File size: 26,667 Bytes
9d92914 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 |
2024-03-26 16:20:27,818 ----------------------------------------------------------------------------------------------------
2024-03-26 16:20:27,818 Model: "SequenceTagger(
(embeddings): TransformerWordEmbeddings(
(model): BertModel(
(embeddings): BertEmbeddings(
(word_embeddings): Embedding(31103, 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=17, bias=True)
(loss_function): CrossEntropyLoss()
)"
2024-03-26 16:20:27,818 ----------------------------------------------------------------------------------------------------
2024-03-26 16:20:27,819 Corpus: 758 train + 94 dev + 96 test sentences
2024-03-26 16:20:27,819 ----------------------------------------------------------------------------------------------------
2024-03-26 16:20:27,819 Train: 758 sentences
2024-03-26 16:20:27,819 (train_with_dev=False, train_with_test=False)
2024-03-26 16:20:27,819 ----------------------------------------------------------------------------------------------------
2024-03-26 16:20:27,819 Training Params:
2024-03-26 16:20:27,819 - learning_rate: "3e-05"
2024-03-26 16:20:27,819 - mini_batch_size: "16"
2024-03-26 16:20:27,819 - max_epochs: "10"
2024-03-26 16:20:27,819 - shuffle: "True"
2024-03-26 16:20:27,819 ----------------------------------------------------------------------------------------------------
2024-03-26 16:20:27,819 Plugins:
2024-03-26 16:20:27,819 - TensorboardLogger
2024-03-26 16:20:27,819 - LinearScheduler | warmup_fraction: '0.1'
2024-03-26 16:20:27,819 ----------------------------------------------------------------------------------------------------
2024-03-26 16:20:27,819 Final evaluation on model from best epoch (best-model.pt)
2024-03-26 16:20:27,819 - metric: "('micro avg', 'f1-score')"
2024-03-26 16:20:27,819 ----------------------------------------------------------------------------------------------------
2024-03-26 16:20:27,819 Computation:
2024-03-26 16:20:27,819 - compute on device: cuda:0
2024-03-26 16:20:27,819 - embedding storage: none
2024-03-26 16:20:27,819 ----------------------------------------------------------------------------------------------------
2024-03-26 16:20:27,819 Model training base path: "flair-co-funer-german_dbmdz_bert_base-bs16-e10-lr3e-05-5"
2024-03-26 16:20:27,819 ----------------------------------------------------------------------------------------------------
2024-03-26 16:20:27,819 ----------------------------------------------------------------------------------------------------
2024-03-26 16:20:27,819 Logging anything other than scalars to TensorBoard is currently not supported.
2024-03-26 16:20:29,303 epoch 1 - iter 4/48 - loss 3.33243891 - time (sec): 1.48 - samples/sec: 1765.97 - lr: 0.000002 - momentum: 0.000000
2024-03-26 16:20:31,987 epoch 1 - iter 8/48 - loss 3.27444886 - time (sec): 4.17 - samples/sec: 1460.39 - lr: 0.000004 - momentum: 0.000000
2024-03-26 16:20:33,821 epoch 1 - iter 12/48 - loss 3.19284867 - time (sec): 6.00 - samples/sec: 1483.98 - lr: 0.000007 - momentum: 0.000000
