File size: 24,309 Bytes
3072fde |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 |
2023-11-16 08:54:28,192 ----------------------------------------------------------------------------------------------------
2023-11-16 08:54:28,194 Model: "SequenceTagger(
(embeddings): TransformerWordEmbeddings(
(model): XLMRobertaModel(
(embeddings): XLMRobertaEmbeddings(
(word_embeddings): Embedding(250003, 1024)
(position_embeddings): Embedding(514, 1024, padding_idx=1)
(token_type_embeddings): Embedding(1, 1024)
(LayerNorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(encoder): XLMRobertaEncoder(
(layer): ModuleList(
(0-23): 24 x XLMRobertaLayer(
(attention): XLMRobertaAttention(
(self): XLMRobertaSelfAttention(
(query): Linear(in_features=1024, out_features=1024, bias=True)
(key): Linear(in_features=1024, out_features=1024, bias=True)
(value): Linear(in_features=1024, out_features=1024, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(output): XLMRobertaSelfOutput(
(dense): Linear(in_features=1024, out_features=1024, bias=True)
(LayerNorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(intermediate): XLMRobertaIntermediate(
(dense): Linear(in_features=1024, out_features=4096, bias=True)
(intermediate_act_fn): GELUActivation()
)
(output): XLMRobertaOutput(
(dense): Linear(in_features=4096, out_features=1024, bias=True)
(LayerNorm): LayerNorm((1024,), eps=1e-05, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
)
)
(pooler): XLMRobertaPooler(
(dense): Linear(in_features=1024, out_features=1024, bias=True)
(activation): Tanh()
)
)
)
(locked_dropout): LockedDropout(p=0.5)
(linear): Linear(in_features=1024, out_features=13, bias=True)
(loss_function): CrossEntropyLoss()
)"
2023-11-16 08:54:28,194 ----------------------------------------------------------------------------------------------------
2023-11-16 08:54:28,194 MultiCorpus: 30000 train + 10000 dev + 10000 test sentences
- ColumnCorpus Corpus: 20000 train + 0 dev + 0 test sentences - /root/.flair/datasets/ner_multi_xtreme/en
- ColumnCorpus Corpus: 10000 train + 10000 dev + 10000 test sentences - /root/.flair/datasets/ner_multi_xtreme/ka
2023-11-16 08:54:28,194 ----------------------------------------------------------------------------------------------------
2023-11-16 08:54:28,194 Train: 30000 sentences
2023-11-16 08:54:28,194 (train_with_dev=False, train_with_test=False)
2023-11-16 08:54:28,194 ----------------------------------------------------------------------------------------------------
2023-11-16 08:54:28,194 Training Params:
2023-11-16 08:54:28,194 - learning_rate: "5e-06"
2023-11-16 08:54:28,194 - mini_batch_size: "4"
2023-11-16 08:54:28,194 - max_epochs: "10"
2023-11-16 08:54:28,194 - shuffle: "True"
2023-11-16 08:54:28,194 ----------------------------------------------------------------------------------------------------
2023-11-16 08:54:28,194 Plugins:
2023-11-16 08:54:28,194 - TensorboardLogger
2023-11-16 08:54:28,194 - LinearScheduler | warmup_fraction: '0.1'
2023-11-16 08:54:28,194 ----------------------------------------------------------------------------------------------------
2023-11-16 08:54:28,194 Final evaluation on model from best epoch (best-model.pt)
2023-11-16 08:54:28,194 - metric: "('micro avg', 'f1-score')"
2023-11-16 08:54:28,194 ----------------------------------------------------------------------------------------------------
2023-11-16 08:54:28,195 Computation:
2023-11-16 08:54:28,195 - compute on device: cuda:0
2023-11-16 08:54:28,195 - embedding storage: none
2023-11-16 08:54:28,195 ----------------------------------------------------------------------------------------------------
2023-11-16 08:54:28,195 Model training base path: "autotrain-flair-georgian-ner-xlm_r_large-bs4-e10-lr5e-06-5"
2023-11-16 08:54:28,195 ----------------------------------------------------------------------------------------------------
2023-11-16 08:54:28,195 ----------------------------------------------------------------------------------------------------
2023-11-16 08:54:28,195 Logging anything other than scalars to TensorBoard is currently not supported.
