File size: 23,972 Bytes
8ac8159 |
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 |
2023-10-17 17:34:10,698 ----------------------------------------------------------------------------------------------------
2023-10-17 17:34:10,699 Model: "SequenceTagger(
(embeddings): TransformerWordEmbeddings(
(model): ElectraModel(
(embeddings): ElectraEmbeddings(
(word_embeddings): Embedding(32001, 768)
(position_embeddings): Embedding(512, 768)
(token_type_embeddings): Embedding(2, 768)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(encoder): ElectraEncoder(
(layer): ModuleList(
(0-11): 12 x ElectraLayer(
(attention): ElectraAttention(
(self): ElectraSelfAttention(
(query): Linear(in_features=768, out_features=768, bias=True)
(key): Linear(in_features=768, out_features=768, bias=True)
(value): Linear(in_features=768, out_features=768, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
)
(output): ElectraSelfOutput(
(dense): Linear(in_features=768, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
(intermediate): ElectraIntermediate(
(dense): Linear(in_features=768, out_features=3072, bias=True)
(intermediate_act_fn): GELUActivation()
)
(output): ElectraOutput(
(dense): Linear(in_features=3072, out_features=768, bias=True)
(LayerNorm): LayerNorm((768,), eps=1e-12, elementwise_affine=True)
(dropout): Dropout(p=0.1, inplace=False)
)
)
)
)
)
)
(locked_dropout): LockedDropout(p=0.5)
(linear): Linear(in_features=768, out_features=17, bias=True)
(loss_function): CrossEntropyLoss()
)"
2023-10-17 17:34:10,699 ----------------------------------------------------------------------------------------------------
2023-10-17 17:34:10,699 MultiCorpus: 1166 train + 165 dev + 415 test sentences
- NER_HIPE_2022 Corpus: 1166 train + 165 dev + 415 test sentences - /root/.flair/datasets/ner_hipe_2022/v2.1/newseye/fi/with_doc_seperator
2023-10-17 17:34:10,699 ----------------------------------------------------------------------------------------------------
2023-10-17 17:34:10,699 Train: 1166 sentences
2023-10-17 17:34:10,699 (train_with_dev=False, train_with_test=False)
2023-10-17 17:34:10,699 ----------------------------------------------------------------------------------------------------
2023-10-17 17:34:10,699 Training Params:
2023-10-17 17:34:10,699 - learning_rate: "3e-05"
2023-10-17 17:34:10,700 - mini_batch_size: "8"
2023-10-17 17:34:10,700 - max_epochs: "10"
2023-10-17 17:34:10,700 - shuffle: "True"
2023-10-17 17:34:10,700 ----------------------------------------------------------------------------------------------------
2023-10-17 17:34:10,700 Plugins:
2023-10-17 17:34:10,700 - TensorboardLogger
2023-10-17 17:34:10,700 - LinearScheduler | warmup_fraction: '0.1'
2023-10-17 17:34:10,700 ----------------------------------------------------------------------------------------------------
2023-10-17 17:34:10,700 Final evaluation on model from best epoch (best-model.pt)
2023-10-17 17:34:10,700 - metric: "('micro avg', 'f1-score')"
2023-10-17 17:34:10,700 ----------------------------------------------------------------------------------------------------
2023-10-17 17:34:10,700 Computation:
2023-10-17 17:34:10,700 - compute on device: cuda:0
2023-10-17 17:34:10,700 - embedding storage: none
2023-10-17 17:34:10,700 ----------------------------------------------------------------------------------------------------
2023-10-17 17:34:10,700 Model training base path: "hmbench-newseye/fi-hmteams/teams-base-historic-multilingual-discriminator-bs8-wsFalse-e10-lr3e-05-poolingfirst-layers-1-crfFalse-1"
2023-10-17 17:34:10,700 ----------------------------------------------------------------------------------------------------
2023-10-17 17:34:10,700 ----------------------------------------------------------------------------------------------------
2023-10-17 17:34:10,700 Logging anything other than scalars to TensorBoard is currently not supported.
