Training input paths: training/void/void_train_image_1500.txt training/void/void_train_sparse_depth_1500.txt training/void/void_train_intrinsics_1500.txt Validation input paths: testing/void/void_test_image_1500.txt testing/void/void_test_sparse_depth_1500.txt testing/void/void_test_intrinsics_1500.txt testing/void/void_test_ground_truth_1500.txt Input settings: n_batch=8 n_height=480 n_width=640 input_channels_image=3 input_channels_depth=2 normalized_image_range=[0.0, 1.0] outlier_removal_kernel_size=7 outlier_removal_threshold=1.50 Sparse to dense pooling settings: min_pool_sizes_sparse_to_dense_pool=[15, 17, 19] max_pool_sizes_sparse_to_dense_pool=[23, 27] n_convolution_sparse_to_dense_pool=3 n_filter_sparse_to_dense_pool=8 Depth network settings: n_filters_encoder_image=[48, 96, 192, 384, 384] n_filters_encoder_depth=[16, 32, 64, 128, 128] resolutions_backprojection=[0, 1, 2, 3] n_filters_decoder=[256, 128, 128, 64, 12] deconv_type=up min_predict_depth=0.10 max_predict_depth=8.00 Weight settings: n_parameter=6957764 n_parameter_depth=6957764 n_parameter_pose=0 weight_initializer=xavier_normal activation_func=leaky_relu Training settings: n_sample=35917 n_epoch=15 n_step=67350 learning_schedule=[0-44890 : 0.0001, 44890-67335 : 5e-05] Augmentation settings: augmentation_schedule=[0--4489 : 1.0] augmentation_random_crop_type=['horizontal', 'vertical', 'anchored'] augmentation_random_flip_type=['none'] augmentation_random_remove_points=[0.3, 0.6] augmentation_random_noise_type=none augmentation_random_noise_spread=-1.0 Loss function settings: w_color=1.5e-01 w_structure=9.5e-01 w_sparse_depth=2.0e+00 w_smoothness=2.0e+00 w_weight_decay_depth=0.0e+00 w_weight_decay_pose=0.0e+00 Evaluation settings: min_evaluate_depth=0.20 max_evaluate_depth=5.00 Checkpoint settings: checkpoint_path=trained_kbnet/void1500/kbnet_model checkpoint_save_frequency=1000 validation_start_step=5000 Tensorboard settings: event_path=trained_kbnet/void1500/kbnet_model/events log_summary_frequency=1000 n_summary_display=4 Hardware settings: device=cuda n_thread=72 Begin training... Step= 1000/67350 Loss=1.32151 Time Elapsed=0.14h Time Remaining=9.27h Step= 2000/67350 Loss=1.05710 Time Elapsed=0.28h Time Remaining=9.09h Step= 3000/67350 Loss=1.47754 Time Elapsed=0.42h Time Remaining=8.94h Step= 4000/67350 Loss=1.12613 Time Elapsed=0.55h Time Remaining=8.79h Step= 5000/67350 Loss=1.19926 Time Elapsed=0.69h Time Remaining=8.64h Training input paths: training/void/void_train_image_1500.txt training/void/void_train_sparse_depth_1500.txt training/void/void_train_intrinsics_1500.txt Validation input paths: testing/void/void_test_image_1500.txt testing/void/void_test_sparse_depth_1500.txt testing/void/void_test_intrinsics_1500.txt testing/void/void_test_ground_truth_1500.txt Input settings: n_batch=8 n_height=480 n_width=640 input_channels_image=3 input_channels_depth=2 normalized_image_range=[0.0, 1.0] outlier_removal_kernel_size=7 outlier_removal_threshold=1.50 Sparse to dense pooling settings: min_pool_sizes_sparse_to_dense_pool=[15, 17, 19] max_pool_sizes_sparse_to_dense_pool=[23, 27] n_convolution_sparse_to_dense_pool=3 n_filter_sparse_to_dense_pool=8 Depth network settings: n_filters_encoder_image=[48, 96, 192, 384, 384] n_filters_encoder_depth=[16, 32, 64, 128, 128] resolutions_backprojection=[0, 1, 2, 3] n_filters_decoder=[256, 128, 128, 64, 12] deconv_type=up min_predict_depth=0.10 max_predict_depth=8.00 Weight settings: n_parameter=6957764 n_parameter_depth=6957764 