File size: 24,017 Bytes
679e0aa |
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 |
2023-10-13 18:25:55,574 ----------------------------------------------------------------------------------------------------
2023-10-13 18:25:55,575 Model: "SequenceTagger(
(embeddings): TransformerWordEmbeddings(
(model): BertModel(
(embeddings): BertEmbeddings(
(word_embeddings): Embedding(32001, 768)
(position_embeddings): Embedding(512, 768)
(token_type_embeddings): Embedding(2, 768)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(encoder): BertEncoder(
(layer): ModuleList(
(0-11): 12 x BertLayer(
(attention): BertAttention(
(self): BertSelfAttention(
(query): Linear(in_features=768, out_features=768, bias=True)
(key): Linear(in_features=768, out_features=768, bias=True)
(value): Linear(in_features=768, out_features=768, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(output): BertSelfOutput(
(dense): Linear(in_features=768, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(intermediate): BertIntermediate(
(dense): Linear(in_features=768, out_features=3072, bias=True)
(intermediate_act_fn): GELUActivation()
)
(output): BertOutput(
(dense): Linear(in_features=3072, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
)
)
(pooler): BertPooler(
(dense): Linear(in_features=768, out_features=768, bias=True)
(activation): Tanh()
)
)
)
(locked_dropout): LockedDropout(p=0.5)
(linear): Linear(in_features=768, out_features=21, bias=True)
(loss_function): CrossEntropyLoss()
)"
2023-10-13 18:25:55,575 ----------------------------------------------------------------------------------------------------
2023-10-13 18:25:55,575 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
2023-10-13 18:25:55,575 ----------------------------------------------------------------------------------------------------
2023-10-13 18:25:55,575 Train: 5901 sentences
2023-10-13 18:25:55,575 (train_with_dev=False, train_with_test=False)
2023-10-13 18:25:55,576 ----------------------------------------------------------------------------------------------------
2023-10-13 18:25:55,576 Training Params:
2023-10-13 18:25:55,576 - learning_rate: "3e-05"
2023-10-13 18:25:55,576 - mini_batch_size: "8"
2023-10-13 18:25:55,576 - max_epochs: "10"
2023-10-13 18:25:55,576 - shuffle: "True"
2023-10-13 18:25:55,576 ----------------------------------------------------------------------------------------------------
2023-10-13 18:25:55,576 Plugins:
2023-10-13 18:25:55,576 - LinearScheduler | warmup_fraction: '0.1'
2023-10-13 18:25:55,576 ----------------------------------------------------------------------------------------------------
2023-10-13 18:25:55,576 Final evaluation on model from best epoch (best-model.pt)
2023-10-13 18:25:55,576 - metric: "('micro avg', 'f1-score')"
2023-10-13 18:25:55,576 ----------------------------------------------------------------------------------------------------
2023-10-13 18:25:55,576 Computation:
2023-10-13 18:25:55,576 - compute on device: cuda:0
2023-10-13 18:25:55,576 - embedding storage: none
2023-10-13 18:25:55,576 ----------------------------------------------------------------------------------------------------
2023-10-13 18:25:55,576 Model training base path: "hmbench-hipe2020/fr-dbmdz/bert-base-historic-multilingual-cased-bs8-wsFalse-e10-lr3e-05-poolingfirst-layers-1-crfFalse-5"
