2023-09-04 14:37:10,223 ---------------------------------------------------------------------------------------------------- 2023-09-04 14:37:10,224 Model: "SequenceTagger( (embeddings): TransformerWordEmbeddings( (model): BertModel( (embeddings): BertEmbeddings( (word_embeddings): Embedding(32001, 768) (position_embeddings): Embedding(512, 768) (token_type_embeddings): Embedding(2, 768) (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True) (dropout): Dropout(p=0.1, inplace=False) ) (encoder): BertEncoder( (layer): ModuleList( (0-11): 12 x BertLayer( (attention): BertAttention( (self): BertSelfAttention( (query): Linear(in_features=768, out_features=768, bias=True) (key): Linear(in_features=768, out_features=768, bias=True) (value): Linear(in_features=768, out_features=768, bias=True) (dropout): Dropout(p=0.1, inplace=False) ) (output): BertSelfOutput( (dense): Linear(in_features=768, out_features=768, bias=True) (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True) (dropout): Dropout(p=0.1, inplace=False) ) ) (intermediate): BertIntermediate( (dense): Linear(in_features=768, out_features=3072, bias=True) (intermediate_act_fn): GELUActivation() ) (output): BertOutput( (dense): Linear(in_features=3072, out_features=768, bias=True) (LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True) (dropout): Dropout(p=0.1, inplace=False) ) ) ) ) (pooler): BertPooler( (dense): Linear(in_features=768, out_features=768, bias=True) (activation): Tanh() ) ) ) (locked_dropout): LockedDropout(p=0.5) (linear): Linear(in_features=768, out_features=21, bias=True) (loss_function): CrossEntropyLoss() )" 2023-09-04 14:37:10,224 ---------------------------------------------------------------------------------------------------- 2023-09-04 14:37:10,224 MultiCorpus: 5901 train + 1287 dev + 1505 test sentences - NER_HIPE_2022 Corpus: 5901 train + 1287 dev + 1505 test sentences - /app/.flair/datasets/ner_hipe_2022/v2.1/hipe2020/fr/with_doc_seperator 2023-09-04 14:37:10,224 ---------------------------------------------------------------------------------------------------- 2023-09-04 14:37:10,224 Train: 5901 sentences 2023-09-04 14:37:10,224 (train_with_dev=False, train_with_test=False) 2023-09-04 14:37:10,224 ---------------------------------------------------------------------------------------------------- 2023-09-04 14:37:10,224 Training Params: 2023-09-04 14:37:10,225 - learning_rate: "5e-05" 2023-09-04 14:37:10,225 - mini_batch_size: "4" 2023-09-04 14:37:10,225 - max_epochs: "10" 2023-09-04 14:37:10,225 - shuffle: "True" 2023-09-04 14:37:10,225 ---------------------------------------------------------------------------------------------------- 2023-09-04 14:37:10,225 Plugins: 2023-09-04 14:37:10,225 - LinearScheduler | warmup_fraction: '0.1' 2023-09-04 14:37:10,225 ---------------------------------------------------------------------------------------------------- 2023-09-04 14:37:10,225 Final