2023-10-11 10:27:33,571 ---------------------------------------------------------------------------------------------------- 2023-10-11 10:27:33,574 Model: "SequenceTagger( (embeddings): ByT5Embeddings( (model): T5EncoderModel( (shared): Embedding(384, 1472) (encoder): T5Stack( (embed_tokens): Embedding(384, 1472) (block): ModuleList( (0): T5Block( (layer): ModuleList( (0): T5LayerSelfAttention( (SelfAttention): T5Attention( (q): Linear(in_features=1472, out_features=384, bias=False) (k): Linear(in_features=1472, out_features=384, bias=False) (v): Linear(in_features=1472, out_features=384, bias=False) (o): Linear(in_features=384, out_features=1472, bias=False) (relative_attention_bias): Embedding(32, 6) ) (layer_norm): FusedRMSNorm(torch.Size([1472]), eps=1e-06, elementwise_affine=True) (dropout): Dropout(p=0.1, inplace=False) ) (1): T5LayerFF( (DenseReluDense): T5DenseGatedActDense( (wi_0): Linear(in_features=1472, out_features=3584, bias=False) (wi_1): Linear(in_features=1472, out_features=3584, bias=False) (wo): Linear(in_features=3584, out_features=1472, bias=False) (dropout): Dropout(p=0.1, inplace=False) (act): NewGELUActivation() ) (layer_norm): FusedRMSNorm(torch.Size([1472]), eps=1e-06, elementwise_affine=True) (dropout): Dropout(p=0.1, inplace=False) ) ) ) (1-11): 11 x T5Block( (layer): ModuleList( (0): T5LayerSelfAttention( (SelfAttention): T5Attention( (q): Linear(in_features=1472, out_features=384, bias=False) (k): Linear(in_features=1472, out_features=384, bias=False) (v): Linear(in_features=1472, out_features=384, bias=False) (o): Linear(in_features=384, out_features=1472, bias=False) ) (layer_norm): FusedRMSNorm(torch.Size([1472]), eps=1e-06, elementwise_affine=True) (dropout): Dropout(p=0.1, inplace=False) ) (1): T5LayerFF( (DenseReluDense): T5DenseGatedActDense( (wi_0): Linear(in_features=1472, out_features=3584, bias=False) (wi_1): Linear(in_features=1472, out_features=3584, bias=False) (wo): Linear(in_features=3584, out_features=1472, bias=False) (dropout): Dropout(p=0.1, inplace=False) (act): NewGELUActivation() ) (layer_norm): FusedRMSNorm(torch.Size([1472]), eps=1e-06, elementwise_affine=True) (dropout): Dropout(p=0.1, inplace=False) ) ) ) ) (final_layer_norm): FusedRMSNorm(torch.Size([1472]), eps=1e-06, elementwise_affine=True) (dropout): Dropout(p=0.1, inplace=False) ) ) ) (locked_dropout): LockedDropout(p=0.5) (linear): Linear(in_features=1472, out_features=17, bias=True) (loss_function): CrossEntropyLoss() )" 2023-10-11 10:27:33,574 ---------------------------------------------------------------------------------------------------- 2023-10-11 10:27:33,574 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 2023-10-11 10:27:33,574 ---------------------------------------------------------------------------------------------------- 2023-10-11 10:27:33,574 Train: 20847 sentences 2023-10-11 10:27:33,575 (train_with_dev=False, train_with_test=False) 2023-10-11 10:27:33,575 ---------------------------------------------------------------------------------------------------- 