File size: 23,992 Bytes
6d28586 |
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 14:21:44,567 ----------------------------------------------------------------------------------------------------
2023-10-17 14:21:44,568 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 14:21:44,568 ----------------------------------------------------------------------------------------------------
2023-10-17 14:21:44,568 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 14:21:44,569 ----------------------------------------------------------------------------------------------------
2023-10-17 14:21:44,569 Train: 7936 sentences
2023-10-17 14:21:44,569 (train_with_dev=False, train_with_test=False)
2023-10-17 14:21:44,569 ----------------------------------------------------------------------------------------------------
2023-10-17 14:21:44,569 Training Params:
2023-10-17 14:21:44,569 - learning_rate: "3e-05"
2023-10-17 14:21:44,569 - mini_batch_size: "4"
2023-10-17 14:21:44,569 - max_epochs: "10"
2023-10-17 14:21:44,569 - shuffle: "True"
2023-10-17 14:21:44,569 ----------------------------------------------------------------------------------------------------
2023-10-17 14:21:44,569 Plugins:
2023-10-17 14:21:44,569 - TensorboardLogger
2023-10-17 14:21:44,569 - LinearScheduler | warmup_fraction: '0.1'
2023-10-17 14:21:44,569 ----------------------------------------------------------------------------------------------------
2023-10-17 14:21:44,569 Final evaluation on model from best epoch (best-model.pt)
2023-10-17 14:21:44,569 - metric: "('micro avg', 'f1-score')"
2023-10-17 14:21:44,569 ----------------------------------------------------------------------------------------------------
2023-10-17 14:21:44,569 Computation:
2023-10-17 14:21:44,569 - compute on device: cuda:0
2023-10-17 14:21:44,569 - embedding storage: none
2023-10-17 14:21:44,569 ----------------------------------------------------------------------------------------------------
2023-10-17 14:21:44,569 Model training base path: "hmbench-icdar/fr-hmteams/teams-base-historic-multilingual-discriminator-bs4-wsFalse-e10-lr3e-05-poolingfirst-layers-1-crfFalse-5"
2023-10-17 14:21:44,569 ----------------------------------------------------------------------------------------------------
2023-10-17 14:21:44,569 ----------------------------------------------------------------------------------------------------
2023-10-17 14:21:44,569 Logging anything other than scalars to TensorBoard is currently not supported.
2023-10-17 14:21:53,339 epoch 1 - iter 198/1984 - loss 2.35272971 - time (sec): 8.77 - samples/sec: 1839.29 - lr: 0.000003 - momentum: 0.000000
2023-10-17 14:22:02,373 epoch 1 - iter 396/1984 - loss 1.35226237 - time (sec): 17.80 - samples/sec: 1885.07 - lr: 0.000006 - momentum: 0.000000
2023-10-17 14:22:11,529 epoch 1 - iter 594/1984 - loss 0.99696504 - time (sec): 26.96 - samples/sec: 1827.49 - lr: 0.000009 - momentum: 0.000000
2023-10-17 14:22:20,576 epoch 1 - iter 792/1984 - loss 0.80282302 - time (sec): 36.01 - samples/sec: 1811.71 - lr: 0.000012 - momentum: 0.000000
