File size: 24,229 Bytes
8cc7f3b |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 |
2023-10-18 20:40:46,083 ----------------------------------------------------------------------------------------------------
2023-10-18 20:40:46,083 Model: "SequenceTagger(
(embeddings): TransformerWordEmbeddings(
(model): BertModel(
(embeddings): BertEmbeddings(
(word_embeddings): Embedding(32001, 128)
(position_embeddings): Embedding(512, 128)
(token_type_embeddings): Embedding(2, 128)
(LayerNorm): LayerNorm((128,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(encoder): BertEncoder(
(layer): ModuleList(
(0-1): 2 x BertLayer(
(attention): BertAttention(
(self): BertSelfAttention(
(query): Linear(in_features=128, out_features=128, bias=True)
(key): Linear(in_features=128, out_features=128, bias=True)
(value): Linear(in_features=128, out_features=128, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(output): BertSelfOutput(
(dense): Linear(in_features=128, out_features=128, bias=True)
(LayerNorm): LayerNorm((128,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(intermediate): BertIntermediate(
(dense): Linear(in_features=128, out_features=512, bias=True)
(intermediate_act_fn): GELUActivation()
)
(output): BertOutput(
(dense): Linear(in_features=512, out_features=128, bias=True)
(LayerNorm): LayerNorm((128,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
)
)
(pooler): BertPooler(
(dense): Linear(in_features=128, out_features=128, bias=True)
(activation): Tanh()
)
)
)
(locked_dropout): LockedDropout(p=0.5)
(linear): Linear(in_features=128, out_features=13, bias=True)
(loss_function): CrossEntropyLoss()
)"
2023-10-18 20:40:46,083 ----------------------------------------------------------------------------------------------------
2023-10-18 20:40:46,083 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
2023-10-18 20:40:46,083 ----------------------------------------------------------------------------------------------------
2023-10-18 20:40:46,083 Train: 7936 sentences
2023-10-18 20:40:46,083 (train_with_dev=False, train_with_test=False)
2023-10-18 20:40:46,083 ----------------------------------------------------------------------------------------------------
2023-10-18 20:40:46,083 Training Params:
2023-10-18 20:40:46,083 - learning_rate: "3e-05"
2023-10-18 20:40:46,084 - mini_batch_size: "4"
2023-10-18 20:40:46,084 - max_epochs: "10"
2023-10-18 20:40:46,084 - shuffle: "True"
2023-10-18 20:40:46,084 ----------------------------------------------------------------------------------------------------
2023-10-18 20:40:46,084 Plugins:
2023-10-18 20:40:46,084 - TensorboardLogger
2023-10-18 20:40:46,084 - LinearScheduler | warmup_fraction: '0.1'
2023-10-18 20:40:46,084 ----------------------------------------------------------------------------------------------------
2023-10-18 20:40:46,084 Final evaluation on model from best epoch (best-model.pt)
2023-10-18 20:40:46,084 - metric: "('micro avg', 'f1-score')"
2023-10-18 20:40:46,084 ----------------------------------------------------------------------------------------------------
2023-10-18 20:40:46,084 Computation:
2023-10-18 20:40:46,084 - compute on device: cuda:0
2023-10-18 20:40:46,084 - embedding storage: none
2023-10-18 20:40:46,084 ----------------------------------------------------------------------------------------------------
2023-10-18 20:40:46,084 Model training base path: "hmbench-icdar/fr-dbmdz/bert-tiny-historic-multilingual-cased-bs4-wsFalse-e10-lr3e-05-poolingfirst-layers-1-crfFalse-2"
2023-10-18 20:40:46,084 ----------------------------------------------------------------------------------------------------
2023-10-18 20:40:46,084 ----------------------------------------------------------------------------------------------------
2023-10-18 20:40:46,084 Logging anything other than scalars to TensorBoard is currently not supported.
