File size: 25,579 Bytes
e293b19 |
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 |
2023-10-11 14:33:43,977 ----------------------------------------------------------------------------------------------------
2023-10-11 14:33:43,979 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()
)"
2023-10-11 14:33:43,980 ----------------------------------------------------------------------------------------------------
2023-10-11 14:33:43,980 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
2023-10-11 14:33:43,980 ----------------------------------------------------------------------------------------------------
2023-10-11 14:33:43,980 Train: 20847 sentences
2023-10-11 14:33:43,980 (train_with_dev=False, train_with_test=False)
2023-10-11 14:33:43,980 ----------------------------------------------------------------------------------------------------
2023-10-11 14:33:43,980 Training Params:
2023-10-11 14:33:43,980 - learning_rate: "0.00016"
2023-10-11 14:33:43,980 - mini_batch_size: "4"
2023-10-11 14:33:43,981 - max_epochs: "10"
2023-10-11 14:33:43,981 - shuffle: "True"
2023-10-11 14:33:43,981 ----------------------------------------------------------------------------------------------------
2023-10-11 14:33:43,981 Plugins:
2023-10-11 14:33:43,981 - TensorboardLogger
2023-10-11 14:33:43,981 - LinearScheduler | warmup_fraction: '0.1'
2023-10-11 14:33:43,981 ----------------------------------------------------------------------------------------------------
2023-10-11 14:33:43,981 Final evaluation on model from best epoch (best-model.pt)
2023-10-11 14:33:43,981 - metric: "('micro avg', 'f1-score')"
2023-10-11 14:33:43,981 ----------------------------------------------------------------------------------------------------
2023-10-11 14:33:43,981 Computation:
2023-10-11 14:33:43,981 - compute on device: cuda:0
2023-10-11 14:33:43,981 - embedding storage: none
2023-10-11 14:33:43,982 ----------------------------------------------------------------------------------------------------
2023-10-11 14:33:43,982 Model training base path: "hmbench-newseye/de-hmbyt5-preliminary/byt5-small-historic-multilingual-span20-flax-bs4-wsFalse-e10-lr0.00016-poolingfirst-layers-1-crfFalse-3"
2023-10-11 14:33:43,982 ----------------------------------------------------------------------------------------------------
2023-10-11 14:33:43,982 ----------------------------------------------------------------------------------------------------
2023-10-11 14:33:43,982 Logging anything other than scalars to TensorBoard is currently not supported.
2023-10-11 14:36:06,547 epoch 1 - iter 521/5212 - loss 2.76662736 - time (sec): 142.56 - samples/sec: 262.14 - lr: 0.000016 - momentum: 0.000000
2023-10-11 14:38:27,452 epoch 1 - iter 1042/5212 - loss 2.28624082 - time (sec): 283.47 - samples/sec: 270.79 - lr: 0.000032 - momentum: 0.000000
2023-10-11 14:40:44,639 epoch 1 - iter 1563/5212 - loss 1.79645371 - time (sec): 420.66 - samples/sec: 269.24 - lr: 0.000048 - momentum: 0.000000
2023-10-11 14:43:01,808 epoch 1 - iter 2084/5212 - loss 1.46859051 - time (sec): 557.82 - samples/sec: 267.94 - lr: 0.000064 - momentum: 0.000000
