File size: 25,540 Bytes
7024ed0 |
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 |
2023-10-10 01:25:46,775 ----------------------------------------------------------------------------------------------------
2023-10-10 01:25:46,778 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 01:25:46,778 ----------------------------------------------------------------------------------------------------
2023-10-10 01:25:46,778 MultiCorpus: 20847 train + 1123 dev + 3350 test sentences
- NER_HIPE_2022 Corpus: 20847 train + 1123 dev + 3350 test sentences - /root/.flair/datasets/ner_hipe_2022/v2.1/newseye/de/with_doc_seperator
2023-10-10 01:25:46,778 ----------------------------------------------------------------------------------------------------
2023-10-10 01:25:46,778 Train: 20847 sentences
2023-10-10 01:25:46,778 (train_with_dev=False, train_with_test=False)
2023-10-10 01:25:46,778 ----------------------------------------------------------------------------------------------------
2023-10-10 01:25:46,779 Training Params:
2023-10-10 01:25:46,779 - learning_rate: "0.00015"
2023-10-10 01:25:46,779 - mini_batch_size: "4"
2023-10-10 01:25:46,779 - max_epochs: "10"
2023-10-10 01:25:46,779 - shuffle: "True"
2023-10-10 01:25:46,779 ----------------------------------------------------------------------------------------------------
2023-10-10 01:25:46,779 Plugins:
2023-10-10 01:25:46,779 - TensorboardLogger
2023-10-10 01:25:46,779 - LinearScheduler | warmup_fraction: '0.1'
2023-10-10 01:25:46,779 ----------------------------------------------------------------------------------------------------
2023-10-10 01:25:46,779 Final evaluation on model from best epoch (best-model.pt)
2023-10-10 01:25:46,779 - metric: "('micro avg', 'f1-score')"
2023-10-10 01:25:46,779 ----------------------------------------------------------------------------------------------------
2023-10-10 01:25:46,780 Computation:
2023-10-10 01:25:46,780 - compute on device: cuda:0
2023-10-10 01:25:46,780 - embedding storage: none
2023-10-10 01:25:46,780 ----------------------------------------------------------------------------------------------------
2023-10-10 01:25:46,780 Model training base path: "hmbench-newseye/de-hmbyt5-preliminary/byt5-small-historic-multilingual-span20-flax-bs4-wsFalse-e10-lr0.00015-poolingfirst-layers-1-crfFalse-1"
2023-10-10 01:25:46,780 ----------------------------------------------------------------------------------------------------
2023-10-10 01:25:46,780 ----------------------------------------------------------------------------------------------------
2023-10-10 01:25:46,780 Logging anything other than scalars to TensorBoard is currently not supported.
2023-10-10 01:28:19,901 epoch 1 - iter 521/5212 - loss 2.78721271 - time (sec): 153.12 - samples/sec: 258.19 - lr: 0.000015 - momentum: 0.000000
2023-10-10 01:30:47,683 epoch 1 - iter 1042/5212 - loss 2.40394453 - time (sec): 300.90 - samples/sec: 246.08 - lr: 0.000030 - momentum: 0.000000
2023-10-10 01:33:20,332 epoch 1 - iter 1563/5212 - loss 1.88420497 - time (sec): 453.55 - samples/sec: 243.15 - lr: 0.000045 - momentum: 0.000000
