Chap -12 ML SYSTEMS

Learn and implement the different types of design patterns and architectural styles when you need to build, serve and release ML models.

1. Model Registry - a custom mlflow like server built using fastapi and postgres db.
2. Model in Image - serving fastapi with kubernetes for production having a pytorch model inside the container.
3. Model Load - serving fastapi with kubernetes for production pulling a pytorch model from s3 inside the container.