File size: 24,076 Bytes
cd935fd |
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 |
2023-10-17 11:52:54,786 ----------------------------------------------------------------------------------------------------
2023-10-17 11:52:54,787 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 11:52:54,787 ----------------------------------------------------------------------------------------------------
2023-10-17 11:52:54,787 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 11:52:54,787 ----------------------------------------------------------------------------------------------------
2023-10-17 11:52:54,787 Train: 14465 sentences
2023-10-17 11:52:54,788 (train_with_dev=False, train_with_test=False)
2023-10-17 11:52:54,788 ----------------------------------------------------------------------------------------------------
2023-10-17 11:52:54,788 Training Params:
2023-10-17 11:52:54,788 - learning_rate: "5e-05"
2023-10-17 11:52:54,788 - mini_batch_size: "4"
2023-10-17 11:52:54,788 - max_epochs: "10"
2023-10-17 11:52:54,788 - shuffle: "True"
2023-10-17 11:52:54,788 ----------------------------------------------------------------------------------------------------
2023-10-17 11:52:54,788 Plugins:
2023-10-17 11:52:54,788 - TensorboardLogger
2023-10-17 11:52:54,788 - LinearScheduler | warmup_fraction: '0.1'
2023-10-17 11:52:54,788 ----------------------------------------------------------------------------------------------------
2023-10-17 11:52:54,788 Final evaluation on model from best epoch (best-model.pt)
2023-10-17 11:52:54,788 - metric: "('micro avg', 'f1-score')"
2023-10-17 11:52:54,788 ----------------------------------------------------------------------------------------------------
2023-10-17 11:52:54,788 Computation:
2023-10-17 11:52:54,789 - compute on device: cuda:0
2023-10-17 11:52:54,789 - embedding storage: none
2023-10-17 11:52:54,789 ----------------------------------------------------------------------------------------------------
2023-10-17 11:52:54,789 Model training base path: "hmbench-letemps/fr-hmteams/teams-base-historic-multilingual-discriminator-bs4-wsFalse-e10-lr5e-05-poolingfirst-layers-1-crfFalse-2"
2023-10-17 11:52:54,789 ----------------------------------------------------------------------------------------------------
2023-10-17 11:52:54,789 ----------------------------------------------------------------------------------------------------
2023-10-17 11:52:54,789 Logging anything other than scalars to TensorBoard is currently not supported.
2023-10-17 11:53:17,101 epoch 1 - iter 361/3617 - loss 1.45041553 - time (sec): 22.31 - samples/sec: 1698.14 - lr: 0.000005 - momentum: 0.000000
2023-10-17 11:53:38,886 epoch 1 - iter 722/3617 - loss 0.83716082 - time (sec): 44.10 - samples/sec: 1686.06 - lr: 0.000010 - momentum: 0.000000
2023-10-17 11:54:02,640 epoch 1 - iter 1083/3617 - loss 0.59781487 - time (sec): 67.85 - samples/sec: 1681.62 - lr: 0.000015 - momentum: 0.000000
2023-10-17 11:54:24,936 epoch 1 - iter 1444/3617 - loss 0.48149327 - time (sec): 90.15 - samples/sec: 1694.00 - lr: 0.000020 - momentum: 0.000000
