File size: 24,107 Bytes
87b5f20 |
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 |
2023-10-23 14:59:24,306 ----------------------------------------------------------------------------------------------------
2023-10-23 14:59:24,307 Model: "SequenceTagger(
(embeddings): TransformerWordEmbeddings(
(model): BertModel(
(embeddings): BertEmbeddings(
(word_embeddings): Embedding(64001, 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): BertEncoder(
(layer): ModuleList(
(0-11): 12 x BertLayer(
(attention): BertAttention(
(self): BertSelfAttention(
(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): BertSelfOutput(
(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): BertIntermediate(
(dense): Linear(in_features=768, out_features=3072, bias=True)
(intermediate_act_fn): GELUActivation()
)
(output): BertOutput(
(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)
)
)
)
)
(pooler): BertPooler(
(dense): Linear(in_features=768, out_features=768, bias=True)
(activation): Tanh()
)
)
)
(locked_dropout): LockedDropout(p=0.5)
(linear): Linear(in_features=768, out_features=25, bias=True)
(loss_function): CrossEntropyLoss()
)"
2023-10-23 14:59:24,307 ----------------------------------------------------------------------------------------------------
2023-10-23 14:59:24,307 MultiCorpus: 1100 train + 206 dev + 240 test sentences
- NER_HIPE_2022 Corpus: 1100 train + 206 dev + 240 test sentences - /root/.flair/datasets/ner_hipe_2022/v2.1/ajmc/de/with_doc_seperator
2023-10-23 14:59:24,307 ----------------------------------------------------------------------------------------------------
2023-10-23 14:59:24,307 Train: 1100 sentences
2023-10-23 14:59:24,307 (train_with_dev=False, train_with_test=False)
2023-10-23 14:59:24,307 ----------------------------------------------------------------------------------------------------
2023-10-23 14:59:24,307 Training Params:
2023-10-23 14:59:24,307 - learning_rate: "5e-05"
2023-10-23 14:59:24,307 - mini_batch_size: "8"
2023-10-23 14:59:24,307 - max_epochs: "10"
2023-10-23 14:59:24,307 - shuffle: "True"
2023-10-23 14:59:24,307 ----------------------------------------------------------------------------------------------------
2023-10-23 14:59:24,308 Plugins:
2023-10-23 14:59:24,308 - TensorboardLogger
2023-10-23 14:59:24,308 - LinearScheduler | warmup_fraction: '0.1'
2023-10-23 14:59:24,308 ----------------------------------------------------------------------------------------------------
2023-10-23 14:59:24,308 Final evaluation on model from best epoch (best-model.pt)
2023-10-23 14:59:24,308 - metric: "('micro avg', 'f1-score')"
2023-10-23 14:59:24,308 ----------------------------------------------------------------------------------------------------
2023-10-23 14:59:24,308 Computation:
2023-10-23 14:59:24,308 - compute on device: cuda:0
2023-10-23 14:59:24,308 - embedding storage: none
2023-10-23 14:59:24,308 ----------------------------------------------------------------------------------------------------
2023-10-23 14:59:24,308 Model training base path: "hmbench-ajmc/de-dbmdz/bert-base-historic-multilingual-64k-td-cased-bs8-wsFalse-e10-lr5e-05-poolingfirst-layers-1-crfFalse-1"
2023-10-23 14:59:24,308 ----------------------------------------------------------------------------------------------------
2023-10-23 14:59:24,308 ----------------------------------------------------------------------------------------------------
2023-10-23 14:59:24,308 Logging anything other than scalars to TensorBoard is currently not supported.
