Hugging Face and AMD : LLMs Out of the Box Acceleration 

2024-02-09 | 02:30 PM - 03:00 PM | Demo Stage 2


Hugging Face is the default location for Opensource AI models these days. Hugging Face and AMD have been collaborating over last several months to enable all the models on Hugging Face to work out of the box on AMD platforms. In this talk you will get a glimpse of the ease of use, performance optimizations and some performance benchmarks of LLMs on AMD GPUs and also witness how easy it is to download a LLM model and execute it locally on a Ryzen ™AI laptop without internet connectivity.