#🏙│ai-images
1 messages · Page 21 of 1
murzilka
cristiano ronaldo art
Sorry, but there was an error, please try again
Possible causes of the error:
- You have not paid mikus-coins
- You have been a member for less than 1 week
- This channel is not for creating art
If you cannot create art, please contact the server administration
love how it memorised mocking spongebob. I didn't even use an image input.
thats cute i like the lizard to
Thanks
this was the first message that popped up, it's at 9:11 AM for me
this was several years ago
They look very nice
When you send your message or when I sended my message
How to stop laughing?
Ayo? @distant solstice level 9 !!! 
thx!
Ping the GPT4oMod in #🤖│bots with the prompt of the image you want
Niji journey attempt generating images that look like real life
looks like ring doorbell footage
that's not necessarily a bad thing
he just mad
furry
"you're a boykisser arent you"
How i can make an ia picture?
Left: generation from 2022, Right: Same prompt, 2024
mrmodz ahh arts
They like Ump45, I like Ump9
Genocider syo!! (I think I need to add more prompts it's not that accurate)
lel
go to general chat and chat there until your username becomes green
Thanks mate
Ayo? @stuck swallow level 1 !!! 
Oh it's prob working now
or there you go
Just wanted to share one of my favourite pieces.
First draft created from Leonardo with style reference. Manual editing then upscaler with SD (octans plus a refiner which I forgot the name)
Nice
If that isn't called beautiful and don't know what is
Good to hear that!!! Most of my art is similar style to this. Anime kinda theme.
I mainly post them on my IG. But will visit and share them here often too
Imagine animating that!
Ayo? @steel haven level 11 !!! 
rave tonight in the church brothers
Woahh amazing 💕
No U
in bottom right broseph broestar is a certified gooning goober with devious intentions
aaaam My face really isn't a very good idea xdd
😏
🍿
you sure do like popcorn
I just like reacting to drama in my server tbh. Popcorn is pretty lit though
Fair, I do like a good batch of movie theater popcorn every once in a while
.
Crowded Moscow, 1980s
Flux.1 Pro
Suburban Paris, 1990s
Flux.1 Pro
Peace
Flux.1 Dev
Flux Dev:
Make her fat then you have a average American
Illustrations of an elf, Flux.1 Pro
Stratosphere
Purr-gramming
New York's skyline
Girl
Skyscraper selfie
Flux.1 Dev portrait with nails
New Year's Eve thing, Pro
And ofc freedom fighting, trust
fal.ai, replicate for flux.1 Pro, u can use hf spaces for flux.1 Dev
😮
so its better for irl stuff than for anime
I have Claude AI write scripts I run in blender to make models that I warp, render, and edit further in gimp
Controlnet 
looks like someone slapped a hole in the pc
Controlnet goated
fart fetishists: 😈
@brittle ocean

owo
How come I cannae post images
May the sacrifices begin
/imagine
this shit is NAWT midjourney dawg
what's the best huggingspace for ai art
Basement
what is your prompt
bro they left
I'm disappointed
lol
how did you get the words to be so clear
@shadow bane hey can you stop spamming yt videos and just post to #1159290752195633273?
cute pfp 
new lora
Zombie
Is the text legit sd or Photoshop
It's flux through comfy, so in essence, SD only
Flux seems to be doing a good job
Has anyone started working on animatediff/controlnets for flux?
hi
Are the ai photos getting more boring or is that just me
lol
Ayo? @quick vessel level 1 !!! 
That's what I was looking for
a blue cat in plane background
Is the plane flying towards Amerika
NO
No 11/9

Ayo? @stoic heron level 12 !!! 
@shadow bane again -> #1159290752195633273
A cat on a plane
This chat does not allow commands to create pictures, try writing this in #🤖│bots something like "@shadow bane draw bla bla bla"
Ayo? @robust flame level 3 !!! 
Dog on plane
/imagine
Hmm
Ayo? @torn ruin level 1 !!! 
@torn ruin why did you send me friend request?
we should not be joking about that, should we..
