Functionality of the vector database retriever

Questions
1

Does the vector database retriever always take the prompt from the Q&A LLM node as basis for the vector search or can I define a custom vector query and insert the found documents via a dynamic variable?

In my case I want to query helpful and relevant document chunks from my vector database and append these to a predefined LLM prompt.
However, the LLM prompt itself has nothing to do with the actual vector search and I would love this to be two seperate processes. I am not sure whether that is possible on n8n.

I appreciate any ideas, tips & tricks! Thank you in advance! :slight_smile:

2

It looks like your topic is missing some important information. Could you provide the following if applicable.

  • n8n version:
  • Database (default: SQLite):
  • n8n EXECUTIONS_PROCESS setting (default: own, main):
  • Running n8n via (Docker, npm, n8n cloud, desktop app):
  • Operating system:
3

Solved. You can use “Get Many” in the vector database node

2 Likes
4

This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.