File size: 24,056 Bytes
eb83e71 |
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 |
2023-10-14 18:32:53,099 ----------------------------------------------------------------------------------------------------
2023-10-14 18:32:53,100 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()
)"
2023-10-14 18:32:53,100 ----------------------------------------------------------------------------------------------------
2023-10-14 18:32:53,100 MultiCorpus: 14465 train + 1392 dev + 2432 test sentences
- NER_HIPE_2022 Corpus: 14465 train + 1392 dev + 2432 test sentences - /root/.flair/datasets/ner_hipe_2022/v2.1/letemps/fr/with_doc_seperator
2023-10-14 18:32:53,100 ----------------------------------------------------------------------------------------------------
2023-10-14 18:32:53,100 Train: 14465 sentences
2023-10-14 18:32:53,100 (train_with_dev=False, train_with_test=False)
2023-10-14 18:32:53,100 ----------------------------------------------------------------------------------------------------
2023-10-14 18:32:53,100 Training Params:
2023-10-14 18:32:53,101 - learning_rate: "3e-05"
2023-10-14 18:32:53,101 - mini_batch_size: "4"
2023-10-14 18:32:53,101 - max_epochs: "10"
2023-10-14 18:32:53,101 - shuffle: "True"
2023-10-14 18:32:53,101 ----------------------------------------------------------------------------------------------------
2023-10-14 18:32:53,101 Plugins:
2023-10-14 18:32:53,101 - LinearScheduler | warmup_fraction: '0.1'
2023-10-14 18:32:53,101 ----------------------------------------------------------------------------------------------------
2023-10-14 18:32:53,101 Final evaluation on model from best epoch (best-model.pt)
2023-10-14 18:32:53,101 - metric: "('micro avg', 'f1-score')"
2023-10-14 18:32:53,101 ----------------------------------------------------------------------------------------------------
2023-10-14 18:32:53,101 Computation:
2023-10-14 18:32:53,101 - compute on device: cuda:0
2023-10-14 18:32:53,101 - embedding storage: none
2023-10-14 18:32:53,101 ----------------------------------------------------------------------------------------------------
2023-10-14 18:32:53,101 Model training base path: "hmbench-letemps/fr-dbmdz/bert-base-historic-multilingual-cased-bs4-wsFalse-e10-lr3e-05-poolingfirst-layers-1-crfFalse-1"
2023-10-14 18:32:53,101 ----------------------------------------------------------------------------------------------------
2023-10-14 18:32:53,101 ----------------------------------------------------------------------------------------------------
2023-10-14 18:33:11,011 epoch 1 - iter 361/3617 - loss 1.49793342 - time (sec): 17.91 - samples/sec: 2144.90 - lr: 0.000003 - momentum: 0.000000
2023-10-14 18:33:28,000 epoch 1 - iter 722/3617 - loss 0.85914597 - time (sec): 34.90 - samples/sec: 2185.64 - lr: 0.000006 - momentum: 0.000000
2023-10-14 18:33:44,513 epoch 1 - iter 1083/3617 - loss 0.63406158 - time (sec): 51.41 - samples/sec: 2213.73 - lr: 0.000009 - momentum: 0.000000
2023-10-14 18:34:01,032 epoch 1 - iter 1444/3617 - loss 0.50536750 - time (sec): 67.93 - samples/sec: 2252.43 - lr: 0.000012 - momentum: 0.000000
