<!-- 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_og_simkd_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: 267.1454
- Accuracy: 0.6807
- Brier Loss: 0.6059
- Nll: 2.5092
- F1 Micro: 0.6807
- F1 Macro: 0.6792
- Ece: 0.2988
- Aurc: 0.1779
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 |
|---|---|---|---|---|---|---|---|---|---|---|
| 286.6037 | 1.0 | 1000 | 286.3978 | 0.242 | 1.0585 | 5.0180 | 0.242 | 0.1919 | 0.3885 | 0.6070 |
| 284.5917 | 2.0 | 2000 | 285.0526 | 0.235 | 1.4192 | 6.2048 | 0.235 | 0.1678 | 0.6914 | 0.6366 |
| 284.1567 | 3.0 | 3000 | 283.4989 | 0.3705 | 1.0456 | 4.7503 | 0.3705 | 0.2880 | 0.4669 | 0.4145 |
| 282.6679 | 4.0 | 4000 | 282.5618 | 0.4042 | 0.8940 | 4.1927 | 0.4042 | 0.3644 | 0.3629 | 0.3572 |
| 282.2283 | 5.0 | 5000 | 281.9135 | 0.418 | 0.9976 | 3.8856 | 0.418 | 0.3686 | 0.4631 | 0.3778 |
| 281.3193 | 6.0 | 6000 | 279.9180 | 0.4723 | 0.8755 | 3.4852 | 0.4723 | 0.4312 | 0.3960 | 0.2962 |
| 280.7993 | 7.0 | 7000 | 279.2325 | 0.5038 | 0.8411 | 3.3760 | 0.5038 | 0.4635 | 0.3844 | 0.2753 |
| 279.8249 | 8.0 | 8000 | 278.4682 | 0.5268 | 0.8078 | 3.1572 | 0.5268 | 0.4894 | 0.3705 | 0.2620 |
| 278.8243 | 9.0 | 9000 | 278.2146 | 0.5268 | 0.8245 | 3.2631 | 0.5268 | 0.5043 | 0.3819 | 0.2729 |
| 278.1676 | 10.0 | 10000 | 276.9399 | 0.5607 | 0.7853 | 3.0151 | 0.5607 | 0.5390 | 0.3741 | 0.2275 |
| 276.8185 | 11.0 | 11000 | 276.3879 | 0.5697 | 0.7659 | 2.9137 | 0.5697 | 0.5520 | 0.3660 | 0.2221 |
| 276.0937 | 12.0 | 12000 | 275.9589 | 0.5777 | 0.7626 | 2.9855 | 0.5777 | 0.5643 | 0.3606 | 0.2360 |
| 276.0743 | 13.0 | 13000 | 275.6118 | 0.5675 | 0.7938 | 3.2975 | 0.5675 | 0.5545 | 0.3852 | 0.2320 |
| 275.008 | 14.0 | 14000 | 275.0585 | 0.6 | 0.7359 | 2.8607 | 0.6 | 0.5861 | 0.3517 | 0.2142 |
| 274.483 | 15.0 | 15000 | 274.0515 | 0.6292 | 0.6738 | 2.7667 | 0.6292 | 0.6262 | 0.3215 | 0.1904 |
| 273.261 | 16.0 | 16000 | 273.7844 | 0.6312 | 0.6819 | 2.7219 | 0.6312 | 0.6296 | 0.3286 | 0.2048 |
| 272.9319 | 17.0 | 17000 | 273.4691 | 0.6198 | 0.7009 | 2.8745 | 0.6198 | 0.6160 | 0.3410 | 0.2134 |
| 272.456 | 18.0 | 18000 | 273.1716 | 0.6195 | 0.7071 | 2.8631 | 0.6195 | 0.6223 | 0.3440 | 0.2140 |
| 272.0481 | 19.0 | 19000 | 272.5084 | 0.6322 | 0.6864 | 2.7598 | 0.6322 | 0.6292 | 0.3362 | 0.2119 |
| 271.0429 | 20.0 | 20000 | 272.1741 | 0.6365 | 0.6830 | 2.8104 | 0.6365 | 0.6300 | 0.3345 | 0.2185 |
| 271.0098 | 21.0 | 21000 | 271.8972 | 0.649 | 0.6569 | 2.8558 | 0.649 | 0.6477 | 0.3221 | 0.2076 |
| 270.1226 | 22.0 | 22000 | 271.3564 | 0.639 | 0.6850 | 3.0353 | 0.639 | 0.6326 | 0.3372 | 0.2275 |
| 269.8644 | 23.0 | 23000 | 271.2604 | 0.6332 | 0.6903 | 2.9472 | 0.6332 | 0.6330 | 0.3400 | 0.2367 |
| 269.6737 | 24.0 | 24000 | 270.9163 | 0.6485 | 0.6622 | 2.8937 | 0.6485 | 0.6477 | 0.3258 | 0.2139 |
