The Problem

We are using Qdrant as our vector store and recently have updated our collection to use quantization in order to reduce memory utilization and improve response times. Here are the relevant docs: https://qdrant.tech/documentation/guides/quantization/

When working with quantized vectors, Qdrant allows for search parameters to enable rescoring and oversampling, which improves accuracy while querying quantized vectors.

Additionally these search parameters allow one to control other index specific parameters while performing search or query requests. More info on query and search

Proposed Solution

Allow for these params to be set when using the QdrantVectorStore object and using the query method