As the excitement around ChatGPT and other large language models (LLMs) continues to grow, many are exploring how to leverage these technologies in their own domains to achieve more targeted and advanced responses. Among the emerging techniques, Retrieval-Augmented Generation (RAG) stands out as one of the most impactful. In this 2-hour lecture, I will introduce the fundamental concepts of RAG through engaging slides and simple demos. We’ll also delve into some advanced techniques within RAG, providing a deeper understanding of how it works.
The session will focus on demonstrating how to build RAG systems using UCLA’s Hoffman2 resources, specifically tailored for the UCLA community. No prerequisites are required, making this lecture accessible to all.