Ever wished you had a bigger machine for training fancy machine learning (ML) models? Or needed a bunch of machines to quickly find the best model and parameter settings for your problem? How do you track all these models and select the best one to deploy to your end-users? And how do we keep things reproducible, so we know how any given model was trained and with what data?
Nowadays, many cloud providers offer fancy MLOps suites with tools that promise to help you solve all of these problems. In this code breakfast, we’ll explore Google’s offering, Vertex AI, and see how Google’s tools can help us do scalable and reproducible machine learning in practice.