Take advantage of insta-scaling for live inferencing

Learn how to deploy an ML container with FastAPI and deploy it to Azure Container Apps with GitHub Actions. This repository gives you a good starting point with a Dockerfile, GitHub Actions workflow, and Python code that already works for generating text using GPT-2 with HuggingFace Transformers.

First, you'll go through an architectural overview of the end-result, then you will go through every configuration item needed to set the automation right between Azure and GitHub Actions and the GitHub Container Registry. Finally, you'll see how to deploy and find a few crucial requirements needed for everything to work, like ingress ports and setting the right amount of RAM and CPU.

In this video lesson, I'll go over the details with an example repository you can use for reference including the following learning objectives:

• Use GitHub Container Registry to push a built container
• Use the Azure CLI in a GItHub Action to authenticate to Azure
• How to generate an Azure Service Principal and a Personal Access Token to authenticate
• Configure Azure Container Apps to correctly serve a model with enough resources
• Look at deployment logs to ensure things are working right

• Example repository
• Practical MLOps book
• MLOps Maturity Model
• Packaging ML models
• MLOps packaging: HuggingFace and Docker
• MLOps packaging: HuggingFace and Azure Container Registry