File size: 24,213 Bytes
dc89aad |
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 |
2023-10-17 10:18:58,581 ----------------------------------------------------------------------------------------------------
2023-10-17 10:18:58,582 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 10:18:58,583 ----------------------------------------------------------------------------------------------------
2023-10-17 10:18:58,583 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 10:18:58,583 ----------------------------------------------------------------------------------------------------
2023-10-17 10:18:58,583 Train: 20847 sentences
2023-10-17 10:18:58,583 (train_with_dev=False, train_with_test=False)
2023-10-17 10:18:58,583 ----------------------------------------------------------------------------------------------------
2023-10-17 10:18:58,583 Training Params:
2023-10-17 10:18:58,583 - learning_rate: "5e-05"
2023-10-17 10:18:58,583 - mini_batch_size: "8"
2023-10-17 10:18:58,583 - max_epochs: "10"
2023-10-17 10:18:58,584 - shuffle: "True"
2023-10-17 10:18:58,584 ----------------------------------------------------------------------------------------------------
2023-10-17 10:18:58,584 Plugins:
2023-10-17 10:18:58,584 - TensorboardLogger
2023-10-17 10:18:58,584 - LinearScheduler | warmup_fraction: '0.1'
2023-10-17 10:18:58,584 ----------------------------------------------------------------------------------------------------
2023-10-17 10:18:58,584 Final evaluation on model from best epoch (best-model.pt)
2023-10-17 10:18:58,584 - metric: "('micro avg', 'f1-score')"
2023-10-17 10:18:58,584 ----------------------------------------------------------------------------------------------------
2023-10-17 10:18:58,584 Computation:
2023-10-17 10:18:58,584 - compute on device: cuda:0
2023-10-17 10:18:58,584 - embedding storage: none
2023-10-17 10:18:58,584 ----------------------------------------------------------------------------------------------------
2023-10-17 10:18:58,584 Model training base path: "hmbench-newseye/de-hmteams/teams-base-historic-multilingual-discriminator-bs8-wsFalse-e10-lr5e-05-poolingfirst-layers-1-crfFalse-1"
2023-10-17 10:18:58,584 ----------------------------------------------------------------------------------------------------
2023-10-17 10:18:58,585 ----------------------------------------------------------------------------------------------------
2023-10-17 10:18:58,585 Logging anything other than scalars to TensorBoard is currently not supported.
2023-10-17 10:19:25,620 epoch 1 - iter 260/2606 - loss 1.88523681 - time (sec): 27.03 - samples/sec: 1244.18 - lr: 0.000005 - momentum: 0.000000
2023-10-17 10:19:52,497 epoch 1 - iter 520/2606 - loss 1.12382223 - time (sec): 53.91 - samples/sec: 1283.76 - lr: 0.000010 - momentum: 0.000000
2023-10-17 10:20:20,300 epoch 1 - iter 780/2606 - loss 0.83719157 - time (sec): 81.71 - samples/sec: 1310.34 - lr: 0.000015 - momentum: 0.000000
2023-10-17 10:20:47,239 epoch 1 - iter 1040/2606 - loss 0.68970036 - time (sec): 108.65 - samples/sec: 1336.82 - lr: 0.000020 - momentum: 0.000000
