#🏞|general-with-images

1 messages · Page 36 of 1

glossy herald
#

I don't have a prompt ready for that

#

let's make one

sleek ingot
#

space is cool too hehe

#

castronauticat

glossy herald
#

trying for more realistic

sleek ingot
#

that looks good and form fitting

#

im just blown away

dense tapir
glossy herald
#

forgot my suit, just chillin' on mars

sleek ingot
#

lol love it 😄

glossy herald
#

got it

sleek ingot
#

LOL

#

magical

glossy herald
#

some wtf too

sleek ingot
#

😄

glossy herald
#

I love how the helmet got cat hears

sleek ingot
#

LOL YES

dense tapir
sleek ingot
#

freaking amazing

ripe cedar
#

I use blender and can generate line-art and depth maps from complex 3d scenes. But I don't know how to make the color consistent from different angles

#

I use ControlNet too

glossy herald
#

I'm not sure on that one, it seems like a complicated thing to get yeah

#

with some training maybe ?

#

nah

ripe cedar
#

I want to make comics so training for each scene is way too much.

glossy herald
#

yeah it's really not practical to train each scene for the background itself
for the characters it can be though, quality wise

dense tapir
ripe cedar
#

I can generate line-art for characters too

#

And put them into scribble

glossy herald
#

but to keep the colors on the background... maybe by img2img, populating by hand the good colors beforehand, but that again seems like quite the chore

ripe cedar
#

For example this is generated in 3d, no AI used

dense tapir
glossy herald
#

I did a very small 4 panels comic, and the bg coudn't stay consistent, I needed so many seeds

glossy herald
ripe cedar
#

Scribble doesn't quite seem to understand the depth of scribble, but the depth doesn't get the intricate details of scribble. If I could get both that would be great

#

Yeah 3d+AI is really powerful. I just need a workflow that works

dense tapir
glossy herald
#

really good for architecture

dense tapir
ripe cedar
glossy herald
#

let me try some

#

mlsd

dense tapir
#

Notice the flag has another star since it is in the future?

#

*2 I meant

sleek ingot
#

haha yep I see those

wispy nest
#

Hi there. I am new to SD (1st time). How do i start generating images?

glossy herald
#

hello there hardy

#

welcome around

#

we don't have a bot for it, but you got other ways

#

here is the faq

crystal deltaBOT
#
FAQ: How do I generate images? Is there a bot on the server?

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!

dense tapir
glossy herald
ripe cedar
wispy nest
dense tapir
glossy herald
# ripe cedar How would that change when taken from other angles, distances, and perspectives?

well, right now, I used a very generic prompt, "a building in an american street".
To keep it more stable, I would need to over specify materials for a start
Then, I could use a second extension like "couple latent", that let me use multiple prompts on multiple parts of the picture, to be able to push for specific things in some places
Lastly, personally, I would also go into img2img, and "seed the image" with the right colors. Meaning I would use that image you shared as controlnet, and use it also as base picture in img2img after having painted it with colors I want, just "putting the right color palette" out there, so that the AI can morph those colors easily while following your controlnets

ripe cedar
#

I see, thanks. I'm going to have to do a lot of experimenting.

wispy nest
glossy herald
#

you may want to try some other tools, like comfyUI, if you are making a very specific but also very similar pipeline from a picture to the next.
It's a lot more complicated UI, but it lets you over specify how you want the picture made.
Like, you can pre sample the background, the characters separately, sum up the latents in a single bigger latent picture, finish the computing, while pushing different controlnets on everything.
Sure, you'll take 10 times more time setting your pipeline up, but then you'll just be spiting out content

glossy herald
#

1/ you can use the website the bot linked

#

there you can do some free pictures for free credits

#

2/ you can install the thing on your computer

#

3/ you can, if your computer is too old to run it, use a cloud service like google colab to run it

chilly lance
smoky oak
#

I am getting so hyped for SD 3.0 man

#

if its even half as good as that SDXL demo, and it can be trained, SD is gonna be for sure on top of AI gen again

peak knoll
#

/prompt

ripe cedar
# glossy herald you may want to try some other tools, like comfyUI, if you are making a very spe...

I can generate line-art for characters and also their facial expressions (rigged characters). And color them (but without shading because with shadow they look bad). I think SD can't make facial expressions with fine control yet (I think there was one that came out just today though). So would character Lora be better or maybe just trying to use AI to color the line art generation? I'm just trying to think of ways to combine the different techniques

wispy nest
crystal deltaBOT
smoky oak
#

One of the devs on twitter said "very soon"

#

so here is hoping man

wispy nest
#

about 3.0 exactly or sdxl?

smoky oak
#

@glossy herald I see the name change haha

smoky oak
wispy nest
smoky oak
#

one of the admin was in here chatting about it yesterday

wispy nest
#

so...what's exactly happening - SDXL release first and then, after some time - 3.0?
Or SDXL won't even be a thing on it's own?

smoky oak
#

so theoretically... We could be in for the biggest image gen improvement the industry has ever seen. Hell, even just that SDXL demo was a bigger improvement than all refined versions previously put together

glossy herald
smoky oak
#

my guess at least

#

its probably just a base for 3.0

wispy nest
#

hmmmm

smoky oak
#

so really, we could have an even bigger improvement than the already monumental improvements with SDXL

#

And I don't say that lightly

wispy nest
#

yea both options sounds interesting anyway

smoky oak
#

SDXL is unprecedented levels of improvement

#

I'd go as far as to say with my testing, its text encoder is equal to or greater in power than midjourneys

#

it listens so well to what I ask from it. It as a base got less things wrong than all of the others got right combined

