2023-10-12 22:05:12,690 ---------------------------------------------------------------------------------------------------- 2023-10-12 22:05:12,692 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=13, bias=True) (loss_function): CrossEntropyLoss() )" 2023-10-12 22:05:12,692 ---------------------------------------------------------------------------------------------------- 2023-10-12 22:05:12,692 MultiCorpus: 7936 train + 992 dev + 992 test sentences - NER_ICDAR_EUROPEANA Corpus: 7936 train + 992 dev + 992 test sentences - /root/.flair/datasets/ner_icdar_europeana/fr 2023-10-12 22:05:12,692 ---------------------------------------------------------------------------------------------------- 2023-10-12 22:05:12,693 Train: 7936 sentences 2023-10-12 22:05:12,693 (train_with_dev=False, train_with_test=False) 2023-10-12 22:05:12,693 ---------------------------------------------------------------------------------------------------- 2023-10-12 22:05:12,693 Training Params: 2023-10-12 22:05:12,693 - learning_rate: "0.00016" 2023-10-12 22:05:12,693 - mini_batch_size: "8" 2023-10-12 22:05:12,693 - max_epochs: "10" 2023-10-12 22:05:12,693 - shuffle: "True" 2023-10-12 22:05:12,693 ---------------------------------------------------------------------------------------------------- 2023-10-12 22:05:12,693 Plugins: 2023-10-12 22:05:12,693 - TensorboardLogger 2023-10-12 22:05:12,693 - LinearScheduler | warmup_fraction: '0.1' 2023-10-12 22:05:12,693 ---------------------------------------------------------------------------------------------------- 2023-10-12 22:05:12,693 Final evaluation on model from best epoch (best-model.pt) 2023-10-12 22:05:12,694 - metric: "('micro avg', 'f1-score')" 2023-10-12 22:05:12,694 ---------------------------------------------------------------------------------------------------- 2023-10-12 22:05:12,694 Computation: 2023-10-12 22:05:12,694 - compute on device: cuda:0 2023-10-12 22:05:12,694 - embedding storage: none 2023-10-12 22:05:12,694 ---------------------------------------------------------------------------------------------------- 2023-10-12 22:05:12,694 Model training base path: "hmbench-icdar/fr-hmbyt5-preliminary/byt5-small-historic-multilingual-span20-flax-bs8-wsFalse-e10-lr0.00016-poolingfirst-layers-1-crfFalse-3" 2023-10-12 22:05:12,694 ---------------------------------------------------------------------------------------------------- 2023-10-12 22:05:12,694 ---------------------------------------------------------------------------------------------------- 2023-10-12 22:05:12,694 Logging anything other than scalars to TensorBoard is currently not supported. 2023-10-12 22:06:02,586 epoch 1 - iter 99/992 - loss 2.53866750 - time (sec): 49.89 - samples/sec: 321.99 - lr: 0.000016 - momentum: 0.000000 2023-10-12 22:06:51,649 epoch 1 - iter 198/992 - loss 2.44203287 - time (sec): 98.95 - samples/sec: 321.09 - lr: 0.000032 - momentum: 0.000000 2023-10-12 22:07:42,422 epoch 1 - iter 297/992 - loss 2.21386372 - time (sec): 149.73 - samples/sec: 325.56 - lr: 0.000048 - momentum: 0.000000 2023-10-12 22:08:33,193 epoch 1 - iter 396/992 - loss 1.97014077 - time (sec): 200.50 - samples/sec: 325.47 - lr: 0.000064 - momentum: 0.000000 2023-10-12 22:09:25,877 epoch 1 - iter 495/992 - loss 1.71310585 - time (sec): 253.18 - samples/sec: 324.30 - lr: 0.000080 - momentum: 0.000000 2023-10-12 22:10:17,664 epoch 1 - iter 594/992 - loss 1.49579279 - time (sec): 304.97 - samples/sec: 322.49 - lr: 0.000096 - momentum: 0.000000 2023-10-12 22:11:05,041 epoch 1 - iter 693/992 - loss 1.33675130 - time (sec): 352.35 - samples/sec: 323.61 - lr: 0.000112 - momentum: 0.000000 2023-10-12 22:11:53,190 epoch 1 - iter 792/992 - loss 1.20207715 - time (sec): 400.49 - samples/sec: 325.89 - lr: 0.000128 - momentum: 0.000000 2023-10-12 22:12:41,655 epoch 1 - iter 891/992 - loss 1.09115549 - time (sec): 448.96 - samples/sec: 328.46 - lr: 0.000144 - momentum: 0.000000 2023-10-12 22:13:30,799 epoch 1 - iter 990/992 - loss 1.00336945 - time (sec): 498.10 - samples/sec: 328.65 - lr: 0.000160 - momentum: 0.000000 2023-10-12 22:13:31,769 ---------------------------------------------------------------------------------------------------- 2023-10-12 22:13:31,769 EPOCH 1 done: loss 1.0021 - lr: 0.000160 2023-10-12 22:13:55,906 DEV : loss 0.16479156911373138 - f1-score (micro avg) 0.4881 2023-10-12 22:13:55,944 saving best model 2023-10-12 22:13:56,845 ---------------------------------------------------------------------------------------------------- 2023-10-12 22:14:45,524 epoch 2 - iter 99/992 - loss 0.16749492 - time (sec): 48.68 - samples/sec: 345.97 - lr: 0.000158 - momentum: 0.000000 2023-10-12 22:15:33,867 epoch 2 - iter 198/992 - loss 0.16513327 - time (sec): 97.02 - samples/sec: 340.72 - lr: 0.000156 - momentum: 0.000000 2023-10-12 22:16:21,989 epoch 2 - iter 297/992 - loss 0.16266570 - time (sec): 145.14 - samples/sec: 340.03 - lr: 0.000155 - momentum: 0.000000 2023-10-12 22:17:09,262 epoch 2 - iter 396/992 - loss 0.15468695 - time (sec): 192.42 - samples/sec: 343.81 - lr: 0.000153 - momentum: 0.000000 2023-10-12 22:17:57,533 epoch 2 - iter 495/992 - loss 0.15190678 - time (sec): 240.69 - samples/sec: 341.26 - lr: 0.000151 - momentum: 0.000000 2023-10-12 22:18:44,949 epoch 2 - iter 594/992 - loss 0.14770062 - time (sec): 288.10 - samples/sec: 343.88 - lr: 0.000149 - momentum: 0.000000 2023-10-12 22:19:32,810 epoch 2 - iter 693/992 - loss 0.14365046 - time (sec): 335.96 - samples/sec: 344.85 - lr: 0.000148 - momentum: 0.000000 2023-10-12 22:20:20,690 epoch 2 - iter 792/992 - loss 0.14111146 - time (sec): 383.84 - samples/sec: 342.67 - lr: 0.000146 - momentum: 0.000000 2023-10-12 22:21:07,617 epoch 2 - iter 891/992 - loss 0.13769670 - time (sec): 430.77 - samples/sec: 342.53 - lr: 0.000144 - momentum: 0.000000 2023-10-12 22:21:54,279 epoch 2 - iter 990/992 - loss 0.13551472 - time (sec): 477.43 - samples/sec: 342.97 - lr: 0.000142 - momentum: 0.000000 2023-10-12 22:21:55,197 ---------------------------------------------------------------------------------------------------- 2023-10-12 22:21:55,197 EPOCH 2 done: loss 0.1354 - lr: 0.000142 2023-10-12 22:22:20,768 DEV : loss 0.08708374202251434 - f1-score (micro avg) 0.7258 2023-10-12 22:22:20,817 saving best model 