File size: 23,939 Bytes
d6a806b |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 |
2023-10-13 20:53:18,545 ----------------------------------------------------------------------------------------------------
2023-10-13 20:53:18,546 Model: "SequenceTagger(
(embeddings): TransformerWordEmbeddings(
(model): BertModel(
(embeddings): BertEmbeddings(
(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): BertEncoder(
(layer): ModuleList(
(0-11): 12 x BertLayer(
(attention): BertAttention(
(self): BertSelfAttention(
(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): BertSelfOutput(
(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): BertIntermediate(
(dense): Linear(in_features=768, out_features=3072, bias=True)
(intermediate_act_fn): GELUActivation()
)
(output): BertOutput(
(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)
)
)
)
)
(pooler): BertPooler(
(dense): Linear(in_features=768, out_features=768, bias=True)
(activation): Tanh()
)
)
)
(locked_dropout): LockedDropout(p=0.5)
(linear): Linear(in_features=768, out_features=13, bias=True)
(loss_function): CrossEntropyLoss()
)"
2023-10-13 20:53:18,546 ----------------------------------------------------------------------------------------------------
2023-10-13 20:53:18,547 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
2023-10-13 20:53:18,547 ----------------------------------------------------------------------------------------------------
2023-10-13 20:53:18,547 Train: 7936 sentences
2023-10-13 20:53:18,547 (train_with_dev=False, train_with_test=False)
2023-10-13 20:53:18,547 ----------------------------------------------------------------------------------------------------
2023-10-13 20:53:18,547 Training Params:
2023-10-13 20:53:18,547 - learning_rate: "3e-05"
2023-10-13 20:53:18,547 - mini_batch_size: "4"
2023-10-13 20:53:18,547 - max_epochs: "10"
2023-10-13 20:53:18,547 - shuffle: "True"
2023-10-13 20:53:18,547 ----------------------------------------------------------------------------------------------------
2023-10-13 20:53:18,547 Plugins:
2023-10-13 20:53:18,547 - LinearScheduler | warmup_fraction: '0.1'
2023-10-13 20:53:18,547 ----------------------------------------------------------------------------------------------------
2023-10-13 20:53:18,547 Final evaluation on model from best epoch (best-model.pt)
2023-10-13 20:53:18,547 - metric: "('micro avg', 'f1-score')"
2023-10-13 20:53:18,547 ----------------------------------------------------------------------------------------------------
2023-10-13 20:53:18,547 Computation:
2023-10-13 20:53:18,547 - compute on device: cuda:0
2023-10-13 20:53:18,547 - embedding storage: none
2023-10-13 20:53:18,547 ----------------------------------------------------------------------------------------------------
2023-10-13 20:53:18,547 Model training base path: "hmbench-icdar/fr-dbmdz/bert-base-historic-multilingual-cased-bs4-wsFalse-e10-lr3e-05-poolingfirst-layers-1-crfFalse-1"
2023-10-13 20:53:18,547 ----------------------------------------------------------------------------------------------------
2023-10-13 20:53:18,547 ----------------------------------------------------------------------------------------------------