Ayo? @slate nacelle level 1 !!! 
Pretah good
Respite
Do you guys know of any free video upscalers?
I had a chat with someone a while ago that was also looking for one. Thing is no one is making any anymore, or they're super out of date.
@gritty haven Здесь это не работает попробуй использовать #🤖│bots или Закреплённые сообщение
A person who is create a image of Lamborghini
@trim lake It doesn't work here, try using #🤖│bots or Pinned message
@jolly cargo -> #1159290752195633273
Ayo? @solemn raven level 1 !!! 
/ai
Ayo? @mossy wasp level 1 !!! 
Draw sexy Latina egirl
a couple of things i made with flux pro
Dang what program yall use?
Stable Diffusion
Mode: Flux
my specialty is stickers >:3
yay
@uncut slate Interesting? 😮
please, promote yourself here https://discord.com/channels/1159260121998827560/1159290752195633273
he's depressed now
Face Card don’t decline
@steel haven can you stop with the sentences there making 0 sense
hey @bronze portal u can use #1159290752195633273 pwease
sure thx for letting me know
Yes, I also had a Website Somewhere so you could Use Flux Faster
Ayo? @graceful lark level 50 !!! 
Wait
@bronze portal try using this: https://fastflux.ai/
Also on the civitai website you can use Flux and Share Photos, but for this you need Local Currency "Buzz"
- Perhaps from these Some sites (Online / Website) Will be Flux
#🏙│ai-images message
you can also right click "open image in new tab" right click again "copy image" instead of screenshotting the page
makes it look better
nah The main thing is just to show that it works
right?
I guess
Who needs quality? The main thing is that the picture is there xdd
do you guys had a web recomendation to upscale the image?
for sure thanks
Ayo? @bronze portal level 1 !!! 
Generate cat
sorry but our generations are taking longer than usual today, please be patient
cat
Me gustaría una ia que cundo ponga mis dibujos me le ponga movimiento
true
Generate Iann Dior (rapper) 1950s style, old image, black and white
Generate Monster Sora from Kingdom Hearts 3
(I changed the text a little)
this is AI?
yeah yume is a king of making emotes in sd
yes yes :3
Yes it's true
????
i dont work with flux
you can train flux though
just not entirely for free
Here's how to train your face into the new AI models. Here's the link for $10 free at Replicate: https://replicate.com/invites/572ae3c6-1385-41b3-afdc-1e99e11836b8 (NOT Sponsored)
Discover More From Me:
🛠️ Explore thousands of AI Tools: https://futuretools.io/
📰 Weekly Newsletter: https://www.futuretools.io/newsletter
🎙️ The Next Wave Podcast: ...
this shows how you can make a flux lora model for free (limited amount times)
There's free lora training on civitai. Costs 1000 credits I think, and you get 25 credits for free each day (more if you also posts pictures)
Did they just increase it or is it for flux training?
pony -500-600 buzz
flux -2000-3000buzz
@languid bobcat
sd .15 idk
Starting with 500 credits? Is this a joke?
@magic radish
as you can see I don't have much buzz but I can get it using my LoRA/img or Just this
huh???
Maybe you want me to mention you?
no why did you do that?
I will return them to you and if necessary I will try to send you more
I don't need it.
If I open temp mail and send karadi to myself
I found a bug
It's working
@graceful lark I'll send you 100 coins with the new tamp mail I just opened, see if you'll come.
credit disappeared lol
I was looking at my original account lol
@magic radish can I give you back the buzz?
Did you come?
yeah an no
By the way, there are 100 ships that I made by opening an account from temgmail.
Credit
Haha
well, I can make unlimited credits with temp gmail
gives you 100 kradi when you first open it
It's working
the only problem is that it's 100 credits.
If someone in charge of civitai is here, I didn't do it on purpose, it was a test.
and that's your deficit.
Kéž bych ti ukázal svůj účet; já ti posílám peníze, ty mě.
Překlad byl proveden umělou inteligencí.
welp I don't think I need them bc You want to create Flux LoRA Right?
