File size: 24,231 Bytes
6442547 |
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 |
2023-10-17 11:07:44,037 ----------------------------------------------------------------------------------------------------
2023-10-17 11:07:44,040 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=17, bias=True)
(loss_function): CrossEntropyLoss()
)"
2023-10-17 11:07:44,040 ----------------------------------------------------------------------------------------------------
2023-10-17 11:07:44,040 MultiCorpus: 20847 train + 1123 dev + 3350 test sentences
- NER_HIPE_2022 Corpus: 20847 train + 1123 dev + 3350 test sentences - /root/.flair/datasets/ner_hipe_2022/v2.1/newseye/de/with_doc_seperator
2023-10-17 11:07:44,040 ----------------------------------------------------------------------------------------------------
2023-10-17 11:07:44,041 Train: 20847 sentences
2023-10-17 11:07:44,041 (train_with_dev=False, train_with_test=False)
2023-10-17 11:07:44,041 ----------------------------------------------------------------------------------------------------
2023-10-17 11:07:44,041 Training Params:
2023-10-17 11:07:44,041 - learning_rate: "3e-05"
2023-10-17 11:07:44,041 - mini_batch_size: "4"
2023-10-17 11:07:44,041 - max_epochs: "10"
2023-10-17 11:07:44,041 - shuffle: "True"
2023-10-17 11:07:44,041 ----------------------------------------------------------------------------------------------------
2023-10-17 11:07:44,041 Plugins:
2023-10-17 11:07:44,041 - TensorboardLogger
2023-10-17 11:07:44,042 - LinearScheduler | warmup_fraction: '0.1'
2023-10-17 11:07:44,042 ----------------------------------------------------------------------------------------------------
2023-10-17 11:07:44,042 Final evaluation on model from best epoch (best-model.pt)
2023-10-17 11:07:44,042 - metric: "('micro avg', 'f1-score')"
2023-10-17 11:07:44,042 ----------------------------------------------------------------------------------------------------
2023-10-17 11:07:44,042 Computation:
2023-10-17 11:07:44,042 - compute on device: cuda:0
2023-10-17 11:07:44,042 - embedding storage: none
2023-10-17 11:07:44,042 ----------------------------------------------------------------------------------------------------
2023-10-17 11:07:44,042 Model training base path: "hmbench-newseye/de-hmteams/teams-base-historic-multilingual-discriminator-bs4-wsFalse-e10-lr3e-05-poolingfirst-layers-1-crfFalse-1"
2023-10-17 11:07:44,042 ----------------------------------------------------------------------------------------------------
2023-10-17 11:07:44,043 ----------------------------------------------------------------------------------------------------
2023-10-17 11:07:44,043 Logging anything other than scalars to TensorBoard is currently not supported.
2023-10-17 11:08:26,363 epoch 1 - iter 521/5212 - loss 1.89163776 - time (sec): 42.32 - samples/sec: 798.01 - lr: 0.000003 - momentum: 0.000000
2023-10-17 11:09:08,855 epoch 1 - iter 1042/5212 - loss 1.13479704 - time (sec): 84.81 - samples/sec: 817.76 - lr: 0.000006 - momentum: 0.000000
2023-10-17 11:09:54,477 epoch 1 - iter 1563/5212 - loss 0.84671118 - time (sec): 130.43 - samples/sec: 826.16 - lr: 0.000009 - momentum: 0.000000