2023-10-13 20:53:28,676 epoch 1 - iter 198/1984 - loss 1.88770718 - time (sec): 10.13 - samples/sec: 1587.68 - lr: 0.000003 - momentum: 0.000000
2023-10-13 20:53:37,508 epoch 1 - iter 396/1984 - loss 1.12484439 - time (sec): 18.96 - samples/sec: 1717.72 - lr: 0.000006 - momentum: 0.000000
2023-10-13 20:53:46,295 epoch 1 - iter 594/1984 - loss 0.83522090 - time (sec): 27.75 - samples/sec: 1772.35 - lr: 0.000009 - momentum: 0.000000
2023-10-13 20:53:54,959 epoch 1 - iter 792/1984 - loss 0.67679270 - time (sec): 36.41 - samples/sec: 1793.37 - lr: 0.000012 - momentum: 0.000000
2023-10-13 20:54:03,579 epoch 1 - iter 990/1984 - loss 0.57703210 - time (sec): 45.03 - samples/sec: 1807.87 - lr: 0.000015 - momentum: 0.000000
2023-10-13 20:54:12,160 epoch 1 - iter 1188/1984 - loss 0.50510842 - time (sec): 53.61 - samples/sec: 1819.73 - lr: 0.000018 - momentum: 0.000000
2023-10-13 20:54:20,861 epoch 1 - iter 1386/1984 - loss 0.45213988 - time (sec): 62.31 - samples/sec: 1840.43 - lr: 0.000021 - momentum: 0.000000
2023-10-13 20:54:30,522 epoch 1 - iter 1584/1984 - loss 0.41118920 - time (sec): 71.97 - samples/sec: 1836.31 - lr: 0.000024 - momentum: 0.000000
2023-10-13 20:54:41,080 epoch 1 - iter 1782/1984 - loss 0.38328017 - time (sec): 82.53 - samples/sec: 1794.37 - lr: 0.000027 - momentum: 0.000000
2023-10-13 20:54:50,731 epoch 1 - iter 1980/1984 - loss 0.35941678 - time (sec): 92.18 - samples/sec: 1776.70 - lr: 0.000030 - momentum: 0.000000
2023-10-13 20:54:50,895 ----------------------------------------------------------------------------------------------------
2023-10-13 20:54:50,895 EPOCH 1 done: loss 0.3590 - lr: 0.000030
2023-10-13 20:54:53,891 DEV : loss 0.10573934763669968 - f1-score (micro avg) 0.6591
2023-10-13 20:54:53,910 saving best model
2023-10-13 20:54:54,647 ----------------------------------------------------------------------------------------------------
2023-10-13 20:55:03,748 epoch 2 - iter 198/1984 - loss 0.12106577 - time (sec): 9.10 - samples/sec: 1896.84 - lr: 0.000030 - momentum: 0.000000
2023-10-13 20:55:12,515 epoch 2 - iter 396/1984 - loss 0.11964675 - time (sec): 17.87 - samples/sec: 1814.78 - lr: 0.000029 - momentum: 0.000000
2023-10-13 20:55:21,655 epoch 2 - iter 594/1984 - loss 0.11534460 - time (sec): 27.01 - samples/sec: 1845.31 - lr: 0.000029 - momentum: 0.000000
2023-10-13 20:55:30,327 epoch 2 - iter 792/1984 - loss 0.11408993 - time (sec): 35.68 - samples/sec: 1814.80 - lr: 0.000029 - momentum: 0.000000
2023-10-13 20:55:39,134 epoch 2 - iter 990/1984 - loss 0.11433753 - time (sec): 44.49 - samples/sec: 1846.50 - lr: 0.000028 - momentum: 0.000000
2023-10-13 20:55:47,726 epoch 2 - iter 1188/1984 - loss 0.11332009 - time (sec): 53.08 - samples/sec: 1848.33 - lr: 0.000028 - momentum: 0.000000
2023-10-13 20:55:57,506 epoch 2 - iter 1386/1984 - loss 0.11366621 - time (sec): 62.86 - samples/sec: 1823.09 - lr: 0.000028 - momentum: 0.000000
2023-10-13 20:56:08,043 epoch 2 - iter 1584/1984 - loss 0.11325580 - time (sec): 73.39 - samples/sec: 1778.70 - lr: 0.000027 - momentum: 0.000000
2023-10-13 20:56:18,565 epoch 2 - iter 1782/1984 - loss 0.11283397 - time (sec): 83.92 - samples/sec: 1755.40 - lr: 0.000027 - momentum: 0.000000
