File size: 24,025 Bytes
379eb25 |
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 |
2023-10-17 10:39:35,326 ----------------------------------------------------------------------------------------------------
2023-10-17 10:39:35,327 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=25, bias=True)
(loss_function): CrossEntropyLoss()
)"
2023-10-17 10:39:35,327 ----------------------------------------------------------------------------------------------------
2023-10-17 10:39:35,327 MultiCorpus: 966 train + 219 dev + 204 test sentences
- NER_HIPE_2022 Corpus: 966 train + 219 dev + 204 test sentences - /root/.flair/datasets/ner_hipe_2022/v2.1/ajmc/fr/with_doc_seperator
2023-10-17 10:39:35,327 ----------------------------------------------------------------------------------------------------
2023-10-17 10:39:35,327 Train: 966 sentences
2023-10-17 10:39:35,327 (train_with_dev=False, train_with_test=False)
2023-10-17 10:39:35,327 ----------------------------------------------------------------------------------------------------
2023-10-17 10:39:35,327 Training Params:
2023-10-17 10:39:35,327 - learning_rate: "3e-05"
2023-10-17 10:39:35,327 - mini_batch_size: "8"
2023-10-17 10:39:35,327 - max_epochs: "10"
2023-10-17 10:39:35,327 - shuffle: "True"
2023-10-17 10:39:35,327 ----------------------------------------------------------------------------------------------------
2023-10-17 10:39:35,327 Plugins:
2023-10-17 10:39:35,327 - TensorboardLogger
2023-10-17 10:39:35,327 - LinearScheduler | warmup_fraction: '0.1'
2023-10-17 10:39:35,327 ----------------------------------------------------------------------------------------------------
2023-10-17 10:39:35,327 Final evaluation on model from best epoch (best-model.pt)
2023-10-17 10:39:35,327 - metric: "('micro avg', 'f1-score')"
2023-10-17 10:39:35,328 ----------------------------------------------------------------------------------------------------
2023-10-17 10:39:35,328 Computation:
2023-10-17 10:39:35,328 - compute on device: cuda:0
2023-10-17 10:39:35,328 - embedding storage: none
2023-10-17 10:39:35,328 ----------------------------------------------------------------------------------------------------
2023-10-17 10:39:35,328 Model training base path: "hmbench-ajmc/fr-hmteams/teams-base-historic-multilingual-discriminator-bs8-wsFalse-e10-lr3e-05-poolingfirst-layers-1-crfFalse-2"
2023-10-17 10:39:35,328 ----------------------------------------------------------------------------------------------------
2023-10-17 10:39:35,328 ----------------------------------------------------------------------------------------------------
2023-10-17 10:39:35,328 Logging anything other than scalars to TensorBoard is currently not supported.
2023-10-17 10:39:36,109 epoch 1 - iter 12/121 - loss 4.25722385 - time (sec): 0.78 - samples/sec: 2872.89 - lr: 0.000003 - momentum: 0.000000
2023-10-17 10:39:36,809 epoch 1 - iter 24/121 - loss 4.00564937 - time (sec): 1.48 - samples/sec: 3154.85 - lr: 0.000006 - momentum: 0.000000
2023-10-17 10:39:37,642 epoch 1 - iter 36/121 - loss 3.40093382 - time (sec): 2.31 - samples/sec: 3172.82 - lr: 0.000009 - momentum: 0.000000
