File size: 24,192 Bytes
db0a7ad |
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 |
2023-10-17 20:46:43,325 ----------------------------------------------------------------------------------------------------
2023-10-17 20:46:43,326 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=21, bias=True)
(loss_function): CrossEntropyLoss()
)"
2023-10-17 20:46:43,329 ----------------------------------------------------------------------------------------------------
2023-10-17 20:46:43,329 MultiCorpus: 5901 train + 1287 dev + 1505 test sentences
- NER_HIPE_2022 Corpus: 5901 train + 1287 dev + 1505 test sentences - /root/.flair/datasets/ner_hipe_2022/v2.1/hipe2020/fr/with_doc_seperator
2023-10-17 20:46:43,329 ----------------------------------------------------------------------------------------------------
2023-10-17 20:46:43,329 Train: 5901 sentences
2023-10-17 20:46:43,329 (train_with_dev=False, train_with_test=False)
2023-10-17 20:46:43,329 ----------------------------------------------------------------------------------------------------
2023-10-17 20:46:43,329 Training Params:
2023-10-17 20:46:43,329 - learning_rate: "5e-05"
2023-10-17 20:46:43,329 - mini_batch_size: "8"
2023-10-17 20:46:43,329 - max_epochs: "10"
2023-10-17 20:46:43,329 - shuffle: "True"
2023-10-17 20:46:43,329 ----------------------------------------------------------------------------------------------------
2023-10-17 20:46:43,329 Plugins:
2023-10-17 20:46:43,329 - TensorboardLogger
2023-10-17 20:46:43,329 - LinearScheduler | warmup_fraction: '0.1'
2023-10-17 20:46:43,329 ----------------------------------------------------------------------------------------------------
2023-10-17 20:46:43,329 Final evaluation on model from best epoch (best-model.pt)
2023-10-17 20:46:43,329 - metric: "('micro avg', 'f1-score')"
2023-10-17 20:46:43,329 ----------------------------------------------------------------------------------------------------
2023-10-17 20:46:43,329 Computation:
2023-10-17 20:46:43,329 - compute on device: cuda:0
2023-10-17 20:46:43,329 - embedding storage: none
2023-10-17 20:46:43,329 ----------------------------------------------------------------------------------------------------
2023-10-17 20:46:43,329 Model training base path: "hmbench-hipe2020/fr-hmteams/teams-base-historic-multilingual-discriminator-bs8-wsFalse-e10-lr5e-05-poolingfirst-layers-1-crfFalse-2"
2023-10-17 20:46:43,330 ----------------------------------------------------------------------------------------------------
2023-10-17 20:46:43,330 ----------------------------------------------------------------------------------------------------
2023-10-17 20:46:43,330 Logging anything other than scalars to TensorBoard is currently not supported.
2023-10-17 20:46:48,122 epoch 1 - iter 73/738 - loss 3.27628587 - time (sec): 4.79 - samples/sec: 3348.31 - lr: 0.000005 - momentum: 0.000000
2023-10-17 20:46:53,518 epoch 1 - iter 146/738 - loss 1.88031680 - time (sec): 10.19 - samples/sec: 3434.81 - lr: 0.000010 - momentum: 0.000000
2023-10-17 20:46:58,077 epoch 1 - iter 219/738 - loss 1.45313720 - time (sec): 14.75 - samples/sec: 3382.83 - lr: 0.000015 - momentum: 0.000000