2023-10-17 10:21:14,600 epoch 1 - iter 1300/2606 - loss 0.59439782 - time (sec): 136.01 - samples/sec: 1353.21 - lr: 0.000025 - momentum: 0.000000
2023-10-17 10:21:42,308 epoch 1 - iter 1560/2606 - loss 0.52755960 - time (sec): 163.72 - samples/sec: 1362.27 - lr: 0.000030 - momentum: 0.000000
2023-10-17 10:22:10,369 epoch 1 - iter 1820/2606 - loss 0.48679454 - time (sec): 191.78 - samples/sec: 1346.47 - lr: 0.000035 - momentum: 0.000000
2023-10-17 10:22:37,175 epoch 1 - iter 2080/2606 - loss 0.45521738 - time (sec): 218.59 - samples/sec: 1338.76 - lr: 0.000040 - momentum: 0.000000
2023-10-17 10:23:04,173 epoch 1 - iter 2340/2606 - loss 0.42400062 - time (sec): 245.59 - samples/sec: 1345.15 - lr: 0.000045 - momentum: 0.000000
2023-10-17 10:23:31,264 epoch 1 - iter 2600/2606 - loss 0.40097181 - time (sec): 272.68 - samples/sec: 1344.06 - lr: 0.000050 - momentum: 0.000000
2023-10-17 10:23:31,864 ----------------------------------------------------------------------------------------------------
2023-10-17 10:23:31,864 EPOCH 1 done: loss 0.4002 - lr: 0.000050
2023-10-17 10:23:39,220 DEV : loss 0.11381553113460541 - f1-score (micro avg) 0.3171
2023-10-17 10:23:39,272 saving best model
2023-10-17 10:23:39,802 ----------------------------------------------------------------------------------------------------
2023-10-17 10:24:08,030 epoch 2 - iter 260/2606 - loss 0.17285748 - time (sec): 28.23 - samples/sec: 1355.61 - lr: 0.000049 - momentum: 0.000000
2023-10-17 10:24:35,777 epoch 2 - iter 520/2606 - loss 0.20526777 - time (sec): 55.97 - samples/sec: 1329.33 - lr: 0.000049 - momentum: 0.000000
2023-10-17 10:25:03,854 epoch 2 - iter 780/2606 - loss 0.19400558 - time (sec): 84.05 - samples/sec: 1333.96 - lr: 0.000048 - momentum: 0.000000
2023-10-17 10:25:30,896 epoch 2 - iter 1040/2606 - loss 0.19145026 - time (sec): 111.09 - samples/sec: 1322.86 - lr: 0.000048 - momentum: 0.000000
2023-10-17 10:25:59,200 epoch 2 - iter 1300/2606 - loss 0.18904313 - time (sec): 139.40 - samples/sec: 1313.04 - lr: 0.000047 - momentum: 0.000000
2023-10-17 10:26:25,856 epoch 2 - iter 1560/2606 - loss 0.18594503 - time (sec): 166.05 - samples/sec: 1319.31 - lr: 0.000047 - momentum: 0.000000
2023-10-17 10:26:54,154 epoch 2 - iter 1820/2606 - loss 0.17938094 - time (sec): 194.35 - samples/sec: 1331.79 - lr: 0.000046 - momentum: 0.000000
2023-10-17 10:27:21,812 epoch 2 - iter 2080/2606 - loss 0.17717797 - time (sec): 222.01 - samples/sec: 1330.89 - lr: 0.000046 - momentum: 0.000000
2023-10-17 10:27:47,173 epoch 2 - iter 2340/2606 - loss 0.17325202 - time (sec): 247.37 - samples/sec: 1329.74 - lr: 0.000045 - momentum: 0.000000
2023-10-17 10:28:13,350 epoch 2 - iter 2600/2606 - loss 0.16972430 - time (sec): 273.55 - samples/sec: 1339.96 - lr: 0.000044 - momentum: 0.000000
2023-10-17 10:28:14,078 ----------------------------------------------------------------------------------------------------
2023-10-17 10:28:14,078 EPOCH 2 done: loss 0.1695 - lr: 0.000044
2023-10-17 10:28:26,197 DEV : loss 0.223122239112854 - f1-score (micro avg) 0.3392
2023-10-17 10:28:26,258 saving best model
2023-10-17 10:28:27,707 ----------------------------------------------------------------------------------------------------
