Run and scale LLMs on Quest, Northwestern’s high-performance computer, leveraging GPUs and using popular tools like Ollama and Hugging Face.
Learn how to use Northwestern's Quest high-performance computing resources to run large language models (LLMs) for research. This hands-on workshop introduces the different ways to deploy and use LLMs on Quest, including through Hugging Face and Ollama. You'll learn when each approach is most appropriate and how to choose the right workflow for your research.
Through guided examples, you'll configure jobs, select appropriate GPU resources, and explore the key options that affect performance, scalability, and cost. By the end of the workshop, you'll understand how to run LLMs efficiently on Quest and choose the deployment strategy that best fits your research needs.
Prerequisites: Basic familiarity with coding in Python (other programming language OK, but workshop will be taught in Python). Participants should either have an active Quest account or be eligible to obtain one (we will provide temporary access if they don't have one but are eligible).
Leticia Vega
Email