V1.0

Beta, the first real portrait version in the series, does not require VEA and only requires a simple hint text to generate authentic Asian portraits. Note that since the refine model does not have a good refining effect on Asian images, it is not recommended to use a refiner for refining; you can do 1024x1024