<!-- 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-base_rvl-cdip-small_rvl_cdip-NK1000_hint_rand
This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the None dataset. It achieves the following results on the evaluation set:
- Loss: 74.2895
- Accuracy: 0.6627
- Brier Loss: 0.6224
- Nll: 3.3689
- F1 Micro: 0.6627
- F1 Macro: 0.6637
- Ece: 0.3019
- Aurc: 0.1471
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: 16
- eval_batch_size: 16
- 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: 50
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy | Brier Loss | Nll | F1 Micro | F1 Macro | Ece | Aurc |
---|---|---|---|---|---|---|---|---|---|---|
77.1398 | 1.0 | 1000 | 76.8452 | 0.1923 | 0.8714 | 4.4881 | 0.1923 | 0.1046 | 0.0814 | 0.6378 |
75.9619 | 2.0 | 2000 | 75.9373 | 0.3513 | 0.7709 | 3.0790 | 0.3513 | 0.3110 | 0.0596 | 0.4537 |
75.5047 | 3.0 | 3000 | 75.7291 | 0.4233 | 0.7112 | 3.0280 | 0.4233 | 0.3913 | 0.0610 | 0.3648 |
75.4727 | 4.0 | 4000 | 75.5639 | 0.4163 | 0.7147 | 3.0030 | 0.4163 | 0.3863 | 0.0669 | 0.3662 |
75.146 | 5.0 | 5000 | 75.4176 | 0.467 | 0.6695 | 2.8545 | 0.467 | 0.4530 | 0.0563 | 0.3180 |
74.8201 | 6.0 | 6000 | 74.8222 | 0.5275 | 0.6023 | 2.6409 | 0.5275 | 0.5201 | 0.0587 | 0.2448 |
74.4727 | 7.0 | 7000 | 74.6341 | 0.5403 | 0.5930 | 2.5700 | 0.5403 | 0.5312 | 0.0707 | 0.2393 |
74.1392 | 8.0 | 8000 | 74.6029 | 0.5615 | 0.5669 | 2.5716 | 0.5615 | 0.5496 | 0.0666 | 0.2142 |
74.165 | 9.0 | 9000 | 74.4072 | 0.5863 | 0.5479 | 2.5087 | 0.5863 | 0.5793 | 0.0689 | 0.1969 |
73.8821 | 10.0 | 10000 | 74.2595 | 0.5817 | 0.5517 | 2.4910 | 0.5817 | 0.5802 | 0.0733 | 0.1973 |
73.6199 | 11.0 | 11000 | 74.2044 | 0.61 | 0.5233 | 2.4183 | 0.61 | 0.6001 | 0.0853 | 0.1722 |
73.4772 | 12.0 | 12000 | 73.9341 | 0.593 | 0.5520 | 2.4592 | 0.593 | 0.5873 | 0.1244 | 0.1847 |
73.2445 | 13.0 | 13000 | 73.9870 | 0.614 | 0.5368 | 2.5577 | 0.614 | 0.6093 | 0.1303 | 0.1706 |
73.1468 | 14.0 | 14000 | 73.9027 | 0.6212 | 0.5368 | 2.6082 | 0.6212 | 0.6192 | 0.1340 | 0.1691 |
72.9154 | 15.0 | 15000 | 73.7745 | 0.6298 | 0.5353 | 2.5866 | 0.6298 | 0.6260 | 0.1564 | 0.1598 |
72.7416 | 16.0 | 16000 | 73.8946 | 0.6225 | 0.5528 | 2.5893 | 0.6225 | 0.6245 | 0.1810 | 0.1679 |
72.4708 | 17.0 | 17000 | 73.9445 | 0.62 | 0.5757 | 2.7436 | 0.62 | 0.6217 | 0.2044 | 0.1714 |
72.5169 | 18.0 | 18000 | 73.7757 | 0.6262 | 0.5741 | 2.6894 | 0.6262 | 0.6292 | 0.2139 | 0.1655 |
72.2021 | 19.0 | 19000 | 73.9482 | 0.6192 | 0.6063 | 2.8813 | 0.6192 | 0.6175 | 0.2534 | 0.1706 |
72.1296 | 20.0 | 20000 | 73.9725 | 0.6185 | 0.6135 | 2.9223 | 0.6185 | 0.6223 | 0.2495 | 0.1736 |
72.1903 | 21.0 | 21000 | 74.0277 | 0.6285 | 0.6091 | 2.8760 | 0.6285 | 0.6307 | 0.2588 | 0.1638 |
71.9868 | 22.0 | 22000 | 74.1811 | 0.6218 | 0.6317 | 3.0858 | 0.6218 | 0.6229 | 0.2792 | 0.1698 |
