Skip to content
Permalink
Branch: master
Find file Copy path
Find file Copy path
Fetching contributors…
Cannot retrieve contributors at this time
1206 lines (1205 sloc) 92.8 KB
dataset.py:read_dataset: reading data from data/snli_1.0/train.txt
dataset.py:read_dataset: read 549367 examples
dataset.py:read_dataset: reading data from data/snli_1.0/dev.txt
dataset.py:read_dataset: read 9842 examples
main.py:train_model: {'gpu_id': 1, 'train_file': 'data/snli_1.0/train.txt', 'test_file': 'data/snli_1.0/dev.txt', 'max_num_examples': -1, 'batch_size': 32, 'print_interval': 5000, 'mode': 'train', 'lr': 0.025, 'epochs': 300, 'embedding': 'glove', 'embedding_source': 'glove.840B.300d', 'embedding_size': 300, 'hidden_size': 200, 'output_dir': 'output/snli-intra', 'model_dir': './output', 'seed': 0, 'dropout': 0.2, 'weight_decay': 1e-05, 'intra_attention': True}
main.py:train_model: [Epoch 0 Batch 5000/17173] loss=1.0521, acc=0.4061
main.py:train_model: [Epoch 0 Batch 10000/17173] loss=0.8471, acc=0.6180
main.py:train_model: [Epoch 0 Batch 15000/17173] loss=0.7841, acc=0.6586
main.py:train_model: [Epoch 0] valid loss=0.6971, valid acc=0.7128, best valid acc=0.7128
main.py:train_model: [Epoch 1 Batch 5000/17173] loss=0.7292, acc=0.6912
main.py:train_model: [Epoch 1 Batch 10000/17173] loss=0.7067, acc=0.7016
main.py:train_model: [Epoch 1 Batch 15000/17173] loss=0.6899, acc=0.7106
main.py:train_model: [Epoch 1] valid loss=0.6229, valid acc=0.7467, best valid acc=0.7467
main.py:train_model: [Epoch 2 Batch 5000/17173] loss=0.6706, acc=0.7219
main.py:train_model: [Epoch 2 Batch 10000/17173] loss=0.6614, acc=0.7274
main.py:train_model: [Epoch 2 Batch 15000/17173] loss=0.6552, acc=0.7298
main.py:train_model: [Epoch 2] valid loss=0.5881, valid acc=0.7628, best valid acc=0.7628
main.py:train_model: [Epoch 3 Batch 5000/17173] loss=0.6412, acc=0.7362
main.py:train_model: [Epoch 3 Batch 10000/17173] loss=0.6338, acc=0.7404
main.py:train_model: [Epoch 3 Batch 15000/17173] loss=0.6290, acc=0.7421
main.py:train_model: [Epoch 3] valid loss=0.5572, valid acc=0.7772, best valid acc=0.7772
main.py:train_model: [Epoch 4 Batch 5000/17173] loss=0.6169, acc=0.7488
main.py:train_model: [Epoch 4 Batch 10000/17173] loss=0.6116, acc=0.7509
main.py:train_model: [Epoch 4 Batch 15000/17173] loss=0.6082, acc=0.7520
main.py:train_model: [Epoch 4] valid loss=0.5378, valid acc=0.7847, best valid acc=0.7847
main.py:train_model: [Epoch 5 Batch 5000/17173] loss=0.5993, acc=0.7567
main.py:train_model: [Epoch 5 Batch 10000/17173] loss=0.5980, acc=0.7559
main.py:train_model: [Epoch 5 Batch 15000/17173] loss=0.5884, acc=0.7616
main.py:train_model: [Epoch 5] valid loss=0.5385, valid acc=0.7868, best valid acc=0.7868
main.py:train_model: [Epoch 6 Batch 5000/17173] loss=0.5830, acc=0.7643
main.py:train_model: [Epoch 6 Batch 10000/17173] loss=0.5816, acc=0.7641
main.py:train_model: [Epoch 6 Batch 15000/17173] loss=0.5827, acc=0.7646
main.py:train_model: [Epoch 6] valid loss=0.5201, valid acc=0.7941, best valid acc=0.7941
main.py:train_model: [Epoch 7 Batch 5000/17173] loss=0.5738, acc=0.7674
main.py:train_model: [Epoch 7 Batch 10000/17173] loss=0.5755, acc=0.7685
main.py:train_model: [Epoch 7 Batch 15000/17173] loss=0.5719, acc=0.7699
main.py:train_model: [Epoch 7] valid loss=0.5183, valid acc=0.7964, best valid acc=0.7964
main.py:train_model: [Epoch 8 Batch 5000/17173] loss=0.5675, acc=0.7714
main.py:train_model: [Epoch 8 Batch 10000/17173] loss=0.5671, acc=0.7715
main.py:train_model: [Epoch 8 Batch 15000/17173] loss=0.5665, acc=0.7711
main.py:train_model: [Epoch 8] valid loss=0.5071, valid acc=0.7977, best valid acc=0.7977
main.py:train_model: [Epoch 9 Batch 5000/17173] loss=0.5627, acc=0.7725
main.py:train_model: [Epoch 9 Batch 10000/17173] loss=0.5577, acc=0.7771
main.py:train_model: [Epoch 9 Batch 15000/17173] loss=0.5584, acc=0.7749
main.py:train_model: [Epoch 9] valid loss=0.5011, valid acc=0.8041, best valid acc=0.8041
main.py:train_model: [Epoch 10 Batch 5000/17173] loss=0.5553, acc=0.7759
main.py:train_model: [Epoch 10 Batch 10000/17173] loss=0.5535, acc=0.7770
main.py:train_model: [Epoch 10 Batch 15000/17173] loss=0.5541, acc=0.7774
main.py:train_model: [Epoch 10] valid loss=0.4922, valid acc=0.8060, best valid acc=0.8060
main.py:train_model: [Epoch 11 Batch 5000/17173] loss=0.5495, acc=0.7802
main.py:train_model: [Epoch 11 Batch 10000/17173] loss=0.5526, acc=0.7779
main.py:train_model: [Epoch 11 Batch 15000/17173] loss=0.5464, acc=0.7822
main.py:train_model: [Epoch 11] valid loss=0.4925, valid acc=0.8075, best valid acc=0.8075
main.py:train_model: [Epoch 12 Batch 5000/17173] loss=0.5435, acc=0.7807
main.py:train_model: [Epoch 12 Batch 10000/17173] loss=0.5440, acc=0.7819
main.py:train_model: [Epoch 12 Batch 15000/17173] loss=0.5437, acc=0.7829
main.py:train_model: [Epoch 12] valid loss=0.4847, valid acc=0.8081, best valid acc=0.8081
main.py:train_model: [Epoch 13 Batch 5000/17173] loss=0.5379, acc=0.7845
main.py:train_model: [Epoch 13 Batch 10000/17173] loss=0.5399, acc=0.7838
main.py:train_model: [Epoch 13 Batch 15000/17173] loss=0.5377, acc=0.7847
main.py:train_model: [Epoch 13] valid loss=0.4828, valid acc=0.8119, best valid acc=0.8119
main.py:train_model: [Epoch 14 Batch 5000/17173] loss=0.5350, acc=0.7850
main.py:train_model: [Epoch 14 Batch 10000/17173] loss=0.5357, acc=0.7862
main.py:train_model: [Epoch 14 Batch 15000/17173] loss=0.5356, acc=0.7861
main.py:train_model: [Epoch 14] valid loss=0.4832, valid acc=0.8117, best valid acc=0.8119
main.py:train_model: [Epoch 15 Batch 5000/17173] loss=0.5321, acc=0.7875
main.py:train_model: [Epoch 15 Batch 10000/17173] loss=0.5292, acc=0.7890
main.py:train_model: [Epoch 15 Batch 15000/17173] loss=0.5314, acc=0.7882
main.py:train_model: [Epoch 15] valid loss=0.4774, valid acc=0.8142, best valid acc=0.8142
main.py:train_model: [Epoch 16 Batch 5000/17173] loss=0.5263, acc=0.7907
main.py:train_model: [Epoch 16 Batch 10000/17173] loss=0.5303, acc=0.7882
main.py:train_model: [Epoch 16 Batch 15000/17173] loss=0.5261, acc=0.7910
main.py:train_model: [Epoch 16] valid loss=0.4763, valid acc=0.8144, best valid acc=0.8144
main.py:train_model: [Epoch 17 Batch 5000/17173] loss=0.5242, acc=0.7919
main.py:train_model: [Epoch 17 Batch 10000/17173] loss=0.5212, acc=0.7925
main.py:train_model: [Epoch 17 Batch 15000/17173] loss=0.5239, acc=0.7913
main.py:train_model: [Epoch 17] valid loss=0.4774, valid acc=0.8161, best valid acc=0.8161
main.py:train_model: [Epoch 18 Batch 5000/17173] loss=0.5194, acc=0.7940
main.py:train_model: [Epoch 18 Batch 10000/17173] loss=0.5220, acc=0.7912
main.py:train_model: [Epoch 18 Batch 15000/17173] loss=0.5211, acc=0.7934
main.py:train_model: [Epoch 18] valid loss=0.4691, valid acc=0.8160, best valid acc=0.8161
main.py:train_model: [Epoch 19 Batch 5000/17173] loss=0.5196, acc=0.7933
main.py:train_model: [Epoch 19 Batch 10000/17173] loss=0.5162, acc=0.7957
main.py:train_model: [Epoch 19 Batch 15000/17173] loss=0.5182, acc=0.7935
main.py:train_model: [Epoch 19] valid loss=0.4639, valid acc=0.8176, best valid acc=0.8176
main.py:train_model: [Epoch 20 Batch 5000/17173] loss=0.5147, acc=0.7965
main.py:train_model: [Epoch 20 Batch 10000/17173] loss=0.5138, acc=0.7958
main.py:train_model: [Epoch 20 Batch 15000/17173] loss=0.5148, acc=0.7965
main.py:train_model: [Epoch 20] valid loss=0.4631, valid acc=0.8186, best valid acc=0.8186
main.py:train_model: [Epoch 21 Batch 5000/17173] loss=0.5102, acc=0.7978
main.py:train_model: [Epoch 21 Batch 10000/17173] loss=0.5124, acc=0.7965
main.py:train_model: [Epoch 21 Batch 15000/17173] loss=0.5115, acc=0.7968
main.py:train_model: [Epoch 21] valid loss=0.4590, valid acc=0.8183, best valid acc=0.8186
main.py:train_model: [Epoch 22 Batch 5000/17173] loss=0.5078, acc=0.7983
main.py:train_model: [Epoch 22 Batch 10000/17173] loss=0.5086, acc=0.7985
main.py:train_model: [Epoch 22 Batch 15000/17173] loss=0.5092, acc=0.7978
main.py:train_model: [Epoch 22] valid loss=0.4575, valid acc=0.8183, best valid acc=0.8186
main.py:train_model: [Epoch 23 Batch 5000/17173] loss=0.5063, acc=0.7986
main.py:train_model: [Epoch 23 Batch 10000/17173] loss=0.5063, acc=0.7995
main.py:train_model: [Epoch 23 Batch 15000/17173] loss=0.5084, acc=0.7986
main.py:train_model: [Epoch 23] valid loss=0.4517, valid acc=0.8214, best valid acc=0.8214
main.py:train_model: [Epoch 24 Batch 5000/17173] loss=0.5032, acc=0.8002
main.py:train_model: [Epoch 24 Batch 10000/17173] loss=0.5058, acc=0.7992
main.py:train_model: [Epoch 24 Batch 15000/17173] loss=0.5033, acc=0.8005
main.py:train_model: [Epoch 24] valid loss=0.4535, valid acc=0.8221, best valid acc=0.8221
main.py:train_model: [Epoch 25 Batch 5000/17173] loss=0.5024, acc=0.8010
main.py:train_model: [Epoch 25 Batch 10000/17173] loss=0.5017, acc=0.8024
main.py:train_model: [Epoch 25 Batch 15000/17173] loss=0.5018, acc=0.8015
main.py:train_model: [Epoch 25] valid loss=0.4545, valid acc=0.8225, best valid acc=0.8225
main.py:train_model: [Epoch 26 Batch 5000/17173] loss=0.5028, acc=0.8006
main.py:train_model: [Epoch 26 Batch 10000/17173] loss=0.4966, acc=0.8045
main.py:train_model: [Epoch 26 Batch 15000/17173] loss=0.4992, acc=0.8027
main.py:train_model: [Epoch 26] valid loss=0.4502, valid acc=0.8250, best valid acc=0.8250
main.py:train_model: [Epoch 27 Batch 5000/17173] loss=0.4968, acc=0.8036
main.py:train_model: [Epoch 27 Batch 10000/17173] loss=0.4978, acc=0.8031
main.py:train_model: [Epoch 27 Batch 15000/17173] loss=0.5002, acc=0.8021
main.py:train_model: [Epoch 27] valid loss=0.4494, valid acc=0.8244, best valid acc=0.8250
main.py:train_model: [Epoch 28 Batch 5000/17173] loss=0.4934, acc=0.8061
main.py:train_model: [Epoch 28 Batch 10000/17173] loss=0.4946, acc=0.8044
main.py:train_model: [Epoch 28 Batch 15000/17173] loss=0.5002, acc=0.8023
main.py:train_model: [Epoch 28] valid loss=0.4496, valid acc=0.8251, best valid acc=0.8251
main.py:train_model: [Epoch 29 Batch 5000/17173] loss=0.4961, acc=0.8046
main.py:train_model: [Epoch 29 Batch 10000/17173] loss=0.4960, acc=0.8048
main.py:train_model: [Epoch 29 Batch 15000/17173] loss=0.4937, acc=0.8054
main.py:train_model: [Epoch 29] valid loss=0.4504, valid acc=0.8264, best valid acc=0.8264
main.py:train_model: [Epoch 30 Batch 5000/17173] loss=0.4927, acc=0.8062
main.py:train_model: [Epoch 30 Batch 10000/17173] loss=0.4941, acc=0.8039
main.py:train_model: [Epoch 30 Batch 15000/17173] loss=0.4950, acc=0.8041
main.py:train_model: [Epoch 30] valid loss=0.4476, valid acc=0.8260, best valid acc=0.8264
main.py:train_model: [Epoch 31 Batch 5000/17173] loss=0.4886, acc=0.8080
main.py:train_model: [Epoch 31 Batch 10000/17173] loss=0.4941, acc=0.8042
main.py:train_model: [Epoch 31 Batch 15000/17173] loss=0.4933, acc=0.8058
main.py:train_model: [Epoch 31] valid loss=0.4468, valid acc=0.8279, best valid acc=0.8279
main.py:train_model: [Epoch 32 Batch 5000/17173] loss=0.4872, acc=0.8079
main.py:train_model: [Epoch 32 Batch 10000/17173] loss=0.4902, acc=0.8072
main.py:train_model: [Epoch 32 Batch 15000/17173] loss=0.4881, acc=0.8079
main.py:train_model: [Epoch 32] valid loss=0.4396, valid acc=0.8290, best valid acc=0.8290
main.py:train_model: [Epoch 33 Batch 5000/17173] loss=0.4866, acc=0.8089
main.py:train_model: [Epoch 33 Batch 10000/17173] loss=0.4908, acc=0.8062
main.py:train_model: [Epoch 33 Batch 15000/17173] loss=0.4856, acc=0.8097
main.py:train_model: [Epoch 33] valid loss=0.4433, valid acc=0.8289, best valid acc=0.8290
main.py:train_model: [Epoch 34 Batch 5000/17173] loss=0.4865, acc=0.8095
main.py:train_model: [Epoch 34 Batch 10000/17173] loss=0.4870, acc=0.8093
main.py:train_model: [Epoch 34 Batch 15000/17173] loss=0.4866, acc=0.8089
main.py:train_model: [Epoch 34] valid loss=0.4442, valid acc=0.8286, best valid acc=0.8290
main.py:train_model: [Epoch 35 Batch 5000/17173] loss=0.4844, acc=0.8100
main.py:train_model: [Epoch 35 Batch 10000/17173] loss=0.4864, acc=0.8090