#

SDXL misses less than all the others hit

#

and I say that with ~75 multi version tests

wispy nest
#

I didn't test it that far, but yea, some aspects looks like a decent improvement

smoky oak
#

decent is an understatement. The improvement between 2.1 and SDXL is several times bigger than 1.4 to 2.1

#

Also, SDXL can do weapons

wispy nest
#

yea, tested it a bit too, but holding objects is still pretty bad

smoky oak
#

realllyyy good weapons

smoky oak
dense tapir
ripe cedar
#

What are the ? on 3.0 so far? Like the unknowns that could be really bad

smoky oak
#

1.5 is unusably bad as a base, and SDXL as a base kicks 1.5 finetuned in some ways

#

so imagine how far we can finetune SDXL

#

SDXL listens insanely well, from my testing

wispy nest
#

thinking my imagination doesn't go too far

smoky oak
#

what I am saying is 1.5 base is so horrible that it gave this for "polygon vector art of a tree frog"

#

and even with it being that trash, we can get images like this out of it finetuned

wispy nest
#

yea I remember your images, I see where you're going with it

smoky oak
wispy nest
#

Listening to prompt and understanding it is one of most important things yea

smoky oak
#

that same trash base was able to be finetuned to give results like that

wispy nest
#

I think I had decent results with 1.5 or 2 base models when I started, but it was really hard to prompt it lol

dense tapir
smoky oak
#

so a base that was so unreliable its basically unusable can achieve near photorealistic quality and 8k+ images pretty easily when finetuned. So one can only imagine how much better we could make SDXl/3.0 with the same training

wispy nest
#

we'll see thinking

smoky oak
#

Cause to say SDXL blows the other bases out of the water would be a huge understatement 😅

#

This comparison between 1.5 base and SDXL 2.2.2 beta shows what I mean

It was supposed to be a night time photograph of a husky puppy sitting on a beach. All 1.5 got was husky puppy, while SDXL 2.2.2 got everything I asked for

dense tapir
#

That isn't nighttime that is dusk

smoky oak
#

its a properly exposed night time image in SDXL

dense tapir
#

then it sucks

smoky oak
#

it just doesn't have noise offset, chill out lol

#

still the only SD model that can do anything past sunset

dense tapir
#

When I ask for nighttime I want nighttime not that

smoky oak
#

(without noise offset)

smoky oak
dense tapir
#

I still really, honestly, do not like noise offset

smoky oak
#

you're the first person to say that the SDXL image didn't look like a properly exposed night time image. And as a photographer of several years, its exactly that 😅

#

now if you wanted an under exposed night time image, then thats a different thing

dense tapir
#

Not in the mood to argue so you continue oin with the overhype train 24x7 and we shall see. I do hope it is as fantastic as you say.

smoky oak
#

Its not overhype if I am showing actual examples

#

I am not saying anything that isn't easily observable lol

#

it handles very specific prompting properly, something none of the others can even do on a basic level, so IDK

#

I have a feeling you'll find some reason to hate it anyways. Seems like everything with SD has a "I do not like" or "is pretty bad" attached to it, regardless of its comparable use or quality

#

like noise offset. It does what is needed, and it does it reliably, but you don't like it

wispy nest
#
  • to that part
    we shall see. I do hope it is as fantastic as you say.
smoky oak
#

I found a way to basically gain a view into what images a model was trained on

#

generate with no positive or negative and the biases literally show themselves

#

protogen infinity

#

faetastic

#

much more flat, interesting

smoky oak
#

Anybody know any good alternatives to ESRGAN_4x?

Something that can give extreme detail like ESRGAN_4x, but without the absurdly oversharpened look

static tusk
smoky oak
#

Oh, I was meaning ones that were external. None of the built in ones have nearly the detail of ESRGAN_4x unfortunately

static tusk
#

LDSR is very vram intensive, so I need tiled vae for that

smoky oak
#

here, let me try something

#

Gonna compare all of the upscalers again

static tusk
#

I tried esrgan_4x for a while and found shadows to be too harsh

smoky oak
#

yeah, it makes thinsg crunchy, but damn is the detail out of it so much cleaner and higher density than the others

#

ESRGAN 4x looks higher detail than all the others at 2x the res

static tusk
#

ideally you want to upscale at 4x not 2x with an upscaler that has_4x at the end

smoky oak
#

a 4x upscale at the same time using ultimate SD is a good way to get huge seams

#

I incrementally upscale

static tusk
#

yeah you def wont get the best quality that way as the upscalers are built for 4x as mentioned

#

or they are made for 2x etc

dense tapir
wispy nest
dense tapir
stoic pivot
#

Four images designed after snapshots taken within Second Life.

dense tapir
smoky oak
#

Barrack O'Bummer

dense tapir
#

Yeah, but this lora is smashing it

#

2,160 steps, 3 epochs, 6 images D-Adaption

smoky oak
#

I only said that cause I thought it was a fail-

#

whats the LoRA of?

dense tapir
#

nope a raging success

#

a cartoon

#

wtf?!?

odd locust
#

cute studen

dense tapir
#

weird, I am training now with the same settings and the images are all jacked up

#

I'll see when it hits epoch 1

smoky oak
#

@dense tapirBTW, are you training on collab again?

stoic pivot
#

This was interesting: I asked for the text "Great Work" in a image, and as normal AI did fail (SD 2.1), but I locked the seed and ran it through "Hires. fix" and forgot I had LATENT on so it became unsharp, I then selected 4x-Ultrasharp and the text was fixed. So is I first run a upscale with LATEN and then Switch to Ultrasharp I do get better result.

dense tapir
smoky oak
#

ahh

dense tapir
#

I am using my port of the kohya_ss

#

since colab is headless anything designed for a display will not work

#

having ripping good losses now in 2.1 for 2.x, lol

#

I mean 0.1x kinda losses

#

instead of 0.4x

#

that really is interesting. Hmmm

#

not sure how it managed to fix it post fact though

#

anyone know how to make kohya change sample steps for its images?