2023-10-13 18:25:55,576 ----------------------------------------------------------------------------------------------------
2023-10-13 18:25:55,576 ----------------------------------------------------------------------------------------------------
2023-10-13 18:26:00,188 epoch 1 - iter 73/738 - loss 2.77948811 - time (sec): 4.61 - samples/sec: 3504.97 - lr: 0.000003 - momentum: 0.000000
2023-10-13 18:26:04,956 epoch 1 - iter 146/738 - loss 1.83651071 - time (sec): 9.38 - samples/sec: 3464.52 - lr: 0.000006 - momentum: 0.000000
2023-10-13 18:26:09,786 epoch 1 - iter 219/738 - loss 1.38880604 - time (sec): 14.21 - samples/sec: 3420.58 - lr: 0.000009 - momentum: 0.000000
2023-10-13 18:26:14,516 epoch 1 - iter 292/738 - loss 1.13857844 - time (sec): 18.94 - samples/sec: 3407.12 - lr: 0.000012 - momentum: 0.000000
2023-10-13 18:26:19,737 epoch 1 - iter 365/738 - loss 0.98269160 - time (sec): 24.16 - samples/sec: 3360.28 - lr: 0.000015 - momentum: 0.000000
2023-10-13 18:26:24,365 epoch 1 - iter 438/738 - loss 0.87412221 - time (sec): 28.79 - samples/sec: 3348.74 - lr: 0.000018 - momentum: 0.000000
2023-10-13 18:26:29,369 epoch 1 - iter 511/738 - loss 0.78141679 - time (sec): 33.79 - samples/sec: 3367.22 - lr: 0.000021 - momentum: 0.000000
2023-10-13 18:26:34,377 epoch 1 - iter 584/738 - loss 0.70772215 - time (sec): 38.80 - samples/sec: 3383.66 - lr: 0.000024 - momentum: 0.000000
2023-10-13 18:26:39,145 epoch 1 - iter 657/738 - loss 0.65339163 - time (sec): 43.57 - samples/sec: 3387.69 - lr: 0.000027 - momentum: 0.000000
2023-10-13 18:26:44,294 epoch 1 - iter 730/738 - loss 0.60764205 - time (sec): 48.72 - samples/sec: 3380.12 - lr: 0.000030 - momentum: 0.000000
2023-10-13 18:26:44,855 ----------------------------------------------------------------------------------------------------
2023-10-13 18:26:44,856 EPOCH 1 done: loss 0.6037 - lr: 0.000030
2023-10-13 18:26:51,028 DEV : loss 0.15163348615169525 - f1-score (micro avg) 0.6833
2023-10-13 18:26:51,056 saving best model
2023-10-13 18:26:51,480 ----------------------------------------------------------------------------------------------------
2023-10-13 18:26:56,596 epoch 2 - iter 73/738 - loss 0.14905787 - time (sec): 5.11 - samples/sec: 3263.37 - lr: 0.000030 - momentum: 0.000000
2023-10-13 18:27:01,383 epoch 2 - iter 146/738 - loss 0.14670637 - time (sec): 9.90 - samples/sec: 3334.54 - lr: 0.000029 - momentum: 0.000000
2023-10-13 18:27:06,404 epoch 2 - iter 219/738 - loss 0.14929708 - time (sec): 14.92 - samples/sec: 3331.31 - lr: 0.000029 - momentum: 0.000000
2023-10-13 18:27:10,955 epoch 2 - iter 292/738 - loss 0.14176362 - time (sec): 19.47 - samples/sec: 3345.22 - lr: 0.000029 - momentum: 0.000000
2023-10-13 18:27:16,010 epoch 2 - iter 365/738 - loss 0.13682994 - time (sec): 24.53 - samples/sec: 3390.21 - lr: 0.000028 - momentum: 0.000000
2023-10-13 18:27:22,282 epoch 2 - iter 438/738 - loss 0.13774552 - time (sec): 30.80 - samples/sec: 3372.32 - lr: 0.000028 - momentum: 0.000000
2023-10-13 18:27:26,751 epoch 2 - iter 511/738 - loss 0.13433078 - time (sec): 35.27 - samples/sec: 3375.37 - lr: 0.000028 - momentum: 0.000000