2023-10-14 18:34:17,641 epoch 1 - iter 1805/3617 - loss 0.43148914 - time (sec): 84.54 - samples/sec: 2252.23 - lr: 0.000015 - momentum: 0.000000
2023-10-14 18:34:33,662 epoch 1 - iter 2166/3617 - loss 0.38143642 - time (sec): 100.56 - samples/sec: 2265.58 - lr: 0.000018 - momentum: 0.000000
2023-10-14 18:34:50,400 epoch 1 - iter 2527/3617 - loss 0.34330335 - time (sec): 117.30 - samples/sec: 2260.05 - lr: 0.000021 - momentum: 0.000000
2023-10-14 18:35:06,830 epoch 1 - iter 2888/3617 - loss 0.31397767 - time (sec): 133.73 - samples/sec: 2279.40 - lr: 0.000024 - momentum: 0.000000
2023-10-14 18:35:23,147 epoch 1 - iter 3249/3617 - loss 0.29300894 - time (sec): 150.04 - samples/sec: 2278.06 - lr: 0.000027 - momentum: 0.000000
2023-10-14 18:35:39,897 epoch 1 - iter 3610/3617 - loss 0.27489902 - time (sec): 166.80 - samples/sec: 2273.67 - lr: 0.000030 - momentum: 0.000000
2023-10-14 18:35:40,216 ----------------------------------------------------------------------------------------------------
2023-10-14 18:35:40,217 EPOCH 1 done: loss 0.2746 - lr: 0.000030
2023-10-14 18:35:45,572 DEV : loss 0.1469813734292984 - f1-score (micro avg) 0.5839
2023-10-14 18:35:45,603 saving best model
2023-10-14 18:35:46,156 ----------------------------------------------------------------------------------------------------
2023-10-14 18:36:05,275 epoch 2 - iter 361/3617 - loss 0.09657648 - time (sec): 19.12 - samples/sec: 1994.77 - lr: 0.000030 - momentum: 0.000000
2023-10-14 18:36:24,315 epoch 2 - iter 722/3617 - loss 0.09526196 - time (sec): 38.16 - samples/sec: 2002.34 - lr: 0.000029 - momentum: 0.000000
2023-10-14 18:36:43,401 epoch 2 - iter 1083/3617 - loss 0.09794662 - time (sec): 57.24 - samples/sec: 2007.27 - lr: 0.000029 - momentum: 0.000000
2023-10-14 18:37:02,359 epoch 2 - iter 1444/3617 - loss 0.10000382 - time (sec): 76.20 - samples/sec: 1996.00 - lr: 0.000029 - momentum: 0.000000
2023-10-14 18:37:21,435 epoch 2 - iter 1805/3617 - loss 0.09984349 - time (sec): 95.28 - samples/sec: 1984.72 - lr: 0.000028 - momentum: 0.000000
2023-10-14 18:37:40,211 epoch 2 - iter 2166/3617 - loss 0.09842580 - time (sec): 114.05 - samples/sec: 1997.07 - lr: 0.000028 - momentum: 0.000000
2023-10-14 18:37:59,219 epoch 2 - iter 2527/3617 - loss 0.09886238 - time (sec): 133.06 - samples/sec: 1999.42 - lr: 0.000028 - momentum: 0.000000
2023-10-14 18:38:18,114 epoch 2 - iter 2888/3617 - loss 0.09750237 - time (sec): 151.96 - samples/sec: 1997.68 - lr: 0.000027 - momentum: 0.000000
2023-10-14 18:38:36,932 epoch 2 - iter 3249/3617 - loss 0.09684901 - time (sec): 170.77 - samples/sec: 2000.95 - lr: 0.000027 - momentum: 0.000000
2023-10-14 18:38:55,778 epoch 2 - iter 3610/3617 - loss 0.09693424 - time (sec): 189.62 - samples/sec: 2000.05 - lr: 0.000027 - momentum: 0.000000
2023-10-14 18:38:56,131 ----------------------------------------------------------------------------------------------------
2023-10-14 18:38:56,131 EPOCH 2 done: loss 0.0970 - lr: 0.000027
2023-10-14 18:39:02,429 DEV : loss 0.14265793561935425 - f1-score (micro avg) 0.6193
2023-10-14 18:39:02,457 saving best model
2023-10-14 18:39:02,982 ----------------------------------------------------------------------------------------------------
