知识卡片
组内基线替代额外价值模型
组相对方法对同一题采样多个回答,以奖励均值作为基线形成相对学习信号,从而避免为此训练单独价值模型。节省这部分开销并未消除多条生成、可微评分、优化器状态和激活所需资源。
这张卡片看不懂?
这张卡片已通过人工审查,不再接受失败反馈。
已有反馈 · 0 条
还没有人提交反馈。
在整书笔记中的位置
原书出处 · 1 处
- · 查看原书依据
does not require a separate value model to estimate a value function. Instead, GRPO derives its learning signal from relative comparisons within a group of sampled responses, which substantially reduces computational overhead.
参考来源
- 位置:《Build a Reasoning Model (From Scratch)》第06章《强化学习训练推理模型》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Build a Reasoning Model (From Scratch)/_epub-src/OEBPS/Text/chapter-6.html)
- 结论依据:摘录中关于组内基线替代额外价值模型的定义、比较或条件支撑卡片;原书表述不严谨的范围已在覆盖记录逐项限定,不把案例结果外推为普遍结论。
- 原始内容:does not require a separate value model to estimate a value function. Instead, GRPO derives its learning signal from relative comparisons within a group of sampled responses, which substantially reduces computational overhead.