Skip to content

Instantly share code, notes, and snippets.

@gangliao

gangliao/ASVspoof2017-CNN-Training

Created April 17, 2019 12:16
Show Gist options
  • Save gangliao/9a388fb74de769b9f8ced46a31898c5c to your computer and use it in GitHub Desktop.
Save gangliao/9a388fb74de769b9f8ced46a31898c5c to your computer and use it in GitHub Desktop.
Download ZIP
Raw
ASVspoof2017-CNN-Training
Parameters: {'no_cuda': True, 'train': True, 'test_batch_size': 10, 'batch_size': 10, 'epochs': 41, 'seed': 1, 'lr': 0.0001, 'log_interval': 10, 'eval': True, 'cuda': False, 'momentum': 0.5, 'output_dir': 'model-output/5cnn-2fc-mfm-768x400'}
Conv1: (32, 1, 3, 3)
Conv2: (48, 16, 3, 3)
Conv3: (64, 24, 3, 3)
Fc1: (128, 1512)
Fc2: (2, 128)
Train Epoch: 1 [0/3016 (0%)] Loss: 0.679156
Train Epoch: 1 [100/3016 (3%)] Loss: 0.690277
Train Epoch: 1 [200/3016 (7%)] Loss: 0.671204
Train Epoch: 1 [300/3016 (10%)] Loss: 0.726018
Train Epoch: 1 [400/3016 (13%)] Loss: 0.701729
Train Epoch: 1 [500/3016 (17%)] Loss: 0.671664
Train Epoch: 1 [600/3016 (20%)] Loss: 0.719431
Train Epoch: 1 [700/3016 (23%)] Loss: 0.724697
Train Epoch: 1 [800/3016 (26%)] Loss: 0.697546
Train Epoch: 1 [900/3016 (30%)] Loss: 0.687959
Train Epoch: 1 [1000/3016 (33%)] Loss: 0.672536
Train Epoch: 1 [1100/3016 (36%)] Loss: 0.703507
Train Epoch: 1 [1200/3016 (40%)] Loss: 0.684229
Train Epoch: 1 [1300/3016 (43%)] Loss: 0.721094
Train Epoch: 1 [1400/3016 (46%)] Loss: 0.679275
Train Epoch: 1 [1500/3016 (50%)] Loss: 0.675661
Train Epoch: 1 [1600/3016 (53%)] Loss: 0.721359
Train Epoch: 1 [1700/3016 (56%)] Loss: 0.725118
Train Epoch: 1 [1800/3016 (60%)] Loss: 0.737032
Train Epoch: 1 [1900/3016 (63%)] Loss: 0.693116
Train Epoch: 1 [2000/3016 (66%)] Loss: 0.706081
Train Epoch: 1 [2100/3016 (70%)] Loss: 0.705836
Train Epoch: 1 [2200/3016 (73%)] Loss: 0.699134
Train Epoch: 1 [2300/3016 (76%)] Loss: 0.725073
Train Epoch: 1 [2400/3016 (79%)] Loss: 0.698008
Train Epoch: 1 [2500/3016 (83%)] Loss: 0.681030
Train Epoch: 1 [2600/3016 (86%)] Loss: 0.710976
Train Epoch: 1 [2700/3016 (89%)] Loss: 0.695863
Train Epoch: 1 [2800/3016 (93%)] Loss: 0.678028
Train Epoch: 1 [2900/3016 (96%)] Loss: 0.717460
Train Epoch: 1 [3000/3016 (99%)] Loss: 0.664573
Total loss = 0.069742
/usr/local/lib/python2.7/site-packages/torch/nn/_reduction.py:49: UserWarning: size_average and reduce args will be deprecated, please use reduction='sum' instead.
warnings.warn(warning.format(ret))
Dev loss is 0.699042
Train Epoch: 2 [0/3016 (0%)] Loss: 0.699921
Train Epoch: 2 [100/3016 (3%)] Loss: 0.676137
Train Epoch: 2 [200/3016 (7%)] Loss: 0.679473
Train Epoch: 2 [300/3016 (10%)] Loss: 0.697464
Train Epoch: 2 [400/3016 (13%)] Loss: 0.660784
Train Epoch: 2 [500/3016 (17%)] Loss: 0.692990
Train Epoch: 2 [600/3016 (20%)] Loss: 0.725979
Train Epoch: 2 [700/3016 (23%)] Loss: 0.686657
Train Epoch: 2 [800/3016 (26%)] Loss: 0.696012
Train Epoch: 2 [900/3016 (30%)] Loss: 0.686633
Train Epoch: 2 [1000/3016 (33%)] Loss: 0.684405
Train Epoch: 2 [1100/3016 (36%)] Loss: 0.685548
Train Epoch: 2 [1200/3016 (40%)] Loss: 0.679128
Train Epoch: 2 [1300/3016 (43%)] Loss: 0.659065
Train Epoch: 2 [1400/3016 (46%)] Loss: 0.710440
Train Epoch: 2 [1500/3016 (50%)] Loss: 0.705534
Train Epoch: 2 [1600/3016 (53%)] Loss: 0.717113
Train Epoch: 2 [1700/3016 (56%)] Loss: 0.699242
Train Epoch: 2 [1800/3016 (60%)] Loss: 0.696675
Train Epoch: 2 [1900/3016 (63%)] Loss: 0.702990
Train Epoch: 2 [2000/3016 (66%)] Loss: 0.710766
Train Epoch: 2 [2100/3016 (70%)] Loss: 0.677398
Train Epoch: 2 [2200/3016 (73%)] Loss: 0.694263
Train Epoch: 2 [2300/3016 (76%)] Loss: 0.712979
Train Epoch: 2 [2400/3016 (79%)] Loss: 0.670982
Train Epoch: 2 [2500/3016 (83%)] Loss: 0.706408
Train Epoch: 2 [2600/3016 (86%)] Loss: 0.621356
Train Epoch: 2 [2700/3016 (89%)] Loss: 0.653619
Train Epoch: 2 [2800/3016 (93%)] Loss: 0.678805
Train Epoch: 2 [2900/3016 (96%)] Loss: 0.667799
Train Epoch: 2 [3000/3016 (99%)] Loss: 0.644149
Total loss = 0.069009
Dev loss is 0.670288
Train Epoch: 3 [0/3016 (0%)] Loss: 0.655485
Train Epoch: 3 [100/3016 (3%)] Loss: 0.634798
Train Epoch: 3 [200/3016 (7%)] Loss: 0.594128
Train Epoch: 3 [300/3016 (10%)] Loss: 0.620971
Train Epoch: 3 [400/3016 (13%)] Loss: 0.637233
Train Epoch: 3 [500/3016 (17%)] Loss: 0.478179
Train Epoch: 3 [600/3016 (20%)] Loss: 0.478510
Train Epoch: 3 [700/3016 (23%)] Loss: 0.604270
Train Epoch: 3 [800/3016 (26%)] Loss: 0.514479
Train Epoch: 3 [900/3016 (30%)] Loss: 0.627764
Train Epoch: 3 [1000/3016 (33%)] Loss: 0.387671
Train Epoch: 3 [1100/3016 (36%)] Loss: 0.482797
Train Epoch: 3 [1200/3016 (40%)] Loss: 0.422659
Train Epoch: 3 [1300/3016 (43%)] Loss: 0.792615
Train Epoch: 3 [1400/3016 (46%)] Loss: 0.756887
Train Epoch: 3 [1500/3016 (50%)] Loss: 0.576640
Train Epoch: 3 [1600/3016 (53%)] Loss: 0.561137
Train Epoch: 3 [1700/3016 (56%)] Loss: 0.475662
Train Epoch: 3 [1800/3016 (60%)] Loss: 0.275708
Train Epoch: 3 [1900/3016 (63%)] Loss: 0.542504
Train Epoch: 3 [2000/3016 (66%)] Loss: 0.342978
Train Epoch: 3 [2100/3016 (70%)] Loss: 0.182694
Train Epoch: 3 [2200/3016 (73%)] Loss: 0.199350
Train Epoch: 3 [2300/3016 (76%)] Loss: 0.088886
