README.md

---
license: mit
pipeline_tag: image-classification
tags:
- image-classification
- pytorch
- dog-breed
- mobilenet
- computer-vision
datasets:
- custom
library_name: transformers
---

# Dog Breed Classifier

This is a MobileNetV2 model fine-tuned for dog breed classification with 120 breed classes.

## Supported Breeds

- Chihuahua
- Japanese Spaniel
- Maltese Dog
- Pekinese
- Shih-Tzu
- Blenheim Spaniel
- Papillon
- Toy Terrier
- Rhodesian Ridgeback
- Afghan Hound
- ... and 110 more breeds

## Model Details

- **Architecture**: MobileNetV2 with custom classifier
- **Input Size**: 224x224 pixels
- **Number of Classes**: 120
- **Framework**: PyTorch

## Usage

### Via Hugging Face Inference API

```python
import requests

API_URL = "https://api-inference.huggingface.co/models/valentinocc/dog-breed-classifier"
headers = {"Authorization": "Bearer YOUR_HF_TOKEN"}

def query(filename):
    with open(filename, "rb") as f:
        data = f.read()
    response = requests.post(API_URL, headers=headers, data=data)
    return response.json()

# Predict dog breed
output = query("dog_image.jpg")
print(f"Predicted breed: {output[0]['label']} ({output[0]['score']:.2%} confidence)")
```

### Django Integration

```python
from utils.huggingface_client import HuggingFaceClient

hf_client = HuggingFaceClient()
result = hf_client.classify_image(image_file, "valentinocc/dog-breed-classifier")
```

## Training Details

This model was trained on a custom dataset of dog breed images using transfer learning on MobileNetV2.
The model achieves good performance across 120 different dog breeds.

## Model Performance

- Input: RGB images, 224x224 pixels
- Output: Probability distribution over 120 dog breeds
- Architecture: MobileNetV2 backbone + custom classifier