File size: 24,119 Bytes
b33cb9a |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 |
2023-10-17 15:29:10,928 ----------------------------------------------------------------------------------------------------
2023-10-17 15:29:10,930 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=13, bias=True)
(loss_function): CrossEntropyLoss()
)"
2023-10-17 15:29:10,930 ----------------------------------------------------------------------------------------------------
2023-10-17 15:29:10,930 MultiCorpus: 14465 train + 1392 dev + 2432 test sentences
- NER_HIPE_2022 Corpus: 14465 train + 1392 dev + 2432 test sentences - /root/.flair/datasets/ner_hipe_2022/v2.1/letemps/fr/with_doc_seperator
2023-10-17 15:29:10,930 ----------------------------------------------------------------------------------------------------
2023-10-17 15:29:10,930 Train: 14465 sentences
2023-10-17 15:29:10,930 (train_with_dev=False, train_with_test=False)
2023-10-17 15:29:10,930 ----------------------------------------------------------------------------------------------------
2023-10-17 15:29:10,930 Training Params:
2023-10-17 15:29:10,930 - learning_rate: "3e-05"
2023-10-17 15:29:10,930 - mini_batch_size: "4"
2023-10-17 15:29:10,930 - max_epochs: "10"
2023-10-17 15:29:10,930 - shuffle: "True"
2023-10-17 15:29:10,930 ----------------------------------------------------------------------------------------------------
2023-10-17 15:29:10,930 Plugins:
2023-10-17 15:29:10,930 - TensorboardLogger
2023-10-17 15:29:10,930 - LinearScheduler | warmup_fraction: '0.1'
2023-10-17 15:29:10,930 ----------------------------------------------------------------------------------------------------
2023-10-17 15:29:10,930 Final evaluation on model from best epoch (best-model.pt)
2023-10-17 15:29:10,930 - metric: "('micro avg', 'f1-score')"
2023-10-17 15:29:10,930 ----------------------------------------------------------------------------------------------------
2023-10-17 15:29:10,931 Computation:
2023-10-17 15:29:10,931 - compute on device: cuda:0
2023-10-17 15:29:10,931 - embedding storage: none
2023-10-17 15:29:10,931 ----------------------------------------------------------------------------------------------------
2023-10-17 15:29:10,931 Model training base path: "hmbench-letemps/fr-hmteams/teams-base-historic-multilingual-discriminator-bs4-wsFalse-e10-lr3e-05-poolingfirst-layers-1-crfFalse-4"
2023-10-17 15:29:10,931 ----------------------------------------------------------------------------------------------------
2023-10-17 15:29:10,931 ----------------------------------------------------------------------------------------------------
2023-10-17 15:29:10,931 Logging anything other than scalars to TensorBoard is currently not supported.
2023-10-17 15:29:33,582 epoch 1 - iter 361/3617 - loss 1.95125825 - time (sec): 22.65 - samples/sec: 1618.02 - lr: 0.000003 - momentum: 0.000000
2023-10-17 15:29:55,459 epoch 1 - iter 722/3617 - loss 1.06563924 - time (sec): 44.53 - samples/sec: 1701.95 - lr: 0.000006 - momentum: 0.000000
2023-10-17 15:30:17,409 epoch 1 - iter 1083/3617 - loss 0.76511000 - time (sec): 66.48 - samples/sec: 1710.02 - lr: 0.000009 - momentum: 0.000000