Train Epoch: 3 [2400/3016 (79%)] Loss: 0.043269
Train Epoch: 3 [2500/3016 (83%)] Loss: 0.230999
Train Epoch: 3 [2600/3016 (86%)] Loss: 0.768863
Train Epoch: 3 [2700/3016 (89%)] Loss: 0.353805
Train Epoch: 3 [2800/3016 (93%)] Loss: 0.297991
Train Epoch: 3 [2900/3016 (96%)] Loss: 0.079424
Train Epoch: 3 [3000/3016 (99%)] Loss: 0.216512
Total loss = 0.042735
Dev loss is 0.444525
Train Epoch: 4 [0/3016 (0%)] Loss: 0.260617
Train Epoch: 4 [100/3016 (3%)] Loss: 0.300901
Train Epoch: 4 [200/3016 (7%)] Loss: 0.117927
Train Epoch: 4 [300/3016 (10%)] Loss: 0.265370
Train Epoch: 4 [400/3016 (13%)] Loss: 0.351818
Train Epoch: 4 [500/3016 (17%)] Loss: 0.127038
Train Epoch: 4 [600/3016 (20%)] Loss: 0.142760
Train Epoch: 4 [700/3016 (23%)] Loss: 0.201735
Train Epoch: 4 [800/3016 (26%)] Loss: 0.158253
Train Epoch: 4 [900/3016 (30%)] Loss: 0.566348
Train Epoch: 4 [1000/3016 (33%)] Loss: 0.325857
Train Epoch: 4 [1100/3016 (36%)] Loss: 0.033522
Train Epoch: 4 [1200/3016 (40%)] Loss: 0.139091
Train Epoch: 4 [1300/3016 (43%)] Loss: 0.294740
Train Epoch: 4 [1400/3016 (46%)] Loss: 0.235377
Train Epoch: 4 [1500/3016 (50%)] Loss: 0.045502
Train Epoch: 4 [1600/3016 (53%)] Loss: 0.338375
Train Epoch: 4 [1700/3016 (56%)] Loss: 0.581512
Train Epoch: 4 [1800/3016 (60%)] Loss: 0.340500
Train Epoch: 4 [1900/3016 (63%)] Loss: 0.257161
Train Epoch: 4 [2000/3016 (66%)] Loss: 0.089565
Train Epoch: 4 [2100/3016 (70%)] Loss: 0.432748
Train Epoch: 4 [2200/3016 (73%)] Loss: 0.129520
Train Epoch: 4 [2300/3016 (76%)] Loss: 0.061777
Train Epoch: 4 [2400/3016 (79%)] Loss: 0.576540
Train Epoch: 4 [2500/3016 (83%)] Loss: 0.600931
Train Epoch: 4 [2600/3016 (86%)] Loss: 0.040597
Train Epoch: 4 [2700/3016 (89%)] Loss: 0.210642
Train Epoch: 4 [2800/3016 (93%)] Loss: 0.226367
Train Epoch: 4 [2900/3016 (96%)] Loss: 0.104761
Train Epoch: 4 [3000/3016 (99%)] Loss: 0.148170
Total loss = 0.022532
Dev loss is 0.430629
Train Epoch: 5 [0/3016 (0%)] Loss: 0.378802
Train Epoch: 5 [100/3016 (3%)] Loss: 0.082958
Train Epoch: 5 [200/3016 (7%)] Loss: 0.116265
Train Epoch: 5 [300/3016 (10%)] Loss: 0.216133
Train Epoch: 5 [400/3016 (13%)] Loss: 0.047992
Train Epoch: 5 [500/3016 (17%)] Loss: 0.114575
Train Epoch: 5 [600/3016 (20%)] Loss: 0.095891
Train Epoch: 5 [700/3016 (23%)] Loss: 0.021858
Train Epoch: 5 [800/3016 (26%)] Loss: 0.015020
Train Epoch: 5 [900/3016 (30%)] Loss: 0.038211
Train Epoch: 5 [1000/3016 (33%)] Loss: 0.129926
Train Epoch: 5 [1100/3016 (36%)] Loss: 0.033511
Train Epoch: 5 [1200/3016 (40%)] Loss: 0.041682
Train Epoch: 5 [1300/3016 (43%)] Loss: 0.789258
Train Epoch: 5 [1400/3016 (46%)] Loss: 0.025230
Train Epoch: 5 [1500/3016 (50%)] Loss: 0.132880
Train Epoch: 5 [1600/3016 (53%)] Loss: 0.139767
Train Epoch: 5 [1700/3016 (56%)] Loss: 0.474614
Train Epoch: 5 [1800/3016 (60%)] Loss: 0.142562
Train Epoch: 5 [1900/3016 (63%)] Loss: 0.238532
Train Epoch: 5 [2000/3016 (66%)] Loss: 0.357306
Train Epoch: 5 [2100/3016 (70%)] Loss: 0.101770
Train Epoch: 5 [2200/3016 (73%)] Loss: 0.122533
Train Epoch: 5 [2300/3016 (76%)] Loss: 0.033133
Train Epoch: 5 [2400/3016 (79%)] Loss: 0.512635
Train Epoch: 5 [2500/3016 (83%)] Loss: 0.070486
Train Epoch: 5 [2600/3016 (86%)] Loss: 0.073064
Train Epoch: 5 [2700/3016 (89%)] Loss: 0.113680
Train Epoch: 5 [2800/3016 (93%)] Loss: 0.075620
Train Epoch: 5 [2900/3016 (96%)] Loss: 0.504819
Train Epoch: 5 [3000/3016 (99%)] Loss: 0.050784
Total loss = 0.015950
Dev loss is 0.392736
Train Epoch: 6 [0/3016 (0%)] Loss: 0.176268
Train Epoch: 6 [100/3016 (3%)] Loss: 0.189820
Train Epoch: 6 [200/3016 (7%)] Loss: 0.029553
Train Epoch: 6 [300/3016 (10%)] Loss: 0.056507
Train Epoch: 6 [400/3016 (13%)] Loss: 0.054776
Train Epoch: 6 [500/3016 (17%)] Loss: 0.512563
Train Epoch: 6 [600/3016 (20%)] Loss: 0.076084
Train Epoch: 6 [700/3016 (23%)] Loss: 0.051120
Train Epoch: 6 [800/3016 (26%)] Loss: 0.237902
Train Epoch: 6 [900/3016 (30%)] Loss: 0.069138
Train Epoch: 6 [1000/3016 (33%)] Loss: 0.144652
Train Epoch: 6 [1100/3016 (36%)] Loss: 0.355421
Train Epoch: 6 [1200/3016 (40%)] Loss: 0.308849
Train Epoch: 6 [1300/3016 (43%)] Loss: 0.100481
Train Epoch: 6 [1400/3016 (46%)] Loss: 0.255426
Train Epoch: 6 [1500/3016 (50%)] Loss: 0.015510
Train Epoch: 6 [1600/3016 (53%)] Loss: 0.056703
Train Epoch: 6 [1700/3016 (56%)] Loss: 0.551097
Train Epoch: 6 [1800/3016 (60%)] Loss: 0.064695
Train Epoch: 6 [1900/3016 (63%)] Loss: 0.026484
Train Epoch: 6 [2000/3016 (66%)] Loss: 0.010729
Train Epoch: 6 [2100/3016 (70%)] Loss: 0.258938
Train Epoch: 6 [2200/3016 (73%)] Loss: 0.063525
Train Epoch: 6 [2300/3016 (76%)] Loss: 0.149133
Train Epoch: 6 [2400/3016 (79%)] Loss: 0.036696
Train Epoch: 6 [2500/3016 (83%)] Loss: 0.007787
Train Epoch: 6 [2600/3016 (86%)] Loss: 0.088395
Train Epoch: 6 [2700/3016 (89%)] Loss: 0.047400
Train Epoch: 6 [2800/3016 (93%)] Loss: 0.040817
Train Epoch: 6 [2900/3016 (96%)] Loss: 0.114001
Train Epoch: 6 [3000/3016 (99%)] Loss: 0.256422
Total loss = 0.013483
Dev loss is 0.406156
Train Epoch: 7 [0/3016 (0%)] Loss: 0.096361
Train Epoch: 7 [100/3016 (3%)] Loss: 0.016592
Train Epoch: 7 [200/3016 (7%)] Loss: 0.040744
Train Epoch: 7 [300/3016 (10%)] Loss: 0.191493
Train Epoch: 7 [400/3016 (13%)] Loss: 0.151431
