This project illustrates the integration of MLFlow, DVC, CML and Streamlit. All these are tools for model management and deployment.
In this demo, different tools that are developed for the management of a machine learning lifecycle were used to test their utility and how each of them fit in in the grand scheme of Machine Learning Operations(MLOps). There are different parts to a model lifecycle, but this demo is focused mainly on model management and deployment. The tools tested in this demo are:
- DVC
- MLflow
- Streamlit
- CML
- Docker
All of the packages used in this project can be installed with the Python package installer pip and are contained in the requirements.txt file
pip install requirements.txt
An app was built with streamlit to serve the model and make it available for users.