2023-10-17 10:39:38,404 epoch 1 - iter 48/121 - loss 2.78062785 - time (sec): 3.07 - samples/sec: 3143.94 - lr: 0.000012 - momentum: 0.000000
2023-10-17 10:39:39,161 epoch 1 - iter 60/121 - loss 2.35231184 - time (sec): 3.83 - samples/sec: 3199.31 - lr: 0.000015 - momentum: 0.000000
2023-10-17 10:39:39,973 epoch 1 - iter 72/121 - loss 2.06127860 - time (sec): 4.64 - samples/sec: 3147.72 - lr: 0.000018 - momentum: 0.000000
2023-10-17 10:39:40,735 epoch 1 - iter 84/121 - loss 1.81693509 - time (sec): 5.41 - samples/sec: 3200.90 - lr: 0.000021 - momentum: 0.000000
2023-10-17 10:39:41,470 epoch 1 - iter 96/121 - loss 1.67323399 - time (sec): 6.14 - samples/sec: 3170.67 - lr: 0.000024 - momentum: 0.000000
2023-10-17 10:39:42,303 epoch 1 - iter 108/121 - loss 1.51544682 - time (sec): 6.97 - samples/sec: 3172.12 - lr: 0.000027 - momentum: 0.000000
2023-10-17 10:39:43,069 epoch 1 - iter 120/121 - loss 1.39049965 - time (sec): 7.74 - samples/sec: 3180.24 - lr: 0.000030 - momentum: 0.000000
2023-10-17 10:39:43,123 ----------------------------------------------------------------------------------------------------
2023-10-17 10:39:43,124 EPOCH 1 done: loss 1.3838 - lr: 0.000030
2023-10-17 10:39:43,777 DEV : loss 0.251099556684494 - f1-score (micro avg) 0.5079
2023-10-17 10:39:43,787 saving best model
2023-10-17 10:39:44,242 ----------------------------------------------------------------------------------------------------
2023-10-17 10:39:44,933 epoch 2 - iter 12/121 - loss 0.26699978 - time (sec): 0.69 - samples/sec: 3362.95 - lr: 0.000030 - momentum: 0.000000
2023-10-17 10:39:45,686 epoch 2 - iter 24/121 - loss 0.27485034 - time (sec): 1.44 - samples/sec: 3098.99 - lr: 0.000029 - momentum: 0.000000
2023-10-17 10:39:46,440 epoch 2 - iter 36/121 - loss 0.27317472 - time (sec): 2.20 - samples/sec: 3168.93 - lr: 0.000029 - momentum: 0.000000
2023-10-17 10:39:47,147 epoch 2 - iter 48/121 - loss 0.25723374 - time (sec): 2.90 - samples/sec: 3131.61 - lr: 0.000029 - momentum: 0.000000
2023-10-17 10:39:47,919 epoch 2 - iter 60/121 - loss 0.24909349 - time (sec): 3.68 - samples/sec: 3140.85 - lr: 0.000028 - momentum: 0.000000
2023-10-17 10:39:48,756 epoch 2 - iter 72/121 - loss 0.23902434 - time (sec): 4.51 - samples/sec: 3138.67 - lr: 0.000028 - momentum: 0.000000
2023-10-17 10:39:49,474 epoch 2 - iter 84/121 - loss 0.23187052 - time (sec): 5.23 - samples/sec: 3187.32 - lr: 0.000028 - momentum: 0.000000
2023-10-17 10:39:50,215 epoch 2 - iter 96/121 - loss 0.22252914 - time (sec): 5.97 - samples/sec: 3228.69 - lr: 0.000027 - momentum: 0.000000
2023-10-17 10:39:51,016 epoch 2 - iter 108/121 - loss 0.21712543 - time (sec): 6.77 - samples/sec: 3218.14 - lr: 0.000027 - momentum: 0.000000
2023-10-17 10:39:51,746 epoch 2 - iter 120/121 - loss 0.21132328 - time (sec): 7.50 - samples/sec: 3262.76 - lr: 0.000027 - momentum: 0.000000
2023-10-17 10:39:51,826 ----------------------------------------------------------------------------------------------------
2023-10-17 10:39:51,826 EPOCH 2 done: loss 0.2094 - lr: 0.000027
2023-10-17 10:39:52,611 DEV : loss 0.14902380108833313 - f1-score (micro avg) 0.75