2023-10-17 10:28:56,388 epoch 3 - iter 260/2606 - loss 0.11196216 - time (sec): 28.68 - samples/sec: 1321.95 - lr: 0.000044 - momentum: 0.000000
2023-10-17 10:29:23,668 epoch 3 - iter 520/2606 - loss 0.11653677 - time (sec): 55.96 - samples/sec: 1336.99 - lr: 0.000043 - momentum: 0.000000
2023-10-17 10:29:50,270 epoch 3 - iter 780/2606 - loss 0.11963140 - time (sec): 82.56 - samples/sec: 1343.12 - lr: 0.000043 - momentum: 0.000000
2023-10-17 10:30:17,420 epoch 3 - iter 1040/2606 - loss 0.11700179 - time (sec): 109.71 - samples/sec: 1335.04 - lr: 0.000042 - momentum: 0.000000
2023-10-17 10:30:44,113 epoch 3 - iter 1300/2606 - loss 0.11677412 - time (sec): 136.40 - samples/sec: 1341.76 - lr: 0.000042 - momentum: 0.000000
2023-10-17 10:31:10,804 epoch 3 - iter 1560/2606 - loss 0.12086406 - time (sec): 163.09 - samples/sec: 1333.15 - lr: 0.000041 - momentum: 0.000000
2023-10-17 10:31:37,531 epoch 3 - iter 1820/2606 - loss 0.11878217 - time (sec): 189.82 - samples/sec: 1335.69 - lr: 0.000041 - momentum: 0.000000
2023-10-17 10:32:05,066 epoch 3 - iter 2080/2606 - loss 0.11924890 - time (sec): 217.36 - samples/sec: 1335.00 - lr: 0.000040 - momentum: 0.000000
2023-10-17 10:32:34,760 epoch 3 - iter 2340/2606 - loss 0.11857763 - time (sec): 247.05 - samples/sec: 1333.10 - lr: 0.000039 - momentum: 0.000000
2023-10-17 10:33:03,869 epoch 3 - iter 2600/2606 - loss 0.11654712 - time (sec): 276.16 - samples/sec: 1327.20 - lr: 0.000039 - momentum: 0.000000
2023-10-17 10:33:04,518 ----------------------------------------------------------------------------------------------------
2023-10-17 10:33:04,518 EPOCH 3 done: loss 0.1163 - lr: 0.000039
2023-10-17 10:33:16,361 DEV : loss 0.18563708662986755 - f1-score (micro avg) 0.356
2023-10-17 10:33:16,417 saving best model
2023-10-17 10:33:17,825 ----------------------------------------------------------------------------------------------------
2023-10-17 10:33:45,981 epoch 4 - iter 260/2606 - loss 0.08809572 - time (sec): 28.15 - samples/sec: 1322.94 - lr: 0.000038 - momentum: 0.000000
2023-10-17 10:34:12,346 epoch 4 - iter 520/2606 - loss 0.08738428 - time (sec): 54.52 - samples/sec: 1348.70 - lr: 0.000038 - momentum: 0.000000
2023-10-17 10:34:37,823 epoch 4 - iter 780/2606 - loss 0.08778704 - time (sec): 79.99 - samples/sec: 1355.87 - lr: 0.000037 - momentum: 0.000000
2023-10-17 10:35:04,756 epoch 4 - iter 1040/2606 - loss 0.08611474 - time (sec): 106.93 - samples/sec: 1346.45 - lr: 0.000037 - momentum: 0.000000
2023-10-17 10:35:30,915 epoch 4 - iter 1300/2606 - loss 0.08825421 - time (sec): 133.09 - samples/sec: 1343.11 - lr: 0.000036 - momentum: 0.000000
2023-10-17 10:35:56,677 epoch 4 - iter 1560/2606 - loss 0.08795449 - time (sec): 158.85 - samples/sec: 1342.10 - lr: 0.000036 - momentum: 0.000000
2023-10-17 10:36:24,361 epoch 4 - iter 1820/2606 - loss 0.08855088 - time (sec): 186.53 - samples/sec: 1348.57 - lr: 0.000035 - momentum: 0.000000
2023-10-17 10:36:51,682 epoch 4 - iter 2080/2606 - loss 0.08701037 - time (sec): 213.85 - samples/sec: 1356.64 - lr: 0.000034 - momentum: 0.000000
