vit-tiny-patch16-224-finetuned-papsmear
This model is a fine-tuned version of WinKawaks/vit-tiny-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 0.2121
- Accuracy: 0.9338
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: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 15
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 1.4005 | 0.9935 | 38 | 1.2214 | 0.5294 |
| 0.8877 | 1.9869 | 76 | 1.0727 | 0.6691 |
| 0.603 | 2.9804 | 114 | 0.6807 | 0.7574 |
| 0.465 | 4.0 | 153 | 0.6485 | 0.7574 |
| 0.432 | 4.9935 | 191 | 0.5024 | 0.8015 |
| 0.2957 | 5.9869 | 229 | 0.4485 | 0.8162 |
| 0.2203 | 6.9804 | 267 | 0.3850 | 0.8529 |
| 0.236 | 8.0 | 306 | 0.3628 | 0.8456 |
| 0.1857 | 8.9935 | 344 | 0.2930 | 0.8824 |
| 0.1907 | 9.9869 | 382 | 0.2121 | 0.9338 |
| 0.1546 | 10.9804 | 420 | 0.2242 | 0.9265 |
| 0.1375 | 12.0 | 459 | 0.1918 | 0.9191 |
| 0.1237 | 12.9935 | 497 | 0.1809 | 0.9338 |
| 0.1637 | 13.9869 | 535 | 0.1774 | 0.9338 |
| 0.0803 | 14.9020 | 570 | 0.1882 | 0.9338 |
Framework versions
- Transformers 4.44.2
- Pytorch 2.4.0+cu121
- Datasets 3.0.0
- Tokenizers 0.19.1