2023-10-13 20:56:28,780 epoch 2 - iter 1980/1984 - loss 0.11160017 - time (sec): 94.13 - samples/sec: 1739.67 - lr: 0.000027 - momentum: 0.000000
2023-10-13 20:56:28,992 ----------------------------------------------------------------------------------------------------
2023-10-13 20:56:28,992 EPOCH 2 done: loss 0.1115 - lr: 0.000027
2023-10-13 20:56:32,481 DEV : loss 0.1008807048201561 - f1-score (micro avg) 0.7066
2023-10-13 20:56:32,506 saving best model
2023-10-13 20:56:33,183 ----------------------------------------------------------------------------------------------------
2023-10-13 20:56:43,466 epoch 3 - iter 198/1984 - loss 0.07917549 - time (sec): 10.28 - samples/sec: 1593.09 - lr: 0.000026 - momentum: 0.000000
2023-10-13 20:56:52,745 epoch 3 - iter 396/1984 - loss 0.07784288 - time (sec): 19.56 - samples/sec: 1629.87 - lr: 0.000026 - momentum: 0.000000
2023-10-13 20:57:02,097 epoch 3 - iter 594/1984 - loss 0.07362250 - time (sec): 28.91 - samples/sec: 1709.82 - lr: 0.000026 - momentum: 0.000000
2023-10-13 20:57:11,220 epoch 3 - iter 792/1984 - loss 0.07809181 - time (sec): 38.04 - samples/sec: 1752.17 - lr: 0.000025 - momentum: 0.000000
2023-10-13 20:57:20,434 epoch 3 - iter 990/1984 - loss 0.08051703 - time (sec): 47.25 - samples/sec: 1753.43 - lr: 0.000025 - momentum: 0.000000
2023-10-13 20:57:29,676 epoch 3 - iter 1188/1984 - loss 0.08164550 - time (sec): 56.49 - samples/sec: 1752.99 - lr: 0.000025 - momentum: 0.000000
2023-10-13 20:57:38,750 epoch 3 - iter 1386/1984 - loss 0.08050737 - time (sec): 65.57 - samples/sec: 1753.84 - lr: 0.000024 - momentum: 0.000000
2023-10-13 20:57:47,724 epoch 3 - iter 1584/1984 - loss 0.08301541 - time (sec): 74.54 - samples/sec: 1764.39 - lr: 0.000024 - momentum: 0.000000
2023-10-13 20:57:56,733 epoch 3 - iter 1782/1984 - loss 0.08278631 - time (sec): 83.55 - samples/sec: 1768.46 - lr: 0.000024 - momentum: 0.000000
2023-10-13 20:58:05,774 epoch 3 - iter 1980/1984 - loss 0.08178188 - time (sec): 92.59 - samples/sec: 1766.44 - lr: 0.000023 - momentum: 0.000000
2023-10-13 20:58:05,956 ----------------------------------------------------------------------------------------------------
2023-10-13 20:58:05,956 EPOCH 3 done: loss 0.0817 - lr: 0.000023
2023-10-13 20:58:09,397 DEV : loss 0.14147229492664337 - f1-score (micro avg) 0.7464
2023-10-13 20:58:09,417 saving best model
2023-10-13 20:58:09,979 ----------------------------------------------------------------------------------------------------
2023-10-13 20:58:18,977 epoch 4 - iter 198/1984 - loss 0.05757599 - time (sec): 8.99 - samples/sec: 1838.23 - lr: 0.000023 - momentum: 0.000000
2023-10-13 20:58:28,176 epoch 4 - iter 396/1984 - loss 0.06403415 - time (sec): 18.19 - samples/sec: 1827.43 - lr: 0.000023 - momentum: 0.000000
2023-10-13 20:58:37,156 epoch 4 - iter 594/1984 - loss 0.06415865 - time (sec): 27.17 - samples/sec: 1805.48 - lr: 0.000022 - momentum: 0.000000
2023-10-13 20:58:46,425 epoch 4 - iter 792/1984 - loss 0.06303083 - time (sec): 36.44 - samples/sec: 1785.64 - lr: 0.000022 - momentum: 0.000000
2023-10-13 20:58:55,437 epoch 4 - iter 990/1984 - loss 0.06259595 - time (sec): 45.45 - samples/sec: 1791.39 - lr: 0.000022 - momentum: 0.000000
