File size: 25,113 Bytes
9c29c9b |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 |
2023-10-12 20:36:34,417 ----------------------------------------------------------------------------------------------------
2023-10-12 20:36:34,419 Model: "SequenceTagger(
(embeddings): ByT5Embeddings(
(model): T5EncoderModel(
(shared): Embedding(384, 1472)
(encoder): T5Stack(
(embed_tokens): Embedding(384, 1472)
(block): ModuleList(
(0): T5Block(
(layer): ModuleList(
(0): T5LayerSelfAttention(
(SelfAttention): T5Attention(
(q): Linear(in_features=1472, out_features=384, bias=False)
(k): Linear(in_features=1472, out_features=384, bias=False)
(v): Linear(in_features=1472, out_features=384, bias=False)
(o): Linear(in_features=384, out_features=1472, bias=False)
(relative_attention_bias): Embedding(32, 6)
)
(layer_norm): FusedRMSNorm(torch.Size([1472]), eps=1e-06, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(1): T5LayerFF(
(DenseReluDense): T5DenseGatedActDense(
(wi_0): Linear(in_features=1472, out_features=3584, bias=False)
(wi_1): Linear(in_features=1472, out_features=3584, bias=False)
(wo): Linear(in_features=3584, out_features=1472, bias=False)
(dropout): Dropout(p=0.1, inplace=False)
(act): NewGELUActivation()
)
(layer_norm): FusedRMSNorm(torch.Size([1472]), eps=1e-06, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
)
(1-11): 11 x T5Block(
(layer): ModuleList(
(0): T5LayerSelfAttention(
(SelfAttention): T5Attention(
(q): Linear(in_features=1472, out_features=384, bias=False)
(k): Linear(in_features=1472, out_features=384, bias=False)
(v): Linear(in_features=1472, out_features=384, bias=False)
(o): Linear(in_features=384, out_features=1472, bias=False)
)
(layer_norm): FusedRMSNorm(torch.Size([1472]), eps=1e-06, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(1): T5LayerFF(
(DenseReluDense): T5DenseGatedActDense(
(wi_0): Linear(in_features=1472, out_features=3584, bias=False)
(wi_1): Linear(in_features=1472, out_features=3584, bias=False)
(wo): Linear(in_features=3584, out_features=1472, bias=False)
(dropout): Dropout(p=0.1, inplace=False)
(act): NewGELUActivation()
)
(layer_norm): FusedRMSNorm(torch.Size([1472]), eps=1e-06, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
)
)
(final_layer_norm): FusedRMSNorm(torch.Size([1472]), eps=1e-06, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
)
(locked_dropout): LockedDropout(p=0.5)
(linear): Linear(in_features=1472, out_features=13, bias=True)
(loss_function): CrossEntropyLoss()
)"
2023-10-12 20:36:34,420 ----------------------------------------------------------------------------------------------------
2023-10-12 20:36:34,420 MultiCorpus: 7936 train + 992 dev + 992 test sentences
- NER_ICDAR_EUROPEANA Corpus: 7936 train + 992 dev + 992 test sentences - /root/.flair/datasets/ner_icdar_europeana/fr
2023-10-12 20:36:34,420 ----------------------------------------------------------------------------------------------------
2023-10-12 20:36:34,420 Train: 7936 sentences
