well i think for training a person no captions better :d
well i think for training a person no captions better :d
Mdepends, yours are usually overtrained 
Xokay i made the unet file inside of models then restarted/updated but still getting same error saying "All available backends failed to load the model"
Fi had this error once
Fso annoying
Fcan you move file into diffusion_models
Ffolder and restart?
F
Xjust did this should i restart pc also?
Xgetting the same error with the flux dev file into into diffusion_models
Xthis is so crazy!
Fmore detailed posted here https://www.reddit.com/r/FluxAI/comments/1fc23v5/i_have_compared_captions_generated_by_internvl28b/
Reddit
Explore this post and more from the FluxAI community
Fcan you post this in here super important
FFi will ping the dev once you post there with details
Fhappened to me as well
Fshow screenshot of folders
Ffiles too
Mwhat is this suppose to mean? did you use a lora where you didn't train it with an instance prompt?
Ffull images : https://www.reddit.com/r/FluxAI/comments/1fc27e0/i_have_compared_captions_generated_by_internvl28b/
Reddit
Explore this post and more from the FluxAI community
Fit was trained with ohwx man only
F150 epoch - 2250 steps
Mand for the two images next to it did you use ohwx man?
Mor are you just comparing prompts?
Fno i used the prompts
Fgenerated without lora :d
Fjust comparing prompts
Mthanks for clarifying, cause i didn't understand at all
Xjust posted details @Dr. Furkan Gözükara
Fadded a clarification comment to post : https://www.reddit.com/r/FluxAI/comments/1fc27e0/i_have_compared_captions_generated_by_internvl28b/
Reddit
Explore this post and more from the FluxAI community
Fgonna ping now
Freply to guy who replied you
D@Dr. Furkan Gözükara
why does it show 52/100 epoch? if I put only 25, what did I do wrong?
why does it show 52/100 epoch? if I put only 25, what did I do wrong?
F4 GPUs = 4x25
Dso I'm going to have 100 versions of loras?
FHow many epoch you save
DSave Every N Epoch = 1,
i thought it would be only 25 versions. but now it will be 100.... Then I should keep the best one, right? the one with the best result. That will be the one I will use for my work.
i thought it would be only 25 versions. but now it will be 100.... Then I should keep the best one, right? the one with the best result. That will be the one I will use for my work.
Mwhat are you trainign for 10k steps?
�Is there a kohya discord server?
D800 imgs
Mthere is an imposter amongnst us
DWhy do you say that?
FSadly nope
FYes it will save 100
Sfor my case
I had to remove " --fast " option. i have, desktop and laptop both 4090 , and same problem.
I had to remove " --fast " option. i have, desktop and laptop both 4090 , and same problem.
Xi tried this and still getting same issue
Xdid you have unet file inside of models?
AFile "C:\Users\chris\AppData\Local\Programs\Python\Python310\lib\runpy.py", line 196, in _run_module_as_main
return _run_code(code, main_globals, None,
File "C:\Users\chris\AppData\Local\Programs\Python\Python310\lib\runpy.py", line 86, in _run_code
exec(code, run_globals)
File "D:\Kohya_GUI_Flux_Installer_30\kohya_ss\venv\Scripts\accelerate.EXE__main__.py", line 7, in <module>
File "D:\Kohya_GUI_Flux_Installer_30\kohya_ss\venv\lib\site-packages\accelerate\commands\accelerate_cli.py", line 48, in main
args.func(args)
File "D:\Kohya_GUI_Flux_Installer_30\kohya_ss\venv\lib\site-packages\accelerate\commands\launch.py", line 1106, in launch_command
simple_launcher(args)
File "D:\Kohya_GUI_Flux_Installer_30\kohya_ss\venv\lib\site-packages\accelerate\commands\launch.py", line 704, in simple_launcher
raise subprocess.CalledProcessError(returncode=process.returncode, cmd=cmd)
subprocess.CalledProcessError: Command '['D:\Kohya_GUI_Flux_Installer_30\kohya_ss\venv\Scripts\python.exe', 'D:/Kohya_GUI_Flux_Installer_30/kohya_ss/sd-scripts/flux_train_network.py', '--config_file', '/home/Ubuntu/apps/StableSwarmUI/Models/Lora/config_lora-20240908-123856.toml']' returned non-zero exit status 3221225477.
12:40:57-496469 INFO Training has ended.
return _run_code(code, main_globals, None,
File "C:\Users\chris\AppData\Local\Programs\Python\Python310\lib\runpy.py", line 86, in _run_code
exec(code, run_globals)
File "D:\Kohya_GUI_Flux_Installer_30\kohya_ss\venv\Scripts\accelerate.EXE__main__.py", line 7, in <module>
File "D:\Kohya_GUI_Flux_Installer_30\kohya_ss\venv\lib\site-packages\accelerate\commands\accelerate_cli.py", line 48, in main
args.func(args)
File "D:\Kohya_GUI_Flux_Installer_30\kohya_ss\venv\lib\site-packages\accelerate\commands\launch.py", line 1106, in launch_command
simple_launcher(args)
File "D:\Kohya_GUI_Flux_Installer_30\kohya_ss\venv\lib\site-packages\accelerate\commands\launch.py", line 704, in simple_launcher
raise subprocess.CalledProcessError(returncode=process.returncode, cmd=cmd)
subprocess.CalledProcessError: Command '['D:\Kohya_GUI_Flux_Installer_30\kohya_ss\venv\Scripts\python.exe', 'D:/Kohya_GUI_Flux_Installer_30/kohya_ss/sd-scripts/flux_train_network.py', '--config_file', '/home/Ubuntu/apps/StableSwarmUI/Models/Lora/config_lora-20240908-123856.toml']' returned non-zero exit status 3221225477.
12:40:57-496469 INFO Training has ended.
Agot the above when I started my training
JWhen T5 / clip-l testing commencing? Need info on what to use, LR to use, etc
Mt5 very sensetive, kohya is implementing different learning rates soon for each