2023-10-17 20:47:03,371 epoch 1 - iter 292/738 - loss 1.18325529 - time (sec): 20.04 - samples/sec: 3327.96 - lr: 0.000020 - momentum: 0.000000
2023-10-17 20:47:08,279 epoch 1 - iter 365/738 - loss 1.01353206 - time (sec): 24.95 - samples/sec: 3318.60 - lr: 0.000025 - momentum: 0.000000
2023-10-17 20:47:13,277 epoch 1 - iter 438/738 - loss 0.89228860 - time (sec): 29.95 - samples/sec: 3300.93 - lr: 0.000030 - momentum: 0.000000
2023-10-17 20:47:18,383 epoch 1 - iter 511/738 - loss 0.79797751 - time (sec): 35.05 - samples/sec: 3292.44 - lr: 0.000035 - momentum: 0.000000
2023-10-17 20:47:22,981 epoch 1 - iter 584/738 - loss 0.72397208 - time (sec): 39.65 - samples/sec: 3309.17 - lr: 0.000039 - momentum: 0.000000
2023-10-17 20:47:28,169 epoch 1 - iter 657/738 - loss 0.66245224 - time (sec): 44.84 - samples/sec: 3301.83 - lr: 0.000044 - momentum: 0.000000
2023-10-17 20:47:33,415 epoch 1 - iter 730/738 - loss 0.61711909 - time (sec): 50.08 - samples/sec: 3273.24 - lr: 0.000049 - momentum: 0.000000
2023-10-17 20:47:34,237 ----------------------------------------------------------------------------------------------------
2023-10-17 20:47:34,237 EPOCH 1 done: loss 0.6106 - lr: 0.000049
2023-10-17 20:47:41,563 DEV : loss 0.1268201768398285 - f1-score (micro avg) 0.7404
2023-10-17 20:47:41,593 saving best model
2023-10-17 20:47:41,993 ----------------------------------------------------------------------------------------------------
2023-10-17 20:47:46,689 epoch 2 - iter 73/738 - loss 0.12468557 - time (sec): 4.69 - samples/sec: 3176.81 - lr: 0.000049 - momentum: 0.000000
2023-10-17 20:47:52,431 epoch 2 - iter 146/738 - loss 0.14402353 - time (sec): 10.44 - samples/sec: 3172.64 - lr: 0.000049 - momentum: 0.000000
2023-10-17 20:47:57,788 epoch 2 - iter 219/738 - loss 0.14540226 - time (sec): 15.79 - samples/sec: 3176.56 - lr: 0.000048 - momentum: 0.000000
2023-10-17 20:48:03,569 epoch 2 - iter 292/738 - loss 0.13897723 - time (sec): 21.57 - samples/sec: 3163.61 - lr: 0.000048 - momentum: 0.000000
2023-10-17 20:48:09,446 epoch 2 - iter 365/738 - loss 0.13575281 - time (sec): 27.45 - samples/sec: 3135.71 - lr: 0.000047 - momentum: 0.000000
2023-10-17 20:48:14,813 epoch 2 - iter 438/738 - loss 0.13129263 - time (sec): 32.82 - samples/sec: 3119.78 - lr: 0.000047 - momentum: 0.000000
2023-10-17 20:48:20,077 epoch 2 - iter 511/738 - loss 0.12744292 - time (sec): 38.08 - samples/sec: 3088.08 - lr: 0.000046 - momentum: 0.000000
2023-10-17 20:48:25,072 epoch 2 - iter 584/738 - loss 0.12567829 - time (sec): 43.08 - samples/sec: 3080.94 - lr: 0.000046 - momentum: 0.000000
2023-10-17 20:48:30,070 epoch 2 - iter 657/738 - loss 0.12273838 - time (sec): 48.08 - samples/sec: 3087.93 - lr: 0.000045 - momentum: 0.000000
2023-10-17 20:48:35,401 epoch 2 - iter 730/738 - loss 0.12256728 - time (sec): 53.41 - samples/sec: 3088.76 - lr: 0.000045 - momentum: 0.000000
2023-10-17 20:48:35,899 ----------------------------------------------------------------------------------------------------
2023-10-17 20:48:35,899 EPOCH 2 done: loss 0.1224 - lr: 0.000045
2023-10-17 20:48:47,359 DEV : loss 0.1086253672838211 - f1-score (micro avg) 0.7764
2023-10-17 20:48:47,392 saving best model
