File size: 23,946 Bytes
e53270a |
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 10:37:47,590 ----------------------------------------------------------------------------------------------------
2023-10-17 10:37:47,591 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 10:37:47,591 ----------------------------------------------------------------------------------------------------
2023-10-17 10:37:47,591 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
2023-10-17 10:37:47,591 ----------------------------------------------------------------------------------------------------
2023-10-17 10:37:47,591 Train: 7936 sentences
2023-10-17 10:37:47,591 (train_with_dev=False, train_with_test=False)
2023-10-17 10:37:47,592 ----------------------------------------------------------------------------------------------------
2023-10-17 10:37:47,592 Training Params:
2023-10-17 10:37:47,592 - learning_rate: "3e-05"
2023-10-17 10:37:47,592 - mini_batch_size: "4"
2023-10-17 10:37:47,592 - max_epochs: "10"
2023-10-17 10:37:47,592 - shuffle: "True"
2023-10-17 10:37:47,592 ----------------------------------------------------------------------------------------------------
2023-10-17 10:37:47,592 Plugins:
2023-10-17 10:37:47,592 - TensorboardLogger
2023-10-17 10:37:47,592 - LinearScheduler | warmup_fraction: '0.1'
2023-10-17 10:37:47,592 ----------------------------------------------------------------------------------------------------
2023-10-17 10:37:47,592 Final evaluation on model from best epoch (best-model.pt)
2023-10-17 10:37:47,592 - metric: "('micro avg', 'f1-score')"
2023-10-17 10:37:47,592 ----------------------------------------------------------------------------------------------------
2023-10-17 10:37:47,592 Computation:
2023-10-17 10:37:47,592 - compute on device: cuda:0
2023-10-17 10:37:47,592 - embedding storage: none
2023-10-17 10:37:47,592 ----------------------------------------------------------------------------------------------------
2023-10-17 10:37:47,592 Model training base path: "hmbench-icdar/fr-hmteams/teams-base-historic-multilingual-discriminator-bs4-wsFalse-e10-lr3e-05-poolingfirst-layers-1-crfFalse-1"
2023-10-17 10:37:47,592 ----------------------------------------------------------------------------------------------------
2023-10-17 10:37:47,592 ----------------------------------------------------------------------------------------------------
2023-10-17 10:37:47,592 Logging anything other than scalars to TensorBoard is currently not supported.
2023-10-17 10:37:57,473 epoch 1 - iter 198/1984 - loss 1.96262558 - time (sec): 9.88 - samples/sec: 1637.74 - lr: 0.000003 - momentum: 0.000000
2023-10-17 10:38:06,276 epoch 1 - iter 396/1984 - loss 1.14353968 - time (sec): 18.68 - samples/sec: 1779.11 - lr: 0.000006 - momentum: 0.000000
2023-10-17 10:38:15,247 epoch 1 - iter 594/1984 - loss 0.85110538 - time (sec): 27.65 - samples/sec: 1794.33 - lr: 0.000009 - momentum: 0.000000
2023-10-17 10:38:23,912 epoch 1 - iter 792/1984 - loss 0.69733723 - time (sec): 36.32 - samples/sec: 1800.50 - lr: 0.000012 - momentum: 0.000000
