File size: 24,135 Bytes
10c3208 |
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 |
2023-10-17 14:24:14,252 ----------------------------------------------------------------------------------------------------
2023-10-17 14:24:14,254 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 14:24:14,254 ----------------------------------------------------------------------------------------------------
2023-10-17 14:24:14,254 MultiCorpus: 20847 train + 1123 dev + 3350 test sentences
- NER_HIPE_2022 Corpus: 20847 train + 1123 dev + 3350 test sentences - /root/.flair/datasets/ner_hipe_2022/v2.1/newseye/de/with_doc_seperator
2023-10-17 14:24:14,254 ----------------------------------------------------------------------------------------------------
2023-10-17 14:24:14,254 Train: 20847 sentences
2023-10-17 14:24:14,254 (train_with_dev=False, train_with_test=False)
2023-10-17 14:24:14,254 ----------------------------------------------------------------------------------------------------
2023-10-17 14:24:14,254 Training Params:
2023-10-17 14:24:14,255 - learning_rate: "5e-05"
2023-10-17 14:24:14,255 - mini_batch_size: "8"
2023-10-17 14:24:14,255 - max_epochs: "10"
2023-10-17 14:24:14,255 - shuffle: "True"
2023-10-17 14:24:14,255 ----------------------------------------------------------------------------------------------------
2023-10-17 14:24:14,255 Plugins:
2023-10-17 14:24:14,255 - TensorboardLogger
2023-10-17 14:24:14,255 - LinearScheduler | warmup_fraction: '0.1'
2023-10-17 14:24:14,255 ----------------------------------------------------------------------------------------------------
2023-10-17 14:24:14,255 Final evaluation on model from best epoch (best-model.pt)
2023-10-17 14:24:14,255 - metric: "('micro avg', 'f1-score')"
2023-10-17 14:24:14,255 ----------------------------------------------------------------------------------------------------
2023-10-17 14:24:14,255 Computation:
2023-10-17 14:24:14,255 - compute on device: cuda:0
2023-10-17 14:24:14,256 - embedding storage: none
2023-10-17 14:24:14,256 ----------------------------------------------------------------------------------------------------
2023-10-17 14:24:14,256 Model training base path: "hmbench-newseye/de-hmteams/teams-base-historic-multilingual-discriminator-bs8-wsFalse-e10-lr5e-05-poolingfirst-layers-1-crfFalse-2"
2023-10-17 14:24:14,256 ----------------------------------------------------------------------------------------------------
2023-10-17 14:24:14,256 ----------------------------------------------------------------------------------------------------
2023-10-17 14:24:14,256 Logging anything other than scalars to TensorBoard is currently not supported.
2023-10-17 14:24:41,540 epoch 1 - iter 260/2606 - loss 1.85327979 - time (sec): 27.28 - samples/sec: 1310.12 - lr: 0.000005 - momentum: 0.000000
2023-10-17 14:25:09,627 epoch 1 - iter 520/2606 - loss 1.10728031 - time (sec): 55.37 - samples/sec: 1331.11 - lr: 0.000010 - momentum: 0.000000
2023-10-17 14:25:36,534 epoch 1 - iter 780/2606 - loss 0.83992496 - time (sec): 82.28 - samples/sec: 1359.25 - lr: 0.000015 - momentum: 0.000000