2023-10-18 20:40:49,484 epoch 1 - iter 198/1984 - loss 2.40181665 - time (sec): 3.40 - samples/sec: 4847.36 - lr: 0.000003 - momentum: 0.000000
2023-10-18 20:40:52,929 epoch 1 - iter 396/1984 - loss 2.06051017 - time (sec): 6.85 - samples/sec: 4779.10 - lr: 0.000006 - momentum: 0.000000
2023-10-18 20:40:56,097 epoch 1 - iter 594/1984 - loss 1.65005864 - time (sec): 10.01 - samples/sec: 4974.57 - lr: 0.000009 - momentum: 0.000000
2023-10-18 20:40:59,185 epoch 1 - iter 792/1984 - loss 1.36396891 - time (sec): 13.10 - samples/sec: 5050.84 - lr: 0.000012 - momentum: 0.000000
2023-10-18 20:41:02,224 epoch 1 - iter 990/1984 - loss 1.19110595 - time (sec): 16.14 - samples/sec: 5155.96 - lr: 0.000015 - momentum: 0.000000
2023-10-18 20:41:05,316 epoch 1 - iter 1188/1984 - loss 1.06964597 - time (sec): 19.23 - samples/sec: 5168.26 - lr: 0.000018 - momentum: 0.000000
2023-10-18 20:41:08,345 epoch 1 - iter 1386/1984 - loss 0.97250875 - time (sec): 22.26 - samples/sec: 5200.44 - lr: 0.000021 - momentum: 0.000000
2023-10-18 20:41:11,364 epoch 1 - iter 1584/1984 - loss 0.89537197 - time (sec): 25.28 - samples/sec: 5207.18 - lr: 0.000024 - momentum: 0.000000
2023-10-18 20:41:14,412 epoch 1 - iter 1782/1984 - loss 0.83846567 - time (sec): 28.33 - samples/sec: 5199.51 - lr: 0.000027 - momentum: 0.000000
2023-10-18 20:41:17,456 epoch 1 - iter 1980/1984 - loss 0.79010219 - time (sec): 31.37 - samples/sec: 5217.57 - lr: 0.000030 - momentum: 0.000000
2023-10-18 20:41:17,514 ----------------------------------------------------------------------------------------------------
2023-10-18 20:41:17,514 EPOCH 1 done: loss 0.7888 - lr: 0.000030
2023-10-18 20:41:19,039 DEV : loss 0.22708812355995178 - f1-score (micro avg) 0.2642
2023-10-18 20:41:19,058 saving best model
2023-10-18 20:41:19,093 ----------------------------------------------------------------------------------------------------
2023-10-18 20:41:22,126 epoch 2 - iter 198/1984 - loss 0.32835931 - time (sec): 3.03 - samples/sec: 5320.46 - lr: 0.000030 - momentum: 0.000000
2023-10-18 20:41:25,167 epoch 2 - iter 396/1984 - loss 0.29962721 - time (sec): 6.07 - samples/sec: 5482.59 - lr: 0.000029 - momentum: 0.000000
2023-10-18 20:41:28,151 epoch 2 - iter 594/1984 - loss 0.29692658 - time (sec): 9.06 - samples/sec: 5412.42 - lr: 0.000029 - momentum: 0.000000
2023-10-18 20:41:31,207 epoch 2 - iter 792/1984 - loss 0.29657996 - time (sec): 12.11 - samples/sec: 5383.77 - lr: 0.000029 - momentum: 0.000000
2023-10-18 20:41:34,225 epoch 2 - iter 990/1984 - loss 0.28962444 - time (sec): 15.13 - samples/sec: 5372.89 - lr: 0.000028 - momentum: 0.000000
2023-10-18 20:41:37,219 epoch 2 - iter 1188/1984 - loss 0.28843578 - time (sec): 18.13 - samples/sec: 5362.62 - lr: 0.000028 - momentum: 0.000000
2023-10-18 20:41:40,292 epoch 2 - iter 1386/1984 - loss 0.29008589 - time (sec): 21.20 - samples/sec: 5333.12 - lr: 0.000028 - momentum: 0.000000
2023-10-18 20:41:43,095 epoch 2 - iter 1584/1984 - loss 0.28766926 - time (sec): 24.00 - samples/sec: 5415.25 - lr: 0.000027 - momentum: 0.000000
2023-10-18 20:41:46,153 epoch 2 - iter 1782/1984 - loss 0.28669108 - time (sec): 27.06 - samples/sec: 5436.91 - lr: 0.000027 - momentum: 0.000000