2023-10-17 20:48:47,962 ----------------------------------------------------------------------------------------------------
2023-10-17 20:48:52,731 epoch 3 - iter 73/738 - loss 0.07098121 - time (sec): 4.77 - samples/sec: 3112.70 - lr: 0.000044 - momentum: 0.000000
2023-10-17 20:48:57,645 epoch 3 - iter 146/738 - loss 0.07973832 - time (sec): 9.68 - samples/sec: 3156.26 - lr: 0.000043 - momentum: 0.000000
2023-10-17 20:49:02,953 epoch 3 - iter 219/738 - loss 0.07961799 - time (sec): 14.99 - samples/sec: 3169.33 - lr: 0.000043 - momentum: 0.000000
2023-10-17 20:49:07,745 epoch 3 - iter 292/738 - loss 0.07850589 - time (sec): 19.78 - samples/sec: 3236.22 - lr: 0.000042 - momentum: 0.000000
2023-10-17 20:49:12,567 epoch 3 - iter 365/738 - loss 0.07402743 - time (sec): 24.60 - samples/sec: 3265.52 - lr: 0.000042 - momentum: 0.000000
2023-10-17 20:49:17,298 epoch 3 - iter 438/738 - loss 0.08001352 - time (sec): 29.33 - samples/sec: 3275.71 - lr: 0.000041 - momentum: 0.000000
2023-10-17 20:49:23,094 epoch 3 - iter 511/738 - loss 0.07748337 - time (sec): 35.13 - samples/sec: 3272.26 - lr: 0.000041 - momentum: 0.000000
2023-10-17 20:49:28,090 epoch 3 - iter 584/738 - loss 0.07564249 - time (sec): 40.12 - samples/sec: 3282.58 - lr: 0.000040 - momentum: 0.000000
2023-10-17 20:49:33,175 epoch 3 - iter 657/738 - loss 0.07343420 - time (sec): 45.21 - samples/sec: 3285.72 - lr: 0.000040 - momentum: 0.000000
2023-10-17 20:49:38,072 epoch 3 - iter 730/738 - loss 0.07366471 - time (sec): 50.11 - samples/sec: 3286.92 - lr: 0.000039 - momentum: 0.000000
2023-10-17 20:49:38,600 ----------------------------------------------------------------------------------------------------
2023-10-17 20:49:38,601 EPOCH 3 done: loss 0.0739 - lr: 0.000039
2023-10-17 20:49:49,997 DEV : loss 0.11555362492799759 - f1-score (micro avg) 0.8198
2023-10-17 20:49:50,031 saving best model
2023-10-17 20:49:50,487 ----------------------------------------------------------------------------------------------------
2023-10-17 20:49:55,834 epoch 4 - iter 73/738 - loss 0.04160062 - time (sec): 5.34 - samples/sec: 3175.79 - lr: 0.000038 - momentum: 0.000000
2023-10-17 20:50:01,239 epoch 4 - iter 146/738 - loss 0.04439446 - time (sec): 10.74 - samples/sec: 3270.24 - lr: 0.000038 - momentum: 0.000000
2023-10-17 20:50:06,206 epoch 4 - iter 219/738 - loss 0.04370494 - time (sec): 15.71 - samples/sec: 3268.86 - lr: 0.000037 - momentum: 0.000000
2023-10-17 20:50:11,401 epoch 4 - iter 292/738 - loss 0.04811555 - time (sec): 20.90 - samples/sec: 3253.28 - lr: 0.000037 - momentum: 0.000000
2023-10-17 20:50:16,097 epoch 4 - iter 365/738 - loss 0.04926683 - time (sec): 25.60 - samples/sec: 3256.26 - lr: 0.000036 - momentum: 0.000000
2023-10-17 20:50:20,727 epoch 4 - iter 438/738 - loss 0.04889972 - time (sec): 30.23 - samples/sec: 3275.71 - lr: 0.000036 - momentum: 0.000000
2023-10-17 20:50:25,568 epoch 4 - iter 511/738 - loss 0.04922740 - time (sec): 35.07 - samples/sec: 3286.15 - lr: 0.000035 - momentum: 0.000000
2023-10-17 20:50:30,582 epoch 4 - iter 584/738 - loss 0.04769097 - time (sec): 40.09 - samples/sec: 3295.61 - lr: 0.000035 - momentum: 0.000000