2023-10-17 14:26:04,493 epoch 1 - iter 1040/2606 - loss 0.68848892 - time (sec): 110.24 - samples/sec: 1356.82 - lr: 0.000020 - momentum: 0.000000
2023-10-17 14:26:31,440 epoch 1 - iter 1300/2606 - loss 0.59467316 - time (sec): 137.18 - samples/sec: 1364.66 - lr: 0.000025 - momentum: 0.000000
2023-10-17 14:26:57,934 epoch 1 - iter 1560/2606 - loss 0.53878529 - time (sec): 163.68 - samples/sec: 1360.76 - lr: 0.000030 - momentum: 0.000000
2023-10-17 14:27:25,638 epoch 1 - iter 1820/2606 - loss 0.48970257 - time (sec): 191.38 - samples/sec: 1363.10 - lr: 0.000035 - momentum: 0.000000
2023-10-17 14:27:53,036 epoch 1 - iter 2080/2606 - loss 0.45602118 - time (sec): 218.78 - samples/sec: 1349.95 - lr: 0.000040 - momentum: 0.000000
2023-10-17 14:28:20,128 epoch 1 - iter 2340/2606 - loss 0.42695326 - time (sec): 245.87 - samples/sec: 1350.40 - lr: 0.000045 - momentum: 0.000000
2023-10-17 14:28:46,892 epoch 1 - iter 2600/2606 - loss 0.40509791 - time (sec): 272.63 - samples/sec: 1345.44 - lr: 0.000050 - momentum: 0.000000
2023-10-17 14:28:47,527 ----------------------------------------------------------------------------------------------------
2023-10-17 14:28:47,527 EPOCH 1 done: loss 0.4046 - lr: 0.000050
2023-10-17 14:28:55,176 DEV : loss 0.14022067189216614 - f1-score (micro avg) 0.2959
2023-10-17 14:28:55,235 saving best model
2023-10-17 14:28:55,813 ----------------------------------------------------------------------------------------------------
2023-10-17 14:29:25,084 epoch 2 - iter 260/2606 - loss 0.17806520 - time (sec): 29.27 - samples/sec: 1262.84 - lr: 0.000049 - momentum: 0.000000
2023-10-17 14:29:52,834 epoch 2 - iter 520/2606 - loss 0.18056473 - time (sec): 57.02 - samples/sec: 1291.61 - lr: 0.000049 - momentum: 0.000000
2023-10-17 14:30:21,288 epoch 2 - iter 780/2606 - loss 0.17470102 - time (sec): 85.47 - samples/sec: 1294.70 - lr: 0.000048 - momentum: 0.000000
2023-10-17 14:30:48,148 epoch 2 - iter 1040/2606 - loss 0.17221863 - time (sec): 112.33 - samples/sec: 1308.82 - lr: 0.000048 - momentum: 0.000000
2023-10-17 14:31:16,076 epoch 2 - iter 1300/2606 - loss 0.16753259 - time (sec): 140.26 - samples/sec: 1324.35 - lr: 0.000047 - momentum: 0.000000
2023-10-17 14:31:43,316 epoch 2 - iter 1560/2606 - loss 0.16685811 - time (sec): 167.50 - samples/sec: 1327.25 - lr: 0.000047 - momentum: 0.000000
2023-10-17 14:32:10,687 epoch 2 - iter 1820/2606 - loss 0.16650143 - time (sec): 194.87 - samples/sec: 1322.48 - lr: 0.000046 - momentum: 0.000000
2023-10-17 14:32:39,701 epoch 2 - iter 2080/2606 - loss 0.16181597 - time (sec): 223.89 - samples/sec: 1313.90 - lr: 0.000046 - momentum: 0.000000
2023-10-17 14:33:08,969 epoch 2 - iter 2340/2606 - loss 0.15958413 - time (sec): 253.15 - samples/sec: 1308.43 - lr: 0.000045 - momentum: 0.000000
2023-10-17 14:33:36,676 epoch 2 - iter 2600/2606 - loss 0.16113565 - time (sec): 280.86 - samples/sec: 1306.23 - lr: 0.000044 - momentum: 0.000000
2023-10-17 14:33:37,189 ----------------------------------------------------------------------------------------------------
2023-10-17 14:33:37,189 EPOCH 2 done: loss 0.1610 - lr: 0.000044
2023-10-17 14:33:49,462 DEV : loss 0.16424185037612915 - f1-score (micro avg) 0.2857
