File size: 24,268 Bytes
29e9179 |
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 |
2023-10-25 18:20:32,214 ----------------------------------------------------------------------------------------------------
2023-10-25 18:20:32,215 Model: "SequenceTagger(
(embeddings): TransformerWordEmbeddings(
(model): BertModel(
(embeddings): BertEmbeddings(
(word_embeddings): Embedding(64001, 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=17, bias=True)
(loss_function): CrossEntropyLoss()
)"
2023-10-25 18:20:32,216 ----------------------------------------------------------------------------------------------------
2023-10-25 18:20:32,216 MultiCorpus: 7142 train + 698 dev + 2570 test sentences
- NER_HIPE_2022 Corpus: 7142 train + 698 dev + 2570 test sentences - /root/.flair/datasets/ner_hipe_2022/v2.1/newseye/fr/with_doc_seperator
2023-10-25 18:20:32,216 ----------------------------------------------------------------------------------------------------
2023-10-25 18:20:32,216 Train: 7142 sentences
2023-10-25 18:20:32,216 (train_with_dev=False, train_with_test=False)
2023-10-25 18:20:32,216 ----------------------------------------------------------------------------------------------------
2023-10-25 18:20:32,216 Training Params:
2023-10-25 18:20:32,216 - learning_rate: "3e-05"
2023-10-25 18:20:32,216 - mini_batch_size: "4"
2023-10-25 18:20:32,216 - max_epochs: "10"
2023-10-25 18:20:32,216 - shuffle: "True"
2023-10-25 18:20:32,216 ----------------------------------------------------------------------------------------------------
2023-10-25 18:20:32,216 Plugins:
2023-10-25 18:20:32,216 - TensorboardLogger
2023-10-25 18:20:32,216 - LinearScheduler | warmup_fraction: '0.1'
2023-10-25 18:20:32,216 ----------------------------------------------------------------------------------------------------
2023-10-25 18:20:32,216 Final evaluation on model from best epoch (best-model.pt)
2023-10-25 18:20:32,216 - metric: "('micro avg', 'f1-score')"
2023-10-25 18:20:32,216 ----------------------------------------------------------------------------------------------------
2023-10-25 18:20:32,216 Computation:
2023-10-25 18:20:32,216 - compute on device: cuda:0
2023-10-25 18:20:32,216 - embedding storage: none
2023-10-25 18:20:32,217 ----------------------------------------------------------------------------------------------------
2023-10-25 18:20:32,217 Model training base path: "hmbench-newseye/fr-dbmdz/bert-base-historic-multilingual-64k-td-cased-bs4-wsFalse-e10-lr3e-05-poolingfirst-layers-1-crfFalse-5"
2023-10-25 18:20:32,217 ----------------------------------------------------------------------------------------------------
2023-10-25 18:20:32,217 ----------------------------------------------------------------------------------------------------
2023-10-25 18:20:32,217 Logging anything other than scalars to TensorBoard is currently not supported.
2023-10-25 18:20:41,829 epoch 1 - iter 178/1786 - loss 1.62894423 - time (sec): 9.61 - samples/sec: 2374.31 - lr: 0.000003 - momentum: 0.000000
2023-10-25 18:20:51,370 epoch 1 - iter 356/1786 - loss 1.05338819 - time (sec): 19.15 - samples/sec: 2440.03 - lr: 0.000006 - momentum: 0.000000
