<!-- 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_tobacco
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: 0.7442
- Accuracy: 0.815
- Brier Loss: 0.3076
- Nll: 1.1877
- F1 Micro: 0.815
- F1 Macro: 0.7942
- Ece: 0.2072
- Aurc: 0.0734
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: 2e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 16
- total_train_batch_size: 128
- 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 | 0.96 | 6 | 2.3082 | 0.085 | 0.9012 | 6.2672 | 0.085 | 0.0735 | 0.1625 | 0.9316 |
No log | 1.96 | 12 | 2.2872 | 0.14 | 0.8970 | 4.8533 | 0.14 | 0.0885 | 0.1958 | 0.8912 |
No log | 2.96 | 18 | 2.2562 | 0.225 | 0.8906 | 4.5559 | 0.225 | 0.1319 | 0.2527 | 0.8101 |
No log | 3.96 | 24 | 2.2107 | 0.265 | 0.8808 | 4.3151 | 0.265 | 0.1614 | 0.2710 | 0.6990 |
No log | 4.96 | 30 | 2.1433 | 0.3 | 0.8654 | 4.1825 | 0.3 | 0.1615 | 0.2943 | 0.6102 |
No log | 5.96 | 36 | 2.0764 | 0.325 | 0.8493 | 3.6715 | 0.325 | 0.1696 | 0.3160 | 0.4502 |
No log | 6.96 | 42 | 2.0012 | 0.375 | 0.8287 | 3.5534 | 0.375 | 0.1901 | 0.3542 | 0.3791 |
No log | 7.96 | 48 | 1.9197 | 0.41 | 0.8041 | 3.3582 | 0.41 | 0.2136 | 0.3528 | 0.3342 |
No log | 8.96 | 54 | 1.8379 | 0.45 | 0.7767 | 3.1997 | 0.45 | 0.2279 | 0.3709 | 0.2872 |
No log | 9.96 | 60 | 1.7538 | 0.535 | 0.7475 | 2.9586 | 0.535 | 0.3755 | 0.4024 | 0.2508 |
No log | 10.96 | 66 | 1.6634 | 0.57 | 0.7132 | 2.6969 | 0.57 | 0.4025 | 0.4182 | 0.2183 |
No log | 11.96 | 72 | 1.5952 | 0.61 | 0.6842 | 2.4519 | 0.61 | 0.4427 | 0.4153 | 0.1882 |
No log | 12.96 | 78 | 1.5205 | 0.655 | 0.6554 | 1.9703 | 0.655 | 0.5306 | 0.4572 | 0.1651 |
No log | 13.96 | 84 | 1.4566 | 0.67 | 0.6308 | 1.7832 | 0.67 | 0.5458 | 0.4240 | 0.1514 |
No log | 14.96 | 90 | 1.4009 | 0.685 | 0.6074 | 1.8217 | 0.685 | 0.5641 | 0.4221 | 0.1406 |
No log | 15.96 | 96 | 1.3520 | 0.7 | 0.5866 | 1.6223 | 0.7 | 0.5896 | 0.4107 | 0.1304 |
No log | 16.96 | 102 | 1.3220 | 0.7 | 0.5741 | 1.4452 | 0.7 | 0.5865 | 0.4029 | 0.1225 |
No log | 17.96 | 108 | 1.2764 | 0.705 | 0.5522 | 1.4534 | 0.705 | 0.6076 | 0.3805 | 0.1269 |
No log | 18.96 | 114 | 1.2448 | 0.72 | 0.5378 | 1.4843 | 0.72 | 0.6321 | 0.3724 | 0.1193 |
No log | 19.96 | 120 | 1.2049 | 0.74 | 0.5210 | 1.2527 | 0.74 | 0.6471 | 0.3947 | 0.1039 |
No log | 20.96 | 126 | 1.1712 | 0.74 | 0.5057 | 1.1657 | 0.74 | 0.6464 | 0.3833 | 0.0955 |
No log | 21.96 | 132 | 1.1453 | 0.735 | 0.4936 | 1.0277 | 0.735 | 0.6597 | 0.3628 | 0.1015 |