2023-10-17 10:39:52,617 saving best model
2023-10-17 10:39:53,122 ----------------------------------------------------------------------------------------------------
2023-10-17 10:39:53,897 epoch 3 - iter 12/121 - loss 0.12685344 - time (sec): 0.77 - samples/sec: 3347.70 - lr: 0.000026 - momentum: 0.000000
2023-10-17 10:39:54,635 epoch 3 - iter 24/121 - loss 0.13397141 - time (sec): 1.51 - samples/sec: 3458.76 - lr: 0.000026 - momentum: 0.000000
2023-10-17 10:39:55,449 epoch 3 - iter 36/121 - loss 0.13558235 - time (sec): 2.33 - samples/sec: 3364.66 - lr: 0.000026 - momentum: 0.000000
2023-10-17 10:39:56,155 epoch 3 - iter 48/121 - loss 0.13474115 - time (sec): 3.03 - samples/sec: 3288.55 - lr: 0.000025 - momentum: 0.000000
2023-10-17 10:39:56,871 epoch 3 - iter 60/121 - loss 0.13146380 - time (sec): 3.75 - samples/sec: 3298.28 - lr: 0.000025 - momentum: 0.000000
2023-10-17 10:39:57,695 epoch 3 - iter 72/121 - loss 0.13411500 - time (sec): 4.57 - samples/sec: 3238.48 - lr: 0.000025 - momentum: 0.000000
2023-10-17 10:39:58,415 epoch 3 - iter 84/121 - loss 0.13560617 - time (sec): 5.29 - samples/sec: 3284.45 - lr: 0.000024 - momentum: 0.000000
2023-10-17 10:39:59,143 epoch 3 - iter 96/121 - loss 0.13216345 - time (sec): 6.02 - samples/sec: 3237.82 - lr: 0.000024 - momentum: 0.000000
2023-10-17 10:39:59,872 epoch 3 - iter 108/121 - loss 0.12725540 - time (sec): 6.75 - samples/sec: 3254.99 - lr: 0.000024 - momentum: 0.000000
2023-10-17 10:40:00,620 epoch 3 - iter 120/121 - loss 0.12351647 - time (sec): 7.50 - samples/sec: 3282.36 - lr: 0.000023 - momentum: 0.000000
2023-10-17 10:40:00,666 ----------------------------------------------------------------------------------------------------
2023-10-17 10:40:00,666 EPOCH 3 done: loss 0.1229 - lr: 0.000023
2023-10-17 10:40:01,626 DEV : loss 0.14644193649291992 - f1-score (micro avg) 0.7974
2023-10-17 10:40:01,632 saving best model
2023-10-17 10:40:02,148 ----------------------------------------------------------------------------------------------------
2023-10-17 10:40:02,872 epoch 4 - iter 12/121 - loss 0.11855343 - time (sec): 0.72 - samples/sec: 3318.24 - lr: 0.000023 - momentum: 0.000000
2023-10-17 10:40:03,569 epoch 4 - iter 24/121 - loss 0.11086880 - time (sec): 1.42 - samples/sec: 3358.02 - lr: 0.000023 - momentum: 0.000000
2023-10-17 10:40:04,292 epoch 4 - iter 36/121 - loss 0.09714372 - time (sec): 2.14 - samples/sec: 3198.21 - lr: 0.000022 - momentum: 0.000000
2023-10-17 10:40:05,039 epoch 4 - iter 48/121 - loss 0.09071078 - time (sec): 2.89 - samples/sec: 3157.80 - lr: 0.000022 - momentum: 0.000000
2023-10-17 10:40:05,821 epoch 4 - iter 60/121 - loss 0.09150231 - time (sec): 3.67 - samples/sec: 3188.65 - lr: 0.000022 - momentum: 0.000000
2023-10-17 10:40:06,595 epoch 4 - iter 72/121 - loss 0.08403862 - time (sec): 4.44 - samples/sec: 3289.27 - lr: 0.000021 - momentum: 0.000000
2023-10-17 10:40:07,362 epoch 4 - iter 84/121 - loss 0.08179926 - time (sec): 5.21 - samples/sec: 3286.29 - lr: 0.000021 - momentum: 0.000000
2023-10-17 10:40:08,181 epoch 4 - iter 96/121 - loss 0.08496227 - time (sec): 6.03 - samples/sec: 3289.16 - lr: 0.000021 - momentum: 0.000000
