Is 1.1 trained from the same SFT model as 1.0?
#18
by
chujiezheng
- opened
Thanks for your work. Is gemma-1.1-7b-it trained from the same SFT model as gemma-7b-it?
Hi
@chujiezheng
, Gemma-1.1-7b-itis an improved version of Gemma-7b-it that benefits from additional training and optimization using Reinforcement Learning from Human Feedback (RLHF). Please have a look at this gemma-1.1-7b-it release notes for more details. Thank you.