<!-- 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-tiny_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: 267.6730
- Accuracy: 0.6705
- Brier Loss: 0.6262
- Nll: 2.7104
- F1 Micro: 0.6705
- F1 Macro: 0.6721
- Ece: 0.3087
- Aurc: 0.1976
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 |
|---|---|---|---|---|---|---|---|---|---|---|
| 286.7271 | 1.0 | 1000 | 285.5399 | 0.2112 | 1.1285 | 5.2382 | 0.2112 | 0.1362 | 0.4400 | 0.6668 |
| 284.6535 | 2.0 | 2000 | 284.8639 | 0.2365 | 1.1876 | 6.1414 | 0.2365 | 0.1846 | 0.5026 | 0.6043 |
| 283.982 | 3.0 | 3000 | 284.8751 | 0.2555 | 1.2913 | 6.7626 | 0.2555 | 0.2072 | 0.5840 | 0.6111 |
| 283.8947 | 4.0 | 4000 | 283.0353 | 0.3585 | 1.0748 | 4.2918 | 0.3585 | 0.3100 | 0.4921 | 0.4239 |
| 282.5615 | 5.0 | 5000 | 282.0369 | 0.3852 | 1.0142 | 4.7413 | 0.3852 | 0.3432 | 0.4558 | 0.3983 |
| 281.6467 | 6.0 | 6000 | 280.8857 | 0.428 | 0.9539 | 4.1971 | 0.428 | 0.3797 | 0.4329 | 0.3427 |
| 280.8835 | 7.0 | 7000 | 279.7836 | 0.4288 | 1.0391 | 3.9288 | 0.4288 | 0.4012 | 0.4994 | 0.3565 |
| 279.5518 | 8.0 | 8000 | 278.7849 | 0.5198 | 0.8045 | 3.0811 | 0.5198 | 0.4977 | 0.3699 | 0.2454 |
| 278.6091 | 9.0 | 9000 | 278.3536 | 0.5155 | 0.8487 | 3.1204 | 0.5155 | 0.4977 | 0.4004 | 0.2587 |
| 277.9435 | 10.0 | 10000 | 277.6002 | 0.5258 | 0.8346 | 3.3232 | 0.5258 | 0.4899 | 0.3923 | 0.2693 |
| 277.646 | 11.0 | 11000 | 276.9034 | 0.5285 | 0.8510 | 3.1019 | 0.5285 | 0.5010 | 0.4079 | 0.2804 |
| 276.6211 | 12.0 | 12000 | 276.8536 | 0.5555 | 0.7899 | 3.0560 | 0.5555 | 0.5446 | 0.3760 | 0.2266 |
| 276.1643 | 13.0 | 13000 | 275.8300 | 0.5685 | 0.7767 | 3.1275 | 0.5685 | 0.5412 | 0.3730 | 0.2267 |
| 275.7773 | 14.0 | 14000 | 275.0154 | 0.5833 | 0.7536 | 2.9981 | 0.5833 | 0.5645 | 0.3603 | 0.2357 |
| 274.971 | 15.0 | 15000 | 275.1284 | 0.6008 | 0.7210 | 2.8953 | 0.6008 | 0.5920 | 0.3414 | 0.2059 |
| 274.6605 | 16.0 | 16000 | 273.9564 | 0.6132 | 0.7168 | 2.8476 | 0.6132 | 0.5968 | 0.3479 | 0.2272 |
| 273.7713 | 17.0 | 17000 | 273.3493 | 0.5995 | 0.7409 | 2.8991 | 0.5995 | 0.5901 | 0.3607 | 0.2272 |
| 272.7905 | 18.0 | 18000 | 273.5748 | 0.598 | 0.7367 | 2.7778 | 0.598 | 0.5858 | 0.3565 | 0.2102 |
| 273.134 | 19.0 | 19000 | 272.6561 | 0.6158 | 0.7128 | 2.8084 | 0.6158 | 0.6023 | 0.3494 | 0.2132 |
| 271.8558 | 20.0 | 20000 | 272.4530 | 0.618 | 0.7139 | 2.9767 | 0.618 | 0.6077 | 0.3480 | 0.2177 |
| 271.9448 | 21.0 | 21000 | 272.1698 | 0.619 | 0.7164 | 2.9459 | 0.619 | 0.6133 | 0.3510 | 0.2256 |
| 270.9343 | 22.0 | 22000 | 272.2906 | 0.6235 | 0.7087 | 2.9843 | 0.6235 | 0.6181 | 0.3452 | 0.2248 |
| 270.6012 | 23.0 | 23000 | 271.5266 | 0.6382 | 0.6781 | 2.9158 | 0.6382 | 0.6352 | 0.3324 | 0.2110 |
| 270.3184 | 24.0 | 24000 | 271.1095 | 0.634 | 0.6922 | 2.9734 | 0.634 | 0.6287 | 0.3348 | 0.2162 |