main.py:train_model: [Epoch 35 Batch 15000/17173] loss=0.4869, acc=0.8088
main.py:train_model: [Epoch 35] valid loss=0.4345, valid acc=0.8313, best valid acc=0.8313
main.py:train_model: [Epoch 36 Batch 5000/17173] loss=0.4858, acc=0.8086
main.py:train_model: [Epoch 36 Batch 10000/17173] loss=0.4836, acc=0.8097
main.py:train_model: [Epoch 36 Batch 15000/17173] loss=0.4824, acc=0.8103
main.py:train_model: [Epoch 36] valid loss=0.4407, valid acc=0.8294, best valid acc=0.8313
main.py:train_model: [Epoch 37 Batch 5000/17173] loss=0.4802, acc=0.8124
main.py:train_model: [Epoch 37 Batch 10000/17173] loss=0.4822, acc=0.8100
main.py:train_model: [Epoch 37 Batch 15000/17173] loss=0.4824, acc=0.8100
main.py:train_model: [Epoch 37] valid loss=0.4391, valid acc=0.8291, best valid acc=0.8313
main.py:train_model: [Epoch 38 Batch 5000/17173] loss=0.4817, acc=0.8112
main.py:train_model: [Epoch 38 Batch 10000/17173] loss=0.4811, acc=0.8109
main.py:train_model: [Epoch 38 Batch 15000/17173] loss=0.4810, acc=0.8104
main.py:train_model: [Epoch 38] valid loss=0.4360, valid acc=0.8308, best valid acc=0.8313
main.py:train_model: [Epoch 39 Batch 5000/17173] loss=0.4803, acc=0.8105
main.py:train_model: [Epoch 39 Batch 10000/17173] loss=0.4803, acc=0.8103
main.py:train_model: [Epoch 39 Batch 15000/17173] loss=0.4801, acc=0.8121
main.py:train_model: [Epoch 39] valid loss=0.4359, valid acc=0.8308, best valid acc=0.8313
main.py:train_model: [Epoch 40 Batch 5000/17173] loss=0.4792, acc=0.8117
main.py:train_model: [Epoch 40 Batch 10000/17173] loss=0.4776, acc=0.8119
main.py:train_model: [Epoch 40 Batch 15000/17173] loss=0.4790, acc=0.8121
main.py:train_model: [Epoch 40] valid loss=0.4354, valid acc=0.8315, best valid acc=0.8315
main.py:train_model: [Epoch 41 Batch 5000/17173] loss=0.4768, acc=0.8124
main.py:train_model: [Epoch 41 Batch 10000/17173] loss=0.4768, acc=0.8127
main.py:train_model: [Epoch 41 Batch 15000/17173] loss=0.4783, acc=0.8126
main.py:train_model: [Epoch 41] valid loss=0.4356, valid acc=0.8311, best valid acc=0.8315
main.py:train_model: [Epoch 42 Batch 5000/17173] loss=0.4749, acc=0.8135
main.py:train_model: [Epoch 42 Batch 10000/17173] loss=0.4782, acc=0.8125
main.py:train_model: [Epoch 42 Batch 15000/17173] loss=0.4767, acc=0.8129
main.py:train_model: [Epoch 42] valid loss=0.4339, valid acc=0.8313, best valid acc=0.8315
main.py:train_model: [Epoch 43 Batch 5000/17173] loss=0.4726, acc=0.8151
main.py:train_model: [Epoch 43 Batch 10000/17173] loss=0.4780, acc=0.8124
main.py:train_model: [Epoch 43 Batch 15000/17173] loss=0.4777, acc=0.8125
main.py:train_model: [Epoch 43] valid loss=0.4342, valid acc=0.8301, best valid acc=0.8315
main.py:train_model: [Epoch 44 Batch 5000/17173] loss=0.4743, acc=0.8140
main.py:train_model: [Epoch 44 Batch 10000/17173] loss=0.4754, acc=0.8133
main.py:train_model: [Epoch 44 Batch 15000/17173] loss=0.4747, acc=0.8140
main.py:train_model: [Epoch 44] valid loss=0.4328, valid acc=0.8333, best valid acc=0.8333
main.py:train_model: [Epoch 45 Batch 5000/17173] loss=0.4731, acc=0.8143
main.py:train_model: [Epoch 45 Batch 10000/17173] loss=0.4752, acc=0.8125
main.py:train_model: [Epoch 45 Batch 15000/17173] loss=0.4741, acc=0.8146
main.py:train_model: [Epoch 45] valid loss=0.4281, valid acc=0.8348, best valid acc=0.8348
main.py:train_model: [Epoch 46 Batch 5000/17173] loss=0.4736, acc=0.8144
main.py:train_model: [Epoch 46 Batch 10000/17173] loss=0.4755, acc=0.8145
main.py:train_model: [Epoch 46 Batch 15000/17173] loss=0.4708, acc=0.8163
main.py:train_model: [Epoch 46] valid loss=0.4265, valid acc=0.8348, best valid acc=0.8348
main.py:train_model: [Epoch 47 Batch 5000/17173] loss=0.4730, acc=0.8145
main.py:train_model: [Epoch 47 Batch 10000/17173] loss=0.4718, acc=0.8160
main.py:train_model: [Epoch 47 Batch 15000/17173] loss=0.4715, acc=0.8151
main.py:train_model: [Epoch 47] valid loss=0.4336, valid acc=0.8325, best valid acc=0.8348
main.py:train_model: [Epoch 48 Batch 5000/17173] loss=0.4700, acc=0.8163
main.py:train_model: [Epoch 48 Batch 10000/17173] loss=0.4726, acc=0.8158
main.py:train_model: [Epoch 48 Batch 15000/17173] loss=0.4726, acc=0.8158
main.py:train_model: [Epoch 48] valid loss=0.4241, valid acc=0.8343, best valid acc=0.8348
main.py:train_model: [Epoch 49 Batch 5000/17173] loss=0.4680, acc=0.8168
main.py:train_model: [Epoch 49 Batch 10000/17173] loss=0.4681, acc=0.8164
main.py:train_model: [Epoch 49 Batch 15000/17173] loss=0.4704, acc=0.8162
main.py:train_model: [Epoch 49] valid loss=0.4280, valid acc=0.8351, best valid acc=0.8351
main.py:train_model: [Epoch 50 Batch 5000/17173] loss=0.4694, acc=0.8158
main.py:train_model: [Epoch 50 Batch 10000/17173] loss=0.4692, acc=0.8167
main.py:train_model: [Epoch 50 Batch 15000/17173] loss=0.4698, acc=0.8166
main.py:train_model: [Epoch 50] valid loss=0.4281, valid acc=0.8345, best valid acc=0.8351
main.py:train_model: [Epoch 51 Batch 5000/17173] loss=0.4673, acc=0.8169
main.py:train_model: [Epoch 51 Batch 10000/17173] loss=0.4696, acc=0.8157
main.py:train_model: [Epoch 51 Batch 15000/17173] loss=0.4689, acc=0.8158
main.py:train_model: [Epoch 51] valid loss=0.4279, valid acc=0.8371, best valid acc=0.8371
main.py:train_model: [Epoch 52 Batch 5000/17173] loss=0.4678, acc=0.8172
main.py:train_model: [Epoch 52 Batch 10000/17173] loss=0.4683, acc=0.8171
main.py:train_model: [Epoch 52 Batch 15000/17173] loss=0.4672, acc=0.8176
main.py:train_model: [Epoch 52] valid loss=0.4226, valid acc=0.8351, best valid acc=0.8371
main.py:train_model: [Epoch 53 Batch 5000/17173] loss=0.4689, acc=0.8168
main.py:train_model: [Epoch 53 Batch 10000/17173] loss=0.4624, acc=0.8197
main.py:train_model: [Epoch 53 Batch 15000/17173] loss=0.4689, acc=0.8172
main.py:train_model: [Epoch 53] valid loss=0.4223, valid acc=0.8363, best valid acc=0.8371
main.py:train_model: [Epoch 54 Batch 5000/17173] loss=0.4652, acc=0.8181
main.py:train_model: [Epoch 54 Batch 10000/17173] loss=0.4660, acc=0.8179
main.py:train_model: [Epoch 54 Batch 15000/17173] loss=0.4654, acc=0.8180
main.py:train_model: [Epoch 54] valid loss=0.4206, valid acc=0.8379, best valid acc=0.8379
main.py:train_model: [Epoch 55 Batch 5000/17173] loss=0.4625, acc=0.8188
main.py:train_model: [Epoch 55 Batch 10000/17173] loss=0.4646, acc=0.8179
main.py:train_model: [Epoch 55 Batch 15000/17173] loss=0.4636, acc=0.8184
main.py:train_model: [Epoch 55] valid loss=0.4181, valid acc=0.8381, best valid acc=0.8381
main.py:train_model: [Epoch 56 Batch 5000/17173] loss=0.4635, acc=0.8194
main.py:train_model: [Epoch 56 Batch 10000/17173] loss=0.4680, acc=0.8173
main.py:train_model: [Epoch 56 Batch 15000/17173] loss=0.4630, acc=0.8187
main.py:train_model: [Epoch 56] valid loss=0.4215, valid acc=0.8385, best valid acc=0.8385
main.py:train_model: [Epoch 57 Batch 5000/17173] loss=0.4616, acc=0.8202
main.py:train_model: [Epoch 57 Batch 10000/17173] loss=0.4630, acc=0.8179
main.py:train_model: [Epoch 57 Batch 15000/17173] loss=0.4636, acc=0.8189
main.py:train_model: [Epoch 57] valid loss=0.4216, valid acc=0.8367, best valid acc=0.8385
main.py:train_model: [Epoch 58 Batch 5000/17173] loss=0.4625, acc=0.8180
main.py:train_model: [Epoch 58 Batch 10000/17173] loss=0.4628, acc=0.8190
main.py:train_model: [Epoch 58 Batch 15000/17173] loss=0.4608, acc=0.8205
main.py:train_model: [Epoch 58] valid loss=0.4227, valid acc=0.8382, best valid acc=0.8385
main.py:train_model: [Epoch 59 Batch 5000/17173] loss=0.4608, acc=0.8207
main.py:train_model: [Epoch 59 Batch 10000/17173] loss=0.4612, acc=0.8191
main.py:train_model: [Epoch 59 Batch 15000/17173] loss=0.4611, acc=0.8197
main.py:train_model: [Epoch 59] valid loss=0.4215, valid acc=0.8371, best valid acc=0.8385
main.py:train_model: [Epoch 60 Batch 5000/17173] loss=0.4606, acc=0.8199
main.py:train_model: [Epoch 60 Batch 10000/17173] loss=0.4615, acc=0.8197
main.py:train_model: [Epoch 60 Batch 15000/17173] loss=0.4603, acc=0.8192
main.py:train_model: [Epoch 60] valid loss=0.4193, valid acc=0.8384, best valid acc=0.8385
main.py:train_model: [Epoch 61 Batch 5000/17173] loss=0.4614, acc=0.8185
main.py:train_model: [Epoch 61 Batch 10000/17173] loss=0.4597, acc=0.8203
main.py:train_model: [Epoch 61 Batch 15000/17173] loss=0.4614, acc=0.8200
main.py:train_model: [Epoch 61] valid loss=0.4145, valid acc=0.8393, best valid acc=0.8393
main.py:train_model: [Epoch 62 Batch 5000/17173] loss=0.4604, acc=0.8200
main.py:train_model: [Epoch 62 Batch 10000/17173] loss=0.4575, acc=0.8209
main.py:train_model: [Epoch 62 Batch 15000/17173] loss=0.4606, acc=0.8207
main.py:train_model: [Epoch 62] valid loss=0.4198, valid acc=0.8395, best valid acc=0.8395
main.py:train_model: [Epoch 63 Batch 5000/17173] loss=0.4584, acc=0.8205
main.py:train_model: [Epoch 63 Batch 10000/17173] loss=0.4597, acc=0.8203
main.py:train_model: [Epoch 63 Batch 15000/17173] loss=0.4583, acc=0.8221
main.py:train_model: [Epoch 63] valid loss=0.4225, valid acc=0.8378, best valid acc=0.8395
main.py:train_model: [Epoch 64 Batch 5000/17173] loss=0.4585, acc=0.8221
main.py:train_model: [Epoch 64 Batch 10000/17173] loss=0.4573, acc=0.8213
main.py:train_model: [Epoch 64 Batch 15000/17173] loss=0.4575, acc=0.8216
main.py:train_model: [Epoch 64] valid loss=0.4232, valid acc=0.8385, best valid acc=0.8395
main.py:train_model: [Epoch 65 Batch 5000/17173] loss=0.4599, acc=0.8204
main.py:train_model: [Epoch 65 Batch 10000/17173] loss=0.4589, acc=0.8208
main.py:train_model: [Epoch 65 Batch 15000/17173] loss=0.4566, acc=0.8217
main.py:train_model: [Epoch 65] valid loss=0.4190, valid acc=0.8414, best valid acc=0.8414
main.py:train_model: [Epoch 66 Batch 5000/17173] loss=0.4542, acc=0.8230
main.py:train_model: [Epoch 66 Batch 10000/17173] loss=0.4605, acc=0.8203
main.py:train_model: [Epoch 66 Batch 15000/17173] loss=0.4566, acc=0.8223
main.py:train_model: [Epoch 66] valid loss=0.4172, valid acc=0.8403, best valid acc=0.8414
main.py:train_model: [Epoch 67 Batch 5000/17173] loss=0.4552, acc=0.8224
main.py:train_model: [Epoch 67 Batch 10000/17173] loss=0.4585, acc=0.8216
main.py:train_model: [Epoch 67 Batch 15000/17173] loss=0.4554, acc=0.8223
main.py:train_model: [Epoch 67] valid loss=0.4165, valid acc=0.8405, best valid acc=0.8414
main.py:train_model: [Epoch 68 Batch 5000/17173] loss=0.4565, acc=0.8215
main.py:train_model: [Epoch 68 Batch 10000/17173] loss=0.4565, acc=0.8214
main.py:train_model: [Epoch 68 Batch 15000/17173] loss=0.4546, acc=0.8236
main.py:train_model: [Epoch 68] valid loss=0.4200, valid acc=0.8388, best valid acc=0.8414
main.py:train_model: [Epoch 69 Batch 5000/17173] loss=0.4555, acc=0.8223
main.py:train_model: [Epoch 69 Batch 10000/17173] loss=0.4538, acc=0.8232
main.py:train_model: [Epoch 69 Batch 15000/17173] loss=0.4563, acc=0.8222
main.py:train_model: [Epoch 69] valid loss=0.4144, valid acc=0.8401, best valid acc=0.8414
main.py:train_model: [Epoch 70 Batch 5000/17173] loss=0.4539, acc=0.8231
main.py:train_model: [Epoch 70 Batch 10000/17173] loss=0.4543, acc=0.8222
main.py:train_model: [Epoch 70 Batch 15000/17173] loss=0.4542, acc=0.8232
main.py:train_model: [Epoch 70] valid loss=0.4167, valid acc=0.8416, best valid acc=0.8416
main.py:train_model: [Epoch 71 Batch 5000/17173] loss=0.4537, acc=0.8232
main.py:train_model: [Epoch 71 Batch 10000/17173] loss=0.4539, acc=0.8228
main.py:train_model: [Epoch 71 Batch 15000/17173] loss=0.4545, acc=0.8229
main.py:train_model: [Epoch 71] valid loss=0.4178, valid acc=0.8416, best valid acc=0.8416
main.py:train_model: [Epoch 72 Batch 5000/17173] loss=0.4520, acc=0.8233
main.py:train_model: [Epoch 72 Batch 10000/17173] loss=0.4576, acc=0.8216
main.py:train_model: [Epoch 72 Batch 15000/17173] loss=0.4497, acc=0.8243
main.py:train_model: [Epoch 72] valid loss=0.4176, valid acc=0.8423, best valid acc=0.8423
main.py:train_model: [Epoch 73 Batch 5000/17173] loss=0.4517, acc=0.8226
main.py:train_model: [Epoch 73 Batch 10000/17173] loss=0.4521, acc=0.8248
main.py:train_model: [Epoch 73 Batch 15000/17173] loss=0.4518, acc=0.8243
