File size: 25,444 Bytes
d75353b |
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 |
2023-10-14 06:00:02,910 ----------------------------------------------------------------------------------------------------
2023-10-14 06:00:02,912 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()
)"
2023-10-14 06:00:02,912 ----------------------------------------------------------------------------------------------------
2023-10-14 06:00:02,912 MultiCorpus: 14465 train + 1392 dev + 2432 test sentences
- NER_HIPE_2022 Corpus: 14465 train + 1392 dev + 2432 test sentences - /root/.flair/datasets/ner_hipe_2022/v2.1/letemps/fr/with_doc_seperator
2023-10-14 06:00:02,913 ----------------------------------------------------------------------------------------------------
2023-10-14 06:00:02,913 Train: 14465 sentences
2023-10-14 06:00:02,913 (train_with_dev=False, train_with_test=False)
2023-10-14 06:00:02,913 ----------------------------------------------------------------------------------------------------
2023-10-14 06:00:02,913 Training Params:
2023-10-14 06:00:02,913 - learning_rate: "0.00016"
2023-10-14 06:00:02,913 - mini_batch_size: "4"
2023-10-14 06:00:02,913 - max_epochs: "10"
2023-10-14 06:00:02,913 - shuffle: "True"
2023-10-14 06:00:02,913 ----------------------------------------------------------------------------------------------------
2023-10-14 06:00:02,913 Plugins:
2023-10-14 06:00:02,913 - TensorboardLogger
2023-10-14 06:00:02,913 - LinearScheduler | warmup_fraction: '0.1'
2023-10-14 06:00:02,913 ----------------------------------------------------------------------------------------------------
2023-10-14 06:00:02,913 Final evaluation on model from best epoch (best-model.pt)
2023-10-14 06:00:02,914 - metric: "('micro avg', 'f1-score')"
2023-10-14 06:00:02,914 ----------------------------------------------------------------------------------------------------
2023-10-14 06:00:02,914 Computation:
2023-10-14 06:00:02,914 - compute on device: cuda:0
2023-10-14 06:00:02,914 - embedding storage: none
2023-10-14 06:00:02,914 ----------------------------------------------------------------------------------------------------
2023-10-14 06:00:02,914 Model training base path: "hmbench-letemps/fr-hmbyt5-preliminary/byt5-small-historic-multilingual-span20-flax-bs4-wsFalse-e10-lr0.00016-poolingfirst-layers-1-crfFalse-3"
2023-10-14 06:00:02,914 ----------------------------------------------------------------------------------------------------
2023-10-14 06:00:02,914 ----------------------------------------------------------------------------------------------------
2023-10-14 06:00:02,914 Logging anything other than scalars to TensorBoard is currently not supported.
2023-10-14 06:01:53,445 epoch 1 - iter 361/3617 - loss 2.47239461 - time (sec): 110.53 - samples/sec: 341.23 - lr: 0.000016 - momentum: 0.000000
2023-10-14 06:03:37,266 epoch 1 - iter 722/3617 - loss 2.07484525 - time (sec): 214.35 - samples/sec: 347.51 - lr: 0.000032 - momentum: 0.000000
2023-10-14 06:05:25,472 epoch 1 - iter 1083/3617 - loss 1.61456832 - time (sec): 322.56 - samples/sec: 350.12 - lr: 0.000048 - momentum: 0.000000
