File size: 25,320 Bytes
0a8e007 |
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 |
2023-10-12 17:24:51,784 ----------------------------------------------------------------------------------------------------
2023-10-12 17:24:51,787 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 17:24:51,787 ----------------------------------------------------------------------------------------------------
2023-10-12 17:24:51,787 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 17:24:51,787 ----------------------------------------------------------------------------------------------------
2023-10-12 17:24:51,787 Train: 7936 sentences
2023-10-12 17:24:51,787 (train_with_dev=False, train_with_test=False)
2023-10-12 17:24:51,787 ----------------------------------------------------------------------------------------------------
2023-10-12 17:24:51,787 Training Params:
2023-10-12 17:24:51,787 - learning_rate: "0.00015"
2023-10-12 17:24:51,787 - mini_batch_size: "4"
2023-10-12 17:24:51,788 - max_epochs: "10"
2023-10-12 17:24:51,788 - shuffle: "True"
2023-10-12 17:24:51,788 ----------------------------------------------------------------------------------------------------
2023-10-12 17:24:51,788 Plugins:
2023-10-12 17:24:51,788 - TensorboardLogger
2023-10-12 17:24:51,788 - LinearScheduler | warmup_fraction: '0.1'
2023-10-12 17:24:51,788 ----------------------------------------------------------------------------------------------------
2023-10-12 17:24:51,788 Final evaluation on model from best epoch (best-model.pt)
2023-10-12 17:24:51,788 - metric: "('micro avg', 'f1-score')"
2023-10-12 17:24:51,788 ----------------------------------------------------------------------------------------------------
2023-10-12 17:24:51,788 Computation:
2023-10-12 17:24:51,788 - compute on device: cuda:0
2023-10-12 17:24:51,788 - embedding storage: none
2023-10-12 17:24:51,788 ----------------------------------------------------------------------------------------------------
2023-10-12 17:24:51,788 Model training base path: "hmbench-icdar/fr-hmbyt5-preliminary/byt5-small-historic-multilingual-span20-flax-bs4-wsFalse-e10-lr0.00015-poolingfirst-layers-1-crfFalse-2"
2023-10-12 17:24:51,789 ----------------------------------------------------------------------------------------------------
2023-10-12 17:24:51,789 ----------------------------------------------------------------------------------------------------
2023-10-12 17:24:51,789 Logging anything other than scalars to TensorBoard is currently not supported.
2023-10-12 17:25:42,845 epoch 1 - iter 198/1984 - loss 2.55536644 - time (sec): 51.05 - samples/sec: 322.35 - lr: 0.000015 - momentum: 0.000000
2023-10-12 17:26:33,679 epoch 1 - iter 396/1984 - loss 2.41138788 - time (sec): 101.89 - samples/sec: 306.71 - lr: 0.000030 - momentum: 0.000000
2023-10-12 17:27:25,852 epoch 1 - iter 594/1984 - loss 2.08406574 - time (sec): 154.06 - samples/sec: 311.46 - lr: 0.000045 - momentum: 0.000000
2023-10-12 17:28:17,880 epoch 1 - iter 792/1984 - loss 1.76238592 - time (sec): 206.09 - samples/sec: 309.43 - lr: 0.000060 - momentum: 0.000000
