<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
vit-small_tobacco3482_kd_CEKD_t5.0_a0.9
This model is a fine-tuned version of WinKawaks/vit-small-patch16-224 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.5373
- Accuracy: 0.85
- Brier Loss: 0.2432
- Nll: 1.1157
- F1 Micro: 0.85
- F1 Macro: 0.8450
- Ece: 0.1621
- Aurc: 0.0427
Model description
More information needed
Intended uses & limitations
More information needed
Training and evaluation data
More information needed
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.0001
- train_batch_size: 128
- eval_batch_size: 128
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 100
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Brier Loss | Nll | F1 Micro | F1 Macro | Ece | Aurc |
|---|---|---|---|---|---|---|---|---|---|---|
| No log | 1.0 | 7 | 2.1036 | 0.215 | 0.8753 | 5.3195 | 0.2150 | 0.1264 | 0.2571 | 0.6923 |
| No log | 2.0 | 14 | 1.6952 | 0.405 | 0.7407 | 3.4929 | 0.405 | 0.2416 | 0.2907 | 0.4040 |
| No log | 3.0 | 21 | 1.1843 | 0.62 | 0.5633 | 2.0113 | 0.62 | 0.5725 | 0.2740 | 0.2014 |
| No log | 4.0 | 28 | 0.8797 | 0.71 | 0.4080 | 1.7043 | 0.7100 | 0.6683 | 0.2024 | 0.1125 |
| No log | 5.0 | 35 | 0.8570 | 0.715 | 0.3837 | 1.6476 | 0.715 | 0.7280 | 0.2189 | 0.1079 |
| No log | 6.0 | 42 | 0.7484 | 0.775 | 0.3285 | 1.5962 | 0.775 | 0.7668 | 0.1873 | 0.0816 |
| No log | 7.0 | 49 | 0.7337 | 0.79 | 0.3131 | 1.5377 | 0.79 | 0.7779 | 0.1904 | 0.0771 |
| No log | 8.0 | 56 | 0.6709 | 0.795 | 0.3012 | 1.2156 | 0.795 | 0.7776 | 0.1939 | 0.0761 |
| No log | 9.0 | 63 | 0.6901 | 0.795 | 0.3069 | 1.4725 | 0.795 | 0.7916 | 0.1882 | 0.0769 |
| No log | 10.0 | 70 | 0.7960 | 0.75 | 0.3586 | 1.4426 | 0.75 | 0.7406 | 0.1868 | 0.0976 |
| No log | 11.0 | 77 | 0.7489 | 0.77 | 0.3296 | 1.6202 | 0.7700 | 0.7794 | 0.2020 | 0.0878 |
| No log | 12.0 | 84 | 0.7068 | 0.785 | 0.3270 | 1.4127 | 0.785 | 0.7812 | 0.1922 | 0.0759 |
| No log | 13.0 | 91 | 0.6687 | 0.79 | 0.3050 | 1.3820 | 0.79 | 0.7945 | 0.1818 | 0.0625 |
| No log | 14.0 | 98 | 0.6052 | 0.79 | 0.2854 | 1.0602 | 0.79 | 0.7716 | 0.1702 | 0.0590 |
| No log | 15.0 | 105 | 0.6369 | 0.795 | 0.2959 | 1.0580 | 0.795 | 0.7953 | 0.1709 | 0.0603 |
| No log | 16.0 | 112 | 0.6204 | 0.81 | 0.2816 | 1.1886 | 0.81 | 0.8050 | 0.1657 | 0.0702 |
| No log | 17.0 | 119 | 0.5648 | 0.83 | 0.2475 | 1.2506 | 0.83 | 0.8241 | 0.1347 | 0.0612 |
| No log | 18.0 | 126 | 0.5849 | 0.83 | 0.2672 | 1.2245 | 0.83 | 0.8155 | 0.1646 | 0.0601 |
| No log | 19.0 | 133 | 0.5536 | 0.835 | 0.2475 | 1.0514 | 0.835 | 0.8254 | 0.1683 | 0.0531 |