2023-10-11 10:27:33,575 Training Params: 2023-10-11 10:27:33,575 - learning_rate: "0.00015" 2023-10-11 10:27:33,575 - mini_batch_size: "4" 2023-10-11 10:27:33,575 - max_epochs: "10" 2023-10-11 10:27:33,575 - shuffle: "True" 2023-10-11 10:27:33,575 ---------------------------------------------------------------------------------------------------- 2023-10-11 10:27:33,575 Plugins: 2023-10-11 10:27:33,575 - TensorboardLogger 2023-10-11 10:27:33,575 - LinearScheduler | warmup_fraction: '0.1' 2023-10-11 10:27:33,575 ---------------------------------------------------------------------------------------------------- 2023-10-11 10:27:33,575 Final evaluation on model from best epoch (best-model.pt) 2023-10-11 10:27:33,576 - metric: "('micro avg', 'f1-score')" 2023-10-11 10:27:33,576 ---------------------------------------------------------------------------------------------------- 2023-10-11 10:27:33,576 Computation: 2023-10-11 10:27:33,576 - compute on device: cuda:0 2023-10-11 10:27:33,576 - embedding storage: none 2023-10-11 10:27:33,576 ---------------------------------------------------------------------------------------------------- 2023-10-11 10:27:33,576 Model training base path: "hmbench-newseye/de-hmbyt5-preliminary/byt5-small-historic-multilingual-span20-flax-bs4-wsFalse-e10-lr0.00015-poolingfirst-layers-1-crfFalse-3" 2023-10-11 10:27:33,576 ---------------------------------------------------------------------------------------------------- 2023-10-11 10:27:33,576 ---------------------------------------------------------------------------------------------------- 2023-10-11 10:27:33,576 Logging anything other than scalars to TensorBoard is currently not supported. 2023-10-11 10:29:55,156 epoch 1 - iter 521/5212 - loss 2.77308848 - time (sec): 141.58 - samples/sec: 263.97 - lr: 0.000015 - momentum: 0.000000 2023-10-11 10:32:16,905 epoch 1 - iter 1042/5212 - loss 2.32230492 - time (sec): 283.33 - samples/sec: 270.92 - lr: 0.000030 - momentum: 0.000000 2023-10-11 10:34:35,422 epoch 1 - iter 1563/5212 - loss 1.84040633 - time (sec): 421.84 - samples/sec: 268.49 - lr: 0.000045 - momentum: 0.000000 2023-10-11 10:36:54,353 epoch 1 - iter 2084/5212 - loss 1.50205626 - time (sec): 560.77 - samples/sec: 266.53 - lr: 0.000060 - momentum: 0.000000 2023-10-11 10:39:16,176 epoch 1 - iter 2605/5212 - loss 1.29871005 - time (sec): 702.60 - samples/sec: 266.48 - lr: 0.000075 - momentum: 0.000000 2023-10-11 10:41:36,574 epoch 1 - iter 3126/5212 - loss 1.14819274 - time (sec): 842.99 - samples/sec: 265.07 - lr: 0.000090 - momentum: 0.000000 2023-10-11 10:43:56,922 epoch 1 - iter 3647/5212 - loss 1.03459929 - time (sec): 983.34 - samples/sec: 263.04 - lr: 0.000105 - momentum: 0.000000 2023-10-11 10:46:16,004 epoch 1 - iter 4168/5212 - loss 0.95149666 - time (sec): 1122.43 - samples/sec: 261.30 - lr: 0.000120 - momentum: 0.000000 2023-10-11 10:48:36,869 epoch 1 - iter 4689/5212 - loss 0.87440388 - time (sec): 1263.29 - samples/sec: 262.05 - lr: 