n_parameter_pose=0 weight_initializer=xavier_normal activation_func=leaky_relu Training settings: n_sample=35917 n_epoch=15 n_step=67350 learning_schedule=[0-44890 : 0.0001, 44890-67335 : 5e-05] Augmentation settings: augmentation_schedule=[0--4489 : 1.0] augmentation_random_crop_type=['horizontal', 'vertical', 'anchored'] augmentation_random_flip_type=['none'] augmentation_random_remove_points=[0.3, 0.6] augmentation_random_noise_type=none augmentation_random_noise_spread=-1.0 Loss function settings: w_color=1.5e-01 w_structure=9.5e-01 w_sparse_depth=2.0e+00 w_smoothness=2.0e+00 w_weight_decay_depth=0.0e+00 w_weight_decay_pose=0.0e+00 Evaluation settings: min_evaluate_depth=0.20 max_evaluate_depth=5.00 Checkpoint settings: checkpoint_path=trained_kbnet/void1500/kbnet_model checkpoint_save_frequency=1000 validation_start_step=5000 Tensorboard settings: event_path=trained_kbnet/void1500/kbnet_model/events log_summary_frequency=1000 n_summary_display=4 Hardware settings: device=cuda n_thread=72 Begin training... Training input paths: training/void/void_train_image_1500.txt training/void/void_train_sparse_depth_1500.txt training/void/void_train_intrinsics_1500.txt Validation input paths: testing/void/void_test_image_1500.txt testing/void/void_test_sparse_depth_1500.txt testing/void/void_test_intrinsics_1500.txt testing/void/void_test_ground_truth_1500.txt Input settings: n_batch=8 n_height=480 n_width=640 input_channels_image=3 input_channels_depth=2 normalized_image_range=[0.0, 1.0] outlier_removal_kernel_size=7 outlier_removal_threshold=1.50 Sparse to dense pooling settings: min_pool_sizes_sparse_to_dense_pool=[15, 17, 19] max_pool_sizes_sparse_to_dense_pool=[23, 27] n_convolution_sparse_to_dense_pool=3 n_filter_sparse_to_dense_pool=8 Depth network settings: n_filters_encoder_image=[48, 96, 192, 384, 384] n_filters_encoder_depth=[16, 32, 64, 128, 128] resolutions_backprojection=[0, 1, 2, 3] n_filters_decoder=[256, 128, 128, 64, 12] deconv_type=up min_predict_depth=0.10 max_predict_depth=8.00 Weight settings: n_parameter=6957764 n_parameter_depth=6957764 n_parameter_pose=0 weight_initializer=xavier_normal activation_func=leaky_relu Training settings: n_sample=35917 n_epoch=15 n_step=67350 learning_schedule=[0-44890 : 0.0001, 44890-67335 : 5e-05] Augmentation settings: augmentation_schedule=[0--4489 : 1.0] augmentation_random_crop_type=['horizontal', 'vertical', 'anchored'] augmentation_random_flip_type=['none'] augmentation_random_remove_points=[0.3, 0.6] augmentation_random_noise_type=none augmentation_random_noise_spread=-1.0 Loss function settings: w_color=1.5e-01 w_structure=9.5e-01 w_sparse_depth=2.0e+00 w_smoothness=2.0e+00 w_weight_decay_depth=0.0e+00 w_weight_decay_pose=0.0e+00 Evaluation settings: min_evaluate_depth=0.20 max_evaluate_depth=5.00 Checkpoint settings: checkpoint_path=trained_kbnet/void1500/kbnet_model checkpoint_save_frequency=1000 validation_start_step=5000 Tensorboard settings: event_path=trained_kbnet/void1500/kbnet_model/events log_summary_frequency=1000 n_summary_display=4 Hardware settings: device=cuda n_thread=72 Begin training... Step= 1000/67350 Loss=1.15377 Time Elapsed=0.14h Time Remaining=9.59h Step= 2000/67350 Loss=1.17042 Time Elapsed=0.29h Time Remaining=9.42h Step= 3000/67350 Loss=1.34876 Time Elapsed=0.43h Time Remaining=9.26h Step= 4000/67350 Loss=1.01620 Time Elapsed=0.57h Time Remaining=9.11h Step= 5000/67350 Loss=1.11315 Time Elapsed=0.72h Time Remaining=8.97h Validation results: Step MAE RMSE iMAE