Getting relevance scores when searching chunks use Qdrant with bge-m3

General
, ,
1

How to get relevance score when search chunks use qdrant + bge-m3

Example code:

class ExternalKnowledgeBase:
    def __init__(self):
        self.vector_db = Qdrant(
            embedder=OllamaEmbedder(id="bge-m3", dimensions=1024),
            url=os.getenv("QDRANT_HOST"),
            port=os.getenv("QDRANT_PORT"),
            collection=os.getenv("EXTERNAL_COLLECTION_NAME"),
            api_key=os.getenv("QDRANT_API_KEY") or None
        )
        self.knowledge_base = TextKnowledgeBase(
            path=paths if paths else "knowledge_base/txt",
            vector_db=self.vector_db,
            num_documents=3,
        )

# Usage:
documents = external_kb.knowledge_base.vector_db.search(query, limit=10)

I want to retrieve the relevance score between the query and the returned documents from Qdrant.
How can I access the score value in this case?

Thanks all.

2

Hi @phucpx247, thanks for reaching out and supporting Agno. I’ve shared this with the team, we’re working through all requests one by one and will get back to you soon.If it’s urgent, please let us know. We appreciate your patience!

3

Hi @phucpx247 relevance scores is a great suggestion, its not yet implemented in our framework right now, but i’ve forwarded this to the team and we’ll get to it in sometime