swin-small-patch4-window7-224-finetuned-isic217
This model is a fine-tuned version of microsoft/swin-small-patch4-window7-224 on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 1.9417
- Accuracy: 0.5455
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: 2
- eval_batch_size: 2
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 8
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 30
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 2.1844 | 0.9796 | 24 | 2.1103 | 0.1364 |
| 2.0018 | 2.0 | 49 | 1.8737 | 0.2727 |
| 1.6474 | 2.9796 | 73 | 1.9019 | 0.2727 |
| 1.3757 | 4.0 | 98 | 1.7487 | 0.3636 |
| 1.1526 | 4.9796 | 122 | 1.7576 | 0.4091 |
| 0.9161 | 6.0 | 147 | 1.5886 | 0.5 |
| 0.7568 | 6.9796 | 171 | 1.8935 | 0.4545 |
| 0.4024 | 8.0 | 196 | 1.6767 | 0.4545 |
| 0.814 | 8.9796 | 220 | 1.7112 | 0.3636 |
| 0.4346 | 10.0 | 245 | 1.9364 | 0.4091 |
| 0.3456 | 10.9796 | 269 | 1.9417 | 0.5455 |
| 0.228 | 12.0 | 294 | 2.1569 | 0.4091 |
| 0.1681 | 12.9796 | 318 | 2.0565 | 0.4545 |
| 0.1498 | 14.0 | 343 | 2.0701 | 0.3636 |
| 0.1599 | 14.9796 | 367 | 2.4973 | 0.5 |
| 0.3856 | 16.0 | 392 | 2.2473 | 0.4545 |
| 0.2529 | 16.9796 | 416 | 2.0918 | 0.4545 |
| 0.0557 | 18.0 | 441 | 1.9596 | 0.5455 |
| 0.0895 | 18.9796 | 465 | 2.5522 | 0.4545 |
| 0.0719 | 20.0 | 490 | 2.2938 | 0.5 |
| 0.0764 | 20.9796 | 514 | 2.6754 | 0.4545 |
| 0.1301 | 22.0 | 539 | 2.5287 | 0.4545 |
| 0.1205 | 22.9796 | 563 | 2.7532 | 0.4091 |
| 0.1013 | 24.0 | 588 | 2.6988 | 0.4545 |
| 0.0777 | 24.9796 | 612 | 2.9345 | 0.4091 |
| 0.1807 | 26.0 | 637 | 2.9981 | 0.4545 |
| 0.0298 | 26.9796 | 661 | 2.8549 | 0.4545 |
| 0.0589 | 28.0 | 686 | 2.6967 | 0.4545 |
| 0.0896 | 28.9796 | 710 | 2.6903 | 0.4545 |
| 0.0218 | 29.3878 | 720 | 2.6902 | 0.4545 |
Framework versions
- Transformers 4.42.0.dev0
- Pytorch 2.3.0+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1