The beauty I know about from the movie Fallen Angels.
The network is trained on photos with a resolution of 768x768. Best result, network strength 0.8, resolution 768x1024. Negative promt is very important, there is a problem with clones.
Michelle_Reis.safetensors