Sure thing, once you go WAN you never WAN'na go back
Sure thing, once you go WAN you never WAN'na go back
Pi hope so! 
XDoes anyone know if flux full fine tuning / dreambooth trained models allow nsfw generations?
Fi think 350 good
Fi trained with 256 images
Thave you tried creating character checkpoints from krea? I'm at 10000 steps and the model is clearly ruined by now: all stylized outputs became blurry mangled and photorealistic
Tno comparable character similiarity was reached like with flux-dev
TI'll try with different datasets but I don't think this process works with krea like with flux dev.
Fi think you dont need detailed captions
Fwith flux detailed captions makes sense only at general fine tuning with multiple concepts
Fand it is not doable with flux dev
Fbut with qwen i am hopeful
A200epoch and 4000 and you should be good timson
T4000 what?
TI mean it's kinda working but the result is much much worse than flux dev for me in both character similarity and model creativity
Asteps
Aidk guys im getting good results
Sme too actually, krea is my new go-to for finetunings
CI did so far just Lora. Not trained yet full model dreambooth
�ok i'll do krea next for comparison. doing my first fine tune locally, done loads on massedcompute. got 5090 and 128gb ram. went with the 24gb config. 25 pics, 200 epochs looks like its going to be around 6h.
Shey @Dr. Furkan Gözükara, could you give me some advice i have a big checkpoint to train (220 epochs with 300+ images) and it would take 88hours with a L40S. Is there a gpu I can rent what would significantly speed up the training (without going multi gpu, i read here that it impacts the quality of the training) ? Thanks
Fwell 220 epoch is too big for 300 images
Fit is a single concept right?
Fthere isnt much you can do to speed up even batch size wont help much
Funless multiple gpu
S@Dr. Furkan Gözükara It's for an editorial style model so the dataset isn't very homogenous... I went with 220 because i'm training for flux krea
Swould multiple gpu hurt the quality of the training ?
Fyou can do this
F4 gpus
Fbatch size 1
Fit will minimally impact
Fso your total batch size will be 4
Fmake learning rate = learning rate * 2
SOk @Dr. Furkan Gözükara thanks a lot. Is there a patreon post where you explain how to set up multiple gpu for runpod ?
FYouTubeSECourses
Unlock the power of FLUX LoRA training, even if you're short on GPUs or looking to boost speed and scale! This comprehensive guide takes you from novice to expert, showing you how to use Kohya GUI for creating top-notch FLUX LoRAs in the cloud. We'll cover everything: maximizing quality, optimizing speed, and finding the best deals. With our exc...
Syou're the best
SJust to make sure, I was using 4e-06, what would that make x2 ?
F4e-06 is 2x of 2e-06
STHanks 
S@Dr. Furkan Gözükara there are no 4xGPU config file in the latest kohya zip file
Fnot for dreambooth
Fit needs 80gb+
Flora has it
F
Fbut you can follow logic of lora and test
Sok will do thanks
�my first image was ok, the 2nd one looked like this, why?
�hmm, maybe 16bit enabled with the t5xxx_enconly caused it?
�or this happens, normal image fine, yolo face segment gets pixelated. whats wrong
