File size: 25,355 Bytes
35e202b |
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 |
2023-10-14 14:40:59,478 ----------------------------------------------------------------------------------------------------
2023-10-14 14:40:59,480 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()
)"
2023-10-14 14:40:59,481 ----------------------------------------------------------------------------------------------------
2023-10-14 14:40:59,481 MultiCorpus: 14465 train + 1392 dev + 2432 test sentences
- NER_HIPE_2022 Corpus: 14465 train + 1392 dev + 2432 test sentences - /root/.flair/datasets/ner_hipe_2022/v2.1/letemps/fr/with_doc_seperator
2023-10-14 14:40:59,481 ----------------------------------------------------------------------------------------------------
2023-10-14 14:40:59,481 Train: 14465 sentences
2023-10-14 14:40:59,481 (train_with_dev=False, train_with_test=False)
2023-10-14 14:40:59,481 ----------------------------------------------------------------------------------------------------
2023-10-14 14:40:59,481 Training Params:
2023-10-14 14:40:59,481 - learning_rate: "0.00015"
2023-10-14 14:40:59,481 - mini_batch_size: "4"
2023-10-14 14:40:59,481 - max_epochs: "10"
2023-10-14 14:40:59,481 - shuffle: "True"
2023-10-14 14:40:59,481 ----------------------------------------------------------------------------------------------------
2023-10-14 14:40:59,481 Plugins:
2023-10-14 14:40:59,481 - TensorboardLogger
2023-10-14 14:40:59,482 - LinearScheduler | warmup_fraction: '0.1'
2023-10-14 14:40:59,482 ----------------------------------------------------------------------------------------------------
2023-10-14 14:40:59,482 Final evaluation on model from best epoch (best-model.pt)
2023-10-14 14:40:59,482 - metric: "('micro avg', 'f1-score')"
2023-10-14 14:40:59,482 ----------------------------------------------------------------------------------------------------
2023-10-14 14:40:59,482 Computation:
2023-10-14 14:40:59,482 - compute on device: cuda:0
2023-10-14 14:40:59,482 - embedding storage: none
2023-10-14 14:40:59,482 ----------------------------------------------------------------------------------------------------
2023-10-14 14:40:59,482 Model training base path: "hmbench-letemps/fr-hmbyt5-preliminary/byt5-small-historic-multilingual-span20-flax-bs4-wsFalse-e10-lr0.00015-poolingfirst-layers-1-crfFalse-4"
2023-10-14 14:40:59,482 ----------------------------------------------------------------------------------------------------
2023-10-14 14:40:59,482 ----------------------------------------------------------------------------------------------------
2023-10-14 14:40:59,482 Logging anything other than scalars to TensorBoard is currently not supported.
2023-10-14 14:42:41,066 epoch 1 - iter 361/3617 - loss 2.50686900 - time (sec): 101.58 - samples/sec: 360.10 - lr: 0.000015 - momentum: 0.000000
2023-10-14 14:44:22,406 epoch 1 - iter 722/3617 - loss 2.10752883 - time (sec): 202.92 - samples/sec: 369.39 - lr: 0.000030 - momentum: 0.000000
2023-10-14 14:46:12,979 epoch 1 - iter 1083/3617 - loss 1.65683609 - time (sec): 313.49 - samples/sec: 363.76 - lr: 0.000045 - momentum: 0.000000