2023-10-18 20:41:49,222 epoch 2 - iter 1980/1984 - loss 0.28366818 - time (sec): 30.13 - samples/sec: 5432.08 - lr: 0.000027 - momentum: 0.000000
2023-10-18 20:41:49,279 ----------------------------------------------------------------------------------------------------
2023-10-18 20:41:49,279 EPOCH 2 done: loss 0.2838 - lr: 0.000027
2023-10-18 20:41:51,481 DEV : loss 0.19872356951236725 - f1-score (micro avg) 0.3621
2023-10-18 20:41:51,500 saving best model
2023-10-18 20:41:51,532 ----------------------------------------------------------------------------------------------------
2023-10-18 20:41:54,500 epoch 3 - iter 198/1984 - loss 0.24485053 - time (sec): 2.97 - samples/sec: 5641.36 - lr: 0.000026 - momentum: 0.000000
2023-10-18 20:41:57,549 epoch 3 - iter 396/1984 - loss 0.25674337 - time (sec): 6.02 - samples/sec: 5432.56 - lr: 0.000026 - momentum: 0.000000
2023-10-18 20:42:00,576 epoch 3 - iter 594/1984 - loss 0.25987147 - time (sec): 9.04 - samples/sec: 5408.81 - lr: 0.000026 - momentum: 0.000000
2023-10-18 20:42:03,632 epoch 3 - iter 792/1984 - loss 0.25653396 - time (sec): 12.10 - samples/sec: 5357.86 - lr: 0.000025 - momentum: 0.000000
2023-10-18 20:42:06,666 epoch 3 - iter 990/1984 - loss 0.24885921 - time (sec): 15.13 - samples/sec: 5396.16 - lr: 0.000025 - momentum: 0.000000
2023-10-18 20:42:09,707 epoch 3 - iter 1188/1984 - loss 0.25191638 - time (sec): 18.17 - samples/sec: 5341.92 - lr: 0.000025 - momentum: 0.000000
2023-10-18 20:42:12,610 epoch 3 - iter 1386/1984 - loss 0.25739981 - time (sec): 21.08 - samples/sec: 5388.28 - lr: 0.000024 - momentum: 0.000000
2023-10-18 20:42:15,681 epoch 3 - iter 1584/1984 - loss 0.25300334 - time (sec): 24.15 - samples/sec: 5398.44 - lr: 0.000024 - momentum: 0.000000
2023-10-18 20:42:18,737 epoch 3 - iter 1782/1984 - loss 0.24809037 - time (sec): 27.20 - samples/sec: 5403.39 - lr: 0.000024 - momentum: 0.000000
2023-10-18 20:42:21,795 epoch 3 - iter 1980/1984 - loss 0.24646940 - time (sec): 30.26 - samples/sec: 5404.26 - lr: 0.000023 - momentum: 0.000000
2023-10-18 20:42:21,867 ----------------------------------------------------------------------------------------------------
2023-10-18 20:42:21,867 EPOCH 3 done: loss 0.2463 - lr: 0.000023
2023-10-18 20:42:23,662 DEV : loss 0.18124178051948547 - f1-score (micro avg) 0.3771
2023-10-18 20:42:23,680 saving best model
2023-10-18 20:42:23,714 ----------------------------------------------------------------------------------------------------
2023-10-18 20:42:26,738 epoch 4 - iter 198/1984 - loss 0.23822861 - time (sec): 3.02 - samples/sec: 5210.29 - lr: 0.000023 - momentum: 0.000000
2023-10-18 20:42:29,815 epoch 4 - iter 396/1984 - loss 0.23209028 - time (sec): 6.10 - samples/sec: 5352.30 - lr: 0.000023 - momentum: 0.000000
2023-10-18 20:42:32,825 epoch 4 - iter 594/1984 - loss 0.22861673 - time (sec): 9.11 - samples/sec: 5254.98 - lr: 0.000022 - momentum: 0.000000
2023-10-18 20:42:35,855 epoch 4 - iter 792/1984 - loss 0.22608576 - time (sec): 12.14 - samples/sec: 5233.89 - lr: 0.000022 - momentum: 0.000000
2023-10-18 20:42:38,904 epoch 4 - iter 990/1984 - loss 0.22959492 - time (sec): 15.19 - samples/sec: 5306.19 - lr: 0.000022 - momentum: 0.000000
2023-10-18 20:42:41,627 epoch 4 - iter 1188/1984 - loss 0.22608383 - time (sec): 17.91 - samples/sec: 5413.99 - lr: 0.000021 - momentum: 0.000000