2023-10-12 17:29:09,576 epoch 1 - iter 990/1984 - loss 1.49063858 - time (sec): 257.79 - samples/sec: 311.56 - lr: 0.000075 - momentum: 0.000000
2023-10-12 17:30:03,660 epoch 1 - iter 1188/1984 - loss 1.28530974 - time (sec): 311.87 - samples/sec: 308.92 - lr: 0.000090 - momentum: 0.000000
2023-10-12 17:30:54,807 epoch 1 - iter 1386/1984 - loss 1.13019078 - time (sec): 363.02 - samples/sec: 313.91 - lr: 0.000105 - momentum: 0.000000
2023-10-12 17:31:46,000 epoch 1 - iter 1584/1984 - loss 1.01185667 - time (sec): 414.21 - samples/sec: 316.00 - lr: 0.000120 - momentum: 0.000000
2023-10-12 17:32:36,962 epoch 1 - iter 1782/1984 - loss 0.92502162 - time (sec): 465.17 - samples/sec: 315.93 - lr: 0.000135 - momentum: 0.000000
2023-10-12 17:33:30,270 epoch 1 - iter 1980/1984 - loss 0.84977154 - time (sec): 518.48 - samples/sec: 315.76 - lr: 0.000150 - momentum: 0.000000
2023-10-12 17:33:31,299 ----------------------------------------------------------------------------------------------------
2023-10-12 17:33:31,300 EPOCH 1 done: loss 0.8485 - lr: 0.000150
2023-10-12 17:33:56,893 DEV : loss 0.16075488924980164 - f1-score (micro avg) 0.5739
2023-10-12 17:33:56,945 saving best model
2023-10-12 17:33:57,979 ----------------------------------------------------------------------------------------------------
2023-10-12 17:34:51,494 epoch 2 - iter 198/1984 - loss 0.19639994 - time (sec): 53.51 - samples/sec: 310.19 - lr: 0.000148 - momentum: 0.000000
2023-10-12 17:35:42,055 epoch 2 - iter 396/1984 - loss 0.17680493 - time (sec): 104.07 - samples/sec: 313.66 - lr: 0.000147 - momentum: 0.000000
2023-10-12 17:36:33,064 epoch 2 - iter 594/1984 - loss 0.16946243 - time (sec): 155.08 - samples/sec: 316.75 - lr: 0.000145 - momentum: 0.000000
2023-10-12 17:37:25,539 epoch 2 - iter 792/1984 - loss 0.16300567 - time (sec): 207.56 - samples/sec: 314.80 - lr: 0.000143 - momentum: 0.000000
2023-10-12 17:38:22,072 epoch 2 - iter 990/1984 - loss 0.15336618 - time (sec): 264.09 - samples/sec: 312.60 - lr: 0.000142 - momentum: 0.000000
2023-10-12 17:39:17,838 epoch 2 - iter 1188/1984 - loss 0.14671314 - time (sec): 319.86 - samples/sec: 308.67 - lr: 0.000140 - momentum: 0.000000
2023-10-12 17:40:13,646 epoch 2 - iter 1386/1984 - loss 0.14319007 - time (sec): 375.66 - samples/sec: 304.39 - lr: 0.000138 - momentum: 0.000000
2023-10-12 17:41:10,943 epoch 2 - iter 1584/1984 - loss 0.13786458 - time (sec): 432.96 - samples/sec: 304.59 - lr: 0.000137 - momentum: 0.000000
2023-10-12 17:42:07,865 epoch 2 - iter 1782/1984 - loss 0.13502095 - time (sec): 489.88 - samples/sec: 303.54 - lr: 0.000135 - momentum: 0.000000
2023-10-12 17:43:01,351 epoch 2 - iter 1980/1984 - loss 0.13266822 - time (sec): 543.37 - samples/sec: 301.33 - lr: 0.000133 - momentum: 0.000000
2023-10-12 17:43:02,387 ----------------------------------------------------------------------------------------------------
2023-10-12 17:43:02,387 EPOCH 2 done: loss 0.1326 - lr: 0.000133
2023-10-12 17:43:28,049 DEV : loss 0.08952224254608154 - f1-score (micro avg) 0.7284
2023-10-12 17:43:28,095 saving best model
