 (0.2927) loss 3.1314 (3.0550) grad_norm 2.7720 (nan) [2021-04-16 17:54:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][890/1251] eta 0:01:45 lr 0.000047 time 0.3121 (0.2925) loss 3.1242 (3.0548) grad_norm 2.8220 (nan) [2021-04-16 17:54:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][900/1251] eta 0:01:42 lr 0.000047 time 0.2899 (0.2924) loss 2.2660 (3.0564) grad_norm 2.7982 (nan) [2021-04-16 17:54:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][910/1251] eta 0:01:39 lr 0.000047 time 0.2853 (0.2923) loss 3.3415 (3.0551) grad_norm 2.9026 (nan) [2021-04-16 17:54:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][920/1251] eta 0:01:36 lr 0.000047 time 0.2732 (0.2921) loss 2.9683 (3.0599) grad_norm 2.7946 (nan) [2021-04-16 17:54:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][930/1251] eta 0:01:33 lr 0.000047 time 0.2814 (0.2921) loss 2.5116 (3.0602) grad_norm 3.0498 (nan) [2021-04-16 17:54:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][940/1251] eta 0:01:30 lr 0.000047 time 0.2819 (0.2920) loss 2.8311 (3.0592) grad_norm 3.4027 (nan) [2021-04-16 17:54:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][950/1251] eta 0:01:27 lr 0.000047 time 0.2822 (0.2919) loss 3.0028 (3.0593) grad_norm 3.1906 (nan) [2021-04-16 17:54:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][960/1251] eta 0:01:24 lr 0.000047 time 0.2936 (0.2918) loss 3.3944 (3.0623) grad_norm 3.1312 (nan) [2021-04-16 17:54:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][970/1251] eta 0:01:21 lr 0.000047 time 0.2862 (0.2916) loss 2.4830 (3.0592) grad_norm 2.9146 (nan) [2021-04-16 17:54:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][980/1251] eta 0:01:19 lr 0.000047 time 0.2656 (0.2916) loss 3.3878 (3.0613) grad_norm 3.1126 (nan) [2021-04-16 17:54:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][990/1251] eta 0:01:16 lr 0.000047 time 0.2624 (0.2914) loss 2.6903 (3.0583) grad_norm 3.0630 (nan) [2021-04-16 17:54:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][1000/1251] eta 0:01:13 lr 0.000047 time 0.2748 (0.2913) loss 3.1674 (3.0551) grad_norm 2.7543 (nan) [2021-04-16 17:54:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][1010/1251] eta 0:01:10 lr 0.000047 time 0.2851 (0.2912) loss 1.9095 (3.0547) grad_norm 2.7125 (nan) [2021-04-16 17:54:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][1020/1251] eta 0:01:07 lr 0.000047 time 0.2680 (0.2911) loss 3.3847 (3.0540) grad_norm 2.9632 (nan) [2021-04-16 17:54:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][1030/1251] eta 0:01:04 lr 0.000047 time 0.2758 (0.2909) loss 3.4155 (3.0530) grad_norm 3.2133 (nan) [2021-04-16 17:54:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][1040/1251] eta 0:01:01 lr 0.000047 time 0.3046 (0.2908) loss 3.4070 (3.0530) grad_norm 3.2367 (nan) [2021-04-16 17:54:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][1050/1251] eta 0:00:58 lr 0.000047 time 0.2645 (0.2907) loss 2.9653 (3.0554) grad_norm 3.5245 (nan) [2021-04-16 17:54:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][1060/1251] eta 0:00:55 lr 0.000047 time 0.2820 (0.2906) loss 3.3376 (3.0552) grad_norm 3.2638 (nan) [2021-04-16 17:54:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][1070/1251] eta 0:00:52 lr 0.000047 time 0.2728 (0.2906) loss 2.6634 (3.0550) grad_norm 3.3240 (nan) [2021-04-16 17:55:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][1080/1251] eta 0:00:49 lr 0.000047 time 0.2620 (0.2904) loss 2.8509 (3.0551) grad_norm 3.1494 (nan) [2021-04-16 17:55:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][1090/1251] eta 0:00:46 lr 0.000047 time 0.2983 (0.2904) loss 2.7483 (3.0531) grad_norm 3.4322 (nan) [2021-04-16 17:55:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][1100/1251] eta 0:00:43 lr 0.000047 time 0.2577 (0.2902) loss 3.0754 (3.0549) grad_norm 3.0734 (nan) [2021-04-16 17:55:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][1110/1251] eta 0:00:40 lr 0.000047 time 0.2960 (0.2901) loss 3.1417 (3.0568) grad_norm 2.7097 (nan) [2021-04-16 17:55:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][1120/1251] eta 0:00:37 lr 0.000047 time 0.2882 (0.2900) loss 3.1669 (3.0556) grad_norm 2.8002 (nan) [2021-04-16 17:55:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][1130/1251] eta 0:00:35 lr 0.000047 time 0.2774 (0.2900) loss 2.1718 (3.0547) grad_norm 2.5952 (nan) [2021-04-16 17:55:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][1140/1251] eta 0:00:32 lr 0.000047 time 0.2823 (0.2900) loss 1.9133 (3.0533) grad_norm 2.5543 (nan) [2021-04-16 17:55:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][1150/1251] eta 0:00:29 lr 0.000047 time 0.2991 (0.2901) loss 3.5161 (3.0523) grad_norm 3.0522 (nan) [2021-04-16 17:55:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][1160/1251] eta 0:00:26 lr 0.000047 time 0.2913 (0.2901) loss 2.9109 (3.0503) grad_norm 2.9356 (nan) [2021-04-16 17:55:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][1170/1251] eta 0:00:23 lr 0.000047 time 0.2676 (0.2900) loss 3.5511 (3.0507) grad_norm 2.9700 (nan) [2021-04-16 17:55:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][1180/1251] eta 0:00:20 lr 0.000047 time 0.2979 (0.2899) loss 2.7135 (3.0484) grad_norm 2.9665 (nan) [2021-04-16 17:55:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][1190/1251] eta 0:00:17 lr 0.000047 time 0.2839 (0.2898) loss 2.6590 (3.0498) grad_norm 2.7092 (nan) [2021-04-16 17:55:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][1200/1251] eta 0:00:14 lr 0.000047 time 0.2796 (0.2897) loss 3.5834 (3.0519) grad_norm 2.8612 (nan) [2021-04-16 17:55:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][1210/1251] eta 0:00:11 lr 0.000047 time 0.2905 (0.2896) loss 2.0883 (3.0514) grad_norm 3.0218 (nan) [2021-04-16 17:55:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][1220/1251] eta 0:00:08 lr 0.000047 time 0.2752 (0.2895) loss 3.2796 (3.0509) grad_norm 2.8672 (nan) [2021-04-16 17:55:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][1230/1251] eta 0:00:06 lr 0.000047 time 0.2847 (0.2895) loss 2.6430 (3.0508) grad_norm 2.7858 (nan) [2021-04-16 17:55:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][1240/1251] eta 0:00:03 lr 0.000047 time 0.2483 (0.2894) loss 3.2846 (3.0493) grad_norm 2.7573 (nan) [2021-04-16 17:55:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [262/300][1250/1251] eta 0:00:00 lr 0.000047 time 0.2486 (0.2890) loss 2.0763 (3.0494) grad_norm 2.5842 (nan) [2021-04-16 17:55:54 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 262 training takes 0:06:07 [2021-04-16 17:55:54 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_262.pth saving...... [2021-04-16 17:56:24 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_262.pth saved !!! [2021-04-16 17:56:25 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.188 (1.188) Loss 0.9317 (0.9317) Acc@1 78.809 (78.809) Acc@5 93.848 (93.848) [2021-04-16 17:56:27 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.273 (0.267) Loss 0.8142 (0.8266) Acc@1 80.664 (80.620) Acc@5 95.508 (95.419) [2021-04-16 17:56:29 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.337 (0.247) Loss 0.9185 (0.8250) Acc@1 78.711 (80.738) Acc@5 94.629 (95.364) [2021-04-16 17:56:31 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.120 (0.220) Loss 0.7897 (0.8248) Acc@1 82.324 (80.727) Acc@5 95.801 (95.423) [2021-04-16 17:56:33 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.218) Loss 0.8116 (0.8243) Acc@1 81.836 (80.752) Acc@5 95.508 (95.403) [2021-04-16 17:56:45 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.728 Acc@5 95.358 [2021-04-16 17:56:45 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 80.7% [2021-04-16 17:56:45 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 80.78% [2021-04-16 17:57:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][0/1251] eta 6:14:34 lr 0.000047 time 17.9650 (17.9650) loss 2.3138 (2.3138) grad_norm 2.8629 (2.8629) [2021-04-16 17:57:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][10/1251] eta 0:38:59 lr 0.000047 time 0.2764 (1.8851) loss 3.3687 (2.9807) grad_norm 3.3445 (3.1576) [2021-04-16 17:57:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][20/1251] eta 0:22:55 lr 0.000047 time 0.2769 (1.1176) loss 3.3017 (3.1329) grad_norm 3.1239 (3.1112) [2021-04-16 17:57:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][30/1251] eta 0:17:13 lr 0.000047 time 0.2881 (0.8468) loss 3.4190 (3.0981) grad_norm 2.8437 (3.1258) [2021-04-16 17:57:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][40/1251] eta 0:14:19 lr 0.000047 time 0.2553 (0.7097) loss 2.3848 (3.0711) grad_norm 2.9756 (3.1742) [2021-04-16 17:57:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][50/1251] eta 0:12:29 lr 0.000047 time 0.2781 (0.6243) loss 2.1781 (3.0099) grad_norm 2.7967 (3.1248) [2021-04-16 17:57:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][60/1251] eta 0:11:19 lr 0.000047 time 0.4474 (0.5707) loss 2.3040 (2.9772) grad_norm 3.2090 (3.0982) [2021-04-16 17:57:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][70/1251] eta 0:10:24 lr 0.000047 time 0.2661 (0.5285) loss 3.2627 (2.9789) grad_norm 3.1255 (3.0833) [2021-04-16 17:57:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][80/1251] eta 0:09:44 lr 0.000047 time 0.2725 (0.4994) loss 3.2030 (3.0014) grad_norm 3.4455 (3.0839) [2021-04-16 17:57:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][90/1251] eta 0:09:11 lr 0.000047 time 0.2646 (0.4747) loss 3.2658 (3.0388) grad_norm 3.0979 (3.1084) [2021-04-16 17:57:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][100/1251] eta 0:08:43 lr 0.000047 time 0.2750 (0.4552) loss 2.8063 (3.0171) grad_norm 3.2594 (3.1038) [2021-04-16 17:57:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][110/1251] eta 0:08:21 lr 0.000047 time 0.3047 (0.4397) loss 2.9448 (3.0292) grad_norm 5.0576 (3.1419) [2021-04-16 17:57:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][120/1251] eta 0:08:01 lr 0.000047 time 0.2691 (0.4259) loss 3.0336 (3.0181) grad_norm 2.8395 (3.1357) [2021-04-16 17:57:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][130/1251] eta 0:07:45 lr 0.000046 time 0.2665 (0.4152) loss 3.3497 (2.9982) grad_norm 3.5623 (3.1268) [2021-04-16 17:57:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][140/1251] eta 0:07:30 lr 0.000046 time 0.2775 (0.4053) loss 3.0833 (3.0173) grad_norm 2.7131 (3.1008) [2021-04-16 17:57:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][150/1251] eta 0:07:17 lr 0.000046 time 0.2559 (0.3969) loss 1.9082 (3.0249) grad_norm 3.3041 (3.0852) [2021-04-16 17:57:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][160/1251] eta 0:07:04 lr 0.000046 time 0.2677 (0.3893) loss 3.3747 (3.0362) grad_norm 2.4462 (3.0631) [2021-04-16 17:57:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][170/1251] eta 0:06:53 lr 0.000046 time 0.2854 (0.3827) loss 3.4421 (3.0432) grad_norm 3.0560 (3.0513) [2021-04-16 17:57:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][180/1251] eta 0:06:43 lr 0.000046 time 0.2900 (0.3771) loss 2.5692 (3.0412) grad_norm 3.0217 (3.0507) [2021-04-16 17:57:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][190/1251] eta 0:06:34 lr 0.000046 time 0.2427 (0.3718) loss 4.0425 (3.0536) grad_norm 2.7552 (3.0429) [2021-04-16 17:57:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][200/1251] eta 0:06:26 lr 0.000046 time 0.2679 (0.3673) loss 3.4613 (3.0622) grad_norm 4.2166 (3.0449) [2021-04-16 17:58:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][210/1251] eta 0:06:18 lr 0.000046 time 0.2959 (0.3632) loss 1.9130 (3.0633) grad_norm 2.8082 (3.0472) [2021-04-16 17:58:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][220/1251] eta 0:06:10 lr 0.000046 time 0.2708 (0.3591) loss 3.3776 (3.0634) grad_norm 3.6096 (3.0524) [2021-04-16 17:58:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][230/1251] eta 0:06:02 lr 0.000046 time 0.2737 (0.3554) loss 3.1602 (3.0595) grad_norm 3.4221 (3.0427) [2021-04-16 17:58:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][240/1251] eta 0:05:55 lr 0.000046 time 0.2800 (0.3520) loss 2.3887 (3.0613) grad_norm 2.8857 (3.0408) [2021-04-16 17:58:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][250/1251] eta 0:05:49 lr 0.000046 time 0.2674 (0.3488) loss 2.5565 (3.0638) grad_norm 2.6424 (3.0457) [2021-04-16 17:58:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][260/1251] eta 0:05:42 lr 0.000046 time 0.2598 (0.3460) loss 3.9016 (3.0694) grad_norm 2.8595 (3.0380) [2021-04-16 17:58:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][270/1251] eta 0:05:36 lr 0.000046 time 0.2850 (0.3434) loss 3.0155 (3.0655) grad_norm 3.1943 (3.0326) [2021-04-16 17:58:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][280/1251] eta 0:05:31 lr 0.000046 time 0.2638 (0.3410) loss 2.7577 (3.0678) grad_norm 3.4747 (3.0354) [2021-04-16 17:58:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][290/1251] eta 0:05:25 lr 0.000046 time 0.2802 (0.3389) loss 3.2523 (3.0613) grad_norm 2.9640 (3.0290) [2021-04-16 17:58:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][300/1251] eta 0:05:20 lr 0.000046 time 0.2870 (0.3368) loss 2.7185 (3.0567) grad_norm 2.8014 (3.0271) [2021-04-16 17:58:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][310/1251] eta 0:05:15 lr 0.000046 time 0.2633 (0.3350) loss 3.5891 (3.0582) grad_norm 4.4126 (3.0338) [2021-04-16 17:58:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][320/1251] eta 0:05:10 lr 0.000046 time 0.2554 (0.3335) loss 3.6524 (3.0623) grad_norm 3.0620 (3.0382) [2021-04-16 17:58:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][330/1251] eta 0:05:05 lr 0.000046 time 0.2884 (0.3318) loss 2.5959 (3.0618) grad_norm 2.6226 (3.0413) [2021-04-16 17:58:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][340/1251] eta 0:05:01 lr 0.000046 time 0.2716 (0.3307) loss 3.0074 (3.0649) grad_norm 2.6691 (3.0412) [2021-04-16 17:58:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][350/1251] eta 0:04:56 lr 0.000046 time 0.2784 (0.3295) loss 2.8202 (3.0621) grad_norm 2.7538 (3.0354) [2021-04-16 17:58:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][360/1251] eta 0:04:52 lr 0.000046 time 0.2976 (0.3281) loss 3.6708 (3.0589) grad_norm 3.0034 (3.0387) [2021-04-16 17:58:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][370/1251] eta 0:04:48 lr 0.000046 time 0.2798 (0.3274) loss 3.3160 (3.0587) grad_norm 2.8867 (3.0446) [2021-04-16 17:58:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][380/1251] eta 0:04:43 lr 0.000046 time 0.2673 (0.3260) loss 3.1110 (3.0619) grad_norm 2.9760 (3.0498) [2021-04-16 17:58:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][390/1251] eta 0:04:39 lr 0.000046 time 0.2746 (0.3247) loss 3.4886 (3.0652) grad_norm 2.6494 (3.0581) [2021-04-16 17:58:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][400/1251] eta 0:04:35 lr 0.000046 time 0.2680 (0.3235) loss 3.0523 (3.0608) grad_norm 3.0252 (3.0550) [2021-04-16 17:58:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][410/1251] eta 0:04:31 lr 0.000046 time 0.2696 (0.3225) loss 3.4970 (3.0603) grad_norm 2.9321 (3.0675) [2021-04-16 17:59:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][420/1251] eta 0:04:27 lr 0.000046 time 0.2866 (0.3215) loss 3.6601 (3.0599) grad_norm 3.6444 (3.0672) [2021-04-16 17:59:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][430/1251] eta 0:04:22 lr 0.000046 time 0.2923 (0.3203) loss 3.1602 (3.0635) grad_norm 2.7588 (3.0628) [2021-04-16 17:59:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][440/1251] eta 0:04:18 lr 0.000046 time 0.2790 (0.3193) loss 3.3897 (3.0615) grad_norm 3.1245 (3.0648) [2021-04-16 17:59:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][450/1251] eta 0:04:14 lr 0.000046 time 0.2457 (0.3183) loss 2.6006 (3.0585) grad_norm 2.7109 (3.0646) [2021-04-16 17:59:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][460/1251] eta 0:04:11 lr 0.000046 time 0.2601 (0.3174) loss 3.2573 (3.0551) grad_norm 3.2358 (3.0660) [2021-04-16 17:59:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][470/1251] eta 0:04:07 lr 0.000046 time 0.2973 (0.3166) loss 3.7313 (3.0559) grad_norm 2.8077 (3.0687) [2021-04-16 17:59:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][480/1251] eta 0:04:03 lr 0.000046 time 0.2627 (0.3157) loss 3.4874 (3.0555) grad_norm 2.8630 (3.0737) [2021-04-16 17:59:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][490/1251] eta 0:03:59 lr 0.000046 time 0.2628 (0.3151) loss 2.0338 (3.0588) grad_norm 3.7261 (3.0784) [2021-04-16 17:59:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][500/1251] eta 0:03:56 lr 0.000046 time 0.2694 (0.3144) loss 3.7163 (3.0622) grad_norm 3.1974 (3.0820) [2021-04-16 17:59:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][510/1251] eta 0:03:52 lr 0.000046 time 0.2829 (0.3137) loss 2.4714 (3.0613) grad_norm 3.1927 (3.0833) [2021-04-16 17:59:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][520/1251] eta 0:03:48 lr 0.000046 time 0.2540 (0.3129) loss 3.3582 (3.0614) grad_norm 3.2398 (3.0820) [2021-04-16 17:59:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][530/1251] eta 0:03:45 lr 0.000046 time 0.2559 (0.3122) loss 3.9752 (3.0654) grad_norm 2.5340 (3.0781) [2021-04-16 17:59:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][540/1251] eta 0:03:41 lr 0.000046 time 0.2856 (0.3119) loss 3.5444 (3.0665) grad_norm 6.1293 (3.0855) [2021-04-16 17:59:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][550/1251] eta 0:03:38 lr 0.000046 time 0.2717 (0.3114) loss 2.3552 (3.0654) grad_norm 2.9027 (3.0964) [2021-04-16 17:59:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][560/1251] eta 0:03:34 lr 0.000046 time 0.2690 (0.3108) loss 2.1702 (3.0645) grad_norm 2.5299 (3.0951) [2021-04-16 17:59:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][570/1251] eta 0:03:31 lr 0.000046 time 0.2703 (0.3104) loss 3.5358 (3.0658) grad_norm 2.8037 (3.0936) [2021-04-16 17:59:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][580/1251] eta 0:03:28 lr 0.000046 time 0.2870 (0.3101) loss 3.1842 (3.0649) grad_norm 3.0954 (3.0939) [2021-04-16 17:59:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][590/1251] eta 0:03:24 lr 0.000046 time 0.2548 (0.3094) loss 3.1765 (3.0659) grad_norm 2.7986 (3.0944) [2021-04-16 17:59:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][600/1251] eta 0:03:21 lr 0.000046 time 0.2605 (0.3089) loss 2.8894 (3.0634) grad_norm 3.9654 (3.0963) [2021-04-16 17:59:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][610/1251] eta 0:03:17 lr 0.000046 time 0.2587 (0.3083) loss 3.4509 (3.0601) grad_norm 3.6281 (3.0961) [2021-04-16 17:59:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][620/1251] eta 0:03:14 lr 0.000046 time 0.2792 (0.3080) loss 3.4711 (3.0620) grad_norm 3.2624 (3.0960) [2021-04-16 17:59:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][630/1251] eta 0:03:10 lr 0.000046 time 0.2755 (0.3076) loss 2.9895 (3.0585) grad_norm 3.1531 (3.0924) [2021-04-16 18:00:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][640/1251] eta 0:03:07 lr 0.000046 time 0.2644 (0.3073) loss 3.6973 (3.0597) grad_norm 2.7776 (3.0914) [2021-04-16 18:00:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][650/1251] eta 0:03:04 lr 0.000046 time 0.2807 (0.3069) loss 2.9375 (3.0614) grad_norm 3.1454 (3.0932) [2021-04-16 18:00:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][660/1251] eta 0:03:01 lr 0.000046 time 0.2478 (0.3064) loss 1.7563 (3.0590) grad_norm 2.7142 (3.0916) [2021-04-16 18:00:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][670/1251] eta 0:02:57 lr 0.000046 time 0.2843 (0.3060) loss 3.1443 (3.0542) grad_norm 2.5857 (3.0918) [2021-04-16 18:00:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][680/1251] eta 0:02:54 lr 0.000046 time 0.2661 (0.3054) loss 3.3481 (3.0549) grad_norm 2.8399 (3.0902) [2021-04-16 18:00:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][690/1251] eta 0:02:51 lr 0.000046 time 0.2715 (0.3051) loss 3.1274 (3.0556) grad_norm 2.9576 (3.0884) [2021-04-16 18:00:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][700/1251] eta 0:02:47 lr 0.000046 time 0.4100 (0.3049) loss 3.5185 (3.0550) grad_norm 3.1570 (3.0883) [2021-04-16 18:00:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][710/1251] eta 0:02:44 lr 0.000046 time 0.2588 (0.3045) loss 3.7471 (3.0568) grad_norm 3.2911 (3.0864) [2021-04-16 18:00:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][720/1251] eta 0:02:41 lr 0.000046 time 0.2692 (0.3040) loss 1.9302 (3.0524) grad_norm 2.8398 (3.0859) [2021-04-16 18:00:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][730/1251] eta 0:02:38 lr 0.000046 time 0.2709 (0.3037) loss 3.3710 (3.0497) grad_norm 2.9533 (3.0856) [2021-04-16 18:00:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][740/1251] eta 0:02:34 lr 0.000046 time 0.2859 (0.3033) loss 2.9654 (3.0468) grad_norm 2.8786 (3.0844) [2021-04-16 18:00:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][750/1251] eta 0:02:31 lr 0.000046 time 0.2663 (0.3029) loss 3.4721 (3.0477) grad_norm 2.8787 (3.0841) [2021-04-16 18:00:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][760/1251] eta 0:02:28 lr 0.000046 time 0.2821 (0.3025) loss 3.4906 (3.0492) grad_norm 2.7981 (3.0845) [2021-04-16 18:00:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][770/1251] eta 0:02:25 lr 0.000045 time 0.2657 (0.3021) loss 3.7238 (3.0482) grad_norm 2.7346 (3.0804) [2021-04-16 18:00:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][780/1251] eta 0:02:22 lr 0.000045 time 0.2603 (0.3017) loss 3.0765 (3.0481) grad_norm 3.8109 (3.0813) [2021-04-16 18:00:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][790/1251] eta 0:02:18 lr 0.000045 time 0.2665 (0.3014) loss 3.3951 (3.0493) grad_norm 3.0884 (3.0796) [2021-04-16 18:00:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][800/1251] eta 0:02:15 lr 0.000045 time 0.3027 (0.3010) loss 2.6435 (3.0483) grad_norm 2.7499 (3.0762) [2021-04-16 18:00:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][810/1251] eta 0:02:12 lr 0.000045 time 0.2806 (0.3008) loss 3.3734 (3.0485) grad_norm 2.9425 (3.0763) [2021-04-16 18:00:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][820/1251] eta 0:02:09 lr 0.000045 time 0.2821 (0.3004) loss 3.0085 (3.0483) grad_norm 2.7382 (3.0855) [2021-04-16 18:00:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][830/1251] eta 0:02:06 lr 0.000045 time 0.2523 (0.3001) loss 2.3168 (3.0475) grad_norm 2.3368 (3.0849) [2021-04-16 18:00:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][840/1251] eta 0:02:03 lr 0.000045 time 0.2983 (0.2999) loss 3.3202 (3.0466) grad_norm 2.9643 (3.0841) [2021-04-16 18:01:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][850/1251] eta 0:02:00 lr 0.000045 time 0.2560 (0.2997) loss 3.5972 (3.0500) grad_norm 2.6888 (3.0849) [2021-04-16 18:01:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][860/1251] eta 0:01:57 lr 0.000045 time 0.2945 (0.2995) loss 3.5456 (3.0484) grad_norm 2.6669 (3.0862) [2021-04-16 18:01:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][870/1251] eta 0:01:54 lr 0.000045 time 0.2796 (0.2993) loss 2.7720 (3.0485) grad_norm 3.3614 (3.0878) [2021-04-16 18:01:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][880/1251] eta 0:01:50 lr 0.000045 time 0.2505 (0.2990) loss 3.3823 (3.0496) grad_norm 3.2086 (3.0894) [2021-04-16 18:01:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][890/1251] eta 0:01:47 lr 0.000045 time 0.2728 (0.2989) loss 2.5051 (3.0461) grad_norm 4.6605 (3.0899) [2021-04-16 18:01:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][900/1251] eta 0:01:44 lr 0.000045 time 0.3138 (0.2986) loss 3.3550 (3.0477) grad_norm 2.8126 (3.0873) [2021-04-16 18:01:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][910/1251] eta 0:01:41 lr 0.000045 time 0.2858 (0.2985) loss 2.4512 (3.0474) grad_norm 3.0199 (3.0896) [2021-04-16 18:01:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][920/1251] eta 0:01:38 lr 0.000045 time 0.2703 (0.2984) loss 2.0439 (3.0488) grad_norm 3.2020 (3.0892) [2021-04-16 18:01:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][930/1251] eta 0:01:35 lr 0.000045 time 0.2798 (0.2983) loss 2.1858 (3.0469) grad_norm 2.8766 (3.0881) [2021-04-16 18:01:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][940/1251] eta 0:01:32 lr 0.000045 time 0.2787 (0.2982) loss 3.1987 (3.0473) grad_norm 3.0388 (3.0889) [2021-04-16 18:01:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][950/1251] eta 0:01:29 lr 0.000045 time 0.2720 (0.2981) loss 2.7155 (3.0460) grad_norm 2.6676 (3.0881) [2021-04-16 18:01:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][960/1251] eta 0:01:26 lr 0.000045 time 0.2477 (0.2978) loss 3.4295 (3.0474) grad_norm 3.5615 (3.0870) [2021-04-16 18:01:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][970/1251] eta 0:01:23 lr 0.000045 time 0.2866 (0.2976) loss 2.2523 (3.0445) grad_norm 2.7826 (3.0864) [2021-04-16 18:01:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][980/1251] eta 0:01:20 lr 0.000045 time 0.2606 (0.2973) loss 3.3830 (3.0458) grad_norm 2.9049 (3.0884) [2021-04-16 18:01:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][990/1251] eta 0:01:17 lr 0.000045 time 0.2694 (0.2971) loss 3.6278 (3.0475) grad_norm 3.1479 (3.0864) [2021-04-16 18:01:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][1000/1251] eta 0:01:14 lr 0.000045 time 0.2640 (0.2969) loss 3.5168 (3.0494) grad_norm 3.0820 (3.0860) [2021-04-16 18:01:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][1010/1251] eta 0:01:11 lr 0.000045 time 0.3039 (0.2968) loss 2.6998 (3.0505) grad_norm 3.0894 (3.0849) [2021-04-16 18:01:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][1020/1251] eta 0:01:08 lr 0.000045 time 0.2926 (0.2966) loss 3.0484 (3.0506) grad_norm 2.8077 (3.0857) [2021-04-16 18:01:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][1030/1251] eta 0:01:05 lr 0.000045 time 0.2774 (0.2963) loss 2.4972 (3.0493) grad_norm 2.6919 (3.0858) [2021-04-16 18:01:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][1040/1251] eta 0:01:02 lr 0.000045 time 0.2729 (0.2961) loss 2.1475 (3.0494) grad_norm 3.2236 (3.0866) [2021-04-16 18:01:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][1050/1251] eta 0:00:59 lr 0.000045 time 0.2708 (0.2959) loss 3.7871 (3.0509) grad_norm 3.1092 (3.0868) [2021-04-16 18:01:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][1060/1251] eta 0:00:56 lr 0.000045 time 0.2707 (0.2957) loss 2.5905 (3.0518) grad_norm 3.9193 (3.0890) [2021-04-16 18:02:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][1070/1251] eta 0:00:53 lr 0.000045 time 0.2844 (0.2957) loss 3.0758 (3.0524) grad_norm 4.2204 (3.0919) [2021-04-16 18:02:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][1080/1251] eta 0:00:50 lr 0.000045 time 0.2627 (0.2955) loss 3.2955 (3.0559) grad_norm 2.9092 (3.0911) [2021-04-16 18:02:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][1090/1251] eta 0:00:47 lr 0.000045 time 0.2716 (0.2953) loss 3.4181 (3.0572) grad_norm 3.3338 (3.0904) [2021-04-16 18:02:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][1100/1251] eta 0:00:44 lr 0.000045 time 0.3193 (0.2952) loss 3.4660 (3.0605) grad_norm 2.9289 (3.0901) [2021-04-16 18:02:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][1110/1251] eta 0:00:41 lr 0.000045 time 0.2678 (0.2950) loss 2.7899 (3.0598) grad_norm 2.8232 (3.0902) [2021-04-16 18:02:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][1120/1251] eta 0:00:38 lr 0.000045 time 0.2803 (0.2950) loss 3.3478 (3.0609) grad_norm 2.7175 (3.0897) [2021-04-16 18:02:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][1130/1251] eta 0:00:35 lr 0.000045 time 0.2696 (0.2948) loss 2.9648 (3.0611) grad_norm 2.8430 (3.0897) [2021-04-16 18:02:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][1140/1251] eta 0:00:32 lr 0.000045 time 0.2853 (0.2946) loss 2.7809 (3.0607) grad_norm 3.3553 (3.0889) [2021-04-16 18:02:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][1150/1251] eta 0:00:29 lr 0.000045 time 0.2621 (0.2946) loss 2.6478 (3.0607) grad_norm 2.8877 (3.0881) [2021-04-16 18:02:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][1160/1251] eta 0:00:26 lr 0.000045 time 0.2838 (0.2946) loss 3.2196 (3.0592) grad_norm 3.1897 (3.0872) [2021-04-16 18:02:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][1170/1251] eta 0:00:23 lr 0.000045 time 0.2840 (0.2945) loss 2.3963 (3.0566) grad_norm 3.0194 (3.0854) [2021-04-16 18:02:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][1180/1251] eta 0:00:20 lr 0.000045 time 0.2722 (0.2944) loss 3.8964 (3.0563) grad_norm 2.7289 (3.0867) [2021-04-16 18:02:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][1190/1251] eta 0:00:17 lr 0.000045 time 0.2732 (0.2943) loss 2.3659 (3.0532) grad_norm 3.1144 (3.0885) [2021-04-16 18:02:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][1200/1251] eta 0:00:15 lr 0.000045 time 0.2676 (0.2941) loss 3.7272 (3.0527) grad_norm 3.0824 (3.0895) [2021-04-16 18:02:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][1210/1251] eta 0:00:12 lr 0.000045 time 0.2603 (0.2940) loss 3.6472 (3.0536) grad_norm 2.7785 (3.0871) [2021-04-16 18:02:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][1220/1251] eta 0:00:09 lr 0.000045 time 0.2822 (0.2940) loss 3.2567 (3.0530) grad_norm 3.3123 (3.0876) [2021-04-16 18:02:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][1230/1251] eta 0:00:06 lr 0.000045 time 0.2602 (0.2940) loss 2.5528 (3.0538) grad_norm 2.6834 (3.0871) [2021-04-16 18:02:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][1240/1251] eta 0:00:03 lr 0.000045 time 0.2479 (0.2937) loss 2.3835 (3.0539) grad_norm 2.8404 (3.0872) [2021-04-16 18:02:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [263/300][1250/1251] eta 0:00:00 lr 0.000045 time 0.2485 (0.2933) loss 3.4940 (3.0535) grad_norm 3.1013 (3.0869) [2021-04-16 18:03:01 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 263 training takes 0:06:16 [2021-04-16 18:03:01 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_263.pth saving...... [2021-04-16 18:03:13 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_263.pth saved !!! [2021-04-16 18:03:15 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.186 (1.186) Loss 0.8368 (0.8368) Acc@1 80.176 (80.176) Acc@5 96.094 (96.094) [2021-04-16 18:03:16 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.138 (0.239) Loss 0.8173 (0.8202) Acc@1 81.641 (80.762) Acc@5 96.094 (95.632) [2021-04-16 18:03:18 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.136 (0.222) Loss 0.8050 (0.8263) Acc@1 81.641 (80.994) Acc@5 94.727 (95.392) [2021-04-16 18:03:21 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.462 (0.244) Loss 0.7829 (0.8226) Acc@1 81.543 (81.020) Acc@5 96.484 (95.391) [2021-04-16 18:03:22 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.092 (0.219) Loss 0.7608 (0.8241) Acc@1 82.422 (80.890) Acc@5 96.387 (95.441) [2021-04-16 18:03:32 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.852 Acc@5 95.400 [2021-04-16 18:03:32 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 80.9% [2021-04-16 18:03:32 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 80.85% [2021-04-16 18:03:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][0/1251] eta 2:39:13 lr 0.000045 time 7.6365 (7.6365) loss 2.6097 (2.6097) grad_norm 3.4016 (3.4016) [2021-04-16 18:03:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][10/1251] eta 0:19:30 lr 0.000045 time 0.3176 (0.9432) loss 2.8304 (2.8940) grad_norm 2.9241 (3.0213) [2021-04-16 18:03:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][20/1251] eta 0:12:49 lr 0.000045 time 0.3063 (0.6253) loss 3.0001 (2.8739) grad_norm 3.2359 (3.0415) [2021-04-16 18:03:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][30/1251] eta 0:10:29 lr 0.000045 time 0.2847 (0.5157) loss 2.7733 (2.9017) grad_norm 2.9488 (3.0503) [2021-04-16 18:03:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][40/1251] eta 0:09:14 lr 0.000045 time 0.2815 (0.4580) loss 2.8402 (2.9042) grad_norm 3.0515 (3.0358) [2021-04-16 18:03:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][50/1251] eta 0:08:27 lr 0.000045 time 0.2596 (0.4227) loss 3.4111 (2.8863) grad_norm 2.9774 (2.9881) [2021-04-16 18:03:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][60/1251] eta 0:07:56 lr 0.000045 time 0.2622 (0.3998) loss 2.1794 (2.8929) grad_norm 3.4426 (2.9756) [2021-04-16 18:03:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][70/1251] eta 0:07:31 lr 0.000045 time 0.2732 (0.3822) loss 2.7959 (2.8775) grad_norm 2.8679 (2.9598) [2021-04-16 18:04:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][80/1251] eta 0:07:13 lr 0.000045 time 0.2846 (0.3698) loss 2.4233 (2.8878) grad_norm 3.3182 (2.9640) [2021-04-16 18:04:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][90/1251] eta 0:06:58 lr 0.000045 time 0.2612 (0.3603) loss 2.7447 (2.8951) grad_norm 3.2125 (2.9673) [2021-04-16 18:04:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][100/1251] eta 0:06:45 lr 0.000045 time 0.2584 (0.3520) loss 3.4233 (2.8980) grad_norm 3.0519 (2.9911) [2021-04-16 18:04:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][110/1251] eta 0:06:33 lr 0.000045 time 0.2496 (0.3452) loss 3.0095 (2.8810) grad_norm 2.5362 (2.9976) [2021-04-16 18:04:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][120/1251] eta 0:06:25 lr 0.000045 time 0.2843 (0.3409) loss 3.6235 (2.8962) grad_norm 2.6949 (3.0007) [2021-04-16 18:04:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][130/1251] eta 0:06:17 lr 0.000045 time 0.2928 (0.3367) loss 2.3804 (2.8868) grad_norm 3.3520 (3.0067) [2021-04-16 18:04:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][140/1251] eta 0:06:11 lr 0.000045 time 0.2822 (0.3341) loss 2.7057 (2.9024) grad_norm 3.3977 (3.0100) [2021-04-16 18:04:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][150/1251] eta 0:06:04 lr 0.000045 time 0.2921 (0.3311) loss 3.6033 (2.9028) grad_norm 3.1405 (3.0118) [2021-04-16 18:04:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][160/1251] eta 0:05:57 lr 0.000045 time 0.2762 (0.3275) loss 3.4795 (2.9087) grad_norm 3.0294 (3.0198) [2021-04-16 18:04:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][170/1251] eta 0:05:50 lr 0.000045 time 0.3091 (0.3247) loss 3.2483 (2.9036) grad_norm 2.4963 (3.0177) [2021-04-16 18:04:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][180/1251] eta 0:05:44 lr 0.000044 time 0.2737 (0.3220) loss 3.2189 (2.9141) grad_norm 2.3529 (3.0194) [2021-04-16 18:04:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][190/1251] eta 0:05:39 lr 0.000044 time 0.2734 (0.3199) loss 3.2538 (2.9304) grad_norm 3.0451 (3.0229) [2021-04-16 18:04:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][200/1251] eta 0:05:34 lr 0.000044 time 0.2614 (0.3179) loss 3.6112 (2.9447) grad_norm 2.9148 (3.0285) [2021-04-16 18:04:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][210/1251] eta 0:05:29 lr 0.000044 time 0.2738 (0.3161) loss 3.3400 (2.9535) grad_norm 2.6253 (3.0281) [2021-04-16 18:04:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][220/1251] eta 0:05:24 lr 0.000044 time 0.2798 (0.3144) loss 2.4702 (2.9630) grad_norm 2.9498 (3.0285) [2021-04-16 18:04:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][230/1251] eta 0:05:19 lr 0.000044 time 0.2731 (0.3128) loss 3.0783 (2.9670) grad_norm 2.7093 (3.0285) [2021-04-16 18:04:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][240/1251] eta 0:05:14 lr 0.000044 time 0.2952 (0.3113) loss 3.6159 (2.9825) grad_norm 3.0448 (3.0272) [2021-04-16 18:04:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][250/1251] eta 0:05:10 lr 0.000044 time 0.2848 (0.3100) loss 3.3092 (2.9883) grad_norm 3.6996 (3.0247) [2021-04-16 18:04:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][260/1251] eta 0:05:05 lr 0.000044 time 0.2702 (0.3086) loss 2.5123 (2.9835) grad_norm 3.2812 (3.0280) [2021-04-16 18:04:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][270/1251] eta 0:05:01 lr 0.000044 time 0.2702 (0.3076) loss 3.0790 (2.9785) grad_norm 3.1624 (3.0266) [2021-04-16 18:04:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][280/1251] eta 0:04:57 lr 0.000044 time 0.2429 (0.3068) loss 2.8163 (2.9840) grad_norm 3.2654 (3.0254) [2021-04-16 18:05:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][290/1251] eta 0:04:54 lr 0.000044 time 0.2706 (0.3065) loss 2.9281 (2.9920) grad_norm 2.9820 (3.0336) [2021-04-16 18:05:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][300/1251] eta 0:04:50 lr 0.000044 time 0.2756 (0.3055) loss 3.3672 (2.9840) grad_norm 2.7104 (3.0314) [2021-04-16 18:05:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][310/1251] eta 0:04:46 lr 0.000044 time 0.3008 (0.3049) loss 3.0023 (2.9818) grad_norm 2.9965 (3.0442) [2021-04-16 18:05:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][320/1251] eta 0:04:43 lr 0.000044 time 0.3001 (0.3044) loss 2.1701 (2.9801) grad_norm 3.3507 (3.0438) [2021-04-16 18:05:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][330/1251] eta 0:04:39 lr 0.000044 time 0.2787 (0.3036) loss 3.2565 (2.9844) grad_norm 2.9469 (3.0547) [2021-04-16 18:05:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][340/1251] eta 0:04:35 lr 0.000044 time 0.2890 (0.3028) loss 2.7963 (2.9834) grad_norm 2.8888 (3.0538) [2021-04-16 18:05:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][350/1251] eta 0:04:32 lr 0.000044 time 0.2747 (0.3020) loss 3.0531 (2.9864) grad_norm 3.1271 (3.0510) [2021-04-16 18:05:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][360/1251] eta 0:04:29 lr 0.000044 time 0.2952 (0.3024) loss 2.2000 (2.9851) grad_norm 3.3639 (3.0471) [2021-04-16 18:05:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][370/1251] eta 0:04:25 lr 0.000044 time 0.2822 (0.3017) loss 2.3669 (2.9787) grad_norm 2.3347 (3.0415) [2021-04-16 18:05:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][380/1251] eta 0:04:22 lr 0.000044 time 0.2813 (0.3012) loss 1.9818 (2.9783) grad_norm 2.4236 (3.0438) [2021-04-16 18:05:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][390/1251] eta 0:04:19 lr 0.000044 time 0.2861 (0.3011) loss 3.2679 (2.9800) grad_norm 2.9907 (3.0439) [2021-04-16 18:05:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][400/1251] eta 0:04:15 lr 0.000044 time 0.2826 (0.3006) loss 2.4144 (2.9813) grad_norm 2.7956 (3.0451) [2021-04-16 18:05:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][410/1251] eta 0:04:12 lr 0.000044 time 0.2938 (0.3000) loss 3.2564 (2.9838) grad_norm 2.6829 (3.0440) [2021-04-16 18:05:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][420/1251] eta 0:04:09 lr 0.000044 time 0.2514 (0.2998) loss 2.5611 (2.9819) grad_norm 3.6101 (3.0448) [2021-04-16 18:05:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][430/1251] eta 0:04:05 lr 0.000044 time 0.2586 (0.2996) loss 3.4162 (2.9852) grad_norm 3.4192 (3.0456) [2021-04-16 18:05:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][440/1251] eta 0:04:02 lr 0.000044 time 0.2933 (0.2993) loss 2.7041 (2.9769) grad_norm 3.2065 (3.0430) [2021-04-16 18:05:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][450/1251] eta 0:03:59 lr 0.000044 time 0.2984 (0.2990) loss 2.9822 (2.9818) grad_norm 3.3783 (3.0389) [2021-04-16 18:05:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][460/1251] eta 0:03:56 lr 0.000044 time 0.2887 (0.2987) loss 1.8030 (2.9840) grad_norm 2.5001 (3.0366) [2021-04-16 18:05:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][470/1251] eta 0:03:52 lr 0.000044 time 0.2939 (0.2983) loss 3.4765 (2.9882) grad_norm 3.2028 (3.0361) [2021-04-16 18:05:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][480/1251] eta 0:03:49 lr 0.000044 time 0.2820 (0.2979) loss 3.0559 (2.9900) grad_norm 3.1175 (3.0336) [2021-04-16 18:05:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][490/1251] eta 0:03:46 lr 0.000044 time 0.2594 (0.2974) loss 3.5202 (2.9890) grad_norm 3.3053 (3.0298) [2021-04-16 18:06:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][500/1251] eta 0:03:43 lr 0.000044 time 0.2970 (0.2971) loss 2.1478 (2.9869) grad_norm 3.1519 (3.0335) [2021-04-16 18:06:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][510/1251] eta 0:03:39 lr 0.000044 time 0.2549 (0.2967) loss 3.7312 (2.9900) grad_norm 3.4729 (3.0341) [2021-04-16 18:06:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][520/1251] eta 0:03:36 lr 0.000044 time 0.2837 (0.2964) loss 3.5076 (2.9952) grad_norm 2.7799 (3.0316) [2021-04-16 18:06:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][530/1251] eta 0:03:33 lr 0.000044 time 0.2610 (0.2960) loss 3.0990 (2.9972) grad_norm 2.9411 (3.0284) [2021-04-16 18:06:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][540/1251] eta 0:03:30 lr 0.000044 time 0.2634 (0.2958) loss 3.2146 (3.0013) grad_norm 3.0988 (3.0247) [2021-04-16 18:06:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][550/1251] eta 0:03:27 lr 0.000044 time 0.2666 (0.2955) loss 3.1820 (2.9987) grad_norm 2.8346 (3.0235) [2021-04-16 18:06:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][560/1251] eta 0:03:23 lr 0.000044 time 0.2856 (0.2952) loss 3.3149 (3.0013) grad_norm 3.0290 (3.0244) [2021-04-16 18:06:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][570/1251] eta 0:03:21 lr 0.000044 time 0.4515 (0.2953) loss 3.7682 (2.9981) grad_norm 3.0311 (3.0256) [2021-04-16 18:06:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][580/1251] eta 0:03:18 lr 0.000044 time 0.2665 (0.2951) loss 3.5140 (2.9997) grad_norm 2.8339 (3.0270) [2021-04-16 18:06:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][590/1251] eta 0:03:14 lr 0.000044 time 0.2850 (0.2948) loss 2.8705 (3.0032) grad_norm 3.0536 (3.0275) [2021-04-16 18:06:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][600/1251] eta 0:03:11 lr 0.000044 time 0.2690 (0.2945) loss 3.3149 (3.0063) grad_norm 2.9017 (3.0270) [2021-04-16 18:06:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][610/1251] eta 0:03:08 lr 0.000044 time 0.2683 (0.2942) loss 3.3264 (3.0036) grad_norm 3.0433 (3.0279) [2021-04-16 18:06:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][620/1251] eta 0:03:05 lr 0.000044 time 0.2974 (0.2939) loss 3.0188 (3.0041) grad_norm 2.7674 (3.0271) [2021-04-16 18:06:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][630/1251] eta 0:03:02 lr 0.000044 time 0.2995 (0.2937) loss 3.1992 (3.0040) grad_norm 3.4094 (3.0442) [2021-04-16 18:06:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][640/1251] eta 0:02:59 lr 0.000044 time 0.2917 (0.2935) loss 2.2090 (3.0007) grad_norm 3.0641 (3.0485) [2021-04-16 18:06:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][650/1251] eta 0:02:56 lr 0.000044 time 0.3084 (0.2933) loss 3.3120 (2.9978) grad_norm 3.3638 (3.0488) [2021-04-16 18:06:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][660/1251] eta 0:02:53 lr 0.000044 time 0.2625 (0.2931) loss 3.3703 (2.9989) grad_norm 2.5271 (3.0489) [2021-04-16 18:06:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][670/1251] eta 0:02:50 lr 0.000044 time 0.2819 (0.2931) loss 3.2440 (3.0000) grad_norm 3.0911 (3.0520) [2021-04-16 18:06:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][680/1251] eta 0:02:47 lr 0.000044 time 0.2565 (0.2929) loss 3.4959 (3.0016) grad_norm 3.0251 (3.0569) [2021-04-16 18:06:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][690/1251] eta 0:02:44 lr 0.000044 time 0.2790 (0.2929) loss 3.8474 (3.0039) grad_norm 3.8591 (3.0562) [2021-04-16 18:06:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][700/1251] eta 0:02:41 lr 0.000044 time 0.2766 (0.2927) loss 3.3208 (3.0065) grad_norm 3.2993 (3.0620) [2021-04-16 18:07:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][710/1251] eta 0:02:38 lr 0.000044 time 0.2964 (0.2925) loss 2.9584 (3.0083) grad_norm 3.0076 (3.0644) [2021-04-16 18:07:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][720/1251] eta 0:02:35 lr 0.000044 time 0.2903 (0.2926) loss 2.0095 (3.0083) grad_norm 3.5585 (3.0642) [2021-04-16 18:07:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][730/1251] eta 0:02:32 lr 0.000044 time 0.2883 (0.2925) loss 3.4328 (3.0109) grad_norm 3.4669 (3.0646) [2021-04-16 18:07:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][740/1251] eta 0:02:29 lr 0.000044 time 0.2654 (0.2923) loss 2.2235 (3.0104) grad_norm 2.9562 (3.0673) [2021-04-16 18:07:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][750/1251] eta 0:02:26 lr 0.000044 time 0.2802 (0.2922) loss 3.2000 (3.0100) grad_norm 2.8619 (3.0692) [2021-04-16 18:07:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][760/1251] eta 0:02:23 lr 0.000044 time 0.2687 (0.2920) loss 3.1156 (3.0108) grad_norm 2.7174 (3.0705) [2021-04-16 18:07:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][770/1251] eta 0:02:20 lr 0.000044 time 0.2950 (0.2919) loss 3.1112 (3.0141) grad_norm 2.5077 (3.0689) [2021-04-16 18:07:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][780/1251] eta 0:02:17 lr 0.000044 time 0.2739 (0.2917) loss 2.4605 (3.0119) grad_norm 2.8844 (3.0678) [2021-04-16 18:07:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][790/1251] eta 0:02:14 lr 0.000044 time 0.2926 (0.2916) loss 2.8778 (3.0144) grad_norm 3.4113 (3.0663) [2021-04-16 18:07:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][800/1251] eta 0:02:11 lr 0.000044 time 0.2751 (0.2915) loss 1.8664 (3.0133) grad_norm 2.8199 (3.0653) [2021-04-16 18:07:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][810/1251] eta 0:02:08 lr 0.000044 time 0.2970 (0.2913) loss 2.5883 (3.0165) grad_norm 2.8565 (3.0658) [2021-04-16 18:07:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][820/1251] eta 0:02:05 lr 0.000044 time 0.3022 (0.2913) loss 3.2264 (3.0176) grad_norm 2.4709 (3.0637) [2021-04-16 18:07:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][830/1251] eta 0:02:02 lr 0.000044 time 0.2645 (0.2911) loss 3.1452 (3.0165) grad_norm 2.6452 (3.0647) [2021-04-16 18:07:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][840/1251] eta 0:01:59 lr 0.000043 time 0.2628 (0.2909) loss 2.6624 (3.0180) grad_norm 2.8651 (3.0654) [2021-04-16 18:07:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][850/1251] eta 0:01:56 lr 0.000043 time 0.2769 (0.2909) loss 2.5363 (3.0189) grad_norm 2.9051 (3.0649) [2021-04-16 18:07:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][860/1251] eta 0:01:53 lr 0.000043 time 0.2879 (0.2906) loss 1.8347 (3.0184) grad_norm 2.6531 (3.0624) [2021-04-16 18:07:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][870/1251] eta 0:01:50 lr 0.000043 time 0.2852 (0.2906) loss 3.5185 (3.0195) grad_norm 3.0699 (3.0608) [2021-04-16 18:07:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][880/1251] eta 0:01:47 lr 0.000043 time 0.3025 (0.2905) loss 3.6998 (3.0214) grad_norm 3.2960 (3.0603) [2021-04-16 18:07:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][890/1251] eta 0:01:44 lr 0.000043 time 0.3130 (0.2906) loss 2.7252 (3.0208) grad_norm 2.6005 (3.0589) [2021-04-16 18:07:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][900/1251] eta 0:01:41 lr 0.000043 time 0.2502 (0.2905) loss 3.0439 (3.0232) grad_norm 3.5639 (3.0601) [2021-04-16 18:07:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][910/1251] eta 0:01:39 lr 0.000043 time 0.2982 (0.2904) loss 1.8429 (3.0230) grad_norm 3.9636 (3.0629) [2021-04-16 18:07:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][920/1251] eta 0:01:36 lr 0.000043 time 0.2703 (0.2902) loss 3.2212 (3.0240) grad_norm 2.9236 (3.0635) [2021-04-16 18:08:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][930/1251] eta 0:01:33 lr 0.000043 time 0.2859 (0.2903) loss 3.2808 (3.0257) grad_norm 3.4857 (3.0620) [2021-04-16 18:08:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][940/1251] eta 0:01:30 lr 0.000043 time 0.2601 (0.2905) loss 2.2631 (3.0246) grad_norm 2.8149 (3.0613) [2021-04-16 18:08:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][950/1251] eta 0:01:27 lr 0.000043 time 0.2687 (0.2904) loss 2.8740 (3.0247) grad_norm 3.8880 (3.0724) [2021-04-16 18:08:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][960/1251] eta 0:01:24 lr 0.000043 time 0.2484 (0.2903) loss 3.6012 (3.0255) grad_norm 2.9493 (3.0731) [2021-04-16 18:08:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][970/1251] eta 0:01:21 lr 0.000043 time 0.2662 (0.2902) loss 2.3836 (3.0221) grad_norm 2.5063 (3.0723) [2021-04-16 18:08:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][980/1251] eta 0:01:18 lr 0.000043 time 0.2657 (0.2901) loss 2.5821 (3.0189) grad_norm 3.1280 (3.0706) [2021-04-16 18:08:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][990/1251] eta 0:01:15 lr 0.000043 time 0.2990 (0.2901) loss 2.4261 (3.0188) grad_norm 3.0562 (3.0715) [2021-04-16 18:08:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][1000/1251] eta 0:01:12 lr 0.000043 time 0.2675 (0.2900) loss 2.4028 (3.0175) grad_norm 2.7871 (3.0701) [2021-04-16 18:08:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][1010/1251] eta 0:01:09 lr 0.000043 time 0.2654 (0.2900) loss 2.7503 (3.0178) grad_norm 2.7723 (3.0684) [2021-04-16 18:08:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][1020/1251] eta 0:01:06 lr 0.000043 time 0.2632 (0.2899) loss 3.5280 (3.0205) grad_norm 3.3006 (3.0702) [2021-04-16 18:08:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][1030/1251] eta 0:01:04 lr 0.000043 time 0.2772 (0.2898) loss 2.7886 (3.0202) grad_norm 2.9250 (3.0697) [2021-04-16 18:08:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][1040/1251] eta 0:01:01 lr 0.000043 time 0.2764 (0.2897) loss 3.4296 (3.0200) grad_norm 3.8150 (3.0712) [2021-04-16 18:08:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][1050/1251] eta 0:00:58 lr 0.000043 time 0.2876 (0.2896) loss 2.9960 (3.0201) grad_norm 2.9669 (3.0725) [2021-04-16 18:08:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][1060/1251] eta 0:00:55 lr 0.000043 time 0.2885 (0.2894) loss 3.1941 (3.0209) grad_norm 3.0325 (3.0724) [2021-04-16 18:08:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][1070/1251] eta 0:00:52 lr 0.000043 time 0.2991 (0.2893) loss 2.9163 (3.0211) grad_norm 2.9611 (3.0739) [2021-04-16 18:08:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][1080/1251] eta 0:00:49 lr 0.000043 time 0.2909 (0.2892) loss 3.1832 (3.0240) grad_norm 2.9537 (3.0767) [2021-04-16 18:08:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][1090/1251] eta 0:00:46 lr 0.000043 time 0.2621 (0.2891) loss 3.3397 (3.0211) grad_norm 3.2990 (3.0759) [2021-04-16 18:08:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][1100/1251] eta 0:00:43 lr 0.000043 time 0.2608 (0.2890) loss 3.5503 (3.0206) grad_norm 3.3483 (3.0765) [2021-04-16 18:08:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][1110/1251] eta 0:00:40 lr 0.000043 time 0.2547 (0.2889) loss 2.5704 (3.0213) grad_norm 2.9874 (3.0745) [2021-04-16 18:08:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][1120/1251] eta 0:00:37 lr 0.000043 time 0.2693 (0.2888) loss 3.3446 (3.0228) grad_norm 2.6738 (3.0736) [2021-04-16 18:08:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][1130/1251] eta 0:00:34 lr 0.000043 time 0.2535 (0.2887) loss 2.6615 (3.0224) grad_norm 3.0927 (3.0722) [2021-04-16 18:09:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][1140/1251] eta 0:00:32 lr 0.000043 time 0.2975 (0.2887) loss 3.4687 (3.0239) grad_norm 2.8640 (3.0712) [2021-04-16 18:09:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][1150/1251] eta 0:00:29 lr 0.000043 time 0.2896 (0.2887) loss 1.7937 (3.0224) grad_norm 3.1011 (3.0699) [2021-04-16 18:09:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][1160/1251] eta 0:00:26 lr 0.000043 time 0.2777 (0.2885) loss 2.8906 (3.0213) grad_norm 2.7893 (3.0684) [2021-04-16 18:09:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][1170/1251] eta 0:00:23 lr 0.000043 time 0.2880 (0.2885) loss 3.2040 (3.0215) grad_norm 2.8766 (3.0706) [2021-04-16 18:09:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][1180/1251] eta 0:00:20 lr 0.000043 time 0.2664 (0.2885) loss 3.4490 (3.0199) grad_norm 3.7522 (3.0708) [2021-04-16 18:09:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][1190/1251] eta 0:00:17 lr 0.000043 time 0.2884 (0.2883) loss 3.1257 (3.0204) grad_norm 2.7678 (3.0695) [2021-04-16 18:09:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][1200/1251] eta 0:00:14 lr 0.000043 time 0.2939 (0.2883) loss 1.7725 (3.0187) grad_norm 3.0617 (3.0681) [2021-04-16 18:09:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][1210/1251] eta 0:00:11 lr 0.000043 time 0.2742 (0.2881) loss 3.0705 (3.0197) grad_norm 2.9611 (3.0674) [2021-04-16 18:09:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][1220/1251] eta 0:00:08 lr 0.000043 time 0.2682 (0.2880) loss 3.2302 (3.0209) grad_norm 2.9387 (3.0679) [2021-04-16 18:09:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][1230/1251] eta 0:00:06 lr 0.000043 time 0.2726 (0.2879) loss 3.1295 (3.0195) grad_norm 3.2563 (3.0669) [2021-04-16 18:09:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][1240/1251] eta 0:00:03 lr 0.000043 time 0.3232 (0.2878) loss 2.8074 (3.0167) grad_norm 3.5643 (3.0685) [2021-04-16 18:09:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [264/300][1250/1251] eta 0:00:00 lr 0.000043 time 0.2484 (0.2875) loss 3.6663 (3.0169) grad_norm 3.7336 (3.0686) [2021-04-16 18:09:37 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 264 training takes 0:06:04 [2021-04-16 18:09:37 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_264.pth saving...... [2021-04-16 18:09:48 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_264.pth saved !!! [2021-04-16 18:09:49 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.250 (1.250) Loss 0.8585 (0.8585) Acc@1 81.348 (81.348) Acc@5 94.922 (94.922) [2021-04-16 18:09:51 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.142 (0.272) Loss 0.8328 (0.8294) Acc@1 81.348 (80.788) Acc@5 95.215 (95.472) [2021-04-16 18:09:53 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.104 (0.249) Loss 0.7729 (0.8281) Acc@1 82.129 (81.045) Acc@5 96.289 (95.322) [2021-04-16 18:09:55 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.125 (0.233) Loss 0.8413 (0.8344) Acc@1 79.590 (80.815) Acc@5 95.312 (95.209) [2021-04-16 18:09:57 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.164 (0.225) Loss 0.8121 (0.8297) Acc@1 81.641 (80.890) Acc@5 95.605 (95.312) [2021-04-16 18:10:10 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.834 Acc@5 95.320 [2021-04-16 18:10:10 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 80.8% [2021-04-16 18:10:10 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 80.85% [2021-04-16 18:10:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][0/1251] eta 3:57:52 lr 0.000043 time 11.4088 (11.4088) loss 2.4087 (2.4087) grad_norm 3.0999 (3.0999) [2021-04-16 18:10:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][10/1251] eta 0:26:44 lr 0.000043 time 0.2949 (1.2928) loss 3.6596 (2.8163) grad_norm 2.6256 (3.4244) [2021-04-16 18:10:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][20/1251] eta 0:16:39 lr 0.000043 time 0.2861 (0.8118) loss 2.1934 (2.8041) grad_norm 3.0342 (3.3444) [2021-04-16 18:10:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][30/1251] eta 0:13:00 lr 0.000043 time 0.2661 (0.6390) loss 2.4847 (2.9004) grad_norm 2.7962 (3.2215) [2021-04-16 18:10:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][40/1251] eta 0:11:06 lr 0.000043 time 0.2690 (0.5502) loss 3.0782 (2.9268) grad_norm 3.1238 (3.2167) [2021-04-16 18:10:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][50/1251] eta 0:09:55 lr 0.000043 time 0.2752 (0.4962) loss 3.0620 (2.9403) grad_norm 3.2461 (3.1899) [2021-04-16 18:10:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][60/1251] eta 0:09:10 lr 0.000043 time 0.3856 (0.4624) loss 3.5706 (2.9177) grad_norm 3.3928 (3.1619) [2021-04-16 18:10:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][70/1251] eta 0:08:33 lr 0.000043 time 0.2616 (0.4352) loss 2.6012 (2.9392) grad_norm 2.6619 (3.1481) [2021-04-16 18:10:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][80/1251] eta 0:08:06 lr 0.000043 time 0.2761 (0.4159) loss 3.2338 (2.9759) grad_norm 2.9601 (3.1474) [2021-04-16 18:10:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][90/1251] eta 0:07:46 lr 0.000043 time 0.2781 (0.4015) loss 2.6916 (2.9928) grad_norm 2.9790 (3.1348) [2021-04-16 18:10:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][100/1251] eta 0:07:28 lr 0.000043 time 0.2700 (0.3900) loss 3.4382 (2.9703) grad_norm 3.3232 (3.1209) [2021-04-16 18:10:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][110/1251] eta 0:07:13 lr 0.000043 time 0.2838 (0.3797) loss 2.8547 (2.9990) grad_norm 2.3830 (3.1064) [2021-04-16 18:10:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][120/1251] eta 0:06:59 lr 0.000043 time 0.2764 (0.3710) loss 2.2514 (3.0048) grad_norm 2.7049 (3.1077) [2021-04-16 18:10:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][130/1251] eta 0:06:48 lr 0.000043 time 0.2749 (0.3642) loss 2.9310 (2.9980) grad_norm 3.2629 (3.1022) [2021-04-16 18:11:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][140/1251] eta 0:06:39 lr 0.000043 time 0.2603 (0.3592) loss 2.3593 (2.9906) grad_norm 3.0536 (3.0958) [2021-04-16 18:11:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][150/1251] eta 0:06:30 lr 0.000043 time 0.2911 (0.3546) loss 3.2075 (2.9984) grad_norm 3.0055 (3.0910) [2021-04-16 18:11:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][160/1251] eta 0:06:22 lr 0.000043 time 0.2873 (0.3504) loss 3.3583 (3.0061) grad_norm 4.1852 (3.1016) [2021-04-16 18:11:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][170/1251] eta 0:06:14 lr 0.000043 time 0.2654 (0.3460) loss 3.6533 (3.0113) grad_norm 2.9547 (3.1161) [2021-04-16 18:11:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][180/1251] eta 0:06:07 lr 0.000043 time 0.2711 (0.3430) loss 2.3442 (3.0199) grad_norm 2.6799 (3.1233) [2021-04-16 18:11:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][190/1251] eta 0:06:00 lr 0.000043 time 0.2625 (0.3400) loss 2.8312 (3.0199) grad_norm 2.9254 (3.1418) [2021-04-16 18:11:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][200/1251] eta 0:05:54 lr 0.000043 time 0.2743 (0.3378) loss 3.0338 (3.0204) grad_norm 2.9810 (3.1334) [2021-04-16 18:11:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][210/1251] eta 0:05:48 lr 0.000043 time 0.2705 (0.3348) loss 3.0362 (3.0312) grad_norm 2.9988 (3.1194) [2021-04-16 18:11:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][220/1251] eta 0:05:43 lr 0.000043 time 0.2897 (0.3327) loss 3.3116 (3.0345) grad_norm 2.6934 (3.1058) [2021-04-16 18:11:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][230/1251] eta 0:05:37 lr 0.000043 time 0.2686 (0.3303) loss 3.2960 (3.0299) grad_norm 2.6792 (3.1032) [2021-04-16 18:11:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][240/1251] eta 0:05:31 lr 0.000043 time 0.2606 (0.3280) loss 3.5638 (3.0276) grad_norm 2.7569 (3.1021) [2021-04-16 18:11:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][250/1251] eta 0:05:26 lr 0.000043 time 0.2764 (0.3262) loss 3.3091 (3.0247) grad_norm 2.8172 (3.1046) [2021-04-16 18:11:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][260/1251] eta 0:05:21 lr 0.000042 time 0.2742 (0.3244) loss 2.8026 (3.0237) grad_norm 3.4023 (3.1058) [2021-04-16 18:11:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][270/1251] eta 0:05:16 lr 0.000042 time 0.2904 (0.3229) loss 3.3078 (3.0257) grad_norm 2.9013 (3.1024) [2021-04-16 18:11:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][280/1251] eta 0:05:11 lr 0.000042 time 0.2799 (0.3212) loss 3.0011 (3.0246) grad_norm 2.9440 (3.0994) [2021-04-16 18:11:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][290/1251] eta 0:05:07 lr 0.000042 time 0.2597 (0.3195) loss 2.7544 (3.0247) grad_norm 3.0701 (3.0954) [2021-04-16 18:11:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][300/1251] eta 0:05:02 lr 0.000042 time 0.2643 (0.3181) loss 3.5514 (3.0199) grad_norm 3.2823 (3.1007) [2021-04-16 18:11:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][310/1251] eta 0:04:58 lr 0.000042 time 0.2921 (0.3167) loss 2.5905 (3.0197) grad_norm 3.2432 (3.0988) [2021-04-16 18:11:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][320/1251] eta 0:04:53 lr 0.000042 time 0.2792 (0.3154) loss 2.8430 (3.0180) grad_norm 3.3267 (3.1002) [2021-04-16 18:11:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][330/1251] eta 0:04:49 lr 0.000042 time 0.2918 (0.3143) loss 3.0354 (3.0167) grad_norm 2.7440 (3.1018) [2021-04-16 18:11:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][340/1251] eta 0:04:45 lr 0.000042 time 0.3039 (0.3133) loss 2.3164 (3.0171) grad_norm 3.2968 (3.1045) [2021-04-16 18:11:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][350/1251] eta 0:04:41 lr 0.000042 time 0.2606 (0.3123) loss 3.0714 (3.0156) grad_norm 3.0635 (3.1102) [2021-04-16 18:12:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][360/1251] eta 0:04:37 lr 0.000042 time 0.2460 (0.3114) loss 3.5141 (3.0204) grad_norm 3.0012 (3.1125) [2021-04-16 18:12:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][370/1251] eta 0:04:33 lr 0.000042 time 0.2808 (0.3109) loss 2.1746 (3.0166) grad_norm 3.1171 (3.1089) [2021-04-16 18:12:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][380/1251] eta 0:04:29 lr 0.000042 time 0.2723 (0.3098) loss 3.7404 (3.0221) grad_norm 2.6901 (3.1058) [2021-04-16 18:12:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][390/1251] eta 0:04:26 lr 0.000042 time 0.2419 (0.3090) loss 3.5127 (3.0207) grad_norm 2.3436 (3.1009) [2021-04-16 18:12:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][400/1251] eta 0:04:22 lr 0.000042 time 0.2873 (0.3081) loss 3.3424 (3.0220) grad_norm 2.7407 (3.0993) [2021-04-16 18:12:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][410/1251] eta 0:04:18 lr 0.000042 time 0.2566 (0.3073) loss 2.0400 (3.0234) grad_norm 2.7803 (3.1000) [2021-04-16 18:12:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][420/1251] eta 0:04:14 lr 0.000042 time 0.2854 (0.3066) loss 3.0303 (3.0246) grad_norm 3.0996 (3.0985) [2021-04-16 18:12:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][430/1251] eta 0:04:11 lr 0.000042 time 0.2645 (0.3059) loss 2.3391 (3.0253) grad_norm 2.9311 (3.0951) [2021-04-16 18:12:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][440/1251] eta 0:04:07 lr 0.000042 time 0.2803 (0.3052) loss 3.4494 (3.0225) grad_norm 3.5870 (3.0977) [2021-04-16 18:12:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][450/1251] eta 0:04:04 lr 0.000042 time 0.3911 (0.3050) loss 3.2880 (3.0266) grad_norm 3.0149 (3.1038) [2021-04-16 18:12:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][460/1251] eta 0:04:00 lr 0.000042 time 0.2647 (0.3043) loss 2.4970 (3.0258) grad_norm 2.9258 (3.1035) [2021-04-16 18:12:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][470/1251] eta 0:03:57 lr 0.000042 time 0.2472 (0.3037) loss 2.2438 (3.0245) grad_norm 3.1166 (3.1006) [2021-04-16 18:12:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][480/1251] eta 0:03:53 lr 0.000042 time 0.2999 (0.3033) loss 3.2399 (3.0260) grad_norm 3.3249 (3.1013) [2021-04-16 18:12:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][490/1251] eta 0:03:50 lr 0.000042 time 0.2888 (0.3032) loss 3.0045 (3.0219) grad_norm 4.2129 (3.1081) [2021-04-16 18:12:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][500/1251] eta 0:03:47 lr 0.000042 time 0.2656 (0.3027) loss 3.0353 (3.0257) grad_norm 2.8246 (3.1056) [2021-04-16 18:12:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][510/1251] eta 0:03:43 lr 0.000042 time 0.2529 (0.3021) loss 3.1839 (3.0212) grad_norm 2.6149 (3.1000) [2021-04-16 18:12:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][520/1251] eta 0:03:40 lr 0.000042 time 0.2597 (0.3016) loss 3.4322 (3.0174) grad_norm 2.7123 (3.0978) [2021-04-16 18:12:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][530/1251] eta 0:03:37 lr 0.000042 time 0.3066 (0.3012) loss 3.3436 (3.0169) grad_norm 3.1330 (3.0963) [2021-04-16 18:12:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][540/1251] eta 0:03:33 lr 0.000042 time 0.2642 (0.3007) loss 3.3954 (3.0215) grad_norm 3.1789 (3.0973) [2021-04-16 18:12:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][550/1251] eta 0:03:30 lr 0.000042 time 0.3013 (0.3003) loss 2.6163 (3.0208) grad_norm 3.4092 (3.0962) [2021-04-16 18:12:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][560/1251] eta 0:03:27 lr 0.000042 time 0.2582 (0.2998) loss 3.5013 (3.0270) grad_norm 2.9209 (3.0932) [2021-04-16 18:13:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][570/1251] eta 0:03:24 lr 0.000042 time 0.2512 (0.2999) loss 2.9504 (3.0244) grad_norm 2.8778 (3.0946) [2021-04-16 18:13:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][580/1251] eta 0:03:21 lr 0.000042 time 0.2860 (0.2996) loss 3.5387 (3.0253) grad_norm 3.0891 (3.0954) [2021-04-16 18:13:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][590/1251] eta 0:03:17 lr 0.000042 time 0.2713 (0.2992) loss 3.4834 (3.0265) grad_norm 3.0018 (3.0942) [2021-04-16 18:13:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][600/1251] eta 0:03:14 lr 0.000042 time 0.3170 (0.2988) loss 2.9750 (3.0274) grad_norm 3.2185 (3.0949) [2021-04-16 18:13:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][610/1251] eta 0:03:11 lr 0.000042 time 0.2989 (0.2985) loss 3.4184 (3.0296) grad_norm 2.7330 (3.0930) [2021-04-16 18:13:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][620/1251] eta 0:03:08 lr 0.000042 time 0.3244 (0.2982) loss 1.9757 (3.0261) grad_norm 3.3824 (3.0918) [2021-04-16 18:13:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][630/1251] eta 0:03:04 lr 0.000042 time 0.2611 (0.2979) loss 3.0557 (3.0301) grad_norm 2.5388 (3.0909) [2021-04-16 18:13:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][640/1251] eta 0:03:01 lr 0.000042 time 0.2859 (0.2978) loss 3.1688 (3.0323) grad_norm 5.8036 (3.0946) [2021-04-16 18:13:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][650/1251] eta 0:02:58 lr 0.000042 time 0.2680 (0.2974) loss 3.4967 (3.0319) grad_norm 2.5772 (3.0932) [2021-04-16 18:13:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][660/1251] eta 0:02:55 lr 0.000042 time 0.2779 (0.2971) loss 3.5306 (3.0351) grad_norm 2.5175 (3.0914) [2021-04-16 18:13:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][670/1251] eta 0:02:52 lr 0.000042 time 0.2661 (0.2970) loss 3.0804 (3.0385) grad_norm 3.1697 (3.0910) [2021-04-16 18:13:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][680/1251] eta 0:02:49 lr 0.000042 time 0.2638 (0.2967) loss 2.4525 (3.0347) grad_norm 3.4938 (3.0910) [2021-04-16 18:13:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][690/1251] eta 0:02:46 lr 0.000042 time 0.2597 (0.2964) loss 3.3061 (3.0347) grad_norm 3.6253 (3.0934) [2021-04-16 18:13:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][700/1251] eta 0:02:43 lr 0.000042 time 0.2624 (0.2961) loss 2.3847 (3.0336) grad_norm 3.9101 (3.0955) [2021-04-16 18:13:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][710/1251] eta 0:02:40 lr 0.000042 time 0.2857 (0.2959) loss 2.3957 (3.0305) grad_norm 3.2948 (3.0960) [2021-04-16 18:13:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][720/1251] eta 0:02:37 lr 0.000042 time 0.2629 (0.2961) loss 2.9842 (3.0291) grad_norm 2.8256 (3.0962) [2021-04-16 18:13:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][730/1251] eta 0:02:34 lr 0.000042 time 0.2667 (0.2960) loss 3.2978 (3.0298) grad_norm 3.2348 (3.0934) [2021-04-16 18:13:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][740/1251] eta 0:02:31 lr 0.000042 time 0.2879 (0.2958) loss 2.2897 (3.0257) grad_norm 2.7729 (3.0935) [2021-04-16 18:13:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][750/1251] eta 0:02:28 lr 0.000042 time 0.2653 (0.2956) loss 2.7308 (3.0244) grad_norm 2.9589 (3.0925) [2021-04-16 18:13:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][760/1251] eta 0:02:25 lr 0.000042 time 0.2627 (0.2954) loss 3.5441 (3.0255) grad_norm 3.1805 (3.0932) [2021-04-16 18:13:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][770/1251] eta 0:02:21 lr 0.000042 time 0.2718 (0.2952) loss 3.0774 (3.0288) grad_norm 3.0859 (3.0931) [2021-04-16 18:14:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][780/1251] eta 0:02:18 lr 0.000042 time 0.2788 (0.2951) loss 3.5872 (3.0319) grad_norm 2.9482 (3.0928) [2021-04-16 18:14:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][790/1251] eta 0:02:15 lr 0.000042 time 0.2633 (0.2950) loss 2.9663 (3.0310) grad_norm 3.0240 (3.0905) [2021-04-16 18:14:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][800/1251] eta 0:02:12 lr 0.000042 time 0.3002 (0.2948) loss 3.0765 (3.0346) grad_norm 2.8290 (3.0889) [2021-04-16 18:14:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][810/1251] eta 0:02:09 lr 0.000042 time 0.2990 (0.2946) loss 3.3568 (3.0346) grad_norm 3.2216 (3.0867) [2021-04-16 18:14:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][820/1251] eta 0:02:06 lr 0.000042 time 0.2801 (0.2944) loss 3.3288 (3.0365) grad_norm 2.5000 (3.0839) [2021-04-16 18:14:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][830/1251] eta 0:02:03 lr 0.000042 time 0.2908 (0.2942) loss 2.9419 (3.0340) grad_norm 3.5898 (3.0839) [2021-04-16 18:14:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][840/1251] eta 0:02:00 lr 0.000042 time 0.2652 (0.2940) loss 2.9904 (3.0348) grad_norm 2.6037 (3.0828) [2021-04-16 18:14:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][850/1251] eta 0:01:57 lr 0.000042 time 0.2764 (0.2938) loss 3.0707 (3.0357) grad_norm 2.7340 (3.0842) [2021-04-16 18:14:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][860/1251] eta 0:01:54 lr 0.000042 time 0.2720 (0.2937) loss 3.4662 (3.0372) grad_norm 3.0394 (3.0847) [2021-04-16 18:14:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][870/1251] eta 0:01:51 lr 0.000042 time 0.2772 (0.2935) loss 3.7210 (3.0380) grad_norm 3.2340 (3.0831) [2021-04-16 18:14:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][880/1251] eta 0:01:48 lr 0.000042 time 0.2874 (0.2934) loss 3.6939 (3.0372) grad_norm 3.2221 (3.0810) [2021-04-16 18:14:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][890/1251] eta 0:01:45 lr 0.000042 time 0.2480 (0.2933) loss 3.4293 (3.0388) grad_norm 3.4019 (3.0829) [2021-04-16 18:14:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][900/1251] eta 0:01:42 lr 0.000042 time 0.2754 (0.2931) loss 3.2208 (3.0385) grad_norm 3.9701 (3.0849) [2021-04-16 18:14:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][910/1251] eta 0:01:39 lr 0.000042 time 0.2780 (0.2930) loss 3.6453 (3.0387) grad_norm 3.0785 (3.0840) [2021-04-16 18:14:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][920/1251] eta 0:01:36 lr 0.000042 time 0.2729 (0.2928) loss 2.1449 (3.0372) grad_norm 3.2446 (3.0834) [2021-04-16 18:14:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][930/1251] eta 0:01:33 lr 0.000042 time 0.2769 (0.2926) loss 3.7640 (3.0375) grad_norm 2.7857 (3.0815) [2021-04-16 18:14:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][940/1251] eta 0:01:30 lr 0.000041 time 0.2803 (0.2925) loss 2.7151 (3.0396) grad_norm 2.6559 (3.0823) [2021-04-16 18:14:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][950/1251] eta 0:01:28 lr 0.000041 time 0.2712 (0.2924) loss 2.8962 (3.0390) grad_norm 3.4116 (3.0831) [2021-04-16 18:14:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][960/1251] eta 0:01:25 lr 0.000041 time 0.2562 (0.2923) loss 2.5381 (3.0360) grad_norm 2.7070 (3.0851) [2021-04-16 18:14:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][970/1251] eta 0:01:22 lr 0.000041 time 0.2682 (0.2921) loss 3.2621 (3.0362) grad_norm 2.9887 (3.0841) [2021-04-16 18:14:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][980/1251] eta 0:01:19 lr 0.000041 time 0.2672 (0.2919) loss 2.4841 (3.0353) grad_norm 3.7092 (3.0849) [2021-04-16 18:14:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][990/1251] eta 0:01:16 lr 0.000041 time 0.2598 (0.2917) loss 2.6864 (3.0339) grad_norm 3.2239 (3.0844) [2021-04-16 18:15:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][1000/1251] eta 0:01:13 lr 0.000041 time 0.2771 (0.2915) loss 3.1410 (3.0355) grad_norm 3.0115 (3.0855) [2021-04-16 18:15:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][1010/1251] eta 0:01:10 lr 0.000041 time 0.2840 (0.2915) loss 2.6531 (3.0331) grad_norm 2.6639 (3.0842) [2021-04-16 18:15:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][1020/1251] eta 0:01:07 lr 0.000041 time 0.2855 (0.2914) loss 2.5844 (3.0312) grad_norm 3.4495 (3.0838) [2021-04-16 18:15:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][1030/1251] eta 0:01:04 lr 0.000041 time 0.2859 (0.2914) loss 3.6095 (3.0314) grad_norm 3.3451 (3.0833) [2021-04-16 18:15:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][1040/1251] eta 0:01:01 lr 0.000041 time 0.2663 (0.2913) loss 3.1862 (3.0308) grad_norm 2.7862 (3.0812) [2021-04-16 18:15:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][1050/1251] eta 0:00:58 lr 0.000041 time 0.2696 (0.2911) loss 2.1692 (3.0300) grad_norm 3.2157 (3.0800) [2021-04-16 18:15:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][1060/1251] eta 0:00:55 lr 0.000041 time 0.2877 (0.2910) loss 3.5815 (3.0305) grad_norm 2.8970 (3.0807) [2021-04-16 18:15:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][1070/1251] eta 0:00:52 lr 0.000041 time 0.2496 (0.2910) loss 2.4786 (3.0297) grad_norm 2.5699 (3.0793) [2021-04-16 18:15:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][1080/1251] eta 0:00:49 lr 0.000041 time 0.2738 (0.2909) loss 3.0452 (3.0288) grad_norm 5.2280 (3.0794) [2021-04-16 18:15:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][1090/1251] eta 0:00:46 lr 0.000041 time 0.2679 (0.2908) loss 3.5170 (3.0294) grad_norm 3.1560 (3.0787) [2021-04-16 18:15:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][1100/1251] eta 0:00:43 lr 0.000041 time 0.2646 (0.2906) loss 3.4269 (3.0324) grad_norm 3.6339 (3.0781) [2021-04-16 18:15:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][1110/1251] eta 0:00:40 lr 0.000041 time 0.2450 (0.2905) loss 3.5949 (3.0334) grad_norm 2.8436 (3.0782) [2021-04-16 18:15:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][1120/1251] eta 0:00:38 lr 0.000041 time 0.2766 (0.2904) loss 3.0483 (3.0348) grad_norm 3.4006 (3.0783) [2021-04-16 18:15:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][1130/1251] eta 0:00:35 lr 0.000041 time 0.2960 (0.2902) loss 3.2542 (3.0359) grad_norm 4.3276 (3.0794) [2021-04-16 18:15:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][1140/1251] eta 0:00:32 lr 0.000041 time 0.2813 (0.2901) loss 3.5302 (3.0356) grad_norm 3.4819 (3.0789) [2021-04-16 18:15:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][1150/1251] eta 0:00:29 lr 0.000041 time 0.2647 (0.2903) loss 3.2136 (3.0342) grad_norm 3.0830 (3.0781) [2021-04-16 18:15:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][1160/1251] eta 0:00:26 lr 0.000041 time 0.2784 (0.2901) loss 3.7383 (3.0351) grad_norm 3.2061 (3.0792) [2021-04-16 18:15:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][1170/1251] eta 0:00:23 lr 0.000041 time 0.2708 (0.2900) loss 2.9748 (3.0353) grad_norm 2.7711 (3.0787) [2021-04-16 18:15:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][1180/1251] eta 0:00:20 lr 0.000041 time 0.2586 (0.2898) loss 3.2837 (3.0347) grad_norm 2.8621 (3.0782) [2021-04-16 18:15:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][1190/1251] eta 0:00:17 lr 0.000041 time 0.3130 (0.2898) loss 2.9631 (3.0342) grad_norm 2.7953 (3.0793) [2021-04-16 18:15:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][1200/1251] eta 0:00:14 lr 0.000041 time 0.2702 (0.2896) loss 2.9063 (3.0302) grad_norm 2.6372 (3.0771) [2021-04-16 18:16:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][1210/1251] eta 0:00:11 lr 0.000041 time 0.2888 (0.2896) loss 2.8064 (3.0295) grad_norm 2.9251 (3.0772) [2021-04-16 18:16:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][1220/1251] eta 0:00:08 lr 0.000041 time 0.2694 (0.2895) loss 3.3626 (3.0280) grad_norm 3.4999 (3.0774) [2021-04-16 18:16:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][1230/1251] eta 0:00:06 lr 0.000041 time 0.2854 (0.2894) loss 2.9001 (3.0285) grad_norm 2.9847 (3.0767) [2021-04-16 18:16:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][1240/1251] eta 0:00:03 lr 0.000041 time 0.3276 (0.2893) loss 3.4568 (3.0304) grad_norm 3.4242 (inf) [2021-04-16 18:16:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [265/300][1250/1251] eta 0:00:00 lr 0.000041 time 0.2492 (0.2890) loss 3.2683 (3.0295) grad_norm 2.9275 (inf) [2021-04-16 18:16:16 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 265 training takes 0:06:06 [2021-04-16 18:16:16 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_265.pth saving...... [2021-04-16 18:16:35 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_265.pth saved !!! [2021-04-16 18:16:36 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.253 (1.253) Loss 0.7858 (0.7858) Acc@1 81.641 (81.641) Acc@5 95.312 (95.312) [2021-04-16 18:16:37 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.155 (0.254) Loss 0.8982 (0.8330) Acc@1 79.004 (80.327) Acc@5 94.727 (95.162) [2021-04-16 18:16:39 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.107 (0.226) Loss 0.8313 (0.8330) Acc@1 79.395 (80.422) Acc@5 96.094 (95.299) [2021-04-16 18:16:41 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.151 (0.216) Loss 0.8067 (0.8315) Acc@1 80.762 (80.447) Acc@5 95.215 (95.325) [2021-04-16 18:16:44 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.218) Loss 0.8705 (0.8276) Acc@1 79.883 (80.664) Acc@5 94.922 (95.351) [2021-04-16 18:16:55 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.748 Acc@5 95.388 [2021-04-16 18:16:55 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 80.7% [2021-04-16 18:16:55 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 80.85% [2021-04-16 18:17:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][0/1251] eta 4:38:25 lr 0.000041 time 13.3539 (13.3539) loss 3.0909 (3.0909) grad_norm 2.6931 (2.6931) [2021-04-16 18:17:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][10/1251] eta 0:30:37 lr 0.000041 time 0.4800 (1.4803) loss 2.9553 (2.9368) grad_norm 3.2381 (2.8745) [2021-04-16 18:17:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][20/1251] eta 0:18:37 lr 0.000041 time 0.2767 (0.9076) loss 3.1557 (2.9868) grad_norm 2.6462 (2.9097) [2021-04-16 18:17:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][30/1251] eta 0:14:18 lr 0.000041 time 0.2675 (0.7033) loss 3.6270 (2.9948) grad_norm 2.8504 (2.8882) [2021-04-16 18:17:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][40/1251] eta 0:12:03 lr 0.000041 time 0.2669 (0.5978) loss 3.3840 (3.0071) grad_norm 4.4445 (2.9288) [2021-04-16 18:17:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][50/1251] eta 0:10:43 lr 0.000041 time 0.2658 (0.5357) loss 3.6660 (3.0437) grad_norm 2.8759 (2.9454) [2021-04-16 18:17:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][60/1251] eta 0:09:48 lr 0.000041 time 0.2783 (0.4941) loss 3.0171 (3.0003) grad_norm 3.0201 (2.9416) [2021-04-16 18:17:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][70/1251] eta 0:09:07 lr 0.000041 time 0.2643 (0.4634) loss 3.2121 (3.0168) grad_norm 2.8324 (2.9356) [2021-04-16 18:17:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][80/1251] eta 0:08:36 lr 0.000041 time 0.2582 (0.4410) loss 2.4316 (3.0148) grad_norm 3.8196 (2.9436) [2021-04-16 18:17:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][90/1251] eta 0:08:11 lr 0.000041 time 0.2895 (0.4230) loss 3.5461 (2.9893) grad_norm 2.8252 (2.9544) [2021-04-16 18:17:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][100/1251] eta 0:07:50 lr 0.000041 time 0.2836 (0.4085) loss 2.5798 (2.9890) grad_norm 2.9102 (3.0017) [2021-04-16 18:17:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][110/1251] eta 0:07:32 lr 0.000041 time 0.2894 (0.3965) loss 3.3880 (3.0022) grad_norm 2.6902 (3.0048) [2021-04-16 18:17:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][120/1251] eta 0:07:16 lr 0.000041 time 0.2690 (0.3864) loss 3.5465 (2.9995) grad_norm 2.9018 (3.0107) [2021-04-16 18:17:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][130/1251] eta 0:07:03 lr 0.000041 time 0.2726 (0.3782) loss 2.3692 (3.0081) grad_norm 3.3146 (3.0159) [2021-04-16 18:17:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][140/1251] eta 0:06:52 lr 0.000041 time 0.3059 (0.3711) loss 3.5477 (2.9969) grad_norm 3.1773 (3.0186) [2021-04-16 18:17:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][150/1251] eta 0:06:41 lr 0.000041 time 0.2861 (0.3645) loss 2.6641 (3.0072) grad_norm 2.9776 (3.0220) [2021-04-16 18:17:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][160/1251] eta 0:06:32 lr 0.000041 time 0.2643 (0.3597) loss 2.8479 (3.0013) grad_norm 2.7203 (3.0243) [2021-04-16 18:17:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][170/1251] eta 0:06:23 lr 0.000041 time 0.2679 (0.3549) loss 3.5059 (3.0142) grad_norm 3.8073 (3.0268) [2021-04-16 18:17:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][180/1251] eta 0:06:15 lr 0.000041 time 0.2818 (0.3506) loss 2.4709 (3.0178) grad_norm 4.0150 (3.0346) [2021-04-16 18:18:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][190/1251] eta 0:06:08 lr 0.000041 time 0.2777 (0.3471) loss 2.5622 (3.0192) grad_norm 2.5241 (3.0456) [2021-04-16 18:18:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][200/1251] eta 0:06:01 lr 0.000041 time 0.2586 (0.3436) loss 2.4362 (3.0256) grad_norm 2.6172 (3.0509) [2021-04-16 18:18:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][210/1251] eta 0:05:54 lr 0.000041 time 0.2644 (0.3403) loss 1.9233 (3.0123) grad_norm 3.2296 (3.0523) [2021-04-16 18:18:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][220/1251] eta 0:05:47 lr 0.000041 time 0.2651 (0.3373) loss 2.7506 (3.0098) grad_norm 2.7159 (3.0446) [2021-04-16 18:18:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][230/1251] eta 0:05:41 lr 0.000041 time 0.2645 (0.3346) loss 2.6352 (3.0173) grad_norm 2.8855 (3.0439) [2021-04-16 18:18:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][240/1251] eta 0:05:35 lr 0.000041 time 0.2745 (0.3321) loss 2.9227 (3.0002) grad_norm 3.7799 (3.0384) [2021-04-16 18:18:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][250/1251] eta 0:05:30 lr 0.000041 time 0.2749 (0.3302) loss 2.2263 (2.9946) grad_norm 2.4816 (3.0332) [2021-04-16 18:18:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][260/1251] eta 0:05:25 lr 0.000041 time 0.2573 (0.3283) loss 2.0538 (3.0009) grad_norm 2.9949 (3.0292) [2021-04-16 18:18:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][270/1251] eta 0:05:20 lr 0.000041 time 0.2743 (0.3265) loss 3.1744 (3.0042) grad_norm 3.2545 (3.0264) [2021-04-16 18:18:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][280/1251] eta 0:05:15 lr 0.000041 time 0.2612 (0.3247) loss 2.7384 (3.0057) grad_norm 2.9489 (3.0228) [2021-04-16 18:18:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][290/1251] eta 0:05:11 lr 0.000041 time 0.2899 (0.3237) loss 3.3343 (3.0043) grad_norm 2.8029 (3.0223) [2021-04-16 18:18:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][300/1251] eta 0:05:06 lr 0.000041 time 0.2879 (0.3221) loss 3.6020 (3.0093) grad_norm 2.6103 (3.0183) [2021-04-16 18:18:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][310/1251] eta 0:05:01 lr 0.000041 time 0.3047 (0.3207) loss 3.4943 (3.0037) grad_norm 2.8196 (3.0186) [2021-04-16 18:18:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][320/1251] eta 0:04:57 lr 0.000041 time 0.2795 (0.3193) loss 2.5744 (3.0073) grad_norm 2.8572 (3.0122) [2021-04-16 18:18:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][330/1251] eta 0:04:53 lr 0.000041 time 0.2914 (0.3182) loss 3.1809 (3.0078) grad_norm 2.8457 (3.0177) [2021-04-16 18:18:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][340/1251] eta 0:04:48 lr 0.000041 time 0.3126 (0.3170) loss 3.6868 (3.0002) grad_norm 3.0275 (3.0184) [2021-04-16 18:18:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][350/1251] eta 0:04:44 lr 0.000041 time 0.2789 (0.3159) loss 3.6026 (3.0066) grad_norm 2.6951 (3.0206) [2021-04-16 18:18:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][360/1251] eta 0:04:41 lr 0.000041 time 0.2638 (0.3158) loss 3.7959 (3.0140) grad_norm 3.1567 (3.0187) [2021-04-16 18:18:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][370/1251] eta 0:04:37 lr 0.000041 time 0.2799 (0.3146) loss 3.4362 (3.0133) grad_norm 2.9666 (3.0177) [2021-04-16 18:18:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][380/1251] eta 0:04:33 lr 0.000040 time 0.2991 (0.3138) loss 2.2191 (3.0133) grad_norm 3.2273 (3.0207) [2021-04-16 18:18:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][390/1251] eta 0:04:29 lr 0.000040 time 0.2638 (0.3128) loss 3.0676 (3.0095) grad_norm 3.3443 (3.0210) [2021-04-16 18:19:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][400/1251] eta 0:04:25 lr 0.000040 time 0.2686 (0.3120) loss 2.5491 (3.0129) grad_norm 3.6305 (3.0203) [2021-04-16 18:19:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][410/1251] eta 0:04:21 lr 0.000040 time 0.2852 (0.3111) loss 3.1539 (3.0126) grad_norm 3.3694 (3.0221) [2021-04-16 18:19:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][420/1251] eta 0:04:18 lr 0.000040 time 0.2886 (0.3106) loss 3.3329 (3.0176) grad_norm 5.0092 (3.0285) [2021-04-16 18:19:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][430/1251] eta 0:04:14 lr 0.000040 time 0.2831 (0.3098) loss 2.9923 (3.0157) grad_norm 3.1996 (3.0302) [2021-04-16 18:19:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][440/1251] eta 0:04:10 lr 0.000040 time 0.2872 (0.3090) loss 3.4174 (3.0161) grad_norm 3.0312 (3.0307) [2021-04-16 18:19:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][450/1251] eta 0:04:07 lr 0.000040 time 0.3606 (0.3086) loss 2.0563 (3.0152) grad_norm 3.3735 (3.0304) [2021-04-16 18:19:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][460/1251] eta 0:04:03 lr 0.000040 time 0.2988 (0.3078) loss 2.1542 (3.0132) grad_norm 3.2880 (3.0356) [2021-04-16 18:19:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][470/1251] eta 0:03:59 lr 0.000040 time 0.2844 (0.3071) loss 3.3280 (3.0152) grad_norm 2.6776 (3.0336) [2021-04-16 18:19:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][480/1251] eta 0:03:56 lr 0.000040 time 0.2959 (0.3064) loss 2.4960 (3.0136) grad_norm 4.6136 (3.0360) [2021-04-16 18:19:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][490/1251] eta 0:03:52 lr 0.000040 time 0.2716 (0.3058) loss 3.4021 (3.0123) grad_norm 2.8780 (3.0359) [2021-04-16 18:19:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][500/1251] eta 0:03:49 lr 0.000040 time 0.2667 (0.3052) loss 3.2536 (3.0105) grad_norm 2.8985 (3.0360) [2021-04-16 18:19:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][510/1251] eta 0:03:45 lr 0.000040 time 0.2658 (0.3046) loss 3.3258 (3.0101) grad_norm 2.6371 (3.0355) [2021-04-16 18:19:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][520/1251] eta 0:03:42 lr 0.000040 time 0.2739 (0.3043) loss 3.9401 (3.0150) grad_norm 3.4433 (3.0357) [2021-04-16 18:19:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][530/1251] eta 0:03:39 lr 0.000040 time 0.2927 (0.3038) loss 3.3775 (3.0137) grad_norm 3.2875 (3.0390) [2021-04-16 18:19:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][540/1251] eta 0:03:35 lr 0.000040 time 0.2954 (0.3036) loss 2.4919 (3.0150) grad_norm 2.9242 (3.0386) [2021-04-16 18:19:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][550/1251] eta 0:03:32 lr 0.000040 time 0.2784 (0.3033) loss 3.8964 (3.0185) grad_norm 2.7517 (3.0345) [2021-04-16 18:19:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][560/1251] eta 0:03:29 lr 0.000040 time 0.2605 (0.3029) loss 3.2479 (3.0157) grad_norm 2.8568 (3.0416) [2021-04-16 18:19:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][570/1251] eta 0:03:25 lr 0.000040 time 0.3133 (0.3025) loss 2.8452 (3.0122) grad_norm 2.8743 (3.0422) [2021-04-16 18:19:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][580/1251] eta 0:03:22 lr 0.000040 time 0.2836 (0.3024) loss 3.4170 (3.0136) grad_norm 2.9098 (3.0438) [2021-04-16 18:19:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][590/1251] eta 0:03:19 lr 0.000040 time 0.2673 (0.3022) loss 2.2970 (3.0123) grad_norm 3.3710 (3.0458) [2021-04-16 18:19:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][600/1251] eta 0:03:16 lr 0.000040 time 0.2728 (0.3019) loss 2.9404 (3.0159) grad_norm 3.3558 (3.0471) [2021-04-16 18:19:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][610/1251] eta 0:03:13 lr 0.000040 time 0.2611 (0.3014) loss 2.8659 (3.0182) grad_norm 3.9245 (3.0461) [2021-04-16 18:20:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][620/1251] eta 0:03:09 lr 0.000040 time 0.2687 (0.3010) loss 3.1766 (3.0136) grad_norm 2.9077 (3.0432) [2021-04-16 18:20:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][630/1251] eta 0:03:06 lr 0.000040 time 0.2968 (0.3006) loss 3.5730 (3.0130) grad_norm 3.0050 (3.0460) [2021-04-16 18:20:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][640/1251] eta 0:03:03 lr 0.000040 time 0.2793 (0.3003) loss 3.0531 (3.0096) grad_norm 3.1486 (3.0475) [2021-04-16 18:20:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][650/1251] eta 0:03:00 lr 0.000040 time 0.2585 (0.2998) loss 3.5194 (3.0123) grad_norm 3.1468 (3.0480) [2021-04-16 18:20:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][660/1251] eta 0:02:56 lr 0.000040 time 0.2897 (0.2995) loss 3.5285 (3.0109) grad_norm 3.2646 (3.0479) [2021-04-16 18:20:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][670/1251] eta 0:02:53 lr 0.000040 time 0.2822 (0.2991) loss 3.4445 (3.0111) grad_norm 3.1080 (3.0496) [2021-04-16 18:20:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][680/1251] eta 0:02:50 lr 0.000040 time 0.2666 (0.2987) loss 2.9157 (3.0101) grad_norm 3.0511 (3.0518) [2021-04-16 18:20:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][690/1251] eta 0:02:47 lr 0.000040 time 0.2824 (0.2985) loss 2.4597 (3.0108) grad_norm 2.8658 (3.0525) [2021-04-16 18:20:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][700/1251] eta 0:02:44 lr 0.000040 time 0.2650 (0.2981) loss 3.5154 (3.0090) grad_norm 3.0469 (3.0550) [2021-04-16 18:20:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][710/1251] eta 0:02:41 lr 0.000040 time 0.2587 (0.2978) loss 3.0625 (3.0104) grad_norm 3.3195 (3.0543) [2021-04-16 18:20:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][720/1251] eta 0:02:37 lr 0.000040 time 0.2776 (0.2975) loss 2.8916 (3.0084) grad_norm 3.1299 (3.0560) [2021-04-16 18:20:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][730/1251] eta 0:02:34 lr 0.000040 time 0.2586 (0.2972) loss 3.2841 (3.0090) grad_norm 3.4057 (3.0542) [2021-04-16 18:20:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][740/1251] eta 0:02:31 lr 0.000040 time 0.2608 (0.2970) loss 3.8337 (3.0119) grad_norm 2.7311 (3.0524) [2021-04-16 18:20:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][750/1251] eta 0:02:28 lr 0.000040 time 0.2829 (0.2968) loss 2.0277 (3.0080) grad_norm 3.2569 (3.0558) [2021-04-16 18:20:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][760/1251] eta 0:02:25 lr 0.000040 time 0.2644 (0.2966) loss 3.3416 (3.0116) grad_norm 3.4494 (3.0564) [2021-04-16 18:20:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][770/1251] eta 0:02:22 lr 0.000040 time 0.2641 (0.2964) loss 2.9923 (3.0116) grad_norm 2.9155 (3.0582) [2021-04-16 18:20:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][780/1251] eta 0:02:19 lr 0.000040 time 0.2751 (0.2961) loss 3.3996 (3.0092) grad_norm 2.8012 (3.0590) [2021-04-16 18:20:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][790/1251] eta 0:02:16 lr 0.000040 time 0.2712 (0.2959) loss 3.0546 (3.0097) grad_norm 3.0133 (3.0572) [2021-04-16 18:20:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][800/1251] eta 0:02:13 lr 0.000040 time 0.2605 (0.2956) loss 2.0472 (3.0108) grad_norm 2.5836 (3.0559) [2021-04-16 18:20:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][810/1251] eta 0:02:10 lr 0.000040 time 0.2805 (0.2954) loss 2.8078 (3.0124) grad_norm 3.3652 (3.0572) [2021-04-16 18:20:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][820/1251] eta 0:02:07 lr 0.000040 time 0.2688 (0.2952) loss 2.2292 (3.0143) grad_norm 2.7505 (3.0556) [2021-04-16 18:21:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][830/1251] eta 0:02:04 lr 0.000040 time 0.2717 (0.2951) loss 3.5241 (3.0206) grad_norm 3.0300 (3.0561) [2021-04-16 18:21:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][840/1251] eta 0:02:01 lr 0.000040 time 0.2584 (0.2948) loss 3.1646 (3.0159) grad_norm 3.3195 (3.0570) [2021-04-16 18:21:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][850/1251] eta 0:01:58 lr 0.000040 time 0.2781 (0.2947) loss 3.4110 (3.0187) grad_norm 3.7853 (3.0578) [2021-04-16 18:21:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][860/1251] eta 0:01:55 lr 0.000040 time 0.3018 (0.2945) loss 3.4249 (3.0165) grad_norm 3.1122 (3.0570) [2021-04-16 18:21:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][870/1251] eta 0:01:52 lr 0.000040 time 0.2739 (0.2944) loss 3.5071 (3.0183) grad_norm 3.9631 (3.0578) [2021-04-16 18:21:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][880/1251] eta 0:01:49 lr 0.000040 time 0.2829 (0.2943) loss 2.9817 (3.0151) grad_norm 2.5945 (3.0600) [2021-04-16 18:21:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][890/1251] eta 0:01:46 lr 0.000040 time 0.2647 (0.2940) loss 3.4954 (3.0171) grad_norm 3.2158 (3.0602) [2021-04-16 18:21:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][900/1251] eta 0:01:43 lr 0.000040 time 0.2747 (0.2939) loss 1.9715 (3.0175) grad_norm 3.0210 (3.0613) [2021-04-16 18:21:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][910/1251] eta 0:01:40 lr 0.000040 time 0.2708 (0.2937) loss 2.2519 (3.0135) grad_norm 2.7588 (3.0607) [2021-04-16 18:21:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][920/1251] eta 0:01:37 lr 0.000040 time 0.2667 (0.2935) loss 3.0278 (3.0135) grad_norm 2.8479 (3.0600) [2021-04-16 18:21:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][930/1251] eta 0:01:34 lr 0.000040 time 0.2868 (0.2934) loss 2.8998 (3.0146) grad_norm 2.8820 (3.0601) [2021-04-16 18:21:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][940/1251] eta 0:01:31 lr 0.000040 time 0.2656 (0.2935) loss 3.7218 (3.0139) grad_norm 3.1774 (3.0581) [2021-04-16 18:21:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][950/1251] eta 0:01:28 lr 0.000040 time 0.2689 (0.2933) loss 3.3683 (3.0111) grad_norm 3.0040 (3.0574) [2021-04-16 18:21:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][960/1251] eta 0:01:25 lr 0.000040 time 0.2851 (0.2931) loss 2.8695 (3.0104) grad_norm 3.0907 (3.0590) [2021-04-16 18:21:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][970/1251] eta 0:01:22 lr 0.000040 time 0.2604 (0.2929) loss 3.2098 (3.0093) grad_norm 2.9852 (3.0582) [2021-04-16 18:21:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][980/1251] eta 0:01:19 lr 0.000040 time 0.2771 (0.2928) loss 2.4415 (3.0081) grad_norm 4.2332 (3.0605) [2021-04-16 18:21:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][990/1251] eta 0:01:16 lr 0.000040 time 0.2593 (0.2927) loss 3.5486 (3.0112) grad_norm 2.8017 (3.0645) [2021-04-16 18:21:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][1000/1251] eta 0:01:13 lr 0.000040 time 0.2558 (0.2925) loss 2.4263 (3.0101) grad_norm 2.8390 (3.0630) [2021-04-16 18:21:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][1010/1251] eta 0:01:10 lr 0.000040 time 0.2519 (0.2924) loss 2.4193 (3.0114) grad_norm 3.1580 (3.0645) [2021-04-16 18:21:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][1020/1251] eta 0:01:07 lr 0.000040 time 0.2745 (0.2923) loss 3.5912 (3.0110) grad_norm 2.5811 (3.0634) [2021-04-16 18:21:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][1030/1251] eta 0:01:04 lr 0.000040 time 0.3076 (0.2922) loss 3.2060 (3.0107) grad_norm 2.7824 (3.0624) [2021-04-16 18:21:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][1040/1251] eta 0:01:01 lr 0.000040 time 0.2647 (0.2920) loss 3.0275 (3.0123) grad_norm 3.5219 (3.0627) [2021-04-16 18:22:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][1050/1251] eta 0:00:58 lr 0.000040 time 0.2719 (0.2919) loss 2.9497 (3.0111) grad_norm 2.5706 (3.0621) [2021-04-16 18:22:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][1060/1251] eta 0:00:55 lr 0.000040 time 0.2591 (0.2917) loss 3.3684 (3.0141) grad_norm 3.4982 (3.0631) [2021-04-16 18:22:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][1070/1251] eta 0:00:52 lr 0.000040 time 0.2769 (0.2915) loss 2.0559 (3.0128) grad_norm 2.7238 (3.0637) [2021-04-16 18:22:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][1080/1251] eta 0:00:49 lr 0.000040 time 0.2639 (0.2913) loss 2.4928 (3.0118) grad_norm 3.3901 (3.0656) [2021-04-16 18:22:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][1090/1251] eta 0:00:46 lr 0.000039 time 0.2495 (0.2911) loss 3.1208 (3.0128) grad_norm 2.9397 (3.0650) [2021-04-16 18:22:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][1100/1251] eta 0:00:43 lr 0.000039 time 0.2679 (0.2910) loss 2.8160 (3.0134) grad_norm 2.9779 (3.0636) [2021-04-16 18:22:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][1110/1251] eta 0:00:41 lr 0.000039 time 0.2761 (0.2909) loss 2.8941 (3.0135) grad_norm 3.9268 (3.0656) [2021-04-16 18:22:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][1120/1251] eta 0:00:38 lr 0.000039 time 0.2693 (0.2908) loss 3.8623 (3.0148) grad_norm 3.1137 (3.0665) [2021-04-16 18:22:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][1130/1251] eta 0:00:35 lr 0.000039 time 0.2756 (0.2908) loss 2.7290 (3.0146) grad_norm 3.2329 (3.0667) [2021-04-16 18:22:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][1140/1251] eta 0:00:32 lr 0.000039 time 0.2830 (0.2907) loss 2.5159 (3.0142) grad_norm 2.8908 (3.0684) [2021-04-16 18:22:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][1150/1251] eta 0:00:29 lr 0.000039 time 0.2700 (0.2905) loss 3.0731 (3.0151) grad_norm 3.1654 (3.0697) [2021-04-16 18:22:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][1160/1251] eta 0:00:26 lr 0.000039 time 0.2608 (0.2906) loss 1.8841 (3.0136) grad_norm 2.8865 (3.0688) [2021-04-16 18:22:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][1170/1251] eta 0:00:23 lr 0.000039 time 0.2535 (0.2906) loss 3.4801 (3.0132) grad_norm 3.8123 (3.0696) [2021-04-16 18:22:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][1180/1251] eta 0:00:20 lr 0.000039 time 0.2667 (0.2905) loss 3.2957 (3.0109) grad_norm 2.5732 (3.0682) [2021-04-16 18:22:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][1190/1251] eta 0:00:17 lr 0.000039 time 0.2627 (0.2903) loss 3.1776 (3.0115) grad_norm 2.6896 (3.0671) [2021-04-16 18:22:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][1200/1251] eta 0:00:14 lr 0.000039 time 0.2737 (0.2903) loss 3.1159 (3.0109) grad_norm 2.9759 (3.0680) [2021-04-16 18:22:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][1210/1251] eta 0:00:11 lr 0.000039 time 0.2893 (0.2902) loss 2.6291 (3.0077) grad_norm 3.3021 (3.0680) [2021-04-16 18:22:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][1220/1251] eta 0:00:08 lr 0.000039 time 0.2689 (0.2900) loss 3.1965 (3.0080) grad_norm 3.8897 (3.0695) [2021-04-16 18:22:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][1230/1251] eta 0:00:06 lr 0.000039 time 0.2676 (0.2899) loss 2.4356 (3.0077) grad_norm 3.0817 (3.0701) [2021-04-16 18:22:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][1240/1251] eta 0:00:03 lr 0.000039 time 0.2496 (0.2897) loss 2.7544 (3.0054) grad_norm 2.8191 (3.0711) [2021-04-16 18:22:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [266/300][1250/1251] eta 0:00:00 lr 0.000039 time 0.2521 (0.2894) loss 3.8237 (3.0054) grad_norm 3.4593 (3.0730) [2021-04-16 18:23:05 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 266 training takes 0:06:09 [2021-04-16 18:23:05 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_266.pth saving...... [2021-04-16 18:23:24 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_266.pth saved !!! [2021-04-16 18:23:25 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.146 (1.146) Loss 0.9022 (0.9022) Acc@1 78.711 (78.711) Acc@5 94.141 (94.141) [2021-04-16 18:23:27 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.155 (0.260) Loss 0.7998 (0.8365) Acc@1 82.422 (80.850) Acc@5 94.141 (95.144) [2021-04-16 18:23:29 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.076 (0.243) Loss 0.7597 (0.8313) Acc@1 82.129 (80.813) Acc@5 95.801 (95.322) [2021-04-16 18:23:31 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.116 (0.224) Loss 0.7865 (0.8306) Acc@1 81.641 (80.894) Acc@5 96.094 (95.347) [2021-04-16 18:23:33 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.220) Loss 0.8128 (0.8276) Acc@1 80.273 (80.871) Acc@5 95.020 (95.370) [2021-04-16 18:23:55 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.888 Acc@5 95.334 [2021-04-16 18:23:55 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 80.9% [2021-04-16 18:23:55 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 80.89% [2021-04-16 18:24:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][0/1251] eta 3:58:16 lr 0.000039 time 11.4278 (11.4278) loss 3.3946 (3.3946) grad_norm 3.5080 (3.5080) [2021-04-16 18:24:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][10/1251] eta 0:26:35 lr 0.000039 time 0.2654 (1.2857) loss 3.3155 (2.8212) grad_norm 2.9429 (3.1594) [2021-04-16 18:24:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][20/1251] eta 0:16:31 lr 0.000039 time 0.2915 (0.8056) loss 2.3876 (2.8387) grad_norm 2.9109 (3.0291) [2021-04-16 18:24:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][30/1251] eta 0:12:56 lr 0.000039 time 0.2677 (0.6360) loss 2.8301 (2.8691) grad_norm 3.0115 (3.0302) [2021-04-16 18:24:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][40/1251] eta 0:11:07 lr 0.000039 time 0.2698 (0.5508) loss 3.0228 (2.9394) grad_norm 2.9026 (3.0290) [2021-04-16 18:24:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][50/1251] eta 0:09:57 lr 0.000039 time 0.2695 (0.4972) loss 2.6590 (2.9147) grad_norm 3.1435 (3.0522) [2021-04-16 18:24:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][60/1251] eta 0:09:11 lr 0.000039 time 0.4091 (0.4629) loss 2.9754 (2.9413) grad_norm 3.0223 (3.0193) [2021-04-16 18:24:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][70/1251] eta 0:08:35 lr 0.000039 time 0.2755 (0.4364) loss 3.3468 (2.9182) grad_norm 4.0913 (3.0289) [2021-04-16 18:24:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][80/1251] eta 0:08:10 lr 0.000039 time 0.2635 (0.4188) loss 3.0622 (2.9594) grad_norm 3.3623 (3.0592) [2021-04-16 18:24:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][90/1251] eta 0:07:48 lr 0.000039 time 0.2578 (0.4034) loss 3.6206 (2.9605) grad_norm 3.5268 (3.0574) [2021-04-16 18:24:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][100/1251] eta 0:07:29 lr 0.000039 time 0.2675 (0.3908) loss 3.4055 (2.9842) grad_norm 3.4115 (3.0579) [2021-04-16 18:24:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][110/1251] eta 0:07:14 lr 0.000039 time 0.3017 (0.3809) loss 2.3800 (3.0031) grad_norm 2.8186 (3.0667) [2021-04-16 18:24:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][120/1251] eta 0:07:00 lr 0.000039 time 0.2501 (0.3721) loss 3.3700 (3.0107) grad_norm 3.2502 (3.1017) [2021-04-16 18:24:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][130/1251] eta 0:06:48 lr 0.000039 time 0.2701 (0.3647) loss 2.6036 (3.0101) grad_norm 3.1738 (3.1034) [2021-04-16 18:24:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][140/1251] eta 0:06:38 lr 0.000039 time 0.2970 (0.3587) loss 3.3417 (3.0114) grad_norm 3.2027 (3.1009) [2021-04-16 18:24:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][150/1251] eta 0:06:30 lr 0.000039 time 0.2509 (0.3543) loss 3.4325 (3.0066) grad_norm 2.9670 (3.1199) [2021-04-16 18:24:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][160/1251] eta 0:06:22 lr 0.000039 time 0.2691 (0.3505) loss 3.1358 (3.0036) grad_norm 3.1899 (3.1175) [2021-04-16 18:24:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][170/1251] eta 0:06:14 lr 0.000039 time 0.2786 (0.3461) loss 3.4157 (3.0004) grad_norm 2.6712 (3.1149) [2021-04-16 18:24:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][180/1251] eta 0:06:06 lr 0.000039 time 0.2698 (0.3420) loss 3.4938 (3.0173) grad_norm 3.2142 (3.1143) [2021-04-16 18:25:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][190/1251] eta 0:05:59 lr 0.000039 time 0.2994 (0.3387) loss 3.4931 (3.0298) grad_norm 3.3735 (3.1196) [2021-04-16 18:25:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][200/1251] eta 0:05:53 lr 0.000039 time 0.2547 (0.3364) loss 3.3783 (3.0362) grad_norm 3.2112 (3.1166) [2021-04-16 18:25:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][210/1251] eta 0:05:47 lr 0.000039 time 0.2754 (0.3334) loss 3.2357 (3.0310) grad_norm 2.8035 (3.1109) [2021-04-16 18:25:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][220/1251] eta 0:05:41 lr 0.000039 time 0.2668 (0.3308) loss 2.7125 (3.0389) grad_norm 2.7272 (3.1127) [2021-04-16 18:25:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][230/1251] eta 0:05:35 lr 0.000039 time 0.2548 (0.3284) loss 3.5604 (3.0352) grad_norm 3.8978 (3.1216) [2021-04-16 18:25:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][240/1251] eta 0:05:29 lr 0.000039 time 0.2746 (0.3262) loss 3.4908 (3.0289) grad_norm 4.7570 (3.1255) [2021-04-16 18:25:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][250/1251] eta 0:05:25 lr 0.000039 time 0.2823 (0.3248) loss 3.3863 (3.0255) grad_norm 2.7215 (3.1249) [2021-04-16 18:25:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][260/1251] eta 0:05:19 lr 0.000039 time 0.2557 (0.3228) loss 3.8720 (3.0278) grad_norm 2.9337 (3.1147) [2021-04-16 18:25:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][270/1251] eta 0:05:14 lr 0.000039 time 0.2547 (0.3209) loss 3.3390 (3.0342) grad_norm 2.9084 (3.1055) [2021-04-16 18:25:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][280/1251] eta 0:05:10 lr 0.000039 time 0.2821 (0.3195) loss 2.2703 (3.0394) grad_norm 3.7816 (3.1044) [2021-04-16 18:25:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][290/1251] eta 0:05:05 lr 0.000039 time 0.2595 (0.3179) loss 2.3892 (3.0374) grad_norm 3.0318 (3.1034) [2021-04-16 18:25:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][300/1251] eta 0:05:00 lr 0.000039 time 0.2967 (0.3165) loss 2.8355 (3.0352) grad_norm 3.6765 (3.1020) [2021-04-16 18:25:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][310/1251] eta 0:04:56 lr 0.000039 time 0.2778 (0.3151) loss 3.1596 (3.0399) grad_norm 3.0969 (3.1071) [2021-04-16 18:25:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][320/1251] eta 0:04:52 lr 0.000039 time 0.2728 (0.3143) loss 2.4533 (3.0452) grad_norm 2.6481 (3.0986) [2021-04-16 18:25:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][330/1251] eta 0:04:48 lr 0.000039 time 0.2570 (0.3131) loss 1.9486 (3.0381) grad_norm 3.0351 (3.0955) [2021-04-16 18:25:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][340/1251] eta 0:04:44 lr 0.000039 time 0.2549 (0.3126) loss 1.8345 (3.0387) grad_norm 2.7791 (3.0979) [2021-04-16 18:25:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][350/1251] eta 0:04:40 lr 0.000039 time 0.2563 (0.3114) loss 2.4274 (3.0379) grad_norm 2.7497 (3.0948) [2021-04-16 18:25:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][360/1251] eta 0:04:37 lr 0.000039 time 0.2844 (0.3110) loss 3.0164 (3.0333) grad_norm 2.9770 (3.0912) [2021-04-16 18:25:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][370/1251] eta 0:04:33 lr 0.000039 time 0.2860 (0.3107) loss 3.3841 (3.0246) grad_norm 3.3614 (3.0887) [2021-04-16 18:25:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][380/1251] eta 0:04:29 lr 0.000039 time 0.2843 (0.3098) loss 2.6083 (3.0246) grad_norm 2.9075 (3.0887) [2021-04-16 18:25:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][390/1251] eta 0:04:26 lr 0.000039 time 0.2453 (0.3090) loss 3.3982 (3.0267) grad_norm 3.2375 (3.0867) [2021-04-16 18:25:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][400/1251] eta 0:04:22 lr 0.000039 time 0.2833 (0.3083) loss 3.3291 (3.0269) grad_norm 2.9532 (3.0879) [2021-04-16 18:26:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][410/1251] eta 0:04:18 lr 0.000039 time 0.2771 (0.3075) loss 2.3778 (3.0209) grad_norm 2.8060 (3.0911) [2021-04-16 18:26:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][420/1251] eta 0:04:14 lr 0.000039 time 0.2460 (0.3067) loss 3.3725 (3.0248) grad_norm 3.0848 (3.0940) [2021-04-16 18:26:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][430/1251] eta 0:04:11 lr 0.000039 time 0.2736 (0.3060) loss 1.8700 (3.0272) grad_norm 3.8874 (3.0957) [2021-04-16 18:26:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][440/1251] eta 0:04:07 lr 0.000039 time 0.2656 (0.3054) loss 3.7492 (3.0305) grad_norm 3.4479 (3.0949) [2021-04-16 18:26:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][450/1251] eta 0:04:04 lr 0.000039 time 0.2680 (0.3047) loss 3.3823 (3.0315) grad_norm 2.9606 (3.0985) [2021-04-16 18:26:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][460/1251] eta 0:04:00 lr 0.000039 time 0.2734 (0.3040) loss 3.3160 (3.0279) grad_norm 3.2373 (3.0985) [2021-04-16 18:26:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][470/1251] eta 0:03:57 lr 0.000039 time 0.2940 (0.3036) loss 3.5030 (3.0271) grad_norm 2.8331 (3.1002) [2021-04-16 18:26:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][480/1251] eta 0:03:53 lr 0.000039 time 0.2788 (0.3031) loss 3.1258 (3.0239) grad_norm 3.0263 (3.0955) [2021-04-16 18:26:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][490/1251] eta 0:03:50 lr 0.000039 time 0.2793 (0.3025) loss 3.8204 (3.0228) grad_norm 2.8618 (3.0929) [2021-04-16 18:26:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][500/1251] eta 0:03:46 lr 0.000039 time 0.2797 (0.3020) loss 3.2997 (3.0254) grad_norm 2.7803 (3.0914) [2021-04-16 18:26:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][510/1251] eta 0:03:43 lr 0.000039 time 0.2567 (0.3015) loss 3.4223 (3.0311) grad_norm 3.5016 (3.0910) [2021-04-16 18:26:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][520/1251] eta 0:03:40 lr 0.000039 time 0.2925 (0.3011) loss 3.3005 (3.0331) grad_norm 3.0885 (3.0889) [2021-04-16 18:26:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][530/1251] eta 0:03:36 lr 0.000039 time 0.3092 (0.3006) loss 2.6022 (3.0301) grad_norm 2.8159 (3.0898) [2021-04-16 18:26:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][540/1251] eta 0:03:33 lr 0.000039 time 0.2671 (0.3001) loss 3.1015 (3.0246) grad_norm 2.5302 (3.0895) [2021-04-16 18:26:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][550/1251] eta 0:03:30 lr 0.000038 time 0.2734 (0.2997) loss 3.5661 (3.0279) grad_norm 3.0481 (3.0968) [2021-04-16 18:26:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][560/1251] eta 0:03:26 lr 0.000038 time 0.2697 (0.2992) loss 3.8768 (3.0274) grad_norm 3.1964 (3.0970) [2021-04-16 18:26:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][570/1251] eta 0:03:23 lr 0.000038 time 0.2710 (0.2990) loss 2.2955 (3.0286) grad_norm 2.8624 (3.0956) [2021-04-16 18:26:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][580/1251] eta 0:03:20 lr 0.000038 time 0.2670 (0.2986) loss 3.4666 (3.0305) grad_norm 3.1828 (3.0930) [2021-04-16 18:26:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][590/1251] eta 0:03:17 lr 0.000038 time 0.2694 (0.2982) loss 2.7222 (3.0289) grad_norm 2.9163 (3.0882) [2021-04-16 18:26:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][600/1251] eta 0:03:13 lr 0.000038 time 0.2985 (0.2978) loss 2.9223 (3.0292) grad_norm 2.8831 (3.0901) [2021-04-16 18:26:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][610/1251] eta 0:03:10 lr 0.000038 time 0.2658 (0.2974) loss 2.6340 (3.0314) grad_norm 3.0403 (3.0894) [2021-04-16 18:27:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][620/1251] eta 0:03:07 lr 0.000038 time 0.2617 (0.2971) loss 3.5460 (3.0291) grad_norm 3.0403 (3.0895) [2021-04-16 18:27:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][630/1251] eta 0:03:04 lr 0.000038 time 0.2824 (0.2968) loss 3.6530 (3.0246) grad_norm 3.3908 (3.0892) [2021-04-16 18:27:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][640/1251] eta 0:03:01 lr 0.000038 time 0.2639 (0.2966) loss 2.9807 (3.0278) grad_norm 3.4056 (3.0902) [2021-04-16 18:27:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][650/1251] eta 0:02:58 lr 0.000038 time 0.2793 (0.2962) loss 3.1017 (3.0281) grad_norm 3.2519 (3.0916) [2021-04-16 18:27:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][660/1251] eta 0:02:54 lr 0.000038 time 0.2540 (0.2959) loss 3.1882 (3.0249) grad_norm 3.0148 (3.0962) [2021-04-16 18:27:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][670/1251] eta 0:02:51 lr 0.000038 time 0.2626 (0.2956) loss 3.1403 (3.0254) grad_norm 2.9642 (3.0955) [2021-04-16 18:27:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][680/1251] eta 0:02:48 lr 0.000038 time 0.2775 (0.2952) loss 2.9559 (3.0223) grad_norm 2.9079 (inf) [2021-04-16 18:27:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][690/1251] eta 0:02:45 lr 0.000038 time 0.2891 (0.2950) loss 2.1023 (3.0216) grad_norm 2.8644 (inf) [2021-04-16 18:27:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][700/1251] eta 0:02:42 lr 0.000038 time 0.2640 (0.2946) loss 3.1099 (3.0202) grad_norm 2.7740 (inf) [2021-04-16 18:27:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][710/1251] eta 0:02:39 lr 0.000038 time 0.2998 (0.2944) loss 3.2748 (3.0242) grad_norm 2.7098 (inf) [2021-04-16 18:27:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][720/1251] eta 0:02:36 lr 0.000038 time 0.4023 (0.2943) loss 3.7863 (3.0258) grad_norm 3.0441 (inf) [2021-04-16 18:27:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][730/1251] eta 0:02:33 lr 0.000038 time 0.4313 (0.2945) loss 2.4207 (3.0208) grad_norm 3.2880 (inf) [2021-04-16 18:27:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][740/1251] eta 0:02:30 lr 0.000038 time 0.3053 (0.2943) loss 3.7736 (3.0227) grad_norm 2.5935 (inf) [2021-04-16 18:27:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][750/1251] eta 0:02:27 lr 0.000038 time 0.2721 (0.2940) loss 3.4839 (3.0263) grad_norm 3.2685 (inf) [2021-04-16 18:27:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][760/1251] eta 0:02:24 lr 0.000038 time 0.2547 (0.2938) loss 2.8252 (3.0237) grad_norm 3.4480 (inf) [2021-04-16 18:27:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][770/1251] eta 0:02:21 lr 0.000038 time 0.2646 (0.2935) loss 3.3050 (3.0265) grad_norm 3.1752 (inf) [2021-04-16 18:27:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][780/1251] eta 0:02:18 lr 0.000038 time 0.2682 (0.2934) loss 3.4030 (3.0313) grad_norm 4.2140 (inf) [2021-04-16 18:27:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][790/1251] eta 0:02:15 lr 0.000038 time 0.2759 (0.2932) loss 3.0144 (3.0319) grad_norm 2.8626 (inf) [2021-04-16 18:27:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][800/1251] eta 0:02:12 lr 0.000038 time 0.2682 (0.2929) loss 2.8040 (3.0337) grad_norm 3.3855 (inf) [2021-04-16 18:27:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][810/1251] eta 0:02:09 lr 0.000038 time 0.2889 (0.2928) loss 3.2031 (3.0357) grad_norm 2.9485 (inf) [2021-04-16 18:27:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][820/1251] eta 0:02:06 lr 0.000038 time 0.2721 (0.2925) loss 3.0441 (3.0348) grad_norm 2.9356 (inf) [2021-04-16 18:27:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][830/1251] eta 0:02:03 lr 0.000038 time 0.2719 (0.2924) loss 2.6797 (3.0351) grad_norm 2.8697 (inf) [2021-04-16 18:28:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][840/1251] eta 0:02:00 lr 0.000038 time 0.2775 (0.2923) loss 3.2461 (3.0364) grad_norm 2.4315 (inf) [2021-04-16 18:28:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][850/1251] eta 0:01:57 lr 0.000038 time 0.3002 (0.2921) loss 3.3785 (3.0377) grad_norm 2.8169 (inf) [2021-04-16 18:28:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][860/1251] eta 0:01:54 lr 0.000038 time 0.2877 (0.2920) loss 2.6315 (3.0383) grad_norm 2.9255 (inf) [2021-04-16 18:28:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][870/1251] eta 0:01:51 lr 0.000038 time 0.2729 (0.2919) loss 3.2445 (3.0397) grad_norm 2.8121 (inf) [2021-04-16 18:28:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][880/1251] eta 0:01:48 lr 0.000038 time 0.2761 (0.2917) loss 3.2993 (3.0401) grad_norm 2.7469 (inf) [2021-04-16 18:28:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][890/1251] eta 0:01:45 lr 0.000038 time 0.2597 (0.2915) loss 2.5341 (3.0374) grad_norm 3.1408 (inf) [2021-04-16 18:28:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][900/1251] eta 0:01:42 lr 0.000038 time 0.3043 (0.2915) loss 3.3803 (3.0395) grad_norm 2.5505 (inf) [2021-04-16 18:28:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][910/1251] eta 0:01:39 lr 0.000038 time 0.2693 (0.2913) loss 3.0470 (3.0390) grad_norm 2.3573 (inf) [2021-04-16 18:28:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][920/1251] eta 0:01:36 lr 0.000038 time 0.2636 (0.2912) loss 3.4606 (3.0406) grad_norm 2.8242 (inf) [2021-04-16 18:28:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][930/1251] eta 0:01:33 lr 0.000038 time 0.2718 (0.2912) loss 3.3208 (3.0384) grad_norm 3.1337 (inf) [2021-04-16 18:28:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][940/1251] eta 0:01:30 lr 0.000038 time 0.3235 (0.2912) loss 2.9228 (3.0376) grad_norm 2.9728 (inf) [2021-04-16 18:28:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][950/1251] eta 0:01:27 lr 0.000038 time 0.2545 (0.2910) loss 3.0827 (3.0384) grad_norm 2.7808 (inf) [2021-04-16 18:28:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][960/1251] eta 0:01:24 lr 0.000038 time 0.2803 (0.2908) loss 2.9146 (3.0401) grad_norm 3.5828 (inf) [2021-04-16 18:28:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][970/1251] eta 0:01:21 lr 0.000038 time 0.2708 (0.2907) loss 2.7400 (3.0403) grad_norm 2.7227 (inf) [2021-04-16 18:28:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][980/1251] eta 0:01:18 lr 0.000038 time 0.2829 (0.2906) loss 3.5135 (3.0403) grad_norm 3.1073 (inf) [2021-04-16 18:28:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][990/1251] eta 0:01:15 lr 0.000038 time 0.2639 (0.2904) loss 3.4767 (3.0392) grad_norm 2.5526 (inf) [2021-04-16 18:28:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][1000/1251] eta 0:01:12 lr 0.000038 time 0.2434 (0.2903) loss 2.6863 (3.0379) grad_norm 2.6437 (inf) [2021-04-16 18:28:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][1010/1251] eta 0:01:09 lr 0.000038 time 0.2669 (0.2902) loss 3.2667 (3.0365) grad_norm 2.9396 (inf) [2021-04-16 18:28:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][1020/1251] eta 0:01:06 lr 0.000038 time 0.2965 (0.2900) loss 3.3596 (3.0382) grad_norm 3.3184 (inf) [2021-04-16 18:28:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][1030/1251] eta 0:01:04 lr 0.000038 time 0.2583 (0.2899) loss 3.6780 (3.0394) grad_norm 2.9674 (inf) [2021-04-16 18:28:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][1040/1251] eta 0:01:01 lr 0.000038 time 0.2922 (0.2898) loss 3.1357 (3.0378) grad_norm 3.1515 (inf) [2021-04-16 18:29:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][1050/1251] eta 0:00:58 lr 0.000038 time 0.2733 (0.2896) loss 2.7786 (3.0382) grad_norm 2.9884 (inf) [2021-04-16 18:29:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][1060/1251] eta 0:00:55 lr 0.000038 time 0.2784 (0.2895) loss 3.6177 (3.0362) grad_norm 3.7790 (inf) [2021-04-16 18:29:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][1070/1251] eta 0:00:52 lr 0.000038 time 0.3102 (0.2894) loss 2.8892 (3.0386) grad_norm 2.5218 (inf) [2021-04-16 18:29:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][1080/1251] eta 0:00:49 lr 0.000038 time 0.2721 (0.2893) loss 2.4211 (3.0388) grad_norm 2.6304 (inf) [2021-04-16 18:29:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][1090/1251] eta 0:00:46 lr 0.000038 time 0.3041 (0.2892) loss 2.7073 (3.0408) grad_norm 3.0440 (inf) [2021-04-16 18:29:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][1100/1251] eta 0:00:43 lr 0.000038 time 0.2899 (0.2891) loss 2.3019 (3.0397) grad_norm 2.7406 (inf) [2021-04-16 18:29:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][1110/1251] eta 0:00:40 lr 0.000038 time 0.2915 (0.2889) loss 2.7840 (3.0380) grad_norm 3.3585 (inf) [2021-04-16 18:29:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][1120/1251] eta 0:00:37 lr 0.000038 time 0.2529 (0.2888) loss 2.6363 (3.0373) grad_norm 3.3110 (inf) [2021-04-16 18:29:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][1130/1251] eta 0:00:34 lr 0.000038 time 0.2800 (0.2887) loss 3.4856 (3.0374) grad_norm 3.2578 (inf) [2021-04-16 18:29:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][1140/1251] eta 0:00:32 lr 0.000038 time 0.2689 (0.2885) loss 3.2378 (3.0380) grad_norm 2.8404 (inf) [2021-04-16 18:29:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][1150/1251] eta 0:00:29 lr 0.000038 time 0.2573 (0.2885) loss 3.2882 (3.0380) grad_norm 2.5968 (inf) [2021-04-16 18:29:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][1160/1251] eta 0:00:26 lr 0.000038 time 0.2910 (0.2883) loss 2.9907 (3.0363) grad_norm 3.4640 (inf) [2021-04-16 18:29:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][1170/1251] eta 0:00:23 lr 0.000038 time 0.2723 (0.2882) loss 2.6250 (3.0348) grad_norm 3.1777 (inf) [2021-04-16 18:29:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][1180/1251] eta 0:00:20 lr 0.000038 time 0.2799 (0.2881) loss 2.8345 (3.0345) grad_norm 2.8736 (inf) [2021-04-16 18:29:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][1190/1251] eta 0:00:17 lr 0.000038 time 0.3863 (0.2881) loss 2.5726 (3.0336) grad_norm 3.1297 (inf) [2021-04-16 18:29:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][1200/1251] eta 0:00:14 lr 0.000038 time 0.2819 (0.2880) loss 3.1969 (3.0352) grad_norm 3.5911 (inf) [2021-04-16 18:29:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][1210/1251] eta 0:00:11 lr 0.000038 time 0.2897 (0.2879) loss 1.5366 (3.0306) grad_norm 2.4957 (inf) [2021-04-16 18:29:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][1220/1251] eta 0:00:08 lr 0.000038 time 0.2625 (0.2879) loss 1.9664 (3.0298) grad_norm 3.3240 (inf) [2021-04-16 18:29:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][1230/1251] eta 0:00:06 lr 0.000038 time 0.2689 (0.2878) loss 3.0512 (3.0295) grad_norm 3.2456 (inf) [2021-04-16 18:29:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][1240/1251] eta 0:00:03 lr 0.000038 time 0.2480 (0.2876) loss 3.3670 (3.0293) grad_norm 3.0364 (inf) [2021-04-16 18:29:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [267/300][1250/1251] eta 0:00:00 lr 0.000038 time 0.2504 (0.2873) loss 3.4773 (3.0295) grad_norm 3.2051 (inf) [2021-04-16 18:30:15 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 267 training takes 0:06:19 [2021-04-16 18:30:15 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_267.pth saving...... [2021-04-16 18:30:41 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_267.pth saved !!! [2021-04-16 18:30:42 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.107 (1.107) Loss 0.7875 (0.7875) Acc@1 82.422 (82.422) Acc@5 95.410 (95.410) [2021-04-16 18:30:43 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.128 (0.223) Loss 0.8554 (0.8166) Acc@1 80.371 (81.188) Acc@5 94.727 (95.552) [2021-04-16 18:30:46 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.367 (0.244) Loss 0.8062 (0.8207) Acc@1 81.250 (81.115) Acc@5 95.996 (95.452) [2021-04-16 18:30:48 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.160 (0.219) Loss 0.8238 (0.8170) Acc@1 79.102 (81.042) Acc@5 96.191 (95.552) [2021-04-16 18:30:50 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.217) Loss 0.9246 (0.8258) Acc@1 80.078 (80.907) Acc@5 93.848 (95.413) [2021-04-16 18:31:06 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.880 Acc@5 95.434 [2021-04-16 18:31:06 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 80.9% [2021-04-16 18:31:06 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 80.89% [2021-04-16 18:31:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][0/1251] eta 6:02:53 lr 0.000038 time 17.4051 (17.4051) loss 3.5022 (3.5022) grad_norm 2.9950 (2.9950) [2021-04-16 18:31:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][10/1251] eta 0:37:49 lr 0.000038 time 0.2701 (1.8288) loss 3.2669 (2.9229) grad_norm 2.7762 (3.1410) [2021-04-16 18:31:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][20/1251] eta 0:22:22 lr 0.000038 time 0.2945 (1.0903) loss 3.6062 (2.9272) grad_norm 2.7392 (3.1605) [2021-04-16 18:31:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][30/1251] eta 0:16:49 lr 0.000037 time 0.2731 (0.8272) loss 3.4260 (3.0248) grad_norm 3.2178 (3.1078) [2021-04-16 18:31:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][40/1251] eta 0:14:06 lr 0.000037 time 0.3578 (0.6986) loss 2.8642 (2.9875) grad_norm 3.5569 (3.1353) [2021-04-16 18:31:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][50/1251] eta 0:12:19 lr 0.000037 time 0.2599 (0.6155) loss 3.3355 (2.9701) grad_norm 3.2619 (3.1535) [2021-04-16 18:31:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][60/1251] eta 0:11:06 lr 0.000037 time 0.2666 (0.5598) loss 3.2629 (3.0157) grad_norm 3.0547 (3.1381) [2021-04-16 18:31:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][70/1251] eta 0:10:14 lr 0.000037 time 0.2844 (0.5199) loss 2.7297 (2.9956) grad_norm 2.6888 (3.1155) [2021-04-16 18:31:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][80/1251] eta 0:09:33 lr 0.000037 time 0.2798 (0.4899) loss 2.7762 (2.9627) grad_norm 2.7218 (3.0951) [2021-04-16 18:31:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][90/1251] eta 0:09:01 lr 0.000037 time 0.2900 (0.4665) loss 2.2430 (2.9563) grad_norm 3.0962 (3.0917) [2021-04-16 18:31:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][100/1251] eta 0:08:35 lr 0.000037 time 0.2979 (0.4483) loss 3.0965 (2.9491) grad_norm 2.7822 (3.0947) [2021-04-16 18:31:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][110/1251] eta 0:08:14 lr 0.000037 time 0.3068 (0.4335) loss 3.1454 (2.9557) grad_norm 2.8437 (3.0851) [2021-04-16 18:31:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][120/1251] eta 0:07:56 lr 0.000037 time 0.2883 (0.4216) loss 2.6517 (2.9475) grad_norm 2.7691 (3.0764) [2021-04-16 18:31:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][130/1251] eta 0:07:40 lr 0.000037 time 0.2922 (0.4108) loss 3.6171 (2.9637) grad_norm 3.6672 (3.0805) [2021-04-16 18:32:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][140/1251] eta 0:07:28 lr 0.000037 time 0.2744 (0.4034) loss 2.6398 (2.9714) grad_norm 3.6876 (3.0882) [2021-04-16 18:32:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][150/1251] eta 0:07:15 lr 0.000037 time 0.4002 (0.3959) loss 2.2174 (2.9683) grad_norm 3.4693 (3.0969) [2021-04-16 18:32:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][160/1251] eta 0:07:03 lr 0.000037 time 0.2645 (0.3886) loss 3.6429 (2.9759) grad_norm 2.5690 (3.0965) [2021-04-16 18:32:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][170/1251] eta 0:06:53 lr 0.000037 time 0.2698 (0.3822) loss 3.4166 (2.9765) grad_norm 2.6568 (3.0915) [2021-04-16 18:32:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][180/1251] eta 0:06:43 lr 0.000037 time 0.2853 (0.3766) loss 3.3575 (2.9777) grad_norm 3.4548 (3.0914) [2021-04-16 18:32:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][190/1251] eta 0:06:34 lr 0.000037 time 0.2954 (0.3717) loss 3.2253 (2.9963) grad_norm 2.6180 (3.0806) [2021-04-16 18:32:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][200/1251] eta 0:06:25 lr 0.000037 time 0.2744 (0.3669) loss 2.0778 (2.9881) grad_norm 5.2324 (3.0926) [2021-04-16 18:32:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][210/1251] eta 0:06:17 lr 0.000037 time 0.3056 (0.3627) loss 3.1913 (2.9870) grad_norm 3.5503 (3.0935) [2021-04-16 18:32:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][220/1251] eta 0:06:10 lr 0.000037 time 0.2744 (0.3591) loss 2.0401 (2.9896) grad_norm 3.3466 (3.0996) [2021-04-16 18:32:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][230/1251] eta 0:06:02 lr 0.000037 time 0.2922 (0.3554) loss 3.1486 (2.9913) grad_norm 3.0582 (3.1063) [2021-04-16 18:32:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][240/1251] eta 0:05:56 lr 0.000037 time 0.2934 (0.3527) loss 3.1687 (2.9984) grad_norm 3.5767 (3.1037) [2021-04-16 18:32:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][250/1251] eta 0:05:50 lr 0.000037 time 0.3079 (0.3497) loss 3.2956 (2.9892) grad_norm 2.6483 (3.1016) [2021-04-16 18:32:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][260/1251] eta 0:05:43 lr 0.000037 time 0.2644 (0.3469) loss 3.7430 (2.9970) grad_norm 2.7311 (3.0994) [2021-04-16 18:32:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][270/1251] eta 0:05:37 lr 0.000037 time 0.2730 (0.3443) loss 3.1956 (3.0061) grad_norm 3.0584 (3.1126) [2021-04-16 18:32:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][280/1251] eta 0:05:32 lr 0.000037 time 0.2677 (0.3421) loss 3.0068 (3.0035) grad_norm 2.9069 (3.1107) [2021-04-16 18:32:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][290/1251] eta 0:05:26 lr 0.000037 time 0.2495 (0.3398) loss 2.8043 (3.0031) grad_norm 2.8416 (3.1055) [2021-04-16 18:32:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][300/1251] eta 0:05:21 lr 0.000037 time 0.3001 (0.3379) loss 3.6001 (3.0129) grad_norm 3.0397 (3.1067) [2021-04-16 18:32:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][310/1251] eta 0:05:16 lr 0.000037 time 0.2923 (0.3359) loss 3.2392 (3.0098) grad_norm 3.0258 (3.1067) [2021-04-16 18:32:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][320/1251] eta 0:05:10 lr 0.000037 time 0.2634 (0.3340) loss 3.7704 (3.0153) grad_norm 2.8465 (3.1043) [2021-04-16 18:32:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][330/1251] eta 0:05:06 lr 0.000037 time 0.2575 (0.3324) loss 2.5020 (3.0069) grad_norm 3.0946 (3.0996) [2021-04-16 18:32:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][340/1251] eta 0:05:01 lr 0.000037 time 0.2433 (0.3308) loss 3.7004 (3.0140) grad_norm 3.0996 (3.1019) [2021-04-16 18:33:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][350/1251] eta 0:04:56 lr 0.000037 time 0.2781 (0.3294) loss 2.4357 (3.0118) grad_norm 3.2834 (3.1035) [2021-04-16 18:33:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][360/1251] eta 0:04:52 lr 0.000037 time 0.2552 (0.3283) loss 3.2139 (3.0186) grad_norm 3.0816 (3.1044) [2021-04-16 18:33:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][370/1251] eta 0:04:48 lr 0.000037 time 0.2839 (0.3272) loss 1.8929 (3.0222) grad_norm 3.0301 (3.1064) [2021-04-16 18:33:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][380/1251] eta 0:04:43 lr 0.000037 time 0.3135 (0.3260) loss 2.2946 (3.0195) grad_norm 2.9921 (3.1105) [2021-04-16 18:33:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][390/1251] eta 0:04:40 lr 0.000037 time 0.3133 (0.3252) loss 3.4053 (3.0196) grad_norm 3.0972 (3.1155) [2021-04-16 18:33:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][400/1251] eta 0:04:35 lr 0.000037 time 0.2996 (0.3240) loss 3.5048 (3.0213) grad_norm 3.2393 (3.1169) [2021-04-16 18:33:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][410/1251] eta 0:04:31 lr 0.000037 time 0.2604 (0.3228) loss 3.7007 (3.0223) grad_norm 2.6450 (3.1185) [2021-04-16 18:33:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][420/1251] eta 0:04:27 lr 0.000037 time 0.2958 (0.3223) loss 2.7034 (3.0226) grad_norm 2.5039 (3.1172) [2021-04-16 18:33:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][430/1251] eta 0:04:23 lr 0.000037 time 0.2881 (0.3214) loss 2.5255 (3.0205) grad_norm 3.5626 (3.1175) [2021-04-16 18:33:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][440/1251] eta 0:04:19 lr 0.000037 time 0.2743 (0.3204) loss 3.1357 (3.0201) grad_norm 2.7844 (3.1155) [2021-04-16 18:33:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][450/1251] eta 0:04:15 lr 0.000037 time 0.2754 (0.3194) loss 3.7561 (3.0255) grad_norm 3.0340 (3.1163) [2021-04-16 18:33:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][460/1251] eta 0:04:12 lr 0.000037 time 0.2898 (0.3187) loss 3.3740 (3.0276) grad_norm 2.9314 (3.1164) [2021-04-16 18:33:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][470/1251] eta 0:04:08 lr 0.000037 time 0.2714 (0.3178) loss 2.9968 (3.0299) grad_norm 2.6338 (3.1222) [2021-04-16 18:33:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][480/1251] eta 0:04:04 lr 0.000037 time 0.2696 (0.3169) loss 3.1134 (3.0264) grad_norm 2.7297 (3.1235) [2021-04-16 18:33:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][490/1251] eta 0:04:00 lr 0.000037 time 0.2710 (0.3161) loss 2.3523 (3.0298) grad_norm 2.9594 (3.1245) [2021-04-16 18:33:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][500/1251] eta 0:03:56 lr 0.000037 time 0.3107 (0.3153) loss 3.4757 (3.0297) grad_norm 3.1089 (3.1228) [2021-04-16 18:33:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][510/1251] eta 0:03:53 lr 0.000037 time 0.2706 (0.3146) loss 3.4339 (3.0317) grad_norm 3.1138 (3.1238) [2021-04-16 18:33:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][520/1251] eta 0:03:49 lr 0.000037 time 0.2737 (0.3140) loss 3.0867 (3.0356) grad_norm 3.1135 (3.1237) [2021-04-16 18:33:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][530/1251] eta 0:03:45 lr 0.000037 time 0.2902 (0.3134) loss 3.4835 (3.0333) grad_norm 5.7482 (3.1276) [2021-04-16 18:33:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][540/1251] eta 0:03:42 lr 0.000037 time 0.2656 (0.3127) loss 3.3386 (3.0315) grad_norm 2.6782 (3.1252) [2021-04-16 18:33:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][550/1251] eta 0:03:38 lr 0.000037 time 0.4165 (0.3123) loss 2.8725 (3.0383) grad_norm 2.9723 (3.1244) [2021-04-16 18:34:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][560/1251] eta 0:03:35 lr 0.000037 time 0.2500 (0.3117) loss 3.2396 (3.0400) grad_norm 3.0142 (3.1220) [2021-04-16 18:34:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][570/1251] eta 0:03:32 lr 0.000037 time 0.2622 (0.3115) loss 3.2429 (3.0398) grad_norm 2.9331 (3.1186) [2021-04-16 18:34:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][580/1251] eta 0:03:28 lr 0.000037 time 0.2588 (0.3112) loss 3.5292 (3.0416) grad_norm 3.0700 (3.1178) [2021-04-16 18:34:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][590/1251] eta 0:03:25 lr 0.000037 time 0.3281 (0.3107) loss 3.3501 (3.0443) grad_norm 2.6334 (3.1183) [2021-04-16 18:34:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][600/1251] eta 0:03:21 lr 0.000037 time 0.2659 (0.3101) loss 3.3380 (3.0433) grad_norm 3.4140 (3.1178) [2021-04-16 18:34:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][610/1251] eta 0:03:18 lr 0.000037 time 0.2824 (0.3098) loss 3.4661 (3.0427) grad_norm 4.0573 (3.1204) [2021-04-16 18:34:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][620/1251] eta 0:03:15 lr 0.000037 time 0.2751 (0.3096) loss 3.4629 (3.0423) grad_norm 2.7650 (3.1203) [2021-04-16 18:34:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][630/1251] eta 0:03:11 lr 0.000037 time 0.2797 (0.3091) loss 3.4813 (3.0445) grad_norm 3.1687 (3.1217) [2021-04-16 18:34:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][640/1251] eta 0:03:08 lr 0.000037 time 0.2492 (0.3085) loss 3.0811 (3.0459) grad_norm 2.9431 (3.1188) [2021-04-16 18:34:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][650/1251] eta 0:03:05 lr 0.000037 time 0.2685 (0.3080) loss 3.4229 (3.0443) grad_norm 2.9691 (3.1191) [2021-04-16 18:34:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][660/1251] eta 0:03:01 lr 0.000037 time 0.2806 (0.3075) loss 3.6555 (3.0469) grad_norm 2.8123 (3.1173) [2021-04-16 18:34:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][670/1251] eta 0:02:58 lr 0.000037 time 0.2809 (0.3072) loss 2.9099 (3.0483) grad_norm 3.1158 (3.1150) [2021-04-16 18:34:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][680/1251] eta 0:02:55 lr 0.000037 time 0.2636 (0.3066) loss 3.1165 (3.0518) grad_norm 3.2144 (3.1150) [2021-04-16 18:34:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][690/1251] eta 0:02:51 lr 0.000037 time 0.2570 (0.3064) loss 3.0333 (3.0539) grad_norm 2.8850 (3.1121) [2021-04-16 18:34:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][700/1251] eta 0:02:48 lr 0.000037 time 0.2817 (0.3061) loss 2.6265 (3.0533) grad_norm 2.7422 (3.1110) [2021-04-16 18:34:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][710/1251] eta 0:02:45 lr 0.000037 time 0.2955 (0.3057) loss 2.6460 (3.0526) grad_norm 2.7463 (3.1136) [2021-04-16 18:34:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][720/1251] eta 0:02:42 lr 0.000037 time 0.2716 (0.3057) loss 3.5965 (3.0506) grad_norm 3.1512 (3.1160) [2021-04-16 18:34:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][730/1251] eta 0:02:39 lr 0.000037 time 0.2415 (0.3055) loss 2.6681 (3.0473) grad_norm 3.0342 (3.1132) [2021-04-16 18:34:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][740/1251] eta 0:02:35 lr 0.000037 time 0.3020 (0.3051) loss 2.7635 (3.0460) grad_norm 2.4366 (3.1116) [2021-04-16 18:34:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][750/1251] eta 0:02:32 lr 0.000037 time 0.2565 (0.3047) loss 3.0186 (3.0485) grad_norm 3.3293 (3.1115) [2021-04-16 18:34:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][760/1251] eta 0:02:29 lr 0.000037 time 0.3043 (0.3043) loss 3.3134 (3.0415) grad_norm 2.8078 (3.1094) [2021-04-16 18:35:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][770/1251] eta 0:02:26 lr 0.000036 time 0.2568 (0.3040) loss 3.5012 (3.0412) grad_norm 2.8662 (3.1096) [2021-04-16 18:35:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][780/1251] eta 0:02:23 lr 0.000036 time 0.2956 (0.3036) loss 3.4615 (3.0419) grad_norm 3.0605 (3.1100) [2021-04-16 18:35:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][790/1251] eta 0:02:19 lr 0.000036 time 0.2607 (0.3033) loss 3.5195 (3.0450) grad_norm 2.8123 (3.1092) [2021-04-16 18:35:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][800/1251] eta 0:02:16 lr 0.000036 time 0.2621 (0.3029) loss 2.7477 (3.0424) grad_norm 2.8571 (3.1097) [2021-04-16 18:35:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][810/1251] eta 0:02:13 lr 0.000036 time 0.2655 (0.3028) loss 3.6863 (3.0441) grad_norm 4.2247 (3.1149) [2021-04-16 18:35:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][820/1251] eta 0:02:10 lr 0.000036 time 0.2749 (0.3024) loss 3.0235 (3.0451) grad_norm 2.8135 (3.1133) [2021-04-16 18:35:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][830/1251] eta 0:02:07 lr 0.000036 time 0.2945 (0.3021) loss 3.3985 (3.0469) grad_norm 2.8611 (3.1169) [2021-04-16 18:35:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][840/1251] eta 0:02:04 lr 0.000036 time 0.2649 (0.3019) loss 2.4051 (3.0467) grad_norm 3.0470 (3.1172) [2021-04-16 18:35:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][850/1251] eta 0:02:00 lr 0.000036 time 0.2804 (0.3017) loss 3.6614 (3.0489) grad_norm 3.4466 (3.1188) [2021-04-16 18:35:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][860/1251] eta 0:01:57 lr 0.000036 time 0.2859 (0.3015) loss 2.3340 (3.0454) grad_norm 2.6356 (3.1187) [2021-04-16 18:35:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][870/1251] eta 0:01:54 lr 0.000036 time 0.2939 (0.3012) loss 3.3096 (3.0433) grad_norm 4.3177 (3.1196) [2021-04-16 18:35:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][880/1251] eta 0:01:51 lr 0.000036 time 0.2616 (0.3010) loss 2.5855 (3.0418) grad_norm 3.2059 (3.1222) [2021-04-16 18:35:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][890/1251] eta 0:01:48 lr 0.000036 time 0.2453 (0.3007) loss 3.0420 (3.0407) grad_norm 2.9285 (3.1222) [2021-04-16 18:35:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][900/1251] eta 0:01:45 lr 0.000036 time 0.2891 (0.3006) loss 2.1353 (3.0384) grad_norm 3.3227 (3.1218) [2021-04-16 18:35:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][910/1251] eta 0:01:42 lr 0.000036 time 0.2685 (0.3003) loss 3.4781 (3.0357) grad_norm 2.8922 (3.1219) [2021-04-16 18:35:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][920/1251] eta 0:01:39 lr 0.000036 time 0.2746 (0.3000) loss 3.0539 (3.0340) grad_norm 3.3564 (3.1229) [2021-04-16 18:35:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][930/1251] eta 0:01:36 lr 0.000036 time 0.2729 (0.2997) loss 2.2012 (3.0321) grad_norm 2.8622 (3.1234) [2021-04-16 18:35:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][940/1251] eta 0:01:33 lr 0.000036 time 0.2673 (0.2996) loss 3.2578 (3.0323) grad_norm 2.9977 (3.1236) [2021-04-16 18:35:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][950/1251] eta 0:01:30 lr 0.000036 time 0.2744 (0.2994) loss 3.1132 (3.0338) grad_norm 3.1381 (3.1264) [2021-04-16 18:35:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][960/1251] eta 0:01:27 lr 0.000036 time 0.2895 (0.2992) loss 2.2745 (3.0343) grad_norm 3.3014 (3.1271) [2021-04-16 18:35:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][970/1251] eta 0:01:24 lr 0.000036 time 0.2801 (0.2991) loss 3.6100 (3.0355) grad_norm 3.0216 (3.1271) [2021-04-16 18:35:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][980/1251] eta 0:01:20 lr 0.000036 time 0.2704 (0.2989) loss 3.1236 (3.0368) grad_norm 2.8777 (3.1261) [2021-04-16 18:36:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][990/1251] eta 0:01:17 lr 0.000036 time 0.2696 (0.2988) loss 2.9066 (3.0371) grad_norm 3.0130 (3.1259) [2021-04-16 18:36:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][1000/1251] eta 0:01:14 lr 0.000036 time 0.2991 (0.2986) loss 2.1954 (3.0358) grad_norm 2.6952 (3.1268) [2021-04-16 18:36:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][1010/1251] eta 0:01:11 lr 0.000036 time 0.2875 (0.2984) loss 3.9199 (3.0366) grad_norm 3.2067 (3.1260) [2021-04-16 18:36:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][1020/1251] eta 0:01:08 lr 0.000036 time 0.3130 (0.2983) loss 3.2386 (3.0369) grad_norm 2.7753 (3.1245) [2021-04-16 18:36:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][1030/1251] eta 0:01:05 lr 0.000036 time 0.2713 (0.2980) loss 3.0191 (3.0367) grad_norm 2.8661 (3.1259) [2021-04-16 18:36:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][1040/1251] eta 0:01:02 lr 0.000036 time 0.2710 (0.2978) loss 3.1921 (3.0382) grad_norm 3.4980 (3.1275) [2021-04-16 18:36:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][1050/1251] eta 0:00:59 lr 0.000036 time 0.2863 (0.2976) loss 2.2054 (3.0381) grad_norm 3.0633 (3.1272) [2021-04-16 18:36:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][1060/1251] eta 0:00:56 lr 0.000036 time 0.2794 (0.2975) loss 2.2623 (3.0357) grad_norm 3.1200 (3.1274) [2021-04-16 18:36:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][1070/1251] eta 0:00:53 lr 0.000036 time 0.2977 (0.2973) loss 3.5157 (3.0354) grad_norm 3.1609 (3.1278) [2021-04-16 18:36:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][1080/1251] eta 0:00:50 lr 0.000036 time 0.2778 (0.2972) loss 3.2923 (3.0348) grad_norm 3.5252 (3.1253) [2021-04-16 18:36:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][1090/1251] eta 0:00:47 lr 0.000036 time 0.2934 (0.2970) loss 3.4766 (3.0339) grad_norm 2.8224 (3.1229) [2021-04-16 18:36:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][1100/1251] eta 0:00:44 lr 0.000036 time 0.2979 (0.2970) loss 3.6279 (3.0338) grad_norm 3.6932 (3.1233) [2021-04-16 18:36:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][1110/1251] eta 0:00:41 lr 0.000036 time 0.2999 (0.2968) loss 2.3369 (3.0313) grad_norm 2.8321 (3.1226) [2021-04-16 18:36:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][1120/1251] eta 0:00:38 lr 0.000036 time 0.2624 (0.2966) loss 3.1407 (3.0306) grad_norm 2.9601 (3.1240) [2021-04-16 18:36:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][1130/1251] eta 0:00:35 lr 0.000036 time 0.2629 (0.2966) loss 3.1785 (3.0303) grad_norm 3.6099 (3.1271) [2021-04-16 18:36:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][1140/1251] eta 0:00:32 lr 0.000036 time 0.2756 (0.2964) loss 2.7161 (3.0285) grad_norm 2.7987 (3.1278) [2021-04-16 18:36:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][1150/1251] eta 0:00:29 lr 0.000036 time 0.2744 (0.2964) loss 2.1865 (3.0279) grad_norm 2.9814 (3.1272) [2021-04-16 18:36:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][1160/1251] eta 0:00:26 lr 0.000036 time 0.3088 (0.2964) loss 2.5296 (3.0277) grad_norm 3.1609 (3.1315) [2021-04-16 18:36:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][1170/1251] eta 0:00:24 lr 0.000036 time 0.2758 (0.2964) loss 3.2779 (3.0263) grad_norm 3.0437 (3.1313) [2021-04-16 18:36:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][1180/1251] eta 0:00:21 lr 0.000036 time 0.2651 (0.2962) loss 3.0360 (3.0277) grad_norm 2.9104 (3.1326) [2021-04-16 18:36:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][1190/1251] eta 0:00:18 lr 0.000036 time 0.3443 (0.2961) loss 1.8906 (3.0235) grad_norm 3.0654 (3.1324) [2021-04-16 18:37:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][1200/1251] eta 0:00:15 lr 0.000036 time 0.2654 (0.2960) loss 2.6815 (3.0227) grad_norm 2.6419 (3.1322) [2021-04-16 18:37:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][1210/1251] eta 0:00:12 lr 0.000036 time 0.2709 (0.2959) loss 3.4503 (3.0213) grad_norm 3.0120 (3.1319) [2021-04-16 18:37:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][1220/1251] eta 0:00:09 lr 0.000036 time 0.2530 (0.2958) loss 2.1700 (3.0196) grad_norm 2.7115 (3.1320) [2021-04-16 18:37:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][1230/1251] eta 0:00:06 lr 0.000036 time 0.2627 (0.2957) loss 3.1768 (3.0203) grad_norm 3.2559 (3.1368) [2021-04-16 18:37:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][1240/1251] eta 0:00:03 lr 0.000036 time 0.2483 (0.2955) loss 2.9201 (3.0181) grad_norm 2.8814 (3.1377) [2021-04-16 18:37:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [268/300][1250/1251] eta 0:00:00 lr 0.000036 time 0.2484 (0.2951) loss 3.1390 (3.0175) grad_norm 3.3027 (3.1358) [2021-04-16 18:37:21 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 268 training takes 0:06:15 [2021-04-16 18:37:21 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_268.pth saving...... [2021-04-16 18:37:40 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_268.pth saved !!! [2021-04-16 18:37:41 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.108 (1.108) Loss 0.9139 (0.9139) Acc@1 79.980 (79.980) Acc@5 94.238 (94.238) [2021-04-16 18:37:42 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.198 (0.206) Loss 0.9634 (0.8376) Acc@1 77.930 (81.081) Acc@5 94.141 (95.224) [2021-04-16 18:37:45 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.122 (0.243) Loss 0.7947 (0.8263) Acc@1 81.348 (80.934) Acc@5 95.215 (95.401) [2021-04-16 18:37:47 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.150 (0.226) Loss 0.8004 (0.8332) Acc@1 82.031 (80.910) Acc@5 96.387 (95.344) [2021-04-16 18:37:49 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.218) Loss 0.8111 (0.8339) Acc@1 80.469 (80.936) Acc@5 96.289 (95.360) [2021-04-16 18:38:03 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.986 Acc@5 95.384 [2021-04-16 18:38:03 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 81.0% [2021-04-16 18:38:03 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 80.99% [2021-04-16 18:38:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][0/1251] eta 6:19:49 lr 0.000036 time 18.2170 (18.2170) loss 3.4254 (3.4254) grad_norm 3.1745 (3.1745) [2021-04-16 18:38:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][10/1251] eta 0:39:24 lr 0.000036 time 0.2859 (1.9050) loss 3.4118 (3.2830) grad_norm 2.5401 (3.4012) [2021-04-16 18:38:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][20/1251] eta 0:23:11 lr 0.000036 time 0.2763 (1.1307) loss 3.6379 (3.1977) grad_norm 2.8601 (3.2568) [2021-04-16 18:38:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][30/1251] eta 0:17:23 lr 0.000036 time 0.2758 (0.8548) loss 3.2122 (3.1082) grad_norm 3.5231 (3.2382) [2021-04-16 18:38:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][40/1251] eta 0:14:23 lr 0.000036 time 0.2558 (0.7132) loss 2.6508 (3.1054) grad_norm 2.7991 (3.2124) [2021-04-16 18:38:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][50/1251] eta 0:12:34 lr 0.000036 time 0.2652 (0.6285) loss 3.8507 (3.0717) grad_norm 2.5801 (3.1520) [2021-04-16 18:38:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][60/1251] eta 0:11:20 lr 0.000036 time 0.2706 (0.5712) loss 3.1640 (3.0845) grad_norm 2.8711 (3.1377) [2021-04-16 18:38:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][70/1251] eta 0:10:26 lr 0.000036 time 0.2925 (0.5304) loss 3.2772 (3.1007) grad_norm 2.8919 (3.1287) [2021-04-16 18:38:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][80/1251] eta 0:09:44 lr 0.000036 time 0.2883 (0.4989) loss 3.6980 (3.1021) grad_norm 2.8782 (3.1345) [2021-04-16 18:38:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][90/1251] eta 0:09:10 lr 0.000036 time 0.2752 (0.4744) loss 3.4087 (3.0941) grad_norm 2.8186 (3.1241) [2021-04-16 18:38:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][100/1251] eta 0:08:44 lr 0.000036 time 0.2877 (0.4559) loss 3.6583 (3.0781) grad_norm 2.7512 (3.1309) [2021-04-16 18:38:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][110/1251] eta 0:08:22 lr 0.000036 time 0.2789 (0.4402) loss 2.3230 (3.0435) grad_norm 3.1751 (3.1428) [2021-04-16 18:38:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][120/1251] eta 0:08:02 lr 0.000036 time 0.2715 (0.4267) loss 2.8426 (3.0476) grad_norm 3.1596 (3.1444) [2021-04-16 18:38:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][130/1251] eta 0:07:45 lr 0.000036 time 0.2673 (0.4150) loss 1.8687 (3.0315) grad_norm 3.3179 (3.1365) [2021-04-16 18:39:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][140/1251] eta 0:07:30 lr 0.000036 time 0.2725 (0.4052) loss 3.4481 (3.0542) grad_norm 2.9365 (3.1493) [2021-04-16 18:39:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][150/1251] eta 0:07:16 lr 0.000036 time 0.2845 (0.3966) loss 2.7822 (3.0495) grad_norm 2.8303 (3.1444) [2021-04-16 18:39:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][160/1251] eta 0:07:05 lr 0.000036 time 0.3012 (0.3902) loss 3.1444 (3.0486) grad_norm 3.1027 (3.1466) [2021-04-16 18:39:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][170/1251] eta 0:06:54 lr 0.000036 time 0.2888 (0.3835) loss 3.7276 (3.0426) grad_norm 2.9718 (3.1365) [2021-04-16 18:39:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][180/1251] eta 0:06:45 lr 0.000036 time 0.2814 (0.3785) loss 2.5951 (3.0340) grad_norm 3.7597 (3.1375) [2021-04-16 18:39:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][190/1251] eta 0:06:35 lr 0.000036 time 0.2604 (0.3729) loss 3.0816 (3.0276) grad_norm 4.5333 (3.1442) [2021-04-16 18:39:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][200/1251] eta 0:06:27 lr 0.000036 time 0.2643 (0.3684) loss 3.6375 (3.0163) grad_norm 2.8557 (3.1441) [2021-04-16 18:39:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][210/1251] eta 0:06:18 lr 0.000036 time 0.2816 (0.3641) loss 2.5791 (3.0222) grad_norm 2.9812 (3.1413) [2021-04-16 18:39:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][220/1251] eta 0:06:11 lr 0.000036 time 0.3004 (0.3605) loss 3.2056 (3.0120) grad_norm 3.4915 (3.1489) [2021-04-16 18:39:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][230/1251] eta 0:06:04 lr 0.000036 time 0.2822 (0.3574) loss 2.2786 (3.0034) grad_norm 3.1051 (3.1476) [2021-04-16 18:39:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][240/1251] eta 0:05:57 lr 0.000036 time 0.2726 (0.3539) loss 2.6065 (2.9978) grad_norm 3.4851 (3.1501) [2021-04-16 18:39:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][250/1251] eta 0:05:51 lr 0.000036 time 0.2755 (0.3507) loss 2.8428 (2.9920) grad_norm 3.4625 (3.1530) [2021-04-16 18:39:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][260/1251] eta 0:05:44 lr 0.000036 time 0.2681 (0.3480) loss 2.4798 (2.9952) grad_norm 3.2457 (3.1506) [2021-04-16 18:39:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][270/1251] eta 0:05:38 lr 0.000035 time 0.2765 (0.3453) loss 1.9766 (2.9913) grad_norm 3.0573 (3.1453) [2021-04-16 18:39:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][280/1251] eta 0:05:33 lr 0.000035 time 0.2996 (0.3430) loss 3.6984 (2.9927) grad_norm 3.2726 (3.1479) [2021-04-16 18:39:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][290/1251] eta 0:05:27 lr 0.000035 time 0.2707 (0.3406) loss 2.1128 (2.9934) grad_norm 3.2324 (3.1492) [2021-04-16 18:39:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][300/1251] eta 0:05:21 lr 0.000035 time 0.2687 (0.3384) loss 2.3019 (2.9930) grad_norm 2.6997 (3.1431) [2021-04-16 18:39:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][310/1251] eta 0:05:16 lr 0.000035 time 0.3209 (0.3367) loss 1.9912 (2.9883) grad_norm 3.3783 (3.1489) [2021-04-16 18:39:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][320/1251] eta 0:05:11 lr 0.000035 time 0.2456 (0.3347) loss 2.9951 (2.9862) grad_norm 2.7789 (3.1505) [2021-04-16 18:39:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][330/1251] eta 0:05:06 lr 0.000035 time 0.2865 (0.3330) loss 3.5540 (2.9870) grad_norm 3.0952 (3.1524) [2021-04-16 18:39:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][340/1251] eta 0:05:02 lr 0.000035 time 0.2669 (0.3316) loss 2.6818 (2.9825) grad_norm 3.3683 (3.1502) [2021-04-16 18:39:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][350/1251] eta 0:04:57 lr 0.000035 time 0.2897 (0.3305) loss 2.2503 (2.9890) grad_norm 2.8858 (3.1440) [2021-04-16 18:40:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][360/1251] eta 0:04:53 lr 0.000035 time 0.2628 (0.3298) loss 2.8940 (2.9877) grad_norm 3.9419 (3.1483) [2021-04-16 18:40:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][370/1251] eta 0:04:49 lr 0.000035 time 0.2889 (0.3283) loss 3.3687 (2.9923) grad_norm 3.4287 (3.1468) [2021-04-16 18:40:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][380/1251] eta 0:04:44 lr 0.000035 time 0.3014 (0.3272) loss 2.7316 (2.9885) grad_norm 3.1627 (3.1493) [2021-04-16 18:40:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][390/1251] eta 0:04:40 lr 0.000035 time 0.2486 (0.3263) loss 2.7726 (2.9928) grad_norm 2.9126 (3.1524) [2021-04-16 18:40:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][400/1251] eta 0:04:36 lr 0.000035 time 0.2779 (0.3252) loss 2.8355 (2.9899) grad_norm 2.9484 (3.1535) [2021-04-16 18:40:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][410/1251] eta 0:04:32 lr 0.000035 time 0.2646 (0.3240) loss 3.6791 (2.9971) grad_norm 3.2021 (3.1536) [2021-04-16 18:40:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][420/1251] eta 0:04:28 lr 0.000035 time 0.2883 (0.3228) loss 1.8133 (2.9931) grad_norm 2.9498 (3.1465) [2021-04-16 18:40:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][430/1251] eta 0:04:24 lr 0.000035 time 0.2874 (0.3217) loss 2.6678 (2.9971) grad_norm 2.7985 (3.1413) [2021-04-16 18:40:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][440/1251] eta 0:04:20 lr 0.000035 time 0.2916 (0.3207) loss 3.3438 (2.9901) grad_norm 2.9972 (3.1410) [2021-04-16 18:40:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][450/1251] eta 0:04:16 lr 0.000035 time 0.2770 (0.3196) loss 2.6804 (2.9856) grad_norm 3.4048 (3.1458) [2021-04-16 18:40:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][460/1251] eta 0:04:12 lr 0.000035 time 0.2979 (0.3190) loss 2.6364 (2.9812) grad_norm 2.9069 (3.1412) [2021-04-16 18:40:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][470/1251] eta 0:04:08 lr 0.000035 time 0.2605 (0.3181) loss 2.3063 (2.9791) grad_norm 2.5753 (3.1377) [2021-04-16 18:40:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][480/1251] eta 0:04:04 lr 0.000035 time 0.2440 (0.3172) loss 3.3876 (2.9807) grad_norm 3.4257 (3.1381) [2021-04-16 18:40:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][490/1251] eta 0:04:01 lr 0.000035 time 0.2961 (0.3167) loss 3.1373 (2.9831) grad_norm 3.2057 (3.1346) [2021-04-16 18:40:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][500/1251] eta 0:03:57 lr 0.000035 time 0.2717 (0.3158) loss 3.3791 (2.9824) grad_norm 3.1339 (3.1294) [2021-04-16 18:40:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][510/1251] eta 0:03:53 lr 0.000035 time 0.2798 (0.3151) loss 3.3772 (2.9791) grad_norm 3.2263 (3.1252) [2021-04-16 18:40:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][520/1251] eta 0:03:49 lr 0.000035 time 0.2590 (0.3145) loss 2.1734 (2.9753) grad_norm 2.5063 (3.1223) [2021-04-16 18:40:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][530/1251] eta 0:03:46 lr 0.000035 time 0.2819 (0.3138) loss 3.1238 (2.9734) grad_norm 2.7142 (3.1205) [2021-04-16 18:40:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][540/1251] eta 0:03:42 lr 0.000035 time 0.2836 (0.3134) loss 3.6581 (2.9749) grad_norm 3.0139 (3.1199) [2021-04-16 18:40:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][550/1251] eta 0:03:39 lr 0.000035 time 0.3031 (0.3129) loss 2.1435 (2.9780) grad_norm 3.4656 (3.1210) [2021-04-16 18:40:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][560/1251] eta 0:03:35 lr 0.000035 time 0.2766 (0.3123) loss 2.1674 (2.9833) grad_norm 3.1726 (3.1196) [2021-04-16 18:41:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][570/1251] eta 0:03:32 lr 0.000035 time 0.2729 (0.3119) loss 3.1209 (2.9858) grad_norm 3.2657 (3.1168) [2021-04-16 18:41:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][580/1251] eta 0:03:29 lr 0.000035 time 0.2772 (0.3117) loss 3.1079 (2.9828) grad_norm 3.0044 (3.1179) [2021-04-16 18:41:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][590/1251] eta 0:03:25 lr 0.000035 time 0.2864 (0.3112) loss 2.8981 (2.9793) grad_norm 3.9902 (3.1194) [2021-04-16 18:41:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][600/1251] eta 0:03:22 lr 0.000035 time 0.2725 (0.3106) loss 3.3405 (2.9788) grad_norm 3.2474 (3.1197) [2021-04-16 18:41:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][610/1251] eta 0:03:18 lr 0.000035 time 0.2838 (0.3102) loss 3.6219 (2.9751) grad_norm 2.8415 (3.1170) [2021-04-16 18:41:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][620/1251] eta 0:03:15 lr 0.000035 time 0.2655 (0.3097) loss 3.3752 (2.9733) grad_norm 2.8531 (3.1170) [2021-04-16 18:41:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][630/1251] eta 0:03:12 lr 0.000035 time 0.2834 (0.3093) loss 3.5644 (2.9797) grad_norm 5.0167 (3.1202) [2021-04-16 18:41:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][640/1251] eta 0:03:08 lr 0.000035 time 0.2541 (0.3087) loss 2.8338 (2.9805) grad_norm 3.0034 (3.1184) [2021-04-16 18:41:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][650/1251] eta 0:03:05 lr 0.000035 time 0.2885 (0.3083) loss 3.7700 (2.9825) grad_norm 3.9569 (3.1195) [2021-04-16 18:41:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][660/1251] eta 0:03:01 lr 0.000035 time 0.2805 (0.3078) loss 3.0603 (2.9806) grad_norm 4.1713 (3.1205) [2021-04-16 18:41:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][670/1251] eta 0:02:58 lr 0.000035 time 0.3011 (0.3074) loss 3.3975 (2.9828) grad_norm 2.9110 (3.1209) [2021-04-16 18:41:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][680/1251] eta 0:02:55 lr 0.000035 time 0.2788 (0.3070) loss 4.0093 (2.9797) grad_norm 2.9758 (3.1186) [2021-04-16 18:41:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][690/1251] eta 0:02:51 lr 0.000035 time 0.2606 (0.3065) loss 3.9470 (2.9848) grad_norm 2.6509 (3.1195) [2021-04-16 18:41:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][700/1251] eta 0:02:48 lr 0.000035 time 0.2783 (0.3060) loss 3.1699 (2.9857) grad_norm 3.0487 (3.1180) [2021-04-16 18:41:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][710/1251] eta 0:02:45 lr 0.000035 time 0.3054 (0.3057) loss 2.9709 (2.9858) grad_norm 2.9024 (3.1185) [2021-04-16 18:41:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][720/1251] eta 0:02:42 lr 0.000035 time 0.2634 (0.3053) loss 2.5915 (2.9832) grad_norm 2.9438 (3.1178) [2021-04-16 18:41:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][730/1251] eta 0:02:38 lr 0.000035 time 0.3743 (0.3050) loss 2.7128 (2.9828) grad_norm 3.1437 (3.1190) [2021-04-16 18:41:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][740/1251] eta 0:02:35 lr 0.000035 time 0.3002 (0.3046) loss 3.4292 (2.9840) grad_norm 3.0582 (3.1195) [2021-04-16 18:41:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][750/1251] eta 0:02:32 lr 0.000035 time 0.2495 (0.3042) loss 2.4203 (2.9822) grad_norm 3.0474 (3.1189) [2021-04-16 18:41:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][760/1251] eta 0:02:29 lr 0.000035 time 0.2593 (0.3040) loss 2.8340 (2.9861) grad_norm 2.6759 (3.1186) [2021-04-16 18:41:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][770/1251] eta 0:02:26 lr 0.000035 time 0.2514 (0.3036) loss 2.5186 (2.9810) grad_norm 2.9074 (3.1202) [2021-04-16 18:42:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][780/1251] eta 0:02:22 lr 0.000035 time 0.2558 (0.3034) loss 2.6387 (2.9804) grad_norm 2.6380 (3.1172) [2021-04-16 18:42:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][790/1251] eta 0:02:19 lr 0.000035 time 0.2726 (0.3030) loss 2.6194 (2.9797) grad_norm 3.8568 (3.1184) [2021-04-16 18:42:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][800/1251] eta 0:02:16 lr 0.000035 time 0.2735 (0.3028) loss 3.4385 (2.9823) grad_norm 2.8487 (3.1175) [2021-04-16 18:42:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][810/1251] eta 0:02:13 lr 0.000035 time 0.2661 (0.3024) loss 2.9110 (2.9833) grad_norm 2.9785 (3.1172) [2021-04-16 18:42:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][820/1251] eta 0:02:10 lr 0.000035 time 0.3075 (0.3022) loss 3.2475 (2.9826) grad_norm 3.2257 (3.1153) [2021-04-16 18:42:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][830/1251] eta 0:02:07 lr 0.000035 time 0.2857 (0.3019) loss 3.0541 (2.9801) grad_norm 3.3867 (3.1159) [2021-04-16 18:42:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][840/1251] eta 0:02:03 lr 0.000035 time 0.2764 (0.3015) loss 2.3874 (2.9809) grad_norm 3.3063 (3.1149) [2021-04-16 18:42:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][850/1251] eta 0:02:00 lr 0.000035 time 0.2908 (0.3012) loss 2.2227 (2.9826) grad_norm 2.7964 (3.1156) [2021-04-16 18:42:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][860/1251] eta 0:01:57 lr 0.000035 time 0.2985 (0.3010) loss 2.8285 (2.9823) grad_norm 2.8631 (3.1156) [2021-04-16 18:42:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][870/1251] eta 0:01:54 lr 0.000035 time 0.2907 (0.3008) loss 3.5348 (2.9828) grad_norm 2.8956 (3.1150) [2021-04-16 18:42:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][880/1251] eta 0:01:51 lr 0.000035 time 0.2974 (0.3004) loss 2.3417 (2.9830) grad_norm 3.2819 (3.1142) [2021-04-16 18:42:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][890/1251] eta 0:01:48 lr 0.000035 time 0.2795 (0.3003) loss 3.0718 (2.9864) grad_norm 3.2590 (3.1130) [2021-04-16 18:42:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][900/1251] eta 0:01:45 lr 0.000035 time 0.2824 (0.3000) loss 3.1681 (2.9877) grad_norm 3.2679 (3.1141) [2021-04-16 18:42:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][910/1251] eta 0:01:42 lr 0.000035 time 0.2998 (0.2997) loss 4.0324 (2.9857) grad_norm 3.2752 (3.1181) [2021-04-16 18:42:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][920/1251] eta 0:01:39 lr 0.000035 time 0.2789 (0.2998) loss 3.4416 (2.9872) grad_norm 3.2329 (3.1181) [2021-04-16 18:42:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][930/1251] eta 0:01:36 lr 0.000035 time 0.2826 (0.2996) loss 2.9359 (2.9899) grad_norm 2.8042 (3.1196) [2021-04-16 18:42:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][940/1251] eta 0:01:33 lr 0.000035 time 0.2903 (0.2995) loss 3.4616 (2.9910) grad_norm 3.3377 (3.1200) [2021-04-16 18:42:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][950/1251] eta 0:01:30 lr 0.000035 time 0.2792 (0.2993) loss 3.1852 (2.9928) grad_norm 3.2286 (3.1243) [2021-04-16 18:42:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][960/1251] eta 0:01:27 lr 0.000035 time 0.2870 (0.2991) loss 3.3001 (2.9950) grad_norm 3.1394 (3.1247) [2021-04-16 18:42:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][970/1251] eta 0:01:24 lr 0.000035 time 0.2947 (0.2991) loss 3.7370 (2.9970) grad_norm 3.1876 (3.1245) [2021-04-16 18:42:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][980/1251] eta 0:01:20 lr 0.000035 time 0.2881 (0.2989) loss 3.3059 (2.9976) grad_norm 3.1062 (3.1239) [2021-04-16 18:42:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][990/1251] eta 0:01:17 lr 0.000035 time 0.3024 (0.2987) loss 3.0590 (2.9988) grad_norm 3.0358 (3.1229) [2021-04-16 18:43:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][1000/1251] eta 0:01:14 lr 0.000035 time 0.2656 (0.2985) loss 3.0441 (2.9989) grad_norm 3.2660 (3.1210) [2021-04-16 18:43:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][1010/1251] eta 0:01:11 lr 0.000035 time 0.2920 (0.2983) loss 3.5934 (3.0002) grad_norm 3.2775 (3.1211) [2021-04-16 18:43:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][1020/1251] eta 0:01:08 lr 0.000035 time 0.2597 (0.2980) loss 2.7810 (3.0022) grad_norm 3.2767 (3.1202) [2021-04-16 18:43:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][1030/1251] eta 0:01:05 lr 0.000035 time 0.2630 (0.2979) loss 3.5974 (2.9997) grad_norm 2.7513 (3.1192) [2021-04-16 18:43:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][1040/1251] eta 0:01:02 lr 0.000034 time 0.2662 (0.2977) loss 3.0464 (3.0007) grad_norm 3.5587 (3.1209) [2021-04-16 18:43:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][1050/1251] eta 0:00:59 lr 0.000034 time 0.2592 (0.2976) loss 2.2394 (3.0011) grad_norm 2.9369 (3.1212) [2021-04-16 18:43:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][1060/1251] eta 0:00:56 lr 0.000034 time 0.2883 (0.2974) loss 1.9809 (3.0007) grad_norm 2.7708 (3.1220) [2021-04-16 18:43:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][1070/1251] eta 0:00:53 lr 0.000034 time 0.2674 (0.2974) loss 3.1617 (3.0000) grad_norm 2.9355 (3.1250) [2021-04-16 18:43:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][1080/1251] eta 0:00:50 lr 0.000034 time 0.2642 (0.2972) loss 2.6961 (2.9997) grad_norm 3.9100 (3.1272) [2021-04-16 18:43:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][1090/1251] eta 0:00:47 lr 0.000034 time 0.2478 (0.2971) loss 2.8052 (2.9999) grad_norm 3.0414 (3.1266) [2021-04-16 18:43:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][1100/1251] eta 0:00:44 lr 0.000034 time 0.2789 (0.2969) loss 2.6163 (3.0002) grad_norm 2.6627 (3.1256) [2021-04-16 18:43:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][1110/1251] eta 0:00:41 lr 0.000034 time 0.2683 (0.2967) loss 2.7777 (3.0019) grad_norm 2.8166 (3.1248) [2021-04-16 18:43:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][1120/1251] eta 0:00:38 lr 0.000034 time 0.2706 (0.2967) loss 3.0570 (3.0012) grad_norm 3.0857 (3.1252) [2021-04-16 18:43:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][1130/1251] eta 0:00:35 lr 0.000034 time 0.2828 (0.2965) loss 2.5132 (2.9993) grad_norm 2.8572 (3.1244) [2021-04-16 18:43:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][1140/1251] eta 0:00:32 lr 0.000034 time 0.2850 (0.2964) loss 2.2959 (2.9995) grad_norm 3.0423 (3.1262) [2021-04-16 18:43:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][1150/1251] eta 0:00:29 lr 0.000034 time 0.2607 (0.2963) loss 3.0632 (2.9987) grad_norm 4.3711 (3.1271) [2021-04-16 18:43:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][1160/1251] eta 0:00:26 lr 0.000034 time 0.2758 (0.2962) loss 3.5588 (2.9976) grad_norm 3.0592 (3.1270) [2021-04-16 18:43:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][1170/1251] eta 0:00:23 lr 0.000034 time 0.2922 (0.2960) loss 3.2142 (2.9977) grad_norm 2.8049 (3.1266) [2021-04-16 18:43:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][1180/1251] eta 0:00:21 lr 0.000034 time 0.2673 (0.2958) loss 3.5467 (2.9973) grad_norm 3.9990 (3.1277) [2021-04-16 18:43:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][1190/1251] eta 0:00:18 lr 0.000034 time 0.2493 (0.2957) loss 2.9923 (2.9980) grad_norm 3.6415 (3.1313) [2021-04-16 18:43:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][1200/1251] eta 0:00:15 lr 0.000034 time 0.2792 (0.2955) loss 2.9408 (2.9976) grad_norm 2.8040 (3.1321) [2021-04-16 18:44:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][1210/1251] eta 0:00:12 lr 0.000034 time 0.2609 (0.2953) loss 3.1611 (2.9976) grad_norm 3.4665 (3.1313) [2021-04-16 18:44:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][1220/1251] eta 0:00:09 lr 0.000034 time 0.2419 (0.2952) loss 3.5707 (2.9967) grad_norm 3.4390 (3.1307) [2021-04-16 18:44:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][1230/1251] eta 0:00:06 lr 0.000034 time 0.2694 (0.2951) loss 3.1821 (2.9967) grad_norm 3.2711 (3.1316) [2021-04-16 18:44:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][1240/1251] eta 0:00:03 lr 0.000034 time 0.2483 (0.2949) loss 3.5407 (2.9981) grad_norm 3.4996 (3.1330) [2021-04-16 18:44:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [269/300][1250/1251] eta 0:00:00 lr 0.000034 time 0.2480 (0.2945) loss 2.7763 (2.9988) grad_norm 3.1960 (3.1345) [2021-04-16 18:44:17 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 269 training takes 0:06:13 [2021-04-16 18:44:17 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_269.pth saving...... [2021-04-16 18:44:33 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_269.pth saved !!! [2021-04-16 18:44:34 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.123 (1.123) Loss 0.8404 (0.8404) Acc@1 80.273 (80.273) Acc@5 94.629 (94.629) [2021-04-16 18:44:35 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.114 (0.228) Loss 0.8106 (0.8447) Acc@1 81.934 (80.584) Acc@5 95.215 (94.904) [2021-04-16 18:44:37 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.220 (0.219) Loss 0.8658 (0.8361) Acc@1 80.273 (80.585) Acc@5 94.727 (95.159) [2021-04-16 18:44:40 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.104 (0.222) Loss 0.8340 (0.8255) Acc@1 80.371 (80.724) Acc@5 94.531 (95.372) [2021-04-16 18:44:42 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.221) Loss 0.8717 (0.8231) Acc@1 80.469 (80.714) Acc@5 94.727 (95.427) [2021-04-16 18:45:02 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.880 Acc@5 95.434 [2021-04-16 18:45:02 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 80.9% [2021-04-16 18:45:02 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 80.99% [2021-04-16 18:45:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][0/1251] eta 8:02:17 lr 0.000034 time 23.1313 (23.1313) loss 3.5585 (3.5585) grad_norm 3.1424 (3.1424) [2021-04-16 18:45:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][10/1251] eta 0:48:36 lr 0.000034 time 0.2800 (2.3503) loss 2.7789 (2.9689) grad_norm 2.9265 (3.1706) [2021-04-16 18:45:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][20/1251] eta 0:27:59 lr 0.000034 time 0.3008 (1.3647) loss 3.5074 (3.0678) grad_norm 2.8355 (3.1260) [2021-04-16 18:45:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][30/1251] eta 0:20:37 lr 0.000034 time 0.2662 (1.0134) loss 3.6114 (3.0351) grad_norm 2.8802 (3.1977) [2021-04-16 18:45:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][40/1251] eta 0:16:48 lr 0.000034 time 0.2687 (0.8326) loss 3.0098 (3.0261) grad_norm 2.8191 (3.1487) [2021-04-16 18:45:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][50/1251] eta 0:14:29 lr 0.000034 time 0.2764 (0.7241) loss 1.8872 (3.0375) grad_norm 2.7150 (3.1000) [2021-04-16 18:45:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][60/1251] eta 0:12:57 lr 0.000034 time 0.3000 (0.6530) loss 2.4853 (3.0369) grad_norm 3.2079 (3.0972) [2021-04-16 18:45:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][70/1251] eta 0:11:48 lr 0.000034 time 0.2703 (0.6002) loss 3.6431 (3.0655) grad_norm 2.7560 (3.0925) [2021-04-16 18:45:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][80/1251] eta 0:10:55 lr 0.000034 time 0.2679 (0.5600) loss 3.3247 (3.0611) grad_norm 2.9787 (3.0649) [2021-04-16 18:45:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][90/1251] eta 0:10:13 lr 0.000034 time 0.2790 (0.5287) loss 2.2168 (3.0530) grad_norm 2.5549 (3.0544) [2021-04-16 18:45:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][100/1251] eta 0:09:39 lr 0.000034 time 0.2909 (0.5039) loss 3.7240 (3.0485) grad_norm 3.2797 (3.0655) [2021-04-16 18:45:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][110/1251] eta 0:09:13 lr 0.000034 time 0.4422 (0.4855) loss 3.3834 (3.0485) grad_norm 2.8898 (3.1004) [2021-04-16 18:45:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][120/1251] eta 0:08:50 lr 0.000034 time 0.3146 (0.4687) loss 3.4842 (3.0527) grad_norm 2.5743 (3.0884) [2021-04-16 18:46:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][130/1251] eta 0:08:29 lr 0.000034 time 0.2559 (0.4549) loss 3.4661 (3.0581) grad_norm 2.8316 (3.0858) [2021-04-16 18:46:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][140/1251] eta 0:08:12 lr 0.000034 time 0.2750 (0.4430) loss 3.3025 (3.0627) grad_norm 2.7927 (3.0690) [2021-04-16 18:46:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][150/1251] eta 0:07:56 lr 0.000034 time 0.2852 (0.4330) loss 2.8450 (3.0432) grad_norm 4.7472 (3.0962) [2021-04-16 18:46:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][160/1251] eta 0:07:41 lr 0.000034 time 0.2735 (0.4230) loss 3.3416 (3.0388) grad_norm 3.5685 (3.0969) [2021-04-16 18:46:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][170/1251] eta 0:07:28 lr 0.000034 time 0.3109 (0.4146) loss 3.1784 (3.0310) grad_norm 3.0084 (3.0980) [2021-04-16 18:46:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][180/1251] eta 0:07:16 lr 0.000034 time 0.2946 (0.4076) loss 2.8525 (3.0254) grad_norm 2.9660 (3.0968) [2021-04-16 18:46:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][190/1251] eta 0:07:05 lr 0.000034 time 0.2784 (0.4008) loss 2.1346 (3.0197) grad_norm 2.6122 (3.1035) [2021-04-16 18:46:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][200/1251] eta 0:06:54 lr 0.000034 time 0.2614 (0.3948) loss 2.3562 (3.0159) grad_norm 2.8935 (3.1196) [2021-04-16 18:46:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][210/1251] eta 0:06:45 lr 0.000034 time 0.2884 (0.3893) loss 3.2419 (3.0254) grad_norm 3.2446 (3.1149) [2021-04-16 18:46:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][220/1251] eta 0:06:36 lr 0.000034 time 0.2699 (0.3843) loss 1.9664 (3.0219) grad_norm 3.0306 (3.1095) [2021-04-16 18:46:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][230/1251] eta 0:06:27 lr 0.000034 time 0.2705 (0.3797) loss 3.5932 (3.0184) grad_norm 3.3985 (3.1054) [2021-04-16 18:46:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][240/1251] eta 0:06:19 lr 0.000034 time 0.2504 (0.3754) loss 3.5743 (3.0218) grad_norm 3.8421 (3.1214) [2021-04-16 18:46:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][250/1251] eta 0:06:11 lr 0.000034 time 0.3275 (0.3715) loss 3.4189 (3.0186) grad_norm 3.3501 (3.1216) [2021-04-16 18:46:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][260/1251] eta 0:06:04 lr 0.000034 time 0.2626 (0.3677) loss 3.2349 (3.0132) grad_norm 2.5333 (3.1223) [2021-04-16 18:46:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][270/1251] eta 0:05:57 lr 0.000034 time 0.2841 (0.3643) loss 2.2009 (3.0138) grad_norm 4.0719 (3.1260) [2021-04-16 18:46:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][280/1251] eta 0:05:50 lr 0.000034 time 0.3111 (0.3611) loss 3.5066 (3.0129) grad_norm 2.8356 (3.1276) [2021-04-16 18:46:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][290/1251] eta 0:05:44 lr 0.000034 time 0.2717 (0.3585) loss 2.4583 (3.0056) grad_norm 3.0861 (3.1272) [2021-04-16 18:46:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][300/1251] eta 0:05:38 lr 0.000034 time 0.2680 (0.3563) loss 2.3696 (3.0041) grad_norm 3.3844 (3.1217) [2021-04-16 18:46:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][310/1251] eta 0:05:32 lr 0.000034 time 0.2874 (0.3537) loss 1.9120 (3.0077) grad_norm 3.0765 (3.1225) [2021-04-16 18:46:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][320/1251] eta 0:05:27 lr 0.000034 time 0.2729 (0.3514) loss 2.6582 (3.0066) grad_norm 3.0968 (3.1234) [2021-04-16 18:46:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][330/1251] eta 0:05:21 lr 0.000034 time 0.2948 (0.3492) loss 3.4397 (3.0054) grad_norm 3.0014 (3.1238) [2021-04-16 18:47:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][340/1251] eta 0:05:16 lr 0.000034 time 0.2756 (0.3470) loss 2.9616 (3.0068) grad_norm 3.2630 (3.1233) [2021-04-16 18:47:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][350/1251] eta 0:05:11 lr 0.000034 time 0.2956 (0.3455) loss 3.3525 (3.0067) grad_norm 3.2747 (3.1242) [2021-04-16 18:47:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][360/1251] eta 0:05:06 lr 0.000034 time 0.2922 (0.3439) loss 2.7499 (3.0080) grad_norm 2.7744 (3.1243) [2021-04-16 18:47:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][370/1251] eta 0:05:01 lr 0.000034 time 0.2852 (0.3425) loss 3.2402 (3.0075) grad_norm 2.9256 (3.1205) [2021-04-16 18:47:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][380/1251] eta 0:04:56 lr 0.000034 time 0.2972 (0.3408) loss 2.1774 (3.0074) grad_norm 3.0994 (3.1226) [2021-04-16 18:47:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][390/1251] eta 0:04:51 lr 0.000034 time 0.2694 (0.3391) loss 3.0822 (3.0116) grad_norm 3.2289 (3.1216) [2021-04-16 18:47:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][400/1251] eta 0:04:47 lr 0.000034 time 0.2551 (0.3376) loss 3.1532 (3.0153) grad_norm 3.5529 (3.1252) [2021-04-16 18:47:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][410/1251] eta 0:04:42 lr 0.000034 time 0.2709 (0.3361) loss 3.1909 (3.0147) grad_norm 3.7243 (3.1286) [2021-04-16 18:47:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][420/1251] eta 0:04:38 lr 0.000034 time 0.2884 (0.3346) loss 2.0574 (3.0115) grad_norm 2.9559 (3.1302) [2021-04-16 18:47:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][430/1251] eta 0:04:33 lr 0.000034 time 0.2759 (0.3333) loss 2.4069 (3.0057) grad_norm 2.5110 (3.1299) [2021-04-16 18:47:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][440/1251] eta 0:04:29 lr 0.000034 time 0.2821 (0.3320) loss 3.3853 (2.9990) grad_norm 2.9461 (3.1304) [2021-04-16 18:47:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][450/1251] eta 0:04:24 lr 0.000034 time 0.3139 (0.3308) loss 3.1205 (2.9977) grad_norm 2.8893 (3.1332) [2021-04-16 18:47:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][460/1251] eta 0:04:20 lr 0.000034 time 0.2524 (0.3296) loss 2.6112 (2.9976) grad_norm 2.8085 (3.1398) [2021-04-16 18:47:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][470/1251] eta 0:04:16 lr 0.000034 time 0.2914 (0.3286) loss 2.2381 (2.9948) grad_norm 2.8972 (3.1406) [2021-04-16 18:47:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][480/1251] eta 0:04:12 lr 0.000034 time 0.2647 (0.3275) loss 2.8505 (2.9936) grad_norm 2.7994 (3.1480) [2021-04-16 18:47:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][490/1251] eta 0:04:08 lr 0.000034 time 0.2759 (0.3264) loss 2.3376 (2.9897) grad_norm 2.7363 (3.1467) [2021-04-16 18:47:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][500/1251] eta 0:04:04 lr 0.000034 time 0.2691 (0.3254) loss 3.1868 (2.9946) grad_norm 3.6478 (3.1546) [2021-04-16 18:47:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][510/1251] eta 0:04:00 lr 0.000034 time 0.2670 (0.3244) loss 2.5352 (2.9967) grad_norm 2.6121 (3.1540) [2021-04-16 18:47:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][520/1251] eta 0:03:56 lr 0.000034 time 0.2807 (0.3235) loss 3.4035 (2.9969) grad_norm 3.1415 (3.1535) [2021-04-16 18:47:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][530/1251] eta 0:03:52 lr 0.000034 time 0.2819 (0.3228) loss 3.5519 (3.0002) grad_norm 2.9430 (3.1526) [2021-04-16 18:47:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][540/1251] eta 0:03:48 lr 0.000034 time 0.2727 (0.3218) loss 3.4727 (3.0014) grad_norm 2.9872 (3.1504) [2021-04-16 18:47:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][550/1251] eta 0:03:45 lr 0.000034 time 0.4090 (0.3212) loss 3.4798 (3.0078) grad_norm 3.0055 (3.1561) [2021-04-16 18:48:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][560/1251] eta 0:03:41 lr 0.000034 time 0.2724 (0.3204) loss 2.5539 (3.0099) grad_norm 3.7521 (3.1590) [2021-04-16 18:48:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][570/1251] eta 0:03:37 lr 0.000034 time 0.2828 (0.3196) loss 3.2548 (3.0057) grad_norm 2.9969 (3.1595) [2021-04-16 18:48:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][580/1251] eta 0:03:34 lr 0.000033 time 0.2925 (0.3194) loss 2.1025 (3.0044) grad_norm 3.2407 (3.1591) [2021-04-16 18:48:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][590/1251] eta 0:03:30 lr 0.000033 time 0.2486 (0.3188) loss 2.3638 (3.0062) grad_norm 2.6247 (3.1579) [2021-04-16 18:48:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][600/1251] eta 0:03:27 lr 0.000033 time 0.2985 (0.3181) loss 2.9271 (3.0038) grad_norm 3.1358 (3.1569) [2021-04-16 18:48:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][610/1251] eta 0:03:23 lr 0.000033 time 0.2856 (0.3174) loss 2.0540 (2.9975) grad_norm 3.0990 (3.1554) [2021-04-16 18:48:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][620/1251] eta 0:03:19 lr 0.000033 time 0.2631 (0.3168) loss 3.3534 (2.9984) grad_norm 3.6813 (3.1558) [2021-04-16 18:48:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][630/1251] eta 0:03:16 lr 0.000033 time 0.2857 (0.3162) loss 3.3615 (2.9974) grad_norm 3.1153 (3.1555) [2021-04-16 18:48:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][640/1251] eta 0:03:12 lr 0.000033 time 0.2710 (0.3156) loss 3.1279 (2.9963) grad_norm 3.4589 (3.1547) [2021-04-16 18:48:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][650/1251] eta 0:03:09 lr 0.000033 time 0.2756 (0.3151) loss 1.8753 (2.9956) grad_norm 3.1258 (3.1524) [2021-04-16 18:48:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][660/1251] eta 0:03:05 lr 0.000033 time 0.2779 (0.3145) loss 2.7532 (2.9928) grad_norm 3.0779 (3.1523) [2021-04-16 18:48:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][670/1251] eta 0:03:02 lr 0.000033 time 0.2546 (0.3139) loss 2.5677 (2.9935) grad_norm 3.2645 (3.1535) [2021-04-16 18:48:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][680/1251] eta 0:02:58 lr 0.000033 time 0.2796 (0.3133) loss 2.5218 (2.9954) grad_norm 3.4873 (3.1548) [2021-04-16 18:48:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][690/1251] eta 0:02:55 lr 0.000033 time 0.2608 (0.3130) loss 3.5147 (2.9958) grad_norm 2.9260 (3.1553) [2021-04-16 18:48:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][700/1251] eta 0:02:52 lr 0.000033 time 0.2498 (0.3125) loss 3.1202 (2.9962) grad_norm 3.2633 (3.1533) [2021-04-16 18:48:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][710/1251] eta 0:02:48 lr 0.000033 time 0.2625 (0.3122) loss 3.3660 (2.9967) grad_norm 2.9538 (3.1528) [2021-04-16 18:48:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][720/1251] eta 0:02:45 lr 0.000033 time 0.2672 (0.3121) loss 3.6641 (2.9991) grad_norm 3.1790 (3.1523) [2021-04-16 18:48:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][730/1251] eta 0:02:42 lr 0.000033 time 0.2998 (0.3116) loss 2.9758 (2.9995) grad_norm 4.1227 (3.1514) [2021-04-16 18:48:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][740/1251] eta 0:02:39 lr 0.000033 time 0.2917 (0.3112) loss 3.4248 (3.0013) grad_norm 3.1311 (3.1636) [2021-04-16 18:48:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][750/1251] eta 0:02:35 lr 0.000033 time 0.2692 (0.3107) loss 2.4118 (3.0013) grad_norm 4.2432 (3.1653) [2021-04-16 18:48:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][760/1251] eta 0:02:32 lr 0.000033 time 0.2722 (0.3105) loss 3.2749 (3.0005) grad_norm 2.5412 (3.1662) [2021-04-16 18:49:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][770/1251] eta 0:02:29 lr 0.000033 time 0.2802 (0.3100) loss 2.7844 (3.0008) grad_norm 3.5295 (3.1691) [2021-04-16 18:49:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][780/1251] eta 0:02:25 lr 0.000033 time 0.2956 (0.3096) loss 2.7035 (3.0016) grad_norm 2.9605 (3.1716) [2021-04-16 18:49:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][790/1251] eta 0:02:22 lr 0.000033 time 0.2770 (0.3093) loss 3.4511 (2.9993) grad_norm 2.8776 (3.1710) [2021-04-16 18:49:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][800/1251] eta 0:02:19 lr 0.000033 time 0.2779 (0.3089) loss 3.8013 (2.9983) grad_norm 3.1237 (3.1693) [2021-04-16 18:49:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][810/1251] eta 0:02:16 lr 0.000033 time 0.2582 (0.3085) loss 3.8370 (2.9995) grad_norm 3.8394 (3.1684) [2021-04-16 18:49:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][820/1251] eta 0:02:12 lr 0.000033 time 0.2646 (0.3081) loss 2.0859 (2.9998) grad_norm 2.8215 (3.1670) [2021-04-16 18:49:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][830/1251] eta 0:02:09 lr 0.000033 time 0.2673 (0.3077) loss 3.4686 (3.0011) grad_norm 3.0257 (3.1646) [2021-04-16 18:49:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][840/1251] eta 0:02:06 lr 0.000033 time 0.2893 (0.3074) loss 3.6131 (3.0016) grad_norm 3.2894 (3.1632) [2021-04-16 18:49:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][850/1251] eta 0:02:03 lr 0.000033 time 0.2940 (0.3070) loss 2.8525 (3.0014) grad_norm 2.7639 (3.1617) [2021-04-16 18:49:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][860/1251] eta 0:01:59 lr 0.000033 time 0.2691 (0.3066) loss 3.2787 (3.0016) grad_norm 3.2893 (3.1620) [2021-04-16 18:49:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][870/1251] eta 0:01:56 lr 0.000033 time 0.2666 (0.3064) loss 3.0602 (3.0004) grad_norm 3.2169 (3.1647) [2021-04-16 18:49:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][880/1251] eta 0:01:53 lr 0.000033 time 0.2697 (0.3061) loss 2.8873 (2.9997) grad_norm 2.9453 (3.1614) [2021-04-16 18:49:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][890/1251] eta 0:01:50 lr 0.000033 time 0.2577 (0.3059) loss 3.2277 (2.9998) grad_norm 4.2993 (3.1611) [2021-04-16 18:49:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][900/1251] eta 0:01:47 lr 0.000033 time 0.2711 (0.3056) loss 3.3186 (3.0015) grad_norm 2.7855 (3.1613) [2021-04-16 18:49:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][910/1251] eta 0:01:44 lr 0.000033 time 0.2692 (0.3052) loss 3.0561 (3.0039) grad_norm 3.3690 (3.1620) [2021-04-16 18:49:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][920/1251] eta 0:01:40 lr 0.000033 time 0.2686 (0.3049) loss 2.6126 (3.0048) grad_norm 2.8367 (nan) [2021-04-16 18:49:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][930/1251] eta 0:01:37 lr 0.000033 time 0.2750 (0.3047) loss 3.1000 (3.0027) grad_norm 3.4098 (nan) [2021-04-16 18:49:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][940/1251] eta 0:01:34 lr 0.000033 time 0.2945 (0.3047) loss 3.1324 (3.0050) grad_norm 3.0952 (nan) [2021-04-16 18:49:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][950/1251] eta 0:01:31 lr 0.000033 time 0.2670 (0.3044) loss 2.6605 (3.0063) grad_norm 3.1330 (nan) [2021-04-16 18:49:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][960/1251] eta 0:01:28 lr 0.000033 time 0.2532 (0.3041) loss 2.3054 (3.0075) grad_norm 2.7037 (nan) [2021-04-16 18:49:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][970/1251] eta 0:01:25 lr 0.000033 time 0.2795 (0.3038) loss 3.1970 (3.0090) grad_norm 3.7764 (nan) [2021-04-16 18:50:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][980/1251] eta 0:01:22 lr 0.000033 time 0.2834 (0.3035) loss 3.6675 (3.0109) grad_norm 3.3495 (nan) [2021-04-16 18:50:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][990/1251] eta 0:01:19 lr 0.000033 time 0.2649 (0.3032) loss 2.8836 (3.0081) grad_norm 3.3328 (nan) [2021-04-16 18:50:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][1000/1251] eta 0:01:16 lr 0.000033 time 0.2725 (0.3029) loss 2.0512 (3.0071) grad_norm 2.5582 (nan) [2021-04-16 18:50:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][1010/1251] eta 0:01:12 lr 0.000033 time 0.2773 (0.3027) loss 2.7831 (3.0070) grad_norm 2.6670 (nan) [2021-04-16 18:50:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][1020/1251] eta 0:01:09 lr 0.000033 time 0.3196 (0.3024) loss 3.4186 (3.0075) grad_norm 5.6191 (nan) [2021-04-16 18:50:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][1030/1251] eta 0:01:06 lr 0.000033 time 0.2804 (0.3022) loss 3.5892 (3.0079) grad_norm 3.0897 (nan) [2021-04-16 18:50:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][1040/1251] eta 0:01:03 lr 0.000033 time 0.2828 (0.3019) loss 3.4412 (3.0064) grad_norm 3.1277 (nan) [2021-04-16 18:50:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][1050/1251] eta 0:01:00 lr 0.000033 time 0.2630 (0.3017) loss 3.5387 (3.0071) grad_norm 2.9146 (nan) [2021-04-16 18:50:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][1060/1251] eta 0:00:57 lr 0.000033 time 0.2830 (0.3015) loss 2.9854 (3.0080) grad_norm 3.2902 (nan) [2021-04-16 18:50:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][1070/1251] eta 0:00:54 lr 0.000033 time 0.2451 (0.3013) loss 3.0519 (3.0093) grad_norm 2.8392 (nan) [2021-04-16 18:50:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][1080/1251] eta 0:00:51 lr 0.000033 time 0.2678 (0.3010) loss 1.9155 (3.0057) grad_norm 2.6201 (nan) [2021-04-16 18:50:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][1090/1251] eta 0:00:48 lr 0.000033 time 0.2747 (0.3008) loss 2.9060 (3.0072) grad_norm 2.9928 (nan) [2021-04-16 18:50:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][1100/1251] eta 0:00:45 lr 0.000033 time 0.2625 (0.3006) loss 3.4145 (3.0059) grad_norm 3.2842 (nan) [2021-04-16 18:50:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][1110/1251] eta 0:00:42 lr 0.000033 time 0.2659 (0.3004) loss 2.4869 (3.0057) grad_norm 2.8949 (nan) [2021-04-16 18:50:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][1120/1251] eta 0:00:39 lr 0.000033 time 0.2791 (0.3002) loss 3.5143 (3.0073) grad_norm 2.8377 (nan) [2021-04-16 18:50:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][1130/1251] eta 0:00:36 lr 0.000033 time 0.2836 (0.3001) loss 3.4229 (3.0082) grad_norm 3.4799 (nan) [2021-04-16 18:50:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][1140/1251] eta 0:00:33 lr 0.000033 time 0.2711 (0.2999) loss 2.4460 (3.0098) grad_norm 2.9738 (nan) [2021-04-16 18:50:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][1150/1251] eta 0:00:30 lr 0.000033 time 0.2832 (0.2997) loss 3.3673 (3.0096) grad_norm 3.2744 (nan) [2021-04-16 18:50:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][1160/1251] eta 0:00:27 lr 0.000033 time 0.2645 (0.2997) loss 2.8447 (3.0097) grad_norm 3.4239 (nan) [2021-04-16 18:50:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][1170/1251] eta 0:00:24 lr 0.000033 time 0.2908 (0.2995) loss 3.3193 (3.0116) grad_norm 2.9437 (nan) [2021-04-16 18:50:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][1180/1251] eta 0:00:21 lr 0.000033 time 0.2913 (0.2993) loss 3.4109 (3.0120) grad_norm 3.1638 (nan) [2021-04-16 18:50:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][1190/1251] eta 0:00:18 lr 0.000033 time 0.2953 (0.2992) loss 2.6044 (3.0137) grad_norm 3.3176 (nan) [2021-04-16 18:51:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][1200/1251] eta 0:00:15 lr 0.000033 time 0.2950 (0.2991) loss 3.0236 (3.0127) grad_norm 2.9354 (nan) [2021-04-16 18:51:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][1210/1251] eta 0:00:12 lr 0.000033 time 0.2786 (0.2989) loss 3.0441 (3.0122) grad_norm 3.3379 (nan) [2021-04-16 18:51:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][1220/1251] eta 0:00:09 lr 0.000033 time 0.2600 (0.2987) loss 2.1735 (3.0121) grad_norm 2.7000 (nan) [2021-04-16 18:51:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][1230/1251] eta 0:00:06 lr 0.000033 time 0.2859 (0.2985) loss 2.5352 (3.0123) grad_norm 3.3328 (nan) [2021-04-16 18:51:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][1240/1251] eta 0:00:03 lr 0.000033 time 0.2481 (0.2983) loss 3.1924 (3.0113) grad_norm 3.2782 (nan) [2021-04-16 18:51:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [270/300][1250/1251] eta 0:00:00 lr 0.000033 time 0.2760 (0.2979) loss 3.6940 (3.0146) grad_norm 3.1625 (nan) [2021-04-16 18:51:19 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 270 training takes 0:06:17 [2021-04-16 18:51:19 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_270.pth saving...... [2021-04-16 18:51:34 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_270.pth saved !!! [2021-04-16 18:51:51 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 16.563 (16.563) Loss 0.8589 (0.8589) Acc@1 80.566 (80.566) Acc@5 94.531 (94.531) [2021-04-16 18:51:52 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.136 (1.622) Loss 0.7357 (0.8183) Acc@1 81.934 (80.939) Acc@5 96.777 (95.339) [2021-04-16 18:51:55 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.138 (0.981) Loss 0.7913 (0.8174) Acc@1 81.055 (80.980) Acc@5 95.801 (95.401) [2021-04-16 18:51:57 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.187 (0.731) Loss 0.8013 (0.8157) Acc@1 81.152 (80.903) Acc@5 95.312 (95.483) [2021-04-16 18:51:58 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.594) Loss 0.8535 (0.8176) Acc@1 79.785 (81.009) Acc@5 95.508 (95.482) [2021-04-16 18:52:17 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 81.000 Acc@5 95.406 [2021-04-16 18:52:17 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 81.0% [2021-04-16 18:52:17 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 81.00% [2021-04-16 18:53:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][0/1251] eta 18:04:34 lr 0.000033 time 52.0177 (52.0177) loss 3.5088 (3.5088) grad_norm 3.0498 (3.0498) [2021-04-16 18:53:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][10/1251] eta 1:42:54 lr 0.000033 time 0.2755 (4.9754) loss 3.4109 (3.2385) grad_norm 3.4436 (3.2035) [2021-04-16 18:53:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][20/1251] eta 0:56:10 lr 0.000033 time 0.2753 (2.7383) loss 2.1944 (3.1077) grad_norm 3.1784 (3.1099) [2021-04-16 18:53:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][30/1251] eta 0:39:36 lr 0.000033 time 0.2913 (1.9462) loss 3.0490 (3.1163) grad_norm 3.2483 (3.1090) [2021-04-16 18:53:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][40/1251] eta 0:31:03 lr 0.000033 time 0.2687 (1.5387) loss 3.7634 (3.0781) grad_norm 3.6305 (3.1413) [2021-04-16 18:53:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][50/1251] eta 0:25:51 lr 0.000033 time 0.2611 (1.2919) loss 2.4077 (2.9770) grad_norm 3.3410 (3.1310) [2021-04-16 18:53:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][60/1251] eta 0:22:23 lr 0.000033 time 0.2906 (1.1279) loss 1.9559 (2.9557) grad_norm 3.0892 (3.1239) [2021-04-16 18:53:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][70/1251] eta 0:19:50 lr 0.000033 time 0.2660 (1.0079) loss 2.4784 (2.9516) grad_norm 3.2871 (3.1888) [2021-04-16 18:53:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][80/1251] eta 0:17:57 lr 0.000033 time 0.2492 (0.9202) loss 3.1741 (2.9758) grad_norm 3.1198 (3.1921) [2021-04-16 18:53:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][90/1251] eta 0:16:25 lr 0.000033 time 0.2607 (0.8491) loss 3.1601 (2.9710) grad_norm 2.8062 (3.1740) [2021-04-16 18:53:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][100/1251] eta 0:15:14 lr 0.000033 time 0.2731 (0.7944) loss 3.2869 (2.9467) grad_norm 3.1777 (3.1935) [2021-04-16 18:53:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][110/1251] eta 0:14:13 lr 0.000033 time 0.2970 (0.7481) loss 3.3615 (2.9522) grad_norm 3.1443 (3.1868) [2021-04-16 18:53:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][120/1251] eta 0:13:22 lr 0.000033 time 0.2881 (0.7095) loss 3.7431 (2.9396) grad_norm 3.0877 (3.1978) [2021-04-16 18:53:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][130/1251] eta 0:12:39 lr 0.000032 time 0.2946 (0.6776) loss 3.5251 (2.9448) grad_norm 3.0242 (3.1883) [2021-04-16 18:53:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][140/1251] eta 0:12:02 lr 0.000032 time 0.2637 (0.6502) loss 3.0102 (2.9482) grad_norm 3.2447 (3.1858) [2021-04-16 18:53:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][150/1251] eta 0:11:29 lr 0.000032 time 0.2783 (0.6266) loss 2.9005 (2.9612) grad_norm 4.6664 (3.2050) [2021-04-16 18:53:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][160/1251] eta 0:10:59 lr 0.000032 time 0.2698 (0.6049) loss 3.4623 (2.9620) grad_norm 3.0852 (3.1943) [2021-04-16 18:53:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][170/1251] eta 0:10:33 lr 0.000032 time 0.2642 (0.5861) loss 2.3775 (2.9703) grad_norm 3.9667 (3.1892) [2021-04-16 18:54:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][180/1251] eta 0:10:09 lr 0.000032 time 0.2544 (0.5688) loss 3.1588 (2.9714) grad_norm 3.2852 (3.1868) [2021-04-16 18:54:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][190/1251] eta 0:09:47 lr 0.000032 time 0.2610 (0.5535) loss 2.9855 (2.9750) grad_norm 3.2507 (3.1843) [2021-04-16 18:54:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][200/1251] eta 0:09:27 lr 0.000032 time 0.2668 (0.5399) loss 2.3146 (2.9721) grad_norm 3.3089 (3.1799) [2021-04-16 18:54:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][210/1251] eta 0:09:09 lr 0.000032 time 0.2676 (0.5274) loss 2.2075 (2.9627) grad_norm 3.5900 (3.1900) [2021-04-16 18:54:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][220/1251] eta 0:08:52 lr 0.000032 time 0.2664 (0.5161) loss 3.2567 (2.9686) grad_norm 3.1730 (3.1939) [2021-04-16 18:54:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][230/1251] eta 0:08:36 lr 0.000032 time 0.2818 (0.5057) loss 1.8515 (2.9721) grad_norm 2.8718 (3.1859) [2021-04-16 18:54:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][240/1251] eta 0:08:21 lr 0.000032 time 0.2681 (0.4960) loss 3.2522 (2.9746) grad_norm 2.8846 (3.1879) [2021-04-16 18:54:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][250/1251] eta 0:08:07 lr 0.000032 time 0.2688 (0.4875) loss 2.4708 (2.9647) grad_norm 2.6689 (3.1735) [2021-04-16 18:54:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][260/1251] eta 0:07:55 lr 0.000032 time 0.2953 (0.4798) loss 3.2409 (2.9768) grad_norm 3.1814 (3.1747) [2021-04-16 18:54:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][270/1251] eta 0:07:43 lr 0.000032 time 0.2872 (0.4723) loss 2.0628 (2.9710) grad_norm 3.1672 (3.1689) [2021-04-16 18:54:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][280/1251] eta 0:07:31 lr 0.000032 time 0.3070 (0.4653) loss 3.0469 (2.9709) grad_norm 4.1591 (3.1753) [2021-04-16 18:54:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][290/1251] eta 0:07:20 lr 0.000032 time 0.2782 (0.4588) loss 2.1971 (2.9686) grad_norm 3.4723 (3.1712) [2021-04-16 18:54:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][300/1251] eta 0:07:10 lr 0.000032 time 0.2698 (0.4526) loss 3.0110 (2.9790) grad_norm 3.1230 (3.1707) [2021-04-16 18:54:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][310/1251] eta 0:07:00 lr 0.000032 time 0.2759 (0.4469) loss 3.3891 (2.9794) grad_norm 3.2719 (3.1666) [2021-04-16 18:54:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][320/1251] eta 0:06:51 lr 0.000032 time 0.2768 (0.4415) loss 2.4601 (2.9835) grad_norm 3.6812 (3.1657) [2021-04-16 18:54:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][330/1251] eta 0:06:42 lr 0.000032 time 0.2702 (0.4369) loss 3.4379 (2.9905) grad_norm 3.9691 (3.1640) [2021-04-16 18:54:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][340/1251] eta 0:06:34 lr 0.000032 time 0.2645 (0.4326) loss 3.8194 (2.9958) grad_norm 3.0577 (3.1686) [2021-04-16 18:54:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][350/1251] eta 0:06:25 lr 0.000032 time 0.2891 (0.4281) loss 3.6454 (3.0009) grad_norm 3.2097 (3.1695) [2021-04-16 18:54:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][360/1251] eta 0:06:18 lr 0.000032 time 0.2494 (0.4245) loss 3.1125 (3.0044) grad_norm 4.3641 (3.1725) [2021-04-16 18:54:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][370/1251] eta 0:06:10 lr 0.000032 time 0.3025 (0.4208) loss 3.1008 (3.0019) grad_norm 3.0747 (3.1657) [2021-04-16 18:54:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][380/1251] eta 0:06:03 lr 0.000032 time 0.2449 (0.4169) loss 2.9845 (3.0083) grad_norm 3.3904 (3.1605) [2021-04-16 18:54:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][390/1251] eta 0:05:56 lr 0.000032 time 0.2991 (0.4135) loss 2.9897 (3.0069) grad_norm 2.7529 (3.1607) [2021-04-16 18:55:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][400/1251] eta 0:05:49 lr 0.000032 time 0.2687 (0.4104) loss 2.1709 (3.0002) grad_norm 3.5278 (3.1639) [2021-04-16 18:55:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][410/1251] eta 0:05:42 lr 0.000032 time 0.2817 (0.4073) loss 1.7648 (3.0038) grad_norm 2.7730 (3.1624) [2021-04-16 18:55:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][420/1251] eta 0:05:35 lr 0.000032 time 0.2621 (0.4043) loss 3.6607 (3.0026) grad_norm 3.6489 (3.1632) [2021-04-16 18:55:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][430/1251] eta 0:05:29 lr 0.000032 time 0.2636 (0.4014) loss 3.6036 (3.0056) grad_norm 3.4824 (3.1626) [2021-04-16 18:55:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][440/1251] eta 0:05:23 lr 0.000032 time 0.2908 (0.3988) loss 2.7461 (3.0057) grad_norm 2.8669 (3.1631) [2021-04-16 18:55:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][450/1251] eta 0:05:17 lr 0.000032 time 0.2934 (0.3960) loss 2.6892 (3.0064) grad_norm 4.0237 (3.1663) [2021-04-16 18:55:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][460/1251] eta 0:05:11 lr 0.000032 time 0.2537 (0.3935) loss 3.5080 (3.0019) grad_norm 4.4856 (3.1637) [2021-04-16 18:55:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][470/1251] eta 0:05:05 lr 0.000032 time 0.3008 (0.3911) loss 3.4300 (3.0066) grad_norm 2.8283 (3.1645) [2021-04-16 18:55:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][480/1251] eta 0:04:59 lr 0.000032 time 0.2746 (0.3886) loss 3.0916 (3.0060) grad_norm 3.1620 (3.1616) [2021-04-16 18:55:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][490/1251] eta 0:04:54 lr 0.000032 time 0.2998 (0.3863) loss 3.3817 (3.0047) grad_norm 2.5619 (3.1613) [2021-04-16 18:55:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][500/1251] eta 0:04:48 lr 0.000032 time 0.2883 (0.3841) loss 3.2526 (3.0088) grad_norm 2.9938 (3.1662) [2021-04-16 18:55:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][510/1251] eta 0:04:43 lr 0.000032 time 0.2991 (0.3821) loss 2.1775 (3.0111) grad_norm 2.9809 (3.1691) [2021-04-16 18:55:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][520/1251] eta 0:04:38 lr 0.000032 time 0.3165 (0.3804) loss 3.5277 (3.0101) grad_norm 2.9945 (3.1658) [2021-04-16 18:55:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][530/1251] eta 0:04:32 lr 0.000032 time 0.2448 (0.3784) loss 3.4935 (3.0116) grad_norm 2.7815 (3.1631) [2021-04-16 18:55:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][540/1251] eta 0:04:27 lr 0.000032 time 0.2961 (0.3766) loss 3.2685 (3.0095) grad_norm 2.9984 (3.1590) [2021-04-16 18:55:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][550/1251] eta 0:04:22 lr 0.000032 time 0.2761 (0.3748) loss 3.5139 (3.0048) grad_norm 2.8398 (3.1597) [2021-04-16 18:55:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][560/1251] eta 0:04:17 lr 0.000032 time 0.2697 (0.3730) loss 2.0985 (3.0012) grad_norm 3.3704 (3.1593) [2021-04-16 18:55:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][570/1251] eta 0:04:13 lr 0.000032 time 0.2557 (0.3717) loss 3.9008 (2.9962) grad_norm 2.9029 (3.1616) [2021-04-16 18:55:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][580/1251] eta 0:04:08 lr 0.000032 time 0.2762 (0.3705) loss 2.5243 (2.9982) grad_norm 3.3561 (3.1644) [2021-04-16 18:55:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][590/1251] eta 0:04:03 lr 0.000032 time 0.2887 (0.3689) loss 3.6549 (3.0019) grad_norm 3.2984 (3.1646) [2021-04-16 18:55:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][600/1251] eta 0:03:59 lr 0.000032 time 0.2709 (0.3673) loss 3.6065 (3.0035) grad_norm 3.1202 (3.1629) [2021-04-16 18:56:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][610/1251] eta 0:03:54 lr 0.000032 time 0.2548 (0.3657) loss 3.6459 (3.0031) grad_norm 3.1853 (3.1635) [2021-04-16 18:56:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][620/1251] eta 0:03:49 lr 0.000032 time 0.2948 (0.3643) loss 2.6386 (3.0063) grad_norm 2.9578 (3.1599) [2021-04-16 18:56:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][630/1251] eta 0:03:45 lr 0.000032 time 0.2748 (0.3629) loss 3.7068 (3.0112) grad_norm 3.0856 (3.1582) [2021-04-16 18:56:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][640/1251] eta 0:03:40 lr 0.000032 time 0.2532 (0.3616) loss 3.5833 (3.0079) grad_norm 2.9684 (3.1553) [2021-04-16 18:56:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][650/1251] eta 0:03:36 lr 0.000032 time 0.3293 (0.3606) loss 3.5256 (3.0056) grad_norm 3.2804 (3.1554) [2021-04-16 18:56:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][660/1251] eta 0:03:32 lr 0.000032 time 0.2802 (0.3593) loss 2.7836 (3.0028) grad_norm 2.7304 (3.1573) [2021-04-16 18:56:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][670/1251] eta 0:03:28 lr 0.000032 time 0.2974 (0.3580) loss 2.4046 (3.0036) grad_norm 2.7452 (3.1613) [2021-04-16 18:56:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][680/1251] eta 0:03:23 lr 0.000032 time 0.2785 (0.3569) loss 3.5689 (3.0052) grad_norm 2.8341 (3.1616) [2021-04-16 18:56:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][690/1251] eta 0:03:19 lr 0.000032 time 0.2517 (0.3559) loss 2.3767 (3.0066) grad_norm 2.9362 (3.1653) [2021-04-16 18:56:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][700/1251] eta 0:03:15 lr 0.000032 time 0.4033 (0.3549) loss 3.4838 (3.0071) grad_norm 2.9520 (3.1683) [2021-04-16 18:56:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][710/1251] eta 0:03:11 lr 0.000032 time 0.2796 (0.3539) loss 2.9460 (3.0092) grad_norm 2.9548 (3.1720) [2021-04-16 18:56:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][720/1251] eta 0:03:07 lr 0.000032 time 0.3135 (0.3530) loss 3.0888 (3.0021) grad_norm 3.4116 (3.1736) [2021-04-16 18:56:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][730/1251] eta 0:03:03 lr 0.000032 time 0.2562 (0.3521) loss 3.2099 (3.0006) grad_norm 3.5587 (3.1753) [2021-04-16 18:56:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][740/1251] eta 0:02:59 lr 0.000032 time 0.2703 (0.3510) loss 3.3068 (3.0024) grad_norm 2.9799 (3.1734) [2021-04-16 18:56:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][750/1251] eta 0:02:55 lr 0.000032 time 0.2887 (0.3501) loss 2.4659 (3.0012) grad_norm 3.0853 (3.1720) [2021-04-16 18:56:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][760/1251] eta 0:02:51 lr 0.000032 time 0.2613 (0.3491) loss 2.1794 (2.9989) grad_norm 2.6599 (3.1693) [2021-04-16 18:56:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][770/1251] eta 0:02:47 lr 0.000032 time 0.2644 (0.3482) loss 2.9471 (2.9992) grad_norm 3.5269 (3.1667) [2021-04-16 18:56:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][780/1251] eta 0:02:43 lr 0.000032 time 0.2646 (0.3473) loss 3.2248 (3.0024) grad_norm 2.7869 (3.1669) [2021-04-16 18:56:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][790/1251] eta 0:02:39 lr 0.000032 time 0.2496 (0.3464) loss 3.2864 (3.0009) grad_norm 3.1973 (3.1667) [2021-04-16 18:56:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][800/1251] eta 0:02:35 lr 0.000032 time 0.2825 (0.3455) loss 2.7270 (3.0017) grad_norm 2.9153 (3.1657) [2021-04-16 18:56:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][810/1251] eta 0:02:31 lr 0.000032 time 0.2753 (0.3447) loss 3.6147 (3.0008) grad_norm 3.6937 (3.1654) [2021-04-16 18:56:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][820/1251] eta 0:02:28 lr 0.000032 time 0.2850 (0.3438) loss 3.3351 (3.0032) grad_norm 2.9807 (3.1650) [2021-04-16 18:57:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][830/1251] eta 0:02:24 lr 0.000032 time 0.2561 (0.3435) loss 3.3897 (3.0032) grad_norm 3.4712 (3.1622) [2021-04-16 18:57:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][840/1251] eta 0:02:20 lr 0.000032 time 0.3043 (0.3428) loss 2.6203 (3.0034) grad_norm 3.4233 (3.1611) [2021-04-16 18:57:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][850/1251] eta 0:02:17 lr 0.000032 time 0.2739 (0.3420) loss 2.3442 (3.0008) grad_norm 3.0076 (3.1603) [2021-04-16 18:57:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][860/1251] eta 0:02:13 lr 0.000032 time 0.2764 (0.3413) loss 2.8929 (3.0028) grad_norm 3.2304 (3.1603) [2021-04-16 18:57:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][870/1251] eta 0:02:09 lr 0.000032 time 0.2777 (0.3405) loss 2.2381 (3.0014) grad_norm 2.6195 (3.1613) [2021-04-16 18:57:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][880/1251] eta 0:02:06 lr 0.000032 time 0.2835 (0.3398) loss 3.4674 (3.0011) grad_norm 2.7535 (3.1612) [2021-04-16 18:57:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][890/1251] eta 0:02:02 lr 0.000032 time 0.2819 (0.3391) loss 2.8945 (3.0015) grad_norm 2.8787 (3.1632) [2021-04-16 18:57:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][900/1251] eta 0:01:58 lr 0.000032 time 0.2926 (0.3384) loss 3.7568 (3.0013) grad_norm 3.1762 (3.1653) [2021-04-16 18:57:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][910/1251] eta 0:01:55 lr 0.000032 time 0.2904 (0.3378) loss 2.3911 (3.0003) grad_norm 2.7487 (3.1628) [2021-04-16 18:57:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][920/1251] eta 0:01:51 lr 0.000032 time 0.2828 (0.3373) loss 2.9399 (2.9963) grad_norm 2.6466 (3.1599) [2021-04-16 18:57:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][930/1251] eta 0:01:48 lr 0.000032 time 0.2556 (0.3369) loss 3.5943 (2.9979) grad_norm 3.0705 (3.1593) [2021-04-16 18:57:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][940/1251] eta 0:01:44 lr 0.000032 time 0.2454 (0.3363) loss 3.3166 (2.9987) grad_norm 3.0290 (3.1569) [2021-04-16 18:57:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][950/1251] eta 0:01:41 lr 0.000031 time 0.3000 (0.3358) loss 2.6399 (2.9993) grad_norm 2.7561 (3.1590) [2021-04-16 18:57:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][960/1251] eta 0:01:37 lr 0.000031 time 0.3039 (0.3352) loss 3.3328 (2.9998) grad_norm 3.0078 (3.1584) [2021-04-16 18:57:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][970/1251] eta 0:01:34 lr 0.000031 time 0.2635 (0.3346) loss 3.2124 (3.0003) grad_norm 3.5226 (3.1581) [2021-04-16 18:57:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][980/1251] eta 0:01:30 lr 0.000031 time 0.2481 (0.3340) loss 3.5383 (2.9992) grad_norm 3.4785 (3.1573) [2021-04-16 18:57:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][990/1251] eta 0:01:27 lr 0.000031 time 0.2822 (0.3334) loss 3.3821 (2.9999) grad_norm 3.1762 (3.1588) [2021-04-16 18:57:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][1000/1251] eta 0:01:23 lr 0.000031 time 0.2663 (0.3329) loss 3.4265 (3.0001) grad_norm 3.9474 (3.1592) [2021-04-16 18:57:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][1010/1251] eta 0:01:20 lr 0.000031 time 0.3063 (0.3323) loss 2.7961 (3.0008) grad_norm 3.3112 (3.1584) [2021-04-16 18:57:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][1020/1251] eta 0:01:16 lr 0.000031 time 0.2507 (0.3318) loss 3.2167 (3.0013) grad_norm 2.8720 (3.1567) [2021-04-16 18:57:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][1030/1251] eta 0:01:13 lr 0.000031 time 0.2729 (0.3313) loss 3.5970 (3.0034) grad_norm 2.9385 (3.1546) [2021-04-16 18:58:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][1040/1251] eta 0:01:09 lr 0.000031 time 0.2816 (0.3309) loss 3.6996 (3.0039) grad_norm 3.0224 (3.1533) [2021-04-16 18:58:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][1050/1251] eta 0:01:06 lr 0.000031 time 0.2615 (0.3303) loss 3.2136 (3.0051) grad_norm 3.0422 (3.1526) [2021-04-16 18:58:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][1060/1251] eta 0:01:02 lr 0.000031 time 0.2665 (0.3298) loss 2.1189 (3.0032) grad_norm 3.0682 (3.1540) [2021-04-16 18:58:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][1070/1251] eta 0:00:59 lr 0.000031 time 0.2874 (0.3294) loss 2.6740 (3.0029) grad_norm 2.8580 (3.1523) [2021-04-16 18:58:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][1080/1251] eta 0:00:56 lr 0.000031 time 0.3001 (0.3289) loss 3.3254 (3.0049) grad_norm 3.4692 (3.1530) [2021-04-16 18:58:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][1090/1251] eta 0:00:52 lr 0.000031 time 0.2571 (0.3286) loss 3.0877 (3.0060) grad_norm 2.7903 (3.1520) [2021-04-16 18:58:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][1100/1251] eta 0:00:49 lr 0.000031 time 0.2583 (0.3281) loss 3.1493 (3.0072) grad_norm 3.1220 (3.1537) [2021-04-16 18:58:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][1110/1251] eta 0:00:46 lr 0.000031 time 0.2878 (0.3277) loss 3.4030 (3.0084) grad_norm 3.0090 (3.1523) [2021-04-16 18:58:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][1120/1251] eta 0:00:42 lr 0.000031 time 0.2798 (0.3273) loss 3.1197 (3.0077) grad_norm 3.0274 (3.1539) [2021-04-16 18:58:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][1130/1251] eta 0:00:39 lr 0.000031 time 0.2599 (0.3269) loss 2.0176 (3.0062) grad_norm 3.1621 (3.1530) [2021-04-16 18:58:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][1140/1251] eta 0:00:36 lr 0.000031 time 0.2568 (0.3266) loss 3.4299 (3.0052) grad_norm 3.2023 (3.1518) [2021-04-16 18:58:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][1150/1251] eta 0:00:32 lr 0.000031 time 0.2872 (0.3263) loss 2.7894 (3.0039) grad_norm 3.3562 (3.1532) [2021-04-16 18:58:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][1160/1251] eta 0:00:29 lr 0.000031 time 0.2830 (0.3261) loss 2.9612 (3.0016) grad_norm 3.5997 (3.1529) [2021-04-16 18:58:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][1170/1251] eta 0:00:26 lr 0.000031 time 0.2654 (0.3257) loss 3.1376 (3.0029) grad_norm 3.1862 (3.1538) [2021-04-16 18:58:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][1180/1251] eta 0:00:23 lr 0.000031 time 0.2682 (0.3253) loss 2.8349 (3.0040) grad_norm 4.1040 (3.1585) [2021-04-16 18:58:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][1190/1251] eta 0:00:19 lr 0.000031 time 0.2918 (0.3249) loss 2.4436 (3.0039) grad_norm 3.0291 (3.1608) [2021-04-16 18:58:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][1200/1251] eta 0:00:16 lr 0.000031 time 0.2932 (0.3245) loss 2.9909 (3.0045) grad_norm 3.0421 (3.1608) [2021-04-16 18:58:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][1210/1251] eta 0:00:13 lr 0.000031 time 0.2663 (0.3240) loss 3.0006 (3.0041) grad_norm 3.3718 (3.1630) [2021-04-16 18:58:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][1220/1251] eta 0:00:10 lr 0.000031 time 0.2885 (0.3237) loss 2.8343 (3.0045) grad_norm 2.7359 (3.1623) [2021-04-16 18:58:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][1230/1251] eta 0:00:06 lr 0.000031 time 0.2728 (0.3234) loss 3.8404 (3.0052) grad_norm 3.4593 (3.1635) [2021-04-16 18:58:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][1240/1251] eta 0:00:03 lr 0.000031 time 0.2478 (0.3229) loss 2.0453 (3.0054) grad_norm 2.9529 (3.1636) [2021-04-16 18:59:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [271/300][1250/1251] eta 0:00:00 lr 0.000031 time 0.2482 (0.3224) loss 2.2624 (3.0049) grad_norm 3.8980 (3.1646) [2021-04-16 18:59:10 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 271 training takes 0:06:53 [2021-04-16 18:59:10 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_271.pth saving...... [2021-04-16 18:59:29 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_271.pth saved !!! [2021-04-16 18:59:31 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.205 (1.205) Loss 0.8626 (0.8626) Acc@1 80.664 (80.664) Acc@5 94.727 (94.727) [2021-04-16 18:59:32 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.120 (0.226) Loss 0.7655 (0.8086) Acc@1 81.934 (81.428) Acc@5 95.801 (95.481) [2021-04-16 18:59:34 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.305 (0.226) Loss 0.8692 (0.8121) Acc@1 80.078 (81.245) Acc@5 94.629 (95.457) [2021-04-16 18:59:37 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.113 (0.233) Loss 0.7836 (0.8142) Acc@1 80.273 (81.083) Acc@5 96.680 (95.442) [2021-04-16 18:59:38 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.216) Loss 