MLFlow in Action

This project contains the exploration of Mlflow, a tool used in MLops to track experiments, runs of a Machine Learning model and sample steps to deploy ML models in production using AWS.

Some of the components explored:

• Tracking: Track experiments, runs. Log inputs, parameters, evaluation metrics, datasets.
• Mlflow models: Model signature and associated functions.
• Mlflow registry: Registering model through UI/API/mlflow client, MLProject file.
• Mlflow Client: Capabilities of mlflow client
• Mlflow CLI: CLI commands to run mlflow functionalities.
• Mlflow with AWS: PoC ro run mlflow project on AWS.