2023-10-12 17:43:38,922 ----------------------------------------------------------------------------------------------------
2023-10-12 17:44:36,088 epoch 3 - iter 198/1984 - loss 0.08127744 - time (sec): 57.16 - samples/sec: 306.32 - lr: 0.000132 - momentum: 0.000000
2023-10-12 17:45:31,021 epoch 3 - iter 396/1984 - loss 0.08641176 - time (sec): 112.09 - samples/sec: 313.84 - lr: 0.000130 - momentum: 0.000000
2023-10-12 17:46:25,458 epoch 3 - iter 594/1984 - loss 0.08807778 - time (sec): 166.53 - samples/sec: 305.34 - lr: 0.000128 - momentum: 0.000000
2023-10-12 17:47:20,350 epoch 3 - iter 792/1984 - loss 0.08782288 - time (sec): 221.42 - samples/sec: 302.47 - lr: 0.000127 - momentum: 0.000000
2023-10-12 17:48:15,286 epoch 3 - iter 990/1984 - loss 0.08496675 - time (sec): 276.36 - samples/sec: 300.64 - lr: 0.000125 - momentum: 0.000000
2023-10-12 17:49:07,276 epoch 3 - iter 1188/1984 - loss 0.08486362 - time (sec): 328.35 - samples/sec: 301.13 - lr: 0.000123 - momentum: 0.000000
2023-10-12 17:50:03,093 epoch 3 - iter 1386/1984 - loss 0.08414590 - time (sec): 384.16 - samples/sec: 301.44 - lr: 0.000122 - momentum: 0.000000
2023-10-12 17:50:55,630 epoch 3 - iter 1584/1984 - loss 0.08148581 - time (sec): 436.70 - samples/sec: 304.47 - lr: 0.000120 - momentum: 0.000000
2023-10-12 17:51:51,927 epoch 3 - iter 1782/1984 - loss 0.08077785 - time (sec): 493.00 - samples/sec: 301.11 - lr: 0.000118 - momentum: 0.000000
2023-10-12 17:52:47,909 epoch 3 - iter 1980/1984 - loss 0.08075089 - time (sec): 548.98 - samples/sec: 298.09 - lr: 0.000117 - momentum: 0.000000
2023-10-12 17:52:49,212 ----------------------------------------------------------------------------------------------------
2023-10-12 17:52:49,213 EPOCH 3 done: loss 0.0808 - lr: 0.000117
2023-10-12 17:53:19,578 DEV : loss 0.10335122048854828 - f1-score (micro avg) 0.7415
2023-10-12 17:53:19,624 saving best model
2023-10-12 17:53:22,236 ----------------------------------------------------------------------------------------------------
2023-10-12 17:54:20,197 epoch 4 - iter 198/1984 - loss 0.04733732 - time (sec): 57.95 - samples/sec: 283.02 - lr: 0.000115 - momentum: 0.000000
2023-10-12 17:55:13,014 epoch 4 - iter 396/1984 - loss 0.05655464 - time (sec): 110.77 - samples/sec: 298.62 - lr: 0.000113 - momentum: 0.000000
2023-10-12 17:56:08,456 epoch 4 - iter 594/1984 - loss 0.05633847 - time (sec): 166.21 - samples/sec: 301.27 - lr: 0.000112 - momentum: 0.000000
2023-10-12 17:57:00,472 epoch 4 - iter 792/1984 - loss 0.05909280 - time (sec): 218.23 - samples/sec: 301.30 - lr: 0.000110 - momentum: 0.000000
2023-10-12 17:57:54,142 epoch 4 - iter 990/1984 - loss 0.06026255 - time (sec): 271.90 - samples/sec: 306.35 - lr: 0.000108 - momentum: 0.000000
2023-10-12 17:58:46,218 epoch 4 - iter 1188/1984 - loss 0.05861280 - time (sec): 323.97 - samples/sec: 309.30 - lr: 0.000107 - momentum: 0.000000
2023-10-12 17:59:38,647 epoch 4 - iter 1386/1984 - loss 0.05877059 - time (sec): 376.40 - samples/sec: 307.34 - lr: 0.000105 - momentum: 0.000000
2023-10-12 18:00:33,949 epoch 4 - iter 1584/1984 - loss 0.05806944 - time (sec): 431.71 - samples/sec: 306.80 - lr: 0.000103 - momentum: 0.000000
