Hello everyone,
I’m encountering a challenge with an index that contains numerous namespaces, each dedicated to individual users. When a user requests a query, I need to load the vectorestore with that specific user’s namespace to answer the question. Unfortunately, I haven’t been able to locate a method for achieving this based on the documentation provided.
I’m utilizing Langchain for this purpose.
Here’s a simplified version of my code:
embeddings = OpenAIEmbeddings()
vectorstore = Pinecone(
pinecone_index, embeddings.embed_query, "text"
)
chatgpt_chain = ConversationalRetrievalChain.from_llm(
model,
retriever=vectorstore.as_retriever(),
memory=memory
)
Any insights or guidance would be greatly appreciated. Thank you!