#🏙│ai-images
1 messages · Page 27 of 1
otay? can u check #🌇│ai-videos? ^^
ah u del otay
sowwy
@rustic ingot URL!!!!
-# Not eddy just eat im sowwy
thx bro
(i have never generated an ai image locally on my machine in my life)
Hey guys illustrious 1.0 sucks so don't bother using it
Not that you can actually get it locally or anything like that
Legally anyway
Okay it's not as bad
as I thought
but it's not the like... Thing we were looking for.
Merges/finetunes are better, still
I gotta ask what model, these are great.
Sowwy? What does that mean?
sorry
Okay I was just curious
does someone has the mfcg pdxl model, it got deletd from civitai and i cant find a place to get it
This is one of the craziest AI generated images I've ever done.
https://media.discordapp.net/attachments/1159290421676085258/1297145264527904768/pixai-1804324057623631354-Anya_and_Goku.png?ex=67adca2a&is=67ac78aa&hm=8df3331150cc45124252e80ca34c34a867189d1cfdc65014541ccbf6331e3837&
Hello friend! How are you doing on this fine morning?
@steel haven AI Vidéo
Make picture about Putin
average ai hub members with a combined braincell of .5
Is not a video though
Yeah stop dragging us down with you
bro left
Or I'll make your the star of the next Sarah McLachlan commercial.
damn that's crazy bro
No it's basic math
damn that's crazy
Hey i might be late but Animagine XL 4.0 is out will soon test it.
Make picture about War
@median dawn not for video channel
yeah pwease del
mm nope
k
If I don't say it, someone from the helpers or moderation will.
They'll do it anyway
Easier for who
mods / helper
“ramthrust”
What’s your favorite spongieboy episode
Wha do you think about Bluey? 😮
There is no such thing as too old for the spobbert!!
Im 17 and it’s like my favorite show!
I'm 70
GRANDPA
ye
Do you have a car?
I was gonna get one but I got a 4090 instead
Oh, I still have the Honda but still
forgot how decent my old model merge was
nahh bros in the backrooms
thats so cool
inpaint / img2img + my crooked hands
Well uh… I hope she has a nice flight!
So is she pilot flying this flight?
hi
later...
😮
mb you should send it to another channel? -> #1159290730200707153
is this hand-drawn, not AI-generated?
Yeah
It’s just a silly thing I drew in response to ramthrust’s generation that I’m not planning on finishing. I just sent it here so I could directly reply to him lol
c'mon it's not that hard to make it using ai
Your art is looking good, but this is not where you upload a "real" art. For real art, go to #1159290730200707153 and attach a response message link there.
Nah, I find it hard to make good AI generated image, especially about two or more different people in one image. I can only draw by hands better instead. 
You will have to talk a little bit until you can send an image here. Your Weights link has your account invite code attached.
there's no way thats ai
read the metadata then
im just saying thats kinda crazy
🙂 ✨ Satan wants to hire you 👺 🤝 🥸
🦊
Can’t believe Mr. Krab cakes would cause such a scene at GameStop smh
i made this with imagefx
i would show you rest of it
but most of em are mostly fat
so you want me the show fat ai image or no
what ya think?
but instead of kobeni
it's freddy fazbear
or more like
freddy fatbear
🟥 🐼
Try gen my Sona! Pweeeease
can it be done with prompting only you think?
Or does it require a lora
Axolotl in this style
https://civitai.com/models/841226?modelVersionId=1398128
mikus, yellow fur, yellow eyes, black tip tail, black ears, white hair, black sweter, gray shorts
"Tell Me What You Want..?!" from music Phonk Funk Brazill!
yea i didn't remember the artist's name
i know, relax, its just that the style was remembering me of someone
yea i remember him now
Don't ask anything, just enjoy
Otay...
yeah I like to use ai with other peoples styles to see characters that will never get drawn by them in that style. Of course out of respect of the original artist I don't post it anywhere other than discord.
Yeah it's gay time yay!
omg AI did this
LOL
Is that flux or what
nah not rlly
thats like 13% bf flexing
the fingers are kindaa weird
Btw yall that was not generated by ai, im sorry for lying i just wanted to see what yall thought about the image
no its oki! ^^
TLC TV stars (Kate Gosselin (Kate Plus 8), Amy Roloff (Little People, Big World), Michelle Duggar (19 Kids and Counting), Stacy London (What Not to Wear) and Sophie & Katherine (DC Cupcakes)) as IVE EMPATHY concept pictures
This isn't a place to post that, or even the right server.
srry i will delete it
I think she likes popcorn a little too much!
do you guys like it? and who misses these TLC shows?
Low quality real-life picture can enough to trick people to think it's AI.
this is pretty cool idk
Tyler, The Creator? 
No that's not tyler
He just looked more like that guy.
I know lol, just being sarcastic.
Also what was even the prompt for this lol
woman eating popcorn
crazy work
nope?
-# prompt: masterpiece, best quality, newest, absurdres, highres, 1girl, large breasts, shoving handfuls of popcorn into mouth, ahegao face "> <", tongue peeking out ":t", butter dripping down chin, laughing mid-bite, popcorn scattered on thighs, oversized hoodie slipping off shoulder
-# negative: child, kid, loli, blurry, watermark, smudgy, (bad quality:1.15), (worst quality:1.3), censored, (multiple views:1.2), jpeg artifacts, (patreon username, patreon logo:1.1), Halo,
-# Parameters: Steps: 35, Sampler: Euler, CFG scale: 5.8, Seed: 549225115021189, Model: ILLUSTRIOUS\Illustrious-Black-Magic-Clean27.safetensors, Model hash: 72e4f83d56, VAE: sdxl_vae.safetensors, VAE hash: 235745af8d, Hashes: {"model": "72e4f83d56", "vae": "235745af8d"}, Batch index: 2, Batch size: 5
Lol
u can check https://www.sdimage.info/
Yeah I know, there are quite a few things for doing that.
pony
-# model: il
Yeah looks nice
thank you
Those don't seem ai generated
yes it is i'm not lying
How? If so, how were they generated?
Ok let me refactor that question, are they deepfakes, face swaps, ai generated images, ai altered images?
ai generated images (i use weights)
Show the proof of them by screenshot your whole "My Creations" on Weights. 
Idk, to me because of the artifacts in the images, they seem fishy
im so sorry 
If you delete your messages again, I think you have a problem here.
i wont do it again
anyways goodnight
girl why tf would you have a problem with deleted messages be serious rn
🔥
Thought this channel disappeared lol
AI-generated by someone else, not me
lumberjack?
more like lumberbear
hehehe ^^
i can show more of the ai image i did if anyone is interested.
since i did that fat mewtwo ai image