2023-10-11 14:45:26,500 epoch 1 - iter 2605/5212 - loss 1.27159482 - time (sec): 702.52 - samples/sec: 266.51 - lr: 0.000080 - momentum: 0.000000
2023-10-11 14:47:48,763 epoch 1 - iter 3126/5212 - loss 1.12283149 - time (sec): 844.78 - samples/sec: 264.51 - lr: 0.000096 - momentum: 0.000000
2023-10-11 14:50:12,100 epoch 1 - iter 3647/5212 - loss 1.01175361 - time (sec): 988.12 - samples/sec: 261.77 - lr: 0.000112 - momentum: 0.000000
2023-10-11 14:52:34,157 epoch 1 - iter 4168/5212 - loss 0.93063939 - time (sec): 1130.17 - samples/sec: 259.51 - lr: 0.000128 - momentum: 0.000000
2023-10-11 14:54:53,773 epoch 1 - iter 4689/5212 - loss 0.85556421 - time (sec): 1269.79 - samples/sec: 260.71 - lr: 0.000144 - momentum: 0.000000
2023-10-11 14:57:19,976 epoch 1 - iter 5210/5212 - loss 0.79086304 - time (sec): 1415.99 - samples/sec: 259.36 - lr: 0.000160 - momentum: 0.000000
2023-10-11 14:57:20,519 ----------------------------------------------------------------------------------------------------
2023-10-11 14:57:20,519 EPOCH 1 done: loss 0.7907 - lr: 0.000160
2023-10-11 14:57:56,736 DEV : loss 0.13207665085792542 - f1-score (micro avg) 0.3237
2023-10-11 14:57:56,789 saving best model
2023-10-11 14:57:57,698 ----------------------------------------------------------------------------------------------------
2023-10-11 15:00:22,791 epoch 2 - iter 521/5212 - loss 0.19332337 - time (sec): 145.09 - samples/sec: 254.59 - lr: 0.000158 - momentum: 0.000000
2023-10-11 15:02:45,528 epoch 2 - iter 1042/5212 - loss 0.18215094 - time (sec): 287.83 - samples/sec: 254.97 - lr: 0.000156 - momentum: 0.000000
2023-10-11 15:05:10,335 epoch 2 - iter 1563/5212 - loss 0.18444577 - time (sec): 432.63 - samples/sec: 260.62 - lr: 0.000155 - momentum: 0.000000
2023-10-11 15:07:31,990 epoch 2 - iter 2084/5212 - loss 0.18224841 - time (sec): 574.29 - samples/sec: 260.31 - lr: 0.000153 - momentum: 0.000000
2023-10-11 15:09:52,786 epoch 2 - iter 2605/5212 - loss 0.17760403 - time (sec): 715.09 - samples/sec: 258.27 - lr: 0.000151 - momentum: 0.000000
2023-10-11 15:12:16,111 epoch 2 - iter 3126/5212 - loss 0.17321912 - time (sec): 858.41 - samples/sec: 258.53 - lr: 0.000149 - momentum: 0.000000
2023-10-11 15:14:34,077 epoch 2 - iter 3647/5212 - loss 0.17155961 - time (sec): 996.38 - samples/sec: 256.43 - lr: 0.000148 - momentum: 0.000000
2023-10-11 15:16:56,628 epoch 2 - iter 4168/5212 - loss 0.16731547 - time (sec): 1138.93 - samples/sec: 256.39 - lr: 0.000146 - momentum: 0.000000
2023-10-11 15:19:19,074 epoch 2 - iter 4689/5212 - loss 0.16554814 - time (sec): 1281.37 - samples/sec: 257.92 - lr: 0.000144 - momentum: 0.000000
2023-10-11 15:21:37,400 epoch 2 - iter 5210/5212 - loss 0.16192292 - time (sec): 1419.70 - samples/sec: 258.75 - lr: 0.000142 - momentum: 0.000000
2023-10-11 15:21:37,838 ----------------------------------------------------------------------------------------------------
2023-10-11 15:21:37,838 EPOCH 2 done: loss 0.1619 - lr: 0.000142
2023-10-11 15:22:17,665 DEV : loss 0.16001686453819275 - f1-score (micro avg) 0.3627
2023-10-11 15:22:17,722 saving best model
