File size: 25,321 Bytes
50b5e2c |
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 |
2023-10-13 12:01:40,581 ----------------------------------------------------------------------------------------------------
2023-10-13 12:01:40,583 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()
)"
2023-10-13 12:01:40,583 ----------------------------------------------------------------------------------------------------
2023-10-13 12:01:40,583 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
2023-10-13 12:01:40,583 ----------------------------------------------------------------------------------------------------
2023-10-13 12:01:40,583 Train: 7936 sentences
2023-10-13 12:01:40,584 (train_with_dev=False, train_with_test=False)
2023-10-13 12:01:40,584 ----------------------------------------------------------------------------------------------------
2023-10-13 12:01:40,584 Training Params:
2023-10-13 12:01:40,584 - learning_rate: "0.00015"
2023-10-13 12:01:40,584 - mini_batch_size: "4"
2023-10-13 12:01:40,584 - max_epochs: "10"
2023-10-13 12:01:40,584 - shuffle: "True"
2023-10-13 12:01:40,584 ----------------------------------------------------------------------------------------------------
2023-10-13 12:01:40,584 Plugins:
2023-10-13 12:01:40,584 - TensorboardLogger
2023-10-13 12:01:40,584 - LinearScheduler | warmup_fraction: '0.1'
2023-10-13 12:01:40,584 ----------------------------------------------------------------------------------------------------
2023-10-13 12:01:40,584 Final evaluation on model from best epoch (best-model.pt)
2023-10-13 12:01:40,584 - metric: "('micro avg', 'f1-score')"
2023-10-13 12:01:40,584 ----------------------------------------------------------------------------------------------------
2023-10-13 12:01:40,585 Computation:
2023-10-13 12:01:40,585 - compute on device: cuda:0
2023-10-13 12:01:40,585 - embedding storage: none
2023-10-13 12:01:40,585 ----------------------------------------------------------------------------------------------------
2023-10-13 12:01:40,585 Model training base path: "hmbench-icdar/fr-hmbyt5-preliminary/byt5-small-historic-multilingual-span20-flax-bs4-wsFalse-e10-lr0.00015-poolingfirst-layers-1-crfFalse-5"
2023-10-13 12:01:40,585 ----------------------------------------------------------------------------------------------------
2023-10-13 12:01:40,585 ----------------------------------------------------------------------------------------------------
2023-10-13 12:01:40,585 Logging anything other than scalars to TensorBoard is currently not supported.
2023-10-13 12:02:36,986 epoch 1 - iter 198/1984 - loss 2.53587256 - time (sec): 56.40 - samples/sec: 313.46 - lr: 0.000015 - momentum: 0.000000
2023-10-13 12:03:30,634 epoch 1 - iter 396/1984 - loss 2.35663481 - time (sec): 110.05 - samples/sec: 303.24 - lr: 0.000030 - momentum: 0.000000
2023-10-13 12:04:25,848 epoch 1 - iter 594/1984 - loss 2.03833143 - time (sec): 165.26 - samples/sec: 306.16 - lr: 0.000045 - momentum: 0.000000
2023-10-13 12:05:19,291 epoch 1 - iter 792/1984 - loss 1.75219619 - time (sec): 218.70 - samples/sec: 299.99 - lr: 0.000060 - momentum: 0.000000
