SL-CvT
This model is a fine-tuned version of microsoft/cvt-13 on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 0.3430
- F1: 0.9298
- Roc Auc: 0.9777
- Accuracy: 0.9317
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: 5e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- gradient_accumulation_steps: 4
- 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 | F1 | Roc Auc | Accuracy |
---|---|---|---|---|---|---|
1.2379 | 1.0 | 60 | 1.0716 | 0.6422 | 0.7323 | 0.7246 |
1.0186 | 2.0 | 120 | 0.8477 | 0.6425 | 0.7879 | 0.7293 |
0.9433 | 3.0 | 180 | 0.7473 | 0.7060 | 0.8454 | 0.7538 |
0.8644 | 4.0 | 240 | 0.6831 | 0.7188 | 0.8696 | 0.7663 |
0.7985 | 5.0 | 300 | 0.6420 | 0.7409 | 0.8943 | 0.7799 |
0.7322 | 6.0 | 360 | 0.5713 | 0.7886 | 0.9196 | 0.8101 |
0.725 | 7.0 | 420 | 0.5311 | 0.7989 | 0.9324 | 0.8190 |
0.6529 | 8.0 | 480 | 0.5246 | 0.7852 | 0.9404 | 0.8117 |
0.6224 | 9.0 | 540 | 0.4598 | 0.8282 | 0.9517 | 0.8440 |
0.6315 | 10.0 | 600 | 0.4363 | 0.8457 | 0.9585 | 0.8529 |
0.5651 | 11.0 | 660 | 0.4437 | 0.8323 | 0.9564 | 0.8503 |
0.574 | 12.0 | 720 | 0.4003 | 0.8531 | 0.9617 | 0.8638 |
0.5269 | 13.0 | 780 | 0.3901 | 0.8676 | 0.9671 | 0.8722 |
0.5138 | 14.0 | 840 | 0.3984 | 0.8607 | 0.9685 | 0.8732 |
0.4839 | 15.0 | 900 | 0.3763 | 0.8683 | 0.9701 | 0.8769 |
0.463 | 16.0 | 960 | 0.3398 | 0.8837 | 0.9718 | 0.8894 |
0.4767 | 17.0 | 1020 | 0.3293 | 0.8846 | 0.9738 | 0.8915 |
0.4985 | 18.0 | 1080 | 0.3350 | 0.8852 | 0.9763 | 0.8863 |
0.4657 | 19.0 | 1140 | 0.3369 | 0.8872 | 0.9746 | 0.8951 |
0.4514 | 20.0 | 1200 | 0.3213 | 0.8880 | 0.9750 | 0.8925 |
0.4207 | 21.0 | 1260 | 0.3175 | 0.8943 | 0.9771 | 0.8978 |
0.4522 | 22.0 | 1320 | 0.3229 | 0.8970 | 0.9767 | 0.8983 |
0.4328 | 23.0 | 1380 | 0.3121 | 0.8948 | 0.9791 | 0.8978 |
0.3942 | 24.0 | 1440 | 0.3111 | 0.8993 | 0.9765 | 0.9030 |
0.4414 | 25.0 | 1500 | 0.3062 | 0.9032 | 0.9763 | 0.9061 |
0.3608 | 26.0 | 1560 | 0.3099 | 0.8997 | 0.9787 | 0.9014 |
0.3729 | 27.0 | 1620 | 0.3050 | 0.9029 | 0.9783 | 0.9082 |
0.393 | 28.0 | 1680 | 0.2970 | 0.9090 | 0.9797 | 0.9108 |
0.402 | 29.0 | 1740 | 0.2986 | 0.9087 | 0.9793 | 0.9113 |
0.3697 | 30.0 | 1800 | 0.3384 | 0.8968 | 0.9769 | 0.9025 |
