File size: 25,198 Bytes
522fb74 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 |
2023-10-06 11:18:54,722 ----------------------------------------------------------------------------------------------------
2023-10-06 11:18:54,723 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): T5LayerNorm()
(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): T5LayerNorm()
(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): T5LayerNorm()
(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): T5LayerNorm()
(dropout): Dropout(p=0.1, inplace=False)
)
)
)
)
(final_layer_norm): T5LayerNorm()
(dropout): Dropout(p=0.1, inplace=False)
)
)
)
(locked_dropout): LockedDropout(p=0.5)
(linear): Linear(in_features=1472, out_features=25, bias=True)
(loss_function): CrossEntropyLoss()
)"
2023-10-06 11:18:54,723 ----------------------------------------------------------------------------------------------------
2023-10-06 11:18:54,723 MultiCorpus: 1214 train + 266 dev + 251 test sentences
- NER_HIPE_2022 Corpus: 1214 train + 266 dev + 251 test sentences - /app/.flair/datasets/ner_hipe_2022/v2.1/ajmc/en/with_doc_seperator
2023-10-06 11:18:54,723 ----------------------------------------------------------------------------------------------------
2023-10-06 11:18:54,723 Train: 1214 sentences
2023-10-06 11:18:54,723 (train_with_dev=False, train_with_test=False)
2023-10-06 11:18:54,723 ----------------------------------------------------------------------------------------------------
2023-10-06 11:18:54,724 Training Params:
2023-10-06 11:18:54,724 - learning_rate: "0.00016"
2023-10-06 11:18:54,724 - mini_batch_size: "4"
2023-10-06 11:18:54,724 - max_epochs: "10"
2023-10-06 11:18:54,724 - shuffle: "True"
2023-10-06 11:18:54,724 ----------------------------------------------------------------------------------------------------
2023-10-06 11:18:54,724 Plugins:
2023-10-06 11:18:54,724 - TensorboardLogger
2023-10-06 11:18:54,724 - LinearScheduler | warmup_fraction: '0.1'
2023-10-06 11:18:54,724 ----------------------------------------------------------------------------------------------------
2023-10-06 11:18:54,724 Final evaluation on model from best epoch (best-model.pt)
2023-10-06 11:18:54,724 - metric: "('micro avg', 'f1-score')"
2023-10-06 11:18:54,724 ----------------------------------------------------------------------------------------------------
2023-10-06 11:18:54,724 Computation:
2023-10-06 11:18:54,724 - compute on device: cuda:0
2023-10-06 11:18:54,724 - embedding storage: none
2023-10-06 11:18:54,724 ----------------------------------------------------------------------------------------------------
2023-10-06 11:18:54,724 Model training base path: "hmbench-ajmc/en-hmbyt5-preliminary/byt5-small-historic-multilingual-span20-flax-bs4-wsFalse-e10-lr0.00016-poolingfirst-layers-1-crfFalse-2"
2023-10-06 11:18:54,725 ----------------------------------------------------------------------------------------------------
2023-10-06 11:18:54,725 ----------------------------------------------------------------------------------------------------
2023-10-06 11:18:54,725 Logging anything other than scalars to TensorBoard is currently not supported.