2023-10-17 11:54:47,775 epoch 1 - iter 1805/3617 - loss 0.41319463 - time (sec): 112.98 - samples/sec: 1690.70 - lr: 0.000025 - momentum: 0.000000
2023-10-17 11:55:11,378 epoch 1 - iter 2166/3617 - loss 0.36558875 - time (sec): 136.59 - samples/sec: 1672.75 - lr: 0.000030 - momentum: 0.000000
2023-10-17 11:55:35,040 epoch 1 - iter 2527/3617 - loss 0.32938220 - time (sec): 160.25 - samples/sec: 1663.32 - lr: 0.000035 - momentum: 0.000000
2023-10-17 11:55:59,195 epoch 1 - iter 2888/3617 - loss 0.30249835 - time (sec): 184.40 - samples/sec: 1658.74 - lr: 0.000040 - momentum: 0.000000
2023-10-17 11:56:22,092 epoch 1 - iter 3249/3617 - loss 0.28225761 - time (sec): 207.30 - samples/sec: 1657.19 - lr: 0.000045 - momentum: 0.000000
2023-10-17 11:56:45,459 epoch 1 - iter 3610/3617 - loss 0.26717410 - time (sec): 230.67 - samples/sec: 1644.02 - lr: 0.000050 - momentum: 0.000000
2023-10-17 11:56:45,909 ----------------------------------------------------------------------------------------------------
2023-10-17 11:56:45,909 EPOCH 1 done: loss 0.2669 - lr: 0.000050
2023-10-17 11:56:51,334 DEV : loss 0.11786916851997375 - f1-score (micro avg) 0.5019
2023-10-17 11:56:51,374 saving best model
2023-10-17 11:56:51,877 ----------------------------------------------------------------------------------------------------
2023-10-17 11:57:15,292 epoch 2 - iter 361/3617 - loss 0.10557221 - time (sec): 23.41 - samples/sec: 1662.15 - lr: 0.000049 - momentum: 0.000000
2023-10-17 11:57:37,011 epoch 2 - iter 722/3617 - loss 0.10278944 - time (sec): 45.13 - samples/sec: 1710.38 - lr: 0.000049 - momentum: 0.000000
2023-10-17 11:57:58,661 epoch 2 - iter 1083/3617 - loss 0.10655278 - time (sec): 66.78 - samples/sec: 1709.09 - lr: 0.000048 - momentum: 0.000000
2023-10-17 11:58:21,263 epoch 2 - iter 1444/3617 - loss 0.10643025 - time (sec): 89.38 - samples/sec: 1696.68 - lr: 0.000048 - momentum: 0.000000
2023-10-17 11:58:44,442 epoch 2 - iter 1805/3617 - loss 0.10457147 - time (sec): 112.56 - samples/sec: 1674.98 - lr: 0.000047 - momentum: 0.000000
2023-10-17 11:59:07,714 epoch 2 - iter 2166/3617 - loss 0.10556253 - time (sec): 135.84 - samples/sec: 1671.06 - lr: 0.000047 - momentum: 0.000000
2023-10-17 11:59:30,682 epoch 2 - iter 2527/3617 - loss 0.10440273 - time (sec): 158.80 - samples/sec: 1673.80 - lr: 0.000046 - momentum: 0.000000
2023-10-17 11:59:51,173 epoch 2 - iter 2888/3617 - loss 0.10398260 - time (sec): 179.29 - samples/sec: 1692.95 - lr: 0.000046 - momentum: 0.000000
2023-10-17 12:00:09,116 epoch 2 - iter 3249/3617 - loss 0.10460700 - time (sec): 197.24 - samples/sec: 1736.35 - lr: 0.000045 - momentum: 0.000000
2023-10-17 12:00:30,262 epoch 2 - iter 3610/3617 - loss 0.10618003 - time (sec): 218.38 - samples/sec: 1736.86 - lr: 0.000044 - momentum: 0.000000
2023-10-17 12:00:30,682 ----------------------------------------------------------------------------------------------------
2023-10-17 12:00:30,682 EPOCH 2 done: loss 0.1063 - lr: 0.000044
2023-10-17 12:00:37,743 DEV : loss 0.15698249638080597 - f1-score (micro avg) 0.6292
2023-10-17 12:00:37,792 saving best model
