<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Moe on skulldc</title><link>https://skulldc.net/tags/moe/</link><description>Recent content in Moe on skulldc</description><generator>Hugo</generator><language>en-us</language><lastBuildDate>Tue, 03 Nov 2026 10:27:42 +0000</lastBuildDate><atom:link href="https://skulldc.net/tags/moe/index.xml" rel="self" type="application/rss+xml"/><item><title>Colibri runs huge mixture-of-experts models by streaming experts from disk</title><link>https://skulldc.net/2026/11/colibri-runs-huge-mixture-of-experts-models-by-streaming-experts-from/</link><pubDate>Tue, 03 Nov 2026 10:27:42 +0000</pubDate><guid>https://skulldc.net/2026/11/colibri-runs-huge-mixture-of-experts-models-by-streaming-experts-from/</guid><description>A C inference engine keeps an MoE model&amp;rsquo;s dense part in RAM, reads routed experts from disk on demand and puts the hottest ones on a GPU if present.</description></item></channel></rss>