2023-10-14 06:07:08,611 epoch 1 - iter 1444/3617 - loss 1.27675526 - time (sec): 425.69 - samples/sec: 355.51 - lr: 0.000064 - momentum: 0.000000
2023-10-14 06:08:50,479 epoch 1 - iter 1805/3617 - loss 1.05993658 - time (sec): 527.56 - samples/sec: 358.18 - lr: 0.000080 - momentum: 0.000000
2023-10-14 06:10:29,696 epoch 1 - iter 2166/3617 - loss 0.91231426 - time (sec): 626.78 - samples/sec: 361.53 - lr: 0.000096 - momentum: 0.000000
2023-10-14 06:12:09,104 epoch 1 - iter 2527/3617 - loss 0.80617937 - time (sec): 726.19 - samples/sec: 362.44 - lr: 0.000112 - momentum: 0.000000
2023-10-14 06:13:53,682 epoch 1 - iter 2888/3617 - loss 0.72280386 - time (sec): 830.77 - samples/sec: 362.49 - lr: 0.000128 - momentum: 0.000000
2023-10-14 06:15:37,793 epoch 1 - iter 3249/3617 - loss 0.65137338 - time (sec): 934.88 - samples/sec: 364.42 - lr: 0.000144 - momentum: 0.000000
2023-10-14 06:17:19,931 epoch 1 - iter 3610/3617 - loss 0.59733239 - time (sec): 1037.01 - samples/sec: 365.74 - lr: 0.000160 - momentum: 0.000000
2023-10-14 06:17:21,610 ----------------------------------------------------------------------------------------------------
2023-10-14 06:17:21,611 EPOCH 1 done: loss 0.5965 - lr: 0.000160
2023-10-14 06:18:02,444 DEV : loss 0.12650439143180847 - f1-score (micro avg) 0.6067
2023-10-14 06:18:02,513 saving best model
2023-10-14 06:18:03,590 ----------------------------------------------------------------------------------------------------
2023-10-14 06:19:58,085 epoch 2 - iter 361/3617 - loss 0.09250931 - time (sec): 114.49 - samples/sec: 331.01 - lr: 0.000158 - momentum: 0.000000
2023-10-14 06:21:42,643 epoch 2 - iter 722/3617 - loss 0.09407297 - time (sec): 219.05 - samples/sec: 342.51 - lr: 0.000156 - momentum: 0.000000
2023-10-14 06:23:32,345 epoch 2 - iter 1083/3617 - loss 0.09368204 - time (sec): 328.75 - samples/sec: 353.59 - lr: 0.000155 - momentum: 0.000000
2023-10-14 06:25:14,764 epoch 2 - iter 1444/3617 - loss 0.09240527 - time (sec): 431.17 - samples/sec: 359.43 - lr: 0.000153 - momentum: 0.000000
2023-10-14 06:26:54,551 epoch 2 - iter 1805/3617 - loss 0.09322341 - time (sec): 530.96 - samples/sec: 363.88 - lr: 0.000151 - momentum: 0.000000
2023-10-14 06:28:32,658 epoch 2 - iter 2166/3617 - loss 0.09164435 - time (sec): 629.06 - samples/sec: 365.00 - lr: 0.000149 - momentum: 0.000000
2023-10-14 06:30:13,945 epoch 2 - iter 2527/3617 - loss 0.09049817 - time (sec): 730.35 - samples/sec: 365.65 - lr: 0.000148 - momentum: 0.000000
2023-10-14 06:31:53,651 epoch 2 - iter 2888/3617 - loss 0.09055334 - time (sec): 830.06 - samples/sec: 366.60 - lr: 0.000146 - momentum: 0.000000
2023-10-14 06:33:36,995 epoch 2 - iter 3249/3617 - loss 0.09003575 - time (sec): 933.40 - samples/sec: 366.95 - lr: 0.000144 - momentum: 0.000000
2023-10-14 06:35:21,055 epoch 2 - iter 3610/3617 - loss 0.08946027 - time (sec): 1037.46 - samples/sec: 365.54 - lr: 0.000142 - momentum: 0.000000
2023-10-14 06:35:22,984 ----------------------------------------------------------------------------------------------------
2023-10-14 06:35:22,984 EPOCH 2 done: loss 0.0894 - lr: 0.000142
2023-10-14 06:36:03,645 DEV : loss 0.11830706894397736 - f1-score (micro avg) 0.6257
