File size: 25,537 Bytes
23a9d2a |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 |
2023-10-11 06:39:23,992 ----------------------------------------------------------------------------------------------------
2023-10-11 06:39:23,994 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=17, bias=True)
(loss_function): CrossEntropyLoss()
)"
2023-10-11 06:39:23,994 ----------------------------------------------------------------------------------------------------
2023-10-11 06:39:23,995 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
2023-10-11 06:39:23,995 ----------------------------------------------------------------------------------------------------
2023-10-11 06:39:23,995 Train: 7142 sentences
2023-10-11 06:39:23,995 (train_with_dev=False, train_with_test=False)
2023-10-11 06:39:23,995 ----------------------------------------------------------------------------------------------------
2023-10-11 06:39:23,995 Training Params:
2023-10-11 06:39:23,995 - learning_rate: "0.00016"
2023-10-11 06:39:23,995 - mini_batch_size: "4"
2023-10-11 06:39:23,995 - max_epochs: "10"
2023-10-11 06:39:23,995 - shuffle: "True"
2023-10-11 06:39:23,995 ----------------------------------------------------------------------------------------------------
2023-10-11 06:39:23,995 Plugins:
2023-10-11 06:39:23,995 - TensorboardLogger
2023-10-11 06:39:23,995 - LinearScheduler | warmup_fraction: '0.1'
2023-10-11 06:39:23,995 ----------------------------------------------------------------------------------------------------
2023-10-11 06:39:23,996 Final evaluation on model from best epoch (best-model.pt)
2023-10-11 06:39:23,996 - metric: "('micro avg', 'f1-score')"
2023-10-11 06:39:23,996 ----------------------------------------------------------------------------------------------------
2023-10-11 06:39:23,996 Computation:
2023-10-11 06:39:23,996 - compute on device: cuda:0
2023-10-11 06:39:23,996 - embedding storage: none
2023-10-11 06:39:23,996 ----------------------------------------------------------------------------------------------------
2023-10-11 06:39:23,996 Model training base path: "hmbench-newseye/fr-hmbyt5-preliminary/byt5-small-historic-multilingual-span20-flax-bs4-wsFalse-e10-lr0.00016-poolingfirst-layers-1-crfFalse-2"
2023-10-11 06:39:23,996 ----------------------------------------------------------------------------------------------------
2023-10-11 06:39:23,996 ----------------------------------------------------------------------------------------------------
2023-10-11 06:39:23,996 Logging anything other than scalars to TensorBoard is currently not supported.
2023-10-11 06:40:19,912 epoch 1 - iter 178/1786 - loss 2.82900261 - time (sec): 55.91 - samples/sec: 474.93 - lr: 0.000016 - momentum: 0.000000
2023-10-11 06:41:14,029 epoch 1 - iter 356/1786 - loss 2.67404787 - time (sec): 110.03 - samples/sec: 470.57 - lr: 0.000032 - momentum: 0.000000
2023-10-11 06:42:09,632 epoch 1 - iter 534/1786 - loss 2.38436083 - time (sec): 165.63 - samples/sec: 468.96 - lr: 0.000048 - momentum: 0.000000
2023-10-11 06:43:05,404 epoch 1 - iter 712/1786 - loss 2.06699231 - time (sec): 221.41 - samples/sec: 464.97 - lr: 0.000064 - momentum: 0.000000
