<!-- 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. -->
225-tiny_tobacco3482_kd
This model is a fine-tuned version of WinKawaks/vit-tiny-patch16-224 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.2991
- Accuracy: 0.775
- Brier Loss: 0.3491
- Nll: 1.2196
- F1 Micro: 0.775
- F1 Macro: 0.7302
- Ece: 0.2602
- Aurc: 0.0644
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: 64
- eval_batch_size: 64
- 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 | 13 | 1.6344 | 0.23 | 0.8900 | 7.7885 | 0.23 | 0.1633 | 0.2747 | 0.7585 |
No log | 2.0 | 26 | 1.0824 | 0.385 | 0.7943 | 4.1668 | 0.3850 | 0.2795 | 0.3160 | 0.4560 |
No log | 3.0 | 39 | 0.8639 | 0.535 | 0.6762 | 3.0145 | 0.535 | 0.4086 | 0.3233 | 0.2913 |
No log | 4.0 | 52 | 0.7309 | 0.595 | 0.5956 | 2.2236 | 0.595 | 0.4646 | 0.3075 | 0.1944 |
No log | 5.0 | 65 | 0.6374 | 0.67 | 0.5211 | 2.1759 | 0.67 | 0.5737 | 0.2898 | 0.1450 |
No log | 6.0 | 78 | 0.6720 | 0.685 | 0.4833 | 2.2861 | 0.685 | 0.5860 | 0.2904 | 0.1331 |
No log | 7.0 | 91 | 0.6097 | 0.675 | 0.4767 | 2.3133 | 0.675 | 0.5733 | 0.2622 | 0.1519 |
No log | 8.0 | 104 | 0.5206 | 0.705 | 0.4301 | 1.8228 | 0.705 | 0.6164 | 0.2603 | 0.1038 |
No log | 9.0 | 117 | 0.5486 | 0.715 | 0.4414 | 1.8451 | 0.715 | 0.6444 | 0.2583 | 0.1063 |
No log | 10.0 | 130 | 0.5067 | 0.7 | 0.4171 | 1.7759 | 0.7 | 0.6325 | 0.2611 | 0.1071 |
No log | 11.0 | 143 | 0.4612 | 0.745 | 0.4017 | 1.4919 | 0.745 | 0.6635 | 0.2840 | 0.0838 |
No log | 12.0 | 156 | 0.4785 | 0.745 | 0.4204 | 1.8579 | 0.745 | 0.6750 | 0.2542 | 0.0979 |
No log | 13.0 | 169 | 0.4518 | 0.715 | 0.4036 | 1.5697 | 0.715 | 0.6496 | 0.2744 | 0.1002 |
No log | 14.0 | 182 | 0.5081 | 0.7 | 0.4294 | 1.9850 | 0.7 | 0.6514 | 0.2364 | 0.1225 |
No log | 15.0 | 195 | 0.4415 | 0.705 | 0.3994 | 1.7828 | 0.705 | 0.6301 | 0.2380 | 0.0992 |
No log | 16.0 | 208 | 0.3859 | 0.73 | 0.3832 | 1.3431 | 0.7300 | 0.6516 | 0.2548 | 0.0817 |
No log | 17.0 | 221 | 0.3869 | 0.75 | 0.3832 | 1.2075 | 0.75 | 0.6651 | 0.2622 | 0.0758 |
No log | 18.0 | 234 | 0.3637 | 0.755 | 0.3770 | 1.2290 | 0.755 | 0.7108 | 0.2569 | 0.0687 |
No log | 19.0 | 247 | 0.3933 | 0.745 | 0.3700 | 1.4931 | 0.745 | 0.6812 | 0.2434 | 0.0799 |
No log | 20.0 | 260 | 0.3540 | 0.745 | 0.3721 | 1.1910 | 0.745 | 0.6702 | 0.2208 | 0.0760 |
No log | 21.0 | 273 | 0.3560 | 0.77 | 0.3718 | 1.1248 | 0.7700 | 0.7142 | 0.2731 | 0.0743 |
No log | 22.0 | 286 | 0.3530 | 0.74 | 0.3758 | 1.4213 | 0.74 | 0.6902 | 0.2326 | 0.0768 |