2023-10-17 14:33:49,517 ----------------------------------------------------------------------------------------------------
2023-10-17 14:34:18,273 epoch 3 - iter 260/2606 - loss 0.13142213 - time (sec): 28.75 - samples/sec: 1297.78 - lr: 0.000044 - momentum: 0.000000
2023-10-17 14:34:45,491 epoch 3 - iter 520/2606 - loss 0.13076265 - time (sec): 55.97 - samples/sec: 1298.02 - lr: 0.000043 - momentum: 0.000000
2023-10-17 14:35:13,286 epoch 3 - iter 780/2606 - loss 0.12346611 - time (sec): 83.77 - samples/sec: 1272.88 - lr: 0.000043 - momentum: 0.000000
2023-10-17 14:35:43,054 epoch 3 - iter 1040/2606 - loss 0.12496563 - time (sec): 113.53 - samples/sec: 1282.05 - lr: 0.000042 - momentum: 0.000000
2023-10-17 14:36:09,740 epoch 3 - iter 1300/2606 - loss 0.11945146 - time (sec): 140.22 - samples/sec: 1309.70 - lr: 0.000042 - momentum: 0.000000
2023-10-17 14:36:37,350 epoch 3 - iter 1560/2606 - loss 0.11698591 - time (sec): 167.83 - samples/sec: 1317.13 - lr: 0.000041 - momentum: 0.000000
2023-10-17 14:37:03,654 epoch 3 - iter 1820/2606 - loss 0.11691007 - time (sec): 194.13 - samples/sec: 1311.04 - lr: 0.000041 - momentum: 0.000000
2023-10-17 14:37:31,807 epoch 3 - iter 2080/2606 - loss 0.11634199 - time (sec): 222.29 - samples/sec: 1314.95 - lr: 0.000040 - momentum: 0.000000
2023-10-17 14:38:00,450 epoch 3 - iter 2340/2606 - loss 0.11599351 - time (sec): 250.93 - samples/sec: 1310.02 - lr: 0.000039 - momentum: 0.000000
2023-10-17 14:38:28,873 epoch 3 - iter 2600/2606 - loss 0.11610612 - time (sec): 279.35 - samples/sec: 1311.26 - lr: 0.000039 - momentum: 0.000000
2023-10-17 14:38:29,622 ----------------------------------------------------------------------------------------------------
2023-10-17 14:38:29,622 EPOCH 3 done: loss 0.1159 - lr: 0.000039
2023-10-17 14:38:40,939 DEV : loss 0.16876201331615448 - f1-score (micro avg) 0.3775
2023-10-17 14:38:41,009 saving best model
2023-10-17 14:38:42,469 ----------------------------------------------------------------------------------------------------
2023-10-17 14:39:11,890 epoch 4 - iter 260/2606 - loss 0.07361030 - time (sec): 29.42 - samples/sec: 1265.45 - lr: 0.000038 - momentum: 0.000000
2023-10-17 14:39:39,718 epoch 4 - iter 520/2606 - loss 0.07660504 - time (sec): 57.24 - samples/sec: 1283.14 - lr: 0.000038 - momentum: 0.000000
2023-10-17 14:40:07,171 epoch 4 - iter 780/2606 - loss 0.07902913 - time (sec): 84.70 - samples/sec: 1289.64 - lr: 0.000037 - momentum: 0.000000
2023-10-17 14:40:34,548 epoch 4 - iter 1040/2606 - loss 0.08287969 - time (sec): 112.07 - samples/sec: 1295.86 - lr: 0.000037 - momentum: 0.000000
2023-10-17 14:41:01,134 epoch 4 - iter 1300/2606 - loss 0.08241178 - time (sec): 138.66 - samples/sec: 1322.92 - lr: 0.000036 - momentum: 0.000000
2023-10-17 14:41:28,978 epoch 4 - iter 1560/2606 - loss 0.08250940 - time (sec): 166.50 - samples/sec: 1317.76 - lr: 0.000036 - momentum: 0.000000
2023-10-17 14:41:56,865 epoch 4 - iter 1820/2606 - loss 0.08213798 - time (sec): 194.39 - samples/sec: 1319.13 - lr: 0.000035 - momentum: 0.000000
2023-10-17 14:42:24,148 epoch 4 - iter 2080/2606 - loss 0.08166340 - time (sec): 221.67 - samples/sec: 1314.38 - lr: 0.000034 - momentum: 0.000000