0.7630 (0.8170) Acc@1 82.324 (81.033) Acc@5 96.387 (95.496) [2021-04-16 19:00:02 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.996 Acc@5 95.474 [2021-04-16 19:00:02 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 81.0% [2021-04-16 19:00:02 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 81.00% [2021-04-16 19:00:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][0/1251] eta 2:12:41 lr 0.000031 time 6.3643 (6.3643) loss 3.6612 (3.6612) grad_norm 3.8180 (3.8180) [2021-04-16 19:00:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][10/1251] eta 0:17:04 lr 0.000031 time 0.2810 (0.8256) loss 3.5037 (3.0339) grad_norm 2.8505 (3.3260) [2021-04-16 19:00:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][20/1251] eta 0:11:34 lr 0.000031 time 0.2861 (0.5646) loss 2.5764 (2.9213) grad_norm 3.3143 (3.3414) [2021-04-16 19:00:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][30/1251] eta 0:09:34 lr 0.000031 time 0.2740 (0.4708) loss 3.3327 (2.9444) grad_norm 2.7059 (3.3644) [2021-04-16 19:00:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][40/1251] eta 0:08:33 lr 0.000031 time 0.3137 (0.4240) loss 3.2350 (2.9504) grad_norm 2.9971 (3.3052) [2021-04-16 19:00:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][50/1251] eta 0:07:53 lr 0.000031 time 0.2929 (0.3944) loss 3.4489 (2.9591) grad_norm 3.2211 (3.2349) [2021-04-16 19:00:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][60/1251] eta 0:07:25 lr 0.000031 time 0.2714 (0.3744) loss 3.2089 (2.9652) grad_norm 2.6497 (3.2075) [2021-04-16 19:00:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][70/1251] eta 0:07:06 lr 0.000031 time 0.2881 (0.3615) loss 3.0970 (2.9511) grad_norm 2.5537 (3.1863) [2021-04-16 19:00:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][80/1251] eta 0:06:51 lr 0.000031 time 0.3167 (0.3512) loss 3.3459 (2.9732) grad_norm 3.0551 (3.2235) [2021-04-16 19:00:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][90/1251] eta 0:06:37 lr 0.000031 time 0.2578 (0.3426) loss 3.2781 (2.9532) grad_norm 3.1113 (3.2156) [2021-04-16 19:00:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][100/1251] eta 0:06:26 lr 0.000031 time 0.2701 (0.3358) loss 3.4988 (2.9504) grad_norm 3.0180 (3.2223) [2021-04-16 19:00:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][110/1251] eta 0:06:16 lr 0.000031 time 0.2646 (0.3301) loss 3.2235 (2.9565) grad_norm 2.9785 (3.2085) [2021-04-16 19:00:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][120/1251] eta 0:06:10 lr 0.000031 time 0.2741 (0.3274) loss 3.2908 (2.9586) grad_norm 2.6979 (3.1859) [2021-04-16 19:00:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][130/1251] eta 0:06:02 lr 0.000031 time 0.3090 (0.3238) loss 3.3320 (2.9571) grad_norm 2.5548 (3.1595) [2021-04-16 19:00:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][140/1251] eta 0:05:58 lr 0.000031 time 0.2545 (0.3224) loss 3.3516 (2.9549) grad_norm 3.0332 (3.1434) [2021-04-16 19:00:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][150/1251] eta 0:05:51 lr 0.000031 time 0.2675 (0.3192) loss 2.9062 (2.9519) grad_norm 4.0912 (3.1476) [2021-04-16 19:00:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][160/1251] eta 0:05:45 lr 0.000031 time 0.2838 (0.3167) loss 2.1800 (2.9547) grad_norm 3.0547 (3.1424) [2021-04-16 19:00:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][170/1251] eta 0:05:40 lr 0.000031 time 0.2601 (0.3149) loss 2.7532 (2.9594) grad_norm 3.3756 (3.1432) [2021-04-16 19:00:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][180/1251] eta 0:05:35 lr 0.000031 time 0.2695 (0.3129) loss 3.2032 (2.9604) grad_norm 3.1531 (3.1515) [2021-04-16 19:01:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][190/1251] eta 0:05:29 lr 0.000031 time 0.2865 (0.3109) loss 2.9510 (2.9609) grad_norm 2.8340 (3.1561) [2021-04-16 19:01:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][200/1251] eta 0:05:24 lr 0.000031 time 0.2711 (0.3091) loss 1.9547 (2.9451) grad_norm 3.0804 (3.1665) [2021-04-16 19:01:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][210/1251] eta 0:05:20 lr 0.000031 time 0.2647 (0.3074) loss 3.2026 (2.9593) grad_norm 3.1936 (3.1643) [2021-04-16 19:01:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][220/1251] eta 0:05:15 lr 0.000031 time 0.2587 (0.3060) loss 3.3836 (2.9648) grad_norm 3.2876 (3.1560) [2021-04-16 19:01:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][230/1251] eta 0:05:10 lr 0.000031 time 0.2565 (0.3045) loss 3.0579 (2.9641) grad_norm 2.9879 (3.1649) [2021-04-16 19:01:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][240/1251] eta 0:05:06 lr 0.000031 time 0.2803 (0.3034) loss 2.6816 (2.9656) grad_norm 7.2997 (3.1917) [2021-04-16 19:01:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][250/1251] eta 0:05:02 lr 0.000031 time 0.2686 (0.3024) loss 2.9107 (2.9583) grad_norm 2.6502 (3.1905) [2021-04-16 19:01:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][260/1251] eta 0:04:58 lr 0.000031 time 0.2918 (0.3015) loss 2.6124 (2.9598) grad_norm 3.4062 (3.1891) [2021-04-16 19:01:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][270/1251] eta 0:04:54 lr 0.000031 time 0.2980 (0.3006) loss 2.3303 (2.9629) grad_norm 3.6879 (3.1967) [2021-04-16 19:01:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][280/1251] eta 0:04:51 lr 0.000031 time 0.2564 (0.2997) loss 3.5132 (2.9649) grad_norm 3.0602 (3.1965) [2021-04-16 19:01:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][290/1251] eta 0:04:47 lr 0.000031 time 0.2726 (0.2988) loss 3.4874 (2.9655) grad_norm 3.0064 (3.2167) [2021-04-16 19:01:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][300/1251] eta 0:04:43 lr 0.000031 time 0.2636 (0.2984) loss 3.4530 (2.9711) grad_norm 3.2187 (3.2144) [2021-04-16 19:01:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][310/1251] eta 0:04:39 lr 0.000031 time 0.2830 (0.2975) loss 2.9901 (2.9750) grad_norm 3.7787 (3.2099) [2021-04-16 19:01:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][320/1251] eta 0:04:36 lr 0.000031 time 0.2920 (0.2971) loss 2.2841 (2.9699) grad_norm 3.4004 (3.2045) [2021-04-16 19:01:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][330/1251] eta 0:04:33 lr 0.000031 time 0.3002 (0.2965) loss 2.9950 (2.9693) grad_norm 3.6547 (3.2112) [2021-04-16 19:01:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][340/1251] eta 0:04:29 lr 0.000031 time 0.2642 (0.2963) loss 3.4006 (2.9685) grad_norm 3.0468 (3.2118) [2021-04-16 19:01:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][350/1251] eta 0:04:27 lr 0.000031 time 0.3757 (0.2965) loss 3.4565 (2.9659) grad_norm 3.0989 (3.2092) [2021-04-16 19:01:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][360/1251] eta 0:04:24 lr 0.000031 time 0.2822 (0.2963) loss 3.4612 (2.9679) grad_norm 2.7684 (3.2057) [2021-04-16 19:01:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][370/1251] eta 0:04:20 lr 0.000031 time 0.2925 (0.2962) loss 2.9963 (2.9618) grad_norm 3.9754 (3.2084) [2021-04-16 19:01:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][380/1251] eta 0:04:17 lr 0.000031 time 0.2885 (0.2957) loss 2.3818 (2.9625) grad_norm 5.5655 (3.2145) [2021-04-16 19:01:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][390/1251] eta 0:04:14 lr 0.000031 time 0.2898 (0.2953) loss 3.1335 (2.9611) grad_norm 2.9530 (3.2118) [2021-04-16 19:02:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][400/1251] eta 0:04:11 lr 0.000031 time 0.2552 (0.2950) loss 3.5660 (2.9551) grad_norm 3.3524 (3.2191) [2021-04-16 19:02:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][410/1251] eta 0:04:07 lr 0.000031 time 0.2579 (0.2945) loss 3.5489 (2.9568) grad_norm 3.0898 (3.2193) [2021-04-16 19:02:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][420/1251] eta 0:04:04 lr 0.000031 time 0.2761 (0.2940) loss 1.8615 (2.9552) grad_norm 3.3069 (3.2199) [2021-04-16 19:02:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][430/1251] eta 0:04:01 lr 0.000031 time 0.2692 (0.2940) loss 3.7501 (2.9558) grad_norm 3.2651 (3.2197) [2021-04-16 19:02:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][440/1251] eta 0:03:58 lr 0.000031 time 0.2918 (0.2936) loss 2.9946 (2.9594) grad_norm 3.2714 (3.2207) [2021-04-16 19:02:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][450/1251] eta 0:03:54 lr 0.000031 time 0.2918 (0.2932) loss 3.2062 (2.9617) grad_norm 3.3438 (3.2184) [2021-04-16 19:02:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][460/1251] eta 0:03:51 lr 0.000031 time 0.2636 (0.2929) loss 3.3477 (2.9616) grad_norm 3.3354 (3.2207) [2021-04-16 19:02:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][470/1251] eta 0:03:48 lr 0.000031 time 0.2670 (0.2925) loss 2.9523 (2.9678) grad_norm 5.5855 (3.2275) [2021-04-16 19:02:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][480/1251] eta 0:03:45 lr 0.000031 time 0.2909 (0.2925) loss 3.0665 (2.9696) grad_norm 2.5545 (3.2220) [2021-04-16 19:02:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][490/1251] eta 0:03:42 lr 0.000031 time 0.2782 (0.2922) loss 2.7547 (2.9676) grad_norm 4.0499 (3.2192) [2021-04-16 19:02:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][500/1251] eta 0:03:39 lr 0.000031 time 0.2662 (0.2919) loss 2.8304 (2.9720) grad_norm 2.9397 (3.2164) [2021-04-16 19:02:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][510/1251] eta 0:03:36 lr 0.000031 time 0.2613 (0.2915) loss 3.2139 (2.9731) grad_norm 3.1132 (3.2168) [2021-04-16 19:02:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][520/1251] eta 0:03:32 lr 0.000031 time 0.2648 (0.2912) loss 3.0611 (2.9741) grad_norm 3.1959 (3.2144) [2021-04-16 19:02:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][530/1251] eta 0:03:29 lr 0.000030 time 0.2839 (0.2910) loss 3.4010 (2.9746) grad_norm 3.3237 (3.2111) [2021-04-16 19:02:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][540/1251] eta 0:03:26 lr 0.000030 time 0.2917 (0.2908) loss 2.9366 (2.9748) grad_norm 3.0218 (3.2121) [2021-04-16 19:02:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][550/1251] eta 0:03:23 lr 0.000030 time 0.3044 (0.2905) loss 3.2883 (2.9792) grad_norm 2.8133 (3.2073) [2021-04-16 19:02:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][560/1251] eta 0:03:20 lr 0.000030 time 0.2604 (0.2904) loss 2.3816 (2.9794) grad_norm 3.1484 (3.2064) [2021-04-16 19:02:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][570/1251] eta 0:03:17 lr 0.000030 time 0.4006 (0.2906) loss 3.3917 (2.9818) grad_norm 3.1255 (3.2012) [2021-04-16 19:02:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][580/1251] eta 0:03:14 lr 0.000030 time 0.2723 (0.2904) loss 2.1832 (2.9813) grad_norm 3.0060 (3.2008) [2021-04-16 19:02:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][590/1251] eta 0:03:11 lr 0.000030 time 0.2430 (0.2901) loss 2.1122 (2.9791) grad_norm 3.0075 (3.1961) [2021-04-16 19:02:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][600/1251] eta 0:03:08 lr 0.000030 time 0.2813 (0.2900) loss 3.1678 (2.9785) grad_norm 2.5111 (3.1908) [2021-04-16 19:02:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][610/1251] eta 0:03:05 lr 0.000030 time 0.2801 (0.2897) loss 2.3875 (2.9756) grad_norm 2.7559 (3.1901) [2021-04-16 19:03:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][620/1251] eta 0:03:02 lr 0.000030 time 0.2996 (0.2895) loss 3.3066 (2.9742) grad_norm 3.3180 (3.1868) [2021-04-16 19:03:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][630/1251] eta 0:02:59 lr 0.000030 time 0.2757 (0.2893) loss 2.1491 (2.9733) grad_norm 3.5093 (3.1865) [2021-04-16 19:03:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][640/1251] eta 0:02:56 lr 0.000030 time 0.2774 (0.2891) loss 3.5073 (2.9741) grad_norm 3.2871 (3.1855) [2021-04-16 19:03:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][650/1251] eta 0:02:53 lr 0.000030 time 0.2539 (0.2889) loss 3.2043 (2.9775) grad_norm 2.9912 (3.1825) [2021-04-16 19:03:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][660/1251] eta 0:02:50 lr 0.000030 time 0.2830 (0.2887) loss 2.4937 (2.9825) grad_norm 3.4640 (3.1797) [2021-04-16 19:03:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][670/1251] eta 0:02:47 lr 0.000030 time 0.2767 (0.2885) loss 3.1838 (2.9809) grad_norm 3.0071 (3.1810) [2021-04-16 19:03:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][680/1251] eta 0:02:44 lr 0.000030 time 0.2634 (0.2882) loss 2.7489 (2.9807) grad_norm 3.1400 (3.1804) [2021-04-16 19:03:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][690/1251] eta 0:02:41 lr 0.000030 time 0.2654 (0.2882) loss 2.8624 (2.9855) grad_norm 3.5969 (3.1790) [2021-04-16 19:03:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][700/1251] eta 0:02:38 lr 0.000030 time 0.2625 (0.2880) loss 3.3447 (2.9878) grad_norm 3.2194 (3.1792) [2021-04-16 19:03:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][710/1251] eta 0:02:35 lr 0.000030 time 0.2475 (0.2878) loss 3.2466 (2.9849) grad_norm 3.8100 (3.1807) [2021-04-16 19:03:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][720/1251] eta 0:02:32 lr 0.000030 time 0.2909 (0.2880) loss 3.4710 (2.9847) grad_norm 3.4067 (3.1807) [2021-04-16 19:03:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][730/1251] eta 0:02:29 lr 0.000030 time 0.2924 (0.2878) loss 3.4462 (2.9873) grad_norm 3.1704 (3.1798) [2021-04-16 19:03:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][740/1251] eta 0:02:26 lr 0.000030 time 0.2725 (0.2876) loss 3.1253 (2.9889) grad_norm 3.0285 (3.1812) [2021-04-16 19:03:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][750/1251] eta 0:02:24 lr 0.000030 time 0.2719 (0.2875) loss 3.4426 (2.9881) grad_norm 3.1859 (3.1845) [2021-04-16 19:03:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][760/1251] eta 0:02:21 lr 0.000030 time 0.2644 (0.2874) loss 3.5842 (2.9902) grad_norm 3.3681 (3.1875) [2021-04-16 19:03:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][770/1251] eta 0:02:18 lr 0.000030 time 0.2571 (0.2872) loss 2.8043 (2.9909) grad_norm 2.9387 (3.1911) [2021-04-16 19:03:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][780/1251] eta 0:02:15 lr 0.000030 time 0.3039 (0.2870) loss 3.6221 (2.9924) grad_norm 2.9607 (3.1921) [2021-04-16 19:03:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][790/1251] eta 0:02:12 lr 0.000030 time 0.2602 (0.2869) loss 3.4455 (2.9936) grad_norm 3.8590 (3.1930) [2021-04-16 19:03:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][800/1251] eta 0:02:09 lr 0.000030 time 0.2833 (0.2868) loss 3.4805 (2.9945) grad_norm 3.2759 (3.1910) [2021-04-16 19:03:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][810/1251] eta 0:02:06 lr 0.000030 time 0.2589 (0.2866) loss 2.9659 (2.9938) grad_norm 3.7796 (3.1898) [2021-04-16 19:03:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][820/1251] eta 0:02:03 lr 0.000030 time 0.3103 (0.2865) loss 1.9159 (2.9925) grad_norm 5.2153 (3.1914) [2021-04-16 19:04:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][830/1251] eta 0:02:00 lr 0.000030 time 0.2685 (0.2864) loss 3.9201 (2.9936) grad_norm 2.9737 (3.1911) [2021-04-16 19:04:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][840/1251] eta 0:01:57 lr 0.000030 time 0.2839 (0.2863) loss 3.2053 (2.9948) grad_norm 3.5446 (3.1883) [2021-04-16 19:04:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][850/1251] eta 0:01:54 lr 0.000030 time 0.2662 (0.2861) loss 2.1605 (2.9943) grad_norm 3.0321 (3.1877) [2021-04-16 19:04:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][860/1251] eta 0:01:51 lr 0.000030 time 0.2660 (0.2860) loss 2.4386 (2.9931) grad_norm 3.5052 (3.1950) [2021-04-16 19:04:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][870/1251] eta 0:01:48 lr 0.000030 time 0.2668 (0.2858) loss 3.4363 (2.9959) grad_norm 2.6269 (3.1955) [2021-04-16 19:04:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][880/1251] eta 0:01:46 lr 0.000030 time 0.2896 (0.2859) loss 3.2068 (2.9974) grad_norm 3.1900 (3.1937) [2021-04-16 19:04:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][890/1251] eta 0:01:43 lr 0.000030 time 0.2781 (0.2858) loss 2.9173 (3.0000) grad_norm 3.0449 (3.1960) [2021-04-16 19:04:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][900/1251] eta 0:01:40 lr 0.000030 time 0.2715 (0.2857) loss 3.3517 (3.0013) grad_norm 3.3007 (3.2008) [2021-04-16 19:04:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][910/1251] eta 0:01:37 lr 0.000030 time 0.2682 (0.2856) loss 3.6594 (3.0012) grad_norm 3.4734 (3.2015) [2021-04-16 19:04:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][920/1251] eta 0:01:34 lr 0.000030 time 0.2883 (0.2856) loss 3.2916 (3.0039) grad_norm 3.5871 (3.2032) [2021-04-16 19:04:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][930/1251] eta 0:01:31 lr 0.000030 time 0.2572 (0.2859) loss 3.0910 (3.0037) grad_norm 2.9038 (3.2033) [2021-04-16 19:04:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][940/1251] eta 0:01:28 lr 0.000030 time 0.2660 (0.2859) loss 3.0127 (3.0047) grad_norm 3.0854 (3.2028) [2021-04-16 19:04:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][950/1251] eta 0:01:26 lr 0.000030 time 0.2760 (0.2858) loss 2.1434 (3.0017) grad_norm 2.7143 (3.2026) [2021-04-16 19:04:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][960/1251] eta 0:01:23 lr 0.000030 time 0.3029 (0.2856) loss 2.0630 (3.0005) grad_norm 2.8157 (3.2024) [2021-04-16 19:04:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][970/1251] eta 0:01:20 lr 0.000030 time 0.2808 (0.2856) loss 3.3053 (3.0006) grad_norm 2.7861 (3.2011) [2021-04-16 19:04:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][980/1251] eta 0:01:17 lr 0.000030 time 0.2850 (0.2855) loss 2.1024 (3.0005) grad_norm 2.7896 (3.1993) [2021-04-16 19:04:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][990/1251] eta 0:01:14 lr 0.000030 time 0.2995 (0.2855) loss 2.4105 (3.0001) grad_norm 3.1224 (3.1973) [2021-04-16 19:04:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][1000/1251] eta 0:01:11 lr 0.000030 time 0.2730 (0.2854) loss 3.2363 (3.0006) grad_norm 2.7900 (3.1960) [2021-04-16 19:04:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][1010/1251] eta 0:01:08 lr 0.000030 time 0.2901 (0.2854) loss 2.6311 (3.0002) grad_norm 2.9530 (3.1967) [2021-04-16 19:04:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][1020/1251] eta 0:01:05 lr 0.000030 time 0.2762 (0.2853) loss 3.0973 (3.0002) grad_norm 3.2471 (3.1961) [2021-04-16 19:04:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][1030/1251] eta 0:01:03 lr 0.000030 time 0.2851 (0.2852) loss 3.0469 (3.0006) grad_norm 2.9062 (3.1973) [2021-04-16 19:04:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][1040/1251] eta 0:01:00 lr 0.000030 time 0.2906 (0.2851) loss 3.4323 (3.0001) grad_norm 3.0832 (3.1979) [2021-04-16 19:05:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][1050/1251] eta 0:00:57 lr 0.000030 time 0.4296 (0.2852) loss 3.2965 (2.9984) grad_norm 2.8181 (3.1963) [2021-04-16 19:05:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][1060/1251] eta 0:00:54 lr 0.000030 time 0.2727 (0.2850) loss 2.5439 (2.9979) grad_norm 3.5389 (3.1971) [2021-04-16 19:05:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][1070/1251] eta 0:00:51 lr 0.000030 time 0.2945 (0.2850) loss 2.9346 (2.9971) grad_norm 3.8898 (3.1964) [2021-04-16 19:05:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][1080/1251] eta 0:00:48 lr 0.000030 time 0.2747 (0.2849) loss 3.6296 (2.9963) grad_norm 3.1090 (3.1989) [2021-04-16 19:05:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][1090/1251] eta 0:00:45 lr 0.000030 time 0.2578 (0.2848) loss 3.6738 (2.9970) grad_norm 3.3495 (3.1996) [2021-04-16 19:05:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][1100/1251] eta 0:00:42 lr 0.000030 time 0.2723 (0.2847) loss 2.7518 (2.9974) grad_norm 3.6876 (3.1984) [2021-04-16 19:05:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][1110/1251] eta 0:00:40 lr 0.000030 time 0.2627 (0.2846) loss 2.0298 (2.9955) grad_norm 2.9534 (3.1990) [2021-04-16 19:05:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][1120/1251] eta 0:00:37 lr 0.000030 time 0.2757 (0.2846) loss 3.2485 (2.9980) grad_norm 3.6115 (3.2013) [2021-04-16 19:05:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][1130/1251] eta 0:00:34 lr 0.000030 time 0.2424 (0.2845) loss 2.5012 (2.9982) grad_norm 3.8717 (3.2009) [2021-04-16 19:05:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][1140/1251] eta 0:00:31 lr 0.000030 time 0.2625 (0.2844) loss 2.1675 (2.9973) grad_norm 3.4779 (3.1991) [2021-04-16 19:05:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][1150/1251] eta 0:00:28 lr 0.000030 time 0.2656 (0.2845) loss 3.4155 (2.9997) grad_norm 3.1564 (3.2009) [2021-04-16 19:05:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][1160/1251] eta 0:00:25 lr 0.000030 time 0.2682 (0.2844) loss 2.2730 (2.9994) grad_norm 3.0073 (3.2008) [2021-04-16 19:05:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][1170/1251] eta 0:00:23 lr 0.000030 time 0.2932 (0.2844) loss 2.6679 (2.9992) grad_norm 2.9179 (3.2021) [2021-04-16 19:05:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][1180/1251] eta 0:00:20 lr 0.000030 time 0.2798 (0.2843) loss 2.8135 (3.0007) grad_norm 3.9494 (3.2040) [2021-04-16 19:05:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][1190/1251] eta 0:00:17 lr 0.000030 time 0.2686 (0.2842) loss 3.2700 (3.0020) grad_norm 2.8762 (3.2036) [2021-04-16 19:05:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][1200/1251] eta 0:00:14 lr 0.000030 time 0.2656 (0.2841) loss 3.4377 (3.0033) grad_norm 3.3601 (3.2050) [2021-04-16 19:05:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][1210/1251] eta 0:00:11 lr 0.000030 time 0.3012 (0.2841) loss 3.4981 (3.0054) grad_norm 2.8190 (3.2062) [2021-04-16 19:05:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][1220/1251] eta 0:00:08 lr 0.000030 time 0.2745 (0.2839) loss 2.5186 (3.0074) grad_norm 3.0368 (3.2046) [2021-04-16 19:05:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][1230/1251] eta 0:00:05 lr 0.000030 time 0.2854 (0.2839) loss 3.5949 (3.0091) grad_norm 3.0073 (3.2050) [2021-04-16 19:05:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][1240/1251] eta 0:00:03 lr 0.000030 time 0.2484 (0.2837) loss 2.6583 (3.0072) grad_norm 2.7718 (3.2090) [2021-04-16 19:05:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [272/300][1250/1251] eta 0:00:00 lr 0.000030 time 0.2478 (0.2834) loss 2.8047 (3.0082) grad_norm 3.2319 (3.2094) [2021-04-16 19:06:05 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 272 training takes 0:06:03 [2021-04-16 19:06:05 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_272.pth saving...... [2021-04-16 19:06:26 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_272.pth saved !!! [2021-04-16 19:06:27 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.153 (1.153) Loss 0.7706 (0.7706) Acc@1 81.250 (81.250) Acc@5 95.703 (95.703) [2021-04-16 19:06:28 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.141 (0.209) Loss 0.7346 (0.8291) Acc@1 83.203 (80.850) Acc@5 96.191 (95.188) [2021-04-16 19:06:30 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.180 (0.204) Loss 0.7740 (0.8277) Acc@1 81.445 (80.808) Acc@5 95.996 (95.238) [2021-04-16 19:06:33 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.195 (0.231) Loss 0.8365 (0.8242) Acc@1 80.078 (80.844) Acc@5 96.289 (95.376) [2021-04-16 19:06:35 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.224) Loss 0.8693 (0.8209) Acc@1 81.348 (80.940) Acc@5 94.727 (95.486) [2021-04-16 19:06:55 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.966 Acc@5 95.506 [2021-04-16 19:06:55 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 81.0% [2021-04-16 19:06:55 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 81.00% [2021-04-16 19:07:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][0/1251] eta 4:52:30 lr 0.000030 time 14.0291 (14.0291) loss 3.0883 (3.0883) grad_norm 3.2908 (3.2908) [2021-04-16 19:07:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][10/1251] eta 0:31:32 lr 0.000030 time 0.2848 (1.5247) loss 2.8679 (2.8888) grad_norm 4.7193 (3.2660) [2021-04-16 19:07:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][20/1251] eta 0:19:04 lr 0.000030 time 0.2667 (0.9296) loss 2.7139 (2.8173) grad_norm 3.0578 (3.2442) [2021-04-16 19:07:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][30/1251] eta 0:14:39 lr 0.000030 time 0.2687 (0.7205) loss 3.5007 (2.9415) grad_norm 3.1264 (3.2911) [2021-04-16 19:07:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][40/1251] eta 0:12:25 lr 0.000030 time 0.2781 (0.6153) loss 3.8800 (2.9263) grad_norm 5.3414 (3.3160) [2021-04-16 19:07:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][50/1251] eta 0:11:00 lr 0.000030 time 0.2497 (0.5501) loss 3.6268 (2.9561) grad_norm 2.9319 (3.2858) [2021-04-16 19:07:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][60/1251] eta 0:10:02 lr 0.000030 time 0.2842 (0.5061) loss 3.2211 (2.9497) grad_norm 2.9161 (3.2812) [2021-04-16 19:07:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][70/1251] eta 0:09:21 lr 0.000030 time 0.2928 (0.4751) loss 2.2880 (2.9286) grad_norm 3.8837 (3.2762) [2021-04-16 19:07:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][80/1251] eta 0:08:47 lr 0.000030 time 0.2740 (0.4502) loss 3.3746 (2.9289) grad_norm 3.3503 (3.2907) [2021-04-16 19:07:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][90/1251] eta 0:08:22 lr 0.000030 time 0.2779 (0.4324) loss 3.1092 (2.9348) grad_norm 3.2730 (3.3005) [2021-04-16 19:07:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][100/1251] eta 0:08:00 lr 0.000030 time 0.2573 (0.4174) loss 3.3891 (2.9570) grad_norm 3.1881 (3.3142) [2021-04-16 19:07:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][110/1251] eta 0:07:42 lr 0.000030 time 0.3136 (0.4050) loss 2.4420 (2.9632) grad_norm 3.4984 (3.3270) [2021-04-16 19:07:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][120/1251] eta 0:07:27 lr 0.000030 time 0.2859 (0.3956) loss 2.4447 (2.9528) grad_norm 4.1439 (3.3216) [2021-04-16 19:07:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][130/1251] eta 0:07:16 lr 0.000030 time 0.4361 (0.3890) loss 3.0165 (2.9658) grad_norm 3.0274 (3.3232) [2021-04-16 19:07:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][140/1251] eta 0:07:04 lr 0.000029 time 0.2795 (0.3823) loss 2.4046 (2.9605) grad_norm 3.1153 (3.3303) [2021-04-16 19:07:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][150/1251] eta 0:06:53 lr 0.000029 time 0.2971 (0.3755) loss 3.4094 (2.9679) grad_norm 2.8301 (3.3060) [2021-04-16 19:07:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][160/1251] eta 0:06:44 lr 0.000029 time 0.3015 (0.3704) loss 3.3716 (2.9707) grad_norm 3.0545 (3.2860) [2021-04-16 19:07:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][170/1251] eta 0:06:34 lr 0.000029 time 0.2711 (0.3650) loss 2.1643 (2.9760) grad_norm 2.9204 (3.2675) [2021-04-16 19:08:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][180/1251] eta 0:06:26 lr 0.000029 time 0.2837 (0.3604) loss 3.0693 (2.9680) grad_norm 3.2539 (3.2611) [2021-04-16 19:08:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][190/1251] eta 0:06:17 lr 0.000029 time 0.2785 (0.3560) loss 2.9781 (2.9771) grad_norm 3.0535 (3.2576) [2021-04-16 19:08:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][200/1251] eta 0:06:09 lr 0.000029 time 0.2588 (0.3517) loss 3.2882 (2.9821) grad_norm 3.3287 (3.2591) [2021-04-16 19:08:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][210/1251] eta 0:06:02 lr 0.000029 time 0.2656 (0.3483) loss 3.3743 (2.9935) grad_norm 3.6053 (3.2653) [2021-04-16 19:08:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][220/1251] eta 0:05:56 lr 0.000029 time 0.2702 (0.3458) loss 3.6696 (2.9976) grad_norm 3.6088 (3.2585) [2021-04-16 19:08:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][230/1251] eta 0:05:50 lr 0.000029 time 0.2892 (0.3431) loss 2.7049 (2.9854) grad_norm 2.9404 (3.2572) [2021-04-16 19:08:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][240/1251] eta 0:05:44 lr 0.000029 time 0.2813 (0.3406) loss 3.5525 (2.9944) grad_norm 3.2271 (3.2620) [2021-04-16 19:08:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][250/1251] eta 0:05:38 lr 0.000029 time 0.2664 (0.3383) loss 2.8458 (2.9875) grad_norm 3.9820 (3.2579) [2021-04-16 19:08:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][260/1251] eta 0:05:33 lr 0.000029 time 0.3881 (0.3362) loss 3.4269 (2.9828) grad_norm 4.6868 (3.2629) [2021-04-16 19:08:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][270/1251] eta 0:05:27 lr 0.000029 time 0.2893 (0.3340) loss 2.1332 (2.9810) grad_norm 2.8805 (3.2595) [2021-04-16 19:08:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][280/1251] eta 0:05:22 lr 0.000029 time 0.3044 (0.3325) loss 3.2118 (2.9891) grad_norm 3.0314 (3.2595) [2021-04-16 19:08:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][290/1251] eta 0:05:17 lr 0.000029 time 0.3061 (0.3307) loss 3.2612 (3.0004) grad_norm 3.3084 (3.2606) [2021-04-16 19:08:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][300/1251] eta 0:05:12 lr 0.000029 time 0.2790 (0.3288) loss 3.0468 (2.9955) grad_norm 2.9409 (3.2611) [2021-04-16 19:08:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][310/1251] eta 0:05:07 lr 0.000029 time 0.2566 (0.3272) loss 3.2410 (3.0002) grad_norm 2.8609 (3.2544) [2021-04-16 19:08:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][320/1251] eta 0:05:03 lr 0.000029 time 0.2592 (0.3255) loss 1.9356 (2.9980) grad_norm 3.4787 (3.2545) [2021-04-16 19:08:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][330/1251] eta 0:04:58 lr 0.000029 time 0.3014 (0.3243) loss 3.3177 (3.0054) grad_norm 3.0033 (3.2562) [2021-04-16 19:08:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][340/1251] eta 0:04:54 lr 0.000029 time 0.2427 (0.3228) loss 3.2077 (3.0035) grad_norm 2.7315 (3.2502) [2021-04-16 19:08:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][350/1251] eta 0:04:49 lr 0.000029 time 0.2536 (0.3214) loss 3.0516 (3.0049) grad_norm 3.4135 (3.2499) [2021-04-16 19:08:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][360/1251] eta 0:04:45 lr 0.000029 time 0.2726 (0.3209) loss 2.9336 (3.0008) grad_norm 2.9716 (3.2450) [2021-04-16 19:08:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][370/1251] eta 0:04:41 lr 0.000029 time 0.2676 (0.3200) loss 2.6845 (2.9932) grad_norm 3.3389 (3.2464) [2021-04-16 19:08:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][380/1251] eta 0:04:37 lr 0.000029 time 0.2766 (0.3189) loss 3.0227 (2.9866) grad_norm 2.7777 (3.2436) [2021-04-16 19:08:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][390/1251] eta 0:04:33 lr 0.000029 time 0.2467 (0.3177) loss 2.6950 (2.9857) grad_norm 3.9354 (3.2457) [2021-04-16 19:09:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][400/1251] eta 0:04:29 lr 0.000029 time 0.2822 (0.3168) loss 3.1417 (2.9744) grad_norm 3.3134 (3.2455) [2021-04-16 19:09:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][410/1251] eta 0:04:25 lr 0.000029 time 0.2978 (0.3161) loss 3.2869 (2.9775) grad_norm 3.0974 (3.2571) [2021-04-16 19:09:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][420/1251] eta 0:04:22 lr 0.000029 time 0.2809 (0.3153) loss 3.4359 (2.9834) grad_norm 3.5639 (3.2633) [2021-04-16 19:09:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][430/1251] eta 0:04:18 lr 0.000029 time 0.2631 (0.3144) loss 3.3807 (2.9730) grad_norm 2.7464 (3.2553) [2021-04-16 19:09:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][440/1251] eta 0:04:14 lr 0.000029 time 0.2701 (0.3135) loss 2.5738 (2.9749) grad_norm 2.9850 (3.2579) [2021-04-16 19:09:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][450/1251] eta 0:04:10 lr 0.000029 time 0.2918 (0.3129) loss 2.3383 (2.9756) grad_norm 3.0790 (3.2566) [2021-04-16 19:09:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][460/1251] eta 0:04:06 lr 0.000029 time 0.2636 (0.3122) loss 3.2719 (2.9711) grad_norm 3.7495 (3.2572) [2021-04-16 19:09:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][470/1251] eta 0:04:03 lr 0.000029 time 0.2717 (0.3115) loss 3.4472 (2.9729) grad_norm 2.9521 (3.2575) [2021-04-16 19:09:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][480/1251] eta 0:03:59 lr 0.000029 time 0.2775 (0.3109) loss 3.4855 (2.9781) grad_norm 4.9200 (3.2575) [2021-04-16 19:09:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][490/1251] eta 0:03:56 lr 0.000029 time 0.2905 (0.3102) loss 3.1924 (2.9808) grad_norm 3.1166 (3.2550) [2021-04-16 19:09:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][500/1251] eta 0:03:52 lr 0.000029 time 0.2584 (0.3095) loss 2.5824 (2.9790) grad_norm 2.9983 (3.2495) [2021-04-16 19:09:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][510/1251] eta 0:03:48 lr 0.000029 time 0.2912 (0.3090) loss 3.6487 (2.9781) grad_norm 3.8746 (3.2487) [2021-04-16 19:09:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][520/1251] eta 0:03:45 lr 0.000029 time 0.2655 (0.3087) loss 3.5002 (2.9814) grad_norm 3.1696 (3.2467) [2021-04-16 19:09:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][530/1251] eta 0:03:42 lr 0.000029 time 0.2593 (0.3080) loss 3.4102 (2.9825) grad_norm 4.0915 (3.2452) [2021-04-16 19:09:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][540/1251] eta 0:03:38 lr 0.000029 time 0.2874 (0.3075) loss 3.5386 (2.9866) grad_norm 3.4101 (3.2513) [2021-04-16 19:09:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][550/1251] eta 0:03:35 lr 0.000029 time 0.2654 (0.3070) loss 2.2160 (2.9799) grad_norm 3.2090 (3.2548) [2021-04-16 19:09:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][560/1251] eta 0:03:31 lr 0.000029 time 0.3056 (0.3066) loss 3.1471 (2.9799) grad_norm 10.9067 (3.2747) [2021-04-16 19:09:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][570/1251] eta 0:03:28 lr 0.000029 time 0.2863 (0.3065) loss 2.6866 (2.9782) grad_norm 3.5017 (3.2733) [2021-04-16 19:09:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][580/1251] eta 0:03:25 lr 0.000029 time 0.2755 (0.3062) loss 2.1738 (2.9792) grad_norm 3.4129 (3.2725) [2021-04-16 19:09:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][590/1251] eta 0:03:22 lr 0.000029 time 0.2867 (0.3058) loss 3.1760 (2.9815) grad_norm 3.4032 (3.2738) [2021-04-16 19:09:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][600/1251] eta 0:03:18 lr 0.000029 time 0.2588 (0.3053) loss 3.4617 (2.9824) grad_norm 2.8516 (3.2702) [2021-04-16 19:10:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][610/1251] eta 0:03:15 lr 0.000029 time 0.2782 (0.3050) loss 2.1776 (2.9796) grad_norm 2.7023 (3.2677) [2021-04-16 19:10:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][620/1251] eta 0:03:12 lr 0.000029 time 0.2839 (0.3045) loss 3.0666 (2.9749) grad_norm 3.0686 (3.2660) [2021-04-16 19:10:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][630/1251] eta 0:03:08 lr 0.000029 time 0.2936 (0.3041) loss 3.1809 (2.9796) grad_norm 2.8317 (3.2636) [2021-04-16 19:10:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][640/1251] eta 0:03:05 lr 0.000029 time 0.2792 (0.3036) loss 2.7557 (2.9818) grad_norm 2.9579 (3.2595) [2021-04-16 19:10:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][650/1251] eta 0:03:02 lr 0.000029 time 0.2834 (0.3032) loss 2.9850 (2.9838) grad_norm 3.2577 (3.2576) [2021-04-16 19:10:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][660/1251] eta 0:02:58 lr 0.000029 time 0.2979 (0.3028) loss 3.1169 (2.9859) grad_norm 3.0230 (3.2559) [2021-04-16 19:10:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][670/1251] eta 0:02:55 lr 0.000029 time 0.2518 (0.3026) loss 3.2294 (2.9872) grad_norm 3.3098 (3.2551) [2021-04-16 19:10:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][680/1251] eta 0:02:52 lr 0.000029 time 0.2721 (0.3022) loss 2.3742 (2.9855) grad_norm 2.8591 (3.2550) [2021-04-16 19:10:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][690/1251] eta 0:02:49 lr 0.000029 time 0.2762 (0.3020) loss 2.0236 (2.9839) grad_norm 2.9386 (3.2540) [2021-04-16 19:10:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][700/1251] eta 0:02:46 lr 0.000029 time 0.2735 (0.3017) loss 3.3953 (2.9836) grad_norm 3.1143 (3.2553) [2021-04-16 19:10:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][710/1251] eta 0:02:43 lr 0.000029 time 0.2906 (0.3017) loss 3.1049 (2.9816) grad_norm 3.0207 (3.2535) [2021-04-16 19:10:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][720/1251] eta 0:02:40 lr 0.000029 time 0.2843 (0.3015) loss 3.5283 (2.9814) grad_norm 2.6344 (3.2528) [2021-04-16 19:10:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][730/1251] eta 0:02:36 lr 0.000029 time 0.2728 (0.3012) loss 3.7326 (2.9824) grad_norm 3.6193 (3.2559) [2021-04-16 19:10:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][740/1251] eta 0:02:33 lr 0.000029 time 0.2668 (0.3009) loss 2.6662 (2.9817) grad_norm 3.3855 (3.2531) [2021-04-16 19:10:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][750/1251] eta 0:02:30 lr 0.000029 time 0.2582 (0.3007) loss 3.3295 (2.9847) grad_norm 3.2368 (3.2511) [2021-04-16 19:10:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][760/1251] eta 0:02:27 lr 0.000029 time 0.2808 (0.3004) loss 2.8634 (2.9869) grad_norm 3.2167 (3.2484) [2021-04-16 19:10:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][770/1251] eta 0:02:24 lr 0.000029 time 0.2866 (0.3001) loss 3.3914 (2.9896) grad_norm 4.0367 (3.2523) [2021-04-16 19:10:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][780/1251] eta 0:02:21 lr 0.000029 time 0.2549 (0.2997) loss 2.4111 (2.9863) grad_norm 3.0849 (3.2511) [2021-04-16 19:10:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][790/1251] eta 0:02:18 lr 0.000029 time 0.2743 (0.2996) loss 2.9201 (2.9860) grad_norm 2.6218 (3.2488) [2021-04-16 19:10:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][800/1251] eta 0:02:15 lr 0.000029 time 0.2699 (0.2993) loss 3.3256 (2.9872) grad_norm 3.5830 (3.2459) [2021-04-16 19:10:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][810/1251] eta 0:02:11 lr 0.000029 time 0.2871 (0.2990) loss 3.0880 (2.9894) grad_norm 3.6492 (3.2468) [2021-04-16 19:11:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][820/1251] eta 0:02:08 lr 0.000029 time 0.2942 (0.2988) loss 2.5944 (2.9884) grad_norm 2.9942 (3.2517) [2021-04-16 19:11:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][830/1251] eta 0:02:05 lr 0.000029 time 0.2818 (0.2986) loss 2.9803 (2.9871) grad_norm 3.0719 (3.2510) [2021-04-16 19:11:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][840/1251] eta 0:02:02 lr 0.000029 time 0.2778 (0.2985) loss 2.7433 (2.9900) grad_norm 3.7011 (3.2482) [2021-04-16 19:11:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][850/1251] eta 0:01:59 lr 0.000029 time 0.2950 (0.2982) loss 2.9318 (2.9911) grad_norm 2.8895 (3.2455) [2021-04-16 19:11:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][860/1251] eta 0:01:56 lr 0.000029 time 0.2957 (0.2980) loss 2.3419 (2.9918) grad_norm 3.3940 (3.2473) [2021-04-16 19:11:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][870/1251] eta 0:01:53 lr 0.000029 time 0.2721 (0.2977) loss 2.6634 (2.9926) grad_norm 3.2717 (3.2496) [2021-04-16 19:11:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][880/1251] eta 0:01:50 lr 0.000029 time 0.2679 (0.2975) loss 2.0651 (2.9916) grad_norm 2.7525 (3.2482) [2021-04-16 19:11:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][890/1251] eta 0:01:47 lr 0.000029 time 0.2604 (0.2973) loss 2.2195 (2.9909) grad_norm 2.6593 (3.2483) [2021-04-16 19:11:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][900/1251] eta 0:01:44 lr 0.000029 time 0.2660 (0.2970) loss 3.1351 (2.9922) grad_norm 2.9523 (nan) [2021-04-16 19:11:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][910/1251] eta 0:01:41 lr 0.000029 time 0.2770 (0.2969) loss 2.4408 (2.9929) grad_norm 3.1539 (nan) [2021-04-16 19:11:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][920/1251] eta 0:01:38 lr 0.000029 time 0.3070 (0.2967) loss 3.2998 (2.9925) grad_norm 2.9352 (nan) [2021-04-16 19:11:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][930/1251] eta 0:01:35 lr 0.000029 time 0.2658 (0.2966) loss 3.1508 (2.9959) grad_norm 3.1701 (nan) [2021-04-16 19:11:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][940/1251] eta 0:01:32 lr 0.000029 time 0.2963 (0.2967) loss 2.9883 (2.9954) grad_norm 3.4329 (nan) [2021-04-16 19:11:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][950/1251] eta 0:01:29 lr 0.000029 time 0.2577 (0.2966) loss 2.1298 (2.9943) grad_norm 3.5444 (nan) [2021-04-16 19:11:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][960/1251] eta 0:01:26 lr 0.000029 time 0.2887 (0.2964) loss 2.8414 (2.9961) grad_norm 2.9370 (nan) [2021-04-16 19:11:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][970/1251] eta 0:01:23 lr 0.000029 time 0.2825 (0.2962) loss 3.3926 (2.9974) grad_norm 2.9592 (nan) [2021-04-16 19:11:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][980/1251] eta 0:01:20 lr 0.000029 time 0.2685 (0.2960) loss 2.8175 (2.9976) grad_norm 4.6752 (nan) [2021-04-16 19:11:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][990/1251] eta 0:01:17 lr 0.000029 time 0.2726 (0.2959) loss 3.2418 (2.9988) grad_norm 2.7814 (nan) [2021-04-16 19:11:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][1000/1251] eta 0:01:14 lr 0.000029 time 0.2946 (0.2957) loss 3.2605 (2.9973) grad_norm 3.0844 (nan) [2021-04-16 19:11:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][1010/1251] eta 0:01:11 lr 0.000029 time 0.2680 (0.2955) loss 2.0581 (2.9972) grad_norm 2.6397 (nan) [2021-04-16 19:11:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][1020/1251] eta 0:01:08 lr 0.000028 time 0.2679 (0.2954) loss 3.1670 (2.9944) grad_norm 3.9327 (nan) [2021-04-16 19:11:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][1030/1251] eta 0:01:05 lr 0.000028 time 0.3105 (0.2952) loss 2.7012 (2.9935) grad_norm 3.6897 (nan) [2021-04-16 19:12:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][1040/1251] eta 0:01:02 lr 0.000028 time 0.2689 (0.2950) loss 3.3638 (2.9939) grad_norm 3.1952 (nan) [2021-04-16 19:12:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][1050/1251] eta 0:00:59 lr 0.000028 time 0.2482 (0.2948) loss 3.1891 (2.9961) grad_norm 2.8887 (nan) [2021-04-16 19:12:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][1060/1251] eta 0:00:56 lr 0.000028 time 0.2725 (0.2947) loss 2.2404 (2.9956) grad_norm 3.5938 (nan) [2021-04-16 19:12:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][1070/1251] eta 0:00:53 lr 0.000028 time 0.2804 (0.2945) loss 3.0110 (2.9954) grad_norm 3.4007 (nan) [2021-04-16 19:12:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][1080/1251] eta 0:00:50 lr 0.000028 time 0.2587 (0.2944) loss 3.4015 (2.9951) grad_norm 3.2771 (nan) [2021-04-16 19:12:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][1090/1251] eta 0:00:47 lr 0.000028 time 0.2518 (0.2944) loss 3.0474 (2.9964) grad_norm 3.2095 (nan) [2021-04-16 19:12:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][1100/1251] eta 0:00:44 lr 0.000028 time 0.2531 (0.2942) loss 3.0527 (2.9965) grad_norm 2.7257 (nan) [2021-04-16 19:12:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][1110/1251] eta 0:00:41 lr 0.000028 time 0.2541 (0.2940) loss 2.4904 (2.9941) grad_norm 2.9854 (nan) [2021-04-16 19:12:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][1120/1251] eta 0:00:38 lr 0.000028 time 0.3162 (0.2939) loss 3.3400 (2.9951) grad_norm 3.2129 (nan) [2021-04-16 19:12:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][1130/1251] eta 0:00:35 lr 0.000028 time 0.2888 (0.2939) loss 2.8602 (2.9956) grad_norm 3.6319 (nan) [2021-04-16 19:12:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][1140/1251] eta 0:00:32 lr 0.000028 time 0.2667 (0.2938) loss 3.4872 (2.9945) grad_norm 3.5630 (nan) [2021-04-16 19:12:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][1150/1251] eta 0:00:29 lr 0.000028 time 0.2764 (0.2939) loss 3.2767 (2.9926) grad_norm 3.1928 (nan) [2021-04-16 19:12:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][1160/1251] eta 0:00:26 lr 0.000028 time 0.2792 (0.2938) loss 3.1955 (2.9925) grad_norm 3.2079 (nan) [2021-04-16 19:12:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][1170/1251] eta 0:00:23 lr 0.000028 time 0.3168 (0.2937) loss 2.5807 (2.9935) grad_norm 3.0659 (nan) [2021-04-16 19:12:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][1180/1251] eta 0:00:20 lr 0.000028 time 0.2913 (0.2935) loss 3.5511 (2.9918) grad_norm 3.1742 (nan) [2021-04-16 19:12:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][1190/1251] eta 0:00:17 lr 0.000028 time 0.2991 (0.2934) loss 2.3153 (2.9901) grad_norm 2.9521 (nan) [2021-04-16 19:12:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][1200/1251] eta 0:00:14 lr 0.000028 time 0.2584 (0.2933) loss 3.5601 (2.9906) grad_norm 2.8611 (nan) [2021-04-16 19:12:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][1210/1251] eta 0:00:12 lr 0.000028 time 0.3059 (0.2932) loss 2.1024 (2.9909) grad_norm 3.5652 (nan) [2021-04-16 19:12:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][1220/1251] eta 0:00:09 lr 0.000028 time 0.2867 (0.2931) loss 3.3212 (2.9901) grad_norm 2.9684 (nan) [2021-04-16 19:12:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][1230/1251] eta 0:00:06 lr 0.000028 time 0.2981 (0.2930) loss 3.6251 (2.9902) grad_norm 3.3511 (nan) [2021-04-16 19:12:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][1240/1251] eta 0:00:03 lr 0.000028 time 0.2477 (0.2928) loss 3.7314 (2.9910) grad_norm 3.1558 (nan) [2021-04-16 19:13:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [273/300][1250/1251] eta 0:00:00 lr 0.000028 time 0.2479 (0.2924) loss 2.0568 (2.9913) grad_norm 2.9263 (nan) [2021-04-16 19:13:20 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 273 training takes 0:06:25 [2021-04-16 19:13:20 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_273.pth saving...... [2021-04-16 19:13:52 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_273.pth saved !!! [2021-04-16 19:13:53 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.089 (1.089) Loss 0.8346 (0.8346) Acc@1 81.836 (81.836) Acc@5 95.605 (95.605) [2021-04-16 19:13:54 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.141 (0.244) Loss 0.8463 (0.8212) Acc@1 78.906 (80.726) Acc@5 95.117 (95.437) [2021-04-16 19:13:56 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.170 (0.231) Loss 0.8060 (0.8236) Acc@1 81.250 (80.878) Acc@5 95.703 (95.368) [2021-04-16 19:13:59 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.107 (0.244) Loss 0.7675 (0.8195) Acc@1 82.617 (80.831) Acc@5 95.312 (95.451) [2021-04-16 19:14:01 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.221) Loss 0.8114 (0.8108) Acc@1 80.469 (81.052) Acc@5 95.801 (95.541) [2021-04-16 19:14:31 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.940 Acc@5 95.472 [2021-04-16 19:14:31 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 80.9% [2021-04-16 19:14:31 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 81.00% [2021-04-16 19:14:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][0/1251] eta 5:48:33 lr 0.000028 time 16.7177 (16.7177) loss 2.4349 (2.4349) grad_norm 3.0157 (3.0157) [2021-04-16 19:14:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][10/1251] eta 0:36:44 lr 0.000028 time 0.3942 (1.7765) loss 3.3437 (2.7211) grad_norm 2.8431 (3.0788) [2021-04-16 19:14:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][20/1251] eta 0:21:48 lr 0.000028 time 0.2745 (1.0629) loss 2.1383 (2.8364) grad_norm 2.9081 (3.1830) [2021-04-16 19:14:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][30/1251] eta 0:16:27 lr 0.000028 time 0.2871 (0.8090) loss 2.5059 (2.9368) grad_norm 2.9512 (3.1900) [2021-04-16 19:14:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][40/1251] eta 0:13:46 lr 0.000028 time 0.3080 (0.6823) loss 2.5250 (2.9681) grad_norm 3.0962 (3.1547) [2021-04-16 19:15:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][50/1251] eta 0:12:02 lr 0.000028 time 0.2787 (0.6016) loss 3.1815 (2.9405) grad_norm 3.4067 (3.2087) [2021-04-16 19:15:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][60/1251] eta 0:10:52 lr 0.000028 time 0.2614 (0.5482) loss 2.1518 (2.8936) grad_norm 2.9028 (3.3347) [2021-04-16 19:15:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][70/1251] eta 0:10:03 lr 0.000028 time 0.2794 (0.5106) loss 2.6575 (2.9127) grad_norm 3.3829 (3.3254) [2021-04-16 19:15:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][80/1251] eta 0:09:25 lr 0.000028 time 0.2854 (0.4826) loss 2.7542 (2.9312) grad_norm 3.5242 (3.3217) [2021-04-16 19:15:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][90/1251] eta 0:08:54 lr 0.000028 time 0.2989 (0.4604) loss 3.3634 (2.9613) grad_norm 3.7792 (3.3118) [2021-04-16 19:15:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][100/1251] eta 0:08:29 lr 0.000028 time 0.2622 (0.4424) loss 3.3034 (2.9590) grad_norm 3.6043 (3.2844) [2021-04-16 19:15:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][110/1251] eta 0:08:07 lr 0.000028 time 0.2625 (0.4273) loss 3.9346 (2.9783) grad_norm 3.1414 (3.3125) [2021-04-16 19:15:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][120/1251] eta 0:07:48 lr 0.000028 time 0.2606 (0.4145) loss 2.9982 (2.9809) grad_norm 2.7593 (3.3172) [2021-04-16 19:15:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][130/1251] eta 0:07:33 lr 0.000028 time 0.3021 (0.4043) loss 3.6914 (2.9927) grad_norm 3.6182 (3.3310) [2021-04-16 19:15:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][140/1251] eta 0:07:18 lr 0.000028 time 0.2805 (0.3951) loss 3.4969 (3.0052) grad_norm 2.9874 (3.3292) [2021-04-16 19:15:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][150/1251] eta 0:07:08 lr 0.000028 time 0.3033 (0.3894) loss 2.6990 (3.0084) grad_norm 3.1199 (3.3117) [2021-04-16 19:15:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][160/1251] eta 0:06:58 lr 0.000028 time 0.2917 (0.3832) loss 3.4100 (3.0085) grad_norm 2.9091 (3.3107) [2021-04-16 19:15:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][170/1251] eta 0:06:47 lr 0.000028 time 0.3103 (0.3772) loss 3.5040 (3.0107) grad_norm 2.9229 (3.3005) [2021-04-16 19:15:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][180/1251] eta 0:06:38 lr 0.000028 time 0.2884 (0.3718) loss 3.2459 (3.0110) grad_norm 2.9061 (3.2896) [2021-04-16 19:15:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][190/1251] eta 0:06:29 lr 0.000028 time 0.2913 (0.3669) loss 3.3473 (3.0048) grad_norm 3.5050 (3.2917) [2021-04-16 19:15:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][200/1251] eta 0:06:20 lr 0.000028 time 0.2593 (0.3624) loss 1.7972 (2.9927) grad_norm 2.7330 (3.2907) [2021-04-16 19:15:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][210/1251] eta 0:06:13 lr 0.000028 time 0.3007 (0.3585) loss 3.4202 (2.9795) grad_norm 3.1448 (3.3007) [2021-04-16 19:15:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][220/1251] eta 0:06:05 lr 0.000028 time 0.2680 (0.3548) loss 1.9243 (2.9695) grad_norm 3.1999 (3.2884) [2021-04-16 19:15:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][230/1251] eta 0:05:58 lr 0.000028 time 0.2734 (0.3513) loss 3.7066 (2.9741) grad_norm 3.4930 (3.2870) [2021-04-16 19:15:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][240/1251] eta 0:05:52 lr 0.000028 time 0.2848 (0.3483) loss 3.2558 (2.9736) grad_norm 3.6071 (3.2818) [2021-04-16 19:15:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][250/1251] eta 0:05:45 lr 0.000028 time 0.2892 (0.3454) loss 2.1555 (2.9730) grad_norm 4.3864 (3.2900) [2021-04-16 19:16:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][260/1251] eta 0:05:39 lr 0.000028 time 0.2681 (0.3427) loss 3.2069 (2.9825) grad_norm 3.1271 (3.2908) [2021-04-16 19:16:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][270/1251] eta 0:05:33 lr 0.000028 time 0.2540 (0.3403) loss 2.8723 (2.9781) grad_norm 3.0832 (3.2841) [2021-04-16 19:16:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][280/1251] eta 0:05:28 lr 0.000028 time 0.2512 (0.3382) loss 2.2072 (2.9708) grad_norm 2.9404 (3.2841) [2021-04-16 19:16:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][290/1251] eta 0:05:22 lr 0.000028 time 0.2966 (0.3361) loss 2.5088 (2.9723) grad_norm 3.2367 (3.2778) [2021-04-16 19:16:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][300/1251] eta 0:05:18 lr 0.000028 time 0.2710 (0.3346) loss 3.4328 (2.9674) grad_norm 4.3468 (3.2776) [2021-04-16 19:16:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][310/1251] eta 0:05:13 lr 0.000028 time 0.2836 (0.3327) loss 2.5644 (2.9584) grad_norm 2.7739 (3.2767) [2021-04-16 19:16:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][320/1251] eta 0:05:08 lr 0.000028 time 0.2866 (0.3309) loss 3.0635 (2.9615) grad_norm 2.8144 (3.2853) [2021-04-16 19:16:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][330/1251] eta 0:05:03 lr 0.000028 time 0.3024 (0.3296) loss 2.7964 (2.9587) grad_norm 5.2471 (3.2918) [2021-04-16 19:16:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][340/1251] eta 0:04:58 lr 0.000028 time 0.2553 (0.3281) loss 3.1762 (2.9675) grad_norm 2.6497 (3.2878) [2021-04-16 19:16:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][350/1251] eta 0:04:54 lr 0.000028 time 0.3076 (0.3271) loss 2.8587 (2.9724) grad_norm 3.4738 (3.2912) [2021-04-16 19:16:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][360/1251] eta 0:04:50 lr 0.000028 time 0.2739 (0.3260) loss 2.4789 (2.9724) grad_norm 2.8731 (3.2854) [2021-04-16 19:16:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][370/1251] eta 0:04:46 lr 0.000028 time 0.2809 (0.3247) loss 2.2769 (2.9736) grad_norm 3.8640 (3.2910) [2021-04-16 19:16:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][380/1251] eta 0:04:41 lr 0.000028 time 0.2681 (0.3234) loss 3.3461 (2.9684) grad_norm 3.8669 (3.2980) [2021-04-16 19:16:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][390/1251] eta 0:04:37 lr 0.000028 time 0.2720 (0.3221) loss 2.3530 (2.9696) grad_norm 3.2335 (3.3080) [2021-04-16 19:16:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][400/1251] eta 0:04:33 lr 0.000028 time 0.2985 (0.3212) loss 3.2708 (2.9714) grad_norm 3.0291 (3.3053) [2021-04-16 19:16:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][410/1251] eta 0:04:29 lr 0.000028 time 0.2979 (0.3202) loss 3.1168 (2.9661) grad_norm 2.5190 (3.2972) [2021-04-16 19:16:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][420/1251] eta 0:04:25 lr 0.000028 time 0.2698 (0.3192) loss 3.4855 (2.9665) grad_norm 4.5062 (3.2914) [2021-04-16 19:16:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][430/1251] eta 0:04:21 lr 0.000028 time 0.3003 (0.3184) loss 3.3865 (2.9772) grad_norm 2.7207 (3.2905) [2021-04-16 19:16:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][440/1251] eta 0:04:17 lr 0.000028 time 0.2697 (0.3176) loss 2.0236 (2.9762) grad_norm 3.3936 (3.2915) [2021-04-16 19:16:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][450/1251] eta 0:04:13 lr 0.000028 time 0.2627 (0.3167) loss 3.6627 (2.9750) grad_norm 2.8957 (3.2913) [2021-04-16 19:16:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][460/1251] eta 0:04:09 lr 0.000028 time 0.2765 (0.3158) loss 2.8678 (2.9790) grad_norm 2.9418 (3.2879) [2021-04-16 19:16:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][470/1251] eta 0:04:06 lr 0.000028 time 0.2741 (0.3150) loss 2.3117 (2.9836) grad_norm 3.3145 (3.2889) [2021-04-16 19:17:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][480/1251] eta 0:04:02 lr 0.000028 time 0.2927 (0.3142) loss 2.9889 (2.9845) grad_norm 3.0490 (3.2876) [2021-04-16 19:17:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][490/1251] eta 0:03:58 lr 0.000028 time 0.2812 (0.3137) loss 3.3762 (2.9790) grad_norm 3.1607 (3.2864) [2021-04-16 19:17:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][500/1251] eta 0:03:55 lr 0.000028 time 0.2950 (0.3131) loss 2.8680 (2.9778) grad_norm 3.4984 (3.2857) [2021-04-16 19:17:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][510/1251] eta 0:03:51 lr 0.000028 time 0.2664 (0.3125) loss 3.2087 (2.9767) grad_norm 3.3607 (3.2836) [2021-04-16 19:17:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][520/1251] eta 0:03:48 lr 0.000028 time 0.3061 (0.3120) loss 2.8388 (2.9748) grad_norm 3.6736 (3.2813) [2021-04-16 19:17:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][530/1251] eta 0:03:44 lr 0.000028 time 0.2570 (0.3115) loss 3.4681 (2.9750) grad_norm 2.6006 (3.2779) [2021-04-16 19:17:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][540/1251] eta 0:03:41 lr 0.000028 time 0.2810 (0.3113) loss 3.0509 (2.9729) grad_norm 2.7135 (3.2745) [2021-04-16 19:17:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][550/1251] eta 0:03:37 lr 0.000028 time 0.2761 (0.3107) loss 2.9028 (2.9699) grad_norm 3.1050 (3.2672) [2021-04-16 19:17:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][560/1251] eta 0:03:34 lr 0.000028 time 0.2870 (0.3101) loss 3.2039 (2.9736) grad_norm 3.5649 (3.2624) [2021-04-16 19:17:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][570/1251] eta 0:03:30 lr 0.000028 time 0.2544 (0.3095) loss 3.2311 (2.9709) grad_norm 2.8807 (3.2626) [2021-04-16 19:17:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][580/1251] eta 0:03:27 lr 0.000028 time 0.2747 (0.3094) loss 1.5833 (2.9688) grad_norm 3.0696 (3.2602) [2021-04-16 19:17:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][590/1251] eta 0:03:24 lr 0.000028 time 0.2773 (0.3090) loss 2.8907 (2.9665) grad_norm 3.1081 (3.2565) [2021-04-16 19:17:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][600/1251] eta 0:03:20 lr 0.000028 time 0.2703 (0.3085) loss 3.3642 (2.9678) grad_norm 2.9594 (3.2550) [2021-04-16 19:17:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][610/1251] eta 0:03:17 lr 0.000028 time 0.2750 (0.3080) loss 2.8926 (2.9638) grad_norm 2.9121 (3.2549) [2021-04-16 19:17:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][620/1251] eta 0:03:14 lr 0.000028 time 0.2529 (0.3077) loss 2.6215 (2.9607) grad_norm 3.3278 (3.2754) [2021-04-16 19:17:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][630/1251] eta 0:03:10 lr 0.000028 time 0.2793 (0.3071) loss 2.6871 (2.9596) grad_norm 2.7529 (3.2703) [2021-04-16 19:17:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][640/1251] eta 0:03:07 lr 0.000028 time 0.2623 (0.3067) loss 3.3165 (2.9578) grad_norm 3.5892 (3.2703) [2021-04-16 19:17:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][650/1251] eta 0:03:04 lr 0.000028 time 0.2823 (0.3062) loss 2.6451 (2.9560) grad_norm 3.0551 (3.2684) [2021-04-16 19:17:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][660/1251] eta 0:03:00 lr 0.000028 time 0.2838 (0.3058) loss 2.7178 (2.9561) grad_norm 3.0372 (3.2669) [2021-04-16 19:17:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][670/1251] eta 0:02:57 lr 0.000027 time 0.2663 (0.3053) loss 3.0852 (2.9552) grad_norm 3.6731 (3.2642) [2021-04-16 19:17:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][680/1251] eta 0:02:54 lr 0.000027 time 0.2666 (0.3050) loss 2.8671 (2.9582) grad_norm 3.1114 (3.2635) [2021-04-16 19:18:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][690/1251] eta 0:02:50 lr 0.000027 time 0.2654 (0.3045) loss 3.4734 (2.9607) grad_norm 3.1484 (3.2605) [2021-04-16 19:18:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][700/1251] eta 0:02:47 lr 0.000027 time 0.2916 (0.3042) loss 3.3787 (2.9587) grad_norm 2.9444 (3.2585) [2021-04-16 19:18:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][710/1251] eta 0:02:44 lr 0.000027 time 0.2893 (0.3038) loss 3.6289 (2.9592) grad_norm 3.2392 (3.2573) [2021-04-16 19:18:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][720/1251] eta 0:02:41 lr 0.000027 time 0.2720 (0.3036) loss 3.0883 (2.9603) grad_norm 3.3026 (3.2529) [2021-04-16 19:18:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][730/1251] eta 0:02:38 lr 0.000027 time 0.3175 (0.3035) loss 1.8595 (2.9577) grad_norm 3.4103 (3.2534) [2021-04-16 19:18:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][740/1251] eta 0:02:34 lr 0.000027 time 0.2615 (0.3032) loss 3.3496 (2.9576) grad_norm 2.8433 (3.2512) [2021-04-16 19:18:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][750/1251] eta 0:02:31 lr 0.000027 time 0.2793 (0.3028) loss 2.7763 (2.9602) grad_norm 3.4619 (3.2496) [2021-04-16 19:18:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][760/1251] eta 0:02:28 lr 0.000027 time 0.2903 (0.3026) loss 3.2537 (2.9613) grad_norm 2.6862 (3.2475) [2021-04-16 19:18:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][770/1251] eta 0:02:25 lr 0.000027 time 0.2908 (0.3024) loss 3.2654 (2.9603) grad_norm 3.7813 (3.2459) [2021-04-16 19:18:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][780/1251] eta 0:02:22 lr 0.000027 time 0.3046 (0.3021) loss 2.6933 (2.9615) grad_norm 2.9108 (3.2447) [2021-04-16 19:18:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][790/1251] eta 0:02:19 lr 0.000027 time 0.2650 (0.3017) loss 2.6303 (2.9602) grad_norm 3.5488 (3.2425) [2021-04-16 19:18:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][800/1251] eta 0:02:15 lr 0.000027 time 0.2940 (0.3014) loss 1.9736 (2.9576) grad_norm 3.1592 (3.2424) [2021-04-16 19:18:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][810/1251] eta 0:02:12 lr 0.000027 time 0.2782 (0.3012) loss 3.2774 (2.9596) grad_norm 3.4376 (3.2412) [2021-04-16 19:18:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][820/1251] eta 0:02:09 lr 0.000027 time 0.2911 (0.3009) loss 3.2863 (2.9589) grad_norm 2.9186 (3.2394) [2021-04-16 19:18:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][830/1251] eta 0:02:06 lr 0.000027 time 0.2812 (0.3006) loss 3.2495 (2.9595) grad_norm 2.9631 (3.2383) [2021-04-16 19:18:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][840/1251] eta 0:02:03 lr 0.000027 time 0.2803 (0.3003) loss 2.9314 (2.9590) grad_norm 3.0899 (3.2356) [2021-04-16 19:18:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][850/1251] eta 0:02:00 lr 0.000027 time 0.2875 (0.3001) loss 2.2634 (2.9586) grad_norm 5.6207 (3.2400) [2021-04-16 19:18:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][860/1251] eta 0:01:57 lr 0.000027 time 0.2524 (0.2998) loss 2.2182 (2.9555) grad_norm 3.2312 (3.2388) [2021-04-16 19:18:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][870/1251] eta 0:01:54 lr 0.000027 time 0.2951 (0.2999) loss 2.3152 (2.9541) grad_norm 3.4650 (3.2381) [2021-04-16 19:18:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][880/1251] eta 0:01:51 lr 0.000027 time 0.2601 (0.2998) loss 3.0461 (2.9548) grad_norm 3.9412 (3.2363) [2021-04-16 19:18:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][890/1251] eta 0:01:48 lr 0.000027 time 0.2946 (0.2995) loss 3.6881 (2.9582) grad_norm 2.9694 (3.2372) [2021-04-16 19:19:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][900/1251] eta 0:01:45 lr 0.000027 time 0.2518 (0.2993) loss 1.9725 (2.9586) grad_norm 2.8668 (3.2356) [2021-04-16 19:19:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][910/1251] eta 0:01:41 lr 0.000027 time 0.2684 (0.2990) loss 3.2736 (2.9592) grad_norm 3.2182 (3.2333) [2021-04-16 19:19:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][920/1251] eta 0:01:38 lr 0.000027 time 0.2523 (0.2988) loss 3.2806 (2.9578) grad_norm 4.1925 (3.2334) [2021-04-16 19:19:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][930/1251] eta 0:01:35 lr 0.000027 time 0.2981 (0.2987) loss 2.8887 (2.9553) grad_norm 2.6534 (3.2327) [2021-04-16 19:19:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][940/1251] eta 0:01:32 lr 0.000027 time 0.2641 (0.2985) loss 3.5525 (2.9565) grad_norm 3.0894 (3.2295) [2021-04-16 19:19:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][950/1251] eta 0:01:29 lr 0.000027 time 0.2973 (0.2984) loss 2.8675 (2.9577) grad_norm 3.1937 (3.2332) [2021-04-16 19:19:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][960/1251] eta 0:01:26 lr 0.000027 time 0.3099 (0.2982) loss 3.5757 (2.9589) grad_norm 3.2485 (3.2341) [2021-04-16 19:19:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][970/1251] eta 0:01:23 lr 0.000027 time 0.2564 (0.2980) loss 3.0973 (2.9616) grad_norm 3.1685 (3.2330) [2021-04-16 19:19:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][980/1251] eta 0:01:20 lr 0.000027 time 0.2738 (0.2978) loss 2.6670 (2.9629) grad_norm 2.5799 (3.2315) [2021-04-16 19:19:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][990/1251] eta 0:01:17 lr 0.000027 time 0.2672 (0.2976) loss 2.3371 (2.9640) grad_norm 2.9980 (3.2318) [2021-04-16 19:19:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][1000/1251] eta 0:01:14 lr 0.000027 time 0.2796 (0.2974) loss 2.5797 (2.9656) grad_norm 3.2431 (3.2315) [2021-04-16 19:19:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][1010/1251] eta 0:01:11 lr 0.000027 time 0.2715 (0.2971) loss 3.0143 (2.9645) grad_norm 3.5189 (3.2318) [2021-04-16 19:19:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][1020/1251] eta 0:01:08 lr 0.000027 time 0.2787 (0.2971) loss 2.8445 (2.9649) grad_norm 3.2849 (3.2352) [2021-04-16 19:19:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][1030/1251] eta 0:01:05 lr 0.000027 time 0.2924 (0.2969) loss 2.7091 (2.9663) grad_norm 3.3752 (3.2339) [2021-04-16 19:19:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][1040/1251] eta 0:01:02 lr 0.000027 time 0.3003 (0.2967) loss 2.0781 (2.9668) grad_norm 3.2996 (3.2354) [2021-04-16 19:19:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][1050/1251] eta 0:00:59 lr 0.000027 time 0.2889 (0.2966) loss 3.0960 (2.9700) grad_norm 3.3215 (3.2345) [2021-04-16 19:19:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][1060/1251] eta 0:00:56 lr 0.000027 time 0.2798 (0.2964) loss 2.2097 (2.9684) grad_norm 3.4154 (3.2330) [2021-04-16 19:19:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][1070/1251] eta 0:00:53 lr 0.000027 time 0.2873 (0.2964) loss 3.1843 (2.9682) grad_norm 3.1491 (3.2328) [2021-04-16 19:19:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][1080/1251] eta 0:00:50 lr 0.000027 time 0.3246 (0.2962) loss 3.1834 (2.9693) grad_norm 2.6680 (3.2313) [2021-04-16 19:19:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][1090/1251] eta 0:00:47 lr 0.000027 time 0.2886 (0.2961) loss 2.3348 (2.9701) grad_norm 3.6511 (3.2305) [2021-04-16 19:19:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][1100/1251] eta 0:00:44 lr 0.000027 time 0.2844 (0.2959) loss 2.8167 (2.9701) grad_norm 3.2372 (3.2305) [2021-04-16 19:20:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][1110/1251] eta 0:00:41 lr 0.000027 time 0.2902 (0.2958) loss 3.5920 (2.9718) grad_norm 2.8864 (3.2299) [2021-04-16 19:20:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][1120/1251] eta 0:00:38 lr 0.000027 time 0.2550 (0.2956) loss 2.3834 (2.9716) grad_norm 3.4089 (3.2291) [2021-04-16 19:20:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][1130/1251] eta 0:00:35 lr 0.000027 time 0.2572 (0.2955) loss 3.2420 (2.9726) grad_norm 3.2915 (3.2298) [2021-04-16 19:20:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][1140/1251] eta 0:00:32 lr 0.000027 time 0.3014 (0.2953) loss 2.8107 (2.9724) grad_norm 2.7294 (3.2293) [2021-04-16 19:20:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][1150/1251] eta 0:00:29 lr 0.000027 time 0.3623 (0.2954) loss 2.7015 (2.9715) grad_norm 2.9991 (3.2279) [2021-04-16 19:20:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][1160/1251] eta 0:00:26 lr 0.000027 time 0.2940 (0.2953) loss 3.0557 (2.9721) grad_norm 3.6099 (3.2270) [2021-04-16 19:20:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][1170/1251] eta 0:00:23 lr 0.000027 time 0.2743 (0.2953) loss 3.0203 (2.9714) grad_norm 2.6866 (3.2281) [2021-04-16 19:20:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][1180/1251] eta 0:00:20 lr 0.000027 time 0.2828 (0.2951) loss 3.3296 (2.9740) grad_norm 3.1127 (3.2289) [2021-04-16 19:20:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][1190/1251] eta 0:00:17 lr 0.000027 time 0.2801 (0.2951) loss 3.1666 (2.9749) grad_norm 2.6657 (3.2285) [2021-04-16 19:20:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][1200/1251] eta 0:00:15 lr 0.000027 time 0.2624 (0.2949) loss 2.1820 (2.9763) grad_norm 4.1384 (3.2314) [2021-04-16 19:20:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][1210/1251] eta 0:00:12 lr 0.000027 time 0.2764 (0.2947) loss 3.2849 (2.9755) grad_norm 3.4822 (3.2313) [2021-04-16 19:20:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][1220/1251] eta 0:00:09 lr 0.000027 time 0.2592 (0.2945) loss 2.1388 (2.9749) grad_norm 3.1519 (3.2299) [2021-04-16 19:20:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][1230/1251] eta 0:00:06 lr 0.000027 time 0.2904 (0.2944) loss 3.1139 (2.9770) grad_norm 2.7578 (3.2313) [2021-04-16 19:20:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][1240/1251] eta 0:00:03 lr 0.000027 time 0.2491 (0.2942) loss 2.8109 (2.9750) grad_norm 3.5075 (3.2337) [2021-04-16 19:20:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [274/300][1250/1251] eta 0:00:00 lr 0.000027 time 0.2486 (0.2938) loss 2.6449 (2.9758) grad_norm 3.6476 (3.2338) [2021-04-16 19:20:45 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 274 training takes 0:06:14 [2021-04-16 19:20:45 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_274.pth saving...... [2021-04-16 19:21:00 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_274.pth saved !!! [2021-04-16 19:21:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.196 (1.196) Loss 0.8057 (0.8057) Acc@1 80.957 (80.957) Acc@5 95.312 (95.312) [2021-04-16 19:21:03 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.106 (0.218) Loss 0.8343 (0.8113) Acc@1 80.469 (81.259) Acc@5 96.094 (95.676) [2021-04-16 19:21:06 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.341 (0.263) Loss 0.8781 (0.8170) Acc@1 79.395 (81.180) Acc@5 95.312 (95.629) [2021-04-16 19:21:07 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.125 (0.223) Loss 0.8225 (0.8239) Acc@1 80.273 (81.058) Acc@5 95.312 (95.470) [2021-04-16 19:21:09 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.215) Loss 0.8401 (0.8283) Acc@1 80.957 (80.914) Acc@5 95.215 (95.422) [2021-04-16 19:21:24 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 80.968 Acc@5 95.420 [2021-04-16 19:21:24 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 81.0% [2021-04-16 19:21:24 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 81.00% [2021-04-16 19:21:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][0/1251] eta 3:37:26 lr 0.000027 time 10.4291 (10.4291) loss 3.3838 (3.3838) grad_norm 3.4131 (3.4131) [2021-04-16 19:21:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][10/1251] eta 0:24:40 lr 0.000027 time 0.2676 (1.1929) loss 3.0738 (3.0792) grad_norm 3.0570 (3.0301) [2021-04-16 19:21:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][20/1251] eta 0:15:32 lr 0.000027 time 0.2761 (0.7572) loss 3.1636 (3.0681) grad_norm 3.5945 (3.1416) [2021-04-16 19:21:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][30/1251] eta 0:12:14 lr 0.000027 time 0.2960 (0.6016) loss 3.3725 (3.1591) grad_norm 3.1106 (3.2166) [2021-04-16 19:21:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][40/1251] eta 0:10:32 lr 0.000027 time 0.2813 (0.5222) loss 1.7903 (3.1080) grad_norm 3.5991 (3.2023) [2021-04-16 19:21:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][50/1251] eta 0:09:29 lr 0.000027 time 0.2779 (0.4746) loss 3.3626 (3.0527) grad_norm 3.1417 (3.2137) [2021-04-16 19:21:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][60/1251] eta 0:08:47 lr 0.000027 time 0.2810 (0.4432) loss 3.1936 (3.0645) grad_norm 2.8254 (3.2098) [2021-04-16 19:21:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][70/1251] eta 0:08:16 lr 0.000027 time 0.2701 (0.4206) loss 3.0653 (3.0697) grad_norm 3.4236 (3.2293) [2021-04-16 19:21:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][80/1251] eta 0:07:54 lr 0.000027 time 0.2703 (0.4050) loss 2.8778 (3.0530) grad_norm 3.2432 (3.2026) [2021-04-16 19:22:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][90/1251] eta 0:07:33 lr 0.000027 time 0.2905 (0.3910) loss 3.5691 (3.0516) grad_norm 5.6154 (3.2140) [2021-04-16 19:22:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][100/1251] eta 0:07:16 lr 0.000027 time 0.2562 (0.3793) loss 3.3111 (3.0573) grad_norm 3.3881 (3.2096) [2021-04-16 19:22:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][110/1251] eta 0:07:02 lr 0.000027 time 0.2610 (0.3699) loss 2.4673 (3.0479) grad_norm 4.5863 (3.2230) [2021-04-16 19:22:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][120/1251] eta 0:06:49 lr 0.000027 time 0.2876 (0.3623) loss 3.0880 (3.0463) grad_norm 3.0625 (3.2326) [2021-04-16 19:22:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][130/1251] eta 0:06:40 lr 0.000027 time 0.2962 (0.3572) loss 3.3623 (3.0555) grad_norm 2.9980 (3.2542) [2021-04-16 19:22:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][140/1251] eta 0:06:30 lr 0.000027 time 0.2691 (0.3512) loss 3.0233 (3.0539) grad_norm 2.6937 (3.2670) [2021-04-16 19:22:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][150/1251] eta 0:06:22 lr 0.000027 time 0.2828 (0.3474) loss 3.3525 (3.0481) grad_norm 3.2933 (3.2638) [2021-04-16 19:22:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][160/1251] eta 0:06:14 lr 0.000027 time 0.2767 (0.3429) loss 3.2074 (3.0308) grad_norm 3.3257 (3.2652) [2021-04-16 19:22:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][170/1251] eta 0:06:06 lr 0.000027 time 0.2829 (0.3390) loss 2.7214 (3.0311) grad_norm 2.7659 (3.2591) [2021-04-16 19:22:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][180/1251] eta 0:05:59 lr 0.000027 time 0.2771 (0.3356) loss 3.5913 (3.0268) grad_norm 3.0001 (3.2513) [2021-04-16 19:22:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][190/1251] eta 0:05:52 lr 0.000027 time 0.2713 (0.3325) loss 3.0720 (3.0225) grad_norm 2.7155 (3.2466) [2021-04-16 19:22:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][200/1251] eta 0:05:46 lr 0.000027 time 0.2582 (0.3297) loss 3.2337 (3.0122) grad_norm 2.7074 (3.2406) [2021-04-16 19:22:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][210/1251] eta 0:05:40 lr 0.000027 time 0.2623 (0.3273) loss 3.0928 (3.0189) grad_norm 3.5257 (3.2394) [2021-04-16 19:22:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][220/1251] eta 0:05:35 lr 0.000027 time 0.2777 (0.3251) loss 3.1012 (3.0089) grad_norm 2.9642 (3.2410) [2021-04-16 19:22:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][230/1251] eta 0:05:30 lr 0.000027 time 0.2717 (0.3239) loss 3.2518 (3.0128) grad_norm 3.4579 (3.2641) [2021-04-16 19:22:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][240/1251] eta 0:05:25 lr 0.000027 time 0.2964 (0.3219) loss 3.3472 (3.0172) grad_norm 3.3557 (3.2664) [2021-04-16 19:22:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][250/1251] eta 0:05:20 lr 0.000027 time 0.2801 (0.3200) loss 3.6368 (3.0207) grad_norm 3.1636 (3.2613) [2021-04-16 19:22:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][260/1251] eta 0:05:15 lr 0.000027 time 0.2792 (0.3183) loss 2.8528 (3.0116) grad_norm 3.5793 (3.2558) [2021-04-16 19:22:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][270/1251] eta 0:05:10 lr 0.000027 time 0.2770 (0.3166) loss 3.4402 (3.0122) grad_norm 3.3223 (3.2540) [2021-04-16 19:22:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][280/1251] eta 0:05:05 lr 0.000027 time 0.2817 (0.3151) loss 3.0024 (3.0115) grad_norm 2.9264 (3.2564) [2021-04-16 19:22:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][290/1251] eta 0:05:01 lr 0.000027 time 0.2912 (0.3142) loss 3.2294 (3.0095) grad_norm 3.1741 (3.2503) [2021-04-16 19:22:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][300/1251] eta 0:04:57 lr 0.000027 time 0.2711 (0.3127) loss 2.0421 (3.0058) grad_norm 3.5074 (3.2514) [2021-04-16 19:23:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][310/1251] eta 0:04:53 lr 0.000027 time 0.2762 (0.3117) loss 3.4677 (3.0087) grad_norm 3.7969 (3.2514) [2021-04-16 19:23:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][320/1251] eta 0:04:49 lr 0.000027 time 0.2570 (0.3109) loss 3.5594 (3.0079) grad_norm 3.1421 (3.2570) [2021-04-16 19:23:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][330/1251] eta 0:04:45 lr 0.000027 time 0.2831 (0.3098) loss 2.0308 (3.0009) grad_norm 3.4734 (3.2559) [2021-04-16 19:23:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][340/1251] eta 0:04:41 lr 0.000027 time 0.2725 (0.3089) loss 3.5822 (3.0062) grad_norm 3.2514 (3.2546) [2021-04-16 19:23:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][350/1251] eta 0:04:37 lr 0.000026 time 0.2723 (0.3081) loss 2.2403 (3.0096) grad_norm 3.1160 (3.2524) [2021-04-16 19:23:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][360/1251] eta 0:04:34 lr 0.000026 time 0.2961 (0.3081) loss 3.4496 (3.0069) grad_norm 3.1176 (3.2523) [2021-04-16 19:23:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][370/1251] eta 0:04:31 lr 0.000026 time 0.2836 (0.3082) loss 3.1002 (3.0116) grad_norm 2.7056 (3.2526) [2021-04-16 19:23:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][380/1251] eta 0:04:27 lr 0.000026 time 0.2542 (0.3073) loss 3.4537 (3.0147) grad_norm 2.9793 (3.2545) [2021-04-16 19:23:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][390/1251] eta 0:04:24 lr 0.000026 time 0.2880 (0.3071) loss 3.3792 (3.0136) grad_norm 3.5356 (3.2561) [2021-04-16 19:23:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][400/1251] eta 0:04:20 lr 0.000026 time 0.2757 (0.3067) loss 3.2486 (3.0176) grad_norm 3.7776 (3.2564) [2021-04-16 19:23:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][410/1251] eta 0:04:17 lr 0.000026 time 0.2733 (0.3058) loss 2.9975 (3.0179) grad_norm 3.2500 (3.2560) [2021-04-16 19:23:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][420/1251] eta 0:04:13 lr 0.000026 time 0.2545 (0.3052) loss 3.1743 (3.0232) grad_norm 3.7679 (3.2605) [2021-04-16 19:23:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][430/1251] eta 0:04:10 lr 0.000026 time 0.2986 (0.3047) loss 2.5553 (3.0262) grad_norm 3.1725 (3.2574) [2021-04-16 19:23:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][440/1251] eta 0:04:06 lr 0.000026 time 0.2764 (0.3040) loss 3.7751 (3.0255) grad_norm 2.7398 (3.2566) [2021-04-16 19:23:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][450/1251] eta 0:04:03 lr 0.000026 time 0.2547 (0.3034) loss 3.4558 (3.0210) grad_norm 3.1004 (3.2555) [2021-04-16 19:23:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][460/1251] eta 0:03:59 lr 0.000026 time 0.2655 (0.3028) loss 3.0148 (3.0174) grad_norm 3.2044 (3.2545) [2021-04-16 19:23:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][470/1251] eta 0:03:56 lr 0.000026 time 0.2824 (0.3023) loss 2.2390 (3.0147) grad_norm 2.9663 (3.2551) [2021-04-16 19:23:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][480/1251] eta 0:03:52 lr 0.000026 time 0.2804 (0.3017) loss 2.7406 (3.0139) grad_norm 2.5769 (3.2526) [2021-04-16 19:23:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][490/1251] eta 0:03:49 lr 0.000026 time 0.3057 (0.3012) loss 3.1427 (3.0156) grad_norm 3.3992 (3.2522) [2021-04-16 19:23:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][500/1251] eta 0:03:45 lr 0.000026 time 0.2707 (0.3008) loss 2.4206 (3.0136) grad_norm 2.9295 (3.2484) [2021-04-16 19:23:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][510/1251] eta 0:03:42 lr 0.000026 time 0.2804 (0.3003) loss 2.5462 (3.0091) grad_norm 2.7555 (3.2480) [2021-04-16 19:24:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][520/1251] eta 0:03:39 lr 0.000026 time 0.3011 (0.2999) loss 3.3112 (3.0083) grad_norm 2.8484 (3.2455) [2021-04-16 19:24:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][530/1251] eta 0:03:35 lr 0.000026 time 0.2855 (0.2995) loss 3.3922 (3.0068) grad_norm 3.3603 (3.2457) [2021-04-16 19:24:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][540/1251] eta 0:03:32 lr 0.000026 time 0.2545 (0.2991) loss 3.5249 (3.0089) grad_norm 3.0044 (3.2446) [2021-04-16 19:24:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][550/1251] eta 0:03:29 lr 0.000026 time 0.2897 (0.2989) loss 3.5604 (3.0128) grad_norm 5.2233 (3.2477) [2021-04-16 19:24:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][560/1251] eta 0:03:26 lr 0.000026 time 0.2873 (0.2984) loss 3.4509 (3.0147) grad_norm 2.6886 (3.2483) [2021-04-16 19:24:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][570/1251] eta 0:03:23 lr 0.000026 time 0.2622 (0.2983) loss 3.1041 (3.0119) grad_norm 2.9253 (3.2473) [2021-04-16 19:24:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][580/1251] eta 0:03:20 lr 0.000026 time 0.2754 (0.2981) loss 3.2583 (3.0113) grad_norm 2.9804 (3.2458) [2021-04-16 19:24:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][590/1251] eta 0:03:16 lr 0.000026 time 0.2940 (0.2979) loss 3.4513 (3.0177) grad_norm 3.7101 (3.2479) [2021-04-16 19:24:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][600/1251] eta 0:03:13 lr 0.000026 time 0.2569 (0.2976) loss 3.2173 (3.0182) grad_norm 3.1517 (3.2459) [2021-04-16 19:24:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][610/1251] eta 0:03:10 lr 0.000026 time 0.2876 (0.2973) loss 3.0666 (3.0149) grad_norm 3.5742 (3.2450) [2021-04-16 19:24:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][620/1251] eta 0:03:07 lr 0.000026 time 0.2545 (0.2969) loss 3.3273 (3.0167) grad_norm 3.0365 (3.2441) [2021-04-16 19:24:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][630/1251] eta 0:03:04 lr 0.000026 time 0.2978 (0.2965) loss 2.8726 (3.0135) grad_norm 2.9349 (3.2428) [2021-04-16 19:24:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][640/1251] eta 0:03:00 lr 0.000026 time 0.2504 (0.2962) loss 2.0338 (3.0126) grad_norm 2.9938 (3.2434) [2021-04-16 19:24:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][650/1251] eta 0:02:57 lr 0.000026 time 0.4276 (0.2961) loss 2.2089 (3.0134) grad_norm 3.2445 (3.2401) [2021-04-16 19:24:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][660/1251] eta 0:02:54 lr 0.000026 time 0.2748 (0.2957) loss 2.5771 (3.0157) grad_norm 3.5028 (3.2396) [2021-04-16 19:24:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][670/1251] eta 0:02:51 lr 0.000026 time 0.2819 (0.2954) loss 3.6906 (3.0201) grad_norm 3.4690 (3.2375) [2021-04-16 19:24:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][680/1251] eta 0:02:48 lr 0.000026 time 0.2742 (0.2950) loss 3.3535 (3.0201) grad_norm 3.2790 (3.2379) [2021-04-16 19:24:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][690/1251] eta 0:02:45 lr 0.000026 time 0.2905 (0.2948) loss 3.1728 (3.0229) grad_norm 3.0785 (3.2358) [2021-04-16 19:24:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][700/1251] eta 0:02:42 lr 0.000026 time 0.3899 (0.2948) loss 3.3179 (3.0244) grad_norm 3.3570 (3.2398) [2021-04-16 19:24:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][710/1251] eta 0:02:39 lr 0.000026 time 0.2820 (0.2945) loss 3.1729 (3.0236) grad_norm 3.2953 (3.2376) [2021-04-16 19:24:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][720/1251] eta 0:02:36 lr 0.000026 time 0.2791 (0.2942) loss 3.2743 (3.0246) grad_norm 3.5283 (3.2388) [2021-04-16 19:24:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][730/1251] eta 0:02:33 lr 0.000026 time 0.2937 (0.2942) loss 2.6393 (3.0244) grad_norm 3.1576 (3.2390) [2021-04-16 19:25:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][740/1251] eta 0:02:30 lr 0.000026 time 0.2624 (0.2939) loss 2.5832 (3.0263) grad_norm 3.2898 (3.2391) [2021-04-16 19:25:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][750/1251] eta 0:02:27 lr 0.000026 time 0.2774 (0.2937) loss 2.6665 (3.0252) grad_norm 3.0153 (3.2382) [2021-04-16 19:25:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][760/1251] eta 0:02:24 lr 0.000026 time 0.2445 (0.2934) loss 3.5159 (3.0285) grad_norm 3.0396 (3.2396) [2021-04-16 19:25:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][770/1251] eta 0:02:21 lr 0.000026 time 0.2871 (0.2932) loss 3.0816 (3.0270) grad_norm 2.7647 (3.2395) [2021-04-16 19:25:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][780/1251] eta 0:02:17 lr 0.000026 time 0.2508 (0.2929) loss 3.1114 (3.0239) grad_norm 4.9744 (3.2402) [2021-04-16 19:25:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][790/1251] eta 0:02:14 lr 0.000026 time 0.2547 (0.2927) loss 3.0705 (3.0249) grad_norm 3.4106 (3.2383) [2021-04-16 19:25:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][800/1251] eta 0:02:11 lr 0.000026 time 0.2957 (0.2925) loss 2.9415 (3.0281) grad_norm 3.5165 (3.2370) [2021-04-16 19:25:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][810/1251] eta 0:02:08 lr 0.000026 time 0.2700 (0.2925) loss 3.3325 (3.0267) grad_norm 3.4593 (3.2358) [2021-04-16 19:25:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][820/1251] eta 0:02:05 lr 0.000026 time 0.2895 (0.2923) loss 1.9187 (3.0281) grad_norm 3.0939 (3.2380) [2021-04-16 19:25:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][830/1251] eta 0:02:02 lr 0.000026 time 0.2656 (0.2920) loss 2.2127 (3.0270) grad_norm 3.8954 (3.2406) [2021-04-16 19:25:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][840/1251] eta 0:01:59 lr 0.000026 time 0.2873 (0.2918) loss 2.5682 (3.0259) grad_norm 3.5544 (3.2413) [2021-04-16 19:25:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][850/1251] eta 0:01:56 lr 0.000026 time 0.2687 (0.2916) loss 3.3001 (3.0269) grad_norm 3.5683 (3.2410) [2021-04-16 19:25:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][860/1251] eta 0:01:53 lr 0.000026 time 0.3040 (0.2914) loss 3.3775 (3.0243) grad_norm 3.2925 (3.2390) [2021-04-16 19:25:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][870/1251] eta 0:01:51 lr 0.000026 time 0.2807 (0.2913) loss 2.0440 (3.0240) grad_norm 3.1854 (3.2422) [2021-04-16 19:25:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][880/1251] eta 0:01:48 lr 0.000026 time 0.2917 (0.2912) loss 3.1028 (3.0239) grad_norm 2.7802 (3.2507) [2021-04-16 19:25:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][890/1251] eta 0:01:45 lr 0.000026 time 0.2677 (0.2911) loss 3.3126 (3.0249) grad_norm 3.0428 (3.2521) [2021-04-16 19:25:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][900/1251] eta 0:01:42 lr 0.000026 time 0.2693 (0.2909) loss 1.8701 (3.0213) grad_norm 3.0689 (3.2498) [2021-04-16 19:25:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][910/1251] eta 0:01:39 lr 0.000026 time 0.2754 (0.2907) loss 2.3328 (3.0221) grad_norm 4.7355 (3.2523) [2021-04-16 19:25:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][920/1251] eta 0:01:36 lr 0.000026 time 0.2729 (0.2906) loss 2.5826 (3.0203) grad_norm 4.5629 (3.2531) [2021-04-16 19:25:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][930/1251] eta 0:01:33 lr 0.000026 time 0.2911 (0.2905) loss 2.2094 (3.0213) grad_norm 3.0681 (3.2531) [2021-04-16 19:25:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][940/1251] eta 0:01:30 lr 0.000026 time 0.4428 (0.2908) loss 3.2383 (3.0225) grad_norm 3.2864 (3.2514) [2021-04-16 19:26:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][950/1251] eta 0:01:27 lr 0.000026 time 0.2622 (0.2907) loss 3.2813 (3.0213) grad_norm 3.2654 (3.2520) [2021-04-16 19:26:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][960/1251] eta 0:01:24 lr 0.000026 time 0.2722 (0.2905) loss 2.4992 (3.0180) grad_norm 3.0888 (3.2517) [2021-04-16 19:26:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][970/1251] eta 0:01:21 lr 0.000026 time 0.2811 (0.2905) loss 3.3286 (3.0198) grad_norm 3.0829 (3.2505) [2021-04-16 19:26:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][980/1251] eta 0:01:18 lr 0.000026 time 0.2817 (0.2904) loss 2.6669 (3.0176) grad_norm 3.3870 (3.2490) [2021-04-16 19:26:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][990/1251] eta 0:01:15 lr 0.000026 time 0.3101 (0.2904) loss 3.3453 (3.0180) grad_norm 4.0113 (3.2504) [2021-04-16 19:26:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][1000/1251] eta 0:01:12 lr 0.000026 time 0.2633 (0.2902) loss 2.5567 (3.0164) grad_norm 2.8212 (3.2510) [2021-04-16 19:26:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][1010/1251] eta 0:01:09 lr 0.000026 time 0.2740 (0.2900) loss 3.3625 (3.0169) grad_norm 2.6836 (3.2505) [2021-04-16 19:26:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][1020/1251] eta 0:01:06 lr 0.000026 time 0.2683 (0.2899) loss 4.0499 (3.0168) grad_norm 2.9357 (3.2516) [2021-04-16 19:26:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][1030/1251] eta 0:01:04 lr 0.000026 time 0.2838 (0.2898) loss 3.1231 (3.0182) grad_norm 3.3135 (3.2504) [2021-04-16 19:26:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][1040/1251] eta 0:01:01 lr 0.000026 time 0.2603 (0.2896) loss 3.4672 (3.0186) grad_norm 2.9561 (3.2485) [2021-04-16 19:26:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][1050/1251] eta 0:00:58 lr 0.000026 time 0.2835 (0.2895) loss 3.5505 (3.0172) grad_norm 2.8670 (3.2486) [2021-04-16 19:26:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][1060/1251] eta 0:00:55 lr 0.000026 time 0.2852 (0.2894) loss 3.4870 (3.0177) grad_norm 2.8829 (3.2485) [2021-04-16 19:26:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][1070/1251] eta 0:00:52 lr 0.000026 time 0.2772 (0.2893) loss 3.0527 (3.0161) grad_norm 2.7325 (3.2470) [2021-04-16 19:26:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][1080/1251] eta 0:00:49 lr 0.000026 time 0.2668 (0.2892) loss 2.9556 (3.0148) grad_norm 3.2010 (3.2477) [2021-04-16 19:26:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][1090/1251] eta 0:00:46 lr 0.000026 time 0.2871 (0.2891) loss 3.1351 (3.0139) grad_norm 4.7690 (3.2490) [2021-04-16 19:26:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][1100/1251] eta 0:00:43 lr 0.000026 time 0.2961 (0.2891) loss 1.8531 (3.0140) grad_norm 3.4319 (3.2488) [2021-04-16 19:26:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][1110/1251] eta 0:00:40 lr 0.000026 time 0.2795 (0.2889) loss 3.0691 (3.0117) grad_norm 2.9363 (3.2490) [2021-04-16 19:26:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][1120/1251] eta 0:00:37 lr 0.000026 time 0.2756 (0.2888) loss 3.5932 (3.0110) grad_norm 3.1992 (3.2482) [2021-04-16 19:26:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][1130/1251] eta 0:00:34 lr 0.000026 time 0.2668 (0.2887) loss 3.3816 (3.0112) grad_norm 3.4991 (3.2493) [2021-04-16 19:26:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][1140/1251] eta 0:00:32 lr 0.000026 time 0.2777 (0.2886) loss 2.9508 (3.0124) grad_norm 2.8799 (3.2482) [2021-04-16 19:26:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][1150/1251] eta 0:00:29 lr 0.000026 time 0.2888 (0.2885) loss 3.2714 (3.0126) grad_norm 2.8473 (3.2455) [2021-04-16 19:26:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][1160/1251] eta 0:00:26 lr 0.000026 time 0.2756 (0.2886) loss 3.5745 (3.0125) grad_norm 3.4354 (3.2458) [2021-04-16 19:27:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][1170/1251] eta 0:00:23 lr 0.000026 time 0.2722 (0.2885) loss 3.3645 (3.0138) grad_norm 3.8208 (3.2453) [2021-04-16 19:27:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][1180/1251] eta 0:00:20 lr 0.000026 time 0.2720 (0.2883) loss 1.9837 (3.0105) grad_norm 2.9007 (3.2487) [2021-04-16 19:27:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][1190/1251] eta 0:00:17 lr 0.000026 time 0.2750 (0.2883) loss 3.1903 (3.0121) grad_norm 3.3604 (3.2500) [2021-04-16 19:27:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][1200/1251] eta 0:00:14 lr 0.000026 time 0.2914 (0.2882) loss 3.4586 (3.0142) grad_norm 3.7744 (3.2514) [2021-04-16 19:27:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][1210/1251] eta 0:00:11 lr 0.000026 time 0.3041 (0.2881) loss 2.6880 (3.0166) grad_norm 3.1749 (3.2521) [2021-04-16 19:27:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][1220/1251] eta 0:00:08 lr 0.000026 time 0.2876 (0.2881) loss 2.8649 (3.0168) grad_norm 3.1259 (3.2511) [2021-04-16 19:27:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][1230/1251] eta 0:00:06 lr 0.000026 time 0.2641 (0.2881) loss 3.0908 (3.0179) grad_norm 3.9147 (3.2527) [2021-04-16 19:27:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][1240/1251] eta 0:00:03 lr 0.000026 time 0.2480 (0.2879) loss 2.9270 (3.0176) grad_norm 3.0682 (3.2536) [2021-04-16 19:27:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [275/300][1250/1251] eta 0:00:00 lr 0.000026 time 0.2525 (0.2876) loss 2.6309 (3.0153) grad_norm 5.4703 (3.2562) [2021-04-16 19:27:28 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 275 training takes 0:06:04 [2021-04-16 19:27:28 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_275.pth saving...... [2021-04-16 19:27:51 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_275.pth saved !!! [2021-04-16 19:27:52 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.122 (1.122) Loss 0.7734 (0.7734) Acc@1 80.957 (80.957) Acc@5 95.703 (95.703) [2021-04-16 19:27:53 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.162 (0.255) Loss 0.9389 (0.8309) Acc@1 79.883 (80.993) Acc@5 93.652 (95.082) [2021-04-16 19:27:56 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.258 (0.241) Loss 0.8268 (0.8192) Acc@1 80.664 (81.036) Acc@5 95.508 (95.392) [2021-04-16 19:27:58 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.108 (0.224) Loss 0.7685 (0.8141) Acc@1 81.348 (81.124) Acc@5 95.605 (95.435) [2021-04-16 19:27:59 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.222 (0.217) Loss 0.8967 (0.8177) Acc@1 80.371 (81.048) Acc@5 94.727 (95.472) [2021-04-16 19:28:19 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 81.062 Acc@5 95.486 [2021-04-16 19:28:19 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 81.1% [2021-04-16 19:28:19 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 81.06% [2021-04-16 19:28:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][0/1251] eta 3:26:14 lr 0.000026 time 9.8917 (9.8917) loss 3.4446 (3.4446) grad_norm 2.8576 (2.8576) [2021-04-16 19:28:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][10/1251] eta 0:23:47 lr 0.000026 time 0.2905 (1.1500) loss 2.7089 (2.9999) grad_norm 3.0756 (3.0078) [2021-04-16 19:28:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][20/1251] eta 0:15:14 lr 0.000026 time 0.2431 (0.7432) loss 3.4443 (3.0333) grad_norm 3.3437 (3.1363) [2021-04-16 19:28:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][30/1251] eta 0:12:06 lr 0.000026 time 0.2757 (0.5952) loss 3.5170 (2.9726) grad_norm 3.8533 (3.1281) [2021-04-16 19:28:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][40/1251] eta 0:10:29 lr 0.000026 time 0.2694 (0.5196) loss 2.6912 (3.0055) grad_norm 2.7043 (3.1465) [2021-04-16 19:28:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][50/1251] eta 0:09:26 lr 0.000025 time 0.2629 (0.4720) loss 2.9215 (2.9674) grad_norm 2.4540 (3.1026) [2021-04-16 19:28:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][60/1251] eta 0:08:44 lr 0.000025 time 0.2708 (0.4400) loss 3.3545 (2.9447) grad_norm 2.8347 (3.1219) [2021-04-16 19:28:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][70/1251] eta 0:08:12 lr 0.000025 time 0.2875 (0.4170) loss 2.5105 (2.9423) grad_norm 2.8319 (3.1298) [2021-04-16 19:28:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][80/1251] eta 0:07:50 lr 0.000025 time 0.2932 (0.4018) loss 1.9801 (2.9040) grad_norm 3.9384 (3.1556) [2021-04-16 19:28:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][90/1251] eta 0:07:32 lr 0.000025 time 0.2916 (0.3896) loss 3.1290 (2.9045) grad_norm 3.2328 (3.1726) [2021-04-16 19:28:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][100/1251] eta 0:07:15 lr 0.000025 time 0.2610 (0.3785) loss 3.0673 (2.9000) grad_norm 3.4675 (3.1834) [2021-04-16 19:29:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][110/1251] eta 0:07:01 lr 0.000025 time 0.3175 (0.3697) loss 3.1225 (2.9060) grad_norm 2.9645 (3.1970) [2021-04-16 19:29:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][120/1251] eta 0:06:49 lr 0.000025 time 0.2782 (0.3618) loss 2.9095 (2.9006) grad_norm 3.8672 (3.2034) [2021-04-16 19:29:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][130/1251] eta 0:06:38 lr 0.000025 time 0.2630 (0.3554) loss 2.9824 (2.9125) grad_norm 2.6295 (3.1886) [2021-04-16 19:29:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][140/1251] eta 0:06:31 lr 0.000025 time 0.2650 (0.3520) loss 1.5771 (2.9128) grad_norm 2.7808 (3.1921) [2021-04-16 19:29:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][150/1251] eta 0:06:24 lr 0.000025 time 0.5098 (0.3491) loss 2.8813 (2.9271) grad_norm 3.6238 (3.1920) [2021-04-16 19:29:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][160/1251] eta 0:06:15 lr 0.000025 time 0.2765 (0.3444) loss 3.4829 (2.9406) grad_norm 2.6113 (3.1797) [2021-04-16 19:29:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][170/1251] eta 0:06:08 lr 0.000025 time 0.2725 (0.3406) loss 2.0503 (2.9339) grad_norm 3.1743 (3.1835) [2021-04-16 19:29:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][180/1251] eta 0:06:01 lr 0.000025 time 0.2989 (0.3380) loss 3.2205 (2.9376) grad_norm 4.3289 (3.1983) [2021-04-16 19:29:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][190/1251] eta 0:05:55 lr 0.000025 time 0.2858 (0.3347) loss 2.8389 (2.9469) grad_norm 3.4348 (3.1970) [2021-04-16 19:29:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][200/1251] eta 0:05:48 lr 0.000025 time 0.2669 (0.3316) loss 3.5129 (2.9470) grad_norm 2.9965 (3.2005) [2021-04-16 19:29:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][210/1251] eta 0:05:42 lr 0.000025 time 0.2754 (0.3289) loss 3.1873 (2.9414) grad_norm 3.2781 (3.1981) [2021-04-16 19:29:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][220/1251] eta 0:05:36 lr 0.000025 time 0.2872 (0.3265) loss 3.3028 (2.9578) grad_norm 2.7422 (3.1977) [2021-04-16 19:29:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][230/1251] eta 0:05:31 lr 0.000025 time 0.2710 (0.3244) loss 2.6444 (2.9578) grad_norm 3.8070 (3.2045) [2021-04-16 19:29:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][240/1251] eta 0:05:25 lr 0.000025 time 0.2674 (0.3223) loss 3.6487 (2.9571) grad_norm 2.8335 (3.2026) [2021-04-16 19:29:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][250/1251] eta 0:05:22 lr 0.000025 time 0.2655 (0.3224) loss 2.3043 (2.9524) grad_norm 4.4712 (3.2173) [2021-04-16 19:29:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][260/1251] eta 0:05:17 lr 0.000025 time 0.2588 (0.3208) loss 3.5068 (2.9488) grad_norm 3.2379 (3.2143) [2021-04-16 19:29:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][270/1251] eta 0:05:13 lr 0.000025 time 0.2858 (0.3191) loss 3.1640 (2.9563) grad_norm 2.8141 (3.2180) [2021-04-16 19:29:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][280/1251] eta 0:05:08 lr 0.000025 time 0.2768 (0.3176) loss 2.7646 (2.9543) grad_norm 2.6768 (3.2164) [2021-04-16 19:29:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][290/1251] eta 0:05:03 lr 0.000025 time 0.2731 (0.3162) loss 3.4483 (2.9630) grad_norm 3.3987 (3.2221) [2021-04-16 19:29:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][300/1251] eta 0:04:59 lr 0.000025 time 0.2727 (0.3149) loss 2.4605 (2.9609) grad_norm 3.3206 (3.2254) [2021-04-16 19:29:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][310/1251] eta 0:04:55 lr 0.000025 time 0.2749 (0.3135) loss 3.0839 (2.9639) grad_norm 3.2476 (3.2187) [2021-04-16 19:29:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][320/1251] eta 0:04:50 lr 0.000025 time 0.2955 (0.3123) loss 2.6003 (2.9658) grad_norm 2.7061 (3.2223) [2021-04-16 19:30:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][330/1251] eta 0:04:46 lr 0.000025 time 0.2880 (0.3112) loss 2.1364 (2.9686) grad_norm 3.4815 (3.2184) [2021-04-16 19:30:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][340/1251] eta 0:04:42 lr 0.000025 time 0.2669 (0.3100) loss 2.7203 (2.9656) grad_norm 3.3329 (3.2144) [2021-04-16 19:30:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][350/1251] eta 0:04:38 lr 0.000025 time 0.2695 (0.3091) loss 3.2675 (2.9678) grad_norm 3.4684 (3.2139) [2021-04-16 19:30:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][360/1251] eta 0:04:35 lr 0.000025 time 0.2889 (0.3092) loss 2.6153 (2.9631) grad_norm 2.8760 (3.2147) [2021-04-16 19:30:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][370/1251] eta 0:04:31 lr 0.000025 time 0.2805 (0.3084) loss 2.5639 (2.9633) grad_norm 3.2705 (3.2191) [2021-04-16 19:30:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][380/1251] eta 0:04:27 lr 0.000025 time 0.2778 (0.3076) loss 3.6697 (2.9643) grad_norm 3.4522 (3.2180) [2021-04-16 19:30:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][390/1251] eta 0:04:24 lr 0.000025 time 0.2543 (0.3067) loss 1.6530 (2.9555) grad_norm 3.1028 (3.2186) [2021-04-16 19:30:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][400/1251] eta 0:04:20 lr 0.000025 time 0.2661 (0.3060) loss 2.5678 (2.9600) grad_norm 3.6528 (3.2180) [2021-04-16 19:30:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][410/1251] eta 0:04:16 lr 0.000025 time 0.2682 (0.3053) loss 3.4075 (2.9578) grad_norm 3.5745 (3.2234) [2021-04-16 19:30:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][420/1251] eta 0:04:13 lr 0.000025 time 0.2631 (0.3045) loss 2.1566 (2.9602) grad_norm 3.3078 (3.2246) [2021-04-16 19:30:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][430/1251] eta 0:04:09 lr 0.000025 time 0.2897 (0.3044) loss 3.3282 (2.9607) grad_norm 3.6579 (3.2381) [2021-04-16 19:30:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][440/1251] eta 0:04:06 lr 0.000025 time 0.2577 (0.3038) loss 2.5593 (2.9636) grad_norm 3.1581 (3.2430) [2021-04-16 19:30:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][450/1251] eta 0:04:02 lr 0.000025 time 0.2432 (0.3031) loss 3.1905 (2.9675) grad_norm 3.5232 (3.2431) [2021-04-16 19:30:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][460/1251] eta 0:03:59 lr 0.000025 time 0.2970 (0.3027) loss 2.6216 (2.9650) grad_norm 9.0218 (3.2584) [2021-04-16 19:30:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][470/1251] eta 0:03:55 lr 0.000025 time 0.2600 (0.3021) loss 3.2744 (2.9652) grad_norm 3.1611 (3.2693) [2021-04-16 19:30:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][480/1251] eta 0:03:52 lr 0.000025 time 0.2764 (0.3017) loss 2.1116 (2.9646) grad_norm 3.5582 (3.2685) [2021-04-16 19:30:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][490/1251] eta 0:03:49 lr 0.000025 time 0.3065 (0.3011) loss 2.5576 (2.9611) grad_norm 3.0023 (3.2664) [2021-04-16 19:30:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][500/1251] eta 0:03:45 lr 0.000025 time 0.3071 (0.3006) loss 3.1368 (2.9619) grad_norm 3.3509 (3.2674) [2021-04-16 19:30:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][510/1251] eta 0:03:42 lr 0.000025 time 0.2946 (0.3002) loss 1.9814 (2.9589) grad_norm 2.8875 (3.2664) [2021-04-16 19:30:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][520/1251] eta 0:03:39 lr 0.000025 time 0.3093 (0.2998) loss 3.5797 (2.9580) grad_norm 3.5178 (3.2623) [2021-04-16 19:30:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][530/1251] eta 0:03:35 lr 0.000025 time 0.2546 (0.2993) loss 2.8337 (2.9644) grad_norm 3.3958 (3.2628) [2021-04-16 19:31:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][540/1251] eta 0:03:32 lr 0.000025 time 0.2697 (0.2989) loss 3.3621 (2.9635) grad_norm 3.4882 (3.2625) [2021-04-16 19:31:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][550/1251] eta 0:03:29 lr 0.000025 time 0.2953 (0.2985) loss 1.9348 (2.9625) grad_norm 3.7572 (3.2745) [2021-04-16 19:31:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][560/1251] eta 0:03:26 lr 0.000025 time 0.2655 (0.2982) loss 3.4950 (2.9624) grad_norm 3.4328 (3.3036) [2021-04-16 19:31:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][570/1251] eta 0:03:23 lr 0.000025 time 0.3214 (0.2982) loss 2.6122 (2.9645) grad_norm 3.2824 (3.3050) [2021-04-16 19:31:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][580/1251] eta 0:03:19 lr 0.000025 time 0.2730 (0.2978) loss 2.1734 (2.9634) grad_norm 3.5758 (3.3045) [2021-04-16 19:31:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][590/1251] eta 0:03:16 lr 0.000025 time 0.2682 (0.2976) loss 2.6130 (2.9654) grad_norm 3.4151 (3.3045) [2021-04-16 19:31:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][600/1251] eta 0:03:13 lr 0.000025 time 0.2935 (0.2973) loss 2.7946 (2.9695) grad_norm 2.6485 (3.3027) [2021-04-16 19:31:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][610/1251] eta 0:03:10 lr 0.000025 time 0.2898 (0.2970) loss 3.0531 (2.9687) grad_norm 3.0689 (3.3013) [2021-04-16 19:31:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][620/1251] eta 0:03:07 lr 0.000025 time 0.2488 (0.2966) loss 2.9211 (2.9706) grad_norm 3.6002 (3.3014) [2021-04-16 19:31:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][630/1251] eta 0:03:03 lr 0.000025 time 0.2644 (0.2962) loss 2.3863 (2.9714) grad_norm 3.2131 (3.3049) [2021-04-16 19:31:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][640/1251] eta 0:03:00 lr 0.000025 time 0.2867 (0.2960) loss 2.8157 (2.9720) grad_norm 3.2754 (3.3013) [2021-04-16 19:31:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][650/1251] eta 0:02:57 lr 0.000025 time 0.2766 (0.2959) loss 1.9683 (2.9700) grad_norm 2.9001 (3.3106) [2021-04-16 19:31:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][660/1251] eta 0:02:54 lr 0.000025 time 0.2900 (0.2956) loss 3.1220 (2.9693) grad_norm 3.9799 (3.3083) [2021-04-16 19:31:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][670/1251] eta 0:02:51 lr 0.000025 time 0.2958 (0.2955) loss 3.2578 (2.9699) grad_norm 3.0722 (3.3124) [2021-04-16 19:31:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][680/1251] eta 0:02:48 lr 0.000025 time 0.2874 (0.2952) loss 2.7882 (2.9701) grad_norm 3.0160 (3.3108) [2021-04-16 19:31:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][690/1251] eta 0:02:45 lr 0.000025 time 0.2692 (0.2949) loss 2.8929 (2.9705) grad_norm 3.5167 (3.3120) [2021-04-16 19:31:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][700/1251] eta 0:02:42 lr 0.000025 time 0.2934 (0.2946) loss 3.3830 (2.9730) grad_norm 3.2292 (3.3111) [2021-04-16 19:31:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][710/1251] eta 0:02:39 lr 0.000025 time 0.2518 (0.2944) loss 2.6475 (2.9746) grad_norm 2.9204 (3.3099) [2021-04-16 19:31:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][720/1251] eta 0:02:36 lr 0.000025 time 0.2468 (0.2946) loss 3.1336 (2.9776) grad_norm 2.8068 (3.3169) [2021-04-16 19:31:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][730/1251] eta 0:02:33 lr 0.000025 time 0.2542 (0.2947) loss 2.7591 (2.9815) grad_norm 3.6927 (3.3155) [2021-04-16 19:31:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][740/1251] eta 0:02:30 lr 0.000025 time 0.3456 (0.2945) loss 2.4459 (2.9818) grad_norm 2.6142 (3.3157) [2021-04-16 19:32:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][750/1251] eta 0:02:27 lr 0.000025 time 0.2773 (0.2942) loss 3.1018 (2.9782) grad_norm 4.1819 (3.3199) [2021-04-16 19:32:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][760/1251] eta 0:02:24 lr 0.000025 time 0.2692 (0.2940) loss 2.5854 (2.9756) grad_norm 2.7953 (3.3185) [2021-04-16 19:32:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][770/1251] eta 0:02:21 lr 0.000025 time 0.2897 (0.2938) loss 3.3997 (2.9735) grad_norm 2.9767 (3.3158) [2021-04-16 19:32:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][780/1251] eta 0:02:18 lr 0.000025 time 0.2586 (0.2936) loss 2.2377 (2.9740) grad_norm 3.4118 (3.3169) [2021-04-16 19:32:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][790/1251] eta 0:02:15 lr 0.000025 time 0.2522 (0.2933) loss 2.2211 (2.9782) grad_norm 3.3859 (3.3137) [2021-04-16 19:32:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][800/1251] eta 0:02:12 lr 0.000025 time 0.2534 (0.2932) loss 3.6059 (2.9787) grad_norm 3.8339 (3.3126) [2021-04-16 19:32:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][810/1251] eta 0:02:09 lr 0.000025 time 0.2699 (0.2930) loss 2.8990 (2.9797) grad_norm 3.0771 (3.3143) [2021-04-16 19:32:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][820/1251] eta 0:02:06 lr 0.000025 time 0.2901 (0.2929) loss 2.8236 (2.9811) grad_norm 3.0324 (3.3142) [2021-04-16 19:32:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][830/1251] eta 0:02:03 lr 0.000025 time 0.2623 (0.2926) loss 3.0330 (2.9831) grad_norm 3.8229 (3.3130) [2021-04-16 19:32:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][840/1251] eta 0:02:00 lr 0.000025 time 0.2900 (0.2924) loss 2.8081 (2.9833) grad_norm 8.4772 (3.3174) [2021-04-16 19:32:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][850/1251] eta 0:01:57 lr 0.000025 time 0.2507 (0.2923) loss 3.5113 (2.9853) grad_norm 