2023-10-23 14:59:25,027 epoch 1 - iter 13/138 - loss 3.66543780 - time (sec): 0.72 - samples/sec: 3026.20 - lr: 0.000004 - momentum: 0.000000
2023-10-23 14:59:25,745 epoch 1 - iter 26/138 - loss 3.00984760 - time (sec): 1.44 - samples/sec: 2969.12 - lr: 0.000009 - momentum: 0.000000
2023-10-23 14:59:26,450 epoch 1 - iter 39/138 - loss 2.39180029 - time (sec): 2.14 - samples/sec: 2875.04 - lr: 0.000014 - momentum: 0.000000
2023-10-23 14:59:27,243 epoch 1 - iter 52/138 - loss 1.93634022 - time (sec): 2.93 - samples/sec: 2845.79 - lr: 0.000018 - momentum: 0.000000
2023-10-23 14:59:27,962 epoch 1 - iter 65/138 - loss 1.68757988 - time (sec): 3.65 - samples/sec: 2879.97 - lr: 0.000023 - momentum: 0.000000
2023-10-23 14:59:28,690 epoch 1 - iter 78/138 - loss 1.48023065 - time (sec): 4.38 - samples/sec: 2886.24 - lr: 0.000028 - momentum: 0.000000
2023-10-23 14:59:29,408 epoch 1 - iter 91/138 - loss 1.32188926 - time (sec): 5.10 - samples/sec: 2900.66 - lr: 0.000033 - momentum: 0.000000
2023-10-23 14:59:30,130 epoch 1 - iter 104/138 - loss 1.18393768 - time (sec): 5.82 - samples/sec: 2915.39 - lr: 0.000037 - momentum: 0.000000
2023-10-23 14:59:30,841 epoch 1 - iter 117/138 - loss 1.07131939 - time (sec): 6.53 - samples/sec: 2945.55 - lr: 0.000042 - momentum: 0.000000
2023-10-23 14:59:31,603 epoch 1 - iter 130/138 - loss 0.98230483 - time (sec): 7.29 - samples/sec: 2950.92 - lr: 0.000047 - momentum: 0.000000
2023-10-23 14:59:32,041 ----------------------------------------------------------------------------------------------------
2023-10-23 14:59:32,041 EPOCH 1 done: loss 0.9408 - lr: 0.000047
2023-10-23 14:59:32,466 DEV : loss 0.17025014758110046 - f1-score (micro avg) 0.7773
2023-10-23 14:59:32,472 saving best model
2023-10-23 14:59:32,868 ----------------------------------------------------------------------------------------------------
2023-10-23 14:59:33,590 epoch 2 - iter 13/138 - loss 0.14784199 - time (sec): 0.72 - samples/sec: 2802.71 - lr: 0.000050 - momentum: 0.000000
2023-10-23 14:59:34,321 epoch 2 - iter 26/138 - loss 0.11871101 - time (sec): 1.45 - samples/sec: 2935.30 - lr: 0.000049 - momentum: 0.000000
2023-10-23 14:59:35,042 epoch 2 - iter 39/138 - loss 0.14691239 - time (sec): 2.17 - samples/sec: 2897.04 - lr: 0.000048 - momentum: 0.000000
2023-10-23 14:59:35,847 epoch 2 - iter 52/138 - loss 0.15325754 - time (sec): 2.98 - samples/sec: 2875.85 - lr: 0.000048 - momentum: 0.000000
2023-10-23 14:59:36,572 epoch 2 - iter 65/138 - loss 0.15032886 - time (sec): 3.70 - samples/sec: 2879.84 - lr: 0.000047 - momentum: 0.000000
2023-10-23 14:59:37,291 epoch 2 - iter 78/138 - loss 0.15293320 - time (sec): 4.42 - samples/sec: 2905.21 - lr: 0.000047 - momentum: 0.000000
2023-10-23 14:59:38,012 epoch 2 - iter 91/138 - loss 0.14945325 - time (sec): 5.14 - samples/sec: 2934.83 - lr: 0.000046 - momentum: 0.000000
2023-10-23 14:59:38,796 epoch 2 - iter 104/138 - loss 0.15342696 - time (sec): 5.93 - samples/sec: 2916.57 - lr: 0.000046 - momentum: 0.000000
2023-10-23 14:59:39,528 epoch 2 - iter 117/138 - loss 0.14987829 - time (sec): 6.66 - samples/sec: 2915.77 - lr: 0.000045 - momentum: 0.000000
2023-10-23 14:59:40,253 epoch 2 - iter 130/138 - loss 0.15091087 - time (sec): 7.38 - samples/sec: 2922.83 - lr: 0.000045 - momentum: 0.000000