| No log | 20.0 | 140 | 0.5689 | 0.835 | 0.2513 | 1.2369 | 0.835 | 0.8437 | 0.1722 | 0.0489 |
| No log | 21.0 | 147 | 0.5540 | 0.83 | 0.2485 | 1.2139 | 0.83 | 0.8165 | 0.1641 | 0.0608 |
| No log | 22.0 | 154 | 0.5352 | 0.835 | 0.2402 | 1.0108 | 0.835 | 0.8295 | 0.1408 | 0.0430 |
| No log | 23.0 | 161 | 0.5380 | 0.84 | 0.2403 | 1.2280 | 0.8400 | 0.8347 | 0.1405 | 0.0436 |
| No log | 24.0 | 168 | 0.5422 | 0.835 | 0.2471 | 1.0204 | 0.835 | 0.8324 | 0.1606 | 0.0445 |
| No log | 25.0 | 175 | 0.5342 | 0.85 | 0.2404 | 1.0767 | 0.85 | 0.8487 | 0.1469 | 0.0432 |
| No log | 26.0 | 182 | 0.5374 | 0.84 | 0.2429 | 1.0774 | 0.8400 | 0.8334 | 0.1420 | 0.0462 |
| No log | 27.0 | 189 | 0.5311 | 0.85 | 0.2395 | 1.0748 | 0.85 | 0.8487 | 0.1439 | 0.0446 |
| No log | 28.0 | 196 | 0.5298 | 0.85 | 0.2384 | 1.1337 | 0.85 | 0.8487 | 0.1570 | 0.0437 |
| No log | 29.0 | 203 | 0.5387 | 0.845 | 0.2435 | 1.1319 | 0.845 | 0.8424 | 0.1539 | 0.0458 |
| No log | 30.0 | 210 | 0.5361 | 0.85 | 0.2430 | 1.0648 | 0.85 | 0.8450 | 0.1679 | 0.0431 |
| No log | 31.0 | 217 | 0.5339 | 0.85 | 0.2413 | 1.0676 | 0.85 | 0.8487 | 0.1646 | 0.0428 |
| No log | 32.0 | 224 | 0.5345 | 0.85 | 0.2421 | 1.0709 | 0.85 | 0.8487 | 0.1476 | 0.0440 |
| No log | 33.0 | 231 | 0.5343 | 0.85 | 0.2421 | 1.1236 | 0.85 | 0.8450 | 0.1621 | 0.0431 |
| No log | 34.0 | 238 | 0.5353 | 0.845 | 0.2426 | 1.1244 | 0.845 | 0.8424 | 0.1710 | 0.0428 |
| No log | 35.0 | 245 | 0.5346 | 0.85 | 0.2423 | 1.0649 | 0.85 | 0.8487 | 0.1520 | 0.0440 |
| No log | 36.0 | 252 | 0.5356 | 0.855 | 0.2422 | 1.1241 | 0.855 | 0.8517 | 0.1814 | 0.0429 |
| No log | 37.0 | 259 | 0.5357 | 0.85 | 0.2426 | 1.1237 | 0.85 | 0.8450 | 0.1670 | 0.0425 |
| No log | 38.0 | 266 | 0.5356 | 0.845 | 0.2426 | 1.1226 | 0.845 | 0.8419 | 0.1607 | 0.0435 |
| No log | 39.0 | 273 | 0.5347 | 0.855 | 0.2420 | 1.0739 | 0.855 | 0.8517 | 0.1597 | 0.0427 |
| No log | 40.0 | 280 | 0.5356 | 0.855 | 0.2423 | 1.1203 | 0.855 | 0.8517 | 0.1676 | 0.0435 |
| No log | 41.0 | 287 | 0.5365 | 0.85 | 0.2431 | 1.1199 | 0.85 | 0.8450 | 0.1780 | 0.0429 |
| No log | 42.0 | 294 | 0.5356 | 0.85 | 0.2426 | 1.1173 | 0.85 | 0.8450 | 0.1653 | 0.0430 |
| No log | 43.0 | 301 | 0.5363 | 0.85 | 0.2428 | 1.1189 | 0.85 | 0.8450 | 0.1550 | 0.0435 |
| No log | 44.0 | 308 | 0.5345 | 0.85 | 0.2418 | 1.1193 | 0.85 | 0.8450 | 0.1590 | 0.0428 |
| No log | 45.0 | 315 | 0.5374 | 0.85 | 0.2435 | 1.1202 | 0.85 | 0.8450 | 0.1633 | 0.0435 |
| No log | 46.0 | 322 | 0.5355 | 0.85 | 0.2423 | 1.1183 | 0.85 | 0.8450 | 0.1564 | 0.0428 |
