Unlocking Data with Generative AI and RAG

10

Key RAG Components in LangChain

This chapter takes an in-depth look at the key technical components that we have been talking about as they relate to LangChain and retrieval-augmented generation (RAG). As a refresher, the key technical components of our RAG system, in order of how they are used, are vector stores, retrievers, and large language models (LLMs). We will step through the latest version of our code, last seen in Chapter 8, Code lab 8.3. We will focus on each of these core components, and we will show the various options for each component using LangChain in the code. Naturally, a lot of this discussion will highlight differences among each option and discuss the different scenarios in which one option might be better over another. ...