2023-10-12 20:36:34,420 (train_with_dev=False, train_with_test=False)
2023-10-12 20:36:34,421 ----------------------------------------------------------------------------------------------------
2023-10-12 20:36:34,421 Training Params:
2023-10-12 20:36:34,421 - learning_rate: "0.00015"
2023-10-12 20:36:34,421 - mini_batch_size: "8"
2023-10-12 20:36:34,421 - max_epochs: "10"
2023-10-12 20:36:34,421 - shuffle: "True"
2023-10-12 20:36:34,421 ----------------------------------------------------------------------------------------------------
2023-10-12 20:36:34,421 Plugins:
2023-10-12 20:36:34,421 - TensorboardLogger
2023-10-12 20:36:34,421 - LinearScheduler | warmup_fraction: '0.1'
2023-10-12 20:36:34,422 ----------------------------------------------------------------------------------------------------
2023-10-12 20:36:34,422 Final evaluation on model from best epoch (best-model.pt)
2023-10-12 20:36:34,422 - metric: "('micro avg', 'f1-score')"
2023-10-12 20:36:34,422 ----------------------------------------------------------------------------------------------------
2023-10-12 20:36:34,422 Computation:
2023-10-12 20:36:34,422 - compute on device: cuda:0
2023-10-12 20:36:34,422 - embedding storage: none
2023-10-12 20:36:34,422 ----------------------------------------------------------------------------------------------------
2023-10-12 20:36:34,422 Model training base path: "hmbench-icdar/fr-hmbyt5-preliminary/byt5-small-historic-multilingual-span20-flax-bs8-wsFalse-e10-lr0.00015-poolingfirst-layers-1-crfFalse-3"
2023-10-12 20:36:34,422 ----------------------------------------------------------------------------------------------------
2023-10-12 20:36:34,423 ----------------------------------------------------------------------------------------------------
2023-10-12 20:36:34,423 Logging anything other than scalars to TensorBoard is currently not supported.
2023-10-12 20:37:23,526 epoch 1 - iter 99/992 - loss 2.53940455 - time (sec): 49.10 - samples/sec: 327.17 - lr: 0.000015 - momentum: 0.000000
2023-10-12 20:38:09,552 epoch 1 - iter 198/992 - loss 2.45209354 - time (sec): 95.13 - samples/sec: 334.01 - lr: 0.000030 - momentum: 0.000000
2023-10-12 20:38:56,579 epoch 1 - iter 297/992 - loss 2.23771454 - time (sec): 142.15 - samples/sec: 342.90 - lr: 0.000045 - momentum: 0.000000
2023-10-12 20:39:47,556 epoch 1 - iter 396/992 - loss 2.00240804 - time (sec): 193.13 - samples/sec: 337.89 - lr: 0.000060 - momentum: 0.000000
2023-10-12 20:40:37,028 epoch 1 - iter 495/992 - loss 1.75078774 - time (sec): 242.60 - samples/sec: 338.44 - lr: 0.000075 - momentum: 0.000000
2023-10-12 20:41:27,744 epoch 1 - iter 594/992 - loss 1.53136792 - time (sec): 293.32 - samples/sec: 335.30 - lr: 0.000090 - momentum: 0.000000
2023-10-12 20:42:16,877 epoch 1 - iter 693/992 - loss 1.36980903 - time (sec): 342.45 - samples/sec: 332.96 - lr: 0.000105 - momentum: 0.000000
2023-10-12 20:43:05,181 epoch 1 - iter 792/992 - loss 1.23246121 - time (sec): 390.76 - samples/sec: 334.01 - lr: 0.000120 - momentum: 0.000000