No log | 22.96 | 138 | 1.1094 | 0.745 | 0.4771 | 1.0003 | 0.745 | 0.6667 | 0.3841 | 0.0938 |
No log | 23.96 | 144 | 1.0803 | 0.75 | 0.4628 | 1.0334 | 0.75 | 0.6972 | 0.3490 | 0.0891 |
No log | 24.96 | 150 | 1.0658 | 0.755 | 0.4559 | 1.0092 | 0.755 | 0.6937 | 0.3536 | 0.0925 |
No log | 25.96 | 156 | 1.0345 | 0.765 | 0.4423 | 0.9971 | 0.765 | 0.7356 | 0.3661 | 0.0852 |
No log | 26.96 | 162 | 1.0133 | 0.76 | 0.4323 | 0.9206 | 0.76 | 0.7302 | 0.3343 | 0.0791 |
No log | 27.96 | 168 | 0.9927 | 0.775 | 0.4225 | 0.9015 | 0.775 | 0.7433 | 0.3457 | 0.0794 |
No log | 28.96 | 174 | 0.9789 | 0.765 | 0.4152 | 0.8946 | 0.765 | 0.7282 | 0.3337 | 0.0818 |
No log | 29.96 | 180 | 0.9509 | 0.78 | 0.4025 | 0.9323 | 0.78 | 0.7565 | 0.3135 | 0.0733 |
No log | 30.96 | 186 | 0.9388 | 0.79 | 0.3968 | 0.8616 | 0.79 | 0.7642 | 0.3353 | 0.0694 |
No log | 31.96 | 192 | 0.9316 | 0.78 | 0.3927 | 0.8636 | 0.78 | 0.7588 | 0.3426 | 0.0739 |
No log | 32.96 | 198 | 0.9197 | 0.79 | 0.3876 | 0.8581 | 0.79 | 0.7656 | 0.3042 | 0.0800 |
No log | 33.96 | 204 | 0.9020 | 0.775 | 0.3792 | 0.8458 | 0.775 | 0.7543 | 0.2872 | 0.0744 |
No log | 34.96 | 210 | 0.8833 | 0.785 | 0.3694 | 0.8288 | 0.785 | 0.7619 | 0.3305 | 0.0663 |
No log | 35.96 | 216 | 0.8684 | 0.795 | 0.3624 | 0.8462 | 0.795 | 0.7779 | 0.3184 | 0.0690 |
No log | 36.96 | 222 | 0.8608 | 0.79 | 0.3584 | 0.8860 | 0.79 | 0.7707 | 0.2790 | 0.0709 |
No log | 37.96 | 228 | 0.8586 | 0.79 | 0.3587 | 0.8954 | 0.79 | 0.7724 | 0.3153 | 0.0754 |
No log | 38.96 | 234 | 0.8470 | 0.79 | 0.3515 | 0.8822 | 0.79 | 0.7684 | 0.3075 | 0.0726 |
No log | 39.96 | 240 | 0.8288 | 0.79 | 0.3434 | 0.8192 | 0.79 | 0.7700 | 0.2700 | 0.0648 |
No log | 40.96 | 246 | 0.8255 | 0.8 | 0.3426 | 0.8191 | 0.8000 | 0.7808 | 0.2760 | 0.0727 |
No log | 41.96 | 252 | 0.8247 | 0.8 | 0.3411 | 0.8876 | 0.8000 | 0.7737 | 0.2903 | 0.0701 |
No log | 42.96 | 258 | 0.8196 | 0.8 | 0.3389 | 0.8841 | 0.8000 | 0.7786 | 0.2768 | 0.0727 |
No log | 43.96 | 264 | 0.8118 | 0.805 | 0.3351 | 0.9510 | 0.805 | 0.7806 | 0.2620 | 0.0685 |
No log | 44.96 | 270 | 0.8127 | 0.795 | 0.3352 | 1.0119 | 0.795 | 0.7705 | 0.2650 | 0.0707 |
No log | 45.96 | 276 | 0.7968 | 0.8 | 0.3285 | 1.0041 | 0.8000 | 0.7788 | 0.2734 | 0.0665 |
No log | 46.96 | 282 | 0.7946 | 0.81 | 0.3274 | 1.0647 | 0.81 | 0.7921 | 0.2765 | 0.0703 |
No log | 47.96 | 288 | 0.7996 | 0.805 | 0.3298 | 1.0108 | 0.805 | 0.7867 | 0.2772 | 0.0714 |