2023-10-06 11:19:06,277 epoch 1 - iter 30/304 - loss 3.23280502 - time (sec): 11.55 - samples/sec: 267.77 - lr: 0.000015 - momentum: 0.000000
2023-10-06 11:19:17,928 epoch 1 - iter 60/304 - loss 3.22267428 - time (sec): 23.20 - samples/sec: 265.71 - lr: 0.000031 - momentum: 0.000000
2023-10-06 11:19:29,792 epoch 1 - iter 90/304 - loss 3.20164486 - time (sec): 35.07 - samples/sec: 267.18 - lr: 0.000047 - momentum: 0.000000
2023-10-06 11:19:42,315 epoch 1 - iter 120/304 - loss 3.14013980 - time (sec): 47.59 - samples/sec: 267.71 - lr: 0.000063 - momentum: 0.000000
2023-10-06 11:19:54,285 epoch 1 - iter 150/304 - loss 3.04148856 - time (sec): 59.56 - samples/sec: 265.87 - lr: 0.000078 - momentum: 0.000000
2023-10-06 11:20:06,390 epoch 1 - iter 180/304 - loss 2.92038907 - time (sec): 71.66 - samples/sec: 265.36 - lr: 0.000094 - momentum: 0.000000
2023-10-06 11:20:17,808 epoch 1 - iter 210/304 - loss 2.80417689 - time (sec): 83.08 - samples/sec: 262.65 - lr: 0.000110 - momentum: 0.000000
2023-10-06 11:20:29,601 epoch 1 - iter 240/304 - loss 2.66758780 - time (sec): 94.88 - samples/sec: 261.89 - lr: 0.000126 - momentum: 0.000000
2023-10-06 11:20:40,947 epoch 1 - iter 270/304 - loss 2.53596647 - time (sec): 106.22 - samples/sec: 259.49 - lr: 0.000142 - momentum: 0.000000
2023-10-06 11:20:53,095 epoch 1 - iter 300/304 - loss 2.37932678 - time (sec): 118.37 - samples/sec: 259.27 - lr: 0.000157 - momentum: 0.000000
2023-10-06 11:20:54,412 ----------------------------------------------------------------------------------------------------
2023-10-06 11:20:54,413 EPOCH 1 done: loss 2.3668 - lr: 0.000157
2023-10-06 11:21:02,136 DEV : loss 0.8940157890319824 - f1-score (micro avg) 0.0
2023-10-06 11:21:02,142 ----------------------------------------------------------------------------------------------------
2023-10-06 11:21:14,199 epoch 2 - iter 30/304 - loss 0.84398850 - time (sec): 12.06 - samples/sec: 255.15 - lr: 0.000158 - momentum: 0.000000
2023-10-06 11:21:26,240 epoch 2 - iter 60/304 - loss 0.71960364 - time (sec): 24.10 - samples/sec: 251.12 - lr: 0.000157 - momentum: 0.000000
2023-10-06 11:21:38,825 epoch 2 - iter 90/304 - loss 0.69902046 - time (sec): 36.68 - samples/sec: 250.94 - lr: 0.000155 - momentum: 0.000000
2023-10-06 11:21:50,702 epoch 2 - iter 120/304 - loss 0.65342144 - time (sec): 48.56 - samples/sec: 251.76 - lr: 0.000153 - momentum: 0.000000
2023-10-06 11:22:02,399 epoch 2 - iter 150/304 - loss 0.61513172 - time (sec): 60.26 - samples/sec: 250.32 - lr: 0.000151 - momentum: 0.000000
2023-10-06 11:22:14,971 epoch 2 - iter 180/304 - loss 0.57390342 - time (sec): 72.83 - samples/sec: 251.94 - lr: 0.000150 - momentum: 0.000000
2023-10-06 11:22:27,394 epoch 2 - iter 210/304 - loss 0.53395295 - time (sec): 85.25 - samples/sec: 251.38 - lr: 0.000148 - momentum: 0.000000
2023-10-06 11:22:38,646 epoch 2 - iter 240/304 - loss 0.51499294 - time (sec): 96.50 - samples/sec: 251.41 - lr: 0.000146 - momentum: 0.000000
2023-10-06 11:22:50,525 epoch 2 - iter 270/304 - loss 0.49138422 - time (sec): 108.38 - samples/sec: 253.49 - lr: 0.000144 - momentum: 0.000000
2023-10-06 11:23:01,692 epoch 2 - iter 300/304 - loss 0.47629857 - time (sec): 119.55 - samples/sec: 255.71 - lr: 0.000143 - momentum: 0.000000
