LinkedIn Google Cloud Professional Machine Learning Engineering Certificate Preparation

This repository is part of my effort to pass the Google Cloud Professional Machine Learning Engineering Certificate

It follow the course given by Noah Gift on LinkedIn Learning: https://www.linkedin.com/learning/paths/prepare-for-the-google-cloud-professional-machine-learning-engineer-certification

Here you can find my notes on the exam: https://gortium.github.io/Slit-box/LinkedIn-Google-Cloud-Professional-Machine-Learning-Engineer-Certification

1. Framing ML Problems

Here you can find my notes on this chapter: https://gortium.github.io/Slit-box/Framing-ML-Problems

In addition to the terms and definitions, in this chapter, Noah Gift showed us multiple practical tricks to allow better CI\CD workflow in MLOps:

Github codespaces

Codespace allow online, and always clean environment. It also allow the use of cloud computing to build code or train ML models. It use docker and vscode.

I got to play with codespace, but saddly I was not able to get access to a GPU so I ended up running the code on my own machine.

devcontainer.json

This file allow us to define a custom codespace (online vscode) deployed within our custom docker and our personalized vscode extention. That way we always have a clean development environment to work in.

Dockerfile

This file allow us to define our custom environment where our codespace will be installed in.

Github actions

Github action allow automated workflow to be executed by github manualy, or at every push. Similar to Jenkins.

Makefile

This file make it easier to call the different CI/CD steps of our develpment workflow.

cicd.yml

This file allow us to call ours Makefile from github automaticaly, or manualy.