evaluation on model from best epoch (best-model.pt) 2023-09-04 14:37:10,225 - metric: "('micro avg', 'f1-score')" 2023-09-04 14:37:10,225 ---------------------------------------------------------------------------------------------------- 2023-09-04 14:37:10,225 Computation: 2023-09-04 14:37:10,225 - compute on device: cuda:0 2023-09-04 14:37:10,225 - embedding storage: none 2023-09-04 14:37:10,225 ---------------------------------------------------------------------------------------------------- 2023-09-04 14:37:10,225 Model training base path: "hmbench-hipe2020/fr-dbmdz/bert-base-historic-multilingual-cased-bs4-wsFalse-e10-lr5e-05-poolingfirst-layers-1-crfFalse-3" 2023-09-04 14:37:10,225 ---------------------------------------------------------------------------------------------------- 2023-09-04 14:37:10,225 ---------------------------------------------------------------------------------------------------- 2023-09-04 14:37:24,926 epoch 1 - iter 147/1476 - loss 2.05618754 - time (sec): 14.70 - samples/sec: 1042.48 - lr: 0.000005 - momentum: 0.000000 2023-09-04 14:37:40,726 epoch 1 - iter 294/1476 - loss 1.24591273 - time (sec): 30.50 - samples/sec: 1073.51 - lr: 0.000010 - momentum: 0.000000 2023-09-04 14:37:55,882 epoch 1 - iter 441/1476 - loss 0.96595507 - time (sec): 45.66 - samples/sec: 1053.60 - lr: 0.000015 - momentum: 0.000000 2023-09-04 14:38:11,770 epoch 1 - iter 588/1476 - loss 0.79301248 - time (sec): 61.54 - samples/sec: 1048.20 - lr: 0.000020 - momentum: 0.000000 2023-09-04 14:38:27,739 epoch 1 - iter 735/1476 - loss 0.68162646 - time (sec): 77.51 - samples/sec: 1050.21 - lr: 0.000025 - momentum: 0.000000 2023-09-04 14:38:44,370 epoch 1 - iter 882/1476 - loss 0.59898767 - time (sec): 94.14 - samples/sec: 1052.37 - lr: 0.000030 - momentum: 0.000000 2023-09-04 14:39:01,175 epoch 1 - iter 1029/1476 - loss 0.54029039 - time (sec): 110.95 - samples/sec: 1049.30 - lr: 0.000035 - momentum: 0.000000 2023-09-04 14:39:16,723 epoch 1 - iter 1176/1476 - loss 0.49996949 - time (sec): 126.50 - samples/sec: 1045.64 - lr: 0.000040 - momentum: 0.000000 2023-09-04 14:39:33,925 epoch 1 - iter 1323/1476 - loss 0.46069672 - time (sec): 143.70 - samples/sec: 1044.79 - lr: 0.000045 - momentum: 0.000000 2023-09-04 14:39:49,637 epoch 1 - iter 1470/1476 - loss 0.43367918 - time (sec): 159.41 - samples/sec: 1040.69 - lr: 0.000050 - momentum: 0.000000 2023-09-04 14:39:50,208 ---------------------------------------------------------------------------------------------------- 2023-09-04 14:39:50,208 EPOCH 1 done: loss 0.4331 - lr: 0.000050 2023-09-04 14:40:04,539 DEV : loss 0.14273743331432343 - f1-score (micro avg) 0.7029 2023-09-04 14:40:04,567 saving best model 2023-09-04 14:40:05,059 ---------------------------------------------------------------------------------------------------- 2023-09-04 14:40:20,954 epoch 2 - iter 147/1476 - loss 0.15116667 - time (sec): 15.89 - samples/sec: 1053.78 - lr: 