How about original LLaMA Chat prompt format?
#3
by
gotzmann
- opened
What are the benefits of using ChatML (OpenAI?) prompt format with models that are supposed to be used with native LLaMA v2 Chat format? Maybe it's better to save the default format for any tunes on models based on LLaMA Chat (like StellarBright)?
It is good enough for ChatGPT, it is good enough for me.
Yeah, but LLaMA Chat already saw 400G tokens with concrete prompting (over 2T tokens with no prompting), so you basically ruining that gigantic pile of additional fine-tune learning with different prompting IMHO. And maybe that's the part of problems with Dolphin.
It is not.
I really wish more models were ChatML. Thanks for adopting it and making my life easier. 👍
Thank you for the kind words :-)
ehartford
changed discussion status to
closed