Kacper Łukawski

The project seems not to be maintained anymore, but since we're doing something similar at Qdrant (https://github.com/qdrant/qdrant), I think I may answer that question. Those tools are rather designed to...

@raulcarlomagno, if you could please request that feature on Airlfow's Github, we can handle that together!

@dagthomas @hp0404 @YLFxGen This has been implemented in #5446.

@ngundotra Thanks for mentioning that. It should be fixed by #1114

@ngundotra I just realized you already made another PR #1102. Thanks a lot!

@hopkins385 Why do you want to set the vector size externally? It should be derived from the embeddings, as far as I know.

@c0sogi Thanks for handling that! Do you see a chance not to repeat the batching process but unify it for sync and async? Can we work it out together?

@c0sogi It's a good idea to switch to the list of points instead of the batch, to make things consistent. I'm unsure if we should throw `NotImplementedError` for async in...

Async API was introduced in #7704.

@otwieracz It seems the `similarity_search_by_vector` should return `List[Document]`, not `List[Tuple[Document, float]]`. That should be consistent with the supertype.