2023-10-13 20:59:04,506 epoch 4 - iter 1188/1984 - loss 0.06147075 - time (sec): 54.52 - samples/sec: 1791.86 - lr: 0.000021 - momentum: 0.000000
2023-10-13 20:59:13,860 epoch 4 - iter 1386/1984 - loss 0.06547581 - time (sec): 63.88 - samples/sec: 1784.41 - lr: 0.000021 - momentum: 0.000000
2023-10-13 20:59:23,431 epoch 4 - iter 1584/1984 - loss 0.06402360 - time (sec): 73.45 - samples/sec: 1769.42 - lr: 0.000021 - momentum: 0.000000
2023-10-13 20:59:32,955 epoch 4 - iter 1782/1984 - loss 0.06275327 - time (sec): 82.97 - samples/sec: 1766.30 - lr: 0.000020 - momentum: 0.000000
2023-10-13 20:59:42,166 epoch 4 - iter 1980/1984 - loss 0.06390400 - time (sec): 92.18 - samples/sec: 1776.30 - lr: 0.000020 - momentum: 0.000000
2023-10-13 20:59:42,347 ----------------------------------------------------------------------------------------------------
2023-10-13 20:59:42,347 EPOCH 4 done: loss 0.0639 - lr: 0.000020
2023-10-13 20:59:46,341 DEV : loss 0.14689981937408447 - f1-score (micro avg) 0.7489
2023-10-13 20:59:46,360 saving best model
2023-10-13 20:59:46,969 ----------------------------------------------------------------------------------------------------
2023-10-13 20:59:56,191 epoch 5 - iter 198/1984 - loss 0.04771070 - time (sec): 9.22 - samples/sec: 1804.52 - lr: 0.000020 - momentum: 0.000000
2023-10-13 21:00:05,681 epoch 5 - iter 396/1984 - loss 0.04649118 - time (sec): 18.71 - samples/sec: 1761.13 - lr: 0.000019 - momentum: 0.000000
2023-10-13 21:00:15,005 epoch 5 - iter 594/1984 - loss 0.04834649 - time (sec): 28.03 - samples/sec: 1757.30 - lr: 0.000019 - momentum: 0.000000
2023-10-13 21:00:24,352 epoch 5 - iter 792/1984 - loss 0.04615015 - time (sec): 37.38 - samples/sec: 1783.56 - lr: 0.000019 - momentum: 0.000000
2023-10-13 21:00:33,340 epoch 5 - iter 990/1984 - loss 0.04648506 - time (sec): 46.37 - samples/sec: 1794.36 - lr: 0.000018 - momentum: 0.000000
2023-10-13 21:00:42,270 epoch 5 - iter 1188/1984 - loss 0.04468332 - time (sec): 55.30 - samples/sec: 1794.29 - lr: 0.000018 - momentum: 0.000000
2023-10-13 21:00:51,141 epoch 5 - iter 1386/1984 - loss 0.04510172 - time (sec): 64.17 - samples/sec: 1793.83 - lr: 0.000018 - momentum: 0.000000
2023-10-13 21:01:00,022 epoch 5 - iter 1584/1984 - loss 0.04472638 - time (sec): 73.05 - samples/sec: 1801.23 - lr: 0.000017 - momentum: 0.000000
2023-10-13 21:01:09,020 epoch 5 - iter 1782/1984 - loss 0.04570004 - time (sec): 82.05 - samples/sec: 1804.48 - lr: 0.000017 - momentum: 0.000000
2023-10-13 21:01:17,633 epoch 5 - iter 1980/1984 - loss 0.04591442 - time (sec): 90.66 - samples/sec: 1804.42 - lr: 0.000017 - momentum: 0.000000
2023-10-13 21:01:17,818 ----------------------------------------------------------------------------------------------------
2023-10-13 21:01:17,818 EPOCH 5 done: loss 0.0460 - lr: 0.000017
2023-10-13 21:01:21,249 DEV : loss 0.16444256901741028 - f1-score (micro avg) 0.7731
2023-10-13 21:01:21,269 saving best model
2023-10-13 21:01:21,874 ----------------------------------------------------------------------------------------------------
2023-10-13 21:01:30,982 epoch 6 - iter 198/1984 - loss 0.03290051 - time (sec): 9.11 - samples/sec: 1821.35 - lr: 0.000016 - momentum: 0.000000