2023-10-17 10:38:32,916 epoch 1 - iter 990/1984 - loss 0.59220786 - time (sec): 45.32 - samples/sec: 1805.64 - lr: 0.000015 - momentum: 0.000000
2023-10-17 10:38:41,867 epoch 1 - iter 1188/1984 - loss 0.51720323 - time (sec): 54.27 - samples/sec: 1817.61 - lr: 0.000018 - momentum: 0.000000
2023-10-17 10:38:50,819 epoch 1 - iter 1386/1984 - loss 0.46529206 - time (sec): 63.23 - samples/sec: 1815.63 - lr: 0.000021 - momentum: 0.000000
2023-10-17 10:38:59,939 epoch 1 - iter 1584/1984 - loss 0.42634007 - time (sec): 72.35 - samples/sec: 1814.85 - lr: 0.000024 - momentum: 0.000000
2023-10-17 10:39:09,082 epoch 1 - iter 1782/1984 - loss 0.39543817 - time (sec): 81.49 - samples/sec: 1806.96 - lr: 0.000027 - momentum: 0.000000
2023-10-17 10:39:18,133 epoch 1 - iter 1980/1984 - loss 0.37002480 - time (sec): 90.54 - samples/sec: 1807.55 - lr: 0.000030 - momentum: 0.000000
2023-10-17 10:39:18,315 ----------------------------------------------------------------------------------------------------
2023-10-17 10:39:18,315 EPOCH 1 done: loss 0.3695 - lr: 0.000030
2023-10-17 10:39:22,176 DEV : loss 0.09817007929086685 - f1-score (micro avg) 0.7151
2023-10-17 10:39:22,205 saving best model
2023-10-17 10:39:22,659 ----------------------------------------------------------------------------------------------------
2023-10-17 10:39:32,314 epoch 2 - iter 198/1984 - loss 0.11703805 - time (sec): 9.65 - samples/sec: 1751.29 - lr: 0.000030 - momentum: 0.000000
2023-10-17 10:39:41,399 epoch 2 - iter 396/1984 - loss 0.12135474 - time (sec): 18.74 - samples/sec: 1767.42 - lr: 0.000029 - momentum: 0.000000
2023-10-17 10:39:50,558 epoch 2 - iter 594/1984 - loss 0.12116589 - time (sec): 27.90 - samples/sec: 1778.79 - lr: 0.000029 - momentum: 0.000000
2023-10-17 10:39:59,873 epoch 2 - iter 792/1984 - loss 0.11894017 - time (sec): 37.21 - samples/sec: 1785.54 - lr: 0.000029 - momentum: 0.000000
2023-10-17 10:40:08,857 epoch 2 - iter 990/1984 - loss 0.11639961 - time (sec): 46.20 - samples/sec: 1787.34 - lr: 0.000028 - momentum: 0.000000
2023-10-17 10:40:18,317 epoch 2 - iter 1188/1984 - loss 0.11443201 - time (sec): 55.66 - samples/sec: 1784.94 - lr: 0.000028 - momentum: 0.000000
2023-10-17 10:40:27,227 epoch 2 - iter 1386/1984 - loss 0.11365206 - time (sec): 64.57 - samples/sec: 1796.15 - lr: 0.000028 - momentum: 0.000000
2023-10-17 10:40:36,279 epoch 2 - iter 1584/1984 - loss 0.11390623 - time (sec): 73.62 - samples/sec: 1788.69 - lr: 0.000027 - momentum: 0.000000
2023-10-17 10:40:45,274 epoch 2 - iter 1782/1984 - loss 0.11339437 - time (sec): 82.61 - samples/sec: 1782.61 - lr: 0.000027 - momentum: 0.000000
2023-10-17 10:40:54,345 epoch 2 - iter 1980/1984 - loss 0.11417783 - time (sec): 91.68 - samples/sec: 1784.96 - lr: 0.000027 - momentum: 0.000000
2023-10-17 10:40:54,529 ----------------------------------------------------------------------------------------------------
2023-10-17 10:40:54,529 EPOCH 2 done: loss 0.1142 - lr: 0.000027
2023-10-17 10:40:58,257 DEV : loss 0.09690196067094803 - f1-score (micro avg) 0.751
2023-10-17 10:40:58,286 saving best model
