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

Implementation of Imagen, Google's text-to-image neural network, in PyTorch

github.com

Post reply on HN
Print view View original

71–80 of 123 posts

Re: Implementation of Imagen, Google's text-to-image neural network, in PyTorch

#71

Post by adamhowell »

nautilus12 wrote: post #66
adamhowell wrote: post #57

Earlier quoted context omitted.

I'm currently working fulltime on AI-powered design suite Accomplice ( https://accomplice.ai ) and if you ask me on a good day I would tell you I do think there's already huge commercial potential. On a bad day, though ;) My current approach is a "model marketplace" ( https://accomplice.ai/models ) where the most popular open source text-to-image models (VQGAN+CLIP, Disco Diffusion, DALL-E Mega coming soon…), sit alo…

Why would people do this when aspiring artists are practically giving their real photos and paintings for free on places like deviant art? Why further commoditize something that's already been commoditize to practically free?

Well, I'm focusing on business use cases – stock images and how Accomplice could be useful for marketing and content creation.

i.e. The ability to easily take your logo and stylize it: https://accomplice.ai/@adam/iterations/2bcc90ad-3237-486a-8d...

Create a photorealistic avatar whenever you need it: https://accomplice.ai/@adam/iterations/988b7d54-dc39-43b1-b5...

Easily remove the background of a photo: https://accomplice.ai/models/97746c4b-c6f0-49cb-ae1b-859716b...

Upscale a photo: https://accomplice.ai/models/bd4619ee-8202-4cf0-a04e-291820f...

Etc etc. AI can make all this stuff easier. And you have a sense of ownership over what you create. All in one place where you can collaborate on all of it with your team. I feel like that's valuable. It's certainly a tool I've always wanted.

But, also, as a bit of an aside – if the Googles and OpenAIs of the world are just going to bite every artist's style anyway with a mostly black box service and training set… it feels like the option for an artist to train/finetune their own model, promote it and possibly make money off of that is worth trying.

Re: Implementation of Imagen, Google's text-to-image neural network, in PyTorch

#72

Post by Kleto »

aqme28 wrote: post #59
thorum wrote: post #47

Earlier quoted context omitted.

Does anyone actually want porn of fake, AI generated “people”? Seems like most of the demand would be deepfakes of real people, which is both highly unethical and a good way to get your business sued out of existence.

There’s also seemingly unlimited content of real people produced consensually by those people. What can DALL-E generate that you can’t find on pornhub or onlyfans??

They cost money.

They involve real woman (ethical concerns).

They might not show exactly what you are looking for.

Like the good scene is to short or the quality is too bad or there is only one video.

Also variation.

ML porn should be a good thing. Only thing Im not sure is about people creating pedo porn. But even that is better than real pedo porn :|

Re: Implementation of Imagen, Google's text-to-image neural network, in PyTorch

#73

Post by benstrumental »

29athrowaway wrote: post #35

I guess the stock photos industry will be disrupted by this.

Definitely. With the right training set, video game assets too. Imagine being able to generate N variations of any asset in any video game style...

Re: Implementation of Imagen, Google's text-to-image neural network, in PyTorch

#74

Post by aero-glide2 »

jfoster wrote: post #60
jonathankoren wrote: post #40

Of course the implementation isn’t the issue. It’s the training data and the compute machines. Open source is pretty meaningless here

In fact, this kind of reverses things, doesn't it? Open source is built on the assumption that you can do more with source code than with binaries. In the case of AI models, the computed weights of models are what's valuable, and the source code used to achieve them is less useful.

How much would it cost in training to match dall-e 2?

Re: Implementation of Imagen, Google's text-to-image neural network, in PyTorch

#75

Post by rockemsockem »

sailingparrot wrote: post #53
Yenrabbit wrote: post #44

Earlier quoted context omitted.

Four or five figures I'd guess? I'm not clued up on costs/performance for TPU stuff to give a better estimate, but guessing at a week on a 256 TPU pod, call it $30k?

