File size: 24,062 Bytes
17984cb |
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 |
2023-10-25 14:22:05,423 ----------------------------------------------------------------------------------------------------
2023-10-25 14:22:05,424 Model: "SequenceTagger(
(embeddings): TransformerWordEmbeddings(
(model): BertModel(
(embeddings): BertEmbeddings(
(word_embeddings): Embedding(64001, 768)
(position_embeddings): Embedding(512, 768)
(token_type_embeddings): Embedding(2, 768)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(encoder): BertEncoder(
(layer): ModuleList(
(0-11): 12 x BertLayer(
(attention): BertAttention(
(self): BertSelfAttention(
(query): Linear(in_features=768, out_features=768, bias=True)
(key): Linear(in_features=768, out_features=768, bias=True)
(value): Linear(in_features=768, out_features=768, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(output): BertSelfOutput(
(dense): Linear(in_features=768, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(intermediate): BertIntermediate(
(dense): Linear(in_features=768, out_features=3072, bias=True)
(intermediate_act_fn): GELUActivation()
)
(output): BertOutput(
(dense): Linear(in_features=3072, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
)
)
(pooler): BertPooler(
(dense): Linear(in_features=768, out_features=768, bias=True)
(activation): Tanh()
)
)
)
(locked_dropout): LockedDropout(p=0.5)
(linear): Linear(in_features=768, out_features=17, bias=True)
(loss_function): CrossEntropyLoss()
)"
2023-10-25 14:22:05,424 ----------------------------------------------------------------------------------------------------
2023-10-25 14:22:05,424 MultiCorpus: 7142 train + 698 dev + 2570 test sentences
- NER_HIPE_2022 Corpus: 7142 train + 698 dev + 2570 test sentences - /root/.flair/datasets/ner_hipe_2022/v2.1/newseye/fr/with_doc_seperator
2023-10-25 14:22:05,424 ----------------------------------------------------------------------------------------------------
2023-10-25 14:22:05,424 Train: 7142 sentences
2023-10-25 14:22:05,424 (train_with_dev=False, train_with_test=False)
2023-10-25 14:22:05,424 ----------------------------------------------------------------------------------------------------
2023-10-25 14:22:05,424 Training Params:
2023-10-25 14:22:05,425 - learning_rate: "5e-05"
2023-10-25 14:22:05,425 - mini_batch_size: "8"
2023-10-25 14:22:05,425 - max_epochs: "10"
2023-10-25 14:22:05,425 - shuffle: "True"
2023-10-25 14:22:05,425 ----------------------------------------------------------------------------------------------------
2023-10-25 14:22:05,425 Plugins:
2023-10-25 14:22:05,425 - TensorboardLogger
2023-10-25 14:22:05,425 - LinearScheduler | warmup_fraction: '0.1'
2023-10-25 14:22:05,425 ----------------------------------------------------------------------------------------------------
2023-10-25 14:22:05,425 Final evaluation on model from best epoch (best-model.pt)
2023-10-25 14:22:05,425 - metric: "('micro avg', 'f1-score')"
2023-10-25 14:22:05,425 ----------------------------------------------------------------------------------------------------
2023-10-25 14:22:05,425 Computation:
2023-10-25 14:22:05,425 - compute on device: cuda:0
2023-10-25 14:22:05,425 - embedding storage: none
2023-10-25 14:22:05,425 ----------------------------------------------------------------------------------------------------
2023-10-25 14:22:05,425 Model training base path: "hmbench-newseye/fr-dbmdz/bert-base-historic-multilingual-64k-td-cased-bs8-wsFalse-e10-lr5e-05-poolingfirst-layers-1-crfFalse-1"
2023-10-25 14:22:05,425 ----------------------------------------------------------------------------------------------------
2023-10-25 14:22:05,425 ----------------------------------------------------------------------------------------------------
2023-10-25 14:22:05,425 Logging anything other than scalars to TensorBoard is currently not supported.
