<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Local-Llama on SmartStack: AI, Self-Hosting &amp; Smart Finance</title><link>https://www.smart-stacking.com/tags/local-llama/</link><description>Recent content in Local-Llama on SmartStack: AI, Self-Hosting &amp; Smart Finance</description><generator>Hugo</generator><language>en</language><copyright>&amp;copy; 2026 SmartStack</copyright><lastBuildDate>Fri, 11 Sep 2026 16:00:04 +0800</lastBuildDate><atom:link href="https://www.smart-stacking.com/tags/local-llama/index.xml" rel="self" type="application/rss+xml"/><item><title>Replicating V4.1 Flash Prefill on Qwen: How Someone Pulled Fast Prefill Tricks on KV</title><link>https://www.smart-stacking.com/posts/2026-09-11-someone-apparently-managed-to-kind-of-replicate-what-v41-flash-does-on-kv-for-fast-prefill-on-qwen/</link><pubDate>Fri, 11 Sep 2026 16:00:04 +0800</pubDate><guid>https://www.smart-stacking.com/posts/2026-09-11-someone-apparently-managed-to-kind-of-replicate-what-v41-flash-does-on-kv-for-fast-prefill-on-qwen/</guid><description>Explore how one Redditor replicated V4.1 flash prefill behavior on KV for Qwen, including the practical steps to try it yourself.</description></item></channel></rss>