File size: 24,017 Bytes
8a2b9e0 |
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 |
2023-10-23 15:07:12,704 ----------------------------------------------------------------------------------------------------
2023-10-23 15:07:12,705 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 15:07:12,705 ----------------------------------------------------------------------------------------------------
2023-10-23 15:07:12,705 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 15:07:12,705 ----------------------------------------------------------------------------------------------------
2023-10-23 15:07:12,705 Train: 1100 sentences
2023-10-23 15:07:12,705 (train_with_dev=False, train_with_test=False)
2023-10-23 15:07:12,705 ----------------------------------------------------------------------------------------------------
2023-10-23 15:07:12,705 Training Params:
2023-10-23 15:07:12,705 - learning_rate: "3e-05"
2023-10-23 15:07:12,705 - mini_batch_size: "8"
2023-10-23 15:07:12,705 - max_epochs: "10"
2023-10-23 15:07:12,705 - shuffle: "True"
2023-10-23 15:07:12,705 ----------------------------------------------------------------------------------------------------
2023-10-23 15:07:12,705 Plugins:
2023-10-23 15:07:12,705 - TensorboardLogger
2023-10-23 15:07:12,706 - LinearScheduler | warmup_fraction: '0.1'
2023-10-23 15:07:12,706 ----------------------------------------------------------------------------------------------------
2023-10-23 15:07:12,706 Final evaluation on model from best epoch (best-model.pt)
2023-10-23 15:07:12,706 - metric: "('micro avg', 'f1-score')"
2023-10-23 15:07:12,706 ----------------------------------------------------------------------------------------------------
2023-10-23 15:07:12,706 Computation:
2023-10-23 15:07:12,706 - compute on device: cuda:0
2023-10-23 15:07:12,706 - embedding storage: none
2023-10-23 15:07:12,706 ----------------------------------------------------------------------------------------------------
2023-10-23 15:07:12,706 Model training base path: "hmbench-ajmc/de-dbmdz/bert-base-historic-multilingual-64k-td-cased-bs8-wsFalse-e10-lr3e-05-poolingfirst-layers-1-crfFalse-2"
2023-10-23 15:07:12,706 ----------------------------------------------------------------------------------------------------
2023-10-23 15:07:12,706 ----------------------------------------------------------------------------------------------------
2023-10-23 15:07:12,706 Logging anything other than scalars to TensorBoard is currently not supported.
2023-10-23 15:07:13,429 epoch 1 - iter 13/138 - loss 3.37467106 - time (sec): 0.72 - samples/sec: 2971.73 - lr: 0.000003 - momentum: 0.000000
2023-10-23 15:07:14,143 epoch 1 - iter 26/138 - loss 2.96938082 - time (sec): 1.44 - samples/sec: 2907.62 - lr: 0.000005 - momentum: 0.000000
2023-10-23 15:07:14,848 epoch 1 - iter 39/138 - loss 2.44657037 - time (sec): 2.14 - samples/sec: 2846.49 - lr: 0.000008 - momentum: 0.000000
2023-10-23 15:07:15,559 epoch 1 - iter 52/138 - loss 2.06681506 - time (sec): 2.85 - samples/sec: 2825.81 - lr: 0.000011 - momentum: 0.000000