Train Epoch: 7 [500/3016 (17%)] Loss: 0.122236
Train Epoch: 7 [600/3016 (20%)] Loss: 0.030362
Train Epoch: 7 [700/3016 (23%)] Loss: 0.068052
Train Epoch: 7 [800/3016 (26%)] Loss: 0.168277
Train Epoch: 7 [900/3016 (30%)] Loss: 0.390652
Train Epoch: 7 [1000/3016 (33%)] Loss: 0.282669
Train Epoch: 7 [1100/3016 (36%)] Loss: 0.032958
Train Epoch: 7 [1200/3016 (40%)] Loss: 0.046842
Train Epoch: 7 [1300/3016 (43%)] Loss: 0.104486
Train Epoch: 7 [1400/3016 (46%)] Loss: 0.021333
Train Epoch: 7 [1500/3016 (50%)] Loss: 0.215297
Train Epoch: 7 [1600/3016 (53%)] Loss: 0.157663
Train Epoch: 7 [1700/3016 (56%)] Loss: 0.190788
Train Epoch: 7 [1800/3016 (60%)] Loss: 0.090378
Train Epoch: 7 [1900/3016 (63%)] Loss: 0.032881
Train Epoch: 7 [2000/3016 (66%)] Loss: 0.027175
Train Epoch: 7 [2100/3016 (70%)] Loss: 0.013607
Train Epoch: 7 [2200/3016 (73%)] Loss: 0.099538
Train Epoch: 7 [2300/3016 (76%)] Loss: 0.060741
Train Epoch: 7 [2400/3016 (79%)] Loss: 0.082169
Train Epoch: 7 [2500/3016 (83%)] Loss: 0.006002
Train Epoch: 7 [2600/3016 (86%)] Loss: 0.337770
Train Epoch: 7 [2700/3016 (89%)] Loss: 0.030200
Train Epoch: 7 [2800/3016 (93%)] Loss: 0.011409
Train Epoch: 7 [2900/3016 (96%)] Loss: 0.172654
Train Epoch: 7 [3000/3016 (99%)] Loss: 0.126154
Total loss = 0.010454
Dev loss is 0.414909
Train Epoch: 8 [0/3016 (0%)] Loss: 0.103129
Train Epoch: 8 [100/3016 (3%)] Loss: 0.003298
Train Epoch: 8 [200/3016 (7%)] Loss: 0.008852
Train Epoch: 8 [300/3016 (10%)] Loss: 0.196912
Train Epoch: 8 [400/3016 (13%)] Loss: 0.263751
Train Epoch: 8 [500/3016 (17%)] Loss: 0.105734
Train Epoch: 8 [600/3016 (20%)] Loss: 0.209386
Train Epoch: 8 [700/3016 (23%)] Loss: 0.006484
Train Epoch: 8 [800/3016 (26%)] Loss: 0.100592
Train Epoch: 8 [900/3016 (30%)] Loss: 0.007448
Train Epoch: 8 [1000/3016 (33%)] Loss: 0.017933
Train Epoch: 8 [1100/3016 (36%)] Loss: 0.005104
Train Epoch: 8 [1200/3016 (40%)] Loss: 0.002109
Train Epoch: 8 [1300/3016 (43%)] Loss: 0.238951
Train Epoch: 8 [1400/3016 (46%)] Loss: 0.216928
Train Epoch: 8 [1500/3016 (50%)] Loss: 0.004355
Train Epoch: 8 [1600/3016 (53%)] Loss: 0.010543
Train Epoch: 8 [1700/3016 (56%)] Loss: 0.263727
Train Epoch: 8 [1800/3016 (60%)] Loss: 0.003844
Train Epoch: 8 [1900/3016 (63%)] Loss: 0.016518
Train Epoch: 8 [2000/3016 (66%)] Loss: 0.074659
Train Epoch: 8 [2100/3016 (70%)] Loss: 0.222038
Train Epoch: 8 [2200/3016 (73%)] Loss: 0.037002
Train Epoch: 8 [2300/3016 (76%)] Loss: 0.051943
Train Epoch: 8 [2400/3016 (79%)] Loss: 0.075705
Train Epoch: 8 [2500/3016 (83%)] Loss: 0.012986
Train Epoch: 8 [2600/3016 (86%)] Loss: 0.018453
Train Epoch: 8 [2700/3016 (89%)] Loss: 0.362083
Train Epoch: 8 [2800/3016 (93%)] Loss: 0.058363
Train Epoch: 8 [2900/3016 (96%)] Loss: 0.106565
Train Epoch: 8 [3000/3016 (99%)] Loss: 0.004482
Total loss = 0.009533
Dev loss is 0.478652
Train Epoch: 9 [0/3016 (0%)] Loss: 0.076874
Train Epoch: 9 [100/3016 (3%)] Loss: 0.058420
Train Epoch: 9 [200/3016 (7%)] Loss: 0.012103
Train Epoch: 9 [300/3016 (10%)] Loss: 0.024800
Train Epoch: 9 [400/3016 (13%)] Loss: 0.018379
Train Epoch: 9 [500/3016 (17%)] Loss: 0.006518
Train Epoch: 9 [600/3016 (20%)] Loss: 0.033437
Train Epoch: 9 [700/3016 (23%)] Loss: 0.072976
Train Epoch: 9 [800/3016 (26%)] Loss: 0.004449
Train Epoch: 9 [900/3016 (30%)] Loss: 0.212501
Train Epoch: 9 [1000/3016 (33%)] Loss: 0.093750
Train Epoch: 9 [1100/3016 (36%)] Loss: 0.004820
Train Epoch: 9 [1200/3016 (40%)] Loss: 0.335263
Train Epoch: 9 [1300/3016 (43%)] Loss: 0.297213
Train Epoch: 9 [1400/3016 (46%)] Loss: 0.128418
Train Epoch: 9 [1500/3016 (50%)] Loss: 0.084203
Train Epoch: 9 [1600/3016 (53%)] Loss: 0.020025
Train Epoch: 9 [1700/3016 (56%)] Loss: 0.019358
Train Epoch: 9 [1800/3016 (60%)] Loss: 0.002119
Train Epoch: 9 [1900/3016 (63%)] Loss: 0.037216
Train Epoch: 9 [2000/3016 (66%)] Loss: 0.066230
Train Epoch: 9 [2100/3016 (70%)] Loss: 0.151359
Train Epoch: 9 [2200/3016 (73%)] Loss: 0.051862
Train Epoch: 9 [2300/3016 (76%)] Loss: 0.067594
Train Epoch: 9 [2400/3016 (79%)] Loss: 0.056922
Train Epoch: 9 [2500/3016 (83%)] Loss: 0.021739
Train Epoch: 9 [2600/3016 (86%)] Loss: 0.056440
Train Epoch: 9 [2700/3016 (89%)] Loss: 0.080899
Train Epoch: 9 [2800/3016 (93%)] Loss: 0.042757
Train Epoch: 9 [2900/3016 (96%)] Loss: 0.017540
Train Epoch: 9 [3000/3016 (99%)] Loss: 0.507161
Total loss = 0.008187
Dev loss is 0.544138
Train Epoch: 10 [0/3016 (0%)] Loss: 0.023152
Train Epoch: 10 [100/3016 (3%)] Loss: 0.022705
Train Epoch: 10 [200/3016 (7%)] Loss: 0.000479
Train Epoch: 10 [300/3016 (10%)] Loss: 0.043663
Train Epoch: 10 [400/3016 (13%)] Loss: 0.136275
Train Epoch: 10 [500/3016 (17%)] Loss: 0.012166
Train Epoch: 10 [600/3016 (20%)] Loss: 0.011573
Train Epoch: 10 [700/3016 (23%)] Loss: 0.070756
Train Epoch: 10 [800/3016 (26%)] Loss: 0.037343
Train Epoch: 10 [900/3016 (30%)] Loss: 0.007444
Train Epoch: 10 [1000/3016 (33%)] Loss: 0.018495
Train Epoch: 10 [1100/3016 (36%)] Loss: 0.098513
Train Epoch: 10 [1200/3016 (40%)] Loss: 0.194199
Train Epoch: 10 [1300/3016 (43%)] Loss: 0.050926
Train Epoch: 10 [1400/3016 (46%)] Loss: 0.017892
Train Epoch: 10 [1500/3016 (50%)] Loss: 0.426654