2023-10-17 10:40:08,894 epoch 4 - iter 108/121 - loss 0.08355173 - time (sec): 6.74 - samples/sec: 3304.38 - lr: 0.000020 - momentum: 0.000000
2023-10-17 10:40:09,660 epoch 4 - iter 120/121 - loss 0.08565829 - time (sec): 7.51 - samples/sec: 3276.12 - lr: 0.000020 - momentum: 0.000000
2023-10-17 10:40:09,706 ----------------------------------------------------------------------------------------------------
2023-10-17 10:40:09,707 EPOCH 4 done: loss 0.0857 - lr: 0.000020
2023-10-17 10:40:10,514 DEV : loss 0.14408640563488007 - f1-score (micro avg) 0.8069
2023-10-17 10:40:10,521 saving best model
2023-10-17 10:40:11,050 ----------------------------------------------------------------------------------------------------
2023-10-17 10:40:11,843 epoch 5 - iter 12/121 - loss 0.06469600 - time (sec): 0.79 - samples/sec: 3149.16 - lr: 0.000020 - momentum: 0.000000
2023-10-17 10:40:12,700 epoch 5 - iter 24/121 - loss 0.06320875 - time (sec): 1.65 - samples/sec: 3129.05 - lr: 0.000019 - momentum: 0.000000
2023-10-17 10:40:13,483 epoch 5 - iter 36/121 - loss 0.05527648 - time (sec): 2.43 - samples/sec: 3190.52 - lr: 0.000019 - momentum: 0.000000
2023-10-17 10:40:14,248 epoch 5 - iter 48/121 - loss 0.05466742 - time (sec): 3.20 - samples/sec: 3259.54 - lr: 0.000019 - momentum: 0.000000
2023-10-17 10:40:15,013 epoch 5 - iter 60/121 - loss 0.05439576 - time (sec): 3.96 - samples/sec: 3233.41 - lr: 0.000018 - momentum: 0.000000
2023-10-17 10:40:15,756 epoch 5 - iter 72/121 - loss 0.05872498 - time (sec): 4.70 - samples/sec: 3267.51 - lr: 0.000018 - momentum: 0.000000
2023-10-17 10:40:16,427 epoch 5 - iter 84/121 - loss 0.05848788 - time (sec): 5.37 - samples/sec: 3280.41 - lr: 0.000018 - momentum: 0.000000
2023-10-17 10:40:17,153 epoch 5 - iter 96/121 - loss 0.05846039 - time (sec): 6.10 - samples/sec: 3272.26 - lr: 0.000017 - momentum: 0.000000
2023-10-17 10:40:17,891 epoch 5 - iter 108/121 - loss 0.05853485 - time (sec): 6.84 - samples/sec: 3269.18 - lr: 0.000017 - momentum: 0.000000
2023-10-17 10:40:18,638 epoch 5 - iter 120/121 - loss 0.05724472 - time (sec): 7.59 - samples/sec: 3250.89 - lr: 0.000017 - momentum: 0.000000
2023-10-17 10:40:18,684 ----------------------------------------------------------------------------------------------------
2023-10-17 10:40:18,685 EPOCH 5 done: loss 0.0571 - lr: 0.000017
2023-10-17 10:40:19,474 DEV : loss 0.15564198791980743 - f1-score (micro avg) 0.8015
2023-10-17 10:40:19,480 ----------------------------------------------------------------------------------------------------
2023-10-17 10:40:20,222 epoch 6 - iter 12/121 - loss 0.04969618 - time (sec): 0.74 - samples/sec: 3225.61 - lr: 0.000016 - momentum: 0.000000
2023-10-17 10:40:20,960 epoch 6 - iter 24/121 - loss 0.04644119 - time (sec): 1.48 - samples/sec: 3078.69 - lr: 0.000016 - momentum: 0.000000
2023-10-17 10:40:21,753 epoch 6 - iter 36/121 - loss 0.04209743 - time (sec): 2.27 - samples/sec: 3148.81 - lr: 0.000016 - momentum: 0.000000
2023-10-17 10:40:22,501 epoch 6 - iter 48/121 - loss 0.04711594 - time (sec): 3.02 - samples/sec: 3220.15 - lr: 0.000015 - momentum: 0.000000
