README

Overview

This is an image classification project using the MNIST dataset with PyTorch, Docker and GitHub Actions.

The project contains:

• models/mnist_model.py - The PyTorch model definition
• notebooks/train.ipynb - Notebook to train the model
• app/app.py - Flask app to serve predictions
• Dockerfile and docker-compose.yml to containerize the app
• .github/workflows/main.yml - CI/CD workflow with GitHub Actions

Running the app

To run the Docker containers: docker-compose up --build

The app will be served at http://localhost:5000

Jupyter notebook can be accessed at http://localhost:8888

To retrain the model, run train.ipynb, then rebuild the Docker image.

CI/CD Pipeline

The GitHub Actions workflow trains the model on new code changes, evaluates it, and automatically deploys the Docker image.

Workflow runs can be seen under the Actions tab.