Find answers from the community

Updated 6 months ago

How to extract nodes with embeddings?

I extract nodes from a vector store:

Plain Text

filters = MetadataFilters(
        filters=[MetadataFilter(key="file_name", value=document.file_name)])
nodes = vector_store.get_nodes(filters=filters)

I'm using Qdrant vector store btw.

How can I get embeddings from those nodes? Do I need to embed them again? @Logan M bro can you help with this one, please?

3 comments

ppikachu8887867

Using qdrant's native library, I can do this:

Plain Text

client.scroll(with_vectors=True, filters=..., collection_name=...)

Is there a way to do it in llamaindex?

ppikachu8887867

I think get_nodes function should get kwargs and then send those here:

Attachment

LLogan M

yea thatd be a good improvement

Add a reply