#🍥|anime
1 messages · Page 35 of 1
Working well
(1girl, solo:1.1), pink hair, high ponytail, detailed glowing yellow eyes, high-waist skirts, turtleneck, bare shoulders, jacket, forehead, light smile, head tilt, nose blush, (spoken heart:1.2),
(from front:1.2), hands in pockets,
Negative prompt: AuroraNegative, (low quality, worst quality:1.4), (bad anatomy), extra digit, fewer digits, (extra arms:1.2), bad hands, by (bad-artist:0.6), bad-image-v2-39000, loli, monochrome, greyscale, nsfw, nude,
Steps: 20, Sampler: DPM++ 2M Karras, CFG scale: 22, Seed: 1801917921, Size: 512x768, Model hash: 4d651c7638, Denoising strength: 0.53, Clip skip: 2, ENSD: 31337, Hires upscale: 2, Hires steps: 10, Hires upscaler: 4x-AnimeSharp, Dynamic thresholding enabled: True, Mimic scale: 7, Threshold percentile: 96, Mimic mode: Half Cosine Up, Mimic scale minimum: 4, CFG mode: Half Cosine Up, CFG scale minimum: 4```
thank you 
All I could think of when I saw that
Should be in the png info, but I can give it if you're unable to access it.
@cyan burrow what this mean?
I gues you need to download the config file by clicking the blue text
and then put it into your models directory
or wherever teh checkpoint is
exists?
First time I saw this
Love all the little details in this
whats that res, holy
anyone tried ABG?
fr
Believe it or not, it's not big enough lol. I need it to fit on my ultrawide quad. 3440x1440
Like this one
What are you using to upscale it? The built in upscaler or something else?
Tiled diffusion (multidiffusion), with Tiled VAE. Denoise of 0.35-ish. Mostly only quality-based tags in img2img for upscale. Works wonders.
Thank you 
And usually 4x_fatal_anime as the upscaler model.
Or sometimes ultrasharp and sometimes anime6b depending on the style.
I'm also a fan of this one 
yeah those look awesome
All Aurora model. I've been testing its limits and putting together images for a review. https://civitai.com/models/40199/aurora
keeps getting better every day, my time with dark gemini
Final ultrares version if anybody wanted it.
wow
Aurora is magic
anime model 
im truly in love
which is better Kenshi or Aurora
I just realized u don't use neg embedding or detailed background neither
You're in luck! Ask the creators of both @quick hill lol
better quality than my average generation
@magic wave ?, thanks for lora btw
semirealistic > kenshi
anime > aurora
Oh I do. auroranegative is in there.
I should update Kenshi 
damn
famous person
@quick hill hi
hi 

thanks probably aurora then
it's a bit weird to use though, very versatile and different than other models
I will say I never got Kenshi to work near as well with all things as I am Aurora. In my mind Aurora replaces most models.
only problem with aurora is that you can't schizo prompt as much as other models and expect it to work
you mean you cant be specific with it?
you have to be specific with it 
oh nice then
i usually make my ocs with it so a lot of specific tags
and i use a lot of those too
copy paste
I also have my OC I use too
Oops. I upscaled this one too big by accident lol.
I didn't know if could maintain its clarity at that high of a resize
😭😭
[Tiled VAE]: split to 3x5 = 15 tiles. Optimal tile size 1408x1312, original tile size 1536x1536
Lol
as long as its not a character I think it should be fine at very widddeee res
how to make public link?

xDDDD#
that's prob in webui-user.bat
ive never use this before lmao
better?
my poor gamer eyes
haha
perfect 👌🏻

you need same connection
Then head to "Virus & threat protection"
Under "Virus & threat protection settings" click "Manage settings"
Scroll to the bottom of the page and go to "Exclusions"
Click "Add an exclusion" and select "Folder"
Add the folder which contains 'Stable Diffusion Files'
This will assure that stable diffusion connects to the internet as long as your internet is actually stable.```
try this
maybe antiviruses might've blocked gradio
But SLE 
Can someone pls anime lineart controlnet this rq?

