<?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>Slime：架构与实践目录 on dextermayhewjd</title><link>https://dextermayhewjd.github.io/projects/slime/</link><description>Recent content in Slime：架构与实践目录 on dextermayhewjd</description><generator>Hugo</generator><language>zh-cn</language><atom:link href="https://dextermayhewjd.github.io/projects/slime/index.xml" rel="self" type="application/rss+xml"/><item><title>Slime 源码阅读顺序</title><link>https://dextermayhewjd.github.io/projects/slime/code-reading/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>https://dextermayhewjd.github.io/projects/slime/code-reading/</guid><description>从模块职责到任务、模型、评测、Trace、并发、部署与训练交接的源码阅读路线。</description></item><item><title>Slime 全仓代码复用与项目路线</title><link>https://dextermayhewjd.github.io/projects/slime/reuse-roadmap/</link><pubDate>Mon, 01 Jan 0001 00:00:00 +0000</pubDate><guid>https://dextermayhewjd.github.io/projects/slime/reuse-roadmap/</guid><description>梳理 Slime 全仓可复用的执行、评测、Trace、数据队列、模型服务与训练能力，以及 API、Kubernetes 和自部署模型的接入边界。</description></item></channel></rss>