2023-10-17 17:34:12,102 epoch 1 - iter 14/146 - loss 3.41781629 - time (sec): 1.40 - samples/sec: 2706.86 - lr: 0.000003 - momentum: 0.000000
2023-10-17 17:34:13,479 epoch 1 - iter 28/146 - loss 3.19523070 - time (sec): 2.78 - samples/sec: 2743.78 - lr: 0.000006 - momentum: 0.000000
2023-10-17 17:34:15,696 epoch 1 - iter 42/146 - loss 2.72247714 - time (sec): 5.00 - samples/sec: 2741.67 - lr: 0.000008 - momentum: 0.000000
2023-10-17 17:34:17,030 epoch 1 - iter 56/146 - loss 2.24817868 - time (sec): 6.33 - samples/sec: 2826.85 - lr: 0.000011 - momentum: 0.000000
2023-10-17 17:34:18,216 epoch 1 - iter 70/146 - loss 1.97425059 - time (sec): 7.52 - samples/sec: 2872.38 - lr: 0.000014 - momentum: 0.000000
2023-10-17 17:34:19,771 epoch 1 - iter 84/146 - loss 1.76062958 - time (sec): 9.07 - samples/sec: 2841.10 - lr: 0.000017 - momentum: 0.000000
2023-10-17 17:34:21,105 epoch 1 - iter 98/146 - loss 1.56691924 - time (sec): 10.40 - samples/sec: 2868.07 - lr: 0.000020 - momentum: 0.000000
2023-10-17 17:34:22,445 epoch 1 - iter 112/146 - loss 1.41624902 - time (sec): 11.74 - samples/sec: 2905.87 - lr: 0.000023 - momentum: 0.000000
2023-10-17 17:34:23,861 epoch 1 - iter 126/146 - loss 1.29548011 - time (sec): 13.16 - samples/sec: 2919.54 - lr: 0.000026 - momentum: 0.000000
2023-10-17 17:34:25,272 epoch 1 - iter 140/146 - loss 1.20756763 - time (sec): 14.57 - samples/sec: 2901.50 - lr: 0.000029 - momentum: 0.000000
2023-10-17 17:34:25,909 ----------------------------------------------------------------------------------------------------
2023-10-17 17:34:25,909 EPOCH 1 done: loss 1.1629 - lr: 0.000029
2023-10-17 17:34:26,796 DEV : loss 0.21301493048667908 - f1-score (micro avg) 0.4358
2023-10-17 17:34:26,804 saving best model
2023-10-17 17:34:27,174 ----------------------------------------------------------------------------------------------------
2023-10-17 17:34:29,054 epoch 2 - iter 14/146 - loss 0.26935291 - time (sec): 1.88 - samples/sec: 2422.00 - lr: 0.000030 - momentum: 0.000000
2023-10-17 17:34:30,577 epoch 2 - iter 28/146 - loss 0.23373936 - time (sec): 3.40 - samples/sec: 2491.77 - lr: 0.000029 - momentum: 0.000000
2023-10-17 17:34:31,942 epoch 2 - iter 42/146 - loss 0.22380424 - time (sec): 4.77 - samples/sec: 2659.63 - lr: 0.000029 - momentum: 0.000000
2023-10-17 17:34:33,819 epoch 2 - iter 56/146 - loss 0.20873831 - time (sec): 6.64 - samples/sec: 2685.69 - lr: 0.000029 - momentum: 0.000000
2023-10-17 17:34:35,150 epoch 2 - iter 70/146 - loss 0.20731025 - time (sec): 7.97 - samples/sec: 2772.54 - lr: 0.000028 - momentum: 0.000000
2023-10-17 17:34:36,476 epoch 2 - iter 84/146 - loss 0.20314886 - time (sec): 9.30 - samples/sec: 2888.49 - lr: 0.000028 - momentum: 0.000000
2023-10-17 17:34:37,622 epoch 2 - iter 98/146 - loss 0.20692578 - time (sec): 10.45 - samples/sec: 2901.97 - lr: 0.000028 - momentum: 0.000000
2023-10-17 17:34:38,837 epoch 2 - iter 112/146 - loss 0.21286406 - time (sec): 11.66 - samples/sec: 2898.05 - lr: 0.000027 - momentum: 0.000000
2023-10-17 17:34:40,431 epoch 2 - iter 126/146 - loss 0.21773797 - time (sec): 13.26 - samples/sec: 2902.32 - lr: 0.000027 - momentum: 0.000000
