2023-10-17 11:07:44,037 ---------------------------------------------------------------------------------------------------- 2023-10-17 11:07:44,040 Model: "SequenceTagger( (embeddings): TransformerWordEmbeddings( (model): ElectraModel( (embeddings): ElectraEmbeddings( (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): ElectraEncoder( (layer): ModuleList( (0-11): 12 x ElectraLayer( (attention): ElectraAttention( (self): ElectraSelfAttention( (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): ElectraSelfOutput( (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): ElectraIntermediate( (dense): Linear(in_features=768, out_features=3072, bias=True) (intermediate_act_fn): GELUActivation() ) (output): ElectraOutput( (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) ) ) ) ) ) ) (locked_dropout): LockedDropout(p=0.5) (linear): Linear(in_features=768, out_features=17, bias=True) (loss_function): CrossEntropyLoss() )" 2023-10-17 11:07:44,040 ---------------------------------------------------------------------------------------------------- 2023-10-17 11:07:44,040 MultiCorpus: 20847 train + 1123 dev + 3350 test sentences - NER_HIPE_2022 Corpus: 20847 train + 1123 dev + 3350 test sentences - /root/.flair/datasets/ner_hipe_2022/v2.1/newseye/de/with_doc_seperator 2023-10-17 11:07:44,040 ---------------------------------------------------------------------------------------------------- 2023-10-17 11:07:44,041 Train: 20847 sentences 2023-10-17 11:07:44,041 (train_with_dev=False, train_with_test=False) 2023-10-17 11:07:44,041 ---------------------------------------------------------------------------------------------------- 2023-10-17 11:07:44,041 Training Params: 2023-10-17 11:07:44,041 - learning_rate: "3e-05" 2023-10-17 11:07:44,041 - mini_batch_size: "4" 2023-10-17 11:07:44,041 - max_epochs: "10" 2023-10-17 11:07:44,041 - shuffle: "True" 2023-10-17 11:07:44,041 ---------------------------------------------------------------------------------------------------- 2023-10-17 11:07:44,041 Plugins: 2023-10-17 11:07:44,041 - TensorboardLogger 2023-10-17 11:07:44,042 - LinearScheduler | warmup_fraction: '0.1' 2023-10-17 11:07:44,042 ---------------------------------------------------------------------------------------------------- 2023-10-17 11:07:44,042 Final evaluation on model from best epoch (best-model.pt) 2023-10-17 11:07:44,042 - metric: "('micro avg', 'f1-score')" 2023-10-17 11:07:44,042 ---------------------------------------------------------------------------------------------------- 2023-10-17 11:07:44,042 Computation: 2023-10-17 11:07:44,042 - compute on device: cuda:0 2023-10-17 11:07:44,042 - embedding storage: none 2023-10-17 11:07:44,042 ---------------------------------------------------------------------------------------------------- 2023-10-17 11:07:44,042 Model training base path: "hmbench-newseye/de-hmteams/teams-base-historic-multilingual-discriminator-bs4-wsFalse-e10-lr3e-05-poolingfirst-layers-1-crfFalse-1" 2023-10-17 11:07:44,042 ---------------------------------------------------------------------------------------------------- 2023-10-17 11:07:44,043 ---------------------------------------------------------------------------------------------------- 2023-10-17 11:07:44,043 Logging anything other than scalars to TensorBoard is currently not supported. 