#🏙│ai-images
1 messages · Page 12 of 1
I am very new to this, I normally use VoiceAI stuff so yeh but the ImageAI really intrigued me
Yeah XL means they are sdxl models, to get good results, check out the model page and check to see the recommended sampling options and other settings to set
SDXL does require more GPU horsepower that SD 1.5, so whats your GPU?
My specs right now look like this:
NVidia KFA2 GeForce RTX 4070
AMD Ryzen 7 7700X 8-Core
Corsair Vengeance 32GB (2x16) RAM 6000 MHz
ASUS TUF Gaming X670E-Plus
Samsung 970 EVO Plus 2TB SSD (+More SSDs)
Oh perfect, have you set launch arguments yet in your webui-user.bat file?
I think I have only installed Automatic111, downloaded two models to start with, and now I have unlocked Clip Skip.
And I ran the update thing
I followed the steps thoroughly for installtion tho, but the steps for setting it up after installation were kinda missing so that's how I ended up getting help from people in #✦│chat and now in #🏙│ai-images
I actually have a guide that helps you set up A1111 so if you'd like to check that out, you can. This definitely does take some setup: https://civitai.com/articles/2982
I made it specifically for this server and for the people who keep asking where and how can I generate images
It also helps out on how to prompt things properly for realistic and anime/cartoon models
I'm assuming I gotta be at this part right now, correct?
I think by now you should be at 4. but refer back to 3 for prompting help
Thank you so much for helping by the way
No problem! I'm always happy to help.
Oh, and one more thing. SDXL Models require a special model to make images generate properly called a VAE. It helps with making colors vibrant and fixes generating: https://huggingface.co/stabilityai/sdxl-vae/tree/main
It will go in your VAE folder
/sdxl-vae
Help with what? You just linked to the forum and not a post
Do I take the NVidia one or the Ryzen one for my setup?
Nvidia
I'm assuming Imma take the file on the bottom?
yep
Double check to make sure I did it correctly
Oh, what file did you open? webui.bat or webui-user.bat
whoops meant to ping the other guy
Vae placed
it should look like this in webui-user.bat
yep good
I think I accidentally got the wrong one
I got the user one
the one on the right
the one on the left
you just don't have file name extensions turned on in windows which is fine
yep looks good, close the webui cmd window if it is open currently and relaunch
and you shouldn't need to relaunch again unless you add extensions or turn off your computer
Hold on open it in notepad I just want to see what it links to as I don't have that file, if it links to webui-user.bat then go ahead
or else just run webui-user.bat
yep, that's fine to run
I am now here
And to enable your VAE go to settings, VAE, and select it from the dropdown list
and xformers got installed so everything is ready to go
go to the same settings as you set the clip skip
and add here vae thing
Now in section 3 of the guide, there is some prompting help and a recommended negative prompt to put in as a test, but basically, Positive is things you want to see in your image, and Negative is what you don't want to see.
Aight I'm exciteeeeddd
sd_vae
you don't have to set that vae setting if you don't want to, but it's up to user preference as VAE can be found in settings unlike clip skip
its useful
True, i guess i've just gotten used to going in settings to change it
if he wanted to change to regular sd
Swag

is this the model you downloaded: https://civitai.com/models/119012/bluepencil-xl ?
Merge everything. Recommend Negative Embeddings: unaestheticXL use stable-diffusion-webui v1.6.0 or newer See HuggingFace for a list of the models....
and are using
Merge everything. Recommend Negative Embeddings: unaestheticXL use stable-diffusion-webui v1.6.0 or newer See HuggingFace for a list of the models....
Model Introduction (英文部分) I Contents In this introduction, you'll learn about: Model information (see Section II ); Instructions for use (see Secti...
the recommended sampling params are DPM++ 3M SDE Karras, 25 steps, and for XL models make sure your resolution is at least 640 x 960 with Hires fix turned on with the ESRGAN 4x Anime model
for blue_pencil at least
and you should be set to generate
Do these settings automatically save into the model or should I note this somewhere so I can put it in everytime I use the model again?
note it
under the generate button, theres an arrow which will grab all the generation parameters including prompt, neg prompt, seed, sampling parameters from the last image generated
even after a relaunch
very helpful
oh didt know about it
but if you do want multiple characters and their prompts you want to keep, you may want to put it into a text doc
lots of people don't, it is a very small button
Do I also note these settings or should these be good to go now?
I'm planning to try out various models so I noted it just in case then
in my testing of XL models a denoising strength of 0.4 and hires steps of 15 should be good, forgot about those
That should be the last thing, and now you can generate as many images as you want
LET'S GOOOOOOOOOOOOOOOOOO
Achievement Unlocked: Infinite Waifus 

