Hello, I have an Issues about Qdrant database vectore store, i would like my users, can delete or add file on a specific index for personalise it, it's easy to handle that with add logic, but for delete i need to delete every chunks (stored on qdrant) on a specifi adress with only the doc name as information. I arrive to delete some chunks but can anyone know how i can fetch every chunk with only doc name? Thanks you

Hello, I'm trying to use a query pipeline to rewrite a query, then use a vector store and a reranker to retrieve the nodes. I'd like to pass the reranked nodes into the second template but I'm getting the following error: ValueError: Only one root is supported.

Hi there, I'm currently in the process of selecting a vector database for my project. Could you please provide some information on the following: Compatibility with asynchronous operations? The practicality and usefulness of hybrid search? Factors influencing the choice between different technologies? I'm considering options like Redis, Chroma, pgvector, Milvus, and Qdrant. Also, just to provide some context, I'll be working with around 5,000 PDFs, totaling approximately 110,000 chunks. Your insights would be greatly appreciated. Thanks!