2023-10-14 06:36:03,706 saving best model
2023-10-14 06:36:09,241 ----------------------------------------------------------------------------------------------------
2023-10-14 06:37:52,453 epoch 3 - iter 361/3617 - loss 0.06677874 - time (sec): 103.21 - samples/sec: 362.87 - lr: 0.000140 - momentum: 0.000000
2023-10-14 06:39:35,134 epoch 3 - iter 722/3617 - loss 0.06628824 - time (sec): 205.89 - samples/sec: 374.82 - lr: 0.000139 - momentum: 0.000000
2023-10-14 06:41:18,032 epoch 3 - iter 1083/3617 - loss 0.06459732 - time (sec): 308.79 - samples/sec: 372.39 - lr: 0.000137 - momentum: 0.000000
2023-10-14 06:43:02,845 epoch 3 - iter 1444/3617 - loss 0.06301256 - time (sec): 413.60 - samples/sec: 369.24 - lr: 0.000135 - momentum: 0.000000
2023-10-14 06:44:48,454 epoch 3 - iter 1805/3617 - loss 0.06356181 - time (sec): 519.21 - samples/sec: 370.00 - lr: 0.000133 - momentum: 0.000000
2023-10-14 06:46:34,140 epoch 3 - iter 2166/3617 - loss 0.06323142 - time (sec): 624.90 - samples/sec: 368.82 - lr: 0.000132 - momentum: 0.000000
2023-10-14 06:48:23,285 epoch 3 - iter 2527/3617 - loss 0.06443387 - time (sec): 734.04 - samples/sec: 363.27 - lr: 0.000130 - momentum: 0.000000
2023-10-14 06:50:15,844 epoch 3 - iter 2888/3617 - loss 0.06348206 - time (sec): 846.60 - samples/sec: 358.62 - lr: 0.000128 - momentum: 0.000000
2023-10-14 06:52:05,894 epoch 3 - iter 3249/3617 - loss 0.06319253 - time (sec): 956.65 - samples/sec: 355.86 - lr: 0.000126 - momentum: 0.000000
2023-10-14 06:53:56,573 epoch 3 - iter 3610/3617 - loss 0.06352064 - time (sec): 1067.33 - samples/sec: 355.17 - lr: 0.000124 - momentum: 0.000000
2023-10-14 06:53:58,615 ----------------------------------------------------------------------------------------------------
2023-10-14 06:53:58,615 EPOCH 3 done: loss 0.0634 - lr: 0.000124
2023-10-14 06:54:40,585 DEV : loss 0.18002015352249146 - f1-score (micro avg) 0.6241
2023-10-14 06:54:40,652 ----------------------------------------------------------------------------------------------------
2023-10-14 06:56:25,424 epoch 4 - iter 361/3617 - loss 0.04589010 - time (sec): 104.77 - samples/sec: 349.69 - lr: 0.000123 - momentum: 0.000000
2023-10-14 06:58:11,340 epoch 4 - iter 722/3617 - loss 0.04524784 - time (sec): 210.69 - samples/sec: 357.08 - lr: 0.000121 - momentum: 0.000000
2023-10-14 07:00:01,015 epoch 4 - iter 1083/3617 - loss 0.04459280 - time (sec): 320.36 - samples/sec: 352.15 - lr: 0.000119 - momentum: 0.000000
2023-10-14 07:01:47,053 epoch 4 - iter 1444/3617 - loss 0.04241941 - time (sec): 426.40 - samples/sec: 351.10 - lr: 0.000117 - momentum: 0.000000
2023-10-14 07:03:29,823 epoch 4 - iter 1805/3617 - loss 0.04232486 - time (sec): 529.17 - samples/sec: 354.82 - lr: 0.000116 - momentum: 0.000000
2023-10-14 07:05:12,160 epoch 4 - iter 2166/3617 - loss 0.04384413 - time (sec): 631.51 - samples/sec: 359.79 - lr: 0.000114 - momentum: 0.000000
2023-10-14 07:06:59,230 epoch 4 - iter 2527/3617 - loss 0.04460396 - time (sec): 738.58 - samples/sec: 360.08 - lr: 0.000112 - momentum: 0.000000
2023-10-14 07:08:41,529 epoch 4 - iter 2888/3617 - loss 0.04544588 - time (sec): 840.88 - samples/sec: 360.44 - lr: 0.000110 - momentum: 0.000000