| No log | 47.0 | 329 | 0.5354 | 0.85 | 0.2425 | 1.1176 | 0.85 | 0.8450 | 0.1509 | 0.0429 |
| No log | 48.0 | 336 | 0.5369 | 0.85 | 0.2433 | 1.1177 | 0.85 | 0.8450 | 0.1517 | 0.0432 |
| No log | 49.0 | 343 | 0.5361 | 0.85 | 0.2428 | 1.1182 | 0.85 | 0.8450 | 0.1490 | 0.0428 |
| No log | 50.0 | 350 | 0.5364 | 0.85 | 0.2431 | 1.1179 | 0.85 | 0.8450 | 0.1654 | 0.0430 |
| No log | 51.0 | 357 | 0.5365 | 0.85 | 0.2428 | 1.1185 | 0.85 | 0.8450 | 0.1729 | 0.0432 |
| No log | 52.0 | 364 | 0.5364 | 0.85 | 0.2430 | 1.1165 | 0.85 | 0.8450 | 0.1614 | 0.0429 |
| No log | 53.0 | 371 | 0.5362 | 0.85 | 0.2429 | 1.1167 | 0.85 | 0.8450 | 0.1694 | 0.0430 |
| No log | 54.0 | 378 | 0.5369 | 0.85 | 0.2432 | 1.1170 | 0.85 | 0.8450 | 0.1597 | 0.0432 |
| No log | 55.0 | 385 | 0.5368 | 0.85 | 0.2430 | 1.1168 | 0.85 | 0.8450 | 0.1670 | 0.0429 |
| No log | 56.0 | 392 | 0.5367 | 0.85 | 0.2430 | 1.1180 | 0.85 | 0.8450 | 0.1619 | 0.0430 |
| No log | 57.0 | 399 | 0.5364 | 0.85 | 0.2429 | 1.1163 | 0.85 | 0.8450 | 0.1649 | 0.0429 |
| No log | 58.0 | 406 | 0.5364 | 0.85 | 0.2430 | 1.1156 | 0.85 | 0.8450 | 0.1611 | 0.0429 |
| No log | 59.0 | 413 | 0.5365 | 0.85 | 0.2428 | 1.1163 | 0.85 | 0.8450 | 0.1591 | 0.0429 |
| No log | 60.0 | 420 | 0.5364 | 0.85 | 0.2429 | 1.1155 | 0.85 | 0.8450 | 0.1588 | 0.0429 |
| No log | 61.0 | 427 | 0.5370 | 0.85 | 0.2432 | 1.1158 | 0.85 | 0.8450 | 0.1772 | 0.0432 |
| No log | 62.0 | 434 | 0.5367 | 0.85 | 0.2429 | 1.1167 | 0.85 | 0.8450 | 0.1622 | 0.0429 |
| No log | 63.0 | 441 | 0.5362 | 0.85 | 0.2428 | 1.1162 | 0.85 | 0.8450 | 0.1503 | 0.0428 |
| No log | 64.0 | 448 | 0.5372 | 0.85 | 0.2433 | 1.1161 | 0.85 | 0.8450 | 0.1616 | 0.0432 |
| No log | 65.0 | 455 | 0.5371 | 0.85 | 0.2431 | 1.1162 | 0.85 | 0.8450 | 0.1499 | 0.0429 |
| No log | 66.0 | 462 | 0.5367 | 0.85 | 0.2430 | 1.1160 | 0.85 | 0.8450 | 0.1591 | 0.0427 |
| No log | 67.0 | 469 | 0.5367 | 0.85 | 0.2430 | 1.1164 | 0.85 | 0.8450 | 0.1562 | 0.0428 |
| No log | 68.0 | 476 | 0.5368 | 0.85 | 0.2430 | 1.1168 | 0.85 | 0.8450 | 0.1556 | 0.0427 |
| No log | 69.0 | 483 | 0.5368 | 0.85 | 0.2431 | 1.1158 | 0.85 | 0.8450 | 0.1593 | 0.0428 |
| No log | 70.0 | 490 | 0.5372 | 0.85 | 0.2432 | 1.1162 | 0.85 | 0.8450 | 0.1628 | 0.0428 |
| No log | 71.0 | 497 | 0.5371 | 0.85 | 0.2432 | 1.1163 | 0.85 | 0.8450 | 0.1599 | 0.0429 |
| 0.1708 | 72.0 | 504 | 0.5370 | 0.85 | 0.2430 | 1.1161 | 0.85 | 0.8450 | 0.1559 | 0.0430 |
| 0.1708 | 73.0 | 511 | 0.5372 | 0.85 | 0.2433 | 1.1154 | 0.85 | 0.8450 | 0.1556 | 0.0428 |
| 0.1708 | 74.0 | 518 | 0.5370 | 0.85 | 0.2429 | 1.1165 | 0.85 | 0.8450 | 0.1540 | 0.0428 |