2023-10-17 15:30:39,364 epoch 1 - iter 1444/3617 - loss 0.60696697 - time (sec): 88.43 - samples/sec: 1727.41 - lr: 0.000012 - momentum: 0.000000
2023-10-17 15:31:01,163 epoch 1 - iter 1805/3617 - loss 0.51265186 - time (sec): 110.23 - samples/sec: 1720.07 - lr: 0.000015 - momentum: 0.000000
2023-10-17 15:31:23,044 epoch 1 - iter 2166/3617 - loss 0.44808148 - time (sec): 132.11 - samples/sec: 1726.30 - lr: 0.000018 - momentum: 0.000000
2023-10-17 15:31:44,775 epoch 1 - iter 2527/3617 - loss 0.40140253 - time (sec): 153.84 - samples/sec: 1730.76 - lr: 0.000021 - momentum: 0.000000
2023-10-17 15:32:07,037 epoch 1 - iter 2888/3617 - loss 0.36462796 - time (sec): 176.10 - samples/sec: 1736.66 - lr: 0.000024 - momentum: 0.000000
2023-10-17 15:32:28,719 epoch 1 - iter 3249/3617 - loss 0.33777977 - time (sec): 197.79 - samples/sec: 1734.26 - lr: 0.000027 - momentum: 0.000000
2023-10-17 15:32:50,322 epoch 1 - iter 3610/3617 - loss 0.31676041 - time (sec): 219.39 - samples/sec: 1729.14 - lr: 0.000030 - momentum: 0.000000
2023-10-17 15:32:50,726 ----------------------------------------------------------------------------------------------------
2023-10-17 15:32:50,727 EPOCH 1 done: loss 0.3164 - lr: 0.000030
2023-10-17 15:32:56,275 DEV : loss 0.12850520014762878 - f1-score (micro avg) 0.6421
2023-10-17 15:32:56,346 saving best model
2023-10-17 15:32:56,847 ----------------------------------------------------------------------------------------------------
2023-10-17 15:33:18,719 epoch 2 - iter 361/3617 - loss 0.10580571 - time (sec): 21.87 - samples/sec: 1777.08 - lr: 0.000030 - momentum: 0.000000
2023-10-17 15:33:40,709 epoch 2 - iter 722/3617 - loss 0.10075377 - time (sec): 43.86 - samples/sec: 1746.67 - lr: 0.000029 - momentum: 0.000000
2023-10-17 15:34:02,957 epoch 2 - iter 1083/3617 - loss 0.09535189 - time (sec): 66.11 - samples/sec: 1739.56 - lr: 0.000029 - momentum: 0.000000
2023-10-17 15:34:24,854 epoch 2 - iter 1444/3617 - loss 0.09554567 - time (sec): 88.00 - samples/sec: 1723.65 - lr: 0.000029 - momentum: 0.000000
2023-10-17 15:34:46,605 epoch 2 - iter 1805/3617 - loss 0.09771623 - time (sec): 109.76 - samples/sec: 1719.58 - lr: 0.000028 - momentum: 0.000000
2023-10-17 15:35:08,360 epoch 2 - iter 2166/3617 - loss 0.10009878 - time (sec): 131.51 - samples/sec: 1711.85 - lr: 0.000028 - momentum: 0.000000
2023-10-17 15:35:31,060 epoch 2 - iter 2527/3617 - loss 0.10203157 - time (sec): 154.21 - samples/sec: 1704.34 - lr: 0.000028 - momentum: 0.000000
2023-10-17 15:35:55,057 epoch 2 - iter 2888/3617 - loss 0.09981740 - time (sec): 178.21 - samples/sec: 1694.39 - lr: 0.000027 - momentum: 0.000000
2023-10-17 15:36:17,897 epoch 2 - iter 3249/3617 - loss 0.09868014 - time (sec): 201.05 - samples/sec: 1687.84 - lr: 0.000027 - momentum: 0.000000
2023-10-17 15:36:40,787 epoch 2 - iter 3610/3617 - loss 0.09884508 - time (sec): 223.94 - samples/sec: 1692.66 - lr: 0.000027 - momentum: 0.000000
2023-10-17 15:36:41,221 ----------------------------------------------------------------------------------------------------
2023-10-17 15:36:41,222 EPOCH 2 done: loss 0.0989 - lr: 0.000027
2023-10-17 15:36:48,360 DEV : loss 0.1303558647632599 - f1-score (micro avg) 0.6286
