swin-tiny-patch4-window7-224-ve-U11-b-12
This model is a fine-tuned version of microsoft/swin-tiny-patch4-window7-224 on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 0.9473
- Accuracy: 0.5435
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: 12
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| No log | 0.92 | 6 | 1.3839 | 0.1304 |
| 1.3821 | 2.0 | 13 | 1.3524 | 0.2391 |
| 1.3821 | 2.92 | 19 | 1.2898 | 0.3043 |
| 1.2875 | 4.0 | 26 | 1.1721 | 0.4348 |
| 1.1072 | 4.92 | 32 | 1.1018 | 0.4348 |
| 1.1072 | 6.0 | 39 | 1.0327 | 0.4783 |
| 0.9941 | 6.92 | 45 | 0.9920 | 0.4565 |
| 0.9132 | 8.0 | 52 | 0.9473 | 0.5435 |
| 0.9132 | 8.92 | 58 | 0.9522 | 0.5217 |
| 0.849 | 10.0 | 65 | 0.9478 | 0.5217 |
| 0.8124 | 10.92 | 71 | 0.9506 | 0.5217 |
| 0.8124 | 11.08 | 72 | 0.9505 | 0.5217 |
Framework versions
- Transformers 4.36.2
- Pytorch 2.1.2+cu118
- Datasets 2.16.1
- Tokenizers 0.15.0