2023-10-12 18:01:25,964 epoch 4 - iter 1782/1984 - loss 0.05889933 - time (sec): 483.72 - samples/sec: 305.46 - lr: 0.000102 - momentum: 0.000000
2023-10-12 18:02:18,642 epoch 4 - iter 1980/1984 - loss 0.05870714 - time (sec): 536.40 - samples/sec: 304.90 - lr: 0.000100 - momentum: 0.000000
2023-10-12 18:02:19,847 ----------------------------------------------------------------------------------------------------
2023-10-12 18:02:19,848 EPOCH 4 done: loss 0.0587 - lr: 0.000100
2023-10-12 18:02:44,009 DEV : loss 0.12530608475208282 - f1-score (micro avg) 0.7549
2023-10-12 18:02:44,047 saving best model
2023-10-12 18:02:47,075 ----------------------------------------------------------------------------------------------------
2023-10-12 18:03:45,754 epoch 5 - iter 198/1984 - loss 0.03684763 - time (sec): 58.67 - samples/sec: 278.92 - lr: 0.000098 - momentum: 0.000000
2023-10-12 18:04:42,683 epoch 5 - iter 396/1984 - loss 0.03542942 - time (sec): 115.60 - samples/sec: 283.18 - lr: 0.000097 - momentum: 0.000000
2023-10-12 18:05:33,118 epoch 5 - iter 594/1984 - loss 0.03923047 - time (sec): 166.04 - samples/sec: 291.45 - lr: 0.000095 - momentum: 0.000000
2023-10-12 18:06:24,269 epoch 5 - iter 792/1984 - loss 0.04070365 - time (sec): 217.19 - samples/sec: 297.96 - lr: 0.000093 - momentum: 0.000000
2023-10-12 18:07:16,126 epoch 5 - iter 990/1984 - loss 0.03866158 - time (sec): 269.05 - samples/sec: 300.33 - lr: 0.000092 - momentum: 0.000000
2023-10-12 18:08:11,590 epoch 5 - iter 1188/1984 - loss 0.04048969 - time (sec): 324.51 - samples/sec: 300.36 - lr: 0.000090 - momentum: 0.000000
2023-10-12 18:09:06,534 epoch 5 - iter 1386/1984 - loss 0.04244996 - time (sec): 379.45 - samples/sec: 301.32 - lr: 0.000088 - momentum: 0.000000
2023-10-12 18:10:03,788 epoch 5 - iter 1584/1984 - loss 0.04340599 - time (sec): 436.71 - samples/sec: 299.41 - lr: 0.000087 - momentum: 0.000000
2023-10-12 18:10:55,718 epoch 5 - iter 1782/1984 - loss 0.04308350 - time (sec): 488.64 - samples/sec: 302.17 - lr: 0.000085 - momentum: 0.000000
2023-10-12 18:11:49,526 epoch 5 - iter 1980/1984 - loss 0.04432984 - time (sec): 542.44 - samples/sec: 301.87 - lr: 0.000083 - momentum: 0.000000
2023-10-12 18:11:50,632 ----------------------------------------------------------------------------------------------------
2023-10-12 18:11:50,632 EPOCH 5 done: loss 0.0443 - lr: 0.000083
2023-10-12 18:12:18,411 DEV : loss 0.1436772644519806 - f1-score (micro avg) 0.7392
2023-10-12 18:12:18,452 ----------------------------------------------------------------------------------------------------
2023-10-12 18:13:10,838 epoch 6 - iter 198/1984 - loss 0.03726587 - time (sec): 52.38 - samples/sec: 316.11 - lr: 0.000082 - momentum: 0.000000
2023-10-12 18:14:06,410 epoch 6 - iter 396/1984 - loss 0.02956098 - time (sec): 107.96 - samples/sec: 303.22 - lr: 0.000080 - momentum: 0.000000
2023-10-12 18:14:58,937 epoch 6 - iter 594/1984 - loss 0.03136481 - time (sec): 160.48 - samples/sec: 306.32 - lr: 0.000078 - momentum: 0.000000
2023-10-12 18:15:54,784 epoch 6 - iter 792/1984 - loss 0.03209960 - time (sec): 216.33 - samples/sec: 303.47 - lr: 0.000077 - momentum: 0.000000