2023-10-17 15:36:48,400 ----------------------------------------------------------------------------------------------------
2023-10-17 15:37:12,493 epoch 3 - iter 361/3617 - loss 0.07737891 - time (sec): 24.09 - samples/sec: 1589.18 - lr: 0.000026 - momentum: 0.000000
2023-10-17 15:37:36,454 epoch 3 - iter 722/3617 - loss 0.07441406 - time (sec): 48.05 - samples/sec: 1596.55 - lr: 0.000026 - momentum: 0.000000
2023-10-17 15:38:00,123 epoch 3 - iter 1083/3617 - loss 0.07366381 - time (sec): 71.72 - samples/sec: 1590.65 - lr: 0.000026 - momentum: 0.000000
2023-10-17 15:38:23,568 epoch 3 - iter 1444/3617 - loss 0.07419654 - time (sec): 95.17 - samples/sec: 1597.42 - lr: 0.000025 - momentum: 0.000000
2023-10-17 15:38:45,945 epoch 3 - iter 1805/3617 - loss 0.07452173 - time (sec): 117.54 - samples/sec: 1612.73 - lr: 0.000025 - momentum: 0.000000
2023-10-17 15:39:09,076 epoch 3 - iter 2166/3617 - loss 0.07529523 - time (sec): 140.67 - samples/sec: 1615.95 - lr: 0.000025 - momentum: 0.000000
2023-10-17 15:39:32,158 epoch 3 - iter 2527/3617 - loss 0.07621308 - time (sec): 163.76 - samples/sec: 1623.77 - lr: 0.000024 - momentum: 0.000000
2023-10-17 15:39:55,149 epoch 3 - iter 2888/3617 - loss 0.07678395 - time (sec): 186.75 - samples/sec: 1621.56 - lr: 0.000024 - momentum: 0.000000
2023-10-17 15:40:19,157 epoch 3 - iter 3249/3617 - loss 0.07812160 - time (sec): 210.76 - samples/sec: 1613.85 - lr: 0.000024 - momentum: 0.000000
2023-10-17 15:40:42,962 epoch 3 - iter 3610/3617 - loss 0.07806970 - time (sec): 234.56 - samples/sec: 1617.10 - lr: 0.000023 - momentum: 0.000000
2023-10-17 15:40:43,372 ----------------------------------------------------------------------------------------------------
2023-10-17 15:40:43,373 EPOCH 3 done: loss 0.0781 - lr: 0.000023
2023-10-17 15:40:49,799 DEV : loss 0.18371737003326416 - f1-score (micro avg) 0.6438
2023-10-17 15:40:49,840 saving best model
2023-10-17 15:40:50,420 ----------------------------------------------------------------------------------------------------
2023-10-17 15:41:13,032 epoch 4 - iter 361/3617 - loss 0.04344296 - time (sec): 22.61 - samples/sec: 1659.04 - lr: 0.000023 - momentum: 0.000000
2023-10-17 15:41:36,100 epoch 4 - iter 722/3617 - loss 0.05011872 - time (sec): 45.68 - samples/sec: 1662.49 - lr: 0.000023 - momentum: 0.000000
2023-10-17 15:41:59,022 epoch 4 - iter 1083/3617 - loss 0.05362355 - time (sec): 68.60 - samples/sec: 1674.43 - lr: 0.000022 - momentum: 0.000000
2023-10-17 15:42:22,769 epoch 4 - iter 1444/3617 - loss 0.05564807 - time (sec): 92.35 - samples/sec: 1644.11 - lr: 0.000022 - momentum: 0.000000
2023-10-17 15:42:46,279 epoch 4 - iter 1805/3617 - loss 0.05342898 - time (sec): 115.86 - samples/sec: 1623.47 - lr: 0.000022 - momentum: 0.000000
2023-10-17 15:43:08,965 epoch 4 - iter 2166/3617 - loss 0.05432027 - time (sec): 138.54 - samples/sec: 1643.99 - lr: 0.000021 - momentum: 0.000000
2023-10-17 15:43:30,566 epoch 4 - iter 2527/3617 - loss 0.05329553 - time (sec): 160.14 - samples/sec: 1656.72 - lr: 0.000021 - momentum: 0.000000
2023-10-17 15:43:52,348 epoch 4 - iter 2888/3617 - loss 0.05372951 - time (sec): 181.93 - samples/sec: 1663.55 - lr: 0.000021 - momentum: 0.000000