iRMSE 5000 57.022 115.433 32.515 76.328 Best results: Step MAE RMSE iMAE iRMSE 5000 57.022 115.433 32.515 76.328 Step= 6000/67350 Loss=1.09535 Time Elapsed=0.87h Time Remaining=8.90h Validation results: Step MAE RMSE iMAE iRMSE 6000 68.721 127.336 40.059 76.804 Best results: Step MAE RMSE iMAE iRMSE 5000 57.022 115.433 32.515 76.328 Step= 7000/67350 Loss=1.08475 Time Elapsed=1.02h Time Remaining=8.80h Validation results: Step MAE RMSE iMAE iRMSE 7000 51.801 109.523 30.879 65.347 Best results: Step MAE RMSE iMAE iRMSE 7000 51.801 109.523 30.879 65.347 Step= 8000/67350 Loss=1.20288 Time Elapsed=1.17h Time Remaining=8.69h Validation results: Step MAE RMSE iMAE iRMSE 8000 62.362 124.378 37.552 74.723 Best results: Step MAE RMSE iMAE iRMSE 7000 51.801 109.523 30.879 65.347 Step= 9000/67350 Loss=1.27384 Time Elapsed=1.32h Time Remaining=8.58h Validation results: Step MAE RMSE iMAE iRMSE 9000 84.994 148.822 50.429 88.866 Best results: Step MAE RMSE iMAE iRMSE 7000 51.801 109.523 30.879 65.347 Step= 10000/67350 Loss=1.17693 Time Elapsed=1.47h Time Remaining=8.45h Validation results: Step MAE RMSE iMAE iRMSE 10000 59.241 126.209 32.530 68.380 Best results: Step MAE RMSE iMAE iRMSE 7000 51.801 109.523 30.879 65.347 Step= 11000/67350 Loss=1.20359 Time Elapsed=1.62h Time Remaining=8.31h Validation results: Step MAE RMSE iMAE iRMSE 11000 71.718 134.747 41.915 79.413 Best results: Step MAE RMSE iMAE iRMSE 7000 51.801 109.523 30.879 65.347 Step= 12000/67350 Loss=1.25831 Time Elapsed=1.77h Time Remaining=8.18h Validation results: Step MAE RMSE iMAE iRMSE 12000 49.350 99.850 28.267 55.779 Best results: Step MAE RMSE iMAE iRMSE 12000 49.350 99.850 28.267 55.779 Step= 13000/67350 Loss=1.28071 Time Elapsed=1.92h Time Remaining=8.04h Validation results: Step MAE RMSE iMAE iRMSE 13000 54.168 108.818 31.483 63.443 Best results: Step MAE RMSE iMAE iRMSE 12000 49.350 99.850 28.267 55.779 Step= 14000/67350 Loss=1.06663 Time Elapsed=2.07h Time Remaining=7.90h Validation results: Step MAE RMSE iMAE iRMSE 14000 56.878 122.163 31.236 66.569 Best results: Step MAE RMSE iMAE iRMSE 12000 49.350 99.850 28.267 55.779 Step= 15000/67350 Loss=1.15728 Time Elapsed=2.22h Time Remaining=7.76h Validation results: Step MAE RMSE iMAE iRMSE 15000 43.464 95.555 24.712 53.424 Best results: Step MAE RMSE iMAE iRMSE 15000 43.464 95.555 24.712 53.424 Step= 16000/67350 Loss=0.94745 Time Elapsed=2.37h Time Remaining=7.62h Validation results: Step MAE RMSE iMAE iRMSE 16000 42.144 94.957 23.650 52.560 Best results: Step MAE RMSE iMAE iRMSE 16000 42.144 94.957 23.650 52.560 Step= 17000/67350 Loss=1.37077 Time Elapsed=2.52h Time Remaining=7.47h Validation results: Step MAE RMSE iMAE iRMSE 17000 59.516 110.355 34.302 62.407 Best results: Step MAE RMSE iMAE iRMSE 16000 42.144 94.957 23.650 52.560 Step= 18000/67350 Loss=1.21326 Time Elapsed=2.68h Time Remaining=7.33h Validation results: Step MAE RMSE iMAE iRMSE 18000 43.529 94.813 25.179 52.336 Best results: Step MAE RMSE iMAE iRMSE 16000 42.144 94.957 23.650 52.560 Step= 19000/67350 Loss=0.89081 Time Elapsed=2.83h Time Remaining=7.19h Validation results: Step MAE RMSE iMAE iRMSE 19000 57.932 108.276 32.193 61.657 Best results: Step MAE RMSE iMAE iRMSE 16000 42.144 94.957 23.650 52.560 Step= 20000/67350 Loss=1.18904 Time Elapsed=3.52h Time Remaining=8.33h Validation results: Step MAE RMSE iMAE iRMSE 20000 52.496 97.130 29.876 54.368 Best results: Step MAE RMSE iMAE iRMSE 16000 42.144 94.957 23.650 52.560 Step= 21000/67350 Loss=0.97192 Time