2023-10-18 20:42:44,722 epoch 4 - iter 1386/1984 - loss 0.22511283 - time (sec): 21.01 - samples/sec: 5393.16 - lr: 0.000021 - momentum: 0.000000
2023-10-18 20:42:47,953 epoch 4 - iter 1584/1984 - loss 0.22675445 - time (sec): 24.24 - samples/sec: 5359.52 - lr: 0.000021 - momentum: 0.000000
2023-10-18 20:42:51,209 epoch 4 - iter 1782/1984 - loss 0.22351402 - time (sec): 27.49 - samples/sec: 5338.20 - lr: 0.000020 - momentum: 0.000000
2023-10-18 20:42:54,429 epoch 4 - iter 1980/1984 - loss 0.22289582 - time (sec): 30.71 - samples/sec: 5326.93 - lr: 0.000020 - momentum: 0.000000
2023-10-18 20:42:54,494 ----------------------------------------------------------------------------------------------------
2023-10-18 20:42:54,494 EPOCH 4 done: loss 0.2227 - lr: 0.000020
2023-10-18 20:42:56,316 DEV : loss 0.1685991734266281 - f1-score (micro avg) 0.4076
2023-10-18 20:42:56,334 saving best model
2023-10-18 20:42:56,368 ----------------------------------------------------------------------------------------------------
2023-10-18 20:42:59,420 epoch 5 - iter 198/1984 - loss 0.18916448 - time (sec): 3.05 - samples/sec: 5486.17 - lr: 0.000020 - momentum: 0.000000
2023-10-18 20:43:02,418 epoch 5 - iter 396/1984 - loss 0.19556003 - time (sec): 6.05 - samples/sec: 5364.05 - lr: 0.000019 - momentum: 0.000000
2023-10-18 20:43:05,400 epoch 5 - iter 594/1984 - loss 0.19887775 - time (sec): 9.03 - samples/sec: 5281.04 - lr: 0.000019 - momentum: 0.000000
2023-10-18 20:43:08,412 epoch 5 - iter 792/1984 - loss 0.19819805 - time (sec): 12.04 - samples/sec: 5354.70 - lr: 0.000019 - momentum: 0.000000
2023-10-18 20:43:11,457 epoch 5 - iter 990/1984 - loss 0.20054697 - time (sec): 15.09 - samples/sec: 5333.89 - lr: 0.000018 - momentum: 0.000000
2023-10-18 20:43:14,512 epoch 5 - iter 1188/1984 - loss 0.20228121 - time (sec): 18.14 - samples/sec: 5324.58 - lr: 0.000018 - momentum: 0.000000
2023-10-18 20:43:17,524 epoch 5 - iter 1386/1984 - loss 0.20335431 - time (sec): 21.15 - samples/sec: 5352.50 - lr: 0.000018 - momentum: 0.000000
2023-10-18 20:43:20,519 epoch 5 - iter 1584/1984 - loss 0.20160468 - time (sec): 24.15 - samples/sec: 5381.60 - lr: 0.000017 - momentum: 0.000000
2023-10-18 20:43:23,580 epoch 5 - iter 1782/1984 - loss 0.20444659 - time (sec): 27.21 - samples/sec: 5400.66 - lr: 0.000017 - momentum: 0.000000
2023-10-18 20:43:26,633 epoch 5 - iter 1980/1984 - loss 0.20797252 - time (sec): 30.26 - samples/sec: 5405.93 - lr: 0.000017 - momentum: 0.000000
2023-10-18 20:43:26,699 ----------------------------------------------------------------------------------------------------
2023-10-18 20:43:26,699 EPOCH 5 done: loss 0.2080 - lr: 0.000017
2023-10-18 20:43:28,526 DEV : loss 0.16076675057411194 - f1-score (micro avg) 0.4272
2023-10-18 20:43:28,544 saving best model
2023-10-18 20:43:28,580 ----------------------------------------------------------------------------------------------------
2023-10-18 20:43:31,611 epoch 6 - iter 198/1984 - loss 0.18626775 - time (sec): 3.03 - samples/sec: 5385.04 - lr: 0.000016 - momentum: 0.000000
2023-10-18 20:43:34,762 epoch 6 - iter 396/1984 - loss 0.18358758 - time (sec): 6.18 - samples/sec: 5174.46 - lr: 0.000016 - momentum: 0.000000