2023-10-17 15:44:13,841 epoch 4 - iter 3249/3617 - loss 0.05394526 - time (sec): 203.42 - samples/sec: 1677.99 - lr: 0.000020 - momentum: 0.000000
2023-10-17 15:44:35,306 epoch 4 - iter 3610/3617 - loss 0.05421478 - time (sec): 224.88 - samples/sec: 1687.06 - lr: 0.000020 - momentum: 0.000000
2023-10-17 15:44:35,696 ----------------------------------------------------------------------------------------------------
2023-10-17 15:44:35,697 EPOCH 4 done: loss 0.0542 - lr: 0.000020
2023-10-17 15:44:42,870 DEV : loss 0.22881226241588593 - f1-score (micro avg) 0.6538
2023-10-17 15:44:42,910 saving best model
2023-10-17 15:44:43,492 ----------------------------------------------------------------------------------------------------
2023-10-17 15:45:06,049 epoch 5 - iter 361/3617 - loss 0.03180280 - time (sec): 22.56 - samples/sec: 1690.26 - lr: 0.000020 - momentum: 0.000000
2023-10-17 15:45:27,669 epoch 5 - iter 722/3617 - loss 0.03588777 - time (sec): 44.17 - samples/sec: 1717.32 - lr: 0.000019 - momentum: 0.000000
2023-10-17 15:45:48,939 epoch 5 - iter 1083/3617 - loss 0.04146070 - time (sec): 65.45 - samples/sec: 1726.42 - lr: 0.000019 - momentum: 0.000000
2023-10-17 15:46:12,476 epoch 5 - iter 1444/3617 - loss 0.03840423 - time (sec): 88.98 - samples/sec: 1696.95 - lr: 0.000019 - momentum: 0.000000
2023-10-17 15:46:35,825 epoch 5 - iter 1805/3617 - loss 0.03842592 - time (sec): 112.33 - samples/sec: 1673.84 - lr: 0.000018 - momentum: 0.000000
2023-10-17 15:46:57,750 epoch 5 - iter 2166/3617 - loss 0.03790309 - time (sec): 134.26 - samples/sec: 1675.11 - lr: 0.000018 - momentum: 0.000000
2023-10-17 15:47:19,326 epoch 5 - iter 2527/3617 - loss 0.04008082 - time (sec): 155.83 - samples/sec: 1692.22 - lr: 0.000018 - momentum: 0.000000
2023-10-17 15:47:42,076 epoch 5 - iter 2888/3617 - loss 0.03976002 - time (sec): 178.58 - samples/sec: 1690.27 - lr: 0.000017 - momentum: 0.000000
2023-10-17 15:48:04,606 epoch 5 - iter 3249/3617 - loss 0.04136127 - time (sec): 201.11 - samples/sec: 1692.09 - lr: 0.000017 - momentum: 0.000000
2023-10-17 15:48:26,428 epoch 5 - iter 3610/3617 - loss 0.04096447 - time (sec): 222.93 - samples/sec: 1701.22 - lr: 0.000017 - momentum: 0.000000
2023-10-17 15:48:26,886 ----------------------------------------------------------------------------------------------------
2023-10-17 15:48:26,887 EPOCH 5 done: loss 0.0409 - lr: 0.000017
2023-10-17 15:48:33,200 DEV : loss 0.30138152837753296 - f1-score (micro avg) 0.6299
2023-10-17 15:48:33,240 ----------------------------------------------------------------------------------------------------
2023-10-17 15:48:55,394 epoch 6 - iter 361/3617 - loss 0.02934544 - time (sec): 22.15 - samples/sec: 1719.54 - lr: 0.000016 - momentum: 0.000000
2023-10-17 15:49:17,504 epoch 6 - iter 722/3617 - loss 0.02571943 - time (sec): 44.26 - samples/sec: 1685.96 - lr: 0.000016 - momentum: 0.000000
2023-10-17 15:49:39,822 epoch 6 - iter 1083/3617 - loss 0.02841514 - time (sec): 66.58 - samples/sec: 1690.02 - lr: 0.000016 - momentum: 0.000000
2023-10-17 15:50:02,839 epoch 6 - iter 1444/3617 - loss 0.02630164 - time (sec): 89.60 - samples/sec: 1700.67 - lr: 0.000015 - momentum: 0.000000
