File size: 25,279 Bytes
5708ee7 |
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 12:43:36,931 ----------------------------------------------------------------------------------------------------
2023-10-12 12:43:36,933 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 12:43:36,933 ----------------------------------------------------------------------------------------------------
2023-10-12 12:43:36,933 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 12:43:36,933 ----------------------------------------------------------------------------------------------------
2023-10-12 12:43:36,933 Train: 7936 sentences
2023-10-12 12:43:36,933 (train_with_dev=False, train_with_test=False)
2023-10-12 12:43:36,934 ----------------------------------------------------------------------------------------------------
2023-10-12 12:43:36,934 Training Params:
2023-10-12 12:43:36,934 - learning_rate: "0.00016"
2023-10-12 12:43:36,934 - mini_batch_size: "4"
2023-10-12 12:43:36,934 - max_epochs: "10"
2023-10-12 12:43:36,934 - shuffle: "True"
2023-10-12 12:43:36,934 ----------------------------------------------------------------------------------------------------
2023-10-12 12:43:36,934 Plugins:
2023-10-12 12:43:36,934 - TensorboardLogger
2023-10-12 12:43:36,934 - LinearScheduler | warmup_fraction: '0.1'
2023-10-12 12:43:36,934 ----------------------------------------------------------------------------------------------------
2023-10-12 12:43:36,934 Final evaluation on model from best epoch (best-model.pt)
2023-10-12 12:43:36,934 - metric: "('micro avg', 'f1-score')"
2023-10-12 12:43:36,934 ----------------------------------------------------------------------------------------------------
2023-10-12 12:43:36,934 Computation:
2023-10-12 12:43:36,935 - compute on device: cuda:0
2023-10-12 12:43:36,935 - embedding storage: none
2023-10-12 12:43:36,935 ----------------------------------------------------------------------------------------------------
2023-10-12 12:43:36,935 Model training base path: "hmbench-icdar/fr-hmbyt5-preliminary/byt5-small-historic-multilingual-span20-flax-bs4-wsFalse-e10-lr0.00016-poolingfirst-layers-1-crfFalse-1"
2023-10-12 12:43:36,935 ----------------------------------------------------------------------------------------------------
2023-10-12 12:43:36,935 ----------------------------------------------------------------------------------------------------
2023-10-12 12:43:36,935 Logging anything other than scalars to TensorBoard is currently not supported.
2023-10-12 12:44:29,444 epoch 1 - iter 198/1984 - loss 2.57872388 - time (sec): 52.51 - samples/sec: 294.57 - lr: 0.000016 - momentum: 0.000000
2023-10-12 12:45:23,159 epoch 1 - iter 396/1984 - loss 2.42977020 - time (sec): 106.22 - samples/sec: 296.79 - lr: 0.000032 - momentum: 0.000000
2023-10-12 12:46:17,601 epoch 1 - iter 594/1984 - loss 2.09651161 - time (sec): 160.66 - samples/sec: 298.18 - lr: 0.000048 - momentum: 0.000000
2023-10-12 12:47:19,176 epoch 1 - iter 792/1984 - loss 1.74658842 - time (sec): 222.24 - samples/sec: 289.43 - lr: 0.000064 - momentum: 0.000000