2023-10-17 12:00:38,379 ----------------------------------------------------------------------------------------------------
2023-10-17 12:01:00,850 epoch 3 - iter 361/3617 - loss 0.08022559 - time (sec): 22.47 - samples/sec: 1636.31 - lr: 0.000044 - momentum: 0.000000
2023-10-17 12:01:23,625 epoch 3 - iter 722/3617 - loss 0.08069320 - time (sec): 45.24 - samples/sec: 1663.23 - lr: 0.000043 - momentum: 0.000000
2023-10-17 12:01:46,101 epoch 3 - iter 1083/3617 - loss 0.08181083 - time (sec): 67.72 - samples/sec: 1672.77 - lr: 0.000043 - momentum: 0.000000
2023-10-17 12:02:09,383 epoch 3 - iter 1444/3617 - loss 0.08325852 - time (sec): 91.00 - samples/sec: 1661.25 - lr: 0.000042 - momentum: 0.000000
2023-10-17 12:02:27,763 epoch 3 - iter 1805/3617 - loss 0.08237466 - time (sec): 109.38 - samples/sec: 1726.56 - lr: 0.000042 - momentum: 0.000000
2023-10-17 12:02:49,091 epoch 3 - iter 2166/3617 - loss 0.08389650 - time (sec): 130.71 - samples/sec: 1745.30 - lr: 0.000041 - momentum: 0.000000
2023-10-17 12:03:11,251 epoch 3 - iter 2527/3617 - loss 0.08517465 - time (sec): 152.87 - samples/sec: 1731.56 - lr: 0.000041 - momentum: 0.000000
2023-10-17 12:03:33,655 epoch 3 - iter 2888/3617 - loss 0.08551924 - time (sec): 175.27 - samples/sec: 1728.77 - lr: 0.000040 - momentum: 0.000000
2023-10-17 12:03:56,927 epoch 3 - iter 3249/3617 - loss 0.08551615 - time (sec): 198.55 - samples/sec: 1722.72 - lr: 0.000039 - momentum: 0.000000
2023-10-17 12:04:19,219 epoch 3 - iter 3610/3617 - loss 0.08533098 - time (sec): 220.84 - samples/sec: 1717.15 - lr: 0.000039 - momentum: 0.000000
2023-10-17 12:04:19,653 ----------------------------------------------------------------------------------------------------
2023-10-17 12:04:19,654 EPOCH 3 done: loss 0.0853 - lr: 0.000039
2023-10-17 12:04:26,008 DEV : loss 0.20511414110660553 - f1-score (micro avg) 0.6161
2023-10-17 12:04:26,049 ----------------------------------------------------------------------------------------------------
2023-10-17 12:04:49,116 epoch 4 - iter 361/3617 - loss 0.05873537 - time (sec): 23.07 - samples/sec: 1674.62 - lr: 0.000038 - momentum: 0.000000
2023-10-17 12:05:12,376 epoch 4 - iter 722/3617 - loss 0.06004159 - time (sec): 46.33 - samples/sec: 1645.73 - lr: 0.000038 - momentum: 0.000000
2023-10-17 12:05:35,066 epoch 4 - iter 1083/3617 - loss 0.06324927 - time (sec): 69.02 - samples/sec: 1673.37 - lr: 0.000037 - momentum: 0.000000
2023-10-17 12:05:57,755 epoch 4 - iter 1444/3617 - loss 0.06295022 - time (sec): 91.70 - samples/sec: 1671.02 - lr: 0.000037 - momentum: 0.000000
2023-10-17 12:06:20,466 epoch 4 - iter 1805/3617 - loss 0.06317630 - time (sec): 114.42 - samples/sec: 1668.82 - lr: 0.000036 - momentum: 0.000000
2023-10-17 12:06:42,695 epoch 4 - iter 2166/3617 - loss 0.06379107 - time (sec): 136.64 - samples/sec: 1673.67 - lr: 0.000036 - momentum: 0.000000
2023-10-17 12:07:04,646 epoch 4 - iter 2527/3617 - loss 0.06397835 - time (sec): 158.60 - samples/sec: 1681.56 - lr: 0.000035 - momentum: 0.000000
2023-10-17 12:07:27,971 epoch 4 - iter 2888/3617 - loss 0.06348311 - time (sec): 181.92 - samples/sec: 1678.47 - lr: 0.000034 - momentum: 0.000000