Train Epoch: 10 [1600/3016 (53%)] Loss: 0.039550
Train Epoch: 10 [1700/3016 (56%)] Loss: 0.112380
Train Epoch: 10 [1800/3016 (60%)] Loss: 0.026773
Train Epoch: 10 [1900/3016 (63%)] Loss: 0.030351
Train Epoch: 10 [2000/3016 (66%)] Loss: 0.004027
Train Epoch: 10 [2100/3016 (70%)] Loss: 0.003668
Train Epoch: 10 [2200/3016 (73%)] Loss: 0.033695
Train Epoch: 10 [2300/3016 (76%)] Loss: 0.049829
Train Epoch: 10 [2400/3016 (79%)] Loss: 0.000978
Train Epoch: 10 [2500/3016 (83%)] Loss: 0.402267
Train Epoch: 10 [2600/3016 (86%)] Loss: 0.049310
Train Epoch: 10 [2700/3016 (89%)] Loss: 0.261600
Train Epoch: 10 [2800/3016 (93%)] Loss: 0.033538
Train Epoch: 10 [2900/3016 (96%)] Loss: 0.132974
Train Epoch: 10 [3000/3016 (99%)] Loss: 0.051098
Total loss = 0.006717
Dev loss is 0.341777
Train Epoch: 11 [0/3016 (0%)] Loss: 0.022969
Train Epoch: 11 [100/3016 (3%)] Loss: 0.015212
Train Epoch: 11 [200/3016 (7%)] Loss: 0.107095
Train Epoch: 11 [300/3016 (10%)] Loss: 0.001546
Train Epoch: 11 [400/3016 (13%)] Loss: 0.002972
Train Epoch: 11 [500/3016 (17%)] Loss: 0.021759
Train Epoch: 11 [600/3016 (20%)] Loss: 0.014909
Train Epoch: 11 [700/3016 (23%)] Loss: 0.000440
Train Epoch: 11 [800/3016 (26%)] Loss: 0.013423
Train Epoch: 11 [900/3016 (30%)] Loss: 0.001969
Train Epoch: 11 [1000/3016 (33%)] Loss: 0.006954
Train Epoch: 11 [1100/3016 (36%)] Loss: 0.041093
Train Epoch: 11 [1200/3016 (40%)] Loss: 0.002497
Train Epoch: 11 [1300/3016 (43%)] Loss: 0.017694
Train Epoch: 11 [1400/3016 (46%)] Loss: 0.019259
Train Epoch: 11 [1500/3016 (50%)] Loss: 0.006534
Train Epoch: 11 [1600/3016 (53%)] Loss: 0.011240
Train Epoch: 11 [1700/3016 (56%)] Loss: 0.197181
Train Epoch: 11 [1800/3016 (60%)] Loss: 0.002715
Train Epoch: 11 [1900/3016 (63%)] Loss: 0.283998
Train Epoch: 11 [2000/3016 (66%)] Loss: 0.148242
Train Epoch: 11 [2100/3016 (70%)] Loss: 0.078169
Train Epoch: 11 [2200/3016 (73%)] Loss: 0.098966
Train Epoch: 11 [2300/3016 (76%)] Loss: 0.131934
Train Epoch: 11 [2400/3016 (79%)] Loss: 0.007414
Train Epoch: 11 [2500/3016 (83%)] Loss: 0.011926
Train Epoch: 11 [2600/3016 (86%)] Loss: 0.057112
Train Epoch: 11 [2700/3016 (89%)] Loss: 0.001415
Train Epoch: 11 [2800/3016 (93%)] Loss: 0.002563
Train Epoch: 11 [2900/3016 (96%)] Loss: 0.038082
Train Epoch: 11 [3000/3016 (99%)] Loss: 0.046146
Total loss = 0.005029
Dev loss is 0.313025
Train Epoch: 12 [0/3016 (0%)] Loss: 0.000203
Train Epoch: 12 [100/3016 (3%)] Loss: 0.002901
Train Epoch: 12 [200/3016 (7%)] Loss: 0.125867
Train Epoch: 12 [300/3016 (10%)] Loss: 0.097349
Train Epoch: 12 [400/3016 (13%)] Loss: 0.039880
Train Epoch: 12 [500/3016 (17%)] Loss: 0.154855
Train Epoch: 12 [600/3016 (20%)] Loss: 0.360564
Train Epoch: 12 [700/3016 (23%)] Loss: 0.007179
Train Epoch: 12 [800/3016 (26%)] Loss: 0.109012
Train Epoch: 12 [900/3016 (30%)] Loss: 0.465744
Train Epoch: 12 [1000/3016 (33%)] Loss: 0.023963
Train Epoch: 12 [1100/3016 (36%)] Loss: 0.012986
Train Epoch: 12 [1200/3016 (40%)] Loss: 0.026066
Train Epoch: 12 [1300/3016 (43%)] Loss: 0.186547
Train Epoch: 12 [1400/3016 (46%)] Loss: 0.035706
Train Epoch: 12 [1500/3016 (50%)] Loss: 0.020995
Train Epoch: 12 [1600/3016 (53%)] Loss: 0.152514
Train Epoch: 12 [1700/3016 (56%)] Loss: 0.194344
Train Epoch: 12 [1800/3016 (60%)] Loss: 0.002521
Train Epoch: 12 [1900/3016 (63%)] Loss: 0.014468
Train Epoch: 12 [2000/3016 (66%)] Loss: 0.009552
Train Epoch: 12 [2100/3016 (70%)] Loss: 0.004636
Train Epoch: 12 [2200/3016 (73%)] Loss: 0.026562
Train Epoch: 12 [2300/3016 (76%)] Loss: 0.360037
Train Epoch: 12 [2400/3016 (79%)] Loss: 0.017579
Train Epoch: 12 [2500/3016 (83%)] Loss: 0.003016
Train Epoch: 12 [2600/3016 (86%)] Loss: 0.012689
Train Epoch: 12 [2700/3016 (89%)] Loss: 0.002242
Train Epoch: 12 [2800/3016 (93%)] Loss: 0.040326
Train Epoch: 12 [2900/3016 (96%)] Loss: 0.012377
Train Epoch: 12 [3000/3016 (99%)] Loss: 0.009420
Total loss = 0.005528
Dev loss is 0.381998
Train Epoch: 13 [0/3016 (0%)] Loss: 0.011153
Train Epoch: 13 [100/3016 (3%)] Loss: 0.049022
Train Epoch: 13 [200/3016 (7%)] Loss: 0.008721
Train Epoch: 13 [300/3016 (10%)] Loss: 0.002679
Train Epoch: 13 [400/3016 (13%)] Loss: 0.003897
Train Epoch: 13 [500/3016 (17%)] Loss: 0.021533
Train Epoch: 13 [600/3016 (20%)] Loss: 0.003566
Train Epoch: 13 [700/3016 (23%)] Loss: 0.084661
Train Epoch: 13 [800/3016 (26%)] Loss: 0.001812
Train Epoch: 13 [900/3016 (30%)] Loss: 0.019079
Train Epoch: 13 [1000/3016 (33%)] Loss: 0.047911
Train Epoch: 13 [1100/3016 (36%)] Loss: 0.002531
Train Epoch: 13 [1200/3016 (40%)] Loss: 0.006521
Train Epoch: 13 [1300/3016 (43%)] Loss: 0.199952
Train Epoch: 13 [1400/3016 (46%)] Loss: 0.089214
Train Epoch: 13 [1500/3016 (50%)] Loss: 0.035623
Train Epoch: 13 [1600/3016 (53%)] Loss: 0.007520
Train Epoch: 13 [1700/3016 (56%)] Loss: 0.001268
Train Epoch: 13 [1800/3016 (60%)] Loss: 0.008082
Train Epoch: 13 [1900/3016 (63%)] Loss: 0.006925
Train Epoch: 13 [2000/3016 (66%)] Loss: 0.124590
Train Epoch: 13 [2100/3016 (70%)] Loss: 0.009314
Train Epoch: 13 [2200/3016 (73%)] Loss: 0.003824
Train Epoch: 13 [2300/3016 (76%)] Loss: 0.026436
Train Epoch: 13 [2400/3016 (79%)] Loss: 0.003261
Train Epoch: 13 [2500/3016 (83%)] Loss: 0.008816