2023-11-16 08:56:03,042 epoch 1 - iter 750/7500 - loss 2.73514177 - time (sec): 94.85 - samples/sec: 251.28 - lr: 0.000000 - momentum: 0.000000
2023-11-16 08:57:34,557 epoch 1 - iter 1500/7500 - loss 2.27024326 - time (sec): 186.36 - samples/sec: 258.14 - lr: 0.000001 - momentum: 0.000000
2023-11-16 08:59:07,587 epoch 1 - iter 2250/7500 - loss 1.96361468 - time (sec): 279.39 - samples/sec: 259.21 - lr: 0.000001 - momentum: 0.000000
2023-11-16 09:00:41,392 epoch 1 - iter 3000/7500 - loss 1.71842298 - time (sec): 373.20 - samples/sec: 259.19 - lr: 0.000002 - momentum: 0.000000
2023-11-16 09:02:13,057 epoch 1 - iter 3750/7500 - loss 1.50867437 - time (sec): 464.86 - samples/sec: 260.24 - lr: 0.000002 - momentum: 0.000000
2023-11-16 09:03:44,283 epoch 1 - iter 4500/7500 - loss 1.34975880 - time (sec): 556.09 - samples/sec: 261.24 - lr: 0.000003 - momentum: 0.000000
2023-11-16 09:05:17,789 epoch 1 - iter 5250/7500 - loss 1.23153815 - time (sec): 649.59 - samples/sec: 261.16 - lr: 0.000003 - momentum: 0.000000
2023-11-16 09:06:51,967 epoch 1 - iter 6000/7500 - loss 1.13901611 - time (sec): 743.77 - samples/sec: 260.11 - lr: 0.000004 - momentum: 0.000000
2023-11-16 09:08:25,989 epoch 1 - iter 6750/7500 - loss 1.06518907 - time (sec): 837.79 - samples/sec: 259.14 - lr: 0.000004 - momentum: 0.000000
2023-11-16 09:10:02,189 epoch 1 - iter 7500/7500 - loss 1.00335008 - time (sec): 933.99 - samples/sec: 257.81 - lr: 0.000005 - momentum: 0.000000
2023-11-16 09:10:02,191 ----------------------------------------------------------------------------------------------------
2023-11-16 09:10:02,191 EPOCH 1 done: loss 1.0034 - lr: 0.000005
2023-11-16 09:10:29,151 DEV : loss 0.25209933519363403 - f1-score (micro avg) 0.818
2023-11-16 09:10:30,799 saving best model
2023-11-16 09:10:32,550 ----------------------------------------------------------------------------------------------------
2023-11-16 09:12:05,182 epoch 2 - iter 750/7500 - loss 0.40758191 - time (sec): 92.63 - samples/sec: 255.89 - lr: 0.000005 - momentum: 0.000000
2023-11-16 09:13:40,646 epoch 2 - iter 1500/7500 - loss 0.41354608 - time (sec): 188.09 - samples/sec: 252.51 - lr: 0.000005 - momentum: 0.000000
2023-11-16 09:15:13,127 epoch 2 - iter 2250/7500 - loss 0.41469889 - time (sec): 280.57 - samples/sec: 256.53 - lr: 0.000005 - momentum: 0.000000
2023-11-16 09:16:46,617 epoch 2 - iter 3000/7500 - loss 0.40864404 - time (sec): 374.06 - samples/sec: 257.65 - lr: 0.000005 - momentum: 0.000000
2023-11-16 09:18:19,301 epoch 2 - iter 3750/7500 - loss 0.40728475 - time (sec): 466.75 - samples/sec: 258.79 - lr: 0.000005 - momentum: 0.000000
2023-11-16 09:19:50,423 epoch 2 - iter 4500/7500 - loss 0.40431483 - time (sec): 557.87 - samples/sec: 259.91 - lr: 0.000005 - momentum: 0.000000
2023-11-16 09:21:22,320 epoch 2 - iter 5250/7500 - loss 0.40010819 - time (sec): 649.77 - samples/sec: 260.46 - lr: 0.000005 - momentum: 0.000000
2023-11-16 09:22:53,693 epoch 2 - iter 6000/7500 - loss 0.39994412 - time (sec): 741.14 - samples/sec: 260.74 - lr: 0.000005 - momentum: 0.000000
2023-11-16 09:24:27,071 epoch 2 - iter 6750/7500 - loss 0.40164576 - time (sec): 834.52 - samples/sec: 260.23 - lr: 0.000005 - momentum: 0.000000