2023-10-17 15:50:25,636 epoch 6 - iter 1805/3617 - loss 0.02616624 - time (sec): 112.39 - samples/sec: 1689.51 - lr: 0.000015 - momentum: 0.000000
2023-10-17 15:50:47,660 epoch 6 - iter 2166/3617 - loss 0.02849047 - time (sec): 134.42 - samples/sec: 1697.10 - lr: 0.000015 - momentum: 0.000000
2023-10-17 15:51:09,936 epoch 6 - iter 2527/3617 - loss 0.02883754 - time (sec): 156.69 - samples/sec: 1683.02 - lr: 0.000014 - momentum: 0.000000
2023-10-17 15:51:32,228 epoch 6 - iter 2888/3617 - loss 0.02863689 - time (sec): 178.99 - samples/sec: 1686.43 - lr: 0.000014 - momentum: 0.000000
2023-10-17 15:51:54,353 epoch 6 - iter 3249/3617 - loss 0.02895811 - time (sec): 201.11 - samples/sec: 1691.66 - lr: 0.000014 - momentum: 0.000000
2023-10-17 15:52:16,274 epoch 6 - iter 3610/3617 - loss 0.02847322 - time (sec): 223.03 - samples/sec: 1700.56 - lr: 0.000013 - momentum: 0.000000
2023-10-17 15:52:16,684 ----------------------------------------------------------------------------------------------------
2023-10-17 15:52:16,685 EPOCH 6 done: loss 0.0285 - lr: 0.000013
2023-10-17 15:52:23,863 DEV : loss 0.3001513183116913 - f1-score (micro avg) 0.6594
2023-10-17 15:52:23,903 saving best model
2023-10-17 15:52:24,490 ----------------------------------------------------------------------------------------------------
2023-10-17 15:52:46,074 epoch 7 - iter 361/3617 - loss 0.02066184 - time (sec): 21.58 - samples/sec: 1680.61 - lr: 0.000013 - momentum: 0.000000
2023-10-17 15:53:07,793 epoch 7 - iter 722/3617 - loss 0.01708605 - time (sec): 43.30 - samples/sec: 1686.78 - lr: 0.000013 - momentum: 0.000000
2023-10-17 15:53:29,760 epoch 7 - iter 1083/3617 - loss 0.01965216 - time (sec): 65.27 - samples/sec: 1679.03 - lr: 0.000012 - momentum: 0.000000
2023-10-17 15:53:52,055 epoch 7 - iter 1444/3617 - loss 0.01952788 - time (sec): 87.56 - samples/sec: 1697.54 - lr: 0.000012 - momentum: 0.000000
2023-10-17 15:54:14,114 epoch 7 - iter 1805/3617 - loss 0.02017047 - time (sec): 109.62 - samples/sec: 1720.19 - lr: 0.000012 - momentum: 0.000000
2023-10-17 15:54:36,183 epoch 7 - iter 2166/3617 - loss 0.02020291 - time (sec): 131.69 - samples/sec: 1730.58 - lr: 0.000011 - momentum: 0.000000
2023-10-17 15:54:58,335 epoch 7 - iter 2527/3617 - loss 0.01975895 - time (sec): 153.84 - samples/sec: 1723.15 - lr: 0.000011 - momentum: 0.000000
2023-10-17 15:55:20,465 epoch 7 - iter 2888/3617 - loss 0.01957773 - time (sec): 175.97 - samples/sec: 1718.41 - lr: 0.000011 - momentum: 0.000000
2023-10-17 15:55:42,750 epoch 7 - iter 3249/3617 - loss 0.01965841 - time (sec): 198.26 - samples/sec: 1717.48 - lr: 0.000010 - momentum: 0.000000
2023-10-17 15:56:04,884 epoch 7 - iter 3610/3617 - loss 0.01945380 - time (sec): 220.39 - samples/sec: 1720.30 - lr: 0.000010 - momentum: 0.000000
2023-10-17 15:56:05,307 ----------------------------------------------------------------------------------------------------
2023-10-17 15:56:05,307 EPOCH 7 done: loss 0.0194 - lr: 0.000010
2023-10-17 15:56:11,598 DEV : loss 0.3526001274585724 - f1-score (micro avg) 0.653