No log | 48.96 | 294 | 0.7971 | 0.805 | 0.3283 | 1.0728 | 0.805 | 0.7816 | 0.2756 | 0.0732 |
No log | 49.96 | 300 | 0.7950 | 0.8 | 0.3278 | 1.0694 | 0.8000 | 0.7758 | 0.2540 | 0.0750 |
No log | 50.96 | 306 | 0.7826 | 0.8 | 0.3222 | 1.0211 | 0.8000 | 0.7784 | 0.2596 | 0.0643 |
No log | 51.96 | 312 | 0.7933 | 0.795 | 0.3273 | 1.0680 | 0.795 | 0.7712 | 0.2619 | 0.0764 |
No log | 52.96 | 318 | 0.7883 | 0.805 | 0.3247 | 1.0730 | 0.805 | 0.7834 | 0.2426 | 0.0712 |
No log | 53.96 | 324 | 0.7811 | 0.815 | 0.3219 | 1.0623 | 0.815 | 0.7913 | 0.2259 | 0.0716 |
No log | 54.96 | 330 | 0.7784 | 0.815 | 0.3203 | 1.0657 | 0.815 | 0.7917 | 0.2797 | 0.0690 |
No log | 55.96 | 336 | 0.7827 | 0.81 | 0.3219 | 1.0770 | 0.81 | 0.7885 | 0.2491 | 0.0752 |
No log | 56.96 | 342 | 0.7701 | 0.815 | 0.3166 | 1.0614 | 0.815 | 0.7913 | 0.2664 | 0.0689 |
No log | 57.96 | 348 | 0.7748 | 0.815 | 0.3187 | 1.0699 | 0.815 | 0.7913 | 0.2487 | 0.0722 |
No log | 58.96 | 354 | 0.7669 | 0.815 | 0.3155 | 1.0607 | 0.815 | 0.7919 | 0.2482 | 0.0685 |
No log | 59.96 | 360 | 0.7721 | 0.81 | 0.3180 | 1.0746 | 0.81 | 0.7859 | 0.2385 | 0.0730 |
No log | 60.96 | 366 | 0.7645 | 0.815 | 0.3145 | 1.0650 | 0.815 | 0.7913 | 0.2468 | 0.0688 |
No log | 61.96 | 372 | 0.7672 | 0.815 | 0.3157 | 1.0782 | 0.815 | 0.7913 | 0.2228 | 0.0728 |
No log | 62.96 | 378 | 0.7625 | 0.82 | 0.3139 | 1.0673 | 0.82 | 0.8025 | 0.2323 | 0.0688 |
No log | 63.96 | 384 | 0.7627 | 0.81 | 0.3144 | 1.1893 | 0.81 | 0.7892 | 0.2236 | 0.0710 |
No log | 64.96 | 390 | 0.7629 | 0.815 | 0.3141 | 1.1934 | 0.815 | 0.7972 | 0.2277 | 0.0707 |
No log | 65.96 | 396 | 0.7569 | 0.81 | 0.3118 | 1.1003 | 0.81 | 0.7866 | 0.2577 | 0.0696 |
No log | 66.96 | 402 | 0.7619 | 0.815 | 0.3136 | 1.1365 | 0.815 | 0.7919 | 0.2562 | 0.0732 |
No log | 67.96 | 408 | 0.7565 | 0.815 | 0.3114 | 1.1325 | 0.815 | 0.7919 | 0.2467 | 0.0694 |
No log | 68.96 | 414 | 0.7558 | 0.815 | 0.3117 | 1.1895 | 0.815 | 0.7972 | 0.2453 | 0.0705 |
No log | 69.96 | 420 | 0.7550 | 0.815 | 0.3111 | 1.1924 | 0.815 | 0.7972 | 0.2107 | 0.0709 |
No log | 70.96 | 426 | 0.7573 | 0.805 | 0.3123 | 1.1886 | 0.805 | 0.7795 | 0.2476 | 0.0737 |
No log | 71.96 | 432 | 0.7521 | 0.81 | 0.3099 | 1.1911 | 0.81 | 0.7866 | 0.2117 | 0.0698 |
No log | 72.96 | 438 | 0.7542 | 0.81 | 0.3112 | 1.1878 | 0.81 | 0.7827 | 0.2332 | 0.0726 |
No log | 73.96 | 444 | 0.7509 | 0.815 | 0.3096 | 1.1880 | 0.815 | 0.7899 | 0.2364 | 0.0709 |
No log | 74.96 | 450 | 0.7526 | 0.81 | 0.3105 | 1.1889 | 0.81 | 0.7827 | 0.2453 | 0.0724 |