2023-10-23 14:59:40,709 ----------------------------------------------------------------------------------------------------
2023-10-23 14:59:40,709 EPOCH 2 done: loss 0.1508 - lr: 0.000045
2023-10-23 14:59:41,247 DEV : loss 0.12781710922718048 - f1-score (micro avg) 0.8111
2023-10-23 14:59:41,252 saving best model
2023-10-23 14:59:41,793 ----------------------------------------------------------------------------------------------------
2023-10-23 14:59:42,587 epoch 3 - iter 13/138 - loss 0.08645294 - time (sec): 0.79 - samples/sec: 2604.66 - lr: 0.000044 - momentum: 0.000000
2023-10-23 14:59:43,329 epoch 3 - iter 26/138 - loss 0.08482620 - time (sec): 1.53 - samples/sec: 2803.15 - lr: 0.000043 - momentum: 0.000000
2023-10-23 14:59:44,087 epoch 3 - iter 39/138 - loss 0.07421142 - time (sec): 2.29 - samples/sec: 2951.16 - lr: 0.000043 - momentum: 0.000000
2023-10-23 14:59:44,821 epoch 3 - iter 52/138 - loss 0.07684756 - time (sec): 3.02 - samples/sec: 2890.92 - lr: 0.000042 - momentum: 0.000000
2023-10-23 14:59:45,575 epoch 3 - iter 65/138 - loss 0.07988944 - time (sec): 3.78 - samples/sec: 2892.17 - lr: 0.000042 - momentum: 0.000000
2023-10-23 14:59:46,359 epoch 3 - iter 78/138 - loss 0.08173290 - time (sec): 4.56 - samples/sec: 2809.17 - lr: 0.000041 - momentum: 0.000000
2023-10-23 14:59:47,106 epoch 3 - iter 91/138 - loss 0.08449295 - time (sec): 5.31 - samples/sec: 2813.00 - lr: 0.000041 - momentum: 0.000000
2023-10-23 14:59:47,859 epoch 3 - iter 104/138 - loss 0.08365887 - time (sec): 6.06 - samples/sec: 2815.61 - lr: 0.000040 - momentum: 0.000000
2023-10-23 14:59:48,597 epoch 3 - iter 117/138 - loss 0.08635961 - time (sec): 6.80 - samples/sec: 2814.13 - lr: 0.000040 - momentum: 0.000000
2023-10-23 14:59:49,362 epoch 3 - iter 130/138 - loss 0.08894879 - time (sec): 7.57 - samples/sec: 2840.59 - lr: 0.000039 - momentum: 0.000000
2023-10-23 14:59:49,833 ----------------------------------------------------------------------------------------------------
2023-10-23 14:59:49,833 EPOCH 3 done: loss 0.0867 - lr: 0.000039
2023-10-23 14:59:50,542 DEV : loss 0.12280590832233429 - f1-score (micro avg) 0.8447
2023-10-23 14:59:50,547 saving best model
2023-10-23 14:59:51,053 ----------------------------------------------------------------------------------------------------
2023-10-23 14:59:51,872 epoch 4 - iter 13/138 - loss 0.05732552 - time (sec): 0.82 - samples/sec: 2575.68 - lr: 0.000038 - momentum: 0.000000
2023-10-23 14:59:52,594 epoch 4 - iter 26/138 - loss 0.05523775 - time (sec): 1.54 - samples/sec: 2899.82 - lr: 0.000038 - momentum: 0.000000
2023-10-23 14:59:53,306 epoch 4 - iter 39/138 - loss 0.05726177 - time (sec): 2.25 - samples/sec: 2827.15 - lr: 0.000037 - momentum: 0.000000
2023-10-23 14:59:54,023 epoch 4 - iter 52/138 - loss 0.05790800 - time (sec): 2.97 - samples/sec: 2893.97 - lr: 0.000037 - momentum: 0.000000
2023-10-23 14:59:54,753 epoch 4 - iter 65/138 - loss 0.05696300 - time (sec): 3.70 - samples/sec: 2898.53 - lr: 0.000036 - momentum: 0.000000
2023-10-23 14:59:55,463 epoch 4 - iter 78/138 - loss 0.05800066 - time (sec): 4.41 - samples/sec: 2861.98 - lr: 0.000036 - momentum: 0.000000