2023-10-17 11:10:37,574 epoch 1 - iter 2084/5212 - loss 0.69980888 - time (sec): 173.53 - samples/sec: 838.22 - lr: 0.000012 - momentum: 0.000000
2023-10-17 11:11:21,695 epoch 1 - iter 2605/5212 - loss 0.60610850 - time (sec): 217.65 - samples/sec: 848.09 - lr: 0.000015 - momentum: 0.000000
2023-10-17 11:12:04,464 epoch 1 - iter 3126/5212 - loss 0.54043037 - time (sec): 260.42 - samples/sec: 857.81 - lr: 0.000018 - momentum: 0.000000
2023-10-17 11:12:47,850 epoch 1 - iter 3647/5212 - loss 0.49949347 - time (sec): 303.81 - samples/sec: 851.51 - lr: 0.000021 - momentum: 0.000000
2023-10-17 11:13:31,805 epoch 1 - iter 4168/5212 - loss 0.46823322 - time (sec): 347.76 - samples/sec: 842.57 - lr: 0.000024 - momentum: 0.000000
2023-10-17 11:14:16,895 epoch 1 - iter 4689/5212 - loss 0.43915469 - time (sec): 392.85 - samples/sec: 842.42 - lr: 0.000027 - momentum: 0.000000
2023-10-17 11:15:00,022 epoch 1 - iter 5210/5212 - loss 0.41555785 - time (sec): 435.98 - samples/sec: 842.27 - lr: 0.000030 - momentum: 0.000000
2023-10-17 11:15:00,190 ----------------------------------------------------------------------------------------------------
2023-10-17 11:15:00,190 EPOCH 1 done: loss 0.4153 - lr: 0.000030
2023-10-17 11:15:07,630 DEV : loss 0.11844930797815323 - f1-score (micro avg) 0.2469
2023-10-17 11:15:07,684 saving best model
2023-10-17 11:15:08,240 ----------------------------------------------------------------------------------------------------
2023-10-17 11:15:51,077 epoch 2 - iter 521/5212 - loss 0.18677525 - time (sec): 42.84 - samples/sec: 893.85 - lr: 0.000030 - momentum: 0.000000
2023-10-17 11:16:34,185 epoch 2 - iter 1042/5212 - loss 0.18467584 - time (sec): 85.94 - samples/sec: 867.54 - lr: 0.000029 - momentum: 0.000000
2023-10-17 11:17:17,438 epoch 2 - iter 1563/5212 - loss 0.18459225 - time (sec): 129.20 - samples/sec: 868.65 - lr: 0.000029 - momentum: 0.000000
2023-10-17 11:18:00,970 epoch 2 - iter 2084/5212 - loss 0.18949291 - time (sec): 172.73 - samples/sec: 854.13 - lr: 0.000029 - momentum: 0.000000
2023-10-17 11:18:45,977 epoch 2 - iter 2605/5212 - loss 0.18960565 - time (sec): 217.74 - samples/sec: 842.99 - lr: 0.000028 - momentum: 0.000000
2023-10-17 11:19:29,280 epoch 2 - iter 3126/5212 - loss 0.18834557 - time (sec): 261.04 - samples/sec: 840.64 - lr: 0.000028 - momentum: 0.000000
2023-10-17 11:20:11,948 epoch 2 - iter 3647/5212 - loss 0.18442261 - time (sec): 303.71 - samples/sec: 854.40 - lr: 0.000028 - momentum: 0.000000
2023-10-17 11:20:55,872 epoch 2 - iter 4168/5212 - loss 0.18307206 - time (sec): 347.63 - samples/sec: 851.42 - lr: 0.000027 - momentum: 0.000000
2023-10-17 11:21:38,580 epoch 2 - iter 4689/5212 - loss 0.17918821 - time (sec): 390.34 - samples/sec: 844.82 - lr: 0.000027 - momentum: 0.000000
2023-10-17 11:22:20,682 epoch 2 - iter 5210/5212 - loss 0.17650046 - time (sec): 432.44 - samples/sec: 849.48 - lr: 0.000027 - momentum: 0.000000
2023-10-17 11:22:20,832 ----------------------------------------------------------------------------------------------------
2023-10-17 11:22:20,833 EPOCH 2 done: loss 0.1765 - lr: 0.000027
2023-10-17 11:22:32,839 DEV : loss 0.23894941806793213 - f1-score (micro avg) 0.3469