2023-10-17 20:50:35,683 epoch 4 - iter 657/738 - loss 0.04770829 - time (sec): 45.19 - samples/sec: 3308.46 - lr: 0.000034 - momentum: 0.000000
2023-10-17 20:50:40,337 epoch 4 - iter 730/738 - loss 0.04843947 - time (sec): 49.84 - samples/sec: 3304.22 - lr: 0.000033 - momentum: 0.000000
2023-10-17 20:50:40,884 ----------------------------------------------------------------------------------------------------
2023-10-17 20:50:40,885 EPOCH 4 done: loss 0.0485 - lr: 0.000033
2023-10-17 20:50:52,292 DEV : loss 0.1695607602596283 - f1-score (micro avg) 0.8296
2023-10-17 20:50:52,322 saving best model
2023-10-17 20:50:52,822 ----------------------------------------------------------------------------------------------------
2023-10-17 20:50:58,011 epoch 5 - iter 73/738 - loss 0.03608107 - time (sec): 5.19 - samples/sec: 3363.48 - lr: 0.000033 - momentum: 0.000000
2023-10-17 20:51:02,971 epoch 5 - iter 146/738 - loss 0.03199147 - time (sec): 10.15 - samples/sec: 3357.55 - lr: 0.000032 - momentum: 0.000000
2023-10-17 20:51:08,337 epoch 5 - iter 219/738 - loss 0.03313355 - time (sec): 15.51 - samples/sec: 3354.59 - lr: 0.000032 - momentum: 0.000000
2023-10-17 20:51:13,302 epoch 5 - iter 292/738 - loss 0.03443954 - time (sec): 20.48 - samples/sec: 3348.04 - lr: 0.000031 - momentum: 0.000000
2023-10-17 20:51:18,206 epoch 5 - iter 365/738 - loss 0.03367944 - time (sec): 25.38 - samples/sec: 3362.98 - lr: 0.000031 - momentum: 0.000000
2023-10-17 20:51:23,370 epoch 5 - iter 438/738 - loss 0.03267726 - time (sec): 30.55 - samples/sec: 3345.35 - lr: 0.000030 - momentum: 0.000000
2023-10-17 20:51:27,818 epoch 5 - iter 511/738 - loss 0.03446430 - time (sec): 34.99 - samples/sec: 3341.96 - lr: 0.000030 - momentum: 0.000000
2023-10-17 20:51:32,447 epoch 5 - iter 584/738 - loss 0.03493802 - time (sec): 39.62 - samples/sec: 3334.16 - lr: 0.000029 - momentum: 0.000000
2023-10-17 20:51:37,619 epoch 5 - iter 657/738 - loss 0.03468427 - time (sec): 44.80 - samples/sec: 3305.72 - lr: 0.000028 - momentum: 0.000000
2023-10-17 20:51:42,797 epoch 5 - iter 730/738 - loss 0.03596301 - time (sec): 49.97 - samples/sec: 3302.27 - lr: 0.000028 - momentum: 0.000000
2023-10-17 20:51:43,259 ----------------------------------------------------------------------------------------------------
2023-10-17 20:51:43,259 EPOCH 5 done: loss 0.0361 - lr: 0.000028
2023-10-17 20:51:54,959 DEV : loss 0.18621283769607544 - f1-score (micro avg) 0.8343
2023-10-17 20:51:54,989 saving best model
2023-10-17 20:51:55,450 ----------------------------------------------------------------------------------------------------
2023-10-17 20:52:00,421 epoch 6 - iter 73/738 - loss 0.03314020 - time (sec): 4.97 - samples/sec: 3267.43 - lr: 0.000027 - momentum: 0.000000
2023-10-17 20:52:05,581 epoch 6 - iter 146/738 - loss 0.03144585 - time (sec): 10.13 - samples/sec: 3177.13 - lr: 0.000027 - momentum: 0.000000
2023-10-17 20:52:10,154 epoch 6 - iter 219/738 - loss 0.02846193 - time (sec): 14.70 - samples/sec: 3204.07 - lr: 0.000026 - momentum: 0.000000
2023-10-17 20:52:15,181 epoch 6 - iter 292/738 - loss 0.02617097 - time (sec): 19.73 - samples/sec: 3233.54 - lr: 0.000026 - momentum: 0.000000