2023-10-13 18:27:31,778 epoch 2 - iter 584/738 - loss 0.13450948 - time (sec): 40.30 - samples/sec: 3373.81 - lr: 0.000027 - momentum: 0.000000
2023-10-13 18:27:35,982 epoch 2 - iter 657/738 - loss 0.13257424 - time (sec): 44.50 - samples/sec: 3378.23 - lr: 0.000027 - momentum: 0.000000
2023-10-13 18:27:40,535 epoch 2 - iter 730/738 - loss 0.13177134 - time (sec): 49.05 - samples/sec: 3363.30 - lr: 0.000027 - momentum: 0.000000
2023-10-13 18:27:40,955 ----------------------------------------------------------------------------------------------------
2023-10-13 18:27:40,956 EPOCH 2 done: loss 0.1311 - lr: 0.000027
2023-10-13 18:27:52,212 DEV : loss 0.11227148026227951 - f1-score (micro avg) 0.792
2023-10-13 18:27:52,241 saving best model
2023-10-13 18:27:52,725 ----------------------------------------------------------------------------------------------------
2023-10-13 18:27:57,928 epoch 3 - iter 73/738 - loss 0.07307088 - time (sec): 5.20 - samples/sec: 3558.65 - lr: 0.000026 - momentum: 0.000000
2023-10-13 18:28:03,006 epoch 3 - iter 146/738 - loss 0.06721425 - time (sec): 10.27 - samples/sec: 3411.70 - lr: 0.000026 - momentum: 0.000000
2023-10-13 18:28:07,946 epoch 3 - iter 219/738 - loss 0.07015625 - time (sec): 15.21 - samples/sec: 3402.24 - lr: 0.000026 - momentum: 0.000000
2023-10-13 18:28:13,475 epoch 3 - iter 292/738 - loss 0.07577362 - time (sec): 20.74 - samples/sec: 3372.37 - lr: 0.000025 - momentum: 0.000000
2023-10-13 18:28:18,038 epoch 3 - iter 365/738 - loss 0.07658825 - time (sec): 25.31 - samples/sec: 3364.70 - lr: 0.000025 - momentum: 0.000000
2023-10-13 18:28:22,940 epoch 3 - iter 438/738 - loss 0.07349264 - time (sec): 30.21 - samples/sec: 3352.76 - lr: 0.000025 - momentum: 0.000000
2023-10-13 18:28:27,523 epoch 3 - iter 511/738 - loss 0.07289690 - time (sec): 34.79 - samples/sec: 3357.71 - lr: 0.000024 - momentum: 0.000000
2023-10-13 18:28:32,227 epoch 3 - iter 584/738 - loss 0.07223721 - time (sec): 39.49 - samples/sec: 3361.46 - lr: 0.000024 - momentum: 0.000000
2023-10-13 18:28:37,075 epoch 3 - iter 657/738 - loss 0.07202171 - time (sec): 44.34 - samples/sec: 3354.44 - lr: 0.000024 - momentum: 0.000000
2023-10-13 18:28:41,868 epoch 3 - iter 730/738 - loss 0.07214547 - time (sec): 49.14 - samples/sec: 3356.10 - lr: 0.000023 - momentum: 0.000000
2023-10-13 18:28:42,345 ----------------------------------------------------------------------------------------------------
2023-10-13 18:28:42,345 EPOCH 3 done: loss 0.0726 - lr: 0.000023
2023-10-13 18:28:53,535 DEV : loss 0.11804373562335968 - f1-score (micro avg) 0.8005
2023-10-13 18:28:53,563 saving best model
2023-10-13 18:28:54,108 ----------------------------------------------------------------------------------------------------
2023-10-13 18:28:59,136 epoch 4 - iter 73/738 - loss 0.04640357 - time (sec): 5.02 - samples/sec: 3178.15 - lr: 0.000023 - momentum: 0.000000
2023-10-13 18:29:04,912 epoch 4 - iter 146/738 - loss 0.05188936 - time (sec): 10.80 - samples/sec: 3286.27 - lr: 0.000023 - momentum: 0.000000
2023-10-13 18:29:10,063 epoch 4 - iter 219/738 - loss 0.04957858 - time (sec): 15.95 - samples/sec: 3226.45 - lr: 0.000022 - momentum: 0.000000
2023-10-13 18:29:14,644 epoch 4 - iter 292/738 - loss 0.05013571 - time (sec): 20.53 - samples/sec: 3217.01 - lr: 0.000022 - momentum: 0.000000