Train Epoch: 13 [2600/3016 (86%)] Loss: 0.003625
Train Epoch: 13 [2700/3016 (89%)] Loss: 0.031985
Train Epoch: 13 [2800/3016 (93%)] Loss: 0.003903
Train Epoch: 13 [2900/3016 (96%)] Loss: 0.004454
Train Epoch: 13 [3000/3016 (99%)] Loss: 0.007493
Total loss = 0.003437
Dev loss is 0.396537
Train Epoch: 14 [0/3016 (0%)] Loss: 0.005231
Train Epoch: 14 [100/3016 (3%)] Loss: 0.043764
Train Epoch: 14 [200/3016 (7%)] Loss: 0.002211
Train Epoch: 14 [300/3016 (10%)] Loss: 0.013407
Train Epoch: 14 [400/3016 (13%)] Loss: 0.010574
Train Epoch: 14 [500/3016 (17%)] Loss: 0.004530
Train Epoch: 14 [600/3016 (20%)] Loss: 0.080314
Train Epoch: 14 [700/3016 (23%)] Loss: 0.059645
Train Epoch: 14 [800/3016 (26%)] Loss: 0.002677
Train Epoch: 14 [900/3016 (30%)] Loss: 0.024123
Train Epoch: 14 [1000/3016 (33%)] Loss: 0.034711
Train Epoch: 14 [1100/3016 (36%)] Loss: 0.003979
Train Epoch: 14 [1200/3016 (40%)] Loss: 0.009792
Train Epoch: 14 [1300/3016 (43%)] Loss: 0.004013
Train Epoch: 14 [1400/3016 (46%)] Loss: 0.000076
Train Epoch: 14 [1500/3016 (50%)] Loss: 0.007084
Train Epoch: 14 [1600/3016 (53%)] Loss: 0.004951
Train Epoch: 14 [1700/3016 (56%)] Loss: 0.000399
Train Epoch: 14 [1800/3016 (60%)] Loss: 0.001259
Train Epoch: 14 [1900/3016 (63%)] Loss: 0.028447
Train Epoch: 14 [2000/3016 (66%)] Loss: 0.000113
Train Epoch: 14 [2100/3016 (70%)] Loss: 0.026755
Train Epoch: 14 [2200/3016 (73%)] Loss: 0.014674
Train Epoch: 14 [2300/3016 (76%)] Loss: 0.033025
Train Epoch: 14 [2400/3016 (79%)] Loss: 0.011009
Train Epoch: 14 [2500/3016 (83%)] Loss: 0.316729
Train Epoch: 14 [2600/3016 (86%)] Loss: 0.001204
Train Epoch: 14 [2700/3016 (89%)] Loss: 0.047659
Train Epoch: 14 [2800/3016 (93%)] Loss: 0.034001
Train Epoch: 14 [2900/3016 (96%)] Loss: 0.064962
Train Epoch: 14 [3000/3016 (99%)] Loss: 0.006653
Total loss = 0.003583
Dev loss is 0.417001
Train Epoch: 15 [0/3016 (0%)] Loss: 0.004181
Train Epoch: 15 [100/3016 (3%)] Loss: 0.001045
Train Epoch: 15 [200/3016 (7%)] Loss: 0.041663
Train Epoch: 15 [300/3016 (10%)] Loss: 0.000172
Train Epoch: 15 [400/3016 (13%)] Loss: 0.002401
Train Epoch: 15 [500/3016 (17%)] Loss: 0.130078
Train Epoch: 15 [600/3016 (20%)] Loss: 0.003962
Train Epoch: 15 [700/3016 (23%)] Loss: 0.001663
Train Epoch: 15 [800/3016 (26%)] Loss: 0.035960
Train Epoch: 15 [900/3016 (30%)] Loss: 0.007867
Train Epoch: 15 [1000/3016 (33%)] Loss: 0.033631
Train Epoch: 15 [1100/3016 (36%)] Loss: 0.007026
Train Epoch: 15 [1200/3016 (40%)] Loss: 0.002230
Train Epoch: 15 [1300/3016 (43%)] Loss: 0.004235
Train Epoch: 15 [1400/3016 (46%)] Loss: 0.005650
Train Epoch: 15 [1500/3016 (50%)] Loss: 0.010916
Train Epoch: 15 [1600/3016 (53%)] Loss: 0.003661
Train Epoch: 15 [1700/3016 (56%)] Loss: 0.000074
Train Epoch: 15 [1800/3016 (60%)] Loss: 0.038834
Train Epoch: 15 [1900/3016 (63%)] Loss: 0.001572
Train Epoch: 15 [2000/3016 (66%)] Loss: 0.115292
Train Epoch: 15 [2100/3016 (70%)] Loss: 0.001270
Train Epoch: 15 [2200/3016 (73%)] Loss: 0.001106
Train Epoch: 15 [2300/3016 (76%)] Loss: 0.064665
Train Epoch: 15 [2400/3016 (79%)] Loss: 0.003573
Train Epoch: 15 [2500/3016 (83%)] Loss: 0.000929
Train Epoch: 15 [2600/3016 (86%)] Loss: 0.003251
Train Epoch: 15 [2700/3016 (89%)] Loss: 0.002830
Train Epoch: 15 [2800/3016 (93%)] Loss: 0.012175
Train Epoch: 15 [2900/3016 (96%)] Loss: 0.000596
Train Epoch: 15 [3000/3016 (99%)] Loss: 0.003428
Total loss = 0.002692
Dev loss is 0.499989
Train Epoch: 16 [0/3016 (0%)] Loss: 0.000200
Train Epoch: 16 [100/3016 (3%)] Loss: 0.000152
Train Epoch: 16 [200/3016 (7%)] Loss: 0.000500
Train Epoch: 16 [300/3016 (10%)] Loss: 0.000336
Train Epoch: 16 [400/3016 (13%)] Loss: 0.000527
Train Epoch: 16 [500/3016 (17%)] Loss: 0.002347
Train Epoch: 16 [600/3016 (20%)] Loss: 0.002068
Train Epoch: 16 [700/3016 (23%)] Loss: 0.033420
Train Epoch: 16 [800/3016 (26%)] Loss: 0.220372
Train Epoch: 16 [900/3016 (30%)] Loss: 0.003268
Train Epoch: 16 [1000/3016 (33%)] Loss: 0.011723
Train Epoch: 16 [1100/3016 (36%)] Loss: 0.075612
Train Epoch: 16 [1200/3016 (40%)] Loss: 0.002177
Train Epoch: 16 [1300/3016 (43%)] Loss: 0.002535
Train Epoch: 16 [1400/3016 (46%)] Loss: 0.001816
Train Epoch: 16 [1500/3016 (50%)] Loss: 0.001947
Train Epoch: 16 [1600/3016 (53%)] Loss: 0.000372
Train Epoch: 16 [1700/3016 (56%)] Loss: 0.001275
Train Epoch: 16 [1800/3016 (60%)] Loss: 0.005868
Train Epoch: 16 [1900/3016 (63%)] Loss: 0.006879
Train Epoch: 16 [2000/3016 (66%)] Loss: 0.000209
Train Epoch: 16 [2100/3016 (70%)] Loss: 0.026446
Train Epoch: 16 [2200/3016 (73%)] Loss: 0.002588
Train Epoch: 16 [2300/3016 (76%)] Loss: 0.001907
Train Epoch: 16 [2400/3016 (79%)] Loss: 0.001105
Train Epoch: 16 [2500/3016 (83%)] Loss: 0.000270
Train Epoch: 16 [2600/3016 (86%)] Loss: 0.016225
Train Epoch: 16 [2700/3016 (89%)] Loss: 0.004016
Train Epoch: 16 [2800/3016 (93%)] Loss: 0.002019
Train Epoch: 16 [2900/3016 (96%)] Loss: 0.014452
Train Epoch: 16 [3000/3016 (99%)] Loss: 0.002016
Total loss = 0.003216
Dev loss is 0.545849
Train Epoch: 17 [0/3016 (0%)] Loss: 0.001359
Train Epoch: 17 [100/3016 (3%)] Loss: 0.016942
Train Epoch: 17 [200/3016 (7%)] Loss: 0.004978
Train Epoch: 17 [300/3016 (10%)] Loss: 0.000125