2023-10-23 15:07:16,303 epoch 1 - iter 65/138 - loss 1.78927097 - time (sec): 3.60 - samples/sec: 2878.28 - lr: 0.000014 - momentum: 0.000000
2023-10-23 15:07:17,052 epoch 1 - iter 78/138 - loss 1.56584748 - time (sec): 4.34 - samples/sec: 2921.38 - lr: 0.000017 - momentum: 0.000000
2023-10-23 15:07:17,771 epoch 1 - iter 91/138 - loss 1.40260405 - time (sec): 5.06 - samples/sec: 2943.91 - lr: 0.000020 - momentum: 0.000000
2023-10-23 15:07:18,483 epoch 1 - iter 104/138 - loss 1.27483414 - time (sec): 5.78 - samples/sec: 2966.95 - lr: 0.000022 - momentum: 0.000000
2023-10-23 15:07:19,221 epoch 1 - iter 117/138 - loss 1.16935322 - time (sec): 6.51 - samples/sec: 2958.64 - lr: 0.000025 - momentum: 0.000000
2023-10-23 15:07:19,945 epoch 1 - iter 130/138 - loss 1.07390114 - time (sec): 7.24 - samples/sec: 2982.84 - lr: 0.000028 - momentum: 0.000000
2023-10-23 15:07:20,374 ----------------------------------------------------------------------------------------------------
2023-10-23 15:07:20,374 EPOCH 1 done: loss 1.0347 - lr: 0.000028
2023-10-23 15:07:20,968 DEV : loss 0.19735337793827057 - f1-score (micro avg) 0.6963
2023-10-23 15:07:20,974 saving best model
2023-10-23 15:07:21,373 ----------------------------------------------------------------------------------------------------
2023-10-23 15:07:22,079 epoch 2 - iter 13/138 - loss 0.15888997 - time (sec): 0.70 - samples/sec: 2975.19 - lr: 0.000030 - momentum: 0.000000
2023-10-23 15:07:22,792 epoch 2 - iter 26/138 - loss 0.17882334 - time (sec): 1.42 - samples/sec: 2950.17 - lr: 0.000029 - momentum: 0.000000
2023-10-23 15:07:23,500 epoch 2 - iter 39/138 - loss 0.18258217 - time (sec): 2.13 - samples/sec: 2905.83 - lr: 0.000029 - momentum: 0.000000
2023-10-23 15:07:24,220 epoch 2 - iter 52/138 - loss 0.19444341 - time (sec): 2.85 - samples/sec: 2957.20 - lr: 0.000029 - momentum: 0.000000
2023-10-23 15:07:24,929 epoch 2 - iter 65/138 - loss 0.19897156 - time (sec): 3.55 - samples/sec: 2954.44 - lr: 0.000028 - momentum: 0.000000
2023-10-23 15:07:25,646 epoch 2 - iter 78/138 - loss 0.19171456 - time (sec): 4.27 - samples/sec: 2927.05 - lr: 0.000028 - momentum: 0.000000
2023-10-23 15:07:26,357 epoch 2 - iter 91/138 - loss 0.17941956 - time (sec): 4.98 - samples/sec: 2916.55 - lr: 0.000028 - momentum: 0.000000
2023-10-23 15:07:27,083 epoch 2 - iter 104/138 - loss 0.17639339 - time (sec): 5.71 - samples/sec: 2970.36 - lr: 0.000028 - momentum: 0.000000
2023-10-23 15:07:27,812 epoch 2 - iter 117/138 - loss 0.16912778 - time (sec): 6.44 - samples/sec: 2974.91 - lr: 0.000027 - momentum: 0.000000
2023-10-23 15:07:28,534 epoch 2 - iter 130/138 - loss 0.17110248 - time (sec): 7.16 - samples/sec: 2990.69 - lr: 0.000027 - momentum: 0.000000
2023-10-23 15:07:28,976 ----------------------------------------------------------------------------------------------------
2023-10-23 15:07:28,976 EPOCH 2 done: loss 0.1705 - lr: 0.000027
2023-10-23 15:07:29,512 DEV : loss 0.122329480946064 - f1-score (micro avg) 0.8271
2023-10-23 15:07:29,518 saving best model