2023-10-12 22:22:23,472 ---------------------------------------------------------------------------------------------------- 2023-10-12 22:23:12,687 epoch 3 - iter 99/992 - loss 0.07926069 - time (sec): 49.21 - samples/sec: 348.70 - lr: 0.000140 - momentum: 0.000000 2023-10-12 22:24:00,548 epoch 3 - iter 198/992 - loss 0.08319851 - time (sec): 97.07 - samples/sec: 341.76 - lr: 0.000139 - momentum: 0.000000 2023-10-12 22:24:48,441 epoch 3 - iter 297/992 - loss 0.07989106 - time (sec): 144.96 - samples/sec: 337.38 - lr: 0.000137 - momentum: 0.000000 2023-10-12 22:25:35,848 epoch 3 - iter 396/992 - loss 0.07688124 - time (sec): 192.37 - samples/sec: 336.27 - lr: 0.000135 - momentum: 0.000000 2023-10-12 22:26:24,723 epoch 3 - iter 495/992 - loss 0.07665381 - time (sec): 241.24 - samples/sec: 337.07 - lr: 0.000133 - momentum: 0.000000 2023-10-12 22:27:13,187 epoch 3 - iter 594/992 - loss 0.07779634 - time (sec): 289.71 - samples/sec: 336.66 - lr: 0.000132 - momentum: 0.000000 2023-10-12 22:28:07,961 epoch 3 - iter 693/992 - loss 0.07743573 - time (sec): 344.48 - samples/sec: 329.27 - lr: 0.000130 - momentum: 0.000000 2023-10-12 22:28:59,877 epoch 3 - iter 792/992 - loss 0.07691112 - time (sec): 396.40 - samples/sec: 327.63 - lr: 0.000128 - momentum: 0.000000 2023-10-12 22:29:48,612 epoch 3 - iter 891/992 - loss 0.07602786 - time (sec): 445.13 - samples/sec: 329.00 - lr: 0.000126 - momentum: 0.000000 2023-10-12 22:30:37,961 epoch 3 - iter 990/992 - loss 0.07645169 - time (sec): 494.48 - samples/sec: 331.20 - lr: 0.000125 - momentum: 0.000000 2023-10-12 22:30:38,911 ---------------------------------------------------------------------------------------------------- 2023-10-12 22:30:38,912 EPOCH 3 done: loss 0.0765 - lr: 0.000125 2023-10-12 22:31:03,320 DEV : loss 0.08860880136489868 - f1-score (micro avg) 0.7359 2023-10-12 22:31:03,362 saving best model 2023-10-12 22:31:05,994 ---------------------------------------------------------------------------------------------------- 2023-10-12 22:31:55,030 epoch 4 - iter 99/992 - loss 0.05266051 - time (sec): 49.03 - samples/sec: 355.22 - lr: 0.000123 - momentum: 0.000000 2023-10-12 22:32:41,837 epoch 4 - iter 198/992 - loss 0.05108763 - time (sec): 95.84 - samples/sec: 346.53 - lr: 0.000121 - momentum: 0.000000 2023-10-12 22:33:29,635 epoch 4 - iter 297/992 - loss 0.05139942 - time (sec): 143.63 - samples/sec: 344.00 - lr: 0.000119 - momentum: 0.000000 2023-10-12 22:34:17,373 epoch 4 - iter 396/992 - loss 0.04998338 - time (sec): 191.37 - samples/sec: 343.45 - lr: 0.000117 - momentum: 0.000000 2023-10-12 22:35:05,184 epoch 4 - iter 495/992 - loss 0.05009155 - time (sec): 239.18 - samples/sec: 343.37 - lr: 0.000116 - momentum: 0.000000 2023-10-12 22:35:52,496 epoch 4 - iter 594/992 - loss 0.05093715 - time (sec): 286.49 - samples/sec: 342.57 - lr: 0.000114 - momentum: 0.000000 2023-10-12 22:36:39,570 epoch 4 - iter 693/992 - loss 0.05159004 - time (sec): 333.57 - samples/sec: 340.90 - lr: 0.000112 - momentum: 0.000000 2023-10-12 22:37:32,868 epoch 4 - iter 792/992 - loss 0.05202951 - time (sec): 386.87 - samples/sec: 337.64 - lr: 0.000110 - momentum: 0.000000 2023-10-12 22:38:26,673 epoch 4 - iter 891/992 - loss 0.05194402 - time (sec): 440.67 - samples/sec: 333.04 - lr: 0.000109 - momentum: 0.000000 2023-10-12 22:39:21,320 epoch 4 - iter 990/992 - loss 0.05169748 - time (sec): 495.32 - samples/sec: 330.46 - lr: 0.000107 - momentum: 0.000000 2023-10-12 22:39:22,402 ---------------------------------------------------------------------------------------------------- 2023-10-12 22:39:22,402 EPOCH 4 done: loss 0.0519 - lr: 0.000107 2023-10-12 22:39:49,852 DEV : loss 0.10738497972488403 - f1-score (micro avg) 0.7452 2023-10-12 22:39:49,898 saving best model 2023-10-12 22:39:52,641 ---------------------------------------------------------------------------------------------------- 2023-10-12 22:40:45,393 epoch 5 - iter 99/992 - loss 0.03744860 - time (sec): 52.75 - samples/sec: 294.32 - lr: 0.000105 - momentum: 0.000000 2023-10-12 22:41:36,500 epoch 5 - iter 198/992 - loss 0.03310035 - time (sec): 103.85 - samples/sec: 307.26 - lr: 0.000103 - momentum: 0.000000 2023-10-12 22:42:23,692 epoch 5 - iter 297/992 - loss 0.03625772 - time (sec): 151.05 - samples/sec: 320.74 - lr: 0.000101 - momentum: 0.000000 2023-10-12 22:43:11,285 epoch 5 - iter 396/992 - loss 0.03810834 - time (sec): 198.64 - samples/sec: 320.58 - lr: 0.000100 - momentum: 0.000000 2023-10-12 22:43:59,504 epoch 5 - iter 495/992 - loss 0.03758287 - time (sec): 246.86 - samples/sec: 321.20 - lr: 0.000098 - momentum: 0.000000 2023-10-12 22:44:48,523 epoch 5 - iter 594/992 - loss 0.03713120 - time (sec): 295.88 - samples/sec: 324.15 - lr: 0.000096 - momentum: 0.000000 2023-10-12 22:45:37,246 epoch 5 - iter 693/992 - loss 0.03650804 - time (sec): 344.60 - samples/sec: 330.08 - lr: 0.000094 - momentum: 0.000000 2023-10-12 22:46:26,597 epoch 5 - iter 792/992 - loss 0.03660515 - time (sec): 393.95 - samples/sec: 330.86 - lr: 0.000093 - momentum: 0.000000 2023-10-12 22:47:16,389 epoch 5 - iter 891/992 - loss 0.03830079 - time (sec): 443.74 - samples/sec: 330.21 - lr: 0.000091 - momentum: 0.000000 2023-10-12 22:48:05,714 epoch 5 - iter 990/992 - loss 0.03848700 - time (sec): 493.07 - samples/sec: 331.93 - lr: 0.000089 - momentum: 0.000000 2023-10-12 22:48:06,780 ---------------------------------------------------------------------------------------------------- 2023-10-12 22:48:06,781 EPOCH 5 done: loss 0.0385 - lr: 0.000089 2023-10-12 22:48:32,775 DEV : loss 0.12247787415981293 - f1-score (micro avg) 0.765 2023-10-12 22:48:32,824 saving best model 2023-10-12 22:48:35,498 ---------------------------------------------------------------------------------------------------- 2023-10-12 22:49:23,615 epoch 6 - iter 99/992 - loss 0.03956974 - time (sec): 48.10 - samples/sec: 337.82 - lr: 0.000087 - momentum: 0.000000 2023-10-12 22:50:13,289 epoch 6 - iter 198/992 - loss 0.03197861 - time (sec): 97.78 - samples/sec: 329.46 - lr: 0.000085 - momentum: 0.000000 2023-10-12 22:51:03,269 epoch 6 - iter 297/992 - loss 0.03209333 - time (sec): 147.76 - samples/sec: 330.30 - lr: 0.000084 - momentum: 0.000000 2023-10-12 22:51:57,716 epoch 6 - iter 396/992 - loss 0.03241736 - time (sec): 202.21 - samples/sec: 323.77 - lr: 0.000082 - momentum: 0.000000 2023-10-12 22:52:50,732 epoch 6 - iter 495/992 - loss 0.03193044 - time (sec): 255.22 - samples/sec: 321.77 - lr: 0.000080 - momentum: 0.000000 2023-10-12 22:53:41,437 epoch 6 - iter 594/992 - loss 0.03185070 - time (sec): 305.93 - samples/sec: 320.29 - lr: 0.000078 - momentum: 0.000000 2023-10-12 22:54:31,597 epoch 6 - iter 693/992 - loss 0.03141795 - time (sec): 356.09 - samples/sec: 323.41 - lr: 0.000077 - momentum: 0.000000 2023-10-12 22:55:20,841 epoch 6 - iter 792/992 - loss 0.02995177 - time (sec): 405.33 - samples/sec: 324.97 - lr: 0.000075 - momentum: 0.000000 2023-10-12 22:56:13,061 epoch 6 - iter 891/992 - loss 0.03069202 - time (sec): 457.55 - samples/sec: 324.19 - lr: 0.000073 - momentum: 0.000000 2023-10-12 22:57:04,066 epoch 6 - iter 990/992 - loss 0.03019184 - time (sec): 508.56 - samples/sec: 322.03 - lr: 0.000071 - momentum: 0.000000 2023-10-12 22:57:05,223 ---------------------------------------------------------------------------------------------------- 2023-10-12 22:57:05,223 EPOCH 6 done: loss 0.0302 - lr: 0.000071 2023-10-12 22:57:31,123 DEV : loss 0.15935450792312622 - f1-score (micro avg) 0.7639 2023-10-12 22:57:31,171 ---------------------------------------------------------------------------------------------------- 2023-10-12 22:58:24,920 epoch 7 - iter 99/992 - loss 0.01651467 - time (sec): 53.75 - samples/sec: 300.62 - lr: 0.000069 - momentum: 0.000000 2023-10-12 22:59:19,604 epoch 7 - iter 198/992 - loss 0.01743649 - time (sec): 108.43 - samples/sec: 303.71 - lr: 0.000068 - momentum: 0.000000 2023-10-12 23:00:15,752 epoch 7 - iter 297/992 - loss 0.01956424 - time (sec): 164.58 - samples/sec: 297.51 - lr: 0.000066 - momentum: 0.000000 2023-10-12 23:01:10,441 epoch 7 - iter 396/992 - loss 0.02209278 - time (sec): 219.27 - samples/sec: 298.43 - lr: 0.000064 - momentum: 0.000000 2023-10-12 23:02:00,796 epoch 7 - iter 495/992 - loss 0.02209311 - time (sec): 269.62 - samples/sec: 302.36 - lr: 0.000062 - momentum: 0.000000 2023-10-12 23:02:53,389 epoch 7 - iter 594/992 - loss 0.02242793 - time (sec): 322.22 - samples/sec: 305.28 - lr: 0.000061 - momentum: 0.000000 2023-10-12 23:03:41,135 epoch 7 - iter 693/992 - loss 0.02205529 - time (sec): 369.96 - samples/sec: 310.06 - lr: 0.000059 - momentum: 0.000000 2023-10-12 23:04:29,920 epoch 7 - iter 792/992 - loss 0.02172005 - time (sec): 418.75 - samples/sec: 313.47 - lr: 0.000057 - momentum: 0.000000 2023-10-12 23:05:18,911 epoch 7 - iter 891/992 - loss 0.02166572 - time (sec): 467.74 - samples/sec: 314.91 - lr: 0.000055 - momentum: 0.000000 2023-10-12 23:06:07,894 epoch 7 - iter 990/992 - loss 0.02245169 - time (sec): 516.72 - samples/sec: 316.80 - lr: 0.000053 - momentum: 0.000000 2023-10-12 23:06:08,942 ---------------------------------------------------------------------------------------------------- 2023-10-12 23:06:08,943 EPOCH 7 done: loss 0.0224 - lr: 0.000053 2023-10-12 23:06:35,694 DEV : loss 0.1707056164741516 - f1-score (micro avg) 0.7589 2023-10-12 23:06:35,736 ---------------------------------------------------------------------------------------------------- 