0.000049 - momentum: 0.000000 2023-09-04 14:40:37,208 epoch 2 - iter 294/1476 - loss 0.15377335 - time (sec): 32.15 - samples/sec: 1047.39 - lr: 0.000049 - momentum: 0.000000 2023-09-04 14:40:54,506 epoch 2 - iter 441/1476 - loss 0.15144934 - time (sec): 49.44 - samples/sec: 1055.90 - lr: 0.000048 - momentum: 0.000000 2023-09-04 14:41:09,751 epoch 2 - iter 588/1476 - loss 0.14514190 - time (sec): 64.69 - samples/sec: 1048.12 - lr: 0.000048 - momentum: 0.000000 2023-09-04 14:41:24,951 epoch 2 - iter 735/1476 - loss 0.14434703 - time (sec): 79.89 - samples/sec: 1044.53 - lr: 0.000047 - momentum: 0.000000 2023-09-04 14:41:42,059 epoch 2 - iter 882/1476 - loss 0.14095490 - time (sec): 97.00 - samples/sec: 1047.48 - lr: 0.000047 - momentum: 0.000000 2023-09-04 14:41:58,137 epoch 2 - iter 1029/1476 - loss 0.13849409 - time (sec): 113.08 - samples/sec: 1045.08 - lr: 0.000046 - momentum: 0.000000 2023-09-04 14:42:13,722 epoch 2 - iter 1176/1476 - loss 0.13600706 - time (sec): 128.66 - samples/sec: 1042.32 - lr: 0.000046 - momentum: 0.000000 2023-09-04 14:42:29,232 epoch 2 - iter 1323/1476 - loss 0.13703274 - time (sec): 144.17 - samples/sec: 1040.78 - lr: 0.000045 - momentum: 0.000000 2023-09-04 14:42:44,323 epoch 2 - iter 1470/1476 - loss 0.13820196 - time (sec): 159.26 - samples/sec: 1041.95 - lr: 0.000044 - momentum: 0.000000 2023-09-04 14:42:44,869 ---------------------------------------------------------------------------------------------------- 2023-09-04 14:42:44,870 EPOCH 2 done: loss 0.1381 - lr: 0.000044 2023-09-04 14:43:02,487 DEV : loss 0.14158931374549866 - f1-score (micro avg) 0.745 2023-09-04 14:43:02,516 saving best model 2023-09-04 14:43:03,854 ---------------------------------------------------------------------------------------------------- 2023-09-04 14:43:18,808 epoch 3 - iter 147/1476 - loss 0.07909018 - time (sec): 14.95 - samples/sec: 1023.67 - lr: 0.000044 - momentum: 0.000000 2023-09-04 14:43:34,437 epoch 3 - iter 294/1476 - loss 0.08191834 - time (sec): 30.58 - samples/sec: 1045.44 - lr: 0.000043 - momentum: 0.000000 2023-09-04 14:43:50,338 epoch 3 - iter 441/1476 - loss 0.08482425 - time (sec): 46.48 - samples/sec: 1047.71 - lr: 0.000043 - momentum: 0.000000 2023-09-04 14:44:07,807 epoch 3 - iter 588/1476 - loss 0.08414229 - time (sec): 63.95 - samples/sec: 1063.02 - lr: 0.000042 - momentum: 0.000000 2023-09-04 14:44:23,918 epoch 3 - iter 735/1476 - loss 0.09109558 - time (sec): 80.06 - samples/sec: 1055.09 - lr: 0.000042 - momentum: 0.000000 2023-09-04 14:44:39,104 epoch 3 - iter 882/1476 - loss 0.08957885 - time (sec): 95.25 - samples/sec: 1049.41 - lr: 0.000041 - momentum: 0.000000 2023-09-04 14:44:55,966 epoch 3 - iter 1029/1476 - loss 0.09154050 - time (sec): 112.11 - samples/sec: 1050.53 - lr: 0.000041 - momentum: 0.000000 2023-09-04 14:45:11,162 epoch 3 - iter 1176/1476 - loss 0.09076679 - time (sec): 127.31 - samples/sec: 1047.21 - lr: 0.000040 - momentum: 