In RAG, As part of retriever model- we are retrieving the relevant information from external knowledge source (i.e. vector database) and this database is always updating with new updates.

In application implementation, are we storing data in databse (with relevant information)? if 'yes' then how we maintating this database as updated?

  • $\begingroup$ Hi @Tovlk, welcome to the site. Your question is not very clear. Do you mean that your knowledge database is constantly being updated and are you asking how to keep your vector database in sync with the knowledge database? $\endgroup$
    – noe
    Feb 23 at 8:46
  • $\begingroup$ As per this statement- RAG provides a solution to mitigate hallucinations by augmenting LLMs with external knowledge such as databases. 1. how these databases are loaded with data? 2. which data it stores? 3. do we load this data manually? $\endgroup$
    – Tovlk
    Feb 23 at 10:34
  • $\begingroup$ The answers to these questions are defined by the person creating the RAG system. For instance, you can have automated process ingesting all the documents of a company into the knowledge database or you can already have such a database from before and just query it. $\endgroup$
    – noe
    Feb 23 at 11:27
  • $\begingroup$ means are you referring task specific dataset or other? $\endgroup$
    – Tovlk
    Feb 26 at 1:42


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.