2023-10-06 11:23:03,107 ----------------------------------------------------------------------------------------------------
2023-10-06 11:23:03,107 EPOCH 2 done: loss 0.4747 - lr: 0.000143
2023-10-06 11:23:10,353 DEV : loss 0.3135835826396942 - f1-score (micro avg) 0.4828
2023-10-06 11:23:10,362 saving best model
2023-10-06 11:23:11,216 ----------------------------------------------------------------------------------------------------
2023-10-06 11:23:22,416 epoch 3 - iter 30/304 - loss 0.25890649 - time (sec): 11.20 - samples/sec: 275.67 - lr: 0.000141 - momentum: 0.000000
2023-10-06 11:23:33,344 epoch 3 - iter 60/304 - loss 0.23641783 - time (sec): 22.13 - samples/sec: 269.90 - lr: 0.000139 - momentum: 0.000000
2023-10-06 11:23:44,494 epoch 3 - iter 90/304 - loss 0.23071244 - time (sec): 33.28 - samples/sec: 267.13 - lr: 0.000137 - momentum: 0.000000
2023-10-06 11:23:57,028 epoch 3 - iter 120/304 - loss 0.22950828 - time (sec): 45.81 - samples/sec: 272.85 - lr: 0.000135 - momentum: 0.000000
2023-10-06 11:24:08,444 epoch 3 - iter 150/304 - loss 0.23056855 - time (sec): 57.23 - samples/sec: 272.32 - lr: 0.000134 - momentum: 0.000000
2023-10-06 11:24:19,960 epoch 3 - iter 180/304 - loss 0.21582432 - time (sec): 68.74 - samples/sec: 271.92 - lr: 0.000132 - momentum: 0.000000
2023-10-06 11:24:30,924 epoch 3 - iter 210/304 - loss 0.20806992 - time (sec): 79.71 - samples/sec: 270.87 - lr: 0.000130 - momentum: 0.000000
2023-10-06 11:24:42,169 epoch 3 - iter 240/304 - loss 0.20498504 - time (sec): 90.95 - samples/sec: 272.14 - lr: 0.000128 - momentum: 0.000000
2023-10-06 11:24:53,388 epoch 3 - iter 270/304 - loss 0.19803346 - time (sec): 102.17 - samples/sec: 271.72 - lr: 0.000127 - momentum: 0.000000
2023-10-06 11:25:04,448 epoch 3 - iter 300/304 - loss 0.19724389 - time (sec): 113.23 - samples/sec: 270.55 - lr: 0.000125 - momentum: 0.000000
2023-10-06 11:25:05,785 ----------------------------------------------------------------------------------------------------
2023-10-06 11:25:05,785 EPOCH 3 done: loss 0.1970 - lr: 0.000125
2023-10-06 11:25:12,953 DEV : loss 0.1839980185031891 - f1-score (micro avg) 0.6935
2023-10-06 11:25:12,962 saving best model
2023-10-06 11:25:17,292 ----------------------------------------------------------------------------------------------------
2023-10-06 11:25:28,279 epoch 4 - iter 30/304 - loss 0.12397772 - time (sec): 10.98 - samples/sec: 256.54 - lr: 0.000123 - momentum: 0.000000
2023-10-06 11:25:39,385 epoch 4 - iter 60/304 - loss 0.12952066 - time (sec): 22.09 - samples/sec: 258.98 - lr: 0.000121 - momentum: 0.000000
2023-10-06 11:25:50,956 epoch 4 - iter 90/304 - loss 0.12637612 - time (sec): 33.66 - samples/sec: 265.05 - lr: 0.000119 - momentum: 0.000000
2023-10-06 11:26:02,254 epoch 4 - iter 120/304 - loss 0.12132022 - time (sec): 44.96 - samples/sec: 265.79 - lr: 0.000118 - momentum: 0.000000
2023-10-06 11:26:14,424 epoch 4 - iter 150/304 - loss 0.12165008 - time (sec): 57.13 - samples/sec: 269.05 - lr: 0.000116 - momentum: 0.000000
2023-10-06 11:26:26,441 epoch 4 - iter 180/304 - loss 0.11739291 - time (sec): 69.15 - samples/sec: 271.36 - lr: 0.000114 - momentum: 0.000000
2023-10-06 11:26:37,875 epoch 4 - iter 210/304 - loss 0.11738931 - time (sec): 80.58 - samples/sec: 271.10 - lr: 0.000112 - momentum: 0.000000