2023-10-17 14:42:51,904 epoch 4 - iter 2340/2606 - loss 0.08556568 - time (sec): 249.43 - samples/sec: 1323.51 - lr: 0.000034 - momentum: 0.000000
2023-10-17 14:43:18,720 epoch 4 - iter 2600/2606 - loss 0.08865792 - time (sec): 276.24 - samples/sec: 1326.87 - lr: 0.000033 - momentum: 0.000000
2023-10-17 14:43:19,407 ----------------------------------------------------------------------------------------------------
2023-10-17 14:43:19,408 EPOCH 4 done: loss 0.0885 - lr: 0.000033
2023-10-17 14:43:30,616 DEV : loss 0.3115158677101135 - f1-score (micro avg) 0.3586
2023-10-17 14:43:30,674 ----------------------------------------------------------------------------------------------------
2023-10-17 14:43:58,778 epoch 5 - iter 260/2606 - loss 0.05863728 - time (sec): 28.10 - samples/sec: 1338.70 - lr: 0.000033 - momentum: 0.000000
2023-10-17 14:44:26,178 epoch 5 - iter 520/2606 - loss 0.06625558 - time (sec): 55.50 - samples/sec: 1348.15 - lr: 0.000032 - momentum: 0.000000
2023-10-17 14:44:54,516 epoch 5 - iter 780/2606 - loss 0.06844922 - time (sec): 83.84 - samples/sec: 1359.79 - lr: 0.000032 - momentum: 0.000000
2023-10-17 14:45:23,817 epoch 5 - iter 1040/2606 - loss 0.06831173 - time (sec): 113.14 - samples/sec: 1346.93 - lr: 0.000031 - momentum: 0.000000
2023-10-17 14:45:53,393 epoch 5 - iter 1300/2606 - loss 0.06365912 - time (sec): 142.72 - samples/sec: 1319.89 - lr: 0.000031 - momentum: 0.000000
2023-10-17 14:46:23,598 epoch 5 - iter 1560/2606 - loss 0.06547327 - time (sec): 172.92 - samples/sec: 1305.77 - lr: 0.000030 - momentum: 0.000000
2023-10-17 14:46:50,708 epoch 5 - iter 1820/2606 - loss 0.06474902 - time (sec): 200.03 - samples/sec: 1311.72 - lr: 0.000029 - momentum: 0.000000
2023-10-17 14:47:18,016 epoch 5 - iter 2080/2606 - loss 0.06304130 - time (sec): 227.34 - samples/sec: 1306.77 - lr: 0.000029 - momentum: 0.000000
2023-10-17 14:47:44,935 epoch 5 - iter 2340/2606 - loss 0.06258972 - time (sec): 254.26 - samples/sec: 1308.18 - lr: 0.000028 - momentum: 0.000000
2023-10-17 14:48:11,273 epoch 5 - iter 2600/2606 - loss 0.06247218 - time (sec): 280.60 - samples/sec: 1306.66 - lr: 0.000028 - momentum: 0.000000
2023-10-17 14:48:11,872 ----------------------------------------------------------------------------------------------------
2023-10-17 14:48:11,872 EPOCH 5 done: loss 0.0626 - lr: 0.000028
2023-10-17 14:48:22,903 DEV : loss 0.35754862427711487 - f1-score (micro avg) 0.3492
2023-10-17 14:48:22,961 ----------------------------------------------------------------------------------------------------
2023-10-17 14:48:50,441 epoch 6 - iter 260/2606 - loss 0.04153328 - time (sec): 27.48 - samples/sec: 1300.53 - lr: 0.000027 - momentum: 0.000000
2023-10-17 14:49:19,441 epoch 6 - iter 520/2606 - loss 0.04069889 - time (sec): 56.48 - samples/sec: 1292.76 - lr: 0.000027 - momentum: 0.000000
2023-10-17 14:49:47,061 epoch 6 - iter 780/2606 - loss 0.04221365 - time (sec): 84.10 - samples/sec: 1265.34 - lr: 0.000026 - momentum: 0.000000
2023-10-17 14:50:14,629 epoch 6 - iter 1040/2606 - loss 0.04286026 - time (sec): 111.66 - samples/sec: 1276.99 - lr: 0.000026 - momentum: 0.000000