No log | 23.0 | 299 | 0.3419 | 0.745 | 0.3699 | 1.2528 | 0.745 | 0.6714 | 0.2324 | 0.0765 |
No log | 24.0 | 312 | 0.3302 | 0.775 | 0.3595 | 1.3338 | 0.775 | 0.7120 | 0.2521 | 0.0665 |
No log | 25.0 | 325 | 0.3533 | 0.775 | 0.3672 | 1.4609 | 0.775 | 0.7167 | 0.2482 | 0.0740 |
No log | 26.0 | 338 | 0.3416 | 0.775 | 0.3684 | 1.1575 | 0.775 | 0.7124 | 0.2601 | 0.0732 |
No log | 27.0 | 351 | 0.3463 | 0.75 | 0.3714 | 1.1053 | 0.75 | 0.6868 | 0.2512 | 0.0808 |
No log | 28.0 | 364 | 0.3298 | 0.775 | 0.3605 | 1.2108 | 0.775 | 0.6986 | 0.2537 | 0.0668 |
No log | 29.0 | 377 | 0.3278 | 0.77 | 0.3645 | 1.1893 | 0.7700 | 0.7013 | 0.2447 | 0.0765 |
No log | 30.0 | 390 | 0.3165 | 0.78 | 0.3608 | 1.1615 | 0.78 | 0.7285 | 0.2472 | 0.0712 |
No log | 31.0 | 403 | 0.3212 | 0.765 | 0.3571 | 1.1317 | 0.765 | 0.6999 | 0.2497 | 0.0725 |
No log | 32.0 | 416 | 0.3119 | 0.765 | 0.3581 | 1.0644 | 0.765 | 0.6881 | 0.2285 | 0.0675 |
No log | 33.0 | 429 | 0.3229 | 0.765 | 0.3523 | 1.2937 | 0.765 | 0.7138 | 0.2517 | 0.0658 |
No log | 34.0 | 442 | 0.3193 | 0.78 | 0.3660 | 1.1849 | 0.78 | 0.7329 | 0.2686 | 0.0700 |
No log | 35.0 | 455 | 0.3088 | 0.775 | 0.3556 | 1.1613 | 0.775 | 0.7071 | 0.2640 | 0.0680 |
No log | 36.0 | 468 | 0.3113 | 0.785 | 0.3508 | 1.1715 | 0.785 | 0.7501 | 0.2443 | 0.0656 |
No log | 37.0 | 481 | 0.3113 | 0.79 | 0.3526 | 1.2334 | 0.79 | 0.7388 | 0.2580 | 0.0639 |
No log | 38.0 | 494 | 0.3077 | 0.755 | 0.3528 | 1.1152 | 0.755 | 0.6973 | 0.2401 | 0.0692 |
0.2783 | 39.0 | 507 | 0.3064 | 0.775 | 0.3567 | 1.2289 | 0.775 | 0.7370 | 0.2417 | 0.0696 |
0.2783 | 40.0 | 520 | 0.3063 | 0.77 | 0.3521 | 1.2437 | 0.7700 | 0.7232 | 0.2396 | 0.0688 |
0.2783 | 41.0 | 533 | 0.3042 | 0.77 | 0.3541 | 1.2490 | 0.7700 | 0.7234 | 0.2470 | 0.0682 |
0.2783 | 42.0 | 546 | 0.2999 | 0.77 | 0.3486 | 1.1626 | 0.7700 | 0.7082 | 0.2491 | 0.0638 |
0.2783 | 43.0 | 559 | 0.3020 | 0.77 | 0.3515 | 1.2141 | 0.7700 | 0.7312 | 0.2570 | 0.0687 |
0.2783 | 44.0 | 572 | 0.3024 | 0.775 | 0.3502 | 1.2184 | 0.775 | 0.7168 | 0.2568 | 0.0648 |
0.2783 | 45.0 | 585 | 0.3002 | 0.78 | 0.3517 | 1.2189 | 0.78 | 0.7364 | 0.2673 | 0.0644 |
0.2783 | 46.0 | 598 | 0.3022 | 0.775 | 0.3511 | 1.1594 | 0.775 | 0.7266 | 0.2538 | 0.0661 |
0.2783 | 47.0 | 611 | 0.2974 | 0.775 | 0.3464 | 1.2157 | 0.775 | 0.7238 | 0.2630 | 0.0627 |
0.2783 | 48.0 | 624 | 0.3003 | 0.78 | 0.3519 | 1.1584 | 0.78 | 0.7318 | 0.2413 | 0.0666 |