2023-10-11 15:22:20,336 ----------------------------------------------------------------------------------------------------
2023-10-11 15:24:36,175 epoch 3 - iter 521/5212 - loss 0.09560052 - time (sec): 135.83 - samples/sec: 257.42 - lr: 0.000140 - momentum: 0.000000
2023-10-11 15:26:53,189 epoch 3 - iter 1042/5212 - loss 0.10144802 - time (sec): 272.85 - samples/sec: 260.58 - lr: 0.000139 - momentum: 0.000000
2023-10-11 15:29:13,216 epoch 3 - iter 1563/5212 - loss 0.09999435 - time (sec): 412.88 - samples/sec: 258.87 - lr: 0.000137 - momentum: 0.000000
2023-10-11 15:31:36,953 epoch 3 - iter 2084/5212 - loss 0.10591845 - time (sec): 556.61 - samples/sec: 261.87 - lr: 0.000135 - momentum: 0.000000
2023-10-11 15:33:57,505 epoch 3 - iter 2605/5212 - loss 0.11024910 - time (sec): 697.16 - samples/sec: 263.88 - lr: 0.000133 - momentum: 0.000000
2023-10-11 15:36:17,912 epoch 3 - iter 3126/5212 - loss 0.10967581 - time (sec): 837.57 - samples/sec: 262.61 - lr: 0.000132 - momentum: 0.000000
2023-10-11 15:38:38,424 epoch 3 - iter 3647/5212 - loss 0.10727296 - time (sec): 978.08 - samples/sec: 261.21 - lr: 0.000130 - momentum: 0.000000
2023-10-11 15:40:59,896 epoch 3 - iter 4168/5212 - loss 0.10769948 - time (sec): 1119.55 - samples/sec: 261.35 - lr: 0.000128 - momentum: 0.000000
2023-10-11 15:43:20,757 epoch 3 - iter 4689/5212 - loss 0.10703189 - time (sec): 1260.42 - samples/sec: 260.75 - lr: 0.000126 - momentum: 0.000000
2023-10-11 15:45:46,414 epoch 3 - iter 5210/5212 - loss 0.10585857 - time (sec): 1406.07 - samples/sec: 261.24 - lr: 0.000124 - momentum: 0.000000
2023-10-11 15:45:46,886 ----------------------------------------------------------------------------------------------------
2023-10-11 15:45:46,886 EPOCH 3 done: loss 0.1060 - lr: 0.000124
2023-10-11 15:46:27,270 DEV : loss 0.21690955758094788 - f1-score (micro avg) 0.3633
2023-10-11 15:46:27,322 saving best model
2023-10-11 15:46:29,931 ----------------------------------------------------------------------------------------------------
2023-10-11 15:48:54,172 epoch 4 - iter 521/5212 - loss 0.07319839 - time (sec): 144.24 - samples/sec: 244.92 - lr: 0.000123 - momentum: 0.000000
2023-10-11 15:51:17,181 epoch 4 - iter 1042/5212 - loss 0.07369046 - time (sec): 287.25 - samples/sec: 250.39 - lr: 0.000121 - momentum: 0.000000
2023-10-11 15:53:41,453 epoch 4 - iter 1563/5212 - loss 0.07241720 - time (sec): 431.52 - samples/sec: 253.60 - lr: 0.000119 - momentum: 0.000000
2023-10-11 15:56:04,766 epoch 4 - iter 2084/5212 - loss 0.07187332 - time (sec): 574.83 - samples/sec: 251.80 - lr: 0.000117 - momentum: 0.000000
2023-10-11 15:58:30,038 epoch 4 - iter 2605/5212 - loss 0.07110536 - time (sec): 720.10 - samples/sec: 256.81 - lr: 0.000116 - momentum: 0.000000
2023-10-11 16:00:49,154 epoch 4 - iter 3126/5212 - loss 0.07215527 - time (sec): 859.22 - samples/sec: 256.21 - lr: 0.000114 - momentum: 0.000000
2023-10-11 16:03:11,988 epoch 4 - iter 3647/5212 - loss 0.07322367 - time (sec): 1002.05 - samples/sec: 257.20 - lr: 0.000112 - momentum: 0.000000
2023-10-11 16:05:38,070 epoch 4 - iter 4168/5212 - loss 0.07331747 - time (sec): 1148.14 - samples/sec: 259.47 - lr: 0.000110 - momentum: 0.000000
