Lora trained on generated content. She is not a real person.
First attempt to train ZIT lora.
Use at 0.7-1.5 depending on the composition.
Use at low weight (0.2-0.5) for realistic, high weight for painted style.
Milena_20260216180133.safetensors