What's your learning rate
What's your learning rate
EFlux. It's this config setup.
JAre you able to generate images while training?
ENo, but I have three computers.
JI think your learning rate is extremely low
Jnow I havent trained Flux, but 5e-05 is pretty low
Jmeaning 0.00005
Jtry learning rate 2e-4
EYes, 5e-5 looks low compared to typical SDXL LoRAs. But for Flux1 it’s the standard recommendation (e.g. Furkan’s config). It trains slower, but keeps things stable since Flux is more sensitive than SDXL/1.5. But maybe I'm wrong.
JYeah I am coming mostly from SD 1.5 , SDXL and Wan
Ji still would try going for 2e-4 with flux and see if you get decent results around 1500-3000 steps
JWorth a shot if you are not happy with your results
Jtry to find out if you can generate images while it's training (that means , the training algorithm stops for a moment to generate an image based on a prompt set)
Jso every 100 steps generate two images, one with prompt identical to a training image, and another one with something you may want to prompt later
Jthe later is more important
JYes , samples 
EThank you very much!
PGuys, can you help? I don't understand why the image is not properly generated... What am i missing? I don't understand how wan works...
Aplay with the return_noise and add_noise
Atheres something wrong there
PThanks for the advice - i used the same setting from base workflow text to video - but the result didn't change much
Atry 4+4 steps
Jtry this workflow
JJonXL Multi Wan 2.1 + 2.2 Advanced V5 AIO Workflow - T2I T2V I2V FLF2V (Trim, Extend, Upscale, Interpolate) * V5 released! Now supports Wan 2.1 wit...
Jits better if you use the 4 steps Lora
JDownload links are in the workflows' notes
PThanks, i definitely will! Wan is a new territory for me 
JSure 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