Elapsed=4.39h Time Remaining=9.69h Validation results: Step MAE RMSE iMAE iRMSE 21000 40.237 91.283 23.482 50.298 Best results: Step MAE RMSE iMAE iRMSE 21000 40.237 91.283 23.482 50.298 Step= 22000/67350 Loss=1.04763 Time Elapsed=5.13h Time Remaining=10.58h Validation results: Step MAE RMSE iMAE iRMSE 22000 50.142 107.232 28.743 58.937 Best results: Step MAE RMSE iMAE iRMSE 21000 40.237 91.283 23.482 50.298 Step= 23000/67350 Loss=1.23116 Time Elapsed=5.85h Time Remaining=11.28h Validation results: Step MAE RMSE iMAE iRMSE 23000 42.949 92.527 23.344 49.113 Best results: Step MAE RMSE iMAE iRMSE 21000 40.237 91.283 23.482 50.298 Step= 24000/67350 Loss=1.24651 Time Elapsed=6.49h Time Remaining=11.73h Validation results: Step MAE RMSE iMAE iRMSE 24000 52.003 106.014 30.887 61.163 Best results: Step MAE RMSE iMAE iRMSE 21000 40.237 91.283 23.482 50.298 Step= 25000/67350 Loss=1.03224 Time Elapsed=7.03h Time Remaining=11.91h Validation results: Step MAE RMSE iMAE iRMSE 25000 41.103 90.397 22.700 47.765 Best results: Step MAE RMSE iMAE iRMSE 25000 41.103 90.397 22.700 47.765 Training input paths: training/void/void_train_image_1500.txt training/void/void_train_sparse_depth_1500.txt training/void/void_train_intrinsics_1500.txt Validation input paths: testing/void/void_test_image_1500.txt testing/void/void_test_sparse_depth_1500.txt testing/void/void_test_intrinsics_1500.txt testing/void/void_test_ground_truth_1500.txt Input settings: n_batch=8 n_height=480 n_width=640 input_channels_image=3 input_channels_depth=2 normalized_image_range=[0.0, 1.0] outlier_removal_kernel_size=7 outlier_removal_threshold=1.50 Sparse to dense pooling settings: min_pool_sizes_sparse_to_dense_pool=[15, 17, 19] max_pool_sizes_sparse_to_dense_pool=[23, 27] n_convolution_sparse_to_dense_pool=3 n_filter_sparse_to_dense_pool=8 Depth network settings: n_filters_encoder_image=[48, 96, 192, 384, 384] n_filters_encoder_depth=[16, 32, 64, 128, 128] resolutions_backprojection=[0, 1, 2, 3] n_filters_decoder=[256, 128, 128, 64, 12] deconv_type=up min_predict_depth=0.10 max_predict_depth=8.00 Weight settings: n_parameter=6957764 n_parameter_depth=6957764 n_parameter_pose=0 weight_initializer=xavier_normal activation_func=leaky_relu Training settings: n_sample=35917 n_epoch=15 n_step=67350 learning_schedule=[0-44890 : 0.0001, 44890-67335 : 5e-05] Augmentation settings: augmentation_schedule=[0--4489 : 1.0] augmentation_random_crop_type=['horizontal', 'vertical', 'anchored'] augmentation_random_flip_type=['none'] augmentation_random_remove_points=[0.3, 0.6] augmentation_random_noise_type=none augmentation_random_noise_spread=-1.0 Loss function settings: w_color=1.5e-01 w_structure=9.5e-01 w_sparse_depth=2.0e+00 w_smoothness=2.0e+00 w_weight_decay_depth=0.0e+00 w_weight_decay_pose=0.0e+00 Evaluation settings: min_evaluate_depth=0.20 max_evaluate_depth=5.00 Checkpoint settings: checkpoint_path=trained_kbnet/void1500/kbnet_model checkpoint_save_frequency=1000 validation_start_step=5000 Tensorboard settings: event_path=trained_kbnet/void1500/kbnet_model/events log_summary_frequency=1000 n_summary_display=4 Hardware settings: device=cuda n_thread=16 Begin training... Step= 1000/67350 Loss=1.31976 Time Elapsed=0.14h Time Remaining=9.60h Step= 2000/67350 Loss=1.36713 Time Elapsed=0.29h Time Remaining=9.41h Step= 3000/67350 Loss=1.19186 Time Elapsed=0.43h Time Remaining=9.25h Step= 4000/67350 Loss=1.10720 Time Elapsed=0.57h Time Remaining=9.09h Step= 5000/67350 Loss=1.30009 Time Elapsed=0.72h Time