2023-10-18 20:43:37,867 epoch 6 - iter 594/1984 - loss 0.18721054 - time (sec): 9.29 - samples/sec: 5277.32 - lr: 0.000016 - momentum: 0.000000
2023-10-18 20:43:40,881 epoch 6 - iter 792/1984 - loss 0.18658538 - time (sec): 12.30 - samples/sec: 5288.77 - lr: 0.000015 - momentum: 0.000000
2023-10-18 20:43:43,940 epoch 6 - iter 990/1984 - loss 0.18950152 - time (sec): 15.36 - samples/sec: 5274.00 - lr: 0.000015 - momentum: 0.000000
2023-10-18 20:43:46,978 epoch 6 - iter 1188/1984 - loss 0.18810016 - time (sec): 18.40 - samples/sec: 5298.65 - lr: 0.000015 - momentum: 0.000000
2023-10-18 20:43:49,811 epoch 6 - iter 1386/1984 - loss 0.19049194 - time (sec): 21.23 - samples/sec: 5340.47 - lr: 0.000014 - momentum: 0.000000
2023-10-18 20:43:52,777 epoch 6 - iter 1584/1984 - loss 0.19125360 - time (sec): 24.20 - samples/sec: 5349.24 - lr: 0.000014 - momentum: 0.000000
2023-10-18 20:43:55,574 epoch 6 - iter 1782/1984 - loss 0.19255913 - time (sec): 26.99 - samples/sec: 5394.96 - lr: 0.000014 - momentum: 0.000000
2023-10-18 20:43:58,643 epoch 6 - iter 1980/1984 - loss 0.19515696 - time (sec): 30.06 - samples/sec: 5444.12 - lr: 0.000013 - momentum: 0.000000
2023-10-18 20:43:58,710 ----------------------------------------------------------------------------------------------------
2023-10-18 20:43:58,710 EPOCH 6 done: loss 0.1950 - lr: 0.000013
2023-10-18 20:44:00,528 DEV : loss 0.1576894223690033 - f1-score (micro avg) 0.4482
2023-10-18 20:44:00,546 saving best model
2023-10-18 20:44:00,581 ----------------------------------------------------------------------------------------------------
2023-10-18 20:44:03,606 epoch 7 - iter 198/1984 - loss 0.20041330 - time (sec): 3.02 - samples/sec: 5330.01 - lr: 0.000013 - momentum: 0.000000
2023-10-18 20:44:06,304 epoch 7 - iter 396/1984 - loss 0.19185999 - time (sec): 5.72 - samples/sec: 5607.62 - lr: 0.000013 - momentum: 0.000000
2023-10-18 20:44:09,246 epoch 7 - iter 594/1984 - loss 0.19074333 - time (sec): 8.66 - samples/sec: 5651.53 - lr: 0.000012 - momentum: 0.000000
2023-10-18 20:44:12,277 epoch 7 - iter 792/1984 - loss 0.19616294 - time (sec): 11.70 - samples/sec: 5593.71 - lr: 0.000012 - momentum: 0.000000
2023-10-18 20:44:15,367 epoch 7 - iter 990/1984 - loss 0.19168219 - time (sec): 14.79 - samples/sec: 5536.96 - lr: 0.000012 - momentum: 0.000000
2023-10-18 20:44:18,405 epoch 7 - iter 1188/1984 - loss 0.18962249 - time (sec): 17.82 - samples/sec: 5504.98 - lr: 0.000011 - momentum: 0.000000
2023-10-18 20:44:21,422 epoch 7 - iter 1386/1984 - loss 0.18950612 - time (sec): 20.84 - samples/sec: 5474.31 - lr: 0.000011 - momentum: 0.000000
2023-10-18 20:44:24,452 epoch 7 - iter 1584/1984 - loss 0.18837304 - time (sec): 23.87 - samples/sec: 5449.01 - lr: 0.000011 - momentum: 0.000000
2023-10-18 20:44:27,644 epoch 7 - iter 1782/1984 - loss 0.18498108 - time (sec): 27.06 - samples/sec: 5459.72 - lr: 0.000010 - momentum: 0.000000
2023-10-18 20:44:30,700 epoch 7 - iter 1980/1984 - loss 0.18632929 - time (sec): 30.12 - samples/sec: 5433.74 - lr: 0.000010 - momentum: 0.000000
2023-10-18 20:44:30,761 ----------------------------------------------------------------------------------------------------
2023-10-18 20:44:30,761 EPOCH 7 done: loss 0.1860 - lr: 0.000010
