My second Lora
Description
FAQ
Comments (7)
put "different-looking-pupils" in negative. reason should be clear.
but all in all this model is very well trained. will have my fun with it in the future :D
It is giving me this error only with this lora, I don't know if it is normal :NansException: A tensor with all NaNs was produced in Unet. Use --disable-nan-check commandline argument to disable this check.
That happens if you try to produce an image larger than what the training data had in it. I have a lora where i can't go over 550 in image size because I didn't train with enough options available, and that'll do it.
@xSatouMatsuzakax What resolution do you use? because I use 512x512 and 768x768 and it still gives an error
Sorry, the Lora didn't work i deleted the version and im trying to fix it
@spam905 a pity, I'm looking forward to try the V2 version :3
Could you please tell me the model file used in 2.0 and 1.0? (I would also like a download link if possible)