2023-10-10 01:35:50,068 epoch 1 - iter 2084/5212 - loss 1.56398813 - time (sec): 603.29 - samples/sec: 238.94 - lr: 0.000060 - momentum: 0.000000
2023-10-10 01:38:28,783 epoch 1 - iter 2605/5212 - loss 1.32922150 - time (sec): 762.00 - samples/sec: 239.39 - lr: 0.000075 - momentum: 0.000000
2023-10-10 01:41:01,466 epoch 1 - iter 3126/5212 - loss 1.16575348 - time (sec): 914.68 - samples/sec: 240.69 - lr: 0.000090 - momentum: 0.000000
2023-10-10 01:43:35,466 epoch 1 - iter 3647/5212 - loss 1.04771990 - time (sec): 1068.68 - samples/sec: 239.12 - lr: 0.000105 - momentum: 0.000000
2023-10-10 01:46:05,151 epoch 1 - iter 4168/5212 - loss 0.94981352 - time (sec): 1218.37 - samples/sec: 240.08 - lr: 0.000120 - momentum: 0.000000
2023-10-10 01:48:42,725 epoch 1 - iter 4689/5212 - loss 0.86675606 - time (sec): 1375.94 - samples/sec: 240.43 - lr: 0.000135 - momentum: 0.000000
2023-10-10 01:51:13,055 epoch 1 - iter 5210/5212 - loss 0.80285395 - time (sec): 1526.27 - samples/sec: 240.62 - lr: 0.000150 - momentum: 0.000000
2023-10-10 01:51:13,587 ----------------------------------------------------------------------------------------------------
2023-10-10 01:51:13,588 EPOCH 1 done: loss 0.8026 - lr: 0.000150
2023-10-10 01:51:50,130 DEV : loss 0.13000161945819855 - f1-score (micro avg) 0.2937
2023-10-10 01:51:50,181 saving best model
2023-10-10 01:51:51,139 ----------------------------------------------------------------------------------------------------
2023-10-10 01:54:26,458 epoch 2 - iter 521/5212 - loss 0.17727772 - time (sec): 155.32 - samples/sec: 256.21 - lr: 0.000148 - momentum: 0.000000
2023-10-10 01:57:04,108 epoch 2 - iter 1042/5212 - loss 0.18552352 - time (sec): 312.97 - samples/sec: 251.81 - lr: 0.000147 - momentum: 0.000000
2023-10-10 01:59:40,241 epoch 2 - iter 1563/5212 - loss 0.17656777 - time (sec): 469.10 - samples/sec: 247.34 - lr: 0.000145 - momentum: 0.000000
2023-10-10 02:02:10,466 epoch 2 - iter 2084/5212 - loss 0.17347903 - time (sec): 619.32 - samples/sec: 243.06 - lr: 0.000143 - momentum: 0.000000
2023-10-10 02:04:44,612 epoch 2 - iter 2605/5212 - loss 0.17247510 - time (sec): 773.47 - samples/sec: 241.26 - lr: 0.000142 - momentum: 0.000000
2023-10-10 02:07:16,051 epoch 2 - iter 3126/5212 - loss 0.17019344 - time (sec): 924.91 - samples/sec: 240.53 - lr: 0.000140 - momentum: 0.000000
2023-10-10 02:09:48,370 epoch 2 - iter 3647/5212 - loss 0.16742648 - time (sec): 1077.23 - samples/sec: 240.03 - lr: 0.000138 - momentum: 0.000000
2023-10-10 02:12:23,147 epoch 2 - iter 4168/5212 - loss 0.16216771 - time (sec): 1232.01 - samples/sec: 240.31 - lr: 0.000137 - momentum: 0.000000
2023-10-10 02:15:01,539 epoch 2 - iter 4689/5212 - loss 0.15863022 - time (sec): 1390.40 - samples/sec: 239.62 - lr: 0.000135 - momentum: 0.000000
2023-10-10 02:17:30,737 epoch 2 - iter 5210/5212 - loss 0.15545238 - time (sec): 1539.60 - samples/sec: 238.55 - lr: 0.000133 - momentum: 0.000000
2023-10-10 02:17:31,246 ----------------------------------------------------------------------------------------------------
2023-10-10 02:17:31,247 EPOCH 2 done: loss 0.1554 - lr: 0.000133
2023-10-10 02:18:13,857 DEV : loss 0.1568019837141037 - f1-score (micro avg) 0.3643
2023-10-10 02:18:13,912 saving best model