2023-10-12 20:43:58,462 epoch 1 - iter 891/992 - loss 1.11904567 - time (sec): 444.04 - samples/sec: 332.10 - lr: 0.000135 - momentum: 0.000000
2023-10-12 20:44:49,457 epoch 1 - iter 990/992 - loss 1.02887479 - time (sec): 495.03 - samples/sec: 330.68 - lr: 0.000150 - momentum: 0.000000
2023-10-12 20:44:50,413 ----------------------------------------------------------------------------------------------------
2023-10-12 20:44:50,413 EPOCH 1 done: loss 1.0276 - lr: 0.000150
2023-10-12 20:45:15,991 DEV : loss 0.16679786145687103 - f1-score (micro avg) 0.492
2023-10-12 20:45:16,031 saving best model
2023-10-12 20:45:16,981 ----------------------------------------------------------------------------------------------------
2023-10-12 20:46:06,496 epoch 2 - iter 99/992 - loss 0.17077119 - time (sec): 49.51 - samples/sec: 340.13 - lr: 0.000148 - momentum: 0.000000
2023-10-12 20:46:56,296 epoch 2 - iter 198/992 - loss 0.16840803 - time (sec): 99.31 - samples/sec: 332.86 - lr: 0.000147 - momentum: 0.000000
2023-10-12 20:47:44,810 epoch 2 - iter 297/992 - loss 0.16705195 - time (sec): 147.83 - samples/sec: 333.86 - lr: 0.000145 - momentum: 0.000000
2023-10-12 20:48:32,058 epoch 2 - iter 396/992 - loss 0.15897276 - time (sec): 195.08 - samples/sec: 339.12 - lr: 0.000143 - momentum: 0.000000
2023-10-12 20:49:21,939 epoch 2 - iter 495/992 - loss 0.15604786 - time (sec): 244.96 - samples/sec: 335.31 - lr: 0.000142 - momentum: 0.000000
2023-10-12 20:50:12,969 epoch 2 - iter 594/992 - loss 0.15226274 - time (sec): 295.99 - samples/sec: 334.72 - lr: 0.000140 - momentum: 0.000000
2023-10-12 20:51:01,723 epoch 2 - iter 693/992 - loss 0.14825625 - time (sec): 344.74 - samples/sec: 336.07 - lr: 0.000138 - momentum: 0.000000
2023-10-12 20:51:55,514 epoch 2 - iter 792/992 - loss 0.14565453 - time (sec): 398.53 - samples/sec: 330.04 - lr: 0.000137 - momentum: 0.000000
2023-10-12 20:52:44,252 epoch 2 - iter 891/992 - loss 0.14261297 - time (sec): 447.27 - samples/sec: 329.90 - lr: 0.000135 - momentum: 0.000000
2023-10-12 20:53:32,986 epoch 2 - iter 990/992 - loss 0.14067570 - time (sec): 496.00 - samples/sec: 330.13 - lr: 0.000133 - momentum: 0.000000
2023-10-12 20:53:33,941 ----------------------------------------------------------------------------------------------------
2023-10-12 20:53:33,942 EPOCH 2 done: loss 0.1406 - lr: 0.000133
2023-10-12 20:53:59,736 DEV : loss 0.09202314913272858 - f1-score (micro avg) 0.7419
2023-10-12 20:53:59,777 saving best model
2023-10-12 20:54:02,426 ----------------------------------------------------------------------------------------------------
2023-10-12 20:54:56,733 epoch 3 - iter 99/992 - loss 0.08481471 - time (sec): 54.29 - samples/sec: 316.06 - lr: 0.000132 - momentum: 0.000000
2023-10-12 20:55:48,391 epoch 3 - iter 198/992 - loss 0.08839525 - time (sec): 105.95 - samples/sec: 313.11 - lr: 0.000130 - momentum: 0.000000
2023-10-12 20:56:36,336 epoch 3 - iter 297/992 - loss 0.08581007 - time (sec): 153.89 - samples/sec: 317.80 - lr: 0.000128 - momentum: 0.000000
