TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

AI model for near-instant image creation on consumer-grade hardware

187 点作者 giuliomagnifico5 个月前

13 条评论

Sharlin5 个月前
For those wondering, it's an adversarially distilled SDXL finetune, not a new base model.
评论 #42382458 未加载
vidarh5 个月前
My favorite test of image models:<p>Drawing of the inside of a cylinder.<p>That&#x27;s usually bad enough. Then try to specify size, and specify things you want to place inside the cylinder relative to the specified size.<p>(e.g. try to approximate an O&#x27;Neill cylinder)<p>I love generative AI models, but they&#x27;re <i>really</i> bad at that, and this one is no exception, but the speed makes playing around with prompt variations to try to see if I get somewhere a lot easier (I&#x27;m not getting anywhere...)
评论 #42384759 未加载
评论 #42386223 未加载
评论 #42386699 未加载
评论 #42387050 未加载
quikoa5 个月前
Github: <a href="https:&#x2F;&#x2F;chendaryen.github.io&#x2F;NitroFusion.github.io&#x2F;" rel="nofollow">https:&#x2F;&#x2F;chendaryen.github.io&#x2F;NitroFusion.github.io&#x2F;</a><p>Paper: <a href="https:&#x2F;&#x2F;arxiv.org&#x2F;html&#x2F;2412.02030v2" rel="nofollow">https:&#x2F;&#x2F;arxiv.org&#x2F;html&#x2F;2412.02030v2</a>
iLoveOncall5 个月前
&gt; Instant image generation that responds as users type – a first in the field<p>Stable Diffusion Turbo has been able to do this for more than a year, even on my &quot;mere&quot; RTX 3080.
评论 #42383267 未加载
ajdjspaj5 个月前
What does consumer-grade mean in this context - is this referring to an M1 MacBook or a tower full of GPUs? I couldn&#x27;t find in the paper or README.
评论 #42384054 未加载
ericra5 个月前
I wasn&#x27;t able to get many decent results after playing with the demo for some time. I guess my question is...what exactly is this for? I was able to get substantially better results about 2 years ago running SD 2 locally on a gaming laptop. Sure, the images took 30 seconds or so each, but the quality was better than I could get in the demo. Not sure what the point of instantly generating a ton of bad quality images is.<p>What am I missing?
评论 #42383779 未加载
评论 #42385277 未加载
betenoire5 个月前
Here is the demo <a href="https:&#x2F;&#x2F;huggingface.co&#x2F;spaces&#x2F;ChenDY&#x2F;NitroFusion_1step_T2I" rel="nofollow">https:&#x2F;&#x2F;huggingface.co&#x2F;spaces&#x2F;ChenDY&#x2F;NitroFusion_1step_T2I</a><p>I&#x27;m unable to get anything that looks as good as the images in the README, what&#x27;s the trick for good image prompts?
评论 #42380773 未加载
评论 #42380654 未加载
评论 #42380420 未加载
评论 #42386176 未加载
tgsovlerkhgsel5 个月前
The models seem to have gotten to a point where even something I can run locally will give decent results in a reasonable time. What is currently &quot;the best&quot; (both from an output quality and ease of installation perspective) setup to just play with local a) image generation, b) image editing?
评论 #42382347 未加载
评论 #42383167 未加载
评论 #42383586 未加载
评论 #42382915 未加载
评论 #42382573 未加载
评论 #42382376 未加载
nprateem5 个月前
The devil&#x27;s in the details as always. A &quot;cartoon of a cat eating an icecream on a unicycle&quot; doesn&#x27;t bring back any of the 6-pawed mutant cats riding a unicycle, etc. Still, impressive speed.
评论 #42391329 未加载
wruza5 个月前
Isn’t this a year old news?<p>It was called LCM&#x2F;Turbo in SD and it generated absolute crap most of the times, just like this one. Which is likely yet another “ground-breaking” finetune of SD.
musicale5 个月前
&gt; Surrey announces world&#x27;s first AI model for near-instant image creation on consumer-grade hardware<p>Kind of like what you can do on an iPhone?
smusamashah5 个月前
Stream Diffusion already exists and gives you images as you type. Worked fine on RTX 3080
gloosx5 个月前
creation... wow, they really love themselves by choosing that vocabulary; to create is divine, and this AI model is merely generating.