0.000000 2023-09-04 14:45:27,181 epoch 3 - iter 1323/1476 - loss 0.09308992 - time (sec): 143.33 - samples/sec: 1044.41 - lr: 0.000039 - momentum: 0.000000 2023-09-04 14:45:42,897 epoch 3 - iter 1470/1476 - loss 0.09133680 - time (sec): 159.04 - samples/sec: 1042.38 - lr: 0.000039 - momentum: 0.000000 2023-09-04 14:45:43,490 ---------------------------------------------------------------------------------------------------- 2023-09-04 14:45:43,490 EPOCH 3 done: loss 0.0911 - lr: 0.000039 2023-09-04 14:46:01,014 DEV : loss 0.17265217006206512 - f1-score (micro avg) 0.8048 2023-09-04 14:46:01,044 saving best model 2023-09-04 14:46:02,399 ---------------------------------------------------------------------------------------------------- 2023-09-04 14:46:18,579 epoch 4 - iter 147/1476 - loss 0.05488287 - time (sec): 16.18 - samples/sec: 1072.38 - lr: 0.000038 - momentum: 0.000000 2023-09-04 14:46:34,303 epoch 4 - iter 294/1476 - loss 0.05976011 - time (sec): 31.90 - samples/sec: 1069.41 - lr: 0.000038 - momentum: 0.000000 2023-09-04 14:46:50,518 epoch 4 - iter 441/1476 - loss 0.06373628 - time (sec): 48.12 - samples/sec: 1061.96 - lr: 0.000037 - momentum: 0.000000 2023-09-04 14:47:06,011 epoch 4 - iter 588/1476 - loss 0.06645004 - time (sec): 63.61 - samples/sec: 1051.93 - lr: 0.000037 - momentum: 0.000000 2023-09-04 14:47:21,374 epoch 4 - iter 735/1476 - loss 0.06599255 - time (sec): 78.97 - samples/sec: 1038.79 - lr: 0.000036 - momentum: 0.000000 2023-09-04 14:47:38,022 epoch 4 - iter 882/1476 - loss 0.06576644 - time (sec): 95.62 - samples/sec: 1036.93 - lr: 0.000036 - momentum: 0.000000 2023-09-04 14:47:55,574 epoch 4 - iter 1029/1476 - loss 0.06729119 - time (sec): 113.17 - samples/sec: 1046.71 - lr: 0.000035 - momentum: 0.000000 2023-09-04 14:48:10,738 epoch 4 - iter 1176/1476 - loss 0.06894602 - time (sec): 128.34 - samples/sec: 1041.22 - lr: 0.000034 - momentum: 0.000000 2023-09-04 14:48:26,358 epoch 4 - iter 1323/1476 - loss 0.06675886 - time (sec): 143.96 - samples/sec: 1037.88 - lr: 0.000034 - momentum: 0.000000 2023-09-04 14:48:42,090 epoch 4 - iter 1470/1476 - loss 0.06721274 - time (sec): 159.69 - samples/sec: 1039.16 - lr: 0.000033 - momentum: 0.000000 2023-09-04 14:48:42,677 ---------------------------------------------------------------------------------------------------- 2023-09-04 14:48:42,677 EPOCH 4 done: loss 0.0671 - lr: 0.000033 2023-09-04 14:49:00,334 DEV : loss 0.16335846483707428 - f1-score (micro avg) 0.8012 2023-09-04 14:49:00,363 ---------------------------------------------------------------------------------------------------- 2023-09-04 14:49:17,015 epoch 5 - iter 147/1476 - loss 0.06622027 - time (sec): 16.65 - samples/sec: 1101.78 - lr: 0.000033 - momentum: 0.000000 2023-09-04 14:49:33,050 epoch 5 - iter 294/1476 - loss 0.06049103 - time (sec): 32.69 - samples/sec: 1066.31 - lr: 0.000032 - momentum: 0.000000 2023-09-04 14:49:49,544 epoch 5 - iter 441/1476 - loss 0.06035344 - time (sec): 49.18 - samples/sec: 1053.86 - lr: 0.000032 - momentum: 0.000000 2023-09-04 14:50:04,750 epoch 5 - iter 588/1476 - loss 0.05825611 - time (sec): 64.39 - samples/sec: 1044.43 - lr: 0.000031 - momentum: 0.000000 2023-09-04 14:50:20,107 epoch 5 - iter 735/1476 - loss 0.05285723 - time (sec): 79.74 - samples/sec: 1040.13 - lr: 0.000031 - momentum: 0.000000 2023-09-04 14:50:35,859 epoch 5 - iter 882/1476 - loss 0.05168682 - time (sec): 95.50 - samples/sec: 1033.23 - lr: 0.000030 - momentum: 0.000000 2023-09-04 14:50:51,736 epoch 5 - iter 1029/1476 - loss 0.05200072 - time (sec): 111.37 - samples/sec: 1033.81 - lr: 0.000029 - momentum: 0.000000 2023-09-04 14:51:07,167 epoch 5 - iter 1176/1476 - loss 0.05045599 - time (sec): 126.80 - samples/sec: 1030.80 - lr: 0.000029 - momentum: 0.000000 2023-09-04 14:51:22,293 epoch 5 - iter 1323/1476 - loss 0.04917480 - time (sec): 141.93 - samples/sec: 1032.43 - lr: 0.000028 - momentum: 0.000000 2023-09-04 14:51:39,898 epoch 5 - iter 1470/1476 - loss 0.04897805 - time (sec): 159.53 - samples/sec: 1039.41 - lr: 0.000028 - momentum: 0.000000 2023-09-04 14:51:40,447 ---------------------------------------------------------------------------------------------------- 2023-09-04 14:51:40,447 EPOCH 5 done: loss 0.0493 - lr: 0.000028 2023-09-04 14:51:58,018 DEV : loss 0.20086710155010223 - f1-score (micro avg) 0.8012 2023-09-04 14:51:58,049 ---------------------------------------------------------------------------------------------------- 2023-09-04 14:52:14,830 epoch 6 - iter 147/1476 - loss 0.04119602 - time (sec): 16.78 - samples/sec: 1079.19 - lr: 0.000027 - momentum: 0.000000 2023-09-04 14:52:31,506 epoch 6 - iter 294/1476 - loss 0.03794587 - time (sec): 33.46 - samples/sec: 1062.89 - lr: 0.000027 - momentum: 0.000000 2023-09-04 14:52:48,721 epoch 6 - iter 441/1476 - loss 0.03922280 - time (sec): 50.67 - samples/sec: 1069.35 - lr: 0.000026 - momentum: 0.000000 2023-09-04 14:53:04,208 epoch 6 - iter 588/1476 - loss 0.03951086 - time (sec): 66.16 - samples/sec: 1059.47 - lr: 0.000026 - momentum: 0.000000 2023-09-04 14:53:19,700 epoch 6 - iter 735/1476 - loss 0.03815692 - time (sec): 81.65 - samples/sec: 1054.77 - lr: 0.000025 - momentum: 0.000000 2023-09-04 14:53:35,869 epoch 6 - iter 882/1476 - loss 0.03838285 - time (sec): 97.82 - samples/sec: 1048.69 - lr: 0.000024 - momentum: 0.000000 2023-09-04 14:53:51,797 epoch 6 - iter 1029/1476 - loss 0.03735327 - time (sec): 113.75 - samples/sec: 1043.81 - lr: 0.000024 - momentum: 0.000000 2023-09-04 14:54:06,862 epoch 6 - iter 1176/1476 - loss 0.03668080 - time (sec): 128.81 - samples/sec: 1039.34 - lr: 0.000023 - momentum: 0.000000 2023-09-04 14:54:22,232 epoch 6 - iter 1323/1476 - loss 0.03692170 - time (sec): 144.18 - samples/sec: 1037.47 - lr: 0.000023 - momentum: 0.000000 2023-09-04 14:54:37,600 epoch 6 - iter 1470/1476 - loss 0.03657753 - time (sec): 159.55 - samples/sec: 1039.55 - lr: 0.000022 - momentum: 0.000000 2023-09-04 14:54:38,166 ---------------------------------------------------------------------------------------------------- 2023-09-04 14:54:38,167 EPOCH 6 done: loss 0.0365 - lr: 0.000022 2023-09-04 14:54:55,735 DEV : loss 0.2203623205423355 - f1-score (micro avg) 0.8002 2023-09-04 14:54:55,763 ---------------------------------------------------------------------------------------------------- 2023-09-04 14:55:10,434 epoch 7 - iter 147/1476 - loss 0.01642304 - time (sec): 14.67 - samples/sec: 1057.19 - lr: 0.000022 - momentum: 0.000000 2023-09-04 14:55:27,264 epoch 7 - iter 294/1476 - loss 0.01866675 - time (sec): 31.50 - samples/sec: 1055.62 - lr: 0.000021 - momentum: 0.000000 2023-09-04 14:55:41,753 epoch 7 - iter 441/1476 - loss 0.02584884 - time (sec): 45.99 - samples/sec: 1044.53 - lr: 0.000021 - momentum: 0.000000 2023-09-04 14:55:58,660 epoch 7 - iter 588/1476 - loss 0.02482793 - time (sec): 62.90 - samples/sec: 1044.61 - lr: 0.000020 - momentum: 0.000000 2023-09-04 14:56:13,767 epoch 7 - iter 735/1476 - loss 0.02488488 - time (sec): 78.00 - samples/sec: 1040.00 - lr: 0.000019 - momentum: 0.000000 2023-09-04 14:56:28,980 epoch 7 - iter 882/1476 - loss 0.02516712 - time (sec): 93.22 - samples/sec: 1035.35 - lr: 0.000019 - momentum: 0.000000 2023-09-04 14:56:48,543 epoch 7 - iter 1029/1476 - loss 0.02511757 - time (sec): 112.78 - samples/sec: 1050.84 - lr: 0.000018 - momentum: 0.000000 2023-09-04 14:57:03,738 epoch 7 - iter 1176/1476 - loss 0.02533099 - time (sec): 127.97 - samples/sec: 1045.97 - lr: 0.000018 - momentum: 0.000000 2023-09-04 14:57:19,755 epoch 7 - iter 1323/1476 - loss 0.02626783 - time (sec): 143.99 - samples/sec: 1043.52 - lr: 0.000017 - momentum: 0.000000 2023-09-04 14:57:35,009 epoch 7 - iter 1470/1476 - loss 0.02547683 - time (sec): 159.24 - samples/sec: 1041.10 - lr: 0.000017 - momentum: 0.000000 2023-09-04 14:57:35,579 ---------------------------------------------------------------------------------------------------- 2023-09-04 14:57:35,579 EPOCH 7 done: loss 0.0255 - lr: 0.000017 2023-09-04 14:57:53,132 DEV : loss 0.22496868669986725 - f1-score (micro avg) 0.8145 2023-09-04 14:57:53,160 saving best model 2023-09-04 14:57:54,555 ---------------------------------------------------------------------------------------------------- 2023-09-04 14:58:10,369 epoch 8 - iter 147/1476 - loss 0.01023728 - time (sec): 15.81 - samples/sec: 1072.20 - lr: 0.000016 - momentum: 0.000000 2023-09-04 14:58:26,766 epoch 8 - iter 294/1476 - loss 0.01604023 - time (sec): 32.21 - samples/sec: 1066.02 - lr: 0.000016 - momentum: 0.000000 2023-09-04 14:58:43,662 epoch 8 - iter 441/1476 - loss 0.02054301 - time (sec): 49.11 - samples/sec: 1072.53 - lr: 0.000015 - momentum: 0.000000 2023-09-04 14:58:59,911 epoch 8 - iter 588/1476 - loss 0.02120702 - time (sec): 65.35 - samples/sec: 1066.61 - lr: 0.000014 - momentum: 0.000000 2023-09-04 14:59:15,831 