<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>kserve on Yang Cao</title><link>https://yc2984.github.io/tags/kserve/</link><description>Recent content in kserve on Yang Cao</description><generator>Hugo</generator><language>en-gb</language><lastBuildDate>Tue, 02 Apr 2024 10:00:00 +0200</lastBuildDate><atom:link href="https://yc2984.github.io/tags/kserve/index.xml" rel="self" type="application/rss+xml"/><item><title>Serving models on Kubernetes with KServe</title><link>https://yc2984.github.io/posts/serving-models-on-kubernetes-with-kserve/</link><pubDate>Tue, 02 Apr 2024 10:00:00 +0200</pubDate><guid>https://yc2984.github.io/posts/serving-models-on-kubernetes-with-kserve/</guid><description>Scale-to-zero sounds free until you meet the cold start. What KServe actually buys you, what it quietly costs, and why we turned the headline feature off in the end.</description></item></channel></rss>