And if you upload constantly to here, you can earn the SD creator role, which is higher than the Model maker role and even the model master role
i never thought of that ty for the idea
Np
👀
And right now you are just in the shallow waters of SD, theres still so much more like Video generation, Custom Models like character, styles, etc.
After conquering the powers of VoiceAI and now in my Goku training arc for ImageAI, maybe picking up new stuff in the future too, ayo??
And yeah shallow waters fr fr, I only started this like two hours ago I think
Definitely combine both and make your own anime 👀
I'd say my favorite thing to do with SD is colorize manga panels, it definitely gives new life into the images
I already use the VoiceAI to make videos as anime characters in VRChat and uploading said vids, so that doesn't even sound like a bad idea tbh
Do you use the img2img for that and just use prompts that indicate a colorized version of said panel that you used as image, or?
Trying to get an understanding of it here
No it's an extension called ControlNet, it can be used to control the pose of the character, grab a style from an image and apply it to your image, and some other things

👀
Nice, just make sure to not send crazy lewd images so you don't get banned
YEAH, one came with an exposed nipple even though I didnt use any nudity-related terms in my prompts aside from "black lingerie" and I already made sure to skip that one in my copy+paste, to not post it here on accident, lmfao
SPEAKING OF WHICH, if I want to cover them more, I'm assuming I can describe outfits in large detail right?
Yep, it's because locally ran SD wont filter any images and the models are trained on NFSW images and SFW images
Yeah things like hoodies, shirts, jackets, work
So for example if I add into the mix: "blue formal suit, red tie, white shirt, shirt and suit tucked into formal pants, confident posture" for example, and then put "black suit" in negative prompt, that should work yes? And in similar way; you can detail the entire outfit specifically how you want it to be even in smaller details right?
yeah things like that work, if you want to emphasize a parameter in your prompt more than other parameters, you can highlight it and press ctrl and the up arrow or down arrow and set it to some thing like 1.4 and it'll show something like (blue formal suit:1.4)
wow just noticed the size of my SD folder
Daaamn
I have been doing this since 11/15/22
so it kinda makes sense
and I also do train models
R.I.P.
Dont worry, testing phase is over and am now trying to actually work on things and these upcoming ladies will be mostly fully covered lmfao
Smash
You wanna see a failed attempt on that one that could have become so much better if it didn't mess up? XD
The arm of the one in the suit going into the clothes of the other is what messed this one up, if her arm was positioned normal then this one could have been perfection, lmfao
I guess you should learn something new, inpainting. Go to your img2img tab and then the inpainting tab, load the generation params from that image using PNG Info, increase the CFG Scale to 0.75. Then draw over the area you want to fix and it'll try to fix it
Ayo? @tawny peak level 12 !!! 
I like their silly toothpaste hair lol
The hairstyle looks kinda like Alear from Fire Emblem Engage
oh and make sure to change the resolution to match the image, and the sampling settings
This was the least messed up version I could make of it for now. XD
Will maybe give it more attempts after bit of sleep, this AI stuff got me so focussed I forgot it's already past5AM at this point, lmfao
It did fix the arm thing but the suit seems a bit more out of shape now, may need to look into this more tomorrow. At least it still looks a lot better than the original one, lol
Thank you so much again for all the help and assistance you have given me today and hope to make u proud of your student one day, sensei! LMAO. But for now; goodniiiighttt!! 😄 @tawny peak
Lmao, goodnight
Can somebody help with SD Deforum? I do everything correctly ig but still there is something wrong with result
have 500k won!?
#howtofightmanhwa
eating one on my pfp
trick is to dive face first into the watermelon
it also just comes out good as a shine when taking a picture
its yandere(no blood no gore) just yandere face expression, NSFW?
sample
Ayo? @quaint nexus level 1 !!! 
Ayo? @wicked scroll level 13 !!! 
Iceybois
How many reincarnations am I gonna make of this boi? YES
Bro unlocked superpowers now somehow, lmfao
Smash