I have already spent more than 10,000 buzz to create 3 LoRA Flux but 2 of them are familiar
I don't think I can do it from there.
I'd rather use colab or something.
:o
flux.1-dev acii art lora
I rembember I had an extension for comfyui that allowed you to put your final generated image into an ascii art converter, though I forgot what it was called
oh that seems cool lol
never used local things as i can't do that
You could select which ascii characters you wanted to use in the output as well
really? what hardware do you have?
hell yeah 🔥
actually integrated graphics aren't all too bad, theres some projects like fastsdcpu and openvino sd webui that actually are accelerated with intels open vino that's in core processors from 6th to 14th gen
really? i never heard of those 2 projects tbh
I've done a test of a i7 10th gen with fastsdcpu and it can make a 512 x 512 sd 1.5 image in 30 seconds or less
i deffo cant run flux.1 dev tho
that seems good
unfortunately it prolly will take more with my i3
from what i googled seems like fastsdcpu is the most updated one and seems to have flux (ig schnell only) too
lol i was just about to say that
lmao
yeah the webui one is outdated
https://github.com/rupeshs/fastsdcpu/tree/main?tab=readme-ov-file#android damn on android too
yep, and there's also a colab version of it as well that actually uses the gpu: https://colab.research.google.com/drive/1SuAqskB-_gjWLYNRFENAkIXZ1aoyINqL?usp=sharing
oh thats good, even if i thought the meaning of it was to run on cpu, or idk if it has any other benefits of running that instead of comfyui or automatic111 on cloud
i remember those 2 were banned on colab and idk if theres a new encrypted colab, but i remember seeing comfyui on kaggle
I think right now it's mainly a proof of concept kinda like onnxstream (which is a version of sd that runs in less than 297 mb ram)
yeah any webui that utilizes opening a port and accessing another page is banned for free tiers only
yea webuis are banned on kaggle & colab but usually people just encrypt so it gets less caught
like for rvc colabs
yeah the best colab for sd is: https://colab.research.google.com/github/R3gm/SD_diffusers_interactive/blob/main/Stable_diffusion_interactive_notebook.ipynb
No flux, but it has decent performance with SD 1.5 and SDXL
and get's updated quite a bit
i remember (if im not wrong) that it was kinda buggy
yeah it is a little
btw do yk any kaggle or colab notebook for training a flux lora?
maybe its possible with T4x2
considering the 30gb of vram
damn more than 30gb? 😭
I’m just talking free tier, because one t4 is 16gb of vram and you need at least 24
I’ve also haven’t needed to use colab at all since i just run webuis locally
So my knowledge on them is iffy
oh cool my iphone stable diffusion app supports importing flux models now
are you on mac?
could u share that notebook please? maybe it could run on kaggle
no, iphone
u can use sd & flux on iphone?
yeah even sd3 and stable cascade
not that you’d want to
damn didnt think that u could do that locally on iphone
especially flux.1-dev
imagine training lora on phone
idk much about apple as i use windows pc and android phones
yeah, even the iphone x from 2017 can run it slow but fine
yeah it supports that, i haven’t tried it yet though
rip to phone then
It does get very hot i will say that

Maybe my adhd made me AI addicted again
like the splatoon ones
Guys can somebody share a workflow for this - https://www.tiktok.com/@vblank_digital/video/7397809937730030881?q=ai ocean dance&t=1724683976821
I'm not sure but you can try runway or lumalab I think
thx dude
It looks very high quality!
How long is negative prompt to make hand like this 😭
Or whitch embedding
no embedding.
im using a new model
that im really loving
just made another
Flox
very uwu
imagine being a furry
said the gay furry uwu
kawaii usually means cute in internet culture
depending on the context it may have a different definitions for locals that do use the word itself but yes for the most part it does indeed mean cute in Japanese
nope, the word itself can be used with different nuance in a sentence. At least thats what another japanese dude was telling me.
yeah again, its context based. he didnt really go further into it
probably im not sure. He just told me it doesnt always mean cute even when used on its own
meanwhile my negative prompt
how u make that it pretty cool
The creature of the water
Ayo? @near perch level 3 !!! 
bro what is wrong you
Lol u are stupid
@keen bramble / @brittle ocean / @quaint tiger WHAT HAPPENED??!?!?!?!?!?!
only recent should
So.. what happened that I missed?
