Google’s Practitioners Guide to MLOps

Deep dive of MLOps processes — Part 1

This post is based on Google’s 2021 published white paper:

Practitioners guide to MLOps: A framework for continuous delivery and automation of machine learning.

The first article I wrote provided an Overview of the MLOps lifecycle and core capabilities and can be read here. When finished you can go to part 2 of the deep dive.

With this post, I will start a series about each process covered in the paper, specifically, the first three stages: ML Development, Training Operationalization, and Continuous Training.

You can read it either chronologically from the start or pick a topic of interest in the table of contents below.

Table of Contents

• Introduction
• ML Development
• Training Operationalization
• Continuous Training
• Final Notes

Introduction to the Deep Dive of MLOps Processes

While the first part of this two-part article series, gave only a brief outline of all the core steps of an ML workflow, this one will dig into a little more detail.

Covered will be the key tasks and flows of control between them. Additionally, the main artifacts created by those…