2023-10-17 17:34:42,013 epoch 2 - iter 140/146 - loss 0.21049606 - time (sec): 14.84 - samples/sec: 2882.52 - lr: 0.000027 - momentum: 0.000000
2023-10-17 17:34:42,634 ----------------------------------------------------------------------------------------------------
2023-10-17 17:34:42,634 EPOCH 2 done: loss 0.2088 - lr: 0.000027
2023-10-17 17:34:43,905 DEV : loss 0.12763920426368713 - f1-score (micro avg) 0.6166
2023-10-17 17:34:43,910 saving best model
2023-10-17 17:34:44,373 ----------------------------------------------------------------------------------------------------
2023-10-17 17:34:45,986 epoch 3 - iter 14/146 - loss 0.13733881 - time (sec): 1.61 - samples/sec: 2783.38 - lr: 0.000026 - momentum: 0.000000
2023-10-17 17:34:47,467 epoch 3 - iter 28/146 - loss 0.12832365 - time (sec): 3.09 - samples/sec: 2830.11 - lr: 0.000026 - momentum: 0.000000
2023-10-17 17:34:48,938 epoch 3 - iter 42/146 - loss 0.13952727 - time (sec): 4.56 - samples/sec: 2899.24 - lr: 0.000026 - momentum: 0.000000
2023-10-17 17:34:50,224 epoch 3 - iter 56/146 - loss 0.13596454 - time (sec): 5.85 - samples/sec: 2943.42 - lr: 0.000025 - momentum: 0.000000
2023-10-17 17:34:51,562 epoch 3 - iter 70/146 - loss 0.13330878 - time (sec): 7.19 - samples/sec: 2939.38 - lr: 0.000025 - momentum: 0.000000
2023-10-17 17:34:52,922 epoch 3 - iter 84/146 - loss 0.12656844 - time (sec): 8.55 - samples/sec: 2946.73 - lr: 0.000025 - momentum: 0.000000
2023-10-17 17:34:54,523 epoch 3 - iter 98/146 - loss 0.12323148 - time (sec): 10.15 - samples/sec: 2950.36 - lr: 0.000024 - momentum: 0.000000
2023-10-17 17:34:56,140 epoch 3 - iter 112/146 - loss 0.12354407 - time (sec): 11.76 - samples/sec: 2947.41 - lr: 0.000024 - momentum: 0.000000
2023-10-17 17:34:57,428 epoch 3 - iter 126/146 - loss 0.12251103 - time (sec): 13.05 - samples/sec: 2950.33 - lr: 0.000024 - momentum: 0.000000
2023-10-17 17:34:58,780 epoch 3 - iter 140/146 - loss 0.12318821 - time (sec): 14.40 - samples/sec: 2926.48 - lr: 0.000024 - momentum: 0.000000
2023-10-17 17:34:59,602 ----------------------------------------------------------------------------------------------------
2023-10-17 17:34:59,602 EPOCH 3 done: loss 0.1235 - lr: 0.000024
2023-10-17 17:35:00,844 DEV : loss 0.11239827424287796 - f1-score (micro avg) 0.7169
2023-10-17 17:35:00,848 saving best model
2023-10-17 17:35:01,320 ----------------------------------------------------------------------------------------------------
2023-10-17 17:35:02,428 epoch 4 - iter 14/146 - loss 0.10523599 - time (sec): 1.11 - samples/sec: 3124.48 - lr: 0.000023 - momentum: 0.000000
2023-10-17 17:35:03,652 epoch 4 - iter 28/146 - loss 0.09148276 - time (sec): 2.33 - samples/sec: 3164.43 - lr: 0.000023 - momentum: 0.000000
2023-10-17 17:35:05,339 epoch 4 - iter 42/146 - loss 0.07551795 - time (sec): 4.02 - samples/sec: 3099.36 - lr: 0.000022 - momentum: 0.000000
2023-10-17 17:35:06,930 epoch 4 - iter 56/146 - loss 0.07430281 - time (sec): 5.61 - samples/sec: 2987.42 - lr: 0.000022 - momentum: 0.000000
2023-10-17 17:35:08,428 epoch 4 - iter 70/146 - loss 0.07274444 - time (sec): 7.11 - samples/sec: 2959.51 - lr: 0.000022 - momentum: 0.000000
2023-10-17 17:35:09,989 epoch 4 - iter 84/146 - loss 0.07061816 - time (sec): 8.67 - samples/sec: 2956.86 - lr: 0.000021 - momentum: 0.000000