2024-03-26 16:20:35,385 epoch 1 - iter 16/48 - loss 3.06760041 - time (sec): 7.57 - samples/sec: 1590.22 - lr: 0.000009 - momentum: 0.000000
2024-03-26 16:20:37,530 epoch 1 - iter 20/48 - loss 2.93913875 - time (sec): 9.71 - samples/sec: 1552.75 - lr: 0.000012 - momentum: 0.000000
2024-03-26 16:20:40,287 epoch 1 - iter 24/48 - loss 2.77773408 - time (sec): 12.47 - samples/sec: 1479.70 - lr: 0.000014 - momentum: 0.000000
2024-03-26 16:20:41,904 epoch 1 - iter 28/48 - loss 2.66904977 - time (sec): 14.08 - samples/sec: 1490.07 - lr: 0.000017 - momentum: 0.000000
2024-03-26 16:20:43,986 epoch 1 - iter 32/48 - loss 2.55099320 - time (sec): 16.17 - samples/sec: 1486.98 - lr: 0.000019 - momentum: 0.000000
2024-03-26 16:20:45,522 epoch 1 - iter 36/48 - loss 2.46407907 - time (sec): 17.70 - samples/sec: 1508.20 - lr: 0.000022 - momentum: 0.000000
2024-03-26 16:20:48,292 epoch 1 - iter 40/48 - loss 2.35518313 - time (sec): 20.47 - samples/sec: 1460.67 - lr: 0.000024 - momentum: 0.000000
2024-03-26 16:20:49,477 epoch 1 - iter 44/48 - loss 2.27229820 - time (sec): 21.66 - samples/sec: 1484.13 - lr: 0.000027 - momentum: 0.000000
2024-03-26 16:20:51,300 epoch 1 - iter 48/48 - loss 2.20094714 - time (sec): 23.48 - samples/sec: 1468.08 - lr: 0.000029 - momentum: 0.000000
2024-03-26 16:20:51,300 ----------------------------------------------------------------------------------------------------
2024-03-26 16:20:51,301 EPOCH 1 done: loss 2.2009 - lr: 0.000029
2024-03-26 16:20:52,115 DEV : loss 0.8398668169975281 - f1-score (micro avg) 0.4601
2024-03-26 16:20:52,116 saving best model
2024-03-26 16:20:52,380 ----------------------------------------------------------------------------------------------------
2024-03-26 16:20:55,073 epoch 2 - iter 4/48 - loss 0.97764540 - time (sec): 2.69 - samples/sec: 1282.27 - lr: 0.000030 - momentum: 0.000000
2024-03-26 16:20:56,909 epoch 2 - iter 8/48 - loss 0.92169240 - time (sec): 4.53 - samples/sec: 1353.08 - lr: 0.000030 - momentum: 0.000000
2024-03-26 16:20:58,797 epoch 2 - iter 12/48 - loss 0.87356675 - time (sec): 6.42 - samples/sec: 1390.54 - lr: 0.000029 - momentum: 0.000000
2024-03-26 16:21:01,443 epoch 2 - iter 16/48 - loss 0.79202489 - time (sec): 9.06 - samples/sec: 1396.31 - lr: 0.000029 - momentum: 0.000000
2024-03-26 16:21:02,817 epoch 2 - iter 20/48 - loss 0.76156184 - time (sec): 10.44 - samples/sec: 1435.48 - lr: 0.000029 - momentum: 0.000000
2024-03-26 16:21:05,616 epoch 2 - iter 24/48 - loss 0.71680551 - time (sec): 13.24 - samples/sec: 1352.27 - lr: 0.000028 - momentum: 0.000000
2024-03-26 16:21:07,217 epoch 2 - iter 28/48 - loss 0.70169294 - time (sec): 14.84 - samples/sec: 1385.86 - lr: 0.000028 - momentum: 0.000000
2024-03-26 16:21:09,224 epoch 2 - iter 32/48 - loss 0.66814375 - time (sec): 16.84 - samples/sec: 1378.44 - lr: 0.000028 - momentum: 0.000000
2024-03-26 16:21:10,971 epoch 2 - iter 36/48 - loss 0.64979725 - time (sec): 18.59 - samples/sec: 1409.85 - lr: 0.000028 - momentum: 0.000000
2024-03-26 16:21:13,315 epoch 2 - iter 40/48 - loss 0.63823796 - time (sec): 20.94 - samples/sec: 1398.22 - lr: 0.000027 - momentum: 0.000000