2023-10-06 11:26:48,927 epoch 4 - iter 240/304 - loss 0.11367717 - time (sec): 91.63 - samples/sec: 269.89 - lr: 0.000111 - momentum: 0.000000
2023-10-06 11:27:00,083 epoch 4 - iter 270/304 - loss 0.11160862 - time (sec): 102.79 - samples/sec: 268.84 - lr: 0.000109 - momentum: 0.000000
2023-10-06 11:27:11,401 epoch 4 - iter 300/304 - loss 0.10767421 - time (sec): 114.11 - samples/sec: 267.63 - lr: 0.000107 - momentum: 0.000000
2023-10-06 11:27:12,975 ----------------------------------------------------------------------------------------------------
2023-10-06 11:27:12,976 EPOCH 4 done: loss 0.1083 - lr: 0.000107
2023-10-06 11:27:20,539 DEV : loss 0.14643196761608124 - f1-score (micro avg) 0.8098
2023-10-06 11:27:20,547 saving best model
2023-10-06 11:27:24,869 ----------------------------------------------------------------------------------------------------
2023-10-06 11:27:36,104 epoch 5 - iter 30/304 - loss 0.05893097 - time (sec): 11.23 - samples/sec: 265.28 - lr: 0.000105 - momentum: 0.000000
2023-10-06 11:27:47,508 epoch 5 - iter 60/304 - loss 0.06093974 - time (sec): 22.64 - samples/sec: 257.44 - lr: 0.000103 - momentum: 0.000000
2023-10-06 11:28:00,110 epoch 5 - iter 90/304 - loss 0.06490850 - time (sec): 35.24 - samples/sec: 263.51 - lr: 0.000102 - momentum: 0.000000
2023-10-06 11:28:12,105 epoch 5 - iter 120/304 - loss 0.06402128 - time (sec): 47.23 - samples/sec: 263.33 - lr: 0.000100 - momentum: 0.000000
2023-10-06 11:28:24,220 epoch 5 - iter 150/304 - loss 0.06709653 - time (sec): 59.35 - samples/sec: 263.14 - lr: 0.000098 - momentum: 0.000000
2023-10-06 11:28:36,465 epoch 5 - iter 180/304 - loss 0.06578819 - time (sec): 71.59 - samples/sec: 261.93 - lr: 0.000096 - momentum: 0.000000
2023-10-06 11:28:47,721 epoch 5 - iter 210/304 - loss 0.06346072 - time (sec): 82.85 - samples/sec: 257.61 - lr: 0.000094 - momentum: 0.000000
2023-10-06 11:28:59,136 epoch 5 - iter 240/304 - loss 0.06379421 - time (sec): 94.27 - samples/sec: 256.46 - lr: 0.000093 - momentum: 0.000000
2023-10-06 11:29:11,538 epoch 5 - iter 270/304 - loss 0.06374858 - time (sec): 106.67 - samples/sec: 256.63 - lr: 0.000091 - momentum: 0.000000
2023-10-06 11:29:23,757 epoch 5 - iter 300/304 - loss 0.06675049 - time (sec): 118.89 - samples/sec: 257.48 - lr: 0.000089 - momentum: 0.000000
2023-10-06 11:29:25,281 ----------------------------------------------------------------------------------------------------
2023-10-06 11:29:25,282 EPOCH 5 done: loss 0.0663 - lr: 0.000089
2023-10-06 11:29:33,343 DEV : loss 0.142156183719635 - f1-score (micro avg) 0.8237
2023-10-06 11:29:33,350 saving best model
2023-10-06 11:29:37,683 ----------------------------------------------------------------------------------------------------
2023-10-06 11:29:49,439 epoch 6 - iter 30/304 - loss 0.04719743 - time (sec): 11.75 - samples/sec: 250.80 - lr: 0.000087 - momentum: 0.000000
2023-10-06 11:30:01,514 epoch 6 - iter 60/304 - loss 0.05141651 - time (sec): 23.83 - samples/sec: 251.54 - lr: 0.000085 - momentum: 0.000000
2023-10-06 11:30:13,127 epoch 6 - iter 90/304 - loss 0.05337061 - time (sec): 35.44 - samples/sec: 252.16 - lr: 0.000084 - momentum: 0.000000