2023-10-10 02:18:16,652 ----------------------------------------------------------------------------------------------------
2023-10-10 02:20:47,238 epoch 3 - iter 521/5212 - loss 0.09227579 - time (sec): 150.58 - samples/sec: 242.13 - lr: 0.000132 - momentum: 0.000000
2023-10-10 02:23:18,633 epoch 3 - iter 1042/5212 - loss 0.10115948 - time (sec): 301.98 - samples/sec: 236.47 - lr: 0.000130 - momentum: 0.000000
2023-10-10 02:25:54,573 epoch 3 - iter 1563/5212 - loss 0.10224575 - time (sec): 457.92 - samples/sec: 241.87 - lr: 0.000128 - momentum: 0.000000
2023-10-10 02:28:26,479 epoch 3 - iter 2084/5212 - loss 0.10442613 - time (sec): 609.82 - samples/sec: 236.77 - lr: 0.000127 - momentum: 0.000000
2023-10-10 02:30:55,098 epoch 3 - iter 2605/5212 - loss 0.10550793 - time (sec): 758.44 - samples/sec: 234.00 - lr: 0.000125 - momentum: 0.000000
2023-10-10 02:33:31,679 epoch 3 - iter 3126/5212 - loss 0.10488651 - time (sec): 915.02 - samples/sec: 237.84 - lr: 0.000123 - momentum: 0.000000
2023-10-10 02:36:05,961 epoch 3 - iter 3647/5212 - loss 0.10697333 - time (sec): 1069.30 - samples/sec: 239.46 - lr: 0.000122 - momentum: 0.000000
2023-10-10 02:38:37,130 epoch 3 - iter 4168/5212 - loss 0.10672325 - time (sec): 1220.47 - samples/sec: 240.95 - lr: 0.000120 - momentum: 0.000000
2023-10-10 02:41:12,922 epoch 3 - iter 4689/5212 - loss 0.10596268 - time (sec): 1376.27 - samples/sec: 241.08 - lr: 0.000118 - momentum: 0.000000
2023-10-10 02:43:44,745 epoch 3 - iter 5210/5212 - loss 0.10482979 - time (sec): 1528.09 - samples/sec: 240.42 - lr: 0.000117 - momentum: 0.000000
2023-10-10 02:43:45,188 ----------------------------------------------------------------------------------------------------
2023-10-10 02:43:45,188 EPOCH 3 done: loss 0.1048 - lr: 0.000117
2023-10-10 02:44:26,403 DEV : loss 0.26355189085006714 - f1-score (micro avg) 0.3544
2023-10-10 02:44:26,460 ----------------------------------------------------------------------------------------------------
2023-10-10 02:47:01,262 epoch 4 - iter 521/5212 - loss 0.06438853 - time (sec): 154.80 - samples/sec: 238.52 - lr: 0.000115 - momentum: 0.000000
2023-10-10 02:49:38,517 epoch 4 - iter 1042/5212 - loss 0.06254566 - time (sec): 312.05 - samples/sec: 231.26 - lr: 0.000113 - momentum: 0.000000
2023-10-10 02:52:14,492 epoch 4 - iter 1563/5212 - loss 0.06423763 - time (sec): 468.03 - samples/sec: 230.64 - lr: 0.000112 - momentum: 0.000000
2023-10-10 02:54:44,857 epoch 4 - iter 2084/5212 - loss 0.06688261 - time (sec): 618.39 - samples/sec: 232.20 - lr: 0.000110 - momentum: 0.000000
2023-10-10 02:57:20,257 epoch 4 - iter 2605/5212 - loss 0.07115747 - time (sec): 773.79 - samples/sec: 234.30 - lr: 0.000108 - momentum: 0.000000
2023-10-10 02:59:54,457 epoch 4 - iter 3126/5212 - loss 0.06965197 - time (sec): 927.99 - samples/sec: 238.97 - lr: 0.000107 - momentum: 0.000000
2023-10-10 03:02:33,869 epoch 4 - iter 3647/5212 - loss 0.06885936 - time (sec): 1087.41 - samples/sec: 235.99 - lr: 0.000105 - momentum: 0.000000
2023-10-10 03:05:03,715 epoch 4 - iter 4168/5212 - loss 0.07019996 - time (sec): 1237.25 - samples/sec: 236.43 - lr: 0.000103 - momentum: 0.000000
