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.
It's really great but if you know how the person talks then it's noticeable. Shit is one winded run-on sentence.
Supposedly you're supposed to use more punctuation.
Wouldn't know since I'm looking to see if someone who made a decent Patrick Bateman can help me out a bit since my own attempt sucked and I don't have a dynamic IP for making new accounts.
 
Last edited:
Ayo check out that new ai voicebot, the niggas are making the stanley parable guy talk about total nigger death and some guy made CWC sing that songs about killing all the gays and the faggots.

Neural Network Generated MATI:

Null explain Sneed's Feed and Seed:

Supposedly you're supposed to use more punctuation.
Wouldn't know since I'm looking to see if someone who made a decent Patrick Bateman can help me out a bit since my own attempt sucked and I don't have a dynamic IP for making new accounts.
Is there something about vocaroo that I'm not getting, or is it just the middleman for another system that does the vocal imitation?
 
Some more stuff I've done recently, apprantly anything v4 is a thing so I'm using that permanently now. Anyways please enjoy these.
anything v4/4.5 is a troll model that's basically just abyssorangemix. i'd recommend using that if you like the style and keep using anything v3 instead
1674403573086-635b369d55b3a42577b95c18.png
1673671777140-62f2b20aeb9e8a5f05cf9a9d.jpeg
1674567622764-635b369d55b3a42577b95c18.png
https://huggingface.co/andite/anything-v4.0/discussions/30#63cd5f8aaf312b5264ad0b7a
https://huggingface.co/andite/anything-v4.0/discussions/28#63cfe10ec85f64d616d6d122
https://huggingface.co/andite/anything-v4.0/discussions/4#63c23463f9453420b5def52b
 
Supposedly you're supposed to use more punctuation.
Wouldn't know since I'm looking to see if someone who made a decent Patrick Bateman can help me out a bit since my own attempt sucked and I don't have a dynamic IP for making new accounts.
From the ones I hear on /pol/ proper punctuation doesn't seem to work either. Like there's no natural sounding pause, it kinda just stops then continues on like it's talking in one large breath. Still amazing though. I guess it works better if the person you're mimicking with the robo waifu already talks like that themselves like Ben Shapiro or a German.
 

I tried making robo-Null, but it came out kind of fucked up.
In the future streamers will be able to hook up a chatbot to react to what's on stream and have it sound exactly like them, but for now the voice tech is more "ghoulish marionette" than "flawless impersonation."

flowers7.pngglades.pngbirb1.pngMorning glory.pngfaerie.png
Here's some more flowers since this is the art thread

Edit: I didn't notice until just now that @777Flux did the same thing just a few posts earlier lmfao
 
Last edited:
Neural Network Generated MATI:

Null explain Sneed's Feed and Seed:
Null is closing KiwiFarms:
For Null, it's the McChicken:
Null is a Navy Seal:
Can you get that thing to pronounce "Arpejlio" or some of the other Nullisms?
 
Anyone has a suggestion for good models? I have seen the list and tbh I am a bit lazy to test every model
Well, you've already got Anything V3 which is a pretty decent general use model (haven't used the leaked NAI model at all since getting AnythingV3). Dreamshaper is still my current favorite to mess around with, though I've also slowed down considerably in testing other models.
 
Ok, a nice tard wrangler aimed me in this direction rather than let me continue to scribble on the walls in another thread.

Been fiddling with this. It's pretty nice but the default has me running it on CPU due to no Nvidia. Getting anywhere from 11 to 45s/it and thought I was fine. Whelp.

Looking at the Isekai Too guide and... I'm guessing this isn't for Windows? Cause I don't have a dolphin command, for example. Anyone got AMD GPUs working inside of Windows 11?
 
Last edited:
I went looking for AMD info because I didn't want to give wrong advice about it being possible since the method I thought everyone was using only seems to work on Linux and not WSL.

It looks like the way most people are going on windows is a package called SHARK: https://github.com/nod-ai/SHARK/ since the previous shim package doesn't work under WSL anymore.

Here's the tutorial that lead to that, too:


I actually had 3 different older Python versions.
>python environment

jungle

Really if you're playing with any deep learning stuff at this point I feel like conda on windows is the only way to keep it sane. On linux you still get the OS python interpreters and such but on windows it's all in a nice portable package even though that package can bloat pretty badly.
 
Last edited:
svc.jpg
I made this and posted it on the Fediverse less than a week ago and in that time it's gone viral and spread and so far over the capital I Internet that it's made it's way back here and Kiwis have posted it in multiple image threads without even knowing where it came from. It's been interesting to see how it's been edited as it's been passed along: scaled up, saturation boosted, text altered, and translated to other languages. A guess then it fits the real definition of a "meme" then, eh?
 
Back