Now generate an image of a burning shrimp. 
Yummy. 🔥
beeg snake~
What is this lol
Why is everyone making funny stuff 😭
no more melting until...
Sometimes, you could've refrain from posting fat inflation generated images and instead posting something normal like foods and people.
literally 5090 owners
Price inflation situation is crazy.
well i can't post any of my normal arts in the art section since the art section is gone
i also can't stream so i just mostly watched you guy's streams on the vc
i thought gif were images
so that sucks.
anyway this is a oreo spongecake flavor
this is a limited edition oreo flavor
only available in certain countries
sweet
ai'nt no way...
hey there, glad you are having fun here but please refrain from posting any sexually suggestive content over here
this place is intended for sfw content only and we are trying to refine and clean up whats posted here
plus the lumberjack bear is wearing clothes
ah i see where the confusion is, well inflation is a fetish so some people may consider it nsfw cuz of that
hence why im suggesting you avoid it you know
ooooooh....
bro went crazy for cartoon girls 😭
Too many to delete
i dont even know

mario kart 9 is awesome
Cute!
Hi
Picture of a Korean baby
I wanted a picture, not a gif
im pretty sure that's not how it works o- o
though why do u need a pic of a korean baby o -o
smhh
-# @keen bramble bte i love ur banner
@graceful lark please keep your messages appropriate
yeah sowwy
Want to know how things work here
Are you all AI
How can I help you?
Are you an AI
How can I help you?
I am an artificial superintelligence capable of autonomous response. How can I help you?
It's like if we took every person on this server, stripped them of all their social skills and you're what's left.

