vince-lam / ml_service

Deploying Machine Learning Models with Django

This project follows the tutorial at https://deploymachinelearning.com/

The web service allows Machine Learning models to be available with REST API. The scope of the project includes:

• keeping the ML models in the web service
• allowing several ML models to be avalable at the same endpoint with different versions
• allowing many endpoint addresses to be defined
• storing information about requests to the ML models, allowing for future model testing and auditing
• testing of ML code and server code
• running A/B tests between different versions of ML models.

The project structure

In the research directory there are:

• code for training machine learning models
• code for simulating A/B testing

In the backend directory there is a Django application.

In the docker directory there are dockerfiles for running the service in the container.