vit_fold_1
This model is a fine-tuned version of google/vit-base-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 1.0952
- Accuracy: 0.9690
- F1 Score: 0.9709
- Recall: 0.9694
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: 1e-05
- train_batch_size: 64
- eval_batch_size: 64
- seed: 42
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 150
- num_epochs: 100
- label_smoothing_factor: 0.15
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 Score | Recall |
|---|---|---|---|---|---|---|
| 2.9282 | 1.0 | 19 | 2.7689 | 0.3310 | 0.2729 | 0.2884 |
| 2.7470 | 2.0 | 38 | 2.5989 | 0.4448 | 0.3540 | 0.3755 |
| 2.4449 | 3.0 | 57 | 2.3265 | 0.5931 | 0.5856 | 0.5613 |
| 2.0224 | 4.0 | 76 | 2.0013 | 0.7379 | 0.7505 | 0.7314 |
| 1.6546 | 5.0 | 95 | 1.6556 | 0.8276 | 0.8368 | 0.8288 |
| 1.3650 | 6.0 | 114 | 1.3947 | 0.9 | 0.9044 | 0.9030 |
| 1.1791 | 7.0 | 133 | 1.2692 | 0.9207 | 0.9228 | 0.9237 |
| 1.1519 | 8.0 | 152 | 1.2026 | 0.9414 | 0.9428 | 0.9398 |
| 1.1252 | 9.0 | 171 | 1.1660 | 0.9448 | 0.9465 | 0.9448 |
| 1.0789 | 10.0 | 190 | 1.1463 | 0.9483 | 0.9501 | 0.9497 |
| 1.0771 | 11.0 | 209 | 1.1329 | 0.9517 | 0.9533 | 0.9509 |
| 1.0519 | 12.0 | 228 | 1.1264 | 0.9552 | 0.9569 | 0.9533 |
| 1.0340 | 13.0 | 247 | 1.1138 | 0.9586 | 0.9610 | 0.9619 |
| 1.0224 | 14.0 | 266 | 1.1073 | 0.9655 | 0.9666 | 0.9644 |
| 0.9936 | 15.0 | 285 | 1.1274 | 0.9586 | 0.9593 | 0.9558 |
| 0.9975 | 16.0 | 304 | 1.1038 | 0.9655 | 0.9666 | 0.9644 |
| 0.9908 | 17.0 | 323 | 1.1107 | 0.9621 | 0.9631 | 0.9595 |
| 0.9870 | 18.0 | 342 | 1.1084 | 0.9655 | 0.9663 | 0.9631 |
| 0.9862 | 19.0 | 361 | 1.1079 | 0.9655 | 0.9663 | 0.9631 |
| 0.9739 | 20.0 | 380 | 1.0912 | 0.9655 | 0.9666 | 0.9644 |
| 0.9810 | 21.0 | 399 | 1.0898 | 0.9690 | 0.9698 | 0.9680 |
| 0.9865 | 22.0 | 418 | 1.0878 | 0.9655 | 0.9668 | 0.9656 |
| 0.9694 | 23.0 | 437 | 1.0842 | 0.9655 | 0.9668 | 0.9656 |
| 0.9649 | 24.0 | 456 | 1.0958 | 0.9655 | 0.9666 | 0.9644 |
| 0.9710 | 25.0 | 475 | 1.0934 | 0.9690 | 0.9698 | 0.9680 |
| 0.9672 | 26.0 | 494 | 1.0923 | 0.9655 | 0.9666 | 0.9644 |
| 0.9653 | 27.0 | 513 | 1.0932 | 0.9690 | 0.9698 | 0.9680 |
| 0.9622 | 28.0 | 532 | 1.0820 | 0.9655 | 0.9669 | 0.9668 |
| 0.9677 | 29.0 | 551 | 1.0928 | 0.9655 | 0.9666 | 0.9644 |
| 0.9639 | 30.0 | 570 | 1.0834 | 0.9655 | 0.9666 | 0.9644 |
| 0.9671 | 31.0 | 589 | 1.0952 | 0.9690 | 0.9709 | 0.9694 |
| 0.9635 | 32.0 | 608 | 1.0909 | 0.9655 | 0.9682 | 0.9694 |
| 0.9710 | 33.0 | 627 | 1.0919 | 0.9690 | 0.9709 | 0.9694 |
| 0.9614 | 34.0 | 646 | 1.0951 | 0.9655 | 0.9682 | 0.9694 |
| 0.9637 | 35.0 | 665 | 1.0940 | 0.9690 | 0.9709 | 0.9694 |
| 0.9618 | 36.0 | 684 | 1.0847 | 0.9690 | 0.9700 | 0.9693 |
| 0.9578 | 37.0 | 703 | 1.0871 | 0.9655 | 0.9668 | 0.9656 |
| 0.9646 | 38.0 | 722 | 1.0910 | 0.9655 | 0.9671 | 0.9680 |
Framework versions
- Transformers 5.0.0
- Pytorch 2.10.0+cu128
- Datasets 5.0.0
- Tokenizers 0.22.2