2023-10-23 14:59:56,181 epoch 4 - iter 91/138 - loss 0.05510131 - time (sec): 5.12 - samples/sec: 2878.87 - lr: 0.000035 - momentum: 0.000000
2023-10-23 14:59:56,894 epoch 4 - iter 104/138 - loss 0.05698036 - time (sec): 5.84 - samples/sec: 2905.22 - lr: 0.000035 - momentum: 0.000000
2023-10-23 14:59:57,614 epoch 4 - iter 117/138 - loss 0.05516748 - time (sec): 6.56 - samples/sec: 2934.96 - lr: 0.000034 - momentum: 0.000000
2023-10-23 14:59:58,324 epoch 4 - iter 130/138 - loss 0.05554848 - time (sec): 7.27 - samples/sec: 2943.58 - lr: 0.000034 - momentum: 0.000000
2023-10-23 14:59:58,765 ----------------------------------------------------------------------------------------------------
2023-10-23 14:59:58,765 EPOCH 4 done: loss 0.0548 - lr: 0.000034
2023-10-23 14:59:59,315 DEV : loss 0.13340476155281067 - f1-score (micro avg) 0.8653
2023-10-23 14:59:59,320 saving best model
2023-10-23 14:59:59,850 ----------------------------------------------------------------------------------------------------
2023-10-23 15:00:00,561 epoch 5 - iter 13/138 - loss 0.02190968 - time (sec): 0.71 - samples/sec: 2839.29 - lr: 0.000033 - momentum: 0.000000
2023-10-23 15:00:01,278 epoch 5 - iter 26/138 - loss 0.03255026 - time (sec): 1.42 - samples/sec: 2948.57 - lr: 0.000032 - momentum: 0.000000
2023-10-23 15:00:01,994 epoch 5 - iter 39/138 - loss 0.04017228 - time (sec): 2.14 - samples/sec: 2944.89 - lr: 0.000032 - momentum: 0.000000
2023-10-23 15:00:02,723 epoch 5 - iter 52/138 - loss 0.05156922 - time (sec): 2.87 - samples/sec: 2939.16 - lr: 0.000031 - momentum: 0.000000
2023-10-23 15:00:03,439 epoch 5 - iter 65/138 - loss 0.05092015 - time (sec): 3.58 - samples/sec: 2910.39 - lr: 0.000031 - momentum: 0.000000
2023-10-23 15:00:04,172 epoch 5 - iter 78/138 - loss 0.04808824 - time (sec): 4.32 - samples/sec: 2931.55 - lr: 0.000030 - momentum: 0.000000
2023-10-23 15:00:04,888 epoch 5 - iter 91/138 - loss 0.04539131 - time (sec): 5.03 - samples/sec: 2964.95 - lr: 0.000030 - momentum: 0.000000
2023-10-23 15:00:05,600 epoch 5 - iter 104/138 - loss 0.04367704 - time (sec): 5.75 - samples/sec: 2983.60 - lr: 0.000029 - momentum: 0.000000
2023-10-23 15:00:06,328 epoch 5 - iter 117/138 - loss 0.04343764 - time (sec): 6.47 - samples/sec: 2981.65 - lr: 0.000029 - momentum: 0.000000
2023-10-23 15:00:07,061 epoch 5 - iter 130/138 - loss 0.04444769 - time (sec): 7.21 - samples/sec: 2988.59 - lr: 0.000028 - momentum: 0.000000
2023-10-23 15:00:07,496 ----------------------------------------------------------------------------------------------------
2023-10-23 15:00:07,496 EPOCH 5 done: loss 0.0447 - lr: 0.000028
2023-10-23 15:00:08,032 DEV : loss 0.13237358629703522 - f1-score (micro avg) 0.8854
2023-10-23 15:00:08,038 saving best model
2023-10-23 15:00:08,590 ----------------------------------------------------------------------------------------------------
2023-10-23 15:00:09,309 epoch 6 - iter 13/138 - loss 0.03222312 - time (sec): 0.72 - samples/sec: 3156.71 - lr: 0.000027 - momentum: 0.000000
2023-10-23 15:00:10,022 epoch 6 - iter 26/138 - loss 0.02407024 - time (sec): 1.43 - samples/sec: 2975.50 - lr: 0.000027 - momentum: 0.000000
2023-10-23 15:00:10,738 epoch 6 - iter 39/138 - loss 0.02068903 - time (sec): 2.15 - samples/sec: 3069.60 - lr: 0.000026 - momentum: 0.000000