2023-10-13 12:06:17,514 epoch 1 - iter 990/1984 - loss 1.50365684 - time (sec): 276.93 - samples/sec: 294.62 - lr: 0.000075 - momentum: 0.000000
2023-10-13 12:07:11,217 epoch 1 - iter 1188/1984 - loss 1.30904393 - time (sec): 330.63 - samples/sec: 295.02 - lr: 0.000090 - momentum: 0.000000
2023-10-13 12:08:05,806 epoch 1 - iter 1386/1984 - loss 1.15357563 - time (sec): 385.22 - samples/sec: 296.78 - lr: 0.000105 - momentum: 0.000000
2023-10-13 12:09:02,915 epoch 1 - iter 1584/1984 - loss 1.03648294 - time (sec): 442.33 - samples/sec: 294.58 - lr: 0.000120 - momentum: 0.000000
2023-10-13 12:10:00,529 epoch 1 - iter 1782/1984 - loss 0.93126337 - time (sec): 499.94 - samples/sec: 296.04 - lr: 0.000135 - momentum: 0.000000
2023-10-13 12:11:00,429 epoch 1 - iter 1980/1984 - loss 0.85719194 - time (sec): 559.84 - samples/sec: 292.50 - lr: 0.000150 - momentum: 0.000000
2023-10-13 12:11:01,586 ----------------------------------------------------------------------------------------------------
2023-10-13 12:11:01,587 EPOCH 1 done: loss 0.8562 - lr: 0.000150
2023-10-13 12:11:27,268 DEV : loss 0.12921574711799622 - f1-score (micro avg) 0.6608
2023-10-13 12:11:27,311 saving best model
2023-10-13 12:11:28,286 ----------------------------------------------------------------------------------------------------
2023-10-13 12:12:25,572 epoch 2 - iter 198/1984 - loss 0.15246598 - time (sec): 57.28 - samples/sec: 288.91 - lr: 0.000148 - momentum: 0.000000
2023-10-13 12:13:23,204 epoch 2 - iter 396/1984 - loss 0.13947723 - time (sec): 114.92 - samples/sec: 289.19 - lr: 0.000147 - momentum: 0.000000
2023-10-13 12:14:19,334 epoch 2 - iter 594/1984 - loss 0.13236118 - time (sec): 171.05 - samples/sec: 293.49 - lr: 0.000145 - momentum: 0.000000
2023-10-13 12:15:15,682 epoch 2 - iter 792/1984 - loss 0.13139937 - time (sec): 227.39 - samples/sec: 289.49 - lr: 0.000143 - momentum: 0.000000
2023-10-13 12:16:13,148 epoch 2 - iter 990/1984 - loss 0.12747303 - time (sec): 284.86 - samples/sec: 288.91 - lr: 0.000142 - momentum: 0.000000
2023-10-13 12:17:07,128 epoch 2 - iter 1188/1984 - loss 0.12581150 - time (sec): 338.84 - samples/sec: 290.91 - lr: 0.000140 - momentum: 0.000000
2023-10-13 12:18:01,605 epoch 2 - iter 1386/1984 - loss 0.12386868 - time (sec): 393.32 - samples/sec: 292.02 - lr: 0.000138 - momentum: 0.000000
2023-10-13 12:18:57,699 epoch 2 - iter 1584/1984 - loss 0.12091584 - time (sec): 449.41 - samples/sec: 291.07 - lr: 0.000137 - momentum: 0.000000
2023-10-13 12:19:53,333 epoch 2 - iter 1782/1984 - loss 0.11952771 - time (sec): 505.04 - samples/sec: 289.53 - lr: 0.000135 - momentum: 0.000000
2023-10-13 12:20:51,124 epoch 2 - iter 1980/1984 - loss 0.11686668 - time (sec): 562.84 - samples/sec: 290.88 - lr: 0.000133 - momentum: 0.000000
2023-10-13 12:20:52,222 ----------------------------------------------------------------------------------------------------
2023-10-13 12:20:52,223 EPOCH 2 done: loss 0.1168 - lr: 0.000133
2023-10-13 12:21:22,532 DEV : loss 0.08594389259815216 - f1-score (micro avg) 0.7326
2023-10-13 12:21:22,586 saving best model