2023-10-25 14:22:11,296 epoch 1 - iter 89/893 - loss 2.22174821 - time (sec): 5.87 - samples/sec: 4255.73 - lr: 0.000005 - momentum: 0.000000
2023-10-25 14:22:17,176 epoch 1 - iter 178/893 - loss 1.34900675 - time (sec): 11.75 - samples/sec: 4297.23 - lr: 0.000010 - momentum: 0.000000
2023-10-25 14:22:23,077 epoch 1 - iter 267/893 - loss 1.01310745 - time (sec): 17.65 - samples/sec: 4246.13 - lr: 0.000015 - momentum: 0.000000
2023-10-25 14:22:28,941 epoch 1 - iter 356/893 - loss 0.81673971 - time (sec): 23.52 - samples/sec: 4266.25 - lr: 0.000020 - momentum: 0.000000
2023-10-25 14:22:34,652 epoch 1 - iter 445/893 - loss 0.69153729 - time (sec): 29.23 - samples/sec: 4280.94 - lr: 0.000025 - momentum: 0.000000
2023-10-25 14:22:40,228 epoch 1 - iter 534/893 - loss 0.60855944 - time (sec): 34.80 - samples/sec: 4283.75 - lr: 0.000030 - momentum: 0.000000
2023-10-25 14:22:46,157 epoch 1 - iter 623/893 - loss 0.53908676 - time (sec): 40.73 - samples/sec: 4303.65 - lr: 0.000035 - momentum: 0.000000
2023-10-25 14:22:51,797 epoch 1 - iter 712/893 - loss 0.49305012 - time (sec): 46.37 - samples/sec: 4290.84 - lr: 0.000040 - momentum: 0.000000
2023-10-25 14:22:57,456 epoch 1 - iter 801/893 - loss 0.45348166 - time (sec): 52.03 - samples/sec: 4296.11 - lr: 0.000045 - momentum: 0.000000
2023-10-25 14:23:03,344 epoch 1 - iter 890/893 - loss 0.42527446 - time (sec): 57.92 - samples/sec: 4279.37 - lr: 0.000050 - momentum: 0.000000
2023-10-25 14:23:03,553 ----------------------------------------------------------------------------------------------------
2023-10-25 14:23:03,554 EPOCH 1 done: loss 0.4241 - lr: 0.000050
2023-10-25 14:23:07,042 DEV : loss 0.10224700719118118 - f1-score (micro avg) 0.7363
2023-10-25 14:23:07,064 saving best model
2023-10-25 14:23:07,584 ----------------------------------------------------------------------------------------------------
2023-10-25 14:23:13,537 epoch 2 - iter 89/893 - loss 0.10807083 - time (sec): 5.95 - samples/sec: 4146.32 - lr: 0.000049 - momentum: 0.000000
2023-10-25 14:23:19,506 epoch 2 - iter 178/893 - loss 0.10717817 - time (sec): 11.92 - samples/sec: 4148.68 - lr: 0.000049 - momentum: 0.000000
2023-10-25 14:23:25,279 epoch 2 - iter 267/893 - loss 0.10936825 - time (sec): 17.69 - samples/sec: 4151.97 - lr: 0.000048 - momentum: 0.000000
2023-10-25 14:23:31,126 epoch 2 - iter 356/893 - loss 0.11237855 - time (sec): 23.54 - samples/sec: 4166.50 - lr: 0.000048 - momentum: 0.000000
2023-10-25 14:23:36,586 epoch 2 - iter 445/893 - loss 0.10585519 - time (sec): 29.00 - samples/sec: 4220.66 - lr: 0.000047 - momentum: 0.000000
2023-10-25 14:23:42,078 epoch 2 - iter 534/893 - loss 0.10593438 - time (sec): 34.49 - samples/sec: 4248.78 - lr: 0.000047 - momentum: 0.000000
2023-10-25 14:23:47,781 epoch 2 - iter 623/893 - loss 0.10798060 - time (sec): 40.20 - samples/sec: 4281.71 - lr: 0.000046 - momentum: 0.000000
2023-10-25 14:23:53,567 epoch 2 - iter 712/893 - loss 0.10704577 - time (sec): 45.98 - samples/sec: 4307.87 - lr: 0.000046 - momentum: 0.000000
2023-10-25 14:23:59,120 epoch 2 - iter 801/893 - loss 0.10429214 - time (sec): 51.53 - samples/sec: 4347.28 - lr: 0.000045 - momentum: 0.000000