2023-10-17 14:22:29,709 epoch 1 - iter 990/1984 - loss 0.68889213 - time (sec): 45.14 - samples/sec: 1785.58 - lr: 0.000015 - momentum: 0.000000
2023-10-17 14:22:38,778 epoch 1 - iter 1188/1984 - loss 0.59950974 - time (sec): 54.21 - samples/sec: 1788.54 - lr: 0.000018 - momentum: 0.000000
2023-10-17 14:22:48,278 epoch 1 - iter 1386/1984 - loss 0.53152483 - time (sec): 63.71 - samples/sec: 1787.61 - lr: 0.000021 - momentum: 0.000000
2023-10-17 14:22:57,520 epoch 1 - iter 1584/1984 - loss 0.48053194 - time (sec): 72.95 - samples/sec: 1789.09 - lr: 0.000024 - momentum: 0.000000
2023-10-17 14:23:06,591 epoch 1 - iter 1782/1984 - loss 0.44152631 - time (sec): 82.02 - samples/sec: 1792.58 - lr: 0.000027 - momentum: 0.000000
2023-10-17 14:23:16,377 epoch 1 - iter 1980/1984 - loss 0.40853255 - time (sec): 91.81 - samples/sec: 1782.55 - lr: 0.000030 - momentum: 0.000000
2023-10-17 14:23:16,556 ----------------------------------------------------------------------------------------------------
2023-10-17 14:23:16,557 EPOCH 1 done: loss 0.4079 - lr: 0.000030
2023-10-17 14:23:19,804 DEV : loss 0.09482365101575851 - f1-score (micro avg) 0.7296
2023-10-17 14:23:19,827 saving best model
2023-10-17 14:23:20,267 ----------------------------------------------------------------------------------------------------
2023-10-17 14:23:29,741 epoch 2 - iter 198/1984 - loss 0.10656840 - time (sec): 9.47 - samples/sec: 1799.02 - lr: 0.000030 - momentum: 0.000000
2023-10-17 14:23:38,991 epoch 2 - iter 396/1984 - loss 0.10818698 - time (sec): 18.72 - samples/sec: 1785.10 - lr: 0.000029 - momentum: 0.000000
2023-10-17 14:23:48,043 epoch 2 - iter 594/1984 - loss 0.11021072 - time (sec): 27.77 - samples/sec: 1791.14 - lr: 0.000029 - momentum: 0.000000
2023-10-17 14:23:57,001 epoch 2 - iter 792/1984 - loss 0.11163153 - time (sec): 36.73 - samples/sec: 1776.91 - lr: 0.000029 - momentum: 0.000000
2023-10-17 14:24:06,134 epoch 2 - iter 990/1984 - loss 0.11183860 - time (sec): 45.87 - samples/sec: 1789.70 - lr: 0.000028 - momentum: 0.000000
2023-10-17 14:24:15,206 epoch 2 - iter 1188/1984 - loss 0.11055508 - time (sec): 54.94 - samples/sec: 1783.11 - lr: 0.000028 - momentum: 0.000000
2023-10-17 14:24:24,164 epoch 2 - iter 1386/1984 - loss 0.11239117 - time (sec): 63.90 - samples/sec: 1783.76 - lr: 0.000028 - momentum: 0.000000
2023-10-17 14:24:33,227 epoch 2 - iter 1584/1984 - loss 0.11240077 - time (sec): 72.96 - samples/sec: 1784.23 - lr: 0.000027 - momentum: 0.000000
2023-10-17 14:24:42,472 epoch 2 - iter 1782/1984 - loss 0.11194093 - time (sec): 82.20 - samples/sec: 1791.13 - lr: 0.000027 - momentum: 0.000000
2023-10-17 14:24:51,688 epoch 2 - iter 1980/1984 - loss 0.11315701 - time (sec): 91.42 - samples/sec: 1790.60 - lr: 0.000027 - momentum: 0.000000
2023-10-17 14:24:51,873 ----------------------------------------------------------------------------------------------------
2023-10-17 14:24:51,873 EPOCH 2 done: loss 0.1131 - lr: 0.000027
2023-10-17 14:24:55,803 DEV : loss 0.09371839463710785 - f1-score (micro avg) 0.7416
2023-10-17 14:24:55,825 saving best model
2023-10-17 14:24:56,405 ----------------------------------------------------------------------------------------------------