2024-03-26 16:21:15,494 epoch 2 - iter 44/48 - loss 0.61211544 - time (sec): 23.11 - samples/sec: 1403.26 - lr: 0.000027 - momentum: 0.000000
2024-03-26 16:21:16,745 epoch 2 - iter 48/48 - loss 0.60381110 - time (sec): 24.37 - samples/sec: 1414.81 - lr: 0.000027 - momentum: 0.000000
2024-03-26 16:21:16,746 ----------------------------------------------------------------------------------------------------
2024-03-26 16:21:16,746 EPOCH 2 done: loss 0.6038 - lr: 0.000027
2024-03-26 16:21:17,672 DEV : loss 0.35631129145622253 - f1-score (micro avg) 0.7369
2024-03-26 16:21:17,673 saving best model
2024-03-26 16:21:18,130 ----------------------------------------------------------------------------------------------------
2024-03-26 16:21:19,212 epoch 3 - iter 4/48 - loss 0.42050435 - time (sec): 1.08 - samples/sec: 2063.24 - lr: 0.000026 - momentum: 0.000000
2024-03-26 16:21:21,086 epoch 3 - iter 8/48 - loss 0.36994000 - time (sec): 2.96 - samples/sec: 1667.12 - lr: 0.000026 - momentum: 0.000000
2024-03-26 16:21:23,303 epoch 3 - iter 12/48 - loss 0.33936035 - time (sec): 5.17 - samples/sec: 1657.72 - lr: 0.000026 - momentum: 0.000000
2024-03-26 16:21:25,257 epoch 3 - iter 16/48 - loss 0.33771789 - time (sec): 7.13 - samples/sec: 1596.10 - lr: 0.000026 - momentum: 0.000000
2024-03-26 16:21:27,102 epoch 3 - iter 20/48 - loss 0.33282889 - time (sec): 8.97 - samples/sec: 1580.23 - lr: 0.000025 - momentum: 0.000000
2024-03-26 16:21:29,041 epoch 3 - iter 24/48 - loss 0.31782340 - time (sec): 10.91 - samples/sec: 1536.45 - lr: 0.000025 - momentum: 0.000000
2024-03-26 16:21:32,217 epoch 3 - iter 28/48 - loss 0.30362057 - time (sec): 14.09 - samples/sec: 1420.98 - lr: 0.000025 - momentum: 0.000000
2024-03-26 16:21:33,721 epoch 3 - iter 32/48 - loss 0.30431378 - time (sec): 15.59 - samples/sec: 1444.75 - lr: 0.000025 - momentum: 0.000000
2024-03-26 16:21:37,002 epoch 3 - iter 36/48 - loss 0.29035892 - time (sec): 18.87 - samples/sec: 1374.51 - lr: 0.000024 - momentum: 0.000000
2024-03-26 16:21:39,388 epoch 3 - iter 40/48 - loss 0.28751496 - time (sec): 21.26 - samples/sec: 1376.57 - lr: 0.000024 - momentum: 0.000000
2024-03-26 16:21:41,517 epoch 3 - iter 44/48 - loss 0.28096716 - time (sec): 23.39 - samples/sec: 1372.11 - lr: 0.000024 - momentum: 0.000000
2024-03-26 16:21:43,103 epoch 3 - iter 48/48 - loss 0.27892438 - time (sec): 24.97 - samples/sec: 1380.42 - lr: 0.000023 - momentum: 0.000000
2024-03-26 16:21:43,103 ----------------------------------------------------------------------------------------------------
2024-03-26 16:21:43,103 EPOCH 3 done: loss 0.2789 - lr: 0.000023
2024-03-26 16:21:44,031 DEV : loss 0.23076379299163818 - f1-score (micro avg) 0.8511
2024-03-26 16:21:44,033 saving best model
2024-03-26 16:21:44,473 ----------------------------------------------------------------------------------------------------
2024-03-26 16:21:47,385 epoch 4 - iter 4/48 - loss 0.14723859 - time (sec): 2.91 - samples/sec: 1280.87 - lr: 0.000023 - momentum: 0.000000
2024-03-26 16:21:48,880 epoch 4 - iter 8/48 - loss 0.19096714 - time (sec): 4.41 - samples/sec: 1411.21 - lr: 0.000023 - momentum: 0.000000