2023-10-23 15:07:30,073 ----------------------------------------------------------------------------------------------------
2023-10-23 15:07:30,808 epoch 3 - iter 13/138 - loss 0.08335533 - time (sec): 0.73 - samples/sec: 2991.57 - lr: 0.000026 - momentum: 0.000000
2023-10-23 15:07:31,532 epoch 3 - iter 26/138 - loss 0.08264747 - time (sec): 1.45 - samples/sec: 2908.96 - lr: 0.000026 - momentum: 0.000000
2023-10-23 15:07:32,255 epoch 3 - iter 39/138 - loss 0.07886078 - time (sec): 2.18 - samples/sec: 2788.01 - lr: 0.000026 - momentum: 0.000000
2023-10-23 15:07:32,965 epoch 3 - iter 52/138 - loss 0.08829144 - time (sec): 2.89 - samples/sec: 2915.05 - lr: 0.000025 - momentum: 0.000000
2023-10-23 15:07:33,706 epoch 3 - iter 65/138 - loss 0.08097315 - time (sec): 3.63 - samples/sec: 2919.45 - lr: 0.000025 - momentum: 0.000000
2023-10-23 15:07:34,428 epoch 3 - iter 78/138 - loss 0.09177425 - time (sec): 4.35 - samples/sec: 2985.50 - lr: 0.000025 - momentum: 0.000000
2023-10-23 15:07:35,146 epoch 3 - iter 91/138 - loss 0.09142755 - time (sec): 5.07 - samples/sec: 2969.81 - lr: 0.000025 - momentum: 0.000000
2023-10-23 15:07:35,861 epoch 3 - iter 104/138 - loss 0.09832228 - time (sec): 5.78 - samples/sec: 2959.21 - lr: 0.000024 - momentum: 0.000000
2023-10-23 15:07:36,586 epoch 3 - iter 117/138 - loss 0.09428056 - time (sec): 6.51 - samples/sec: 2934.42 - lr: 0.000024 - momentum: 0.000000
2023-10-23 15:07:37,315 epoch 3 - iter 130/138 - loss 0.09309894 - time (sec): 7.24 - samples/sec: 2981.29 - lr: 0.000024 - momentum: 0.000000
2023-10-23 15:07:37,749 ----------------------------------------------------------------------------------------------------
2023-10-23 15:07:37,749 EPOCH 3 done: loss 0.0929 - lr: 0.000024
2023-10-23 15:07:38,296 DEV : loss 0.11146871745586395 - f1-score (micro avg) 0.844
2023-10-23 15:07:38,302 saving best model
2023-10-23 15:07:38,852 ----------------------------------------------------------------------------------------------------
2023-10-23 15:07:39,565 epoch 4 - iter 13/138 - loss 0.07417365 - time (sec): 0.71 - samples/sec: 2998.80 - lr: 0.000023 - momentum: 0.000000
2023-10-23 15:07:40,286 epoch 4 - iter 26/138 - loss 0.06986805 - time (sec): 1.43 - samples/sec: 2968.61 - lr: 0.000023 - momentum: 0.000000
2023-10-23 15:07:41,007 epoch 4 - iter 39/138 - loss 0.08114707 - time (sec): 2.15 - samples/sec: 3128.61 - lr: 0.000022 - momentum: 0.000000
2023-10-23 15:07:41,736 epoch 4 - iter 52/138 - loss 0.08438909 - time (sec): 2.88 - samples/sec: 3084.25 - lr: 0.000022 - momentum: 0.000000
2023-10-23 15:07:42,461 epoch 4 - iter 65/138 - loss 0.08351933 - time (sec): 3.61 - samples/sec: 3031.35 - lr: 0.000022 - momentum: 0.000000
2023-10-23 15:07:43,184 epoch 4 - iter 78/138 - loss 0.08021593 - time (sec): 4.33 - samples/sec: 3031.07 - lr: 0.000021 - momentum: 0.000000
2023-10-23 15:07:43,906 epoch 4 - iter 91/138 - loss 0.07395782 - time (sec): 5.05 - samples/sec: 2999.11 - lr: 0.000021 - momentum: 0.000000
2023-10-23 15:07:44,620 epoch 4 - iter 104/138 - loss 0.06842096 - time (sec): 5.76 - samples/sec: 3001.16 - lr: 0.000021 - momentum: 0.000000