main.py:train_model: [Epoch 73] valid loss=0.4179, valid acc=0.8422, best valid acc=0.8423
main.py:train_model: [Epoch 74 Batch 5000/17173] loss=0.4533, acc=0.8229
main.py:train_model: [Epoch 74 Batch 10000/17173] loss=0.4511, acc=0.8246
main.py:train_model: [Epoch 74 Batch 15000/17173] loss=0.4515, acc=0.8251
main.py:train_model: [Epoch 74] valid loss=0.4113, valid acc=0.8422, best valid acc=0.8423
main.py:train_model: [Epoch 75 Batch 5000/17173] loss=0.4483, acc=0.8247
main.py:train_model: [Epoch 75 Batch 10000/17173] loss=0.4526, acc=0.8239
main.py:train_model: [Epoch 75 Batch 15000/17173] loss=0.4516, acc=0.8250
main.py:train_model: [Epoch 75] valid loss=0.4182, valid acc=0.8400, best valid acc=0.8423
main.py:train_model: [Epoch 76 Batch 5000/17173] loss=0.4500, acc=0.8254
main.py:train_model: [Epoch 76 Batch 10000/17173] loss=0.4519, acc=0.8238
main.py:train_model: [Epoch 76 Batch 15000/17173] loss=0.4501, acc=0.8253
main.py:train_model: [Epoch 76] valid loss=0.4146, valid acc=0.8432, best valid acc=0.8432
main.py:train_model: [Epoch 77 Batch 5000/17173] loss=0.4518, acc=0.8239
main.py:train_model: [Epoch 77 Batch 10000/17173] loss=0.4510, acc=0.8242
main.py:train_model: [Epoch 77 Batch 15000/17173] loss=0.4502, acc=0.8250
main.py:train_model: [Epoch 77] valid loss=0.4151, valid acc=0.8406, best valid acc=0.8432
main.py:train_model: [Epoch 78 Batch 5000/17173] loss=0.4515, acc=0.8233
main.py:train_model: [Epoch 78 Batch 10000/17173] loss=0.4496, acc=0.8250
main.py:train_model: [Epoch 78 Batch 15000/17173] loss=0.4501, acc=0.8257
main.py:train_model: [Epoch 78] valid loss=0.4115, valid acc=0.8443, best valid acc=0.8443
main.py:train_model: [Epoch 79 Batch 5000/17173] loss=0.4496, acc=0.8255
main.py:train_model: [Epoch 79 Batch 10000/17173] loss=0.4494, acc=0.8248
main.py:train_model: [Epoch 79 Batch 15000/17173] loss=0.4507, acc=0.8239
main.py:train_model: [Epoch 79] valid loss=0.4138, valid acc=0.8433, best valid acc=0.8443
main.py:train_model: [Epoch 80 Batch 5000/17173] loss=0.4501, acc=0.8249
main.py:train_model: [Epoch 80 Batch 10000/17173] loss=0.4490, acc=0.8242
main.py:train_model: [Epoch 80 Batch 15000/17173] loss=0.4495, acc=0.8256
main.py:train_model: [Epoch 80] valid loss=0.4124, valid acc=0.8424, best valid acc=0.8443
main.py:train_model: [Epoch 81 Batch 5000/17173] loss=0.4460, acc=0.8265
main.py:train_model: [Epoch 81 Batch 10000/17173] loss=0.4509, acc=0.8246
main.py:train_model: [Epoch 81 Batch 15000/17173] loss=0.4493, acc=0.8251
main.py:train_model: [Epoch 81] valid loss=0.4145, valid acc=0.8406, best valid acc=0.8443
main.py:train_model: [Epoch 82 Batch 5000/17173] loss=0.4461, acc=0.8275
main.py:train_model: [Epoch 82 Batch 10000/17173] loss=0.4518, acc=0.8239
main.py:train_model: [Epoch 82 Batch 15000/17173] loss=0.4471, acc=0.8263
main.py:train_model: [Epoch 82] valid loss=0.4118, valid acc=0.8424, best valid acc=0.8443
main.py:train_model: [Epoch 83 Batch 5000/17173] loss=0.4438, acc=0.8272
main.py:train_model: [Epoch 83 Batch 10000/17173] loss=0.4480, acc=0.8256
main.py:train_model: [Epoch 83 Batch 15000/17173] loss=0.4499, acc=0.8255
main.py:train_model: [Epoch 83] valid loss=0.4102, valid acc=0.8433, best valid acc=0.8443
main.py:train_model: [Epoch 84 Batch 5000/17173] loss=0.4472, acc=0.8257
main.py:train_model: [Epoch 84 Batch 10000/17173] loss=0.4462, acc=0.8258
main.py:train_model: [Epoch 84 Batch 15000/17173] loss=0.4488, acc=0.8249
main.py:train_model: [Epoch 84] valid loss=0.4123, valid acc=0.8428, best valid acc=0.8443
main.py:train_model: [Epoch 85 Batch 5000/17173] loss=0.4471, acc=0.8254
main.py:train_model: [Epoch 85 Batch 10000/17173] loss=0.4464, acc=0.8264
main.py:train_model: [Epoch 85 Batch 15000/17173] loss=0.4460, acc=0.8273
main.py:train_model: [Epoch 85] valid loss=0.4106, valid acc=0.8445, best valid acc=0.8445
main.py:train_model: [Epoch 86 Batch 5000/17173] loss=0.4471, acc=0.8259
main.py:train_model: [Epoch 86 Batch 10000/17173] loss=0.4464, acc=0.8269
main.py:train_model: [Epoch 86 Batch 15000/17173] loss=0.4463, acc=0.8259
main.py:train_model: [Epoch 86] valid loss=0.4124, valid acc=0.8455, best valid acc=0.8455
main.py:train_model: [Epoch 87 Batch 5000/17173] loss=0.4440, acc=0.8269
main.py:train_model: [Epoch 87 Batch 10000/17173] loss=0.4483, acc=0.8257
main.py:train_model: [Epoch 87 Batch 15000/17173] loss=0.4466, acc=0.8255
main.py:train_model: [Epoch 87] valid loss=0.4075, valid acc=0.8453, best valid acc=0.8455
main.py:train_model: [Epoch 88 Batch 5000/17173] loss=0.4466, acc=0.8265
main.py:train_model: [Epoch 88 Batch 10000/17173] loss=0.4459, acc=0.8266
main.py:train_model: [Epoch 88 Batch 15000/17173] loss=0.4450, acc=0.8268
main.py:train_model: [Epoch 88] valid loss=0.4082, valid acc=0.8457, best valid acc=0.8457
main.py:train_model: [Epoch 89 Batch 5000/17173] loss=0.4455, acc=0.8259
main.py:train_model: [Epoch 89 Batch 10000/17173] loss=0.4455, acc=0.8270
main.py:train_model: [Epoch 89 Batch 15000/17173] loss=0.4442, acc=0.8270
main.py:train_model: [Epoch 89] valid loss=0.4118, valid acc=0.8456, best valid acc=0.8457
main.py:train_model: [Epoch 90 Batch 5000/17173] loss=0.4450, acc=0.8268
main.py:train_model: [Epoch 90 Batch 10000/17173] loss=0.4449, acc=0.8272
main.py:train_model: [Epoch 90 Batch 15000/17173] loss=0.4442, acc=0.8266
main.py:train_model: [Epoch 90] valid loss=0.4125, valid acc=0.8445, best valid acc=0.8457
main.py:train_model: [Epoch 91 Batch 5000/17173] loss=0.4422, acc=0.8272
main.py:train_model: [Epoch 91 Batch 10000/17173] loss=0.4457, acc=0.8280
main.py:train_model: [Epoch 91 Batch 15000/17173] loss=0.4460, acc=0.8274
main.py:train_model: [Epoch 91] valid loss=0.4080, valid acc=0.8452, best valid acc=0.8457
main.py:train_model: [Epoch 92 Batch 5000/17173] loss=0.4445, acc=0.8280
main.py:train_model: [Epoch 92 Batch 10000/17173] loss=0.4431, acc=0.8285
main.py:train_model: [Epoch 92 Batch 15000/17173] loss=0.4434, acc=0.8266
main.py:train_model: [Epoch 92] valid loss=0.4095, valid acc=0.8458, best valid acc=0.8458
main.py:train_model: [Epoch 93 Batch 5000/17173] loss=0.4399, acc=0.8292
main.py:train_model: [Epoch 93 Batch 10000/17173] loss=0.4454, acc=0.8270
main.py:train_model: [Epoch 93 Batch 15000/17173] loss=0.4452, acc=0.8277
main.py:train_model: [Epoch 93] valid loss=0.4133, valid acc=0.8444, best valid acc=0.8458
main.py:train_model: [Epoch 94 Batch 5000/17173] loss=0.4400, acc=0.8293
main.py:train_model: [Epoch 94 Batch 10000/17173] loss=0.4447, acc=0.8267
main.py:train_model: [Epoch 94 Batch 15000/17173] loss=0.4455, acc=0.8262
main.py:train_model: [Epoch 94] valid loss=0.4035, valid acc=0.8488, best valid acc=0.8488
main.py:train_model: [Epoch 95 Batch 5000/17173] loss=0.4427, acc=0.8278
main.py:train_model: [Epoch 95 Batch 10000/17173] loss=0.4434, acc=0.8280
main.py:train_model: [Epoch 95 Batch 15000/17173] loss=0.4426, acc=0.8284
main.py:train_model: [Epoch 95] valid loss=0.4092, valid acc=0.8452, best valid acc=0.8488
main.py:train_model: [Epoch 96 Batch 5000/17173] loss=0.4429, acc=0.8274
main.py:train_model: [Epoch 96 Batch 10000/17173] loss=0.4417, acc=0.8286
main.py:train_model: [Epoch 96 Batch 15000/17173] loss=0.4426, acc=0.8282
main.py:train_model: [Epoch 96] valid loss=0.4089, valid acc=0.8459, best valid acc=0.8488
main.py:train_model: [Epoch 97 Batch 5000/17173] loss=0.4417, acc=0.8279
main.py:train_model: [Epoch 97 Batch 10000/17173] loss=0.4409, acc=0.8297
main.py:train_model: [Epoch 97 Batch 15000/17173] loss=0.4433, acc=0.8276
main.py:train_model: [Epoch 97] valid loss=0.4058, valid acc=0.8462, best valid acc=0.8488
main.py:train_model: [Epoch 98 Batch 5000/17173] loss=0.4405, acc=0.8282
main.py:train_model: [Epoch 98 Batch 10000/17173] loss=0.4430, acc=0.8273
main.py:train_model: [Epoch 98 Batch 15000/17173] loss=0.4435, acc=0.8279
main.py:train_model: [Epoch 98] valid loss=0.4088, valid acc=0.8453, best valid acc=0.8488
main.py:train_model: [Epoch 99 Batch 5000/17173] loss=0.4420, acc=0.8292
main.py:train_model: [Epoch 99 Batch 10000/17173] loss=0.4394, acc=0.8282
main.py:train_model: [Epoch 99 Batch 15000/17173] loss=0.4409, acc=0.8279
main.py:train_model: [Epoch 99] valid loss=0.4063, valid acc=0.8459, best valid acc=0.8488
main.py:train_model: [Epoch 100 Batch 5000/17173] loss=0.4405, acc=0.8293
main.py:train_model: [Epoch 100 Batch 10000/17173] loss=0.4416, acc=0.8286
main.py:train_model: [Epoch 100 Batch 15000/17173] loss=0.4430, acc=0.8275
main.py:train_model: [Epoch 100] valid loss=0.4093, valid acc=0.8456, best valid acc=0.8488
main.py:train_model: [Epoch 101 Batch 5000/17173] loss=0.4383, acc=0.8288
main.py:train_model: [Epoch 101 Batch 10000/17173] loss=0.4407, acc=0.8295
main.py:train_model: [Epoch 101 Batch 15000/17173] loss=0.4435, acc=0.8271
main.py:train_model: [Epoch 101] valid loss=0.4059, valid acc=0.8472, best valid acc=0.8488
main.py:train_model: [Epoch 102 Batch 5000/17173] loss=0.4421, acc=0.8295
main.py:train_model: [Epoch 102 Batch 10000/17173] loss=0.4401, acc=0.8304
main.py:train_model: [Epoch 102 Batch 15000/17173] loss=0.4394, acc=0.8296
main.py:train_model: [Epoch 102] valid loss=0.4096, valid acc=0.8459, best valid acc=0.8488
main.py:train_model: [Epoch 103 Batch 5000/17173] loss=0.4400, acc=0.8293
main.py:train_model: [Epoch 103 Batch 10000/17173] loss=0.4395, acc=0.8295
main.py:train_model: [Epoch 103 Batch 15000/17173] loss=0.4437, acc=0.8277
main.py:train_model: [Epoch 103] valid loss=0.4067, valid acc=0.8475, best valid acc=0.8488
main.py:train_model: [Epoch 104 Batch 5000/17173] loss=0.4376, acc=0.8306
main.py:train_model: [Epoch 104 Batch 10000/17173] loss=0.4395, acc=0.8295
main.py:train_model: [Epoch 104 Batch 15000/17173] loss=0.4411, acc=0.8294
main.py:train_model: [Epoch 104] valid loss=0.4025, valid acc=0.8467, best valid acc=0.8488
main.py:train_model: [Epoch 105 Batch 5000/17173] loss=0.4390, acc=0.8290
main.py:train_model: [Epoch 105 Batch 10000/17173] loss=0.4388, acc=0.8305
main.py:train_model: [Epoch 105 Batch 15000/17173] loss=0.4421, acc=0.8279
main.py:train_model: [Epoch 105] valid loss=0.4055, valid acc=0.8467, best valid acc=0.8488
main.py:train_model: [Epoch 106 Batch 5000/17173] loss=0.4384, acc=0.8300
main.py:train_model: [Epoch 106 Batch 10000/17173] loss=0.4371, acc=0.8307
main.py:train_model: [Epoch 106 Batch 15000/17173] loss=0.4421, acc=0.8280
main.py:train_model: [Epoch 106] valid loss=0.4023, valid acc=0.8490, best valid acc=0.8490
main.py:train_model: [Epoch 107 Batch 5000/17173] loss=0.4390, acc=0.8293
main.py:train_model: [Epoch 107 Batch 10000/17173] loss=0.4398, acc=0.8288
main.py:train_model: [Epoch 107 Batch 15000/17173] loss=0.4386, acc=0.8298
main.py:train_model: [Epoch 107] valid loss=0.4063, valid acc=0.8465, best valid acc=0.8490
main.py:train_model: [Epoch 108 Batch 5000/17173] loss=0.4412, acc=0.8289
main.py:train_model: [Epoch 108 Batch 10000/17173] loss=0.4353, acc=0.8304
main.py:train_model: [Epoch 108 Batch 15000/17173] loss=0.4401, acc=0.8288
main.py:train_model: [Epoch 108] valid loss=0.4083, valid acc=0.8461, best valid acc=0.8490
main.py:train_model: [Epoch 109 Batch 5000/17173] loss=0.4388, acc=0.8297
main.py:train_model: [Epoch 109 Batch 10000/17173] loss=0.4400, acc=0.8296
main.py:train_model: [Epoch 109 Batch 15000/17173] loss=0.4390, acc=0.8285
main.py:train_model: [Epoch 109] valid loss=0.4062, valid acc=0.8471, best valid acc=0.8490
main.py:train_model: [Epoch 110 Batch 5000/17173] loss=0.4356, acc=0.8310
main.py:train_model: [Epoch 110 Batch 10000/17173] loss=0.4396, acc=0.8299
main.py:train_model: [Epoch 110 Batch 15000/17173] loss=0.4399, acc=0.8292
main.py:train_model: [Epoch 110] valid loss=0.4076, valid acc=0.8485, best valid acc=0.8490
main.py:train_model: [Epoch 111 Batch 5000/17173] loss=0.4371, acc=0.8298
main.py:train_model: [Epoch 111 Batch 10000/17173] loss=0.4355, acc=0.8308
main.py:train_model: [Epoch 111 Batch 15000/17173] loss=0.4405, acc=0.8286