2023-10-17 17:35:11,818 epoch 4 - iter 98/146 - loss 0.07687115 - time (sec): 10.50 - samples/sec: 2916.54 - lr: 0.000021 - momentum: 0.000000
2023-10-17 17:35:13,168 epoch 4 - iter 112/146 - loss 0.07669767 - time (sec): 11.85 - samples/sec: 2902.91 - lr: 0.000021 - momentum: 0.000000
2023-10-17 17:35:14,638 epoch 4 - iter 126/146 - loss 0.08004943 - time (sec): 13.32 - samples/sec: 2924.14 - lr: 0.000021 - momentum: 0.000000
2023-10-17 17:35:16,115 epoch 4 - iter 140/146 - loss 0.07957329 - time (sec): 14.79 - samples/sec: 2899.39 - lr: 0.000020 - momentum: 0.000000
2023-10-17 17:35:16,652 ----------------------------------------------------------------------------------------------------
2023-10-17 17:35:16,652 EPOCH 4 done: loss 0.0787 - lr: 0.000020
2023-10-17 17:35:17,960 DEV : loss 0.12767371535301208 - f1-score (micro avg) 0.7347
2023-10-17 17:35:17,965 saving best model
2023-10-17 17:35:18,422 ----------------------------------------------------------------------------------------------------
2023-10-17 17:35:20,198 epoch 5 - iter 14/146 - loss 0.06713213 - time (sec): 1.77 - samples/sec: 2816.55 - lr: 0.000020 - momentum: 0.000000
2023-10-17 17:35:22,147 epoch 5 - iter 28/146 - loss 0.05002344 - time (sec): 3.72 - samples/sec: 2629.73 - lr: 0.000019 - momentum: 0.000000
2023-10-17 17:35:23,492 epoch 5 - iter 42/146 - loss 0.06121285 - time (sec): 5.07 - samples/sec: 2780.50 - lr: 0.000019 - momentum: 0.000000
2023-10-17 17:35:24,688 epoch 5 - iter 56/146 - loss 0.06588173 - time (sec): 6.26 - samples/sec: 2820.49 - lr: 0.000019 - momentum: 0.000000
2023-10-17 17:35:26,222 epoch 5 - iter 70/146 - loss 0.06594720 - time (sec): 7.80 - samples/sec: 2802.19 - lr: 0.000018 - momentum: 0.000000
2023-10-17 17:35:27,674 epoch 5 - iter 84/146 - loss 0.06721579 - time (sec): 9.25 - samples/sec: 2827.08 - lr: 0.000018 - momentum: 0.000000
2023-10-17 17:35:29,074 epoch 5 - iter 98/146 - loss 0.06330378 - time (sec): 10.65 - samples/sec: 2880.43 - lr: 0.000018 - momentum: 0.000000
2023-10-17 17:35:30,422 epoch 5 - iter 112/146 - loss 0.06266183 - time (sec): 12.00 - samples/sec: 2895.08 - lr: 0.000018 - momentum: 0.000000
2023-10-17 17:35:31,794 epoch 5 - iter 126/146 - loss 0.06018288 - time (sec): 13.37 - samples/sec: 2886.31 - lr: 0.000017 - momentum: 0.000000
2023-10-17 17:35:33,215 epoch 5 - iter 140/146 - loss 0.05869652 - time (sec): 14.79 - samples/sec: 2898.02 - lr: 0.000017 - momentum: 0.000000
2023-10-17 17:35:33,682 ----------------------------------------------------------------------------------------------------
2023-10-17 17:35:33,683 EPOCH 5 done: loss 0.0575 - lr: 0.000017
2023-10-17 17:35:35,033 DEV : loss 0.11119718104600906 - f1-score (micro avg) 0.7366
2023-10-17 17:35:35,042 saving best model
2023-10-17 17:35:35,505 ----------------------------------------------------------------------------------------------------
2023-10-17 17:35:36,875 epoch 6 - iter 14/146 - loss 0.04824613 - time (sec): 1.37 - samples/sec: 3188.47 - lr: 0.000016 - momentum: 0.000000
2023-10-17 17:35:38,086 epoch 6 - iter 28/146 - loss 0.03975259 - time (sec): 2.58 - samples/sec: 3016.19 - lr: 0.000016 - momentum: 0.000000
