GitHub AI Model Integration Example

This project demonstrates integration with the GitHub AI model inference endpoint using the Azure.AI.OpenAI SDK.

Prerequisites

1. .NET 8.0 SDK
2. GitHub Personal Access Token with models:read permissions

Setup

1. Set your GitHub token as an environment variable:

   PowerShell:

   $Env:GITHUB_TOKEN="<your-github-token-goes-here>"

   Command Prompt:

   set GITHUB_TOKEN=<your-github-token-goes-here>

   Bash:

   export GITHUB_TOKEN="<your-github-token-goes-here>"

2. Restore dependencies:

   dotnet restore

Running the Examples

Run the project using:

dotnet run

The program will demonstrate several examples:

1. Basic chat completion
2. Multi-turn conversation
3. Streaming output
4. Tool usage with a flight information example

Each example will be clearly marked in the console output.

Features

• Basic chat completion with the GPT-4o model
• Multi-turn conversation support
• Streaming responses for better user experience
• Function calling with a flight information example

Notes

• Make sure your GitHub token has the required models:read permissions, or the requests will return unauthorized errors.
• This code uses the Azure.AI.OpenAI SDK version 1.0.0-beta.9, not the OpenAI SDK.
• The endpoint used is https://models.inference.ai.azure.com for accessing GitHub AI models.