File size: 25,329 Bytes
a0a4edb |
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 |
2023-10-10 21:12:02,183 ----------------------------------------------------------------------------------------------------
2023-10-10 21:12:02,186 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=17, bias=True)
(loss_function): CrossEntropyLoss()
)"
2023-10-10 21:12:02,186 ----------------------------------------------------------------------------------------------------
2023-10-10 21:12:02,186 MultiCorpus: 7142 train + 698 dev + 2570 test sentences
- NER_HIPE_2022 Corpus: 7142 train + 698 dev + 2570 test sentences - /root/.flair/datasets/ner_hipe_2022/v2.1/newseye/fr/with_doc_seperator
2023-10-10 21:12:02,186 ----------------------------------------------------------------------------------------------------
2023-10-10 21:12:02,186 Train: 7142 sentences
2023-10-10 21:12:02,186 (train_with_dev=False, train_with_test=False)
2023-10-10 21:12:02,186 ----------------------------------------------------------------------------------------------------
2023-10-10 21:12:02,186 Training Params:
2023-10-10 21:12:02,187 - learning_rate: "0.00016"
2023-10-10 21:12:02,187 - mini_batch_size: "8"
2023-10-10 21:12:02,187 - max_epochs: "10"
2023-10-10 21:12:02,187 - shuffle: "True"
2023-10-10 21:12:02,187 ----------------------------------------------------------------------------------------------------
2023-10-10 21:12:02,187 Plugins:
2023-10-10 21:12:02,187 - TensorboardLogger
2023-10-10 21:12:02,187 - LinearScheduler | warmup_fraction: '0.1'
2023-10-10 21:12:02,187 ----------------------------------------------------------------------------------------------------
2023-10-10 21:12:02,187 Final evaluation on model from best epoch (best-model.pt)
2023-10-10 21:12:02,187 - metric: "('micro avg', 'f1-score')"
2023-10-10 21:12:02,187 ----------------------------------------------------------------------------------------------------
2023-10-10 21:12:02,187 Computation:
2023-10-10 21:12:02,187 - compute on device: cuda:0
2023-10-10 21:12:02,188 - embedding storage: none
2023-10-10 21:12:02,188 ----------------------------------------------------------------------------------------------------
2023-10-10 21:12:02,188 Model training base path: "hmbench-newseye/fr-hmbyt5-preliminary/byt5-small-historic-multilingual-span20-flax-bs8-wsFalse-e10-lr0.00016-poolingfirst-layers-1-crfFalse-1"
2023-10-10 21:12:02,188 ----------------------------------------------------------------------------------------------------
2023-10-10 21:12:02,188 ----------------------------------------------------------------------------------------------------
2023-10-10 21:12:02,188 Logging anything other than scalars to TensorBoard is currently not supported.
2023-10-10 21:12:55,812 epoch 1 - iter 89/893 - loss 2.82831623 - time (sec): 53.62 - samples/sec: 475.14 - lr: 0.000016 - momentum: 0.000000
2023-10-10 21:13:49,028 epoch 1 - iter 178/893 - loss 2.77409191 - time (sec): 106.84 - samples/sec: 471.84 - lr: 0.000032 - momentum: 0.000000
2023-10-10 21:14:41,947 epoch 1 - iter 267/893 - loss 2.58893902 - time (sec): 159.76 - samples/sec: 469.74 - lr: 0.000048 - momentum: 0.000000