Remaining=8.96h Validation results: Step MAE RMSE iMAE iRMSE 5000 83.507 143.069 50.119 87.915 Best results: Step MAE RMSE iMAE iRMSE 5000 83.507 143.069 50.119 87.915 Step= 6000/67350 Loss=0.94178 Time Elapsed=0.87h Time Remaining=8.89h Validation results: Step MAE RMSE iMAE iRMSE 6000 56.986 109.462 32.784 65.296 Best results: Step MAE RMSE iMAE iRMSE 6000 56.986 109.462 32.784 65.296 Step= 7000/67350 Loss=1.19433 Time Elapsed=1.02h Time Remaining=8.79h Validation results: Step MAE RMSE iMAE iRMSE 7000 56.202 109.871 32.635 65.920 Best results: Step MAE RMSE iMAE iRMSE 6000 56.986 109.462 32.784 65.296 Step= 8000/67350 Loss=1.19096 Time Elapsed=1.17h Time Remaining=8.67h Validation results: Step MAE RMSE iMAE iRMSE 8000 50.518 103.431 30.544 62.677 Best results: Step MAE RMSE iMAE iRMSE 8000 50.518 103.431 30.544 62.677 Step= 9000/67350 Loss=1.42468 Time Elapsed=1.32h Time Remaining=8.56h Validation results: Step MAE RMSE iMAE iRMSE 9000 54.272 118.077 31.309 67.710 Best results: Step MAE RMSE iMAE iRMSE 8000 50.518 103.431 30.544 62.677 Step= 10000/67350 Loss=1.02636 Time Elapsed=1.47h Time Remaining=8.43h Validation results: Step MAE RMSE iMAE iRMSE 10000 48.897 100.474 30.132 61.457 Best results: Step MAE RMSE iMAE iRMSE 10000 48.897 100.474 30.132 61.457 Step= 11000/67350 Loss=0.99607 Time Elapsed=1.62h Time Remaining=8.30h Validation results: Step MAE RMSE iMAE iRMSE 11000 47.878 102.048 28.248 59.684 Best results: Step MAE RMSE iMAE iRMSE 11000 47.878 102.048 28.248 59.684 Step= 12000/67350 Loss=1.25286 Time Elapsed=1.77h Time Remaining=8.17h Validation results: Step MAE RMSE iMAE iRMSE 12000 64.796 117.208 36.744 68.819 Best results: Step MAE RMSE iMAE iRMSE 11000 47.878 102.048 28.248 59.684 Step= 13000/67350 Loss=1.14606 Time Elapsed=1.92h Time Remaining=8.04h Validation results: Step MAE RMSE iMAE iRMSE 13000 57.056 112.719 35.438 67.135 Best results: Step MAE RMSE iMAE iRMSE 11000 47.878 102.048 28.248 59.684 Step= 14000/67350 Loss=1.15299 Time Elapsed=2.07h Time Remaining=7.90h Validation results: Step MAE RMSE iMAE iRMSE 14000 41.940 94.975 25.432 55.666 Best results: Step MAE RMSE iMAE iRMSE 14000 41.940 94.975 25.432 55.666 Step= 15000/67350 Loss=1.42613 Time Elapsed=2.22h Time Remaining=7.76h Validation results: Step MAE RMSE iMAE iRMSE 15000 50.547 98.100 28.462 56.330 Best results: Step MAE RMSE iMAE iRMSE 14000 41.940 94.975 25.432 55.666 Step= 16000/67350 Loss=1.15945 Time Elapsed=2.38h Time Remaining=7.62h Validation results: Step MAE RMSE iMAE iRMSE 16000 40.226 93.590 23.643 53.103 Best results: Step MAE RMSE iMAE iRMSE 16000 40.226 93.590 23.643 53.103 Step= 17000/67350 Loss=1.15369 Time Elapsed=2.53h Time Remaining=7.48h Validation results: Step MAE RMSE iMAE iRMSE 17000 46.936 96.496 27.347 55.846 Best results: Step MAE RMSE iMAE iRMSE 16000 40.226 93.590 23.643 53.103 Step= 18000/67350 Loss=1.10486 Time Elapsed=2.68h Time Remaining=7.34h Validation results: Step MAE RMSE iMAE iRMSE 18000 48.935 97.662 25.642 52.489 Best results: Step MAE RMSE iMAE iRMSE 16000 40.226 93.590 23.643 53.103 Step= 19000/67350 Loss=1.19928 Time Elapsed=2.83h Time Remaining=7.20h Validation results: Step MAE RMSE iMAE iRMSE 19000 55.123 102.991 28.569 54.553 Best results: Step MAE RMSE iMAE iRMSE 16000 40.226 93.590 23.643 53.103 Step= 20000/67350 Loss=1.28250 Time Elapsed=2.98h Time Remaining=7.06h Validation results: Step MAE RMSE iMAE iRMSE 20000 52.255 104.068 28.089 57.014 Best results: Step MAE RMSE iMAE iRMSE 16000 40.226 93.590 