Retrieval Augmented Generation

This week we will be talking about Retrieval Augmented Generation – also known as RAG. The basic premise is to use an existing LLM to generate content based on an existing collection of documents. This removes the need to fine-tune an LLM, which is cost prohibitive and hardware constrained.

Over the next several sessions, we will be diving deeper into separate components needed for RAG – hopefully resulting in a chat-based Q&A service for the NASA Technical Report Server.  We were introduced to this data during our submission for the 2022 NASA SpaceApps Challenge – where we placed 2nd. Our submission was a semantic search based on the abstracts for the NSTR dataset of 10,000 papers.

For this session, we will cover an overview of RAG as well as walk through the submission from the challenge. I have recently updated it to work with current versions of the dependent libraries – so we can demonstrate the starting point for this project.

Also – if you missed last week’s session on AI & Video – you can find the recording here – https://hsv.ai/2024/02/22/ai-breakthroughs-in-video/

Links:

• RAG Paper – https://arxiv.org/pdf/2005.11401.pdf
• Huntsville AI 2022 SpaceApps Submission – https://github.com/HSV-AI/spaceapps2022

Details:

• Date – 03/06/2024
• Time – 6-7pm
• Location – HudsonAlpha
• Address –  601 Genome Way Northwest, Huntsville, AL 35806
• Zoom –https://us02web.zoom.us/j/82845221441?pwd=cEdZYXBlVm1naC82U0dpZlljV1B6Zz09

Hope to see you soon!
-J.