2023-10-25 18:21:00,897 epoch 1 - iter 534/1786 - loss 0.81520377 - time (sec): 28.68 - samples/sec: 2483.98 - lr: 0.000009 - momentum: 0.000000
2023-10-25 18:21:10,205 epoch 1 - iter 712/1786 - loss 0.65815637 - time (sec): 37.99 - samples/sec: 2589.41 - lr: 0.000012 - momentum: 0.000000
2023-10-25 18:21:18,795 epoch 1 - iter 890/1786 - loss 0.56620539 - time (sec): 46.58 - samples/sec: 2635.83 - lr: 0.000015 - momentum: 0.000000
2023-10-25 18:21:27,476 epoch 1 - iter 1068/1786 - loss 0.50136101 - time (sec): 55.26 - samples/sec: 2648.12 - lr: 0.000018 - momentum: 0.000000
2023-10-25 18:21:36,384 epoch 1 - iter 1246/1786 - loss 0.45154314 - time (sec): 64.17 - samples/sec: 2662.89 - lr: 0.000021 - momentum: 0.000000
2023-10-25 18:21:45,468 epoch 1 - iter 1424/1786 - loss 0.41094756 - time (sec): 73.25 - samples/sec: 2680.96 - lr: 0.000024 - momentum: 0.000000
2023-10-25 18:21:54,948 epoch 1 - iter 1602/1786 - loss 0.37891662 - time (sec): 82.73 - samples/sec: 2692.41 - lr: 0.000027 - momentum: 0.000000
2023-10-25 18:22:04,798 epoch 1 - iter 1780/1786 - loss 0.35716707 - time (sec): 92.58 - samples/sec: 2677.72 - lr: 0.000030 - momentum: 0.000000
2023-10-25 18:22:05,143 ----------------------------------------------------------------------------------------------------
2023-10-25 18:22:05,143 EPOCH 1 done: loss 0.3565 - lr: 0.000030
2023-10-25 18:22:08,918 DEV : loss 0.10421743243932724 - f1-score (micro avg) 0.7273
2023-10-25 18:22:08,940 saving best model
2023-10-25 18:22:09,391 ----------------------------------------------------------------------------------------------------
2023-10-25 18:22:19,092 epoch 2 - iter 178/1786 - loss 0.11382450 - time (sec): 9.70 - samples/sec: 2665.09 - lr: 0.000030 - momentum: 0.000000
2023-10-25 18:22:28,505 epoch 2 - iter 356/1786 - loss 0.11818253 - time (sec): 19.11 - samples/sec: 2534.69 - lr: 0.000029 - momentum: 0.000000
2023-10-25 18:22:37,505 epoch 2 - iter 534/1786 - loss 0.11752290 - time (sec): 28.11 - samples/sec: 2607.90 - lr: 0.000029 - momentum: 0.000000
2023-10-25 18:22:46,687 epoch 2 - iter 712/1786 - loss 0.11710291 - time (sec): 37.29 - samples/sec: 2656.23 - lr: 0.000029 - momentum: 0.000000
2023-10-25 18:22:55,874 epoch 2 - iter 890/1786 - loss 0.11761515 - time (sec): 46.48 - samples/sec: 2626.69 - lr: 0.000028 - momentum: 0.000000
2023-10-25 18:23:04,857 epoch 2 - iter 1068/1786 - loss 0.11880463 - time (sec): 55.46 - samples/sec: 2647.77 - lr: 0.000028 - momentum: 0.000000
2023-10-25 18:23:13,642 epoch 2 - iter 1246/1786 - loss 0.11830693 - time (sec): 64.25 - samples/sec: 2669.76 - lr: 0.000028 - momentum: 0.000000
2023-10-25 18:23:22,746 epoch 2 - iter 1424/1786 - loss 0.11678831 - time (sec): 73.35 - samples/sec: 2706.42 - lr: 0.000027 - momentum: 0.000000
2023-10-25 18:23:31,977 epoch 2 - iter 1602/1786 - loss 0.11600197 - time (sec): 82.58 - samples/sec: 2700.85 - lr: 0.000027 - momentum: 0.000000
2023-10-25 18:23:41,217 epoch 2 - iter 1780/1786 - loss 0.11609587 - time (sec): 91.82 - samples/sec: 2701.26 - lr: 0.000027 - momentum: 0.000000
2023-10-25 18:23:41,525 ----------------------------------------------------------------------------------------------------
2023-10-25 18:23:41,526 EPOCH 2 done: loss 0.1160 - lr: 0.000027
