/cyb/ - Cyberpunk Fiction and Fact

>>2182
>It makes sense you don't always get the same results, even when all things are equal.
Running same seed and prompt I get same image every time. Also running others prompt and seeds (I seen people post) I get same image they get when using the same model they used.

>>2182
I don't think the AI is updating its own model with data from the pictures it generates.
I hope it isn't. I don't want my AI to give itself dementia by using its garbled output as new input.
By now its garbled output probably outnumbers the number of images it was trained on.

Does anyone have any good images? I tried doing a cursory search on ponerpics and the first few pages were horrid dogshit.

The era of abstract merchants has reached a new golden age.
Merchants are being produced faster than the ADL could ever hope to flag them.

>>2234
How do you make these?
I've got Stable Diffusion working, what's the secret?

>>2235
Idk. I got them from Discord.
If you find out any methods, do share. I want to mass produce merchants.

>>1884
That's not what I meant when I said I wanted racks on racks.

>>2241
Racks on racks on racks on racks. Stacked rat stacks in hats on mats.

Pixiv has lost its collective mind recently (more than before).
It's impossible to use the site without being recommended pics of [i]pregnant toddlers!

>>2248
Oh fuck. Don't some models use data taken from pixiv or anime booru sites?

>>2250
>you will never haiy her lcat or her spry aliet hany
Why live?

I got the new Everything V3 model.
It seems more varied than the previous smaller Waifu model I was using.
But it also loves to wash out the colours of anything I make with it. So I often have to saturate this desaturated art output until it looks right.

"Twilight Sparkle"

>>2273
Nice. Not knowing prompt but adding "(unicorn)" to prompt and "(text), devil" to negative prompt might fix some of the devil horns and and text. Love how AI art is evolving with better and better models.

>>2275
Wow. It looks fantastic.
Is this using upscaling and inpainting or...? if you know.

>>2274
Speaking of AI evolution how do I update my installation of Stable Diffusion? I heard the new version has no limit on text input and is better at drawing hands.

>>2277
The way I do it is that I used git to get the initial code and run an git update in the bat file used to start SD.

So initially do an
>git clone https://github.com/AUTOMATIC1111/stable-diffusion-webui

Then in bat file I do
>...
>set COMMANDLINE_ARGS=--ckpt pony_sfw_80k_safe_and_suggestive_500rating_plus-pruned.ckpt --lowvram --opt-split-attention
>
>cd stable-diffusion-webui
>git pull
>call webui.bat
>cd ..

Wtf is he doing to edit the AI generated art and get new ones like that? https://m.youtube.com/watch?v=aFnGO74KsF0

>>2281
I assumed he did, but looked at description and he is using the Krita plugin for Stable Diffusion that runs an img2img with prompt. It can take basic input and you have to tell it what it should draw and weight how much it should resemble the input image. I am not fully sure how the weights should be but I think the Krita plugin is fairly easy to use.
First get Krita https://krita.org/
Then install the Krita plugin https://github.com/sddebz/stable-diffusion-krita-plugin

You can do the same in Stable Diffusion webui directly by going to the img2img tab.

>>2281
>>2282
I see that he also uses inpaint to fix areas.
I don't have full knowledge on how to use it effectively as my computer takes forever to generate a single image so I can't really play around with it without it taking forever.

>>2283
Have you tried using that setting where your generator makes the image tiny and then scales it up?

>>2287
I have not tried upscaling or any of the fancy options as the biggest I can generate is 512x512 and even that is pushing my card to the limits. And it takes a couple of minutes to generate an image. So experimenting with settings and what it does is an exercise in patience and remembering what I did half an hour ago and if my changes actually had an impact.

>>2288
Damn, that sucks. Back when I had to make do with a piece of shit laptop that would take multiple minutes to boot up and do basic tasks like open the file explorer or open a webpage, I ran Fallout New Vegas at barely 10 frames a second even with the lowest graphics settings and as few mods as possible. The game was borderline unplayable especially when combat started so I had to rely on companions basically doing all combat for me. The lag didn't fuck their attacks up as much. Even the fucking word processor lagged with that thing. A word processor!

Does anyone actually believe GPU prices are dropping down anytime soon?
>>2288
Specs?
>>2291
Lmao
My trashtop struggles to run Halo 2 on anything but low settings.
Android-x86 is game-changing tho. if your phone is old and shitty like mine.
I couldn't have possibly run the PC version of Honkai on that thing.

>>2292
>Does anyone actually believe GPU prices are dropping down anytime soon?
There was talk that GPU prices could drop when Etherium did their change and no longer required top end GPU to mine coins (or something like that). But I assume the Manufacturers/Shops have gotten used to getting paid top dollars so sadly it might take a good while before it drops.

>Specs?
My GPU is an GTX 1060 3GB so it is on the low end to be able to run it at all.

>>2293
>But I assume the Manufacturers/Shops have gotten used to getting paid top dollars
I guess so. I remember some Anons warning about this before.
>GTX 1060 3GB
>Low-end
Welp, I guess I shouldn't even bother.

>>2294
>Welp, I guess I shouldn't even bother.
I think minimum recommended is 4GB card and 3GB is lowest that they had been able to run it on. Still the AI is able to run on consumer grade GPU compared to others that need 40GB+ cards to run so it is much better than what it could have been. But been a while since I read the FAQ and could be they have managed to lower the spec needed (looks like 2GB vram is lowest now) People have been able to run it on CPU only but significantly slower but it is possibilities.

