<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>GGUF on 知识铺的博客</title>
    <link>https://index.zshipu.com/geek/tags/GGUF/</link>
    <description>Recent content in GGUF on 知识铺的博客</description>
    <generator>Hugo -- gohugo.io</generator>
    <language>zh-CN</language>
    <lastBuildDate>Thu, 20 Aug 2026 00:00:00 +0000</lastBuildDate>
    <atom:link href="https://index.zshipu.com/geek/tags/GGUF/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Unsloth Dynamic 3.0 GGUF发布：同体积下准确率提升10%</title>
      <link>https://index.zshipu.com/geek/post/20260820/Unsloth-Dynamic-3.0-GGUF%E5%8F%91%E5%B8%83%E5%90%8C%E4%BD%93%E7%A7%AF%E4%B8%8B%E5%87%86%E7%A1%AE%E7%8E%87%E6%8F%90%E5%8D%8710/</link>
      <pubDate>Thu, 20 Aug 2026 00:00:00 +0000</pubDate>
      <guid>https://index.zshipu.com/geek/post/20260820/Unsloth-Dynamic-3.0-GGUF%E5%8F%91%E5%B8%83%E5%90%8C%E4%BD%93%E7%A7%AF%E4%B8%8B%E5%87%86%E7%A1%AE%E7%8E%87%E6%8F%90%E5%8D%8710/</guid>
      <description>Unsloth Dynamic 3.0 GGUF发布：同体积下准确率提升10% Unsloth 今天发布了 Dynamic 3.0 量化方案，首发 Qwen3.8-27B 量化版本。相比 v2.0，新版本在同等磁盘空间下 top-1% 准确率提升超过 10%，KL Divergence 指标全面优于所有其他量化提供商。 改进原理：更聪明的量化 量化技术通过降低模型参数的精度来减小文件体积，但往往以牺牲准确率为代价。U</description>
    </item>
  </channel>
</rss>