2023-10-12 20:57:25,973 epoch 3 - iter 396/992 - loss 0.08230993 - time (sec): 203.53 - samples/sec: 317.82 - lr: 0.000127 - momentum: 0.000000
2023-10-12 20:58:15,723 epoch 3 - iter 495/992 - loss 0.08191812 - time (sec): 253.28 - samples/sec: 321.05 - lr: 0.000125 - momentum: 0.000000
2023-10-12 20:59:05,929 epoch 3 - iter 594/992 - loss 0.08257402 - time (sec): 303.49 - samples/sec: 321.37 - lr: 0.000123 - momentum: 0.000000
2023-10-12 20:59:53,635 epoch 3 - iter 693/992 - loss 0.08166994 - time (sec): 351.19 - samples/sec: 322.98 - lr: 0.000122 - momentum: 0.000000
2023-10-12 21:00:42,974 epoch 3 - iter 792/992 - loss 0.08101549 - time (sec): 400.53 - samples/sec: 324.25 - lr: 0.000120 - momentum: 0.000000
2023-10-12 21:01:32,280 epoch 3 - iter 891/992 - loss 0.08026505 - time (sec): 449.84 - samples/sec: 325.56 - lr: 0.000118 - momentum: 0.000000
2023-10-12 21:02:23,827 epoch 3 - iter 990/992 - loss 0.08033861 - time (sec): 501.38 - samples/sec: 326.64 - lr: 0.000117 - momentum: 0.000000
2023-10-12 21:02:24,696 ----------------------------------------------------------------------------------------------------
2023-10-12 21:02:24,696 EPOCH 3 done: loss 0.0803 - lr: 0.000117
2023-10-12 21:02:49,665 DEV : loss 0.08990765362977982 - f1-score (micro avg) 0.7604
2023-10-12 21:02:49,705 saving best model
2023-10-12 21:02:52,306 ----------------------------------------------------------------------------------------------------
2023-10-12 21:03:41,885 epoch 4 - iter 99/992 - loss 0.05709566 - time (sec): 49.57 - samples/sec: 351.31 - lr: 0.000115 - momentum: 0.000000
2023-10-12 21:04:31,315 epoch 4 - iter 198/992 - loss 0.05384534 - time (sec): 99.00 - samples/sec: 335.44 - lr: 0.000113 - momentum: 0.000000
2023-10-12 21:05:20,247 epoch 4 - iter 297/992 - loss 0.05406144 - time (sec): 147.94 - samples/sec: 334.00 - lr: 0.000112 - momentum: 0.000000
2023-10-12 21:06:07,442 epoch 4 - iter 396/992 - loss 0.05302751 - time (sec): 195.13 - samples/sec: 336.83 - lr: 0.000110 - momentum: 0.000000
2023-10-12 21:06:53,930 epoch 4 - iter 495/992 - loss 0.05258435 - time (sec): 241.62 - samples/sec: 339.91 - lr: 0.000108 - momentum: 0.000000
2023-10-12 21:07:42,776 epoch 4 - iter 594/992 - loss 0.05291144 - time (sec): 290.47 - samples/sec: 337.89 - lr: 0.000107 - momentum: 0.000000
2023-10-12 21:08:31,390 epoch 4 - iter 693/992 - loss 0.05440967 - time (sec): 339.08 - samples/sec: 335.36 - lr: 0.000105 - momentum: 0.000000
2023-10-12 21:09:21,204 epoch 4 - iter 792/992 - loss 0.05467823 - time (sec): 388.89 - samples/sec: 335.88 - lr: 0.000103 - momentum: 0.000000
2023-10-12 21:10:13,890 epoch 4 - iter 891/992 - loss 0.05499260 - time (sec): 441.58 - samples/sec: 332.35 - lr: 0.000102 - momentum: 0.000000
2023-10-12 21:11:07,827 epoch 4 - iter 990/992 - loss 0.05480641 - time (sec): 495.52 - samples/sec: 330.33 - lr: 0.000100 - momentum: 0.000000
2023-10-12 21:11:08,918 ----------------------------------------------------------------------------------------------------