Train Epoch: 17 [400/3016 (13%)] Loss: 0.000900
Train Epoch: 17 [500/3016 (17%)] Loss: 0.001493
Train Epoch: 17 [600/3016 (20%)] Loss: 0.000022
Train Epoch: 17 [700/3016 (23%)] Loss: 0.004877
Train Epoch: 17 [800/3016 (26%)] Loss: 0.000383
Train Epoch: 17 [900/3016 (30%)] Loss: 0.002079
Train Epoch: 17 [1000/3016 (33%)] Loss: 0.156728
Train Epoch: 17 [1100/3016 (36%)] Loss: 0.011431
Train Epoch: 17 [1200/3016 (40%)] Loss: 0.002888
Train Epoch: 17 [1300/3016 (43%)] Loss: 0.032387
Train Epoch: 17 [1400/3016 (46%)] Loss: 0.000583
Train Epoch: 17 [1500/3016 (50%)] Loss: 0.000869
Train Epoch: 17 [1600/3016 (53%)] Loss: 0.002255
Train Epoch: 17 [1700/3016 (56%)] Loss: 0.114670
Train Epoch: 17 [1800/3016 (60%)] Loss: 0.002090
Train Epoch: 17 [1900/3016 (63%)] Loss: 0.026033
Train Epoch: 17 [2000/3016 (66%)] Loss: 0.072197
Train Epoch: 17 [2100/3016 (70%)] Loss: 0.005592
Train Epoch: 17 [2200/3016 (73%)] Loss: 0.018023
Train Epoch: 17 [2300/3016 (76%)] Loss: 0.084198
Train Epoch: 17 [2400/3016 (79%)] Loss: 0.001905
Train Epoch: 17 [2500/3016 (83%)] Loss: 0.000787
Train Epoch: 17 [2600/3016 (86%)] Loss: 0.075382
Train Epoch: 17 [2700/3016 (89%)] Loss: 0.033196
Train Epoch: 17 [2800/3016 (93%)] Loss: 0.004381
Train Epoch: 17 [2900/3016 (96%)] Loss: 0.022334
Train Epoch: 17 [3000/3016 (99%)] Loss: 0.001308
Total loss = 0.003008
Dev loss is 0.591007
Train Epoch: 18 [0/3016 (0%)] Loss: 0.016350
Train Epoch: 18 [100/3016 (3%)] Loss: 0.026286
Train Epoch: 18 [200/3016 (7%)] Loss: 0.013728
Train Epoch: 18 [300/3016 (10%)] Loss: 0.032300
Train Epoch: 18 [400/3016 (13%)] Loss: 0.005726
Train Epoch: 18 [500/3016 (17%)] Loss: 0.000295
Train Epoch: 18 [600/3016 (20%)] Loss: 0.011204
Train Epoch: 18 [700/3016 (23%)] Loss: 0.000508
Train Epoch: 18 [800/3016 (26%)] Loss: 0.003936
Train Epoch: 18 [900/3016 (30%)] Loss: 0.003268
Train Epoch: 18 [1000/3016 (33%)] Loss: 0.009851
Train Epoch: 18 [1100/3016 (36%)] Loss: 0.012176
Train Epoch: 18 [1200/3016 (40%)] Loss: 0.022570
Train Epoch: 18 [1300/3016 (43%)] Loss: 0.000564
Train Epoch: 18 [1400/3016 (46%)] Loss: 0.004621
Train Epoch: 18 [1500/3016 (50%)] Loss: 0.020412
Train Epoch: 18 [1600/3016 (53%)] Loss: 0.002892
Train Epoch: 18 [1700/3016 (56%)] Loss: 0.001978
Train Epoch: 18 [1800/3016 (60%)] Loss: 0.108579
Train Epoch: 18 [1900/3016 (63%)] Loss: 0.251623
Train Epoch: 18 [2000/3016 (66%)] Loss: 0.112387
Train Epoch: 18 [2100/3016 (70%)] Loss: 0.003740
Train Epoch: 18 [2200/3016 (73%)] Loss: 0.002455
Train Epoch: 18 [2300/3016 (76%)] Loss: 0.004258
Train Epoch: 18 [2400/3016 (79%)] Loss: 0.000113
Train Epoch: 18 [2500/3016 (83%)] Loss: 0.005034
Train Epoch: 18 [2600/3016 (86%)] Loss: 0.231828
Train Epoch: 18 [2700/3016 (89%)] Loss: 0.000206
Train Epoch: 18 [2800/3016 (93%)] Loss: 0.045239
Train Epoch: 18 [2900/3016 (96%)] Loss: 0.000764
Train Epoch: 18 [3000/3016 (99%)] Loss: 0.003373
Total loss = 0.002596
Dev loss is 0.400398
Train Epoch: 19 [0/3016 (0%)] Loss: 0.000432
Train Epoch: 19 [100/3016 (3%)] Loss: 0.004571
Train Epoch: 19 [200/3016 (7%)] Loss: 0.006059
Train Epoch: 19 [300/3016 (10%)] Loss: 0.001207
Train Epoch: 19 [400/3016 (13%)] Loss: 0.003042
Train Epoch: 19 [500/3016 (17%)] Loss: 0.008797
Train Epoch: 19 [600/3016 (20%)] Loss: 0.031010
Train Epoch: 19 [700/3016 (23%)] Loss: 0.000114
Train Epoch: 19 [800/3016 (26%)] Loss: 0.029825
Train Epoch: 19 [900/3016 (30%)] Loss: 0.000267
Train Epoch: 19 [1000/3016 (33%)] Loss: 0.000412
Train Epoch: 19 [1100/3016 (36%)] Loss: 0.001060
Train Epoch: 19 [1200/3016 (40%)] Loss: 0.016338
Train Epoch: 19 [1300/3016 (43%)] Loss: 0.000226
Train Epoch: 19 [1400/3016 (46%)] Loss: 0.058061
Train Epoch: 19 [1500/3016 (50%)] Loss: 0.226102
Train Epoch: 19 [1600/3016 (53%)] Loss: 0.000285
Train Epoch: 19 [1700/3016 (56%)] Loss: 0.014304
Train Epoch: 19 [1800/3016 (60%)] Loss: 0.001574
Train Epoch: 19 [1900/3016 (63%)] Loss: 0.000221
Train Epoch: 19 [2000/3016 (66%)] Loss: 0.002335
Train Epoch: 19 [2100/3016 (70%)] Loss: 0.002716
Train Epoch: 19 [2200/3016 (73%)] Loss: 0.000467
Train Epoch: 19 [2300/3016 (76%)] Loss: 0.000345
Train Epoch: 19 [2400/3016 (79%)] Loss: 0.006737
Train Epoch: 19 [2500/3016 (83%)] Loss: 0.003426
Train Epoch: 19 [2600/3016 (86%)] Loss: 0.000216
Train Epoch: 19 [2700/3016 (89%)] Loss: 0.017138
Train Epoch: 19 [2800/3016 (93%)] Loss: 0.000393
Train Epoch: 19 [2900/3016 (96%)] Loss: 0.000942
Train Epoch: 19 [3000/3016 (99%)] Loss: 0.015560
Total loss = 0.002357
Dev loss is 0.375816
Train Epoch: 20 [0/3016 (0%)] Loss: 0.004915
Train Epoch: 20 [100/3016 (3%)] Loss: 0.021066
Train Epoch: 20 [200/3016 (7%)] Loss: 0.177734
Train Epoch: 20 [300/3016 (10%)] Loss: 0.006871
Train Epoch: 20 [400/3016 (13%)] Loss: 0.027458
Train Epoch: 20 [500/3016 (17%)] Loss: 0.004393
Train Epoch: 20 [600/3016 (20%)] Loss: 0.012841
Train Epoch: 20 [700/3016 (23%)] Loss: 0.000572
Train Epoch: 20 [800/3016 (26%)] Loss: 0.016672
Train Epoch: 20 [900/3016 (30%)] Loss: 0.102734
Train Epoch: 20 [1000/3016 (33%)] Loss: 0.001811
Train Epoch: 20 [1100/3016 (36%)] Loss: 0.002679
Train Epoch: 20 [1200/3016 (40%)] Loss: 0.000108
Train Epoch: 20 [1300/3016 (43%)] Loss: 0.002293