2023-10-14 14:47:57,110 epoch 1 - iter 1444/3617 - loss 1.31501097 - time (sec): 417.63 - samples/sec: 364.72 - lr: 0.000060 - momentum: 0.000000
2023-10-14 14:49:44,051 epoch 1 - iter 1805/3617 - loss 1.09620008 - time (sec): 524.57 - samples/sec: 361.95 - lr: 0.000075 - momentum: 0.000000
2023-10-14 14:51:28,614 epoch 1 - iter 2166/3617 - loss 0.93978591 - time (sec): 629.13 - samples/sec: 364.30 - lr: 0.000090 - momentum: 0.000000
2023-10-14 14:53:09,495 epoch 1 - iter 2527/3617 - loss 0.82889421 - time (sec): 730.01 - samples/sec: 366.34 - lr: 0.000105 - momentum: 0.000000
2023-10-14 14:54:50,772 epoch 1 - iter 2888/3617 - loss 0.74534435 - time (sec): 831.29 - samples/sec: 367.53 - lr: 0.000120 - momentum: 0.000000
2023-10-14 14:56:30,263 epoch 1 - iter 3249/3617 - loss 0.67952585 - time (sec): 930.78 - samples/sec: 367.24 - lr: 0.000135 - momentum: 0.000000
2023-10-14 14:58:13,235 epoch 1 - iter 3610/3617 - loss 0.62441070 - time (sec): 1033.75 - samples/sec: 366.94 - lr: 0.000150 - momentum: 0.000000
2023-10-14 14:58:14,910 ----------------------------------------------------------------------------------------------------
2023-10-14 14:58:14,910 EPOCH 1 done: loss 0.6239 - lr: 0.000150
2023-10-14 14:58:53,380 DEV : loss 0.11717528849840164 - f1-score (micro avg) 0.5796
2023-10-14 14:58:53,438 saving best model
2023-10-14 14:58:54,356 ----------------------------------------------------------------------------------------------------
2023-10-14 15:00:36,235 epoch 2 - iter 361/3617 - loss 0.10878939 - time (sec): 101.88 - samples/sec: 370.35 - lr: 0.000148 - momentum: 0.000000
2023-10-14 15:02:20,856 epoch 2 - iter 722/3617 - loss 0.10496323 - time (sec): 206.50 - samples/sec: 369.37 - lr: 0.000147 - momentum: 0.000000
2023-10-14 15:04:10,294 epoch 2 - iter 1083/3617 - loss 0.10512196 - time (sec): 315.94 - samples/sec: 358.08 - lr: 0.000145 - momentum: 0.000000
2023-10-14 15:05:52,218 epoch 2 - iter 1444/3617 - loss 0.10247324 - time (sec): 417.86 - samples/sec: 360.21 - lr: 0.000143 - momentum: 0.000000
2023-10-14 15:07:32,739 epoch 2 - iter 1805/3617 - loss 0.10015258 - time (sec): 518.38 - samples/sec: 362.60 - lr: 0.000142 - momentum: 0.000000
2023-10-14 15:09:12,203 epoch 2 - iter 2166/3617 - loss 0.09832530 - time (sec): 617.84 - samples/sec: 366.60 - lr: 0.000140 - momentum: 0.000000
2023-10-14 15:10:55,993 epoch 2 - iter 2527/3617 - loss 0.09600180 - time (sec): 721.63 - samples/sec: 368.59 - lr: 0.000138 - momentum: 0.000000
2023-10-14 15:12:42,577 epoch 2 - iter 2888/3617 - loss 0.09489153 - time (sec): 828.22 - samples/sec: 367.36 - lr: 0.000137 - momentum: 0.000000
2023-10-14 15:14:25,871 epoch 2 - iter 3249/3617 - loss 0.09188276 - time (sec): 931.51 - samples/sec: 367.52 - lr: 0.000135 - momentum: 0.000000
2023-10-14 15:16:09,934 epoch 2 - iter 3610/3617 - loss 0.09054350 - time (sec): 1035.58 - samples/sec: 366.37 - lr: 0.000133 - momentum: 0.000000
2023-10-14 15:16:11,806 ----------------------------------------------------------------------------------------------------
2023-10-14 15:16:11,807 EPOCH 2 done: loss 0.0909 - lr: 0.000133
2023-10-14 15:16:51,726 DEV : loss 0.11103517562150955 - f1-score (micro avg) 0.6208