23.643 53.103 Step= 21000/67350 Loss=1.15070 Time Elapsed=3.13h Time Remaining=6.91h Validation results: Step MAE RMSE iMAE iRMSE 21000 45.847 94.807 26.011 52.601 Best results: Step MAE RMSE iMAE iRMSE 16000 40.226 93.590 23.643 53.103 Step= 22000/67350 Loss=1.11121 Time Elapsed=3.28h Time Remaining=6.77h Validation results: Step MAE RMSE iMAE iRMSE 22000 62.277 106.170 36.172 59.259 Best results: Step MAE RMSE iMAE iRMSE 16000 40.226 93.590 23.643 53.103 Step= 23000/67350 Loss=1.06806 Time Elapsed=3.44h Time Remaining=6.62h Validation results: Step MAE RMSE iMAE iRMSE 23000 39.287 91.200 21.771 49.035 Best results: Step MAE RMSE iMAE iRMSE 23000 39.287 91.200 21.771 49.035 Step= 24000/67350 Loss=1.07366 Time Elapsed=3.59h Time Remaining=6.48h Validation results: Step MAE RMSE iMAE iRMSE 24000 38.680 92.896 22.124 50.923 Best results: Step MAE RMSE iMAE iRMSE 23000 39.287 91.200 21.771 49.035 Step= 25000/67350 Loss=1.24281 Time Elapsed=3.74h Time Remaining=6.33h Validation results: Step MAE RMSE iMAE iRMSE 25000 49.258 96.377 28.800 55.543 Best results: Step MAE RMSE iMAE iRMSE 23000 39.287 91.200 21.771 49.035 Step= 26000/67350 Loss=1.02737 Time Elapsed=3.89h Time Remaining=6.19h Validation results: Step MAE RMSE iMAE iRMSE 26000 61.390 103.956 31.938 53.080 Best results: Step MAE RMSE iMAE iRMSE 23000 39.287 91.200 21.771 49.035 Step= 27000/67350 Loss=1.12811 Time Elapsed=4.04h Time Remaining=6.04h Validation results: Step MAE RMSE iMAE iRMSE 27000 43.465 96.232 24.394 52.952 Best results: Step MAE RMSE iMAE iRMSE 23000 39.287 91.200 21.771 49.035 Step= 28000/67350 Loss=1.42430 Time Elapsed=4.19h Time Remaining=5.89h Validation results: Step MAE RMSE iMAE iRMSE 28000 44.451 93.810 26.406 52.565 Best results: Step MAE RMSE iMAE iRMSE 23000 39.287 91.200 21.771 49.035 Step= 29000/67350 Loss=1.10990 Time Elapsed=4.34h Time Remaining=5.74h Validation results: Step MAE RMSE iMAE iRMSE 29000 37.081 89.913 21.858 50.037 Best results: Step MAE RMSE iMAE iRMSE 23000 39.287 91.200 21.771 49.035 Step= 30000/67350 Loss=0.91196 Time Elapsed=4.49h Time Remaining=5.59h Validation results: Step MAE RMSE iMAE iRMSE 30000 52.768 95.429 28.044 51.903 Best results: Step MAE RMSE iMAE iRMSE 23000 39.287 91.200 21.771 49.035 Step= 31000/67350 Loss=1.16290 Time Elapsed=4.64h Time Remaining=5.44h Validation results: Step MAE RMSE iMAE iRMSE 31000 37.620 90.123 21.817 49.765 Best results: Step MAE RMSE iMAE iRMSE 23000 39.287 91.200 21.771 49.035 Step= 32000/67350 Loss=1.35568 Time Elapsed=4.79h Time Remaining=5.30h Validation results: Step MAE RMSE iMAE iRMSE 32000 39.054 91.514 22.144 49.043 Best results: Step MAE RMSE iMAE iRMSE 23000 39.287 91.200 21.771 49.035 Step= 33000/67350 Loss=1.35563 Time Elapsed=4.95h Time Remaining=5.15h Validation results: Step MAE RMSE iMAE iRMSE 33000 73.064 118.048 45.117 74.119 Best results: Step MAE RMSE iMAE iRMSE 23000 39.287 91.200 21.771 49.035 Step= 34000/67350 Loss=1.10823 Time Elapsed=5.10h Time Remaining=5.00h Validation results: Step MAE RMSE iMAE iRMSE 34000 54.828 101.743 29.063 54.119 Best results: Step MAE RMSE iMAE iRMSE 23000 39.287 91.200 21.771 49.035 Step= 35000/67350 Loss=1.14810 Time Elapsed=5.25h Time Remaining=4.85h Validation results: Step MAE RMSE iMAE iRMSE 35000 58.381 108.522 30.960 56.427 Best results: Step MAE RMSE iMAE iRMSE 23000 39.287 91.200 21.771 49.035 Step= 36000/67350 Loss=1.33982 Time Elapsed=5.40h Time Remaining=4.70h Validation results: Step MAE RMSE iMAE iRMSE 36000 45.489 96.607 