2023-10-17 17:35:39,837 epoch 6 - iter 42/146 - loss 0.04375931 - time (sec): 4.33 - samples/sec: 2885.12 - lr: 0.000016 - momentum: 0.000000
2023-10-17 17:35:41,314 epoch 6 - iter 56/146 - loss 0.04373476 - time (sec): 5.81 - samples/sec: 2902.21 - lr: 0.000015 - momentum: 0.000000
2023-10-17 17:35:42,542 epoch 6 - iter 70/146 - loss 0.04210856 - time (sec): 7.04 - samples/sec: 2889.54 - lr: 0.000015 - momentum: 0.000000
2023-10-17 17:35:44,084 epoch 6 - iter 84/146 - loss 0.04064450 - time (sec): 8.58 - samples/sec: 2827.82 - lr: 0.000015 - momentum: 0.000000
2023-10-17 17:35:45,532 epoch 6 - iter 98/146 - loss 0.04063501 - time (sec): 10.03 - samples/sec: 2811.44 - lr: 0.000015 - momentum: 0.000000
2023-10-17 17:35:47,178 epoch 6 - iter 112/146 - loss 0.04206829 - time (sec): 11.67 - samples/sec: 2824.16 - lr: 0.000014 - momentum: 0.000000
2023-10-17 17:35:48,636 epoch 6 - iter 126/146 - loss 0.04214994 - time (sec): 13.13 - samples/sec: 2847.92 - lr: 0.000014 - momentum: 0.000000
2023-10-17 17:35:50,492 epoch 6 - iter 140/146 - loss 0.04140993 - time (sec): 14.99 - samples/sec: 2860.81 - lr: 0.000014 - momentum: 0.000000
2023-10-17 17:35:51,009 ----------------------------------------------------------------------------------------------------
2023-10-17 17:35:51,009 EPOCH 6 done: loss 0.0411 - lr: 0.000014
2023-10-17 17:35:52,347 DEV : loss 0.12232689559459686 - f1-score (micro avg) 0.7387
2023-10-17 17:35:52,353 saving best model
2023-10-17 17:35:52,811 ----------------------------------------------------------------------------------------------------
2023-10-17 17:35:54,469 epoch 7 - iter 14/146 - loss 0.02256882 - time (sec): 1.66 - samples/sec: 2879.40 - lr: 0.000013 - momentum: 0.000000
2023-10-17 17:35:56,042 epoch 7 - iter 28/146 - loss 0.03288779 - time (sec): 3.23 - samples/sec: 2693.38 - lr: 0.000013 - momentum: 0.000000
2023-10-17 17:35:57,619 epoch 7 - iter 42/146 - loss 0.03598142 - time (sec): 4.81 - samples/sec: 2800.42 - lr: 0.000012 - momentum: 0.000000
2023-10-17 17:35:59,128 epoch 7 - iter 56/146 - loss 0.03345291 - time (sec): 6.31 - samples/sec: 2852.01 - lr: 0.000012 - momentum: 0.000000
2023-10-17 17:36:00,762 epoch 7 - iter 70/146 - loss 0.03144006 - time (sec): 7.95 - samples/sec: 2865.15 - lr: 0.000012 - momentum: 0.000000
2023-10-17 17:36:02,090 epoch 7 - iter 84/146 - loss 0.03099728 - time (sec): 9.28 - samples/sec: 2910.56 - lr: 0.000012 - momentum: 0.000000
2023-10-17 17:36:03,582 epoch 7 - iter 98/146 - loss 0.03123906 - time (sec): 10.77 - samples/sec: 2920.04 - lr: 0.000011 - momentum: 0.000000
2023-10-17 17:36:04,878 epoch 7 - iter 112/146 - loss 0.03086938 - time (sec): 12.06 - samples/sec: 2954.52 - lr: 0.000011 - momentum: 0.000000
2023-10-17 17:36:06,038 epoch 7 - iter 126/146 - loss 0.02983269 - time (sec): 13.23 - samples/sec: 2952.40 - lr: 0.000011 - momentum: 0.000000
2023-10-17 17:36:07,324 epoch 7 - iter 140/146 - loss 0.03052926 - time (sec): 14.51 - samples/sec: 2955.31 - lr: 0.000010 - momentum: 0.000000
2023-10-17 17:36:07,903 ----------------------------------------------------------------------------------------------------
2023-10-17 17:36:07,903 EPOCH 7 done: loss 0.0303 - lr: 0.000010