why is his nipples see through
Un cane

She is now a marketable plushie
H
H
Sexo
2 and 3 are fine don't post the rest tho 👍
dream massage? 
base SDXL?
Nope, Bing Image Creator
damn, looks cool
Ayo? @summer wind level 1 !!! 
Ty

whats the command for stable difffn
what
like what command to put for generating that img
which image
this
oh
(masterpiece, best quality, high resolution), white hands emerging from a singularity, sleeping face, many hands, vortex, intricate detail, 8k, hdr
just to type this? im asking like the input process
im just a NEWBIE lol
/ and whatever
command
bruh thats too hard
i cant make it out
just pls tell me in personal
what the command is
but i mae it with this
that ofc just a sentence
we don't generate images on this discord, we just talk about it/instruct or post it
ohhhhhhh
i thought wecan generate it here
lol
sry for wasting ur time
i'd seen that smwhere
mb
i was confused af😭
for that you need to use sources like google AI, Bing AI, Tensor.art etc.
dalle is not that bad
this aint no Khazar women
Women dont exist in AI Hub
Do you guys know if AnimationDiffusion works with AMD currently?
What prompt do i need to make one image like this style?
Most likely you'll need a lora
On bing image you can't
We cant help you much, you'll just have to ask the prompter what his settings where
Okay if bing then what his prompts were
The most important being the "by:artist" tags
Is there an artist of this style?
Or you can try googling fantasy art, looking for which artists draw in that style yep
Any that you know
oushie
alr
thankyou for the help
Np
the Mario ones
?
It was a miss click
I know this channel is mostly Stable Diffusion oriented, but 1 out of 4 times I post something that isn’t generated with SD, people think it’s a Lora or severely good prompting
Anyone know? (Just bumping in case someone else sees it now)
I mean there is directml A1111, there is ROCM which works best with Linux, there are possibilities, but I myself have 0 knowledge about it, besides some loose readings of conversations in other discords...
some people use M1/M2 Macs with AnimationDiffusion and ComfyUI, so I guess technically it should be possible