2023-11-16 09:26:00,518 epoch 2 - iter 7500/7500 - loss 0.40139034 - time (sec): 927.97 - samples/sec: 259.49 - lr: 0.000004 - momentum: 0.000000
2023-11-16 09:26:00,523 ----------------------------------------------------------------------------------------------------
2023-11-16 09:26:00,523 EPOCH 2 done: loss 0.4014 - lr: 0.000004
2023-11-16 09:26:28,590 DEV : loss 0.27011075615882874 - f1-score (micro avg) 0.8685
2023-11-16 09:26:30,477 saving best model
2023-11-16 09:26:32,489 ----------------------------------------------------------------------------------------------------
2023-11-16 09:28:06,865 epoch 3 - iter 750/7500 - loss 0.34664813 - time (sec): 94.37 - samples/sec: 259.00 - lr: 0.000004 - momentum: 0.000000
2023-11-16 09:29:40,525 epoch 3 - iter 1500/7500 - loss 0.35802918 - time (sec): 188.03 - samples/sec: 256.41 - lr: 0.000004 - momentum: 0.000000
2023-11-16 09:31:13,248 epoch 3 - iter 2250/7500 - loss 0.34949160 - time (sec): 280.76 - samples/sec: 259.11 - lr: 0.000004 - momentum: 0.000000
2023-11-16 09:32:44,431 epoch 3 - iter 3000/7500 - loss 0.34400653 - time (sec): 371.94 - samples/sec: 261.74 - lr: 0.000004 - momentum: 0.000000
2023-11-16 09:34:18,607 epoch 3 - iter 3750/7500 - loss 0.34736997 - time (sec): 466.12 - samples/sec: 259.21 - lr: 0.000004 - momentum: 0.000000
2023-11-16 09:35:51,979 epoch 3 - iter 4500/7500 - loss 0.34780959 - time (sec): 559.49 - samples/sec: 258.64 - lr: 0.000004 - momentum: 0.000000
2023-11-16 09:37:25,102 epoch 3 - iter 5250/7500 - loss 0.34688026 - time (sec): 652.61 - samples/sec: 258.12 - lr: 0.000004 - momentum: 0.000000
2023-11-16 09:38:58,439 epoch 3 - iter 6000/7500 - loss 0.34475842 - time (sec): 745.95 - samples/sec: 257.81 - lr: 0.000004 - momentum: 0.000000
2023-11-16 09:40:34,113 epoch 3 - iter 6750/7500 - loss 0.34349763 - time (sec): 841.62 - samples/sec: 256.89 - lr: 0.000004 - momentum: 0.000000
2023-11-16 09:42:10,676 epoch 3 - iter 7500/7500 - loss 0.34117216 - time (sec): 938.18 - samples/sec: 256.66 - lr: 0.000004 - momentum: 0.000000
2023-11-16 09:42:10,679 ----------------------------------------------------------------------------------------------------
2023-11-16 09:42:10,679 EPOCH 3 done: loss 0.3412 - lr: 0.000004
2023-11-16 09:42:37,676 DEV : loss 0.2926769554615021 - f1-score (micro avg) 0.8843
2023-11-16 09:42:39,581 saving best model
2023-11-16 09:42:41,582 ----------------------------------------------------------------------------------------------------
2023-11-16 09:44:14,986 epoch 4 - iter 750/7500 - loss 0.28935096 - time (sec): 93.40 - samples/sec: 255.70 - lr: 0.000004 - momentum: 0.000000
2023-11-16 09:45:47,989 epoch 4 - iter 1500/7500 - loss 0.29093752 - time (sec): 186.40 - samples/sec: 257.90 - lr: 0.000004 - momentum: 0.000000
2023-11-16 09:47:19,359 epoch 4 - iter 2250/7500 - loss 0.29888625 - time (sec): 277.77 - samples/sec: 259.38 - lr: 0.000004 - momentum: 0.000000
2023-11-16 09:48:51,599 epoch 4 - iter 3000/7500 - loss 0.30114670 - time (sec): 370.01 - samples/sec: 258.77 - lr: 0.000004 - momentum: 0.000000
2023-11-16 09:50:23,345 epoch 4 - iter 3750/7500 - loss 0.30007515 - time (sec): 461.76 - samples/sec: 259.19 - lr: 0.000004 - momentum: 0.000000
2023-11-16 09:51:56,749 epoch 4 - iter 4500/7500 - loss 0.29780444 - time (sec): 555.16 - samples/sec: 259.74 - lr: 0.000004 - momentum: 0.000000