0.000135 - momentum: 0.000000 2023-10-11 10:50:57,081 epoch 1 - iter 5210/5212 - loss 0.80881280 - time (sec): 1403.50 - samples/sec: 261.66 - lr: 0.000150 - momentum: 0.000000 2023-10-11 10:50:57,613 ---------------------------------------------------------------------------------------------------- 2023-10-11 10:50:57,613 EPOCH 1 done: loss 0.8086 - lr: 0.000150 2023-10-11 10:51:33,753 DEV : loss 0.1371876299381256 - f1-score (micro avg) 0.3336 2023-10-11 10:51:33,806 saving best model 2023-10-11 10:51:34,714 ---------------------------------------------------------------------------------------------------- 2023-10-11 10:53:54,909 epoch 2 - iter 521/5212 - loss 0.19618675 - time (sec): 140.19 - samples/sec: 263.48 - lr: 0.000148 - momentum: 0.000000 2023-10-11 10:56:12,183 epoch 2 - iter 1042/5212 - loss 0.18458622 - time (sec): 277.47 - samples/sec: 264.49 - lr: 0.000147 - momentum: 0.000000 2023-10-11 10:58:36,985 epoch 2 - iter 1563/5212 - loss 0.18961166 - time (sec): 422.27 - samples/sec: 267.02 - lr: 0.000145 - momentum: 0.000000 2023-10-11 11:00:58,974 epoch 2 - iter 2084/5212 - loss 0.18529432 - time (sec): 564.26 - samples/sec: 264.94 - lr: 0.000143 - momentum: 0.000000 2023-10-11 11:03:19,697 epoch 2 - iter 2605/5212 - loss 0.18045804 - time (sec): 704.98 - samples/sec: 261.97 - lr: 0.000142 - momentum: 0.000000 2023-10-11 11:05:42,377 epoch 2 - iter 3126/5212 - loss 0.17752518 - time (sec): 847.66 - samples/sec: 261.81 - lr: 0.000140 - momentum: 0.000000 2023-10-11 11:08:02,418 epoch 2 - iter 3647/5212 - loss 0.17641629 - time (sec): 987.70 - samples/sec: 258.68 - lr: 0.000138 - momentum: 0.000000 2023-10-11 11:10:25,754 epoch 2 - iter 4168/5212 - loss 0.17189078 - time (sec): 1131.04 - samples/sec: 258.18 - lr: 0.000137 - momentum: 0.000000 2023-10-11 11:12:50,890 epoch 2 - iter 4689/5212 - loss 0.16869126 - time (sec): 1276.17 - samples/sec: 258.97 - lr: 0.000135 - momentum: 0.000000 2023-10-11 11:15:14,550 epoch 2 - iter 5210/5212 - loss 0.16487858 - time (sec): 1419.83 - samples/sec: 258.73 - lr: 0.000133 - momentum: 0.000000 2023-10-11 11:15:14,995 ---------------------------------------------------------------------------------------------------- 2023-10-11 11:15:14,995 EPOCH 2 done: loss 0.1649 - lr: 0.000133 2023-10-11 11:15:54,268 DEV : loss 0.14473062753677368 - f1-score (micro avg) 0.3188 2023-10-11 11:15:54,319 ---------------------------------------------------------------------------------------------------- 2023-10-11 11:18:13,777 epoch 3 - iter 521/5212 - loss 0.10458077 - time (sec): 139.46 - samples/sec: 250.74 - lr: 0.000132 - momentum: 0.000000 2023-10-11 11:20:33,676 epoch 3 - iter 1042/5212 - loss 0.10612993 - time (sec): 279.36 - samples/sec: 254.51 - lr: 0.000130 - momentum: 0.000000 2023-10-11 11:22:53,147 epoch 3 - iter 1563/5212 - loss 0.10354381 - time (sec): 418.83 - samples/sec: 255.19 - lr: 0.000128 - momentum: 0.000000 2023-10-11 11:25:16,345 epoch 3 - iter 2084/5212 - loss 0.10944179 - time (sec): 562.02 - samples/sec: 