3.3889 (3.3191) [2021-04-16 19:32:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][860/1251] eta 0:01:54 lr 0.000025 time 0.3203 (0.2922) loss 1.9186 (2.9819) grad_norm 3.0817 (3.3166) [2021-04-16 19:32:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][870/1251] eta 0:01:51 lr 0.000025 time 0.2741 (0.2921) loss 3.5474 (2.9832) grad_norm 3.1366 (3.3157) [2021-04-16 19:32:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][880/1251] eta 0:01:48 lr 0.000025 time 0.2701 (0.2920) loss 2.6850 (2.9824) grad_norm 3.6197 (3.3155) [2021-04-16 19:32:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][890/1251] eta 0:01:45 lr 0.000025 time 0.2721 (0.2918) loss 2.5902 (2.9810) grad_norm 3.2443 (3.3148) [2021-04-16 19:32:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][900/1251] eta 0:01:42 lr 0.000025 time 0.2785 (0.2917) loss 3.2726 (2.9847) grad_norm 2.9745 (3.3197) [2021-04-16 19:32:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][910/1251] eta 0:01:39 lr 0.000025 time 0.2827 (0.2915) loss 2.8675 (2.9844) grad_norm 2.7397 (3.3184) [2021-04-16 19:32:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][920/1251] eta 0:01:36 lr 0.000025 time 0.2829 (0.2913) loss 3.0612 (2.9832) grad_norm 3.1354 (3.3177) [2021-04-16 19:32:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][930/1251] eta 0:01:33 lr 0.000025 time 0.2991 (0.2912) loss 3.2208 (2.9851) grad_norm 2.8096 (3.3189) [2021-04-16 19:32:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][940/1251] eta 0:01:30 lr 0.000025 time 0.2645 (0.2912) loss 2.6124 (2.9864) grad_norm 2.8719 (3.3189) [2021-04-16 19:32:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][950/1251] eta 0:01:27 lr 0.000025 time 0.2854 (0.2911) loss 2.4201 (2.9862) grad_norm 3.9746 (3.3172) [2021-04-16 19:32:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][960/1251] eta 0:01:24 lr 0.000025 time 0.2949 (0.2910) loss 3.5206 (2.9880) grad_norm 4.0439 (3.3171) [2021-04-16 19:33:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][970/1251] eta 0:01:21 lr 0.000025 time 0.2709 (0.2908) loss 2.9543 (2.9902) grad_norm 3.2245 (3.3173) [2021-04-16 19:33:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][980/1251] eta 0:01:18 lr 0.000025 time 0.2835 (0.2907) loss 2.7587 (2.9904) grad_norm 2.8054 (3.3149) [2021-04-16 19:33:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][990/1251] eta 0:01:15 lr 0.000025 time 0.2882 (0.2905) loss 3.0932 (2.9899) grad_norm 2.9623 (3.3152) [2021-04-16 19:33:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][1000/1251] eta 0:01:12 lr 0.000025 time 0.2683 (0.2904) loss 3.3568 (2.9910) grad_norm 3.3173 (3.3139) [2021-04-16 19:33:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][1010/1251] eta 0:01:09 lr 0.000025 time 0.2938 (0.2904) loss 2.5689 (2.9894) grad_norm 2.9333 (3.3152) [2021-04-16 19:33:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][1020/1251] eta 0:01:07 lr 0.000025 time 0.2809 (0.2902) loss 3.3464 (2.9904) grad_norm 2.9922 (3.3154) [2021-04-16 19:33:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][1030/1251] eta 0:01:04 lr 0.000025 time 0.2743 (0.2901) loss 3.3985 (2.9913) grad_norm 3.1163 (3.3172) [2021-04-16 19:33:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][1040/1251] eta 0:01:01 lr 0.000024 time 0.2699 (0.2900) loss 2.7565 (2.9909) grad_norm 3.0899 (3.3154) [2021-04-16 19:33:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][1050/1251] eta 0:00:58 lr 0.000024 time 0.2621 (0.2898) loss 3.5750 (2.9905) grad_norm 2.9666 (3.3135) [2021-04-16 19:33:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][1060/1251] eta 0:00:55 lr 0.000024 time 0.2448 (0.2897) loss 2.9179 (2.9914) grad_norm 3.2047 (3.3122) [2021-04-16 19:33:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][1070/1251] eta 0:00:52 lr 0.000024 time 0.2760 (0.2896) loss 2.3804 (2.9921) grad_norm 3.3177 (3.3111) [2021-04-16 19:33:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][1080/1251] eta 0:00:49 lr 0.000024 time 0.2582 (0.2894) loss 3.6201 (2.9934) grad_norm 2.8896 (3.3103) [2021-04-16 19:33:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][1090/1251] eta 0:00:46 lr 0.000024 time 0.3073 (0.2894) loss 2.0996 (2.9931) grad_norm 3.2424 (3.3089) [2021-04-16 19:33:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][1100/1251] eta 0:00:43 lr 0.000024 time 0.2961 (0.2893) loss 2.5065 (2.9940) grad_norm 3.2264 (3.3081) [2021-04-16 19:33:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][1110/1251] eta 0:00:40 lr 0.000024 time 0.2953 (0.2892) loss 2.7011 (2.9949) grad_norm 2.7705 (3.3075) [2021-04-16 19:33:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][1120/1251] eta 0:00:37 lr 0.000024 time 0.2861 (0.2890) loss 3.3680 (2.9939) grad_norm 2.6782 (3.3045) [2021-04-16 19:33:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][1130/1251] eta 0:00:34 lr 0.000024 time 0.2668 (0.2889) loss 3.0819 (2.9956) grad_norm 3.1021 (3.3038) [2021-04-16 19:33:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][1140/1251] eta 0:00:32 lr 0.000024 time 0.3097 (0.2888) loss 2.6268 (2.9962) grad_norm 3.5041 (3.3031) [2021-04-16 19:33:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][1150/1251] eta 0:00:29 lr 0.000024 time 0.2932 (0.2889) loss 2.7070 (2.9962) grad_norm 3.4506 (3.3031) [2021-04-16 19:33:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][1160/1251] eta 0:00:26 lr 0.000024 time 0.2868 (0.2888) loss 3.0980 (2.9971) grad_norm 3.1757 (3.3017) [2021-04-16 19:33:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][1170/1251] eta 0:00:23 lr 0.000024 time 0.2627 (0.2888) loss 3.2469 (2.9973) grad_norm 3.2876 (3.3012) [2021-04-16 19:34:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][1180/1251] eta 0:00:20 lr 0.000024 time 0.2846 (0.2887) loss 3.1678 (2.9986) grad_norm 3.2401 (3.2993) [2021-04-16 19:34:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][1190/1251] eta 0:00:17 lr 0.000024 time 0.2955 (0.2887) loss 3.3754 (3.0008) grad_norm 2.9636 (3.2994) [2021-04-16 19:34:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][1200/1251] eta 0:00:14 lr 0.000024 time 0.2920 (0.2886) loss 3.3707 (2.9996) grad_norm 3.5821 (3.2992) [2021-04-16 19:34:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][1210/1251] eta 0:00:11 lr 0.000024 time 0.2452 (0.2885) loss 3.1076 (2.9980) grad_norm 3.4740 (3.3011) [2021-04-16 19:34:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][1220/1251] eta 0:00:08 lr 0.000024 time 0.2930 (0.2885) loss 3.4146 (3.0001) grad_norm 3.4039 (3.3019) [2021-04-16 19:34:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][1230/1251] eta 0:00:06 lr 0.000024 time 0.2663 (0.2884) loss 3.0273 (2.9997) grad_norm 3.5174 (3.3040) [2021-04-16 19:34:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][1240/1251] eta 0:00:03 lr 0.000024 time 0.3411 (0.2883) loss 2.2958 (3.0006) grad_norm 3.4316 (3.3031) [2021-04-16 19:34:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [276/300][1250/1251] eta 0:00:00 lr 0.000024 time 0.2485 (0.2880) loss 3.3756 (3.0009) grad_norm 3.1628 (3.3012) [2021-04-16 19:34:27 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 276 training takes 0:06:07 [2021-04-16 19:34:27 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_276.pth saving...... [2021-04-16 19:34:59 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_276.pth saved !!! [2021-04-16 19:35:00 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.159 (1.159) Loss 0.7861 (0.7861) Acc@1 81.543 (81.543) Acc@5 95.508 (95.508) [2021-04-16 19:35:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.327 (0.229) Loss 0.8092 (0.8260) Acc@1 81.348 (80.771) Acc@5 95.215 (95.241) [2021-04-16 19:35:04 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.244 (0.242) Loss 0.8415 (0.8121) Acc@1 79.980 (81.110) Acc@5 95.215 (95.489) [2021-04-16 19:35:07 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.100 (0.253) Loss 0.8239 (0.8182) Acc@1 80.859 (80.954) Acc@5 95.703 (95.439) [2021-04-16 19:35:08 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.219) Loss 0.7774 (0.8139) Acc@1 81.250 (81.136) Acc@5 96.387 (95.482) [2021-04-16 19:35:30 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 81.022 Acc@5 95.432 [2021-04-16 19:35:30 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 81.0% [2021-04-16 19:35:30 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 81.06% [2021-04-16 19:35:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][0/1251] eta 3:29:33 lr 0.000024 time 10.0509 (10.0509) loss 3.2751 (3.2751) grad_norm 2.7852 (2.7852) [2021-04-16 19:35:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][10/1251] eta 0:24:02 lr 0.000024 time 0.2502 (1.1623) loss 3.3487 (3.0639) grad_norm 3.3314 (3.0518) [2021-04-16 19:35:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][20/1251] eta 0:15:11 lr 0.000024 time 0.2706 (0.7407) loss 3.2021 (2.9269) grad_norm 2.9647 (3.1522) [2021-04-16 19:35:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][30/1251] eta 0:12:04 lr 0.000024 time 0.2879 (0.5933) loss 3.2423 (2.9638) grad_norm 2.9234 (3.1539) [2021-04-16 19:35:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][40/1251] eta 0:10:28 lr 0.000024 time 0.3053 (0.5187) loss 2.7193 (2.9919) grad_norm 3.2674 (3.1971) [2021-04-16 19:35:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][50/1251] eta 0:09:25 lr 0.000024 time 0.2627 (0.4706) loss 3.3987 (2.9788) grad_norm 3.8837 (3.2201) [2021-04-16 19:35:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][60/1251] eta 0:08:42 lr 0.000024 time 0.2925 (0.4384) loss 3.4725 (2.9369) grad_norm 3.6917 (3.2127) [2021-04-16 19:35:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][70/1251] eta 0:08:11 lr 0.000024 time 0.3020 (0.4158) loss 2.1737 (2.9500) grad_norm 3.7238 (3.2232) [2021-04-16 19:36:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][80/1251] eta 0:07:46 lr 0.000024 time 0.2760 (0.3987) loss 2.9285 (2.9195) grad_norm 3.2086 (3.2425) [2021-04-16 19:36:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][90/1251] eta 0:07:27 lr 0.000024 time 0.2834 (0.3856) loss 3.5023 (2.9370) grad_norm 3.2291 (3.2316) [2021-04-16 19:36:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][100/1251] eta 0:07:12 lr 0.000024 time 0.2966 (0.3758) loss 3.3320 (2.9429) grad_norm 3.4116 (3.2351) [2021-04-16 19:36:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][110/1251] eta 0:06:59 lr 0.000024 time 0.2594 (0.3672) loss 2.5536 (2.9255) grad_norm 2.8739 (3.2091) [2021-04-16 19:36:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][120/1251] eta 0:06:47 lr 0.000024 time 0.2722 (0.3602) loss 3.3934 (2.9229) grad_norm 2.9909 (3.1956) [2021-04-16 19:36:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][130/1251] eta 0:06:36 lr 0.000024 time 0.2682 (0.3539) loss 3.5173 (2.9318) grad_norm 2.9570 (3.1828) [2021-04-16 19:36:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][140/1251] eta 0:06:29 lr 0.000024 time 0.3185 (0.3503) loss 3.5991 (2.9343) grad_norm 3.1830 (3.1697) [2021-04-16 19:36:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][150/1251] eta 0:06:20 lr 0.000024 time 0.2564 (0.3460) loss 3.3569 (2.9364) grad_norm 3.2024 (3.1855) [2021-04-16 19:36:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][160/1251] eta 0:06:13 lr 0.000024 time 0.2687 (0.3425) loss 2.5941 (2.9439) grad_norm 3.0244 (3.2034) [2021-04-16 19:36:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][170/1251] eta 0:06:06 lr 0.000024 time 0.2724 (0.3390) loss 2.5159 (2.9439) grad_norm 3.1563 (3.2036) [2021-04-16 19:36:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][180/1251] eta 0:05:59 lr 0.000024 time 0.2784 (0.3357) loss 3.0091 (2.9582) grad_norm 3.5867 (3.2112) [2021-04-16 19:36:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][190/1251] eta 0:05:52 lr 0.000024 time 0.2682 (0.3324) loss 1.8285 (2.9613) grad_norm 2.8452 (3.2062) [2021-04-16 19:36:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][200/1251] eta 0:05:46 lr 0.000024 time 0.2759 (0.3299) loss 3.4310 (2.9619) grad_norm 2.8046 (3.2100) [2021-04-16 19:36:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][210/1251] eta 0:05:40 lr 0.000024 time 0.2536 (0.3274) loss 3.1165 (2.9567) grad_norm 2.9053 (3.2252) [2021-04-16 19:36:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][220/1251] eta 0:05:35 lr 0.000024 time 0.2579 (0.3257) loss 2.4061 (2.9518) grad_norm 3.3634 (3.2434) [2021-04-16 19:36:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][230/1251] eta 0:05:30 lr 0.000024 time 0.3269 (0.3240) loss 2.6202 (2.9461) grad_norm 3.5807 (3.2412) [2021-04-16 19:36:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][240/1251] eta 0:05:25 lr 0.000024 time 0.3133 (0.3223) loss 2.8740 (2.9399) grad_norm 3.1592 (3.2368) [2021-04-16 19:36:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][250/1251] eta 0:05:21 lr 0.000024 time 0.3046 (0.3208) loss 3.3270 (2.9368) grad_norm 2.9842 (3.2315) [2021-04-16 19:36:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][260/1251] eta 0:05:16 lr 0.000024 time 0.2938 (0.3191) loss 2.9307 (2.9350) grad_norm 3.6092 (3.2310) [2021-04-16 19:36:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][270/1251] eta 0:05:11 lr 0.000024 time 0.2768 (0.3175) loss 2.9323 (2.9481) grad_norm 3.8118 (3.2298) [2021-04-16 19:36:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][280/1251] eta 0:05:06 lr 0.000024 time 0.2750 (0.3160) loss 2.6955 (2.9478) grad_norm 2.9445 (3.2312) [2021-04-16 19:37:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][290/1251] eta 0:05:02 lr 0.000024 time 0.2872 (0.3147) loss 3.3788 (2.9529) grad_norm 3.2944 (3.2301) [2021-04-16 19:37:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][300/1251] eta 0:04:58 lr 0.000024 time 0.2890 (0.3134) loss 1.8058 (2.9507) grad_norm 4.2761 (3.2456) [2021-04-16 19:37:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][310/1251] eta 0:04:53 lr 0.000024 time 0.2719 (0.3123) loss 3.3093 (2.9553) grad_norm 4.0490 (3.2453) [2021-04-16 19:37:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][320/1251] eta 0:04:49 lr 0.000024 time 0.2720 (0.3113) loss 3.3600 (2.9544) grad_norm 8.7595 (3.2617) [2021-04-16 19:37:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][330/1251] eta 0:04:45 lr 0.000024 time 0.2812 (0.3103) loss 3.2314 (2.9530) grad_norm 3.5302 (3.2598) [2021-04-16 19:37:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][340/1251] eta 0:04:41 lr 0.000024 time 0.2747 (0.3095) loss 3.1590 (2.9473) grad_norm 3.4028 (3.2604) [2021-04-16 19:37:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][350/1251] eta 0:04:38 lr 0.000024 time 0.2931 (0.3086) loss 3.0670 (2.9531) grad_norm 3.9019 (3.2620) [2021-04-16 19:37:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][360/1251] eta 0:04:34 lr 0.000024 time 0.3108 (0.3080) loss 2.9988 (2.9462) grad_norm 2.9034 (3.2630) [2021-04-16 19:37:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][370/1251] eta 0:04:31 lr 0.000024 time 0.2569 (0.3081) loss 2.7747 (2.9537) grad_norm 2.8265 (3.2646) [2021-04-16 19:37:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][380/1251] eta 0:04:27 lr 0.000024 time 0.2649 (0.3074) loss 2.9705 (2.9546) grad_norm 2.7532 (3.2604) [2021-04-16 19:37:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][390/1251] eta 0:04:24 lr 0.000024 time 0.2981 (0.3068) loss 2.7446 (2.9613) grad_norm 3.4253 (3.2610) [2021-04-16 19:37:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][400/1251] eta 0:04:20 lr 0.000024 time 0.2861 (0.3060) loss 2.0453 (2.9570) grad_norm 3.0032 (3.2581) [2021-04-16 19:37:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][410/1251] eta 0:04:16 lr 0.000024 time 0.2669 (0.3054) loss 3.0106 (2.9634) grad_norm 3.1316 (3.2592) [2021-04-16 19:37:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][420/1251] eta 0:04:13 lr 0.000024 time 0.2751 (0.3047) loss 3.0200 (2.9631) grad_norm 2.7536 (3.2603) [2021-04-16 19:37:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][430/1251] eta 0:04:09 lr 0.000024 time 0.2746 (0.3044) loss 3.3956 (2.9700) grad_norm 3.1894 (3.2557) [2021-04-16 19:37:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][440/1251] eta 0:04:06 lr 0.000024 time 0.2935 (0.3038) loss 3.2550 (2.9673) grad_norm 3.4523 (3.2554) [2021-04-16 19:37:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][450/1251] eta 0:04:03 lr 0.000024 time 0.3942 (0.3035) loss 1.9996 (2.9697) grad_norm 3.1948 (3.2577) [2021-04-16 19:37:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][460/1251] eta 0:03:59 lr 0.000024 time 0.2843 (0.3029) loss 3.8437 (2.9739) grad_norm 2.9382 (3.2650) [2021-04-16 19:37:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][470/1251] eta 0:03:56 lr 0.000024 time 0.2600 (0.3023) loss 3.3643 (2.9696) grad_norm 3.3284 (3.2603) [2021-04-16 19:37:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][480/1251] eta 0:03:52 lr 0.000024 time 0.2659 (0.3018) loss 3.0924 (2.9673) grad_norm 3.0189 (3.2665) [2021-04-16 19:37:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][490/1251] eta 0:03:49 lr 0.000024 time 0.2698 (0.3011) loss 3.2502 (2.9674) grad_norm 3.7729 (3.2743) [2021-04-16 19:38:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][500/1251] eta 0:03:45 lr 0.000024 time 0.2589 (0.3006) loss 3.1225 (2.9667) grad_norm 3.0942 (3.2711) [2021-04-16 19:38:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][510/1251] eta 0:03:42 lr 0.000024 time 0.2778 (0.3002) loss 3.0156 (2.9668) grad_norm 3.1797 (3.2708) [2021-04-16 19:38:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][520/1251] eta 0:03:39 lr 0.000024 time 0.2792 (0.2997) loss 2.2511 (2.9671) grad_norm 3.4307 (3.2680) [2021-04-16 19:38:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][530/1251] eta 0:03:35 lr 0.000024 time 0.2862 (0.2994) loss 3.3749 (2.9704) grad_norm 3.2683 (3.2698) [2021-04-16 19:38:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][540/1251] eta 0:03:32 lr 0.000024 time 0.2512 (0.2992) loss 3.3532 (2.9731) grad_norm 3.1022 (3.2647) [2021-04-16 19:38:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][550/1251] eta 0:03:29 lr 0.000024 time 0.2876 (0.2990) loss 3.3970 (2.9714) grad_norm 2.6541 (3.2647) [2021-04-16 19:38:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][560/1251] eta 0:03:26 lr 0.000024 time 0.2697 (0.2986) loss 2.2137 (2.9693) grad_norm 2.7748 (3.2652) [2021-04-16 19:38:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][570/1251] eta 0:03:23 lr 0.000024 time 0.2656 (0.2987) loss 3.4409 (2.9725) grad_norm 3.6732 (3.2675) [2021-04-16 19:38:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][580/1251] eta 0:03:20 lr 0.000024 time 0.2801 (0.2983) loss 2.2334 (2.9742) grad_norm 3.4764 (nan) [2021-04-16 19:38:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][590/1251] eta 0:03:16 lr 0.000024 time 0.2952 (0.2979) loss 3.4069 (2.9752) grad_norm 2.8937 (nan) [2021-04-16 19:38:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][600/1251] eta 0:03:13 lr 0.000024 time 0.2586 (0.2977) loss 3.0390 (2.9769) grad_norm 3.2452 (nan) [2021-04-16 19:38:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][610/1251] eta 0:03:10 lr 0.000024 time 0.2788 (0.2974) loss 3.0058 (2.9816) grad_norm 2.8334 (nan) [2021-04-16 19:38:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][620/1251] eta 0:03:07 lr 0.000024 time 0.2826 (0.2970) loss 3.0036 (2.9830) grad_norm 3.1429 (nan) [2021-04-16 19:38:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][630/1251] eta 0:03:04 lr 0.000024 time 0.2760 (0.2968) loss 3.2865 (2.9844) grad_norm 2.9639 (nan) [2021-04-16 19:38:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][640/1251] eta 0:03:01 lr 0.000024 time 0.2725 (0.2965) loss 2.2417 (2.9809) grad_norm 4.0420 (nan) [2021-04-16 19:38:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][650/1251] eta 0:02:58 lr 0.000024 time 0.2795 (0.2963) loss 2.1291 (2.9741) grad_norm 3.3445 (nan) [2021-04-16 19:38:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][660/1251] eta 0:02:54 lr 0.000024 time 0.2597 (0.2960) loss 2.4643 (2.9716) grad_norm 3.0184 (nan) [2021-04-16 19:38:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][670/1251] eta 0:02:51 lr 0.000024 time 0.2810 (0.2957) loss 2.0906 (2.9743) grad_norm 2.8740 (nan) [2021-04-16 19:38:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][680/1251] eta 0:02:48 lr 0.000024 time 0.2813 (0.2955) loss 2.2653 (2.9771) grad_norm 2.9006 (nan) [2021-04-16 19:38:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][690/1251] eta 0:02:45 lr 0.000024 time 0.2778 (0.2952) loss 2.6113 (2.9790) grad_norm 3.4408 (nan) [2021-04-16 19:38:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][700/1251] eta 0:02:42 lr 0.000024 time 0.2840 (0.2950) loss 3.0760 (2.9749) grad_norm 3.5641 (nan) [2021-04-16 19:38:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][710/1251] eta 0:02:39 lr 0.000024 time 0.2774 (0.2947) loss 3.4356 (2.9740) grad_norm 3.8329 (nan) [2021-04-16 19:39:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][720/1251] eta 0:02:36 lr 0.000024 time 0.2875 (0.2949) loss 3.2892 (2.9741) grad_norm 3.1457 (nan) [2021-04-16 19:39:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][730/1251] eta 0:02:33 lr 0.000024 time 0.3734 (0.2947) loss 3.4454 (2.9758) grad_norm 3.0837 (nan) [2021-04-16 19:39:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][740/1251] eta 0:02:30 lr 0.000024 time 0.2701 (0.2944) loss 3.0678 (2.9738) grad_norm 3.4760 (nan) [2021-04-16 19:39:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][750/1251] eta 0:02:27 lr 0.000024 time 0.2857 (0.2943) loss 2.8631 (2.9719) grad_norm 2.9732 (nan) [2021-04-16 19:39:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][760/1251] eta 0:02:24 lr 0.000024 time 0.2987 (0.2941) loss 3.3343 (2.9740) grad_norm 3.7945 (nan) [2021-04-16 19:39:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][770/1251] eta 0:02:21 lr 0.000024 time 0.2820 (0.2939) loss 3.1108 (2.9740) grad_norm 3.8732 (nan) [2021-04-16 19:39:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][780/1251] eta 0:02:18 lr 0.000024 time 0.2850 (0.2938) loss 3.7646 (2.9740) grad_norm 3.1888 (nan) [2021-04-16 19:39:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][790/1251] eta 0:02:15 lr 0.000024 time 0.3117 (0.2938) loss 3.6107 (2.9741) grad_norm 3.1769 (nan) [2021-04-16 19:39:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][800/1251] eta 0:02:12 lr 0.000024 time 0.2949 (0.2936) loss 3.7679 (2.9762) grad_norm 2.6871 (nan) [2021-04-16 19:39:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][810/1251] eta 0:02:09 lr 0.000023 time 0.2707 (0.2934) loss 2.1102 (2.9768) grad_norm 2.9594 (nan) [2021-04-16 19:39:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][820/1251] eta 0:02:06 lr 0.000023 time 0.2855 (0.2932) loss 3.0711 (2.9772) grad_norm 3.4145 (nan) [2021-04-16 19:39:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][830/1251] eta 0:02:03 lr 0.000023 time 0.2735 (0.2930) loss 3.2227 (2.9805) grad_norm 3.1085 (nan) [2021-04-16 19:39:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][840/1251] eta 0:02:00 lr 0.000023 time 0.2782 (0.2928) loss 3.0781 (2.9786) grad_norm 3.1339 (nan) [2021-04-16 19:39:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][850/1251] eta 0:01:57 lr 0.000023 time 0.2613 (0.2926) loss 3.1750 (2.9812) grad_norm 3.2150 (nan) [2021-04-16 19:39:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][860/1251] eta 0:01:54 lr 0.000023 time 0.2686 (0.2925) loss 3.5607 (2.9826) grad_norm 2.9684 (nan) [2021-04-16 19:39:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][870/1251] eta 0:01:51 lr 0.000023 time 0.2704 (0.2924) loss 3.0098 (2.9836) grad_norm 2.8427 (nan) [2021-04-16 19:39:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][880/1251] eta 0:01:48 lr 0.000023 time 0.2793 (0.2922) loss 2.3232 (2.9843) grad_norm 2.7702 (nan) [2021-04-16 19:39:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][890/1251] eta 0:01:45 lr 0.000023 time 0.2881 (0.2920) loss 2.2167 (2.9845) grad_norm 3.2502 (nan) [2021-04-16 19:39:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][900/1251] eta 0:01:42 lr 0.000023 time 0.2802 (0.2919) loss 3.2546 (2.9844) grad_norm 3.4464 (nan) [2021-04-16 19:39:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][910/1251] eta 0:01:39 lr 0.000023 time 0.2853 (0.2917) loss 2.9764 (2.9845) grad_norm 3.0905 (nan) [2021-04-16 19:39:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][920/1251] eta 0:01:36 lr 0.000023 time 0.2947 (0.2916) loss 3.3642 (2.9862) grad_norm 3.2266 (nan) [2021-04-16 19:40:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][930/1251] eta 0:01:33 lr 0.000023 time 0.3016 (0.2915) loss 3.1091 (2.9875) grad_norm 2.8832 (nan) [2021-04-16 19:40:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][940/1251] eta 0:01:30 lr 0.000023 time 0.4066 (0.2916) loss 3.0199 (2.9865) grad_norm 3.1927 (nan) [2021-04-16 19:40:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][950/1251] eta 0:01:27 lr 0.000023 time 0.2700 (0.2916) loss 3.3804 (2.9883) grad_norm 3.1032 (nan) [2021-04-16 19:40:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][960/1251] eta 0:01:24 lr 0.000023 time 0.2739 (0.2915) loss 3.3636 (2.9899) grad_norm 2.9066 (nan) [2021-04-16 19:40:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][970/1251] eta 0:01:21 lr 0.000023 time 0.2527 (0.2913) loss 3.4572 (2.9901) grad_norm 3.6810 (nan) [2021-04-16 19:40:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][980/1251] eta 0:01:18 lr 0.000023 time 0.3053 (0.2912) loss 3.2203 (2.9917) grad_norm 2.9882 (nan) [2021-04-16 19:40:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][990/1251] eta 0:01:15 lr 0.000023 time 0.3120 (0.2911) loss 3.2058 (2.9895) grad_norm 2.8740 (nan) [2021-04-16 19:40:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][1000/1251] eta 0:01:13 lr 0.000023 time 0.2686 (0.2910) loss 3.2688 (2.9907) grad_norm 3.3005 (nan) [2021-04-16 19:40:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][1010/1251] eta 0:01:10 lr 0.000023 time 0.2954 (0.2910) loss 3.3448 (2.9935) grad_norm 3.0069 (nan) [2021-04-16 19:40:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][1020/1251] eta 0:01:07 lr 0.000023 time 0.2641 (0.2908) loss 3.1729 (2.9964) grad_norm 3.2114 (nan) [2021-04-16 19:40:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][1030/1251] eta 0:01:04 lr 0.000023 time 0.2572 (0.2913) loss 3.0313 (2.9966) grad_norm 2.7616 (nan) [2021-04-16 19:40:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][1040/1251] eta 0:01:01 lr 0.000023 time 0.2765 (0.2912) loss 3.0330 (2.9987) grad_norm 2.9025 (nan) [2021-04-16 19:40:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][1050/1251] eta 0:00:58 lr 0.000023 time 0.2835 (0.2911) loss 2.5934 (2.9967) grad_norm 3.1121 (nan) [2021-04-16 19:40:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][1060/1251] eta 0:00:55 lr 0.000023 time 0.2852 (0.2910) loss 2.9525 (2.9962) grad_norm 3.0872 (nan) [2021-04-16 19:40:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][1070/1251] eta 0:00:52 lr 0.000023 time 0.2754 (0.2908) loss 3.3980 (2.9948) grad_norm 3.0382 (nan) [2021-04-16 19:40:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][1080/1251] eta 0:00:49 lr 0.000023 time 0.2985 (0.2907) loss 2.4411 (2.9915) grad_norm 3.6809 (nan) [2021-04-16 19:40:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][1090/1251] eta 0:00:46 lr 0.000023 time 0.2690 (0.2906) loss 3.5216 (2.9904) grad_norm 3.0828 (nan) [2021-04-16 19:40:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][1100/1251] eta 0:00:43 lr 0.000023 time 0.3030 (0.2905) loss 2.3833 (2.9895) grad_norm 3.5183 (nan) [2021-04-16 19:40:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][1110/1251] eta 0:00:40 lr 0.000023 time 0.2724 (0.2904) loss 3.0042 (2.9901) grad_norm 3.2854 (nan) [2021-04-16 19:40:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][1120/1251] eta 0:00:38 lr 0.000023 time 0.2842 (0.2904) loss 3.1912 (2.9909) grad_norm 4.6383 (nan) [2021-04-16 19:40:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][1130/1251] eta 0:00:35 lr 0.000023 time 0.2808 (0.2904) loss 3.2545 (2.9915) grad_norm 3.5062 (nan) [2021-04-16 19:41:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][1140/1251] eta 0:00:32 lr 0.000023 time 0.2524 (0.2903) loss 2.9820 (2.9920) grad_norm 3.4811 (nan) [2021-04-16 19:41:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][1150/1251] eta 0:00:29 lr 0.000023 time 0.2704 (0.2904) loss 2.9758 (2.9917) grad_norm 4.9708 (nan) [2021-04-16 19:41:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][1160/1251] eta 0:00:26 lr 0.000023 time 0.2792 (0.2903) loss 3.5305 (2.9926) grad_norm 5.4576 (nan) [2021-04-16 19:41:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][1170/1251] eta 0:00:23 lr 0.000023 time 0.2872 (0.2903) loss 3.3308 (2.9930) grad_norm 3.2824 (nan) [2021-04-16 19:41:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][1180/1251] eta 0:00:20 lr 0.000023 time 0.2957 (0.2902) loss 3.4773 (2.9937) grad_norm 3.5320 (nan) [2021-04-16 19:41:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][1190/1251] eta 0:00:17 lr 0.000023 time 0.3075 (0.2901) loss 3.2478 (2.9922) grad_norm 2.8788 (nan) [2021-04-16 19:41:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][1200/1251] eta 0:00:14 lr 0.000023 time 0.2537 (0.2900) loss 2.2750 (2.9925) grad_norm 2.9932 (nan) [2021-04-16 19:41:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][1210/1251] eta 0:00:11 lr 0.000023 time 0.2738 (0.2898) loss 2.7482 (2.9917) grad_norm 2.9838 (nan) [2021-04-16 19:41:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][1220/1251] eta 0:00:08 lr 0.000023 time 0.2895 (0.2897) loss 3.1155 (2.9909) grad_norm 2.9561 (nan) [2021-04-16 19:41:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][1230/1251] eta 0:00:06 lr 0.000023 time 0.2704 (0.2897) loss 2.8344 (2.9912) grad_norm 3.1441 (nan) [2021-04-16 19:41:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][1240/1251] eta 0:00:03 lr 0.000023 time 0.2612 (0.2895) loss 2.5394 (2.9912) grad_norm 3.1303 (nan) [2021-04-16 19:41:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [277/300][1250/1251] eta 0:00:00 lr 0.000023 time 0.2480 (0.2892) loss 2.8211 (2.9914) grad_norm 2.7296 (nan) [2021-04-16 19:41:41 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 277 training takes 0:06:11 [2021-04-16 19:41:41 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_277.pth saving...... [2021-04-16 19:41:55 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_277.pth saved !!! [2021-04-16 19:41:56 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.121 (1.121) Loss 0.8195 (0.8195) Acc@1 80.762 (80.762) Acc@5 95.605 (95.605) [2021-04-16 19:41:58 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.311 (0.279) Loss 0.7720 (0.7984) Acc@1 81.543 (81.348) Acc@5 96.289 (95.703) [2021-04-16 19:42:00 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.100 (0.247) Loss 0.8144 (0.8147) Acc@1 81.348 (80.952) Acc@5 95.312 (95.475) [2021-04-16 19:42:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.137 (0.226) Loss 0.8381 (0.8148) Acc@1 80.078 (80.944) Acc@5 95.703 (95.549) [2021-04-16 19:42:04 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.090 (0.214) Loss 0.7802 (0.8129) Acc@1 82.617 (81.064) Acc@5 95.215 (95.520) [2021-04-16 19:42:20 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 81.038 Acc@5 95.468 [2021-04-16 19:42:20 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 81.0% [2021-04-16 19:42:20 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 81.06% [2021-04-16 19:42:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][0/1251] eta 4:05:30 lr 0.000023 time 11.7751 (11.7751) loss 3.1836 (3.1836) grad_norm 4.8242 (4.8242) [2021-04-16 19:42:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][10/1251] eta 0:27:22 lr 0.000023 time 0.2708 (1.3232) loss 2.2988 (2.8513) grad_norm 2.6924 (3.4023) [2021-04-16 19:42:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][20/1251] eta 0:17:01 lr 0.000023 time 0.2679 (0.8301) loss 3.4422 (2.9923) grad_norm 3.1856 (3.3201) [2021-04-16 19:42:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][30/1251] eta 0:13:15 lr 0.000023 time 0.2676 (0.6516) loss 1.8118 (2.9761) grad_norm 3.0367 (3.4118) [2021-04-16 19:42:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][40/1251] eta 0:11:16 lr 0.000023 time 0.2577 (0.5590) loss 3.2994 (3.0053) grad_norm 3.0033 (3.3317) [2021-04-16 19:42:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][50/1251] eta 0:10:06 lr 0.000023 time 0.2758 (0.5046) loss 2.9510 (2.9956) grad_norm 3.4330 (3.4304) [2021-04-16 19:42:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][60/1251] eta 0:09:16 lr 0.000023 time 0.2778 (0.4673) loss 2.7834 (2.9783) grad_norm 2.6661 (3.3581) [2021-04-16 19:42:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][70/1251] eta 0:08:41 lr 0.000023 time 0.2804 (0.4414) loss 2.6789 (2.9844) grad_norm 3.0951 (3.3263) [2021-04-16 19:42:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][80/1251] eta 0:08:12 lr 0.000023 time 0.2566 (0.4207) loss 2.8550 (2.9828) grad_norm 2.9883 (3.3237) [2021-04-16 19:42:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][90/1251] eta 0:07:50 lr 0.000023 time 0.2973 (0.4056) loss 2.0669 (2.9786) grad_norm 3.5761 (3.3252) [2021-04-16 19:42:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][100/1251] eta 0:07:33 lr 0.000023 time 0.2883 (0.3939) loss 2.9793 (2.9899) grad_norm 3.5848 (3.3017) [2021-04-16 19:43:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][110/1251] eta 0:07:17 lr 0.000023 time 0.2845 (0.3835) loss 3.7705 (2.9896) grad_norm 3.3017 (3.2978) [2021-04-16 19:43:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][120/1251] eta 0:07:03 lr 0.000023 time 0.2670 (0.3746) loss 2.2575 (2.9828) grad_norm 3.1782 (3.2881) [2021-04-16 19:43:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][130/1251] eta 0:06:51 lr 0.000023 time 0.2792 (0.3670) loss 3.3722 (2.9765) grad_norm 4.0570 (3.3292) [2021-04-16 19:43:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][140/1251] eta 0:06:40 lr 0.000023 time 0.2908 (0.3605) loss 2.8446 (2.9656) grad_norm 2.8945 (3.3251) [2021-04-16 19:43:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][150/1251] eta 0:06:32 lr 0.000023 time 0.4578 (0.3567) loss 3.2074 (2.9697) grad_norm 2.8797 (3.3094) [2021-04-16 19:43:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][160/1251] eta 0:06:24 lr 0.000023 time 0.2970 (0.3526) loss 2.8065 (2.9688) grad_norm 3.1066 (3.3118) [2021-04-16 19:43:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][170/1251] eta 0:06:16 lr 0.000023 time 0.3016 (0.3482) loss 3.4940 (2.9750) grad_norm 3.5094 (3.3100) [2021-04-16 19:43:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][180/1251] eta 0:06:08 lr 0.000023 time 0.2701 (0.3441) loss 3.3903 (2.9687) grad_norm 2.8436 (3.3135) [2021-04-16 19:43:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][190/1251] eta 0:06:01 lr 0.000023 time 0.3008 (0.3409) loss 1.9463 (2.9617) grad_norm 3.3513 (3.3253) [2021-04-16 19:43:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][200/1251] eta 0:05:55 lr 0.000023 time 0.2804 (0.3380) loss 2.9946 (2.9620) grad_norm 2.8928 (3.3206) [2021-04-16 19:43:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][210/1251] eta 0:05:48 lr 0.000023 time 0.2807 (0.3351) loss 3.1622 (2.9605) grad_norm 3.3145 (3.3209) [2021-04-16 19:43:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][220/1251] eta 0:05:43 lr 0.000023 time 0.2815 (0.3330) loss 3.0984 (2.9670) grad_norm 3.1484 (3.3264) [2021-04-16 19:43:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][230/1251] eta 0:05:37 lr 0.000023 time 0.2838 (0.3308) loss 3.5711 (2.9693) grad_norm 3.5374 (3.3203) [2021-04-16 19:43:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][240/1251] eta 0:05:32 lr 0.000023 time 0.2462 (0.3285) loss 2.9452 (2.9632) grad_norm 3.2009 (3.3101) [2021-04-16 19:43:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][250/1251] eta 0:05:27 lr 0.000023 time 0.2645 (0.3269) loss 3.1395 (2.9561) grad_norm 3.3908 (3.3197) [2021-04-16 19:43:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][260/1251] eta 0:05:21 lr 0.000023 time 0.2813 (0.3248) loss 3.0667 (2.9503) grad_norm 3.4434 (3.3322) [2021-04-16 19:43:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][270/1251] eta 0:05:16 lr 0.000023 time 0.2596 (0.3231) loss 2.2609 (2.9441) grad_norm 3.1093 (3.3284) [2021-04-16 19:43:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][280/1251] eta 0:05:11 lr 0.000023 time 0.2815 (0.3213) loss 3.3757 (2.9557) grad_norm 3.0452 (3.3313) [2021-04-16 19:43:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][290/1251] eta 0:05:07 lr 0.000023 time 0.2878 (0.3202) loss 3.0577 (2.9596) grad_norm 3.0122 (3.3606) [2021-04-16 19:43:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][300/1251] eta 0:05:03 lr 0.000023 time 0.2601 (0.3188) loss 2.8087 (2.9557) grad_norm 4.5335 (3.3564) [2021-04-16 19:43:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][310/1251] eta 0:04:59 lr 0.000023 time 0.3009 (0.3179) loss 3.5583 (2.9625) grad_norm 2.9122 (3.3585) [2021-04-16 19:44:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][320/1251] eta 0:04:54 lr 0.000023 time 0.2549 (0.3167) loss 2.5328 (2.9668) grad_norm 2.9495 (3.3597) [2021-04-16 19:44:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][330/1251] eta 0:04:51 lr 0.000023 time 0.3042 (0.3160) loss 3.7174 (2.9693) grad_norm 3.9277 (3.3587) [2021-04-16 19:44:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][340/1251] eta 0:04:46 lr 0.000023 time 0.2921 (0.3149) loss 3.4432 (2.9730) grad_norm 5.5079 (3.3550) [2021-04-16 19:44:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][350/1251] eta 0:04:43 lr 0.000023 time 0.2613 (0.3141) loss 3.1934 (2.9760) grad_norm 2.8464 (3.3570) [2021-04-16 19:44:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][360/1251] eta 0:04:39 lr 0.000023 time 0.2823 (0.3138) loss 3.0147 (2.9743) grad_norm 2.8663 (3.3598) [2021-04-16 19:44:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][370/1251] eta 0:04:35 lr 0.000023 time 0.2839 (0.3127) loss 3.1179 (2.9720) grad_norm 2.9093 (3.3597) [2021-04-16 19:44:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][380/1251] eta 0:04:31 lr 0.000023 time 0.2666 (0.3117) loss 3.1257 (2.9685) grad_norm 2.9294 (3.3529) [2021-04-16 19:44:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][390/1251] eta 0:04:27 lr 0.000023 time 0.2750 (0.3108) loss 2.2220 (2.9685) grad_norm 3.6477 (3.3530) [2021-04-16 19:44:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][400/1251] eta 0:04:23 lr 0.000023 time 0.2778 (0.3099) loss 3.3348 (2.9725) grad_norm 2.8137 (3.3494) [2021-04-16 19:44:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][410/1251] eta 0:04:20 lr 0.000023 time 0.2962 (0.3092) loss 1.9457 (2.9670) grad_norm 3.5662 (3.3549) [2021-04-16 19:44:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][420/1251] eta 0:04:16 lr 0.000023 time 0.2710 (0.3085) loss 2.6215 (2.9660) grad_norm 3.8652 (3.3635) [2021-04-16 19:44:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][430/1251] eta 0:04:12 lr 0.000023 time 0.2619 (0.3076) loss 2.5409 (2.9703) grad_norm 3.1354 (3.3677) [2021-04-16 19:44:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][440/1251] eta 0:04:09 lr 0.000023 time 0.2829 (0.3070) loss 2.8804 (2.9729) grad_norm 3.8653 (3.3846) [2021-04-16 19:44:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][450/1251] eta 0:04:05 lr 0.000023 time 0.2629 (0.3064) loss 3.5377 (2.9750) grad_norm 2.7113 (3.3810) [2021-04-16 19:44:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][460/1251] eta 0:04:01 lr 0.000023 time 0.2820 (0.3058) loss 3.5162 (2.9721) grad_norm 2.7407 (3.3768) [2021-04-16 19:44:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][470/1251] eta 0:03:58 lr 0.000023 time 0.2902 (0.3051) loss 3.4579 (2.9731) grad_norm 4.3909 (3.3801) [2021-04-16 19:44:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][480/1251] eta 0:03:54 lr 0.000023 time 0.2864 (0.3045) loss 2.8155 (2.9743) grad_norm 4.1904 (3.3819) [2021-04-16 19:44:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][490/1251] eta 0:03:51 lr 0.000023 time 0.2805 (0.3041) loss 3.0839 (2.9756) grad_norm 2.5973 (3.3779) [2021-04-16 19:44:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][500/1251] eta 0:03:47 lr 0.000023 time 0.2620 (0.3034) loss 3.2212 (2.9757) grad_norm 3.1805 (3.3784) [2021-04-16 19:44:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][510/1251] eta 0:03:44 lr 0.000023 time 0.2721 (0.3030) loss 3.2236 (2.9791) grad_norm 2.9582 (3.3742) [2021-04-16 19:44:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][520/1251] eta 0:03:41 lr 0.000023 time 0.2444 (0.3024) loss 2.9837 (2.9771) grad_norm 3.5760 (3.3716) [2021-04-16 19:45:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][530/1251] eta 0:03:37 lr 0.000023 time 0.2677 (0.3019) loss 2.9509 (2.9747) grad_norm 2.7929 (3.3682) [2021-04-16 19:45:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][540/1251] eta 0:03:34 lr 0.000023 time 0.2647 (0.3015) loss 3.3235 (2.9740) grad_norm 3.4004 (3.3685) [2021-04-16 19:45:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][550/1251] eta 0:03:31 lr 0.000023 time 0.2802 (0.3011) loss 3.1421 (2.9722) grad_norm 3.0442 (3.3657) [2021-04-16 19:45:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][560/1251] eta 0:03:27 lr 0.000023 time 0.3062 (0.3007) loss 2.6493 (2.9765) grad_norm 3.2873 (3.3670) [2021-04-16 19:45:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][570/1251] eta 0:03:24 lr 0.000023 time 0.2796 (0.3002) loss 3.5219 (2.9796) grad_norm 3.2664 (3.3601) [2021-04-16 19:45:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][580/1251] eta 0:03:21 lr 0.000023 time 0.2936 (0.3003) loss 2.0989 (2.9820) grad_norm 3.1115 (3.3583) [2021-04-16 19:45:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][590/1251] eta 0:03:18 lr 0.000023 time 0.2529 (0.3000) loss 2.8680 (2.9839) grad_norm 3.2846 (3.3555) [2021-04-16 19:45:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][600/1251] eta 0:03:15 lr 0.000023 time 0.2617 (0.2996) loss 2.1593 (2.9790) grad_norm 5.0943 (3.3559) [2021-04-16 19:45:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][610/1251] eta 0:03:11 lr 0.000023 time 0.2996 (0.2994) loss 3.1544 (2.9797) grad_norm 3.3409 (3.3533) [2021-04-16 19:45:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][620/1251] eta 0:03:08 lr 0.000022 time 0.2713 (0.2991) loss 3.0139 (2.9828) grad_norm 3.0004 (3.3517) [2021-04-16 19:45:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][630/1251] eta 0:03:05 lr 0.000022 time 0.2753 (0.2988) loss 2.5193 (2.9816) grad_norm 3.7961 (3.3500) [2021-04-16 19:45:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][640/1251] eta 0:03:02 lr 0.000022 time 0.2793 (0.2986) loss 1.7172 (2.9790) grad_norm 3.0574 (3.3491) [2021-04-16 19:45:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][650/1251] eta 0:02:59 lr 0.000022 time 0.2578 (0.2982) loss 3.0669 (2.9862) grad_norm 3.1571 (3.3510) [2021-04-16 19:45:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][660/1251] eta 0:02:56 lr 0.000022 time 0.2663 (0.2978) loss 3.4641 (2.9869) grad_norm 3.0288 (3.3479) [2021-04-16 19:45:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][670/1251] eta 0:02:52 lr 0.000022 time 0.2864 (0.2976) loss 3.3691 (2.9894) grad_norm 3.7330 (3.3470) [2021-04-16 19:45:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][680/1251] eta 0:02:49 lr 0.000022 time 0.2965 (0.2973) loss 3.3999 (2.9916) grad_norm 3.1462 (3.3447) [2021-04-16 19:45:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][690/1251] eta 0:02:46 lr 0.000022 time 0.2766 (0.2970) loss 2.6115 (2.9945) grad_norm 3.4629 (3.3426) [2021-04-16 19:45:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][700/1251] eta 0:02:43 lr 0.000022 time 0.2534 (0.2968) loss 2.1939 (2.9913) grad_norm 2.9866 (3.3443) [2021-04-16 19:45:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][710/1251] eta 0:02:40 lr 0.000022 time 0.2749 (0.2965) loss 2.4286 (2.9900) grad_norm 3.3438 (3.3434) [2021-04-16 19:45:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][720/1251] eta 0:02:37 lr 0.000022 time 0.2779 (0.2963) loss 2.6225 (2.9880) grad_norm 3.1878 (3.3418) [2021-04-16 19:45:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][730/1251] eta 0:02:34 lr 0.000022 time 0.2893 (0.2963) loss 2.9502 (2.9908) grad_norm 2.9913 (3.3419) [2021-04-16 19:45:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][740/1251] eta 0:02:31 lr 0.000022 time 0.2863 (0.2961) loss 3.1862 (2.9921) grad_norm 3.2190 (3.3404) [2021-04-16 19:46:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][750/1251] eta 0:02:28 lr 0.000022 time 0.2614 (0.2959) loss 3.1363 (2.9940) grad_norm 3.1677 (3.3391) [2021-04-16 19:46:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][760/1251] eta 0:02:25 lr 0.000022 time 0.2655 (0.2958) loss 3.6846 (2.9936) grad_norm 3.1807 (3.3356) [2021-04-16 19:46:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][770/1251] eta 0:02:22 lr 0.000022 time 0.2965 (0.2956) loss 1.8296 (2.9929) grad_norm 2.8578 (3.3359) [2021-04-16 19:46:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][780/1251] eta 0:02:19 lr 0.000022 time 0.2658 (0.2953) loss 3.1429 (2.9941) grad_norm 2.8352 (3.3342) [2021-04-16 19:46:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][790/1251] eta 0:02:16 lr 0.000022 time 0.2535 (0.2951) loss 2.4015 (2.9932) grad_norm 2.9092 (3.3362) [2021-04-16 19:46:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][800/1251] eta 0:02:12 lr 0.000022 time 0.2736 (0.2949) loss 2.7621 (2.9945) grad_norm 3.1922 (3.3345) [2021-04-16 19:46:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][810/1251] eta 0:02:10 lr 0.000022 time 0.2979 (0.2948) loss 2.1756 (2.9940) grad_norm 3.1383 (3.3355) [2021-04-16 19:46:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][820/1251] eta 0:02:07 lr 0.000022 time 0.2904 (0.2948) loss 2.5825 (2.9903) grad_norm 3.2223 (3.3412) [2021-04-16 19:46:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][830/1251] eta 0:02:04 lr 0.000022 time 0.2663 (0.2946) loss 3.1452 (2.9856) grad_norm 3.8488 (3.3411) [2021-04-16 19:46:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][840/1251] eta 0:02:01 lr 0.000022 time 0.2645 (0.2944) loss 3.8302 (2.9859) grad_norm 3.3136 (3.3382) [2021-04-16 19:46:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][850/1251] eta 0:01:57 lr 0.000022 time 0.2661 (0.2942) loss 3.1969 (2.9833) grad_norm 3.5606 (3.3379) [2021-04-16 19:46:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][860/1251] eta 0:01:54 lr 0.000022 time 0.2827 (0.2940) loss 2.7078 (2.9841) grad_norm 2.8416 (3.3380) [2021-04-16 19:46:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][870/1251] eta 0:01:51 lr 0.000022 time 0.2744 (0.2938) loss 2.5541 (2.9815) grad_norm 4.2900 (3.3402) [2021-04-16 19:46:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][880/1251] eta 0:01:49 lr 0.000022 time 0.4136 (0.2939) loss 3.2218 (2.9823) grad_norm 2.9545 (3.3383) [2021-04-16 19:46:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][890/1251] eta 0:01:46 lr 0.000022 time 0.2632 (0.2936) loss 2.8849 (2.9796) grad_norm 3.1434 (3.3351) [2021-04-16 19:46:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][900/1251] eta 0:01:43 lr 0.000022 time 0.4353 (0.2937) loss 2.5524 (2.9801) grad_norm 2.9799 (3.3339) [2021-04-16 19:46:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][910/1251] eta 0:01:40 lr 0.000022 time 0.2909 (0.2936) loss 2.0846 (2.9783) grad_norm 3.0494 (3.3338) [2021-04-16 19:46:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][920/1251] eta 0:01:37 lr 0.000022 time 0.2786 (0.2934) loss 3.1869 (2.9787) grad_norm 3.3629 (3.3312) [2021-04-16 19:46:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][930/1251] eta 0:01:34 lr 0.000022 time 0.2884 (0.2934) loss 2.5572 (2.9804) grad_norm 3.2253 (3.3321) [2021-04-16 19:46:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][940/1251] eta 0:01:31 lr 0.000022 time 0.2906 (0.2935) loss 2.2019 (2.9815) grad_norm 2.6609 (3.3330) [2021-04-16 19:46:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][950/1251] eta 0:01:28 lr 0.000022 time 0.2580 (0.2933) loss 3.2871 (2.9842) grad_norm 3.2083 (3.3308) [2021-04-16 19:47:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][960/1251] eta 0:01:25 lr 0.000022 time 0.3040 (0.2933) loss 3.6301 (2.9863) grad_norm 2.6538 (3.3296) [2021-04-16 19:47:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][970/1251] eta 0:01:22 lr 0.000022 time 0.2987 (0.2931) loss 2.8836 (2.9845) grad_norm 4.1150 (3.3291) [2021-04-16 19:47:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][980/1251] eta 0:01:19 lr 0.000022 time 0.2487 (0.2930) loss 3.4304 (2.9854) grad_norm 3.3022 (3.3272) [2021-04-16 19:47:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][990/1251] eta 0:01:16 lr 0.000022 time 0.2722 (0.2928) loss 3.5351 (2.9871) grad_norm 3.2625 (3.3251) [2021-04-16 19:47:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][1000/1251] eta 0:01:13 lr 0.000022 time 0.3121 (0.2927) loss 3.1537 (2.9880) grad_norm 2.7103 (3.3240) [2021-04-16 19:47:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][1010/1251] eta 0:01:10 lr 0.000022 time 0.2437 (0.2925) loss 3.3887 (2.9869) grad_norm 3.1572 (3.3233) [2021-04-16 19:47:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][1020/1251] eta 0:01:07 lr 0.000022 time 0.2930 (0.2923) loss 2.3875 (2.9867) grad_norm 3.0665 (3.3223) [2021-04-16 19:47:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][1030/1251] eta 0:01:04 lr 0.000022 time 0.2602 (0.2922) loss 3.1506 (2.9867) grad_norm 3.0476 (3.3203) [2021-04-16 19:47:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][1040/1251] eta 0:01:01 lr 0.000022 time 0.2617 (0.2920) loss 3.4966 (2.9873) grad_norm 2.9520 (3.3177) [2021-04-16 19:47:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][1050/1251] eta 0:00:58 lr 0.000022 time 0.2685 (0.2919) loss 2.1520 (2.9836) grad_norm 3.0276 (3.3172) [2021-04-16 19:47:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][1060/1251] eta 0:00:55 lr 0.000022 time 0.2784 (0.2917) loss 2.3707 (2.9842) grad_norm 3.1727 (3.3157) [2021-04-16 19:47:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][1070/1251] eta 0:00:52 lr 0.000022 time 0.2785 (0.2917) loss 2.7846 (2.9848) grad_norm 3.3424 (3.3147) [2021-04-16 19:47:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][1080/1251] eta 0:00:49 lr 0.000022 time 0.2650 (0.2916) loss 2.8061 (2.9858) grad_norm 3.7636 (3.3167) [2021-04-16 19:47:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][1090/1251] eta 0:00:46 lr 0.000022 time 0.2789 (0.2914) loss 3.2725 (2.9851) grad_norm 3.2311 (3.3183) [2021-04-16 19:47:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][1100/1251] eta 0:00:43 lr 0.000022 time 0.2703 (0.2913) loss 2.7620 (2.9864) grad_norm 3.8396 (3.3204) [2021-04-16 19:47:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][1110/1251] eta 0:00:41 lr 0.000022 time 0.3031 (0.2912) loss 3.8343 (2.9863) grad_norm 3.2662 (3.3199) [2021-04-16 19:47:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][1120/1251] eta 0:00:38 lr 0.000022 time 0.2570 (0.2911) loss 2.9367 (2.9863) grad_norm 2.7249 (3.3191) [2021-04-16 19:47:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][1130/1251] eta 0:00:35 lr 0.000022 time 0.2926 (0.2909) loss 3.4695 (2.9861) grad_norm 4.0686 (3.3250) [2021-04-16 19:47:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][1140/1251] eta 0:00:32 lr 0.000022 time 0.2841 (0.2909) loss 3.1472 (2.9872) grad_norm 2.8239 (3.3252) [2021-04-16 19:47:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][1150/1251] eta 0:00:29 lr 0.000022 time 0.3031 (0.2908) loss 3.5228 (2.9851) grad_norm 2.9975 (3.3263) [2021-04-16 19:47:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][1160/1251] eta 0:00:26 lr 0.000022 time 0.2586 (0.2909) loss 3.4529 (2.9839) grad_norm 4.1898 (3.3263) [2021-04-16 19:48:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][1170/1251] eta 0:00:23 lr 0.000022 time 0.2552 (0.2908) loss 2.2293 (2.9834) grad_norm 3.0803 (3.3264) [2021-04-16 19:48:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][1180/1251] eta 0:00:20 lr 0.000022 time 0.2570 (0.2907) loss 3.0244 (2.9841) grad_norm 3.2803 (3.3262) [2021-04-16 19:48:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][1190/1251] eta 0:00:17 lr 0.000022 time 0.2904 (0.2906) loss 1.9319 (2.9855) grad_norm 3.1785 (3.3241) [2021-04-16 19:48:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][1200/1251] eta 0:00:14 lr 0.000022 time 0.2921 (0.2905) loss 3.7844 (2.9887) grad_norm 3.1680 (3.3236) [2021-04-16 19:48:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][1210/1251] eta 0:00:11 lr 0.000022 time 0.2605 (0.2904) loss 3.2236 (2.9892) grad_norm 3.1434 (3.3216) [2021-04-16 19:48:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][1220/1251] eta 0:00:09 lr 0.000022 time 0.2750 (0.2904) loss 3.3305 (2.9883) grad_norm 3.0087 (3.3207) [2021-04-16 19:48:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][1230/1251] eta 0:00:06 lr 0.000022 time 0.3005 (0.2904) loss 3.2904 (2.9891) grad_norm 3.4225 (3.3224) [2021-04-16 19:48:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][1240/1251] eta 0:00:03 lr 0.000022 time 0.2479 (0.2902) loss 3.1631 (2.9912) grad_norm 2.9245 (3.3215) [2021-04-16 19:48:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [278/300][1250/1251] eta 0:00:00 lr 0.000022 time 0.2480 (0.2899) loss 2.4196 (2.9924) grad_norm 3.0619 (3.3217) [2021-04-16 19:48:28 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 278 training takes 0:06:08 [2021-04-16 19:48:28 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_278.pth saving...... [2021-04-16 19:48:45 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_278.pth saved !!! [2021-04-16 19:48:47 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.116 (1.116) Loss 0.8073 (0.8073) Acc@1 81.543 (81.543) Acc@5 95.996 (95.996) [2021-04-16 19:48:48 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.686 (0.280) Loss 0.8029 (0.8046) Acc@1 82.422 (81.729) Acc@5 94.922 (95.588) [2021-04-16 19:48:50 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.114 (0.208) Loss 0.8923 (0.8181) Acc@1 80.176 (81.129) Acc@5 94.336 (95.410) [2021-04-16 19:48:52 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.112 (0.224) Loss 0.8481 (0.8189) Acc@1 80.078 (81.023) Acc@5 96.582 (95.489) [2021-04-16 19:48:54 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.214) Loss 0.7966 (0.8194) Acc@1 81.348 (81.038) Acc@5 95.996 (95.501) [2021-04-16 19:49:09 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 81.034 Acc@5 95.498 [2021-04-16 19:49:09 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 81.0% [2021-04-16 19:49:09 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 81.06% [2021-04-16 19:49:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][0/1251] eta 6:52:44 lr 0.000022 time 19.7954 (19.7954) loss 3.2389 (3.2389) grad_norm 3.4741 (3.4741) [2021-04-16 19:49:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][10/1251] eta 0:42:35 lr 0.000022 time 0.3981 (2.0595) loss 3.2443 (2.9884) grad_norm 3.1261 (3.4276) [2021-04-16 19:49:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][20/1251] eta 0:24:53 lr 0.000022 time 0.2801 (1.2134) loss 2.3010 (2.9773) grad_norm 2.8384 (3.2888) [2021-04-16 19:49:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][30/1251] eta 0:18:42 lr 0.000022 time 0.2924 (0.9189) loss 3.1978 (2.9138) grad_norm 3.1449 (3.3052) [2021-04-16 19:49:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][40/1251] eta 0:15:23 lr 0.000022 time 0.2453 (0.7623) loss 3.9826 (2.9113) grad_norm 3.3864 (3.3489) [2021-04-16 19:49:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][50/1251] eta 0:13:21 lr 0.000022 time 0.2767 (0.6675) loss 3.1943 (2.9295) grad_norm 3.3339 (3.3463) [2021-04-16 19:49:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][60/1251] eta 0:11:58 lr 0.000022 time 0.3020 (0.6036) loss 3.5834 (2.8948) grad_norm 2.9178 (3.3387) [2021-04-16 19:49:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][70/1251] eta 0:11:00 lr 0.000022 time 0.2782 (0.5590) loss 2.5510 (2.8967) grad_norm 4.1551 (3.3454) [2021-04-16 19:49:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][80/1251] eta 0:10:14 lr 0.000022 time 0.2649 (0.5248) loss 3.2643 (2.9411) grad_norm 2.9983 (3.3379) [2021-04-16 19:49:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][90/1251] eta 0:09:37 lr 0.000022 time 0.2703 (0.4974) loss 2.9941 (2.9661) grad_norm 2.9887 (3.3175) [2021-04-16 19:49:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][100/1251] eta 0:09:09 lr 0.000022 time 0.2961 (0.4776) loss 3.0019 (2.9733) grad_norm 2.7319 (3.3176) [2021-04-16 19:50:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][110/1251] eta 0:08:45 lr 0.000022 time 0.4089 (0.4608) loss 2.9335 (2.9690) grad_norm 3.7097 (3.3028) [2021-04-16 19:50:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][120/1251] eta 0:08:23 lr 0.000022 time 0.2803 (0.4455) loss 1.8162 (2.9586) grad_norm 3.4196 (3.3187) [2021-04-16 19:50:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][130/1251] eta 0:08:06 lr 0.000022 time 0.2746 (0.4340) loss 2.8986 (2.9649) grad_norm 3.2313 (3.3387) [2021-04-16 19:50:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][140/1251] eta 0:07:50 lr 0.000022 time 0.3043 (0.4233) loss 3.8397 (2.9773) grad_norm 4.3667 (3.3268) [2021-04-16 19:50:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][150/1251] eta 0:07:37 lr 0.000022 time 0.2763 (0.4154) loss 1.8539 (2.9810) grad_norm 2.9209 (3.3236) [2021-04-16 19:50:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][160/1251] eta 0:07:23 lr 0.000022 time 0.2853 (0.4068) loss 3.3685 (2.9738) grad_norm 2.8928 (3.3235) [2021-04-16 19:50:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][170/1251] eta 0:07:11 lr 0.000022 time 0.2617 (0.3992) loss 3.4277 (2.9815) grad_norm 3.4879 (3.3346) [2021-04-16 19:50:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][180/1251] eta 0:07:00 lr 0.000022 time 0.3363 (0.3928) loss 3.2105 (2.9964) grad_norm 3.0350 (3.3280) [2021-04-16 19:50:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][190/1251] eta 0:06:50 lr 0.000022 time 0.2723 (0.3868) loss 2.5708 (2.9948) grad_norm 3.5214 (3.3456) [2021-04-16 19:50:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][200/1251] eta 0:06:40 lr 0.000022 time 0.2833 (0.3815) loss 3.1378 (2.9908) grad_norm 2.8081 (3.3402) [2021-04-16 19:50:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][210/1251] eta 0:06:32 lr 0.000022 time 0.2752 (0.3769) loss 2.6036 (2.9780) grad_norm 2.8211 (3.3312) [2021-04-16 19:50:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][220/1251] eta 0:06:24 lr 0.000022 time 0.2711 (0.3725) loss 2.8487 (2.9861) grad_norm 3.4377 (3.3279) [2021-04-16 19:50:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][230/1251] eta 0:06:16 lr 0.000022 time 0.2814 (0.3683) loss 3.6602 (2.9890) grad_norm 2.7170 (3.3168) [2021-04-16 19:50:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][240/1251] eta 0:06:08 lr 0.000022 time 0.2789 (0.3647) loss 2.7730 (2.9829) grad_norm 4.3978 (3.3156) [2021-04-16 19:50:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][250/1251] eta 0:06:01 lr 0.000022 time 0.2959 (0.3612) loss 3.1144 (2.9841) grad_norm 3.2357 (3.3079) [2021-04-16 19:50:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][260/1251] eta 0:05:54 lr 0.000022 time 0.2696 (0.3579) loss 2.9939 (2.9798) grad_norm 4.8537 (3.3221) [2021-04-16 19:50:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][270/1251] eta 0:05:48 lr 0.000022 time 0.2768 (0.3548) loss 2.4194 (2.9795) grad_norm 2.8007 (nan) [2021-04-16 19:50:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][280/1251] eta 0:05:41 lr 0.000022 time 0.2638 (0.3521) loss 3.3859 (2.9752) grad_norm 4.2021 (nan) [2021-04-16 19:50:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][290/1251] eta 0:05:35 lr 0.000022 time 0.2759 (0.3496) loss 2.4559 (2.9780) grad_norm 2.8296 (nan) [2021-04-16 19:50:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][300/1251] eta 0:05:30 lr 0.000022 time 0.2635 (0.3472) loss 2.8742 (2.9775) grad_norm 2.8842 (nan) [2021-04-16 19:50:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][310/1251] eta 0:05:24 lr 0.000022 time 0.2825 (0.3451) loss 1.8683 (2.9755) grad_norm 2.8979 (nan) [2021-04-16 19:50:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][320/1251] eta 0:05:19 lr 0.000022 time 0.2880 (0.3429) loss 2.7633 (2.9801) grad_norm 3.1195 (nan) [2021-04-16 19:51:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][330/1251] eta 0:05:13 lr 0.000022 time 0.2844 (0.3408) loss 3.4393 (2.9772) grad_norm 3.7675 (nan) [2021-04-16 19:51:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][340/1251] eta 0:05:08 lr 0.000022 time 0.2680 (0.3391) loss 3.2555 (2.9742) grad_norm 3.1767 (nan) [2021-04-16 19:51:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][350/1251] eta 0:05:03 lr 0.000022 time 0.2992 (0.3374) loss 3.0975 (2.9704) grad_norm 2.7425 (nan) [2021-04-16 19:51:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][360/1251] eta 0:04:59 lr 0.000022 time 0.3030 (0.3362) loss 3.3260 (2.9701) grad_norm 3.2732 (nan) [2021-04-16 19:51:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][370/1251] eta 0:04:55 lr 0.000022 time 0.2555 (0.3355) loss 2.5666 (2.9737) grad_norm 3.3634 (nan) [2021-04-16 19:51:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][380/1251] eta 0:04:50 lr 0.000022 time 0.3167 (0.3341) loss 3.0127 (2.9824) grad_norm 2.8889 (nan) [2021-04-16 19:51:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][390/1251] eta 0:04:47 lr 0.000022 time 0.4153 (0.3342) loss 3.5044 (2.9838) grad_norm 3.4695 (nan) [2021-04-16 19:51:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][400/1251] eta 0:04:43 lr 0.000022 time 0.2895 (0.3330) loss 3.0399 (2.9853) grad_norm 3.5944 (nan) [2021-04-16 19:51:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][410/1251] eta 0:04:38 lr 0.000022 time 0.2940 (0.3317) loss 3.2333 (2.9847) grad_norm 2.6856 (nan) [2021-04-16 19:51:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][420/1251] eta 0:04:34 lr 0.000022 time 0.2787 (0.3304) loss 3.3106 (2.9866) grad_norm 2.7220 (nan) [2021-04-16 19:51:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][430/1251] eta 0:04:30 lr 0.000022 time 0.2920 (0.3292) loss 3.3029 (2.9934) grad_norm 3.5416 (nan) [2021-04-16 19:51:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][440/1251] eta 0:04:25 lr 0.000022 time 0.2514 (0.3279) loss 2.5095 (2.9891) grad_norm 2.7579 (nan) [2021-04-16 19:51:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][450/1251] eta 0:04:21 lr 0.000022 time 0.2878 (0.3268) loss 2.8921 (2.9968) grad_norm 3.1807 (nan) [2021-04-16 19:51:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][460/1251] eta 0:04:17 lr 0.000022 time 0.3015 (0.3257) loss 2.7810 (3.0006) grad_norm 3.0205 (nan) [2021-04-16 19:51:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][470/1251] eta 0:04:13 lr 0.000022 time 0.2656 (0.3248) loss 3.0508 (3.0019) grad_norm 3.0194 (nan) [2021-04-16 19:51:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][480/1251] eta 0:04:09 lr 0.000021 time 0.2664 (0.3237) loss 2.0337 (2.9915) grad_norm 3.6019 (nan) [2021-04-16 19:51:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][490/1251] eta 0:04:05 lr 0.000021 time 0.2680 (0.3229) loss 3.1548 (2.9827) grad_norm 3.2283 (nan) [2021-04-16 19:51:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][500/1251] eta 0:04:01 lr 0.000021 time 0.2925 (0.3222) loss 2.5791 (2.9811) grad_norm 2.9034 (nan) [2021-04-16 19:51:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][510/1251] eta 0:03:58 lr 0.000021 time 0.2616 (0.3213) loss 3.2505 (2.9827) grad_norm 2.8002 (nan) [2021-04-16 19:51:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][520/1251] eta 0:03:54 lr 0.000021 time 0.2630 (0.3204) loss 2.1365 (2.9802) grad_norm 2.7467 (nan) [2021-04-16 19:51:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][530/1251] eta 0:03:50 lr 0.000021 time 0.2804 (0.3196) loss 3.2160 (2.9804) grad_norm 3.3804 (nan) [2021-04-16 19:52:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][540/1251] eta 0:03:46 lr 0.000021 time 0.2709 (0.3187) loss 3.2627 (2.9803) grad_norm 3.2894 (nan) [2021-04-16 19:52:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][550/1251] eta 0:03:43 lr 0.000021 time 0.2883 (0.3182) loss 2.5611 (2.9797) grad_norm 3.1118 (nan) [2021-04-16 19:52:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][560/1251] eta 0:03:39 lr 0.000021 time 0.2778 (0.3174) loss 3.4543 (2.9844) grad_norm 2.7917 (nan) [2021-04-16 19:52:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][570/1251] eta 0:03:35 lr 0.000021 time 0.2473 (0.3172) loss 3.3231 (2.9837) grad_norm 3.4608 (nan) [2021-04-16 19:52:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][580/1251] eta 0:03:32 lr 0.000021 time 0.2885 (0.3168) loss 3.4920 (2.9869) grad_norm 3.1093 (nan) [2021-04-16 19:52:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][590/1251] eta 0:03:29 lr 0.000021 time 0.2882 (0.3163) loss 2.7083 (2.9867) grad_norm 3.7132 (nan) [2021-04-16 19:52:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][600/1251] eta 0:03:25 lr 0.000021 time 0.2588 (0.3156) loss 2.1988 (2.9841) grad_norm 2.8510 (nan) [2021-04-16 19:52:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][610/1251] eta 0:03:22 lr 0.000021 time 0.2760 (0.3153) loss 3.6332 (2.9844) grad_norm 2.9004 (nan) [2021-04-16 19:52:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][620/1251] eta 0:03:18 lr 0.000021 time 0.2909 (0.3147) loss 1.8370 (2.9834) grad_norm 2.8108 (nan) [2021-04-16 19:52:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][630/1251] eta 0:03:15 lr 0.000021 time 0.2943 (0.3141) loss 3.0532 (2.9839) grad_norm 3.2408 (nan) [2021-04-16 19:52:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][640/1251] eta 0:03:11 lr 0.000021 time 0.2579 (0.3136) loss 3.5518 (2.9842) grad_norm 3.8654 (nan) [2021-04-16 19:52:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][650/1251] eta 0:03:08 lr 0.000021 time 0.3065 (0.3131) loss 3.3937 (2.9855) grad_norm 2.6033 (nan) [2021-04-16 19:52:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][660/1251] eta 0:03:04 lr 0.000021 time 0.2755 (0.3126) loss 2.3004 (2.9793) grad_norm 3.2908 (nan) [2021-04-16 19:52:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][670/1251] eta 0:03:01 lr 0.000021 time 0.2866 (0.3120) loss 3.5574 (2.9817) grad_norm 3.6016 (nan) [2021-04-16 19:52:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][680/1251] eta 0:02:57 lr 0.000021 time 0.2720 (0.3117) loss 3.0440 (2.9819) grad_norm 3.0954 (nan) [2021-04-16 19:52:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][690/1251] eta 0:02:54 lr 0.000021 time 0.2554 (0.3113) loss 3.1670 (2.9821) grad_norm 4.1976 (nan) [2021-04-16 19:52:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][700/1251] eta 0:02:51 lr 0.000021 time 0.4049 (0.3110) loss 3.2975 (2.9845) grad_norm 3.6635 (nan) [2021-04-16 19:52:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][710/1251] eta 0:02:47 lr 0.000021 time 0.2731 (0.3104) loss 2.9131 (2.9823) grad_norm 3.5247 (nan) [2021-04-16 19:52:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][720/1251] eta 0:02:44 lr 0.000021 time 0.2826 (0.3101) loss 3.3885 (2.9848) grad_norm 3.3986 (nan) [2021-04-16 19:52:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][730/1251] eta 0:02:41 lr 0.000021 time 0.2624 (0.3099) loss 2.5659 (2.9837) grad_norm 4.5482 (nan) [2021-04-16 19:52:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][740/1251] eta 0:02:38 lr 0.000021 time 0.2458 (0.3094) loss 2.8735 (2.9827) grad_norm 3.0878 (nan) [2021-04-16 19:53:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][750/1251] eta 0:02:34 lr 0.000021 time 0.2784 (0.3090) loss 3.3715 (2.9819) grad_norm 3.0207 (nan) [2021-04-16 19:53:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][760/1251] eta 0:02:31 lr 0.000021 time 0.2933 (0.3086) loss 2.9361 (2.9819) grad_norm 2.9575 (nan) [2021-04-16 19:53:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][770/1251] eta 0:02:28 lr 0.000021 time 0.2621 (0.3081) loss 3.3293 (2.9847) grad_norm 3.7080 (nan) [2021-04-16 19:53:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][780/1251] eta 0:02:24 lr 0.000021 time 0.2816 (0.3077) loss 2.6007 (2.9834) grad_norm 2.9996 (nan) [2021-04-16 19:53:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][790/1251] eta 0:02:21 lr 0.000021 time 0.2844 (0.3074) loss 2.2020 (2.9819) grad_norm 3.0371 (nan) [2021-04-16 19:53:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][800/1251] eta 0:02:18 lr 0.000021 time 0.2710 (0.3071) loss 2.8351 (2.9811) grad_norm 2.8327 (nan) [2021-04-16 19:53:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][810/1251] eta 0:02:15 lr 0.000021 time 0.2705 (0.3068) loss 2.9897 (2.9837) grad_norm 3.3248 (nan) [2021-04-16 19:53:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][820/1251] eta 0:02:12 lr 0.000021 time 0.2908 (0.3065) loss 2.3589 (2.9832) grad_norm 3.3197 (nan) [2021-04-16 19:53:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][830/1251] eta 0:02:08 lr 0.000021 time 0.2942 (0.3061) loss 3.1084 (2.9811) grad_norm 2.8362 (nan) [2021-04-16 19:53:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][840/1251] eta 0:02:05 lr 0.000021 time 0.2699 (0.3057) loss 3.0791 (2.9801) grad_norm 3.1143 (nan) [2021-04-16 19:53:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][850/1251] eta 0:02:02 lr 0.000021 time 0.2581 (0.3054) loss 2.8929 (2.9831) grad_norm 4.0284 (nan) [2021-04-16 19:53:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][860/1251] eta 0:01:59 lr 0.000021 time 0.2515 (0.3050) loss 2.5429 (2.9801) grad_norm 2.4622 (nan) [2021-04-16 19:53:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][870/1251] eta 0:01:56 lr 0.000021 time 0.2821 (0.3047) loss 3.0143 (2.9794) grad_norm 2.8766 (nan) [2021-04-16 19:53:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][880/1251] eta 0:01:52 lr 0.000021 time 0.2418 (0.3045) loss 3.6035 (2.9809) grad_norm 3.2122 (nan) [2021-04-16 19:53:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][890/1251] eta 0:01:49 lr 0.000021 time 0.2796 (0.3042) loss 1.8409 (2.9807) grad_norm 2.8858 (nan) [2021-04-16 19:53:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][900/1251] eta 0:01:46 lr 0.000021 time 0.2973 (0.3040) loss 3.4606 (2.9829) grad_norm 3.6872 (nan) [2021-04-16 19:53:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][910/1251] eta 0:01:43 lr 0.000021 time 0.2757 (0.3037) loss 3.2335 (2.9813) grad_norm 3.0793 (nan) [2021-04-16 19:53:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][920/1251] eta 0:01:40 lr 0.000021 time 0.2841 (0.3034) loss 3.4337 (2.9846) grad_norm 2.9150 (nan) [2021-04-16 19:53:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][930/1251] eta 0:01:37 lr 0.000021 time 0.2694 (0.3033) loss 3.4058 (2.9864) grad_norm 2.9159 (nan) [2021-04-16 19:53:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][940/1251] eta 0:01:34 lr 0.000021 time 0.3883 (0.3034) loss 3.3843 (2.9853) grad_norm 3.0749 (nan) [2021-04-16 19:53:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][950/1251] eta 0:01:31 lr 0.000021 time 0.2755 (0.3031) loss 3.3051 (2.9853) grad_norm 3.2546 (nan) [2021-04-16 19:54:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][960/1251] eta 0:01:28 lr 0.000021 time 0.2974 (0.3029) loss 3.4825 (2.9829) grad_norm 3.5236 (nan) [2021-04-16 19:54:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][970/1251] eta 0:01:25 lr 0.000021 time 0.2736 (0.3026) loss 2.3561 (2.9833) grad_norm 3.1695 (nan) [2021-04-16 19:54:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][980/1251] eta 0:01:21 lr 0.000021 time 0.2734 (0.3024) loss 3.5992 (2.9845) grad_norm 3.6556 (nan) [2021-04-16 19:54:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][990/1251] eta 0:01:18 lr 0.000021 time 0.2955 (0.3022) loss 2.3758 (2.9824) grad_norm 3.1306 (nan) [2021-04-16 19:54:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][1000/1251] eta 0:01:15 lr 0.000021 time 0.2529 (0.3020) loss 2.8637 (2.9817) grad_norm 3.5827 (nan) [2021-04-16 19:54:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][1010/1251] eta 0:01:12 lr 0.000021 time 0.2978 (0.3018) loss 2.9532 (2.9817) grad_norm 2.9074 (nan) [2021-04-16 19:54:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][1020/1251] eta 0:01:09 lr 0.000021 time 0.2635 (0.3016) loss 3.3331 (2.9814) grad_norm 3.3944 (nan) [2021-04-16 19:54:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][1030/1251] eta 0:01:06 lr 0.000021 time 0.2879 (0.3014) loss 2.5118 (2.9812) grad_norm 2.9492 (nan) [2021-04-16 19:54:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][1040/1251] eta 0:01:03 lr 0.000021 time 0.3050 (0.3012) loss 2.9498 (2.9834) grad_norm 2.9358 (nan) [2021-04-16 19:54:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][1050/1251] eta 0:01:00 lr 0.000021 time 0.2758 (0.3009) loss 2.7619 (2.9850) grad_norm 2.6602 (nan) [2021-04-16 19:54:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][1060/1251] eta 0:00:57 lr 0.000021 time 0.2591 (0.3007) loss 2.2212 (2.9861) grad_norm 3.0188 (nan) [2021-04-16 19:54:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][1070/1251] eta 0:00:54 lr 0.000021 time 0.2906 (0.3005) loss 3.3724 (2.9875) grad_norm 2.9043 (nan) [2021-04-16 19:54:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][1080/1251] eta 0:00:51 lr 0.000021 time 0.2734 (0.3003) loss 3.5131 (2.9872) grad_norm 3.4150 (nan) [2021-04-16 19:54:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][1090/1251] eta 0:00:48 lr 0.000021 time 0.2852 (0.3001) loss 2.1731 (2.9888) grad_norm 2.9114 (nan) [2021-04-16 19:54:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][1100/1251] eta 0:00:45 lr 0.000021 time 0.2875 (0.3000) loss 3.3705 (2.9890) grad_norm 2.7377 (nan) [2021-04-16 19:54:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][1110/1251] eta 0:00:42 lr 0.000021 time 0.2602 (0.2998) loss 3.2360 (2.9900) grad_norm 2.9883 (nan) [2021-04-16 19:54:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][1120/1251] eta 0:00:39 lr 0.000021 time 0.2770 (0.2996) loss 3.0117 (2.9898) grad_norm 5.3061 (nan) [2021-04-16 19:54:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][1130/1251] eta 0:00:36 lr 0.000021 time 0.2725 (0.2994) loss 3.7752 (2.9900) grad_norm 2.8141 (nan) [2021-04-16 19:54:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][1140/1251] eta 0:00:33 lr 0.000021 time 0.2705 (0.2993) loss 2.9682 (2.9899) grad_norm 3.2563 (nan) [2021-04-16 19:54:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][1150/1251] eta 0:00:30 lr 0.000021 time 0.3073 (0.2992) loss 3.1257 (2.9904) grad_norm 2.6016 (nan) [2021-04-16 19:54:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][1160/1251] eta 0:00:27 lr 0.000021 time 0.2763 (0.2991) loss 3.0731 (2.9905) grad_norm 3.3492 (nan) [2021-04-16 19:54:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][1170/1251] eta 0:00:24 lr 0.000021 time 0.2658 (0.2990) loss 2.6113 (2.9911) grad_norm 3.2443 (nan) [2021-04-16 19:55:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][1180/1251] eta 0:00:21 lr 0.000021 time 0.2676 (0.2990) loss 3.3434 (2.9926) grad_norm 3.5061 (nan) [2021-04-16 19:55:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][1190/1251] eta 0:00:18 lr 0.000021 time 0.2963 (0.2989) loss 3.0383 (2.9926) grad_norm 3.0195 (nan) [2021-04-16 19:55:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][1200/1251] eta 0:00:15 lr 0.000021 time 0.2806 (0.2987) loss 3.5541 (2.9929) grad_norm 2.9776 (nan) [2021-04-16 19:55:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][1210/1251] eta 0:00:12 lr 0.000021 time 0.2586 (0.2985) loss 2.0317 (2.9903) grad_norm 3.2030 (nan) [2021-04-16 19:55:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][1220/1251] eta 0:00:09 lr 0.000021 time 0.2426 (0.2985) loss 3.9116 (2.9921) grad_norm 3.4832 (nan) [2021-04-16 19:55:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][1230/1251] eta 0:00:06 lr 0.000021 time 0.2924 (0.2983) loss 3.9153 (2.9916) grad_norm 3.4769 (nan) [2021-04-16 19:55:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][1240/1251] eta 0:00:03 lr 0.000021 time 0.2480 (0.2981) loss 3.2614 (2.9934) grad_norm 4.3381 (nan) [2021-04-16 19:55:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [279/300][1250/1251] eta 0:00:00 lr 0.000021 time 0.2473 (0.2977) loss 3.2877 (2.9945) grad_norm 2.9132 (nan) [2021-04-16 19:55:27 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 279 training takes 0:06:18 [2021-04-16 19:55:27 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_279.pth saving...... [2021-04-16 19:55:41 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_279.pth saved !!! [2021-04-16 19:55:42 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.167 (1.167) Loss 0.7825 (0.7825) Acc@1 81.738 (81.738) Acc@5 95.410 (95.410) [2021-04-16 19:55:44 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.450 (0.241) Loss 0.8430 (0.8276) Acc@1 80.566 (80.717) Acc@5 94.727 (95.224) [2021-04-16 19:55:46 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.435 (0.237) Loss 0.7375 (0.8279) Acc@1 81.934 (80.757) Acc@5 96.387 (95.140) [2021-04-16 19:55:48 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.528 (0.238) Loss 0.8497 (0.8197) Acc@1 78.125 (80.929) Acc@5 95.117 (95.363) [2021-04-16 19:55:50 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.214) Loss 0.8310 (0.8179) Acc@1 80.664 (80.971) Acc@5 95.312 (95.417) [2021-04-16 19:56:09 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 81.046 Acc@5 95.440 [2021-04-16 19:56:09 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 81.0% [2021-04-16 19:56:09 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 81.06% [2021-04-16 19:56:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][0/1251] eta 6:50:08 lr 0.000021 time 19.6708 (19.6708) loss 3.7458 (3.7458) grad_norm 4.3202 (4.3202) [2021-04-16 19:56:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][10/1251] eta 0:42:03 lr 0.000021 time 0.2631 (2.0333) loss 3.1225 (3.1132) grad_norm 3.2395 (3.2209) [2021-04-16 19:56:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][20/1251] eta 0:24:36 lr 0.000021 time 0.3067 (1.1995) loss 2.5981 (3.0378) grad_norm 3.7818 (3.4009) [2021-04-16 19:56:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][30/1251] eta 0:18:20 lr 0.000021 time 0.2832 (0.9017) loss 3.3701 (3.1071) grad_norm 3.4782 (3.4112) [2021-04-16 19:56:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][40/1251] eta 0:15:08 lr 0.000021 time 0.2941 (0.7506) loss 2.5664 (3.1140) grad_norm 3.2564 (3.3789) [2021-04-16 19:56:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][50/1251] eta 0:13:10 lr 0.000021 time 0.2993 (0.6582) loss 3.6216 (3.0961) grad_norm 2.8261 (3.3425) [2021-04-16 19:56:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][60/1251] eta 0:11:48 lr 0.000021 time 0.2698 (0.5953) loss 2.7435 (3.0868) grad_norm 3.3115 (3.3192) [2021-04-16 19:56:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][70/1251] eta 0:10:49 lr 0.000021 time 0.2610 (0.5504) loss 2.3143 (3.0857) grad_norm 3.4178 (3.3574) [2021-04-16 19:56:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][80/1251] eta 0:10:05 lr 0.000021 time 0.2763 (0.5167) loss 3.4352 (3.0392) grad_norm 5.2108 (3.4333) [2021-04-16 19:56:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][90/1251] eta 0:09:30 lr 0.000021 time 0.2774 (0.4910) loss 3.1599 (3.0507) grad_norm 3.1453 (3.4053) [2021-04-16 19:56:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][100/1251] eta 0:09:00 lr 0.000021 time 0.2685 (0.4699) loss 3.6343 (3.0459) grad_norm 3.2478 (3.3956) [2021-04-16 19:56:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][110/1251] eta 0:08:35 lr 0.000021 time 0.2522 (0.4522) loss 2.9882 (3.0483) grad_norm 3.8992 (3.3965) [2021-04-16 19:57:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][120/1251] eta 0:08:15 lr 0.000021 time 0.2730 (0.4377) loss 3.2742 (3.0371) grad_norm 3.4735 (3.4076) [2021-04-16 19:57:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][130/1251] eta 0:07:56 lr 0.000021 time 0.2918 (0.4254) loss 2.2927 (3.0369) grad_norm 3.9841 (3.4070) [2021-04-16 19:57:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][140/1251] eta 0:07:42 lr 0.000021 time 0.2714 (0.4166) loss 3.3797 (3.0332) grad_norm 3.4981 (3.4294) [2021-04-16 19:57:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][150/1251] eta 0:07:28 lr 0.000021 time 0.3725 (0.4077) loss 2.2782 (3.0167) grad_norm 3.6777 (3.4194) [2021-04-16 19:57:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][160/1251] eta 0:07:16 lr 0.000021 time 0.2934 (0.3997) loss 2.3808 (3.0194) grad_norm 3.4036 (3.4425) [2021-04-16 19:57:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][170/1251] eta 0:07:03 lr 0.000021 time 0.2561 (0.3922) loss 2.3088 (3.0052) grad_norm 3.3198 (3.4351) [2021-04-16 19:57:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][180/1251] eta 0:06:53 lr 0.000021 time 0.2718 (0.3857) loss 3.6316 (3.0056) grad_norm 3.7403 (3.4251) [2021-04-16 19:57:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][190/1251] eta 0:06:43 lr 0.000021 time 0.2885 (0.3800) loss 3.7636 (3.0048) grad_norm 3.5441 (3.4067) [2021-04-16 19:57:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][200/1251] eta 0:06:33 lr 0.000021 time 0.2696 (0.3747) loss 3.1281 (3.0107) grad_norm 3.2392 (3.3919) [2021-04-16 19:57:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][210/1251] eta 0:06:25 lr 0.000021 time 0.2820 (0.3700) loss 2.2055 (3.0030) grad_norm 3.1926 (3.3880) [2021-04-16 19:57:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][220/1251] eta 0:06:17 lr 0.000021 time 0.2861 (0.3660) loss 3.1800 (2.9996) grad_norm 2.9270 (3.3784) [2021-04-16 19:57:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][230/1251] eta 0:06:09 lr 0.000021 time 0.2677 (0.3621) loss 2.9780 (2.9959) grad_norm 3.4367 (3.3776) [2021-04-16 19:57:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][240/1251] eta 0:06:02 lr 0.000021 time 0.2654 (0.3588) loss 1.8466 (2.9864) grad_norm 2.9458 (3.3864) [2021-04-16 19:57:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][250/1251] eta 0:05:55 lr 0.000021 time 0.2601 (0.3555) loss 3.0034 (2.9878) grad_norm 4.1037 (3.3857) [2021-04-16 19:57:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][260/1251] eta 0:05:49 lr 0.000021 time 0.2907 (0.3523) loss 3.8155 (2.9912) grad_norm 3.1046 (3.3840) [2021-04-16 19:57:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][270/1251] eta 0:05:42 lr 0.000021 time 0.2830 (0.3493) loss 3.5474 (3.0050) grad_norm 3.0041 (3.3821) [2021-04-16 19:57:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][280/1251] eta 0:05:36 lr 0.000021 time 0.2850 (0.3466) loss 1.5503 (3.0040) grad_norm 2.8229 (3.3813) [2021-04-16 19:57:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][290/1251] eta 0:05:30 lr 0.000021 time 0.2600 (0.3441) loss 2.6860 (3.0001) grad_norm 2.9693 (3.3714) [2021-04-16 19:57:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][300/1251] eta 0:05:25 lr 0.000021 time 0.2671 (0.3420) loss 2.7427 (3.0056) grad_norm 3.3988 (3.3759) [2021-04-16 19:57:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][310/1251] eta 0:05:20 lr 0.000021 time 0.2906 (0.3401) loss 3.4810 (3.0003) grad_norm 3.8789 (3.3713) [2021-04-16 19:57:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][320/1251] eta 0:05:14 lr 0.000021 time 0.2680 (0.3380) loss 3.3864 (3.0032) grad_norm 2.7763 (3.3697) [2021-04-16 19:58:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][330/1251] eta 0:05:09 lr 0.000021 time 0.2771 (0.3361) loss 3.8289 (3.0035) grad_norm 3.3689 (3.3670) [2021-04-16 19:58:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][340/1251] eta 0:05:05 lr 0.000021 time 0.2926 (0.3348) loss 3.4736 (3.0085) grad_norm 3.0751 (3.3640) [2021-04-16 19:58:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][350/1251] eta 0:05:00 lr 0.000021 time 0.2799 (0.3331) loss 2.7225 (3.0063) grad_norm 3.1704 (3.3616) [2021-04-16 19:58:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][360/1251] eta 0:04:55 lr 0.000021 time 0.2877 (0.3322) loss 2.6965 (3.0067) grad_norm 2.7976 (3.3574) [2021-04-16 19:58:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][370/1251] eta 0:04:51 lr 0.000021 time 0.2589 (0.3311) loss 2.1663 (3.0035) grad_norm 2.8890 (3.3496) [2021-04-16 19:58:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][380/1251] eta 0:04:47 lr 0.000020 time 0.3112 (0.3304) loss 2.9873 (3.0053) grad_norm 3.1748 (3.3455) [2021-04-16 19:58:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][390/1251] eta 0:04:43 lr 0.000020 time 0.2895 (0.3291) loss 2.8897 (3.0122) grad_norm 4.3581 (3.3587) [2021-04-16 19:58:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][400/1251] eta 0:04:39 lr 0.000020 time 0.2690 (0.3284) loss 3.0465 (3.0125) grad_norm 3.1784 (3.3725) [2021-04-16 19:58:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][410/1251] eta 0:04:35 lr 0.000020 time 0.2864 (0.3271) loss 2.4509 (3.0062) grad_norm 3.3599 (3.3697) [2021-04-16 19:58:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][420/1251] eta 0:04:30 lr 0.000020 time 0.2843 (0.3258) loss 2.8139 (2.9990) grad_norm 3.8204 (3.3667) [2021-04-16 19:58:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][430/1251] eta 0:04:26 lr 0.000020 time 0.2967 (0.3248) loss 2.8909 (2.9958) grad_norm 2.8034 (3.3594) [2021-04-16 19:58:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][440/1251] eta 0:04:22 lr 0.000020 time 0.2985 (0.3241) loss 3.1499 (2.9975) grad_norm 3.3017 (3.3600) [2021-04-16 19:58:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][450/1251] eta 0:04:18 lr 0.000020 time 0.2443 (0.3231) loss 3.0216 (2.9969) grad_norm 2.9143 (3.3573) [2021-04-16 19:58:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][460/1251] eta 0:04:14 lr 0.000020 time 0.2902 (0.3220) loss 3.3705 (2.9952) grad_norm 2.9257 (3.3604) [2021-04-16 19:58:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][470/1251] eta 0:04:10 lr 0.000020 time 0.2934 (0.3210) loss 2.0728 (2.9950) grad_norm 3.0226 (3.3603) [2021-04-16 19:58:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][480/1251] eta 0:04:06 lr 0.000020 time 0.2677 (0.3201) loss 2.9089 (2.9873) grad_norm 3.1569 (3.3561) [2021-04-16 19:58:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][490/1251] eta 0:04:03 lr 0.000020 time 0.2760 (0.3195) loss 3.0531 (2.9907) grad_norm 2.9797 (3.3522) [2021-04-16 19:58:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][500/1251] eta 0:03:59 lr 0.000020 time 0.3045 (0.3186) loss 2.3083 (2.9908) grad_norm 2.8914 (3.3530) [2021-04-16 19:58:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][510/1251] eta 0:03:55 lr 0.000020 time 0.2701 (0.3178) loss 2.5654 (2.9877) grad_norm 3.1483 (3.3516) [2021-04-16 19:58:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][520/1251] eta 0:03:51 lr 0.000020 time 0.2669 (0.3170) loss 2.9615 (2.9828) grad_norm 3.1044 (3.3554) [2021-04-16 19:58:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][530/1251] eta 0:03:47 lr 0.000020 time 0.2784 (0.3162) loss 3.2830 (2.9827) grad_norm 4.1009 (3.3540) [2021-04-16 19:59:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][540/1251] eta 0:03:44 lr 0.000020 time 0.2682 (0.3155) loss 3.4337 (2.9824) grad_norm 2.7190 (3.3524) [2021-04-16 19:59:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][550/1251] eta 0:03:40 lr 0.000020 time 0.2795 (0.3147) loss 3.1754 (2.9798) grad_norm 3.4326 (3.3523) [2021-04-16 19:59:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][560/1251] eta 0:03:36 lr 0.000020 time 0.2728 (0.3140) loss 3.1876 (2.9747) grad_norm 3.6528 (3.3527) [2021-04-16 19:59:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][570/1251] eta 0:03:33 lr 0.000020 time 0.3949 (0.3137) loss 3.4950 (2.9759) grad_norm 3.0177 (3.3517) [2021-04-16 19:59:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][580/1251] eta 0:03:30 lr 0.000020 time 0.2741 (0.3133) loss 2.1147 (2.9755) grad_norm 3.2884 (3.3490) [2021-04-16 19:59:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][590/1251] eta 0:03:26 lr 0.000020 time 0.2639 (0.3126) loss 1.9506 (2.9710) grad_norm 2.7166 (3.3471) [2021-04-16 19:59:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][600/1251] eta 0:03:23 lr 0.000020 time 0.2772 (0.3120) loss 3.1752 (2.9738) grad_norm 3.2038 (3.3488) [2021-04-16 19:59:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][610/1251] eta 0:03:19 lr 0.000020 time 0.2845 (0.3113) loss 2.6932 (2.9744) grad_norm 3.6189 (3.3489) [2021-04-16 19:59:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][620/1251] eta 0:03:16 lr 0.000020 time 0.2765 (0.3108) loss 3.2566 (2.9718) grad_norm 3.3523 (3.3493) [2021-04-16 19:59:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][630/1251] eta 0:03:12 lr 0.000020 time 0.2830 (0.3102) loss 3.2927 (2.9735) grad_norm 3.5056 (3.3494) [2021-04-16 19:59:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][640/1251] eta 0:03:09 lr 0.000020 time 0.2548 (0.3097) loss 3.1977 (2.9738) grad_norm 4.1702 (3.3514) [2021-04-16 19:59:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][650/1251] eta 0:03:05 lr 0.000020 time 0.2877 (0.3092) loss 3.5400 (2.9760) grad_norm 2.7134 (3.3484) [2021-04-16 19:59:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][660/1251] eta 0:03:02 lr 0.000020 time 0.2654 (0.3086) loss 2.2492 (2.9783) grad_norm 4.4216 (3.3529) [2021-04-16 19:59:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][670/1251] eta 0:02:59 lr 0.000020 time 0.2782 (0.3081) loss 2.1907 (2.9770) grad_norm 2.8561 (3.3516) [2021-04-16 19:59:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][680/1251] eta 0:02:55 lr 0.000020 time 0.2899 (0.3077) loss 3.8175 (2.9785) grad_norm 2.9634 (3.3499) [2021-04-16 19:59:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][690/1251] eta 0:02:52 lr 0.000020 time 0.2629 (0.3073) loss 1.8868 (2.9761) grad_norm 3.2840 (3.3517) [2021-04-16 19:59:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][700/1251] eta 0:02:49 lr 0.000020 time 0.2709 (0.3068) loss 3.0603 (2.9761) grad_norm 3.5256 (3.3504) [2021-04-16 19:59:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][710/1251] eta 0:02:45 lr 0.000020 time 0.2736 (0.3064) loss 2.3424 (2.9744) grad_norm 3.0763 (3.3475) [2021-04-16 19:59:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][720/1251] eta 0:02:42 lr 0.000020 time 0.2698 (0.3063) loss 3.2926 (2.9747) grad_norm 3.6927 (3.3466) [2021-04-16 19:59:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][730/1251] eta 0:02:39 lr 0.000020 time 0.2775 (0.3060) loss 3.5116 (2.9724) grad_norm 3.8216 (3.3470) [2021-04-16 19:59:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][740/1251] eta 0:02:36 lr 0.000020 time 0.2916 (0.3056) loss 3.2012 (2.9714) grad_norm 3.6030 (3.3502) [2021-04-16 19:59:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][750/1251] eta 0:02:32 lr 0.000020 time 0.2953 (0.3052) loss 2.3891 (2.9712) grad_norm 3.1435 (3.3490) [2021-04-16 20:00:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][760/1251] eta 0:02:29 lr 0.000020 time 0.2780 (0.3050) loss 2.7763 (2.9686) grad_norm 2.8310 (3.3516) [2021-04-16 20:00:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][770/1251] eta 0:02:26 lr 0.000020 time 0.3030 (0.3046) loss 3.2826 (2.9690) grad_norm 3.1980 (3.3519) [2021-04-16 20:00:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][780/1251] eta 0:02:23 lr 0.000020 time 0.2510 (0.3042) loss 2.4913 (2.9663) grad_norm 3.2945 (3.3534) [2021-04-16 20:00:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][790/1251] eta 0:02:20 lr 0.000020 time 0.2837 (0.3039) loss 3.2800 (2.9670) grad_norm 3.2733 (3.3547) [2021-04-16 20:00:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][800/1251] eta 0:02:16 lr 0.000020 time 0.2758 (0.3036) loss 3.3080 (2.9670) grad_norm 3.1255 (3.3530) [2021-04-16 20:00:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][810/1251] eta 0:02:13 lr 0.000020 time 0.2690 (0.3033) loss 3.8372 (2.9694) grad_norm 3.2663 (3.3519) [2021-04-16 20:00:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][820/1251] eta 0:02:10 lr 0.000020 time 0.2692 (0.3030) loss 3.6937 (2.9714) grad_norm 3.7547 (3.3539) [2021-04-16 20:00:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][830/1251] eta 0:02:07 lr 0.000020 time 0.2750 (0.3027) loss 2.8521 (2.9712) grad_norm 3.2060 (3.3547) [2021-04-16 20:00:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][840/1251] eta 0:02:04 lr 0.000020 time 0.2887 (0.3024) loss 3.6820 (2.9714) grad_norm 3.0359 (3.3534) [2021-04-16 20:00:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][850/1251] eta 0:02:01 lr 0.000020 time 0.2682 (0.3022) loss 1.9710 (2.9697) grad_norm 3.4629 (3.3515) [2021-04-16 20:00:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][860/1251] eta 0:01:58 lr 0.000020 time 0.2578 (0.3019) loss 3.4458 (2.9706) grad_norm 3.1942 (3.3497) [2021-04-16 20:00:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][870/1251] eta 0:01:54 lr 0.000020 time 0.2893 (0.3016) loss 3.3077 (2.9684) grad_norm 2.7249 (3.3462) [2021-04-16 20:00:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][880/1251] eta 0:01:51 lr 0.000020 time 0.2811 (0.3013) loss 3.4786 (2.9706) grad_norm 3.5041 (3.3481) [2021-04-16 20:00:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][890/1251] eta 0:01:48 lr 0.000020 time 0.2909 (0.3010) loss 3.5617 (2.9700) grad_norm 3.8803 (3.3505) [2021-04-16 20:00:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][900/1251] eta 0:01:45 lr 0.000020 time 0.2790 (0.3007) loss 2.8145 (2.9676) grad_norm 3.7533 (3.3530) [2021-04-16 20:00:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][910/1251] eta 0:01:42 lr 0.000020 time 0.2687 (0.3005) loss 2.5313 (2.9678) grad_norm 3.8685 (3.3550) [2021-04-16 20:00:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][920/1251] eta 0:01:39 lr 0.000020 time 0.2899 (0.3003) loss 3.0404 (2.9671) grad_norm 2.8214 (3.3536) [2021-04-16 20:00:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][930/1251] eta 0:01:36 lr 0.000020 time 0.2737 (0.3002) loss 2.6031 (2.9646) grad_norm 3.6464 (3.3529) [2021-04-16 20:00:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][940/1251] eta 0:01:33 lr 0.000020 time 0.2902 (0.3002) loss 3.1802 (2.9663) grad_norm 2.8072 (3.3525) [2021-04-16 20:00:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][950/1251] eta 0:01:30 lr 0.000020 time 0.2942 (0.3000) loss 3.4272 (2.9685) grad_norm 3.0422 (3.3498) [2021-04-16 20:00:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][960/1251] eta 0:01:27 lr 0.000020 time 0.3059 (0.2998) loss 1.8714 (2.9685) grad_norm 3.4394 (3.3477) [2021-04-16 20:01:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][970/1251] eta 0:01:24 lr 0.000020 time 0.2895 (0.2996) loss 3.1651 (2.9688) grad_norm 3.4479 (3.3508) [2021-04-16 20:01:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][980/1251] eta 0:01:21 lr 0.000020 time 0.2811 (0.2995) loss 3.0941 (2.9700) grad_norm 3.0827 (3.3524) [2021-04-16 20:01:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][990/1251] eta 0:01:18 lr 0.000020 time 0.3053 (0.2992) loss 2.3286 (2.9703) grad_norm 9.2312 (3.3602) [2021-04-16 20:01:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][1000/1251] eta 0:01:15 lr 0.000020 time 0.2603 (0.2990) loss 3.3103 (2.9704) grad_norm 3.1112 (3.3576) [2021-04-16 20:01:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][1010/1251] eta 0:01:12 lr 0.000020 time 0.2886 (0.2988) loss 3.1703 (2.9689) grad_norm 2.9173 (3.3559) [2021-04-16 20:01:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][1020/1251] eta 0:01:08 lr 0.000020 time 0.2587 (0.2986) loss 2.9714 (2.9696) grad_norm 3.3439 (3.3532) [2021-04-16 20:01:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][1030/1251] eta 0:01:05 lr 0.000020 time 0.2735 (0.2984) loss 2.2994 (2.9690) grad_norm 3.5001 (3.3523) [2021-04-16 20:01:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][1040/1251] eta 0:01:02 lr 0.000020 time 0.2774 (0.2982) loss 3.1967 (2.9678) grad_norm 2.6986 (3.3506) [2021-04-16 20:01:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][1050/1251] eta 0:00:59 lr 0.000020 time 0.2649 (0.2979) loss 3.4552 (2.9707) grad_norm 2.9785 (inf) [2021-04-16 20:01:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][1060/1251] eta 0:00:56 lr 0.000020 time 0.2966 (0.2977) loss 3.1913 (2.9707) grad_norm 3.2358 (inf) [2021-04-16 20:01:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][1070/1251] eta 0:00:53 lr 0.000020 time 0.2522 (0.2977) loss 2.4624 (2.9708) grad_norm 2.7190 (inf) [2021-04-16 20:01:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][1080/1251] eta 0:00:50 lr 0.000020 time 0.2767 (0.2974) loss 2.3485 (2.9699) grad_norm 3.1023 (inf) [2021-04-16 20:01:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][1090/1251] eta 0:00:47 lr 0.000020 time 0.2726 (0.2972) loss 2.5762 (2.9709) grad_norm 3.0326 (inf) [2021-04-16 20:01:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][1100/1251] eta 0:00:44 lr 0.000020 time 0.2713 (0.2971) loss 3.2835 (2.9715) grad_norm 3.2610 (inf) [2021-04-16 20:01:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][1110/1251] eta 0:00:41 lr 0.000020 time 0.2763 (0.2969) loss 3.4245 (2.9720) grad_norm 3.2943 (inf) [2021-04-16 20:01:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][1120/1251] eta 0:00:38 lr 0.000020 time 0.2771 (0.2967) loss 2.9334 (2.9723) grad_norm 3.0839 (inf) [2021-04-16 20:01:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][1130/1251] eta 0:00:35 lr 0.000020 time 0.2882 (0.2966) loss 2.9676 (2.9730) grad_norm 4.4792 (inf) [2021-04-16 20:01:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][1140/1251] eta 0:00:32 lr 0.000020 time 0.2602 (0.2964) loss 3.6888 (2.9739) grad_norm 2.8073 (inf) [2021-04-16 20:01:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][1150/1251] eta 0:00:29 lr 0.000020 time 0.2655 (0.2965) loss 3.2929 (2.9732) grad_norm 3.8485 (inf) [2021-04-16 20:01:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][1160/1251] eta 0:00:26 lr 0.000020 time 0.2813 (0.2964) loss 3.7621 (2.9749) grad_norm 3.1420 (inf) [2021-04-16 20:01:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][1170/1251] eta 0:00:23 lr 0.000020 time 0.2535 (0.2962) loss 3.4689 (2.9740) grad_norm 3.7443 (inf) [2021-04-16 20:01:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][1180/1251] eta 0:00:21 lr 0.000020 time 0.2659 (0.2961) loss 2.7375 (2.9749) grad_norm 2.8916 (inf) [2021-04-16 20:02:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][1190/1251] eta 0:00:18 lr 0.000020 time 0.2808 (0.2959) loss 2.2114 (2.9747) grad_norm 4.0251 (inf) [2021-04-16 20:02:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][1200/1251] eta 0:00:15 lr 0.000020 time 0.2978 (0.2957) loss 3.7497 (2.9762) grad_norm 3.1545 (inf) [2021-04-16 20:02:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][1210/1251] eta 0:00:12 lr 0.000020 time 0.2603 (0.2956) loss 2.8481 (2.9746) grad_norm 3.0230 (inf) [2021-04-16 20:02:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][1220/1251] eta 0:00:09 lr 0.000020 time 0.2766 (0.2954) loss 3.3535 (2.9747) grad_norm 3.1184 (inf) [2021-04-16 20:02:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][1230/1251] eta 0:00:06 lr 0.000020 time 0.3012 (0.2953) loss 2.9178 (2.9753) grad_norm 2.9850 (inf) [2021-04-16 20:02:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][1240/1251] eta 0:00:03 lr 0.000020 time 0.2503 (0.2950) loss 3.0314 (2.9743) grad_norm 2.8746 (inf) [2021-04-16 20:02:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [280/300][1250/1251] eta 0:00:00 lr 0.000020 time 0.2479 (0.2946) loss 2.5554 (2.9733) grad_norm 3.2020 (inf) [2021-04-16 20:02:25 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 280 training takes 0:06:16 [2021-04-16 20:02:25 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_280.pth saving...... [2021-04-16 20:02:43 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_280.pth saved !!! [2021-04-16 20:02:44 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.189 (1.189) Loss 0.8035 (0.8035) Acc@1 80.957 (80.957) Acc@5 95.215 (95.215) [2021-04-16 20:02:46 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.116 (0.249) Loss 0.7958 (0.8271) Acc@1 81.445 (80.593) Acc@5 95.605 (95.330) [2021-04-16 20:02:48 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.115 (0.236) Loss 0.7622 (0.8157) Acc@1 83.887 (81.083) Acc@5 96.387 (95.480) [2021-04-16 20:02:51 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.084 (0.247) Loss 0.8009 (0.8165) Acc@1 81.055 (81.092) Acc@5 96.484 (95.520) [2021-04-16 20:02:52 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.213) Loss 0.8293 (0.8187) Acc@1 81.738 (81.093) Acc@5 95.215 (95.498) [2021-04-16 20:03:13 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 81.104 Acc@5 95.490 [2021-04-16 20:03:13 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 81.1% [2021-04-16 20:03:13 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 81.10% [2021-04-16 20:03:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][0/1251] eta 4:04:56 lr 0.000020 time 11.7480 (11.7480) loss 3.3639 (3.3639) grad_norm 3.5011 (3.5011) [2021-04-16 20:03:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][10/1251] eta 0:27:13 lr 0.000020 time 0.2830 (1.3165) loss 2.9025 (3.0717) grad_norm 2.8702 (3.0606) [2021-04-16 20:03:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][20/1251] eta 0:16:54 lr 0.000020 time 0.3091 (0.8240) loss 3.3441 (3.0933) grad_norm 3.2373 (3.1316) [2021-04-16 20:03:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][30/1251] eta 0:13:10 lr 0.000020 time 0.2771 (0.6474) loss 2.2527 (2.9930) grad_norm 3.3688 (3.2929) [2021-04-16 20:03:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][40/1251] eta 0:11:14 lr 0.000020 time 0.2730 (0.5568) loss 3.1260 (2.9693) grad_norm 3.1067 (3.2771) [2021-04-16 20:03:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][50/1251] eta 0:10:04 lr 0.000020 time 0.2868 (0.5032) loss 3.3896 (3.0120) grad_norm 2.9072 (3.2776) [2021-04-16 20:03:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][60/1251] eta 0:09:14 lr 0.000020 time 0.2450 (0.4659) loss 3.2310 (3.0479) grad_norm 3.4392 (3.2483) [2021-04-16 20:03:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][70/1251] eta 0:08:39 lr 0.000020 time 0.2914 (0.4396) loss 3.5805 (2.9964) grad_norm 3.7976 (3.2626) [2021-04-16 20:03:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][80/1251] eta 0:08:11 lr 0.000020 time 0.3023 (0.4196) loss 3.0926 (2.9847) grad_norm 3.7828 (3.3069) [2021-04-16 20:03:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][90/1251] eta 0:07:49 lr 0.000020 time 0.2795 (0.4048) loss 2.6732 (2.9938) grad_norm 3.6698 (3.3123) [2021-04-16 20:03:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][100/1251] eta 0:07:31 lr 0.000020 time 0.2701 (0.3921) loss 3.4652 (2.9940) grad_norm 3.5260 (3.3123) [2021-04-16 20:03:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][110/1251] eta 0:07:16 lr 0.000020 time 0.2800 (0.3822) loss 3.0719 (2.9825) grad_norm 3.1274 (3.3153) [2021-04-16 20:03:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][120/1251] eta 0:07:03 lr 0.000020 time 0.2675 (0.3744) loss 3.2525 (2.9901) grad_norm 3.0641 (3.3066) [2021-04-16 20:04:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][130/1251] eta 0:06:51 lr 0.000020 time 0.2849 (0.3667) loss 3.3625 (2.9794) grad_norm 3.3238 (3.2905) [2021-04-16 20:04:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][140/1251] eta 0:06:42 lr 0.000020 time 0.2753 (0.3623) loss 1.9398 (2.9869) grad_norm 3.5691 (3.3119) [2021-04-16 20:04:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][150/1251] eta 0:06:33 lr 0.000020 time 0.2831 (0.3570) loss 2.8305 (3.0038) grad_norm 3.3672 (3.3256) [2021-04-16 20:04:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][160/1251] eta 0:06:25 lr 0.000020 time 0.2782 (0.3531) loss 3.1660 (3.0041) grad_norm 3.5876 (3.3165) [2021-04-16 20:04:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][170/1251] eta 0:06:17 lr 0.000020 time 0.2623 (0.3493) loss 3.1176 (3.0075) grad_norm 10.3156 (3.3474) [2021-04-16 20:04:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][180/1251] eta 0:06:11 lr 0.000020 time 0.4216 (0.3468) loss 2.9253 (3.0018) grad_norm 3.8530 (3.3626) [2021-04-16 20:04:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][190/1251] eta 0:06:03 lr 0.000020 time 0.2711 (0.3428) loss 2.9672 (3.0059) grad_norm 3.3484 (3.3650) [2021-04-16 20:04:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][200/1251] eta 0:05:57 lr 0.000020 time 0.2936 (0.3397) loss 3.0931 (2.9988) grad_norm 3.8164 (3.3650) [2021-04-16 20:04:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][210/1251] eta 0:05:50 lr 0.000020 time 0.2792 (0.3369) loss 2.7974 (2.9853) grad_norm 4.6698 (3.3685) [2021-04-16 20:04:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][220/1251] eta 0:05:44 lr 0.000020 time 0.2907 (0.3341) loss 3.1057 (2.9909) grad_norm 3.8179 (3.3733) [2021-04-16 20:04:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][230/1251] eta 0:05:38 lr 0.000020 time 0.3021 (0.3316) loss 3.2936 (2.9959) grad_norm 3.2724 (3.3640) [2021-04-16 20:04:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][240/1251] eta 0:05:32 lr 0.000020 time 0.2670 (0.3292) loss 3.5308 (3.0053) grad_norm 2.9906 (3.3521) [2021-04-16 20:04:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][250/1251] eta 0:05:27 lr 0.000020 time 0.2643 (0.3271) loss 3.7148 (3.0185) grad_norm 3.1688 (3.3510) [2021-04-16 20:04:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][260/1251] eta 0:05:22 lr 0.000020 time 0.2605 (0.3251) loss 3.2092 (3.0176) grad_norm 2.8743 (3.3447) [2021-04-16 20:04:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][270/1251] eta 0:05:17 lr 0.000020 time 0.2648 (0.3234) loss 3.3199 (3.0301) grad_norm 3.0837 (3.3360) [2021-04-16 20:04:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][280/1251] eta 0:05:12 lr 0.000020 time 0.2693 (0.3218) loss 2.6334 (3.0223) grad_norm 3.3670 (3.3327) [2021-04-16 20:04:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][290/1251] eta 0:05:07 lr 0.000020 time 0.2731 (0.3204) loss 1.8994 (3.0208) grad_norm 3.0317 (3.3311) [2021-04-16 20:04:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][300/1251] eta 0:05:03 lr 0.000020 time 0.2483 (0.3189) loss 2.5839 (3.0008) grad_norm 5.4239 (3.3318) [2021-04-16 20:04:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][310/1251] eta 0:04:58 lr 0.000020 time 0.2902 (0.3176) loss 2.0874 (3.0043) grad_norm 3.4753 (3.3322) [2021-04-16 20:04:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][320/1251] eta 0:04:54 lr 0.000020 time 0.3081 (0.3162) loss 3.5311 (3.0089) grad_norm 3.9023 (3.3338) [2021-04-16 20:04:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][330/1251] eta 0:04:50 lr 0.000019 time 0.2713 (0.3149) loss 3.5228 (3.0082) grad_norm 3.1461 (3.3347) [2021-04-16 20:05:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][340/1251] eta 0:04:45 lr 0.000019 time 0.2678 (0.3139) loss 3.5011 (3.0122) grad_norm 4.8340 (3.3339) [2021-04-16 20:05:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][350/1251] eta 0:04:41 lr 0.000019 time 0.2437 (0.3128) loss 3.0462 (3.0097) grad_norm 3.4730 (3.3331) [2021-04-16 20:05:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][360/1251] eta 0:04:38 lr 0.000019 time 0.2783 (0.3127) loss 3.1545 (3.0105) grad_norm 5.5638 (3.3359) [2021-04-16 20:05:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][370/1251] eta 0:04:35 lr 0.000019 time 0.3093 (0.3122) loss 2.2916 (2.9998) grad_norm 4.4506 (3.3334) [2021-04-16 20:05:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][380/1251] eta 0:04:31 lr 0.000019 time 0.2783 (0.3114) loss 3.3098 (3.0072) grad_norm 3.4912 (3.3341) [2021-04-16 20:05:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][390/1251] eta 0:04:27 lr 0.000019 time 0.2781 (0.3105) loss 3.7732 (3.0092) grad_norm 2.8593 (3.3315) [2021-04-16 20:05:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][400/1251] eta 0:04:23 lr 0.000019 time 0.2721 (0.3096) loss 3.0223 (3.0117) grad_norm 3.2650 (3.3309) [2021-04-16 20:05:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][410/1251] eta 0:04:19 lr 0.000019 time 0.3000 (0.3091) loss 2.2894 (3.0098) grad_norm 3.0805 (3.3245) [2021-04-16 20:05:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][420/1251] eta 0:04:16 lr 0.000019 time 