Qdrant Python for Hybrid Search

snippets.py

## Recommended Imports
from qdrant_client import QdrantClient
from qdrant_client.models import (
    Distance,
    NamedSparseVector,
    NamedVector,
    SparseVector,
    PointStruct,
    SearchRequest,
    SparseIndexParams,
    SparseVectorParams,
    VectorParams,
    ScoredPoint,
)

## Creating a collection
client.create_collection(
    collection_name,
    vectors_config={
        "text-dense": VectorParams(
            size=1024,  # OpenAI Embeddings
            distance=Distance.COSINE,
        )
    },
    sparse_vectors_config={
        "text-sparse": SparseVectorParams(
            index=SparseIndexParams(
                on_disk=False,
            )
        )
    },
)

## Creating Points
points = []
for idx, (text, sparse_vector, dense_vector) in enumerate(zip(product_texts, sparse_vectors, dense_vectors)):
    sparse_vector = SparseVector(indices=sparse_vector.indices.tolist(), values=sparse_vector.values.tolist())
    point = PointStruct(
        id=idx,
        payload={"text": text, "product_id": rows[idx]["product_id"]},  # Add any additional payload if necessary
        vector={
            "text-sparse": sparse_vector,
            "text-dense": dense_vector,
        },
    )
    points.append(point)

## Upsert
client.upsert(collection_name, points)