blond sex toys
buddy whaaaaaaaaat
wat
buddy whaaaaaat
boi wadahel

Upper body Crayon/colored pencil scratchy style doodle of a black colored fuzzy Gengar from Pokemon with orange eyes, wearing a dark purple jacket, orange wool shirt, black jeans, orange/purple striped Santa hat, smirking, blushing, dynamic pose, purple/orange vaporwave background, overcast, unconventional, (image only)
Yo that’s sick
How can i ask for the ai generate a picture that looks like a screenshot gameplay, like you did there?
try something like, "text, heads up display" or something like that
I usually do something like "a video game screenshot from a N64 game"
I heard the other one was just “Mario Odyssey style”
@odd lance ogm!!!!!!!!
THIS TWO ARE AMAZING!!!!!!
i'm also sending these 2 to loli art channel
No
only like explicit stuff is nsfw
what kind of work do you have? xD
Oh this is sick asf lemme pin this
buddy what makes you think that 😭
i mean... like.... it has a bit tad much skin showing so....
Hi i need a logo for a clothing brand called 7 Ate 9
Creame una imagen
yeah, I can't even post my negative embedding from civitai here because half the page is nsfw images
i could use some more stars 🥰
maybe in a country that has Sharia law, but seriously mate, the rest of the world sure won't call that image NSFW O.o
wait what someone has sent?
just some digital art gothy/emoy girl with black half top, something like this: belly was a little more visible
lol why it got deleted
imagine if we could make loras for bing ai generator
it's never gonna happen, since it ain't open source, but just imagine
@rustic raft
wait that's not ai, right?
AND I NEED THIS PLUSH
the bunny ogmogmgm
FR
i am unable to look at this image and NOT think about the 2010's
i edited the message and got deleted and forgot it would delete if edited :3
bit
lyiar 
that's wild
🥶
any news or leak regarding NovelAI v3? <@&1159579982683701379>
i dont use novelai so
rip
After the v2 leak u can be sure that there won't be v3 leak
rip
so this what it is? idk https://huggingface.co/spaces/nyanko7/NovelAI-V3
asking bc big figure asking me
and trying to gather as much things as possible
i love gradio
LMFAO
I didn't use it, but it's possible for membership i think
The model weight itself isn't available so rip ig 
this used a novelai API but the account has since been restricted
so its brokey
rip
how many chance it will be leaked?
obviously speaking in therms of possibility, i dont want it, i just want to get updated on the news
don't use novelai as well and why are you asking for leaks 💀
cuz the models aren't half bad
It's like getting the current 1.5 SD level of advancements but with a x10 better model
Everyone wants another leak
But they probably won't do the same mistake again
And got better security
good for them
Your account has been restricted due to unusual activity.
Free generations are unavailable.```
yeah that didn't work out for long 
lmfao
yeah I know what you mean, though that kind of style is still pretty prevalent with ~20y girls around here
ye
my waifu wholeheartedly agrees, if you find it, let her know 
Can anyone please tell me how to make ai pics?
You can’t ping a channel
Thank u ✨
How was it possible to get my huggingface account banned?
I have never done anything bad with it
[From NovelAI]
Your account has been restricted due to unusual activity.
Free generations are unavailable.
how tf 😭
that's from novelAI, nothing to do with HF
is this nsfw? 😆
in what world?
no idea 
Yes too much shoulders 😡😡😡
yo guys i got this new action figure
banned
spoooooooky 
also pretty spooky 
tsundere-san
Nice hands

Smoky field
we when like... uh... me when i like... uhhh.... like me when... uh.. uh uh..... big brain... uh.. forehead.... uhh... big brain forehead... big head... me like.. me like... uhh... yeah
morbius
you good?
cheeseburger
why did you show me this
Cheese
Ayo? @next magnet level 1 !!! 
you listing your favorite foods? 
1 more

idk i just using lora
Recommended Weight: 0.6 - 1 No Trigger Word Suggested prompt: 1 girl, closed mouth, tears, wavy mouth, tearing up, pout, :i, meme, :t, parody, frow...
stream starto
they aint even in ai hub
Discord didn't refresh that on my side, I could see the AINoob role and everything still, now after a restart it's just a dead id number 

I got some cute results, so I'm not complaining 

cat saying woof 
@rustic raft I need your cuteness overload reaction here 
KHIDFSGKJUHFSDRKJHLSFDKHUKFSEIJOH
CATTTTTTTTTTTTTTTTTTTTTT