2023-10-17 15:56:11,642 ----------------------------------------------------------------------------------------------------
2023-10-17 15:56:34,023 epoch 8 - iter 361/3617 - loss 0.01442396 - time (sec): 22.38 - samples/sec: 1663.89 - lr: 0.000010 - momentum: 0.000000
2023-10-17 15:56:56,449 epoch 8 - iter 722/3617 - loss 0.01297978 - time (sec): 44.81 - samples/sec: 1646.69 - lr: 0.000009 - momentum: 0.000000
2023-10-17 15:57:19,213 epoch 8 - iter 1083/3617 - loss 0.01197713 - time (sec): 67.57 - samples/sec: 1644.68 - lr: 0.000009 - momentum: 0.000000
2023-10-17 15:57:42,219 epoch 8 - iter 1444/3617 - loss 0.01296548 - time (sec): 90.58 - samples/sec: 1661.07 - lr: 0.000009 - momentum: 0.000000
2023-10-17 15:58:04,626 epoch 8 - iter 1805/3617 - loss 0.01361106 - time (sec): 112.98 - samples/sec: 1668.32 - lr: 0.000008 - momentum: 0.000000
2023-10-17 15:58:26,765 epoch 8 - iter 2166/3617 - loss 0.01333120 - time (sec): 135.12 - samples/sec: 1671.29 - lr: 0.000008 - momentum: 0.000000
2023-10-17 15:58:48,975 epoch 8 - iter 2527/3617 - loss 0.01337065 - time (sec): 157.33 - samples/sec: 1681.32 - lr: 0.000008 - momentum: 0.000000
2023-10-17 15:59:11,245 epoch 8 - iter 2888/3617 - loss 0.01298466 - time (sec): 179.60 - samples/sec: 1687.72 - lr: 0.000007 - momentum: 0.000000
2023-10-17 15:59:32,888 epoch 8 - iter 3249/3617 - loss 0.01264441 - time (sec): 201.24 - samples/sec: 1687.78 - lr: 0.000007 - momentum: 0.000000
2023-10-17 15:59:54,791 epoch 8 - iter 3610/3617 - loss 0.01285488 - time (sec): 223.15 - samples/sec: 1698.74 - lr: 0.000007 - momentum: 0.000000
2023-10-17 15:59:55,221 ----------------------------------------------------------------------------------------------------
2023-10-17 15:59:55,221 EPOCH 8 done: loss 0.0129 - lr: 0.000007
2023-10-17 16:00:01,661 DEV : loss 0.38147813081741333 - f1-score (micro avg) 0.6595
2023-10-17 16:00:01,703 saving best model
2023-10-17 16:00:02,302 ----------------------------------------------------------------------------------------------------
2023-10-17 16:00:24,155 epoch 9 - iter 361/3617 - loss 0.00411798 - time (sec): 21.85 - samples/sec: 1665.06 - lr: 0.000006 - momentum: 0.000000
2023-10-17 16:00:45,993 epoch 9 - iter 722/3617 - loss 0.00798323 - time (sec): 43.69 - samples/sec: 1695.04 - lr: 0.000006 - momentum: 0.000000
2023-10-17 16:01:08,507 epoch 9 - iter 1083/3617 - loss 0.00861031 - time (sec): 66.20 - samples/sec: 1696.57 - lr: 0.000006 - momentum: 0.000000
2023-10-17 16:01:32,055 epoch 9 - iter 1444/3617 - loss 0.00771193 - time (sec): 89.75 - samples/sec: 1681.98 - lr: 0.000005 - momentum: 0.000000
2023-10-17 16:01:53,809 epoch 9 - iter 1805/3617 - loss 0.00769213 - time (sec): 111.51 - samples/sec: 1693.76 - lr: 0.000005 - momentum: 0.000000
2023-10-17 16:02:16,235 epoch 9 - iter 2166/3617 - loss 0.00792339 - time (sec): 133.93 - samples/sec: 1694.24 - lr: 0.000005 - momentum: 0.000000
2023-10-17 16:02:39,818 epoch 9 - iter 2527/3617 - loss 0.00790490 - time (sec): 157.51 - samples/sec: 1695.75 - lr: 0.000004 - momentum: 0.000000
2023-10-17 16:03:02,857 epoch 9 - iter 2888/3617 - loss 0.00773309 - time (sec): 180.55 - samples/sec: 1691.08 - lr: 0.000004 - momentum: 0.000000