2023-10-11 06:44:01,010 epoch 1 - iter 890/1786 - loss 1.80287645 - time (sec): 277.01 - samples/sec: 459.40 - lr: 0.000080 - momentum: 0.000000
2023-10-11 06:44:56,410 epoch 1 - iter 1068/1786 - loss 1.59842649 - time (sec): 332.41 - samples/sec: 460.39 - lr: 0.000096 - momentum: 0.000000
2023-10-11 06:45:50,950 epoch 1 - iter 1246/1786 - loss 1.43900727 - time (sec): 386.95 - samples/sec: 459.78 - lr: 0.000112 - momentum: 0.000000
2023-10-11 06:46:43,346 epoch 1 - iter 1424/1786 - loss 1.31683155 - time (sec): 439.35 - samples/sec: 457.72 - lr: 0.000127 - momentum: 0.000000
2023-10-11 06:47:34,731 epoch 1 - iter 1602/1786 - loss 1.21390367 - time (sec): 490.73 - samples/sec: 457.50 - lr: 0.000143 - momentum: 0.000000
2023-10-11 06:48:27,379 epoch 1 - iter 1780/1786 - loss 1.12581674 - time (sec): 543.38 - samples/sec: 456.87 - lr: 0.000159 - momentum: 0.000000
2023-10-11 06:48:28,795 ----------------------------------------------------------------------------------------------------
2023-10-11 06:48:28,795 EPOCH 1 done: loss 1.1242 - lr: 0.000159
2023-10-11 06:48:48,043 DEV : loss 0.21339952945709229 - f1-score (micro avg) 0.4704
2023-10-11 06:48:48,073 saving best model
2023-10-11 06:48:48,971 ----------------------------------------------------------------------------------------------------
2023-10-11 06:49:40,643 epoch 2 - iter 178/1786 - loss 0.20730434 - time (sec): 51.67 - samples/sec: 486.40 - lr: 0.000158 - momentum: 0.000000
2023-10-11 06:50:33,708 epoch 2 - iter 356/1786 - loss 0.19312469 - time (sec): 104.73 - samples/sec: 491.93 - lr: 0.000156 - momentum: 0.000000
2023-10-11 06:51:25,317 epoch 2 - iter 534/1786 - loss 0.18172318 - time (sec): 156.34 - samples/sec: 488.33 - lr: 0.000155 - momentum: 0.000000
2023-10-11 06:52:16,339 epoch 2 - iter 712/1786 - loss 0.17387846 - time (sec): 207.37 - samples/sec: 484.71 - lr: 0.000153 - momentum: 0.000000
2023-10-11 06:53:07,904 epoch 2 - iter 890/1786 - loss 0.16698512 - time (sec): 258.93 - samples/sec: 481.09 - lr: 0.000151 - momentum: 0.000000
2023-10-11 06:53:58,781 epoch 2 - iter 1068/1786 - loss 0.16051997 - time (sec): 309.81 - samples/sec: 479.18 - lr: 0.000149 - momentum: 0.000000
2023-10-11 06:54:49,112 epoch 2 - iter 1246/1786 - loss 0.15405773 - time (sec): 360.14 - samples/sec: 477.91 - lr: 0.000148 - momentum: 0.000000
2023-10-11 06:55:41,208 epoch 2 - iter 1424/1786 - loss 0.14923648 - time (sec): 412.23 - samples/sec: 478.67 - lr: 0.000146 - momentum: 0.000000
2023-10-11 06:56:35,914 epoch 2 - iter 1602/1786 - loss 0.14482467 - time (sec): 466.94 - samples/sec: 478.44 - lr: 0.000144 - momentum: 0.000000
2023-10-11 06:57:29,822 epoch 2 - iter 1780/1786 - loss 0.13951975 - time (sec): 520.85 - samples/sec: 476.15 - lr: 0.000142 - momentum: 0.000000
2023-10-11 06:57:31,487 ----------------------------------------------------------------------------------------------------
2023-10-11 06:57:31,487 EPOCH 2 done: loss 0.1394 - lr: 0.000142
2023-10-11 06:57:53,454 DEV : loss 0.12035670131444931 - f1-score (micro avg) 0.7478
2023-10-11 06:57:53,485 saving best model