2023-10-13 12:21:25,326 ----------------------------------------------------------------------------------------------------
2023-10-13 12:22:22,376 epoch 3 - iter 198/1984 - loss 0.06835268 - time (sec): 57.05 - samples/sec: 282.30 - lr: 0.000132 - momentum: 0.000000
2023-10-13 12:23:21,479 epoch 3 - iter 396/1984 - loss 0.07762823 - time (sec): 116.15 - samples/sec: 279.06 - lr: 0.000130 - momentum: 0.000000
2023-10-13 12:24:16,710 epoch 3 - iter 594/1984 - loss 0.07827809 - time (sec): 171.38 - samples/sec: 283.05 - lr: 0.000128 - momentum: 0.000000
2023-10-13 12:25:12,056 epoch 3 - iter 792/1984 - loss 0.07869207 - time (sec): 226.73 - samples/sec: 286.28 - lr: 0.000127 - momentum: 0.000000
2023-10-13 12:26:07,407 epoch 3 - iter 990/1984 - loss 0.07826522 - time (sec): 282.08 - samples/sec: 287.01 - lr: 0.000125 - momentum: 0.000000
2023-10-13 12:27:04,007 epoch 3 - iter 1188/1984 - loss 0.07782816 - time (sec): 338.68 - samples/sec: 288.60 - lr: 0.000123 - momentum: 0.000000
2023-10-13 12:28:00,792 epoch 3 - iter 1386/1984 - loss 0.07814075 - time (sec): 395.46 - samples/sec: 288.89 - lr: 0.000122 - momentum: 0.000000
2023-10-13 12:28:58,604 epoch 3 - iter 1584/1984 - loss 0.07681156 - time (sec): 453.27 - samples/sec: 288.09 - lr: 0.000120 - momentum: 0.000000
2023-10-13 12:29:54,823 epoch 3 - iter 1782/1984 - loss 0.07566942 - time (sec): 509.49 - samples/sec: 288.85 - lr: 0.000118 - momentum: 0.000000
2023-10-13 12:30:54,387 epoch 3 - iter 1980/1984 - loss 0.07589565 - time (sec): 569.06 - samples/sec: 287.58 - lr: 0.000117 - momentum: 0.000000
2023-10-13 12:30:55,447 ----------------------------------------------------------------------------------------------------
2023-10-13 12:30:55,448 EPOCH 3 done: loss 0.0758 - lr: 0.000117
2023-10-13 12:31:22,850 DEV : loss 0.09770967811346054 - f1-score (micro avg) 0.7546
2023-10-13 12:31:22,896 saving best model
2023-10-13 12:31:25,680 ----------------------------------------------------------------------------------------------------
2023-10-13 12:32:22,939 epoch 4 - iter 198/1984 - loss 0.05842859 - time (sec): 57.25 - samples/sec: 287.82 - lr: 0.000115 - momentum: 0.000000
2023-10-13 12:33:20,230 epoch 4 - iter 396/1984 - loss 0.05467250 - time (sec): 114.55 - samples/sec: 285.40 - lr: 0.000113 - momentum: 0.000000
2023-10-13 12:34:20,580 epoch 4 - iter 594/1984 - loss 0.05230249 - time (sec): 174.90 - samples/sec: 279.68 - lr: 0.000112 - momentum: 0.000000
2023-10-13 12:35:21,055 epoch 4 - iter 792/1984 - loss 0.05292552 - time (sec): 235.37 - samples/sec: 277.36 - lr: 0.000110 - momentum: 0.000000
2023-10-13 12:36:20,546 epoch 4 - iter 990/1984 - loss 0.05358516 - time (sec): 294.86 - samples/sec: 279.46 - lr: 0.000108 - momentum: 0.000000
2023-10-13 12:37:21,108 epoch 4 - iter 1188/1984 - loss 0.05228562 - time (sec): 355.42 - samples/sec: 279.04 - lr: 0.000107 - momentum: 0.000000
2023-10-13 12:38:21,766 epoch 4 - iter 1386/1984 - loss 0.05219360 - time (sec): 416.08 - samples/sec: 276.24 - lr: 0.000105 - momentum: 0.000000
2023-10-13 12:39:19,041 epoch 4 - iter 1584/1984 - loss 0.05282172 - time (sec): 473.36 - samples/sec: 277.03 - lr: 0.000103 - momentum: 0.000000