2023-10-25 14:24:04,739 epoch 2 - iter 890/893 - loss 0.10359464 - time (sec): 57.15 - samples/sec: 4337.39 - lr: 0.000044 - momentum: 0.000000
2023-10-25 14:24:04,955 ----------------------------------------------------------------------------------------------------
2023-10-25 14:24:04,955 EPOCH 2 done: loss 0.1033 - lr: 0.000044
2023-10-25 14:24:10,135 DEV : loss 0.0969092845916748 - f1-score (micro avg) 0.7786
2023-10-25 14:24:10,157 saving best model
2023-10-25 14:24:10,855 ----------------------------------------------------------------------------------------------------
2023-10-25 14:24:16,400 epoch 3 - iter 89/893 - loss 0.05486912 - time (sec): 5.54 - samples/sec: 4434.09 - lr: 0.000044 - momentum: 0.000000
2023-10-25 14:24:22,119 epoch 3 - iter 178/893 - loss 0.06296271 - time (sec): 11.26 - samples/sec: 4548.70 - lr: 0.000043 - momentum: 0.000000
2023-10-25 14:24:27,693 epoch 3 - iter 267/893 - loss 0.06237710 - time (sec): 16.83 - samples/sec: 4480.75 - lr: 0.000043 - momentum: 0.000000
2023-10-25 14:24:33,190 epoch 3 - iter 356/893 - loss 0.07078572 - time (sec): 22.33 - samples/sec: 4456.89 - lr: 0.000042 - momentum: 0.000000
2023-10-25 14:24:38,702 epoch 3 - iter 445/893 - loss 0.06966703 - time (sec): 27.84 - samples/sec: 4421.33 - lr: 0.000042 - momentum: 0.000000
2023-10-25 14:24:44,432 epoch 3 - iter 534/893 - loss 0.06853292 - time (sec): 33.57 - samples/sec: 4407.07 - lr: 0.000041 - momentum: 0.000000
2023-10-25 14:24:50,191 epoch 3 - iter 623/893 - loss 0.06693004 - time (sec): 39.33 - samples/sec: 4415.91 - lr: 0.000041 - momentum: 0.000000
2023-10-25 14:24:55,886 epoch 3 - iter 712/893 - loss 0.06730109 - time (sec): 45.03 - samples/sec: 4413.01 - lr: 0.000040 - momentum: 0.000000
2023-10-25 14:25:01,340 epoch 3 - iter 801/893 - loss 0.06610240 - time (sec): 50.48 - samples/sec: 4414.05 - lr: 0.000039 - momentum: 0.000000
2023-10-25 14:25:06,891 epoch 3 - iter 890/893 - loss 0.06743181 - time (sec): 56.03 - samples/sec: 4428.50 - lr: 0.000039 - momentum: 0.000000
2023-10-25 14:25:07,061 ----------------------------------------------------------------------------------------------------
2023-10-25 14:25:07,061 EPOCH 3 done: loss 0.0674 - lr: 0.000039
2023-10-25 14:25:10,865 DEV : loss 0.11728406697511673 - f1-score (micro avg) 0.7507
2023-10-25 14:25:10,887 ----------------------------------------------------------------------------------------------------
2023-10-25 14:25:16,941 epoch 4 - iter 89/893 - loss 0.05172042 - time (sec): 6.05 - samples/sec: 4265.61 - lr: 0.000038 - momentum: 0.000000
2023-10-25 14:25:22,824 epoch 4 - iter 178/893 - loss 0.04890968 - time (sec): 11.93 - samples/sec: 4188.03 - lr: 0.000038 - momentum: 0.000000
2023-10-25 14:25:28,627 epoch 4 - iter 267/893 - loss 0.04835057 - time (sec): 17.74 - samples/sec: 4153.56 - lr: 0.000037 - momentum: 0.000000
2023-10-25 14:25:34,625 epoch 4 - iter 356/893 - loss 0.04648503 - time (sec): 23.74 - samples/sec: 4206.05 - lr: 0.000037 - momentum: 0.000000
2023-10-25 14:25:41,203 epoch 4 - iter 445/893 - loss 0.04728395 - time (sec): 30.31 - samples/sec: 4135.62 - lr: 0.000036 - momentum: 0.000000
2023-10-25 14:25:46,943 epoch 4 - iter 534/893 - loss 0.04862880 - time (sec): 36.05 - samples/sec: 4139.07 - lr: 0.000036 - momentum: 0.000000