2023-10-11 06:57:56,090 ----------------------------------------------------------------------------------------------------
2023-10-11 06:58:49,895 epoch 3 - iter 178/1786 - loss 0.07648250 - time (sec): 53.80 - samples/sec: 480.38 - lr: 0.000140 - momentum: 0.000000
2023-10-11 06:59:42,427 epoch 3 - iter 356/1786 - loss 0.07475891 - time (sec): 106.33 - samples/sec: 457.73 - lr: 0.000139 - momentum: 0.000000
2023-10-11 07:00:36,639 epoch 3 - iter 534/1786 - loss 0.07760944 - time (sec): 160.54 - samples/sec: 461.01 - lr: 0.000137 - momentum: 0.000000
2023-10-11 07:01:30,450 epoch 3 - iter 712/1786 - loss 0.07885752 - time (sec): 214.36 - samples/sec: 458.24 - lr: 0.000135 - momentum: 0.000000
2023-10-11 07:02:24,329 epoch 3 - iter 890/1786 - loss 0.07553472 - time (sec): 268.23 - samples/sec: 458.26 - lr: 0.000133 - momentum: 0.000000
2023-10-11 07:03:19,934 epoch 3 - iter 1068/1786 - loss 0.07803250 - time (sec): 323.84 - samples/sec: 457.48 - lr: 0.000132 - momentum: 0.000000
2023-10-11 07:04:18,636 epoch 3 - iter 1246/1786 - loss 0.07541694 - time (sec): 382.54 - samples/sec: 456.93 - lr: 0.000130 - momentum: 0.000000
2023-10-11 07:05:16,926 epoch 3 - iter 1424/1786 - loss 0.07566700 - time (sec): 440.83 - samples/sec: 452.91 - lr: 0.000128 - momentum: 0.000000
2023-10-11 07:06:14,488 epoch 3 - iter 1602/1786 - loss 0.07521801 - time (sec): 498.39 - samples/sec: 449.09 - lr: 0.000126 - momentum: 0.000000
2023-10-11 07:07:11,671 epoch 3 - iter 1780/1786 - loss 0.07599665 - time (sec): 555.58 - samples/sec: 446.70 - lr: 0.000125 - momentum: 0.000000
2023-10-11 07:07:13,371 ----------------------------------------------------------------------------------------------------
2023-10-11 07:07:13,371 EPOCH 3 done: loss 0.0763 - lr: 0.000125
2023-10-11 07:07:36,874 DEV : loss 0.12293554842472076 - f1-score (micro avg) 0.7694
2023-10-11 07:07:36,908 saving best model
2023-10-11 07:07:39,619 ----------------------------------------------------------------------------------------------------
2023-10-11 07:08:32,498 epoch 4 - iter 178/1786 - loss 0.06175889 - time (sec): 52.87 - samples/sec: 468.15 - lr: 0.000123 - momentum: 0.000000
2023-10-11 07:09:25,152 epoch 4 - iter 356/1786 - loss 0.05291025 - time (sec): 105.53 - samples/sec: 453.32 - lr: 0.000121 - momentum: 0.000000
2023-10-11 07:10:25,644 epoch 4 - iter 534/1786 - loss 0.05155940 - time (sec): 166.02 - samples/sec: 445.22 - lr: 0.000119 - momentum: 0.000000
2023-10-11 07:11:24,877 epoch 4 - iter 712/1786 - loss 0.05289251 - time (sec): 225.25 - samples/sec: 446.72 - lr: 0.000117 - momentum: 0.000000
2023-10-11 07:12:16,721 epoch 4 - iter 890/1786 - loss 0.05247741 - time (sec): 277.10 - samples/sec: 446.69 - lr: 0.000116 - momentum: 0.000000
2023-10-11 07:13:08,568 epoch 4 - iter 1068/1786 - loss 0.05140588 - time (sec): 328.94 - samples/sec: 451.81 - lr: 0.000114 - momentum: 0.000000
2023-10-11 07:14:02,792 epoch 4 - iter 1246/1786 - loss 0.05313146 - time (sec): 383.17 - samples/sec: 455.46 - lr: 0.000112 - momentum: 0.000000
2023-10-11 07:14:58,620 epoch 4 - iter 1424/1786 - loss 0.05344596 - time (sec): 439.00 - samples/sec: 452.79 - lr: 0.000110 - momentum: 0.000000