2023-10-17 10:40:23,295 epoch 6 - iter 60/121 - loss 0.04764666 - time (sec): 3.81 - samples/sec: 3223.55 - lr: 0.000015 - momentum: 0.000000
2023-10-17 10:40:24,073 epoch 6 - iter 72/121 - loss 0.04477288 - time (sec): 4.59 - samples/sec: 3214.32 - lr: 0.000015 - momentum: 0.000000
2023-10-17 10:40:24,807 epoch 6 - iter 84/121 - loss 0.04169976 - time (sec): 5.33 - samples/sec: 3205.75 - lr: 0.000014 - momentum: 0.000000
2023-10-17 10:40:25,540 epoch 6 - iter 96/121 - loss 0.04232871 - time (sec): 6.06 - samples/sec: 3233.25 - lr: 0.000014 - momentum: 0.000000
2023-10-17 10:40:26,304 epoch 6 - iter 108/121 - loss 0.04135200 - time (sec): 6.82 - samples/sec: 3234.30 - lr: 0.000014 - momentum: 0.000000
2023-10-17 10:40:27,015 epoch 6 - iter 120/121 - loss 0.04338782 - time (sec): 7.53 - samples/sec: 3270.66 - lr: 0.000013 - momentum: 0.000000
2023-10-17 10:40:27,061 ----------------------------------------------------------------------------------------------------
2023-10-17 10:40:27,061 EPOCH 6 done: loss 0.0434 - lr: 0.000013
2023-10-17 10:40:27,828 DEV : loss 0.158291295170784 - f1-score (micro avg) 0.8204
2023-10-17 10:40:27,833 saving best model
2023-10-17 10:40:28,330 ----------------------------------------------------------------------------------------------------
2023-10-17 10:40:29,054 epoch 7 - iter 12/121 - loss 0.04068845 - time (sec): 0.72 - samples/sec: 3223.97 - lr: 0.000013 - momentum: 0.000000
2023-10-17 10:40:29,841 epoch 7 - iter 24/121 - loss 0.02967958 - time (sec): 1.51 - samples/sec: 3316.62 - lr: 0.000013 - momentum: 0.000000
2023-10-17 10:40:30,547 epoch 7 - iter 36/121 - loss 0.02930156 - time (sec): 2.21 - samples/sec: 3237.98 - lr: 0.000012 - momentum: 0.000000
2023-10-17 10:40:31,282 epoch 7 - iter 48/121 - loss 0.03417830 - time (sec): 2.95 - samples/sec: 3281.96 - lr: 0.000012 - momentum: 0.000000
2023-10-17 10:40:32,050 epoch 7 - iter 60/121 - loss 0.03460234 - time (sec): 3.72 - samples/sec: 3308.07 - lr: 0.000012 - momentum: 0.000000
2023-10-17 10:40:32,784 epoch 7 - iter 72/121 - loss 0.03483963 - time (sec): 4.45 - samples/sec: 3336.06 - lr: 0.000011 - momentum: 0.000000
2023-10-17 10:40:33,557 epoch 7 - iter 84/121 - loss 0.03363358 - time (sec): 5.22 - samples/sec: 3322.19 - lr: 0.000011 - momentum: 0.000000
2023-10-17 10:40:34,261 epoch 7 - iter 96/121 - loss 0.03264425 - time (sec): 5.93 - samples/sec: 3317.81 - lr: 0.000011 - momentum: 0.000000
2023-10-17 10:40:35,055 epoch 7 - iter 108/121 - loss 0.03194370 - time (sec): 6.72 - samples/sec: 3313.03 - lr: 0.000010 - momentum: 0.000000
2023-10-17 10:40:35,820 epoch 7 - iter 120/121 - loss 0.03311731 - time (sec): 7.49 - samples/sec: 3286.01 - lr: 0.000010 - momentum: 0.000000
2023-10-17 10:40:35,875 ----------------------------------------------------------------------------------------------------
2023-10-17 10:40:35,876 EPOCH 7 done: loss 0.0330 - lr: 0.000010
2023-10-17 10:40:36,669 DEV : loss 0.1770782619714737 - f1-score (micro avg) 0.8185
2023-10-17 10:40:36,681 ----------------------------------------------------------------------------------------------------