2023-10-13 12:40:16,271 epoch 4 - iter 1782/1984 - loss 0.05363652 - time (sec): 530.59 - samples/sec: 277.86 - lr: 0.000102 - momentum: 0.000000
2023-10-13 12:41:14,283 epoch 4 - iter 1980/1984 - loss 0.05409466 - time (sec): 588.60 - samples/sec: 278.11 - lr: 0.000100 - momentum: 0.000000
2023-10-13 12:41:15,420 ----------------------------------------------------------------------------------------------------
2023-10-13 12:41:15,421 EPOCH 4 done: loss 0.0540 - lr: 0.000100
2023-10-13 12:41:43,934 DEV : loss 0.13024039566516876 - f1-score (micro avg) 0.7696
2023-10-13 12:41:43,978 saving best model
2023-10-13 12:41:49,138 ----------------------------------------------------------------------------------------------------
2023-10-13 12:42:46,154 epoch 5 - iter 198/1984 - loss 0.03679399 - time (sec): 57.01 - samples/sec: 294.51 - lr: 0.000098 - momentum: 0.000000
2023-10-13 12:43:43,295 epoch 5 - iter 396/1984 - loss 0.03905502 - time (sec): 114.15 - samples/sec: 292.02 - lr: 0.000097 - momentum: 0.000000
2023-10-13 12:44:40,686 epoch 5 - iter 594/1984 - loss 0.04296223 - time (sec): 171.54 - samples/sec: 291.74 - lr: 0.000095 - momentum: 0.000000
2023-10-13 12:45:35,672 epoch 5 - iter 792/1984 - loss 0.04216427 - time (sec): 226.53 - samples/sec: 293.31 - lr: 0.000093 - momentum: 0.000000
2023-10-13 12:46:31,117 epoch 5 - iter 990/1984 - loss 0.04149098 - time (sec): 281.97 - samples/sec: 292.56 - lr: 0.000092 - momentum: 0.000000
2023-10-13 12:47:30,500 epoch 5 - iter 1188/1984 - loss 0.04224225 - time (sec): 341.36 - samples/sec: 287.90 - lr: 0.000090 - momentum: 0.000000
2023-10-13 12:48:25,231 epoch 5 - iter 1386/1984 - loss 0.04138722 - time (sec): 396.09 - samples/sec: 288.37 - lr: 0.000088 - momentum: 0.000000
2023-10-13 12:49:21,590 epoch 5 - iter 1584/1984 - loss 0.04151736 - time (sec): 452.45 - samples/sec: 290.10 - lr: 0.000087 - momentum: 0.000000
2023-10-13 12:50:20,385 epoch 5 - iter 1782/1984 - loss 0.04213374 - time (sec): 511.24 - samples/sec: 288.80 - lr: 0.000085 - momentum: 0.000000
2023-10-13 12:51:16,349 epoch 5 - iter 1980/1984 - loss 0.04118791 - time (sec): 567.21 - samples/sec: 288.75 - lr: 0.000083 - momentum: 0.000000
2023-10-13 12:51:17,450 ----------------------------------------------------------------------------------------------------
2023-10-13 12:51:17,451 EPOCH 5 done: loss 0.0411 - lr: 0.000083
2023-10-13 12:51:45,096 DEV : loss 0.1498626172542572 - f1-score (micro avg) 0.7658
2023-10-13 12:51:45,140 ----------------------------------------------------------------------------------------------------
2023-10-13 12:52:42,428 epoch 6 - iter 198/1984 - loss 0.02535367 - time (sec): 57.29 - samples/sec: 298.24 - lr: 0.000082 - momentum: 0.000000
2023-10-13 12:53:43,797 epoch 6 - iter 396/1984 - loss 0.02460873 - time (sec): 118.65 - samples/sec: 282.58 - lr: 0.000080 - momentum: 0.000000
2023-10-13 12:54:39,904 epoch 6 - iter 594/1984 - loss 0.02590999 - time (sec): 174.76 - samples/sec: 282.27 - lr: 0.000078 - momentum: 0.000000
