swin-base-patch4-window7-224-in22k-finetuned-brain-tumor-skullStrippded_04
This model is a fine-tuned version of microsoft/swin-base-patch4-window7-224-in22k on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 0.0470
- Accuracy: 0.9843
- F1 Score: 0.9844
- Precision: 0.9846
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: 0.0001
- train_batch_size: 100
- eval_batch_size: 100
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 400
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Accuracy | F1 Score | Validation Loss | Precision |
|---|---|---|---|---|---|---|
| 1.2662 | 1.0 | 16 | 0.8370 | 0.8309 | 0.4424 | 0.8464 |
| 0.3778 | 0.98 | 20 | 0.2700 | 0.9062 | 0.9067 | 0.9072 |
| 0.2377 | 2.0 | 41 | 0.2035 | 0.9229 | 0.9234 | 0.9269 |
| 0.1201 | 2.98 | 61 | 0.1345 | 0.9465 | 0.9467 | 0.9512 |
| 0.0774 | 4.0 | 82 | 0.1229 | 0.9612 | 0.9618 | 0.9643 |
| 0.0495 | 4.98 | 102 | 0.0562 | 0.9813 | 0.9815 | 0.9816 |
| 0.0358 | 6.0 | 123 | 0.0470 | 0.9843 | 0.9844 | 0.9846 |
| 0.0228 | 6.98 | 143 | 0.0447 | 0.9833 | 0.9834 | 0.9836 |
| 0.0181 | 8.0 | 164 | 0.0465 | 0.9828 | 0.9830 | 0.9831 |
| 0.0132 | 8.98 | 184 | 0.0436 | 0.9833 | 0.9835 | 0.9836 |
| 0.0126 | 9.76 | 200 | 0.0461 | 0.9838 | 0.9840 | 0.9840 |
Framework versions
- Transformers 4.27.4
- Pytorch 2.0.0+cu117
- Datasets 2.11.0
- Tokenizers 0.13.3