<?xml version="1.0" encoding="UTF-8"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Yestino - The Signal · FlyDSL FlexAttention</title><link>https://yestino.com/zh-CN/entities/flydsl-flexattention-ae01d0</link><description>与 FlyDSL FlexAttention 相关的全部事件</description><language>zh-CN</language><atom:link href="https://yestino.com/zh-CN/entities/flydsl-flexattention-ae01d0/feed.xml" rel="self" type="application/rss+xml"/><item><title>trunk/2208903bd2a09f30786d54b3d2426c98013d0159: [Inductor] Add gfx950 FlyDSL FlexAttention forward (#194309)</title><link>https://yestino.com/zh-CN/events/trunk-2208903bd2a09f30786d54b3d2426c98013d0159-inductor-add-1610cb</link><guid isPermaLink="true">https://yestino.com/zh-CN/events/trunk-2208903bd2a09f30786d54b3d2426c98013d0159-inductor-add-1610cb</guid><pubDate>Sat, 10 Oct 2026 04:16:54 GMT</pubDate><description>I reran these benchmarks on the pinned PyTorch main snapshot and reviewed the results; the updated tables document the final decode coverage and performance.
来源：PyTorch Releases</description></item></channel></rss>