2023-10-10 03:07:37,272 epoch 4 - iter 4689/5212 - loss 0.07063262 - time (sec): 1390.81 - samples/sec: 237.63 - lr: 0.000102 - momentum: 0.000000
2023-10-10 03:10:08,879 epoch 4 - iter 5210/5212 - loss 0.07185046 - time (sec): 1542.42 - samples/sec: 238.20 - lr: 0.000100 - momentum: 0.000000
2023-10-10 03:10:09,314 ----------------------------------------------------------------------------------------------------
2023-10-10 03:10:09,314 EPOCH 4 done: loss 0.0719 - lr: 0.000100
2023-10-10 03:10:57,812 DEV : loss 0.328808069229126 - f1-score (micro avg) 0.3675
2023-10-10 03:10:57,879 saving best model
2023-10-10 03:11:09,277 ----------------------------------------------------------------------------------------------------
2023-10-10 03:13:41,434 epoch 5 - iter 521/5212 - loss 0.05178094 - time (sec): 152.15 - samples/sec: 227.02 - lr: 0.000098 - momentum: 0.000000
2023-10-10 03:16:13,044 epoch 5 - iter 1042/5212 - loss 0.05498718 - time (sec): 303.76 - samples/sec: 234.98 - lr: 0.000097 - momentum: 0.000000
2023-10-10 03:18:45,690 epoch 5 - iter 1563/5212 - loss 0.05426323 - time (sec): 456.41 - samples/sec: 241.01 - lr: 0.000095 - momentum: 0.000000
2023-10-10 03:21:23,172 epoch 5 - iter 2084/5212 - loss 0.05256449 - time (sec): 613.89 - samples/sec: 242.02 - lr: 0.000093 - momentum: 0.000000
2023-10-10 03:23:56,778 epoch 5 - iter 2605/5212 - loss 0.05299146 - time (sec): 767.50 - samples/sec: 238.88 - lr: 0.000092 - momentum: 0.000000
2023-10-10 03:26:30,782 epoch 5 - iter 3126/5212 - loss 0.05454252 - time (sec): 921.50 - samples/sec: 239.64 - lr: 0.000090 - momentum: 0.000000
2023-10-10 03:29:07,368 epoch 5 - iter 3647/5212 - loss 0.05456496 - time (sec): 1078.09 - samples/sec: 240.64 - lr: 0.000088 - momentum: 0.000000
2023-10-10 03:31:43,290 epoch 5 - iter 4168/5212 - loss 0.05333736 - time (sec): 1234.01 - samples/sec: 239.80 - lr: 0.000087 - momentum: 0.000000
2023-10-10 03:34:15,429 epoch 5 - iter 4689/5212 - loss 0.05223842 - time (sec): 1386.15 - samples/sec: 238.43 - lr: 0.000085 - momentum: 0.000000
2023-10-10 03:36:49,643 epoch 5 - iter 5210/5212 - loss 0.05325230 - time (sec): 1540.36 - samples/sec: 238.48 - lr: 0.000083 - momentum: 0.000000
2023-10-10 03:36:50,135 ----------------------------------------------------------------------------------------------------
2023-10-10 03:36:50,135 EPOCH 5 done: loss 0.0532 - lr: 0.000083
2023-10-10 03:37:32,491 DEV : loss 0.3098498284816742 - f1-score (micro avg) 0.3954
2023-10-10 03:37:32,548 saving best model
2023-10-10 03:37:35,444 ----------------------------------------------------------------------------------------------------
2023-10-10 03:40:16,955 epoch 6 - iter 521/5212 - loss 0.03216296 - time (sec): 161.51 - samples/sec: 227.66 - lr: 0.000082 - momentum: 0.000000
2023-10-10 03:42:49,627 epoch 6 - iter 1042/5212 - loss 0.03763248 - time (sec): 314.18 - samples/sec: 225.89 - lr: 0.000080 - momentum: 0.000000
2023-10-10 03:45:24,399 epoch 6 - iter 1563/5212 - loss 0.03476631 - time (sec): 468.95 - samples/sec: 234.51 - lr: 0.000078 - momentum: 0.000000
2023-10-10 03:48:01,947 epoch 6 - iter 2084/5212 - loss 0.03495034 - time (sec): 626.50 - samples/sec: 232.55 - lr: 0.000077 - momentum: 0.000000
