← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 6766.e · 生成模型:GPT-6

存储类型上限与索引上限须分开

理解等级:中级

向量扩展可对数据类型、维数、非零项以及具体索引分别设限制。某索引最多1000个非零元素,不等于稀疏向量只有1000维;应按安装版本查实际约束,再考虑降维、量化或其他索引,不能照搬单一数字。

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

      How does pgvector compare to proprietary alternatives on benchmarks for throughput, latency, and cost? If you need to compete with the big boys at web scale, you will need to install both pgvector and pgvectorscale plugins. This combination of plugins will give you a significant performance and infrastructure cost advantage over almost any proprietary alternative.48 However, the pgvector field does come with some limitations on

    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:摘录中关于存储类型上限与索引上限须分开的定义、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:How does pgvector compare to proprietary alternatives on benchmarks for throughput, latency, and cost? If you need to compete with the big boys at web scale, you will need to install both pgvector and pgvectorscale plugins. This combination of plugins will give you a significant performance and infrastructure cost advantage over almost any proprietary alternative.48 However, the pgvector field does come with some limitations on