I don't talk like a big boy. I talk like a man. 
Why did you guys troll him lol
It's in my nature
Good nature
How did you do that? I want to know
i just made a random ai image generator make it
i used an image to prompt and then tweaked it a little to have michael jordan in the image
Quick call mommy
Misono Mika drinking tea. 
ooh @prime plover!this chat is for AI Content! 
For some reason, the admins deleted the channel
ai
ok sorry
hello guys am i at the right place to ask a question ?
puppy in situation
this place is not for non AI generated shitposts
I understand
imma go get an ai image rn
anime girl in banana suit
ai generated:)
To send a normal image and picture, you can just send it in #✦│chat
ok tanx
hi me
@zinc portal
Which app
???
Which app are you using.?
I use ComfyUI to make the images, with a modified artboard, and with the Flux Dev fp16 Model
Asuma Toki and Sanae Kochiya sitting in a park.
Whats?
Someone ban him and delete that, what he did is prohibited by law
@lament crystal Ban
no dragon.. bruh
@graceful lark Can you create another AI Meme Art for me? I need another one as well
???
huh
@keen bramble
how refined is that lora damn
which model did you train it on
Illustrious
fifth-mix-illustrious-absinthe-sdxl
ooh wait
@wild valley its u xdd
a
I need permission to upload my work on AI HUB nut I can, can you help me
get lvl 1
Talk a bit more in a chat. That's what I can help.
A cute kitten dressed in a bumblebee costume, standing upright with big, expressive eyes. The costume has black and yellow stripes, tiny wings on the back, and little antennae on the hood. The kitten looks adorable and slightly curious, with fluffy fur and a soft, warm lighting background. The image has a polished, cartoonish style similar to a digital illustration. Only one kitten in the image.
That's just giving the prompt words for everyone.
That won't make a bot here to generate for you.
sounds like a nonsense bot
Silly bot. ✅
Use the /image command
Here's your generated image!
"A cute kitten dressed in a bumblebee costume, standing upright with big, expressive eyes. The costume has black and yellow stripes, tiny wings on the back, and little antennae on the hood. The kitten looks adorable and slightly curious, with fluffy fur and a soft, warm lighting background. The image has a polished, cartoonish style similar to a digital illustration. Only one kitten in the image."
but in #🤖│bots 
yeah?
Oh no, now people are going to be spamming image commands in this channel
@keen bramble
Down boy!
friend wanted me to generate Holo playing the switch, turned out very cute
This is not where you provoke a bot to generate an image for you.
You asked like if this is the typical Discord AI server or something. Use Weights bot to generate an image in #🤖│bots and forward the message to here.
Another thing, you'll have to talk a bit in #✦│chat or #🧬│ai-chat to get image permission to send an image here.
you throw some coins to a pond and nothing will happen
how did you do that 😭
thanks!
question, what UI do you use to generate your images?
Colab
ok, thank you!
Please don't drop generations here that are rather suggestive. The buldge is showing too much
Hitori dressing as Arona.
how to make sdxl fast it takes 1 min per img
nvm i fixed it
sefyu french rappeur
/imagine prompt:cover for the two sides of book with 22/22 centimetres, pyramid turned upside down with crimson hue, landscape of crimson desert, 4k
cover for the two sides of book with 22/22 centimetres, pyramid turned upside down with crimson hue, landscape of crimson desert, 4k
use /image in #🤖│bots

A random dude in white shirt, Son Goku and Anya-chan. Generated on Pixai.art.
身着一袭优雅长裙,轻盈的布料随着步伐微微摇曳,仿佛将星光穿在了身上。精致的剪裁勾勒出她曼妙的身姿,细腻的蕾丝点缀在领口与袖间,透出几分柔美与神秘。妆容淡雅却不失光彩,一抹红唇点亮了整个面庞,眉眼间流转着自信与从容。发丝如瀑,轻轻垂落,几缕微卷的发梢在耳畔轻舞,增添了几分灵动与妩媚。每一步都仿佛带着风,举手投足间尽显优雅与气质,令人不禁屏息凝神,沉醉于她的美丽之中
She was dressed in an elegant long dress, and the light fabric swayed slightly with her steps, as if she was wearing the starlight on her body. The delicate cut outlines her graceful figure, and the delicate lace is embellished on the neckline and sleeves, revealing a bit of femininity and mystery. The makeup is elegant but not dull, a touch of red lips lights up the whole face, and there is confidence and calmness flowing between the eyebrows and eyes. The hair is like a waterfall, falling gently, and a few slightly curly ends of the hair dance lightly in the ears, adding a bit of agility and charm. Every step seems to carry the wind, and the gestures are full of elegance and temperament, which makes people take their breath and indulge in her beauty.
抱歉,这里不适合写你的故事
sorry this is not a right place to write your story
AI image made with Flux and some loras, but don't ask how I achieved that quality, I just moved a random parameter and it fell into perfection
@keen bramble
lemme microwave myself after that
what is the most up to date to train a lora locally with 12gb vram
you want hasky image? 😮
If u want to sure
otay! 
i need details on that first one... how
ayo
@graceful lark what parameters and how many images did you use to train this lora
Ooh AI chan
I accidentally added a Flip picture so this model is not very good
Honestly, I don't remember
I'll write to you about it in a day, otay?
is it sd3.5?
Any ai that can make a existing ai generated iage become more realistic?
Ni.
Anyone have for fragpunk?
I don't have that game. If you mean by a game, let's go to #✦│chat. But if you mean by a style image model, try search in CivitAI.
what did u use??
My personal homebrew. :)
@graceful lark trained a new lora of my OC
well.. honestly im not sure
im not pro 
dataset (59 img)
@keen bramble can u send more img? 😮
dm?
doing some tests to reach the limits
used 67 images
😮
doing some wild stuff like this is good to test how much it distorts
can i test too 😮
sure
Generated with ComfyUI
My best one so far
||<@&1159293140440723499>||
This image has been automatically detected as NSFW content and removed.
||
Hentai: 82.43%
Drawing: 16.61%
Porn: 0.84%
Neutral: 0.10%
Sexy: 0.02%
||
@harsh moss
rockyandthepawpatrol (1284317225884450878)
None

