Upload README.md with huggingface_hub
Browse files
README.md
ADDED
|
@@ -0,0 +1,27 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
tags:
|
| 4 |
+
- computer-vision
|
| 5 |
+
- image-classification
|
| 6 |
+
- food101
|
| 7 |
+
- cnn-vit
|
| 8 |
+
---
|
| 9 |
+
|
| 10 |
+
# Hybrid Food Image Classifier (CNN + DeiT)
|
| 11 |
+
|
| 12 |
+
This model combines ResNet50 (CNN) and DeiT-Base (ViT) with an adaptive fusion head.
|
| 13 |
+
|
| 14 |
+
## Files
|
| 15 |
+
- best_model.pth: Trained PyTorch checkpoint
|
| 16 |
+
- real_class_mapping.json: Human-readable class names
|
| 17 |
+
- config.yaml: Project configuration
|
| 18 |
+
- food101_class_names.json: Original class names
|
| 19 |
+
|
| 20 |
+
## Quick Usage
|
| 21 |
+
```python
|
| 22 |
+
from huggingface_hub import hf_hub_download
|
| 23 |
+
ckpt = hf_hub_download(repo_id="codealchemist01/food-image-classifier-hybrid", filename="best_model.pth")
|
| 24 |
+
```
|
| 25 |
+
|
| 26 |
+
## Notes
|
| 27 |
+
Validation accuracy plateaued around ~82.5%. Use best_model for inference.
|