2023-10-14 15:16:51,794 saving best model
2023-10-14 15:16:57,345 ----------------------------------------------------------------------------------------------------
2023-10-14 15:18:47,930 epoch 3 - iter 361/3617 - loss 0.06378214 - time (sec): 110.58 - samples/sec: 346.95 - lr: 0.000132 - momentum: 0.000000
2023-10-14 15:20:27,753 epoch 3 - iter 722/3617 - loss 0.06540151 - time (sec): 210.40 - samples/sec: 356.29 - lr: 0.000130 - momentum: 0.000000
2023-10-14 15:22:11,280 epoch 3 - iter 1083/3617 - loss 0.06698159 - time (sec): 313.93 - samples/sec: 359.05 - lr: 0.000128 - momentum: 0.000000
2023-10-14 15:23:52,034 epoch 3 - iter 1444/3617 - loss 0.06526504 - time (sec): 414.68 - samples/sec: 366.94 - lr: 0.000127 - momentum: 0.000000
2023-10-14 15:25:35,189 epoch 3 - iter 1805/3617 - loss 0.06525753 - time (sec): 517.84 - samples/sec: 368.07 - lr: 0.000125 - momentum: 0.000000
2023-10-14 15:27:21,741 epoch 3 - iter 2166/3617 - loss 0.06493734 - time (sec): 624.39 - samples/sec: 367.06 - lr: 0.000123 - momentum: 0.000000
2023-10-14 15:29:02,837 epoch 3 - iter 2527/3617 - loss 0.06477906 - time (sec): 725.49 - samples/sec: 368.41 - lr: 0.000122 - momentum: 0.000000
2023-10-14 15:30:43,338 epoch 3 - iter 2888/3617 - loss 0.06486032 - time (sec): 825.99 - samples/sec: 367.73 - lr: 0.000120 - momentum: 0.000000
2023-10-14 15:32:21,513 epoch 3 - iter 3249/3617 - loss 0.06429360 - time (sec): 924.16 - samples/sec: 369.65 - lr: 0.000118 - momentum: 0.000000
2023-10-14 15:34:00,097 epoch 3 - iter 3610/3617 - loss 0.06473516 - time (sec): 1022.75 - samples/sec: 370.86 - lr: 0.000117 - momentum: 0.000000
2023-10-14 15:34:01,771 ----------------------------------------------------------------------------------------------------
2023-10-14 15:34:01,772 EPOCH 3 done: loss 0.0648 - lr: 0.000117
2023-10-14 15:34:42,878 DEV : loss 0.1630961298942566 - f1-score (micro avg) 0.6158
2023-10-14 15:34:42,948 ----------------------------------------------------------------------------------------------------
2023-10-14 15:36:23,601 epoch 4 - iter 361/3617 - loss 0.04594123 - time (sec): 100.65 - samples/sec: 363.48 - lr: 0.000115 - momentum: 0.000000
2023-10-14 15:38:03,903 epoch 4 - iter 722/3617 - loss 0.04874699 - time (sec): 200.95 - samples/sec: 370.90 - lr: 0.000113 - momentum: 0.000000
2023-10-14 15:39:47,633 epoch 4 - iter 1083/3617 - loss 0.04891474 - time (sec): 304.68 - samples/sec: 372.99 - lr: 0.000112 - momentum: 0.000000
2023-10-14 15:41:28,223 epoch 4 - iter 1444/3617 - loss 0.04690533 - time (sec): 405.27 - samples/sec: 370.54 - lr: 0.000110 - momentum: 0.000000
2023-10-14 15:43:08,244 epoch 4 - iter 1805/3617 - loss 0.04706884 - time (sec): 505.29 - samples/sec: 371.67 - lr: 0.000108 - momentum: 0.000000
2023-10-14 15:44:53,122 epoch 4 - iter 2166/3617 - loss 0.04732382 - time (sec): 610.17 - samples/sec: 371.22 - lr: 0.000107 - momentum: 0.000000
2023-10-14 15:46:34,153 epoch 4 - iter 2527/3617 - loss 0.04755205 - time (sec): 711.20 - samples/sec: 372.43 - lr: 0.000105 - momentum: 0.000000
2023-10-14 15:48:16,775 epoch 4 - iter 2888/3617 - loss 0.04692890 - time (sec): 813.82 - samples/sec: 373.54 - lr: 0.000103 - momentum: 0.000000
