Also if you feel like constant learning rate is cooking your image, switch to cosine
Also if you feel like constant learning rate is cooking your image, switch to cosine
YI'm trying to see if its possible to just pass in the link of a huggingface dataset instead of making an image folder and linking it with captions
Yespecially on the gui side
LYou know what, I’ll bet that’s the biggest issue on time. Been using 1 on Alpha
FIt shouldnt be hmmm, are you usingMemory efficient attention >
LI don’t think so
Fare you training locally or runpod
Yon google colab
Fhow about gaming? noticing any issues with the 4090?
Fahh never used google collab, you should try kaggle or runpod
LNot a bit, crushes Baldur’s 3 on max
Ythey should work the same I think
Ydid work with kaggle before and it's pretty similar
Fyou should be good then, if training is still slow lemme know
Falso do you have harware-acclerated gpu scheduling turned on?
Fdamn might have to wait till somene gets back to you 
idk google collab
idk google collabLDo you change the LR scheduler?
LAnd B16 or F16?
Yah okay, honestly I'm only looking at this
FYep metioned it above ^ Also if you feel like constant learning rate is cooking your image, switch to cosine
Fbf16 since you're on a 4090
Ftry putting it like this /stabilityai/stable-diffusion-xl-base-1.0
Yah, that's chill, I think it wants me to download it and give it the full path. I can do that
Ythe issue is honestly with the dataset, if I just give it this Yorth/dalleTestDataFiltered2, it will make life much easier than downloading the dataset, separating the image folder, finding a way to link them with the captions without messing things up
Y:sigh:
LI’m playing around and the numbers are dropping. For reference are your settings based on 512 by 512? I’m using 768 by 1344
FMine is 1024x1024, and on 1536x1536 it takes around 2 and half hours
Fcan you share me your cmd, like a secreenshot of it
Fyou can dm if you want
LWhat’s a cmd?
Fthe terminal where the kohay gui is
Fwhere it tells you the progress of training
LSure
Falso your kohya could be bad, so maybe reinstall it as well, and reconfigure accelerate
L
LIt’s much better with your suggestions
Fwhats your vram usage?
LHow do I check that?
Fsearch for task manager
Fthen go to performance and then to gpu
LIt’s tinkering at 100 percent is that normal
Fokay cancel the training, and enable gradient checkpointing
Falso 2080 steps is a little low, so maybe up your repeats by 10 or epochs to 14
LThis Konya version doesn’t have all the older one has, is there a hidden spot for gradient checkpoints
Fwhich version are you on?
L22.1.0
Fthe min snr gamma is optional sometimes it gives better results
Fand then i have noise ofset at 0.0357 for sdxl
Fupgrade to the latest version when you get a chance
LFound it, do I not have the latest?
F22.3.0 is the latest