what checkpoint and lora do you use btw
@shell star tbh I'm not super good at this 
Besides that fluke with 4 I can't ever get it to make a background
u are fr
i flooded it with bg prompts and flooded negative prompts with stuff like "white background" "blank background"
I think it's an issue with the input image at this point
Same prompts with text to img
then why the hell it gave u white backgrounds lol
(i didnt bother to inpaint those 3)
@shadow bane still remember this masterpiece
I use looots of different checkpoints, one of my gotos is AlbedoXL though, since it's very versatile
I think Aniverse and Animerge are pretty versatile Anime models for 1.5
Lora I mainly just use a add-detail Lora at a low 0.5 strength
depending on what I want to see, I will then add Loras accordingly
at first i thought it's cause the image had a white background so I made it a transparent bg png
khazar made it
but it didnt change anything
v1, v2, v3
that stop sign either turns into a leg or doesnt appear at all
Aniverse over Anitoon then?
yeah I was chastized for it not really being greek enough 
try what works for your prompts, that's really the biggest factor, you might have a better experience the way you write your prompts or what "scenes" you're trying to accomplish.
it's hard to recommend stuff, beyond "well it makes beautiful images with almost any prompt"(but might not follow your prompt that well), or "it can follow prompts super well and is very versatile", etc.
@shadow bane for example do you have a lora or checkpoint you feel would work well with this?
I need to sort some stuff for squishy, then I will try to see what I can find 😉
🙏
@ruby lagoon what tool are you using atm for AI images?
ok but which UI from that, they have different choices no? 
this one
ah so A1111, ok that makes it pretty simple since that one is the most used one on the internet atm, though my knowledge of it is very rudimentary 
@tawny peak @brittle ocean you use A1111 right? do you have tips for Lineart/etc. img2img?
Yeah, I use the ControlNet Lineart Model to colorize images, if that's what you're looking for
just controlnet
@ruby lagoon ^
No matter how many times I try inpainting the magic braids it refuses to die
try turning up your CFG scale
I like the dithering effect that bing ai does
whenever I see things like that I struggle to find what one would call that style 
CRT?
just dithering?! 
dithering or video degragation is also something that works
dithering or halftones
cat
looks veeeeery good already, of course the side shot is still a little wonky, cause the AI has no idea what to do 
playing around with MileHighStyler a little
cat
this gave me an idea
oh gosh 😭
||please dm it||
id rather show it

epic
Question: I'm trying to use ai to make a reference sheet (just the character) for a commission later.
How do I do it while making sure the AI artifacts don't show up and make the person mad I used AI for the reference?
Is this possible?
use controlnet for reference pose then just normal prompt
I need to get better images of peter for the dataset😭
How do you do it?
a1111 and lora trained by me
As someone who draws I don't think this is possible unless the artist is just oblivious
Reference sheets are made to be closely examined
At first glance it looks good, but if you look closely there's artifacts everywhere in the clothes and faces that they'll pick up on
I think you should just say "i used ai just for the reference sheet, but I mean im still commissioning you right now arent I"
Or tell them you got that image off google and had no idea it was ai generated LOL
I was just wondering and testing the latest AI stuff.
Yeah ai art is incredible but as of rn there will always be artifacts
Unless like a human goes through the whole thing and manually do touch ups
it is only a matter of time
Yeah thats why I said as of rn
Ai?
Ayo? @outer rose level 1 !!! 
ye
Ayo? @brittle ocean level 32 !!! 
is that seriously AI generated?
Yeah
that's crazy 😭
It's more obvious when you look at the smaller details, and the hand in the bottom left corner
yeah, even patterns are hard for it
when do you think AI will be able to make hands correctly?
Ayo? @kindred viper level 60 !!! 
It can already, it's just very uncommon
sick
or the big detail: watermarks 
Ayo? @languid bobcat level 11 !!! 
Not too bad:
/help
No Category:
help Shows this message
Type /help command for more info on a command.
You can also type /help category for more info on a category.
Hey! We have built a custom robot here. Now we would like to use AI to create images of our robot in different environments and situations. What would be the best tools to train a image generator to use our robot?
https://ibb.co/q11K12w
1st 3 images out of 10 (at <t:1702889400:t> the 2nd 3 images are posted)
How to make image?
/goku
tried stable diffusion for the first time owo
gyatt 
if you need help with AI image generation, you can check out #🔍│help-ai-art
there is also a list of commands in the pinned comments there
in #🔍│help-ai-art you can see pinned commands that give you links to platforms where you can generate images
😳
tick tock heavy like a brinks truck
“Yes ladies, I know I have a 6 pack…”
yo wtf
F R U I T B O W L
tf? ayo
hi