2023-10-17 20:52:20,401 epoch 6 - iter 365/738 - loss 0.02459012 - time (sec): 24.95 - samples/sec: 3217.27 - lr: 0.000025 - momentum: 0.000000
2023-10-17 20:52:24,933 epoch 6 - iter 438/738 - loss 0.02420569 - time (sec): 29.48 - samples/sec: 3251.94 - lr: 0.000025 - momentum: 0.000000
2023-10-17 20:52:29,485 epoch 6 - iter 511/738 - loss 0.02455905 - time (sec): 34.03 - samples/sec: 3278.27 - lr: 0.000024 - momentum: 0.000000
2023-10-17 20:52:34,344 epoch 6 - iter 584/738 - loss 0.02479615 - time (sec): 38.89 - samples/sec: 3268.69 - lr: 0.000023 - momentum: 0.000000
2023-10-17 20:52:40,085 epoch 6 - iter 657/738 - loss 0.02602451 - time (sec): 44.63 - samples/sec: 3301.26 - lr: 0.000023 - momentum: 0.000000
2023-10-17 20:52:45,132 epoch 6 - iter 730/738 - loss 0.02514643 - time (sec): 49.68 - samples/sec: 3302.41 - lr: 0.000022 - momentum: 0.000000
2023-10-17 20:52:45,851 ----------------------------------------------------------------------------------------------------
2023-10-17 20:52:45,852 EPOCH 6 done: loss 0.0252 - lr: 0.000022
2023-10-17 20:52:57,524 DEV : loss 0.19898028671741486 - f1-score (micro avg) 0.8438
2023-10-17 20:52:57,557 saving best model
2023-10-17 20:52:58,043 ----------------------------------------------------------------------------------------------------
2023-10-17 20:53:02,903 epoch 7 - iter 73/738 - loss 0.01475665 - time (sec): 4.86 - samples/sec: 3133.95 - lr: 0.000022 - momentum: 0.000000
2023-10-17 20:53:07,464 epoch 7 - iter 146/738 - loss 0.01079426 - time (sec): 9.42 - samples/sec: 3313.04 - lr: 0.000021 - momentum: 0.000000
2023-10-17 20:53:12,052 epoch 7 - iter 219/738 - loss 0.01197725 - time (sec): 14.01 - samples/sec: 3264.54 - lr: 0.000021 - momentum: 0.000000
2023-10-17 20:53:16,929 epoch 7 - iter 292/738 - loss 0.01292926 - time (sec): 18.88 - samples/sec: 3284.79 - lr: 0.000020 - momentum: 0.000000
2023-10-17 20:53:21,733 epoch 7 - iter 365/738 - loss 0.01589341 - time (sec): 23.69 - samples/sec: 3296.02 - lr: 0.000020 - momentum: 0.000000
2023-10-17 20:53:26,782 epoch 7 - iter 438/738 - loss 0.01627966 - time (sec): 28.74 - samples/sec: 3340.14 - lr: 0.000019 - momentum: 0.000000
2023-10-17 20:53:32,829 epoch 7 - iter 511/738 - loss 0.01882907 - time (sec): 34.79 - samples/sec: 3352.44 - lr: 0.000018 - momentum: 0.000000
2023-10-17 20:53:37,503 epoch 7 - iter 584/738 - loss 0.01875664 - time (sec): 39.46 - samples/sec: 3353.64 - lr: 0.000018 - momentum: 0.000000
2023-10-17 20:53:42,403 epoch 7 - iter 657/738 - loss 0.01898628 - time (sec): 44.36 - samples/sec: 3350.33 - lr: 0.000017 - momentum: 0.000000
2023-10-17 20:53:47,332 epoch 7 - iter 730/738 - loss 0.01903040 - time (sec): 49.29 - samples/sec: 3341.12 - lr: 0.000017 - momentum: 0.000000
2023-10-17 20:53:47,894 ----------------------------------------------------------------------------------------------------
2023-10-17 20:53:47,894 EPOCH 7 done: loss 0.0194 - lr: 0.000017
2023-10-17 20:53:59,398 DEV : loss 0.2004670649766922 - f1-score (micro avg) 0.8497
2023-10-17 20:53:59,432 saving best model
2023-10-17 20:53:59,916 ----------------------------------------------------------------------------------------------------