2023-10-25 14:25:52,737 epoch 4 - iter 623/893 - loss 0.04803766 - time (sec): 41.85 - samples/sec: 4149.38 - lr: 0.000035 - momentum: 0.000000
2023-10-25 14:25:58,381 epoch 4 - iter 712/893 - loss 0.04864147 - time (sec): 47.49 - samples/sec: 4175.34 - lr: 0.000034 - momentum: 0.000000
2023-10-25 14:26:04,318 epoch 4 - iter 801/893 - loss 0.04871377 - time (sec): 53.43 - samples/sec: 4193.31 - lr: 0.000034 - momentum: 0.000000
2023-10-25 14:26:09,795 epoch 4 - iter 890/893 - loss 0.04850206 - time (sec): 58.91 - samples/sec: 4210.00 - lr: 0.000033 - momentum: 0.000000
2023-10-25 14:26:09,973 ----------------------------------------------------------------------------------------------------
2023-10-25 14:26:09,973 EPOCH 4 done: loss 0.0486 - lr: 0.000033
2023-10-25 14:26:14,041 DEV : loss 0.15373246371746063 - f1-score (micro avg) 0.7653
2023-10-25 14:26:14,065 ----------------------------------------------------------------------------------------------------
2023-10-25 14:26:20,047 epoch 5 - iter 89/893 - loss 0.03855375 - time (sec): 5.98 - samples/sec: 4417.22 - lr: 0.000033 - momentum: 0.000000
2023-10-25 14:26:25,827 epoch 5 - iter 178/893 - loss 0.04024206 - time (sec): 11.76 - samples/sec: 4380.57 - lr: 0.000032 - momentum: 0.000000
2023-10-25 14:26:31,615 epoch 5 - iter 267/893 - loss 0.03624782 - time (sec): 17.55 - samples/sec: 4351.23 - lr: 0.000032 - momentum: 0.000000
2023-10-25 14:26:37,230 epoch 5 - iter 356/893 - loss 0.03803371 - time (sec): 23.16 - samples/sec: 4321.77 - lr: 0.000031 - momentum: 0.000000
2023-10-25 14:26:42,808 epoch 5 - iter 445/893 - loss 0.03757448 - time (sec): 28.74 - samples/sec: 4285.27 - lr: 0.000031 - momentum: 0.000000
2023-10-25 14:26:48,363 epoch 5 - iter 534/893 - loss 0.03668820 - time (sec): 34.30 - samples/sec: 4293.91 - lr: 0.000030 - momentum: 0.000000
2023-10-25 14:26:53,894 epoch 5 - iter 623/893 - loss 0.03705404 - time (sec): 39.83 - samples/sec: 4274.67 - lr: 0.000029 - momentum: 0.000000
2023-10-25 14:26:59,729 epoch 5 - iter 712/893 - loss 0.03634319 - time (sec): 45.66 - samples/sec: 4325.05 - lr: 0.000029 - momentum: 0.000000
2023-10-25 14:27:05,488 epoch 5 - iter 801/893 - loss 0.03547607 - time (sec): 51.42 - samples/sec: 4340.40 - lr: 0.000028 - momentum: 0.000000
2023-10-25 14:27:10,939 epoch 5 - iter 890/893 - loss 0.03563294 - time (sec): 56.87 - samples/sec: 4360.64 - lr: 0.000028 - momentum: 0.000000
2023-10-25 14:27:11,109 ----------------------------------------------------------------------------------------------------
2023-10-25 14:27:11,109 EPOCH 5 done: loss 0.0355 - lr: 0.000028
2023-10-25 14:27:15,806 DEV : loss 0.16320814192295074 - f1-score (micro avg) 0.7874
2023-10-25 14:27:15,826 saving best model
2023-10-25 14:27:16,529 ----------------------------------------------------------------------------------------------------
2023-10-25 14:27:22,218 epoch 6 - iter 89/893 - loss 0.02617113 - time (sec): 5.69 - samples/sec: 4700.94 - lr: 0.000027 - momentum: 0.000000
2023-10-25 14:27:28,467 epoch 6 - iter 178/893 - loss 0.02520431 - time (sec): 11.94 - samples/sec: 4332.41 - lr: 0.000027 - momentum: 0.000000
2023-10-25 14:27:33,984 epoch 6 - iter 267/893 - loss 0.02638010 - time (sec): 17.45 - samples/sec: 4320.59 - lr: 0.000026 - momentum: 0.000000