How does one describe how this looks like? sorta like a miniature? Mind sharing what in your prompt achieved that? Or was it just random?
how do I add LyCoris to automatic1111?
Put it in your models/LyCORIS folder and load it the same way you do a lora
alright, and do I have to download anything for it?
cause idk what to download from here
oh alright, thanks for the help!
No problem
wait in lycoris folder?
i put this lycoris models in lora folder
Hey can someone explain what stable diffusion is? How to use it? It's types etc. and from where to get the model as well as how I can train my own models ?
Stable diffusion is an AI image generator. It’s leagues different from VoiceAI/RVC and you can run it from a webui or use an online service like bing ai or civitai. I made a guide here you can check out if you’re interested: https://civitai.com/articles/2982
Totally Out of Touch...
what'd it say
Ayo? @gleaming rock level 6 !!! 
https://youtube.com/shorts/0cY3QkTh_7A?si=nKBM_JOCOb9fbgc0
Any open source tools for this?
🚀 We Are Hiring! APPLY HERE: https://techwiser.com/we-are-hiring
🌟 Please leave a LIKE ❤️ and SUBSCRIBE For More Videos Like This! 🌟
🌟 Follow Us Other Social Media Platforms 🌟
🐦 Twitter: https://twitter.com/TechWiser
👩👧👦 Facebook: https://www.facebook.com/tec...
I mean sure, but I don't know about an open source app that does all that for you, there are tools out there that you can use to accomplish it though, like SVD, animatediff and co.
but this level of quality is pretty hard to do
If you're not that knowledgeable about these things, you probably won't have a open source tool for this for a while.
So far they have always cought up, so just give it some time.
Ayo? @shadow bane level 28 !!! 
is that the famous elite english I hear so much about? 
yeahh
"A anime scene of a train passing by. The train has motion blur. Bokeh effect. ((It's raining)). The ambient is watery and dark."
Cyberrealistic model with sdxl example LoRA.
I used some negative arguments, but they were the common negative arguments such as: "Bad quality, worst quality,... "
enma ┬─┬ノ( º _ ºノ)
What
I drive
Ayo? @vivid mango level 6 !!! 
this is only a place to post images, you can’t generate any here. if you want to find programs to generate images, go to #🔍│help-ai-art
please read the message above
the logo of gta vi is made with ai?
#🏙│ai-images ur mom
ur mom

#🏙│ai-images Степа баран купил себе процессор Интел и любит его
Извините, здесь нельзя создавать изображения. У меня есть руководство, но оно только на английском языке, но я думаю, что с помощью перевода вы сможете его в некоторой степени понять.
how can make text with ai image?
Ayo? @nimble steppe level 1 !!! 
Yea... im totally done giving out presents after christmas
Dall-E 3 is pretty okay with text, you can get it with bing AI
Teahkook
Putin 4k
yo @shadow bane is this true?
might be true for their Lora, it depends on the checkpoint/Lora etc. prompting and weighting is very dependend on those things, how they were tagged and trained
most creators will tell you what they know could help you with their models
i mean for me it looked like input for dataset and it worked the best on 0.85 so i put 0.85 in the description
so did that supposed wonder-negative help you with 1.0?
Realistic image of a woman, sitting on her bed, wearing pink headphones, with ahegao face, warm light
u cant generate images here
Oh perdón
chin dimple 
@kindred viper 🤨
Leave me aloe 😭😭😭
Not me 🤭

Wtf
David Finch! 