2023-10-17 14:50:44,251 epoch 6 - iter 1300/2606 - loss 0.04244173 - time (sec): 141.29 - samples/sec: 1281.41 - lr: 0.000025 - momentum: 0.000000
2023-10-17 14:51:13,157 epoch 6 - iter 1560/2606 - loss 0.04269662 - time (sec): 170.19 - samples/sec: 1275.44 - lr: 0.000024 - momentum: 0.000000
2023-10-17 14:51:40,843 epoch 6 - iter 1820/2606 - loss 0.04514594 - time (sec): 197.88 - samples/sec: 1283.68 - lr: 0.000024 - momentum: 0.000000
2023-10-17 14:52:08,773 epoch 6 - iter 2080/2606 - loss 0.04536499 - time (sec): 225.81 - samples/sec: 1297.42 - lr: 0.000023 - momentum: 0.000000
2023-10-17 14:52:36,271 epoch 6 - iter 2340/2606 - loss 0.04581858 - time (sec): 253.31 - samples/sec: 1296.31 - lr: 0.000023 - momentum: 0.000000
2023-10-17 14:53:03,992 epoch 6 - iter 2600/2606 - loss 0.04576906 - time (sec): 281.03 - samples/sec: 1303.50 - lr: 0.000022 - momentum: 0.000000
2023-10-17 14:53:04,741 ----------------------------------------------------------------------------------------------------
2023-10-17 14:53:04,741 EPOCH 6 done: loss 0.0457 - lr: 0.000022
2023-10-17 14:53:15,359 DEV : loss 0.3733910918235779 - f1-score (micro avg) 0.3597
2023-10-17 14:53:15,411 ----------------------------------------------------------------------------------------------------
2023-10-17 14:53:43,296 epoch 7 - iter 260/2606 - loss 0.03044292 - time (sec): 27.88 - samples/sec: 1313.32 - lr: 0.000022 - momentum: 0.000000
2023-10-17 14:54:10,604 epoch 7 - iter 520/2606 - loss 0.02603500 - time (sec): 55.19 - samples/sec: 1346.10 - lr: 0.000021 - momentum: 0.000000
2023-10-17 14:54:37,579 epoch 7 - iter 780/2606 - loss 0.02879241 - time (sec): 82.17 - samples/sec: 1329.28 - lr: 0.000021 - momentum: 0.000000
2023-10-17 14:55:06,764 epoch 7 - iter 1040/2606 - loss 0.02855577 - time (sec): 111.35 - samples/sec: 1329.11 - lr: 0.000020 - momentum: 0.000000
2023-10-17 14:55:34,201 epoch 7 - iter 1300/2606 - loss 0.02824816 - time (sec): 138.79 - samples/sec: 1343.78 - lr: 0.000019 - momentum: 0.000000
2023-10-17 14:56:01,313 epoch 7 - iter 1560/2606 - loss 0.03022833 - time (sec): 165.90 - samples/sec: 1346.40 - lr: 0.000019 - momentum: 0.000000
2023-10-17 14:56:29,641 epoch 7 - iter 1820/2606 - loss 0.02870834 - time (sec): 194.23 - samples/sec: 1345.69 - lr: 0.000018 - momentum: 0.000000
2023-10-17 14:56:56,189 epoch 7 - iter 2080/2606 - loss 0.02859220 - time (sec): 220.78 - samples/sec: 1350.30 - lr: 0.000018 - momentum: 0.000000
2023-10-17 14:57:23,440 epoch 7 - iter 2340/2606 - loss 0.02938788 - time (sec): 248.03 - samples/sec: 1338.16 - lr: 0.000017 - momentum: 0.000000
2023-10-17 14:57:49,641 epoch 7 - iter 2600/2606 - loss 0.02928523 - time (sec): 274.23 - samples/sec: 1338.32 - lr: 0.000017 - momentum: 0.000000
2023-10-17 14:57:50,166 ----------------------------------------------------------------------------------------------------
2023-10-17 14:57:50,166 EPOCH 7 done: loss 0.0293 - lr: 0.000017
2023-10-17 14:58:01,362 DEV : loss 0.39208441972732544 - f1-score (micro avg) 0.376
2023-10-17 14:58:01,418 ----------------------------------------------------------------------------------------------------
