This workshop will cover an introduction to LLMs and how to use them for inference on UVA HPC systems. Hands-on exercises will be included in the workshop. We will cover LLM set up/installation, HPC resource allocation, model selection from a hub such as Hugging Face, using LLMs for inference, and creating Slurm scripts for LLM jobs.
For UVA users only. An HPC account is recommended. You should be familiar with Python and general HPC concepts. Familiarity with LLMs will be helpful but is not necessary.