2023-10-17 12:07:49,326 epoch 4 - iter 3249/3617 - loss 0.06347814 - time (sec): 203.28 - samples/sec: 1686.73 - lr: 0.000034 - momentum: 0.000000
2023-10-17 12:08:11,036 epoch 4 - iter 3610/3617 - loss 0.06347373 - time (sec): 224.98 - samples/sec: 1686.43 - lr: 0.000033 - momentum: 0.000000
2023-10-17 12:08:11,436 ----------------------------------------------------------------------------------------------------
2023-10-17 12:08:11,436 EPOCH 4 done: loss 0.0634 - lr: 0.000033
2023-10-17 12:08:18,597 DEV : loss 0.22120679914951324 - f1-score (micro avg) 0.6187
2023-10-17 12:08:18,639 ----------------------------------------------------------------------------------------------------
2023-10-17 12:08:42,405 epoch 5 - iter 361/3617 - loss 0.03776153 - time (sec): 23.77 - samples/sec: 1596.80 - lr: 0.000033 - momentum: 0.000000
2023-10-17 12:09:05,646 epoch 5 - iter 722/3617 - loss 0.04135391 - time (sec): 47.01 - samples/sec: 1629.07 - lr: 0.000032 - momentum: 0.000000
2023-10-17 12:09:28,421 epoch 5 - iter 1083/3617 - loss 0.04270991 - time (sec): 69.78 - samples/sec: 1637.87 - lr: 0.000032 - momentum: 0.000000
2023-10-17 12:09:51,769 epoch 5 - iter 1444/3617 - loss 0.04260540 - time (sec): 93.13 - samples/sec: 1635.70 - lr: 0.000031 - momentum: 0.000000
2023-10-17 12:10:14,299 epoch 5 - iter 1805/3617 - loss 0.04110131 - time (sec): 115.66 - samples/sec: 1656.86 - lr: 0.000031 - momentum: 0.000000
2023-10-17 12:10:36,976 epoch 5 - iter 2166/3617 - loss 0.04177018 - time (sec): 138.34 - samples/sec: 1668.53 - lr: 0.000030 - momentum: 0.000000
2023-10-17 12:10:59,264 epoch 5 - iter 2527/3617 - loss 0.04324074 - time (sec): 160.62 - samples/sec: 1663.73 - lr: 0.000029 - momentum: 0.000000
2023-10-17 12:11:21,101 epoch 5 - iter 2888/3617 - loss 0.04411956 - time (sec): 182.46 - samples/sec: 1658.59 - lr: 0.000029 - momentum: 0.000000
2023-10-17 12:11:45,148 epoch 5 - iter 3249/3617 - loss 0.04348395 - time (sec): 206.51 - samples/sec: 1652.71 - lr: 0.000028 - momentum: 0.000000
2023-10-17 12:12:07,783 epoch 5 - iter 3610/3617 - loss 0.04478114 - time (sec): 229.14 - samples/sec: 1654.40 - lr: 0.000028 - momentum: 0.000000
2023-10-17 12:12:08,198 ----------------------------------------------------------------------------------------------------
2023-10-17 12:12:08,198 EPOCH 5 done: loss 0.0447 - lr: 0.000028
2023-10-17 12:12:14,607 DEV : loss 0.2750839591026306 - f1-score (micro avg) 0.6418
2023-10-17 12:12:14,652 saving best model
2023-10-17 12:12:15,246 ----------------------------------------------------------------------------------------------------
2023-10-17 12:12:37,546 epoch 6 - iter 361/3617 - loss 0.02896062 - time (sec): 22.30 - samples/sec: 1678.63 - lr: 0.000027 - momentum: 0.000000
2023-10-17 12:13:01,122 epoch 6 - iter 722/3617 - loss 0.02877240 - time (sec): 45.87 - samples/sec: 1629.00 - lr: 0.000027 - momentum: 0.000000
2023-10-17 12:13:23,419 epoch 6 - iter 1083/3617 - loss 0.03088311 - time (sec): 68.17 - samples/sec: 1654.59 - lr: 0.000026 - momentum: 0.000000
2023-10-17 12:13:45,707 epoch 6 - iter 1444/3617 - loss 0.03181763 - time (sec): 90.46 - samples/sec: 1672.21 - lr: 0.000026 - momentum: 0.000000