2023-10-17 11:08:26,363 epoch 1 - iter 521/5212 - loss 1.89163776 - time (sec): 42.32 - samples/sec: 798.01 - lr: 0.000003 - momentum: 0.000000 2023-10-17 11:09:08,855 epoch 1 - iter 1042/5212 - loss 1.13479704 - time (sec): 84.81 - samples/sec: 817.76 - lr: 0.000006 - momentum: 0.000000 2023-10-17 11:09:54,477 epoch 1 - iter 1563/5212 - loss 0.84671118 - time (sec): 130.43 - samples/sec: 826.16 - lr: 0.000009 - momentum: 0.000000 2023-10-17 11:10:37,574 epoch 1 - iter 2084/5212 - loss 0.69980888 - time (sec): 173.53 - samples/sec: 838.22 - lr: 0.000012 - momentum: 0.000000 2023-10-17 11:11:21,695 epoch 1 - iter 2605/5212 - loss 0.60610850 - time (sec): 217.65 - samples/sec: 848.09 - lr: 0.000015 - momentum: 0.000000 2023-10-17 11:12:04,464 epoch 1 - iter 3126/5212 - loss 0.54043037 - time (sec): 260.42 - samples/sec: 857.81 - lr: 0.000018 - momentum: 0.000000 2023-10-17 11:12:47,850 epoch 1 - iter 3647/5212 - loss 0.49949347 - time (sec): 303.81 - samples/sec: 851.51 - lr: 0.000021 - momentum: 0.000000 2023-10-17 11:13:31,805 epoch 1 - iter 4168/5212 - loss 0.46823322 - time (sec): 347.76 - samples/sec: 842.57 - lr: 0.000024 - momentum: 0.000000 2023-10-17 11:14:16,895 epoch 1 - iter 4689/5212 - loss 0.43915469 - time (sec): 392.85 - samples/sec: 842.42 - lr: 0.000027 - momentum: 0.000000 2023-10-17 11:15:00,022 epoch 1 - iter 5210/5212 - loss 0.41555785 - time (sec): 435.98 - samples/sec: 842.27 - lr: 0.000030 - momentum: 0.000000 2023-10-17 11:15:00,190 ---------------------------------------------------------------------------------------------------- 2023-10-17 11:15:00,190 EPOCH 1 done: loss 0.4153 - lr: 0.000030 2023-10-17 11:15:07,630 DEV : loss 0.11844930797815323 - f1-score (micro avg) 0.2469 2023-10-17 11:15:07,684 saving best model 2023-10-17 11:15:08,240 ---------------------------------------------------------------------------------------------------- 2023-10-17 11:15:51,077 epoch 2 - iter 521/5212 - loss 0.18677525 - time (sec): 42.84 - samples/sec: 893.85 - lr: 0.000030 - momentum: 0.000000 2023-10-17 11:16:34,185 epoch 2 - iter 1042/5212 - loss 0.18467584 - time (sec): 85.94 - samples/sec: 867.54 - lr: 0.000029 - momentum: 0.000000 2023-10-17 11:17:17,438 epoch 2 - iter 1563/5212 - loss 0.18459225 - time (sec): 129.20 - samples/sec: 868.65 - lr: 0.000029 - momentum: 0.000000 2023-10-17 11:18:00,970 epoch 2 - iter 2084/5212 - loss 0.18949291 - time (sec): 172.73 - samples/sec: 854.13 - lr: 0.000029 - momentum: 0.000000 2023-10-17 11:18:45,977 epoch 2 - iter 2605/5212 - loss 0.18960565 - time (sec): 217.74 - samples/sec: 842.99 - lr: 0.000028 - momentum: 0.000000 2023-10-17 11:19:29,280 epoch 2 - iter 3126/5212 - loss 0.18834557 - time (sec): 261.04 - samples/sec: 840.64 - lr: 0.000028 - momentum: 0.000000 2023-10-17 11:20:11,948 epoch 2 - iter 3647/5212 - loss 0.18442261 - time (sec): 303.71 - samples/sec: 854.40 - lr: 0.000028 - momentum: 0.000000 2023-10-17 11:20:55,872 epoch 2 - iter 4168/5212 - loss 0.18307206 - time (sec): 347.63 - samples/sec: 851.42 - lr: 0.000027 - momentum: 0.000000 2023-10-17 11:21:38,580 epoch 2 - iter 4689/5212 - loss 0.17918821 - time (sec): 390.34 - samples/sec: 844.82 - lr: 0.000027 - momentum: 0.000000 2023-10-17 