If I could get controlnet to not brick my SD venv then I would. Until yesterday things worked fine with controlnet and I enjoyed its benefits.
Hmm
I should have left ending steps to 1 probably
Yeah maybe try that
That's just when I forgot to enable cnet
I guess it's time for "Borrowing Smol Sus's prompts"
me who accidentally put 6.0 instead of 0.6 weight on lora
Modern art
my prompt is almost always the same, no good idea at the moment
Same

so good 
And there goes the hair color
still good tho
Wdym?, mine just bsod so lowkey afraid this happened
You're killin my heart
mmmm weird, i havent encountered that but i havent booted back into sd since bsod, ill keep an eye on it
been trying to learn controlnet today
Yeah, it was a fantastic extension when it worked for me.
This was controlnet 1.1 or 1.0?
Im using 1.1 so idk if that matters or not
I just woke up, you want thicc lips LoRA? 
Hmm....
It was the latest from the extensions tab.
'Twas partly how I made this
Yeah ig, not necessarily THICC but like strongly visible since most anime models dont do that well
And better support for all kinda of colored lips
yeah idk which one that is i manually installed it in the extensions tab with https://github.com/lllyasviel/ControlNet-v1-1-nightly.git
youll need the 1.1 models
https://civitai.com/models/38784/controlnet-11-models
try this if youre using the older one
Hmm. I'll have to try that then. I'll do it later tonight though, as I'd rather not redownload the 5 GB venv atm if it goes wrong, lol. Thanks for the link!
lol understandable, hope this helps
Oh wow. It was updated 7 hours ago.
It'll already be different from the one I tried to install.
yeah i saw someone mention it here today so i decided mightr as well learn the new one instead the old and new one since this my first time messing with it today
really neat, need to mess with it more but after getting bsod critical error i think its break time
Aurora can support a lot of colored lips I think, and there is also a lipstick LoRA 
But if I make a LoRA, I'll be thicc bimbo lips 

So they will be THICC™️
and hot pink
based


Some failed diffusions
More prompts
I guess Clyde will approve this one
Is this an oc of urs? Or just a prompt
just prompting but.... it's my waifu
i claimed it

I see i see
The heck are these doing here when generating? I thought these were supposed to be taken care of with VAEs? They're still slightly showing up at the end of the gen.
is that the end of the generation?
What even are those
No, the start
That's the end.
See above the left eyebrow? The mark is still there... among the others.
probably just denoising but not enough?
maybe more steps?
It's 50 steps on 2M Karras Test--it should be enough. This didn't start happening (or at least I haven't noticed it) until the last hour. 
hmmm, that's weird, no idea
For those who want this beautiful prompt stolen from Ashen
mksks style, masterpiece, best quality, detailed background, cozy atmosphere, warm colors, soft lighting, BREAK 1girl, solo, :), looking at viewer, casual clothing, oversized sweater, leggings, white socks, messy bun, brown hair, blue eyes, holding a mug of hot cocoa, BREAK indoors, living room, plush couch, knit blanket, throw pillows, bookshelf, fireplace with crackling fire, framed photos, houseplants, large window with a view of a peaceful snowy landscape, string lights,(vase:1.5)
the lips 

I was trying to protect her forehead 

I am! It is quite good imo
@gloomy scroll what model are you using for those portraits, if you don't mind me asking?
Aurora
tyty
what do you mean? 
I have like 30 checkpoint models... trying to fine-tune what I like, then going to remove a bunch. I wish A1111 had some sort of filtering for checkpoints.
https://aigc.ioclab.com/sd-showcase/light_controlnet.html we could be seeing a lighting ControlNet soon. Would be convenient
if you want, you can put it into sub-folder
and it will appear in auto1111 as well
oh - tyvm @lyric zealot

oof thats cute
yeah that is probably one of my favorites ever haha
would make for a perfect vinyl sticker
Well I'm just not having much luck with my SD lately. Wanted to try SAG, but it doesn't apply properly.
error: modules/sd_samplers_kdiffusion.py: patch does not apply```
Reee i keep ooming[out of memory>.>] with controlnet lineart
I'm going to guess wrong. 2, 3.
what it means to "train" your model on waifu
Beautiful. And I'd like to think that's accurate.
I love that the starboard is 95% Anime :3