| 0.1708 | 75.0 | 525 | 0.5371 | 0.85 | 0.2431 | 1.1161 | 0.85 | 0.8450 | 0.1616 | 0.0427 |
| 0.1708 | 76.0 | 532 | 0.5369 | 0.85 | 0.2431 | 1.1161 | 0.85 | 0.8450 | 0.1619 | 0.0427 |
| 0.1708 | 77.0 | 539 | 0.5369 | 0.85 | 0.2430 | 1.1156 | 0.85 | 0.8450 | 0.1623 | 0.0429 |
| 0.1708 | 78.0 | 546 | 0.5372 | 0.85 | 0.2432 | 1.1158 | 0.85 | 0.8450 | 0.1619 | 0.0427 |
| 0.1708 | 79.0 | 553 | 0.5375 | 0.85 | 0.2433 | 1.1162 | 0.85 | 0.8450 | 0.1688 | 0.0429 |
| 0.1708 | 80.0 | 560 | 0.5372 | 0.85 | 0.2432 | 1.1160 | 0.85 | 0.8450 | 0.1623 | 0.0429 |
| 0.1708 | 81.0 | 567 | 0.5373 | 0.85 | 0.2432 | 1.1162 | 0.85 | 0.8450 | 0.1620 | 0.0428 |
| 0.1708 | 82.0 | 574 | 0.5374 | 0.85 | 0.2433 | 1.1160 | 0.85 | 0.8450 | 0.1622 | 0.0428 |
| 0.1708 | 83.0 | 581 | 0.5372 | 0.85 | 0.2432 | 1.1159 | 0.85 | 0.8450 | 0.1622 | 0.0428 |
| 0.1708 | 84.0 | 588 | 0.5371 | 0.85 | 0.2431 | 1.1157 | 0.85 | 0.8450 | 0.1621 | 0.0427 |
| 0.1708 | 85.0 | 595 | 0.5372 | 0.85 | 0.2432 | 1.1158 | 0.85 | 0.8450 | 0.1687 | 0.0426 |
| 0.1708 | 86.0 | 602 | 0.5372 | 0.85 | 0.2432 | 1.1157 | 0.85 | 0.8450 | 0.1619 | 0.0426 |
| 0.1708 | 87.0 | 609 | 0.5374 | 0.85 | 0.2432 | 1.1159 | 0.85 | 0.8450 | 0.1687 | 0.0428 |
| 0.1708 | 88.0 | 616 | 0.5373 | 0.85 | 0.2432 | 1.1160 | 0.85 | 0.8450 | 0.1620 | 0.0427 |
| 0.1708 | 89.0 | 623 | 0.5373 | 0.85 | 0.2432 | 1.1157 | 0.85 | 0.8450 | 0.1620 | 0.0427 |
| 0.1708 | 90.0 | 630 | 0.5373 | 0.85 | 0.2432 | 1.1156 | 0.85 | 0.8450 | 0.1620 | 0.0427 |
| 0.1708 | 91.0 | 637 | 0.5372 | 0.85 | 0.2432 | 1.1156 | 0.85 | 0.8450 | 0.1620 | 0.0427 |
| 0.1708 | 92.0 | 644 | 0.5373 | 0.85 | 0.2432 | 1.1157 | 0.85 | 0.8450 | 0.1620 | 0.0427 |
| 0.1708 | 93.0 | 651 | 0.5372 | 0.85 | 0.2432 | 1.1156 | 0.85 | 0.8450 | 0.1620 | 0.0427 |
| 0.1708 | 94.0 | 658 | 0.5373 | 0.85 | 0.2432 | 1.1158 | 0.85 | 0.8450 | 0.1620 | 0.0427 |
| 0.1708 | 95.0 | 665 | 0.5373 | 0.85 | 0.2432 | 1.1157 | 0.85 | 0.8450 | 0.1621 | 0.0427 |
| 0.1708 | 96.0 | 672 | 0.5372 | 0.85 | 0.2432 | 1.1157 | 0.85 | 0.8450 | 0.1621 | 0.0427 |
| 0.1708 | 97.0 | 679 | 0.5372 | 0.85 | 0.2432 | 1.1157 | 0.85 | 0.8450 | 0.1620 | 0.0427 |
| 0.1708 | 98.0 | 686 | 0.5373 | 0.85 | 0.2432 | 1.1157 | 0.85 | 0.8450 | 0.1621 | 0.0427 |
| 0.1708 | 99.0 | 693 | 0.5373 | 0.85 | 0.2432 | 1.1157 | 0.85 | 0.8450 | 0.1621 | 0.0427 |
| 0.1708 | 100.0 | 700 | 0.5373 | 0.85 | 0.2432 | 1.1157 | 0.85 | 0.8450 | 0.1621 | 0.0427 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.13.1.post200
- Datasets 2.9.0
- Tokenizers 0.13.2