2023-10-12 12:48:12,650 epoch 1 - iter 990/1984 - loss 1.47898744 - time (sec): 275.71 - samples/sec: 291.16 - lr: 0.000080 - momentum: 0.000000
2023-10-12 12:49:03,780 epoch 1 - iter 1188/1984 - loss 1.27073991 - time (sec): 326.84 - samples/sec: 296.79 - lr: 0.000096 - momentum: 0.000000
2023-10-12 12:49:58,979 epoch 1 - iter 1386/1984 - loss 1.11087775 - time (sec): 382.04 - samples/sec: 299.34 - lr: 0.000112 - momentum: 0.000000
2023-10-12 12:51:00,244 epoch 1 - iter 1584/1984 - loss 0.99932054 - time (sec): 443.31 - samples/sec: 295.65 - lr: 0.000128 - momentum: 0.000000
2023-10-12 12:51:58,406 epoch 1 - iter 1782/1984 - loss 0.91329740 - time (sec): 501.47 - samples/sec: 293.81 - lr: 0.000144 - momentum: 0.000000
2023-10-12 12:52:52,746 epoch 1 - iter 1980/1984 - loss 0.83963990 - time (sec): 555.81 - samples/sec: 294.49 - lr: 0.000160 - momentum: 0.000000
2023-10-12 12:52:53,861 ----------------------------------------------------------------------------------------------------
2023-10-12 12:52:53,862 EPOCH 1 done: loss 0.8384 - lr: 0.000160
2023-10-12 12:53:19,383 DEV : loss 0.14055970311164856 - f1-score (micro avg) 0.6499
2023-10-12 12:53:19,431 saving best model
2023-10-12 12:53:20,360 ----------------------------------------------------------------------------------------------------
2023-10-12 12:54:14,361 epoch 2 - iter 198/1984 - loss 0.16528251 - time (sec): 54.00 - samples/sec: 305.04 - lr: 0.000158 - momentum: 0.000000
2023-10-12 12:55:10,170 epoch 2 - iter 396/1984 - loss 0.14401493 - time (sec): 109.81 - samples/sec: 297.71 - lr: 0.000156 - momentum: 0.000000
2023-10-12 12:56:08,489 epoch 2 - iter 594/1984 - loss 0.13710523 - time (sec): 168.13 - samples/sec: 291.73 - lr: 0.000155 - momentum: 0.000000
2023-10-12 12:57:06,323 epoch 2 - iter 792/1984 - loss 0.13661458 - time (sec): 225.96 - samples/sec: 290.86 - lr: 0.000153 - momentum: 0.000000
2023-10-12 12:58:01,146 epoch 2 - iter 990/1984 - loss 0.13287502 - time (sec): 280.78 - samples/sec: 293.36 - lr: 0.000151 - momentum: 0.000000
2023-10-12 12:58:55,564 epoch 2 - iter 1188/1984 - loss 0.13028398 - time (sec): 335.20 - samples/sec: 293.98 - lr: 0.000149 - momentum: 0.000000
2023-10-12 12:59:49,914 epoch 2 - iter 1386/1984 - loss 0.13061150 - time (sec): 389.55 - samples/sec: 295.42 - lr: 0.000148 - momentum: 0.000000
2023-10-12 13:00:43,318 epoch 2 - iter 1584/1984 - loss 0.12716756 - time (sec): 442.96 - samples/sec: 295.30 - lr: 0.000146 - momentum: 0.000000
2023-10-12 13:01:41,149 epoch 2 - iter 1782/1984 - loss 0.12425762 - time (sec): 500.79 - samples/sec: 293.51 - lr: 0.000144 - momentum: 0.000000
2023-10-12 13:02:40,598 epoch 2 - iter 1980/1984 - loss 0.12247253 - time (sec): 560.24 - samples/sec: 291.85 - lr: 0.000142 - momentum: 0.000000
2023-10-12 13:02:41,892 ----------------------------------------------------------------------------------------------------
2023-10-12 13:02:41,893 EPOCH 2 done: loss 0.1223 - lr: 0.000142
2023-10-12 13:03:09,818 DEV : loss 0.0859101414680481 - f1-score (micro avg) 0.7275
2023-10-12 13:03:09,869 saving best model