2023-10-06 11:30:25,134 epoch 6 - iter 120/304 - loss 0.05387818 - time (sec): 47.45 - samples/sec: 251.55 - lr: 0.000082 - momentum: 0.000000
2023-10-06 11:30:37,112 epoch 6 - iter 150/304 - loss 0.05061363 - time (sec): 59.43 - samples/sec: 252.88 - lr: 0.000080 - momentum: 0.000000
2023-10-06 11:30:49,479 epoch 6 - iter 180/304 - loss 0.05163472 - time (sec): 71.79 - samples/sec: 255.63 - lr: 0.000078 - momentum: 0.000000
2023-10-06 11:31:00,754 epoch 6 - iter 210/304 - loss 0.05441294 - time (sec): 83.07 - samples/sec: 256.26 - lr: 0.000077 - momentum: 0.000000
2023-10-06 11:31:12,972 epoch 6 - iter 240/304 - loss 0.04984931 - time (sec): 95.29 - samples/sec: 258.76 - lr: 0.000075 - momentum: 0.000000
2023-10-06 11:31:24,420 epoch 6 - iter 270/304 - loss 0.04806941 - time (sec): 106.73 - samples/sec: 259.32 - lr: 0.000073 - momentum: 0.000000
2023-10-06 11:31:35,735 epoch 6 - iter 300/304 - loss 0.04815844 - time (sec): 118.05 - samples/sec: 260.13 - lr: 0.000071 - momentum: 0.000000
2023-10-06 11:31:36,953 ----------------------------------------------------------------------------------------------------
2023-10-06 11:31:36,954 EPOCH 6 done: loss 0.0480 - lr: 0.000071
2023-10-06 11:31:43,942 DEV : loss 0.15544277429580688 - f1-score (micro avg) 0.8353
2023-10-06 11:31:43,948 saving best model
2023-10-06 11:31:48,274 ----------------------------------------------------------------------------------------------------
2023-10-06 11:31:59,658 epoch 7 - iter 30/304 - loss 0.04314236 - time (sec): 11.38 - samples/sec: 271.21 - lr: 0.000069 - momentum: 0.000000
2023-10-06 11:32:11,020 epoch 7 - iter 60/304 - loss 0.02656087 - time (sec): 22.74 - samples/sec: 273.22 - lr: 0.000068 - momentum: 0.000000
2023-10-06 11:32:22,477 epoch 7 - iter 90/304 - loss 0.03582258 - time (sec): 34.20 - samples/sec: 275.07 - lr: 0.000066 - momentum: 0.000000
2023-10-06 11:32:34,161 epoch 7 - iter 120/304 - loss 0.04110032 - time (sec): 45.88 - samples/sec: 273.62 - lr: 0.000064 - momentum: 0.000000
2023-10-06 11:32:44,970 epoch 7 - iter 150/304 - loss 0.04287401 - time (sec): 56.69 - samples/sec: 271.07 - lr: 0.000062 - momentum: 0.000000
2023-10-06 11:32:56,196 epoch 7 - iter 180/304 - loss 0.03931426 - time (sec): 67.92 - samples/sec: 269.55 - lr: 0.000061 - momentum: 0.000000
2023-10-06 11:33:07,609 epoch 7 - iter 210/304 - loss 0.03759564 - time (sec): 79.33 - samples/sec: 269.62 - lr: 0.000059 - momentum: 0.000000
2023-10-06 11:33:18,767 epoch 7 - iter 240/304 - loss 0.03621065 - time (sec): 90.49 - samples/sec: 268.77 - lr: 0.000057 - momentum: 0.000000
2023-10-06 11:33:30,772 epoch 7 - iter 270/304 - loss 0.03523973 - time (sec): 102.50 - samples/sec: 268.88 - lr: 0.000055 - momentum: 0.000000
2023-10-06 11:33:42,020 epoch 7 - iter 300/304 - loss 0.03705883 - time (sec): 113.74 - samples/sec: 269.06 - lr: 0.000054 - momentum: 0.000000
2023-10-06 11:33:43,436 ----------------------------------------------------------------------------------------------------
2023-10-06 11:33:43,437 EPOCH 7 done: loss 0.0367 - lr: 0.000054
2023-10-06 11:33:50,541 DEV : loss 0.1492297500371933 - f1-score (micro avg) 0.8435
2023-10-06 11:33:50,548 saving best model
2023-10-06 11:33:54,879 ----------------------------------------------------------------------------------------------------