0.2783 | 49.0 | 637 | 0.2990 | 0.77 | 0.3492 | 1.2187 | 0.7700 | 0.7136 | 0.2401 | 0.0643 |
0.2783 | 50.0 | 650 | 0.3019 | 0.765 | 0.3516 | 1.2254 | 0.765 | 0.7180 | 0.2409 | 0.0673 |
0.2783 | 51.0 | 663 | 0.2991 | 0.77 | 0.3499 | 1.2186 | 0.7700 | 0.7145 | 0.2566 | 0.0646 |
0.2783 | 52.0 | 676 | 0.2990 | 0.77 | 0.3507 | 1.2204 | 0.7700 | 0.7207 | 0.2360 | 0.0651 |
0.2783 | 53.0 | 689 | 0.2982 | 0.765 | 0.3488 | 1.1663 | 0.765 | 0.7042 | 0.2338 | 0.0643 |
0.2783 | 54.0 | 702 | 0.2969 | 0.775 | 0.3485 | 1.1667 | 0.775 | 0.7302 | 0.2586 | 0.0642 |
0.2783 | 55.0 | 715 | 0.2989 | 0.775 | 0.3487 | 1.2181 | 0.775 | 0.7302 | 0.2670 | 0.0647 |
0.2783 | 56.0 | 728 | 0.2991 | 0.77 | 0.3499 | 1.2208 | 0.7700 | 0.7136 | 0.2339 | 0.0650 |
0.2783 | 57.0 | 741 | 0.2986 | 0.775 | 0.3487 | 1.2162 | 0.775 | 0.7302 | 0.2415 | 0.0639 |
0.2783 | 58.0 | 754 | 0.2985 | 0.77 | 0.3490 | 1.2183 | 0.7700 | 0.7207 | 0.2547 | 0.0647 |
0.2783 | 59.0 | 767 | 0.2993 | 0.77 | 0.3494 | 1.2218 | 0.7700 | 0.7136 | 0.2417 | 0.0649 |
0.2783 | 60.0 | 780 | 0.2983 | 0.77 | 0.3487 | 1.2185 | 0.7700 | 0.7207 | 0.2555 | 0.0646 |
0.2783 | 61.0 | 793 | 0.2989 | 0.775 | 0.3492 | 1.2182 | 0.775 | 0.7302 | 0.2444 | 0.0645 |
0.2783 | 62.0 | 806 | 0.2987 | 0.775 | 0.3487 | 1.2174 | 0.775 | 0.7302 | 0.2438 | 0.0642 |
0.2783 | 63.0 | 819 | 0.2987 | 0.775 | 0.3490 | 1.2198 | 0.775 | 0.7302 | 0.2508 | 0.0646 |
0.2783 | 64.0 | 832 | 0.2989 | 0.775 | 0.3494 | 1.2195 | 0.775 | 0.7302 | 0.2609 | 0.0646 |
0.2783 | 65.0 | 845 | 0.2990 | 0.775 | 0.3492 | 1.2177 | 0.775 | 0.7302 | 0.2528 | 0.0644 |
0.2783 | 66.0 | 858 | 0.2992 | 0.775 | 0.3493 | 1.2193 | 0.775 | 0.7302 | 0.2537 | 0.0646 |
0.2783 | 67.0 | 871 | 0.2990 | 0.775 | 0.3493 | 1.2199 | 0.775 | 0.7302 | 0.2510 | 0.0647 |
0.2783 | 68.0 | 884 | 0.2991 | 0.775 | 0.3495 | 1.2199 | 0.775 | 0.7302 | 0.2476 | 0.0646 |
0.2783 | 69.0 | 897 | 0.2989 | 0.775 | 0.3491 | 1.2187 | 0.775 | 0.7302 | 0.2606 | 0.0646 |
0.2783 | 70.0 | 910 | 0.2987 | 0.775 | 0.3490 | 1.2187 | 0.775 | 0.7302 | 0.2436 | 0.0642 |
0.2783 | 71.0 | 923 | 0.2990 | 0.775 | 0.3491 | 1.2190 | 0.775 | 0.7302 | 0.2510 | 0.0646 |
0.2783 | 72.0 | 936 | 0.2990 | 0.775 | 0.3492 | 1.2191 | 0.775 | 0.7302 | 0.2541 | 0.0646 |
0.2783 | 73.0 | 949 | 0.2990 | 0.775 | 0.3491 | 1.2176 | 0.775 | 0.7302 | 0.2509 | 0.0647 |
0.2783 | 74.0 | 962 | 0.2990 | 0.775 | 0.3493 | 1.2203 | 0.775 | 0.7302 | 0.2600 | 0.0643 |
0.2783 | 75.0 | 975 | 0.2989 | 0.775 | 0.3492 | 1.2203 | 0.775 | 0.7302 | 0.2665 | 0.0643 |