2023-10-13 21:01:40,070 epoch 6 - iter 396/1984 - loss 0.03321950 - time (sec): 18.19 - samples/sec: 1858.59 - lr: 0.000016 - momentum: 0.000000
2023-10-13 21:01:49,239 epoch 6 - iter 594/1984 - loss 0.03311731 - time (sec): 27.36 - samples/sec: 1815.85 - lr: 0.000016 - momentum: 0.000000
2023-10-13 21:01:58,266 epoch 6 - iter 792/1984 - loss 0.03061596 - time (sec): 36.39 - samples/sec: 1801.60 - lr: 0.000015 - momentum: 0.000000
2023-10-13 21:02:07,399 epoch 6 - iter 990/1984 - loss 0.03289182 - time (sec): 45.52 - samples/sec: 1799.81 - lr: 0.000015 - momentum: 0.000000
2023-10-13 21:02:16,317 epoch 6 - iter 1188/1984 - loss 0.03218297 - time (sec): 54.44 - samples/sec: 1807.03 - lr: 0.000015 - momentum: 0.000000
2023-10-13 21:02:25,287 epoch 6 - iter 1386/1984 - loss 0.03348283 - time (sec): 63.41 - samples/sec: 1815.90 - lr: 0.000014 - momentum: 0.000000
2023-10-13 21:02:34,604 epoch 6 - iter 1584/1984 - loss 0.03361440 - time (sec): 72.73 - samples/sec: 1807.15 - lr: 0.000014 - momentum: 0.000000
2023-10-13 21:02:43,949 epoch 6 - iter 1782/1984 - loss 0.03357184 - time (sec): 82.07 - samples/sec: 1796.19 - lr: 0.000014 - momentum: 0.000000
2023-10-13 21:02:53,471 epoch 6 - iter 1980/1984 - loss 0.03361402 - time (sec): 91.59 - samples/sec: 1786.91 - lr: 0.000013 - momentum: 0.000000
2023-10-13 21:02:53,648 ----------------------------------------------------------------------------------------------------
2023-10-13 21:02:53,648 EPOCH 6 done: loss 0.0336 - lr: 0.000013
2023-10-13 21:02:57,584 DEV : loss 0.20081490278244019 - f1-score (micro avg) 0.7536
2023-10-13 21:02:57,604 ----------------------------------------------------------------------------------------------------
2023-10-13 21:03:06,697 epoch 7 - iter 198/1984 - loss 0.02416311 - time (sec): 9.09 - samples/sec: 1689.96 - lr: 0.000013 - momentum: 0.000000
2023-10-13 21:03:15,884 epoch 7 - iter 396/1984 - loss 0.02612239 - time (sec): 18.28 - samples/sec: 1761.83 - lr: 0.000013 - momentum: 0.000000
2023-10-13 21:03:24,877 epoch 7 - iter 594/1984 - loss 0.02503909 - time (sec): 27.27 - samples/sec: 1752.06 - lr: 0.000012 - momentum: 0.000000
2023-10-13 21:03:33,834 epoch 7 - iter 792/1984 - loss 0.02544103 - time (sec): 36.23 - samples/sec: 1765.44 - lr: 0.000012 - momentum: 0.000000
2023-10-13 21:03:42,811 epoch 7 - iter 990/1984 - loss 0.02481477 - time (sec): 45.21 - samples/sec: 1795.09 - lr: 0.000012 - momentum: 0.000000
2023-10-13 21:03:52,055 epoch 7 - iter 1188/1984 - loss 0.02586965 - time (sec): 54.45 - samples/sec: 1808.49 - lr: 0.000011 - momentum: 0.000000
2023-10-13 21:04:01,424 epoch 7 - iter 1386/1984 - loss 0.02624069 - time (sec): 63.82 - samples/sec: 1781.30 - lr: 0.000011 - momentum: 0.000000
2023-10-13 21:04:10,658 epoch 7 - iter 1584/1984 - loss 0.02564693 - time (sec): 73.05 - samples/sec: 1788.89 - lr: 0.000011 - momentum: 0.000000
2023-10-13 21:04:19,928 epoch 7 - iter 1782/1984 - loss 0.02628502 - time (sec): 82.32 - samples/sec: 1786.06 - lr: 0.000010 - momentum: 0.000000
2023-10-13 21:04:29,051 epoch 7 - iter 1980/1984 - loss 0.02588685 - time (sec): 91.45 - samples/sec: 1788.99 - lr: 0.000010 - momentum: 0.000000
2023-10-13 21:04:29,240 ----------------------------------------------------------------------------------------------------