2023-10-13 12:55:35,769 epoch 6 - iter 792/1984 - loss 0.02897049 - time (sec): 230.63 - samples/sec: 285.87 - lr: 0.000077 - momentum: 0.000000
2023-10-13 12:56:29,696 epoch 6 - iter 990/1984 - loss 0.02985477 - time (sec): 284.55 - samples/sec: 289.08 - lr: 0.000075 - momentum: 0.000000
2023-10-13 12:57:23,689 epoch 6 - iter 1188/1984 - loss 0.02976773 - time (sec): 338.55 - samples/sec: 291.14 - lr: 0.000073 - momentum: 0.000000
2023-10-13 12:58:17,578 epoch 6 - iter 1386/1984 - loss 0.02995438 - time (sec): 392.44 - samples/sec: 292.99 - lr: 0.000072 - momentum: 0.000000
2023-10-13 12:59:11,837 epoch 6 - iter 1584/1984 - loss 0.03040644 - time (sec): 446.70 - samples/sec: 293.10 - lr: 0.000070 - momentum: 0.000000
2023-10-13 13:00:05,389 epoch 6 - iter 1782/1984 - loss 0.03000902 - time (sec): 500.25 - samples/sec: 294.19 - lr: 0.000068 - momentum: 0.000000
2023-10-13 13:00:58,136 epoch 6 - iter 1980/1984 - loss 0.03131165 - time (sec): 552.99 - samples/sec: 296.04 - lr: 0.000067 - momentum: 0.000000
2023-10-13 13:00:59,167 ----------------------------------------------------------------------------------------------------
2023-10-13 13:00:59,167 EPOCH 6 done: loss 0.0314 - lr: 0.000067
2023-10-13 13:01:25,953 DEV : loss 0.1546768993139267 - f1-score (micro avg) 0.7604
2023-10-13 13:01:25,994 ----------------------------------------------------------------------------------------------------
2023-10-13 13:02:20,487 epoch 7 - iter 198/1984 - loss 0.02164230 - time (sec): 54.49 - samples/sec: 302.56 - lr: 0.000065 - momentum: 0.000000
2023-10-13 13:03:15,597 epoch 7 - iter 396/1984 - loss 0.02187396 - time (sec): 109.60 - samples/sec: 294.20 - lr: 0.000063 - momentum: 0.000000
2023-10-13 13:04:11,813 epoch 7 - iter 594/1984 - loss 0.02064500 - time (sec): 165.82 - samples/sec: 296.65 - lr: 0.000062 - momentum: 0.000000
2023-10-13 13:05:08,654 epoch 7 - iter 792/1984 - loss 0.02239191 - time (sec): 222.66 - samples/sec: 292.47 - lr: 0.000060 - momentum: 0.000000
2023-10-13 13:06:06,207 epoch 7 - iter 990/1984 - loss 0.02117540 - time (sec): 280.21 - samples/sec: 290.85 - lr: 0.000058 - momentum: 0.000000
2023-10-13 13:07:00,612 epoch 7 - iter 1188/1984 - loss 0.02163675 - time (sec): 334.62 - samples/sec: 292.24 - lr: 0.000057 - momentum: 0.000000
2023-10-13 13:07:58,537 epoch 7 - iter 1386/1984 - loss 0.02215893 - time (sec): 392.54 - samples/sec: 290.62 - lr: 0.000055 - momentum: 0.000000
2023-10-13 13:08:52,925 epoch 7 - iter 1584/1984 - loss 0.02198376 - time (sec): 446.93 - samples/sec: 290.89 - lr: 0.000053 - momentum: 0.000000
2023-10-13 13:09:48,604 epoch 7 - iter 1782/1984 - loss 0.02278283 - time (sec): 502.61 - samples/sec: 292.61 - lr: 0.000052 - momentum: 0.000000
2023-10-13 13:10:47,760 epoch 7 - iter 1980/1984 - loss 0.02389101 - time (sec): 561.76 - samples/sec: 291.52 - lr: 0.000050 - momentum: 0.000000
2023-10-13 13:10:48,845 ----------------------------------------------------------------------------------------------------
2023-10-13 13:10:48,845 EPOCH 7 done: loss 0.0239 - lr: 0.000050
2023-10-13 13:11:15,612 DEV : loss 0.183299720287323 - f1-score (micro avg) 0.7642
