search

vanna-ai / vanna

🤖 Chat with your SQL database 📊. Accurate Text-to-SQL Generation via LLMs using RAG 🔄.
https://vanna.ai/docs/
MIT License
12.06k stars 967 forks source link

No option to increase fetch limit for Weviate Vector DB #615

Closed talrejanikhil closed 3 months ago

talrejanikhil commented 3 months ago

Is your feature request related to a problem? Please describe. As of now, the no. of results that can be fetched for Weviate DB is hard coded to 3. This should be flexible like it is for other vector stores

Describe the solution you'd like The no. of results fetched for Weviate DB should be changeable using config. Here is an example for ChromaDB

self.n_results_sql = config.get("n_results_sql", config.get("n_results", 10))
self.n_results_documentation = config.get("n_results_documentation", config.get("n_results", 10))
self.n_results_ddl = config.get("n_results_ddl", config.get("n_results", 10))

And pinecone

self.n_results = config.get("n_results", 10)

Describe alternatives you've considered Using a different vector store

zainhoda commented 3 months ago

I agree with this. Ideally these should be configurations in the base class that is implemented for every vector store.

talrejanikhil commented 3 months ago

Could I do a quick fix for Weviate for now? I am not sure what is the plan for having it across all vector stores

zainhoda commented 3 months ago

@talrejanikhil you have a couple of options -- if you want to submit a PR you're more than welcome to do so. The other alternative is to follow the setup for "Other Vector Database" and then just copy/paste the implementation from Weaviate and then make the necessary modifications.

talrejanikhil commented 3 months ago

I am unable to push changes to my branch. Can you please grant access?

zainhoda commented 3 months ago

You have to first fork and then submit a PR from the fork

talrejanikhil commented 3 months ago

https://github.com/vanna-ai/vanna/pull/616