Stable Diffusion, NovelAI, Machine Learning Art - AI art generation discussion and image dump

  • 🐕 I am attempting to get the site runnning as fast as possible. If you are experiencing slow page load times, please report it.
At last, my GPU upgrade is complete.
View attachment 3751926
Meanwhile I have to live with my poor old 1060 having 1/10th of that speed. :[
1666290470248.png
 
This program is pretty incredible. I have not yet used inpainting, but the text to image alone produces such high quality art. I had a friend request some psychedelic cyber space cat art. I also tested out the random artist feature requesting what Death would look like.

Cyber Space Cats:
00099-1376358193-masterpiece, best quality, masterpiece, psychedelic, cybernetic space cats.png
00098-4014430970-masterpiece, best quality, masterpiece, psychedelic, cybernetic space cats.png

Death:
00067-1359864781-masterpiece, best quality, masterpiece, Death, Shotaro Ishinomori.png

ps.
If anyone can help me with troubleshooting the NovaAi weight configuration with Asuka sitting in a chair crossed legs it would be greatly appreciate it. I followed the listed guide and get different results. I am aware this thread is sfw, so I rather not post my results in this thread. Send me a DM. Thank you!
 
Can someone help me out here? I'd like to make a very scary bloody border collie with a blood red background and dim lighting with img2img using NovelAI, but I have no idea what I'm doing and so far the results aren't really that impressive. Could I get some hints from people who know how to use NovelAI properly?
2022-10-20 20.51.41 127.0.0.1 a276c81759c6.png
 
Can someone help me out here? I'd like to make a very scary bloody border collie with a blood red background and dim lighting with img2img using NovelAI, but I have no idea what I'm doing and so far the results aren't really that impressive. Could I get some hints from people who know how to use NovelAI properly?
View attachment 3752781
Still learning, but I hear that tokens that come first in the prompt are checked more heavily, and you could try sandwiching the words you want emphasised with (). If that still doesn't work you could try inpainting and masking the background, then give a prompt like red background.
 
Can someone help a brainlet?
I make a batch of say 9 images with a random seed. I want to know the seed for one of those 9 images. But all 9 images show the same seed. Is it impossible to generate a single image again or with different settings, if it comes from a batch?
 
What does inpainting do? For me it just seems to erase parts of the picture.
 
What does inpainting do? For me it just seems to erase parts of the picture.
It basically reworks the mask (the part you draw in with the masking tool) into your prompt as best as it can. Here's a tutorial vid for beginners:

Not sure what the major difference in masked content is.

On that note, does anyone know if this—removing people from photos and filling in the space with what's behind them—is possible on the web UI?
 
Can someone help me out here? I'd like to make a very scary bloody border collie with a blood red background and dim lighting with img2img using NovelAI, but I have no idea what I'm doing and so far the results aren't really that impressive. Could I get some hints from people who know how to use NovelAI properly?
View attachment 3752781
Try to lower CFG scale and denoising strength.
Also bigger batch size would mean it will generate multiple pics at once, if your videocard can handle that, increasing the odds of a good result.
 
Really wish GPUS didn't cost an arm and a leg, I regret not upgrading earlier. Especially when they cost like 200% more than they should for no reason at all.
I was never happier I bought my 3060 when I did. All that work would've been too much for a 1060.
1666245694486488.png1666261401148800.png
 
Last edited by a moderator:
Things will change, they always did. Computers changed the world, and AI will probably change it in very impactful ways again. I see a lot of potential to improve the quality of life for everyone, if used correctly.
Since when have the "powers that be" ever used technology as a means to improve the quality of life for the masses? I see AI as a weapon of potentially apocalyptic consequence. The manipulation, subterfuge, distraction, exploitation and derision capable with it could bring about a dystopic societal shift of the kind only seen so far in the most pessimistic views produced for science-fiction. I think the people making this technology under the principle of "benefiting humanity" have no idea how much damage it actually has the potential to cause and everyone else participating in making it knows exactly how much damage it has the potential to cause and for them it is the great driving force in their quest to make it a reality.
 
Last edited:
Since when have the "powers that be" ever used technology as a means to improve the quality of life for the masses? I see AI as a weapon of potentially apocalyptic consequence. The manipulation, subterfuge, distraction, exploitation and derision capable with it could bring about a dystopic societal shift of the kind only seen so far in the most pessimistic views produced for science-fiction. I think the people making this technology, at least the ones doing it under the guise of "benefiting humanity" have no idea how much damage it actually has the potential to cause.
I'll take doomposting for 500 Alex.

If it's such a fucking dystopia, just kill yourself. Before the "powers that be" take that away from you too. Whining about shit effectively does nothing. All you can do is help your friends and family.

Bro trust me, when the aliens come down for the centurial checkup things will only get worse.
 
I draw sometimes, I use AI for inspiration purposes only.
Like, when I'm going through an art block, I tell the AI some weird ass prompt, and then I redraw the prompt in my own art style from scratch.

Sometimes I'm thinking about Barack Obama playing Metal Slug 2 in the outer space, I'm still unable to think where everything should be and how they should like, but the AI gives me what could've been a sketch. A pre-sketch actually.
 
Back