2023-10-13 21:04:29,240 EPOCH 7 done: loss 0.0258 - lr: 0.000010
2023-10-13 21:04:32,591 DEV : loss 0.19746360182762146 - f1-score (micro avg) 0.7682
2023-10-13 21:04:32,611 ----------------------------------------------------------------------------------------------------
2023-10-13 21:04:42,128 epoch 8 - iter 198/1984 - loss 0.01260002 - time (sec): 9.52 - samples/sec: 1716.88 - lr: 0.000010 - momentum: 0.000000
2023-10-13 21:04:51,190 epoch 8 - iter 396/1984 - loss 0.01660345 - time (sec): 18.58 - samples/sec: 1773.92 - lr: 0.000009 - momentum: 0.000000
2023-10-13 21:05:00,514 epoch 8 - iter 594/1984 - loss 0.01395853 - time (sec): 27.90 - samples/sec: 1750.34 - lr: 0.000009 - momentum: 0.000000
2023-10-13 21:05:09,660 epoch 8 - iter 792/1984 - loss 0.01619325 - time (sec): 37.05 - samples/sec: 1776.28 - lr: 0.000009 - momentum: 0.000000
2023-10-13 21:05:18,644 epoch 8 - iter 990/1984 - loss 0.01676123 - time (sec): 46.03 - samples/sec: 1778.63 - lr: 0.000008 - momentum: 0.000000
2023-10-13 21:05:27,517 epoch 8 - iter 1188/1984 - loss 0.01715802 - time (sec): 54.91 - samples/sec: 1771.93 - lr: 0.000008 - momentum: 0.000000
2023-10-13 21:05:36,488 epoch 8 - iter 1386/1984 - loss 0.01772604 - time (sec): 63.88 - samples/sec: 1784.87 - lr: 0.000008 - momentum: 0.000000
2023-10-13 21:05:45,810 epoch 8 - iter 1584/1984 - loss 0.01778083 - time (sec): 73.20 - samples/sec: 1797.68 - lr: 0.000007 - momentum: 0.000000
2023-10-13 21:05:54,770 epoch 8 - iter 1782/1984 - loss 0.01759045 - time (sec): 82.16 - samples/sec: 1794.86 - lr: 0.000007 - momentum: 0.000000
2023-10-13 21:06:04,075 epoch 8 - iter 1980/1984 - loss 0.01768193 - time (sec): 91.46 - samples/sec: 1790.18 - lr: 0.000007 - momentum: 0.000000
2023-10-13 21:06:04,255 ----------------------------------------------------------------------------------------------------
2023-10-13 21:06:04,255 EPOCH 8 done: loss 0.0177 - lr: 0.000007
2023-10-13 21:06:07,624 DEV : loss 0.2036585807800293 - f1-score (micro avg) 0.7669
2023-10-13 21:06:07,644 ----------------------------------------------------------------------------------------------------
2023-10-13 21:06:16,928 epoch 9 - iter 198/1984 - loss 0.01121945 - time (sec): 9.28 - samples/sec: 1803.90 - lr: 0.000006 - momentum: 0.000000
2023-10-13 21:06:25,941 epoch 9 - iter 396/1984 - loss 0.00942507 - time (sec): 18.30 - samples/sec: 1791.88 - lr: 0.000006 - momentum: 0.000000
2023-10-13 21:06:34,795 epoch 9 - iter 594/1984 - loss 0.01115411 - time (sec): 27.15 - samples/sec: 1791.55 - lr: 0.000006 - momentum: 0.000000
2023-10-13 21:06:43,917 epoch 9 - iter 792/1984 - loss 0.01115229 - time (sec): 36.27 - samples/sec: 1813.41 - lr: 0.000005 - momentum: 0.000000
2023-10-13 21:06:53,144 epoch 9 - iter 990/1984 - loss 0.01139535 - time (sec): 45.50 - samples/sec: 1813.79 - lr: 0.000005 - momentum: 0.000000
2023-10-13 21:07:02,159 epoch 9 - iter 1188/1984 - loss 0.01109025 - time (sec): 54.51 - samples/sec: 1799.43 - lr: 0.000005 - momentum: 0.000000
2023-10-13 21:07:11,471 epoch 9 - iter 1386/1984 - loss 0.01120806 - time (sec): 63.83 - samples/sec: 1788.63 - lr: 0.000004 - momentum: 0.000000
2023-10-13 21:07:20,727 epoch 9 - iter 1584/1984 - loss 0.01155852 - time (sec): 73.08 - samples/sec: 1792.50 - lr: 0.000004 - momentum: 0.000000