2023-10-17 10:40:37,429 epoch 8 - iter 12/121 - loss 0.02318625 - time (sec): 0.75 - samples/sec: 3466.20 - lr: 0.000010 - momentum: 0.000000
2023-10-17 10:40:38,254 epoch 8 - iter 24/121 - loss 0.02179879 - time (sec): 1.57 - samples/sec: 3154.01 - lr: 0.000009 - momentum: 0.000000
2023-10-17 10:40:38,978 epoch 8 - iter 36/121 - loss 0.01865314 - time (sec): 2.30 - samples/sec: 3244.27 - lr: 0.000009 - momentum: 0.000000
2023-10-17 10:40:39,799 epoch 8 - iter 48/121 - loss 0.02183121 - time (sec): 3.12 - samples/sec: 3222.49 - lr: 0.000009 - momentum: 0.000000
2023-10-17 10:40:40,533 epoch 8 - iter 60/121 - loss 0.02294907 - time (sec): 3.85 - samples/sec: 3239.56 - lr: 0.000008 - momentum: 0.000000
2023-10-17 10:40:41,312 epoch 8 - iter 72/121 - loss 0.02283248 - time (sec): 4.63 - samples/sec: 3244.31 - lr: 0.000008 - momentum: 0.000000
2023-10-17 10:40:42,039 epoch 8 - iter 84/121 - loss 0.02507082 - time (sec): 5.36 - samples/sec: 3259.64 - lr: 0.000008 - momentum: 0.000000
2023-10-17 10:40:42,799 epoch 8 - iter 96/121 - loss 0.02396940 - time (sec): 6.12 - samples/sec: 3244.75 - lr: 0.000008 - momentum: 0.000000
2023-10-17 10:40:43,575 epoch 8 - iter 108/121 - loss 0.02353728 - time (sec): 6.89 - samples/sec: 3224.57 - lr: 0.000007 - momentum: 0.000000
2023-10-17 10:40:44,343 epoch 8 - iter 120/121 - loss 0.02497422 - time (sec): 7.66 - samples/sec: 3211.10 - lr: 0.000007 - momentum: 0.000000
2023-10-17 10:40:44,397 ----------------------------------------------------------------------------------------------------
2023-10-17 10:40:44,398 EPOCH 8 done: loss 0.0249 - lr: 0.000007
2023-10-17 10:40:45,177 DEV : loss 0.19383026659488678 - f1-score (micro avg) 0.8306
2023-10-17 10:40:45,183 saving best model
2023-10-17 10:40:45,701 ----------------------------------------------------------------------------------------------------
2023-10-17 10:40:46,480 epoch 9 - iter 12/121 - loss 0.02257339 - time (sec): 0.77 - samples/sec: 3397.27 - lr: 0.000006 - momentum: 0.000000
2023-10-17 10:40:47,200 epoch 9 - iter 24/121 - loss 0.02788200 - time (sec): 1.48 - samples/sec: 3334.14 - lr: 0.000006 - momentum: 0.000000
2023-10-17 10:40:47,941 epoch 9 - iter 36/121 - loss 0.02369329 - time (sec): 2.23 - samples/sec: 3205.46 - lr: 0.000006 - momentum: 0.000000
2023-10-17 10:40:48,712 epoch 9 - iter 48/121 - loss 0.02520442 - time (sec): 3.00 - samples/sec: 3205.41 - lr: 0.000006 - momentum: 0.000000
2023-10-17 10:40:49,476 epoch 9 - iter 60/121 - loss 0.02227888 - time (sec): 3.76 - samples/sec: 3217.52 - lr: 0.000005 - momentum: 0.000000
2023-10-17 10:40:50,296 epoch 9 - iter 72/121 - loss 0.02140296 - time (sec): 4.58 - samples/sec: 3213.21 - lr: 0.000005 - momentum: 0.000000
2023-10-17 10:40:50,996 epoch 9 - iter 84/121 - loss 0.01988658 - time (sec): 5.28 - samples/sec: 3224.77 - lr: 0.000005 - momentum: 0.000000
2023-10-17 10:40:51,791 epoch 9 - iter 96/121 - loss 0.02156552 - time (sec): 6.08 - samples/sec: 3214.58 - lr: 0.000004 - momentum: 0.000000
2023-10-17 10:40:52,531 epoch 9 - iter 108/121 - loss 0.02013316 - time (sec): 6.82 - samples/sec: 3232.85 - lr: 0.000004 - momentum: 0.000000