2023-10-12 21:11:08,919 EPOCH 4 done: loss 0.0550 - lr: 0.000100
2023-10-12 21:11:34,092 DEV : loss 0.11015438288450241 - f1-score (micro avg) 0.7516
2023-10-12 21:11:34,132 ----------------------------------------------------------------------------------------------------
2023-10-12 21:12:29,220 epoch 5 - iter 99/992 - loss 0.03490950 - time (sec): 55.09 - samples/sec: 281.83 - lr: 0.000098 - momentum: 0.000000
2023-10-12 21:13:19,557 epoch 5 - iter 198/992 - loss 0.03110268 - time (sec): 105.42 - samples/sec: 302.69 - lr: 0.000097 - momentum: 0.000000
2023-10-12 21:14:08,872 epoch 5 - iter 297/992 - loss 0.03567964 - time (sec): 154.74 - samples/sec: 313.09 - lr: 0.000095 - momentum: 0.000000
2023-10-12 21:15:00,906 epoch 5 - iter 396/992 - loss 0.03894388 - time (sec): 206.77 - samples/sec: 307.97 - lr: 0.000093 - momentum: 0.000000
2023-10-12 21:15:50,989 epoch 5 - iter 495/992 - loss 0.03871853 - time (sec): 256.85 - samples/sec: 308.70 - lr: 0.000092 - momentum: 0.000000
2023-10-12 21:16:42,127 epoch 5 - iter 594/992 - loss 0.03866059 - time (sec): 307.99 - samples/sec: 311.40 - lr: 0.000090 - momentum: 0.000000
2023-10-12 21:17:33,881 epoch 5 - iter 693/992 - loss 0.03769547 - time (sec): 359.75 - samples/sec: 316.19 - lr: 0.000088 - momentum: 0.000000
2023-10-12 21:18:22,487 epoch 5 - iter 792/992 - loss 0.03775340 - time (sec): 408.35 - samples/sec: 319.19 - lr: 0.000087 - momentum: 0.000000
2023-10-12 21:19:12,969 epoch 5 - iter 891/992 - loss 0.03942820 - time (sec): 458.83 - samples/sec: 319.35 - lr: 0.000085 - momentum: 0.000000
2023-10-12 21:20:05,611 epoch 5 - iter 990/992 - loss 0.03993968 - time (sec): 511.48 - samples/sec: 319.98 - lr: 0.000083 - momentum: 0.000000
2023-10-12 21:20:06,684 ----------------------------------------------------------------------------------------------------
2023-10-12 21:20:06,685 EPOCH 5 done: loss 0.0399 - lr: 0.000083
2023-10-12 21:20:34,573 DEV : loss 0.12319868057966232 - f1-score (micro avg) 0.7494
2023-10-12 21:20:34,627 ----------------------------------------------------------------------------------------------------
2023-10-12 21:21:25,537 epoch 6 - iter 99/992 - loss 0.03811252 - time (sec): 50.91 - samples/sec: 319.23 - lr: 0.000082 - momentum: 0.000000
2023-10-12 21:22:16,408 epoch 6 - iter 198/992 - loss 0.03228055 - time (sec): 101.78 - samples/sec: 316.51 - lr: 0.000080 - momentum: 0.000000
2023-10-12 21:23:07,761 epoch 6 - iter 297/992 - loss 0.03228149 - time (sec): 153.13 - samples/sec: 318.71 - lr: 0.000078 - momentum: 0.000000
2023-10-12 21:24:00,236 epoch 6 - iter 396/992 - loss 0.03210612 - time (sec): 205.61 - samples/sec: 318.41 - lr: 0.000077 - momentum: 0.000000
2023-10-12 21:24:53,252 epoch 6 - iter 495/992 - loss 0.03193389 - time (sec): 258.62 - samples/sec: 317.54 - lr: 0.000075 - momentum: 0.000000
2023-10-12 21:25:42,174 epoch 6 - iter 594/992 - loss 0.03222843 - time (sec): 307.54 - samples/sec: 318.60 - lr: 0.000073 - momentum: 0.000000