0.2683 (0.3085) loss 3.3857 (3.0111) grad_norm 2.7773 (3.3230) [2021-04-16 20:05:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][430/1251] eta 0:04:12 lr 0.000019 time 0.2624 (0.3076) loss 2.0106 (3.0107) grad_norm 4.6704 (3.3316) [2021-04-16 20:05:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][440/1251] eta 0:04:08 lr 0.000019 time 0.2955 (0.3069) loss 3.3256 (3.0164) grad_norm 3.0562 (3.3278) [2021-04-16 20:05:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][450/1251] eta 0:04:05 lr 0.000019 time 0.2763 (0.3062) loss 3.6568 (3.0169) grad_norm 3.1490 (3.3215) [2021-04-16 20:05:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][460/1251] eta 0:04:01 lr 0.000019 time 0.2742 (0.3055) loss 3.2102 (3.0138) grad_norm 2.8599 (3.3194) [2021-04-16 20:05:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][470/1251] eta 0:03:58 lr 0.000019 time 0.2956 (0.3048) loss 3.3742 (3.0103) grad_norm 3.2243 (3.3205) [2021-04-16 20:05:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][480/1251] eta 0:03:54 lr 0.000019 time 0.2978 (0.3043) loss 2.0635 (3.0082) grad_norm 3.0871 (3.3190) [2021-04-16 20:05:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][490/1251] eta 0:03:51 lr 0.000019 time 0.2702 (0.3037) loss 1.7924 (2.9960) grad_norm 3.5452 (3.3186) [2021-04-16 20:05:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][500/1251] eta 0:03:47 lr 0.000019 time 0.2779 (0.3032) loss 3.2310 (2.9962) grad_norm 3.0122 (3.3180) [2021-04-16 20:05:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][510/1251] eta 0:03:44 lr 0.000019 time 0.2876 (0.3027) loss 2.3726 (2.9919) grad_norm 3.8498 (3.3181) [2021-04-16 20:05:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][520/1251] eta 0:03:41 lr 0.000019 time 0.2813 (0.3024) loss 2.3895 (2.9941) grad_norm 2.9069 (3.3189) [2021-04-16 20:05:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][530/1251] eta 0:03:37 lr 0.000019 time 0.3015 (0.3019) loss 3.3466 (2.9913) grad_norm 2.8702 (3.3172) [2021-04-16 20:05:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][540/1251] eta 0:03:34 lr 0.000019 time 0.2746 (0.3014) loss 3.0735 (2.9899) grad_norm 3.2748 (3.3176) [2021-04-16 20:05:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][550/1251] eta 0:03:30 lr 0.000019 time 0.2580 (0.3009) loss 2.4010 (2.9877) grad_norm 3.2932 (3.3151) [2021-04-16 20:06:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][560/1251] eta 0:03:27 lr 0.000019 time 0.2975 (0.3005) loss 2.9829 (2.9874) grad_norm 2.8653 (3.3172) [2021-04-16 20:06:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][570/1251] eta 0:03:24 lr 0.000019 time 0.2651 (0.3003) loss 2.6139 (2.9842) grad_norm 3.4163 (3.3210) [2021-04-16 20:06:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][580/1251] eta 0:03:21 lr 0.000019 time 0.2701 (0.3004) loss 2.6926 (2.9868) grad_norm 3.0582 (3.3196) [2021-04-16 20:06:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][590/1251] eta 0:03:18 lr 0.000019 time 0.2755 (0.3001) loss 2.4834 (2.9826) grad_norm 4.0108 (3.3206) [2021-04-16 20:06:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][600/1251] eta 0:03:15 lr 0.000019 time 0.2821 (0.2997) loss 3.3072 (2.9785) grad_norm 3.4304 (3.3193) [2021-04-16 20:06:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][610/1251] eta 0:03:11 lr 0.000019 time 0.2573 (0.2993) loss 3.2304 (2.9809) grad_norm 2.9527 (3.3192) [2021-04-16 20:06:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][620/1251] eta 0:03:08 lr 0.000019 time 0.2860 (0.2991) loss 2.6020 (2.9752) grad_norm 3.4431 (3.3260) [2021-04-16 20:06:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][630/1251] eta 0:03:05 lr 0.000019 time 0.2520 (0.2988) loss 3.1321 (2.9757) grad_norm 3.5987 (3.3315) [2021-04-16 20:06:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][640/1251] eta 0:03:02 lr 0.000019 time 0.2801 (0.2984) loss 3.1699 (2.9731) grad_norm 3.4352 (3.3289) [2021-04-16 20:06:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][650/1251] eta 0:02:59 lr 0.000019 time 0.2476 (0.2982) loss 2.9254 (2.9718) grad_norm 3.3023 (3.3304) [2021-04-16 20:06:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][660/1251] eta 0:02:56 lr 0.000019 time 0.2856 (0.2979) loss 2.3677 (2.9663) grad_norm 2.8456 (3.3318) [2021-04-16 20:06:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][670/1251] eta 0:02:52 lr 0.000019 time 0.2861 (0.2975) loss 3.3245 (2.9731) grad_norm 3.0501 (3.3307) [2021-04-16 20:06:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][680/1251] eta 0:02:49 lr 0.000019 time 0.2730 (0.2973) loss 2.5393 (2.9737) grad_norm 3.2589 (3.3277) [2021-04-16 20:06:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][690/1251] eta 0:02:46 lr 0.000019 time 0.2691 (0.2971) loss 3.3101 (2.9721) grad_norm 3.0631 (3.3256) [2021-04-16 20:06:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][700/1251] eta 0:02:43 lr 0.000019 time 0.2741 (0.2970) loss 3.6980 (2.9698) grad_norm 3.0766 (3.3247) [2021-04-16 20:06:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][710/1251] eta 0:02:40 lr 0.000019 time 0.2653 (0.2968) loss 2.8401 (2.9725) grad_norm 3.0155 (3.3227) [2021-04-16 20:06:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][720/1251] eta 0:02:37 lr 0.000019 time 0.2848 (0.2969) loss 2.8551 (2.9679) grad_norm 3.0573 (3.3213) [2021-04-16 20:06:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][730/1251] eta 0:02:34 lr 0.000019 time 0.4329 (0.2968) loss 2.7497 (2.9679) grad_norm 3.3592 (3.3202) [2021-04-16 20:06:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][740/1251] eta 0:02:31 lr 0.000019 time 0.4621 (0.2966) loss 3.5768 (2.9682) grad_norm 3.1024 (3.3190) [2021-04-16 20:06:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][750/1251] eta 0:02:28 lr 0.000019 time 0.2716 (0.2964) loss 3.2001 (2.9710) grad_norm 5.1310 (3.3245) [2021-04-16 20:06:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][760/1251] eta 0:02:25 lr 0.000019 time 0.2738 (0.2963) loss 3.0785 (2.9708) grad_norm 3.0640 (3.3267) [2021-04-16 20:07:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][770/1251] eta 0:02:22 lr 0.000019 time 0.2635 (0.2960) loss 2.5983 (2.9705) grad_norm 3.2224 (3.3271) [2021-04-16 20:07:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][780/1251] eta 0:02:19 lr 0.000019 time 0.3114 (0.2958) loss 2.0796 (2.9676) grad_norm 3.5464 (3.3262) [2021-04-16 20:07:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][790/1251] eta 0:02:16 lr 0.000019 time 0.2826 (0.2956) loss 3.0167 (2.9663) grad_norm 3.9453 (3.3260) [2021-04-16 20:07:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][800/1251] eta 0:02:13 lr 0.000019 time 0.2685 (0.2953) loss 3.0760 (2.9638) grad_norm 4.6104 (3.3286) [2021-04-16 20:07:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][810/1251] eta 0:02:10 lr 0.000019 time 0.2869 (0.2952) loss 3.0798 (2.9641) grad_norm 4.3996 (3.3305) [2021-04-16 20:07:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][820/1251] eta 0:02:07 lr 0.000019 time 0.2817 (0.2949) loss 3.3861 (2.9621) grad_norm 3.0092 (3.3311) [2021-04-16 20:07:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][830/1251] eta 0:02:04 lr 0.000019 time 0.2668 (0.2947) loss 3.3255 (2.9645) grad_norm 3.3526 (3.3312) [2021-04-16 20:07:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][840/1251] eta 0:02:01 lr 0.000019 time 0.2917 (0.2947) loss 3.3847 (2.9682) grad_norm 3.3494 (3.3300) [2021-04-16 20:07:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][850/1251] eta 0:01:58 lr 0.000019 time 0.2606 (0.2945) loss 1.8459 (2.9693) grad_norm 3.4885 (3.3305) [2021-04-16 20:07:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][860/1251] eta 0:01:55 lr 0.000019 time 0.2760 (0.2943) loss 3.5949 (2.9710) grad_norm 3.1296 (3.3306) [2021-04-16 20:07:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][870/1251] eta 0:01:52 lr 0.000019 time 0.2978 (0.2942) loss 3.5468 (2.9718) grad_norm 3.1267 (3.3306) [2021-04-16 20:07:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][880/1251] eta 0:01:49 lr 0.000019 time 0.2595 (0.2940) loss 1.8336 (2.9713) grad_norm 3.7431 (3.3300) [2021-04-16 20:07:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][890/1251] eta 0:01:46 lr 0.000019 time 0.2595 (0.2938) loss 3.2077 (2.9741) grad_norm 3.0095 (3.3282) [2021-04-16 20:07:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][900/1251] eta 0:01:43 lr 0.000019 time 0.2513 (0.2936) loss 3.0655 (2.9766) grad_norm 3.5700 (3.3314) [2021-04-16 20:07:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][910/1251] eta 0:01:40 lr 0.000019 time 0.2730 (0.2934) loss 3.0400 (2.9798) grad_norm 2.6425 (3.3304) [2021-04-16 20:07:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][920/1251] eta 0:01:37 lr 0.000019 time 0.2491 (0.2932) loss 3.1613 (2.9787) grad_norm 3.3696 (3.3302) [2021-04-16 20:07:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][930/1251] eta 0:01:34 lr 0.000019 time 0.2812 (0.2933) loss 3.5188 (2.9790) grad_norm 2.7307 (3.3277) [2021-04-16 20:07:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][940/1251] eta 0:01:31 lr 0.000019 time 0.3027 (0.2932) loss 3.5712 (2.9808) grad_norm 2.7478 (3.3267) [2021-04-16 20:07:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][950/1251] eta 0:01:28 lr 0.000019 time 0.2863 (0.2932) loss 3.3760 (2.9801) grad_norm 3.4680 (3.3252) [2021-04-16 20:07:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][960/1251] eta 0:01:25 lr 0.000019 time 0.2950 (0.2930) loss 3.5560 (2.9811) grad_norm 2.8704 (3.3269) [2021-04-16 20:07:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][970/1251] eta 0:01:22 lr 0.000019 time 0.2943 (0.2929) loss 2.9736 (2.9813) grad_norm 3.8174 (3.3280) [2021-04-16 20:08:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][980/1251] eta 0:01:19 lr 0.000019 time 0.2657 (0.2927) loss 2.7633 (2.9799) grad_norm 3.4198 (3.3305) [2021-04-16 20:08:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][990/1251] eta 0:01:16 lr 0.000019 time 0.2663 (0.2926) loss 2.7802 (2.9804) grad_norm 3.5970 (3.3307) [2021-04-16 20:08:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][1000/1251] eta 0:01:13 lr 0.000019 time 0.2724 (0.2924) loss 2.2339 (2.9787) grad_norm 3.4547 (3.3309) [2021-04-16 20:08:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][1010/1251] eta 0:01:10 lr 0.000019 time 0.2599 (0.2923) loss 2.7692 (2.9772) grad_norm 3.1284 (3.3332) [2021-04-16 20:08:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][1020/1251] eta 0:01:07 lr 0.000019 time 0.3099 (0.2923) loss 2.2454 (2.9766) grad_norm 2.9726 (3.3330) [2021-04-16 20:08:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][1030/1251] eta 0:01:04 lr 0.000019 time 0.2752 (0.2922) loss 2.7279 (2.9764) grad_norm 2.7905 (3.3333) [2021-04-16 20:08:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][1040/1251] eta 0:01:01 lr 0.000019 time 0.3003 (0.2921) loss 2.8967 (2.9763) grad_norm 4.0515 (3.3344) [2021-04-16 20:08:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][1050/1251] eta 0:00:58 lr 0.000019 time 0.2700 (0.2919) loss 3.0846 (2.9778) grad_norm 3.6805 (3.3362) [2021-04-16 20:08:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][1060/1251] eta 0:00:55 lr 0.000019 time 0.2737 (0.2918) loss 2.9606 (2.9781) grad_norm 3.1598 (3.3350) [2021-04-16 20:08:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][1070/1251] eta 0:00:52 lr 0.000019 time 0.2472 (0.2916) loss 3.0392 (2.9789) grad_norm 2.5751 (3.3335) [2021-04-16 20:08:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][1080/1251] eta 0:00:49 lr 0.000019 time 0.2838 (0.2916) loss 3.4763 (2.9812) grad_norm 3.4024 (3.3325) [2021-04-16 20:08:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][1090/1251] eta 0:00:46 lr 0.000019 time 0.2941 (0.2914) loss 2.8809 (2.9811) grad_norm 3.4317 (3.3317) [2021-04-16 20:08:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][1100/1251] eta 0:00:43 lr 0.000019 time 0.3077 (0.2913) loss 2.8175 (2.9813) grad_norm 3.3605 (3.3328) [2021-04-16 20:08:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][1110/1251] eta 0:00:41 lr 0.000019 time 0.2674 (0.2912) loss 3.8476 (2.9841) grad_norm 3.4175 (3.3317) [2021-04-16 20:08:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][1120/1251] eta 0:00:38 lr 0.000019 time 0.2673 (0.2910) loss 3.1227 (2.9848) grad_norm 3.2996 (3.3302) [2021-04-16 20:08:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][1130/1251] eta 0:00:35 lr 0.000019 time 0.2802 (0.2909) loss 3.5861 (2.9840) grad_norm 3.0104 (3.3282) [2021-04-16 20:08:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][1140/1251] eta 0:00:32 lr 0.000019 time 0.2736 (0.2910) loss 2.8563 (2.9835) grad_norm 3.3653 (3.3281) [2021-04-16 20:08:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][1150/1251] eta 0:00:29 lr 0.000019 time 0.2758 (0.2910) loss 3.0988 (2.9842) grad_norm 3.0879 (3.3270) [2021-04-16 20:08:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][1160/1251] eta 0:00:26 lr 0.000019 time 0.3023 (0.2910) loss 3.4890 (2.9849) grad_norm 3.3952 (3.3279) [2021-04-16 20:08:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][1170/1251] eta 0:00:23 lr 0.000019 time 0.2664 (0.2909) loss 3.0475 (2.9855) grad_norm 2.7225 (3.3265) [2021-04-16 20:08:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][1180/1251] eta 0:00:20 lr 0.000019 time 0.2743 (0.2908) loss 2.3878 (2.9836) grad_norm 2.8773 (3.3266) [2021-04-16 20:08:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][1190/1251] eta 0:00:17 lr 0.000019 time 0.3035 (0.2906) loss 2.8466 (2.9862) grad_norm 2.6412 (3.3281) [2021-04-16 20:09:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][1200/1251] eta 0:00:14 lr 0.000019 time 0.2757 (0.2905) loss 3.0515 (2.9874) grad_norm 3.0857 (3.3277) [2021-04-16 20:09:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][1210/1251] eta 0:00:11 lr 0.000019 time 0.2763 (0.2903) loss 3.7118 (2.9904) grad_norm 9.2335 (3.3326) [2021-04-16 20:09:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][1220/1251] eta 0:00:08 lr 0.000019 time 0.2861 (0.2902) loss 2.6906 (2.9887) grad_norm 3.2267 (3.3326) [2021-04-16 20:09:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][1230/1251] eta 0:00:06 lr 0.000019 time 0.2921 (0.2901) loss 3.3039 (2.9888) grad_norm 3.4704 (3.3338) [2021-04-16 20:09:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][1240/1251] eta 0:00:03 lr 0.000019 time 0.2495 (0.2900) loss 3.4848 (2.9886) grad_norm 3.5481 (3.3339) [2021-04-16 20:09:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [281/300][1250/1251] eta 0:00:00 lr 0.000019 time 0.2478 (0.2897) loss 2.9643 (2.9874) grad_norm 3.0681 (3.3328) [2021-04-16 20:09:24 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 281 training takes 0:06:11 [2021-04-16 20:09:25 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_281.pth saving...... [2021-04-16 20:09:35 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_281.pth saved !!! [2021-04-16 20:09:56 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 21.738 (21.738) Loss 0.8079 (0.8079) Acc@1 81.445 (81.445) Acc@5 95.020 (95.020) [2021-04-16 20:09:58 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.117 (2.121) Loss 0.8817 (0.8150) Acc@1 80.762 (81.046) Acc@5 94.824 (95.437) [2021-04-16 20:10:00 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.128 (1.192) Loss 0.7899 (0.8152) Acc@1 82.617 (81.069) Acc@5 95.312 (95.508) [2021-04-16 20:10:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.126 (0.897) Loss 0.8313 (0.8181) Acc@1 82.031 (81.124) Acc@5 95.312 (95.451) [2021-04-16 20:10:04 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.724) Loss 0.7639 (0.8204) Acc@1 82.520 (81.017) Acc@5 96.191 (95.420) [2021-04-16 20:10:22 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 81.038 Acc@5 95.454 [2021-04-16 20:10:22 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 81.0% [2021-04-16 20:10:22 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 81.10% [2021-04-16 20:10:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][0/1251] eta 7:18:49 lr 0.000019 time 21.0466 (21.0466) loss 3.1234 (3.1234) grad_norm 3.2166 (3.2166) [2021-04-16 20:10:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][10/1251] eta 0:44:38 lr 0.000019 time 0.2627 (2.1581) loss 3.1715 (2.9413) grad_norm 3.3004 (3.2934) [2021-04-16 20:10:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][20/1251] eta 0:25:56 lr 0.000019 time 0.2779 (1.2641) loss 2.9609 (3.0295) grad_norm 3.5650 (3.2825) [2021-04-16 20:10:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][30/1251] eta 0:19:15 lr 0.000019 time 0.2804 (0.9464) loss 2.8814 (3.0258) grad_norm 2.9538 (3.2895) [2021-04-16 20:10:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][40/1251] eta 0:15:48 lr 0.000019 time 0.2914 (0.7833) loss 1.5550 (2.9663) grad_norm 2.9703 (3.2654) [2021-04-16 20:10:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][50/1251] eta 0:13:42 lr 0.000019 time 0.2736 (0.6847) loss 2.9894 (2.9667) grad_norm 3.0553 (3.2845) [2021-04-16 20:10:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][60/1251] eta 0:12:18 lr 0.000019 time 0.3986 (0.6201) loss 3.6153 (2.9939) grad_norm 3.0352 (3.2601) [2021-04-16 20:11:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][70/1251] eta 0:11:14 lr 0.000019 time 0.2785 (0.5711) loss 2.6737 (2.9935) grad_norm 2.8196 (3.2706) [2021-04-16 20:11:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][80/1251] eta 0:10:25 lr 0.000019 time 0.2846 (0.5341) loss 2.4658 (2.9867) grad_norm 3.1981 (3.2793) [2021-04-16 20:11:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][90/1251] eta 0:09:47 lr 0.000019 time 0.2678 (0.5060) loss 3.4861 (2.9970) grad_norm 2.9479 (3.2771) [2021-04-16 20:11:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][100/1251] eta 0:09:16 lr 0.000019 time 0.3019 (0.4838) loss 3.2748 (3.0060) grad_norm 2.9265 (3.2736) [2021-04-16 20:11:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][110/1251] eta 0:08:50 lr 0.000019 time 0.2503 (0.4649) loss 2.5777 (2.9810) grad_norm 2.9039 (3.3987) [2021-04-16 20:11:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][120/1251] eta 0:08:28 lr 0.000019 time 0.2850 (0.4493) loss 3.0366 (2.9732) grad_norm 2.7346 (3.3776) [2021-04-16 20:11:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][130/1251] eta 0:08:09 lr 0.000019 time 0.2743 (0.4364) loss 3.6762 (2.9864) grad_norm 3.1127 (3.3626) [2021-04-16 20:11:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][140/1251] eta 0:07:54 lr 0.000019 time 0.2869 (0.4275) loss 3.1567 (3.0021) grad_norm 3.1366 (3.3775) [2021-04-16 20:11:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][150/1251] eta 0:07:40 lr 0.000019 time 0.2921 (0.4183) loss 2.5901 (3.0173) grad_norm 3.5501 (3.3677) [2021-04-16 20:11:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][160/1251] eta 0:07:27 lr 0.000019 time 0.2574 (0.4105) loss 3.3701 (3.0234) grad_norm 3.0402 (3.3539) [2021-04-16 20:11:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][170/1251] eta 0:07:15 lr 0.000019 time 0.2512 (0.4025) loss 3.2323 (3.0151) grad_norm 3.8964 (3.3571) [2021-04-16 20:11:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][180/1251] eta 0:07:04 lr 0.000019 time 0.2808 (0.3961) loss 2.8030 (3.0030) grad_norm 3.3414 (3.3561) [2021-04-16 20:11:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][190/1251] eta 0:06:53 lr 0.000019 time 0.2510 (0.3898) loss 2.8784 (3.0056) grad_norm 2.8956 (3.3635) [2021-04-16 20:11:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][200/1251] eta 0:06:43 lr 0.000019 time 0.2668 (0.3844) loss 3.5808 (3.0030) grad_norm 2.7676 (3.3580) [2021-04-16 20:11:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][210/1251] eta 0:06:34 lr 0.000019 time 0.2796 (0.3791) loss 3.4264 (3.0023) grad_norm 3.6024 (3.3588) [2021-04-16 20:11:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][220/1251] eta 0:06:26 lr 0.000019 time 0.2768 (0.3744) loss 3.2872 (3.0055) grad_norm 3.5334 (3.3609) [2021-04-16 20:11:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][230/1251] eta 0:06:17 lr 0.000019 time 0.2799 (0.3701) loss 3.9356 (3.0113) grad_norm 2.8900 (3.3475) [2021-04-16 20:11:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][240/1251] eta 0:06:10 lr 0.000019 time 0.2679 (0.3662) loss 2.9055 (3.0146) grad_norm 3.3044 (3.3451) [2021-04-16 20:11:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][250/1251] eta 0:06:02 lr 0.000019 time 0.2738 (0.3625) loss 3.0411 (3.0192) grad_norm 3.1136 (3.3450) [2021-04-16 20:11:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][260/1251] eta 0:05:56 lr 0.000019 time 0.2775 (0.3592) loss 2.5671 (3.0158) grad_norm 2.7458 (3.3458) [2021-04-16 20:11:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][270/1251] eta 0:05:49 lr 0.000019 time 0.2598 (0.3564) loss 3.1868 (3.0117) grad_norm 3.0226 (3.3401) [2021-04-16 20:12:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][280/1251] eta 0:05:43 lr 0.000019 time 0.2659 (0.3536) loss 2.5135 (3.0136) grad_norm 2.6242 (3.3374) [2021-04-16 20:12:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][290/1251] eta 0:05:37 lr 0.000019 time 0.2681 (0.3509) loss 1.7689 (3.0048) grad_norm 2.9540 (3.3393) [2021-04-16 20:12:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][300/1251] eta 0:05:31 lr 0.000019 time 0.2961 (0.3486) loss 2.8173 (3.0092) grad_norm 3.3306 (3.3352) [2021-04-16 20:12:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][310/1251] eta 0:05:25 lr 0.000019 time 0.2511 (0.3461) loss 3.2456 (3.0077) grad_norm 3.3464 (3.3409) [2021-04-16 20:12:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][320/1251] eta 0:05:20 lr 0.000019 time 0.2756 (0.3439) loss 2.0169 (3.0047) grad_norm 2.9793 (3.3364) [2021-04-16 20:12:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][330/1251] eta 0:05:14 lr 0.000019 time 0.2816 (0.3418) loss 3.1085 (2.9987) grad_norm 2.7871 (3.3323) [2021-04-16 20:12:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][340/1251] eta 0:05:10 lr 0.000019 time 0.2820 (0.3403) loss 3.4019 (3.0028) grad_norm 2.7808 (3.3340) [2021-04-16 20:12:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][350/1251] eta 0:05:05 lr 0.000018 time 0.2831 (0.3387) loss 3.4473 (3.0062) grad_norm 3.3790 (3.3331) [2021-04-16 20:12:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][360/1251] eta 0:05:00 lr 0.000018 time 0.2787 (0.3373) loss 2.5631 (2.9998) grad_norm 3.0057 (3.3352) [2021-04-16 20:12:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][370/1251] eta 0:04:56 lr 0.000018 time 0.2719 (0.3362) loss 2.1082 (2.9988) grad_norm 3.5144 (3.3328) [2021-04-16 20:12:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][380/1251] eta 0:04:51 lr 0.000018 time 0.2749 (0.3345) loss 3.4717 (2.9910) grad_norm 3.0972 (3.3286) [2021-04-16 20:12:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][390/1251] eta 0:04:46 lr 0.000018 time 0.2728 (0.3330) loss 3.3838 (2.9857) grad_norm 3.0820 (3.3274) [2021-04-16 20:12:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][400/1251] eta 0:04:42 lr 0.000018 time 0.3068 (0.3316) loss 2.2072 (2.9800) grad_norm 3.3558 (3.3249) [2021-04-16 20:12:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][410/1251] eta 0:04:37 lr 0.000018 time 0.2792 (0.3302) loss 1.9098 (2.9774) grad_norm 2.8948 (3.3225) [2021-04-16 20:12:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][420/1251] eta 0:04:33 lr 0.000018 time 0.2943 (0.3289) loss 3.2915 (2.9795) grad_norm 2.8360 (3.3208) [2021-04-16 20:12:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][430/1251] eta 0:04:29 lr 0.000018 time 0.2880 (0.3277) loss 3.0381 (2.9832) grad_norm 4.7542 (3.3255) [2021-04-16 20:12:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][440/1251] eta 0:04:24 lr 0.000018 time 0.2711 (0.3266) loss 2.4644 (2.9851) grad_norm 3.1559 (3.3243) [2021-04-16 20:12:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][450/1251] eta 0:04:20 lr 0.000018 time 0.2582 (0.3255) loss 3.4387 (2.9875) grad_norm 2.8259 (3.3251) [2021-04-16 20:12:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][460/1251] eta 0:04:16 lr 0.000018 time 0.2782 (0.3244) loss 2.2238 (2.9869) grad_norm 3.9498 (3.3296) [2021-04-16 20:12:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][470/1251] eta 0:04:12 lr 0.000018 time 0.2462 (0.3233) loss 3.0896 (2.9800) grad_norm 2.7680 (3.3248) [2021-04-16 20:12:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][480/1251] eta 0:04:08 lr 0.000018 time 0.2920 (0.3226) loss 2.6648 (2.9818) grad_norm 3.5058 (3.3273) [2021-04-16 20:13:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][490/1251] eta 0:04:04 lr 0.000018 time 0.2789 (0.3218) loss 2.9866 (2.9834) grad_norm 3.1141 (3.3256) [2021-04-16 20:13:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][500/1251] eta 0:04:01 lr 0.000018 time 0.2666 (0.3209) loss 2.7455 (2.9847) grad_norm 2.8973 (3.3207) [2021-04-16 20:13:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][510/1251] eta 0:03:57 lr 0.000018 time 0.2836 (0.3201) loss 3.3107 (2.9853) grad_norm 2.8407 (3.3176) [2021-04-16 20:13:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][520/1251] eta 0:03:53 lr 0.000018 time 0.2620 (0.3192) loss 2.9367 (2.9915) grad_norm 3.2500 (3.3203) [2021-04-16 20:13:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][530/1251] eta 0:03:49 lr 0.000018 time 0.2858 (0.3185) loss 3.3761 (2.9937) grad_norm 3.9749 (3.3169) [2021-04-16 20:13:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][540/1251] eta 0:03:46 lr 0.000018 time 0.2883 (0.3179) loss 2.7616 (2.9895) grad_norm 3.4022 (3.3127) [2021-04-16 20:13:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][550/1251] eta 0:03:42 lr 0.000018 time 0.2743 (0.3172) loss 3.5431 (2.9931) grad_norm 3.0914 (3.3151) [2021-04-16 20:13:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][560/1251] eta 0:03:38 lr 0.000018 time 0.2832 (0.3165) loss 2.9684 (2.9928) grad_norm 2.8638 (3.3110) [2021-04-16 20:13:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][570/1251] eta 0:03:34 lr 0.000018 time 0.2791 (0.3157) loss 3.1643 (2.9935) grad_norm 3.1177 (3.3095) [2021-04-16 20:13:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][580/1251] eta 0:03:31 lr 0.000018 time 0.2889 (0.3156) loss 2.2375 (2.9960) grad_norm 3.5201 (3.3079) [2021-04-16 20:13:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][590/1251] eta 0:03:28 lr 0.000018 time 0.2849 (0.3150) loss 3.2000 (3.0003) grad_norm 3.1291 (3.3091) [2021-04-16 20:13:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][600/1251] eta 0:03:24 lr 0.000018 time 0.2422 (0.3143) loss 3.0708 (3.0026) grad_norm 2.9770 (3.3043) [2021-04-16 20:13:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][610/1251] eta 0:03:21 lr 0.000018 time 0.2675 (0.3137) loss 3.3543 (3.0061) grad_norm 3.6843 (3.3050) [2021-04-16 20:13:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][620/1251] eta 0:03:17 lr 0.000018 time 0.2887 (0.3132) loss 2.9999 (3.0044) grad_norm 5.0126 (3.3069) [2021-04-16 20:13:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][630/1251] eta 0:03:14 lr 0.000018 time 0.2731 (0.3125) loss 1.8801 (3.0039) grad_norm 3.0080 (3.3107) [2021-04-16 20:13:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][640/1251] eta 0:03:10 lr 0.000018 time 0.2798 (0.3121) loss 2.9767 (3.0018) grad_norm 2.6814 (3.3120) [2021-04-16 20:13:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][650/1251] eta 0:03:07 lr 0.000018 time 0.2947 (0.3116) loss 2.8124 (2.9982) grad_norm 2.8701 (3.3107) [2021-04-16 20:13:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][660/1251] eta 0:03:03 lr 0.000018 time 0.2850 (0.3110) loss 3.6530 (2.9991) grad_norm 3.4317 (3.3114) [2021-04-16 20:13:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][670/1251] eta 0:03:00 lr 0.000018 time 0.2888 (0.3105) loss 3.9370 (2.9970) grad_norm 3.6267 (3.3101) [2021-04-16 20:13:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][680/1251] eta 0:02:57 lr 0.000018 time 0.2692 (0.3100) loss 3.4166 (2.9999) grad_norm 3.4369 (3.3112) [2021-04-16 20:13:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][690/1251] eta 0:02:53 lr 0.000018 time 0.2621 (0.3095) loss 3.3401 (3.0041) grad_norm 3.6588 (3.3127) [2021-04-16 20:13:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][700/1251] eta 0:02:50 lr 0.000018 time 0.2697 (0.3091) loss 3.4422 (3.0085) grad_norm 3.3095 (3.3139) [2021-04-16 20:14:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][710/1251] eta 0:02:47 lr 0.000018 time 0.2831 (0.3088) loss 3.7327 (3.0075) grad_norm 3.6325 (3.3134) [2021-04-16 20:14:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][720/1251] eta 0:02:43 lr 0.000018 time 0.2520 (0.3086) loss 1.9934 (3.0049) grad_norm 3.1195 (3.3212) [2021-04-16 20:14:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][730/1251] eta 0:02:40 lr 0.000018 time 0.2843 (0.3082) loss 2.6663 (3.0028) grad_norm 3.6382 (3.3231) [2021-04-16 20:14:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][740/1251] eta 0:02:37 lr 0.000018 time 0.2720 (0.3079) loss 3.1778 (3.0020) grad_norm 3.4727 (3.3236) [2021-04-16 20:14:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][750/1251] eta 0:02:34 lr 0.000018 time 0.2712 (0.3075) loss 3.1340 (3.0030) grad_norm 3.1832 (3.3228) [2021-04-16 20:14:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][760/1251] eta 0:02:30 lr 0.000018 time 0.2912 (0.3071) loss 2.9244 (3.0047) grad_norm 3.8843 (3.3238) [2021-04-16 20:14:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][770/1251] eta 0:02:27 lr 0.000018 time 0.2602 (0.3066) loss 3.2060 (3.0038) grad_norm 3.4532 (3.3254) [2021-04-16 20:14:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][780/1251] eta 0:02:24 lr 0.000018 time 0.2892 (0.3063) loss 2.8552 (3.0039) grad_norm 3.0252 (3.3230) [2021-04-16 20:14:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][790/1251] eta 0:02:21 lr 0.000018 time 0.2869 (0.3059) loss 2.9753 (3.0041) grad_norm 4.7506 (3.3232) [2021-04-16 20:14:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][800/1251] eta 0:02:17 lr 0.000018 time 0.2749 (0.3056) loss 2.9595 (3.0021) grad_norm 2.8537 (3.3209) [2021-04-16 20:14:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][810/1251] eta 0:02:14 lr 0.000018 time 0.2941 (0.3053) loss 2.9635 (3.0036) grad_norm 3.3150 (3.3203) [2021-04-16 20:14:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][820/1251] eta 0:02:11 lr 0.000018 time 0.2864 (0.3049) loss 2.9147 (3.0033) grad_norm 2.8437 (3.3193) [2021-04-16 20:14:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][830/1251] eta 0:02:08 lr 0.000018 time 0.2667 (0.3045) loss 3.0708 (3.0043) grad_norm 2.8075 (3.3189) [2021-04-16 20:14:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][840/1251] eta 0:02:05 lr 0.000018 time 0.2677 (0.3042) loss 3.3937 (3.0047) grad_norm 3.1904 (3.3178) [2021-04-16 20:14:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][850/1251] eta 0:02:01 lr 0.000018 time 0.2448 (0.3039) loss 3.4969 (3.0058) grad_norm 3.2192 (3.3173) [2021-04-16 20:14:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][860/1251] eta 0:01:58 lr 0.000018 time 0.2926 (0.3037) loss 3.2079 (3.0044) grad_norm 3.6516 (3.3178) [2021-04-16 20:14:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][870/1251] eta 0:01:55 lr 0.000018 time 0.2451 (0.3034) loss 2.9707 (3.0035) grad_norm 3.6694 (3.3199) [2021-04-16 20:14:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][880/1251] eta 0:01:52 lr 0.000018 time 0.2762 (0.3031) loss 2.3757 (3.0025) grad_norm 3.5180 (3.3206) [2021-04-16 20:14:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][890/1251] eta 0:01:49 lr 0.000018 time 0.3133 (0.3028) loss 2.2347 (3.0018) grad_norm 3.0709 (3.3162) [2021-04-16 20:14:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][900/1251] eta 0:01:46 lr 0.000018 time 0.2533 (0.3025) loss 2.8218 (3.0003) grad_norm 3.0610 (3.3166) [2021-04-16 20:14:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][910/1251] eta 0:01:43 lr 0.000018 time 0.2583 (0.3022) loss 3.1858 (3.0006) grad_norm 3.7849 (3.3159) [2021-04-16 20:15:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][920/1251] eta 0:01:39 lr 0.000018 time 0.2664 (0.3020) loss 2.1547 (2.9997) grad_norm 3.0449 (3.3160) [2021-04-16 20:15:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][930/1251] eta 0:01:36 lr 0.000018 time 0.2709 (0.3018) loss 3.4892 (3.0005) grad_norm 2.8172 (3.3137) [2021-04-16 20:15:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][940/1251] eta 0:01:33 lr 0.000018 time 0.2859 (0.3018) loss 3.2719 (2.9991) grad_norm 3.3549 (3.3129) [2021-04-16 20:15:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][950/1251] eta 0:01:30 lr 0.000018 time 0.2602 (0.3015) loss 3.4496 (3.0001) grad_norm 3.2734 (3.3137) [2021-04-16 20:15:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][960/1251] eta 0:01:27 lr 0.000018 time 0.2545 (0.3012) loss 2.1521 (3.0000) grad_norm 2.8784 (3.3106) [2021-04-16 20:15:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][970/1251] eta 0:01:24 lr 0.000018 time 0.2757 (0.3010) loss 2.2112 (3.0009) grad_norm 3.2361 (3.3103) [2021-04-16 20:15:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][980/1251] eta 0:01:21 lr 0.000018 time 0.2754 (0.3007) loss 3.4054 (2.9985) grad_norm 3.5278 (3.3108) [2021-04-16 20:15:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][990/1251] eta 0:01:18 lr 0.000018 time 0.2618 (0.3005) loss 2.9099 (2.9988) grad_norm 3.1702 (3.3089) [2021-04-16 20:15:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][1000/1251] eta 0:01:15 lr 0.000018 time 0.2783 (0.3002) loss 3.4265 (2.9998) grad_norm 2.9282 (3.3093) [2021-04-16 20:15:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][1010/1251] eta 0:01:12 lr 0.000018 time 0.2775 (0.3000) loss 3.4770 (3.0010) grad_norm 2.8662 (3.3103) [2021-04-16 20:15:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][1020/1251] eta 0:01:09 lr 0.000018 time 0.2929 (0.2997) loss 3.2425 (3.0032) grad_norm 3.0171 (3.3079) [2021-04-16 20:15:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][1030/1251] eta 0:01:06 lr 0.000018 time 0.2750 (0.2995) loss 2.7864 (3.0040) grad_norm 3.6993 (3.3074) [2021-04-16 20:15:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][1040/1251] eta 0:01:03 lr 0.000018 time 0.2754 (0.2993) loss 2.9849 (3.0039) grad_norm 4.1031 (3.3075) [2021-04-16 20:15:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][1050/1251] eta 0:01:00 lr 0.000018 time 0.2903 (0.2990) loss 3.2185 (3.0056) grad_norm 3.6426 (3.3062) [2021-04-16 20:15:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][1060/1251] eta 0:00:57 lr 0.000018 time 0.2726 (0.2988) loss 2.5793 (3.0049) grad_norm 2.8648 (3.3058) [2021-04-16 20:15:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][1070/1251] eta 0:00:54 lr 0.000018 time 0.2800 (0.2987) loss 3.3088 (3.0046) grad_norm 2.9303 (3.3048) [2021-04-16 20:15:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][1080/1251] eta 0:00:51 lr 0.000018 time 0.2760 (0.2985) loss 3.3822 (3.0038) grad_norm 2.8775 (3.3056) [2021-04-16 20:15:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][1090/1251] eta 0:00:48 lr 0.000018 time 0.2769 (0.2982) loss 3.1500 (3.0033) grad_norm 2.9567 (3.3052) [2021-04-16 20:15:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][1100/1251] eta 0:00:44 lr 0.000018 time 0.2685 (0.2980) loss 2.4408 (3.0027) grad_norm 3.0278 (3.3043) [2021-04-16 20:15:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][1110/1251] eta 0:00:41 lr 0.000018 time 0.2950 (0.2979) loss 2.1623 (3.0026) grad_norm 3.7345 (3.3038) [2021-04-16 20:15:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][1120/1251] eta 0:00:39 lr 0.000018 time 0.2784 (0.2978) loss 2.9029 (3.0023) grad_norm 2.7900 (3.3045) [2021-04-16 20:15:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][1130/1251] eta 0:00:36 lr 0.000018 time 0.2802 (0.2976) loss 2.9960 (3.0014) grad_norm 3.8921 (3.3060) [2021-04-16 20:16:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][1140/1251] eta 0:00:33 lr 0.000018 time 0.2765 (0.2974) loss 1.8095 (3.0003) grad_norm 2.6413 (3.3043) [2021-04-16 20:16:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][1150/1251] eta 0:00:30 lr 0.000018 time 0.2717 (0.2972) loss 2.2859 (3.0006) grad_norm 3.4628 (3.3032) [2021-04-16 20:16:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][1160/1251] eta 0:00:27 lr 0.000018 time 0.2886 (0.2973) loss 2.9806 (2.9993) grad_norm 3.1455 (3.3032) [2021-04-16 20:16:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][1170/1251] eta 0:00:24 lr 0.000018 time 0.2587 (0.2971) loss 3.3896 (3.0005) grad_norm 3.7607 (3.3026) [2021-04-16 20:16:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][1180/1251] eta 0:00:21 lr 0.000018 time 0.2923 (0.2969) loss 3.0127 (3.0002) grad_norm 3.3110 (3.3017) [2021-04-16 20:16:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][1190/1251] eta 0:00:18 lr 0.000018 time 0.2786 (0.2967) loss 3.3925 (3.0007) grad_norm 3.1859 (3.3015) [2021-04-16 20:16:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][1200/1251] eta 0:00:15 lr 0.000018 time 0.2617 (0.2965) loss 3.0267 (3.0018) grad_norm 2.9732 (3.3013) [2021-04-16 20:16:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][1210/1251] eta 0:00:12 lr 0.000018 time 0.2680 (0.2963) loss 3.2576 (2.9997) grad_norm 3.2739 (3.2997) [2021-04-16 20:16:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][1220/1251] eta 0:00:09 lr 0.000018 time 0.2786 (0.2963) loss 2.7019 (2.9994) grad_norm 3.2709 (3.3012) [2021-04-16 20:16:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][1230/1251] eta 0:00:06 lr 0.000018 time 0.2661 (0.2961) loss 3.1733 (3.0002) grad_norm 3.0950 (3.3012) [2021-04-16 20:16:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][1240/1251] eta 0:00:03 lr 0.000018 time 0.2479 (0.2959) loss 3.0376 (2.9978) grad_norm 3.1844 (3.3010) [2021-04-16 20:16:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [282/300][1250/1251] eta 0:00:00 lr 0.000018 time 0.2481 (0.2955) loss 2.7443 (2.9973) grad_norm 3.3200 (3.3010) [2021-04-16 20:16:37 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 282 training takes 0:06:15 [2021-04-16 20:16:37 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_282.pth saving...... [2021-04-16 20:16:54 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_282.pth saved !!! [2021-04-16 20:16:55 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.094 (1.094) Loss 0.7440 (0.7440) Acc@1 82.910 (82.910) Acc@5 96.191 (96.191) [2021-04-16 20:16:57 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.410 (0.219) Loss 0.8317 (0.8113) Acc@1 80.957 (80.735) Acc@5 95.020 (95.632) [2021-04-16 20:16:59 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.151 (0.227) Loss 0.7837 (0.8125) Acc@1 82.227 (80.966) Acc@5 95.898 (95.499) [2021-04-16 20:17:01 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.170 (0.225) Loss 0.8623 (0.8149) Acc@1 80.469 (81.029) Acc@5 94.629 (95.473) [2021-04-16 20:17:03 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.221) Loss 0.8176 (0.8133) Acc@1 80.469 (81.109) Acc@5 96.094 (95.541) [2021-04-16 20:17:24 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 81.134 Acc@5 95.554 [2021-04-16 20:17:24 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 81.1% [2021-04-16 20:17:24 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 81.13% [2021-04-16 20:17:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][0/1251] eta 3:00:17 lr 0.000018 time 8.6471 (8.6471) loss 3.1251 (3.1251) grad_norm 3.0153 (3.0153) [2021-04-16 20:17:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][10/1251] eta 0:21:19 lr 0.000018 time 0.2637 (1.0310) loss 3.4140 (2.9419) grad_norm 3.4864 (3.2364) [2021-04-16 20:17:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][20/1251] eta 0:13:49 lr 0.000018 time 0.3004 (0.6741) loss 3.2504 (3.0804) grad_norm 3.2558 (3.3600) [2021-04-16 20:17:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][30/1251] eta 0:11:06 lr 0.000018 time 0.2590 (0.5460) loss 3.5966 (3.0803) grad_norm 3.7812 (3.3465) [2021-04-16 20:17:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][40/1251] eta 0:09:42 lr 0.000018 time 0.2916 (0.4808) loss 2.4156 (3.0803) grad_norm 3.3918 (3.3041) [2021-04-16 20:17:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][50/1251] eta 0:08:49 lr 0.000018 time 0.2875 (0.4408) loss 2.5266 (3.0502) grad_norm 3.0813 (3.3222) [2021-04-16 20:17:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][60/1251] eta 0:08:14 lr 0.000018 time 0.2896 (0.4148) loss 3.4439 (3.0466) grad_norm 4.1281 (3.3370) [2021-04-16 20:17:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][70/1251] eta 0:07:46 lr 0.000018 time 0.2609 (0.3953) loss 2.9707 (3.0361) grad_norm 3.4415 (3.3548) [2021-04-16 20:17:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][80/1251] eta 0:07:25 lr 0.000018 time 0.2569 (0.3805) loss 1.7596 (3.0376) grad_norm 2.4376 (3.3376) [2021-04-16 20:17:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][90/1251] eta 0:07:09 lr 0.000018 time 0.2995 (0.3696) loss 2.9345 (3.0255) grad_norm 3.1034 (3.3515) [2021-04-16 20:18:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][100/1251] eta 0:06:55 lr 0.000018 time 0.2821 (0.3609) loss 2.4409 (2.9714) grad_norm 3.2348 (3.3305) [2021-04-16 20:18:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][110/1251] eta 0:06:43 lr 0.000018 time 0.3047 (0.3533) loss 3.0393 (2.9885) grad_norm 4.2534 (3.3375) [2021-04-16 20:18:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][120/1251] eta 0:06:33 lr 0.000018 time 0.2786 (0.3477) loss 2.9096 (2.9829) grad_norm 3.4641 (3.3239) [2021-04-16 20:18:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][130/1251] eta 0:06:24 lr 0.000018 time 0.2696 (0.3429) loss 3.5426 (2.9925) grad_norm 2.9451 (3.3120) [2021-04-16 20:18:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][140/1251] eta 0:06:16 lr 0.000018 time 0.2915 (0.3391) loss 3.2601 (2.9896) grad_norm 3.2665 (3.3101) [2021-04-16 20:18:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][150/1251] eta 0:06:09 lr 0.000018 time 0.2683 (0.3359) loss 2.8378 (2.9825) grad_norm 3.6348 (3.3030) [2021-04-16 20:18:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][160/1251] eta 0:06:02 lr 0.000018 time 0.2803 (0.3324) loss 3.2459 (2.9971) grad_norm 3.3564 (3.3133) [2021-04-16 20:18:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][170/1251] eta 0:05:55 lr 0.000018 time 0.2804 (0.3292) loss 3.1097 (2.9900) grad_norm 3.0206 (3.3116) [2021-04-16 20:18:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][180/1251] eta 0:05:49 lr 0.000018 time 0.2568 (0.3261) loss 2.8991 (2.9842) grad_norm 2.9065 (3.2989) [2021-04-16 20:18:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][190/1251] eta 0:05:43 lr 0.000018 time 0.2775 (0.3240) loss 3.2833 (2.9894) grad_norm 3.2060 (3.2945) [2021-04-16 20:18:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][200/1251] eta 0:05:37 lr 0.000018 time 0.2694 (0.3215) loss 2.5139 (2.9691) grad_norm 3.0511 (3.2910) [2021-04-16 20:18:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][210/1251] eta 0:05:32 lr 0.000018 time 0.2676 (0.3194) loss 2.8471 (2.9664) grad_norm 3.7490 (3.2928) [2021-04-16 20:18:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][220/1251] eta 0:05:27 lr 0.000018 time 0.2479 (0.3175) loss 3.1420 (2.9691) grad_norm 3.0825 (3.3082) [2021-04-16 20:18:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][230/1251] eta 0:05:22 lr 0.000018 time 0.2595 (0.3156) loss 1.9193 (2.9713) grad_norm 2.9607 (3.3115) [2021-04-16 20:18:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][240/1251] eta 0:05:17 lr 0.000018 time 0.3012 (0.3140) loss 2.8471 (2.9696) grad_norm 3.4844 (3.3250) [2021-04-16 20:18:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][250/1251] eta 0:05:12 lr 0.000018 time 0.2653 (0.3126) loss 3.1740 (2.9639) grad_norm 3.3015 (3.3300) [2021-04-16 20:18:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][260/1251] eta 0:05:08 lr 0.000018 time 0.2920 (0.3114) loss 3.2358 (2.9683) grad_norm 3.0755 (3.3347) [2021-04-16 20:18:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][270/1251] eta 0:05:04 lr 0.000018 time 0.2721 (0.3099) loss 3.0651 (2.9699) grad_norm 3.1433 (3.3365) [2021-04-16 20:18:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][280/1251] eta 0:04:59 lr 0.000018 time 0.2453 (0.3087) loss 3.1897 (2.9691) grad_norm 2.9611 (3.3464) [2021-04-16 20:18:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][290/1251] eta 0:04:55 lr 0.000018 time 0.2788 (0.3079) loss 2.9702 (2.9656) grad_norm 3.7890 (3.3779) [2021-04-16 20:18:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][300/1251] eta 0:04:51 lr 0.000018 time 0.2630 (0.3070) loss 2.2279 (2.9716) grad_norm 2.7287 (3.3732) [2021-04-16 20:18:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][310/1251] eta 0:04:48 lr 0.000018 time 0.2940 (0.3062) loss 2.7990 (2.9607) grad_norm 2.9457 (3.3663) [2021-04-16 20:19:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][320/1251] eta 0:04:44 lr 0.000018 time 0.3029 (0.3058) loss 1.9057 (2.9699) grad_norm 2.7991 (3.3720) [2021-04-16 20:19:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][330/1251] eta 0:04:40 lr 0.000018 time 0.2603 (0.3050) loss 3.1788 (2.9754) grad_norm 2.9524 (3.3667) [2021-04-16 20:19:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][340/1251] eta 0:04:37 lr 0.000018 time 0.2886 (0.3042) loss 2.8251 (2.9825) grad_norm 3.8579 (3.3650) [2021-04-16 20:19:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][350/1251] eta 0:04:33 lr 0.000018 time 0.2831 (0.3034) loss 2.3054 (2.9792) grad_norm 3.4754 (3.3604) [2021-04-16 20:19:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][360/1251] eta 0:04:30 lr 0.000018 time 0.2708 (0.3032) loss 2.5739 (2.9743) grad_norm 3.5977 (3.3656) [2021-04-16 20:19:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][370/1251] eta 0:04:26 lr 0.000018 time 0.3078 (0.3029) loss 2.9167 (2.9815) grad_norm 3.5362 (3.3741) [2021-04-16 20:19:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][380/1251] eta 0:04:23 lr 0.000018 time 0.2545 (0.3021) loss 3.2137 (2.9848) grad_norm 3.0498 (3.3709) [2021-04-16 20:19:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][390/1251] eta 0:04:19 lr 0.000018 time 0.2984 (0.3014) loss 2.7187 (2.9914) grad_norm 3.5516 (3.3750) [2021-04-16 20:19:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][400/1251] eta 0:04:16 lr 0.000018 time 0.2767 (0.3010) loss 3.3587 (2.9989) grad_norm 3.6875 (3.3702) [2021-04-16 20:19:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][410/1251] eta 0:04:12 lr 0.000018 time 0.3048 (0.3004) loss 3.0971 (2.9939) grad_norm 3.7835 (inf) [2021-04-16 20:19:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][420/1251] eta 0:04:09 lr 0.000018 time 0.2673 (0.3000) loss 3.6823 (2.9969) grad_norm 3.5407 (inf) [2021-04-16 20:19:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][430/1251] eta 0:04:05 lr 0.000018 time 0.2603 (0.2995) loss 3.2924 (3.0026) grad_norm 2.8553 (inf) [2021-04-16 20:19:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][440/1251] eta 0:04:02 lr 0.000018 time 0.3280 (0.2991) loss 3.1765 (3.0025) grad_norm 3.3754 (inf) [2021-04-16 20:19:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][450/1251] eta 0:03:59 lr 0.000017 time 0.2982 (0.2986) loss 3.3483 (2.9927) grad_norm 3.0784 (inf) [2021-04-16 20:19:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][460/1251] eta 0:03:55 lr 0.000017 time 0.2885 (0.2980) loss 3.3176 (2.9956) grad_norm 2.8511 (inf) [2021-04-16 20:19:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][470/1251] eta 0:03:52 lr 0.000017 time 0.2729 (0.2976) loss 2.1504 (2.9918) grad_norm 3.3021 (inf) [2021-04-16 20:19:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][480/1251] eta 0:03:49 lr 0.000017 time 0.3169 (0.2972) loss 3.3626 (2.9885) grad_norm 3.2808 (inf) [2021-04-16 20:19:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][490/1251] eta 0:03:45 lr 0.000017 time 0.2453 (0.2969) loss 3.0978 (2.9919) grad_norm 3.6881 (inf) [2021-04-16 20:19:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][500/1251] eta 0:03:42 lr 0.000017 time 0.2980 (0.2964) loss 3.2185 (2.9943) grad_norm 3.5321 (inf) [2021-04-16 20:19:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][510/1251] eta 0:03:39 lr 0.000017 time 0.2939 (0.2961) loss 3.2314 (2.9948) grad_norm 4.5087 (inf) [2021-04-16 20:19:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][520/1251] eta 0:03:36 lr 0.000017 time 0.2829 (0.2956) loss 3.0322 (2.9982) grad_norm 3.4015 (inf) [2021-04-16 20:20:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][530/1251] eta 0:03:32 lr 0.000017 time 0.2712 (0.2952) loss 3.4857 (2.9919) grad_norm 3.2838 (inf) [2021-04-16 20:20:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][540/1251] eta 0:03:29 lr 0.000017 time 0.2972 (0.2950) loss 2.8838 (2.9927) grad_norm 3.5323 (inf) [2021-04-16 20:20:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][550/1251] eta 0:03:26 lr 0.000017 time 0.2952 (0.2947) loss 2.4071 (2.9915) grad_norm 3.3463 (inf) [2021-04-16 20:20:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][560/1251] eta 0:03:23 lr 0.000017 time 0.2650 (0.2944) loss 3.3738 (2.9894) grad_norm 2.7902 (inf) [2021-04-16 20:20:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][570/1251] eta 0:03:20 lr 0.000017 time 0.2649 (0.2943) loss 3.4990 (2.9878) grad_norm 3.0501 (inf) [2021-04-16 20:20:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][580/1251] eta 0:03:17 lr 0.000017 time 0.2767 (0.2944) loss 2.0789 (2.9871) grad_norm 3.2317 (inf) [2021-04-16 20:20:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][590/1251] eta 0:03:14 lr 0.000017 time 0.2619 (0.2941) loss 3.1409 (2.9861) grad_norm 3.0187 (inf) [2021-04-16 20:20:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][600/1251] eta 0:03:11 lr 0.000017 time 0.2699 (0.2938) loss 3.4468 (2.9829) grad_norm 3.4168 (inf) [2021-04-16 20:20:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][610/1251] eta 0:03:08 lr 0.000017 time 0.2935 (0.2935) loss 3.2023 (2.9830) grad_norm 3.7822 (inf) [2021-04-16 20:20:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][620/1251] eta 0:03:05 lr 0.000017 time 0.2844 (0.2932) loss 2.4484 (2.9855) grad_norm 3.1269 (inf) [2021-04-16 20:20:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][630/1251] eta 0:03:01 lr 0.000017 time 0.2524 (0.2929) loss 3.1454 (2.9881) grad_norm 2.9733 (inf) [2021-04-16 20:20:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][640/1251] eta 0:02:58 lr 0.000017 time 0.2781 (0.2928) loss 3.3162 (2.9886) grad_norm 3.7756 (inf) [2021-04-16 20:20:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][650/1251] eta 0:02:55 lr 0.000017 time 0.2569 (0.2926) loss 3.3628 (2.9906) grad_norm 3.1786 (inf) [2021-04-16 20:20:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][660/1251] eta 0:02:52 lr 0.000017 time 0.2571 (0.2922) loss 3.0438 (2.9892) grad_norm 2.9579 (inf) [2021-04-16 20:20:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][670/1251] eta 0:02:49 lr 0.000017 time 0.2615 (0.2919) loss 2.3525 (2.9879) grad_norm 3.1131 (inf) [2021-04-16 20:20:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][680/1251] eta 0:02:46 lr 0.000017 time 0.2599 (0.2917) loss 2.8530 (2.9830) grad_norm 2.8251 (inf) [2021-04-16 20:20:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][690/1251] eta 0:02:43 lr 0.000017 time 0.2828 (0.2916) loss 2.7153 (2.9833) grad_norm 3.4909 (inf) [2021-04-16 20:20:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][700/1251] eta 0:02:40 lr 0.000017 time 0.3759 (0.2916) loss 3.4445 (2.9841) grad_norm 3.2343 (inf) [2021-04-16 20:20:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][710/1251] eta 0:02:37 lr 0.000017 time 0.2607 (0.2913) loss 3.7667 (2.9855) grad_norm 4.4018 (inf) [2021-04-16 20:20:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][720/1251] eta 0:02:34 lr 0.000017 time 0.2565 (0.2912) loss 3.8210 (2.9891) grad_norm 2.7902 (inf) [2021-04-16 20:20:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][730/1251] eta 0:02:31 lr 0.000017 time 0.2644 (0.2911) loss 3.4222 (2.9945) grad_norm 3.1418 (inf) [2021-04-16 20:20:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][740/1251] eta 0:02:28 lr 0.000017 time 0.2784 (0.2909) loss 2.9140 (2.9953) grad_norm 3.4521 (inf) [2021-04-16 20:21:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][750/1251] eta 0:02:25 lr 0.000017 time 0.3037 (0.2908) loss 3.7931 (2.9947) grad_norm 3.3051 (inf) [2021-04-16 20:21:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][760/1251] eta 0:02:22 lr 0.000017 time 0.2782 (0.2907) loss 3.6155 (2.9939) grad_norm 3.3538 (inf) [2021-04-16 20:21:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][770/1251] eta 0:02:19 lr 0.000017 time 0.2738 (0.2905) loss 2.0618 (2.9910) grad_norm 3.3048 (inf) [2021-04-16 20:21:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][780/1251] eta 0:02:16 lr 0.000017 time 0.2708 (0.2903) loss 3.2356 (2.9925) grad_norm 3.4685 (inf) [2021-04-16 20:21:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][790/1251] eta 0:02:13 lr 0.000017 time 0.2757 (0.2901) loss 3.2751 (2.9880) grad_norm 2.7350 (inf) [2021-04-16 20:21:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][800/1251] eta 0:02:10 lr 0.000017 time 0.2944 (0.2901) loss 3.4681 (2.9913) grad_norm 3.0927 (inf) [2021-04-16 20:21:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][810/1251] eta 0:02:07 lr 0.000017 time 0.2923 (0.2900) loss 3.5112 (2.9920) grad_norm 3.3686 (inf) [2021-04-16 20:21:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][820/1251] eta 0:02:04 lr 0.000017 time 0.2723 (0.2898) loss 2.9426 (2.9902) grad_norm 2.8274 (inf) [2021-04-16 20:21:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][830/1251] eta 0:02:01 lr 0.000017 time 0.3330 (0.2897) loss 3.5692 (2.9893) grad_norm 3.2451 (inf) [2021-04-16 20:21:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][840/1251] eta 0:01:58 lr 0.000017 time 0.2682 (0.2895) loss 2.7754 (2.9861) grad_norm 3.8088 (inf) [2021-04-16 20:21:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][850/1251] eta 0:01:55 lr 0.000017 time 0.2742 (0.2893) loss 3.1719 (2.9862) grad_norm 2.7425 (inf) [2021-04-16 20:21:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][860/1251] eta 0:01:53 lr 0.000017 time 0.2990 (0.2891) loss 3.4166 (2.9853) grad_norm 3.8522 (inf) [2021-04-16 20:21:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][870/1251] eta 0:01:50 lr 0.000017 time 0.2686 (0.2890) loss 2.5273 (2.9848) grad_norm 2.7663 (inf) [2021-04-16 20:21:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][880/1251] eta 0:01:47 lr 0.000017 time 0.2758 (0.2888) loss 3.3741 (2.9856) grad_norm 2.9612 (inf) [2021-04-16 20:21:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][890/1251] eta 0:01:44 lr 0.000017 time 0.2504 (0.2889) loss 2.7347 (2.9876) grad_norm 4.1518 (inf) [2021-04-16 20:21:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][900/1251] eta 0:01:41 lr 0.000017 time 0.2751 (0.2887) loss 2.9859 (2.9897) grad_norm 3.0459 (inf) [2021-04-16 20:21:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][910/1251] eta 0:01:38 lr 0.000017 time 0.2536 (0.2886) loss 2.5665 (2.9891) grad_norm 2.9976 (inf) [2021-04-16 20:21:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][920/1251] eta 0:01:35 lr 0.000017 time 0.2817 (0.2884) loss 3.4189 (2.9859) grad_norm 3.2791 (inf) [2021-04-16 20:21:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][930/1251] eta 0:01:32 lr 0.000017 time 0.2607 (0.2884) loss 2.7361 (2.9836) grad_norm 3.7010 (inf) [2021-04-16 20:21:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][940/1251] eta 0:01:29 lr 0.000017 time 0.2790 (0.2885) loss 3.3996 (2.9824) grad_norm 2.8899 (inf) [2021-04-16 20:21:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][950/1251] eta 0:01:26 lr 0.000017 time 0.2869 (0.2883) loss 2.4940 (2.9846) grad_norm 3.8536 (inf) [2021-04-16 20:22:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][960/1251] eta 0:01:23 lr 0.000017 time 0.2630 (0.2882) loss 3.8680 (2.9847) grad_norm 3.4380 (inf) [2021-04-16 20:22:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][970/1251] eta 0:01:20 lr 0.000017 time 0.2686 (0.2881) loss 3.1311 (2.9836) grad_norm 3.1258 (inf) [2021-04-16 20:22:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][980/1251] eta 0:01:18 lr 0.000017 time 0.2583 (0.2880) loss 3.2251 (2.9835) grad_norm 4.2820 (inf) [2021-04-16 20:22:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][990/1251] eta 0:01:15 lr 0.000017 time 0.2787 (0.2880) loss 2.4087 (2.9832) grad_norm 3.3939 (inf) [2021-04-16 20:22:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][1000/1251] eta 0:01:12 lr 0.000017 time 0.2978 (0.2879) loss 2.5869 (2.9839) grad_norm 3.0498 (inf) [2021-04-16 20:22:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][1010/1251] eta 0:01:09 lr 0.000017 time 0.2604 (0.2877) loss 3.2321 (2.9827) grad_norm 2.8102 (inf) [2021-04-16 20:22:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][1020/1251] eta 0:01:06 lr 0.000017 time 0.2639 (0.2876) loss 2.8661 (2.9847) grad_norm 3.1464 (inf) [2021-04-16 20:22:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][1030/1251] eta 0:01:03 lr 0.000017 time 0.2563 (0.2875) loss 3.1976 (2.9863) grad_norm 3.0537 (inf) [2021-04-16 20:22:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][1040/1251] eta 0:01:00 lr 0.000017 time 0.3003 (0.2875) loss 2.2908 (2.9850) grad_norm 4.7593 (inf) [2021-04-16 20:22:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][1050/1251] eta 0:00:57 lr 0.000017 time 0.2737 (0.2874) loss 2.9030 (2.9858) grad_norm 4.4038 (inf) [2021-04-16 20:22:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][1060/1251] eta 0:00:54 lr 0.000017 time 0.2733 (0.2873) loss 3.4767 (2.9868) grad_norm 3.0359 (inf) [2021-04-16 20:22:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][1070/1251] eta 0:00:51 lr 0.000017 time 0.3096 (0.2872) loss 2.7236 (2.9875) grad_norm 2.9998 (inf) [2021-04-16 20:22:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][1080/1251] eta 0:00:49 lr 0.000017 time 0.2700 (0.2871) loss 3.4647 (2.9884) grad_norm 3.2515 (inf) [2021-04-16 20:22:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][1090/1251] eta 0:00:46 lr 0.000017 time 0.2693 (0.2870) loss 3.1634 (2.9874) grad_norm 3.2279 (inf) [2021-04-16 20:22:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][1100/1251] eta 0:00:43 lr 0.000017 time 0.2917 (0.2869) loss 3.4473 (2.9853) grad_norm 3.0517 (inf) [2021-04-16 20:22:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][1110/1251] eta 0:00:40 lr 0.000017 time 0.2694 (0.2867) loss 3.3131 (2.9850) grad_norm 3.1774 (inf) [2021-04-16 20:22:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][1120/1251] eta 0:00:37 lr 0.000017 time 0.2697 (0.2866) loss 3.4725 (2.9859) grad_norm 3.3771 (inf) [2021-04-16 20:22:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][1130/1251] eta 0:00:34 lr 0.000017 time 0.2725 (0.2865) loss 2.8760 (2.9878) grad_norm 2.7973 (inf) [2021-04-16 20:22:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][1140/1251] eta 0:00:31 lr 0.000017 time 0.2574 (0.2864) loss 3.2210 (2.9890) grad_norm 3.9092 (inf) [2021-04-16 20:22:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][1150/1251] eta 0:00:28 lr 0.000017 time 0.2587 (0.2864) loss 3.4525 (2.9891) grad_norm 3.6621 (inf) [2021-04-16 20:22:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][1160/1251] eta 0:00:26 lr 0.000017 time 0.2918 (0.2865) loss 2.9429 (2.9860) grad_norm 2.5590 (inf) [2021-04-16 20:22:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][1170/1251] eta 0:00:23 lr 0.000017 time 0.2537 (0.2864) loss 3.4165 (2.9862) grad_norm 3.1089 (inf) [2021-04-16 20:23:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][1180/1251] eta 0:00:20 lr 0.000017 time 0.3038 (0.2864) loss 2.7091 (2.9857) grad_norm 3.2708 (inf) [2021-04-16 20:23:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][1190/1251] eta 0:00:17 lr 0.000017 time 0.2738 (0.2863) loss 3.6844 (2.9857) grad_norm 3.3966 (inf) [2021-04-16 20:23:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][1200/1251] eta 0:00:14 lr 0.000017 time 0.2736 (0.2862) loss 3.4209 (2.9847) grad_norm 3.2305 (inf) [2021-04-16 20:23:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][1210/1251] eta 0:00:11 lr 0.000017 time 0.2927 (0.2861) loss 3.0314 (2.9856) grad_norm 3.7396 (inf) [2021-04-16 20:23:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][1220/1251] eta 0:00:08 lr 0.000017 time 0.2969 (0.2860) loss 3.3672 (2.9851) grad_norm 2.8039 (inf) [2021-04-16 20:23:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][1230/1251] eta 0:00:06 lr 0.000017 time 0.2534 (0.2859) loss 3.2108 (2.9851) grad_norm 3.5684 (inf) [2021-04-16 20:23:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][1240/1251] eta 0:00:03 lr 0.000017 time 0.2487 (0.2857) loss 3.3398 (2.9857) grad_norm 3.4190 (inf) [2021-04-16 20:23:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [283/300][1250/1251] eta 0:00:00 lr 0.000017 time 0.2484 (0.2854) loss 2.8030 (2.9878) grad_norm 3.4429 (inf) [2021-04-16 20:23:30 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 283 training takes 0:06:06 [2021-04-16 20:23:30 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_283.pth saving...... [2021-04-16 20:23:52 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_283.pth saved !!! [2021-04-16 20:23:53 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.129 (1.129) Loss 0.8270 (0.8270) Acc@1 80.469 (80.469) Acc@5 95.215 (95.215) [2021-04-16 20:23:55 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.257 (0.287) Loss 0.8260 (0.8254) Acc@1 81.543 (80.877) Acc@5 95.117 (95.490) [2021-04-16 20:23:57 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.491 (0.238) Loss 0.7513 (0.8150) Acc@1 83.203 (81.041) Acc@5 95.508 (95.508) [2021-04-16 20:23:59 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.137 (0.225) Loss 0.8264 (0.8121) Acc@1 81.543 (81.143) Acc@5 94.824 (95.524) [2021-04-16 20:24:00 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.212) Loss 0.8675 (0.8118) Acc@1 80.078 (81.083) Acc@5 94.434 (95.482) [2021-04-16 20:24:17 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 81.108 Acc@5 95.468 [2021-04-16 20:24:17 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 81.1% [2021-04-16 20:24:17 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 81.13% [2021-04-16 20:24:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][0/1251] eta 3:39:33 lr 0.000017 time 10.5300 (10.5300) loss 3.2311 (3.2311) grad_norm 3.4311 (3.4311) [2021-04-16 20:24:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][10/1251] eta 0:24:55 lr 0.000017 time 0.3003 (1.2048) loss 2.9111 (2.7139) grad_norm 2.7845 (3.5158) [2021-04-16 20:24:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][20/1251] eta 0:15:41 lr 0.000017 time 0.2729 (0.7649) loss 3.1877 (2.9480) grad_norm 3.6460 (3.4570) [2021-04-16 20:24:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][30/1251] eta 0:12:23 lr 0.000017 time 0.2725 (0.6086) loss 3.2134 (2.9709) grad_norm 3.1426 (3.7247) [2021-04-16 20:24:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][40/1251] eta 0:10:39 lr 0.000017 time 0.3054 (0.5283) loss 3.3463 (3.0069) grad_norm 2.7989 (3.5920) [2021-04-16 20:24:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][50/1251] eta 0:09:35 lr 0.000017 time 0.2823 (0.4796) loss 3.1308 (2.9817) grad_norm 3.2290 (3.5286) [2021-04-16 20:24:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][60/1251] eta 0:08:51 lr 0.000017 time 0.3038 (0.4464) loss 2.3100 (2.9737) grad_norm 2.9160 (3.4670) [2021-04-16 20:24:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][70/1251] eta 0:08:19 lr 0.000017 time 0.3149 (0.4226) loss 3.0294 (2.9475) grad_norm 3.9729 (3.4268) [2021-04-16 20:24:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][80/1251] eta 0:07:55 lr 0.000017 time 0.2967 (0.4060) loss 3.4726 (2.9387) grad_norm 3.3097 (3.3989) [2021-04-16 20:24:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][90/1251] eta 0:07:35 lr 0.000017 time 0.2745 (0.3919) loss 2.9855 (2.9076) grad_norm 3.3807 (3.3804) [2021-04-16 20:24:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][100/1251] eta 0:07:19 lr 0.000017 time 0.2665 (0.3818) loss 3.2219 (2.9174) grad_norm 3.8646 (3.3627) [2021-04-16 20:24:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][110/1251] eta 0:07:05 lr 0.000017 time 0.2680 (0.3725) loss 2.7349 (2.9133) grad_norm 3.5453 (3.3506) [2021-04-16 20:25:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][120/1251] eta 0:06:52 lr 0.000017 time 0.2702 (0.3647) loss 3.3665 (2.9106) grad_norm 4.8773 (3.3690) [2021-04-16 20:25:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][130/1251] eta 0:06:41 lr 0.000017 time 0.2871 (0.3582) loss 3.2034 (2.9261) grad_norm 3.3126 (3.3651) [2021-04-16 20:25:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][140/1251] eta 0:06:33 lr 0.000017 time 0.2864 (0.3542) loss 2.8684 (2.9317) grad_norm 3.4917 (3.3670) [2021-04-16 20:25:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][150/1251] eta 0:06:24 lr 0.000017 time 0.3204 (0.3491) loss 2.4599 (2.9409) grad_norm 5.4232 (3.4014) [2021-04-16 20:25:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][160/1251] eta 0:06:16 lr 0.000017 time 0.2733 (0.3449) loss 2.3000 (2.9465) grad_norm 4.4311 (3.4237) [2021-04-16 20:25:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][170/1251] eta 0:06:08 lr 0.000017 time 0.2697 (0.3412) loss 3.5209 (2.9402) grad_norm 4.5003 (3.4392) [2021-04-16 20:25:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][180/1251] eta 0:06:01 lr 0.000017 time 0.2773 (0.3376) loss 3.2609 (2.9238) grad_norm 3.1365 (3.4362) [2021-04-16 20:25:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][190/1251] eta 0:05:54 lr 0.000017 time 0.2786 (0.3344) loss 3.1282 (2.9328) grad_norm 2.9485 (3.4370) [2021-04-16 20:25:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][200/1251] eta 0:05:48 lr 0.000017 time 0.2529 (0.3319) loss 2.6713 (2.9409) grad_norm 2.7040 (3.4346) [2021-04-16 20:25:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][210/1251] eta 0:05:42 lr 0.000017 time 0.2784 (0.3294) loss 2.2302 (2.9427) grad_norm 3.0297 (3.4438) [2021-04-16 20:25:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][220/1251] eta 0:05:37 lr 0.000017 time 0.2634 (0.3269) loss 3.1537 (2.9550) grad_norm 2.9514 (3.4328) [2021-04-16 20:25:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][230/1251] eta 0:05:31 lr 0.000017 time 0.2540 (0.3246) loss 3.2258 (2.9553) grad_norm 2.7513 (3.4271) [2021-04-16 20:25:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][240/1251] eta 0:05:26 lr 0.000017 time 0.2458 (0.3230) loss 3.5056 (2.9592) grad_norm 3.1684 (3.4168) [2021-04-16 20:25:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][250/1251] eta 0:05:21 lr 0.000017 time 0.2807 (0.3209) loss 3.2165 (2.9629) grad_norm 4.2935 (3.4197) [2021-04-16 20:25:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][260/1251] eta 0:05:16 lr 0.000017 time 0.3152 (0.3197) loss 2.7021 (2.9655) grad_norm 8.4248 (3.4400) [2021-04-16 20:25:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][270/1251] eta 0:05:12 lr 0.000017 time 0.2658 (0.3183) loss 2.8104 (2.9667) grad_norm 3.2946 (3.4355) [2021-04-16 20:25:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][280/1251] eta 0:05:07 lr 0.000017 time 0.2759 (0.3169) loss 1.8143 (2.9515) grad_norm 3.1651 (3.4343) [2021-04-16 20:25:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][290/1251] eta 0:05:03 lr 0.000017 time 0.2751 (0.3156) loss 2.2022 (2.9373) grad_norm 9.0213 (3.4672) [2021-04-16 20:25:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][300/1251] eta 0:04:59 lr 0.000017 time 0.2748 (0.3148) loss 2.8703 (2.9355) grad_norm 2.8840 (3.4616) [2021-04-16 20:25:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][310/1251] eta 0:04:55 lr 0.000017 time 0.2924 (0.3139) loss 3.2716 (2.9358) grad_norm 3.7187 (3.4962) [2021-04-16 20:25:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][320/1251] eta 0:04:51 lr 0.000017 time 0.2605 (0.3129) loss 3.4601 (2.9400) grad_norm 3.0073 (3.5028) [2021-04-16 20:26:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][330/1251] eta 0:04:47 lr 0.000017 time 0.2878 (0.3122) loss 2.7052 (2.9323) grad_norm 2.7803 (3.5011) [2021-04-16 20:26:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][340/1251] eta 0:04:43 lr 0.000017 time 0.2874 (0.3114) loss 3.2128 (2.9332) grad_norm 3.8323 (3.4951) [2021-04-16 20:26:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][350/1251] eta 0:04:39 lr 0.000017 time 0.2591 (0.3105) loss 3.4733 (2.9257) grad_norm 2.9471 (3.4892) [2021-04-16 20:26:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][360/1251] eta 0:04:36 lr 0.000017 time 0.2836 (0.3102) loss 3.1658 (2.9258) grad_norm 3.3418 (3.4794) [2021-04-16 20:26:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][370/1251] eta 0:04:33 lr 0.000017 time 0.2731 (0.3099) loss 3.2568 (2.9308) grad_norm 2.9557 (3.4764) [2021-04-16 20:26:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][380/1251] eta 0:04:29 lr 0.000017 time 0.2942 (0.3090) loss 2.5889 (2.9257) grad_norm 3.2922 (3.4762) [2021-04-16 20:26:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][390/1251] eta 0:04:25 lr 0.000017 time 0.3058 (0.3086) loss 3.0370 (2.9242) grad_norm 3.3670 (3.4673) [2021-04-16 20:26:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][400/1251] eta 0:04:21 lr 0.000017 time 0.2815 (0.3078) loss 2.0885 (2.9177) grad_norm 3.3475 (3.4581) [2021-04-16 20:26:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][410/1251] eta 0:04:18 lr 0.000017 time 0.2502 (0.3069) loss 3.0200 (2.9151) grad_norm 2.7794 (3.4503) [2021-04-16 20:26:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][420/1251] eta 0:04:14 lr 0.000017 time 0.2667 (0.3063) loss 2.5207 (2.9166) grad_norm 3.2613 (3.4512) [2021-04-16 20:26:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][430/1251] eta 0:04:10 lr 0.000017 time 0.2622 (0.3057) loss 3.5713 (2.9228) grad_norm 2.6392 (3.4433) [2021-04-16 20:26:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][440/1251] eta 0:04:07 lr 0.000017 time 0.2735 (0.3051) loss 2.5729 (2.9234) grad_norm 3.6240 (3.4381) [2021-04-16 20:26:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][450/1251] eta 0:04:03 lr 0.000017 time 0.2860 (0.3045) loss 1.9946 (2.9245) grad_norm 3.0297 (3.4454) [2021-04-16 20:26:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][460/1251] eta 0:04:00 lr 0.000017 time 0.2979 (0.3041) loss 3.1104 (2.9256) grad_norm 3.0544 (3.4421) [2021-04-16 20:26:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][470/1251] eta 0:03:57 lr 0.000017 time 0.2901 (0.3037) loss 2.7610 (2.9250) grad_norm 3.2056 (3.4377) [2021-04-16 20:26:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][480/1251] eta 0:03:53 lr 0.000017 time 0.2667 (0.3032) loss 3.3770 (2.9292) grad_norm 3.0067 (3.4369) [2021-04-16 20:26:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][490/1251] eta 0:03:50 lr 0.000017 time 0.2767 (0.3029) loss 3.3627 (2.9301) grad_norm 3.7889 (3.4372) [2021-04-16 20:26:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][500/1251] eta 0:03:47 lr 0.000017 time 0.2701 (0.3025) loss 2.7799 (2.9291) grad_norm 5.4505 (3.4439) [2021-04-16 20:26:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][510/1251] eta 0:03:43 lr 0.000017 time 0.2807 (0.3019) loss 3.4395 (2.9355) grad_norm 2.8795 (3.4407) [2021-04-16 20:26:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][520/1251] eta 0:03:40 lr 0.000017 time 0.2491 (0.3015) loss 3.1107 (2.9352) grad_norm 2.7688 (3.4383) [2021-04-16 20:26:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][530/1251] eta 0:03:37 lr 0.000017 time 0.2931 (0.3011) loss 2.3343 (2.9348) grad_norm 3.8038 (3.4348) [2021-04-16 20:27:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][540/1251] eta 0:03:33 lr 0.000017 time 0.2805 (0.3007) loss 2.5233 (2.9311) grad_norm 3.7411 (3.4337) [2021-04-16 20:27:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][550/1251] eta 0:03:30 lr 0.000017 time 0.2859 (0.3003) loss 3.7808 (2.9328) grad_norm 2.9217 (3.4384) [2021-04-16 20:27:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][560/1251] eta 0:03:27 lr 0.000017 time 0.2854 (0.2998) loss 3.2025 (2.9350) grad_norm 2.7954 (3.4381) [2021-04-16 20:27:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][570/1251] eta 0:03:24 lr 0.000017 time 0.4102 (0.3001) loss 3.6970 (2.9413) grad_norm 4.3505 (3.4322) [2021-04-16 20:27:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][580/1251] eta 0:03:21 lr 0.000017 time 0.2726 (0.2997) loss 2.1985 (2.9368) grad_norm 3.8352 (3.4302) [2021-04-16 20:27:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][590/1251] eta 0:03:17 lr 0.000017 time 0.2654 (0.2993) loss 2.8911 (2.9352) grad_norm 3.1327 (3.4274) [2021-04-16 20:27:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][600/1251] eta 0:03:14 lr 0.000017 time 0.2891 (0.2990) loss 2.6218 (2.9367) grad_norm 3.6773 (3.4271) [2021-04-16 20:27:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][610/1251] eta 0:03:11 lr 0.000017 time 0.2865 (0.2986) loss 3.1060 (2.9360) grad_norm 3.3632 (3.4224) [2021-04-16 20:27:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][620/1251] eta 0:03:08 lr 0.000017 time 0.2761 (0.2982) loss 3.4774 (2.9407) grad_norm 3.0277 (3.4218) [2021-04-16 20:27:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][630/1251] eta 0:03:05 lr 0.000017 time 0.2826 (0.2979) loss 2.1481 (2.9417) grad_norm 3.1120 (3.4205) [2021-04-16 20:27:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][640/1251] eta 0:03:01 lr 0.000016 time 0.2802 (0.2976) loss 3.4639 (2.9446) grad_norm 3.4918 (3.4188) [2021-04-16 20:27:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][650/1251] eta 0:02:58 lr 0.000016 time 0.2871 (0.2974) loss 2.6492 (2.9479) grad_norm 2.7956 (3.4190) [2021-04-16 20:27:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][660/1251] eta 0:02:55 lr 0.000016 time 0.2757 (0.2971) loss 2.7860 (2.9479) grad_norm 3.6236 (3.4184) [2021-04-16 20:27:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][670/1251] eta 0:02:52 lr 0.000016 time 0.2808 (0.2969) loss 2.6455 (2.9497) grad_norm 3.6766 (3.4254) [2021-04-16 20:27:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][680/1251] eta 0:02:49 lr 0.000016 time 0.2951 (0.2968) loss 2.6576 (2.9507) grad_norm 2.8868 (3.4237) [2021-04-16 20:27:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][690/1251] eta 0:02:46 lr 0.000016 time 0.2780 (0.2965) loss 3.5847 (2.9533) grad_norm 3.0162 (3.4228) [2021-04-16 20:27:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][700/1251] eta 0:02:43 lr 0.000016 time 0.2754 (0.2963) loss 3.1117 (2.9570) grad_norm 3.5384 (3.4223) [2021-04-16 20:27:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][710/1251] eta 0:02:40 lr 0.000016 time 0.2930 (0.2961) loss 3.8494 (2.9604) grad_norm 3.6404 (3.4240) [2021-04-16 20:27:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][720/1251] eta 0:02:37 lr 0.000016 time 0.2823 (0.2962) loss 2.7078 (2.9632) grad_norm 3.0810 (3.4232) [2021-04-16 20:27:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][730/1251] eta 0:02:34 lr 0.000016 time 0.2545 (0.2960) loss 2.9262 (2.9630) grad_norm 3.0781 (3.4229) [2021-04-16 20:27:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][740/1251] eta 0:02:31 lr 0.000016 time 0.2704 (0.2957) loss 3.2389 (2.9630) grad_norm 3.0750 (3.4215) [2021-04-16 20:27:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][750/1251] eta 0:02:28 lr 0.000016 time 0.2572 (0.2955) loss 2.7817 (2.9637) grad_norm 3.8582 (3.4207) [2021-04-16 20:28:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][760/1251] eta 0:02:24 lr 0.000016 time 0.2906 (0.2953) loss 2.7572 (2.9624) grad_norm 3.7173 (3.4237) [2021-04-16 20:28:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][770/1251] eta 0:02:21 lr 0.000016 time 0.2732 (0.2951) loss 2.4397 (2.9620) grad_norm 3.0489 (3.4260) [2021-04-16 20:28:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][780/1251] eta 0:02:18 lr 0.000016 time 0.2799 (0.2950) loss 2.8420 (2.9604) grad_norm 4.9801 (3.4317) [2021-04-16 20:28:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][790/1251] eta 0:02:15 lr 0.000016 time 0.3009 (0.2949) loss 3.1052 (2.9609) grad_norm 3.9196 (3.4307) [2021-04-16 20:28:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][800/1251] eta 0:02:12 lr 0.000016 time 0.2745 (0.2948) loss 2.8213 (2.9570) grad_norm 3.8797 (3.4300) [2021-04-16 20:28:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][810/1251] eta 0:02:09 lr 0.000016 time 0.2913 (0.2946) loss 3.3512 (2.9609) grad_norm 3.3042 (3.4262) [2021-04-16 20:28:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][820/1251] eta 0:02:06 lr 0.000016 time 0.2865 (0.2944) loss 3.1428 (2.9647) grad_norm 3.0921 (3.4273) [2021-04-16 20:28:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][830/1251] eta 0:02:03 lr 0.000016 time 0.3074 (0.2944) loss 2.8642 (2.9645) grad_norm 3.7109 (3.4253) [2021-04-16 20:28:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][840/1251] eta 0:02:00 lr 0.000016 time 0.2642 (0.2942) loss 3.2563 (2.9635) grad_norm 3.8040 (3.4256) [2021-04-16 20:28:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][850/1251] eta 0:01:57 lr 0.000016 time 0.2960 (0.2941) loss 3.1366 (2.9618) grad_norm 2.8757 (3.4255) [2021-04-16 20:28:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][860/1251] eta 0:01:54 lr 0.000016 time 0.2892 (0.2940) loss 3.0283 (2.9620) grad_norm 3.5141 (3.4236) [2021-04-16 20:28:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][870/1251] eta 0:01:51 lr 0.000016 time 0.2863 (0.2939) loss 3.7867 (2.9631) grad_norm 3.7530 (3.4249) [2021-04-16 20:28:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][880/1251] eta 0:01:48 lr 0.000016 time 0.2814 (0.2937) loss 2.6747 (2.9627) grad_norm 3.4519 (3.4235) [2021-04-16 20:28:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][890/1251] eta 0:01:45 lr 0.000016 time 0.2650 (0.2935) loss 2.5865 (2.9604) grad_norm 2.9267 (3.4215) [2021-04-16 20:28:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][900/1251] eta 0:01:42 lr 0.000016 time 0.2612 (0.2933) loss 1.9736 (2.9608) grad_norm 3.4479 (3.4208) [2021-04-16 20:28:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][910/1251] eta 0:01:39 lr 0.000016 time 0.2779 (0.2931) loss 3.3541 (2.9626) grad_norm 3.4422 (3.4194) [2021-04-16 20:28:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][920/1251] eta 0:01:36 lr 0.000016 time 0.2822 (0.2930) loss 3.2660 (2.9633) grad_norm 3.6035 (3.4189) [2021-04-16 20:28:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][930/1251] eta 0:01:34 lr 0.000016 time 0.2714 (0.2929) loss 3.0477 (2.9652) grad_norm 3.6651 (3.4181) [2021-04-16 20:28:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][940/1251] eta 0:01:31 lr 0.000016 time 0.2561 (0.2931) loss 3.1353 (2.9669) grad_norm 4.5131 (3.4185) [2021-04-16 20:28:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][950/1251] eta 0:01:28 lr 0.000016 time 0.2768 (0.2929) loss 3.3031 (2.9658) grad_norm 3.1768 (3.4172) [2021-04-16 20:28:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][960/1251] eta 0:01:25 lr 0.000016 time 0.2969 (0.2927) loss 3.2100 (2.9679) grad_norm 3.3426 (3.4173) [2021-04-16 20:29:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][970/1251] eta 0:01:22 lr 0.000016 time 0.2805 (0.2925) loss 3.1633 (2.9666) grad_norm 3.3332 (3.4151) [2021-04-16 20:29:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][980/1251] eta 0:01:19 lr 0.000016 time 0.2743 (0.2924) loss 2.0145 (2.9654) grad_norm 2.8873 (3.4139) [2021-04-16 20:29:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][990/1251] eta 0:01:16 lr 0.000016 time 0.3021 (0.2922) loss 3.0143 (2.9643) grad_norm 2.8962 (3.4122) [2021-04-16 20:29:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][1000/1251] eta 0:01:13 lr 0.000016 time 0.2949 (0.2921) loss 2.5639 (2.9633) grad_norm 2.7721 (3.4111) [2021-04-16 20:29:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][1010/1251] eta 0:01:10 lr 0.000016 time 0.2964 (0.2920) loss 2.9084 (2.9636) grad_norm 3.1739 (3.4104) [2021-04-16 20:29:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][1020/1251] eta 0:01:07 lr 0.000016 time 0.2585 (0.2917) loss 3.0738 (2.9630) grad_norm 3.2289 (3.4099) [2021-04-16 20:29:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][1030/1251] eta 0:01:04 lr 0.000016 time 0.2704 (0.2917) loss 2.7264 (2.9643) grad_norm 3.0971 (3.4099) [2021-04-16 20:29:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][1040/1251] eta 0:01:01 lr 0.000016 time 0.2519 (0.2915) loss 3.1575 (2.9650) grad_norm 3.6305 (3.4083) [2021-04-16 20:29:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][1050/1251] eta 0:00:58 lr 0.000016 time 0.2627 (0.2914) loss 2.3463 (2.9630) grad_norm 3.9687 (3.4078) [2021-04-16 20:29:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][1060/1251] eta 0:00:55 lr 0.000016 time 0.2717 (0.2913) loss 3.2925 (2.9631) grad_norm 3.0769 (3.4092) [2021-04-16 20:29:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][1070/1251] eta 0:00:52 lr 0.000016 time 0.2509 (0.2913) loss 2.3986 (2.9637) grad_norm 3.1747 (3.4082) [2021-04-16 20:29:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][1080/1251] eta 0:00:49 lr 0.000016 time 0.2662 (0.2912) loss 3.2928 (2.9651) grad_norm 3.1126 (3.4084) [2021-04-16 20:29:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][1090/1251] eta 0:00:46 lr 0.000016 time 0.2629 (0.2910) loss 2.4193 (2.9673) grad_norm 7.5902 (3.4121) [2021-04-16 20:29:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][1100/1251] eta 0:00:43 lr 0.000016 time 0.2890 (0.2909) loss 2.6180 (2.9688) grad_norm 3.7422 (3.4103) [2021-04-16 20:29:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][1110/1251] eta 0:00:40 lr 0.000016 time 0.2438 (0.2907) loss 3.1640 (2.9682) grad_norm 3.2539 (3.4092) [2021-04-16 20:29:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][1120/1251] eta 0:00:38 lr 0.000016 time 0.2743 (0.2906) loss 3.0098 (2.9658) grad_norm 3.5241 (3.4099) [2021-04-16 20:29:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][1130/1251] eta 0:00:35 lr 0.000016 time 0.2796 (0.2905) loss 3.1175 (2.9656) grad_norm 3.1521 (3.4107) [2021-04-16 20:29:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][1140/1251] eta 0:00:32 lr 0.000016 time 0.2955 (0.2905) loss 3.4319 (2.9685) grad_norm 3.5416 (3.4107) [2021-04-16 20:29:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][1150/1251] eta 0:00:29 lr 0.000016 time 0.2701 (0.2906) loss 3.1932 (2.9693) grad_norm 3.0711 (3.4100) [2021-04-16 20:29:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][1160/1251] eta 0:00:26 lr 0.000016 time 0.2868 (0.2905) loss 3.2279 (2.9697) grad_norm 3.1235 (3.4082) [2021-04-16 20:29:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][1170/1251] eta 0:00:23 lr 0.000016 time 0.2872 (0.2905) loss 3.0360 (2.9726) grad_norm 3.6689 (3.4093) [2021-04-16 20:30:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][1180/1251] eta 0:00:20 lr 0.000016 time 0.2449 (0.2904) loss 3.2459 (2.9707) grad_norm 3.8105 (3.4096) [2021-04-16 20:30:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][1190/1251] eta 0:00:17 lr 0.000016 time 0.2659 (0.2903) loss 2.8075 (2.9714) grad_norm 3.2322 (3.4091) [2021-04-16 20:30:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][1200/1251] eta 0:00:14 lr 0.000016 time 0.2836 (0.2902) loss 3.3483 (2.9712) grad_norm 3.7381 (3.4082) [2021-04-16 20:30:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][1210/1251] eta 0:00:11 lr 0.000016 time 0.3132 (0.2901) loss 3.7582 (2.9738) grad_norm 3.2885 (3.4115) [2021-04-16 20:30:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][1220/1251] eta 0:00:08 lr 0.000016 time 0.2991 (0.2900) loss 2.4384 (2.9736) grad_norm 2.9370 (3.4172) [2021-04-16 20:30:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][1230/1251] eta 0:00:06 lr 0.000016 time 0.3104 (0.2901) loss 3.2207 (2.9738) grad_norm 3.2004 (3.4188) [2021-04-16 20:30:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][1240/1251] eta 0:00:03 lr 0.000016 time 0.2485 (0.2899) loss 3.1615 (2.9724) grad_norm 2.5600 (3.4173) [2021-04-16 20:30:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [284/300][1250/1251] eta 0:00:00 lr 0.000016 time 0.2485 (0.2896) loss 3.6504 (2.9726) grad_norm 3.3625 (3.4177) [2021-04-16 20:30:25 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 284 training takes 0:06:07 [2021-04-16 20:30:25 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_284.pth saving...... [2021-04-16 20:30:37 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_284.pth saved !!! [2021-04-16 20:30:38 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.240 (1.240) Loss 0.8093 (0.8093) Acc@1 81.348 (81.348) Acc@5 95.898 (95.898) [2021-04-16 20:30:40 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.115 (0.244) Loss 0.7940 (0.8162) Acc@1 81.348 (80.868) Acc@5 95.508 (95.517) [2021-04-16 20:30:42 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.157 (0.234) Loss 0.7638 (0.8159) Acc@1 82.520 (81.092) Acc@5 95.508 (95.461) [2021-04-16 20:30:45 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.074 (0.258) Loss 0.8888 (0.8113) Acc@1 79.199 (81.124) Acc@5 95.117 (95.549) [2021-04-16 20:30:46 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.076 (0.224) Loss 0.8464 (0.8145) Acc@1 80.078 (81.009) Acc@5 95.410 (95.524) [2021-04-16 20:31:09 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 81.082 Acc@5 95.478 [2021-04-16 20:31:09 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 81.1% [2021-04-16 20:31:09 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 81.13% [2021-04-16 20:31:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][0/1251] eta 1:59:28 lr 0.000016 time 5.7305 (5.7305) loss 3.1928 (3.1928) grad_norm 3.2237 (3.2237) [2021-04-16 20:31:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][10/1251] eta 0:15:52 lr 0.000016 time 0.2730 (0.7678) loss 2.0183 (2.9898) grad_norm 3.4369 (3.4541) [2021-04-16 20:31:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][20/1251] eta 0:10:56 lr 0.000016 time 0.2607 (0.5334) loss 3.2661 (3.0946) grad_norm 3.1519 (3.3953) [2021-04-16 20:31:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][30/1251] eta 0:09:10 lr 0.000016 time 0.2698 (0.4512) loss 2.0169 (3.0814) grad_norm 3.0288 (3.3179) [2021-04-16 20:31:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][40/1251] eta 0:08:14 lr 0.000016 time 0.2792 (0.4085) loss 3.3510 (3.0425) grad_norm 3.9579 (3.3605) [2021-04-16 20:31:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][50/1251] eta 0:07:40 lr 0.000016 time 0.2929 (0.3835) loss 3.1577 (3.0466) grad_norm 3.6132 (3.4057) [2021-04-16 20:31:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][60/1251] eta 0:07:16 lr 0.000016 time 0.3055 (0.3664) loss 3.0015 (3.0729) grad_norm 2.8778 (3.3862) [2021-04-16 20:31:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][70/1251] eta 0:06:57 lr 0.000016 time 0.2728 (0.3533) loss 2.9484 (3.0650) grad_norm 4.0611 (3.3873) [2021-04-16 20:31:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][80/1251] eta 0:06:43 lr 0.000016 time 0.2884 (0.3446) loss 2.6434 (3.0438) grad_norm 3.4760 (3.3604) [2021-04-16 20:31:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][90/1251] eta 0:06:30 lr 0.000016 time 0.2601 (0.3368) loss 2.7102 (3.0429) grad_norm 3.3133 (3.3512) [2021-04-16 20:31:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][100/1251] eta 0:06:20 lr 0.000016 time 0.2638 (0.3303) loss 3.0613 (3.0444) grad_norm 3.1541 (3.3383) [2021-04-16 20:31:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][110/1251] eta 0:06:11 lr 0.000016 time 0.2833 (0.3255) loss 2.9221 (3.0398) grad_norm 3.1580 (3.3183) [2021-04-16 20:31:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][120/1251] eta 0:06:03 lr 0.000016 time 0.2712 (0.3214) loss 3.3001 (3.0328) grad_norm 3.3549 (3.3188) [2021-04-16 20:31:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][130/1251] eta 0:05:56 lr 0.000016 time 0.2712 (0.3179) loss 2.7763 (3.0062) grad_norm 3.1470 (3.3084) [2021-04-16 20:31:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][140/1251] eta 0:05:52 lr 0.000016 time 0.2804 (0.3173) loss 2.8668 (2.9993) grad_norm 3.4879 (3.2982) [2021-04-16 20:31:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][150/1251] eta 0:05:46 lr 0.000016 time 0.2755 (0.3149) loss 2.7666 (3.0007) grad_norm 3.1237 (3.3106) [2021-04-16 20:32:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][160/1251] eta 0:05:42 lr 0.000016 time 0.3018 (0.3139) loss 2.2064 (2.9979) grad_norm 2.9815 (3.3142) [2021-04-16 20:32:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][170/1251] eta 0:05:37 lr 0.000016 time 0.3093 (0.3119) loss 2.7539 (3.0079) grad_norm 2.7990 (3.3075) [2021-04-16 20:32:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][180/1251] eta 0:05:32 lr 0.000016 time 0.2807 (0.3103) loss 3.8887 (3.0158) grad_norm 3.4414 (3.3113) [2021-04-16 20:32:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][190/1251] eta 0:05:27 lr 0.000016 time 0.2783 (0.3085) loss 3.1172 (3.0080) grad_norm 3.2160 (3.3081) [2021-04-16 20:32:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][200/1251] eta 0:05:22 lr 0.000016 time 0.2707 (0.3071) loss 3.0975 (3.0133) grad_norm 3.7439 (3.3115) [2021-04-16 20:32:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][210/1251] eta 0:05:18 lr 0.000016 time 0.2576 (0.3059) loss 3.2197 (3.0094) grad_norm 3.4951 (3.3118) [2021-04-16 20:32:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][220/1251] eta 0:05:14 lr 0.000016 time 0.2910 (0.3046) loss 1.9683 (2.9927) grad_norm 3.1211 (3.3015) [2021-04-16 20:32:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][230/1251] eta 0:05:09 lr 0.000016 time 0.2791 (0.3035) loss 2.7352 (2.9930) grad_norm 3.2564 (3.3097) [2021-04-16 20:32:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][240/1251] eta 0:05:06 lr 0.000016 time 0.2996 (0.3027) loss 3.7330 (2.9987) grad_norm 3.1558 (3.3127) [2021-04-16 20:32:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][250/1251] eta 0:05:01 lr 0.000016 time 0.2857 (0.3016) loss 3.0926 (2.9983) grad_norm 3.5932 (3.3146) [2021-04-16 20:32:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][260/1251] eta 0:04:57 lr 0.000016 time 0.2718 (0.3006) loss 3.4305 (2.9994) grad_norm 3.5370 (3.3146) [2021-04-16 20:32:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][270/1251] eta 0:04:54 lr 0.000016 time 0.2809 (0.2997) loss 2.8968 (2.9973) grad_norm 3.6148 (3.3126) [2021-04-16 20:32:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][280/1251] eta 0:04:50 lr 0.000016 time 0.2522 (0.2989) loss 3.6759 (3.0082) grad_norm 2.7818 (3.3203) [2021-04-16 20:32:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][290/1251] eta 0:04:46 lr 0.000016 time 0.2718 (0.2985) loss 3.0357 (2.9977) grad_norm 3.3047 (3.3207) [2021-04-16 20:32:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][300/1251] eta 0:04:43 lr 0.000016 time 0.2782 (0.2978) loss 2.0680 (2.9998) grad_norm 2.9681 (3.3206) [2021-04-16 20:32:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][310/1251] eta 0:04:39 lr 0.000016 time 0.2604 (0.2971) loss 3.5648 (3.0122) grad_norm 3.2702 (3.3235) [2021-04-16 20:32:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][320/1251] eta 0:04:35 lr 0.000016 time 0.2705 (0.2964) loss 3.2043 (3.0081) grad_norm 3.0762 (3.3278) [2021-04-16 20:32:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][330/1251] eta 0:04:32 lr 0.000016 time 0.2470 (0.2957) loss 3.1153 (2.9976) grad_norm 3.7023 (3.3334) [2021-04-16 20:32:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][340/1251] eta 0:04:28 lr 0.000016 time 0.2873 (0.2952) loss 3.3034 (2.9956) grad_norm 3.4809 (3.3321) [2021-04-16 20:32:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][350/1251] eta 0:04:25 lr 0.000016 time 0.2785 (0.2948) loss 2.7656 (2.9932) grad_norm 3.2957 (3.3299) [2021-04-16 20:32:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][360/1251] eta 0:04:22 lr 0.000016 time 0.2700 (0.2951) loss 3.0515 (2.9927) grad_norm 3.7103 (3.3327) [2021-04-16 20:32:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][370/1251] eta 0:04:19 lr 0.000016 time 0.3032 (0.2950) loss 3.2043 (2.9915) grad_norm 4.3136 (3.3340) [2021-04-16 20:33:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][380/1251] eta 0:04:16 lr 0.000016 time 0.2591 (0.2945) loss 2.8137 (2.9918) grad_norm 3.6682 (3.3322) [2021-04-16 20:33:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][390/1251] eta 0:04:13 lr 0.000016 time 0.2971 (0.2940) loss 3.3291 (2.9933) grad_norm 2.8877 (3.3388) [2021-04-16 20:33:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][400/1251] eta 0:04:09 lr 0.000016 time 0.3026 (0.2935) loss 2.9642 (2.9983) grad_norm 3.1030 (3.3462) [2021-04-16 20:33:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][410/1251] eta 0:04:06 lr 0.000016 time 0.2878 (0.2933) loss 3.3535 (2.9984) grad_norm 3.0745 (3.3484) [2021-04-16 20:33:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][420/1251] eta 0:04:03 lr 0.000016 time 0.2761 (0.2931) loss 1.8927 (2.9984) grad_norm 2.9261 (3.3470) [2021-04-16 20:33:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][430/1251] eta 0:04:00 lr 0.000016 time 0.2729 (0.2932) loss 3.2669 (3.0019) grad_norm 3.4105 (3.3468) [2021-04-16 20:33:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][440/1251] eta 0:03:57 lr 0.000016 time 0.2788 (0.2928) loss 2.6206 (3.0019) grad_norm 3.3455 (3.3458) [2021-04-16 20:33:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][450/1251] eta 0:03:54 lr 0.000016 time 0.2910 (0.2924) loss 2.7189 (2.9993) grad_norm 2.9379 (3.3470) [2021-04-16 20:33:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][460/1251] eta 0:03:51 lr 0.000016 time 0.2882 (0.2922) loss 2.5795 (2.9942) grad_norm 2.9486 (3.3479) [2021-04-16 20:33:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][470/1251] eta 0:03:48 lr 0.000016 time 0.2716 (0.2922) loss 2.9710 (2.9887) grad_norm 2.7405 (3.3418) [2021-04-16 20:33:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][480/1251] eta 0:03:45 lr 0.000016 time 0.2909 (0.2919) loss 2.2989 (2.9811) grad_norm 2.9851 (3.3417) [2021-04-16 20:33:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][490/1251] eta 0:03:41 lr 0.000016 time 0.2728 (0.2916) loss 2.4729 (2.9817) grad_norm 3.0092 (3.3433) [2021-04-16 20:33:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][500/1251] eta 0:03:38 lr 0.000016 time 0.2698 (0.2913) loss 2.9518 (2.9814) grad_norm 3.2861 (3.3420) [2021-04-16 20:33:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][510/1251] eta 0:03:35 lr 0.000016 time 0.2921 (0.2910) loss 3.0897 (2.9779) grad_norm 3.1136 (3.3448) [2021-04-16 20:33:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][520/1251] eta 0:03:32 lr 0.000016 time 0.2786 (0.2909) loss 2.2425 (2.9748) grad_norm 3.4330 (3.3495) [2021-04-16 20:33:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][530/1251] eta 0:03:29 lr 0.000016 time 0.2900 (0.2906) loss 3.0165 (2.9749) grad_norm 3.1915 (3.3474) [2021-04-16 20:33:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][540/1251] eta 0:03:26 lr 0.000016 time 0.2637 (0.2907) loss 2.5463 (2.9688) grad_norm 3.8110 (3.3518) [2021-04-16 20:33:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][550/1251] eta 0:03:23 lr 0.000016 time 0.2772 (0.2904) loss 2.6263 (2.9674) grad_norm 3.1916 (3.3548) [2021-04-16 20:33:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][560/1251] eta 0:03:20 lr 0.000016 time 0.2828 (0.2902) loss 3.2159 (2.9709) grad_norm 3.4542 (3.3541) [2021-04-16 20:33:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][570/1251] eta 0:03:17 lr 0.000016 time 0.2626 (0.2902) loss 3.4551 (2.9743) grad_norm 3.3027 (3.3543) [2021-04-16 20:33:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][580/1251] eta 0:03:14 lr 0.000016 time 0.2686 (0.2901) loss 2.5590 (2.9775) grad_norm 2.9950 (3.3554) [2021-04-16 20:34:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][590/1251] eta 0:03:11 lr 0.000016 time 0.2909 (0.2898) loss 3.2954 (2.9763) grad_norm 3.0419 (3.3566) [2021-04-16 20:34:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][600/1251] eta 0:03:08 lr 0.000016 time 0.2840 (0.2897) loss 3.3678 (2.9771) grad_norm 3.2204 (3.3555) [2021-04-16 20:34:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][610/1251] eta 0:03:05 lr 0.000016 time 0.2949 (0.2895) loss 2.8011 (2.9777) grad_norm 3.1299 (3.3529) [2021-04-16 20:34:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][620/1251] eta 0:03:02 lr 0.000016 time 0.2789 (0.2893) loss 3.0943 (2.9780) grad_norm 3.0259 (3.3522) [2021-04-16 20:34:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][630/1251] eta 0:02:59 lr 0.000016 time 0.2862 (0.2892) loss 3.1680 (2.9817) grad_norm 3.6577 (3.3536) [2021-04-16 20:34:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][640/1251] eta 0:02:56 lr 0.000016 time 0.2868 (0.2891) loss 3.0594 (2.9847) grad_norm 3.3490 (3.3569) [2021-04-16 20:34:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][650/1251] eta 0:02:53 lr 0.000016 time 0.2677 (0.2889) loss 3.1907 (2.9841) grad_norm 3.2007 (3.3557) [2021-04-16 20:34:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][660/1251] eta 0:02:50 lr 0.000016 time 0.2780 (0.2888) loss 3.4569 (2.9825) grad_norm 3.0004 (3.3565) [2021-04-16 20:34:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][670/1251] eta 0:02:47 lr 0.000016 time 0.2841 (0.2887) loss 3.3279 (2.9791) grad_norm 3.1853 (3.3561) [2021-04-16 20:34:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][680/1251] eta 0:02:44 lr 0.000016 time 0.3022 (0.2885) loss 2.3429 (2.9804) grad_norm 3.4051 (3.3552) [2021-04-16 20:34:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][690/1251] eta 0:02:41 lr 0.000016 time 0.2648 (0.2884) loss 3.5214 (2.9836) grad_norm 2.9554 (3.3528) [2021-04-16 20:34:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][700/1251] eta 0:02:38 lr 0.000016 time 0.2766 (0.2882) loss 2.8335 (2.9791) grad_norm 3.1531 (3.3492) [2021-04-16 20:34:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][710/1251] eta 0:02:35 lr 0.000016 time 0.2706 (0.2880) loss 3.2797 (2.9790) grad_norm 2.6303 (3.3517) [2021-04-16 20:34:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][720/1251] eta 0:02:33 lr 0.000016 time 0.2665 (0.2882) loss 3.4577 (2.9806) grad_norm 3.5957 (3.3542) [2021-04-16 20:34:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][730/1251] eta 0:02:30 lr 0.000016 time 0.3876 (0.2881) loss 2.4226 (2.9819) grad_norm 3.3437 (3.3554) [2021-04-16 20:34:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][740/1251] eta 0:02:27 lr 0.000016 time 0.2795 (0.2880) loss 2.8953 (2.9821) grad_norm 2.6615 (3.3558) [2021-04-16 20:34:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][750/1251] eta 0:02:24 lr 0.000016 time 0.3049 (0.2880) loss 3.2849 (2.9831) grad_norm 3.1012 (3.3554) [2021-04-16 20:34:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][760/1251] eta 0:02:21 lr 0.000016 time 0.2842 (0.2881) loss 2.7403 (2.9828) grad_norm 3.2505 (3.3618) [2021-04-16 20:34:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][770/1251] eta 0:02:18 lr 0.000016 time 0.2854 (0.2879) loss 2.3700 (2.9793) grad_norm 3.0483 (3.3621) [2021-04-16 20:34:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][780/1251] eta 0:02:15 lr 0.000016 time 0.3036 (0.2878) loss 3.4271 (2.9802) grad_norm 3.5426 (3.3608) [2021-04-16 20:34:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][790/1251] eta 0:02:12 lr 0.000016 time 0.2621 (0.2877) loss 2.7953 (2.9826) grad_norm 3.0251 (3.3580) [2021-04-16 20:35:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][800/1251] eta 0:02:09 lr 0.000016 time 0.2904 (0.2877) loss 3.0207 (2.9846) grad_norm 2.9146 (3.3619) [2021-04-16 20:35:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][810/1251] eta 0:02:06 lr 0.000016 time 0.2577 (0.2877) loss 3.5637 (2.9871) grad_norm 3.4889 (3.3595) [2021-04-16 20:35:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][820/1251] eta 0:02:03 lr 0.000016 time 0.3002 (0.2875) loss 3.1613 (2.9889) grad_norm 3.4707 (3.3606) [2021-04-16 20:35:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][830/1251] eta 0:02:01 lr 0.000016 time 0.2973 (0.2875) loss 3.4015 (2.9898) grad_norm 3.1908 (3.3607) [2021-04-16 20:35:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][840/1251] eta 0:01:58 lr 0.000016 time 0.2652 (0.2873) loss 3.9362 (2.9912) grad_norm 3.7463 (3.3624) [2021-04-16 20:35:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][850/1251] eta 0:01:55 lr 0.000016 time 0.2781 (0.2871) loss 2.6652 (2.9881) grad_norm 5.0687 (3.3661) [2021-04-16 20:35:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][860/1251] eta 0:01:52 lr 0.000016 time 0.2677 (0.2870) loss 3.6260 (2.9868) grad_norm 3.3983 (3.3687) [2021-04-16 20:35:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][870/1251] eta 0:01:49 lr 0.000016 time 0.2566 (0.2869) loss 3.2453 (2.9872) grad_norm 3.6668 (3.3683) [2021-04-16 20:35:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][880/1251] eta 0:01:46 lr 0.000016 time 0.2952 (0.2868) loss 3.0105 (2.9888) grad_norm 3.6254 (3.3703) [2021-04-16 20:35:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][890/1251] eta 0:01:43 lr 0.000016 time 0.2782 (0.2867) loss 3.1879 (2.9905) grad_norm 2.9314 (3.3680) [2021-04-16 20:35:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][900/1251] eta 0:01:40 lr 0.000016 time 0.2747 (0.2866) loss 2.8202 (2.9889) grad_norm 3.4614 (3.3665) [2021-04-16 20:35:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][910/1251] eta 0:01:37 lr 0.000016 time 0.2710 (0.2866) loss 1.8413 (2.9887) grad_norm 3.1104 (3.3674) [2021-04-16 20:35:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][920/1251] eta 0:01:34 lr 0.000016 time 0.2805 (0.2866) loss 2.4226 (2.9878) grad_norm 3.5509 (3.3643) [2021-04-16 20:35:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][930/1251] eta 0:01:31 lr 0.000016 time 0.2770 (0.2866) loss 2.8319 (2.9869) grad_norm 4.5959 (3.3634) [2021-04-16 20:35:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][940/1251] eta 0:01:29 lr 0.000016 time 0.2731 (0.2867) loss 2.6289 (2.9872) grad_norm 3.9570 (3.3632) [2021-04-16 20:35:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][950/1251] eta 0:01:26 lr 0.000015 time 0.2760 (0.2867) loss 2.8326 (2.9887) grad_norm 3.2309 (3.3621) [2021-04-16 20:35:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][960/1251] eta 0:01:23 lr 0.000015 time 0.2764 (0.2867) loss 2.4919 (2.9880) grad_norm 3.0239 (3.3622) [2021-04-16 20:35:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][970/1251] eta 0:01:20 lr 0.000015 time 0.3082 (0.2866) loss 2.7277 (2.9847) grad_norm 3.9281 (3.3634) [2021-04-16 20:35:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][980/1251] eta 0:01:17 lr 0.000015 time 0.2802 (0.2864) loss 1.9801 (2.9823) grad_norm 3.2988 (3.3646) [2021-04-16 20:35:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][990/1251] eta 0:01:14 lr 0.000015 time 0.2769 (0.2864) loss 2.7527 (2.9829) grad_norm 3.6930 (3.3661) [2021-04-16 20:35:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][1000/1251] eta 0:01:11 lr 0.000015 time 0.2874 (0.2864) loss 1.7548 (2.9814) grad_norm 3.0216 (3.3707) [2021-04-16 20:35:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][1010/1251] eta 0:01:09 lr 0.000015 time 0.3062 (0.2864) loss 2.9146 (2.9827) grad_norm 3.5471 (3.3697) [2021-04-16 20:36:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][1020/1251] eta 0:01:06 lr 0.000015 time 0.2961 (0.2863) loss 2.8116 (2.9815) grad_norm 4.7721 (3.3711) [2021-04-16 20:36:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][1030/1251] eta 0:01:03 lr 0.000015 time 0.2921 (0.2863) loss 3.6000 (2.9844) grad_norm 3.0605 (3.3692) [2021-04-16 20:36:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][1040/1251] eta 0:01:00 lr 0.000015 time 0.2553 (0.2862) loss 1.7908 (2.9834) grad_norm 3.4215 (3.3695) [2021-04-16 20:36:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][1050/1251] eta 0:00:57 lr 0.000015 time 0.2812 (0.2862) loss 3.6459 (2.9845) grad_norm 4.1373 (3.3707) [2021-04-16 20:36:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][1060/1251] eta 0:00:54 lr 0.000015 time 0.2914 (0.2861) loss 3.1142 (2.9858) grad_norm 3.5749 (3.3710) [2021-04-16 20:36:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][1070/1251] eta 0:00:51 lr 0.000015 time 0.2893 (0.2860) loss 3.8860 (2.9872) grad_norm 3.4621 (3.3739) [2021-04-16 20:36:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][1080/1251] eta 0:00:48 lr 0.000015 time 0.2645 (0.2860) loss 3.2780 (2.9865) grad_norm 3.2750 (3.3756) [2021-04-16 20:36:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][1090/1251] eta 0:00:46 lr 0.000015 time 0.2722 (0.2859) loss 2.7603 (2.9858) grad_norm 2.8630 (3.3763) [2021-04-16 20:36:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][1100/1251] eta 0:00:43 lr 0.000015 time 0.2969 (0.2859) loss 2.8285 (2.9868) grad_norm 3.4711 (3.3770) [2021-04-16 20:36:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][1110/1251] eta 0:00:40 lr 0.000015 time 0.2798 (0.2859) loss 2.7361 (2.9890) grad_norm 2.9311 (3.3785) [2021-04-16 20:36:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][1120/1251] eta 0:00:37 lr 0.000015 time 0.2865 (0.2859) loss 2.8355 (2.9901) grad_norm 3.6810 (3.3775) [2021-04-16 20:36:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][1130/1251] eta 0:00:34 lr 0.000015 time 0.3019 (0.2858) loss 3.0255 (2.9894) grad_norm 2.9973 (3.3755) [2021-04-16 20:36:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][1140/1251] eta 0:00:31 lr 0.000015 time 0.2838 (0.2859) loss 2.5277 (2.9897) grad_norm 3.2715 (3.3774) [2021-04-16 20:36:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][1150/1251] eta 0:00:28 lr 0.000015 time 0.2626 (0.2859) loss 2.2822 (2.9888) grad_norm 3.6353 (3.3783) [2021-04-16 20:36:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][1160/1251] eta 0:00:26 lr 0.000015 time 0.2850 (0.2858) loss 3.2601 (2.9910) grad_norm 3.3116 (3.3811) [2021-04-16 20:36:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][1170/1251] eta 0:00:23 lr 0.000015 time 0.2436 (0.2857) loss 3.7326 (2.9898) grad_norm 3.5526 (3.3809) [2021-04-16 20:36:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][1180/1251] eta 0:00:20 lr 0.000015 time 0.2770 (0.2856) loss 2.2759 (2.9882) grad_norm 4.6115 (3.3830) [2021-04-16 20:36:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][1190/1251] eta 0:00:17 lr 0.000015 time 0.2696 (0.2855) loss 1.7282 (2.9880) grad_norm 2.8852 (3.3815) [2021-04-16 20:36:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][1200/1251] eta 0:00:14 lr 0.000015 time 0.2618 (0.2855) loss 3.0222 (2.9895) grad_norm 2.8720 (3.3809) [2021-04-16 20:36:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][1210/1251] eta 0:00:11 lr 0.000015 time 0.2740 (0.2853) loss 2.8504 (2.9871) grad_norm 3.5664 (3.3807) [2021-04-16 20:36:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][1220/1251] eta 0:00:08 lr 0.000015 time 0.3061 (0.2853) loss 2.8324 (2.9870) grad_norm 3.2677 (3.3807) [2021-04-16 20:37:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][1230/1251] eta 0:00:05 lr 0.000015 time 0.2501 (0.2852) loss 2.7356 (2.9870) grad_norm 3.0757 (3.3800) [2021-04-16 20:37:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][1240/1251] eta 0:00:03 lr 0.000015 time 0.2480 (0.2851) loss 3.3930 (2.9875) grad_norm 2.6378 (3.3803) [2021-04-16 20:37:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [285/300][1250/1251] eta 0:00:00 lr 0.000015 time 0.2484 (0.2848) loss 2.5935 (2.9893) grad_norm 2.9703 (3.3797) [2021-04-16 20:37:21 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 285 training takes 0:06:11 [2021-04-16 20:37:21 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_285.pth saving...... [2021-04-16 20:37:42 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_285.pth saved !!! [2021-04-16 20:37:43 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.073 (1.073) Loss 0.8816 (0.8816) Acc@1 80.078 (80.078) Acc@5 94.629 (94.629) [2021-04-16 20:37:44 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.103 (0.236) Loss 0.8196 (0.8283) Acc@1 80.371 (80.877) Acc@5 95.898 (95.392) [2021-04-16 20:37:46 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.137 (0.225) Loss 0.7702 (0.8102) Acc@1 81.543 (81.222) Acc@5 95.703 (95.564) [2021-04-16 20:37:49 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.109 (0.229) Loss 0.7309 (0.8111) Acc@1 83.105 (81.237) Acc@5 