English
Share
Sign In
Combination of RAG, Embedding Model, and VectorDB
Naver Webtoon, Yang Young-soon <Denma>
It's a combination of three techniques that makes you want to drip. If you've read this guide, you probably remember RAG (Retrieval-Augmented Generation). Then you may naturally be curious. I heard that RAG also provides richer answers by referring to existing information? So what is the role of Embedding Model and VectorDB? And so. Sorry if you weren't curious. But I was curious.
The embedding model is responsible for converting words or sentences into an array of numbers that a computer can understand, that is, a vector. This is similar to teaching a computer a language, and the computer can understand the meaning of sentences or words through these vectors.
VectorDB is a database that stores and manages these converted vectors. In other words, it can be said to be like a ‘vocabulary’ of the language learned by the computer. Here, many words and sentences are organized in vector form, so you can quickly look them up when you need them.
RAG) is an artificial intelligence model that searches for existing stored information (for example, information stored in VectorDB) when answering a question and generates an answer based on it. It is similar to a student recalling what he or she studied to solve a test question. RAG references existing information to produce more accurate and useful answers.
In other words, VectorDB can be a repository of information that the RAG model can reference. Of course, the same applies to other databases. RAG searches vector information stored in VectorDB to generate answers to specific questions and uses that information to create new answers. In this process, RAG 'understands' the information in VectorDB and constructs its own answer based on it.
In this way, RAG leverages existing knowledge to provide richer, more accurate information. However, the reason why RAG is attracting attention is that artificial intelligence does not simply provide predetermined answers, but is innovative in that it creates new answers based on stored knowledge.
ⓒ 2023. Haebom, all rights reserved.
The source is indicated and may be used for commercial purposes with the permission of the copyright holder.
The images used are for illustrative purposes only, and all rights to the images belong to the original authors indicated in the source.