2023-10-17 11:22:32,893 saving best model
2023-10-17 11:22:34,316 ----------------------------------------------------------------------------------------------------
2023-10-17 11:23:15,325 epoch 3 - iter 521/5212 - loss 0.11437386 - time (sec): 41.01 - samples/sec: 925.03 - lr: 0.000026 - momentum: 0.000000
2023-10-17 11:23:57,365 epoch 3 - iter 1042/5212 - loss 0.12459624 - time (sec): 83.04 - samples/sec: 902.84 - lr: 0.000026 - momentum: 0.000000
2023-10-17 11:24:39,204 epoch 3 - iter 1563/5212 - loss 0.12923542 - time (sec): 124.88 - samples/sec: 889.63 - lr: 0.000026 - momentum: 0.000000
2023-10-17 11:25:20,547 epoch 3 - iter 2084/5212 - loss 0.13146091 - time (sec): 166.23 - samples/sec: 882.27 - lr: 0.000025 - momentum: 0.000000
2023-10-17 11:26:01,412 epoch 3 - iter 2605/5212 - loss 0.12864210 - time (sec): 207.09 - samples/sec: 886.34 - lr: 0.000025 - momentum: 0.000000
2023-10-17 11:26:42,942 epoch 3 - iter 3126/5212 - loss 0.13399268 - time (sec): 248.62 - samples/sec: 875.92 - lr: 0.000025 - momentum: 0.000000
2023-10-17 11:27:25,033 epoch 3 - iter 3647/5212 - loss 0.13321294 - time (sec): 290.71 - samples/sec: 874.05 - lr: 0.000024 - momentum: 0.000000
2023-10-17 11:28:06,996 epoch 3 - iter 4168/5212 - loss 0.13283195 - time (sec): 332.68 - samples/sec: 874.06 - lr: 0.000024 - momentum: 0.000000
2023-10-17 11:28:50,011 epoch 3 - iter 4689/5212 - loss 0.13502035 - time (sec): 375.69 - samples/sec: 878.62 - lr: 0.000024 - momentum: 0.000000
2023-10-17 11:29:31,516 epoch 3 - iter 5210/5212 - loss 0.13202538 - time (sec): 417.20 - samples/sec: 880.09 - lr: 0.000023 - momentum: 0.000000
2023-10-17 11:29:31,671 ----------------------------------------------------------------------------------------------------
2023-10-17 11:29:31,671 EPOCH 3 done: loss 0.1319 - lr: 0.000023
2023-10-17 11:29:43,652 DEV : loss 0.24874247610569 - f1-score (micro avg) 0.351
2023-10-17 11:29:43,706 saving best model
2023-10-17 11:29:45,126 ----------------------------------------------------------------------------------------------------
2023-10-17 11:30:29,111 epoch 4 - iter 521/5212 - loss 0.09645922 - time (sec): 43.98 - samples/sec: 847.79 - lr: 0.000023 - momentum: 0.000000
2023-10-17 11:31:14,228 epoch 4 - iter 1042/5212 - loss 0.09575057 - time (sec): 89.10 - samples/sec: 827.29 - lr: 0.000023 - momentum: 0.000000
2023-10-17 11:31:56,651 epoch 4 - iter 1563/5212 - loss 0.09440810 - time (sec): 131.52 - samples/sec: 826.16 - lr: 0.000022 - momentum: 0.000000
2023-10-17 11:32:38,800 epoch 4 - iter 2084/5212 - loss 0.09200648 - time (sec): 173.67 - samples/sec: 830.24 - lr: 0.000022 - momentum: 0.000000
2023-10-17 11:33:22,254 epoch 4 - iter 2605/5212 - loss 0.09507978 - time (sec): 217.12 - samples/sec: 824.50 - lr: 0.000022 - momentum: 0.000000
2023-10-17 11:34:03,455 epoch 4 - iter 3126/5212 - loss 0.09574149 - time (sec): 258.33 - samples/sec: 827.19 - lr: 0.000021 - momentum: 0.000000
2023-10-17 11:34:45,413 epoch 4 - iter 3647/5212 - loss 0.09612564 - time (sec): 300.28 - samples/sec: 839.08 - lr: 0.000021 - momentum: 0.000000
2023-10-17 11:35:28,560 epoch 4 - iter 4168/5212 - loss 0.09673298 - time (sec): 343.43 - samples/sec: 846.54 - lr: 0.000021 - momentum: 0.000000
