['/home/yogeshchandrasekharuni/Developer/food-ordering/PURE/run_entity.py', '--do_train', '--do_eval', '--learning_rate=1e-5', '--task_learning_rate=5e-4', '--train_batch_size=16', '--context_window=0', '--task=food', '--data_dir=./custom_data', '--model=bert-base-uncased', '--output_dir=./custom_data_out']
Namespace(bert_model_dir=None, bertadam=False, context_window=0, data_dir='./custom_data', dev_data='./custom_data/dev.json', dev_pred_filename='ent_pred_dev.json', do_eval=True, do_train=True, eval_batch_size=32, eval_per_epoch=1, eval_test=False, learning_rate=1e-05, max_span_length=8, model='bert-base-uncased', num_epoch=100, output_dir='./custom_data_out', print_loss_step=100, seed=0, task='food', task_learning_rate=0.0005, test_data='./custom_data/test.json', test_pred_filename='ent_pred_test.json', train_batch_size=16, train_data='./custom_data/train.json', train_shuffle=False, use_albert=False, warmup_proportion=0.1)
Moving to CUDA...
# GPUs = 1
# Overlap: 0
Extracted 11284 samples from 11284 documents, with 44018 NER labels, 17.283 avg input length, 52 max length
Max Length: 52, max NER: 12
# Overlap: 0
Extracted 44972 samples from 44972 documents, with 173425 NER labels, 17.131 avg input length, 55 max length
Max Length: 55, max NER: 12
Epoch=0, iter=99, loss=243.44804
Epoch=0, iter=199, loss=235.79827
Epoch=0, iter=299, loss=233.29755
Epoch=0, iter=399, loss=202.52836
Epoch=0, iter=499, loss=118.44091
Epoch=0, iter=599, loss=41.34291
Epoch=0, iter=699, loss=25.83691
Epoch=0, iter=799, loss=24.32704
Epoch=0, iter=899, loss=23.66524
Epoch=0, iter=999, loss=23.12145
Epoch=0, iter=1099, loss=22.90508
Epoch=0, iter=1199, loss=22.97754
Epoch=0, iter=1299, loss=21.99249
Epoch=0, iter=1399, loss=22.37460
Epoch=0, iter=1499, loss=22.32461
Epoch=0, iter=1599, loss=21.33955
Epoch=0, iter=1699, loss=20.31395
Epoch=0, iter=1799, loss=19.90072
Epoch=0, iter=1899, loss=19.10657
Epoch=0, iter=1999, loss=17.86468
Epoch=0, iter=2099, loss=17.70574
Epoch=0, iter=2199, loss=15.95796
Epoch=0, iter=2299, loss=15.67962
Epoch=0, iter=2399, loss=14.26441
Epoch=0, iter=2499, loss=13.86101
Epoch=0, iter=2599, loss=12.75729
Epoch=0, iter=2699, loss=12.65991
Epoch=0, iter=2799, loss=12.01459
Evaluating...
Accuracy: 0.973822
Cor: 14929, Pred TOT: 21103, Gold TOT: 44018
P: 0.70743, R: 0.33916, F1: 0.45850
Used time: 97.782789
!!! Best valid (epoch=0): 45.85
Saving model to ./custom_data_out...
Epoch=1, iter=88, loss=11.36566
Epoch=1, iter=188, loss=10.61195
Epoch=1, iter=288, loss=10.59556
Epoch=1, iter=388, loss=9.86007
Epoch=1, iter=488, loss=9.40349
Epoch=1, iter=588, loss=9.20889
Epoch=1, iter=688, loss=8.78756
Epoch=1, iter=788, loss=8.56425
Epoch=1, iter=888, loss=8.51660
Epoch=1, iter=988, loss=7.91102
Epoch=1, iter=1088, loss=7.51398
Epoch=1, iter=1188, loss=7.35832
Epoch=1, iter=1288, loss=6.69688
Epoch=1, iter=1388, loss=6.77910
Epoch=1, iter=1488, loss=6.58863
Epoch=1, iter=1588, loss=6.22025
Epoch=1, iter=1688, loss=5.78497
Epoch=1, iter=1788, loss=5.83271
Epoch=1, iter=1888, loss=5.79998
Epoch=1, iter=1988, loss=5.42660
Epoch=1, iter=2088, loss=5.18899
Epoch=1, iter=2188, loss=4.97301
Epoch=1, iter=2288, loss=5.00884
Epoch=1, iter=2388, loss=4.51213
Epoch=1, iter=2488, loss=4.79700
Epoch=1, iter=2588, loss=4.24864
Epoch=1, iter=2688, loss=4.32532
Epoch=1, iter=2788, loss=4.28969
Evaluating...
Accuracy: 0.990703
Cor: 34145, Pred TOT: 36355, Gold TOT: 44018
P: 0.93921, R: 0.77571, F1: 0.84966
Used time: 101.091194
!!! Best valid (epoch=1): 84.97
Saving model to ./custom_data_out...