0.3502 | 31.0 | 1860 | 0.3035 | 0.9058 | 0.9789 | 0.9103 |
0.3653 | 32.0 | 1920 | 0.3127 | 0.9024 | 0.9788 | 0.9025 |
0.3898 | 33.0 | 1980 | 0.3222 | 0.9050 | 0.9778 | 0.9061 |
0.317 | 34.0 | 2040 | 0.3013 | 0.9124 | 0.9798 | 0.9139 |
0.3166 | 35.0 | 2100 | 0.3185 | 0.9095 | 0.9775 | 0.9134 |
0.3771 | 36.0 | 2160 | 0.3067 | 0.9049 | 0.9782 | 0.9066 |
0.3487 | 37.0 | 2220 | 0.2948 | 0.9118 | 0.9801 | 0.9134 |
0.3202 | 38.0 | 2280 | 0.2916 | 0.9168 | 0.9788 | 0.9186 |
0.3163 | 39.0 | 2340 | 0.3149 | 0.9141 | 0.9777 | 0.9155 |
0.3605 | 40.0 | 2400 | 0.2964 | 0.9192 | 0.9797 | 0.9207 |
0.3636 | 41.0 | 2460 | 0.3142 | 0.9111 | 0.9810 | 0.9134 |
0.3454 | 42.0 | 2520 | 0.3133 | 0.9111 | 0.9792 | 0.9113 |
0.3561 | 43.0 | 2580 | 0.3090 | 0.9073 | 0.9804 | 0.9077 |
0.3136 | 44.0 | 2640 | 0.3236 | 0.9144 | 0.9782 | 0.9176 |
0.3529 | 45.0 | 2700 | 0.3054 | 0.9175 | 0.9800 | 0.9202 |
0.2987 | 46.0 | 2760 | 0.2944 | 0.9222 | 0.9802 | 0.9233 |
0.2966 | 47.0 | 2820 | 0.3215 | 0.9201 | 0.9786 | 0.9233 |
0.3203 | 48.0 | 2880 | 0.3150 | 0.9219 | 0.9797 | 0.9244 |
0.2821 | 49.0 | 2940 | 0.3072 | 0.9273 | 0.9800 | 0.9291 |
0.2852 | 50.0 | 3000 | 0.3265 | 0.9155 | 0.9792 | 0.9176 |
0.3544 | 51.0 | 3060 | 0.3175 | 0.9150 | 0.9802 | 0.9150 |
0.3327 | 52.0 | 3120 | 0.3134 | 0.9222 | 0.9802 | 0.9244 |
0.2877 | 53.0 | 3180 | 0.3222 | 0.9154 | 0.9805 | 0.9165 |
0.3089 | 54.0 | 3240 | 0.3045 | 0.9248 | 0.9811 | 0.9259 |
0.2904 | 55.0 | 3300 | 0.3301 | 0.9175 | 0.9787 | 0.9186 |
0.2821 | 56.0 | 3360 | 0.3069 | 0.9206 | 0.9810 | 0.9218 |
0.321 | 57.0 | 3420 | 0.3209 | 0.9254 | 0.9800 | 0.9270 |
0.2995 | 58.0 | 3480 | 0.3281 | 0.9202 | 0.9802 | 0.9233 |
0.2683 | 59.0 | 3540 | 0.3263 | 0.9174 | 0.9802 | 0.9202 |
0.3021 | 60.0 | 3600 | 0.3484 | 0.9170 | 0.9788 | 0.9186 |
0.3262 | 61.0 | 3660 | 0.3270 | 0.9151 | 0.9807 | 0.9165 |
0.2329 | 62.0 | 3720 | 0.3280 | 0.9211 | 0.9807 | 0.9233 |
0.2935 | 63.0 | 3780 | 0.3296 | 0.9244 | 0.9807 | 0.9264 |
0.2856 | 64.0 | 3840 | 0.3323 | 0.9209 | 0.9811 | 0.9218 |
0.2829 | 65.0 | 3900 | 0.3390 | 0.9200 | 0.9802 | 0.9218 |
0.3044 | 66.0 | 3960 | 0.3324 | 0.9215 | 0.9799 | 0.9228 |
0.2767 | 67.0 | 4020 | 0.3496 | 0.9150 | 0.9778 | 0.9160 |
0.2936 | 68.0 | 4080 | 0.3378 | 0.9257 | 0.9790 | 0.9275 |
