Software Engineering Courses (SECourses)•3y ago

Ah thank you very much for the insight. Will do another fresh install. But wouldn't xformers be dep

Ah thank you very much for the insight. Will do another fresh install.

But wouldn't xformers be dependent on what GPU you use.

I use an RTX 3060 with 12GB

CCPI_MrRagingBull 🇬🇧Ah thank you very much for the insight. Will do another fresh install. But woul...

I A•3/17/23, 5:47 PM

I honestly am not an expert on this, but I read somewhere that you don't need to add --xformers in your webui-user file, as of today. I will find the link and share it later, so that Dr. Furkan can give us a proper technical answer. So far I am using A1111 without adding anything to the webui, and it is handling more ControlNet tabs than before!

II A I honestly am not an expert on this, but I read somewhere that you don't need to...

CPI_MrRagingBull 🇬🇧OP•3/17/23, 5:48 PM

Good to know. Will definitely check this out later. Appreciate the messages

CCPI_MrRagingBull 🇬🇧@Furkan Gözükara SECourses I had to reinstall auto1111 and followed the steps on...

FORTUNE VARO•3/17/23, 6:25 PM

I have the same error

II A I don't think that is necessary anymore, I think even --xformers are not needed....

CPI_MrRagingBull 🇬🇧OP•3/17/23, 6:26 PM

Read these messages. @FORTUNE VARO

Zono50•3/17/23, 6:36 PM

should i be using anaconda for stable diffusion? seems everytime i run automatic1111 i'm using like 5 gigs of vram before i even start training

Zono50•3/17/23, 6:41 PM

i love making hoes on my computer bro

Waldo Faldo•3/17/23, 8:12 PM

Hi again, I've configured the model and generated 300 images using generated class images button, but when I press train button it start to create more than 8000 class images. I don't know why it generate them again, and why too many.

Waldo Faldo•3/17/23, 8:12 PM

re-processing images: waldo-imagenes: : 0it [00:00, ?it/s] We need a total of 7500 class images.
Pre-processing images: waldo-imagenes: : 0it [00:00, ?it/s]
Generating 7500 class images for training...

easyaskovic•3/17/23, 8:36 PM

Greetings everyone!

I followed your tutorials and was able to successfully run Dreambooth training on RunPod through webui. However, I am unable to view the training progress in the terminal window as shown in your videos. I have tried clicking on all available terminals under the terminals tab in Jupyter, but I couldn't find it anywhere. Could you please guide me on how to access the training progress in the terminal? Many thanks

WWaldo Faldo re-processing images: waldo-imagenes: : 0it [00:00, ?it/s] We...

Waldo Faldo•3/17/23, 8:38 PM

I fix sit, photo folder problem.

FORTUNE VARO•3/17/23, 8:57 PM

Hi! trying to install Dreambooth again, yesterday I followed this tutorial and ended messing everything at a point that auto 1111 didn´t open, I format everything and reinstall everything now I have webUI auto1111 running fine again, I´m not a programmer so everything is a little confusing to me. Now, with the fresh start I try to install DB again and have some errors again: Checking Dreambooth requirements...
Failed to install Dreambooth requirements.
[+] torch version 1.13.1+cu117 installed.
[+] torchvision version 0.14.1+cu117 installed.
[!] xformers NOT installed.
[!] accelerate version 0.12.0 installed.
[+] bitsandbytes version 0.35.4 installed.
[!] diffusers NOT installed.
[!] transformers version 4.25.1 installed.
Launch errors detected:
xformers not installed.
Incorrect version of accelerate installed.
diffusers not installed.
Incorrect version of transformers installed.

FORTUNE VARO•3/17/23, 8:58 PM

This time I want to be more careful messing and downgrading the code. Please help!

FORTUNE VARO•3/17/23, 9:02 PM

I´m trying to follow this instructions https://github.com/d8ahazard/sd_dreambooth_extension#if-you-are-having-issues-with-requirements-after-installing-look-below
But not sure how to "specify the command line argument:" No idea what that is

. I´m a digital artist not a programmer

GitHub

GitHub - d8ahazard/sd_dreambooth_extension

Contribute to d8ahazard/sd_dreambooth_extension development by creating an account on GitHub.

II A I honestly am not an expert on this, but I read somewhere that you don't need to...

CPI_MrRagingBull 🇬🇧OP•3/17/23, 9:23 PM

I dont know what it is but i am not managing to get textual inversions to work after the reinstall.

CPI_MrRagingBull 🇬🇧OP•3/17/23, 9:23 PM

I am using the same settings as i used to but its not working

EEssaBalls This graph from dreambooth indicating losses is reliable?

Furkan Gözükara SECourses•3/17/23, 9:38 PM

very low loss. unfortunately loss doesnt mean much but lower loss supposed to be better

Sstavsap are you guts changing the learning rates in textural inversion or using constant...

Furkan Gözükara SECourses•3/17/23, 9:38 PM

i use constant

Nn8hawk_hi there. I am asking myself (and you here) what would be the best way to chieve...

Furkan Gözükara SECourses•3/17/23, 9:38 PM

my upcoming video is about this. teach yourself and a style to the base model. that is a very good way to achieve this

WWaldo Faldo I got these Errors

Furkan Gözükara SECourses•3/17/23, 9:39 PM

i recorded a new dreambooth extension install video. currently processing it. hopefully will be published today. now it supports torch version 2 as well. i will also show that

CCPI_MrRagingBull 🇬🇧@Furkan Gözükara SECourses I had to reinstall auto1111 and followed the steps on...

Furkan Gözükara SECourses•3/17/23, 9:40 PM

today i am explaining how to install latest version. will be published in few hours hopefully post processing