2023-10-10 03:50:33,387 epoch 6 - iter 2605/5212 - loss 0.03541576 - time (sec): 777.94 - samples/sec: 234.85 - lr: 0.000075 - momentum: 0.000000
2023-10-10 03:53:03,173 epoch 6 - iter 3126/5212 - loss 0.03439095 - time (sec): 927.72 - samples/sec: 233.04 - lr: 0.000073 - momentum: 0.000000
2023-10-10 03:55:38,848 epoch 6 - iter 3647/5212 - loss 0.03483107 - time (sec): 1083.40 - samples/sec: 233.74 - lr: 0.000072 - momentum: 0.000000
2023-10-10 03:58:12,884 epoch 6 - iter 4168/5212 - loss 0.03532550 - time (sec): 1237.44 - samples/sec: 235.10 - lr: 0.000070 - momentum: 0.000000
2023-10-10 04:00:56,870 epoch 6 - iter 4689/5212 - loss 0.03575938 - time (sec): 1401.42 - samples/sec: 235.54 - lr: 0.000068 - momentum: 0.000000
2023-10-10 04:03:29,722 epoch 6 - iter 5210/5212 - loss 0.03599569 - time (sec): 1554.27 - samples/sec: 236.34 - lr: 0.000067 - momentum: 0.000000
2023-10-10 04:03:30,188 ----------------------------------------------------------------------------------------------------
2023-10-10 04:03:30,188 EPOCH 6 done: loss 0.0360 - lr: 0.000067
2023-10-10 04:04:12,941 DEV : loss 0.3672010004520416 - f1-score (micro avg) 0.3823
2023-10-10 04:04:13,004 ----------------------------------------------------------------------------------------------------
2023-10-10 04:06:49,841 epoch 7 - iter 521/5212 - loss 0.01901470 - time (sec): 156.83 - samples/sec: 245.25 - lr: 0.000065 - momentum: 0.000000
2023-10-10 04:09:26,255 epoch 7 - iter 1042/5212 - loss 0.02194304 - time (sec): 313.25 - samples/sec: 244.81 - lr: 0.000063 - momentum: 0.000000
2023-10-10 04:12:00,065 epoch 7 - iter 1563/5212 - loss 0.02292694 - time (sec): 467.06 - samples/sec: 239.09 - lr: 0.000062 - momentum: 0.000000
2023-10-10 04:14:34,985 epoch 7 - iter 2084/5212 - loss 0.02286490 - time (sec): 621.98 - samples/sec: 239.67 - lr: 0.000060 - momentum: 0.000000
2023-10-10 04:17:07,483 epoch 7 - iter 2605/5212 - loss 0.02158969 - time (sec): 774.48 - samples/sec: 241.38 - lr: 0.000058 - momentum: 0.000000
2023-10-10 04:19:42,460 epoch 7 - iter 3126/5212 - loss 0.02437178 - time (sec): 929.45 - samples/sec: 241.11 - lr: 0.000057 - momentum: 0.000000
2023-10-10 04:22:14,067 epoch 7 - iter 3647/5212 - loss 0.02420860 - time (sec): 1081.06 - samples/sec: 240.27 - lr: 0.000055 - momentum: 0.000000
2023-10-10 04:24:44,274 epoch 7 - iter 4168/5212 - loss 0.02497424 - time (sec): 1231.27 - samples/sec: 240.58 - lr: 0.000053 - momentum: 0.000000
2023-10-10 04:27:18,230 epoch 7 - iter 4689/5212 - loss 0.02500752 - time (sec): 1385.22 - samples/sec: 240.06 - lr: 0.000052 - momentum: 0.000000
2023-10-10 04:29:46,445 epoch 7 - iter 5210/5212 - loss 0.02487125 - time (sec): 1533.44 - samples/sec: 239.57 - lr: 0.000050 - momentum: 0.000000
2023-10-10 04:29:46,901 ----------------------------------------------------------------------------------------------------
2023-10-10 04:29:46,902 EPOCH 7 done: loss 0.0249 - lr: 0.000050
2023-10-10 04:30:31,180 DEV : loss 0.43393340706825256 - f1-score (micro avg) 0.3819
2023-10-10 04:30:31,244 ----------------------------------------------------------------------------------------------------
