Software Engineering Courses (SECourses)•3y ago

what are the two packages we need to upgrade in stable diffusion for faster processing? can we not j

what are the two packages we need to upgrade in stable diffusion for faster processing? can we not just pip install --upgrade then the package names?

Zono50OP•3/21/23, 10:20 PM

transformers and can't remember the other one

ZZono50 what are the two packages we need to upgrade in stable diffusion for faster proc...

Furkan Gözükara SECourses•3/21/23, 10:22 PM

well you need pre compiled wheel

Furkan Gözükara SECourses•3/21/23, 10:23 PM

so you cant just execute install command not yet

Zono50OP•3/22/23, 1:20 AM

is there any prompt to activate face restoration?

FORTUNE VARO•3/22/23, 2:56 AM

Hi! I´m learning how to train models following this tutorials. I´m training my first one and in the preview output there are 4 images the first one is super random, nightmarish, the second one looks like the person I´m trying to model and then there are two graphics that I don´t understand and don´t know how to real:

FORTUNE VARO•3/22/23, 2:56 AM

https://ibb.co/YfGg59C

ImgBB

Screenshot-2023-03-21-205056 hosted at ImgBB

Image Screenshot-2023-03-21-205056 hosted in ImgBB

FORTUNE VARO•3/22/23, 2:56 AM

Is this common?

Stevens•3/22/23, 3:08 AM

Hi I am having all sorts of problems so i decided to try and make an embed because it seemed easier but they are all coming out like they were painted by a madman. Is this normal?

FORTUNE VARO•3/22/23, 3:10 AM

I read somewhere that the latest Dreambooth is just broken, no matter what you do, what parameters you adjust the results would be nightmarish

FORTUNE VARO•3/22/23, 3:11 AM

It´s so frustrating! Spent days trying for a model to work and to understand the app but the results are always garbage after hours of waiting

SStevens Hi I am having all sorts of problems so i decided to try and make an embed beca...

FORTUNE VARO•3/22/23, 3:12 AM

What´s an embed?

FFORTUNE VARO What´s an embed?

Stevens•3/22/23, 3:37 AM

An embed is a way of training that, according to my research, allows you to make a model based on one of the models (say, 1.5) and then you can use that face in any model that is based on 1.5.

So it would allow one to put the character they trained into a lot of things. Unless they look like claymation horrors from hell's basement.

FORTUNE VARO•3/22/23, 3:37 AM

Ho yes! That´s what I´m doing!

FORTUNE VARO•3/22/23, 3:38 AM

I´m starting to get interesting results FINALLY!!!

Stevens•3/22/23, 3:41 AM

does "interesting" mean melted abomination? Cause if so, me too!

Stevens•3/22/23, 3:42 AM

The Doc has a video on training a LORA but he never mentions an embed so maybe it's trash?

FORTUNE VARO•3/22/23, 3:43 AM

I think that "Save your Model Frequency" Is a very important value, I did just that, the memory usage goes crazy because at the end of the training you have lots of models. I assumed as a noob that the more the code works on a model the better it gets but that´s not the case, I think its called overtraining so... I ended with 10 versions of a model and apparently the third one is working very good, the fourth one gives bad results the fifth one very bad results and so on...

SStevens The Doc has a video on training a LORA but he never mentions an embed so maybe i...

FORTUNE VARO•3/22/23, 3:44 AM

No! realistic results!

Stevens•3/22/23, 3:45 AM

pfft, if that's what you're going for, I suppose. But it's blurry nonsense for me! LET'S GO!

SStevens The Doc has a video on training a LORA but he never mentions an embed so maybe i...

FORTUNE VARO•3/22/23, 3:45 AM

Ho! no, I he mentions that LORA is more difficult so I´m starting with embed 1.5

FORTUNE VARO•3/22/23, 3:46 AM

Stevens•3/22/23, 3:47 AM

your pic looks more like my subject than mine does

Stevens•3/22/23, 3:47 AM

that is dope

FORTUNE VARO•3/22/23, 3:47 AM

PROMT: painting of p3p34gu1L4r by rembrandt, high resolution. 8 k 3 d render trending on art station 4k intricate artwork masterpiece digital painting concept illustration greg rutkowski beeple alphonse mucha cinematic lighting atmospheric light glow octane unreal engine photorealistic hyperrealism photography complex detailed matte print fantasy style atmosphere surreal portrait canon eos c 300 ƒ 1 5 0 mm f 2 - 6 sony alpha n

SStevens your pic looks more like my subject than mine does

FORTUNE VARO•3/22/23, 3:47 AM

hahaha

FORTUNE VARO•3/22/23, 3:47 AM

Try this:
I think that "Save your Model Frequency" Is a very important value, I did just that, the memory usage goes crazy because at the end of the training you have lots of models. I assumed as a noob that the more the code works on a model the better it gets but that´s not the case, I think its called overtraining so... I ended with 10 versions of a model and apparently the third one is working very good, the fourth one gives bad results the fifth one very bad results and so on...