2023-10-17 14:25:05,416 epoch 3 - iter 198/1984 - loss 0.08688426 - time (sec): 9.00 - samples/sec: 1688.59 - lr: 0.000026 - momentum: 0.000000
2023-10-17 14:25:14,414 epoch 3 - iter 396/1984 - loss 0.08535351 - time (sec): 18.00 - samples/sec: 1749.88 - lr: 0.000026 - momentum: 0.000000
2023-10-17 14:25:23,215 epoch 3 - iter 594/1984 - loss 0.08392751 - time (sec): 26.80 - samples/sec: 1804.77 - lr: 0.000026 - momentum: 0.000000
2023-10-17 14:25:31,824 epoch 3 - iter 792/1984 - loss 0.08103732 - time (sec): 35.41 - samples/sec: 1840.55 - lr: 0.000025 - momentum: 0.000000
2023-10-17 14:25:40,506 epoch 3 - iter 990/1984 - loss 0.08337144 - time (sec): 44.09 - samples/sec: 1849.78 - lr: 0.000025 - momentum: 0.000000
2023-10-17 14:25:49,403 epoch 3 - iter 1188/1984 - loss 0.08370851 - time (sec): 52.99 - samples/sec: 1861.28 - lr: 0.000025 - momentum: 0.000000
2023-10-17 14:25:58,429 epoch 3 - iter 1386/1984 - loss 0.08381845 - time (sec): 62.01 - samples/sec: 1844.58 - lr: 0.000024 - momentum: 0.000000
2023-10-17 14:26:07,378 epoch 3 - iter 1584/1984 - loss 0.08471920 - time (sec): 70.96 - samples/sec: 1839.09 - lr: 0.000024 - momentum: 0.000000
2023-10-17 14:26:16,525 epoch 3 - iter 1782/1984 - loss 0.08345150 - time (sec): 80.11 - samples/sec: 1841.82 - lr: 0.000024 - momentum: 0.000000
2023-10-17 14:26:25,615 epoch 3 - iter 1980/1984 - loss 0.08467558 - time (sec): 89.20 - samples/sec: 1834.64 - lr: 0.000023 - momentum: 0.000000
2023-10-17 14:26:25,813 ----------------------------------------------------------------------------------------------------
2023-10-17 14:26:25,813 EPOCH 3 done: loss 0.0846 - lr: 0.000023
2023-10-17 14:26:29,205 DEV : loss 0.12102336436510086 - f1-score (micro avg) 0.7527
2023-10-17 14:26:29,226 saving best model
2023-10-17 14:26:29,740 ----------------------------------------------------------------------------------------------------
2023-10-17 14:26:38,903 epoch 4 - iter 198/1984 - loss 0.06092971 - time (sec): 9.16 - samples/sec: 1722.01 - lr: 0.000023 - momentum: 0.000000
2023-10-17 14:26:48,164 epoch 4 - iter 396/1984 - loss 0.06579458 - time (sec): 18.42 - samples/sec: 1768.71 - lr: 0.000023 - momentum: 0.000000
2023-10-17 14:26:57,322 epoch 4 - iter 594/1984 - loss 0.06716833 - time (sec): 27.58 - samples/sec: 1757.12 - lr: 0.000022 - momentum: 0.000000
2023-10-17 14:27:06,588 epoch 4 - iter 792/1984 - loss 0.06375569 - time (sec): 36.84 - samples/sec: 1767.60 - lr: 0.000022 - momentum: 0.000000
2023-10-17 14:27:15,756 epoch 4 - iter 990/1984 - loss 0.06406550 - time (sec): 46.01 - samples/sec: 1774.90 - lr: 0.000022 - momentum: 0.000000
2023-10-17 14:27:24,732 epoch 4 - iter 1188/1984 - loss 0.06699263 - time (sec): 54.99 - samples/sec: 1778.23 - lr: 0.000021 - momentum: 0.000000
2023-10-17 14:27:33,952 epoch 4 - iter 1386/1984 - loss 0.06689916 - time (sec): 64.21 - samples/sec: 1779.83 - lr: 0.000021 - momentum: 0.000000
2023-10-17 14:27:42,808 epoch 4 - iter 1584/1984 - loss 0.06814155 - time (sec): 73.06 - samples/sec: 1785.97 - lr: 0.000021 - momentum: 0.000000
