Find answers from the community

Updated 2 months ago

This may be an issue for the Neo4J folks

This may be an issue for the Neo4J folks
Attachment
Screenshot_2024-06-20_at_7.43.54_AM.png
L
g
19 comments
I'm not fully sure I undertsand this error
The types all make sense under the hood, unless you don't have the latest version of the neo4j vector store
Neo4jVectorStore extends BasePydanticVectorStore
Just noticed this in VSCode; I'm using

these versions:

Plain Text
llama-index-llms-openai = "^0.1.22"
llama-index-vector-stores-neo4jvector = "^0.1.4"
llama-index-embeddings-openai = "^0.1.10"
neo4j = "^5.21.0"
llama-index-graph-stores-neo4j = "^0.2.4"
llama-index-embeddings-bedrock = "^0.2.1"
Probably you want llama-index-vector-stores-neo4jvector = "^0.1.5" if its not already installed
Oh weird, maybe it didn't publish. Wll try a manual publish
should be published now
Plain Text
# TODO: Temp copy of VectorStore for pydantic, can't mix with runtime_checkable
class BasePydanticVectorStore(BaseComponent, ABC):
    """Abstract vector store protocol."""

    stores_text: bool
    is_embedding_query: bool = True

    class Config:
        arbitrary_types_allowed = True

    @property
    @abstractmethod
    def client(self) -> Any:
        """Get client."""


I don't see an implementation of client in Neo4jVectorStore
Missing

Plain Text
    def client(self) -> neo4j.GraphDatabase.driver:
        return self._driver
?
Yeah, so that's missing in Neo4jVectorStore at the moment.
Easy pr to contribute πŸ˜‰
Oooh, you guys run a lot of CI.

Ping me if you want changes otherwise merge away.
so much CI lol
Pushed one small nit (should have the property decorator) but looks good!
Add a reply
Sign up and join the conversation on Discord