#

Man, it is suffering for some odd reason

stoic pivot
#

02325 = Original Image.
02327 = Image use LATENT trick to upscale.
02326 = Image upscaled with 4x-Ultra sharp

4x-Ultrasharp have more details but the image mage with the LATEN trick is closer to the original.

dense tapir
shut silo
smoky oak
#

@dense tapirPardon my botherance, but could you please remind me how to get that Ai layer that makes image gen several times faster?

#

I don't remember where we talked about it, otherwise I would search for the link to it

#

oh, I think I found it

#

VoltaML

dense tapir
#

Yep, volta

wispy nest
#

i get terrible results following the cow-girl prompt

#

sadge

dense tapir
#

4070 is due out in a couple of weeks and it is almost priced right for those who want 12GB cards.

#

50 less and it would be the best bang for the buck (price to performance) of all the 4k lineup

static tusk
#

I wish more tech review channels on YT were focused on SD and reviewed all the cards as they were released. sadly they only do gaming

dense tapir
#

I have said the same thing for six months now.

#

sucks

wispy nest
dense tapir
#

I just ran across that a few days ago but didn't really read it. Sounds a bit of more of the same at a cursory glance.

fickle gorge
#

While I imagine the 4070 would be considerably faster than a 2060, is there be something the 2060 can't do that the 4070 can?

dense tapir
#

12gb

#

if you wish to train that is valuable

shut silo
#

whats the cheapest gpu for 24gb vram either 3090 used or a 7900xtx?

dense tapir
#

just to squirt out pics then stay 2060

fickle gorge
#

But the 2060 has a 12 GB version

shut silo
#

got a 2080ti and it cant do shit

wispy nest
#

What kind of shit are we talking thinking

dense tapir
#

personally, stay away from AMD

#

for SD

wispy nest
#

and for rtx...yea

#

and for any new technology basically thinking

dense tapir
#

unless you know Linux then go for it

shut silo
#

@dense tapir had a 69xt didnt try sd on it

dense tapir
shut silo
#

reee

fickle gorge
#

Because if the difference between 4070 and 2060 (12 GB) is just speed, then I wouldn't need to spend as much... I'm used with slowness on SD already

wispy nest
#

comfyui is so good but theres no face restoration... auto111 is so slow for me and crashes my xorg a lot

dense tapir
#

if I had any other card I could buy that would gen and train and had at least 16gb that is a new card I would seriously look at it. Frankly Nvidia has all of our parts in a vise and they know it.

#

I can get two of those three but not all three

fickle gorge
#

Is Intel Arc usable with SD at all?

dense tapir
#

Worst of all of them

#

horror stories for gen and forget training last I was told

wispy nest
# wispy nest comfyui is so good but theres no face restoration... auto111 is so slow for me a...

You don't need face restoration if you're using updated vae and not generating persons too far away.

But well, even if you actually need to fix face - you can just throw image you want in SD or here
https://huggingface.co/spaces/sczhou/CodeFormer

Buuuut again, it doesn't work for all styles and images, often it's just better to get right resolution for your gen.
If you know you're going for full body - go for higher resolution. It'll fix face

#

kk

#

generated some images using stable horde it was horrible

#

thinking no idea what that is

dense tapir
#

distributed computing

wispy nest
#

army of pcs helping each other make images

#

thinking smart words doesn't help dumb persons

wispy nest
fickle gorge
#

I'm divided on whether I should delete some pictures. I like so many of them. But there are thousands of them and even the file manager is struggling

#

(ignore the ingrish)

dense tapir
#

wth? My lora keeps chopping off her legs :/

ripe cedar
dense tapir
#

she has no legs below the knees

ripe cedar
fickle gorge
#

Were there pictures showing the feet?

fickle gorge
dense tapir
#

The pic above her was less steps and look at it

wispy nest
#

dude wtf

#

nothing I do is making my images non full-body

fickle gorge
# dense tapir

This one? There is no feet too, but it makes sense on this one

dense tapir
#

trying a different prompt

wispy nest
#

but in this case you shouldn't specify clothes or details below upper body or below face

#

and probably shouldn't go too low on cfg

ripe cedar
#

Anyone know if comfyUI can save presets for different types of workflows?

dense tapir
#

Different prompt

wispy nest
#

thinking maybe I'm saying dumb things

ripe cedar
#

I haven't tried. But I thought it might be possible to have a preset based on if someone wanted certain ControlNet nodes, models, etc already ready. Instead of having to make new nodes each time.