2023-10-17 14:27:51,811 epoch 4 - iter 1782/1984 - loss 0.06896506 - time (sec): 82.07 - samples/sec: 1795.00 - lr: 0.000020 - momentum: 0.000000
2023-10-17 14:28:00,740 epoch 4 - iter 1980/1984 - loss 0.06896094 - time (sec): 91.00 - samples/sec: 1798.71 - lr: 0.000020 - momentum: 0.000000
2023-10-17 14:28:00,926 ----------------------------------------------------------------------------------------------------
2023-10-17 14:28:00,926 EPOCH 4 done: loss 0.0691 - lr: 0.000020
2023-10-17 14:28:04,626 DEV : loss 0.1571902483701706 - f1-score (micro avg) 0.7542
2023-10-17 14:28:04,649 saving best model
2023-10-17 14:28:05,140 ----------------------------------------------------------------------------------------------------
2023-10-17 14:28:14,051 epoch 5 - iter 198/1984 - loss 0.04531813 - time (sec): 8.91 - samples/sec: 1796.32 - lr: 0.000020 - momentum: 0.000000
2023-10-17 14:28:23,211 epoch 5 - iter 396/1984 - loss 0.04647017 - time (sec): 18.07 - samples/sec: 1815.98 - lr: 0.000019 - momentum: 0.000000
2023-10-17 14:28:32,522 epoch 5 - iter 594/1984 - loss 0.04475855 - time (sec): 27.38 - samples/sec: 1830.73 - lr: 0.000019 - momentum: 0.000000
2023-10-17 14:28:41,587 epoch 5 - iter 792/1984 - loss 0.04669245 - time (sec): 36.45 - samples/sec: 1818.97 - lr: 0.000019 - momentum: 0.000000
2023-10-17 14:28:50,763 epoch 5 - iter 990/1984 - loss 0.04912804 - time (sec): 45.62 - samples/sec: 1834.17 - lr: 0.000018 - momentum: 0.000000
2023-10-17 14:28:59,777 epoch 5 - iter 1188/1984 - loss 0.04867525 - time (sec): 54.64 - samples/sec: 1821.02 - lr: 0.000018 - momentum: 0.000000
2023-10-17 14:29:09,023 epoch 5 - iter 1386/1984 - loss 0.04927445 - time (sec): 63.88 - samples/sec: 1821.15 - lr: 0.000018 - momentum: 0.000000
2023-10-17 14:29:18,033 epoch 5 - iter 1584/1984 - loss 0.04954311 - time (sec): 72.89 - samples/sec: 1809.93 - lr: 0.000017 - momentum: 0.000000
2023-10-17 14:29:27,001 epoch 5 - iter 1782/1984 - loss 0.04932246 - time (sec): 81.86 - samples/sec: 1803.75 - lr: 0.000017 - momentum: 0.000000
2023-10-17 14:29:36,222 epoch 5 - iter 1980/1984 - loss 0.04937008 - time (sec): 91.08 - samples/sec: 1796.77 - lr: 0.000017 - momentum: 0.000000
2023-10-17 14:29:36,399 ----------------------------------------------------------------------------------------------------
2023-10-17 14:29:36,399 EPOCH 5 done: loss 0.0493 - lr: 0.000017
2023-10-17 14:29:39,976 DEV : loss 0.16390633583068848 - f1-score (micro avg) 0.7745
2023-10-17 14:29:40,003 saving best model
2023-10-17 14:29:40,607 ----------------------------------------------------------------------------------------------------
2023-10-17 14:29:49,628 epoch 6 - iter 198/1984 - loss 0.03464468 - time (sec): 9.02 - samples/sec: 1836.48 - lr: 0.000016 - momentum: 0.000000
2023-10-17 14:29:58,781 epoch 6 - iter 396/1984 - loss 0.03116501 - time (sec): 18.17 - samples/sec: 1841.59 - lr: 0.000016 - momentum: 0.000000
2023-10-17 14:30:07,928 epoch 6 - iter 594/1984 - loss 0.03152775 - time (sec): 27.32 - samples/sec: 1836.49 - lr: 0.000016 - momentum: 0.000000