2023-10-25 18:23:46,583 DEV : loss 0.10009025037288666 - f1-score (micro avg) 0.7704
2023-10-25 18:23:46,604 saving best model
2023-10-25 18:23:47,260 ----------------------------------------------------------------------------------------------------
2023-10-25 18:23:56,854 epoch 3 - iter 178/1786 - loss 0.06189937 - time (sec): 9.59 - samples/sec: 2612.13 - lr: 0.000026 - momentum: 0.000000
2023-10-25 18:24:06,469 epoch 3 - iter 356/1786 - loss 0.06991416 - time (sec): 19.21 - samples/sec: 2562.50 - lr: 0.000026 - momentum: 0.000000
2023-10-25 18:24:15,985 epoch 3 - iter 534/1786 - loss 0.07505640 - time (sec): 28.72 - samples/sec: 2562.39 - lr: 0.000026 - momentum: 0.000000
2023-10-25 18:24:25,614 epoch 3 - iter 712/1786 - loss 0.07238654 - time (sec): 38.35 - samples/sec: 2574.30 - lr: 0.000025 - momentum: 0.000000
2023-10-25 18:24:35,379 epoch 3 - iter 890/1786 - loss 0.07251223 - time (sec): 48.12 - samples/sec: 2571.82 - lr: 0.000025 - momentum: 0.000000
2023-10-25 18:24:45,040 epoch 3 - iter 1068/1786 - loss 0.07322538 - time (sec): 57.78 - samples/sec: 2580.46 - lr: 0.000025 - momentum: 0.000000
2023-10-25 18:24:54,249 epoch 3 - iter 1246/1786 - loss 0.07152561 - time (sec): 66.98 - samples/sec: 2609.81 - lr: 0.000024 - momentum: 0.000000
2023-10-25 18:25:03,582 epoch 3 - iter 1424/1786 - loss 0.07161969 - time (sec): 76.32 - samples/sec: 2579.33 - lr: 0.000024 - momentum: 0.000000
2023-10-25 18:25:12,455 epoch 3 - iter 1602/1786 - loss 0.07144466 - time (sec): 85.19 - samples/sec: 2609.06 - lr: 0.000024 - momentum: 0.000000
2023-10-25 18:25:21,964 epoch 3 - iter 1780/1786 - loss 0.07137444 - time (sec): 94.70 - samples/sec: 2618.28 - lr: 0.000023 - momentum: 0.000000
2023-10-25 18:25:22,284 ----------------------------------------------------------------------------------------------------
2023-10-25 18:25:22,284 EPOCH 3 done: loss 0.0713 - lr: 0.000023
2023-10-25 18:25:27,382 DEV : loss 0.13084866106510162 - f1-score (micro avg) 0.7918
2023-10-25 18:25:27,404 saving best model
2023-10-25 18:25:28,076 ----------------------------------------------------------------------------------------------------
2023-10-25 18:25:36,796 epoch 4 - iter 178/1786 - loss 0.04603763 - time (sec): 8.72 - samples/sec: 2822.79 - lr: 0.000023 - momentum: 0.000000
2023-10-25 18:25:45,700 epoch 4 - iter 356/1786 - loss 0.04595061 - time (sec): 17.62 - samples/sec: 2858.45 - lr: 0.000023 - momentum: 0.000000
2023-10-25 18:25:54,721 epoch 4 - iter 534/1786 - loss 0.05049014 - time (sec): 26.64 - samples/sec: 2796.43 - lr: 0.000022 - momentum: 0.000000
2023-10-25 18:26:04,026 epoch 4 - iter 712/1786 - loss 0.05319326 - time (sec): 35.95 - samples/sec: 2750.89 - lr: 0.000022 - momentum: 0.000000
2023-10-25 18:26:13,466 epoch 4 - iter 890/1786 - loss 0.05285730 - time (sec): 45.39 - samples/sec: 2713.26 - lr: 0.000022 - momentum: 0.000000
2023-10-25 18:26:22,918 epoch 4 - iter 1068/1786 - loss 0.05374856 - time (sec): 54.84 - samples/sec: 2711.40 - lr: 0.000021 - momentum: 0.000000
2023-10-25 18:26:32,404 epoch 4 - iter 1246/1786 - loss 0.05460603 - time (sec): 64.33 - samples/sec: 2687.26 - lr: 0.000021 - momentum: 0.000000