I asked mrmodz to checkout some users bad behaviors, he decided to ban him but accidentally banned me instead xD
hey guys can you help me to find a free ai that can change text into image
damn embed fail
use propee channels
what would've been the proper channel for sharing my ai generation?
here
but i'm not trying to promote anything? i'd just post the mp3 file but i'm not allowed to attach files
Ayo? @shadow bane level 1 !!! 
ah nvm here we go
@tawny peak flux dev work on 12 gb gpu but the time 
Then use #🧬│ai-chat, this channel is for stable diffusion / DALL-E or any derivatives, if you'd like to request a channel to be made for generative ai music make a post in #1159516963014451302
hope that helps
🪦 rip
It'll just be a matter of time until they figure out how to lower vram reqs and how to speed it up, just like how SDXL 0.9 was meh and 1.0 is amazing
yeah soon youll need server gpu to render something in normal time
Sd 1.5 has gotten so low in terms of requirements it can a 3 x 4 grid of 512 x 512 images can be rendered in realtime on a 2060 6gb
Using Stream diffusion
Which uses TensorRT
hmm
And I guess I should also mention OnnxStream which can generate 1.5 and XL images in 512 mb of ram on a raspberry pi zero 2 w
Sure it can take up to 1.5 hours for a single image, but it works...
The Crew LoRA, will train a better described version later.
@tawny peak community made faster flux model
You can used it as Schnell , DEV, NF4 (tested on forge) https://civitai.com/images/24458443 thanks for @Dunc4n1dah0 You can used it as Schnell or D...

How much batch size can you run at once? Is it limited by vram or RAM? Is there a code to iterate through the batch on lowvram mode?
idk read the description and probably by vram
noooo man pweeeeeeeease
Makima but fox?
Ayo? @shadow bane level 1 !!! 
Btw, check my Instagram and deviantart if you haven't. Been making a lot of stuff there.
Sample
no
for some reasons she reminds me of makima
remove the fox features, have a more redhair and you get makima
No, I physically can't
aaaah
But I've gotten to the point where I have novelai locally basically
huuuuuuuuuh
:P
lol
Like... how does this work?
I want to sleep just understand this + stupidity it’s me you’ll get used to it I hope
SDXL model, tweaks, and lotsa prompting
Hmmmmmm
Interesting
In any case, until I go to bed
I don't want to think too much
See you later!
goodnight
same, but it's fun to create things
I wonder if someone has made Nagito Komaeda already
this is one of my favorites, enjoyed doing the lighting and other effects, as the RAW one is way different
https://www.deviantart.com/lunelovehearn/art/the-Sunlight-song-1050575999
gigacat
Oh no please tell me it’s not you Twitter account something like “AI Brothers” Which makes videos with AI
You mean 𝕏??
oh btw I have a dumb Question!
Yeah ?
Floating point
Just if it's lower then it offers lower precision but can be more efficient
So worse but faster
mhm otay ty ❤️
elsisi
@brittle ocean lol I haven't updated onetrainer in a while:
@tawny peak do you have colab flux? 😮
So that I can use LoRA
no
I can't figure out how to AI generate a horse without legs
I don't think its possible
even a single leg
HOW DO I TEXT TO AI IMAGE (╯°□°)╯︵ ┻━┻
guys
how do i type in text
where
do i type in text
to geenrate ai
chat gpt
it starts like this
helloooo chat gpt could you please generate me an image of a hamburger plssssssssss
thnk bro
Ayo? @arctic meteor level 1 !!! 
applio
(watermarked since its for my personal use :3 is an OC)
aint no way thats a.i
it is.
what did you use for it
combination of novelai and ForgeSD
oh okay thx 🙂
Ayo? @fresh lynx level 2 !!! 