2023-10-17 14:30:17,082 epoch 6 - iter 792/1984 - loss 0.03071046 - time (sec): 36.47 - samples/sec: 1832.41 - lr: 0.000015 - momentum: 0.000000
2023-10-17 14:30:26,291 epoch 6 - iter 990/1984 - loss 0.03185003 - time (sec): 45.68 - samples/sec: 1820.68 - lr: 0.000015 - momentum: 0.000000
2023-10-17 14:30:35,672 epoch 6 - iter 1188/1984 - loss 0.03355453 - time (sec): 55.06 - samples/sec: 1823.91 - lr: 0.000015 - momentum: 0.000000
2023-10-17 14:30:44,728 epoch 6 - iter 1386/1984 - loss 0.03464690 - time (sec): 64.12 - samples/sec: 1807.92 - lr: 0.000014 - momentum: 0.000000
2023-10-17 14:30:53,876 epoch 6 - iter 1584/1984 - loss 0.03696921 - time (sec): 73.27 - samples/sec: 1795.05 - lr: 0.000014 - momentum: 0.000000
2023-10-17 14:31:02,724 epoch 6 - iter 1782/1984 - loss 0.03702742 - time (sec): 82.11 - samples/sec: 1799.57 - lr: 0.000014 - momentum: 0.000000
2023-10-17 14:31:11,783 epoch 6 - iter 1980/1984 - loss 0.03663836 - time (sec): 91.17 - samples/sec: 1795.02 - lr: 0.000013 - momentum: 0.000000
2023-10-17 14:31:11,968 ----------------------------------------------------------------------------------------------------
2023-10-17 14:31:11,968 EPOCH 6 done: loss 0.0367 - lr: 0.000013
2023-10-17 14:31:15,510 DEV : loss 0.19935466349124908 - f1-score (micro avg) 0.7647
2023-10-17 14:31:15,540 ----------------------------------------------------------------------------------------------------
2023-10-17 14:31:26,185 epoch 7 - iter 198/1984 - loss 0.02511532 - time (sec): 10.64 - samples/sec: 1492.77 - lr: 0.000013 - momentum: 0.000000
2023-10-17 14:31:36,808 epoch 7 - iter 396/1984 - loss 0.02353593 - time (sec): 21.27 - samples/sec: 1529.14 - lr: 0.000013 - momentum: 0.000000
2023-10-17 14:31:46,481 epoch 7 - iter 594/1984 - loss 0.02725110 - time (sec): 30.94 - samples/sec: 1579.91 - lr: 0.000012 - momentum: 0.000000
2023-10-17 14:31:55,734 epoch 7 - iter 792/1984 - loss 0.02742823 - time (sec): 40.19 - samples/sec: 1651.96 - lr: 0.000012 - momentum: 0.000000
2023-10-17 14:32:05,005 epoch 7 - iter 990/1984 - loss 0.02499010 - time (sec): 49.46 - samples/sec: 1704.82 - lr: 0.000012 - momentum: 0.000000
2023-10-17 14:32:14,065 epoch 7 - iter 1188/1984 - loss 0.02517202 - time (sec): 58.52 - samples/sec: 1716.61 - lr: 0.000011 - momentum: 0.000000
2023-10-17 14:32:22,809 epoch 7 - iter 1386/1984 - loss 0.02492481 - time (sec): 67.27 - samples/sec: 1727.30 - lr: 0.000011 - momentum: 0.000000
2023-10-17 14:32:31,645 epoch 7 - iter 1584/1984 - loss 0.02477025 - time (sec): 76.10 - samples/sec: 1737.03 - lr: 0.000011 - momentum: 0.000000
2023-10-17 14:32:40,907 epoch 7 - iter 1782/1984 - loss 0.02572704 - time (sec): 85.37 - samples/sec: 1736.07 - lr: 0.000010 - momentum: 0.000000
2023-10-17 14:32:49,995 epoch 7 - iter 1980/1984 - loss 0.02491306 - time (sec): 94.45 - samples/sec: 1732.11 - lr: 0.000010 - momentum: 0.000000
2023-10-17 14:32:50,174 ----------------------------------------------------------------------------------------------------
2023-10-17 14:32:50,174 EPOCH 7 done: loss 0.0250 - lr: 0.000010
2023-10-17 14:32:53,683 DEV : loss 0.20214584469795227 - f1-score (micro avg) 0.7623