2023-10-14 15:49:58,502 epoch 4 - iter 3249/3617 - loss 0.04641577 - time (sec): 915.55 - samples/sec: 372.42 - lr: 0.000102 - momentum: 0.000000
2023-10-14 15:51:37,654 epoch 4 - iter 3610/3617 - loss 0.04580765 - time (sec): 1014.70 - samples/sec: 373.81 - lr: 0.000100 - momentum: 0.000000
2023-10-14 15:51:39,377 ----------------------------------------------------------------------------------------------------
2023-10-14 15:51:39,378 EPOCH 4 done: loss 0.0458 - lr: 0.000100
2023-10-14 15:52:18,475 DEV : loss 0.21207064390182495 - f1-score (micro avg) 0.6575
2023-10-14 15:52:18,532 saving best model
2023-10-14 15:52:21,265 ----------------------------------------------------------------------------------------------------
2023-10-14 15:53:59,089 epoch 5 - iter 361/3617 - loss 0.02339736 - time (sec): 97.82 - samples/sec: 384.30 - lr: 0.000098 - momentum: 0.000000
2023-10-14 15:55:43,726 epoch 5 - iter 722/3617 - loss 0.02553085 - time (sec): 202.46 - samples/sec: 387.78 - lr: 0.000097 - momentum: 0.000000
2023-10-14 15:57:28,602 epoch 5 - iter 1083/3617 - loss 0.02625521 - time (sec): 307.33 - samples/sec: 381.45 - lr: 0.000095 - momentum: 0.000000
2023-10-14 15:59:07,613 epoch 5 - iter 1444/3617 - loss 0.02682998 - time (sec): 406.34 - samples/sec: 378.67 - lr: 0.000093 - momentum: 0.000000
2023-10-14 16:00:54,191 epoch 5 - iter 1805/3617 - loss 0.02777093 - time (sec): 512.92 - samples/sec: 371.65 - lr: 0.000092 - momentum: 0.000000
2023-10-14 16:02:34,896 epoch 5 - iter 2166/3617 - loss 0.02927279 - time (sec): 613.63 - samples/sec: 373.11 - lr: 0.000090 - momentum: 0.000000
2023-10-14 16:04:12,788 epoch 5 - iter 2527/3617 - loss 0.02908364 - time (sec): 711.52 - samples/sec: 374.65 - lr: 0.000088 - momentum: 0.000000
2023-10-14 16:05:50,506 epoch 5 - iter 2888/3617 - loss 0.02954096 - time (sec): 809.24 - samples/sec: 375.84 - lr: 0.000087 - momentum: 0.000000
2023-10-14 16:07:30,269 epoch 5 - iter 3249/3617 - loss 0.03004065 - time (sec): 909.00 - samples/sec: 375.39 - lr: 0.000085 - momentum: 0.000000
2023-10-14 16:09:09,000 epoch 5 - iter 3610/3617 - loss 0.03067994 - time (sec): 1007.73 - samples/sec: 376.30 - lr: 0.000083 - momentum: 0.000000
2023-10-14 16:09:10,704 ----------------------------------------------------------------------------------------------------
2023-10-14 16:09:10,705 EPOCH 5 done: loss 0.0306 - lr: 0.000083
2023-10-14 16:09:49,418 DEV : loss 0.2788721024990082 - f1-score (micro avg) 0.6128
2023-10-14 16:09:49,475 ----------------------------------------------------------------------------------------------------
2023-10-14 16:11:32,473 epoch 6 - iter 361/3617 - loss 0.02077887 - time (sec): 103.00 - samples/sec: 381.48 - lr: 0.000082 - momentum: 0.000000
2023-10-14 16:13:14,265 epoch 6 - iter 722/3617 - loss 0.02070222 - time (sec): 204.79 - samples/sec: 373.48 - lr: 0.000080 - momentum: 0.000000
2023-10-14 16:14:54,315 epoch 6 - iter 1083/3617 - loss 0.01942465 - time (sec): 304.84 - samples/sec: 374.09 - lr: 0.000078 - momentum: 0.000000
2023-10-14 16:16:32,568 epoch 6 - iter 1444/3617 - loss 0.02052837 - time (sec): 403.09 - samples/sec: 375.08 - lr: 0.000077 - momentum: 0.000000