259.35 - lr: 0.000127 - momentum: 0.000000 2023-10-11 11:27:35,775 epoch 3 - iter 2605/5212 - loss 0.11310926 - time (sec): 701.45 - samples/sec: 262.27 - lr: 0.000125 - momentum: 0.000000 2023-10-11 11:29:54,384 epoch 3 - iter 3126/5212 - loss 0.10867516 - time (sec): 840.06 - samples/sec: 261.83 - lr: 0.000123 - momentum: 0.000000 2023-10-11 11:32:12,976 epoch 3 - iter 3647/5212 - loss 0.10736809 - time (sec): 978.66 - samples/sec: 261.05 - lr: 0.000122 - momentum: 0.000000 2023-10-11 11:34:32,157 epoch 3 - iter 4168/5212 - loss 0.10821714 - time (sec): 1117.84 - samples/sec: 261.76 - lr: 0.000120 - momentum: 0.000000 2023-10-11 11:36:50,238 epoch 3 - iter 4689/5212 - loss 0.10784416 - time (sec): 1255.92 - samples/sec: 261.68 - lr: 0.000118 - momentum: 0.000000 2023-10-11 11:39:11,437 epoch 3 - iter 5210/5212 - loss 0.10805328 - time (sec): 1397.12 - samples/sec: 262.92 - lr: 0.000117 - momentum: 0.000000 2023-10-11 11:39:11,884 ---------------------------------------------------------------------------------------------------- 2023-10-11 11:39:11,885 EPOCH 3 done: loss 0.1082 - lr: 0.000117 2023-10-11 11:39:50,899 DEV : loss 0.18280969560146332 - f1-score (micro avg) 0.4057 2023-10-11 11:39:50,958 saving best model 2023-10-11 11:39:53,591 ---------------------------------------------------------------------------------------------------- 2023-10-11 11:42:14,816 epoch 4 - iter 521/5212 - loss 0.07278263 - time (sec): 141.22 - samples/sec: 250.15 - lr: 0.000115 - momentum: 0.000000 2023-10-11 11:44:36,537 epoch 4 - iter 1042/5212 - loss 0.07239825 - time (sec): 282.94 - samples/sec: 254.20 - lr: 0.000113 - momentum: 0.000000 2023-10-11 11:46:57,338 epoch 4 - iter 1563/5212 - loss 0.07231137 - time (sec): 423.74 - samples/sec: 258.25 - lr: 0.000112 - momentum: 0.000000 2023-10-11 11:49:15,233 epoch 4 - iter 2084/5212 - loss 0.07343325 - time (sec): 561.64 - samples/sec: 257.71 - lr: 0.000110 - momentum: 0.000000 2023-10-11 11:51:38,449 epoch 4 - iter 2605/5212 - loss 0.07181772 - time (sec): 704.85 - samples/sec: 262.36 - lr: 0.000108 - momentum: 0.000000 2023-10-11 11:53:56,748 epoch 4 - iter 3126/5212 - loss 0.07331153 - time (sec): 843.15 - samples/sec: 261.09 - lr: 0.000107 - momentum: 0.000000 2023-10-11 11:56:18,345 epoch 4 - iter 3647/5212 - loss 0.07316098 - time (sec): 984.75 - samples/sec: 261.72 - lr: 0.000105 - momentum: 0.000000 2023-10-11 11:58:40,109 epoch 4 - iter 4168/5212 - loss 0.07306211 - time (sec): 1126.51 - samples/sec: 264.45 - lr: 0.000103 - momentum: 0.000000 2023-10-11 12:00:54,299 epoch 4 - iter 4689/5212 - loss 0.07216413 - time (sec): 1260.70 - samples/sec: 263.41 - lr: 0.000102 - momentum: 0.000000 2023-10-11 12:03:10,241 epoch 4 - iter 5210/5212 - loss 0.07359050 - time (sec): 1396.65 - samples/sec: 263.05 - lr: 0.000100 - momentum: 0.000000 2023-10-11 12:03:10,639 ---------------------------------------------------------------------------------------------------- 2023-10-11 12:03:10,640 EPOCH 4 done: loss 0.0736 - lr: 0.000100 