2023-10-17 12:14:07,684 epoch 6 - iter 1805/3617 - loss 0.03280478 - time (sec): 112.44 - samples/sec: 1689.30 - lr: 0.000025 - momentum: 0.000000
2023-10-17 12:14:30,091 epoch 6 - iter 2166/3617 - loss 0.03257395 - time (sec): 134.84 - samples/sec: 1691.26 - lr: 0.000024 - momentum: 0.000000
2023-10-17 12:14:52,451 epoch 6 - iter 2527/3617 - loss 0.03273840 - time (sec): 157.20 - samples/sec: 1693.80 - lr: 0.000024 - momentum: 0.000000
2023-10-17 12:15:16,751 epoch 6 - iter 2888/3617 - loss 0.03337305 - time (sec): 181.50 - samples/sec: 1675.04 - lr: 0.000023 - momentum: 0.000000
2023-10-17 12:15:39,289 epoch 6 - iter 3249/3617 - loss 0.03410165 - time (sec): 204.04 - samples/sec: 1671.99 - lr: 0.000023 - momentum: 0.000000
2023-10-17 12:15:59,015 epoch 6 - iter 3610/3617 - loss 0.03367272 - time (sec): 223.77 - samples/sec: 1695.44 - lr: 0.000022 - momentum: 0.000000
2023-10-17 12:15:59,460 ----------------------------------------------------------------------------------------------------
2023-10-17 12:15:59,460 EPOCH 6 done: loss 0.0337 - lr: 0.000022
2023-10-17 12:16:06,723 DEV : loss 0.3224092423915863 - f1-score (micro avg) 0.6278
2023-10-17 12:16:06,764 ----------------------------------------------------------------------------------------------------
2023-10-17 12:16:30,012 epoch 7 - iter 361/3617 - loss 0.01207785 - time (sec): 23.25 - samples/sec: 1632.13 - lr: 0.000022 - momentum: 0.000000
2023-10-17 12:16:52,430 epoch 7 - iter 722/3617 - loss 0.01588108 - time (sec): 45.66 - samples/sec: 1649.57 - lr: 0.000021 - momentum: 0.000000
2023-10-17 12:17:14,472 epoch 7 - iter 1083/3617 - loss 0.01914681 - time (sec): 67.71 - samples/sec: 1671.24 - lr: 0.000021 - momentum: 0.000000
2023-10-17 12:17:38,387 epoch 7 - iter 1444/3617 - loss 0.02093559 - time (sec): 91.62 - samples/sec: 1648.96 - lr: 0.000020 - momentum: 0.000000
2023-10-17 12:18:00,039 epoch 7 - iter 1805/3617 - loss 0.02030467 - time (sec): 113.27 - samples/sec: 1670.47 - lr: 0.000019 - momentum: 0.000000
2023-10-17 12:18:22,176 epoch 7 - iter 2166/3617 - loss 0.02042856 - time (sec): 135.41 - samples/sec: 1675.92 - lr: 0.000019 - momentum: 0.000000
2023-10-17 12:18:45,119 epoch 7 - iter 2527/3617 - loss 0.01975550 - time (sec): 158.35 - samples/sec: 1676.74 - lr: 0.000018 - momentum: 0.000000
2023-10-17 12:19:04,588 epoch 7 - iter 2888/3617 - loss 0.01949421 - time (sec): 177.82 - samples/sec: 1705.58 - lr: 0.000018 - momentum: 0.000000
2023-10-17 12:19:22,591 epoch 7 - iter 3249/3617 - loss 0.01975215 - time (sec): 195.82 - samples/sec: 1746.76 - lr: 0.000017 - momentum: 0.000000
2023-10-17 12:19:44,496 epoch 7 - iter 3610/3617 - loss 0.02006605 - time (sec): 217.73 - samples/sec: 1742.14 - lr: 0.000017 - momentum: 0.000000
2023-10-17 12:19:44,901 ----------------------------------------------------------------------------------------------------
2023-10-17 12:19:44,901 EPOCH 7 done: loss 0.0202 - lr: 0.000017
2023-10-17 12:19:51,251 DEV : loss 0.317545086145401 - f1-score (micro avg) 0.6272
2023-10-17 12:19:51,295 ----------------------------------------------------------------------------------------------------