Not based on image, because I'd want to make images rather than reverse engineer based on another image

wispy nest
#

I like SD's way, I'm storing my prompts as images, easier to see what it represents without reading

dense tapir
static tusk
dense tapir
wispy nest
#

After I generated tons of demon girls I can't look at diablo 4 logo anymore, it looks like it's made by AI now...and it doesn't even look like good gen 🥺

ripe cedar
#

I think the halo that news sites love to use aren't so subtle any more. It's something they slap on all over

wispy nest
#

as reference lol

static tusk
#

lilith is def a 3d model

dense tapir
wispy nest
dense tapir
#

I love the spare robot leg

#

black tyres are so 2050s

shadow surge
#

John wick futuristic

dense tapir
grizzled sage
#

using controlnet with jsut a white sphere works to make very discord profile picture worthy images

#

sorry. i mean white circle

dense tapir
summer viper
brave horizon
#

hh

woven dawn
summer viper
#

like in a fairy tale

grizzled sage
#

reminds me of the galloping goose trail i used to walk every day to work

#

the circle control gets awesome framing on lora portraits too

summer viper
grizzled sage
#

or hair, i guess

dense tapir
harsh dove
#

Robot red

dense tapir
grizzled sage
dense tapir
dense tapir
smoky oak
#

@dense tapirI am looking to trying VoltaML for insane speed gens

#

aiming for sub 1 second 512x512 images

grizzled sage
smoky oak
#

ahhh haha

#

I see now lol

#

that explains the ROUND

#

@dense tapirWould you possibly be able to help me source tensorRT 8.4.1.2 and PyTorch 1.12 cu11.x?

#

I am not sure how to ensure I have those exact versions

dense tapir
#

that is a bit older version of both

smoky oak
#

yeah, I am just not sure how to force those specific versions in my SD install is all

plain heath
#

anywhere I can put intentionally unaesthetic images like this

smoky oak
#

I wonder how many times multiplication I can expect to my SD gens

smoky oak
plain heath
#

i seriously want to crush it so bad

dense tapir
#

to see if you have it pip show nvidia-tensorrt

smoky oak
#

is that a command I run in the root?

dense tapir
#

yes

#

sorry, cat said dinner time

smoky oak
#

alright, running in the root of the SD install

#

well, my second SD install

nocturne pebble
#

Can anyone tell me how i can generate images ?

smoky oak
#

@dense tapirSays I have no tensorrt

dense tapir
#

pip install tensorRT==8.4.1.2

smoky oak
dense tapir
#

which cuda do you have?

smoky oak
#

I have

#

11.8? I think

#

how do I check again, sorry about all the questions lol

dense tapir
#

in the root nvcc -V

smoky oak
#

there we go

#

yeah, 11.8

dense tapir
#

thought so

smoky oak
#

ok, so I need a Cuda version at 11 or higher

#

which I am good on

dense tapir
#

I do not even see that version

smoky oak
#

So I meet the cuda requirement

#

and I meet the driver requirement

#

just need the pytorch and tensor RT

grizzled sage
#

i got 11.8 and 12 installed but i dont think i needed 12. it was an effort to boost my ada card

#

new pytorch came out this week i heard

smoky oak
#

I need specifically 1.12 pytorch

dense tapir
#

python -c "import torch; print(torch.version)"

smoky oak
#

so

dense tapir
#

my native at the desktop is 1.13.0

#

+cpu

#

in auto I dunno

grizzled sage
smoky oak
#

python -c "import torch; print(1.12)"

#

is that how I do it?

dense tapir
#

no

grizzled sage
#

compiles models to optimal tensors

dense tapir
#

version literal

grizzled sage
#

useless to us. distinctions. lol.

smoky oak
#

pytorch 2 is worse than 1.x for now

#

until auto implements it

grizzled sage
#

lol

smoky oak
#

its around 2it/s slower for me

grizzled sage
#

there are branches with it i think

dense tapir
#

python -c "import torch; print(torch.version)"

#

fuck discord formating

smoky oak
#

so type it exactly like that, I am not replacing any variables?

dense tapir
#

yes

smoky oak
#

2.0.0+cu118

dense tapir
#

shit, lol

smoky oak
#

this was the SD install I did for the pytorch 2 one lol

dense tapir
#

get out of that thing

#

zap it

grizzled sage
#

oh its just a pull request and it is filled with people using it and misunderstanding the purpose lol

smoky oak
#

brb

grizzled sage
#

slower not good. we want that compiling feature, plus also speed on base models is good too. you got a 4090 yeh?

#

one day the webui with all it's extensions and sloppy ui goodness will know compiled models

smoky oak
grizzled sage
#

ooo 3060ti kay

tawdry carbon
#

Hi

smoky oak
#

this change should hopefully see my 3060ti above 40it/s

smoky oak
static tusk
#

but can you still use embeddings or train with it?

smoky oak
#

yes

dense tapir
#

yes

smoky oak
#

yes to what I asked, or what they asked?

dense tapir
#

both

smoky oak
#

ok

grizzled sage
#

yes

static tusk
#

lol

smoky oak
#

I am trying to install VoltaML, which can speed up SD by several times

dense tapir
#

wish my card had tensor cores but not on pascal cards as they came with Turing cards next gen

smoky oak
#

some things it speeds up by 10x+ over pytorch

grizzled sage
#

volta is awesome yeah. works on 4090 tho too lol. it recompiles a model and i haven't seen a lot of research on how well that performs

smoky oak
#

so I should be able to do sub second image gens

dense tapir
smoky oak
#

@dense tapircan you please remind me how to force install pytorch 1.12?

#

I deleted the venv

grizzled sage
#

stability was talking about fast diffusion too. i think its the same idea, compiled models. but the stuff is really buggy to work out maybe. tbh i been really busy doing image batches and experiments and haven't dug into much news on that topic lately

dense tapir
#

just rerun auto

grizzled sage
#

--reinstall-pytorch but also deleting it does the job too

dense tapir
#

then check it

smoky oak
#

it will install specifically 1.12?