2023-10-12 13:03:10,961 ----------------------------------------------------------------------------------------------------
2023-10-12 13:04:04,131 epoch 3 - iter 198/1984 - loss 0.07669844 - time (sec): 53.17 - samples/sec: 296.03 - lr: 0.000140 - momentum: 0.000000
2023-10-12 13:04:57,976 epoch 3 - iter 396/1984 - loss 0.07779340 - time (sec): 107.01 - samples/sec: 298.62 - lr: 0.000139 - momentum: 0.000000
2023-10-12 13:05:53,748 epoch 3 - iter 594/1984 - loss 0.07883378 - time (sec): 162.78 - samples/sec: 299.80 - lr: 0.000137 - momentum: 0.000000
2023-10-12 13:06:53,717 epoch 3 - iter 792/1984 - loss 0.07827758 - time (sec): 222.75 - samples/sec: 292.63 - lr: 0.000135 - momentum: 0.000000
2023-10-12 13:07:52,154 epoch 3 - iter 990/1984 - loss 0.07681269 - time (sec): 281.19 - samples/sec: 289.45 - lr: 0.000133 - momentum: 0.000000
2023-10-12 13:08:51,767 epoch 3 - iter 1188/1984 - loss 0.07663353 - time (sec): 340.80 - samples/sec: 285.97 - lr: 0.000132 - momentum: 0.000000
2023-10-12 13:09:47,186 epoch 3 - iter 1386/1984 - loss 0.07734261 - time (sec): 396.22 - samples/sec: 285.72 - lr: 0.000130 - momentum: 0.000000
2023-10-12 13:10:44,118 epoch 3 - iter 1584/1984 - loss 0.07604814 - time (sec): 453.15 - samples/sec: 288.78 - lr: 0.000128 - momentum: 0.000000
2023-10-12 13:11:40,292 epoch 3 - iter 1782/1984 - loss 0.07555546 - time (sec): 509.33 - samples/sec: 290.07 - lr: 0.000126 - momentum: 0.000000
2023-10-12 13:12:38,011 epoch 3 - iter 1980/1984 - loss 0.07603113 - time (sec): 567.05 - samples/sec: 288.67 - lr: 0.000125 - momentum: 0.000000
2023-10-12 13:12:39,271 ----------------------------------------------------------------------------------------------------
2023-10-12 13:12:39,271 EPOCH 3 done: loss 0.0760 - lr: 0.000125
2023-10-12 13:13:08,663 DEV : loss 0.10180744528770447 - f1-score (micro avg) 0.7625
2023-10-12 13:13:08,710 saving best model
2023-10-12 13:13:11,483 ----------------------------------------------------------------------------------------------------
2023-10-12 13:14:09,654 epoch 4 - iter 198/1984 - loss 0.04829137 - time (sec): 58.16 - samples/sec: 294.70 - lr: 0.000123 - momentum: 0.000000
2023-10-12 13:15:02,659 epoch 4 - iter 396/1984 - loss 0.05173970 - time (sec): 111.17 - samples/sec: 306.37 - lr: 0.000121 - momentum: 0.000000
2023-10-12 13:15:57,996 epoch 4 - iter 594/1984 - loss 0.05651052 - time (sec): 166.51 - samples/sec: 300.16 - lr: 0.000119 - momentum: 0.000000
2023-10-12 13:16:53,481 epoch 4 - iter 792/1984 - loss 0.05517049 - time (sec): 221.99 - samples/sec: 296.01 - lr: 0.000117 - momentum: 0.000000
2023-10-12 13:17:52,233 epoch 4 - iter 990/1984 - loss 0.05634728 - time (sec): 280.74 - samples/sec: 291.99 - lr: 0.000116 - momentum: 0.000000
2023-10-12 13:18:49,209 epoch 4 - iter 1188/1984 - loss 0.05607999 - time (sec): 337.72 - samples/sec: 290.31 - lr: 0.000114 - momentum: 0.000000
2023-10-12 13:19:46,175 epoch 4 - iter 1386/1984 - loss 0.05551788 - time (sec): 394.69 - samples/sec: 290.32 - lr: 0.000112 - momentum: 0.000000
2023-10-12 13:20:44,641 epoch 4 - iter 1584/1984 - loss 0.05700501 - time (sec): 453.15 - samples/sec: 288.52 - lr: 0.000110 - momentum: 0.000000