2023-10-12 21:26:32,031 epoch 6 - iter 693/992 - loss 0.03237733 - time (sec): 357.40 - samples/sec: 322.22 - lr: 0.000072 - momentum: 0.000000
2023-10-12 21:27:23,817 epoch 6 - iter 792/992 - loss 0.03084620 - time (sec): 409.19 - samples/sec: 321.91 - lr: 0.000070 - momentum: 0.000000
2023-10-12 21:28:16,696 epoch 6 - iter 891/992 - loss 0.03172155 - time (sec): 462.07 - samples/sec: 321.02 - lr: 0.000068 - momentum: 0.000000
2023-10-12 21:29:06,526 epoch 6 - iter 990/992 - loss 0.03169299 - time (sec): 511.90 - samples/sec: 319.93 - lr: 0.000067 - momentum: 0.000000
2023-10-12 21:29:07,472 ----------------------------------------------------------------------------------------------------
2023-10-12 21:29:07,473 EPOCH 6 done: loss 0.0317 - lr: 0.000067
2023-10-12 21:29:34,854 DEV : loss 0.1523526906967163 - f1-score (micro avg) 0.745
2023-10-12 21:29:34,898 ----------------------------------------------------------------------------------------------------
2023-10-12 21:30:29,230 epoch 7 - iter 99/992 - loss 0.01836233 - time (sec): 54.33 - samples/sec: 297.39 - lr: 0.000065 - momentum: 0.000000
2023-10-12 21:31:25,213 epoch 7 - iter 198/992 - loss 0.01888426 - time (sec): 110.31 - samples/sec: 298.53 - lr: 0.000063 - momentum: 0.000000
2023-10-12 21:32:16,337 epoch 7 - iter 297/992 - loss 0.02103087 - time (sec): 161.44 - samples/sec: 303.30 - lr: 0.000062 - momentum: 0.000000
2023-10-12 21:33:10,363 epoch 7 - iter 396/992 - loss 0.02296196 - time (sec): 215.46 - samples/sec: 303.71 - lr: 0.000060 - momentum: 0.000000
2023-10-12 21:34:04,597 epoch 7 - iter 495/992 - loss 0.02381151 - time (sec): 269.70 - samples/sec: 302.28 - lr: 0.000058 - momentum: 0.000000
2023-10-12 21:34:59,557 epoch 7 - iter 594/992 - loss 0.02454321 - time (sec): 324.66 - samples/sec: 302.98 - lr: 0.000057 - momentum: 0.000000
2023-10-12 21:35:53,242 epoch 7 - iter 693/992 - loss 0.02362879 - time (sec): 378.34 - samples/sec: 303.19 - lr: 0.000055 - momentum: 0.000000
2023-10-12 21:36:41,463 epoch 7 - iter 792/992 - loss 0.02358871 - time (sec): 426.56 - samples/sec: 307.73 - lr: 0.000053 - momentum: 0.000000
2023-10-12 21:37:29,049 epoch 7 - iter 891/992 - loss 0.02327723 - time (sec): 474.15 - samples/sec: 310.66 - lr: 0.000052 - momentum: 0.000000
2023-10-12 21:38:17,393 epoch 7 - iter 990/992 - loss 0.02414669 - time (sec): 522.49 - samples/sec: 313.30 - lr: 0.000050 - momentum: 0.000000
2023-10-12 21:38:18,316 ----------------------------------------------------------------------------------------------------
2023-10-12 21:38:18,316 EPOCH 7 done: loss 0.0241 - lr: 0.000050
2023-10-12 21:38:44,375 DEV : loss 0.16238392889499664 - f1-score (micro avg) 0.7575
2023-10-12 21:38:44,416 ----------------------------------------------------------------------------------------------------
2023-10-12 21:39:32,311 epoch 8 - iter 99/992 - loss 0.02098412 - time (sec): 47.89 - samples/sec: 345.39 - lr: 0.000048 - momentum: 0.000000
