Building a machine learning model that runs locally on a laptop probably isn't generating any value, you have to get that model into production. This talk will focus on getting Data Scientist and Data Engineers more comfortable with the Machine Learning Lifecycle, and how the open source tool MLflow can help. Let's take our machine learning models to production and beyond!
- What is the machine learning lifecycle?
- Why should I care about this?
What is MLflow?
- High-level overview of this open source Python project
What is model tracking?
- Demo how MLflow can easily be used to track and record experiments
How to build a reproducible project?
- Demo how to use MLflow to be able to reproduce model building
How to create models that can be run anywhere?
- Demo building a model with Apache Spark and deploy on a non-Apache Spark cluster.