A large portion of the training images has been replaced, making the generated 3D-style images much closer to the character’s appearance in the film.
By rewriting the prompt structure, the LoRA now has greater versatility.
大量更换了用于训练的图片,现在生成3D风格的图片会和电影里面的形象更加接近。
通过重新写提示词,lora的泛用性更强。
我很高兴能创建这个LORA,这是我第一次制作并发布LORA,我真的很喜欢Pawbert!