How is it 82% hentai but only .84% porn
Crazy.
porn is irl or just not a drawing
Ah
cuz yiff or anything furry art related is more hentai than porn
the image was full-body boykisser-ish
it's not using SD or anything to detect
no its fine
It was SFW
not necessarily nsfw but I see why it flagged it
there was nothing bad in it
More likely the word was included one in commect section of the image, so maybe that's why.
yeah cuz full body
^
better safe than sorry :p
full body + furry - clothing = probably nsfw
literally
I don't think this will get flagged~
I used DPM++ 3M SDE GPU as the Sampler and Karras as the Scheduler with 40 steps and 6 CFG
🙀 , what even is that?
Yes
wha model (checkpoint) u use? 😮
Specifically not Boykisser, but it’s called Femboy Fur on CivitAI website
But I modified the prompt to make it look like boykisser
ooh wait i mean
like
uuugh
Stable Diffusion 1.5 (name?)
SDXL? (Pony, IL, Noobai)
PonyV6 XL
ooh 😮
I recommend trying iL.
I Created LoRA for Il Boykisser!
-# also have for pony v6 but work so badly
I will try it in the future
how to create or generate images here
A very relaxed monkey with his favorite music.
What song is the monkey listening to that makes it so relaxed?
Generated in CivitAI
Use Weights bot to generate an image in #🤖│bots
This channel #🏙│ai-images is not where you provoke a bot to do for you.
generated by Gwenwithgreeneyes as part of Maintenance Mode contest, August 2024.
Lil Wayne dressed as Arona from Blue Archive.
octane
can I create a manga ?
from AI only
@vocal vigil
i know python, CSS and AI development
also video editing and 3d vfx etc
should I apply for a job or wait for a degree rn ?
i'm 17


This is not where you give your own idea about a story as words bruh. For talking, it would be better to go to #✦│chat.
This channel #🏙│ai-images here is where you share an AI generated image, AI-generated manga illustration for example, not a word essay generated by ChatGPT. 
y asking some career advice in this fking channel? 
@keen bramble Do you mind if I show you who I caught today? 
She bit off my wires!
im just tired..
Guess the anime characters in ai style
I'm back with another image, bruh
😮
||<@&1159293140440723499>||
This image has been automatically detected as NSFW content and removed.
||
Hentai: 79.65%
Drawing: 17.86%
Porn: 1.94%
Neutral: 0.43%
Sexy: 0.12%
||
@harsh moss
rockyandthepawpatrol (1284317225884450878)
Hmm....
AGAIN???!!!
WTF??? It's just a thing sitting on a bench drinking boba tea!
It's safe dw
I think, it's cause is blushed
😳
Maybe xd
@sterile pelican / @pure dagger try this site! ->
https://pawgge.rs/
-# welp or just try find other github 🤷♂️ 😅
I didn't make that bot lol
yes, I'm sorry for the inconvenience. When I get the time to finish training the proper model I'll do that. Until then, please be patient and reach out in DMs or via ping if it gets completely out of hand.
pika + boykisser? @~@
You're telling me to use Pika with my image?
no no i mean look like pika + boykisser
Ah, no, I just used ComfyUI, and tweaked the prompt. If you want, you can give me a prompt to generate an image for you?
@keen bramble i want cookies pweeease
oow ty! ❤️
but rn i use colab!
I tried using your Boykisser model, but I did not get good results
ooh im sowwy
you use pony?
bc il work well!
Yes
This is your model LoRA I used:
It's okay, you're still trying
ooh wait you use score_9, score_8?
Uh, no, I thought those were not needed, so I removed them
noo!
these tags make the picture more detailed! and beautiful
just use this tags
Ok
So, should I use only one, or multiple?
Wait, should I use score_9, or score_9_up?
This is using the "score" tags
Using a Boykisser LoRA model, (not yours)
genius
ik! its yume lora!
||<@&1159293140440723499>||
This image has been automatically detected as NSFW content and removed.
||
Hentai: 95.81%
Drawing: 3.97%
Porn: 0.21%
Neutral: 0.01%
Sexy: 0.00%
||
@pallid thicket
two_bit_bone_berry (1106954024562991104)
None
: (
What
Peak fr.
Actual waste of space, I swear
Let me try somethin
||<@&1159293140440723499>||
This image has been automatically detected as NSFW content and removed.
||
Hentai: 93.95%
Drawing: 5.85%
Porn: 0.18%
Neutral: 0.01%
Sexy: 0.00%
||
@pallid thicket
two_bit_bone_berry (1106954024562991104)
Now lets see if this works
Nope, my hypothesis was wrong
How do I turn these notifs off bruh

