vit-base-patch16-224-in21k
This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.1026
- Accuracy: 0.982
Model description
This model is a fine-tuned version of google/vit-base-patch16-224-in21k which discriminates cats from dogs.
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: 0.0002
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 4
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 0.177 | 0.5 | 500 | 0.2100 | 0.9435 |
| 0.1515 | 1.0 | 1000 | 0.0710 | 0.975 |
| 0.0443 | 1.5 | 1500 | 0.2043 | 0.9535 |
| 0.0625 | 2.0 | 2000 | 0.0898 | 0.9745 |
| 0.0181 | 2.5 | 2500 | 0.0961 | 0.9805 |
| 0.0091 | 3.0 | 3000 | 0.1049 | 0.982 |
| 0.0016 | 3.5 | 3500 | 0.1066 | 0.981 |
| 0.0015 | 4.0 | 4000 | 0.1026 | 0.982 |
Framework versions
- Transformers 4.25.1
- Pytorch 1.13.0+cu116
- Datasets 2.8.0
- Tokenizers 0.13.2