prepare to execute patience. It's BUGGY
does anyone know what to use use for animal crossing stuff like this? this was done with AI but idk their workflow
Buggy how? 
Well, it fails to startup sometimes for me, and I have to kill it and try again.
It will get duplicate entries in its dropdown boxes for upscaler selection and stuff.
Some of the default settings are pretty weird and I'll need to find a way to adjust. Not a bug, just weird.
Passing stuff between screens has crashed it for me. I like to pass prompts or images between txt2img and img2img and pnginfo, and do it often. But in this it is not reliable.
Hmm, seems it may use more vram, too. Can't generate as high of a res with this.
wow.. by a lot
lemme know what you think
It may take a sec. I'm going through the settings one by one. 😅
ah
change jpg to png is one I remember
add in your desired samplers
but yeah I couldn't even generate 1536x1536 in vlad's
so back to auto for me

As in upscale? Or do you use 2.x?
just a hires fix
Oh hey Nubby is here. 
which is just img2img 😛
Shopping for prompts
This error occurs when your sd_samplers_kdiffusion.py has been modified, such as changing/adding the modified DPM++ 2M Karras sampler or when you've already applied the SAG patch.
Ahhh. So I'd have to change it back, apply SAG, then add the changes.
oh god that girl is stuck in her walls
Dang, wish I could get the sharpness from higher denoise, but without such a big change 😐
Yep. Not positive if SAG has conflicts with the DPM++ 2M Karras v2 sampler. I haven't added the v2 sampler back in since adding SAG.
SAG can help a lot with sharpness.
lol its "art" dang it
does anyone have good controlnet tut/guide i can study
dis be complicated asf
someone's gotta warn the woman painting before it's too late 
i keep geetting weird stuff like this when trying to fill in the canny stencil
looks cool at least
mistakes and/or unintentional results are where peak art comes from
yeah i agree it looks really neat
honestly looks cool but I hear ya
even at a low denoise img2img/hires?
i was able to get the canny from into but thats where i hit a ceiling in what i want
How low of denoising are you talking?
I wouldn't go lower than 0.15, otherwise you're better off just having a straight upscaling without SD.
Then almost certainly not. The issue there is simply that you are taking a relatively blurry upscaled image and not allowing it to be changed enough to sharpen it. For Hires fix, you really should be no lower than around 0.4, and probably closer to 0.5 denoising.
up to about 0.2, I get a similar result. beyond that I start getting changes I dont like.
For fun....
Yeah, Hires fix is supposed to have pretty noticeable changes. The whole point is to set a basic composition of the image with the increased coherency that low res generations have, then upscale the image and pass it through image to image to add way more details than would have been possible at the low resolution.
Nubby do you remember the quicksetting for VAE? I wish there was a dropdown menu for this.
I think it's just SD_VAE. I would check but I don't have SD loaded right now. If that doesn't work, just change the vae in settings and hit apply, it will show the variable name at the top of the settings menu and then you can add that to the quicksettings list.
I'm an idiot. I thought it would be too obscure to search but people are talking about it. I guess SD AI usage is even more popular than I realized. Heh. It is indeed sd_vae. Thanks!
yeah and that's my typical workflow, but this one was just too good. 😄
getting somewhere
I could probably vector-fy it and then enlarge...
SD Upscale would probably be the better option for upscaling that instead of Hires fix then. SD Upscale will usually work well between 0.1-0.2 denoising.
ah good call
ill flip over
You tried vlad's yet, @supple raptor ? I hit a couple issues and put it on hold, but some seem to like it.
waiting for rocm support 
I tried it a few days ago and also ran into some issues that caused me to jump back to the UX fork, then SAG came out and needed either Vlad or Auto so I switched again to regular old Auto1111
tech support #2 now? 
i rlly liked how tiled diff and vae was in that fork
I actually also want help with updating webui 
just git pull nerd
it's gonna break so hard