2023-10-12 13:21:46,416 epoch 4 - iter 1782/1984 - loss 0.05565357 - time (sec): 514.93 - samples/sec: 286.92 - lr: 0.000109 - momentum: 0.000000
2023-10-12 13:22:42,025 epoch 4 - iter 1980/1984 - loss 0.05607850 - time (sec): 570.53 - samples/sec: 287.00 - lr: 0.000107 - momentum: 0.000000
2023-10-12 13:22:43,130 ----------------------------------------------------------------------------------------------------
2023-10-12 13:22:43,130 EPOCH 4 done: loss 0.0560 - lr: 0.000107
2023-10-12 13:23:11,653 DEV : loss 0.12707574665546417 - f1-score (micro avg) 0.7592
2023-10-12 13:23:11,704 ----------------------------------------------------------------------------------------------------
2023-10-12 13:24:12,175 epoch 5 - iter 198/1984 - loss 0.04138412 - time (sec): 60.47 - samples/sec: 267.17 - lr: 0.000105 - momentum: 0.000000
2023-10-12 13:25:08,068 epoch 5 - iter 396/1984 - loss 0.03535050 - time (sec): 116.36 - samples/sec: 277.92 - lr: 0.000103 - momentum: 0.000000
2023-10-12 13:26:04,766 epoch 5 - iter 594/1984 - loss 0.03648738 - time (sec): 173.06 - samples/sec: 281.12 - lr: 0.000101 - momentum: 0.000000
2023-10-12 13:27:01,501 epoch 5 - iter 792/1984 - loss 0.03686215 - time (sec): 229.79 - samples/sec: 283.29 - lr: 0.000100 - momentum: 0.000000
2023-10-12 13:27:53,816 epoch 5 - iter 990/1984 - loss 0.03625039 - time (sec): 282.11 - samples/sec: 288.19 - lr: 0.000098 - momentum: 0.000000
2023-10-12 13:28:46,678 epoch 5 - iter 1188/1984 - loss 0.03960505 - time (sec): 334.97 - samples/sec: 292.34 - lr: 0.000096 - momentum: 0.000000
2023-10-12 13:29:40,681 epoch 5 - iter 1386/1984 - loss 0.04001738 - time (sec): 388.97 - samples/sec: 294.20 - lr: 0.000094 - momentum: 0.000000
2023-10-12 13:30:35,002 epoch 5 - iter 1584/1984 - loss 0.04082094 - time (sec): 443.29 - samples/sec: 296.26 - lr: 0.000093 - momentum: 0.000000
2023-10-12 13:31:28,676 epoch 5 - iter 1782/1984 - loss 0.04148610 - time (sec): 496.97 - samples/sec: 297.49 - lr: 0.000091 - momentum: 0.000000
2023-10-12 13:32:21,692 epoch 5 - iter 1980/1984 - loss 0.04251132 - time (sec): 549.99 - samples/sec: 297.50 - lr: 0.000089 - momentum: 0.000000
2023-10-12 13:32:22,804 ----------------------------------------------------------------------------------------------------
2023-10-12 13:32:22,804 EPOCH 5 done: loss 0.0425 - lr: 0.000089
2023-10-12 13:32:48,621 DEV : loss 0.14472655951976776 - f1-score (micro avg) 0.7461
2023-10-12 13:32:48,680 ----------------------------------------------------------------------------------------------------
2023-10-12 13:33:42,335 epoch 6 - iter 198/1984 - loss 0.02627233 - time (sec): 53.65 - samples/sec: 291.97 - lr: 0.000087 - momentum: 0.000000
2023-10-12 13:34:34,598 epoch 6 - iter 396/1984 - loss 0.02760424 - time (sec): 105.92 - samples/sec: 302.33 - lr: 0.000085 - momentum: 0.000000
2023-10-12 13:35:26,843 epoch 6 - iter 594/1984 - loss 0.02865242 - time (sec): 158.16 - samples/sec: 304.60 - lr: 0.000084 - momentum: 0.000000
2023-10-12 13:36:21,745 epoch 6 - iter 792/1984 - loss 0.02818054 - time (sec): 213.06 - samples/sec: 305.87 - lr: 0.000082 - momentum: 0.000000