main.py:train_model: [Epoch 111] valid loss=0.4029, valid acc=0.8480, best valid acc=0.8490
main.py:train_model: [Epoch 112 Batch 5000/17173] loss=0.4341, acc=0.8333
main.py:train_model: [Epoch 112 Batch 10000/17173] loss=0.4399, acc=0.8286
main.py:train_model: [Epoch 112 Batch 15000/17173] loss=0.4380, acc=0.8306
main.py:train_model: [Epoch 112] valid loss=0.4019, valid acc=0.8471, best valid acc=0.8490
main.py:train_model: [Epoch 113 Batch 5000/17173] loss=0.4362, acc=0.8304
main.py:train_model: [Epoch 113 Batch 10000/17173] loss=0.4394, acc=0.8287
main.py:train_model: [Epoch 113 Batch 15000/17173] loss=0.4344, acc=0.8316
main.py:train_model: [Epoch 113] valid loss=0.4003, valid acc=0.8493, best valid acc=0.8493
main.py:train_model: [Epoch 114 Batch 5000/17173] loss=0.4356, acc=0.8302
main.py:train_model: [Epoch 114 Batch 10000/17173] loss=0.4344, acc=0.8324
main.py:train_model: [Epoch 114 Batch 15000/17173] loss=0.4412, acc=0.8286
main.py:train_model: [Epoch 114] valid loss=0.4043, valid acc=0.8487, best valid acc=0.8493
main.py:train_model: [Epoch 115 Batch 5000/17173] loss=0.4338, acc=0.8327
main.py:train_model: [Epoch 115 Batch 10000/17173] loss=0.4381, acc=0.8302
main.py:train_model: [Epoch 115 Batch 15000/17173] loss=0.4356, acc=0.8309
main.py:train_model: [Epoch 115] valid loss=0.4061, valid acc=0.8472, best valid acc=0.8493
main.py:train_model: [Epoch 116 Batch 5000/17173] loss=0.4348, acc=0.8319
main.py:train_model: [Epoch 116 Batch 10000/17173] loss=0.4336, acc=0.8318
main.py:train_model: [Epoch 116 Batch 15000/17173] loss=0.4408, acc=0.8294
main.py:train_model: [Epoch 116] valid loss=0.4041, valid acc=0.8497, best valid acc=0.8497
main.py:train_model: [Epoch 117 Batch 5000/17173] loss=0.4340, acc=0.8321
main.py:train_model: [Epoch 117 Batch 10000/17173] loss=0.4367, acc=0.8299
main.py:train_model: [Epoch 117 Batch 15000/17173] loss=0.4381, acc=0.8302
main.py:train_model: [Epoch 117] valid loss=0.4093, valid acc=0.8474, best valid acc=0.8497
main.py:train_model: [Epoch 118 Batch 5000/17173] loss=0.4356, acc=0.8316
main.py:train_model: [Epoch 118 Batch 10000/17173] loss=0.4341, acc=0.8316
main.py:train_model: [Epoch 118 Batch 15000/17173] loss=0.4379, acc=0.8294
main.py:train_model: [Epoch 118] valid loss=0.4066, valid acc=0.8471, best valid acc=0.8497
main.py:train_model: [Epoch 119 Batch 5000/17173] loss=0.4330, acc=0.8329
main.py:train_model: [Epoch 119 Batch 10000/17173] loss=0.4392, acc=0.8298
main.py:train_model: [Epoch 119 Batch 15000/17173] loss=0.4335, acc=0.8318
main.py:train_model: [Epoch 119] valid loss=0.4031, valid acc=0.8511, best valid acc=0.8511
main.py:train_model: [Epoch 120 Batch 5000/17173] loss=0.4325, acc=0.8328
main.py:train_model: [Epoch 120 Batch 10000/17173] loss=0.4382, acc=0.8298
main.py:train_model: [Epoch 120 Batch 15000/17173] loss=0.4362, acc=0.8305
main.py:train_model: [Epoch 120] valid loss=0.4045, valid acc=0.8492, best valid acc=0.8511
main.py:train_model: [Epoch 121 Batch 5000/17173] loss=0.4303, acc=0.8335
main.py:train_model: [Epoch 121 Batch 10000/17173] loss=0.4359, acc=0.8314
main.py:train_model: [Epoch 121 Batch 15000/17173] loss=0.4382, acc=0.8303
main.py:train_model: [Epoch 121] valid loss=0.4063, valid acc=0.8472, best valid acc=0.8511
main.py:train_model: [Epoch 122 Batch 5000/17173] loss=0.4320, acc=0.8326
main.py:train_model: [Epoch 122 Batch 10000/17173] loss=0.4343, acc=0.8314
main.py:train_model: [Epoch 122 Batch 15000/17173] loss=0.4357, acc=0.8308
main.py:train_model: [Epoch 122] valid loss=0.4021, valid acc=0.8476, best valid acc=0.8511
main.py:train_model: [Epoch 123 Batch 5000/17173] loss=0.4334, acc=0.8327
main.py:train_model: [Epoch 123 Batch 10000/17173] loss=0.4341, acc=0.8320
main.py:train_model: [Epoch 123 Batch 15000/17173] loss=0.4352, acc=0.8317
main.py:train_model: [Epoch 123] valid loss=0.4050, valid acc=0.8483, best valid acc=0.8511
main.py:train_model: [Epoch 124 Batch 5000/17173] loss=0.4361, acc=0.8307
main.py:train_model: [Epoch 124 Batch 10000/17173] loss=0.4346, acc=0.8302
main.py:train_model: [Epoch 124 Batch 15000/17173] loss=0.4335, acc=0.8320
main.py:train_model: [Epoch 124] valid loss=0.4017, valid acc=0.8478, best valid acc=0.8511
main.py:train_model: [Epoch 125 Batch 5000/17173] loss=0.4333, acc=0.8309
main.py:train_model: [Epoch 125 Batch 10000/17173] loss=0.4328, acc=0.8331
main.py:train_model: [Epoch 125 Batch 15000/17173] loss=0.4374, acc=0.8298
main.py:train_model: [Epoch 125] valid loss=0.4032, valid acc=0.8494, best valid acc=0.8511
main.py:train_model: [Epoch 126 Batch 5000/17173] loss=0.4329, acc=0.8323
main.py:train_model: [Epoch 126 Batch 10000/17173] loss=0.4352, acc=0.8304
main.py:train_model: [Epoch 126 Batch 15000/17173] loss=0.4359, acc=0.8317
main.py:train_model: [Epoch 126] valid loss=0.4020, valid acc=0.8488, best valid acc=0.8511
main.py:train_model: [Epoch 127 Batch 5000/17173] loss=0.4327, acc=0.8322
main.py:train_model: [Epoch 127 Batch 10000/17173] loss=0.4323, acc=0.8333
main.py:train_model: [Epoch 127 Batch 15000/17173] loss=0.4353, acc=0.8321
main.py:train_model: [Epoch 127] valid loss=0.4045, valid acc=0.8478, best valid acc=0.8511
main.py:train_model: [Epoch 128 Batch 5000/17173] loss=0.4296, acc=0.8342
main.py:train_model: [Epoch 128 Batch 10000/17173] loss=0.4346, acc=0.8309
main.py:train_model: [Epoch 128 Batch 15000/17173] loss=0.4357, acc=0.8315
main.py:train_model: [Epoch 128] valid loss=0.4004, valid acc=0.8502, best valid acc=0.8511
main.py:train_model: [Epoch 129 Batch 5000/17173] loss=0.4321, acc=0.8320
main.py:train_model: [Epoch 129 Batch 10000/17173] loss=0.4320, acc=0.8333
main.py:train_model: [Epoch 129 Batch 15000/17173] loss=0.4336, acc=0.8330
main.py:train_model: [Epoch 129] valid loss=0.4035, valid acc=0.8493, best valid acc=0.8511
main.py:train_model: [Epoch 130 Batch 5000/17173] loss=0.4336, acc=0.8321
main.py:train_model: [Epoch 130 Batch 10000/17173] loss=0.4338, acc=0.8319
main.py:train_model: [Epoch 130 Batch 15000/17173] loss=0.4317, acc=0.8333
main.py:train_model: [Epoch 130] valid loss=0.3999, valid acc=0.8504, best valid acc=0.8511
main.py:train_model: [Epoch 131 Batch 5000/17173] loss=0.4296, acc=0.8345
main.py:train_model: [Epoch 131 Batch 10000/17173] loss=0.4307, acc=0.8333
main.py:train_model: [Epoch 131 Batch 15000/17173] loss=0.4346, acc=0.8318
main.py:train_model: [Epoch 131] valid loss=0.4005, valid acc=0.8499, best valid acc=0.8511
main.py:train_model: [Epoch 132 Batch 5000/17173] loss=0.4299, acc=0.8334
main.py:train_model: [Epoch 132 Batch 10000/17173] loss=0.4336, acc=0.8325
main.py:train_model: [Epoch 132 Batch 15000/17173] loss=0.4314, acc=0.8332
main.py:train_model: [Epoch 132] valid loss=0.4013, valid acc=0.8487, best valid acc=0.8511
main.py:train_model: [Epoch 133 Batch 5000/17173] loss=0.4294, acc=0.8333
main.py:train_model: [Epoch 133 Batch 10000/17173] loss=0.4328, acc=0.8316
main.py:train_model: [Epoch 133 Batch 15000/17173] loss=0.4324, acc=0.8314
main.py:train_model: [Epoch 133] valid loss=0.4032, valid acc=0.8486, best valid acc=0.8511
main.py:train_model: [Epoch 134 Batch 5000/17173] loss=0.4300, acc=0.8343
main.py:train_model: [Epoch 134 Batch 10000/17173] loss=0.4324, acc=0.8328
main.py:train_model: [Epoch 134 Batch 15000/17173] loss=0.4346, acc=0.8318
main.py:train_model: [Epoch 134] valid loss=0.4045, valid acc=0.8486, best valid acc=0.8511
main.py:train_model: [Epoch 135 Batch 5000/17173] loss=0.4309, acc=0.8329
main.py:train_model: [Epoch 135 Batch 10000/17173] loss=0.4308, acc=0.8328
main.py:train_model: [Epoch 135 Batch 15000/17173] loss=0.4318, acc=0.8324
main.py:train_model: [Epoch 135] valid loss=0.3994, valid acc=0.8514, best valid acc=0.8514
main.py:train_model: [Epoch 136 Batch 5000/17173] loss=0.4305, acc=0.8338
main.py:train_model: [Epoch 136 Batch 10000/17173] loss=0.4326, acc=0.8330
main.py:train_model: [Epoch 136 Batch 15000/17173] loss=0.4325, acc=0.8331
main.py:train_model: [Epoch 136] valid loss=0.4052, valid acc=0.8480, best valid acc=0.8514
main.py:train_model: [Epoch 137 Batch 5000/17173] loss=0.4321, acc=0.8326
main.py:train_model: [Epoch 137 Batch 10000/17173] loss=0.4319, acc=0.8332
main.py:train_model: [Epoch 137 Batch 15000/17173] loss=0.4316, acc=0.8334
main.py:train_model: [Epoch 137] valid loss=0.4065, valid acc=0.8469, best valid acc=0.8514
main.py:train_model: [Epoch 138 Batch 5000/17173] loss=0.4285, acc=0.8342
main.py:train_model: [Epoch 138 Batch 10000/17173] loss=0.4327, acc=0.8335
main.py:train_model: [Epoch 138 Batch 15000/17173] loss=0.4317, acc=0.8317
main.py:train_model: [Epoch 138] valid loss=0.4014, valid acc=0.8498, best valid acc=0.8514
main.py:train_model: [Epoch 139 Batch 5000/17173] loss=0.4297, acc=0.8338
main.py:train_model: [Epoch 139 Batch 10000/17173] loss=0.4314, acc=0.8331
main.py:train_model: [Epoch 139 Batch 15000/17173] loss=0.4310, acc=0.8334
main.py:train_model: [Epoch 139] valid loss=0.3998, valid acc=0.8515, best valid acc=0.8515
main.py:train_model: [Epoch 140 Batch 5000/17173] loss=0.4261, acc=0.8347
main.py:train_model: [Epoch 140 Batch 10000/17173] loss=0.4306, acc=0.8331
main.py:train_model: [Epoch 140 Batch 15000/17173] loss=0.4316, acc=0.8330
main.py:train_model: [Epoch 140] valid loss=0.4014, valid acc=0.8508, best valid acc=0.8515
main.py:train_model: [Epoch 141 Batch 5000/17173] loss=0.4285, acc=0.8337
main.py:train_model: [Epoch 141 Batch 10000/17173] loss=0.4302, acc=0.8325
main.py:train_model: [Epoch 141 Batch 15000/17173] loss=0.4313, acc=0.8326
main.py:train_model: [Epoch 141] valid loss=0.3977, valid acc=0.8510, best valid acc=0.8515
main.py:train_model: [Epoch 142 Batch 5000/17173] loss=0.4275, acc=0.8354
main.py:train_model: [Epoch 142 Batch 10000/17173] loss=0.4305, acc=0.8331
main.py:train_model: [Epoch 142 Batch 15000/17173] loss=0.4312, acc=0.8324
main.py:train_model: [Epoch 142] valid loss=0.4032, valid acc=0.8488, best valid acc=0.8515
main.py:train_model: [Epoch 143 Batch 5000/17173] loss=0.4280, acc=0.8346
main.py:train_model: [Epoch 143 Batch 10000/17173] loss=0.4275, acc=0.8348
main.py:train_model: [Epoch 143 Batch 15000/17173] loss=0.4309, acc=0.8326
main.py:train_model: [Epoch 143] valid loss=0.3989, valid acc=0.8517, best valid acc=0.8517
main.py:train_model: [Epoch 144 Batch 5000/17173] loss=0.4298, acc=0.8342
main.py:train_model: [Epoch 144 Batch 10000/17173] loss=0.4299, acc=0.8336
main.py:train_model: [Epoch 144 Batch 15000/17173] loss=0.4298, acc=0.8331
main.py:train_model: [Epoch 144] valid loss=0.4034, valid acc=0.8496, best valid acc=0.8517
main.py:train_model: [Epoch 145 Batch 5000/17173] loss=0.4292, acc=0.8333
main.py:train_model: [Epoch 145 Batch 10000/17173] loss=0.4289, acc=0.8337
main.py:train_model: [Epoch 145 Batch 15000/17173] loss=0.4329, acc=0.8319
main.py:train_model: [Epoch 145] valid loss=0.3981, valid acc=0.8528, best valid acc=0.8528
main.py:train_model: [Epoch 146 Batch 5000/17173] loss=0.4251, acc=0.8357
main.py:train_model: [Epoch 146 Batch 10000/17173] loss=0.4311, acc=0.8332
main.py:train_model: [Epoch 146 Batch 15000/17173] loss=0.4280, acc=0.8341
main.py:train_model: [Epoch 146] valid loss=0.4013, valid acc=0.8520, best valid acc=0.8528
main.py:train_model: [Epoch 147 Batch 5000/17173] loss=0.4258, acc=0.8364
main.py:train_model: [Epoch 147 Batch 10000/17173] loss=0.4293, acc=0.8336
main.py:train_model: [Epoch 147 Batch 15000/17173] loss=0.4309, acc=0.8337
main.py:train_model: [Epoch 147] valid loss=0.3991, valid acc=0.8496, best valid acc=0.8528
main.py:train_model: [Epoch 148 Batch 5000/17173] loss=0.4279, acc=0.8347
main.py:train_model: [Epoch 148 Batch 10000/17173] loss=0.4301, acc=0.8335
main.py:train_model: [Epoch 148 Batch 15000/17173] loss=0.4269, acc=0.8340
main.py:train_model: [Epoch 148] valid loss=0.3974, valid acc=0.8514, best valid acc=0.8528
main.py:train_model: [Epoch 149 Batch 5000/17173] loss=0.4263, acc=0.8358
main.py:train_model: [Epoch 149 Batch 10000/17173] loss=0.4287, acc=0.8344
