TypeError: only integer scalar arrays can be converted to a scalar index about qdrant-client HOT 7 CLOSED

Ahh I see. We will fix this soon.

@dabasmoti you can try using our client-server implementation which we anyways recommend for production. It works fine. Just use:

client = QdrantClient('localhost')

from qdrant-client.

Yep,
you are right.
It does work with the server.

from qdrant-client.

I took the liberty to transfer the issue to the qdrant Python client repository.

from qdrant-client.

Hi @dabasmoti

We have a bug on our side.
In case you need to use local mode and don't want to wait for the release, you can sort indices and corresponding values before vector creation, e.g.:

OK:

SparseVector(
    indices=[0, 5],
    values=[0.3, 0.2]
)

ERROR:

SparseVector(
    indices=[5, 0],
    values=[0.2, 0.3]
)

from qdrant-client.

Good catch. I think you can do this in the meantime to make it work.

client.upsert(
    collection_name=COLLECTION_NAME,
    points=[
        models.PointStruct(
            id=1,
            payload={},  # Add any additional payload if necessary
            vector={
                "text": models.SparseVector(
                    indices=np.array([1479876206, 696618268]),
                    values=np.array([0.5, 0.6])
                )
            },
        )
    ],
)

@joein @agourlay Please have a look

It happens because in the source code sorted_indices is a list and arr[list_of_indices_to_extract] works with numpy arrays but not python arrays. Here sorted_indices = [0, 1] and vector.indices = [1479876206, 696618268].

We need to fix local mode implementation for this. I think client-server mode works fine.

from qdrant-client.

@KShivendu
I tried it before.
It doesn't solve it.
same error

from qdrant-client.

available as of qdrant-client==1.7.1

from qdrant-client.