#

alright, I'll try it

grizzled sage
#

or --reinstall-torch actually, iirc

dense tapir
#

probably not

grizzled sage
#

don't matter cause deleted

#

these days it installs 1.13.1

smoky oak
#

I am running it, making venv

smoky oak
dense tapir
#

yes, then activate that venv and check

smoky oak
#

hope its an easy fix

smoky oak
grizzled sage
#

should just be going into venv and force downgrade maybe

smoky oak
#

it was just a problem to try to downgrade 2.x to 1.x, wasn't it?

dense tapir
#

yeah, that 1.12 is old

grizzled sage
#

changing the version in requirement_versions.txt maybe

smoky oak
#

I am just letting it do what it do

#

general awareness has yet to let me down on these mutual experiments lol

dense tapir
#

Pytorch is never implicitly mentioned only pytorch_lightning

smoky oak
#

They tell me how to do it, and I let them test vicariously through me haha

grizzled sage
#

yeah if you don't specifically need 1.12

smoky oak
#

I do

#

very specifically 1.12

grizzled sage
#

yeh lotsa paths to get there

dense tapir
#

All this shit is very restricted to a single version of things

smoky oak
#

yeah

#

but for these speed gains, its worth it IMO

dense tapir
#

Know what that says? they are hacked versions

grizzled sage
#

pytorch is called torch in python, for some weird reason

smoky oak
#

I still don't believe it can be that fast 😅

dense tapir
#

update breaks the hack like game cheating

smoky oak
#

it shows "Shell infrastructure host" using a lot of CPU, not command line

#

is that normal?

grizzled sage
#

i was reading that you can just make a new virtual environment in the same folder, call it something else, and point a copy of the launch at the new virtual env. but that's still out my level with python. i'm grinding to get there

smoky oak
#

I know nothing about code. Its something that has never clicked for me

#

luckily, I can rely on other people to help me understand the little I tend to need

dense tapir
#

what version exactly do you need of pytorch?

smoky oak
#

1.12 cu11.x

grizzled sage
#

11.8 would be best for compiling i imagine

#

might only be 116 though

smoky oak
#

oops

dense tapir
smoky oak
#

I paused the command line lmao

#

thats why it was frozen lol

#

there we go, now python is moving

dense tapir
#

see, 11.6 was the last version of cuda for that ancient 1.12.1

smoky oak
#

so I am gonna have to downgrade cuda?

dense tapir
#

yes

#

this is why I hate fucking with this shit as it snowballs

smoky oak
#

alright so run that exact command when the venv is done?

grizzled sage
#

if you have the 11.8 sdk installed, it should work fine with torch's 116

dense tapir
#

yeah

smoky oak
#

alright, its at web ui requirements

#

ok, its done

dense tapir
#

should and does is two different things

smoky oak
#

oh damn it

#

it did the middleware error again

dense tapir
#

ahh, ffs

smoky oak
#

this is the third time this has happened with my SD installs lol

#

do we remember the fix?

dense tapir
#

auto is a big pile of shit but it is the only real shit we have yet

#

I don't

#

I keep having to redo venv a lot lately

smoky oak
#

just a sec

#

I know I can find it

#

@dense tapirFound the fix

#

here is hoping

dense tapir
#

k

smoky oak
#

it fixed it before

#

I really wanna get sub second gens for the hell of it haha

#

it fixed it :>

#

no more middleware error. Tis a good day

dense tapir
smoky oak
#

god damn it

#

2.0.0

dense tapir
#

from root of auto but activate the new venv

#

he must have updated you

smoky oak
#

:C

#

can i clone my other install and do this on that install?

grizzled sage
#

you can see what the ui uses by scrolling to bottom

dense tapir
#

yeah, do that

smoky oak
#

alrighty

dense tapir
#

2/0 is slower

#

2.0

smoky oak
#

yeah

#

problematic moment

#

excuse me?

#

230GB!?

dense tapir
#

huh?

smoky oak
#

the copy is saying its 230 GB

dense tapir
#

venv is 230?

smoky oak
#

oh damn, I have 131GB models lol

#

oops

dense tapir
#

lol

smoky oak
#

ohhh

#

I can just paste the venv?

grizzled sage
#

yeh the venv is what is operating really

dense tapir
#

inside root activate that venv

smoky oak
#

I was copying everything 😅

#

ok, much easier lol

grizzled sage
#

venv transplants are fun

dense tapir
#

once it is activated

#

now those last two I do believe is in the requirements and he will screw you next git pull

smoky oak
#

I should be fine as long as I never update this install, right?

dense tapir
#

you're safe

#

I just checked

smoky oak
#

a good day indeed lol

grizzled sage
dense tapir
#

not via requirements it will not

smoky oak
#

what the hell is up with copying this venv

#

why is it so damn slow

grizzled sage
#

lots of small files

smoky oak
#

its saying 0 bytes/s

#

and I have super fast drives

#

its a full speed gen 3 nvme

#

its saying 3 hours

#

dies dead

grizzled sage
smoky oak
#

sighhh

#

pretty sure the copy just crashed

#

oh nevermind

#

it just froze one it got to the parts to replace

#

ok, this is at least faster

#

my drives are 4000/4000

#

ok sweet, the rest got up to 1.8GB/s lol

#

its done now

grizzled sage
#

lots of small files. probably was determining how to cache that and you had just stopped the bigger ocpying op, might've been cleaning up after that in the bg still

smoky oak
#

@dense tapirok, i give up

#

even copied the venv

#

pytorch 2.0.0

#

its saying my main install is 2.0.0, which isn't possible

#

cause I never updated

#

bro

#

@dense tapir your command is pulling wrong info

#

it just told me my main install is 2.0.0, but when I ran it

dense tapir
#

did you activate the venv

smoky oak
#

its 1.13.1

smoky oak
dense tapir
#

no

smoky oak
#

oh-

dense tapir
#

"./venv/script/activate"

smoky oak
#

but regardless, its even saying my main install is 2.0.0

#

oh rightttt

#

sorry 😅

dense tapir
#

your main machine has 2.0 on it

#

you infected it

#

lol

smoky oak
#

its 1.13.1

dense tapir
#

correct

smoky oak
#

and your command even after activating it, is still saying 2.0.0

#

even tho if I run it and check it, its 1.13.1

dense tapir
#

that is why python uses virtual environments.