2023-10-17 17:36:09,300 DEV : loss 0.11899629980325699 - f1-score (micro avg) 0.7489
2023-10-17 17:36:09,306 saving best model
2023-10-17 17:36:09,770 ----------------------------------------------------------------------------------------------------
2023-10-17 17:36:11,285 epoch 8 - iter 14/146 - loss 0.01724715 - time (sec): 1.51 - samples/sec: 2998.30 - lr: 0.000010 - momentum: 0.000000
2023-10-17 17:36:12,868 epoch 8 - iter 28/146 - loss 0.02957612 - time (sec): 3.10 - samples/sec: 2991.71 - lr: 0.000009 - momentum: 0.000000
2023-10-17 17:36:14,548 epoch 8 - iter 42/146 - loss 0.02774801 - time (sec): 4.78 - samples/sec: 2859.21 - lr: 0.000009 - momentum: 0.000000
2023-10-17 17:36:15,854 epoch 8 - iter 56/146 - loss 0.03053498 - time (sec): 6.08 - samples/sec: 2902.91 - lr: 0.000009 - momentum: 0.000000
2023-10-17 17:36:17,216 epoch 8 - iter 70/146 - loss 0.02811800 - time (sec): 7.44 - samples/sec: 2871.92 - lr: 0.000009 - momentum: 0.000000
2023-10-17 17:36:18,479 epoch 8 - iter 84/146 - loss 0.02621154 - time (sec): 8.71 - samples/sec: 2883.41 - lr: 0.000008 - momentum: 0.000000
2023-10-17 17:36:19,931 epoch 8 - iter 98/146 - loss 0.02607277 - time (sec): 10.16 - samples/sec: 2892.05 - lr: 0.000008 - momentum: 0.000000
2023-10-17 17:36:21,382 epoch 8 - iter 112/146 - loss 0.02474799 - time (sec): 11.61 - samples/sec: 2879.71 - lr: 0.000008 - momentum: 0.000000
2023-10-17 17:36:22,964 epoch 8 - iter 126/146 - loss 0.02303571 - time (sec): 13.19 - samples/sec: 2861.56 - lr: 0.000007 - momentum: 0.000000
2023-10-17 17:36:24,668 epoch 8 - iter 140/146 - loss 0.02411078 - time (sec): 14.90 - samples/sec: 2877.30 - lr: 0.000007 - momentum: 0.000000
2023-10-17 17:36:25,140 ----------------------------------------------------------------------------------------------------
2023-10-17 17:36:25,140 EPOCH 8 done: loss 0.0239 - lr: 0.000007
2023-10-17 17:36:26,405 DEV : loss 0.1307678520679474 - f1-score (micro avg) 0.7686
2023-10-17 17:36:26,410 saving best model
2023-10-17 17:36:26,879 ----------------------------------------------------------------------------------------------------
2023-10-17 17:36:28,444 epoch 9 - iter 14/146 - loss 0.01901972 - time (sec): 1.56 - samples/sec: 2641.86 - lr: 0.000006 - momentum: 0.000000
2023-10-17 17:36:29,830 epoch 9 - iter 28/146 - loss 0.02385709 - time (sec): 2.95 - samples/sec: 2719.97 - lr: 0.000006 - momentum: 0.000000
2023-10-17 17:36:31,047 epoch 9 - iter 42/146 - loss 0.02182791 - time (sec): 4.16 - samples/sec: 2735.27 - lr: 0.000006 - momentum: 0.000000
2023-10-17 17:36:32,276 epoch 9 - iter 56/146 - loss 0.01826679 - time (sec): 5.39 - samples/sec: 2787.55 - lr: 0.000006 - momentum: 0.000000
2023-10-17 17:36:33,957 epoch 9 - iter 70/146 - loss 0.01817023 - time (sec): 7.08 - samples/sec: 2849.20 - lr: 0.000005 - momentum: 0.000000
2023-10-17 17:36:35,421 epoch 9 - iter 84/146 - loss 0.01820976 - time (sec): 8.54 - samples/sec: 2893.14 - lr: 0.000005 - momentum: 0.000000
2023-10-17 17:36:36,738 epoch 9 - iter 98/146 - loss 0.01827366 - time (sec): 9.86 - samples/sec: 2888.66 - lr: 0.000005 - momentum: 0.000000
2023-10-17 17:36:38,078 epoch 9 - iter 112/146 - loss 0.01689217 - time (sec): 11.20 - samples/sec: 2925.35 - lr: 0.000004 - momentum: 0.000000
