Works best at 1.1 to 1.2 weight.
Trained on 31 Photos, mostly headshots, 2400 steps.
Willa_w1ll4-flux.safetensors