Yeah ong
@uncut slatewhat does dat say
without squinting your eyes
answer correctly or !/ban@uncut slate 🤭

/demote @kindred viper
no it wasnt a test I was giving a hint to lista
oh no 😭
It reads "smack my ass like a drum"

https://civitai.com/images/4750494 this is what this guy rendered on 1.0 eyes are fucked up
destroyed another image in audacity

what do you mean by audacity? that’s audio editing software
You can datawarp audacity using the raw document function
huh, neat
I think you can datawarp in most programs
surprised the file is even readable after the fact
Again what are you using to make it look like that
Audacity?
Applied bandpass and echo
To a tiff file
hello to all
Hi
how are you
Fine, you?
fine
Ayo? @acoustic helm level 1 !!! 
@tawny peak hello
hi? again
you was boy
uh, yeah
ok
#imagine a cute little baby orange
#image a
@autumn moat this goes INSANELY hard
nggghhggmghgghghhhhhhh only if I could screenshot it.....
download the image 
Start
Wat
#🏙│ai-images print lego man
the postal dude lora https://civitai.com/models/238019/postal-dude?modelVersionId=268379
Main character from the Postal game series. No fixed trigger word, prompting for example images was in vein of "3D CGI man, sunglasses, goatee, hol...
Idc if you download it, no credit needed
how to use pls help
Ayo? @rocky canyon level 1 !!! 
thank you
make me a car with a ball
Ayo? @woeful creek level 1 !!! 
lols
cute😍
Synthwave Taylor lit af
You can’t generate images like that here
ok
I have a very sneaking suspicion abt u
whisperIthinkthey'reaferretwhisper
nuh nuh, my gaydar is picking up smthing


What
What do you suspect about me
mmmm
Tell me
i think ur gay
Hell no
no offense lmao just so u know

magical indeed
nft
lebron jame
lebron jame
lebron jame
fr
eh
Original message was deleted
ahaha
#🏙│ai-images make an image of levi hugging hange
Is there something better than Pika(dot)art in animating still images?
make an image of Spider-Man
it's one of the most easy online tools, I don't know about any others, the rest I know is local/collab like Animatediff or StableDiffusionVideo etc.
Merry Christmas :)
this one's meh but i like the sky
"ripped gengar doesn't exist, he can't hurt you"
ripped gengar:
kinda mid but eh

Elaina art
bread
Ayo? @shadow bane level 1 !!! 
breaaddd
Ayo? @alpine junco level 6 !!! 
How to do?
How
@shadow bane here are some links to online tools that you can use for image generation:
#🔍│help-ai-art message
Ayo? @wet scarab level 1 !!! 
also posting this in loli art channel
original prompt generated some mid results imo
Sheesh
this will be @autumn moat in 2015 
Ayo? @kindred viper level 64 !!! 
Poopmaster has a timemachine?! 
Ayo? @shadow bane level 29 !!! 
hai Khazar
you didnt know?

i knew you would send that
RAAHH
bownling ball
/generate among us r34 3 boys 1 girl feet and skibidi out of toilet ultra realistic 5 fingers
what AI did you use for this, its real nice
AI? SD with Lora's
meh
how? 😭
we don't have generations on this server, you can use online tools if you want:
#🔍│help-ai-art message
you need level 1 to send images in #🏙│ai-images
Oh
Ayo? @wraith thorn level 2 !!! 
Bro’s got the kicks
what is the AI you use ?
Bing Image Creator (Dall-E 3)
thanks
it's free ?
Yes
Just make a Microsoft account and go to https://bing.com/create
Yes
why i write michael jackson american singer I have content warning ?
u cant generate celebrities
on bing
ok thanks
Explain him in great detail
trying to make more art that looks like my character forres
but i only have 1 art piece of him ahahah
so its hard >:3
Yeah I've also made a single image lora as well, it's not great but surprisingly versatile with prompting and doesn't keep a style or anything.
it works here and there
Ayo? @quaint tiger level 37 !!! 
and having a controlnet referance is neat too
did you split it up into multiple images with different sections or just train one single image
just trained a single image :P
check out this
Part one had some issues where 2 addresses and fixes them so no need to worry about 1
yume how do you even make lora's ive always wondered on how to make one
maybe dreambooth
kohya-ss
hope that answers it XD
also