2023-10-17 10:40:58,864 ----------------------------------------------------------------------------------------------------
2023-10-17 10:41:08,525 epoch 3 - iter 198/1984 - loss 0.08243740 - time (sec): 9.66 - samples/sec: 1715.82 - lr: 0.000026 - momentum: 0.000000
2023-10-17 10:41:17,969 epoch 3 - iter 396/1984 - loss 0.08405210 - time (sec): 19.10 - samples/sec: 1720.39 - lr: 0.000026 - momentum: 0.000000
2023-10-17 10:41:28,471 epoch 3 - iter 594/1984 - loss 0.08612073 - time (sec): 29.60 - samples/sec: 1675.84 - lr: 0.000026 - momentum: 0.000000
2023-10-17 10:41:37,701 epoch 3 - iter 792/1984 - loss 0.08283051 - time (sec): 38.83 - samples/sec: 1710.41 - lr: 0.000025 - momentum: 0.000000
2023-10-17 10:41:46,586 epoch 3 - iter 990/1984 - loss 0.08274479 - time (sec): 47.72 - samples/sec: 1733.22 - lr: 0.000025 - momentum: 0.000000
2023-10-17 10:41:55,624 epoch 3 - iter 1188/1984 - loss 0.08251098 - time (sec): 56.76 - samples/sec: 1734.90 - lr: 0.000025 - momentum: 0.000000
2023-10-17 10:42:04,447 epoch 3 - iter 1386/1984 - loss 0.08323712 - time (sec): 65.58 - samples/sec: 1750.52 - lr: 0.000024 - momentum: 0.000000
2023-10-17 10:42:13,596 epoch 3 - iter 1584/1984 - loss 0.08296027 - time (sec): 74.73 - samples/sec: 1755.28 - lr: 0.000024 - momentum: 0.000000
2023-10-17 10:42:22,756 epoch 3 - iter 1782/1984 - loss 0.08247111 - time (sec): 83.89 - samples/sec: 1763.01 - lr: 0.000024 - momentum: 0.000000
2023-10-17 10:42:31,763 epoch 3 - iter 1980/1984 - loss 0.08393188 - time (sec): 92.90 - samples/sec: 1761.21 - lr: 0.000023 - momentum: 0.000000
2023-10-17 10:42:31,952 ----------------------------------------------------------------------------------------------------
2023-10-17 10:42:31,952 EPOCH 3 done: loss 0.0841 - lr: 0.000023
2023-10-17 10:42:35,645 DEV : loss 0.11184526234865189 - f1-score (micro avg) 0.7592
2023-10-17 10:42:35,669 saving best model
2023-10-17 10:42:36,275 ----------------------------------------------------------------------------------------------------
2023-10-17 10:42:45,911 epoch 4 - iter 198/1984 - loss 0.06238332 - time (sec): 9.63 - samples/sec: 1704.13 - lr: 0.000023 - momentum: 0.000000
2023-10-17 10:42:55,157 epoch 4 - iter 396/1984 - loss 0.06005836 - time (sec): 18.88 - samples/sec: 1808.83 - lr: 0.000023 - momentum: 0.000000
2023-10-17 10:43:04,798 epoch 4 - iter 594/1984 - loss 0.06300783 - time (sec): 28.52 - samples/sec: 1785.86 - lr: 0.000022 - momentum: 0.000000
2023-10-17 10:43:15,294 epoch 4 - iter 792/1984 - loss 0.06418959 - time (sec): 39.02 - samples/sec: 1728.50 - lr: 0.000022 - momentum: 0.000000
2023-10-17 10:43:25,402 epoch 4 - iter 990/1984 - loss 0.06410527 - time (sec): 49.12 - samples/sec: 1703.66 - lr: 0.000022 - momentum: 0.000000
2023-10-17 10:43:34,556 epoch 4 - iter 1188/1984 - loss 0.06627853 - time (sec): 58.28 - samples/sec: 1704.06 - lr: 0.000021 - momentum: 0.000000
2023-10-17 10:43:43,828 epoch 4 - iter 1386/1984 - loss 0.06493629 - time (sec): 67.55 - samples/sec: 1701.29 - lr: 0.000021 - momentum: 0.000000
2023-10-17 10:43:53,244 epoch 4 - iter 1584/1984 - loss 0.06437724 - time (sec): 76.97 - samples/sec: 1702.26 - lr: 0.000021 - momentum: 0.000000