already broke once, that's why i revert back
oop well im directml so i cant relate i got my own problems 
What's the UX fork? This one? https://github.com/anapnoe/stable-diffusion-webui-ux
yeah there's some discussions about memory leaks in vlad's... that's a no-go for me. I do too many higher-res generations.
Yep. Functionally it's the same as Auto before the somewhat recent Gradio version update, but with many UI improvements.
https://civitai.com/models/27574/super-easy-ai-installer-tool
anyone use this? sucks i cant use anything here that i wanna play with cos amd xD
Yeah I heard that this UX one might collaborate with Vlad, but we'll see.
I have AMD.. what's the issue?
oh... duh... you probably mean an AMD GPU. 🙂
yeah haha sorry
as with most of the fun stuff its mostly built on nvidia's cuda
was hoping to play with kandinsky 
Yeah, I heard that too. I'll try it again at some point, but for now I don't feel like Vlad is worth the extra hassle.
It shows a lot of promise. Will be interesting to see what happens.
I haven't really tried any of the other major players ... invoke or comfy or whatever.
Other than that other big one that came out when it all blew up... can't remember the name. Brainfart.
Doesn't Auto1111 support rocm natively now? https://github.com/AUTOMATIC1111/stable-diffusion-webui/wiki/Install-and-Run-on-AMD-GPUs
HKNY's or something...
HLKY I think.
I used HLKY first, then Auto, I've also tried out Comfy a bit.
not yet 😛
still the fork im using
Is COmfy the one that's node based?
I cant use comfyui for instance
nor vlad's fork
right below that it says
WHOOPS... when you accidentally leave a high denoise on SD Upscale. 😄
i have rx 6600
If you're going to try it, I would start by just installing Auto1111 alongside the repo you already have, test it out for a bit to make sure everything works as it should and if you're happy with it then you can fully swap. You can even symlink your embeddings, models, and outputs folder from your old repo to the Auto1111 install so that they will be shared.
you lost me at symlinks
x;
so what youre saying before i got confused
install this repo in the same folder as where i keep ishqy?
lilke below or smth
I use sandboxing for all this... makes for easy switching and reinstalls and such. Something to consider.
You can git clone it to the folder you're inside in that screenshot that'll add another folder there for Auto1111.
Fantastic thats simple for my smol braim
Ima give this a go
im hoping this will allow me to use some features that are broken on the fork im using like clip
"Don't look behind me, look at this fire instead."
wait is this for linux or windows
shouldve asked this first since im windows
It should be Windows I think since there are linux instructions below that.
yeah saw that and it made me think about that real quick
the power of friendship and anime, very glad you pointed this out lol
Happy to help 
weirdly enough its running off my cpu lol

new installation of auto1111?
maybe it is for linux afterall and failed to mention, more mind power than i have available to figure it out
yee
not a big deal the other fork still works perfectly fine
Im surprised Clyde allow this tbh
Symlinks don't seem to work with Vlad, unless there was an issue in my setting it up. This is the end of what I see after symlinking the main folders to my auto1111 setup and starting vlad's webui.bat
FileExistsError: [WinError 183] Cannot create a file when that file already exists: 'D:\\0\\SD-Vlad\\automatic\\models\\hypernetworks'
good or trash?
very slightly kinda reminds me of the character viney from the owl hosue
except without a pony tail
not sure why, I guess it's the art style and facial structure + brown hair
No idea who that is, but right on. 😄
this person
I see
Oie
just went and checked out some of the latest model/lora releases
pulled a bunch
so... sigh...
omg guys guys guys
idk im stupid but this is so cool to me im sorry 💀
im on a github page (thats my lora)
?

bwm typing fast?? never heard of it
You have to delete the existing directory where you're adding the new symlink. That error is telling you that D:\0\SD-Vlad\automatic\models\hypernetworks already exists so it can't create the symlink because it would share the same exact filepath. (To be clear, the one that needs deleted is the empty directory that's in the new install not the one from your old install that has all of your things in it.) Alternatively, you can move all of those files outside of both directories and then symlink them into both installs which is extra work, but would allow saving your models to a different drive for example as well as making it a little harder to accidentally delete your models if you decide to delete the install to start fresh or something.
Wow, that's a lot of text! Thanks. 😅 Sadly, the symlinks were already created successfully and in place. It was vlad's SD that was telling me that for some reason.
Hmm... I would think that error could probably be ignored then since it should find the directory anyways. Probably just an error coded into the setup script.
Except for the fact that it said "Press any button to continue" and then promptly exited, lol.

