Overview
Save Big on Coursera Plus. 7,000+ courses at $160 off. Limited Time Only!
Explore the advanced technique of RAG fusion in this 13-minute video tutorial. Learn how to improve the effectiveness of Retrieval-Augmented Generation (RAG) models by combining multiple search queries and ranking results using reciprocal rank fusion. Dive into the implementation process using LangChain, starting with an overview of RAG and progressing through the steps of reproducing RAG fusion. Download necessary data and utilities, set up the retriever, create the query chain, and understand the reciprocal rank fusion process. Conclude by constructing the final chain, gaining practical insights into applying RAG fusion in various scenarios. Access the provided Colab notebook, blog post, and original code repository for additional resources and hands-on practice.
Syllabus
Intro
RAG Overview
RAG Fusion Blog Post
Rag Fusion Diagram
Code Time
Reproducing RAG Fusion with LangChain
Download Data and Utils
Setup Retriever
Create the Query Chain
Reciprocal Rank Fusion
Make the Final Chain
Taught by
Sam Witteveen