2023-10-12 21:40:21,125 epoch 8 - iter 198/992 - loss 0.02001400 - time (sec): 96.71 - samples/sec: 330.15 - lr: 0.000047 - momentum: 0.000000
2023-10-12 21:41:11,175 epoch 8 - iter 297/992 - loss 0.01778574 - time (sec): 146.76 - samples/sec: 332.28 - lr: 0.000045 - momentum: 0.000000
2023-10-12 21:42:00,933 epoch 8 - iter 396/992 - loss 0.01851136 - time (sec): 196.51 - samples/sec: 333.89 - lr: 0.000043 - momentum: 0.000000
2023-10-12 21:42:47,636 epoch 8 - iter 495/992 - loss 0.01977844 - time (sec): 243.22 - samples/sec: 337.22 - lr: 0.000042 - momentum: 0.000000
2023-10-12 21:43:35,853 epoch 8 - iter 594/992 - loss 0.01974787 - time (sec): 291.43 - samples/sec: 336.66 - lr: 0.000040 - momentum: 0.000000
2023-10-12 21:44:21,954 epoch 8 - iter 693/992 - loss 0.01908591 - time (sec): 337.54 - samples/sec: 338.15 - lr: 0.000038 - momentum: 0.000000
2023-10-12 21:45:09,680 epoch 8 - iter 792/992 - loss 0.01817565 - time (sec): 385.26 - samples/sec: 340.05 - lr: 0.000037 - momentum: 0.000000
2023-10-12 21:45:56,435 epoch 8 - iter 891/992 - loss 0.01823549 - time (sec): 432.02 - samples/sec: 340.11 - lr: 0.000035 - momentum: 0.000000
2023-10-12 21:46:44,381 epoch 8 - iter 990/992 - loss 0.01881597 - time (sec): 479.96 - samples/sec: 340.91 - lr: 0.000033 - momentum: 0.000000
2023-10-12 21:46:45,354 ----------------------------------------------------------------------------------------------------
2023-10-12 21:46:45,354 EPOCH 8 done: loss 0.0188 - lr: 0.000033
2023-10-12 21:47:10,756 DEV : loss 0.1798100620508194 - f1-score (micro avg) 0.7504
2023-10-12 21:47:10,796 ----------------------------------------------------------------------------------------------------
2023-10-12 21:47:57,915 epoch 9 - iter 99/992 - loss 0.01330830 - time (sec): 47.12 - samples/sec: 328.12 - lr: 0.000032 - momentum: 0.000000
2023-10-12 21:48:46,407 epoch 9 - iter 198/992 - loss 0.01224787 - time (sec): 95.61 - samples/sec: 322.93 - lr: 0.000030 - momentum: 0.000000
2023-10-12 21:49:35,505 epoch 9 - iter 297/992 - loss 0.01382402 - time (sec): 144.71 - samples/sec: 326.64 - lr: 0.000028 - momentum: 0.000000
2023-10-12 21:50:25,041 epoch 9 - iter 396/992 - loss 0.01437517 - time (sec): 194.24 - samples/sec: 331.02 - lr: 0.000027 - momentum: 0.000000
2023-10-12 21:51:12,360 epoch 9 - iter 495/992 - loss 0.01368674 - time (sec): 241.56 - samples/sec: 334.81 - lr: 0.000025 - momentum: 0.000000
2023-10-12 21:51:59,940 epoch 9 - iter 594/992 - loss 0.01453578 - time (sec): 289.14 - samples/sec: 340.98 - lr: 0.000023 - momentum: 0.000000
2023-10-12 21:52:46,742 epoch 9 - iter 693/992 - loss 0.01553987 - time (sec): 335.94 - samples/sec: 343.65 - lr: 0.000022 - momentum: 0.000000
2023-10-12 21:53:34,899 epoch 9 - iter 792/992 - loss 0.01569527 - time (sec): 384.10 - samples/sec: 343.96 - lr: 0.000020 - momentum: 0.000000
2023-10-12 21:54:22,482 epoch 9 - iter 891/992 - loss 0.01579337 - time (sec): 431.68 - samples/sec: 344.20 - lr: 0.000018 - momentum: 0.000000