2023-10-17 10:37:19,341 epoch 4 - iter 2340/2606 - loss 0.08716051 - time (sec): 241.51 - samples/sec: 1361.96 - lr: 0.000034 - momentum: 0.000000
2023-10-17 10:37:47,256 epoch 4 - iter 2600/2606 - loss 0.08575731 - time (sec): 269.43 - samples/sec: 1361.07 - lr: 0.000033 - momentum: 0.000000
2023-10-17 10:37:47,818 ----------------------------------------------------------------------------------------------------
2023-10-17 10:37:47,818 EPOCH 4 done: loss 0.0856 - lr: 0.000033
2023-10-17 10:37:58,718 DEV : loss 0.2579371929168701 - f1-score (micro avg) 0.3733
2023-10-17 10:37:58,776 saving best model
2023-10-17 10:38:00,188 ----------------------------------------------------------------------------------------------------
2023-10-17 10:38:27,739 epoch 5 - iter 260/2606 - loss 0.05100938 - time (sec): 27.55 - samples/sec: 1367.58 - lr: 0.000033 - momentum: 0.000000
2023-10-17 10:38:53,642 epoch 5 - iter 520/2606 - loss 0.04902557 - time (sec): 53.45 - samples/sec: 1337.47 - lr: 0.000032 - momentum: 0.000000
2023-10-17 10:39:20,996 epoch 5 - iter 780/2606 - loss 0.05197574 - time (sec): 80.80 - samples/sec: 1340.12 - lr: 0.000032 - momentum: 0.000000
2023-10-17 10:39:47,084 epoch 5 - iter 1040/2606 - loss 0.05396406 - time (sec): 106.89 - samples/sec: 1332.05 - lr: 0.000031 - momentum: 0.000000
2023-10-17 10:40:16,075 epoch 5 - iter 1300/2606 - loss 0.05874220 - time (sec): 135.88 - samples/sec: 1339.43 - lr: 0.000031 - momentum: 0.000000
2023-10-17 10:40:43,125 epoch 5 - iter 1560/2606 - loss 0.05952086 - time (sec): 162.93 - samples/sec: 1361.63 - lr: 0.000030 - momentum: 0.000000
2023-10-17 10:41:09,653 epoch 5 - iter 1820/2606 - loss 0.06088136 - time (sec): 189.46 - samples/sec: 1363.89 - lr: 0.000029 - momentum: 0.000000
2023-10-17 10:41:36,811 epoch 5 - iter 2080/2606 - loss 0.06080196 - time (sec): 216.62 - samples/sec: 1365.76 - lr: 0.000029 - momentum: 0.000000
2023-10-17 10:42:02,485 epoch 5 - iter 2340/2606 - loss 0.06099854 - time (sec): 242.29 - samples/sec: 1363.26 - lr: 0.000028 - momentum: 0.000000
2023-10-17 10:42:29,443 epoch 5 - iter 2600/2606 - loss 0.06045113 - time (sec): 269.25 - samples/sec: 1362.05 - lr: 0.000028 - momentum: 0.000000
2023-10-17 10:42:29,970 ----------------------------------------------------------------------------------------------------
2023-10-17 10:42:29,970 EPOCH 5 done: loss 0.0604 - lr: 0.000028
2023-10-17 10:42:40,709 DEV : loss 0.3068985044956207 - f1-score (micro avg) 0.4156
2023-10-17 10:42:40,761 saving best model
2023-10-17 10:42:42,155 ----------------------------------------------------------------------------------------------------
2023-10-17 10:43:08,497 epoch 6 - iter 260/2606 - loss 0.04273978 - time (sec): 26.34 - samples/sec: 1415.83 - lr: 0.000027 - momentum: 0.000000
2023-10-17 10:43:34,516 epoch 6 - iter 520/2606 - loss 0.04270582 - time (sec): 52.36 - samples/sec: 1379.74 - lr: 0.000027 - momentum: 0.000000
2023-10-17 10:44:00,440 epoch 6 - iter 780/2606 - loss 0.04178911 - time (sec): 78.28 - samples/sec: 1368.55 - lr: 0.000026 - momentum: 0.000000
2023-10-17 10:44:27,949 epoch 6 - iter 1040/2606 - loss 0.03962571 - time (sec): 105.79 - samples/sec: 1385.07 - lr: 0.000026 - momentum: 0.000000