2023-10-11 07:15:55,417 epoch 4 - iter 1602/1786 - loss 0.05307124 - time (sec): 495.79 - samples/sec: 450.00 - lr: 0.000109 - momentum: 0.000000
2023-10-11 07:16:48,635 epoch 4 - iter 1780/1786 - loss 0.05340878 - time (sec): 549.01 - samples/sec: 452.21 - lr: 0.000107 - momentum: 0.000000
2023-10-11 07:16:50,081 ----------------------------------------------------------------------------------------------------
2023-10-11 07:16:50,081 EPOCH 4 done: loss 0.0533 - lr: 0.000107
2023-10-11 07:17:11,531 DEV : loss 0.15573516488075256 - f1-score (micro avg) 0.7904
2023-10-11 07:17:11,562 saving best model
2023-10-11 07:17:14,160 ----------------------------------------------------------------------------------------------------
2023-10-11 07:18:09,045 epoch 5 - iter 178/1786 - loss 0.04583648 - time (sec): 54.87 - samples/sec: 453.51 - lr: 0.000105 - momentum: 0.000000
2023-10-11 07:19:01,699 epoch 5 - iter 356/1786 - loss 0.04299453 - time (sec): 107.53 - samples/sec: 442.06 - lr: 0.000103 - momentum: 0.000000
2023-10-11 07:19:53,208 epoch 5 - iter 534/1786 - loss 0.04267254 - time (sec): 159.04 - samples/sec: 455.11 - lr: 0.000101 - momentum: 0.000000
2023-10-11 07:20:50,338 epoch 5 - iter 712/1786 - loss 0.04135181 - time (sec): 216.17 - samples/sec: 450.26 - lr: 0.000100 - momentum: 0.000000
2023-10-11 07:21:51,130 epoch 5 - iter 890/1786 - loss 0.04091209 - time (sec): 276.96 - samples/sec: 437.86 - lr: 0.000098 - momentum: 0.000000
2023-10-11 07:22:51,933 epoch 5 - iter 1068/1786 - loss 0.03936224 - time (sec): 337.76 - samples/sec: 433.05 - lr: 0.000096 - momentum: 0.000000
2023-10-11 07:23:51,852 epoch 5 - iter 1246/1786 - loss 0.03988643 - time (sec): 397.68 - samples/sec: 434.38 - lr: 0.000094 - momentum: 0.000000
2023-10-11 07:24:47,391 epoch 5 - iter 1424/1786 - loss 0.04034676 - time (sec): 453.22 - samples/sec: 436.01 - lr: 0.000093 - momentum: 0.000000
2023-10-11 07:25:43,025 epoch 5 - iter 1602/1786 - loss 0.03940605 - time (sec): 508.85 - samples/sec: 436.78 - lr: 0.000091 - momentum: 0.000000
2023-10-11 07:26:37,195 epoch 5 - iter 1780/1786 - loss 0.03943045 - time (sec): 563.02 - samples/sec: 440.61 - lr: 0.000089 - momentum: 0.000000
2023-10-11 07:26:38,788 ----------------------------------------------------------------------------------------------------
2023-10-11 07:26:38,788 EPOCH 5 done: loss 0.0393 - lr: 0.000089
2023-10-11 07:26:59,794 DEV : loss 0.16520686447620392 - f1-score (micro avg) 0.7995
2023-10-11 07:26:59,824 saving best model
2023-10-11 07:27:02,424 ----------------------------------------------------------------------------------------------------
2023-10-11 07:27:57,899 epoch 6 - iter 178/1786 - loss 0.03094883 - time (sec): 55.47 - samples/sec: 446.32 - lr: 0.000087 - momentum: 0.000000
2023-10-11 07:28:53,808 epoch 6 - iter 356/1786 - loss 0.03037107 - time (sec): 111.38 - samples/sec: 447.38 - lr: 0.000085 - momentum: 0.000000
2023-10-11 07:29:48,736 epoch 6 - iter 534/1786 - loss 0.02899602 - time (sec): 166.31 - samples/sec: 447.72 - lr: 0.000084 - momentum: 0.000000
2023-10-11 07:30:44,570 epoch 6 - iter 712/1786 - loss 0.02900632 - time (sec): 222.14 - samples/sec: 446.71 - lr: 0.000082 - momentum: 0.000000