main.py:train_model: [Epoch 149 Batch 15000/17173] loss=0.4290, acc=0.8332
main.py:train_model: [Epoch 149] valid loss=0.3971, valid acc=0.8519, best valid acc=0.8528
main.py:train_model: [Epoch 150 Batch 5000/17173] loss=0.4297, acc=0.8334
main.py:train_model: [Epoch 150 Batch 10000/17173] loss=0.4266, acc=0.8353
main.py:train_model: [Epoch 150 Batch 15000/17173] loss=0.4271, acc=0.8349
main.py:train_model: [Epoch 150] valid loss=0.3970, valid acc=0.8508, best valid acc=0.8528
main.py:train_model: [Epoch 151 Batch 5000/17173] loss=0.4277, acc=0.8347
main.py:train_model: [Epoch 151 Batch 10000/17173] loss=0.4272, acc=0.8353
main.py:train_model: [Epoch 151 Batch 15000/17173] loss=0.4315, acc=0.8328
main.py:train_model: [Epoch 151] valid loss=0.3978, valid acc=0.8515, best valid acc=0.8528
main.py:train_model: [Epoch 152 Batch 5000/17173] loss=0.4284, acc=0.8340
main.py:train_model: [Epoch 152 Batch 10000/17173] loss=0.4272, acc=0.8344
main.py:train_model: [Epoch 152 Batch 15000/17173] loss=0.4306, acc=0.8340
main.py:train_model: [Epoch 152] valid loss=0.3992, valid acc=0.8495, best valid acc=0.8528
main.py:train_model: [Epoch 153 Batch 5000/17173] loss=0.4288, acc=0.8335
main.py:train_model: [Epoch 153 Batch 10000/17173] loss=0.4275, acc=0.8351
main.py:train_model: [Epoch 153 Batch 15000/17173] loss=0.4290, acc=0.8345
main.py:train_model: [Epoch 153] valid loss=0.3954, valid acc=0.8539, best valid acc=0.8539
main.py:train_model: [Epoch 154 Batch 5000/17173] loss=0.4300, acc=0.8336
main.py:train_model: [Epoch 154 Batch 10000/17173] loss=0.4271, acc=0.8348
main.py:train_model: [Epoch 154 Batch 15000/17173] loss=0.4260, acc=0.8349
main.py:train_model: [Epoch 154] valid loss=0.3968, valid acc=0.8529, best valid acc=0.8539
main.py:train_model: [Epoch 155 Batch 5000/17173] loss=0.4249, acc=0.8365
main.py:train_model: [Epoch 155 Batch 10000/17173] loss=0.4269, acc=0.8345
main.py:train_model: [Epoch 155 Batch 15000/17173] loss=0.4308, acc=0.8335
main.py:train_model: [Epoch 155] valid loss=0.3972, valid acc=0.8529, best valid acc=0.8539
main.py:train_model: [Epoch 156 Batch 5000/17173] loss=0.4281, acc=0.8345
main.py:train_model: [Epoch 156 Batch 10000/17173] loss=0.4258, acc=0.8356
main.py:train_model: [Epoch 156 Batch 15000/17173] loss=0.4284, acc=0.8338
main.py:train_model: [Epoch 156] valid loss=0.3988, valid acc=0.8527, best valid acc=0.8539
main.py:train_model: [Epoch 157 Batch 5000/17173] loss=0.4244, acc=0.8353
main.py:train_model: [Epoch 157 Batch 10000/17173] loss=0.4274, acc=0.8349
main.py:train_model: [Epoch 157 Batch 15000/17173] loss=0.4296, acc=0.8349
main.py:train_model: [Epoch 157] valid loss=0.3985, valid acc=0.8525, best valid acc=0.8539
main.py:train_model: [Epoch 158 Batch 5000/17173] loss=0.4283, acc=0.8334
main.py:train_model: [Epoch 158 Batch 10000/17173] loss=0.4247, acc=0.8359
main.py:train_model: [Epoch 158 Batch 15000/17173] loss=0.4283, acc=0.8349
main.py:train_model: [Epoch 158] valid loss=0.3981, valid acc=0.8516, best valid acc=0.8539
main.py:train_model: [Epoch 159 Batch 5000/17173] loss=0.4264, acc=0.8353
main.py:train_model: [Epoch 159 Batch 10000/17173] loss=0.4284, acc=0.8340
main.py:train_model: [Epoch 159 Batch 15000/17173] loss=0.4253, acc=0.8355
main.py:train_model: [Epoch 159] valid loss=0.3963, valid acc=0.8539, best valid acc=0.8539
main.py:train_model: [Epoch 160 Batch 5000/17173] loss=0.4246, acc=0.8356
main.py:train_model: [Epoch 160 Batch 10000/17173] loss=0.4268, acc=0.8343
main.py:train_model: [Epoch 160 Batch 15000/17173] loss=0.4293, acc=0.8337
main.py:train_model: [Epoch 160] valid loss=0.3966, valid acc=0.8537, best valid acc=0.8539
main.py:train_model: [Epoch 161 Batch 5000/17173] loss=0.4238, acc=0.8362
main.py:train_model: [Epoch 161 Batch 10000/17173] loss=0.4241, acc=0.8358
main.py:train_model: [Epoch 161 Batch 15000/17173] loss=0.4298, acc=0.8346
main.py:train_model: [Epoch 161] valid loss=0.3994, valid acc=0.8517, best valid acc=0.8539
main.py:train_model: [Epoch 162 Batch 5000/17173] loss=0.4265, acc=0.8355
main.py:train_model: [Epoch 162 Batch 10000/17173] loss=0.4262, acc=0.8353
main.py:train_model: [Epoch 162 Batch 15000/17173] loss=0.4272, acc=0.8338
main.py:train_model: [Epoch 162] valid loss=0.3977, valid acc=0.8527, best valid acc=0.8539
main.py:train_model: [Epoch 163 Batch 5000/17173] loss=0.4261, acc=0.8358
main.py:train_model: [Epoch 163 Batch 10000/17173] loss=0.4252, acc=0.8355
main.py:train_model: [Epoch 163 Batch 15000/17173] loss=0.4250, acc=0.8356
main.py:train_model: [Epoch 163] valid loss=0.3968, valid acc=0.8519, best valid acc=0.8539
main.py:train_model: [Epoch 164 Batch 5000/17173] loss=0.4266, acc=0.8342
main.py:train_model: [Epoch 164 Batch 10000/17173] loss=0.4271, acc=0.8337
main.py:train_model: [Epoch 164 Batch 15000/17173] loss=0.4265, acc=0.8354
main.py:train_model: [Epoch 164] valid loss=0.3967, valid acc=0.8527, best valid acc=0.8539
main.py:train_model: [Epoch 165 Batch 5000/17173] loss=0.4243, acc=0.8362
main.py:train_model: [Epoch 165 Batch 10000/17173] loss=0.4261, acc=0.8360
main.py:train_model: [Epoch 165 Batch 15000/17173] loss=0.4256, acc=0.8364
main.py:train_model: [Epoch 165] valid loss=0.3941, valid acc=0.8537, best valid acc=0.8539
main.py:train_model: [Epoch 166 Batch 5000/17173] loss=0.4261, acc=0.8356
main.py:train_model: [Epoch 166 Batch 10000/17173] loss=0.4249, acc=0.8352
main.py:train_model: [Epoch 166 Batch 15000/17173] loss=0.4254, acc=0.8346
main.py:train_model: [Epoch 166] valid loss=0.3934, valid acc=0.8540, best valid acc=0.8540
main.py:train_model: [Epoch 167 Batch 5000/17173] loss=0.4245, acc=0.8353
main.py:train_model: [Epoch 167 Batch 10000/17173] loss=0.4275, acc=0.8353
main.py:train_model: [Epoch 167 Batch 15000/17173] loss=0.4227, acc=0.8360
main.py:train_model: [Epoch 167] valid loss=0.3984, valid acc=0.8516, best valid acc=0.8540
main.py:train_model: [Epoch 168 Batch 5000/17173] loss=0.4230, acc=0.8360
main.py:train_model: [Epoch 168 Batch 10000/17173] loss=0.4266, acc=0.8349
main.py:train_model: [Epoch 168 Batch 15000/17173] loss=0.4251, acc=0.8355
main.py:train_model: [Epoch 168] valid loss=0.3966, valid acc=0.8531, best valid acc=0.8540
main.py:train_model: [Epoch 169 Batch 5000/17173] loss=0.4253, acc=0.8359
main.py:train_model: [Epoch 169 Batch 10000/17173] loss=0.4239, acc=0.8361
main.py:train_model: [Epoch 169 Batch 15000/17173] loss=0.4244, acc=0.8364
main.py:train_model: [Epoch 169] valid loss=0.3991, valid acc=0.8515, best valid acc=0.8540
main.py:train_model: [Epoch 170 Batch 5000/17173] loss=0.4246, acc=0.8353
main.py:train_model: [Epoch 170 Batch 10000/17173] loss=0.4258, acc=0.8349
main.py:train_model: [Epoch 170 Batch 15000/17173] loss=0.4259, acc=0.8350
main.py:train_model: [Epoch 170] valid loss=0.3991, valid acc=0.8511, best valid acc=0.8540
main.py:train_model: [Epoch 171 Batch 5000/17173] loss=0.4242, acc=0.8364
main.py:train_model: [Epoch 171 Batch 10000/17173] loss=0.4238, acc=0.8363
main.py:train_model: [Epoch 171 Batch 15000/17173] loss=0.4244, acc=0.8359
main.py:train_model: [Epoch 171] valid loss=0.3961, valid acc=0.8525, best valid acc=0.8540
main.py:train_model: [Epoch 172 Batch 5000/17173] loss=0.4248, acc=0.8352
main.py:train_model: [Epoch 172 Batch 10000/17173] loss=0.4213, acc=0.8383
main.py:train_model: [Epoch 172 Batch 15000/17173] loss=0.4262, acc=0.8353
main.py:train_model: [Epoch 172] valid loss=0.3951, valid acc=0.8522, best valid acc=0.8540
main.py:train_model: [Epoch 173 Batch 5000/17173] loss=0.4249, acc=0.8350
main.py:train_model: [Epoch 173 Batch 10000/17173] loss=0.4245, acc=0.8363
main.py:train_model: [Epoch 173 Batch 15000/17173] loss=0.4252, acc=0.8361
main.py:train_model: [Epoch 173] valid loss=0.3977, valid acc=0.8531, best valid acc=0.8540
main.py:train_model: [Epoch 174 Batch 5000/17173] loss=0.4221, acc=0.8372
main.py:train_model: [Epoch 174 Batch 10000/17173] loss=0.4231, acc=0.8365
main.py:train_model: [Epoch 174 Batch 15000/17173] loss=0.4276, acc=0.8344
main.py:train_model: [Epoch 174] valid loss=0.3965, valid acc=0.8549, best valid acc=0.8549
main.py:train_model: [Epoch 175 Batch 5000/17173] loss=0.4230, acc=0.8368
main.py:train_model: [Epoch 175 Batch 10000/17173] loss=0.4249, acc=0.8364
main.py:train_model: [Epoch 175 Batch 15000/17173] loss=0.4244, acc=0.8361
main.py:train_model: [Epoch 175] valid loss=0.3986, valid acc=0.8535, best valid acc=0.8549
main.py:train_model: [Epoch 176 Batch 5000/17173] loss=0.4224, acc=0.8372
main.py:train_model: [Epoch 176 Batch 10000/17173] loss=0.4245, acc=0.8363
main.py:train_model: [Epoch 176 Batch 15000/17173] loss=0.4236, acc=0.8352
main.py:train_model: [Epoch 176] valid loss=0.3958, valid acc=0.8542, best valid acc=0.8549
main.py:train_model: [Epoch 177 Batch 5000/17173] loss=0.4241, acc=0.8365
main.py:train_model: [Epoch 177 Batch 10000/17173] loss=0.4237, acc=0.8366
main.py:train_model: [Epoch 177 Batch 15000/17173] loss=0.4262, acc=0.8350
main.py:train_model: [Epoch 177] valid loss=0.3945, valid acc=0.8536, best valid acc=0.8549
main.py:train_model: [Epoch 178 Batch 5000/17173] loss=0.4218, acc=0.8369
main.py:train_model: [Epoch 178 Batch 10000/17173] loss=0.4274, acc=0.8352
main.py:train_model: [Epoch 178 Batch 15000/17173] loss=0.4233, acc=0.8368
main.py:train_model: [Epoch 178] valid loss=0.3928, valid acc=0.8548, best valid acc=0.8549
main.py:train_model: [Epoch 179 Batch 5000/17173] loss=0.4222, acc=0.8373
main.py:train_model: [Epoch 179 Batch 10000/17173] loss=0.4241, acc=0.8356
main.py:train_model: [Epoch 179 Batch 15000/17173] loss=0.4219, acc=0.8373
main.py:train_model: [Epoch 179] valid loss=0.3930, valid acc=0.8518, best valid acc=0.8549
main.py:train_model: [Epoch 180 Batch 5000/17173] loss=0.4241, acc=0.8356
main.py:train_model: [Epoch 180 Batch 10000/17173] loss=0.4205, acc=0.8376
main.py:train_model: [Epoch 180 Batch 15000/17173] loss=0.4245, acc=0.8360
main.py:train_model: [Epoch 180] valid loss=0.3984, valid acc=0.8523, best valid acc=0.8549
main.py:train_model: [Epoch 181 Batch 5000/17173] loss=0.4253, acc=0.8352
main.py:train_model: [Epoch 181 Batch 10000/17173] loss=0.4193, acc=0.8389
main.py:train_model: [Epoch 181 Batch 15000/17173] loss=0.4249, acc=0.8361
main.py:train_model: [Epoch 181] valid loss=0.3961, valid acc=0.8533, best valid acc=0.8549
main.py:train_model: [Epoch 182 Batch 5000/17173] loss=0.4232, acc=0.8369
main.py:train_model: [Epoch 182 Batch 10000/17173] loss=0.4225, acc=0.8367
main.py:train_model: [Epoch 182 Batch 15000/17173] loss=0.4240, acc=0.8357
main.py:train_model: [Epoch 182] valid loss=0.3949, valid acc=0.8537, best valid acc=0.8549
main.py:train_model: [Epoch 183 Batch 5000/17173] loss=0.4209, acc=0.8370
main.py:train_model: [Epoch 183 Batch 10000/17173] loss=0.4203, acc=0.8383
main.py:train_model: [Epoch 183 Batch 15000/17173] loss=0.4283, acc=0.8341
main.py:train_model: [Epoch 183] valid loss=0.3972, valid acc=0.8525, best valid acc=0.8549
main.py:train_model: [Epoch 184 Batch 5000/17173] loss=0.4203, acc=0.8378
main.py:train_model: [Epoch 184 Batch 10000/17173] loss=0.4227, acc=0.8371
main.py:train_model: [Epoch 184 Batch 15000/17173] loss=0.4255, acc=0.8351
main.py:train_model: [Epoch 184] valid loss=0.3909, valid acc=0.8544, best valid acc=0.8549
main.py:train_model: [Epoch 185 Batch 5000/17173] loss=0.4232, acc=0.8368
main.py:train_model: [Epoch 185 Batch 10000/17173] loss=0.4228, acc=0.8366
main.py:train_model: [Epoch 185 Batch 15000/17173] loss=0.4215, acc=0.8378
main.py:train_model: [Epoch 185] valid loss=0.3971, valid acc=0.8533, best valid acc=0.8549
main.py:train_model: [Epoch 186 Batch 5000/17173] loss=0.4229, acc=0.8362
main.py:train_model: [Epoch 186 Batch 10000/17173] loss=0.4244, acc=0.8360
main.py:train_model: [Epoch 186 Batch 15000/17173] loss=0.4227, acc=0.8353
main.py:train_model: [Epoch 186] valid loss=0.3946, valid acc=0.8543, best valid acc=0.8549