2023-10-23 15:00:11,452 epoch 6 - iter 52/138 - loss 0.03086218 - time (sec): 2.86 - samples/sec: 3087.93 - lr: 0.000026 - momentum: 0.000000
2023-10-23 15:00:12,173 epoch 6 - iter 65/138 - loss 0.03106766 - time (sec): 3.58 - samples/sec: 3080.79 - lr: 0.000025 - momentum: 0.000000
2023-10-23 15:00:12,884 epoch 6 - iter 78/138 - loss 0.03356739 - time (sec): 4.29 - samples/sec: 3029.68 - lr: 0.000025 - momentum: 0.000000
2023-10-23 15:00:13,599 epoch 6 - iter 91/138 - loss 0.03089337 - time (sec): 5.01 - samples/sec: 3050.62 - lr: 0.000024 - momentum: 0.000000
2023-10-23 15:00:14,300 epoch 6 - iter 104/138 - loss 0.03442090 - time (sec): 5.71 - samples/sec: 3025.17 - lr: 0.000024 - momentum: 0.000000
2023-10-23 15:00:15,035 epoch 6 - iter 117/138 - loss 0.03202038 - time (sec): 6.44 - samples/sec: 3021.98 - lr: 0.000023 - momentum: 0.000000
2023-10-23 15:00:15,771 epoch 6 - iter 130/138 - loss 0.03533535 - time (sec): 7.18 - samples/sec: 3006.32 - lr: 0.000023 - momentum: 0.000000
2023-10-23 15:00:16,197 ----------------------------------------------------------------------------------------------------
2023-10-23 15:00:16,197 EPOCH 6 done: loss 0.0364 - lr: 0.000023
2023-10-23 15:00:16,729 DEV : loss 0.15275059640407562 - f1-score (micro avg) 0.8938
2023-10-23 15:00:16,735 saving best model
2023-10-23 15:00:17,279 ----------------------------------------------------------------------------------------------------
2023-10-23 15:00:18,001 epoch 7 - iter 13/138 - loss 0.00527848 - time (sec): 0.72 - samples/sec: 2744.44 - lr: 0.000022 - momentum: 0.000000
2023-10-23 15:00:18,717 epoch 7 - iter 26/138 - loss 0.01820801 - time (sec): 1.44 - samples/sec: 2926.12 - lr: 0.000021 - momentum: 0.000000
2023-10-23 15:00:19,427 epoch 7 - iter 39/138 - loss 0.01831966 - time (sec): 2.15 - samples/sec: 2962.30 - lr: 0.000021 - momentum: 0.000000
2023-10-23 15:00:20,142 epoch 7 - iter 52/138 - loss 0.02742555 - time (sec): 2.86 - samples/sec: 3044.56 - lr: 0.000020 - momentum: 0.000000
2023-10-23 15:00:20,875 epoch 7 - iter 65/138 - loss 0.02380596 - time (sec): 3.59 - samples/sec: 3024.39 - lr: 0.000020 - momentum: 0.000000
2023-10-23 15:00:21,601 epoch 7 - iter 78/138 - loss 0.02668274 - time (sec): 4.32 - samples/sec: 3007.31 - lr: 0.000019 - momentum: 0.000000
2023-10-23 15:00:22,323 epoch 7 - iter 91/138 - loss 0.02478362 - time (sec): 5.04 - samples/sec: 3037.01 - lr: 0.000019 - momentum: 0.000000
2023-10-23 15:00:23,037 epoch 7 - iter 104/138 - loss 0.02409141 - time (sec): 5.76 - samples/sec: 3014.13 - lr: 0.000018 - momentum: 0.000000
2023-10-23 15:00:23,746 epoch 7 - iter 117/138 - loss 0.02390261 - time (sec): 6.46 - samples/sec: 3022.82 - lr: 0.000018 - momentum: 0.000000
2023-10-23 15:00:24,451 epoch 7 - iter 130/138 - loss 0.02278232 - time (sec): 7.17 - samples/sec: 3011.96 - lr: 0.000017 - momentum: 0.000000
2023-10-23 15:00:24,880 ----------------------------------------------------------------------------------------------------
2023-10-23 15:00:24,881 EPOCH 7 done: loss 0.0232 - lr: 0.000017
2023-10-23 15:00:25,413 DEV : loss 0.1645553708076477 - f1-score (micro avg) 0.883