2023-10-14 16:18:15,025 epoch 6 - iter 1805/3617 - loss 0.02128468 - time (sec): 505.55 - samples/sec: 372.79 - lr: 0.000075 - momentum: 0.000000
2023-10-14 16:19:57,182 epoch 6 - iter 2166/3617 - loss 0.02145412 - time (sec): 607.70 - samples/sec: 373.59 - lr: 0.000073 - momentum: 0.000000
2023-10-14 16:21:37,405 epoch 6 - iter 2527/3617 - loss 0.02127559 - time (sec): 707.93 - samples/sec: 373.85 - lr: 0.000072 - momentum: 0.000000
2023-10-14 16:23:17,748 epoch 6 - iter 2888/3617 - loss 0.02174407 - time (sec): 808.27 - samples/sec: 375.50 - lr: 0.000070 - momentum: 0.000000
2023-10-14 16:24:58,379 epoch 6 - iter 3249/3617 - loss 0.02165837 - time (sec): 908.90 - samples/sec: 375.13 - lr: 0.000068 - momentum: 0.000000
2023-10-14 16:26:39,103 epoch 6 - iter 3610/3617 - loss 0.02119253 - time (sec): 1009.63 - samples/sec: 375.80 - lr: 0.000067 - momentum: 0.000000
2023-10-14 16:26:41,074 ----------------------------------------------------------------------------------------------------
2023-10-14 16:26:41,074 EPOCH 6 done: loss 0.0212 - lr: 0.000067
2023-10-14 16:27:20,798 DEV : loss 0.313930869102478 - f1-score (micro avg) 0.6284
2023-10-14 16:27:20,855 ----------------------------------------------------------------------------------------------------
2023-10-14 16:29:06,678 epoch 7 - iter 361/3617 - loss 0.01422485 - time (sec): 105.82 - samples/sec: 383.75 - lr: 0.000065 - momentum: 0.000000
2023-10-14 16:30:54,879 epoch 7 - iter 722/3617 - loss 0.01422789 - time (sec): 214.02 - samples/sec: 363.13 - lr: 0.000063 - momentum: 0.000000
2023-10-14 16:32:43,802 epoch 7 - iter 1083/3617 - loss 0.01393561 - time (sec): 322.94 - samples/sec: 357.57 - lr: 0.000062 - momentum: 0.000000
2023-10-14 16:34:23,506 epoch 7 - iter 1444/3617 - loss 0.01434256 - time (sec): 422.65 - samples/sec: 361.08 - lr: 0.000060 - momentum: 0.000000
2023-10-14 16:36:06,389 epoch 7 - iter 1805/3617 - loss 0.01398391 - time (sec): 525.53 - samples/sec: 361.44 - lr: 0.000058 - momentum: 0.000000
2023-10-14 16:37:52,259 epoch 7 - iter 2166/3617 - loss 0.01422903 - time (sec): 631.40 - samples/sec: 360.88 - lr: 0.000057 - momentum: 0.000000
2023-10-14 16:39:35,485 epoch 7 - iter 2527/3617 - loss 0.01414726 - time (sec): 734.63 - samples/sec: 362.33 - lr: 0.000055 - momentum: 0.000000
2023-10-14 16:41:19,107 epoch 7 - iter 2888/3617 - loss 0.01450497 - time (sec): 838.25 - samples/sec: 364.58 - lr: 0.000053 - momentum: 0.000000
2023-10-14 16:43:02,248 epoch 7 - iter 3249/3617 - loss 0.01550406 - time (sec): 941.39 - samples/sec: 364.18 - lr: 0.000052 - momentum: 0.000000
2023-10-14 16:44:48,987 epoch 7 - iter 3610/3617 - loss 0.01571456 - time (sec): 1048.13 - samples/sec: 361.60 - lr: 0.000050 - momentum: 0.000000
2023-10-14 16:44:51,074 ----------------------------------------------------------------------------------------------------
2023-10-14 16:44:51,074 EPOCH 7 done: loss 0.0158 - lr: 0.000050
2023-10-14 16:45:30,018 DEV : loss 0.3257623016834259 - f1-score (micro avg) 0.6263
2023-10-14 16:45:30,075 ----------------------------------------------------------------------------------------------------