2023-10-25 18:26:42,081 epoch 4 - iter 1424/1786 - loss 0.05410043 - time (sec): 74.00 - samples/sec: 2681.28 - lr: 0.000021 - momentum: 0.000000
2023-10-25 18:26:51,822 epoch 4 - iter 1602/1786 - loss 0.05349229 - time (sec): 83.74 - samples/sec: 2668.08 - lr: 0.000020 - momentum: 0.000000
2023-10-25 18:27:01,217 epoch 4 - iter 1780/1786 - loss 0.05316503 - time (sec): 93.14 - samples/sec: 2663.63 - lr: 0.000020 - momentum: 0.000000
2023-10-25 18:27:01,526 ----------------------------------------------------------------------------------------------------
2023-10-25 18:27:01,527 EPOCH 4 done: loss 0.0532 - lr: 0.000020
2023-10-25 18:27:06,049 DEV : loss 0.16789670288562775 - f1-score (micro avg) 0.7829
2023-10-25 18:27:06,070 ----------------------------------------------------------------------------------------------------
2023-10-25 18:27:15,697 epoch 5 - iter 178/1786 - loss 0.05389475 - time (sec): 9.63 - samples/sec: 2455.06 - lr: 0.000020 - momentum: 0.000000
2023-10-25 18:27:25,190 epoch 5 - iter 356/1786 - loss 0.04260056 - time (sec): 19.12 - samples/sec: 2632.79 - lr: 0.000019 - momentum: 0.000000
2023-10-25 18:27:34,006 epoch 5 - iter 534/1786 - loss 0.04052769 - time (sec): 27.93 - samples/sec: 2688.86 - lr: 0.000019 - momentum: 0.000000
2023-10-25 18:27:43,195 epoch 5 - iter 712/1786 - loss 0.04001294 - time (sec): 37.12 - samples/sec: 2703.06 - lr: 0.000019 - momentum: 0.000000
2023-10-25 18:27:52,669 epoch 5 - iter 890/1786 - loss 0.03975722 - time (sec): 46.60 - samples/sec: 2699.77 - lr: 0.000018 - momentum: 0.000000
2023-10-25 18:28:02,256 epoch 5 - iter 1068/1786 - loss 0.03974050 - time (sec): 56.18 - samples/sec: 2654.81 - lr: 0.000018 - momentum: 0.000000
2023-10-25 18:28:11,497 epoch 5 - iter 1246/1786 - loss 0.03952798 - time (sec): 65.43 - samples/sec: 2666.55 - lr: 0.000018 - momentum: 0.000000
2023-10-25 18:28:20,489 epoch 5 - iter 1424/1786 - loss 0.03903402 - time (sec): 74.42 - samples/sec: 2667.01 - lr: 0.000017 - momentum: 0.000000
2023-10-25 18:28:29,211 epoch 5 - iter 1602/1786 - loss 0.03844957 - time (sec): 83.14 - samples/sec: 2669.18 - lr: 0.000017 - momentum: 0.000000
2023-10-25 18:28:38,265 epoch 5 - iter 1780/1786 - loss 0.03829347 - time (sec): 92.19 - samples/sec: 2687.67 - lr: 0.000017 - momentum: 0.000000
2023-10-25 18:28:38,578 ----------------------------------------------------------------------------------------------------
2023-10-25 18:28:38,578 EPOCH 5 done: loss 0.0382 - lr: 0.000017
2023-10-25 18:28:44,071 DEV : loss 0.19442911446094513 - f1-score (micro avg) 0.7802
2023-10-25 18:28:44,094 ----------------------------------------------------------------------------------------------------
2023-10-25 18:28:53,424 epoch 6 - iter 178/1786 - loss 0.03398055 - time (sec): 9.33 - samples/sec: 2700.33 - lr: 0.000016 - momentum: 0.000000
2023-10-25 18:29:02,610 epoch 6 - iter 356/1786 - loss 0.03316916 - time (sec): 18.51 - samples/sec: 2757.68 - lr: 0.000016 - momentum: 0.000000
2023-10-25 18:29:12,181 epoch 6 - iter 534/1786 - loss 0.03040477 - time (sec): 28.09 - samples/sec: 2648.55 - lr: 0.000016 - momentum: 0.000000
2023-10-25 18:29:21,873 epoch 6 - iter 712/1786 - loss 0.03212632 - time (sec): 37.78 - samples/sec: 2628.07 - lr: 0.000015 - momentum: 0.000000