0.2884 | 69.0 | 4140 | 0.3493 | 0.9227 | 0.9790 | 0.9249 |
0.2906 | 70.0 | 4200 | 0.3408 | 0.9259 | 0.9794 | 0.9275 |
0.2542 | 71.0 | 4260 | 0.3559 | 0.9233 | 0.9769 | 0.9249 |
0.2557 | 72.0 | 4320 | 0.3481 | 0.9237 | 0.9779 | 0.9254 |
0.2266 | 73.0 | 4380 | 0.3518 | 0.9208 | 0.9781 | 0.9223 |
0.2771 | 74.0 | 4440 | 0.3544 | 0.9231 | 0.9776 | 0.9254 |
0.2747 | 75.0 | 4500 | 0.3469 | 0.9270 | 0.9780 | 0.9285 |
0.2443 | 76.0 | 4560 | 0.3513 | 0.9216 | 0.9767 | 0.9233 |
0.2859 | 77.0 | 4620 | 0.3456 | 0.9234 | 0.9771 | 0.9254 |
0.2677 | 78.0 | 4680 | 0.3474 | 0.9239 | 0.9780 | 0.9254 |
0.2492 | 79.0 | 4740 | 0.3513 | 0.9235 | 0.9778 | 0.9254 |
0.2532 | 80.0 | 4800 | 0.3524 | 0.9210 | 0.9773 | 0.9233 |
0.2646 | 81.0 | 4860 | 0.3529 | 0.9240 | 0.9784 | 0.9238 |
0.2842 | 82.0 | 4920 | 0.3433 | 0.9260 | 0.9777 | 0.9280 |
0.2872 | 83.0 | 4980 | 0.3584 | 0.9272 | 0.9771 | 0.9285 |
0.2678 | 84.0 | 5040 | 0.3430 | 0.9298 | 0.9777 | 0.9317 |
0.2705 | 85.0 | 5100 | 0.3534 | 0.9268 | 0.9777 | 0.9291 |
0.2605 | 86.0 | 5160 | 0.3574 | 0.9272 | 0.9777 | 0.9296 |
0.2572 | 87.0 | 5220 | 0.3426 | 0.9273 | 0.9781 | 0.9291 |
0.2646 | 88.0 | 5280 | 0.3472 | 0.9234 | 0.9789 | 0.9244 |
0.2831 | 89.0 | 5340 | 0.3433 | 0.9272 | 0.9779 | 0.9291 |
0.277 | 90.0 | 5400 | 0.3441 | 0.9263 | 0.9789 | 0.9280 |
0.2584 | 91.0 | 5460 | 0.3432 | 0.9236 | 0.9788 | 0.9249 |
0.2703 | 92.0 | 5520 | 0.3409 | 0.9248 | 0.9789 | 0.9259 |
0.2811 | 93.0 | 5580 | 0.3449 | 0.9215 | 0.9795 | 0.9228 |
0.2786 | 94.0 | 5640 | 0.3465 | 0.9260 | 0.9789 | 0.9280 |
0.267 | 95.0 | 5700 | 0.3472 | 0.9260 | 0.9791 | 0.9275 |
0.2695 | 96.0 | 5760 | 0.3500 | 0.9268 | 0.9786 | 0.9285 |
0.279 | 97.0 | 5820 | 0.3582 | 0.9249 | 0.9782 | 0.9270 |
0.2774 | 98.0 | 5880 | 0.3486 | 0.9251 | 0.9790 | 0.9270 |
0.2512 | 99.0 | 5940 | 0.3514 | 0.9287 | 0.9786 | 0.9306 |
0.2218 | 100.0 | 6000 | 0.3482 | 0.9269 | 0.9789 | 0.9285 |
Framework versions
- Transformers 4.29.2
- Pytorch 2.0.0+cu118
- Datasets 2.12.0
- Tokenizers 0.13.3
- Downloads last month
- 20
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.
Evaluation results
- F1 on imagefolderself-reported0.930
- Accuracy on imagefolderself-reported0.932