2023-10-12 21:55:10,034 epoch 9 - iter 990/992 - loss 0.01620441 - time (sec): 479.24 - samples/sec: 341.45 - lr: 0.000017 - momentum: 0.000000
2023-10-12 21:55:10,982 ----------------------------------------------------------------------------------------------------
2023-10-12 21:55:10,982 EPOCH 9 done: loss 0.0162 - lr: 0.000017
2023-10-12 21:55:37,009 DEV : loss 0.18687152862548828 - f1-score (micro avg) 0.7575
2023-10-12 21:55:37,055 ----------------------------------------------------------------------------------------------------
2023-10-12 21:56:24,933 epoch 10 - iter 99/992 - loss 0.00770149 - time (sec): 47.88 - samples/sec: 344.62 - lr: 0.000015 - momentum: 0.000000
2023-10-12 21:57:12,835 epoch 10 - iter 198/992 - loss 0.00989841 - time (sec): 95.78 - samples/sec: 346.95 - lr: 0.000013 - momentum: 0.000000
2023-10-12 21:58:00,731 epoch 10 - iter 297/992 - loss 0.01166225 - time (sec): 143.67 - samples/sec: 348.74 - lr: 0.000012 - momentum: 0.000000
2023-10-12 21:58:46,948 epoch 10 - iter 396/992 - loss 0.01166889 - time (sec): 189.89 - samples/sec: 348.12 - lr: 0.000010 - momentum: 0.000000
2023-10-12 21:59:35,210 epoch 10 - iter 495/992 - loss 0.01103474 - time (sec): 238.15 - samples/sec: 346.06 - lr: 0.000008 - momentum: 0.000000
2023-10-12 22:00:23,313 epoch 10 - iter 594/992 - loss 0.01167473 - time (sec): 286.26 - samples/sec: 343.43 - lr: 0.000007 - momentum: 0.000000
2023-10-12 22:01:12,909 epoch 10 - iter 693/992 - loss 0.01237909 - time (sec): 335.85 - samples/sec: 342.47 - lr: 0.000005 - momentum: 0.000000
2023-10-12 22:01:59,272 epoch 10 - iter 792/992 - loss 0.01251453 - time (sec): 382.21 - samples/sec: 342.30 - lr: 0.000004 - momentum: 0.000000
2023-10-12 22:02:46,892 epoch 10 - iter 891/992 - loss 0.01256245 - time (sec): 429.83 - samples/sec: 343.86 - lr: 0.000002 - momentum: 0.000000
2023-10-12 22:03:34,464 epoch 10 - iter 990/992 - loss 0.01261074 - time (sec): 477.41 - samples/sec: 342.85 - lr: 0.000000 - momentum: 0.000000
2023-10-12 22:03:35,396 ----------------------------------------------------------------------------------------------------
2023-10-12 22:03:35,397 EPOCH 10 done: loss 0.0126 - lr: 0.000000
2023-10-12 22:04:01,331 DEV : loss 0.19634360074996948 - f1-score (micro avg) 0.7592
2023-10-12 22:04:02,299 ----------------------------------------------------------------------------------------------------
2023-10-12 22:04:02,301 Loading model from best epoch ...
2023-10-12 22:04:05,962 SequenceTagger predicts: Dictionary with 13 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
2023-10-12 22:04:30,881
Results:
- F-score (micro) 0.7525
- F-score (macro) 0.6889
- Accuracy 0.6284
By class:
precision recall f1-score support
LOC 0.7875 0.8092 0.7982 655
PER 0.7255 0.8296 0.7741 223
ORG 0.4595 0.5354 0.4945 127
micro avg 0.7277 0.7791 0.7525 1005
macro avg 0.6575 0.7247 0.6889 1005
weighted avg 0.7323 0.7791 0.7545 1005
2023-10-12 22:04:30,881 ----------------------------------------------------------------------------------------------------
|