So the higher the batch size, the lower amount of steps you’ll need to achieve a quality result?
So the higher the batch size, the lower amount of steps you’ll need to achieve a quality result?
HLike if I rented an H200 and set the batch count to 14, I would likely need fewer than 4K steps to get the same quality as batch size 7?
Fbatch size meaning is at 1 step it process more images
Fso we are still processing like equal number of samples (the model trained number of images)
Fbut the model weights are updated lesser
Fsince batch size changes averaged
HGotcha, thank you for clarifying!
Yhey just joined to try get better settings for kohya finetune of flux and see discord pop up so thought I'd ask here. 48gb RTX 6000 Ada should be enough to finetune right? going mad here keeps getting CUDA out of memory even with 512x512, batch 1, fp16, adam8, gradient_checkpointing and everything i can find to reduce memory, and feel i must be doing something wrong 
its really odd
its really oddYnm i used one of the scripts from the patreon for 48gb and it works - god only knows which setting it is 
Thanks! much worth the patreon sub!
Thanks! much worth the patreon sub!
R@Furkan Gözükara SECourses The onetrainer sdxl config files load into fine tune, not lora. is this the intent?
Fyep i dont have lora config sadly
PHi, I've been trying to merge 2 Flux LoRas via Kohya SS together and I keep failing.
Say, if normally I set these 2 Loras to the strenght of 1 and 0.6, what ratios should I put in there? And then, sometimes it merges a lora to the size of 24Gb or 40Gb+ if I switch the save precision so I'm not really sure what settings in general there would be the best
Say, if normally I set these 2 Loras to the strenght of 1 and 0.6, what ratios should I put in there? And then, sometimes it merges a lora to the size of 24Gb or 40Gb+ if I switch the save precision so I'm not really sure what settings in general there would be the best
CIs ohwx an especially good instance prompt or would using any another string of rare special characters would be just as good?
Fohwx good
Frare token
RJust got around to dreambooth videos and I gotta say this is amazing we can do this. You are a god send my friend...Keep up the great work!
AIs it a bad idea to use regularisation dataset with a cartoon dataset for fine tuning?
Fit depends. for flux we dont use. for sdxl if your reg images are also cartoon good
LI admire your work. As a Patreon member and Massed Compute user ($100+ spent), I'm experiencing an issue with my Dreambooth training.
My training with 277 properly prepared images, using background filling, is progressing using the "Fine Tuning / DreamBooth: 256 Training Images & Batch Size is 1 : Best Epoch is 70" setting. This should finish within the hour on Massed Compute, totaling 277 x 70 steps. I have checkpoint intervals every 10 epochs and have six models so far.
The first checkpoint showed excellent results, but subsequent epochs have produced increasingly blurry images.
Would it be possible to discuss this with you, perhaps via Skype? I would appreciate any guidance you can offer, either freely or through a specialized support option.
My training with 277 properly prepared images, using background filling, is progressing using the "Fine Tuning / DreamBooth: 256 Training Images & Batch Size is 1 : Best Epoch is 70" setting. This should finish within the hour on Massed Compute, totaling 277 x 70 steps. I have checkpoint intervals every 10 epochs and have six models so far.
The first checkpoint showed excellent results, but subsequent epochs have produced increasingly blurry images.
Would it be possible to discuss this with you, perhaps via Skype? I would appreciate any guidance you can offer, either freely or through a specialized support option.
Fthank you so much and welcome. replied you from email
�can you explain this, why are the eyes so bad? did sdxl lora training, 30 train pics, all great quality, many close ups on face, 150 epochs, repeat 1, 1024 resolution, biglovexl2 as base model.
�training on civitai
�training-data, i see network rank 32. don't know if its possible to change that on training
train-data.txt1.41KB
Fare you using our config_?
Fwhich config name?
Falso some people reporting bad results
Fi am gonna train now to see
�this was civitai training, I used what you said about ohwx woman token, 150 epochs, repeat 1.
�where can we download flux dev fast now? black labs has it gated, so can't use the download link in script...
Fah i dont know
Fi trained on my local pc yesterday
Four config
Flatest kohya
Fperfect results
�Oh awesome, you recommend kohya for sdxl lora and checkpoint training now?
Fi tested flux
Fbut we have sdxl dreambooth config too
�Ok, so onetrainer gives slightly better results for sdxlin your opinion?
AHey Doc, how you doing today, I was just wondering if it’s possible to create a stylised flux model in place of a stylised lora
LI am pleased to inform you that the dreambooth training has produced excellent results. It seems the previous difficulty arose from an incorrect learning rate.
I have encountered a remaining challenge: all generated images tend to resemble the model, irrespective of the prompt utilized. Although the model was trained using the prompt "ohwx" and the class "woman," this issue persists even when the prompt is omitted or when a different subject, such as "man," is specified.
Could you please offer some guidance on the potential cause of this occurrence? My objective is to generate images of other women/man in certain scenes who do not precisely replicate the trained model.
I have encountered a remaining challenge: all generated images tend to resemble the model, irrespective of the prompt utilized. Although the model was trained using the prompt "ohwx" and the class "woman," this issue persists even when the prompt is omitted or when a different subject, such as "man," is specified.
Could you please offer some guidance on the potential cause of this occurrence? My objective is to generate images of other women/man in certain scenes who do not precisely replicate the trained model.
Fsurely possible
Fyes it will make everyone same
Fthat is biggest issue of flux
Fit bleeds horribly
1Have you had a chance to try hidream yet? I’ve been testing loras with the same data set as flux and it’s so much better. Not sure how it will cope with multiple subjects concepts but for my purpose it’s miles ahead.
Fi didnt train yet
Fplanning to research it deeply
Fcurrently flux is just perfect except multi subject
