Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

SDXL Turbo: A Real-Time Text-to-Image Generation Model

stability.ai

Post reply on HN
Print view View original

101–110 of 157 posts

Re: SDXL Turbo: A Real-Time Text-to-Image Generation Model

#101

Post by viraptor »

Jackson__ wrote: post #62
kmeisthax wrote: post #48

Noncommercial use - aside from being one of my licensing pet peeves - seems to indicate that the money is drying up. My guess is that the investors over at Stability are tired of subsidizing the part of the generative AI market that OpenAI refuses to touch[0]. The thing is, I'm not entirely sure there's a paying portion of the market? Yes, I've heard of people paying for ChatGPT because it answers programming questio…

The only truly successful commercial use of SDXL I know of is by NovelAI. Said company appears to have used an 256xH100 cluster to finetune it to produce anime art. Open source efforts to produce a similar model seem to have failed due to the extreme compute requirements for finetuning. For example, Waifu Diffusion using 8XA40[0] have not managed to bend SDXL to their will after potentially months of training. If you…

What do you mean by extreme requirements? There's lots of SDXL fine tunings available at civit, like https://civitai.com/models/119012/bluepencil-xl for anime. The relevant discords for models/apps are full of people doing this at home.

Or are you looking at some very specific definition / threshold for fine tuning here?

Re: SDXL Turbo: A Real-Time Text-to-Image Generation Model

#102

Post by thot_experiment »

refulgentis wrote: post #70
__loam wrote: post #66

Earlier quoted context omitted.

[flagged]

?

I can't reply to dead comment, but I'll reply to yours RE above:

> You can rip off other people's work faster than ever!

I have views on IP that mean I reject the ripping off premise on it's face, BUT IF I DID ACCEPT IT

Who am I ripping off? What artist is being denied work if i reskin a VR world on the fly with AI? nobody was going to be painting a scene's worth of textures in a few seconds, the AI is enabling new cool things that haven't been seen before, y'all are absolutely out to lunch replying to a comment specifically calling out brand new things that weren't possible before with this garbage.

Re: SDXL Turbo: A Real-Time Text-to-Image Generation Model

#103

Post by hospitalJail »

torginus wrote: post #91
kmeisthax wrote: post #48

Noncommercial use - aside from being one of my licensing pet peeves - seems to indicate that the money is drying up. My guess is that the investors over at Stability are tired of subsidizing the part of the generative AI market that OpenAI refuses to touch[0]. The thing is, I'm not entirely sure there's a paying portion of the market? Yes, I've heard of people paying for ChatGPT because it answers programming questio…

Isn't literally every imagegen AI that's not DALL-E or Midjourney based on Stable Diffusion?

Are we sure that those arent based on stable diffusion?

No code black box, and we get to tease the closed source companies for wrapping FOSS stuff.

Midjourny I'm most convinced is just a SD with a fine-tuned model. That would explain why everything looks like pixar and can't follow the prompt.

Re: SDXL Turbo: A Real-Time Text-to-Image Generation Model

#104

Post by donkeyd »

viraptor wrote: post #101
Jackson__ wrote: post #62

Earlier quoted context omitted.

The only truly successful commercial use of SDXL I know of is by NovelAI. Said company appears to have used an 256xH100 cluster to finetune it to produce anime art. Open source efforts to produce a similar model seem to have failed due to the extreme compute requirements for finetuning. For example, Waifu Diffusion using 8XA40[0] have not managed to bend SDXL to their will after potentially months of training. If you…

What do you mean by extreme requirements? There's lots of SDXL fine tunings available at civit, like https://civitai.com/models/119012/bluepencil-xl for anime. The relevant discords for models/apps are full of people doing this at home. Or are you looking at some very specific definition / threshold for fine tuning here?

I'm just going out on a limb here, but a paid service needs to be good with limited input. I've used SD locally quite a lot and it takes quite a bit of work through x/y plots to find combinations of settings that produce good images somewhat consistently. Even when using decent fine tunings from CivitAI.

When I use a decent paid service, pretty much every prompt gives me a good response out of the box. Which is good, because otherwise I'd have no use for paid services, since I can run it all locally. This causes me to go to a paid service whenever I want something quick, but don't need full control. When I do want full control, I stick to my local solution, but that takes a lot more time.

Re: SDXL Turbo: A Real-Time Text-to-Image Generation Model

#105

Post by fragmede »

LordDragonfang wrote: post #6
KennyBlanken wrote: post #5

Earlier quoted context omitted.

...which requires you to sign in. That nice little text box invites you until you actually click to enter some text and get a registration box thrust at you People that design a UX where the user tricked into a registration 'ambush' need to be punched in the face.

Luckily this one accepts burner emails just fine, without any intrusive other data collection (name, etc) http://grr.la

That's a useful website, but holy shit it is utterly unusable without an adblocker.

Re: SDXL Turbo: A Real-Time Text-to-Image Generation Model

#106

Post by __loam »

visarga wrote: post #98
__loam wrote: post #90

Earlier quoted context omitted.