I tried that a few times just to be sure.
That's just dumb then
LOL
Would probably have to delete the symlink, let it create the directory like it wants, then delete that directory and recreate the symlink.
ᶦ ᵈᶦᵈ ᵗʰᵃᵗ ᵗʰᵒᵘᵍʰ
I'm running it straight from the webui.bat so every time it runs through the install process before starting SD--this is likely the issue. I wish that there was a bat that skipped the install process.
wait, you mean you are trying to create the symlink everytime you start auto1111?


I doubleclick the webui.bat to start SD.
yup yup
The symlinks were created by the time I started it.
after starting it once, then try deleting that hypernetworks folder
then do symlink?
and it doesnt work that way?
Yeah. Thanks though, heh.
just thinking out loud, but maybe removing the check_pip section?
that doesnt make sense tho
ignore me 

Usually with a problem like this, I often enjoy others attempts to solve it being met with just as much confoundation as myself.
It helps alleviate the frustration, lol.

Now to try to refine this prompt...
i thought of a way that might work
but it's gonna be a pain to do everytime you want to start vlad webui
Well, I brute forced just moved the model/network folders over to vlad's for now, so now that I have it running i'll just play around with it to even see if I want to continue using it.
does this extension help with low vram?
probably doesnt do much to vram, just let you know how much it's using?
f

and is this the one you download to use control net?
my controlnet extension is probably really outdated, so don't quote me on this, but looks correct
as long as it works man
Getting closeeeeeer
Someone posted a Renaissance Oil Painting model on Civitai and I absolutely love it, lol. Most of the deities in my story emerge during the Renaissance and I love seeing that its still able to keep al lot of the facial details of each character despite being t2i from their prompts
uh, help. junji ito has appeared unbidden again


hogwarts classroom
LOL. One of these successive gens is not like the others.
I never managed to get this lora to work properly :C

at least with the 1 model I tried it with 

That's it. I may have to put "horror, scary" in the neg.
we might not have solve world hunger, but at least we managed to put a cigarette at waifu mouth

even her shirt print is smoking 👀

I complain a lot, but I'm also getting gens like these... 
super cool t-shirt tho 

ahhh it's done loading
Is not sad or depressing! It's awesome! 
this one looks like the plot for an ending of an anime
right at the ending of an arc as they escape a den
POW research?
TYTY, you too 
u better join

nick pain
finally got controlnet to work?
Not yet. I'm making the last of the images for my review of Aurora, then I'll install the latest controlnet to Vlads. 
why is she so cool 
you can go stay in the same cell as her


well, don't mind if I do..
hmm... maybe I should try to add the copied city to the background of this one.
Nubby 
Smol!
the copied city? 
One punch man
@hearty kindle
Currently, there is no bot on the server that generates images. However, there are plenty of other ways such as the official https://beta.dreamstudio.ai/ website or by running Stable Diffusion locally using your own hardware! Check out #1080946152318443610 for more details! You can also stop by #1025467151206854736 for any issues you experience while using DreamStudio or #🤝|tech-support for any problems you encounter while installing it locally!
One of the areas from NieR: Automata

ooooo tasty depth map
do it 
I'll try it tomorrow. Gonna boot up the chatbot and go watch tv for now.
Anybody know how to mix in helltaker lora with other lora
Like this
Yet when i do it Its bad like it mutated
What strength do i need to put in the lora prompt?
levels gotta calibrate your numbers and prompt
0.6?
depends on the lora you're using , some lora play well with others, some don't, its on a case by case basis
Interesting
i think it's about testing, if you are using 2 lora, test each lora first, if they are working on their own, then include both, but start with much lesser strength with both
What's the minimum strength?
probably like 0.00000000~1?
but that will probably be like 0
so, test from 0.1 and up?
if you wanna test a bunch of strength, you can use plot xyz for that
Aight ima test it out thx for the help
How to make money online
First, make a nsfw model, then sell it online
what prompt was this?
sometimes it depends on the model as well, some LoRAs deepfry the image on certain models
I did it based on experience with my model and intuition. Then when it fails it goes to trial and error
being admitted to the hospital in 4 hours
surgery tomorrow morning
everything alright?
All the best buddy!
Sending virtual hugs! 
thank you
that sucks, I hope the surgery goes well, wish u the best of luck 
yeah if it doesn't that only means one thing and I don't want that
Think positive! 
I'm a realist, I'm incapable of thinking positive or negative at most times
aye thank you
some eyes I've made...gonna make a ton more of every eyecolor there is and then make a lora
forgot those
Living Ships do battle in Davey Jones Ocean
ofc I add freckles on accident 
ahhh, definitely accident
I could use one of those right now
raiden :1.0 , helltaker :0.55
bunger


what's a good free remote desktop software for windows 10/11, so I can access my comp at home and keep activating SD if it crashes
I still use chrome remote desktop on my android phone, but I've heard its changing. It still works great for me though.
I dunno what that is
It's pretty decent for doing simple things