How do I A.I Album cover guys
Dude that's rlly good
ima try making my oc if possible and ai can do that shit right
but thats sick
@quaint tiger this okay? for my second try at generating it, seems kinda mid but close enough to the image i had in my head 🥹
Wish the hair were a bit lighter in a sky blue tint as well as the muzzle and inside ear lol
what do you use for generating art?
Recently discovered Perchance two days ago
That's the website - Perchance.org 😭 I was recommended that by a friend since it had no advertisements or generating limits and such-
ohhh
ya
i use novel (paid) and SD Forge (runs on my local machine)
Is it hard to get set up?

I'm texting on my phone -.-' lemme log into my laptop
lemme be specific then 😭 gaming laptop
We're watching
lol sorry
kinda.
Ayo? @steel haven level 12 !!! 
so like where would i go again to check gpu
like we talking about gpu memory or?
system information
devmgmt.msc
task manager
oh I'm in amd software
wtf windows
i dun use 10 so idk
Ah it's def amd
you're all-but-fucked sir.
+1
BUT IT'S ACTUALLY AWESOME FOR GAMES
i could check someplace else like the omen gaming hub
I'll def give that to AMD
amd isnt good for ai
glgl
so thats a bit of a bummer
all this ai stuff is very much focused on nvidia gpus power mostly
if you ever do feel like making furry art and you can miss 25 bucks a month, i recommend novel
its incredibly versatile
mhm?
oh hey
thats a rtx 4070
laptop version is always a lil weaker, but you should be able to run ai with it
yeah it was my graduation present from my parents in May
It should be pretty simple to setup
Theres instructions on the github for Forge SD
oki ty
Scroll way down for the instructions
Thats just how github works
X3
Do be patient when you install it though
It takes a while
And sometimes it just indicates nothing
Just let it run and it should be fine
sure
should i click to download Forge with CUDA 12.1 + Pytorch 2.3.1 bcuz its recommended?
Yes, i personally go always with the recommended route
alrighty.
then from there it tells me to uncompress and update.bat then run.bat after
Welp that's fine.
Again I can always stick with Perchance 🥲
lol, just click more info, then run anyways will show up
it's just a too new application to show up in microsoft's database
idk if this goes here but well..
https://www.deviantart.com/lunelovehearn/art/A-ray-of-Hope-1094789787
YAWN
A
Valla
Acabo de crear una imagen para mi cancion super linda
La mejor imagen de todos y le agrege movimiento
@onyx relic pwease -> #1159290752195633273
(Este canal está destinado únicamente a Video / Imagen (En palabras simples, Medios), así que envíe este video aquí (Sin música, un video simple sin indicar el “Autor del video”) o use enlaces de yt en el canal #1159290752195633273 )
huuuuuuh
damn why the right arm is muscly
She's just right handed
yeah the she should use left hand to do and right to scroll
hi
oh hi khoya
azure, this is only because the program isnt in the database of windows, its a perfectly safe file.
- press more info
- press run anyway
i understand
zamn…
Nagito is so cool
Cute
what ai is this? where can i try it? it look so cool
This is Stable Diffusion, using ❤️❤️❤️❤️❤️mix v52-XL
i wanna see how to make realistic-looking ai art, what image generators would you guys recommend?
Depends on what you are trying to make. Flux is good for realistic. Sdxl / pony checkpoints have options but all feel uncanny. I don’t think it will let me post examples.
White hair female is one flux with my Lora and sdxl pony for the other with my lora. Cat is flux rest sdxl checkpoints
someone save this cat 😭🙏
hey guys
Ayo? @barren dust level 2 !!! 
Hi
looks good, idk how you do that tho.
meanwhile me, doing cute stuff
Got flux to work on an Intel Arc A750. Have elmo as a cashier
anyone knows a good free app for phone or mac where i can vectorize images?
i Use AI Arta on iphone to create images, but havent found a good vectorizer yet. Thx for the help
Ayo? @stoic fog level 2 !!! 
first try real girl
second try real girl
Stable diff
Safetensor : https://civitai.com/models/25694/epicrealism
card : Radeon RX 6750XT
Other metadata
Guidance:6.5
Steps:20
Sampler:DPM++ 2M Karras
Seed:2946753955
What’s your prompt positive and negative? Looks you might need to work on that to correct some of the issues. Like the long torso.
https://civitai.green/models/650185/flux-black-magic-xl
Shameless self promo
It's a flux lora to make anime not hideous on it.
