SGLang Summit Workshop by LMSYS: Local AI Coding Lab

​A SGLang Summit ticket is required to attend the event and get approved on Luma. Both Full and Expo passes are accepted.

​Ticket 🎫 https://luma.com/sglang-summit

​Build a complete local AI coding workflow, from serving a model to solving a coding challenge with an AI agent.

​Join the SGLang and NVIDIA Local AI teams for a two-hour hands-on workshop. Working in an environment with only local model access, you’ll deploy Qwen3.8-27B with SGLang, test your endpoint, connect a coding agent, and put it to work on a programming task.

​What You’ll Do

• ​Explore local AI hardware: Get a brief introduction from NVIDIA to hardware considerations for running models locally.

• ​Deploy Qwen3.8-27B: Configure SGLang and launch a model server.

• ​Test your endpoint: Send inference requests and verify your setup.

• ​Connect a coding agent: Point a local coding agent at your model endpoint.

• ​Take on a coding challenge: Use your agent to modify, debug, or optimize code. Potential exercises include mini-sglang optimization.

​Who Should Join

​Developers interested in local AI, model serving, and coding agents. Familiarity with Python, the command line, and basic Docker usage is recommended.