Azure with Machine Learning Workshop

Welcome to the Azure with Machine Learning workshop! This repository contains the code and instructions for an interactive coding session focused on cloud-native development on Azure, serverless functions, and machine learning.

Project Overview

In this project, you'll create a serverless Azure function that integrates with a machine learning model. The function will accept an image upload and return information about the image, such as age and emotions.

Prerequisites

Before you begin, ensure you have the following:

Find a Team

• This is a Team Exercise. Form Groups of 2-3 students.

Azure Account

• Each team needs a free Azure account:
  • Create Azure Account https://azure.microsoft.com/free/?ref=microsoft.com&utm_source=microsoft.com&utm_medium=docs&utm_campaign=visualstudio
  • If eligible, use the Azure for Students offer. https://azure.microsoft.com/en-gb/free/students/

Tools and Software

Automatic Setup

• If you have nix installed, use the flake: from project-root run nix develop. Everything will be set up for you. You can use the setup.sh script to install nix.

Alternatively

• Azure CLI: Install Azure CLI https://learn.microsoft.com/en-us/cli/azure/install-azure-cli
• Azure Functions Core Tools: Install Azure Functions Core Tools https://docs.microsoft.com/en-us/azure/azure-functions/functions-run-local?tabs=v4%2Cwindows%2Ccsharp%2Cportal%2Cbash#install-the-azure-functions-core-tools
• Visual Studio Code: Download Visual Studio Code https://code.visualstudio.com/
  • Install these extensions:
    • REST Client https://marketplace.visualstudio.com/items?itemName=humao.rest-client
    • Python https://marketplace.visualstudio.com/items?itemName=ms-python.python
    • Azure Functions https://marketplace.visualstudio.com/items?itemName=ms-azuretools.vscode-azurefunctions
• Python 3.11 on PATH: Download Python 3.11 https://www.python.org/downloads/release/python-3110/
  • Install these Python packages:
    • azure-functions
    • opencv-python
    • onnxruntime
    • numpy
  • Set up a virtual environment in the ./HttpTrigger directory:

    python3.11 -m venv .venv

Getting Started

Follow these steps to set up and run the project:

1. Login to Azure

• Open the terminal in Visual Studio Code and run:

  az login

• This will open a browser window for you to log in.

2. Set Up the Project

• Clone this repository and navigate to the project directory.
• Open the folder starting with 01 in Visual Studio Code.
• Follow the instructions in the subsequent folders (02, 03, etc.).

Troubleshooting

If you encounter any issues, please reach out to us for assistance. We're here to help you succeed!