OpenRLHF icon indicating copy to clipboard operation
OpenRLHF copied to clipboard

agent_func、reward_func时--normalization_reward参数不起作用,对吗

Open YSQ-boop opened this issue 6 months ago • 1 comments

agent_func、reward_func时--normalization_reward参数不起作用,对吗 我是否也可以理解为normalization_reward其实是normalization_value,只有包含critic model时才生效

YSQ-boop avatar Aug 16 '25 05:08 YSQ-boop

hijkzzz avatar Aug 16 '25 15:08 hijkzzz