What is RAG? And how does it relate to Agents?

General
1

Hi! I’ve heard a lot about RAG and I want to know if anyone knows what it is and has any cool resources to learn it because I’ve heard that it’s popular with agentic frameworks?

2

According to Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks(https://dl.acm.org/doi/abs/10.5555/3495724.3496517) RAG seems to be about a new method to train LLM’s on data that they weren’t initially trained on. This results in the LLM being more accurate in it’s responses and reduces AI hallucinations. However I still am unsure how it relates to agentic frameworks, if anyone is aware of them.

Hope this helps.

3

Thanks! Does anyone have any resources on this?

4

I would strongly suggest using RAG with langchain.
Here is a tutorial - Build a Retrieval Augmented Generation (RAG) App: Part 1 | 🦜️🔗 LangChain

Langchain is a tool that allows you to connect and create complex agentic workflows. The value add of langchain is that it allows you to easily debug, and more easily understand what the LLM is doing. It also does a good job of wrapping away all the boilerplate code.