2023-10-17 10:44:55,668 epoch 6 - iter 1300/2606 - loss 0.04045489 - time (sec): 133.51 - samples/sec: 1387.29 - lr: 0.000025 - momentum: 0.000000
2023-10-17 10:45:22,831 epoch 6 - iter 1560/2606 - loss 0.04097251 - time (sec): 160.67 - samples/sec: 1383.91 - lr: 0.000024 - momentum: 0.000000
2023-10-17 10:45:50,498 epoch 6 - iter 1820/2606 - loss 0.04012529 - time (sec): 188.34 - samples/sec: 1371.39 - lr: 0.000024 - momentum: 0.000000
2023-10-17 10:46:17,568 epoch 6 - iter 2080/2606 - loss 0.04115226 - time (sec): 215.41 - samples/sec: 1363.74 - lr: 0.000023 - momentum: 0.000000
2023-10-17 10:46:46,617 epoch 6 - iter 2340/2606 - loss 0.04163284 - time (sec): 244.46 - samples/sec: 1352.33 - lr: 0.000023 - momentum: 0.000000
2023-10-17 10:47:13,563 epoch 6 - iter 2600/2606 - loss 0.04297284 - time (sec): 271.40 - samples/sec: 1350.77 - lr: 0.000022 - momentum: 0.000000
2023-10-17 10:47:14,120 ----------------------------------------------------------------------------------------------------
2023-10-17 10:47:14,120 EPOCH 6 done: loss 0.0429 - lr: 0.000022
2023-10-17 10:47:24,944 DEV : loss 0.3329330384731293 - f1-score (micro avg) 0.3972
2023-10-17 10:47:24,995 ----------------------------------------------------------------------------------------------------
2023-10-17 10:47:52,020 epoch 7 - iter 260/2606 - loss 0.03145306 - time (sec): 27.02 - samples/sec: 1399.79 - lr: 0.000022 - momentum: 0.000000
2023-10-17 10:48:19,096 epoch 7 - iter 520/2606 - loss 0.02828661 - time (sec): 54.10 - samples/sec: 1391.68 - lr: 0.000021 - momentum: 0.000000
2023-10-17 10:48:46,758 epoch 7 - iter 780/2606 - loss 0.02853558 - time (sec): 81.76 - samples/sec: 1374.54 - lr: 0.000021 - momentum: 0.000000
2023-10-17 10:49:16,621 epoch 7 - iter 1040/2606 - loss 0.02992094 - time (sec): 111.62 - samples/sec: 1335.70 - lr: 0.000020 - momentum: 0.000000
2023-10-17 10:49:44,050 epoch 7 - iter 1300/2606 - loss 0.03061991 - time (sec): 139.05 - samples/sec: 1327.08 - lr: 0.000019 - momentum: 0.000000
2023-10-17 10:50:10,681 epoch 7 - iter 1560/2606 - loss 0.02982746 - time (sec): 165.68 - samples/sec: 1326.21 - lr: 0.000019 - momentum: 0.000000
2023-10-17 10:50:37,937 epoch 7 - iter 1820/2606 - loss 0.03066087 - time (sec): 192.94 - samples/sec: 1325.75 - lr: 0.000018 - momentum: 0.000000
2023-10-17 10:51:06,450 epoch 7 - iter 2080/2606 - loss 0.03008073 - time (sec): 221.45 - samples/sec: 1332.50 - lr: 0.000018 - momentum: 0.000000
2023-10-17 10:51:34,169 epoch 7 - iter 2340/2606 - loss 0.03142644 - time (sec): 249.17 - samples/sec: 1331.26 - lr: 0.000017 - momentum: 0.000000
2023-10-17 10:52:00,747 epoch 7 - iter 2600/2606 - loss 0.03141193 - time (sec): 275.75 - samples/sec: 1330.53 - lr: 0.000017 - momentum: 0.000000
2023-10-17 10:52:01,309 ----------------------------------------------------------------------------------------------------
2023-10-17 10:52:01,309 EPOCH 7 done: loss 0.0314 - lr: 0.000017
2023-10-17 10:52:12,353 DEV : loss 0.363223135471344 - f1-score (micro avg) 0.4096
2023-10-17 10:52:12,417 ----------------------------------------------------------------------------------------------------