28.834 57.877 Best results: Step MAE RMSE iMAE iRMSE 23000 39.287 91.200 21.771 49.035 Step= 37000/67350 Loss=1.10611 Time Elapsed=5.55h Time Remaining=4.55h Validation results: Step MAE RMSE iMAE iRMSE 37000 36.053 87.053 21.757 48.577 Best results: Step MAE RMSE iMAE iRMSE 37000 36.053 87.053 21.757 48.577 Step= 38000/67350 Loss=1.20461 Time Elapsed=5.70h Time Remaining=4.40h Validation results: Step MAE RMSE iMAE iRMSE 38000 41.594 95.467 24.892 53.605 Best results: Step MAE RMSE iMAE iRMSE 37000 36.053 87.053 21.757 48.577 Step= 39000/67350 Loss=1.27347 Time Elapsed=5.85h Time Remaining=4.25h Validation results: Step MAE RMSE iMAE iRMSE 39000 41.069 97.383 23.789 52.479 Best results: Step MAE RMSE iMAE iRMSE 37000 36.053 87.053 21.757 48.577 Step= 40000/67350 Loss=1.12445 Time Elapsed=6.00h Time Remaining=4.10h Validation results: Step MAE RMSE iMAE iRMSE 40000 42.858 91.291 24.559 49.803 Best results: Step MAE RMSE iMAE iRMSE 37000 36.053 87.053 21.757 48.577 Step= 41000/67350 Loss=1.07163 Time Elapsed=6.15h Time Remaining=3.96h Validation results: Step MAE RMSE iMAE iRMSE 41000 88.463 127.677 45.026 67.878 Best results: Step MAE RMSE iMAE iRMSE 37000 36.053 87.053 21.757 48.577 Step= 42000/67350 Loss=0.99703 Time Elapsed=6.31h Time Remaining=3.81h Validation results: Step MAE RMSE iMAE iRMSE 42000 36.534 88.508 20.737 46.541 Best results: Step MAE RMSE iMAE iRMSE 37000 36.053 87.053 21.757 48.577 Step= 43000/67350 Loss=0.91656 Time Elapsed=6.46h Time Remaining=3.66h Validation results: Step MAE RMSE iMAE iRMSE 43000 39.948 97.482 22.532 51.563 Best results: Step MAE RMSE iMAE iRMSE 37000 36.053 87.053 21.757 48.577 Step= 44000/67350 Loss=1.14309 Time Elapsed=6.61h Time Remaining=3.51h Validation results: Step MAE RMSE iMAE iRMSE 44000 72.528 112.965 39.649 63.479 Best results: Step MAE RMSE iMAE iRMSE 37000 36.053 87.053 21.757 48.577 Step= 45000/67350 Loss=1.23675 Time Elapsed=6.85h Time Remaining=3.40h Validation results: Step MAE RMSE iMAE iRMSE 45000 38.373 92.862 22.512 50.501 Best results: Step MAE RMSE iMAE iRMSE 37000 36.053 87.053 21.757 48.577 Step= 46000/67350 Loss=1.14691 Time Elapsed=7.05h Time Remaining=3.27h Validation results: Step MAE RMSE iMAE iRMSE 46000 39.201 90.388 22.828 49.554 Best results: Step MAE RMSE iMAE iRMSE 37000 36.053 87.053 21.757 48.577 Step= 47000/67350 Loss=1.04185 Time Elapsed=7.20h Time Remaining=3.12h Validation results: Step MAE RMSE iMAE iRMSE 47000 40.466 94.780 23.289 51.195 Best results: Step MAE RMSE iMAE iRMSE 37000 36.053 87.053 21.757 48.577 Step= 48000/67350 Loss=1.04116 Time Elapsed=7.35h Time Remaining=2.96h Validation results: Step MAE RMSE iMAE iRMSE 48000 44.548 103.424 24.605 54.154 Best results: Step MAE RMSE iMAE iRMSE 37000 36.053 87.053 21.757 48.577 Step= 49000/67350 Loss=1.06950 Time Elapsed=7.50h Time Remaining=2.81h Validation results: Step MAE RMSE iMAE iRMSE 49000 41.380 97.486 22.488 50.736 Best results: Step MAE RMSE iMAE iRMSE 37000 36.053 87.053 21.757 48.577 Step= 50000/67350 Loss=1.28999 Time Elapsed=7.65h Time Remaining=2.66h Validation results: Step MAE RMSE iMAE iRMSE 50000 38.145 90.552 22.216 49.257 Best results: Step MAE RMSE iMAE iRMSE 37000 36.053 87.053 21.757 48.577 Step= 51000/67350 Loss=1.07065 Time Elapsed=7.80h Time Remaining=2.50h Validation results: Step MAE RMSE iMAE iRMSE 51000 37.850 90.879 21.841 48.832 Best results: Step MAE RMSE iMAE iRMSE 37000 36.053 87.053 21.757 48.577 Step= 52000/67350 Loss=1.12818 Time Elapsed=7.96h Time Remaining=2.35h