2023-10-12 18:16:47,593 epoch 6 - iter 990/1984 - loss 0.03287748 - time (sec): 269.14 - samples/sec: 306.09 - lr: 0.000075 - momentum: 0.000000
2023-10-12 18:17:44,743 epoch 6 - iter 1188/1984 - loss 0.03282659 - time (sec): 326.29 - samples/sec: 300.73 - lr: 0.000073 - momentum: 0.000000
2023-10-12 18:18:38,567 epoch 6 - iter 1386/1984 - loss 0.03292277 - time (sec): 380.11 - samples/sec: 301.47 - lr: 0.000072 - momentum: 0.000000
2023-10-12 18:19:30,537 epoch 6 - iter 1584/1984 - loss 0.03367855 - time (sec): 432.08 - samples/sec: 302.95 - lr: 0.000070 - momentum: 0.000000
2023-10-12 18:20:27,404 epoch 6 - iter 1782/1984 - loss 0.03441227 - time (sec): 488.95 - samples/sec: 301.11 - lr: 0.000068 - momentum: 0.000000
2023-10-12 18:21:20,899 epoch 6 - iter 1980/1984 - loss 0.03431832 - time (sec): 542.45 - samples/sec: 301.61 - lr: 0.000067 - momentum: 0.000000
2023-10-12 18:21:21,962 ----------------------------------------------------------------------------------------------------
2023-10-12 18:21:21,962 EPOCH 6 done: loss 0.0342 - lr: 0.000067
2023-10-12 18:21:48,190 DEV : loss 0.18953415751457214 - f1-score (micro avg) 0.7546
2023-10-12 18:21:48,236 ----------------------------------------------------------------------------------------------------
2023-10-12 18:22:42,513 epoch 7 - iter 198/1984 - loss 0.01880897 - time (sec): 54.27 - samples/sec: 300.69 - lr: 0.000065 - momentum: 0.000000
2023-10-12 18:23:36,788 epoch 7 - iter 396/1984 - loss 0.02129238 - time (sec): 108.55 - samples/sec: 310.17 - lr: 0.000063 - momentum: 0.000000
2023-10-12 18:24:27,787 epoch 7 - iter 594/1984 - loss 0.02038739 - time (sec): 159.55 - samples/sec: 308.46 - lr: 0.000062 - momentum: 0.000000
2023-10-12 18:25:18,964 epoch 7 - iter 792/1984 - loss 0.02267150 - time (sec): 210.73 - samples/sec: 305.18 - lr: 0.000060 - momentum: 0.000000
2023-10-12 18:26:11,225 epoch 7 - iter 990/1984 - loss 0.02325419 - time (sec): 262.99 - samples/sec: 308.16 - lr: 0.000058 - momentum: 0.000000
2023-10-12 18:27:02,156 epoch 7 - iter 1188/1984 - loss 0.02443802 - time (sec): 313.92 - samples/sec: 310.04 - lr: 0.000057 - momentum: 0.000000
2023-10-12 18:27:53,992 epoch 7 - iter 1386/1984 - loss 0.02362788 - time (sec): 365.75 - samples/sec: 310.65 - lr: 0.000055 - momentum: 0.000000
2023-10-12 18:28:46,404 epoch 7 - iter 1584/1984 - loss 0.02445248 - time (sec): 418.17 - samples/sec: 310.92 - lr: 0.000053 - momentum: 0.000000
2023-10-12 18:29:42,092 epoch 7 - iter 1782/1984 - loss 0.02416577 - time (sec): 473.85 - samples/sec: 310.85 - lr: 0.000052 - momentum: 0.000000
2023-10-12 18:30:37,599 epoch 7 - iter 1980/1984 - loss 0.02497708 - time (sec): 529.36 - samples/sec: 308.90 - lr: 0.000050 - momentum: 0.000000
2023-10-12 18:30:38,861 ----------------------------------------------------------------------------------------------------
2023-10-12 18:30:38,862 EPOCH 7 done: loss 0.0251 - lr: 0.000050
2023-10-12 18:31:11,064 DEV : loss 0.2033829540014267 - f1-score (micro avg) 0.735