2024-03-26 16:21:51,392 epoch 4 - iter 12/48 - loss 0.17787461 - time (sec): 6.92 - samples/sec: 1344.27 - lr: 0.000023 - momentum: 0.000000
2024-03-26 16:21:54,042 epoch 4 - iter 16/48 - loss 0.17213388 - time (sec): 9.57 - samples/sec: 1326.82 - lr: 0.000022 - momentum: 0.000000
2024-03-26 16:21:56,290 epoch 4 - iter 20/48 - loss 0.16683254 - time (sec): 11.82 - samples/sec: 1335.55 - lr: 0.000022 - momentum: 0.000000
2024-03-26 16:21:57,785 epoch 4 - iter 24/48 - loss 0.16783017 - time (sec): 13.31 - samples/sec: 1369.58 - lr: 0.000022 - momentum: 0.000000
2024-03-26 16:22:00,142 epoch 4 - iter 28/48 - loss 0.17152331 - time (sec): 15.67 - samples/sec: 1355.74 - lr: 0.000022 - momentum: 0.000000
2024-03-26 16:22:03,081 epoch 4 - iter 32/48 - loss 0.17071354 - time (sec): 18.61 - samples/sec: 1346.75 - lr: 0.000021 - momentum: 0.000000
2024-03-26 16:22:04,712 epoch 4 - iter 36/48 - loss 0.17256077 - time (sec): 20.24 - samples/sec: 1371.45 - lr: 0.000021 - momentum: 0.000000
2024-03-26 16:22:05,702 epoch 4 - iter 40/48 - loss 0.17602230 - time (sec): 21.23 - samples/sec: 1414.58 - lr: 0.000021 - momentum: 0.000000
2024-03-26 16:22:07,134 epoch 4 - iter 44/48 - loss 0.17600155 - time (sec): 22.66 - samples/sec: 1435.65 - lr: 0.000020 - momentum: 0.000000
2024-03-26 16:22:08,005 epoch 4 - iter 48/48 - loss 0.17965021 - time (sec): 23.53 - samples/sec: 1464.92 - lr: 0.000020 - momentum: 0.000000
2024-03-26 16:22:08,006 ----------------------------------------------------------------------------------------------------
2024-03-26 16:22:08,006 EPOCH 4 done: loss 0.1797 - lr: 0.000020
2024-03-26 16:22:08,929 DEV : loss 0.19128400087356567 - f1-score (micro avg) 0.8865
2024-03-26 16:22:08,930 saving best model
2024-03-26 16:22:09,357 ----------------------------------------------------------------------------------------------------
2024-03-26 16:22:11,177 epoch 5 - iter 4/48 - loss 0.14799718 - time (sec): 1.82 - samples/sec: 1578.71 - lr: 0.000020 - momentum: 0.000000
2024-03-26 16:22:13,039 epoch 5 - iter 8/48 - loss 0.13649489 - time (sec): 3.68 - samples/sec: 1685.38 - lr: 0.000020 - momentum: 0.000000
2024-03-26 16:22:16,132 epoch 5 - iter 12/48 - loss 0.12703718 - time (sec): 6.77 - samples/sec: 1420.24 - lr: 0.000019 - momentum: 0.000000
2024-03-26 16:22:17,451 epoch 5 - iter 16/48 - loss 0.12204170 - time (sec): 8.09 - samples/sec: 1469.19 - lr: 0.000019 - momentum: 0.000000
2024-03-26 16:22:19,722 epoch 5 - iter 20/48 - loss 0.13938063 - time (sec): 10.37 - samples/sec: 1453.83 - lr: 0.000019 - momentum: 0.000000
2024-03-26 16:22:21,860 epoch 5 - iter 24/48 - loss 0.13765159 - time (sec): 12.50 - samples/sec: 1421.75 - lr: 0.000018 - momentum: 0.000000
2024-03-26 16:22:23,223 epoch 5 - iter 28/48 - loss 0.14191204 - time (sec): 13.87 - samples/sec: 1463.46 - lr: 0.000018 - momentum: 0.000000
2024-03-26 16:22:24,594 epoch 5 - iter 32/48 - loss 0.14166725 - time (sec): 15.24 - samples/sec: 1496.23 - lr: 0.000018 - momentum: 0.000000
2024-03-26 16:22:26,716 epoch 5 - iter 36/48 - loss 0.14084111 - time (sec): 17.36 - samples/sec: 1487.67 - lr: 0.000018 - momentum: 0.000000