2023-10-25 14:27:39,878 epoch 6 - iter 356/893 - loss 0.02643239 - time (sec): 23.35 - samples/sec: 4349.87 - lr: 0.000026 - momentum: 0.000000
2023-10-25 14:27:45,656 epoch 6 - iter 445/893 - loss 0.02697284 - time (sec): 29.13 - samples/sec: 4332.53 - lr: 0.000025 - momentum: 0.000000
2023-10-25 14:27:51,208 epoch 6 - iter 534/893 - loss 0.02717458 - time (sec): 34.68 - samples/sec: 4355.89 - lr: 0.000024 - momentum: 0.000000
2023-10-25 14:27:56,893 epoch 6 - iter 623/893 - loss 0.02638170 - time (sec): 40.36 - samples/sec: 4340.97 - lr: 0.000024 - momentum: 0.000000
2023-10-25 14:28:02,464 epoch 6 - iter 712/893 - loss 0.02646998 - time (sec): 45.93 - samples/sec: 4345.70 - lr: 0.000023 - momentum: 0.000000
2023-10-25 14:28:08,304 epoch 6 - iter 801/893 - loss 0.02574741 - time (sec): 51.77 - samples/sec: 4344.49 - lr: 0.000023 - momentum: 0.000000
2023-10-25 14:28:13,959 epoch 6 - iter 890/893 - loss 0.02590536 - time (sec): 57.43 - samples/sec: 4318.36 - lr: 0.000022 - momentum: 0.000000
2023-10-25 14:28:14,125 ----------------------------------------------------------------------------------------------------
2023-10-25 14:28:14,126 EPOCH 6 done: loss 0.0259 - lr: 0.000022
2023-10-25 14:28:17,933 DEV : loss 0.18096403777599335 - f1-score (micro avg) 0.7885
2023-10-25 14:28:17,953 saving best model
2023-10-25 14:28:18,694 ----------------------------------------------------------------------------------------------------
2023-10-25 14:28:25,439 epoch 7 - iter 89/893 - loss 0.01678159 - time (sec): 6.74 - samples/sec: 3409.22 - lr: 0.000022 - momentum: 0.000000
2023-10-25 14:28:31,190 epoch 7 - iter 178/893 - loss 0.02086138 - time (sec): 12.49 - samples/sec: 3832.17 - lr: 0.000021 - momentum: 0.000000
2023-10-25 14:28:37,291 epoch 7 - iter 267/893 - loss 0.02080217 - time (sec): 18.59 - samples/sec: 4067.89 - lr: 0.000021 - momentum: 0.000000
2023-10-25 14:28:43,102 epoch 7 - iter 356/893 - loss 0.02143983 - time (sec): 24.40 - samples/sec: 4179.14 - lr: 0.000020 - momentum: 0.000000
2023-10-25 14:28:48,654 epoch 7 - iter 445/893 - loss 0.02119625 - time (sec): 29.96 - samples/sec: 4154.44 - lr: 0.000019 - momentum: 0.000000
2023-10-25 14:28:54,735 epoch 7 - iter 534/893 - loss 0.02007361 - time (sec): 36.04 - samples/sec: 4119.27 - lr: 0.000019 - momentum: 0.000000
2023-10-25 14:29:00,403 epoch 7 - iter 623/893 - loss 0.02020608 - time (sec): 41.71 - samples/sec: 4157.17 - lr: 0.000018 - momentum: 0.000000
2023-10-25 14:29:06,389 epoch 7 - iter 712/893 - loss 0.02032914 - time (sec): 47.69 - samples/sec: 4157.75 - lr: 0.000018 - momentum: 0.000000
2023-10-25 14:29:12,245 epoch 7 - iter 801/893 - loss 0.02130873 - time (sec): 53.55 - samples/sec: 4154.56 - lr: 0.000017 - momentum: 0.000000
2023-10-25 14:29:18,268 epoch 7 - iter 890/893 - loss 0.02140375 - time (sec): 59.57 - samples/sec: 4164.21 - lr: 0.000017 - momentum: 0.000000
2023-10-25 14:29:18,452 ----------------------------------------------------------------------------------------------------
2023-10-25 14:29:18,452 EPOCH 7 done: loss 0.0214 - lr: 0.000017
2023-10-25 14:29:22,302 DEV : loss 0.19584135711193085 - f1-score (micro avg) 0.7874