2023-10-17 14:58:29,386 epoch 8 - iter 260/2606 - loss 0.01804159 - time (sec): 27.97 - samples/sec: 1244.37 - lr: 0.000016 - momentum: 0.000000
2023-10-17 14:58:55,174 epoch 8 - iter 520/2606 - loss 0.01839017 - time (sec): 53.75 - samples/sec: 1315.63 - lr: 0.000016 - momentum: 0.000000
2023-10-17 14:59:21,421 epoch 8 - iter 780/2606 - loss 0.01982613 - time (sec): 80.00 - samples/sec: 1364.08 - lr: 0.000015 - momentum: 0.000000
2023-10-17 14:59:48,015 epoch 8 - iter 1040/2606 - loss 0.02178505 - time (sec): 106.60 - samples/sec: 1358.26 - lr: 0.000014 - momentum: 0.000000
2023-10-17 15:00:15,164 epoch 8 - iter 1300/2606 - loss 0.02262431 - time (sec): 133.74 - samples/sec: 1348.86 - lr: 0.000014 - momentum: 0.000000
2023-10-17 15:00:42,859 epoch 8 - iter 1560/2606 - loss 0.02170814 - time (sec): 161.44 - samples/sec: 1343.38 - lr: 0.000013 - momentum: 0.000000
2023-10-17 15:01:10,224 epoch 8 - iter 1820/2606 - loss 0.02174864 - time (sec): 188.80 - samples/sec: 1350.02 - lr: 0.000013 - momentum: 0.000000
2023-10-17 15:01:38,355 epoch 8 - iter 2080/2606 - loss 0.02146193 - time (sec): 216.93 - samples/sec: 1343.56 - lr: 0.000012 - momentum: 0.000000
2023-10-17 15:02:07,541 epoch 8 - iter 2340/2606 - loss 0.02265158 - time (sec): 246.12 - samples/sec: 1338.24 - lr: 0.000012 - momentum: 0.000000
2023-10-17 15:02:34,852 epoch 8 - iter 2600/2606 - loss 0.02262534 - time (sec): 273.43 - samples/sec: 1340.03 - lr: 0.000011 - momentum: 0.000000
2023-10-17 15:02:35,458 ----------------------------------------------------------------------------------------------------
2023-10-17 15:02:35,458 EPOCH 8 done: loss 0.0227 - lr: 0.000011
2023-10-17 15:02:47,842 DEV : loss 0.4110426604747772 - f1-score (micro avg) 0.3964
2023-10-17 15:02:47,905 saving best model
2023-10-17 15:02:49,402 ----------------------------------------------------------------------------------------------------
2023-10-17 15:03:16,976 epoch 9 - iter 260/2606 - loss 0.01333780 - time (sec): 27.57 - samples/sec: 1333.90 - lr: 0.000011 - momentum: 0.000000
2023-10-17 15:03:44,619 epoch 9 - iter 520/2606 - loss 0.01398241 - time (sec): 55.21 - samples/sec: 1356.45 - lr: 0.000010 - momentum: 0.000000
2023-10-17 15:04:11,530 epoch 9 - iter 780/2606 - loss 0.01371935 - time (sec): 82.12 - samples/sec: 1357.95 - lr: 0.000009 - momentum: 0.000000
2023-10-17 15:04:38,342 epoch 9 - iter 1040/2606 - loss 0.01381938 - time (sec): 108.93 - samples/sec: 1354.11 - lr: 0.000009 - momentum: 0.000000
2023-10-17 15:05:04,756 epoch 9 - iter 1300/2606 - loss 0.01407551 - time (sec): 135.35 - samples/sec: 1366.05 - lr: 0.000008 - momentum: 0.000000
2023-10-17 15:05:33,009 epoch 9 - iter 1560/2606 - loss 0.01393126 - time (sec): 163.60 - samples/sec: 1357.67 - lr: 0.000008 - momentum: 0.000000
2023-10-17 15:06:00,019 epoch 9 - iter 1820/2606 - loss 0.01360054 - time (sec): 190.61 - samples/sec: 1344.36 - lr: 0.000007 - momentum: 0.000000
2023-10-17 15:06:29,986 epoch 9 - iter 2080/2606 - loss 0.01397962 - time (sec): 220.58 - samples/sec: 1349.15 - lr: 0.000007 - momentum: 0.000000