2023-11-16 09:53:29,783 epoch 4 - iter 5250/7500 - loss 0.29632423 - time (sec): 648.20 - samples/sec: 260.29 - lr: 0.000004 - momentum: 0.000000
2023-11-16 09:55:03,332 epoch 4 - iter 6000/7500 - loss 0.29715629 - time (sec): 741.75 - samples/sec: 259.64 - lr: 0.000003 - momentum: 0.000000
2023-11-16 09:56:39,271 epoch 4 - iter 6750/7500 - loss 0.29789543 - time (sec): 837.69 - samples/sec: 258.76 - lr: 0.000003 - momentum: 0.000000
2023-11-16 09:58:14,979 epoch 4 - iter 7500/7500 - loss 0.30080354 - time (sec): 933.39 - samples/sec: 257.98 - lr: 0.000003 - momentum: 0.000000
2023-11-16 09:58:14,982 ----------------------------------------------------------------------------------------------------
2023-11-16 09:58:14,982 EPOCH 4 done: loss 0.3008 - lr: 0.000003
2023-11-16 09:58:42,391 DEV : loss 0.24168777465820312 - f1-score (micro avg) 0.8958
2023-11-16 09:58:44,896 saving best model
2023-11-16 09:58:47,890 ----------------------------------------------------------------------------------------------------
2023-11-16 10:00:21,193 epoch 5 - iter 750/7500 - loss 0.24325762 - time (sec): 93.30 - samples/sec: 254.74 - lr: 0.000003 - momentum: 0.000000
2023-11-16 10:01:54,947 epoch 5 - iter 1500/7500 - loss 0.24699916 - time (sec): 187.05 - samples/sec: 256.42 - lr: 0.000003 - momentum: 0.000000
2023-11-16 10:03:33,373 epoch 5 - iter 2250/7500 - loss 0.24105182 - time (sec): 285.48 - samples/sec: 251.76 - lr: 0.000003 - momentum: 0.000000
2023-11-16 10:05:10,591 epoch 5 - iter 3000/7500 - loss 0.24548635 - time (sec): 382.70 - samples/sec: 250.75 - lr: 0.000003 - momentum: 0.000000
2023-11-16 10:06:45,160 epoch 5 - iter 3750/7500 - loss 0.24697996 - time (sec): 477.27 - samples/sec: 252.00 - lr: 0.000003 - momentum: 0.000000
2023-11-16 10:08:18,840 epoch 5 - iter 4500/7500 - loss 0.24902921 - time (sec): 570.95 - samples/sec: 252.17 - lr: 0.000003 - momentum: 0.000000
2023-11-16 10:09:54,495 epoch 5 - iter 5250/7500 - loss 0.24900570 - time (sec): 666.60 - samples/sec: 251.98 - lr: 0.000003 - momentum: 0.000000
2023-11-16 10:11:28,132 epoch 5 - iter 6000/7500 - loss 0.25246330 - time (sec): 760.24 - samples/sec: 253.26 - lr: 0.000003 - momentum: 0.000000
2023-11-16 10:13:02,532 epoch 5 - iter 6750/7500 - loss 0.25090384 - time (sec): 854.64 - samples/sec: 253.77 - lr: 0.000003 - momentum: 0.000000
2023-11-16 10:14:33,980 epoch 5 - iter 7500/7500 - loss 0.25096782 - time (sec): 946.09 - samples/sec: 254.52 - lr: 0.000003 - momentum: 0.000000
2023-11-16 10:14:33,983 ----------------------------------------------------------------------------------------------------
2023-11-16 10:14:33,983 EPOCH 5 done: loss 0.2510 - lr: 0.000003
2023-11-16 10:15:01,770 DEV : loss 0.30133897066116333 - f1-score (micro avg) 0.8909
2023-11-16 10:15:04,505 ----------------------------------------------------------------------------------------------------
2023-11-16 10:16:43,786 epoch 6 - iter 750/7500 - loss 0.21043668 - time (sec): 99.28 - samples/sec: 246.62 - lr: 0.000003 - momentum: 0.000000
2023-11-16 10:18:21,438 epoch 6 - iter 1500/7500 - loss 0.21551137 - time (sec): 196.93 - samples/sec: 244.08 - lr: 0.000003 - momentum: 0.000000