Score tags primarily provide different styles and quality. What you choose is up to your preference
sol futuristico
This is with two-pass upscaling, using RealESRGANx4
Bad mod
@~@
Asuma Toki and Moo Deng the hippo. 

@keen bramble
I have returned... XD
AI-chan and a big fish. 
How to make ai pics
I have to get image perms again... >_>
I retrained my furry model again... this time 59 images, 190 epochs, took forever but it looks even better now.
1920 x 1088 generations too
why 8 pixels more height?
welp
oh wait maybe I'm wrong o.o it lets me do 1920 x 1080??
i remember trying it before and getting errors
O_o
I'd rather use any resolution with the same 16:9 aspect ratio
it's generating an image now, I'll see if it doesn't error out when it finishes
2560x1440 is divisible by 32
it can also make 2560 x 1440p images but that's like the limit... i think it's a limitation of Flux
after that it starts repeating stuff
u can use tiling strategy/hi-res fix
What program did u use to train it?
it
Kohya ss
then why do whenever i train mine mine always look deepfried with kohyass
What base model are you training on?
i think its my settings
AnythingV5
last base i used
@shadow bane what settings do u use
There's a lot of settings o.o and I've only gotten amazing results from Flux
ah that might be why
i dont train flux
cuz i used to train 2 years ago and now so much has changed
With flux, i don't crop or resize the training images or anything, and I enable bucketing
ah i train 512
Stable Diffusion 1.5 model uses 512x512. XL and Flux typically use 1024x1024.
I train 1024 x 1024 but my training images are multiple sizes and all larger than that
||<@&1159293140440723499>||
This image has been automatically detected as NSFW content and removed.
||
Sexy: 90.63%
Neutral: 8.81%
Porn: 0.50%
Drawing: 0.06%
Hentai: 0.01%
||
@zinc portal
rxfaelgrafikart (1119830598983360594)
None
LMAOO
must be the neck
Sexy
MIKU OMG
pick up a pencil
pick up a pencil
pick up a pencil
pick up a pencil
pick up a pencil
AI ART ISNT REAL ART
AI ART ISNT REAL ART
AI ART ISNT REAL ART
AI ART ISNT REAL ART
AI ART ISNT REAL ART
Tell that to yourself also, but I draw art. 
bro got automodded 
Real artists would never complain about picking up the pencil. 
AI learns the same way people do... i think
except much faster
if you grow up watching cartoons and then learn to draw, then your drawings will be influenced by stuff you've seen / experienced / watched
they could be afraid of having their works stolen as training data, and instead put some ridiculuous watermarks
damn
Art is art, no matter what type
@keen bramble *bite*
Retraining my model at 1280 x 1280 now to see if it can look even better. If my GPU doesn't explode.
For what SDXL?
Flux
You shouldn't need anything higher than 1024x1024
I don't need it, but... the 1024 x 1024 results are already amazing and I want to see if it could improve even more
XD
But yeesh, 7 seconds per iteration is slow af.
I'm leaving a couple gigs of VRAM available to play games while training
Ah, i usually just let it go full blast and just do something else while it's going
i think it's only 1 or 2 seconds per iteration slower this way
With my 3060 training SDXL based models I usually also turn off my displays and it gets up to around 3 sec/it at 896 resolution
4 With both displays on
this is dreambooth with the full size 24GB flux model, and 1280x1280 so that's why it's so slow x.x
That poor GPU
@tawny peak 😭
hewwo! can i "conncet" mu loras to checkpoint on colab?
i try this one but i have problem wich pip install so any other? 😅
just numpy not work
literally full
What error does it give
Using cached numpy-1.20.3.zip (7.8 MB)
Installing build dependencies ... done
Getting requirements to build wheel ... done
Preparing metadata (pyproject.toml) ... done
Building wheels for collected packages: numpy
error: subprocess-exited-with-error
× Building wheel for numpy (pyproject.toml) did not run successfully.
│ exit code: 1
╰─> See above for output.
note: This error originates from a subprocess, and is likely not a problem with pip.
Building wheel for numpy (pyproject.toml) ... error
ERROR: Failed building wheel for numpy
Failed to build numpy
ERROR: Failed to build installable wheels for some pyproject.toml based projects (numpy)```
i try install 1.20.3
2.2.4
1.26.4
1.19.5
and other but not work
just error
I can't even seem to get pytorch installed on the collab
Try my lora
Zundamon singing at a concert. 
I prompted ChatGPT to remake a TF2 meme in a cartoony style, the result was NOT disappointing.
Ooow yeah!
Amazing
Somebody knows a link to download Topaz Video IA free (crack) ?
Zundamon + Arona 