2023-10-10 04:33:08,702 epoch 8 - iter 521/5212 - loss 0.02118548 - time (sec): 157.46 - samples/sec: 229.44 - lr: 0.000048 - momentum: 0.000000
2023-10-10 04:35:41,639 epoch 8 - iter 1042/5212 - loss 0.01953101 - time (sec): 310.39 - samples/sec: 235.18 - lr: 0.000047 - momentum: 0.000000
2023-10-10 04:38:18,307 epoch 8 - iter 1563/5212 - loss 0.01941108 - time (sec): 467.06 - samples/sec: 237.33 - lr: 0.000045 - momentum: 0.000000
2023-10-10 04:40:52,233 epoch 8 - iter 2084/5212 - loss 0.01883043 - time (sec): 620.99 - samples/sec: 236.34 - lr: 0.000043 - momentum: 0.000000
2023-10-10 04:43:26,171 epoch 8 - iter 2605/5212 - loss 0.01862250 - time (sec): 774.92 - samples/sec: 236.00 - lr: 0.000042 - momentum: 0.000000
2023-10-10 04:46:02,246 epoch 8 - iter 3126/5212 - loss 0.01937239 - time (sec): 931.00 - samples/sec: 236.09 - lr: 0.000040 - momentum: 0.000000
2023-10-10 04:48:33,038 epoch 8 - iter 3647/5212 - loss 0.01908805 - time (sec): 1081.79 - samples/sec: 235.79 - lr: 0.000038 - momentum: 0.000000
2023-10-10 04:51:12,633 epoch 8 - iter 4168/5212 - loss 0.01889720 - time (sec): 1241.39 - samples/sec: 236.80 - lr: 0.000037 - momentum: 0.000000
2023-10-10 04:53:47,905 epoch 8 - iter 4689/5212 - loss 0.01843401 - time (sec): 1396.66 - samples/sec: 236.95 - lr: 0.000035 - momentum: 0.000000
2023-10-10 04:56:21,396 epoch 8 - iter 5210/5212 - loss 0.01859840 - time (sec): 1550.15 - samples/sec: 236.93 - lr: 0.000033 - momentum: 0.000000
2023-10-10 04:56:21,934 ----------------------------------------------------------------------------------------------------
2023-10-10 04:56:21,934 EPOCH 8 done: loss 0.0186 - lr: 0.000033
2023-10-10 04:57:03,167 DEV : loss 0.46257734298706055 - f1-score (micro avg) 0.3734
2023-10-10 04:57:03,238 ----------------------------------------------------------------------------------------------------
2023-10-10 04:59:37,592 epoch 9 - iter 521/5212 - loss 0.01100351 - time (sec): 154.35 - samples/sec: 244.77 - lr: 0.000032 - momentum: 0.000000
2023-10-10 05:02:12,911 epoch 9 - iter 1042/5212 - loss 0.01142614 - time (sec): 309.67 - samples/sec: 244.21 - lr: 0.000030 - momentum: 0.000000
2023-10-10 05:04:45,707 epoch 9 - iter 1563/5212 - loss 0.01200825 - time (sec): 462.47 - samples/sec: 238.44 - lr: 0.000028 - momentum: 0.000000
2023-10-10 05:07:25,374 epoch 9 - iter 2084/5212 - loss 0.01167246 - time (sec): 622.13 - samples/sec: 236.70 - lr: 0.000027 - momentum: 0.000000
2023-10-10 05:10:00,448 epoch 9 - iter 2605/5212 - loss 0.01211194 - time (sec): 777.21 - samples/sec: 236.18 - lr: 0.000025 - momentum: 0.000000
2023-10-10 05:12:35,180 epoch 9 - iter 3126/5212 - loss 0.01227832 - time (sec): 931.94 - samples/sec: 236.70 - lr: 0.000023 - momentum: 0.000000
2023-10-10 05:15:13,447 epoch 9 - iter 3647/5212 - loss 0.01232401 - time (sec): 1090.21 - samples/sec: 235.04 - lr: 0.000022 - momentum: 0.000000
2023-10-10 05:17:46,800 epoch 9 - iter 4168/5212 - loss 0.01205408 - time (sec): 1243.56 - samples/sec: 234.71 - lr: 0.000020 - momentum: 0.000000
2023-10-10 05:20:19,024 epoch 9 - iter 4689/5212 - loss 0.01137052 - time (sec): 1395.78 - samples/sec: 235.39 - lr: 0.000018 - momentum: 0.000000
