Failed to create collection Error: Unknown arguments: ['payload_schema']

[urbanscribe]

Running 1.8.2

Keep getting this error when trying to define the payload schema in a new collection

` def create_collection(self):
"""
Creates a collection with the specified vectors and payload configuration.

    Parameters:
        self: The object instance
        vectors_config: The configuration for the vectors
        payload_config: The configuration for the payload
    
    Returns:
        None
    """
    vectors_config = {
        "vectors": {
            "size": 768,  # The dimensionality of the vectors
            "distance": "Cosine"  # The distance metric for vector comparison
        }
    }
    payload_config = {
        "user_id": {"type": "integer"},
        "company_id": {"type": "integer"},
        "project_id": {"type": "integer"}
    }
    optimizers_config = {
        "max_optimization_threads": 2  # Adjust this value as needed
    }

    try:
        self.qdrant_client.create_collection(
            collection_name=self.collection_name,
            vectors_config=vectors_config,
            payload_schema=payload_config,
            optimizers_config=optimizers_config
        )
        
    except Exception as e:
        st.error(
            f"Failed to create collection '{self.collection_name}'. Error: {e}")`

Any thoughts?

[joein]

hey @urbanscribe

create_collection does not have payload_schema argument
payload indices created separately via create_payload_index

you can find an example here

[urbanscribe]

thanks, what confusing is you can fetch payload_schema from a CollectionInfo

 status: "CollectionStatus" = Field(..., description="Current statistics and configuration of the collection")
   optimizer_status: "OptimizersStatus" = Field(
       ..., description="Current statistics and configuration of the collection"
   )
   vectors_count: Optional[int] = Field(
       default=None,
       description="Approximate number of vectors in collection. All vectors in collection are available for querying. Calculated as `points_count x vectors_per_point`. Where `vectors_per_point` is a number of named vectors in schema.",
   )
   indexed_vectors_count: Optional[int] = Field(
       default=None,
       description="Approximate number of indexed vectors in the collection. Indexed vectors in large segments are faster to query, as it is stored in a specialized vector index.",
   )
   points_count: Optional[int] = Field(
       default=None,
       description="Approximate number of points (vectors + payloads) in collection. Each point could be accessed by unique id.",
   )
   segments_count: int = Field(
       ..., description="Number of segments in collection. Each segment has independent vector as payload indexes"
   )
   config: "CollectionConfig" = Field(..., description="Current statistics and configuration of the collection")
   payload_schema: Dict[str, "PayloadIndexInfo"] = Field(..., description="Types of stored payload")

so what are the best practices for managing the payload schema

perhaps there is a confusion on my part between setting a payload schema for a collection and creating an index inside that collection. I am looking to do the former, I know how to do the latter.

[joein]

payload_schema in CollectionInfo shows payload indices created for that collection

You don’t need (and currently can’t) set a payload schema to describe all the fields in your payload, if that is what you’re looking for

[urbanscribe]

Thanks.

Does that mean every time I upsert a file with any new payload/metadata field/value it will accept it and thus defining a schema for validation ahead of time is not necessary or supported?

Payload items, schema and indexing are used a little interchangeably throughout the docs.

Thanks in advance.

[joein]

Yes, it is correct

Payload items, schema and indexing are used a little interchangeably throughout the docs.

Could you provide some links which causes confusion please?