epoch 8 - iter 735/1476 - loss 0.01961425 - time (sec): 81.27 - samples/sec: 1057.21 - lr: 0.000014 - momentum: 0.000000 2023-09-04 14:59:32,105 epoch 8 - iter 882/1476 - loss 0.01762423 - time (sec): 97.55 - samples/sec: 1056.12 - lr: 0.000013 - momentum: 0.000000 2023-09-04 14:59:47,161 epoch 8 - iter 1029/1476 - loss 0.01750371 - time (sec): 112.60 - samples/sec: 1053.32 - lr: 0.000013 - momentum: 0.000000 2023-09-04 15:00:02,623 epoch 8 - iter 1176/1476 - loss 0.01730821 - time (sec): 128.07 - samples/sec: 1047.60 - lr: 0.000012 - momentum: 0.000000 2023-09-04 15:00:17,134 epoch 8 - iter 1323/1476 - loss 0.01677759 - time (sec): 142.58 - samples/sec: 1043.05 - lr: 0.000012 - momentum: 0.000000 2023-09-04 15:00:33,610 epoch 8 - iter 1470/1476 - loss 0.01729207 - time (sec): 159.05 - samples/sec: 1042.65 - lr: 0.000011 - momentum: 0.000000 2023-09-04 15:00:34,162 ---------------------------------------------------------------------------------------------------- 2023-09-04 15:00:34,162 EPOCH 8 done: loss 0.0172 - lr: 0.000011 2023-09-04 15:00:51,816 DEV : loss 0.20783096551895142 - f1-score (micro avg) 0.824 2023-09-04 15:00:51,845 saving best model 2023-09-04 15:00:53,192 ---------------------------------------------------------------------------------------------------- 2023-09-04 15:01:10,602 epoch 9 - iter 147/1476 - loss 0.01460204 - time (sec): 17.41 - samples/sec: 1033.48 - lr: 0.000011 - momentum: 0.000000 2023-09-04 15:01:26,925 epoch 9 - iter 294/1476 - loss 0.01077039 - time (sec): 33.73 - samples/sec: 1040.12 - lr: 0.000010 - momentum: 0.000000 2023-09-04 15:01:42,542 epoch 9 - iter 441/1476 - loss 0.01075809 - time (sec): 49.35 - samples/sec: 1025.31 - lr: 0.000009 - momentum: 0.000000 2023-09-04 15:01:58,486 epoch 9 - iter 588/1476 - loss 0.01065765 - time (sec): 65.29 - samples/sec: 1027.65 - lr: 0.000009 - momentum: 0.000000 2023-09-04 15:02:13,294 epoch 9 - iter 735/1476 - loss 0.01255226 - time (sec): 80.10 - samples/sec: 1020.74 - lr: 0.000008 - momentum: 0.000000 2023-09-04 15:02:29,925 epoch 9 - iter 882/1476 - loss 0.01202051 - time (sec): 96.73 - samples/sec: 1023.62 - lr: 0.000008 - momentum: 0.000000 2023-09-04 15:02:46,444 epoch 9 - iter 1029/1476 - loss 0.01258727 - time (sec): 113.25 - samples/sec: 1030.10 - lr: 0.000007 - momentum: 0.000000 2023-09-04 15:03:01,498 epoch 9 - iter 1176/1476 - loss 0.01267641 - time (sec): 128.30 - samples/sec: 1035.35 - lr: 0.000007 - momentum: 0.000000 2023-09-04 15:03:16,907 epoch 9 - iter 1323/1476 - loss 0.01223326 - time (sec): 143.71 - samples/sec: 1036.81 - lr: 0.000006 - momentum: 0.000000 2023-09-04 15:03:33,731 epoch 9 - iter 1470/1476 - loss 0.01227625 - time (sec): 160.54 - samples/sec: 1033.07 - lr: 0.000006 - momentum: 0.000000 2023-09-04 15:03:34,325 ---------------------------------------------------------------------------------------------------- 2023-09-04 15:03:34,325 EPOCH 9 done: loss 0.0122 - lr: 