During this process, however, any history or traceable development that lead to the latest model may get lost since models usually are not being versioned continuously and their performance is not being stored. Results may then no longer be reproducible.

In this workshop, we will introduce MLflow and show how you can integrate this model management framework into the ML lifecycle, from tracking experiments to eventually deploying models.

Upon completion, you will acquire an overview of MLflow and know how to use and integrate its functionalities into your modeling procedure.

Please bring your own laptops to the workshop. Some common Python knowledge is necessary and familiarity with Python ML libraries (at least scikit-learn) is advantageous. Please clone the GitHub repository https://github.com/amld/AMLD_2020_MLflow, take care of the prerequisites mentioned and follow the setup instructions prior to the workshop.
 
At the end of the workshop, participants will get the chance to work with different data sets and implement MLflow in the ML lifecycle. Participants are invited to bring their own data/use cases, otherwise some data sets will be provided.