96.484 (95.555) [2021-04-16 20:37:51 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.248 (0.223) Loss 0.8041 (0.8131) Acc@1 81.348 (81.167) Acc@5 95.801 (95.513) [2021-04-16 20:38:18 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 81.082 Acc@5 95.490 [2021-04-16 20:38:18 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 81.1% [2021-04-16 20:38:18 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 81.13% [2021-04-16 20:38:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][0/1251] eta 0:29:39 lr 0.000015 time 1.4225 (1.4225) loss 3.2067 (3.2067) grad_norm 4.0621 (4.0621) [2021-04-16 20:38:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][10/1251] eta 0:08:00 lr 0.000015 time 0.2756 (0.3868) loss 3.2963 (3.0785) grad_norm 3.3924 (4.5061) [2021-04-16 20:38:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][20/1251] eta 0:06:53 lr 0.000015 time 0.2700 (0.3355) loss 2.8584 (3.0487) grad_norm 3.9369 (4.0403) [2021-04-16 20:38:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][30/1251] eta 0:06:27 lr 0.000015 time 0.2831 (0.3171) loss 2.8805 (2.9868) grad_norm 3.2093 (3.8592) [2021-04-16 20:38:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][40/1251] eta 0:06:11 lr 0.000015 time 0.2594 (0.3071) loss 3.2278 (2.9854) grad_norm 3.5321 (3.7393) [2021-04-16 20:38:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][50/1251] eta 0:06:01 lr 0.000015 time 0.2828 (0.3011) loss 3.5105 (2.9588) grad_norm 3.2773 (3.7249) [2021-04-16 20:38:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][60/1251] eta 0:05:53 lr 0.000015 time 0.2762 (0.2968) loss 3.4086 (2.9377) grad_norm 2.9463 (3.6544) [2021-04-16 20:38:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][70/1251] eta 0:05:46 lr 0.000015 time 0.2685 (0.2935) loss 2.9157 (2.9350) grad_norm 4.3707 (3.6376) [2021-04-16 20:38:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][80/1251] eta 0:05:42 lr 0.000015 time 0.2708 (0.2927) loss 3.4775 (2.9509) grad_norm 3.4059 (3.6091) [2021-04-16 20:38:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][90/1251] eta 0:05:37 lr 0.000015 time 0.2504 (0.2908) loss 3.2604 (2.9541) grad_norm 3.5549 (3.5782) [2021-04-16 20:38:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][100/1251] eta 0:05:32 lr 0.000015 time 0.2670 (0.2888) loss 3.6089 (2.9740) grad_norm 3.0884 (3.5457) [2021-04-16 20:38:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][110/1251] eta 0:05:28 lr 0.000015 time 0.2648 (0.2876) loss 3.6264 (2.9627) grad_norm 3.5294 (3.5282) [2021-04-16 20:38:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][120/1251] eta 0:05:23 lr 0.000015 time 0.2553 (0.2863) loss 2.9278 (2.9725) grad_norm 3.2931 (3.5142) [2021-04-16 20:38:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][130/1251] eta 0:05:20 lr 0.000015 time 0.2696 (0.2855) loss 3.1263 (2.9905) grad_norm 3.1599 (3.5013) [2021-04-16 20:38:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][140/1251] eta 0:05:16 lr 0.000015 time 0.2681 (0.2852) loss 3.2692 (3.0052) grad_norm 3.7798 (3.5158) [2021-04-16 20:39:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][150/1251] eta 0:05:14 lr 0.000015 time 0.2713 (0.2861) loss 2.9089 (3.0086) grad_norm 3.3360 (3.4970) [2021-04-16 20:39:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][160/1251] eta 0:05:12 lr 0.000015 time 0.2625 (0.2865) loss 3.4541 (3.0209) grad_norm 3.0707 (3.4921) [2021-04-16 20:39:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][170/1251] eta 0:05:08 lr 0.000015 time 0.2783 (0.2857) loss 3.0134 (3.0079) grad_norm 3.3162 (3.4754) [2021-04-16 20:39:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][180/1251] eta 0:05:05 lr 0.000015 time 0.2791 (0.2853) loss 2.4708 (3.0016) grad_norm 3.6045 (3.4835) [2021-04-16 20:39:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][190/1251] eta 0:05:02 lr 0.000015 time 0.2935 (0.2851) loss 3.3765 (2.9910) grad_norm 3.0204 (3.4764) [2021-04-16 20:39:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][200/1251] eta 0:04:58 lr 0.000015 time 0.2694 (0.2844) loss 3.0828 (2.9937) grad_norm 3.6196 (3.4725) [2021-04-16 20:39:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][210/1251] eta 0:04:55 lr 0.000015 time 0.2513 (0.2839) loss 2.8270 (2.9968) grad_norm 4.0424 (3.4765) [2021-04-16 20:39:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][220/1251] eta 0:04:53 lr 0.000015 time 0.2548 (0.2843) loss 2.7088 (2.9937) grad_norm 5.0687 (3.4778) [2021-04-16 20:39:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][230/1251] eta 0:04:50 lr 0.000015 time 0.3179 (0.2841) loss 3.2039 (2.9967) grad_norm 3.2634 (3.4869) [2021-04-16 20:39:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][240/1251] eta 0:04:46 lr 0.000015 time 0.2697 (0.2837) loss 3.3027 (2.9993) grad_norm 3.5908 (3.4855) [2021-04-16 20:39:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][250/1251] eta 0:04:44 lr 0.000015 time 0.2775 (0.2838) loss 3.2203 (2.9987) grad_norm 3.9200 (3.4829) [2021-04-16 20:39:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][260/1251] eta 0:04:40 lr 0.000015 time 0.2851 (0.2835) loss 3.0884 (2.9897) grad_norm 2.9563 (3.4791) [2021-04-16 20:39:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][270/1251] eta 0:04:37 lr 0.000015 time 0.2824 (0.2833) loss 3.1537 (2.9882) grad_norm 2.7187 (3.4743) [2021-04-16 20:39:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][280/1251] eta 0:04:34 lr 0.000015 time 0.2556 (0.2831) loss 3.3463 (2.9808) grad_norm 2.8402 (3.4726) [2021-04-16 20:39:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][290/1251] eta 0:04:31 lr 0.000015 time 0.2869 (0.2830) loss 3.0030 (2.9780) grad_norm 2.9472 (3.4688) [2021-04-16 20:39:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][300/1251] eta 0:04:28 lr 0.000015 time 0.2657 (0.2828) loss 3.0204 (2.9798) grad_norm 2.9482 (3.4670) [2021-04-16 20:39:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][310/1251] eta 0:04:25 lr 0.000015 time 0.2846 (0.2826) loss 1.8810 (2.9745) grad_norm 2.8679 (3.4614) [2021-04-16 20:39:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][320/1251] eta 0:04:23 lr 0.000015 time 0.2773 (0.2832) loss 2.0077 (2.9794) grad_norm 3.5063 (3.4557) [2021-04-16 20:39:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][330/1251] eta 0:04:20 lr 0.000015 time 0.2722 (0.2829) loss 2.6990 (2.9810) grad_norm 3.1330 (3.4501) [2021-04-16 20:39:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][340/1251] eta 0:04:17 lr 0.000015 time 0.2731 (0.2827) loss 3.2018 (2.9837) grad_norm 2.8971 (3.4513) [2021-04-16 20:39:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][350/1251] eta 0:04:14 lr 0.000015 time 0.2650 (0.2828) loss 2.1104 (2.9847) grad_norm 3.0112 (3.4510) [2021-04-16 20:40:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][360/1251] eta 0:04:12 lr 0.000015 time 0.3053 (0.2831) loss 3.4479 (2.9904) grad_norm 3.2007 (3.4502) [2021-04-16 20:40:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][370/1251] eta 0:04:09 lr 0.000015 time 0.2613 (0.2828) loss 1.7898 (2.9879) grad_norm 3.7160 (3.4489) [2021-04-16 20:40:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][380/1251] eta 0:04:06 lr 0.000015 time 0.2759 (0.2827) loss 2.0776 (2.9841) grad_norm 3.1561 (3.4483) [2021-04-16 20:40:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][390/1251] eta 0:04:03 lr 0.000015 time 0.2676 (0.2826) loss 3.5772 (2.9804) grad_norm 3.6701 (3.4593) [2021-04-16 20:40:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][400/1251] eta 0:04:00 lr 0.000015 time 0.2654 (0.2825) loss 2.3881 (2.9834) grad_norm 3.3065 (3.4624) [2021-04-16 20:40:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][410/1251] eta 0:03:57 lr 0.000015 time 0.2649 (0.2824) loss 2.7832 (2.9850) grad_norm 4.5861 (3.4648) [2021-04-16 20:40:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][420/1251] eta 0:03:54 lr 0.000015 time 0.2899 (0.2824) loss 3.2278 (2.9841) grad_norm 3.3513 (3.4568) [2021-04-16 20:40:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][430/1251] eta 0:03:51 lr 0.000015 time 0.2695 (0.2821) loss 3.0363 (2.9884) grad_norm 3.0979 (3.4569) [2021-04-16 20:40:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][440/1251] eta 0:03:48 lr 0.000015 time 0.2871 (0.2820) loss 3.0830 (2.9902) grad_norm 2.8167 (3.4596) [2021-04-16 20:40:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][450/1251] eta 0:03:45 lr 0.000015 time 0.2883 (0.2819) loss 3.1016 (2.9887) grad_norm 3.8137 (3.4599) [2021-04-16 20:40:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][460/1251] eta 0:03:42 lr 0.000015 time 0.2746 (0.2817) loss 3.1037 (2.9869) grad_norm 3.5278 (3.4585) [2021-04-16 20:40:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][470/1251] eta 0:03:39 lr 0.000015 time 0.2655 (0.2815) loss 2.1321 (2.9836) grad_norm 3.6999 (3.4573) [2021-04-16 20:40:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][480/1251] eta 0:03:37 lr 0.000015 time 0.2731 (0.2815) loss 3.0759 (2.9818) grad_norm 3.6950 (3.4561) [2021-04-16 20:40:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][490/1251] eta 0:03:34 lr 0.000015 time 0.2872 (0.2813) loss 3.1230 (2.9766) grad_norm 3.6777 (nan) [2021-04-16 20:40:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][500/1251] eta 0:03:31 lr 0.000015 time 0.2762 (0.2813) loss 3.7418 (2.9745) grad_norm 3.6884 (nan) [2021-04-16 20:40:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][510/1251] eta 0:03:28 lr 0.000015 time 0.2618 (0.2813) loss 3.5366 (2.9761) grad_norm 2.8542 (nan) [2021-04-16 20:40:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][520/1251] eta 0:03:25 lr 0.000015 time 0.2868 (0.2813) loss 3.7782 (2.9792) grad_norm 3.4411 (nan) [2021-04-16 20:40:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][530/1251] eta 0:03:22 lr 0.000015 time 0.2857 (0.2813) loss 3.3793 (2.9821) grad_norm 3.1402 (nan) [2021-04-16 20:40:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][540/1251] eta 0:03:19 lr 0.000015 time 0.2785 (0.2812) loss 3.2319 (2.9803) grad_norm 3.5727 (nan) [2021-04-16 20:40:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][550/1251] eta 0:03:17 lr 0.000015 time 0.2601 (0.2811) loss 2.6886 (2.9761) grad_norm 3.4260 (nan) [2021-04-16 20:40:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][560/1251] eta 0:03:14 lr 0.000015 time 0.2793 (0.2810) loss 3.2441 (2.9812) grad_norm 3.9393 (nan) [2021-04-16 20:40:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][570/1251] eta 0:03:11 lr 0.000015 time 0.2725 (0.2811) loss 3.2807 (2.9840) grad_norm 3.3815 (nan) [2021-04-16 20:41:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][580/1251] eta 0:03:09 lr 0.000015 time 0.2806 (0.2818) loss 3.3177 (2.9850) grad_norm 2.9999 (nan) [2021-04-16 20:41:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][590/1251] eta 0:03:06 lr 0.000015 time 0.2767 (0.2817) loss 1.8511 (2.9874) grad_norm 3.5682 (nan) [2021-04-16 20:41:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][600/1251] eta 0:03:03 lr 0.000015 time 0.2727 (0.2816) loss 3.1822 (2.9876) grad_norm 3.4414 (nan) [2021-04-16 20:41:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][610/1251] eta 0:03:00 lr 0.000015 time 0.3115 (0.2816) loss 2.0919 (2.9866) grad_norm 3.2978 (nan) [2021-04-16 20:41:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][620/1251] eta 0:02:57 lr 0.000015 time 0.2800 (0.2814) loss 3.1908 (2.9842) grad_norm 3.1213 (nan) [2021-04-16 20:41:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][630/1251] eta 0:02:54 lr 0.000015 time 0.2849 (0.2814) loss 3.5863 (2.9848) grad_norm 3.7052 (nan) [2021-04-16 20:41:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][640/1251] eta 0:02:51 lr 0.000015 time 0.2895 (0.2814) loss 2.4792 (2.9859) grad_norm 3.2289 (nan) [2021-04-16 20:41:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][650/1251] eta 0:02:49 lr 0.000015 time 0.2561 (0.2816) loss 3.3636 (2.9878) grad_norm 4.2557 (nan) [2021-04-16 20:41:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][660/1251] eta 0:02:46 lr 0.000015 time 0.2706 (0.2815) loss 3.6575 (2.9879) grad_norm 3.3461 (nan) [2021-04-16 20:41:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][670/1251] eta 0:02:43 lr 0.000015 time 0.2610 (0.2814) loss 2.7309 (2.9911) grad_norm 4.3285 (nan) [2021-04-16 20:41:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][680/1251] eta 0:02:40 lr 0.000015 time 0.2893 (0.2812) loss 3.4137 (2.9930) grad_norm 3.3986 (nan) [2021-04-16 20:41:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][690/1251] eta 0:02:37 lr 0.000015 time 0.3001 (0.2812) loss 2.7449 (2.9951) grad_norm 4.3878 (nan) [2021-04-16 20:41:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][700/1251] eta 0:02:34 lr 0.000015 time 0.2824 (0.2811) loss 2.2551 (2.9948) grad_norm 3.1362 (nan) [2021-04-16 20:41:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][710/1251] eta 0:02:31 lr 0.000015 time 0.2820 (0.2809) loss 3.0595 (2.9974) grad_norm 3.1169 (nan) [2021-04-16 20:41:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][720/1251] eta 0:02:29 lr 0.000015 time 0.2701 (0.2810) loss 3.4070 (2.9952) grad_norm 3.4157 (nan) [2021-04-16 20:41:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][730/1251] eta 0:02:26 lr 0.000015 time 0.2789 (0.2810) loss 2.9399 (2.9953) grad_norm 3.6829 (nan) [2021-04-16 20:41:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][740/1251] eta 0:02:23 lr 0.000015 time 0.2616 (0.2811) loss 2.7796 (2.9916) grad_norm 3.2577 (nan) [2021-04-16 20:41:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][750/1251] eta 0:02:20 lr 0.000015 time 0.2746 (0.2811) loss 2.5623 (2.9883) grad_norm 3.3614 (nan) [2021-04-16 20:41:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][760/1251] eta 0:02:18 lr 0.000015 time 0.2960 (0.2811) loss 2.8568 (2.9881) grad_norm 3.3219 (nan) [2021-04-16 20:41:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][770/1251] eta 0:02:15 lr 0.000015 time 0.2549 (0.2811) loss 3.3452 (2.9875) grad_norm 3.4317 (nan) [2021-04-16 20:41:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][780/1251] eta 0:02:12 lr 0.000015 time 0.2751 (0.2810) loss 3.2661 (2.9870) grad_norm 2.9761 (nan) [2021-04-16 20:42:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][790/1251] eta 0:02:09 lr 0.000015 time 0.2612 (0.2811) loss 2.9679 (2.9868) grad_norm 3.3509 (nan) [2021-04-16 20:42:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][800/1251] eta 0:02:06 lr 0.000015 time 0.2476 (0.2810) loss 2.5767 (2.9854) grad_norm 3.5202 (nan) [2021-04-16 20:42:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][810/1251] eta 0:02:03 lr 0.000015 time 0.2619 (0.2810) loss 2.5356 (2.9833) grad_norm 3.2641 (nan) [2021-04-16 20:42:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][820/1251] eta 0:02:01 lr 0.000015 time 0.2569 (0.2809) loss 3.3219 (2.9876) grad_norm 3.1346 (nan) [2021-04-16 20:42:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][830/1251] eta 0:01:58 lr 0.000015 time 0.2497 (0.2809) loss 2.3812 (2.9876) grad_norm 3.7587 (nan) [2021-04-16 20:42:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][840/1251] eta 0:01:55 lr 0.000015 time 0.2915 (0.2809) loss 3.1105 (2.9862) grad_norm 3.4283 (nan) [2021-04-16 20:42:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][850/1251] eta 0:01:52 lr 0.000015 time 0.2734 (0.2808) loss 2.6770 (2.9864) grad_norm 3.2666 (nan) [2021-04-16 20:42:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][860/1251] eta 0:01:49 lr 0.000015 time 0.2717 (0.2808) loss 2.3239 (2.9851) grad_norm 3.4545 (nan) [2021-04-16 20:42:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][870/1251] eta 0:01:46 lr 0.000015 time 0.2901 (0.2808) loss 2.5739 (2.9815) grad_norm 2.9882 (nan) [2021-04-16 20:42:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][880/1251] eta 0:01:44 lr 0.000015 time 0.3012 (0.2807) loss 2.4781 (2.9793) grad_norm 3.6703 (nan) [2021-04-16 20:42:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][890/1251] eta 0:01:41 lr 0.000015 time 0.2700 (0.2807) loss 2.9741 (2.9801) grad_norm 2.7561 (nan) [2021-04-16 20:42:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][900/1251] eta 0:01:38 lr 0.000015 time 0.2535 (0.2807) loss 3.6592 (2.9830) grad_norm 3.0269 (nan) [2021-04-16 20:42:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][910/1251] eta 0:01:35 lr 0.000015 time 0.2576 (0.2806) loss 2.1502 (2.9811) grad_norm 2.9833 (nan) [2021-04-16 20:42:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][920/1251] eta 0:01:32 lr 0.000015 time 0.2711 (0.2806) loss 3.1744 (2.9808) grad_norm 3.1404 (nan) [2021-04-16 20:42:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][930/1251] eta 0:01:30 lr 0.000015 time 0.2642 (0.2807) loss 3.0350 (2.9773) grad_norm 2.8592 (nan) [2021-04-16 20:42:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][940/1251] eta 0:01:27 lr 0.000015 time 0.2602 (0.2808) loss 2.9350 (2.9771) grad_norm 2.9066 (nan) [2021-04-16 20:42:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][950/1251] eta 0:01:24 lr 0.000015 time 0.2802 (0.2807) loss 3.4108 (2.9788) grad_norm 3.3072 (nan) [2021-04-16 20:42:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][960/1251] eta 0:01:21 lr 0.000015 time 0.2696 (0.2807) loss 2.0808 (2.9759) grad_norm 2.8772 (nan) [2021-04-16 20:42:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][970/1251] eta 0:01:18 lr 0.000015 time 0.2462 (0.2806) loss 3.3531 (2.9765) grad_norm 3.9209 (nan) [2021-04-16 20:42:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][980/1251] eta 0:01:16 lr 0.000015 time 0.2630 (0.2806) loss 2.9024 (2.9765) grad_norm 3.4716 (nan) [2021-04-16 20:42:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][990/1251] eta 0:01:13 lr 0.000015 time 0.2972 (0.2805) loss 2.5756 (2.9742) grad_norm 2.7668 (nan) [2021-04-16 20:42:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][1000/1251] eta 0:01:10 lr 0.000015 time 0.2833 (0.2804) loss 2.6447 (2.9705) grad_norm 3.0943 (nan) [2021-04-16 20:43:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][1010/1251] eta 0:01:07 lr 0.000015 time 0.3025 (0.2804) loss 2.9739 (2.9704) grad_norm 3.1751 (nan) [2021-04-16 20:43:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][1020/1251] eta 0:01:04 lr 0.000015 time 0.2957 (0.2803) loss 3.2974 (2.9695) grad_norm 3.0204 (nan) [2021-04-16 20:43:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][1030/1251] eta 0:01:01 lr 0.000015 time 0.2878 (0.2803) loss 2.7233 (2.9680) grad_norm 3.3096 (nan) [2021-04-16 20:43:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][1040/1251] eta 0:00:59 lr 0.000015 time 0.2713 (0.2803) loss 3.1229 (2.9672) grad_norm 3.3417 (nan) [2021-04-16 20:43:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][1050/1251] eta 0:00:56 lr 0.000015 time 0.2969 (0.2804) loss 3.2804 (2.9678) grad_norm 3.5752 (nan) [2021-04-16 20:43:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][1060/1251] eta 0:00:53 lr 0.000015 time 0.2733 (0.2803) loss 2.0007 (2.9673) grad_norm 2.8811 (nan) [2021-04-16 20:43:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][1070/1251] eta 0:00:50 lr 0.000015 time 0.2579 (0.2803) loss 2.7625 (2.9664) grad_norm 3.3470 (nan) [2021-04-16 20:43:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][1080/1251] eta 0:00:47 lr 0.000015 time 0.2705 (0.2803) loss 3.4727 (2.9683) grad_norm 3.1059 (nan) [2021-04-16 20:43:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][1090/1251] eta 0:00:45 lr 0.000015 time 0.2784 (0.2802) loss 3.1803 (2.9684) grad_norm 2.9316 (nan) [2021-04-16 20:43:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][1100/1251] eta 0:00:42 lr 0.000015 time 0.2648 (0.2801) loss 3.4308 (2.9690) grad_norm 3.5860 (nan) [2021-04-16 20:43:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][1110/1251] eta 0:00:39 lr 0.000015 time 0.2642 (0.2801) loss 2.5864 (2.9687) grad_norm 3.7243 (nan) [2021-04-16 20:43:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][1120/1251] eta 0:00:36 lr 0.000015 time 0.2653 (0.2801) loss 2.8350 (2.9681) grad_norm 3.0982 (nan) [2021-04-16 20:43:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][1130/1251] eta 0:00:33 lr 0.000015 time 0.2730 (0.2800) loss 3.5526 (2.9690) grad_norm 3.8141 (nan) [2021-04-16 20:43:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][1140/1251] eta 0:00:31 lr 0.000015 time 0.2652 (0.2800) loss 3.7217 (2.9702) grad_norm 3.6855 (nan) [2021-04-16 20:43:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][1150/1251] eta 0:00:28 lr 0.000015 time 0.2519 (0.2801) loss 3.0307 (2.9682) grad_norm 3.6565 (nan) [2021-04-16 20:43:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][1160/1251] eta 0:00:25 lr 0.000015 time 0.2893 (0.2803) loss 2.7524 (2.9676) grad_norm 3.9798 (nan) [2021-04-16 20:43:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][1170/1251] eta 0:00:22 lr 0.000015 time 0.2864 (0.2803) loss 3.3222 (2.9686) grad_norm 3.6491 (nan) [2021-04-16 20:43:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][1180/1251] eta 0:00:19 lr 0.000015 time 0.2806 (0.2803) loss 3.2318 (2.9660) grad_norm 3.5573 (nan) [2021-04-16 20:43:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][1190/1251] eta 0:00:17 lr 0.000015 time 0.2686 (0.2802) loss 3.5394 (2.9676) grad_norm 2.8750 (nan) [2021-04-16 20:43:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][1200/1251] eta 0:00:14 lr 0.000015 time 0.2815 (0.2802) loss 3.4068 (2.9688) grad_norm 3.2479 (nan) [2021-04-16 20:43:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][1210/1251] eta 0:00:11 lr 0.000015 time 0.2861 (0.2802) loss 2.5434 (2.9677) grad_norm 3.3859 (nan) [2021-04-16 20:44:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][1220/1251] eta 0:00:08 lr 0.000015 time 0.2619 (0.2801) loss 3.1487 (2.9682) grad_norm 2.9588 (nan) [2021-04-16 20:44:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][1230/1251] eta 0:00:05 lr 0.000015 time 0.2868 (0.2803) loss 3.0753 (2.9691) grad_norm 4.2316 (nan) [2021-04-16 20:44:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][1240/1251] eta 0:00:03 lr 0.000015 time 0.2679 (0.2802) loss 3.2928 (2.9662) grad_norm 2.9715 (nan) [2021-04-16 20:44:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [286/300][1250/1251] eta 0:00:00 lr 0.000015 time 0.2586 (0.2800) loss 3.0776 (2.9668) grad_norm 2.8916 (nan) [2021-04-16 20:44:15 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 286 training takes 0:05:56 [2021-04-16 20:44:15 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_286.pth saving...... [2021-04-16 20:44:33 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_286.pth saved !!! [2021-04-16 20:44:34 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.056 (1.056) Loss 0.8071 (0.8071) Acc@1 82.520 (82.520) Acc@5 95.215 (95.215) [2021-04-16 20:44:36 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.174 (0.274) Loss 0.9041 (0.8350) Acc@1 79.004 (80.593) Acc@5 93.750 (95.135) [2021-04-16 20:44:38 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.273 (0.234) Loss 0.8760 (0.8280) Acc@1 79.492 (80.748) Acc@5 94.434 (95.359) [2021-04-16 20:44:40 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.120 (0.218) Loss 0.7945 (0.8234) Acc@1 81.641 (80.771) Acc@5 95.801 (95.376) [2021-04-16 20:44:42 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.213) Loss 0.7813 (0.8154) Acc@1 81.836 (81.057) Acc@5 95.898 (95.448) [2021-04-16 20:45:02 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 81.136 Acc@5 95.450 [2021-04-16 20:45:02 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 81.1% [2021-04-16 20:45:02 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 81.14% [2021-04-16 20:45:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][0/1251] eta 1:40:47 lr 0.000015 time 4.8343 (4.8343) loss 3.0912 (3.0912) grad_norm 3.6028 (3.6028) [2021-04-16 20:45:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][10/1251] eta 0:14:32 lr 0.000015 time 0.4179 (0.7032) loss 3.2998 (3.1969) grad_norm 3.3217 (3.4670) [2021-04-16 20:45:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][20/1251] eta 0:10:15 lr 0.000015 time 0.2643 (0.5000) loss 2.7882 (3.1055) grad_norm 3.3190 (3.3626) [2021-04-16 20:45:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][30/1251] eta 0:08:43 lr 0.000015 time 0.2664 (0.4286) loss 3.2573 (2.9887) grad_norm 3.3447 (3.3599) [2021-04-16 20:45:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][40/1251] eta 0:07:57 lr 0.000015 time 0.2805 (0.3941) loss 3.2137 (3.0475) grad_norm 3.4717 (3.3299) [2021-04-16 20:45:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][50/1251] eta 0:07:28 lr 0.000015 time 0.2647 (0.3730) loss 2.2026 (2.9760) grad_norm 3.3732 (3.3370) [2021-04-16 20:45:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][60/1251] eta 0:07:08 lr 0.000015 time 0.2723 (0.3594) loss 1.9797 (2.9781) grad_norm 3.2818 (3.3421) [2021-04-16 20:45:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][70/1251] eta 0:06:50 lr 0.000015 time 0.2495 (0.3477) loss 3.3632 (2.9665) grad_norm 3.2861 (3.3261) [2021-04-16 20:45:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][80/1251] eta 0:06:37 lr 0.000015 time 0.2480 (0.3393) loss 3.0274 (2.9694) grad_norm 3.2819 (3.3576) [2021-04-16 20:45:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][90/1251] eta 0:06:30 lr 0.000015 time 0.2979 (0.3361) loss 2.8670 (2.9458) grad_norm 3.8430 (3.3606) [2021-04-16 20:45:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][100/1251] eta 0:06:20 lr 0.000015 time 0.2974 (0.3308) loss 3.3100 (2.9698) grad_norm 3.2305 (3.3529) [2021-04-16 20:45:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][110/1251] eta 0:06:12 lr 0.000015 time 0.2893 (0.3264) loss 3.0113 (2.9753) grad_norm 3.0515 (3.3383) [2021-04-16 20:45:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][120/1251] eta 0:06:04 lr 0.000015 time 0.2863 (0.3226) loss 3.0457 (2.9894) grad_norm 3.4234 (3.3382) [2021-04-16 20:45:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][130/1251] eta 0:05:59 lr 0.000015 time 0.2843 (0.3206) loss 3.2235 (3.0094) grad_norm 2.9922 (3.4016) [2021-04-16 20:45:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][140/1251] eta 0:05:53 lr 0.000015 time 0.2946 (0.3185) loss 3.1404 (2.9880) grad_norm 5.2070 (3.4324) [2021-04-16 20:45:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][150/1251] eta 0:05:47 lr 0.000014 time 0.2627 (0.3158) loss 3.4422 (2.9837) grad_norm 3.2365 (3.4237) [2021-04-16 20:45:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][160/1251] eta 0:05:41 lr 0.000014 time 0.2635 (0.3133) loss 3.0941 (2.9888) grad_norm 3.9966 (3.4089) [2021-04-16 20:45:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][170/1251] eta 0:05:36 lr 0.000014 time 0.2688 (0.3112) loss 2.7056 (2.9999) grad_norm 3.9042 (3.3981) [2021-04-16 20:45:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][180/1251] eta 0:05:31 lr 0.000014 time 0.2545 (0.3092) loss 2.4136 (2.9990) grad_norm 3.2065 (3.3868) [2021-04-16 20:46:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][190/1251] eta 0:05:26 lr 0.000014 time 0.2847 (0.3077) loss 3.0539 (2.9850) grad_norm 3.1396 (3.3786) [2021-04-16 20:46:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][200/1251] eta 0:05:21 lr 0.000014 time 0.2767 (0.3063) loss 2.9744 (2.9922) grad_norm 2.9373 (3.3758) [2021-04-16 20:46:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][210/1251] eta 0:05:17 lr 0.000014 time 0.2787 (0.3049) loss 2.7927 (3.0014) grad_norm 3.3458 (3.3782) [2021-04-16 20:46:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][220/1251] eta 0:05:13 lr 0.000014 time 0.2709 (0.3038) loss 3.3048 (3.0020) grad_norm 3.4801 (3.3849) [2021-04-16 20:46:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][230/1251] eta 0:05:09 lr 0.000014 time 0.2763 (0.3027) loss 3.7496 (3.0048) grad_norm 3.4856 (3.3876) [2021-04-16 20:46:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][240/1251] eta 0:05:04 lr 0.000014 time 0.2789 (0.3015) loss 3.2917 (3.0086) grad_norm 3.6086 (3.3913) [2021-04-16 20:46:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][250/1251] eta 0:05:00 lr 0.000014 time 0.2865 (0.3005) loss 2.2992 (3.0024) grad_norm 3.3180 (3.3951) [2021-04-16 20:46:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][260/1251] eta 0:04:56 lr 0.000014 time 0.2890 (0.2996) loss 3.2497 (3.0087) grad_norm 3.9059 (3.4020) [2021-04-16 20:46:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][270/1251] eta 0:04:53 lr 0.000014 time 0.2784 (0.2987) loss 3.0899 (3.0049) grad_norm 3.9045 (3.4074) [2021-04-16 20:46:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][280/1251] eta 0:04:49 lr 0.000014 time 0.3001 (0.2981) loss 3.3098 (3.0061) grad_norm 6.0786 (3.4285) [2021-04-16 20:46:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][290/1251] eta 0:04:45 lr 0.000014 time 0.2689 (0.2974) loss 3.0462 (3.0007) grad_norm 3.4107 (3.4427) [2021-04-16 20:46:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][300/1251] eta 0:04:42 lr 0.000014 time 0.2660 (0.2970) loss 3.6545 (2.9960) grad_norm 3.3674 (3.4383) [2021-04-16 20:46:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][310/1251] eta 0:04:38 lr 0.000014 time 0.2820 (0.2964) loss 3.3960 (2.9974) grad_norm 3.1011 (3.4312) [2021-04-16 20:46:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][320/1251] eta 0:04:35 lr 0.000014 time 0.2745 (0.2958) loss 3.3142 (2.9988) grad_norm 3.2863 (3.4352) [2021-04-16 20:46:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][330/1251] eta 0:04:31 lr 0.000014 time 0.2874 (0.2953) loss 3.6146 (3.0011) grad_norm 3.6960 (3.4445) [2021-04-16 20:46:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][340/1251] eta 0:04:28 lr 0.000014 time 0.2665 (0.2947) loss 3.4319 (3.0028) grad_norm 3.2855 (3.4437) [2021-04-16 20:46:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][350/1251] eta 0:04:25 lr 0.000014 time 0.2633 (0.2944) loss 3.3333 (3.0055) grad_norm 3.3155 (3.4496) [2021-04-16 20:46:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][360/1251] eta 0:04:22 lr 0.000014 time 0.2611 (0.2944) loss 3.2492 (3.0106) grad_norm 3.3174 (3.4531) [2021-04-16 20:46:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][370/1251] eta 0:04:19 lr 0.000014 time 0.2831 (0.2942) loss 3.7267 (3.0122) grad_norm 3.4155 (3.4621) [2021-04-16 20:46:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][380/1251] eta 0:04:15 lr 0.000014 time 0.2646 (0.2938) loss 3.1204 (3.0152) grad_norm nan (nan) [2021-04-16 20:46:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][390/1251] eta 0:04:12 lr 0.000014 time 0.2906 (0.2934) loss 3.5339 (3.0069) grad_norm 4.0787 (nan) [2021-04-16 20:47:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][400/1251] eta 0:04:09 lr 0.000014 time 0.2785 (0.2929) loss 2.8282 (3.0028) grad_norm 2.9609 (nan) [2021-04-16 20:47:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][410/1251] eta 0:04:06 lr 0.000014 time 0.3295 (0.2928) loss 2.3114 (3.0009) grad_norm 3.1072 (nan) [2021-04-16 20:47:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][420/1251] eta 0:04:03 lr 0.000014 time 0.2784 (0.2925) loss 2.5975 (3.0015) grad_norm 3.8316 (nan) [2021-04-16 20:47:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][430/1251] eta 0:04:00 lr 0.000014 time 0.2555 (0.2925) loss 2.8846 (2.9967) grad_norm 2.8497 (nan) [2021-04-16 20:47:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][440/1251] eta 0:03:56 lr 0.000014 time 0.2817 (0.2922) loss 1.9148 (2.9967) grad_norm 2.8452 (nan) [2021-04-16 20:47:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][450/1251] eta 0:03:53 lr 0.000014 time 0.2732 (0.2921) loss 3.1441 (3.0009) grad_norm 3.1630 (nan) [2021-04-16 20:47:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][460/1251] eta 0:03:50 lr 0.000014 time 0.2742 (0.2916) loss 3.2026 (2.9996) grad_norm 2.6903 (nan) [2021-04-16 20:47:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][470/1251] eta 0:03:47 lr 0.000014 time 0.2869 (0.2914) loss 2.2704 (2.9983) grad_norm 3.5453 (nan) [2021-04-16 20:47:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][480/1251] eta 0:03:44 lr 0.000014 time 0.2878 (0.2911) loss 2.9527 (2.9992) grad_norm 3.3605 (nan) [2021-04-16 20:47:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][490/1251] eta 0:03:41 lr 0.000014 time 0.2458 (0.2908) loss 3.2365 (3.0058) grad_norm 3.1283 (nan) [2021-04-16 20:47:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][500/1251] eta 0:03:38 lr 0.000014 time 0.2761 (0.2905) loss 2.7090 (3.0089) grad_norm 2.6790 (nan) [2021-04-16 20:47:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][510/1251] eta 0:03:35 lr 0.000014 time 0.2866 (0.2903) loss 2.6579 (3.0065) grad_norm 3.6802 (nan) [2021-04-16 20:47:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][520/1251] eta 0:03:32 lr 0.000014 time 0.2901 (0.2901) loss 3.0282 (3.0062) grad_norm 2.8793 (nan) [2021-04-16 20:47:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][530/1251] eta 0:03:29 lr 0.000014 time 0.2947 (0.2899) loss 2.4245 (3.0054) grad_norm 3.0662 (nan) [2021-04-16 20:47:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][540/1251] eta 0:03:25 lr 0.000014 time 0.2623 (0.2897) loss 3.6444 (3.0088) grad_norm 2.8949 (nan) [2021-04-16 20:47:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][550/1251] eta 0:03:22 lr 0.000014 time 0.2727 (0.2894) loss 2.6708 (3.0043) grad_norm 3.2434 (nan) [2021-04-16 20:47:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][560/1251] eta 0:03:19 lr 0.000014 time 0.2705 (0.2893) loss 3.0746 (3.0024) grad_norm 3.4534 (nan) [2021-04-16 20:47:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][570/1251] eta 0:03:17 lr 0.000014 time 0.2534 (0.2893) loss 3.1368 (3.0032) grad_norm 3.3869 (nan) [2021-04-16 20:47:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][580/1251] eta 0:03:14 lr 0.000014 time 0.2809 (0.2896) loss 2.9421 (3.0038) grad_norm 3.2242 (nan) [2021-04-16 20:47:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][590/1251] eta 0:03:11 lr 0.000014 time 0.2646 (0.2896) loss 3.1202 (3.0051) grad_norm 3.1572 (nan) [2021-04-16 20:47:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][600/1251] eta 0:03:08 lr 0.000014 time 0.2571 (0.2893) loss 1.8069 (3.0015) grad_norm 3.3648 (nan) [2021-04-16 20:47:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][610/1251] eta 0:03:05 lr 0.000014 time 0.2762 (0.2892) loss 3.1687 (3.0001) grad_norm 3.2961 (nan) [2021-04-16 20:48:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][620/1251] eta 0:03:02 lr 0.000014 time 0.2801 (0.2890) loss 2.8807 (2.9980) grad_norm 3.2868 (nan) [2021-04-16 20:48:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][630/1251] eta 0:02:59 lr 0.000014 time 0.2554 (0.2889) loss 4.0390 (2.9985) grad_norm 3.7052 (nan) [2021-04-16 20:48:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][640/1251] eta 0:02:56 lr 0.000014 time 0.2890 (0.2888) loss 2.8972 (2.9982) grad_norm 2.9569 (nan) [2021-04-16 20:48:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][650/1251] eta 0:02:53 lr 0.000014 time 0.2800 (0.2887) loss 2.2957 (2.9962) grad_norm 3.2182 (nan) [2021-04-16 20:48:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][660/1251] eta 0:02:50 lr 0.000014 time 0.2617 (0.2885) loss 2.9344 (2.9954) grad_norm 3.7255 (nan) [2021-04-16 20:48:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][670/1251] eta 0:02:47 lr 0.000014 time 0.2761 (0.2886) loss 2.6815 (2.9956) grad_norm 3.0985 (nan) [2021-04-16 20:48:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][680/1251] eta 0:02:44 lr 0.000014 time 0.2714 (0.2884) loss 3.0931 (2.9956) grad_norm 3.3380 (nan) [2021-04-16 20:48:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][690/1251] eta 0:02:41 lr 0.000014 time 0.2637 (0.2882) loss 2.7138 (2.9950) grad_norm 2.8537 (nan) [2021-04-16 20:48:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][700/1251] eta 0:02:38 lr 0.000014 time 0.4066 (0.2883) loss 2.4905 (2.9924) grad_norm 3.3529 (nan) [2021-04-16 20:48:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][710/1251] eta 0:02:35 lr 0.000014 time 0.3658 (0.2882) loss 3.2697 (2.9896) grad_norm 4.2161 (nan) [2021-04-16 20:48:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][720/1251] eta 0:02:32 lr 0.000014 time 0.3035 (0.2881) loss 3.2303 (2.9900) grad_norm 3.2468 (nan) [2021-04-16 20:48:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][730/1251] eta 0:02:29 lr 0.000014 time 0.2830 (0.2879) loss 3.3759 (2.9878) grad_norm 3.2231 (nan) [2021-04-16 20:48:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][740/1251] eta 0:02:26 lr 0.000014 time 0.2699 (0.2876) loss 3.1653 (2.9898) grad_norm 3.4314 (nan) [2021-04-16 20:48:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][750/1251] eta 0:02:24 lr 0.000014 time 0.2727 (0.2875) loss 2.5609 (2.9868) grad_norm 3.1303 (nan) [2021-04-16 20:48:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][760/1251] eta 0:02:21 lr 0.000014 time 0.2787 (0.2874) loss 2.2516 (2.9874) grad_norm 3.1535 (nan) [2021-04-16 20:48:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][770/1251] eta 0:02:18 lr 0.000014 time 0.2781 (0.2873) loss 3.3830 (2.9857) grad_norm 4.9307 (nan) [2021-04-16 20:48:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][780/1251] eta 0:02:15 lr 0.000014 time 0.2859 (0.2871) loss 2.9039 (2.9852) grad_norm 3.3481 (nan) [2021-04-16 20:48:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][790/1251] eta 0:02:12 lr 0.000014 time 0.2737 (0.2871) loss 2.6298 (2.9833) grad_norm 2.6504 (nan) [2021-04-16 20:48:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][800/1251] eta 0:02:09 lr 0.000014 time 0.2682 (0.2869) loss 3.2920 (2.9819) grad_norm 2.8263 (nan) [2021-04-16 20:48:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][810/1251] eta 0:02:06 lr 0.000014 time 0.2576 (0.2868) loss 2.1407 (2.9809) grad_norm 2.9201 (nan) [2021-04-16 20:48:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][820/1251] eta 0:02:03 lr 0.000014 time 0.2819 (0.2867) loss 3.3144 (2.9799) grad_norm 3.4121 (nan) [2021-04-16 20:49:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][830/1251] eta 0:02:00 lr 0.000014 time 0.2622 (0.2866) loss 2.6100 (2.9795) grad_norm 3.8606 (nan) [2021-04-16 20:49:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][840/1251] eta 0:01:57 lr 0.000014 time 0.2689 (0.2865) loss 3.2326 (2.9799) grad_norm 3.9473 (nan) [2021-04-16 20:49:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][850/1251] eta 0:01:54 lr 0.000014 time 0.2894 (0.2864) loss 3.1441 (2.9823) grad_norm 3.1092 (nan) [2021-04-16 20:49:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][860/1251] eta 0:01:51 lr 0.000014 time 0.3036 (0.2864) loss 3.0633 (2.9836) grad_norm 3.2890 (nan) [2021-04-16 20:49:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][870/1251] eta 0:01:49 lr 0.000014 time 0.3019 (0.2863) loss 3.9624 (2.9847) grad_norm 3.3360 (nan) [2021-04-16 20:49:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][880/1251] eta 0:01:46 lr 0.000014 time 0.2722 (0.2861) loss 3.1235 (2.9858) grad_norm 2.9564 (nan) [2021-04-16 20:49:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][890/1251] eta 0:01:43 lr 0.000014 time 0.2885 (0.2860) loss 2.6536 (2.9808) grad_norm 3.1830 (nan) [2021-04-16 20:49:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][900/1251] eta 0:01:40 lr 0.000014 time 0.2802 (0.2860) loss 3.2164 (2.9790) grad_norm 3.5584 (nan) [2021-04-16 20:49:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][910/1251] eta 0:01:37 lr 0.000014 time 0.2803 (0.2860) loss 3.4890 (2.9804) grad_norm 2.9237 (nan) [2021-04-16 20:49:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][920/1251] eta 0:01:34 lr 0.000014 time 0.2846 (0.2859) loss 2.9555 (2.9781) grad_norm 3.7505 (nan) [2021-04-16 20:49:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][930/1251] eta 0:01:31 lr 0.000014 time 0.2696 (0.2860) loss 3.2793 (2.9780) grad_norm 3.0884 (nan) [2021-04-16 20:49:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][940/1251] eta 0:01:28 lr 0.000014 time 0.2780 (0.2862) loss 3.1080 (2.9775) grad_norm 3.7357 (nan) [2021-04-16 20:49:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][950/1251] eta 0:01:26 lr 0.000014 time 0.3035 (0.2861) loss 2.4478 (2.9786) grad_norm 2.7508 (nan) [2021-04-16 20:49:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][960/1251] eta 0:01:23 lr 0.000014 time 0.2721 (0.2861) loss 3.6187 (2.9770) grad_norm 2.7955 (nan) [2021-04-16 20:49:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][970/1251] eta 0:01:20 lr 0.000014 time 0.3127 (0.2860) loss 3.5406 (2.9760) grad_norm 3.2885 (nan) [2021-04-16 20:49:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][980/1251] eta 0:01:17 lr 0.000014 time 0.2705 (0.2859) loss 3.4341 (2.9782) grad_norm 2.9576 (nan) [2021-04-16 20:49:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][990/1251] eta 0:01:14 lr 0.000014 time 0.2856 (0.2858) loss 3.1827 (2.9777) grad_norm 4.2738 (nan) [2021-04-16 20:49:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][1000/1251] eta 0:01:11 lr 0.000014 time 0.2633 (0.2857) loss 2.0841 (2.9785) grad_norm 4.5182 (nan) [2021-04-16 20:49:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][1010/1251] eta 0:01:08 lr 0.000014 time 0.3062 (0.2857) loss 2.1931 (2.9763) grad_norm 3.3527 (nan) [2021-04-16 20:49:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][1020/1251] eta 0:01:05 lr 0.000014 time 0.2789 (0.2856) loss 3.3772 (2.9771) grad_norm 3.2304 (nan) [2021-04-16 20:49:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][1030/1251] eta 0:01:03 lr 0.000014 time 0.2683 (0.2855) loss 3.2050 (2.9755) grad_norm 3.0130 (nan) [2021-04-16 20:49:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][1040/1251] eta 0:01:00 lr 0.000014 time 0.2478 (0.2854) loss 2.7652 (2.9767) grad_norm 3.8814 (nan) [2021-04-16 20:50:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][1050/1251] eta 0:00:57 lr 0.000014 time 0.2953 (0.2853) loss 2.4065 (2.9772) grad_norm 3.1950 (nan) [2021-04-16 20:50:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][1060/1251] eta 0:00:54 lr 0.000014 time 0.2860 (0.2853) loss 3.2450 (2.9751) grad_norm 2.8657 (nan) [2021-04-16 20:50:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][1070/1251] eta 0:00:51 lr 0.000014 time 0.2709 (0.2853) loss 3.2905 (2.9734) grad_norm 3.3548 (nan) [2021-04-16 20:50:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][1080/1251] eta 0:00:48 lr 0.000014 time 0.2970 (0.2852) loss 2.9579 (2.9760) grad_norm 3.1770 (nan) [2021-04-16 20:50:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][1090/1251] eta 0:00:45 lr 0.000014 time 0.2911 (0.2852) loss 1.9688 (2.9760) grad_norm 3.2888 (nan) [2021-04-16 20:50:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][1100/1251] eta 0:00:43 lr 0.000014 time 0.2793 (0.2851) loss 2.0534 (2.9736) grad_norm 3.3661 (nan) [2021-04-16 20:50:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][1110/1251] eta 0:00:40 lr 0.000014 time 0.2730 (0.2851) loss 3.4948 (2.9751) grad_norm 3.9013 (nan) [2021-04-16 20:50:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][1120/1251] eta 0:00:37 lr 0.000014 time 0.2778 (0.2850) loss 3.0898 (2.9760) grad_norm 3.0719 (nan) [2021-04-16 20:50:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][1130/1251] eta 0:00:34 lr 0.000014 time 0.2529 (0.2849) loss 2.7568 (2.9743) grad_norm 3.2157 (nan) [2021-04-16 20:50:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][1140/1251] eta 0:00:31 lr 0.000014 time 0.2984 (0.2849) loss 2.6510 (2.9719) grad_norm 2.8412 (nan) [2021-04-16 20:50:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][1150/1251] eta 0:00:28 lr 0.000014 time 0.2566 (0.2850) loss 2.1858 (2.9686) grad_norm 3.4811 (nan) [2021-04-16 20:50:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][1160/1251] eta 0:00:25 lr 0.000014 time 0.3083 (0.2851) loss 3.1275 (2.9676) grad_norm 2.9600 (nan) [2021-04-16 20:50:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][1170/1251] eta 0:00:23 lr 0.000014 time 0.2718 (0.2852) loss 3.1532 (2.9674) grad_norm 3.1595 (nan) [2021-04-16 20:50:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][1180/1251] eta 0:00:20 lr 0.000014 time 0.2572 (0.2852) loss 2.2546 (2.9665) grad_norm 3.4944 (nan) [2021-04-16 20:50:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][1190/1251] eta 0:00:17 lr 0.000014 time 0.2883 (0.2852) loss 1.9716 (2.9659) grad_norm 3.5904 (nan) [2021-04-16 20:50:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][1200/1251] eta 0:00:14 lr 0.000014 time 0.2669 (0.2852) loss 3.0828 (2.9655) grad_norm 2.9755 (nan) [2021-04-16 20:50:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][1210/1251] eta 0:00:11 lr 0.000014 time 0.2798 (0.2852) loss 2.1673 (2.9627) grad_norm 3.0867 (nan) [2021-04-16 20:50:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][1220/1251] eta 0:00:08 lr 0.000014 time 0.2693 (0.2851) loss 2.3843 (2.9627) grad_norm 3.1863 (nan) [2021-04-16 20:50:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][1230/1251] eta 0:00:05 lr 0.000014 time 0.2733 (0.2851) loss 2.2088 (2.9605) grad_norm 3.6172 (nan) [2021-04-16 20:50:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][1240/1251] eta 0:00:03 lr 0.000014 time 0.3575 (0.2850) loss 2.7217 (2.9611) grad_norm 3.4730 (nan) [2021-04-16 20:50:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [287/300][1250/1251] eta 0:00:00 lr 0.000014 time 0.2680 (0.2847) loss 2.2768 (2.9608) grad_norm 3.0616 (nan) [2021-04-16 20:51:04 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 287 training takes 0:06:01 [2021-04-16 20:51:04 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_287.pth saving...... [2021-04-16 20:51:25 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_287.pth saved !!! [2021-04-16 20:51:26 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.198 (1.198) Loss 0.7948 (0.7948) Acc@1 80.762 (80.762) Acc@5 96.094 (96.094) [2021-04-16 20:51:27 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.102 (0.212) Loss 0.8318 (0.8037) Acc@1 80.664 (81.348) Acc@5 95.312 (95.508) [2021-04-16 20:51:30 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.300 (0.249) Loss 0.6829 (0.8015) Acc@1 83.496 (81.510) Acc@5 96.875 (95.564) [2021-04-16 20:51:32 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.083 (0.238) Loss 0.8074 (0.8047) Acc@1 80.469 (81.530) Acc@5 95.312 (95.514) [2021-04-16 20:51:34 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.219) Loss 0.8350 (0.8070) Acc@1 80.469 (81.398) Acc@5 95.312 (95.520) [2021-04-16 20:51:53 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 81.206 Acc@5 95.488 [2021-04-16 20:51:53 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 81.2% [2021-04-16 20:51:53 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 81.21% [2021-04-16 20:52:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][0/1251] eta 8:23:27 lr 0.000014 time 24.1471 (24.1471) loss 3.5272 (3.5272) grad_norm 2.9387 (2.9387) [2021-04-16 20:52:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][10/1251] eta 0:50:35 lr 0.000014 time 0.2913 (2.4464) loss 2.3198 (3.0152) grad_norm 3.0182 (3.1824) [2021-04-16 20:52:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][20/1251] eta 0:29:04 lr 0.000014 time 0.2933 (1.4174) loss 3.1652 (2.9686) grad_norm 4.0917 (3.2961) [2021-04-16 20:52:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][30/1251] eta 0:21:21 lr 0.000014 time 0.2775 (1.0493) loss 3.5217 (3.0228) grad_norm 3.6154 (3.2902) [2021-04-16 20:52:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][40/1251] eta 0:17:22 lr 0.000014 time 0.2435 (0.8608) loss 3.6778 (2.9838) grad_norm 3.6387 (3.2895) [2021-04-16 20:52:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][50/1251] eta 0:14:56 lr 0.000014 time 0.2738 (0.7468) loss 3.1337 (2.9480) grad_norm 3.0658 (3.2524) [2021-04-16 20:52:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][60/1251] eta 0:13:17 lr 0.000014 time 0.2851 (0.6694) loss 3.3790 (2.9790) grad_norm 3.3853 (3.2736) [2021-04-16 20:52:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][70/1251] eta 0:12:05 lr 0.000014 time 0.2580 (0.6147) loss 3.4021 (2.9771) grad_norm 2.6482 (3.2641) [2021-04-16 20:52:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][80/1251] eta 0:11:10 lr 0.000014 time 0.2646 (0.5726) loss 3.3594 (2.9738) grad_norm 3.6338 (3.3053) [2021-04-16 20:52:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][90/1251] eta 0:10:27 lr 0.000014 time 0.2842 (0.5403) loss 2.9640 (2.9662) grad_norm 3.6949 (3.3189) [2021-04-16 20:52:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][100/1251] eta 0:09:51 lr 0.000014 time 0.2608 (0.5137) loss 2.3867 (2.9266) grad_norm 3.1506 (3.3106) [2021-04-16 20:52:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][110/1251] eta 0:09:21 lr 0.000014 time 0.2893 (0.4925) loss 3.1750 (2.9360) grad_norm 2.9249 (3.3225) [2021-04-16 20:52:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][120/1251] eta 0:08:57 lr 0.000014 time 0.2769 (0.4750) loss 3.3097 (2.9447) grad_norm 3.2382 (3.3260) [2021-04-16 20:52:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][130/1251] eta 0:08:35 lr 0.000014 time 0.2736 (0.4599) loss 2.5561 (2.9223) grad_norm 3.4072 (3.3230) [2021-04-16 20:52:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][140/1251] eta 0:08:18 lr 0.000014 time 0.2545 (0.4490) loss 2.1314 (2.9325) grad_norm 3.9204 (3.3309) [2021-04-16 20:52:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][150/1251] eta 0:08:03 lr 0.000014 time 0.2552 (0.4387) loss 2.7344 (2.9557) grad_norm 3.6557 (3.3364) [2021-04-16 20:53:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][160/1251] eta 0:07:47 lr 0.000014 time 0.2698 (0.4283) loss 2.2817 (2.9551) grad_norm 3.1104 (3.3310) [2021-04-16 20:53:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][170/1251] eta 0:07:33 lr 0.000014 time 0.2772 (0.4194) loss 3.6808 (2.9598) grad_norm 3.2522 (3.3332) [2021-04-16 20:53:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][180/1251] eta 0:07:21 lr 0.000014 time 0.2781 (0.4119) loss 2.1840 (2.9400) grad_norm 3.2058 (3.3297) [2021-04-16 20:53:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][190/1251] eta 0:07:09 lr 0.000014 time 0.2727 (0.4047) loss 2.9507 (2.9399) grad_norm 3.4584 (3.3274) [2021-04-16 20:53:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][200/1251] eta 0:06:58 lr 0.000014 time 0.2648 (0.3982) loss 3.5258 (2.9449) grad_norm 3.2811 (3.3260) [2021-04-16 20:53:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][210/1251] eta 0:06:48 lr 0.000014 time 0.2998 (0.3927) loss 3.2715 (2.9555) grad_norm 3.5545 (3.3290) [2021-04-16 20:53:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][220/1251] eta 0:06:39 lr 0.000014 time 0.2792 (0.3880) loss 3.3162 (2.9494) grad_norm 3.1420 (3.3278) [2021-04-16 20:53:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][230/1251] eta 0:06:31 lr 0.000014 time 0.3016 (0.3834) loss 3.4985 (2.9502) grad_norm 4.5671 (3.3329) [2021-04-16 20:53:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][240/1251] eta 0:06:23 lr 0.000014 time 0.2874 (0.3792) loss 2.1175 (2.9534) grad_norm 2.8322 (3.3310) [2021-04-16 20:53:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][250/1251] eta 0:06:15 lr 0.000014 time 0.2902 (0.3751) loss 1.9377 (2.9419) grad_norm 3.4718 (3.3280) [2021-04-16 20:53:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][260/1251] eta 0:06:07 lr 0.000014 time 0.2649 (0.3711) loss 3.2635 (2.9490) grad_norm 4.2832 (3.3341) [2021-04-16 20:53:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][270/1251] eta 0:06:00 lr 0.000014 time 0.2910 (0.3677) loss 3.5620 (2.9612) grad_norm 2.8419 (3.3323) [2021-04-16 20:53:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][280/1251] eta 0:05:53 lr 0.000014 time 0.2657 (0.3644) loss 2.7924 (2.9609) grad_norm 3.4172 (3.3391) [2021-04-16 20:53:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][290/1251] eta 0:05:47 lr 0.000014 time 0.2463 (0.3614) loss 2.4832 (2.9589) grad_norm 3.1507 (3.3348) [2021-04-16 20:53:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][300/1251] eta 0:05:41 lr 0.000014 time 0.2777 (0.3590) loss 3.5824 (2.9592) grad_norm 3.1270 (3.3301) [2021-04-16 20:53:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][310/1251] eta 0:05:35 lr 0.000014 time 0.2699 (0.3564) loss 3.3678 (2.9651) grad_norm 3.1614 (3.3305) [2021-04-16 20:53:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][320/1251] eta 0:05:29 lr 0.000014 time 0.2881 (0.3538) loss 2.6165 (2.9633) grad_norm 2.9629 (3.3305) [2021-04-16 20:53:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][330/1251] eta 0:05:23 lr 0.000014 time 0.2714 (0.3515) loss 1.9733 (2.9580) grad_norm 3.0232 (3.3279) [2021-04-16 20:53:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][340/1251] eta 0:05:18 lr 0.000014 time 0.2765 (0.3492) loss 2.8080 (2.9616) grad_norm 3.2427 (3.3273) [2021-04-16 20:53:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][350/1251] eta 0:05:12 lr 0.000014 time 0.2726 (0.3470) loss 3.2181 (2.9616) grad_norm 2.8506 (3.3307) [2021-04-16 20:53:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][360/1251] eta 0:05:08 lr 0.000014 time 0.2889 (0.3459) loss 2.0701 (2.9579) grad_norm 11.3785 (3.3501) [2021-04-16 20:54:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][370/1251] eta 0:05:03 lr 0.000014 time 0.2710 (0.3447) loss 3.2262 (2.9532) grad_norm 3.7350 (3.3548) [2021-04-16 20:54:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][380/1251] eta 0:04:58 lr 0.000014 time 0.2555 (0.3428) loss 3.4790 (2.9534) grad_norm 3.0083 (3.3553) [2021-04-16 20:54:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][390/1251] eta 0:04:54 lr 0.000014 time 0.2731 (0.3415) loss 3.2714 (2.9566) grad_norm 3.6128 (3.3593) [2021-04-16 20:54:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][400/1251] eta 0:04:49 lr 0.000014 time 0.2983 (0.3400) loss 2.8977 (2.9542) grad_norm 3.1590 (3.3739) [2021-04-16 20:54:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][410/1251] eta 0:04:44 lr 0.000014 time 0.2567 (0.3386) loss 3.1227 (2.9556) grad_norm 3.3565 (3.3822) [2021-04-16 20:54:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][420/1251] eta 0:04:40 lr 0.000014 time 0.2860 (0.3372) loss 2.1397 (2.9514) grad_norm 3.7628 (3.3930) [2021-04-16 20:54:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][430/1251] eta 0:04:35 lr 0.000014 time 0.2980 (0.3359) loss 2.9052 (2.9514) grad_norm 4.1489 (3.3952) [2021-04-16 20:54:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][440/1251] eta 0:04:31 lr 0.000014 time 0.2775 (0.3346) loss 2.8141 (2.9553) grad_norm 2.9918 (3.3966) [2021-04-16 20:54:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][450/1251] eta 0:04:27 lr 0.000014 time 0.4694 (0.3336) loss 3.6029 (2.9521) grad_norm 4.4692 (3.3974) [2021-04-16 20:54:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][460/1251] eta 0:04:22 lr 0.000014 time 0.2574 (0.3323) loss 3.2253 (2.9555) grad_norm 3.2832 (3.3981) [2021-04-16 20:54:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][470/1251] eta 0:04:18 lr 0.000014 time 0.2476 (0.3312) loss 2.8344 (2.9559) grad_norm 3.0645 (3.3984) [2021-04-16 20:54:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][480/1251] eta 0:04:14 lr 0.000014 time 0.2793 (0.3301) loss 2.0892 (2.9585) grad_norm 3.5205 (3.3946) [2021-04-16 20:54:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][490/1251] eta 0:04:10 lr 0.000014 time 0.3136 (0.3293) loss 3.1361 (2.9570) grad_norm 3.7965 (3.3958) [2021-04-16 20:54:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][500/1251] eta 0:04:06 lr 0.000014 time 0.2973 (0.3283) loss 2.5083 (2.9590) grad_norm 2.6822 (3.3897) [2021-04-16 20:54:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][510/1251] eta 0:04:02 lr 0.000014 time 0.2764 (0.3273) loss 3.2965 (2.9616) grad_norm 3.0498 (3.3856) [2021-04-16 20:54:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][520/1251] eta 0:03:58 lr 0.000014 time 0.2527 (0.3265) loss 3.4583 (2.9622) grad_norm 3.5405 (3.3833) [2021-04-16 20:54:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][530/1251] eta 0:03:54 lr 0.000014 time 0.2896 (0.3256) loss 3.1025 (2.9591) grad_norm 3.5926 (3.3827) [2021-04-16 20:54:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][540/1251] eta 0:03:50 lr 0.000014 time 0.2754 (0.3247) loss 3.7883 (2.9587) grad_norm 3.1014 (3.3833) [2021-04-16 20:54:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][550/1251] eta 0:03:47 lr 0.000014 time 0.2834 (0.3240) loss 2.1255 (2.9580) grad_norm 3.0682 (3.3839) [2021-04-16 20:54:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][560/1251] eta 0:03:43 lr 0.000014 time 0.2814 (0.3232) loss 3.0237 (2.9612) grad_norm 3.2683 (3.3853) [2021-04-16 20:54:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][570/1251] eta 0:03:39 lr 0.000014 time 0.4172 (0.3230) loss 2.1588 (2.9531) grad_norm 3.0786 (3.3841) [2021-04-16 20:55:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][580/1251] eta 0:03:36 lr 0.000014 time 0.2513 (0.3221) loss 2.8231 (2.9523) grad_norm 3.1454 (3.3810) [2021-04-16 20:55:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][590/1251] eta 0:03:32 lr 0.000014 time 0.2596 (0.3213) loss 3.3289 (2.9544) grad_norm 3.3920 (3.3801) [2021-04-16 20:55:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][600/1251] eta 0:03:28 lr 0.000014 time 0.2508 (0.3205) loss 3.2013 (2.9531) grad_norm 3.3289 (3.3862) [2021-04-16 20:55:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][610/1251] eta 0:03:25 lr 0.000014 time 0.2591 (0.3198) loss 2.9665 (2.9551) grad_norm 3.0585 (3.3892) [2021-04-16 20:55:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][620/1251] eta 0:03:21 lr 0.000014 time 0.2803 (0.3191) loss 3.0458 (2.9542) grad_norm 4.1924 (3.3900) [2021-04-16 20:55:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][630/1251] eta 0:03:17 lr 0.000014 time 0.2898 (0.3184) loss 2.2152 (2.9529) grad_norm 2.9690 (3.3874) [2021-04-16 20:55:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][640/1251] eta 0:03:14 lr 0.000014 time 0.2884 (0.3177) loss 3.5262 (2.9542) grad_norm 3.4583 (3.3860) [2021-04-16 20:55:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][650/1251] eta 0:03:10 lr 0.000014 time 0.2763 (0.3171) loss 2.9264 (2.9547) grad_norm 3.4488 (3.3831) [2021-04-16 20:55:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][660/1251] eta 0:03:07 lr 0.000014 time 0.2715 (0.3165) loss 2.5618 (2.9522) grad_norm 3.0292 (3.3794) [2021-04-16 20:55:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][670/1251] eta 0:03:03 lr 0.000014 time 0.2729 (0.3160) loss 3.3014 (2.9535) grad_norm 3.3505 (3.3797) [2021-04-16 20:55:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][680/1251] eta 0:03:00 lr 0.000014 time 0.2833 (0.3154) loss 3.5522 (2.9531) grad_norm 3.8843 (3.3848) [2021-04-16 20:55:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][690/1251] eta 0:02:56 lr 0.000014 time 0.2911 (0.3148) loss 3.0311 (2.9509) grad_norm 3.2029 (3.3816) [2021-04-16 20:55:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][700/1251] eta 0:02:53 lr 0.000014 time 0.2931 (0.3142) loss 2.5813 (2.9489) grad_norm 3.3528 (3.3818) [2021-04-16 20:55:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][710/1251] eta 0:02:49 lr 0.000014 time 0.2892 (0.3138) loss 2.5092 (2.9492) grad_norm 3.3203 (3.3824) [2021-04-16 20:55:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][720/1251] eta 0:02:46 lr 0.000014 time 0.2931 (0.3136) loss 3.0429 (2.9496) grad_norm 3.5304 (3.3843) [2021-04-16 20:55:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][730/1251] eta 0:02:43 lr 0.000014 time 0.2765 (0.3131) loss 2.0242 (2.9492) grad_norm 2.8984 (3.3836) [2021-04-16 20:55:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][740/1251] eta 0:02:39 lr 0.000014 time 0.2914 (0.3126) loss 2.9248 (2.9449) grad_norm 2.9366 (3.3820) [2021-04-16 20:55:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][750/1251] eta 0:02:36 lr 0.000014 time 0.2899 (0.3122) loss 3.4203 (2.9450) grad_norm 2.7800 (3.3786) [2021-04-16 20:55:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][760/1251] eta 0:02:33 lr 0.000014 time 0.2853 (0.3117) loss 2.0394 (2.9439) grad_norm 3.5200 (3.3766) [2021-04-16 20:55:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][770/1251] eta 0:02:29 lr 0.000014 time 0.2532 (0.3112) loss 2.6141 (2.9406) grad_norm 3.4123 (3.3793) [2021-04-16 20:55:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][780/1251] eta 0:02:26 lr 0.000014 time 0.2838 (0.3108) loss 2.4531 (2.9420) grad_norm 3.4307 (3.3805) [2021-04-16 20:55:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][790/1251] eta 0:02:23 lr 0.000014 time 0.2673 (0.3104) loss 3.3659 (2.9430) grad_norm 3.3178 (3.3803) [2021-04-16 20:56:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][800/1251] eta 0:02:19 lr 0.000013 time 0.3039 (0.3100) loss 3.0885 (2.9404) grad_norm 3.1925 (3.3807) [2021-04-16 20:56:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][810/1251] eta 0:02:16 lr 0.000013 time 0.2900 (0.3096) loss 2.9995 (2.9383) grad_norm 4.0543 (3.3833) [2021-04-16 20:56:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][820/1251] eta 0:02:13 lr 0.000013 time 0.2874 (0.3092) loss 3.4111 (2.9412) grad_norm 4.1594 (3.3853) [2021-04-16 20:56:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][830/1251] eta 0:02:10 lr 0.000013 time 0.3005 (0.3089) loss 2.4450 (2.9406) grad_norm 3.0650 (3.3842) [2021-04-16 20:56:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][840/1251] eta 0:02:06 lr 0.000013 time 0.3058 (0.3086) loss 3.1887 (2.9398) grad_norm 3.5559 (3.3837) [2021-04-16 20:56:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][850/1251] eta 0:02:03 lr 0.000013 time 0.2900 (0.3082) loss 3.4248 (2.9420) grad_norm 4.3852 (3.3857) [2021-04-16 20:56:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][860/1251] eta 0:02:00 lr 0.000013 time 0.2851 (0.3078) loss 3.6970 (2.9430) grad_norm 3.0051 (3.3914) [2021-04-16 20:56:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][870/1251] eta 0:01:57 lr 0.000013 time 0.2749 (0.3074) loss 2.7434 (2.9454) grad_norm 3.5138 (3.3957) [2021-04-16 20:56:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][880/1251] eta 0:01:53 lr 0.000013 time 0.2815 (0.3071) loss 3.2076 (2.9476) grad_norm 3.3587 (3.3948) [2021-04-16 20:56:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][890/1251] eta 0:01:50 lr 0.000013 time 0.2805 (0.3068) loss 2.0756 (2.9485) grad_norm 3.3764 (3.3956) [2021-04-16 20:56:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][900/1251] eta 0:01:47 lr 0.000013 time 0.3000 (0.3066) loss 2.8998 (2.9502) grad_norm 3.0641 (3.3990) [2021-04-16 20:56:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][910/1251] eta 0:01:44 lr 0.000013 time 0.2680 (0.3063) loss 2.7462 (2.9503) grad_norm 4.6265 (3.4035) [2021-04-16 20:56:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][920/1251] eta 0:01:41 lr 0.000013 time 0.2887 (0.3060) loss 3.7184 (2.9512) grad_norm 3.5313 (3.4073) [2021-04-16 20:56:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][930/1251] eta 0:01:38 lr 0.000013 time 0.2708 (0.3059) loss 2.7966 (2.9509) grad_norm 3.3451 (3.4099) [2021-04-16 20:56:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][940/1251] eta 0:01:35 lr 0.000013 time 0.4387 (0.3060) loss 3.0412 (2.9519) grad_norm 3.0028 (3.4100) [2021-04-16 20:56:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][950/1251] eta 0:01:32 lr 0.000013 time 0.2770 (0.3057) loss 2.9895 (2.9539) grad_norm 3.6833 (3.4101) [2021-04-16 20:56:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][960/1251] eta 0:01:28 lr 0.000013 time 0.2843 (0.3054) loss 2.6488 (2.9514) grad_norm 6.7031 (3.4157) [2021-04-16 20:56:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][970/1251] eta 0:01:25 lr 0.000013 time 0.2637 (0.3053) loss 3.0095 (2.9490) grad_norm 4.1348 (3.4183) [2021-04-16 20:56:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][980/1251] eta 0:01:22 lr 0.000013 time 0.2954 (0.3050) loss 3.5434 (2.9511) grad_norm 3.2827 (3.4172) [2021-04-16 20:56:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][990/1251] eta 0:01:19 lr 0.000013 time 0.2690 (0.3048) loss 3.0151 (2.9510) grad_norm 4.1044 (3.4186) [2021-04-16 20:56:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][1000/1251] eta 0:01:16 lr 0.000013 time 0.2701 (0.3044) loss 3.0973 (2.9529) grad_norm 3.6704 (3.4176) [2021-04-16 20:57:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][1010/1251] eta 0:01:13 lr 0.000013 time 0.2658 (0.3042) loss 2.7911 (2.9532) grad_norm 3.5561 (3.4159) [2021-04-16 20:57:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][1020/1251] eta 0:01:10 lr 0.000013 time 0.2782 (0.3039) loss 3.5246 (2.9532) grad_norm 3.7727 (3.4184) [2021-04-16 20:57:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][1030/1251] eta 0:01:07 lr 0.000013 time 0.2571 (0.3038) loss 3.0816 (2.9516) grad_norm 2.9606 (3.4176) [2021-04-16 20:57:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][1040/1251] eta 0:01:04 lr 0.000013 time 0.2661 (0.3034) loss 2.4461 (2.9510) grad_norm 3.1228 (3.4191) [2021-04-16 20:57:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][1050/1251] eta 0:01:00 lr 0.000013 time 0.2920 (0.3032) loss 3.1043 (2.9516) grad_norm 3.3955 (3.4199) [2021-04-16 20:57:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][1060/1251] eta 0:00:57 lr 0.000013 time 0.2631 (0.3029) loss 2.9672 (2.9524) grad_norm 3.2415 (3.4225) [2021-04-16 20:57:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][1070/1251] eta 0:00:54 lr 0.000013 time 0.2511 (0.3027) loss 2.6898 (2.9517) grad_norm 3.5244 (3.4255) [2021-04-16 20:57:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][1080/1251] eta 0:00:51 lr 0.000013 time 0.2668 (0.3024) loss 3.1476 (2.9540) grad_norm 3.3328 (3.4247) [2021-04-16 20:57:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][1090/1251] eta 0:00:48 lr 0.000013 time 0.2698 (0.3023) loss 3.5108 (2.9545) grad_norm 3.4376 (3.4260) [2021-04-16 20:57:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][1100/1251] eta 0:00:45 lr 0.000013 time 0.3004 (0.3021) loss 3.0601 (2.9553) grad_norm 3.4756 (3.4249) [2021-04-16 20:57:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][1110/1251] eta 0:00:42 lr 0.000013 time 0.2720 (0.3018) loss 2.9240 (2.9537) grad_norm 3.4365 (3.4244) [2021-04-16 20:57:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][1120/1251] eta 0:00:39 lr 0.000013 time 0.2645 (0.3016) loss 2.5086 (2.9520) grad_norm 3.1585 (3.4223) [2021-04-16 20:57:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][1130/1251] eta 0:00:36 lr 0.000013 time 0.2713 (0.3014) loss 3.2212 (2.9520) grad_norm 3.3720 (3.4211) [2021-04-16 20:57:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][1140/1251] eta 0:00:33 lr 0.000013 time 0.2838 (0.3012) loss 3.0186 (2.9554) grad_norm 3.7960 (3.4188) [2021-04-16 20:57:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][1150/1251] eta 0:00:30 lr 0.000013 time 0.2965 (0.3012) loss 2.7231 (2.9530) grad_norm 3.3821 (3.4179) [2021-04-16 20:57:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][1160/1251] eta 0:00:27 lr 0.000013 time 0.2631 (0.3009) loss 3.0396 (2.9538) grad_norm 3.9135 (3.4188) [2021-04-16 20:57:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][1170/1251] eta 0:00:24 lr 0.000013 time 0.3004 (0.3007) loss 2.6659 (2.9538) grad_norm 3.6642 (3.4214) [2021-04-16 20:57:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][1180/1251] eta 0:00:21 lr 0.000013 time 0.2706 (0.3005) loss 2.3778 (2.9546) grad_norm 3.2842 (3.4220) [2021-04-16 20:57:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][1190/1251] eta 0:00:18 lr 0.000013 time 0.2616 (0.3003) loss 2.2605 (2.9526) grad_norm 3.0865 (3.4215) [2021-04-16 20:57:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][1200/1251] eta 0:00:15 lr 0.000013 time 0.2671 (0.3001) loss 3.2446 (2.9541) grad_norm 3.2013 (3.4200) [2021-04-16 20:57:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][1210/1251] eta 0:00:12 lr 0.000013 time 0.2737 (0.2999) loss 3.1549 (2.9559) grad_norm 2.9038 (3.4225) [2021-04-16 20:57:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][1220/1251] eta 0:00:09 lr 0.000013 time 0.2698 (0.2997) loss 2.2577 (2.9567) grad_norm 3.3891 (3.4218) [2021-04-16 20:58:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][1230/1251] eta 0:00:06 lr 0.000013 time 0.2972 (0.2995) loss 3.3192 (2.9574) grad_norm 3.5578 (3.4222) [2021-04-16 20:58:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][1240/1251] eta 0:00:03 lr 0.000013 time 0.2481 (0.2993) loss 2.8965 (2.9584) grad_norm 3.1525 (3.4212) [2021-04-16 20:58:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [288/300][1250/1251] eta 0:00:00 lr 0.000013 time 0.2446 (0.2989) loss 3.3005 (2.9603) grad_norm 3.2620 (3.4211) [2021-04-16 20:58:12 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 288 training takes 0:06:18 [2021-04-16 20:58:12 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_288.pth saving...... [2021-04-16 20:58:25 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_288.pth saved !!! [2021-04-16 20:58:26 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.165 (1.165) Loss 0.8351 (0.8351) Acc@1 79.199 (79.199) Acc@5 95.312 (95.312) [2021-04-16 20:58:27 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.073 (0.200) Loss 0.7905 (0.8025) Acc@1 82.031 (81.481) Acc@5 95.898 (95.827) [2021-04-16 20:58:30 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.804 (0.250) Loss 0.8243 (0.8015) Acc@1 79.297 (81.399) Acc@5 95.312 (95.750) [2021-04-16 20:58:33 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.078 (0.257) Loss 0.8544 (0.8053) Acc@1 80.469 (81.357) Acc@5 94.922 (95.621) [2021-04-16 20:58:34 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.221) Loss 0.8241 (0.8082) Acc@1 79.883 (81.257) Acc@5 95.215 (95.582) [2021-04-16 20:58:54 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 81.126 Acc@5 95.500 [2021-04-16 20:58:54 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 81.1% [2021-04-16 20:58:54 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 81.21% [2021-04-16 20:59:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][0/1251] eta 4:56:46 lr 0.000013 time 14.2340 (14.2340) loss 3.1114 (3.1114) grad_norm 3.0178 (3.0178) [2021-04-16 20:59:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][10/1251] eta 0:32:02 lr 0.000013 time 0.3577 (1.5494) loss 3.0194 (2.9087) grad_norm 3.7067 (3.2617) [2021-04-16 20:59:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][20/1251] eta 0:19:20 lr 0.000013 time 0.2918 (0.9430) loss 3.6187 (2.9779) grad_norm 3.9390 (3.3674) [2021-04-16 20:59:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][30/1251] eta 0:14:52 lr 0.000013 time 0.2829 (0.7309) loss 2.7955 (2.9359) grad_norm 3.3517 (3.3333) [2021-04-16 20:59:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][40/1251] eta 0:12:31 lr 0.000013 time 0.2929 (0.6203) loss 2.4823 (2.8423) grad_norm 3.7426 (3.3286) [2021-04-16 20:59:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][50/1251] eta 0:11:03 lr 0.000013 time 0.2756 (0.5528) loss 2.9384 (2.8778) grad_norm 4.1336 (3.3669) [2021-04-16 20:59:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][60/1251] eta 0:10:04 lr 0.000013 time 0.3006 (0.5079) loss 3.0297 (2.8940) grad_norm 3.4659 (3.3733) [2021-04-16 20:59:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][70/1251] eta 0:09:22 lr 0.000013 time 0.2693 (0.4760) loss 3.3445 (2.8935) grad_norm 3.3534 (3.4044) [2021-04-16 20:59:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][80/1251] eta 0:08:49 lr 0.000013 time 0.2878 (0.4518) loss 3.2296 (2.8782) grad_norm 3.2903 (3.4161) [2021-04-16 20:59:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][90/1251] eta 0:08:23 lr 0.000013 time 0.2946 (0.4333) loss 3.3363 (2.8896) grad_norm 3.8761 (3.4157) [2021-04-16 20:59:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][100/1251] eta 0:08:01 lr 0.000013 time 0.3007 (0.4182) loss 3.1500 (2.8999) grad_norm 2.8883 (3.4022) [2021-04-16 20:59:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][110/1251] eta 0:07:43 lr 0.000013 time 0.3134 (0.4058) loss 2.4866 (2.8985) grad_norm 4.1228 (3.4028) [2021-04-16 20:59:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][120/1251] eta 0:07:27 lr 0.000013 time 0.2960 (0.3956) loss 2.3026 (2.8942) grad_norm 2.7880 (3.3866) [2021-04-16 20:59:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][130/1251] eta 0:07:14 lr 0.000013 time 0.2980 (0.3879) loss 3.4211 (2.8698) grad_norm 3.3042 (3.3940) [2021-04-16 20:59:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][140/1251] eta 0:07:04 lr 0.000013 time 0.2420 (0.3821) loss 2.6719 (2.8785) grad_norm 3.7066 (3.3971) [2021-04-16 20:59:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][150/1251] eta 0:06:53 lr 0.000013 time 0.2936 (0.3760) loss 3.0615 (2.8773) grad_norm 3.2914 (3.3873) [2021-04-16 20:59:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][160/1251] eta 0:06:45 lr 0.000013 time 0.2958 (0.3712) loss 2.5639 (2.8854) grad_norm 3.8643 (3.3961) [2021-04-16 20:59:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][170/1251] eta 0:06:35 lr 0.000013 time 0.2756 (0.3661) loss 2.9690 (2.8783) grad_norm 3.3538 (3.3946) [2021-04-16 20:59:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][180/1251] eta 0:06:27 lr 0.000013 time 0.2665 (0.3617) loss 2.1652 (2.8699) grad_norm 4.0045 (3.3917) [2021-04-16 21:00:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][190/1251] eta 0:06:19 lr 0.000013 time 0.2890 (0.3572) loss 3.2131 (2.8766) grad_norm 3.9090 (3.4027) [2021-04-16 21:00:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][200/1251] eta 0:06:11 lr 0.000013 time 0.2835 (0.3535) loss 3.0079 (2.8810) grad_norm 2.8183 (3.4084) [2021-04-16 21:00:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][210/1251] eta 0:06:04 lr 0.000013 time 0.2658 (0.3502) loss 3.5584 (2.8860) grad_norm 4.3262 (3.4133) [2021-04-16 21:00:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][220/1251] eta 0:05:57 lr 0.000013 time 0.2791 (0.3470) loss 2.1462 (2.8910) grad_norm 2.9977 (3.4199) [2021-04-16 21:00:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][230/1251] eta 0:05:51 lr 0.000013 time 0.2750 (0.3440) loss 2.4975 (2.8947) grad_norm 3.2922 (3.4220) [2021-04-16 21:00:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][240/1251] eta 0:05:45 lr 0.000013 time 0.2788 (0.3416) loss 2.9717 (2.8980) grad_norm 2.8951 (3.4203) [2021-04-16 21:00:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][250/1251] eta 0:05:39 lr 0.000013 time 0.2598 (0.3396) loss 2.9846 (2.9084) grad_norm 2.7448 (3.4201) [2021-04-16 21:00:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][260/1251] eta 0:05:34 lr 0.000013 time 0.2854 (0.3373) loss 2.3247 (2.8973) grad_norm 3.1997 (3.4192) [2021-04-16 21:00:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][270/1251] eta 0:05:28 lr 0.000013 time 0.2880 (0.3349) loss 2.5400 (2.8983) grad_norm 3.5594 (3.4149) [2021-04-16 21:00:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][280/1251] eta 0:05:23 lr 0.000013 time 0.2738 (0.3329) loss 2.4831 (2.8970) grad_norm 3.1980 (3.4152) [2021-04-16 21:00:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][290/1251] eta 0:05:18 lr 0.000013 time 0.2760 (0.3313) loss 3.1791 (2.9006) grad_norm 3.5261 (3.4194) [2021-04-16 21:00:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][300/1251] eta 0:05:13 lr 0.000013 time 0.2664 (0.3299) loss 3.2369 (2.9070) grad_norm 3.5620 (3.4203) [2021-04-16 21:00:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][310/1251] eta 0:05:08 lr 0.000013 time 0.2706 (0.3283) loss 1.8611 (2.8990) grad_norm 3.3417 (3.4253) [2021-04-16 21:00:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][320/1251] eta 0:05:04 lr 0.000013 time 0.2754 (0.3267) loss 1.7737 (2.9106) grad_norm 3.2331 (3.4256) [2021-04-16 21:00:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][330/1251] eta 0:04:59 lr 0.000013 time 0.2690 (0.3254) loss 3.0110 (2.9159) grad_norm 5.2185 (3.4294) [2021-04-16 21:00:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][340/1251] eta 0:04:55 lr 0.000013 time 0.2798 (0.3240) loss 3.1690 (2.9202) grad_norm 3.8672 (3.4266) [2021-04-16 21:00:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][350/1251] eta 0:04:50 lr 0.000013 time 0.2553 (0.3227) loss 3.2657 (2.9254) grad_norm 4.9900 (3.4292) [2021-04-16 21:00:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][360/1251] eta 0:04:46 lr 0.000013 time 0.2816 (0.3217) loss 3.2993 (2.9304) grad_norm 3.1331 (3.4276) [2021-04-16 21:00:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][370/1251] eta 0:04:42 lr 0.000013 time 0.2636 (0.3209) loss 1.9127 (2.9292) grad_norm 4.3967 (3.4322) [2021-04-16 21:00:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][380/1251] eta 0:04:38 lr 0.000013 time 0.2712 (0.3195) loss 3.2951 (2.9320) grad_norm 3.2730 (3.4308) [2021-04-16 21:00:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][390/1251] eta 0:04:34 lr 0.000013 time 0.2962 (0.3186) loss 2.5918 (2.9330) grad_norm 3.3973 (3.4303) [2021-04-16 21:01:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][400/1251] eta 0:04:30 lr 0.000013 time 0.2791 (0.3175) loss 3.2898 (2.9369) grad_norm 2.9499 (3.4348) [2021-04-16 21:01:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][410/1251] eta 0:04:26 lr 0.000013 time 0.2729 (0.3166) loss 2.2924 (2.9399) grad_norm 4.8126 (3.4415) [2021-04-16 21:01:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][420/1251] eta 0:04:22 lr 0.000013 time 0.2703 (0.3158) loss 2.9773 (2.9422) grad_norm 3.5419 (3.4471) [2021-04-16 21:01:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][430/1251] eta 0:04:18 lr 0.000013 time 0.2739 (0.3151) loss 3.4872 (2.9398) grad_norm 3.6693 (3.4473) [2021-04-16 21:01:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][440/1251] eta 0:04:14 lr 0.000013 time 0.2714 (0.3143) loss 3.1003 (2.9424) grad_norm 3.2522 (3.4578) [2021-04-16 21:01:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][450/1251] eta 0:04:11 lr 0.000013 time 0.2828 (0.3136) loss 1.9834 (2.9423) grad_norm 3.3587 (3.4575) [2021-04-16 21:01:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][460/1251] eta 0:04:07 lr 0.000013 time 0.2926 (0.3128) loss 3.3566 (2.9438) grad_norm 3.1010 (3.4594) [2021-04-16 21:01:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][470/1251] eta 0:04:03 lr 0.000013 time 0.2879 (0.3120) loss 3.3906 (2.9480) grad_norm 3.3332 (3.4610) [2021-04-16 21:01:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][480/1251] eta 0:03:59 lr 0.000013 time 0.2744 (0.3112) loss 3.6079 (2.9511) grad_norm 3.2729 (3.4624) [2021-04-16 21:01:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][490/1251] eta 0:03:56 lr 0.000013 time 0.2649 (0.3105) loss 3.7293 (2.9527) grad_norm 3.3325 (3.4621) [2021-04-16 21:01:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][500/1251] eta 0:03:52 lr 0.000013 time 0.3159 (0.3100) loss 3.0405 (2.9550) grad_norm 2.9634 (3.4601) [2021-04-16 21:01:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][510/1251] eta 0:03:49 lr 0.000013 time 0.2764 (0.3094) loss 2.2495 (2.9565) grad_norm 3.0428 (3.4575) [2021-04-16 21:01:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][520/1251] eta 0:03:45 lr 0.000013 time 0.2735 (0.3089) loss 2.3024 (2.9554) grad_norm 3.1223 (3.4607) [2021-04-16 21:01:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][530/1251] eta 0:03:42 lr 0.000013 time 0.2795 (0.3084) loss 2.8762 (2.9511) grad_norm 3.0757 (3.4610) [2021-04-16 21:01:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][540/1251] eta 0:03:38 lr 0.000013 time 0.2900 (0.3080) loss 2.4250 (2.9515) grad_norm 3.6914 (3.4639) [2021-04-16 21:01:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][550/1251] eta 0:03:35 lr 0.000013 time 0.2639 (0.3074) loss 2.2103 (2.9561) grad_norm 3.3780 (3.4617) [2021-04-16 21:01:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][560/1251] eta 0:03:32 lr 0.000013 time 0.2625 (0.3069) loss 2.7392 (2.9517) grad_norm 3.5358 (3.4642) [2021-04-16 21:01:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][570/1251] eta 0:03:28 lr 0.000013 time 0.2667 (0.3067) loss 3.2048 (2.9529) grad_norm 3.0410 (3.4670) [2021-04-16 21:01:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][580/1251] eta 0:03:25 lr 0.000013 time 0.2788 (0.3065) loss 3.2264 (2.9548) grad_norm 3.0985 (3.4677) [2021-04-16 21:01:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][590/1251] eta 0:03:22 lr 0.000013 time 0.2747 (0.3062) loss 3.3022 (2.9570) grad_norm 3.3993 (3.4650) [2021-04-16 21:01:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][600/1251] eta 0:03:18 lr 0.000013 time 0.2583 (0.3057) loss 3.3942 (2.9534) grad_norm 3.7102 (3.4658) [2021-04-16 21:02:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][610/1251] eta 0:03:15 lr 0.000013 time 0.2751 (0.3053) loss 3.2347 (2.9535) grad_norm 3.4985 (3.4666) [2021-04-16 21:02:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][620/1251] eta 0:03:12 lr 0.000013 time 0.2599 (0.3048) loss 3.0959 (2.9531) grad_norm 3.7738 (3.4669) [2021-04-16 21:02:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][630/1251] eta 0:03:09 lr 0.000013 time 0.2858 (0.3044) loss 3.3292 (2.9563) grad_norm 2.9352 (3.4661) [2021-04-16 21:02:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][640/1251] eta 0:03:05 lr 0.000013 time 0.2857 (0.3039) loss 3.4052 (2.9532) grad_norm 2.6660 (3.4631) [2021-04-16 21:02:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][650/1251] eta 0:03:02 lr 0.000013 time 0.2871 (0.3035) loss 3.4877 (2.9525) grad_norm 3.5847 (3.4684) [2021-04-16 21:02:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][660/1251] eta 0:02:59 lr 0.000013 time 0.2966 (0.3031) loss 2.2059 (2.9533) grad_norm 4.2488 (3.4672) [2021-04-16 21:02:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][670/1251] eta 0:02:55 lr 0.000013 time 0.2745 (0.3028) loss 3.4483 (2.9545) grad_norm 3.7798 (3.4648) [2021-04-16 21:02:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][680/1251] eta 0:02:52 lr 0.000013 time 0.2920 (0.3024) loss 2.3773 (2.9543) grad_norm 2.9942 (3.4622) [2021-04-16 21:02:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][690/1251] eta 0:02:49 lr 0.000013 time 0.2865 (0.3020) loss 2.0656 (2.9528) grad_norm 3.0096 (3.4587) [2021-04-16 21:02:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][700/1251] eta 0:02:46 lr 0.000013 time 0.2680 (0.3019) loss 3.1018 (2.9511) grad_norm 3.4628 (3.4574) [2021-04-16 21:02:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][710/1251] eta 0:02:43 lr 0.000013 time 0.2821 (0.3016) loss 3.0365 (2.9498) grad_norm 3.2127 (3.4587) [2021-04-16 21:02:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][720/1251] eta 0:02:40 lr 0.000013 time 0.2798 (0.3017) loss 3.2082 (2.9534) grad_norm 2.9194 (3.4561) [2021-04-16 21:02:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][730/1251] eta 0:02:37 lr 0.000013 time 0.4680 (0.3016) loss 3.1797 (2.9579) grad_norm 3.1497 (3.4539) [2021-04-16 21:02:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][740/1251] eta 0:02:33 lr 0.000013 time 0.2694 (0.3012) loss 3.2835 (2.9575) grad_norm 3.2825 (3.4513) [2021-04-16 21:02:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][750/1251] eta 0:02:30 lr 0.000013 time 0.2871 (0.3010) loss 3.3185 (2.9588) grad_norm 3.1070 (3.4540) [2021-04-16 21:02:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][760/1251] eta 0:02:27 lr 0.000013 time 0.2776 (0.3008) loss 3.2743 (2.9617) grad_norm 3.2701 (3.4553) [2021-04-16 21:02:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][770/1251] eta 0:02:24 lr 0.000013 time 0.2735 (0.3005) loss 2.9755 (2.9632) grad_norm 3.2616 (3.4534) [2021-04-16 21:02:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][780/1251] eta 0:02:21 lr 0.000013 time 0.2802 (0.3002) loss 3.6080 (2.9622) grad_norm 3.5490 (3.4520) [2021-04-16 21:02:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][790/1251] eta 0:02:18 lr 0.000013 time 0.2684 (0.2999) loss 3.4284 (2.9639) grad_norm 3.5353 (3.4611) [2021-04-16 21:02:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][800/1251] eta 0:02:15 lr 0.000013 time 0.2879 (0.2997) loss 3.1075 (2.9649) grad_norm 3.3628 (3.4635) [2021-04-16 21:02:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][810/1251] eta 0:02:12 lr 0.000013 time 0.2804 (0.2996) loss 3.4905 (2.9662) grad_norm 4.4138 (3.4648) [2021-04-16 21:03:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][820/1251] eta 0:02:09 lr 0.000013 time 0.2633 (0.2994) loss 3.4016 (2.9663) grad_norm 3.5668 (3.4624) [2021-04-16 21:03:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][830/1251] eta 0:02:05 lr 0.000013 time 0.2861 (0.2992) loss 3.0908 (2.9668) grad_norm 3.2925 (3.4623) [2021-04-16 21:03:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][840/1251] eta 0:02:02 lr 0.000013 time 0.2768 (0.2990) loss 3.3471 (2.9657) grad_norm 3.2347 (3.4606) [2021-04-16 21:03:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][850/1251] eta 0:01:59 lr 0.000013 time 0.2927 (0.2988) loss 2.2682 (2.9651) grad_norm 3.3988 (3.4600) [2021-04-16 21:03:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][860/1251] eta 0:01:56 lr 0.000013 time 0.2949 (0.2986) loss 3.5428 (2.9667) grad_norm 2.8938 (3.4576) [2021-04-16 21:03:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][870/1251] eta 0:01:53 lr 0.000013 time 0.2832 (0.2985) loss 2.2432 (2.9678) grad_norm 3.1067 (3.4556) [2021-04-16 21:03:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][880/1251] eta 0:01:50 lr 0.000013 time 0.2894 (0.2983) loss 3.2782 (2.9662) grad_norm 3.1957 (3.4526) [2021-04-16 21:03:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][890/1251] eta 0:01:47 lr 0.000013 time 0.2664 (0.2981) loss 3.0511 (2.9665) grad_norm 3.4342 (3.4504) [2021-04-16 21:03:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][900/1251] eta 0:01:44 lr 0.000013 time 0.2864 (0.2979) loss 2.6198 (2.9659) grad_norm 4.1866 (3.4521) [2021-04-16 21:03:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][910/1251] eta 0:01:41 lr 0.000013 time 0.2712 (0.2977) loss 3.1237 (2.9687) grad_norm 4.0862 (3.4512) [2021-04-16 21:03:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][920/1251] eta 0:01:38 lr 0.000013 time 0.2821 (0.2975) loss 3.2323 (2.9696) grad_norm 4.0998 (3.4500) [2021-04-16 21:03:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][930/1251] eta 0:01:35 lr 0.000013 time 0.2646 (0.2973) loss 3.1309 (2.9710) grad_norm 4.9976 (3.4484) [2021-04-16 21:03:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][940/1251] eta 0:01:32 lr 0.000013 time 0.2852 (0.2973) loss 3.0934 (2.9694) grad_norm 3.1296 (3.4467) [2021-04-16 21:03:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][950/1251] eta 0:01:29 lr 0.000013 time 0.2817 (0.2972) loss 2.9417 (2.9695) grad_norm 3.4405 (3.4457) [2021-04-16 21:03:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][960/1251] eta 0:01:26 lr 0.000013 time 0.2734 (0.2970) loss 3.1248 (2.9691) grad_norm 3.0325 (3.4418) [2021-04-16 21:03:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][970/1251] eta 0:01:23 lr 0.000013 time 0.3028 (0.2969) loss 2.3533 (2.9687) grad_norm 3.2829 (3.4428) [2021-04-16 21:03:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][980/1251] eta 0:01:20 lr 0.000013 time 0.2806 (0.2967) loss 3.5103 (2.9703) grad_norm 3.0135 (3.4412) [2021-04-16 21:03:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][990/1251] eta 0:01:17 lr 0.000013 time 0.2627 (0.2966) loss 1.9948 (2.9700) grad_norm 3.3150 (3.4406) [2021-04-16 21:03:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][1000/1251] eta 0:01:14 lr 0.000013 time 0.2777 (0.2964) loss 3.4929 (2.9705) grad_norm 3.7557 (3.4433) [2021-04-16 21:03:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][1010/1251] eta 0:01:11 lr 0.000013 time 0.2826 (0.2962) loss 3.4684 (2.9705) grad_norm 3.1269 (3.4449) [2021-04-16 21:03:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][1020/1251] eta 0:01:08 lr 0.000013 time 0.3031 (0.2961) loss 3.1261 (2.9697) grad_norm 5.2622 (3.4460) [2021-04-16 21:03:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][1030/1251] eta 0:01:05 lr 0.000013 time 0.2978 (0.2959) loss 3.8877 (2.9710) grad_norm 3.1547 (3.4476) [2021-04-16 21:04:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][1040/1251] eta 0:01:02 lr 0.000013 time 0.2885 (0.2957) loss 3.3915 (2.9728) grad_norm 3.7291 (3.4464) [2021-04-16 21:04:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][1050/1251] eta 0:00:59 lr 0.000013 time 0.2803 (0.2956) loss 2.8579 (2.9721) grad_norm 3.3847 (3.4445) [2021-04-16 21:04:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][1060/1251] eta 0:00:56 lr 0.000013 time 0.3050 (0.2955) loss 3.0922 (2.9709) grad_norm 3.4493 (3.4440) [2021-04-16 21:04:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][1070/1251] eta 0:00:53 lr 0.000013 time 0.2695 (0.2954) loss 3.2105 (2.9724) grad_norm 2.7441 (3.4434) [2021-04-16 21:04:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][1080/1251] eta 0:00:50 lr 0.000013 time 0.2930 (0.2953) loss 3.3044 (2.9700) grad_norm 3.3272 (3.4436) [2021-04-16 21:04:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][1090/1251] eta 0:00:47 lr 0.000013 time 0.2525 (0.2952) loss 3.2646 (2.9703) grad_norm 3.7984 (3.4502) [2021-04-16 21:04:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][1100/1251] eta 0:00:44 lr 0.000013 time 0.2629 (0.2950) loss 1.8856 (2.9658) grad_norm 2.9812 (3.4489) [2021-04-16 21:04:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][1110/1251] eta 0:00:41 lr 0.000013 time 0.2753 (0.2948) loss 2.1368 (2.9638) grad_norm 3.2788 (3.4507) [2021-04-16 21:04:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][1120/1251] eta 0:00:38 lr 0.000013 time 0.2772 (0.2948) loss 2.2439 (2.9633) grad_norm 3.4483 (3.4516) [2021-04-16 21:04:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][1130/1251] eta 0:00:35 lr 0.000013 time 0.3033 (0.2946) loss 3.1770 (2.9643) grad_norm 3.2998 (3.4536) [2021-04-16 21:04:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][1140/1251] eta 0:00:32 lr 0.000013 time 0.2683 (0.2947) loss 2.3157 (2.9629) grad_norm 3.4872 (3.4566) [2021-04-16 21:04:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][1150/1251] eta 0:00:29 lr 0.000013 time 0.2716 (0.2947) loss 1.9704 (2.9612) grad_norm 3.4581 (3.4573) [2021-04-16 21:04:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][1160/1251] eta 0:00:26 lr 0.000013 time 0.2583 (0.2945) loss 3.1636 (2.9613) grad_norm 3.2385 (3.4553) [2021-04-16 21:04:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][1170/1251] eta 0:00:23 lr 0.000013 time 0.2634 (0.2944) loss 3.5719 (2.9610) grad_norm 3.9061 (3.4539) [2021-04-16 21:04:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][1180/1251] eta 0:00:20 lr 0.000013 time 0.2710 (0.2943) loss 2.0522 (2.9595) grad_norm 3.8555 (3.4544) [2021-04-16 21:04:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][1190/1251] eta 0:00:17 lr 0.000013 time 0.2839 (0.2942) loss 3.3857 (2.9598) grad_norm 3.1563 (3.4539) [2021-04-16 21:04:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][1200/1251] eta 0:00:14 lr 0.000013 time 0.2731 (0.2940) loss 3.0954 (2.9583) grad_norm 3.4677 (3.4537) [2021-04-16 21:04:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][1210/1251] eta 0:00:12 lr 0.000013 time 0.2710 (0.2939) loss 3.2011 (2.9587) grad_norm 3.4049 (3.4523) [2021-04-16 21:04:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][1220/1251] eta 0:00:09 lr 0.000013 time 0.3070 (0.2939) loss 3.4417 (2.9603) grad_norm 3.1796 (3.4517) [2021-04-16 21:04:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][1230/1251] eta 0:00:06 lr 0.000013 time 0.2782 (0.2937) loss 2.7747 (2.9600) grad_norm 3.2532 (3.4513) [2021-04-16 21:04:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][1240/1251] eta 0:00:03 lr 0.000013 time 0.2480 (0.2935) loss 3.5294 (2.9620) grad_norm 3.3801 (3.4524) [2021-04-16 21:05:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [289/300][1250/1251] eta 0:00:00 lr 0.000013 time 0.2495 (0.2931) loss 2.6308 (2.9622) grad_norm 3.0292 (3.4513) [2021-04-16 21:05:05 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 289 training takes 0:06:11 [2021-04-16 21:05:05 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_289.pth saving...... [2021-04-16 21:05:19 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_289.pth saved !!! [2021-04-16 21:05:20 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.076 (1.076) Loss 0.8288 (0.8288) Acc@1 80.566 (80.566) Acc@5 95.801 (95.801) [2021-04-16 21:05:22 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.269 (0.283) Loss 0.7692 (0.8053) Acc@1 83.008 (81.392) Acc@5 95.898 (95.694) [2021-04-16 21:05:24 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.159 (0.225) Loss 0.7786 (0.8055) Acc@1 83.301 (81.501) Acc@5 95.410 (95.587) [2021-04-16 21:05:25 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.122 (0.206) Loss 0.8588 (0.8170) Acc@1 80.078 (81.137) Acc@5 94.824 (95.571) [2021-04-16 21:05:27 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.103 (0.211) Loss 0.7599 (0.8151) Acc@1 83.398 (81.219) Acc@5 95.898 (95.486) [2021-04-16 21:05:44 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 81.238 Acc@5 95.524 [2021-04-16 21:05:44 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 81.2% [2021-04-16 21:05:44 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 81.24% [2021-04-16 21:05:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][0/1251] eta 2:46:24 lr 0.000013 time 7.9809 (7.9809) loss 3.5193 (3.5193) grad_norm 3.2540 (3.2540) [2021-04-16 21:05:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][10/1251] eta 0:20:12 lr 0.000013 time 0.2764 (0.9768) loss 2.8307 (3.1561) grad_norm 3.0702 (3.3222) [2021-04-16 21:05:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][20/1251] eta 0:13:12 lr 0.000013 time 0.2570 (0.6436) loss 3.1918 (3.0363) grad_norm 3.5543 (3.6592) [2021-04-16 21:06:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][30/1251] eta 0:10:44 lr 0.000013 time 0.2772 (0.5277) loss 3.1345 (3.0705) grad_norm 2.7187 (3.5158) [2021-04-16 21:06:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][40/1251] eta 0:09:25 lr 0.000013 time 0.2840 (0.4672) loss 2.5754 (3.0535) grad_norm 3.0781 (3.5100) [2021-04-16 21:06:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][50/1251] eta 0:08:36 lr 0.000013 time 0.2939 (0.4299) loss 3.0386 (3.0639) grad_norm 2.7262 (3.5116) [2021-04-16 21:06:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][60/1251] eta 0:08:03 lr 0.000013 time 0.2849 (0.4057) loss 2.9510 (3.0698) grad_norm 3.9565 (3.5298) [2021-04-16 21:06:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][70/1251] eta 0:07:37 lr 0.000013 time 0.2590 (0.3870) loss 2.2455 (3.0704) grad_norm 3.2367 (3.5044) [2021-04-16 21:06:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][80/1251] eta 0:07:18 lr 0.000013 time 0.2983 (0.3744) loss 3.3139 (3.0414) grad_norm 3.5864 (3.5384) [2021-04-16 21:06:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][90/1251] eta 0:07:03 lr 0.000013 time 0.2661 (0.3651) loss 3.1358 (3.0392) grad_norm 3.4901 (3.5265) [2021-04-16 21:06:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][100/1251] eta 0:06:50 lr 0.000013 time 0.2848 (0.3563) loss 2.9908 (3.0493) grad_norm 3.3874 (3.5289) [2021-04-16 21:06:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][110/1251] eta 0:06:38 lr 0.000013 time 0.2910 (0.3493) loss 2.8503 (3.0463) grad_norm 2.9223 (3.5450) [2021-04-16 21:06:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][120/1251] eta 0:06:28 lr 0.000013 time 0.2486 (0.3437) loss 3.3326 (3.0353) grad_norm 2.9955 (3.5436) [2021-04-16 21:06:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][130/1251] eta 0:06:19 lr 0.000013 time 0.2711 (0.3389) loss 2.4466 (3.0225) grad_norm 4.8976 (3.5963) [2021-04-16 21:06:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][140/1251] eta 0:06:11 lr 0.000013 time 0.2644 (0.3343) loss 3.0141 (2.9990) grad_norm 3.6724 (3.6182) [2021-04-16 21:06:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][150/1251] eta 0:06:04 lr 0.000013 time 0.2774 (0.3309) loss 3.3390 (2.9854) grad_norm 9.2310 (3.6672) [2021-04-16 21:06:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][160/1251] eta 0:05:58 lr 0.000013 time 0.3065 (0.3285) loss 3.4122 (2.9789) grad_norm 2.8231 (3.6466) [2021-04-16 21:06:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][170/1251] eta 0:05:51 lr 0.000013 time 0.2797 (0.3253) loss 3.6706 (2.9884) grad_norm 3.5599 (3.6407) [2021-04-16 21:06:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][180/1251] eta 0:05:45 lr 0.000013 time 0.2602 (0.3225) loss 2.6229 (2.9871) grad_norm 4.1254 (3.6248) [2021-04-16 21:06:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][190/1251] eta 0:05:39 lr 0.000013 time 0.2570 (0.3201) loss 2.2213 (2.9815) grad_norm 3.1965 (3.6082) [2021-04-16 21:06:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][200/1251] eta 0:05:34 lr 0.000013 time 0.2942 (0.3182) loss 2.7642 (2.9815) grad_norm 4.2316 (3.6058) [2021-04-16 21:06:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][210/1251] eta 0:05:29 lr 0.000013 time 0.2679 (0.3165) loss 2.5426 (2.9771) grad_norm 3.1696 (3.5934) [2021-04-16 21:06:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][220/1251] eta 0:05:24 lr 0.000013 time 0.2538 (0.3147) loss 2.8933 (2.9764) grad_norm 3.3103 (3.5830) [2021-04-16 21:06:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][230/1251] eta 0:05:19 lr 0.000013 time 0.2588 (0.3134) loss 2.8472 (2.9671) grad_norm 3.6148 (3.5733) [2021-04-16 21:06:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][240/1251] eta 0:05:15 lr 0.000013 time 0.2715 (0.3117) loss 3.2807 (2.9655) grad_norm 3.7823 (3.5797) [2021-04-16 21:07:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][250/1251] eta 0:05:10 lr 0.000013 time 0.2780 (0.3106) loss 3.3798 (2.9700) grad_norm 4.2960 (3.5759) [2021-04-16 21:07:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][260/1251] eta 0:05:06 lr 0.000013 time 0.2888 (0.3092) loss 2.2204 (2.9642) grad_norm 3.8252 (3.5706) [2021-04-16 21:07:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][270/1251] eta 0:05:01 lr 0.000013 time 0.2789 (0.3077) loss 3.3565 (2.9726) grad_norm 3.1922 (3.5665) [2021-04-16 21:07:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][280/1251] eta 0:04:57 lr 0.000013 time 0.2693 (0.3068) loss 2.1833 (2.9685) grad_norm 3.0858 (3.5659) [2021-04-16 21:07:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][290/1251] eta 0:04:53 lr 0.000013 time 0.2774 (0.3058) loss 3.0530 (2.9669) grad_norm 3.3092 (3.5618) [2021-04-16 21:07:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][300/1251] eta 0:04:49 lr 0.000013 time 0.2585 (0.3048) loss 2.8609 (2.9619) grad_norm 3.4939 (3.5507) [2021-04-16 21:07:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][310/1251] eta 0:04:45 lr 0.000013 time 0.2731 (0.3037) loss 2.9796 (2.9576) grad_norm 3.4853 (3.5500) [2021-04-16 21:07:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][320/1251] eta 0:04:42 lr 0.000013 time 0.3994 (0.3039) loss 1.8935 (2.9517) grad_norm 2.8068 (3.5445) [2021-04-16 21:07:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][330/1251] eta 0:04:39 lr 0.000013 time 0.2791 (0.3032) loss 3.6314 (2.9545) grad_norm 3.1583 (3.5459) [2021-04-16 21:07:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][340/1251] eta 0:04:35 lr 0.000013 time 0.2861 (0.3026) loss 2.7899 (2.9552) grad_norm 3.5932 (3.5402) [2021-04-16 21:07:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][350/1251] eta 0:04:32 lr 0.000013 time 0.2519 (0.3020) loss 2.6884 (2.9571) grad_norm 3.2720 (3.5408) [2021-04-16 21:07:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][360/1251] eta 0:04:28 lr 0.000013 time 0.2656 (0.3019) loss 2.3464 (2.9558) grad_norm 3.6943 (3.5424) [2021-04-16 21:07:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][370/1251] eta 0:04:25 lr 0.000013 time 0.2986 (0.3016) loss 3.3239 (2.9625) grad_norm 3.6361 (3.5404) [2021-04-16 21:07:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][380/1251] eta 0:04:22 lr 0.000013 time 0.2721 (0.3010) loss 2.5562 (2.9630) grad_norm 3.0244 (3.5371) [2021-04-16 21:07:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][390/1251] eta 0:04:18 lr 0.000013 time 0.2570 (0.3004) loss 2.7705 (2.9648) grad_norm 3.3790 (3.5288) [2021-04-16 21:07:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][400/1251] eta 0:04:15 lr 0.000013 time 0.2801 (0.2999) loss 3.0351 (2.9659) grad_norm 3.2210 (3.5268) [2021-04-16 21:07:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][410/1251] eta 0:04:11 lr 0.000013 time 0.2886 (0.2992) loss 3.1325 (2.9648) grad_norm 3.2697 (3.5278) [2021-04-16 21:07:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][420/1251] eta 0:04:08 lr 0.000013 time 0.2596 (0.2987) loss 3.2073 (2.9691) grad_norm 3.4491 (3.5214) [2021-04-16 21:07:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][430/1251] eta 0:04:04 lr 0.000013 time 0.2792 (0.2981) loss 3.3622 (2.9649) grad_norm 3.4769 (3.5252) [2021-04-16 21:07:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][440/1251] eta 0:04:01 lr 0.000013 time 0.2843 (0.2975) loss 1.8378 (2.9655) grad_norm 2.7045 (3.5187) [2021-04-16 21:07:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][450/1251] eta 0:03:57 lr 0.000013 time 0.2745 (0.2970) loss 3.4732 (2.9689) grad_norm 2.8538 (3.5120) [2021-04-16 21:08:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][460/1251] eta 0:03:54 lr 0.000013 time 0.2992 (0.2965) loss 2.5211 (2.9702) grad_norm 4.0725 (3.5159) [2021-04-16 21:08:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][470/1251] eta 0:03:51 lr 0.000013 time 0.2824 (0.2961) loss 2.6043 (2.9673) grad_norm 3.6499 (3.5088) [2021-04-16 21:08:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][480/1251] eta 0:03:47 lr 0.000013 time 0.2545 (0.2957) loss 3.1460 (2.9681) grad_norm 3.4127 (3.5070) [2021-04-16 21:08:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][490/1251] eta 0:03:44 lr 0.000013 time 0.3059 (0.2953) loss 3.5123 (2.9660) grad_norm 3.6322 (3.5027) [2021-04-16 21:08:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][500/1251] eta 0:03:41 lr 0.000012 time 0.2865 (0.2949) loss 2.0307 (2.9579) grad_norm 3.7529 (3.4990) [2021-04-16 21:08:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][510/1251] eta 0:03:38 lr 0.000012 time 0.2778 (0.2946) loss 3.1654 (2.9617) grad_norm 4.9865 (3.5009) [2021-04-16 21:08:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][520/1251] eta 0:03:35 lr 0.000012 time 0.3032 (0.2943) loss 3.6218 (2.9618) grad_norm 3.6230 (3.5044) [2021-04-16 21:08:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][530/1251] eta 0:03:31 lr 0.000012 time 0.2863 (0.2940) loss 3.5929 (2.9619) grad_norm 3.2500 (3.5039) [2021-04-16 21:08:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][540/1251] eta 0:03:28 lr 0.000012 time 0.2685 (0.2939) loss 3.2677 (2.9620) grad_norm 3.1926 (3.5029) [2021-04-16 21:08:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][550/1251] eta 0:03:26 lr 0.000012 time 0.2822 (0.2940) loss 3.0277 (2.9610) grad_norm 4.1575 (3.5047) [2021-04-16 21:08:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][560/1251] eta 0:03:22 lr 0.000012 time 0.2750 (0.2937) loss 3.1544 (2.9632) grad_norm 3.5213 (3.5024) [2021-04-16 21:08:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][570/1251] eta 0:03:19 lr 0.000012 time 0.2986 (0.2936) loss 3.1711 (2.9655) grad_norm 3.6474 (3.5015) [2021-04-16 21:08:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][580/1251] eta 0:03:17 lr 0.000012 time 0.2819 (0.2938) loss 2.3806 (2.9655) grad_norm 3.6597 (3.5000) [2021-04-16 21:08:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][590/1251] eta 0:03:13 lr 0.000012 time 0.2742 (0.2935) loss 2.2239 (2.9682) grad_norm 3.4511 (3.5015) [2021-04-16 21:08:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][600/1251] eta 0:03:10 lr 0.000012 time 0.2928 (0.2932) loss 3.0149 (2.9654) grad_norm 3.9706 (3.5003) [2021-04-16 21:08:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][610/1251] eta 0:03:07 lr 0.000012 time 0.2930 (0.2932) loss 3.2673 (2.9633) grad_norm 3.1730 (3.5059) [2021-04-16 21:08:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][620/1251] eta 0:03:04 lr 0.000012 time 0.2823 (0.2930) loss 2.0404 (2.9627) grad_norm 4.0623 (3.5068) [2021-04-16 21:08:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][630/1251] eta 0:03:01 lr 0.000012 time 0.2427 (0.2926) loss 3.4133 (2.9624) grad_norm 3.6138 (3.5069) [2021-04-16 21:08:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][640/1251] eta 0:02:58 lr 0.000012 time 0.3249 (0.2924) loss 2.4306 (2.9615) grad_norm 5.5897 (3.5127) [2021-04-16 21:08:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][650/1251] eta 0:02:55 lr 0.000012 time 0.3030 (0.2922) loss 3.3491 (2.9639) grad_norm 6.2768 (3.5161) [2021-04-16 21:08:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][660/1251] eta 0:02:52 lr 0.000012 time 0.2607 (0.2919) loss 3.4843 (2.9606) grad_norm 3.4983 (3.5128) [2021-04-16 21:08:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][670/1251] eta 0:02:49 lr 0.000012 time 0.2890 (0.2919) loss 2.2845 (2.9596) grad_norm 3.4692 (3.5151) [2021-04-16 21:09:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][680/1251] eta 0:02:46 lr 0.000012 time 0.3091 (0.2917) loss 2.1768 (2.9607) grad_norm 3.6443 (3.5184) [2021-04-16 21:09:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][690/1251] eta 0:02:43 lr 0.000012 time 0.2806 (0.2914) loss 2.0390 (2.9589) grad_norm 3.3418 (3.5260) [2021-04-16 21:09:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][700/1251] eta 0:02:40 lr 0.000012 time 0.3202 (0.2913) loss 2.1590 (2.9600) grad_norm 3.5995 (3.5239) [2021-04-16 21:09:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][710/1251] eta 0:02:37 lr 0.000012 time 0.3040 (0.2911) loss 3.0646 (2.9612) grad_norm 3.2441 (3.5261) [2021-04-16 21:09:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][720/1251] eta 0:02:34 lr 0.000012 time 0.3115 (0.2910) loss 3.3091 (2.9641) grad_norm 4.4797 (3.5253) [2021-04-16 21:09:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][730/1251] eta 0:02:31 lr 0.000012 time 0.2887 (0.2908) loss 3.5283 (2.9645) grad_norm 3.9726 (3.5316) [2021-04-16 21:09:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][740/1251] eta 0:02:28 lr 0.000012 time 0.2906 (0.2907) loss 2.9419 (2.9670) grad_norm 3.2035 (3.5292) [2021-04-16 21:09:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][750/1251] eta 0:02:25 lr 0.000012 time 0.2829 (0.2905) loss 2.2962 (2.9676) grad_norm 3.7591 (3.5282) [2021-04-16 21:09:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][760/1251] eta 0:02:22 lr 0.000012 time 0.2947 (0.2904) loss 2.8240 (2.9647) grad_norm 3.5351 (3.5275) [2021-04-16 21:09:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][770/1251] eta 0:02:19 lr 0.000012 time 0.2715 (0.2901) loss 3.1011 (2.9686) grad_norm 3.2848 (3.5258) [2021-04-16 21:09:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][780/1251] eta 0:02:16 lr 0.000012 time 0.2815 (0.2900) loss 3.0485 (2.9689) grad_norm 3.0748 (3.5316) [2021-04-16 21:09:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][790/1251] eta 0:02:13 lr 0.000012 time 0.3056 (0.2898) loss 3.1021 (2.9681) grad_norm 3.0830 (3.5279) [2021-04-16 21:09:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][800/1251] eta 0:02:10 lr 0.000012 time 0.2898 (0.2896) loss 2.2140 (2.9683) grad_norm 3.3300 (3.5243) [2021-04-16 21:09:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][810/1251] eta 0:02:07 lr 0.000012 time 0.2839 (0.2895) loss 2.3575 (2.9700) grad_norm 3.2888 (3.5251) [2021-04-16 21:09:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][820/1251] eta 0:02:04 lr 0.000012 time 0.2740 (0.2894) loss 3.4149 (2.9691) grad_norm 3.1872 (3.5215) [2021-04-16 21:09:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][830/1251] eta 0:02:01 lr 0.000012 time 0.2744 (0.2893) loss 3.4990 (2.9688) grad_norm 3.4330 (3.5195) [2021-04-16 21:09:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][840/1251] eta 0:01:58 lr 0.000012 time 0.3095 (0.2891) loss 2.5259 (2.9685) grad_norm 3.5165 (3.5178) [2021-04-16 21:09:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][850/1251] eta 0:01:55 lr 0.000012 time 0.2818 (0.2890) loss 3.4396 (2.9674) grad_norm 3.0986 (3.5162) [2021-04-16 21:09:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][860/1251] eta 0:01:52 lr 0.000012 time 0.2575 (0.2889) loss 3.2160 (2.9644) grad_norm 4.0486 (3.5187) [2021-04-16 21:09:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][870/1251] eta 0:01:50 lr 