说明：  WGAN是GAN的一种变种，通过用Wasserstein距离又叫Earth-Mover（EM）距离来衡量，生成数据与真实数据之间的分布举例，解决了GAN训练不稳定的问题
(Wgan is a variant of GaN. By measuring Wasserstein distance, also known as earth mover (EM) distance, wgan generates an example of distribution between data and real data to solve the problem of unstable training in GaN)