2023-10-10 21:15:34,481 epoch 1 - iter 356/893 - loss 2.35977449 - time (sec): 212.29 - samples/sec: 468.65 - lr: 0.000064 - momentum: 0.000000
2023-10-10 21:16:29,178 epoch 1 - iter 445/893 - loss 2.09834694 - time (sec): 266.99 - samples/sec: 470.82 - lr: 0.000080 - momentum: 0.000000
2023-10-10 21:17:21,664 epoch 1 - iter 534/893 - loss 1.88678651 - time (sec): 319.47 - samples/sec: 465.84 - lr: 0.000095 - momentum: 0.000000
2023-10-10 21:18:14,012 epoch 1 - iter 623/893 - loss 1.70594199 - time (sec): 371.82 - samples/sec: 463.66 - lr: 0.000111 - momentum: 0.000000
2023-10-10 21:19:05,643 epoch 1 - iter 712/893 - loss 1.54482663 - time (sec): 423.45 - samples/sec: 466.76 - lr: 0.000127 - momentum: 0.000000
2023-10-10 21:19:56,253 epoch 1 - iter 801/893 - loss 1.41640141 - time (sec): 474.06 - samples/sec: 471.26 - lr: 0.000143 - momentum: 0.000000
2023-10-10 21:20:46,274 epoch 1 - iter 890/893 - loss 1.31458924 - time (sec): 524.08 - samples/sec: 473.16 - lr: 0.000159 - momentum: 0.000000
2023-10-10 21:20:47,807 ----------------------------------------------------------------------------------------------------
2023-10-10 21:20:47,807 EPOCH 1 done: loss 1.3116 - lr: 0.000159
2023-10-10 21:21:07,697 DEV : loss 0.283738374710083 - f1-score (micro avg) 0.2694
2023-10-10 21:21:07,728 saving best model
2023-10-10 21:21:08,577 ----------------------------------------------------------------------------------------------------
2023-10-10 21:22:00,389 epoch 2 - iter 89/893 - loss 0.31517413 - time (sec): 51.81 - samples/sec: 510.50 - lr: 0.000158 - momentum: 0.000000
2023-10-10 21:22:49,744 epoch 2 - iter 178/893 - loss 0.30864388 - time (sec): 101.16 - samples/sec: 498.63 - lr: 0.000156 - momentum: 0.000000
2023-10-10 21:23:41,490 epoch 2 - iter 267/893 - loss 0.29140884 - time (sec): 152.91 - samples/sec: 485.83 - lr: 0.000155 - momentum: 0.000000
2023-10-10 21:24:33,788 epoch 2 - iter 356/893 - loss 0.27172420 - time (sec): 205.21 - samples/sec: 484.09 - lr: 0.000153 - momentum: 0.000000
2023-10-10 21:25:26,569 epoch 2 - iter 445/893 - loss 0.25547354 - time (sec): 257.99 - samples/sec: 481.26 - lr: 0.000151 - momentum: 0.000000
2023-10-10 21:26:18,173 epoch 2 - iter 534/893 - loss 0.24503363 - time (sec): 309.59 - samples/sec: 478.48 - lr: 0.000149 - momentum: 0.000000
2023-10-10 21:27:10,956 epoch 2 - iter 623/893 - loss 0.23162160 - time (sec): 362.38 - samples/sec: 478.84 - lr: 0.000148 - momentum: 0.000000
2023-10-10 21:28:04,603 epoch 2 - iter 712/893 - loss 0.22016367 - time (sec): 416.02 - samples/sec: 479.91 - lr: 0.000146 - momentum: 0.000000
2023-10-10 21:28:55,788 epoch 2 - iter 801/893 - loss 0.21046665 - time (sec): 467.21 - samples/sec: 478.78 - lr: 0.000144 - momentum: 0.000000
2023-10-10 21:29:46,943 epoch 2 - iter 890/893 - loss 0.20217199 - time (sec): 518.36 - samples/sec: 478.66 - lr: 0.000142 - momentum: 0.000000
2023-10-10 21:29:48,400 ----------------------------------------------------------------------------------------------------
2023-10-10 21:29:48,400 EPOCH 2 done: loss 0.2019 - lr: 0.000142
2023-10-10 21:30:10,827 DEV : loss 0.11071376502513885 - f1-score (micro avg) 0.7305