2023-10-17 10:44:02,367 epoch 4 - iter 1782/1984 - loss 0.06554549 - time (sec): 86.09 - samples/sec: 1715.79 - lr: 0.000020 - momentum: 0.000000
2023-10-17 10:44:11,335 epoch 4 - iter 1980/1984 - loss 0.06532216 - time (sec): 95.06 - samples/sec: 1722.76 - lr: 0.000020 - momentum: 0.000000
2023-10-17 10:44:11,528 ----------------------------------------------------------------------------------------------------
2023-10-17 10:44:11,528 EPOCH 4 done: loss 0.0653 - lr: 0.000020
2023-10-17 10:44:15,291 DEV : loss 0.14710469543933868 - f1-score (micro avg) 0.7764
2023-10-17 10:44:15,314 saving best model
2023-10-17 10:44:15,834 ----------------------------------------------------------------------------------------------------
2023-10-17 10:44:24,991 epoch 5 - iter 198/1984 - loss 0.05310506 - time (sec): 9.16 - samples/sec: 1776.04 - lr: 0.000020 - momentum: 0.000000
2023-10-17 10:44:33,692 epoch 5 - iter 396/1984 - loss 0.05550943 - time (sec): 17.86 - samples/sec: 1876.49 - lr: 0.000019 - momentum: 0.000000
2023-10-17 10:44:42,752 epoch 5 - iter 594/1984 - loss 0.05128920 - time (sec): 26.92 - samples/sec: 1865.39 - lr: 0.000019 - momentum: 0.000000
2023-10-17 10:44:51,952 epoch 5 - iter 792/1984 - loss 0.05178507 - time (sec): 36.12 - samples/sec: 1875.12 - lr: 0.000019 - momentum: 0.000000
2023-10-17 10:45:01,073 epoch 5 - iter 990/1984 - loss 0.05015438 - time (sec): 45.24 - samples/sec: 1860.06 - lr: 0.000018 - momentum: 0.000000
2023-10-17 10:45:10,073 epoch 5 - iter 1188/1984 - loss 0.05003515 - time (sec): 54.24 - samples/sec: 1843.52 - lr: 0.000018 - momentum: 0.000000
2023-10-17 10:45:19,019 epoch 5 - iter 1386/1984 - loss 0.05017555 - time (sec): 63.18 - samples/sec: 1842.13 - lr: 0.000018 - momentum: 0.000000
2023-10-17 10:45:27,581 epoch 5 - iter 1584/1984 - loss 0.05031975 - time (sec): 71.75 - samples/sec: 1842.93 - lr: 0.000017 - momentum: 0.000000
2023-10-17 10:45:36,697 epoch 5 - iter 1782/1984 - loss 0.05007302 - time (sec): 80.86 - samples/sec: 1832.85 - lr: 0.000017 - momentum: 0.000000
2023-10-17 10:45:46,443 epoch 5 - iter 1980/1984 - loss 0.04916707 - time (sec): 90.61 - samples/sec: 1805.91 - lr: 0.000017 - momentum: 0.000000
2023-10-17 10:45:46,629 ----------------------------------------------------------------------------------------------------
2023-10-17 10:45:46,629 EPOCH 5 done: loss 0.0493 - lr: 0.000017
2023-10-17 10:45:50,160 DEV : loss 0.20937786996364594 - f1-score (micro avg) 0.761
2023-10-17 10:45:50,185 ----------------------------------------------------------------------------------------------------
2023-10-17 10:46:00,578 epoch 6 - iter 198/1984 - loss 0.03889353 - time (sec): 10.39 - samples/sec: 1582.61 - lr: 0.000016 - momentum: 0.000000
2023-10-17 10:46:09,604 epoch 6 - iter 396/1984 - loss 0.04124372 - time (sec): 19.42 - samples/sec: 1681.97 - lr: 0.000016 - momentum: 0.000000
2023-10-17 10:46:18,836 epoch 6 - iter 594/1984 - loss 0.04058583 - time (sec): 28.65 - samples/sec: 1753.98 - lr: 0.000016 - momentum: 0.000000
2023-10-17 10:46:27,818 epoch 6 - iter 792/1984 - loss 0.04104349 - time (sec): 37.63 - samples/sec: 1770.74 - lr: 0.000015 - momentum: 0.000000