2023-10-23 15:00:25,418 ----------------------------------------------------------------------------------------------------
2023-10-23 15:00:26,141 epoch 8 - iter 13/138 - loss 0.03132831 - time (sec): 0.72 - samples/sec: 3036.44 - lr: 0.000016 - momentum: 0.000000
2023-10-23 15:00:26,851 epoch 8 - iter 26/138 - loss 0.03228913 - time (sec): 1.43 - samples/sec: 3076.10 - lr: 0.000016 - momentum: 0.000000
2023-10-23 15:00:27,565 epoch 8 - iter 39/138 - loss 0.02750764 - time (sec): 2.15 - samples/sec: 2978.19 - lr: 0.000015 - momentum: 0.000000
2023-10-23 15:00:28,292 epoch 8 - iter 52/138 - loss 0.02386624 - time (sec): 2.87 - samples/sec: 3066.68 - lr: 0.000015 - momentum: 0.000000
2023-10-23 15:00:29,005 epoch 8 - iter 65/138 - loss 0.02192605 - time (sec): 3.59 - samples/sec: 3071.37 - lr: 0.000014 - momentum: 0.000000
2023-10-23 15:00:29,709 epoch 8 - iter 78/138 - loss 0.01935845 - time (sec): 4.29 - samples/sec: 3028.11 - lr: 0.000014 - momentum: 0.000000
2023-10-23 15:00:30,419 epoch 8 - iter 91/138 - loss 0.01854524 - time (sec): 5.00 - samples/sec: 3005.90 - lr: 0.000013 - momentum: 0.000000
2023-10-23 15:00:31,127 epoch 8 - iter 104/138 - loss 0.01754654 - time (sec): 5.71 - samples/sec: 3037.43 - lr: 0.000013 - momentum: 0.000000
2023-10-23 15:00:31,844 epoch 8 - iter 117/138 - loss 0.01644821 - time (sec): 6.42 - samples/sec: 3044.82 - lr: 0.000012 - momentum: 0.000000
2023-10-23 15:00:32,546 epoch 8 - iter 130/138 - loss 0.01595208 - time (sec): 7.13 - samples/sec: 3027.96 - lr: 0.000012 - momentum: 0.000000
2023-10-23 15:00:32,977 ----------------------------------------------------------------------------------------------------
2023-10-23 15:00:32,977 EPOCH 8 done: loss 0.0162 - lr: 0.000012
2023-10-23 15:00:33,512 DEV : loss 0.16538017988204956 - f1-score (micro avg) 0.8924
2023-10-23 15:00:33,517 ----------------------------------------------------------------------------------------------------
2023-10-23 15:00:34,230 epoch 9 - iter 13/138 - loss 0.02788626 - time (sec): 0.71 - samples/sec: 2745.49 - lr: 0.000011 - momentum: 0.000000
2023-10-23 15:00:34,951 epoch 9 - iter 26/138 - loss 0.01954824 - time (sec): 1.43 - samples/sec: 3002.40 - lr: 0.000010 - momentum: 0.000000
2023-10-23 15:00:35,659 epoch 9 - iter 39/138 - loss 0.01360116 - time (sec): 2.14 - samples/sec: 2984.85 - lr: 0.000010 - momentum: 0.000000
2023-10-23 15:00:36,369 epoch 9 - iter 52/138 - loss 0.01102750 - time (sec): 2.85 - samples/sec: 2994.57 - lr: 0.000009 - momentum: 0.000000
2023-10-23 15:00:37,087 epoch 9 - iter 65/138 - loss 0.01089929 - time (sec): 3.57 - samples/sec: 3021.95 - lr: 0.000009 - momentum: 0.000000
2023-10-23 15:00:37,808 epoch 9 - iter 78/138 - loss 0.01437039 - time (sec): 4.29 - samples/sec: 3072.07 - lr: 0.000008 - momentum: 0.000000
2023-10-23 15:00:38,526 epoch 9 - iter 91/138 - loss 0.01440842 - time (sec): 5.01 - samples/sec: 3068.49 - lr: 0.000008 - momentum: 0.000000
2023-10-23 15:00:39,241 epoch 9 - iter 104/138 - loss 0.01322011 - time (sec): 5.72 - samples/sec: 3013.70 - lr: 0.000007 - momentum: 0.000000
2023-10-23 15:00:39,951 epoch 9 - iter 117/138 - loss 0.01204757 - time (sec): 6.43 - samples/sec: 2979.71 - lr: 0.000007 - momentum: 0.000000
