my_food_model
This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the food101 dataset. It achieves the following results on the evaluation set:
- Loss: 0.4724
- Accuracy: 0.933
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: 16
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 64
- 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 | Validation Loss | Accuracy |
|---|---|---|---|---|
| 3.262 | 0.99 | 62 | 3.0234 | 0.775 |
| 1.8087 | 2.0 | 125 | 1.6821 | 0.853 |
| 1.2098 | 2.99 | 187 | 1.1582 | 0.91 |
| 0.8913 | 4.0 | 250 | 0.8991 | 0.92 |
| 0.7071 | 4.99 | 312 | 0.7349 | 0.912 |
| 0.5607 | 6.0 | 375 | 0.6197 | 0.921 |
| 0.4785 | 6.99 | 437 | 0.5506 | 0.929 |
| 0.3926 | 8.0 | 500 | 0.5015 | 0.93 |
| 0.3906 | 8.99 | 562 | 0.4902 | 0.927 |
| 0.3866 | 9.92 | 620 | 0.4724 | 0.933 |
Framework versions
- Transformers 4.28.0
- Pytorch 2.0.1+cu118
- Datasets 2.13.0
- Tokenizers 0.13.3