Hello, I'm new here. Can I post AI Art here?
yea you just need to be lvl 1 which you can get just from typing
how can i generate a picture
Check pins
Is this a flux generation or sdxl?
Is there a Google Colab where it supports “safetensors”? I want to generate ai images.
TRY 3
In these I used an XL and in another stable diffusion
I'm missing the upscale in the one of the girl with red hair
^
Hey Guys!
What AI u use for create photos?
check pins
Ayo? @exotic ivy level 1 !!! 
hi how to send images
Sorry, for replying late, but I used Google's latest Imagen 3 model from the ImageFX labs. You can find it here: https://aitestkitchen.withgoogle.com/tools/image-fx. But, in order to access it, you need to fill out a form and wait approx. 5 days, more or less, then, you will receive an email that you got access.
It sure though can create accurate depictions of copyright characters, with no distortions whatsoever.🤣
Thanks for the info. Have you tried flux?
Hi everyone
checkpoint : ||https://civitai.com/models/101055/sd-xl ||
lycoris : ||https://civitai.com/models/353018/egirl-discord-profile-picture ||
res : 512x768 but upsacled
😐
I have tried, using https://piclumen.com/app/image-generator/create, but I don't like it
Btw, do you think this is from actual Minecraft, or AI-generated?
AI, the trees in the distance give it away
da lora is nice
(there are no hands for a reason)
then just use a hand fix embed
the lora is bad it got trained to make bad hand i think : /
looks like ai
Alexia Gavira from the BBEU!
actually kiiinda happy with my protag Harvey Seagate, went through a couple different ais and prototypes to get him accurate to what I wanted, aaaand, yeah. I like it. Kinda what i wanted, not photo realistic, but with some degree of realism. Overall it’ll do. I think I’m gonna use it. The eyes are a little eh, but pretty good for ai.
Anybody knows what's the song title?
Download a LORA and fix its eyes
I KNOW ITS VERY LONG BUT WHICH IS BEST I TRAINED IT FOR 7300 STEPS
Ayo? @safe summit level 11 !!! 
good idea, thanks
Much faster than fixing it manually like i was gonna try
try this one
https://civitai.com/models/119399?modelVersionId=855772
not made my me
good for realistic, 3d, cartoon and anime
LoRA of Perfect Eyes (XL) Welcome to my corner! I'm creating Dreambooths, LyCORIS, and LORAs. (guide: https://civitai.com/articles/7/dreambooth-lyc...
Flux random test, comfyUI
yeah no problem, with 100 steps you were just adding more pointless latents to an already finished image
Marge Simpson drowning in a pile of potatoes in the kitchen
it's pretty good but also very demanding
i'm barely able to run it on my system
i got it running with 8gb vram on a laptop 🤔 you just gotta use the right model and parameters
i feel like i've returned to my laptop era, cus each generation takes 40-120 seconds
my old laptop with rtx 3050 4gb used to be this slow with sdxl models
how fast is the generation on your laptop?
hmm
but i got a few 30 second ones
what gpu do you have
hmmmmmmmmm
my 4070 s definitely should be faster, considering that i have a desktop gpu
how did you optimize it? do u use comfyui?
i use forge
Aight thanks, I'll try it if I won't die
i got gens in around 60 seconds which is really good. gonna try comfy in a bit
Ooh, nice
mondongo
How do I make art
Homer caught you in 4K:
pika, luma labs, runway, kling
No entiendo
Así se llama la IA que puede darle vida a tu foto
Made this with Microsofts copilot early release
Idk how i managed to pull it off
But I did inform microsofts insider tab abt how it can generate racism
Still though was funny
Wao como lo hiciste
Ayo? @onyx relic level 2 !!! 
