Hi~ I'm Juaner.

This is a model that reduces realism while enhancing the aesthetic appeal of images. The higher the weight, the stronger the face retouching effect will be, making faces more symmetrical and poses more like those of fashion models.

However, due to issues with the dataset, at a weight of 1, the face sharpening effect is very strong, and the facial features become more Western-looking, which I don't recommend.

Recommended prompts:
Natural language

For example:
A photo of a young Chinese woman with sleek, long black hair, Laugh Out Loud, dressed in a classic denim-on-denim outfit with a twist. She is posed in a studio with vibrant lighting and a retro boombox as a prop. The camera captures her in a candid laugh, bringing a lively energy.

Recommended base model:
flux dev

Recommended weight: 0.3-0.6 (I personally really like the quality at 0.3, retaining realism while being a bit more aesthetically pleasing than the original flux images)

Recommended size: 1080*1984

Please adjust flexibly based on the results.

Have fun, and welcome everyone to share their images.

If you have any commercial licensing or model customization needs, you can contact me at [email protected].

ღ( ´･ᴗ･` )

hi~我是卷儿。

这是一个会减少真实感提升图片美观度的模型，越高的权重就会越加强脸部ps的感觉，人脸会变得更加对称，拍站姿势会更加偏向于平面模特风格。

但是由于数据集的问题，权重为1的时候人脸锐化感会很强而且人物脸部会更加西方，我个人不推荐这个参数。

推荐prompt：

自然语言

比如：
A photo of a young Chinese woman with sleek, long black hair, Laugh Out Loud,dressed in a classic denim-on-denim outfit with a twist. She is posed in a studio with vibrant lighting and a retro boombox as a prop. The camera captures her in a candid laugh, bringing a lively energy

推荐底模：

flux dev

推荐权重：0.3-0.6（个人非常喜欢0.3的质感，保留真实又会比flux的原图美观一点）

推荐尺寸：1080*1984

请根据结果灵活调整

have fun, 欢迎大家返图。

如果有商业授权或者模型定制需求可以联系： [email protected]

ღ( ´･ᴗ･` )比心