2023-10-14 16:47:09,244 epoch 8 - iter 361/3617 - loss 0.01142318 - time (sec): 99.17 - samples/sec: 387.29 - lr: 0.000048 - momentum: 0.000000
2023-10-14 16:48:57,539 epoch 8 - iter 722/3617 - loss 0.01106006 - time (sec): 207.46 - samples/sec: 374.10 - lr: 0.000047 - momentum: 0.000000
2023-10-14 16:50:47,572 epoch 8 - iter 1083/3617 - loss 0.01148474 - time (sec): 317.49 - samples/sec: 363.91 - lr: 0.000045 - momentum: 0.000000
2023-10-14 16:52:29,277 epoch 8 - iter 1444/3617 - loss 0.01109820 - time (sec): 419.20 - samples/sec: 363.08 - lr: 0.000043 - momentum: 0.000000
2023-10-14 16:54:07,215 epoch 8 - iter 1805/3617 - loss 0.01060742 - time (sec): 517.14 - samples/sec: 368.74 - lr: 0.000042 - momentum: 0.000000
2023-10-14 16:55:45,728 epoch 8 - iter 2166/3617 - loss 0.01013602 - time (sec): 615.65 - samples/sec: 368.66 - lr: 0.000040 - momentum: 0.000000
2023-10-14 16:57:28,600 epoch 8 - iter 2527/3617 - loss 0.01003184 - time (sec): 718.52 - samples/sec: 370.30 - lr: 0.000038 - momentum: 0.000000
2023-10-14 16:59:08,971 epoch 8 - iter 2888/3617 - loss 0.01030870 - time (sec): 818.89 - samples/sec: 370.18 - lr: 0.000037 - momentum: 0.000000
2023-10-14 17:00:48,363 epoch 8 - iter 3249/3617 - loss 0.01013006 - time (sec): 918.29 - samples/sec: 371.85 - lr: 0.000035 - momentum: 0.000000
2023-10-14 17:02:27,074 epoch 8 - iter 3610/3617 - loss 0.00973135 - time (sec): 1017.00 - samples/sec: 372.91 - lr: 0.000033 - momentum: 0.000000
2023-10-14 17:02:28,759 ----------------------------------------------------------------------------------------------------
2023-10-14 17:02:28,759 EPOCH 8 done: loss 0.0097 - lr: 0.000033
2023-10-14 17:03:08,171 DEV : loss 0.3519401252269745 - f1-score (micro avg) 0.6383
2023-10-14 17:03:08,238 ----------------------------------------------------------------------------------------------------
2023-10-14 17:04:56,121 epoch 9 - iter 361/3617 - loss 0.00400351 - time (sec): 107.88 - samples/sec: 337.43 - lr: 0.000032 - momentum: 0.000000
2023-10-14 17:06:44,549 epoch 9 - iter 722/3617 - loss 0.00553294 - time (sec): 216.31 - samples/sec: 337.08 - lr: 0.000030 - momentum: 0.000000
2023-10-14 17:08:26,029 epoch 9 - iter 1083/3617 - loss 0.00718701 - time (sec): 317.79 - samples/sec: 350.01 - lr: 0.000028 - momentum: 0.000000
2023-10-14 17:10:14,915 epoch 9 - iter 1444/3617 - loss 0.00747014 - time (sec): 426.67 - samples/sec: 351.11 - lr: 0.000027 - momentum: 0.000000
2023-10-14 17:12:13,127 epoch 9 - iter 1805/3617 - loss 0.00700743 - time (sec): 544.89 - samples/sec: 346.36 - lr: 0.000025 - momentum: 0.000000
2023-10-14 17:13:56,710 epoch 9 - iter 2166/3617 - loss 0.00753126 - time (sec): 648.47 - samples/sec: 348.52 - lr: 0.000023 - momentum: 0.000000
2023-10-14 17:15:35,725 epoch 9 - iter 2527/3617 - loss 0.00713809 - time (sec): 747.48 - samples/sec: 353.09 - lr: 0.000022 - momentum: 0.000000
2023-10-14 17:17:15,835 epoch 9 - iter 2888/3617 - loss 0.00710439 - time (sec): 847.60 - samples/sec: 358.03 - lr: 0.000020 - momentum: 0.000000
2023-10-14 17:18:53,848 epoch 9 - iter 3249/3617 - loss 0.00691053 - time (sec): 945.61 - samples/sec: 361.20 - lr: 0.000018 - momentum: 0.000000