0.2783 | 76.0 | 988 | 0.2991 | 0.775 | 0.3492 | 1.2193 | 0.775 | 0.7302 | 0.2601 | 0.0643 |
0.0005 | 77.0 | 1001 | 0.2991 | 0.775 | 0.3491 | 1.2201 | 0.775 | 0.7302 | 0.2598 | 0.0645 |
0.0005 | 78.0 | 1014 | 0.2991 | 0.775 | 0.3490 | 1.2198 | 0.775 | 0.7302 | 0.2441 | 0.0645 |
0.0005 | 79.0 | 1027 | 0.2991 | 0.775 | 0.3492 | 1.2182 | 0.775 | 0.7302 | 0.2513 | 0.0645 |
0.0005 | 80.0 | 1040 | 0.2992 | 0.775 | 0.3491 | 1.2183 | 0.775 | 0.7302 | 0.2514 | 0.0645 |
0.0005 | 81.0 | 1053 | 0.2992 | 0.775 | 0.3492 | 1.2196 | 0.775 | 0.7302 | 0.2584 | 0.0646 |
0.0005 | 82.0 | 1066 | 0.2992 | 0.775 | 0.3493 | 1.2199 | 0.775 | 0.7302 | 0.2520 | 0.0646 |
0.0005 | 83.0 | 1079 | 0.2991 | 0.775 | 0.3491 | 1.2191 | 0.775 | 0.7302 | 0.2514 | 0.0643 |
0.0005 | 84.0 | 1092 | 0.2991 | 0.775 | 0.3491 | 1.2194 | 0.775 | 0.7302 | 0.2516 | 0.0645 |
0.0005 | 85.0 | 1105 | 0.2990 | 0.775 | 0.3491 | 1.2188 | 0.775 | 0.7302 | 0.2585 | 0.0645 |
0.0005 | 86.0 | 1118 | 0.2991 | 0.775 | 0.3492 | 1.2193 | 0.775 | 0.7302 | 0.2584 | 0.0645 |
0.0005 | 87.0 | 1131 | 0.2991 | 0.775 | 0.3491 | 1.2201 | 0.775 | 0.7302 | 0.2667 | 0.0643 |
0.0005 | 88.0 | 1144 | 0.2991 | 0.775 | 0.3492 | 1.2199 | 0.775 | 0.7302 | 0.2516 | 0.0645 |
0.0005 | 89.0 | 1157 | 0.2990 | 0.775 | 0.3491 | 1.2193 | 0.775 | 0.7302 | 0.2603 | 0.0644 |
0.0005 | 90.0 | 1170 | 0.2990 | 0.775 | 0.3492 | 1.2197 | 0.775 | 0.7302 | 0.2536 | 0.0645 |
0.0005 | 91.0 | 1183 | 0.2990 | 0.775 | 0.3491 | 1.2201 | 0.775 | 0.7302 | 0.2668 | 0.0644 |
0.0005 | 92.0 | 1196 | 0.2991 | 0.775 | 0.3491 | 1.2190 | 0.775 | 0.7302 | 0.2533 | 0.0644 |
0.0005 | 93.0 | 1209 | 0.2991 | 0.775 | 0.3492 | 1.2192 | 0.775 | 0.7302 | 0.2602 | 0.0645 |
0.0005 | 94.0 | 1222 | 0.2991 | 0.775 | 0.3492 | 1.2193 | 0.775 | 0.7302 | 0.2533 | 0.0645 |
0.0005 | 95.0 | 1235 | 0.2991 | 0.775 | 0.3491 | 1.2192 | 0.775 | 0.7302 | 0.2533 | 0.0644 |
0.0005 | 96.0 | 1248 | 0.2991 | 0.775 | 0.3491 | 1.2196 | 0.775 | 0.7302 | 0.2668 | 0.0644 |
0.0005 | 97.0 | 1261 | 0.2991 | 0.775 | 0.3492 | 1.2196 | 0.775 | 0.7302 | 0.2602 | 0.0644 |
0.0005 | 98.0 | 1274 | 0.2991 | 0.775 | 0.3491 | 1.2194 | 0.775 | 0.7302 | 0.2533 | 0.0644 |
0.0005 | 99.0 | 1287 | 0.2991 | 0.775 | 0.3491 | 1.2195 | 0.775 | 0.7302 | 0.2602 | 0.0644 |
0.0005 | 100.0 | 1300 | 0.2991 | 0.775 | 0.3491 | 1.2196 | 0.775 | 0.7302 | 0.2602 | 0.0644 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.13.1.post200
- Datasets 2.9.0
- Tokenizers 0.13.2