2023-10-17 11:36:11,639 epoch 4 - iter 4689/5212 - loss 0.09605759 - time (sec): 386.51 - samples/sec: 853.19 - lr: 0.000020 - momentum: 0.000000
2023-10-17 11:36:54,277 epoch 4 - iter 5210/5212 - loss 0.09435830 - time (sec): 429.15 - samples/sec: 855.98 - lr: 0.000020 - momentum: 0.000000
2023-10-17 11:36:54,441 ----------------------------------------------------------------------------------------------------
2023-10-17 11:36:54,441 EPOCH 4 done: loss 0.0943 - lr: 0.000020
2023-10-17 11:37:06,587 DEV : loss 0.2750011384487152 - f1-score (micro avg) 0.3813
2023-10-17 11:37:06,641 saving best model
2023-10-17 11:37:08,118 ----------------------------------------------------------------------------------------------------
2023-10-17 11:37:52,422 epoch 5 - iter 521/5212 - loss 0.05806410 - time (sec): 44.30 - samples/sec: 850.98 - lr: 0.000020 - momentum: 0.000000
2023-10-17 11:38:35,833 epoch 5 - iter 1042/5212 - loss 0.05768321 - time (sec): 87.71 - samples/sec: 816.42 - lr: 0.000019 - momentum: 0.000000
2023-10-17 11:39:21,545 epoch 5 - iter 1563/5212 - loss 0.06370432 - time (sec): 133.42 - samples/sec: 813.72 - lr: 0.000019 - momentum: 0.000000
2023-10-17 11:40:04,393 epoch 5 - iter 2084/5212 - loss 0.06198607 - time (sec): 176.27 - samples/sec: 810.85 - lr: 0.000019 - momentum: 0.000000
2023-10-17 11:40:50,143 epoch 5 - iter 2605/5212 - loss 0.06397044 - time (sec): 222.02 - samples/sec: 820.57 - lr: 0.000018 - momentum: 0.000000
2023-10-17 11:41:34,615 epoch 5 - iter 3126/5212 - loss 0.06343104 - time (sec): 266.49 - samples/sec: 834.78 - lr: 0.000018 - momentum: 0.000000
2023-10-17 11:42:17,862 epoch 5 - iter 3647/5212 - loss 0.06370732 - time (sec): 309.74 - samples/sec: 834.87 - lr: 0.000018 - momentum: 0.000000
2023-10-17 11:43:00,416 epoch 5 - iter 4168/5212 - loss 0.06361989 - time (sec): 352.29 - samples/sec: 841.76 - lr: 0.000017 - momentum: 0.000000
2023-10-17 11:43:41,207 epoch 5 - iter 4689/5212 - loss 0.06404810 - time (sec): 393.08 - samples/sec: 842.08 - lr: 0.000017 - momentum: 0.000000
2023-10-17 11:44:23,171 epoch 5 - iter 5210/5212 - loss 0.06343811 - time (sec): 435.05 - samples/sec: 844.47 - lr: 0.000017 - momentum: 0.000000
2023-10-17 11:44:23,319 ----------------------------------------------------------------------------------------------------
2023-10-17 11:44:23,320 EPOCH 5 done: loss 0.0635 - lr: 0.000017
2023-10-17 11:44:34,163 DEV : loss 0.34400203824043274 - f1-score (micro avg) 0.3937
2023-10-17 11:44:34,220 saving best model
2023-10-17 11:44:35,623 ----------------------------------------------------------------------------------------------------
2023-10-17 11:45:19,252 epoch 6 - iter 521/5212 - loss 0.05472897 - time (sec): 43.62 - samples/sec: 855.22 - lr: 0.000016 - momentum: 0.000000
2023-10-17 11:46:00,081 epoch 6 - iter 1042/5212 - loss 0.05370031 - time (sec): 84.45 - samples/sec: 855.73 - lr: 0.000016 - momentum: 0.000000
2023-10-17 11:46:41,972 epoch 6 - iter 1563/5212 - loss 0.04710872 - time (sec): 126.34 - samples/sec: 849.29 - lr: 0.000016 - momentum: 0.000000
2023-10-17 11:47:27,063 epoch 6 - iter 2084/5212 - loss 0.04824334 - time (sec): 171.44 - samples/sec: 855.69 - lr: 0.000015 - momentum: 0.000000