>Nvidia guide: https://rentry.org/voldy
>CPU guide: https://rentry.org/cputard
>AMD guide: https://rentry.org/sdamd
(think this is still the relevant guides)

Can my hp laptop run Crysis Stable Diffusion?
Asking for a poorfag fren that might want to generate mares.

>>2296
I think if you run the CPU version it could work. Not sure though as I haven't tried the CPU one, but it might be worth checking out. If it has an Nvidia GPU with 2GB or more vram it should run according to the guide without problem.

Isn't there a command you can add to the .bat file to make it generate images slower while using less memory at once?

>>2298
Yes, the "--lowvram --opt-split-attention" in the COMMANDLINE_ARGS is the command line args for lowram systems. The "--ckpt " is to select what model it uses (the pony model in this example)
>set COMMANDLINE_ARGS=--ckpt pony_sfw_80k_safe_and_suggestive_500rating_plus-pruned.ckpt --lowvram --opt-split-attention

Wait do you mark a word or phrase with (((these))) or !!!these!!! to increase or decrease the importance the AI gives to those words in the request?

>>2300
Yes you have a few weighting and other tricks to do in prompt. Cant remember them all or find the guide that had it all listed but I guess it is in here somewhere https://rentry.org/sdgoldmine
(((increased importance))) - the more paratheses the more importance it should be given

https://boards.4chan.org/pol/thread/406797368
Quality thread.

>>2307
>pic 3
The AI reinterpreted two hoes into one waifu, but one of the hoes is fat enough to be two women.

>>2302
Love the AI renderings. Looking forward to getting "colorized" videos in AI rendering. So much good too look forward to.

Has anyone by the way run any of the CWC comics through any of the AI's yet?

>>2317
nice

How are you AI generating porn?

>>2320
There is an NSFW model but can't find link to it . I assume that is the one being used
...Found the link
>https://desuarchive.org/mlp/thread/39110334/#39114433
>>NSFW Pony Model Weights.
>>https://drive.google.com/drive/folders/14JyQE36wYABH-0TSV_HBEsBJ3r8ZITrS?usp=sharing

why

Just updated my Stable Diffusion install, I see new buttons and features but the limit on prompts is still there.
How do I get rid of that text limit?

>>2326
>>2325
If you want to get rid of text being generated in image add "text" to negative prompt. At least it usually works to get rid of mangled signatures and such in images.

>>2327
There's a limit on the number of prompts I can give the AI at once. It's around 75 words. How do I break this limit on input text?

>>2329
Not sure, but have you tested if the limit is on number of keywords individual words or if it is on prompt sections.
Like:
>word1, word2, word3, ...
>full sentence 1, full sentence 2, ....
...
>pony, flower, lake
>pony standing in field by a lake surrounded by flowers, stars, moon, (unicorn), by Peter Elson
(I have not run these examples so I have no idea what they will produce)

Only limit I can see is on full input, but I think this is a limit in Stable Diffusion that the webui can't circumvent. But could be that there is a limit on the number of "sections" i.e. keywords, keysentences.

>request man in blue jacket with white flames
Wow, whoever made this AI model loves men.

>>2330
Also I updated my Stable Diffusion installation after wiping my PC. Now the text limit gets slightly higher whenever I go over the limit. Seems to be missing key words now and then, unless I'm using the wrong terms for some things like pony_ears.

Holy mother of God.

>>2334
Does putting Text in the negative prompt get rid of that watermark?
What about mixing the model with other models?

Art Subreddit Bans Guy Because His Work “Looks Like It was AI-Generated”
https://nichegamer.com/art-subreddit-bans-artist-style-ai/

Who are your favourite artists to use for art generation?

>type "horse vagina" over and over
>get this
why

>>2342
I thought combining AI models would make it better at producing horse pussy, not worse. What the hell is this?

>>2342
>>2343
Looks like you need different generating terms or a differently learned model.

>>2344
When I wrote
>blonde
no blonde hair
when I wrote (((((blonde))))) to emphasize the tag above all others I got that weird yellow pic.

Twilight Sparkle

Rainbow Dash
I'm bored now. The art is coming out crappy. I don't think I'll do the rest of the mane six.

>>2347
Are you using the Waifu Model or Pony Model?
>>2321 (Pony model NSFW)
>>2167 (Pony Model SFW)

>>2348
I don't remember if I combined those with the other models or not but I'm sure Everything V3 and an anime waifu model are in there.
The mixed model can't make good porn.

>>2349
Just wondered as the rendering looked a bit like the ones I got when I tried to generate MLP using Waifu model.
>pic 1 Waifu Model (v1.2)
>pic 2 Pony Model

>twilight_sparkle, mlp, seductive
>Steps: 20, Sampler: Euler a, CFG scale: 7, Seed: 3967284458, Size: 512x512
>Not the best seed I guess

>>2350
And just to have it added too here is the same prompt and seed in NSFW Pony Model

>>2351
Oh my, are they holding hooves?! How lewd!

So I tried generating something with the (((((arcane style))))) tag and got this.
Turns out if you overrate Arcane's importance you get weird satanic looking shit.

Nigger Sam Hyde.

>>2447
This is interesting. I wonder what it means for the industry as a whole, particularly for artist who only partly use AI tools in their art process.

>>2481
I believe that from the very moment they use it, their human intellectual property goes in flames.

F

>>3114
B-but, is that Rainbow Dash, or is she not?

>Jonathan Blow on AI art and tech
https://www.youtube.com/watch?v=DP7kguY2Z6U
In a nutshell: "Mediocre artists will loose their job, and the worthy will remain".