<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Magika on Zayn's Blog</title><link>https://blog.treesir.pub/tags/magika/</link><description>Recent content in Magika on Zayn's Blog</description><generator>Hugo -- gohugo.io</generator><language>zh-cn</language><managingEditor>yangzun@treesir.pub (Zayn)</managingEditor><webMaster>yangzun@treesir.pub (Zayn)</webMaster><copyright>2021-2026 Zayn</copyright><lastBuildDate>Mon, 08 Jun 2026 21:45:00 +0800</lastBuildDate><atom:link href="https://blog.treesir.pub/tags/magika/index.xml" rel="self" type="application/rss+xml"/><item><title>基于 macvlan 的 JupyterLab GPU 调试实例部署复盘</title><link>https://blog.treesir.pub/posts/jupyterlab-gpu-macvlan-deployment/</link><pubDate>Mon, 08 Jun 2026 21:45:00 +0800</pubDate><author>yangzun@treesir.pub (Zayn)</author><guid>https://blog.treesir.pub/posts/jupyterlab-gpu-macvlan-deployment/</guid><description>记录一个用于 AI 数据集调试的 JupyterLab 实例：容器通过 macvlan 获得独立内网 IP，并在具备 NVIDIA GPU 时支持 Notebook 直接调用 CUDA。</description></item></channel></rss>