2023-10-17 14:32:53,716 ----------------------------------------------------------------------------------------------------
2023-10-17 14:33:02,786 epoch 8 - iter 198/1984 - loss 0.01165556 - time (sec): 9.07 - samples/sec: 1793.90 - lr: 0.000010 - momentum: 0.000000
2023-10-17 14:33:11,842 epoch 8 - iter 396/1984 - loss 0.01479537 - time (sec): 18.13 - samples/sec: 1782.05 - lr: 0.000009 - momentum: 0.000000
2023-10-17 14:33:20,675 epoch 8 - iter 594/1984 - loss 0.01696006 - time (sec): 26.96 - samples/sec: 1809.49 - lr: 0.000009 - momentum: 0.000000
2023-10-17 14:33:29,808 epoch 8 - iter 792/1984 - loss 0.01648117 - time (sec): 36.09 - samples/sec: 1809.14 - lr: 0.000009 - momentum: 0.000000
2023-10-17 14:33:38,982 epoch 8 - iter 990/1984 - loss 0.01744306 - time (sec): 45.27 - samples/sec: 1798.74 - lr: 0.000008 - momentum: 0.000000
2023-10-17 14:33:48,084 epoch 8 - iter 1188/1984 - loss 0.01772065 - time (sec): 54.37 - samples/sec: 1786.89 - lr: 0.000008 - momentum: 0.000000
2023-10-17 14:33:57,336 epoch 8 - iter 1386/1984 - loss 0.01868319 - time (sec): 63.62 - samples/sec: 1786.50 - lr: 0.000008 - momentum: 0.000000
2023-10-17 14:34:06,576 epoch 8 - iter 1584/1984 - loss 0.01859987 - time (sec): 72.86 - samples/sec: 1790.02 - lr: 0.000007 - momentum: 0.000000
2023-10-17 14:34:17,150 epoch 8 - iter 1782/1984 - loss 0.01846760 - time (sec): 83.43 - samples/sec: 1753.15 - lr: 0.000007 - momentum: 0.000000
2023-10-17 14:34:27,610 epoch 8 - iter 1980/1984 - loss 0.01868024 - time (sec): 93.89 - samples/sec: 1742.66 - lr: 0.000007 - momentum: 0.000000
2023-10-17 14:34:27,831 ----------------------------------------------------------------------------------------------------
2023-10-17 14:34:27,831 EPOCH 8 done: loss 0.0187 - lr: 0.000007
2023-10-17 14:34:31,286 DEV : loss 0.22779549658298492 - f1-score (micro avg) 0.7628
2023-10-17 14:34:31,310 ----------------------------------------------------------------------------------------------------
2023-10-17 14:34:41,901 epoch 9 - iter 198/1984 - loss 0.01906451 - time (sec): 10.59 - samples/sec: 1589.50 - lr: 0.000006 - momentum: 0.000000
2023-10-17 14:34:51,675 epoch 9 - iter 396/1984 - loss 0.01555467 - time (sec): 20.36 - samples/sec: 1658.32 - lr: 0.000006 - momentum: 0.000000
2023-10-17 14:35:01,204 epoch 9 - iter 594/1984 - loss 0.01521979 - time (sec): 29.89 - samples/sec: 1678.39 - lr: 0.000006 - momentum: 0.000000
2023-10-17 14:35:10,391 epoch 9 - iter 792/1984 - loss 0.01491455 - time (sec): 39.08 - samples/sec: 1697.20 - lr: 0.000005 - momentum: 0.000000
2023-10-17 14:35:19,561 epoch 9 - iter 990/1984 - loss 0.01387345 - time (sec): 48.25 - samples/sec: 1725.91 - lr: 0.000005 - momentum: 0.000000
2023-10-17 14:35:28,646 epoch 9 - iter 1188/1984 - loss 0.01431168 - time (sec): 57.33 - samples/sec: 1728.81 - lr: 0.000005 - momentum: 0.000000
2023-10-17 14:35:37,776 epoch 9 - iter 1386/1984 - loss 0.01452248 - time (sec): 66.46 - samples/sec: 1739.53 - lr: 0.000004 - momentum: 0.000000
2023-10-17 14:35:46,951 epoch 9 - iter 1584/1984 - loss 0.01440257 - time (sec): 75.64 - samples/sec: 1742.47 - lr: 0.000004 - momentum: 0.000000