Validation results: Step MAE RMSE iMAE iRMSE 52000 40.694 97.432 22.711 51.080 Best results: Step MAE RMSE iMAE iRMSE 37000 36.053 87.053 21.757 48.577 Step= 53000/67350 Loss=1.13346 Time Elapsed=8.12h Time Remaining=2.20h Validation results: Step MAE RMSE iMAE iRMSE 53000 39.561 94.193 22.963 51.218 Best results: Step MAE RMSE iMAE iRMSE 37000 36.053 87.053 21.757 48.577 Step= 54000/67350 Loss=1.21016 Time Elapsed=8.37h Time Remaining=2.07h Validation results: Step MAE RMSE iMAE iRMSE 54000 41.995 98.750 24.078 53.133 Best results: Step MAE RMSE iMAE iRMSE 37000 36.053 87.053 21.757 48.577 Step= 55000/67350 Loss=0.99507 Time Elapsed=8.59h Time Remaining=1.93h Validation results: Step MAE RMSE iMAE iRMSE 55000 37.000 91.791 20.797 48.230 Best results: Step MAE RMSE iMAE iRMSE 37000 36.053 87.053 21.757 48.577 Step= 56000/67350 Loss=1.04803 Time Elapsed=8.74h Time Remaining=1.77h Validation results: Step MAE RMSE iMAE iRMSE 56000 39.830 95.706 22.984 50.606 Best results: Step MAE RMSE iMAE iRMSE 37000 36.053 87.053 21.757 48.577 Step= 57000/67350 Loss=1.11132 Time Elapsed=8.89h Time Remaining=1.61h Validation results: Step MAE RMSE iMAE iRMSE 57000 38.975 90.080 21.806 47.744 Best results: Step MAE RMSE iMAE iRMSE 37000 36.053 87.053 21.757 48.577 Step= 58000/67350 Loss=0.82085 Time Elapsed=9.04h Time Remaining=1.46h Validation results: Step MAE RMSE iMAE iRMSE 58000 37.845 91.695 21.825 49.654 Best results: Step MAE RMSE iMAE iRMSE 37000 36.053 87.053 21.757 48.577 Step= 59000/67350 Loss=1.05060 Time Elapsed=9.19h Time Remaining=1.30h Validation results: Step MAE RMSE iMAE iRMSE 59000 37.128 88.384 21.165 46.696 Best results: Step MAE RMSE iMAE iRMSE 37000 36.053 87.053 21.757 48.577 Step= 60000/67350 Loss=1.22494 Time Elapsed=9.34h Time Remaining=1.14h Validation results: Step MAE RMSE iMAE iRMSE 60000 38.559 93.583 21.597 48.988 Best results: Step MAE RMSE iMAE iRMSE 37000 36.053 87.053 21.757 48.577 Step= 61000/67350 Loss=0.92884 Time Elapsed=9.49h Time Remaining=0.99h Validation results: Step MAE RMSE iMAE iRMSE 61000 40.313 93.850 22.505 49.408 Best results: Step MAE RMSE iMAE iRMSE 37000 36.053 87.053 21.757 48.577 Step= 62000/67350 Loss=1.16037 Time Elapsed=9.64h Time Remaining=0.83h Validation results: Step MAE RMSE iMAE iRMSE 62000 36.436 90.489 20.463 47.454 Best results: Step MAE RMSE iMAE iRMSE 37000 36.053 87.053 21.757 48.577 Step= 63000/67350 Loss=1.12444 Time Elapsed=9.79h Time Remaining=0.68h Validation results: Step MAE RMSE iMAE iRMSE 63000 35.596 89.272 20.327 46.497 Best results: Step MAE RMSE iMAE iRMSE 63000 35.596 89.272 20.327 46.497 Step= 64000/67350 Loss=1.04234 Time Elapsed=9.94h Time Remaining=0.52h Validation results: Step MAE RMSE iMAE iRMSE 64000 37.002 90.253 21.012 47.850 Best results: Step MAE RMSE iMAE iRMSE 63000 35.596 89.272 20.327 46.497 Step= 65000/67350 Loss=0.99864 Time Elapsed=10.09h Time Remaining=0.36h Validation results: Step MAE RMSE iMAE iRMSE 65000 40.389 93.391 23.443 51.245 Best results: Step MAE RMSE iMAE iRMSE 63000 35.596 89.272 20.327 46.497 Step= 66000/67350 Loss=1.10018 Time Elapsed=10.24h Time Remaining=0.21h Validation results: Step MAE RMSE iMAE iRMSE 66000 39.457 95.390 22.597 49.893 Best results: Step MAE RMSE iMAE iRMSE 63000 35.596 89.272 20.327 46.497 Step= 67000/67350 Loss=1.23952 Time Elapsed=10.39h Time Remaining=0.05h Validation results: Step MAE RMSE iMAE iRMSE 67000 36.761 91.658 20.631 47.741 Best results: Step MAE RMSE iMAE iRMSE 63000 35.596 89.272 20.327 46.497