https://hackerpaper.kaipfister.de/p/2204.08583/Collected HN comments and submissions linking this paper.enMon, 02 Nov 2026 11:16:22 GMThttps://news.ycombinator.com/item?id=32420067urn:hackerpapers:mention:2204.08583:32420067Thu, 11 Aug 2022 00:12:42 GMT<p>Stable Diffusion produces substantially higher quality images in most context, but is much more expensive to produce. The genius of VQGAN-CLIP is that it showed that you could take two pre-existing models and combine them to get text-to-image synthesis to work at all. By contrast, models like DALL-E and Stable Diffusion require extremely expensive pretraining. There&#x27;s a discussion of this in the VQGAN-CLIP paper, see in particular 6.1 &quot;Efficiency as a Value&quot; https://arxiv.org/abs/2204.08583 Disclaimer: I&#x27;m one of the authors of the VQGAN-CLIP paper and was tangentially involved with Stable…</p><p><a href="https://hackerpaper.kaipfister.de/p/2204.08583/">Paper and discussion history</a>. A mention is not an endorsement.</p>cs.CVhttps://news.ycombinator.com/item?id=31091725urn:hackerpapers:submission:2204.08583:31091725Wed, 20 Apr 2022 01:32:44 GMT<p>VQGAN-CLIP: Open Domain Image Generation/Editing with Natural Language Guidance</p><p></p><p>17 points; 1 discussion reply on HN. <a href="https://hackerpaper.kaipfister.de/p/2204.08583/">Paper and source history</a>. A submission is not an endorsement.</p>cs.CV