2023-10-14 07:10:25,972 epoch 4 - iter 3249/3617 - loss 0.04605584 - time (sec): 945.32 - samples/sec: 361.87 - lr: 0.000108 - momentum: 0.000000
2023-10-14 07:12:06,260 epoch 4 - iter 3610/3617 - loss 0.04631529 - time (sec): 1045.61 - samples/sec: 362.78 - lr: 0.000107 - momentum: 0.000000
2023-10-14 07:12:07,916 ----------------------------------------------------------------------------------------------------
2023-10-14 07:12:07,916 EPOCH 4 done: loss 0.0463 - lr: 0.000107
2023-10-14 07:12:47,955 DEV : loss 0.2199789583683014 - f1-score (micro avg) 0.6424
2023-10-14 07:12:48,014 saving best model
2023-10-14 07:12:49,081 ----------------------------------------------------------------------------------------------------
2023-10-14 07:14:30,776 epoch 5 - iter 361/3617 - loss 0.02732386 - time (sec): 101.69 - samples/sec: 378.30 - lr: 0.000105 - momentum: 0.000000
2023-10-14 07:16:19,492 epoch 5 - iter 722/3617 - loss 0.02918770 - time (sec): 210.41 - samples/sec: 363.23 - lr: 0.000103 - momentum: 0.000000
2023-10-14 07:18:11,252 epoch 5 - iter 1083/3617 - loss 0.03233304 - time (sec): 322.17 - samples/sec: 352.73 - lr: 0.000101 - momentum: 0.000000
2023-10-14 07:20:01,023 epoch 5 - iter 1444/3617 - loss 0.03284681 - time (sec): 431.94 - samples/sec: 347.83 - lr: 0.000100 - momentum: 0.000000
2023-10-14 07:21:44,113 epoch 5 - iter 1805/3617 - loss 0.03214533 - time (sec): 535.03 - samples/sec: 353.10 - lr: 0.000098 - momentum: 0.000000
2023-10-14 07:23:24,045 epoch 5 - iter 2166/3617 - loss 0.03217787 - time (sec): 634.96 - samples/sec: 356.66 - lr: 0.000096 - momentum: 0.000000
2023-10-14 07:25:04,587 epoch 5 - iter 2527/3617 - loss 0.03195592 - time (sec): 735.50 - samples/sec: 357.47 - lr: 0.000094 - momentum: 0.000000
2023-10-14 07:26:45,009 epoch 5 - iter 2888/3617 - loss 0.03247486 - time (sec): 835.93 - samples/sec: 359.86 - lr: 0.000092 - momentum: 0.000000
2023-10-14 07:28:30,143 epoch 5 - iter 3249/3617 - loss 0.03317830 - time (sec): 941.06 - samples/sec: 361.44 - lr: 0.000091 - momentum: 0.000000
2023-10-14 07:30:10,900 epoch 5 - iter 3610/3617 - loss 0.03266682 - time (sec): 1041.82 - samples/sec: 364.12 - lr: 0.000089 - momentum: 0.000000
2023-10-14 07:30:12,602 ----------------------------------------------------------------------------------------------------
2023-10-14 07:30:12,603 EPOCH 5 done: loss 0.0327 - lr: 0.000089
2023-10-14 07:30:54,210 DEV : loss 0.24780842661857605 - f1-score (micro avg) 0.6605
2023-10-14 07:30:54,277 saving best model
2023-10-14 07:30:59,067 ----------------------------------------------------------------------------------------------------
2023-10-14 07:32:41,778 epoch 6 - iter 361/3617 - loss 0.01937433 - time (sec): 102.70 - samples/sec: 382.13 - lr: 0.000087 - momentum: 0.000000
2023-10-14 07:34:22,211 epoch 6 - iter 722/3617 - loss 0.01983213 - time (sec): 203.13 - samples/sec: 379.33 - lr: 0.000085 - momentum: 0.000000
2023-10-14 07:36:00,973 epoch 6 - iter 1083/3617 - loss 0.02234270 - time (sec): 301.89 - samples/sec: 379.02 - lr: 0.000084 - momentum: 0.000000
2023-10-14 07:37:43,093 epoch 6 - iter 1444/3617 - loss 0.02403649 - time (sec): 404.01 - samples/sec: 373.77 - lr: 0.000082 - momentum: 0.000000
