convnext-base-3e-5
This model is a fine-tuned version of facebook/convnextv2-base-22k-384 on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 0.2406
- Accuracy: 0.9391
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: 3e-05
- train_batch_size: 16
- eval_batch_size: 16
- 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 |
|---|---|---|---|---|
| 0.4903 | 1.0 | 1099 | 0.3690 | 0.8911 |
| 0.3615 | 2.0 | 2198 | 0.2986 | 0.9169 |
| 0.2686 | 3.0 | 3297 | 0.2662 | 0.9304 |
| 0.246 | 4.0 | 4396 | 0.2340 | 0.9396 |
| 0.201 | 5.0 | 5495 | 0.2588 | 0.9412 |
| 0.1757 | 6.0 | 6594 | 0.2436 | 0.9439 |
| 0.1699 | 7.0 | 7693 | 0.2558 | 0.9443 |
| 0.1178 | 8.0 | 8792 | 0.2512 | 0.9479 |
| 0.1167 | 9.0 | 9891 | 0.2536 | 0.9479 |
| 0.0889 | 10.0 | 10990 | 0.2536 | 0.9495 |
Framework versions
- Transformers 4.39.3
- Pytorch 2.1.2
- Datasets 2.18.0
- Tokenizers 0.15.2