not sure for flux but what about increasing "Gradient accumulate steps" ? probably at least - 2, you
not sure for flux but what about increasing "Gradient accumulate steps" ? probably at least - 2, you think would help?
Fno i dont suggest
Fbatch size bigger = lower quality
Fbest is batch size 2 max
Fand multi gpu
Falso you need to change LR
Fthis is so slow it should get better
Fon RTX A6000 on massed compute fast config of rank 1 gets like 4.5 second it - 5 second it on multi gpu
Yaight let me check
Yimma make the LR more aggressive
Y0.00005 atm
Fye with multi gpu and batch size you need to increase
Fi explained logic in coming video :d
Yok thank u!
Gbrother whats the proper structure to do 1000 images with an accompanying description in text file form. I think I want to do another lora of 1000 images with descriptions this time to see if it trains better and to see if those trained words allow me to generate the things in the descriptions
GI dont need help making the description, I will make a python script for that since I already have metadata for each file
Gjust want to know the proper structure and how to make sure Kohya knows the description is there or is it automatic if the txt file has the same name as the image?
Ffilename.jpg + filename.txt
Fin the same folder
Fselect captions extension txt
Fthats it
Fit doesnt have to be jpg can be png too 
Gthats it? where is that captions extention txt?
Gpng is what I do
Gthere is a switch or its automatic?
Fthat is it
Fmy configs are by default .txt
Fdo ctrl f and type .txt to see
G
Fexample style dataset published
Ftrue
Gso how does it do it with no captions now? I do see when I do dataset preparations it does add 2 additional files per image however, in the image folder. so as long as the description matches the filename we are good. also do i put the instance prompt and class prompt at the beginning of each description. I will set it up now
Gthanks
Freads folder name
Fyou really should watch video
Fi tell all of these and more
FYouTubeSECourses
Ultimate Kohya GUI FLUX LoRA training tutorial. This tutorial is product of non-stop 9 days research and training. I have trained over 73 FLUX LoRA models and analyzed all to prepare this tutorial video. The research still going on and hopefully the results will be significantly improved and latest configs and findings will be shared. Please wat...
GI thought u did without captions on these newer lora trainings
Fyes i didnt use
Fit reduces likeliness of a person
Fcurrently for style i am testing
MI have figured out regularzation images
Mstill testing, but quite close
Fearly checkpoints arrived
Flooks like LR is too low. 125 epochs and not learnt
Fmaybe i am wrong looks like captioned one works
Fwith lora
F
Fleft org - right lora
