Nothing. Could be the python version? witch version are you using, me . 3.10.6
Nothing. Could be the python version? witch version are you using, me . 3.10.6
NI rember adding that line to the webuiuser.bat :
set "REQS_FILE=.\extensions\sd_dreambooth_extension\requirements.txt"
set "REQS_FILE=.\extensions\sd_dreambooth_extension\requirements.txt"
NGitHub
Contribute to d8ahazard/sd_dreambooth_extension development by creating an account on GitHub.
NGitHub
Contribute to d8ahazard/sd_dreambooth_extension development by creating an account on GitHub.
Nthere...
Nyour python version should be ok.
WOk, let me check thanks....
N"After installing via the WebUI, it is recommended to set the above flags and re-launch the entire Stable-diffusion-webui, not just reload it."
WThis means presing the webui-server.bat to re-launch all right?
Nafter closing the old cmd-window.
the name was imho webui-user.bat
the name was imho webui-user.bat
WYes sorry
WThe command is already there: set "REQS_FILE=.\extensions\sd_dreambooth_extension\requirements.txt". Nothing, same problem. Is there any alternative extension I can use? Is Dreambooth the most popular?
W"We also need a newer version of diffusers, as SD-WebUI uses version 0.3.0, while DB training requires >= 0.10.0. Not having the right diffusers version is the cause of the 'UNet2DConditionModel' object has no attribute ' enable_gradient_checkpointing' error message, as well as safety checker warnings"
WMay be this is the problem
Nmy version of diffusers is 0.10.2
C@Furkan Gözükara SECourses I had to reinstall auto1111 and followed the steps on downgrading cuda. But when i downgrade and start webui i get this error
Ci get same error when i upgrade to cuda 117 which you mention in here https://gist.github.com/FurkanGozukara/e2db853d2016a4a9ae2cc32dc41d730a
WHow can It check the Diffusers and Updated?
Npip show diffusers
WI still are not familiarised with not WebUI environment to know where I have to put that command sorry. This is my first day
NI made $23,000! I am better!
Njust type that in the cmd-window
Wnot found
Wprompt have to be in the stable diffusion folder?
WAnyway it is not found, strange A1111 webUi doesn't install it automatically
N\stable-diffusion-webui\venv\Scripts
W:\A1111\stable-diffusion-webui\venv\Scripts>pip show diffusers
WARNING: Ignoring invalid distribution -rotobuf (c:\a1111\stable-diffusion-webui\venv\lib\site-packages)
WARNING: Package(s) not found: diffusers
WARNING: Ignoring invalid distribution -rotobuf (c:\a1111\stable-diffusion-webui\venv\lib\site-packages)
WARNING: Package(s) not found: diffusers
Wok I could install them using "pip install --upgrade diffusers[torch]". Last version is 0.14.0
WI've reinstalled the Dreambooth, no error right know but the Dream tap is no there, arrggg
Ndid you restart the webui-user.bat
Wyes
Nclosed the browser window?
WI using chome and edge, nothing
IYes, 1.5, but also others. Same results. I spent all night trying different settings. In my case changing bf16 to fp16 fixed it. The wizard may suggest one option, but you have to test it first to see if it actually works, before you do the whole training.
IYou are getting much better results than I got, even with more steps, and so on. Can you resume your the settings that work for you? I know they can't be replicated, but it helps us understand how training works.
II don't think that is necessary anymore, I think even --xformers are not needed. Major changes happened last night while I was working, so I did a fresh reinstall of A1111 too, and it worked much much better without any downgrade, without --xformers. The only problem is the damn Dreambooth, so we have to wait for Dr. Furkan's new tutorial for that.
CAh 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
But wouldn't xformers be dependent on what GPU you use.
I use an RTX 3060 with 12GB
II 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!
CGood to know. Will definitely check this out later. Appreciate the messages
FI have the same error
CRead these messages. @FORTUNE VARO
Zshould 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
Zi love making hoes on my computer bro
WHi 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.
Wre-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...
Pre-processing images: waldo-imagenes: : 0it [00:00, ?it/s]
Generating 7500 class images for training...
EGreetings 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
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 thanksWI fix sit, photo folder problem. 
FHi! 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.
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.
FThis time I want to be more careful messing and downgrading the code. Please help!
FI´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
But not sure how to "specify the command line argument:" No idea what that is
. I´m a digital artist not a programmerGitHub
Contribute to d8ahazard/sd_dreambooth_extension development by creating an account on GitHub.
CI dont know what it is but i am not managing to get textual inversions to work after the reinstall.