2024-03-26 16:22:28,542 epoch 5 - iter 40/48 - loss 0.13988440 - time (sec): 19.18 - samples/sec: 1486.03 - lr: 0.000017 - momentum: 0.000000
2024-03-26 16:22:30,541 epoch 5 - iter 44/48 - loss 0.13722618 - time (sec): 21.18 - samples/sec: 1498.59 - lr: 0.000017 - momentum: 0.000000
2024-03-26 16:22:32,649 epoch 5 - iter 48/48 - loss 0.13324400 - time (sec): 23.29 - samples/sec: 1480.01 - lr: 0.000017 - momentum: 0.000000
2024-03-26 16:22:32,649 ----------------------------------------------------------------------------------------------------
2024-03-26 16:22:32,650 EPOCH 5 done: loss 0.1332 - lr: 0.000017
2024-03-26 16:22:33,569 DEV : loss 0.18336626887321472 - f1-score (micro avg) 0.8929
2024-03-26 16:22:33,571 saving best model
2024-03-26 16:22:34,022 ----------------------------------------------------------------------------------------------------
2024-03-26 16:22:35,910 epoch 6 - iter 4/48 - loss 0.10824307 - time (sec): 1.89 - samples/sec: 1454.79 - lr: 0.000017 - momentum: 0.000000
2024-03-26 16:22:38,630 epoch 6 - iter 8/48 - loss 0.11635535 - time (sec): 4.61 - samples/sec: 1379.60 - lr: 0.000016 - momentum: 0.000000
2024-03-26 16:22:40,511 epoch 6 - iter 12/48 - loss 0.11860529 - time (sec): 6.49 - samples/sec: 1391.86 - lr: 0.000016 - momentum: 0.000000
2024-03-26 16:22:41,984 epoch 6 - iter 16/48 - loss 0.12373574 - time (sec): 7.96 - samples/sec: 1453.46 - lr: 0.000016 - momentum: 0.000000
2024-03-26 16:22:44,659 epoch 6 - iter 20/48 - loss 0.11554259 - time (sec): 10.64 - samples/sec: 1370.95 - lr: 0.000015 - momentum: 0.000000
2024-03-26 16:22:47,307 epoch 6 - iter 24/48 - loss 0.10657385 - time (sec): 13.28 - samples/sec: 1344.95 - lr: 0.000015 - momentum: 0.000000
2024-03-26 16:22:49,757 epoch 6 - iter 28/48 - loss 0.10502829 - time (sec): 15.73 - samples/sec: 1318.85 - lr: 0.000015 - momentum: 0.000000
2024-03-26 16:22:51,129 epoch 6 - iter 32/48 - loss 0.11414547 - time (sec): 17.11 - samples/sec: 1362.93 - lr: 0.000015 - momentum: 0.000000
2024-03-26 16:22:52,981 epoch 6 - iter 36/48 - loss 0.10975802 - time (sec): 18.96 - samples/sec: 1373.68 - lr: 0.000014 - momentum: 0.000000
2024-03-26 16:22:53,959 epoch 6 - iter 40/48 - loss 0.11001287 - time (sec): 19.94 - samples/sec: 1415.04 - lr: 0.000014 - momentum: 0.000000
2024-03-26 16:22:56,456 epoch 6 - iter 44/48 - loss 0.10762077 - time (sec): 22.43 - samples/sec: 1387.00 - lr: 0.000014 - momentum: 0.000000
2024-03-26 16:22:59,205 epoch 6 - iter 48/48 - loss 0.10325396 - time (sec): 25.18 - samples/sec: 1368.93 - lr: 0.000014 - momentum: 0.000000
2024-03-26 16:22:59,205 ----------------------------------------------------------------------------------------------------
2024-03-26 16:22:59,205 EPOCH 6 done: loss 0.1033 - lr: 0.000014
2024-03-26 16:23:00,104 DEV : loss 0.186451718211174 - f1-score (micro avg) 0.8952
2024-03-26 16:23:00,105 saving best model
2024-03-26 16:23:00,568 ----------------------------------------------------------------------------------------------------