It's not worth using it anyway considering how many resources you need
To run it
SD3 takes less time and resources the quality is only a little bit bad IN SOME IMAGES + sd3 2b is being used while we use flux 12b
same problem as sd3 the light poles are deformed see there is not that huge of a difference
hello is does anyone here know how you do seg anything with svd someone a few months ago before luma sent a gif where individual leaves and hairs moved without any morphing they said they did it with lcm and seg anything
hewo
mhm idk
😮
I want to live that dream. I love snow but my city it doesn't snow.
I love them please could you make in 9:16? To get wallpapers in that aesthetic 🤩
just made a new pony model
it's really good if you like anime, super promise.
bonus if you know how to actually use pony
Bella Ramsey 01 - IA Art 3072 x 3840
Does she look like Bella Ramsey or not? How much do you give her on a scale from 1 to 10?
10 sir
Como lograste que todas quedarán igual
Yo quiero que mi imágenes sean iguales y diferentes estilos
Entiendo, gracias, pero yo no se tampoco hablar english 😅😅
Billie Eilish 01 - IA Art 1792 x 3840
How much do you look like Billie on a scale of 1 to 10?
Utilizo el mismo mensaje + creo el estilo LoRA y para que otros entiendan cómo cambia el estilo, trato de usar siempre el mismo mensaje
@maiden hearth -> #🤖│bots
https://suno.com/song/96d0393c-6a9c-4893-b623-3d6c78f25c7a
https://suno.com/song/991a2d94-b60d-4153-a581-fda1274f7993
phonk dark trap aggressive hip hop horrorcore heavy bass distorted beats creepy atmosphere haunting sounds eerie soundsc song. Listen and make your own with Suno.
lo-fi hip hop jazzhop chillhop relax study music calming music atmospheric soundscapes vinyl crackle rain sounds song. Listen and make your own with Suno.
@graceful lark Draw any image
Ayo? @maiden hearth level 1 !!! 
what are you talking about? What do you want me to do?
I don't do anything that is +18
Ayo? @vapid basin level 1 !!! 
Bella Ramsey 03 - AI Art 1792 x 3840
This is the best AI photo I have so far of Bella Ramsey, how much do you give her on a scale from 1 to 10?
Here I leave you the process I use to make realistic photos.
And I also leave you a test of the new AI that is used to edit faces and give them the expression you want.
@graceful lark have you heard of illustrious? It's a beta anime sdxl checkpoint with crazy range
Uuum noope
Sowwy 😅

NO NO IT'S OKAY it's just really cool and I wanted to let you know about it
Like it's in beta but I think it could be better than pony
Tbh lately I don't do Anime pictures very often, so I don't use these models very often
ahhhh
looks cool, i'll try it out when it's updated a bit more
the next update is on the 30th
I mean i dont even know how pdxl got so popular, it's one of the worst trained models i've ever seen, not saying the quality is bad at all, but just the way it's trained and you having to put the score_9, score_8, etc stuff at the beginning is so stupid and just makes prompts unreasonably long and break with slightly too large prompts
oh yeah definitely
i meant pony has a lot of range out the gate
and illustrious has CRAZY range
Does illustrious break lora compatibility like pdxl?
nope!
i made a scout lora and it's pretty dang good... it just has some quality issues but its to be expected for a beta model
guessing it’s using sdxl base training then
good range in styles too
yeah it's an sdxl model
Cool, actually may try it later since i have tons of sdxl trained loras
lora was trained on 38 images 
actually i haven't tried it with other loras made on sdxl 🤔
For some reason some of my best loras were made with only one image
so i actually can't say
oh dang lmao
hands also look decent for a beta model
it messed up on a bunch of the other ones but yeah 😭 you just gotta prompt it right
i’ll test to see if the sdxl hand embeddings work on it
but it has artist names baked into it so you can prompt artists
ill train something on it
but later
it look plastic rn
peak
Ayo? @scenic skiff level 3 !!! 
hello guys

