知识卡片
训练阶段优化不同目标
预训练通过后续token预测学习语言规律;指令微调使用任务示例提高遵循请求的能力;偏好训练进一步改变输出选择。三者目标不同,不能把语言流畅、遵循指令与符合偏好当作同一指标。
结构图:
flowchart TD
A[初始模型] --> B[预训练:预测后续词元]
B --> C[指令微调:学习任务响应]
C --> D[偏好训练:调整输出选择]
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
- · 理解推理模型
原书出处 · 1 处
- · 查看原书依据
In short, we can think of pre-training as "raw language prediction" (via next-token prediction) that gives the LLM some basic properties and capabilities to produce coherent texts. The post-training stage then improves the task understanding of LLMs via instruction tuning and refines the LLM to create answers with preferred stylistic choices via preference tuning.
参考来源
- 位置:《Build a Reasoning Model (From Scratch)》第01章《理解推理模型》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Build a Reasoning Model (From Scratch)/_epub-src/OEBPS/Text/chapter-1.html)
- 结论依据:摘录中关于训练阶段优化不同目标的定义、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:In short, we can think of pre-training as "raw language prediction" (via next-token prediction) that gives the LLM some basic properties and capabilities to produce coherent texts. The post-training stage then improves the task understanding of LLMs via instruction tuning and refines the LLM to create answers with preferred stylistic choices via preference tuning.