2023-10-14 18:39:21,781 epoch 3 - iter 361/3617 - loss 0.06955254 - time (sec): 18.79 - samples/sec: 1884.93 - lr: 0.000026 - momentum: 0.000000
2023-10-14 18:39:40,836 epoch 3 - iter 722/3617 - loss 0.06922338 - time (sec): 37.85 - samples/sec: 1943.55 - lr: 0.000026 - momentum: 0.000000
2023-10-14 18:39:59,620 epoch 3 - iter 1083/3617 - loss 0.07232891 - time (sec): 56.63 - samples/sec: 1967.42 - lr: 0.000026 - momentum: 0.000000
2023-10-14 18:40:18,296 epoch 3 - iter 1444/3617 - loss 0.07277401 - time (sec): 75.31 - samples/sec: 2002.36 - lr: 0.000025 - momentum: 0.000000
2023-10-14 18:40:36,983 epoch 3 - iter 1805/3617 - loss 0.07138983 - time (sec): 94.00 - samples/sec: 2006.08 - lr: 0.000025 - momentum: 0.000000
2023-10-14 18:40:55,720 epoch 3 - iter 2166/3617 - loss 0.07379404 - time (sec): 112.73 - samples/sec: 2016.27 - lr: 0.000025 - momentum: 0.000000
2023-10-14 18:41:14,768 epoch 3 - iter 2527/3617 - loss 0.07306798 - time (sec): 131.78 - samples/sec: 2017.47 - lr: 0.000024 - momentum: 0.000000
2023-10-14 18:41:33,623 epoch 3 - iter 2888/3617 - loss 0.07400133 - time (sec): 150.64 - samples/sec: 2014.28 - lr: 0.000024 - momentum: 0.000000
2023-10-14 18:41:52,426 epoch 3 - iter 3249/3617 - loss 0.07385904 - time (sec): 169.44 - samples/sec: 2012.00 - lr: 0.000024 - momentum: 0.000000
2023-10-14 18:42:11,275 epoch 3 - iter 3610/3617 - loss 0.07300804 - time (sec): 188.29 - samples/sec: 2014.74 - lr: 0.000023 - momentum: 0.000000
2023-10-14 18:42:11,633 ----------------------------------------------------------------------------------------------------
2023-10-14 18:42:11,633 EPOCH 3 done: loss 0.0730 - lr: 0.000023
2023-10-14 18:42:17,786 DEV : loss 0.19656439125537872 - f1-score (micro avg) 0.6213
2023-10-14 18:42:17,815 saving best model
2023-10-14 18:42:18,413 ----------------------------------------------------------------------------------------------------
2023-10-14 18:42:37,058 epoch 4 - iter 361/3617 - loss 0.04567558 - time (sec): 18.64 - samples/sec: 2104.23 - lr: 0.000023 - momentum: 0.000000
2023-10-14 18:42:55,665 epoch 4 - iter 722/3617 - loss 0.04907222 - time (sec): 37.25 - samples/sec: 2057.31 - lr: 0.000023 - momentum: 0.000000
2023-10-14 18:43:14,466 epoch 4 - iter 1083/3617 - loss 0.04822914 - time (sec): 56.05 - samples/sec: 2037.22 - lr: 0.000022 - momentum: 0.000000
2023-10-14 18:43:31,749 epoch 4 - iter 1444/3617 - loss 0.04834016 - time (sec): 73.33 - samples/sec: 2053.38 - lr: 0.000022 - momentum: 0.000000
2023-10-14 18:43:48,091 epoch 4 - iter 1805/3617 - loss 0.04851364 - time (sec): 89.68 - samples/sec: 2110.20 - lr: 0.000022 - momentum: 0.000000
2023-10-14 18:44:04,310 epoch 4 - iter 2166/3617 - loss 0.04902986 - time (sec): 105.90 - samples/sec: 2140.07 - lr: 0.000021 - momentum: 0.000000
2023-10-14 18:44:20,694 epoch 4 - iter 2527/3617 - loss 0.04948457 - time (sec): 122.28 - samples/sec: 2166.31 - lr: 0.000021 - momentum: 0.000000
2023-10-14 18:44:36,979 epoch 4 - iter 2888/3617 - loss 0.05023602 - time (sec): 138.57 - samples/sec: 2189.47 - lr: 0.000021 - momentum: 0.000000