2023-10-14 07:39:32,914 epoch 6 - iter 1805/3617 - loss 0.02399559 - time (sec): 513.83 - samples/sec: 365.51 - lr: 0.000080 - momentum: 0.000000
2023-10-14 07:41:13,200 epoch 6 - iter 2166/3617 - loss 0.02304717 - time (sec): 614.12 - samples/sec: 366.92 - lr: 0.000078 - momentum: 0.000000
2023-10-14 07:42:57,050 epoch 6 - iter 2527/3617 - loss 0.02261731 - time (sec): 717.97 - samples/sec: 369.09 - lr: 0.000076 - momentum: 0.000000
2023-10-14 07:44:39,865 epoch 6 - iter 2888/3617 - loss 0.02304199 - time (sec): 820.78 - samples/sec: 370.40 - lr: 0.000075 - momentum: 0.000000
2023-10-14 07:46:20,394 epoch 6 - iter 3249/3617 - loss 0.02338086 - time (sec): 921.31 - samples/sec: 369.43 - lr: 0.000073 - momentum: 0.000000
2023-10-14 07:48:04,126 epoch 6 - iter 3610/3617 - loss 0.02327186 - time (sec): 1025.04 - samples/sec: 369.81 - lr: 0.000071 - momentum: 0.000000
2023-10-14 07:48:06,205 ----------------------------------------------------------------------------------------------------
2023-10-14 07:48:06,205 EPOCH 6 done: loss 0.0232 - lr: 0.000071
2023-10-14 07:48:44,950 DEV : loss 0.2599741816520691 - f1-score (micro avg) 0.6625
2023-10-14 07:48:45,008 saving best model
2023-10-14 07:48:46,022 ----------------------------------------------------------------------------------------------------
2023-10-14 07:50:26,907 epoch 7 - iter 361/3617 - loss 0.01112762 - time (sec): 100.88 - samples/sec: 380.64 - lr: 0.000069 - momentum: 0.000000
2023-10-14 07:52:10,004 epoch 7 - iter 722/3617 - loss 0.01292842 - time (sec): 203.98 - samples/sec: 373.47 - lr: 0.000068 - momentum: 0.000000
2023-10-14 07:53:56,320 epoch 7 - iter 1083/3617 - loss 0.01286273 - time (sec): 310.30 - samples/sec: 365.83 - lr: 0.000066 - momentum: 0.000000
2023-10-14 07:55:39,361 epoch 7 - iter 1444/3617 - loss 0.01371638 - time (sec): 413.34 - samples/sec: 370.42 - lr: 0.000064 - momentum: 0.000000
2023-10-14 07:57:19,068 epoch 7 - iter 1805/3617 - loss 0.01339942 - time (sec): 513.04 - samples/sec: 371.39 - lr: 0.000062 - momentum: 0.000000
2023-10-14 07:59:01,846 epoch 7 - iter 2166/3617 - loss 0.01421913 - time (sec): 615.82 - samples/sec: 370.62 - lr: 0.000060 - momentum: 0.000000
2023-10-14 08:00:47,048 epoch 7 - iter 2527/3617 - loss 0.01439729 - time (sec): 721.02 - samples/sec: 370.52 - lr: 0.000059 - momentum: 0.000000
2023-10-14 08:02:27,477 epoch 7 - iter 2888/3617 - loss 0.01432233 - time (sec): 821.45 - samples/sec: 370.63 - lr: 0.000057 - momentum: 0.000000
2023-10-14 08:04:08,239 epoch 7 - iter 3249/3617 - loss 0.01466633 - time (sec): 922.21 - samples/sec: 371.36 - lr: 0.000055 - momentum: 0.000000
2023-10-14 08:05:56,328 epoch 7 - iter 3610/3617 - loss 0.01454880 - time (sec): 1030.30 - samples/sec: 368.20 - lr: 0.000053 - momentum: 0.000000
2023-10-14 08:05:58,204 ----------------------------------------------------------------------------------------------------
2023-10-14 08:05:58,204 EPOCH 7 done: loss 0.0145 - lr: 0.000053
2023-10-14 08:06:46,371 DEV : loss 0.303059846162796 - f1-score (micro avg) 0.6603
