Need help with Pinecone, or want to use your experience to help other Pinecone users? This is the place.
Also see our docs. Paying customers can contact support@pinecone.io for priority support.
I´m creating a chatbot with access to a custom knowledge Vector DB using Pinecone. One question I have is if the answer for the query is not found in the pinecone index, can you provide an instructions to the language model to provide the answer with the existing trained data it has? For example if a user query _“who is the first man to land on the moon?” is a generic query that can be answered with existing trained data of the LLM, but if the user ask " top resturants with rooftop in mexico city" it can used the pinecone index?
If you have got solution how to resolve this type of error, can you plz tell me