2023-10-23 15:00:40,696 epoch 9 - iter 130/138 - loss 0.01145643 - time (sec): 7.18 - samples/sec: 3030.74 - lr: 0.000006 - momentum: 0.000000
2023-10-23 15:00:41,121 ----------------------------------------------------------------------------------------------------
2023-10-23 15:00:41,121 EPOCH 9 done: loss 0.0113 - lr: 0.000006
2023-10-23 15:00:41,656 DEV : loss 0.17684748768806458 - f1-score (micro avg) 0.8894
2023-10-23 15:00:41,661 ----------------------------------------------------------------------------------------------------
2023-10-23 15:00:42,380 epoch 10 - iter 13/138 - loss 0.00794737 - time (sec): 0.72 - samples/sec: 2884.96 - lr: 0.000005 - momentum: 0.000000
2023-10-23 15:00:43,098 epoch 10 - iter 26/138 - loss 0.00943552 - time (sec): 1.44 - samples/sec: 2951.74 - lr: 0.000005 - momentum: 0.000000
2023-10-23 15:00:43,822 epoch 10 - iter 39/138 - loss 0.00724063 - time (sec): 2.16 - samples/sec: 2948.75 - lr: 0.000004 - momentum: 0.000000
2023-10-23 15:00:44,535 epoch 10 - iter 52/138 - loss 0.00610859 - time (sec): 2.87 - samples/sec: 2948.72 - lr: 0.000004 - momentum: 0.000000
2023-10-23 15:00:45,237 epoch 10 - iter 65/138 - loss 0.00718335 - time (sec): 3.58 - samples/sec: 2912.38 - lr: 0.000003 - momentum: 0.000000
2023-10-23 15:00:45,949 epoch 10 - iter 78/138 - loss 0.00609465 - time (sec): 4.29 - samples/sec: 2974.64 - lr: 0.000003 - momentum: 0.000000
2023-10-23 15:00:46,661 epoch 10 - iter 91/138 - loss 0.00724821 - time (sec): 5.00 - samples/sec: 2992.73 - lr: 0.000002 - momentum: 0.000000
2023-10-23 15:00:47,375 epoch 10 - iter 104/138 - loss 0.00669768 - time (sec): 5.71 - samples/sec: 3002.23 - lr: 0.000002 - momentum: 0.000000
2023-10-23 15:00:48,096 epoch 10 - iter 117/138 - loss 0.01007155 - time (sec): 6.43 - samples/sec: 3012.47 - lr: 0.000001 - momentum: 0.000000
2023-10-23 15:00:48,811 epoch 10 - iter 130/138 - loss 0.00905841 - time (sec): 7.15 - samples/sec: 3019.54 - lr: 0.000000 - momentum: 0.000000
2023-10-23 15:00:49,255 ----------------------------------------------------------------------------------------------------
2023-10-23 15:00:49,256 EPOCH 10 done: loss 0.0087 - lr: 0.000000
2023-10-23 15:00:49,787 DEV : loss 0.17404906451702118 - f1-score (micro avg) 0.8951
2023-10-23 15:00:49,792 saving best model
2023-10-23 15:00:50,737 ----------------------------------------------------------------------------------------------------
2023-10-23 15:00:50,738 Loading model from best epoch ...
2023-10-23 15:00:52,533 SequenceTagger predicts: Dictionary with 25 tags: O, S-scope, B-scope, E-scope, I-scope, S-pers, B-pers, E-pers, I-pers, S-work, B-work, E-work, I-work, S-loc, B-loc, E-loc, I-loc, S-object, B-object, E-object, I-object, S-date, B-date, E-date, I-date
2023-10-23 15:00:53,190
Results:
- F-score (micro) 0.9026
- F-score (macro) 0.8381
- Accuracy 0.8407
By class:
precision recall f1-score support
scope 0.8876 0.8977 0.8927 176
pers 1.0000 0.9297 0.9636 128
work 0.8182 0.8514 0.8344 74
object 1.0000 1.0000 1.0000 2
loc 0.5000 0.5000 0.5000 2
micro avg 0.9074 0.8979 0.9026 382
macro avg 0.8412 0.8358 0.8381 382
weighted avg 0.9104 0.8979 0.9036 382
2023-10-23 15:00:53,190 ----------------------------------------------------------------------------------------------------
|