2023-10-23 15:07:45,339 epoch 4 - iter 117/138 - loss 0.06532291 - time (sec): 6.48 - samples/sec: 3016.64 - lr: 0.000021 - momentum: 0.000000
2023-10-23 15:07:46,062 epoch 4 - iter 130/138 - loss 0.06170555 - time (sec): 7.21 - samples/sec: 3004.71 - lr: 0.000020 - momentum: 0.000000
2023-10-23 15:07:46,496 ----------------------------------------------------------------------------------------------------
2023-10-23 15:07:46,497 EPOCH 4 done: loss 0.0591 - lr: 0.000020
2023-10-23 15:07:47,039 DEV : loss 0.14198631048202515 - f1-score (micro avg) 0.8429
2023-10-23 15:07:47,045 ----------------------------------------------------------------------------------------------------
2023-10-23 15:07:47,788 epoch 5 - iter 13/138 - loss 0.06644428 - time (sec): 0.74 - samples/sec: 2745.04 - lr: 0.000020 - momentum: 0.000000
2023-10-23 15:07:48,512 epoch 5 - iter 26/138 - loss 0.04553320 - time (sec): 1.47 - samples/sec: 2952.14 - lr: 0.000019 - momentum: 0.000000
2023-10-23 15:07:49,230 epoch 5 - iter 39/138 - loss 0.04521235 - time (sec): 2.18 - samples/sec: 2950.05 - lr: 0.000019 - momentum: 0.000000
2023-10-23 15:07:49,949 epoch 5 - iter 52/138 - loss 0.04484997 - time (sec): 2.90 - samples/sec: 2990.83 - lr: 0.000019 - momentum: 0.000000
2023-10-23 15:07:50,661 epoch 5 - iter 65/138 - loss 0.04792661 - time (sec): 3.61 - samples/sec: 2919.57 - lr: 0.000018 - momentum: 0.000000
2023-10-23 15:07:51,387 epoch 5 - iter 78/138 - loss 0.04659725 - time (sec): 4.34 - samples/sec: 2989.06 - lr: 0.000018 - momentum: 0.000000
2023-10-23 15:07:52,103 epoch 5 - iter 91/138 - loss 0.04544001 - time (sec): 5.06 - samples/sec: 3014.62 - lr: 0.000018 - momentum: 0.000000
2023-10-23 15:07:52,819 epoch 5 - iter 104/138 - loss 0.04228585 - time (sec): 5.77 - samples/sec: 2995.04 - lr: 0.000018 - momentum: 0.000000
2023-10-23 15:07:53,539 epoch 5 - iter 117/138 - loss 0.04306651 - time (sec): 6.49 - samples/sec: 2988.26 - lr: 0.000017 - momentum: 0.000000
2023-10-23 15:07:54,265 epoch 5 - iter 130/138 - loss 0.04583492 - time (sec): 7.22 - samples/sec: 3009.23 - lr: 0.000017 - momentum: 0.000000
2023-10-23 15:07:54,703 ----------------------------------------------------------------------------------------------------
2023-10-23 15:07:54,704 EPOCH 5 done: loss 0.0454 - lr: 0.000017
2023-10-23 15:07:55,244 DEV : loss 0.1429574191570282 - f1-score (micro avg) 0.8729
2023-10-23 15:07:55,250 saving best model
2023-10-23 15:07:55,797 ----------------------------------------------------------------------------------------------------
2023-10-23 15:07:56,548 epoch 6 - iter 13/138 - loss 0.05313669 - time (sec): 0.74 - samples/sec: 3211.35 - lr: 0.000016 - momentum: 0.000000
2023-10-23 15:07:57,283 epoch 6 - iter 26/138 - loss 0.03327101 - time (sec): 1.48 - samples/sec: 3195.35 - lr: 0.000016 - momentum: 0.000000
2023-10-23 15:07:58,021 epoch 6 - iter 39/138 - loss 0.04254490 - time (sec): 2.22 - samples/sec: 3140.84 - lr: 0.000016 - momentum: 0.000000
2023-10-23 15:07:58,758 epoch 6 - iter 52/138 - loss 0.03923725 - time (sec): 2.95 - samples/sec: 2936.13 - lr: 0.000015 - momentum: 0.000000
