I’m guessing there’s a mix. The smallest version is 700 million, possibly the one used to generate the time data reported, but the largest (or not?) still runs with 8gb. If I remember correctly SD3 is supposed to have multiple versions, starting from 800 millions and going up, so this is going to be interesting.
And thanks for all the fishes!
My understanding (potentially obsolete!) is that the integrated gpu of the Deck isn’t supported by, uh… whatever the AMD equivalent of Cuda was called, also there’s a chance you’d have to install it manually too.
It amuses me how that sentence could be a worst case scenario on InfoWars lol
He needs a supporting cast to pull it off, I’m thinking Pedro Pascal… ideally as a bounty hunter.
I’ve never had issues with Cura losing my stuff (been using it with an Ender 3 since… 2019 I think) but the backup plugin is connected with your account. It’s a wonderful thing! New computer? Login, restore backup, everything is like you left it.
I do make sure to use it before any update because I have seen the same kind of posts as you have.
Perfectly sensible is the only thing I can say about it :)
Depending on your hardware you might be able to generate images without using any service! For all its quirks, downsides and moral stuff, it can be a lot of fun to play with. If you’re on Apple hardware, try my favorite tool, Draw Things https://apps.apple.com/it/app/draw-things-ai-generation/id6444050820?l=en-GB that is free, without ads and worst case scenario it runs like crap on your device if it’s not very modern. If you’re not using Apple gear, Enfugue (that still supports Macs) https://github.com/painebenjamin/app.enfugue.ai is quite simple and easy to install.
And obviously what’s probably the most popular option, the behemoth that is known as automatic1111 stable diffusion webui https://github.com/AUTOMATIC1111/stable-diffusion-webui that’s not the most user friendly solution but…
That’s actually a good example, “any of them”. Unless the model is trained by someone with specific interest in Star Trek, the Enterprise has… what, 60 years of different reinterpretations? I enjoy generative AI, and made let’s call it an extension to be able to make the old Fiat Panda. Without it, you’d get either the new one, or randomness. And with my efforts now it’s possible to get-ish it. My dataset wasn’t the greatest quality or variety, plus I didn’t put a ton of effort into tagging the single pictures so it’s not very flexible or particularly good. It works sufficiently well and I’m happy with it, but I’m very aware of its shortcomings… namely that I didn’t tag any picture with the orientation of the car. It is however a car so it has a basic framework to show it, while the Enterprise? Not the most simple subject to reproduce. Anyway you don’t really care, sorry for that minor wall of text :D
Holy shit, no kidding. Some actors it likes, others… I once tried to make a group photo with all the captains, I tried many different models, lots of tricks (plenty of advancements since then, I could probably pull it off nowadays) and it was a disaster. I’ll post one or two of those if I can find them again, absolute waste of an afternoon.
In Voyager I had to “delete” my account from the interface and then add it back again before it worked again. Now in fairness the button shows “log out” while doing so, but still it was an unusual step so probably nice for others to know. :)
So let me get this straight… The thing is trained on text harvested from the internet. Turns out that (Star Trek aside) if you ask politely it’s more likely to access the data leading to the answer? Anyone surprised by the fact that asking nicely, is more likely that the human on the other side of the connection will write the answer, that is then harvested for AI training?