main.py:train_model: [Epoch 187 Batch 5000/17173] loss=0.4249, acc=0.8362
main.py:train_model: [Epoch 187 Batch 10000/17173] loss=0.4234, acc=0.8361
main.py:train_model: [Epoch 187 Batch 15000/17173] loss=0.4199, acc=0.8383
main.py:train_model: [Epoch 187] valid loss=0.3979, valid acc=0.8527, best valid acc=0.8549
main.py:train_model: [Epoch 188 Batch 5000/17173] loss=0.4214, acc=0.8373
main.py:train_model: [Epoch 188 Batch 10000/17173] loss=0.4260, acc=0.8354
main.py:train_model: [Epoch 188 Batch 15000/17173] loss=0.4214, acc=0.8367
main.py:train_model: [Epoch 188] valid loss=0.3931, valid acc=0.8561, best valid acc=0.8561
main.py:train_model: [Epoch 189 Batch 5000/17173] loss=0.4214, acc=0.8377
main.py:train_model: [Epoch 189 Batch 10000/17173] loss=0.4233, acc=0.8359
main.py:train_model: [Epoch 189 Batch 15000/17173] loss=0.4218, acc=0.8371
main.py:train_model: [Epoch 189] valid loss=0.3969, valid acc=0.8526, best valid acc=0.8561
main.py:train_model: [Epoch 190 Batch 5000/17173] loss=0.4197, acc=0.8384
main.py:train_model: [Epoch 190 Batch 10000/17173] loss=0.4228, acc=0.8360
main.py:train_model: [Epoch 190 Batch 15000/17173] loss=0.4225, acc=0.8366
main.py:train_model: [Epoch 190] valid loss=0.3960, valid acc=0.8532, best valid acc=0.8561
main.py:train_model: [Epoch 191 Batch 5000/17173] loss=0.4184, acc=0.8384
main.py:train_model: [Epoch 191 Batch 10000/17173] loss=0.4235, acc=0.8360
main.py:train_model: [Epoch 191 Batch 15000/17173] loss=0.4219, acc=0.8373
main.py:train_model: [Epoch 191] valid loss=0.3938, valid acc=0.8551, best valid acc=0.8561
main.py:train_model: [Epoch 192 Batch 5000/17173] loss=0.4188, acc=0.8388
main.py:train_model: [Epoch 192 Batch 10000/17173] loss=0.4226, acc=0.8371
main.py:train_model: [Epoch 192 Batch 15000/17173] loss=0.4245, acc=0.8372
main.py:train_model: [Epoch 192] valid loss=0.3952, valid acc=0.8551, best valid acc=0.8561
main.py:train_model: [Epoch 193 Batch 5000/17173] loss=0.4209, acc=0.8382
main.py:train_model: [Epoch 193 Batch 10000/17173] loss=0.4230, acc=0.8357
main.py:train_model: [Epoch 193 Batch 15000/17173] loss=0.4200, acc=0.8383
main.py:train_model: [Epoch 193] valid loss=0.3911, valid acc=0.8552, best valid acc=0.8561
main.py:train_model: [Epoch 194 Batch 5000/17173] loss=0.4234, acc=0.8367
main.py:train_model: [Epoch 194 Batch 10000/17173] loss=0.4199, acc=0.8380
main.py:train_model: [Epoch 194 Batch 15000/17173] loss=0.4226, acc=0.8379
main.py:train_model: [Epoch 194] valid loss=0.3993, valid acc=0.8541, best valid acc=0.8561
main.py:train_model: [Epoch 195 Batch 5000/17173] loss=0.4217, acc=0.8370
main.py:train_model: [Epoch 195 Batch 10000/17173] loss=0.4227, acc=0.8369
main.py:train_model: [Epoch 195 Batch 15000/17173] loss=0.4183, acc=0.8388
main.py:train_model: [Epoch 195] valid loss=0.3953, valid acc=0.8532, best valid acc=0.8561
main.py:train_model: [Epoch 196 Batch 5000/17173] loss=0.4233, acc=0.8373
main.py:train_model: [Epoch 196 Batch 10000/17173] loss=0.4199, acc=0.8385
main.py:train_model: [Epoch 196 Batch 15000/17173] loss=0.4202, acc=0.8378
main.py:train_model: [Epoch 196] valid loss=0.3965, valid acc=0.8545, best valid acc=0.8561
main.py:train_model: [Epoch 197 Batch 5000/17173] loss=0.4202, acc=0.8372
main.py:train_model: [Epoch 197 Batch 10000/17173] loss=0.4233, acc=0.8360
main.py:train_model: [Epoch 197 Batch 15000/17173] loss=0.4213, acc=0.8375
main.py:train_model: [Epoch 197] valid loss=0.3974, valid acc=0.8538, best valid acc=0.8561
main.py:train_model: [Epoch 198 Batch 5000/17173] loss=0.4184, acc=0.8385
main.py:train_model: [Epoch 198 Batch 10000/17173] loss=0.4216, acc=0.8367
main.py:train_model: [Epoch 198 Batch 15000/17173] loss=0.4229, acc=0.8362
main.py:train_model: [Epoch 198] valid loss=0.3923, valid acc=0.8562, best valid acc=0.8562
main.py:train_model: [Epoch 199 Batch 5000/17173] loss=0.4195, acc=0.8377
main.py:train_model: [Epoch 199 Batch 10000/17173] loss=0.4165, acc=0.8402
main.py:train_model: [Epoch 199 Batch 15000/17173] loss=0.4214, acc=0.8389
main.py:train_model: [Epoch 199] valid loss=0.3902, valid acc=0.8551, best valid acc=0.8562
main.py:train_model: [Epoch 200 Batch 5000/17173] loss=0.4211, acc=0.8387
main.py:train_model: [Epoch 200 Batch 10000/17173] loss=0.4198, acc=0.8384
main.py:train_model: [Epoch 200 Batch 15000/17173] loss=0.4220, acc=0.8383
main.py:train_model: [Epoch 200] valid loss=0.3930, valid acc=0.8555, best valid acc=0.8562
main.py:train_model: [Epoch 201 Batch 5000/17173] loss=0.4188, acc=0.8388
main.py:train_model: [Epoch 201 Batch 10000/17173] loss=0.4212, acc=0.8375
main.py:train_model: [Epoch 201 Batch 15000/17173] loss=0.4197, acc=0.8374
main.py:train_model: [Epoch 201] valid loss=0.3932, valid acc=0.8538, best valid acc=0.8562
main.py:train_model: [Epoch 202 Batch 5000/17173] loss=0.4225, acc=0.8360
main.py:train_model: [Epoch 202 Batch 10000/17173] loss=0.4211, acc=0.8370
main.py:train_model: [Epoch 202 Batch 15000/17173] loss=0.4194, acc=0.8381
main.py:train_model: [Epoch 202] valid loss=0.3970, valid acc=0.8541, best valid acc=0.8562
main.py:train_model: [Epoch 203 Batch 5000/17173] loss=0.4195, acc=0.8385
main.py:train_model: [Epoch 203 Batch 10000/17173] loss=0.4211, acc=0.8377
main.py:train_model: [Epoch 203 Batch 15000/17173] loss=0.4192, acc=0.8388
main.py:train_model: [Epoch 203] valid loss=0.3961, valid acc=0.8540, best valid acc=0.8562
main.py:train_model: [Epoch 204 Batch 5000/17173] loss=0.4224, acc=0.8372
main.py:train_model: [Epoch 204 Batch 10000/17173] loss=0.4165, acc=0.8401
main.py:train_model: [Epoch 204 Batch 15000/17173] loss=0.4228, acc=0.8368
main.py:train_model: [Epoch 204] valid loss=0.3951, valid acc=0.8527, best valid acc=0.8562
main.py:train_model: [Epoch 205 Batch 5000/17173] loss=0.4216, acc=0.8377
main.py:train_model: [Epoch 205 Batch 10000/17173] loss=0.4207, acc=0.8377
main.py:train_model: [Epoch 205 Batch 15000/17173] loss=0.4174, acc=0.8388
main.py:train_model: [Epoch 205] valid loss=0.3953, valid acc=0.8555, best valid acc=0.8562
main.py:train_model: [Epoch 206 Batch 5000/17173] loss=0.4193, acc=0.8385
main.py:train_model: [Epoch 206 Batch 10000/17173] loss=0.4188, acc=0.8391
main.py:train_model: [Epoch 206 Batch 15000/17173] loss=0.4239, acc=0.8359
main.py:train_model: [Epoch 206] valid loss=0.3949, valid acc=0.8532, best valid acc=0.8562
main.py:train_model: [Epoch 207 Batch 5000/17173] loss=0.4199, acc=0.8377
main.py:train_model: [Epoch 207 Batch 10000/17173] loss=0.4205, acc=0.8380
main.py:train_model: [Epoch 207 Batch 15000/17173] loss=0.4206, acc=0.8381
main.py:train_model: [Epoch 207] valid loss=0.3919, valid acc=0.8548, best valid acc=0.8562
main.py:train_model: [Epoch 208 Batch 5000/17173] loss=0.4205, acc=0.8372
main.py:train_model: [Epoch 208 Batch 10000/17173] loss=0.4205, acc=0.8377
main.py:train_model: [Epoch 208 Batch 15000/17173] loss=0.4217, acc=0.8382
main.py:train_model: [Epoch 208] valid loss=0.3909, valid acc=0.8551, best valid acc=0.8562
main.py:train_model: [Epoch 209 Batch 5000/17173] loss=0.4198, acc=0.8379
main.py:train_model: [Epoch 209 Batch 10000/17173] loss=0.4182, acc=0.8393
main.py:train_model: [Epoch 209 Batch 15000/17173] loss=0.4200, acc=0.8383
main.py:train_model: [Epoch 209] valid loss=0.3930, valid acc=0.8545, best valid acc=0.8562
main.py:train_model: [Epoch 210 Batch 5000/17173] loss=0.4197, acc=0.8392
main.py:train_model: [Epoch 210 Batch 10000/17173] loss=0.4211, acc=0.8376
main.py:train_model: [Epoch 210 Batch 15000/17173] loss=0.4206, acc=0.8381
main.py:train_model: [Epoch 210] valid loss=0.3902, valid acc=0.8550, best valid acc=0.8562
main.py:train_model: [Epoch 211 Batch 5000/17173] loss=0.4201, acc=0.8379
main.py:train_model: [Epoch 211 Batch 10000/17173] loss=0.4221, acc=0.8371
main.py:train_model: [Epoch 211 Batch 15000/17173] loss=0.4177, acc=0.8390
main.py:train_model: [Epoch 211] valid loss=0.3923, valid acc=0.8526, best valid acc=0.8562
main.py:train_model: [Epoch 212 Batch 5000/17173] loss=0.4181, acc=0.8383
main.py:train_model: [Epoch 212 Batch 10000/17173] loss=0.4221, acc=0.8374
main.py:train_model: [Epoch 212 Batch 15000/17173] loss=0.4195, acc=0.8377
main.py:train_model: [Epoch 212] valid loss=0.3917, valid acc=0.8534, best valid acc=0.8562
main.py:train_model: [Epoch 213 Batch 5000/17173] loss=0.4166, acc=0.8399
main.py:train_model: [Epoch 213 Batch 10000/17173] loss=0.4189, acc=0.8385
main.py:train_model: [Epoch 213 Batch 15000/17173] loss=0.4194, acc=0.8388
main.py:train_model: [Epoch 213] valid loss=0.3905, valid acc=0.8546, best valid acc=0.8562
main.py:train_model: [Epoch 214 Batch 5000/17173] loss=0.4174, acc=0.8392
main.py:train_model: [Epoch 214 Batch 10000/17173] loss=0.4239, acc=0.8366
main.py:train_model: [Epoch 214 Batch 15000/17173] loss=0.4150, acc=0.8403
main.py:train_model: [Epoch 214] valid loss=0.3927, valid acc=0.8527, best valid acc=0.8562
main.py:train_model: [Epoch 215 Batch 5000/17173] loss=0.4210, acc=0.8376
main.py:train_model: [Epoch 215 Batch 10000/17173] loss=0.4146, acc=0.8411
main.py:train_model: [Epoch 215 Batch 15000/17173] loss=0.4200, acc=0.8375
main.py:train_model: [Epoch 215] valid loss=0.3880, valid acc=0.8569, best valid acc=0.8569
main.py:train_model: [Epoch 216 Batch 5000/17173] loss=0.4141, acc=0.8411
main.py:train_model: [Epoch 216 Batch 10000/17173] loss=0.4187, acc=0.8385
main.py:train_model: [Epoch 216 Batch 15000/17173] loss=0.4213, acc=0.8367
main.py:train_model: [Epoch 216] valid loss=0.3913, valid acc=0.8558, best valid acc=0.8569
main.py:train_model: [Epoch 217 Batch 5000/17173] loss=0.4162, acc=0.8394
main.py:train_model: [Epoch 217 Batch 10000/17173] loss=0.4151, acc=0.8393
main.py:train_model: [Epoch 217 Batch 15000/17173] loss=0.4215, acc=0.8380
main.py:train_model: [Epoch 217] valid loss=0.3937, valid acc=0.8534, best valid acc=0.8569
main.py:train_model: [Epoch 218 Batch 5000/17173] loss=0.4168, acc=0.8395
main.py:train_model: [Epoch 218 Batch 10000/17173] loss=0.4201, acc=0.8384
main.py:train_model: [Epoch 218 Batch 15000/17173] loss=0.4170, acc=0.8394
main.py:train_model: [Epoch 218] valid loss=0.3888, valid acc=0.8566, best valid acc=0.8569
main.py:train_model: [Epoch 219 Batch 5000/17173] loss=0.4169, acc=0.8392
main.py:train_model: [Epoch 219 Batch 10000/17173] loss=0.4196, acc=0.8383
main.py:train_model: [Epoch 219 Batch 15000/17173] loss=0.4184, acc=0.8381
main.py:train_model: [Epoch 219] valid loss=0.3943, valid acc=0.8545, best valid acc=0.8569
main.py:train_model: [Epoch 220 Batch 5000/17173] loss=0.4175, acc=0.8380
main.py:train_model: [Epoch 220 Batch 10000/17173] loss=0.4189, acc=0.8392
main.py:train_model: [Epoch 220 Batch 15000/17173] loss=0.4196, acc=0.8383
main.py:train_model: [Epoch 220] valid loss=0.3949, valid acc=0.8537, best valid acc=0.8569
main.py:train_model: [Epoch 221 Batch 5000/17173] loss=0.4174, acc=0.8393
main.py:train_model: [Epoch 221 Batch 10000/17173] loss=0.4191, acc=0.8380
main.py:train_model: [Epoch 221 Batch 15000/17173] loss=0.4169, acc=0.8390
main.py:train_model: [Epoch 221] valid loss=0.3928, valid acc=0.8531, best valid acc=0.8569
main.py:train_model: [Epoch 222 Batch 5000/17173] loss=0.4179, acc=0.8397
main.py:train_model: [Epoch 222 Batch 10000/17173] loss=0.4159, acc=0.8408
main.py:train_model: [Epoch 222 Batch 15000/17173] loss=0.4192, acc=0.8376
main.py:train_model: [Epoch 222] valid loss=0.3890, valid acc=0.8560, best valid acc=0.8569
main.py:train_model: [Epoch 223 Batch 5000/17173] loss=0.4163, acc=0.8395
main.py:train_model: [Epoch 223 Batch 10000/17173] loss=0.4202, acc=0.8373
main.py:train_model: [Epoch 223 Batch 15000/17173] loss=0.4190, acc=0.8383
main.py:train_model: [Epoch 223] valid loss=0.3907, valid acc=0.8545, best valid acc=0.8569
main.py:train_model: [Epoch 224 Batch 5000/17173] loss=0.4158, acc=0.8400
main.py:train_model: [Epoch 224 Batch 10000/17173] loss=0.4185, acc=0.8383
main.py:train_model: [Epoch 224 Batch 15000/17173] loss=0.4207, acc=0.8380