0.000006 2023-09-04 15:03:52,069 DEV : loss 0.21233657002449036 - f1-score (micro avg) 0.8289 2023-09-04 15:03:52,098 saving best model 2023-09-04 15:03:53,446 ---------------------------------------------------------------------------------------------------- 2023-09-04 15:04:09,986 epoch 10 - iter 147/1476 - loss 0.00749491 - time (sec): 16.54 - samples/sec: 1086.47 - lr: 0.000005 - momentum: 0.000000 2023-09-04 15:04:25,082 epoch 10 - iter 294/1476 - loss 0.00470430 - time (sec): 31.64 - samples/sec: 1045.80 - lr: 0.000004 - momentum: 0.000000 2023-09-04 15:04:41,524 epoch 10 - iter 441/1476 - loss 0.00519274 - time (sec): 48.08 - samples/sec: 1044.30 - lr: 0.000004 - momentum: 0.000000 2023-09-04 15:04:57,211 epoch 10 - iter 588/1476 - loss 0.00563998 - time (sec): 63.76 - samples/sec: 1028.16 - lr: 0.000003 - momentum: 0.000000 2023-09-04 15:05:14,292 epoch 10 - iter 735/1476 - loss 0.00672732 - time (sec): 80.85 - samples/sec: 1045.90 - lr: 0.000003 - momentum: 0.000000 2023-09-04 15:05:29,635 epoch 10 - iter 882/1476 - loss 0.00659474 - time (sec): 96.19 - samples/sec: 1044.69 - lr: 0.000002 - momentum: 0.000000 2023-09-04 15:05:45,611 epoch 10 - iter 1029/1476 - loss 0.00694730 - time (sec): 112.16 - samples/sec: 1044.52 - lr: 0.000002 - momentum: 0.000000 2023-09-04 15:06:01,416 epoch 10 - iter 1176/1476 - loss 0.00708200 - time (sec): 127.97 - samples/sec: 1049.01 - lr: 0.000001 - momentum: 0.000000 2023-09-04 15:06:16,546 epoch 10 - iter 1323/1476 - loss 0.00670365 - time (sec): 143.10 - samples/sec: 1047.88 - lr: 0.000001 - momentum: 0.000000 2023-09-04 15:06:32,906 epoch 10 - iter 1470/1476 - loss 0.00690620 - time (sec): 159.46 - samples/sec: 1040.07 - lr: 0.000000 - momentum: 0.000000 2023-09-04 15:06:33,489 ---------------------------------------------------------------------------------------------------- 2023-09-04 15:06:33,489 EPOCH 10 done: loss 0.0069 - lr: 0.000000 2023-09-04 15:06:51,186 DEV : loss 0.22755871713161469 - f1-score (micro avg) 0.8257 2023-09-04 15:06:51,686 ---------------------------------------------------------------------------------------------------- 2023-09-04 15:06:51,687 Loading model from best epoch ... 2023-09-04 15:06:53,562 SequenceTagger predicts: Dictionary with 21 tags: O, S-loc, B-loc, E-loc, I-loc, S-pers, B-pers, E-pers, I-pers, S-org, B-org, E-org, I-org, S-time, B-time, E-time, I-time, S-prod, B-prod, E-prod, I-prod 2023-09-04 15:07:08,305 Results: - F-score (micro) 0.7862 - F-score (macro) 0.6798 - Accuracy 0.6733 By class: precision recall f1-score support loc 0.8512 0.8531 0.8522 858 pers 0.7548 0.8026 0.7780 537 org 0.5461 0.5833 0.5641 132 time 0.5224 0.6481 0.5785 54 prod 0.6667 0.5902 0.6261 61 micro avg 0.7744 0.7984 0.7862 1642 macro avg 0.6682 0.6955 0.6798 1642 weighted avg 0.7775 0.7984 0.7873 1642 2023-09-04 15:07:08,305 ----------------------------------------------------------------------------------------------------