Conversation
Does anyone play with LLMs, preferably using python / llama_cpp? Is there good way to make it select from multiple choices? The model I'm using is llama-3.2-1b-instruct-q4_k_m.gguf. I'd like to take a sentence from user, and determine if he is a) asking for time, or b) asking for phone call, or ... or h) something else. chatgpt is powerful enough that I can just explain that in a prompt, but there should be way to do that directly with better results. (One option would be to ask for probabilities of multiple utterances, I'm sure there are others.)
0
1
0