main.py:train_model: [Epoch 224] valid loss=0.3949, valid acc=0.8524, best valid acc=0.8569
main.py:train_model: [Epoch 225 Batch 5000/17173] loss=0.4171, acc=0.8378
main.py:train_model: [Epoch 225 Batch 10000/17173] loss=0.4174, acc=0.8388
main.py:train_model: [Epoch 225 Batch 15000/17173] loss=0.4180, acc=0.8396
main.py:train_model: [Epoch 225] valid loss=0.3932, valid acc=0.8541, best valid acc=0.8569
main.py:train_model: [Epoch 226 Batch 5000/17173] loss=0.4181, acc=0.8395
main.py:train_model: [Epoch 226 Batch 10000/17173] loss=0.4156, acc=0.8400
main.py:train_model: [Epoch 226 Batch 15000/17173] loss=0.4204, acc=0.8371
main.py:train_model: [Epoch 226] valid loss=0.3962, valid acc=0.8522, best valid acc=0.8569
main.py:train_model: [Epoch 227 Batch 5000/17173] loss=0.4158, acc=0.8399
main.py:train_model: [Epoch 227 Batch 10000/17173] loss=0.4178, acc=0.8388
main.py:train_model: [Epoch 227 Batch 15000/17173] loss=0.4184, acc=0.8384
main.py:train_model: [Epoch 227] valid loss=0.3900, valid acc=0.8535, best valid acc=0.8569
main.py:train_model: [Epoch 228 Batch 5000/17173] loss=0.4164, acc=0.8405
main.py:train_model: [Epoch 228 Batch 10000/17173] loss=0.4174, acc=0.8395
main.py:train_model: [Epoch 228 Batch 15000/17173] loss=0.4191, acc=0.8384
main.py:train_model: [Epoch 228] valid loss=0.3943, valid acc=0.8536, best valid acc=0.8569
main.py:train_model: [Epoch 229 Batch 5000/17173] loss=0.4141, acc=0.8408
main.py:train_model: [Epoch 229 Batch 10000/17173] loss=0.4181, acc=0.8389
main.py:train_model: [Epoch 229 Batch 15000/17173] loss=0.4182, acc=0.8390
main.py:train_model: [Epoch 229] valid loss=0.3901, valid acc=0.8555, best valid acc=0.8569
main.py:train_model: [Epoch 230 Batch 5000/17173] loss=0.4164, acc=0.8404
main.py:train_model: [Epoch 230 Batch 10000/17173] loss=0.4154, acc=0.8393
main.py:train_model: [Epoch 230 Batch 15000/17173] loss=0.4190, acc=0.8383
main.py:train_model: [Epoch 230] valid loss=0.3924, valid acc=0.8546, best valid acc=0.8569
main.py:train_model: [Epoch 231 Batch 5000/17173] loss=0.4175, acc=0.8396
main.py:train_model: [Epoch 231 Batch 10000/17173] loss=0.4180, acc=0.8386
main.py:train_model: [Epoch 231 Batch 15000/17173] loss=0.4169, acc=0.8393
main.py:train_model: [Epoch 231] valid loss=0.3893, valid acc=0.8554, best valid acc=0.8569
main.py:train_model: [Epoch 232 Batch 5000/17173] loss=0.4161, acc=0.8396
main.py:train_model: [Epoch 232 Batch 10000/17173] loss=0.4181, acc=0.8388
main.py:train_model: [Epoch 232 Batch 15000/17173] loss=0.4171, acc=0.8388
main.py:train_model: [Epoch 232] valid loss=0.3888, valid acc=0.8571, best valid acc=0.8571
main.py:train_model: [Epoch 233 Batch 5000/17173] loss=0.4141, acc=0.8403
main.py:train_model: [Epoch 233 Batch 10000/17173] loss=0.4145, acc=0.8400
main.py:train_model: [Epoch 233 Batch 15000/17173] loss=0.4212, acc=0.8377
main.py:train_model: [Epoch 233] valid loss=0.3912, valid acc=0.8545, best valid acc=0.8571
main.py:train_model: [Epoch 234 Batch 5000/17173] loss=0.4164, acc=0.8398
main.py:train_model: [Epoch 234 Batch 10000/17173] loss=0.4167, acc=0.8393
main.py:train_model: [Epoch 234 Batch 15000/17173] loss=0.4157, acc=0.8407
main.py:train_model: [Epoch 234] valid loss=0.3910, valid acc=0.8568, best valid acc=0.8571
main.py:train_model: [Epoch 235 Batch 5000/17173] loss=0.4162, acc=0.8395
main.py:train_model: [Epoch 235 Batch 10000/17173] loss=0.4174, acc=0.8396
main.py:train_model: [Epoch 235 Batch 15000/17173] loss=0.4162, acc=0.8400
main.py:train_model: [Epoch 235] valid loss=0.3968, valid acc=0.8536, best valid acc=0.8571
main.py:train_model: [Epoch 236 Batch 5000/17173] loss=0.4127, acc=0.8402
main.py:train_model: [Epoch 236 Batch 10000/17173] loss=0.4173, acc=0.8394
main.py:train_model: [Epoch 236 Batch 15000/17173] loss=0.4175, acc=0.8396
main.py:train_model: [Epoch 236] valid loss=0.3913, valid acc=0.8550, best valid acc=0.8571
main.py:train_model: [Epoch 237 Batch 5000/17173] loss=0.4159, acc=0.8391
main.py:train_model: [Epoch 237 Batch 10000/17173] loss=0.4178, acc=0.8388
main.py:train_model: [Epoch 237 Batch 15000/17173] loss=0.4151, acc=0.8403
main.py:train_model: [Epoch 237] valid loss=0.3924, valid acc=0.8549, best valid acc=0.8571
main.py:train_model: [Epoch 238 Batch 5000/17173] loss=0.4170, acc=0.8396
main.py:train_model: [Epoch 238 Batch 10000/17173] loss=0.4163, acc=0.8394
main.py:train_model: [Epoch 238 Batch 15000/17173] loss=0.4177, acc=0.8386
main.py:train_model: [Epoch 238] valid loss=0.3939, valid acc=0.8546, best valid acc=0.8571
main.py:train_model: [Epoch 239 Batch 5000/17173] loss=0.4141, acc=0.8400
main.py:train_model: [Epoch 239 Batch 10000/17173] loss=0.4183, acc=0.8389
main.py:train_model: [Epoch 239 Batch 15000/17173] loss=0.4152, acc=0.8396
main.py:train_model: [Epoch 239] valid loss=0.3953, valid acc=0.8529, best valid acc=0.8571
main.py:train_model: [Epoch 240 Batch 5000/17173] loss=0.4169, acc=0.8399
main.py:train_model: [Epoch 240 Batch 10000/17173] loss=0.4176, acc=0.8385
main.py:train_model: [Epoch 240 Batch 15000/17173] loss=0.4157, acc=0.8402
main.py:train_model: [Epoch 240] valid loss=0.3941, valid acc=0.8538, best valid acc=0.8571
main.py:train_model: [Epoch 241 Batch 5000/17173] loss=0.4158, acc=0.8398
main.py:train_model: [Epoch 241 Batch 10000/17173] loss=0.4171, acc=0.8398
main.py:train_model: [Epoch 241 Batch 15000/17173] loss=0.4179, acc=0.8386
main.py:train_model: [Epoch 241] valid loss=0.3947, valid acc=0.8529, best valid acc=0.8571
main.py:train_model: [Epoch 242 Batch 5000/17173] loss=0.4157, acc=0.8401
main.py:train_model: [Epoch 242 Batch 10000/17173] loss=0.4152, acc=0.8403
main.py:train_model: [Epoch 242 Batch 15000/17173] loss=0.4169, acc=0.8396
main.py:train_model: [Epoch 242] valid loss=0.3883, valid acc=0.8546, best valid acc=0.8571
main.py:train_model: [Epoch 243 Batch 5000/17173] loss=0.4156, acc=0.8389
main.py:train_model: [Epoch 243 Batch 10000/17173] loss=0.4125, acc=0.8411
main.py:train_model: [Epoch 243 Batch 15000/17173] loss=0.4162, acc=0.8407
main.py:train_model: [Epoch 243] valid loss=0.3972, valid acc=0.8503, best valid acc=0.8571
main.py:train_model: [Epoch 244 Batch 5000/17173] loss=0.4161, acc=0.8389
main.py:train_model: [Epoch 244 Batch 10000/17173] loss=0.4156, acc=0.8410
main.py:train_model: [Epoch 244 Batch 15000/17173] loss=0.4170, acc=0.8403
main.py:train_model: [Epoch 244] valid loss=0.3958, valid acc=0.8510, best valid acc=0.8571
main.py:train_model: [Epoch 245 Batch 5000/17173] loss=0.4154, acc=0.8400
main.py:train_model: [Epoch 245 Batch 10000/17173] loss=0.4139, acc=0.8420
main.py:train_model: [Epoch 245 Batch 15000/17173] loss=0.4173, acc=0.8387
main.py:train_model: [Epoch 245] valid loss=0.3950, valid acc=0.8529, best valid acc=0.8571
main.py:train_model: [Epoch 246 Batch 5000/17173] loss=0.4153, acc=0.8405
main.py:train_model: [Epoch 246 Batch 10000/17173] loss=0.4158, acc=0.8392
main.py:train_model: [Epoch 246 Batch 15000/17173] loss=0.4169, acc=0.8394
main.py:train_model: [Epoch 246] valid loss=0.3949, valid acc=0.8549, best valid acc=0.8571
main.py:train_model: [Epoch 247 Batch 5000/17173] loss=0.4158, acc=0.8396
main.py:train_model: [Epoch 247 Batch 10000/17173] loss=0.4161, acc=0.8391
main.py:train_model: [Epoch 247 Batch 15000/17173] loss=0.4150, acc=0.8401
main.py:train_model: [Epoch 247] valid loss=0.3953, valid acc=0.8536, best valid acc=0.8571
main.py:train_model: [Epoch 248 Batch 5000/17173] loss=0.4140, acc=0.8404
main.py:train_model: [Epoch 248 Batch 10000/17173] loss=0.4155, acc=0.8396
main.py:train_model: [Epoch 248 Batch 15000/17173] loss=0.4164, acc=0.8393
main.py:train_model: [Epoch 248] valid loss=0.3920, valid acc=0.8540, best valid acc=0.8571
main.py:train_model: [Epoch 249 Batch 5000/17173] loss=0.4158, acc=0.8399
main.py:train_model: [Epoch 249 Batch 10000/17173] loss=0.4149, acc=0.8398
main.py:train_model: [Epoch 249 Batch 15000/17173] loss=0.4164, acc=0.8385
main.py:train_model: [Epoch 249] valid loss=0.3929, valid acc=0.8532, best valid acc=0.8571
main.py:train_model: [Epoch 250 Batch 5000/17173] loss=0.4140, acc=0.8402
main.py:train_model: [Epoch 250 Batch 10000/17173] loss=0.4145, acc=0.8400
main.py:train_model: [Epoch 250 Batch 15000/17173] loss=0.4175, acc=0.8396
main.py:train_model: [Epoch 250] valid loss=0.3923, valid acc=0.8566, best valid acc=0.8571
main.py:train_model: [Epoch 251 Batch 5000/17173] loss=0.4173, acc=0.8384
main.py:train_model: [Epoch 251 Batch 10000/17173] loss=0.4119, acc=0.8409
main.py:train_model: [Epoch 251 Batch 15000/17173] loss=0.4156, acc=0.8398
main.py:train_model: [Epoch 251] valid loss=0.3914, valid acc=0.8544, best valid acc=0.8571
main.py:train_model: [Epoch 252 Batch 5000/17173] loss=0.4120, acc=0.8409
main.py:train_model: [Epoch 252 Batch 10000/17173] loss=0.4167, acc=0.8394
main.py:train_model: [Epoch 252 Batch 15000/17173] loss=0.4161, acc=0.8403
main.py:train_model: [Epoch 252] valid loss=0.3970, valid acc=0.8531, best valid acc=0.8571
main.py:train_model: [Epoch 253 Batch 5000/17173] loss=0.4122, acc=0.8409
main.py:train_model: [Epoch 253 Batch 10000/17173] loss=0.4167, acc=0.8399
main.py:train_model: [Epoch 253 Batch 15000/17173] loss=0.4148, acc=0.8404
main.py:train_model: [Epoch 253] valid loss=0.3929, valid acc=0.8542, best valid acc=0.8571
main.py:train_model: [Epoch 254 Batch 5000/17173] loss=0.4145, acc=0.8407
main.py:train_model: [Epoch 254 Batch 10000/17173] loss=0.4195, acc=0.8388
main.py:train_model: [Epoch 254 Batch 15000/17173] loss=0.4129, acc=0.8415
main.py:train_model: [Epoch 254] valid loss=0.3906, valid acc=0.8535, best valid acc=0.8571
main.py:train_model: [Epoch 255 Batch 5000/17173] loss=0.4134, acc=0.8412
main.py:train_model: [Epoch 255 Batch 10000/17173] loss=0.4150, acc=0.8392
main.py:train_model: [Epoch 255 Batch 15000/17173] loss=0.4165, acc=0.8390
main.py:train_model: [Epoch 255] valid loss=0.3917, valid acc=0.8538, best valid acc=0.8571
main.py:train_model: [Epoch 256 Batch 5000/17173] loss=0.4159, acc=0.8392
main.py:train_model: [Epoch 256 Batch 10000/17173] loss=0.4115, acc=0.8416
main.py:train_model: [Epoch 256 Batch 15000/17173] loss=0.4175, acc=0.8399
main.py:train_model: [Epoch 256] valid loss=0.3935, valid acc=0.8538, best valid acc=0.8571
main.py:train_model: [Epoch 257 Batch 5000/17173] loss=0.4142, acc=0.8412
main.py:train_model: [Epoch 257 Batch 10000/17173] loss=0.4137, acc=0.8407
main.py:train_model: [Epoch 257 Batch 15000/17173] loss=0.4151, acc=0.8404
main.py:train_model: [Epoch 257] valid loss=0.3955, valid acc=0.8525, best valid acc=0.8571
main.py:train_model: [Epoch 258 Batch 5000/17173] loss=0.4166, acc=0.8393
main.py:train_model: [Epoch 258 Batch 10000/17173] loss=0.4137, acc=0.8411
main.py:train_model: [Epoch 258 Batch 15000/17173] loss=0.4159, acc=0.8403
main.py:train_model: [Epoch 258] valid loss=0.3901, valid acc=0.8549, best valid acc=0.8571
main.py:train_model: [Epoch 259 Batch 5000/17173] loss=0.4133, acc=0.8407
main.py:train_model: [Epoch 259 Batch 10000/17173] loss=0.4174, acc=0.8391
main.py:train_model: [Epoch 259 Batch 15000/17173] loss=0.4149, acc=0.8405
main.py:train_model: [Epoch 259] valid loss=0.3933, valid acc=0.8540, best valid acc=0.8571
main.py:train_model: [Epoch 260 Batch 5000/17173] loss=0.4125, acc=0.8404
main.py:train_model: [Epoch 260 Batch 10000/17173] loss=0.4144, acc=0.8407
main.py:train_model: [Epoch 260 Batch 15000/17173] loss=0.4192, acc=0.8383
main.py:train_model: [Epoch 260] valid loss=0.3891, valid acc=0.8561, best valid acc=0.8571
main.py:train_model: [Epoch 261 Batch 5000/17173] loss=0.4147, acc=0.8410
main.py:train_model: [Epoch 261 Batch 10000/17173] loss=0.4156, acc=0.8397
main.py:train_model: [Epoch 261 Batch 15000/17173] loss=0.4146, acc=0.8406
main.py:train_model: [Epoch 261] valid loss=0.3948, valid acc=0.8530, best valid acc=0.8571
main.py:train_model: [Epoch 262 Batch 5000/17173] loss=0.4149, acc=0.8411
main.py:train_model: [Epoch 262 Batch 10000/17173] loss=0.4136, acc=0.8399
