RLHF: Reinforcement Learning from Human Feedback

RLHF: Reinforcement Learning from Human Feedback(huyenchip.com)

4 points by madisonmay 3 years ago | 1 comment

This is a very well written article. Not in the article, but can we still call models like Alpaca RLHF though? What do we call these models finetune on demonstrations created by other chat bots?

heliophobicdude 3 years ago |

This is a very well written article. Not in the article, but can we still call models like Alpaca RLHF though? What do we call these models finetune on demonstrations created by other chat bots?