Lady Gaga LoRA trained on 40 images. This is 512x512 model and works on 1.5 models.
Works best on Protogen, revAnimated, Dreamshaper and Anything
CFG 5 - 9
Model strength between 0.6 and 0.8 seems to be best, tending towards 0.8
I will also work on a 768x768 model for 1.5
Description
FAQ
Comments (1)
I think the training parameters need to be adjusted or there just needs to be more training or something, this is only a vague resemblance at best, it seems.