Quoted By:
To make XL loras do you need more images of the character?
At first I thought the problem was on the model, but after doing some loras and some working and some being ignored completely despite being trained on the right model my only conclussion is that XL loras require quite more images than normal loras?
Is the only reason I can come with since Im always using the same setting but sometimes works and sometimes not, but the ones that not work are the ones done with under 20 images while the ones that work has 30+
But I ask just in case number of images is irrelevant and I should look for the problem somewhere else