vit_cub
This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the CUB-200-2011 dataset. It achieves the following results on the evaluation set:
- Loss: 1.0122
- Accuracy: 0.7447
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: 0.0002
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 40
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy |
---|---|---|---|---|
0.2462 | 0.33 | 100 | 1.0099 | 0.7573 |
0.3515 | 0.67 | 200 | 0.9446 | 0.7740 |
0.2781 | 1.0 | 300 | 0.9761 | 0.7473 |
0.2071 | 1.33 | 400 | 1.0055 | 0.7431 |
0.223 | 1.67 | 500 | 0.9942 | 0.7531 |
0.1899 | 2.0 | 600 | 1.0806 | 0.7331 |
0.1568 | 2.33 | 700 | 1.1502 | 0.7156 |
Framework versions
- Transformers 4.38.2
- Pytorch 2.2.1+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2
- Downloads last month
- 88