2023-10-10 21:30:10,860 saving best model
2023-10-10 21:30:19,632 ----------------------------------------------------------------------------------------------------
2023-10-10 21:31:11,573 epoch 3 - iter 89/893 - loss 0.09192063 - time (sec): 51.94 - samples/sec: 459.99 - lr: 0.000140 - momentum: 0.000000
2023-10-10 21:32:03,261 epoch 3 - iter 178/893 - loss 0.08757251 - time (sec): 103.62 - samples/sec: 476.56 - lr: 0.000139 - momentum: 0.000000
2023-10-10 21:32:55,633 epoch 3 - iter 267/893 - loss 0.08958475 - time (sec): 156.00 - samples/sec: 474.65 - lr: 0.000137 - momentum: 0.000000
2023-10-10 21:33:46,581 epoch 3 - iter 356/893 - loss 0.09204390 - time (sec): 206.94 - samples/sec: 470.62 - lr: 0.000135 - momentum: 0.000000
2023-10-10 21:34:40,061 epoch 3 - iter 445/893 - loss 0.09030135 - time (sec): 260.42 - samples/sec: 473.36 - lr: 0.000133 - momentum: 0.000000
2023-10-10 21:35:33,000 epoch 3 - iter 534/893 - loss 0.08756515 - time (sec): 313.36 - samples/sec: 472.67 - lr: 0.000132 - momentum: 0.000000
2023-10-10 21:36:28,115 epoch 3 - iter 623/893 - loss 0.08431353 - time (sec): 368.48 - samples/sec: 469.26 - lr: 0.000130 - momentum: 0.000000
2023-10-10 21:37:19,051 epoch 3 - iter 712/893 - loss 0.08334354 - time (sec): 419.41 - samples/sec: 473.01 - lr: 0.000128 - momentum: 0.000000
2023-10-10 21:38:10,924 epoch 3 - iter 801/893 - loss 0.08188831 - time (sec): 471.29 - samples/sec: 477.66 - lr: 0.000126 - momentum: 0.000000
2023-10-10 21:38:59,366 epoch 3 - iter 890/893 - loss 0.08186637 - time (sec): 519.73 - samples/sec: 477.20 - lr: 0.000125 - momentum: 0.000000
2023-10-10 21:39:00,986 ----------------------------------------------------------------------------------------------------
2023-10-10 21:39:00,986 EPOCH 3 done: loss 0.0818 - lr: 0.000125
2023-10-10 21:39:23,280 DEV : loss 0.1093674823641777 - f1-score (micro avg) 0.7573
2023-10-10 21:39:23,310 saving best model
2023-10-10 21:39:29,432 ----------------------------------------------------------------------------------------------------
2023-10-10 21:40:20,072 epoch 4 - iter 89/893 - loss 0.04796913 - time (sec): 50.64 - samples/sec: 492.08 - lr: 0.000123 - momentum: 0.000000
2023-10-10 21:41:10,758 epoch 4 - iter 178/893 - loss 0.05118746 - time (sec): 101.32 - samples/sec: 485.71 - lr: 0.000121 - momentum: 0.000000
2023-10-10 21:42:02,462 epoch 4 - iter 267/893 - loss 0.05230464 - time (sec): 153.03 - samples/sec: 482.23 - lr: 0.000119 - momentum: 0.000000
2023-10-10 21:42:54,950 epoch 4 - iter 356/893 - loss 0.05482020 - time (sec): 205.51 - samples/sec: 482.78 - lr: 0.000117 - momentum: 0.000000
2023-10-10 21:43:49,144 epoch 4 - iter 445/893 - loss 0.05300902 - time (sec): 259.71 - samples/sec: 483.40 - lr: 0.000116 - momentum: 0.000000
2023-10-10 21:44:41,491 epoch 4 - iter 534/893 - loss 0.05241872 - time (sec): 312.05 - samples/sec: 483.03 - lr: 0.000114 - momentum: 0.000000
2023-10-10 21:45:34,715 epoch 4 - iter 623/893 - loss 0.05109125 - time (sec): 365.28 - samples/sec: 484.73 - lr: 0.000112 - momentum: 0.000000