I only need to click things
but when too much is happening my internet can't handle it 
This will work perfectly then.
but I can do without generating for a week or two
okay guys time to sleep
assuming I survive the surgery
which I should
Good night
G'night Smol! 
gn smol
good night smol

👋
Uh, careful with that muzzle direction
Get that out of that girl's hands
she's just looking if there's something stuck in the barrel 
"I was made by AI, it taught me how to hold a weapon"
"Dont worry I can do this! GANBATTE WATASHI WA!1111"


these are better
boom
some simple background brushes and sketches in clip studio paint pro and then img2img and I get these
@cyan burrow daily thank you for the model, its fucking incredible
And great lora support
img1 "I have an idea"
img2 "We need to buy and stockpile all manga before it is destroyed!" His Lt Cmd panics in background
img 3thinks of another idea and Lt Cmd panics at the thought
img 4Continues thinking, aaahh hmmm
img 5 "Ready all our fighter planes. We're going to the moon"
img 6: Lt Cmd: "That's not possible! You can't breathe on the moon! You'll die!"
Img7 : General is struck by the harsh reality "That's impossible!"
Wow, that image you generated is absolutely stunning! You have a real talent for creating lifelike and captivating visuals. Keep up the great work!
thank you
I guess I should get a netflix sub for my time in the hospital
that was chatgpt, what i'd say is, this looks like it's located at the eyrie from game of thrones
would be quite unpleasant to be there desu
my apologizes
it was a social experiment of sorts

In other news... I got SAG working on my regular SD!!! 

Vlads just killed anything that I wanted to upscale, so we're back to regular Auto.

Vlad's did have something that I want in my 1111 though. Off to get the CLIP interrogator I go. 
its all hype, no substance
It had some nice features for sure. The only downside I had was the fact that I couldn't upscale anything because it thought I didn't have enough vram... 
Also, does anybody know what do?
Series of pics generated from a description from my story of Myōken-shū, a (fictional) emergent Buddhist sect in Feudal Japan that believes magic is a method to reach Enlightenment. It eventually gives rise to a group known as Reikizukai (s. reikishi) which became an emergent social class of magical practioners that ended up clashing with the similarly emerging Samurai class for dominance.
Last pics got duplicated
Ehhh, comes and goes, I prefer fatal cause it has nicer colors
it's darker / more contrasty than source image
Yeah kinda
so cinematic, which is doing what it was designed for, animation
i Hate cinema
captivating
she's drinking Cerenkov radiation
drink of the future
what did you prompt to make the drink glow though ?
yesterday many hours were spent trying to figure out glow effect
Master prompter
Though in actuality this might be that one effect where you see patterns where there are none. Only way to actually know is do hundreds if not thousands of comparisons.
I prefer magicians. Taking images from the latent space through invocations and making them manifest.
Its not the first time

omg I can't take it anymore i'm going to proooooooooompt
Ok peeps. Which one do y'all like better?
A/B?
the one with a fish in her mouth
Thanks. That's the last gen then for my Aurora model review!
Pr"oo"mters

B
Looks fucking sick
Aaaaagh, then I may have to break my rule and include both.
i dont like aurora its too generic for my taste
Ah yes. What I posted above is so generic...
💀
The magic is in our fingers my friend. Aurora allows for magic to manifest. Surrender yourself to its power. 