2023-10-11 07:31:40,780 epoch 6 - iter 890/1786 - loss 0.02906074 - time (sec): 278.35 - samples/sec: 446.06 - lr: 0.000080 - momentum: 0.000000
2023-10-11 07:32:38,978 epoch 6 - iter 1068/1786 - loss 0.02883507 - time (sec): 336.55 - samples/sec: 446.30 - lr: 0.000078 - momentum: 0.000000
2023-10-11 07:33:33,658 epoch 6 - iter 1246/1786 - loss 0.02715296 - time (sec): 391.23 - samples/sec: 445.77 - lr: 0.000077 - momentum: 0.000000
2023-10-11 07:34:29,044 epoch 6 - iter 1424/1786 - loss 0.02742899 - time (sec): 446.62 - samples/sec: 446.73 - lr: 0.000075 - momentum: 0.000000
2023-10-11 07:35:22,481 epoch 6 - iter 1602/1786 - loss 0.02809173 - time (sec): 500.05 - samples/sec: 450.53 - lr: 0.000073 - momentum: 0.000000
2023-10-11 07:36:15,937 epoch 6 - iter 1780/1786 - loss 0.02903797 - time (sec): 553.51 - samples/sec: 448.58 - lr: 0.000071 - momentum: 0.000000
2023-10-11 07:36:17,437 ----------------------------------------------------------------------------------------------------
2023-10-11 07:36:17,438 EPOCH 6 done: loss 0.0290 - lr: 0.000071
2023-10-11 07:36:42,730 DEV : loss 0.16866964101791382 - f1-score (micro avg) 0.7947
2023-10-11 07:36:42,764 ----------------------------------------------------------------------------------------------------
2023-10-11 07:37:41,091 epoch 7 - iter 178/1786 - loss 0.02176058 - time (sec): 58.33 - samples/sec: 469.08 - lr: 0.000069 - momentum: 0.000000
2023-10-11 07:38:39,578 epoch 7 - iter 356/1786 - loss 0.02292926 - time (sec): 116.81 - samples/sec: 445.15 - lr: 0.000068 - momentum: 0.000000
2023-10-11 07:39:34,898 epoch 7 - iter 534/1786 - loss 0.02247289 - time (sec): 172.13 - samples/sec: 439.81 - lr: 0.000066 - momentum: 0.000000
2023-10-11 07:40:31,803 epoch 7 - iter 712/1786 - loss 0.02241562 - time (sec): 229.04 - samples/sec: 436.51 - lr: 0.000064 - momentum: 0.000000
2023-10-11 07:41:27,757 epoch 7 - iter 890/1786 - loss 0.02220387 - time (sec): 284.99 - samples/sec: 436.33 - lr: 0.000062 - momentum: 0.000000
2023-10-11 07:42:23,342 epoch 7 - iter 1068/1786 - loss 0.02137649 - time (sec): 340.58 - samples/sec: 440.01 - lr: 0.000061 - momentum: 0.000000
2023-10-11 07:43:18,915 epoch 7 - iter 1246/1786 - loss 0.02099827 - time (sec): 396.15 - samples/sec: 438.43 - lr: 0.000059 - momentum: 0.000000
2023-10-11 07:44:18,289 epoch 7 - iter 1424/1786 - loss 0.02038066 - time (sec): 455.52 - samples/sec: 439.21 - lr: 0.000057 - momentum: 0.000000
2023-10-11 07:45:16,495 epoch 7 - iter 1602/1786 - loss 0.01985291 - time (sec): 513.73 - samples/sec: 437.28 - lr: 0.000055 - momentum: 0.000000
2023-10-11 07:46:09,692 epoch 7 - iter 1780/1786 - loss 0.02003687 - time (sec): 566.93 - samples/sec: 437.53 - lr: 0.000053 - momentum: 0.000000
2023-10-11 07:46:11,314 ----------------------------------------------------------------------------------------------------
2023-10-11 07:46:11,314 EPOCH 7 done: loss 0.0201 - lr: 0.000053
2023-10-11 07:46:32,834 DEV : loss 0.19440826773643494 - f1-score (micro avg) 0.7954
2023-10-11 07:46:32,868 ----------------------------------------------------------------------------------------------------
