← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 6766.a · 生成模型:GPT-6

精确检索仍有条件适用空间

理解等级:中级

穷举距离是通用精确基线,并行批处理可在中小规模很有效;特定维度与数据也存在可剪枝的精确结构。规模扩大后应实测再选择ANN,不能把所有精确查找说成唯有逐项点积,或断言几千向量必然不可行。

这张卡片看不懂?

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

      There is only one way to find the exact nearest neighbor for your query. Remember how we discussed exhaustive search in chapter 4? Back then, we found the nearest neighbor of the search query by computing its dot product with every vector in the database. That was okay because back then, your database included only a couple dozen vectors. It won’t scale to a database

    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:摘录中关于精确检索仍有条件适用空间的定义、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:There is only one way to find the exact nearest neighbor for your query. Remember how we discussed exhaustive search in chapter 4? Back then, we found the nearest neighbor of the search query by computing its dot product with every vector in the database. That was okay because back then, your database included only a couple dozen vectors. It won’t scale to a database