No log | 75.96 | 456 | 0.7488 | 0.81 | 0.3090 | 1.1869 | 0.81 | 0.7827 | 0.2285 | 0.0699 |
No log | 76.96 | 462 | 0.7506 | 0.815 | 0.3097 | 1.1901 | 0.815 | 0.7934 | 0.2547 | 0.0721 |
No log | 77.96 | 468 | 0.7505 | 0.81 | 0.3098 | 1.1876 | 0.81 | 0.7827 | 0.2110 | 0.0724 |
No log | 78.96 | 474 | 0.7487 | 0.815 | 0.3089 | 1.1885 | 0.815 | 0.7934 | 0.2319 | 0.0715 |
No log | 79.96 | 480 | 0.7472 | 0.81 | 0.3083 | 1.1877 | 0.81 | 0.7827 | 0.2310 | 0.0714 |
No log | 80.96 | 486 | 0.7494 | 0.81 | 0.3094 | 1.1877 | 0.81 | 0.7827 | 0.2462 | 0.0738 |
No log | 81.96 | 492 | 0.7466 | 0.815 | 0.3082 | 1.1888 | 0.815 | 0.7922 | 0.2181 | 0.0709 |
No log | 82.96 | 498 | 0.7467 | 0.81 | 0.3083 | 1.1874 | 0.81 | 0.7827 | 0.2454 | 0.0714 |
0.7129 | 83.96 | 504 | 0.7479 | 0.815 | 0.3088 | 1.1888 | 0.815 | 0.7922 | 0.2272 | 0.0741 |
0.7129 | 84.96 | 510 | 0.7456 | 0.81 | 0.3080 | 1.1853 | 0.81 | 0.7847 | 0.2358 | 0.0719 |
0.7129 | 85.96 | 516 | 0.7465 | 0.815 | 0.3082 | 1.1908 | 0.815 | 0.7922 | 0.2322 | 0.0721 |
0.7129 | 86.96 | 522 | 0.7454 | 0.805 | 0.3081 | 1.1848 | 0.805 | 0.7819 | 0.2262 | 0.0719 |
0.7129 | 87.96 | 528 | 0.7471 | 0.815 | 0.3086 | 1.1894 | 0.815 | 0.7922 | 0.2351 | 0.0741 |
0.7129 | 88.96 | 534 | 0.7459 | 0.815 | 0.3082 | 1.1885 | 0.815 | 0.7922 | 0.2159 | 0.0726 |
0.7129 | 89.96 | 540 | 0.7435 | 0.815 | 0.3072 | 1.1861 | 0.815 | 0.7922 | 0.2291 | 0.0712 |
0.7129 | 90.96 | 546 | 0.7454 | 0.81 | 0.3080 | 1.1876 | 0.81 | 0.7847 | 0.2180 | 0.0733 |
0.7129 | 91.96 | 552 | 0.7461 | 0.815 | 0.3083 | 1.1883 | 0.815 | 0.7942 | 0.2308 | 0.0743 |
0.7129 | 92.96 | 558 | 0.7451 | 0.815 | 0.3079 | 1.1883 | 0.815 | 0.7922 | 0.2330 | 0.0734 |
0.7129 | 93.96 | 564 | 0.7434 | 0.815 | 0.3073 | 1.1863 | 0.815 | 0.7942 | 0.2217 | 0.0720 |
0.7129 | 94.96 | 570 | 0.7446 | 0.815 | 0.3077 | 1.1882 | 0.815 | 0.7942 | 0.2400 | 0.0731 |
0.7129 | 95.96 | 576 | 0.7450 | 0.815 | 0.3079 | 1.1882 | 0.815 | 0.7942 | 0.2144 | 0.0735 |
0.7129 | 96.96 | 582 | 0.7440 | 0.815 | 0.3075 | 1.1871 | 0.815 | 0.7942 | 0.2348 | 0.0731 |
0.7129 | 97.96 | 588 | 0.7441 | 0.815 | 0.3076 | 1.1876 | 0.815 | 0.7942 | 0.2225 | 0.0732 |
0.7129 | 98.96 | 594 | 0.7442 | 0.815 | 0.3076 | 1.1877 | 0.815 | 0.7942 | 0.2072 | 0.0734 |
0.7129 | 99.96 | 600 | 0.7442 | 0.815 | 0.3076 | 1.1877 | 0.815 | 0.7942 | 0.2072 | 0.0734 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.13.1.post200
- Datasets 2.9.0
- Tokenizers 0.13.2