11:22:20,682 epoch 2 - iter 5210/5212 - loss 0.17650046 - time (sec): 432.44 - samples/sec: 849.48 - lr: 0.000027 - momentum: 0.000000 2023-10-17 11:22:20,832 ---------------------------------------------------------------------------------------------------- 2023-10-17 11:22:20,833 EPOCH 2 done: loss 0.1765 - lr: 0.000027 2023-10-17 11:22:32,839 DEV : loss 0.23894941806793213 - f1-score (micro avg) 0.3469 2023-10-17 11:22:32,893 saving best model 2023-10-17 11:22:34,316 ---------------------------------------------------------------------------------------------------- 2023-10-17 11:23:15,325 epoch 3 - iter 521/5212 - loss 0.11437386 - time (sec): 41.01 - samples/sec: 925.03 - lr: 0.000026 - momentum: 0.000000 2023-10-17 11:23:57,365 epoch 3 - iter 1042/5212 - loss 0.12459624 - time (sec): 83.04 - samples/sec: 902.84 - lr: 0.000026 - momentum: 0.000000 2023-10-17 11:24:39,204 epoch 3 - iter 1563/5212 - loss 0.12923542 - time (sec): 124.88 - samples/sec: 889.63 - lr: 0.000026 - momentum: 0.000000 2023-10-17 11:25:20,547 epoch 3 - iter 2084/5212 - loss 0.13146091 - time (sec): 166.23 - samples/sec: 882.27 - lr: 0.000025 - momentum: 0.000000 2023-10-17 11:26:01,412 epoch 3 - iter 2605/5212 - loss 0.12864210 - time (sec): 207.09 - samples/sec: 886.34 - lr: 0.000025 - momentum: 0.000000 2023-10-17 11:26:42,942 epoch 3 - iter 3126/5212 - loss 0.13399268 - time (sec): 248.62 - samples/sec: 875.92 - lr: 0.000025 - momentum: 0.000000 2023-10-17 11:27:25,033 epoch 3 - iter 3647/5212 - loss 0.13321294 - time (sec): 290.71 - samples/sec: 874.05 - lr: 0.000024 - momentum: 0.000000 2023-10-17 11:28:06,996 epoch 3 - iter 4168/5212 - loss 0.13283195 - time (sec): 332.68 - samples/sec: 874.06 - lr: 0.000024 - momentum: 0.000000 2023-10-17 11:28:50,011 epoch 3 - iter 4689/5212 - loss 0.13502035 - time (sec): 375.69 - samples/sec: 878.62 - lr: 0.000024 - momentum: 0.000000 2023-10-17 11:29:31,516 epoch 3 - iter 5210/5212 - loss 0.13202538 - time (sec): 417.20 - samples/sec: 880.09 - lr: 0.000023 - momentum: 0.000000 2023-10-17 11:29:31,671 ---------------------------------------------------------------------------------------------------- 2023-10-17 11:29:31,671 EPOCH 3 done: loss 0.1319 - lr: 0.000023 2023-10-17 11:29:43,652 DEV : loss 0.24874247610569 - f1-score (micro avg) 0.351 2023-10-17 11:29:43,706 saving best model 2023-10-17 11:29:45,126 ---------------------------------------------------------------------------------------------------- 2023-10-17 11:30:29,111 epoch 4 - iter 521/5212 - loss 0.09645922 - time (sec): 43.98 - samples/sec: 847.79 - lr: 0.000023 - momentum: 0.000000 2023-10-17 11:31:14,228 epoch 4 - iter 1042/5212 - loss 0.09575057 - time (sec): 89.10 - samples/sec: 827.29 - lr: 0.000023 - momentum: 0.000000 2023-10-17 11:31:56,651 epoch 4 - iter 1563/5212 - loss 0.09440810 - time (sec): 131.52 - samples/sec: 826.16 - lr: 0.000022 - momentum: 0.000000 2023-10-17 11:32:38,800 epoch 4 - iter 2084/5212 - loss 0.09200648 - time (sec): 173.67 - samples/sec: 830.24 - lr: 0.000022 - momentum: 0.000000 2023-10-17 11:33:22,254 epoch 4 - iter 2605/5212 - loss 0.09507978 - time (sec): 217.12 - samples/sec: 824.50 - lr: 0.000022 - momentum: 0.000000 2023-10-17 11:34:03,455 epoch 4 - iter 3126/5212 - loss 0.09574149 - time (sec): 258.33 - samples/sec: 827.19 - lr: 0.000021 - momentum: 0.000000 2023-10-17 11:34:45,413 epoch 4 - iter 3647/5212 - loss 0.09612564 - time (sec): 300.28 - samples/sec: 839.08 - lr: 0.000021 - momentum: 0.000000 2023-10-17 11:35:28,560 epoch 4 - iter 4168/5212 - loss 0.09673298 - time (sec): 343.43 - samples/sec: 846.54 - lr: 0.000021 - momentum: 0.000000 2023-10-17 11:36:11,639 epoch 4 - iter 4689/5212 - loss 0.09605759 - time (sec): 386.51 - samples/sec: 853.19 - lr: 0.000020 - momentum: 0.000000 2023-10-17 11:36:54,277 epoch 4 - iter 5210/5212 - loss 0.09435830 - time (sec): 429.15 - samples/sec: 855.98 - lr: 0.000020 - momentum: 0.000000 2023-10-17 11:36:54,441 ---------------------------------------------------------------------------------------------------- 2023-10-17 11:36:54,441 EPOCH 4 done: loss 0.0943 - lr: 0.000020 2023-10-17 11:37:06,587 DEV : loss 0.2750011384487152 - f1-score (micro avg) 0.3813 2023-10-17 11:37:06,641 saving best model 2023-10-17 11:37:08,118 ---------------------------------------------------------------------------------------------------- 2023-10-17 11:37:52,422 epoch 5 - iter 521/5212 - loss 0.05806410 - time (sec): 44.30 - samples/sec: 850.98 - lr: 0.000020 - momentum: 0.000000 2023-10-17 11:38:35,833 epoch 5 - iter 1042/5212 - loss 0.05768321 - time (sec): 87.71 - samples/sec: 816.42 - lr: 0.000019 - momentum: 0.000000 2023-10-17 11:39:21,545 epoch 5 - iter 1563/5212 - loss 0.06370432 - time (sec): 133.42 - samples/sec: 813.72 - lr: 0.000019 - momentum: 0.000000 2023-10-17 11:40:04,393 epoch 5 - iter 2084/5212 - loss 0.06198607 - time (sec): 176.27 - samples/sec: 810.85 - lr: 0.000019 - momentum: 0.000000 2023-10-17 11:40:50,143 epoch 5 - iter 2605/5212 - loss 0.06397044 - time (sec): 222.02 - samples/sec: 820.57 - lr: 0.000018 - momentum: 0.000000 2023-10-17 11:41:34,615 epoch 5 - iter 3126/5212 - loss 0.06343104 - time (sec): 266.49 - samples/sec: 834.78 - lr: 0.000018 - momentum: 0.000000 2023-10-17 11:42:17,862 epoch 5 - iter 3647/5212 - loss 0.06370732 - time (sec): 309.74 - samples/sec: 834.87 - lr: 0.000018 - momentum: 0.000000 2023-10-17 11:43:00,416 epoch 5 - iter 4168/5212 - loss 0.06361989 - time (sec): 352.29 - samples/sec: 841.76 - lr: 0.000017 - momentum: 0.000000 2023-10-17 11:43:41,207 epoch 5 - iter 4689/5212 - loss 0.06404810 - time (sec): 393.08 - samples/sec: 842.08 - lr: 0.000017 - momentum: 0.000000 2023-10-17 11:44:23,171 epoch 5 - iter 5210/5212 - loss 0.06343811 - time (sec): 435.05 - samples/sec: 844.47 - lr: 0.000017 - momentum: 0.000000 2023-10-17 11:44:23,319 ---------------------------------------------------------------------------------------------------- 2023-10-17 11:44:23,320 EPOCH 5 done: loss 0.0635 - lr: 0.000017 2023-10-17 11:44:34,163 DEV : loss 0.34400203824043274 - f1-score (micro avg) 0.3937 2023-10-17 11:44:34,220 saving best model 2023-10-17 11:44:35,623 ---------------------------------------------------------------------------------------------------- 2023-10-17 11:45:19,252 epoch 6 - iter 521/5212 - loss 0.05472897 - time (sec): 43.62 - samples/sec: 855.22 - lr: 0.000016 - momentum: 0.000000 2023-10-17 11:46:00,081 epoch 6 - iter 1042/5212 - loss 0.05370031 - time (sec): 84.45 - samples/sec: 855.73 - lr: 0.000016 - momentum: 0.000000 2023-10-17 