2023-10-10 21:46:26,148 epoch 4 - iter 712/893 - loss 0.05123523 - time (sec): 416.71 - samples/sec: 483.62 - lr: 0.000110 - momentum: 0.000000
2023-10-10 21:47:16,467 epoch 4 - iter 801/893 - loss 0.05153884 - time (sec): 467.03 - samples/sec: 481.68 - lr: 0.000109 - momentum: 0.000000
2023-10-10 21:48:05,530 epoch 4 - iter 890/893 - loss 0.05165032 - time (sec): 516.09 - samples/sec: 480.77 - lr: 0.000107 - momentum: 0.000000
2023-10-10 21:48:06,971 ----------------------------------------------------------------------------------------------------
2023-10-10 21:48:06,972 EPOCH 4 done: loss 0.0518 - lr: 0.000107
2023-10-10 21:48:29,692 DEV : loss 0.11296474188566208 - f1-score (micro avg) 0.782
2023-10-10 21:48:29,723 saving best model
2023-10-10 21:48:35,782 ----------------------------------------------------------------------------------------------------
2023-10-10 21:49:28,784 epoch 5 - iter 89/893 - loss 0.03597354 - time (sec): 53.00 - samples/sec: 479.80 - lr: 0.000105 - momentum: 0.000000
2023-10-10 21:50:20,053 epoch 5 - iter 178/893 - loss 0.03649107 - time (sec): 104.27 - samples/sec: 466.65 - lr: 0.000103 - momentum: 0.000000
2023-10-10 21:51:12,981 epoch 5 - iter 267/893 - loss 0.03580546 - time (sec): 157.19 - samples/sec: 473.76 - lr: 0.000101 - momentum: 0.000000
2023-10-10 21:52:06,072 epoch 5 - iter 356/893 - loss 0.03754465 - time (sec): 210.29 - samples/sec: 478.75 - lr: 0.000100 - momentum: 0.000000
2023-10-10 21:52:56,908 epoch 5 - iter 445/893 - loss 0.03813979 - time (sec): 261.12 - samples/sec: 472.86 - lr: 0.000098 - momentum: 0.000000
2023-10-10 21:53:47,123 epoch 5 - iter 534/893 - loss 0.03795036 - time (sec): 311.34 - samples/sec: 473.37 - lr: 0.000096 - momentum: 0.000000
2023-10-10 21:54:40,551 epoch 5 - iter 623/893 - loss 0.03843597 - time (sec): 364.76 - samples/sec: 472.83 - lr: 0.000094 - momentum: 0.000000
2023-10-10 21:55:32,349 epoch 5 - iter 712/893 - loss 0.03898741 - time (sec): 416.56 - samples/sec: 475.76 - lr: 0.000093 - momentum: 0.000000
2023-10-10 21:56:23,023 epoch 5 - iter 801/893 - loss 0.03832088 - time (sec): 467.24 - samples/sec: 477.46 - lr: 0.000091 - momentum: 0.000000
2023-10-10 21:57:12,653 epoch 5 - iter 890/893 - loss 0.03812875 - time (sec): 516.87 - samples/sec: 479.88 - lr: 0.000089 - momentum: 0.000000
2023-10-10 21:57:14,185 ----------------------------------------------------------------------------------------------------
2023-10-10 21:57:14,186 EPOCH 5 done: loss 0.0382 - lr: 0.000089
2023-10-10 21:57:36,202 DEV : loss 0.13481374084949493 - f1-score (micro avg) 0.7888
2023-10-10 21:57:36,234 saving best model
2023-10-10 21:57:45,079 ----------------------------------------------------------------------------------------------------
2023-10-10 21:58:34,561 epoch 6 - iter 89/893 - loss 0.02499157 - time (sec): 49.48 - samples/sec: 503.84 - lr: 0.000087 - momentum: 0.000000
2023-10-10 21:59:24,317 epoch 6 - iter 178/893 - loss 0.02831728 - time (sec): 99.23 - samples/sec: 499.20 - lr: 0.000085 - momentum: 0.000000
2023-10-10 22:00:14,930 epoch 6 - iter 267/893 - loss 0.02752569 - time (sec): 149.85 - samples/sec: 501.57 - lr: 0.000084 - momentum: 0.000000