2023-10-12 23:07:27,949 epoch 8 - iter 99/992 - loss 0.01962247 - time (sec): 52.21 - samples/sec: 316.83 - lr: 0.000052 - momentum: 0.000000 2023-10-12 23:08:19,108 epoch 8 - iter 198/992 - loss 0.01859288 - time (sec): 103.37 - samples/sec: 308.87 - lr: 0.000050 - momentum: 0.000000 2023-10-12 23:09:10,193 epoch 8 - iter 297/992 - loss 0.01679401 - time (sec): 154.46 - samples/sec: 315.72 - lr: 0.000048 - momentum: 0.000000 2023-10-12 23:09:59,592 epoch 8 - iter 396/992 - loss 0.01831256 - time (sec): 203.85 - samples/sec: 321.87 - lr: 0.000046 - momentum: 0.000000 2023-10-12 23:10:50,216 epoch 8 - iter 495/992 - loss 0.01858703 - time (sec): 254.48 - samples/sec: 322.29 - lr: 0.000045 - momentum: 0.000000 2023-10-12 23:11:39,881 epoch 8 - iter 594/992 - loss 0.01869986 - time (sec): 304.14 - samples/sec: 322.59 - lr: 0.000043 - momentum: 0.000000 2023-10-12 23:12:32,934 epoch 8 - iter 693/992 - loss 0.01788096 - time (sec): 357.20 - samples/sec: 319.54 - lr: 0.000041 - momentum: 0.000000 2023-10-12 23:13:25,150 epoch 8 - iter 792/992 - loss 0.01720215 - time (sec): 409.41 - samples/sec: 320.00 - lr: 0.000039 - momentum: 0.000000 2023-10-12 23:14:15,414 epoch 8 - iter 891/992 - loss 0.01721473 - time (sec): 459.68 - samples/sec: 319.64 - lr: 0.000037 - momentum: 0.000000 2023-10-12 23:15:07,027 epoch 8 - iter 990/992 - loss 0.01805657 - time (sec): 511.29 - samples/sec: 320.02 - lr: 0.000036 - momentum: 0.000000 2023-10-12 23:15:07,988 ---------------------------------------------------------------------------------------------------- 2023-10-12 23:15:07,988 EPOCH 8 done: loss 0.0181 - lr: 0.000036 2023-10-12 23:15:33,788 DEV : loss 0.1880449652671814 - f1-score (micro avg) 0.7603 2023-10-12 23:15:33,833 ---------------------------------------------------------------------------------------------------- 2023-10-12 23:16:24,732 epoch 9 - iter 99/992 - loss 0.01209024 - time (sec): 50.90 - samples/sec: 303.76 - lr: 0.000034 - momentum: 0.000000 2023-10-12 23:17:12,430 epoch 9 - iter 198/992 - loss 0.01145554 - time (sec): 98.59 - samples/sec: 313.15 - lr: 0.000032 - momentum: 0.000000 2023-10-12 23:18:00,671 epoch 9 - iter 297/992 - loss 0.01364701 - time (sec): 146.83 - samples/sec: 321.91 - lr: 0.000030 - momentum: 0.000000 2023-10-12 23:18:51,788 epoch 9 - iter 396/992 - loss 0.01407126 - time (sec): 197.95 - samples/sec: 324.82 - lr: 0.000029 - momentum: 0.000000 2023-10-12 23:19:45,370 epoch 9 - iter 495/992 - loss 0.01342117 - time (sec): 251.53 - samples/sec: 321.54 - lr: 0.000027 - momentum: 0.000000 2023-10-12 23:20:34,936 epoch 9 - iter 594/992 - loss 0.01418408 - time (sec): 301.10 - samples/sec: 327.44 - lr: 0.000025 - momentum: 0.000000 2023-10-12 23:21:24,448 epoch 9 - iter 693/992 - loss 0.01478215 - time (sec): 350.61 - samples/sec: 329.28 - lr: 0.000023 - momentum: 0.000000 2023-10-12 23:22:12,712 epoch 9 - iter 792/992 - loss 0.01519139 - time (sec): 398.88 - samples/sec: 331.22 - lr: 0.000022 - momentum: 0.000000 2023-10-12 23:23:00,954 epoch 9 - iter 891/992 - loss 0.01476678 - time (sec): 447.12 - samples/sec: 332.32 - lr: 0.000020 - momentum: 0.000000 2023-10-12 23:23:47,907 epoch 9 - iter 990/992 - loss 0.01486460 - time (sec): 494.07 - samples/sec: 331.20 - lr: 0.000018 - momentum: 0.000000 2023-10-12 23:23:48,866 ---------------------------------------------------------------------------------------------------- 2023-10-12 23:23:48,866 EPOCH 9 done: loss 0.0148 - lr: 0.000018 2023-10-12 23:24:14,572 DEV : loss 0.18750455975532532 - f1-score (micro avg) 0.7678 2023-10-12 23:24:14,613 saving best model 2023-10-12 23:24:17,232 ---------------------------------------------------------------------------------------------------- 2023-10-12 23:25:06,168 epoch 10 - iter 99/992 - loss 0.00876544 - time (sec): 48.93 - samples/sec: 337.18 - lr: 0.000016 - momentum: 0.000000 2023-10-12 23:25:52,988 epoch 10 - iter 198/992 - loss 0.00939953 - time (sec): 95.75 - samples/sec: 347.04 - lr: 0.000014 - momentum: 0.000000 2023-10-12 23:26:40,053 epoch 10 - iter 297/992 - loss 0.01057757 - time (sec): 142.82 - samples/sec: 350.83 - lr: 0.000013 - momentum: 0.000000 2023-10-12 23:27:28,303 epoch 10 - iter 396/992 - loss 0.01038951 - time (sec): 191.07 - samples/sec: 345.97 - lr: 0.000011 - momentum: 0.000000 2023-10-12 23:28:16,166 epoch 10 - iter 495/992 - loss 0.01021717 - time (sec): 238.93 - samples/sec: 344.94 - lr: 0.000009 - momentum: 0.000000 2023-10-12 23:29:08,104 epoch 10 - iter 594/992 - loss 0.01086773 - time (sec): 290.87 - samples/sec: 337.98 - lr: 0.000007 - momentum: 0.000000 2023-10-12 23:29:57,166 epoch 10 - iter 693/992 - loss 0.01145449 - time (sec): 339.93 - samples/sec: 338.36 - lr: 0.000006 - momentum: 0.000000 2023-10-12 23:30:45,548 epoch 10 - iter 792/992 - loss 0.01111800 - time (sec): 388.31 - samples/sec: 336.93 - lr: 0.000004 - momentum: 0.000000 2023-10-12 23:31:36,159 epoch 10 - iter 891/992 - loss 0.01070743 - time (sec): 438.92 - samples/sec: 336.74 - lr: 0.000002 - momentum: 0.000000 2023-10-12 23:32:29,104 epoch 10 - iter 990/992 - loss 0.01101518 - time (sec): 491.87 - samples/sec: 332.77 - lr: 0.000000 - momentum: 0.000000 2023-10-12 23:32:30,054 ---------------------------------------------------------------------------------------------------- 2023-10-12 23:32:30,054 EPOCH 10 done: loss 0.0111 - lr: 0.000000 2023-10-12 23:32:55,364 DEV : loss 0.19627229869365692 - f1-score (micro avg) 0.7659 2023-10-12 23:32:56,332 ---------------------------------------------------------------------------------------------------- 2023-10-12 23:32:56,334 Loading model from best epoch ... 2023-10-12 23:32:59,958 SequenceTagger predicts: Dictionary with 13 tags: O, S-PER, B-PER, E-PER, I-PER, S-LOC, B-LOC, E-LOC, I-LOC, S-ORG, B-ORG, E-ORG, I-ORG 2023-10-12 23:33:23,960 Results: - F-score (micro) 0.7784 - F-score (macro) 0.6876 - Accuracy 0.6589 By class: precision recall f1-score support LOC 0.8260 0.8550 0.8402 655 PER 0.7236 0.7982 0.7591 223 ORG 0.5094 0.4252 0.4635 127 micro avg 0.7689 0.7881 0.7784 1005 macro avg 0.6863 0.6928 0.6876 1005 weighted avg 0.7632 0.7881 0.7746 1005 2023-10-12 23:33:23,960 ----------------------------------------------------------------------------------------------------