2023-10-17 14:35:56,334 epoch 9 - iter 1782/1984 - loss 0.01404208 - time (sec): 85.02 - samples/sec: 1738.26 - lr: 0.000004 - momentum: 0.000000
2023-10-17 14:36:05,792 epoch 9 - iter 1980/1984 - loss 0.01392469 - time (sec): 94.48 - samples/sec: 1732.59 - lr: 0.000003 - momentum: 0.000000
2023-10-17 14:36:05,980 ----------------------------------------------------------------------------------------------------
2023-10-17 14:36:05,980 EPOCH 9 done: loss 0.0140 - lr: 0.000003
2023-10-17 14:36:09,397 DEV : loss 0.23514559864997864 - f1-score (micro avg) 0.7616
2023-10-17 14:36:09,418 ----------------------------------------------------------------------------------------------------
2023-10-17 14:36:18,647 epoch 10 - iter 198/1984 - loss 0.00949880 - time (sec): 9.23 - samples/sec: 1832.95 - lr: 0.000003 - momentum: 0.000000
2023-10-17 14:36:27,628 epoch 10 - iter 396/1984 - loss 0.00775092 - time (sec): 18.21 - samples/sec: 1832.79 - lr: 0.000003 - momentum: 0.000000
2023-10-17 14:36:36,789 epoch 10 - iter 594/1984 - loss 0.00839411 - time (sec): 27.37 - samples/sec: 1833.08 - lr: 0.000002 - momentum: 0.000000
2023-10-17 14:36:45,901 epoch 10 - iter 792/1984 - loss 0.00870864 - time (sec): 36.48 - samples/sec: 1796.51 - lr: 0.000002 - momentum: 0.000000
2023-10-17 14:36:55,161 epoch 10 - iter 990/1984 - loss 0.00887633 - time (sec): 45.74 - samples/sec: 1804.60 - lr: 0.000002 - momentum: 0.000000
2023-10-17 14:37:04,293 epoch 10 - iter 1188/1984 - loss 0.00855463 - time (sec): 54.87 - samples/sec: 1791.29 - lr: 0.000001 - momentum: 0.000000
2023-10-17 14:37:13,457 epoch 10 - iter 1386/1984 - loss 0.00875616 - time (sec): 64.04 - samples/sec: 1789.47 - lr: 0.000001 - momentum: 0.000000
2023-10-17 14:37:22,534 epoch 10 - iter 1584/1984 - loss 0.00856302 - time (sec): 73.11 - samples/sec: 1784.41 - lr: 0.000001 - momentum: 0.000000
2023-10-17 14:37:31,887 epoch 10 - iter 1782/1984 - loss 0.00882088 - time (sec): 82.47 - samples/sec: 1770.65 - lr: 0.000000 - momentum: 0.000000
2023-10-17 14:37:41,392 epoch 10 - iter 1980/1984 - loss 0.00856634 - time (sec): 91.97 - samples/sec: 1780.44 - lr: 0.000000 - momentum: 0.000000
2023-10-17 14:37:41,566 ----------------------------------------------------------------------------------------------------
2023-10-17 14:37:41,566 EPOCH 10 done: loss 0.0086 - lr: 0.000000
2023-10-17 14:37:44,996 DEV : loss 0.23482687771320343 - f1-score (micro avg) 0.7613
2023-10-17 14:37:45,439 ----------------------------------------------------------------------------------------------------
2023-10-17 14:37:45,440 Loading model from best epoch ...
2023-10-17 14:37:48,035 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 14:37:50,922
Results:
- F-score (micro) 0.8006
- F-score (macro) 0.7098
- Accuracy 0.6849
By class:
precision recall f1-score support
LOC 0.8376 0.8901 0.8631 655
PER 0.7311 0.7803 0.7549 223
ORG 0.6087 0.4409 0.5114 127
micro avg 0.7924 0.8090 0.8006 1005
macro avg 0.7258 0.7038 0.7098 1005
weighted avg 0.7851 0.8090 0.7946 1005
2023-10-17 14:37:50,922 ----------------------------------------------------------------------------------------------------
|