Train Epoch: 20 [1400/3016 (46%)] Loss: 0.001401
Train Epoch: 20 [1500/3016 (50%)] Loss: 0.075998
Train Epoch: 20 [1600/3016 (53%)] Loss: 0.018791
Train Epoch: 20 [1700/3016 (56%)] Loss: 0.055988
Train Epoch: 20 [1800/3016 (60%)] Loss: 0.015417
Train Epoch: 20 [1900/3016 (63%)] Loss: 0.000407
Train Epoch: 20 [2000/3016 (66%)] Loss: 0.032924
Train Epoch: 20 [2100/3016 (70%)] Loss: 0.000149
Train Epoch: 20 [2200/3016 (73%)] Loss: 0.001055
Train Epoch: 20 [2300/3016 (76%)] Loss: 0.009155
Train Epoch: 20 [2400/3016 (79%)] Loss: 0.001834
Train Epoch: 20 [2500/3016 (83%)] Loss: 0.009599
Train Epoch: 20 [2600/3016 (86%)] Loss: 0.003730
Train Epoch: 20 [2700/3016 (89%)] Loss: 0.000919
Train Epoch: 20 [2800/3016 (93%)] Loss: 0.001343
Train Epoch: 20 [2900/3016 (96%)] Loss: 0.016228
Train Epoch: 20 [3000/3016 (99%)] Loss: 0.014121
Total loss = 0.002741
Dev loss is 0.669932
Train Epoch: 21 [0/3016 (0%)] Loss: 0.015707
Train Epoch: 21 [100/3016 (3%)] Loss: 0.011186
Train Epoch: 21 [200/3016 (7%)] Loss: 0.151382
Train Epoch: 21 [300/3016 (10%)] Loss: 0.004965
Train Epoch: 21 [400/3016 (13%)] Loss: 0.002108
Train Epoch: 21 [500/3016 (17%)] Loss: 0.027960
Train Epoch: 21 [600/3016 (20%)] Loss: 0.005778
Train Epoch: 21 [700/3016 (23%)] Loss: 0.015017
Train Epoch: 21 [800/3016 (26%)] Loss: 0.002000
Train Epoch: 21 [900/3016 (30%)] Loss: 0.000159
Train Epoch: 21 [1000/3016 (33%)] Loss: 0.000586
Train Epoch: 21 [1100/3016 (36%)] Loss: 0.000502
Train Epoch: 21 [1200/3016 (40%)] Loss: 0.005638
Train Epoch: 21 [1300/3016 (43%)] Loss: 0.000660
Train Epoch: 21 [1400/3016 (46%)] Loss: 0.000495
Train Epoch: 21 [1500/3016 (50%)] Loss: 0.002996
Train Epoch: 21 [1600/3016 (53%)] Loss: 0.075466
Train Epoch: 21 [1700/3016 (56%)] Loss: 0.000337
Train Epoch: 21 [1800/3016 (60%)] Loss: 0.000378
Train Epoch: 21 [1900/3016 (63%)] Loss: 0.000302
Train Epoch: 21 [2000/3016 (66%)] Loss: 0.008518
Train Epoch: 21 [2100/3016 (70%)] Loss: 0.007758
Train Epoch: 21 [2200/3016 (73%)] Loss: 0.004881
Train Epoch: 21 [2300/3016 (76%)] Loss: 0.014798
Train Epoch: 21 [2400/3016 (79%)] Loss: 0.000117
Train Epoch: 21 [2500/3016 (83%)] Loss: 0.000147
Train Epoch: 21 [2600/3016 (86%)] Loss: 0.000273
Train Epoch: 21 [2700/3016 (89%)] Loss: 0.289612
Train Epoch: 21 [2800/3016 (93%)] Loss: 0.006285
Train Epoch: 21 [2900/3016 (96%)] Loss: 0.004765
Train Epoch: 21 [3000/3016 (99%)] Loss: 0.000413
Total loss = 0.001659
Dev loss is 1.150287
Train Epoch: 22 [0/3016 (0%)] Loss: 0.000743
Train Epoch: 22 [100/3016 (3%)] Loss: 0.001183
Train Epoch: 22 [200/3016 (7%)] Loss: 0.001599
Train Epoch: 22 [300/3016 (10%)] Loss: 0.000187
Train Epoch: 22 [400/3016 (13%)] Loss: 0.000950
Train Epoch: 22 [500/3016 (17%)] Loss: 0.001318
Train Epoch: 22 [600/3016 (20%)] Loss: 0.002310
Train Epoch: 22 [700/3016 (23%)] Loss: 0.003474
Train Epoch: 22 [800/3016 (26%)] Loss: 0.004722
Train Epoch: 22 [900/3016 (30%)] Loss: 0.005627
Train Epoch: 22 [1000/3016 (33%)] Loss: 0.003032
Train Epoch: 22 [1100/3016 (36%)] Loss: 0.000306
Train Epoch: 22 [1200/3016 (40%)] Loss: 0.002393
Train Epoch: 22 [1300/3016 (43%)] Loss: 0.000120
Train Epoch: 22 [1400/3016 (46%)] Loss: 0.000004
Train Epoch: 22 [1500/3016 (50%)] Loss: 0.000016
Train Epoch: 22 [1600/3016 (53%)] Loss: 0.000094
Train Epoch: 22 [1700/3016 (56%)] Loss: 0.057458
Train Epoch: 22 [1800/3016 (60%)] Loss: 0.000598
Train Epoch: 22 [1900/3016 (63%)] Loss: 0.034401
Train Epoch: 22 [2000/3016 (66%)] Loss: 0.003806
Train Epoch: 22 [2100/3016 (70%)] Loss: 0.000043
Train Epoch: 22 [2200/3016 (73%)] Loss: 0.017811
Train Epoch: 22 [2300/3016 (76%)] Loss: 0.000323
Train Epoch: 22 [2400/3016 (79%)] Loss: 0.008448
Train Epoch: 22 [2500/3016 (83%)] Loss: 0.000057
Train Epoch: 22 [2600/3016 (86%)] Loss: 0.000489
Train Epoch: 22 [2700/3016 (89%)] Loss: 0.001005
Train Epoch: 22 [2800/3016 (93%)] Loss: 0.000879
Train Epoch: 22 [2900/3016 (96%)] Loss: 0.001140
Train Epoch: 22 [3000/3016 (99%)] Loss: 0.000634
Total loss = 0.001822
Dev loss is 0.774249
Train Epoch: 23 [0/3016 (0%)] Loss: 0.007042
Train Epoch: 23 [100/3016 (3%)] Loss: 0.001256
Train Epoch: 23 [200/3016 (7%)] Loss: 0.000032
Train Epoch: 23 [300/3016 (10%)] Loss: 0.000090
Train Epoch: 23 [400/3016 (13%)] Loss: 0.002388
Train Epoch: 23 [500/3016 (17%)] Loss: 0.056072
Train Epoch: 23 [600/3016 (20%)] Loss: 0.000005
Train Epoch: 23 [700/3016 (23%)] Loss: 0.000355
Train Epoch: 23 [800/3016 (26%)] Loss: 0.000123
Train Epoch: 23 [900/3016 (30%)] Loss: 0.009441
Train Epoch: 23 [1000/3016 (33%)] Loss: 0.034089
Train Epoch: 23 [1100/3016 (36%)] Loss: 0.000380
Train Epoch: 23 [1200/3016 (40%)] Loss: 0.000534
Train Epoch: 23 [1300/3016 (43%)] Loss: 0.006908
Train Epoch: 23 [1400/3016 (46%)] Loss: 0.003400
Train Epoch: 23 [1500/3016 (50%)] Loss: 0.004769
Train Epoch: 23 [1600/3016 (53%)] Loss: 0.001651
Train Epoch: 23 [1700/3016 (56%)] Loss: 0.000031
Train Epoch: 23 [1800/3016 (60%)] Loss: 0.000014
Train Epoch: 23 [1900/3016 (63%)] Loss: 0.003701
Train Epoch: 23 [2000/3016 (66%)] Loss: 0.000095
Train Epoch: 23 [2100/3016 (70%)] Loss: 0.007435
Train Epoch: 23 [2200/3016 (73%)] Loss: 0.000001