0.000012 time 0.2846 (0.2887) loss 3.0474 (2.9649) grad_norm 3.8588 (3.5198) [2021-04-16 21:09:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][880/1251] eta 0:01:47 lr 0.000012 time 0.2905 (0.2887) loss 3.3256 (2.9633) grad_norm 3.8369 (3.5191) [2021-04-16 21:10:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][890/1251] eta 0:01:44 lr 0.000012 time 0.2831 (0.2887) loss 3.0858 (2.9647) grad_norm 3.2556 (3.5171) [2021-04-16 21:10:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][900/1251] eta 0:01:41 lr 0.000012 time 0.3063 (0.2886) loss 3.5783 (2.9655) grad_norm 3.6445 (3.5170) [2021-04-16 21:10:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][910/1251] eta 0:01:38 lr 0.000012 time 0.2655 (0.2885) loss 3.0518 (2.9643) grad_norm 3.9998 (3.5164) [2021-04-16 21:10:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][920/1251] eta 0:01:35 lr 0.000012 time 0.3002 (0.2886) loss 3.2987 (2.9649) grad_norm 3.8572 (3.5183) [2021-04-16 21:10:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][930/1251] eta 0:01:32 lr 0.000012 time 0.2817 (0.2884) loss 2.4217 (2.9663) grad_norm 3.7216 (3.5197) [2021-04-16 21:10:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][940/1251] eta 0:01:29 lr 0.000012 time 0.2660 (0.2886) loss 3.3785 (2.9690) grad_norm 4.2421 (3.5183) [2021-04-16 21:10:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][950/1251] eta 0:01:26 lr 0.000012 time 0.2898 (0.2885) loss 2.4663 (2.9703) grad_norm 3.3071 (3.5156) [2021-04-16 21:10:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][960/1251] eta 0:01:23 lr 0.000012 time 0.2705 (0.2884) loss 2.8432 (2.9720) grad_norm 3.7756 (3.5131) [2021-04-16 21:10:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][970/1251] eta 0:01:21 lr 0.000012 time 0.2748 (0.2883) loss 2.9742 (2.9698) grad_norm 3.3580 (3.5127) [2021-04-16 21:10:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][980/1251] eta 0:01:18 lr 0.000012 time 0.2768 (0.2883) loss 3.1923 (2.9701) grad_norm 3.2097 (3.5127) [2021-04-16 21:10:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][990/1251] eta 0:01:15 lr 0.000012 time 0.2591 (0.2881) loss 3.2548 (2.9708) grad_norm 3.5696 (3.5113) [2021-04-16 21:10:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][1000/1251] eta 0:01:12 lr 0.000012 time 0.2548 (0.2880) loss 2.6028 (2.9701) grad_norm 3.5820 (3.5118) [2021-04-16 21:10:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][1010/1251] eta 0:01:09 lr 0.000012 time 0.2681 (0.2878) loss 3.2245 (2.9698) grad_norm 2.9955 (3.5112) [2021-04-16 21:10:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][1020/1251] eta 0:01:06 lr 0.000012 time 0.2898 (0.2878) loss 3.6977 (2.9708) grad_norm 3.4944 (3.5110) [2021-04-16 21:10:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][1030/1251] eta 0:01:03 lr 0.000012 time 0.2878 (0.2877) loss 3.7294 (2.9708) grad_norm 3.3264 (3.5094) [2021-04-16 21:10:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][1040/1251] eta 0:01:00 lr 0.000012 time 0.2749 (0.2876) loss 2.6960 (2.9721) grad_norm 3.6645 (3.5106) [2021-04-16 21:10:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][1050/1251] eta 0:00:57 lr 0.000012 time 0.2708 (0.2875) loss 2.9763 (2.9698) grad_norm 3.4034 (3.5094) [2021-04-16 21:10:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][1060/1251] eta 0:00:54 lr 0.000012 time 0.2721 (0.2874) loss 2.9342 (2.9698) grad_norm 3.8962 (3.5097) [2021-04-16 21:10:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][1070/1251] eta 0:00:51 lr 0.000012 time 0.2516 (0.2872) loss 3.1723 (2.9704) grad_norm 3.2073 (3.5120) [2021-04-16 21:10:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][1080/1251] eta 0:00:49 lr 0.000012 time 0.3041 (0.2872) loss 2.0290 (2.9694) grad_norm 2.8666 (3.5102) [2021-04-16 21:10:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][1090/1251] eta 0:00:46 lr 0.000012 time 0.2645 (0.2871) loss 3.3420 (2.9721) grad_norm 2.7379 (3.5074) [2021-04-16 21:11:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][1100/1251] eta 0:00:43 lr 0.000012 time 0.2704 (0.2870) loss 3.2806 (2.9742) grad_norm 2.8840 (3.5069) [2021-04-16 21:11:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][1110/1251] eta 0:00:40 lr 0.000012 time 0.3067 (0.2869) loss 2.9114 (2.9740) grad_norm 3.4873 (3.5057) [2021-04-16 21:11:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][1120/1251] eta 0:00:37 lr 0.000012 time 0.2947 (0.2870) loss 2.2608 (2.9731) grad_norm 3.7315 (3.5062) [2021-04-16 21:11:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][1130/1251] eta 0:00:34 lr 0.000012 time 0.2651 (0.2869) loss 2.4378 (2.9718) grad_norm 2.9531 (3.5064) [2021-04-16 21:11:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][1140/1251] eta 0:00:31 lr 0.000012 time 0.2671 (0.2869) loss 3.1566 (2.9696) grad_norm 3.3082 (3.5073) [2021-04-16 21:11:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][1150/1251] eta 0:00:28 lr 0.000012 time 0.2798 (0.2868) loss 2.3019 (2.9716) grad_norm 3.1787 (3.5087) [2021-04-16 21:11:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][1160/1251] eta 0:00:26 lr 0.000012 time 0.2730 (0.2870) loss 2.9896 (2.9717) grad_norm 3.3096 (3.5159) [2021-04-16 21:11:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][1170/1251] eta 0:00:23 lr 0.000012 time 0.2795 (0.2869) loss 2.8930 (2.9709) grad_norm 3.5205 (3.5148) [2021-04-16 21:11:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][1180/1251] eta 0:00:20 lr 0.000012 time 0.2627 (0.2869) loss 3.1411 (2.9733) grad_norm 5.0578 (3.5148) [2021-04-16 21:11:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][1190/1251] eta 0:00:17 lr 0.000012 time 0.2766 (0.2868) loss 2.7006 (2.9732) grad_norm 2.9142 (3.5145) [2021-04-16 21:11:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][1200/1251] eta 0:00:14 lr 0.000012 time 0.2802 (0.2867) loss 2.6382 (2.9722) grad_norm 3.7563 (3.5128) [2021-04-16 21:11:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][1210/1251] eta 0:00:11 lr 0.000012 time 0.2825 (0.2866) loss 3.1845 (2.9715) grad_norm 3.0233 (3.5115) [2021-04-16 21:11:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][1220/1251] eta 0:00:08 lr 0.000012 time 0.2785 (0.2865) loss 2.5534 (2.9693) grad_norm 4.6920 (3.5124) [2021-04-16 21:11:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][1230/1251] eta 0:00:06 lr 0.000012 time 0.2904 (0.2865) loss 2.9105 (2.9690) grad_norm 3.0200 (3.5114) [2021-04-16 21:11:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][1240/1251] eta 0:00:03 lr 0.000012 time 0.2478 (0.2864) loss 3.5130 (2.9691) grad_norm 3.6650 (3.5127) [2021-04-16 21:11:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [290/300][1250/1251] eta 0:00:00 lr 0.000012 time 0.2479 (0.2861) loss 3.2211 (2.9681) grad_norm 3.0579 (3.5130) [2021-04-16 21:11:48 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 290 training takes 0:06:04 [2021-04-16 21:11:48 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_290.pth saving...... [2021-04-16 21:12:00 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_290.pth saved !!! [2021-04-16 21:12:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.211 (1.211) Loss 0.7881 (0.7881) Acc@1 82.031 (82.031) Acc@5 95.801 (95.801) [2021-04-16 21:12:03 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.125 (0.220) Loss 0.8345 (0.8214) Acc@1 80.566 (81.312) Acc@5 95.801 (95.526) [2021-04-16 21:12:05 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.110 (0.234) Loss 0.8357 (0.8244) Acc@1 80.566 (81.083) Acc@5 95.410 (95.433) [2021-04-16 21:12:07 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.103 (0.221) Loss 0.8235 (0.8227) Acc@1 81.055 (81.178) Acc@5 94.824 (95.410) [2021-04-16 21:12:09 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.206) Loss 0.8164 (0.8153) Acc@1 81.250 (81.243) Acc@5 95.605 (95.470) [2021-04-16 21:12:19 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 81.150 Acc@5 95.502 [2021-04-16 21:12:19 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 81.2% [2021-04-16 21:12:19 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 81.24% [2021-04-16 21:12:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][0/1251] eta 4:42:48 lr 0.000012 time 13.5640 (13.5640) loss 3.5092 (3.5092) grad_norm 2.9877 (2.9877) [2021-04-16 21:12:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][10/1251] eta 0:30:37 lr 0.000012 time 0.2818 (1.4805) loss 3.3405 (2.8510) grad_norm 4.0913 (3.3464) [2021-04-16 21:12:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][20/1251] eta 0:18:36 lr 0.000012 time 0.2555 (0.9067) loss 3.3509 (2.8799) grad_norm 4.2460 (3.5296) [2021-04-16 21:12:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][30/1251] eta 0:14:18 lr 0.000012 time 0.2633 (0.7033) loss 3.1629 (2.9735) grad_norm 3.3764 (3.5568) [2021-04-16 21:12:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][40/1251] eta 0:12:05 lr 0.000012 time 0.3023 (0.5995) loss 3.6192 (3.0283) grad_norm 4.8508 (3.6223) [2021-04-16 21:12:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][50/1251] eta 0:10:44 lr 0.000012 time 0.2704 (0.5365) loss 3.2342 (3.0261) grad_norm 2.9039 (3.5295) [2021-04-16 21:12:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][60/1251] eta 0:09:48 lr 0.000012 time 0.2935 (0.4941) loss 2.3631 (2.9996) grad_norm 4.1030 (3.5384) [2021-04-16 21:12:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][70/1251] eta 0:09:06 lr 0.000012 time 0.2625 (0.4628) loss 2.9016 (3.0033) grad_norm 3.0806 (3.5201) [2021-04-16 21:12:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][80/1251] eta 0:08:35 lr 0.000012 time 0.2864 (0.4402) loss 2.5765 (3.0153) grad_norm 3.8244 (3.5170) [2021-04-16 21:12:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][90/1251] eta 0:08:10 lr 0.000012 time 0.2666 (0.4223) loss 3.6402 (3.0125) grad_norm 4.2924 (3.5394) [2021-04-16 21:13:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][100/1251] eta 0:07:50 lr 0.000012 time 0.2746 (0.4087) loss 3.0422 (2.9916) grad_norm 3.8984 (inf) [2021-04-16 21:13:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][110/1251] eta 0:07:32 lr 0.000012 time 0.2699 (0.3966) loss 2.6279 (2.9810) grad_norm 4.6804 (inf) [2021-04-16 21:13:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][120/1251] eta 0:07:18 lr 0.000012 time 0.2789 (0.3880) loss 3.2015 (2.9627) grad_norm 3.0211 (inf) [2021-04-16 21:13:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][130/1251] eta 0:07:05 lr 0.000012 time 0.2836 (0.3794) loss 3.3644 (2.9732) grad_norm 3.2566 (inf) [2021-04-16 21:13:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][140/1251] eta 0:06:54 lr 0.000012 time 0.2621 (0.3732) loss 3.3852 (2.9873) grad_norm 2.8719 (inf) [2021-04-16 21:13:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][150/1251] eta 0:06:45 lr 0.000012 time 0.2703 (0.3682) loss 2.1547 (2.9777) grad_norm 3.0255 (inf) [2021-04-16 21:13:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][160/1251] eta 0:06:35 lr 0.000012 time 0.2845 (0.3624) loss 3.1155 (2.9899) grad_norm 3.0716 (inf) [2021-04-16 21:13:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][170/1251] eta 0:06:26 lr 0.000012 time 0.2626 (0.3572) loss 3.0057 (2.9877) grad_norm 3.1074 (inf) [2021-04-16 21:13:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][180/1251] eta 0:06:17 lr 0.000012 time 0.2710 (0.3527) loss 2.5971 (2.9791) grad_norm 3.6478 (inf) [2021-04-16 21:13:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][190/1251] eta 0:06:10 lr 0.000012 time 0.3027 (0.3488) loss 3.4083 (2.9786) grad_norm 3.4719 (inf) [2021-04-16 21:13:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][200/1251] eta 0:06:02 lr 0.000012 time 0.2709 (0.3453) loss 3.4767 (2.9911) grad_norm 3.3453 (inf) [2021-04-16 21:13:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][210/1251] eta 0:05:56 lr 0.000012 time 0.2847 (0.3420) loss 3.0304 (2.9855) grad_norm 3.0285 (inf) [2021-04-16 21:13:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][220/1251] eta 0:05:49 lr 0.000012 time 0.2561 (0.3391) loss 3.2180 (2.9973) grad_norm 3.7334 (inf) [2021-04-16 21:13:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][230/1251] eta 0:05:43 lr 0.000012 time 0.2656 (0.3363) loss 3.0986 (3.0099) grad_norm 3.2851 (inf) [2021-04-16 21:13:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][240/1251] eta 0:05:37 lr 0.000012 time 0.2915 (0.3338) loss 2.8456 (3.0064) grad_norm 3.7200 (inf) [2021-04-16 21:13:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][250/1251] eta 0:05:32 lr 0.000012 time 0.2844 (0.3317) loss 2.6405 (2.9987) grad_norm 3.1341 (inf) [2021-04-16 21:13:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][260/1251] eta 0:05:26 lr 0.000012 time 0.2796 (0.3298) loss 3.3680 (3.0096) grad_norm 3.0531 (inf) [2021-04-16 21:13:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][270/1251] eta 0:05:21 lr 0.000012 time 0.2919 (0.3280) loss 2.5627 (2.9986) grad_norm 3.1723 (inf) [2021-04-16 21:13:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][280/1251] eta 0:05:16 lr 0.000012 time 0.2693 (0.3260) loss 1.9730 (2.9888) grad_norm 3.1188 (inf) [2021-04-16 21:13:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][290/1251] eta 0:05:11 lr 0.000012 time 0.3031 (0.3243) loss 1.9110 (2.9769) grad_norm 5.0149 (inf) [2021-04-16 21:13:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][300/1251] eta 0:05:07 lr 0.000012 time 0.2926 (0.3229) loss 3.7219 (2.9804) grad_norm 4.3194 (inf) [2021-04-16 21:13:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][310/1251] eta 0:05:02 lr 0.000012 time 0.2598 (0.3211) loss 1.7567 (2.9841) grad_norm 3.1997 (inf) [2021-04-16 21:14:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][320/1251] eta 0:04:57 lr 0.000012 time 0.2569 (0.3200) loss 2.9611 (2.9842) grad_norm 3.2341 (inf) [2021-04-16 21:14:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][330/1251] eta 0:04:53 lr 0.000012 time 0.2864 (0.3187) loss 3.2322 (2.9885) grad_norm 3.3280 (inf) [2021-04-16 21:14:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][340/1251] eta 0:04:49 lr 0.000012 time 0.2721 (0.3179) loss 1.9359 (2.9867) grad_norm 4.0796 (inf) [2021-04-16 21:14:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][350/1251] eta 0:04:45 lr 0.000012 time 0.2686 (0.3167) loss 3.2016 (2.9943) grad_norm 3.3700 (inf) [2021-04-16 21:14:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][360/1251] eta 0:04:42 lr 0.000012 time 0.2754 (0.3165) loss 2.6348 (2.9948) grad_norm 3.2739 (inf) [2021-04-16 21:14:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][370/1251] eta 0:04:37 lr 0.000012 time 0.2571 (0.3155) loss 1.7068 (2.9930) grad_norm 3.6196 (inf) [2021-04-16 21:14:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][380/1251] eta 0:04:34 lr 0.000012 time 0.2947 (0.3149) loss 3.3593 (2.9989) grad_norm 3.7853 (inf) [2021-04-16 21:14:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][390/1251] eta 0:04:30 lr 0.000012 time 0.2863 (0.3139) loss 3.2287 (2.9927) grad_norm 2.9413 (inf) [2021-04-16 21:14:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][400/1251] eta 0:04:26 lr 0.000012 time 0.2750 (0.3130) loss 3.5959 (2.9941) grad_norm 3.5351 (inf) [2021-04-16 21:14:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][410/1251] eta 0:04:22 lr 0.000012 time 0.3044 (0.3121) loss 3.2344 (2.9945) grad_norm 5.0514 (inf) [2021-04-16 21:14:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][420/1251] eta 0:04:18 lr 0.000012 time 0.2527 (0.3111) loss 2.1936 (2.9889) grad_norm 4.7991 (inf) [2021-04-16 21:14:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][430/1251] eta 0:04:14 lr 0.000012 time 0.2908 (0.3103) loss 3.2177 (2.9896) grad_norm 3.2758 (inf) [2021-04-16 21:14:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][440/1251] eta 0:04:11 lr 0.000012 time 0.2879 (0.3099) loss 3.3686 (2.9929) grad_norm 5.4416 (inf) [2021-04-16 21:14:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][450/1251] eta 0:04:07 lr 0.000012 time 0.3155 (0.3094) loss 2.9846 (2.9951) grad_norm 4.2843 (inf) [2021-04-16 21:14:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][460/1251] eta 0:04:04 lr 0.000012 time 0.2648 (0.3086) loss 3.3092 (2.9945) grad_norm 3.1224 (inf) [2021-04-16 21:14:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][470/1251] eta 0:04:00 lr 0.000012 time 0.3111 (0.3080) loss 1.7797 (2.9904) grad_norm 3.5353 (inf) [2021-04-16 21:14:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][480/1251] eta 0:03:57 lr 0.000012 time 0.2743 (0.3075) loss 3.3201 (2.9928) grad_norm 3.8860 (inf) [2021-04-16 21:14:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][490/1251] eta 0:03:53 lr 0.000012 time 0.2665 (0.3068) loss 3.1070 (2.9912) grad_norm 3.3983 (inf) [2021-04-16 21:14:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][500/1251] eta 0:03:49 lr 0.000012 time 0.2665 (0.3062) loss 3.3882 (2.9881) grad_norm 3.1565 (inf) [2021-04-16 21:14:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][510/1251] eta 0:03:46 lr 0.000012 time 0.2777 (0.3055) loss 3.6499 (2.9867) grad_norm 3.6154 (inf) [2021-04-16 21:14:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][520/1251] eta 0:03:42 lr 0.000012 time 0.2683 (0.3050) loss 3.3245 (2.9886) grad_norm 2.9449 (inf) [2021-04-16 21:15:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][530/1251] eta 0:03:39 lr 0.000012 time 0.2837 (0.3044) loss 3.7614 (2.9888) grad_norm 4.5103 (inf) [2021-04-16 21:15:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][540/1251] eta 0:03:36 lr 0.000012 time 0.2789 (0.3039) loss 3.4663 (2.9930) grad_norm 3.0107 (inf) [2021-04-16 21:15:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][550/1251] eta 0:03:32 lr 0.000012 time 0.2801 (0.3034) loss 2.7097 (2.9904) grad_norm 3.2043 (inf) [2021-04-16 21:15:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][560/1251] eta 0:03:29 lr 0.000012 time 0.2879 (0.3029) loss 2.9974 (2.9897) grad_norm 3.2559 (inf) [2021-04-16 21:15:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][570/1251] eta 0:03:26 lr 0.000012 time 0.2764 (0.3026) loss 3.3432 (2.9941) grad_norm 3.2790 (inf) [2021-04-16 21:15:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][580/1251] eta 0:03:22 lr 0.000012 time 0.2740 (0.3024) loss 3.3608 (2.9966) grad_norm 4.2230 (inf) [2021-04-16 21:15:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][590/1251] eta 0:03:19 lr 0.000012 time 0.2549 (0.3019) loss 2.5921 (2.9933) grad_norm 3.2434 (inf) [2021-04-16 21:15:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][600/1251] eta 0:03:16 lr 0.000012 time 0.2444 (0.3014) loss 2.8233 (2.9940) grad_norm 3.2675 (inf) [2021-04-16 21:15:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][610/1251] eta 0:03:12 lr 0.000012 time 0.2790 (0.3010) loss 2.1455 (2.9908) grad_norm 3.3137 (inf) [2021-04-16 21:15:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][620/1251] eta 0:03:09 lr 0.000012 time 0.2832 (0.3006) loss 2.7717 (2.9867) grad_norm 4.5263 (inf) [2021-04-16 21:15:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][630/1251] eta 0:03:06 lr 0.000012 time 0.2869 (0.3002) loss 1.7259 (2.9901) grad_norm 3.1059 (inf) [2021-04-16 21:15:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][640/1251] eta 0:03:03 lr 0.000012 time 0.2966 (0.2997) loss 2.6957 (2.9896) grad_norm 3.9828 (inf) [2021-04-16 21:15:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][650/1251] eta 0:02:59 lr 0.000012 time 0.2702 (0.2993) loss 3.5315 (2.9867) grad_norm 3.3561 (inf) [2021-04-16 21:15:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][660/1251] eta 0:02:56 lr 0.000012 time 0.2912 (0.2990) loss 3.1398 (2.9900) grad_norm 3.0672 (inf) [2021-04-16 21:15:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][670/1251] eta 0:02:53 lr 0.000012 time 0.3065 (0.2987) loss 2.7830 (2.9913) grad_norm 6.9875 (inf) [2021-04-16 21:15:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][680/1251] eta 0:02:50 lr 0.000012 time 0.2684 (0.2986) loss 1.9868 (2.9867) grad_norm 3.5185 (inf) [2021-04-16 21:15:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][690/1251] eta 0:02:47 lr 0.000012 time 0.2991 (0.2984) loss 2.2550 (2.9875) grad_norm 3.5728 (inf) [2021-04-16 21:15:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][700/1251] eta 0:02:44 lr 0.000012 time 0.2791 (0.2981) loss 2.5078 (2.9866) grad_norm 3.4503 (inf) [2021-04-16 21:15:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][710/1251] eta 0:02:41 lr 0.000012 time 0.2937 (0.2977) loss 3.3912 (2.9894) grad_norm 3.8998 (inf) [2021-04-16 21:15:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][720/1251] eta 0:02:38 lr 0.000012 time 0.2444 (0.2978) loss 3.0961 (2.9898) grad_norm 3.5409 (inf) [2021-04-16 21:15:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][730/1251] eta 0:02:35 lr 0.000012 time 0.2515 (0.2975) loss 2.4885 (2.9915) grad_norm 3.4477 (inf) [2021-04-16 21:16:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][740/1251] eta 0:02:31 lr 0.000012 time 0.3146 (0.2973) loss 3.6119 (2.9915) grad_norm 3.6910 (inf) [2021-04-16 21:16:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][750/1251] eta 0:02:28 lr 0.000012 time 0.2872 (0.2970) loss 2.7856 (2.9907) grad_norm 2.7058 (inf) [2021-04-16 21:16:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][760/1251] eta 0:02:25 lr 0.000012 time 0.2752 (0.2968) loss 3.3893 (2.9925) grad_norm 3.8417 (inf) [2021-04-16 21:16:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][770/1251] eta 0:02:22 lr 0.000012 time 0.2726 (0.2966) loss 3.0854 (2.9940) grad_norm 3.3408 (inf) [2021-04-16 21:16:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][780/1251] eta 0:02:19 lr 0.000012 time 0.3127 (0.2963) loss 3.2259 (2.9937) grad_norm 3.4906 (inf) [2021-04-16 21:16:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][790/1251] eta 0:02:16 lr 0.000012 time 0.2824 (0.2960) loss 3.5244 (2.9965) grad_norm 3.3474 (inf) [2021-04-16 21:16:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][800/1251] eta 0:02:13 lr 0.000012 time 0.2718 (0.2958) loss 1.7085 (2.9962) grad_norm 4.1419 (inf) [2021-04-16 21:16:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][810/1251] eta 0:02:10 lr 0.000012 time 0.2740 (0.2956) loss 2.6696 (2.9958) grad_norm 2.7445 (inf) [2021-04-16 21:16:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][820/1251] eta 0:02:07 lr 0.000012 time 0.2795 (0.2953) loss 1.9333 (2.9934) grad_norm 3.1017 (inf) [2021-04-16 21:16:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][830/1251] eta 0:02:04 lr 0.000012 time 0.2827 (0.2951) loss 3.1903 (2.9913) grad_norm 3.7578 (inf) [2021-04-16 21:16:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][840/1251] eta 0:02:01 lr 0.000012 time 0.2769 (0.2948) loss 2.9104 (2.9907) grad_norm 2.8860 (inf) [2021-04-16 21:16:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][850/1251] eta 0:01:58 lr 0.000012 time 0.2630 (0.2946) loss 3.3326 (2.9928) grad_norm 3.6492 (inf) [2021-04-16 21:16:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][860/1251] eta 0:01:55 lr 0.000012 time 0.2636 (0.2944) loss 3.0646 (2.9924) grad_norm 3.1582 (inf) [2021-04-16 21:16:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][870/1251] eta 0:01:52 lr 0.000012 time 0.2637 (0.2942) loss 3.0469 (2.9922) grad_norm 2.9884 (inf) [2021-04-16 21:16:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][880/1251] eta 0:01:49 lr 0.000012 time 0.2777 (0.2940) loss 2.8703 (2.9952) grad_norm 3.6872 (inf) [2021-04-16 21:16:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][890/1251] eta 0:01:46 lr 0.000012 time 0.2791 (0.2939) loss 1.9464 (2.9975) grad_norm 3.4334 (inf) [2021-04-16 21:16:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][900/1251] eta 0:01:43 lr 0.000012 time 0.2664 (0.2937) loss 2.9369 (2.9980) grad_norm 3.6825 (inf) [2021-04-16 21:16:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][910/1251] eta 0:01:40 lr 0.000012 time 0.2642 (0.2934) loss 3.1813 (2.9971) grad_norm 3.3907 (inf) [2021-04-16 21:16:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][920/1251] eta 0:01:37 lr 0.000012 time 0.2704 (0.2934) loss 3.2020 (2.9953) grad_norm 3.4439 (inf) [2021-04-16 21:16:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][930/1251] eta 0:01:34 lr 0.000012 time 0.2684 (0.2934) loss 3.1513 (2.9937) grad_norm 2.8723 (inf) [2021-04-16 21:16:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][940/1251] eta 0:01:31 lr 0.000012 time 0.2704 (0.2933) loss 2.9503 (2.9916) grad_norm 3.3466 (inf) [2021-04-16 21:16:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][950/1251] eta 0:01:28 lr 0.000012 time 0.2827 (0.2931) loss 2.9300 (2.9933) grad_norm 3.0589 (inf) [2021-04-16 21:17:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][960/1251] eta 0:01:25 lr 0.000012 time 0.3049 (0.2930) loss 3.2391 (2.9922) grad_norm 3.4171 (inf) [2021-04-16 21:17:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][970/1251] eta 0:01:22 lr 0.000012 time 0.2660 (0.2929) loss 3.3021 (2.9916) grad_norm 4.0835 (inf) [2021-04-16 21:17:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][980/1251] eta 0:01:19 lr 0.000012 time 0.2847 (0.2928) loss 2.9166 (2.9941) grad_norm 3.9134 (inf) [2021-04-16 21:17:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][990/1251] eta 0:01:16 lr 0.000012 time 0.2754 (0.2926) loss 2.9809 (2.9914) grad_norm 3.1685 (inf) [2021-04-16 21:17:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][1000/1251] eta 0:01:13 lr 0.000012 time 0.2850 (0.2924) loss 2.3716 (2.9921) grad_norm 3.1586 (inf) [2021-04-16 21:17:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][1010/1251] eta 0:01:10 lr 0.000012 time 0.2896 (0.2923) loss 2.8197 (2.9933) grad_norm 3.8237 (inf) [2021-04-16 21:17:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][1020/1251] eta 0:01:07 lr 0.000012 time 0.2791 (0.2923) loss 2.3811 (2.9927) grad_norm 3.1424 (inf) [2021-04-16 21:17:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][1030/1251] eta 0:01:04 lr 0.000012 time 0.2793 (0.2922) loss 2.0922 (2.9925) grad_norm 3.3174 (inf) [2021-04-16 21:17:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][1040/1251] eta 0:01:01 lr 0.000012 time 0.2864 (0.2921) loss 3.3970 (2.9921) grad_norm 2.9659 (inf) [2021-04-16 21:17:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][1050/1251] eta 0:00:58 lr 0.000012 time 0.2684 (0.2919) loss 3.2623 (2.9909) grad_norm 2.9455 (inf) [2021-04-16 21:17:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][1060/1251] eta 0:00:55 lr 0.000012 time 0.2649 (0.2917) loss 3.4832 (2.9920) grad_norm 3.0229 (inf) [2021-04-16 21:17:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][1070/1251] eta 0:00:52 lr 0.000012 time 0.2682 (0.2916) loss 2.2551 (2.9892) grad_norm 3.2937 (inf) [2021-04-16 21:17:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][1080/1251] eta 0:00:49 lr 0.000012 time 0.2828 (0.2915) loss 2.7750 (2.9892) grad_norm 3.1585 (inf) [2021-04-16 21:17:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][1090/1251] eta 0:00:46 lr 0.000012 time 0.2462 (0.2913) loss 3.1897 (2.9900) grad_norm 4.7011 (inf) [2021-04-16 21:17:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][1100/1251] eta 0:00:43 lr 0.000012 time 0.2625 (0.2912) loss 3.3557 (2.9899) grad_norm 3.1153 (inf) [2021-04-16 21:17:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][1110/1251] eta 0:00:41 lr 0.000012 time 0.2792 (0.2910) loss 3.4053 (2.9903) grad_norm 3.9206 (inf) [2021-04-16 21:17:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][1120/1251] eta 0:00:38 lr 0.000012 time 0.2867 (0.2909) loss 3.4095 (2.9921) grad_norm 3.4931 (inf) [2021-04-16 21:17:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][1130/1251] eta 0:00:35 lr 0.000012 time 0.2786 (0.2908) loss 2.2943 (2.9919) grad_norm 3.0866 (inf) [2021-04-16 21:17:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][1140/1251] eta 0:00:32 lr 0.000012 time 0.2955 (0.2906) loss 3.5285 (2.9928) grad_norm 3.9634 (inf) [2021-04-16 21:17:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][1150/1251] eta 0:00:29 lr 0.000012 time 0.2636 (0.2905) loss 2.6879 (2.9913) grad_norm 4.0072 (inf) [2021-04-16 21:17:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][1160/1251] eta 0:00:26 lr 0.000012 time 0.2567 (0.2905) loss 2.8295 (2.9904) grad_norm 3.0348 (inf) [2021-04-16 21:17:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][1170/1251] eta 0:00:23 lr 0.000012 time 0.2979 (0.2904) loss 2.5468 (2.9903) grad_norm 3.6334 (inf) [2021-04-16 21:18:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][1180/1251] eta 0:00:20 lr 0.000012 time 0.2870 (0.2903) loss 1.9170 (2.9886) grad_norm 3.2024 (inf) [2021-04-16 21:18:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][1190/1251] eta 0:00:17 lr 0.000012 time 0.2882 (0.2902) loss 2.7625 (2.9885) grad_norm 3.5872 (inf) [2021-04-16 21:18:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][1200/1251] eta 0:00:14 lr 0.000012 time 0.2796 (0.2901) loss 3.4476 (2.9879) grad_norm 3.1969 (inf) [2021-04-16 21:18:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][1210/1251] eta 0:00:11 lr 0.000012 time 0.2834 (0.2900) loss 3.5762 (2.9875) grad_norm 3.2752 (inf) [2021-04-16 21:18:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][1220/1251] eta 0:00:08 lr 0.000012 time 0.2577 (0.2898) loss 2.9587 (2.9892) grad_norm 3.2731 (inf) [2021-04-16 21:18:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][1230/1251] eta 0:00:06 lr 0.000012 time 0.2790 (0.2897) loss 2.6350 (2.9879) grad_norm 2.7379 (inf) [2021-04-16 21:18:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][1240/1251] eta 0:00:03 lr 0.000012 time 0.2478 (0.2895) loss 3.2783 (2.9869) grad_norm 3.4509 (inf) [2021-04-16 21:18:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [291/300][1250/1251] eta 0:00:00 lr 0.000012 time 0.2616 (0.2893) loss 2.0841 (2.9853) grad_norm 3.2399 (inf) [2021-04-16 21:18:34 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 291 training takes 0:06:14 [2021-04-16 21:18:34 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_291.pth saving...... [2021-04-16 21:18:59 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_291.pth saved !!! [2021-04-16 21:19:01 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.167 (1.167) Loss 0.8089 (0.8089) Acc@1 80.859 (80.859) Acc@5 95.605 (95.605) [2021-04-16 21:19:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.107 (0.212) Loss 0.7213 (0.8153) Acc@1 83.008 (81.303) Acc@5 96.973 (95.579) [2021-04-16 21:19:05 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.405 (0.240) Loss 0.7676 (0.8227) Acc@1 82.227 (81.110) Acc@5 95.215 (95.512) [2021-04-16 21:19:07 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.132 (0.228) Loss 0.7441 (0.8223) Acc@1 82.715 (81.196) Acc@5 95.605 (95.467) [2021-04-16 21:19:08 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.218) Loss 0.8490 (0.8181) Acc@1 80.176 (81.255) Acc@5 94.727 (95.491) [2021-04-16 21:19:28 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 81.188 Acc@5 95.524 [2021-04-16 21:19:28 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 81.2% [2021-04-16 21:19:28 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 81.24% [2021-04-16 21:19:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][0/1251] eta 2:51:27 lr 0.000012 time 8.2238 (8.2238) loss 2.3817 (2.3817) grad_norm 3.5612 (3.5612) [2021-04-16 21:19:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][10/1251] eta 0:20:37 lr 0.000012 time 0.2816 (0.9974) loss 3.4713 (2.8439) grad_norm 3.1388 (3.2712) [2021-04-16 21:19:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][20/1251] eta 0:13:22 lr 0.000012 time 0.2658 (0.6518) loss 3.1304 (2.9039) grad_norm 3.2657 (3.3506) [2021-04-16 21:19:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][30/1251] eta 0:10:48 lr 0.000012 time 0.2875 (0.5313) loss 2.9702 (2.8724) grad_norm 4.0619 (3.4805) [2021-04-16 21:19:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][40/1251] eta 0:09:29 lr 0.000012 time 0.2700 (0.4701) loss 2.3069 (2.8711) grad_norm 3.2542 (3.4535) [2021-04-16 21:19:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][50/1251] eta 0:08:39 lr 0.000012 time 0.2779 (0.4324) loss 2.8041 (2.8609) grad_norm 3.4339 (3.5267) [2021-04-16 21:19:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][60/1251] eta 0:08:07 lr 0.000012 time 0.3153 (0.4091) loss 3.1247 (2.8734) grad_norm 3.6049 (3.5341) [2021-04-16 21:19:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][70/1251] eta 0:07:42 lr 0.000012 time 0.2730 (0.3917) loss 3.3732 (2.9143) grad_norm 3.1400 (3.5438) [2021-04-16 21:19:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][80/1251] eta 0:07:22 lr 0.000012 time 0.2746 (0.3783) loss 3.1298 (2.9076) grad_norm 2.9841 (3.5190) [2021-04-16 21:20:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][90/1251] eta 0:07:08 lr 0.000012 time 0.2639 (0.3688) loss 3.6323 (2.9158) grad_norm 2.8354 (3.5050) [2021-04-16 21:20:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][100/1251] eta 0:06:53 lr 0.000012 time 0.2644 (0.3595) loss 3.1762 (2.9074) grad_norm 2.6897 (3.4813) [2021-04-16 21:20:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][110/1251] eta 0:06:42 lr 0.000012 time 0.2882 (0.3526) loss 2.4316 (2.9163) grad_norm 3.4142 (3.4948) [2021-04-16 21:20:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][120/1251] eta 0:06:32 lr 0.000012 time 0.2704 (0.3472) loss 2.3264 (2.9159) grad_norm 3.0671 (3.4718) [2021-04-16 21:20:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][130/1251] eta 0:06:23 lr 0.000012 time 0.2919 (0.3417) loss 3.4338 (2.9249) grad_norm 3.6412 (3.4621) [2021-04-16 21:20:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][140/1251] eta 0:06:16 lr 0.000012 time 0.2737 (0.3393) loss 3.0148 (2.9305) grad_norm 3.0693 (3.4533) [2021-04-16 21:20:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][150/1251] eta 0:06:10 lr 0.000012 time 0.2858 (0.3364) loss 1.8685 (2.9235) grad_norm 2.9171 (3.4472) [2021-04-16 21:20:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][160/1251] eta 0:06:02 lr 0.000012 time 0.2656 (0.3327) loss 3.3710 (2.9204) grad_norm 3.9312 (3.4457) [2021-04-16 21:20:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][170/1251] eta 0:05:56 lr 0.000012 time 0.2659 (0.3302) loss 2.9227 (2.9104) grad_norm 3.4974 (3.4544) [2021-04-16 21:20:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][180/1251] eta 0:05:50 lr 0.000012 time 0.3015 (0.3274) loss 2.9474 (2.9052) grad_norm 3.9929 (3.4497) [2021-04-16 21:20:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][190/1251] eta 0:05:44 lr 0.000012 time 0.2492 (0.3249) loss 2.1940 (2.8958) grad_norm 4.0913 (3.4467) [2021-04-16 21:20:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][200/1251] eta 0:05:38 lr 0.000012 time 0.2726 (0.3225) loss 2.1710 (2.8915) grad_norm 3.6270 (3.4480) [2021-04-16 21:20:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][210/1251] eta 0:05:33 lr 0.000012 time 0.2887 (0.3202) loss 3.4034 (2.8861) grad_norm 3.4235 (3.4450) [2021-04-16 21:20:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][220/1251] eta 0:05:28 lr 0.000012 time 0.2875 (0.3183) loss 2.7856 (2.8979) grad_norm 4.2939 (3.4541) [2021-04-16 21:20:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][230/1251] eta 0:05:23 lr 0.000012 time 0.2840 (0.3170) loss 2.9516 (2.9080) grad_norm 3.4971 (3.4622) [2021-04-16 21:20:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][240/1251] eta 0:05:18 lr 0.000012 time 0.2580 (0.3152) loss 2.9131 (2.9173) grad_norm 4.4204 (3.4690) [2021-04-16 21:20:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][250/1251] eta 0:05:14 lr 0.000012 time 0.3048 (0.3140) loss 3.2432 (2.9217) grad_norm 3.7072 (3.4930) [2021-04-16 21:20:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][260/1251] eta 0:05:09 lr 0.000012 time 0.2478 (0.3123) loss 2.9968 (2.9224) grad_norm 3.9791 (3.5023) [2021-04-16 21:20:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][270/1251] eta 0:05:04 lr 0.000012 time 0.2601 (0.3109) loss 2.3633 (2.9208) grad_norm 3.4898 (3.5012) [2021-04-16 21:20:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][280/1251] eta 0:05:00 lr 0.000012 time 0.2667 (0.3096) loss 3.5563 (2.9293) grad_norm 3.5473 (3.4947) [2021-04-16 21:20:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][290/1251] eta 0:04:56 lr 0.000012 time 0.2899 (0.3085) loss 2.8268 (2.9312) grad_norm 3.1835 (3.4864) [2021-04-16 21:21:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][300/1251] eta 0:04:52 lr 0.000012 time 0.2976 (0.3078) loss 2.9020 (2.9314) grad_norm 3.2061 (3.4770) [2021-04-16 21:21:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][310/1251] eta 0:04:48 lr 0.000012 time 0.2923 (0.3070) loss 3.8191 (2.9331) grad_norm 3.2549 (3.4789) [2021-04-16 21:21:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][320/1251] eta 0:04:45 lr 0.000012 time 0.2697 (0.3063) loss 3.1264 (2.9312) grad_norm 2.9510 (3.4794) [2021-04-16 21:21:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][330/1251] eta 0:04:41 lr 0.000012 time 0.2732 (0.3055) loss 2.8826 (2.9257) grad_norm 3.5633 (3.4820) [2021-04-16 21:21:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][340/1251] eta 0:04:37 lr 0.000012 time 0.2638 (0.3049) loss 3.3865 (2.9346) grad_norm 3.0375 (3.4752) [2021-04-16 21:21:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][350/1251] eta 0:04:33 lr 0.000012 time 0.2863 (0.3040) loss 3.3013 (2.9427) grad_norm 4.0296 (3.4773) [2021-04-16 21:21:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][360/1251] eta 0:04:30 lr 0.000012 time 0.2612 (0.3039) loss 3.0954 (2.9514) grad_norm 3.4327 (3.4731) [2021-04-16 21:21:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][370/1251] eta 0:04:27 lr 0.000012 time 0.2564 (0.3033) loss 2.5796 (2.9539) grad_norm 3.9864 (3.4736) [2021-04-16 21:21:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][380/1251] eta 0:04:23 lr 0.000012 time 0.2699 (0.3025) loss 3.2976 (2.9558) grad_norm 3.2068 (3.4659) [2021-04-16 21:21:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][390/1251] eta 0:04:20 lr 0.000012 time 0.2724 (0.3020) loss 3.4376 (2.9607) grad_norm 3.0179 (3.4597) [2021-04-16 21:21:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][400/1251] eta 0:04:16 lr 0.000012 time 0.2559 (0.3013) loss 2.8363 (2.9554) grad_norm 2.6486 (3.4606) [2021-04-16 21:21:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][410/1251] eta 0:04:12 lr 0.000012 time 0.2726 (0.3007) loss 2.7486 (2.9503) grad_norm 3.3575 (3.4592) [2021-04-16 21:21:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][420/1251] eta 0:04:09 lr 0.000012 time 0.2555 (0.3002) loss 2.6331 (2.9561) grad_norm 3.6047 (3.4563) [2021-04-16 21:21:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][430/1251] eta 0:04:06 lr 0.000012 time 0.2943 (0.2997) loss 1.9928 (2.9495) grad_norm 3.2221 (3.4491) [2021-04-16 21:21:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][440/1251] eta 0:04:02 lr 0.000012 time 0.2749 (0.2991) loss 3.1742 (2.9488) grad_norm 2.9469 (3.4413) [2021-04-16 21:21:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][450/1251] eta 0:03:59 lr 0.000012 time 0.2775 (0.2986) loss 3.0812 (2.9522) grad_norm 4.1124 (3.4450) [2021-04-16 21:21:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][460/1251] eta 0:03:55 lr 0.000012 time 0.2644 (0.2981) loss 3.6885 (2.9585) grad_norm 3.5935 (3.4476) [2021-04-16 21:21:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][470/1251] eta 0:03:52 lr 0.000012 time 0.2859 (0.2978) loss 3.3721 (2.9600) grad_norm 4.0570 (3.4482) [2021-04-16 21:21:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][480/1251] eta 0:03:49 lr 0.000012 time 0.2851 (0.2977) loss 3.4626 (2.9633) grad_norm 3.2183 (3.4459) [2021-04-16 21:21:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][490/1251] eta 0:03:46 lr 0.000012 time 0.2723 (0.2976) loss 3.4278 (2.9681) grad_norm 4.5971 (3.4489) [2021-04-16 21:21:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][500/1251] eta 0:03:43 lr 0.000012 time 0.2916 (0.2974) loss 2.7272 (2.9634) grad_norm 3.0800 (3.4484) [2021-04-16 21:22:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][510/1251] eta 0:03:40 lr 0.000012 time 0.2879 (0.2970) loss 2.3220 (2.9611) grad_norm 3.5740 (3.4456) [2021-04-16 21:22:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][520/1251] eta 0:03:36 lr 0.000012 time 0.2796 (0.2968) loss 3.6362 (2.9598) grad_norm 3.7376 (3.4447) [2021-04-16 21:22:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][530/1251] eta 0:03:33 lr 0.000012 time 0.2809 (0.2966) loss 2.0545 (2.9587) grad_norm 3.8231 (3.4501) [2021-04-16 21:22:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][540/1251] eta 0:03:30 lr 0.000012 time 0.2756 (0.2966) loss 3.2294 (2.9625) grad_norm 3.4102 (3.4467) [2021-04-16 21:22:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][550/1251] eta 0:03:27 lr 0.000012 time 0.2885 (0.2963) loss 3.4430 (2.9618) grad_norm 3.9042 (3.4477) [2021-04-16 21:22:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][560/1251] eta 0:03:24 lr 0.000012 time 0.2880 (0.2959) loss 2.5679 (2.9616) grad_norm 4.9965 (3.4500) [2021-04-16 21:22:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][570/1251] eta 0:03:21 lr 0.000012 time 0.3002 (0.2958) loss 3.3545 (2.9653) grad_norm 3.5354 (3.4478) [2021-04-16 21:22:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][580/1251] eta 0:03:18 lr 0.000012 time 0.2802 (0.2958) loss 3.1618 (2.9652) grad_norm 3.9195 (3.4486) [2021-04-16 21:22:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][590/1251] eta 0:03:15 lr 0.000012 time 0.2925 (0.2954) loss 2.8823 (2.9651) grad_norm 4.5862 (3.4498) [2021-04-16 21:22:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][600/1251] eta 0:03:12 lr 0.000012 time 0.2737 (0.2953) loss 2.6867 (2.9621) grad_norm 3.7036 (3.4536) [2021-04-16 21:22:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][610/1251] eta 0:03:09 lr 0.000012 time 0.2405 (0.2950) loss 2.1884 (2.9604) grad_norm 3.3545 (3.4526) [2021-04-16 21:22:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][620/1251] eta 0:03:05 lr 0.000012 time 0.2654 (0.2947) loss 3.1676 (2.9640) grad_norm 3.6787 (3.4511) [2021-04-16 21:22:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][630/1251] eta 0:03:02 lr 0.000012 time 0.2654 (0.2944) loss 2.8610 (2.9615) grad_norm 3.3344 (3.4501) [2021-04-16 21:22:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][640/1251] eta 0:02:59 lr 0.000012 time 0.2669 (0.2941) loss 2.4497 (2.9609) grad_norm 3.4967 (3.4538) [2021-04-16 21:22:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][650/1251] eta 0:02:56 lr 0.000012 time 0.3029 (0.2939) loss 2.6701 (2.9614) grad_norm 3.4650 (3.4523) [2021-04-16 21:22:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][660/1251] eta 0:02:53 lr 0.000012 time 0.2618 (0.2938) loss 3.0900 (2.9599) grad_norm 3.3631 (3.4542) [2021-04-16 21:22:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][670/1251] eta 0:02:50 lr 0.000012 time 0.3092 (0.2938) loss 3.5232 (2.9602) grad_norm 3.6877 (3.4547) [2021-04-16 21:22:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][680/1251] eta 0:02:47 lr 0.000012 time 0.2825 (0.2937) loss 3.7157 (2.9602) grad_norm 3.0354 (3.4522) [2021-04-16 21:22:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][690/1251] eta 0:02:44 lr 0.000012 time 0.2739 (0.2935) loss 2.6448 (2.9636) grad_norm 3.9123 (3.4501) [2021-04-16 21:22:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][700/1251] eta 0:02:41 lr 0.000012 time 0.2445 (0.2932) loss 3.2649 (2.9646) grad_norm 3.1481 (3.4515) [2021-04-16 21:22:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][710/1251] eta 0:02:38 lr 0.000011 time 0.3055 (0.2931) loss 2.3390 (2.9628) grad_norm 3.4528 (3.4502) [2021-04-16 21:23:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][720/1251] eta 0:02:35 lr 0.000011 time 0.2719 (0.2933) loss 2.8150 (2.9631) grad_norm 3.7063 (3.4500) [2021-04-16 21:23:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][730/1251] eta 0:02:32 lr 0.000011 time 0.2737 (0.2930) loss 2.7141 (2.9650) grad_norm 3.6016 (3.4497) [2021-04-16 21:23:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][740/1251] eta 0:02:29 lr 0.000011 time 0.4814 (0.2931) loss 3.2129 (2.9663) grad_norm 3.2633 (3.4526) [2021-04-16 21:23:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][750/1251] eta 0:02:26 lr 0.000011 time 0.2947 (0.2929) loss 3.2730 (2.9679) grad_norm 3.8011 (3.4558) [2021-04-16 21:23:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][760/1251] eta 0:02:23 lr 0.000011 time 0.2532 (0.2927) loss 2.9434 (2.9641) grad_norm 5.7316 (3.4598) [2021-04-16 21:23:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][770/1251] eta 0:02:20 lr 0.000011 time 0.2735 (0.2925) loss 3.2528 (2.9670) grad_norm 3.6005 (3.4592) [2021-04-16 21:23:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][780/1251] eta 0:02:17 lr 0.000011 time 0.2850 (0.2923) loss 3.5831 (2.9680) grad_norm 2.9876 (3.4599) [2021-04-16 21:23:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][790/1251] eta 0:02:14 lr 0.000011 time 0.2767 (0.2922) loss 3.7654 (2.9662) grad_norm 4.2884 (3.4575) [2021-04-16 21:23:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][800/1251] eta 0:02:11 lr 0.000011 time 0.2766 (0.2922) loss 3.4372 (2.9665) grad_norm 3.2816 (3.4571) [2021-04-16 21:23:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][810/1251] eta 0:02:08 lr 0.000011 time 0.2829 (0.2921) loss 2.1435 (2.9651) grad_norm 5.2427 (3.4596) [2021-04-16 21:23:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][820/1251] eta 0:02:05 lr 0.000011 time 0.2690 (0.2919) loss 2.9666 (2.9654) grad_norm 3.1446 (3.4594) [2021-04-16 21:23:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][830/1251] eta 0:02:02 lr 0.000011 time 0.2948 (0.2918) loss 3.0000 (2.9690) grad_norm 3.1050 (3.4604) [2021-04-16 21:23:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][840/1251] eta 0:01:59 lr 0.000011 time 0.2760 (0.2916) loss 2.6567 (2.9698) grad_norm 2.9568 (3.4591) [2021-04-16 21:23:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][850/1251] eta 0:01:56 lr 0.000011 time 0.2773 (0.2915) loss 3.2270 (2.9691) grad_norm 3.0230 (3.4618) [2021-04-16 21:23:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][860/1251] eta 0:01:53 lr 0.000011 time 0.2734 (0.2913) loss 2.6387 (2.9680) grad_norm 3.0732 (3.4651) [2021-04-16 21:23:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][870/1251] eta 0:01:50 lr 0.000011 time 0.2575 (0.2911) loss 3.0959 (2.9680) grad_norm 3.8833 (3.4668) [2021-04-16 21:23:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][880/1251] eta 0:01:47 lr 0.000011 time 0.2601 (0.2910) loss 3.0688 (2.9687) grad_norm 3.2420 (3.4635) [2021-04-16 21:23:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][890/1251] eta 0:01:45 lr 0.000011 time 0.3044 (0.2909) loss 2.4693 (2.9689) grad_norm 3.6803 (3.4633) [2021-04-16 21:23:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][900/1251] eta 0:01:42 lr 0.000011 time 0.2575 (0.2908) loss 2.5947 (2.9696) grad_norm 3.3316 (3.4607) [2021-04-16 21:23:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][910/1251] eta 0:01:39 lr 0.000011 time 0.2753 (0.2907) loss 3.2355 (2.9700) grad_norm 6.1134 (3.4691) [2021-04-16 21:23:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][920/1251] eta 0:01:36 lr 0.000011 time 0.2646 (0.2906) loss 2.7320 (2.9684) grad_norm 3.0429 (3.4663) [2021-04-16 21:23:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][930/1251] eta 0:01:33 lr 0.000011 time 0.2864 (0.2906) loss 3.2538 (2.9695) grad_norm 3.7454 (3.4676) [2021-04-16 21:24:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][940/1251] eta 0:01:30 lr 0.000011 time 0.2794 (0.2905) loss 3.4096 (2.9725) grad_norm 3.3986 (3.4672) [2021-04-16 21:24:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][950/1251] eta 0:01:27 lr 0.000011 time 0.3018 (0.2904) loss 3.5321 (2.9711) grad_norm 3.4708 (3.4669) [2021-04-16 21:24:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][960/1251] eta 0:01:24 lr 0.000011 time 0.2740 (0.2903) loss 2.1388 (2.9687) grad_norm 2.9406 (3.4674) [2021-04-16 21:24:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][970/1251] eta 0:01:21 lr 0.000011 time 0.2618 (0.2902) loss 3.2829 (2.9696) grad_norm 3.6988 (3.4672) [2021-04-16 21:24:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][980/1251] eta 0:01:18 lr 0.000011 time 0.2672 (0.2900) loss 3.8112 (2.9687) grad_norm 3.2472 (3.4682) [2021-04-16 21:24:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][990/1251] eta 0:01:15 lr 0.000011 time 0.2751 (0.2899) loss 3.1084 (2.9671) grad_norm 3.1010 (3.4660) [2021-04-16 21:24:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][1000/1251] eta 0:01:12 lr 0.000011 time 0.2658 (0.2897) loss 2.6687 (2.9668) grad_norm 3.8878 (3.4682) [2021-04-16 21:24:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][1010/1251] eta 0:01:09 lr 0.000011 time 0.2599 (0.2895) loss 2.9132 (2.9651) grad_norm 3.6615 (3.4672) [2021-04-16 21:24:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][1020/1251] eta 0:01:06 lr 0.000011 time 0.2679 (0.2894) loss 2.2260 (2.9645) grad_norm 2.8412 (3.4671) [2021-04-16 21:24:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][1030/1251] eta 0:01:03 lr 0.000011 time 0.2514 (0.2892) loss 3.6179 (2.9654) grad_norm 2.8550 (3.4657) [2021-04-16 21:24:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][1040/1251] eta 0:01:01 lr 0.000011 time 0.2740 (0.2891) loss 1.8086 (2.9624) grad_norm 3.3864 (3.4730) [2021-04-16 21:24:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][1050/1251] eta 0:00:58 lr 0.000011 time 0.3973 (0.2891) loss 3.2840 (2.9610) grad_norm 4.2838 (3.4716) [2021-04-16 21:24:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][1060/1251] eta 0:00:55 lr 0.000011 time 0.2624 (0.2890) loss 3.3694 (2.9625) grad_norm 3.2403 (3.4687) [2021-04-16 21:24:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][1070/1251] eta 0:00:52 lr 0.000011 time 0.2827 (0.2890) loss 3.3850 (2.9626) grad_norm 3.5251 (3.4674) [2021-04-16 21:24:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][1080/1251] eta 0:00:49 lr 0.000011 time 0.2700 (0.2889) loss 2.9038 (2.9645) grad_norm 4.7370 (3.4695) [2021-04-16 21:24:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][1090/1251] eta 0:00:46 lr 0.000011 time 0.2553 (0.2888) loss 3.4484 (2.9643) grad_norm 3.6108 (3.4728) [2021-04-16 21:24:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][1100/1251] eta 0:00:43 lr 0.000011 time 0.2812 (0.2887) loss 3.0206 (2.9640) grad_norm 3.2818 (3.4728) [2021-04-16 21:24:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][1110/1251] eta 0:00:40 lr 0.000011 time 0.2615 (0.2885) loss 2.9167 (2.9651) grad_norm 3.3167 (3.4739) [2021-04-16 21:24:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][1120/1251] eta 0:00:37 lr 0.000011 time 0.2877 (0.2885) loss 2.9990 (2.9661) grad_norm 3.1226 (3.4738) [2021-04-16 21:24:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][1130/1251] eta 0:00:34 lr 0.000011 time 0.2874 (0.2884) loss 2.9157 (2.9661) grad_norm 3.4817 (3.4727) [2021-04-16 21:24:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][1140/1251] eta 0:00:32 lr 0.000011 time 0.2942 (0.2884) loss 2.9182 (2.9638) grad_norm 3.6698 (3.4735) [2021-04-16 21:25:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][1150/1251] eta 0:00:29 lr 0.000011 time 0.2656 (0.2884) loss 3.0810 (2.9627) grad_norm 2.8544 (3.4725) [2021-04-16 21:25:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][1160/1251] eta 0:00:26 lr 0.000011 time 0.2843 (0.2885) loss 3.1014 (2.9609) grad_norm 2.9219 (3.4720) [2021-04-16 21:25:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][1170/1251] eta 0:00:23 lr 0.000011 time 0.2756 (0.2884) loss 2.9103 (2.9609) grad_norm 3.0151 (3.4713) [2021-04-16 21:25:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][1180/1251] eta 0:00:20 lr 0.000011 time 0.2840 (0.2883) loss 3.6911 (2.9626) grad_norm 3.5321 (3.4717) [2021-04-16 21:25:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][1190/1251] eta 0:00:17 lr 0.000011 time 0.2774 (0.2882) loss 2.6387 (2.9626) grad_norm 4.5545 (3.4727) [2021-04-16 21:25:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][1200/1251] eta 0:00:14 lr 0.000011 time 0.2668 (0.2881) loss 2.3609 (2.9616) grad_norm 4.2892 (3.4718) [2021-04-16 21:25:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][1210/1251] eta 0:00:11 lr 0.000011 time 0.2578 (0.2880) loss 3.2754 (2.9625) grad_norm 3.2828 (3.4711) [2021-04-16 21:25:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][1220/1251] eta 0:00:08 lr 0.000011 time 0.2943 (0.2879) loss 3.0380 (2.9624) grad_norm 3.6740 (3.4716) [2021-04-16 21:25:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][1230/1251] eta 0:00:06 lr 0.000011 time 0.2991 (0.2878) loss 2.3674 (2.9645) grad_norm 2.8738 (3.4703) [2021-04-16 21:25:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][1240/1251] eta 0:00:03 lr 0.000011 time 0.3275 (0.2877) loss 3.2164 (2.9668) grad_norm 3.1678 (3.4693) [2021-04-16 21:25:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [292/300][1250/1251] eta 0:00:00 lr 0.000011 time 0.2485 (0.2874) loss 3.3073 (2.9645) grad_norm 2.9788 (3.4676) [2021-04-16 21:25:34 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 292 training takes 0:06:05 [2021-04-16 21:25:34 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_292.pth saving...... [2021-04-16 21:25:48 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_292.pth saved !!! [2021-04-16 21:25:49 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.116 (1.116) Loss 0.8405 (0.8405) Acc@1 79.297 (79.297) Acc@5 95.410 (95.410) [2021-04-16 21:25:50 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.143 (0.223) Loss 0.8048 (0.8046) Acc@1 80.859 (81.090) Acc@5 96.094 (95.694) [2021-04-16 21:25:53 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.818 (0.245) Loss 0.8104 (0.8098) Acc@1 81.250 (81.027) Acc@5 95.508 (95.592) [2021-04-16 21:25:55 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.418 (0.240) Loss 0.8557 (0.8159) Acc@1 80.371 (80.938) Acc@5 95.020 (95.508) [2021-04-16 21:25:56 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.147 (0.218) Loss 0.8110 (0.8126) Acc@1 80.957 (81.117) Acc@5 95.508 (95.496) [2021-04-16 21:26:17 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 81.226 Acc@5 95.552 [2021-04-16 21:26:17 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 81.2% [2021-04-16 21:26:17 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 81.24% [2021-04-16 21:26:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][0/1251] eta 3:46:34 lr 0.000011 time 10.8670 (10.8670) loss 3.0697 (3.0697) grad_norm 3.9882 (3.9882) [2021-04-16 21:26:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][10/1251] eta 0:25:36 lr 0.000011 time 0.2753 (1.2382) loss 2.9771 (3.0211) grad_norm 3.1416 (3.3958) [2021-04-16 21:26:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][20/1251] eta 0:16:10 lr 0.000011 time 0.2818 (0.7881) loss 3.5544 (3.0037) grad_norm 3.8134 (3.6188) [2021-04-16 21:26:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][30/1251] eta 0:12:42 lr 0.000011 time 0.3021 (0.6248) loss 2.1083 (2.9459) grad_norm 3.5169 (3.5187) [2021-04-16 21:26:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][40/1251] eta 0:10:56 lr 0.000011 time 0.2557 (0.5423) loss 3.2401 (2.9618) grad_norm 4.0265 (3.4869) [2021-04-16 21:26:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][50/1251] eta 0:09:48 lr 0.000011 time 0.2746 (0.4899) loss 2.9406 (2.9312) grad_norm 4.5409 (3.5134) [2021-04-16 21:26:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][60/1251] eta 0:09:01 lr 0.000011 time 0.2886 (0.4550) loss 2.6482 (2.9145) grad_norm 3.0998 (3.4991) [2021-04-16 21:26:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][70/1251] eta 0:08:27 lr 0.000011 time 0.2813 (0.4297) loss 3.4266 (2.9630) grad_norm 4.3404 (3.5025) [2021-04-16 21:26:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][80/1251] eta 0:08:01 lr 0.000011 time 0.2916 (0.4108) loss 3.4080 (2.9460) grad_norm 3.5860 (3.4814) [2021-04-16 21:26:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][90/1251] eta 0:07:40 lr 0.000011 time 0.2864 (0.3963) loss 2.0971 (2.9508) grad_norm 5.5449 (3.5049) [2021-04-16 21:26:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][100/1251] eta 0:07:22 lr 0.000011 time 0.2959 (0.3845) loss 2.9783 (2.9778) grad_norm 3.2408 (3.4729) [2021-04-16 21:26:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][110/1251] eta 0:07:08 lr 0.000011 time 0.2977 (0.3754) loss 3.1032 (2.9825) grad_norm 3.5772 (3.4729) [2021-04-16 21:27:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][120/1251] eta 0:06:55 lr 0.000011 time 0.2733 (0.3669) loss 2.9356 (2.9764) grad_norm 2.9973 (3.4668) [2021-04-16 21:27:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][130/1251] eta 0:06:44 lr 0.000011 time 0.2873 (0.3606) loss 3.4732 (2.9802) grad_norm 3.3959 (3.4645) [2021-04-16 21:27:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][140/1251] eta 0:06:35 lr 0.000011 time 0.2749 (0.3557) loss 2.6980 (2.9715) grad_norm 2.9354 (3.4562) [2021-04-16 21:27:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][150/1251] eta 0:06:25 lr 0.000011 time 0.2824 (0.3504) loss 2.8932 (2.9590) grad_norm 3.7389 (3.4729) [2021-04-16 21:27:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][160/1251] eta 0:06:18 lr 0.000011 time 0.3030 (0.3471) loss 3.1920 (2.9588) grad_norm 2.9946 (3.4584) [2021-04-16 21:27:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][170/1251] eta 0:06:11 lr 0.000011 time 0.2588 (0.3434) loss 2.5786 (2.9478) grad_norm 3.8186 (3.4480) [2021-04-16 21:27:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][180/1251] eta 0:06:04 lr 0.000011 time 0.2473 (0.3401) loss 2.6146 (2.9385) grad_norm 3.3429 (3.4668) [2021-04-16 21:27:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][190/1251] eta 0:05:57 lr 0.000011 time 0.2859 (0.3373) loss 2.6479 (2.9417) grad_norm 3.5246 (3.4556) [2021-04-16 21:27:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][200/1251] eta 0:05:51 lr 0.000011 time 0.2659 (0.3343) loss 2.9754 (2.9427) grad_norm 3.1364 (3.4485) [2021-04-16 21:27:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][210/1251] eta 0:05:44 lr 0.000011 time 0.2641 (0.3314) loss 3.2472 (2.9485) grad_norm 3.0410 (3.4441) [2021-04-16 21:27:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][220/1251] eta 0:05:38 lr 0.000011 time 0.2777 (0.3288) loss 3.0443 (2.9431) grad_norm 3.1831 (3.4436) [2021-04-16 21:27:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][230/1251] eta 0:05:33 lr 0.000011 time 0.2702 (0.3265) loss 2.9584 (2.9445) grad_norm 3.2144 (3.4449) [2021-04-16 21:27:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][240/1251] eta 0:05:28 lr 0.000011 time 0.2833 (0.3248) loss 3.2614 (2.9350) grad_norm 3.2220 (3.4428) [2021-04-16 21:27:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][250/1251] eta 0:05:23 lr 0.000011 time 0.3016 (0.3231) loss 3.2818 (2.9481) grad_norm 3.2473 (3.4543) [2021-04-16 21:27:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][260/1251] eta 0:05:18 lr 0.000011 time 0.4144 (0.3218) loss 1.8955 (2.9455) grad_norm 3.4890 (3.4580) [2021-04-16 21:27:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][270/1251] eta 0:05:14 lr 0.000011 time 0.2864 (0.3204) loss 3.0443 (2.9494) grad_norm 3.3967 (3.4712) [2021-04-16 21:27:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][280/1251] eta 0:05:09 lr 0.000011 time 0.2764 (0.3189) loss 2.1999 (2.9487) grad_norm 2.9974 (nan) [2021-04-16 21:27:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][290/1251] eta 0:05:05 lr 0.000011 time 0.2886 (0.3181) loss 3.5397 (2.9513) grad_norm 3.5639 (nan) [2021-04-16 21:27:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][300/1251] eta 0:05:01 lr 0.000011 time 0.2644 (0.3166) loss 3.3238 (2.9489) grad_norm 3.7656 (nan) [2021-04-16 21:27:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][310/1251] eta 0:04:56 lr 0.000011 time 0.2893 (0.3154) loss 3.1690 (2.9538) grad_norm 3.1194 (nan) [2021-04-16 21:27:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][320/1251] eta 0:04:53 lr 0.000011 time 0.2856 (0.3147) loss 3.3798 (2.9599) grad_norm 3.9485 (nan) [2021-04-16 21:28:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][330/1251] eta 0:04:48 lr 0.000011 time 0.2776 (0.3137) loss 3.1638 (2.9695) grad_norm 3.5772 (nan) [2021-04-16 21:28:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][340/1251] eta 0:04:44 lr 0.000011 time 0.2745 (0.3128) loss 2.6025 (2.9732) grad_norm 3.4861 (nan) [2021-04-16 21:28:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][350/1251] eta 0:04:40 lr 0.000011 time 0.2616 (0.3118) loss 3.3359 (2.9754) grad_norm 2.8176 (nan) [2021-04-16 21:28:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][360/1251] eta 0:04:37 lr 0.000011 time 0.3113 (0.3117) loss 3.3940 (2.9779) grad_norm 3.6389 (nan) [2021-04-16 21:28:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][370/1251] eta 0:04:33 lr 0.000011 time 0.2554 (0.3108) loss 3.0017 (2.9840) grad_norm 3.1032 (nan) [2021-04-16 21:28:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][380/1251] eta 0:04:29 lr 0.000011 time 0.2913 (0.3100) loss 2.1854 (2.9744) grad_norm 3.3471 (nan) [2021-04-16 21:28:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][390/1251] eta 0:04:26 lr 0.000011 time 0.2844 (0.3092) loss 3.4376 (2.9732) grad_norm 3.0857 (nan) [2021-04-16 21:28:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][400/1251] eta 0:04:22 lr 0.000011 time 0.2765 (0.3085) loss 3.0336 (2.9753) grad_norm 3.7371 (nan) [2021-04-16 21:28:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][410/1251] eta 0:04:18 lr 0.000011 time 0.2939 (0.3079) loss 1.9025 (2.9636) grad_norm 2.9796 (nan) [2021-04-16 21:28:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][420/1251] eta 0:04:15 lr 0.000011 time 0.3039 (0.3071) loss 3.0466 (2.9608) grad_norm 2.9443 (nan) [2021-04-16 21:28:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][430/1251] eta 0:04:11 lr 0.000011 time 0.2880 (0.3065) loss 3.4523 (2.9653) grad_norm 3.8173 (nan) [2021-04-16 21:28:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][440/1251] eta 0:04:08 lr 0.000011 time 0.2713 (0.3062) loss 3.1728 (2.9658) grad_norm 3.5316 (nan) [2021-04-16 21:28:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][450/1251] eta 0:04:04 lr 0.000011 time 0.3029 (0.3057) loss 3.2116 (2.9651) grad_norm 2.9412 (nan) [2021-04-16 21:28:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][460/1251] eta 0:04:01 lr 0.000011 time 0.2774 (0.3052) loss 3.7959 (2.9704) grad_norm 3.0428 (nan) [2021-04-16 21:28:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][470/1251] eta 0:03:57 lr 0.000011 time 0.2682 (0.3046) loss 3.5114 (2.9771) grad_norm 3.2652 (nan) [2021-04-16 21:28:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][480/1251] eta 0:03:54 lr 0.000011 time 0.2904 (0.3041) loss 3.2718 (2.9787) grad_norm 2.9973 (nan) [2021-04-16 21:28:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][490/1251] eta 0:03:51 lr 0.000011 time 0.2453 (0.3038) loss 3.6233 (2.9785) grad_norm 3.8847 (nan) [2021-04-16 21:28:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][500/1251] eta 0:03:47 lr 0.000011 time 0.2918 (0.3033) loss 3.5306 (2.9783) grad_norm 3.0207 (nan) [2021-04-16 21:28:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][510/1251] eta 0:03:44 lr 0.000011 time 0.2710 (0.3029) loss 3.0768 (2.9782) grad_norm 3.2402 (nan) [2021-04-16 21:28:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][520/1251] eta 0:03:41 lr 0.000011 time 0.2963 (0.3024) loss 3.5538 (2.9754) grad_norm 3.2790 (nan) [2021-04-16 21:28:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][530/1251] eta 0:03:37 lr 0.000011 time 0.2761 (0.3020) loss 2.5445 (2.9723) grad_norm 3.2472 (nan) [2021-04-16 21:29:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][540/1251] eta 0:03:34 lr 0.000011 time 0.2776 (0.3016) loss 3.2380 (2.9705) grad_norm 4.1736 (nan) [2021-04-16 21:29:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][550/1251] eta 0:03:31 lr 0.000011 time 0.3158 (0.3011) loss 2.0839 (2.9705) grad_norm 4.2126 (nan) [2021-04-16 21:29:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][560/1251] eta 0:03:27 lr 0.000011 time 0.2891 (0.3007) loss 3.0978 (2.9716) grad_norm 2.9884 (nan) [2021-04-16 21:29:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][570/1251] eta 0:03:24 lr 0.000011 time 0.2671 (0.3007) loss 3.2455 (2.9770) grad_norm 3.6365 (nan) [2021-04-16 21:29:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][580/1251] eta 0:03:21 lr 0.000011 time 0.2821 (0.3009) loss 2.4731 (2.9785) grad_norm 3.7088 (nan) [2021-04-16 21:29:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][590/1251] eta 0:03:18 lr 0.000011 time 0.2682 (0.3008) loss 2.2285 (2.9768) grad_norm 3.7234 (nan) [2021-04-16 21:29:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][600/1251] eta 0:03:15 lr 0.000011 time 0.2876 (0.3005) loss 3.6097 (2.9779) grad_norm 3.1932 (nan) [2021-04-16 21:29:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][610/1251] eta 0:03:12 lr 0.000011 time 0.2982 (0.3002) loss 3.1245 (2.9803) grad_norm 3.7313 (nan) [2021-04-16 21:29:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][620/1251] eta 0:03:09 lr 0.000011 time 0.2591 (0.2997) loss 2.9268 (2.9777) grad_norm 3.0531 (nan) [2021-04-16 21:29:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][630/1251] eta 0:03:05 lr 0.000011 time 0.2795 (0.2994) loss 2.9536 (2.9796) grad_norm 3.0633 (nan) [2021-04-16 21:29:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][640/1251] eta 0:03:02 lr 0.000011 time 0.3240 (0.2992) loss 2.0119 (2.9761) grad_norm 3.0980 (nan) [2021-04-16 21:29:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][650/1251] eta 0:02:59 lr 0.000011 time 0.2713 (0.2989) loss 3.1372 (2.9779) grad_norm 3.0930 (nan) [2021-04-16 21:29:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][660/1251] eta 0:02:56 lr 0.000011 time 0.2813 (0.2986) loss 2.7379 (2.9797) grad_norm 3.5598 (nan) [2021-04-16 21:29:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][670/1251] eta 0:02:53 lr 0.000011 time 0.2697 (0.2983) loss 2.3340 (2.9749) grad_norm 3.4882 (nan) [2021-04-16 21:29:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][680/1251] eta 0:02:50 lr 0.000011 time 0.2544 (0.2981) loss 2.9408 (2.9731) grad_norm 3.8929 (nan) [2021-04-16 21:29:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][690/1251] eta 0:02:47 lr 0.000011 time 0.3218 (0.2979) loss 2.4840 (2.9719) grad_norm 3.1798 (nan) [2021-04-16 21:29:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][700/1251] eta 0:02:43 lr 0.000011 time 0.3140 (0.2976) loss 2.9532 (2.9743) grad_norm 3.2000 (nan) [2021-04-16 21:29:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][710/1251] eta 0:02:40 lr 0.000011 time 0.2563 (0.2974) loss 3.2174 (2.9745) grad_norm 3.4294 (nan) [2021-04-16 21:29:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][720/1251] eta 0:02:37 lr 0.000011 time 0.2794 (0.2974) loss 3.4812 (2.9737) grad_norm 3.7118 (nan) [2021-04-16 21:29:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][730/1251] eta 0:02:34 lr 0.000011 time 0.4279 (0.2974) loss 2.7970 (2.9692) grad_norm 3.4675 (nan) [2021-04-16 21:29:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][740/1251] eta 0:02:31 lr 0.000011 time 0.2649 (0.2971) loss 2.1344 (2.9714) grad_norm 3.5439 (nan) [2021-04-16 21:30:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][750/1251] eta 0:02:28 lr 0.000011 time 0.2641 (0.2969) loss 3.3603 (2.9729) grad_norm 3.7325 (nan) [2021-04-16 21:30:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][760/1251] eta 0:02:25 lr 0.000011 time 0.2674 (0.2967) loss 2.2428 (2.9722) grad_norm 3.0206 (nan) [2021-04-16 21:30:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][770/1251] eta 0:02:22 lr 0.000011 time 0.2729 (0.2965) loss 2.9533 (2.9751) grad_norm 3.6678 (nan) [2021-04-16 21:30:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][780/1251] eta 0:02:19 lr 0.000011 time 0.2834 (0.2963) loss 2.3016 (2.9737) grad_norm 3.1635 (nan) [2021-04-16 21:30:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][790/1251] eta 0:02:16 lr 0.000011 time 0.2839 (0.2962) loss 2.7975 (2.9742) grad_norm 3.1679 (nan) [2021-04-16 21:30:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][800/1251] eta 0:02:13 lr 0.000011 time 0.2596 (0.2959) loss 3.0873 (2.9773) grad_norm 3.8886 (nan) [2021-04-16 21:30:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][810/1251] eta 0:02:10 lr 0.000011 time 0.2963 (0.2957) loss 3.3274 (2.9792) grad_norm 3.4493 (nan) [2021-04-16 21:30:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][820/1251] eta 0:02:07 lr 0.000011 time 0.3016 (0.2954) loss 2.9492 (2.9813) grad_norm 3.5837 (nan) [2021-04-16 21:30:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][830/1251] eta 0:02:04 lr 0.000011 time 0.2885 (0.2953) loss 2.9077 (2.9830) grad_norm 3.7918 (nan) [2021-04-16 21:30:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][840/1251] eta 0:02:01 lr 0.000011 time 0.2644 (0.2952) loss 3.2802 (2.9842) grad_norm 3.0209 (nan) [2021-04-16 21:30:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][850/1251] eta 0:01:58 lr 0.000011 time 0.2731 (0.2949) loss 2.3482 (2.9825) grad_norm 4.2703 (nan) [2021-04-16 21:30:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][860/1251] eta 0:01:55 lr 0.000011 time 0.2724 (0.2947) loss 2.7190 (2.9818) grad_norm 3.3052 (nan) [2021-04-16 21:30:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][870/1251] eta 0:01:52 lr 0.000011 time 0.2747 (0.2946) loss 2.9806 (2.9819) grad_norm 3.5205 (nan) [2021-04-16 21:30:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][880/1251] eta 0:01:49 lr 0.000011 time 0.2872 (0.2945) loss 3.0472 (2.9826) grad_norm 2.9032 (nan) [2021-04-16 21:30:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][890/1251] eta 0:01:46 lr 0.000011 time 0.2860 (0.2943) loss 3.3886 (2.9842) grad_norm 3.4514 (nan) [2021-04-16 21:30:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][900/1251] eta 0:01:43 lr 0.000011 time 0.2594 (0.2941) loss 2.5269 (2.9860) grad_norm 2.7821 (nan) [2021-04-16 21:30:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][910/1251] eta 0:01:40 lr 0.000011 time 0.2614 (0.2939) loss 3.1097 (2.9865) grad_norm 3.7021 (nan) [2021-04-16 21:30:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][920/1251] eta 0:01:37 lr 0.000011 time 0.2909 (0.2937) loss 2.9877 (2.9887) grad_norm 2.9370 (nan) [2021-04-16 21:30:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][930/1251] eta 0:01:34 lr 0.000011 time 0.2511 (0.2936) loss 2.7198 (2.9885) grad_norm 4.3365 (nan) [2021-04-16 21:30:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][940/1251] eta 0:01:31 lr 0.000011 time 0.2706 (0.2935) loss 2.3772 (2.9882) grad_norm 2.9946 (nan) [2021-04-16 21:30:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][950/1251] eta 0:01:28 lr 0.000011 time 0.2778 (0.2933) loss 2.8111 (2.9891) grad_norm 3.8889 (nan) [2021-04-16 21:30:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][960/1251] eta 0:01:25 lr 0.000011 time 0.2850 (0.2932) loss 3.3445 (2.9884) grad_norm 4.1825 (nan) [2021-04-16 21:31:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][970/1251] eta 0:01:22 lr 0.000011 time 0.2683 (0.2930) loss 2.2208 (2.9875) grad_norm 2.9480 (nan) [2021-04-16 21:31:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][980/1251] eta 0:01:19 lr 0.000011 time 0.2535 (0.2928) loss 2.0136 (2.9886) grad_norm 3.3540 (nan) [2021-04-16 21:31:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][990/1251] eta 0:01:16 lr 0.000011 time 0.2793 (0.2927) loss 3.5457 (2.9885) grad_norm 3.1181 (nan) [2021-04-16 21:31:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][1000/1251] eta 0:01:13 lr 0.000011 time 0.2891 (0.2926) loss 2.9032 (2.9882) grad_norm 2.9730 (nan) [2021-04-16 21:31:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][1010/1251] eta 0:01:10 lr 0.000011 time 0.2817 (0.2925) loss 3.3635 (2.9880) grad_norm 2.7063 (nan) [2021-04-16 21:31:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][1020/1251] eta 0:01:07 lr 0.000011 time 0.3203 (0.2926) loss 2.4666 (2.9899) grad_norm 4.1580 (nan) [2021-04-16 21:31:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][1030/1251] eta 0:01:04 lr 0.000011 time 0.2826 (0.2925) loss 3.1612 (2.9865) grad_norm 2.9267 (nan) [2021-04-16 21:31:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][1040/1251] eta 0:01:01 lr 0.000011 time 0.2875 (0.2924) loss 3.4736 (2.9847) grad_norm 3.8912 (nan) [2021-04-16 21:31:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][1050/1251] eta 0:00:58 lr 0.000011 time 0.2839 (0.2923) loss 2.8316 (2.9841) grad_norm 3.1584 (nan) [2021-04-16 21:31:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][1060/1251] eta 0:00:55 lr 0.000011 time 0.2699 (0.2921) loss 3.7306 (2.9845) grad_norm 3.5180 (nan) [2021-04-16 21:31:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][1070/1251] eta 0:00:52 lr 0.000011 time 0.2790 (0.2921) loss 3.2571 (2.9854) grad_norm 3.0631 (nan) [2021-04-16 21:31:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][1080/1251] eta 0:00:49 lr 0.000011 time 0.2669 (0.2920) loss 3.2923 (2.9848) grad_norm 3.0723 (nan) [2021-04-16 21:31:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][1090/1251] eta 0:00:46 lr 0.000011 time 0.2515 (0.2919) loss 2.3563 (2.9863) grad_norm 3.3934 (nan) [2021-04-16 21:31:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][1100/1251] eta 0:00:44 lr 0.000011 time 0.2807 (0.2918) loss 2.9194 (2.9853) grad_norm 3.9797 (nan) [2021-04-16 21:31:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][1110/1251] eta 0:00:41 lr 0.000011 time 0.2623 (0.2916) loss 3.5771 (2.9870) grad_norm 4.0569 (nan) [2021-04-16 21:31:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][1120/1251] eta 0:00:38 lr 0.000011 time 0.2912 (0.2915) loss 2.8870 (2.9867) grad_norm 3.3249 (nan) [2021-04-16 21:31:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][1130/1251] eta 0:00:35 lr 0.000011 time 0.2939 (0.2914) loss 3.6153 (2.9868) grad_norm 3.0572 (nan) [2021-04-16 21:31:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][1140/1251] eta 0:00:32 lr 0.000011 time 0.2557 (0.2914) loss 2.2557 (2.9862) grad_norm 3.1113 (nan) [2021-04-16 21:31:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][1150/1251] eta 0:00:29 lr 0.000011 time 0.4312 (0.2916) loss 3.3941 (2.9861) grad_norm 3.9155 (nan) [2021-04-16 21:31:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][1160/1251] eta 0:00:26 lr 0.000011 time 0.3047 (0.2915) loss 3.1039 (2.9855) grad_norm 2.9726 (nan) [2021-04-16 21:31:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][1170/1251] eta 0:00:23 lr 0.000011 time 0.2867 (0.2916) loss 3.1176 (2.9856) grad_norm 2.6774 (nan) [2021-04-16 21:32:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][1180/1251] eta 0:00:20 lr 0.000011 time 0.2548 (0.2915) loss 3.4856 (2.9836) grad_norm 3.6309 (nan) [2021-04-16 21:32:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][1190/1251] eta 0:00:17 lr 0.000011 time 0.2791 (0.2914) loss 3.2959 (2.9828) grad_norm 3.0851 (nan) [2021-04-16 21:32:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][1200/1251] eta 0:00:14 lr 0.000011 time 0.2872 (0.2913) loss 2.1086 (2.9819) grad_norm 3.6312 (nan) [2021-04-16 21:32:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][1210/1251] eta 0:00:11 lr 0.000011 time 0.2686 (0.2911) loss 2.9255 (2.9832) grad_norm 3.2405 (nan) [2021-04-16 21:32:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][1220/1251] eta 0:00:09 lr 0.000011 time 0.3009 (0.2911) loss 3.0880 (2.9842) grad_norm 2.9258 (nan) [2021-04-16 21:32:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][1230/1251] eta 0:00:06 lr 0.000011 time 0.2734 (0.2909) loss 2.5498 (2.9827) grad_norm 3.5384 (nan) [2021-04-16 21:32:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][1240/1251] eta 0:00:03 lr 0.000011 time 0.2482 (0.2908) loss 3.3712 (2.9846) grad_norm 2.9880 (nan) [2021-04-16 21:32:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [293/300][1250/1251] eta 0:00:00 lr 0.000011 time 0.2482 (0.2904) loss 3.5089 (2.9841) grad_norm 3.4096 (nan) [2021-04-16 21:32:29 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 293 training takes 0:06:12 [2021-04-16 21:32:29 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_293.pth saving...... [2021-04-16 21:32:49 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_293.pth saved !!! [2021-04-16 21:32:50 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.107 (1.107) Loss 0.8354 (0.8354) Acc@1 79.492 (79.492) Acc@5 95.312 (95.312) [2021-04-16 21:32:51 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.387 (0.243) Loss 0.8307 (0.8001) Acc@1 79.004 (81.197) Acc@5 95.703 (95.801) [2021-04-16 21:32:54 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.084 (0.249) Loss 0.8645 (0.8020) Acc@1 79.395 (81.157) Acc@5 95.117 (95.652) [2021-04-16 21:32:56 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.170 (0.240) Loss 0.9265 (0.8098) Acc@1 78.809 (81.092) Acc@5 93.262 (95.530) [2021-04-16 21:32:58 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.140 (0.218) Loss 0.8649 (0.8148) Acc@1 80.078 (80.993) Acc@5 95.605 (95.501) [2021-04-16 21:33:15 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 81.134 Acc@5 95.524 [2021-04-16 21:33:15 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 81.1% [2021-04-16 21:33:15 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 81.24% [2021-04-16 21:33:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][0/1251] eta 5:15:17 lr 0.000011 time 15.1217 (15.1217) loss 2.7288 (2.7288) grad_norm 3.9653 (3.9653) [2021-04-16 21:33:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][10/1251] eta 0:33:31 lr 0.000011 time 0.2764 (1.6209) loss 3.7462 (2.9959) grad_norm 3.0271 (3.4877) [2021-04-16 21:33:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][20/1251] eta 0:20:07 lr 0.000011 time 0.3134 (0.9813) loss 2.3283 (3.0440) grad_norm 3.2902 (3.4278) [2021-04-16 21:33:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][30/1251] eta 0:15:19 lr 0.000011 time 0.2836 (0.7532) loss 3.4203 (2.9439) grad_norm 3.1808 (3.4234) [2021-04-16 21:33:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][40/1251] eta 0:12:53 lr 0.000011 time 0.2540 (0.6389) loss 2.5140 (2.9255) grad_norm 3.0340 (3.3851) [2021-04-16 21:33:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][50/1251] eta 0:11:22 lr 0.000011 time 0.2524 (0.5679) loss 2.0070 (2.9441) grad_norm 3.5012 (3.3685) [2021-04-16 21:33:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][60/1251] eta 0:10:21 lr 0.000011 time 0.3638 (0.5216) loss 3.3952 (2.9644) grad_norm 3.0926 (3.3962) [2021-04-16 21:33:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][70/1251] eta 0:09:35 lr 0.000011 time 0.2905 (0.4870) loss 2.9637 (2.9719) grad_norm 3.0433 (3.3974) [2021-04-16 21:33:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][80/1251] eta 0:08:59 lr 0.000011 time 0.2665 (0.4610) loss 3.2320 (2.9782) grad_norm 4.1851 (3.4320) [2021-04-16 21:33:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][90/1251] eta 0:08:31 lr 0.000011 time 0.2768 (0.4407) loss 3.1667 (2.9866) grad_norm 4.1535 (3.4634) [2021-04-16 21:33:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][100/1251] eta 0:08:09 lr 0.000011 time 0.2944 (0.4256) loss 2.2058 (2.9909) grad_norm 3.8005 (3.4840) [2021-04-16 21:34:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][110/1251] eta 0:07:50 lr 0.000011 time 0.3026 (0.4124) loss 3.4893 (2.9822) grad_norm 5.7639 (3.4919) [2021-04-16 21:34:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][120/1251] eta 0:07:33 lr 0.000011 time 0.2932 (0.4010) loss 3.6448 (2.9891) grad_norm 3.7235 (3.4958) [2021-04-16 21:34:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][130/1251] eta 0:07:18 lr 0.000011 time 0.2656 (0.3912) loss 3.0862 (2.9866) grad_norm 2.8631 (3.4795) [2021-04-16 21:34:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][140/1251] eta 0:07:06 lr 0.000011 time 0.3060 (0.3841) loss 3.0475 (2.9702) grad_norm 3.5549 (3.4719) [2021-04-16 21:34:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][150/1251] eta 0:06:55 lr 0.000011 time 0.2484 (0.3777) loss 3.4579 (2.9808) grad_norm 3.3517 (3.4808) [2021-04-16 21:34:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][160/1251] eta 0:06:45 lr 0.000011 time 0.2533 (0.3716) loss 3.0550 (2.9754) grad_norm 3.4410 (3.4793) [2021-04-16 21:34:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][170/1251] eta 0:06:36 lr 0.000011 time 0.2901 (0.3664) loss 3.2053 (2.9634) grad_norm 3.2386 (3.4663) [2021-04-16 21:34:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][180/1251] eta 0:06:27 lr 0.000011 time 0.2987 (0.3620) loss 3.0844 (2.9559) grad_norm 3.5040 (3.4767) [2021-04-16 21:34:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][190/1251] eta 0:06:19 lr 0.000011 time 0.2926 (0.3577) loss 3.2870 (2.9523) grad_norm 3.3369 (3.4697) [2021-04-16 21:34:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][200/1251] eta 0:06:11 lr 0.000011 time 0.2734 (0.3537) loss 3.0929 (2.9533) grad_norm 3.3589 (3.4710) [2021-04-16 21:34:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][210/1251] eta 0:06:04 lr 0.000011 time 0.2722 (0.3503) loss 2.8281 (2.9371) grad_norm 2.7237 (3.4731) [2021-04-16 21:34:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][220/1251] eta 0:05:57 lr 0.000011 time 0.3015 (0.3470) loss 2.9961 (2.9471) grad_norm 3.8364 (3.4724) [2021-04-16 21:34:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][230/1251] eta 0:05:51 lr 0.000011 time 0.2782 (0.3440) loss 3.3739 (2.9413) grad_norm 3.2837 (3.4636) [2021-04-16 21:34:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][240/1251] eta 0:05:45 lr 0.000011 time 0.2759 (0.3416) loss 2.7193 (2.9493) grad_norm 3.0669 (3.4556) [2021-04-16 21:34:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][250/1251] eta 0:05:39 lr 0.000011 time 0.2501 (0.3389) loss 2.6983 (2.9417) grad_norm 3.5905 (3.4589) [2021-04-16 21:34:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][260/1251] eta 0:05:33 lr 0.000011 time 0.2783 (0.3366) loss 3.2693 (2.9428) grad_norm 3.0264 (3.4573) [2021-04-16 21:34:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][270/1251] eta 0:05:28 lr 0.000011 time 0.2889 (0.3344) loss 2.5322 (2.9471) grad_norm 3.0757 (3.4645) [2021-04-16 21:34:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][280/1251] eta 0:05:22 lr 0.000011 time 0.2438 (0.3323) loss 3.2955 (2.9516) grad_norm 3.7003 (3.4644) [2021-04-16 21:34:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][290/1251] eta 0:05:17 lr 0.000011 time 0.2698 (0.3309) loss 3.5969 (2.9545) grad_norm 3.2201 (3.4664) [2021-04-16 21:34:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][300/1251] eta 0:05:12 lr 0.000011 time 0.2709 (0.3291) loss 3.2133 (2.9609) grad_norm 3.4578 (3.4651) [2021-04-16 21:34:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][310/1251] eta 0:05:08 lr 0.000011 time 0.2981 (0.3276) loss 2.6778 (2.9554) grad_norm 2.8923 (3.4619) [2021-04-16 21:35:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][320/1251] eta 0:05:03 lr 0.000011 time 0.2805 (0.3261) loss 2.5716 (2.9565) grad_norm 3.0294 (3.4570) [2021-04-16 21:35:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][330/1251] eta 0:04:59 lr 0.000011 time 0.2774 (0.3247) loss 3.6625 (2.9600) grad_norm 3.2466 (3.4647) [2021-04-16 21:35:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][340/1251] eta 0:04:54 lr 0.000011 time 0.2597 (0.3232) loss 2.5672 (2.9591) grad_norm 3.1862 (3.4720) [2021-04-16 21:35:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][350/1251] eta 0:04:50 lr 0.000011 time 0.2685 (0.3224) loss 3.4631 (2.9541) grad_norm 3.3361 (3.4700) [2021-04-16 21:35:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][360/1251] eta 0:04:46 lr 0.000011 time 0.2696 (0.3218) loss 3.0656 (2.9518) grad_norm 3.9067 (3.4770) [2021-04-16 21:35:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][370/1251] eta 0:04:43 lr 0.000011 time 0.2855 (0.3214) loss 2.9627 (2.9512) grad_norm 3.2537 (3.4806) [2021-04-16 21:35:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][380/1251] eta 0:04:38 lr 0.000011 time 0.2662 (0.3203) loss 3.1999 (2.9531) grad_norm 3.3000 (3.4829) [2021-04-16 21:35:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][390/1251] eta 0:04:34 lr 0.000011 time 0.2806 (0.3192) loss 3.1601 (2.9498) grad_norm 2.8793 (3.4854) [2021-04-16 21:35:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][400/1251] eta 0:04:30 lr 0.000011 time 0.2857 (0.3182) loss 3.0310 (2.9441) grad_norm 3.3559 (3.4850) [2021-04-16 21:35:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][410/1251] eta 0:04:26 lr 0.000011 time 0.2493 (0.3171) loss 3.5489 (2.9402) grad_norm 4.6802 (3.4838) [2021-04-16 21:35:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][420/1251] eta 0:04:22 lr 0.000011 time 0.2683 (0.3161) loss 3.0480 (2.9458) grad_norm 3.4561 (3.4797) [2021-04-16 21:35:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][430/1251] eta 0:04:18 lr 0.000011 time 0.2969 (0.3152) loss 3.1720 (2.9465) grad_norm 3.1864 (3.4783) [2021-04-16 21:35:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][440/1251] eta 0:04:14 lr 0.000011 time 0.2615 (0.3142) loss 3.0694 (2.9504) grad_norm 3.4575 (3.4784) [2021-04-16 21:35:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][450/1251] eta 0:04:10 lr 0.000011 time 0.2718 (0.3133) loss 2.9202 (2.9488) grad_norm 3.3961 (3.4812) [2021-04-16 21:35:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][460/1251] eta 0:04:07 lr 0.000011 time 0.2830 (0.3125) loss 2.9328 (2.9513) grad_norm 3.8254 (3.4825) [2021-04-16 21:35:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][470/1251] eta 0:04:03 lr 0.000011 time 0.2891 (0.3120) loss 3.1309 (2.9488) grad_norm 3.8167 (3.4756) [2021-04-16 21:35:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][480/1251] eta 0:04:00 lr 0.000011 time 0.2595 (0.3115) loss 2.7308 (2.9445) grad_norm 3.1999 (3.4723) [2021-04-16 21:35:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][490/1251] eta 0:03:56 lr 0.000011 time 0.2805 (0.3109) loss 3.2057 (2.9444) grad_norm 3.0069 (3.4646) [2021-04-16 21:35:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][500/1251] eta 0:03:52 lr 0.000011 time 0.2820 (0.3102) loss 3.1600 (2.9409) grad_norm 3.3561 (3.4616) [2021-04-16 21:35:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][510/1251] eta 0:03:49 lr 0.000011 time 0.2973 (0.3094) loss 3.6994 (2.9459) grad_norm 3.4197 (3.4649) [2021-04-16 21:35:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][520/1251] eta 0:03:45 lr 0.000011 time 0.2913 (0.3089) loss 2.8262 (2.9471) grad_norm 4.0445 (3.4640) [2021-04-16 21:35:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][530/1251] eta 0:03:42 lr 0.000011 time 0.2620 (0.3083) loss 3.0118 (2.9493) grad_norm 3.0094 (3.4607) [2021-04-16 21:36:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][540/1251] eta 0:03:38 lr 0.000011 time 0.2953 (0.3078) loss 2.8509 (2.9505) grad_norm 4.0537 (3.4596) [2021-04-16 21:36:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][550/1251] eta 0:03:35 lr 0.000011 time 0.3008 (0.3073) loss 3.1838 (2.9507) grad_norm 2.7529 (3.4618) [2021-04-16 21:36:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][560/1251] eta 0:03:32 lr 0.000011 time 0.2871 (0.3069) loss 3.1144 (2.9479) grad_norm 2.9298 (3.4626) [2021-04-16 21:36:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][570/1251] eta 0:03:28 lr 0.000011 time 0.2654 (0.3063) loss 3.1890 (2.9516) grad_norm 3.1190 (3.4605) [2021-04-16 21:36:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][580/1251] eta 0:03:25 lr 0.000011 time 0.2882 (0.3063) loss 3.2087 (2.9530) grad_norm 3.1330 (3.4625) [2021-04-16 21:36:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][590/1251] eta 0:03:22 lr 0.000011 time 0.2940 (0.3058) loss 2.7287 (2.9564) grad_norm 3.3894 (3.4586) [2021-04-16 21:36:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][600/1251] eta 0:03:18 lr 0.000011 time 0.2681 (0.3053) loss 2.6117 (2.9584) grad_norm 3.3226 (3.4556) [2021-04-16 21:36:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][610/1251] eta 0:03:15 lr 0.000011 time 0.2682 (0.3048) loss 3.0584 (2.9623) grad_norm 3.2704 (3.4555) [2021-04-16 21:36:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][620/1251] eta 0:03:12 lr 0.000011 time 0.2626 (0.3045) loss 2.9625 (2.9641) grad_norm 3.0919 (3.4528) [2021-04-16 21:36:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][630/1251] eta 0:03:08 lr 0.000011 time 0.2765 (0.3040) loss 3.2519 (2.9664) grad_norm 2.9795 (3.4554) [2021-04-16 21:36:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][640/1251] eta 0:03:05 lr 0.000011 time 0.3073 (0.3038) loss 3.5047 (2.9708) grad_norm 3.1489 (3.4691) [2021-04-16 21:36:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][650/1251] eta 0:03:02 lr 0.000011 time 0.2647 (0.3034) loss 3.6356 (2.9690) grad_norm 3.4129 (3.4736) [2021-04-16 21:36:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][660/1251] eta 0:02:59 lr 0.000011 time 0.2994 (0.3031) loss 2.7483 (2.9719) grad_norm 3.2539 (3.4728) [2021-04-16 21:36:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][670/1251] eta 0:02:55 lr 0.000011 time 0.2888 (0.3029) loss 2.8707 (2.9719) grad_norm 3.1219 (3.4752) [2021-04-16 21:36:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][680/1251] eta 0:02:52 lr 0.000011 time 0.2613 (0.3026) loss 3.2631 (2.9743) grad_norm 3.7647 (3.4749) [2021-04-16 21:36:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][690/1251] eta 0:02:49 lr 0.000011 time 0.3024 (0.3022) loss 3.5171 (2.9748) grad_norm 3.5977 (3.4761) [2021-04-16 21:36:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][700/1251] eta 0:02:46 lr 0.000011 time 0.2657 (0.3019) loss 2.9387 (2.9756) grad_norm 3.5160 (3.4763) [2021-04-16 21:36:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][710/1251] eta 0:02:43 lr 0.000011 time 0.2873 (0.3015) loss 3.3707 (2.9770) grad_norm 3.3357 (3.4749) [2021-04-16 21:36:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][720/1251] eta 0:02:40 lr 0.000011 time 0.2587 (0.3015) loss 2.6325 (2.9752) grad_norm 3.2491 (3.4755) [2021-04-16 21:36:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][730/1251] eta 0:02:36 lr 0.000011 time 0.2854 (0.3011) loss 2.7453 (2.9701) grad_norm 2.6484 (3.4725) [2021-04-16 21:36:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][740/1251] eta 0:02:33 lr 0.000011 time 0.2937 (0.3008) loss 3.8560 (2.9713) grad_norm 3.2795 (3.4749) [2021-04-16 21:37:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][750/1251] eta 0:02:30 lr 0.000011 time 0.2694 (0.3004) loss 3.0709 (2.9727) grad_norm 3.0106 (3.4748) [2021-04-16 21:37:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][760/1251] eta 0:02:27 lr 0.000011 time 0.2898 (0.3001) loss 3.2968 (2.9719) grad_norm 3.1460 (3.4736) [2021-04-16 21:37:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][770/1251] eta 0:02:24 lr 0.000011 time 0.2934 (0.2999) loss 2.9527 (2.9713) grad_norm 2.9412 (3.4717) [2021-04-16 21:37:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][780/1251] eta 0:02:21 lr 0.000011 time 0.2810 (0.2997) loss 2.2440 (2.9702) grad_norm 3.2998 (3.4708) [2021-04-16 21:37:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][790/1251] eta 0:02:18 lr 0.000011 time 0.2662 (0.2994) loss 3.2673 (2.9704) grad_norm 4.0015 (3.4708) [2021-04-16 21:37:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][800/1251] eta 0:02:14 lr 0.000011 time 0.2779 (0.2991) loss 3.5683 (2.9740) grad_norm 3.5507 (3.4705) [2021-04-16 21:37:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][810/1251] eta 0:02:11 lr 0.000011 time 0.2489 (0.2988) loss 3.2404 (2.9744) grad_norm 3.3929 (3.4679) [2021-04-16 21:37:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][820/1251] eta 0:02:08 lr 0.000011 time 0.2882 (0.2986) loss 3.4423 (2.9761) grad_norm 3.2581 (3.4733) [2021-04-16 21:37:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][830/1251] eta 0:02:05 lr 0.000011 time 0.2893 (0.2984) loss 2.7146 (2.9764) grad_norm 3.3494 (3.4763) [2021-04-16 21:37:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][840/1251] eta 0:02:02 lr 0.000011 time 0.2510 (0.2982) loss 3.7915 (2.9773) grad_norm 3.2126 (3.4788) [2021-04-16 21:37:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][850/1251] eta 0:01:59 lr 0.000011 time 0.2991 (0.2980) loss 2.7028 (2.9743) grad_norm 3.1386 (3.4782) [2021-04-16 21:37:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][860/1251] eta 0:01:56 lr 0.000011 time 0.2689 (0.2977) loss 3.1069 (2.9747) grad_norm 3.3236 (3.4768) [2021-04-16 21:37:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][870/1251] eta 0:01:53 lr 0.000011 time 0.2899 (0.2977) loss 2.8109 (2.9722) grad_norm 4.1235 (3.4770) [2021-04-16 21:37:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][880/1251] eta 0:01:50 lr 0.000011 time 0.2511 (0.2974) loss 3.5147 (2.9730) grad_norm 3.6304 (3.4780) [2021-04-16 21:37:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][890/1251] eta 0:01:47 lr 0.000011 time 0.2746 (0.2971) loss 3.3624 (2.9738) grad_norm 2.9607 (3.4761) [2021-04-16 21:37:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][900/1251] eta 0:01:44 lr 0.000011 time 0.2628 (0.2969) loss 2.9727 (2.9755) grad_norm 3.1708 (3.4777) [2021-04-16 21:37:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][910/1251] eta 0:01:41 lr 0.000011 time 0.2842 (0.2967) loss 2.9505 (2.9761) grad_norm 3.1058 (3.4753) [2021-04-16 21:37:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][920/1251] eta 0:01:38 lr 0.000011 time 0.2743 (0.2965) loss 3.1958 (2.9756) grad_norm 3.0755 (3.4776) [2021-04-16 21:37:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][930/1251] eta 0:01:35 lr 0.000011 time 0.2597 (0.2964) loss 2.3214 (2.9757) grad_norm 2.9617 (3.4764) [2021-04-16 21:37:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][940/1251] eta 0:01:32 lr 0.000011 time 0.4539 (0.2967) loss 2.9164 (2.9775) grad_norm 3.0379 (3.4768) [2021-04-16 21:37:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][950/1251] eta 0:01:29 lr 0.000011 time 0.2807 (0.2965) loss 2.9027 (2.9776) grad_norm 4.1858 (3.4759) [2021-04-16 21:38:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][960/1251] eta 0:01:26 lr 0.000011 time 0.3044 (0.2964) loss 2.6100 (2.9745) grad_norm 3.7323 (3.4818) [2021-04-16 21:38:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][970/1251] eta 0:01:23 lr 0.000011 time 0.2680 (0.2962) loss 3.1403 (2.9750) grad_norm 3.8036 (3.4811) [2021-04-16 21:38:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][980/1251] eta 0:01:20 lr 0.000011 time 0.2884 (0.2960) loss 3.1555 (2.9737) grad_norm 3.4877 (3.4804) [2021-04-16 21:38:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][990/1251] eta 0:01:17 lr 0.000011 time 0.2906 (0.2958) loss 3.0201 (2.9724) grad_norm 3.0295 (3.4808) [2021-04-16 21:38:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][1000/1251] eta 0:01:14 lr 0.000011 time 0.2823 (0.2955) loss 2.9297 (2.9712) grad_norm 3.5233 (3.4810) [2021-04-16 21:38:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][1010/1251] eta 0:01:11 lr 0.000011 time 0.2827 (0.2953) loss 2.6810 (2.9700) grad_norm 4.5981 (3.4814) [2021-04-16 21:38:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][1020/1251] eta 0:01:08 lr 0.000011 time 0.2580 (0.2951) loss 3.4751 (2.9721) grad_norm 3.5380 (3.4830) [2021-04-16 21:38:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][1030/1251] eta 0:01:05 lr 0.000011 time 0.2914 (0.2950) loss 2.7626 (2.9720) grad_norm 3.0404 (3.4818) [2021-04-16 21:38:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][1040/1251] eta 0:01:02 lr 0.000011 time 0.2773 (0.2948) loss 3.4396 (2.9722) grad_norm 4.3926 (3.4850) [2021-04-16 21:38:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][1050/1251] eta 0:00:59 lr 0.000011 time 0.2893 (0.2946) loss 3.0177 (2.9724) grad_norm 3.3971 (3.4869) [2021-04-16 21:38:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][1060/1251] eta 0:00:56 lr 0.000011 time 0.2779 (0.2944) loss 2.8629 (2.9715) grad_norm 3.2872 (3.4853) [2021-04-16 21:38:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][1070/1251] eta 0:00:53 lr 0.000011 time 0.2980 (0.2943) loss 1.8546 (2.9713) grad_norm 3.5056 (3.4853) [2021-04-16 21:38:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][1080/1251] eta 0:00:50 lr 0.000011 time 0.2744 (0.2941) loss 2.9374 (2.9708) grad_norm 3.2590 (3.4847) [2021-04-16 21:38:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][1090/1251] eta 0:00:47 lr 0.000011 time 0.2688 (0.2940) loss 3.0735 (2.9705) grad_norm 2.9138 (3.4853) [2021-04-16 21:38:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][1100/1251] eta 0:00:44 lr 0.000011 time 0.2740 (0.2938) loss 2.3234 (2.9694) grad_norm 2.9615 (3.4838) [2021-04-16 21:38:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][1110/1251] eta 0:00:41 lr 0.000011 time 0.2902 (0.2937) loss 3.7524 (2.9685) grad_norm 2.8452 (3.4823) [2021-04-16 21:38:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][1120/1251] eta 0:00:38 lr 0.000011 time 0.2623 (0.2935) loss 3.3369 (2.9664) grad_norm 3.1019 (3.4835) [2021-04-16 21:38:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][1130/1251] eta 0:00:35 lr 0.000011 time 0.2630 (0.2933) loss 2.2545 (2.9659) grad_norm 2.9397 (3.4831) [2021-04-16 21:38:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][1140/1251] eta 0:00:32 lr 0.000011 time 0.2691 (0.2932) loss 3.2242 (2.9647) grad_norm 3.1625 (3.4840) [2021-04-16 21:38:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][1150/1251] eta 0:00:29 lr 0.000011 time 0.3000 (0.2931) loss 3.3254 (2.9646) grad_norm 3.1548 (3.4851) [2021-04-16 21:38:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][1160/1251] eta 0:00:26 lr 0.000011 time 0.2700 (0.2931) loss 2.8391 (2.9635) grad_norm 3.0846 (3.4838) [2021-04-16 21:38:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][1170/1251] eta 0:00:23 lr 0.000011 time 0.2598 (0.2930) loss 2.3911 (2.9628) grad_norm 2.9980 (3.4830) [2021-04-16 21:39:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][1180/1251] eta 0:00:20 lr 0.000011 time 0.2941 (0.2929) loss 3.0578 (2.9635) grad_norm 3.3735 (3.4815) [2021-04-16 21:39:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][1190/1251] eta 0:00:17 lr 0.000011 time 0.2927 (0.2929) loss 3.8036 (2.9650) grad_norm 3.0100 (3.4803) [2021-04-16 21:39:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][1200/1251] eta 0:00:14 lr 0.000011 time 0.2661 (0.2928) loss 3.3753 (2.9663) grad_norm 3.5531 (3.4796) [2021-04-16 21:39:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][1210/1251] eta 0:00:11 lr 0.000011 time 0.2463 (0.2926) loss 2.0756 (2.9656) grad_norm 3.3065 (3.4790) [2021-04-16 21:39:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][1220/1251] eta 0:00:09 lr 0.000011 time 0.2922 (0.2925) loss 2.8619 (2.9670) grad_norm 3.5534 (3.4791) [2021-04-16 21:39:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][1230/1251] eta 0:00:06 lr 0.000011 time 0.2751 (0.2924) loss 2.5559 (2.9668) grad_norm 3.4782 (3.4791) [2021-04-16 21:39:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][1240/1251] eta 0:00:03 lr 0.000011 time 0.2474 (0.2922) loss 2.4401 (2.9664) grad_norm 3.8081 (3.4802) [2021-04-16 21:39:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [294/300][1250/1251] eta 0:00:00 lr 0.000011 time 0.2477 (0.2919) loss 3.7108 (2.9659) grad_norm 3.4063 (3.4801) [2021-04-16 21:39:37 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 294 training takes 0:06:22 [2021-04-16 21:39:37 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_294.pth saving...... [2021-04-16 21:39:55 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_294.pth saved !!! [2021-04-16 21:39:56 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.132 (1.132) Loss 0.8462 (0.8462) Acc@1 81.445 (81.445) Acc@5 95.605 (95.605) [2021-04-16 21:39:57 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.211 (0.220) Loss 0.7313 (0.8066) Acc@1 83.301 (81.401) Acc@5 96.387 (95.570) [2021-04-16 21:39:59 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.812 (0.235) Loss 0.8109 (0.8172) Acc@1 80.371 (80.948) Acc@5 96.484 (95.545) [2021-04-16 21:40:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.079 (0.255) Loss 0.8297 (0.8239) Acc@1 80.957 (80.869) Acc@5 95.020 (95.439) [2021-04-16 21:40:04 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.220) Loss 0.8200 (0.8182) Acc@1 81.152 (81.007) Acc@5 95.801 (95.498) [2021-04-16 21:40:28 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 81.076 Acc@5 95.510 [2021-04-16 21:40:28 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 81.1% [2021-04-16 21:40:28 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 81.24% [2021-04-16 21:40:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][0/1251] eta 0:31:28 lr 0.000011 time 1.5099 (1.5099) loss 2.4009 (2.4009) grad_norm 3.3358 (3.3358) [2021-04-16 21:40:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][10/1251] eta 0:07:56 lr 0.000011 time 0.3000 (0.3841) loss 3.3851 (2.8531) grad_norm 4.6752 (3.7082) [2021-04-16 21:40:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][20/1251] eta 0:06:53 lr 0.000011 time 0.2753 (0.3360) loss 3.7185 (3.0243) grad_norm 3.5130 (3.6217) [2021-04-16 21:40:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][30/1251] eta 0:06:26 lr 0.000011 time 0.2643 (0.3162) loss 2.6170 (2.9905) grad_norm 3.7999 (3.4894) [2021-04-16 21:40:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][40/1251] eta 0:06:10 lr 0.000011 time 0.2671 (0.3056) loss 3.4821 (2.9694) grad_norm 3.3881 (3.5723) [2021-04-16 21:40:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][50/1251] eta 0:06:01 lr 0.000011 time 0.2795 (0.3012) loss 3.6768 (2.9856) grad_norm 3.3958 (3.6237) [2021-04-16 21:40:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][60/1251] eta 0:05:57 lr 0.000011 time 0.4268 (0.3003) loss 2.9592 (2.9948) grad_norm 3.2318 (3.6438) [2021-04-16 21:40:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][70/1251] eta 0:05:49 lr 0.000011 time 0.2556 (0.2957) loss 3.3045 (2.9908) grad_norm 3.3754 (3.6071) [2021-04-16 21:40:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][80/1251] eta 0:05:43 lr 0.000011 time 0.2602 (0.2934) loss 2.3812 (3.0202) grad_norm 4.3390 (3.6118) [2021-04-16 21:40:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][90/1251] eta 0:05:38 lr 0.000011 time 0.2545 (0.2915) loss 3.1572 (3.0022) grad_norm 3.1905 (3.6264) [2021-04-16 21:40:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][100/1251] eta 0:05:33 lr 0.000011 time 0.2665 (0.2897) loss 3.4213 (2.9873) grad_norm 3.2445 (3.6214) [2021-04-16 21:41:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][110/1251] eta 0:05:28 lr 0.000011 time 0.2661 (0.2883) loss 3.5332 (2.9649) grad_norm 4.3864 (3.6171) [2021-04-16 21:41:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][120/1251] eta 0:05:25 lr 0.000011 time 0.2664 (0.2874) loss 2.3490 (2.9509) grad_norm 3.2176 (3.5963) [2021-04-16 21:41:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][130/1251] eta 0:05:22 lr 0.000011 time 0.2810 (0.2876) loss 3.2313 (2.9474) grad_norm 3.2511 (3.5978) [2021-04-16 21:41:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][140/1251] eta 0:05:20 lr 0.000011 time 0.2602 (0.2884) loss 2.5780 (2.9464) grad_norm 3.3379 (3.5828) [2021-04-16 21:41:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][150/1251] eta 0:05:18 lr 0.000011 time 0.4325 (0.2890) loss 3.7104 (2.9560) grad_norm 3.2711 (3.5718) [2021-04-16 21:41:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][160/1251] eta 0:05:14 lr 0.000011 time 0.2654 (0.2881) loss 3.3782 (2.9663) grad_norm 3.0966 (3.5607) [2021-04-16 21:41:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][170/1251] eta 0:05:10 lr 0.000011 time 0.2568 (0.2873) loss 3.2784 (2.9780) grad_norm 3.9349 (3.5659) [2021-04-16 21:41:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][180/1251] eta 0:05:06 lr 0.000011 time 0.2711 (0.2865) loss 2.6229 (2.9607) grad_norm 3.0765 (3.5473) [2021-04-16 21:41:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][190/1251] eta 0:05:03 lr 0.000011 time 0.2589 (0.2861) loss 2.3716 (2.9647) grad_norm 2.8497 (3.5383) [2021-04-16 21:41:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][200/1251] eta 0:05:00 lr 0.000011 time 0.2652 (0.2855) loss 3.0468 (2.9632) grad_norm 3.0894 (3.5483) [2021-04-16 21:41:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][210/1251] eta 0:04:56 lr 0.000011 time 0.2824 (0.2852) loss 2.8562 (2.9664) grad_norm 3.7370 (3.5527) [2021-04-16 21:41:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][220/1251] eta 0:04:53 lr 0.000011 time 0.2868 (0.2847) loss 3.4982 (2.9672) grad_norm 3.2722 (3.5440) [2021-04-16 21:41:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][230/1251] eta 0:04:50 lr 0.000011 time 0.2678 (0.2844) loss 3.2350 (2.9753) grad_norm 5.1900 (3.5480) [2021-04-16 21:41:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][240/1251] eta 0:04:47 lr 0.000011 time 0.2908 (0.2839) loss 3.1473 (2.9687) grad_norm 3.2907 (3.5421) [2021-04-16 21:41:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][250/1251] eta 0:04:43 lr 0.000011 time 0.2885 (0.2836) loss 2.8438 (2.9654) grad_norm 5.1226 (3.5428) [2021-04-16 21:41:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][260/1251] eta 0:04:40 lr 0.000011 time 0.2707 (0.2835) loss 3.1207 (2.9638) grad_norm 3.8918 (3.5319) [2021-04-16 21:41:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][270/1251] eta 0:04:37 lr 0.000011 time 0.2739 (0.2832) loss 2.9854 (2.9593) grad_norm 3.8550 (3.5333) [2021-04-16 21:41:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][280/1251] eta 0:04:34 lr 0.000011 time 0.2847 (0.2828) loss 2.7391 (2.9571) grad_norm 3.4153 (3.5369) [2021-04-16 21:41:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][290/1251] eta 0:04:31 lr 0.000011 time 0.3024 (0.2829) loss 3.3477 (2.9613) grad_norm 3.1751 (3.5436) [2021-04-16 21:41:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][300/1251] eta 0:04:28 lr 0.000011 time 0.2888 (0.2827) loss 2.3022 (2.9599) grad_norm 3.8588 (3.5407) [2021-04-16 21:41:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][310/1251] eta 0:04:25 lr 0.000011 time 0.2633 (0.2826) loss 3.6150 (2.9582) grad_norm 4.1071 (3.5419) [2021-04-16 21:41:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][320/1251] eta 0:04:22 lr 0.000011 time 0.2892 (0.2823) loss 3.0031 (2.9611) grad_norm 3.2000 (3.5480) [2021-04-16 21:42:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][330/1251] eta 0:04:19 lr 0.000011 time 0.2804 (0.2820) loss 2.7857 (2.9644) grad_norm 3.7052 (3.5403) [2021-04-16 21:42:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][340/1251] eta 0:04:16 lr 0.000011 time 0.2577 (0.2817) loss 2.8533 (2.9625) grad_norm 3.3765 (3.5369) [2021-04-16 21:42:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][350/1251] eta 0:04:13 lr 0.000011 time 0.3016 (0.2817) loss 3.2401 (2.9684) grad_norm 2.7414 (3.5290) [2021-04-16 21:42:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][360/1251] eta 0:04:11 lr 0.000011 time 0.2748 (0.2818) loss 3.3859 (2.9738) grad_norm 3.0454 (3.5296) [2021-04-16 21:42:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][370/1251] eta 0:04:08 lr 0.000011 time 0.2764 (0.2820) loss 2.9074 (2.9712) grad_norm 3.1156 (3.5289) [2021-04-16 21:42:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][380/1251] eta 0:04:05 lr 0.000011 time 0.2856 (0.2819) loss 3.1448 (2.9707) grad_norm 3.1797 (3.5277) [2021-04-16 21:42:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][390/1251] eta 0:04:02 lr 0.000011 time 0.2462 (0.2820) loss 2.2278 (2.9671) grad_norm 3.0349 (3.5229) [2021-04-16 21:42:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][400/1251] eta 0:03:59 lr 0.000011 time 0.2677 (0.2819) loss 3.2204 (2.9664) grad_norm 3.2396 (3.5206) [2021-04-16 21:42:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][410/1251] eta 0:03:57 lr 0.000011 time 0.2878 (0.2819) loss 2.5773 (2.9629) grad_norm 3.8133 (3.5193) [2021-04-16 21:42:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][420/1251] eta 0:03:54 lr 0.000011 time 0.2786 (0.2822) loss 3.4141 (2.9567) grad_norm 2.9835 (3.5132) [2021-04-16 21:42:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][430/1251] eta 0:03:51 lr 0.000011 time 0.2716 (0.2821) loss 3.2746 (2.9600) grad_norm 3.2653 (3.5110) [2021-04-16 21:42:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][440/1251] eta 0:03:48 lr 0.000011 time 0.2851 (0.2820) loss 2.6366 (2.9586) grad_norm 3.4115 (3.5078) [2021-04-16 21:42:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][450/1251] eta 0:03:45 lr 0.000011 time 0.2789 (0.2818) loss 3.5005 (2.9595) grad_norm 3.4322 (3.5115) [2021-04-16 21:42:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][460/1251] eta 0:03:42 lr 0.000011 time 0.2842 (0.2816) loss 3.0532 (2.9555) grad_norm 4.1931 (3.5182) [2021-04-16 21:42:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][470/1251] eta 0:03:40 lr 0.000011 time 0.2604 (0.2820) loss 2.1752 (2.9529) grad_norm 2.9237 (3.5139) [2021-04-16 21:42:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][480/1251] eta 0:03:37 lr 0.000011 time 0.2872 (0.2821) loss 3.1723 (2.9524) grad_norm 3.2473 (3.5109) [2021-04-16 21:42:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][490/1251] eta 0:03:34 lr 0.000011 time 0.2869 (0.2819) loss 3.1473 (2.9544) grad_norm 2.8551 (3.5035) [2021-04-16 21:42:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][500/1251] eta 0:03:31 lr 0.000011 time 0.2905 (0.2817) loss 2.8480 (2.9535) grad_norm 3.4364 (3.5037) [2021-04-16 21:42:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][510/1251] eta 0:03:28 lr 0.000011 time 0.2583 (0.2815) loss 3.6107 (2.9588) grad_norm 3.1616 (3.5044) [2021-04-16 21:42:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][520/1251] eta 0:03:25 lr 0.000011 time 0.2672 (0.2814) loss 2.2745 (2.9559) grad_norm 3.8273 (3.5019) [2021-04-16 21:42:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][530/1251] eta 0:03:22 lr 0.000011 time 0.2621 (0.2812) loss 1.9492 (2.9517) grad_norm 3.0300 (3.5022) [2021-04-16 21:43:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][540/1251] eta 0:03:19 lr 0.000011 time 0.2738 (0.2811) loss 3.1087 (2.9542) grad_norm 2.8417 (3.4983) [2021-04-16 21:43:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][550/1251] eta 0:03:16 lr 0.000011 time 0.2557 (0.2810) loss 2.8813 (2.9543) grad_norm 3.4775 (3.5023) [2021-04-16 21:43:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][560/1251] eta 0:03:14 lr 0.000011 time 0.2863 (0.2809) loss 2.9381 (2.9562) grad_norm 3.4760 (3.5023) [2021-04-16 21:43:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][570/1251] eta 0:03:11 lr 0.000011 time 0.2744 (0.2808) loss 3.2373 (2.9613) grad_norm 3.2603 (3.5078) [2021-04-16 21:43:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][580/1251] eta 0:03:08 lr 0.000011 time 0.2749 (0.2810) loss 2.3109 (2.9624) grad_norm 3.0202 (3.5070) [2021-04-16 21:43:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][590/1251] eta 0:03:05 lr 0.000011 time 0.2670 (0.2811) loss 2.7617 (2.9622) grad_norm 2.6221 (3.5089) [2021-04-16 21:43:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][600/1251] eta 0:03:02 lr 0.000011 time 0.3011 (0.2811) loss 2.9999 (2.9640) grad_norm 3.4400 (3.5083) [2021-04-16 21:43:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][610/1251] eta 0:03:00 lr 0.000011 time 0.3018 (0.2810) loss 1.8340 (2.9612) grad_norm 2.6794 (3.5045) [2021-04-16 21:43:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][620/1251] eta 0:02:57 lr 0.000011 time 0.2525 (0.2809) loss 2.5517 (2.9593) grad_norm 3.7688 (3.5031) [2021-04-16 21:43:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][630/1251] eta 0:02:54 lr 0.000011 time 0.2767 (0.2808) loss 3.4050 (2.9609) grad_norm 3.0639 (3.5007) [2021-04-16 21:43:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][640/1251] eta 0:02:51 lr 0.000011 time 0.2745 (0.2809) loss 3.1838 (2.9637) grad_norm 3.1715 (3.5008) [2021-04-16 21:43:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][650/1251] eta 0:02:48 lr 0.000011 time 0.2757 (0.2808) loss 2.1824 (2.9605) grad_norm 3.1891 (3.5012) [2021-04-16 21:43:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][660/1251] eta 0:02:45 lr 0.000011 time 0.2901 (0.2808) loss 3.1190 (2.9614) grad_norm 3.8324 (3.5016) [2021-04-16 21:43:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][670/1251] eta 0:02:43 lr 0.000011 time 0.2874 (0.2808) loss 2.6921 (2.9610) grad_norm 2.9879 (3.5017) [2021-04-16 21:43:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][680/1251] eta 0:02:40 lr 0.000011 time 0.2808 (0.2807) loss 3.2070 (2.9578) grad_norm 3.1898 (3.5033) [2021-04-16 21:43:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][690/1251] eta 0:02:37 lr 0.000011 time 0.2985 (0.2806) loss 2.8079 (2.9561) grad_norm 2.9313 (3.5027) [2021-04-16 21:43:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][700/1251] eta 0:02:34 lr 0.000011 time 0.3921 (0.2807) loss 3.0489 (2.9586) grad_norm 3.3408 (3.5022) [2021-04-16 21:43:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][710/1251] eta 0:02:31 lr 0.000011 time 0.3029 (0.2806) loss 3.1160 (2.9576) grad_norm 3.4353 (3.5065) [2021-04-16 21:43:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][720/1251] eta 0:02:29 lr 0.000011 time 0.2708 (0.2808) loss 2.7368 (2.9580) grad_norm 3.2882 (3.5067) [2021-04-16 21:43:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][730/1251] eta 0:02:26 lr 0.000011 time 0.2567 (0.2810) loss 2.3288 (2.9562) grad_norm 3.6531 (3.5066) [2021-04-16 21:43:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][740/1251] eta 0:02:23 lr 0.000011 time 0.2747 (0.2809) loss 2.0492 (2.9585) grad_norm 3.5672 (3.5058) [2021-04-16 21:43:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][750/1251] eta 0:02:20 lr 0.000011 time 0.2994 (0.2808) loss 2.3407 (2.9581) grad_norm 3.0663 (3.5064) [2021-04-16 21:44:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][760/1251] eta 0:02:17 lr 0.000011 time 0.2690 (0.2808) loss 3.3306 (2.9619) grad_norm 4.2148 (nan) [2021-04-16 21:44:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][770/1251] eta 0:02:15 lr 0.000011 time 0.2625 (0.2807) loss 3.5007 (2.9604) grad_norm 3.5893 (nan) [2021-04-16 21:44:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][780/1251] eta 0:02:12 lr 0.000011 time 0.2783 (0.2806) loss 2.8692 (2.9605) grad_norm 2.9930 (nan) [2021-04-16 21:44:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][790/1251] eta 0:02:09 lr 0.000011 time 0.2829 (0.2806) loss 2.6477 (2.9581) grad_norm 3.2620 (nan) [2021-04-16 21:44:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][800/1251] eta 0:02:06 lr 0.000011 time 0.2706 (0.2805) loss 2.8663 (2.9566) grad_norm 4.0211 (nan) [2021-04-16 21:44:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][810/1251] eta 0:02:03 lr 0.000011 time 0.2896 (0.2805) loss 3.6086 (2.9592) grad_norm 4.7063 (nan) [2021-04-16 21:44:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][820/1251] eta 0:02:00 lr 0.000011 time 0.2850 (0.2805) loss 2.8389 (2.9602) grad_norm 4.7834 (nan) [2021-04-16 21:44:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][830/1251] eta 0:01:58 lr 0.000011 time 0.2807 (0.2804) loss 2.5702 (2.9596) grad_norm 2.8581 (nan) [2021-04-16 21:44:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][840/1251] eta 0:01:55 lr 0.000011 time 0.2931 (0.2804) loss 2.6830 (2.9609) grad_norm 4.0594 (nan) [2021-04-16 21:44:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][850/1251] eta 0:01:52 lr 0.000011 time 0.2719 (0.2804) loss 3.5707 (2.9636) grad_norm 4.3284 (nan) [2021-04-16 21:44:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][860/1251] eta 0:01:49 lr 0.000011 time 0.2609 (0.2804) loss 2.7660 (2.9619) grad_norm 3.6448 (nan) [2021-04-16 21:44:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][870/1251] eta 0:01:46 lr 0.000011 time 0.2597 (0.2803) loss 3.4924 (2.9615) grad_norm 3.4371 (nan) [2021-04-16 21:44:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][880/1251] eta 0:01:43 lr 0.000011 time 0.3076 (0.2803) loss 3.3665 (2.9637) grad_norm 3.3651 (nan) [2021-04-16 21:44:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][890/1251] eta 0:01:41 lr 0.000010 time 0.2767 (0.2802) loss 3.2459 (2.9635) grad_norm 3.0927 (nan) [2021-04-16 21:44:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][900/1251] eta 0:01:38 lr 0.000010 time 0.3117 (0.2802) loss 2.0513 (2.9635) grad_norm 3.1976 (nan) [2021-04-16 21:44:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][910/1251] eta 0:01:35 lr 0.000010 time 0.2911 (0.2801) loss 3.2122 (2.9633) grad_norm 4.2543 (nan) [2021-04-16 21:44:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][920/1251] eta 0:01:32 lr 0.000010 time 0.2931 (0.2801) loss 3.7317 (2.9625) grad_norm 3.0970 (nan) [2021-04-16 21:44:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][930/1251] eta 0:01:29 lr 0.000010 time 0.2684 (0.2802) loss 2.9997 (2.9618) grad_norm 3.2759 (nan) [2021-04-16 21:44:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][940/1251] eta 0:01:27 lr 0.000010 time 0.2769 (0.2803) loss 2.8333 (2.9617) grad_norm 3.2192 (nan) [2021-04-16 21:44:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][950/1251] eta 0:01:24 lr 0.000010 time 0.2642 (0.2803) loss 2.5238 (2.9592) grad_norm 3.3355 (nan) [2021-04-16 21:44:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][960/1251] eta 0:01:21 lr 0.000010 time 0.2809 (0.2802) loss 3.0020 (2.9610) grad_norm 3.4814 (nan) [2021-04-16 21:45:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][970/1251] eta 0:01:18 lr 0.000010 time 0.2710 (0.2803) loss 2.5324 (2.9602) grad_norm 3.5642 (nan) [2021-04-16 21:45:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][980/1251] eta 0:01:15 lr 0.000010 time 0.2776 (0.2802) loss 3.4823 (2.9614) grad_norm 3.2119 (nan) [2021-04-16 21:45:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][990/1251] eta 0:01:13 lr 0.000010 time 0.2605 (0.2803) loss 3.1567 (2.9605) grad_norm 3.9082 (nan) [2021-04-16 21:45:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][1000/1251] eta 0:01:10 lr 0.000010 time 0.2846 (0.2803) loss 2.7107 (2.9622) grad_norm 2.7405 (nan) [2021-04-16 21:45:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][1010/1251] eta 0:01:07 lr 0.000010 time 0.2924 (0.2803) loss 3.2040 (2.9630) grad_norm 4.6951 (nan) [2021-04-16 21:45:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][1020/1251] eta 0:01:04 lr 0.000010 time 0.2877 (0.2803) loss 3.3824 (2.9623) grad_norm 3.4787 (nan) [2021-04-16 21:45:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][1030/1251] eta 0:01:01 lr 0.000010 time 0.2523 (0.2802) loss 2.2674 (2.9619) grad_norm 3.2961 (nan) [2021-04-16 21:45:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][1040/1251] eta 0:00:59 lr 0.000010 time 0.2665 (0.2804) loss 3.3810 (2.9616) grad_norm 4.3741 (nan) [2021-04-16 21:45:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][1050/1251] eta 0:00:56 lr 0.000010 time 0.2991 (0.2803) loss 2.9031 (2.9602) grad_norm 3.3244 (nan) [2021-04-16 21:45:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][1060/1251] eta 0:00:53 lr 0.000010 time 0.2889 (0.2802) loss 2.8709 (2.9585) grad_norm 3.0816 (nan) [2021-04-16 21:45:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][1070/1251] eta 0:00:50 lr 0.000010 time 0.2765 (0.2802) loss 3.1238 (2.9593) grad_norm 3.4469 (nan) [2021-04-16 21:45:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][1080/1251] eta 0:00:47 lr 0.000010 time 0.2836 (0.2802) loss 3.3579 (2.9604) grad_norm 2.8852 (nan) [2021-04-16 21:45:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][1090/1251] eta 0:00:45 lr 0.000010 time 0.2819 (0.2802) loss 3.2989 (2.9615) grad_norm 3.0107 (nan) [2021-04-16 21:45:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][1100/1251] eta 0:00:42 lr 0.000010 time 0.2700 (0.2801) loss 3.3938 (2.9629) grad_norm 3.0682 (nan) [2021-04-16 21:45:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][1110/1251] eta 0:00:39 lr 0.000010 time 0.2628 (0.2801) loss 3.2735 (2.9645) grad_norm 3.4293 (nan) [2021-04-16 21:45:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][1120/1251] eta 0:00:36 lr 0.000010 time 0.2974 (0.2800) loss 3.5213 (2.9646) grad_norm 3.1091 (nan) [2021-04-16 21:45:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][1130/1251] eta 0:00:33 lr 0.000010 time 0.2762 (0.2800) loss 3.5133 (2.9655) grad_norm 3.0920 (nan) [2021-04-16 21:45:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][1140/1251] eta 0:00:31 lr 0.000010 time 0.2868 (0.2800) loss 2.6943 (2.9642) grad_norm 4.4189 (nan) [2021-04-16 21:45:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][1150/1251] eta 0:00:28 lr 0.000010 time 0.2894 (0.2799) loss 3.4853 (2.9651) grad_norm 3.5391 (nan) [2021-04-16 21:45:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][1160/1251] eta 0:00:25 lr 0.000010 time 0.2998 (0.2801) loss 2.2074 (2.9640) grad_norm 3.9610 (nan) [2021-04-16 21:45:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][1170/1251] eta 0:00:22 lr 0.000010 time 0.2602 (0.2802) loss 2.6223 (2.9632) grad_norm 4.0007 (nan) [2021-04-16 21:45:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][1180/1251] eta 0:00:19 lr 0.000010 time 0.2787 (0.2801) loss 2.3847 (2.9626) grad_norm 3.1475 (nan) [2021-04-16 21:46:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][1190/1251] eta 0:00:17 lr 0.000010 time 0.3022 (0.2801) loss 3.1304 (2.9632) grad_norm 3.7392 (nan) [2021-04-16 21:46:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][1200/1251] eta 0:00:14 lr 0.000010 time 0.2565 (0.2800) loss 2.6751 (2.9628) grad_norm 3.8801 (nan) [2021-04-16 21:46:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][1210/1251] eta 0:00:11 lr 0.000010 time 0.2943 (0.2800) loss 2.6316 (2.9628) grad_norm 7.8461 (nan) [2021-04-16 21:46:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][1220/1251] eta 0:00:08 lr 0.000010 time 0.2946 (0.2801) loss 3.3860 (2.9633) grad_norm 3.5100 (nan) [2021-04-16 21:46:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][1230/1251] eta 0:00:05 lr 0.000010 time 0.2619 (0.2801) loss 2.4264 (2.9634) grad_norm 3.5993 (nan) [2021-04-16 21:46:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][1240/1251] eta 0:00:03 lr 0.000010 time 0.2472 (0.2800) loss 3.9092 (2.9640) grad_norm 4.7665 (nan) [2021-04-16 21:46:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [295/300][1250/1251] eta 0:00:00 lr 0.000010 time 0.2483 (0.2797) loss 2.5540 (2.9653) grad_norm 3.5408 (nan) [2021-04-16 21:46:23 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 295 training takes 0:05:54 [2021-04-16 21:46:23 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_295.pth saving...... [2021-04-16 21:46:34 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_295.pth saved !!! [2021-04-16 21:46:35 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.116 (1.116) Loss 0.7413 (0.7413) Acc@1 83.789 (83.789) Acc@5 95.703 (95.703) [2021-04-16 21:46:37 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.102 (0.218) Loss 0.8616 (0.8111) Acc@1 80.469 (81.436) Acc@5 95.117 (95.614) [2021-04-16 21:46:39 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.230 (0.238) Loss 0.8519 (0.8112) Acc@1 81.250 (81.264) Acc@5 95.117 (95.559) [2021-04-16 21:46:42 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.288 (0.236) Loss 0.8289 (0.8108) Acc@1 80.371 (81.282) Acc@5 95.117 (95.514) [2021-04-16 21:46:43 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.214) Loss 0.8217 (0.8106) Acc@1 81.055 (81.257) Acc@5 95.898 (95.524) [2021-04-16 21:46:54 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 81.194 Acc@5 95.552 [2021-04-16 21:46:54 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 81.2% [2021-04-16 21:46:54 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 81.24% [2021-04-16 21:47:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][0/1251] eta 3:25:12 lr 0.000010 time 9.8420 (9.8420) loss 3.1038 (3.1038) grad_norm 3.9779 (3.9779) [2021-04-16 21:47:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][10/1251] eta 0:23:45 lr 0.000010 time 0.3075 (1.1489) loss 2.4383 (2.7531) grad_norm 3.0898 (3.3596) [2021-04-16 21:47:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][20/1251] eta 0:15:05 lr 0.000010 time 0.2928 (0.7357) loss 3.5011 (2.8555) grad_norm 3.5442 (3.4702) [2021-04-16 21:47:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][30/1251] eta 0:12:01 lr 0.000010 time 0.2930 (0.5908) loss 2.0631 (2.8034) grad_norm 4.0301 (3.4435) [2021-04-16 21:47:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][40/1251] eta 0:10:21 lr 0.000010 time 0.2746 (0.5133) loss 2.9845 (2.7867) grad_norm 3.1002 (3.3975) [2021-04-16 21:47:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][50/1251] eta 0:09:20 lr 0.000010 time 0.2881 (0.4668) loss 3.1782 (2.8687) grad_norm 3.5477 (3.4459) [2021-04-16 21:47:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][60/1251] eta 0:08:38 lr 0.000010 time 0.2812 (0.4356) loss 3.4135 (2.9120) grad_norm 3.1951 (3.4476) [2021-04-16 21:47:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][70/1251] eta 0:08:08 lr 0.000010 time 0.2692 (0.4133) loss 2.9271 (2.9436) grad_norm 2.9309 (3.4661) [2021-04-16 21:47:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][80/1251] eta 0:07:45 lr 0.000010 time 0.3075 (0.3973) loss 3.7988 (2.9324) grad_norm 3.3827 (3.4780) [2021-04-16 21:47:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][90/1251] eta 0:07:25 lr 0.000010 time 0.2644 (0.3841) loss 3.1819 (2.9479) grad_norm 3.7473 (3.4818) [2021-04-16 21:47:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][100/1251] eta 0:07:09 lr 0.000010 time 0.2705 (0.3733) loss 3.1063 (2.9737) grad_norm 3.4546 (3.4897) [2021-04-16 21:47:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][110/1251] eta 0:06:55 lr 0.000010 time 0.2722 (0.3642) loss 3.5121 (2.9809) grad_norm 3.3541 (3.4735) [2021-04-16 21:47:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][120/1251] eta 0:06:44 lr 0.000010 time 0.2786 (0.3575) loss 3.1810 (2.9824) grad_norm 4.1423 (3.4769) [2021-04-16 21:47:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][130/1251] eta 0:06:34 lr 0.000010 time 0.2791 (0.3515) loss 3.4062 (2.9863) grad_norm 3.7077 (3.4840) [2021-04-16 21:47:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][140/1251] eta 0:06:27 lr 0.000010 time 0.2951 (0.3484) loss 1.9809 (2.9649) grad_norm 3.4093 (3.4757) [2021-04-16 21:47:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][150/1251] eta 0:06:17 lr 0.000010 time 0.2567 (0.3433) loss 3.2367 (2.9814) grad_norm 2.9550 (3.4760) [2021-04-16 21:47:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][160/1251] eta 0:06:10 lr 0.000010 time 0.2774 (0.3392) loss 2.5446 (2.9716) grad_norm 2.8268 (3.4734) [2021-04-16 21:47:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][170/1251] eta 0:06:02 lr 0.000010 time 0.2700 (0.3353) loss 3.0454 (2.9753) grad_norm 3.8007 (3.4763) [2021-04-16 21:47:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][180/1251] eta 0:05:56 lr 0.000010 time 0.2926 (0.3332) loss 3.4590 (2.9742) grad_norm 3.1515 (3.4754) [2021-04-16 21:47:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][190/1251] eta 0:05:50 lr 0.000010 time 0.2524 (0.3301) loss 3.2919 (2.9657) grad_norm 4.5537 (3.4919) [2021-04-16 21:48:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][200/1251] eta 0:05:45 lr 0.000010 time 0.2575 (0.3283) loss 3.0986 (2.9749) grad_norm 3.7914 (3.5118) [2021-04-16 21:48:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][210/1251] eta 0:05:39 lr 0.000010 time 0.2632 (0.3258) loss 1.8821 (2.9650) grad_norm 2.8666 (3.5118) [2021-04-16 21:48:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][220/1251] eta 0:05:33 lr 0.000010 time 0.2616 (0.3236) loss 3.1487 (2.9692) grad_norm 3.5327 (3.5085) [2021-04-16 21:48:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][230/1251] eta 0:05:28 lr 0.000010 time 0.2810 (0.3217) loss 2.9824 (2.9695) grad_norm 3.1642 (3.5053) [2021-04-16 21:48:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][240/1251] eta 0:05:23 lr 0.000010 time 0.2760 (0.3200) loss 2.8731 (2.9650) grad_norm 3.4796 (3.5080) [2021-04-16 21:48:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][250/1251] eta 0:05:19 lr 0.000010 time 0.2761 (0.3189) loss 3.3787 (2.9664) grad_norm 3.5963 (3.5042) [2021-04-16 21:48:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][260/1251] eta 0:05:14 lr 0.000010 time 0.2985 (0.3176) loss 3.2595 (2.9635) grad_norm 3.3460 (3.4976) [2021-04-16 21:48:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][270/1251] eta 0:05:09 lr 0.000010 time 0.2671 (0.3160) loss 3.2454 (2.9571) grad_norm 3.0595 (3.5051) [2021-04-16 21:48:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][280/1251] eta 0:05:05 lr 0.000010 time 0.3028 (0.3146) loss 3.4788 (2.9623) grad_norm 4.3278 (3.5164) [2021-04-16 21:48:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][290/1251] eta 0:05:01 lr 0.000010 time 0.2694 (0.3134) loss 3.3625 (2.9684) grad_norm 3.7555 (3.5144) [2021-04-16 21:48:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][300/1251] eta 0:04:56 lr 0.000010 time 0.2932 (0.3121) loss 2.9696 (2.9682) grad_norm 3.5028 (3.5139) [2021-04-16 21:48:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][310/1251] eta 0:04:52 lr 0.000010 time 0.2792 (0.3110) loss 3.0119 (2.9655) grad_norm 3.2211 (3.5174) [2021-04-16 21:48:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][320/1251] eta 0:04:48 lr 0.000010 time 0.2623 (0.3099) loss 2.1620 (2.9608) grad_norm 3.5190 (3.5232) [2021-04-16 21:48:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][330/1251] eta 0:04:44 lr 0.000010 time 0.2478 (0.3088) loss 2.9222 (2.9605) grad_norm 3.8937 (3.5240) [2021-04-16 21:48:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][340/1251] eta 0:04:40 lr 0.000010 time 0.2831 (0.3079) loss 3.4320 (2.9646) grad_norm 4.0992 (3.5249) [2021-04-16 21:48:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][350/1251] eta 0:04:36 lr 0.000010 time 0.2948 (0.3071) loss 2.9344 (2.9639) grad_norm 4.0054 (3.5219) [2021-04-16 21:48:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][360/1251] eta 0:04:32 lr 0.000010 time 0.2713 (0.3064) loss 3.2430 (2.9626) grad_norm 3.0922 (3.5232) [2021-04-16 21:48:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][370/1251] eta 0:04:29 lr 0.000010 time 0.2878 (0.3060) loss 2.7954 (2.9640) grad_norm 3.1144 (3.5183) [2021-04-16 21:48:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][380/1251] eta 0:04:25 lr 0.000010 time 0.2551 (0.3052) loss 3.6086 (2.9611) grad_norm 3.1290 (3.5143) [2021-04-16 21:48:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][390/1251] eta 0:04:22 lr 0.000010 time 0.2840 (0.3044) loss 2.4254 (2.9613) grad_norm 2.9667 (3.5123) [2021-04-16 21:48:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][400/1251] eta 0:04:18 lr 0.000010 time 0.2618 (0.3037) loss 3.4371 (2.9554) grad_norm 3.3707 (3.5184) [2021-04-16 21:48:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][410/1251] eta 0:04:14 lr 0.000010 time 0.2807 (0.3031) loss 2.4606 (2.9496) grad_norm 4.5057 (3.5179) [2021-04-16 21:49:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][420/1251] eta 0:04:11 lr 0.000010 time 0.2435 (0.3028) loss 2.7300 (2.9525) grad_norm 2.8517 (3.5189) [2021-04-16 21:49:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][430/1251] eta 0:04:08 lr 0.000010 time 0.2918 (0.3023) loss 3.1979 (2.9533) grad_norm 3.5256 (3.5168) [2021-04-16 21:49:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][440/1251] eta 0:04:04 lr 0.000010 time 0.2842 (0.3017) loss 3.6078 (2.9559) grad_norm 2.8739 (3.5189) [2021-04-16 21:49:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][450/1251] eta 0:04:01 lr 0.000010 time 0.2742 (0.3013) loss 1.9657 (2.9521) grad_norm 2.8571 (3.5175) [2021-04-16 21:49:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][460/1251] eta 0:03:57 lr 0.000010 time 0.2809 (0.3008) loss 3.6232 (2.9494) grad_norm 3.7729 (3.5148) [2021-04-16 21:49:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][470/1251] eta 0:03:54 lr 0.000010 time 0.2766 (0.3003) loss 3.3564 (2.9548) grad_norm 3.3722 (3.5155) [2021-04-16 21:49:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][480/1251] eta 0:03:51 lr 0.000010 time 0.2854 (0.2998) loss 3.2678 (2.9528) grad_norm 3.3029 (3.5118) [2021-04-16 21:49:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][490/1251] eta 0:03:47 lr 0.000010 time 0.3004 (0.2995) loss 3.6359 (2.9537) grad_norm 4.2235 (3.5108) [2021-04-16 21:49:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][500/1251] eta 0:03:44 lr 0.000010 time 0.2821 (0.2990) loss 3.1329 (2.9522) grad_norm 4.3119 (3.5138) [2021-04-16 21:49:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][510/1251] eta 0:03:41 lr 0.000010 time 0.3101 (0.2989) loss 3.0763 (2.9540) grad_norm 3.3493 (3.5195) [2021-04-16 21:49:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][520/1251] eta 0:03:38 lr 0.000010 time 0.2811 (0.2984) loss 3.4359 (2.9575) grad_norm 3.9605 (3.5190) [2021-04-16 21:49:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][530/1251] eta 0:03:34 lr 0.000010 time 0.3120 (0.2981) loss 2.6913 (2.9555) grad_norm 3.5539 (3.5144) [2021-04-16 21:49:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][540/1251] eta 0:03:31 lr 0.000010 time 0.2774 (0.2978) loss 2.1400 (2.9507) grad_norm 3.6320 (3.5110) [2021-04-16 21:49:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][550/1251] eta 0:03:28 lr 0.000010 time 0.2853 (0.2974) loss 3.4890 (2.9511) grad_norm 3.0492 (3.5099) [2021-04-16 21:49:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][560/1251] eta 0:03:25 lr 0.000010 time 0.2656 (0.2970) loss 3.1450 (2.9485) grad_norm 4.4825 (3.5112) [2021-04-16 21:49:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][570/1251] eta 0:03:22 lr 0.000010 time 0.2746 (0.2973) loss 2.8049 (2.9484) grad_norm 3.1695 (3.5073) [2021-04-16 21:49:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][580/1251] eta 0:03:19 lr 0.000010 time 0.2653 (0.2972) loss 3.0675 (2.9461) grad_norm 3.6746 (3.5059) [2021-04-16 21:49:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][590/1251] eta 0:03:16 lr 0.000010 time 0.2797 (0.2968) loss 3.4150 (2.9458) grad_norm 3.0397 (3.5132) [2021-04-16 21:49:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][600/1251] eta 0:03:12 lr 0.000010 time 0.2624 (0.2965) loss 3.3072 (2.9489) grad_norm 3.0418 (3.5127) [2021-04-16 21:49:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][610/1251] eta 0:03:09 lr 0.000010 time 0.2779 (0.2961) loss 2.5424 (2.9489) grad_norm 3.4566 (3.5123) [2021-04-16 21:49:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][620/1251] eta 0:03:06 lr 0.000010 time 0.2618 (0.2958) loss 1.8144 (2.9456) grad_norm 3.6631 (3.5099) [2021-04-16 21:50:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][630/1251] eta 0:03:03 lr 0.000010 time 0.2892 (0.2956) loss 2.5993 (2.9465) grad_norm 3.5878 (3.5132) [2021-04-16 21:50:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][640/1251] eta 0:03:00 lr 0.000010 time 0.2678 (0.2952) loss 2.4105 (2.9473) grad_norm 3.6868 (3.5121) [2021-04-16 21:50:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][650/1251] eta 0:02:57 lr 0.000010 time 0.2755 (0.2949) loss 2.1974 (2.9488) grad_norm 3.1614 (3.5092) [2021-04-16 21:50:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][660/1251] eta 0:02:54 lr 0.000010 time 0.2895 (0.2946) loss 3.1501 (2.9495) grad_norm 3.4751 (3.5108) [2021-04-16 21:50:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][670/1251] eta 0:02:51 lr 0.000010 time 0.3133 (0.2944) loss 3.1881 (2.9478) grad_norm 3.6365 (3.5124) [2021-04-16 21:50:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][680/1251] eta 0:02:48 lr 0.000010 time 0.2720 (0.2942) loss 2.7656 (2.9501) grad_norm 4.3045 (3.5126) [2021-04-16 21:50:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][690/1251] eta 0:02:44 lr 0.000010 time 0.2781 (0.2940) loss 2.6691 (2.9452) grad_norm 3.3059 (3.5097) [2021-04-16 21:50:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][700/1251] eta 0:02:41 lr 0.000010 time 0.2760 (0.2937) loss 3.1681 (2.9470) grad_norm 3.3652 (3.5101) [2021-04-16 21:50:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][710/1251] eta 0:02:38 lr 0.000010 time 0.2978 (0.2936) loss 2.3165 (2.9480) grad_norm 2.8568 (3.5088) [2021-04-16 21:50:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][720/1251] eta 0:02:35 lr 0.000010 time 0.2766 (0.2937) loss 3.2892 (2.9498) grad_norm 3.4002 (3.5074) [2021-04-16 21:50:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][730/1251] eta 0:02:32 lr 0.000010 time 0.2870 (0.2936) loss 3.2475 (2.9473) grad_norm 3.0195 (3.5091) [2021-04-16 21:50:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][740/1251] eta 0:02:29 lr 0.000010 time 0.2855 (0.2933) loss 2.4435 (2.9482) grad_norm 3.1885 (3.5101) [2021-04-16 21:50:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][750/1251] eta 0:02:26 lr 0.000010 time 0.2905 (0.2932) loss 2.5889 (2.9508) grad_norm 2.9512 (3.5113) [2021-04-16 21:50:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][760/1251] eta 0:02:23 lr 0.000010 time 0.2698 (0.2933) loss 3.2694 (2.9484) grad_norm 3.5844 (3.5111) [2021-04-16 21:50:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][770/1251] eta 0:02:20 lr 0.000010 time 0.2961 (0.2931) loss 3.1386 (2.9523) grad_norm 3.2315 (3.5117) [2021-04-16 21:50:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][780/1251] eta 0:02:18 lr 0.000010 time 0.2763 (0.2930) loss 3.1606 (2.9549) grad_norm 2.6904 (3.5087) [2021-04-16 21:50:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][790/1251] eta 0:02:15 lr 0.000010 time 0.2603 (0.2929) loss 3.1694 (2.9584) grad_norm 3.3769 (3.5105) [2021-04-16 21:50:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][800/1251] eta 0:02:12 lr 0.000010 time 0.3009 (0.2927) loss 2.7081 (2.9577) grad_norm 3.4733 (3.5075) [2021-04-16 21:50:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][810/1251] eta 0:02:08 lr 0.000010 time 0.2683 (0.2925) loss 2.3592 (2.9590) grad_norm 3.1178 (3.5071) [2021-04-16 21:50:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][820/1251] eta 0:02:06 lr 0.000010 time 0.2839 (0.2924) loss 2.9677 (2.9595) grad_norm 3.0411 (3.5063) [2021-04-16 21:50:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][830/1251] eta 0:02:03 lr 0.000010 time 0.2696 (0.2924) loss 3.7824 (2.9601) grad_norm 3.4668 (3.5057) [2021-04-16 21:50:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][840/1251] eta 0:02:00 lr 0.000010 time 0.2917 (0.2922) loss 2.2026 (2.9587) grad_norm 3.7056 (3.5052) [2021-04-16 21:51:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][850/1251] eta 0:01:57 lr 0.000010 time 0.2831 (0.2920) loss 3.1520 (2.9582) grad_norm 4.5189 (3.5121) [2021-04-16 21:51:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][860/1251] eta 0:01:54 lr 0.000010 time 0.2998 (0.2919) loss 3.6028 (2.9610) grad_norm 3.4956 (3.5153) [2021-04-16 21:51:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][870/1251] eta 0:01:51 lr 0.000010 time 0.2509 (0.2917) loss 3.1442 (2.9619) grad_norm 4.6653 (3.5172) [2021-04-16 21:51:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][880/1251] eta 0:01:48 lr 0.000010 time 0.3096 (0.2916) loss 1.6126 (2.9578) grad_norm 2.9700 (3.5199) [2021-04-16 21:51:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][890/1251] eta 0:01:45 lr 0.000010 time 0.2979 (0.2914) loss 1.8872 (2.9589) grad_norm 3.1239 (3.5174) [2021-04-16 21:51:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][900/1251] eta 0:01:42 lr 0.000010 time 0.2865 (0.2914) loss 1.7917 (2.9589) grad_norm 3.9151 (3.5152) [2021-04-16 21:51:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][910/1251] eta 0:01:39 lr 0.000010 time 0.2713 (0.2912) loss 3.1297 (2.9595) grad_norm 3.4088 (3.5156) [2021-04-16 21:51:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][920/1251] eta 0:01:36 lr 0.000010 time 0.2971 (0.2911) loss 2.6851 (2.9599) grad_norm 4.0990 (3.5147) [2021-04-16 21:51:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][930/1251] eta 0:01:33 lr 0.000010 time 0.2506 (0.2911) loss 2.6884 (2.9595) grad_norm 3.2314 (3.5144) [2021-04-16 21:51:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][940/1251] eta 0:01:30 lr 0.000010 time 0.2959 (0.2911) loss 2.6296 (2.9598) grad_norm 4.0951 (3.5177) [2021-04-16 21:51:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][950/1251] eta 0:01:27 lr 0.000010 time 0.2782 (0.2910) loss 3.4988 (2.9609) grad_norm 2.9968 (3.5145) [2021-04-16 21:51:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][960/1251] eta 0:01:24 lr 0.000010 time 0.2928 (0.2910) loss 2.7145 (2.9599) grad_norm 3.1665 (3.5139) [2021-04-16 21:51:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][970/1251] eta 0:01:21 lr 0.000010 time 0.2548 (0.2908) loss 3.4916 (2.9605) grad_norm 3.3757 (3.5135) [2021-04-16 21:51:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][980/1251] eta 0:01:18 lr 0.000010 time 0.2748 (0.2906) loss 2.9485 (2.9626) grad_norm 3.2100 (3.5203) [2021-04-16 21:51:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][990/1251] eta 0:01:15 lr 0.000010 time 0.2599 (0.2906) loss 3.3391 (2.9633) grad_norm 2.9944 (3.5171) [2021-04-16 21:51:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][1000/1251] eta 0:01:12 lr 0.000010 time 0.2796 (0.2905) loss 3.5545 (2.9642) grad_norm 3.6335 (3.5177) [2021-04-16 21:51:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][1010/1251] eta 0:01:09 lr 0.000010 time 0.2968 (0.2904) loss 1.8585 (2.9633) grad_norm 3.6152 (3.5267) [2021-04-16 21:51:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][1020/1251] eta 0:01:07 lr 0.000010 time 0.2733 (0.2903) loss 3.2082 (2.9649) grad_norm 2.9039 (3.5268) [2021-04-16 21:51:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][1030/1251] eta 0:01:04 lr 0.000010 time 0.2852 (0.2902) loss 3.1886 (2.9640) grad_norm 2.9783 (3.5250) [2021-04-16 21:51:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][1040/1251] eta 0:01:01 lr 0.000010 time 0.2962 (0.2901) loss 3.0628 (2.9628) grad_norm 4.3802 (3.5293) [2021-04-16 21:51:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][1050/1251] eta 0:00:58 lr 0.000010 time 0.2903 (0.2900) loss 3.2596 (2.9625) grad_norm 3.8593 (3.5324) [2021-04-16 21:52:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][1060/1251] eta 0:00:55 lr 0.000010 time 0.2792 (0.2898) loss 2.8486 (2.9625) grad_norm 3.4419 (3.5325) [2021-04-16 21:52:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][1070/1251] eta 0:00:52 lr 0.000010 time 0.2534 (0.2897) loss 3.2602 (2.9638) grad_norm 2.9548 (3.5305) [2021-04-16 21:52:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][1080/1251] eta 0:00:49 lr 0.000010 time 0.2774 (0.2896) loss 3.2297 (2.9637) grad_norm 4.3363 (3.5308) [2021-04-16 21:52:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][1090/1251] eta 0:00:46 lr 0.000010 time 0.2917 (0.2895) loss 3.2882 (2.9637) grad_norm 2.5642 (3.5318) [2021-04-16 21:52:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][1100/1251] eta 0:00:43 lr 0.000010 time 0.2693 (0.2893) loss 2.5360 (2.9630) grad_norm 3.6733 (3.5305) [2021-04-16 21:52:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][1110/1251] eta 0:00:40 lr 0.000010 time 0.2661 (0.2892) loss 3.1363 (2.9647) grad_norm 3.6621 (3.5295) [2021-04-16 21:52:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][1120/1251] eta 0:00:37 lr 0.000010 time 0.3127 (0.2891) loss 2.6335 (2.9649) grad_norm 3.4614 (3.5282) [2021-04-16 21:52:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][1130/1251] eta 0:00:34 lr 0.000010 time 0.2786 (0.2890) loss 3.2608 (2.9655) grad_norm 3.4550 (3.5278) [2021-04-16 21:52:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][1140/1251] eta 0:00:32 lr 0.000010 time 0.2517 (0.2890) loss 2.1893 (2.9630) grad_norm 3.9126 (3.5288) [2021-04-16 21:52:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][1150/1251] eta 0:00:29 lr 0.000010 time 0.2694 (0.2892) loss 3.3297 (2.9635) grad_norm 3.6441 (3.5271) [2021-04-16 21:52:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][1160/1251] eta 0:00:26 lr 0.000010 time 0.2613 (0.2890) loss 2.2685 (2.9612) grad_norm 3.0441 (3.5261) [2021-04-16 21:52:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][1170/1251] eta 0:00:23 lr 0.000010 time 0.2566 (0.2890) loss 2.9493 (2.9643) grad_norm 3.2865 (3.5250) [2021-04-16 21:52:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][1180/1251] eta 0:00:20 lr 0.000010 time 0.2781 (0.2889) loss 2.7998 (2.9640) grad_norm 3.0322 (3.5236) [2021-04-16 21:52:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][1190/1251] eta 0:00:17 lr 0.000010 time 0.2517 (0.2888) loss 2.7414 (2.9642) grad_norm 3.5029 (3.5233) [2021-04-16 21:52:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][1200/1251] eta 0:00:14 lr 0.000010 time 0.2426 (0.2887) loss 2.1091 (2.9627) grad_norm 3.1013 (3.5236) [2021-04-16 21:52:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][1210/1251] eta 0:00:11 lr 0.000010 time 0.2909 (0.2887) loss 2.8544 (2.9629) grad_norm 3.9403 (3.5261) [2021-04-16 21:52:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][1220/1251] eta 0:00:08 lr 0.000010 time 0.2561 (0.2886) loss 2.3405 (2.9607) grad_norm 3.6331 (3.5248) [2021-04-16 21:52:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][1230/1251] eta 0:00:06 lr 0.000010 time 0.2996 (0.2885) loss 2.4097 (2.9606) grad_norm 3.0035 (3.5230) [2021-04-16 21:52:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][1240/1251] eta 0:00:03 lr 0.000010 time 0.2479 (0.2883) loss 3.2741 (2.9623) grad_norm 3.5698 (3.5238) [2021-04-16 21:52:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [296/300][1250/1251] eta 0:00:00 lr 0.000010 time 0.2482 (0.2880) loss 3.6112 (2.9646) grad_norm 4.2143 (3.5228) [2021-04-16 21:52:58 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 296 training takes 0:06:04 [2021-04-16 21:52:58 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_296.pth saving...... [2021-04-16 21:53:05 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_296.pth saved !!! [2021-04-16 21:53:07 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.239 (1.239) Loss 0.8167 (0.8167) Acc@1 80.664 (80.664) Acc@5 96.680 (96.680) [2021-04-16 21:53:08 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.135 (0.202) Loss 0.8656 (0.8047) Acc@1 80.859 (81.365) Acc@5 95.117 (95.890) [2021-04-16 21:53:10 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.124 (0.209) Loss 0.8849 (0.8052) Acc@1 79.883 (81.487) Acc@5 94.727 (95.708) [2021-04-16 21:53:12 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.549 (0.225) Loss 0.7632 (0.8101) Acc@1 82.812 (81.433) Acc@5 95.996 (95.628) [2021-04-16 21:53:14 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.154 (0.213) Loss 0.8760 (0.8176) Acc@1 79.199 (81.195) Acc@5 94.727 (95.491) [2021-04-16 21:53:25 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 81.248 Acc@5 95.500 [2021-04-16 21:53:25 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 81.2% [2021-04-16 21:53:25 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 81.25% [2021-04-16 21:53:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][0/1251] eta 2:14:52 lr 0.000010 time 6.4691 (6.4691) loss 3.4582 (3.4582) grad_norm 2.8192 (2.8192) [2021-04-16 21:53:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][10/1251] eta 0:17:18 lr 0.000010 time 0.2744 (0.8371) loss 3.1550 (3.1030) grad_norm 3.0899 (3.5001) [2021-04-16 21:53:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][20/1251] eta 0:11:46 lr 0.000010 time 0.2680 (0.5738) loss 3.4223 (2.9102) grad_norm 3.1499 (3.4300) [2021-04-16 21:53:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][30/1251] eta 0:09:48 lr 0.000010 time 0.2818 (0.4822) loss 3.0396 (3.0071) grad_norm 3.2033 (3.3943) [2021-04-16 21:53:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][40/1251] eta 0:08:45 lr 0.000010 time 0.2911 (0.4336) loss 2.0532 (2.9379) grad_norm 2.9243 (3.3732) [2021-04-16 21:53:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][50/1251] eta 0:08:04 lr 0.000010 time 0.3074 (0.4037) loss 3.1021 (2.9074) grad_norm 3.6901 (3.3536) [2021-04-16 21:53:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][60/1251] eta 0:07:36 lr 0.000010 time 0.2861 (0.3837) loss 2.0629 (2.9283) grad_norm 3.6914 (3.3682) [2021-04-16 21:53:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][70/1251] eta 0:07:16 lr 0.000010 time 0.2779 (0.3698) loss 2.5068 (2.9325) grad_norm 3.8019 (3.4572) [2021-04-16 21:53:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][80/1251] eta 0:07:00 lr 0.000010 time 0.2802 (0.3588) loss 3.2655 (2.9361) grad_norm 3.2832 (3.4954) [2021-04-16 21:53:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][90/1251] eta 0:06:46 lr 0.000010 time 0.2781 (0.3499) loss 3.1192 (2.9354) grad_norm 2.9106 (3.4987) [2021-04-16 21:54:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][100/1251] eta 0:06:34 lr 0.000010 time 0.2622 (0.3428) loss 2.7137 (2.9303) grad_norm 3.7962 (3.5066) [2021-04-16 21:54:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][110/1251] eta 0:06:24 lr 0.000010 time 0.2765 (0.3369) loss 3.3375 (2.9523) grad_norm 2.8768 (3.5093) [2021-04-16 21:54:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][120/1251] eta 0:06:15 lr 0.000010 time 0.2667 (0.3319) loss 3.0621 (2.9594) grad_norm 4.3189 (3.5196) [2021-04-16 21:54:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][130/1251] eta 0:06:07 lr 0.000010 time 0.2610 (0.3282) loss 3.1303 (2.9464) grad_norm 3.6584 (3.5128) [2021-04-16 21:54:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][140/1251] eta 0:06:02 lr 0.000010 time 0.2693 (0.3260) loss 2.5067 (2.9515) grad_norm 3.4705 (3.5079) [2021-04-16 21:54:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][150/1251] eta 0:05:55 lr 0.000010 time 0.2721 (0.3229) loss 2.9156 (2.9463) grad_norm 3.0505 (3.5262) [2021-04-16 21:54:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][160/1251] eta 0:05:49 lr 0.000010 time 0.2676 (0.3206) loss 3.3437 (2.9425) grad_norm 2.9990 (3.5137) [2021-04-16 21:54:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][170/1251] eta 0:05:43 lr 0.000010 time 0.2975 (0.3181) loss 2.6222 (2.9298) grad_norm 2.9997 (3.5166) [2021-04-16 21:54:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][180/1251] eta 0:05:38 lr 0.000010 time 0.2803 (0.3160) loss 2.9526 (2.9381) grad_norm 3.0924 (3.5044) [2021-04-16 21:54:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][190/1251] eta 0:05:33 lr 0.000010 time 0.2883 (0.3142) loss 2.2173 (2.9364) grad_norm 4.1110 (3.5037) [2021-04-16 21:54:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][200/1251] eta 0:05:28 lr 0.000010 time 0.2726 (0.3130) loss 2.6141 (2.9387) grad_norm 3.2049 (3.5198) [2021-04-16 21:54:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][210/1251] eta 0:05:24 lr 0.000010 time 0.2826 (0.3113) loss 1.9910 (2.9431) grad_norm 3.1981 (3.5090) [2021-04-16 21:54:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][220/1251] eta 0:05:19 lr 0.000010 time 0.2960 (0.3099) loss 3.0802 (2.9571) grad_norm 3.3628 (3.5025) [2021-04-16 21:54:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][230/1251] eta 0:05:14 lr 0.000010 time 0.2936 (0.3085) loss 2.7243 (2.9497) grad_norm 3.2883 (3.5041) [2021-04-16 21:54:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][240/1251] eta 0:05:10 lr 0.000010 time 0.2978 (0.3074) loss 2.0487 (2.9396) grad_norm 3.6592 (3.4946) [2021-04-16 21:54:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][250/1251] eta 0:05:06 lr 0.000010 time 0.2773 (0.3062) loss 3.6916 (2.9407) grad_norm 3.0210 (3.4914) [2021-04-16 21:54:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][260/1251] eta 0:05:02 lr 0.000010 time 0.2925 (0.3051) loss 2.4645 (2.9432) grad_norm 3.7009 (3.4821) [2021-04-16 21:54:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][270/1251] eta 0:04:58 lr 0.000010 time 0.2511 (0.3041) loss 2.0545 (2.9403) grad_norm 3.9749 (3.4806) [2021-04-16 21:54:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][280/1251] eta 0:04:54 lr 0.000010 time 0.2988 (0.3032) loss 2.1578 (2.9315) grad_norm 4.5316 (3.4932) [2021-04-16 21:54:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][290/1251] eta 0:04:50 lr 0.000010 time 0.2875 (0.3023) loss 2.0610 (2.9296) grad_norm 2.3862 (3.4906) [2021-04-16 21:54:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][300/1251] eta 0:04:46 lr 0.000010 time 0.2698 (0.3016) loss 3.1754 (2.9378) grad_norm 3.4020 (3.4943) [2021-04-16 21:54:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][310/1251] eta 0:04:43 lr 0.000010 time 0.2605 (0.3011) loss 2.9619 (2.9374) grad_norm 2.8468 (3.4878) [2021-04-16 21:55:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][320/1251] eta 0:04:39 lr 0.000010 time 0.2714 (0.3005) loss 3.0877 (2.9400) grad_norm 3.0658 (3.4818) [2021-04-16 21:55:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][330/1251] eta 0:04:36 lr 0.000010 time 0.2637 (0.3002) loss 3.4683 (2.9508) grad_norm 3.2658 (3.4758) [2021-04-16 21:55:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][340/1251] eta 0:04:33 lr 0.000010 time 0.2647 (0.2999) loss 3.2223 (2.9474) grad_norm 3.5024 (3.4804) [2021-04-16 21:55:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][350/1251] eta 0:04:29 lr 0.000010 time 0.2969 (0.2994) loss 2.4919 (2.9510) grad_norm 3.6216 (3.4991) [2021-04-16 21:55:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][360/1251] eta 0:04:27 lr 0.000010 time 0.2804 (0.2997) loss 3.5640 (2.9551) grad_norm 3.3569 (3.5068) [2021-04-16 21:55:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][370/1251] eta 0:04:23 lr 0.000010 time 0.2895 (0.2994) loss 2.8615 (2.9502) grad_norm 3.7600 (3.5063) [2021-04-16 21:55:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][380/1251] eta 0:04:20 lr 0.000010 time 0.2789 (0.2988) loss 3.2100 (2.9526) grad_norm 2.9412 (inf) [2021-04-16 21:55:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][390/1251] eta 0:04:16 lr 0.000010 time 0.2837 (0.2982) loss 3.4180 (2.9501) grad_norm 2.7859 (inf) [2021-04-16 21:55:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][400/1251] eta 0:04:13 lr 0.000010 time 0.2716 (0.2977) loss 3.5358 (2.9511) grad_norm 3.0931 (inf) [2021-04-16 21:55:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][410/1251] eta 0:04:10 lr 0.000010 time 0.2589 (0.2973) loss 2.5955 (2.9542) grad_norm 3.1038 (inf) [2021-04-16 21:55:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][420/1251] eta 0:04:06 lr 0.000010 time 0.2752 (0.2969) loss 3.4174 (2.9602) grad_norm 3.9015 (inf) [2021-04-16 21:55:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][430/1251] eta 0:04:03 lr 0.000010 time 0.2614 (0.2965) loss 2.3882 (2.9598) grad_norm 3.1922 (inf) [2021-04-16 21:55:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][440/1251] eta 0:04:00 lr 0.000010 time 0.2594 (0.2961) loss 2.1347 (2.9527) grad_norm 3.2018 (inf) [2021-04-16 21:55:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][450/1251] eta 0:03:56 lr 0.000010 time 0.2771 (0.2958) loss 3.2463 (2.9533) grad_norm 3.4516 (inf) [2021-04-16 21:55:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][460/1251] eta 0:03:53 lr 0.000010 time 0.2858 (0.2954) loss 3.2360 (2.9567) grad_norm 4.2568 (inf) [2021-04-16 21:55:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][470/1251] eta 0:03:50 lr 0.000010 time 0.2891 (0.2950) loss 2.7339 (2.9509) grad_norm 3.5079 (inf) [2021-04-16 21:55:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][480/1251] eta 0:03:47 lr 0.000010 time 0.2853 (0.2946) loss 3.6709 (2.9515) grad_norm 3.0757 (inf) [2021-04-16 21:55:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][490/1251] eta 0:03:43 lr 0.000010 time 0.2932 (0.2943) loss 3.2915 (2.9569) grad_norm 3.1992 (inf) [2021-04-16 21:55:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][500/1251] eta 0:03:40 lr 0.000010 time 0.2734 (0.2939) loss 2.7145 (2.9603) grad_norm 3.4238 (inf) [2021-04-16 21:55:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][510/1251] eta 0:03:37 lr 0.000010 time 0.2542 (0.2935) loss 2.0496 (2.9589) grad_norm 3.4562 (inf) [2021-04-16 21:55:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][520/1251] eta 0:03:34 lr 0.000010 time 0.2721 (0.2932) loss 3.0195 (2.9564) grad_norm 3.8433 (inf) [2021-04-16 21:56:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][530/1251] eta 0:03:31 lr 0.000010 time 0.2944 (0.2931) loss 2.7010 (2.9580) grad_norm 3.3795 (inf) [2021-04-16 21:56:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][540/1251] eta 0:03:28 lr 0.000010 time 0.2909 (0.2929) loss 3.1395 (2.9613) grad_norm 3.7500 (inf) [2021-04-16 21:56:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][550/1251] eta 0:03:25 lr 0.000010 time 0.2502 (0.2929) loss 1.9753 (2.9572) grad_norm 3.3286 (inf) [2021-04-16 21:56:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][560/1251] eta 0:03:22 lr 0.000010 time 0.2718 (0.2927) loss 3.3551 (2.9584) grad_norm 3.2642 (inf) [2021-04-16 21:56:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][570/1251] eta 0:03:19 lr 0.000010 time 0.2791 (0.2927) loss 3.2741 (2.9613) grad_norm 3.1366 (inf) [2021-04-16 21:56:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][580/1251] eta 0:03:16 lr 0.000010 time 0.2848 (0.2927) loss 2.2700 (2.9561) grad_norm 3.0872 (inf) [2021-04-16 21:56:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][590/1251] eta 0:03:13 lr 0.000010 time 0.2812 (0.2925) loss 2.9974 (2.9572) grad_norm 3.3756 (inf) [2021-04-16 21:56:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][600/1251] eta 0:03:10 lr 0.000010 time 0.3900 (0.2925) loss 3.4720 (2.9558) grad_norm 3.4090 (inf) [2021-04-16 21:56:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][610/1251] eta 0:03:07 lr 0.000010 time 0.2803 (0.2923) loss 3.2465 (2.9584) grad_norm 4.3011 (inf) [2021-04-16 21:56:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][620/1251] eta 0:03:04 lr 0.000010 time 0.2757 (0.2922) loss 2.3099 (2.9528) grad_norm 3.8709 (inf) [2021-04-16 21:56:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][630/1251] eta 0:03:01 lr 0.000010 time 0.2840 (0.2920) loss 3.5786 (2.9521) grad_norm 3.5238 (inf) [2021-04-16 21:56:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][640/1251] eta 0:02:58 lr 0.000010 time 0.2707 (0.2918) loss 2.2453 (2.9552) grad_norm 3.5479 (inf) [2021-04-16 21:56:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][650/1251] eta 0:02:55 lr 0.000010 time 0.2879 (0.2916) loss 3.2750 (2.9563) grad_norm 2.8958 (inf) [2021-04-16 21:56:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][660/1251] eta 0:02:52 lr 0.000010 time 0.2952 (0.2915) loss 2.7428 (2.9554) grad_norm 3.2826 (inf) [2021-04-16 21:56:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][670/1251] eta 0:02:49 lr 0.000010 time 0.2972 (0.2913) loss 2.9866 (2.9564) grad_norm 3.2538 (inf) [2021-04-16 21:56:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][680/1251] eta 0:02:46 lr 0.000010 time 0.2683 (0.2911) loss 2.6270 (2.9551) grad_norm 3.7458 (inf) [2021-04-16 21:56:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][690/1251] eta 0:02:43 lr 0.000010 time 0.2782 (0.2909) loss 3.4011 (2.9571) grad_norm 3.3255 (inf) [2021-04-16 21:56:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][700/1251] eta 0:02:40 lr 0.000010 time 0.2930 (0.2908) loss 3.4259 (2.9569) grad_norm 3.0219 (inf) [2021-04-16 21:56:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][710/1251] eta 0:02:37 lr 0.000010 time 0.2717 (0.2906) loss 3.2297 (2.9569) grad_norm 3.6568 (inf) [2021-04-16 21:56:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][720/1251] eta 0:02:34 lr 0.000010 time 0.2883 (0.2908) loss 3.2469 (2.9591) grad_norm 2.9042 (inf) [2021-04-16 21:56:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][730/1251] eta 0:02:31 lr 0.000010 time 0.2544 (0.2907) loss 3.2995 (2.9595) grad_norm 3.0700 (inf) [2021-04-16 21:57:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][740/1251] eta 0:02:28 lr 0.000010 time 0.2813 (0.2906) loss 3.4547 (2.9584) grad_norm 3.2942 (inf) [2021-04-16 21:57:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][750/1251] eta 0:02:25 lr 0.000010 time 0.2653 (0.2904) loss 3.5023 (2.9596) grad_norm 3.5347 (inf) [2021-04-16 21:57:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][760/1251] eta 0:02:22 lr 0.000010 time 0.2960 (0.2903) loss 2.8549 (2.9586) grad_norm 2.9209 (inf) [2021-04-16 21:57:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][770/1251] eta 0:02:19 lr 0.000010 time 0.2742 (0.2901) loss 3.0842 (2.9603) grad_norm 3.1381 (inf) [2021-04-16 21:57:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][780/1251] eta 0:02:16 lr 0.000010 time 0.2705 (0.2901) loss 3.0178 (2.9594) grad_norm 3.9796 (inf) [2021-04-16 21:57:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][790/1251] eta 0:02:13 lr 0.000010 time 0.2576 (0.2900) loss 3.0366 (2.9621) grad_norm 3.4643 (inf) [2021-04-16 21:57:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][800/1251] eta 0:02:10 lr 0.000010 time 0.2765 (0.2898) loss 2.9562 (2.9608) grad_norm 3.0788 (inf) [2021-04-16 21:57:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][810/1251] eta 0:02:07 lr 0.000010 time 0.2608 (0.2896) loss 3.2841 (2.9586) grad_norm 2.9347 (inf) [2021-04-16 21:57:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][820/1251] eta 0:02:04 lr 0.000010 time 0.2752 (0.2895) loss 3.5309 (2.9600) grad_norm 3.2136 (inf) [2021-04-16 21:57:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][830/1251] eta 0:02:01 lr 0.000010 time 0.2572 (0.2893) loss 3.2409 (2.9612) grad_norm 2.9223 (inf) [2021-04-16 21:57:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][840/1251] eta 0:01:58 lr 0.000010 time 0.2999 (0.2892) loss 3.4363 (2.9615) grad_norm 3.0496 (inf) [2021-04-16 21:57:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][850/1251] eta 0:01:55 lr 0.000010 time 0.2741 (0.2892) loss 3.1084 (2.9633) grad_norm 4.7290 (inf) [2021-04-16 21:57:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][860/1251] eta 0:01:53 lr 0.000010 time 0.2895 (0.2892) loss 2.1213 (2.9626) grad_norm 3.6216 (inf) [2021-04-16 21:57:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][870/1251] eta 0:01:50 lr 0.000010 time 0.3045 (0.2891) loss 2.0985 (2.9612) grad_norm 3.8720 (inf) [2021-04-16 21:57:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][880/1251] eta 0:01:47 lr 0.000010 time 0.3605 (0.2891) loss 3.0874 (2.9612) grad_norm 3.1985 (inf) [2021-04-16 21:57:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][890/1251] eta 0:01:44 lr 0.000010 time 0.2718 (0.2890) loss 3.2461 (2.9640) grad_norm 3.4147 (inf) [2021-04-16 21:57:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][900/1251] eta 0:01:41 lr 0.000010 time 0.2854 (0.2889) loss 3.0948 (2.9655) grad_norm 3.1610 (inf) [2021-04-16 21:57:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][910/1251] eta 0:01:38 lr 0.000010 time 0.3075 (0.2888) loss 3.3254 (2.9668) grad_norm 3.1837 (inf) [2021-04-16 21:57:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][920/1251] eta 0:01:35 lr 0.000010 time 0.2816 (0.2888) loss 3.2729 (2.9665) grad_norm 3.7618 (inf) [2021-04-16 21:57:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][930/1251] eta 0:01:32 lr 0.000010 time 0.2755 (0.2887) loss 3.0717 (2.9646) grad_norm 3.8270 (inf) [2021-04-16 21:57:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][940/1251] eta 0:01:29 lr 0.000010 time 0.3067 (0.2889) loss 2.7907 (2.9625) grad_norm 3.0636 (inf) [2021-04-16 21:58:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][950/1251] eta 0:01:26 lr 0.000010 time 0.2864 (0.2889) loss 3.0233 (2.9625) grad_norm 3.3822 (inf) [2021-04-16 21:58:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][960/1251] eta 0:01:24 lr 0.000010 time 0.2904 (0.2888) loss 3.1371 (2.9622) grad_norm 4.4711 (inf) [2021-04-16 21:58:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][970/1251] eta 0:01:21 lr 0.000010 time 0.2481 (0.2886) loss 3.1669 (2.9609) grad_norm 3.3902 (inf) [2021-04-16 21:58:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][980/1251] eta 0:01:18 lr 0.000010 time 0.2604 (0.2885) loss 3.0123 (2.9599) grad_norm 3.2677 (inf) [2021-04-16 21:58:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][990/1251] eta 0:01:15 lr 0.000010 time 0.2925 (0.2884) loss 2.7263 (2.9598) grad_norm 3.3555 (inf) [2021-04-16 21:58:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][1000/1251] eta 0:01:12 lr 0.000010 time 0.2590 (0.2883) loss 2.4137 (2.9615) grad_norm 4.1187 (inf) [2021-04-16 21:58:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][1010/1251] eta 0:01:09 lr 0.000010 time 0.2949 (0.2882) loss 3.1313 (2.9606) grad_norm 3.1674 (inf) [2021-04-16 21:58:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][1020/1251] eta 0:01:06 lr 0.000010 time 0.2515 (0.2882) loss 2.5280 (2.9609) grad_norm 3.0168 (inf) [2021-04-16 21:58:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][1030/1251] eta 0:01:03 lr 0.000010 time 0.3159 (0.2881) loss 3.3180 (2.9613) grad_norm 3.3307 (inf) [2021-04-16 21:58:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][1040/1251] eta 0:01:00 lr 0.000010 time 0.2600 (0.2880) loss 3.1652 (2.9620) grad_norm 3.8104 (inf) [2021-04-16 21:58:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][1050/1251] eta 0:00:57 lr 0.000010 time 0.2870 (0.2880) loss 3.1689 (2.9620) grad_norm 4.1780 (inf) [2021-04-16 21:58:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][1060/1251] eta 0:00:54 lr 0.000010 time 0.2914 (0.2879) loss 2.7383 (2.9627) grad_norm 4.3663 (inf) [2021-04-16 21:58:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][1070/1251] eta 0:00:52 lr 0.000010 time 0.2761 (0.2878) loss 3.7626 (2.9625) grad_norm 3.8153 (inf) [2021-04-16 21:58:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][1080/1251] eta 0:00:49 lr 0.000010 time 0.2882 (0.2877) loss 3.4496 (2.9634) grad_norm 3.1783 (inf) [2021-04-16 21:58:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][1090/1251] eta 0:00:46 lr 0.000010 time 0.2630 (0.2876) loss 3.0931 (2.9626) grad_norm 4.2966 (inf) [2021-04-16 21:58:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][1100/1251] eta 0:00:43 lr 0.000010 time 0.2575 (0.2875) loss 2.9965 (2.9621) grad_norm 3.3866 (inf) [2021-04-16 21:58:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][1110/1251] eta 0:00:40 lr 0.000010 time 0.2932 (0.2874) loss 2.6563 (2.9608) grad_norm 4.3872 (inf) [2021-04-16 21:58:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][1120/1251] eta 0:00:37 lr 0.000010 time 0.2849 (0.2874) loss 3.0036 (2.9611) grad_norm 3.1263 (inf) [2021-04-16 21:58:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][1130/1251] eta 0:00:34 lr 0.000010 time 0.2806 (0.2875) loss 3.1411 (2.9619) grad_norm 4.3452 (inf) [2021-04-16 21:58:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][1140/1251] eta 0:00:31 lr 0.000010 time 0.2821 (0.2875) loss 2.1648 (2.9625) grad_norm 4.1779 (inf) [2021-04-16 21:58:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][1150/1251] eta 0:00:29 lr 0.000010 time 0.2584 (0.2875) loss 3.3854 (2.9634) grad_norm 3.8560 (inf) [2021-04-16 21:58:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][1160/1251] eta 0:00:26 lr 0.000010 time 0.2822 (0.2874) loss 2.1397 (2.9616) grad_norm 3.2428 (inf) [2021-04-16 21:59:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][1170/1251] eta 0:00:23 lr 0.000010 time 0.2545 (0.2873) loss 3.6177 (2.9611) grad_norm 3.6629 (inf) [2021-04-16 21:59:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][1180/1251] eta 0:00:20 lr 0.000010 time 0.2805 (0.2874) loss 3.2963 (2.9639) grad_norm 7.4531 (inf) [2021-04-16 21:59:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][1190/1251] eta 0:00:17 lr 0.000010 time 0.2786 (0.2872) loss 2.8330 (2.9633) grad_norm 4.4794 (inf) [2021-04-16 21:59:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][1200/1251] eta 0:00:14 lr 0.000010 time 0.2914 (0.2872) loss 2.9452 (2.9643) grad_norm 3.7547 (inf) [2021-04-16 21:59:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][1210/1251] eta 0:00:11 lr 0.000010 time 0.2763 (0.2871) loss 1.7221 (2.9628) grad_norm 3.3637 (inf) [2021-04-16 21:59:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][1220/1251] eta 0:00:08 lr 0.000010 time 0.2767 (0.2870) loss 2.3544 (2.9633) grad_norm 3.3725 (inf) [2021-04-16 21:59:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][1230/1251] eta 0:00:06 lr 0.000010 time 0.2749 (0.2869) loss 2.8983 (2.9596) grad_norm 3.9617 (inf) [2021-04-16 21:59:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][1240/1251] eta 0:00:03 lr 0.000010 time 0.2490 (0.2867) loss 3.0388 (2.9594) grad_norm 3.1761 (inf) [2021-04-16 21:59:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [297/300][1250/1251] eta 0:00:00 lr 0.000010 time 0.2482 (0.2864) loss 3.1540 (2.9613) grad_norm 3.1077 (inf) [2021-04-16 21:59:35 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 297 training takes 0:06:10 [2021-04-16 21:59:35 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_297.pth saving...... [2021-04-16 21:59:57 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_297.pth saved !!! [2021-04-16 21:59:58 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.164 (1.164) Loss 0.8129 (0.8129) Acc@1 81.836 (81.836) Acc@5 95.605 (95.605) [2021-04-16 21:59:59 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.096 (0.197) Loss 0.8124 (0.8150) Acc@1 81.445 (81.312) Acc@5 95.703 (95.428) [2021-04-16 22:00:02 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.135 (0.233) Loss 0.8204 (0.8153) Acc@1 79.688 (81.143) Acc@5 95.508 (95.522) [2021-04-16 22:00:04 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.125 (0.237) Loss 0.7893 (0.8122) Acc@1 80.859 (81.133) Acc@5 95.312 (95.549) [2021-04-16 22:00:05 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.074 (0.213) Loss 0.8094 (0.8109) Acc@1 80.859 (81.186) Acc@5 95.703 (95.527) [2021-04-16 22:00:22 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 81.172 Acc@5 95.550 [2021-04-16 22:00:22 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 81.2% [2021-04-16 22:00:22 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 81.25% [2021-04-16 22:00:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][0/1251] eta 3:57:26 lr 0.000010 time 11.3884 (11.3884) loss 2.5122 (2.5122) grad_norm 3.2156 (3.2156) [2021-04-16 22:00:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][10/1251] eta 0:26:49 lr 0.000010 time 0.4253 (1.2969) loss 2.7112 (2.8777) grad_norm 2.9717 (3.4616) [2021-04-16 22:00:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][20/1251] eta 0:16:39 lr 0.000010 time 0.2757 (0.8115) loss 3.6250 (2.8586) grad_norm 3.4075 (3.4717) [2021-04-16 22:00:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][30/1251] eta 0:13:00 lr 0.000010 time 0.2838 (0.6396) loss 3.5081 (2.8858) grad_norm 4.0810 (3.5003) [2021-04-16 22:00:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][40/1251] eta 0:11:06 lr 0.000010 time 0.2622 (0.5505) loss 3.1230 (2.8616) grad_norm 3.7734 (3.5273) [2021-04-16 22:00:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][50/1251] eta 0:09:57 lr 0.000010 time 0.2787 (0.4976) loss 1.9129 (2.8698) grad_norm 2.8272 (3.7233) [2021-04-16 22:00:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][60/1251] eta 0:09:08 lr 0.000010 time 0.2728 (0.4609) loss 3.1626 (2.9023) grad_norm 4.2308 (3.6822) [2021-04-16 22:00:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][70/1251] eta 0:08:35 lr 0.000010 time 0.2752 (0.4361) loss 3.4322 (2.9121) grad_norm 3.3334 (3.6851) [2021-04-16 22:00:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][80/1251] eta 0:08:07 lr 0.000010 time 0.3114 (0.4163) loss 3.2059 (2.9069) grad_norm 3.6968 (3.6314) [2021-04-16 22:00:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][90/1251] eta 0:07:45 lr 0.000010 time 0.2841 (0.4006) loss 1.9620 (2.9103) grad_norm 2.7702 (3.6437) [2021-04-16 22:01:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][100/1251] eta 0:07:27 lr 0.000010 time 0.2703 (0.3885) loss 1.8838 (2.8899) grad_norm 2.8452 (3.6111) [2021-04-16 22:01:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][110/1251] eta 0:07:11 lr 0.000010 time 0.2608 (0.3782) loss 3.1949 (2.8896) grad_norm 3.2009 (3.5838) [2021-04-16 22:01:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][120/1251] eta 0:06:59 lr 0.000010 time 0.2647 (0.3705) loss 3.0271 (2.8807) grad_norm 3.4272 (3.5732) [2021-04-16 22:01:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][130/1251] eta 0:06:47 lr 0.000010 time 0.2794 (0.3633) loss 2.4765 (2.8665) grad_norm 6.2716 (3.5864) [2021-04-16 22:01:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][140/1251] eta 0:06:37 lr 0.000010 time 0.2504 (0.3576) loss 2.2941 (2.8654) grad_norm 3.0778 (3.5854) [2021-04-16 22:01:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][150/1251] eta 0:06:29 lr 0.000010 time 0.3076 (0.3542) loss 2.0637 (2.8766) grad_norm 3.6589 (3.5678) [2021-04-16 22:01:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][160/1251] eta 0:06:22 lr 0.000010 time 0.2437 (0.3503) loss 2.2278 (2.8778) grad_norm 3.5121 (3.5590) [2021-04-16 22:01:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][170/1251] eta 0:06:14 lr 0.000010 time 0.2985 (0.3460) loss 2.5266 (2.8821) grad_norm 4.2649 (3.5571) [2021-04-16 22:01:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][180/1251] eta 0:06:06 lr 0.000010 time 0.2742 (0.3423) loss 3.0519 (2.9002) grad_norm 3.7764 (3.5508) [2021-04-16 22:01:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][190/1251] eta 0:05:59 lr 0.000010 time 0.2793 (0.3388) loss 3.5016 (2.9167) grad_norm 3.9135 (3.5436) [2021-04-16 22:01:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][200/1251] eta 0:05:52 lr 0.000010 time 0.2864 (0.3357) loss 3.4543 (2.9307) grad_norm 3.3792 (3.5377) [2021-04-16 22:01:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][210/1251] eta 0:05:46 lr 0.000010 time 0.2615 (0.3327) loss 2.6184 (2.9261) grad_norm 3.7777 (3.5464) [2021-04-16 22:01:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][220/1251] eta 0:05:40 lr 0.000010 time 0.2948 (0.3300) loss 2.8638 (2.9273) grad_norm 3.0996 (3.5384) [2021-04-16 22:01:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][230/1251] eta 0:05:34 lr 0.000010 time 0.2782 (0.3278) loss 2.9870 (2.9319) grad_norm 3.1246 (3.5268) [2021-04-16 22:01:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][240/1251] eta 0:05:29 lr 0.000010 time 0.2712 (0.3258) loss 2.7702 (2.9347) grad_norm 3.4261 (3.5187) [2021-04-16 22:01:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][250/1251] eta 0:05:24 lr 0.000010 time 0.3108 (0.3239) loss 2.0394 (2.9341) grad_norm 3.2134 (3.5197) [2021-04-16 22:01:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][260/1251] eta 0:05:19 lr 0.000010 time 0.2914 (0.3223) loss 3.4589 (2.9291) grad_norm 3.1804 (3.5118) [2021-04-16 22:01:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][270/1251] eta 0:05:14 lr 0.000010 time 0.2828 (0.3208) loss 3.2575 (2.9382) grad_norm 2.9098 (3.5027) [2021-04-16 22:01:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][280/1251] eta 0:05:09 lr 0.000010 time 0.2745 (0.3191) loss 2.8109 (2.9431) grad_norm 2.8233 (3.4953) [2021-04-16 22:01:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][290/1251] eta 0:05:05 lr 0.000010 time 0.2788 (0.3179) loss 3.1092 (2.9430) grad_norm 3.5375 (3.4922) [2021-04-16 22:01:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][300/1251] eta 0:05:00 lr 0.000010 time 0.2613 (0.3164) loss 2.8508 (2.9492) grad_norm 3.6545 (3.4872) [2021-04-16 22:02:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][310/1251] eta 0:04:56 lr 0.000010 time 0.2783 (0.3151) loss 2.4389 (2.9423) grad_norm 4.1981 (3.4894) [2021-04-16 22:02:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][320/1251] eta 0:04:52 lr 0.000010 time 0.2736 (0.3139) loss 3.5091 (2.9494) grad_norm 2.9107 (3.4805) [2021-04-16 22:02:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][330/1251] eta 0:04:48 lr 0.000010 time 0.2944 (0.3131) loss 3.5453 (2.9494) grad_norm 3.6314 (3.4843) [2021-04-16 22:02:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][340/1251] eta 0:04:44 lr 0.000010 time 0.2837 (0.3125) loss 3.0704 (2.9530) grad_norm 3.4118 (3.4809) [2021-04-16 22:02:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][350/1251] eta 0:04:40 lr 0.000010 time 0.2840 (0.3118) loss 2.5479 (2.9467) grad_norm 3.2141 (3.4758) [2021-04-16 22:02:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][360/1251] eta 0:04:36 lr 0.000010 time 0.2683 (0.3107) loss 3.2916 (2.9474) grad_norm 3.3199 (3.4820) [2021-04-16 22:02:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][370/1251] eta 0:04:32 lr 0.000010 time 0.2813 (0.3097) loss 2.8551 (2.9493) grad_norm 3.7577 (3.4842) [2021-04-16 22:02:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][380/1251] eta 0:04:29 lr 0.000010 time 0.2660 (0.3092) loss 3.1175 (2.9427) grad_norm 3.8799 (3.4931) [2021-04-16 22:02:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][390/1251] eta 0:04:25 lr 0.000010 time 0.2622 (0.3082) loss 3.1387 (2.9435) grad_norm 3.9476 (3.4918) [2021-04-16 22:02:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][400/1251] eta 0:04:21 lr 0.000010 time 0.2696 (0.3073) loss 3.4707 (2.9502) grad_norm 3.7252 (3.4938) [2021-04-16 22:02:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][410/1251] eta 0:04:17 lr 0.000010 time 0.2987 (0.3065) loss 2.5179 (2.9484) grad_norm 3.1638 (3.4864) [2021-04-16 22:02:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][420/1251] eta 0:04:14 lr 0.000010 time 0.2921 (0.3058) loss 3.3354 (2.9472) grad_norm 3.8530 (3.4844) [2021-04-16 22:02:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][430/1251] eta 0:04:10 lr 0.000010 time 0.2719 (0.3050) loss 2.3693 (2.9433) grad_norm 2.9516 (3.4842) [2021-04-16 22:02:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][440/1251] eta 0:04:06 lr 0.000010 time 0.2626 (0.3043) loss 3.4501 (2.9500) grad_norm 4.7613 (3.4979) [2021-04-16 22:02:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][450/1251] eta 0:04:03 lr 0.000010 time 0.2723 (0.3038) loss 2.9873 (2.9499) grad_norm 3.5024 (3.5081) [2021-04-16 22:02:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][460/1251] eta 0:03:59 lr 0.000010 time 0.2594 (0.3030) loss 2.2587 (2.9472) grad_norm 3.5481 (3.5084) [2021-04-16 22:02:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][470/1251] eta 0:03:56 lr 0.000010 time 0.2760 (0.3024) loss 3.2811 (2.9451) grad_norm 3.4229 (3.5127) [2021-04-16 22:02:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][480/1251] eta 0:03:52 lr 0.000010 time 0.2622 (0.3019) loss 3.5077 (2.9406) grad_norm 3.4781 (3.5161) [2021-04-16 22:02:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][490/1251] eta 0:03:49 lr 0.000010 time 0.2621 (0.3012) loss 2.8989 (2.9428) grad_norm 3.8946 (3.5162) [2021-04-16 22:02:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][500/1251] eta 0:03:45 lr 0.000010 time 0.2537 (0.3008) loss 3.3464 (2.9427) grad_norm 3.9056 (3.5132) [2021-04-16 22:02:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][510/1251] eta 0:03:42 lr 0.000010 time 0.2881 (0.3003) loss 3.2779 (2.9423) grad_norm 3.0766 (3.5073) [2021-04-16 22:02:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][520/1251] eta 0:03:39 lr 0.000010 time 0.2914 (0.3000) loss 2.6044 (2.9451) grad_norm 2.8079 (3.5047) [2021-04-16 22:03:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][530/1251] eta 0:03:35 lr 0.000010 time 0.2691 (0.2995) loss 3.5039 (2.9489) grad_norm 2.9842 (3.5052) [2021-04-16 22:03:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][540/1251] eta 0:03:32 lr 0.000010 time 0.2849 (0.2991) loss 3.1894 (2.9470) grad_norm 4.0241 (3.5077) [2021-04-16 22:03:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][550/1251] eta 0:03:29 lr 0.000010 time 0.2779 (0.2988) loss 3.4232 (2.9480) grad_norm 3.3056 (3.5084) [2021-04-16 22:03:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][560/1251] eta 0:03:26 lr 0.000010 time 0.2630 (0.2984) loss 3.1189 (2.9498) grad_norm 3.4807 (3.5067) [2021-04-16 22:03:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][570/1251] eta 0:03:22 lr 0.000010 time 0.2707 (0.2981) loss 2.3428 (2.9448) grad_norm 4.0406 (3.5101) [2021-04-16 22:03:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][580/1251] eta 0:03:20 lr 0.000010 time 0.2643 (0.2982) loss 2.6826 (2.9447) grad_norm 2.9128 (3.5053) [2021-04-16 22:03:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][590/1251] eta 0:03:16 lr 0.000010 time 0.2878 (0.2979) loss 2.1220 (2.9462) grad_norm 4.2361 (3.5045) [2021-04-16 22:03:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][600/1251] eta 0:03:13 lr 0.000010 time 0.2638 (0.2976) loss 2.8605 (2.9484) grad_norm 4.8874 (3.5137) [2021-04-16 22:03:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][610/1251] eta 0:03:10 lr 0.000010 time 0.2508 (0.2972) loss 3.5801 (2.9496) grad_norm 3.3814 (3.5153) [2021-04-16 22:03:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][620/1251] eta 0:03:07 lr 0.000010 time 0.2816 (0.2969) loss 3.4719 (2.9522) grad_norm 3.2609 (3.5153) [2021-04-16 22:03:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][630/1251] eta 0:03:04 lr 0.000010 time 0.2639 (0.2965) loss 3.1784 (2.9520) grad_norm 3.6011 (3.5136) [2021-04-16 22:03:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][640/1251] eta 0:03:01 lr 0.000010 time 0.2569 (0.2965) loss 3.9683 (2.9494) grad_norm 3.4339 (3.5147) [2021-04-16 22:03:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][650/1251] eta 0:02:57 lr 0.000010 time 0.2881 (0.2961) loss 2.9956 (2.9513) grad_norm 3.9989 (3.5132) [2021-04-16 22:03:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][660/1251] eta 0:02:54 lr 0.000010 time 0.2839 (0.2958) loss 2.4365 (2.9493) grad_norm 3.4066 (3.5134) [2021-04-16 22:03:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][670/1251] eta 0:02:51 lr 0.000010 time 0.2753 (0.2955) loss 3.2631 (2.9499) grad_norm 2.9831 (3.5099) [2021-04-16 22:03:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][680/1251] eta 0:02:48 lr 0.000010 time 0.2686 (0.2952) loss 3.4044 (2.9503) grad_norm 3.0183 (3.5107) [2021-04-16 22:03:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][690/1251] eta 0:02:45 lr 0.000010 time 0.2898 (0.2951) loss 3.3791 (2.9476) grad_norm 3.6013 (3.5113) [2021-04-16 22:03:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][700/1251] eta 0:02:42 lr 0.000010 time 0.2791 (0.2949) loss 2.2574 (2.9479) grad_norm 3.3952 (3.5118) [2021-04-16 22:03:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][710/1251] eta 0:02:39 lr 0.000010 time 0.3037 (0.2948) loss 3.3421 (2.9477) grad_norm 3.0736 (3.5114) [2021-04-16 22:03:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][720/1251] eta 0:02:36 lr 0.000010 time 0.2823 (0.2948) loss 3.3087 (2.9500) grad_norm 3.1156 (3.5101) [2021-04-16 22:03:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][730/1251] eta 0:02:33 lr 0.000010 time 0.2531 (0.2948) loss 3.4004 (2.9532) grad_norm 5.2743 (3.5121) [2021-04-16 22:04:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][740/1251] eta 0:02:30 lr 0.000010 time 0.2566 (0.2947) loss 3.1342 (2.9560) grad_norm 3.5221 (3.5146) [2021-04-16 22:04:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][750/1251] eta 0:02:27 lr 0.000010 time 0.2949 (0.2945) loss 1.9776 (2.9530) grad_norm 3.7963 (3.5186) [2021-04-16 22:04:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][760/1251] eta 0:02:24 lr 0.000010 time 0.2849 (0.2944) loss 3.2625 (2.9552) grad_norm 4.4761 (3.5181) [2021-04-16 22:04:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][770/1251] eta 0:02:21 lr 0.000010 time 0.2891 (0.2941) loss 3.0320 (2.9548) grad_norm 3.3132 (3.5191) [2021-04-16 22:04:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][780/1251] eta 0:02:18 lr 0.000010 time 0.2807 (0.2939) loss 3.1638 (2.9539) grad_norm 3.2426 (3.5179) [2021-04-16 22:04:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][790/1251] eta 0:02:15 lr 0.000010 time 0.2648 (0.2937) loss 2.6447 (2.9557) grad_norm 3.4296 (3.5183) [2021-04-16 22:04:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][800/1251] eta 0:02:12 lr 0.000010 time 0.2595 (0.2935) loss 3.2970 (2.9560) grad_norm 3.2708 (3.5144) [2021-04-16 22:04:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][810/1251] eta 0:02:09 lr 0.000010 time 0.2704 (0.2933) loss 2.0324 (2.9525) grad_norm 3.2380 (3.5140) [2021-04-16 22:04:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][820/1251] eta 0:02:06 lr 0.000010 time 0.2811 (0.2930) loss 3.2810 (2.9527) grad_norm 4.2988 (3.5183) [2021-04-16 22:04:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][830/1251] eta 0:02:03 lr 0.000010 time 0.2713 (0.2928) loss 3.3702 (2.9510) grad_norm 4.6117 (3.5198) [2021-04-16 22:04:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][840/1251] eta 0:02:00 lr 0.000010 time 0.2635 (0.2926) loss 2.4069 (2.9507) grad_norm 2.9566 (3.5170) [2021-04-16 22:04:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][850/1251] eta 0:01:57 lr 0.000010 time 0.2736 (0.2923) loss 2.3565 (2.9495) grad_norm 2.9517 (3.5155) [2021-04-16 22:04:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][860/1251] eta 0:01:54 lr 0.000010 time 0.2759 (0.2921) loss 3.1328 (2.9513) grad_norm 4.1846 (3.5164) [2021-04-16 22:04:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][870/1251] eta 0:01:51 lr 0.000010 time 0.2733 (0.2920) loss 2.8734 (2.9530) grad_norm 3.5944 (3.5142) [2021-04-16 22:04:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][880/1251] eta 0:01:48 lr 0.000010 time 0.2696 (0.2918) loss 2.5660 (2.9525) grad_norm 3.5438 (3.5155) [2021-04-16 22:04:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][890/1251] eta 0:01:45 lr 0.000010 time 0.2966 (0.2916) loss 3.0261 (2.9531) grad_norm 2.8994 (3.5115) [2021-04-16 22:04:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][900/1251] eta 0:01:42 lr 0.000010 time 0.2557 (0.2914) loss 3.7904 (2.9521) grad_norm 3.3172 (3.5112) [2021-04-16 22:04:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][910/1251] eta 0:01:39 lr 0.000010 time 0.2723 (0.2914) loss 2.9500 (2.9507) grad_norm 3.3011 (3.5130) [2021-04-16 22:04:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][920/1251] eta 0:01:36 lr 0.000010 time 0.2959 (0.2914) loss 3.9680 (2.9519) grad_norm 3.8141 (3.5157) [2021-04-16 22:04:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][930/1251] eta 0:01:33 lr 0.000010 time 0.2766 (0.2914) loss 3.2890 (2.9525) grad_norm 3.4141 (3.5135) [2021-04-16 22:04:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][940/1251] eta 0:01:30 lr 0.000010 time 0.2685 (0.2912) loss 3.2912 (2.9510) grad_norm 4.0102 (3.5149) [2021-04-16 22:04:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][950/1251] eta 0:01:27 lr 0.000010 time 0.2957 (0.2910) loss 2.8033 (2.9498) grad_norm 3.0938 (3.5161) [2021-04-16 22:05:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][960/1251] eta 0:01:24 lr 0.000010 time 0.2785 (0.2910) loss 3.0513 (2.9501) grad_norm 3.9916 (3.5157) [2021-04-16 22:05:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][970/1251] eta 0:01:21 lr 0.000010 time 0.2861 (0.2908) loss 2.2671 (2.9495) grad_norm 3.3987 (3.5147) [2021-04-16 22:05:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][980/1251] eta 0:01:18 lr 0.000010 time 0.2499 (0.2906) loss 3.3709 (2.9483) grad_norm 3.2874 (3.5135) [2021-04-16 22:05:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][990/1251] eta 0:01:15 lr 0.000010 time 0.3028 (0.2905) loss 2.9271 (2.9488) grad_norm 2.9996 (3.5115) [2021-04-16 22:05:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][1000/1251] eta 0:01:12 lr 0.000010 time 0.2739 (0.2903) loss 2.6429 (2.9475) grad_norm 2.9448 (3.5134) [2021-04-16 22:05:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][1010/1251] eta 0:01:09 lr 0.000010 time 0.2680 (0.2901) loss 3.3247 (2.9490) grad_norm 3.1423 (3.5119) [2021-04-16 22:05:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][1020/1251] eta 0:01:06 lr 0.000010 time 0.2826 (0.2900) loss 3.0179 (2.9474) grad_norm 3.2620 (3.5095) [2021-04-16 22:05:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][1030/1251] eta 0:01:04 lr 0.000010 time 0.2704 (0.2899) loss 3.1884 (2.9478) grad_norm 3.5949 (3.5085) [2021-04-16 22:05:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][1040/1251] eta 0:01:01 lr 0.000010 time 0.2866 (0.2898) loss 3.3574 (2.9464) grad_norm 3.8394 (3.5075) [2021-04-16 22:05:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][1050/1251] eta 0:00:58 lr 0.000010 time 0.2693 (0.2896) loss 3.3241 (2.9481) grad_norm 3.1531 (3.5064) [2021-04-16 22:05:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][1060/1251] eta 0:00:55 lr 0.000010 time 0.2778 (0.2896) loss 3.6545 (2.9504) grad_norm 3.2279 (3.5057) [2021-04-16 22:05:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][1070/1251] eta 0:00:52 lr 0.000010 time 0.2459 (0.2894) loss 3.1881 (2.9486) grad_norm 3.8379 (3.5045) [2021-04-16 22:05:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][1080/1251] eta 0:00:49 lr 0.000010 time 0.2875 (0.2893) loss 3.3975 (2.9488) grad_norm 3.4725 (3.5056) [2021-04-16 22:05:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][1090/1251] eta 0:00:46 lr 0.000010 time 0.2791 (0.2892) loss 2.6154 (2.9493) grad_norm 3.7984 (3.5038) [2021-04-16 22:05:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][1100/1251] eta 0:00:43 lr 0.000010 time 0.2633 (0.2891) loss 1.9607 (2.9473) grad_norm 3.4555 (3.5038) [2021-04-16 22:05:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][1110/1251] eta 0:00:40 lr 0.000010 time 0.2798 (0.2890) loss 2.3357 (2.9485) grad_norm 3.4571 (3.5042) [2021-04-16 22:05:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][1120/1251] eta 0:00:37 lr 0.000010 time 0.2689 (0.2889) loss 2.3519 (2.9486) grad_norm 3.9497 (3.5056) [2021-04-16 22:05:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][1130/1251] eta 0:00:34 lr 0.000010 time 0.2876 (0.2889) loss 2.9960 (2.9495) grad_norm 3.8299 (3.5062) [2021-04-16 22:05:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][1140/1251] eta 0:00:32 lr 0.000010 time 0.3059 (0.2887) loss 2.5936 (2.9486) grad_norm 3.4730 (3.5060) [2021-04-16 22:05:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][1150/1251] eta 0:00:29 lr 0.000010 time 0.2492 (0.2887) loss 3.1651 (2.9487) grad_norm 3.0725 (3.5031) [2021-04-16 22:05:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][1160/1251] eta 0:00:26 lr 0.000010 time 0.2705 (0.2887) loss 2.4693 (2.9482) grad_norm 3.2061 (3.5005) [2021-04-16 22:06:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][1170/1251] eta 0:00:23 lr 0.000010 time 0.2757 (0.2888) loss 2.8223 (2.9473) grad_norm 3.3036 (3.4989) [2021-04-16 22:06:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][1180/1251] eta 0:00:20 lr 0.000010 time 0.2644 (0.2887) loss 1.7913 (2.9468) grad_norm 3.4071 (3.4990) [2021-04-16 22:06:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][1190/1251] eta 0:00:17 lr 0.000010 time 0.2867 (0.2885) loss 2.7269 (2.9469) grad_norm 3.0395 (3.4979) [2021-04-16 22:06:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][1200/1251] eta 0:00:14 lr 0.000010 time 0.2435 (0.2884) loss 3.8258 (2.9485) grad_norm 3.5840 (3.4966) [2021-04-16 22:06:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][1210/1251] eta 0:00:11 lr 0.000010 time 0.2552 (0.2883) loss 3.1714 (2.9504) grad_norm 3.3983 (3.4978) [2021-04-16 22:06:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][1220/1251] eta 0:00:08 lr 0.000010 time 0.2655 (0.2883) loss 2.1539 (2.9492) grad_norm 4.0388 (3.4973) [2021-04-16 22:06:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][1230/1251] eta 0:00:06 lr 0.000010 time 0.2965 (0.2882) loss 3.1108 (2.9495) grad_norm 3.6896 (3.5042) [2021-04-16 22:06:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][1240/1251] eta 0:00:03 lr 0.000010 time 0.2725 (0.2880) loss 3.5439 (2.9503) grad_norm 3.2940 (3.5034) [2021-04-16 22:06:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [298/300][1250/1251] eta 0:00:00 lr 0.000010 time 0.2481 (0.2877) loss 2.6230 (2.9494) grad_norm 2.7679 (3.5033) [2021-04-16 22:06:27 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 298 training takes 0:06:04 [2021-04-16 22:06:27 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_298.pth saving...... [2021-04-16 22:06:47 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_298.pth saved !!! [2021-04-16 22:06:49 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 2.038 (2.038) Loss 0.8030 (0.8030) Acc@1 82.324 (82.324) Acc@5 95.605 (95.605) [2021-04-16 22:06:51 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.110 (0.314) Loss 0.7617 (0.8101) Acc@1 82.910 (81.543) Acc@5 95.605 (95.481) [2021-04-16 22:06:54 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.500 (0.316) Loss 0.8496 (0.8223) Acc@1 80.371 (81.283) Acc@5 95.312 (95.368) [2021-04-16 22:06:56 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.141 (0.275) Loss 0.8202 (0.8162) Acc@1 80.371 (81.250) Acc@5 94.824 (95.410) [2021-04-16 22:06:57 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.124 (0.249) Loss 0.8105 (0.8113) Acc@1 81.445 (81.307) Acc@5 95.801 (95.503) [2021-04-16 22:07:08 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 81.228 Acc@5 95.518 [2021-04-16 22:07:08 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 81.2% [2021-04-16 22:07:08 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 81.25% [2021-04-16 22:07:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][0/1251] eta 5:43:33 lr 0.000010 time 16.4776 (16.4776) loss 3.4841 (3.4841) grad_norm 3.1518 (3.1518) [2021-04-16 22:07:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][10/1251] eta 0:36:08 lr 0.000010 time 0.2747 (1.7475) loss 2.3167 (3.0441) grad_norm 4.4286 (3.4743) [2021-04-16 22:07:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][20/1251] eta 0:21:27 lr 0.000010 time 0.2615 (1.0462) loss 2.4633 (2.9086) grad_norm 3.3087 (3.4674) [2021-04-16 22:07:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][30/1251] eta 0:16:15 lr 0.000010 time 0.2574 (0.7986) loss 2.7135 (2.9007) grad_norm 3.8155 (3.4677) [2021-04-16 22:07:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][40/1251] eta 0:13:34 lr 0.000010 time 0.2547 (0.6726) loss 3.7672 (2.9273) grad_norm 3.4822 (3.4979) [2021-04-16 22:07:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][50/1251] eta 0:11:54 lr 0.000010 time 0.2798 (0.5952) loss 2.4247 (2.9023) grad_norm 3.0358 (3.5110) [2021-04-16 22:07:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][60/1251] eta 0:10:47 lr 0.000010 time 0.2701 (0.5434) loss 3.0018 (2.8819) grad_norm 3.4917 (3.5362) [2021-04-16 22:07:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][70/1251] eta 0:09:57 lr 0.000010 time 0.2688 (0.5058) loss 3.1744 (2.9054) grad_norm 3.4405 (3.5579) [2021-04-16 22:07:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][80/1251] eta 0:09:18 lr 0.000010 time 0.2742 (0.4773) loss 3.5978 (2.9149) grad_norm 3.9746 (3.5952) [2021-04-16 22:07:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][90/1251] eta 0:08:48 lr 0.000010 time 0.2615 (0.4556) loss 2.7302 (2.9574) grad_norm 3.7117 (3.5994) [2021-04-16 22:07:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][100/1251] eta 0:08:26 lr 0.000010 time 0.2760 (0.4398) loss 3.1615 (2.9536) grad_norm 3.1569 (3.6551) [2021-04-16 22:07:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][110/1251] eta 0:08:05 lr 0.000010 time 0.2705 (0.4254) loss 3.6097 (2.9607) grad_norm 3.4193 (3.6358) [2021-04-16 22:07:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][120/1251] eta 0:07:49 lr 0.000010 time 0.2672 (0.4148) loss 2.6345 (2.9563) grad_norm 3.7624 (3.6494) [2021-04-16 22:08:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][130/1251] eta 0:07:35 lr 0.000010 time 0.2868 (0.4060) loss 3.3931 (2.9543) grad_norm 3.3019 (3.6389) [2021-04-16 22:08:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][140/1251] eta 0:07:22 lr 0.000010 time 0.3069 (0.3980) loss 3.1599 (2.9715) grad_norm 3.3251 (3.6188) [2021-04-16 22:08:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][150/1251] eta 0:07:09 lr 0.000010 time 0.2769 (0.3900) loss 3.0192 (2.9584) grad_norm 3.3810 (3.5936) [2021-04-16 22:08:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][160/1251] eta 0:06:57 lr 0.000010 time 0.2471 (0.3831) loss 3.0727 (2.9474) grad_norm 3.6850 (3.5892) [2021-04-16 22:08:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][170/1251] eta 0:06:48 lr 0.000010 time 0.2742 (0.3776) loss 2.7968 (2.9455) grad_norm 4.8433 (3.5925) [2021-04-16 22:08:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][180/1251] eta 0:06:38 lr 0.000010 time 0.2785 (0.3725) loss 3.4442 (2.9598) grad_norm 3.9825 (3.5974) [2021-04-16 22:08:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][190/1251] eta 0:06:30 lr 0.000010 time 0.2963 (0.3678) loss 2.7901 (2.9568) grad_norm 3.3414 (3.6099) [2021-04-16 22:08:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][200/1251] eta 0:06:21 lr 0.000010 time 0.2516 (0.3631) loss 2.1571 (2.9584) grad_norm 2.9674 (3.6018) [2021-04-16 22:08:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][210/1251] eta 0:06:13 lr 0.000010 time 0.2702 (0.3590) loss 3.2750 (2.9613) grad_norm 3.4999 (3.5955) [2021-04-16 22:08:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][220/1251] eta 0:06:06 lr 0.000010 time 0.3047 (0.3554) loss 2.3458 (2.9529) grad_norm 4.1323 (3.5896) [2021-04-16 22:08:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][230/1251] eta 0:05:59 lr 0.000010 time 0.2824 (0.3522) loss 3.2015 (2.9590) grad_norm 8.6171 (3.6085) [2021-04-16 22:08:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][240/1251] eta 0:05:52 lr 0.000010 time 0.2620 (0.3490) loss 3.0823 (2.9548) grad_norm 3.3784 (3.6081) [2021-04-16 22:08:35 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][250/1251] eta 0:05:46 lr 0.000010 time 0.3134 (0.3463) loss 2.7475 (2.9382) grad_norm 3.0375 (3.5972) [2021-04-16 22:08:38 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][260/1251] eta 0:05:40 lr 0.000010 time 0.2770 (0.3435) loss 3.7322 (2.9398) grad_norm 3.0332 (3.5877) [2021-04-16 22:08:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][270/1251] eta 0:05:34 lr 0.000010 time 0.2964 (0.3413) loss 3.1248 (2.9415) grad_norm 4.8974 (3.6058) [2021-04-16 22:08:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][280/1251] eta 0:05:29 lr 0.000010 time 0.3014 (0.3391) loss 2.0923 (2.9399) grad_norm 3.3678 (3.5967) [2021-04-16 22:08:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][290/1251] eta 0:05:23 lr 0.000010 time 0.2726 (0.3371) loss 2.4436 (2.9392) grad_norm 3.2947 (3.5871) [2021-04-16 22:08:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][300/1251] eta 0:05:18 lr 0.000010 time 0.2871 (0.3352) loss 2.8446 (2.9474) grad_norm 3.5426 (3.5931) [2021-04-16 22:08:52 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][310/1251] eta 0:05:13 lr 0.000010 time 0.2739 (0.3334) loss 3.8259 (2.9429) grad_norm 3.5734 (3.5884) [2021-04-16 22:08:55 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][320/1251] eta 0:05:08 lr 0.000010 time 0.2549 (0.3318) loss 3.3066 (2.9391) grad_norm 3.4618 (3.5899) [2021-04-16 22:08:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][330/1251] eta 0:05:04 lr 0.000010 time 0.2879 (0.3303) loss 3.5657 (2.9432) grad_norm 3.0800 (3.5920) [2021-04-16 22:09:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][340/1251] eta 0:04:59 lr 0.000010 time 0.2934 (0.3287) loss 3.3508 (2.9419) grad_norm 2.6007 (3.5895) [2021-04-16 22:09:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][350/1251] eta 0:04:54 lr 0.000010 time 0.2456 (0.3273) loss 3.1211 (2.9416) grad_norm 4.2728 (3.5952) [2021-04-16 22:09:06 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][360/1251] eta 0:04:50 lr 0.000010 time 0.2703 (0.3264) loss 3.5151 (2.9443) grad_norm 4.4921 (3.5921) [2021-04-16 22:09:09 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][370/1251] eta 0:04:46 lr 0.000010 time 0.3069 (0.3257) loss 3.3897 (2.9450) grad_norm 3.0433 (3.5941) [2021-04-16 22:09:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][380/1251] eta 0:04:42 lr 0.000010 time 0.2823 (0.3245) loss 2.1869 (2.9434) grad_norm 3.3438 (3.5944) [2021-04-16 22:09:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][390/1251] eta 0:04:38 lr 0.000010 time 0.2939 (0.3235) loss 2.6061 (2.9470) grad_norm 4.3984 (3.5910) [2021-04-16 22:09:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][400/1251] eta 0:04:34 lr 0.000010 time 0.2488 (0.3226) loss 2.0944 (2.9444) grad_norm 4.6573 (3.5926) [2021-04-16 22:09:20 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][410/1251] eta 0:04:30 lr 0.000010 time 0.2892 (0.3214) loss 2.5456 (2.9473) grad_norm 3.4587 (3.5935) [2021-04-16 22:09:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][420/1251] eta 0:04:26 lr 0.000010 time 0.3007 (0.3204) loss 3.3383 (2.9460) grad_norm 3.0628 (3.5822) [2021-04-16 22:09:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][430/1251] eta 0:04:22 lr 0.000010 time 0.2967 (0.3194) loss 4.2008 (2.9499) grad_norm 3.3084 (nan) [2021-04-16 22:09:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][440/1251] eta 0:04:18 lr 0.000010 time 0.2809 (0.3184) loss 3.1631 (2.9546) grad_norm 3.1240 (nan) [2021-04-16 22:09:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][450/1251] eta 0:04:14 lr 0.000010 time 0.2927 (0.3176) loss 2.9868 (2.9559) grad_norm 3.0603 (nan) [2021-04-16 22:09:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][460/1251] eta 0:04:10 lr 0.000010 time 0.3049 (0.3171) loss 1.6274 (2.9556) grad_norm 3.0465 (nan) [2021-04-16 22:09:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][470/1251] eta 0:04:07 lr 0.000010 time 0.2925 (0.3163) loss 2.2580 (2.9523) grad_norm 3.5597 (nan) [2021-04-16 22:09:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][480/1251] eta 0:04:03 lr 0.000010 time 0.2489 (0.3154) loss 3.2235 (2.9543) grad_norm 3.5147 (nan) [2021-04-16 22:09:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][490/1251] eta 0:03:59 lr 0.000010 time 0.2681 (0.3149) loss 3.4740 (2.9590) grad_norm 4.7530 (nan) [2021-04-16 22:09:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][500/1251] eta 0:03:55 lr 0.000010 time 0.2879 (0.3141) loss 2.1688 (2.9568) grad_norm 3.8349 (nan) [2021-04-16 22:09:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][510/1251] eta 0:03:52 lr 0.000010 time 0.2848 (0.3133) loss 1.9640 (2.9599) grad_norm 3.4434 (nan) [2021-04-16 22:09:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][520/1251] eta 0:03:48 lr 0.000010 time 0.2663 (0.3126) loss 3.5391 (2.9587) grad_norm 5.4074 (nan) [2021-04-16 22:09:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][530/1251] eta 0:03:44 lr 0.000010 time 0.2790 (0.3120) loss 3.0678 (2.9591) grad_norm 3.0439 (nan) [2021-04-16 22:09:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][540/1251] eta 0:03:41 lr 0.000010 time 0.2798 (0.3114) loss 3.7050 (2.9592) grad_norm 3.5908 (nan) [2021-04-16 22:09:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][550/1251] eta 0:03:37 lr 0.000010 time 0.2555 (0.3107) loss 2.3841 (2.9631) grad_norm 3.3497 (nan) [2021-04-16 22:10:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][560/1251] eta 0:03:34 lr 0.000010 time 0.2855 (0.3101) loss 1.7251 (2.9595) grad_norm 3.7884 (nan) [2021-04-16 22:10:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][570/1251] eta 0:03:31 lr 0.000010 time 0.2989 (0.3099) loss 3.4253 (2.9579) grad_norm 2.7496 (nan) [2021-04-16 22:10:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][580/1251] eta 0:03:27 lr 0.000010 time 0.2845 (0.3095) loss 2.8415 (2.9586) grad_norm 3.0487 (nan) [2021-04-16 22:10:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][590/1251] eta 0:03:24 lr 0.000010 time 0.3056 (0.3090) loss 3.5830 (2.9613) grad_norm 3.8624 (nan) [2021-04-16 22:10:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][600/1251] eta 0:03:20 lr 0.000010 time 0.2846 (0.3085) loss 3.4076 (2.9643) grad_norm 3.5190 (nan) [2021-04-16 22:10:16 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][610/1251] eta 0:03:17 lr 0.000010 time 0.2791 (0.3079) loss 2.9976 (2.9635) grad_norm 3.2683 (nan) [2021-04-16 22:10:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][620/1251] eta 0:03:13 lr 0.000010 time 0.2760 (0.3074) loss 2.7935 (2.9614) grad_norm 3.9224 (nan) [2021-04-16 22:10:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][630/1251] eta 0:03:10 lr 0.000010 time 0.2738 (0.3070) loss 1.8946 (2.9607) grad_norm 3.9847 (nan) [2021-04-16 22:10:24 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][640/1251] eta 0:03:07 lr 0.000010 time 0.2777 (0.3065) loss 3.4449 (2.9623) grad_norm 3.3769 (nan) [2021-04-16 22:10:27 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][650/1251] eta 0:03:03 lr 0.000010 time 0.2681 (0.3060) loss 2.1826 (2.9623) grad_norm 4.1844 (nan) [2021-04-16 22:10:30 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][660/1251] eta 0:03:00 lr 0.000010 time 0.3019 (0.3056) loss 2.7411 (2.9636) grad_norm 3.3982 (nan) [2021-04-16 22:10:33 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][670/1251] eta 0:02:57 lr 0.000010 time 0.2855 (0.3052) loss 3.4958 (2.9649) grad_norm 3.1348 (nan) [2021-04-16 22:10:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][680/1251] eta 0:02:54 lr 0.000010 time 0.2883 (0.3050) loss 3.0544 (2.9635) grad_norm 3.5763 (nan) [2021-04-16 22:10:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][690/1251] eta 0:02:50 lr 0.000010 time 0.2802 (0.3047) loss 2.5357 (2.9655) grad_norm 2.9244 (nan) [2021-04-16 22:10:41 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][700/1251] eta 0:02:47 lr 0.000010 time 0.4134 (0.3045) loss 2.7500 (2.9665) grad_norm 3.3917 (nan) [2021-04-16 22:10:44 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][710/1251] eta 0:02:44 lr 0.000010 time 0.2906 (0.3040) loss 3.3711 (2.9678) grad_norm 4.5081 (nan) [2021-04-16 22:10:47 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][720/1251] eta 0:02:41 lr 0.000010 time 0.2631 (0.3038) loss 3.7282 (2.9728) grad_norm 3.6501 (nan) [2021-04-16 22:10:50 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][730/1251] eta 0:02:38 lr 0.000010 time 0.2839 (0.3035) loss 2.6867 (2.9715) grad_norm 3.7283 (nan) [2021-04-16 22:10:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][740/1251] eta 0:02:35 lr 0.000010 time 0.2904 (0.3034) loss 3.3261 (2.9732) grad_norm 3.2278 (nan) [2021-04-16 22:10:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][750/1251] eta 0:02:31 lr 0.000010 time 0.2729 (0.3031) loss 3.2235 (2.9727) grad_norm 3.1842 (nan) [2021-04-16 22:10:58 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][760/1251] eta 0:02:28 lr 0.000010 time 0.2787 (0.3027) loss 3.3821 (2.9713) grad_norm 3.2808 (nan) [2021-04-16 22:11:01 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][770/1251] eta 0:02:25 lr 0.000010 time 0.2823 (0.3024) loss 2.6664 (2.9704) grad_norm 3.0643 (nan) [2021-04-16 22:11:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][780/1251] eta 0:02:22 lr 0.000010 time 0.2748 (0.3021) loss 3.2971 (2.9694) grad_norm 3.0518 (nan) [2021-04-16 22:11:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][790/1251] eta 0:02:19 lr 0.000010 time 0.2921 (0.3018) loss 2.6678 (2.9659) grad_norm 3.2593 (nan) [2021-04-16 22:11:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][800/1251] eta 0:02:16 lr 0.000010 time 0.3037 (0.3016) loss 2.3444 (2.9656) grad_norm 3.6631 (nan) [2021-04-16 22:11:12 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][810/1251] eta 0:02:12 lr 0.000010 time 0.2677 (0.3012) loss 3.3842 (2.9658) grad_norm 3.2220 (nan) [2021-04-16 22:11:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][820/1251] eta 0:02:09 lr 0.000010 time 0.2842 (0.3009) loss 2.5411 (2.9665) grad_norm 4.4135 (nan) [2021-04-16 22:11:18 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][830/1251] eta 0:02:06 lr 0.000010 time 0.2865 (0.3006) loss 3.3360 (2.9678) grad_norm 4.1068 (nan) [2021-04-16 22:11:21 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][840/1251] eta 0:02:03 lr 0.000010 time 0.2888 (0.3004) loss 3.0118 (2.9692) grad_norm 3.9003 (nan) [2021-04-16 22:11:23 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][850/1251] eta 0:02:00 lr 0.000010 time 0.2948 (0.3001) loss 2.7218 (2.9710) grad_norm 3.4716 (nan) [2021-04-16 22:11:26 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][860/1251] eta 0:01:57 lr 0.000010 time 0.2925 (0.2999) loss 3.4111 (2.9746) grad_norm 2.9878 (nan) [2021-04-16 22:11:29 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][870/1251] eta 0:01:54 lr 0.000010 time 0.2844 (0.2996) loss 3.1102 (2.9751) grad_norm 3.3954 (nan) [2021-04-16 22:11:32 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][880/1251] eta 0:01:51 lr 0.000010 time 0.2668 (0.2993) loss 2.2076 (2.9747) grad_norm 3.5437 (nan) [2021-04-16 22:11:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][890/1251] eta 0:01:47 lr 0.000010 time 0.2713 (0.2991) loss 3.0186 (2.9747) grad_norm 3.5920 (nan) [2021-04-16 22:11:37 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][900/1251] eta 0:01:44 lr 0.000010 time 0.2943 (0.2988) loss 3.0394 (2.9729) grad_norm 3.1676 (nan) [2021-04-16 22:11:40 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][910/1251] eta 0:01:41 lr 0.000010 time 0.2806 (0.2986) loss 3.0295 (2.9743) grad_norm 4.5112 (nan) [2021-04-16 22:11:43 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][920/1251] eta 0:01:38 lr 0.000010 time 0.2706 (0.2983) loss 3.0547 (2.9760) grad_norm 3.4098 (nan) [2021-04-16 22:11:46 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][930/1251] eta 0:01:35 lr 0.000010 time 0.2767 (0.2981) loss 2.0017 (2.9769) grad_norm 3.0704 (nan) [2021-04-16 22:11:49 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][940/1251] eta 0:01:32 lr 0.000010 time 0.2910 (0.2982) loss 3.2792 (2.9770) grad_norm 3.3816 (nan) [2021-04-16 22:11:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][950/1251] eta 0:01:29 lr 0.000010 time 0.2781 (0.2979) loss 3.3608 (2.9781) grad_norm 4.9254 (nan) [2021-04-16 22:11:54 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][960/1251] eta 0:01:26 lr 0.000010 time 0.2786 (0.2977) loss 2.3708 (2.9768) grad_norm 3.2597 (nan) [2021-04-16 22:11:57 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][970/1251] eta 0:01:23 lr 0.000010 time 0.2701 (0.2975) loss 2.9622 (2.9765) grad_norm 3.7797 (nan) [2021-04-16 22:12:00 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][980/1251] eta 0:01:20 lr 0.000010 time 0.2914 (0.2974) loss 3.0717 (2.9747) grad_norm 3.5898 (nan) [2021-04-16 22:12:03 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][990/1251] eta 0:01:17 lr 0.000010 time 0.2716 (0.2973) loss 2.6709 (2.9749) grad_norm 3.4582 (nan) [2021-04-16 22:12:05 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][1000/1251] eta 0:01:14 lr 0.000010 time 0.2697 (0.2971) loss 2.9782 (2.9731) grad_norm 4.1173 (nan) [2021-04-16 22:12:08 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][1010/1251] eta 0:01:11 lr 0.000010 time 0.2556 (0.2969) loss 3.3529 (2.9732) grad_norm 4.8788 (nan) [2021-04-16 22:12:11 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][1020/1251] eta 0:01:08 lr 0.000010 time 0.2804 (0.2968) loss 3.2633 (2.9753) grad_norm 3.9980 (nan) [2021-04-16 22:12:14 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][1030/1251] eta 0:01:05 lr 0.000010 time 0.2577 (0.2966) loss 3.5127 (2.9733) grad_norm 3.3436 (nan) [2021-04-16 22:12:17 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][1040/1251] eta 0:01:02 lr 0.000010 time 0.2720 (0.2965) loss 2.2572 (2.9709) grad_norm 3.6270 (nan) [2021-04-16 22:12:19 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][1050/1251] eta 0:00:59 lr 0.000010 time 0.2722 (0.2963) loss 2.6939 (2.9711) grad_norm 3.3436 (nan) [2021-04-16 22:12:22 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][1060/1251] eta 0:00:56 lr 0.000010 time 0.2574 (0.2961) loss 2.2471 (2.9680) grad_norm 2.7985 (nan) [2021-04-16 22:12:25 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][1070/1251] eta 0:00:53 lr 0.000010 time 0.2993 (0.2961) loss 2.7497 (2.9675) grad_norm 3.3208 (nan) [2021-04-16 22:12:28 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][1080/1251] eta 0:00:50 lr 0.000010 time 0.2834 (0.2960) loss 3.5962 (2.9679) grad_norm 4.2411 (nan) [2021-04-16 22:12:31 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][1090/1251] eta 0:00:47 lr 0.000010 time 0.2766 (0.2959) loss 3.3602 (2.9692) grad_norm 2.9455 (nan) [2021-04-16 22:12:34 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][1100/1251] eta 0:00:44 lr 0.000010 time 0.2903 (0.2957) loss 3.1657 (2.9710) grad_norm 3.0902 (nan) [2021-04-16 22:12:36 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][1110/1251] eta 0:00:41 lr 0.000010 time 0.2774 (0.2956) loss 2.5958 (2.9694) grad_norm 3.5996 (nan) [2021-04-16 22:12:39 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][1120/1251] eta 0:00:38 lr 0.000010 time 0.2707 (0.2954) loss 2.7454 (2.9703) grad_norm 3.1783 (nan) [2021-04-16 22:12:42 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][1130/1251] eta 0:00:35 lr 0.000010 time 0.2779 (0.2953) loss 2.3676 (2.9696) grad_norm 3.5495 (nan) [2021-04-16 22:12:45 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][1140/1251] eta 0:00:32 lr 0.000010 time 0.2717 (0.2951) loss 1.9471 (2.9674) grad_norm 3.8998 (nan) [2021-04-16 22:12:48 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][1150/1251] eta 0:00:29 lr 0.000010 time 0.2839 (0.2951) loss 3.3926 (2.9670) grad_norm 3.7625 (nan) [2021-04-16 22:12:51 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][1160/1251] eta 0:00:26 lr 0.000010 time 0.2526 (0.2950) loss 2.8482 (2.9660) grad_norm 3.9539 (nan) [2021-04-16 22:12:53 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][1170/1251] eta 0:00:23 lr 0.000010 time 0.3028 (0.2949) loss 2.8109 (2.9655) grad_norm 3.6347 (nan) [2021-04-16 22:12:56 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][1180/1251] eta 0:00:20 lr 0.000010 time 0.2740 (0.2947) loss 3.0481 (2.9655) grad_norm 3.5018 (nan) [2021-04-16 22:12:59 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][1190/1251] eta 0:00:17 lr 0.000010 time 0.2714 (0.2946) loss 2.5945 (2.9665) grad_norm 3.6522 (nan) [2021-04-16 22:13:02 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][1200/1251] eta 0:00:15 lr 0.000010 time 0.2421 (0.2944) loss 3.3664 (2.9658) grad_norm 3.4272 (nan) [2021-04-16 22:13:04 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][1210/1251] eta 0:00:12 lr 0.000010 time 0.2864 (0.2943) loss 3.0617 (2.9651) grad_norm 3.5248 (nan) [2021-04-16 22:13:07 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][1220/1251] eta 0:00:09 lr 0.000010 time 0.2536 (0.2941) loss 2.8032 (2.9657) grad_norm 3.4971 (nan) [2021-04-16 22:13:10 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][1230/1251] eta 0:00:06 lr 0.000010 time 0.2717 (0.2940) loss 2.9414 (2.9639) grad_norm 3.8009 (nan) [2021-04-16 22:13:13 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][1240/1251] eta 0:00:03 lr 0.000010 time 0.2489 (0.2937) loss 3.5120 (2.9629) grad_norm 3.6568 (nan) [2021-04-16 22:13:15 swin_tiny_patch4_window7_224] (main.py 231): INFO Train: [299/300][1250/1251] eta 0:00:00 lr 0.000010 time 0.2487 (0.2934) loss 3.6834 (2.9625) grad_norm 3.4666 (nan) [2021-04-16 22:13:22 swin_tiny_patch4_window7_224] (main.py 238): INFO EPOCH 299 training takes 0:06:13 [2021-04-16 22:13:22 swin_tiny_patch4_window7_224] (utils.py 143): INFO ./ckpt_epoch_299.pth saving...... [2021-04-16 22:13:45 swin_tiny_patch4_window7_224] (utils.py 145): INFO ./ckpt_epoch_299.pth saved !!! [2021-04-16 22:13:46 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [0/49] Time 1.034 (1.034) Loss 0.7496 (0.7496) Acc@1 82.812 (82.812) Acc@5 95.801 (95.801) [2021-04-16 22:13:48 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [10/49] Time 0.704 (0.266) Loss 0.8734 (0.8130) Acc@1 79.688 (81.303) Acc@5 94.629 (95.543) [2021-04-16 22:13:50 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [20/49] Time 0.154 (0.237) Loss 0.8315 (0.8160) Acc@1 79.590 (81.157) Acc@5 95.508 (95.526) [2021-04-16 22:13:53 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [30/49] Time 0.258 (0.241) Loss 0.8308 (0.8115) Acc@1 82.227 (81.307) Acc@5 94.336 (95.486) [2021-04-16 22:13:54 swin_tiny_patch4_window7_224] (main.py 278): INFO Test: [40/49] Time 0.073 (0.219) Loss 0.8382 (0.8057) Acc@1 80.762 (81.357) Acc@5 95.312 (95.565) [2021-04-16 22:14:09 swin_tiny_patch4_window7_224] (main.py 284): INFO * Acc@1 81.286 Acc@5 95.544 [2021-04-16 22:14:09 swin_tiny_patch4_window7_224] (main.py 151): INFO Accuracy of the network on the 50000 test images: 81.3% [2021-04-16 22:14:09 swin_tiny_patch4_window7_224] (main.py 153): INFO Max accuracy: 81.29%