2023-10-11 12:03:49,861 DEV : loss 0.27816441655158997 - f1-score (micro avg) 0.3524 2023-10-11 12:03:49,914 ---------------------------------------------------------------------------------------------------- 2023-10-11 12:06:07,583 epoch 5 - iter 521/5212 - loss 0.04138265 - time (sec): 137.67 - samples/sec: 260.75 - lr: 0.000098 - momentum: 0.000000 2023-10-11 12:08:31,272 epoch 5 - iter 1042/5212 - loss 0.04749933 - time (sec): 281.36 - samples/sec: 261.65 - lr: 0.000097 - momentum: 0.000000 2023-10-11 12:10:54,820 epoch 5 - iter 1563/5212 - loss 0.04798952 - time (sec): 424.90 - samples/sec: 257.07 - lr: 0.000095 - momentum: 0.000000 2023-10-11 12:13:20,187 epoch 5 - iter 2084/5212 - loss 0.05106117 - time (sec): 570.27 - samples/sec: 255.58 - lr: 0.000093 - momentum: 0.000000 2023-10-11 12:15:47,123 epoch 5 - iter 2605/5212 - loss 0.05002478 - time (sec): 717.21 - samples/sec: 256.72 - lr: 0.000092 - momentum: 0.000000 2023-10-11 12:18:10,619 epoch 5 - iter 3126/5212 - loss 0.05055487 - time (sec): 860.70 - samples/sec: 254.97 - lr: 0.000090 - momentum: 0.000000 2023-10-11 12:20:35,848 epoch 5 - iter 3647/5212 - loss 0.05188152 - time (sec): 1005.93 - samples/sec: 254.69 - lr: 0.000088 - momentum: 0.000000 2023-10-11 12:22:59,869 epoch 5 - iter 4168/5212 - loss 0.05088261 - time (sec): 1149.95 - samples/sec: 253.41 - lr: 0.000087 - momentum: 0.000000 2023-10-11 12:25:25,685 epoch 5 - iter 4689/5212 - loss 0.05025009 - time (sec): 1295.77 - samples/sec: 253.90 - lr: 0.000085 - momentum: 0.000000 2023-10-11 12:27:51,044 epoch 5 - iter 5210/5212 - loss 0.05118220 - time (sec): 1441.13 - samples/sec: 254.89 - lr: 0.000083 - momentum: 0.000000 2023-10-11 12:27:51,510 ---------------------------------------------------------------------------------------------------- 2023-10-11 12:27:51,511 EPOCH 5 done: loss 0.0512 - lr: 0.000083 2023-10-11 12:28:32,525 DEV : loss 0.3333892226219177 - f1-score (micro avg) 0.3863 2023-10-11 12:28:32,579 ---------------------------------------------------------------------------------------------------- 2023-10-11 12:30:52,575 epoch 6 - iter 521/5212 - loss 0.03246002 - time (sec): 139.99 - samples/sec: 241.08 - lr: 0.000082 - momentum: 0.000000 2023-10-11 12:33:13,663 epoch 6 - iter 1042/5212 - loss 0.03096266 - time (sec): 281.08 - samples/sec: 243.35 - lr: 0.000080 - momentum: 0.000000 2023-10-11 12:35:38,110 epoch 6 - iter 1563/5212 - loss 0.03241169 - time (sec): 425.53 - samples/sec: 245.97 - lr: 0.000078 - momentum: 0.000000 2023-10-11 12:38:01,041 epoch 6 - iter 2084/5212 - loss 0.03229808 - time (sec): 568.46 - samples/sec: 247.46 - lr: 0.000077 - momentum: 0.000000 2023-10-11 12:40:25,250 epoch 6 - iter 2605/5212 - loss 0.03266383 - time (sec): 712.67 - samples/sec: 249.02 - lr: 0.000075 - momentum: 0.000000 2023-10-11 12:42:46,535 epoch 6 - iter 3126/5212 - loss 0.03217009 - time (sec): 853.95 - samples/sec: 249.45 - lr: 0.000073 - momentum: 0.000000 2023-10-11 12:45:11,195 epoch 6 - iter 3647/5212 - loss 0.03269803 - time (sec): 998.61 - samples/sec: 252.89 - lr: 0.000072 - momentum: 0.000000 2023-10-11 12:47:35,975 epoch 6 - iter 4168/5212 - loss 0.03242644 - time (sec): 1143.39 - samples/sec: 253.88 - lr: 0.000070 - momentum: 0.000000 2023-10-11 12:50:00,875 epoch 6 - iter 4689/5212 - loss 0.03391785 - time (sec): 1288.29 - samples/sec: 255.87 - lr: 0.000068 - momentum: 0.000000 2023-10-11 12:52:21,466 epoch 6 - iter 5210/5212 - loss 0.03471456 - time (sec): 1428.89 - samples/sec: 256.95 - lr: 0.000067 - momentum: 0.000000 2023-10-11 12:52:22,095 ---------------------------------------------------------------------------------------------------- 2023-10-11 12:52:22,095 EPOCH 6 done: loss 0.0347 - lr: 0.000067 2023-10-11 12:53:00,341 DEV : loss 0.397626131772995 - f1-score (micro avg) 0.3773 2023-10-11 12:53:00,393 ---------------------------------------------------------------------------------------------------- 2023-10-11 12:55:23,221 epoch 7 - iter 521/5212 - loss 0.02766582 - time (sec): 142.83 - samples/sec: 280.80 - lr: 0.000065 - momentum: 0.000000 2023-10-11 12:57:43,563 epoch 7 - iter 1042/5212 - loss 0.02549282 - time (sec): 283.17 - samples/sec: 268.30 - lr: 0.000063 - momentum: 0.000000 2023-10-11 13:00:06,987 epoch 7 - iter 1563/5212 - loss 0.02298030 - time (sec): 426.59 - samples/sec: 265.97 - lr: 0.000062 - momentum: 0.000000 2023-10-11 13:02:33,264 epoch 7 - iter 2084/5212 - loss 0.02542458 - time (sec): 572.87 - samples/sec: 265.28 - lr: 0.000060 - momentum: 0.000000 2023-10-11 13:04:53,874 epoch 7 - iter 2605/5212 - loss 0.02658181 - time (sec): 713.48 - samples/sec: 260.46 - lr: 0.000058 - momentum: 0.000000 2023-10-11 13:07:19,303 epoch 7 - iter 3126/5212 - loss 0.02678770 - time (sec): 858.91 - samples/sec: 260.17 - lr: 0.000057 - momentum: 0.000000 2023-10-11 13:09:41,217 epoch 7 - iter 3647/5212 - loss 0.02682370 - time (sec): 1000.82 - samples/sec: 259.27 - lr: 0.000055 - momentum: 0.000000 2023-10-11 13:12:01,858 epoch 7 - iter 4168/5212 - loss 0.02663508 - time (sec): 1141.46 - samples/sec: 258.26 - lr: 0.000053 - momentum: 0.000000 2023-10-11 13:14:23,833 epoch 7 - iter 4689/5212 - loss 0.02646033 - time (sec): 1283.44 - samples/sec: 258.05 - lr: 0.000052 - momentum: 0.000000 2023-10-11 13:16:45,213 epoch 7 - iter 5210/5212 - loss 0.02638106 - time (sec): 1424.82 - samples/sec: 257.84 - lr: 0.000050 - momentum: 0.000000 2023-10-11 13:16:45,633 ---------------------------------------------------------------------------------------------------- 2023-10-11 13:16:45,633 EPOCH 7 done: loss 0.0264 - lr: 0.000050 2023-10-11 13:17:23,898 DEV : loss 0.39803504943847656 - f1-score (micro avg) 0.3855 2023-10-11 13:17:23,949 ---------------------------------------------------------------------------------------------------- 2023-10-11 13:19:45,732 epoch 8 - iter 521/5212 - loss 0.01602735 - time (sec): 141.78 - samples/sec: 259.43 - lr: 0.000048 - momentum: 0.000000 2023-10-11 13:22:08,337 epoch 8 - iter 1042/5212 - loss 0.01754525 - time (sec): 284.39 - samples/sec: 260.76 - lr: 0.000047 - momentum: 0.000000 2023-10-11 13:24:29,046 epoch 8 - iter 1563/5212 - loss 0.01837014 - time (sec): 425.09 - samples/sec: 259.41 - lr: 0.000045 - momentum: 0.000000 2023-10-11 13:26:54,017 epoch 8 - iter 2084/5212 - loss 0.01699882 - time (sec): 570.07 - samples/sec: 258.05 - lr: 0.000043 - momentum: 0.000000 2023-10-11 13:29:16,314 epoch 8 - iter 2605/5212 - loss 0.01790932 - time (sec): 712.36 - samples/sec: 259.27 - lr: 0.000042 - momentum: 0.000000 2023-10-11 13:31:37,281 epoch 8 - iter 3126/5212 - loss 0.01863312 - time (sec): 853.33 - samples/sec: 259.21 - lr: 0.000040 - momentum: 0.000000 2023-10-11 13:33:56,013 epoch 8 - iter 3647/5212 - loss 0.01818979 - time (sec): 992.06 - samples/sec: 259.09 - lr: 0.000038 - momentum: 0.000000 2023-10-11 13:36:17,040 epoch 8 - iter 4168/5212 - loss 0.01807363 - time (sec): 1133.09 - samples/sec: 259.49 - lr: 0.000037 - momentum: 0.000000 2023-10-11 13:38:38,320 epoch 8 - iter 4689/5212 - loss 0.01736008 - time (sec): 1274.37 - samples/sec: 260.41 - lr: 0.000035 - momentum: 0.000000 2023-10-11 13:40:56,604 epoch 8 - iter 5210/5212 - loss 0.01727055 - time (sec): 1412.65 - samples/sec: 259.87 - lr: 0.000033 - momentum: 0.000000 2023-10-11 13:40:57,310 ---------------------------------------------------------------------------------------------------- 2023-10-11 13:40:57,311 EPOCH 8 done: loss 0.0173 - lr: 0.000033 2023-10-11 13:41:35,861 DEV : loss 0.434541255235672 - f1-score (micro avg) 0.413 2023-10-11 13:41:35,914 saving best model 2023-10-11 13:41:38,523 ---------------------------------------------------------------------------------------------------- 2023-10-11 13:44:00,885 epoch 9 - iter 521/5212 - loss 0.01391910 - time (sec): 142.36 - samples/sec: 269.72 - lr: 0.000032 - momentum: 0.000000 2023-10-11 13:46:21,602 epoch 9 - iter 1042/5212 - loss 0.01227543 - time (sec): 283.07 - samples/sec: 268.26 - lr: 0.000030 - momentum: 0.000000 2023-10-11 13:48:44,428 epoch 9 - iter 1563/5212 - loss 0.01127380 - time (sec): 425.90 - samples/sec: 261.83 - lr: 0.000028 - momentum: 0.000000 2023-10-11 13:51:05,408 epoch 9 - iter 2084/5212 - loss 0.01170911 - time (sec): 566.88 - samples/sec: 257.72 - lr: 0.000027 - momentum: 0.000000 2023-10-11 13:53:27,927 epoch 9 - iter 2605/5212 - loss 0.01188654 - time (sec): 709.40 - samples/sec: 258.85 - lr: 0.000025 - momentum: 0.000000 2023-10-11 13:55:49,808 epoch 9 - iter 3126/5212 - loss 0.01195441 - time (sec): 851.28 - samples/sec: 257.46 - lr: 0.000023 - momentum: 0.000000 2023-10-11 13:58:19,120 epoch 9 - iter 3647/5212 - loss 0.01196011 - time (sec): 1000.59 - samples/sec: 255.97 - lr: 0.000022 - momentum: 0.000000 2023-10-11 14:00:47,719 epoch 9 - iter 4168/5212 - loss 0.01172818 - time (sec): 1149.19 - samples/sec: 255.46 - lr: 0.000020 - momentum: 0.000000 2023-10-11 14:03:20,778 epoch 9 - iter 4689/5212 - loss 0.01252167 - time (sec): 1302.25 - samples/sec: 254.08 - lr: 0.000018 - momentum: 0.000000 2023-10-11 14:05:43,342 epoch 9 - iter 