2023-10-13 18:29:19,570 epoch 4 - iter 365/738 - loss 0.04848137 - time (sec): 25.46 - samples/sec: 3251.80 - lr: 0.000022 - momentum: 0.000000
2023-10-13 18:29:24,227 epoch 4 - iter 438/738 - loss 0.04768082 - time (sec): 30.11 - samples/sec: 3283.31 - lr: 0.000021 - momentum: 0.000000
2023-10-13 18:29:28,806 epoch 4 - iter 511/738 - loss 0.04744479 - time (sec): 34.69 - samples/sec: 3284.39 - lr: 0.000021 - momentum: 0.000000
2023-10-13 18:29:33,378 epoch 4 - iter 584/738 - loss 0.04764204 - time (sec): 39.27 - samples/sec: 3303.23 - lr: 0.000021 - momentum: 0.000000
2023-10-13 18:29:38,231 epoch 4 - iter 657/738 - loss 0.04732965 - time (sec): 44.12 - samples/sec: 3300.01 - lr: 0.000020 - momentum: 0.000000
2023-10-13 18:29:43,735 epoch 4 - iter 730/738 - loss 0.04657406 - time (sec): 49.62 - samples/sec: 3319.62 - lr: 0.000020 - momentum: 0.000000
2023-10-13 18:29:44,198 ----------------------------------------------------------------------------------------------------
2023-10-13 18:29:44,198 EPOCH 4 done: loss 0.0466 - lr: 0.000020
2023-10-13 18:29:55,341 DEV : loss 0.14324556291103363 - f1-score (micro avg) 0.8153
2023-10-13 18:29:55,370 saving best model
2023-10-13 18:29:55,914 ----------------------------------------------------------------------------------------------------
2023-10-13 18:30:00,745 epoch 5 - iter 73/738 - loss 0.03463417 - time (sec): 4.83 - samples/sec: 3466.69 - lr: 0.000020 - momentum: 0.000000
2023-10-13 18:30:05,323 epoch 5 - iter 146/738 - loss 0.03424645 - time (sec): 9.40 - samples/sec: 3330.81 - lr: 0.000019 - momentum: 0.000000
2023-10-13 18:30:10,542 epoch 5 - iter 219/738 - loss 0.03250383 - time (sec): 14.62 - samples/sec: 3279.94 - lr: 0.000019 - momentum: 0.000000
2023-10-13 18:30:15,573 epoch 5 - iter 292/738 - loss 0.03238949 - time (sec): 19.65 - samples/sec: 3278.44 - lr: 0.000019 - momentum: 0.000000
2023-10-13 18:30:20,752 epoch 5 - iter 365/738 - loss 0.03263209 - time (sec): 24.83 - samples/sec: 3282.40 - lr: 0.000018 - momentum: 0.000000
2023-10-13 18:30:25,879 epoch 5 - iter 438/738 - loss 0.03210855 - time (sec): 29.96 - samples/sec: 3280.02 - lr: 0.000018 - momentum: 0.000000
2023-10-13 18:30:30,870 epoch 5 - iter 511/738 - loss 0.03349194 - time (sec): 34.95 - samples/sec: 3286.12 - lr: 0.000018 - momentum: 0.000000
2023-10-13 18:30:36,087 epoch 5 - iter 584/738 - loss 0.03354639 - time (sec): 40.17 - samples/sec: 3282.79 - lr: 0.000017 - momentum: 0.000000
2023-10-13 18:30:41,586 epoch 5 - iter 657/738 - loss 0.03422851 - time (sec): 45.67 - samples/sec: 3256.83 - lr: 0.000017 - momentum: 0.000000
2023-10-13 18:30:46,158 epoch 5 - iter 730/738 - loss 0.03467585 - time (sec): 50.24 - samples/sec: 3276.48 - lr: 0.000017 - momentum: 0.000000
2023-10-13 18:30:46,772 ----------------------------------------------------------------------------------------------------
2023-10-13 18:30:46,773 EPOCH 5 done: loss 0.0344 - lr: 0.000017
2023-10-13 18:30:57,925 DEV : loss 0.16191978752613068 - f1-score (micro avg) 0.8316
2023-10-13 18:30:57,954 saving best model
2023-10-13 18:30:58,448 ----------------------------------------------------------------------------------------------------
2023-10-13 18:31:03,446 epoch 6 - iter 73/738 - loss 0.03187475 - time (sec): 5.00 - samples/sec: 3395.25 - lr: 0.000016 - momentum: 0.000000