2023-10-17 10:40:53,309 epoch 9 - iter 120/121 - loss 0.01851873 - time (sec): 7.59 - samples/sec: 3240.54 - lr: 0.000004 - momentum: 0.000000
2023-10-17 10:40:53,365 ----------------------------------------------------------------------------------------------------
2023-10-17 10:40:53,366 EPOCH 9 done: loss 0.0186 - lr: 0.000004
2023-10-17 10:40:54,229 DEV : loss 0.1929878145456314 - f1-score (micro avg) 0.8363
2023-10-17 10:40:54,235 saving best model
2023-10-17 10:40:54,758 ----------------------------------------------------------------------------------------------------
2023-10-17 10:40:55,485 epoch 10 - iter 12/121 - loss 0.03533969 - time (sec): 0.72 - samples/sec: 3361.11 - lr: 0.000003 - momentum: 0.000000
2023-10-17 10:40:56,257 epoch 10 - iter 24/121 - loss 0.03073965 - time (sec): 1.49 - samples/sec: 3256.62 - lr: 0.000003 - momentum: 0.000000
2023-10-17 10:40:57,027 epoch 10 - iter 36/121 - loss 0.02357388 - time (sec): 2.26 - samples/sec: 3159.50 - lr: 0.000003 - momentum: 0.000000
2023-10-17 10:40:57,804 epoch 10 - iter 48/121 - loss 0.02346600 - time (sec): 3.04 - samples/sec: 3225.02 - lr: 0.000002 - momentum: 0.000000
2023-10-17 10:40:58,572 epoch 10 - iter 60/121 - loss 0.02331925 - time (sec): 3.81 - samples/sec: 3251.15 - lr: 0.000002 - momentum: 0.000000
2023-10-17 10:40:59,365 epoch 10 - iter 72/121 - loss 0.02129171 - time (sec): 4.60 - samples/sec: 3191.76 - lr: 0.000002 - momentum: 0.000000
2023-10-17 10:41:00,107 epoch 10 - iter 84/121 - loss 0.02077126 - time (sec): 5.34 - samples/sec: 3216.04 - lr: 0.000001 - momentum: 0.000000
2023-10-17 10:41:00,884 epoch 10 - iter 96/121 - loss 0.01902530 - time (sec): 6.12 - samples/sec: 3199.96 - lr: 0.000001 - momentum: 0.000000
2023-10-17 10:41:01,678 epoch 10 - iter 108/121 - loss 0.01742089 - time (sec): 6.92 - samples/sec: 3187.95 - lr: 0.000001 - momentum: 0.000000
2023-10-17 10:41:02,493 epoch 10 - iter 120/121 - loss 0.01642575 - time (sec): 7.73 - samples/sec: 3181.68 - lr: 0.000000 - momentum: 0.000000
2023-10-17 10:41:02,541 ----------------------------------------------------------------------------------------------------
2023-10-17 10:41:02,541 EPOCH 10 done: loss 0.0164 - lr: 0.000000
2023-10-17 10:41:03,340 DEV : loss 0.19614964723587036 - f1-score (micro avg) 0.8416
2023-10-17 10:41:03,345 saving best model
2023-10-17 10:41:04,256 ----------------------------------------------------------------------------------------------------
2023-10-17 10:41:04,257 Loading model from best epoch ...
2023-10-17 10:41:05,657 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
2023-10-17 10:41:06,551
Results:
- F-score (micro) 0.8263
- F-score (macro) 0.563
- Accuracy 0.7225
By class:
precision recall f1-score support
pers 0.8531 0.8777 0.8652 139
scope 0.8444 0.8837 0.8636 129
work 0.7111 0.8000 0.7529 80
loc 0.6667 0.2222 0.3333 9
date 0.0000 0.0000 0.0000 3
micro avg 0.8140 0.8389 0.8263 360
macro avg 0.6151 0.5567 0.5630 360
weighted avg 0.8067 0.8389 0.8192 360
2023-10-17 10:41:06,552 ----------------------------------------------------------------------------------------------------
|