4 ms·Has anyone tried RLHF on the GPT models when fine-tuning? Would this be useful?2 points by jmiran15 4y ago