71.9677 | 23.0 | 23000 | 74.1227 | 0.6222 | 0.6442 | 3.0329 | 0.6222 | 0.6214 | 0.2872 | 0.1764 |
71.8254 | 24.0 | 24000 | 74.2927 | 0.6282 | 0.6412 | 3.1773 | 0.6282 | 0.6220 | 0.2928 | 0.1739 |
71.7948 | 25.0 | 25000 | 74.1580 | 0.626 | 0.6498 | 3.1230 | 0.626 | 0.6286 | 0.3007 | 0.1703 |
71.6915 | 26.0 | 26000 | 74.1776 | 0.6335 | 0.6367 | 3.1272 | 0.6335 | 0.6351 | 0.2937 | 0.1655 |
71.4526 | 27.0 | 27000 | 74.4076 | 0.6335 | 0.6519 | 3.3331 | 0.6335 | 0.6318 | 0.3023 | 0.1749 |
71.2967 | 28.0 | 28000 | 74.1954 | 0.6402 | 0.6361 | 3.1669 | 0.6402 | 0.6392 | 0.2995 | 0.1618 |
71.4139 | 29.0 | 29000 | 74.2737 | 0.6342 | 0.6454 | 3.0744 | 0.6342 | 0.6347 | 0.3070 | 0.1626 |
71.3204 | 30.0 | 30000 | 74.2779 | 0.652 | 0.6277 | 3.2286 | 0.652 | 0.6550 | 0.2956 | 0.1572 |
71.4168 | 31.0 | 31000 | 74.3630 | 0.6458 | 0.6386 | 3.2327 | 0.6458 | 0.6463 | 0.3032 | 0.1594 |
71.387 | 32.0 | 32000 | 74.4710 | 0.6522 | 0.6383 | 3.3193 | 0.6522 | 0.6526 | 0.3016 | 0.1610 |
71.2382 | 33.0 | 33000 | 74.4096 | 0.652 | 0.6275 | 3.3440 | 0.652 | 0.6522 | 0.2977 | 0.1584 |
71.1387 | 34.0 | 34000 | 74.2451 | 0.6512 | 0.6316 | 3.2834 | 0.6512 | 0.6525 | 0.3022 | 0.1555 |
71.0904 | 35.0 | 35000 | 74.2640 | 0.6525 | 0.6341 | 3.1942 | 0.6525 | 0.6518 | 0.3023 | 0.1521 |
70.9615 | 36.0 | 36000 | 74.1828 | 0.6565 | 0.6239 | 3.1805 | 0.6565 | 0.6568 | 0.3014 | 0.1516 |
71.0673 | 37.0 | 37000 | 74.3405 | 0.6498 | 0.6341 | 3.3365 | 0.6498 | 0.6518 | 0.3071 | 0.1556 |
71.0009 | 38.0 | 38000 | 74.2596 | 0.6595 | 0.6296 | 3.3359 | 0.6595 | 0.6622 | 0.2991 | 0.1512 |
70.8441 | 39.0 | 39000 | 74.2837 | 0.6593 | 0.6254 | 3.3852 | 0.6593 | 0.6609 | 0.3005 | 0.1537 |
70.8273 | 40.0 | 40000 | 74.3321 | 0.6567 | 0.6342 | 3.3111 | 0.6567 | 0.6589 | 0.3068 | 0.1544 |
70.8931 | 41.0 | 41000 | 74.3478 | 0.662 | 0.6253 | 3.3022 | 0.662 | 0.6604 | 0.3029 | 0.1474 |
70.8954 | 42.0 | 42000 | 74.2638 | 0.6613 | 0.6275 | 3.3811 | 0.6613 | 0.6612 | 0.3033 | 0.1499 |
70.7389 | 43.0 | 43000 | 74.2531 | 0.6633 | 0.6221 | 3.3627 | 0.6633 | 0.6650 | 0.2998 | 0.1489 |
70.7911 | 44.0 | 44000 | 74.3263 | 0.6587 | 0.6299 | 3.3918 | 0.6587 | 0.6588 | 0.3037 | 0.1496 |
70.8719 | 45.0 | 45000 | 74.2778 | 0.6627 | 0.6236 | 3.3826 | 0.6627 | 0.6641 | 0.3009 | 0.1480 |
70.7289 | 46.0 | 46000 | 74.2760 | 0.6625 | 0.6201 | 3.3467 | 0.6625 | 0.6635 | 0.3016 | 0.1469 |
70.8773 | 47.0 | 47000 | 74.2709 | 0.6643 | 0.6185 | 3.3370 | 0.6643 | 0.6660 | 0.2989 | 0.1476 |
70.6951 | 48.0 | 48000 | 74.2857 | 0.6643 | 0.6218 | 3.3545 | 0.6643 | 0.6648 | 0.2995 | 0.1477 |
70.8059 | 49.0 | 49000 | 74.3124 | 0.6623 | 0.6228 | 3.3592 | 0.6623 | 0.6634 | 0.3020 | 0.1470 |
70.6955 | 50.0 | 50000 | 74.2895 | 0.6627 | 0.6224 | 3.3689 | 0.6627 | 0.6637 | 0.3019 | 0.1471 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.13.1.post200
- Datasets 2.9.0
- Tokenizers 0.13.2