| 268.3083 | 25.0 | 25000 | 270.3471 | 0.6528 | 0.6590 | 2.7873 | 0.6528 | 0.6550 | 0.3231 | 0.2228 |
| 268.6058 | 26.0 | 26000 | 270.2531 | 0.659 | 0.6377 | 2.7500 | 0.659 | 0.6599 | 0.3125 | 0.1980 |
| 268.5694 | 27.0 | 27000 | 270.0281 | 0.6535 | 0.6510 | 2.7183 | 0.6535 | 0.6502 | 0.3210 | 0.2112 |
| 267.5742 | 28.0 | 28000 | 269.6303 | 0.664 | 0.6327 | 2.6630 | 0.664 | 0.6619 | 0.3109 | 0.1974 |
| 267.4235 | 29.0 | 29000 | 269.3493 | 0.6607 | 0.6417 | 2.7860 | 0.6607 | 0.6568 | 0.3162 | 0.2074 |
| 267.1017 | 30.0 | 30000 | 269.1249 | 0.675 | 0.6152 | 2.6205 | 0.675 | 0.6760 | 0.3013 | 0.1923 |
| 266.7395 | 31.0 | 31000 | 268.8958 | 0.6685 | 0.6281 | 2.7126 | 0.6685 | 0.6638 | 0.3086 | 0.1943 |
| 266.3374 | 32.0 | 32000 | 268.6245 | 0.6703 | 0.6224 | 2.7028 | 0.6703 | 0.6686 | 0.3065 | 0.1900 |
| 266.3529 | 33.0 | 33000 | 268.4537 | 0.6697 | 0.6240 | 2.6593 | 0.6697 | 0.6683 | 0.3066 | 0.1964 |
| 266.1322 | 34.0 | 34000 | 268.1314 | 0.678 | 0.6096 | 2.6485 | 0.678 | 0.6784 | 0.3008 | 0.1857 |
| 265.3824 | 35.0 | 35000 | 268.1505 | 0.6707 | 0.6242 | 2.5832 | 0.6707 | 0.6696 | 0.3058 | 0.1916 |
| 265.5754 | 36.0 | 36000 | 267.9319 | 0.676 | 0.6155 | 2.6208 | 0.676 | 0.6761 | 0.3014 | 0.1908 |
| 265.6115 | 37.0 | 37000 | 268.0886 | 0.679 | 0.6093 | 2.6068 | 0.679 | 0.6795 | 0.2991 | 0.1796 |
| 264.8437 | 38.0 | 38000 | 267.9896 | 0.6783 | 0.6113 | 2.5873 | 0.6783 | 0.6765 | 0.3000 | 0.1805 |
| 264.8028 | 39.0 | 39000 | 267.5381 | 0.68 | 0.6048 | 2.5007 | 0.68 | 0.6771 | 0.2974 | 0.1771 |
| 264.8063 | 40.0 | 40000 | 267.6070 | 0.6763 | 0.6127 | 2.5359 | 0.6763 | 0.6751 | 0.3030 | 0.1821 |
| 264.7481 | 41.0 | 41000 | 267.4914 | 0.6837 | 0.6000 | 2.5214 | 0.6837 | 0.6809 | 0.2942 | 0.1830 |
| 264.6455 | 42.0 | 42000 | 267.6581 | 0.6857 | 0.5968 | 2.5211 | 0.6857 | 0.6856 | 0.2919 | 0.1741 |
| 264.0388 | 43.0 | 43000 | 267.3815 | 0.6797 | 0.6035 | 2.5123 | 0.6797 | 0.6795 | 0.2973 | 0.1773 |
| 264.3585 | 44.0 | 44000 | 267.3548 | 0.6847 | 0.5997 | 2.5583 | 0.6847 | 0.6851 | 0.2943 | 0.1769 |
| 263.7822 | 45.0 | 45000 | 267.0005 | 0.682 | 0.6043 | 2.5023 | 0.682 | 0.6793 | 0.2966 | 0.1788 |
| 263.9765 | 46.0 | 46000 | 267.2113 | 0.6853 | 0.5955 | 2.5256 | 0.6853 | 0.6816 | 0.2922 | 0.1737 |
| 264.1576 | 47.0 | 47000 | 267.1731 | 0.6833 | 0.6002 | 2.5071 | 0.6833 | 0.6825 | 0.2951 | 0.1768 |
| 263.8688 | 48.0 | 48000 | 267.0122 | 0.6843 | 0.5980 | 2.5328 | 0.6843 | 0.6830 | 0.2942 | 0.1781 |
| 263.8963 | 49.0 | 49000 | 266.8628 | 0.6843 | 0.6021 | 2.5231 | 0.6843 | 0.6831 | 0.2957 | 0.1782 |
| 264.2061 | 50.0 | 50000 | 267.1454 | 0.6807 | 0.6059 | 2.5092 | 0.6807 | 0.6792 | 0.2988 | 0.1779 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.13.1.post200
- Datasets 2.9.0
- Tokenizers 0.13.2