#

well, no idea but that is the command to check it

#

other one is to open a python teminal after you activate venv and type
import torch
print(torch.version)

#

import torch
print(torch.__version__)

#

or whatever. I despise formatting

smoky oak
#

this is proving to be a lot more tedious than I originally expected

dense tapir
#

there we go

smoky oak
#

alright, how do I open a python terminal?

dense tapir
#

oh, it is trust me it is hell

#

venv activated? just type python

smoky oak
#

ok, I'm in now

#

checking

#

I don't know man

#

still saying 2.0.0

#

when its not lmao

#

this is the install its saying is 2.0.0

dense tapir
#

I believe the pyterminal so no idea

#

not gonna worry about it myself

smoky oak
#

assuming it was 1.13.1 like the program actually running says, what would I do from here?

dense tapir
smoky oak
#

ok, i tried that and it says it needs an additional arg

dense tapir
smoky oak
#

ok great

#

everything is broken

#

sighhhh

#

I can't activate the venv now

smoky oak
#

WTF

#

ok, i fixed it

#

for whatever reason it was called "Scripts" in my venv

dense tapir
#

now you run my previous commands it will say 1.13

smoky oak
#

weird how it just worked previously

#

also, the extra arg it needed was the URL lol

#

I didn't realize I needed them at the same time

#

it uninstalled torch 2.0.0 :>

dense tapir
#

yep

smoky oak
#

ok, its done

#

checking

#

awesome!

#

1.12.1

dense tapir
#

yep

smoky oak
#

ok, is there anything else?

dense tapir
#

beats me

smoky oak
#

so run this in the root?

dense tapir
#

yes

smoky oak
#

alrighty, lets see

#

should be the moment of truth I think

#

ok, its done

#

do I have the tensor RT now?

dense tapir
#

I think so

smoky oak
#

alright, lets try it

#

lets see if I fly

dense tapir
#

do they have anyway to see/test?

smoky oak
#

I have no idea

#

I guess if it goes fast thats a good sign haha

dense tapir
#

should be on their site some way to see

smoky oak
#

this is the site

dense tapir
#

yep, that is the site

smoky oak
#

oh my god

#

dude

#

They have a "fast installer" now

dense tapir
#

LOL

smoky oak
#

interesting

#

this suggests getting rid of xformers

#

its showing its slower

static tusk
#

torch 2 doesnt need xformers

smoky oak
#

not on torch 2

#

ok, looks like volta is not working

#

same normal speeds

#

gonna try their fast installer, screw this lol

dense tapir
#

ticket it then because I suspect you must be be on 11.6 not 11.8 as a lot of this shit is not backwards compatible.

static tusk
#

Ive been using torch 2 for a little bit now and instead of xformers I use --opt-sdp-attention

dense tapir
#

yes, run that

smoky oak
#

ohhh

#

its a "fast installer" through docker

dense tapir
#

will that work for non tensorcore cards?

#

pure

smoky oak
static tusk
smoky oak
#

it says NVIDIA GPU with CUDA support

dense tapir
#

yeah, T4 listed as they told me it would be much faster around 60% on it

#

A100 is the 400%

#

great, fucking docker

smoky oak
#

they have a step by step guide

#

its 5 steps

dense tapir
#

grab it

smoky oak
#

but I have no idea how to use docker

dense tapir
#

I might do this on colab

#

nor I

smoky oak
#

oh great, so neither of us know how to use docker?

dense tapir
smoky oak
#

it does

#

a massive amount

dense tapir
#

I never have, no.

static tusk
#

yeah the sdp attention command does the same thing

smoky oak
#

for me it allowed me to go from 768x1024 on my GPU to doing 2560x1080 gens no problem

dense tapir
#

oh, cool. I wonder if it is as unstable as xformers?

smoky oak
#

I have never had any xformers issues, personally-

#

and I have no xformers noise either

static tusk
#

no issues here, but I dont train so not sure

dense tapir
#

even the devs of it on FB research said it jitters and can't be helped

static tusk
#

like my gens arent visibly different

smoky oak
#

I have even pixel peeped mine, and no difference

#

same seed is same image

dense tapir
smoky oak
#

mine shows 1.13.1 still as well, even though I am on 1.12.1 now

dense tapir
#

2.0 sucked and made me so slow as I was already slow I ditched it

#

you have to rerun it all

static tusk
#

what Im using now
python: 3.10.9  •  torch: 2.0.0+cu118  •  xformers: N/A  •  gradio: 3.23.0  •  commit: 22bcc7be  •  checkpoint: 639d0db70f

smoky oak
#

why torch 2.0.0?

