<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>重排器 on 办公AI智能小助手</title>
    <link>https://blog.qife122.com/tags/%E9%87%8D%E6%8E%92%E5%99%A8/</link>
    <description>Recent content in 重排器 on 办公AI智能小助手</description>
    <generator>Hugo</generator>
    <language>zh-cn</language>
    <copyright>qife</copyright>
    <lastBuildDate>Wed, 08 Oct 2025 17:56:28 +0800</lastBuildDate>
    <atom:link href="https://blog.qife122.com/tags/%E9%87%8D%E6%8E%92%E5%99%A8/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>重排器：比LLM更快更省的检索优化技术</title>
      <link>https://blog.qife122.com/p/%E9%87%8D%E6%8E%92%E5%99%A8%E6%AF%94llm%E6%9B%B4%E5%BF%AB%E6%9B%B4%E7%9C%81%E7%9A%84%E6%A3%80%E7%B4%A2%E4%BC%98%E5%8C%96%E6%8A%80%E6%9C%AF/</link>
      <pubDate>Wed, 08 Oct 2025 17:56:28 +0800</pubDate>
      <guid>https://blog.qife122.com/p/%E9%87%8D%E6%8E%92%E5%99%A8%E6%AF%94llm%E6%9B%B4%E5%BF%AB%E6%9B%B4%E7%9C%81%E7%9A%84%E6%A3%80%E7%B4%A2%E4%BC%98%E5%8C%96%E6%8A%80%E6%9C%AF/</guid>
      <description>&lt;p&gt;重排器比大型语言模型（LLM）快几个数量级，且成本更低。在适度规模（约40亿参数）的交叉编码器上，使用廉价GPU（如A10G）的典型开箱即用延迟将低于50毫秒。您无法在该硬件上运行复杂的大型语言模型，且未经调优的情况下，延迟至少需要数百毫秒。&lt;/p&gt;</description>
    </item>
  </channel>
</rss>
