1000 DOWNLOAD AND TRAINING IMAGES SHARED!
v2:
Trained on 4092 image for 30 step / image, in total for 122760 step.
Recommended to use both v1 and v2 combined!
Update:
-Can adapt to hair color
-More detailed
v1:
Trained on 400 image for 200 step / image, in total for 40000 step.
All the images were created with ADetailer.
THERE IS A BETTER VERSION THAT I WILL RELEASE IF THIS GOT ENOUGH ATTENTION!
Please share, download, and upload images using this LoRA!
Pro:
This LoRA can generate cat ears with checkpoints that are otherwise are unable.
Con:
This LoRA was trained only on images of females.