Hi!
Is it possible to add a text document (like a book) as added knowledge to one of the ChatGPT-like models like Dolly 2.0 without retraining the whole model? I realize some context can be added as input in a query, but that seems to be quite limited in size.