vit-base-patch16-224-finalterm
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: 0.3547
- Accuracy: 0.8812
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: 15
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 1.3999 | 1.0 | 10 | 1.1607 | 0.5094 |
| 0.993 | 2.0 | 20 | 0.7807 | 0.7031 |
| 0.6819 | 3.0 | 30 | 0.5753 | 0.8063 |
| 0.5485 | 4.0 | 40 | 0.6475 | 0.7594 |
| 0.463 | 5.0 | 50 | 0.4393 | 0.8406 |
| 0.3929 | 6.0 | 60 | 0.4067 | 0.8625 |
| 0.3636 | 7.0 | 70 | 0.3626 | 0.8875 |
| 0.3719 | 8.0 | 80 | 0.3613 | 0.8875 |
| 0.343 | 9.0 | 90 | 0.3624 | 0.8781 |
| 0.3297 | 10.0 | 100 | 0.3800 | 0.8625 |
| 0.2948 | 11.0 | 110 | 0.3320 | 0.8938 |
| 0.33 | 12.0 | 120 | 0.3481 | 0.8781 |
| 0.3281 | 13.0 | 130 | 0.3418 | 0.8875 |
| 0.3 | 14.0 | 140 | 0.3425 | 0.8844 |
| 0.3014 | 15.0 | 150 | 0.3547 | 0.8812 |
Framework versions
- Transformers 4.41.2
- Pytorch 2.3.0+cu121
- Datasets 2.19.2
- Tokenizers 0.19.1