These models can't exist without the training sets. Their value is entirely derived from existing data. The ml architecture does not matter at all. Sure, throw enough compute and data at a problem, do a little parallelization, and you can extract plenty of patterns. Does that mean the ml engineers understand art? Or are they just using glorified brute force to alienate people who actually make things from their labor…

If these creations inspire such violent disgust, then it's likely that you perceive them as authentic art. If AI images were devoid of meaning or value, they wouldn't have sparked such passion. You cannot claim ownership over culture, nor can AI. Culture is a collaborative process, and no one can barricade themselves from the input of others. Artists using AI are simply exercising their right to contribute to the col…

This is, imo, an extremely naive take. You claim culture is a collaborative process, yet AI only takes from the communities that produce art. It gives nothing back. You claim AI produces culture, but all it does is atomize our society, promising personal yet meaningless experiences for everyone. There's no shared culture if everyone is just consuming individualized streams of content. It's simultaneously homogenizing too, producing uninteresting torrents of homogenous images from the same model. This also harms culture, stamping out uniqueness under the weight of thousands of meaningless images flooding online art spaces. You claim the only art that's off limits to AI is that which remains unpublished, yet you continue to use the labor of others without permission, discouraging people from publishing their work in the absence of any protection for that work. You claim AI will make art more open, yet most of these models are built and operated by massive corporations with closed source code. They steal from the public and cry out fair use while trying to build walled gardens they can monopolize.

So I'm sorry but there's an argument to every point you're making. I trust the artists I speak to far more than the proponents of this technology. At least they're striving for something genuinely instead of making disengenuous claims about "democratization".

Re: SDXL Turbo: A Real-Time Text-to-Image Generation Model

#107

Post by __loam »

thot_experiment wrote: post #102
refulgentis wrote: post #70

Earlier quoted context omitted.

?

I can't reply to dead comment, but I'll reply to yours RE above: > You can rip off other people's work faster than ever! I have views on IP that mean I reject the ripping off premise on it's face, BUT IF I DID ACCEPT IT Who am I ripping off? What artist is being denied work if i reskin a VR world on the fly with AI? nobody was going to be painting a scene's worth of textures in a few seconds, the AI is enabling new c…

It's so interesting to me that hackernews will flag any comment that dissents to the use of this technology. Supposedly it's irrelevant to the conversation.

The value of these models is derived from the training set, not the ml model. Take away the training data and the model does nothing. So who cares if you reskin your vr world on the fly with AI? I'd argue many artists whose work was ingested into these models without consent care very much about that. Many of them have voiced their concerns publicly. So yeah, go ahead and use this bullshit. But you don't get to just ignore the ethics around it. If you want people like me to shut up and let you enjoy your automated slop, use licensed training sets. Until then, this technology is built on exploitation and alienation.

Also, you don't just get to dismiss intellectual property because you don't like it. It would be awesome if we lived in some utopia where people didn't need to leverage their skills to eat and pay rent. We don't live in that economy, and I don't think corporate AI is going to get us there unless Microsoft shareholders suddenly become bleeding hearts. Nearly every single dev on this site makes their living off of proprietary code, so it's really rich for y'all to just dismiss the concerns of people whose work is being used for this without permission.

Re: SDXL Turbo: A Real-Time Text-to-Image Generation Model

#108

Post by donkeyd »

adventured wrote: post #61
yreg wrote: post #29

Earlier quoted context omitted.

> there was a court ruling What jurisdiction? USA?

Yes - https://www.reuters.com/legal/ai-generated-art-cannot-receiv...

Actually, no. This is just another example of a headline leaving out important details of the actual case. In this case the plaintiff actually named the AI as the producer, not themselves. From the case:

"the sole issue of whether a work generated entirely by an artificial system absent human involvement"

This leaves a lot of wiggle room for AI created art with some form of human involvement. I.E. import the generated image into Photoshop and edit it. Perform some inpainting to improve certain parts. Possibly even prompting and configuring things in Automatic1111 might be regarded as human involvement.

This is going to bring many legal procedures before there's a clear answer to whether AI generated art can be copyrighted.

Re: SDXL Turbo: A Real-Time Text-to-Image Generation Model

#109

Post by donkeyd »

liuliu wrote: post #25

Instruction here for how to use it on iPhone / iPad / Mac today: https://twitter.com/drawthingsapp/status/1729633231526404400 (note that you need to convert 8-bit version yourself if you want to use it on < 8GiB devices).

I remember when you shared the first version on HN, I was totally impressed with what my little phone could do. It was my first step of many into the incredible world of SD. I never expected this app to be maintained this long and especially this actively. And all of this for free.

So thank you very much for your work!

PS, will you be adding Turbo directly to the app in the near future?

Re: SDXL Turbo: A Real-Time Text-to-Image Generation Model

#110

Post by emadm »

kmeisthax wrote: post #48

Noncommercial use - aside from being one of my licensing pet peeves - seems to indicate that the money is drying up. My guess is that the investors over at Stability are tired of subsidizing the part of the generative AI market that OpenAI refuses to touch[0]. The thing is, I'm not entirely sure there's a paying portion of the market? Yes, I've heard of people paying for ChatGPT because it answers programming questio…

We build the best video, image and other models with more downloads and usage than anyone.

It is quite revolutionary for creative industry which is a few hundred billion in size, which is a reasonable market globally.

Post reply on HN