2023-10-10 22:01:06,016 epoch 6 - iter 356/893 - loss 0.02810037 - time (sec): 200.93 - samples/sec: 493.70 - lr: 0.000082 - momentum: 0.000000
2023-10-10 22:01:56,477 epoch 6 - iter 445/893 - loss 0.02752164 - time (sec): 251.39 - samples/sec: 489.39 - lr: 0.000080 - momentum: 0.000000
2023-10-10 22:02:47,904 epoch 6 - iter 534/893 - loss 0.02786169 - time (sec): 302.82 - samples/sec: 488.32 - lr: 0.000078 - momentum: 0.000000
2023-10-10 22:03:40,128 epoch 6 - iter 623/893 - loss 0.02761400 - time (sec): 355.05 - samples/sec: 490.71 - lr: 0.000077 - momentum: 0.000000
2023-10-10 22:04:30,228 epoch 6 - iter 712/893 - loss 0.02792058 - time (sec): 405.15 - samples/sec: 491.45 - lr: 0.000075 - momentum: 0.000000
2023-10-10 22:05:21,277 epoch 6 - iter 801/893 - loss 0.02858658 - time (sec): 456.19 - samples/sec: 492.50 - lr: 0.000073 - momentum: 0.000000
2023-10-10 22:06:11,127 epoch 6 - iter 890/893 - loss 0.02899685 - time (sec): 506.04 - samples/sec: 490.13 - lr: 0.000071 - momentum: 0.000000
2023-10-10 22:06:12,768 ----------------------------------------------------------------------------------------------------
2023-10-10 22:06:12,769 EPOCH 6 done: loss 0.0289 - lr: 0.000071
2023-10-10 22:06:34,365 DEV : loss 0.16970570385456085 - f1-score (micro avg) 0.7684
2023-10-10 22:06:34,396 ----------------------------------------------------------------------------------------------------
2023-10-10 22:07:25,120 epoch 7 - iter 89/893 - loss 0.01834186 - time (sec): 50.72 - samples/sec: 500.26 - lr: 0.000069 - momentum: 0.000000
2023-10-10 22:08:14,621 epoch 7 - iter 178/893 - loss 0.02010496 - time (sec): 100.22 - samples/sec: 486.08 - lr: 0.000068 - momentum: 0.000000
2023-10-10 22:09:05,764 epoch 7 - iter 267/893 - loss 0.02031563 - time (sec): 151.37 - samples/sec: 490.13 - lr: 0.000066 - momentum: 0.000000
2023-10-10 22:09:56,686 epoch 7 - iter 356/893 - loss 0.02131291 - time (sec): 202.29 - samples/sec: 489.17 - lr: 0.000064 - momentum: 0.000000
2023-10-10 22:10:46,378 epoch 7 - iter 445/893 - loss 0.02126158 - time (sec): 251.98 - samples/sec: 487.98 - lr: 0.000062 - momentum: 0.000000
2023-10-10 22:11:36,949 epoch 7 - iter 534/893 - loss 0.02104177 - time (sec): 302.55 - samples/sec: 490.11 - lr: 0.000061 - momentum: 0.000000
2023-10-10 22:12:28,625 epoch 7 - iter 623/893 - loss 0.02189809 - time (sec): 354.23 - samples/sec: 488.81 - lr: 0.000059 - momentum: 0.000000
2023-10-10 22:13:19,313 epoch 7 - iter 712/893 - loss 0.02158960 - time (sec): 404.91 - samples/sec: 485.19 - lr: 0.000057 - momentum: 0.000000
2023-10-10 22:14:11,525 epoch 7 - iter 801/893 - loss 0.02192818 - time (sec): 457.13 - samples/sec: 487.63 - lr: 0.000055 - momentum: 0.000000
2023-10-10 22:15:02,258 epoch 7 - iter 890/893 - loss 0.02245709 - time (sec): 507.86 - samples/sec: 488.52 - lr: 0.000053 - momentum: 0.000000
2023-10-10 22:15:03,904 ----------------------------------------------------------------------------------------------------
2023-10-10 22:15:03,904 EPOCH 7 done: loss 0.0224 - lr: 0.000053
2023-10-10 22:15:27,246 DEV : loss 0.16878585517406464 - f1-score (micro avg) 0.781
2023-10-10 22:15:27,280 ----------------------------------------------------------------------------------------------------