You are off by an order of magnitude at least. 256 TPUs-v4 (not pod), would cost you around 20k$/day. They actually used 512 TPUs (256 for base model + 128 for each of the two superresolution models). Assuming an average training time of 1 week as you said, that gives us about 280k$. It's also most likely trained for longer than a week, the base model for Dalle-2 was trained for 100-200k GPU hours, so between 2-4x lo…

While that is what they did, they also used a batch size of 2048 while training. This is just to speed training up, not a hard requirement. It's easy for Google to justify more money on compute to save engineer iteration loops.

I'll have to read the paper for more details, but it would almost certainly cost less (and take longer) to train a model like this in a more resource constrained situation than Google faces .

Re: Implementation of Imagen, Google's text-to-image neural network, in PyTorch

#76

Post by gbasin »

adamhowell wrote: post #57
nmfisher wrote: post #42

Earlier quoted context omitted.

Two reasons: 1) Even though it's all technically very impressive, so far there's not a huge amount of commercialization potential here. OpenAI is charging for its GPT-3 model but its revenue is probably negligible next to the hardware costs (sunk + ongoing) to train it in the first place, let alone the researcher salaries they're paying 2) Most of the stunning examples are cherry-picked. These things fail much more o…

I'm currently working fulltime on AI-powered design suite Accomplice ( https://accomplice.ai ) and if you ask me on a good day I would tell you I do think there's already huge commercial potential. On a bad day, though ;) My current approach is a "model marketplace" ( https://accomplice.ai/models ) where the most popular open source text-to-image models (VQGAN+CLIP, Disco Diffusion, DALL-E Mega coming soon…), sit alo…

Cool project! Will follow along

Re: Implementation of Imagen, Google's text-to-image neural network, in PyTorch

#78

Post by jermaustin1 »

adamhowell wrote: post #70
TekMol wrote: post #65

Earlier quoted context omitted.

I wanted to try your site, but after clicking on the link in the email, it tries to send me through some redirectingat.com link which is blocked by my ad blocker. What does that link do?

Update: Ah, found the "Tracking" setting in Sendgrid that I thought I had already turned off. Off for sure now. Thanks again for the heads up! The confirmation link should only be going to accomplice.ai unless Sendgrid is doing some link tracking that I've just forgotten about. Could you forward that email to adam at accomplice dot ai if you get a chance. Thanks for letting me know!

That will get turned back on automatically in an "update".

Same thing happened to me multiple times across multiple platforms: SendGrid, Mandrill, MailJet, MailGun. I always turn off the tracking (enabled by default on all of them), but magically its back on a few weeks/months later. I've given up finding a solution and just revisit my settings every few months to check on it.

Re: Implementation of Imagen, Google's text-to-image neural network, in PyTorch

#79

Post by 29athrowaway »

lajamerr wrote: post #69
29athrowaway wrote: post #68

Earlier quoted context omitted.

How can this be the most upvoted comment? What you describe is already illegal on many jurisdictions.

Which part is illegal? Are you talking about porn? Or using potential copyright material in an infringing manner? Or perhaps the possibility of "deepfake" porn?

In some jurisdictions, deepfakes of any kind are treated as a form of defamation. In others, adult deepfakes specifically are illegal.

Re: Implementation of Imagen, Google's text-to-image neural network, in PyTorch

#80

Post by gfodor »

nmfisher wrote: post #42
TekMol wrote: post #16

What is the reason Google published their research details about Imagen? Why don't they just keep their findings to themselfes and build products on top of them? Public companies can't do stuff just for the fun of it, right? So there must be some commercial reasoning behind it?

Two reasons: 1) Even though it's all technically very impressive, so far there's not a huge amount of commercialization potential here. OpenAI is charging for its GPT-3 model but its revenue is probably negligible next to the hardware costs (sunk + ongoing) to train it in the first place, let alone the researcher salaries they're paying 2) Most of the stunning examples are cherry-picked. These things fail much more o…

Seems exactly false. DALL-E 2 seems to end much of the illustrator industry and if the endless array of Twitter posts from early adopters are any indication it works great.
Post reply on HN