2023-10-14 08:06:46,454 ----------------------------------------------------------------------------------------------------
2023-10-14 08:08:30,534 epoch 8 - iter 361/3617 - loss 0.00660663 - time (sec): 104.08 - samples/sec: 356.89 - lr: 0.000052 - momentum: 0.000000
2023-10-14 08:10:12,010 epoch 8 - iter 722/3617 - loss 0.00864086 - time (sec): 205.55 - samples/sec: 366.15 - lr: 0.000050 - momentum: 0.000000
2023-10-14 08:11:54,676 epoch 8 - iter 1083/3617 - loss 0.00759540 - time (sec): 308.22 - samples/sec: 372.41 - lr: 0.000048 - momentum: 0.000000
2023-10-14 08:13:35,504 epoch 8 - iter 1444/3617 - loss 0.00839154 - time (sec): 409.05 - samples/sec: 372.79 - lr: 0.000046 - momentum: 0.000000
2023-10-14 08:15:20,638 epoch 8 - iter 1805/3617 - loss 0.00820435 - time (sec): 514.18 - samples/sec: 372.20 - lr: 0.000044 - momentum: 0.000000
2023-10-14 08:17:02,003 epoch 8 - iter 2166/3617 - loss 0.00843406 - time (sec): 615.55 - samples/sec: 371.47 - lr: 0.000043 - momentum: 0.000000
2023-10-14 08:18:41,684 epoch 8 - iter 2527/3617 - loss 0.00913255 - time (sec): 715.23 - samples/sec: 372.84 - lr: 0.000041 - momentum: 0.000000
2023-10-14 08:20:30,352 epoch 8 - iter 2888/3617 - loss 0.00931845 - time (sec): 823.90 - samples/sec: 369.40 - lr: 0.000039 - momentum: 0.000000
2023-10-14 08:22:10,317 epoch 8 - iter 3249/3617 - loss 0.00922798 - time (sec): 923.86 - samples/sec: 369.93 - lr: 0.000037 - momentum: 0.000000
2023-10-14 08:23:58,744 epoch 8 - iter 3610/3617 - loss 0.00921932 - time (sec): 1032.29 - samples/sec: 367.62 - lr: 0.000036 - momentum: 0.000000
2023-10-14 08:24:00,299 ----------------------------------------------------------------------------------------------------
2023-10-14 08:24:00,299 EPOCH 8 done: loss 0.0092 - lr: 0.000036
2023-10-14 08:24:39,988 DEV : loss 0.3336223363876343 - f1-score (micro avg) 0.6534
2023-10-14 08:24:40,055 ----------------------------------------------------------------------------------------------------
2023-10-14 08:26:22,702 epoch 9 - iter 361/3617 - loss 0.00528020 - time (sec): 102.65 - samples/sec: 377.08 - lr: 0.000034 - momentum: 0.000000
2023-10-14 08:28:04,835 epoch 9 - iter 722/3617 - loss 0.00615366 - time (sec): 204.78 - samples/sec: 381.57 - lr: 0.000032 - momentum: 0.000000
2023-10-14 08:29:42,817 epoch 9 - iter 1083/3617 - loss 0.00574831 - time (sec): 302.76 - samples/sec: 381.88 - lr: 0.000030 - momentum: 0.000000
2023-10-14 08:31:21,510 epoch 9 - iter 1444/3617 - loss 0.00593530 - time (sec): 401.45 - samples/sec: 383.40 - lr: 0.000028 - momentum: 0.000000
2023-10-14 08:33:07,164 epoch 9 - iter 1805/3617 - loss 0.00576089 - time (sec): 507.11 - samples/sec: 377.05 - lr: 0.000027 - momentum: 0.000000
2023-10-14 08:34:51,885 epoch 9 - iter 2166/3617 - loss 0.00642936 - time (sec): 611.83 - samples/sec: 375.27 - lr: 0.000025 - momentum: 0.000000
2023-10-14 08:36:35,388 epoch 9 - iter 2527/3617 - loss 0.00589172 - time (sec): 715.33 - samples/sec: 373.83 - lr: 0.000023 - momentum: 0.000000
2023-10-14 08:38:16,323 epoch 9 - iter 2888/3617 - loss 0.00596664 - time (sec): 816.27 - samples/sec: 372.98 - lr: 0.000021 - momentum: 0.000000
2023-10-14 08:40:00,323 epoch 9 - iter 3249/3617 - loss 0.00596886 - time (sec): 920.27 - samples/sec: 370.19 - lr: 0.000020 - momentum: 0.000000