2023-10-13 18:31:08,103 epoch 6 - iter 146/738 - loss 0.02607637 - time (sec): 9.65 - samples/sec: 3340.93 - lr: 0.000016 - momentum: 0.000000
2023-10-13 18:31:12,977 epoch 6 - iter 219/738 - loss 0.02584962 - time (sec): 14.53 - samples/sec: 3330.61 - lr: 0.000016 - momentum: 0.000000
2023-10-13 18:31:17,890 epoch 6 - iter 292/738 - loss 0.02764999 - time (sec): 19.44 - samples/sec: 3322.47 - lr: 0.000015 - momentum: 0.000000
2023-10-13 18:31:22,652 epoch 6 - iter 365/738 - loss 0.02678067 - time (sec): 24.20 - samples/sec: 3311.91 - lr: 0.000015 - momentum: 0.000000
2023-10-13 18:31:27,144 epoch 6 - iter 438/738 - loss 0.02654707 - time (sec): 28.69 - samples/sec: 3322.25 - lr: 0.000015 - momentum: 0.000000
2023-10-13 18:31:32,359 epoch 6 - iter 511/738 - loss 0.02516894 - time (sec): 33.91 - samples/sec: 3334.53 - lr: 0.000014 - momentum: 0.000000
2023-10-13 18:31:37,209 epoch 6 - iter 584/738 - loss 0.02524705 - time (sec): 38.76 - samples/sec: 3337.29 - lr: 0.000014 - momentum: 0.000000
2023-10-13 18:31:42,168 epoch 6 - iter 657/738 - loss 0.02616650 - time (sec): 43.72 - samples/sec: 3335.83 - lr: 0.000014 - momentum: 0.000000
2023-10-13 18:31:47,146 epoch 6 - iter 730/738 - loss 0.02554515 - time (sec): 48.70 - samples/sec: 3369.83 - lr: 0.000013 - momentum: 0.000000
2023-10-13 18:31:47,871 ----------------------------------------------------------------------------------------------------
2023-10-13 18:31:47,871 EPOCH 6 done: loss 0.0256 - lr: 0.000013
2023-10-13 18:31:59,141 DEV : loss 0.18813824653625488 - f1-score (micro avg) 0.8172
2023-10-13 18:31:59,170 ----------------------------------------------------------------------------------------------------
2023-10-13 18:32:04,119 epoch 7 - iter 73/738 - loss 0.01945020 - time (sec): 4.95 - samples/sec: 3505.84 - lr: 0.000013 - momentum: 0.000000
2023-10-13 18:32:09,953 epoch 7 - iter 146/738 - loss 0.02190146 - time (sec): 10.78 - samples/sec: 3316.79 - lr: 0.000013 - momentum: 0.000000
2023-10-13 18:32:15,106 epoch 7 - iter 219/738 - loss 0.01858043 - time (sec): 15.93 - samples/sec: 3324.70 - lr: 0.000012 - momentum: 0.000000
2023-10-13 18:32:20,277 epoch 7 - iter 292/738 - loss 0.01910198 - time (sec): 21.11 - samples/sec: 3364.95 - lr: 0.000012 - momentum: 0.000000
2023-10-13 18:32:24,500 epoch 7 - iter 365/738 - loss 0.01972075 - time (sec): 25.33 - samples/sec: 3391.31 - lr: 0.000012 - momentum: 0.000000
2023-10-13 18:32:29,452 epoch 7 - iter 438/738 - loss 0.02062901 - time (sec): 30.28 - samples/sec: 3382.30 - lr: 0.000011 - momentum: 0.000000
2023-10-13 18:32:34,183 epoch 7 - iter 511/738 - loss 0.01984974 - time (sec): 35.01 - samples/sec: 3373.91 - lr: 0.000011 - momentum: 0.000000
2023-10-13 18:32:38,775 epoch 7 - iter 584/738 - loss 0.01906639 - time (sec): 39.60 - samples/sec: 3372.83 - lr: 0.000011 - momentum: 0.000000
2023-10-13 18:32:43,455 epoch 7 - iter 657/738 - loss 0.01891734 - time (sec): 44.28 - samples/sec: 3373.12 - lr: 0.000010 - momentum: 0.000000
2023-10-13 18:32:48,038 epoch 7 - iter 730/738 - loss 0.01895225 - time (sec): 48.87 - samples/sec: 3368.29 - lr: 0.000010 - momentum: 0.000000
2023-10-13 18:32:48,534 ----------------------------------------------------------------------------------------------------