2023-10-17 10:46:36,799 epoch 6 - iter 990/1984 - loss 0.03959461 - time (sec): 46.61 - samples/sec: 1789.31 - lr: 0.000015 - momentum: 0.000000
2023-10-17 10:46:45,771 epoch 6 - iter 1188/1984 - loss 0.03893122 - time (sec): 55.58 - samples/sec: 1793.74 - lr: 0.000015 - momentum: 0.000000
2023-10-17 10:46:54,632 epoch 6 - iter 1386/1984 - loss 0.03835506 - time (sec): 64.45 - samples/sec: 1789.43 - lr: 0.000014 - momentum: 0.000000
2023-10-17 10:47:03,663 epoch 6 - iter 1584/1984 - loss 0.03800712 - time (sec): 73.48 - samples/sec: 1785.14 - lr: 0.000014 - momentum: 0.000000
2023-10-17 10:47:12,792 epoch 6 - iter 1782/1984 - loss 0.03800912 - time (sec): 82.61 - samples/sec: 1783.20 - lr: 0.000014 - momentum: 0.000000
2023-10-17 10:47:21,770 epoch 6 - iter 1980/1984 - loss 0.03804788 - time (sec): 91.58 - samples/sec: 1786.04 - lr: 0.000013 - momentum: 0.000000
2023-10-17 10:47:21,950 ----------------------------------------------------------------------------------------------------
2023-10-17 10:47:21,950 EPOCH 6 done: loss 0.0380 - lr: 0.000013
2023-10-17 10:47:25,502 DEV : loss 0.20233358442783356 - f1-score (micro avg) 0.7609
2023-10-17 10:47:25,529 ----------------------------------------------------------------------------------------------------
2023-10-17 10:47:35,922 epoch 7 - iter 198/1984 - loss 0.02769256 - time (sec): 10.39 - samples/sec: 1568.34 - lr: 0.000013 - momentum: 0.000000
2023-10-17 10:47:45,345 epoch 7 - iter 396/1984 - loss 0.02809458 - time (sec): 19.81 - samples/sec: 1656.25 - lr: 0.000013 - momentum: 0.000000
2023-10-17 10:47:54,754 epoch 7 - iter 594/1984 - loss 0.02925218 - time (sec): 29.22 - samples/sec: 1696.02 - lr: 0.000012 - momentum: 0.000000
2023-10-17 10:48:03,765 epoch 7 - iter 792/1984 - loss 0.02875332 - time (sec): 38.23 - samples/sec: 1717.27 - lr: 0.000012 - momentum: 0.000000
2023-10-17 10:48:12,807 epoch 7 - iter 990/1984 - loss 0.02726148 - time (sec): 47.28 - samples/sec: 1733.61 - lr: 0.000012 - momentum: 0.000000
2023-10-17 10:48:21,730 epoch 7 - iter 1188/1984 - loss 0.02657124 - time (sec): 56.20 - samples/sec: 1751.52 - lr: 0.000011 - momentum: 0.000000
2023-10-17 10:48:30,737 epoch 7 - iter 1386/1984 - loss 0.02702982 - time (sec): 65.21 - samples/sec: 1759.96 - lr: 0.000011 - momentum: 0.000000
2023-10-17 10:48:39,970 epoch 7 - iter 1584/1984 - loss 0.02650602 - time (sec): 74.44 - samples/sec: 1750.60 - lr: 0.000011 - momentum: 0.000000
2023-10-17 10:48:49,162 epoch 7 - iter 1782/1984 - loss 0.02844952 - time (sec): 83.63 - samples/sec: 1762.76 - lr: 0.000010 - momentum: 0.000000
2023-10-17 10:48:58,295 epoch 7 - iter 1980/1984 - loss 0.02825356 - time (sec): 92.76 - samples/sec: 1764.63 - lr: 0.000010 - momentum: 0.000000
2023-10-17 10:48:58,477 ----------------------------------------------------------------------------------------------------
2023-10-17 10:48:58,477 EPOCH 7 done: loss 0.0282 - lr: 0.000010
2023-10-17 10:49:01,991 DEV : loss 0.21014443039894104 - f1-score (micro avg) 0.7593
