vit-base-augment
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: 0.3257
- Accuracy: 0.9143
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: 64
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 0.678 | 1.0 | 275 | 0.5015 | 0.8624 |
| 0.4558 | 2.0 | 550 | 0.4184 | 0.8859 |
| 0.3474 | 3.0 | 825 | 0.3892 | 0.8907 |
| 0.2925 | 4.0 | 1100 | 0.3692 | 0.8974 |
| 0.2376 | 5.0 | 1375 | 0.3615 | 0.9034 |
| 0.1907 | 6.0 | 1650 | 0.3533 | 0.9046 |
| 0.1605 | 7.0 | 1925 | 0.3385 | 0.9133 |
| 0.138 | 8.0 | 2200 | 0.3296 | 0.9165 |
| 0.1288 | 9.0 | 2475 | 0.3323 | 0.9149 |
| 0.1415 | 10.0 | 2750 | 0.3319 | 0.9165 |
Framework versions
- Transformers 4.39.3
- Pytorch 2.1.2
- Datasets 2.18.0
- Tokenizers 0.15.2