2023-10-25 18:29:31,591 epoch 6 - iter 890/1786 - loss 0.02992094 - time (sec): 47.50 - samples/sec: 2632.24 - lr: 0.000015 - momentum: 0.000000
2023-10-25 18:29:41,332 epoch 6 - iter 1068/1786 - loss 0.02930104 - time (sec): 57.24 - samples/sec: 2618.93 - lr: 0.000015 - momentum: 0.000000
2023-10-25 18:29:50,819 epoch 6 - iter 1246/1786 - loss 0.02972192 - time (sec): 66.72 - samples/sec: 2622.50 - lr: 0.000014 - momentum: 0.000000
2023-10-25 18:30:00,534 epoch 6 - iter 1424/1786 - loss 0.02944805 - time (sec): 76.44 - samples/sec: 2594.65 - lr: 0.000014 - momentum: 0.000000
2023-10-25 18:30:09,882 epoch 6 - iter 1602/1786 - loss 0.02958451 - time (sec): 85.79 - samples/sec: 2626.36 - lr: 0.000014 - momentum: 0.000000
2023-10-25 18:30:18,857 epoch 6 - iter 1780/1786 - loss 0.03003769 - time (sec): 94.76 - samples/sec: 2618.30 - lr: 0.000013 - momentum: 0.000000
2023-10-25 18:30:19,162 ----------------------------------------------------------------------------------------------------
2023-10-25 18:30:19,163 EPOCH 6 done: loss 0.0300 - lr: 0.000013
2023-10-25 18:30:23,457 DEV : loss 0.18333885073661804 - f1-score (micro avg) 0.7925
2023-10-25 18:30:23,480 saving best model
2023-10-25 18:30:24,184 ----------------------------------------------------------------------------------------------------
2023-10-25 18:30:34,775 epoch 7 - iter 178/1786 - loss 0.02162460 - time (sec): 10.59 - samples/sec: 2448.54 - lr: 0.000013 - momentum: 0.000000
2023-10-25 18:30:44,255 epoch 7 - iter 356/1786 - loss 0.01908762 - time (sec): 20.07 - samples/sec: 2453.18 - lr: 0.000013 - momentum: 0.000000
2023-10-25 18:30:53,682 epoch 7 - iter 534/1786 - loss 0.01976296 - time (sec): 29.50 - samples/sec: 2516.85 - lr: 0.000012 - momentum: 0.000000
2023-10-25 18:31:02,712 epoch 7 - iter 712/1786 - loss 0.02174272 - time (sec): 38.53 - samples/sec: 2564.31 - lr: 0.000012 - momentum: 0.000000
2023-10-25 18:31:11,779 epoch 7 - iter 890/1786 - loss 0.02278711 - time (sec): 47.59 - samples/sec: 2635.47 - lr: 0.000012 - momentum: 0.000000
2023-10-25 18:31:20,900 epoch 7 - iter 1068/1786 - loss 0.02154762 - time (sec): 56.71 - samples/sec: 2659.10 - lr: 0.000011 - momentum: 0.000000
2023-10-25 18:31:29,668 epoch 7 - iter 1246/1786 - loss 0.02092074 - time (sec): 65.48 - samples/sec: 2699.75 - lr: 0.000011 - momentum: 0.000000
2023-10-25 18:31:38,781 epoch 7 - iter 1424/1786 - loss 0.02100336 - time (sec): 74.60 - samples/sec: 2665.33 - lr: 0.000011 - momentum: 0.000000
2023-10-25 18:31:48,049 epoch 7 - iter 1602/1786 - loss 0.02189701 - time (sec): 83.86 - samples/sec: 2662.93 - lr: 0.000010 - momentum: 0.000000
2023-10-25 18:31:57,161 epoch 7 - iter 1780/1786 - loss 0.02145412 - time (sec): 92.98 - samples/sec: 2665.12 - lr: 0.000010 - momentum: 0.000000
2023-10-25 18:31:57,481 ----------------------------------------------------------------------------------------------------
2023-10-25 18:31:57,481 EPOCH 7 done: loss 0.0214 - lr: 0.000010
2023-10-25 18:32:01,934 DEV : loss 0.19056054949760437 - f1-score (micro avg) 0.8075
2023-10-25 18:32:01,956 saving best model
2023-10-25 18:32:02,611 ----------------------------------------------------------------------------------------------------
