24gb 24GB_GPU_22900MB_7.3_second_it_Tier_1, maybe i dont understand, but i think i am slow few times
24gb 24GB_GPU_22900MB_7.3_second_it_Tier_1, maybe i dont understand, but i think i am slow few timess than was in example
Fyou can do this
Fsearch block swap in gradio
Fincrease number by 1
G@Furkan Gözükara SECourses i have a data set of images + mask (using sam 2) , will dreamboth train accordingly?
Fif you do properly yes should do
Four masks are same as onetrainer generates i believe
WHi @Furkan Gözükara SECourses my lora training from your config today get error, dont know why yesterday still working well
Wthis is my error:
message.txt8.47KB
WI use mass compute
Si'll try now thank you. Yesterday when was tryin in the beginning i had 112 adn then 34 it/s for few minutes and probalistic tiome for learning was 30 min bit then again speed reduced to 6
Stoday i have this:
Slora flux, config for best finetune 24gb, all is like in video images is cpoped with tools in video 1024x1024 115 images
EHi! This guy can apparently train dreambooth models on 12 gb vram, any thoughts on how that might be possible? https://www.linkedin.com/pulse/comprehensive-guide-midjourney-sref-style-training-4-valsecchi-cxi6f/?trackingId=ihrUD9hrgJh%2F1CaCHPBg%2BA%3D%3D
S@Furkan Gözükara SECourses i finished learning with 2 batch size, it is good speed or not? Thank you for anser and help
message.txt29.48KB
Fwe can dreambooth on 6 gb vram
FYouTubeSECourses
If you want to train FLUX with maximum possible quality, this is the tutorial looking for. In this comprehensive tutorial, you will learn how to install Kohya GUI and use it to fully Fine-Tune / DreamBooth FLUX model. After that how to use SwarmUI to compare generated checkpoints / models and find the very best one to generate most amazing image...
Fgreat speed
Fbecause batch size 2 = 2x speed
ESo confused, thought you needed 80 vram
Fit is very old info
Fnow block swap technique is well developed and established
Fall you need is 64 GB RAM + 6 GB VRAM
EWoops okey, does it work on kohya v46?
Fyes
Flatest zip file on patreon are up to date
SThank you! Results of generation are amazing even without yolo for now, thank you very much for guide, tomorrow will try with yolo. This is truly the best guide 
FThanks a lot. I recently researched a better workflow but this performed best 
NSounds good
SWith yolo incredible results very good! I jusst don't understand how you gain approximately spee 1.3 s/it on 3090, i saw post on X.com
Fi can implement flash attention - may help. also teacache
Fbut both may reduce quality
Fbut it is already that speed on my rtx 3090 ti
F
Sthank you i will try tomorrow
SI see movement in GPU and RAM usage
S
Fwhich config?
Fyour config wrong
F8-bit AdamW optimizer
Fplease follow tutorial dont skipo
FYouTubeSECourses
If you want to train FLUX with maximum possible quality, this is the tutorial looking for. In this comprehensive tutorial, you will learn how to install Kohya GUI and use it to fully Fine-Tune / DreamBooth FLUX model. After that how to use SwarmUI to compare generated checkpoints / models and find the very best one to generate most amazing image...
SOkay will ReDo
SHello @Furkan Gözükara SECourses , after all tries i have few questions:
- if i use 15 batch size on a100 i need use approximately reduce x2 epochs? I have 137 photo, i think need near 100 epochs with 7 batch size, if batch size 15 can try approx 60 epochs?
- if my photos from iphone them size approx 1 mb, yours photo are 5mb is it very important in this case, i dont see very big difference in learning (using 137 preprocessed photos) avr_loss near 0.4 always +-
Now i'm doing upscaling with SUPIR, and will try after it will finish. - tell please can i fix freezes with scrolling screen with ThinLink macos m4 max?
S
Fepoch doesnt change according to batch size
Fepoch is related to number of image you have usually
Fmb of photos doesnt matter what matter is resolution like 1024x1024 - best
Fif you dont downscale it will auto
FFawesome