2023-10-12 18:31:11,110 ----------------------------------------------------------------------------------------------------
2023-10-12 18:32:08,080 epoch 8 - iter 198/1984 - loss 0.01615564 - time (sec): 56.97 - samples/sec: 290.72 - lr: 0.000048 - momentum: 0.000000
2023-10-12 18:33:04,407 epoch 8 - iter 396/1984 - loss 0.01873509 - time (sec): 113.29 - samples/sec: 297.31 - lr: 0.000047 - momentum: 0.000000
2023-10-12 18:34:00,564 epoch 8 - iter 594/1984 - loss 0.01784955 - time (sec): 169.45 - samples/sec: 304.92 - lr: 0.000045 - momentum: 0.000000
2023-10-12 18:34:54,773 epoch 8 - iter 792/1984 - loss 0.01716047 - time (sec): 223.66 - samples/sec: 301.53 - lr: 0.000043 - momentum: 0.000000
2023-10-12 18:35:48,316 epoch 8 - iter 990/1984 - loss 0.01676986 - time (sec): 277.20 - samples/sec: 299.97 - lr: 0.000042 - momentum: 0.000000
2023-10-12 18:36:43,833 epoch 8 - iter 1188/1984 - loss 0.01653262 - time (sec): 332.72 - samples/sec: 296.71 - lr: 0.000040 - momentum: 0.000000
2023-10-12 18:37:37,320 epoch 8 - iter 1386/1984 - loss 0.01775393 - time (sec): 386.21 - samples/sec: 295.96 - lr: 0.000038 - momentum: 0.000000
2023-10-12 18:38:30,408 epoch 8 - iter 1584/1984 - loss 0.01695689 - time (sec): 439.30 - samples/sec: 296.91 - lr: 0.000037 - momentum: 0.000000
2023-10-12 18:39:23,967 epoch 8 - iter 1782/1984 - loss 0.01771032 - time (sec): 492.85 - samples/sec: 298.27 - lr: 0.000035 - momentum: 0.000000
2023-10-12 18:40:16,256 epoch 8 - iter 1980/1984 - loss 0.01950296 - time (sec): 545.14 - samples/sec: 299.99 - lr: 0.000033 - momentum: 0.000000
2023-10-12 18:40:17,576 ----------------------------------------------------------------------------------------------------
2023-10-12 18:40:17,577 EPOCH 8 done: loss 0.0195 - lr: 0.000033
2023-10-12 18:40:48,348 DEV : loss 0.20685595273971558 - f1-score (micro avg) 0.7506
2023-10-12 18:40:48,388 ----------------------------------------------------------------------------------------------------
2023-10-12 18:41:43,210 epoch 9 - iter 198/1984 - loss 0.01469186 - time (sec): 54.82 - samples/sec: 297.92 - lr: 0.000032 - momentum: 0.000000
2023-10-12 18:42:34,820 epoch 9 - iter 396/1984 - loss 0.01231871 - time (sec): 106.43 - samples/sec: 307.79 - lr: 0.000030 - momentum: 0.000000
2023-10-12 18:43:28,721 epoch 9 - iter 594/1984 - loss 0.01418213 - time (sec): 160.33 - samples/sec: 308.89 - lr: 0.000028 - momentum: 0.000000
2023-10-12 18:44:21,248 epoch 9 - iter 792/1984 - loss 0.01449414 - time (sec): 212.86 - samples/sec: 309.93 - lr: 0.000027 - momentum: 0.000000
2023-10-12 18:45:15,929 epoch 9 - iter 990/1984 - loss 0.01310674 - time (sec): 267.54 - samples/sec: 311.53 - lr: 0.000025 - momentum: 0.000000
2023-10-12 18:46:08,149 epoch 9 - iter 1188/1984 - loss 0.01385134 - time (sec): 319.76 - samples/sec: 312.20 - lr: 0.000023 - momentum: 0.000000
2023-10-12 18:46:58,627 epoch 9 - iter 1386/1984 - loss 0.01328832 - time (sec): 370.24 - samples/sec: 313.82 - lr: 0.000022 - momentum: 0.000000
2023-10-12 18:47:51,195 epoch 9 - iter 1584/1984 - loss 0.01329775 - time (sec): 422.80 - samples/sec: 313.30 - lr: 0.000020 - momentum: 0.000000