11:46:41,972 epoch 6 - iter 1563/5212 - loss 0.04710872 - time (sec): 126.34 - samples/sec: 849.29 - lr: 0.000016 - momentum: 0.000000 2023-10-17 11:47:27,063 epoch 6 - iter 2084/5212 - loss 0.04824334 - time (sec): 171.44 - samples/sec: 855.69 - lr: 0.000015 - momentum: 0.000000 2023-10-17 11:48:09,332 epoch 6 - iter 2605/5212 - loss 0.04632066 - time (sec): 213.70 - samples/sec: 868.35 - lr: 0.000015 - momentum: 0.000000 2023-10-17 11:48:50,366 epoch 6 - iter 3126/5212 - loss 0.04608767 - time (sec): 254.74 - samples/sec: 874.91 - lr: 0.000015 - momentum: 0.000000 2023-10-17 11:49:32,284 epoch 6 - iter 3647/5212 - loss 0.04516211 - time (sec): 296.66 - samples/sec: 871.68 - lr: 0.000014 - momentum: 0.000000 2023-10-17 11:50:13,897 epoch 6 - iter 4168/5212 - loss 0.04805567 - time (sec): 338.27 - samples/sec: 869.50 - lr: 0.000014 - momentum: 0.000000 2023-10-17 11:50:56,045 epoch 6 - iter 4689/5212 - loss 0.04758346 - time (sec): 380.42 - samples/sec: 871.32 - lr: 0.000014 - momentum: 0.000000 2023-10-17 11:51:37,913 epoch 6 - iter 5210/5212 - loss 0.04826135 - time (sec): 422.29 - samples/sec: 869.98 - lr: 0.000013 - momentum: 0.000000 2023-10-17 11:51:38,067 ---------------------------------------------------------------------------------------------------- 2023-10-17 11:51:38,068 EPOCH 6 done: loss 0.0483 - lr: 0.000013 2023-10-17 11:51:49,240 DEV : loss 0.2987309396266937 - f1-score (micro avg) 0.3914 2023-10-17 11:51:49,296 ---------------------------------------------------------------------------------------------------- 2023-10-17 11:52:31,138 epoch 7 - iter 521/5212 - loss 0.03418427 - time (sec): 41.84 - samples/sec: 904.38 - lr: 0.000013 - momentum: 0.000000 2023-10-17 11:53:13,452 epoch 7 - iter 1042/5212 - loss 0.03052773 - time (sec): 84.15 - samples/sec: 895.56 - lr: 0.000013 - momentum: 0.000000 2023-10-17 11:53:57,642 epoch 7 - iter 1563/5212 - loss 0.03183996 - time (sec): 128.34 - samples/sec: 877.20 - lr: 0.000012 - momentum: 0.000000 2023-10-17 11:54:40,075 epoch 7 - iter 2084/5212 - loss 0.03089690 - time (sec): 170.78 - samples/sec: 876.39 - lr: 0.000012 - momentum: 0.000000 2023-10-17 11:55:22,339 epoch 7 - iter 2605/5212 - loss 0.03370678 - time (sec): 213.04 - samples/sec: 867.96 - lr: 0.000012 - momentum: 0.000000 2023-10-17 11:56:03,952 epoch 7 - iter 3126/5212 - loss 0.03259067 - time (sec): 254.65 - samples/sec: 864.70 - lr: 0.000011 - momentum: 0.000000 2023-10-17 11:56:47,386 epoch 7 - iter 3647/5212 - loss 0.03324339 - time (sec): 298.09 - samples/sec: 858.90 - lr: 0.000011 - momentum: 0.000000 2023-10-17 11:57:31,079 epoch 7 - iter 4168/5212 - loss 0.03204679 - time (sec): 341.78 - samples/sec: 865.26 - lr: 0.000011 - momentum: 0.000000 2023-10-17 11:58:13,042 epoch 7 - iter 4689/5212 - loss 0.03270349 - time (sec): 383.74 - samples/sec: 866.76 - lr: 0.000010 - momentum: 0.000000 2023-10-17 11:58:56,293 epoch 7 - iter 5210/5212 - loss 0.03210456 - time (sec): 426.99 - samples/sec: 860.40 - lr: 0.000010 - momentum: 0.000000 2023-10-17 11:58:56,458 ---------------------------------------------------------------------------------------------------- 2023-10-17 11:58:56,458 EPOCH 7 done: loss 0.0321 - lr: 0.000010 2023-10-17 11:59:07,929 DEV : loss 0.4514279067516327 - f1-score (micro avg) 