2023-10-10 05:23:03,373 epoch 9 - iter 5210/5212 - loss 0.01116459 - time (sec): 1560.13 - samples/sec: 235.41 - lr: 0.000017 - momentum: 0.000000
2023-10-10 05:23:03,935 ----------------------------------------------------------------------------------------------------
2023-10-10 05:23:03,935 EPOCH 9 done: loss 0.0112 - lr: 0.000017
2023-10-10 05:23:45,741 DEV : loss 0.49680188298225403 - f1-score (micro avg) 0.386
2023-10-10 05:23:45,810 ----------------------------------------------------------------------------------------------------
2023-10-10 05:26:19,459 epoch 10 - iter 521/5212 - loss 0.00618823 - time (sec): 153.65 - samples/sec: 239.16 - lr: 0.000015 - momentum: 0.000000
2023-10-10 05:28:50,815 epoch 10 - iter 1042/5212 - loss 0.00778641 - time (sec): 305.00 - samples/sec: 236.39 - lr: 0.000013 - momentum: 0.000000
2023-10-10 05:31:22,417 epoch 10 - iter 1563/5212 - loss 0.00880970 - time (sec): 456.60 - samples/sec: 231.53 - lr: 0.000012 - momentum: 0.000000
2023-10-10 05:34:00,744 epoch 10 - iter 2084/5212 - loss 0.00838074 - time (sec): 614.93 - samples/sec: 234.01 - lr: 0.000010 - momentum: 0.000000
2023-10-10 05:36:40,115 epoch 10 - iter 2605/5212 - loss 0.00876966 - time (sec): 774.30 - samples/sec: 238.88 - lr: 0.000008 - momentum: 0.000000
2023-10-10 05:39:13,455 epoch 10 - iter 3126/5212 - loss 0.00882349 - time (sec): 927.64 - samples/sec: 237.17 - lr: 0.000007 - momentum: 0.000000
2023-10-10 05:41:47,684 epoch 10 - iter 3647/5212 - loss 0.00895006 - time (sec): 1081.87 - samples/sec: 238.01 - lr: 0.000005 - momentum: 0.000000
2023-10-10 05:44:19,241 epoch 10 - iter 4168/5212 - loss 0.00888338 - time (sec): 1233.43 - samples/sec: 239.39 - lr: 0.000003 - momentum: 0.000000
2023-10-10 05:46:50,896 epoch 10 - iter 4689/5212 - loss 0.00858734 - time (sec): 1385.08 - samples/sec: 240.18 - lr: 0.000002 - momentum: 0.000000
2023-10-10 05:49:20,398 epoch 10 - iter 5210/5212 - loss 0.00861144 - time (sec): 1534.59 - samples/sec: 239.31 - lr: 0.000000 - momentum: 0.000000
2023-10-10 05:49:20,941 ----------------------------------------------------------------------------------------------------
2023-10-10 05:49:20,942 EPOCH 10 done: loss 0.0086 - lr: 0.000000
2023-10-10 05:50:01,786 DEV : loss 0.507175087928772 - f1-score (micro avg) 0.382
2023-10-10 05:50:02,826 ----------------------------------------------------------------------------------------------------
2023-10-10 05:50:02,828 Loading model from best epoch ...
2023-10-10 05:50:07,141 SequenceTagger predicts: Dictionary with 17 tags: O, S-LOC, B-LOC, E-LOC, I-LOC, S-PER, B-PER, E-PER, I-PER, S-ORG, B-ORG, E-ORG, I-ORG, S-HumanProd, B-HumanProd, E-HumanProd, I-HumanProd
2023-10-10 05:51:51,254
Results:
- F-score (micro) 0.4873
- F-score (macro) 0.327
- Accuracy 0.3265
By class:
precision recall f1-score support
LOC 0.5032 0.6400 0.5635 1214
PER 0.4093 0.5025 0.4511 808
ORG 0.2982 0.2890 0.2935 353
HumanProd 0.0000 0.0000 0.0000 15
micro avg 0.4456 0.5377 0.4873 2390
macro avg 0.3027 0.3579 0.3270 2390
weighted avg 0.4380 0.5377 0.4821 2390
2023-10-10 05:51:51,255 ----------------------------------------------------------------------------------------------------
|