old panther
Hello,
I am trying to test out creating a bot that can have a conversation with a user about a corpus of text. Is it the case that there is no way for a model to hold tokens in memory? Meaning in the various bots that I see around that let you discuss content with them, they need to pass the content each time to the bot for a given answer?
What are strategies for handling this?
Are embeddings useful here?
deft socket
