AI_Agent_Instructions.md

AI AGENT TASK: Upgrade Fashion MNIST Repository

TASK SUMMARY

Transform https://github.com/millenniumsingha/StyleNet from a basic ML notebook into a production-ready application with CNN model, REST API, and web interface.

EXECUTION STEPS

PHASE 1: Setup Structure

# Create directories
mkdir -p src api app tests models notebooks legacy

# Move original notebook
mv StyleNet.ipynb legacy/

# Create placeholder files
touch src/__init__.py api/__init__.py tests/__init__.py models/.gitkeep

PHASE 2: Create Core Files

Create these files with contents from the main guide (FashionMNIST_Upgrade_Guide.md):

Priority 1 (Required for functionality):

1. requirements.txt
2. src/config.py
3. src/model.py
4. src/train.py
5. src/predict.py
6. api/main.py
7. api/schemas.py
8. app/streamlit_app.py

Priority 2 (Required for deployment): 9. Dockerfile 10. docker-compose.yml 11. .gitignore 12. .dockerignore 13. README.md

Priority 3 (Quality assurance): 14. tests/test_model.py 15. tests/test_api.py

PHASE 3: Validate

# Install dependencies
pip install -r requirements.txt

# Train model (should achieve ~92% accuracy)
python -m src.train --model cnn --epochs 15

# Run tests
pytest tests/ -v

# Test API
uvicorn api.main:app --port 8000 &
curl http://localhost:8000/health

# Test Streamlit
streamlit run app/streamlit_app.py --server.port 8501

PHASE 4: Commit

git add .
git commit -m "Upgrade to production-ready ML application

- Added CNN model architecture (92%+ accuracy)
- Added FastAPI REST API with /predict endpoint
- Added Streamlit web interface
- Added Docker containerization
- Added comprehensive tests
- Restructured project for production use
- Archived original notebook in legacy/"

git push origin master

EXPECTED FINAL STRUCTURE

StyleNet/
├── README.md
├── requirements.txt
├── Dockerfile
├── docker-compose.yml
├── .gitignore
├── .dockerignore
├── src/
│   ├── __init__.py
│   ├── config.py
│   ├── model.py
│   ├── train.py
│   └── predict.py
├── api/
│   ├── __init__.py
│   ├── main.py
│   └── schemas.py
├── app/
│   └── streamlit_app.py
├── tests/
│   ├── __init__.py
│   ├── test_model.py
│   └── test_api.py
├── models/
│   └── .gitkeep
├── notebooks/
├── images/
│   └── (keep existing)
└── legacy/
    └── StyleNet.ipynb

SUCCESS CRITERIA

• Model trains to 92%+ test accuracy
• API responds at http://localhost:8000/docs
• Streamlit loads at http://localhost:8501
• All tests pass
• Docker builds successfully
• README renders correctly on GitHub

REFERENCE

See FashionMNIST_Upgrade_Guide.md for complete file contents.