2023-10-25 14:29:22,325 ----------------------------------------------------------------------------------------------------
2023-10-25 14:29:28,254 epoch 8 - iter 89/893 - loss 0.01269647 - time (sec): 5.93 - samples/sec: 4175.01 - lr: 0.000016 - momentum: 0.000000
2023-10-25 14:29:34,170 epoch 8 - iter 178/893 - loss 0.01103982 - time (sec): 11.84 - samples/sec: 4330.76 - lr: 0.000016 - momentum: 0.000000
2023-10-25 14:29:40,172 epoch 8 - iter 267/893 - loss 0.01056955 - time (sec): 17.84 - samples/sec: 4320.95 - lr: 0.000015 - momentum: 0.000000
2023-10-25 14:29:46,086 epoch 8 - iter 356/893 - loss 0.01165606 - time (sec): 23.76 - samples/sec: 4333.99 - lr: 0.000014 - momentum: 0.000000
2023-10-25 14:29:51,817 epoch 8 - iter 445/893 - loss 0.01217482 - time (sec): 29.49 - samples/sec: 4280.96 - lr: 0.000014 - momentum: 0.000000
2023-10-25 14:29:57,619 epoch 8 - iter 534/893 - loss 0.01360158 - time (sec): 35.29 - samples/sec: 4255.30 - lr: 0.000013 - momentum: 0.000000
2023-10-25 14:30:03,211 epoch 8 - iter 623/893 - loss 0.01358451 - time (sec): 40.88 - samples/sec: 4280.98 - lr: 0.000013 - momentum: 0.000000
2023-10-25 14:30:09,010 epoch 8 - iter 712/893 - loss 0.01326304 - time (sec): 46.68 - samples/sec: 4279.43 - lr: 0.000012 - momentum: 0.000000
2023-10-25 14:30:14,815 epoch 8 - iter 801/893 - loss 0.01363656 - time (sec): 52.49 - samples/sec: 4280.60 - lr: 0.000012 - momentum: 0.000000
2023-10-25 14:30:20,749 epoch 8 - iter 890/893 - loss 0.01344285 - time (sec): 58.42 - samples/sec: 4244.81 - lr: 0.000011 - momentum: 0.000000
2023-10-25 14:30:20,929 ----------------------------------------------------------------------------------------------------
2023-10-25 14:30:20,929 EPOCH 8 done: loss 0.0134 - lr: 0.000011
2023-10-25 14:30:26,071 DEV : loss 0.21516965329647064 - f1-score (micro avg) 0.796
2023-10-25 14:30:26,096 saving best model
2023-10-25 14:30:26,883 ----------------------------------------------------------------------------------------------------
2023-10-25 14:30:32,597 epoch 9 - iter 89/893 - loss 0.01268252 - time (sec): 5.71 - samples/sec: 4284.31 - lr: 0.000011 - momentum: 0.000000
2023-10-25 14:30:38,584 epoch 9 - iter 178/893 - loss 0.01397325 - time (sec): 11.70 - samples/sec: 4298.37 - lr: 0.000010 - momentum: 0.000000
2023-10-25 14:30:44,380 epoch 9 - iter 267/893 - loss 0.01243272 - time (sec): 17.49 - samples/sec: 4354.73 - lr: 0.000009 - momentum: 0.000000
2023-10-25 14:30:49,932 epoch 9 - iter 356/893 - loss 0.01305437 - time (sec): 23.05 - samples/sec: 4343.10 - lr: 0.000009 - momentum: 0.000000
2023-10-25 14:30:55,731 epoch 9 - iter 445/893 - loss 0.01171325 - time (sec): 28.85 - samples/sec: 4364.70 - lr: 0.000008 - momentum: 0.000000
2023-10-25 14:31:01,450 epoch 9 - iter 534/893 - loss 0.01082927 - time (sec): 34.56 - samples/sec: 4348.58 - lr: 0.000008 - momentum: 0.000000
2023-10-25 14:31:07,263 epoch 9 - iter 623/893 - loss 0.01010485 - time (sec): 40.38 - samples/sec: 4345.62 - lr: 0.000007 - momentum: 0.000000
2023-10-25 14:31:12,771 epoch 9 - iter 712/893 - loss 0.01004408 - time (sec): 45.88 - samples/sec: 4345.42 - lr: 0.000007 - momentum: 0.000000
2023-10-25 14:31:18,573 epoch 9 - iter 801/893 - loss 0.01006474 - time (sec): 51.69 - samples/sec: 4340.94 - lr: 0.000006 - momentum: 0.000000