Epoch=2, iter=77, loss=4.11897
Epoch=2, iter=177, loss=3.93580
Epoch=2, iter=277, loss=4.09057
Epoch=2, iter=377, loss=4.04855
Epoch=2, iter=477, loss=3.67340
Epoch=2, iter=577, loss=3.70883
Epoch=2, iter=677, loss=3.42234
Epoch=2, iter=777, loss=3.61931
Epoch=2, iter=877, loss=3.53848
Epoch=2, iter=977, loss=3.27525
Epoch=2, iter=1077, loss=3.22880
Epoch=2, iter=1177, loss=3.33254
Epoch=2, iter=1277, loss=3.03457
Epoch=2, iter=1377, loss=3.18838
Epoch=2, iter=1477, loss=3.05527
Epoch=2, iter=1577, loss=3.08948
Epoch=2, iter=1677, loss=2.82804
Epoch=2, iter=1777, loss=3.00084
Epoch=2, iter=1877, loss=2.98393
Epoch=2, iter=1977, loss=2.79067
Epoch=2, iter=2077, loss=2.85030
Epoch=2, iter=2177, loss=2.71113
Epoch=2, iter=2277, loss=2.65620
Epoch=2, iter=2377, loss=2.55023
Epoch=2, iter=2477, loss=2.66498
Epoch=2, iter=2577, loss=2.45346
Epoch=2, iter=2677, loss=2.45250
Epoch=2, iter=2777, loss=2.49077
Evaluating...
Accuracy: 0.993408
Cor: 36716, Pred TOT: 37983, Gold TOT: 44018
P: 0.96664, R: 0.83411, F1: 0.89550
Used time: 99.946823
!!! Best valid (epoch=2): 89.55
Saving model to ./custom_data_out...
Epoch=3, iter=66, loss=2.44420
Epoch=3, iter=166, loss=2.39805
Epoch=3, iter=266, loss=2.45792
Epoch=3, iter=366, loss=2.53090
Epoch=3, iter=466, loss=2.25996
Epoch=3, iter=566, loss=2.26043
Epoch=3, iter=666, loss=2.17943
Epoch=3, iter=766, loss=2.19351
Epoch=3, iter=866, loss=2.40202
Epoch=3, iter=966, loss=2.09336
Epoch=3, iter=1066, loss=2.08938
Epoch=3, iter=1166, loss=2.16629
Epoch=3, iter=1266, loss=1.93024
Epoch=3, iter=1366, loss=2.07213
Epoch=3, iter=1466, loss=1.93386
Epoch=3, iter=1566, loss=1.99428
Epoch=3, iter=1666, loss=1.87277
Epoch=3, iter=1766, loss=1.92387
Epoch=3, iter=1866, loss=1.82625
Epoch=3, iter=1966, loss=1.82094
Epoch=3, iter=2066, loss=1.74876
Epoch=3, iter=2166, loss=1.67695
Epoch=3, iter=2266, loss=1.61129
Epoch=3, iter=2366, loss=1.68518
Epoch=3, iter=2466, loss=1.56917
Epoch=3, iter=2566, loss=1.49777
Epoch=3, iter=2666, loss=1.50936
Epoch=3, iter=2766, loss=1.59259
Evaluating...
Accuracy: 0.996364
Cor: 40160, Pred TOT: 41149, Gold TOT: 44018
P: 0.97597, R: 0.91235, F1: 0.94309
Used time: 99.399823
!!! Best valid (epoch=3): 94.31
Saving model to ./custom_data_out...
Epoch=4, iter=55, loss=1.51106
Epoch=4, iter=155, loss=1.43524
Epoch=4, iter=255, loss=1.50080
Epoch=4, iter=355, loss=1.53660
Epoch=4, iter=455, loss=1.40172
Epoch=4, iter=555, loss=1.34934
Epoch=4, iter=655, loss=1.31810
Epoch=4, iter=755, loss=1.38727
Epoch=4, iter=855, loss=1.45979
Epoch=4, iter=955, loss=1.18226
Epoch=4, iter=1055, loss=1.31470
Epoch=4, iter=1155, loss=1.33524
Epoch=4, iter=1255, loss=1.19242
Epoch=4, iter=1355, loss=1.22670
Epoch=4, iter=1455, loss=1.11627
Epoch=4, iter=1555, loss=1.17806
Epoch=4, iter=1655, loss=1.05301
Epoch=4, iter=1755, loss=1.20741
Epoch=4, iter=1855, loss=1.14403
Epoch=4, iter=1955, loss=1.10366
Epoch=4, iter=2055, loss=0.99180
Epoch=4, iter=2155, loss=1.07554
Epoch=4, iter=2255, loss=0.96035
Epoch=4, iter=2355, loss=1.06883
Epoch=4, iter=2455, loss=0.98427
Epoch=4, iter=2555, loss=1.02981
Epoch=4, iter=2655, loss=0.98175
Epoch=4, iter=2755, loss=0.96598
Evaluating...
Accuracy: 0.997638
Cor: 41425, Pred TOT: 42095, Gold TOT: 44018
P: 0.98408, R: 0.94109, F1: 0.96211
Used time: 98.526833
!!! Best valid (epoch=4): 96.21
Saving model to ./custom_data_out...
Epoch=5, iter=44, loss=1.04638
Epoch=5, iter=144, loss=0.93053
Epoch=5, iter=244, loss=0.93799
Epoch=5, iter=344, loss=0.98231
Epoch=5, iter=444, loss=0.95388
Epoch=5, iter=544, loss=0.92944
Epoch=5, iter=644, loss=0.87617
Epoch=5, iter=744, loss=0.90396
Epoch=5, iter=844, loss=0.90082
Epoch=5, iter=944, loss=0.81391
Epoch=5, iter=1044, loss=0.85680
Epoch=5, iter=1144, loss=0.90453
Epoch=5, iter=1244, loss=0.82384