2023-10-17 10:52:39,557 epoch 8 - iter 260/2606 - loss 0.01939361 - time (sec): 27.14 - samples/sec: 1321.47 - lr: 0.000016 - momentum: 0.000000
2023-10-17 10:53:07,789 epoch 8 - iter 520/2606 - loss 0.01739651 - time (sec): 55.37 - samples/sec: 1290.53 - lr: 0.000016 - momentum: 0.000000
2023-10-17 10:53:36,222 epoch 8 - iter 780/2606 - loss 0.01873777 - time (sec): 83.80 - samples/sec: 1267.00 - lr: 0.000015 - momentum: 0.000000
2023-10-17 10:54:05,116 epoch 8 - iter 1040/2606 - loss 0.01864960 - time (sec): 112.70 - samples/sec: 1263.19 - lr: 0.000014 - momentum: 0.000000
2023-10-17 10:54:32,882 epoch 8 - iter 1300/2606 - loss 0.01961418 - time (sec): 140.46 - samples/sec: 1267.02 - lr: 0.000014 - momentum: 0.000000
2023-10-17 10:55:00,732 epoch 8 - iter 1560/2606 - loss 0.01972925 - time (sec): 168.31 - samples/sec: 1280.96 - lr: 0.000013 - momentum: 0.000000
2023-10-17 10:55:28,259 epoch 8 - iter 1820/2606 - loss 0.01993508 - time (sec): 195.84 - samples/sec: 1291.10 - lr: 0.000013 - momentum: 0.000000
2023-10-17 10:55:55,992 epoch 8 - iter 2080/2606 - loss 0.02035476 - time (sec): 223.57 - samples/sec: 1305.07 - lr: 0.000012 - momentum: 0.000000
2023-10-17 10:56:23,834 epoch 8 - iter 2340/2606 - loss 0.02030965 - time (sec): 251.41 - samples/sec: 1312.25 - lr: 0.000012 - momentum: 0.000000
2023-10-17 10:56:52,233 epoch 8 - iter 2600/2606 - loss 0.02034546 - time (sec): 279.81 - samples/sec: 1310.76 - lr: 0.000011 - momentum: 0.000000
2023-10-17 10:56:52,873 ----------------------------------------------------------------------------------------------------
2023-10-17 10:56:52,874 EPOCH 8 done: loss 0.0203 - lr: 0.000011
2023-10-17 10:57:05,587 DEV : loss 0.393916517496109 - f1-score (micro avg) 0.406
2023-10-17 10:57:05,657 ----------------------------------------------------------------------------------------------------
2023-10-17 10:57:32,201 epoch 9 - iter 260/2606 - loss 0.00999242 - time (sec): 26.54 - samples/sec: 1281.27 - lr: 0.000011 - momentum: 0.000000
2023-10-17 10:57:59,545 epoch 9 - iter 520/2606 - loss 0.01555837 - time (sec): 53.89 - samples/sec: 1322.90 - lr: 0.000010 - momentum: 0.000000
2023-10-17 10:58:26,352 epoch 9 - iter 780/2606 - loss 0.01445469 - time (sec): 80.69 - samples/sec: 1300.78 - lr: 0.000009 - momentum: 0.000000
2023-10-17 10:58:54,326 epoch 9 - iter 1040/2606 - loss 0.01476011 - time (sec): 108.67 - samples/sec: 1284.97 - lr: 0.000009 - momentum: 0.000000
2023-10-17 10:59:21,870 epoch 9 - iter 1300/2606 - loss 0.01578290 - time (sec): 136.21 - samples/sec: 1293.88 - lr: 0.000008 - momentum: 0.000000
2023-10-17 10:59:50,972 epoch 9 - iter 1560/2606 - loss 0.01579451 - time (sec): 165.31 - samples/sec: 1287.08 - lr: 0.000008 - momentum: 0.000000
2023-10-17 11:00:20,149 epoch 9 - iter 1820/2606 - loss 0.01527993 - time (sec): 194.49 - samples/sec: 1286.46 - lr: 0.000007 - momentum: 0.000000
2023-10-17 11:00:49,010 epoch 9 - iter 2080/2606 - loss 0.01478612 - time (sec): 223.35 - samples/sec: 1291.50 - lr: 0.000007 - momentum: 0.000000