2023-10-13 13:11:15,659 ----------------------------------------------------------------------------------------------------
2023-10-13 13:12:11,816 epoch 8 - iter 198/1984 - loss 0.01535571 - time (sec): 56.16 - samples/sec: 293.44 - lr: 0.000048 - momentum: 0.000000
2023-10-13 13:13:07,794 epoch 8 - iter 396/1984 - loss 0.01489584 - time (sec): 112.13 - samples/sec: 296.02 - lr: 0.000047 - momentum: 0.000000
2023-10-13 13:14:01,252 epoch 8 - iter 594/1984 - loss 0.01439117 - time (sec): 165.59 - samples/sec: 297.78 - lr: 0.000045 - momentum: 0.000000
2023-10-13 13:14:54,994 epoch 8 - iter 792/1984 - loss 0.01550966 - time (sec): 219.33 - samples/sec: 300.88 - lr: 0.000043 - momentum: 0.000000
2023-10-13 13:15:53,380 epoch 8 - iter 990/1984 - loss 0.01519291 - time (sec): 277.72 - samples/sec: 296.15 - lr: 0.000042 - momentum: 0.000000
2023-10-13 13:16:45,061 epoch 8 - iter 1188/1984 - loss 0.01508252 - time (sec): 329.40 - samples/sec: 299.29 - lr: 0.000040 - momentum: 0.000000
2023-10-13 13:17:41,489 epoch 8 - iter 1386/1984 - loss 0.01547066 - time (sec): 385.83 - samples/sec: 297.82 - lr: 0.000038 - momentum: 0.000000
2023-10-13 13:18:34,335 epoch 8 - iter 1584/1984 - loss 0.01553468 - time (sec): 438.67 - samples/sec: 297.60 - lr: 0.000037 - momentum: 0.000000
2023-10-13 13:19:27,559 epoch 8 - iter 1782/1984 - loss 0.01564675 - time (sec): 491.90 - samples/sec: 298.83 - lr: 0.000035 - momentum: 0.000000
2023-10-13 13:20:20,410 epoch 8 - iter 1980/1984 - loss 0.01495060 - time (sec): 544.75 - samples/sec: 300.60 - lr: 0.000033 - momentum: 0.000000
2023-10-13 13:20:21,482 ----------------------------------------------------------------------------------------------------
2023-10-13 13:20:21,483 EPOCH 8 done: loss 0.0151 - lr: 0.000033
2023-10-13 13:20:48,505 DEV : loss 0.20292401313781738 - f1-score (micro avg) 0.7682
2023-10-13 13:20:48,555 ----------------------------------------------------------------------------------------------------
2023-10-13 13:21:39,886 epoch 9 - iter 198/1984 - loss 0.01309539 - time (sec): 51.33 - samples/sec: 307.26 - lr: 0.000032 - momentum: 0.000000
2023-10-13 13:22:31,621 epoch 9 - iter 396/1984 - loss 0.01271783 - time (sec): 103.06 - samples/sec: 308.62 - lr: 0.000030 - momentum: 0.000000
2023-10-13 13:23:25,218 epoch 9 - iter 594/1984 - loss 0.01154190 - time (sec): 156.66 - samples/sec: 308.95 - lr: 0.000028 - momentum: 0.000000
2023-10-13 13:24:19,703 epoch 9 - iter 792/1984 - loss 0.01142730 - time (sec): 211.15 - samples/sec: 308.12 - lr: 0.000027 - momentum: 0.000000
2023-10-13 13:25:13,668 epoch 9 - iter 990/1984 - loss 0.01281801 - time (sec): 265.11 - samples/sec: 306.46 - lr: 0.000025 - momentum: 0.000000
2023-10-13 13:26:07,047 epoch 9 - iter 1188/1984 - loss 0.01251343 - time (sec): 318.49 - samples/sec: 301.02 - lr: 0.000023 - momentum: 0.000000
2023-10-13 13:27:02,940 epoch 9 - iter 1386/1984 - loss 0.01204291 - time (sec): 374.38 - samples/sec: 302.67 - lr: 0.000022 - momentum: 0.000000
2023-10-13 13:27:56,110 epoch 9 - iter 1584/1984 - loss 0.01200449 - time (sec): 427.55 - samples/sec: 303.59 - lr: 0.000020 - momentum: 0.000000
