bioemerl

Be warned, prompt processing is slow

Mixtral GPTQ can run on a 3090

Mistral 7b can run on most modern gpus

That’s ( fixed, messed up mah conversion) .1wh for a second of 3090 time/ 30 images a second.

If a 3090 drew 3 watt hours in 1/30th of a second it would melt.

Possibly off by one order of magnitude though… Editing post to see, and it looks like I was. 300 images per charge instead of 3000.

This is outdated in a big way with stable diffusion turbo and the recent LCM models that can render images at 30fps on a 3090.

360w * 1s /60 seconds a minute / 60 minutes an hour = .1 wh/image

30 images a second? .033 wh

A phone battery is 3000 mah * 3.5volts = 10.5 wh

318 images per phone charge

My math is probably off, but you get the idea.

Oh yeah, I’ve had a few brown recluses crawling around in mine.

Also less chance of spiders when you go to bed

The communists are policing speech? Who would have guessed

Do you want AI to exclusively be in the hands of big companies and the government?

Do you want the future of technology locked behind pay walls and censored so that you can’t use it to do anything they don’t want you to do?

If you think AI regulation comes in the form of making sure big companies can’t do bad things to you, you haven’t been paying attention.

Big AI regulation fan.

I’d say good riddance but the replacement is worse

Because I browse the root website and don’t pay any attention to the sub communities. Are you wanting to speak exclusively to like minded individuals who never criticize or oppose your thoughts?

People who want to use Facebook can pay, deal with ads, or just not use the site. This isn’t some huge scandal and nothing is really wrong here

Then don’t use it. I don’t use Facebook every day.

Oh my God you have to pay for a service you use?

The scandal! The outrage! The horror!

This could be fine if it didn’t immediately send all of your data to the internet.

But as is, fuck that and fuck you Microsoft.

Windows told me I don’t have permission to do something. On my computer. As an administrator. Using the command line.

Fuck Windows, fuck Microsoft and their controlling asses, and fuck co-pilot and Open AI for contributing to artificial intelligence not only being closed source and proprietary, but encouraging the United States government to make it literally illegal to do it on the open source field as well.

Because I've seen these things work and I know for damn sure that if those people have a chance to vote for me they wouldn't do it in a million years.

I've learned better than the trust narratives of how you should be a good person.

All I will get is yet more rounds of talk about how everyone else needs help while all the issues I actually care about are left to wither and die.

So fuck them, I'm voting based on the issues that I feel are important and impact my life.

It doesn't really matter whether the original data is present in the model

Yeah it does. One of the arguments people make is that AI models are just a form of compression, and as a result distributing the model is akin to distributing all the component parts. This fact invalidates that argument.

This isn't a slam dunk argument that there's nothing wrong with what an AI does even if we grant it is transformative. It may also simply be proving that the copyright law we have fails to protect artists in the new era of AI.

If we change the law to make it illegal it's illegal.

Over fitting is an issue for the images that were overfit. But note in that article that those images mostly appeared many times in the data set.

People who own the rights to one of those images have a valid argument. Everyone else doesn't.

It is illegal to use copyrighted material period outside of fair use, and this is most certainly not.

Yeah it is. Even assuming fair use applied, fair use is largely a question of how much a work is transformed and (a billion images) -> AI model is just about the most transformative use case out there.

And this assumes this matters when they're literally not copying the original work (barring over fitting). It's a public internet download. The "copy" is made by Facebook or whoever you uploaded the image to.

The model doesn't contain the original artwork or parts of it. Stable diffusion literally has one byte per image of training data.

They use a ton of data as reference points. It’s literally in the name of the technology.

Reference is the wrong word.

They learn the patterns that exist in data and are able to predict future patterns.

They don't actually reference the source material during generation (barring over itting which can happen and is roughly akin to a human memorizing something and reproducing it).