知识卡片
文本过滤影响模型学到的格式
书稿中正文、代码、标题和脚注标记的保留方式,会改变训练分布及生成习惯。先确认希望模型产出什么,再选择材料并保留转换记录;剔除部分格式不自动修复正文事实,也不保证其余标记不会被学会。
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
In this chapter, you’ll do a simpler version of fine-tuning—simply feeding your model a particular text corpus to adjust its generation patterns. Some people use this type of fine-tuning, for example, to make the model speak like their favorite character from the movies. In this case, this dataset would be this very book, parsed into a database of lines. Begin by loading it from the
参考来源
- 位置:《Natural Language Processing in Action, Second Edition》第10章《真实世界中的大语言模型》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Natural Language Processing in Action, Second Edition/_epub-src/OEBPS/Text/chapter-10.html)
- 结论依据:摘录中关于文本过滤影响模型学到的格式的定义、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 this chapter, you’ll do a simpler version of fine-tuning—simply feeding your model a particular text corpus to adjust its generation patterns. Some people use this type of fine-tuning, for example, to make the model speak like their favorite character from the movies. In this case, this dataset would be this very book, parsed into a database of lines. Begin by loading it from the