2023-10-14 18:44:54,061 epoch 4 - iter 3249/3617 - loss 0.05065579 - time (sec): 155.65 - samples/sec: 2194.24 - lr: 0.000020 - momentum: 0.000000
2023-10-14 18:45:12,956 epoch 4 - iter 3610/3617 - loss 0.05150400 - time (sec): 174.54 - samples/sec: 2172.39 - lr: 0.000020 - momentum: 0.000000
2023-10-14 18:45:13,320 ----------------------------------------------------------------------------------------------------
2023-10-14 18:45:13,320 EPOCH 4 done: loss 0.0514 - lr: 0.000020
2023-10-14 18:45:18,927 DEV : loss 0.27270472049713135 - f1-score (micro avg) 0.622
2023-10-14 18:45:18,960 saving best model
2023-10-14 18:45:19,676 ----------------------------------------------------------------------------------------------------
2023-10-14 18:45:36,826 epoch 5 - iter 361/3617 - loss 0.03804518 - time (sec): 17.15 - samples/sec: 2296.61 - lr: 0.000020 - momentum: 0.000000
2023-10-14 18:45:55,785 epoch 5 - iter 722/3617 - loss 0.03841622 - time (sec): 36.11 - samples/sec: 2135.41 - lr: 0.000019 - momentum: 0.000000
2023-10-14 18:46:14,740 epoch 5 - iter 1083/3617 - loss 0.03924782 - time (sec): 55.06 - samples/sec: 2090.46 - lr: 0.000019 - momentum: 0.000000
2023-10-14 18:46:33,750 epoch 5 - iter 1444/3617 - loss 0.04097206 - time (sec): 74.07 - samples/sec: 2051.11 - lr: 0.000019 - momentum: 0.000000
2023-10-14 18:46:53,472 epoch 5 - iter 1805/3617 - loss 0.03960546 - time (sec): 93.79 - samples/sec: 2036.03 - lr: 0.000018 - momentum: 0.000000
2023-10-14 18:47:12,427 epoch 5 - iter 2166/3617 - loss 0.04004957 - time (sec): 112.75 - samples/sec: 2034.50 - lr: 0.000018 - momentum: 0.000000
2023-10-14 18:47:31,472 epoch 5 - iter 2527/3617 - loss 0.03876434 - time (sec): 131.79 - samples/sec: 2032.00 - lr: 0.000018 - momentum: 0.000000
2023-10-14 18:47:50,424 epoch 5 - iter 2888/3617 - loss 0.03792790 - time (sec): 150.75 - samples/sec: 2021.21 - lr: 0.000017 - momentum: 0.000000
2023-10-14 18:48:09,228 epoch 5 - iter 3249/3617 - loss 0.03873215 - time (sec): 169.55 - samples/sec: 2022.58 - lr: 0.000017 - momentum: 0.000000
2023-10-14 18:48:28,017 epoch 5 - iter 3610/3617 - loss 0.03865296 - time (sec): 188.34 - samples/sec: 2013.85 - lr: 0.000017 - momentum: 0.000000
2023-10-14 18:48:28,376 ----------------------------------------------------------------------------------------------------
2023-10-14 18:48:28,377 EPOCH 5 done: loss 0.0386 - lr: 0.000017
2023-10-14 18:48:33,939 DEV : loss 0.32902124524116516 - f1-score (micro avg) 0.6482
2023-10-14 18:48:33,971 saving best model
2023-10-14 18:48:34,499 ----------------------------------------------------------------------------------------------------
2023-10-14 18:48:53,479 epoch 6 - iter 361/3617 - loss 0.02506183 - time (sec): 18.98 - samples/sec: 2000.72 - lr: 0.000016 - momentum: 0.000000
2023-10-14 18:49:12,488 epoch 6 - iter 722/3617 - loss 0.02531762 - time (sec): 37.98 - samples/sec: 1980.58 - lr: 0.000016 - momentum: 0.000000
2023-10-14 18:49:31,432 epoch 6 - iter 1083/3617 - loss 0.02644327 - time (sec): 56.93 - samples/sec: 1964.83 - lr: 0.000016 - momentum: 0.000000
2023-10-14 18:49:50,306 epoch 6 - iter 1444/3617 - loss 0.02682723 - time (sec): 75.80 - samples/sec: 1972.03 - lr: 0.000015 - momentum: 0.000000