2023-10-13 13:28:51,579 epoch 9 - iter 1782/1984 - loss 0.01219217 - time (sec): 483.02 - samples/sec: 304.57 - lr: 0.000018 - momentum: 0.000000
2023-10-13 13:29:45,819 epoch 9 - iter 1980/1984 - loss 0.01242827 - time (sec): 537.26 - samples/sec: 304.48 - lr: 0.000017 - momentum: 0.000000
2023-10-13 13:29:47,047 ----------------------------------------------------------------------------------------------------
2023-10-13 13:29:47,048 EPOCH 9 done: loss 0.0124 - lr: 0.000017
2023-10-13 13:30:13,711 DEV : loss 0.21697697043418884 - f1-score (micro avg) 0.7617
2023-10-13 13:30:13,764 ----------------------------------------------------------------------------------------------------
2023-10-13 13:31:08,460 epoch 10 - iter 198/1984 - loss 0.00658872 - time (sec): 54.69 - samples/sec: 309.49 - lr: 0.000015 - momentum: 0.000000
2023-10-13 13:32:01,825 epoch 10 - iter 396/1984 - loss 0.00947681 - time (sec): 108.06 - samples/sec: 301.78 - lr: 0.000013 - momentum: 0.000000
2023-10-13 13:32:55,149 epoch 10 - iter 594/1984 - loss 0.00879204 - time (sec): 161.38 - samples/sec: 300.65 - lr: 0.000012 - momentum: 0.000000
2023-10-13 13:33:51,084 epoch 10 - iter 792/1984 - loss 0.00808015 - time (sec): 217.32 - samples/sec: 296.34 - lr: 0.000010 - momentum: 0.000000
2023-10-13 13:34:48,261 epoch 10 - iter 990/1984 - loss 0.00746892 - time (sec): 274.49 - samples/sec: 295.76 - lr: 0.000008 - momentum: 0.000000
2023-10-13 13:35:43,142 epoch 10 - iter 1188/1984 - loss 0.00746131 - time (sec): 329.38 - samples/sec: 297.49 - lr: 0.000007 - momentum: 0.000000
2023-10-13 13:36:37,996 epoch 10 - iter 1386/1984 - loss 0.00742811 - time (sec): 384.23 - samples/sec: 298.45 - lr: 0.000005 - momentum: 0.000000
2023-10-13 13:37:32,826 epoch 10 - iter 1584/1984 - loss 0.00777138 - time (sec): 439.06 - samples/sec: 299.37 - lr: 0.000003 - momentum: 0.000000
2023-10-13 13:38:29,578 epoch 10 - iter 1782/1984 - loss 0.00774151 - time (sec): 495.81 - samples/sec: 298.21 - lr: 0.000002 - momentum: 0.000000
2023-10-13 13:39:24,305 epoch 10 - iter 1980/1984 - loss 0.00806446 - time (sec): 550.54 - samples/sec: 297.17 - lr: 0.000000 - momentum: 0.000000
2023-10-13 13:39:25,573 ----------------------------------------------------------------------------------------------------
2023-10-13 13:39:25,574 EPOCH 10 done: loss 0.0080 - lr: 0.000000
2023-10-13 13:39:51,477 DEV : loss 0.22803443670272827 - f1-score (micro avg) 0.7611
2023-10-13 13:39:52,471 ----------------------------------------------------------------------------------------------------
2023-10-13 13:39:52,473 Loading model from best epoch ...
2023-10-13 13:39:57,370 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
2023-10-13 13:40:22,958
Results:
- F-score (micro) 0.7832
- F-score (macro) 0.6902
- Accuracy 0.6628
By class:
precision recall f1-score support
LOC 0.8351 0.8656 0.8501 655
PER 0.7284 0.7937 0.7597 223
ORG 0.4828 0.4409 0.4609 127
micro avg 0.7707 0.7960 0.7832 1005
macro avg 0.6821 0.7001 0.6902 1005
weighted avg 0.7669 0.7960 0.7808 1005
2023-10-13 13:40:22,958 ----------------------------------------------------------------------------------------------------
|