2023-10-12 18:48:46,351 epoch 9 - iter 1782/1984 - loss 0.01366045 - time (sec): 477.96 - samples/sec: 309.53 - lr: 0.000018 - momentum: 0.000000
2023-10-12 18:49:40,513 epoch 9 - iter 1980/1984 - loss 0.01397599 - time (sec): 532.12 - samples/sec: 307.31 - lr: 0.000017 - momentum: 0.000000
2023-10-12 18:49:41,587 ----------------------------------------------------------------------------------------------------
2023-10-12 18:49:41,588 EPOCH 9 done: loss 0.0140 - lr: 0.000017
2023-10-12 18:50:07,293 DEV : loss 0.2238703966140747 - f1-score (micro avg) 0.7411
2023-10-12 18:50:07,343 ----------------------------------------------------------------------------------------------------
2023-10-12 18:51:01,512 epoch 10 - iter 198/1984 - loss 0.01050438 - time (sec): 54.17 - samples/sec: 309.79 - lr: 0.000015 - momentum: 0.000000
2023-10-12 18:51:53,307 epoch 10 - iter 396/1984 - loss 0.00772763 - time (sec): 105.96 - samples/sec: 305.80 - lr: 0.000013 - momentum: 0.000000
2023-10-12 18:52:48,019 epoch 10 - iter 594/1984 - loss 0.01012820 - time (sec): 160.67 - samples/sec: 299.59 - lr: 0.000012 - momentum: 0.000000
2023-10-12 18:53:41,519 epoch 10 - iter 792/1984 - loss 0.00947119 - time (sec): 214.17 - samples/sec: 301.96 - lr: 0.000010 - momentum: 0.000000
2023-10-12 18:54:40,509 epoch 10 - iter 990/1984 - loss 0.01022609 - time (sec): 273.16 - samples/sec: 300.88 - lr: 0.000008 - momentum: 0.000000
2023-10-12 18:55:34,282 epoch 10 - iter 1188/1984 - loss 0.00989781 - time (sec): 326.94 - samples/sec: 302.53 - lr: 0.000007 - momentum: 0.000000
2023-10-12 18:56:29,254 epoch 10 - iter 1386/1984 - loss 0.00983913 - time (sec): 381.91 - samples/sec: 300.51 - lr: 0.000005 - momentum: 0.000000
2023-10-12 18:57:21,637 epoch 10 - iter 1584/1984 - loss 0.01002305 - time (sec): 434.29 - samples/sec: 301.31 - lr: 0.000003 - momentum: 0.000000
2023-10-12 18:58:16,687 epoch 10 - iter 1782/1984 - loss 0.00997923 - time (sec): 489.34 - samples/sec: 301.28 - lr: 0.000002 - momentum: 0.000000
2023-10-12 18:59:09,635 epoch 10 - iter 1980/1984 - loss 0.01053401 - time (sec): 542.29 - samples/sec: 301.94 - lr: 0.000000 - momentum: 0.000000
2023-10-12 18:59:10,685 ----------------------------------------------------------------------------------------------------
2023-10-12 18:59:10,685 EPOCH 10 done: loss 0.0105 - lr: 0.000000
2023-10-12 18:59:40,730 DEV : loss 0.22921979427337646 - f1-score (micro avg) 0.7538
2023-10-12 18:59:41,817 ----------------------------------------------------------------------------------------------------
2023-10-12 18:59:41,819 Loading model from best epoch ...
2023-10-12 18:59:46,867 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 19:00:16,825
Results:
- F-score (micro) 0.763
- F-score (macro) 0.6832
- Accuracy 0.6361
By class:
precision recall f1-score support
LOC 0.7887 0.8489 0.8176 655
PER 0.7465 0.7265 0.7364 223
ORG 0.5421 0.4567 0.4957 127
micro avg 0.7541 0.7721 0.7630 1005
macro avg 0.6924 0.6773 0.6832 1005
weighted avg 0.7481 0.7721 0.7589 1005
2023-10-12 19:00:16,826 ----------------------------------------------------------------------------------------------------
|