2023-10-17 11:48:09,332 epoch 6 - iter 2605/5212 - loss 0.04632066 - time (sec): 213.70 - samples/sec: 868.35 - lr: 0.000015 - momentum: 0.000000
2023-10-17 11:48:50,366 epoch 6 - iter 3126/5212 - loss 0.04608767 - time (sec): 254.74 - samples/sec: 874.91 - lr: 0.000015 - momentum: 0.000000
2023-10-17 11:49:32,284 epoch 6 - iter 3647/5212 - loss 0.04516211 - time (sec): 296.66 - samples/sec: 871.68 - lr: 0.000014 - momentum: 0.000000
2023-10-17 11:50:13,897 epoch 6 - iter 4168/5212 - loss 0.04805567 - time (sec): 338.27 - samples/sec: 869.50 - lr: 0.000014 - momentum: 0.000000
2023-10-17 11:50:56,045 epoch 6 - iter 4689/5212 - loss 0.04758346 - time (sec): 380.42 - samples/sec: 871.32 - lr: 0.000014 - momentum: 0.000000
2023-10-17 11:51:37,913 epoch 6 - iter 5210/5212 - loss 0.04826135 - time (sec): 422.29 - samples/sec: 869.98 - lr: 0.000013 - momentum: 0.000000
2023-10-17 11:51:38,067 ----------------------------------------------------------------------------------------------------
2023-10-17 11:51:38,068 EPOCH 6 done: loss 0.0483 - lr: 0.000013
2023-10-17 11:51:49,240 DEV : loss 0.2987309396266937 - f1-score (micro avg) 0.3914
2023-10-17 11:51:49,296 ----------------------------------------------------------------------------------------------------
2023-10-17 11:52:31,138 epoch 7 - iter 521/5212 - loss 0.03418427 - time (sec): 41.84 - samples/sec: 904.38 - lr: 0.000013 - momentum: 0.000000
2023-10-17 11:53:13,452 epoch 7 - iter 1042/5212 - loss 0.03052773 - time (sec): 84.15 - samples/sec: 895.56 - lr: 0.000013 - momentum: 0.000000
2023-10-17 11:53:57,642 epoch 7 - iter 1563/5212 - loss 0.03183996 - time (sec): 128.34 - samples/sec: 877.20 - lr: 0.000012 - momentum: 0.000000
2023-10-17 11:54:40,075 epoch 7 - iter 2084/5212 - loss 0.03089690 - time (sec): 170.78 - samples/sec: 876.39 - lr: 0.000012 - momentum: 0.000000
2023-10-17 11:55:22,339 epoch 7 - iter 2605/5212 - loss 0.03370678 - time (sec): 213.04 - samples/sec: 867.96 - lr: 0.000012 - momentum: 0.000000
2023-10-17 11:56:03,952 epoch 7 - iter 3126/5212 - loss 0.03259067 - time (sec): 254.65 - samples/sec: 864.70 - lr: 0.000011 - momentum: 0.000000
2023-10-17 11:56:47,386 epoch 7 - iter 3647/5212 - loss 0.03324339 - time (sec): 298.09 - samples/sec: 858.90 - lr: 0.000011 - momentum: 0.000000
2023-10-17 11:57:31,079 epoch 7 - iter 4168/5212 - loss 0.03204679 - time (sec): 341.78 - samples/sec: 865.26 - lr: 0.000011 - momentum: 0.000000
2023-10-17 11:58:13,042 epoch 7 - iter 4689/5212 - loss 0.03270349 - time (sec): 383.74 - samples/sec: 866.76 - lr: 0.000010 - momentum: 0.000000
2023-10-17 11:58:56,293 epoch 7 - iter 5210/5212 - loss 0.03210456 - time (sec): 426.99 - samples/sec: 860.40 - lr: 0.000010 - momentum: 0.000000
2023-10-17 11:58:56,458 ----------------------------------------------------------------------------------------------------
2023-10-17 11:58:56,458 EPOCH 7 done: loss 0.0321 - lr: 0.000010
2023-10-17 11:59:07,929 DEV : loss 0.4514279067516327 - f1-score (micro avg) 0.3873
2023-10-17 11:59:08,000 ----------------------------------------------------------------------------------------------------