2023-10-25 18:32:12,182 epoch 8 - iter 178/1786 - loss 0.02612736 - time (sec): 9.57 - samples/sec: 2588.30 - lr: 0.000010 - momentum: 0.000000
2023-10-25 18:32:21,872 epoch 8 - iter 356/1786 - loss 0.01894543 - time (sec): 19.26 - samples/sec: 2516.13 - lr: 0.000009 - momentum: 0.000000
2023-10-25 18:32:31,455 epoch 8 - iter 534/1786 - loss 0.01672658 - time (sec): 28.84 - samples/sec: 2550.32 - lr: 0.000009 - momentum: 0.000000
2023-10-25 18:32:40,981 epoch 8 - iter 712/1786 - loss 0.01545569 - time (sec): 38.37 - samples/sec: 2583.47 - lr: 0.000009 - momentum: 0.000000
2023-10-25 18:32:50,491 epoch 8 - iter 890/1786 - loss 0.01420894 - time (sec): 47.88 - samples/sec: 2582.39 - lr: 0.000008 - momentum: 0.000000
2023-10-25 18:32:59,910 epoch 8 - iter 1068/1786 - loss 0.01478084 - time (sec): 57.30 - samples/sec: 2565.98 - lr: 0.000008 - momentum: 0.000000
2023-10-25 18:33:09,307 epoch 8 - iter 1246/1786 - loss 0.01595451 - time (sec): 66.69 - samples/sec: 2572.39 - lr: 0.000008 - momentum: 0.000000
2023-10-25 18:33:18,275 epoch 8 - iter 1424/1786 - loss 0.01587470 - time (sec): 75.66 - samples/sec: 2598.46 - lr: 0.000007 - momentum: 0.000000
2023-10-25 18:33:27,281 epoch 8 - iter 1602/1786 - loss 0.01586368 - time (sec): 84.67 - samples/sec: 2636.76 - lr: 0.000007 - momentum: 0.000000
2023-10-25 18:33:36,357 epoch 8 - iter 1780/1786 - loss 0.01584675 - time (sec): 93.74 - samples/sec: 2645.49 - lr: 0.000007 - momentum: 0.000000
2023-10-25 18:33:36,660 ----------------------------------------------------------------------------------------------------
2023-10-25 18:33:36,660 EPOCH 8 done: loss 0.0158 - lr: 0.000007
2023-10-25 18:33:42,265 DEV : loss 0.19766351580619812 - f1-score (micro avg) 0.8038
2023-10-25 18:33:42,287 ----------------------------------------------------------------------------------------------------
2023-10-25 18:33:51,969 epoch 9 - iter 178/1786 - loss 0.01028428 - time (sec): 9.68 - samples/sec: 2633.27 - lr: 0.000006 - momentum: 0.000000
2023-10-25 18:34:01,084 epoch 9 - iter 356/1786 - loss 0.00955984 - time (sec): 18.80 - samples/sec: 2571.59 - lr: 0.000006 - momentum: 0.000000
2023-10-25 18:34:10,368 epoch 9 - iter 534/1786 - loss 0.01151104 - time (sec): 28.08 - samples/sec: 2581.54 - lr: 0.000006 - momentum: 0.000000
2023-10-25 18:34:19,236 epoch 9 - iter 712/1786 - loss 0.01207848 - time (sec): 36.95 - samples/sec: 2625.24 - lr: 0.000005 - momentum: 0.000000
2023-10-25 18:34:28,301 epoch 9 - iter 890/1786 - loss 0.01138794 - time (sec): 46.01 - samples/sec: 2612.46 - lr: 0.000005 - momentum: 0.000000
2023-10-25 18:34:37,234 epoch 9 - iter 1068/1786 - loss 0.01116143 - time (sec): 54.95 - samples/sec: 2672.29 - lr: 0.000005 - momentum: 0.000000
2023-10-25 18:34:46,369 epoch 9 - iter 1246/1786 - loss 0.01155486 - time (sec): 64.08 - samples/sec: 2669.89 - lr: 0.000004 - momentum: 0.000000
2023-10-25 18:34:55,509 epoch 9 - iter 1424/1786 - loss 0.01114669 - time (sec): 73.22 - samples/sec: 2685.06 - lr: 0.000004 - momentum: 0.000000
2023-10-25 18:35:04,412 epoch 9 - iter 1602/1786 - loss 0.01110376 - time (sec): 82.12 - samples/sec: 2697.08 - lr: 0.000004 - momentum: 0.000000