2023-10-11 16:07:57,178 epoch 4 - iter 4689/5212 - loss 0.07365219 - time (sec): 1287.24 - samples/sec: 257.98 - lr: 0.000108 - momentum: 0.000000
2023-10-11 16:10:18,604 epoch 4 - iter 5210/5212 - loss 0.07301371 - time (sec): 1428.67 - samples/sec: 257.16 - lr: 0.000107 - momentum: 0.000000
2023-10-11 16:10:19,011 ----------------------------------------------------------------------------------------------------
2023-10-11 16:10:19,011 EPOCH 4 done: loss 0.0730 - lr: 0.000107
2023-10-11 16:10:58,573 DEV : loss 0.2578122615814209 - f1-score (micro avg) 0.4037
2023-10-11 16:10:58,626 saving best model
2023-10-11 16:11:01,254 ----------------------------------------------------------------------------------------------------
2023-10-11 16:13:24,831 epoch 5 - iter 521/5212 - loss 0.04596828 - time (sec): 143.57 - samples/sec: 250.03 - lr: 0.000105 - momentum: 0.000000
2023-10-11 16:15:52,154 epoch 5 - iter 1042/5212 - loss 0.05176257 - time (sec): 290.89 - samples/sec: 253.07 - lr: 0.000103 - momentum: 0.000000
2023-10-11 16:18:26,212 epoch 5 - iter 1563/5212 - loss 0.05301496 - time (sec): 444.95 - samples/sec: 245.48 - lr: 0.000101 - momentum: 0.000000
2023-10-11 16:20:56,560 epoch 5 - iter 2084/5212 - loss 0.05310164 - time (sec): 595.30 - samples/sec: 244.83 - lr: 0.000100 - momentum: 0.000000
2023-10-11 16:23:29,574 epoch 5 - iter 2605/5212 - loss 0.05219774 - time (sec): 748.32 - samples/sec: 246.05 - lr: 0.000098 - momentum: 0.000000
2023-10-11 16:25:52,444 epoch 5 - iter 3126/5212 - loss 0.05195458 - time (sec): 891.19 - samples/sec: 246.25 - lr: 0.000096 - momentum: 0.000000
2023-10-11 16:28:14,834 epoch 5 - iter 3647/5212 - loss 0.05186622 - time (sec): 1033.58 - samples/sec: 247.88 - lr: 0.000094 - momentum: 0.000000
2023-10-11 16:30:31,671 epoch 5 - iter 4168/5212 - loss 0.05189342 - time (sec): 1170.41 - samples/sec: 248.98 - lr: 0.000092 - momentum: 0.000000
2023-10-11 16:32:51,894 epoch 5 - iter 4689/5212 - loss 0.05074979 - time (sec): 1310.64 - samples/sec: 251.02 - lr: 0.000091 - momentum: 0.000000
2023-10-11 16:35:18,427 epoch 5 - iter 5210/5212 - loss 0.05240335 - time (sec): 1457.17 - samples/sec: 252.09 - lr: 0.000089 - momentum: 0.000000
2023-10-11 16:35:18,890 ----------------------------------------------------------------------------------------------------
2023-10-11 16:35:18,891 EPOCH 5 done: loss 0.0524 - lr: 0.000089
2023-10-11 16:35:57,847 DEV : loss 0.3427276015281677 - f1-score (micro avg) 0.3733
2023-10-11 16:35:57,906 ----------------------------------------------------------------------------------------------------
2023-10-11 16:38:16,413 epoch 6 - iter 521/5212 - loss 0.03453889 - time (sec): 138.50 - samples/sec: 243.67 - lr: 0.000087 - momentum: 0.000000
2023-10-11 16:40:36,448 epoch 6 - iter 1042/5212 - loss 0.03594338 - time (sec): 278.54 - samples/sec: 245.57 - lr: 0.000085 - momentum: 0.000000
2023-10-11 16:42:56,379 epoch 6 - iter 1563/5212 - loss 0.03623246 - time (sec): 418.47 - samples/sec: 250.12 - lr: 0.000084 - momentum: 0.000000
2023-10-11 16:45:13,717 epoch 6 - iter 2084/5212 - loss 0.03614444 - time (sec): 555.81 - samples/sec: 253.09 - lr: 0.000082 - momentum: 0.000000
