RAG with Web Search

Retrieval augmented generation (RAG) is a popular way to use current or proprietary data with Large Language Models (LLMs). There are many articles describing how to perform RAG. Typically, they involve encoding data as vectors and storing the vectors in a database. The database is queried and the data is placed in the context where it is tokenized (converted to vectors) along with the prompt to the LLM. At its simplest, RAG is placing data in the prompt for the …

articles context current data database encoding language language models large language large language models llms popular proprietary rag retrieval retrieval augmented generation search vectors web web search