2023-10-17 17:36:39,820 epoch 9 - iter 126/146 - loss 0.01694711 - time (sec): 12.94 - samples/sec: 2937.51 - lr: 0.000004 - momentum: 0.000000
2023-10-17 17:36:41,252 epoch 9 - iter 140/146 - loss 0.01782277 - time (sec): 14.37 - samples/sec: 2934.83 - lr: 0.000004 - momentum: 0.000000
2023-10-17 17:36:42,148 ----------------------------------------------------------------------------------------------------
2023-10-17 17:36:42,148 EPOCH 9 done: loss 0.0181 - lr: 0.000004
2023-10-17 17:36:43,479 DEV : loss 0.13454537093639374 - f1-score (micro avg) 0.7615
2023-10-17 17:36:43,487 ----------------------------------------------------------------------------------------------------
2023-10-17 17:36:45,042 epoch 10 - iter 14/146 - loss 0.01570550 - time (sec): 1.55 - samples/sec: 3010.72 - lr: 0.000003 - momentum: 0.000000
2023-10-17 17:36:46,505 epoch 10 - iter 28/146 - loss 0.01766356 - time (sec): 3.02 - samples/sec: 3095.27 - lr: 0.000003 - momentum: 0.000000
2023-10-17 17:36:47,949 epoch 10 - iter 42/146 - loss 0.02308068 - time (sec): 4.46 - samples/sec: 3045.17 - lr: 0.000003 - momentum: 0.000000
2023-10-17 17:36:49,330 epoch 10 - iter 56/146 - loss 0.01941746 - time (sec): 5.84 - samples/sec: 2986.50 - lr: 0.000002 - momentum: 0.000000
2023-10-17 17:36:50,825 epoch 10 - iter 70/146 - loss 0.01612324 - time (sec): 7.34 - samples/sec: 2950.05 - lr: 0.000002 - momentum: 0.000000
2023-10-17 17:36:52,332 epoch 10 - iter 84/146 - loss 0.01574558 - time (sec): 8.84 - samples/sec: 2940.92 - lr: 0.000002 - momentum: 0.000000
2023-10-17 17:36:53,845 epoch 10 - iter 98/146 - loss 0.01635377 - time (sec): 10.36 - samples/sec: 2912.53 - lr: 0.000001 - momentum: 0.000000
2023-10-17 17:36:55,386 epoch 10 - iter 112/146 - loss 0.01613895 - time (sec): 11.90 - samples/sec: 2916.89 - lr: 0.000001 - momentum: 0.000000
2023-10-17 17:36:56,988 epoch 10 - iter 126/146 - loss 0.01578624 - time (sec): 13.50 - samples/sec: 2917.22 - lr: 0.000001 - momentum: 0.000000
2023-10-17 17:36:58,336 epoch 10 - iter 140/146 - loss 0.01493481 - time (sec): 14.85 - samples/sec: 2919.38 - lr: 0.000000 - momentum: 0.000000
2023-10-17 17:36:58,751 ----------------------------------------------------------------------------------------------------
2023-10-17 17:36:58,751 EPOCH 10 done: loss 0.0147 - lr: 0.000000
2023-10-17 17:37:00,104 DEV : loss 0.13558898866176605 - f1-score (micro avg) 0.7582
2023-10-17 17:37:00,455 ----------------------------------------------------------------------------------------------------
2023-10-17 17:37:00,456 Loading model from best epoch ...
2023-10-17 17:37:01,822 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
2023-10-17 17:37:04,514
Results:
- F-score (micro) 0.7634
- F-score (macro) 0.6812
- Accuracy 0.635
By class:
precision recall f1-score support
PER 0.8362 0.8506 0.8433 348
LOC 0.6292 0.8582 0.7261 261
ORG 0.4737 0.3462 0.4000 52
HumanProd 0.7391 0.7727 0.7556 22
micro avg 0.7198 0.8126 0.7634 683
macro avg 0.6695 0.7069 0.6812 683
weighted avg 0.7264 0.8126 0.7619 683
2023-10-17 17:37:04,515 ----------------------------------------------------------------------------------------------------
|