<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>The Missing U for Efficient Diffusion Models · HackerPapers discussions</title><link>https://hackerpaper.kaipfister.de/p/2310.20092/</link><description>Collected HN comments and submissions linking this paper.</description><language>en</language><lastBuildDate>Mon, 02 Nov 2026 11:16:22 GMT</lastBuildDate><atom:link href="https://hackerpaper.kaipfister.de/p/2310.20092/feed.xml" rel="self" type="application/rss+xml" /><item><title>The Missing U for Efficient Diffusion Models — mentioned by orbital-decay</title><link>https://news.ycombinator.com/item?id=38352800</link><guid isPermaLink="false">urn:hackerpapers:mention:2310.20092:38352800</guid><pubDate>Mon, 20 Nov 2023 19:06:31 GMT</pubDate><description>&lt;p&gt;Commissions are professional work. If you need speed to get things done, you can pay to rent the GPU and the cost will be negligible to what you earn, even if you&amp;#x27;re just starting. This is really not that much of a barrier compared to the hoops hobbyist 3D artists had to jump through 20 years ago. Regardless, you can run SD on a several years old laptop just fine, this is entirely within the reach of most; yes you won&amp;#x27;t be getting realtime updates but that&amp;#x27;s not really necessary. And that&amp;#x27;s only the beginning. SD was trained using really poor data; everybody is doing that on semi-synthetic…&lt;/p&gt;&lt;p&gt;&lt;a href="https://hackerpaper.kaipfister.de/p/2310.20092/"&gt;Paper and discussion history&lt;/a&gt;. A mention is not an endorsement.&lt;/p&gt;</description><category>cs.LG</category></item></channel></rss>