#

its slower

#

even auto himself said its slower

#

until he implements it

dense tapir
#

yep, he has to then it will speed up

#

for all of us

static tusk
#

the difference is so small, so I left it installed

dense tapir
#

even the torch devs said that

#

once auto implements it correctly it will be a nice gain

smoky oak
#

10.5 it/s to 8.5 it/s

static tusk
#

really its almost unnoticeable when I tested

smoky oak
#

brb

dense tapir
#

I went from 20 to 32s and 40s to 56s (512x512 and 768x768)

smoky oak
#

restarting PC. trying to get docker to work

#

I want SPEED

static tusk
#

maybe 5% if that

dense tapir
#

you must have a fast card

static tusk
#

3080 10gb

dense tapir
#

1060 here sytan has a 3060

smoky oak
#

Grrrr

#

Windows updates

#

My bane

#

I have a *3060ti :p

#

The not slow one, but rather the VRAM cucked one

flat pike
#

i installed xformers and now all is much slower (1070 Laptop) i delete the --xformers entry from webui, but its slower even weith out activated xformers 😦 someone has an idea what i can do? the basic 1.5 was faster for me.

smoky oak
#

Weird

#

Never heard of that

#

UPDATE WINDOWS

#

COME ON

#

It's at 99%

dense tapir
#

I heard of that because I had that when auto first came out. I lost a smidge (like 2 secs) with 1.5

smoky oak
#

Ok, nevermind

#

This is just asking way too much

#

It needs me to install docker, but in order to install docker, I have to install Ubuntu, but in order to do that I need to install some virtualization in my PowerShell, which required me to restart. But now in order for me to utilize that virtualization, I need to turn off my computer and enable the virtual machine platform Windows feature in my BIOS. I don't see what part of this is an easy and fast installer lmao

#

Oh waitttt

#

I messed up

dense tapir
#

wtf? You have to be using wsl2 on windows to install this shit? ffs

smoky oak
#

I'm running the local install, not the docker install

smoky oak
#

I thought that was the link in order to install docker, but I clicked on the wrong hyperlink

#

Man, it's just an exe lmao

#

I was doing so much extra, and for nothing lol

dense tapir
#

see, I would love to install this on colab for training because the t4 has tensor cores due to it being a Turing based card

smoky oak
#

I am seeing if I can figure this all out

#

I am trash at code, but I am a damn fast learner, so I am following the guide

dense tapir
#

colab is ubunto but an old version

#

*unbuntu

smoky oak
#

STILL USES WSL 2

#

bruhhh

#

I'm just gonna try it

dense tapir
#

yeah, fuck that shit, and besides I can't use it locally

smoky oak
#

worse comes to worse, its just a program I have to delete

dense tapir
#

WHOA, wtf?

smoky oak
#

Yeah

#

its its own web UI

#

I thought I said that

#

looks like this

dense tapir
#

I need to contact them

#

for realz

#

Will this not work for training?

smoky oak
#

no clue

dense tapir
#

meh, apparently not

smoky oak
#

oh, ok

#

so the docker one is the more annoying one

#

interesting

dense tapir
#

this is one of the errors I get related to OOM on the vramRuntimeError: Given groups=1, weight of size [320, 320, 1, 1], expected input[1, 2, 9216, 320] to have 320 channels, but got 2 channels instead

#

I X it out and restart and can go for awhile

smoky oak
#

oh weird

dense tapir
smoky oak
#

@dense tapirMan, this is a pain in my ass

dense tapir
#

Yep

smoky oak
#

i absolutely love having no info on all of this

#

It just says to set a username and password

#

and then says nothing as to how to do it lol

dense tapir
#

I wish we could merge lohas

#

These are coming out exceptionally well

static tusk
smoky oak
#

yeah, i figured that out eventually, no thanks to literally ALL of the 6 pages of documentation I looked at

wispy nest
smoky oak
#

@static tuskPlease tell me you know how the hell I can paste something into linux

#

Cause I can't figure it out, and this is an annoying command

static tusk
#

the limited usage ive had was just ctrl v

smoky oak
#

I wish it was so simple

dense tapir
#

shift insert I think

smoky oak
#

nope, that plays an error sound

#

shift insert, control insert, right click, control v, control shift v, they all don't work

dense tapir
#

ctrl shift + v. if neither then windows is fucking around

#

then windows is fucking around and why I would never run wsj

#

wsl

static tusk
smoky oak
#

thank god

#

thank you

#

@static tuskTHANK YOU

#

right lmao

#

99% of the crap people claim to be "how AI actually works" is just BS thats perpetuated to make more people angry

#

my ex boyfriend is a digital artist. He makes all money off coms, and he was sooo mad when I said I was using SD. But now that he has seen how it works, and I explained all the rest, he constantly asks me to make reference images for his art

#

once people know how it works, they typically chill out lol

dense tapir
#

see that arm up? one of my pics has that so how do I stop that?

#

training info

smoky oak
#

@dense tapirI GOT IT

dense tapir
#

sweet

smoky oak
#

That took a longgg time lol

dense tapir
#

way too complicated. test it yet?

smoky oak
#

hahaha

#

no

#

cause they don't say how to load models into it

dense tapir
#

whale shit

#

tensorrt from nvidia is only for inference

smoky oak
#

Currently trying to figure out where ubuntu is located on my PC

#

@dense tapiroh goddddd

#

it can only use models it can directly download from hugging face

#

ugghhhhh

flat pike
#

is there an compare about gpu spped from diffrent model avaiable?

#

speed

#

especially for sd 1.5

dense tapir
#

for a valid speed comparison you need to remove 512x512 from your mind and concentrate more on 768x768

#

3 times slower about

smoky oak
#

3x slower?