2023-10-18 20:44:32,576 DEV : loss 0.158553346991539 - f1-score (micro avg) 0.4563
2023-10-18 20:44:32,594 saving best model
2023-10-18 20:44:32,627 ----------------------------------------------------------------------------------------------------
2023-10-18 20:44:35,383 epoch 8 - iter 198/1984 - loss 0.18600477 - time (sec): 2.75 - samples/sec: 6073.68 - lr: 0.000010 - momentum: 0.000000
2023-10-18 20:44:38,367 epoch 8 - iter 396/1984 - loss 0.17748589 - time (sec): 5.74 - samples/sec: 5766.31 - lr: 0.000009 - momentum: 0.000000
2023-10-18 20:44:41,403 epoch 8 - iter 594/1984 - loss 0.17785171 - time (sec): 8.77 - samples/sec: 5787.16 - lr: 0.000009 - momentum: 0.000000
2023-10-18 20:44:44,402 epoch 8 - iter 792/1984 - loss 0.18108144 - time (sec): 11.77 - samples/sec: 5648.04 - lr: 0.000009 - momentum: 0.000000
2023-10-18 20:44:47,271 epoch 8 - iter 990/1984 - loss 0.17884882 - time (sec): 14.64 - samples/sec: 5596.01 - lr: 0.000008 - momentum: 0.000000
2023-10-18 20:44:50,250 epoch 8 - iter 1188/1984 - loss 0.18579308 - time (sec): 17.62 - samples/sec: 5569.41 - lr: 0.000008 - momentum: 0.000000
2023-10-18 20:44:53,279 epoch 8 - iter 1386/1984 - loss 0.18431354 - time (sec): 20.65 - samples/sec: 5527.58 - lr: 0.000008 - momentum: 0.000000
2023-10-18 20:44:56,298 epoch 8 - iter 1584/1984 - loss 0.18224221 - time (sec): 23.67 - samples/sec: 5503.83 - lr: 0.000007 - momentum: 0.000000
2023-10-18 20:44:59,335 epoch 8 - iter 1782/1984 - loss 0.18145094 - time (sec): 26.71 - samples/sec: 5496.95 - lr: 0.000007 - momentum: 0.000000
2023-10-18 20:45:02,392 epoch 8 - iter 1980/1984 - loss 0.18085147 - time (sec): 29.76 - samples/sec: 5496.75 - lr: 0.000007 - momentum: 0.000000
2023-10-18 20:45:02,463 ----------------------------------------------------------------------------------------------------
2023-10-18 20:45:02,463 EPOCH 8 done: loss 0.1807 - lr: 0.000007
2023-10-18 20:45:04,686 DEV : loss 0.15737873315811157 - f1-score (micro avg) 0.4634
2023-10-18 20:45:04,704 saving best model
2023-10-18 20:45:04,739 ----------------------------------------------------------------------------------------------------
2023-10-18 20:45:07,834 epoch 9 - iter 198/1984 - loss 0.17913017 - time (sec): 3.09 - samples/sec: 5424.89 - lr: 0.000006 - momentum: 0.000000
2023-10-18 20:45:10,867 epoch 9 - iter 396/1984 - loss 0.18063398 - time (sec): 6.13 - samples/sec: 5637.56 - lr: 0.000006 - momentum: 0.000000
2023-10-18 20:45:13,885 epoch 9 - iter 594/1984 - loss 0.17485929 - time (sec): 9.14 - samples/sec: 5510.69 - lr: 0.000006 - momentum: 0.000000
2023-10-18 20:45:16,911 epoch 9 - iter 792/1984 - loss 0.17546060 - time (sec): 12.17 - samples/sec: 5414.96 - lr: 0.000005 - momentum: 0.000000
2023-10-18 20:45:19,931 epoch 9 - iter 990/1984 - loss 0.17552482 - time (sec): 15.19 - samples/sec: 5424.88 - lr: 0.000005 - momentum: 0.000000
2023-10-18 20:45:23,005 epoch 9 - iter 1188/1984 - loss 0.18049553 - time (sec): 18.26 - samples/sec: 5403.10 - lr: 0.000005 - momentum: 0.000000
2023-10-18 20:45:26,042 epoch 9 - iter 1386/1984 - loss 0.18155530 - time (sec): 21.30 - samples/sec: 5357.94 - lr: 0.000004 - momentum: 0.000000
2023-10-18 20:45:29,095 epoch 9 - iter 1584/1984 - loss 0.17860347 - time (sec): 24.36 - samples/sec: 5354.22 - lr: 0.000004 - momentum: 0.000000