2023-10-23 15:07:59,478 epoch 6 - iter 65/138 - loss 0.03746846 - time (sec): 3.67 - samples/sec: 2937.94 - lr: 0.000015 - momentum: 0.000000
2023-10-23 15:08:00,197 epoch 6 - iter 78/138 - loss 0.03435769 - time (sec): 4.39 - samples/sec: 2994.45 - lr: 0.000015 - momentum: 0.000000
2023-10-23 15:08:00,945 epoch 6 - iter 91/138 - loss 0.03640010 - time (sec): 5.14 - samples/sec: 2998.68 - lr: 0.000015 - momentum: 0.000000
2023-10-23 15:08:01,694 epoch 6 - iter 104/138 - loss 0.03377304 - time (sec): 5.89 - samples/sec: 2984.79 - lr: 0.000014 - momentum: 0.000000
2023-10-23 15:08:02,446 epoch 6 - iter 117/138 - loss 0.03257520 - time (sec): 6.64 - samples/sec: 2957.25 - lr: 0.000014 - momentum: 0.000000
2023-10-23 15:08:03,180 epoch 6 - iter 130/138 - loss 0.03234360 - time (sec): 7.38 - samples/sec: 2954.25 - lr: 0.000014 - momentum: 0.000000
2023-10-23 15:08:03,655 ----------------------------------------------------------------------------------------------------
2023-10-23 15:08:03,656 EPOCH 6 done: loss 0.0321 - lr: 0.000014
2023-10-23 15:08:04,195 DEV : loss 0.1479213535785675 - f1-score (micro avg) 0.8709
2023-10-23 15:08:04,200 ----------------------------------------------------------------------------------------------------
2023-10-23 15:08:04,966 epoch 7 - iter 13/138 - loss 0.02597140 - time (sec): 0.76 - samples/sec: 2514.83 - lr: 0.000013 - momentum: 0.000000
2023-10-23 15:08:05,726 epoch 7 - iter 26/138 - loss 0.01437665 - time (sec): 1.52 - samples/sec: 2761.37 - lr: 0.000013 - momentum: 0.000000
2023-10-23 15:08:06,471 epoch 7 - iter 39/138 - loss 0.01851670 - time (sec): 2.27 - samples/sec: 2850.88 - lr: 0.000012 - momentum: 0.000000
2023-10-23 15:08:07,196 epoch 7 - iter 52/138 - loss 0.02480619 - time (sec): 2.99 - samples/sec: 2820.13 - lr: 0.000012 - momentum: 0.000000
2023-10-23 15:08:07,948 epoch 7 - iter 65/138 - loss 0.02149033 - time (sec): 3.75 - samples/sec: 2910.10 - lr: 0.000012 - momentum: 0.000000
2023-10-23 15:08:08,703 epoch 7 - iter 78/138 - loss 0.02322504 - time (sec): 4.50 - samples/sec: 2906.22 - lr: 0.000012 - momentum: 0.000000
2023-10-23 15:08:09,465 epoch 7 - iter 91/138 - loss 0.02397668 - time (sec): 5.26 - samples/sec: 2870.09 - lr: 0.000011 - momentum: 0.000000
2023-10-23 15:08:10,222 epoch 7 - iter 104/138 - loss 0.02779713 - time (sec): 6.02 - samples/sec: 2892.16 - lr: 0.000011 - momentum: 0.000000
2023-10-23 15:08:10,971 epoch 7 - iter 117/138 - loss 0.02625708 - time (sec): 6.77 - samples/sec: 2907.51 - lr: 0.000011 - momentum: 0.000000
2023-10-23 15:08:11,719 epoch 7 - iter 130/138 - loss 0.02663558 - time (sec): 7.52 - samples/sec: 2877.48 - lr: 0.000010 - momentum: 0.000000
2023-10-23 15:08:12,167 ----------------------------------------------------------------------------------------------------
2023-10-23 15:08:12,168 EPOCH 7 done: loss 0.0262 - lr: 0.000010
2023-10-23 15:08:12,700 DEV : loss 0.1471543312072754 - f1-score (micro avg) 0.8889
2023-10-23 15:08:12,706 saving best model
2023-10-23 15:08:13,258 ----------------------------------------------------------------------------------------------------