0.3873 2023-10-17 11:59:08,000 ---------------------------------------------------------------------------------------------------- 2023-10-17 11:59:50,456 epoch 8 - iter 521/5212 - loss 0.02287542 - time (sec): 42.45 - samples/sec: 845.18 - lr: 0.000010 - momentum: 0.000000 2023-10-17 12:00:35,464 epoch 8 - iter 1042/5212 - loss 0.02032053 - time (sec): 87.46 - samples/sec: 818.29 - lr: 0.000009 - momentum: 0.000000 2023-10-17 12:01:18,589 epoch 8 - iter 1563/5212 - loss 0.02247712 - time (sec): 130.59 - samples/sec: 814.40 - lr: 0.000009 - momentum: 0.000000 2023-10-17 12:02:01,597 epoch 8 - iter 2084/5212 - loss 0.02124931 - time (sec): 173.59 - samples/sec: 821.16 - lr: 0.000009 - momentum: 0.000000 2023-10-17 12:02:44,007 epoch 8 - iter 2605/5212 - loss 0.02199294 - time (sec): 216.00 - samples/sec: 826.62 - lr: 0.000008 - momentum: 0.000000 2023-10-17 12:03:28,062 epoch 8 - iter 3126/5212 - loss 0.02206727 - time (sec): 260.06 - samples/sec: 829.98 - lr: 0.000008 - momentum: 0.000000 2023-10-17 12:04:11,244 epoch 8 - iter 3647/5212 - loss 0.02198914 - time (sec): 303.24 - samples/sec: 835.71 - lr: 0.000008 - momentum: 0.000000 2023-10-17 12:04:53,240 epoch 8 - iter 4168/5212 - loss 0.02133287 - time (sec): 345.24 - samples/sec: 846.62 - lr: 0.000007 - momentum: 0.000000 2023-10-17 12:05:35,938 epoch 8 - iter 4689/5212 - loss 0.02089455 - time (sec): 387.93 - samples/sec: 851.73 - lr: 0.000007 - momentum: 0.000000 2023-10-17 12:06:18,629 epoch 8 - iter 5210/5212 - loss 0.02139060 - time (sec): 430.63 - samples/sec: 853.09 - lr: 0.000007 - momentum: 0.000000 2023-10-17 12:06:18,786 ---------------------------------------------------------------------------------------------------- 2023-10-17 12:06:18,787 EPOCH 8 done: loss 0.0214 - lr: 0.000007 2023-10-17 12:06:31,446 DEV : loss 0.42482689023017883 - f1-score (micro avg) 0.4045 2023-10-17 12:06:31,520 saving best model 2023-10-17 12:06:33,030 ---------------------------------------------------------------------------------------------------- 2023-10-17 12:07:14,963 epoch 9 - iter 521/5212 - loss 0.01446293 - time (sec): 41.93 - samples/sec: 811.46 - lr: 0.000006 - momentum: 0.000000 2023-10-17 12:07:57,416 epoch 9 - iter 1042/5212 - loss 0.01733620 - time (sec): 84.38 - samples/sec: 845.43 - lr: 0.000006 - momentum: 0.000000 2023-10-17 12:08:38,808 epoch 9 - iter 1563/5212 - loss 0.01496632 - time (sec): 125.78 - samples/sec: 836.07 - lr: 0.000006 - momentum: 0.000000 2023-10-17 12:09:20,895 epoch 9 - iter 2084/5212 - loss 0.01584419 - time (sec): 167.86 - samples/sec: 834.70 - lr: 0.000005 - momentum: 0.000000 2023-10-17 12:10:02,618 epoch 9 - iter 2605/5212 - loss 0.01574388 - time (sec): 209.59 - samples/sec: 843.01 - lr: 0.000005 - momentum: 0.000000 2023-10-17 12:10:44,627 epoch 9 - iter 3126/5212 - loss 0.01792002 - time (sec): 251.59 - samples/sec: 847.62 - lr: 0.000005 - momentum: 0.000000 2023-10-17 12:11:27,348 epoch 9 - iter 3647/5212 - loss 0.01713778 - time (sec): 294.32 - samples/sec: 851.35 - lr: 0.000004 - momentum: 0.000000 2023-10-17 12:12:10,707 epoch 9 - iter 4168/5212 - loss 0.01688059 - time (sec): 337.68 - samples/sec: 855.37 - lr: 0.000004 - momentum: 0.000000 2023-10-17 12:12:53,136 epoch 9 - iter 4689/5212 - loss 0.01654568 - time (sec): 380.10 - samples/sec: 860.95 - lr: 0.000004 - momentum: 0.000000 2023-10-17 12:13:36,164 epoch 9 - iter 5210/5212 - loss 0.01654318 - time (sec): 423.13 - samples/sec: 868.21 - lr: 0.000003 - momentum: 0.000000 2023-10-17 12:13:36,312 ---------------------------------------------------------------------------------------------------- 2023-10-17 12:13:36,313 EPOCH 9 done: loss 0.0165 - lr: 0.000003 2023-10-17 12:13:48,791 DEV : loss 0.41042855381965637 - f1-score (micro avg) 0.416 2023-10-17 12:13:48,867 saving best model 2023-10-17 12:13:50,323 ---------------------------------------------------------------------------------------------------- 2023-10-17 12:14:33,294 epoch 10 - iter 521/5212 - loss 0.00781776 - time (sec): 42.96 - samples/sec: 865.88 - lr: 0.000003 - momentum: 0.000000 2023-10-17 12:15:15,785 epoch 10 - iter 1042/5212 - loss 0.00793401 - time (sec): 85.46 - samples/sec: 852.16 - lr: 0.000003 - momentum: 0.000000 2023-10-17 12:15:58,787 epoch 10 - iter 1563/5212 - loss 0.00880507 - time (sec): 128.46 - samples/sec: 831.97 - lr: 0.000002 - momentum: 0.000000 2023-10-17 12:16:41,351 epoch 10 - iter 2084/5212 - loss 0.00972971 - time (sec): 171.02 - samples/sec: 843.50 - lr: 0.000002 - momentum: 0.000000 2023-10-17 12:17:24,604 epoch 10 - iter 2605/5212 - loss 0.01024931 - time (sec): 214.27 - samples/sec: 848.53 - lr: 0.000002 - momentum: 0.000000 2023-10-17 12:18:05,735 epoch 10 - iter 3126/5212 - loss 0.00986422 - time (sec): 255.41 - samples/sec: 849.94 - lr: 0.000001 - momentum: 0.000000 2023-10-17 12:18:49,745 epoch 10 - iter 3647/5212 - loss 0.00950244 - time (sec): 299.42 - samples/sec: 845.97 - lr: 0.000001 - momentum: 0.000000 2023-10-17 12:19:32,787 epoch 10 - iter 4168/5212 - loss 0.00935274 - time (sec): 342.46 - samples/sec: 847.08 - lr: 0.000001 - momentum: 0.000000 2023-10-17 12:20:15,120 epoch 10 - iter 4689/5212 - loss 0.00939995 - time (sec): 384.79 - samples/sec: 855.40 - lr: 0.000000 - momentum: 0.000000 2023-10-17 12:20:57,845 epoch 10 - iter 5210/5212 - loss 0.00928220 - time (sec): 427.52 - samples/sec: 859.13 - lr: 0.000000 - momentum: 0.000000 2023-10-17 12:20:58,002 ---------------------------------------------------------------------------------------------------- 2023-10-17 12:20:58,002 EPOCH 10 done: loss 0.0093 - lr: 0.000000 2023-10-17 12:21:11,081 DEV : loss 0.4939973056316376 - f1-score (micro avg) 0.3985 2023-10-17 12:21:11,724 ---------------------------------------------------------------------------------------------------- 2023-10-17 12:21:11,727 Loading model from best epoch ... 2023-10-17 12:21:14,296 SequenceTagger predicts: Dictionary with 17 tags: O, S-LOC, B-LOC, E-LOC, I-LOC, S-PER, B-PER, E-PER, I-PER, S-ORG, B-ORG, E-ORG, I-ORG, S-HumanProd, B-HumanProd, E-HumanProd, I-HumanProd 2023-10-17 12:21:35,379 Results: - F-score (micro) 0.4827 - F-score (macro) 0.3288 - Accuracy 0.3215 By class: precision recall f1-score support LOC 0.5341 0.5939 0.5624 1214 PER 0.4248 0.4542 0.4390 808 ORG 0.3102 0.3173 0.3137 353 HumanProd 0.0000 0.0000 0.0000 15 micro avg 0.4648 0.5021 0.4827 2390 macro avg 0.3173 0.3413 0.3288 2390 weighted avg 0.4607 0.5021 0.4804 2390 2023-10-17 12:21:35,379 ----------------------------------------------------------------------------------------------------