2023-10-10 22:16:19,605 epoch 8 - iter 89/893 - loss 0.01907382 - time (sec): 52.32 - samples/sec: 469.29 - lr: 0.000052 - momentum: 0.000000
2023-10-10 22:17:10,982 epoch 8 - iter 178/893 - loss 0.01745040 - time (sec): 103.70 - samples/sec: 467.76 - lr: 0.000050 - momentum: 0.000000
2023-10-10 22:18:02,971 epoch 8 - iter 267/893 - loss 0.01934552 - time (sec): 155.69 - samples/sec: 462.81 - lr: 0.000048 - momentum: 0.000000
2023-10-10 22:18:55,894 epoch 8 - iter 356/893 - loss 0.01839673 - time (sec): 208.61 - samples/sec: 470.60 - lr: 0.000046 - momentum: 0.000000
2023-10-10 22:19:49,404 epoch 8 - iter 445/893 - loss 0.01773640 - time (sec): 262.12 - samples/sec: 468.09 - lr: 0.000045 - momentum: 0.000000
2023-10-10 22:20:42,717 epoch 8 - iter 534/893 - loss 0.01726269 - time (sec): 315.44 - samples/sec: 462.59 - lr: 0.000043 - momentum: 0.000000
2023-10-10 22:21:35,308 epoch 8 - iter 623/893 - loss 0.01741947 - time (sec): 368.03 - samples/sec: 464.14 - lr: 0.000041 - momentum: 0.000000
2023-10-10 22:22:26,901 epoch 8 - iter 712/893 - loss 0.01689766 - time (sec): 419.62 - samples/sec: 465.07 - lr: 0.000039 - momentum: 0.000000
2023-10-10 22:23:20,021 epoch 8 - iter 801/893 - loss 0.01726890 - time (sec): 472.74 - samples/sec: 467.86 - lr: 0.000037 - momentum: 0.000000
2023-10-10 22:24:13,002 epoch 8 - iter 890/893 - loss 0.01701346 - time (sec): 525.72 - samples/sec: 471.25 - lr: 0.000036 - momentum: 0.000000
2023-10-10 22:24:14,799 ----------------------------------------------------------------------------------------------------
2023-10-10 22:24:14,799 EPOCH 8 done: loss 0.0171 - lr: 0.000036
2023-10-10 22:24:38,300 DEV : loss 0.183110311627388 - f1-score (micro avg) 0.7858
2023-10-10 22:24:38,331 ----------------------------------------------------------------------------------------------------
2023-10-10 22:25:29,529 epoch 9 - iter 89/893 - loss 0.01593252 - time (sec): 51.20 - samples/sec: 486.33 - lr: 0.000034 - momentum: 0.000000
2023-10-10 22:26:22,330 epoch 9 - iter 178/893 - loss 0.01574775 - time (sec): 104.00 - samples/sec: 469.56 - lr: 0.000032 - momentum: 0.000000
2023-10-10 22:27:13,972 epoch 9 - iter 267/893 - loss 0.01645156 - time (sec): 155.64 - samples/sec: 480.85 - lr: 0.000030 - momentum: 0.000000
2023-10-10 22:28:04,643 epoch 9 - iter 356/893 - loss 0.01555361 - time (sec): 206.31 - samples/sec: 473.98 - lr: 0.000029 - momentum: 0.000000
2023-10-10 22:28:57,093 epoch 9 - iter 445/893 - loss 0.01523364 - time (sec): 258.76 - samples/sec: 470.80 - lr: 0.000027 - momentum: 0.000000
2023-10-10 22:29:48,190 epoch 9 - iter 534/893 - loss 0.01528624 - time (sec): 309.86 - samples/sec: 471.86 - lr: 0.000025 - momentum: 0.000000
2023-10-10 22:30:38,406 epoch 9 - iter 623/893 - loss 0.01463531 - time (sec): 360.07 - samples/sec: 473.08 - lr: 0.000023 - momentum: 0.000000
2023-10-10 22:31:30,450 epoch 9 - iter 712/893 - loss 0.01411081 - time (sec): 412.12 - samples/sec: 475.02 - lr: 0.000022 - momentum: 0.000000
2023-10-10 22:32:22,181 epoch 9 - iter 801/893 - loss 0.01404007 - time (sec): 463.85 - samples/sec: 476.81 - lr: 0.000020 - momentum: 0.000000