2023-10-11 07:47:26,349 epoch 8 - iter 178/1786 - loss 0.02118880 - time (sec): 53.48 - samples/sec: 450.21 - lr: 0.000052 - momentum: 0.000000
2023-10-11 07:48:21,828 epoch 8 - iter 356/1786 - loss 0.01874784 - time (sec): 108.96 - samples/sec: 445.93 - lr: 0.000050 - momentum: 0.000000
2023-10-11 07:49:16,307 epoch 8 - iter 534/1786 - loss 0.01722056 - time (sec): 163.44 - samples/sec: 450.09 - lr: 0.000048 - momentum: 0.000000
2023-10-11 07:50:13,820 epoch 8 - iter 712/1786 - loss 0.01921550 - time (sec): 220.95 - samples/sec: 449.54 - lr: 0.000046 - momentum: 0.000000
2023-10-11 07:51:10,663 epoch 8 - iter 890/1786 - loss 0.01831347 - time (sec): 277.79 - samples/sec: 444.64 - lr: 0.000044 - momentum: 0.000000
2023-10-11 07:52:09,449 epoch 8 - iter 1068/1786 - loss 0.01900935 - time (sec): 336.58 - samples/sec: 440.68 - lr: 0.000043 - momentum: 0.000000
2023-10-11 07:53:04,351 epoch 8 - iter 1246/1786 - loss 0.01828550 - time (sec): 391.48 - samples/sec: 441.25 - lr: 0.000041 - momentum: 0.000000
2023-10-11 07:54:00,162 epoch 8 - iter 1424/1786 - loss 0.01722443 - time (sec): 447.29 - samples/sec: 443.60 - lr: 0.000039 - momentum: 0.000000
2023-10-11 07:54:56,327 epoch 8 - iter 1602/1786 - loss 0.01696995 - time (sec): 503.46 - samples/sec: 442.39 - lr: 0.000037 - momentum: 0.000000
2023-10-11 07:55:53,224 epoch 8 - iter 1780/1786 - loss 0.01668282 - time (sec): 560.35 - samples/sec: 442.21 - lr: 0.000036 - momentum: 0.000000
2023-10-11 07:55:55,029 ----------------------------------------------------------------------------------------------------
2023-10-11 07:55:55,030 EPOCH 8 done: loss 0.0167 - lr: 0.000036
2023-10-11 07:56:17,077 DEV : loss 0.1955002099275589 - f1-score (micro avg) 0.803
2023-10-11 07:56:17,108 saving best model
2023-10-11 07:56:19,773 ----------------------------------------------------------------------------------------------------
2023-10-11 07:57:19,182 epoch 9 - iter 178/1786 - loss 0.00753763 - time (sec): 59.40 - samples/sec: 434.60 - lr: 0.000034 - momentum: 0.000000
2023-10-11 07:58:14,298 epoch 9 - iter 356/1786 - loss 0.01214607 - time (sec): 114.52 - samples/sec: 438.67 - lr: 0.000032 - momentum: 0.000000
2023-10-11 07:59:07,909 epoch 9 - iter 534/1786 - loss 0.00983533 - time (sec): 168.13 - samples/sec: 445.20 - lr: 0.000030 - momentum: 0.000000
2023-10-11 08:00:01,122 epoch 9 - iter 712/1786 - loss 0.01002213 - time (sec): 221.34 - samples/sec: 448.23 - lr: 0.000028 - momentum: 0.000000
2023-10-11 08:00:56,745 epoch 9 - iter 890/1786 - loss 0.00888437 - time (sec): 276.97 - samples/sec: 447.41 - lr: 0.000027 - momentum: 0.000000
2023-10-11 08:01:52,410 epoch 9 - iter 1068/1786 - loss 0.00895876 - time (sec): 332.63 - samples/sec: 446.60 - lr: 0.000025 - momentum: 0.000000
2023-10-11 08:02:51,663 epoch 9 - iter 1246/1786 - loss 0.00900417 - time (sec): 391.89 - samples/sec: 437.63 - lr: 0.000023 - momentum: 0.000000
2023-10-11 08:03:51,512 epoch 9 - iter 1424/1786 - loss 0.01017587 - time (sec): 451.73 - samples/sec: 436.07 - lr: 0.000021 - momentum: 0.000000
