GitHub Models
This is a demo video for the launch of GitHub Models. The video covers the following specific topics:
- GitHub Models on the GitHub Marketplace: a curated collection of top AI models available under your existing GitHub account entitlements
- Exploring GPT-4o in the interactive playground, including adjusting the system prompt and temperature parameter
- Side-by-side model comparison: switching from GPT-4o to phi-3 mini instruct and evaluating responses to the same prompt
- Navigating model detail pages (README, evaluation, and transparency tabs) to inform model selection
- Getting started in code using the built-in code samples and GitHub Codespaces, pre-configured with SDKs and no API key setup required
- Updating the streaming sample to target a specific model (phi-3 mini 4K instruct) and customising prompts
- Using GitHub Models from the GitHub CLI: generating first-year computer science quiz questions with an AI model call
- Combining GitHub CLI commands with model calls, such as summarising the last 10 commits of a repository using phi-3 mini instruct
Related Content
How to use Copilot Workspace for inspiration
GitHub
Demonstrates GitHub Copilot Workspace starting from a template repository to scaffold a to-do application, showing how AI can jumpstart creativity when starting a new project. The video walks through the full specification-to-plan workflow, iterating with a live app preview, and switching between Copilot Workspace and GitHub Codespaces to make direct code edits. It also covers the share feature for collaborating on work in progress and the automated first deployment via GitHub Actions.
Code from your phone with Github Copilot Workspace
GitHubShows GitHub Copilot Workspace running on the GitHub mobile app, allowing developers to capture ideas and iterate on code from a phone or tablet. The video covers using device dictation for task input, reviewing and editing the AI-generated specification and plan, inspecting code changes via the built-in diff view, running unit tests from the integrated terminal, and creating a pull request with an auto-generated description—all from a mobile device.
ToolUp Days #15
Chris and Matt pick up inside GitHub Codespaces, getting the Player State service running locally on the first attempt thanks to Dapr's environment-variable secret store and Codespaces secrets. They then wire up the World Events Engine with Azure Storage Queue bindings and Table Storage state before hitting a stubborn .NET minimal API routing bug — becoming the cliffhanger that sets up the next session.