2023-10-10 22:33:14,754 epoch 9 - iter 890/893 - loss 0.01366814 - time (sec): 516.42 - samples/sec: 480.08 - lr: 0.000018 - momentum: 0.000000
2023-10-10 22:33:16,429 ----------------------------------------------------------------------------------------------------
2023-10-10 22:33:16,430 EPOCH 9 done: loss 0.0137 - lr: 0.000018
2023-10-10 22:33:39,324 DEV : loss 0.19475506246089935 - f1-score (micro avg) 0.7882
2023-10-10 22:33:39,354 ----------------------------------------------------------------------------------------------------
2023-10-10 22:34:30,591 epoch 10 - iter 89/893 - loss 0.01252786 - time (sec): 51.23 - samples/sec: 492.58 - lr: 0.000016 - momentum: 0.000000
2023-10-10 22:35:22,707 epoch 10 - iter 178/893 - loss 0.01292130 - time (sec): 103.35 - samples/sec: 476.58 - lr: 0.000014 - momentum: 0.000000
2023-10-10 22:36:14,189 epoch 10 - iter 267/893 - loss 0.01258759 - time (sec): 154.83 - samples/sec: 466.66 - lr: 0.000013 - momentum: 0.000000
2023-10-10 22:37:09,377 epoch 10 - iter 356/893 - loss 0.01217698 - time (sec): 210.02 - samples/sec: 469.59 - lr: 0.000011 - momentum: 0.000000
2023-10-10 22:38:01,970 epoch 10 - iter 445/893 - loss 0.01170757 - time (sec): 262.61 - samples/sec: 475.29 - lr: 0.000009 - momentum: 0.000000
2023-10-10 22:38:54,782 epoch 10 - iter 534/893 - loss 0.01179636 - time (sec): 315.43 - samples/sec: 471.41 - lr: 0.000007 - momentum: 0.000000
2023-10-10 22:39:47,080 epoch 10 - iter 623/893 - loss 0.01205257 - time (sec): 367.72 - samples/sec: 476.23 - lr: 0.000006 - momentum: 0.000000
2023-10-10 22:40:38,533 epoch 10 - iter 712/893 - loss 0.01244404 - time (sec): 419.18 - samples/sec: 473.93 - lr: 0.000004 - momentum: 0.000000
2023-10-10 22:41:28,980 epoch 10 - iter 801/893 - loss 0.01197188 - time (sec): 469.62 - samples/sec: 473.93 - lr: 0.000002 - momentum: 0.000000
2023-10-10 22:42:22,734 epoch 10 - iter 890/893 - loss 0.01189933 - time (sec): 523.38 - samples/sec: 473.94 - lr: 0.000000 - momentum: 0.000000
2023-10-10 22:42:24,281 ----------------------------------------------------------------------------------------------------
2023-10-10 22:42:24,282 EPOCH 10 done: loss 0.0119 - lr: 0.000000
2023-10-10 22:42:46,904 DEV : loss 0.20220361649990082 - f1-score (micro avg) 0.7859
2023-10-10 22:42:47,805 ----------------------------------------------------------------------------------------------------
2023-10-10 22:42:47,807 Loading model from best epoch ...
2023-10-10 22:42:52,366 SequenceTagger predicts: Dictionary with 17 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, S-HumanProd, B-HumanProd, E-HumanProd, I-HumanProd
2023-10-10 22:44:04,150
Results:
- F-score (micro) 0.6943
- F-score (macro) 0.6066
- Accuracy 0.5472
By class:
precision recall f1-score support
LOC 0.7046 0.7014 0.7030 1095
PER 0.7580 0.7737 0.7658 1012
ORG 0.4698 0.5658 0.5133 357
HumanProd 0.3509 0.6061 0.4444 33
micro avg 0.6793 0.7101 0.6943 2497
macro avg 0.5708 0.6617 0.6066 2497
weighted avg 0.6880 0.7101 0.6979 2497
2023-10-10 22:44:04,150 ----------------------------------------------------------------------------------------------------
|