From the course: Advanced Guide to ChatGPT, Embeddings, and Other Large Language Models (LLMs)

Unlock this course with a free trial

Join today to access over 25,200 courses taught by industry experts.

Introduction to retrieval-augmented generation (RAG)

Introduction to retrieval-augmented generation (RAG) - ChatGPT Tutorial

From the course: Advanced Guide to ChatGPT, Embeddings, and Other Large Language Models (LLMs)

Start my 1-month free trial Buy for my team

Introduction to retrieval-augmented generation (RAG)

“

- It is time to put a lot of what we learned in the last couple of sessions into a real-life application. We've seen semantic search and we've seen how prompt engineering can drastically change the outputs of our generative AIs. Putting these two together often results in a system called retrieval-augmented generation, or RAG. A RAG system, at its simplest, is an interactive application, usually a chatbot, but technically not always, that combines retrieved and generally static information, i.e. semantic search, with a dynamic conversation generator. In our case, generative AI, like a GPT, Claude, Cohere, Llama, name it. The premise of how this is going to work, this figure looks familiar, it's because it's effectively a ripoff of the semantic search system that I showed you a couple lessons ago. A basic RAG system will involve someone asking a question, or a query on the left. The system retrieving results, i.e. through semantic search from some Pinecone database, other vector…

Contents