Using revanimated model, (the cartoon faces work better for low vram) Lora:multipleinsertions:0.7 ( for nude […]