2023-10-25 18:35:13,124 epoch 9 - iter 1780/1786 - loss 0.01074968 - time (sec): 90.84 - samples/sec: 2731.45 - lr: 0.000003 - momentum: 0.000000
2023-10-25 18:35:13,422 ----------------------------------------------------------------------------------------------------
2023-10-25 18:35:13,423 EPOCH 9 done: loss 0.0108 - lr: 0.000003
2023-10-25 18:35:17,832 DEV : loss 0.21357937157154083 - f1-score (micro avg) 0.8104
2023-10-25 18:35:17,855 saving best model
2023-10-25 18:35:18,543 ----------------------------------------------------------------------------------------------------
2023-10-25 18:35:28,085 epoch 10 - iter 178/1786 - loss 0.00637564 - time (sec): 9.54 - samples/sec: 2632.83 - lr: 0.000003 - momentum: 0.000000
2023-10-25 18:35:37,383 epoch 10 - iter 356/1786 - loss 0.00589645 - time (sec): 18.84 - samples/sec: 2586.16 - lr: 0.000003 - momentum: 0.000000
2023-10-25 18:35:46,529 epoch 10 - iter 534/1786 - loss 0.00533950 - time (sec): 27.98 - samples/sec: 2649.13 - lr: 0.000002 - momentum: 0.000000
2023-10-25 18:35:55,356 epoch 10 - iter 712/1786 - loss 0.00617715 - time (sec): 36.81 - samples/sec: 2702.22 - lr: 0.000002 - momentum: 0.000000
2023-10-25 18:36:04,322 epoch 10 - iter 890/1786 - loss 0.00620859 - time (sec): 45.78 - samples/sec: 2721.41 - lr: 0.000002 - momentum: 0.000000
2023-10-25 18:36:13,367 epoch 10 - iter 1068/1786 - loss 0.00641849 - time (sec): 54.82 - samples/sec: 2715.20 - lr: 0.000001 - momentum: 0.000000
2023-10-25 18:36:22,629 epoch 10 - iter 1246/1786 - loss 0.00676503 - time (sec): 64.08 - samples/sec: 2702.32 - lr: 0.000001 - momentum: 0.000000
2023-10-25 18:36:32,265 epoch 10 - iter 1424/1786 - loss 0.00651546 - time (sec): 73.72 - samples/sec: 2700.44 - lr: 0.000001 - momentum: 0.000000
2023-10-25 18:36:41,692 epoch 10 - iter 1602/1786 - loss 0.00660118 - time (sec): 83.15 - samples/sec: 2690.29 - lr: 0.000000 - momentum: 0.000000
2023-10-25 18:36:50,765 epoch 10 - iter 1780/1786 - loss 0.00682484 - time (sec): 92.22 - samples/sec: 2689.05 - lr: 0.000000 - momentum: 0.000000
2023-10-25 18:36:51,073 ----------------------------------------------------------------------------------------------------
2023-10-25 18:36:51,074 EPOCH 10 done: loss 0.0068 - lr: 0.000000
2023-10-25 18:36:56,195 DEV : loss 0.21570290625095367 - f1-score (micro avg) 0.8096
2023-10-25 18:36:56,710 ----------------------------------------------------------------------------------------------------
2023-10-25 18:36:56,712 Loading model from best epoch ...
2023-10-25 18:36:58,657 SequenceTagger predicts: Dictionary with 17 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, S-HumanProd, B-HumanProd, E-HumanProd, I-HumanProd
2023-10-25 18:37:13,089
Results:
- F-score (micro) 0.6996
- F-score (macro) 0.6261
- Accuracy 0.5539
By class:
precision recall f1-score support
LOC 0.6930 0.7050 0.6990 1095
PER 0.7827 0.7796 0.7812 1012
ORG 0.4655 0.5854 0.5186 357
HumanProd 0.3966 0.6970 0.5055 33
micro avg 0.6820 0.7181 0.6996 2497
macro avg 0.5844 0.6918 0.6261 2497
weighted avg 0.6929 0.7181 0.7039 2497
2023-10-25 18:37:13,089 ----------------------------------------------------------------------------------------------------
|