2023-10-13 18:32:48,534 EPOCH 7 done: loss 0.0192 - lr: 0.000010
2023-10-13 18:32:59,738 DEV : loss 0.21069413423538208 - f1-score (micro avg) 0.8205
2023-10-13 18:32:59,768 ----------------------------------------------------------------------------------------------------
2023-10-13 18:33:04,738 epoch 8 - iter 73/738 - loss 0.01208919 - time (sec): 4.97 - samples/sec: 3546.99 - lr: 0.000010 - momentum: 0.000000
2023-10-13 18:33:09,349 epoch 8 - iter 146/738 - loss 0.01011562 - time (sec): 9.58 - samples/sec: 3428.49 - lr: 0.000009 - momentum: 0.000000
2023-10-13 18:33:14,711 epoch 8 - iter 219/738 - loss 0.01197427 - time (sec): 14.94 - samples/sec: 3453.24 - lr: 0.000009 - momentum: 0.000000
2023-10-13 18:33:19,729 epoch 8 - iter 292/738 - loss 0.01073791 - time (sec): 19.96 - samples/sec: 3356.40 - lr: 0.000009 - momentum: 0.000000
2023-10-13 18:33:24,003 epoch 8 - iter 365/738 - loss 0.01273368 - time (sec): 24.23 - samples/sec: 3359.97 - lr: 0.000008 - momentum: 0.000000
2023-10-13 18:33:28,829 epoch 8 - iter 438/738 - loss 0.01246504 - time (sec): 29.06 - samples/sec: 3347.32 - lr: 0.000008 - momentum: 0.000000
2023-10-13 18:33:33,827 epoch 8 - iter 511/738 - loss 0.01250396 - time (sec): 34.06 - samples/sec: 3368.55 - lr: 0.000008 - momentum: 0.000000
2023-10-13 18:33:38,771 epoch 8 - iter 584/738 - loss 0.01223467 - time (sec): 39.00 - samples/sec: 3356.88 - lr: 0.000007 - momentum: 0.000000
2023-10-13 18:33:43,688 epoch 8 - iter 657/738 - loss 0.01262949 - time (sec): 43.92 - samples/sec: 3350.21 - lr: 0.000007 - momentum: 0.000000
2023-10-13 18:33:48,706 epoch 8 - iter 730/738 - loss 0.01263386 - time (sec): 48.94 - samples/sec: 3352.69 - lr: 0.000007 - momentum: 0.000000
2023-10-13 18:33:49,394 ----------------------------------------------------------------------------------------------------
2023-10-13 18:33:49,395 EPOCH 8 done: loss 0.0125 - lr: 0.000007
2023-10-13 18:34:00,541 DEV : loss 0.1914064884185791 - f1-score (micro avg) 0.8327
2023-10-13 18:34:00,571 saving best model
2023-10-13 18:34:01,118 ----------------------------------------------------------------------------------------------------
2023-10-13 18:34:05,943 epoch 9 - iter 73/738 - loss 0.01636345 - time (sec): 4.82 - samples/sec: 3310.78 - lr: 0.000006 - momentum: 0.000000
2023-10-13 18:34:10,703 epoch 9 - iter 146/738 - loss 0.01343032 - time (sec): 9.58 - samples/sec: 3354.67 - lr: 0.000006 - momentum: 0.000000
2023-10-13 18:34:15,064 epoch 9 - iter 219/738 - loss 0.01024467 - time (sec): 13.94 - samples/sec: 3361.07 - lr: 0.000006 - momentum: 0.000000
2023-10-13 18:34:20,316 epoch 9 - iter 292/738 - loss 0.01040151 - time (sec): 19.20 - samples/sec: 3322.59 - lr: 0.000005 - momentum: 0.000000
2023-10-13 18:34:25,011 epoch 9 - iter 365/738 - loss 0.00980650 - time (sec): 23.89 - samples/sec: 3318.50 - lr: 0.000005 - momentum: 0.000000
2023-10-13 18:34:29,967 epoch 9 - iter 438/738 - loss 0.00914193 - time (sec): 28.85 - samples/sec: 3308.67 - lr: 0.000005 - momentum: 0.000000
2023-10-13 18:34:35,040 epoch 9 - iter 511/738 - loss 0.00900502 - time (sec): 33.92 - samples/sec: 3333.78 - lr: 0.000004 - momentum: 0.000000
2023-10-13 18:34:40,548 epoch 9 - iter 584/738 - loss 0.00995733 - time (sec): 39.43 - samples/sec: 3334.23 - lr: 0.000004 - momentum: 0.000000