2024-03-26 16:23:02,694 epoch 7 - iter 4/48 - loss 0.05956705 - time (sec): 2.12 - samples/sec: 1369.90 - lr: 0.000013 - momentum: 0.000000
2024-03-26 16:23:04,375 epoch 7 - iter 8/48 - loss 0.06304602 - time (sec): 3.80 - samples/sec: 1400.81 - lr: 0.000013 - momentum: 0.000000
2024-03-26 16:23:05,781 epoch 7 - iter 12/48 - loss 0.09022166 - time (sec): 5.21 - samples/sec: 1456.53 - lr: 0.000013 - momentum: 0.000000
2024-03-26 16:23:07,632 epoch 7 - iter 16/48 - loss 0.08283968 - time (sec): 7.06 - samples/sec: 1503.32 - lr: 0.000012 - momentum: 0.000000
2024-03-26 16:23:09,888 epoch 7 - iter 20/48 - loss 0.09229244 - time (sec): 9.32 - samples/sec: 1555.37 - lr: 0.000012 - momentum: 0.000000
2024-03-26 16:23:11,222 epoch 7 - iter 24/48 - loss 0.08973072 - time (sec): 10.65 - samples/sec: 1599.53 - lr: 0.000012 - momentum: 0.000000
2024-03-26 16:23:13,436 epoch 7 - iter 28/48 - loss 0.08953758 - time (sec): 12.87 - samples/sec: 1550.28 - lr: 0.000012 - momentum: 0.000000
2024-03-26 16:23:15,290 epoch 7 - iter 32/48 - loss 0.09068748 - time (sec): 14.72 - samples/sec: 1546.78 - lr: 0.000011 - momentum: 0.000000
2024-03-26 16:23:17,277 epoch 7 - iter 36/48 - loss 0.08845022 - time (sec): 16.71 - samples/sec: 1512.95 - lr: 0.000011 - momentum: 0.000000
2024-03-26 16:23:20,044 epoch 7 - iter 40/48 - loss 0.08448801 - time (sec): 19.47 - samples/sec: 1495.60 - lr: 0.000011 - momentum: 0.000000
2024-03-26 16:23:21,520 epoch 7 - iter 44/48 - loss 0.08574476 - time (sec): 20.95 - samples/sec: 1511.66 - lr: 0.000010 - momentum: 0.000000
2024-03-26 16:23:23,638 epoch 7 - iter 48/48 - loss 0.08296352 - time (sec): 23.07 - samples/sec: 1494.36 - lr: 0.000010 - momentum: 0.000000
2024-03-26 16:23:23,639 ----------------------------------------------------------------------------------------------------
2024-03-26 16:23:23,639 EPOCH 7 done: loss 0.0830 - lr: 0.000010
2024-03-26 16:23:24,535 DEV : loss 0.17565500736236572 - f1-score (micro avg) 0.8989
2024-03-26 16:23:24,536 saving best model
2024-03-26 16:23:24,982 ----------------------------------------------------------------------------------------------------
2024-03-26 16:23:27,264 epoch 8 - iter 4/48 - loss 0.09509616 - time (sec): 2.28 - samples/sec: 1223.00 - lr: 0.000010 - momentum: 0.000000
2024-03-26 16:23:28,791 epoch 8 - iter 8/48 - loss 0.06551246 - time (sec): 3.81 - samples/sec: 1427.41 - lr: 0.000010 - momentum: 0.000000
2024-03-26 16:23:31,677 epoch 8 - iter 12/48 - loss 0.05936792 - time (sec): 6.69 - samples/sec: 1344.49 - lr: 0.000009 - momentum: 0.000000
2024-03-26 16:23:34,104 epoch 8 - iter 16/48 - loss 0.06172484 - time (sec): 9.12 - samples/sec: 1347.09 - lr: 0.000009 - momentum: 0.000000
2024-03-26 16:23:35,538 epoch 8 - iter 20/48 - loss 0.05961344 - time (sec): 10.55 - samples/sec: 1407.47 - lr: 0.000009 - momentum: 0.000000
2024-03-26 16:23:36,941 epoch 8 - iter 24/48 - loss 0.06299755 - time (sec): 11.96 - samples/sec: 1479.82 - lr: 0.000009 - momentum: 0.000000
2024-03-26 16:23:38,250 epoch 8 - iter 28/48 - loss 0.06330688 - time (sec): 13.27 - samples/sec: 1541.24 - lr: 0.000008 - momentum: 0.000000