2023-10-11 16:47:33,364 epoch 6 - iter 2605/5212 - loss 0.03645808 - time (sec): 695.46 - samples/sec: 255.18 - lr: 0.000080 - momentum: 0.000000
2023-10-11 16:49:50,937 epoch 6 - iter 3126/5212 - loss 0.03572045 - time (sec): 833.03 - samples/sec: 255.72 - lr: 0.000078 - momentum: 0.000000
2023-10-11 16:52:13,879 epoch 6 - iter 3647/5212 - loss 0.03557647 - time (sec): 975.97 - samples/sec: 258.76 - lr: 0.000076 - momentum: 0.000000
2023-10-11 16:54:39,082 epoch 6 - iter 4168/5212 - loss 0.03568794 - time (sec): 1121.17 - samples/sec: 258.91 - lr: 0.000075 - momentum: 0.000000
2023-10-11 16:57:09,848 epoch 6 - iter 4689/5212 - loss 0.03525994 - time (sec): 1271.94 - samples/sec: 259.16 - lr: 0.000073 - momentum: 0.000000
2023-10-11 16:59:33,202 epoch 6 - iter 5210/5212 - loss 0.03514270 - time (sec): 1415.29 - samples/sec: 259.42 - lr: 0.000071 - momentum: 0.000000
2023-10-11 16:59:33,838 ----------------------------------------------------------------------------------------------------
2023-10-11 16:59:33,838 EPOCH 6 done: loss 0.0351 - lr: 0.000071
2023-10-11 17:00:14,726 DEV : loss 0.4193594753742218 - f1-score (micro avg) 0.3791
2023-10-11 17:00:14,781 ----------------------------------------------------------------------------------------------------
2023-10-11 17:02:44,618 epoch 7 - iter 521/5212 - loss 0.02269233 - time (sec): 149.83 - samples/sec: 267.66 - lr: 0.000069 - momentum: 0.000000
2023-10-11 17:05:04,738 epoch 7 - iter 1042/5212 - loss 0.02635316 - time (sec): 289.95 - samples/sec: 262.02 - lr: 0.000068 - momentum: 0.000000
2023-10-11 17:07:27,032 epoch 7 - iter 1563/5212 - loss 0.02308396 - time (sec): 432.25 - samples/sec: 262.49 - lr: 0.000066 - momentum: 0.000000
2023-10-11 17:09:48,465 epoch 7 - iter 2084/5212 - loss 0.02400360 - time (sec): 573.68 - samples/sec: 264.91 - lr: 0.000064 - momentum: 0.000000
2023-10-11 17:12:06,814 epoch 7 - iter 2605/5212 - loss 0.02542856 - time (sec): 712.03 - samples/sec: 260.99 - lr: 0.000062 - momentum: 0.000000
2023-10-11 17:14:31,094 epoch 7 - iter 3126/5212 - loss 0.02631394 - time (sec): 856.31 - samples/sec: 260.96 - lr: 0.000060 - momentum: 0.000000
2023-10-11 17:16:53,608 epoch 7 - iter 3647/5212 - loss 0.02640215 - time (sec): 998.82 - samples/sec: 259.79 - lr: 0.000059 - momentum: 0.000000
2023-10-11 17:19:14,821 epoch 7 - iter 4168/5212 - loss 0.02608981 - time (sec): 1140.04 - samples/sec: 258.59 - lr: 0.000057 - momentum: 0.000000
2023-10-11 17:21:40,650 epoch 7 - iter 4689/5212 - loss 0.02628534 - time (sec): 1285.87 - samples/sec: 257.56 - lr: 0.000055 - momentum: 0.000000
2023-10-11 17:24:06,529 epoch 7 - iter 5210/5212 - loss 0.02607886 - time (sec): 1431.75 - samples/sec: 256.60 - lr: 0.000053 - momentum: 0.000000
2023-10-11 17:24:06,965 ----------------------------------------------------------------------------------------------------
2023-10-11 17:24:06,965 EPOCH 7 done: loss 0.0261 - lr: 0.000053
2023-10-11 17:24:47,377 DEV : loss 0.4072570204734802 - f1-score (micro avg) 0.3823
2023-10-11 17:24:47,431 ----------------------------------------------------------------------------------------------------