2023-10-13 18:34:45,124 epoch 9 - iter 657/738 - loss 0.00904693 - time (sec): 44.00 - samples/sec: 3336.14 - lr: 0.000004 - momentum: 0.000000
2023-10-13 18:34:49,897 epoch 9 - iter 730/738 - loss 0.00901017 - time (sec): 48.78 - samples/sec: 3355.24 - lr: 0.000003 - momentum: 0.000000
2023-10-13 18:34:50,765 ----------------------------------------------------------------------------------------------------
2023-10-13 18:34:50,765 EPOCH 9 done: loss 0.0089 - lr: 0.000003
2023-10-13 18:35:02,006 DEV : loss 0.20760603249073029 - f1-score (micro avg) 0.8264
2023-10-13 18:35:02,036 ----------------------------------------------------------------------------------------------------
2023-10-13 18:35:06,611 epoch 10 - iter 73/738 - loss 0.00505717 - time (sec): 4.57 - samples/sec: 3318.85 - lr: 0.000003 - momentum: 0.000000
2023-10-13 18:35:11,830 epoch 10 - iter 146/738 - loss 0.00742873 - time (sec): 9.79 - samples/sec: 3315.90 - lr: 0.000003 - momentum: 0.000000
2023-10-13 18:35:17,652 epoch 10 - iter 219/738 - loss 0.00650006 - time (sec): 15.62 - samples/sec: 3154.32 - lr: 0.000002 - momentum: 0.000000
2023-10-13 18:35:22,973 epoch 10 - iter 292/738 - loss 0.00677815 - time (sec): 20.94 - samples/sec: 3170.26 - lr: 0.000002 - momentum: 0.000000
2023-10-13 18:35:28,293 epoch 10 - iter 365/738 - loss 0.00621081 - time (sec): 26.26 - samples/sec: 3208.44 - lr: 0.000002 - momentum: 0.000000
2023-10-13 18:35:32,788 epoch 10 - iter 438/738 - loss 0.00657852 - time (sec): 30.75 - samples/sec: 3245.23 - lr: 0.000001 - momentum: 0.000000
2023-10-13 18:35:37,205 epoch 10 - iter 511/738 - loss 0.00821402 - time (sec): 35.17 - samples/sec: 3273.58 - lr: 0.000001 - momentum: 0.000000
2023-10-13 18:35:42,271 epoch 10 - iter 584/738 - loss 0.00795187 - time (sec): 40.23 - samples/sec: 3268.92 - lr: 0.000001 - momentum: 0.000000
2023-10-13 18:35:47,325 epoch 10 - iter 657/738 - loss 0.00749875 - time (sec): 45.29 - samples/sec: 3296.73 - lr: 0.000000 - momentum: 0.000000
2023-10-13 18:35:51,858 epoch 10 - iter 730/738 - loss 0.00725227 - time (sec): 49.82 - samples/sec: 3308.48 - lr: 0.000000 - momentum: 0.000000
2023-10-13 18:35:52,320 ----------------------------------------------------------------------------------------------------
2023-10-13 18:35:52,320 EPOCH 10 done: loss 0.0072 - lr: 0.000000
2023-10-13 18:36:03,499 DEV : loss 0.20923016965389252 - f1-score (micro avg) 0.8323
2023-10-13 18:36:03,920 ----------------------------------------------------------------------------------------------------
2023-10-13 18:36:03,922 Loading model from best epoch ...
2023-10-13 18:36:05,272 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
2023-10-13 18:36:11,234
Results:
- F-score (micro) 0.8013
- F-score (macro) 0.7115
- Accuracy 0.6915
By class:
precision recall f1-score support
loc 0.8692 0.8671 0.8681 858
pers 0.7301 0.8212 0.7730 537
org 0.6170 0.6591 0.6374 132
prod 0.7119 0.6885 0.7000 61
time 0.5500 0.6111 0.5789 54
micro avg 0.7831 0.8203 0.8013 1642
macro avg 0.6956 0.7294 0.7115 1642
weighted avg 0.7871 0.8203 0.8027 1642
2023-10-13 18:36:11,234 ----------------------------------------------------------------------------------------------------
|