2023-10-14 18:50:09,249 epoch 6 - iter 1805/3617 - loss 0.02642346 - time (sec): 94.75 - samples/sec: 1987.90 - lr: 0.000015 - momentum: 0.000000
2023-10-14 18:50:28,143 epoch 6 - iter 2166/3617 - loss 0.02688605 - time (sec): 113.64 - samples/sec: 1995.11 - lr: 0.000015 - momentum: 0.000000
2023-10-14 18:50:47,094 epoch 6 - iter 2527/3617 - loss 0.02597740 - time (sec): 132.59 - samples/sec: 1991.28 - lr: 0.000014 - momentum: 0.000000
2023-10-14 18:51:06,110 epoch 6 - iter 2888/3617 - loss 0.02657743 - time (sec): 151.61 - samples/sec: 2004.75 - lr: 0.000014 - momentum: 0.000000
2023-10-14 18:51:24,765 epoch 6 - iter 3249/3617 - loss 0.02701549 - time (sec): 170.26 - samples/sec: 2001.72 - lr: 0.000014 - momentum: 0.000000
2023-10-14 18:51:43,525 epoch 6 - iter 3610/3617 - loss 0.02721336 - time (sec): 189.02 - samples/sec: 2006.30 - lr: 0.000013 - momentum: 0.000000
2023-10-14 18:51:43,876 ----------------------------------------------------------------------------------------------------
2023-10-14 18:51:43,877 EPOCH 6 done: loss 0.0272 - lr: 0.000013
2023-10-14 18:51:50,080 DEV : loss 0.3020953834056854 - f1-score (micro avg) 0.6316
2023-10-14 18:51:50,109 ----------------------------------------------------------------------------------------------------
2023-10-14 18:52:09,031 epoch 7 - iter 361/3617 - loss 0.01566413 - time (sec): 18.92 - samples/sec: 2050.15 - lr: 0.000013 - momentum: 0.000000
2023-10-14 18:52:28,180 epoch 7 - iter 722/3617 - loss 0.01602060 - time (sec): 38.07 - samples/sec: 2024.60 - lr: 0.000013 - momentum: 0.000000
2023-10-14 18:52:47,111 epoch 7 - iter 1083/3617 - loss 0.01755375 - time (sec): 57.00 - samples/sec: 2017.47 - lr: 0.000012 - momentum: 0.000000
2023-10-14 18:53:05,964 epoch 7 - iter 1444/3617 - loss 0.01898152 - time (sec): 75.85 - samples/sec: 2014.71 - lr: 0.000012 - momentum: 0.000000
2023-10-14 18:53:24,801 epoch 7 - iter 1805/3617 - loss 0.02029282 - time (sec): 94.69 - samples/sec: 2013.35 - lr: 0.000012 - momentum: 0.000000
2023-10-14 18:53:43,723 epoch 7 - iter 2166/3617 - loss 0.01945143 - time (sec): 113.61 - samples/sec: 2011.18 - lr: 0.000011 - momentum: 0.000000
2023-10-14 18:54:02,484 epoch 7 - iter 2527/3617 - loss 0.02008824 - time (sec): 132.37 - samples/sec: 2009.24 - lr: 0.000011 - momentum: 0.000000
2023-10-14 18:54:21,323 epoch 7 - iter 2888/3617 - loss 0.02026028 - time (sec): 151.21 - samples/sec: 2011.04 - lr: 0.000011 - momentum: 0.000000
2023-10-14 18:54:40,125 epoch 7 - iter 3249/3617 - loss 0.02016038 - time (sec): 170.01 - samples/sec: 2007.85 - lr: 0.000010 - momentum: 0.000000
2023-10-14 18:54:58,847 epoch 7 - iter 3610/3617 - loss 0.02025047 - time (sec): 188.74 - samples/sec: 2009.28 - lr: 0.000010 - momentum: 0.000000
2023-10-14 18:54:59,208 ----------------------------------------------------------------------------------------------------
2023-10-14 18:54:59,209 EPOCH 7 done: loss 0.0203 - lr: 0.000010
2023-10-14 18:55:05,497 DEV : loss 0.33693569898605347 - f1-score (micro avg) 0.6245
2023-10-14 18:55:05,525 ----------------------------------------------------------------------------------------------------
