Add model card with metadata for inference API

# GoVed-AI: Cattle Breed Classification Model

This model is a PyTorch-based ResNet-18 image classification model specifically designed for identifying cattle breeds. The model can classify between 41 different cattle breeds, including both Indian breeds (like Gir, Sahiwal, Tharparkar) and international breeds (like Holstein Friesian, Jersey, Brown Swiss).

## Model Details

- **Architecture**: ResNet-18
- **Framework**: PyTorch
- **Task**: Image Classification
- **Number of Classes**: 41 cattle breeds
- **Input**: RGB images (224x224 pixels)
- **Output**: Breed classification with confidence scores

## Supported Breeds

The model can classify the following cattle breeds:
- Alambadi, Amritmahal, Ayrshire, Banni, Bargur, Bhadawari, Brown Swiss
- Dangi, Deoni, Gir, Guernsey, Hallikar, Hariana, Holstein Friesian
- Jaffrabadi, Jersey, Kangayam, Kankrej, Kasargod, Kenkatha, Kherigarh
- Khillari, Krishna Valley, Malnad Gidda, Mehsana, Murrah, Nagori, Nagpuri
- Nili Ravi, Nimari, Ongole, Pulikulam, Rathi, Red Dane, Red Sindhi
- Sahiwal, Surti, Tharparkar, Toda, Umblachery, Vechur

## Usage

This model includes custom inference code and can be used for cattle breed identification in agricultural applications, livestock management systems, and veterinary practices.

## Files

- `bovine_model.pth`: Pre-trained model weights
- `inference.py`: Custom inference script
- `model_config.json`: Model configuration
- `requirements.txt`: Dependencies (torch, torchvision, Pillow)

README.md

@@ -0,0 +1,3 @@
+---
+pipeline_tag: image-classification
+---