2023-10-12 13:37:16,009 epoch 6 - iter 990/1984 - loss 0.02691623 - time (sec): 267.33 - samples/sec: 303.51 - lr: 0.000080 - momentum: 0.000000
2023-10-12 13:38:10,624 epoch 6 - iter 1188/1984 - loss 0.02670693 - time (sec): 321.94 - samples/sec: 303.88 - lr: 0.000078 - momentum: 0.000000
2023-10-12 13:39:05,104 epoch 6 - iter 1386/1984 - loss 0.02728730 - time (sec): 376.42 - samples/sec: 304.70 - lr: 0.000077 - momentum: 0.000000
2023-10-12 13:40:00,408 epoch 6 - iter 1584/1984 - loss 0.02914735 - time (sec): 431.73 - samples/sec: 302.80 - lr: 0.000075 - momentum: 0.000000
2023-10-12 13:40:54,644 epoch 6 - iter 1782/1984 - loss 0.02922904 - time (sec): 485.96 - samples/sec: 303.18 - lr: 0.000073 - momentum: 0.000000
2023-10-12 13:41:50,958 epoch 6 - iter 1980/1984 - loss 0.02961781 - time (sec): 542.28 - samples/sec: 301.72 - lr: 0.000071 - momentum: 0.000000
2023-10-12 13:41:52,135 ----------------------------------------------------------------------------------------------------
2023-10-12 13:41:52,136 EPOCH 6 done: loss 0.0297 - lr: 0.000071
2023-10-12 13:42:19,073 DEV : loss 0.17014847695827484 - f1-score (micro avg) 0.7612
2023-10-12 13:42:19,119 ----------------------------------------------------------------------------------------------------
2023-10-12 13:43:11,962 epoch 7 - iter 198/1984 - loss 0.01757743 - time (sec): 52.84 - samples/sec: 308.61 - lr: 0.000069 - momentum: 0.000000
2023-10-12 13:44:03,780 epoch 7 - iter 396/1984 - loss 0.02000202 - time (sec): 104.66 - samples/sec: 315.03 - lr: 0.000068 - momentum: 0.000000
2023-10-12 13:44:57,934 epoch 7 - iter 594/1984 - loss 0.01900677 - time (sec): 158.81 - samples/sec: 308.16 - lr: 0.000066 - momentum: 0.000000
2023-10-12 13:45:54,285 epoch 7 - iter 792/1984 - loss 0.02132227 - time (sec): 215.16 - samples/sec: 305.10 - lr: 0.000064 - momentum: 0.000000
2023-10-12 13:46:51,268 epoch 7 - iter 990/1984 - loss 0.02065637 - time (sec): 272.15 - samples/sec: 299.69 - lr: 0.000062 - momentum: 0.000000
2023-10-12 13:47:50,086 epoch 7 - iter 1188/1984 - loss 0.02083419 - time (sec): 330.96 - samples/sec: 295.66 - lr: 0.000061 - momentum: 0.000000
2023-10-12 13:48:47,692 epoch 7 - iter 1386/1984 - loss 0.02064694 - time (sec): 388.57 - samples/sec: 295.32 - lr: 0.000059 - momentum: 0.000000
2023-10-12 13:49:41,352 epoch 7 - iter 1584/1984 - loss 0.02141756 - time (sec): 442.23 - samples/sec: 293.14 - lr: 0.000057 - momentum: 0.000000
2023-10-12 13:50:36,232 epoch 7 - iter 1782/1984 - loss 0.02092148 - time (sec): 497.11 - samples/sec: 294.32 - lr: 0.000055 - momentum: 0.000000
2023-10-12 13:51:31,622 epoch 7 - iter 1980/1984 - loss 0.02052943 - time (sec): 552.50 - samples/sec: 296.11 - lr: 0.000053 - momentum: 0.000000
2023-10-12 13:51:32,742 ----------------------------------------------------------------------------------------------------
2023-10-12 13:51:32,742 EPOCH 7 done: loss 0.0205 - lr: 0.000053
2023-10-12 13:51:59,673 DEV : loss 0.2085985541343689 - f1-score (micro avg) 0.7496
2023-10-12 13:51:59,718 ----------------------------------------------------------------------------------------------------