| 269.5019 | 25.0 | 25000 | 270.8806 | 0.644 | 0.6683 | 2.8735 | 0.644 | 0.6359 | 0.3258 | 0.2123 |
| 269.5113 | 26.0 | 26000 | 270.6180 | 0.6445 | 0.6650 | 2.6933 | 0.6445 | 0.6418 | 0.3271 | 0.2032 |
| 269.1238 | 27.0 | 27000 | 270.1308 | 0.6445 | 0.6712 | 2.8097 | 0.6445 | 0.6462 | 0.3290 | 0.2128 |
| 268.424 | 28.0 | 28000 | 269.7667 | 0.6352 | 0.6872 | 2.9166 | 0.6352 | 0.6314 | 0.3371 | 0.2231 |
| 268.4034 | 29.0 | 29000 | 270.0039 | 0.6455 | 0.6685 | 2.7765 | 0.6455 | 0.6459 | 0.3273 | 0.2097 |
| 268.3632 | 30.0 | 30000 | 270.0340 | 0.6448 | 0.6741 | 2.8602 | 0.6448 | 0.6455 | 0.3291 | 0.2178 |
| 268.1831 | 31.0 | 31000 | 269.3010 | 0.6597 | 0.6467 | 2.7502 | 0.6597 | 0.6571 | 0.3176 | 0.2053 |
| 268.0006 | 32.0 | 32000 | 269.4335 | 0.652 | 0.6583 | 2.8213 | 0.652 | 0.6457 | 0.3236 | 0.2081 |
| 267.5016 | 33.0 | 33000 | 269.2711 | 0.654 | 0.6530 | 2.8720 | 0.654 | 0.6517 | 0.3199 | 0.2090 |
| 267.177 | 34.0 | 34000 | 268.7774 | 0.661 | 0.6402 | 2.7718 | 0.661 | 0.6589 | 0.3137 | 0.1979 |
| 266.8408 | 35.0 | 35000 | 268.8279 | 0.6478 | 0.6640 | 2.8626 | 0.6478 | 0.6472 | 0.3271 | 0.2204 |
| 266.1984 | 36.0 | 36000 | 268.3442 | 0.6635 | 0.6378 | 2.7999 | 0.6635 | 0.6611 | 0.3128 | 0.2079 |
| 266.1338 | 37.0 | 37000 | 268.5704 | 0.66 | 0.6430 | 2.8314 | 0.66 | 0.6576 | 0.3165 | 0.2039 |
| 266.6958 | 38.0 | 38000 | 268.1453 | 0.6635 | 0.6415 | 2.7881 | 0.6635 | 0.6627 | 0.3147 | 0.2106 |
| 265.6171 | 39.0 | 39000 | 268.1818 | 0.6635 | 0.6398 | 2.7602 | 0.6635 | 0.6641 | 0.3142 | 0.2025 |
| 265.8238 | 40.0 | 40000 | 268.1265 | 0.6637 | 0.6390 | 2.8178 | 0.6637 | 0.6648 | 0.3151 | 0.2016 |
| 265.4164 | 41.0 | 41000 | 267.8777 | 0.6663 | 0.6304 | 2.7649 | 0.6663 | 0.6664 | 0.3113 | 0.2012 |
| 265.6293 | 42.0 | 42000 | 267.8370 | 0.6683 | 0.6285 | 2.7730 | 0.6683 | 0.6677 | 0.3108 | 0.2023 |
| 265.6068 | 43.0 | 43000 | 267.7586 | 0.665 | 0.6348 | 2.7612 | 0.665 | 0.6649 | 0.3126 | 0.1992 |
| 265.2131 | 44.0 | 44000 | 268.0432 | 0.667 | 0.6293 | 2.7217 | 0.667 | 0.6669 | 0.3094 | 0.1885 |
| 265.1312 | 45.0 | 45000 | 267.6967 | 0.6653 | 0.6316 | 2.6899 | 0.6653 | 0.6637 | 0.3127 | 0.2000 |
| 265.371 | 46.0 | 46000 | 267.5307 | 0.668 | 0.6317 | 2.7472 | 0.668 | 0.6684 | 0.3105 | 0.2000 |
| 264.9213 | 47.0 | 47000 | 267.5887 | 0.672 | 0.6214 | 2.6635 | 0.672 | 0.6720 | 0.3063 | 0.1935 |
| 265.1304 | 48.0 | 48000 | 267.4995 | 0.6735 | 0.6220 | 2.7437 | 0.6735 | 0.6730 | 0.3049 | 0.1958 |
| 264.6242 | 49.0 | 49000 | 267.2600 | 0.6723 | 0.6236 | 2.8222 | 0.6723 | 0.6713 | 0.3074 | 0.1974 |
| 265.1563 | 50.0 | 50000 | 267.6730 | 0.6705 | 0.6262 | 2.7104 | 0.6705 | 0.6721 | 0.3087 | 0.1976 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.13.1.post200
- Datasets 2.9.0
- Tokenizers 0.13.2