2023-10-17 11:01:17,393 epoch 9 - iter 2340/2606 - loss 0.01475537 - time (sec): 251.73 - samples/sec: 1297.41 - lr: 0.000006 - momentum: 0.000000
2023-10-17 11:01:46,450 epoch 9 - iter 2600/2606 - loss 0.01458389 - time (sec): 280.79 - samples/sec: 1305.91 - lr: 0.000006 - momentum: 0.000000
2023-10-17 11:01:47,114 ----------------------------------------------------------------------------------------------------
2023-10-17 11:01:47,114 EPOCH 9 done: loss 0.0147 - lr: 0.000006
2023-10-17 11:02:00,212 DEV : loss 0.5024428367614746 - f1-score (micro avg) 0.3891
2023-10-17 11:02:00,276 ----------------------------------------------------------------------------------------------------
2023-10-17 11:02:28,755 epoch 10 - iter 260/2606 - loss 0.00777663 - time (sec): 28.48 - samples/sec: 1304.69 - lr: 0.000005 - momentum: 0.000000
2023-10-17 11:02:56,881 epoch 10 - iter 520/2606 - loss 0.00852951 - time (sec): 56.60 - samples/sec: 1283.24 - lr: 0.000004 - momentum: 0.000000
2023-10-17 11:03:24,114 epoch 10 - iter 780/2606 - loss 0.00914261 - time (sec): 83.84 - samples/sec: 1274.15 - lr: 0.000004 - momentum: 0.000000
2023-10-17 11:03:50,849 epoch 10 - iter 1040/2606 - loss 0.00961380 - time (sec): 110.57 - samples/sec: 1302.75 - lr: 0.000003 - momentum: 0.000000
2023-10-17 11:04:20,070 epoch 10 - iter 1300/2606 - loss 0.00995380 - time (sec): 139.79 - samples/sec: 1298.95 - lr: 0.000003 - momentum: 0.000000
2023-10-17 11:04:47,597 epoch 10 - iter 1560/2606 - loss 0.00946199 - time (sec): 167.32 - samples/sec: 1295.54 - lr: 0.000002 - momentum: 0.000000
2023-10-17 11:05:14,358 epoch 10 - iter 1820/2606 - loss 0.00930366 - time (sec): 194.08 - samples/sec: 1303.79 - lr: 0.000002 - momentum: 0.000000
2023-10-17 11:05:41,338 epoch 10 - iter 2080/2606 - loss 0.00918369 - time (sec): 221.06 - samples/sec: 1310.36 - lr: 0.000001 - momentum: 0.000000
2023-10-17 11:06:09,838 epoch 10 - iter 2340/2606 - loss 0.00912808 - time (sec): 249.56 - samples/sec: 1315.21 - lr: 0.000001 - momentum: 0.000000
2023-10-17 11:06:39,144 epoch 10 - iter 2600/2606 - loss 0.00928698 - time (sec): 278.87 - samples/sec: 1315.25 - lr: 0.000000 - momentum: 0.000000
2023-10-17 11:06:39,685 ----------------------------------------------------------------------------------------------------
2023-10-17 11:06:39,685 EPOCH 10 done: loss 0.0093 - lr: 0.000000
2023-10-17 11:06:51,588 DEV : loss 0.5263164043426514 - f1-score (micro avg) 0.3942
2023-10-17 11:06:52,178 ----------------------------------------------------------------------------------------------------
2023-10-17 11:06:52,180 Loading model from best epoch ...
2023-10-17 11:06:54,538 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 11:07:15,105
Results:
- F-score (micro) 0.4845
- F-score (macro) 0.3222
- Accuracy 0.3241
By class:
precision recall f1-score support
LOC 0.5629 0.5972 0.5795 1214
PER 0.4140 0.4406 0.4269 808
ORG 0.3077 0.2606 0.2822 353
HumanProd 0.0000 0.0000 0.0000 15
micro avg 0.4784 0.4908 0.4845 2390
macro avg 0.3211 0.3246 0.3222 2390
weighted avg 0.4713 0.4908 0.4804 2390
2023-10-17 11:07:15,105 ----------------------------------------------------------------------------------------------------
|