2023-10-12 13:52:54,707 epoch 8 - iter 198/1984 - loss 0.01667937 - time (sec): 54.99 - samples/sec: 307.84 - lr: 0.000052 - momentum: 0.000000
2023-10-12 13:53:50,334 epoch 8 - iter 396/1984 - loss 0.01438657 - time (sec): 110.61 - samples/sec: 291.56 - lr: 0.000050 - momentum: 0.000000
2023-10-12 13:54:45,708 epoch 8 - iter 594/1984 - loss 0.01421150 - time (sec): 165.99 - samples/sec: 287.68 - lr: 0.000048 - momentum: 0.000000
2023-10-12 13:55:41,536 epoch 8 - iter 792/1984 - loss 0.01432064 - time (sec): 221.82 - samples/sec: 287.50 - lr: 0.000046 - momentum: 0.000000
2023-10-12 13:56:38,668 epoch 8 - iter 990/1984 - loss 0.01369389 - time (sec): 278.95 - samples/sec: 288.69 - lr: 0.000045 - momentum: 0.000000
2023-10-12 13:57:37,033 epoch 8 - iter 1188/1984 - loss 0.01537915 - time (sec): 337.31 - samples/sec: 290.16 - lr: 0.000043 - momentum: 0.000000
2023-10-12 13:58:34,687 epoch 8 - iter 1386/1984 - loss 0.01524152 - time (sec): 394.97 - samples/sec: 288.48 - lr: 0.000041 - momentum: 0.000000
2023-10-12 13:59:30,605 epoch 8 - iter 1584/1984 - loss 0.01517160 - time (sec): 450.88 - samples/sec: 289.72 - lr: 0.000039 - momentum: 0.000000
2023-10-12 14:00:28,078 epoch 8 - iter 1782/1984 - loss 0.01553226 - time (sec): 508.36 - samples/sec: 288.49 - lr: 0.000037 - momentum: 0.000000
2023-10-12 14:01:27,127 epoch 8 - iter 1980/1984 - loss 0.01498313 - time (sec): 567.41 - samples/sec: 288.59 - lr: 0.000036 - momentum: 0.000000
2023-10-12 14:01:28,169 ----------------------------------------------------------------------------------------------------
2023-10-12 14:01:28,170 EPOCH 8 done: loss 0.0150 - lr: 0.000036
2023-10-12 14:01:58,438 DEV : loss 0.21621058881282806 - f1-score (micro avg) 0.757
2023-10-12 14:01:58,492 ----------------------------------------------------------------------------------------------------
2023-10-12 14:02:56,912 epoch 9 - iter 198/1984 - loss 0.01321939 - time (sec): 58.42 - samples/sec: 294.64 - lr: 0.000034 - momentum: 0.000000
2023-10-12 14:03:55,154 epoch 9 - iter 396/1984 - loss 0.01312927 - time (sec): 116.66 - samples/sec: 288.45 - lr: 0.000032 - momentum: 0.000000
2023-10-12 14:04:49,193 epoch 9 - iter 594/1984 - loss 0.01098511 - time (sec): 170.70 - samples/sec: 296.55 - lr: 0.000030 - momentum: 0.000000
2023-10-12 14:05:43,288 epoch 9 - iter 792/1984 - loss 0.01215431 - time (sec): 224.79 - samples/sec: 294.40 - lr: 0.000029 - momentum: 0.000000
2023-10-12 14:06:43,526 epoch 9 - iter 990/1984 - loss 0.01151383 - time (sec): 285.03 - samples/sec: 290.06 - lr: 0.000027 - momentum: 0.000000
2023-10-12 14:07:40,132 epoch 9 - iter 1188/1984 - loss 0.01094490 - time (sec): 341.64 - samples/sec: 289.99 - lr: 0.000025 - momentum: 0.000000
2023-10-12 14:08:35,756 epoch 9 - iter 1386/1984 - loss 0.01030513 - time (sec): 397.26 - samples/sec: 291.31 - lr: 0.000023 - momentum: 0.000000
2023-10-12 14:09:31,788 epoch 9 - iter 1584/1984 - loss 0.01081881 - time (sec): 453.29 - samples/sec: 289.50 - lr: 0.000021 - momentum: 0.000000