2023-10-14 18:55:24,336 epoch 8 - iter 361/3617 - loss 0.01327463 - time (sec): 18.81 - samples/sec: 1964.88 - lr: 0.000010 - momentum: 0.000000
2023-10-14 18:55:43,223 epoch 8 - iter 722/3617 - loss 0.01473725 - time (sec): 37.70 - samples/sec: 1989.05 - lr: 0.000009 - momentum: 0.000000
2023-10-14 18:56:02,086 epoch 8 - iter 1083/3617 - loss 0.01272106 - time (sec): 56.56 - samples/sec: 2011.45 - lr: 0.000009 - momentum: 0.000000
2023-10-14 18:56:20,916 epoch 8 - iter 1444/3617 - loss 0.01306734 - time (sec): 75.39 - samples/sec: 2011.73 - lr: 0.000009 - momentum: 0.000000
2023-10-14 18:56:39,693 epoch 8 - iter 1805/3617 - loss 0.01336643 - time (sec): 94.17 - samples/sec: 2006.28 - lr: 0.000008 - momentum: 0.000000
2023-10-14 18:56:58,759 epoch 8 - iter 2166/3617 - loss 0.01331363 - time (sec): 113.23 - samples/sec: 1982.43 - lr: 0.000008 - momentum: 0.000000
2023-10-14 18:57:17,808 epoch 8 - iter 2527/3617 - loss 0.01360786 - time (sec): 132.28 - samples/sec: 1995.37 - lr: 0.000008 - momentum: 0.000000
2023-10-14 18:57:36,722 epoch 8 - iter 2888/3617 - loss 0.01309021 - time (sec): 151.20 - samples/sec: 1993.35 - lr: 0.000007 - momentum: 0.000000
2023-10-14 18:57:55,357 epoch 8 - iter 3249/3617 - loss 0.01286633 - time (sec): 169.83 - samples/sec: 2004.16 - lr: 0.000007 - momentum: 0.000000
2023-10-14 18:58:14,255 epoch 8 - iter 3610/3617 - loss 0.01337274 - time (sec): 188.73 - samples/sec: 2008.80 - lr: 0.000007 - momentum: 0.000000
2023-10-14 18:58:14,611 ----------------------------------------------------------------------------------------------------
2023-10-14 18:58:14,611 EPOCH 8 done: loss 0.0134 - lr: 0.000007
2023-10-14 18:58:20,871 DEV : loss 0.3728400766849518 - f1-score (micro avg) 0.6292
2023-10-14 18:58:20,900 ----------------------------------------------------------------------------------------------------
2023-10-14 18:58:39,836 epoch 9 - iter 361/3617 - loss 0.00636871 - time (sec): 18.93 - samples/sec: 2004.28 - lr: 0.000006 - momentum: 0.000000
2023-10-14 18:58:58,882 epoch 9 - iter 722/3617 - loss 0.00907563 - time (sec): 37.98 - samples/sec: 2028.07 - lr: 0.000006 - momentum: 0.000000
2023-10-14 18:59:17,616 epoch 9 - iter 1083/3617 - loss 0.00828252 - time (sec): 56.72 - samples/sec: 2036.27 - lr: 0.000006 - momentum: 0.000000
2023-10-14 18:59:36,465 epoch 9 - iter 1444/3617 - loss 0.00898628 - time (sec): 75.56 - samples/sec: 2012.65 - lr: 0.000005 - momentum: 0.000000
2023-10-14 18:59:55,560 epoch 9 - iter 1805/3617 - loss 0.00817358 - time (sec): 94.66 - samples/sec: 2003.84 - lr: 0.000005 - momentum: 0.000000
2023-10-14 19:00:14,343 epoch 9 - iter 2166/3617 - loss 0.00817291 - time (sec): 113.44 - samples/sec: 2002.35 - lr: 0.000005 - momentum: 0.000000
2023-10-14 19:00:33,183 epoch 9 - iter 2527/3617 - loss 0.00832167 - time (sec): 132.28 - samples/sec: 1994.85 - lr: 0.000004 - momentum: 0.000000
2023-10-14 19:00:52,035 epoch 9 - iter 2888/3617 - loss 0.00797829 - time (sec): 151.13 - samples/sec: 1996.91 - lr: 0.000004 - momentum: 0.000000
2023-10-14 19:01:10,800 epoch 9 - iter 3249/3617 - loss 0.00799275 - time (sec): 169.90 - samples/sec: 2003.98 - lr: 0.000004 - momentum: 0.000000