2023-10-17 20:54:04,698 epoch 8 - iter 73/738 - loss 0.01638856 - time (sec): 4.78 - samples/sec: 3252.98 - lr: 0.000016 - momentum: 0.000000
2023-10-17 20:54:09,704 epoch 8 - iter 146/738 - loss 0.01268644 - time (sec): 9.79 - samples/sec: 3215.86 - lr: 0.000016 - momentum: 0.000000
2023-10-17 20:54:14,523 epoch 8 - iter 219/738 - loss 0.01154489 - time (sec): 14.61 - samples/sec: 3238.88 - lr: 0.000015 - momentum: 0.000000
2023-10-17 20:54:19,163 epoch 8 - iter 292/738 - loss 0.01121476 - time (sec): 19.25 - samples/sec: 3253.95 - lr: 0.000015 - momentum: 0.000000
2023-10-17 20:54:25,326 epoch 8 - iter 365/738 - loss 0.01237795 - time (sec): 25.41 - samples/sec: 3256.44 - lr: 0.000014 - momentum: 0.000000
2023-10-17 20:54:31,141 epoch 8 - iter 438/738 - loss 0.01216074 - time (sec): 31.22 - samples/sec: 3254.28 - lr: 0.000013 - momentum: 0.000000
2023-10-17 20:54:36,019 epoch 8 - iter 511/738 - loss 0.01129103 - time (sec): 36.10 - samples/sec: 3253.41 - lr: 0.000013 - momentum: 0.000000
2023-10-17 20:54:41,054 epoch 8 - iter 584/738 - loss 0.01142640 - time (sec): 41.14 - samples/sec: 3261.75 - lr: 0.000012 - momentum: 0.000000
2023-10-17 20:54:45,906 epoch 8 - iter 657/738 - loss 0.01201789 - time (sec): 45.99 - samples/sec: 3247.76 - lr: 0.000012 - momentum: 0.000000
2023-10-17 20:54:50,232 epoch 8 - iter 730/738 - loss 0.01177942 - time (sec): 50.31 - samples/sec: 3270.00 - lr: 0.000011 - momentum: 0.000000
2023-10-17 20:54:50,766 ----------------------------------------------------------------------------------------------------
2023-10-17 20:54:50,767 EPOCH 8 done: loss 0.0118 - lr: 0.000011
2023-10-17 20:55:02,150 DEV : loss 0.198430597782135 - f1-score (micro avg) 0.8567
2023-10-17 20:55:02,181 saving best model
2023-10-17 20:55:02,675 ----------------------------------------------------------------------------------------------------
2023-10-17 20:55:07,868 epoch 9 - iter 73/738 - loss 0.01519568 - time (sec): 5.19 - samples/sec: 3455.39 - lr: 0.000011 - momentum: 0.000000
2023-10-17 20:55:12,960 epoch 9 - iter 146/738 - loss 0.00937816 - time (sec): 10.28 - samples/sec: 3348.66 - lr: 0.000010 - momentum: 0.000000
2023-10-17 20:55:17,958 epoch 9 - iter 219/738 - loss 0.00848427 - time (sec): 15.28 - samples/sec: 3265.03 - lr: 0.000010 - momentum: 0.000000
2023-10-17 20:55:23,209 epoch 9 - iter 292/738 - loss 0.00799314 - time (sec): 20.53 - samples/sec: 3258.21 - lr: 0.000009 - momentum: 0.000000
2023-10-17 20:55:27,918 epoch 9 - iter 365/738 - loss 0.00807520 - time (sec): 25.24 - samples/sec: 3272.15 - lr: 0.000008 - momentum: 0.000000
2023-10-17 20:55:33,094 epoch 9 - iter 438/738 - loss 0.00784412 - time (sec): 30.42 - samples/sec: 3286.92 - lr: 0.000008 - momentum: 0.000000
2023-10-17 20:55:38,313 epoch 9 - iter 511/738 - loss 0.00889544 - time (sec): 35.64 - samples/sec: 3274.78 - lr: 0.000007 - momentum: 0.000000
2023-10-17 20:55:43,202 epoch 9 - iter 584/738 - loss 0.00926012 - time (sec): 40.52 - samples/sec: 3281.89 - lr: 0.000007 - momentum: 0.000000
2023-10-17 20:55:48,350 epoch 9 - iter 657/738 - loss 0.00863704 - time (sec): 45.67 - samples/sec: 3283.07 - lr: 0.000006 - momentum: 0.000000
2023-10-17 20:55:52,805 epoch 9 - iter 730/738 - loss 0.00798481 - time (sec): 50.13 - samples/sec: 3290.70 - lr: 0.000006 - momentum: 0.000000
2023-10-17 20:55:53,313 ----------------------------------------------------------------------------------------------------
2023-10-17 20:55:53,313 EPOCH 9 done: loss 0.0079 - lr: 0.000006
2023-10-17 20:56:04,772 DEV : loss 0.2114827036857605 - f1-score (micro avg) 0.8592
2023-10-17 20:56:04,806 saving best model
2023-10-17 20:56:05,225 ----------------------------------------------------------------------------------------------------
2023-10-17 20:56:10,423 epoch 10 - iter 73/738 - loss 0.00697079 - time (sec): 5.20 - samples/sec: 3137.08 - lr: 0.000005 - momentum: 0.000000
2023-10-17 20:56:16,026 epoch 10 - iter 146/738 - loss 0.00806024 - time (sec): 10.80 - samples/sec: 3238.10 - lr: 0.000004 - momentum: 0.000000
2023-10-17 20:56:20,994 epoch 10 - iter 219/738 - loss 0.00854713 - time (sec): 15.77 - samples/sec: 3195.31 - lr: 0.000004 - momentum: 0.000000
2023-10-17 20:56:26,151 epoch 10 - iter 292/738 - loss 0.00725008 - time (sec): 20.92 - samples/sec: 3224.88 - lr: 0.000003 - momentum: 0.000000
2023-10-17 20:56:30,811 epoch 10 - iter 365/738 - loss 0.00633495 - time (sec): 25.58 - samples/sec: 3248.59 - lr: 0.000003 - momentum: 0.000000
2023-10-17 20:56:35,376 epoch 10 - iter 438/738 - loss 0.00772199 - time (sec): 30.15 - samples/sec: 3274.22 - lr: 0.000002 - momentum: 0.000000
2023-10-17 20:56:40,499 epoch 10 - iter 511/738 - loss 0.00693419 - time (sec): 35.27 - samples/sec: 3250.26 - lr: 0.000002 - momentum: 0.000000
2023-10-17 20:56:45,335 epoch 10 - iter 584/738 - loss 0.00661804 - time (sec): 40.11 - samples/sec: 3269.64 - lr: 0.000001 - momentum: 0.000000
2023-10-17 20:56:50,238 epoch 10 - iter 657/738 - loss 0.00621935 - time (sec): 45.01 - samples/sec: 3277.65 - lr: 0.000001 - momentum: 0.000000
2023-10-17 20:56:55,354 epoch 10 - iter 730/738 - loss 0.00579273 - time (sec): 50.13 - samples/sec: 3289.07 - lr: 0.000000 - momentum: 0.000000
2023-10-17 20:56:55,849 ----------------------------------------------------------------------------------------------------
2023-10-17 20:56:55,849 EPOCH 10 done: loss 0.0057 - lr: 0.000000
2023-10-17 20:57:07,458 DEV : loss 0.2092556357383728 - f1-score (micro avg) 0.8575
2023-10-17 20:57:07,845 ----------------------------------------------------------------------------------------------------
2023-10-17 20:57:07,846 Loading model from best epoch ...
2023-10-17 20:57:09,560 SequenceTagger predicts: Dictionary with 21 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, S-time, B-time, E-time, I-time, S-prod, B-prod, E-prod, I-prod
2023-10-17 20:57:15,748
Results:
- F-score (micro) 0.8061
- F-score (macro) 0.7226
- Accuracy 0.6955
By class:
precision recall f1-score support
loc 0.8573 0.8753 0.8662 858
pers 0.7504 0.8175 0.7825 537
org 0.6579 0.5682 0.6098 132
time 0.5806 0.6667 0.6207 54
prod 0.8333 0.6557 0.7339 61
micro avg 0.7958 0.8167 0.8061 1642
macro avg 0.7359 0.7167 0.7226 1642
weighted avg 0.7963 0.8167 0.8052 1642
2023-10-17 20:57:15,748 ----------------------------------------------------------------------------------------------------
|