2023-10-17 10:49:02,014 ----------------------------------------------------------------------------------------------------
2023-10-17 10:49:10,679 epoch 8 - iter 198/1984 - loss 0.01310734 - time (sec): 8.66 - samples/sec: 1889.48 - lr: 0.000010 - momentum: 0.000000
2023-10-17 10:49:19,332 epoch 8 - iter 396/1984 - loss 0.01617564 - time (sec): 17.32 - samples/sec: 1870.85 - lr: 0.000009 - momentum: 0.000000
2023-10-17 10:49:28,226 epoch 8 - iter 594/1984 - loss 0.01587113 - time (sec): 26.21 - samples/sec: 1908.46 - lr: 0.000009 - momentum: 0.000000
2023-10-17 10:49:36,913 epoch 8 - iter 792/1984 - loss 0.01491232 - time (sec): 34.90 - samples/sec: 1893.99 - lr: 0.000009 - momentum: 0.000000
2023-10-17 10:49:45,623 epoch 8 - iter 990/1984 - loss 0.01558421 - time (sec): 43.61 - samples/sec: 1907.85 - lr: 0.000008 - momentum: 0.000000
2023-10-17 10:49:54,766 epoch 8 - iter 1188/1984 - loss 0.01606079 - time (sec): 52.75 - samples/sec: 1894.03 - lr: 0.000008 - momentum: 0.000000
2023-10-17 10:50:03,951 epoch 8 - iter 1386/1984 - loss 0.01694365 - time (sec): 61.94 - samples/sec: 1866.52 - lr: 0.000008 - momentum: 0.000000
2023-10-17 10:50:13,283 epoch 8 - iter 1584/1984 - loss 0.01809506 - time (sec): 71.27 - samples/sec: 1832.48 - lr: 0.000007 - momentum: 0.000000
2023-10-17 10:50:22,600 epoch 8 - iter 1782/1984 - loss 0.01857775 - time (sec): 80.58 - samples/sec: 1826.78 - lr: 0.000007 - momentum: 0.000000
2023-10-17 10:50:31,758 epoch 8 - iter 1980/1984 - loss 0.01906275 - time (sec): 89.74 - samples/sec: 1823.40 - lr: 0.000007 - momentum: 0.000000
2023-10-17 10:50:31,947 ----------------------------------------------------------------------------------------------------
2023-10-17 10:50:31,947 EPOCH 8 done: loss 0.0191 - lr: 0.000007
2023-10-17 10:50:35,465 DEV : loss 0.2338264435529709 - f1-score (micro avg) 0.7676
2023-10-17 10:50:35,489 ----------------------------------------------------------------------------------------------------
2023-10-17 10:50:44,512 epoch 9 - iter 198/1984 - loss 0.01453751 - time (sec): 9.02 - samples/sec: 1755.79 - lr: 0.000006 - momentum: 0.000000
2023-10-17 10:50:53,683 epoch 9 - iter 396/1984 - loss 0.01196746 - time (sec): 18.19 - samples/sec: 1825.23 - lr: 0.000006 - momentum: 0.000000
2023-10-17 10:51:03,022 epoch 9 - iter 594/1984 - loss 0.01160464 - time (sec): 27.53 - samples/sec: 1831.16 - lr: 0.000006 - momentum: 0.000000
2023-10-17 10:51:12,214 epoch 9 - iter 792/1984 - loss 0.01106266 - time (sec): 36.72 - samples/sec: 1815.83 - lr: 0.000005 - momentum: 0.000000
2023-10-17 10:51:21,261 epoch 9 - iter 990/1984 - loss 0.01089225 - time (sec): 45.77 - samples/sec: 1803.84 - lr: 0.000005 - momentum: 0.000000
2023-10-17 10:51:30,381 epoch 9 - iter 1188/1984 - loss 0.01111371 - time (sec): 54.89 - samples/sec: 1798.55 - lr: 0.000005 - momentum: 0.000000
2023-10-17 10:51:39,368 epoch 9 - iter 1386/1984 - loss 0.01130474 - time (sec): 63.88 - samples/sec: 1796.78 - lr: 0.000004 - momentum: 0.000000
2023-10-17 10:51:48,561 epoch 9 - iter 1584/1984 - loss 0.01217727 - time (sec): 73.07 - samples/sec: 1800.21 - lr: 0.000004 - momentum: 0.000000
