ML Model CI/CD Pipeline

This repository demonstrates a simple CI/CD pipeline for a Machine Learning project using GitHub Actions. The project includes a Convolutional Neural Network (CNN) model trained on the MNIST dataset, with automated testing and model validation.

Project Structure

├── .github/ │ └── workflows/ │ └── ml-pipeline.yml ├── model/ │ ├── init.py │ └── network.py ├── tests/ │ └── test_model.py ├── .gitignore ├── README.md ├── train.py └── requirements.txt

Features

• Simple CNN architecture for MNIST digit classification
• Automated model training
• Model validation and testing
• GitHub Actions integration for CI/CD
• CPU-only implementation for better compatibility

Model Architecture

The model is a simple CNN with:

• 2 convolutional layers
• 2 max-pooling layers
• 2 fully connected layers
• Less than 25,000 parameters
• Input shape: 28x28
• Output: 10 classes (digits 0-9)

Requirements

• Python 3.8+
• PyTorch (CPU version)
• torchvision
• pytest

Local Setup

1. Create and activate a virtual environment: python -m venv venv source venv/bin/activate # On Windows: venv\Scripts\activate

2. Install dependencies: bash pip install -r requirements.txt

3. Train the model: bash python train.py

4. Run tests: bash pytest tests/

CI/CD Pipeline

The GitHub Actions workflow (ml-pipeline.yml) automatically:

1. Sets up Python environment
2. Installs dependencies
3. Trains the model
4. Runs validation tests
5. Archives the trained model

Tests

The pipeline validates that the model:

• Has less than 25,000 parameters
• Accepts 28x28 input images
• Outputs 10 classes
• Achieves >95% accuracy on test set

Model Artifacts

Trained models are saved with timestamps and accuracy metrics in the saved_models/ directory. The naming format is:

model_timestamp_accuracy.pth

Contributing

1. Fork the repository
2. Create your feature branch (git checkout -b feature/AmazingFeature)
3. Commit your changes (git commit -m 'Add some AmazingFeature')
4. Push to the branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

License

This project is licensed under the MIT License - see the LICENSE file for details.

Notes

• The model is configured for CPU-only training to ensure compatibility with GitHub Actions
• Training is limited to 1 epoch for demonstration purposes
• For production use, consider increasing the number of epochs and adjusting the architecture