2023-10-14 08:41:46,467 epoch 9 - iter 3610/3617 - loss 0.00585104 - time (sec): 1026.41 - samples/sec: 369.47 - lr: 0.000018 - momentum: 0.000000
2023-10-14 08:41:48,372 ----------------------------------------------------------------------------------------------------
2023-10-14 08:41:48,373 EPOCH 9 done: loss 0.0059 - lr: 0.000018
2023-10-14 08:42:29,867 DEV : loss 0.37816229462623596 - f1-score (micro avg) 0.6517
2023-10-14 08:42:29,939 ----------------------------------------------------------------------------------------------------
2023-10-14 08:44:16,008 epoch 10 - iter 361/3617 - loss 0.00562974 - time (sec): 106.07 - samples/sec: 355.80 - lr: 0.000016 - momentum: 0.000000
2023-10-14 08:45:57,224 epoch 10 - iter 722/3617 - loss 0.00588847 - time (sec): 207.28 - samples/sec: 356.86 - lr: 0.000014 - momentum: 0.000000
2023-10-14 08:47:37,036 epoch 10 - iter 1083/3617 - loss 0.00580326 - time (sec): 307.09 - samples/sec: 366.26 - lr: 0.000012 - momentum: 0.000000
2023-10-14 08:49:18,323 epoch 10 - iter 1444/3617 - loss 0.00517980 - time (sec): 408.38 - samples/sec: 365.73 - lr: 0.000011 - momentum: 0.000000
2023-10-14 08:50:59,735 epoch 10 - iter 1805/3617 - loss 0.00496664 - time (sec): 509.79 - samples/sec: 368.87 - lr: 0.000009 - momentum: 0.000000
2023-10-14 08:52:39,423 epoch 10 - iter 2166/3617 - loss 0.00448679 - time (sec): 609.48 - samples/sec: 369.82 - lr: 0.000007 - momentum: 0.000000
2023-10-14 08:54:21,432 epoch 10 - iter 2527/3617 - loss 0.00434716 - time (sec): 711.49 - samples/sec: 372.23 - lr: 0.000005 - momentum: 0.000000
2023-10-14 08:56:01,306 epoch 10 - iter 2888/3617 - loss 0.00431111 - time (sec): 811.36 - samples/sec: 372.31 - lr: 0.000004 - momentum: 0.000000
2023-10-14 08:57:43,937 epoch 10 - iter 3249/3617 - loss 0.00426079 - time (sec): 914.00 - samples/sec: 373.34 - lr: 0.000002 - momentum: 0.000000
2023-10-14 08:59:25,667 epoch 10 - iter 3610/3617 - loss 0.00423353 - time (sec): 1015.73 - samples/sec: 373.10 - lr: 0.000000 - momentum: 0.000000
2023-10-14 08:59:27,802 ----------------------------------------------------------------------------------------------------
2023-10-14 08:59:27,802 EPOCH 10 done: loss 0.0042 - lr: 0.000000
2023-10-14 09:00:08,268 DEV : loss 0.3906834125518799 - f1-score (micro avg) 0.6573
2023-10-14 09:00:09,310 ----------------------------------------------------------------------------------------------------
2023-10-14 09:00:09,312 Loading model from best epoch ...
2023-10-14 09:00:13,705 SequenceTagger predicts: Dictionary with 13 tags: O, S-loc, B-loc, E-loc, I-loc, S-pers, B-pers, E-pers, I-pers, S-org, B-org, E-org, I-org
2023-10-14 09:01:11,888
Results:
- F-score (micro) 0.6379
- F-score (macro) 0.4852
- Accuracy 0.4792
By class:
precision recall f1-score support
loc 0.6534 0.7242 0.6870 591
pers 0.5650 0.7423 0.6416 357
org 0.1702 0.1013 0.1270 79
micro avg 0.5986 0.6826 0.6379 1027
macro avg 0.4629 0.5226 0.4852 1027
weighted avg 0.5855 0.6826 0.6282 1027
2023-10-14 09:01:11,888 ----------------------------------------------------------------------------------------------------
|