5210/5212 - loss 0.01279053 - time (sec): 1444.81 - samples/sec: 254.13 - lr: 0.000017 - momentum: 0.000000 2023-10-11 14:05:43,953 ---------------------------------------------------------------------------------------------------- 2023-10-11 14:05:43,953 EPOCH 9 done: loss 0.0128 - lr: 0.000017 2023-10-11 14:06:24,418 DEV : loss 0.48711156845092773 - f1-score (micro avg) 0.3852 2023-10-11 14:06:24,474 ---------------------------------------------------------------------------------------------------- 2023-10-11 14:08:53,718 epoch 10 - iter 521/5212 - loss 0.00536186 - time (sec): 149.24 - samples/sec: 244.44 - lr: 0.000015 - momentum: 0.000000 2023-10-11 14:11:12,907 epoch 10 - iter 1042/5212 - loss 0.00727275 - time (sec): 288.43 - samples/sec: 249.23 - lr: 0.000013 - momentum: 0.000000 2023-10-11 14:13:34,227 epoch 10 - iter 1563/5212 - loss 0.00704063 - time (sec): 429.75 - samples/sec: 252.94 - lr: 0.000012 - momentum: 0.000000 2023-10-11 14:15:56,744 epoch 10 - iter 2084/5212 - loss 0.00715484 - time (sec): 572.27 - samples/sec: 251.17 - lr: 0.000010 - momentum: 0.000000 2023-10-11 14:18:24,269 epoch 10 - iter 2605/5212 - loss 0.00703148 - time (sec): 719.79 - samples/sec: 254.39 - lr: 0.000008 - momentum: 0.000000 2023-10-11 14:20:45,693 epoch 10 - iter 3126/5212 - loss 0.00712965 - time (sec): 861.22 - samples/sec: 254.36 - lr: 0.000007 - momentum: 0.000000 2023-10-11 14:23:09,963 epoch 10 - iter 3647/5212 - loss 0.00771973 - time (sec): 1005.49 - samples/sec: 254.06 - lr: 0.000005 - momentum: 0.000000 2023-10-11 14:25:37,329 epoch 10 - iter 4168/5212 - loss 0.00782038 - time (sec): 1152.85 - samples/sec: 251.90 - lr: 0.000003 - momentum: 0.000000 2023-10-11 14:28:04,663 epoch 10 - iter 4689/5212 - loss 0.00789966 - time (sec): 1300.19 - samples/sec: 253.75 - lr: 0.000002 - momentum: 0.000000 2023-10-11 14:30:30,685 epoch 10 - iter 5210/5212 - loss 0.00775810 - time (sec): 1446.21 - samples/sec: 254.04 - lr: 0.000000 - momentum: 0.000000 2023-10-11 14:30:31,090 ---------------------------------------------------------------------------------------------------- 2023-10-11 14:30:31,090 EPOCH 10 done: loss 0.0078 - lr: 0.000000 2023-10-11 14:31:09,345 DEV : loss 0.4862366318702698 - f1-score (micro avg) 0.3966 2023-10-11 14:31:10,297 ---------------------------------------------------------------------------------------------------- 2023-10-11 14:31:10,299 Loading model from best epoch ... 2023-10-11 14:31:14,039 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 2023-10-11 14:32:55,343 Results: - F-score (micro) 0.4259 - F-score (macro) 0.2963 - Accuracy 0.2744 By class: precision recall f1-score support LOC 0.4982 0.4596 0.4781 1214 PER 0.3957 0.4369 0.4153 808 ORG 0.2918 0.2918 0.2918 353 HumanProd 0.0000 0.0000 0.0000 15 micro avg 0.4275 0.4243 0.4259 2390 macro avg 0.2964 0.2971 0.2963 2390 weighted avg 0.4300 0.4243 0.4264 2390 2023-10-11 14:32:55,344 ----------------------------------------------------------------------------------------------------