so for this to work, we need to update all ai apps or just this triton, i will wait until its clear
so for this to work, we need to update all ai apps or just this triton, i will wait until its clear what exsactly to do.
Fyou need to just install it into venvs
Fof your apps
Fi will include it to my installers
Fslowly
Fhopefully i will make a triton tutorial today
Ftime to update some installers
DI Installed the "Kohya_GUI_Flux_Installer_v40" script which works perfectly for flux.
However I also want to train an sdxl model.
When trying to do so I get the error "ModuleNotFoundError: No module named 'xformers'".
Is pip install xformers the correct way?
However I also want to train an sdxl model.
When trying to do so I get the error "ModuleNotFoundError: No module named 'xformers'".
Is pip install xformers the correct way?
Ndreambooth fine-tuning should i choose float? i remember lora is float.
Fcurrently xformers not supported on torch 2.5
Fso make a seperate install and dont run torch 2.5
Ffloat is best
Fbut fp16 also works decent
ZNo matter what Rank config I use, I run out of VRAM 
Lora Training/16GB Vram, using Rank 5
Idle 0.5/16GB used
Are there any updates in branch sd3-flux.1 that are not accounted for in the .json files? Settings wise? I must be missing something
Lora Training/16GB Vram, using Rank 5Idle 0.5/16GB used
Are there any updates in branch sd3-flux.1 that are not accounted for in the .json files? Settings wise? I must be missing something
Nthanks
Ferror is easy
Fyou are loading into dreambooth
For at least once loaded
Fget a fresh config from zip
Fload into lora
ZI don't think so, it's flux_network script
Z
Z
Fif you loaded once it becomes corrupted
Zok, let me start fresh 
Nim trying to train a gaming setup lora (use dreambooth extract) i have 500+ images should i decrease images?
Fwell it depends on purpose
Fif you read latest work 256 images worked very well
Ffor style training more better usually
ZSame result with fresh .json... and it's way over
Fplease message me your json file
Ffrom private message
Falso show your vram before starting training
F@Zet your config is dreambooth corrupted config
Fwhy you dont follow what i say
DFresh install of Kohya_GUI_Flux_Installer_v40.zip and Kohya_FLUX_DreamBooth_v7.zip both yield:
ERROR: Cannot install -r requirements.txt (line 1), -r requirements.txt (line 10), diffusers[torch]==0.25.0 and huggingface-hub==0.24.5 because these package versions have conflicting dependencies.
ERROR: Cannot install -r requirements.txt (line 1), -r requirements.txt (line 10), diffusers[torch]==0.25.0 and huggingface-hub==0.24.5 because these package versions have conflicting dependencies.
Zchange requirements.txt huggingface-hub==0.24.5 to huggingface-hub==0.25.2
ZI am, or at least I am trying my very best
Fthis looks like bmaltasi error
Fwe need to report this
ZYes, some of the newer packages require huggingface-hub>=0.25. And careful, 0.26 removes cached_download
Zthat will need attention as well
Fi am making a fresh install to test
D@Dr. Furkan Gözükara New test Flux-De-Distilled with this settings the lora at 160 epochs looks severely undertrained, may be Flux-De-Distilled requieres higher LR
DI test it on the de-distilled model and on regular Flux-Dev
DI'm going to resume the training with LR 0.0001
Dand see if it converges
Ftotally could be
Fi will research it later hopefully
FGitHub
SD3 flux branch is broken when doing fresh install @kohya-ss @bmaltais Collecting wcwidth>=0.2.5 Using cached wcwidth-0.2.13-py2.py3-none-any.whl (34 kB) INFO: pip is looking at multiple version...
DI gonna resume from 160 epochs at 0.001 for 10 epochs and test if get signs of converging, I will report