2023-10-13 21:07:29,843 epoch 9 - iter 1782/1984 - loss 0.01241535 - time (sec): 82.20 - samples/sec: 1798.57 - lr: 0.000004 - momentum: 0.000000
2023-10-13 21:07:38,723 epoch 9 - iter 1980/1984 - loss 0.01230984 - time (sec): 91.08 - samples/sec: 1796.01 - lr: 0.000003 - momentum: 0.000000
2023-10-13 21:07:38,909 ----------------------------------------------------------------------------------------------------
2023-10-13 21:07:38,909 EPOCH 9 done: loss 0.0123 - lr: 0.000003
2023-10-13 21:07:42,702 DEV : loss 0.21578332781791687 - f1-score (micro avg) 0.7644
2023-10-13 21:07:42,722 ----------------------------------------------------------------------------------------------------
2023-10-13 21:07:52,175 epoch 10 - iter 198/1984 - loss 0.01257622 - time (sec): 9.45 - samples/sec: 1803.26 - lr: 0.000003 - momentum: 0.000000
2023-10-13 21:08:01,164 epoch 10 - iter 396/1984 - loss 0.01088765 - time (sec): 18.44 - samples/sec: 1777.14 - lr: 0.000003 - momentum: 0.000000
2023-10-13 21:08:10,229 epoch 10 - iter 594/1984 - loss 0.00982544 - time (sec): 27.51 - samples/sec: 1774.23 - lr: 0.000002 - momentum: 0.000000
2023-10-13 21:08:19,951 epoch 10 - iter 792/1984 - loss 0.00945824 - time (sec): 37.23 - samples/sec: 1758.02 - lr: 0.000002 - momentum: 0.000000
2023-10-13 21:08:29,388 epoch 10 - iter 990/1984 - loss 0.00969098 - time (sec): 46.66 - samples/sec: 1746.07 - lr: 0.000002 - momentum: 0.000000
2023-10-13 21:08:38,684 epoch 10 - iter 1188/1984 - loss 0.00953138 - time (sec): 55.96 - samples/sec: 1750.67 - lr: 0.000001 - momentum: 0.000000
2023-10-13 21:08:47,952 epoch 10 - iter 1386/1984 - loss 0.00928403 - time (sec): 65.23 - samples/sec: 1752.64 - lr: 0.000001 - momentum: 0.000000
2023-10-13 21:08:56,924 epoch 10 - iter 1584/1984 - loss 0.00941987 - time (sec): 74.20 - samples/sec: 1764.57 - lr: 0.000001 - momentum: 0.000000
2023-10-13 21:09:06,116 epoch 10 - iter 1782/1984 - loss 0.00912027 - time (sec): 83.39 - samples/sec: 1774.72 - lr: 0.000000 - momentum: 0.000000
2023-10-13 21:09:15,364 epoch 10 - iter 1980/1984 - loss 0.00877729 - time (sec): 92.64 - samples/sec: 1766.45 - lr: 0.000000 - momentum: 0.000000
2023-10-13 21:09:15,565 ----------------------------------------------------------------------------------------------------
2023-10-13 21:09:15,565 EPOCH 10 done: loss 0.0088 - lr: 0.000000
2023-10-13 21:09:18,982 DEV : loss 0.22608202695846558 - f1-score (micro avg) 0.767
2023-10-13 21:09:19,478 ----------------------------------------------------------------------------------------------------
2023-10-13 21:09:19,479 Loading model from best epoch ...
2023-10-13 21:09:21,398 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
2023-10-13 21:09:24,698
Results:
- F-score (micro) 0.7846
- F-score (macro) 0.6793
- Accuracy 0.6641
By class:
precision recall f1-score support
LOC 0.8554 0.8489 0.8521 655
PER 0.7449 0.8117 0.7768 223
ORG 0.4694 0.3622 0.4089 127
micro avg 0.7901 0.7791 0.7846 1005
macro avg 0.6899 0.6742 0.6793 1005
weighted avg 0.7821 0.7791 0.7794 1005
2023-10-13 21:09:24,699 ----------------------------------------------------------------------------------------------------
|