V2 updated:

V2 utilizes a wider variety of images with different styles as training data, resulting in more diverse and enriched outcomes when generating content.

V2采用更多不同风格的图片作为训练素材，使得模型抽卡结果更加丰富！