<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:content="http://purl.org/rss/1.0/modules/content/">
  <channel>
    <title>R1 on Osman&#39;s Odyssey: Byte &amp; Build</title>
    <link>https://www.ahmadosman.com/tags/r1/</link>
    <description>Recent content in R1 on Osman&#39;s Odyssey: Byte &amp; Build</description>
    <image>
      <title>Osman&#39;s Odyssey: Byte &amp; Build</title>
      <url>https://www.ahmadosman.com/logo/byte-and-build.png</url>
      <link>https://www.ahmadosman.com/logo/byte-and-build.png</link>
    </image>
    <generator>Hugo -- 0.145.0</generator>
    <language>en-us</language>
    <lastBuildDate>Fri, 14 Feb 2025 02:56:56 -0600</lastBuildDate>
    <atom:link href="https://www.ahmadosman.com/tags/r1/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>Key Highlights From Running DeepSeek R-1 671B on 14x RTX 3090s &#43; Epyc 7713 &amp; 512GB RAM</title>
      <link>https://www.ahmadosman.com/blog/r1-ktransformers-inference-livestream/</link>
      <pubDate>Fri, 14 Feb 2025 02:56:56 -0600</pubDate>
      <guid>https://www.ahmadosman.com/blog/r1-ktransformers-inference-livestream/</guid>
      <description>Key takeaways from livestreaming DeepSeek R-1 671B (4-bit) on a 14x RTX 3090 basement AI server. See how KTransformers crushed llama.cpp in prompt eval speeds, compare setups, and get real-world insights into massive LLM inference with vLLM, ExLlamaV2, and more.</description>
    </item>
  </channel>
</rss>