Stevens•3/22/23, 3:48 AM

I am saving every 50 steps so i have like a million, each worse than the next!

Stevens•3/22/23, 3:48 AM

FORTUNE VARO•3/22/23, 3:48 AM

damn!

FORTUNE VARO•3/22/23, 3:50 AM

I know how frustrating that is, believe me... I spent 4 days watching nightmarish distorted monsters until now

FORTUNE VARO•3/22/23, 3:50 AM

How many photos are you using?

Stevens•3/22/23, 3:52 AM

Stevens•3/22/23, 3:53 AM

19?

FORTUNE VARO•3/22/23, 3:54 AM

thats a good number I think

Stevens•3/22/23, 3:54 AM

I am sure my settings are wrong

Stevens•3/22/23, 3:54 AM

sooo many buttons

FORTUNE VARO•3/22/23, 3:55 AM

Remember that the photos should be in 512X512 and they need to be very different from each other, different angles, lenses, backgrounds, clothes etc

SStevens sooo many buttons

FORTUNE VARO•3/22/23, 3:57 AM

Yes! Don´t lose hope!

FORTUNE VARO•3/22/23, 3:59 AM

I have follow this tutorial https://www.youtube.com/watch?v=Bdl-jWR3Ukc&t=3153s 4 times in the last 3 days, from beginning to end, every time a notice something new, so much to learn!

YouTubeSECourses - Software Engineering Courses

Zero To Hero Stable Diffusion DreamBooth Tutorial By Using Automati...

Our Discord : https://discord.gg/HbqgGaZVmr. The most advanced tutorial of Stable Diffusion Dreambooth Training. If I have been of assistance to you and you would like to show your support for my work, please consider becoming a patron on

https://www.patreon.com/SECourses

Playlist of Stable Diffusion Tutorials, Automatic1111 and Google Colab ...

Sιɳɠυʅαɾ Sιɳɠυʅαɾιƚყ•3/22/23, 5:03 AM

is anyone experienced with lora training here?

Fabricatedgirls•3/22/23, 5:43 AM

Is there some bug in A1111 dreambooth training where the checkpoints it generates just ignore all the training? it generates samples fine but then try to use the checkpoint to generate images and the model you generate works but just ignores all the training instance tokens completely

Fabricatedgirls•3/22/23, 5:44 AM

or resuming training is broken or something, not sure wth is going on

FFORTUNE VARO It´s so frustrating! Spent days trying for a model to work and to understand the...

FORTUNE VARO•3/22/23, 5:45 AM

yes

FORTUNE VARO•3/22/23, 5:51 AM

I read somewhere that the latest Dreambooth is just broken, no matter what you do, what parameters you adjust the results would be nightmarish
It´s so frustrating! Spent days trying for a model to work and to understand the app but the results are always garbage after hours of waiting

FFORTUNE VARO I read somewhere that the latest Dreambooth is just broken, no matter what you d...

Fabricatedgirls•3/22/23, 5:51 AM

how do you mean garbage?

FORTUNE VARO•3/22/23, 5:52 AM

distorted nightmarish

Fabricatedgirls•3/22/23, 5:53 AM

thats not exactly the issue i'm facing, it is training, its generating good samples, but when i try to use the checkpoints in SD to generate images, it ignores the training and just treats it like its the underlying model with no training at all.

Fabricatedgirls•3/22/23, 5:53 AM

its not doing a crappy warped version of what i ask it to do, its just ignoring the training altogether

EssaBalls•3/22/23, 7:41 AM

I got error like this while training hypernetwork:
"A tensor with all NaNs was produced in Unet. This could be either because there's not enough precision to represent the picture, or because your video card does not support half type. Try setting the "Upcast cross attention layer to float32" option in Settings > Stable Diffusion or using the --no-half commandline argument to fix this. Use --disable-nan-check commandline argument to disable this check."
I tried to add --no-half or set the "Upcast cross attention layer to float32" but it wont help. Wheni used --disable-nan-check then error disappeard BUT my samples are all black.

FFORTUNE VARO Hi! I´m learning how to train models following this tutorials. I´m training my f...

Furkan Gözükara SECourses•3/22/23, 7:48 AM

hi welcome

what are the two packages we need to upgrade in stable diffusion for faster processing? can we not j

Similar Threads

Similar Threads

Similar Threads