2023-10-06 11:34:06,054 epoch 8 - iter 30/304 - loss 0.03218697 - time (sec): 11.17 - samples/sec: 258.75 - lr: 0.000052 - momentum: 0.000000
2023-10-06 11:34:17,624 epoch 8 - iter 60/304 - loss 0.02731734 - time (sec): 22.74 - samples/sec: 264.88 - lr: 0.000050 - momentum: 0.000000
2023-10-06 11:34:29,003 epoch 8 - iter 90/304 - loss 0.02188931 - time (sec): 34.12 - samples/sec: 267.10 - lr: 0.000048 - momentum: 0.000000
2023-10-06 11:34:40,978 epoch 8 - iter 120/304 - loss 0.02915435 - time (sec): 46.10 - samples/sec: 270.00 - lr: 0.000046 - momentum: 0.000000
2023-10-06 11:34:52,063 epoch 8 - iter 150/304 - loss 0.03216068 - time (sec): 57.18 - samples/sec: 270.89 - lr: 0.000045 - momentum: 0.000000
2023-10-06 11:35:03,065 epoch 8 - iter 180/304 - loss 0.03239949 - time (sec): 68.18 - samples/sec: 269.05 - lr: 0.000043 - momentum: 0.000000
2023-10-06 11:35:14,605 epoch 8 - iter 210/304 - loss 0.03143481 - time (sec): 79.72 - samples/sec: 270.36 - lr: 0.000041 - momentum: 0.000000
2023-10-06 11:35:25,798 epoch 8 - iter 240/304 - loss 0.03117235 - time (sec): 90.92 - samples/sec: 270.52 - lr: 0.000039 - momentum: 0.000000
2023-10-06 11:35:36,506 epoch 8 - iter 270/304 - loss 0.03155368 - time (sec): 101.62 - samples/sec: 269.96 - lr: 0.000038 - momentum: 0.000000
2023-10-06 11:35:48,073 epoch 8 - iter 300/304 - loss 0.02966872 - time (sec): 113.19 - samples/sec: 270.86 - lr: 0.000036 - momentum: 0.000000
2023-10-06 11:35:49,338 ----------------------------------------------------------------------------------------------------
2023-10-06 11:35:49,339 EPOCH 8 done: loss 0.0299 - lr: 0.000036
2023-10-06 11:35:56,469 DEV : loss 0.15260867774486542 - f1-score (micro avg) 0.8501
2023-10-06 11:35:56,478 saving best model
2023-10-06 11:36:00,804 ----------------------------------------------------------------------------------------------------
2023-10-06 11:36:12,364 epoch 9 - iter 30/304 - loss 0.01720815 - time (sec): 11.56 - samples/sec: 265.33 - lr: 0.000034 - momentum: 0.000000
2023-10-06 11:36:23,267 epoch 9 - iter 60/304 - loss 0.01566765 - time (sec): 22.46 - samples/sec: 267.25 - lr: 0.000032 - momentum: 0.000000
2023-10-06 11:36:34,334 epoch 9 - iter 90/304 - loss 0.01381731 - time (sec): 33.53 - samples/sec: 265.83 - lr: 0.000030 - momentum: 0.000000
2023-10-06 11:36:46,087 epoch 9 - iter 120/304 - loss 0.02016463 - time (sec): 45.28 - samples/sec: 269.22 - lr: 0.000029 - momentum: 0.000000
2023-10-06 11:36:57,288 epoch 9 - iter 150/304 - loss 0.02271172 - time (sec): 56.48 - samples/sec: 270.06 - lr: 0.000027 - momentum: 0.000000
2023-10-06 11:37:09,194 epoch 9 - iter 180/304 - loss 0.02697841 - time (sec): 68.39 - samples/sec: 273.42 - lr: 0.000025 - momentum: 0.000000
2023-10-06 11:37:20,137 epoch 9 - iter 210/304 - loss 0.02519416 - time (sec): 79.33 - samples/sec: 271.43 - lr: 0.000023 - momentum: 0.000000
2023-10-06 11:37:31,743 epoch 9 - iter 240/304 - loss 0.02668071 - time (sec): 90.94 - samples/sec: 271.28 - lr: 0.000022 - momentum: 0.000000
2023-10-06 11:37:42,866 epoch 9 - iter 270/304 - loss 0.02689309 - time (sec): 102.06 - samples/sec: 269.99 - lr: 0.000020 - momentum: 0.000000