2023-10-11 08:04:46,799 epoch 9 - iter 1602/1786 - loss 0.01074771 - time (sec): 507.02 - samples/sec: 440.42 - lr: 0.000020 - momentum: 0.000000
2023-10-11 08:05:40,266 epoch 9 - iter 1780/1786 - loss 0.01046865 - time (sec): 560.49 - samples/sec: 442.27 - lr: 0.000018 - momentum: 0.000000
2023-10-11 08:05:42,023 ----------------------------------------------------------------------------------------------------
2023-10-11 08:05:42,023 EPOCH 9 done: loss 0.0104 - lr: 0.000018
2023-10-11 08:06:04,094 DEV : loss 0.21326804161071777 - f1-score (micro avg) 0.7971
2023-10-11 08:06:04,125 ----------------------------------------------------------------------------------------------------
2023-10-11 08:06:58,019 epoch 10 - iter 178/1786 - loss 0.00672035 - time (sec): 53.89 - samples/sec: 436.00 - lr: 0.000016 - momentum: 0.000000
2023-10-11 08:07:53,539 epoch 10 - iter 356/1786 - loss 0.00687932 - time (sec): 109.41 - samples/sec: 441.96 - lr: 0.000014 - momentum: 0.000000
2023-10-11 08:08:52,210 epoch 10 - iter 534/1786 - loss 0.00685882 - time (sec): 168.08 - samples/sec: 439.86 - lr: 0.000012 - momentum: 0.000000
2023-10-11 08:09:49,775 epoch 10 - iter 712/1786 - loss 0.00638074 - time (sec): 225.65 - samples/sec: 433.50 - lr: 0.000011 - momentum: 0.000000
2023-10-11 08:10:46,408 epoch 10 - iter 890/1786 - loss 0.00750964 - time (sec): 282.28 - samples/sec: 439.14 - lr: 0.000009 - momentum: 0.000000
2023-10-11 08:11:42,983 epoch 10 - iter 1068/1786 - loss 0.00846706 - time (sec): 338.86 - samples/sec: 444.69 - lr: 0.000007 - momentum: 0.000000
2023-10-11 08:12:37,467 epoch 10 - iter 1246/1786 - loss 0.00856117 - time (sec): 393.34 - samples/sec: 441.99 - lr: 0.000005 - momentum: 0.000000
2023-10-11 08:13:35,598 epoch 10 - iter 1424/1786 - loss 0.00826761 - time (sec): 451.47 - samples/sec: 441.54 - lr: 0.000004 - momentum: 0.000000
2023-10-11 08:14:36,646 epoch 10 - iter 1602/1786 - loss 0.00793073 - time (sec): 512.52 - samples/sec: 436.65 - lr: 0.000002 - momentum: 0.000000
2023-10-11 08:15:31,095 epoch 10 - iter 1780/1786 - loss 0.00780604 - time (sec): 566.97 - samples/sec: 437.59 - lr: 0.000000 - momentum: 0.000000
2023-10-11 08:15:32,665 ----------------------------------------------------------------------------------------------------
2023-10-11 08:15:32,666 EPOCH 10 done: loss 0.0078 - lr: 0.000000
2023-10-11 08:15:53,506 DEV : loss 0.21688954532146454 - f1-score (micro avg) 0.7973
2023-10-11 08:15:54,599 ----------------------------------------------------------------------------------------------------
2023-10-11 08:15:54,601 Loading model from best epoch ...
2023-10-11 08:15:58,969 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
2023-10-11 08:17:06,782
Results:
- F-score (micro) 0.7048
- F-score (macro) 0.6244
- Accuracy 0.5644
By class:
precision recall f1-score support
LOC 0.7416 0.7078 0.7243 1095
PER 0.7849 0.7826 0.7838 1012
ORG 0.4219 0.5826 0.4894 357
HumanProd 0.4000 0.6667 0.5000 33
micro avg 0.6906 0.7197 0.7048 2497
macro avg 0.5871 0.6849 0.6244 2497
weighted avg 0.7090 0.7197 0.7119 2497
2023-10-11 08:17:06,782 ----------------------------------------------------------------------------------------------------
|