#

maybe on your GPU

dense tapir
#

for me

smoky oak
#

768x768 isn't much slower for me

dense tapir
#

cause you have the ram

#

once it stuffs the vram it is fast

static tusk
dense tapir
#

my point is still valid cause 1024 is coming where even your card will cry

smoky oak
#

@dense tapirwoah, VoltaML has samplers I have never seen

dense tapir
#

yes

#

probably only for their models

smoky oak
#

PNDM

dense tapir
#

I have seen that before

smoky oak
#

DEIS Multistep

dense tapir
#

dunno that one

smoky oak
#

LMSD

dense tapir
#

LSD?

smoky oak
#

UniPCMultistep

dense tapir
#

M is the new shit

summer viper
smoky oak
#

DDPM

#

ok, gonna try DDIM...

dense tapir
#

UniPC with additional stuff

#

DDPM is old like the first one is

smoky oak
#

here comes the moment we see if this was worth it

dense tapir
#

they were discarded by most

smoky oak
#

0.9s/img

#

👁️

#

not as fast as I was hoping, actually

#

ITS

#

THE SAME SPEED AS A1111

#

BRO

flat pike
smoky oak
#

I'm getting 10.7it/s on DDIM

dense tapir
#

I just went to see current prices for a 4090 and FE 2300 USD. that is 1600 MSRP

smoky oak
#

oh dang, this actually has controlnet support lol

#

ohhhh what

summer viper
dense tapir
rocky fiber
summer viper
#

like a real

dense tapir
#

Wonder what it tastes like?

summer viper
#

heheh

dense tapir
smoky oak
#

alright, so looks like VoltaML was a complete and under failure heh...

#

its no faster for me

static tusk
#

sure all the dependencies are right version?

dense tapir
#

I bet not

smoky oak
#

there was nothing for me to interfere with

static tusk
#

slow cpu could be the cause maybe

dense tapir
#

that as well

smoky oak
#

I do have a slow CPU, let me see

dense tapir
#

like me and a 4090 I would bottleneck it all to hell

smoky oak
#

only using 15% CPU when genning

#

also

#

WTF

#

this thing is using 25GB RAM

#

excuse me

dense tapir
#

that is not how you know

summer viper
dense tapir
#

you have 16gb?

summer viper
#

4

#

32 ram

smoky oak
dense tapir
#

ahhh

#

I suspect your cpu

smoky oak
#

oh wait, it doesn't have xformers

#

its installing now

dense tapir
#

doesn't matter if 1% usage what matter is how slow it is getting the shit back and forth from the stuff

smoky oak
#

it says very slight at the bottom you need to manually install xformers

dense tapir
#

WHAT? lol

static tusk
#

yeah its single thread performance

smoky oak
#

ah yes, I love being once again limited by my CPU

#

Slow ass CPU

dense tapir
#

well, it is taking me 2-3m per gen now

#

3m for a DDIM 30, 2m for a DDIM 20

#

I am really about to toss this shit

#

even dpm++ sde karras 8 is taking a min

summer viper
dense tapir
flat pike
#

do not load is no solution, its so damn slow now

summer viper
summer viper
chilly lance
summer viper
flat pike
#

the mighty amphibian Warrior drinks a beer and stay for a while to rest a bit:

#

🙂

summer viper
#

by using "token merging" I was able to improve the performance twice

#

but for some reason the results get worse when the "--medvram" flag is enabled

jagged sluice
smoky oak
#

I love this image lol

summer viper
#

not bad

hardy tulip
#

ah yes, download more vram

summer viper
#

yeah

#

download in internet

jagged sluice
tiny thistle
#

stable diffusion online cuz i cant

hardy tulip
#

oh mah gawsh

summer viper
chilly lance
flat pike
#

ok the presentation pics definatly focus the wrong part , sorry^^

oblique nimbus
flat pike
#

XD maybe

#

but the lora is nice

tropic shell
stoic pivot
#

AI was a great idea to help humanity to be smarter, then it was released on internet and had to work with "Make a studio photo [[tiger|hamster]-[bird|human]] painting".

#

Master prompt No 346; "studio photo of a [[tiger|hamster]-[bird|human]] painting by Salvador Picasso".

#

Seekreet prompt recipe that no one will be able to guess 🙂

hardy tulip
#

where can i find this model?

#

on the anyv4 page there is no -bake version

stoic pivot
#

The seekreet... censored.

flat pike
hardy tulip
hardy tulip
#

read my message again

flat pike
#

ah ok

dry crow
flat pike
#

sory misunderstood it

hardy tulip
#

so he merged the vae to the checkpoint?

dry crow
#

It seems so yes

hardy tulip
#

ah, that clarifies

#

thanks

dry crow
#

No problem 🙂

flat pike
wispy nest
#

Create a logo on the topic Logo for bullet resistant window and door company Logo Text
BallisticShield Aluminium Windows & Doors

flat pike
#

^^there is no bot here, u have to install Stable Diffusion local or use the website

crystal deltaBOT
#
FAQ: How do I generate images? Is there a bot on the server?

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!

halcyon grail
#

Is the only way to do inpainting on Midjourney images,, is to download them into the img to img and do that?

wispy nest
vague oasis
#

mona lisa is not overtrained in SDXL like she was in SD 1.x

#

theyre blurry cause i used 20 steps btw

wispy nest
#

thinking 20 steps usually enough tho

#

for any sampler

glossy herald
#

for 512x512*

#

but yeah

#

higher size tend to require more steps to converge

wispy nest
#

like...how much more samplers need

glossy herald
#

no, I was just stating what I observed by doing a lot more higher resolution recently

wispy nest
#

I've been doing 1024x768 with 25 steps and it was just fine I think

glossy herald
#

I have had to go up to around 50-60 for something good to come up, when doing 960x540

wispy nest
#

on DPM++ SDE Karass

glossy herald
#

well, our millage seems to vary x)

wispy nest
#

What sampler was it?

glossy herald
#

UniPC

#

EulerA seems to also give me that

wispy nest
#

weird, I thought it requires less steps then anything else

wispy nest
glossy herald
wispy nest
#

lemme find the thing again...

wispy nest
#

on anime models even 20 is fine

glossy herald
#

that's what converging means usually, going to that point where any more step doesn't bring change