2023-10-17 11:59:50,456 epoch 8 - iter 521/5212 - loss 0.02287542 - time (sec): 42.45 - samples/sec: 845.18 - lr: 0.000010 - momentum: 0.000000
2023-10-17 12:00:35,464 epoch 8 - iter 1042/5212 - loss 0.02032053 - time (sec): 87.46 - samples/sec: 818.29 - lr: 0.000009 - momentum: 0.000000
2023-10-17 12:01:18,589 epoch 8 - iter 1563/5212 - loss 0.02247712 - time (sec): 130.59 - samples/sec: 814.40 - lr: 0.000009 - momentum: 0.000000
2023-10-17 12:02:01,597 epoch 8 - iter 2084/5212 - loss 0.02124931 - time (sec): 173.59 - samples/sec: 821.16 - lr: 0.000009 - momentum: 0.000000
2023-10-17 12:02:44,007 epoch 8 - iter 2605/5212 - loss 0.02199294 - time (sec): 216.00 - samples/sec: 826.62 - lr: 0.000008 - momentum: 0.000000
2023-10-17 12:03:28,062 epoch 8 - iter 3126/5212 - loss 0.02206727 - time (sec): 260.06 - samples/sec: 829.98 - lr: 0.000008 - momentum: 0.000000
2023-10-17 12:04:11,244 epoch 8 - iter 3647/5212 - loss 0.02198914 - time (sec): 303.24 - samples/sec: 835.71 - lr: 0.000008 - momentum: 0.000000
2023-10-17 12:04:53,240 epoch 8 - iter 4168/5212 - loss 0.02133287 - time (sec): 345.24 - samples/sec: 846.62 - lr: 0.000007 - momentum: 0.000000
2023-10-17 12:05:35,938 epoch 8 - iter 4689/5212 - loss 0.02089455 - time (sec): 387.93 - samples/sec: 851.73 - lr: 0.000007 - momentum: 0.000000
2023-10-17 12:06:18,629 epoch 8 - iter 5210/5212 - loss 0.02139060 - time (sec): 430.63 - samples/sec: 853.09 - lr: 0.000007 - momentum: 0.000000
2023-10-17 12:06:18,786 ----------------------------------------------------------------------------------------------------
2023-10-17 12:06:18,787 EPOCH 8 done: loss 0.0214 - lr: 0.000007
2023-10-17 12:06:31,446 DEV : loss 0.42482689023017883 - f1-score (micro avg) 0.4045
2023-10-17 12:06:31,520 saving best model
2023-10-17 12:06:33,030 ----------------------------------------------------------------------------------------------------
2023-10-17 12:07:14,963 epoch 9 - iter 521/5212 - loss 0.01446293 - time (sec): 41.93 - samples/sec: 811.46 - lr: 0.000006 - momentum: 0.000000
2023-10-17 12:07:57,416 epoch 9 - iter 1042/5212 - loss 0.01733620 - time (sec): 84.38 - samples/sec: 845.43 - lr: 0.000006 - momentum: 0.000000
2023-10-17 12:08:38,808 epoch 9 - iter 1563/5212 - loss 0.01496632 - time (sec): 125.78 - samples/sec: 836.07 - lr: 0.000006 - momentum: 0.000000
2023-10-17 12:09:20,895 epoch 9 - iter 2084/5212 - loss 0.01584419 - time (sec): 167.86 - samples/sec: 834.70 - lr: 0.000005 - momentum: 0.000000
2023-10-17 12:10:02,618 epoch 9 - iter 2605/5212 - loss 0.01574388 - time (sec): 209.59 - samples/sec: 843.01 - lr: 0.000005 - momentum: 0.000000
2023-10-17 12:10:44,627 epoch 9 - iter 3126/5212 - loss 0.01792002 - time (sec): 251.59 - samples/sec: 847.62 - lr: 0.000005 - momentum: 0.000000
2023-10-17 12:11:27,348 epoch 9 - iter 3647/5212 - loss 0.01713778 - time (sec): 294.32 - samples/sec: 851.35 - lr: 0.000004 - momentum: 0.000000
2023-10-17 12:12:10,707 epoch 9 - iter 4168/5212 - loss 0.01688059 - time (sec): 337.68 - samples/sec: 855.37 - lr: 0.000004 - momentum: 0.000000
2023-10-17 12:12:53,136 epoch 9 - iter 4689/5212 - loss 0.01654568 - time (sec): 380.10 - samples/sec: 860.95 - lr: 0.000004 - momentum: 0.000000
2023-10-17 12:13:36,164 epoch 9 - iter 5210/5212 - loss 0.01654318 - time (sec): 423.13 - samples/sec: 868.21 - lr: 0.000003 - momentum: 0.000000
2023-10-17 12:13:36,312 ----------------------------------------------------------------------------------------------------
2023-10-17 12:13:36,313 EPOCH 9 done: loss 0.0165 - lr: 0.000003
2023-10-17 12:13:48,791 DEV : loss 0.41042855381965637 - f1-score (micro avg) 0.416
2023-10-17 12:13:48,867 saving best model
2023-10-17 12:13:50,323 ----------------------------------------------------------------------------------------------------
2023-10-17 12:14:33,294 epoch 10 - iter 521/5212 - loss 0.00781776 - time (sec): 42.96 - samples/sec: 865.88 - lr: 0.000003 - momentum: 0.000000
2023-10-17 12:15:15,785 epoch 10 - iter 1042/5212 - loss 0.00793401 - time (sec): 85.46 - samples/sec: 852.16 - lr: 0.000003 - momentum: 0.000000
2023-10-17 12:15:58,787 epoch 10 - iter 1563/5212 - loss 0.00880507 - time (sec): 128.46 - samples/sec: 831.97 - lr: 0.000002 - momentum: 0.000000
2023-10-17 12:16:41,351 epoch 10 - iter 2084/5212 - loss 0.00972971 - time (sec): 171.02 - samples/sec: 843.50 - lr: 0.000002 - momentum: 0.000000
2023-10-17 12:17:24,604 epoch 10 - iter 2605/5212 - loss 0.01024931 - time (sec): 214.27 - samples/sec: 848.53 - lr: 0.000002 - momentum: 0.000000
2023-10-17 12:18:05,735 epoch 10 - iter 3126/5212 - loss 0.00986422 - time (sec): 255.41 - samples/sec: 849.94 - lr: 0.000001 - momentum: 0.000000
2023-10-17 12:18:49,745 epoch 10 - iter 3647/5212 - loss 0.00950244 - time (sec): 299.42 - samples/sec: 845.97 - lr: 0.000001 - momentum: 0.000000
2023-10-17 12:19:32,787 epoch 10 - iter 4168/5212 - loss 0.00935274 - time (sec): 342.46 - samples/sec: 847.08 - lr: 0.000001 - momentum: 0.000000
2023-10-17 12:20:15,120 epoch 10 - iter 4689/5212 - loss 0.00939995 - time (sec): 384.79 - samples/sec: 855.40 - lr: 0.000000 - momentum: 0.000000
2023-10-17 12:20:57,845 epoch 10 - iter 5210/5212 - loss 0.00928220 - time (sec): 427.52 - samples/sec: 859.13 - lr: 0.000000 - momentum: 0.000000
2023-10-17 12:20:58,002 ----------------------------------------------------------------------------------------------------
2023-10-17 12:20:58,002 EPOCH 10 done: loss 0.0093 - lr: 0.000000
2023-10-17 12:21:11,081 DEV : loss 0.4939973056316376 - f1-score (micro avg) 0.3985
2023-10-17 12:21:11,724 ----------------------------------------------------------------------------------------------------
2023-10-17 12:21:11,727 Loading model from best epoch ...
2023-10-17 12:21:14,296 SequenceTagger predicts: Dictionary with 17 tags: O, S-LOC, B-LOC, E-LOC, I-LOC, S-PER, B-PER, E-PER, I-PER, S-ORG, B-ORG, E-ORG, I-ORG, S-HumanProd, B-HumanProd, E-HumanProd, I-HumanProd
2023-10-17 12:21:35,379
Results:
- F-score (micro) 0.4827
- F-score (macro) 0.3288
- Accuracy 0.3215
By class:
precision recall f1-score support
LOC 0.5341 0.5939 0.5624 1214
PER 0.4248 0.4542 0.4390 808
ORG 0.3102 0.3173 0.3137 353
HumanProd 0.0000 0.0000 0.0000 15
micro avg 0.4648 0.5021 0.4827 2390
macro avg 0.3173 0.3413 0.3288 2390
weighted avg 0.4607 0.5021 0.4804 2390
2023-10-17 12:21:35,379 ----------------------------------------------------------------------------------------------------
|