2023-10-17 15:06:55,256 epoch 9 - iter 2340/2606 - loss 0.01383784 - time (sec): 245.85 - samples/sec: 1344.69 - lr: 0.000006 - momentum: 0.000000
2023-10-17 15:07:21,400 epoch 9 - iter 2600/2606 - loss 0.01329672 - time (sec): 271.99 - samples/sec: 1348.44 - lr: 0.000006 - momentum: 0.000000
2023-10-17 15:07:21,934 ----------------------------------------------------------------------------------------------------
2023-10-17 15:07:21,935 EPOCH 9 done: loss 0.0133 - lr: 0.000006
2023-10-17 15:07:33,497 DEV : loss 0.487269788980484 - f1-score (micro avg) 0.3764
2023-10-17 15:07:33,558 ----------------------------------------------------------------------------------------------------
2023-10-17 15:08:01,040 epoch 10 - iter 260/2606 - loss 0.00522579 - time (sec): 27.48 - samples/sec: 1369.25 - lr: 0.000005 - momentum: 0.000000
2023-10-17 15:08:27,874 epoch 10 - iter 520/2606 - loss 0.00713905 - time (sec): 54.31 - samples/sec: 1357.92 - lr: 0.000004 - momentum: 0.000000
2023-10-17 15:08:54,168 epoch 10 - iter 780/2606 - loss 0.00706736 - time (sec): 80.61 - samples/sec: 1347.44 - lr: 0.000004 - momentum: 0.000000
2023-10-17 15:09:19,702 epoch 10 - iter 1040/2606 - loss 0.00700287 - time (sec): 106.14 - samples/sec: 1349.51 - lr: 0.000003 - momentum: 0.000000
2023-10-17 15:09:47,107 epoch 10 - iter 1300/2606 - loss 0.00826984 - time (sec): 133.55 - samples/sec: 1343.57 - lr: 0.000003 - momentum: 0.000000
2023-10-17 15:10:13,743 epoch 10 - iter 1560/2606 - loss 0.00865486 - time (sec): 160.18 - samples/sec: 1338.36 - lr: 0.000002 - momentum: 0.000000
2023-10-17 15:10:39,805 epoch 10 - iter 1820/2606 - loss 0.00972362 - time (sec): 186.24 - samples/sec: 1342.74 - lr: 0.000002 - momentum: 0.000000
2023-10-17 15:11:07,610 epoch 10 - iter 2080/2606 - loss 0.01024238 - time (sec): 214.05 - samples/sec: 1347.46 - lr: 0.000001 - momentum: 0.000000
2023-10-17 15:11:36,590 epoch 10 - iter 2340/2606 - loss 0.01033785 - time (sec): 243.03 - samples/sec: 1353.12 - lr: 0.000001 - momentum: 0.000000
2023-10-17 15:12:04,385 epoch 10 - iter 2600/2606 - loss 0.01015142 - time (sec): 270.82 - samples/sec: 1354.66 - lr: 0.000000 - momentum: 0.000000
2023-10-17 15:12:04,937 ----------------------------------------------------------------------------------------------------
2023-10-17 15:12:04,937 EPOCH 10 done: loss 0.0101 - lr: 0.000000
2023-10-17 15:12:17,266 DEV : loss 0.5094925761222839 - f1-score (micro avg) 0.3588
2023-10-17 15:12:17,894 ----------------------------------------------------------------------------------------------------
2023-10-17 15:12:17,896 Loading model from best epoch ...
2023-10-17 15:12:20,219 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 15:12:39,854
Results:
- F-score (micro) 0.4493
- F-score (macro) 0.3128
- Accuracy 0.2933
By class:
precision recall f1-score support
LOC 0.4879 0.5132 0.5002 1214
PER 0.4061 0.4814 0.4405 808
ORG 0.3013 0.3201 0.3104 353
HumanProd 0.0000 0.0000 0.0000 15
micro avg 0.4297 0.4707 0.4493 2390
macro avg 0.2988 0.3287 0.3128 2390
weighted avg 0.4296 0.4707 0.4489 2390
2023-10-17 15:12:39,854 ----------------------------------------------------------------------------------------------------
|