2023-10-14 17:20:32,410 epoch 9 - iter 3610/3617 - loss 0.00723383 - time (sec): 1044.17 - samples/sec: 363.12 - lr: 0.000017 - momentum: 0.000000
2023-10-14 17:20:34,181 ----------------------------------------------------------------------------------------------------
2023-10-14 17:20:34,181 EPOCH 9 done: loss 0.0072 - lr: 0.000017
2023-10-14 17:21:13,627 DEV : loss 0.37418004870414734 - f1-score (micro avg) 0.6425
2023-10-14 17:21:13,685 ----------------------------------------------------------------------------------------------------
2023-10-14 17:22:51,217 epoch 10 - iter 361/3617 - loss 0.00288085 - time (sec): 97.53 - samples/sec: 383.84 - lr: 0.000015 - momentum: 0.000000
2023-10-14 17:24:32,451 epoch 10 - iter 722/3617 - loss 0.00430136 - time (sec): 198.76 - samples/sec: 381.28 - lr: 0.000013 - momentum: 0.000000
2023-10-14 17:26:18,156 epoch 10 - iter 1083/3617 - loss 0.00516123 - time (sec): 304.47 - samples/sec: 376.22 - lr: 0.000012 - momentum: 0.000000
2023-10-14 17:27:59,531 epoch 10 - iter 1444/3617 - loss 0.00455380 - time (sec): 405.84 - samples/sec: 374.69 - lr: 0.000010 - momentum: 0.000000
2023-10-14 17:29:41,578 epoch 10 - iter 1805/3617 - loss 0.00417121 - time (sec): 507.89 - samples/sec: 373.65 - lr: 0.000008 - momentum: 0.000000
2023-10-14 17:31:24,980 epoch 10 - iter 2166/3617 - loss 0.00427900 - time (sec): 611.29 - samples/sec: 373.25 - lr: 0.000007 - momentum: 0.000000
2023-10-14 17:33:04,407 epoch 10 - iter 2527/3617 - loss 0.00423939 - time (sec): 710.72 - samples/sec: 374.13 - lr: 0.000005 - momentum: 0.000000
2023-10-14 17:34:44,494 epoch 10 - iter 2888/3617 - loss 0.00423096 - time (sec): 810.81 - samples/sec: 376.32 - lr: 0.000003 - momentum: 0.000000
2023-10-14 17:36:23,538 epoch 10 - iter 3249/3617 - loss 0.00456365 - time (sec): 909.85 - samples/sec: 376.25 - lr: 0.000002 - momentum: 0.000000
2023-10-14 17:38:04,514 epoch 10 - iter 3610/3617 - loss 0.00445018 - time (sec): 1010.83 - samples/sec: 375.28 - lr: 0.000000 - momentum: 0.000000
2023-10-14 17:38:06,359 ----------------------------------------------------------------------------------------------------
2023-10-14 17:38:06,359 EPOCH 10 done: loss 0.0045 - lr: 0.000000
2023-10-14 17:38:48,489 DEV : loss 0.383007675409317 - f1-score (micro avg) 0.6403
2023-10-14 17:38:49,482 ----------------------------------------------------------------------------------------------------
2023-10-14 17:38:49,484 Loading model from best epoch ...
2023-10-14 17:38:53,352 SequenceTagger predicts: Dictionary with 13 tags: O, S-loc, B-loc, E-loc, I-loc, S-pers, B-pers, E-pers, I-pers, S-org, B-org, E-org, I-org
2023-10-14 17:39:53,980
Results:
- F-score (micro) 0.6356
- F-score (macro) 0.4981
- Accuracy 0.4788
By class:
precision recall f1-score support
loc 0.6276 0.7699 0.6915 591
pers 0.5664 0.7171 0.6329 357
org 0.1757 0.1646 0.1699 79
micro avg 0.5787 0.7050 0.6356 1027
macro avg 0.4565 0.5505 0.4981 1027
weighted avg 0.5715 0.7050 0.6310 1027
2023-10-14 17:39:53,980 ----------------------------------------------------------------------------------------------------
|