2023-10-23 15:08:13,992 epoch 8 - iter 13/138 - loss 0.00685940 - time (sec): 0.73 - samples/sec: 3052.76 - lr: 0.000010 - momentum: 0.000000
2023-10-23 15:08:14,719 epoch 8 - iter 26/138 - loss 0.00837478 - time (sec): 1.46 - samples/sec: 3063.30 - lr: 0.000009 - momentum: 0.000000
2023-10-23 15:08:15,449 epoch 8 - iter 39/138 - loss 0.01552073 - time (sec): 2.19 - samples/sec: 2964.82 - lr: 0.000009 - momentum: 0.000000
2023-10-23 15:08:16,191 epoch 8 - iter 52/138 - loss 0.01548429 - time (sec): 2.93 - samples/sec: 2968.22 - lr: 0.000009 - momentum: 0.000000
2023-10-23 15:08:16,913 epoch 8 - iter 65/138 - loss 0.01707087 - time (sec): 3.65 - samples/sec: 2992.16 - lr: 0.000009 - momentum: 0.000000
2023-10-23 15:08:17,643 epoch 8 - iter 78/138 - loss 0.01863647 - time (sec): 4.38 - samples/sec: 3012.24 - lr: 0.000008 - momentum: 0.000000
2023-10-23 15:08:18,388 epoch 8 - iter 91/138 - loss 0.01654718 - time (sec): 5.13 - samples/sec: 3023.60 - lr: 0.000008 - momentum: 0.000000
2023-10-23 15:08:19,120 epoch 8 - iter 104/138 - loss 0.01502671 - time (sec): 5.86 - samples/sec: 2978.80 - lr: 0.000008 - momentum: 0.000000
2023-10-23 15:08:19,834 epoch 8 - iter 117/138 - loss 0.01598256 - time (sec): 6.57 - samples/sec: 2959.46 - lr: 0.000007 - momentum: 0.000000
2023-10-23 15:08:20,554 epoch 8 - iter 130/138 - loss 0.01714523 - time (sec): 7.29 - samples/sec: 2971.22 - lr: 0.000007 - momentum: 0.000000
2023-10-23 15:08:21,007 ----------------------------------------------------------------------------------------------------
2023-10-23 15:08:21,007 EPOCH 8 done: loss 0.0202 - lr: 0.000007
2023-10-23 15:08:21,548 DEV : loss 0.16060495376586914 - f1-score (micro avg) 0.8709
2023-10-23 15:08:21,554 ----------------------------------------------------------------------------------------------------
2023-10-23 15:08:22,282 epoch 9 - iter 13/138 - loss 0.01332285 - time (sec): 0.73 - samples/sec: 2626.51 - lr: 0.000006 - momentum: 0.000000
2023-10-23 15:08:23,018 epoch 9 - iter 26/138 - loss 0.01710958 - time (sec): 1.46 - samples/sec: 2850.06 - lr: 0.000006 - momentum: 0.000000
2023-10-23 15:08:23,760 epoch 9 - iter 39/138 - loss 0.01294442 - time (sec): 2.20 - samples/sec: 2937.72 - lr: 0.000006 - momentum: 0.000000
2023-10-23 15:08:24,471 epoch 9 - iter 52/138 - loss 0.01105454 - time (sec): 2.92 - samples/sec: 2875.90 - lr: 0.000005 - momentum: 0.000000
2023-10-23 15:08:25,187 epoch 9 - iter 65/138 - loss 0.01446311 - time (sec): 3.63 - samples/sec: 2900.29 - lr: 0.000005 - momentum: 0.000000
2023-10-23 15:08:25,912 epoch 9 - iter 78/138 - loss 0.01373708 - time (sec): 4.36 - samples/sec: 2918.40 - lr: 0.000005 - momentum: 0.000000
2023-10-23 15:08:26,634 epoch 9 - iter 91/138 - loss 0.01216404 - time (sec): 5.08 - samples/sec: 2946.47 - lr: 0.000005 - momentum: 0.000000
2023-10-23 15:08:27,354 epoch 9 - iter 104/138 - loss 0.01273020 - time (sec): 5.80 - samples/sec: 2957.94 - lr: 0.000004 - momentum: 0.000000
2023-10-23 15:08:28,075 epoch 9 - iter 117/138 - loss 0.01627844 - time (sec): 6.52 - samples/sec: 2952.13 - lr: 0.000004 - momentum: 0.000000