my prompts are pretty complicated, and it cant do what i want lol, seems like some of the prompts were like stripped out of the model that generate effects
Sounds like you're used to doing things a certain way. Nothing wrong with that. I wouldn't blame the model though--I've seen what it can do given the right direction.
cant believe im a professional prompter lol
1girl
job done
nice
true
almost time to take off 😮
anyone know a good remote lan software? like hamachi
free
and simple
Le fishe au chocolate
Hamachi good
There is also ngrock I think
is hamachi free though?
Ye
I think
But you have to connect each new network you wanna use
If you use ngrock then you can connect from everywhere
https://www.zerotier.com/ Free tier has 25 nodes
well I just want one network to share with this computer and my laptop in order to access my SD remotely
without a gradio account thingy which is unsafe and unstable
lmaooo
no obviously not since I will be at a hospital quite far away
Oh ok
ugh gotta sign in and everything on hamachi and make accounts etc
no time for that
oh well, fuck it then
unless there's a zero-time click for a remote desktop of high quality access that doesn't waste packages
Never heard of that
its pretty good alternative to hamachi
Looks like it
Google remote thingy
you can run it on every platform https://www.zerotier.com/download/
Chrome remote desktop
yeah trying but it's not running properly for even install
I should try it sometime
Use stable horde
And download like a few models and all the extensions
And boom ez setup
Or you can use nocrypt's amazing colab
Comes with most useful extensions
I dont really think I will be in a headspace where I can do all that
I will be going through MAJOR surgery
Oh right lol
the kind where I can die during
and where I'm certain to lose a few organs
If you not die, would you send selfie?
its all doom and gloom for you eh? no bright side at all?
the bright side is if I live
He has us. That's the bright side hehe
so all doom and gloom
so what do you expect from me @subtle crown ? Put on a positive mask for the sake of the world? 😛
world dosnt care about you, you should put positive mask for the sake of yourself
I think there are some times where we should be sad and this is one of those times
Instead of lying to yourself that you are happy
yeah after dealing with this since 2020 and having it come back now with even less chance of survival with one of the most aggressive cancers in the world
I'm tired
so fucking tired
I don't have the energy to fake happiness
gotta go, will beo nline on laptop
Be like guts from berserk my dude
Be like Sisyphus in "The Myth of Sisyphus" by Albert Camus
he needs a big ass sword 😛
He got the other kind of sword if you know what I mean
no use for that, girls only want chads, if your bank account is not 5 million, your not using that sword
both
anyone knows how can i fix this issue?
whenever i press interrupt or skip it doesnt work anymore for some reason
it used to but stopped i think after updating webui
common issue no way to stop it that i know of, other then terminating webui
Ha! Luckily I did include both in my review, freshly posted. Civitai seems to be giving me no end of trouble with it though. Hopefully it'll show up alright.
Let's see if this links correctly: https://civitai.com/posts/169820
Yeah CivitAI very buggy 🥹
no lora?
i really like the image
it looks like real drawing
No 
yea i did that but its annoying that i have to open webui and close it
and takes like 3min each
The only embeds used in the examples were auroranegative, khfb, or sometimes easynegative/bandhandv4. That's it.
Probably my favorite reviews due to the versatility and show its capability 
I tried. Lol
I can see that 
You've got a lot of great examples of the versatility in your model examples though.
yesh 
I hope they fix their database at civit, it's starting to be tiring ^^
Civitai didn't want to accept this one for some reason though.
worse than cylde??
Lol
me too 🥹
too black?

recently, 1 model out of 2 I publish isn't indexed : doesn't show in new models nor on its tags, only on my profile. Super easy x)
Nah, I interrogated it and it shows up with "teen-girl" part of the prompt, so maybe that's part of it.
Them trying to prevent cp and all.
I never get any notifications from users or reviews anymore; Also their UI is much more slower
Yikes. And I thought my review issues were bad.
I still used their UI to do lots of examples, it has some good parts, but it's scary to spend too much time in it when you know it can just get nuked by their mistakes in a few minutes
lol yeah
I think the 1st one might be favorite, there's something to it XD
reminded me of my big sis
Aww
Yeah, it's with SAG.
Oh. Maybe I should've mentioned that. 
Well, the fundemental is the same. SAG just give some extra clarity and contrast.
yeah it's not much diff imo

BUBBLES!
ilove bubbles
Bubbles! :3
so small...
ZOOM IN

Anyway, now that that's accomplished I must rest. And enjoy Aurora for making things that I want to see tomorrow. Goodnight peeps.

goodnight 
im using SAG scale 2 and 1.5 mask, the quality is really nice, as you can see for yourself 😛 BUBBLES!