2023-11-16 10:19:58,967 epoch 6 - iter 2250/7500 - loss 0.21671764 - time (sec): 294.46 - samples/sec: 244.90 - lr: 0.000003 - momentum: 0.000000
2023-11-16 10:21:36,357 epoch 6 - iter 3000/7500 - loss 0.21410789 - time (sec): 391.85 - samples/sec: 245.03 - lr: 0.000003 - momentum: 0.000000
2023-11-16 10:23:11,820 epoch 6 - iter 3750/7500 - loss 0.21190957 - time (sec): 487.31 - samples/sec: 246.63 - lr: 0.000003 - momentum: 0.000000
2023-11-16 10:24:44,519 epoch 6 - iter 4500/7500 - loss 0.21724671 - time (sec): 580.01 - samples/sec: 248.11 - lr: 0.000002 - momentum: 0.000000
2023-11-16 10:26:19,240 epoch 6 - iter 5250/7500 - loss 0.21517436 - time (sec): 674.73 - samples/sec: 249.52 - lr: 0.000002 - momentum: 0.000000
2023-11-16 10:27:52,483 epoch 6 - iter 6000/7500 - loss 0.21502133 - time (sec): 767.97 - samples/sec: 250.59 - lr: 0.000002 - momentum: 0.000000
2023-11-16 10:29:26,618 epoch 6 - iter 6750/7500 - loss 0.21284089 - time (sec): 862.11 - samples/sec: 251.34 - lr: 0.000002 - momentum: 0.000000
2023-11-16 10:31:02,401 epoch 6 - iter 7500/7500 - loss 0.21376183 - time (sec): 957.89 - samples/sec: 251.38 - lr: 0.000002 - momentum: 0.000000
2023-11-16 10:31:02,403 ----------------------------------------------------------------------------------------------------
2023-11-16 10:31:02,404 EPOCH 6 done: loss 0.2138 - lr: 0.000002
2023-11-16 10:31:29,893 DEV : loss 0.2858603894710541 - f1-score (micro avg) 0.9013
2023-11-16 10:31:32,354 saving best model
2023-11-16 10:31:34,717 ----------------------------------------------------------------------------------------------------
2023-11-16 10:33:10,947 epoch 7 - iter 750/7500 - loss 0.18560997 - time (sec): 96.22 - samples/sec: 252.05 - lr: 0.000002 - momentum: 0.000000
2023-11-16 10:34:45,274 epoch 7 - iter 1500/7500 - loss 0.18388762 - time (sec): 190.55 - samples/sec: 254.81 - lr: 0.000002 - momentum: 0.000000
2023-11-16 10:36:20,608 epoch 7 - iter 2250/7500 - loss 0.17300910 - time (sec): 285.89 - samples/sec: 254.67 - lr: 0.000002 - momentum: 0.000000
2023-11-16 10:37:56,122 epoch 7 - iter 3000/7500 - loss 0.18185895 - time (sec): 381.40 - samples/sec: 253.44 - lr: 0.000002 - momentum: 0.000000
2023-11-16 10:39:29,201 epoch 7 - iter 3750/7500 - loss 0.18240739 - time (sec): 474.48 - samples/sec: 253.00 - lr: 0.000002 - momentum: 0.000000
2023-11-16 10:41:03,810 epoch 7 - iter 4500/7500 - loss 0.18167213 - time (sec): 569.09 - samples/sec: 253.57 - lr: 0.000002 - momentum: 0.000000
2023-11-16 10:42:34,871 epoch 7 - iter 5250/7500 - loss 0.18305956 - time (sec): 660.15 - samples/sec: 255.17 - lr: 0.000002 - momentum: 0.000000
2023-11-16 10:44:06,443 epoch 7 - iter 6000/7500 - loss 0.18397991 - time (sec): 751.72 - samples/sec: 256.24 - lr: 0.000002 - momentum: 0.000000
2023-11-16 10:45:40,244 epoch 7 - iter 6750/7500 - loss 0.18284928 - time (sec): 845.52 - samples/sec: 256.33 - lr: 0.000002 - momentum: 0.000000
2023-11-16 10:47:13,133 epoch 7 - iter 7500/7500 - loss 0.18356346 - time (sec): 938.41 - samples/sec: 256.60 - lr: 0.000002 - momentum: 0.000000
2023-11-16 10:47:13,135 ----------------------------------------------------------------------------------------------------
2023-11-16 10:47:13,136 EPOCH 7 done: loss 0.1836 - lr: 0.000002