2023-10-11 17:27:15,714 epoch 8 - iter 521/5212 - loss 0.01684084 - time (sec): 148.28 - samples/sec: 248.06 - lr: 0.000052 - momentum: 0.000000
2023-10-11 17:29:41,615 epoch 8 - iter 1042/5212 - loss 0.01921364 - time (sec): 294.18 - samples/sec: 252.08 - lr: 0.000050 - momentum: 0.000000
2023-10-11 17:32:06,810 epoch 8 - iter 1563/5212 - loss 0.02074029 - time (sec): 439.38 - samples/sec: 250.98 - lr: 0.000048 - momentum: 0.000000
2023-10-11 17:34:30,337 epoch 8 - iter 2084/5212 - loss 0.01866275 - time (sec): 582.90 - samples/sec: 252.36 - lr: 0.000046 - momentum: 0.000000
2023-10-11 17:36:55,146 epoch 8 - iter 2605/5212 - loss 0.01912678 - time (sec): 727.71 - samples/sec: 253.80 - lr: 0.000044 - momentum: 0.000000
2023-10-11 17:39:18,799 epoch 8 - iter 3126/5212 - loss 0.01960685 - time (sec): 871.37 - samples/sec: 253.85 - lr: 0.000043 - momentum: 0.000000
2023-10-11 17:41:42,158 epoch 8 - iter 3647/5212 - loss 0.01902745 - time (sec): 1014.72 - samples/sec: 253.30 - lr: 0.000041 - momentum: 0.000000
2023-10-11 17:44:07,681 epoch 8 - iter 4168/5212 - loss 0.01851736 - time (sec): 1160.25 - samples/sec: 253.42 - lr: 0.000039 - momentum: 0.000000
2023-10-11 17:46:34,426 epoch 8 - iter 4689/5212 - loss 0.01823475 - time (sec): 1306.99 - samples/sec: 253.91 - lr: 0.000037 - momentum: 0.000000
2023-10-11 17:48:56,985 epoch 8 - iter 5210/5212 - loss 0.01875215 - time (sec): 1449.55 - samples/sec: 253.25 - lr: 0.000036 - momentum: 0.000000
2023-10-11 17:48:57,706 ----------------------------------------------------------------------------------------------------
2023-10-11 17:48:57,707 EPOCH 8 done: loss 0.0187 - lr: 0.000036
2023-10-11 17:49:37,133 DEV : loss 0.41436994075775146 - f1-score (micro avg) 0.4115
2023-10-11 17:49:37,186 saving best model
2023-10-11 17:49:39,805 ----------------------------------------------------------------------------------------------------
2023-10-11 17:52:05,310 epoch 9 - iter 521/5212 - loss 0.01196532 - time (sec): 145.50 - samples/sec: 263.89 - lr: 0.000034 - momentum: 0.000000
2023-10-11 17:54:27,376 epoch 9 - iter 1042/5212 - loss 0.01333594 - time (sec): 287.57 - samples/sec: 264.07 - lr: 0.000032 - momentum: 0.000000
2023-10-11 17:56:48,142 epoch 9 - iter 1563/5212 - loss 0.01211283 - time (sec): 428.33 - samples/sec: 260.34 - lr: 0.000030 - momentum: 0.000000
2023-10-11 17:59:07,321 epoch 9 - iter 2084/5212 - loss 0.01222749 - time (sec): 567.51 - samples/sec: 257.43 - lr: 0.000028 - momentum: 0.000000
2023-10-11 18:01:27,565 epoch 9 - iter 2605/5212 - loss 0.01192194 - time (sec): 707.76 - samples/sec: 259.45 - lr: 0.000027 - momentum: 0.000000
2023-10-11 18:03:45,920 epoch 9 - iter 3126/5212 - loss 0.01186385 - time (sec): 846.11 - samples/sec: 259.03 - lr: 0.000025 - momentum: 0.000000
2023-10-11 18:06:04,727 epoch 9 - iter 3647/5212 - loss 0.01241909 - time (sec): 984.92 - samples/sec: 260.05 - lr: 0.000023 - momentum: 0.000000
2023-10-11 18:08:24,518 epoch 9 - iter 4168/5212 - loss 0.01204146 - time (sec): 1124.71 - samples/sec: 261.02 - lr: 0.000021 - momentum: 0.000000