2023-10-17 12:20:13,563 epoch 8 - iter 361/3617 - loss 0.01412422 - time (sec): 22.27 - samples/sec: 1693.14 - lr: 0.000016 - momentum: 0.000000
2023-10-17 12:20:37,743 epoch 8 - iter 722/3617 - loss 0.01808065 - time (sec): 46.45 - samples/sec: 1612.50 - lr: 0.000016 - momentum: 0.000000
2023-10-17 12:21:02,137 epoch 8 - iter 1083/3617 - loss 0.01681718 - time (sec): 70.84 - samples/sec: 1590.05 - lr: 0.000015 - momentum: 0.000000
2023-10-17 12:21:26,146 epoch 8 - iter 1444/3617 - loss 0.01601246 - time (sec): 94.85 - samples/sec: 1600.06 - lr: 0.000014 - momentum: 0.000000
2023-10-17 12:21:49,292 epoch 8 - iter 1805/3617 - loss 0.01568364 - time (sec): 117.99 - samples/sec: 1600.39 - lr: 0.000014 - momentum: 0.000000
2023-10-17 12:22:11,848 epoch 8 - iter 2166/3617 - loss 0.01635340 - time (sec): 140.55 - samples/sec: 1609.35 - lr: 0.000013 - momentum: 0.000000
2023-10-17 12:22:33,797 epoch 8 - iter 2527/3617 - loss 0.01630735 - time (sec): 162.50 - samples/sec: 1622.79 - lr: 0.000013 - momentum: 0.000000
2023-10-17 12:22:55,815 epoch 8 - iter 2888/3617 - loss 0.01515567 - time (sec): 184.52 - samples/sec: 1637.90 - lr: 0.000012 - momentum: 0.000000
2023-10-17 12:23:18,342 epoch 8 - iter 3249/3617 - loss 0.01483412 - time (sec): 207.05 - samples/sec: 1648.57 - lr: 0.000012 - momentum: 0.000000
2023-10-17 12:23:41,314 epoch 8 - iter 3610/3617 - loss 0.01486930 - time (sec): 230.02 - samples/sec: 1648.51 - lr: 0.000011 - momentum: 0.000000
2023-10-17 12:23:41,781 ----------------------------------------------------------------------------------------------------
2023-10-17 12:23:41,781 EPOCH 8 done: loss 0.0148 - lr: 0.000011
2023-10-17 12:23:48,164 DEV : loss 0.3655739724636078 - f1-score (micro avg) 0.6602
2023-10-17 12:23:48,207 saving best model
2023-10-17 12:23:48,840 ----------------------------------------------------------------------------------------------------
2023-10-17 12:24:11,655 epoch 9 - iter 361/3617 - loss 0.00915400 - time (sec): 22.81 - samples/sec: 1667.62 - lr: 0.000011 - momentum: 0.000000
2023-10-17 12:24:34,394 epoch 9 - iter 722/3617 - loss 0.00828080 - time (sec): 45.55 - samples/sec: 1634.29 - lr: 0.000010 - momentum: 0.000000
2023-10-17 12:24:57,398 epoch 9 - iter 1083/3617 - loss 0.00838778 - time (sec): 68.56 - samples/sec: 1624.06 - lr: 0.000009 - momentum: 0.000000
2023-10-17 12:25:21,466 epoch 9 - iter 1444/3617 - loss 0.00854920 - time (sec): 92.62 - samples/sec: 1609.74 - lr: 0.000009 - momentum: 0.000000
2023-10-17 12:25:44,800 epoch 9 - iter 1805/3617 - loss 0.00847143 - time (sec): 115.96 - samples/sec: 1621.53 - lr: 0.000008 - momentum: 0.000000
2023-10-17 12:26:07,965 epoch 9 - iter 2166/3617 - loss 0.00871744 - time (sec): 139.12 - samples/sec: 1625.29 - lr: 0.000008 - momentum: 0.000000
2023-10-17 12:26:31,239 epoch 9 - iter 2527/3617 - loss 0.00823885 - time (sec): 162.40 - samples/sec: 1627.58 - lr: 0.000007 - momentum: 0.000000
2023-10-17 12:26:55,849 epoch 9 - iter 2888/3617 - loss 0.00773025 - time (sec): 187.01 - samples/sec: 1618.47 - lr: 0.000007 - momentum: 0.000000