2023-10-17 16:03:27,163 epoch 9 - iter 3249/3617 - loss 0.00781997 - time (sec): 204.86 - samples/sec: 1673.25 - lr: 0.000004 - momentum: 0.000000
2023-10-17 16:03:50,243 epoch 9 - iter 3610/3617 - loss 0.00767986 - time (sec): 227.94 - samples/sec: 1663.71 - lr: 0.000003 - momentum: 0.000000
2023-10-17 16:03:50,696 ----------------------------------------------------------------------------------------------------
2023-10-17 16:03:50,697 EPOCH 9 done: loss 0.0077 - lr: 0.000003
2023-10-17 16:03:56,943 DEV : loss 0.39987462759017944 - f1-score (micro avg) 0.6503
2023-10-17 16:03:56,984 ----------------------------------------------------------------------------------------------------
2023-10-17 16:04:19,198 epoch 10 - iter 361/3617 - loss 0.00449363 - time (sec): 22.21 - samples/sec: 1699.38 - lr: 0.000003 - momentum: 0.000000
2023-10-17 16:04:42,363 epoch 10 - iter 722/3617 - loss 0.00614855 - time (sec): 45.38 - samples/sec: 1718.53 - lr: 0.000003 - momentum: 0.000000
2023-10-17 16:05:06,457 epoch 10 - iter 1083/3617 - loss 0.00541348 - time (sec): 69.47 - samples/sec: 1661.47 - lr: 0.000002 - momentum: 0.000000
2023-10-17 16:05:29,935 epoch 10 - iter 1444/3617 - loss 0.00603835 - time (sec): 92.95 - samples/sec: 1651.51 - lr: 0.000002 - momentum: 0.000000
2023-10-17 16:05:52,494 epoch 10 - iter 1805/3617 - loss 0.00569937 - time (sec): 115.51 - samples/sec: 1642.98 - lr: 0.000002 - momentum: 0.000000
2023-10-17 16:06:15,075 epoch 10 - iter 2166/3617 - loss 0.00554871 - time (sec): 138.09 - samples/sec: 1654.28 - lr: 0.000001 - momentum: 0.000000
2023-10-17 16:06:37,223 epoch 10 - iter 2527/3617 - loss 0.00529464 - time (sec): 160.24 - samples/sec: 1664.53 - lr: 0.000001 - momentum: 0.000000
2023-10-17 16:07:00,170 epoch 10 - iter 2888/3617 - loss 0.00525690 - time (sec): 183.18 - samples/sec: 1661.69 - lr: 0.000001 - momentum: 0.000000
2023-10-17 16:07:22,855 epoch 10 - iter 3249/3617 - loss 0.00525983 - time (sec): 205.87 - samples/sec: 1667.90 - lr: 0.000000 - momentum: 0.000000
2023-10-17 16:07:45,081 epoch 10 - iter 3610/3617 - loss 0.00525246 - time (sec): 228.09 - samples/sec: 1663.75 - lr: 0.000000 - momentum: 0.000000
2023-10-17 16:07:45,507 ----------------------------------------------------------------------------------------------------
2023-10-17 16:07:45,508 EPOCH 10 done: loss 0.0052 - lr: 0.000000
2023-10-17 16:07:52,788 DEV : loss 0.4126187264919281 - f1-score (micro avg) 0.6545
2023-10-17 16:07:53,330 ----------------------------------------------------------------------------------------------------
2023-10-17 16:07:53,332 Loading model from best epoch ...
2023-10-17 16:07:55,106 SequenceTagger predicts: Dictionary with 13 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
2023-10-17 16:08:03,374
Results:
- F-score (micro) 0.6596
- F-score (macro) 0.5329
- Accuracy 0.5036
By class:
precision recall f1-score support
loc 0.6219 0.8156 0.7057 591
pers 0.5813 0.7815 0.6667 357
org 0.2250 0.2278 0.2264 79
micro avg 0.5835 0.7585 0.6596 1027
macro avg 0.4761 0.6083 0.5329 1027
weighted avg 0.5773 0.7585 0.6553 1027
2023-10-17 16:08:03,374 ----------------------------------------------------------------------------------------------------
|