2023-10-23 15:08:28,814 epoch 9 - iter 130/138 - loss 0.01638689 - time (sec): 7.26 - samples/sec: 2966.35 - lr: 0.000004 - momentum: 0.000000
2023-10-23 15:08:29,252 ----------------------------------------------------------------------------------------------------
2023-10-23 15:08:29,252 EPOCH 9 done: loss 0.0177 - lr: 0.000004
2023-10-23 15:08:29,788 DEV : loss 0.16575126349925995 - f1-score (micro avg) 0.873
2023-10-23 15:08:29,794 ----------------------------------------------------------------------------------------------------
2023-10-23 15:08:30,511 epoch 10 - iter 13/138 - loss 0.02224445 - time (sec): 0.72 - samples/sec: 2949.26 - lr: 0.000003 - momentum: 0.000000
2023-10-23 15:08:31,229 epoch 10 - iter 26/138 - loss 0.01329094 - time (sec): 1.43 - samples/sec: 2876.82 - lr: 0.000003 - momentum: 0.000000
2023-10-23 15:08:31,953 epoch 10 - iter 39/138 - loss 0.01583951 - time (sec): 2.16 - samples/sec: 2954.08 - lr: 0.000002 - momentum: 0.000000
2023-10-23 15:08:32,674 epoch 10 - iter 52/138 - loss 0.01334604 - time (sec): 2.88 - samples/sec: 2962.65 - lr: 0.000002 - momentum: 0.000000
2023-10-23 15:08:33,393 epoch 10 - iter 65/138 - loss 0.01096514 - time (sec): 3.60 - samples/sec: 2943.65 - lr: 0.000002 - momentum: 0.000000
2023-10-23 15:08:34,127 epoch 10 - iter 78/138 - loss 0.01231639 - time (sec): 4.33 - samples/sec: 3033.22 - lr: 0.000002 - momentum: 0.000000
2023-10-23 15:08:34,865 epoch 10 - iter 91/138 - loss 0.01196000 - time (sec): 5.07 - samples/sec: 3093.54 - lr: 0.000001 - momentum: 0.000000
2023-10-23 15:08:35,600 epoch 10 - iter 104/138 - loss 0.01316873 - time (sec): 5.81 - samples/sec: 3074.63 - lr: 0.000001 - momentum: 0.000000
2023-10-23 15:08:36,309 epoch 10 - iter 117/138 - loss 0.01240979 - time (sec): 6.51 - samples/sec: 3021.99 - lr: 0.000001 - momentum: 0.000000
2023-10-23 15:08:37,035 epoch 10 - iter 130/138 - loss 0.01491721 - time (sec): 7.24 - samples/sec: 2991.14 - lr: 0.000000 - momentum: 0.000000
2023-10-23 15:08:37,473 ----------------------------------------------------------------------------------------------------
2023-10-23 15:08:37,473 EPOCH 10 done: loss 0.0146 - lr: 0.000000
2023-10-23 15:08:38,012 DEV : loss 0.1697995662689209 - f1-score (micro avg) 0.8779
2023-10-23 15:08:38,413 ----------------------------------------------------------------------------------------------------
2023-10-23 15:08:38,414 Loading model from best epoch ...
2023-10-23 15:08:40,173 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 15:08:40,847
Results:
- F-score (micro) 0.9089
- F-score (macro) 0.7794
- Accuracy 0.8411
By class:
precision recall f1-score support
scope 0.8814 0.8864 0.8839 176
pers 0.9615 0.9766 0.9690 128
work 0.9385 0.8243 0.8777 74
object 0.5000 0.5000 0.5000 2
loc 1.0000 0.5000 0.6667 2
micro avg 0.9173 0.9005 0.9089 382
macro avg 0.8563 0.7375 0.7794 382
weighted avg 0.9179 0.9005 0.9080 382
2023-10-23 15:08:40,847 ----------------------------------------------------------------------------------------------------
|