2023-10-25 14:31:24,256 epoch 9 - iter 890/893 - loss 0.01007528 - time (sec): 57.37 - samples/sec: 4323.30 - lr: 0.000006 - momentum: 0.000000
2023-10-25 14:31:24,441 ----------------------------------------------------------------------------------------------------
2023-10-25 14:31:24,441 EPOCH 9 done: loss 0.0101 - lr: 0.000006
2023-10-25 14:31:29,063 DEV : loss 0.23022107779979706 - f1-score (micro avg) 0.7853
2023-10-25 14:31:29,084 ----------------------------------------------------------------------------------------------------
2023-10-25 14:31:34,994 epoch 10 - iter 89/893 - loss 0.00710690 - time (sec): 5.91 - samples/sec: 4371.85 - lr: 0.000005 - momentum: 0.000000
2023-10-25 14:31:40,814 epoch 10 - iter 178/893 - loss 0.00583517 - time (sec): 11.73 - samples/sec: 4378.00 - lr: 0.000004 - momentum: 0.000000
2023-10-25 14:31:46,500 epoch 10 - iter 267/893 - loss 0.00596312 - time (sec): 17.41 - samples/sec: 4319.07 - lr: 0.000004 - momentum: 0.000000
2023-10-25 14:31:52,284 epoch 10 - iter 356/893 - loss 0.00716220 - time (sec): 23.20 - samples/sec: 4331.80 - lr: 0.000003 - momentum: 0.000000
2023-10-25 14:31:57,828 epoch 10 - iter 445/893 - loss 0.00782994 - time (sec): 28.74 - samples/sec: 4317.08 - lr: 0.000003 - momentum: 0.000000
2023-10-25 14:32:03,404 epoch 10 - iter 534/893 - loss 0.00818274 - time (sec): 34.32 - samples/sec: 4315.55 - lr: 0.000002 - momentum: 0.000000
2023-10-25 14:32:09,212 epoch 10 - iter 623/893 - loss 0.00842830 - time (sec): 40.13 - samples/sec: 4322.39 - lr: 0.000002 - momentum: 0.000000
2023-10-25 14:32:15,164 epoch 10 - iter 712/893 - loss 0.00783075 - time (sec): 46.08 - samples/sec: 4317.21 - lr: 0.000001 - momentum: 0.000000
2023-10-25 14:32:20,786 epoch 10 - iter 801/893 - loss 0.00742591 - time (sec): 51.70 - samples/sec: 4318.44 - lr: 0.000001 - momentum: 0.000000
2023-10-25 14:32:26,431 epoch 10 - iter 890/893 - loss 0.00733109 - time (sec): 57.35 - samples/sec: 4320.96 - lr: 0.000000 - momentum: 0.000000
2023-10-25 14:32:26,641 ----------------------------------------------------------------------------------------------------
2023-10-25 14:32:26,641 EPOCH 10 done: loss 0.0073 - lr: 0.000000
2023-10-25 14:32:30,523 DEV : loss 0.2292526513338089 - f1-score (micro avg) 0.7891
2023-10-25 14:32:31,069 ----------------------------------------------------------------------------------------------------
2023-10-25 14:32:31,071 Loading model from best epoch ...
2023-10-25 14:32:33,012 SequenceTagger predicts: Dictionary with 17 tags: O, S-PER, B-PER, E-PER, I-PER, S-LOC, B-LOC, E-LOC, I-LOC, S-ORG, B-ORG, E-ORG, I-ORG, S-HumanProd, B-HumanProd, E-HumanProd, I-HumanProd
2023-10-25 14:32:45,073
Results:
- F-score (micro) 0.6915
- F-score (macro) 0.6191
- Accuracy 0.5445
By class:
precision recall f1-score support
LOC 0.7139 0.6767 0.6948 1095
PER 0.7722 0.7737 0.7730 1012
ORG 0.4444 0.5602 0.4957 357
HumanProd 0.4444 0.6061 0.5128 33
micro avg 0.6847 0.6984 0.6915 2497
macro avg 0.5937 0.6542 0.6191 2497
weighted avg 0.6954 0.6984 0.6956 2497
2023-10-25 14:32:45,073 ----------------------------------------------------------------------------------------------------
|