2023-11-16 10:47:40,667 DEV : loss 0.3011305034160614 - f1-score (micro avg) 0.8987
2023-11-16 10:47:42,779 ----------------------------------------------------------------------------------------------------
2023-11-16 10:49:16,539 epoch 8 - iter 750/7500 - loss 0.14330999 - time (sec): 93.76 - samples/sec: 256.81 - lr: 0.000002 - momentum: 0.000000
2023-11-16 10:50:50,862 epoch 8 - iter 1500/7500 - loss 0.14160047 - time (sec): 188.08 - samples/sec: 252.69 - lr: 0.000002 - momentum: 0.000000
2023-11-16 10:52:23,414 epoch 8 - iter 2250/7500 - loss 0.14478770 - time (sec): 280.63 - samples/sec: 256.60 - lr: 0.000002 - momentum: 0.000000
2023-11-16 10:53:56,497 epoch 8 - iter 3000/7500 - loss 0.14930840 - time (sec): 373.71 - samples/sec: 258.22 - lr: 0.000001 - momentum: 0.000000
2023-11-16 10:55:28,192 epoch 8 - iter 3750/7500 - loss 0.15175926 - time (sec): 465.41 - samples/sec: 258.55 - lr: 0.000001 - momentum: 0.000000
2023-11-16 10:57:00,742 epoch 8 - iter 4500/7500 - loss 0.15482872 - time (sec): 557.96 - samples/sec: 259.33 - lr: 0.000001 - momentum: 0.000000
2023-11-16 10:58:32,426 epoch 8 - iter 5250/7500 - loss 0.15110368 - time (sec): 649.64 - samples/sec: 259.55 - lr: 0.000001 - momentum: 0.000000
2023-11-16 11:00:02,971 epoch 8 - iter 6000/7500 - loss 0.15064249 - time (sec): 740.19 - samples/sec: 260.02 - lr: 0.000001 - momentum: 0.000000
2023-11-16 11:01:35,563 epoch 8 - iter 6750/7500 - loss 0.15119609 - time (sec): 832.78 - samples/sec: 260.42 - lr: 0.000001 - momentum: 0.000000
2023-11-16 11:03:09,780 epoch 8 - iter 7500/7500 - loss 0.15276122 - time (sec): 927.00 - samples/sec: 259.76 - lr: 0.000001 - momentum: 0.000000
2023-11-16 11:03:09,783 ----------------------------------------------------------------------------------------------------
2023-11-16 11:03:09,783 EPOCH 8 done: loss 0.1528 - lr: 0.000001
2023-11-16 11:03:37,801 DEV : loss 0.31595587730407715 - f1-score (micro avg) 0.9048
2023-11-16 11:03:40,424 saving best model
2023-11-16 11:03:43,075 ----------------------------------------------------------------------------------------------------
2023-11-16 11:05:19,854 epoch 9 - iter 750/7500 - loss 0.12299891 - time (sec): 96.77 - samples/sec: 253.49 - lr: 0.000001 - momentum: 0.000000
2023-11-16 11:06:56,996 epoch 9 - iter 1500/7500 - loss 0.12314833 - time (sec): 193.92 - samples/sec: 248.21 - lr: 0.000001 - momentum: 0.000000
2023-11-16 11:08:31,883 epoch 9 - iter 2250/7500 - loss 0.13008764 - time (sec): 288.80 - samples/sec: 250.26 - lr: 0.000001 - momentum: 0.000000
2023-11-16 11:10:08,385 epoch 9 - iter 3000/7500 - loss 0.13080561 - time (sec): 385.30 - samples/sec: 250.44 - lr: 0.000001 - momentum: 0.000000
2023-11-16 11:11:42,064 epoch 9 - iter 3750/7500 - loss 0.13273049 - time (sec): 478.98 - samples/sec: 251.53 - lr: 0.000001 - momentum: 0.000000
2023-11-16 11:13:16,037 epoch 9 - iter 4500/7500 - loss 0.13360348 - time (sec): 572.96 - samples/sec: 251.77 - lr: 0.000001 - momentum: 0.000000
2023-11-16 11:14:49,720 epoch 9 - iter 5250/7500 - loss 0.13299160 - time (sec): 666.64 - samples/sec: 252.02 - lr: 0.000001 - momentum: 0.000000
2023-11-16 11:16:23,602 epoch 9 - iter 6000/7500 - loss 0.13133134 - time (sec): 760.52 - samples/sec: 252.84 - lr: 0.000001 - momentum: 0.000000