2023-10-14 19:01:30,189 epoch 9 - iter 3610/3617 - loss 0.00795879 - time (sec): 189.29 - samples/sec: 2003.69 - lr: 0.000003 - momentum: 0.000000
2023-10-14 19:01:30,546 ----------------------------------------------------------------------------------------------------
2023-10-14 19:01:30,546 EPOCH 9 done: loss 0.0079 - lr: 0.000003
2023-10-14 19:01:36,746 DEV : loss 0.3810341954231262 - f1-score (micro avg) 0.6453
2023-10-14 19:01:36,775 ----------------------------------------------------------------------------------------------------
2023-10-14 19:01:55,572 epoch 10 - iter 361/3617 - loss 0.00233359 - time (sec): 18.80 - samples/sec: 2041.41 - lr: 0.000003 - momentum: 0.000000
2023-10-14 19:02:12,702 epoch 10 - iter 722/3617 - loss 0.00511219 - time (sec): 35.93 - samples/sec: 2151.13 - lr: 0.000003 - momentum: 0.000000
2023-10-14 19:02:29,091 epoch 10 - iter 1083/3617 - loss 0.00547615 - time (sec): 52.32 - samples/sec: 2185.00 - lr: 0.000002 - momentum: 0.000000
2023-10-14 19:02:45,447 epoch 10 - iter 1444/3617 - loss 0.00486842 - time (sec): 68.67 - samples/sec: 2216.76 - lr: 0.000002 - momentum: 0.000000
2023-10-14 19:03:02,177 epoch 10 - iter 1805/3617 - loss 0.00509004 - time (sec): 85.40 - samples/sec: 2231.64 - lr: 0.000002 - momentum: 0.000000
2023-10-14 19:03:19,019 epoch 10 - iter 2166/3617 - loss 0.00541113 - time (sec): 102.24 - samples/sec: 2238.08 - lr: 0.000001 - momentum: 0.000000
2023-10-14 19:03:35,397 epoch 10 - iter 2527/3617 - loss 0.00522420 - time (sec): 118.62 - samples/sec: 2236.37 - lr: 0.000001 - momentum: 0.000000
2023-10-14 19:03:51,652 epoch 10 - iter 2888/3617 - loss 0.00553422 - time (sec): 134.88 - samples/sec: 2230.99 - lr: 0.000001 - momentum: 0.000000
2023-10-14 19:04:08,143 epoch 10 - iter 3249/3617 - loss 0.00538172 - time (sec): 151.37 - samples/sec: 2243.70 - lr: 0.000000 - momentum: 0.000000
2023-10-14 19:04:24,734 epoch 10 - iter 3610/3617 - loss 0.00535691 - time (sec): 167.96 - samples/sec: 2258.53 - lr: 0.000000 - momentum: 0.000000
2023-10-14 19:04:25,047 ----------------------------------------------------------------------------------------------------
2023-10-14 19:04:25,047 EPOCH 10 done: loss 0.0053 - lr: 0.000000
2023-10-14 19:04:30,571 DEV : loss 0.395516961812973 - f1-score (micro avg) 0.6422
2023-10-14 19:04:31,008 ----------------------------------------------------------------------------------------------------
2023-10-14 19:04:31,009 Loading model from best epoch ...
2023-10-14 19:04:33,509 SequenceTagger predicts: Dictionary with 13 tags: O, S-loc, B-loc, E-loc, I-loc, S-pers, B-pers, E-pers, I-pers, S-org, B-org, E-org, I-org
2023-10-14 19:04:40,299
Results:
- F-score (micro) 0.6431
- F-score (macro) 0.5165
- Accuracy 0.4908
By class:
precision recall f1-score support
loc 0.6159 0.7868 0.6909 591
pers 0.5281 0.8151 0.6410 357
org 0.2353 0.2025 0.2177 79
micro avg 0.5619 0.7517 0.6431 1027
macro avg 0.4598 0.6015 0.5165 1027
weighted avg 0.5561 0.7517 0.6372 1027
2023-10-14 19:04:40,299 ----------------------------------------------------------------------------------------------------
|