Find answers from the community

Updated 4 months ago

Implementing Advanced Fusion Retriever

At a glance

The community member is trying to use the Advanced Fusion Retriever with a Milvus Vector Store that already has indexed data, but it is not working as the retriever wants to reindex everything every time. The comments suggest that the community member should set up the existing vector store and then use the VectorStoreIndex.from_vector_store() method, but this is failing and showing zero entities. Additionally, there is an error in the query_index method, which is causing a division by zero error.

Useful resources
Implementing Advanced Fusion Retriever with Storage Context
I want to try Advance Fusion Retriever with Milvus Vector Store with already indexed data
https://docs.llamaindex.ai/en/stable/examples/low_level/fusion_retriever.html
It's not working with already indexed data, It wants to reindex all everytime
L
B
3 comments
Setup your existing vector store, then do index = VectorStoreIndex.from_vector_store(vector_store, service_context=service_context)
that's failing showing zero enteties
Error in query_index UUID: 123 token: 123 - division by zero
Add a reply
Sign up and join the conversation on Discord