2023-10-18 20:45:32,172 epoch 9 - iter 1782/1984 - loss 0.17832306 - time (sec): 27.43 - samples/sec: 5365.33 - lr: 0.000004 - momentum: 0.000000
2023-10-18 20:45:35,265 epoch 9 - iter 1980/1984 - loss 0.17617574 - time (sec): 30.53 - samples/sec: 5361.27 - lr: 0.000003 - momentum: 0.000000
2023-10-18 20:45:35,330 ----------------------------------------------------------------------------------------------------
2023-10-18 20:45:35,330 EPOCH 9 done: loss 0.1760 - lr: 0.000003
2023-10-18 20:45:37,172 DEV : loss 0.15737827122211456 - f1-score (micro avg) 0.4735
2023-10-18 20:45:37,192 saving best model
2023-10-18 20:45:37,227 ----------------------------------------------------------------------------------------------------
2023-10-18 20:45:40,620 epoch 10 - iter 198/1984 - loss 0.18505314 - time (sec): 3.39 - samples/sec: 4622.04 - lr: 0.000003 - momentum: 0.000000
2023-10-18 20:45:43,687 epoch 10 - iter 396/1984 - loss 0.17359687 - time (sec): 6.46 - samples/sec: 5035.29 - lr: 0.000003 - momentum: 0.000000
2023-10-18 20:45:46,745 epoch 10 - iter 594/1984 - loss 0.17583619 - time (sec): 9.52 - samples/sec: 5147.81 - lr: 0.000002 - momentum: 0.000000
2023-10-18 20:45:49,746 epoch 10 - iter 792/1984 - loss 0.17469714 - time (sec): 12.52 - samples/sec: 5204.51 - lr: 0.000002 - momentum: 0.000000
2023-10-18 20:45:52,506 epoch 10 - iter 990/1984 - loss 0.17268153 - time (sec): 15.28 - samples/sec: 5359.97 - lr: 0.000002 - momentum: 0.000000
2023-10-18 20:45:55,438 epoch 10 - iter 1188/1984 - loss 0.17585525 - time (sec): 18.21 - samples/sec: 5413.48 - lr: 0.000001 - momentum: 0.000000
2023-10-18 20:45:58,487 epoch 10 - iter 1386/1984 - loss 0.17518554 - time (sec): 21.26 - samples/sec: 5415.87 - lr: 0.000001 - momentum: 0.000000
2023-10-18 20:46:01,428 epoch 10 - iter 1584/1984 - loss 0.17473003 - time (sec): 24.20 - samples/sec: 5442.09 - lr: 0.000001 - momentum: 0.000000
2023-10-18 20:46:04,477 epoch 10 - iter 1782/1984 - loss 0.17456587 - time (sec): 27.25 - samples/sec: 5406.90 - lr: 0.000000 - momentum: 0.000000
2023-10-18 20:46:07,538 epoch 10 - iter 1980/1984 - loss 0.17442832 - time (sec): 30.31 - samples/sec: 5402.43 - lr: 0.000000 - momentum: 0.000000
2023-10-18 20:46:07,607 ----------------------------------------------------------------------------------------------------
2023-10-18 20:46:07,607 EPOCH 10 done: loss 0.1746 - lr: 0.000000
2023-10-18 20:46:09,452 DEV : loss 0.15687035024166107 - f1-score (micro avg) 0.4721
2023-10-18 20:46:09,500 ----------------------------------------------------------------------------------------------------
2023-10-18 20:46:09,501 Loading model from best epoch ...
2023-10-18 20:46:09,584 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
2023-10-18 20:46:11,130
Results:
- F-score (micro) 0.5343
- F-score (macro) 0.3481
- Accuracy 0.4039
By class:
precision recall f1-score support
LOC 0.6647 0.6779 0.6712 655
PER 0.2786 0.5022 0.3584 223
ORG 0.1000 0.0079 0.0146 127
micro avg 0.5157 0.5542 0.5343 1005
macro avg 0.3478 0.3960 0.3481 1005
weighted avg 0.5077 0.5542 0.5188 1005
2023-10-18 20:46:11,130 ----------------------------------------------------------------------------------------------------
|