2023-10-17 12:27:19,397 epoch 9 - iter 3249/3617 - loss 0.00760834 - time (sec): 210.55 - samples/sec: 1614.92 - lr: 0.000006 - momentum: 0.000000
2023-10-17 12:27:43,113 epoch 9 - iter 3610/3617 - loss 0.00769799 - time (sec): 234.27 - samples/sec: 1618.48 - lr: 0.000006 - momentum: 0.000000
2023-10-17 12:27:43,552 ----------------------------------------------------------------------------------------------------
2023-10-17 12:27:43,552 EPOCH 9 done: loss 0.0077 - lr: 0.000006
2023-10-17 12:27:49,962 DEV : loss 0.4163112938404083 - f1-score (micro avg) 0.6467
2023-10-17 12:27:50,004 ----------------------------------------------------------------------------------------------------
2023-10-17 12:28:13,013 epoch 10 - iter 361/3617 - loss 0.00413246 - time (sec): 23.01 - samples/sec: 1593.63 - lr: 0.000005 - momentum: 0.000000
2023-10-17 12:28:35,587 epoch 10 - iter 722/3617 - loss 0.00334843 - time (sec): 45.58 - samples/sec: 1657.68 - lr: 0.000004 - momentum: 0.000000
2023-10-17 12:28:58,768 epoch 10 - iter 1083/3617 - loss 0.00444223 - time (sec): 68.76 - samples/sec: 1625.52 - lr: 0.000004 - momentum: 0.000000
2023-10-17 12:29:22,388 epoch 10 - iter 1444/3617 - loss 0.00421888 - time (sec): 92.38 - samples/sec: 1628.53 - lr: 0.000003 - momentum: 0.000000
2023-10-17 12:29:45,256 epoch 10 - iter 1805/3617 - loss 0.00399761 - time (sec): 115.25 - samples/sec: 1631.71 - lr: 0.000003 - momentum: 0.000000
2023-10-17 12:30:07,935 epoch 10 - iter 2166/3617 - loss 0.00472994 - time (sec): 137.93 - samples/sec: 1641.01 - lr: 0.000002 - momentum: 0.000000
2023-10-17 12:30:31,147 epoch 10 - iter 2527/3617 - loss 0.00456748 - time (sec): 161.14 - samples/sec: 1631.95 - lr: 0.000002 - momentum: 0.000000
2023-10-17 12:30:54,698 epoch 10 - iter 2888/3617 - loss 0.00474495 - time (sec): 184.69 - samples/sec: 1636.72 - lr: 0.000001 - momentum: 0.000000
2023-10-17 12:31:16,709 epoch 10 - iter 3249/3617 - loss 0.00463207 - time (sec): 206.70 - samples/sec: 1652.97 - lr: 0.000001 - momentum: 0.000000
2023-10-17 12:31:40,217 epoch 10 - iter 3610/3617 - loss 0.00465381 - time (sec): 230.21 - samples/sec: 1647.46 - lr: 0.000000 - momentum: 0.000000
2023-10-17 12:31:40,665 ----------------------------------------------------------------------------------------------------
2023-10-17 12:31:40,665 EPOCH 10 done: loss 0.0046 - lr: 0.000000
2023-10-17 12:31:47,136 DEV : loss 0.43618330359458923 - f1-score (micro avg) 0.6512
2023-10-17 12:31:48,461 ----------------------------------------------------------------------------------------------------
2023-10-17 12:31:48,463 Loading model from best epoch ...
2023-10-17 12:31:50,256 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 12:31:58,239
Results:
- F-score (micro) 0.6485
- F-score (macro) 0.4908
- Accuracy 0.4915
By class:
precision recall f1-score support
loc 0.6662 0.7733 0.7157 591
pers 0.5293 0.7339 0.6150 357
org 0.2353 0.1013 0.1416 79
micro avg 0.5984 0.7079 0.6485 1027
macro avg 0.4769 0.5361 0.4908 1027
weighted avg 0.5855 0.7079 0.6366 1027
2023-10-17 12:31:58,239 ----------------------------------------------------------------------------------------------------
|