main.py:train_model: [Epoch 262 Batch 15000/17173] loss=0.4160, acc=0.8405
main.py:train_model: [Epoch 262] valid loss=0.3947, valid acc=0.8535, best valid acc=0.8571
main.py:train_model: [Epoch 263 Batch 5000/17173] loss=0.4148, acc=0.8396
main.py:train_model: [Epoch 263 Batch 10000/17173] loss=0.4134, acc=0.8412
main.py:train_model: [Epoch 263 Batch 15000/17173] loss=0.4136, acc=0.8403
main.py:train_model: [Epoch 263] valid loss=0.3909, valid acc=0.8532, best valid acc=0.8571
main.py:train_model: [Epoch 264 Batch 5000/17173] loss=0.4126, acc=0.8420
main.py:train_model: [Epoch 264 Batch 10000/17173] loss=0.4164, acc=0.8402
main.py:train_model: [Epoch 264 Batch 15000/17173] loss=0.4131, acc=0.8413
main.py:train_model: [Epoch 264] valid loss=0.3886, valid acc=0.8540, best valid acc=0.8571
main.py:train_model: [Epoch 265 Batch 5000/17173] loss=0.4124, acc=0.8410
main.py:train_model: [Epoch 265 Batch 10000/17173] loss=0.4126, acc=0.8417
main.py:train_model: [Epoch 265 Batch 15000/17173] loss=0.4146, acc=0.8396
main.py:train_model: [Epoch 265] valid loss=0.3912, valid acc=0.8541, best valid acc=0.8571
main.py:train_model: [Epoch 266 Batch 5000/17173] loss=0.4146, acc=0.8398
main.py:train_model: [Epoch 266 Batch 10000/17173] loss=0.4136, acc=0.8403
main.py:train_model: [Epoch 266 Batch 15000/17173] loss=0.4140, acc=0.8402
main.py:train_model: [Epoch 266] valid loss=0.3963, valid acc=0.8518, best valid acc=0.8571
main.py:train_model: [Epoch 267 Batch 5000/17173] loss=0.4126, acc=0.8411
main.py:train_model: [Epoch 267 Batch 10000/17173] loss=0.4143, acc=0.8397
main.py:train_model: [Epoch 267 Batch 15000/17173] loss=0.4164, acc=0.8402
main.py:train_model: [Epoch 267] valid loss=0.3940, valid acc=0.8538, best valid acc=0.8571
main.py:train_model: [Epoch 268 Batch 5000/17173] loss=0.4175, acc=0.8389
main.py:train_model: [Epoch 268 Batch 10000/17173] loss=0.4125, acc=0.8418
main.py:train_model: [Epoch 268 Batch 15000/17173] loss=0.4114, acc=0.8412
main.py:train_model: [Epoch 268] valid loss=0.3890, valid acc=0.8546, best valid acc=0.8571
main.py:train_model: [Epoch 269 Batch 5000/17173] loss=0.4111, acc=0.8412
main.py:train_model: [Epoch 269 Batch 10000/17173] loss=0.4169, acc=0.8395
main.py:train_model: [Epoch 269 Batch 15000/17173] loss=0.4143, acc=0.8413
main.py:train_model: [Epoch 269] valid loss=0.3936, valid acc=0.8533, best valid acc=0.8571
main.py:train_model: [Epoch 270 Batch 5000/17173] loss=0.4151, acc=0.8400
main.py:train_model: [Epoch 270 Batch 10000/17173] loss=0.4135, acc=0.8404
main.py:train_model: [Epoch 270 Batch 15000/17173] loss=0.4125, acc=0.8417
main.py:train_model: [Epoch 270] valid loss=0.3940, valid acc=0.8536, best valid acc=0.8571
main.py:train_model: [Epoch 271 Batch 5000/17173] loss=0.4143, acc=0.8408
main.py:train_model: [Epoch 271 Batch 10000/17173] loss=0.4148, acc=0.8409
main.py:train_model: [Epoch 271 Batch 15000/17173] loss=0.4138, acc=0.8404
main.py:train_model: [Epoch 271] valid loss=0.3891, valid acc=0.8566, best valid acc=0.8571
main.py:train_model: [Epoch 272 Batch 5000/17173] loss=0.4116, acc=0.8412
main.py:train_model: [Epoch 272 Batch 10000/17173] loss=0.4152, acc=0.8400
main.py:train_model: [Epoch 272 Batch 15000/17173] loss=0.4141, acc=0.8397
main.py:train_model: [Epoch 272] valid loss=0.3937, valid acc=0.8562, best valid acc=0.8571
main.py:train_model: [Epoch 273 Batch 5000/17173] loss=0.4128, acc=0.8402
main.py:train_model: [Epoch 273 Batch 10000/17173] loss=0.4129, acc=0.8410
main.py:train_model: [Epoch 273 Batch 15000/17173] loss=0.4145, acc=0.8408
main.py:train_model: [Epoch 273] valid loss=0.3928, valid acc=0.8524, best valid acc=0.8571
main.py:train_model: [Epoch 274 Batch 5000/17173] loss=0.4163, acc=0.8400
main.py:train_model: [Epoch 274 Batch 10000/17173] loss=0.4118, acc=0.8417
main.py:train_model: [Epoch 274 Batch 15000/17173] loss=0.4116, acc=0.8417
main.py:train_model: [Epoch 274] valid loss=0.3962, valid acc=0.8527, best valid acc=0.8571
main.py:train_model: [Epoch 275 Batch 5000/17173] loss=0.4143, acc=0.8395
main.py:train_model: [Epoch 275 Batch 10000/17173] loss=0.4134, acc=0.8410
main.py:train_model: [Epoch 275 Batch 15000/17173] loss=0.4141, acc=0.8407
main.py:train_model: [Epoch 275] valid loss=0.3898, valid acc=0.8548, best valid acc=0.8571
main.py:train_model: [Epoch 276 Batch 5000/17173] loss=0.4136, acc=0.8396
main.py:train_model: [Epoch 276 Batch 10000/17173] loss=0.4113, acc=0.8419
main.py:train_model: [Epoch 276 Batch 15000/17173] loss=0.4166, acc=0.8391
main.py:train_model: [Epoch 276] valid loss=0.3895, valid acc=0.8558, best valid acc=0.8571
main.py:train_model: [Epoch 277 Batch 5000/17173] loss=0.4131, acc=0.8420
main.py:train_model: [Epoch 277 Batch 10000/17173] loss=0.4124, acc=0.8404
main.py:train_model: [Epoch 277 Batch 15000/17173] loss=0.4118, acc=0.8420
main.py:train_model: [Epoch 277] valid loss=0.3913, valid acc=0.8541, best valid acc=0.8571
main.py:train_model: [Epoch 278 Batch 5000/17173] loss=0.4160, acc=0.8406
main.py:train_model: [Epoch 278 Batch 10000/17173] loss=0.4124, acc=0.8415
main.py:train_model: [Epoch 278 Batch 15000/17173] loss=0.4126, acc=0.8414
main.py:train_model: [Epoch 278] valid loss=0.3937, valid acc=0.8536, best valid acc=0.8571
main.py:train_model: [Epoch 279 Batch 5000/17173] loss=0.4121, acc=0.8411
main.py:train_model: [Epoch 279 Batch 10000/17173] loss=0.4121, acc=0.8420
main.py:train_model: [Epoch 279 Batch 15000/17173] loss=0.4132, acc=0.8411
main.py:train_model: [Epoch 279] valid loss=0.3934, valid acc=0.8544, best valid acc=0.8571
main.py:train_model: [Epoch 280 Batch 5000/17173] loss=0.4107, acc=0.8421
main.py:train_model: [Epoch 280 Batch 10000/17173] loss=0.4139, acc=0.8415
main.py:train_model: [Epoch 280 Batch 15000/17173] loss=0.4113, acc=0.8416
main.py:train_model: [Epoch 280] valid loss=0.3925, valid acc=0.8535, best valid acc=0.8571
main.py:train_model: [Epoch 281 Batch 5000/17173] loss=0.4129, acc=0.8407
main.py:train_model: [Epoch 281 Batch 10000/17173] loss=0.4123, acc=0.8412
main.py:train_model: [Epoch 281 Batch 15000/17173] loss=0.4146, acc=0.8403
main.py:train_model: [Epoch 281] valid loss=0.3926, valid acc=0.8544, best valid acc=0.8571
main.py:train_model: [Epoch 282 Batch 5000/17173] loss=0.4118, acc=0.8419
main.py:train_model: [Epoch 282 Batch 10000/17173] loss=0.4152, acc=0.8393
main.py:train_model: [Epoch 282 Batch 15000/17173] loss=0.4091, acc=0.8432
main.py:train_model: [Epoch 282] valid loss=0.3933, valid acc=0.8544, best valid acc=0.8571
main.py:train_model: [Epoch 283 Batch 5000/17173] loss=0.4128, acc=0.8412
main.py:train_model: [Epoch 283 Batch 10000/17173] loss=0.4122, acc=0.8421
main.py:train_model: [Epoch 283 Batch 15000/17173] loss=0.4152, acc=0.8399
main.py:train_model: [Epoch 283] valid loss=0.3937, valid acc=0.8553, best valid acc=0.8571
main.py:train_model: [Epoch 284 Batch 5000/17173] loss=0.4107, acc=0.8424
main.py:train_model: [Epoch 284 Batch 10000/17173] loss=0.4123, acc=0.8409
main.py:train_model: [Epoch 284 Batch 15000/17173] loss=0.4155, acc=0.8401
main.py:train_model: [Epoch 284] valid loss=0.3893, valid acc=0.8547, best valid acc=0.8571
main.py:train_model: [Epoch 285 Batch 5000/17173] loss=0.4116, acc=0.8416
main.py:train_model: [Epoch 285 Batch 10000/17173] loss=0.4139, acc=0.8401
main.py:train_model: [Epoch 285 Batch 15000/17173] loss=0.4126, acc=0.8415
main.py:train_model: [Epoch 285] valid loss=0.3903, valid acc=0.8540, best valid acc=0.8571
main.py:train_model: [Epoch 286 Batch 5000/17173] loss=0.4130, acc=0.8410
main.py:train_model: [Epoch 286 Batch 10000/17173] loss=0.4122, acc=0.8415
main.py:train_model: [Epoch 286 Batch 15000/17173] loss=0.4126, acc=0.8413
main.py:train_model: [Epoch 286] valid loss=0.3897, valid acc=0.8572, best valid acc=0.8572
main.py:train_model: [Epoch 287 Batch 5000/17173] loss=0.4146, acc=0.8401
main.py:train_model: [Epoch 287 Batch 10000/17173] loss=0.4119, acc=0.8411
main.py:train_model: [Epoch 287 Batch 15000/17173] loss=0.4111, acc=0.8414
main.py:train_model: [Epoch 287] valid loss=0.3928, valid acc=0.8545, best valid acc=0.8572
main.py:train_model: [Epoch 288 Batch 5000/17173] loss=0.4146, acc=0.8412
main.py:train_model: [Epoch 288 Batch 10000/17173] loss=0.4091, acc=0.8427
main.py:train_model: [Epoch 288 Batch 15000/17173] loss=0.4140, acc=0.8407
main.py:train_model: [Epoch 288] valid loss=0.3911, valid acc=0.8534, best valid acc=0.8572
main.py:train_model: [Epoch 289 Batch 5000/17173] loss=0.4147, acc=0.8397
main.py:train_model: [Epoch 289 Batch 10000/17173] loss=0.4129, acc=0.8411
main.py:train_model: [Epoch 289 Batch 15000/17173] loss=0.4103, acc=0.8429
main.py:train_model: [Epoch 289] valid loss=0.3865, valid acc=0.8559, best valid acc=0.8572
main.py:train_model: [Epoch 290 Batch 5000/17173] loss=0.4119, acc=0.8404
main.py:train_model: [Epoch 290 Batch 10000/17173] loss=0.4114, acc=0.8425
main.py:train_model: [Epoch 290 Batch 15000/17173] loss=0.4124, acc=0.8406
main.py:train_model: [Epoch 290] valid loss=0.3925, valid acc=0.8538, best valid acc=0.8572
main.py:train_model: [Epoch 291 Batch 5000/17173] loss=0.4107, acc=0.8416
main.py:train_model: [Epoch 291 Batch 10000/17173] loss=0.4135, acc=0.8409
main.py:train_model: [Epoch 291 Batch 15000/17173] loss=0.4123, acc=0.8404
main.py:train_model: [Epoch 291] valid loss=0.3866, valid acc=0.8561, best valid acc=0.8572
main.py:train_model: [Epoch 292 Batch 5000/17173] loss=0.4134, acc=0.8418
main.py:train_model: [Epoch 292 Batch 10000/17173] loss=0.4150, acc=0.8406
main.py:train_model: [Epoch 292 Batch 15000/17173] loss=0.4125, acc=0.8406
main.py:train_model: [Epoch 292] valid loss=0.3938, valid acc=0.8532, best valid acc=0.8572
main.py:train_model: [Epoch 293 Batch 5000/17173] loss=0.4132, acc=0.8411
main.py:train_model: [Epoch 293 Batch 10000/17173] loss=0.4123, acc=0.8414
main.py:train_model: [Epoch 293 Batch 15000/17173] loss=0.4111, acc=0.8424
main.py:train_model: [Epoch 293] valid loss=0.3921, valid acc=0.8539, best valid acc=0.8572
main.py:train_model: [Epoch 294 Batch 5000/17173] loss=0.4084, acc=0.8430
main.py:train_model: [Epoch 294 Batch 10000/17173] loss=0.4113, acc=0.8428
main.py:train_model: [Epoch 294 Batch 15000/17173] loss=0.4149, acc=0.8406
main.py:train_model: [Epoch 294] valid loss=0.3883, valid acc=0.8554, best valid acc=0.8572
main.py:train_model: [Epoch 295 Batch 5000/17173] loss=0.4136, acc=0.8401
main.py:train_model: [Epoch 295 Batch 10000/17173] loss=0.4124, acc=0.8417
main.py:train_model: [Epoch 295 Batch 15000/17173] loss=0.4090, acc=0.8437
main.py:train_model: [Epoch 295] valid loss=0.3877, valid acc=0.8562, best valid acc=0.8572
main.py:train_model: [Epoch 296 Batch 5000/17173] loss=0.4080, acc=0.8436
main.py:train_model: [Epoch 296 Batch 10000/17173] loss=0.4109, acc=0.8424
main.py:train_model: [Epoch 296 Batch 15000/17173] loss=0.4162, acc=0.8396
main.py:train_model: [Epoch 296] valid loss=0.3920, valid acc=0.8540, best valid acc=0.8572
main.py:train_model: [Epoch 297 Batch 5000/17173] loss=0.4118, acc=0.8413
main.py:train_model: [Epoch 297 Batch 10000/17173] loss=0.4102, acc=0.8423
main.py:train_model: [Epoch 297 Batch 15000/17173] loss=0.4121, acc=0.8416
main.py:train_model: [Epoch 297] valid loss=0.3905, valid acc=0.8558, best valid acc=0.8572
main.py:train_model: [Epoch 298 Batch 5000/17173] loss=0.4095, acc=0.8426
main.py:train_model: [Epoch 298 Batch 10000/17173] loss=0.4101, acc=0.8420
main.py:train_model: [Epoch 298 Batch 15000/17173] loss=0.4131, acc=0.8412
main.py:train_model: [Epoch 298] valid loss=0.3917, valid acc=0.8543, best valid acc=0.8572
main.py:train_model: [Epoch 299 Batch 5000/17173] loss=0.4107, acc=0.8412
main.py:train_model: [Epoch 299 Batch 10000/17173] loss=0.4134, acc=0.8410
main.py:train_model: [Epoch 299 Batch 15000/17173] loss=0.4124, acc=0.8421
main.py:train_model: [Epoch 299] valid loss=0.3904, valid acc=0.8540, best valid acc=0.8572
You can’t perform that action at this time.