@Dr. Furkan Gözükara if i let it fininsh it with this 1gpu usage it will take more time but everythi
@Dr. Furkan Gözükara if i let it fininsh it with this 1gpu usage it will take more time but everything else will be same like results right /
B?
F4x gpu will reduce quality slightly
Fsince increases batch size
Areduce quality = less similarity with subject
or
reduce quality= lower resolution results / noise appearance?
or
reduce quality= lower resolution results / noise appearance?
Fless similarity
Xroot@9cb546417a6e:/workspace/SUPIR# rm -r supir
rm: cannot remove 'supir': No such file or directory
rm: cannot remove 'supir': No such file or directory
Xim getting that error
Fand type again
Frm -r SUPIR
B@Dr. Furkan Gözükara from where can i get those saved config before traing which we saved before starting the training
Fwhat do you mean
Bbefore starting the training we saved the cofig so i want to use that i got them
Bthanks fro response
Fcheck folders
Fprobably in workspace
X/launch-linux.sh --launch_mode none --host 0.0.0.0
bash: ./launch-linux.sh: No such file or directory
root@9cb546417a6e:/# cd SwarmUI
bash: cd: SwarmUI: No such file or directory
root@9cb546417a6e:/# ./dotnet-install.sh --channel 8.0
cd ..
./launch-linux.sh --launch_mode none --host 0.0.0.0
bash: ./dotnet-install.sh: No such file or directory
bash: ./launch-linux.sh: No such file or directory
root@9cb546417a6e:/# chmod +x install-linux.sh
./install-linux.sh
chmod: cannot access 'install-linux.sh': No such file or directory
bash: ./install-linux.sh: No such file or directory
root@9cb546417a6e:/#
bash: ./launch-linux.sh: No such file or directory
root@9cb546417a6e:/# cd SwarmUI
bash: cd: SwarmUI: No such file or directory
root@9cb546417a6e:/# ./dotnet-install.sh --channel 8.0
cd ..
./launch-linux.sh --launch_mode none --host 0.0.0.0
bash: ./dotnet-install.sh: No such file or directory
bash: ./launch-linux.sh: No such file or directory
root@9cb546417a6e:/# chmod +x install-linux.sh
./install-linux.sh
chmod: cannot access 'install-linux.sh': No such file or directory
bash: ./install-linux.sh: No such file or directory
root@9cb546417a6e:/#
Xgettingthis error while trying to launch swarm_ui
Fyou need to learn folder path reading
Froot@9cb546417a6e:/# cd SwarmUI
bash: cd: SwarmUI: No such file or directory
bash: cd: SwarmUI: No such file or directory
Fit is literally telling you the error
XBut i do have a file named SwarmUI in work space
Fyou are not in workspace
Froot@9cb546417a6e:/#
Xroot@9cb546417a6e:/workspace# cd SwarmUI
root@9cb546417a6e:/workspace/SwarmUI# cd launchtools
bash: cd: launchtools: No such file or directory
root@9cb546417a6e:/workspace/SwarmUI# cd launchtools
bash: cd: launchtools: No such file or directory
XI should have swarm_ui installed already i wasn't getting this error earlier
FYouTubeSECourses
In this video, I demonstrate how to install and use #SwarmUI on cloud services. If you lack a powerful GPU or wish to harness more GPU power, this video is essential. You'll learn how to install and utilize SwarmUI, one of the most powerful Generative AI interfaces, on Massed Compute, RunPod, and Kaggle (which offers free dual T4 GPU access for ...
Xhmmm im following along but i wonder what happened to my swarmui it was working well
Fif not working best is reinstall
Fmake sure to not delete your models
Xdo you have video on how i can upload my model to hugging face?
FYouTubeSECourses
Notebook updated to V5 with huge improvements. If you are having trouble to download models onto your cloud platforms such as RunPod, Google Colab, Kaggle, Massed Compute or anywhere, this tutorial is made for you. Moreover, you can upload your files / models into your private Hugging Face repository and save them permanently privately and downl...
M@Dr. Furkan Gözükara do you know why dreambooth extracted loras for flux for me. At strength 1 likeness is not good but at 1.2 it gets better, i even tried with clamp quntile at 1
Fsame here
Fflux lora extraction not very good
Fmaybe kohya implementation broken :d
Mmost likely also probably with the clip l training will be better, no minium difference option
Fyes clip l improves
Fi tested
Fhe still didnt add
Fsd 3.5 large has it
Mhopefully kohya can add it in, would be pretty nice, wonder if he could do the t5 as well
F100%
Fcan you request?
Fi did several times
FGitHub
Replace SD3Tokenizer with the original CLIP-L/G/T5 tokenizers.
Extend the max token length to 256 for T5XXL.
Refactor caching for latents.
Refactor caching for Text Encoder outputs
Extract arch...
Extend the max token length to 256 for T5XXL.
Refactor caching for latents.
Refactor caching for Text Encoder outputs
Extract arch...
XWan is showing error but no error in terminal of runpod
XThe following models are loaded: ['wan_video_image_encoder'].
Loading models from: models/models_t5_umt5-xxl-enc-bf16.pth
Loading models from: models/models_t5_umt5-xxl-enc-bf16.pth
Xfor wan