convnext-base
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: 0.2818
- Accuracy: 0.9442
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: 0.0001
- 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 |
|---|---|---|---|---|
| 0.5947 | 1.0 | 1099 | 0.4242 | 0.8684 |
| 0.4798 | 2.0 | 2198 | 0.4242 | 0.8728 |
| 0.3625 | 3.0 | 3297 | 0.3553 | 0.9078 |
| 0.2777 | 4.0 | 4396 | 0.3241 | 0.9185 |
| 0.2368 | 5.0 | 5495 | 0.3413 | 0.9245 |
| 0.1635 | 6.0 | 6594 | 0.3116 | 0.9356 |
| 0.1564 | 7.0 | 7693 | 0.2997 | 0.9360 |
| 0.1082 | 8.0 | 8792 | 0.2916 | 0.9451 |
| 0.1146 | 9.0 | 9891 | 0.2963 | 0.9431 |
| 0.0801 | 10.0 | 10990 | 0.2946 | 0.9439 |
Framework versions
- Transformers 4.39.3
- Pytorch 2.1.2
- Datasets 2.18.0
- Tokenizers 0.15.2