2023-10-11 18:10:47,837 epoch 9 - iter 4689/5212 - loss 0.01273801 - time (sec): 1268.03 - samples/sec: 260.93 - lr: 0.000020 - momentum: 0.000000
2023-10-11 18:13:07,309 epoch 9 - iter 5210/5212 - loss 0.01307196 - time (sec): 1407.50 - samples/sec: 260.86 - lr: 0.000018 - momentum: 0.000000
2023-10-11 18:13:07,901 ----------------------------------------------------------------------------------------------------
2023-10-11 18:13:07,901 EPOCH 9 done: loss 0.0131 - lr: 0.000018
2023-10-11 18:13:46,547 DEV : loss 0.4808931350708008 - f1-score (micro avg) 0.3967
2023-10-11 18:13:46,607 ----------------------------------------------------------------------------------------------------
2023-10-11 18:16:05,453 epoch 10 - iter 521/5212 - loss 0.00668555 - time (sec): 138.84 - samples/sec: 262.74 - lr: 0.000016 - momentum: 0.000000
2023-10-11 18:18:22,229 epoch 10 - iter 1042/5212 - loss 0.00706135 - time (sec): 275.62 - samples/sec: 260.81 - lr: 0.000014 - momentum: 0.000000
2023-10-11 18:20:40,291 epoch 10 - iter 1563/5212 - loss 0.00669620 - time (sec): 413.68 - samples/sec: 262.77 - lr: 0.000012 - momentum: 0.000000
2023-10-11 18:22:57,132 epoch 10 - iter 2084/5212 - loss 0.00716732 - time (sec): 550.52 - samples/sec: 261.09 - lr: 0.000011 - momentum: 0.000000
2023-10-11 18:25:18,034 epoch 10 - iter 2605/5212 - loss 0.00737559 - time (sec): 691.43 - samples/sec: 264.83 - lr: 0.000009 - momentum: 0.000000
2023-10-11 18:27:35,108 epoch 10 - iter 3126/5212 - loss 0.00746666 - time (sec): 828.50 - samples/sec: 264.40 - lr: 0.000007 - momentum: 0.000000
2023-10-11 18:29:52,239 epoch 10 - iter 3647/5212 - loss 0.00801610 - time (sec): 965.63 - samples/sec: 264.55 - lr: 0.000005 - momentum: 0.000000
2023-10-11 18:32:08,536 epoch 10 - iter 4168/5212 - loss 0.00797516 - time (sec): 1101.93 - samples/sec: 263.54 - lr: 0.000004 - momentum: 0.000000
2023-10-11 18:34:30,744 epoch 10 - iter 4689/5212 - loss 0.00813383 - time (sec): 1244.13 - samples/sec: 265.19 - lr: 0.000002 - momentum: 0.000000
2023-10-11 18:36:51,864 epoch 10 - iter 5210/5212 - loss 0.00803504 - time (sec): 1385.26 - samples/sec: 265.22 - lr: 0.000000 - momentum: 0.000000
2023-10-11 18:36:52,264 ----------------------------------------------------------------------------------------------------
2023-10-11 18:36:52,265 EPOCH 10 done: loss 0.0080 - lr: 0.000000
2023-10-11 18:37:30,395 DEV : loss 0.47644102573394775 - f1-score (micro avg) 0.4056
2023-10-11 18:37:31,324 ----------------------------------------------------------------------------------------------------
2023-10-11 18:37:31,326 Loading model from best epoch ...
2023-10-11 18:37:35,775 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
2023-10-11 18:39:14,485
Results:
- F-score (micro) 0.4756
- F-score (macro) 0.3235
- Accuracy 0.317
By class:
precision recall f1-score support
LOC 0.4876 0.5972 0.5368 1214
PER 0.4418 0.4790 0.4596 808
ORG 0.2923 0.3031 0.2976 353
HumanProd 0.0000 0.0000 0.0000 15
micro avg 0.4455 0.5100 0.4756 2390
macro avg 0.3054 0.3448 0.3235 2390
weighted avg 0.4402 0.5100 0.4720 2390
2023-10-11 18:39:14,485 ----------------------------------------------------------------------------------------------------
|