why i can't send photos
You'll have to talk a little bit until your name turns light blue, and you'll be able to send image here.
ok i'll see
Nice
woman
woman
Blue sky
Poyo.
Yeah? 😮
you cant post that! look at the rules
Shhh, don't tell anyone 🤫
🤓
what
lvl up
😲
Oh hell nah, we definitely fell for it 😞
hey hey I need someone to make me a youtube mini with my face please contact me pv discord
nice
I was testing Loras all this week, and I managed to get the Flux + Loras Generation Model to give me a High Quality Image.
Image 432 x 960
Drake - More Rizz 
anybody here who likes create character and environment ai arts
wow 😮
/imagine prompt: Um homem e lua cheia
/imagine A young samurai standing under cherry blossoms, manga style, black and white, highly detailed, emotional expression, dynamic pose, cinematic lighting
/imagine A young male warrior, around 18 years old, athletic and strong build, short spiky white hair, bright blue eyes glowing with electricity, heroic and determined expression, wearing futuristic dark armor with silver and blue highlights, standing in a dramatic battle pose, electricity and lightning surrounding his body, a stormy battlefield background with dark clouds and flashes of light, epic manga style, highly detailed, dynamic action shot, cinematic lighting, powerful aura --style manga --v 5 --ar 2:3

xdd
you guys having fun cloud rendering with chatgpt 
Sunaookami Shiroko dressed as Momiji Inubashiri, riding a motorcycle. 

A photo I took when I went to Day out with Thomas three years ago, but prompted using Ai.
One message removed from a suspended account.
One message removed from a suspended account.
An image I prompted using a photo of my Loona plush. :3
sick! also huh we have a common friend for some reason lol
Things are starting to get really cursed out here when combining Nick Wilde and Judy Hopps from Zootopia together to form one character. One example:
This uses 50 Steps
Should I really go for 100 steps, or is it really unnecessary?
Does it do a thing?
This is with 100 steps, took around ~322.74 seconds
Here’s a cute pic I prompted using my Cream the Rabbit plush.
Oh hey popcorn guy is back!
Now this image, with my new ComfyUI workflow, now made by 3-pass samplers and each with it's own CLIP Text Encode nodes
This took ~193.29 seconds
New image. First Pass:
Second Pass:
Third and final pass:
No, no, no, my generations using my 3-pass workflow looks so cursed
Characters become too photorealistic
From this:
To this!
Another one, also went through my 3-pass refinement system in ComfyUI:
TF2 meme I prompted in ChatGPT.
That is coool!
I have plenty more!
Yayyy, please share!!
Here's two memes of Soldier. One that's just him wearing a sad face. The other is based upon a YT video called "Heavy and Soldier Meet Bane".
Here's two pics I prompted of Heavy. They're taken from Gmod shots I've actually taken myself.
And finally, a meme of the Medic with a distorted face. Based on the SEES or SOOS video by notna/ugotaphonenumber.
Ah heck nah she’s gone bananas
can anyone tell me which AI tools are best for text-to-image generation for free?
how can i send my ai images here
Ooh u need 5 lvl
bananas
ramthrust.
what a name