2023-10-06 11:37:54,297 epoch 9 - iter 300/304 - loss 0.02592662 - time (sec): 113.49 - samples/sec: 269.72 - lr: 0.000018 - momentum: 0.000000
2023-10-06 11:37:55,755 ----------------------------------------------------------------------------------------------------
2023-10-06 11:37:55,755 EPOCH 9 done: loss 0.0260 - lr: 0.000018
2023-10-06 11:38:03,089 DEV : loss 0.15542393922805786 - f1-score (micro avg) 0.8548
2023-10-06 11:38:03,096 saving best model
2023-10-06 11:38:07,435 ----------------------------------------------------------------------------------------------------
2023-10-06 11:38:19,068 epoch 10 - iter 30/304 - loss 0.03378871 - time (sec): 11.63 - samples/sec: 261.03 - lr: 0.000016 - momentum: 0.000000
2023-10-06 11:38:30,437 epoch 10 - iter 60/304 - loss 0.02038193 - time (sec): 23.00 - samples/sec: 259.17 - lr: 0.000014 - momentum: 0.000000
2023-10-06 11:38:42,524 epoch 10 - iter 90/304 - loss 0.01635131 - time (sec): 35.09 - samples/sec: 263.31 - lr: 0.000013 - momentum: 0.000000
2023-10-06 11:38:54,185 epoch 10 - iter 120/304 - loss 0.02086686 - time (sec): 46.75 - samples/sec: 262.96 - lr: 0.000011 - momentum: 0.000000
2023-10-06 11:39:06,591 epoch 10 - iter 150/304 - loss 0.01965079 - time (sec): 59.15 - samples/sec: 264.95 - lr: 0.000009 - momentum: 0.000000
2023-10-06 11:39:18,554 epoch 10 - iter 180/304 - loss 0.01872377 - time (sec): 71.12 - samples/sec: 264.36 - lr: 0.000007 - momentum: 0.000000
2023-10-06 11:39:30,013 epoch 10 - iter 210/304 - loss 0.02004221 - time (sec): 82.58 - samples/sec: 264.22 - lr: 0.000006 - momentum: 0.000000
2023-10-06 11:39:41,814 epoch 10 - iter 240/304 - loss 0.02224598 - time (sec): 94.38 - samples/sec: 261.48 - lr: 0.000004 - momentum: 0.000000
2023-10-06 11:39:53,331 epoch 10 - iter 270/304 - loss 0.02227044 - time (sec): 105.89 - samples/sec: 260.31 - lr: 0.000002 - momentum: 0.000000
2023-10-06 11:40:05,369 epoch 10 - iter 300/304 - loss 0.02142454 - time (sec): 117.93 - samples/sec: 259.44 - lr: 0.000000 - momentum: 0.000000
2023-10-06 11:40:06,863 ----------------------------------------------------------------------------------------------------
2023-10-06 11:40:06,863 EPOCH 10 done: loss 0.0212 - lr: 0.000000
2023-10-06 11:40:14,816 DEV : loss 0.15741805732250214 - f1-score (micro avg) 0.8444
2023-10-06 11:40:15,690 ----------------------------------------------------------------------------------------------------
2023-10-06 11:40:15,691 Loading model from best epoch ...
2023-10-06 11:40:18,282 SequenceTagger predicts: Dictionary with 25 tags: O, S-scope, B-scope, E-scope, I-scope, S-pers, B-pers, E-pers, I-pers, S-work, B-work, E-work, I-work, S-loc, B-loc, E-loc, I-loc, S-date, B-date, E-date, I-date, S-object, B-object, E-object, I-object
2023-10-06 11:40:25,525
Results:
- F-score (micro) 0.8116
- F-score (macro) 0.6534
- Accuracy 0.6876
By class:
precision recall f1-score support
scope 0.7564 0.7815 0.7687 151
pers 0.7863 0.9583 0.8638 96
work 0.7981 0.8737 0.8342 95
loc 1.0000 0.6667 0.8000 3
date 0.0000 0.0000 0.0000 3
micro avg 0.7784 0.8477 0.8116 348
macro avg 0.6682 0.6560 0.6534 348
weighted avg 0.7716 0.8477 0.8065 348
2023-10-06 11:40:25,525 ----------------------------------------------------------------------------------------------------
|