2023-11-16 11:17:56,438 epoch 9 - iter 6750/7500 - loss 0.13363917 - time (sec): 853.36 - samples/sec: 253.29 - lr: 0.000001 - momentum: 0.000000
2023-11-16 11:19:29,919 epoch 9 - iter 7500/7500 - loss 0.13122225 - time (sec): 946.84 - samples/sec: 254.32 - lr: 0.000001 - momentum: 0.000000
2023-11-16 11:19:29,921 ----------------------------------------------------------------------------------------------------
2023-11-16 11:19:29,921 EPOCH 9 done: loss 0.1312 - lr: 0.000001
2023-11-16 11:19:57,777 DEV : loss 0.33471381664276123 - f1-score (micro avg) 0.9024
2023-11-16 11:20:00,060 ----------------------------------------------------------------------------------------------------
2023-11-16 11:21:34,000 epoch 10 - iter 750/7500 - loss 0.11743559 - time (sec): 93.94 - samples/sec: 260.10 - lr: 0.000001 - momentum: 0.000000
2023-11-16 11:23:09,621 epoch 10 - iter 1500/7500 - loss 0.12274941 - time (sec): 189.56 - samples/sec: 254.16 - lr: 0.000000 - momentum: 0.000000
2023-11-16 11:24:43,426 epoch 10 - iter 2250/7500 - loss 0.11588386 - time (sec): 283.36 - samples/sec: 255.95 - lr: 0.000000 - momentum: 0.000000
2023-11-16 11:26:20,492 epoch 10 - iter 3000/7500 - loss 0.11238624 - time (sec): 380.43 - samples/sec: 255.77 - lr: 0.000000 - momentum: 0.000000
2023-11-16 11:27:52,662 epoch 10 - iter 3750/7500 - loss 0.11433172 - time (sec): 472.60 - samples/sec: 256.50 - lr: 0.000000 - momentum: 0.000000
2023-11-16 11:29:26,109 epoch 10 - iter 4500/7500 - loss 0.11525478 - time (sec): 566.05 - samples/sec: 256.49 - lr: 0.000000 - momentum: 0.000000
2023-11-16 11:30:58,606 epoch 10 - iter 5250/7500 - loss 0.11793983 - time (sec): 658.54 - samples/sec: 256.98 - lr: 0.000000 - momentum: 0.000000
2023-11-16 11:32:30,455 epoch 10 - iter 6000/7500 - loss 0.11586937 - time (sec): 750.39 - samples/sec: 257.62 - lr: 0.000000 - momentum: 0.000000
2023-11-16 11:34:03,956 epoch 10 - iter 6750/7500 - loss 0.11407474 - time (sec): 843.89 - samples/sec: 257.08 - lr: 0.000000 - momentum: 0.000000
2023-11-16 11:35:37,383 epoch 10 - iter 7500/7500 - loss 0.11399531 - time (sec): 937.32 - samples/sec: 256.90 - lr: 0.000000 - momentum: 0.000000
2023-11-16 11:35:37,386 ----------------------------------------------------------------------------------------------------
2023-11-16 11:35:37,386 EPOCH 10 done: loss 0.1140 - lr: 0.000000
2023-11-16 11:36:05,240 DEV : loss 0.3250023126602173 - f1-score (micro avg) 0.9045
2023-11-16 11:36:08,837 ----------------------------------------------------------------------------------------------------
2023-11-16 11:36:08,840 Loading model from best epoch ...
2023-11-16 11:36:16,806 SequenceTagger predicts: Dictionary with 13 tags: O, S-LOC, B-LOC, E-LOC, I-LOC, S-ORG, B-ORG, E-ORG, I-ORG, S-PER, B-PER, E-PER, I-PER
2023-11-16 11:36:45,465
Results:
- F-score (micro) 0.9042
- F-score (macro) 0.9031
- Accuracy 0.8544
By class:
precision recall f1-score support
LOC 0.9076 0.9106 0.9091 5288
PER 0.9288 0.9419 0.9353 3962
ORG 0.8606 0.8695 0.8650 3807
micro avg 0.9004 0.9081 0.9042 13057
macro avg 0.8990 0.9073 0.9031 13057
weighted avg 0.9004 0.9081 0.9042 13057
2023-11-16 11:36:45,466 ----------------------------------------------------------------------------------------------------
|