CODEWITHSUNDEEP
langchainlarge-language-modelragdspy
o3omoomooa
o3omoomooa

Reputation: 1

Is there a way to implement multiple csv's as RAG?

I recently uploaded a csv and wanted to create a project to analyze the csv with llm.

However, I don't know which RAG to use for RAG through the csv file.

In addition, the resources of the csv file are numbers, not natural language, so it seems too difficult to draw out the performance of RAG.

Does anyone have a good method or idea?

I looked through the Pandas DataFrame and LangChain documentation, but I couldn't find a way to implement a performant RAG for csv.

Upvotes: 0

Views: 294

Answers (2)

Safder Raza
Safder Raza

Reputation: 11

How big are your csv files. If they are small and can fit into the context window, just try putting them in the prompt. Most LLMs do a decent job of answering questions on a csv file if its small enough.

If your csv files are large, then look for text->SQL or text -> dataframe solutions

Upvotes: 0

Jaycee
Jaycee

Reputation: 1

I think the advantage of rag is that it processes unstructured text data. If you want to process csv data, you still need some specific functions.

Upvotes: 0

Related Questions