Train Epoch: 23 [2300/3016 (76%)] Loss: 0.000259
Train Epoch: 23 [2400/3016 (79%)] Loss: 0.000005
Train Epoch: 23 [2500/3016 (83%)] Loss: 0.000582
Train Epoch: 23 [2600/3016 (86%)] Loss: 0.002006
Train Epoch: 23 [2700/3016 (89%)] Loss: 0.000046
Train Epoch: 23 [2800/3016 (93%)] Loss: 0.118996
Train Epoch: 23 [2900/3016 (96%)] Loss: 0.000051
Train Epoch: 23 [3000/3016 (99%)] Loss: 0.000912
Total loss = 0.002208
Dev loss is 0.822080
Train Epoch: 24 [0/3016 (0%)] Loss: 0.000503
Train Epoch: 24 [100/3016 (3%)] Loss: 0.210552
Train Epoch: 24 [200/3016 (7%)] Loss: 0.000371
Train Epoch: 24 [300/3016 (10%)] Loss: 0.094849
Train Epoch: 24 [400/3016 (13%)] Loss: 0.009786
Train Epoch: 24 [500/3016 (17%)] Loss: 0.004728
Train Epoch: 24 [600/3016 (20%)] Loss: 0.002674
Train Epoch: 24 [700/3016 (23%)] Loss: 0.042638
Train Epoch: 24 [800/3016 (26%)] Loss: 0.000117
Train Epoch: 24 [900/3016 (30%)] Loss: 0.001391
Train Epoch: 24 [1000/3016 (33%)] Loss: 0.007430
Train Epoch: 24 [1100/3016 (36%)] Loss: 0.000130
Train Epoch: 24 [1200/3016 (40%)] Loss: 0.010505
Train Epoch: 24 [1300/3016 (43%)] Loss: 0.078833
Train Epoch: 24 [1400/3016 (46%)] Loss: 0.011908
Train Epoch: 24 [1500/3016 (50%)] Loss: 0.000833
Train Epoch: 24 [1600/3016 (53%)] Loss: 0.001674
Train Epoch: 24 [1700/3016 (56%)] Loss: 0.140511
Train Epoch: 24 [1800/3016 (60%)] Loss: 0.001601
Train Epoch: 24 [1900/3016 (63%)] Loss: 0.000212
Train Epoch: 24 [2000/3016 (66%)] Loss: 0.000149
Train Epoch: 24 [2100/3016 (70%)] Loss: 0.392137
Train Epoch: 24 [2200/3016 (73%)] Loss: 0.008956
Train Epoch: 24 [2300/3016 (76%)] Loss: 0.137028
Train Epoch: 24 [2400/3016 (79%)] Loss: 0.004243
Train Epoch: 24 [2500/3016 (83%)] Loss: 0.052916
Train Epoch: 24 [2600/3016 (86%)] Loss: 0.000782
Train Epoch: 24 [2700/3016 (89%)] Loss: 0.290022
Train Epoch: 24 [2800/3016 (93%)] Loss: 0.001200
Train Epoch: 24 [2900/3016 (96%)] Loss: 0.012669
Train Epoch: 24 [3000/3016 (99%)] Loss: 0.008858
Total loss = 0.002562
Dev loss is 1.058864
Train Epoch: 25 [0/3016 (0%)] Loss: 0.004033
Train Epoch: 25 [100/3016 (3%)] Loss: 0.002034
Train Epoch: 25 [200/3016 (7%)] Loss: 0.001714
Train Epoch: 25 [300/3016 (10%)] Loss: 0.001823
Train Epoch: 25 [400/3016 (13%)] Loss: 0.047205
Train Epoch: 25 [500/3016 (17%)] Loss: 0.005502
Train Epoch: 25 [600/3016 (20%)] Loss: 0.000703
Train Epoch: 25 [700/3016 (23%)] Loss: 0.039674
Train Epoch: 25 [800/3016 (26%)] Loss: 0.001295
Train Epoch: 25 [900/3016 (30%)] Loss: 0.000048
Train Epoch: 25 [1000/3016 (33%)] Loss: 0.001093
Train Epoch: 25 [1100/3016 (36%)] Loss: 0.042403
Train Epoch: 25 [1200/3016 (40%)] Loss: 0.020649
Train Epoch: 25 [1300/3016 (43%)] Loss: 0.000474
Train Epoch: 25 [1400/3016 (46%)] Loss: 0.019275
Train Epoch: 25 [1500/3016 (50%)] Loss: 0.001174
Train Epoch: 25 [1600/3016 (53%)] Loss: 0.001793
Train Epoch: 25 [1700/3016 (56%)] Loss: 0.000299
Train Epoch: 25 [1800/3016 (60%)] Loss: 0.002047
Train Epoch: 25 [1900/3016 (63%)] Loss: 0.074177
Train Epoch: 25 [2000/3016 (66%)] Loss: 0.000449
Train Epoch: 25 [2100/3016 (70%)] Loss: 0.064484
Train Epoch: 25 [2200/3016 (73%)] Loss: 0.000750
Train Epoch: 25 [2300/3016 (76%)] Loss: 0.012086
Train Epoch: 25 [2400/3016 (79%)] Loss: 0.000144
Train Epoch: 25 [2500/3016 (83%)] Loss: 0.027945
Train Epoch: 25 [2600/3016 (86%)] Loss: 0.003413
Train Epoch: 25 [2700/3016 (89%)] Loss: 0.001292
Train Epoch: 25 [2800/3016 (93%)] Loss: 0.126377
Train Epoch: 25 [2900/3016 (96%)] Loss: 0.007662
Train Epoch: 25 [3000/3016 (99%)] Loss: 0.000105
Total loss = 0.001509
Dev loss is 0.584196
Train Epoch: 26 [0/3016 (0%)] Loss: 0.003436
Train Epoch: 26 [100/3016 (3%)] Loss: 0.000812
Train Epoch: 26 [200/3016 (7%)] Loss: 0.001562
Train Epoch: 26 [300/3016 (10%)] Loss: 0.001284
Train Epoch: 26 [400/3016 (13%)] Loss: 0.000435
Train Epoch: 26 [500/3016 (17%)] Loss: 0.020246
Train Epoch: 26 [600/3016 (20%)] Loss: 0.009113
Train Epoch: 26 [700/3016 (23%)] Loss: 0.041454
Train Epoch: 26 [800/3016 (26%)] Loss: 0.000041
Train Epoch: 26 [900/3016 (30%)] Loss: 0.001662
Train Epoch: 26 [1000/3016 (33%)] Loss: 0.000103
Train Epoch: 26 [1100/3016 (36%)] Loss: 0.011122
Train Epoch: 26 [1200/3016 (40%)] Loss: 0.007472
Train Epoch: 26 [1300/3016 (43%)] Loss: 0.007819
Train Epoch: 26 [1400/3016 (46%)] Loss: 0.027555
Train Epoch: 26 [1500/3016 (50%)] Loss: 0.085014
Train Epoch: 26 [1600/3016 (53%)] Loss: 0.000604
Train Epoch: 26 [1700/3016 (56%)] Loss: 0.002629
Train Epoch: 26 [1800/3016 (60%)] Loss: 0.000422
Train Epoch: 26 [1900/3016 (63%)] Loss: 0.002501
Train Epoch: 26 [2000/3016 (66%)] Loss: 0.000287
Train Epoch: 26 [2100/3016 (70%)] Loss: 0.008193
Train Epoch: 26 [2200/3016 (73%)] Loss: 0.000504
Train Epoch: 26 [2300/3016 (76%)] Loss: 0.000626
Train Epoch: 26 [2400/3016 (79%)] Loss: 0.000055
Train Epoch: 26 [2500/3016 (83%)] Loss: 0.000033
Train Epoch: 26 [2600/3016 (86%)] Loss: 0.000089
Train Epoch: 26 [2700/3016 (89%)] Loss: 0.028252
Train Epoch: 26 [2800/3016 (93%)] Loss: 0.000105
Train Epoch: 26 [2900/3016 (96%)] Loss: 0.000436
Train Epoch: 26 [3000/3016 (99%)] Loss: 0.000347
Total loss = 0.002394
Dev loss is 1.105142
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment