File size: 23,985 Bytes
b880a8f |
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 |
2023-10-17 10:50:20,328 ----------------------------------------------------------------------------------------------------
2023-10-17 10:50:20,329 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=25, bias=True)
(loss_function): CrossEntropyLoss()
)"
2023-10-17 10:50:20,329 ----------------------------------------------------------------------------------------------------
2023-10-17 10:50:20,329 MultiCorpus: 966 train + 219 dev + 204 test sentences
- NER_HIPE_2022 Corpus: 966 train + 219 dev + 204 test sentences - /root/.flair/datasets/ner_hipe_2022/v2.1/ajmc/fr/with_doc_seperator
2023-10-17 10:50:20,329 ----------------------------------------------------------------------------------------------------
2023-10-17 10:50:20,329 Train: 966 sentences
2023-10-17 10:50:20,329 (train_with_dev=False, train_with_test=False)
2023-10-17 10:50:20,329 ----------------------------------------------------------------------------------------------------
2023-10-17 10:50:20,329 Training Params:
2023-10-17 10:50:20,329 - learning_rate: "5e-05"
2023-10-17 10:50:20,329 - mini_batch_size: "8"
2023-10-17 10:50:20,329 - max_epochs: "10"
2023-10-17 10:50:20,329 - shuffle: "True"
2023-10-17 10:50:20,329 ----------------------------------------------------------------------------------------------------
2023-10-17 10:50:20,329 Plugins:
2023-10-17 10:50:20,329 - TensorboardLogger
2023-10-17 10:50:20,329 - LinearScheduler | warmup_fraction: '0.1'
2023-10-17 10:50:20,329 ----------------------------------------------------------------------------------------------------
2023-10-17 10:50:20,329 Final evaluation on model from best epoch (best-model.pt)
2023-10-17 10:50:20,329 - metric: "('micro avg', 'f1-score')"
2023-10-17 10:50:20,329 ----------------------------------------------------------------------------------------------------
2023-10-17 10:50:20,329 Computation:
2023-10-17 10:50:20,329 - compute on device: cuda:0
2023-10-17 10:50:20,329 - embedding storage: none
2023-10-17 10:50:20,329 ----------------------------------------------------------------------------------------------------
2023-10-17 10:50:20,329 Model training base path: "hmbench-ajmc/fr-hmteams/teams-base-historic-multilingual-discriminator-bs8-wsFalse-e10-lr5e-05-poolingfirst-layers-1-crfFalse-3"
2023-10-17 10:50:20,329 ----------------------------------------------------------------------------------------------------
2023-10-17 10:50:20,329 ----------------------------------------------------------------------------------------------------
2023-10-17 10:50:20,330 Logging anything other than scalars to TensorBoard is currently not supported.
2023-10-17 10:50:21,015 epoch 1 - iter 12/121 - loss 4.32717646 - time (sec): 0.68 - samples/sec: 3534.81 - lr: 0.000005 - momentum: 0.000000
2023-10-17 10:50:21,794 epoch 1 - iter 24/121 - loss 3.79490891 - time (sec): 1.46 - samples/sec: 3486.90 - lr: 0.000010 - momentum: 0.000000
2023-10-17 10:50:22,476 epoch 1 - iter 36/121 - loss 3.02438972 - time (sec): 2.15 - samples/sec: 3542.13 - lr: 0.000014 - momentum: 0.000000
2023-10-17 10:50:23,191 epoch 1 - iter 48/121 - loss 2.44713095 - time (sec): 2.86 - samples/sec: 3522.54 - lr: 0.000019 - momentum: 0.000000
2023-10-17 10:50:23,961 epoch 1 - iter 60/121 - loss 2.02221030 - time (sec): 3.63 - samples/sec: 3481.44 - lr: 0.000024 - momentum: 0.000000
2023-10-17 10:50:24,723 epoch 1 - iter 72/121 - loss 1.76776424 - time (sec): 4.39 - samples/sec: 3425.36 - lr: 0.000029 - momentum: 0.000000
2023-10-17 10:50:25,436 epoch 1 - iter 84/121 - loss 1.58304931 - time (sec): 5.11 - samples/sec: 3404.68 - lr: 0.000034 - momentum: 0.000000
2023-10-17 10:50:26,157 epoch 1 - iter 96/121 - loss 1.43030745 - time (sec): 5.83 - samples/sec: 3397.39 - lr: 0.000039 - momentum: 0.000000
2023-10-17 10:50:26,900 epoch 1 - iter 108/121 - loss 1.31326734 - time (sec): 6.57 - samples/sec: 3364.32 - lr: 0.000044 - momentum: 0.000000
2023-10-17 10:50:27,691 epoch 1 - iter 120/121 - loss 1.20599299 - time (sec): 7.36 - samples/sec: 3340.92 - lr: 0.000049 - momentum: 0.000000
2023-10-17 10:50:27,742 ----------------------------------------------------------------------------------------------------
2023-10-17 10:50:27,743 EPOCH 1 done: loss 1.1989 - lr: 0.000049
2023-10-17 10:50:28,619 DEV : loss 0.2106715589761734 - f1-score (micro avg) 0.6094
2023-10-17 10:50:28,624 saving best model
2023-10-17 10:50:29,074 ----------------------------------------------------------------------------------------------------
2023-10-17 10:50:29,893 epoch 2 - iter 12/121 - loss 0.23738903 - time (sec): 0.82 - samples/sec: 3194.92 - lr: 0.000049 - momentum: 0.000000
2023-10-17 10:50:30,612 epoch 2 - iter 24/121 - loss 0.24865588 - time (sec): 1.54 - samples/sec: 3223.62 - lr: 0.000049 - momentum: 0.000000
2023-10-17 10:50:31,317 epoch 2 - iter 36/121 - loss 0.23597534 - time (sec): 2.24 - samples/sec: 3328.38 - lr: 0.000048 - momentum: 0.000000
2023-10-17 10:50:32,044 epoch 2 - iter 48/121 - loss 0.22613230 - time (sec): 2.97 - samples/sec: 3296.33 - lr: 0.000048 - momentum: 0.000000
2023-10-17 10:50:32,802 epoch 2 - iter 60/121 - loss 0.22136440 - time (sec): 3.73 - samples/sec: 3223.84 - lr: 0.000047 - momentum: 0.000000
2023-10-17 10:50:33,545 epoch 2 - iter 72/121 - loss 0.21509557 - time (sec): 4.47 - samples/sec: 3255.40 - lr: 0.000047 - momentum: 0.000000
2023-10-17 10:50:34,256 epoch 2 - iter 84/121 - loss 0.20744860 - time (sec): 5.18 - samples/sec: 3245.71 - lr: 0.000046 - momentum: 0.000000
2023-10-17 10:50:35,035 epoch 2 - iter 96/121 - loss 0.19170653 - time (sec): 5.96 - samples/sec: 3304.39 - lr: 0.000046 - momentum: 0.000000
2023-10-17 10:50:35,743 epoch 2 - iter 108/121 - loss 0.18452437 - time (sec): 6.67 - samples/sec: 3309.48 - lr: 0.000045 - momentum: 0.000000
2023-10-17 10:50:36,560 epoch 2 - iter 120/121 - loss 0.18185624 - time (sec): 7.48 - samples/sec: 3283.39 - lr: 0.000045 - momentum: 0.000000
2023-10-17 10:50:36,613 ----------------------------------------------------------------------------------------------------
2023-10-17 10:50:36,614 EPOCH 2 done: loss 0.1816 - lr: 0.000045
2023-10-17 10:50:37,356 DEV : loss 0.12012884765863419 - f1-score (micro avg) 0.8149
2023-10-17 10:50:37,361 saving best model
2023-10-17 10:50:37,919 ----------------------------------------------------------------------------------------------------
2023-10-17 10:50:38,615 epoch 3 - iter 12/121 - loss 0.15633326 - time (sec): 0.69 - samples/sec: 3413.70 - lr: 0.000044 - momentum: 0.000000
2023-10-17 10:50:39,379 epoch 3 - iter 24/121 - loss 0.11693087 - time (sec): 1.46 - samples/sec: 3358.81 - lr: 0.000043 - momentum: 0.000000
2023-10-17 10:50:40,100 epoch 3 - iter 36/121 - loss 0.09791334 - time (sec): 2.18 - samples/sec: 3307.00 - lr: 0.000043 - momentum: 0.000000
2023-10-17 10:50:40,777 epoch 3 - iter 48/121 - loss 0.09570602 - time (sec): 2.85 - samples/sec: 3390.21 - lr: 0.000042 - momentum: 0.000000
2023-10-17 10:50:41,463 epoch 3 - iter 60/121 - loss 0.10341246 - time (sec): 3.54 - samples/sec: 3382.12 - lr: 0.000042 - momentum: 0.000000
2023-10-17 10:50:42,164 epoch 3 - iter 72/121 - loss 0.10258523 - time (sec): 4.24 - samples/sec: 3444.02 - lr: 0.000041 - momentum: 0.000000
2023-10-17 10:50:42,911 epoch 3 - iter 84/121 - loss 0.10332344 - time (sec): 4.99 - samples/sec: 3448.10 - lr: 0.000041 - momentum: 0.000000
2023-10-17 10:50:43,640 epoch 3 - iter 96/121 - loss 0.10408491 - time (sec): 5.72 - samples/sec: 3429.02 - lr: 0.000040 - momentum: 0.000000
2023-10-17 10:50:44,345 epoch 3 - iter 108/121 - loss 0.09963223 - time (sec): 6.42 - samples/sec: 3420.37 - lr: 0.000040 - momentum: 0.000000
2023-10-17 10:50:45,131 epoch 3 - iter 120/121 - loss 0.10104994 - time (sec): 7.21 - samples/sec: 3413.52 - lr: 0.000039 - momentum: 0.000000
2023-10-17 10:50:45,199 ----------------------------------------------------------------------------------------------------
2023-10-17 10:50:45,200 EPOCH 3 done: loss 0.1006 - lr: 0.000039
2023-10-17 10:50:45,951 DEV : loss 0.12703241407871246 - f1-score (micro avg) 0.8174
2023-10-17 10:50:45,956 saving best model
2023-10-17 10:50:46,475 ----------------------------------------------------------------------------------------------------
2023-10-17 10:50:47,241 epoch 4 - iter 12/121 - loss 0.07640209 - time (sec): 0.76 - samples/sec: 3140.72 - lr: 0.000038 - momentum: 0.000000
2023-10-17 10:50:48,011 epoch 4 - iter 24/121 - loss 0.06068045 - time (sec): 1.53 - samples/sec: 3274.30 - lr: 0.000038 - momentum: 0.000000
2023-10-17 10:50:48,722 epoch 4 - iter 36/121 - loss 0.06492967 - time (sec): 2.24 - samples/sec: 3320.97 - lr: 0.000037 - momentum: 0.000000
2023-10-17 10:50:49,455 epoch 4 - iter 48/121 - loss 0.07473805 - time (sec): 2.98 - samples/sec: 3346.46 - lr: 0.000037 - momentum: 0.000000
2023-10-17 10:50:50,188 epoch 4 - iter 60/121 - loss 0.07679395 - time (sec): 3.71 - samples/sec: 3309.00 - lr: 0.000036 - momentum: 0.000000
2023-10-17 10:50:50,964 epoch 4 - iter 72/121 - loss 0.07496194 - time (sec): 4.49 - samples/sec: 3326.15 - lr: 0.000036 - momentum: 0.000000
2023-10-17 10:50:51,721 epoch 4 - iter 84/121 - loss 0.07700276 - time (sec): 5.24 - samples/sec: 3289.90 - lr: 0.000035 - momentum: 0.000000
2023-10-17 10:50:52,558 epoch 4 - iter 96/121 - loss 0.07436746 - time (sec): 6.08 - samples/sec: 3265.35 - lr: 0.000035 - momentum: 0.000000
2023-10-17 10:50:53,232 epoch 4 - iter 108/121 - loss 0.07539758 - time (sec): 6.75 - samples/sec: 3275.69 - lr: 0.000034 - momentum: 0.000000
2023-10-17 10:50:53,982 epoch 4 - iter 120/121 - loss 0.07183802 - time (sec): 7.50 - samples/sec: 3261.73 - lr: 0.000034 - momentum: 0.000000
2023-10-17 10:50:54,065 ----------------------------------------------------------------------------------------------------
2023-10-17 10:50:54,065 EPOCH 4 done: loss 0.0712 - lr: 0.000034
2023-10-17 10:50:54,810 DEV : loss 0.15281036496162415 - f1-score (micro avg) 0.7975
2023-10-17 10:50:54,815 ----------------------------------------------------------------------------------------------------
2023-10-17 10:50:55,545 epoch 5 - iter 12/121 - loss 0.08270934 - time (sec): 0.73 - samples/sec: 3631.70 - lr: 0.000033 - momentum: 0.000000
2023-10-17 10:50:56,326 epoch 5 - iter 24/121 - loss 0.06659483 - time (sec): 1.51 - samples/sec: 3368.31 - lr: 0.000032 - momentum: 0.000000
2023-10-17 10:50:57,088 epoch 5 - iter 36/121 - loss 0.06233471 - time (sec): 2.27 - samples/sec: 3299.63 - lr: 0.000032 - momentum: 0.000000
2023-10-17 10:50:57,863 epoch 5 - iter 48/121 - loss 0.06132773 - time (sec): 3.05 - samples/sec: 3308.14 - lr: 0.000031 - momentum: 0.000000
2023-10-17 10:50:58,652 epoch 5 - iter 60/121 - loss 0.05814848 - time (sec): 3.84 - samples/sec: 3243.44 - lr: 0.000031 - momentum: 0.000000
2023-10-17 10:50:59,394 epoch 5 - iter 72/121 - loss 0.05297482 - time (sec): 4.58 - samples/sec: 3267.35 - lr: 0.000030 - momentum: 0.000000
2023-10-17 10:51:00,138 epoch 5 - iter 84/121 - loss 0.05054123 - time (sec): 5.32 - samples/sec: 3291.35 - lr: 0.000030 - momentum: 0.000000
2023-10-17 10:51:00,891 epoch 5 - iter 96/121 - loss 0.04886654 - time (sec): 6.08 - samples/sec: 3257.38 - lr: 0.000029 - momentum: 0.000000
2023-10-17 10:51:01,606 epoch 5 - iter 108/121 - loss 0.04837317 - time (sec): 6.79 - samples/sec: 3269.49 - lr: 0.000029 - momentum: 0.000000
2023-10-17 10:51:02,298 epoch 5 - iter 120/121 - loss 0.04816303 - time (sec): 7.48 - samples/sec: 3277.29 - lr: 0.000028 - momentum: 0.000000
2023-10-17 10:51:02,357 ----------------------------------------------------------------------------------------------------
2023-10-17 10:51:02,357 EPOCH 5 done: loss 0.0481 - lr: 0.000028
2023-10-17 10:51:03,111 DEV : loss 0.14528368413448334 - f1-score (micro avg) 0.8388
2023-10-17 10:51:03,116 saving best model
2023-10-17 10:51:03,632 ----------------------------------------------------------------------------------------------------
2023-10-17 10:51:04,397 epoch 6 - iter 12/121 - loss 0.05046710 - time (sec): 0.76 - samples/sec: 3238.54 - lr: 0.000027 - momentum: 0.000000
2023-10-17 10:51:05,155 epoch 6 - iter 24/121 - loss 0.03305271 - time (sec): 1.52 - samples/sec: 3151.43 - lr: 0.000027 - momentum: 0.000000
2023-10-17 10:51:05,855 epoch 6 - iter 36/121 - loss 0.03931931 - time (sec): 2.22 - samples/sec: 3237.12 - lr: 0.000026 - momentum: 0.000000
2023-10-17 10:51:06,625 epoch 6 - iter 48/121 - loss 0.03731659 - time (sec): 2.99 - samples/sec: 3269.12 - lr: 0.000026 - momentum: 0.000000
2023-10-17 10:51:07,391 epoch 6 - iter 60/121 - loss 0.03486466 - time (sec): 3.75 - samples/sec: 3267.13 - lr: 0.000025 - momentum: 0.000000
2023-10-17 10:51:08,190 epoch 6 - iter 72/121 - loss 0.03218689 - time (sec): 4.55 - samples/sec: 3276.40 - lr: 0.000025 - momentum: 0.000000
2023-10-17 10:51:08,940 epoch 6 - iter 84/121 - loss 0.03699222 - time (sec): 5.30 - samples/sec: 3288.75 - lr: 0.000024 - momentum: 0.000000
2023-10-17 10:51:09,668 epoch 6 - iter 96/121 - loss 0.03583866 - time (sec): 6.03 - samples/sec: 3272.02 - lr: 0.000024 - momentum: 0.000000
2023-10-17 10:51:10,413 epoch 6 - iter 108/121 - loss 0.03569511 - time (sec): 6.78 - samples/sec: 3268.97 - lr: 0.000023 - momentum: 0.000000
2023-10-17 10:51:11,175 epoch 6 - iter 120/121 - loss 0.03576859 - time (sec): 7.54 - samples/sec: 3269.06 - lr: 0.000022 - momentum: 0.000000
2023-10-17 10:51:11,220 ----------------------------------------------------------------------------------------------------
2023-10-17 10:51:11,220 EPOCH 6 done: loss 0.0356 - lr: 0.000022
2023-10-17 10:51:11,970 DEV : loss 0.1540733426809311 - f1-score (micro avg) 0.8438
2023-10-17 10:51:11,975 saving best model
2023-10-17 10:51:12,529 ----------------------------------------------------------------------------------------------------
2023-10-17 10:51:13,352 epoch 7 - iter 12/121 - loss 0.03229565 - time (sec): 0.82 - samples/sec: 3322.96 - lr: 0.000022 - momentum: 0.000000
2023-10-17 10:51:14,033 epoch 7 - iter 24/121 - loss 0.02874215 - time (sec): 1.50 - samples/sec: 3195.16 - lr: 0.000021 - momentum: 0.000000
2023-10-17 10:51:14,762 epoch 7 - iter 36/121 - loss 0.03499203 - time (sec): 2.23 - samples/sec: 3250.18 - lr: 0.000021 - momentum: 0.000000
2023-10-17 10:51:15,468 epoch 7 - iter 48/121 - loss 0.03181783 - time (sec): 2.93 - samples/sec: 3277.16 - lr: 0.000020 - momentum: 0.000000
2023-10-17 10:51:16,232 epoch 7 - iter 60/121 - loss 0.02848033 - time (sec): 3.70 - samples/sec: 3299.28 - lr: 0.000020 - momentum: 0.000000
2023-10-17 10:51:17,088 epoch 7 - iter 72/121 - loss 0.02842693 - time (sec): 4.55 - samples/sec: 3298.42 - lr: 0.000019 - momentum: 0.000000
2023-10-17 10:51:17,859 epoch 7 - iter 84/121 - loss 0.02616365 - time (sec): 5.33 - samples/sec: 3237.01 - lr: 0.000019 - momentum: 0.000000
2023-10-17 10:51:18,666 epoch 7 - iter 96/121 - loss 0.02440421 - time (sec): 6.13 - samples/sec: 3249.58 - lr: 0.000018 - momentum: 0.000000
2023-10-17 10:51:19,407 epoch 7 - iter 108/121 - loss 0.02378147 - time (sec): 6.87 - samples/sec: 3222.47 - lr: 0.000017 - momentum: 0.000000
2023-10-17 10:51:20,149 epoch 7 - iter 120/121 - loss 0.02375173 - time (sec): 7.61 - samples/sec: 3225.51 - lr: 0.000017 - momentum: 0.000000
2023-10-17 10:51:20,199 ----------------------------------------------------------------------------------------------------
2023-10-17 10:51:20,199 EPOCH 7 done: loss 0.0236 - lr: 0.000017
2023-10-17 10:51:20,946 DEV : loss 0.18394897878170013 - f1-score (micro avg) 0.8486
2023-10-17 10:51:20,951 saving best model
2023-10-17 10:51:21,446 ----------------------------------------------------------------------------------------------------
2023-10-17 10:51:22,256 epoch 8 - iter 12/121 - loss 0.01483429 - time (sec): 0.81 - samples/sec: 2850.77 - lr: 0.000016 - momentum: 0.000000
2023-10-17 10:51:23,054 epoch 8 - iter 24/121 - loss 0.01352030 - time (sec): 1.60 - samples/sec: 3133.34 - lr: 0.000016 - momentum: 0.000000
2023-10-17 10:51:23,784 epoch 8 - iter 36/121 - loss 0.01328131 - time (sec): 2.33 - samples/sec: 3155.77 - lr: 0.000015 - momentum: 0.000000
2023-10-17 10:51:24,708 epoch 8 - iter 48/121 - loss 0.02044563 - time (sec): 3.26 - samples/sec: 3066.02 - lr: 0.000015 - momentum: 0.000000
2023-10-17 10:51:25,493 epoch 8 - iter 60/121 - loss 0.01788009 - time (sec): 4.04 - samples/sec: 3057.56 - lr: 0.000014 - momentum: 0.000000
2023-10-17 10:51:26,287 epoch 8 - iter 72/121 - loss 0.01649561 - time (sec): 4.84 - samples/sec: 3080.65 - lr: 0.000014 - momentum: 0.000000
2023-10-17 10:51:27,048 epoch 8 - iter 84/121 - loss 0.01657942 - time (sec): 5.60 - samples/sec: 3117.59 - lr: 0.000013 - momentum: 0.000000
2023-10-17 10:51:27,791 epoch 8 - iter 96/121 - loss 0.01844025 - time (sec): 6.34 - samples/sec: 3112.92 - lr: 0.000013 - momentum: 0.000000
2023-10-17 10:51:28,510 epoch 8 - iter 108/121 - loss 0.01744183 - time (sec): 7.06 - samples/sec: 3138.61 - lr: 0.000012 - momentum: 0.000000
2023-10-17 10:51:29,269 epoch 8 - iter 120/121 - loss 0.01642666 - time (sec): 7.82 - samples/sec: 3150.18 - lr: 0.000011 - momentum: 0.000000
2023-10-17 10:51:29,319 ----------------------------------------------------------------------------------------------------
2023-10-17 10:51:29,319 EPOCH 8 done: loss 0.0165 - lr: 0.000011
2023-10-17 10:51:30,065 DEV : loss 0.2010391652584076 - f1-score (micro avg) 0.8458
2023-10-17 10:51:30,070 ----------------------------------------------------------------------------------------------------
2023-10-17 10:51:30,833 epoch 9 - iter 12/121 - loss 0.00202037 - time (sec): 0.76 - samples/sec: 3322.91 - lr: 0.000011 - momentum: 0.000000
2023-10-17 10:51:31,620 epoch 9 - iter 24/121 - loss 0.00557648 - time (sec): 1.55 - samples/sec: 3299.18 - lr: 0.000010 - momentum: 0.000000
2023-10-17 10:51:32,334 epoch 9 - iter 36/121 - loss 0.00835979 - time (sec): 2.26 - samples/sec: 3355.49 - lr: 0.000010 - momentum: 0.000000
2023-10-17 10:51:33,072 epoch 9 - iter 48/121 - loss 0.00997265 - time (sec): 3.00 - samples/sec: 3296.50 - lr: 0.000009 - momentum: 0.000000
2023-10-17 10:51:33,794 epoch 9 - iter 60/121 - loss 0.01256146 - time (sec): 3.72 - samples/sec: 3290.16 - lr: 0.000009 - momentum: 0.000000
2023-10-17 10:51:34,472 epoch 9 - iter 72/121 - loss 0.01172354 - time (sec): 4.40 - samples/sec: 3276.50 - lr: 0.000008 - momentum: 0.000000
2023-10-17 10:51:35,286 epoch 9 - iter 84/121 - loss 0.01341421 - time (sec): 5.22 - samples/sec: 3266.20 - lr: 0.000008 - momentum: 0.000000
2023-10-17 10:51:36,037 epoch 9 - iter 96/121 - loss 0.01221592 - time (sec): 5.97 - samples/sec: 3278.11 - lr: 0.000007 - momentum: 0.000000
2023-10-17 10:51:36,870 epoch 9 - iter 108/121 - loss 0.01168407 - time (sec): 6.80 - samples/sec: 3266.46 - lr: 0.000006 - momentum: 0.000000
2023-10-17 10:51:37,591 epoch 9 - iter 120/121 - loss 0.01135339 - time (sec): 7.52 - samples/sec: 3264.48 - lr: 0.000006 - momentum: 0.000000
2023-10-17 10:51:37,643 ----------------------------------------------------------------------------------------------------
2023-10-17 10:51:37,644 EPOCH 9 done: loss 0.0113 - lr: 0.000006
2023-10-17 10:51:38,396 DEV : loss 0.21391461789608002 - f1-score (micro avg) 0.8365
2023-10-17 10:51:38,401 ----------------------------------------------------------------------------------------------------
2023-10-17 10:51:39,098 epoch 10 - iter 12/121 - loss 0.00259867 - time (sec): 0.70 - samples/sec: 3364.48 - lr: 0.000005 - momentum: 0.000000
2023-10-17 10:51:39,884 epoch 10 - iter 24/121 - loss 0.01367263 - time (sec): 1.48 - samples/sec: 3276.75 - lr: 0.000005 - momentum: 0.000000
2023-10-17 10:51:40,587 epoch 10 - iter 36/121 - loss 0.01004612 - time (sec): 2.19 - samples/sec: 3290.10 - lr: 0.000004 - momentum: 0.000000
2023-10-17 10:51:41,349 epoch 10 - iter 48/121 - loss 0.00780401 - time (sec): 2.95 - samples/sec: 3354.48 - lr: 0.000004 - momentum: 0.000000
2023-10-17 10:51:42,214 epoch 10 - iter 60/121 - loss 0.00683455 - time (sec): 3.81 - samples/sec: 3306.74 - lr: 0.000003 - momentum: 0.000000
2023-10-17 10:51:42,924 epoch 10 - iter 72/121 - loss 0.00900062 - time (sec): 4.52 - samples/sec: 3303.98 - lr: 0.000003 - momentum: 0.000000
2023-10-17 10:51:43,632 epoch 10 - iter 84/121 - loss 0.00996906 - time (sec): 5.23 - samples/sec: 3332.64 - lr: 0.000002 - momentum: 0.000000
2023-10-17 10:51:44,407 epoch 10 - iter 96/121 - loss 0.00903771 - time (sec): 6.01 - samples/sec: 3329.89 - lr: 0.000001 - momentum: 0.000000
2023-10-17 10:51:45,110 epoch 10 - iter 108/121 - loss 0.00926567 - time (sec): 6.71 - samples/sec: 3323.21 - lr: 0.000001 - momentum: 0.000000
2023-10-17 10:51:45,842 epoch 10 - iter 120/121 - loss 0.00996103 - time (sec): 7.44 - samples/sec: 3313.35 - lr: 0.000000 - momentum: 0.000000
2023-10-17 10:51:45,888 ----------------------------------------------------------------------------------------------------
2023-10-17 10:51:45,889 EPOCH 10 done: loss 0.0099 - lr: 0.000000
2023-10-17 10:51:46,646 DEV : loss 0.21232573688030243 - f1-score (micro avg) 0.85
2023-10-17 10:51:46,651 saving best model
2023-10-17 10:51:47,559 ----------------------------------------------------------------------------------------------------
2023-10-17 10:51:47,560 Loading model from best epoch ...
2023-10-17 10:51:48,927 SequenceTagger predicts: Dictionary with 25 tags: O, S-scope, B-scope, E-scope, I-scope, S-pers, B-pers, E-pers, I-pers, S-work, B-work, E-work, I-work, S-loc, B-loc, E-loc, I-loc, S-object, B-object, E-object, I-object, S-date, B-date, E-date, I-date
2023-10-17 10:51:49,597
Results:
- F-score (micro) 0.832
- F-score (macro) 0.5646
- Accuracy 0.731
By class:
precision recall f1-score support
pers 0.8732 0.8921 0.8826 139
scope 0.8540 0.9070 0.8797 129
work 0.6667 0.8000 0.7273 80
loc 0.6667 0.2222 0.3333 9
date 0.0000 0.0000 0.0000 3
micro avg 0.8122 0.8528 0.8320 360
macro avg 0.6121 0.5643 0.5646 360
weighted avg 0.8080 0.8528 0.8259 360
2023-10-17 10:51:49,597 ----------------------------------------------------------------------------------------------------
|