2024-03-26 16:23:40,419 epoch 8 - iter 32/48 - loss 0.06427299 - time (sec): 15.44 - samples/sec: 1500.93 - lr: 0.000008 - momentum: 0.000000
2024-03-26 16:23:42,951 epoch 8 - iter 36/48 - loss 0.06243502 - time (sec): 17.97 - samples/sec: 1457.19 - lr: 0.000008 - momentum: 0.000000
2024-03-26 16:23:44,885 epoch 8 - iter 40/48 - loss 0.06427326 - time (sec): 19.90 - samples/sec: 1466.47 - lr: 0.000007 - momentum: 0.000000
2024-03-26 16:23:47,020 epoch 8 - iter 44/48 - loss 0.06609326 - time (sec): 22.04 - samples/sec: 1449.54 - lr: 0.000007 - momentum: 0.000000
2024-03-26 16:23:48,631 epoch 8 - iter 48/48 - loss 0.06732946 - time (sec): 23.65 - samples/sec: 1457.77 - lr: 0.000007 - momentum: 0.000000
2024-03-26 16:23:48,631 ----------------------------------------------------------------------------------------------------
2024-03-26 16:23:48,631 EPOCH 8 done: loss 0.0673 - lr: 0.000007
2024-03-26 16:23:49,563 DEV : loss 0.17866738140583038 - f1-score (micro avg) 0.9161
2024-03-26 16:23:49,566 saving best model
2024-03-26 16:23:50,022 ----------------------------------------------------------------------------------------------------
2024-03-26 16:23:52,687 epoch 9 - iter 4/48 - loss 0.05635972 - time (sec): 2.66 - samples/sec: 1313.34 - lr: 0.000007 - momentum: 0.000000
2024-03-26 16:23:54,721 epoch 9 - iter 8/48 - loss 0.04969652 - time (sec): 4.70 - samples/sec: 1359.34 - lr: 0.000006 - momentum: 0.000000
2024-03-26 16:23:57,544 epoch 9 - iter 12/48 - loss 0.04878836 - time (sec): 7.52 - samples/sec: 1293.36 - lr: 0.000006 - momentum: 0.000000
2024-03-26 16:24:00,636 epoch 9 - iter 16/48 - loss 0.06176998 - time (sec): 10.61 - samples/sec: 1266.76 - lr: 0.000006 - momentum: 0.000000
2024-03-26 16:24:01,512 epoch 9 - iter 20/48 - loss 0.06154541 - time (sec): 11.49 - samples/sec: 1355.71 - lr: 0.000006 - momentum: 0.000000
2024-03-26 16:24:03,379 epoch 9 - iter 24/48 - loss 0.05983481 - time (sec): 13.36 - samples/sec: 1349.60 - lr: 0.000005 - momentum: 0.000000
2024-03-26 16:24:05,388 epoch 9 - iter 28/48 - loss 0.05892038 - time (sec): 15.37 - samples/sec: 1362.76 - lr: 0.000005 - momentum: 0.000000
2024-03-26 16:24:06,393 epoch 9 - iter 32/48 - loss 0.06007050 - time (sec): 16.37 - samples/sec: 1427.10 - lr: 0.000005 - momentum: 0.000000
2024-03-26 16:24:07,517 epoch 9 - iter 36/48 - loss 0.05951861 - time (sec): 17.49 - samples/sec: 1480.94 - lr: 0.000004 - momentum: 0.000000
2024-03-26 16:24:08,827 epoch 9 - iter 40/48 - loss 0.05760006 - time (sec): 18.80 - samples/sec: 1508.29 - lr: 0.000004 - momentum: 0.000000
2024-03-26 16:24:11,938 epoch 9 - iter 44/48 - loss 0.06032529 - time (sec): 21.92 - samples/sec: 1471.87 - lr: 0.000004 - momentum: 0.000000
2024-03-26 16:24:13,417 epoch 9 - iter 48/48 - loss 0.05798854 - time (sec): 23.39 - samples/sec: 1473.49 - lr: 0.000004 - momentum: 0.000000
2024-03-26 16:24:13,418 ----------------------------------------------------------------------------------------------------
2024-03-26 16:24:13,418 EPOCH 9 done: loss 0.0580 - lr: 0.000004
2024-03-26 16:24:14,343 DEV : loss 0.1765606552362442 - f1-score (micro avg) 0.9177
