Run GPT-OSS-120B Locally with Lemonade on AMD ROCm™ Software

Опубликовано 17 ноября 2025, 18:30

In this tutorial, Victoria Godsoe, Software Development Manager at AMD, walks through how to run the GPT-OSS-120B model locally on a Strix Halo PC using Lemonade, AMD ROCm™ software, and open-source tools.

Learn how to:
• Launch Lemonade Server with ROCm™ as the backend
• Load and serve a 60B+ parameter model
• Generate and edit a playable Asteroids game using local inference

Whether you're building local LLM workflows or experimenting with open model deployment, this demo shows what’s possible with AMD hardware and the ROCm™ software platform.

Star the repo: github.com/lemonade-sdk/lemonade

Try Lemonade on GitHub Copilot: marketplace.visualstudio.com/i...

Join the AMD Developer Discord: discord.gg/2tYF7hqW

Find the resources you need to develop using AMD products: amd.com/en/developer.html

***

© 2025 Advanced Micro Devices, Inc. All rights reserved. AMD, the AMD Arrow logo, EPYC, ROCm, and AMD Instinct and combinations thereof are trademarks of Advanced Micro Devices, Inc.

GSGSMSTHMSZEXJYI