2023-10-12 14:10:30,471 epoch 9 - iter 1782/1984 - loss 0.01121344 - time (sec): 511.98 - samples/sec: 287.80 - lr: 0.000020 - momentum: 0.000000
2023-10-12 14:11:25,942 epoch 9 - iter 1980/1984 - loss 0.01074014 - time (sec): 567.45 - samples/sec: 288.38 - lr: 0.000018 - momentum: 0.000000
2023-10-12 14:11:27,169 ----------------------------------------------------------------------------------------------------
2023-10-12 14:11:27,169 EPOCH 9 done: loss 0.0107 - lr: 0.000018
2023-10-12 14:11:55,640 DEV : loss 0.22689764201641083 - f1-score (micro avg) 0.7614
2023-10-12 14:11:55,690 ----------------------------------------------------------------------------------------------------
2023-10-12 14:12:53,298 epoch 10 - iter 198/1984 - loss 0.00665429 - time (sec): 57.61 - samples/sec: 289.71 - lr: 0.000016 - momentum: 0.000000
2023-10-12 14:13:51,316 epoch 10 - iter 396/1984 - loss 0.00646911 - time (sec): 115.62 - samples/sec: 283.40 - lr: 0.000014 - momentum: 0.000000
2023-10-12 14:14:49,815 epoch 10 - iter 594/1984 - loss 0.00659430 - time (sec): 174.12 - samples/sec: 283.69 - lr: 0.000013 - momentum: 0.000000
2023-10-12 14:15:46,196 epoch 10 - iter 792/1984 - loss 0.00696637 - time (sec): 230.50 - samples/sec: 286.08 - lr: 0.000011 - momentum: 0.000000
2023-10-12 14:16:42,436 epoch 10 - iter 990/1984 - loss 0.00672449 - time (sec): 286.74 - samples/sec: 287.82 - lr: 0.000009 - momentum: 0.000000
2023-10-12 14:17:38,147 epoch 10 - iter 1188/1984 - loss 0.00643333 - time (sec): 342.45 - samples/sec: 287.35 - lr: 0.000007 - momentum: 0.000000
2023-10-12 14:18:34,399 epoch 10 - iter 1386/1984 - loss 0.00719725 - time (sec): 398.71 - samples/sec: 286.32 - lr: 0.000005 - momentum: 0.000000
2023-10-12 14:19:30,932 epoch 10 - iter 1584/1984 - loss 0.00759850 - time (sec): 455.24 - samples/sec: 286.89 - lr: 0.000004 - momentum: 0.000000
2023-10-12 14:20:26,329 epoch 10 - iter 1782/1984 - loss 0.00756874 - time (sec): 510.64 - samples/sec: 288.34 - lr: 0.000002 - momentum: 0.000000
2023-10-12 14:21:23,506 epoch 10 - iter 1980/1984 - loss 0.00842761 - time (sec): 567.81 - samples/sec: 288.42 - lr: 0.000000 - momentum: 0.000000
2023-10-12 14:21:24,509 ----------------------------------------------------------------------------------------------------
2023-10-12 14:21:24,509 EPOCH 10 done: loss 0.0085 - lr: 0.000000
2023-10-12 14:21:51,035 DEV : loss 0.2328162044286728 - f1-score (micro avg) 0.7578
2023-10-12 14:21:52,053 ----------------------------------------------------------------------------------------------------
2023-10-12 14:21:52,056 Loading model from best epoch ...
2023-10-12 14:21:56,378 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 14:22:22,594
Results:
- F-score (micro) 0.7368
- F-score (macro) 0.6608
- Accuracy 0.6204
By class:
precision recall f1-score support
LOC 0.8380 0.7817 0.8088 655
PER 0.6341 0.8161 0.7137 223
ORG 0.4125 0.5197 0.4599 127
micro avg 0.7183 0.7562 0.7368 1005
macro avg 0.6282 0.7058 0.6608 1005
weighted avg 0.7390 0.7562 0.7436 1005
2023-10-12 14:22:22,594 ----------------------------------------------------------------------------------------------------
|