2024-03-26 16:24:14,344 saving best model
2024-03-26 16:24:14,792 ----------------------------------------------------------------------------------------------------
2024-03-26 16:24:17,626 epoch 10 - iter 4/48 - loss 0.05458151 - time (sec): 2.83 - samples/sec: 1310.58 - lr: 0.000003 - momentum: 0.000000
2024-03-26 16:24:19,584 epoch 10 - iter 8/48 - loss 0.04944972 - time (sec): 4.79 - samples/sec: 1348.75 - lr: 0.000003 - momentum: 0.000000
2024-03-26 16:24:21,756 epoch 10 - iter 12/48 - loss 0.04715640 - time (sec): 6.96 - samples/sec: 1306.11 - lr: 0.000003 - momentum: 0.000000
2024-03-26 16:24:24,192 epoch 10 - iter 16/48 - loss 0.04421213 - time (sec): 9.40 - samples/sec: 1270.61 - lr: 0.000002 - momentum: 0.000000
2024-03-26 16:24:26,739 epoch 10 - iter 20/48 - loss 0.04651412 - time (sec): 11.95 - samples/sec: 1274.41 - lr: 0.000002 - momentum: 0.000000
2024-03-26 16:24:28,153 epoch 10 - iter 24/48 - loss 0.04510724 - time (sec): 13.36 - samples/sec: 1335.07 - lr: 0.000002 - momentum: 0.000000
2024-03-26 16:24:29,025 epoch 10 - iter 28/48 - loss 0.04686140 - time (sec): 14.23 - samples/sec: 1407.49 - lr: 0.000002 - momentum: 0.000000
2024-03-26 16:24:30,944 epoch 10 - iter 32/48 - loss 0.05020359 - time (sec): 16.15 - samples/sec: 1428.10 - lr: 0.000001 - momentum: 0.000000
2024-03-26 16:24:33,193 epoch 10 - iter 36/48 - loss 0.05011506 - time (sec): 18.40 - samples/sec: 1406.00 - lr: 0.000001 - momentum: 0.000000
2024-03-26 16:24:34,840 epoch 10 - iter 40/48 - loss 0.05124147 - time (sec): 20.05 - samples/sec: 1432.59 - lr: 0.000001 - momentum: 0.000000
2024-03-26 16:24:37,982 epoch 10 - iter 44/48 - loss 0.05009590 - time (sec): 23.19 - samples/sec: 1413.36 - lr: 0.000001 - momentum: 0.000000
2024-03-26 16:24:38,698 epoch 10 - iter 48/48 - loss 0.05065391 - time (sec): 23.91 - samples/sec: 1442.04 - lr: 0.000000 - momentum: 0.000000
2024-03-26 16:24:38,698 ----------------------------------------------------------------------------------------------------
2024-03-26 16:24:38,699 EPOCH 10 done: loss 0.0507 - lr: 0.000000
2024-03-26 16:24:39,612 DEV : loss 0.17731061577796936 - f1-score (micro avg) 0.9171
2024-03-26 16:24:39,882 ----------------------------------------------------------------------------------------------------
2024-03-26 16:24:39,883 Loading model from best epoch ...
2024-03-26 16:24:40,622 SequenceTagger predicts: Dictionary with 17 tags: O, S-Unternehmen, B-Unternehmen, E-Unternehmen, I-Unternehmen, S-Auslagerung, B-Auslagerung, E-Auslagerung, I-Auslagerung, S-Ort, B-Ort, E-Ort, I-Ort, S-Software, B-Software, E-Software, I-Software
2024-03-26 16:24:41,491
Results:
- F-score (micro) 0.8926
- F-score (macro) 0.68
- Accuracy 0.8095
By class:
precision recall f1-score support
Unternehmen 0.8876 0.8609 0.8740 266
Auslagerung 0.8566 0.8876 0.8718 249
Ort 0.9635 0.9851 0.9742 134
Software 0.0000 0.0000 0.0000 0
micro avg 0.8885 0.8968 0.8926 649
macro avg 0.6769 0.6834 0.6800 649
weighted avg 0.8914 0.8968 0.8939 649
2024-03-26 16:24:41,491 ----------------------------------------------------------------------------------------------------
|