2023-10-17 10:51:57,852 epoch 9 - iter 1782/1984 - loss 0.01235461 - time (sec): 82.36 - samples/sec: 1793.51 - lr: 0.000004 - momentum: 0.000000
2023-10-17 10:52:07,034 epoch 9 - iter 1980/1984 - loss 0.01325613 - time (sec): 91.54 - samples/sec: 1788.07 - lr: 0.000003 - momentum: 0.000000
2023-10-17 10:52:07,211 ----------------------------------------------------------------------------------------------------
2023-10-17 10:52:07,212 EPOCH 9 done: loss 0.0132 - lr: 0.000003
2023-10-17 10:52:10,636 DEV : loss 0.2385382354259491 - f1-score (micro avg) 0.772
2023-10-17 10:52:10,663 ----------------------------------------------------------------------------------------------------
2023-10-17 10:52:19,890 epoch 10 - iter 198/1984 - loss 0.00637340 - time (sec): 9.23 - samples/sec: 1789.65 - lr: 0.000003 - momentum: 0.000000
2023-10-17 10:52:28,935 epoch 10 - iter 396/1984 - loss 0.00651409 - time (sec): 18.27 - samples/sec: 1759.71 - lr: 0.000003 - momentum: 0.000000
2023-10-17 10:52:38,136 epoch 10 - iter 594/1984 - loss 0.00741839 - time (sec): 27.47 - samples/sec: 1769.92 - lr: 0.000002 - momentum: 0.000000
2023-10-17 10:52:47,372 epoch 10 - iter 792/1984 - loss 0.00875355 - time (sec): 36.71 - samples/sec: 1765.52 - lr: 0.000002 - momentum: 0.000000
2023-10-17 10:52:56,770 epoch 10 - iter 990/1984 - loss 0.00887012 - time (sec): 46.11 - samples/sec: 1747.58 - lr: 0.000002 - momentum: 0.000000
2023-10-17 10:53:06,020 epoch 10 - iter 1188/1984 - loss 0.00809607 - time (sec): 55.36 - samples/sec: 1760.41 - lr: 0.000001 - momentum: 0.000000
2023-10-17 10:53:15,196 epoch 10 - iter 1386/1984 - loss 0.00842195 - time (sec): 64.53 - samples/sec: 1776.32 - lr: 0.000001 - momentum: 0.000000
2023-10-17 10:53:24,107 epoch 10 - iter 1584/1984 - loss 0.00846139 - time (sec): 73.44 - samples/sec: 1784.87 - lr: 0.000001 - momentum: 0.000000
2023-10-17 10:53:32,767 epoch 10 - iter 1782/1984 - loss 0.00886546 - time (sec): 82.10 - samples/sec: 1789.89 - lr: 0.000000 - momentum: 0.000000
2023-10-17 10:53:41,465 epoch 10 - iter 1980/1984 - loss 0.00947778 - time (sec): 90.80 - samples/sec: 1802.94 - lr: 0.000000 - momentum: 0.000000
2023-10-17 10:53:41,637 ----------------------------------------------------------------------------------------------------
2023-10-17 10:53:41,637 EPOCH 10 done: loss 0.0095 - lr: 0.000000
2023-10-17 10:53:45,532 DEV : loss 0.24614956974983215 - f1-score (micro avg) 0.7678
2023-10-17 10:53:45,963 ----------------------------------------------------------------------------------------------------
2023-10-17 10:53:45,964 Loading model from best epoch ...
2023-10-17 10:53:47,827 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
2023-10-17 10:53:51,361
Results:
- F-score (micro) 0.7855
- F-score (macro) 0.7063
- Accuracy 0.6667
By class:
precision recall f1-score support
LOC 0.8135 0.8656 0.8388 655
PER 0.7333 0.7892 0.7603 223
ORG 0.5900 0.4646 0.5198 127
micro avg 0.7734 0.7980 0.7855 1005
macro avg 0.7123 0.7065 0.7063 1005
weighted avg 0.7675 0.7980 0.7810 1005
2023-10-17 10:53:51,362 ----------------------------------------------------------------------------------------------------
|